{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.20009018487900196,
  "eval_steps": 500,
  "global_step": 3328,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "auxiliary_loss_clip": 0.05028445,
      "auxiliary_loss_mlp": 0.02215396,
      "balance_loss_clip": 2.43573999,
      "balance_loss_mlp": 1.76983953,
      "epoch": 6.012325266796934e-05,
      "flos": 24456507091200.0,
      "grad_norm": 55.00561300220404,
      "language_loss": 2.85272503,
      "learning_rate": 0.0,
      "loss": 1.94613922,
      "num_input_tokens_seen": 19155,
      "step": 1,
      "time_per_iteration": 18.059409618377686
    },
    {
      "auxiliary_loss_clip": 0.03380539,
      "auxiliary_loss_mlp": 0.01459449,
      "balance_loss_clip": 1.62786555,
      "balance_loss_mlp": 1.18936849,
      "epoch": 0.00012024650533593868,
      "flos": 20225931246720.0,
      "grad_norm": 34.93149751452764,
      "language_loss": 1.82606053,
      "learning_rate": 4.4628432569317594e-07,
      "loss": 1.87446034,
      "num_input_tokens_seen": 36175,
      "step": 2,
      "time_per_iteration": 2.6318798065185547
    },
    {
      "auxiliary_loss_clip": 0.03320229,
      "auxiliary_loss_mlp": 0.01440978,
      "balance_loss_clip": 1.62577581,
      "balance_loss_mlp": 1.18882656,
      "epoch": 0.000180369758003908,
      "flos": 22309935454080.0,
      "grad_norm": 32.71870482280511,
      "language_loss": 1.57573509,
      "learning_rate": 7.073439208833112e-07,
      "loss": 1.62334716,
      "num_input_tokens_seen": 54870,
      "step": 3,
      "time_per_iteration": 2.6362481117248535
    },
    {
      "auxiliary_loss_clip": 0.03361497,
      "auxiliary_loss_mlp": 0.01451404,
      "balance_loss_clip": 1.62418985,
      "balance_loss_mlp": 1.15500188,
      "epoch": 0.00024049301067187735,
      "flos": 22414650577920.0,
      "grad_norm": 51.2387172839747,
      "language_loss": 1.67362881,
      "learning_rate": 8.925686513863519e-07,
      "loss": 1.72175777,
      "num_input_tokens_seen": 74575,
      "step": 4,
      "time_per_iteration": 2.7070822715759277
    },
    {
      "auxiliary_loss_clip": 0.03402497,
      "auxiliary_loss_mlp": 0.01505358,
      "balance_loss_clip": 1.62493396,
      "balance_loss_mlp": 1.21715808,
      "epoch": 0.0003006162633398467,
      "flos": 21396978449280.0,
      "grad_norm": 56.088721215944275,
      "language_loss": 1.91627169,
      "learning_rate": 1.0362401141348472e-06,
      "loss": 1.96535027,
      "num_input_tokens_seen": 92580,
      "step": 5,
      "time_per_iteration": 2.91436767578125
    },
    {
      "auxiliary_loss_clip": 0.03370454,
      "auxiliary_loss_mlp": 0.01515599,
      "balance_loss_clip": 1.61556244,
      "balance_loss_mlp": 1.22110426,
      "epoch": 0.000360739516007816,
      "flos": 21652375127040.0,
      "grad_norm": 33.397169652317885,
      "language_loss": 1.60591149,
      "learning_rate": 1.153628246576487e-06,
      "loss": 1.65477204,
      "num_input_tokens_seen": 109705,
      "step": 6,
      "time_per_iteration": 2.994969367980957
    },
    {
      "auxiliary_loss_clip": 0.03354239,
      "auxiliary_loss_mlp": 0.01486417,
      "balance_loss_clip": 1.61577415,
      "balance_loss_mlp": 1.20336628,
      "epoch": 0.0004208627686757854,
      "flos": 27159742897920.0,
      "grad_norm": 24.6270766983672,
      "language_loss": 1.53276002,
      "learning_rate": 1.2528784983718962e-06,
      "loss": 1.58116663,
      "num_input_tokens_seen": 129425,
      "step": 7,
      "time_per_iteration": 3.0675876140594482
    },
    {
      "auxiliary_loss_clip": 0.03321216,
      "auxiliary_loss_mlp": 0.0144328,
      "balance_loss_clip": 1.61205018,
      "balance_loss_mlp": 1.16499734,
      "epoch": 0.0004809860213437547,
      "flos": 31319096135040.0,
      "grad_norm": 31.71613063643349,
      "language_loss": 1.43881059,
      "learning_rate": 1.338852977079528e-06,
      "loss": 1.48645568,
      "num_input_tokens_seen": 149210,
      "step": 8,
      "time_per_iteration": 3.172358751296997
    },
    {
      "auxiliary_loss_clip": 0.03368839,
      "auxiliary_loss_mlp": 0.01496105,
      "balance_loss_clip": 1.6120348,
      "balance_loss_mlp": 1.21229148,
      "epoch": 0.000541109274011724,
      "flos": 32160411463680.0,
      "grad_norm": 28.204490849684397,
      "language_loss": 1.4969244,
      "learning_rate": 1.4146878417666224e-06,
      "loss": 1.54557395,
      "num_input_tokens_seen": 169055,
      "step": 9,
      "time_per_iteration": 3.112215280532837
    },
    {
      "auxiliary_loss_clip": 0.03308365,
      "auxiliary_loss_mlp": 0.01475035,
      "balance_loss_clip": 1.61541438,
      "balance_loss_mlp": 1.20647991,
      "epoch": 0.0006012325266796934,
      "flos": 18916808163840.0,
      "grad_norm": 23.420774723604698,
      "language_loss": 1.44714785,
      "learning_rate": 1.4825244398280232e-06,
      "loss": 1.49498188,
      "num_input_tokens_seen": 188045,
      "step": 10,
      "time_per_iteration": 2.9495606422424316
    },
    {
      "auxiliary_loss_clip": 0.03364194,
      "auxiliary_loss_mlp": 0.01494262,
      "balance_loss_clip": 1.62042511,
      "balance_loss_mlp": 1.22036684,
      "epoch": 0.0006613557793476627,
      "flos": 20774861867520.0,
      "grad_norm": 18.353281468858004,
      "language_loss": 1.4520936,
      "learning_rate": 1.5438901072051983e-06,
      "loss": 1.50067806,
      "num_input_tokens_seen": 207035,
      "step": 11,
      "time_per_iteration": 3.0797431468963623
    },
    {
      "auxiliary_loss_clip": 0.03292683,
      "auxiliary_loss_mlp": 0.0145154,
      "balance_loss_clip": 1.60771322,
      "balance_loss_mlp": 1.17554641,
      "epoch": 0.000721479032015632,
      "flos": 16581680997120.0,
      "grad_norm": 16.61869254675767,
      "language_loss": 1.45121813,
      "learning_rate": 1.5999125722696629e-06,
      "loss": 1.49866033,
      "num_input_tokens_seen": 223225,
      "step": 12,
      "time_per_iteration": 2.9887659549713135
    },
    {
      "auxiliary_loss_clip": 0.03321669,
      "auxiliary_loss_mlp": 0.01405912,
      "balance_loss_clip": 1.61740756,
      "balance_loss_mlp": 1.14765704,
      "epoch": 0.0007816022846836014,
      "flos": 23805471144960.0,
      "grad_norm": 14.02187318243825,
      "language_loss": 1.23759985,
      "learning_rate": 1.6514482443788434e-06,
      "loss": 1.28487587,
      "num_input_tokens_seen": 242570,
      "step": 13,
      "time_per_iteration": 3.032742977142334
    },
    {
      "auxiliary_loss_clip": 0.03287474,
      "auxiliary_loss_mlp": 0.01470749,
      "balance_loss_clip": 1.61299658,
      "balance_loss_mlp": 1.20257616,
      "epoch": 0.0008417255373515708,
      "flos": 19172204841600.0,
      "grad_norm": 5.790568956401358,
      "language_loss": 1.20684385,
      "learning_rate": 1.6991628240650723e-06,
      "loss": 1.254426,
      "num_input_tokens_seen": 261215,
      "step": 14,
      "time_per_iteration": 3.002887487411499
    },
    {
      "auxiliary_loss_clip": 0.03272826,
      "auxiliary_loss_mlp": 0.01431255,
      "balance_loss_clip": 1.6181426,
      "balance_loss_mlp": 1.16804111,
      "epoch": 0.00090184879001954,
      "flos": 26395564026240.0,
      "grad_norm": 6.353887091300461,
      "language_loss": 1.12925518,
      "learning_rate": 1.7435840350181584e-06,
      "loss": 1.176296,
      "num_input_tokens_seen": 280035,
      "step": 15,
      "time_per_iteration": 3.0238780975341797
    },
    {
      "auxiliary_loss_clip": 0.03238489,
      "auxiliary_loss_mlp": 0.01411651,
      "balance_loss_clip": 1.60288334,
      "balance_loss_mlp": 1.16197944,
      "epoch": 0.0009619720426875094,
      "flos": 24679500785280.0,
      "grad_norm": 4.670310144758637,
      "language_loss": 1.11125767,
      "learning_rate": 1.7851373027727038e-06,
      "loss": 1.15775907,
      "num_input_tokens_seen": 300265,
      "step": 16,
      "time_per_iteration": 4.605847120285034
    },
    {
      "auxiliary_loss_clip": 0.03223993,
      "auxiliary_loss_mlp": 0.01417304,
      "balance_loss_clip": 1.60910368,
      "balance_loss_mlp": 1.17774093,
      "epoch": 0.0010220952953554788,
      "flos": 18624531196800.0,
      "grad_norm": 8.838429022323517,
      "language_loss": 1.12645221,
      "learning_rate": 1.8241705979033208e-06,
      "loss": 1.17286515,
      "num_input_tokens_seen": 317375,
      "step": 17,
      "time_per_iteration": 4.579033851623535
    },
    {
      "auxiliary_loss_clip": 0.03161492,
      "auxiliary_loss_mlp": 0.01379312,
      "balance_loss_clip": 1.60685277,
      "balance_loss_mlp": 1.1475693,
      "epoch": 0.001082218548023448,
      "flos": 26142537646080.0,
      "grad_norm": 3.823557061532633,
      "language_loss": 1.08069181,
      "learning_rate": 1.860972167459798e-06,
      "loss": 1.12609982,
      "num_input_tokens_seen": 337975,
      "step": 18,
      "time_per_iteration": 3.0132579803466797
    },
    {
      "auxiliary_loss_clip": 0.0318761,
      "auxiliary_loss_mlp": 0.01403306,
      "balance_loss_clip": 1.60585093,
      "balance_loss_mlp": 1.13799417,
      "epoch": 0.0011423418006914173,
      "flos": 19609776322560.0,
      "grad_norm": 4.403621106373983,
      "language_loss": 1.02445412,
      "learning_rate": 1.89578346593066e-06,
      "loss": 1.07036328,
      "num_input_tokens_seen": 356635,
      "step": 19,
      "time_per_iteration": 3.016176462173462
    },
    {
      "auxiliary_loss_clip": 0.0313029,
      "auxiliary_loss_mlp": 0.01342049,
      "balance_loss_clip": 1.60759044,
      "balance_loss_mlp": 1.12155962,
      "epoch": 0.0012024650533593868,
      "flos": 17895365107200.0,
      "grad_norm": 3.958333686933058,
      "language_loss": 1.16706228,
      "learning_rate": 1.928808765521199e-06,
      "loss": 1.21178555,
      "num_input_tokens_seen": 375625,
      "step": 20,
      "time_per_iteration": 3.0274486541748047
    },
    {
      "auxiliary_loss_clip": 0.03118109,
      "auxiliary_loss_mlp": 0.01378536,
      "balance_loss_clip": 1.58886433,
      "balance_loss_mlp": 1.1298182,
      "epoch": 0.001262588306027356,
      "flos": 21252043071360.0,
      "grad_norm": 4.333519066420982,
      "language_loss": 1.06129968,
      "learning_rate": 1.9602224192552076e-06,
      "loss": 1.10626626,
      "num_input_tokens_seen": 394350,
      "step": 21,
      "time_per_iteration": 2.9418578147888184
    },
    {
      "auxiliary_loss_clip": 0.03013912,
      "auxiliary_loss_mlp": 0.0137937,
      "balance_loss_clip": 1.57028937,
      "balance_loss_mlp": 1.14552903,
      "epoch": 0.0013227115586953253,
      "flos": 26104077158400.0,
      "grad_norm": 3.63841390311849,
      "language_loss": 1.05861485,
      "learning_rate": 1.9901744328983746e-06,
      "loss": 1.10254765,
      "num_input_tokens_seen": 413255,
      "step": 22,
      "time_per_iteration": 2.9651288986206055
    },
    {
      "auxiliary_loss_clip": 0.02966296,
      "auxiliary_loss_mlp": 0.01334065,
      "balance_loss_clip": 1.57175612,
      "balance_loss_mlp": 1.12377954,
      "epoch": 0.0013828348113632948,
      "flos": 23951376190080.0,
      "grad_norm": 2.8746130742538347,
      "language_loss": 0.9177655,
      "learning_rate": 2.018794797290208e-06,
      "loss": 0.96076906,
      "num_input_tokens_seen": 433065,
      "step": 23,
      "time_per_iteration": 3.049853563308716
    },
    {
      "auxiliary_loss_clip": 0.02932793,
      "auxiliary_loss_mlp": 0.01362183,
      "balance_loss_clip": 1.56404662,
      "balance_loss_mlp": 1.14236116,
      "epoch": 0.001442958064031264,
      "flos": 15959851724160.0,
      "grad_norm": 3.0897201135857735,
      "language_loss": 1.08192635,
      "learning_rate": 2.046196897962839e-06,
      "loss": 1.12487614,
      "num_input_tokens_seen": 451175,
      "step": 24,
      "time_per_iteration": 3.0543172359466553
    },
    {
      "auxiliary_loss_clip": 0.02823838,
      "auxiliary_loss_mlp": 0.01329007,
      "balance_loss_clip": 1.55692792,
      "balance_loss_mlp": 1.11853111,
      "epoch": 0.0015030813166992333,
      "flos": 18108350801280.0,
      "grad_norm": 4.111246686692462,
      "language_loss": 1.01367807,
      "learning_rate": 2.0724802282696944e-06,
      "loss": 1.05520654,
      "num_input_tokens_seen": 468775,
      "step": 25,
      "time_per_iteration": 3.0059614181518555
    },
    {
      "auxiliary_loss_clip": 0.02818207,
      "auxiliary_loss_mlp": 0.01309454,
      "balance_loss_clip": 1.55974329,
      "balance_loss_mlp": 1.10012197,
      "epoch": 0.0015632045693672028,
      "flos": 22234558763520.0,
      "grad_norm": 2.7163042439620018,
      "language_loss": 1.0669204,
      "learning_rate": 2.0977325700720194e-06,
      "loss": 1.10819697,
      "num_input_tokens_seen": 488530,
      "step": 26,
      "time_per_iteration": 3.1159534454345703
    },
    {
      "auxiliary_loss_clip": 0.0276047,
      "auxiliary_loss_mlp": 0.01325034,
      "balance_loss_clip": 1.54973662,
      "balance_loss_mlp": 1.12533486,
      "epoch": 0.001623327822035172,
      "flos": 23991955580160.0,
      "grad_norm": 2.562596284241794,
      "language_loss": 0.95537072,
      "learning_rate": 2.122031762649933e-06,
      "loss": 0.99622583,
      "num_input_tokens_seen": 510495,
      "step": 27,
      "time_per_iteration": 3.018643617630005
    },
    {
      "auxiliary_loss_clip": 0.02736222,
      "auxiliary_loss_mlp": 0.01311707,
      "balance_loss_clip": 1.55399776,
      "balance_loss_mlp": 1.13089037,
      "epoch": 0.0016834510747031415,
      "flos": 19677647070720.0,
      "grad_norm": 2.42975125432869,
      "language_loss": 1.06393945,
      "learning_rate": 2.1454471497582483e-06,
      "loss": 1.10441875,
      "num_input_tokens_seen": 528605,
      "step": 28,
      "time_per_iteration": 2.9263083934783936
    },
    {
      "auxiliary_loss_clip": 0.0270011,
      "auxiliary_loss_mlp": 0.0131913,
      "balance_loss_clip": 1.53841436,
      "balance_loss_mlp": 1.13297284,
      "epoch": 0.0017435743273711108,
      "flos": 20923819568640.0,
      "grad_norm": 4.42090805909513,
      "language_loss": 1.02493238,
      "learning_rate": 2.1680407726407727e-06,
      "loss": 1.06512475,
      "num_input_tokens_seen": 548515,
      "step": 29,
      "time_per_iteration": 3.0062997341156006
    },
    {
      "auxiliary_loss_clip": 0.0269246,
      "auxiliary_loss_mlp": 0.01312758,
      "balance_loss_clip": 1.53459728,
      "balance_loss_mlp": 1.12631428,
      "epoch": 0.00180369758003908,
      "flos": 19528976678400.0,
      "grad_norm": 3.1534114534186446,
      "language_loss": 1.19265521,
      "learning_rate": 2.189868360711334e-06,
      "loss": 1.23270726,
      "num_input_tokens_seen": 564025,
      "step": 30,
      "time_per_iteration": 2.931145429611206
    },
    {
      "auxiliary_loss_clip": 0.02610377,
      "auxiliary_loss_mlp": 0.01337183,
      "balance_loss_clip": 1.52116311,
      "balance_loss_mlp": 1.15665221,
      "epoch": 0.0018638208327070496,
      "flos": 27453169100160.0,
      "grad_norm": 2.735994596991484,
      "language_loss": 1.02616811,
      "learning_rate": 2.2109801597326265e-06,
      "loss": 1.06564379,
      "num_input_tokens_seen": 583345,
      "step": 31,
      "time_per_iteration": 2.993251085281372
    },
    {
      "auxiliary_loss_clip": 0.02582044,
      "auxiliary_loss_mlp": 0.01331305,
      "balance_loss_clip": 1.522609,
      "balance_loss_mlp": 1.15163302,
      "epoch": 0.0019239440853750188,
      "flos": 13589460380160.0,
      "grad_norm": 3.9056907796043654,
      "language_loss": 0.95266509,
      "learning_rate": 2.2314216284658796e-06,
      "loss": 0.99179864,
      "num_input_tokens_seen": 600010,
      "step": 32,
      "time_per_iteration": 2.9459571838378906
    },
    {
      "auxiliary_loss_clip": 0.02564836,
      "auxiliary_loss_mlp": 0.01302659,
      "balance_loss_clip": 1.51811624,
      "balance_loss_mlp": 1.13586164,
      "epoch": 0.001984067338042988,
      "flos": 11253866336640.0,
      "grad_norm": 3.226486022987097,
      "language_loss": 0.95143497,
      "learning_rate": 2.2512340280885094e-06,
      "loss": 0.99010992,
      "num_input_tokens_seen": 616295,
      "step": 33,
      "time_per_iteration": 2.9855570793151855
    },
    {
      "auxiliary_loss_clip": 0.02421202,
      "auxiliary_loss_mlp": 0.01304214,
      "balance_loss_clip": 1.48474145,
      "balance_loss_mlp": 1.14676213,
      "epoch": 0.0020441905907109576,
      "flos": 22386245898240.0,
      "grad_norm": 2.1714659525821247,
      "language_loss": 0.91547924,
      "learning_rate": 2.270454923596497e-06,
      "loss": 0.9527334,
      "num_input_tokens_seen": 637640,
      "step": 34,
      "time_per_iteration": 2.981541872024536
    },
    {
      "auxiliary_loss_clip": 0.02375249,
      "auxiliary_loss_mlp": 0.01271963,
      "balance_loss_clip": 1.45095515,
      "balance_loss_mlp": 1.11689591,
      "epoch": 0.0021043138433789266,
      "flos": 49778580337920.0,
      "grad_norm": 2.2635429103650386,
      "language_loss": 0.76603377,
      "learning_rate": 2.2891186125067434e-06,
      "loss": 0.80250585,
      "num_input_tokens_seen": 659710,
      "step": 35,
      "time_per_iteration": 3.2267208099365234
    },
    {
      "auxiliary_loss_clip": 0.02347187,
      "auxiliary_loss_mlp": 0.01276388,
      "balance_loss_clip": 1.46356034,
      "balance_loss_mlp": 1.13238275,
      "epoch": 0.002164437096046896,
      "flos": 20557961591040.0,
      "grad_norm": 2.3605884715298506,
      "language_loss": 0.88713098,
      "learning_rate": 2.307256493152974e-06,
      "loss": 0.92336679,
      "num_input_tokens_seen": 679670,
      "step": 36,
      "time_per_iteration": 2.948162078857422
    },
    {
      "auxiliary_loss_clip": 0.02289192,
      "auxiliary_loss_mlp": 0.01338204,
      "balance_loss_clip": 1.45043015,
      "balance_loss_mlp": 1.19105196,
      "epoch": 0.0022245603487148656,
      "flos": 26542295084160.0,
      "grad_norm": 2.4929063351918166,
      "language_loss": 0.93038809,
      "learning_rate": 2.3248973825097614e-06,
      "loss": 0.96666199,
      "num_input_tokens_seen": 700170,
      "step": 37,
      "time_per_iteration": 2.9556422233581543
    },
    {
      "auxiliary_loss_clip": 0.02249098,
      "auxiliary_loss_mlp": 0.01276785,
      "balance_loss_clip": 1.44485605,
      "balance_loss_mlp": 1.15500069,
      "epoch": 0.0022846836013828346,
      "flos": 20338188226560.0,
      "grad_norm": 2.177909778954084,
      "language_loss": 1.03952074,
      "learning_rate": 2.3420677916238357e-06,
      "loss": 1.07477951,
      "num_input_tokens_seen": 718545,
      "step": 38,
      "time_per_iteration": 2.9959065914154053
    },
    {
      "auxiliary_loss_clip": 0.02216028,
      "auxiliary_loss_mlp": 0.01260768,
      "balance_loss_clip": 1.43807542,
      "balance_loss_mlp": 1.13726676,
      "epoch": 0.002344806854050804,
      "flos": 26247575992320.0,
      "grad_norm": 2.22652515093943,
      "language_loss": 0.85297108,
      "learning_rate": 2.358792165262154e-06,
      "loss": 0.887739,
      "num_input_tokens_seen": 739865,
      "step": 39,
      "time_per_iteration": 3.035399913787842
    },
    {
      "auxiliary_loss_clip": 0.02192275,
      "auxiliary_loss_mlp": 0.01250434,
      "balance_loss_clip": 1.4289664,
      "balance_loss_mlp": 1.12216496,
      "epoch": 0.0024049301067187736,
      "flos": 11801539981440.0,
      "grad_norm": 3.258228308703562,
      "language_loss": 0.90279335,
      "learning_rate": 2.3750930912143747e-06,
      "loss": 0.93722045,
      "num_input_tokens_seen": 755770,
      "step": 40,
      "time_per_iteration": 3.060368299484253
    },
    {
      "auxiliary_loss_clip": 0.02142113,
      "auxiliary_loss_mlp": 0.01273783,
      "balance_loss_clip": 1.41895449,
      "balance_loss_mlp": 1.16086745,
      "epoch": 0.0024650533593867426,
      "flos": 20631506688000.0,
      "grad_norm": 3.245861029799582,
      "language_loss": 0.93271625,
      "learning_rate": 2.3909914837471044e-06,
      "loss": 0.9668752,
      "num_input_tokens_seen": 773440,
      "step": 41,
      "time_per_iteration": 2.9518353939056396
    },
    {
      "auxiliary_loss_clip": 0.02105753,
      "auxiliary_loss_mlp": 0.01254821,
      "balance_loss_clip": 1.41097844,
      "balance_loss_mlp": 1.15168142,
      "epoch": 0.002525176612054712,
      "flos": 18406122549120.0,
      "grad_norm": 3.3039479788253536,
      "language_loss": 0.97533798,
      "learning_rate": 2.4065067449483835e-06,
      "loss": 1.0089438,
      "num_input_tokens_seen": 790455,
      "step": 42,
      "time_per_iteration": 2.933177947998047
    },
    {
      "auxiliary_loss_clip": 0.020675,
      "auxiliary_loss_mlp": 0.01298422,
      "balance_loss_clip": 1.41198874,
      "balance_loss_mlp": 1.19189644,
      "epoch": 0.0025852998647226816,
      "flos": 28184023128960.0,
      "grad_norm": 3.15071165872949,
      "language_loss": 0.97562659,
      "learning_rate": 2.4216569070848724e-06,
      "loss": 1.00928593,
      "num_input_tokens_seen": 810645,
      "step": 43,
      "time_per_iteration": 2.9760589599609375
    },
    {
      "auxiliary_loss_clip": 0.02086351,
      "auxiliary_loss_mlp": 0.01314601,
      "balance_loss_clip": 1.41042757,
      "balance_loss_mlp": 1.20283043,
      "epoch": 0.0026454231173906506,
      "flos": 14283110897280.0,
      "grad_norm": 2.3612650137146574,
      "language_loss": 0.93435001,
      "learning_rate": 2.4364587585915504e-06,
      "loss": 0.96835947,
      "num_input_tokens_seen": 827470,
      "step": 44,
      "time_per_iteration": 2.9239895343780518
    },
    {
      "auxiliary_loss_clip": 0.02043996,
      "auxiliary_loss_mlp": 0.01272131,
      "balance_loss_clip": 1.40557313,
      "balance_loss_mlp": 1.17399764,
      "epoch": 0.00270554637005862,
      "flos": 22419211605120.0,
      "grad_norm": 2.1476860292916644,
      "language_loss": 0.98677421,
      "learning_rate": 2.450927955901469e-06,
      "loss": 1.01993537,
      "num_input_tokens_seen": 847285,
      "step": 45,
      "time_per_iteration": 2.9626305103302
    },
    {
      "auxiliary_loss_clip": 0.02018804,
      "auxiliary_loss_mlp": 0.01228873,
      "balance_loss_clip": 1.39126372,
      "balance_loss_mlp": 1.14208817,
      "epoch": 0.0027656696227265896,
      "flos": 23985778440960.0,
      "grad_norm": 1.8862192248435494,
      "language_loss": 1.02800822,
      "learning_rate": 2.465079122983384e-06,
      "loss": 1.06048501,
      "num_input_tokens_seen": 867545,
      "step": 46,
      "time_per_iteration": 2.9913573265075684
    },
    {
      "auxiliary_loss_clip": 0.0198766,
      "auxiliary_loss_mlp": 0.01272862,
      "balance_loss_clip": 1.38388658,
      "balance_loss_mlp": 1.182549,
      "epoch": 0.0028257928753945586,
      "flos": 37669503087360.0,
      "grad_norm": 2.1076645953887696,
      "language_loss": 0.87839413,
      "learning_rate": 2.4789259401737868e-06,
      "loss": 0.9109993,
      "num_input_tokens_seen": 889915,
      "step": 47,
      "time_per_iteration": 3.0189881324768066
    },
    {
      "auxiliary_loss_clip": 0.01949271,
      "auxiliary_loss_mlp": 0.01255947,
      "balance_loss_clip": 1.37360096,
      "balance_loss_mlp": 1.16963911,
      "epoch": 0.002885916128062528,
      "flos": 22454547609600.0,
      "grad_norm": 4.4561049138068,
      "language_loss": 0.87809587,
      "learning_rate": 2.492481223656015e-06,
      "loss": 0.91014802,
      "num_input_tokens_seen": 908975,
      "step": 48,
      "time_per_iteration": 2.863565444946289
    },
    {
      "auxiliary_loss_clip": 0.01949016,
      "auxiliary_loss_mlp": 0.0124182,
      "balance_loss_clip": 1.36337733,
      "balance_loss_mlp": 1.15069616,
      "epoch": 0.0029460393807304976,
      "flos": 27012796358400.0,
      "grad_norm": 2.9451035624229855,
      "language_loss": 0.89691317,
      "learning_rate": 2.5057569967437924e-06,
      "loss": 0.9288215,
      "num_input_tokens_seen": 929810,
      "step": 49,
      "time_per_iteration": 2.9967453479766846
    },
    {
      "auxiliary_loss_clip": 0.0194038,
      "auxiliary_loss_mlp": 0.01234077,
      "balance_loss_clip": 1.35742152,
      "balance_loss_mlp": 1.14996314,
      "epoch": 0.0030061626333984666,
      "flos": 15851832549120.0,
      "grad_norm": 3.162716210197168,
      "language_loss": 0.90914285,
      "learning_rate": 2.51876455396287e-06,
      "loss": 0.94088745,
      "num_input_tokens_seen": 948650,
      "step": 50,
      "time_per_iteration": 2.8832523822784424
    },
    {
      "auxiliary_loss_clip": 0.01938537,
      "auxiliary_loss_mlp": 0.01199505,
      "balance_loss_clip": 1.36240602,
      "balance_loss_mlp": 1.11844242,
      "epoch": 0.003066285886066436,
      "flos": 31827052316160.0,
      "grad_norm": 6.098010360158733,
      "language_loss": 0.86977792,
      "learning_rate": 2.5315145187866316e-06,
      "loss": 0.90115827,
      "num_input_tokens_seen": 966455,
      "step": 51,
      "time_per_iteration": 2.9061717987060547
    },
    {
      "auxiliary_loss_clip": 0.01895637,
      "auxiliary_loss_mlp": 0.01206588,
      "balance_loss_clip": 1.35252357,
      "balance_loss_mlp": 1.12829173,
      "epoch": 0.0031264091387344056,
      "flos": 41427482774400.0,
      "grad_norm": 2.043292881862276,
      "language_loss": 0.95171362,
      "learning_rate": 2.5440168957651953e-06,
      "loss": 0.98273587,
      "num_input_tokens_seen": 988110,
      "step": 52,
      "time_per_iteration": 3.0266616344451904
    },
    {
      "auxiliary_loss_clip": 0.01893195,
      "auxiliary_loss_mlp": 0.01241159,
      "balance_loss_clip": 1.34894896,
      "balance_loss_mlp": 1.16162264,
      "epoch": 0.0031865323914023747,
      "flos": 23440941970560.0,
      "grad_norm": 4.2358840345824635,
      "language_loss": 0.92323011,
      "learning_rate": 2.5562811176888872e-06,
      "loss": 0.95457363,
      "num_input_tokens_seen": 1008550,
      "step": 53,
      "time_per_iteration": 2.8850226402282715
    },
    {
      "auxiliary_loss_clip": 0.01882736,
      "auxiliary_loss_mlp": 0.01197045,
      "balance_loss_clip": 1.35264134,
      "balance_loss_mlp": 1.11669779,
      "epoch": 0.003246655644070344,
      "flos": 14429195510400.0,
      "grad_norm": 2.290226623360683,
      "language_loss": 0.8260113,
      "learning_rate": 2.5683160883431093e-06,
      "loss": 0.85680908,
      "num_input_tokens_seen": 1026840,
      "step": 54,
      "time_per_iteration": 2.9433553218841553
    },
    {
      "auxiliary_loss_clip": 0.01880073,
      "auxiliary_loss_mlp": 0.01210775,
      "balance_loss_clip": 1.34162152,
      "balance_loss_mlp": 1.13233542,
      "epoch": 0.0033067788967383136,
      "flos": 35918247496320.0,
      "grad_norm": 2.911577423572303,
      "language_loss": 0.81303245,
      "learning_rate": 2.580130221340046e-06,
      "loss": 0.84394085,
      "num_input_tokens_seen": 1048875,
      "step": 55,
      "time_per_iteration": 3.0040643215179443
    },
    {
      "auxiliary_loss_clip": 0.01870075,
      "auxiliary_loss_mlp": 0.0120375,
      "balance_loss_clip": 1.33644819,
      "balance_loss_mlp": 1.12521541,
      "epoch": 0.003366902149406283,
      "flos": 22958732862720.0,
      "grad_norm": 2.639118679342801,
      "language_loss": 0.87089968,
      "learning_rate": 2.5917314754514246e-06,
      "loss": 0.90163803,
      "num_input_tokens_seen": 1066435,
      "step": 56,
      "time_per_iteration": 2.830453395843506
    },
    {
      "auxiliary_loss_clip": 0.01869912,
      "auxiliary_loss_mlp": 0.01161425,
      "balance_loss_clip": 1.32921791,
      "balance_loss_mlp": 1.08851671,
      "epoch": 0.003427025402074252,
      "flos": 26582838560640.0,
      "grad_norm": 2.101574700040827,
      "language_loss": 0.92890096,
      "learning_rate": 2.6031273868139713e-06,
      "loss": 0.95921433,
      "num_input_tokens_seen": 1090330,
      "step": 57,
      "time_per_iteration": 7.0071024894714355
    },
    {
      "auxiliary_loss_clip": 0.01833802,
      "auxiliary_loss_mlp": 0.0121675,
      "balance_loss_clip": 1.33333457,
      "balance_loss_mlp": 1.14493799,
      "epoch": 0.0034871486547422216,
      "flos": 23951196622080.0,
      "grad_norm": 14.610065921505914,
      "language_loss": 0.9972856,
      "learning_rate": 2.614325098333948e-06,
      "loss": 1.02779114,
      "num_input_tokens_seen": 1109840,
      "step": 58,
      "time_per_iteration": 2.830960273742676
    },
    {
      "auxiliary_loss_clip": 0.0181804,
      "auxiliary_loss_mlp": 0.01199311,
      "balance_loss_clip": 1.32073379,
      "balance_loss_mlp": 1.12835753,
      "epoch": 0.003547271907410191,
      "flos": 21214983214080.0,
      "grad_norm": 2.120622270947527,
      "language_loss": 0.88172519,
      "learning_rate": 2.625331386578098e-06,
      "loss": 0.91189873,
      "num_input_tokens_seen": 1128415,
      "step": 59,
      "time_per_iteration": 2.8507089614868164
    },
    {
      "auxiliary_loss_clip": 0.01839573,
      "auxiliary_loss_mlp": 0.01163328,
      "balance_loss_clip": 1.32924581,
      "balance_loss_mlp": 1.09075332,
      "epoch": 0.00360739516007816,
      "flos": 16504903676160.0,
      "grad_norm": 2.021991994360373,
      "language_loss": 0.93542433,
      "learning_rate": 2.63615268640451e-06,
      "loss": 0.96545339,
      "num_input_tokens_seen": 1146515,
      "step": 60,
      "time_per_iteration": 2.8517534732818604
    },
    {
      "auxiliary_loss_clip": 0.0181893,
      "auxiliary_loss_mlp": 0.01176948,
      "balance_loss_clip": 1.31414318,
      "balance_loss_mlp": 1.10923755,
      "epoch": 0.0036675184127461296,
      "flos": 19464805031040.0,
      "grad_norm": 2.908283338489548,
      "language_loss": 0.90021706,
      "learning_rate": 2.6467951135575943e-06,
      "loss": 0.9301759,
      "num_input_tokens_seen": 1166330,
      "step": 61,
      "time_per_iteration": 2.8853390216827393
    },
    {
      "auxiliary_loss_clip": 0.01803943,
      "auxiliary_loss_mlp": 0.01142904,
      "balance_loss_clip": 1.31131864,
      "balance_loss_mlp": 1.07581341,
      "epoch": 0.003727641665414099,
      "flos": 20957323979520.0,
      "grad_norm": 1.8428161811646855,
      "language_loss": 0.88479733,
      "learning_rate": 2.657264485425803e-06,
      "loss": 0.91426575,
      "num_input_tokens_seen": 1186010,
      "step": 62,
      "time_per_iteration": 2.8860812187194824
    },
    {
      "auxiliary_loss_clip": 0.01785338,
      "auxiliary_loss_mlp": 0.0116457,
      "balance_loss_clip": 1.30233741,
      "balance_loss_mlp": 1.09504724,
      "epoch": 0.003787764918082068,
      "flos": 18406050721920.0,
      "grad_norm": 2.4385306002926512,
      "language_loss": 0.96280968,
      "learning_rate": 2.6675663401385186e-06,
      "loss": 0.99230874,
      "num_input_tokens_seen": 1204985,
      "step": 63,
      "time_per_iteration": 2.9081404209136963
    },
    {
      "auxiliary_loss_clip": 0.01795068,
      "auxiliary_loss_mlp": 0.01171321,
      "balance_loss_clip": 1.31071985,
      "balance_loss_mlp": 1.10499322,
      "epoch": 0.0038478881707500376,
      "flos": 12459243962880.0,
      "grad_norm": 3.0781639748926697,
      "language_loss": 0.98840165,
      "learning_rate": 2.677705954159056e-06,
      "loss": 1.01806557,
      "num_input_tokens_seen": 1223545,
      "step": 64,
      "time_per_iteration": 2.893603801727295
    },
    {
      "auxiliary_loss_clip": 0.01801311,
      "auxiliary_loss_mlp": 0.01151112,
      "balance_loss_clip": 1.30960393,
      "balance_loss_mlp": 1.08368695,
      "epoch": 0.003908011423418007,
      "flos": 13553334276480.0,
      "grad_norm": 2.4813676281781554,
      "language_loss": 0.85397774,
      "learning_rate": 2.6876883585136904e-06,
      "loss": 0.88350195,
      "num_input_tokens_seen": 1241175,
      "step": 65,
      "time_per_iteration": 2.8768796920776367
    },
    {
      "auxiliary_loss_clip": 0.01777474,
      "auxiliary_loss_mlp": 0.01155217,
      "balance_loss_clip": 1.29563761,
      "balance_loss_mlp": 1.087888,
      "epoch": 0.003968134676085976,
      "flos": 18333475292160.0,
      "grad_norm": 1.8550079005121831,
      "language_loss": 0.85281348,
      "learning_rate": 2.697518353781685e-06,
      "loss": 0.88214046,
      "num_input_tokens_seen": 1259315,
      "step": 66,
      "time_per_iteration": 2.769274950027466
    },
    {
      "auxiliary_loss_clip": 0.01779987,
      "auxiliary_loss_mlp": 0.01151372,
      "balance_loss_clip": 1.29312515,
      "balance_loss_mlp": 1.07650828,
      "epoch": 0.004028257928753946,
      "flos": 20485242506880.0,
      "grad_norm": 2.74895944689593,
      "language_loss": 0.96567476,
      "learning_rate": 2.7072005239581103e-06,
      "loss": 0.99498826,
      "num_input_tokens_seen": 1277055,
      "step": 67,
      "time_per_iteration": 2.889369249343872
    },
    {
      "auxiliary_loss_clip": 0.01752442,
      "auxiliary_loss_mlp": 0.01152779,
      "balance_loss_clip": 1.28765118,
      "balance_loss_mlp": 1.08120584,
      "epoch": 0.004088381181421915,
      "flos": 18843837684480.0,
      "grad_norm": 2.109359538419204,
      "language_loss": 0.94516367,
      "learning_rate": 2.7167392492896727e-06,
      "loss": 0.97421581,
      "num_input_tokens_seen": 1294355,
      "step": 68,
      "time_per_iteration": 2.8107409477233887
    },
    {
      "auxiliary_loss_clip": 0.01747204,
      "auxiliary_loss_mlp": 0.0115424,
      "balance_loss_clip": 1.28511512,
      "balance_loss_mlp": 1.08476448,
      "epoch": 0.004148504434089885,
      "flos": 19427817000960.0,
      "grad_norm": 2.2931216646069092,
      "language_loss": 0.96014255,
      "learning_rate": 2.7261387181735195e-06,
      "loss": 0.98915702,
      "num_input_tokens_seen": 1313525,
      "step": 69,
      "time_per_iteration": 2.8138387203216553
    },
    {
      "auxiliary_loss_clip": 0.01741342,
      "auxiliary_loss_mlp": 0.01160375,
      "balance_loss_clip": 1.28807163,
      "balance_loss_mlp": 1.09581161,
      "epoch": 0.004208627686757853,
      "flos": 20811023884800.0,
      "grad_norm": 2.1764096137707494,
      "language_loss": 0.98070192,
      "learning_rate": 2.7354029381999196e-06,
      "loss": 1.00971913,
      "num_input_tokens_seen": 1330505,
      "step": 70,
      "time_per_iteration": 2.8319084644317627
    },
    {
      "auxiliary_loss_clip": 0.0174721,
      "auxiliary_loss_mlp": 0.01145619,
      "balance_loss_clip": 1.27791202,
      "balance_loss_mlp": 1.07685876,
      "epoch": 0.004268750939425823,
      "flos": 19098623831040.0,
      "grad_norm": 2.9300158782571324,
      "language_loss": 0.94016141,
      "learning_rate": 2.7445357464116983e-06,
      "loss": 0.96908975,
      "num_input_tokens_seen": 1349615,
      "step": 71,
      "time_per_iteration": 2.8469433784484863
    },
    {
      "auxiliary_loss_clip": 0.01815227,
      "auxiliary_loss_mlp": 0.01294388,
      "balance_loss_clip": 1.43495834,
      "balance_loss_mlp": 1.25490558,
      "epoch": 0.004328874192093792,
      "flos": 52439635514880.0,
      "grad_norm": 2.409331683106634,
      "language_loss": 0.65682542,
      "learning_rate": 2.75354081884615e-06,
      "loss": 0.68792164,
      "num_input_tokens_seen": 1410275,
      "step": 72,
      "time_per_iteration": 3.2019593715667725
    },
    {
      "auxiliary_loss_clip": 0.01799527,
      "auxiliary_loss_mlp": 0.01271558,
      "balance_loss_clip": 1.43197393,
      "balance_loss_mlp": 1.2316941,
      "epoch": 0.004388997444761762,
      "flos": 66473239564800.0,
      "grad_norm": 2.25068040880696,
      "language_loss": 0.63694263,
      "learning_rate": 2.7624216794188286e-06,
      "loss": 0.66765356,
      "num_input_tokens_seen": 1473020,
      "step": 73,
      "time_per_iteration": 3.3545596599578857
    },
    {
      "auxiliary_loss_clip": 0.01720805,
      "auxiliary_loss_mlp": 0.01140553,
      "balance_loss_clip": 1.26912856,
      "balance_loss_mlp": 1.07279444,
      "epoch": 0.004449120697429731,
      "flos": 18952970181120.0,
      "grad_norm": 2.554977860093902,
      "language_loss": 0.86212188,
      "learning_rate": 2.771181708202938e-06,
      "loss": 0.89073551,
      "num_input_tokens_seen": 1490385,
      "step": 74,
      "time_per_iteration": 2.823498487472534
    },
    {
      "auxiliary_loss_clip": 0.0172287,
      "auxiliary_loss_mlp": 0.01162493,
      "balance_loss_clip": 1.26811171,
      "balance_loss_mlp": 1.09344697,
      "epoch": 0.004509243950097701,
      "flos": 21105491581440.0,
      "grad_norm": 3.0087618017840105,
      "language_loss": 0.97196102,
      "learning_rate": 2.779824149153005e-06,
      "loss": 1.00081468,
      "num_input_tokens_seen": 1509725,
      "step": 75,
      "time_per_iteration": 2.888415575027466
    },
    {
      "auxiliary_loss_clip": 0.0170198,
      "auxiliary_loss_mlp": 0.01142315,
      "balance_loss_clip": 1.26420689,
      "balance_loss_mlp": 1.07608271,
      "epoch": 0.004569367202765669,
      "flos": 20698730991360.0,
      "grad_norm": 2.6610382542709043,
      "language_loss": 0.87740695,
      "learning_rate": 2.788352117317012e-06,
      "loss": 0.90584993,
      "num_input_tokens_seen": 1527245,
      "step": 76,
      "time_per_iteration": 2.9226863384246826
    },
    {
      "auxiliary_loss_clip": 0.01702512,
      "auxiliary_loss_mlp": 0.01145374,
      "balance_loss_clip": 1.26239479,
      "balance_loss_mlp": 1.07656646,
      "epoch": 0.004629490455433639,
      "flos": 28658474899200.0,
      "grad_norm": 2.4272090643104574,
      "language_loss": 0.91791159,
      "learning_rate": 2.796768605577095e-06,
      "loss": 0.94639051,
      "num_input_tokens_seen": 1548930,
      "step": 77,
      "time_per_iteration": 2.8720929622650146
    },
    {
      "auxiliary_loss_clip": 0.01693018,
      "auxiliary_loss_mlp": 0.01165978,
      "balance_loss_clip": 1.26398146,
      "balance_loss_mlp": 1.09569168,
      "epoch": 0.004689613708101608,
      "flos": 11072409805440.0,
      "grad_norm": 2.2822185142383034,
      "language_loss": 0.9211635,
      "learning_rate": 2.80507649095533e-06,
      "loss": 0.94975346,
      "num_input_tokens_seen": 1565695,
      "step": 78,
      "time_per_iteration": 2.7832391262054443
    },
    {
      "auxiliary_loss_clip": 0.01689271,
      "auxiliary_loss_mlp": 0.01153255,
      "balance_loss_clip": 1.25836253,
      "balance_loss_mlp": 1.08482933,
      "epoch": 0.004749736960769578,
      "flos": 21799106184960.0,
      "grad_norm": 2.263191265943929,
      "language_loss": 0.82771945,
      "learning_rate": 2.813278540517843e-06,
      "loss": 0.85614467,
      "num_input_tokens_seen": 1582625,
      "step": 79,
      "time_per_iteration": 2.7723355293273926
    },
    {
      "auxiliary_loss_clip": 0.01702468,
      "auxiliary_loss_mlp": 0.01130708,
      "balance_loss_clip": 1.26147008,
      "balance_loss_mlp": 1.0609467,
      "epoch": 0.004809860213437547,
      "flos": 19792597570560.0,
      "grad_norm": 1.9992491725405546,
      "language_loss": 0.91272199,
      "learning_rate": 2.8213774169075505e-06,
      "loss": 0.94105375,
      "num_input_tokens_seen": 1601725,
      "step": 80,
      "time_per_iteration": 2.742046356201172
    },
    {
      "auxiliary_loss_clip": 0.01671156,
      "auxiliary_loss_mlp": 0.01144048,
      "balance_loss_clip": 1.25365841,
      "balance_loss_mlp": 1.07371473,
      "epoch": 0.004869983466105517,
      "flos": 26574327037440.0,
      "grad_norm": 2.0371265012476742,
      "language_loss": 0.95241439,
      "learning_rate": 2.829375683533245e-06,
      "loss": 0.9805665,
      "num_input_tokens_seen": 1622420,
      "step": 81,
      "time_per_iteration": 2.8996386528015137
    },
    {
      "auxiliary_loss_clip": 0.01686092,
      "auxiliary_loss_mlp": 0.01147828,
      "balance_loss_clip": 1.25779653,
      "balance_loss_mlp": 1.08149946,
      "epoch": 0.004930106718773485,
      "flos": 12823378087680.0,
      "grad_norm": 2.9441337112970296,
      "language_loss": 0.96288472,
      "learning_rate": 2.8372758094402803e-06,
      "loss": 0.99122393,
      "num_input_tokens_seen": 1640715,
      "step": 82,
      "time_per_iteration": 2.819120407104492
    },
    {
      "auxiliary_loss_clip": 0.01668255,
      "auxiliary_loss_mlp": 0.01156428,
      "balance_loss_clip": 1.2461338,
      "balance_loss_mlp": 1.08709574,
      "epoch": 0.004990229971441455,
      "flos": 25774919902080.0,
      "grad_norm": 2.6601797838877856,
      "language_loss": 0.86762071,
      "learning_rate": 2.84508017388607e-06,
      "loss": 0.89586747,
      "num_input_tokens_seen": 1662210,
      "step": 83,
      "time_per_iteration": 2.7959344387054443
    },
    {
      "auxiliary_loss_clip": 0.01662665,
      "auxiliary_loss_mlp": 0.01154043,
      "balance_loss_clip": 1.24844718,
      "balance_loss_mlp": 1.084234,
      "epoch": 0.005050353224109424,
      "flos": 17457254922240.0,
      "grad_norm": 2.5416281292503986,
      "language_loss": 0.92081314,
      "learning_rate": 2.852791070641559e-06,
      "loss": 0.94898021,
      "num_input_tokens_seen": 1681070,
      "step": 84,
      "time_per_iteration": 2.7176246643066406
    },
    {
      "auxiliary_loss_clip": 0.01647627,
      "auxiliary_loss_mlp": 0.01154949,
      "balance_loss_clip": 1.36429358,
      "balance_loss_mlp": 1.11527622,
      "epoch": 0.005110476476777394,
      "flos": 69805460367360.0,
      "grad_norm": 1.4023430227621099,
      "language_loss": 0.6252538,
      "learning_rate": 2.8604107120381682e-06,
      "loss": 0.65327954,
      "num_input_tokens_seen": 1747140,
      "step": 85,
      "time_per_iteration": 3.296835422515869
    },
    {
      "auxiliary_loss_clip": 0.01649469,
      "auxiliary_loss_mlp": 0.0112642,
      "balance_loss_clip": 1.23797417,
      "balance_loss_mlp": 1.05642033,
      "epoch": 0.005170599729445363,
      "flos": 24790105739520.0,
      "grad_norm": 1.805253124779358,
      "language_loss": 0.90709531,
      "learning_rate": 2.8679412327780482e-06,
      "loss": 0.93485421,
      "num_input_tokens_seen": 1767475,
      "step": 86,
      "time_per_iteration": 2.761484146118164
    },
    {
      "auxiliary_loss_clip": 0.01653351,
      "auxiliary_loss_mlp": 0.01158608,
      "balance_loss_clip": 1.24437881,
      "balance_loss_mlp": 1.08741617,
      "epoch": 0.005230722982113333,
      "flos": 23258048895360.0,
      "grad_norm": 2.3398213465495776,
      "language_loss": 0.81961077,
      "learning_rate": 2.8753846935240833e-06,
      "loss": 0.8477304,
      "num_input_tokens_seen": 1784980,
      "step": 87,
      "time_per_iteration": 2.763185739517212
    },
    {
      "auxiliary_loss_clip": 0.01641581,
      "auxiliary_loss_mlp": 0.01152623,
      "balance_loss_clip": 1.24129367,
      "balance_loss_mlp": 1.08457828,
      "epoch": 0.005290846234781301,
      "flos": 16727909264640.0,
      "grad_norm": 3.1951080427559857,
      "language_loss": 0.95790672,
      "learning_rate": 2.8827430842847267e-06,
      "loss": 0.98584872,
      "num_input_tokens_seen": 1803030,
      "step": 88,
      "time_per_iteration": 2.7855517864227295
    },
    {
      "auxiliary_loss_clip": 0.01658657,
      "auxiliary_loss_mlp": 0.01147064,
      "balance_loss_clip": 1.24130976,
      "balance_loss_mlp": 1.07978201,
      "epoch": 0.005350969487449271,
      "flos": 20886077352960.0,
      "grad_norm": 3.405407923072192,
      "language_loss": 0.86023164,
      "learning_rate": 2.8900183276075957e-06,
      "loss": 0.88828892,
      "num_input_tokens_seen": 1822865,
      "step": 89,
      "time_per_iteration": 2.7517924308776855
    },
    {
      "auxiliary_loss_clip": 0.01647446,
      "auxiliary_loss_mlp": 0.01133456,
      "balance_loss_clip": 1.23541856,
      "balance_loss_mlp": 1.06727123,
      "epoch": 0.00541109274011724,
      "flos": 26209977431040.0,
      "grad_norm": 2.130771496386599,
      "language_loss": 0.9150058,
      "learning_rate": 2.8972122815946455e-06,
      "loss": 0.94281483,
      "num_input_tokens_seen": 1842435,
      "step": 90,
      "time_per_iteration": 2.7526872158050537
    },
    {
      "auxiliary_loss_clip": 0.01629409,
      "auxiliary_loss_mlp": 0.01133822,
      "balance_loss_clip": 1.23219132,
      "balance_loss_mlp": 1.06582534,
      "epoch": 0.00547121599278521,
      "flos": 21178569801600.0,
      "grad_norm": 2.6928798867856796,
      "language_loss": 0.86073506,
      "learning_rate": 2.90432674275074e-06,
      "loss": 0.88836741,
      "num_input_tokens_seen": 1860065,
      "step": 91,
      "time_per_iteration": 2.7995588779449463
    },
    {
      "auxiliary_loss_clip": 0.01628638,
      "auxiliary_loss_mlp": 0.01138916,
      "balance_loss_clip": 1.22774827,
      "balance_loss_mlp": 1.07335091,
      "epoch": 0.005531339245453179,
      "flos": 19718801078400.0,
      "grad_norm": 5.062847798051961,
      "language_loss": 0.87041199,
      "learning_rate": 2.91136344867656e-06,
      "loss": 0.8980875,
      "num_input_tokens_seen": 1878135,
      "step": 92,
      "time_per_iteration": 2.7813079357147217
    },
    {
      "auxiliary_loss_clip": 0.01620799,
      "auxiliary_loss_mlp": 0.01174163,
      "balance_loss_clip": 1.21933174,
      "balance_loss_mlp": 1.10650027,
      "epoch": 0.005591462498121149,
      "flos": 17636089760640.0,
      "grad_norm": 4.340668874696889,
      "language_loss": 0.9210887,
      "learning_rate": 2.918324080615938e-06,
      "loss": 0.94903833,
      "num_input_tokens_seen": 1894895,
      "step": 93,
      "time_per_iteration": 2.7582218647003174
    },
    {
      "auxiliary_loss_clip": 0.0163427,
      "auxiliary_loss_mlp": 0.01153574,
      "balance_loss_clip": 1.22659743,
      "balance_loss_mlp": 1.08238208,
      "epoch": 0.005651585750789117,
      "flos": 20011221699840.0,
      "grad_norm": 4.327341326162078,
      "language_loss": 0.87578797,
      "learning_rate": 2.925210265866963e-06,
      "loss": 0.90366644,
      "num_input_tokens_seen": 1913220,
      "step": 94,
      "time_per_iteration": 2.783581256866455
    },
    {
      "auxiliary_loss_clip": 0.01570285,
      "auxiliary_loss_mlp": 0.01051726,
      "balance_loss_clip": 1.31970167,
      "balance_loss_mlp": 1.01376939,
      "epoch": 0.005711709003457087,
      "flos": 59812957981440.0,
      "grad_norm": 1.3608185384271176,
      "language_loss": 0.68098927,
      "learning_rate": 2.932023580065507e-06,
      "loss": 0.70720935,
      "num_input_tokens_seen": 1970970,
      "step": 95,
      "time_per_iteration": 3.1328847408294678
    },
    {
      "auxiliary_loss_clip": 0.01612519,
      "auxiliary_loss_mlp": 0.01150182,
      "balance_loss_clip": 1.21488237,
      "balance_loss_mlp": 1.08318627,
      "epoch": 0.005771832256125056,
      "flos": 15559591495680.0,
      "grad_norm": 6.736145376327001,
      "language_loss": 0.90221369,
      "learning_rate": 2.9387655493491906e-06,
      "loss": 0.92984068,
      "num_input_tokens_seen": 1988930,
      "step": 96,
      "time_per_iteration": 2.8015241622924805
    },
    {
      "auxiliary_loss_clip": 0.01605814,
      "auxiliary_loss_mlp": 0.01142022,
      "balance_loss_clip": 1.21851277,
      "balance_loss_mlp": 1.08003318,
      "epoch": 0.005831955508793026,
      "flos": 22528380015360.0,
      "grad_norm": 3.8307865500968044,
      "language_loss": 0.89869905,
      "learning_rate": 2.9454376524092147e-06,
      "loss": 0.92617744,
      "num_input_tokens_seen": 2006285,
      "step": 97,
      "time_per_iteration": 4.387299060821533
    },
    {
      "auxiliary_loss_clip": 0.01593214,
      "auxiliary_loss_mlp": 0.01140673,
      "balance_loss_clip": 1.2102325,
      "balance_loss_mlp": 1.07200789,
      "epoch": 0.005892078761460995,
      "flos": 22049834094720.0,
      "grad_norm": 2.291581893082518,
      "language_loss": 0.76274347,
      "learning_rate": 2.952041322436969e-06,
      "loss": 0.79008234,
      "num_input_tokens_seen": 2024905,
      "step": 98,
      "time_per_iteration": 2.751507043838501
    },
    {
      "auxiliary_loss_clip": 0.01533926,
      "auxiliary_loss_mlp": 0.01036775,
      "balance_loss_clip": 1.29271698,
      "balance_loss_mlp": 1.00129879,
      "epoch": 0.005952202014128965,
      "flos": 68539143317760.0,
      "grad_norm": 1.0388395506080574,
      "language_loss": 0.65518898,
      "learning_rate": 2.9585779489718204e-06,
      "loss": 0.68089598,
      "num_input_tokens_seen": 2086220,
      "step": 99,
      "time_per_iteration": 3.3125040531158447
    },
    {
      "auxiliary_loss_clip": 0.01595694,
      "auxiliary_loss_mlp": 0.01142556,
      "balance_loss_clip": 1.21028757,
      "balance_loss_mlp": 1.07217503,
      "epoch": 0.006012325266796933,
      "flos": 22960887678720.0,
      "grad_norm": 2.051483688350497,
      "language_loss": 0.90885437,
      "learning_rate": 2.9650488796560464e-06,
      "loss": 0.93623686,
      "num_input_tokens_seen": 2103365,
      "step": 100,
      "time_per_iteration": 2.7632548809051514
    },
    {
      "auxiliary_loss_clip": 0.01607235,
      "auxiliary_loss_mlp": 0.01150276,
      "balance_loss_clip": 1.21294045,
      "balance_loss_mlp": 1.08394814,
      "epoch": 0.006072448519464903,
      "flos": 17347942857600.0,
      "grad_norm": 2.0181737234491566,
      "language_loss": 0.91081136,
      "learning_rate": 2.971455421902446e-06,
      "loss": 0.9383865,
      "num_input_tokens_seen": 2121995,
      "step": 101,
      "time_per_iteration": 2.7214279174804688
    },
    {
      "auxiliary_loss_clip": 0.015938,
      "auxiliary_loss_mlp": 0.01152009,
      "balance_loss_clip": 1.21248627,
      "balance_loss_mlp": 1.08124638,
      "epoch": 0.006132571772132872,
      "flos": 24681116897280.0,
      "grad_norm": 2.076276442041171,
      "language_loss": 0.90774924,
      "learning_rate": 2.9777988444798075e-06,
      "loss": 0.93520737,
      "num_input_tokens_seen": 2141815,
      "step": 102,
      "time_per_iteration": 2.8389108180999756
    },
    {
      "auxiliary_loss_clip": 0.01588155,
      "auxiliary_loss_mlp": 0.01133785,
      "balance_loss_clip": 1.20914173,
      "balance_loss_mlp": 1.06912589,
      "epoch": 0.006192695024800842,
      "flos": 21465675210240.0,
      "grad_norm": 2.3272829989328456,
      "language_loss": 0.88006896,
      "learning_rate": 2.9840803790210285e-06,
      "loss": 0.90728837,
      "num_input_tokens_seen": 2161125,
      "step": 103,
      "time_per_iteration": 2.768784761428833
    },
    {
      "auxiliary_loss_clip": 0.01588751,
      "auxiliary_loss_mlp": 0.01136216,
      "balance_loss_clip": 1.21138883,
      "balance_loss_mlp": 1.06998372,
      "epoch": 0.006252818277468811,
      "flos": 17420410546560.0,
      "grad_norm": 1.9182889224259552,
      "language_loss": 0.93644351,
      "learning_rate": 2.990301221458371e-06,
      "loss": 0.96369314,
      "num_input_tokens_seen": 2179510,
      "step": 104,
      "time_per_iteration": 2.7109038829803467
    },
    {
      "auxiliary_loss_clip": 0.01579421,
      "auxiliary_loss_mlp": 0.01146524,
      "balance_loss_clip": 1.20086741,
      "balance_loss_mlp": 1.08258009,
      "epoch": 0.006312941530136781,
      "flos": 19099557584640.0,
      "grad_norm": 3.0437899698059367,
      "language_loss": 0.96655375,
      "learning_rate": 2.9964625333900544e-06,
      "loss": 0.99381316,
      "num_input_tokens_seen": 2197870,
      "step": 105,
      "time_per_iteration": 2.7254133224487305
    },
    {
      "auxiliary_loss_clip": 0.01578331,
      "auxiliary_loss_mlp": 0.01158544,
      "balance_loss_clip": 1.20144236,
      "balance_loss_mlp": 1.08768642,
      "epoch": 0.006373064782804749,
      "flos": 24060831909120.0,
      "grad_norm": 3.1837681777002302,
      "language_loss": 0.87119448,
      "learning_rate": 3.002565443382063e-06,
      "loss": 0.89856327,
      "num_input_tokens_seen": 2217495,
      "step": 106,
      "time_per_iteration": 2.7705447673797607
    },
    {
      "auxiliary_loss_clip": 0.01561845,
      "auxiliary_loss_mlp": 0.01143018,
      "balance_loss_clip": 1.18746924,
      "balance_loss_mlp": 1.0751636,
      "epoch": 0.006433188035472719,
      "flos": 18332433797760.0,
      "grad_norm": 2.228856706842439,
      "language_loss": 0.83398581,
      "learning_rate": 3.008611048208843e-06,
      "loss": 0.86103439,
      "num_input_tokens_seen": 2236520,
      "step": 107,
      "time_per_iteration": 2.6885263919830322
    },
    {
      "auxiliary_loss_clip": 0.01469631,
      "auxiliary_loss_mlp": 0.0103327,
      "balance_loss_clip": 1.25210869,
      "balance_loss_mlp": 1.00179863,
      "epoch": 0.006493311288140688,
      "flos": 62562387594240.0,
      "grad_norm": 0.9900995959758047,
      "language_loss": 0.64796811,
      "learning_rate": 3.014600414036285e-06,
      "loss": 0.67299712,
      "num_input_tokens_seen": 2300140,
      "step": 108,
      "time_per_iteration": 3.278621196746826
    },
    {
      "auxiliary_loss_clip": 0.01552898,
      "auxiliary_loss_mlp": 0.01132858,
      "balance_loss_clip": 1.18960094,
      "balance_loss_mlp": 1.06424141,
      "epoch": 0.006553434540808658,
      "flos": 19500141035520.0,
      "grad_norm": 2.019247660217844,
      "language_loss": 0.97709465,
      "learning_rate": 3.0205345775501937e-06,
      "loss": 1.00395215,
      "num_input_tokens_seen": 2317320,
      "step": 109,
      "time_per_iteration": 2.750502347946167
    },
    {
      "auxiliary_loss_clip": 0.01550996,
      "auxiliary_loss_mlp": 0.01140204,
      "balance_loss_clip": 1.19136214,
      "balance_loss_mlp": 1.07430482,
      "epoch": 0.006613557793476627,
      "flos": 21105132445440.0,
      "grad_norm": 1.9540987754213832,
      "language_loss": 0.84243041,
      "learning_rate": 3.0264145470332218e-06,
      "loss": 0.86934245,
      "num_input_tokens_seen": 2337820,
      "step": 110,
      "time_per_iteration": 2.82443904876709
    },
    {
      "auxiliary_loss_clip": 0.01544634,
      "auxiliary_loss_mlp": 0.01151549,
      "balance_loss_clip": 1.18396342,
      "balance_loss_mlp": 1.08493507,
      "epoch": 0.006673681046144597,
      "flos": 26030747543040.0,
      "grad_norm": 2.4580319150483563,
      "language_loss": 0.82940048,
      "learning_rate": 3.032241303393073e-06,
      "loss": 0.85636234,
      "num_input_tokens_seen": 2358560,
      "step": 111,
      "time_per_iteration": 2.8308968544006348
    },
    {
      "auxiliary_loss_clip": 0.0154596,
      "auxiliary_loss_mlp": 0.01133366,
      "balance_loss_clip": 1.18776846,
      "balance_loss_mlp": 1.06970847,
      "epoch": 0.006733804298812566,
      "flos": 23147767163520.0,
      "grad_norm": 2.356589096997363,
      "language_loss": 0.93989801,
      "learning_rate": 3.0380158011446e-06,
      "loss": 0.9666912,
      "num_input_tokens_seen": 2379005,
      "step": 112,
      "time_per_iteration": 2.8007922172546387
    },
    {
      "auxiliary_loss_clip": 0.01549647,
      "auxiliary_loss_mlp": 0.01136979,
      "balance_loss_clip": 1.18394601,
      "balance_loss_mlp": 1.07322621,
      "epoch": 0.006793927551480535,
      "flos": 11764444210560.0,
      "grad_norm": 2.521639841990545,
      "language_loss": 0.79509294,
      "learning_rate": 3.0437389693482466e-06,
      "loss": 0.82195914,
      "num_input_tokens_seen": 2395610,
      "step": 113,
      "time_per_iteration": 2.7599966526031494
    },
    {
      "auxiliary_loss_clip": 0.0153736,
      "auxiliary_loss_mlp": 0.01131524,
      "balance_loss_clip": 1.18028498,
      "balance_loss_mlp": 1.06562555,
      "epoch": 0.006854050804148504,
      "flos": 19171953446400.0,
      "grad_norm": 2.343117351168218,
      "language_loss": 0.93439317,
      "learning_rate": 3.0494117125071475e-06,
      "loss": 0.96108204,
      "num_input_tokens_seen": 2415005,
      "step": 114,
      "time_per_iteration": 2.723540782928467
    },
    {
      "auxiliary_loss_clip": 0.01544971,
      "auxiliary_loss_mlp": 0.01138932,
      "balance_loss_clip": 1.17997146,
      "balance_loss_mlp": 1.07918465,
      "epoch": 0.006914174056816474,
      "flos": 21981891519360.0,
      "grad_norm": 1.9509019191057126,
      "language_loss": 0.9463321,
      "learning_rate": 3.055034911425055e-06,
      "loss": 0.97317111,
      "num_input_tokens_seen": 2433965,
      "step": 115,
      "time_per_iteration": 2.7077698707580566
    },
    {
      "auxiliary_loss_clip": 0.01537699,
      "auxiliary_loss_mlp": 0.01118178,
      "balance_loss_clip": 1.17675614,
      "balance_loss_mlp": 1.05151677,
      "epoch": 0.006974297309484443,
      "flos": 16289152634880.0,
      "grad_norm": 10.363795807176915,
      "language_loss": 0.82148951,
      "learning_rate": 3.0606094240271244e-06,
      "loss": 0.84804827,
      "num_input_tokens_seen": 2451605,
      "step": 116,
      "time_per_iteration": 2.681190013885498
    },
    {
      "auxiliary_loss_clip": 0.01528803,
      "auxiliary_loss_mlp": 0.01126189,
      "balance_loss_clip": 1.17677391,
      "balance_loss_mlp": 1.06219721,
      "epoch": 0.007034420562152413,
      "flos": 26104005331200.0,
      "grad_norm": 2.4150591879391627,
      "language_loss": 0.88368428,
      "learning_rate": 3.0661360861454656e-06,
      "loss": 0.91023421,
      "num_input_tokens_seen": 2472035,
      "step": 117,
      "time_per_iteration": 2.776143789291382
    },
    {
      "auxiliary_loss_clip": 0.01527909,
      "auxiliary_loss_mlp": 0.01146127,
      "balance_loss_clip": 1.17495561,
      "balance_loss_mlp": 1.08041906,
      "epoch": 0.007094543814820382,
      "flos": 14204609723520.0,
      "grad_norm": 2.3639764059040265,
      "language_loss": 0.8454417,
      "learning_rate": 3.071615712271274e-06,
      "loss": 0.87218207,
      "num_input_tokens_seen": 2489285,
      "step": 118,
      "time_per_iteration": 2.7110469341278076
    },
    {
      "auxiliary_loss_clip": 0.01538161,
      "auxiliary_loss_mlp": 0.01163868,
      "balance_loss_clip": 1.1759789,
      "balance_loss_mlp": 1.0984937,
      "epoch": 0.007154667067488351,
      "flos": 14976007228800.0,
      "grad_norm": 2.231843342078736,
      "language_loss": 0.99319011,
      "learning_rate": 3.0770490962752172e-06,
      "loss": 1.02021039,
      "num_input_tokens_seen": 2506460,
      "step": 119,
      "time_per_iteration": 2.674121856689453
    },
    {
      "auxiliary_loss_clip": 0.01540018,
      "auxiliary_loss_mlp": 0.01120611,
      "balance_loss_clip": 1.17242217,
      "balance_loss_mlp": 1.05738258,
      "epoch": 0.00721479032015632,
      "flos": 20193288762240.0,
      "grad_norm": 2.7981733983226764,
      "language_loss": 0.8963809,
      "learning_rate": 3.082437012097686e-06,
      "loss": 0.92298722,
      "num_input_tokens_seen": 2525565,
      "step": 120,
      "time_per_iteration": 2.745962381362915
    },
    {
      "auxiliary_loss_clip": 0.01524916,
      "auxiliary_loss_mlp": 0.01129465,
      "balance_loss_clip": 1.1734432,
      "balance_loss_mlp": 1.06513989,
      "epoch": 0.00727491357282429,
      "flos": 23147228459520.0,
      "grad_norm": 1.797716104424251,
      "language_loss": 0.93491542,
      "learning_rate": 3.0877802144103967e-06,
      "loss": 0.96145928,
      "num_input_tokens_seen": 2546605,
      "step": 121,
      "time_per_iteration": 2.7924466133117676
    },
    {
      "auxiliary_loss_clip": 0.01526294,
      "auxiliary_loss_mlp": 0.0114832,
      "balance_loss_clip": 1.17395604,
      "balance_loss_mlp": 1.08490098,
      "epoch": 0.007335036825492259,
      "flos": 15521669712000.0,
      "grad_norm": 2.3704869501778285,
      "language_loss": 0.90462255,
      "learning_rate": 3.09307943925077e-06,
      "loss": 0.93136871,
      "num_input_tokens_seen": 2560730,
      "step": 122,
      "time_per_iteration": 2.930413246154785
    },
    {
      "auxiliary_loss_clip": 0.01521826,
      "auxiliary_loss_mlp": 0.01146566,
      "balance_loss_clip": 1.1681807,
      "balance_loss_mlp": 1.07861674,
      "epoch": 0.007395160078160229,
      "flos": 24243365848320.0,
      "grad_norm": 2.4867163179710037,
      "language_loss": 0.92660481,
      "learning_rate": 3.0983354046304154e-06,
      "loss": 0.95328873,
      "num_input_tokens_seen": 2579550,
      "step": 123,
      "time_per_iteration": 2.7484309673309326
    },
    {
      "auxiliary_loss_clip": 0.01519363,
      "auxiliary_loss_mlp": 0.01127611,
      "balance_loss_clip": 1.16324139,
      "balance_loss_mlp": 1.0651449,
      "epoch": 0.007455283330828198,
      "flos": 31759792099200.0,
      "grad_norm": 2.366639004459226,
      "language_loss": 0.71187961,
      "learning_rate": 3.103548811118979e-06,
      "loss": 0.73834932,
      "num_input_tokens_seen": 2600390,
      "step": 124,
      "time_per_iteration": 2.8419976234436035
    },
    {
      "auxiliary_loss_clip": 0.01506936,
      "auxiliary_loss_mlp": 0.01125571,
      "balance_loss_clip": 1.16464007,
      "balance_loss_mlp": 1.06167519,
      "epoch": 0.007515406583496167,
      "flos": 26615157822720.0,
      "grad_norm": 2.1632751269766106,
      "language_loss": 0.88450015,
      "learning_rate": 3.108720342404542e-06,
      "loss": 0.91082525,
      "num_input_tokens_seen": 2620770,
      "step": 125,
      "time_per_iteration": 2.823296308517456
    },
    {
      "auxiliary_loss_clip": 0.01522239,
      "auxiliary_loss_mlp": 0.01142214,
      "balance_loss_clip": 1.16456664,
      "balance_loss_mlp": 1.07912827,
      "epoch": 0.007575529836164136,
      "flos": 18223696350720.0,
      "grad_norm": 2.6632616920164067,
      "language_loss": 0.82381976,
      "learning_rate": 3.1138506658316945e-06,
      "loss": 0.85046428,
      "num_input_tokens_seen": 2639900,
      "step": 126,
      "time_per_iteration": 2.7325809001922607
    },
    {
      "auxiliary_loss_clip": 0.015153,
      "auxiliary_loss_mlp": 0.01142869,
      "balance_loss_clip": 1.16330886,
      "balance_loss_mlp": 1.08088017,
      "epoch": 0.007635653088832106,
      "flos": 21580410228480.0,
      "grad_norm": 3.925284628341409,
      "language_loss": 0.6743899,
      "learning_rate": 3.1189404329183404e-06,
      "loss": 0.7009716,
      "num_input_tokens_seen": 2657450,
      "step": 127,
      "time_per_iteration": 2.709821939468384
    },
    {
      "auxiliary_loss_clip": 0.01503057,
      "auxiliary_loss_mlp": 0.01132416,
      "balance_loss_clip": 1.165169,
      "balance_loss_mlp": 1.06861567,
      "epoch": 0.007695776341500075,
      "flos": 25375054723200.0,
      "grad_norm": 2.0535131533503734,
      "language_loss": 0.8819322,
      "learning_rate": 3.1239902798522317e-06,
      "loss": 0.90828693,
      "num_input_tokens_seen": 2678150,
      "step": 128,
      "time_per_iteration": 2.764707565307617
    },
    {
      "auxiliary_loss_clip": 0.01505955,
      "auxiliary_loss_mlp": 0.01144223,
      "balance_loss_clip": 1.16043079,
      "balance_loss_mlp": 1.08042252,
      "epoch": 0.007755899594168045,
      "flos": 22343906741760.0,
      "grad_norm": 2.6427711693827005,
      "language_loss": 0.84719259,
      "learning_rate": 3.129000827968184e-06,
      "loss": 0.87369436,
      "num_input_tokens_seen": 2698290,
      "step": 129,
      "time_per_iteration": 2.7472774982452393
    },
    {
      "auxiliary_loss_clip": 0.01497871,
      "auxiliary_loss_mlp": 0.01130211,
      "balance_loss_clip": 1.15871263,
      "balance_loss_mlp": 1.06655347,
      "epoch": 0.007816022846836013,
      "flos": 22638230784000.0,
      "grad_norm": 2.366492959329914,
      "language_loss": 0.97564614,
      "learning_rate": 3.133972684206866e-06,
      "loss": 1.00192702,
      "num_input_tokens_seen": 2717630,
      "step": 130,
      "time_per_iteration": 2.6955018043518066
    },
    {
      "auxiliary_loss_clip": 0.01492272,
      "auxiliary_loss_mlp": 0.01134965,
      "balance_loss_clip": 1.15630865,
      "balance_loss_mlp": 1.06987715,
      "epoch": 0.007876146099503984,
      "flos": 18182901479040.0,
      "grad_norm": 2.2164470079204572,
      "language_loss": 0.82658112,
      "learning_rate": 3.138906441556014e-06,
      "loss": 0.85285342,
      "num_input_tokens_seen": 2735835,
      "step": 131,
      "time_per_iteration": 2.722247362136841
    },
    {
      "auxiliary_loss_clip": 0.01500937,
      "auxiliary_loss_mlp": 0.01128359,
      "balance_loss_clip": 1.15885806,
      "balance_loss_mlp": 1.06694245,
      "epoch": 0.007936269352171952,
      "flos": 27119486730240.0,
      "grad_norm": 2.7663180664822193,
      "language_loss": 0.82781422,
      "learning_rate": 3.143802679474861e-06,
      "loss": 0.85410714,
      "num_input_tokens_seen": 2756335,
      "step": 132,
      "time_per_iteration": 2.7937612533569336
    },
    {
      "auxiliary_loss_clip": 0.01491919,
      "auxiliary_loss_mlp": 0.01128624,
      "balance_loss_clip": 1.15346444,
      "balance_loss_mlp": 1.0664922,
      "epoch": 0.007996392604839923,
      "flos": 19026335710080.0,
      "grad_norm": 2.182366740159355,
      "language_loss": 0.95499313,
      "learning_rate": 3.1486619643025565e-06,
      "loss": 0.98119843,
      "num_input_tokens_seen": 2775090,
      "step": 133,
      "time_per_iteration": 2.7380354404449463
    },
    {
      "auxiliary_loss_clip": 0.01487746,
      "auxiliary_loss_mlp": 0.0112871,
      "balance_loss_clip": 1.16170454,
      "balance_loss_mlp": 1.06843781,
      "epoch": 0.008056515857507891,
      "flos": 25484151306240.0,
      "grad_norm": 1.8164116645967854,
      "language_loss": 0.73478442,
      "learning_rate": 3.153484849651286e-06,
      "loss": 0.76094896,
      "num_input_tokens_seen": 2795320,
      "step": 134,
      "time_per_iteration": 2.7483408451080322
    },
    {
      "auxiliary_loss_clip": 0.01484621,
      "auxiliary_loss_mlp": 0.01132134,
      "balance_loss_clip": 1.15115011,
      "balance_loss_mlp": 1.06695068,
      "epoch": 0.00811663911017586,
      "flos": 20557566541440.0,
      "grad_norm": 5.027018494085059,
      "language_loss": 0.88792509,
      "learning_rate": 3.1582718767847806e-06,
      "loss": 0.91409266,
      "num_input_tokens_seen": 2812815,
      "step": 135,
      "time_per_iteration": 2.6838128566741943
    },
    {
      "auxiliary_loss_clip": 0.01487119,
      "auxiliary_loss_mlp": 0.0113257,
      "balance_loss_clip": 1.15490174,
      "balance_loss_mlp": 1.06714821,
      "epoch": 0.00817676236284383,
      "flos": 18799738761600.0,
      "grad_norm": 1.9282722528396903,
      "language_loss": 0.89138198,
      "learning_rate": 3.1630235749828485e-06,
      "loss": 0.91757882,
      "num_input_tokens_seen": 2830445,
      "step": 136,
      "time_per_iteration": 2.726475238800049
    },
    {
      "auxiliary_loss_clip": 0.01483417,
      "auxiliary_loss_mlp": 0.01110724,
      "balance_loss_clip": 1.1494019,
      "balance_loss_mlp": 1.05078554,
      "epoch": 0.008236885615511799,
      "flos": 23873593288320.0,
      "grad_norm": 2.2984339413846078,
      "language_loss": 0.84091324,
      "learning_rate": 3.1677404618925676e-06,
      "loss": 0.86685467,
      "num_input_tokens_seen": 2846965,
      "step": 137,
      "time_per_iteration": 7.4708640575408936
    },
    {
      "auxiliary_loss_clip": 0.01481848,
      "auxiliary_loss_mlp": 0.01118837,
      "balance_loss_clip": 1.1500535,
      "balance_loss_mlp": 1.05894589,
      "epoch": 0.00829700886817977,
      "flos": 24643626076800.0,
      "grad_norm": 1.69378413504035,
      "language_loss": 0.9018681,
      "learning_rate": 3.1724230438666953e-06,
      "loss": 0.92787492,
      "num_input_tokens_seen": 2867520,
      "step": 138,
      "time_per_iteration": 4.311830520629883
    },
    {
      "auxiliary_loss_clip": 0.01469655,
      "auxiliary_loss_mlp": 0.01123604,
      "balance_loss_clip": 1.14824438,
      "balance_loss_mlp": 1.05904007,
      "epoch": 0.008357132120847738,
      "flos": 25262007644160.0,
      "grad_norm": 2.1515203004813785,
      "language_loss": 0.91478992,
      "learning_rate": 3.177071816289865e-06,
      "loss": 0.94072247,
      "num_input_tokens_seen": 2885675,
      "step": 139,
      "time_per_iteration": 2.7678122520446777
    },
    {
      "auxiliary_loss_clip": 0.01486799,
      "auxiliary_loss_mlp": 0.01124947,
      "balance_loss_clip": 1.15521085,
      "balance_loss_mlp": 1.06195688,
      "epoch": 0.008417255373515706,
      "flos": 27344898529920.0,
      "grad_norm": 2.305315677890536,
      "language_loss": 0.85667789,
      "learning_rate": 3.181687263893095e-06,
      "loss": 0.88279533,
      "num_input_tokens_seen": 2905960,
      "step": 140,
      "time_per_iteration": 2.8557639122009277
    },
    {
      "auxiliary_loss_clip": 0.01473538,
      "auxiliary_loss_mlp": 0.01122701,
      "balance_loss_clip": 1.14923954,
      "balance_loss_mlp": 1.06166625,
      "epoch": 0.008477378626183677,
      "flos": 17639070589440.0,
      "grad_norm": 2.3443620963590455,
      "language_loss": 0.84346074,
      "learning_rate": 3.186269861057098e-06,
      "loss": 0.86942315,
      "num_input_tokens_seen": 2922780,
      "step": 141,
      "time_per_iteration": 2.7656807899475098
    },
    {
      "auxiliary_loss_clip": 0.01477141,
      "auxiliary_loss_mlp": 0.01135217,
      "balance_loss_clip": 1.14718878,
      "balance_loss_mlp": 1.07360983,
      "epoch": 0.008537501878851645,
      "flos": 13881342297600.0,
      "grad_norm": 2.29020652115343,
      "language_loss": 0.8105557,
      "learning_rate": 3.1908200721048745e-06,
      "loss": 0.83667928,
      "num_input_tokens_seen": 2938765,
      "step": 142,
      "time_per_iteration": 2.747598171234131
    },
    {
      "auxiliary_loss_clip": 0.01378886,
      "auxiliary_loss_mlp": 0.01060004,
      "balance_loss_clip": 1.19240355,
      "balance_loss_mlp": 1.03406358,
      "epoch": 0.008597625131519616,
      "flos": 71248101281280.0,
      "grad_norm": 1.056887207538052,
      "language_loss": 0.66899812,
      "learning_rate": 3.195338351584042e-06,
      "loss": 0.69338703,
      "num_input_tokens_seen": 3006665,
      "step": 143,
      "time_per_iteration": 3.346982002258301
    },
    {
      "auxiliary_loss_clip": 0.01467707,
      "auxiliary_loss_mlp": 0.01123721,
      "balance_loss_clip": 1.14666772,
      "balance_loss_mlp": 1.06273365,
      "epoch": 0.008657748384187584,
      "flos": 17602836744960.0,
      "grad_norm": 2.6467048454978523,
      "language_loss": 0.84356761,
      "learning_rate": 3.1998251445393258e-06,
      "loss": 0.86948192,
      "num_input_tokens_seen": 3024335,
      "step": 144,
      "time_per_iteration": 2.762087345123291
    },
    {
      "auxiliary_loss_clip": 0.01455701,
      "auxiliary_loss_mlp": 0.01114511,
      "balance_loss_clip": 1.14058816,
      "balance_loss_mlp": 1.05085373,
      "epoch": 0.008717871636855555,
      "flos": 19715317459200.0,
      "grad_norm": 1.8692883316747366,
      "language_loss": 0.88353741,
      "learning_rate": 3.204280886775619e-06,
      "loss": 0.90923953,
      "num_input_tokens_seen": 3043300,
      "step": 145,
      "time_per_iteration": 2.7050039768218994
    },
    {
      "auxiliary_loss_clip": 0.01470385,
      "auxiliary_loss_mlp": 0.01121817,
      "balance_loss_clip": 1.14247775,
      "balance_loss_mlp": 1.05873132,
      "epoch": 0.008777994889523523,
      "flos": 24717422568960.0,
      "grad_norm": 1.860830881508538,
      "language_loss": 0.86182559,
      "learning_rate": 3.208706005112005e-06,
      "loss": 0.88774765,
      "num_input_tokens_seen": 3064610,
      "step": 146,
      "time_per_iteration": 2.741013288497925
    },
    {
      "auxiliary_loss_clip": 0.01356998,
      "auxiliary_loss_mlp": 0.01029681,
      "balance_loss_clip": 1.18072379,
      "balance_loss_mlp": 1.00431335,
      "epoch": 0.008838118142191492,
      "flos": 70132067758080.0,
      "grad_norm": 0.8598047517885464,
      "language_loss": 0.60122073,
      "learning_rate": 3.213100917627104e-06,
      "loss": 0.6250875,
      "num_input_tokens_seen": 3130385,
      "step": 147,
      "time_per_iteration": 3.27382230758667
    },
    {
      "auxiliary_loss_clip": 0.01463009,
      "auxiliary_loss_mlp": 0.01123472,
      "balance_loss_clip": 1.14658976,
      "balance_loss_mlp": 1.06548882,
      "epoch": 0.008898241394859462,
      "flos": 20044797937920.0,
      "grad_norm": 1.8116070485228748,
      "language_loss": 0.84620225,
      "learning_rate": 3.2174660338961135e-06,
      "loss": 0.87206709,
      "num_input_tokens_seen": 3149760,
      "step": 148,
      "time_per_iteration": 2.72910475730896
    },
    {
      "auxiliary_loss_clip": 0.01466623,
      "auxiliary_loss_mlp": 0.01144944,
      "balance_loss_clip": 1.14777792,
      "balance_loss_mlp": 1.07985532,
      "epoch": 0.008958364647527431,
      "flos": 10743611685120.0,
      "grad_norm": 2.5530775415688205,
      "language_loss": 0.88680327,
      "learning_rate": 3.2218017552198588e-06,
      "loss": 0.91291893,
      "num_input_tokens_seen": 3164500,
      "step": 149,
      "time_per_iteration": 2.688528537750244
    },
    {
      "auxiliary_loss_clip": 0.01463954,
      "auxiliary_loss_mlp": 0.01114885,
      "balance_loss_clip": 1.14290714,
      "balance_loss_mlp": 1.05728304,
      "epoch": 0.009018487900195401,
      "flos": 29127467802240.0,
      "grad_norm": 2.1996557200804823,
      "language_loss": 0.93269086,
      "learning_rate": 3.226108474846181e-06,
      "loss": 0.95847929,
      "num_input_tokens_seen": 3182455,
      "step": 150,
      "time_per_iteration": 2.7901580333709717
    },
    {
      "auxiliary_loss_clip": 0.01450819,
      "auxiliary_loss_mlp": 0.01114571,
      "balance_loss_clip": 1.13812149,
      "balance_loss_mlp": 1.05839944,
      "epoch": 0.00907861115286337,
      "flos": 32963661354240.0,
      "grad_norm": 4.690239135210318,
      "language_loss": 0.7421813,
      "learning_rate": 3.2303865781839817e-06,
      "loss": 0.7678352,
      "num_input_tokens_seen": 3203995,
      "step": 151,
      "time_per_iteration": 2.79590106010437
    },
    {
      "auxiliary_loss_clip": 0.01463077,
      "auxiliary_loss_mlp": 0.01128244,
      "balance_loss_clip": 1.14311624,
      "balance_loss_mlp": 1.06954527,
      "epoch": 0.009138734405531338,
      "flos": 21762441377280.0,
      "grad_norm": 4.291097242497492,
      "language_loss": 0.88460332,
      "learning_rate": 3.234636443010188e-06,
      "loss": 0.9105165,
      "num_input_tokens_seen": 3222575,
      "step": 152,
      "time_per_iteration": 2.701775550842285
    },
    {
      "auxiliary_loss_clip": 0.01462099,
      "auxiliary_loss_mlp": 0.01122264,
      "balance_loss_clip": 1.14743185,
      "balance_loss_mlp": 1.06275451,
      "epoch": 0.009198857658199309,
      "flos": 20842517134080.0,
      "grad_norm": 3.861411936226758,
      "language_loss": 0.83918798,
      "learning_rate": 3.238858439669943e-06,
      "loss": 0.8650316,
      "num_input_tokens_seen": 3240180,
      "step": 153,
      "time_per_iteration": 2.730654716491699
    },
    {
      "auxiliary_loss_clip": 0.01453756,
      "auxiliary_loss_mlp": 0.01136244,
      "balance_loss_clip": 1.14024806,
      "balance_loss_mlp": 1.07554269,
      "epoch": 0.009258980910867277,
      "flos": 24827381078400.0,
      "grad_norm": 1.8788427995178905,
      "language_loss": 0.89924759,
      "learning_rate": 3.2430529312702712e-06,
      "loss": 0.92514759,
      "num_input_tokens_seen": 3259800,
      "step": 154,
      "time_per_iteration": 2.8150386810302734
    },
    {
      "auxiliary_loss_clip": 0.01457041,
      "auxiliary_loss_mlp": 0.01148182,
      "balance_loss_clip": 1.1422174,
      "balance_loss_mlp": 1.08934021,
      "epoch": 0.009319104163535248,
      "flos": 28767786963840.0,
      "grad_norm": 2.155148564981828,
      "language_loss": 0.89730597,
      "learning_rate": 3.2472202738674737e-06,
      "loss": 0.9233582,
      "num_input_tokens_seen": 3280400,
      "step": 155,
      "time_per_iteration": 2.7780215740203857
    },
    {
      "auxiliary_loss_clip": 0.01462257,
      "auxiliary_loss_mlp": 0.01115972,
      "balance_loss_clip": 1.14140153,
      "balance_loss_mlp": 1.0580368,
      "epoch": 0.009379227416203216,
      "flos": 16582004219520.0,
      "grad_norm": 2.6722626388977986,
      "language_loss": 0.86758631,
      "learning_rate": 3.2513608166485063e-06,
      "loss": 0.8933686,
      "num_input_tokens_seen": 3297600,
      "step": 156,
      "time_per_iteration": 2.7195818424224854
    },
    {
      "auxiliary_loss_clip": 0.01460326,
      "auxiliary_loss_mlp": 0.01116019,
      "balance_loss_clip": 1.14530039,
      "balance_loss_mlp": 1.05770147,
      "epoch": 0.009439350668871187,
      "flos": 18329919845760.0,
      "grad_norm": 2.3212743339319926,
      "language_loss": 0.99652225,
      "learning_rate": 3.2554749021065498e-06,
      "loss": 1.0222857,
      "num_input_tokens_seen": 3313635,
      "step": 157,
      "time_per_iteration": 2.7530624866485596
    },
    {
      "auxiliary_loss_clip": 0.01445494,
      "auxiliary_loss_mlp": 0.01139991,
      "balance_loss_clip": 1.14011836,
      "balance_loss_mlp": 1.08162606,
      "epoch": 0.009499473921539155,
      "flos": 24349912565760.0,
      "grad_norm": 2.2650385025378834,
      "language_loss": 0.88388717,
      "learning_rate": 3.2595628662110186e-06,
      "loss": 0.90974212,
      "num_input_tokens_seen": 3333735,
      "step": 158,
      "time_per_iteration": 2.744640588760376
    },
    {
      "auxiliary_loss_clip": 0.01451838,
      "auxiliary_loss_mlp": 0.01122147,
      "balance_loss_clip": 1.13977575,
      "balance_loss_mlp": 1.0630666,
      "epoch": 0.009559597174207124,
      "flos": 16399326625920.0,
      "grad_norm": 2.1807440045696165,
      "language_loss": 0.86407602,
      "learning_rate": 3.2636250385721982e-06,
      "loss": 0.88981581,
      "num_input_tokens_seen": 3348800,
      "step": 159,
      "time_per_iteration": 2.7330005168914795
    },
    {
      "auxiliary_loss_clip": 0.01441743,
      "auxiliary_loss_mlp": 0.01137796,
      "balance_loss_clip": 1.13474953,
      "balance_loss_mlp": 1.07752383,
      "epoch": 0.009619720426875094,
      "flos": 22856890826880.0,
      "grad_norm": 1.7296815250329798,
      "language_loss": 0.86756837,
      "learning_rate": 3.2676617426007263e-06,
      "loss": 0.89336377,
      "num_input_tokens_seen": 3368595,
      "step": 160,
      "time_per_iteration": 2.844817876815796
    },
    {
      "auxiliary_loss_clip": 0.01447614,
      "auxiliary_loss_mlp": 0.0112266,
      "balance_loss_clip": 1.13978457,
      "balance_loss_mlp": 1.06725168,
      "epoch": 0.009679843679543063,
      "flos": 19135001329920.0,
      "grad_norm": 2.462408333273543,
      "language_loss": 0.91543746,
      "learning_rate": 3.2716732956621042e-06,
      "loss": 0.94114017,
      "num_input_tokens_seen": 3384975,
      "step": 161,
      "time_per_iteration": 2.667666435241699
    },
    {
      "auxiliary_loss_clip": 0.01453392,
      "auxiliary_loss_mlp": 0.01111804,
      "balance_loss_clip": 1.14104879,
      "balance_loss_mlp": 1.05610919,
      "epoch": 0.009739966932211033,
      "flos": 20302995876480.0,
      "grad_norm": 1.7914334411859298,
      "language_loss": 0.91582954,
      "learning_rate": 3.2756600092264203e-06,
      "loss": 0.94148147,
      "num_input_tokens_seen": 3404755,
      "step": 162,
      "time_per_iteration": 2.6779961585998535
    },
    {
      "auxiliary_loss_clip": 0.0131522,
      "auxiliary_loss_mlp": 0.01056953,
      "balance_loss_clip": 1.15019548,
      "balance_loss_mlp": 1.03358769,
      "epoch": 0.009800090184879002,
      "flos": 67034234177280.0,
      "grad_norm": 1.183297200633083,
      "language_loss": 0.72292268,
      "learning_rate": 3.279622189013474e-06,
      "loss": 0.74664438,
      "num_input_tokens_seen": 3467210,
      "step": 163,
      "time_per_iteration": 3.226755142211914
    },
    {
      "auxiliary_loss_clip": 0.01439788,
      "auxiliary_loss_mlp": 0.01116102,
      "balance_loss_clip": 1.13873029,
      "balance_loss_mlp": 1.05921507,
      "epoch": 0.00986021343754697,
      "flos": 17164690646400.0,
      "grad_norm": 3.3372881081540937,
      "language_loss": 0.84684807,
      "learning_rate": 3.283560135133457e-06,
      "loss": 0.87240696,
      "num_input_tokens_seen": 3483220,
      "step": 164,
      "time_per_iteration": 2.768935203552246
    },
    {
      "auxiliary_loss_clip": 0.01430933,
      "auxiliary_loss_mlp": 0.0110117,
      "balance_loss_clip": 1.13048434,
      "balance_loss_mlp": 1.04533219,
      "epoch": 0.00992033669021494,
      "flos": 17749424148480.0,
      "grad_norm": 4.079659732294038,
      "language_loss": 0.89080763,
      "learning_rate": 3.2874741422233565e-06,
      "loss": 0.91612864,
      "num_input_tokens_seen": 3501465,
      "step": 165,
      "time_per_iteration": 2.673292875289917
    },
    {
      "auxiliary_loss_clip": 0.01433192,
      "auxiliary_loss_mlp": 0.01128138,
      "balance_loss_clip": 1.13111067,
      "balance_loss_mlp": 1.06819916,
      "epoch": 0.00998045994288291,
      "flos": 25297164080640.0,
      "grad_norm": 1.7359539169577796,
      "language_loss": 0.79931343,
      "learning_rate": 3.2913644995792465e-06,
      "loss": 0.82492673,
      "num_input_tokens_seen": 3520480,
      "step": 166,
      "time_per_iteration": 2.762742757797241
    },
    {
      "auxiliary_loss_clip": 0.01438026,
      "auxiliary_loss_mlp": 0.01129718,
      "balance_loss_clip": 1.13488948,
      "balance_loss_mlp": 1.07066131,
      "epoch": 0.01004058319555088,
      "flos": 32298954220800.0,
      "grad_norm": 2.3252666324684585,
      "language_loss": 0.92125285,
      "learning_rate": 3.2952314912845914e-06,
      "loss": 0.94693023,
      "num_input_tokens_seen": 3539570,
      "step": 167,
      "time_per_iteration": 2.970964193344116
    },
    {
      "auxiliary_loss_clip": 0.01429698,
      "auxiliary_loss_mlp": 0.01133324,
      "balance_loss_clip": 1.13294363,
      "balance_loss_mlp": 1.07734346,
      "epoch": 0.010100706448218848,
      "flos": 11319941404800.0,
      "grad_norm": 13.512238716069085,
      "language_loss": 0.90781063,
      "learning_rate": 3.299075396334735e-06,
      "loss": 0.93344086,
      "num_input_tokens_seen": 3555465,
      "step": 168,
      "time_per_iteration": 2.8039841651916504
    },
    {
      "auxiliary_loss_clip": 0.01424367,
      "auxiliary_loss_mlp": 0.01104795,
      "balance_loss_clip": 1.12848639,
      "balance_loss_mlp": 1.04700291,
      "epoch": 0.010160829700886819,
      "flos": 29719491765120.0,
      "grad_norm": 1.6705351130563955,
      "language_loss": 0.87173021,
      "learning_rate": 3.3028964887576868e-06,
      "loss": 0.89702177,
      "num_input_tokens_seen": 3578970,
      "step": 169,
      "time_per_iteration": 2.8215444087982178
    },
    {
      "auxiliary_loss_clip": 0.01425902,
      "auxiliary_loss_mlp": 0.01110538,
      "balance_loss_clip": 1.13139379,
      "balance_loss_mlp": 1.05317438,
      "epoch": 0.010220952953554787,
      "flos": 20412343854720.0,
      "grad_norm": 1.7404257397879006,
      "language_loss": 0.84622329,
      "learning_rate": 3.306695037731344e-06,
      "loss": 0.87158769,
      "num_input_tokens_seen": 3597275,
      "step": 170,
      "time_per_iteration": 2.6759181022644043
    },
    {
      "auxiliary_loss_clip": 0.0143612,
      "auxiliary_loss_mlp": 0.01137162,
      "balance_loss_clip": 1.13149834,
      "balance_loss_mlp": 1.07874942,
      "epoch": 0.010281076206222756,
      "flos": 31285124847360.0,
      "grad_norm": 2.174517661608974,
      "language_loss": 0.89936447,
      "learning_rate": 3.3104713076972827e-06,
      "loss": 0.92509729,
      "num_input_tokens_seen": 3618905,
      "step": 171,
      "time_per_iteration": 2.800394058227539
    },
    {
      "auxiliary_loss_clip": 0.01430673,
      "auxiliary_loss_mlp": 0.01108779,
      "balance_loss_clip": 1.1347487,
      "balance_loss_mlp": 1.05382347,
      "epoch": 0.010341199458890726,
      "flos": 21982286568960.0,
      "grad_norm": 1.938241860949196,
      "language_loss": 0.88895655,
      "learning_rate": 3.314225558471224e-06,
      "loss": 0.91435111,
      "num_input_tokens_seen": 3639610,
      "step": 172,
      "time_per_iteration": 2.755190849304199
    },
    {
      "auxiliary_loss_clip": 0.01418638,
      "auxiliary_loss_mlp": 0.01118471,
      "balance_loss_clip": 1.12744904,
      "balance_loss_mlp": 1.06270456,
      "epoch": 0.010401322711558695,
      "flos": 30810529422720.0,
      "grad_norm": 1.7925778946034159,
      "language_loss": 0.80943549,
      "learning_rate": 3.317958045350308e-06,
      "loss": 0.83480656,
      "num_input_tokens_seen": 3664030,
      "step": 173,
      "time_per_iteration": 2.751945734024048
    },
    {
      "auxiliary_loss_clip": 0.01429615,
      "auxiliary_loss_mlp": 0.01107965,
      "balance_loss_clip": 1.13108575,
      "balance_loss_mlp": 1.05534625,
      "epoch": 0.010461445964226665,
      "flos": 24715124098560.0,
      "grad_norm": 2.1644843911099216,
      "language_loss": 0.82763064,
      "learning_rate": 3.3216690192172596e-06,
      "loss": 0.85300648,
      "num_input_tokens_seen": 3683615,
      "step": 174,
      "time_per_iteration": 2.676630735397339
    },
    {
      "auxiliary_loss_clip": 0.01423443,
      "auxiliary_loss_mlp": 0.01120976,
      "balance_loss_clip": 1.12816644,
      "balance_loss_mlp": 1.06523335,
      "epoch": 0.010521569216894634,
      "flos": 27710361457920.0,
      "grad_norm": 2.331494685324117,
      "language_loss": 0.72837007,
      "learning_rate": 3.325358726641591e-06,
      "loss": 0.75381434,
      "num_input_tokens_seen": 3704540,
      "step": 175,
      "time_per_iteration": 2.6876866817474365
    },
    {
      "auxiliary_loss_clip": 0.01425333,
      "auxiliary_loss_mlp": 0.01127215,
      "balance_loss_clip": 1.12866652,
      "balance_loss_mlp": 1.06980324,
      "epoch": 0.010581692469562603,
      "flos": 12458346122880.0,
      "grad_norm": 4.811985773634618,
      "language_loss": 0.97983754,
      "learning_rate": 3.329027409977902e-06,
      "loss": 1.00536299,
      "num_input_tokens_seen": 3721320,
      "step": 176,
      "time_per_iteration": 2.8159937858581543
    },
    {
      "auxiliary_loss_clip": 0.0141033,
      "auxiliary_loss_mlp": 0.01130651,
      "balance_loss_clip": 1.12546706,
      "balance_loss_mlp": 1.07738805,
      "epoch": 0.010641815722230573,
      "flos": 19427601519360.0,
      "grad_norm": 2.8326118759658585,
      "language_loss": 0.76926064,
      "learning_rate": 3.3326753074614087e-06,
      "loss": 0.7946704,
      "num_input_tokens_seen": 3739385,
      "step": 177,
      "time_per_iteration": 5.7707555294036865
    },
    {
      "auxiliary_loss_clip": 0.01421858,
      "auxiliary_loss_mlp": 0.01104718,
      "balance_loss_clip": 1.12455702,
      "balance_loss_mlp": 1.05002475,
      "epoch": 0.010701938974898541,
      "flos": 18332577452160.0,
      "grad_norm": 2.6517911185675014,
      "language_loss": 0.76942402,
      "learning_rate": 3.3363026533007716e-06,
      "loss": 0.79468977,
      "num_input_tokens_seen": 3756360,
      "step": 178,
      "time_per_iteration": 4.337082386016846
    },
    {
      "auxiliary_loss_clip": 0.01430293,
      "auxiliary_loss_mlp": 0.01109414,
      "balance_loss_clip": 1.1303575,
      "balance_loss_mlp": 1.05252683,
      "epoch": 0.010762062227566512,
      "flos": 19203985399680.0,
      "grad_norm": 2.6843360372821925,
      "language_loss": 0.84022826,
      "learning_rate": 3.3399096777683303e-06,
      "loss": 0.86562538,
      "num_input_tokens_seen": 3773930,
      "step": 179,
      "time_per_iteration": 2.6826629638671875
    },
    {
      "auxiliary_loss_clip": 0.01418094,
      "auxiliary_loss_mlp": 0.01108667,
      "balance_loss_clip": 1.12202275,
      "balance_loss_mlp": 1.05158973,
      "epoch": 0.01082218548023448,
      "flos": 31425427370880.0,
      "grad_norm": 2.0256655839140083,
      "language_loss": 0.83674574,
      "learning_rate": 3.3434966072878213e-06,
      "loss": 0.86201334,
      "num_input_tokens_seen": 3793630,
      "step": 180,
      "time_per_iteration": 2.7483785152435303
    },
    {
      "auxiliary_loss_clip": 0.01421326,
      "auxiliary_loss_mlp": 0.01120347,
      "balance_loss_clip": 1.12740374,
      "balance_loss_mlp": 1.0646286,
      "epoch": 0.01088230873290245,
      "flos": 25046436170880.0,
      "grad_norm": 3.253139118534122,
      "language_loss": 0.77958715,
      "learning_rate": 3.3470636645196674e-06,
      "loss": 0.80500388,
      "num_input_tokens_seen": 3813610,
      "step": 181,
      "time_per_iteration": 2.698941469192505
    },
    {
      "auxiliary_loss_clip": 0.01414948,
      "auxiliary_loss_mlp": 0.01130231,
      "balance_loss_clip": 1.12188053,
      "balance_loss_mlp": 1.07577634,
      "epoch": 0.01094243198557042,
      "flos": 22893411980160.0,
      "grad_norm": 2.56637338396407,
      "language_loss": 0.76438594,
      "learning_rate": 3.3506110684439156e-06,
      "loss": 0.78983772,
      "num_input_tokens_seen": 3831390,
      "step": 182,
      "time_per_iteration": 2.6951375007629395
    },
    {
      "auxiliary_loss_clip": 0.01412526,
      "auxiliary_loss_mlp": 0.01126665,
      "balance_loss_clip": 1.12167537,
      "balance_loss_mlp": 1.0702554,
      "epoch": 0.011002555238238388,
      "flos": 17165049782400.0,
      "grad_norm": 2.083158831639218,
      "language_loss": 0.87484097,
      "learning_rate": 3.3541390344409054e-06,
      "loss": 0.90023291,
      "num_input_tokens_seen": 3849705,
      "step": 183,
      "time_per_iteration": 2.733753204345703
    },
    {
      "auxiliary_loss_clip": 0.01415922,
      "auxiliary_loss_mlp": 0.01110585,
      "balance_loss_clip": 1.12529624,
      "balance_loss_mlp": 1.05922985,
      "epoch": 0.011062678490906358,
      "flos": 22310150935680.0,
      "grad_norm": 3.105080129831269,
      "language_loss": 0.86911464,
      "learning_rate": 3.357647774369736e-06,
      "loss": 0.89437973,
      "num_input_tokens_seen": 3869230,
      "step": 184,
      "time_per_iteration": 2.6783828735351562
    },
    {
      "auxiliary_loss_clip": 0.01410648,
      "auxiliary_loss_mlp": 0.01108321,
      "balance_loss_clip": 1.12499499,
      "balance_loss_mlp": 1.05203021,
      "epoch": 0.011122801743574327,
      "flos": 24388373053440.0,
      "grad_norm": 1.8650514063709744,
      "language_loss": 0.83885491,
      "learning_rate": 3.3611374966446085e-06,
      "loss": 0.86404455,
      "num_input_tokens_seen": 3889735,
      "step": 185,
      "time_per_iteration": 2.6863327026367188
    },
    {
      "auxiliary_loss_clip": 0.01419384,
      "auxiliary_loss_mlp": 0.01107812,
      "balance_loss_clip": 1.12355363,
      "balance_loss_mlp": 1.04999495,
      "epoch": 0.011182924996242297,
      "flos": 18150258994560.0,
      "grad_norm": 2.8933407749520743,
      "language_loss": 0.71027243,
      "learning_rate": 3.3646084063091142e-06,
      "loss": 0.73554444,
      "num_input_tokens_seen": 3908855,
      "step": 186,
      "time_per_iteration": 2.819805383682251
    },
    {
      "auxiliary_loss_clip": 0.01415699,
      "auxiliary_loss_mlp": 0.01108312,
      "balance_loss_clip": 1.12262082,
      "balance_loss_mlp": 1.05574071,
      "epoch": 0.011243048248910266,
      "flos": 15486800584320.0,
      "grad_norm": 2.4244794785226733,
      "language_loss": 1.01999915,
      "learning_rate": 3.3680607051085194e-06,
      "loss": 1.04523933,
      "num_input_tokens_seen": 3923865,
      "step": 187,
      "time_per_iteration": 2.65875506401062
    },
    {
      "auxiliary_loss_clip": 0.01404987,
      "auxiliary_loss_mlp": 0.01107995,
      "balance_loss_clip": 1.12269068,
      "balance_loss_mlp": 1.05253887,
      "epoch": 0.011303171501578235,
      "flos": 40916868986880.0,
      "grad_norm": 2.0089158406542524,
      "language_loss": 0.74998611,
      "learning_rate": 3.371494591560139e-06,
      "loss": 0.77511597,
      "num_input_tokens_seen": 3946870,
      "step": 188,
      "time_per_iteration": 2.8631174564361572
    },
    {
      "auxiliary_loss_clip": 0.01298557,
      "auxiliary_loss_mlp": 0.01067058,
      "balance_loss_clip": 1.14124644,
      "balance_loss_mlp": 1.04474187,
      "epoch": 0.011363294754246205,
      "flos": 66302697790080.0,
      "grad_norm": 0.7620731385906954,
      "language_loss": 0.56192517,
      "learning_rate": 3.3749102610218297e-06,
      "loss": 0.5855813,
      "num_input_tokens_seen": 4010005,
      "step": 189,
      "time_per_iteration": 3.2704074382781982
    },
    {
      "auxiliary_loss_clip": 0.01402206,
      "auxiliary_loss_mlp": 0.011217,
      "balance_loss_clip": 1.11730003,
      "balance_loss_mlp": 1.06662548,
      "epoch": 0.011423418006914174,
      "flos": 24900279730560.0,
      "grad_norm": 2.640219984380571,
      "language_loss": 0.95085573,
      "learning_rate": 3.3783079057586833e-06,
      "loss": 0.97609472,
      "num_input_tokens_seen": 4029035,
      "step": 190,
      "time_per_iteration": 2.6898255348205566
    },
    {
      "auxiliary_loss_clip": 0.01405088,
      "auxiliary_loss_mlp": 0.01103893,
      "balance_loss_clip": 1.11979234,
      "balance_loss_mlp": 1.05167961,
      "epoch": 0.011483541259582144,
      "flos": 19791879298560.0,
      "grad_norm": 4.133813113517846,
      "language_loss": 0.8463847,
      "learning_rate": 3.3816877150079665e-06,
      "loss": 0.8714745,
      "num_input_tokens_seen": 4046995,
      "step": 191,
      "time_per_iteration": 2.71589994430542
    },
    {
      "auxiliary_loss_clip": 0.01403196,
      "auxiliary_loss_mlp": 0.01118385,
      "balance_loss_clip": 1.11570346,
      "balance_loss_mlp": 1.06624269,
      "epoch": 0.011543664512250112,
      "flos": 26176939896960.0,
      "grad_norm": 2.0065119945705887,
      "language_loss": 0.91894913,
      "learning_rate": 3.385049875042367e-06,
      "loss": 0.94416493,
      "num_input_tokens_seen": 4065865,
      "step": 192,
      "time_per_iteration": 2.775974988937378
    },
    {
      "auxiliary_loss_clip": 0.01398496,
      "auxiliary_loss_mlp": 0.01118924,
      "balance_loss_clip": 1.11665678,
      "balance_loss_mlp": 1.06117916,
      "epoch": 0.011603787764918083,
      "flos": 23768985905280.0,
      "grad_norm": 2.10033302347605,
      "language_loss": 0.86923265,
      "learning_rate": 3.3883945692315938e-06,
      "loss": 0.89440691,
      "num_input_tokens_seen": 4085305,
      "step": 193,
      "time_per_iteration": 2.792947292327881
    },
    {
      "auxiliary_loss_clip": 0.01402535,
      "auxiliary_loss_mlp": 0.01102276,
      "balance_loss_clip": 1.11514282,
      "balance_loss_mlp": 1.05061066,
      "epoch": 0.011663911017586051,
      "flos": 25954688494080.0,
      "grad_norm": 2.2253165290939076,
      "language_loss": 0.92296255,
      "learning_rate": 3.3917219781023906e-06,
      "loss": 0.94801068,
      "num_input_tokens_seen": 4105185,
      "step": 194,
      "time_per_iteration": 2.6886558532714844
    },
    {
      "auxiliary_loss_clip": 0.01407209,
      "auxiliary_loss_mlp": 0.01108641,
      "balance_loss_clip": 1.11930478,
      "balance_loss_mlp": 1.05630851,
      "epoch": 0.01172403427025402,
      "flos": 17895149625600.0,
      "grad_norm": 2.4241235245311503,
      "language_loss": 0.89768875,
      "learning_rate": 3.3950322793970014e-06,
      "loss": 0.92284721,
      "num_input_tokens_seen": 4123160,
      "step": 195,
      "time_per_iteration": 2.654517889022827
    },
    {
      "auxiliary_loss_clip": 0.01400339,
      "auxiliary_loss_mlp": 0.01114485,
      "balance_loss_clip": 1.11779022,
      "balance_loss_mlp": 1.05981565,
      "epoch": 0.01178415752292199,
      "flos": 17894539094400.0,
      "grad_norm": 3.1130999341447385,
      "language_loss": 0.86019921,
      "learning_rate": 3.3983256481301445e-06,
      "loss": 0.88534749,
      "num_input_tokens_seen": 4140425,
      "step": 196,
      "time_per_iteration": 2.643598794937134
    },
    {
      "auxiliary_loss_clip": 0.01398067,
      "auxiliary_loss_mlp": 0.01107082,
      "balance_loss_clip": 1.11464977,
      "balance_loss_mlp": 1.05308056,
      "epoch": 0.011844280775589959,
      "flos": 22893555634560.0,
      "grad_norm": 3.666533247373141,
      "language_loss": 0.93052697,
      "learning_rate": 3.4016022566445335e-06,
      "loss": 0.95557845,
      "num_input_tokens_seen": 4159555,
      "step": 197,
      "time_per_iteration": 2.7120354175567627
    },
    {
      "auxiliary_loss_clip": 0.01396424,
      "auxiliary_loss_mlp": 0.01112388,
      "balance_loss_clip": 1.11625624,
      "balance_loss_mlp": 1.05943501,
      "epoch": 0.01190440402825793,
      "flos": 26980333441920.0,
      "grad_norm": 1.9614954763997827,
      "language_loss": 0.79043806,
      "learning_rate": 3.4048622746649966e-06,
      "loss": 0.81552619,
      "num_input_tokens_seen": 4180480,
      "step": 198,
      "time_per_iteration": 2.774059772491455
    },
    {
      "auxiliary_loss_clip": 0.0139305,
      "auxiliary_loss_mlp": 0.01120527,
      "balance_loss_clip": 1.11708748,
      "balance_loss_mlp": 1.06821764,
      "epoch": 0.011964527280925898,
      "flos": 20521584092160.0,
      "grad_norm": 1.8823459083646328,
      "language_loss": 0.88239717,
      "learning_rate": 3.4081058693512278e-06,
      "loss": 0.90753293,
      "num_input_tokens_seen": 4198835,
      "step": 199,
      "time_per_iteration": 2.6808881759643555
    },
    {
      "auxiliary_loss_clip": 0.01403709,
      "auxiliary_loss_mlp": 0.0112899,
      "balance_loss_clip": 1.11951399,
      "balance_loss_mlp": 1.07200766,
      "epoch": 0.012024650533593867,
      "flos": 27745984771200.0,
      "grad_norm": 2.0663906916258497,
      "language_loss": 0.81151628,
      "learning_rate": 3.411333205349222e-06,
      "loss": 0.83684325,
      "num_input_tokens_seen": 4219335,
      "step": 200,
      "time_per_iteration": 2.625380516052246
    },
    {
      "auxiliary_loss_clip": 0.0140201,
      "auxiliary_loss_mlp": 0.01104413,
      "balance_loss_clip": 1.11633158,
      "balance_loss_mlp": 1.05048287,
      "epoch": 0.012084773786261837,
      "flos": 10452017076480.0,
      "grad_norm": 2.253120238884594,
      "language_loss": 0.87696433,
      "learning_rate": 3.4145444448414217e-06,
      "loss": 0.90202856,
      "num_input_tokens_seen": 4236940,
      "step": 201,
      "time_per_iteration": 2.6062326431274414
    },
    {
      "auxiliary_loss_clip": 0.01399494,
      "auxiliary_loss_mlp": 0.01115643,
      "balance_loss_clip": 1.11764228,
      "balance_loss_mlp": 1.0614028,
      "epoch": 0.012144897038929806,
      "flos": 23105751229440.0,
      "grad_norm": 2.088192664231089,
      "language_loss": 0.84052485,
      "learning_rate": 3.4177397475956223e-06,
      "loss": 0.86567622,
      "num_input_tokens_seen": 4256755,
      "step": 202,
      "time_per_iteration": 2.6981592178344727
    },
    {
      "auxiliary_loss_clip": 0.01388741,
      "auxiliary_loss_mlp": 0.0111019,
      "balance_loss_clip": 1.11006808,
      "balance_loss_mlp": 1.05771446,
      "epoch": 0.012205020291597776,
      "flos": 21033203460480.0,
      "grad_norm": 1.7861279575653157,
      "language_loss": 0.89964712,
      "learning_rate": 3.4209192710126685e-06,
      "loss": 0.92463642,
      "num_input_tokens_seen": 4276505,
      "step": 203,
      "time_per_iteration": 2.668757438659668
    },
    {
      "auxiliary_loss_clip": 0.01276289,
      "auxiliary_loss_mlp": 0.01095021,
      "balance_loss_clip": 1.12578154,
      "balance_loss_mlp": 1.07470798,
      "epoch": 0.012265143544265745,
      "flos": 68447785075200.0,
      "grad_norm": 1.0265297625980543,
      "language_loss": 0.61255801,
      "learning_rate": 3.4240831701729837e-06,
      "loss": 0.63627112,
      "num_input_tokens_seen": 4330965,
      "step": 204,
      "time_per_iteration": 3.161599636077881
    },
    {
      "auxiliary_loss_clip": 0.01396271,
      "auxiliary_loss_mlp": 0.01111806,
      "balance_loss_clip": 1.11291122,
      "balance_loss_mlp": 1.05930579,
      "epoch": 0.012325266796933715,
      "flos": 17019252478080.0,
      "grad_norm": 2.3248674300118184,
      "language_loss": 0.91324663,
      "learning_rate": 3.4272315978819516e-06,
      "loss": 0.93832743,
      "num_input_tokens_seen": 4348200,
      "step": 205,
      "time_per_iteration": 2.6764047145843506
    },
    {
      "auxiliary_loss_clip": 0.01404558,
      "auxiliary_loss_mlp": 0.0112167,
      "balance_loss_clip": 1.11773109,
      "balance_loss_mlp": 1.06773925,
      "epoch": 0.012385390049601683,
      "flos": 20190056538240.0,
      "grad_norm": 2.1088315130515207,
      "language_loss": 0.89305568,
      "learning_rate": 3.4303647047142043e-06,
      "loss": 0.91831797,
      "num_input_tokens_seen": 4365460,
      "step": 206,
      "time_per_iteration": 2.7157227993011475
    },
    {
      "auxiliary_loss_clip": 0.0139534,
      "auxiliary_loss_mlp": 0.01100957,
      "balance_loss_clip": 1.11176991,
      "balance_loss_mlp": 1.04888678,
      "epoch": 0.012445513302269652,
      "flos": 16253134272000.0,
      "grad_norm": 2.399816031687551,
      "language_loss": 0.95542914,
      "learning_rate": 3.43348263905683e-06,
      "loss": 0.9803921,
      "num_input_tokens_seen": 4383650,
      "step": 207,
      "time_per_iteration": 2.611348867416382
    },
    {
      "auxiliary_loss_clip": 0.01393005,
      "auxiliary_loss_mlp": 0.01117764,
      "balance_loss_clip": 1.11658561,
      "balance_loss_mlp": 1.06497812,
      "epoch": 0.012505636554937622,
      "flos": 23769380954880.0,
      "grad_norm": 1.8144323603981871,
      "language_loss": 0.75985783,
      "learning_rate": 3.436585547151547e-06,
      "loss": 0.78496552,
      "num_input_tokens_seen": 4403765,
      "step": 208,
      "time_per_iteration": 2.7184154987335205
    },
    {
      "auxiliary_loss_clip": 0.0138146,
      "auxiliary_loss_mlp": 0.01108623,
      "balance_loss_clip": 1.11071992,
      "balance_loss_mlp": 1.05576587,
      "epoch": 0.012565759807605591,
      "flos": 30591546157440.0,
      "grad_norm": 2.2326965650696855,
      "language_loss": 0.98386943,
      "learning_rate": 3.4396735731358586e-06,
      "loss": 1.00877023,
      "num_input_tokens_seen": 4421935,
      "step": 209,
      "time_per_iteration": 2.7354249954223633
    },
    {
      "auxiliary_loss_clip": 0.01387012,
      "auxiliary_loss_mlp": 0.0111836,
      "balance_loss_clip": 1.11136842,
      "balance_loss_mlp": 1.06490695,
      "epoch": 0.012625883060273561,
      "flos": 40113511355520.0,
      "grad_norm": 9.084733304650118,
      "language_loss": 0.85514843,
      "learning_rate": 3.4427468590832302e-06,
      "loss": 0.88020217,
      "num_input_tokens_seen": 4441470,
      "step": 210,
      "time_per_iteration": 2.888749122619629
    },
    {
      "auxiliary_loss_clip": 0.01384384,
      "auxiliary_loss_mlp": 0.01121559,
      "balance_loss_clip": 1.11018038,
      "balance_loss_mlp": 1.07115781,
      "epoch": 0.01268600631294153,
      "flos": 27089178629760.0,
      "grad_norm": 3.431917100192063,
      "language_loss": 0.97194636,
      "learning_rate": 3.445805545042314e-06,
      "loss": 0.99700582,
      "num_input_tokens_seen": 4459950,
      "step": 211,
      "time_per_iteration": 2.7465193271636963
    },
    {
      "auxiliary_loss_clip": 0.01393556,
      "auxiliary_loss_mlp": 0.01123542,
      "balance_loss_clip": 1.11511767,
      "balance_loss_mlp": 1.06999326,
      "epoch": 0.012746129565609499,
      "flos": 16982767238400.0,
      "grad_norm": 2.3992368053115163,
      "language_loss": 0.9508543,
      "learning_rate": 3.448849769075239e-06,
      "loss": 0.97602528,
      "num_input_tokens_seen": 4478390,
      "step": 212,
      "time_per_iteration": 2.6340651512145996
    },
    {
      "auxiliary_loss_clip": 0.01381697,
      "auxiliary_loss_mlp": 0.01116386,
      "balance_loss_clip": 1.112149,
      "balance_loss_mlp": 1.06381512,
      "epoch": 0.012806252818277469,
      "flos": 46533476995200.0,
      "grad_norm": 1.701444843398511,
      "language_loss": 0.76078421,
      "learning_rate": 3.4518796672950093e-06,
      "loss": 0.78576505,
      "num_input_tokens_seen": 4501665,
      "step": 213,
      "time_per_iteration": 2.9250640869140625
    },
    {
      "auxiliary_loss_clip": 0.01385821,
      "auxiliary_loss_mlp": 0.01111776,
      "balance_loss_clip": 1.11002433,
      "balance_loss_mlp": 1.06056333,
      "epoch": 0.012866376070945438,
      "flos": 14388616120320.0,
      "grad_norm": 3.5300370267625922,
      "language_loss": 0.86698866,
      "learning_rate": 3.4548953739020187e-06,
      "loss": 0.89196461,
      "num_input_tokens_seen": 4519055,
      "step": 214,
      "time_per_iteration": 2.645289659500122
    },
    {
      "auxiliary_loss_clip": 0.01383455,
      "auxiliary_loss_mlp": 0.01128262,
      "balance_loss_clip": 1.1159339,
      "balance_loss_mlp": 1.07359219,
      "epoch": 0.012926499323613408,
      "flos": 26140813793280.0,
      "grad_norm": 2.14433888305053,
      "language_loss": 0.77582061,
      "learning_rate": 3.4578970212197196e-06,
      "loss": 0.80093777,
      "num_input_tokens_seen": 4540870,
      "step": 215,
      "time_per_iteration": 2.7315175533294678
    },
    {
      "auxiliary_loss_clip": 0.01391951,
      "auxiliary_loss_mlp": 0.01115104,
      "balance_loss_clip": 1.11440635,
      "balance_loss_mlp": 1.0638206,
      "epoch": 0.012986622576281377,
      "flos": 30117202128000.0,
      "grad_norm": 2.2964706747038233,
      "language_loss": 0.90423942,
      "learning_rate": 3.460884739729461e-06,
      "loss": 0.92930996,
      "num_input_tokens_seen": 4560395,
      "step": 216,
      "time_per_iteration": 2.724698781967163
    },
    {
      "auxiliary_loss_clip": 0.01384729,
      "auxiliary_loss_mlp": 0.01113374,
      "balance_loss_clip": 1.10847259,
      "balance_loss_mlp": 1.06096959,
      "epoch": 0.013046745828949347,
      "flos": 13954025468160.0,
      "grad_norm": 3.60062834696173,
      "language_loss": 0.93473232,
      "learning_rate": 3.463858658104523e-06,
      "loss": 0.95971346,
      "num_input_tokens_seen": 4575785,
      "step": 217,
      "time_per_iteration": 5.762276649475098
    },
    {
      "auxiliary_loss_clip": 0.01377712,
      "auxiliary_loss_mlp": 0.0110874,
      "balance_loss_clip": 1.10726643,
      "balance_loss_mlp": 1.05433273,
      "epoch": 0.013106869081617315,
      "flos": 17347835116800.0,
      "grad_norm": 1.943339896357513,
      "language_loss": 0.93811166,
      "learning_rate": 3.4668189032433696e-06,
      "loss": 0.96297616,
      "num_input_tokens_seen": 4594985,
      "step": 218,
      "time_per_iteration": 5.832701206207275
    },
    {
      "auxiliary_loss_clip": 0.01372884,
      "auxiliary_loss_mlp": 0.01106717,
      "balance_loss_clip": 1.10647273,
      "balance_loss_mlp": 1.05552888,
      "epoch": 0.013166992334285284,
      "flos": 25884914325120.0,
      "grad_norm": 2.252873600345955,
      "language_loss": 0.86196327,
      "learning_rate": 3.46976560030214e-06,
      "loss": 0.88675928,
      "num_input_tokens_seen": 4616125,
      "step": 219,
      "time_per_iteration": 2.794581651687622
    },
    {
      "auxiliary_loss_clip": 0.0137885,
      "auxiliary_loss_mlp": 0.01102953,
      "balance_loss_clip": 1.10957599,
      "balance_loss_mlp": 1.05188394,
      "epoch": 0.013227115586953254,
      "flos": 31175956437120.0,
      "grad_norm": 1.897987121161891,
      "language_loss": 0.8748548,
      "learning_rate": 3.4726988727263976e-06,
      "loss": 0.89967287,
      "num_input_tokens_seen": 4637795,
      "step": 220,
      "time_per_iteration": 2.799927234649658
    },
    {
      "auxiliary_loss_clip": 0.01370688,
      "auxiliary_loss_mlp": 0.01115596,
      "balance_loss_clip": 1.10440111,
      "balance_loss_mlp": 1.0679127,
      "epoch": 0.013287238839621223,
      "flos": 20409470766720.0,
      "grad_norm": 3.2557072980071795,
      "language_loss": 0.86437249,
      "learning_rate": 3.475618842282164e-06,
      "loss": 0.88923532,
      "num_input_tokens_seen": 4656835,
      "step": 221,
      "time_per_iteration": 2.7040672302246094
    },
    {
      "auxiliary_loss_clip": 0.01376134,
      "auxiliary_loss_mlp": 0.01116397,
      "balance_loss_clip": 1.10384834,
      "balance_loss_mlp": 1.0637064,
      "epoch": 0.013347362092289193,
      "flos": 14137134024960.0,
      "grad_norm": 2.585706849100757,
      "language_loss": 0.92369294,
      "learning_rate": 3.4785256290862486e-06,
      "loss": 0.94861829,
      "num_input_tokens_seen": 4673015,
      "step": 222,
      "time_per_iteration": 2.6648194789886475
    },
    {
      "auxiliary_loss_clip": 0.01373283,
      "auxiliary_loss_mlp": 0.01106423,
      "balance_loss_clip": 1.10636806,
      "balance_loss_mlp": 1.05156267,
      "epoch": 0.013407485344957162,
      "flos": 21797705554560.0,
      "grad_norm": 7.739608779999776,
      "language_loss": 0.95708215,
      "learning_rate": 3.481419351635897e-06,
      "loss": 0.98187923,
      "num_input_tokens_seen": 4692355,
      "step": 223,
      "time_per_iteration": 2.7261807918548584
    },
    {
      "auxiliary_loss_clip": 0.01374555,
      "auxiliary_loss_mlp": 0.0110963,
      "balance_loss_clip": 1.10768425,
      "balance_loss_mlp": 1.05870414,
      "epoch": 0.013467608597625132,
      "flos": 18621622195200.0,
      "grad_norm": 2.673591615227502,
      "language_loss": 0.88031876,
      "learning_rate": 3.484300126837776e-06,
      "loss": 0.90516055,
      "num_input_tokens_seen": 4710080,
      "step": 224,
      "time_per_iteration": 2.601686477661133
    },
    {
      "auxiliary_loss_clip": 0.01374533,
      "auxiliary_loss_mlp": 0.01103,
      "balance_loss_clip": 1.10679817,
      "balance_loss_mlp": 1.04804444,
      "epoch": 0.013527731850293101,
      "flos": 18552314903040.0,
      "grad_norm": 3.0722216996453535,
      "language_loss": 0.89625597,
      "learning_rate": 3.487168070036317e-06,
      "loss": 0.9210313,
      "num_input_tokens_seen": 4728980,
      "step": 225,
      "time_per_iteration": 2.6677513122558594
    },
    {
      "auxiliary_loss_clip": 0.01369955,
      "auxiliary_loss_mlp": 0.0112021,
      "balance_loss_clip": 1.10561275,
      "balance_loss_mlp": 1.06675696,
      "epoch": 0.01358785510296107,
      "flos": 19165381257600.0,
      "grad_norm": 1.9576206039109396,
      "language_loss": 0.98980033,
      "learning_rate": 3.4900232950414224e-06,
      "loss": 1.01470196,
      "num_input_tokens_seen": 4747020,
      "step": 226,
      "time_per_iteration": 2.8320930004119873
    },
    {
      "auxiliary_loss_clip": 0.01375268,
      "auxiliary_loss_mlp": 0.01110039,
      "balance_loss_clip": 1.10837173,
      "balance_loss_mlp": 1.05572701,
      "epoch": 0.01364797835562904,
      "flos": 23329941966720.0,
      "grad_norm": 2.3303410550109245,
      "language_loss": 0.90965348,
      "learning_rate": 3.4928659141555727e-06,
      "loss": 0.93450654,
      "num_input_tokens_seen": 4765000,
      "step": 227,
      "time_per_iteration": 2.648606061935425
    },
    {
      "auxiliary_loss_clip": 0.01255161,
      "auxiliary_loss_mlp": 0.01079249,
      "balance_loss_clip": 1.11229861,
      "balance_loss_mlp": 1.06017554,
      "epoch": 0.013708101608297009,
      "flos": 70993746097920.0,
      "grad_norm": 0.9472069433514878,
      "language_loss": 0.57650995,
      "learning_rate": 3.4956960382003234e-06,
      "loss": 0.59985405,
      "num_input_tokens_seen": 4833210,
      "step": 228,
      "time_per_iteration": 3.246328592300415
    },
    {
      "auxiliary_loss_clip": 0.01366835,
      "auxiliary_loss_mlp": 0.01117377,
      "balance_loss_clip": 1.10507822,
      "balance_loss_mlp": 1.06711841,
      "epoch": 0.013768224860964979,
      "flos": 16325170997760.0,
      "grad_norm": 2.957038430634678,
      "language_loss": 0.87773621,
      "learning_rate": 3.4985137765422354e-06,
      "loss": 0.90257835,
      "num_input_tokens_seen": 4850120,
      "step": 229,
      "time_per_iteration": 2.6319024562835693
    },
    {
      "auxiliary_loss_clip": 0.01375278,
      "auxiliary_loss_mlp": 0.01098609,
      "balance_loss_clip": 1.10567176,
      "balance_loss_mlp": 1.04873204,
      "epoch": 0.013828348113632948,
      "flos": 20193037367040.0,
      "grad_norm": 4.72663824849547,
      "language_loss": 0.83937395,
      "learning_rate": 3.501319237118231e-06,
      "loss": 0.86411285,
      "num_input_tokens_seen": 4866215,
      "step": 230,
      "time_per_iteration": 2.7026398181915283
    },
    {
      "auxiliary_loss_clip": 0.01373544,
      "auxiliary_loss_mlp": 0.01113683,
      "balance_loss_clip": 1.10701275,
      "balance_loss_mlp": 1.06361556,
      "epoch": 0.013888471366300916,
      "flos": 20741070147840.0,
      "grad_norm": 2.2562202151287867,
      "language_loss": 0.904212,
      "learning_rate": 3.5041125264604056e-06,
      "loss": 0.9290843,
      "num_input_tokens_seen": 4885630,
      "step": 231,
      "time_per_iteration": 2.6424474716186523
    },
    {
      "auxiliary_loss_clip": 0.01377759,
      "auxiliary_loss_mlp": 0.01110232,
      "balance_loss_clip": 1.11118639,
      "balance_loss_mlp": 1.06030726,
      "epoch": 0.013948594618968886,
      "flos": 22090628966400.0,
      "grad_norm": 2.0229562700819215,
      "language_loss": 0.83624899,
      "learning_rate": 3.5068937497203002e-06,
      "loss": 0.86112887,
      "num_input_tokens_seen": 4905570,
      "step": 232,
      "time_per_iteration": 2.621704339981079
    },
    {
      "auxiliary_loss_clip": 0.01377798,
      "auxiliary_loss_mlp": 0.01094369,
      "balance_loss_clip": 1.10229027,
      "balance_loss_mlp": 1.04253721,
      "epoch": 0.014008717871636855,
      "flos": 19063108258560.0,
      "grad_norm": 5.516695444379509,
      "language_loss": 0.74727643,
      "learning_rate": 3.509663010692652e-06,
      "loss": 0.77199805,
      "num_input_tokens_seen": 4923535,
      "step": 233,
      "time_per_iteration": 2.659188747406006
    },
    {
      "auxiliary_loss_clip": 0.01382744,
      "auxiliary_loss_mlp": 0.01125121,
      "balance_loss_clip": 1.1099937,
      "balance_loss_mlp": 1.0723356,
      "epoch": 0.014068841124304825,
      "flos": 14530822064640.0,
      "grad_norm": 2.5763093382937483,
      "language_loss": 0.85633421,
      "learning_rate": 3.512420411838642e-06,
      "loss": 0.88141286,
      "num_input_tokens_seen": 4939200,
      "step": 234,
      "time_per_iteration": 2.610635757446289
    },
    {
      "auxiliary_loss_clip": 0.01374562,
      "auxiliary_loss_mlp": 0.01114672,
      "balance_loss_clip": 1.10890436,
      "balance_loss_mlp": 1.06467605,
      "epoch": 0.014128964376972794,
      "flos": 18077396256000.0,
      "grad_norm": 2.467487286445388,
      "language_loss": 0.89192498,
      "learning_rate": 3.515166054308634e-06,
      "loss": 0.91681731,
      "num_input_tokens_seen": 4956620,
      "step": 235,
      "time_per_iteration": 2.668769359588623
    },
    {
      "auxiliary_loss_clip": 0.01373018,
      "auxiliary_loss_mlp": 0.01131641,
      "balance_loss_clip": 1.11011076,
      "balance_loss_mlp": 1.08073914,
      "epoch": 0.014189087629640764,
      "flos": 25334331678720.0,
      "grad_norm": 2.143165146200321,
      "language_loss": 0.85535377,
      "learning_rate": 3.5179000379644498e-06,
      "loss": 0.88040036,
      "num_input_tokens_seen": 4975650,
      "step": 236,
      "time_per_iteration": 2.7570323944091797
    },
    {
      "auxiliary_loss_clip": 0.01369632,
      "auxiliary_loss_mlp": 0.01100269,
      "balance_loss_clip": 1.10296702,
      "balance_loss_mlp": 1.04905629,
      "epoch": 0.014249210882308733,
      "flos": 36139744713600.0,
      "grad_norm": 2.1351980688483136,
      "language_loss": 0.82550979,
      "learning_rate": 3.520622461401154e-06,
      "loss": 0.85020876,
      "num_input_tokens_seen": 4997415,
      "step": 237,
      "time_per_iteration": 2.811617374420166
    },
    {
      "auxiliary_loss_clip": 0.01369728,
      "auxiliary_loss_mlp": 0.01124352,
      "balance_loss_clip": 1.10659075,
      "balance_loss_mlp": 1.07085085,
      "epoch": 0.014309334134976702,
      "flos": 12932977461120.0,
      "grad_norm": 2.0241581748099313,
      "language_loss": 0.77096599,
      "learning_rate": 3.5233334219683935e-06,
      "loss": 0.79590684,
      "num_input_tokens_seen": 5013905,
      "step": 238,
      "time_per_iteration": 2.8044662475585938
    },
    {
      "auxiliary_loss_clip": 0.01367496,
      "auxiliary_loss_mlp": 0.01111406,
      "balance_loss_clip": 1.10897434,
      "balance_loss_mlp": 1.06343579,
      "epoch": 0.014369457387644672,
      "flos": 20777519473920.0,
      "grad_norm": 1.8300428555870456,
      "language_loss": 0.8707583,
      "learning_rate": 3.526033015791284e-06,
      "loss": 0.89554727,
      "num_input_tokens_seen": 5033645,
      "step": 239,
      "time_per_iteration": 2.681452751159668
    },
    {
      "auxiliary_loss_clip": 0.01353036,
      "auxiliary_loss_mlp": 0.01103184,
      "balance_loss_clip": 1.10036874,
      "balance_loss_mlp": 1.05516672,
      "epoch": 0.01442958064031264,
      "flos": 25848536826240.0,
      "grad_norm": 2.109315431148974,
      "language_loss": 0.93055749,
      "learning_rate": 3.528721337790862e-06,
      "loss": 0.95511973,
      "num_input_tokens_seen": 5052875,
      "step": 240,
      "time_per_iteration": 2.679826021194458
    },
    {
      "auxiliary_loss_clip": 0.01360794,
      "auxiliary_loss_mlp": 0.01103084,
      "balance_loss_clip": 1.10475957,
      "balance_loss_mlp": 1.05611515,
      "epoch": 0.014489703892980611,
      "flos": 28219718269440.0,
      "grad_norm": 3.7136133710916575,
      "language_loss": 0.8482846,
      "learning_rate": 3.531398481704111e-06,
      "loss": 0.87292337,
      "num_input_tokens_seen": 5075005,
      "step": 241,
      "time_per_iteration": 2.679126262664795
    },
    {
      "auxiliary_loss_clip": 0.01359518,
      "auxiliary_loss_mlp": 0.01119602,
      "balance_loss_clip": 1.11010456,
      "balance_loss_mlp": 1.06931913,
      "epoch": 0.01454982714564858,
      "flos": 22490925108480.0,
      "grad_norm": 1.8502491938168453,
      "language_loss": 0.88590866,
      "learning_rate": 3.534064540103573e-06,
      "loss": 0.9106999,
      "num_input_tokens_seen": 5091875,
      "step": 242,
      "time_per_iteration": 2.7366583347320557
    },
    {
      "auxiliary_loss_clip": 0.01359534,
      "auxiliary_loss_mlp": 0.01104713,
      "balance_loss_clip": 1.10356677,
      "balance_loss_mlp": 1.05342889,
      "epoch": 0.014609950398316548,
      "flos": 21653201139840.0,
      "grad_norm": 2.261458758817042,
      "language_loss": 0.86688942,
      "learning_rate": 3.536719604416555e-06,
      "loss": 0.89153194,
      "num_input_tokens_seen": 5111290,
      "step": 243,
      "time_per_iteration": 2.764378070831299
    },
    {
      "auxiliary_loss_clip": 0.01364897,
      "auxiliary_loss_mlp": 0.01106776,
      "balance_loss_clip": 1.10636568,
      "balance_loss_mlp": 1.05656552,
      "epoch": 0.014670073650984519,
      "flos": 21869993675520.0,
      "grad_norm": 1.6964959858678799,
      "language_loss": 0.84256208,
      "learning_rate": 3.5393637649439464e-06,
      "loss": 0.86727887,
      "num_input_tokens_seen": 5132265,
      "step": 244,
      "time_per_iteration": 2.630441188812256
    },
    {
      "auxiliary_loss_clip": 0.01372266,
      "auxiliary_loss_mlp": 0.01115072,
      "balance_loss_clip": 1.10771632,
      "balance_loss_mlp": 1.06328762,
      "epoch": 0.014730196903652487,
      "flos": 23183713699200.0,
      "grad_norm": 8.49550264430495,
      "language_loss": 0.78613877,
      "learning_rate": 3.54199711087864e-06,
      "loss": 0.81101215,
      "num_input_tokens_seen": 5148575,
      "step": 245,
      "time_per_iteration": 2.6991443634033203
    },
    {
      "auxiliary_loss_clip": 0.01371598,
      "auxiliary_loss_mlp": 0.0110404,
      "balance_loss_clip": 1.10405719,
      "balance_loss_mlp": 1.05008554,
      "epoch": 0.014790320156320457,
      "flos": 23222605150080.0,
      "grad_norm": 2.2582939339926305,
      "language_loss": 0.84165329,
      "learning_rate": 3.5446197303235913e-06,
      "loss": 0.86640966,
      "num_input_tokens_seen": 5170415,
      "step": 246,
      "time_per_iteration": 2.726743221282959
    },
    {
      "auxiliary_loss_clip": 0.01365538,
      "auxiliary_loss_mlp": 0.01101456,
      "balance_loss_clip": 1.10242295,
      "balance_loss_mlp": 1.05062532,
      "epoch": 0.014850443408988426,
      "flos": 15815490963840.0,
      "grad_norm": 1.9870849133800452,
      "language_loss": 0.89958012,
      "learning_rate": 3.5472317103095034e-06,
      "loss": 0.92425001,
      "num_input_tokens_seen": 5188565,
      "step": 247,
      "time_per_iteration": 2.5998406410217285
    },
    {
      "auxiliary_loss_clip": 0.01364581,
      "auxiliary_loss_mlp": 0.01098108,
      "balance_loss_clip": 1.09896278,
      "balance_loss_mlp": 1.0489223,
      "epoch": 0.014910566661656396,
      "flos": 22781657790720.0,
      "grad_norm": 2.0527635487774343,
      "language_loss": 0.783005,
      "learning_rate": 3.549833136812155e-06,
      "loss": 0.80763197,
      "num_input_tokens_seen": 5207810,
      "step": 248,
      "time_per_iteration": 2.689784049987793
    },
    {
      "auxiliary_loss_clip": 0.01365896,
      "auxiliary_loss_mlp": 0.01110511,
      "balance_loss_clip": 1.10732806,
      "balance_loss_mlp": 1.06044269,
      "epoch": 0.014970689914324365,
      "flos": 26865023806080.0,
      "grad_norm": 1.9405946352322343,
      "language_loss": 0.83855766,
      "learning_rate": 3.552424094769381e-06,
      "loss": 0.86332172,
      "num_input_tokens_seen": 5226210,
      "step": 249,
      "time_per_iteration": 2.8210339546203613
    },
    {
      "auxiliary_loss_clip": 0.01358179,
      "auxiliary_loss_mlp": 0.01106801,
      "balance_loss_clip": 1.10089588,
      "balance_loss_mlp": 1.05802023,
      "epoch": 0.015030813166992334,
      "flos": 13985662371840.0,
      "grad_norm": 2.0689026358419786,
      "language_loss": 0.93631709,
      "learning_rate": 3.5550046680977174e-06,
      "loss": 0.96096689,
      "num_input_tokens_seen": 5241660,
      "step": 250,
      "time_per_iteration": 2.7074570655822754
    },
    {
      "auxiliary_loss_clip": 0.01368183,
      "auxiliary_loss_mlp": 0.01115393,
      "balance_loss_clip": 1.1065619,
      "balance_loss_mlp": 1.06415713,
      "epoch": 0.015090936419660304,
      "flos": 24717817618560.0,
      "grad_norm": 2.6509740932573127,
      "language_loss": 0.9678722,
      "learning_rate": 3.5575749397087034e-06,
      "loss": 0.99270797,
      "num_input_tokens_seen": 5261090,
      "step": 251,
      "time_per_iteration": 2.6740176677703857
    },
    {
      "auxiliary_loss_clip": 0.01361249,
      "auxiliary_loss_mlp": 0.01108489,
      "balance_loss_clip": 1.10063529,
      "balance_loss_mlp": 1.0597558,
      "epoch": 0.015151059672328273,
      "flos": 25738793798400.0,
      "grad_norm": 1.996044018630987,
      "language_loss": 0.84516245,
      "learning_rate": 3.5601349915248707e-06,
      "loss": 0.86985981,
      "num_input_tokens_seen": 5279175,
      "step": 252,
      "time_per_iteration": 2.7198123931884766
    },
    {
      "auxiliary_loss_clip": 0.01356789,
      "auxiliary_loss_mlp": 0.0111346,
      "balance_loss_clip": 1.1023767,
      "balance_loss_mlp": 1.06346345,
      "epoch": 0.015211182924996243,
      "flos": 21871214737920.0,
      "grad_norm": 2.3132428526475275,
      "language_loss": 0.98516917,
      "learning_rate": 3.5626849044954064e-06,
      "loss": 1.0098716,
      "num_input_tokens_seen": 5296975,
      "step": 253,
      "time_per_iteration": 2.6751561164855957
    },
    {
      "auxiliary_loss_clip": 0.01244193,
      "auxiliary_loss_mlp": 0.01100072,
      "balance_loss_clip": 1.1058414,
      "balance_loss_mlp": 1.08338308,
      "epoch": 0.015271306177664212,
      "flos": 66895080888960.0,
      "grad_norm": 0.8719135194962525,
      "language_loss": 0.55628473,
      "learning_rate": 3.5652247586115167e-06,
      "loss": 0.57972741,
      "num_input_tokens_seen": 5358375,
      "step": 254,
      "time_per_iteration": 3.2305996417999268
    },
    {
      "auxiliary_loss_clip": 0.0136146,
      "auxiliary_loss_mlp": 0.01119692,
      "balance_loss_clip": 1.0985806,
      "balance_loss_mlp": 1.06952846,
      "epoch": 0.01533142943033218,
      "flos": 26834069260800.0,
      "grad_norm": 2.113472843461701,
      "language_loss": 0.90234184,
      "learning_rate": 3.567754632921479e-06,
      "loss": 0.92715329,
      "num_input_tokens_seen": 5377255,
      "step": 255,
      "time_per_iteration": 2.7138473987579346
    },
    {
      "auxiliary_loss_clip": 0.01357311,
      "auxiliary_loss_mlp": 0.01137867,
      "balance_loss_clip": 1.1001389,
      "balance_loss_mlp": 1.08803785,
      "epoch": 0.01539155268300015,
      "flos": 20813753318400.0,
      "grad_norm": 2.320838285045027,
      "language_loss": 0.85392761,
      "learning_rate": 3.5702746055454075e-06,
      "loss": 0.87887937,
      "num_input_tokens_seen": 5395320,
      "step": 256,
      "time_per_iteration": 2.7135775089263916
    },
    {
      "auxiliary_loss_clip": 0.01363873,
      "auxiliary_loss_mlp": 0.0112257,
      "balance_loss_clip": 1.10053098,
      "balance_loss_mlp": 1.07281172,
      "epoch": 0.01545167593566812,
      "flos": 15961862885760.0,
      "grad_norm": 4.480294478847577,
      "language_loss": 0.71472508,
      "learning_rate": 3.5727847536897254e-06,
      "loss": 0.73958945,
      "num_input_tokens_seen": 5411970,
      "step": 257,
      "time_per_iteration": 6.340675592422485
    },
    {
      "auxiliary_loss_clip": 0.01355912,
      "auxiliary_loss_mlp": 0.01112611,
      "balance_loss_clip": 1.10014856,
      "balance_loss_mlp": 1.06280565,
      "epoch": 0.01551179918833609,
      "flos": 22601745544320.0,
      "grad_norm": 2.0292888191897673,
      "language_loss": 0.94713151,
      "learning_rate": 3.5752851536613596e-06,
      "loss": 0.97181678,
      "num_input_tokens_seen": 5430245,
      "step": 258,
      "time_per_iteration": 5.674164772033691
    },
    {
      "auxiliary_loss_clip": 0.01356656,
      "auxiliary_loss_mlp": 0.01113313,
      "balance_loss_clip": 1.09867072,
      "balance_loss_mlp": 1.0645566,
      "epoch": 0.015571922441004058,
      "flos": 22816706486400.0,
      "grad_norm": 2.3215886633849236,
      "language_loss": 0.93037683,
      "learning_rate": 3.577775880881658e-06,
      "loss": 0.95507646,
      "num_input_tokens_seen": 5448905,
      "step": 259,
      "time_per_iteration": 2.6286497116088867
    },
    {
      "auxiliary_loss_clip": 0.01348977,
      "auxiliary_loss_mlp": 0.01102171,
      "balance_loss_clip": 1.10076857,
      "balance_loss_mlp": 1.05625176,
      "epoch": 0.015632045693672027,
      "flos": 18947439486720.0,
      "grad_norm": 1.9575053933526474,
      "language_loss": 0.97368109,
      "learning_rate": 3.5802570099000424e-06,
      "loss": 0.99819261,
      "num_input_tokens_seen": 5466405,
      "step": 260,
      "time_per_iteration": 2.625072717666626
    },
    {
      "auxiliary_loss_clip": 0.01362999,
      "auxiliary_loss_mlp": 0.01127943,
      "balance_loss_clip": 1.1010474,
      "balance_loss_mlp": 1.07940137,
      "epoch": 0.015692168946339995,
      "flos": 29971728046080.0,
      "grad_norm": 2.2828802632863305,
      "language_loss": 0.87807435,
      "learning_rate": 3.5827286144073947e-06,
      "loss": 0.90298378,
      "num_input_tokens_seen": 5487055,
      "step": 261,
      "time_per_iteration": 2.6737279891967773
    },
    {
      "auxiliary_loss_clip": 0.01357008,
      "auxiliary_loss_mlp": 0.01125312,
      "balance_loss_clip": 1.09822345,
      "balance_loss_mlp": 1.07665133,
      "epoch": 0.015752292199007967,
      "flos": 19392085946880.0,
      "grad_norm": 5.057676675675106,
      "language_loss": 0.67100549,
      "learning_rate": 3.5851907672491904e-06,
      "loss": 0.69582868,
      "num_input_tokens_seen": 5506600,
      "step": 262,
      "time_per_iteration": 2.651690721511841
    },
    {
      "auxiliary_loss_clip": 0.01353953,
      "auxiliary_loss_mlp": 0.01135541,
      "balance_loss_clip": 1.09924924,
      "balance_loss_mlp": 1.08499634,
      "epoch": 0.015812415451675936,
      "flos": 20339804338560.0,
      "grad_norm": 3.0820356667611337,
      "language_loss": 0.68077701,
      "learning_rate": 3.587643540438383e-06,
      "loss": 0.70567191,
      "num_input_tokens_seen": 5524350,
      "step": 263,
      "time_per_iteration": 2.6885130405426025
    },
    {
      "auxiliary_loss_clip": 0.01355592,
      "auxiliary_loss_mlp": 0.01116799,
      "balance_loss_clip": 1.09620881,
      "balance_loss_mlp": 1.06766081,
      "epoch": 0.015872538704343905,
      "flos": 17525412979200.0,
      "grad_norm": 3.9089218881424674,
      "language_loss": 0.85002583,
      "learning_rate": 3.590087005168037e-06,
      "loss": 0.87474978,
      "num_input_tokens_seen": 5542145,
      "step": 264,
      "time_per_iteration": 2.6557912826538086
    },
    {
      "auxiliary_loss_clip": 0.01360388,
      "auxiliary_loss_mlp": 0.01102763,
      "balance_loss_clip": 1.10088885,
      "balance_loss_mlp": 1.056319,
      "epoch": 0.015932661957011873,
      "flos": 15260490944640.0,
      "grad_norm": 2.7020928553211476,
      "language_loss": 1.04234743,
      "learning_rate": 3.5925212318237344e-06,
      "loss": 1.06697881,
      "num_input_tokens_seen": 5557920,
      "step": 265,
      "time_per_iteration": 2.6262216567993164
    },
    {
      "auxiliary_loss_clip": 0.01364512,
      "auxiliary_loss_mlp": 0.01120309,
      "balance_loss_clip": 1.1033864,
      "balance_loss_mlp": 1.06835794,
      "epoch": 0.015992785209679845,
      "flos": 20302528999680.0,
      "grad_norm": 3.1220748516520134,
      "language_loss": 0.74914098,
      "learning_rate": 3.5949462899957323e-06,
      "loss": 0.7739892,
      "num_input_tokens_seen": 5576290,
      "step": 266,
      "time_per_iteration": 2.6244583129882812
    },
    {
      "auxiliary_loss_clip": 0.01349738,
      "auxiliary_loss_mlp": 0.0111189,
      "balance_loss_clip": 1.1000762,
      "balance_loss_mlp": 1.06206095,
      "epoch": 0.016052908462347814,
      "flos": 23362368969600.0,
      "grad_norm": 1.8166776194063956,
      "language_loss": 0.90909529,
      "learning_rate": 3.5973622484909068e-06,
      "loss": 0.93371153,
      "num_input_tokens_seen": 5595205,
      "step": 267,
      "time_per_iteration": 2.6753580570220947
    },
    {
      "auxiliary_loss_clip": 0.01359091,
      "auxiliary_loss_mlp": 0.01115968,
      "balance_loss_clip": 1.10122573,
      "balance_loss_mlp": 1.06797481,
      "epoch": 0.016113031715015783,
      "flos": 21286588976640.0,
      "grad_norm": 2.450608875877181,
      "language_loss": 0.85636413,
      "learning_rate": 3.599769175344462e-06,
      "loss": 0.88111478,
      "num_input_tokens_seen": 5612645,
      "step": 268,
      "time_per_iteration": 2.7161567211151123
    },
    {
      "auxiliary_loss_clip": 0.01351132,
      "auxiliary_loss_mlp": 0.01102276,
      "balance_loss_clip": 1.10226274,
      "balance_loss_mlp": 1.05475891,
      "epoch": 0.01617315496768375,
      "flos": 18914689261440.0,
      "grad_norm": 2.1714201716772457,
      "language_loss": 0.88080788,
      "learning_rate": 3.602167137831432e-06,
      "loss": 0.90534198,
      "num_input_tokens_seen": 5628345,
      "step": 269,
      "time_per_iteration": 2.6403756141662598
    },
    {
      "auxiliary_loss_clip": 0.01357907,
      "auxiliary_loss_mlp": 0.01111574,
      "balance_loss_clip": 1.10001528,
      "balance_loss_mlp": 1.06021833,
      "epoch": 0.01623327822035172,
      "flos": 16546488647040.0,
      "grad_norm": 2.5848702107942803,
      "language_loss": 0.97077739,
      "learning_rate": 3.6045562024779565e-06,
      "loss": 0.99547219,
      "num_input_tokens_seen": 5645940,
      "step": 270,
      "time_per_iteration": 2.635546922683716
    },
    {
      "auxiliary_loss_clip": 0.01356007,
      "auxiliary_loss_mlp": 0.01118132,
      "balance_loss_clip": 1.10402,
      "balance_loss_mlp": 1.06918478,
      "epoch": 0.016293401473019692,
      "flos": 23513481486720.0,
      "grad_norm": 2.1115750591463223,
      "language_loss": 0.86112005,
      "learning_rate": 3.606936435072361e-06,
      "loss": 0.8858614,
      "num_input_tokens_seen": 5665690,
      "step": 271,
      "time_per_iteration": 2.6877286434173584
    },
    {
      "auxiliary_loss_clip": 0.013537,
      "auxiliary_loss_mlp": 0.01105687,
      "balance_loss_clip": 1.0962286,
      "balance_loss_mlp": 1.057693,
      "epoch": 0.01635352472568766,
      "flos": 29016072748800.0,
      "grad_norm": 2.5391912683658413,
      "language_loss": 0.81550127,
      "learning_rate": 3.609307900676025e-06,
      "loss": 0.84009504,
      "num_input_tokens_seen": 5683190,
      "step": 272,
      "time_per_iteration": 2.6728365421295166
    },
    {
      "auxiliary_loss_clip": 0.01348527,
      "auxiliary_loss_mlp": 0.01120864,
      "balance_loss_clip": 1.09806561,
      "balance_loss_mlp": 1.07368064,
      "epoch": 0.01641364797835563,
      "flos": 13370513028480.0,
      "grad_norm": 2.3613573538590487,
      "language_loss": 0.81075382,
      "learning_rate": 3.611670663634051e-06,
      "loss": 0.83544779,
      "num_input_tokens_seen": 5699780,
      "step": 273,
      "time_per_iteration": 2.595008134841919
    },
    {
      "auxiliary_loss_clip": 0.01346135,
      "auxiliary_loss_mlp": 0.01105539,
      "balance_loss_clip": 1.09398317,
      "balance_loss_mlp": 1.05749762,
      "epoch": 0.016473771231023598,
      "flos": 18878239935360.0,
      "grad_norm": 2.1979313648400547,
      "language_loss": 0.9131726,
      "learning_rate": 3.614024787585744e-06,
      "loss": 0.9376893,
      "num_input_tokens_seen": 5716980,
      "step": 274,
      "time_per_iteration": 2.684718132019043
    },
    {
      "auxiliary_loss_clip": 0.013432,
      "auxiliary_loss_mlp": 0.01108715,
      "balance_loss_clip": 1.09515727,
      "balance_loss_mlp": 1.06062579,
      "epoch": 0.016533894483691566,
      "flos": 22601637803520.0,
      "grad_norm": 1.9719932168994616,
      "language_loss": 0.88054645,
      "learning_rate": 3.6163703354748927e-06,
      "loss": 0.90506566,
      "num_input_tokens_seen": 5737780,
      "step": 275,
      "time_per_iteration": 2.7204532623291016
    },
    {
      "auxiliary_loss_clip": 0.01346726,
      "auxiliary_loss_mlp": 0.01102856,
      "balance_loss_clip": 1.09623361,
      "balance_loss_mlp": 1.05312169,
      "epoch": 0.01659401773635954,
      "flos": 21507188353920.0,
      "grad_norm": 1.7930545784536995,
      "language_loss": 0.80726624,
      "learning_rate": 3.6187073695598707e-06,
      "loss": 0.83176208,
      "num_input_tokens_seen": 5758330,
      "step": 276,
      "time_per_iteration": 3.04716157913208
    },
    {
      "auxiliary_loss_clip": 0.0133817,
      "auxiliary_loss_mlp": 0.01096103,
      "balance_loss_clip": 1.09588337,
      "balance_loss_mlp": 1.05220985,
      "epoch": 0.016654140989027507,
      "flos": 32850973411200.0,
      "grad_norm": 1.9196343116615175,
      "language_loss": 0.80707026,
      "learning_rate": 3.621035951423551e-06,
      "loss": 0.83141291,
      "num_input_tokens_seen": 5778340,
      "step": 277,
      "time_per_iteration": 2.809645652770996
    },
    {
      "auxiliary_loss_clip": 0.01337061,
      "auxiliary_loss_mlp": 0.0109637,
      "balance_loss_clip": 1.08979487,
      "balance_loss_mlp": 1.04923487,
      "epoch": 0.016714264241695476,
      "flos": 12306228024960.0,
      "grad_norm": 2.3224792061881185,
      "language_loss": 0.80508065,
      "learning_rate": 3.623356141983041e-06,
      "loss": 0.82941496,
      "num_input_tokens_seen": 5794295,
      "step": 278,
      "time_per_iteration": 2.604830741882324
    },
    {
      "auxiliary_loss_clip": 0.01341116,
      "auxiliary_loss_mlp": 0.01101968,
      "balance_loss_clip": 1.09395671,
      "balance_loss_mlp": 1.05585837,
      "epoch": 0.016774387494363444,
      "flos": 27123796362240.0,
      "grad_norm": 2.0021377353660057,
      "language_loss": 0.90582991,
      "learning_rate": 3.6256680014992486e-06,
      "loss": 0.93026078,
      "num_input_tokens_seen": 5814405,
      "step": 279,
      "time_per_iteration": 2.7193243503570557
    },
    {
      "auxiliary_loss_clip": 0.01346095,
      "auxiliary_loss_mlp": 0.01112065,
      "balance_loss_clip": 1.09383631,
      "balance_loss_mlp": 1.06450009,
      "epoch": 0.016834510747031413,
      "flos": 20191493082240.0,
      "grad_norm": 2.9314445951013988,
      "language_loss": 0.94049025,
      "learning_rate": 3.6279715895862713e-06,
      "loss": 0.96507192,
      "num_input_tokens_seen": 5832795,
      "step": 280,
      "time_per_iteration": 2.680924654006958
    },
    {
      "auxiliary_loss_clip": 0.01346658,
      "auxiliary_loss_mlp": 0.01109166,
      "balance_loss_clip": 1.09285879,
      "balance_loss_mlp": 1.06060064,
      "epoch": 0.016894633999699385,
      "flos": 27274262434560.0,
      "grad_norm": 2.6758913403282483,
      "language_loss": 0.74425459,
      "learning_rate": 3.6302669652206183e-06,
      "loss": 0.76881289,
      "num_input_tokens_seen": 5855750,
      "step": 281,
      "time_per_iteration": 2.691152811050415
    },
    {
      "auxiliary_loss_clip": 0.01343371,
      "auxiliary_loss_mlp": 0.01117708,
      "balance_loss_clip": 1.09609079,
      "balance_loss_mlp": 1.0724318,
      "epoch": 0.016954757252367354,
      "flos": 14902964922240.0,
      "grad_norm": 3.4878028680462005,
      "language_loss": 0.80255079,
      "learning_rate": 3.632554186750274e-06,
      "loss": 0.82716167,
      "num_input_tokens_seen": 5872610,
      "step": 282,
      "time_per_iteration": 2.592664957046509
    },
    {
      "auxiliary_loss_clip": 0.01348082,
      "auxiliary_loss_mlp": 0.01118449,
      "balance_loss_clip": 1.09700727,
      "balance_loss_mlp": 1.07114697,
      "epoch": 0.017014880505035322,
      "flos": 21358805270400.0,
      "grad_norm": 2.296781711700251,
      "language_loss": 0.77719986,
      "learning_rate": 3.6348333119035937e-06,
      "loss": 0.80186516,
      "num_input_tokens_seen": 5892985,
      "step": 283,
      "time_per_iteration": 2.6502227783203125
    },
    {
      "auxiliary_loss_clip": 0.01347311,
      "auxiliary_loss_mlp": 0.01092934,
      "balance_loss_clip": 1.0977478,
      "balance_loss_mlp": 1.04804015,
      "epoch": 0.01707500375770329,
      "flos": 35333154858240.0,
      "grad_norm": 2.3467060832193414,
      "language_loss": 0.84246969,
      "learning_rate": 3.6371043977980503e-06,
      "loss": 0.86687213,
      "num_input_tokens_seen": 5914060,
      "step": 284,
      "time_per_iteration": 2.8534958362579346
    },
    {
      "auxiliary_loss_clip": 0.01337962,
      "auxiliary_loss_mlp": 0.01100399,
      "balance_loss_clip": 1.09212708,
      "balance_loss_mlp": 1.05297756,
      "epoch": 0.01713512701037126,
      "flos": 23582070506880.0,
      "grad_norm": 2.7335752956200388,
      "language_loss": 0.96998906,
      "learning_rate": 3.639367500948819e-06,
      "loss": 0.99437273,
      "num_input_tokens_seen": 5932860,
      "step": 285,
      "time_per_iteration": 2.6338655948638916
    },
    {
      "auxiliary_loss_clip": 0.01341319,
      "auxiliary_loss_mlp": 0.01095606,
      "balance_loss_clip": 1.09538078,
      "balance_loss_mlp": 1.05123687,
      "epoch": 0.01719525026303923,
      "flos": 27634661544960.0,
      "grad_norm": 2.294843469150046,
      "language_loss": 0.94079655,
      "learning_rate": 3.6416226772772178e-06,
      "loss": 0.96516573,
      "num_input_tokens_seen": 5952725,
      "step": 286,
      "time_per_iteration": 2.711087942123413
    },
    {
      "auxiliary_loss_clip": 0.01332862,
      "auxiliary_loss_mlp": 0.0109035,
      "balance_loss_clip": 1.08986938,
      "balance_loss_mlp": 1.04409683,
      "epoch": 0.0172553735157072,
      "flos": 26979722910720.0,
      "grad_norm": 1.9277896882465477,
      "language_loss": 0.92464817,
      "learning_rate": 3.643869982119001e-06,
      "loss": 0.94888031,
      "num_input_tokens_seen": 5970560,
      "step": 287,
      "time_per_iteration": 2.640267848968506
    },
    {
      "auxiliary_loss_clip": 0.01338192,
      "auxiliary_loss_mlp": 0.01092315,
      "balance_loss_clip": 1.09039164,
      "balance_loss_mlp": 1.04651475,
      "epoch": 0.01731549676837517,
      "flos": 14056621689600.0,
      "grad_norm": 2.7883535936791035,
      "language_loss": 1.01873291,
      "learning_rate": 3.646109470232502e-06,
      "loss": 1.04303789,
      "num_input_tokens_seen": 5982980,
      "step": 288,
      "time_per_iteration": 2.558312177658081
    },
    {
      "auxiliary_loss_clip": 0.01225082,
      "auxiliary_loss_mlp": 0.01188305,
      "balance_loss_clip": 1.09194219,
      "balance_loss_mlp": 1.17228377,
      "epoch": 0.017375620021043137,
      "flos": 66510694471680.0,
      "grad_norm": 0.9289960013542303,
      "language_loss": 0.63867617,
      "learning_rate": 3.6483411958066417e-06,
      "loss": 0.66281009,
      "num_input_tokens_seen": 6049445,
      "step": 289,
      "time_per_iteration": 3.386254072189331
    },
    {
      "auxiliary_loss_clip": 0.01341215,
      "auxiliary_loss_mlp": 0.01107788,
      "balance_loss_clip": 1.09622383,
      "balance_loss_mlp": 1.06482446,
      "epoch": 0.01743574327371111,
      "flos": 15225154940160.0,
      "grad_norm": 2.368974734045724,
      "language_loss": 0.88156199,
      "learning_rate": 3.6505652124687957e-06,
      "loss": 0.90605205,
      "num_input_tokens_seen": 6064150,
      "step": 290,
      "time_per_iteration": 2.5670948028564453
    },
    {
      "auxiliary_loss_clip": 0.0133848,
      "auxiliary_loss_mlp": 0.010946,
      "balance_loss_clip": 1.09388971,
      "balance_loss_mlp": 1.04965782,
      "epoch": 0.017495866526379078,
      "flos": 25373869574400.0,
      "grad_norm": 2.2011772664145504,
      "language_loss": 0.84472585,
      "learning_rate": 3.6527815732925258e-06,
      "loss": 0.8690567,
      "num_input_tokens_seen": 6083920,
      "step": 291,
      "time_per_iteration": 2.648452043533325
    },
    {
      "auxiliary_loss_clip": 0.01343563,
      "auxiliary_loss_mlp": 0.01115116,
      "balance_loss_clip": 1.10129941,
      "balance_loss_mlp": 1.06607366,
      "epoch": 0.017555989779047047,
      "flos": 26359473836160.0,
      "grad_norm": 1.7675259544479762,
      "language_loss": 0.72679955,
      "learning_rate": 3.6549903308051806e-06,
      "loss": 0.75138628,
      "num_input_tokens_seen": 6105460,
      "step": 292,
      "time_per_iteration": 2.7239537239074707
    },
    {
      "auxiliary_loss_clip": 0.01334066,
      "auxiliary_loss_mlp": 0.01107289,
      "balance_loss_clip": 1.09397244,
      "balance_loss_mlp": 1.06170392,
      "epoch": 0.017616113031715015,
      "flos": 22338807010560.0,
      "grad_norm": 2.419616990787406,
      "language_loss": 0.86866581,
      "learning_rate": 3.6571915369953646e-06,
      "loss": 0.89307928,
      "num_input_tokens_seen": 6122890,
      "step": 293,
      "time_per_iteration": 2.642854690551758
    },
    {
      "auxiliary_loss_clip": 0.01333726,
      "auxiliary_loss_mlp": 0.0110557,
      "balance_loss_clip": 1.09271646,
      "balance_loss_mlp": 1.06086659,
      "epoch": 0.017676236284382984,
      "flos": 20156911263360.0,
      "grad_norm": 2.112624444766753,
      "language_loss": 0.80896151,
      "learning_rate": 3.6593852433202797e-06,
      "loss": 0.83335447,
      "num_input_tokens_seen": 6142890,
      "step": 294,
      "time_per_iteration": 2.598176956176758
    },
    {
      "auxiliary_loss_clip": 0.01334179,
      "auxiliary_loss_mlp": 0.01113433,
      "balance_loss_clip": 1.09030747,
      "balance_loss_mlp": 1.06892014,
      "epoch": 0.017736359537050956,
      "flos": 25223331674880.0,
      "grad_norm": 2.8289841764142416,
      "language_loss": 0.83806521,
      "learning_rate": 3.6615715007129453e-06,
      "loss": 0.86254132,
      "num_input_tokens_seen": 6162030,
      "step": 295,
      "time_per_iteration": 2.750103712081909
    },
    {
      "auxiliary_loss_clip": 0.01339845,
      "auxiliary_loss_mlp": 0.01121984,
      "balance_loss_clip": 1.09978509,
      "balance_loss_mlp": 1.0772326,
      "epoch": 0.017796482789718925,
      "flos": 20338798757760.0,
      "grad_norm": 1.8804378237246864,
      "language_loss": 0.84576106,
      "learning_rate": 3.6637503595892897e-06,
      "loss": 0.87037927,
      "num_input_tokens_seen": 6180540,
      "step": 296,
      "time_per_iteration": 4.154251337051392
    },
    {
      "auxiliary_loss_clip": 0.01337678,
      "auxiliary_loss_mlp": 0.01105295,
      "balance_loss_clip": 1.09463406,
      "balance_loss_mlp": 1.06154561,
      "epoch": 0.017856606042386893,
      "flos": 22379206832640.0,
      "grad_norm": 2.055710812588959,
      "language_loss": 0.87810111,
      "learning_rate": 3.665921869855132e-06,
      "loss": 0.90253091,
      "num_input_tokens_seen": 6199425,
      "step": 297,
      "time_per_iteration": 4.379676103591919
    },
    {
      "auxiliary_loss_clip": 0.0133717,
      "auxiliary_loss_mlp": 0.01103766,
      "balance_loss_clip": 1.09343684,
      "balance_loss_mlp": 1.06004047,
      "epoch": 0.017916729295054862,
      "flos": 20230061310720.0,
      "grad_norm": 2.689351030321763,
      "language_loss": 0.88947791,
      "learning_rate": 3.6680860809130346e-06,
      "loss": 0.91388726,
      "num_input_tokens_seen": 6219170,
      "step": 298,
      "time_per_iteration": 4.1055779457092285
    },
    {
      "auxiliary_loss_clip": 0.01333843,
      "auxiliary_loss_mlp": 0.01121179,
      "balance_loss_clip": 1.09470236,
      "balance_loss_mlp": 1.07499719,
      "epoch": 0.01797685254772283,
      "flos": 19390972625280.0,
      "grad_norm": 1.8935027270905305,
      "language_loss": 0.88550889,
      "learning_rate": 3.6702430416690516e-06,
      "loss": 0.91005915,
      "num_input_tokens_seen": 6237930,
      "step": 299,
      "time_per_iteration": 2.611168622970581
    },
    {
      "auxiliary_loss_clip": 0.0133938,
      "auxiliary_loss_mlp": 0.0110718,
      "balance_loss_clip": 1.09468794,
      "balance_loss_mlp": 1.06130886,
      "epoch": 0.018036975800390802,
      "flos": 24426007528320.0,
      "grad_norm": 4.075580609786654,
      "language_loss": 0.64664406,
      "learning_rate": 3.672392800539357e-06,
      "loss": 0.67110968,
      "num_input_tokens_seen": 6257170,
      "step": 300,
      "time_per_iteration": 2.645603656768799
    },
    {
      "auxiliary_loss_clip": 0.01338559,
      "auxiliary_loss_mlp": 0.01111665,
      "balance_loss_clip": 1.09775913,
      "balance_loss_mlp": 1.06636548,
      "epoch": 0.01809709905305877,
      "flos": 15778933896960.0,
      "grad_norm": 2.5071418214687515,
      "language_loss": 0.87940675,
      "learning_rate": 3.6745354054567686e-06,
      "loss": 0.90390897,
      "num_input_tokens_seen": 6274780,
      "step": 301,
      "time_per_iteration": 2.6035923957824707
    },
    {
      "auxiliary_loss_clip": 0.01238361,
      "auxiliary_loss_mlp": 0.01073699,
      "balance_loss_clip": 1.1100142,
      "balance_loss_mlp": 1.05901265,
      "epoch": 0.01815722230572674,
      "flos": 67348382526720.0,
      "grad_norm": 0.8350739260664176,
      "language_loss": 0.62219667,
      "learning_rate": 3.676670903877158e-06,
      "loss": 0.64531732,
      "num_input_tokens_seen": 6340435,
      "step": 302,
      "time_per_iteration": 3.3307297229766846
    },
    {
      "auxiliary_loss_clip": 0.0132981,
      "auxiliary_loss_mlp": 0.01110918,
      "balance_loss_clip": 1.0910126,
      "balance_loss_mlp": 1.06507051,
      "epoch": 0.01821734555839471,
      "flos": 15485615435520.0,
      "grad_norm": 2.115144575016314,
      "language_loss": 0.89737153,
      "learning_rate": 3.6787993427857567e-06,
      "loss": 0.9217788,
      "num_input_tokens_seen": 6358160,
      "step": 303,
      "time_per_iteration": 2.6773293018341064
    },
    {
      "auxiliary_loss_clip": 0.01335628,
      "auxiliary_loss_mlp": 0.01118481,
      "balance_loss_clip": 1.09579217,
      "balance_loss_mlp": 1.07237101,
      "epoch": 0.018277468811062677,
      "flos": 24097424889600.0,
      "grad_norm": 1.8670669350935472,
      "language_loss": 0.80417514,
      "learning_rate": 3.680920768703364e-06,
      "loss": 0.82871628,
      "num_input_tokens_seen": 6378485,
      "step": 304,
      "time_per_iteration": 2.691347360610962
    },
    {
      "auxiliary_loss_clip": 0.01330802,
      "auxiliary_loss_mlp": 0.01091671,
      "balance_loss_clip": 1.09832263,
      "balance_loss_mlp": 1.04858923,
      "epoch": 0.01833759206373065,
      "flos": 20959335141120.0,
      "grad_norm": 1.6863564291935742,
      "language_loss": 0.82761526,
      "learning_rate": 3.6830352276924415e-06,
      "loss": 0.85184002,
      "num_input_tokens_seen": 6397845,
      "step": 305,
      "time_per_iteration": 2.6883981227874756
    },
    {
      "auxiliary_loss_clip": 0.01330759,
      "auxiliary_loss_mlp": 0.01093908,
      "balance_loss_clip": 1.09012437,
      "balance_loss_mlp": 1.05115986,
      "epoch": 0.018397715316398618,
      "flos": 19390757143680.0,
      "grad_norm": 2.1780708917523297,
      "language_loss": 0.91148543,
      "learning_rate": 3.685142765363119e-06,
      "loss": 0.93573213,
      "num_input_tokens_seen": 6416475,
      "step": 306,
      "time_per_iteration": 2.6465187072753906
    },
    {
      "auxiliary_loss_clip": 0.01324743,
      "auxiliary_loss_mlp": 0.01091696,
      "balance_loss_clip": 1.08900762,
      "balance_loss_mlp": 1.04882836,
      "epoch": 0.018457838569066586,
      "flos": 29132531619840.0,
      "grad_norm": 3.4680205003751072,
      "language_loss": 0.86581063,
      "learning_rate": 3.687243426879095e-06,
      "loss": 0.88997507,
      "num_input_tokens_seen": 6437520,
      "step": 307,
      "time_per_iteration": 2.7787318229675293
    },
    {
      "auxiliary_loss_clip": 0.01326572,
      "auxiliary_loss_mlp": 0.01110018,
      "balance_loss_clip": 1.09346747,
      "balance_loss_mlp": 1.06247783,
      "epoch": 0.018517961821734555,
      "flos": 19208654167680.0,
      "grad_norm": 2.413130156754219,
      "language_loss": 0.71650648,
      "learning_rate": 3.6893372569634466e-06,
      "loss": 0.74087244,
      "num_input_tokens_seen": 6455680,
      "step": 308,
      "time_per_iteration": 2.652973175048828
    },
    {
      "auxiliary_loss_clip": 0.01331912,
      "auxiliary_loss_mlp": 0.01102766,
      "balance_loss_clip": 1.09061241,
      "balance_loss_mlp": 1.05911207,
      "epoch": 0.018578085074402523,
      "flos": 19863018184320.0,
      "grad_norm": 2.1869498369051077,
      "language_loss": 0.91841364,
      "learning_rate": 3.6914242999043395e-06,
      "loss": 0.94276047,
      "num_input_tokens_seen": 6474880,
      "step": 309,
      "time_per_iteration": 2.6613030433654785
    },
    {
      "auxiliary_loss_clip": 0.01339178,
      "auxiliary_loss_mlp": 0.01096668,
      "balance_loss_clip": 1.09145641,
      "balance_loss_mlp": 1.05084395,
      "epoch": 0.018638208327070496,
      "flos": 29606947476480.0,
      "grad_norm": 2.0400456475786353,
      "language_loss": 0.72784412,
      "learning_rate": 3.69350459956065e-06,
      "loss": 0.75220263,
      "num_input_tokens_seen": 6495945,
      "step": 310,
      "time_per_iteration": 2.705345392227173
    },
    {
      "auxiliary_loss_clip": 0.01331019,
      "auxiliary_loss_mlp": 0.01113021,
      "balance_loss_clip": 1.09560525,
      "balance_loss_mlp": 1.06922317,
      "epoch": 0.018698331579738464,
      "flos": 45731555907840.0,
      "grad_norm": 2.1345597100799645,
      "language_loss": 0.74162471,
      "learning_rate": 3.695578199367497e-06,
      "loss": 0.76606506,
      "num_input_tokens_seen": 6519930,
      "step": 311,
      "time_per_iteration": 2.846503496170044
    },
    {
      "auxiliary_loss_clip": 0.01338389,
      "auxiliary_loss_mlp": 0.01104203,
      "balance_loss_clip": 1.09206033,
      "balance_loss_mlp": 1.0609777,
      "epoch": 0.018758454832406433,
      "flos": 20483662308480.0,
      "grad_norm": 3.713635021153945,
      "language_loss": 0.91668129,
      "learning_rate": 3.6976451423416825e-06,
      "loss": 0.94110715,
      "num_input_tokens_seen": 6535070,
      "step": 312,
      "time_per_iteration": 2.598400592803955
    },
    {
      "auxiliary_loss_clip": 0.01339145,
      "auxiliary_loss_mlp": 0.01116197,
      "balance_loss_clip": 1.09512305,
      "balance_loss_mlp": 1.07034922,
      "epoch": 0.0188185780850744,
      "flos": 15777784661760.0,
      "grad_norm": 4.5530066286460045,
      "language_loss": 0.89634913,
      "learning_rate": 3.699705471087043e-06,
      "loss": 0.92090249,
      "num_input_tokens_seen": 6554135,
      "step": 313,
      "time_per_iteration": 2.6944596767425537
    },
    {
      "auxiliary_loss_clip": 0.01340962,
      "auxiliary_loss_mlp": 0.0109941,
      "balance_loss_clip": 1.09381938,
      "balance_loss_mlp": 1.05430174,
      "epoch": 0.018878701337742373,
      "flos": 22455732758400.0,
      "grad_norm": 2.3990870717118455,
      "language_loss": 0.7335974,
      "learning_rate": 3.7017592277997256e-06,
      "loss": 0.75800109,
      "num_input_tokens_seen": 6572275,
      "step": 314,
      "time_per_iteration": 2.6550133228302
    },
    {
      "auxiliary_loss_clip": 0.01329658,
      "auxiliary_loss_mlp": 0.01105546,
      "balance_loss_clip": 1.09075165,
      "balance_loss_mlp": 1.06246412,
      "epoch": 0.018938824590410342,
      "flos": 30993530238720.0,
      "grad_norm": 5.81191681220521,
      "language_loss": 0.89890182,
      "learning_rate": 3.7038064542733654e-06,
      "loss": 0.92325383,
      "num_input_tokens_seen": 6594520,
      "step": 315,
      "time_per_iteration": 2.7121222019195557
    },
    {
      "auxiliary_loss_clip": 0.0133262,
      "auxiliary_loss_mlp": 0.01096177,
      "balance_loss_clip": 1.09287357,
      "balance_loss_mlp": 1.05209303,
      "epoch": 0.01899894784307831,
      "flos": 23258910821760.0,
      "grad_norm": 2.446494284682687,
      "language_loss": 0.80517328,
      "learning_rate": 3.7058471919041945e-06,
      "loss": 0.82946122,
      "num_input_tokens_seen": 6614245,
      "step": 316,
      "time_per_iteration": 2.640573501586914
    },
    {
      "auxiliary_loss_clip": 0.01326654,
      "auxiliary_loss_mlp": 0.01094904,
      "balance_loss_clip": 1.09036672,
      "balance_loss_mlp": 1.05046248,
      "epoch": 0.01905907109574628,
      "flos": 17457901367040.0,
      "grad_norm": 2.3705495670370524,
      "language_loss": 0.90161496,
      "learning_rate": 3.7078814816960605e-06,
      "loss": 0.92583054,
      "num_input_tokens_seen": 6632015,
      "step": 317,
      "time_per_iteration": 2.594388246536255
    },
    {
      "auxiliary_loss_clip": 0.01324014,
      "auxiliary_loss_mlp": 0.01097498,
      "balance_loss_clip": 1.08944559,
      "balance_loss_mlp": 1.05281842,
      "epoch": 0.019119194348414248,
      "flos": 14970225139200.0,
      "grad_norm": 7.443622240044352,
      "language_loss": 0.90836811,
      "learning_rate": 3.709909364265374e-06,
      "loss": 0.93258321,
      "num_input_tokens_seen": 6649015,
      "step": 318,
      "time_per_iteration": 2.6647114753723145
    },
    {
      "auxiliary_loss_clip": 0.01326579,
      "auxiliary_loss_mlp": 0.01092817,
      "balance_loss_clip": 1.0886786,
      "balance_loss_mlp": 1.05102181,
      "epoch": 0.01917931760108222,
      "flos": 25482822503040.0,
      "grad_norm": 2.232217614618188,
      "language_loss": 0.93955356,
      "learning_rate": 3.7119308798459706e-06,
      "loss": 0.9637475,
      "num_input_tokens_seen": 6669225,
      "step": 319,
      "time_per_iteration": 2.6901800632476807
    },
    {
      "auxiliary_loss_clip": 0.01209258,
      "auxiliary_loss_mlp": 0.01057567,
      "balance_loss_clip": 1.08611965,
      "balance_loss_mlp": 1.04288089,
      "epoch": 0.01923944085375019,
      "flos": 71556967353600.0,
      "grad_norm": 1.0009907084180605,
      "language_loss": 0.59817195,
      "learning_rate": 3.7139460682939026e-06,
      "loss": 0.62084019,
      "num_input_tokens_seen": 6725775,
      "step": 320,
      "time_per_iteration": 3.1044812202453613
    },
    {
      "auxiliary_loss_clip": 0.01323701,
      "auxiliary_loss_mlp": 0.01105882,
      "balance_loss_clip": 1.08827436,
      "balance_loss_mlp": 1.06291938,
      "epoch": 0.019299564106418157,
      "flos": 19682495406720.0,
      "grad_norm": 3.6735645336458163,
      "language_loss": 0.89620435,
      "learning_rate": 3.715954969092154e-06,
      "loss": 0.92050016,
      "num_input_tokens_seen": 6744170,
      "step": 321,
      "time_per_iteration": 2.650325298309326
    },
    {
      "auxiliary_loss_clip": 0.01333523,
      "auxiliary_loss_mlp": 0.01118534,
      "balance_loss_clip": 1.09200621,
      "balance_loss_mlp": 1.07440257,
      "epoch": 0.019359687359086126,
      "flos": 24387151991040.0,
      "grad_norm": 2.289334718991835,
      "language_loss": 0.82897186,
      "learning_rate": 3.7179576213552805e-06,
      "loss": 0.85349244,
      "num_input_tokens_seen": 6764565,
      "step": 322,
      "time_per_iteration": 2.65793514251709
    },
    {
      "auxiliary_loss_clip": 0.01332983,
      "auxiliary_loss_mlp": 0.01092262,
      "balance_loss_clip": 1.09035325,
      "balance_loss_mlp": 1.05061018,
      "epoch": 0.019419810611754094,
      "flos": 23951376190080.0,
      "grad_norm": 2.3678949255052912,
      "language_loss": 0.72983897,
      "learning_rate": 3.719954063833981e-06,
      "loss": 0.75409144,
      "num_input_tokens_seen": 6785310,
      "step": 323,
      "time_per_iteration": 2.6827828884124756
    },
    {
      "auxiliary_loss_clip": 0.01321298,
      "auxiliary_loss_mlp": 0.01092254,
      "balance_loss_clip": 1.08474624,
      "balance_loss_mlp": 1.04974401,
      "epoch": 0.019479933864422067,
      "flos": 22160223567360.0,
      "grad_norm": 1.9971507164977458,
      "language_loss": 0.92358303,
      "learning_rate": 3.721944334919596e-06,
      "loss": 0.9477185,
      "num_input_tokens_seen": 6803290,
      "step": 324,
      "time_per_iteration": 2.667363405227661
    },
    {
      "auxiliary_loss_clip": 0.0133014,
      "auxiliary_loss_mlp": 0.01089098,
      "balance_loss_clip": 1.09217644,
      "balance_loss_mlp": 1.04878139,
      "epoch": 0.019540057117090035,
      "flos": 22236821320320.0,
      "grad_norm": 6.407507213214319,
      "language_loss": 0.65127969,
      "learning_rate": 3.7239284726485375e-06,
      "loss": 0.67547202,
      "num_input_tokens_seen": 6822570,
      "step": 325,
      "time_per_iteration": 2.658700466156006
    },
    {
      "auxiliary_loss_clip": 0.01328385,
      "auxiliary_loss_mlp": 0.01109788,
      "balance_loss_clip": 1.09598839,
      "balance_loss_mlp": 1.06675363,
      "epoch": 0.019600180369758004,
      "flos": 23076771932160.0,
      "grad_norm": 1.7177375017641943,
      "language_loss": 0.76394802,
      "learning_rate": 3.72590651470665e-06,
      "loss": 0.78832972,
      "num_input_tokens_seen": 6841910,
      "step": 326,
      "time_per_iteration": 2.6326630115509033
    },
    {
      "auxiliary_loss_clip": 0.01322824,
      "auxiliary_loss_mlp": 0.01103487,
      "balance_loss_clip": 1.09083152,
      "balance_loss_mlp": 1.06040514,
      "epoch": 0.019660303622425972,
      "flos": 25410857604480.0,
      "grad_norm": 2.041100065316132,
      "language_loss": 0.79262185,
      "learning_rate": 3.727878498433505e-06,
      "loss": 0.81688493,
      "num_input_tokens_seen": 6862480,
      "step": 327,
      "time_per_iteration": 2.7195518016815186
    },
    {
      "auxiliary_loss_clip": 0.0132945,
      "auxiliary_loss_mlp": 0.01099712,
      "balance_loss_clip": 1.09292865,
      "balance_loss_mlp": 1.05832207,
      "epoch": 0.01972042687509394,
      "flos": 23657519024640.0,
      "grad_norm": 2.852301933148325,
      "language_loss": 0.80569315,
      "learning_rate": 3.7298444608266328e-06,
      "loss": 0.82998472,
      "num_input_tokens_seen": 6882015,
      "step": 328,
      "time_per_iteration": 2.6789369583129883
    },
    {
      "auxiliary_loss_clip": 0.01327544,
      "auxiliary_loss_mlp": 0.01094059,
      "balance_loss_clip": 1.08719349,
      "balance_loss_mlp": 1.05045235,
      "epoch": 0.019780550127761913,
      "flos": 18223480869120.0,
      "grad_norm": 2.280823996815513,
      "language_loss": 0.93599927,
      "learning_rate": 3.731804438545683e-06,
      "loss": 0.96021533,
      "num_input_tokens_seen": 6899785,
      "step": 329,
      "time_per_iteration": 2.6043548583984375
    },
    {
      "auxiliary_loss_clip": 0.0133329,
      "auxiliary_loss_mlp": 0.0110952,
      "balance_loss_clip": 1.09211767,
      "balance_loss_mlp": 1.06629419,
      "epoch": 0.01984067338042988,
      "flos": 22418780641920.0,
      "grad_norm": 2.788704520584699,
      "language_loss": 0.7476396,
      "learning_rate": 3.7337584679165324e-06,
      "loss": 0.77206767,
      "num_input_tokens_seen": 6918575,
      "step": 330,
      "time_per_iteration": 2.706001043319702
    },
    {
      "auxiliary_loss_clip": 0.0133006,
      "auxiliary_loss_mlp": 0.01115344,
      "balance_loss_clip": 1.09077096,
      "balance_loss_mlp": 1.07280993,
      "epoch": 0.01990079663309785,
      "flos": 17055199013760.0,
      "grad_norm": 4.201650057157668,
      "language_loss": 0.93435889,
      "learning_rate": 3.7357065849353186e-06,
      "loss": 0.95881295,
      "num_input_tokens_seen": 6936965,
      "step": 331,
      "time_per_iteration": 2.6499180793762207
    },
    {
      "auxiliary_loss_clip": 0.01316843,
      "auxiliary_loss_mlp": 0.01085812,
      "balance_loss_clip": 1.08825564,
      "balance_loss_mlp": 1.04563856,
      "epoch": 0.01996091988576582,
      "flos": 15961791058560.0,
      "grad_norm": 2.5475056489813968,
      "language_loss": 0.9293468,
      "learning_rate": 3.737648825272422e-06,
      "loss": 0.95337331,
      "num_input_tokens_seen": 6953475,
      "step": 332,
      "time_per_iteration": 2.5990231037139893
    },
    {
      "auxiliary_loss_clip": 0.01325701,
      "auxiliary_loss_mlp": 0.01091941,
      "balance_loss_clip": 1.09376514,
      "balance_loss_mlp": 1.04902601,
      "epoch": 0.02002104313843379,
      "flos": 23586451966080.0,
      "grad_norm": 2.7319388202061106,
      "language_loss": 0.75380504,
      "learning_rate": 3.739585224276384e-06,
      "loss": 0.77798152,
      "num_input_tokens_seen": 6971630,
      "step": 333,
      "time_per_iteration": 2.6225569248199463
    },
    {
      "auxiliary_loss_clip": 0.01323488,
      "auxiliary_loss_mlp": 0.01083816,
      "balance_loss_clip": 1.08822608,
      "balance_loss_mlp": 1.04249835,
      "epoch": 0.02008116639110176,
      "flos": 34094883352320.0,
      "grad_norm": 3.3732742696494924,
      "language_loss": 0.78797042,
      "learning_rate": 3.7415158169777673e-06,
      "loss": 0.81204355,
      "num_input_tokens_seen": 6992775,
      "step": 334,
      "time_per_iteration": 2.725562572479248
    },
    {
      "auxiliary_loss_clip": 0.01325152,
      "auxiliary_loss_mlp": 0.01093257,
      "balance_loss_clip": 1.08535278,
      "balance_loss_mlp": 1.04867256,
      "epoch": 0.020141289643769728,
      "flos": 19683716469120.0,
      "grad_norm": 1.945115565921162,
      "language_loss": 0.83465719,
      "learning_rate": 3.7434406380929575e-06,
      "loss": 0.8588413,
      "num_input_tokens_seen": 7011425,
      "step": 335,
      "time_per_iteration": 2.638871192932129
    },
    {
      "auxiliary_loss_clip": 0.01322365,
      "auxiliary_loss_mlp": 0.01085854,
      "balance_loss_clip": 1.08842373,
      "balance_loss_mlp": 1.04405963,
      "epoch": 0.020201412896437697,
      "flos": 20740567357440.0,
      "grad_norm": 2.3527147371949058,
      "language_loss": 0.92432821,
      "learning_rate": 3.745359722027911e-06,
      "loss": 0.94841033,
      "num_input_tokens_seen": 7029450,
      "step": 336,
      "time_per_iteration": 2.6654980182647705
    },
    {
      "auxiliary_loss_clip": 0.01321531,
      "auxiliary_loss_mlp": 0.01079695,
      "balance_loss_clip": 1.08577883,
      "balance_loss_mlp": 1.03818631,
      "epoch": 0.020261536149105665,
      "flos": 20266510636800.0,
      "grad_norm": 1.7223490941555537,
      "language_loss": 0.88663971,
      "learning_rate": 3.7472731028818428e-06,
      "loss": 0.91065204,
      "num_input_tokens_seen": 7047555,
      "step": 337,
      "time_per_iteration": 4.246743440628052
    },
    {
      "auxiliary_loss_clip": 0.01312441,
      "auxiliary_loss_mlp": 0.01102336,
      "balance_loss_clip": 1.08320296,
      "balance_loss_mlp": 1.05841899,
      "epoch": 0.020321659401773638,
      "flos": 25848752307840.0,
      "grad_norm": 1.6493597356962735,
      "language_loss": 0.89869279,
      "learning_rate": 3.7491808144508626e-06,
      "loss": 0.92284054,
      "num_input_tokens_seen": 7068185,
      "step": 338,
      "time_per_iteration": 5.869866609573364
    },
    {
      "auxiliary_loss_clip": 0.01321566,
      "auxiliary_loss_mlp": 0.0109858,
      "balance_loss_clip": 1.08546185,
      "balance_loss_mlp": 1.05554605,
      "epoch": 0.020381782654441606,
      "flos": 17495033051520.0,
      "grad_norm": 2.1603069065052694,
      "language_loss": 0.85168982,
      "learning_rate": 3.7510828902315576e-06,
      "loss": 0.87589133,
      "num_input_tokens_seen": 7085955,
      "step": 339,
      "time_per_iteration": 2.603130340576172
    },
    {
      "auxiliary_loss_clip": 0.01328225,
      "auxiliary_loss_mlp": 0.01099064,
      "balance_loss_clip": 1.0902226,
      "balance_loss_mlp": 1.05524242,
      "epoch": 0.020441905907109575,
      "flos": 24243940465920.0,
      "grad_norm": 2.1746002196087817,
      "language_loss": 0.88821882,
      "learning_rate": 3.75297936342452e-06,
      "loss": 0.91249174,
      "num_input_tokens_seen": 7106345,
      "step": 340,
      "time_per_iteration": 2.7247626781463623
    },
    {
      "auxiliary_loss_clip": 0.01322505,
      "auxiliary_loss_mlp": 0.01085559,
      "balance_loss_clip": 1.08594203,
      "balance_loss_mlp": 1.04004502,
      "epoch": 0.020502029159777543,
      "flos": 22233301787520.0,
      "grad_norm": 2.004763613818719,
      "language_loss": 0.88489276,
      "learning_rate": 3.7548702669378253e-06,
      "loss": 0.9089734,
      "num_input_tokens_seen": 7125070,
      "step": 341,
      "time_per_iteration": 2.731411933898926
    },
    {
      "auxiliary_loss_clip": 0.01324734,
      "auxiliary_loss_mlp": 0.01098572,
      "balance_loss_clip": 1.08451748,
      "balance_loss_mlp": 1.05479813,
      "epoch": 0.020562152412445512,
      "flos": 23987861429760.0,
      "grad_norm": 2.3638593093640736,
      "language_loss": 0.80611861,
      "learning_rate": 3.756755633390458e-06,
      "loss": 0.83035159,
      "num_input_tokens_seen": 7144675,
      "step": 342,
      "time_per_iteration": 2.6085095405578613
    },
    {
      "auxiliary_loss_clip": 0.01313805,
      "auxiliary_loss_mlp": 0.01098164,
      "balance_loss_clip": 1.08411694,
      "balance_loss_mlp": 1.05138612,
      "epoch": 0.020622275665113484,
      "flos": 26975305537920.0,
      "grad_norm": 1.727276092160433,
      "language_loss": 0.89612651,
      "learning_rate": 3.7586354951156886e-06,
      "loss": 0.92024612,
      "num_input_tokens_seen": 7165505,
      "step": 343,
      "time_per_iteration": 2.739912509918213
    },
    {
      "auxiliary_loss_clip": 0.01324722,
      "auxiliary_loss_mlp": 0.01096954,
      "balance_loss_clip": 1.09109879,
      "balance_loss_mlp": 1.05518293,
      "epoch": 0.020682398917781453,
      "flos": 22600704049920.0,
      "grad_norm": 2.6902665590614663,
      "language_loss": 0.78381217,
      "learning_rate": 3.7605098841644e-06,
      "loss": 0.80802888,
      "num_input_tokens_seen": 7184605,
      "step": 344,
      "time_per_iteration": 2.638439655303955
    },
    {
      "auxiliary_loss_clip": 0.01310552,
      "auxiliary_loss_mlp": 0.01103983,
      "balance_loss_clip": 1.08375537,
      "balance_loss_mlp": 1.05982804,
      "epoch": 0.02074252217044942,
      "flos": 15013605790080.0,
      "grad_norm": 2.2675296623639114,
      "language_loss": 0.75051636,
      "learning_rate": 3.7623788323083666e-06,
      "loss": 0.77466166,
      "num_input_tokens_seen": 7203065,
      "step": 345,
      "time_per_iteration": 2.581258773803711
    },
    {
      "auxiliary_loss_clip": 0.01316305,
      "auxiliary_loss_mlp": 0.01107937,
      "balance_loss_clip": 1.08855689,
      "balance_loss_mlp": 1.06447339,
      "epoch": 0.02080264542311739,
      "flos": 25337958952320.0,
      "grad_norm": 2.2144688897761395,
      "language_loss": 0.90414572,
      "learning_rate": 3.7642423710434837e-06,
      "loss": 0.92838824,
      "num_input_tokens_seen": 7222995,
      "step": 346,
      "time_per_iteration": 2.6281676292419434
    },
    {
      "auxiliary_loss_clip": 0.01312286,
      "auxiliary_loss_mlp": 0.01096576,
      "balance_loss_clip": 1.08357453,
      "balance_loss_mlp": 1.05621195,
      "epoch": 0.02086276867578536,
      "flos": 24388804016640.0,
      "grad_norm": 3.1106741063140366,
      "language_loss": 0.79133296,
      "learning_rate": 3.7661005315929563e-06,
      "loss": 0.81542158,
      "num_input_tokens_seen": 7244625,
      "step": 347,
      "time_per_iteration": 2.6477038860321045
    },
    {
      "auxiliary_loss_clip": 0.01317665,
      "auxiliary_loss_mlp": 0.01097416,
      "balance_loss_clip": 1.08921003,
      "balance_loss_mlp": 1.05328524,
      "epoch": 0.02092289192845333,
      "flos": 24462205459200.0,
      "grad_norm": 3.7065871267995893,
      "language_loss": 0.71211165,
      "learning_rate": 3.7679533449104354e-06,
      "loss": 0.73626244,
      "num_input_tokens_seen": 7263255,
      "step": 348,
      "time_per_iteration": 2.6215686798095703
    },
    {
      "auxiliary_loss_clip": 0.01319168,
      "auxiliary_loss_mlp": 0.01104109,
      "balance_loss_clip": 1.0859139,
      "balance_loss_mlp": 1.06066906,
      "epoch": 0.0209830151811213,
      "flos": 17451185523840.0,
      "grad_norm": 2.3976328225512495,
      "language_loss": 0.77118891,
      "learning_rate": 3.7698008416831116e-06,
      "loss": 0.79542166,
      "num_input_tokens_seen": 7279275,
      "step": 349,
      "time_per_iteration": 2.60102915763855
    },
    {
      "auxiliary_loss_clip": 0.01304146,
      "auxiliary_loss_mlp": 0.01101496,
      "balance_loss_clip": 1.08412242,
      "balance_loss_mlp": 1.06017756,
      "epoch": 0.021043138433789268,
      "flos": 24573995562240.0,
      "grad_norm": 1.7599420553547571,
      "language_loss": 0.85191035,
      "learning_rate": 3.7716430523347664e-06,
      "loss": 0.87596673,
      "num_input_tokens_seen": 7300180,
      "step": 350,
      "time_per_iteration": 2.7636313438415527
    },
    {
      "auxiliary_loss_clip": 0.01310639,
      "auxiliary_loss_mlp": 0.01090182,
      "balance_loss_clip": 1.08742464,
      "balance_loss_mlp": 1.05015147,
      "epoch": 0.021103261686457236,
      "flos": 24454053072000.0,
      "grad_norm": 2.2188224040826956,
      "language_loss": 0.7998929,
      "learning_rate": 3.773480007028776e-06,
      "loss": 0.82390112,
      "num_input_tokens_seen": 7317430,
      "step": 351,
      "time_per_iteration": 2.651803493499756
    },
    {
      "auxiliary_loss_clip": 0.01318922,
      "auxiliary_loss_mlp": 0.01104903,
      "balance_loss_clip": 1.08851838,
      "balance_loss_mlp": 1.06093884,
      "epoch": 0.021163384939125205,
      "flos": 14683083816960.0,
      "grad_norm": 2.30399977815629,
      "language_loss": 0.8746841,
      "learning_rate": 3.775311735671078e-06,
      "loss": 0.89892232,
      "num_input_tokens_seen": 7334875,
      "step": 352,
      "time_per_iteration": 2.687080144882202
    },
    {
      "auxiliary_loss_clip": 0.01311303,
      "auxiliary_loss_mlp": 0.01101912,
      "balance_loss_clip": 1.0859803,
      "balance_loss_mlp": 1.05861485,
      "epoch": 0.021223508191793177,
      "flos": 24493195918080.0,
      "grad_norm": 2.574621592267882,
      "language_loss": 0.8247534,
      "learning_rate": 3.7771382679130878e-06,
      "loss": 0.84888554,
      "num_input_tokens_seen": 7355185,
      "step": 353,
      "time_per_iteration": 2.7096078395843506
    },
    {
      "auxiliary_loss_clip": 0.01308698,
      "auxiliary_loss_mlp": 0.01092448,
      "balance_loss_clip": 1.08573294,
      "balance_loss_mlp": 1.05160654,
      "epoch": 0.021283631444461146,
      "flos": 24126978804480.0,
      "grad_norm": 1.9591973719581535,
      "language_loss": 0.8089481,
      "learning_rate": 3.7789596331545845e-06,
      "loss": 0.83295953,
      "num_input_tokens_seen": 7374425,
      "step": 354,
      "time_per_iteration": 2.658649444580078
    },
    {
      "auxiliary_loss_clip": 0.01314249,
      "auxiliary_loss_mlp": 0.01095812,
      "balance_loss_clip": 1.08369493,
      "balance_loss_mlp": 1.05218124,
      "epoch": 0.021343754697129114,
      "flos": 25192233475200.0,
      "grad_norm": 2.22170783568627,
      "language_loss": 0.81311834,
      "learning_rate": 3.780775860546545e-06,
      "loss": 0.837219,
      "num_input_tokens_seen": 7394175,
      "step": 355,
      "time_per_iteration": 2.619551420211792
    },
    {
      "auxiliary_loss_clip": 0.01310207,
      "auxiliary_loss_mlp": 0.01090401,
      "balance_loss_clip": 1.08222032,
      "balance_loss_mlp": 1.04851055,
      "epoch": 0.021403877949797083,
      "flos": 17274182279040.0,
      "grad_norm": 2.212340256471132,
      "language_loss": 0.89746779,
      "learning_rate": 3.7825869789939474e-06,
      "loss": 0.92147392,
      "num_input_tokens_seen": 7412645,
      "step": 356,
      "time_per_iteration": 2.5877137184143066
    },
    {
      "auxiliary_loss_clip": 0.01308298,
      "auxiliary_loss_mlp": 0.0108474,
      "balance_loss_clip": 1.08573771,
      "balance_loss_mlp": 1.04191971,
      "epoch": 0.021464001202465055,
      "flos": 30917435276160.0,
      "grad_norm": 1.9878508054592678,
      "language_loss": 0.79956681,
      "learning_rate": 3.784393017158528e-06,
      "loss": 0.82349718,
      "num_input_tokens_seen": 7432275,
      "step": 357,
      "time_per_iteration": 2.781755208969116
    },
    {
      "auxiliary_loss_clip": 0.0130988,
      "auxiliary_loss_mlp": 0.01083565,
      "balance_loss_clip": 1.08250284,
      "balance_loss_mlp": 1.04417801,
      "epoch": 0.021524124455133024,
      "flos": 18186385098240.0,
      "grad_norm": 2.6679617624252137,
      "language_loss": 0.76516652,
      "learning_rate": 3.786194003461506e-06,
      "loss": 0.78910094,
      "num_input_tokens_seen": 7450245,
      "step": 358,
      "time_per_iteration": 2.63144850730896
    },
    {
      "auxiliary_loss_clip": 0.01307251,
      "auxiliary_loss_mlp": 0.01092013,
      "balance_loss_clip": 1.08083165,
      "balance_loss_mlp": 1.04842997,
      "epoch": 0.021584247707800992,
      "flos": 13805786039040.0,
      "grad_norm": 2.344744226979962,
      "language_loss": 0.88770491,
      "learning_rate": 3.787989966086264e-06,
      "loss": 0.91169769,
      "num_input_tokens_seen": 7466845,
      "step": 359,
      "time_per_iteration": 2.641932964324951
    },
    {
      "auxiliary_loss_clip": 0.01315087,
      "auxiliary_loss_mlp": 0.01090441,
      "balance_loss_clip": 1.08486438,
      "balance_loss_mlp": 1.05088758,
      "epoch": 0.02164437096046896,
      "flos": 23294713703040.0,
      "grad_norm": 3.6505103877164804,
      "language_loss": 0.75853801,
      "learning_rate": 3.789780932980997e-06,
      "loss": 0.78259325,
      "num_input_tokens_seen": 7485450,
      "step": 360,
      "time_per_iteration": 2.5901477336883545
    },
    {
      "auxiliary_loss_clip": 0.01203506,
      "auxiliary_loss_mlp": 0.0103078,
      "balance_loss_clip": 1.07682121,
      "balance_loss_mlp": 1.01781011,
      "epoch": 0.02170449421313693,
      "flos": 68899578341760.0,
      "grad_norm": 0.8439708743577624,
      "language_loss": 0.64861441,
      "learning_rate": 3.79156693186132e-06,
      "loss": 0.67095727,
      "num_input_tokens_seen": 7553780,
      "step": 361,
      "time_per_iteration": 3.278409957885742
    },
    {
      "auxiliary_loss_clip": 0.01306068,
      "auxiliary_loss_mlp": 0.01086116,
      "balance_loss_clip": 1.0792098,
      "balance_loss_mlp": 1.04501224,
      "epoch": 0.0217646174658049,
      "flos": 25228539146880.0,
      "grad_norm": 3.144635825096315,
      "language_loss": 0.78844237,
      "learning_rate": 3.7933479902128433e-06,
      "loss": 0.81236422,
      "num_input_tokens_seen": 7574155,
      "step": 362,
      "time_per_iteration": 2.6302051544189453
    },
    {
      "auxiliary_loss_clip": 0.01309585,
      "auxiliary_loss_mlp": 0.01093258,
      "balance_loss_clip": 1.08188891,
      "balance_loss_mlp": 1.05244076,
      "epoch": 0.02182474071847287,
      "flos": 22893124671360.0,
      "grad_norm": 2.019833715135914,
      "language_loss": 0.92474592,
      "learning_rate": 3.7951241352937077e-06,
      "loss": 0.94877434,
      "num_input_tokens_seen": 7592320,
      "step": 363,
      "time_per_iteration": 2.6566081047058105
    },
    {
      "auxiliary_loss_clip": 0.01305173,
      "auxiliary_loss_mlp": 0.01096467,
      "balance_loss_clip": 1.0816617,
      "balance_loss_mlp": 1.05693769,
      "epoch": 0.02188486397114084,
      "flos": 23658991482240.0,
      "grad_norm": 2.282586403147275,
      "language_loss": 0.89844346,
      "learning_rate": 3.7968953941370915e-06,
      "loss": 0.92245984,
      "num_input_tokens_seen": 7611185,
      "step": 364,
      "time_per_iteration": 2.711911201477051
    },
    {
      "auxiliary_loss_clip": 0.01311963,
      "auxiliary_loss_mlp": 0.0109247,
      "balance_loss_clip": 1.08607888,
      "balance_loss_mlp": 1.04955506,
      "epoch": 0.021944987223808807,
      "flos": 21543637680000.0,
      "grad_norm": 1.948927065488749,
      "language_loss": 0.79460645,
      "learning_rate": 3.798661793553676e-06,
      "loss": 0.81865084,
      "num_input_tokens_seen": 7631970,
      "step": 365,
      "time_per_iteration": 2.6396052837371826
    },
    {
      "auxiliary_loss_clip": 0.01306043,
      "auxiliary_loss_mlp": 0.01100405,
      "balance_loss_clip": 1.08267248,
      "balance_loss_mlp": 1.05658317,
      "epoch": 0.022005110476476776,
      "flos": 16070887641600.0,
      "grad_norm": 1.85181498507666,
      "language_loss": 0.84341359,
      "learning_rate": 3.8004233601340808e-06,
      "loss": 0.86747801,
      "num_input_tokens_seen": 7649745,
      "step": 366,
      "time_per_iteration": 2.6278867721557617
    },
    {
      "auxiliary_loss_clip": 0.01312113,
      "auxiliary_loss_mlp": 0.01087574,
      "balance_loss_clip": 1.08304918,
      "balance_loss_mlp": 1.04859269,
      "epoch": 0.022065233729144748,
      "flos": 21433715084160.0,
      "grad_norm": 1.9326288300300676,
      "language_loss": 0.87040466,
      "learning_rate": 3.8021801202512694e-06,
      "loss": 0.89440155,
      "num_input_tokens_seen": 7668830,
      "step": 367,
      "time_per_iteration": 2.6410560607910156
    },
    {
      "auxiliary_loss_clip": 0.01312217,
      "auxiliary_loss_mlp": 0.01096053,
      "balance_loss_clip": 1.08074582,
      "balance_loss_mlp": 1.05335259,
      "epoch": 0.022125356981812717,
      "flos": 21543709507200.0,
      "grad_norm": 2.7247329926128976,
      "language_loss": 0.8487373,
      "learning_rate": 3.803932100062912e-06,
      "loss": 0.87282002,
      "num_input_tokens_seen": 7687240,
      "step": 368,
      "time_per_iteration": 2.652012825012207
    },
    {
      "auxiliary_loss_clip": 0.01312089,
      "auxiliary_loss_mlp": 0.01079926,
      "balance_loss_clip": 1.0801568,
      "balance_loss_mlp": 1.04027653,
      "epoch": 0.022185480234480685,
      "flos": 20704153944960.0,
      "grad_norm": 2.4839328990540794,
      "language_loss": 0.75997221,
      "learning_rate": 3.8056793255137264e-06,
      "loss": 0.78389233,
      "num_input_tokens_seen": 7704440,
      "step": 369,
      "time_per_iteration": 2.601384401321411
    },
    {
      "auxiliary_loss_clip": 0.01306737,
      "auxiliary_loss_mlp": 0.01099274,
      "balance_loss_clip": 1.08232927,
      "balance_loss_mlp": 1.05836105,
      "epoch": 0.022245603487148654,
      "flos": 25193203142400.0,
      "grad_norm": 2.189428421230448,
      "language_loss": 0.82977992,
      "learning_rate": 3.8074218223377844e-06,
      "loss": 0.85383999,
      "num_input_tokens_seen": 7727160,
      "step": 370,
      "time_per_iteration": 2.6538548469543457
    },
    {
      "auxiliary_loss_clip": 0.01306327,
      "auxiliary_loss_mlp": 0.01099594,
      "balance_loss_clip": 1.08127654,
      "balance_loss_mlp": 1.05713177,
      "epoch": 0.022305726739816623,
      "flos": 21395936954880.0,
      "grad_norm": 1.8569755368340455,
      "language_loss": 0.81588483,
      "learning_rate": 3.8091596160607834e-06,
      "loss": 0.83994406,
      "num_input_tokens_seen": 7747730,
      "step": 371,
      "time_per_iteration": 2.6779489517211914
    },
    {
      "auxiliary_loss_clip": 0.01311283,
      "auxiliary_loss_mlp": 0.01093653,
      "balance_loss_clip": 1.08593988,
      "balance_loss_mlp": 1.05169153,
      "epoch": 0.022365849992484595,
      "flos": 22492146170880.0,
      "grad_norm": 2.0622769904034817,
      "language_loss": 0.83493644,
      "learning_rate": 3.8108927320022896e-06,
      "loss": 0.85898578,
      "num_input_tokens_seen": 7766765,
      "step": 372,
      "time_per_iteration": 2.676797866821289
    },
    {
      "auxiliary_loss_clip": 0.01303906,
      "auxiliary_loss_mlp": 0.01091688,
      "balance_loss_clip": 1.08125615,
      "balance_loss_mlp": 1.05022752,
      "epoch": 0.022425973245152563,
      "flos": 17856581397120.0,
      "grad_norm": 2.8569846697004424,
      "language_loss": 0.79004842,
      "learning_rate": 3.8126211952779548e-06,
      "loss": 0.81400436,
      "num_input_tokens_seen": 7784010,
      "step": 373,
      "time_per_iteration": 2.593186616897583
    },
    {
      "auxiliary_loss_clip": 0.01309731,
      "auxiliary_loss_mlp": 0.01087409,
      "balance_loss_clip": 1.08431911,
      "balance_loss_mlp": 1.0448271,
      "epoch": 0.022486096497820532,
      "flos": 15483029656320.0,
      "grad_norm": 2.5442660874947385,
      "language_loss": 0.77622557,
      "learning_rate": 3.8143450308016952e-06,
      "loss": 0.80019701,
      "num_input_tokens_seen": 7801305,
      "step": 374,
      "time_per_iteration": 2.628392457962036
    },
    {
      "auxiliary_loss_clip": 0.0129871,
      "auxiliary_loss_mlp": 0.01076131,
      "balance_loss_clip": 1.07404125,
      "balance_loss_mlp": 1.03395462,
      "epoch": 0.0225462197504885,
      "flos": 27784157950080.0,
      "grad_norm": 1.574507922341891,
      "language_loss": 0.86032569,
      "learning_rate": 3.8160642632878525e-06,
      "loss": 0.88407415,
      "num_input_tokens_seen": 7823965,
      "step": 375,
      "time_per_iteration": 2.6783435344696045
    },
    {
      "auxiliary_loss_clip": 0.01307026,
      "auxiliary_loss_mlp": 0.01102393,
      "balance_loss_clip": 1.08340597,
      "balance_loss_mlp": 1.0590483,
      "epoch": 0.02260634300315647,
      "flos": 19975490645760.0,
      "grad_norm": 2.1279260859120286,
      "language_loss": 0.8901403,
      "learning_rate": 3.817778917253314e-06,
      "loss": 0.91423446,
      "num_input_tokens_seen": 7842115,
      "step": 376,
      "time_per_iteration": 2.621629476547241
    },
    {
      "auxiliary_loss_clip": 0.01306872,
      "auxiliary_loss_mlp": 0.01087647,
      "balance_loss_clip": 1.07870364,
      "balance_loss_mlp": 1.04868913,
      "epoch": 0.02266646625582444,
      "flos": 16028189349120.0,
      "grad_norm": 3.0367767906095917,
      "language_loss": 0.75437558,
      "learning_rate": 3.8194890170196155e-06,
      "loss": 0.77832079,
      "num_input_tokens_seen": 7857830,
      "step": 377,
      "time_per_iteration": 2.5465245246887207
    },
    {
      "auxiliary_loss_clip": 0.01298987,
      "auxiliary_loss_mlp": 0.01093623,
      "balance_loss_clip": 1.08128345,
      "balance_loss_mlp": 1.0517087,
      "epoch": 0.02272658950849241,
      "flos": 20404622430720.0,
      "grad_norm": 2.1955644054597374,
      "language_loss": 0.99231368,
      "learning_rate": 3.8211945867150055e-06,
      "loss": 1.01623976,
      "num_input_tokens_seen": 7875840,
      "step": 378,
      "time_per_iteration": 7.184643983840942
    },
    {
      "auxiliary_loss_clip": 0.01202133,
      "auxiliary_loss_mlp": 0.01040839,
      "balance_loss_clip": 1.0828104,
      "balance_loss_mlp": 1.0283463,
      "epoch": 0.02278671276116038,
      "flos": 69847332647040.0,
      "grad_norm": 0.9608118941287621,
      "language_loss": 0.75395739,
      "learning_rate": 3.822895650276492e-06,
      "loss": 0.7763871,
      "num_input_tokens_seen": 7940190,
      "step": 379,
      "time_per_iteration": 4.961140394210815
    },
    {
      "auxiliary_loss_clip": 0.01308523,
      "auxiliary_loss_mlp": 0.01087195,
      "balance_loss_clip": 1.07820678,
      "balance_loss_mlp": 1.04792738,
      "epoch": 0.022846836013828347,
      "flos": 38508771340800.0,
      "grad_norm": 3.7276648293904375,
      "language_loss": 0.78197825,
      "learning_rate": 3.824592231451859e-06,
      "loss": 0.8059355,
      "num_input_tokens_seen": 7960840,
      "step": 380,
      "time_per_iteration": 2.7892863750457764
    },
    {
      "auxiliary_loss_clip": 0.01301718,
      "auxiliary_loss_mlp": 0.01088822,
      "balance_loss_clip": 1.07955217,
      "balance_loss_mlp": 1.04945946,
      "epoch": 0.02290695926649632,
      "flos": 20959478795520.0,
      "grad_norm": 2.0941800643649855,
      "language_loss": 0.96743369,
      "learning_rate": 3.826284353801652e-06,
      "loss": 0.99133915,
      "num_input_tokens_seen": 7975500,
      "step": 381,
      "time_per_iteration": 2.619854688644409
    },
    {
      "auxiliary_loss_clip": 0.01311313,
      "auxiliary_loss_mlp": 0.01093973,
      "balance_loss_clip": 1.08192921,
      "balance_loss_mlp": 1.0539186,
      "epoch": 0.022967082519164288,
      "flos": 24022407335040.0,
      "grad_norm": 2.122042453210184,
      "language_loss": 0.87664795,
      "learning_rate": 3.827972040701142e-06,
      "loss": 0.90070075,
      "num_input_tokens_seen": 7993880,
      "step": 382,
      "time_per_iteration": 2.617398500442505
    },
    {
      "auxiliary_loss_clip": 0.01304042,
      "auxiliary_loss_mlp": 0.01096828,
      "balance_loss_clip": 1.0821979,
      "balance_loss_mlp": 1.05760849,
      "epoch": 0.023027205771832256,
      "flos": 20997149184000.0,
      "grad_norm": 1.978420170714987,
      "language_loss": 0.84990942,
      "learning_rate": 3.829655315342268e-06,
      "loss": 0.87391812,
      "num_input_tokens_seen": 8012730,
      "step": 383,
      "time_per_iteration": 2.6345314979553223
    },
    {
      "auxiliary_loss_clip": 0.01300873,
      "auxiliary_loss_mlp": 0.0111136,
      "balance_loss_clip": 1.08199024,
      "balance_loss_mlp": 1.0716393,
      "epoch": 0.023087329024500225,
      "flos": 21360816432000.0,
      "grad_norm": 2.0575071112917778,
      "language_loss": 0.83349717,
      "learning_rate": 3.831334200735543e-06,
      "loss": 0.8576194,
      "num_input_tokens_seen": 8031275,
      "step": 384,
      "time_per_iteration": 2.6339902877807617
    },
    {
      "auxiliary_loss_clip": 0.0129979,
      "auxiliary_loss_mlp": 0.010893,
      "balance_loss_clip": 1.08362782,
      "balance_loss_mlp": 1.05255938,
      "epoch": 0.023147452277168194,
      "flos": 21872435800320.0,
      "grad_norm": 1.7828777740185773,
      "language_loss": 0.89289594,
      "learning_rate": 3.8330087197119426e-06,
      "loss": 0.91678685,
      "num_input_tokens_seen": 8051600,
      "step": 385,
      "time_per_iteration": 2.690460205078125
    },
    {
      "auxiliary_loss_clip": 0.01305297,
      "auxiliary_loss_mlp": 0.01118129,
      "balance_loss_clip": 1.08288455,
      "balance_loss_mlp": 1.07926655,
      "epoch": 0.023207575529836166,
      "flos": 18916700423040.0,
      "grad_norm": 1.9487706588237765,
      "language_loss": 0.70157433,
      "learning_rate": 3.83467889492477e-06,
      "loss": 0.72580856,
      "num_input_tokens_seen": 8070600,
      "step": 386,
      "time_per_iteration": 2.681957721710205
    },
    {
      "auxiliary_loss_clip": 0.01305989,
      "auxiliary_loss_mlp": 0.0109088,
      "balance_loss_clip": 1.08441973,
      "balance_loss_mlp": 1.05309081,
      "epoch": 0.023267698782504134,
      "flos": 25046005207680.0,
      "grad_norm": 2.354342660334866,
      "language_loss": 0.87840039,
      "learning_rate": 3.836344748851495e-06,
      "loss": 0.90236908,
      "num_input_tokens_seen": 8090680,
      "step": 387,
      "time_per_iteration": 2.6511123180389404
    },
    {
      "auxiliary_loss_clip": 0.01304298,
      "auxiliary_loss_mlp": 0.01075541,
      "balance_loss_clip": 1.08178413,
      "balance_loss_mlp": 1.03658366,
      "epoch": 0.023327822035172103,
      "flos": 28879217930880.0,
      "grad_norm": 2.2068948332198643,
      "language_loss": 0.8341614,
      "learning_rate": 3.838006303795566e-06,
      "loss": 0.85795981,
      "num_input_tokens_seen": 8114610,
      "step": 388,
      "time_per_iteration": 2.7062034606933594
    },
    {
      "auxiliary_loss_clip": 0.01301997,
      "auxiliary_loss_mlp": 0.01089724,
      "balance_loss_clip": 1.08110905,
      "balance_loss_mlp": 1.05284107,
      "epoch": 0.02338794528784007,
      "flos": 27121533805440.0,
      "grad_norm": 2.1887236217853863,
      "language_loss": 0.93710232,
      "learning_rate": 3.839663581888206e-06,
      "loss": 0.96101958,
      "num_input_tokens_seen": 8133975,
      "step": 389,
      "time_per_iteration": 2.680280923843384
    },
    {
      "auxiliary_loss_clip": 0.01296082,
      "auxiliary_loss_mlp": 0.01083127,
      "balance_loss_clip": 1.0818491,
      "balance_loss_mlp": 1.04397893,
      "epoch": 0.02344806854050804,
      "flos": 21322355944320.0,
      "grad_norm": 1.981860280002506,
      "language_loss": 0.87747037,
      "learning_rate": 3.841316605090178e-06,
      "loss": 0.9012624,
      "num_input_tokens_seen": 8153570,
      "step": 390,
      "time_per_iteration": 2.65970516204834
    },
    {
      "auxiliary_loss_clip": 0.01301203,
      "auxiliary_loss_mlp": 0.01092853,
      "balance_loss_clip": 1.08357048,
      "balance_loss_mlp": 1.0568521,
      "epoch": 0.023508191793176012,
      "flos": 24789997998720.0,
      "grad_norm": 2.134782100250632,
      "language_loss": 0.89370871,
      "learning_rate": 3.842965395193529e-06,
      "loss": 0.91764927,
      "num_input_tokens_seen": 8170075,
      "step": 391,
      "time_per_iteration": 2.620009660720825
    },
    {
      "auxiliary_loss_clip": 0.01296395,
      "auxiliary_loss_mlp": 0.01072264,
      "balance_loss_clip": 1.07956719,
      "balance_loss_mlp": 1.03521371,
      "epoch": 0.02356831504584398,
      "flos": 25995375624960.0,
      "grad_norm": 2.366558958564603,
      "language_loss": 0.86076117,
      "learning_rate": 3.84460997382332e-06,
      "loss": 0.88444775,
      "num_input_tokens_seen": 8190420,
      "step": 392,
      "time_per_iteration": 2.7171695232391357
    },
    {
      "auxiliary_loss_clip": 0.01293283,
      "auxiliary_loss_mlp": 0.01084283,
      "balance_loss_clip": 1.07891107,
      "balance_loss_mlp": 1.04763794,
      "epoch": 0.02362843829851195,
      "flos": 19062461813760.0,
      "grad_norm": 2.038818686720474,
      "language_loss": 0.89096916,
      "learning_rate": 3.8462503624393256e-06,
      "loss": 0.91474473,
      "num_input_tokens_seen": 8208790,
      "step": 393,
      "time_per_iteration": 2.632129669189453
    },
    {
      "auxiliary_loss_clip": 0.01304158,
      "auxiliary_loss_mlp": 0.01102255,
      "balance_loss_clip": 1.08471596,
      "balance_loss_mlp": 1.06279635,
      "epoch": 0.023688561551179918,
      "flos": 16071031296000.0,
      "grad_norm": 1.7920692319020195,
      "language_loss": 0.8156364,
      "learning_rate": 3.84788658233771e-06,
      "loss": 0.83970058,
      "num_input_tokens_seen": 8226885,
      "step": 394,
      "time_per_iteration": 2.5932936668395996
    },
    {
      "auxiliary_loss_clip": 0.01296851,
      "auxiliary_loss_mlp": 0.01088191,
      "balance_loss_clip": 1.07939875,
      "balance_loss_mlp": 1.04920936,
      "epoch": 0.023748684803847887,
      "flos": 21724375939200.0,
      "grad_norm": 4.539737106404062,
      "language_loss": 0.85808635,
      "learning_rate": 3.84951865465269e-06,
      "loss": 0.88193679,
      "num_input_tokens_seen": 8246825,
      "step": 395,
      "time_per_iteration": 2.6112868785858154
    },
    {
      "auxiliary_loss_clip": 0.01194704,
      "auxiliary_loss_mlp": 0.01034684,
      "balance_loss_clip": 1.07210529,
      "balance_loss_mlp": 1.02319229,
      "epoch": 0.02380880805651586,
      "flos": 61926192881280.0,
      "grad_norm": 0.9258089920958834,
      "language_loss": 0.6380353,
      "learning_rate": 3.851146600358172e-06,
      "loss": 0.66032922,
      "num_input_tokens_seen": 8302835,
      "step": 396,
      "time_per_iteration": 3.031489133834839
    },
    {
      "auxiliary_loss_clip": 0.0129188,
      "auxiliary_loss_mlp": 0.01071022,
      "balance_loss_clip": 1.07806754,
      "balance_loss_mlp": 1.03447223,
      "epoch": 0.023868931309183827,
      "flos": 20266331068800.0,
      "grad_norm": 2.3741099598177624,
      "language_loss": 0.83878696,
      "learning_rate": 3.852770440269372e-06,
      "loss": 0.86241591,
      "num_input_tokens_seen": 8320745,
      "step": 397,
      "time_per_iteration": 2.6049532890319824
    },
    {
      "auxiliary_loss_clip": 0.01297108,
      "auxiliary_loss_mlp": 0.01087341,
      "balance_loss_clip": 1.08104038,
      "balance_loss_mlp": 1.04890823,
      "epoch": 0.023929054561851796,
      "flos": 21139103733120.0,
      "grad_norm": 4.6847154905409205,
      "language_loss": 0.84066498,
      "learning_rate": 3.854390195044404e-06,
      "loss": 0.86450952,
      "num_input_tokens_seen": 8339540,
      "step": 398,
      "time_per_iteration": 2.6516692638397217
    },
    {
      "auxiliary_loss_clip": 0.01295876,
      "auxiliary_loss_mlp": 0.01078722,
      "balance_loss_clip": 1.07671928,
      "balance_loss_mlp": 1.04007471,
      "epoch": 0.023989177814519765,
      "flos": 13698521049600.0,
      "grad_norm": 2.80358563189936,
      "language_loss": 0.86029691,
      "learning_rate": 3.856005885185868e-06,
      "loss": 0.88404286,
      "num_input_tokens_seen": 8354890,
      "step": 399,
      "time_per_iteration": 2.5452589988708496
    },
    {
      "auxiliary_loss_clip": 0.01292698,
      "auxiliary_loss_mlp": 0.01090822,
      "balance_loss_clip": 1.08074594,
      "balance_loss_mlp": 1.05308056,
      "epoch": 0.024049301067187733,
      "flos": 26322018929280.0,
      "grad_norm": 2.021318687641168,
      "language_loss": 0.86254489,
      "learning_rate": 3.857617531042398e-06,
      "loss": 0.88638014,
      "num_input_tokens_seen": 8375845,
      "step": 400,
      "time_per_iteration": 2.6626927852630615
    },
    {
      "auxiliary_loss_clip": 0.01299822,
      "auxiliary_loss_mlp": 0.01083301,
      "balance_loss_clip": 1.08346462,
      "balance_loss_mlp": 1.04687035,
      "epoch": 0.024109424319855705,
      "flos": 24425432910720.0,
      "grad_norm": 1.735822397657743,
      "language_loss": 0.79276752,
      "learning_rate": 3.8592251528102065e-06,
      "loss": 0.81659877,
      "num_input_tokens_seen": 8395240,
      "step": 401,
      "time_per_iteration": 2.68418025970459
    },
    {
      "auxiliary_loss_clip": 0.0129275,
      "auxiliary_loss_mlp": 0.01091389,
      "balance_loss_clip": 1.07852793,
      "balance_loss_mlp": 1.05493474,
      "epoch": 0.024169547572523674,
      "flos": 29604397610880.0,
      "grad_norm": 3.889755427752258,
      "language_loss": 0.78890866,
      "learning_rate": 3.8608287705345976e-06,
      "loss": 0.81274998,
      "num_input_tokens_seen": 8416950,
      "step": 402,
      "time_per_iteration": 2.7509379386901855
    },
    {
      "auxiliary_loss_clip": 0.01296434,
      "auxiliary_loss_mlp": 0.01082712,
      "balance_loss_clip": 1.07797897,
      "balance_loss_mlp": 1.04399323,
      "epoch": 0.024229670825191642,
      "flos": 22601458235520.0,
      "grad_norm": 2.49356632429363,
      "language_loss": 0.94936156,
      "learning_rate": 3.86242840411147e-06,
      "loss": 0.97315305,
      "num_input_tokens_seen": 8433660,
      "step": 403,
      "time_per_iteration": 2.5760560035705566
    },
    {
      "auxiliary_loss_clip": 0.0129994,
      "auxiliary_loss_mlp": 0.01091893,
      "balance_loss_clip": 1.07754242,
      "balance_loss_mlp": 1.05315053,
      "epoch": 0.02428979407785961,
      "flos": 18150258994560.0,
      "grad_norm": 2.361656575803209,
      "language_loss": 0.99877387,
      "learning_rate": 3.864024073288798e-06,
      "loss": 1.0226922,
      "num_input_tokens_seen": 8450180,
      "step": 404,
      "time_per_iteration": 2.5966458320617676
    },
    {
      "auxiliary_loss_clip": 0.01298911,
      "auxiliary_loss_mlp": 0.01100127,
      "balance_loss_clip": 1.08096266,
      "balance_loss_mlp": 1.06312442,
      "epoch": 0.024349917330527583,
      "flos": 15304984917120.0,
      "grad_norm": 2.3162348618509276,
      "language_loss": 0.8802169,
      "learning_rate": 3.865615797668091e-06,
      "loss": 0.90420723,
      "num_input_tokens_seen": 8467775,
      "step": 405,
      "time_per_iteration": 2.5728275775909424
    },
    {
      "auxiliary_loss_clip": 0.01306827,
      "auxiliary_loss_mlp": 0.01097881,
      "balance_loss_clip": 1.084512,
      "balance_loss_mlp": 1.06004393,
      "epoch": 0.024410040583195552,
      "flos": 20773892200320.0,
      "grad_norm": 2.7399607903318275,
      "language_loss": 0.93386561,
      "learning_rate": 3.867203596705844e-06,
      "loss": 0.95791268,
      "num_input_tokens_seen": 8486765,
      "step": 406,
      "time_per_iteration": 2.612668991088867
    },
    {
      "auxiliary_loss_clip": 0.01299426,
      "auxiliary_loss_mlp": 0.01088378,
      "balance_loss_clip": 1.08213782,
      "balance_loss_mlp": 1.0500164,
      "epoch": 0.02447016383586352,
      "flos": 21798854789760.0,
      "grad_norm": 2.1742012769968526,
      "language_loss": 0.87128031,
      "learning_rate": 3.86878748971496e-06,
      "loss": 0.89515841,
      "num_input_tokens_seen": 8506515,
      "step": 407,
      "time_per_iteration": 2.5982017517089844
    },
    {
      "auxiliary_loss_clip": 0.01298266,
      "auxiliary_loss_mlp": 0.01083858,
      "balance_loss_clip": 1.08472157,
      "balance_loss_mlp": 1.04630709,
      "epoch": 0.02453028708853149,
      "flos": 33948116380800.0,
      "grad_norm": 2.1458430439144234,
      "language_loss": 0.74102569,
      "learning_rate": 3.8703674958661596e-06,
      "loss": 0.76484692,
      "num_input_tokens_seen": 8528035,
      "step": 408,
      "time_per_iteration": 2.708670139312744
    },
    {
      "auxiliary_loss_clip": 0.01300128,
      "auxiliary_loss_mlp": 0.01089985,
      "balance_loss_clip": 1.08222318,
      "balance_loss_mlp": 1.05233896,
      "epoch": 0.024590410341199458,
      "flos": 21793000872960.0,
      "grad_norm": 2.4878473813549675,
      "language_loss": 0.92509401,
      "learning_rate": 3.871943634189376e-06,
      "loss": 0.94899511,
      "num_input_tokens_seen": 8546455,
      "step": 409,
      "time_per_iteration": 2.665321111679077
    },
    {
      "auxiliary_loss_clip": 0.01296394,
      "auxiliary_loss_mlp": 0.01077538,
      "balance_loss_clip": 1.08126342,
      "balance_loss_mlp": 1.04291987,
      "epoch": 0.02465053359386743,
      "flos": 35114782124160.0,
      "grad_norm": 2.2521095969191722,
      "language_loss": 0.82792604,
      "learning_rate": 3.873515923575128e-06,
      "loss": 0.85166532,
      "num_input_tokens_seen": 8568450,
      "step": 410,
      "time_per_iteration": 2.848928213119507
    },
    {
      "auxiliary_loss_clip": 0.01299459,
      "auxiliary_loss_mlp": 0.01089133,
      "balance_loss_clip": 1.08187068,
      "balance_loss_mlp": 1.05284572,
      "epoch": 0.0247106568465354,
      "flos": 27451409333760.0,
      "grad_norm": 2.1393760271628595,
      "language_loss": 0.77577484,
      "learning_rate": 3.875084382775879e-06,
      "loss": 0.79966074,
      "num_input_tokens_seen": 8589340,
      "step": 411,
      "time_per_iteration": 2.6645278930664062
    },
    {
      "auxiliary_loss_clip": 0.01298341,
      "auxiliary_loss_mlp": 0.0110154,
      "balance_loss_clip": 1.07977521,
      "balance_loss_mlp": 1.06289268,
      "epoch": 0.024770780099203367,
      "flos": 20703794808960.0,
      "grad_norm": 2.2974658872162665,
      "language_loss": 0.86379063,
      "learning_rate": 3.87664903040738e-06,
      "loss": 0.88778943,
      "num_input_tokens_seen": 8607150,
      "step": 412,
      "time_per_iteration": 2.6091151237487793
    },
    {
      "auxiliary_loss_clip": 0.01187014,
      "auxiliary_loss_mlp": 0.01031436,
      "balance_loss_clip": 1.07387948,
      "balance_loss_mlp": 1.02089787,
      "epoch": 0.024830903351871336,
      "flos": 69551859369600.0,
      "grad_norm": 0.8687159185244209,
      "language_loss": 0.5852263,
      "learning_rate": 3.878209884949994e-06,
      "loss": 0.60741079,
      "num_input_tokens_seen": 8669865,
      "step": 413,
      "time_per_iteration": 3.2269625663757324
    },
    {
      "auxiliary_loss_clip": 0.0129043,
      "auxiliary_loss_mlp": 0.01091958,
      "balance_loss_clip": 1.07709181,
      "balance_loss_mlp": 1.05249953,
      "epoch": 0.024891026604539304,
      "flos": 32270477713920.0,
      "grad_norm": 1.8280666153990437,
      "language_loss": 0.80517173,
      "learning_rate": 3.879766964750006e-06,
      "loss": 0.82899559,
      "num_input_tokens_seen": 8690235,
      "step": 414,
      "time_per_iteration": 2.720341444015503
    },
    {
      "auxiliary_loss_clip": 0.01287097,
      "auxiliary_loss_mlp": 0.0109242,
      "balance_loss_clip": 1.0756042,
      "balance_loss_mlp": 1.0556556,
      "epoch": 0.024951149857207276,
      "flos": 18840282238080.0,
      "grad_norm": 2.1921003994701302,
      "language_loss": 0.80227423,
      "learning_rate": 3.881320288020917e-06,
      "loss": 0.82606936,
      "num_input_tokens_seen": 8706295,
      "step": 415,
      "time_per_iteration": 2.6473400592803955
    },
    {
      "auxiliary_loss_clip": 0.01302694,
      "auxiliary_loss_mlp": 0.01082455,
      "balance_loss_clip": 1.08156919,
      "balance_loss_mlp": 1.04497528,
      "epoch": 0.025011273109875245,
      "flos": 15377201210880.0,
      "grad_norm": 2.9318871737289776,
      "language_loss": 0.96236515,
      "learning_rate": 3.882869872844723e-06,
      "loss": 0.9862166,
      "num_input_tokens_seen": 8724200,
      "step": 416,
      "time_per_iteration": 2.596189260482788
    },
    {
      "auxiliary_loss_clip": 0.01291636,
      "auxiliary_loss_mlp": 0.01074465,
      "balance_loss_clip": 1.07628798,
      "balance_loss_mlp": 1.0355792,
      "epoch": 0.025071396362543213,
      "flos": 18915515274240.0,
      "grad_norm": 1.741746736079687,
      "language_loss": 0.77381694,
      "learning_rate": 3.884415737173176e-06,
      "loss": 0.79747796,
      "num_input_tokens_seen": 8744170,
      "step": 417,
      "time_per_iteration": 5.610344171524048
    },
    {
      "auxiliary_loss_clip": 0.01290746,
      "auxiliary_loss_mlp": 0.0109022,
      "balance_loss_clip": 1.08072221,
      "balance_loss_mlp": 1.05264485,
      "epoch": 0.025131519615211182,
      "flos": 25337958952320.0,
      "grad_norm": 1.554385639735456,
      "language_loss": 0.77076226,
      "learning_rate": 3.8859578988290344e-06,
      "loss": 0.79457194,
      "num_input_tokens_seen": 8765120,
      "step": 418,
      "time_per_iteration": 5.837290525436401
    },
    {
      "auxiliary_loss_clip": 0.01297026,
      "auxiliary_loss_mlp": 0.01071197,
      "balance_loss_clip": 1.08019948,
      "balance_loss_mlp": 1.03550553,
      "epoch": 0.02519164286787915,
      "flos": 18953149749120.0,
      "grad_norm": 2.4603268634516207,
      "language_loss": 0.81445098,
      "learning_rate": 3.887496375507294e-06,
      "loss": 0.83813322,
      "num_input_tokens_seen": 8783500,
      "step": 419,
      "time_per_iteration": 2.582590341567993
    },
    {
      "auxiliary_loss_clip": 0.01291114,
      "auxiliary_loss_mlp": 0.01086736,
      "balance_loss_clip": 1.07929599,
      "balance_loss_mlp": 1.04708743,
      "epoch": 0.025251766120547123,
      "flos": 17421092904960.0,
      "grad_norm": 1.8078532084212713,
      "language_loss": 0.73618573,
      "learning_rate": 3.8890311847764065e-06,
      "loss": 0.75996423,
      "num_input_tokens_seen": 8801175,
      "step": 420,
      "time_per_iteration": 2.6739418506622314
    },
    {
      "auxiliary_loss_clip": 0.01290485,
      "auxiliary_loss_mlp": 0.01096292,
      "balance_loss_clip": 1.07605243,
      "balance_loss_mlp": 1.05924153,
      "epoch": 0.02531188937321509,
      "flos": 25045430590080.0,
      "grad_norm": 1.77336014903074,
      "language_loss": 0.79040134,
      "learning_rate": 3.890562344079484e-06,
      "loss": 0.81426907,
      "num_input_tokens_seen": 8820215,
      "step": 421,
      "time_per_iteration": 2.6928632259368896
    },
    {
      "auxiliary_loss_clip": 0.01290689,
      "auxiliary_loss_mlp": 0.01088863,
      "balance_loss_clip": 1.07922924,
      "balance_loss_mlp": 1.04983425,
      "epoch": 0.02537201262588306,
      "flos": 30592228515840.0,
      "grad_norm": 2.2139016136437104,
      "language_loss": 0.8203755,
      "learning_rate": 3.89208987073549e-06,
      "loss": 0.84417105,
      "num_input_tokens_seen": 8839660,
      "step": 422,
      "time_per_iteration": 2.714707851409912
    },
    {
      "auxiliary_loss_clip": 0.01293659,
      "auxiliary_loss_mlp": 0.01078975,
      "balance_loss_clip": 1.07677865,
      "balance_loss_mlp": 1.04430926,
      "epoch": 0.02543213587855103,
      "flos": 26065365275520.0,
      "grad_norm": 2.1259138778576356,
      "language_loss": 0.83458018,
      "learning_rate": 3.893613781940409e-06,
      "loss": 0.85830647,
      "num_input_tokens_seen": 8859280,
      "step": 423,
      "time_per_iteration": 2.652757167816162
    },
    {
      "auxiliary_loss_clip": 0.01287497,
      "auxiliary_loss_mlp": 0.01078335,
      "balance_loss_clip": 1.0742569,
      "balance_loss_mlp": 1.04221487,
      "epoch": 0.025492259131218997,
      "flos": 36022818965760.0,
      "grad_norm": 2.012741083661608,
      "language_loss": 0.74129444,
      "learning_rate": 3.895134094768415e-06,
      "loss": 0.76495278,
      "num_input_tokens_seen": 8880560,
      "step": 424,
      "time_per_iteration": 2.7724521160125732
    },
    {
      "auxiliary_loss_clip": 0.01296446,
      "auxiliary_loss_mlp": 0.01093799,
      "balance_loss_clip": 1.07987142,
      "balance_loss_mlp": 1.05782199,
      "epoch": 0.02555238238388697,
      "flos": 18588045957120.0,
      "grad_norm": 4.623670538116741,
      "language_loss": 0.83193713,
      "learning_rate": 3.896650826173015e-06,
      "loss": 0.85583955,
      "num_input_tokens_seen": 8899155,
      "step": 425,
      "time_per_iteration": 2.608029842376709
    },
    {
      "auxiliary_loss_clip": 0.01292462,
      "auxiliary_loss_mlp": 0.01092376,
      "balance_loss_clip": 1.07259536,
      "balance_loss_mlp": 1.0544672,
      "epoch": 0.025612505636554938,
      "flos": 24243186280320.0,
      "grad_norm": 2.5075767706443566,
      "language_loss": 0.853073,
      "learning_rate": 3.898163992988186e-06,
      "loss": 0.87692136,
      "num_input_tokens_seen": 8917890,
      "step": 426,
      "time_per_iteration": 2.6445271968841553
    },
    {
      "auxiliary_loss_clip": 0.01175923,
      "auxiliary_loss_mlp": 0.01017688,
      "balance_loss_clip": 1.06532824,
      "balance_loss_mlp": 1.00781715,
      "epoch": 0.025672628889222907,
      "flos": 60586941265920.0,
      "grad_norm": 0.8949637292547264,
      "language_loss": 0.57219732,
      "learning_rate": 3.899673611929491e-06,
      "loss": 0.5941335,
      "num_input_tokens_seen": 8978260,
      "step": 427,
      "time_per_iteration": 3.2690517902374268
    },
    {
      "auxiliary_loss_clip": 0.01291989,
      "auxiliary_loss_mlp": 0.01092649,
      "balance_loss_clip": 1.08155811,
      "balance_loss_mlp": 1.05674267,
      "epoch": 0.025732752141890875,
      "flos": 19573255169280.0,
      "grad_norm": 2.4869215225306673,
      "language_loss": 0.88130605,
      "learning_rate": 3.901179699595194e-06,
      "loss": 0.90515244,
      "num_input_tokens_seen": 8994460,
      "step": 428,
      "time_per_iteration": 2.6143813133239746
    },
    {
      "auxiliary_loss_clip": 0.01283603,
      "auxiliary_loss_mlp": 0.0107531,
      "balance_loss_clip": 1.07418942,
      "balance_loss_mlp": 1.03735399,
      "epoch": 0.025792875394558847,
      "flos": 31284262920960.0,
      "grad_norm": 2.067247304638145,
      "language_loss": 0.85790849,
      "learning_rate": 3.902682272467353e-06,
      "loss": 0.88149762,
      "num_input_tokens_seen": 9016670,
      "step": 429,
      "time_per_iteration": 2.749328374862671
    },
    {
      "auxiliary_loss_clip": 0.01288943,
      "auxiliary_loss_mlp": 0.01083888,
      "balance_loss_clip": 1.07337689,
      "balance_loss_mlp": 1.04590786,
      "epoch": 0.025852998647226816,
      "flos": 32379610210560.0,
      "grad_norm": 2.4411876712444034,
      "language_loss": 0.8815223,
      "learning_rate": 3.904181346912895e-06,
      "loss": 0.90525061,
      "num_input_tokens_seen": 9039720,
      "step": 430,
      "time_per_iteration": 2.7483572959899902
    },
    {
      "auxiliary_loss_clip": 0.01290726,
      "auxiliary_loss_mlp": 0.01080495,
      "balance_loss_clip": 1.0803287,
      "balance_loss_mlp": 1.04573333,
      "epoch": 0.025913121899894784,
      "flos": 20193288762240.0,
      "grad_norm": 2.086180078538185,
      "language_loss": 0.84249514,
      "learning_rate": 3.905676939184698e-06,
      "loss": 0.8662073,
      "num_input_tokens_seen": 9059850,
      "step": 431,
      "time_per_iteration": 2.6531126499176025
    },
    {
      "auxiliary_loss_clip": 0.01286945,
      "auxiliary_loss_mlp": 0.01073345,
      "balance_loss_clip": 1.07570636,
      "balance_loss_mlp": 1.03951311,
      "epoch": 0.025973245152562753,
      "flos": 14720430983040.0,
      "grad_norm": 2.681931959502968,
      "language_loss": 0.86511916,
      "learning_rate": 3.907169065422638e-06,
      "loss": 0.88872206,
      "num_input_tokens_seen": 9077590,
      "step": 432,
      "time_per_iteration": 2.7582762241363525
    },
    {
      "auxiliary_loss_clip": 0.01287429,
      "auxiliary_loss_mlp": 0.01072961,
      "balance_loss_clip": 1.07632601,
      "balance_loss_mlp": 1.03891492,
      "epoch": 0.02603336840523072,
      "flos": 30992991534720.0,
      "grad_norm": 1.95596969308187,
      "language_loss": 0.76036298,
      "learning_rate": 3.908657741654636e-06,
      "loss": 0.7839669,
      "num_input_tokens_seen": 9099880,
      "step": 433,
      "time_per_iteration": 2.707771062850952
    },
    {
      "auxiliary_loss_clip": 0.01289436,
      "auxiliary_loss_mlp": 0.01088504,
      "balance_loss_clip": 1.07470191,
      "balance_loss_mlp": 1.04973757,
      "epoch": 0.026093491657898694,
      "flos": 17674262939520.0,
      "grad_norm": 2.157056093147959,
      "language_loss": 0.8979522,
      "learning_rate": 3.910142983797699e-06,
      "loss": 0.92173159,
      "num_input_tokens_seen": 9118620,
      "step": 434,
      "time_per_iteration": 2.5665409564971924
    },
    {
      "auxiliary_loss_clip": 0.01289617,
      "auxiliary_loss_mlp": 0.01096405,
      "balance_loss_clip": 1.07960439,
      "balance_loss_mlp": 1.05904448,
      "epoch": 0.026153614910566662,
      "flos": 17857874286720.0,
      "grad_norm": 2.306071945033866,
      "language_loss": 0.80187833,
      "learning_rate": 3.9116248076589305e-06,
      "loss": 0.82573849,
      "num_input_tokens_seen": 9135655,
      "step": 435,
      "time_per_iteration": 2.614440679550171
    },
    {
      "auxiliary_loss_clip": 0.01285396,
      "auxiliary_loss_mlp": 0.01092207,
      "balance_loss_clip": 1.07367229,
      "balance_loss_mlp": 1.05503798,
      "epoch": 0.02621373816323463,
      "flos": 20011113959040.0,
      "grad_norm": 3.0257040949539356,
      "language_loss": 0.86361396,
      "learning_rate": 3.913103228936546e-06,
      "loss": 0.88739002,
      "num_input_tokens_seen": 9153520,
      "step": 436,
      "time_per_iteration": 2.635033130645752
    },
    {
      "auxiliary_loss_clip": 0.01289558,
      "auxiliary_loss_mlp": 0.01096903,
      "balance_loss_clip": 1.07716811,
      "balance_loss_mlp": 1.06080687,
      "epoch": 0.0262738614159026,
      "flos": 19281193683840.0,
      "grad_norm": 2.4233286399217993,
      "language_loss": 0.74725163,
      "learning_rate": 3.914578263220868e-06,
      "loss": 0.77111626,
      "num_input_tokens_seen": 9170750,
      "step": 437,
      "time_per_iteration": 2.6614880561828613
    },
    {
      "auxiliary_loss_clip": 0.01286403,
      "auxiliary_loss_mlp": 0.01100399,
      "balance_loss_clip": 1.07628679,
      "balance_loss_mlp": 1.06220388,
      "epoch": 0.026333984668570568,
      "flos": 18807208790400.0,
      "grad_norm": 2.79370908187484,
      "language_loss": 0.9131338,
      "learning_rate": 3.916049925995316e-06,
      "loss": 0.93700182,
      "num_input_tokens_seen": 9188430,
      "step": 438,
      "time_per_iteration": 2.674877166748047
    },
    {
      "auxiliary_loss_clip": 0.01169678,
      "auxiliary_loss_mlp": 0.01072518,
      "balance_loss_clip": 1.0602653,
      "balance_loss_mlp": 1.06250465,
      "epoch": 0.02639410792123854,
      "flos": 64572020691840.0,
      "grad_norm": 0.8871275810137318,
      "language_loss": 0.62631273,
      "learning_rate": 3.917518232637377e-06,
      "loss": 0.64873469,
      "num_input_tokens_seen": 9255835,
      "step": 439,
      "time_per_iteration": 3.2527849674224854
    },
    {
      "auxiliary_loss_clip": 0.01296492,
      "auxiliary_loss_mlp": 0.01095184,
      "balance_loss_clip": 1.08175814,
      "balance_loss_mlp": 1.05758572,
      "epoch": 0.02645423117390651,
      "flos": 28473462921600.0,
      "grad_norm": 3.31985956061953,
      "language_loss": 0.75982475,
      "learning_rate": 3.918983198419573e-06,
      "loss": 0.78374153,
      "num_input_tokens_seen": 9276835,
      "step": 440,
      "time_per_iteration": 2.6770262718200684
    },
    {
      "auxiliary_loss_clip": 0.01286342,
      "auxiliary_loss_mlp": 0.01076505,
      "balance_loss_clip": 1.07652593,
      "balance_loss_mlp": 1.04048026,
      "epoch": 0.026514354426574478,
      "flos": 18551237495040.0,
      "grad_norm": 3.0236705091068283,
      "language_loss": 0.83197021,
      "learning_rate": 3.920444838510415e-06,
      "loss": 0.85559869,
      "num_input_tokens_seen": 9295075,
      "step": 441,
      "time_per_iteration": 2.591306209564209
    },
    {
      "auxiliary_loss_clip": 0.01291817,
      "auxiliary_loss_mlp": 0.01086154,
      "balance_loss_clip": 1.07703269,
      "balance_loss_mlp": 1.04829359,
      "epoch": 0.026574477679242446,
      "flos": 20667812359680.0,
      "grad_norm": 2.202684635319811,
      "language_loss": 0.78490162,
      "learning_rate": 3.92190316797534e-06,
      "loss": 0.80868137,
      "num_input_tokens_seen": 9314205,
      "step": 442,
      "time_per_iteration": 2.633054733276367
    },
    {
      "auxiliary_loss_clip": 0.0116251,
      "auxiliary_loss_mlp": 0.01015158,
      "balance_loss_clip": 1.05336332,
      "balance_loss_mlp": 1.0054301,
      "epoch": 0.026634600931910415,
      "flos": 57956125340160.0,
      "grad_norm": 0.9609264438471399,
      "language_loss": 0.64459753,
      "learning_rate": 3.92335820177765e-06,
      "loss": 0.66637421,
      "num_input_tokens_seen": 9367395,
      "step": 443,
      "time_per_iteration": 3.1241400241851807
    },
    {
      "auxiliary_loss_clip": 0.01291897,
      "auxiliary_loss_mlp": 0.01085882,
      "balance_loss_clip": 1.08147204,
      "balance_loss_mlp": 1.04906964,
      "epoch": 0.026694724184578387,
      "flos": 15815131827840.0,
      "grad_norm": 2.121488874389134,
      "language_loss": 0.82093638,
      "learning_rate": 3.924809954779425e-06,
      "loss": 0.84471416,
      "num_input_tokens_seen": 9385185,
      "step": 444,
      "time_per_iteration": 2.6202428340911865
    },
    {
      "auxiliary_loss_clip": 0.0129406,
      "auxiliary_loss_mlp": 0.01082041,
      "balance_loss_clip": 1.07940578,
      "balance_loss_mlp": 1.04263067,
      "epoch": 0.026754847437246355,
      "flos": 23440259612160.0,
      "grad_norm": 2.2213674770888607,
      "language_loss": 0.95689106,
      "learning_rate": 3.9262584417424425e-06,
      "loss": 0.98065209,
      "num_input_tokens_seen": 9403225,
      "step": 445,
      "time_per_iteration": 2.6071228981018066
    },
    {
      "auxiliary_loss_clip": 0.01289866,
      "auxiliary_loss_mlp": 0.01094053,
      "balance_loss_clip": 1.07953668,
      "balance_loss_mlp": 1.05492878,
      "epoch": 0.026814970689914324,
      "flos": 17341801632000.0,
      "grad_norm": 2.775359545549618,
      "language_loss": 0.91932094,
      "learning_rate": 3.9277036773290725e-06,
      "loss": 0.94316012,
      "num_input_tokens_seen": 9420540,
      "step": 446,
      "time_per_iteration": 2.5791916847229004
    },
    {
      "auxiliary_loss_clip": 0.01289847,
      "auxiliary_loss_mlp": 0.01088114,
      "balance_loss_clip": 1.08072042,
      "balance_loss_mlp": 1.05092025,
      "epoch": 0.026875093942582293,
      "flos": 17894718662400.0,
      "grad_norm": 2.0562763127679204,
      "language_loss": 0.79831308,
      "learning_rate": 3.92914567610317e-06,
      "loss": 0.82209271,
      "num_input_tokens_seen": 9438840,
      "step": 447,
      "time_per_iteration": 2.6420843601226807
    },
    {
      "auxiliary_loss_clip": 0.01289397,
      "auxiliary_loss_mlp": 0.01079607,
      "balance_loss_clip": 1.07901013,
      "balance_loss_mlp": 1.04446411,
      "epoch": 0.026935217195250265,
      "flos": 21723980889600.0,
      "grad_norm": 2.231264914467203,
      "language_loss": 0.86402845,
      "learning_rate": 3.930584452530952e-06,
      "loss": 0.8877185,
      "num_input_tokens_seen": 9457215,
      "step": 448,
      "time_per_iteration": 2.590277910232544
    },
    {
      "auxiliary_loss_clip": 0.01282455,
      "auxiliary_loss_mlp": 0.01091099,
      "balance_loss_clip": 1.07706833,
      "balance_loss_mlp": 1.05662322,
      "epoch": 0.026995340447918233,
      "flos": 23622685810560.0,
      "grad_norm": 1.941778256808524,
      "language_loss": 0.88581634,
      "learning_rate": 3.9320200209818755e-06,
      "loss": 0.90955186,
      "num_input_tokens_seen": 9475615,
      "step": 449,
      "time_per_iteration": 2.610065460205078
    },
    {
      "auxiliary_loss_clip": 0.01293472,
      "auxiliary_loss_mlp": 0.01085576,
      "balance_loss_clip": 1.07856452,
      "balance_loss_mlp": 1.04814398,
      "epoch": 0.027055463700586202,
      "flos": 17931275729280.0,
      "grad_norm": 2.199007921978797,
      "language_loss": 0.80395782,
      "learning_rate": 3.933452395729493e-06,
      "loss": 0.8277483,
      "num_input_tokens_seen": 9493975,
      "step": 450,
      "time_per_iteration": 2.637465238571167
    },
    {
      "auxiliary_loss_clip": 0.01284612,
      "auxiliary_loss_mlp": 0.0108001,
      "balance_loss_clip": 1.08025336,
      "balance_loss_mlp": 1.04384232,
      "epoch": 0.02711558695325417,
      "flos": 25118903859840.0,
      "grad_norm": 1.599374223212879,
      "language_loss": 0.81562543,
      "learning_rate": 3.934881590952304e-06,
      "loss": 0.83927161,
      "num_input_tokens_seen": 9514810,
      "step": 451,
      "time_per_iteration": 2.6506927013397217
    },
    {
      "auxiliary_loss_clip": 0.0128567,
      "auxiliary_loss_mlp": 0.01090719,
      "balance_loss_clip": 1.08126068,
      "balance_loss_mlp": 1.0533824,
      "epoch": 0.02717571020592214,
      "flos": 24239559006720.0,
      "grad_norm": 1.9677929562692107,
      "language_loss": 0.77019048,
      "learning_rate": 3.936307620734599e-06,
      "loss": 0.79395437,
      "num_input_tokens_seen": 9533635,
      "step": 452,
      "time_per_iteration": 2.5751442909240723
    },
    {
      "auxiliary_loss_clip": 0.01286865,
      "auxiliary_loss_mlp": 0.01088287,
      "balance_loss_clip": 1.08011293,
      "balance_loss_mlp": 1.05135596,
      "epoch": 0.02723583345859011,
      "flos": 25118939773440.0,
      "grad_norm": 1.7205362750177517,
      "language_loss": 0.72874546,
      "learning_rate": 3.937730499067294e-06,
      "loss": 0.75249696,
      "num_input_tokens_seen": 9555420,
      "step": 453,
      "time_per_iteration": 2.668083667755127
    },
    {
      "auxiliary_loss_clip": 0.01281405,
      "auxiliary_loss_mlp": 0.01083223,
      "balance_loss_clip": 1.07714963,
      "balance_loss_mlp": 1.04748416,
      "epoch": 0.02729595671125808,
      "flos": 42741597847680.0,
      "grad_norm": 1.8353680194819204,
      "language_loss": 0.82419729,
      "learning_rate": 3.939150239848748e-06,
      "loss": 0.84784359,
      "num_input_tokens_seen": 9578950,
      "step": 454,
      "time_per_iteration": 2.8580126762390137
    },
    {
      "auxiliary_loss_clip": 0.01285525,
      "auxiliary_loss_mlp": 0.01077241,
      "balance_loss_clip": 1.07935429,
      "balance_loss_mlp": 1.043648,
      "epoch": 0.02735607996392605,
      "flos": 21430985650560.0,
      "grad_norm": 1.985829769195046,
      "language_loss": 0.75404847,
      "learning_rate": 3.9405668568855866e-06,
      "loss": 0.77767611,
      "num_input_tokens_seen": 9598160,
      "step": 455,
      "time_per_iteration": 2.6593477725982666
    },
    {
      "auxiliary_loss_clip": 0.01282853,
      "auxiliary_loss_mlp": 0.01094959,
      "balance_loss_clip": 1.07477236,
      "balance_loss_mlp": 1.0597918,
      "epoch": 0.027416203216594017,
      "flos": 20851280052480.0,
      "grad_norm": 1.92483069519606,
      "language_loss": 0.80670613,
      "learning_rate": 3.941980363893499e-06,
      "loss": 0.83048427,
      "num_input_tokens_seen": 9616010,
      "step": 456,
      "time_per_iteration": 2.6798384189605713
    },
    {
      "auxiliary_loss_clip": 0.01280135,
      "auxiliary_loss_mlp": 0.01080319,
      "balance_loss_clip": 1.07714963,
      "balance_loss_mlp": 1.0435549,
      "epoch": 0.027476326469261986,
      "flos": 13224500242560.0,
      "grad_norm": 2.171481572134165,
      "language_loss": 0.81587321,
      "learning_rate": 3.9433907744980384e-06,
      "loss": 0.83947778,
      "num_input_tokens_seen": 9634000,
      "step": 457,
      "time_per_iteration": 5.62308406829834
    },
    {
      "auxiliary_loss_clip": 0.01283922,
      "auxiliary_loss_mlp": 0.01084055,
      "balance_loss_clip": 1.07603848,
      "balance_loss_mlp": 1.04891229,
      "epoch": 0.027536449721929958,
      "flos": 24024526237440.0,
      "grad_norm": 2.024184269172234,
      "language_loss": 0.94030929,
      "learning_rate": 3.944798102235412e-06,
      "loss": 0.96398914,
      "num_input_tokens_seen": 9653455,
      "step": 458,
      "time_per_iteration": 5.694372653961182
    },
    {
      "auxiliary_loss_clip": 0.01280807,
      "auxiliary_loss_mlp": 0.01091426,
      "balance_loss_clip": 1.07479525,
      "balance_loss_mlp": 1.05666471,
      "epoch": 0.027596572974597926,
      "flos": 13006055681280.0,
      "grad_norm": 2.356061876390436,
      "language_loss": 0.79279089,
      "learning_rate": 3.9462023605532545e-06,
      "loss": 0.81651318,
      "num_input_tokens_seen": 9669650,
      "step": 459,
      "time_per_iteration": 2.626948595046997
    },
    {
      "auxiliary_loss_clip": 0.01286253,
      "auxiliary_loss_mlp": 0.01081623,
      "balance_loss_clip": 1.08119941,
      "balance_loss_mlp": 1.04278445,
      "epoch": 0.027656696227265895,
      "flos": 26143076350080.0,
      "grad_norm": 2.0583603779546404,
      "language_loss": 0.83362132,
      "learning_rate": 3.947603562811407e-06,
      "loss": 0.85730016,
      "num_input_tokens_seen": 9691415,
      "step": 460,
      "time_per_iteration": 2.7191598415374756
    },
    {
      "auxiliary_loss_clip": 0.01158037,
      "auxiliary_loss_mlp": 0.01054463,
      "balance_loss_clip": 1.05032754,
      "balance_loss_mlp": 1.044402,
      "epoch": 0.027716819479933864,
      "flos": 60697222997760.0,
      "grad_norm": 1.612511499168885,
      "language_loss": 0.7351321,
      "learning_rate": 3.949001722282675e-06,
      "loss": 0.7572571,
      "num_input_tokens_seen": 9755605,
      "step": 461,
      "time_per_iteration": 3.210820436477661
    },
    {
      "auxiliary_loss_clip": 0.01284234,
      "auxiliary_loss_mlp": 0.01079832,
      "balance_loss_clip": 1.08432341,
      "balance_loss_mlp": 1.04700136,
      "epoch": 0.027776942732601832,
      "flos": 31211938886400.0,
      "grad_norm": 2.4500038571081073,
      "language_loss": 0.81596625,
      "learning_rate": 3.950396852153582e-06,
      "loss": 0.839607,
      "num_input_tokens_seen": 9776270,
      "step": 462,
      "time_per_iteration": 2.683197021484375
    },
    {
      "auxiliary_loss_clip": 0.01280414,
      "auxiliary_loss_mlp": 0.0107864,
      "balance_loss_clip": 1.07752454,
      "balance_loss_mlp": 1.0454762,
      "epoch": 0.027837065985269804,
      "flos": 22674644196480.0,
      "grad_norm": 2.258526594266715,
      "language_loss": 0.90062451,
      "learning_rate": 3.951788965525118e-06,
      "loss": 0.92421508,
      "num_input_tokens_seen": 9794465,
      "step": 463,
      "time_per_iteration": 2.641674757003784
    },
    {
      "auxiliary_loss_clip": 0.01151842,
      "auxiliary_loss_mlp": 0.01010002,
      "balance_loss_clip": 1.04755902,
      "balance_loss_mlp": 1.00027454,
      "epoch": 0.027897189237937773,
      "flos": 62182487399040.0,
      "grad_norm": 0.8962796480673014,
      "language_loss": 0.59058654,
      "learning_rate": 3.953178075413476e-06,
      "loss": 0.61220491,
      "num_input_tokens_seen": 9849685,
      "step": 464,
      "time_per_iteration": 3.1129612922668457
    },
    {
      "auxiliary_loss_clip": 0.01292933,
      "auxiliary_loss_mlp": 0.01100533,
      "balance_loss_clip": 1.08296049,
      "balance_loss_mlp": 1.06412649,
      "epoch": 0.02795731249060574,
      "flos": 24493160004480.0,
      "grad_norm": 2.3712654859298055,
      "language_loss": 0.81454253,
      "learning_rate": 3.954564194750784e-06,
      "loss": 0.83847719,
      "num_input_tokens_seen": 9869505,
      "step": 465,
      "time_per_iteration": 2.723144769668579
    },
    {
      "auxiliary_loss_clip": 0.01279938,
      "auxiliary_loss_mlp": 0.01092668,
      "balance_loss_clip": 1.07546401,
      "balance_loss_mlp": 1.05630863,
      "epoch": 0.02801743574327371,
      "flos": 23733003456000.0,
      "grad_norm": 1.9968224423519798,
      "language_loss": 0.78396618,
      "learning_rate": 3.955947336385828e-06,
      "loss": 0.80769229,
      "num_input_tokens_seen": 9890950,
      "step": 466,
      "time_per_iteration": 2.6278555393218994
    },
    {
      "auxiliary_loss_clip": 0.0127853,
      "auxiliary_loss_mlp": 0.01091802,
      "balance_loss_clip": 1.07703936,
      "balance_loss_mlp": 1.05661178,
      "epoch": 0.02807755899594168,
      "flos": 20629100476800.0,
      "grad_norm": 2.010021605622182,
      "language_loss": 0.87699366,
      "learning_rate": 3.957327513084761e-06,
      "loss": 0.90069699,
      "num_input_tokens_seen": 9911265,
      "step": 467,
      "time_per_iteration": 2.6687490940093994
    },
    {
      "auxiliary_loss_clip": 0.01285129,
      "auxiliary_loss_mlp": 0.01112935,
      "balance_loss_clip": 1.07874036,
      "balance_loss_mlp": 1.07576585,
      "epoch": 0.02813768224860965,
      "flos": 19244564789760.0,
      "grad_norm": 2.2302958424490416,
      "language_loss": 0.86091757,
      "learning_rate": 3.958704737531818e-06,
      "loss": 0.88489819,
      "num_input_tokens_seen": 9929025,
      "step": 468,
      "time_per_iteration": 2.5745644569396973
    },
    {
      "auxiliary_loss_clip": 0.01281128,
      "auxiliary_loss_mlp": 0.01085455,
      "balance_loss_clip": 1.07529211,
      "balance_loss_mlp": 1.04857147,
      "epoch": 0.02819780550127762,
      "flos": 20813968800000.0,
      "grad_norm": 2.1866562002509875,
      "language_loss": 0.91690558,
      "learning_rate": 3.9600790223300065e-06,
      "loss": 0.94057143,
      "num_input_tokens_seen": 9945190,
      "step": 469,
      "time_per_iteration": 2.610821008682251
    },
    {
      "auxiliary_loss_clip": 0.0127909,
      "auxiliary_loss_mlp": 0.0110095,
      "balance_loss_clip": 1.07675052,
      "balance_loss_mlp": 1.06482995,
      "epoch": 0.028257928753945588,
      "flos": 19974125928960.0,
      "grad_norm": 2.674428223667968,
      "language_loss": 0.81758964,
      "learning_rate": 3.96145038000181e-06,
      "loss": 0.84139001,
      "num_input_tokens_seen": 9962820,
      "step": 470,
      "time_per_iteration": 2.6004326343536377
    },
    {
      "auxiliary_loss_clip": 0.0128074,
      "auxiliary_loss_mlp": 0.01086643,
      "balance_loss_clip": 1.07482624,
      "balance_loss_mlp": 1.04947352,
      "epoch": 0.028318052006613557,
      "flos": 20484488321280.0,
      "grad_norm": 1.788793606991614,
      "language_loss": 0.93071401,
      "learning_rate": 3.962818822989861e-06,
      "loss": 0.95438784,
      "num_input_tokens_seen": 9982595,
      "step": 471,
      "time_per_iteration": 2.556288719177246
    },
    {
      "auxiliary_loss_clip": 0.01273697,
      "auxiliary_loss_mlp": 0.0110454,
      "balance_loss_clip": 1.07223165,
      "balance_loss_mlp": 1.06884849,
      "epoch": 0.02837817525928153,
      "flos": 28514832410880.0,
      "grad_norm": 1.8550872135639116,
      "language_loss": 0.7613501,
      "learning_rate": 3.964184363657625e-06,
      "loss": 0.78513247,
      "num_input_tokens_seen": 10004645,
      "step": 472,
      "time_per_iteration": 2.667804002761841
    },
    {
      "auxiliary_loss_clip": 0.01280341,
      "auxiliary_loss_mlp": 0.01090649,
      "balance_loss_clip": 1.07279634,
      "balance_loss_mlp": 1.05624473,
      "epoch": 0.028438298511949497,
      "flos": 18551668458240.0,
      "grad_norm": 1.9914661475951314,
      "language_loss": 0.93097353,
      "learning_rate": 3.965547014290071e-06,
      "loss": 0.95468336,
      "num_input_tokens_seen": 10022555,
      "step": 473,
      "time_per_iteration": 2.6402342319488525
    },
    {
      "auxiliary_loss_clip": 0.01287339,
      "auxiliary_loss_mlp": 0.01124194,
      "balance_loss_clip": 1.07773685,
      "balance_loss_mlp": 1.08979011,
      "epoch": 0.028498421764617466,
      "flos": 16910227722240.0,
      "grad_norm": 3.2560638787193237,
      "language_loss": 0.88488632,
      "learning_rate": 3.96690678709433e-06,
      "loss": 0.90900171,
      "num_input_tokens_seen": 10041025,
      "step": 474,
      "time_per_iteration": 2.5853888988494873
    },
    {
      "auxiliary_loss_clip": 0.0127783,
      "auxiliary_loss_mlp": 0.01093132,
      "balance_loss_clip": 1.07535374,
      "balance_loss_mlp": 1.05620146,
      "epoch": 0.028558545017285435,
      "flos": 27778699082880.0,
      "grad_norm": 3.1427023167402006,
      "language_loss": 0.78901398,
      "learning_rate": 3.968263694200355e-06,
      "loss": 0.81272364,
      "num_input_tokens_seen": 10060775,
      "step": 475,
      "time_per_iteration": 2.654519557952881
    },
    {
      "auxiliary_loss_clip": 0.01148107,
      "auxiliary_loss_mlp": 0.01095224,
      "balance_loss_clip": 1.04505777,
      "balance_loss_mlp": 1.08583021,
      "epoch": 0.028618668269953403,
      "flos": 65654367258240.0,
      "grad_norm": 0.9280065830162254,
      "language_loss": 0.66926932,
      "learning_rate": 3.969617747661569e-06,
      "loss": 0.6917026,
      "num_input_tokens_seen": 10120225,
      "step": 476,
      "time_per_iteration": 3.1292569637298584
    },
    {
      "auxiliary_loss_clip": 0.01279748,
      "auxiliary_loss_mlp": 0.01088794,
      "balance_loss_clip": 1.07638311,
      "balance_loss_mlp": 1.05188656,
      "epoch": 0.028678791522621375,
      "flos": 21937074324480.0,
      "grad_norm": 2.985672001195028,
      "language_loss": 0.83807188,
      "learning_rate": 3.970968959455509e-06,
      "loss": 0.86175728,
      "num_input_tokens_seen": 10137880,
      "step": 477,
      "time_per_iteration": 2.651493549346924
    },
    {
      "auxiliary_loss_clip": 0.01284956,
      "auxiliary_loss_mlp": 0.0108711,
      "balance_loss_clip": 1.07924342,
      "balance_loss_mlp": 1.05089426,
      "epoch": 0.028738914775289344,
      "flos": 24572128055040.0,
      "grad_norm": 2.1929055744411943,
      "language_loss": 0.8233152,
      "learning_rate": 3.97231734148446e-06,
      "loss": 0.84703588,
      "num_input_tokens_seen": 10156930,
      "step": 478,
      "time_per_iteration": 2.6986753940582275
    },
    {
      "auxiliary_loss_clip": 0.01277687,
      "auxiliary_loss_mlp": 0.01080644,
      "balance_loss_clip": 1.07448888,
      "balance_loss_mlp": 1.04500043,
      "epoch": 0.028799038027957313,
      "flos": 23257977068160.0,
      "grad_norm": 4.057107988717453,
      "language_loss": 0.81195259,
      "learning_rate": 3.973662905576082e-06,
      "loss": 0.83553594,
      "num_input_tokens_seen": 10176295,
      "step": 479,
      "time_per_iteration": 2.6321041584014893
    },
    {
      "auxiliary_loss_clip": 0.01273765,
      "auxiliary_loss_mlp": 0.01083313,
      "balance_loss_clip": 1.07335579,
      "balance_loss_mlp": 1.04552341,
      "epoch": 0.02885916128062528,
      "flos": 22164102236160.0,
      "grad_norm": 2.352225573775279,
      "language_loss": 0.7335608,
      "learning_rate": 3.975005663484038e-06,
      "loss": 0.75713164,
      "num_input_tokens_seen": 10195790,
      "step": 480,
      "time_per_iteration": 2.650696277618408
    },
    {
      "auxiliary_loss_clip": 0.01273107,
      "auxiliary_loss_mlp": 0.01075586,
      "balance_loss_clip": 1.07424879,
      "balance_loss_mlp": 1.04277968,
      "epoch": 0.02891928453329325,
      "flos": 22932842135040.0,
      "grad_norm": 1.867890428108999,
      "language_loss": 0.87560165,
      "learning_rate": 3.976345626888605e-06,
      "loss": 0.89908862,
      "num_input_tokens_seen": 10218405,
      "step": 481,
      "time_per_iteration": 2.6585533618927
    },
    {
      "auxiliary_loss_clip": 0.01142103,
      "auxiliary_loss_mlp": 0.01017301,
      "balance_loss_clip": 1.04286921,
      "balance_loss_mlp": 1.00895679,
      "epoch": 0.028979407785961222,
      "flos": 57432941792640.0,
      "grad_norm": 0.8486437303263991,
      "language_loss": 0.66030192,
      "learning_rate": 3.9776828073972864e-06,
      "loss": 0.68189597,
      "num_input_tokens_seen": 10271005,
      "step": 482,
      "time_per_iteration": 2.9788918495178223
    },
    {
      "auxiliary_loss_clip": 0.01287904,
      "auxiliary_loss_mlp": 0.01082416,
      "balance_loss_clip": 1.07739437,
      "balance_loss_mlp": 1.04868007,
      "epoch": 0.02903953103862919,
      "flos": 16722737706240.0,
      "grad_norm": 2.6473263724689873,
      "language_loss": 0.7899214,
      "learning_rate": 3.979017216545415e-06,
      "loss": 0.81362462,
      "num_input_tokens_seen": 10288405,
      "step": 483,
      "time_per_iteration": 2.5642752647399902
    },
    {
      "auxiliary_loss_clip": 0.01283775,
      "auxiliary_loss_mlp": 0.01097438,
      "balance_loss_clip": 1.07794189,
      "balance_loss_mlp": 1.06155562,
      "epoch": 0.02909965429129716,
      "flos": 16763640318720.0,
      "grad_norm": 2.6777328906555766,
      "language_loss": 0.75510043,
      "learning_rate": 3.980348865796749e-06,
      "loss": 0.77891254,
      "num_input_tokens_seen": 10306875,
      "step": 484,
      "time_per_iteration": 2.608337640762329
    },
    {
      "auxiliary_loss_clip": 0.0127962,
      "auxiliary_loss_mlp": 0.01081582,
      "balance_loss_clip": 1.07543373,
      "balance_loss_mlp": 1.04760778,
      "epoch": 0.029159777543965128,
      "flos": 19785343023360.0,
      "grad_norm": 2.3457282915841113,
      "language_loss": 0.8378315,
      "learning_rate": 3.9816777665440615e-06,
      "loss": 0.86144352,
      "num_input_tokens_seen": 10323965,
      "step": 485,
      "time_per_iteration": 2.591409921646118
    },
    {
      "auxiliary_loss_clip": 0.01282377,
      "auxiliary_loss_mlp": 0.01084922,
      "balance_loss_clip": 1.08029485,
      "balance_loss_mlp": 1.04956484,
      "epoch": 0.029219900796633096,
      "flos": 19642670202240.0,
      "grad_norm": 2.044831141886674,
      "language_loss": 0.84432101,
      "learning_rate": 3.983003930109732e-06,
      "loss": 0.86799401,
      "num_input_tokens_seen": 10342620,
      "step": 486,
      "time_per_iteration": 2.7101452350616455
    },
    {
      "auxiliary_loss_clip": 0.01276806,
      "auxiliary_loss_mlp": 0.01090739,
      "balance_loss_clip": 1.07363296,
      "balance_loss_mlp": 1.05476189,
      "epoch": 0.02928002404930107,
      "flos": 25885704424320.0,
      "grad_norm": 12.432525192672303,
      "language_loss": 0.88968349,
      "learning_rate": 3.984327367746315e-06,
      "loss": 0.91335887,
      "num_input_tokens_seen": 10364610,
      "step": 487,
      "time_per_iteration": 2.637910842895508
    },
    {
      "auxiliary_loss_clip": 0.01283084,
      "auxiliary_loss_mlp": 0.01069223,
      "balance_loss_clip": 1.07921362,
      "balance_loss_mlp": 1.03677416,
      "epoch": 0.029340147301969037,
      "flos": 20660234590080.0,
      "grad_norm": 2.566388301054309,
      "language_loss": 0.88581878,
      "learning_rate": 3.985648090637122e-06,
      "loss": 0.90934181,
      "num_input_tokens_seen": 10380910,
      "step": 488,
      "time_per_iteration": 2.6569244861602783
    },
    {
      "auxiliary_loss_clip": 0.01275613,
      "auxiliary_loss_mlp": 0.01081415,
      "balance_loss_clip": 1.07419777,
      "balance_loss_mlp": 1.04667735,
      "epoch": 0.029400270554637006,
      "flos": 24428018689920.0,
      "grad_norm": 2.0135021623582503,
      "language_loss": 0.88869834,
      "learning_rate": 3.986966109896785e-06,
      "loss": 0.91226858,
      "num_input_tokens_seen": 10400665,
      "step": 489,
      "time_per_iteration": 2.805555582046509
    },
    {
      "auxiliary_loss_clip": 0.01271096,
      "auxiliary_loss_mlp": 0.01077182,
      "balance_loss_clip": 1.0704807,
      "balance_loss_mlp": 1.04168141,
      "epoch": 0.029460393807304974,
      "flos": 20120892900480.0,
      "grad_norm": 2.807428314395572,
      "language_loss": 0.88554472,
      "learning_rate": 3.988281436571815e-06,
      "loss": 0.90902752,
      "num_input_tokens_seen": 10420150,
      "step": 490,
      "time_per_iteration": 2.612993001937866
    },
    {
      "auxiliary_loss_clip": 0.01276687,
      "auxiliary_loss_mlp": 0.01088031,
      "balance_loss_clip": 1.0729506,
      "balance_loss_mlp": 1.0536747,
      "epoch": 0.029520517059972943,
      "flos": 17675914965120.0,
      "grad_norm": 2.430337539839543,
      "language_loss": 0.91496718,
      "learning_rate": 3.989594081641164e-06,
      "loss": 0.93861437,
      "num_input_tokens_seen": 10438210,
      "step": 491,
      "time_per_iteration": 2.6203627586364746
    },
    {
      "auxiliary_loss_clip": 0.01266864,
      "auxiliary_loss_mlp": 0.01072939,
      "balance_loss_clip": 1.07131863,
      "balance_loss_mlp": 1.03984618,
      "epoch": 0.029580640312640915,
      "flos": 18953185662720.0,
      "grad_norm": 1.9753258841331502,
      "language_loss": 0.85654163,
      "learning_rate": 3.9909040560167675e-06,
      "loss": 0.87993968,
      "num_input_tokens_seen": 10455125,
      "step": 492,
      "time_per_iteration": 2.636378288269043
    },
    {
      "auxiliary_loss_clip": 0.01279009,
      "auxiliary_loss_mlp": 0.01100381,
      "balance_loss_clip": 1.07765996,
      "balance_loss_mlp": 1.06471384,
      "epoch": 0.029640763565308884,
      "flos": 18726121837440.0,
      "grad_norm": 4.076790847855052,
      "language_loss": 0.84615922,
      "learning_rate": 3.992211370544093e-06,
      "loss": 0.86995316,
      "num_input_tokens_seen": 10470990,
      "step": 493,
      "time_per_iteration": 2.6144914627075195
    },
    {
      "auxiliary_loss_clip": 0.01272514,
      "auxiliary_loss_mlp": 0.01074657,
      "balance_loss_clip": 1.07140934,
      "balance_loss_mlp": 1.04042029,
      "epoch": 0.029700886817976852,
      "flos": 20595308757120.0,
      "grad_norm": 1.8084917907335818,
      "language_loss": 0.8658669,
      "learning_rate": 3.99351603600268e-06,
      "loss": 0.88933873,
      "num_input_tokens_seen": 10490685,
      "step": 494,
      "time_per_iteration": 2.7063095569610596
    },
    {
      "auxiliary_loss_clip": 0.01281688,
      "auxiliary_loss_mlp": 0.01084428,
      "balance_loss_clip": 1.07739305,
      "balance_loss_mlp": 1.05279028,
      "epoch": 0.02976101007064482,
      "flos": 22236857233920.0,
      "grad_norm": 7.125038043922513,
      "language_loss": 0.86841047,
      "learning_rate": 3.994818063106668e-06,
      "loss": 0.8920716,
      "num_input_tokens_seen": 10509435,
      "step": 495,
      "time_per_iteration": 2.641700267791748
    },
    {
      "auxiliary_loss_clip": 0.01268945,
      "auxiliary_loss_mlp": 0.01078198,
      "balance_loss_clip": 1.07384837,
      "balance_loss_mlp": 1.04508162,
      "epoch": 0.029821133323312793,
      "flos": 23732644320000.0,
      "grad_norm": 2.201071528053665,
      "language_loss": 0.61988759,
      "learning_rate": 3.99611746250533e-06,
      "loss": 0.64335901,
      "num_input_tokens_seen": 10530050,
      "step": 496,
      "time_per_iteration": 2.6524407863616943
    },
    {
      "auxiliary_loss_clip": 0.01270994,
      "auxiliary_loss_mlp": 0.01089922,
      "balance_loss_clip": 1.07575428,
      "balance_loss_mlp": 1.05680561,
      "epoch": 0.02988125657598076,
      "flos": 22419498913920.0,
      "grad_norm": 1.7538974268426115,
      "language_loss": 0.88820887,
      "learning_rate": 3.997414244783595e-06,
      "loss": 0.91181797,
      "num_input_tokens_seen": 10551370,
      "step": 497,
      "time_per_iteration": 5.648245811462402
    },
    {
      "auxiliary_loss_clip": 0.01277289,
      "auxiliary_loss_mlp": 0.01079642,
      "balance_loss_clip": 1.07670021,
      "balance_loss_mlp": 1.04604888,
      "epoch": 0.02994137982864873,
      "flos": 13845108453120.0,
      "grad_norm": 2.8395997319333204,
      "language_loss": 0.85091698,
      "learning_rate": 3.998708420462557e-06,
      "loss": 0.87448633,
      "num_input_tokens_seen": 10569225,
      "step": 498,
      "time_per_iteration": 4.362173080444336
    },
    {
      "auxiliary_loss_clip": 0.0127249,
      "auxiliary_loss_mlp": 0.01078673,
      "balance_loss_clip": 1.07436109,
      "balance_loss_mlp": 1.04691589,
      "epoch": 0.0300015030813167,
      "flos": 23908354675200.0,
      "grad_norm": 3.2275044857926605,
      "language_loss": 0.77883017,
      "learning_rate": 4e-06,
      "loss": 0.80234182,
      "num_input_tokens_seen": 10586170,
      "step": 499,
      "time_per_iteration": 2.6029655933380127
    },
    {
      "auxiliary_loss_clip": 0.01272525,
      "auxiliary_loss_mlp": 0.01082339,
      "balance_loss_clip": 1.07433248,
      "balance_loss_mlp": 1.04905546,
      "epoch": 0.030061626333984667,
      "flos": 22016796560640.0,
      "grad_norm": 2.244229511477372,
      "language_loss": 0.82687509,
      "learning_rate": 3.9999999620799e-06,
      "loss": 0.85042375,
      "num_input_tokens_seen": 10606205,
      "step": 500,
      "time_per_iteration": 2.6293113231658936
    },
    {
      "auxiliary_loss_clip": 0.01266453,
      "auxiliary_loss_mlp": 0.0108458,
      "balance_loss_clip": 1.07100737,
      "balance_loss_mlp": 1.04922247,
      "epoch": 0.03012174958665264,
      "flos": 23039747988480.0,
      "grad_norm": 3.2569274145363356,
      "language_loss": 0.88086087,
      "learning_rate": 3.9999998483196e-06,
      "loss": 0.90437114,
      "num_input_tokens_seen": 10625995,
      "step": 501,
      "time_per_iteration": 2.601081132888794
    },
    {
      "auxiliary_loss_clip": 0.01273997,
      "auxiliary_loss_mlp": 0.01071746,
      "balance_loss_clip": 1.07361674,
      "balance_loss_mlp": 1.04025102,
      "epoch": 0.030181872839320608,
      "flos": 18953257489920.0,
      "grad_norm": 3.3627001763511855,
      "language_loss": 0.86654103,
      "learning_rate": 3.9999996587191065e-06,
      "loss": 0.88999844,
      "num_input_tokens_seen": 10644105,
      "step": 502,
      "time_per_iteration": 2.5507659912109375
    },
    {
      "auxiliary_loss_clip": 0.01270542,
      "auxiliary_loss_mlp": 0.01081534,
      "balance_loss_clip": 1.07475543,
      "balance_loss_mlp": 1.04827452,
      "epoch": 0.030241996091988577,
      "flos": 16728017005440.0,
      "grad_norm": 2.4572357458963876,
      "language_loss": 0.84281206,
      "learning_rate": 3.999999393278425e-06,
      "loss": 0.86633277,
      "num_input_tokens_seen": 10661090,
      "step": 503,
      "time_per_iteration": 2.618587017059326
    },
    {
      "auxiliary_loss_clip": 0.01262547,
      "auxiliary_loss_mlp": 0.01091143,
      "balance_loss_clip": 1.0710721,
      "balance_loss_mlp": 1.05781209,
      "epoch": 0.030302119344656545,
      "flos": 28621271387520.0,
      "grad_norm": 1.6994359255159197,
      "language_loss": 0.88137805,
      "learning_rate": 3.999999051997567e-06,
      "loss": 0.90491492,
      "num_input_tokens_seen": 10682380,
      "step": 504,
      "time_per_iteration": 2.6794183254241943
    },
    {
      "auxiliary_loss_clip": 0.01264601,
      "auxiliary_loss_mlp": 0.01086749,
      "balance_loss_clip": 1.07040262,
      "balance_loss_mlp": 1.0541091,
      "epoch": 0.030362242597324514,
      "flos": 15669334523520.0,
      "grad_norm": 2.074855698516145,
      "language_loss": 0.786093,
      "learning_rate": 3.9999986348765425e-06,
      "loss": 0.80960649,
      "num_input_tokens_seen": 10699925,
      "step": 505,
      "time_per_iteration": 2.564960479736328
    },
    {
      "auxiliary_loss_clip": 0.01134686,
      "auxiliary_loss_mlp": 0.010147,
      "balance_loss_clip": 1.03763247,
      "balance_loss_mlp": 1.00692737,
      "epoch": 0.030422365849992486,
      "flos": 72125973676800.0,
      "grad_norm": 0.9565689962416369,
      "language_loss": 0.54981297,
      "learning_rate": 3.999998141915371e-06,
      "loss": 0.57130682,
      "num_input_tokens_seen": 10766525,
      "step": 506,
      "time_per_iteration": 3.3345654010772705
    },
    {
      "auxiliary_loss_clip": 0.01266577,
      "auxiliary_loss_mlp": 0.01090299,
      "balance_loss_clip": 1.07119894,
      "balance_loss_mlp": 1.05687308,
      "epoch": 0.030482489102660455,
      "flos": 19427817000960.0,
      "grad_norm": 2.2738865373146684,
      "language_loss": 0.83377159,
      "learning_rate": 3.999997573114069e-06,
      "loss": 0.8573404,
      "num_input_tokens_seen": 10786725,
      "step": 507,
      "time_per_iteration": 2.645613670349121
    },
    {
      "auxiliary_loss_clip": 0.01269938,
      "auxiliary_loss_mlp": 0.01076205,
      "balance_loss_clip": 1.07151937,
      "balance_loss_mlp": 1.04344678,
      "epoch": 0.030542612355328423,
      "flos": 20375822701440.0,
      "grad_norm": 2.375369924968869,
      "language_loss": 0.88842839,
      "learning_rate": 3.999996928472659e-06,
      "loss": 0.91188985,
      "num_input_tokens_seen": 10805390,
      "step": 508,
      "time_per_iteration": 2.617283344268799
    },
    {
      "auxiliary_loss_clip": 0.01272148,
      "auxiliary_loss_mlp": 0.01067206,
      "balance_loss_clip": 1.07232118,
      "balance_loss_mlp": 1.03394616,
      "epoch": 0.030602735607996392,
      "flos": 34677354297600.0,
      "grad_norm": 6.964954749829821,
      "language_loss": 0.71807706,
      "learning_rate": 3.999996207991165e-06,
      "loss": 0.74147063,
      "num_input_tokens_seen": 10828030,
      "step": 509,
      "time_per_iteration": 2.7723498344421387
    },
    {
      "auxiliary_loss_clip": 0.01264594,
      "auxiliary_loss_mlp": 0.01074377,
      "balance_loss_clip": 1.07241154,
      "balance_loss_mlp": 1.04333544,
      "epoch": 0.03066285886066436,
      "flos": 23658668259840.0,
      "grad_norm": 1.9285974370038053,
      "language_loss": 0.82031929,
      "learning_rate": 3.999995411669614e-06,
      "loss": 0.84370899,
      "num_input_tokens_seen": 10845240,
      "step": 510,
      "time_per_iteration": 2.6254217624664307
    },
    {
      "auxiliary_loss_clip": 0.01268793,
      "auxiliary_loss_mlp": 0.01075379,
      "balance_loss_clip": 1.07532823,
      "balance_loss_mlp": 1.04252458,
      "epoch": 0.030722982113332332,
      "flos": 23002975440000.0,
      "grad_norm": 5.706057095430757,
      "language_loss": 0.83572316,
      "learning_rate": 3.999994539508036e-06,
      "loss": 0.85916495,
      "num_input_tokens_seen": 10864325,
      "step": 511,
      "time_per_iteration": 2.613457441329956
    },
    {
      "auxiliary_loss_clip": 0.01269742,
      "auxiliary_loss_mlp": 0.01081314,
      "balance_loss_clip": 1.07207167,
      "balance_loss_mlp": 1.0496521,
      "epoch": 0.0307831053660003,
      "flos": 24750855152640.0,
      "grad_norm": 2.025270681093948,
      "language_loss": 0.82109964,
      "learning_rate": 3.9999935915064655e-06,
      "loss": 0.84461015,
      "num_input_tokens_seen": 10883860,
      "step": 512,
      "time_per_iteration": 2.630404233932495
    },
    {
      "auxiliary_loss_clip": 0.01266054,
      "auxiliary_loss_mlp": 0.01084436,
      "balance_loss_clip": 1.07086158,
      "balance_loss_mlp": 1.05070007,
      "epoch": 0.03084322861866827,
      "flos": 26140885620480.0,
      "grad_norm": 2.500363981205655,
      "language_loss": 0.86933553,
      "learning_rate": 3.9999925676649374e-06,
      "loss": 0.89284045,
      "num_input_tokens_seen": 10904555,
      "step": 513,
      "time_per_iteration": 2.671926259994507
    },
    {
      "auxiliary_loss_clip": 0.01272542,
      "auxiliary_loss_mlp": 0.01080065,
      "balance_loss_clip": 1.07461214,
      "balance_loss_mlp": 1.04744935,
      "epoch": 0.03090335187133624,
      "flos": 18771298168320.0,
      "grad_norm": 1.704575426690477,
      "language_loss": 0.79124331,
      "learning_rate": 3.999991467983491e-06,
      "loss": 0.81476939,
      "num_input_tokens_seen": 10923700,
      "step": 514,
      "time_per_iteration": 2.6158573627471924
    },
    {
      "auxiliary_loss_clip": 0.01265821,
      "auxiliary_loss_mlp": 0.01067844,
      "balance_loss_clip": 1.07397485,
      "balance_loss_mlp": 1.03711247,
      "epoch": 0.030963475124004207,
      "flos": 23221886878080.0,
      "grad_norm": 2.729063628201222,
      "language_loss": 0.77758944,
      "learning_rate": 3.999990292462167e-06,
      "loss": 0.80092615,
      "num_input_tokens_seen": 10942730,
      "step": 515,
      "time_per_iteration": 2.636294364929199
    },
    {
      "auxiliary_loss_clip": 0.0126398,
      "auxiliary_loss_mlp": 0.01072575,
      "balance_loss_clip": 1.06835747,
      "balance_loss_mlp": 1.03874326,
      "epoch": 0.03102359837667218,
      "flos": 42525595411200.0,
      "grad_norm": 2.1228851207681503,
      "language_loss": 0.82452714,
      "learning_rate": 3.999989041101011e-06,
      "loss": 0.84789264,
      "num_input_tokens_seen": 10967120,
      "step": 516,
      "time_per_iteration": 2.8078057765960693
    },
    {
      "auxiliary_loss_clip": 0.01263726,
      "auxiliary_loss_mlp": 0.01073859,
      "balance_loss_clip": 1.0712111,
      "balance_loss_mlp": 1.04090929,
      "epoch": 0.031083721629340148,
      "flos": 21176953689600.0,
      "grad_norm": 1.9016724574566626,
      "language_loss": 0.79088318,
      "learning_rate": 3.999987713900071e-06,
      "loss": 0.81425899,
      "num_input_tokens_seen": 10986775,
      "step": 517,
      "time_per_iteration": 2.5935981273651123
    },
    {
      "auxiliary_loss_clip": 0.0125895,
      "auxiliary_loss_mlp": 0.0107836,
      "balance_loss_clip": 1.07049131,
      "balance_loss_mlp": 1.04629326,
      "epoch": 0.031143844882008116,
      "flos": 29716187713920.0,
      "grad_norm": 1.6829619528007147,
      "language_loss": 0.90798068,
      "learning_rate": 3.999986310859396e-06,
      "loss": 0.93135381,
      "num_input_tokens_seen": 11011360,
      "step": 518,
      "time_per_iteration": 2.6855509281158447
    },
    {
      "auxiliary_loss_clip": 0.01272237,
      "auxiliary_loss_mlp": 0.01097567,
      "balance_loss_clip": 1.07848859,
      "balance_loss_mlp": 1.06230497,
      "epoch": 0.031203968134676085,
      "flos": 23112467072640.0,
      "grad_norm": 1.8835331125391583,
      "language_loss": 0.86759162,
      "learning_rate": 3.999984831979039e-06,
      "loss": 0.89128959,
      "num_input_tokens_seen": 11030150,
      "step": 519,
      "time_per_iteration": 2.628380060195923
    },
    {
      "auxiliary_loss_clip": 0.01265864,
      "auxiliary_loss_mlp": 0.01086943,
      "balance_loss_clip": 1.06901193,
      "balance_loss_mlp": 1.05578136,
      "epoch": 0.03126409138734405,
      "flos": 20954379064320.0,
      "grad_norm": 3.8823628482318164,
      "language_loss": 0.87246573,
      "learning_rate": 3.999983277259057e-06,
      "loss": 0.89599377,
      "num_input_tokens_seen": 11049145,
      "step": 520,
      "time_per_iteration": 2.5850255489349365
    },
    {
      "auxiliary_loss_clip": 0.01269157,
      "auxiliary_loss_mlp": 0.01086266,
      "balance_loss_clip": 1.07231963,
      "balance_loss_mlp": 1.0528394,
      "epoch": 0.031324214640012026,
      "flos": 21650112570240.0,
      "grad_norm": 1.7050130216714323,
      "language_loss": 0.89274424,
      "learning_rate": 3.999981646699509e-06,
      "loss": 0.91629851,
      "num_input_tokens_seen": 11068835,
      "step": 521,
      "time_per_iteration": 2.6412506103515625
    },
    {
      "auxiliary_loss_clip": 0.01263772,
      "auxiliary_loss_mlp": 0.01082584,
      "balance_loss_clip": 1.0717473,
      "balance_loss_mlp": 1.04827595,
      "epoch": 0.03138433789267999,
      "flos": 23441337020160.0,
      "grad_norm": 2.085624200373119,
      "language_loss": 0.71452564,
      "learning_rate": 3.999979940300456e-06,
      "loss": 0.73798925,
      "num_input_tokens_seen": 11088980,
      "step": 522,
      "time_per_iteration": 2.6561174392700195
    },
    {
      "auxiliary_loss_clip": 0.01265725,
      "auxiliary_loss_mlp": 0.01082552,
      "balance_loss_clip": 1.06871116,
      "balance_loss_mlp": 1.05079484,
      "epoch": 0.03144446114534796,
      "flos": 18982164960000.0,
      "grad_norm": 4.223323698032832,
      "language_loss": 0.84758592,
      "learning_rate": 3.999978158061963e-06,
      "loss": 0.87106872,
      "num_input_tokens_seen": 11104300,
      "step": 523,
      "time_per_iteration": 2.608565330505371
    },
    {
      "auxiliary_loss_clip": 0.01271589,
      "auxiliary_loss_mlp": 0.01076253,
      "balance_loss_clip": 1.07193565,
      "balance_loss_mlp": 1.04296994,
      "epoch": 0.031504584398015935,
      "flos": 22637692080000.0,
      "grad_norm": 2.324094801199308,
      "language_loss": 0.89989722,
      "learning_rate": 3.999976299984099e-06,
      "loss": 0.92337573,
      "num_input_tokens_seen": 11123335,
      "step": 524,
      "time_per_iteration": 2.68269944190979
    },
    {
      "auxiliary_loss_clip": 0.01273471,
      "auxiliary_loss_mlp": 0.0108318,
      "balance_loss_clip": 1.07427168,
      "balance_loss_mlp": 1.04944324,
      "epoch": 0.0315647076506839,
      "flos": 25297056339840.0,
      "grad_norm": 2.4635323942475766,
      "language_loss": 0.80114233,
      "learning_rate": 3.999974366066933e-06,
      "loss": 0.82470882,
      "num_input_tokens_seen": 11140880,
      "step": 525,
      "time_per_iteration": 2.6396324634552
    },
    {
      "auxiliary_loss_clip": 0.01264716,
      "auxiliary_loss_mlp": 0.01080959,
      "balance_loss_clip": 1.0681529,
      "balance_loss_mlp": 1.04798603,
      "epoch": 0.03162483090335187,
      "flos": 16982839065600.0,
      "grad_norm": 2.3553733144031948,
      "language_loss": 0.81162,
      "learning_rate": 3.999972356310538e-06,
      "loss": 0.83507675,
      "num_input_tokens_seen": 11158710,
      "step": 526,
      "time_per_iteration": 2.6167168617248535
    },
    {
      "auxiliary_loss_clip": 0.01273987,
      "auxiliary_loss_mlp": 0.01072725,
      "balance_loss_clip": 1.07507181,
      "balance_loss_mlp": 1.03736734,
      "epoch": 0.03168495415601984,
      "flos": 18734489706240.0,
      "grad_norm": 1.9666844995001491,
      "language_loss": 0.81491739,
      "learning_rate": 3.999970270714991e-06,
      "loss": 0.83838451,
      "num_input_tokens_seen": 11177550,
      "step": 527,
      "time_per_iteration": 2.580310821533203
    },
    {
      "auxiliary_loss_clip": 0.01261155,
      "auxiliary_loss_mlp": 0.01080842,
      "balance_loss_clip": 1.06786597,
      "balance_loss_mlp": 1.04717755,
      "epoch": 0.03174507740868781,
      "flos": 21214875473280.0,
      "grad_norm": 1.9105688869262756,
      "language_loss": 0.93801636,
      "learning_rate": 3.999968109280371e-06,
      "loss": 0.96143627,
      "num_input_tokens_seen": 11196230,
      "step": 528,
      "time_per_iteration": 2.5901002883911133
    },
    {
      "auxiliary_loss_clip": 0.01263275,
      "auxiliary_loss_mlp": 0.01071724,
      "balance_loss_clip": 1.06776333,
      "balance_loss_mlp": 1.0387274,
      "epoch": 0.03180520066135578,
      "flos": 24787663614720.0,
      "grad_norm": 1.8924176613796981,
      "language_loss": 0.84130204,
      "learning_rate": 3.99996587200676e-06,
      "loss": 0.86465204,
      "num_input_tokens_seen": 11214935,
      "step": 529,
      "time_per_iteration": 2.593867063522339
    },
    {
      "auxiliary_loss_clip": 0.01266309,
      "auxiliary_loss_mlp": 0.01088988,
      "balance_loss_clip": 1.07501197,
      "balance_loss_mlp": 1.0563724,
      "epoch": 0.03186532391402375,
      "flos": 24864261367680.0,
      "grad_norm": 2.316883777672742,
      "language_loss": 0.90458709,
      "learning_rate": 3.999963558894243e-06,
      "loss": 0.92814004,
      "num_input_tokens_seen": 11235310,
      "step": 530,
      "time_per_iteration": 2.5994982719421387
    },
    {
      "auxiliary_loss_clip": 0.01261024,
      "auxiliary_loss_mlp": 0.0107627,
      "balance_loss_clip": 1.06481552,
      "balance_loss_mlp": 1.04188991,
      "epoch": 0.03192544716669172,
      "flos": 21215055041280.0,
      "grad_norm": 2.2744046769674324,
      "language_loss": 0.76334512,
      "learning_rate": 3.999961169942907e-06,
      "loss": 0.78671807,
      "num_input_tokens_seen": 11254425,
      "step": 531,
      "time_per_iteration": 2.618149757385254
    },
    {
      "auxiliary_loss_clip": 0.01260981,
      "auxiliary_loss_mlp": 0.01064937,
      "balance_loss_clip": 1.0669558,
      "balance_loss_mlp": 1.03143883,
      "epoch": 0.03198557041935969,
      "flos": 24353216616960.0,
      "grad_norm": 2.467757262816931,
      "language_loss": 0.90483695,
      "learning_rate": 3.999958705152843e-06,
      "loss": 0.92809618,
      "num_input_tokens_seen": 11274595,
      "step": 532,
      "time_per_iteration": 2.647947072982788
    },
    {
      "auxiliary_loss_clip": 0.01146464,
      "auxiliary_loss_mlp": 0.01012028,
      "balance_loss_clip": 1.04988623,
      "balance_loss_mlp": 1.00325394,
      "epoch": 0.032045693672027656,
      "flos": 61827367587840.0,
      "grad_norm": 1.9655071928838626,
      "language_loss": 0.57953775,
      "learning_rate": 3.9999561645241445e-06,
      "loss": 0.60112268,
      "num_input_tokens_seen": 11336705,
      "step": 533,
      "time_per_iteration": 3.2502808570861816
    },
    {
      "auxiliary_loss_clip": 0.01260941,
      "auxiliary_loss_mlp": 0.01084263,
      "balance_loss_clip": 1.06724441,
      "balance_loss_mlp": 1.0516715,
      "epoch": 0.03210581692469563,
      "flos": 28401174800640.0,
      "grad_norm": 1.7138682169725878,
      "language_loss": 0.86666048,
      "learning_rate": 3.999953548056907e-06,
      "loss": 0.89011252,
      "num_input_tokens_seen": 11356820,
      "step": 534,
      "time_per_iteration": 2.678739070892334
    },
    {
      "auxiliary_loss_clip": 0.01259554,
      "auxiliary_loss_mlp": 0.01066669,
      "balance_loss_clip": 1.06782031,
      "balance_loss_mlp": 1.03407741,
      "epoch": 0.03216594017736359,
      "flos": 24717709877760.0,
      "grad_norm": 2.12774196295415,
      "language_loss": 0.77627808,
      "learning_rate": 3.999950855751232e-06,
      "loss": 0.79954034,
      "num_input_tokens_seen": 11376645,
      "step": 535,
      "time_per_iteration": 2.7128217220306396
    },
    {
      "auxiliary_loss_clip": 0.01261708,
      "auxiliary_loss_mlp": 0.01081378,
      "balance_loss_clip": 1.06843078,
      "balance_loss_mlp": 1.0485003,
      "epoch": 0.032226063430031565,
      "flos": 31175453646720.0,
      "grad_norm": 3.9913279940153585,
      "language_loss": 0.80939913,
      "learning_rate": 3.999948087607219e-06,
      "loss": 0.83283001,
      "num_input_tokens_seen": 11397310,
      "step": 536,
      "time_per_iteration": 2.7490127086639404
    },
    {
      "auxiliary_loss_clip": 0.01262237,
      "auxiliary_loss_mlp": 0.01075987,
      "balance_loss_clip": 1.06839073,
      "balance_loss_mlp": 1.04167831,
      "epoch": 0.03228618668269954,
      "flos": 32198225506560.0,
      "grad_norm": 1.6888601787189168,
      "language_loss": 0.7009111,
      "learning_rate": 3.999945243624975e-06,
      "loss": 0.72429335,
      "num_input_tokens_seen": 11418475,
      "step": 537,
      "time_per_iteration": 5.5609166622161865
    },
    {
      "auxiliary_loss_clip": 0.0126357,
      "auxiliary_loss_mlp": 0.01084205,
      "balance_loss_clip": 1.07331729,
      "balance_loss_mlp": 1.05161297,
      "epoch": 0.0323463099353675,
      "flos": 22670154996480.0,
      "grad_norm": 2.146306428033486,
      "language_loss": 0.82684958,
      "learning_rate": 3.999942323804607e-06,
      "loss": 0.85032725,
      "num_input_tokens_seen": 11436630,
      "step": 538,
      "time_per_iteration": 2.5465030670166016
    },
    {
      "auxiliary_loss_clip": 0.01269537,
      "auxiliary_loss_mlp": 0.01078099,
      "balance_loss_clip": 1.06987572,
      "balance_loss_mlp": 1.04536414,
      "epoch": 0.032406433188035474,
      "flos": 26905172232960.0,
      "grad_norm": 1.8709064214989917,
      "language_loss": 0.79146457,
      "learning_rate": 3.999939328146225e-06,
      "loss": 0.81494099,
      "num_input_tokens_seen": 11457275,
      "step": 539,
      "time_per_iteration": 4.172123432159424
    },
    {
      "auxiliary_loss_clip": 0.0126143,
      "auxiliary_loss_mlp": 0.01069528,
      "balance_loss_clip": 1.06830835,
      "balance_loss_mlp": 1.03567231,
      "epoch": 0.03246655644070344,
      "flos": 31503928544640.0,
      "grad_norm": 35.59051030008172,
      "language_loss": 0.77379727,
      "learning_rate": 3.999936256649943e-06,
      "loss": 0.79710686,
      "num_input_tokens_seen": 11476925,
      "step": 540,
      "time_per_iteration": 2.5633046627044678
    },
    {
      "auxiliary_loss_clip": 0.01269863,
      "auxiliary_loss_mlp": 0.01073669,
      "balance_loss_clip": 1.07271969,
      "balance_loss_mlp": 1.04124355,
      "epoch": 0.03252667969337141,
      "flos": 23218331431680.0,
      "grad_norm": 2.0489065110302636,
      "language_loss": 0.85458571,
      "learning_rate": 3.999933109315878e-06,
      "loss": 0.878021,
      "num_input_tokens_seen": 11496830,
      "step": 541,
      "time_per_iteration": 2.6079938411712646
    },
    {
      "auxiliary_loss_clip": 0.01258504,
      "auxiliary_loss_mlp": 0.01082451,
      "balance_loss_clip": 1.06961954,
      "balance_loss_mlp": 1.04835749,
      "epoch": 0.032586802946039384,
      "flos": 14757454926720.0,
      "grad_norm": 2.674731240129174,
      "language_loss": 0.89234567,
      "learning_rate": 3.9999298861441496e-06,
      "loss": 0.91575521,
      "num_input_tokens_seen": 11515605,
      "step": 542,
      "time_per_iteration": 2.597036600112915
    },
    {
      "auxiliary_loss_clip": 0.0126351,
      "auxiliary_loss_mlp": 0.01081041,
      "balance_loss_clip": 1.06974792,
      "balance_loss_mlp": 1.04792452,
      "epoch": 0.03264692619870735,
      "flos": 24280677100800.0,
      "grad_norm": 2.2714121360014334,
      "language_loss": 0.71123677,
      "learning_rate": 3.999926587134879e-06,
      "loss": 0.73468232,
      "num_input_tokens_seen": 11536230,
      "step": 543,
      "time_per_iteration": 2.634601354598999
    },
    {
      "auxiliary_loss_clip": 0.01259994,
      "auxiliary_loss_mlp": 0.01088763,
      "balance_loss_clip": 1.06379187,
      "balance_loss_mlp": 1.05545604,
      "epoch": 0.03270704945137532,
      "flos": 22893160584960.0,
      "grad_norm": 4.777521083182084,
      "language_loss": 0.91540575,
      "learning_rate": 3.999923212288192e-06,
      "loss": 0.93889332,
      "num_input_tokens_seen": 11554715,
      "step": 544,
      "time_per_iteration": 2.6173009872436523
    },
    {
      "auxiliary_loss_clip": 0.01264485,
      "auxiliary_loss_mlp": 0.01085684,
      "balance_loss_clip": 1.06989884,
      "balance_loss_mlp": 1.05571437,
      "epoch": 0.032767172704043286,
      "flos": 18041018757120.0,
      "grad_norm": 2.6951315012120025,
      "language_loss": 0.65799558,
      "learning_rate": 3.999919761604216e-06,
      "loss": 0.68149722,
      "num_input_tokens_seen": 11571370,
      "step": 545,
      "time_per_iteration": 2.6500988006591797
    },
    {
      "auxiliary_loss_clip": 0.012623,
      "auxiliary_loss_mlp": 0.0107161,
      "balance_loss_clip": 1.06693912,
      "balance_loss_mlp": 1.0393517,
      "epoch": 0.03282729595671126,
      "flos": 22528739151360.0,
      "grad_norm": 2.2564766449723908,
      "language_loss": 0.92221987,
      "learning_rate": 3.999916235083083e-06,
      "loss": 0.94555902,
      "num_input_tokens_seen": 11588560,
      "step": 546,
      "time_per_iteration": 2.673250913619995
    },
    {
      "auxiliary_loss_clip": 0.01260258,
      "auxiliary_loss_mlp": 0.01077296,
      "balance_loss_clip": 1.06488204,
      "balance_loss_mlp": 1.04313052,
      "epoch": 0.03288741920937923,
      "flos": 20410620001920.0,
      "grad_norm": 2.1923718908590653,
      "language_loss": 0.81706661,
      "learning_rate": 3.999912632724925e-06,
      "loss": 0.84044212,
      "num_input_tokens_seen": 11605685,
      "step": 547,
      "time_per_iteration": 2.725198745727539
    },
    {
      "auxiliary_loss_clip": 0.0126227,
      "auxiliary_loss_mlp": 0.0107871,
      "balance_loss_clip": 1.06794477,
      "balance_loss_mlp": 1.04480648,
      "epoch": 0.032947542462047195,
      "flos": 20777986350720.0,
      "grad_norm": 1.730652582963277,
      "language_loss": 0.81227565,
      "learning_rate": 3.999908954529881e-06,
      "loss": 0.83568549,
      "num_input_tokens_seen": 11626290,
      "step": 548,
      "time_per_iteration": 2.714073419570923
    },
    {
      "auxiliary_loss_clip": 0.01264818,
      "auxiliary_loss_mlp": 0.01084154,
      "balance_loss_clip": 1.06963027,
      "balance_loss_mlp": 1.04870164,
      "epoch": 0.03300766571471517,
      "flos": 19901263190400.0,
      "grad_norm": 3.8540092911047603,
      "language_loss": 0.67460287,
      "learning_rate": 3.999905200498087e-06,
      "loss": 0.69809258,
      "num_input_tokens_seen": 11643950,
      "step": 549,
      "time_per_iteration": 2.6747171878814697
    },
    {
      "auxiliary_loss_clip": 0.0125805,
      "auxiliary_loss_mlp": 0.01076001,
      "balance_loss_clip": 1.06968856,
      "balance_loss_mlp": 1.04236054,
      "epoch": 0.03306778896738313,
      "flos": 17967760968960.0,
      "grad_norm": 1.933615596136007,
      "language_loss": 0.86379111,
      "learning_rate": 3.999901370629689e-06,
      "loss": 0.88713157,
      "num_input_tokens_seen": 11662560,
      "step": 550,
      "time_per_iteration": 2.553386926651001
    },
    {
      "auxiliary_loss_clip": 0.01264951,
      "auxiliary_loss_mlp": 0.01095377,
      "balance_loss_clip": 1.07279766,
      "balance_loss_mlp": 1.06142652,
      "epoch": 0.033127912220051105,
      "flos": 21653380707840.0,
      "grad_norm": 3.1958143211070977,
      "language_loss": 0.8127178,
      "learning_rate": 3.99989746492483e-06,
      "loss": 0.83632112,
      "num_input_tokens_seen": 11682265,
      "step": 551,
      "time_per_iteration": 2.6231682300567627
    },
    {
      "auxiliary_loss_clip": 0.01271579,
      "auxiliary_loss_mlp": 0.0108998,
      "balance_loss_clip": 1.07285261,
      "balance_loss_mlp": 1.05626702,
      "epoch": 0.03318803547271908,
      "flos": 30188376927360.0,
      "grad_norm": 2.9473143774727606,
      "language_loss": 0.86134821,
      "learning_rate": 3.999893483383658e-06,
      "loss": 0.88496381,
      "num_input_tokens_seen": 11699300,
      "step": 552,
      "time_per_iteration": 2.7002694606781006
    },
    {
      "auxiliary_loss_clip": 0.01267081,
      "auxiliary_loss_mlp": 0.01081671,
      "balance_loss_clip": 1.07191086,
      "balance_loss_mlp": 1.04650474,
      "epoch": 0.03324815872538704,
      "flos": 20376038183040.0,
      "grad_norm": 2.990469903058063,
      "language_loss": 0.9301765,
      "learning_rate": 3.999889426006326e-06,
      "loss": 0.95366406,
      "num_input_tokens_seen": 11716955,
      "step": 553,
      "time_per_iteration": 2.6629648208618164
    },
    {
      "auxiliary_loss_clip": 0.01262345,
      "auxiliary_loss_mlp": 0.01077186,
      "balance_loss_clip": 1.06925786,
      "balance_loss_mlp": 1.04149485,
      "epoch": 0.033308281978055014,
      "flos": 24494560634880.0,
      "grad_norm": 2.1924330874053166,
      "language_loss": 0.78881586,
      "learning_rate": 3.999885292792986e-06,
      "loss": 0.8122111,
      "num_input_tokens_seen": 11736130,
      "step": 554,
      "time_per_iteration": 2.668970823287964
    },
    {
      "auxiliary_loss_clip": 0.01258048,
      "auxiliary_loss_mlp": 0.0108557,
      "balance_loss_clip": 1.06745815,
      "balance_loss_mlp": 1.05045104,
      "epoch": 0.03336840523072298,
      "flos": 23400326666880.0,
      "grad_norm": 2.2144550089326938,
      "language_loss": 0.81971425,
      "learning_rate": 3.999881083743795e-06,
      "loss": 0.84315038,
      "num_input_tokens_seen": 11754425,
      "step": 555,
      "time_per_iteration": 2.610807418823242
    },
    {
      "auxiliary_loss_clip": 0.01264442,
      "auxiliary_loss_mlp": 0.0108339,
      "balance_loss_clip": 1.06914032,
      "balance_loss_mlp": 1.04805672,
      "epoch": 0.03342852848339095,
      "flos": 30550571717760.0,
      "grad_norm": 3.7821745066525487,
      "language_loss": 0.88661897,
      "learning_rate": 3.999876798858914e-06,
      "loss": 0.9100973,
      "num_input_tokens_seen": 11772845,
      "step": 556,
      "time_per_iteration": 2.6288907527923584
    },
    {
      "auxiliary_loss_clip": 0.01262553,
      "auxiliary_loss_mlp": 0.01084158,
      "balance_loss_clip": 1.06896496,
      "balance_loss_mlp": 1.04863358,
      "epoch": 0.03348865173605892,
      "flos": 22893304239360.0,
      "grad_norm": 1.974910128087634,
      "language_loss": 0.83708388,
      "learning_rate": 3.999872438138503e-06,
      "loss": 0.860551,
      "num_input_tokens_seen": 11792850,
      "step": 557,
      "time_per_iteration": 2.649401903152466
    },
    {
      "auxiliary_loss_clip": 0.01268198,
      "auxiliary_loss_mlp": 0.01069057,
      "balance_loss_clip": 1.07400489,
      "balance_loss_mlp": 1.03684711,
      "epoch": 0.03354877498872689,
      "flos": 17676022705920.0,
      "grad_norm": 3.176542206824637,
      "language_loss": 0.94202292,
      "learning_rate": 3.999868001582729e-06,
      "loss": 0.96539545,
      "num_input_tokens_seen": 11809670,
      "step": 558,
      "time_per_iteration": 2.550515651702881
    },
    {
      "auxiliary_loss_clip": 0.01258948,
      "auxiliary_loss_mlp": 0.01074291,
      "balance_loss_clip": 1.06591845,
      "balance_loss_mlp": 1.04036427,
      "epoch": 0.03360889824139486,
      "flos": 21652985658240.0,
      "grad_norm": 2.6619487077732384,
      "language_loss": 0.77115649,
      "learning_rate": 3.99986348919176e-06,
      "loss": 0.79448891,
      "num_input_tokens_seen": 11829665,
      "step": 559,
      "time_per_iteration": 2.729597330093384
    },
    {
      "auxiliary_loss_clip": 0.01261947,
      "auxiliary_loss_mlp": 0.01080822,
      "balance_loss_clip": 1.06835234,
      "balance_loss_mlp": 1.04882574,
      "epoch": 0.033669021494062826,
      "flos": 21795730306560.0,
      "grad_norm": 1.945022837871561,
      "language_loss": 0.87472397,
      "learning_rate": 3.9998589009657675e-06,
      "loss": 0.89815164,
      "num_input_tokens_seen": 11848190,
      "step": 560,
      "time_per_iteration": 2.6082279682159424
    },
    {
      "auxiliary_loss_clip": 0.01257198,
      "auxiliary_loss_mlp": 0.0107356,
      "balance_loss_clip": 1.06704283,
      "balance_loss_mlp": 1.04199314,
      "epoch": 0.0337291447467308,
      "flos": 21866222747520.0,
      "grad_norm": 2.4061219554407502,
      "language_loss": 0.81578708,
      "learning_rate": 3.999854236904925e-06,
      "loss": 0.83909464,
      "num_input_tokens_seen": 11864795,
      "step": 561,
      "time_per_iteration": 2.602193832397461
    },
    {
      "auxiliary_loss_clip": 0.01254722,
      "auxiliary_loss_mlp": 0.01076361,
      "balance_loss_clip": 1.06685936,
      "balance_loss_mlp": 1.04422247,
      "epoch": 0.03378926799939877,
      "flos": 24245951627520.0,
      "grad_norm": 1.683217504050761,
      "language_loss": 0.82320511,
      "learning_rate": 3.999849497009409e-06,
      "loss": 0.84651601,
      "num_input_tokens_seen": 11885275,
      "step": 562,
      "time_per_iteration": 2.675872564315796
    },
    {
      "auxiliary_loss_clip": 0.01262146,
      "auxiliary_loss_mlp": 0.01084212,
      "balance_loss_clip": 1.06894755,
      "balance_loss_mlp": 1.0508337,
      "epoch": 0.033849391252066735,
      "flos": 16507812677760.0,
      "grad_norm": 2.262509698135982,
      "language_loss": 0.84285647,
      "learning_rate": 3.999844681279401e-06,
      "loss": 0.86632001,
      "num_input_tokens_seen": 11903595,
      "step": 563,
      "time_per_iteration": 2.586944103240967
    },
    {
      "auxiliary_loss_clip": 0.01258135,
      "auxiliary_loss_mlp": 0.01083866,
      "balance_loss_clip": 1.0675565,
      "balance_loss_mlp": 1.05094075,
      "epoch": 0.03390951450473471,
      "flos": 15669298609920.0,
      "grad_norm": 2.115200912185494,
      "language_loss": 0.94438875,
      "learning_rate": 3.99983978971508e-06,
      "loss": 0.96780878,
      "num_input_tokens_seen": 11917815,
      "step": 564,
      "time_per_iteration": 2.5444440841674805
    },
    {
      "auxiliary_loss_clip": 0.01259509,
      "auxiliary_loss_mlp": 0.01073406,
      "balance_loss_clip": 1.06518865,
      "balance_loss_mlp": 1.03907406,
      "epoch": 0.03396963775740267,
      "flos": 22674787850880.0,
      "grad_norm": 2.6560391741906924,
      "language_loss": 0.94669235,
      "learning_rate": 3.999834822316635e-06,
      "loss": 0.97002149,
      "num_input_tokens_seen": 11936305,
      "step": 565,
      "time_per_iteration": 2.5614171028137207
    },
    {
      "auxiliary_loss_clip": 0.01150452,
      "auxiliary_loss_mlp": 0.01081579,
      "balance_loss_clip": 1.04835606,
      "balance_loss_mlp": 1.07499874,
      "epoch": 0.034029761010070644,
      "flos": 64392683063040.0,
      "grad_norm": 1.0610477485673708,
      "language_loss": 0.54800498,
      "learning_rate": 3.9998297790842535e-06,
      "loss": 0.57032537,
      "num_input_tokens_seen": 11998940,
      "step": 566,
      "time_per_iteration": 3.229137659072876
    },
    {
      "auxiliary_loss_clip": 0.0126129,
      "auxiliary_loss_mlp": 0.01073482,
      "balance_loss_clip": 1.06798041,
      "balance_loss_mlp": 1.03793335,
      "epoch": 0.034089884262738616,
      "flos": 25004204755200.0,
      "grad_norm": 3.1955261820278564,
      "language_loss": 0.76836932,
      "learning_rate": 3.999824660018126e-06,
      "loss": 0.79171705,
      "num_input_tokens_seen": 12018860,
      "step": 567,
      "time_per_iteration": 2.632741928100586
    },
    {
      "auxiliary_loss_clip": 0.01253596,
      "auxiliary_loss_mlp": 0.01083559,
      "balance_loss_clip": 1.06611466,
      "balance_loss_mlp": 1.05153918,
      "epoch": 0.03415000751540658,
      "flos": 28439096584320.0,
      "grad_norm": 2.115683621050472,
      "language_loss": 0.80834144,
      "learning_rate": 3.999819465118447e-06,
      "loss": 0.83171296,
      "num_input_tokens_seen": 12039675,
      "step": 568,
      "time_per_iteration": 2.7206337451934814
    },
    {
      "auxiliary_loss_clip": 0.01254921,
      "auxiliary_loss_mlp": 0.01082401,
      "balance_loss_clip": 1.06888509,
      "balance_loss_mlp": 1.04940367,
      "epoch": 0.034210130768074554,
      "flos": 21468727866240.0,
      "grad_norm": 1.891360159585894,
      "language_loss": 0.86560667,
      "learning_rate": 3.999814194385413e-06,
      "loss": 0.88897985,
      "num_input_tokens_seen": 12057680,
      "step": 569,
      "time_per_iteration": 2.7271673679351807
    },
    {
      "auxiliary_loss_clip": 0.01255135,
      "auxiliary_loss_mlp": 0.01082251,
      "balance_loss_clip": 1.06644094,
      "balance_loss_mlp": 1.04922962,
      "epoch": 0.03427025402074252,
      "flos": 18697501676160.0,
      "grad_norm": 1.6888504559193653,
      "language_loss": 0.95945716,
      "learning_rate": 3.9998088478192255e-06,
      "loss": 0.982831,
      "num_input_tokens_seen": 12076135,
      "step": 570,
      "time_per_iteration": 2.5918867588043213
    },
    {
      "auxiliary_loss_clip": 0.01255487,
      "auxiliary_loss_mlp": 0.0108066,
      "balance_loss_clip": 1.06228065,
      "balance_loss_mlp": 1.0435617,
      "epoch": 0.03433037727341049,
      "flos": 20849987162880.0,
      "grad_norm": 2.39132447086081,
      "language_loss": 0.7964232,
      "learning_rate": 3.9998034254200846e-06,
      "loss": 0.8197847,
      "num_input_tokens_seen": 12094785,
      "step": 571,
      "time_per_iteration": 2.590184450149536
    },
    {
      "auxiliary_loss_clip": 0.01256218,
      "auxiliary_loss_mlp": 0.01091484,
      "balance_loss_clip": 1.06740785,
      "balance_loss_mlp": 1.0565083,
      "epoch": 0.03439050052607846,
      "flos": 25410282986880.0,
      "grad_norm": 2.0738695690993,
      "language_loss": 0.80214274,
      "learning_rate": 3.999797927188199e-06,
      "loss": 0.82561976,
      "num_input_tokens_seen": 12114590,
      "step": 572,
      "time_per_iteration": 2.6862123012542725
    },
    {
      "auxiliary_loss_clip": 0.01263024,
      "auxiliary_loss_mlp": 0.01074173,
      "balance_loss_clip": 1.06995344,
      "balance_loss_mlp": 1.04098535,
      "epoch": 0.03445062377874643,
      "flos": 17640147997440.0,
      "grad_norm": 2.2324763929909284,
      "language_loss": 0.84548658,
      "learning_rate": 3.999792353123774e-06,
      "loss": 0.86885858,
      "num_input_tokens_seen": 12132390,
      "step": 573,
      "time_per_iteration": 2.78487229347229
    },
    {
      "auxiliary_loss_clip": 0.01256326,
      "auxiliary_loss_mlp": 0.01068789,
      "balance_loss_clip": 1.0644815,
      "balance_loss_mlp": 1.03781831,
      "epoch": 0.0345107470314144,
      "flos": 16764502245120.0,
      "grad_norm": 2.576428901855709,
      "language_loss": 0.76602584,
      "learning_rate": 3.999786703227023e-06,
      "loss": 0.78927696,
      "num_input_tokens_seen": 12149035,
      "step": 574,
      "time_per_iteration": 2.5697100162506104
    },
    {
      "auxiliary_loss_clip": 0.01255191,
      "auxiliary_loss_mlp": 0.0107671,
      "balance_loss_clip": 1.06581593,
      "balance_loss_mlp": 1.04502439,
      "epoch": 0.03457087028408237,
      "flos": 14684448533760.0,
      "grad_norm": 2.156110110571344,
      "language_loss": 0.83854586,
      "learning_rate": 3.9997809774981606e-06,
      "loss": 0.86186486,
      "num_input_tokens_seen": 12167530,
      "step": 575,
      "time_per_iteration": 2.596418619155884
    },
    {
      "auxiliary_loss_clip": 0.01249695,
      "auxiliary_loss_mlp": 0.01076053,
      "balance_loss_clip": 1.06684637,
      "balance_loss_mlp": 1.04334211,
      "epoch": 0.03463099353675034,
      "flos": 20011293527040.0,
      "grad_norm": 2.350120742735315,
      "language_loss": 0.83990753,
      "learning_rate": 3.9997751759374025e-06,
      "loss": 0.86316502,
      "num_input_tokens_seen": 12186340,
      "step": 576,
      "time_per_iteration": 5.821930646896362
    },
    {
      "auxiliary_loss_clip": 0.01257114,
      "auxiliary_loss_mlp": 0.01079503,
      "balance_loss_clip": 1.07237518,
      "balance_loss_mlp": 1.04817426,
      "epoch": 0.03469111678941831,
      "flos": 25301150490240.0,
      "grad_norm": 2.138457686407641,
      "language_loss": 0.85803086,
      "learning_rate": 3.99976929854497e-06,
      "loss": 0.88139701,
      "num_input_tokens_seen": 12204090,
      "step": 577,
      "time_per_iteration": 4.225277423858643
    },
    {
      "auxiliary_loss_clip": 0.01253845,
      "auxiliary_loss_mlp": 0.01080214,
      "balance_loss_clip": 1.06869018,
      "balance_loss_mlp": 1.04712176,
      "epoch": 0.034751240042086275,
      "flos": 23259413612160.0,
      "grad_norm": 4.535240156776142,
      "language_loss": 0.72226608,
      "learning_rate": 3.9997633453210845e-06,
      "loss": 0.74560666,
      "num_input_tokens_seen": 12224850,
      "step": 578,
      "time_per_iteration": 4.486239433288574
    },
    {
      "auxiliary_loss_clip": 0.01251871,
      "auxiliary_loss_mlp": 0.01080519,
      "balance_loss_clip": 1.06461096,
      "balance_loss_mlp": 1.04663968,
      "epoch": 0.03481136329475425,
      "flos": 23769237300480.0,
      "grad_norm": 1.9496379050984929,
      "language_loss": 0.77785492,
      "learning_rate": 3.999757316265973e-06,
      "loss": 0.80117887,
      "num_input_tokens_seen": 12244935,
      "step": 579,
      "time_per_iteration": 2.6706583499908447
    },
    {
      "auxiliary_loss_clip": 0.01251647,
      "auxiliary_loss_mlp": 0.01087497,
      "balance_loss_clip": 1.06656826,
      "balance_loss_mlp": 1.05435717,
      "epoch": 0.03487148654742222,
      "flos": 20157521794560.0,
      "grad_norm": 2.054973215074824,
      "language_loss": 0.86841297,
      "learning_rate": 3.999751211379863e-06,
      "loss": 0.8918044,
      "num_input_tokens_seen": 12262140,
      "step": 580,
      "time_per_iteration": 2.639146566390991
    },
    {
      "auxiliary_loss_clip": 0.01256528,
      "auxiliary_loss_mlp": 0.01069029,
      "balance_loss_clip": 1.06636667,
      "balance_loss_mlp": 1.0398469,
      "epoch": 0.034931609800090184,
      "flos": 15669585918720.0,
      "grad_norm": 2.205850105033732,
      "language_loss": 0.82570344,
      "learning_rate": 3.999745030662987e-06,
      "loss": 0.84895897,
      "num_input_tokens_seen": 12280930,
      "step": 581,
      "time_per_iteration": 2.6505649089813232
    },
    {
      "auxiliary_loss_clip": 0.01252942,
      "auxiliary_loss_mlp": 0.01072317,
      "balance_loss_clip": 1.06823969,
      "balance_loss_mlp": 1.04168022,
      "epoch": 0.034991733052758156,
      "flos": 16362374509440.0,
      "grad_norm": 2.1922492117358146,
      "language_loss": 0.7733047,
      "learning_rate": 3.99973877411558e-06,
      "loss": 0.79655731,
      "num_input_tokens_seen": 12299125,
      "step": 582,
      "time_per_iteration": 2.7323596477508545
    },
    {
      "auxiliary_loss_clip": 0.01250253,
      "auxiliary_loss_mlp": 0.01082356,
      "balance_loss_clip": 1.06794167,
      "balance_loss_mlp": 1.04861939,
      "epoch": 0.03505185630542612,
      "flos": 19387309438080.0,
      "grad_norm": 2.1536178016194327,
      "language_loss": 0.87679923,
      "learning_rate": 3.999732441737877e-06,
      "loss": 0.90012532,
      "num_input_tokens_seen": 12316905,
      "step": 583,
      "time_per_iteration": 2.6049294471740723
    },
    {
      "auxiliary_loss_clip": 0.01255473,
      "auxiliary_loss_mlp": 0.01092826,
      "balance_loss_clip": 1.06699181,
      "balance_loss_mlp": 1.06104505,
      "epoch": 0.03511197955809409,
      "flos": 21323828401920.0,
      "grad_norm": 3.7027110169592015,
      "language_loss": 0.81196821,
      "learning_rate": 3.99972603353012e-06,
      "loss": 0.83545119,
      "num_input_tokens_seen": 12335070,
      "step": 584,
      "time_per_iteration": 2.6011815071105957
    },
    {
      "auxiliary_loss_clip": 0.01251161,
      "auxiliary_loss_mlp": 0.01069463,
      "balance_loss_clip": 1.06472683,
      "balance_loss_mlp": 1.03832567,
      "epoch": 0.035172102810762065,
      "flos": 14136595320960.0,
      "grad_norm": 3.067717812226321,
      "language_loss": 0.92399198,
      "learning_rate": 3.999719549492551e-06,
      "loss": 0.94719815,
      "num_input_tokens_seen": 12350315,
      "step": 585,
      "time_per_iteration": 2.5592780113220215
    },
    {
      "auxiliary_loss_clip": 0.01251271,
      "auxiliary_loss_mlp": 0.01077423,
      "balance_loss_clip": 1.06562734,
      "balance_loss_mlp": 1.04552317,
      "epoch": 0.03523222606343003,
      "flos": 20296890564480.0,
      "grad_norm": 2.196660024103635,
      "language_loss": 0.87644351,
      "learning_rate": 3.9997129896254165e-06,
      "loss": 0.89973044,
      "num_input_tokens_seen": 12366030,
      "step": 586,
      "time_per_iteration": 2.5486221313476562
    },
    {
      "auxiliary_loss_clip": 0.01256485,
      "auxiliary_loss_mlp": 0.0108018,
      "balance_loss_clip": 1.06803596,
      "balance_loss_mlp": 1.04918551,
      "epoch": 0.035292349316098,
      "flos": 20375822701440.0,
      "grad_norm": 2.1222089199850878,
      "language_loss": 0.76079381,
      "learning_rate": 3.999706353928965e-06,
      "loss": 0.78416049,
      "num_input_tokens_seen": 12384895,
      "step": 587,
      "time_per_iteration": 2.5923714637756348
    },
    {
      "auxiliary_loss_clip": 0.01257125,
      "auxiliary_loss_mlp": 0.01068649,
      "balance_loss_clip": 1.06683922,
      "balance_loss_mlp": 1.03586686,
      "epoch": 0.03535247256876597,
      "flos": 21468871520640.0,
      "grad_norm": 2.212352192395094,
      "language_loss": 0.78601038,
      "learning_rate": 3.999699642403449e-06,
      "loss": 0.80926806,
      "num_input_tokens_seen": 12404980,
      "step": 588,
      "time_per_iteration": 2.579280138015747
    },
    {
      "auxiliary_loss_clip": 0.0125398,
      "auxiliary_loss_mlp": 0.0107827,
      "balance_loss_clip": 1.06582928,
      "balance_loss_mlp": 1.04367518,
      "epoch": 0.03541259582143394,
      "flos": 23623044946560.0,
      "grad_norm": 2.153589114745919,
      "language_loss": 0.94312829,
      "learning_rate": 3.99969285504912e-06,
      "loss": 0.96645081,
      "num_input_tokens_seen": 12423835,
      "step": 589,
      "time_per_iteration": 2.5964701175689697
    },
    {
      "auxiliary_loss_clip": 0.01256884,
      "auxiliary_loss_mlp": 0.01078108,
      "balance_loss_clip": 1.06697679,
      "balance_loss_mlp": 1.04666042,
      "epoch": 0.03547271907410191,
      "flos": 33726367768320.0,
      "grad_norm": 2.1162556876212695,
      "language_loss": 0.84116042,
      "learning_rate": 3.99968599186624e-06,
      "loss": 0.8645103,
      "num_input_tokens_seen": 12443135,
      "step": 590,
      "time_per_iteration": 2.746436357498169
    },
    {
      "auxiliary_loss_clip": 0.01249398,
      "auxiliary_loss_mlp": 0.01068452,
      "balance_loss_clip": 1.06658125,
      "balance_loss_mlp": 1.03893578,
      "epoch": 0.03553284232676988,
      "flos": 21142695093120.0,
      "grad_norm": 1.984522351394552,
      "language_loss": 0.8684091,
      "learning_rate": 3.999679052855065e-06,
      "loss": 0.89158762,
      "num_input_tokens_seen": 12462895,
      "step": 591,
      "time_per_iteration": 2.692303419113159
    },
    {
      "auxiliary_loss_clip": 0.01250641,
      "auxiliary_loss_mlp": 0.01082122,
      "balance_loss_clip": 1.06297326,
      "balance_loss_mlp": 1.04883862,
      "epoch": 0.03559296557943785,
      "flos": 20046593617920.0,
      "grad_norm": 2.0873185001780783,
      "language_loss": 0.83075488,
      "learning_rate": 3.999672038015861e-06,
      "loss": 0.85408247,
      "num_input_tokens_seen": 12481515,
      "step": 592,
      "time_per_iteration": 2.7822203636169434
    },
    {
      "auxiliary_loss_clip": 0.01146211,
      "auxiliary_loss_mlp": 0.01034159,
      "balance_loss_clip": 1.05013406,
      "balance_loss_mlp": 1.02676773,
      "epoch": 0.035653088832105814,
      "flos": 60334597244160.0,
      "grad_norm": 0.8804992705477848,
      "language_loss": 0.59754086,
      "learning_rate": 3.999664947348893e-06,
      "loss": 0.61934447,
      "num_input_tokens_seen": 12548220,
      "step": 593,
      "time_per_iteration": 3.274080276489258
    },
    {
      "auxiliary_loss_clip": 0.01249386,
      "auxiliary_loss_mlp": 0.0107742,
      "balance_loss_clip": 1.06737614,
      "balance_loss_mlp": 1.04473329,
      "epoch": 0.035713212084773786,
      "flos": 20113135562880.0,
      "grad_norm": 1.8086551314359374,
      "language_loss": 0.87077361,
      "learning_rate": 3.999657780854429e-06,
      "loss": 0.89404166,
      "num_input_tokens_seen": 12566105,
      "step": 594,
      "time_per_iteration": 2.682236671447754
    },
    {
      "auxiliary_loss_clip": 0.012487,
      "auxiliary_loss_mlp": 0.01082358,
      "balance_loss_clip": 1.06235993,
      "balance_loss_mlp": 1.05057716,
      "epoch": 0.03577333533744176,
      "flos": 26285785084800.0,
      "grad_norm": 5.516524335860627,
      "language_loss": 0.83920246,
      "learning_rate": 3.999650538532742e-06,
      "loss": 0.86251307,
      "num_input_tokens_seen": 12586680,
      "step": 595,
      "time_per_iteration": 2.773669481277466
    },
    {
      "auxiliary_loss_clip": 0.01248678,
      "auxiliary_loss_mlp": 0.01090544,
      "balance_loss_clip": 1.06579614,
      "balance_loss_mlp": 1.05850017,
      "epoch": 0.035833458590109724,
      "flos": 10889732211840.0,
      "grad_norm": 2.3448814752825204,
      "language_loss": 0.96041518,
      "learning_rate": 3.999643220384106e-06,
      "loss": 0.98380733,
      "num_input_tokens_seen": 12601605,
      "step": 596,
      "time_per_iteration": 2.6541590690612793
    },
    {
      "auxiliary_loss_clip": 0.01252662,
      "auxiliary_loss_mlp": 0.01081887,
      "balance_loss_clip": 1.0675534,
      "balance_loss_mlp": 1.05165553,
      "epoch": 0.035893581842777696,
      "flos": 22090198003200.0,
      "grad_norm": 2.4353221882859004,
      "language_loss": 0.82993281,
      "learning_rate": 3.999635826408799e-06,
      "loss": 0.85327828,
      "num_input_tokens_seen": 12620365,
      "step": 597,
      "time_per_iteration": 2.7023818492889404
    },
    {
      "auxiliary_loss_clip": 0.01247839,
      "auxiliary_loss_mlp": 0.01079829,
      "balance_loss_clip": 1.0668776,
      "balance_loss_mlp": 1.04766583,
      "epoch": 0.03595370509544566,
      "flos": 23038347358080.0,
      "grad_norm": 2.374757318483944,
      "language_loss": 0.81364304,
      "learning_rate": 3.999628356607101e-06,
      "loss": 0.83691972,
      "num_input_tokens_seen": 12641140,
      "step": 598,
      "time_per_iteration": 2.731229782104492
    },
    {
      "auxiliary_loss_clip": 0.01243692,
      "auxiliary_loss_mlp": 0.01077827,
      "balance_loss_clip": 1.0663228,
      "balance_loss_mlp": 1.04587913,
      "epoch": 0.03601382834811363,
      "flos": 20777734955520.0,
      "grad_norm": 1.817680341814684,
      "language_loss": 0.81172699,
      "learning_rate": 3.999620810979295e-06,
      "loss": 0.83494222,
      "num_input_tokens_seen": 12661080,
      "step": 599,
      "time_per_iteration": 2.710191011428833
    },
    {
      "auxiliary_loss_clip": 0.01250419,
      "auxiliary_loss_mlp": 0.01074577,
      "balance_loss_clip": 1.06356514,
      "balance_loss_mlp": 1.045228,
      "epoch": 0.036073951600781605,
      "flos": 23951627585280.0,
      "grad_norm": 2.3963649020429627,
      "language_loss": 0.8651731,
      "learning_rate": 3.999613189525668e-06,
      "loss": 0.88842309,
      "num_input_tokens_seen": 12678270,
      "step": 600,
      "time_per_iteration": 2.682262420654297
    },
    {
      "auxiliary_loss_clip": 0.01241882,
      "auxiliary_loss_mlp": 0.01084809,
      "balance_loss_clip": 1.05918193,
      "balance_loss_mlp": 1.05297971,
      "epoch": 0.03613407485344957,
      "flos": 18912283050240.0,
      "grad_norm": 2.0308947613075423,
      "language_loss": 0.82355881,
      "learning_rate": 3.999605492246508e-06,
      "loss": 0.84682572,
      "num_input_tokens_seen": 12697295,
      "step": 601,
      "time_per_iteration": 2.6570894718170166
    },
    {
      "auxiliary_loss_clip": 0.01240868,
      "auxiliary_loss_mlp": 0.010708,
      "balance_loss_clip": 1.06129336,
      "balance_loss_mlp": 1.03920949,
      "epoch": 0.03619419810611754,
      "flos": 23038526926080.0,
      "grad_norm": 2.3080142694085555,
      "language_loss": 0.7502507,
      "learning_rate": 3.999597719142107e-06,
      "loss": 0.77336735,
      "num_input_tokens_seen": 12716165,
      "step": 602,
      "time_per_iteration": 2.6434237957000732
    },
    {
      "auxiliary_loss_clip": 0.01239543,
      "auxiliary_loss_mlp": 0.01066859,
      "balance_loss_clip": 1.0604254,
      "balance_loss_mlp": 1.03562629,
      "epoch": 0.03625432135878551,
      "flos": 29457774293760.0,
      "grad_norm": 1.9681237382646195,
      "language_loss": 0.79599822,
      "learning_rate": 3.999589870212761e-06,
      "loss": 0.81906223,
      "num_input_tokens_seen": 12735475,
      "step": 603,
      "time_per_iteration": 2.7201666831970215
    },
    {
      "auxiliary_loss_clip": 0.01244834,
      "auxiliary_loss_mlp": 0.01071177,
      "balance_loss_clip": 1.06545615,
      "balance_loss_mlp": 1.04130292,
      "epoch": 0.03631444461145348,
      "flos": 23508525409920.0,
      "grad_norm": 1.8363641170913294,
      "language_loss": 0.86668456,
      "learning_rate": 3.9995819454587664e-06,
      "loss": 0.88984472,
      "num_input_tokens_seen": 12754540,
      "step": 604,
      "time_per_iteration": 2.60249924659729
    },
    {
      "auxiliary_loss_clip": 0.01248906,
      "auxiliary_loss_mlp": 0.01072985,
      "balance_loss_clip": 1.0674324,
      "balance_loss_mlp": 1.04010737,
      "epoch": 0.03637456786412145,
      "flos": 16618130323200.0,
      "grad_norm": 2.510130211393037,
      "language_loss": 0.80746496,
      "learning_rate": 3.999573944880424e-06,
      "loss": 0.83068383,
      "num_input_tokens_seen": 12773050,
      "step": 605,
      "time_per_iteration": 2.766684055328369
    },
    {
      "auxiliary_loss_clip": 0.01244274,
      "auxiliary_loss_mlp": 0.0107873,
      "balance_loss_clip": 1.0630821,
      "balance_loss_mlp": 1.04846251,
      "epoch": 0.03643469111678942,
      "flos": 15851832549120.0,
      "grad_norm": 2.2216143800596835,
      "language_loss": 0.85942292,
      "learning_rate": 3.9995658684780375e-06,
      "loss": 0.882653,
      "num_input_tokens_seen": 12791240,
      "step": 606,
      "time_per_iteration": 2.6133925914764404
    },
    {
      "auxiliary_loss_clip": 0.01247732,
      "auxiliary_loss_mlp": 0.01077404,
      "balance_loss_clip": 1.06413972,
      "balance_loss_mlp": 1.04588532,
      "epoch": 0.03649481436945739,
      "flos": 23620387340160.0,
      "grad_norm": 2.0684825764003394,
      "language_loss": 0.82179952,
      "learning_rate": 3.999557716251912e-06,
      "loss": 0.84505081,
      "num_input_tokens_seen": 12812245,
      "step": 607,
      "time_per_iteration": 2.6805856227874756
    },
    {
      "auxiliary_loss_clip": 0.01245394,
      "auxiliary_loss_mlp": 0.01073743,
      "balance_loss_clip": 1.06585169,
      "balance_loss_mlp": 1.04317796,
      "epoch": 0.036554937622125354,
      "flos": 21755581879680.0,
      "grad_norm": 2.3717179235904533,
      "language_loss": 0.83567071,
      "learning_rate": 3.999549488202358e-06,
      "loss": 0.8588621,
      "num_input_tokens_seen": 12831085,
      "step": 608,
      "time_per_iteration": 2.6593453884124756
    },
    {
      "auxiliary_loss_clip": 0.01251062,
      "auxiliary_loss_mlp": 0.01073705,
      "balance_loss_clip": 1.06682992,
      "balance_loss_mlp": 1.04006422,
      "epoch": 0.036615060874793326,
      "flos": 17819772935040.0,
      "grad_norm": 2.4795108668903305,
      "language_loss": 0.8201133,
      "learning_rate": 3.999541184329688e-06,
      "loss": 0.84336102,
      "num_input_tokens_seen": 12849115,
      "step": 609,
      "time_per_iteration": 2.6299383640289307
    },
    {
      "auxiliary_loss_clip": 0.01255655,
      "auxiliary_loss_mlp": 0.01091893,
      "balance_loss_clip": 1.07322037,
      "balance_loss_mlp": 1.06158984,
      "epoch": 0.0366751841274613,
      "flos": 26753808320640.0,
      "grad_norm": 1.992640540297191,
      "language_loss": 0.79448462,
      "learning_rate": 3.999532804634215e-06,
      "loss": 0.81796008,
      "num_input_tokens_seen": 12868005,
      "step": 610,
      "time_per_iteration": 2.65120530128479
    },
    {
      "auxiliary_loss_clip": 0.01254423,
      "auxiliary_loss_mlp": 0.01088228,
      "balance_loss_clip": 1.06914616,
      "balance_loss_mlp": 1.05656588,
      "epoch": 0.03673530738012926,
      "flos": 22196960202240.0,
      "grad_norm": 1.9328503999291824,
      "language_loss": 0.87282723,
      "learning_rate": 3.9995243491162575e-06,
      "loss": 0.89625371,
      "num_input_tokens_seen": 12886890,
      "step": 611,
      "time_per_iteration": 2.7398059368133545
    },
    {
      "auxiliary_loss_clip": 0.01248885,
      "auxiliary_loss_mlp": 0.01097673,
      "balance_loss_clip": 1.06917143,
      "balance_loss_mlp": 1.06651139,
      "epoch": 0.036795430632797235,
      "flos": 24681655601280.0,
      "grad_norm": 3.7435200854847266,
      "language_loss": 0.72589231,
      "learning_rate": 3.999515817776136e-06,
      "loss": 0.74935788,
      "num_input_tokens_seen": 12906130,
      "step": 612,
      "time_per_iteration": 2.700406551361084
    },
    {
      "auxiliary_loss_clip": 0.01249112,
      "auxiliary_loss_mlp": 0.01076924,
      "balance_loss_clip": 1.06581926,
      "balance_loss_mlp": 1.04480934,
      "epoch": 0.0368555538854652,
      "flos": 17748921358080.0,
      "grad_norm": 3.0863603820013434,
      "language_loss": 0.79110008,
      "learning_rate": 3.999507210614175e-06,
      "loss": 0.81436038,
      "num_input_tokens_seen": 12925260,
      "step": 613,
      "time_per_iteration": 2.630472183227539
    },
    {
      "auxiliary_loss_clip": 0.01242581,
      "auxiliary_loss_mlp": 0.01090278,
      "balance_loss_clip": 1.06378841,
      "balance_loss_mlp": 1.05961776,
      "epoch": 0.03691567713813317,
      "flos": 20594554571520.0,
      "grad_norm": 2.2015687298668336,
      "language_loss": 0.93885028,
      "learning_rate": 3.9994985276307e-06,
      "loss": 0.96217889,
      "num_input_tokens_seen": 12944590,
      "step": 614,
      "time_per_iteration": 2.6977972984313965
    },
    {
      "auxiliary_loss_clip": 0.01254503,
      "auxiliary_loss_mlp": 0.01081137,
      "balance_loss_clip": 1.07009673,
      "balance_loss_mlp": 1.04732919,
      "epoch": 0.036975800390801145,
      "flos": 33650380546560.0,
      "grad_norm": 3.0661216019279576,
      "language_loss": 0.72932875,
      "learning_rate": 3.999489768826041e-06,
      "loss": 0.75268513,
      "num_input_tokens_seen": 12964785,
      "step": 615,
      "time_per_iteration": 2.697291612625122
    },
    {
      "auxiliary_loss_clip": 0.01250213,
      "auxiliary_loss_mlp": 0.010716,
      "balance_loss_clip": 1.06649876,
      "balance_loss_mlp": 1.04015231,
      "epoch": 0.03703592364346911,
      "flos": 28293694329600.0,
      "grad_norm": 2.9941392641088695,
      "language_loss": 0.81630868,
      "learning_rate": 3.999480934200528e-06,
      "loss": 0.83952683,
      "num_input_tokens_seen": 12986705,
      "step": 616,
      "time_per_iteration": 4.1762495040893555
    },
    {
      "auxiliary_loss_clip": 0.0124999,
      "auxiliary_loss_mlp": 0.01076541,
      "balance_loss_clip": 1.06807041,
      "balance_loss_mlp": 1.0467627,
      "epoch": 0.03709604689613708,
      "flos": 31504215853440.0,
      "grad_norm": 2.320593216419041,
      "language_loss": 0.68178958,
      "learning_rate": 3.999472023754499e-06,
      "loss": 0.70505488,
      "num_input_tokens_seen": 13010560,
      "step": 617,
      "time_per_iteration": 4.224538564682007
    },
    {
      "auxiliary_loss_clip": 0.01254259,
      "auxiliary_loss_mlp": 0.010771,
      "balance_loss_clip": 1.07098567,
      "balance_loss_mlp": 1.04415071,
      "epoch": 0.03715617014880505,
      "flos": 19609381272960.0,
      "grad_norm": 2.245411088847763,
      "language_loss": 0.80595517,
      "learning_rate": 3.99946303748829e-06,
      "loss": 0.82926875,
      "num_input_tokens_seen": 13028935,
      "step": 618,
      "time_per_iteration": 4.200341463088989
    },
    {
      "auxiliary_loss_clip": 0.01257669,
      "auxiliary_loss_mlp": 0.01079294,
      "balance_loss_clip": 1.06808555,
      "balance_loss_mlp": 1.04605901,
      "epoch": 0.03721629340147302,
      "flos": 15924192497280.0,
      "grad_norm": 10.155035046705617,
      "language_loss": 0.91591841,
      "learning_rate": 3.999453975402242e-06,
      "loss": 0.93928802,
      "num_input_tokens_seen": 13046000,
      "step": 619,
      "time_per_iteration": 2.5787301063537598
    },
    {
      "auxiliary_loss_clip": 0.01251145,
      "auxiliary_loss_mlp": 0.01083548,
      "balance_loss_clip": 1.06999123,
      "balance_loss_mlp": 1.05181432,
      "epoch": 0.03727641665414099,
      "flos": 21104090951040.0,
      "grad_norm": 2.0803022158745406,
      "language_loss": 0.94071603,
      "learning_rate": 3.9994448374967e-06,
      "loss": 0.96406299,
      "num_input_tokens_seen": 13062995,
      "step": 620,
      "time_per_iteration": 2.5987205505371094
    },
    {
      "auxiliary_loss_clip": 0.01249568,
      "auxiliary_loss_mlp": 0.0108317,
      "balance_loss_clip": 1.06624317,
      "balance_loss_mlp": 1.0502919,
      "epoch": 0.037336539906808956,
      "flos": 24131683486080.0,
      "grad_norm": 1.7431896174296577,
      "language_loss": 0.77319217,
      "learning_rate": 3.999435623772008e-06,
      "loss": 0.79651952,
      "num_input_tokens_seen": 13084120,
      "step": 621,
      "time_per_iteration": 2.68758225440979
    },
    {
      "auxiliary_loss_clip": 0.01247252,
      "auxiliary_loss_mlp": 0.01071013,
      "balance_loss_clip": 1.06894088,
      "balance_loss_mlp": 1.03792048,
      "epoch": 0.03739666315947693,
      "flos": 22346384780160.0,
      "grad_norm": 2.3852872810563364,
      "language_loss": 0.86546707,
      "learning_rate": 3.999426334228518e-06,
      "loss": 0.88864976,
      "num_input_tokens_seen": 13100035,
      "step": 622,
      "time_per_iteration": 2.607121467590332
    },
    {
      "auxiliary_loss_clip": 0.012499,
      "auxiliary_loss_mlp": 0.01072461,
      "balance_loss_clip": 1.06715882,
      "balance_loss_mlp": 1.04048872,
      "epoch": 0.0374567864121449,
      "flos": 20449511452800.0,
      "grad_norm": 2.2621736327299766,
      "language_loss": 0.90008956,
      "learning_rate": 3.999416968866581e-06,
      "loss": 0.92331314,
      "num_input_tokens_seen": 13118070,
      "step": 623,
      "time_per_iteration": 2.6513512134552
    },
    {
      "auxiliary_loss_clip": 0.01251762,
      "auxiliary_loss_mlp": 0.01090534,
      "balance_loss_clip": 1.07006013,
      "balance_loss_mlp": 1.05844235,
      "epoch": 0.037516909664812866,
      "flos": 19208043636480.0,
      "grad_norm": 2.760597076727266,
      "language_loss": 0.84095174,
      "learning_rate": 3.999407527686551e-06,
      "loss": 0.8643747,
      "num_input_tokens_seen": 13136355,
      "step": 624,
      "time_per_iteration": 2.66623592376709
    },
    {
      "auxiliary_loss_clip": 0.01252431,
      "auxiliary_loss_mlp": 0.01076353,
      "balance_loss_clip": 1.06697702,
      "balance_loss_mlp": 1.04423809,
      "epoch": 0.03757703291748084,
      "flos": 35005218664320.0,
      "grad_norm": 4.259276014089895,
      "language_loss": 0.66778994,
      "learning_rate": 3.999398010688788e-06,
      "loss": 0.69107783,
      "num_input_tokens_seen": 13155435,
      "step": 625,
      "time_per_iteration": 2.7288877964019775
    },
    {
      "auxiliary_loss_clip": 0.01244959,
      "auxiliary_loss_mlp": 0.01076274,
      "balance_loss_clip": 1.06605244,
      "balance_loss_mlp": 1.042943,
      "epoch": 0.0376371561701488,
      "flos": 25483899911040.0,
      "grad_norm": 3.375450269409945,
      "language_loss": 0.77496696,
      "learning_rate": 3.999388417873652e-06,
      "loss": 0.79817927,
      "num_input_tokens_seen": 13174295,
      "step": 626,
      "time_per_iteration": 2.648942470550537
    },
    {
      "auxiliary_loss_clip": 0.01249107,
      "auxiliary_loss_mlp": 0.0108376,
      "balance_loss_clip": 1.06770003,
      "balance_loss_mlp": 1.05200303,
      "epoch": 0.037697279422816775,
      "flos": 18185630912640.0,
      "grad_norm": 2.0480468386724766,
      "language_loss": 0.81463408,
      "learning_rate": 3.999378749241506e-06,
      "loss": 0.83796275,
      "num_input_tokens_seen": 13192500,
      "step": 627,
      "time_per_iteration": 2.6209845542907715
    },
    {
      "auxiliary_loss_clip": 0.01254363,
      "auxiliary_loss_mlp": 0.01084942,
      "balance_loss_clip": 1.07041132,
      "balance_loss_mlp": 1.05215955,
      "epoch": 0.03775740267548475,
      "flos": 24644272521600.0,
      "grad_norm": 1.6934072791943036,
      "language_loss": 0.88809037,
      "learning_rate": 3.999369004792719e-06,
      "loss": 0.91148341,
      "num_input_tokens_seen": 13213470,
      "step": 628,
      "time_per_iteration": 2.7221415042877197
    },
    {
      "auxiliary_loss_clip": 0.01247303,
      "auxiliary_loss_mlp": 0.01080197,
      "balance_loss_clip": 1.0627017,
      "balance_loss_mlp": 1.04765344,
      "epoch": 0.03781752592815271,
      "flos": 21288205088640.0,
      "grad_norm": 2.536151380104699,
      "language_loss": 0.79840028,
      "learning_rate": 3.999359184527658e-06,
      "loss": 0.82167524,
      "num_input_tokens_seen": 13232365,
      "step": 629,
      "time_per_iteration": 2.6535024642944336
    },
    {
      "auxiliary_loss_clip": 0.01249218,
      "auxiliary_loss_mlp": 0.0106958,
      "balance_loss_clip": 1.06675959,
      "balance_loss_mlp": 1.03885961,
      "epoch": 0.037877649180820684,
      "flos": 22089623385600.0,
      "grad_norm": 1.6861994278356789,
      "language_loss": 0.76824844,
      "learning_rate": 3.999349288446696e-06,
      "loss": 0.79143643,
      "num_input_tokens_seen": 13251920,
      "step": 630,
      "time_per_iteration": 2.6175966262817383
    },
    {
      "auxiliary_loss_clip": 0.01254291,
      "auxiliary_loss_mlp": 0.01075963,
      "balance_loss_clip": 1.06833327,
      "balance_loss_mlp": 1.04504025,
      "epoch": 0.03793777243348865,
      "flos": 14501339976960.0,
      "grad_norm": 3.12435515576561,
      "language_loss": 0.91593724,
      "learning_rate": 3.99933931655021e-06,
      "loss": 0.93923974,
      "num_input_tokens_seen": 13267440,
      "step": 631,
      "time_per_iteration": 2.565293788909912
    },
    {
      "auxiliary_loss_clip": 0.01243525,
      "auxiliary_loss_mlp": 0.01087901,
      "balance_loss_clip": 1.06386209,
      "balance_loss_mlp": 1.05356884,
      "epoch": 0.03799789568615662,
      "flos": 21908418249600.0,
      "grad_norm": 1.6822536287963328,
      "language_loss": 0.92157543,
      "learning_rate": 3.999329268838575e-06,
      "loss": 0.94488978,
      "num_input_tokens_seen": 13287850,
      "step": 632,
      "time_per_iteration": 2.6235203742980957
    },
    {
      "auxiliary_loss_clip": 0.01248362,
      "auxiliary_loss_mlp": 0.01067296,
      "balance_loss_clip": 1.06696796,
      "balance_loss_mlp": 1.03613472,
      "epoch": 0.03805801893882459,
      "flos": 24827021942400.0,
      "grad_norm": 2.1097171792430456,
      "language_loss": 0.83139223,
      "learning_rate": 3.999319145312175e-06,
      "loss": 0.85454881,
      "num_input_tokens_seen": 13307760,
      "step": 633,
      "time_per_iteration": 2.6461985111236572
    },
    {
      "auxiliary_loss_clip": 0.01247735,
      "auxiliary_loss_mlp": 0.01079895,
      "balance_loss_clip": 1.06473529,
      "balance_loss_mlp": 1.04811358,
      "epoch": 0.03811814219149256,
      "flos": 30482952364800.0,
      "grad_norm": 1.599115294194595,
      "language_loss": 0.69883299,
      "learning_rate": 3.999308945971392e-06,
      "loss": 0.72210932,
      "num_input_tokens_seen": 13331230,
      "step": 634,
      "time_per_iteration": 2.709033727645874
    },
    {
      "auxiliary_loss_clip": 0.01133204,
      "auxiliary_loss_mlp": 0.01009504,
      "balance_loss_clip": 1.04124916,
      "balance_loss_mlp": 1.00249422,
      "epoch": 0.03817826544416053,
      "flos": 66992577379200.0,
      "grad_norm": 0.893126545279708,
      "language_loss": 0.61645919,
      "learning_rate": 3.999298670816614e-06,
      "loss": 0.63788629,
      "num_input_tokens_seen": 13394760,
      "step": 635,
      "time_per_iteration": 3.2099475860595703
    },
    {
      "auxiliary_loss_clip": 0.01244276,
      "auxiliary_loss_mlp": 0.01072984,
      "balance_loss_clip": 1.06475401,
      "balance_loss_mlp": 1.04129851,
      "epoch": 0.038238388696828496,
      "flos": 20485350247680.0,
      "grad_norm": 2.0563589539657205,
      "language_loss": 0.83629507,
      "learning_rate": 3.9992883198482294e-06,
      "loss": 0.85946769,
      "num_input_tokens_seen": 13412775,
      "step": 636,
      "time_per_iteration": 2.6278960704803467
    },
    {
      "auxiliary_loss_clip": 0.01248078,
      "auxiliary_loss_mlp": 0.01096471,
      "balance_loss_clip": 1.06714165,
      "balance_loss_mlp": 1.06530952,
      "epoch": 0.03829851194949647,
      "flos": 17965893461760.0,
      "grad_norm": 2.346379148367956,
      "language_loss": 0.79578567,
      "learning_rate": 3.999277893066632e-06,
      "loss": 0.81923115,
      "num_input_tokens_seen": 13427835,
      "step": 637,
      "time_per_iteration": 2.646414279937744
    },
    {
      "auxiliary_loss_clip": 0.01247939,
      "auxiliary_loss_mlp": 0.01088528,
      "balance_loss_clip": 1.06356907,
      "balance_loss_mlp": 1.0562222,
      "epoch": 0.03835863520216444,
      "flos": 22456522857600.0,
      "grad_norm": 1.9563283234999833,
      "language_loss": 0.83989692,
      "learning_rate": 3.999267390472215e-06,
      "loss": 0.86326158,
      "num_input_tokens_seen": 13447295,
      "step": 638,
      "time_per_iteration": 2.6416285037994385
    },
    {
      "auxiliary_loss_clip": 0.01253172,
      "auxiliary_loss_mlp": 0.01074704,
      "balance_loss_clip": 1.06563985,
      "balance_loss_mlp": 1.04163575,
      "epoch": 0.038418758454832405,
      "flos": 22164425458560.0,
      "grad_norm": 2.5596504471077224,
      "language_loss": 0.70109725,
      "learning_rate": 3.999256812065381e-06,
      "loss": 0.72437602,
      "num_input_tokens_seen": 13468455,
      "step": 639,
      "time_per_iteration": 2.610682487487793
    },
    {
      "auxiliary_loss_clip": 0.01248829,
      "auxiliary_loss_mlp": 0.01081808,
      "balance_loss_clip": 1.06618333,
      "balance_loss_mlp": 1.04790449,
      "epoch": 0.03847888170750038,
      "flos": 22747435107840.0,
      "grad_norm": 2.5791624605537082,
      "language_loss": 0.85322344,
      "learning_rate": 3.999246157846526e-06,
      "loss": 0.87652987,
      "num_input_tokens_seen": 13489085,
      "step": 640,
      "time_per_iteration": 2.700456380844116
    },
    {
      "auxiliary_loss_clip": 0.01252579,
      "auxiliary_loss_mlp": 0.01083722,
      "balance_loss_clip": 1.06751871,
      "balance_loss_mlp": 1.04934239,
      "epoch": 0.03853900496016834,
      "flos": 22711201263360.0,
      "grad_norm": 2.331268680461456,
      "language_loss": 0.82141805,
      "learning_rate": 3.9992354278160574e-06,
      "loss": 0.84478104,
      "num_input_tokens_seen": 13509120,
      "step": 641,
      "time_per_iteration": 2.6572046279907227
    },
    {
      "auxiliary_loss_clip": 0.0112759,
      "auxiliary_loss_mlp": 0.01008008,
      "balance_loss_clip": 1.03825259,
      "balance_loss_mlp": 1.00095105,
      "epoch": 0.038599128212836314,
      "flos": 70399136355840.0,
      "grad_norm": 0.9037629700551453,
      "language_loss": 0.65444964,
      "learning_rate": 3.999224621974381e-06,
      "loss": 0.67580563,
      "num_input_tokens_seen": 13562005,
      "step": 642,
      "time_per_iteration": 3.199925422668457
    },
    {
      "auxiliary_loss_clip": 0.01246698,
      "auxiliary_loss_mlp": 0.01064563,
      "balance_loss_clip": 1.0651319,
      "balance_loss_mlp": 1.03453398,
      "epoch": 0.03865925146550429,
      "flos": 23295144666240.0,
      "grad_norm": 1.9113268312481755,
      "language_loss": 0.79272145,
      "learning_rate": 3.999213740321906e-06,
      "loss": 0.81583405,
      "num_input_tokens_seen": 13582185,
      "step": 643,
      "time_per_iteration": 2.641437292098999
    },
    {
      "auxiliary_loss_clip": 0.01244786,
      "auxiliary_loss_mlp": 0.01076057,
      "balance_loss_clip": 1.06219232,
      "balance_loss_mlp": 1.04599261,
      "epoch": 0.03871937471817225,
      "flos": 21430446946560.0,
      "grad_norm": 2.2104774200729262,
      "language_loss": 0.8294487,
      "learning_rate": 3.999202782859046e-06,
      "loss": 0.85265714,
      "num_input_tokens_seen": 13599555,
      "step": 644,
      "time_per_iteration": 2.600558280944824
    },
    {
      "auxiliary_loss_clip": 0.01247273,
      "auxiliary_loss_mlp": 0.01074554,
      "balance_loss_clip": 1.06383467,
      "balance_loss_mlp": 1.04193854,
      "epoch": 0.038779497970840224,
      "flos": 34277309550720.0,
      "grad_norm": 1.994902925690418,
      "language_loss": 0.82286513,
      "learning_rate": 3.9991917495862165e-06,
      "loss": 0.8460834,
      "num_input_tokens_seen": 13621160,
      "step": 645,
      "time_per_iteration": 2.6751983165740967
    },
    {
      "auxiliary_loss_clip": 0.01248631,
      "auxiliary_loss_mlp": 0.01070807,
      "balance_loss_clip": 1.06525111,
      "balance_loss_mlp": 1.03890657,
      "epoch": 0.03883962122350819,
      "flos": 22748189293440.0,
      "grad_norm": 2.290384247239265,
      "language_loss": 0.81889713,
      "learning_rate": 3.9991806405038345e-06,
      "loss": 0.84209144,
      "num_input_tokens_seen": 13641915,
      "step": 646,
      "time_per_iteration": 2.6987667083740234
    },
    {
      "auxiliary_loss_clip": 0.01250204,
      "auxiliary_loss_mlp": 0.01078836,
      "balance_loss_clip": 1.06982899,
      "balance_loss_mlp": 1.04791331,
      "epoch": 0.03889974447617616,
      "flos": 21945837242880.0,
      "grad_norm": 1.9171219640425325,
      "language_loss": 0.82015383,
      "learning_rate": 3.999169455612323e-06,
      "loss": 0.84344423,
      "num_input_tokens_seen": 13661410,
      "step": 647,
      "time_per_iteration": 2.590102195739746
    },
    {
      "auxiliary_loss_clip": 0.0124696,
      "auxiliary_loss_mlp": 0.01072111,
      "balance_loss_clip": 1.06628954,
      "balance_loss_mlp": 1.04216528,
      "epoch": 0.03895986772884413,
      "flos": 31504826384640.0,
      "grad_norm": 1.9398424653049293,
      "language_loss": 0.84477997,
      "learning_rate": 3.999158194912106e-06,
      "loss": 0.86797059,
      "num_input_tokens_seen": 13681705,
      "step": 648,
      "time_per_iteration": 2.7516121864318848
    },
    {
      "auxiliary_loss_clip": 0.01244808,
      "auxiliary_loss_mlp": 0.0107293,
      "balance_loss_clip": 1.06524062,
      "balance_loss_mlp": 1.04210222,
      "epoch": 0.0390199909815121,
      "flos": 19901011795200.0,
      "grad_norm": 2.3870859420748136,
      "language_loss": 0.84254295,
      "learning_rate": 3.9991468584036086e-06,
      "loss": 0.86572027,
      "num_input_tokens_seen": 13700400,
      "step": 649,
      "time_per_iteration": 2.6116180419921875
    },
    {
      "auxiliary_loss_clip": 0.01246653,
      "auxiliary_loss_mlp": 0.01073574,
      "balance_loss_clip": 1.06560743,
      "balance_loss_mlp": 1.0416739,
      "epoch": 0.03908011423418007,
      "flos": 21612478095360.0,
      "grad_norm": 2.00775905451926,
      "language_loss": 0.79783499,
      "learning_rate": 3.999135446087263e-06,
      "loss": 0.82103723,
      "num_input_tokens_seen": 13720145,
      "step": 650,
      "time_per_iteration": 2.574939727783203
    },
    {
      "auxiliary_loss_clip": 0.01242721,
      "auxiliary_loss_mlp": 0.01077536,
      "balance_loss_clip": 1.06209707,
      "balance_loss_mlp": 1.04534984,
      "epoch": 0.039140237486848035,
      "flos": 18661411486080.0,
      "grad_norm": 2.334811800093409,
      "language_loss": 0.78698987,
      "learning_rate": 3.9991239579635e-06,
      "loss": 0.81019247,
      "num_input_tokens_seen": 13737500,
      "step": 651,
      "time_per_iteration": 2.5930917263031006
    },
    {
      "auxiliary_loss_clip": 0.0124425,
      "auxiliary_loss_mlp": 0.010838,
      "balance_loss_clip": 1.06317663,
      "balance_loss_mlp": 1.05087411,
      "epoch": 0.03920036073951601,
      "flos": 18661124177280.0,
      "grad_norm": 3.361008988618244,
      "language_loss": 0.87392938,
      "learning_rate": 3.999112394032757e-06,
      "loss": 0.89720988,
      "num_input_tokens_seen": 13754750,
      "step": 652,
      "time_per_iteration": 2.6072869300842285
    },
    {
      "auxiliary_loss_clip": 0.01239638,
      "auxiliary_loss_mlp": 0.01073938,
      "balance_loss_clip": 1.06362963,
      "balance_loss_mlp": 1.0434916,
      "epoch": 0.03926048399218398,
      "flos": 31354468053120.0,
      "grad_norm": 2.6218665998754904,
      "language_loss": 0.79297256,
      "learning_rate": 3.999100754295471e-06,
      "loss": 0.81610829,
      "num_input_tokens_seen": 13771990,
      "step": 653,
      "time_per_iteration": 2.626145362854004
    },
    {
      "auxiliary_loss_clip": 0.01250652,
      "auxiliary_loss_mlp": 0.01075546,
      "balance_loss_clip": 1.06496143,
      "balance_loss_mlp": 1.04374111,
      "epoch": 0.039320607244851945,
      "flos": 29603499770880.0,
      "grad_norm": 2.0720296605490094,
      "language_loss": 0.85909009,
      "learning_rate": 3.999089038752085e-06,
      "loss": 0.88235211,
      "num_input_tokens_seen": 13792750,
      "step": 654,
      "time_per_iteration": 2.6775124073028564
    },
    {
      "auxiliary_loss_clip": 0.01126661,
      "auxiliary_loss_mlp": 0.01016641,
      "balance_loss_clip": 1.03977203,
      "balance_loss_mlp": 1.01001298,
      "epoch": 0.03938073049751992,
      "flos": 66534609951360.0,
      "grad_norm": 0.7366259780501333,
      "language_loss": 0.4997997,
      "learning_rate": 3.999077247403041e-06,
      "loss": 0.52123272,
      "num_input_tokens_seen": 13858570,
      "step": 655,
      "time_per_iteration": 3.3006510734558105
    },
    {
      "auxiliary_loss_clip": 0.01241143,
      "auxiliary_loss_mlp": 0.01076374,
      "balance_loss_clip": 1.0658412,
      "balance_loss_mlp": 1.04680991,
      "epoch": 0.03944085375018788,
      "flos": 23367827836800.0,
      "grad_norm": 4.17474796245144,
      "language_loss": 0.80903178,
      "learning_rate": 3.9990653802487886e-06,
      "loss": 0.83220696,
      "num_input_tokens_seen": 13876335,
      "step": 656,
      "time_per_iteration": 4.228931427001953
    },
    {
      "auxiliary_loss_clip": 0.01251519,
      "auxiliary_loss_mlp": 0.01093573,
      "balance_loss_clip": 1.06740427,
      "balance_loss_mlp": 1.05802524,
      "epoch": 0.039500977002855854,
      "flos": 18548292579840.0,
      "grad_norm": 2.068956760077258,
      "language_loss": 0.76289558,
      "learning_rate": 3.999053437289776e-06,
      "loss": 0.7863465,
      "num_input_tokens_seen": 13892640,
      "step": 657,
      "time_per_iteration": 4.218473434448242
    },
    {
      "auxiliary_loss_clip": 0.0124824,
      "auxiliary_loss_mlp": 0.01076812,
      "balance_loss_clip": 1.06641233,
      "balance_loss_mlp": 1.04522133,
      "epoch": 0.039561100255523826,
      "flos": 25338174433920.0,
      "grad_norm": 2.07475431213476,
      "language_loss": 0.8179062,
      "learning_rate": 3.999041418526457e-06,
      "loss": 0.84115672,
      "num_input_tokens_seen": 13910085,
      "step": 658,
      "time_per_iteration": 2.671675682067871
    },
    {
      "auxiliary_loss_clip": 0.01242678,
      "auxiliary_loss_mlp": 0.01077963,
      "balance_loss_clip": 1.06347871,
      "balance_loss_mlp": 1.0454669,
      "epoch": 0.03962122350819179,
      "flos": 18219889509120.0,
      "grad_norm": 2.2444983110753625,
      "language_loss": 0.90790772,
      "learning_rate": 3.999029323959287e-06,
      "loss": 0.93111408,
      "num_input_tokens_seen": 13928800,
      "step": 659,
      "time_per_iteration": 4.2601988315582275
    },
    {
      "auxiliary_loss_clip": 0.01247633,
      "auxiliary_loss_mlp": 0.01073069,
      "balance_loss_clip": 1.06654835,
      "balance_loss_mlp": 1.04215825,
      "epoch": 0.03968134676085976,
      "flos": 20522230536960.0,
      "grad_norm": 2.2083626038373656,
      "language_loss": 0.79760063,
      "learning_rate": 3.999017153588724e-06,
      "loss": 0.82080764,
      "num_input_tokens_seen": 13948325,
      "step": 660,
      "time_per_iteration": 2.62716007232666
    },
    {
      "auxiliary_loss_clip": 0.01246027,
      "auxiliary_loss_mlp": 0.01077579,
      "balance_loss_clip": 1.0675652,
      "balance_loss_mlp": 1.0456785,
      "epoch": 0.03974147001352773,
      "flos": 22422587483520.0,
      "grad_norm": 1.6747851381362888,
      "language_loss": 0.81757367,
      "learning_rate": 3.999004907415231e-06,
      "loss": 0.8408097,
      "num_input_tokens_seen": 13969090,
      "step": 661,
      "time_per_iteration": 2.645423412322998
    },
    {
      "auxiliary_loss_clip": 0.01119895,
      "auxiliary_loss_mlp": 0.01007167,
      "balance_loss_clip": 1.03320217,
      "balance_loss_mlp": 1.00077713,
      "epoch": 0.0398015932661957,
      "flos": 71128769322240.0,
      "grad_norm": 0.9117564509831767,
      "language_loss": 0.69349593,
      "learning_rate": 3.998992585439272e-06,
      "loss": 0.71476656,
      "num_input_tokens_seen": 14037555,
      "step": 662,
      "time_per_iteration": 3.3032331466674805
    },
    {
      "auxiliary_loss_clip": 0.01249217,
      "auxiliary_loss_mlp": 0.01074722,
      "balance_loss_clip": 1.06995225,
      "balance_loss_mlp": 1.04322648,
      "epoch": 0.03986171651886367,
      "flos": 16800951571200.0,
      "grad_norm": 2.160679749799672,
      "language_loss": 0.82765651,
      "learning_rate": 3.998980187661314e-06,
      "loss": 0.85089582,
      "num_input_tokens_seen": 14055765,
      "step": 663,
      "time_per_iteration": 2.6217782497406006
    },
    {
      "auxiliary_loss_clip": 0.01252759,
      "auxiliary_loss_mlp": 0.01063705,
      "balance_loss_clip": 1.06966817,
      "balance_loss_mlp": 1.03254378,
      "epoch": 0.03992183977153164,
      "flos": 24535068197760.0,
      "grad_norm": 2.19374813563436,
      "language_loss": 0.87302262,
      "learning_rate": 3.998967714081826e-06,
      "loss": 0.89618725,
      "num_input_tokens_seen": 14074195,
      "step": 664,
      "time_per_iteration": 2.6729183197021484
    },
    {
      "auxiliary_loss_clip": 0.01241647,
      "auxiliary_loss_mlp": 0.0106515,
      "balance_loss_clip": 1.06656313,
      "balance_loss_mlp": 1.03346384,
      "epoch": 0.03998196302419961,
      "flos": 15595897167360.0,
      "grad_norm": 2.036983550581997,
      "language_loss": 0.84821391,
      "learning_rate": 3.998955164701281e-06,
      "loss": 0.87128186,
      "num_input_tokens_seen": 14090215,
      "step": 665,
      "time_per_iteration": 2.593832015991211
    },
    {
      "auxiliary_loss_clip": 0.012521,
      "auxiliary_loss_mlp": 0.01085682,
      "balance_loss_clip": 1.06867695,
      "balance_loss_mlp": 1.05223155,
      "epoch": 0.04004208627686758,
      "flos": 25305065072640.0,
      "grad_norm": 2.172699570421913,
      "language_loss": 0.81745672,
      "learning_rate": 3.998942539520158e-06,
      "loss": 0.8408345,
      "num_input_tokens_seen": 14112150,
      "step": 666,
      "time_per_iteration": 2.6743290424346924
    },
    {
      "auxiliary_loss_clip": 0.01241565,
      "auxiliary_loss_mlp": 0.01073617,
      "balance_loss_clip": 1.06443083,
      "balance_loss_mlp": 1.04007161,
      "epoch": 0.04010220952953555,
      "flos": 23475847011840.0,
      "grad_norm": 2.1003520396389828,
      "language_loss": 0.87117827,
      "learning_rate": 3.998929838538932e-06,
      "loss": 0.89433014,
      "num_input_tokens_seen": 14131475,
      "step": 667,
      "time_per_iteration": 2.6147067546844482
    },
    {
      "auxiliary_loss_clip": 0.0124275,
      "auxiliary_loss_mlp": 0.01071583,
      "balance_loss_clip": 1.07009172,
      "balance_loss_mlp": 1.04161382,
      "epoch": 0.04016233278220352,
      "flos": 18617025254400.0,
      "grad_norm": 2.331266403294307,
      "language_loss": 0.80641299,
      "learning_rate": 3.998917061758087e-06,
      "loss": 0.82955635,
      "num_input_tokens_seen": 14146165,
      "step": 668,
      "time_per_iteration": 2.6015820503234863
    },
    {
      "auxiliary_loss_clip": 0.01115034,
      "auxiliary_loss_mlp": 0.01008949,
      "balance_loss_clip": 1.02975297,
      "balance_loss_mlp": 1.00317907,
      "epoch": 0.040222456034871484,
      "flos": 70906194696960.0,
      "grad_norm": 0.7870483750596657,
      "language_loss": 0.60066259,
      "learning_rate": 3.998904209178107e-06,
      "loss": 0.62190247,
      "num_input_tokens_seen": 14215005,
      "step": 669,
      "time_per_iteration": 3.2993202209472656
    },
    {
      "auxiliary_loss_clip": 0.01242272,
      "auxiliary_loss_mlp": 0.01071485,
      "balance_loss_clip": 1.06408751,
      "balance_loss_mlp": 1.04120564,
      "epoch": 0.040282579287539456,
      "flos": 23764712186880.0,
      "grad_norm": 1.7022357666604506,
      "language_loss": 0.86290276,
      "learning_rate": 3.9988912807994785e-06,
      "loss": 0.88604033,
      "num_input_tokens_seen": 14235510,
      "step": 670,
      "time_per_iteration": 2.700657844543457
    },
    {
      "auxiliary_loss_clip": 0.01242087,
      "auxiliary_loss_mlp": 0.01080448,
      "balance_loss_clip": 1.06647801,
      "balance_loss_mlp": 1.05014467,
      "epoch": 0.04034270254020743,
      "flos": 18478518410880.0,
      "grad_norm": 1.8224152334464152,
      "language_loss": 0.75569212,
      "learning_rate": 3.998878276622692e-06,
      "loss": 0.77891749,
      "num_input_tokens_seen": 14254565,
      "step": 671,
      "time_per_iteration": 2.6698572635650635
    },
    {
      "auxiliary_loss_clip": 0.01248936,
      "auxiliary_loss_mlp": 0.01076667,
      "balance_loss_clip": 1.06943047,
      "balance_loss_mlp": 1.04605412,
      "epoch": 0.040402825792875394,
      "flos": 17201858244480.0,
      "grad_norm": 1.9730812981627939,
      "language_loss": 0.92416775,
      "learning_rate": 3.998865196648242e-06,
      "loss": 0.94742376,
      "num_input_tokens_seen": 14271885,
      "step": 672,
      "time_per_iteration": 2.567563533782959
    },
    {
      "auxiliary_loss_clip": 0.01245231,
      "auxiliary_loss_mlp": 0.010776,
      "balance_loss_clip": 1.0677104,
      "balance_loss_mlp": 1.04422188,
      "epoch": 0.040462949045543366,
      "flos": 19172168928000.0,
      "grad_norm": 1.800141829654062,
      "language_loss": 0.90174723,
      "learning_rate": 3.998852040876622e-06,
      "loss": 0.92497551,
      "num_input_tokens_seen": 14289670,
      "step": 673,
      "time_per_iteration": 2.547154426574707
    },
    {
      "auxiliary_loss_clip": 0.01239752,
      "auxiliary_loss_mlp": 0.01084248,
      "balance_loss_clip": 1.06466973,
      "balance_loss_mlp": 1.05184698,
      "epoch": 0.04052307229821133,
      "flos": 24019821555840.0,
      "grad_norm": 2.3989934860433486,
      "language_loss": 0.75016737,
      "learning_rate": 3.998838809308334e-06,
      "loss": 0.7734074,
      "num_input_tokens_seen": 14309285,
      "step": 674,
      "time_per_iteration": 2.681896924972534
    },
    {
      "auxiliary_loss_clip": 0.01249861,
      "auxiliary_loss_mlp": 0.01064308,
      "balance_loss_clip": 1.06744063,
      "balance_loss_mlp": 1.03334963,
      "epoch": 0.0405831955508793,
      "flos": 16436601964800.0,
      "grad_norm": 2.55613513039197,
      "language_loss": 0.78289407,
      "learning_rate": 3.9988255019438766e-06,
      "loss": 0.80603576,
      "num_input_tokens_seen": 14328300,
      "step": 675,
      "time_per_iteration": 2.6965043544769287
    },
    {
      "auxiliary_loss_clip": 0.01241749,
      "auxiliary_loss_mlp": 0.01079652,
      "balance_loss_clip": 1.06532836,
      "balance_loss_mlp": 1.04648817,
      "epoch": 0.040643318803547275,
      "flos": 24279922915200.0,
      "grad_norm": 2.047384767684118,
      "language_loss": 0.76844448,
      "learning_rate": 3.998812118783757e-06,
      "loss": 0.79165846,
      "num_input_tokens_seen": 14346395,
      "step": 676,
      "time_per_iteration": 2.6216623783111572
    },
    {
      "auxiliary_loss_clip": 0.01248147,
      "auxiliary_loss_mlp": 0.01079294,
      "balance_loss_clip": 1.06811619,
      "balance_loss_mlp": 1.04813254,
      "epoch": 0.04070344205621524,
      "flos": 17712076982400.0,
      "grad_norm": 2.318905665785744,
      "language_loss": 0.85139382,
      "learning_rate": 3.9987986598284804e-06,
      "loss": 0.8746683,
      "num_input_tokens_seen": 14364605,
      "step": 677,
      "time_per_iteration": 2.5663015842437744
    },
    {
      "auxiliary_loss_clip": 0.01240385,
      "auxiliary_loss_mlp": 0.01070741,
      "balance_loss_clip": 1.06558609,
      "balance_loss_mlp": 1.03901923,
      "epoch": 0.04076356530888321,
      "flos": 26177658168960.0,
      "grad_norm": 2.5041724349122645,
      "language_loss": 0.76572061,
      "learning_rate": 3.998785125078559e-06,
      "loss": 0.78883183,
      "num_input_tokens_seen": 14385265,
      "step": 678,
      "time_per_iteration": 2.624689817428589
    },
    {
      "auxiliary_loss_clip": 0.01240972,
      "auxiliary_loss_mlp": 0.01072606,
      "balance_loss_clip": 1.06374967,
      "balance_loss_mlp": 1.04242194,
      "epoch": 0.04082368856155118,
      "flos": 35773455772800.0,
      "grad_norm": 1.7096242150987748,
      "language_loss": 0.82139099,
      "learning_rate": 3.998771514534505e-06,
      "loss": 0.84452677,
      "num_input_tokens_seen": 14406090,
      "step": 679,
      "time_per_iteration": 2.7073023319244385
    },
    {
      "auxiliary_loss_clip": 0.01248879,
      "auxiliary_loss_mlp": 0.01064116,
      "balance_loss_clip": 1.07185793,
      "balance_loss_mlp": 1.0340035,
      "epoch": 0.04088381181421915,
      "flos": 28146640049280.0,
      "grad_norm": 1.963288262989073,
      "language_loss": 0.76260424,
      "learning_rate": 3.998757828196835e-06,
      "loss": 0.78573418,
      "num_input_tokens_seen": 14425130,
      "step": 680,
      "time_per_iteration": 2.6767218112945557
    },
    {
      "auxiliary_loss_clip": 0.01244441,
      "auxiliary_loss_mlp": 0.01071738,
      "balance_loss_clip": 1.06458521,
      "balance_loss_mlp": 1.03864551,
      "epoch": 0.04094393506688712,
      "flos": 27597673514880.0,
      "grad_norm": 1.713943858995997,
      "language_loss": 0.83089912,
      "learning_rate": 3.9987440660660685e-06,
      "loss": 0.85406095,
      "num_input_tokens_seen": 14447355,
      "step": 681,
      "time_per_iteration": 2.6386382579803467
    },
    {
      "auxiliary_loss_clip": 0.01244279,
      "auxiliary_loss_mlp": 0.01073303,
      "balance_loss_clip": 1.06438065,
      "balance_loss_mlp": 1.04127121,
      "epoch": 0.04100405831955509,
      "flos": 23112036109440.0,
      "grad_norm": 1.706698119772261,
      "language_loss": 0.71538687,
      "learning_rate": 3.998730228142726e-06,
      "loss": 0.7385627,
      "num_input_tokens_seen": 14466790,
      "step": 682,
      "time_per_iteration": 2.618792772293091
    },
    {
      "auxiliary_loss_clip": 0.01243156,
      "auxiliary_loss_mlp": 0.01078429,
      "balance_loss_clip": 1.06440282,
      "balance_loss_mlp": 1.04781592,
      "epoch": 0.04106418157222306,
      "flos": 20156731695360.0,
      "grad_norm": 1.6947476714586034,
      "language_loss": 0.72599399,
      "learning_rate": 3.998716314427333e-06,
      "loss": 0.74920982,
      "num_input_tokens_seen": 14485195,
      "step": 683,
      "time_per_iteration": 2.676133394241333
    },
    {
      "auxiliary_loss_clip": 0.01241071,
      "auxiliary_loss_mlp": 0.01079531,
      "balance_loss_clip": 1.07077932,
      "balance_loss_mlp": 1.04851258,
      "epoch": 0.041124304824891024,
      "flos": 17420697855360.0,
      "grad_norm": 2.098652785935233,
      "language_loss": 0.81419414,
      "learning_rate": 3.998702324920417e-06,
      "loss": 0.8374002,
      "num_input_tokens_seen": 14503370,
      "step": 684,
      "time_per_iteration": 2.6538476943969727
    },
    {
      "auxiliary_loss_clip": 0.01242791,
      "auxiliary_loss_mlp": 0.0107365,
      "balance_loss_clip": 1.06783867,
      "balance_loss_mlp": 1.04139185,
      "epoch": 0.041184428077558996,
      "flos": 25780163287680.0,
      "grad_norm": 1.5053911947555274,
      "language_loss": 0.90680599,
      "learning_rate": 3.9986882596225085e-06,
      "loss": 0.92997038,
      "num_input_tokens_seen": 14526415,
      "step": 685,
      "time_per_iteration": 2.6541450023651123
    },
    {
      "auxiliary_loss_clip": 0.01244219,
      "auxiliary_loss_mlp": 0.01072481,
      "balance_loss_clip": 1.06659365,
      "balance_loss_mlp": 1.04093838,
      "epoch": 0.04124455133022697,
      "flos": 22964766347520.0,
      "grad_norm": 2.2251875217653185,
      "language_loss": 0.87851977,
      "learning_rate": 3.998674118534141e-06,
      "loss": 0.90168673,
      "num_input_tokens_seen": 14546595,
      "step": 686,
      "time_per_iteration": 2.7298531532287598
    },
    {
      "auxiliary_loss_clip": 0.01247476,
      "auxiliary_loss_mlp": 0.01073385,
      "balance_loss_clip": 1.06586432,
      "balance_loss_mlp": 1.04224789,
      "epoch": 0.04130467458289493,
      "flos": 21289067015040.0,
      "grad_norm": 1.8582614005091855,
      "language_loss": 0.7152915,
      "learning_rate": 3.998659901655851e-06,
      "loss": 0.73850012,
      "num_input_tokens_seen": 14566590,
      "step": 687,
      "time_per_iteration": 2.6284232139587402
    },
    {
      "auxiliary_loss_clip": 0.01243582,
      "auxiliary_loss_mlp": 0.01076448,
      "balance_loss_clip": 1.06979251,
      "balance_loss_mlp": 1.04756403,
      "epoch": 0.041364797835562905,
      "flos": 19974233669760.0,
      "grad_norm": 2.596672934278983,
      "language_loss": 0.86028284,
      "learning_rate": 3.998645608988177e-06,
      "loss": 0.88348317,
      "num_input_tokens_seen": 14585965,
      "step": 688,
      "time_per_iteration": 2.522634506225586
    },
    {
      "auxiliary_loss_clip": 0.01241593,
      "auxiliary_loss_mlp": 0.01079647,
      "balance_loss_clip": 1.06802177,
      "balance_loss_mlp": 1.04908216,
      "epoch": 0.04142492108823087,
      "flos": 21906227520000.0,
      "grad_norm": 2.852238187591699,
      "language_loss": 0.83393514,
      "learning_rate": 3.998631240531661e-06,
      "loss": 0.85714757,
      "num_input_tokens_seen": 14606015,
      "step": 689,
      "time_per_iteration": 2.6140944957733154
    },
    {
      "auxiliary_loss_clip": 0.01238254,
      "auxiliary_loss_mlp": 0.01085009,
      "balance_loss_clip": 1.06293654,
      "balance_loss_mlp": 1.05463421,
      "epoch": 0.04148504434089884,
      "flos": 27639617621760.0,
      "grad_norm": 2.870474577544969,
      "language_loss": 0.68398476,
      "learning_rate": 3.998616796286848e-06,
      "loss": 0.70721734,
      "num_input_tokens_seen": 14629955,
      "step": 690,
      "time_per_iteration": 2.658987522125244
    },
    {
      "auxiliary_loss_clip": 0.01235903,
      "auxiliary_loss_mlp": 0.01075275,
      "balance_loss_clip": 1.0625304,
      "balance_loss_mlp": 1.04565191,
      "epoch": 0.041545167593566815,
      "flos": 20518387781760.0,
      "grad_norm": 1.634289561889102,
      "language_loss": 0.74927461,
      "learning_rate": 3.998602276254286e-06,
      "loss": 0.77238643,
      "num_input_tokens_seen": 14648000,
      "step": 691,
      "time_per_iteration": 2.599957227706909
    },
    {
      "auxiliary_loss_clip": 0.01239089,
      "auxiliary_loss_mlp": 0.01081705,
      "balance_loss_clip": 1.06458938,
      "balance_loss_mlp": 1.04978108,
      "epoch": 0.04160529084623478,
      "flos": 11868907939200.0,
      "grad_norm": 2.123432521314224,
      "language_loss": 0.84469771,
      "learning_rate": 3.998587680434526e-06,
      "loss": 0.86790562,
      "num_input_tokens_seen": 14662235,
      "step": 692,
      "time_per_iteration": 2.5748491287231445
    },
    {
      "auxiliary_loss_clip": 0.01242126,
      "auxiliary_loss_mlp": 0.01076613,
      "balance_loss_clip": 1.06274796,
      "balance_loss_mlp": 1.04409313,
      "epoch": 0.04166541409890275,
      "flos": 14828306503680.0,
      "grad_norm": 2.3463094595874665,
      "language_loss": 0.88948715,
      "learning_rate": 3.99857300882812e-06,
      "loss": 0.91267455,
      "num_input_tokens_seen": 14676065,
      "step": 693,
      "time_per_iteration": 2.569277286529541
    },
    {
      "auxiliary_loss_clip": 0.01245438,
      "auxiliary_loss_mlp": 0.01071471,
      "balance_loss_clip": 1.06845784,
      "balance_loss_mlp": 1.04123962,
      "epoch": 0.04172553735157072,
      "flos": 25808137004160.0,
      "grad_norm": 5.499777597079252,
      "language_loss": 0.81987685,
      "learning_rate": 3.998558261435626e-06,
      "loss": 0.84304595,
      "num_input_tokens_seen": 14694955,
      "step": 694,
      "time_per_iteration": 2.6798722743988037
    },
    {
      "auxiliary_loss_clip": 0.01242101,
      "auxiliary_loss_mlp": 0.01073692,
      "balance_loss_clip": 1.06179321,
      "balance_loss_mlp": 1.04303181,
      "epoch": 0.04178566060423869,
      "flos": 24279815174400.0,
      "grad_norm": 2.051302362473346,
      "language_loss": 0.83672506,
      "learning_rate": 3.9985434382576015e-06,
      "loss": 0.85988301,
      "num_input_tokens_seen": 14715510,
      "step": 695,
      "time_per_iteration": 2.684537649154663
    },
    {
      "auxiliary_loss_clip": 0.01242205,
      "auxiliary_loss_mlp": 0.01080004,
      "balance_loss_clip": 1.06535804,
      "balance_loss_mlp": 1.04822254,
      "epoch": 0.04184578385690666,
      "flos": 18222008411520.0,
      "grad_norm": 2.113561459264794,
      "language_loss": 0.84351176,
      "learning_rate": 3.99852853929461e-06,
      "loss": 0.86673379,
      "num_input_tokens_seen": 14731755,
      "step": 696,
      "time_per_iteration": 4.1141321659088135
    },
    {
      "auxiliary_loss_clip": 0.01238462,
      "auxiliary_loss_mlp": 0.01083207,
      "balance_loss_clip": 1.06265593,
      "balance_loss_mlp": 1.05099702,
      "epoch": 0.041905907109574626,
      "flos": 22776342577920.0,
      "grad_norm": 6.921460264787684,
      "language_loss": 0.93193012,
      "learning_rate": 3.998513564547216e-06,
      "loss": 0.95514685,
      "num_input_tokens_seen": 14750810,
      "step": 697,
      "time_per_iteration": 5.71666693687439
    },
    {
      "auxiliary_loss_clip": 0.01235964,
      "auxiliary_loss_mlp": 0.01074448,
      "balance_loss_clip": 1.06324339,
      "balance_loss_mlp": 1.04495573,
      "epoch": 0.0419660303622426,
      "flos": 20156947176960.0,
      "grad_norm": 2.1002029886241904,
      "language_loss": 0.83775562,
      "learning_rate": 3.998498514015987e-06,
      "loss": 0.86085975,
      "num_input_tokens_seen": 14768435,
      "step": 698,
      "time_per_iteration": 4.194530010223389
    },
    {
      "auxiliary_loss_clip": 0.01239177,
      "auxiliary_loss_mlp": 0.01093516,
      "balance_loss_clip": 1.06274605,
      "balance_loss_mlp": 1.06175828,
      "epoch": 0.042026153614910564,
      "flos": 23076376882560.0,
      "grad_norm": 2.1234669437327955,
      "language_loss": 0.91715962,
      "learning_rate": 3.998483387701495e-06,
      "loss": 0.94048655,
      "num_input_tokens_seen": 14786690,
      "step": 699,
      "time_per_iteration": 2.6399078369140625
    },
    {
      "auxiliary_loss_clip": 0.01113327,
      "auxiliary_loss_mlp": 0.0102038,
      "balance_loss_clip": 1.03020263,
      "balance_loss_mlp": 1.01403797,
      "epoch": 0.042086276867578536,
      "flos": 64495243370880.0,
      "grad_norm": 0.9035134571641164,
      "language_loss": 0.67873394,
      "learning_rate": 3.998468185604312e-06,
      "loss": 0.70007098,
      "num_input_tokens_seen": 14853840,
      "step": 700,
      "time_per_iteration": 3.192026376724243
    },
    {
      "auxiliary_loss_clip": 0.01246765,
      "auxiliary_loss_mlp": 0.01082955,
      "balance_loss_clip": 1.06717515,
      "balance_loss_mlp": 1.05017269,
      "epoch": 0.04214640012024651,
      "flos": 15487016065920.0,
      "grad_norm": 2.2754848646841888,
      "language_loss": 0.884673,
      "learning_rate": 3.998452907725016e-06,
      "loss": 0.90797025,
      "num_input_tokens_seen": 14869580,
      "step": 701,
      "time_per_iteration": 2.5790441036224365
    },
    {
      "auxiliary_loss_clip": 0.01242428,
      "auxiliary_loss_mlp": 0.01080259,
      "balance_loss_clip": 1.06793952,
      "balance_loss_mlp": 1.04833448,
      "epoch": 0.04220652337291447,
      "flos": 23877040993920.0,
      "grad_norm": 2.000128536077818,
      "language_loss": 0.67100394,
      "learning_rate": 3.998437554064184e-06,
      "loss": 0.69423079,
      "num_input_tokens_seen": 14891065,
      "step": 702,
      "time_per_iteration": 2.6247870922088623
    },
    {
      "auxiliary_loss_clip": 0.01107168,
      "auxiliary_loss_mlp": 0.01005563,
      "balance_loss_clip": 1.02512407,
      "balance_loss_mlp": 0.99922067,
      "epoch": 0.042266646625582445,
      "flos": 63795451628160.0,
      "grad_norm": 0.8439205282656718,
      "language_loss": 0.60756463,
      "learning_rate": 3.9984221246224006e-06,
      "loss": 0.62869191,
      "num_input_tokens_seen": 14954815,
      "step": 703,
      "time_per_iteration": 3.1991655826568604
    },
    {
      "auxiliary_loss_clip": 0.01107933,
      "auxiliary_loss_mlp": 0.01006502,
      "balance_loss_clip": 1.02562141,
      "balance_loss_mlp": 0.99973089,
      "epoch": 0.04232676987825041,
      "flos": 50018863345920.0,
      "grad_norm": 1.0471369072250156,
      "language_loss": 0.57677412,
      "learning_rate": 3.9984066194002494e-06,
      "loss": 0.59791845,
      "num_input_tokens_seen": 15003050,
      "step": 704,
      "time_per_iteration": 3.037705659866333
    },
    {
      "auxiliary_loss_clip": 0.01241513,
      "auxiliary_loss_mlp": 0.01072126,
      "balance_loss_clip": 1.06549489,
      "balance_loss_mlp": 1.0406549,
      "epoch": 0.04238689313091838,
      "flos": 21616105368960.0,
      "grad_norm": 2.9488804643242488,
      "language_loss": 0.87553984,
      "learning_rate": 3.998391038398319e-06,
      "loss": 0.89867628,
      "num_input_tokens_seen": 15021990,
      "step": 705,
      "time_per_iteration": 2.6233222484588623
    },
    {
      "auxiliary_loss_clip": 0.01230342,
      "auxiliary_loss_mlp": 0.0107194,
      "balance_loss_clip": 1.0605582,
      "balance_loss_mlp": 1.04204249,
      "epoch": 0.042447016383586354,
      "flos": 19135109070720.0,
      "grad_norm": 2.556815837902013,
      "language_loss": 0.71071029,
      "learning_rate": 3.998375381617201e-06,
      "loss": 0.73373306,
      "num_input_tokens_seen": 15040700,
      "step": 706,
      "time_per_iteration": 2.560434579849243
    },
    {
      "auxiliary_loss_clip": 0.0123412,
      "auxiliary_loss_mlp": 0.01070349,
      "balance_loss_clip": 1.06249404,
      "balance_loss_mlp": 1.03799582,
      "epoch": 0.04250713963625432,
      "flos": 24426007528320.0,
      "grad_norm": 2.0814078632624167,
      "language_loss": 0.93418455,
      "learning_rate": 3.9983596490574875e-06,
      "loss": 0.95722926,
      "num_input_tokens_seen": 15056725,
      "step": 707,
      "time_per_iteration": 2.6130473613739014
    },
    {
      "auxiliary_loss_clip": 0.01237541,
      "auxiliary_loss_mlp": 0.01067908,
      "balance_loss_clip": 1.05994225,
      "balance_loss_mlp": 1.03617477,
      "epoch": 0.04256726288892229,
      "flos": 30367391333760.0,
      "grad_norm": 2.424205580643553,
      "language_loss": 0.81514043,
      "learning_rate": 3.998343840719776e-06,
      "loss": 0.83819497,
      "num_input_tokens_seen": 15077550,
      "step": 708,
      "time_per_iteration": 2.656277894973755
    },
    {
      "auxiliary_loss_clip": 0.01243932,
      "auxiliary_loss_mlp": 0.0108167,
      "balance_loss_clip": 1.06461239,
      "balance_loss_mlp": 1.04934049,
      "epoch": 0.04262738614159026,
      "flos": 16362661818240.0,
      "grad_norm": 2.0883592727868145,
      "language_loss": 0.82027614,
      "learning_rate": 3.998327956604666e-06,
      "loss": 0.8435322,
      "num_input_tokens_seen": 15094955,
      "step": 709,
      "time_per_iteration": 2.5758891105651855
    },
    {
      "auxiliary_loss_clip": 0.01243538,
      "auxiliary_loss_mlp": 0.01071217,
      "balance_loss_clip": 1.06374872,
      "balance_loss_mlp": 1.03960264,
      "epoch": 0.04268750939425823,
      "flos": 20412379768320.0,
      "grad_norm": 2.7686525844665133,
      "language_loss": 0.8502059,
      "learning_rate": 3.99831199671276e-06,
      "loss": 0.87335348,
      "num_input_tokens_seen": 15113395,
      "step": 710,
      "time_per_iteration": 2.571559429168701
    },
    {
      "auxiliary_loss_clip": 0.0124498,
      "auxiliary_loss_mlp": 0.01072229,
      "balance_loss_clip": 1.06788397,
      "balance_loss_mlp": 1.04166365,
      "epoch": 0.0427476326469262,
      "flos": 20302959962880.0,
      "grad_norm": 7.911177124524585,
      "language_loss": 0.84914303,
      "learning_rate": 3.998295961044662e-06,
      "loss": 0.87231517,
      "num_input_tokens_seen": 15132920,
      "step": 711,
      "time_per_iteration": 2.569959878921509
    },
    {
      "auxiliary_loss_clip": 0.01237769,
      "auxiliary_loss_mlp": 0.01074338,
      "balance_loss_clip": 1.06188083,
      "balance_loss_mlp": 1.04229426,
      "epoch": 0.042807755899594166,
      "flos": 21650794928640.0,
      "grad_norm": 1.7189790042473796,
      "language_loss": 0.85439789,
      "learning_rate": 3.9982798496009804e-06,
      "loss": 0.87751901,
      "num_input_tokens_seen": 15153115,
      "step": 712,
      "time_per_iteration": 2.6200509071350098
    },
    {
      "auxiliary_loss_clip": 0.01242397,
      "auxiliary_loss_mlp": 0.01069523,
      "balance_loss_clip": 1.06085837,
      "balance_loss_mlp": 1.03989983,
      "epoch": 0.04286787915226214,
      "flos": 21435007973760.0,
      "grad_norm": 5.490507523621204,
      "language_loss": 0.91178697,
      "learning_rate": 3.998263662382328e-06,
      "loss": 0.93490618,
      "num_input_tokens_seen": 15172770,
      "step": 713,
      "time_per_iteration": 2.6353416442871094
    },
    {
      "auxiliary_loss_clip": 0.01104693,
      "auxiliary_loss_mlp": 0.01006514,
      "balance_loss_clip": 1.02325606,
      "balance_loss_mlp": 0.99955195,
      "epoch": 0.04292800240493011,
      "flos": 66397970615040.0,
      "grad_norm": 0.9310328114407391,
      "language_loss": 0.63725489,
      "learning_rate": 3.9982473993893165e-06,
      "loss": 0.65836698,
      "num_input_tokens_seen": 15240055,
      "step": 714,
      "time_per_iteration": 3.2544445991516113
    },
    {
      "auxiliary_loss_clip": 0.01239175,
      "auxiliary_loss_mlp": 0.01085992,
      "balance_loss_clip": 1.06602359,
      "balance_loss_mlp": 1.05552244,
      "epoch": 0.042988125657598075,
      "flos": 31650264552960.0,
      "grad_norm": 1.8449858143817996,
      "language_loss": 0.75010103,
      "learning_rate": 3.998231060622563e-06,
      "loss": 0.77335274,
      "num_input_tokens_seen": 15261585,
      "step": 715,
      "time_per_iteration": 2.7048466205596924
    },
    {
      "auxiliary_loss_clip": 0.01242734,
      "auxiliary_loss_mlp": 0.01074126,
      "balance_loss_clip": 1.0666225,
      "balance_loss_mlp": 1.04227352,
      "epoch": 0.04304824891026605,
      "flos": 33248468292480.0,
      "grad_norm": 1.9505519101092619,
      "language_loss": 0.72289199,
      "learning_rate": 3.998214646082688e-06,
      "loss": 0.74606061,
      "num_input_tokens_seen": 15281160,
      "step": 716,
      "time_per_iteration": 2.7807397842407227
    },
    {
      "auxiliary_loss_clip": 0.01104303,
      "auxiliary_loss_mlp": 0.01006894,
      "balance_loss_clip": 1.02277207,
      "balance_loss_mlp": 0.99997944,
      "epoch": 0.04310837216293401,
      "flos": 64064782782720.0,
      "grad_norm": 0.9245106661639481,
      "language_loss": 0.65587437,
      "learning_rate": 3.998198155770314e-06,
      "loss": 0.67698634,
      "num_input_tokens_seen": 15344505,
      "step": 717,
      "time_per_iteration": 3.250870943069458
    },
    {
      "auxiliary_loss_clip": 0.01103971,
      "auxiliary_loss_mlp": 0.01009587,
      "balance_loss_clip": 1.02238059,
      "balance_loss_mlp": 1.00267255,
      "epoch": 0.043168495415601985,
      "flos": 61343757849600.0,
      "grad_norm": 0.9849394627593366,
      "language_loss": 0.58785796,
      "learning_rate": 3.998181589686065e-06,
      "loss": 0.60899353,
      "num_input_tokens_seen": 15404050,
      "step": 718,
      "time_per_iteration": 3.0402464866638184
    },
    {
      "auxiliary_loss_clip": 0.0124025,
      "auxiliary_loss_mlp": 0.0107507,
      "balance_loss_clip": 1.06784248,
      "balance_loss_mlp": 1.0424546,
      "epoch": 0.04322861866826996,
      "flos": 20704261685760.0,
      "grad_norm": 1.9557310597444375,
      "language_loss": 0.91440111,
      "learning_rate": 3.99816494783057e-06,
      "loss": 0.9375543,
      "num_input_tokens_seen": 15424190,
      "step": 719,
      "time_per_iteration": 2.6500089168548584
    },
    {
      "auxiliary_loss_clip": 0.01235843,
      "auxiliary_loss_mlp": 0.01072906,
      "balance_loss_clip": 1.06020999,
      "balance_loss_mlp": 1.04296041,
      "epoch": 0.04328874192093792,
      "flos": 30373352991360.0,
      "grad_norm": 1.7057721639328365,
      "language_loss": 0.66461253,
      "learning_rate": 3.99814823020446e-06,
      "loss": 0.68770003,
      "num_input_tokens_seen": 15446500,
      "step": 720,
      "time_per_iteration": 2.673184871673584
    },
    {
      "auxiliary_loss_clip": 0.01234245,
      "auxiliary_loss_mlp": 0.01072069,
      "balance_loss_clip": 1.06111717,
      "balance_loss_mlp": 1.04131258,
      "epoch": 0.043348865173605894,
      "flos": 21944795748480.0,
      "grad_norm": 1.9491363249287763,
      "language_loss": 0.77460182,
      "learning_rate": 3.9981314368083684e-06,
      "loss": 0.79766488,
      "num_input_tokens_seen": 15465830,
      "step": 721,
      "time_per_iteration": 2.6695611476898193
    },
    {
      "auxiliary_loss_clip": 0.01241854,
      "auxiliary_loss_mlp": 0.01087169,
      "balance_loss_clip": 1.06622314,
      "balance_loss_mlp": 1.05719972,
      "epoch": 0.04340898842627386,
      "flos": 15264225959040.0,
      "grad_norm": 2.8383174670702718,
      "language_loss": 0.88298881,
      "learning_rate": 3.998114567642933e-06,
      "loss": 0.90627909,
      "num_input_tokens_seen": 15479985,
      "step": 722,
      "time_per_iteration": 2.661313533782959
    },
    {
      "auxiliary_loss_clip": 0.01244836,
      "auxiliary_loss_mlp": 0.01076885,
      "balance_loss_clip": 1.06665182,
      "balance_loss_mlp": 1.0480125,
      "epoch": 0.04346911167894183,
      "flos": 27965434913280.0,
      "grad_norm": 5.515838365549148,
      "language_loss": 0.84387141,
      "learning_rate": 3.998097622708792e-06,
      "loss": 0.86708868,
      "num_input_tokens_seen": 15501545,
      "step": 723,
      "time_per_iteration": 2.6447954177856445
    },
    {
      "auxiliary_loss_clip": 0.01245825,
      "auxiliary_loss_mlp": 0.01081354,
      "balance_loss_clip": 1.06723523,
      "balance_loss_mlp": 1.05019248,
      "epoch": 0.0435292349316098,
      "flos": 29242202820480.0,
      "grad_norm": 1.7852936089408447,
      "language_loss": 0.82789439,
      "learning_rate": 3.99808060200659e-06,
      "loss": 0.85116619,
      "num_input_tokens_seen": 15521725,
      "step": 724,
      "time_per_iteration": 2.676985263824463
    },
    {
      "auxiliary_loss_clip": 0.0124127,
      "auxiliary_loss_mlp": 0.01087491,
      "balance_loss_clip": 1.06535757,
      "balance_loss_mlp": 1.05609179,
      "epoch": 0.04358935818427777,
      "flos": 20558356640640.0,
      "grad_norm": 2.011685360503238,
      "language_loss": 0.79444051,
      "learning_rate": 3.998063505536971e-06,
      "loss": 0.81772816,
      "num_input_tokens_seen": 15540910,
      "step": 725,
      "time_per_iteration": 2.6241447925567627
    },
    {
      "auxiliary_loss_clip": 0.01251777,
      "auxiliary_loss_mlp": 0.01074923,
      "balance_loss_clip": 1.06783843,
      "balance_loss_mlp": 1.04309392,
      "epoch": 0.04364948143694574,
      "flos": 14464926564480.0,
      "grad_norm": 2.2160842755462817,
      "language_loss": 0.87175703,
      "learning_rate": 3.998046333300584e-06,
      "loss": 0.89502406,
      "num_input_tokens_seen": 15558640,
      "step": 726,
      "time_per_iteration": 2.555551052093506
    },
    {
      "auxiliary_loss_clip": 0.01100917,
      "auxiliary_loss_mlp": 0.01015411,
      "balance_loss_clip": 1.02171838,
      "balance_loss_mlp": 1.00947404,
      "epoch": 0.043709604689613706,
      "flos": 50067268922880.0,
      "grad_norm": 0.908981905466007,
      "language_loss": 0.55868411,
      "learning_rate": 3.998029085298079e-06,
      "loss": 0.5798474,
      "num_input_tokens_seen": 15612975,
      "step": 727,
      "time_per_iteration": 3.375901699066162
    },
    {
      "auxiliary_loss_clip": 0.01245647,
      "auxiliary_loss_mlp": 0.0108809,
      "balance_loss_clip": 1.06717396,
      "balance_loss_mlp": 1.05614173,
      "epoch": 0.04376972794228168,
      "flos": 13991588115840.0,
      "grad_norm": 2.282663852625415,
      "language_loss": 0.82326066,
      "learning_rate": 3.998011761530112e-06,
      "loss": 0.84659809,
      "num_input_tokens_seen": 15631070,
      "step": 728,
      "time_per_iteration": 2.605970621109009
    },
    {
      "auxiliary_loss_clip": 0.01237902,
      "auxiliary_loss_mlp": 0.01073495,
      "balance_loss_clip": 1.06600416,
      "balance_loss_mlp": 1.04321551,
      "epoch": 0.04382985119494965,
      "flos": 22009901149440.0,
      "grad_norm": 2.1303486954703152,
      "language_loss": 0.76890069,
      "learning_rate": 3.997994361997338e-06,
      "loss": 0.7920146,
      "num_input_tokens_seen": 15647825,
      "step": 729,
      "time_per_iteration": 2.652466297149658
    },
    {
      "auxiliary_loss_clip": 0.01243746,
      "auxiliary_loss_mlp": 0.01079207,
      "balance_loss_clip": 1.06438255,
      "balance_loss_mlp": 1.04859376,
      "epoch": 0.043889974447617615,
      "flos": 24206521472640.0,
      "grad_norm": 2.1385115795714107,
      "language_loss": 0.95153189,
      "learning_rate": 3.997976886700417e-06,
      "loss": 0.97476137,
      "num_input_tokens_seen": 15668260,
      "step": 730,
      "time_per_iteration": 2.734614133834839
    },
    {
      "auxiliary_loss_clip": 0.01238581,
      "auxiliary_loss_mlp": 0.01074727,
      "balance_loss_clip": 1.06093788,
      "balance_loss_mlp": 1.04315984,
      "epoch": 0.04395009770028559,
      "flos": 17274541415040.0,
      "grad_norm": 2.333073864238008,
      "language_loss": 0.88456279,
      "learning_rate": 3.997959335640013e-06,
      "loss": 0.90769589,
      "num_input_tokens_seen": 15685630,
      "step": 731,
      "time_per_iteration": 2.5912294387817383
    },
    {
      "auxiliary_loss_clip": 0.01242247,
      "auxiliary_loss_mlp": 0.01076563,
      "balance_loss_clip": 1.06636512,
      "balance_loss_mlp": 1.04757094,
      "epoch": 0.04401022095295355,
      "flos": 12310286261760.0,
      "grad_norm": 3.0398759554531254,
      "language_loss": 0.88683128,
      "learning_rate": 3.997941708816791e-06,
      "loss": 0.9100194,
      "num_input_tokens_seen": 15698645,
      "step": 732,
      "time_per_iteration": 2.5897367000579834
    },
    {
      "auxiliary_loss_clip": 0.01242736,
      "auxiliary_loss_mlp": 0.01087795,
      "balance_loss_clip": 1.06544232,
      "balance_loss_mlp": 1.05646718,
      "epoch": 0.044070344205621524,
      "flos": 20959658363520.0,
      "grad_norm": 2.304959545118842,
      "language_loss": 0.85829747,
      "learning_rate": 3.997924006231419e-06,
      "loss": 0.88160276,
      "num_input_tokens_seen": 15716775,
      "step": 733,
      "time_per_iteration": 2.650681972503662
    },
    {
      "auxiliary_loss_clip": 0.01246603,
      "auxiliary_loss_mlp": 0.01088724,
      "balance_loss_clip": 1.06722379,
      "balance_loss_mlp": 1.05544066,
      "epoch": 0.044130467458289496,
      "flos": 13845288021120.0,
      "grad_norm": 2.207780377909299,
      "language_loss": 0.91189414,
      "learning_rate": 3.9979062278845685e-06,
      "loss": 0.93524742,
      "num_input_tokens_seen": 15733320,
      "step": 734,
      "time_per_iteration": 2.5956180095672607
    },
    {
      "auxiliary_loss_clip": 0.01238395,
      "auxiliary_loss_mlp": 0.01067579,
      "balance_loss_clip": 1.06596422,
      "balance_loss_mlp": 1.03781235,
      "epoch": 0.04419059071095746,
      "flos": 28655063107200.0,
      "grad_norm": 1.9297536072777384,
      "language_loss": 0.77884138,
      "learning_rate": 3.9978883737769125e-06,
      "loss": 0.8019011,
      "num_input_tokens_seen": 15752705,
      "step": 735,
      "time_per_iteration": 2.603809118270874
    },
    {
      "auxiliary_loss_clip": 0.01234188,
      "auxiliary_loss_mlp": 0.01070499,
      "balance_loss_clip": 1.06063068,
      "balance_loss_mlp": 1.04091144,
      "epoch": 0.04425071396362543,
      "flos": 28183304856960.0,
      "grad_norm": 2.266122200005257,
      "language_loss": 0.8832593,
      "learning_rate": 3.9978704439091305e-06,
      "loss": 0.90630615,
      "num_input_tokens_seen": 15772800,
      "step": 736,
      "time_per_iteration": 5.841086149215698
    },
    {
      "auxiliary_loss_clip": 0.01235947,
      "auxiliary_loss_mlp": 0.01081098,
      "balance_loss_clip": 1.06597185,
      "balance_loss_mlp": 1.05165362,
      "epoch": 0.0443108372162934,
      "flos": 23658452778240.0,
      "grad_norm": 1.8984177574034653,
      "language_loss": 0.84481263,
      "learning_rate": 3.997852438281901e-06,
      "loss": 0.8679831,
      "num_input_tokens_seen": 15793665,
      "step": 737,
      "time_per_iteration": 4.1386003494262695
    },
    {
      "auxiliary_loss_clip": 0.01240863,
      "auxiliary_loss_mlp": 0.01072388,
      "balance_loss_clip": 1.0653491,
      "balance_loss_mlp": 1.03961766,
      "epoch": 0.04437096046896137,
      "flos": 33979861025280.0,
      "grad_norm": 2.2366199062134706,
      "language_loss": 0.84712577,
      "learning_rate": 3.997834356895906e-06,
      "loss": 0.87025833,
      "num_input_tokens_seen": 15813175,
      "step": 738,
      "time_per_iteration": 4.447159290313721
    },
    {
      "auxiliary_loss_clip": 0.01098733,
      "auxiliary_loss_mlp": 0.0102196,
      "balance_loss_clip": 1.02144337,
      "balance_loss_mlp": 1.01685739,
      "epoch": 0.04443108372162934,
      "flos": 67397506375680.0,
      "grad_norm": 0.8779518557387592,
      "language_loss": 0.59179878,
      "learning_rate": 3.9978161997518324e-06,
      "loss": 0.61300576,
      "num_input_tokens_seen": 15872050,
      "step": 739,
      "time_per_iteration": 3.0780396461486816
    },
    {
      "auxiliary_loss_clip": 0.012386,
      "auxiliary_loss_mlp": 0.01067387,
      "balance_loss_clip": 1.06604302,
      "balance_loss_mlp": 1.03717899,
      "epoch": 0.04449120697429731,
      "flos": 29752672953600.0,
      "grad_norm": 2.295102845773205,
      "language_loss": 0.91329807,
      "learning_rate": 3.997797966850369e-06,
      "loss": 0.93635798,
      "num_input_tokens_seen": 15891085,
      "step": 740,
      "time_per_iteration": 2.6687562465667725
    },
    {
      "auxiliary_loss_clip": 0.01243424,
      "auxiliary_loss_mlp": 0.01067832,
      "balance_loss_clip": 1.06807768,
      "balance_loss_mlp": 1.03929377,
      "epoch": 0.04455133022696528,
      "flos": 36502119072000.0,
      "grad_norm": 2.0543845689042484,
      "language_loss": 0.71875739,
      "learning_rate": 3.997779658192205e-06,
      "loss": 0.74186987,
      "num_input_tokens_seen": 15914225,
      "step": 741,
      "time_per_iteration": 2.707231283187866
    },
    {
      "auxiliary_loss_clip": 0.01233192,
      "auxiliary_loss_mlp": 0.01084138,
      "balance_loss_clip": 1.062482,
      "balance_loss_mlp": 1.05476475,
      "epoch": 0.044611453479633245,
      "flos": 28803661672320.0,
      "grad_norm": 1.7086571433899975,
      "language_loss": 0.88933527,
      "learning_rate": 3.997761273778037e-06,
      "loss": 0.91250861,
      "num_input_tokens_seen": 15934540,
      "step": 742,
      "time_per_iteration": 2.6647751331329346
    },
    {
      "auxiliary_loss_clip": 0.01237248,
      "auxiliary_loss_mlp": 0.0106534,
      "balance_loss_clip": 1.06481838,
      "balance_loss_mlp": 1.03367805,
      "epoch": 0.04467157673230122,
      "flos": 20010970304640.0,
      "grad_norm": 1.9055071619943689,
      "language_loss": 0.83840811,
      "learning_rate": 3.997742813608561e-06,
      "loss": 0.86143398,
      "num_input_tokens_seen": 15952560,
      "step": 743,
      "time_per_iteration": 2.697864055633545
    },
    {
      "auxiliary_loss_clip": 0.01239398,
      "auxiliary_loss_mlp": 0.01073846,
      "balance_loss_clip": 1.06395566,
      "balance_loss_mlp": 1.04373407,
      "epoch": 0.04473169998496919,
      "flos": 18004964480640.0,
      "grad_norm": 2.2041873634107696,
      "language_loss": 0.80026019,
      "learning_rate": 3.997724277684479e-06,
      "loss": 0.82339263,
      "num_input_tokens_seen": 15970620,
      "step": 744,
      "time_per_iteration": 2.6551101207733154
    },
    {
      "auxiliary_loss_clip": 0.01236158,
      "auxiliary_loss_mlp": 0.01076186,
      "balance_loss_clip": 1.06385589,
      "balance_loss_mlp": 1.04665816,
      "epoch": 0.044791823237637154,
      "flos": 20631722169600.0,
      "grad_norm": 2.139129927663487,
      "language_loss": 0.85502481,
      "learning_rate": 3.99770566600649e-06,
      "loss": 0.87814826,
      "num_input_tokens_seen": 15987325,
      "step": 745,
      "time_per_iteration": 2.6686010360717773
    },
    {
      "auxiliary_loss_clip": 0.01235001,
      "auxiliary_loss_mlp": 0.01066107,
      "balance_loss_clip": 1.06320596,
      "balance_loss_mlp": 1.03594685,
      "epoch": 0.04485194649030513,
      "flos": 31176171918720.0,
      "grad_norm": 1.8251828520192552,
      "language_loss": 0.69291008,
      "learning_rate": 3.997686978575302e-06,
      "loss": 0.71592116,
      "num_input_tokens_seen": 16008310,
      "step": 746,
      "time_per_iteration": 2.6782095432281494
    },
    {
      "auxiliary_loss_clip": 0.01244022,
      "auxiliary_loss_mlp": 0.01081644,
      "balance_loss_clip": 1.07012939,
      "balance_loss_mlp": 1.05000615,
      "epoch": 0.04491206974297309,
      "flos": 26143291831680.0,
      "grad_norm": 3.6053643469900982,
      "language_loss": 0.68531066,
      "learning_rate": 3.997668215391625e-06,
      "loss": 0.70856726,
      "num_input_tokens_seen": 16029620,
      "step": 747,
      "time_per_iteration": 2.6589114665985107
    },
    {
      "auxiliary_loss_clip": 0.0124018,
      "auxiliary_loss_mlp": 0.01083594,
      "balance_loss_clip": 1.0652504,
      "balance_loss_mlp": 1.05183625,
      "epoch": 0.044972192995641064,
      "flos": 20667668705280.0,
      "grad_norm": 1.8376208182131786,
      "language_loss": 0.66778374,
      "learning_rate": 3.997649376456168e-06,
      "loss": 0.69102144,
      "num_input_tokens_seen": 16049065,
      "step": 748,
      "time_per_iteration": 2.674691677093506
    },
    {
      "auxiliary_loss_clip": 0.01243343,
      "auxiliary_loss_mlp": 0.01085665,
      "balance_loss_clip": 1.07101417,
      "balance_loss_mlp": 1.05596995,
      "epoch": 0.045032316248309036,
      "flos": 16106834177280.0,
      "grad_norm": 2.4197486882062322,
      "language_loss": 0.76684916,
      "learning_rate": 3.997630461769647e-06,
      "loss": 0.7901392,
      "num_input_tokens_seen": 16066765,
      "step": 749,
      "time_per_iteration": 2.5940611362457275
    },
    {
      "auxiliary_loss_clip": 0.01243381,
      "auxiliary_loss_mlp": 0.01083303,
      "balance_loss_clip": 1.06892776,
      "balance_loss_mlp": 1.05338168,
      "epoch": 0.045092439500977,
      "flos": 17858843953920.0,
      "grad_norm": 1.926675828378473,
      "language_loss": 0.88739896,
      "learning_rate": 3.997611471332778e-06,
      "loss": 0.91066581,
      "num_input_tokens_seen": 16085980,
      "step": 750,
      "time_per_iteration": 2.551717758178711
    },
    {
      "auxiliary_loss_clip": 0.01238484,
      "auxiliary_loss_mlp": 0.01077419,
      "balance_loss_clip": 1.062783,
      "balance_loss_mlp": 1.04404092,
      "epoch": 0.04515256275364497,
      "flos": 24462815990400.0,
      "grad_norm": 3.4910287963746116,
      "language_loss": 0.74371743,
      "learning_rate": 3.9975924051462825e-06,
      "loss": 0.76687646,
      "num_input_tokens_seen": 16106260,
      "step": 751,
      "time_per_iteration": 2.6299028396606445
    },
    {
      "auxiliary_loss_clip": 0.0123577,
      "auxiliary_loss_mlp": 0.01078322,
      "balance_loss_clip": 1.06347609,
      "balance_loss_mlp": 1.04884171,
      "epoch": 0.04521268600631294,
      "flos": 20916385453440.0,
      "grad_norm": 3.3938056459605583,
      "language_loss": 0.69115144,
      "learning_rate": 3.997573263210883e-06,
      "loss": 0.71429229,
      "num_input_tokens_seen": 16123475,
      "step": 752,
      "time_per_iteration": 2.571223020553589
    },
    {
      "auxiliary_loss_clip": 0.01235899,
      "auxiliary_loss_mlp": 0.01060876,
      "balance_loss_clip": 1.0627141,
      "balance_loss_mlp": 1.03212225,
      "epoch": 0.04527280925898091,
      "flos": 13371374954880.0,
      "grad_norm": 2.69328062598792,
      "language_loss": 0.92126763,
      "learning_rate": 3.997554045527305e-06,
      "loss": 0.94423538,
      "num_input_tokens_seen": 16138335,
      "step": 753,
      "time_per_iteration": 2.6100237369537354
    },
    {
      "auxiliary_loss_clip": 0.01239023,
      "auxiliary_loss_mlp": 0.01080271,
      "balance_loss_clip": 1.06628633,
      "balance_loss_mlp": 1.05116034,
      "epoch": 0.04533293251164888,
      "flos": 23254565276160.0,
      "grad_norm": 4.138305317267875,
      "language_loss": 0.91373456,
      "learning_rate": 3.997534752096277e-06,
      "loss": 0.93692756,
      "num_input_tokens_seen": 16157110,
      "step": 754,
      "time_per_iteration": 2.642747402191162
    },
    {
      "auxiliary_loss_clip": 0.01229195,
      "auxiliary_loss_mlp": 0.01078016,
      "balance_loss_clip": 1.06402516,
      "balance_loss_mlp": 1.04725957,
      "epoch": 0.04539305576431685,
      "flos": 12422004537600.0,
      "grad_norm": 4.559941934311277,
      "language_loss": 0.78558046,
      "learning_rate": 3.997515382918531e-06,
      "loss": 0.80865264,
      "num_input_tokens_seen": 16174155,
      "step": 755,
      "time_per_iteration": 2.6316659450531006
    },
    {
      "auxiliary_loss_clip": 0.01240044,
      "auxiliary_loss_mlp": 0.01081048,
      "balance_loss_clip": 1.06624937,
      "balance_loss_mlp": 1.05099559,
      "epoch": 0.04545317901698482,
      "flos": 16070995382400.0,
      "grad_norm": 2.193539224658874,
      "language_loss": 0.78473848,
      "learning_rate": 3.9974959379948015e-06,
      "loss": 0.80794942,
      "num_input_tokens_seen": 16192240,
      "step": 756,
      "time_per_iteration": 2.6390748023986816
    },
    {
      "auxiliary_loss_clip": 0.01101224,
      "auxiliary_loss_mlp": 0.01013849,
      "balance_loss_clip": 1.02455997,
      "balance_loss_mlp": 1.0089612,
      "epoch": 0.045513302269652785,
      "flos": 66396139021440.0,
      "grad_norm": 0.8202876780471967,
      "language_loss": 0.62756521,
      "learning_rate": 3.997476417325827e-06,
      "loss": 0.64871597,
      "num_input_tokens_seen": 16255775,
      "step": 757,
      "time_per_iteration": 3.2393198013305664
    },
    {
      "auxiliary_loss_clip": 0.01235136,
      "auxiliary_loss_mlp": 0.01071767,
      "balance_loss_clip": 1.06455243,
      "balance_loss_mlp": 1.04346693,
      "epoch": 0.04557342552232076,
      "flos": 21471169991040.0,
      "grad_norm": 1.6528285304744148,
      "language_loss": 0.84211069,
      "learning_rate": 3.997456820912346e-06,
      "loss": 0.86517978,
      "num_input_tokens_seen": 16277015,
      "step": 758,
      "time_per_iteration": 2.6508655548095703
    },
    {
      "auxiliary_loss_clip": 0.01228461,
      "auxiliary_loss_mlp": 0.01067033,
      "balance_loss_clip": 1.05912399,
      "balance_loss_mlp": 1.0391618,
      "epoch": 0.04563354877498873,
      "flos": 23732680233600.0,
      "grad_norm": 2.695805662282291,
      "language_loss": 0.88150775,
      "learning_rate": 3.997437148755101e-06,
      "loss": 0.9044627,
      "num_input_tokens_seen": 16296005,
      "step": 759,
      "time_per_iteration": 2.7782890796661377
    },
    {
      "auxiliary_loss_clip": 0.01240589,
      "auxiliary_loss_mlp": 0.01078815,
      "balance_loss_clip": 1.06747675,
      "balance_loss_mlp": 1.04846466,
      "epoch": 0.045693672027656694,
      "flos": 25735741142400.0,
      "grad_norm": 2.392455009776849,
      "language_loss": 0.73440695,
      "learning_rate": 3.9974174008548405e-06,
      "loss": 0.75760102,
      "num_input_tokens_seen": 16315300,
      "step": 760,
      "time_per_iteration": 2.7138822078704834
    },
    {
      "auxiliary_loss_clip": 0.01240372,
      "auxiliary_loss_mlp": 0.01079791,
      "balance_loss_clip": 1.07095265,
      "balance_loss_mlp": 1.05162191,
      "epoch": 0.045753795280324666,
      "flos": 19719016560000.0,
      "grad_norm": 3.497321311688565,
      "language_loss": 0.81781888,
      "learning_rate": 3.9973975772123105e-06,
      "loss": 0.84102058,
      "num_input_tokens_seen": 16333820,
      "step": 761,
      "time_per_iteration": 2.631303310394287
    },
    {
      "auxiliary_loss_clip": 0.01231969,
      "auxiliary_loss_mlp": 0.01078623,
      "balance_loss_clip": 1.06324267,
      "balance_loss_mlp": 1.04922605,
      "epoch": 0.04581391853299264,
      "flos": 23255786338560.0,
      "grad_norm": 2.0632320043111965,
      "language_loss": 0.79811668,
      "learning_rate": 3.997377677828266e-06,
      "loss": 0.82122266,
      "num_input_tokens_seen": 16355290,
      "step": 762,
      "time_per_iteration": 2.646928071975708
    },
    {
      "auxiliary_loss_clip": 0.01093869,
      "auxiliary_loss_mlp": 0.01027943,
      "balance_loss_clip": 1.01857328,
      "balance_loss_mlp": 1.02288842,
      "epoch": 0.0458740417856606,
      "flos": 64231155601920.0,
      "grad_norm": 1.0128965743658471,
      "language_loss": 0.58723813,
      "learning_rate": 3.9973577027034585e-06,
      "loss": 0.60845619,
      "num_input_tokens_seen": 16415995,
      "step": 763,
      "time_per_iteration": 3.1712563037872314
    },
    {
      "auxiliary_loss_clip": 0.012343,
      "auxiliary_loss_mlp": 0.01082461,
      "balance_loss_clip": 1.06205368,
      "balance_loss_mlp": 1.0531354,
      "epoch": 0.045934165038328575,
      "flos": 20770121272320.0,
      "grad_norm": 4.978761831483118,
      "language_loss": 0.87544954,
      "learning_rate": 3.9973376518386475e-06,
      "loss": 0.89861715,
      "num_input_tokens_seen": 16433120,
      "step": 764,
      "time_per_iteration": 2.5985426902770996
    },
    {
      "auxiliary_loss_clip": 0.01236145,
      "auxiliary_loss_mlp": 0.01087868,
      "balance_loss_clip": 1.06553543,
      "balance_loss_mlp": 1.05854285,
      "epoch": 0.04599428829099654,
      "flos": 30262891691520.0,
      "grad_norm": 2.0894169515773067,
      "language_loss": 0.85966802,
      "learning_rate": 3.997317525234592e-06,
      "loss": 0.88290817,
      "num_input_tokens_seen": 16453360,
      "step": 765,
      "time_per_iteration": 2.6572606563568115
    },
    {
      "auxiliary_loss_clip": 0.01239644,
      "auxiliary_loss_mlp": 0.01077398,
      "balance_loss_clip": 1.06530261,
      "balance_loss_mlp": 1.04573584,
      "epoch": 0.04605441154366451,
      "flos": 23038921975680.0,
      "grad_norm": 2.628046285830335,
      "language_loss": 0.88265938,
      "learning_rate": 3.997297322892056e-06,
      "loss": 0.90582979,
      "num_input_tokens_seen": 16471160,
      "step": 766,
      "time_per_iteration": 2.673226833343506
    },
    {
      "auxiliary_loss_clip": 0.01235506,
      "auxiliary_loss_mlp": 0.0107998,
      "balance_loss_clip": 1.06371713,
      "balance_loss_mlp": 1.05115545,
      "epoch": 0.046114534796332485,
      "flos": 22017407091840.0,
      "grad_norm": 2.343908591401411,
      "language_loss": 0.84302223,
      "learning_rate": 3.997277044811806e-06,
      "loss": 0.86617708,
      "num_input_tokens_seen": 16488940,
      "step": 767,
      "time_per_iteration": 2.683429002761841
    },
    {
      "auxiliary_loss_clip": 0.01236229,
      "auxiliary_loss_mlp": 0.01067844,
      "balance_loss_clip": 1.06769753,
      "balance_loss_mlp": 1.03791094,
      "epoch": 0.04617465804900045,
      "flos": 29862380067840.0,
      "grad_norm": 1.9268984031305718,
      "language_loss": 0.8669976,
      "learning_rate": 3.99725669099461e-06,
      "loss": 0.89003831,
      "num_input_tokens_seen": 16509505,
      "step": 768,
      "time_per_iteration": 2.8125200271606445
    },
    {
      "auxiliary_loss_clip": 0.01234175,
      "auxiliary_loss_mlp": 0.01076069,
      "balance_loss_clip": 1.06150854,
      "balance_loss_mlp": 1.04738712,
      "epoch": 0.04623478130166842,
      "flos": 25630056351360.0,
      "grad_norm": 2.115272554881108,
      "language_loss": 0.75152099,
      "learning_rate": 3.9972362614412395e-06,
      "loss": 0.77462339,
      "num_input_tokens_seen": 16528840,
      "step": 769,
      "time_per_iteration": 2.7286128997802734
    },
    {
      "auxiliary_loss_clip": 0.01229956,
      "auxiliary_loss_mlp": 0.01072391,
      "balance_loss_clip": 1.06326365,
      "balance_loss_mlp": 1.04462695,
      "epoch": 0.04629490455433639,
      "flos": 20449080489600.0,
      "grad_norm": 1.8368669953292174,
      "language_loss": 0.86292851,
      "learning_rate": 3.997215756152471e-06,
      "loss": 0.885952,
      "num_input_tokens_seen": 16548335,
      "step": 770,
      "time_per_iteration": 2.68608021736145
    },
    {
      "auxiliary_loss_clip": 0.01239009,
      "auxiliary_loss_mlp": 0.01072125,
      "balance_loss_clip": 1.06274092,
      "balance_loss_mlp": 1.04284704,
      "epoch": 0.04635502780700436,
      "flos": 23148736830720.0,
      "grad_norm": 2.058802627607224,
      "language_loss": 0.86842889,
      "learning_rate": 3.99719517512908e-06,
      "loss": 0.89154023,
      "num_input_tokens_seen": 16567725,
      "step": 771,
      "time_per_iteration": 2.637509822845459
    },
    {
      "auxiliary_loss_clip": 0.01239449,
      "auxiliary_loss_mlp": 0.01079651,
      "balance_loss_clip": 1.06184912,
      "balance_loss_mlp": 1.04884768,
      "epoch": 0.04641515105967233,
      "flos": 23292020183040.0,
      "grad_norm": 1.87920888608735,
      "language_loss": 0.83691382,
      "learning_rate": 3.997174518371848e-06,
      "loss": 0.8601048,
      "num_input_tokens_seen": 16588175,
      "step": 772,
      "time_per_iteration": 2.745006561279297
    },
    {
      "auxiliary_loss_clip": 0.01236322,
      "auxiliary_loss_mlp": 0.0107061,
      "balance_loss_clip": 1.06672883,
      "balance_loss_mlp": 1.04220271,
      "epoch": 0.046475274312340296,
      "flos": 25115204759040.0,
      "grad_norm": 1.9655107083336736,
      "language_loss": 0.73639083,
      "learning_rate": 3.997153785881557e-06,
      "loss": 0.75946015,
      "num_input_tokens_seen": 16607735,
      "step": 773,
      "time_per_iteration": 2.869290828704834
    },
    {
      "auxiliary_loss_clip": 0.01231219,
      "auxiliary_loss_mlp": 0.01071681,
      "balance_loss_clip": 1.06529772,
      "balance_loss_mlp": 1.04054356,
      "epoch": 0.04653539756500827,
      "flos": 25264916645760.0,
      "grad_norm": 2.096431798380756,
      "language_loss": 0.78228974,
      "learning_rate": 3.997132977658996e-06,
      "loss": 0.80531871,
      "num_input_tokens_seen": 16627225,
      "step": 774,
      "time_per_iteration": 2.6967568397521973
    },
    {
      "auxiliary_loss_clip": 0.01230587,
      "auxiliary_loss_mlp": 0.01069519,
      "balance_loss_clip": 1.06347871,
      "balance_loss_mlp": 1.04131365,
      "epoch": 0.046595520817676234,
      "flos": 35404150089600.0,
      "grad_norm": 2.018140205527256,
      "language_loss": 0.73187691,
      "learning_rate": 3.997112093704952e-06,
      "loss": 0.75487792,
      "num_input_tokens_seen": 16647785,
      "step": 775,
      "time_per_iteration": 2.737140417098999
    },
    {
      "auxiliary_loss_clip": 0.01231996,
      "auxiliary_loss_mlp": 0.01066454,
      "balance_loss_clip": 1.06187618,
      "balance_loss_mlp": 1.03650832,
      "epoch": 0.046655644070344206,
      "flos": 18112516778880.0,
      "grad_norm": 1.668093168561758,
      "language_loss": 0.77180624,
      "learning_rate": 3.997091134020217e-06,
      "loss": 0.7947908,
      "num_input_tokens_seen": 16667555,
      "step": 776,
      "time_per_iteration": 4.154085159301758
    },
    {
      "auxiliary_loss_clip": 0.0122577,
      "auxiliary_loss_mlp": 0.01071334,
      "balance_loss_clip": 1.06031108,
      "balance_loss_mlp": 1.04352236,
      "epoch": 0.04671576732301218,
      "flos": 29205286617600.0,
      "grad_norm": 1.9054628166827923,
      "language_loss": 0.7087816,
      "learning_rate": 3.997070098605585e-06,
      "loss": 0.73175263,
      "num_input_tokens_seen": 16686875,
      "step": 777,
      "time_per_iteration": 4.176887512207031
    },
    {
      "auxiliary_loss_clip": 0.0122979,
      "auxiliary_loss_mlp": 0.01076806,
      "balance_loss_clip": 1.06275606,
      "balance_loss_mlp": 1.04705119,
      "epoch": 0.04677589057568014,
      "flos": 30478319510400.0,
      "grad_norm": 1.8083238359854679,
      "language_loss": 0.77069759,
      "learning_rate": 3.997048987461856e-06,
      "loss": 0.79376352,
      "num_input_tokens_seen": 16706420,
      "step": 778,
      "time_per_iteration": 5.943394422531128
    },
    {
      "auxiliary_loss_clip": 0.01227067,
      "auxiliary_loss_mlp": 0.01064982,
      "balance_loss_clip": 1.06043744,
      "balance_loss_mlp": 1.03563297,
      "epoch": 0.046836013828348115,
      "flos": 20557674282240.0,
      "grad_norm": 2.1737778598926463,
      "language_loss": 0.79181123,
      "learning_rate": 3.997027800589829e-06,
      "loss": 0.81473172,
      "num_input_tokens_seen": 16726390,
      "step": 779,
      "time_per_iteration": 2.611804485321045
    },
    {
      "auxiliary_loss_clip": 0.01219629,
      "auxiliary_loss_mlp": 0.01070238,
      "balance_loss_clip": 1.05842376,
      "balance_loss_mlp": 1.04271269,
      "epoch": 0.04689613708101608,
      "flos": 25447378757760.0,
      "grad_norm": 1.888854926622149,
      "language_loss": 0.77364886,
      "learning_rate": 3.997006537990308e-06,
      "loss": 0.79654753,
      "num_input_tokens_seen": 16748965,
      "step": 780,
      "time_per_iteration": 2.668239116668701
    },
    {
      "auxiliary_loss_clip": 0.012253,
      "auxiliary_loss_mlp": 0.01073321,
      "balance_loss_clip": 1.06098521,
      "balance_loss_mlp": 1.04605746,
      "epoch": 0.04695626033368405,
      "flos": 23001395241600.0,
      "grad_norm": 1.7616538282563206,
      "language_loss": 0.76700419,
      "learning_rate": 3.996985199664099e-06,
      "loss": 0.78999043,
      "num_input_tokens_seen": 16768620,
      "step": 781,
      "time_per_iteration": 2.5979926586151123
    },
    {
      "auxiliary_loss_clip": 0.01236637,
      "auxiliary_loss_mlp": 0.01077479,
      "balance_loss_clip": 1.0639379,
      "balance_loss_mlp": 1.04836786,
      "epoch": 0.047016383586352024,
      "flos": 29133357632640.0,
      "grad_norm": 3.0946494667490856,
      "language_loss": 0.73786414,
      "learning_rate": 3.99696378561201e-06,
      "loss": 0.76100528,
      "num_input_tokens_seen": 16789755,
      "step": 782,
      "time_per_iteration": 2.708855390548706
    },
    {
      "auxiliary_loss_clip": 0.0122968,
      "auxiliary_loss_mlp": 0.01069368,
      "balance_loss_clip": 1.06431556,
      "balance_loss_mlp": 1.04253423,
      "epoch": 0.04707650683901999,
      "flos": 14976330451200.0,
      "grad_norm": 2.1459158015790183,
      "language_loss": 0.80524659,
      "learning_rate": 3.996942295834855e-06,
      "loss": 0.82823706,
      "num_input_tokens_seen": 16807585,
      "step": 783,
      "time_per_iteration": 2.6355738639831543
    },
    {
      "auxiliary_loss_clip": 0.01222415,
      "auxiliary_loss_mlp": 0.01063155,
      "balance_loss_clip": 1.06221437,
      "balance_loss_mlp": 1.03663135,
      "epoch": 0.04713663009168796,
      "flos": 21651118151040.0,
      "grad_norm": 1.9084512066318515,
      "language_loss": 0.81687874,
      "learning_rate": 3.996920730333448e-06,
      "loss": 0.83973444,
      "num_input_tokens_seen": 16827220,
      "step": 784,
      "time_per_iteration": 2.64365291595459
    },
    {
      "auxiliary_loss_clip": 0.01226632,
      "auxiliary_loss_mlp": 0.01074549,
      "balance_loss_clip": 1.0582943,
      "balance_loss_mlp": 1.04719007,
      "epoch": 0.04719675334435593,
      "flos": 21325408600320.0,
      "grad_norm": 3.970707764370453,
      "language_loss": 0.80619848,
      "learning_rate": 3.996899089108607e-06,
      "loss": 0.82921028,
      "num_input_tokens_seen": 16846230,
      "step": 785,
      "time_per_iteration": 2.682971715927124
    },
    {
      "auxiliary_loss_clip": 0.01231621,
      "auxiliary_loss_mlp": 0.01063774,
      "balance_loss_clip": 1.06683421,
      "balance_loss_mlp": 1.03784585,
      "epoch": 0.0472568765970239,
      "flos": 17931383470080.0,
      "grad_norm": 2.074448818096939,
      "language_loss": 0.89784658,
      "learning_rate": 3.996877372161152e-06,
      "loss": 0.92080051,
      "num_input_tokens_seen": 16865325,
      "step": 786,
      "time_per_iteration": 2.6072235107421875
    },
    {
      "auxiliary_loss_clip": 0.01227201,
      "auxiliary_loss_mlp": 0.01069453,
      "balance_loss_clip": 1.05475712,
      "balance_loss_mlp": 1.03912568,
      "epoch": 0.04731699984969187,
      "flos": 18077324428800.0,
      "grad_norm": 6.783818284100465,
      "language_loss": 0.76794451,
      "learning_rate": 3.9968555794919065e-06,
      "loss": 0.79091108,
      "num_input_tokens_seen": 16882930,
      "step": 787,
      "time_per_iteration": 2.595069646835327
    },
    {
      "auxiliary_loss_clip": 0.01233526,
      "auxiliary_loss_mlp": 0.01070856,
      "balance_loss_clip": 1.06563127,
      "balance_loss_mlp": 1.04248405,
      "epoch": 0.047377123102359836,
      "flos": 23185078416000.0,
      "grad_norm": 2.309745026689568,
      "language_loss": 0.81301165,
      "learning_rate": 3.996833711101698e-06,
      "loss": 0.83605546,
      "num_input_tokens_seen": 16900710,
      "step": 788,
      "time_per_iteration": 2.633812427520752
    },
    {
      "auxiliary_loss_clip": 0.01225447,
      "auxiliary_loss_mlp": 0.01078934,
      "balance_loss_clip": 1.06370282,
      "balance_loss_mlp": 1.04934621,
      "epoch": 0.04743724635502781,
      "flos": 22747794243840.0,
      "grad_norm": 2.941245147417381,
      "language_loss": 0.84428835,
      "learning_rate": 3.996811766991355e-06,
      "loss": 0.86733222,
      "num_input_tokens_seen": 16919210,
      "step": 789,
      "time_per_iteration": 2.6711082458496094
    },
    {
      "auxiliary_loss_clip": 0.01230866,
      "auxiliary_loss_mlp": 0.01071483,
      "balance_loss_clip": 1.06367648,
      "balance_loss_mlp": 1.0441606,
      "epoch": 0.04749736960769577,
      "flos": 17238702620160.0,
      "grad_norm": 2.0289407228390615,
      "language_loss": 0.81787878,
      "learning_rate": 3.996789747161709e-06,
      "loss": 0.84090227,
      "num_input_tokens_seen": 16937125,
      "step": 790,
      "time_per_iteration": 2.6136717796325684
    },
    {
      "auxiliary_loss_clip": 0.01224033,
      "auxiliary_loss_mlp": 0.01064065,
      "balance_loss_clip": 1.05880189,
      "balance_loss_mlp": 1.03546715,
      "epoch": 0.047557492860363745,
      "flos": 40479261592320.0,
      "grad_norm": 2.9735437778568965,
      "language_loss": 0.88116109,
      "learning_rate": 3.996767651613597e-06,
      "loss": 0.90404207,
      "num_input_tokens_seen": 16958610,
      "step": 791,
      "time_per_iteration": 2.747586727142334
    },
    {
      "auxiliary_loss_clip": 0.01226267,
      "auxiliary_loss_mlp": 0.01066471,
      "balance_loss_clip": 1.06144643,
      "balance_loss_mlp": 1.03743124,
      "epoch": 0.04761761611303172,
      "flos": 18698004466560.0,
      "grad_norm": 2.1239226540804537,
      "language_loss": 0.90671498,
      "learning_rate": 3.996745480347854e-06,
      "loss": 0.92964232,
      "num_input_tokens_seen": 16977300,
      "step": 792,
      "time_per_iteration": 2.591477870941162
    },
    {
      "auxiliary_loss_clip": 0.01226882,
      "auxiliary_loss_mlp": 0.0107926,
      "balance_loss_clip": 1.05968022,
      "balance_loss_mlp": 1.05225897,
      "epoch": 0.04767773936569968,
      "flos": 20921987975040.0,
      "grad_norm": 1.9120988315570397,
      "language_loss": 0.73246223,
      "learning_rate": 3.996723233365324e-06,
      "loss": 0.75552362,
      "num_input_tokens_seen": 16994950,
      "step": 793,
      "time_per_iteration": 2.6319899559020996
    },
    {
      "auxiliary_loss_clip": 0.01231301,
      "auxiliary_loss_mlp": 0.01070716,
      "balance_loss_clip": 1.06213653,
      "balance_loss_mlp": 1.04146254,
      "epoch": 0.047737862618367655,
      "flos": 23732680233600.0,
      "grad_norm": 1.86347948201136,
      "language_loss": 0.86139679,
      "learning_rate": 3.996700910666847e-06,
      "loss": 0.88441694,
      "num_input_tokens_seen": 17014760,
      "step": 794,
      "time_per_iteration": 2.6835687160491943
    },
    {
      "auxiliary_loss_clip": 0.01228204,
      "auxiliary_loss_mlp": 0.01077895,
      "balance_loss_clip": 1.05969596,
      "balance_loss_mlp": 1.04935622,
      "epoch": 0.04779798587103562,
      "flos": 23695764030720.0,
      "grad_norm": 2.370166301863074,
      "language_loss": 0.69069195,
      "learning_rate": 3.996678512253272e-06,
      "loss": 0.71375293,
      "num_input_tokens_seen": 17032715,
      "step": 795,
      "time_per_iteration": 2.669261932373047
    },
    {
      "auxiliary_loss_clip": 0.01225748,
      "auxiliary_loss_mlp": 0.01076275,
      "balance_loss_clip": 1.06129098,
      "balance_loss_mlp": 1.04756904,
      "epoch": 0.04785810912370359,
      "flos": 23183641872000.0,
      "grad_norm": 1.744925212230271,
      "language_loss": 0.810256,
      "learning_rate": 3.996656038125449e-06,
      "loss": 0.83327615,
      "num_input_tokens_seen": 17052215,
      "step": 796,
      "time_per_iteration": 2.5800065994262695
    },
    {
      "auxiliary_loss_clip": 0.01228235,
      "auxiliary_loss_mlp": 0.01065433,
      "balance_loss_clip": 1.06224668,
      "balance_loss_mlp": 1.03638172,
      "epoch": 0.047918232376371564,
      "flos": 18040623707520.0,
      "grad_norm": 1.979164246440182,
      "language_loss": 0.8128069,
      "learning_rate": 3.996633488284228e-06,
      "loss": 0.83574355,
      "num_input_tokens_seen": 17069225,
      "step": 797,
      "time_per_iteration": 2.58878493309021
    },
    {
      "auxiliary_loss_clip": 0.01100259,
      "auxiliary_loss_mlp": 0.01007215,
      "balance_loss_clip": 1.02779806,
      "balance_loss_mlp": 1.00266171,
      "epoch": 0.04797835562903953,
      "flos": 62442588758400.0,
      "grad_norm": 0.912416075283383,
      "language_loss": 0.64532876,
      "learning_rate": 3.996610862730465e-06,
      "loss": 0.66640353,
      "num_input_tokens_seen": 17126680,
      "step": 798,
      "time_per_iteration": 3.0779380798339844
    },
    {
      "auxiliary_loss_clip": 0.01229665,
      "auxiliary_loss_mlp": 0.01068747,
      "balance_loss_clip": 1.05799031,
      "balance_loss_mlp": 1.04121017,
      "epoch": 0.0480384788817075,
      "flos": 21507296094720.0,
      "grad_norm": 2.0206600610723333,
      "language_loss": 0.91274291,
      "learning_rate": 3.996588161465018e-06,
      "loss": 0.935727,
      "num_input_tokens_seen": 17144835,
      "step": 799,
      "time_per_iteration": 2.660438299179077
    },
    {
      "auxiliary_loss_clip": 0.01230751,
      "auxiliary_loss_mlp": 0.010715,
      "balance_loss_clip": 1.06640434,
      "balance_loss_mlp": 1.04274678,
      "epoch": 0.048098602134375466,
      "flos": 21726710323200.0,
      "grad_norm": 2.0752654205923866,
      "language_loss": 0.86825287,
      "learning_rate": 3.996565384488748e-06,
      "loss": 0.89127541,
      "num_input_tokens_seen": 17165030,
      "step": 800,
      "time_per_iteration": 2.6700456142425537
    },
    {
      "auxiliary_loss_clip": 0.01229893,
      "auxiliary_loss_mlp": 0.01072058,
      "balance_loss_clip": 1.06186771,
      "balance_loss_mlp": 1.04618931,
      "epoch": 0.04815872538704344,
      "flos": 22931082368640.0,
      "grad_norm": 2.5310108886746976,
      "language_loss": 0.83949852,
      "learning_rate": 3.996542531802518e-06,
      "loss": 0.86251807,
      "num_input_tokens_seen": 17184895,
      "step": 801,
      "time_per_iteration": 2.7724695205688477
    },
    {
      "auxiliary_loss_clip": 0.01227846,
      "auxiliary_loss_mlp": 0.010756,
      "balance_loss_clip": 1.06226814,
      "balance_loss_mlp": 1.04847932,
      "epoch": 0.04821884863971141,
      "flos": 43174716042240.0,
      "grad_norm": 1.9607091513106172,
      "language_loss": 0.79818648,
      "learning_rate": 3.996519603407196e-06,
      "loss": 0.82122099,
      "num_input_tokens_seen": 17208225,
      "step": 802,
      "time_per_iteration": 2.861309766769409
    },
    {
      "auxiliary_loss_clip": 0.0122832,
      "auxiliary_loss_mlp": 0.01069086,
      "balance_loss_clip": 1.06392837,
      "balance_loss_mlp": 1.04278886,
      "epoch": 0.048278971892379376,
      "flos": 18620006083200.0,
      "grad_norm": 1.798745906633195,
      "language_loss": 0.86600745,
      "learning_rate": 3.996496599303649e-06,
      "loss": 0.88898146,
      "num_input_tokens_seen": 17226305,
      "step": 803,
      "time_per_iteration": 2.612684965133667
    },
    {
      "auxiliary_loss_clip": 0.01222438,
      "auxiliary_loss_mlp": 0.01063116,
      "balance_loss_clip": 1.06214345,
      "balance_loss_mlp": 1.03643703,
      "epoch": 0.04833909514504735,
      "flos": 20230061310720.0,
      "grad_norm": 5.958214069975319,
      "language_loss": 0.85139012,
      "learning_rate": 3.996473519492753e-06,
      "loss": 0.8742457,
      "num_input_tokens_seen": 17244545,
      "step": 804,
      "time_per_iteration": 2.596965789794922
    },
    {
      "auxiliary_loss_clip": 0.01225485,
      "auxiliary_loss_mlp": 0.0106948,
      "balance_loss_clip": 1.06206632,
      "balance_loss_mlp": 1.04222918,
      "epoch": 0.04839921839771532,
      "flos": 24645170361600.0,
      "grad_norm": 1.9492340448514227,
      "language_loss": 0.85939878,
      "learning_rate": 3.99645036397538e-06,
      "loss": 0.88234842,
      "num_input_tokens_seen": 17265730,
      "step": 805,
      "time_per_iteration": 2.6773781776428223
    },
    {
      "auxiliary_loss_clip": 0.01221339,
      "auxiliary_loss_mlp": 0.01071867,
      "balance_loss_clip": 1.05968738,
      "balance_loss_mlp": 1.04591477,
      "epoch": 0.048459341650383285,
      "flos": 24827452905600.0,
      "grad_norm": 1.8764849579047527,
      "language_loss": 0.68025368,
      "learning_rate": 3.9964271327524085e-06,
      "loss": 0.70318574,
      "num_input_tokens_seen": 17284820,
      "step": 806,
      "time_per_iteration": 2.6270596981048584
    },
    {
      "auxiliary_loss_clip": 0.01221043,
      "auxiliary_loss_mlp": 0.01060505,
      "balance_loss_clip": 1.06064904,
      "balance_loss_mlp": 1.03384972,
      "epoch": 0.04851946490305126,
      "flos": 22163204396160.0,
      "grad_norm": 8.586680684018,
      "language_loss": 0.76488906,
      "learning_rate": 3.9964038258247214e-06,
      "loss": 0.78770459,
      "num_input_tokens_seen": 17305085,
      "step": 807,
      "time_per_iteration": 2.6783089637756348
    },
    {
      "auxiliary_loss_clip": 0.01218859,
      "auxiliary_loss_mlp": 0.01068871,
      "balance_loss_clip": 1.05734789,
      "balance_loss_mlp": 1.04290676,
      "epoch": 0.04857958815571922,
      "flos": 19792022952960.0,
      "grad_norm": 2.4056749627509157,
      "language_loss": 0.86882269,
      "learning_rate": 3.9963804431932005e-06,
      "loss": 0.89170003,
      "num_input_tokens_seen": 17322715,
      "step": 808,
      "time_per_iteration": 2.6447641849517822
    },
    {
      "auxiliary_loss_clip": 0.01227529,
      "auxiliary_loss_mlp": 0.01069446,
      "balance_loss_clip": 1.06140316,
      "balance_loss_mlp": 1.0424329,
      "epoch": 0.048639711408387194,
      "flos": 18697968552960.0,
      "grad_norm": 2.6040733531164424,
      "language_loss": 0.89710444,
      "learning_rate": 3.996356984858732e-06,
      "loss": 0.92007422,
      "num_input_tokens_seen": 17341455,
      "step": 809,
      "time_per_iteration": 2.6679790019989014
    },
    {
      "auxiliary_loss_clip": 0.01226608,
      "auxiliary_loss_mlp": 0.01067211,
      "balance_loss_clip": 1.0643065,
      "balance_loss_mlp": 1.04060316,
      "epoch": 0.048699834661055166,
      "flos": 24863507182080.0,
      "grad_norm": 3.0721319202916324,
      "language_loss": 0.84918916,
      "learning_rate": 3.996333450822208e-06,
      "loss": 0.87212729,
      "num_input_tokens_seen": 17360765,
      "step": 810,
      "time_per_iteration": 2.696772575378418
    },
    {
      "auxiliary_loss_clip": 0.01227202,
      "auxiliary_loss_mlp": 0.01067343,
      "balance_loss_clip": 1.0622344,
      "balance_loss_mlp": 1.04049683,
      "epoch": 0.04875995791372313,
      "flos": 20704010290560.0,
      "grad_norm": 1.8136675943398954,
      "language_loss": 0.80799425,
      "learning_rate": 3.99630984108452e-06,
      "loss": 0.83093977,
      "num_input_tokens_seen": 17380625,
      "step": 811,
      "time_per_iteration": 2.653808355331421
    },
    {
      "auxiliary_loss_clip": 0.01217843,
      "auxiliary_loss_mlp": 0.01070621,
      "balance_loss_clip": 1.05928314,
      "balance_loss_mlp": 1.04466903,
      "epoch": 0.048820081166391104,
      "flos": 18588297352320.0,
      "grad_norm": 1.7193599003225197,
      "language_loss": 0.74634516,
      "learning_rate": 3.9962861556465615e-06,
      "loss": 0.76922977,
      "num_input_tokens_seen": 17399355,
      "step": 812,
      "time_per_iteration": 2.7274649143218994
    },
    {
      "auxiliary_loss_clip": 0.01222659,
      "auxiliary_loss_mlp": 0.01073562,
      "balance_loss_clip": 1.06445217,
      "balance_loss_mlp": 1.04862356,
      "epoch": 0.04888020441905907,
      "flos": 22707322594560.0,
      "grad_norm": 1.9311665765462733,
      "language_loss": 0.90124279,
      "learning_rate": 3.996262394509233e-06,
      "loss": 0.92420495,
      "num_input_tokens_seen": 17418240,
      "step": 813,
      "time_per_iteration": 2.654874801635742
    },
    {
      "auxiliary_loss_clip": 0.0122,
      "auxiliary_loss_mlp": 0.01057827,
      "balance_loss_clip": 1.06157589,
      "balance_loss_mlp": 1.03248262,
      "epoch": 0.04894032767172704,
      "flos": 22784351310720.0,
      "grad_norm": 1.9238840150723209,
      "language_loss": 0.74904704,
      "learning_rate": 3.9962385576734335e-06,
      "loss": 0.77182531,
      "num_input_tokens_seen": 17436250,
      "step": 814,
      "time_per_iteration": 2.7381603717803955
    },
    {
      "auxiliary_loss_clip": 0.01223782,
      "auxiliary_loss_mlp": 0.01069686,
      "balance_loss_clip": 1.06125045,
      "balance_loss_mlp": 1.04289961,
      "epoch": 0.04900045092439501,
      "flos": 25516147345920.0,
      "grad_norm": 2.1966001004582596,
      "language_loss": 0.83816808,
      "learning_rate": 3.9962146451400675e-06,
      "loss": 0.86110282,
      "num_input_tokens_seen": 17455750,
      "step": 815,
      "time_per_iteration": 2.7289621829986572
    },
    {
      "auxiliary_loss_clip": 0.01227011,
      "auxiliary_loss_mlp": 0.01060571,
      "balance_loss_clip": 1.06326818,
      "balance_loss_mlp": 1.0344646,
      "epoch": 0.04906057417706298,
      "flos": 25958136199680.0,
      "grad_norm": 2.3329994981275943,
      "language_loss": 0.90796101,
      "learning_rate": 3.996190656910043e-06,
      "loss": 0.93083686,
      "num_input_tokens_seen": 17474995,
      "step": 816,
      "time_per_iteration": 4.174290180206299
    },
    {
      "auxiliary_loss_clip": 0.01226278,
      "auxiliary_loss_mlp": 0.0105651,
      "balance_loss_clip": 1.06172895,
      "balance_loss_mlp": 1.03054583,
      "epoch": 0.04912069742973095,
      "flos": 18624638937600.0,
      "grad_norm": 2.2253098946667853,
      "language_loss": 0.79834002,
      "learning_rate": 3.996166592984268e-06,
      "loss": 0.82116789,
      "num_input_tokens_seen": 17493395,
      "step": 817,
      "time_per_iteration": 4.2819907665252686
    },
    {
      "auxiliary_loss_clip": 0.01222491,
      "auxiliary_loss_mlp": 0.01072358,
      "balance_loss_clip": 1.06228495,
      "balance_loss_mlp": 1.04563141,
      "epoch": 0.049180820682398915,
      "flos": 23699786353920.0,
      "grad_norm": 1.9292138186207266,
      "language_loss": 0.8532303,
      "learning_rate": 3.996142453363656e-06,
      "loss": 0.8761788,
      "num_input_tokens_seen": 17514565,
      "step": 818,
      "time_per_iteration": 7.687308073043823
    },
    {
      "auxiliary_loss_clip": 0.01228571,
      "auxiliary_loss_mlp": 0.01064433,
      "balance_loss_clip": 1.06170368,
      "balance_loss_mlp": 1.0369786,
      "epoch": 0.04924094393506689,
      "flos": 22420396753920.0,
      "grad_norm": 2.1064810754058407,
      "language_loss": 0.75623614,
      "learning_rate": 3.996118238049124e-06,
      "loss": 0.77916616,
      "num_input_tokens_seen": 17534590,
      "step": 819,
      "time_per_iteration": 2.5708072185516357
    },
    {
      "auxiliary_loss_clip": 0.01227988,
      "auxiliary_loss_mlp": 0.010616,
      "balance_loss_clip": 1.06580663,
      "balance_loss_mlp": 1.03785336,
      "epoch": 0.04930106718773486,
      "flos": 15738246766080.0,
      "grad_norm": 2.8685299631500487,
      "language_loss": 0.85082126,
      "learning_rate": 3.996093947041586e-06,
      "loss": 0.87371719,
      "num_input_tokens_seen": 17551900,
      "step": 820,
      "time_per_iteration": 2.695204973220825
    },
    {
      "auxiliary_loss_clip": 0.01224953,
      "auxiliary_loss_mlp": 0.01065985,
      "balance_loss_clip": 1.06082845,
      "balance_loss_mlp": 1.04037917,
      "epoch": 0.049361190440402825,
      "flos": 26250628648320.0,
      "grad_norm": 1.734636988660555,
      "language_loss": 0.90459162,
      "learning_rate": 3.996069580341966e-06,
      "loss": 0.92750102,
      "num_input_tokens_seen": 17571485,
      "step": 821,
      "time_per_iteration": 2.6284992694854736
    },
    {
      "auxiliary_loss_clip": 0.01222526,
      "auxiliary_loss_mlp": 0.01080357,
      "balance_loss_clip": 1.06015635,
      "balance_loss_mlp": 1.05485809,
      "epoch": 0.0494213136930708,
      "flos": 21252366293760.0,
      "grad_norm": 1.7915267676548876,
      "language_loss": 0.89795959,
      "learning_rate": 3.996045137951188e-06,
      "loss": 0.92098844,
      "num_input_tokens_seen": 17591410,
      "step": 822,
      "time_per_iteration": 2.6085855960845947
    },
    {
      "auxiliary_loss_clip": 0.0122571,
      "auxiliary_loss_mlp": 0.01062887,
      "balance_loss_clip": 1.0639379,
      "balance_loss_mlp": 1.03472972,
      "epoch": 0.04948143694573876,
      "flos": 27965506740480.0,
      "grad_norm": 2.28747155105076,
      "language_loss": 0.67558801,
      "learning_rate": 3.996020619870178e-06,
      "loss": 0.69847399,
      "num_input_tokens_seen": 17612010,
      "step": 823,
      "time_per_iteration": 2.644277572631836
    },
    {
      "auxiliary_loss_clip": 0.01099376,
      "auxiliary_loss_mlp": 0.0100741,
      "balance_loss_clip": 1.0267303,
      "balance_loss_mlp": 1.00266516,
      "epoch": 0.049541560198406734,
      "flos": 66180995533440.0,
      "grad_norm": 1.3456360586087317,
      "language_loss": 0.62254131,
      "learning_rate": 3.995996026099866e-06,
      "loss": 0.64360917,
      "num_input_tokens_seen": 17673430,
      "step": 824,
      "time_per_iteration": 3.230381488800049
    },
    {
      "auxiliary_loss_clip": 0.01228758,
      "auxiliary_loss_mlp": 0.01066541,
      "balance_loss_clip": 1.06346989,
      "balance_loss_mlp": 1.03909945,
      "epoch": 0.049601683451074706,
      "flos": 22892693708160.0,
      "grad_norm": 1.8854339538524305,
      "language_loss": 0.90479428,
      "learning_rate": 3.995971356641185e-06,
      "loss": 0.92774737,
      "num_input_tokens_seen": 17689545,
      "step": 825,
      "time_per_iteration": 2.58868670463562
    },
    {
      "auxiliary_loss_clip": 0.01227734,
      "auxiliary_loss_mlp": 0.01066527,
      "balance_loss_clip": 1.06315517,
      "balance_loss_mlp": 1.03844118,
      "epoch": 0.04966180670374267,
      "flos": 21433643256960.0,
      "grad_norm": 2.307419213246734,
      "language_loss": 0.66851091,
      "learning_rate": 3.9959466114950695e-06,
      "loss": 0.69145352,
      "num_input_tokens_seen": 17705965,
      "step": 826,
      "time_per_iteration": 2.59468412399292
    },
    {
      "auxiliary_loss_clip": 0.01230149,
      "auxiliary_loss_mlp": 0.01069061,
      "balance_loss_clip": 1.06421614,
      "balance_loss_mlp": 1.04216766,
      "epoch": 0.04972192995641064,
      "flos": 23107367341440.0,
      "grad_norm": 1.8316571551414482,
      "language_loss": 0.78298402,
      "learning_rate": 3.995921790662459e-06,
      "loss": 0.80597603,
      "num_input_tokens_seen": 17724580,
      "step": 827,
      "time_per_iteration": 2.7148005962371826
    },
    {
      "auxiliary_loss_clip": 0.01230507,
      "auxiliary_loss_mlp": 0.01079145,
      "balance_loss_clip": 1.06385946,
      "balance_loss_mlp": 1.05119085,
      "epoch": 0.04978205320907861,
      "flos": 40406147458560.0,
      "grad_norm": 1.6017511297862308,
      "language_loss": 0.78696525,
      "learning_rate": 3.995896894144294e-06,
      "loss": 0.81006181,
      "num_input_tokens_seen": 17747755,
      "step": 828,
      "time_per_iteration": 2.86991548538208
    },
    {
      "auxiliary_loss_clip": 0.0121958,
      "auxiliary_loss_mlp": 0.01059689,
      "balance_loss_clip": 1.05939984,
      "balance_loss_mlp": 1.03390431,
      "epoch": 0.04984217646174658,
      "flos": 25228539146880.0,
      "grad_norm": 2.48577103336206,
      "language_loss": 0.83530867,
      "learning_rate": 3.995871921941519e-06,
      "loss": 0.85810131,
      "num_input_tokens_seen": 17768550,
      "step": 829,
      "time_per_iteration": 2.655895948410034
    },
    {
      "auxiliary_loss_clip": 0.01226863,
      "auxiliary_loss_mlp": 0.01080723,
      "balance_loss_clip": 1.06109536,
      "balance_loss_mlp": 1.05068195,
      "epoch": 0.04990229971441455,
      "flos": 15959636242560.0,
      "grad_norm": 2.078538436430036,
      "language_loss": 0.74857247,
      "learning_rate": 3.99584687405508e-06,
      "loss": 0.77164829,
      "num_input_tokens_seen": 17786080,
      "step": 830,
      "time_per_iteration": 2.5820400714874268
    },
    {
      "auxiliary_loss_clip": 0.0122584,
      "auxiliary_loss_mlp": 0.01074077,
      "balance_loss_clip": 1.06154907,
      "balance_loss_mlp": 1.04667115,
      "epoch": 0.04996242296708252,
      "flos": 18405116968320.0,
      "grad_norm": 1.8327841960194244,
      "language_loss": 0.79279459,
      "learning_rate": 3.995821750485929e-06,
      "loss": 0.81579381,
      "num_input_tokens_seen": 17803635,
      "step": 831,
      "time_per_iteration": 2.5980231761932373
    },
    {
      "auxiliary_loss_clip": 0.01173206,
      "auxiliary_loss_mlp": 0.01072743,
      "balance_loss_clip": 1.0542444,
      "balance_loss_mlp": 1.04725623,
      "epoch": 0.05002254621975049,
      "flos": 17858053854720.0,
      "grad_norm": 3.034319898285603,
      "language_loss": 0.91497368,
      "learning_rate": 3.995796551235016e-06,
      "loss": 0.93743312,
      "num_input_tokens_seen": 17822190,
      "step": 832,
      "time_per_iteration": 2.7498815059661865
    },
    {
      "auxiliary_loss_clip": 0.01194428,
      "auxiliary_loss_mlp": 0.01081719,
      "balance_loss_clip": 1.05826366,
      "balance_loss_mlp": 1.05667353,
      "epoch": 0.050082669472418455,
      "flos": 45660273367680.0,
      "grad_norm": 1.887029338258115,
      "language_loss": 0.83167893,
      "learning_rate": 3.9957712763032974e-06,
      "loss": 0.85444039,
      "num_input_tokens_seen": 17846915,
      "step": 833,
      "time_per_iteration": 2.863208770751953
    },
    {
      "auxiliary_loss_clip": 0.01199525,
      "auxiliary_loss_mlp": 0.01061962,
      "balance_loss_clip": 1.05888343,
      "balance_loss_mlp": 1.03468657,
      "epoch": 0.05014279272508643,
      "flos": 37962067363200.0,
      "grad_norm": 2.8753922020214033,
      "language_loss": 0.82409853,
      "learning_rate": 3.995745925691733e-06,
      "loss": 0.84671336,
      "num_input_tokens_seen": 17867270,
      "step": 834,
      "time_per_iteration": 2.7868030071258545
    },
    {
      "auxiliary_loss_clip": 0.01216246,
      "auxiliary_loss_mlp": 0.01064427,
      "balance_loss_clip": 1.06272483,
      "balance_loss_mlp": 1.03672278,
      "epoch": 0.0502029159777544,
      "flos": 20996179516800.0,
      "grad_norm": 2.2306487397141646,
      "language_loss": 0.92186153,
      "learning_rate": 3.995720499401282e-06,
      "loss": 0.94466823,
      "num_input_tokens_seen": 17884880,
      "step": 835,
      "time_per_iteration": 2.6224496364593506
    },
    {
      "auxiliary_loss_clip": 0.01229494,
      "auxiliary_loss_mlp": 0.01074922,
      "balance_loss_clip": 1.06143415,
      "balance_loss_mlp": 1.0464313,
      "epoch": 0.050263039230422364,
      "flos": 15888066393600.0,
      "grad_norm": 2.196832783808158,
      "language_loss": 0.76143622,
      "learning_rate": 3.995694997432911e-06,
      "loss": 0.78448039,
      "num_input_tokens_seen": 17903695,
      "step": 836,
      "time_per_iteration": 2.5648462772369385
    },
    {
      "auxiliary_loss_clip": 0.01211162,
      "auxiliary_loss_mlp": 0.01075977,
      "balance_loss_clip": 1.06259084,
      "balance_loss_mlp": 1.04992962,
      "epoch": 0.050323162483090336,
      "flos": 23732752060800.0,
      "grad_norm": 2.100773352560791,
      "language_loss": 0.83627856,
      "learning_rate": 3.9956694197875855e-06,
      "loss": 0.85914999,
      "num_input_tokens_seen": 17920745,
      "step": 837,
      "time_per_iteration": 2.7420156002044678
    },
    {
      "auxiliary_loss_clip": 0.01198815,
      "auxiliary_loss_mlp": 0.0078439,
      "balance_loss_clip": 1.06345344,
      "balance_loss_mlp": 1.00053763,
      "epoch": 0.0503832857357583,
      "flos": 20266223328000.0,
      "grad_norm": 2.1353335821274477,
      "language_loss": 0.72857559,
      "learning_rate": 3.995643766466275e-06,
      "loss": 0.7484076,
      "num_input_tokens_seen": 17938220,
      "step": 838,
      "time_per_iteration": 2.679177761077881
    },
    {
      "auxiliary_loss_clip": 0.01189223,
      "auxiliary_loss_mlp": 0.01071526,
      "balance_loss_clip": 1.05415273,
      "balance_loss_mlp": 1.04510927,
      "epoch": 0.05044340898842627,
      "flos": 17785011548160.0,
      "grad_norm": 1.8138261016039334,
      "language_loss": 0.83462799,
      "learning_rate": 3.995618037469953e-06,
      "loss": 0.85723549,
      "num_input_tokens_seen": 17957325,
      "step": 839,
      "time_per_iteration": 2.69063663482666
    },
    {
      "auxiliary_loss_clip": 0.01220356,
      "auxiliary_loss_mlp": 0.01069331,
      "balance_loss_clip": 1.05991399,
      "balance_loss_mlp": 1.04411805,
      "epoch": 0.050503532241094246,
      "flos": 22966526113920.0,
      "grad_norm": 1.7513762525269907,
      "language_loss": 0.85775483,
      "learning_rate": 3.995592232799595e-06,
      "loss": 0.88065171,
      "num_input_tokens_seen": 17975875,
      "step": 840,
      "time_per_iteration": 2.6477303504943848
    },
    {
      "auxiliary_loss_clip": 0.01192112,
      "auxiliary_loss_mlp": 0.01064377,
      "balance_loss_clip": 1.05451894,
      "balance_loss_mlp": 1.036291,
      "epoch": 0.05056365549376221,
      "flos": 22776989022720.0,
      "grad_norm": 1.7956760046069329,
      "language_loss": 0.9457823,
      "learning_rate": 3.99556635245618e-06,
      "loss": 0.96834719,
      "num_input_tokens_seen": 17994340,
      "step": 841,
      "time_per_iteration": 2.8354220390319824
    },
    {
      "auxiliary_loss_clip": 0.0122473,
      "auxiliary_loss_mlp": 0.01070125,
      "balance_loss_clip": 1.06219172,
      "balance_loss_mlp": 1.04329097,
      "epoch": 0.05062377874643018,
      "flos": 30916968399360.0,
      "grad_norm": 2.3106044659054104,
      "language_loss": 0.77566791,
      "learning_rate": 3.995540396440688e-06,
      "loss": 0.79861641,
      "num_input_tokens_seen": 18015260,
      "step": 842,
      "time_per_iteration": 2.6909749507904053
    },
    {
      "auxiliary_loss_clip": 0.01214637,
      "auxiliary_loss_mlp": 0.01071033,
      "balance_loss_clip": 1.06270838,
      "balance_loss_mlp": 1.04391265,
      "epoch": 0.05068390199909815,
      "flos": 19647159402240.0,
      "grad_norm": 2.8849837971101864,
      "language_loss": 0.78126526,
      "learning_rate": 3.995514364754105e-06,
      "loss": 0.80412203,
      "num_input_tokens_seen": 18033960,
      "step": 843,
      "time_per_iteration": 2.6534156799316406
    },
    {
      "auxiliary_loss_clip": 0.01212948,
      "auxiliary_loss_mlp": 0.01063612,
      "balance_loss_clip": 1.06317043,
      "balance_loss_mlp": 1.03894806,
      "epoch": 0.05074402525176612,
      "flos": 37962103276800.0,
      "grad_norm": 1.9320015451631862,
      "language_loss": 0.83256191,
      "learning_rate": 3.995488257397417e-06,
      "loss": 0.85532749,
      "num_input_tokens_seen": 18056700,
      "step": 844,
      "time_per_iteration": 2.7682149410247803
    },
    {
      "auxiliary_loss_clip": 0.01216308,
      "auxiliary_loss_mlp": 0.01067162,
      "balance_loss_clip": 1.06307864,
      "balance_loss_mlp": 1.04138875,
      "epoch": 0.05080414850443409,
      "flos": 22054610603520.0,
      "grad_norm": 2.113957107027846,
      "language_loss": 0.77108061,
      "learning_rate": 3.995462074371614e-06,
      "loss": 0.79391527,
      "num_input_tokens_seen": 18075815,
      "step": 845,
      "time_per_iteration": 2.6720399856567383
    },
    {
      "auxiliary_loss_clip": 0.01206643,
      "auxiliary_loss_mlp": 0.01065522,
      "balance_loss_clip": 1.05881417,
      "balance_loss_mlp": 1.03885484,
      "epoch": 0.05086427175710206,
      "flos": 20225787592320.0,
      "grad_norm": 1.8497392628450484,
      "language_loss": 0.87773871,
      "learning_rate": 3.99543581567769e-06,
      "loss": 0.90046036,
      "num_input_tokens_seen": 18095095,
      "step": 846,
      "time_per_iteration": 2.696049690246582
    },
    {
      "auxiliary_loss_clip": 0.01206291,
      "auxiliary_loss_mlp": 0.01069231,
      "balance_loss_clip": 1.06204462,
      "balance_loss_mlp": 1.04330277,
      "epoch": 0.05092439500977003,
      "flos": 15159223526400.0,
      "grad_norm": 1.695550491545423,
      "language_loss": 0.87364423,
      "learning_rate": 3.9954094813166394e-06,
      "loss": 0.89639944,
      "num_input_tokens_seen": 18112675,
      "step": 847,
      "time_per_iteration": 2.666907548904419
    },
    {
      "auxiliary_loss_clip": 0.01175052,
      "auxiliary_loss_mlp": 0.01071976,
      "balance_loss_clip": 1.06267309,
      "balance_loss_mlp": 1.0447005,
      "epoch": 0.050984518262437994,
      "flos": 22055149307520.0,
      "grad_norm": 2.5687168450386637,
      "language_loss": 0.81878662,
      "learning_rate": 3.995383071289462e-06,
      "loss": 0.84125686,
      "num_input_tokens_seen": 18130745,
      "step": 848,
      "time_per_iteration": 2.782135486602783
    },
    {
      "auxiliary_loss_clip": 0.0122638,
      "auxiliary_loss_mlp": 0.01071388,
      "balance_loss_clip": 1.06619906,
      "balance_loss_mlp": 1.04544854,
      "epoch": 0.05104464151510597,
      "flos": 30225329043840.0,
      "grad_norm": 1.678404869397893,
      "language_loss": 0.87187904,
      "learning_rate": 3.995356585597158e-06,
      "loss": 0.89485669,
      "num_input_tokens_seen": 18152410,
      "step": 849,
      "time_per_iteration": 2.787992000579834
    },
    {
      "auxiliary_loss_clip": 0.01220251,
      "auxiliary_loss_mlp": 0.0106131,
      "balance_loss_clip": 1.06049275,
      "balance_loss_mlp": 1.03545308,
      "epoch": 0.05110476476777394,
      "flos": 18332900674560.0,
      "grad_norm": 2.125711462362114,
      "language_loss": 0.8315587,
      "learning_rate": 3.995330024240732e-06,
      "loss": 0.85437429,
      "num_input_tokens_seen": 18170870,
      "step": 850,
      "time_per_iteration": 2.6548752784729004
    },
    {
      "auxiliary_loss_clip": 0.01210598,
      "auxiliary_loss_mlp": 0.01063491,
      "balance_loss_clip": 1.06061506,
      "balance_loss_mlp": 1.0379566,
      "epoch": 0.051164888020441904,
      "flos": 37998732170880.0,
      "grad_norm": 2.2115645013354253,
      "language_loss": 0.65423882,
      "learning_rate": 3.995303387221192e-06,
      "loss": 0.67697972,
      "num_input_tokens_seen": 18191555,
      "step": 851,
      "time_per_iteration": 2.817197322845459
    },
    {
      "auxiliary_loss_clip": 0.0120566,
      "auxiliary_loss_mlp": 0.01075745,
      "balance_loss_clip": 1.05822444,
      "balance_loss_mlp": 1.04761147,
      "epoch": 0.051225011273109876,
      "flos": 23038634666880.0,
      "grad_norm": 2.3720786299251073,
      "language_loss": 0.83587611,
      "learning_rate": 3.995276674539547e-06,
      "loss": 0.8586902,
      "num_input_tokens_seen": 18208620,
      "step": 852,
      "time_per_iteration": 2.685727119445801
    },
    {
      "auxiliary_loss_clip": 0.01193575,
      "auxiliary_loss_mlp": 0.01074152,
      "balance_loss_clip": 1.05924761,
      "balance_loss_mlp": 1.04737723,
      "epoch": 0.05128513452577785,
      "flos": 18259822454400.0,
      "grad_norm": 2.1832763559951234,
      "language_loss": 0.80761266,
      "learning_rate": 3.995249886196811e-06,
      "loss": 0.8302899,
      "num_input_tokens_seen": 18226370,
      "step": 853,
      "time_per_iteration": 2.6078240871429443
    },
    {
      "auxiliary_loss_clip": 0.01222394,
      "auxiliary_loss_mlp": 0.01065268,
      "balance_loss_clip": 1.06223083,
      "balance_loss_mlp": 1.03780222,
      "epoch": 0.05134525777844581,
      "flos": 27198957571200.0,
      "grad_norm": 1.8511550328562763,
      "language_loss": 0.75617325,
      "learning_rate": 3.995223022193999e-06,
      "loss": 0.77904987,
      "num_input_tokens_seen": 18247075,
      "step": 854,
      "time_per_iteration": 2.633543014526367
    },
    {
      "auxiliary_loss_clip": 0.01202415,
      "auxiliary_loss_mlp": 0.01065973,
      "balance_loss_clip": 1.06141627,
      "balance_loss_mlp": 1.03828049,
      "epoch": 0.051405381031113785,
      "flos": 28362247436160.0,
      "grad_norm": 2.04057054323539,
      "language_loss": 0.81722355,
      "learning_rate": 3.99519608253213e-06,
      "loss": 0.83990741,
      "num_input_tokens_seen": 18265680,
      "step": 855,
      "time_per_iteration": 2.760880708694458
    },
    {
      "auxiliary_loss_clip": 0.01076712,
      "auxiliary_loss_mlp": 0.00762392,
      "balance_loss_clip": 1.0358243,
      "balance_loss_mlp": 1.00074518,
      "epoch": 0.05146550428378175,
      "flos": 65618169327360.0,
      "grad_norm": 0.9894594919315515,
      "language_loss": 0.65634769,
      "learning_rate": 3.995169067212227e-06,
      "loss": 0.67473871,
      "num_input_tokens_seen": 18327015,
      "step": 856,
      "time_per_iteration": 6.271182298660278
    },
    {
      "auxiliary_loss_clip": 0.01194232,
      "auxiliary_loss_mlp": 0.01056626,
      "balance_loss_clip": 1.05972147,
      "balance_loss_mlp": 1.02994716,
      "epoch": 0.05152562753644972,
      "flos": 22054861998720.0,
      "grad_norm": 1.8001295724347575,
      "language_loss": 0.77139348,
      "learning_rate": 3.9951419762353116e-06,
      "loss": 0.79390204,
      "num_input_tokens_seen": 18345235,
      "step": 857,
      "time_per_iteration": 4.905239582061768
    },
    {
      "auxiliary_loss_clip": 0.01183581,
      "auxiliary_loss_mlp": 0.01059685,
      "balance_loss_clip": 1.05640614,
      "balance_loss_mlp": 1.03291047,
      "epoch": 0.051585750789117694,
      "flos": 18509544783360.0,
      "grad_norm": 2.111656321737554,
      "language_loss": 0.89194518,
      "learning_rate": 3.995114809602412e-06,
      "loss": 0.91437781,
      "num_input_tokens_seen": 18362350,
      "step": 858,
      "time_per_iteration": 2.7349045276641846
    },
    {
      "auxiliary_loss_clip": 0.01196113,
      "auxiliary_loss_mlp": 0.01060739,
      "balance_loss_clip": 1.06114125,
      "balance_loss_mlp": 1.03398848,
      "epoch": 0.05164587404178566,
      "flos": 23730238108800.0,
      "grad_norm": 2.030377637624243,
      "language_loss": 0.75684321,
      "learning_rate": 3.9950875673145605e-06,
      "loss": 0.77941179,
      "num_input_tokens_seen": 18383390,
      "step": 859,
      "time_per_iteration": 2.7611751556396484
    },
    {
      "auxiliary_loss_clip": 0.01186313,
      "auxiliary_loss_mlp": 0.0107269,
      "balance_loss_clip": 1.05708003,
      "balance_loss_mlp": 1.04354358,
      "epoch": 0.05170599729445363,
      "flos": 16252882876800.0,
      "grad_norm": 2.134655488493178,
      "language_loss": 0.91122925,
      "learning_rate": 3.995060249372788e-06,
      "loss": 0.93381929,
      "num_input_tokens_seen": 18399220,
      "step": 860,
      "time_per_iteration": 2.666740894317627
    },
    {
      "auxiliary_loss_clip": 0.0122488,
      "auxiliary_loss_mlp": 0.01060586,
      "balance_loss_clip": 1.06531346,
      "balance_loss_mlp": 1.03536153,
      "epoch": 0.0517661205471216,
      "flos": 23985922095360.0,
      "grad_norm": 1.7954568874114027,
      "language_loss": 0.82378531,
      "learning_rate": 3.99503285577813e-06,
      "loss": 0.84663993,
      "num_input_tokens_seen": 18419005,
      "step": 861,
      "time_per_iteration": 2.6337814331054688
    },
    {
      "auxiliary_loss_clip": 0.01198486,
      "auxiliary_loss_mlp": 0.01060236,
      "balance_loss_clip": 1.06147969,
      "balance_loss_mlp": 1.03437924,
      "epoch": 0.05182624379978957,
      "flos": 29277718392960.0,
      "grad_norm": 2.5785699637959776,
      "language_loss": 0.78664875,
      "learning_rate": 3.995005386531627e-06,
      "loss": 0.80923599,
      "num_input_tokens_seen": 18440550,
      "step": 862,
      "time_per_iteration": 2.7570109367370605
    },
    {
      "auxiliary_loss_clip": 0.01189664,
      "auxiliary_loss_mlp": 0.01070327,
      "balance_loss_clip": 1.058797,
      "balance_loss_mlp": 1.04547238,
      "epoch": 0.05188636705245754,
      "flos": 24170826332160.0,
      "grad_norm": 1.7880881456146414,
      "language_loss": 0.89090264,
      "learning_rate": 3.9949778416343195e-06,
      "loss": 0.91350257,
      "num_input_tokens_seen": 18461950,
      "step": 863,
      "time_per_iteration": 2.7118866443634033
    },
    {
      "auxiliary_loss_clip": 0.01201772,
      "auxiliary_loss_mlp": 0.01064316,
      "balance_loss_clip": 1.06488204,
      "balance_loss_mlp": 1.0369451,
      "epoch": 0.051946490305125506,
      "flos": 26760703731840.0,
      "grad_norm": 2.081656150811602,
      "language_loss": 0.76119763,
      "learning_rate": 3.9949502210872525e-06,
      "loss": 0.78385854,
      "num_input_tokens_seen": 18480555,
      "step": 864,
      "time_per_iteration": 2.6946637630462646
    },
    {
      "auxiliary_loss_clip": 0.01186585,
      "auxiliary_loss_mlp": 0.01067959,
      "balance_loss_clip": 1.05559874,
      "balance_loss_mlp": 1.04046965,
      "epoch": 0.05200661355779348,
      "flos": 21502519585920.0,
      "grad_norm": 1.9374308734697678,
      "language_loss": 0.7908361,
      "learning_rate": 3.994922524891474e-06,
      "loss": 0.81338149,
      "num_input_tokens_seen": 18499645,
      "step": 865,
      "time_per_iteration": 2.7700579166412354
    },
    {
      "auxiliary_loss_clip": 0.01210067,
      "auxiliary_loss_mlp": 0.01067568,
      "balance_loss_clip": 1.06164694,
      "balance_loss_mlp": 1.04152083,
      "epoch": 0.05206673681046144,
      "flos": 18114492026880.0,
      "grad_norm": 2.269489500676155,
      "language_loss": 0.85860598,
      "learning_rate": 3.994894753048032e-06,
      "loss": 0.88138229,
      "num_input_tokens_seen": 18516810,
      "step": 866,
      "time_per_iteration": 2.659614086151123
    },
    {
      "auxiliary_loss_clip": 0.01186536,
      "auxiliary_loss_mlp": 0.01070465,
      "balance_loss_clip": 1.06327558,
      "balance_loss_mlp": 1.04371393,
      "epoch": 0.052126860063129415,
      "flos": 17524191916800.0,
      "grad_norm": 2.1733876112564565,
      "language_loss": 0.87495244,
      "learning_rate": 3.9948669055579815e-06,
      "loss": 0.89752245,
      "num_input_tokens_seen": 18532510,
      "step": 867,
      "time_per_iteration": 2.740238904953003
    },
    {
      "auxiliary_loss_clip": 0.01167585,
      "auxiliary_loss_mlp": 0.01078445,
      "balance_loss_clip": 1.05696058,
      "balance_loss_mlp": 1.05437636,
      "epoch": 0.05218698331579739,
      "flos": 32598054771840.0,
      "grad_norm": 1.8498678854952728,
      "language_loss": 0.63917863,
      "learning_rate": 3.9948389824223785e-06,
      "loss": 0.66163892,
      "num_input_tokens_seen": 18557380,
      "step": 868,
      "time_per_iteration": 2.9310383796691895
    },
    {
      "auxiliary_loss_clip": 0.01225135,
      "auxiliary_loss_mlp": 0.01069894,
      "balance_loss_clip": 1.06287289,
      "balance_loss_mlp": 1.04173636,
      "epoch": 0.05224710656846535,
      "flos": 22127293774080.0,
      "grad_norm": 2.742912036955754,
      "language_loss": 0.83379138,
      "learning_rate": 3.994810983642281e-06,
      "loss": 0.85674161,
      "num_input_tokens_seen": 18575720,
      "step": 869,
      "time_per_iteration": 2.6453137397766113
    },
    {
      "auxiliary_loss_clip": 0.01216406,
      "auxiliary_loss_mlp": 0.01056401,
      "balance_loss_clip": 1.0645746,
      "balance_loss_mlp": 1.03053236,
      "epoch": 0.052307229821133325,
      "flos": 11145092976000.0,
      "grad_norm": 2.188953802542244,
      "language_loss": 0.87822217,
      "learning_rate": 3.994782909218751e-06,
      "loss": 0.90095031,
      "num_input_tokens_seen": 18592185,
      "step": 870,
      "time_per_iteration": 2.7044875621795654
    },
    {
      "auxiliary_loss_clip": 0.01226316,
      "auxiliary_loss_mlp": 0.01064746,
      "balance_loss_clip": 1.06603277,
      "balance_loss_mlp": 1.03965199,
      "epoch": 0.05236735307380129,
      "flos": 19128070005120.0,
      "grad_norm": 1.975067156516721,
      "language_loss": 0.80651748,
      "learning_rate": 3.994754759152854e-06,
      "loss": 0.82942802,
      "num_input_tokens_seen": 18609560,
      "step": 871,
      "time_per_iteration": 2.6892175674438477
    },
    {
      "auxiliary_loss_clip": 0.0119502,
      "auxiliary_loss_mlp": 0.01064309,
      "balance_loss_clip": 1.0650804,
      "balance_loss_mlp": 1.0396452,
      "epoch": 0.05242747632646926,
      "flos": 20960663944320.0,
      "grad_norm": 1.7402390708810018,
      "language_loss": 0.81330585,
      "learning_rate": 3.994726533445656e-06,
      "loss": 0.83589917,
      "num_input_tokens_seen": 18629405,
      "step": 872,
      "time_per_iteration": 2.8044185638427734
    },
    {
      "auxiliary_loss_clip": 0.0107835,
      "auxiliary_loss_mlp": 0.01020667,
      "balance_loss_clip": 1.03168392,
      "balance_loss_mlp": 1.01515913,
      "epoch": 0.052487599579137234,
      "flos": 65020542842880.0,
      "grad_norm": 0.883483589670371,
      "language_loss": 0.61589074,
      "learning_rate": 3.9946982320982274e-06,
      "loss": 0.63688087,
      "num_input_tokens_seen": 18681480,
      "step": 873,
      "time_per_iteration": 3.1711297035217285
    },
    {
      "auxiliary_loss_clip": 0.01197438,
      "auxiliary_loss_mlp": 0.01056818,
      "balance_loss_clip": 1.06202292,
      "balance_loss_mlp": 1.03120041,
      "epoch": 0.0525477228318052,
      "flos": 23288859786240.0,
      "grad_norm": 2.1995328011281488,
      "language_loss": 0.88965189,
      "learning_rate": 3.994669855111643e-06,
      "loss": 0.91219449,
      "num_input_tokens_seen": 18700390,
      "step": 874,
      "time_per_iteration": 2.8240153789520264
    },
    {
      "auxiliary_loss_clip": 0.01197247,
      "auxiliary_loss_mlp": 0.01063458,
      "balance_loss_clip": 1.0614326,
      "balance_loss_mlp": 1.03682709,
      "epoch": 0.05260784608447317,
      "flos": 32230221546240.0,
      "grad_norm": 1.858649685360537,
      "language_loss": 0.74537963,
      "learning_rate": 3.994641402486977e-06,
      "loss": 0.76798666,
      "num_input_tokens_seen": 18721280,
      "step": 875,
      "time_per_iteration": 2.9111931324005127
    },
    {
      "auxiliary_loss_clip": 0.01206205,
      "auxiliary_loss_mlp": 0.01058912,
      "balance_loss_clip": 1.06306934,
      "balance_loss_mlp": 1.03210175,
      "epoch": 0.052667969337141136,
      "flos": 24463211040000.0,
      "grad_norm": 1.7697857141051123,
      "language_loss": 0.92843151,
      "learning_rate": 3.99461287422531e-06,
      "loss": 0.95108265,
      "num_input_tokens_seen": 18741545,
      "step": 876,
      "time_per_iteration": 2.800252676010132
    },
    {
      "auxiliary_loss_clip": 0.01100151,
      "auxiliary_loss_mlp": 0.01006341,
      "balance_loss_clip": 1.02669787,
      "balance_loss_mlp": 1.0020256,
      "epoch": 0.05272809258980911,
      "flos": 57784329567360.0,
      "grad_norm": 0.8383495859932864,
      "language_loss": 0.62929404,
      "learning_rate": 3.994584270327722e-06,
      "loss": 0.65035897,
      "num_input_tokens_seen": 18801400,
      "step": 877,
      "time_per_iteration": 3.2090368270874023
    },
    {
      "auxiliary_loss_clip": 0.01200578,
      "auxiliary_loss_mlp": 0.0106702,
      "balance_loss_clip": 1.06150424,
      "balance_loss_mlp": 1.03931606,
      "epoch": 0.05278821584247708,
      "flos": 17420805596160.0,
      "grad_norm": 2.042786693643985,
      "language_loss": 0.85383844,
      "learning_rate": 3.994555590795299e-06,
      "loss": 0.87651443,
      "num_input_tokens_seen": 18819670,
      "step": 878,
      "time_per_iteration": 2.823835849761963
    },
    {
      "auxiliary_loss_clip": 0.0122514,
      "auxiliary_loss_mlp": 0.01061117,
      "balance_loss_clip": 1.0635035,
      "balance_loss_mlp": 1.03551078,
      "epoch": 0.052848339095145046,
      "flos": 26137258346880.0,
      "grad_norm": 1.7462717669338121,
      "language_loss": 0.83076209,
      "learning_rate": 3.9945268356291275e-06,
      "loss": 0.8536247,
      "num_input_tokens_seen": 18840580,
      "step": 879,
      "time_per_iteration": 2.743673086166382
    },
    {
      "auxiliary_loss_clip": 0.0119139,
      "auxiliary_loss_mlp": 0.01066471,
      "balance_loss_clip": 1.06152987,
      "balance_loss_mlp": 1.04013824,
      "epoch": 0.05290846234781302,
      "flos": 16472081623680.0,
      "grad_norm": 1.9601789563010765,
      "language_loss": 0.84284604,
      "learning_rate": 3.9944980048302985e-06,
      "loss": 0.86542469,
      "num_input_tokens_seen": 18859295,
      "step": 880,
      "time_per_iteration": 2.7560529708862305
    },
    {
      "auxiliary_loss_clip": 0.01184956,
      "auxiliary_loss_mlp": 0.01065063,
      "balance_loss_clip": 1.05969453,
      "balance_loss_mlp": 1.03887296,
      "epoch": 0.05296858560048098,
      "flos": 19865173000320.0,
      "grad_norm": 2.4477328752698564,
      "language_loss": 0.86870736,
      "learning_rate": 3.994469098399906e-06,
      "loss": 0.89120758,
      "num_input_tokens_seen": 18877485,
      "step": 881,
      "time_per_iteration": 2.855395555496216
    },
    {
      "auxiliary_loss_clip": 0.01207858,
      "auxiliary_loss_mlp": 0.01070235,
      "balance_loss_clip": 1.05984437,
      "balance_loss_mlp": 1.04238808,
      "epoch": 0.053028708853148955,
      "flos": 24388588535040.0,
      "grad_norm": 1.7611192020675561,
      "language_loss": 0.87967896,
      "learning_rate": 3.994440116339046e-06,
      "loss": 0.90245986,
      "num_input_tokens_seen": 18898275,
      "step": 882,
      "time_per_iteration": 2.8480119705200195
    },
    {
      "auxiliary_loss_clip": 0.01224906,
      "auxiliary_loss_mlp": 0.01057944,
      "balance_loss_clip": 1.06268644,
      "balance_loss_mlp": 1.03059733,
      "epoch": 0.05308883210581693,
      "flos": 36393166143360.0,
      "grad_norm": 2.3555018967788635,
      "language_loss": 0.69469339,
      "learning_rate": 3.994411058648816e-06,
      "loss": 0.71752191,
      "num_input_tokens_seen": 18920665,
      "step": 883,
      "time_per_iteration": 2.8808236122131348
    },
    {
      "auxiliary_loss_clip": 0.01166777,
      "auxiliary_loss_mlp": 0.01063991,
      "balance_loss_clip": 1.05333591,
      "balance_loss_mlp": 1.03855157,
      "epoch": 0.05314895535848489,
      "flos": 22855095146880.0,
      "grad_norm": 2.039016812023355,
      "language_loss": 0.76100993,
      "learning_rate": 3.994381925330319e-06,
      "loss": 0.78331757,
      "num_input_tokens_seen": 18939835,
      "step": 884,
      "time_per_iteration": 2.8462212085723877
    },
    {
      "auxiliary_loss_clip": 0.01172569,
      "auxiliary_loss_mlp": 0.01066856,
      "balance_loss_clip": 1.06269383,
      "balance_loss_mlp": 1.04147613,
      "epoch": 0.053209078611152864,
      "flos": 12860330204160.0,
      "grad_norm": 1.9865896222141148,
      "language_loss": 0.86195529,
      "learning_rate": 3.994352716384659e-06,
      "loss": 0.88434947,
      "num_input_tokens_seen": 18958405,
      "step": 885,
      "time_per_iteration": 2.7825753688812256
    },
    {
      "auxiliary_loss_clip": 0.0118405,
      "auxiliary_loss_mlp": 0.01068976,
      "balance_loss_clip": 1.05229151,
      "balance_loss_mlp": 1.04203486,
      "epoch": 0.05326920186382083,
      "flos": 12164596698240.0,
      "grad_norm": 2.608647457747672,
      "language_loss": 0.85971159,
      "learning_rate": 3.994323431812945e-06,
      "loss": 0.88224185,
      "num_input_tokens_seen": 18975445,
      "step": 886,
      "time_per_iteration": 2.7393639087677
    },
    {
      "auxiliary_loss_clip": 0.0117343,
      "auxiliary_loss_mlp": 0.01065966,
      "balance_loss_clip": 1.05620933,
      "balance_loss_mlp": 1.03879774,
      "epoch": 0.0533293251164888,
      "flos": 22704485420160.0,
      "grad_norm": 2.040002880698432,
      "language_loss": 0.8961553,
      "learning_rate": 3.994294071616286e-06,
      "loss": 0.91854936,
      "num_input_tokens_seen": 18991930,
      "step": 887,
      "time_per_iteration": 2.8606581687927246
    },
    {
      "auxiliary_loss_clip": 0.01144444,
      "auxiliary_loss_mlp": 0.01072438,
      "balance_loss_clip": 1.04453194,
      "balance_loss_mlp": 1.04411352,
      "epoch": 0.053389448369156774,
      "flos": 26940939200640.0,
      "grad_norm": 2.062562868466936,
      "language_loss": 0.74852538,
      "learning_rate": 3.994264635795796e-06,
      "loss": 0.77069414,
      "num_input_tokens_seen": 19009790,
      "step": 888,
      "time_per_iteration": 2.8675312995910645
    },
    {
      "auxiliary_loss_clip": 0.01164085,
      "auxiliary_loss_mlp": 0.01072324,
      "balance_loss_clip": 1.05659473,
      "balance_loss_mlp": 1.04525173,
      "epoch": 0.05344957162182474,
      "flos": 25556331686400.0,
      "grad_norm": 1.7884280759117637,
      "language_loss": 0.88440782,
      "learning_rate": 3.994235124352592e-06,
      "loss": 0.9067719,
      "num_input_tokens_seen": 19030170,
      "step": 889,
      "time_per_iteration": 2.9419636726379395
    },
    {
      "auxiliary_loss_clip": 0.0121577,
      "auxiliary_loss_mlp": 0.0105125,
      "balance_loss_clip": 1.06085157,
      "balance_loss_mlp": 1.02607334,
      "epoch": 0.05350969487449271,
      "flos": 19719591177600.0,
      "grad_norm": 1.9333059575084248,
      "language_loss": 0.88386381,
      "learning_rate": 3.994205537287791e-06,
      "loss": 0.90653402,
      "num_input_tokens_seen": 19048075,
      "step": 890,
      "time_per_iteration": 2.7030327320098877
    },
    {
      "auxiliary_loss_clip": 0.01195034,
      "auxiliary_loss_mlp": 0.01069003,
      "balance_loss_clip": 1.05835462,
      "balance_loss_mlp": 1.04450595,
      "epoch": 0.053569818127160676,
      "flos": 27016351804800.0,
      "grad_norm": 2.435204176890571,
      "language_loss": 0.93450797,
      "learning_rate": 3.994175874602517e-06,
      "loss": 0.95714831,
      "num_input_tokens_seen": 19067465,
      "step": 891,
      "time_per_iteration": 2.81527042388916
    },
    {
      "auxiliary_loss_clip": 0.01190797,
      "auxiliary_loss_mlp": 0.01066955,
      "balance_loss_clip": 1.05605483,
      "balance_loss_mlp": 1.03909576,
      "epoch": 0.05362994137982865,
      "flos": 13188338225280.0,
      "grad_norm": 2.3400199158693087,
      "language_loss": 0.71625131,
      "learning_rate": 3.994146136297893e-06,
      "loss": 0.73882878,
      "num_input_tokens_seen": 19085505,
      "step": 892,
      "time_per_iteration": 2.825984239578247
    },
    {
      "auxiliary_loss_clip": 0.01191313,
      "auxiliary_loss_mlp": 0.0078394,
      "balance_loss_clip": 1.05727172,
      "balance_loss_mlp": 1.00024366,
      "epoch": 0.05369006463249662,
      "flos": 28658008022400.0,
      "grad_norm": 1.6058100223173828,
      "language_loss": 0.82331586,
      "learning_rate": 3.994116322375049e-06,
      "loss": 0.84306836,
      "num_input_tokens_seen": 19104360,
      "step": 893,
      "time_per_iteration": 2.8618266582489014
    },
    {
      "auxiliary_loss_clip": 0.01192677,
      "auxiliary_loss_mlp": 0.01063531,
      "balance_loss_clip": 1.0572021,
      "balance_loss_mlp": 1.03850877,
      "epoch": 0.053750187885164585,
      "flos": 28913153304960.0,
      "grad_norm": 2.0228714136718122,
      "language_loss": 0.82052565,
      "learning_rate": 3.994086432835114e-06,
      "loss": 0.84308773,
      "num_input_tokens_seen": 19124680,
      "step": 894,
      "time_per_iteration": 2.8347885608673096
    },
    {
      "auxiliary_loss_clip": 0.0120111,
      "auxiliary_loss_mlp": 0.01065233,
      "balance_loss_clip": 1.0570271,
      "balance_loss_mlp": 1.03997254,
      "epoch": 0.05381031113783256,
      "flos": 15158828476800.0,
      "grad_norm": 2.260594705980758,
      "language_loss": 0.76133072,
      "learning_rate": 3.994056467679221e-06,
      "loss": 0.78399414,
      "num_input_tokens_seen": 19142895,
      "step": 895,
      "time_per_iteration": 2.7288858890533447
    },
    {
      "auxiliary_loss_clip": 0.01200143,
      "auxiliary_loss_mlp": 0.01060588,
      "balance_loss_clip": 1.06422663,
      "balance_loss_mlp": 1.03547084,
      "epoch": 0.05387043439050053,
      "flos": 21835232288640.0,
      "grad_norm": 2.0450623179174974,
      "language_loss": 0.86767507,
      "learning_rate": 3.9940264269085065e-06,
      "loss": 0.89028239,
      "num_input_tokens_seen": 19163125,
      "step": 896,
      "time_per_iteration": 4.404265642166138
    },
    {
      "auxiliary_loss_clip": 0.0122203,
      "auxiliary_loss_mlp": 0.00782931,
      "balance_loss_clip": 1.06062579,
      "balance_loss_mlp": 1.0002867,
      "epoch": 0.053930557643168495,
      "flos": 17310308382720.0,
      "grad_norm": 3.0866230440609805,
      "language_loss": 0.8797363,
      "learning_rate": 3.9939963105241115e-06,
      "loss": 0.89978594,
      "num_input_tokens_seen": 19179385,
      "step": 897,
      "time_per_iteration": 4.843130588531494
    },
    {
      "auxiliary_loss_clip": 0.01201639,
      "auxiliary_loss_mlp": 0.01063724,
      "balance_loss_clip": 1.05896854,
      "balance_loss_mlp": 1.03658032,
      "epoch": 0.05399068089583647,
      "flos": 17348481561600.0,
      "grad_norm": 1.8270040910241792,
      "language_loss": 0.90170419,
      "learning_rate": 3.993966118527175e-06,
      "loss": 0.92435783,
      "num_input_tokens_seen": 19198725,
      "step": 898,
      "time_per_iteration": 2.695235252380371
    },
    {
      "auxiliary_loss_clip": 0.01200189,
      "auxiliary_loss_mlp": 0.01076438,
      "balance_loss_clip": 1.05787873,
      "balance_loss_mlp": 1.05105805,
      "epoch": 0.05405080414850443,
      "flos": 17486952491520.0,
      "grad_norm": 2.793625116693953,
      "language_loss": 0.91544139,
      "learning_rate": 3.993935850918845e-06,
      "loss": 0.93820769,
      "num_input_tokens_seen": 19212380,
      "step": 899,
      "time_per_iteration": 2.7509548664093018
    },
    {
      "auxiliary_loss_clip": 0.01186479,
      "auxiliary_loss_mlp": 0.01068594,
      "balance_loss_clip": 1.05614042,
      "balance_loss_mlp": 1.04154527,
      "epoch": 0.054110927401172404,
      "flos": 24496787278080.0,
      "grad_norm": 1.983572968760697,
      "language_loss": 0.75742769,
      "learning_rate": 3.9939055077002665e-06,
      "loss": 0.77997845,
      "num_input_tokens_seen": 19232235,
      "step": 900,
      "time_per_iteration": 2.771371364593506
    },
    {
      "auxiliary_loss_clip": 0.01211506,
      "auxiliary_loss_mlp": 0.01058176,
      "balance_loss_clip": 1.05839145,
      "balance_loss_mlp": 1.03401244,
      "epoch": 0.054171050653840376,
      "flos": 22930040874240.0,
      "grad_norm": 2.192527627735503,
      "language_loss": 0.74331856,
      "learning_rate": 3.993875088872592e-06,
      "loss": 0.76601535,
      "num_input_tokens_seen": 19251460,
      "step": 901,
      "time_per_iteration": 2.859912157058716
    },
    {
      "auxiliary_loss_clip": 0.01177502,
      "auxiliary_loss_mlp": 0.01065445,
      "balance_loss_clip": 1.0569309,
      "balance_loss_mlp": 1.04166329,
      "epoch": 0.05423117390650834,
      "flos": 12933192942720.0,
      "grad_norm": 2.352700712836257,
      "language_loss": 0.85287452,
      "learning_rate": 3.9938445944369745e-06,
      "loss": 0.87530404,
      "num_input_tokens_seen": 19269060,
      "step": 902,
      "time_per_iteration": 2.7940642833709717
    },
    {
      "auxiliary_loss_clip": 0.01161069,
      "auxiliary_loss_mlp": 0.01066664,
      "balance_loss_clip": 1.04903233,
      "balance_loss_mlp": 1.04112983,
      "epoch": 0.05429129715917631,
      "flos": 19901335017600.0,
      "grad_norm": 1.9620711230312637,
      "language_loss": 0.86385572,
      "learning_rate": 3.993814024394569e-06,
      "loss": 0.88613302,
      "num_input_tokens_seen": 19288620,
      "step": 903,
      "time_per_iteration": 2.9258980751037598
    },
    {
      "auxiliary_loss_clip": 0.0121005,
      "auxiliary_loss_mlp": 0.01059616,
      "balance_loss_clip": 1.06094384,
      "balance_loss_mlp": 1.03534508,
      "epoch": 0.05435142041184428,
      "flos": 16908611610240.0,
      "grad_norm": 2.175127974944855,
      "language_loss": 0.74927866,
      "learning_rate": 3.993783378746537e-06,
      "loss": 0.7719754,
      "num_input_tokens_seen": 19306615,
      "step": 904,
      "time_per_iteration": 2.7239954471588135
    },
    {
      "auxiliary_loss_clip": 0.01208402,
      "auxiliary_loss_mlp": 0.01067543,
      "balance_loss_clip": 1.06052148,
      "balance_loss_mlp": 1.04325962,
      "epoch": 0.05441154366451225,
      "flos": 23948323534080.0,
      "grad_norm": 2.5191963984804535,
      "language_loss": 0.85946918,
      "learning_rate": 3.993752657494039e-06,
      "loss": 0.88222867,
      "num_input_tokens_seen": 19321680,
      "step": 905,
      "time_per_iteration": 2.693896532058716
    },
    {
      "auxiliary_loss_clip": 0.01198232,
      "auxiliary_loss_mlp": 0.01078072,
      "balance_loss_clip": 1.06483209,
      "balance_loss_mlp": 1.05400348,
      "epoch": 0.05447166691718022,
      "flos": 19975382904960.0,
      "grad_norm": 1.7753581401878566,
      "language_loss": 0.74413162,
      "learning_rate": 3.993721860638241e-06,
      "loss": 0.7668947,
      "num_input_tokens_seen": 19339760,
      "step": 906,
      "time_per_iteration": 2.6679019927978516
    },
    {
      "auxiliary_loss_clip": 0.01192373,
      "auxiliary_loss_mlp": 0.01064381,
      "balance_loss_clip": 1.05954027,
      "balance_loss_mlp": 1.0397284,
      "epoch": 0.05453179016984819,
      "flos": 24936513575040.0,
      "grad_norm": 2.3037248114268896,
      "language_loss": 0.87340188,
      "learning_rate": 3.993690988180309e-06,
      "loss": 0.89596951,
      "num_input_tokens_seen": 19359585,
      "step": 907,
      "time_per_iteration": 2.7363240718841553
    },
    {
      "auxiliary_loss_clip": 0.01205519,
      "auxiliary_loss_mlp": 0.01068463,
      "balance_loss_clip": 1.0616293,
      "balance_loss_mlp": 1.04332149,
      "epoch": 0.05459191342251616,
      "flos": 18115102558080.0,
      "grad_norm": 1.6666873589767146,
      "language_loss": 0.86928803,
      "learning_rate": 3.9936600401214165e-06,
      "loss": 0.89202785,
      "num_input_tokens_seen": 19378590,
      "step": 908,
      "time_per_iteration": 2.6266026496887207
    },
    {
      "auxiliary_loss_clip": 0.01198848,
      "auxiliary_loss_mlp": 0.01067336,
      "balance_loss_clip": 1.05974221,
      "balance_loss_mlp": 1.04107404,
      "epoch": 0.054652036675184125,
      "flos": 19208295031680.0,
      "grad_norm": 2.1282794409977215,
      "language_loss": 0.89792144,
      "learning_rate": 3.9936290164627345e-06,
      "loss": 0.92058325,
      "num_input_tokens_seen": 19397910,
      "step": 909,
      "time_per_iteration": 2.7163166999816895
    },
    {
      "auxiliary_loss_clip": 0.01200393,
      "auxiliary_loss_mlp": 0.01073374,
      "balance_loss_clip": 1.06157839,
      "balance_loss_mlp": 1.04742169,
      "epoch": 0.0547121599278521,
      "flos": 16325745615360.0,
      "grad_norm": 2.095924869989121,
      "language_loss": 0.70949811,
      "learning_rate": 3.99359791720544e-06,
      "loss": 0.73223579,
      "num_input_tokens_seen": 19415950,
      "step": 910,
      "time_per_iteration": 2.6697354316711426
    },
    {
      "auxiliary_loss_clip": 0.01187784,
      "auxiliary_loss_mlp": 0.01054671,
      "balance_loss_clip": 1.05651259,
      "balance_loss_mlp": 1.02975583,
      "epoch": 0.05477228318052007,
      "flos": 20339014239360.0,
      "grad_norm": 1.6633724338567386,
      "language_loss": 0.83651805,
      "learning_rate": 3.993566742350714e-06,
      "loss": 0.85894263,
      "num_input_tokens_seen": 19435275,
      "step": 911,
      "time_per_iteration": 2.692798137664795
    },
    {
      "auxiliary_loss_clip": 0.01187113,
      "auxiliary_loss_mlp": 0.01073028,
      "balance_loss_clip": 1.05334687,
      "balance_loss_mlp": 1.04719508,
      "epoch": 0.054832406433188034,
      "flos": 21973092687360.0,
      "grad_norm": 2.283907419545301,
      "language_loss": 0.76320881,
      "learning_rate": 3.993535491899736e-06,
      "loss": 0.78581023,
      "num_input_tokens_seen": 19452090,
      "step": 912,
      "time_per_iteration": 2.6653189659118652
    },
    {
      "auxiliary_loss_clip": 0.01186313,
      "auxiliary_loss_mlp": 0.01051652,
      "balance_loss_clip": 1.05707574,
      "balance_loss_mlp": 1.0271548,
      "epoch": 0.054892529685856006,
      "flos": 16398931576320.0,
      "grad_norm": 2.366460016615147,
      "language_loss": 0.82826668,
      "learning_rate": 3.993504165853694e-06,
      "loss": 0.85064626,
      "num_input_tokens_seen": 19470865,
      "step": 913,
      "time_per_iteration": 2.6826348304748535
    },
    {
      "auxiliary_loss_clip": 0.01194515,
      "auxiliary_loss_mlp": 0.01060483,
      "balance_loss_clip": 1.0581125,
      "balance_loss_mlp": 1.03651023,
      "epoch": 0.05495265293852397,
      "flos": 23912341084800.0,
      "grad_norm": 3.3338391252510586,
      "language_loss": 0.8373239,
      "learning_rate": 3.993472764213772e-06,
      "loss": 0.85987389,
      "num_input_tokens_seen": 19492145,
      "step": 914,
      "time_per_iteration": 2.7358829975128174
    },
    {
      "auxiliary_loss_clip": 0.0120705,
      "auxiliary_loss_mlp": 0.0078227,
      "balance_loss_clip": 1.06039774,
      "balance_loss_mlp": 1.00027478,
      "epoch": 0.055012776191191944,
      "flos": 23586954756480.0,
      "grad_norm": 2.520244909384168,
      "language_loss": 0.90146536,
      "learning_rate": 3.9934412869811655e-06,
      "loss": 0.92135859,
      "num_input_tokens_seen": 19511015,
      "step": 915,
      "time_per_iteration": 2.9398341178894043
    },
    {
      "auxiliary_loss_clip": 0.01201461,
      "auxiliary_loss_mlp": 0.01059252,
      "balance_loss_clip": 1.06274199,
      "balance_loss_mlp": 1.03558862,
      "epoch": 0.055072899443859916,
      "flos": 17528501548800.0,
      "grad_norm": 2.182721785653499,
      "language_loss": 0.89710975,
      "learning_rate": 3.993409734157064e-06,
      "loss": 0.91971689,
      "num_input_tokens_seen": 19529040,
      "step": 916,
      "time_per_iteration": 2.7210159301757812
    },
    {
      "auxiliary_loss_clip": 0.01175226,
      "auxiliary_loss_mlp": 0.01066073,
      "balance_loss_clip": 1.05741024,
      "balance_loss_mlp": 1.04103947,
      "epoch": 0.05513302269652788,
      "flos": 21687172427520.0,
      "grad_norm": 1.7899379897310368,
      "language_loss": 0.8016991,
      "learning_rate": 3.993378105742666e-06,
      "loss": 0.82411212,
      "num_input_tokens_seen": 19549540,
      "step": 917,
      "time_per_iteration": 2.7923104763031006
    },
    {
      "auxiliary_loss_clip": 0.01139072,
      "auxiliary_loss_mlp": 0.0105947,
      "balance_loss_clip": 1.05135942,
      "balance_loss_mlp": 1.03414989,
      "epoch": 0.05519314594919585,
      "flos": 21613340021760.0,
      "grad_norm": 2.106744179667805,
      "language_loss": 0.79437333,
      "learning_rate": 3.9933464017391705e-06,
      "loss": 0.81635869,
      "num_input_tokens_seen": 19567570,
      "step": 918,
      "time_per_iteration": 2.8051092624664307
    },
    {
      "auxiliary_loss_clip": 0.01196947,
      "auxiliary_loss_mlp": 0.01055679,
      "balance_loss_clip": 1.05616307,
      "balance_loss_mlp": 1.03166997,
      "epoch": 0.05525326920186382,
      "flos": 21798567480960.0,
      "grad_norm": 2.454030193031321,
      "language_loss": 0.89019686,
      "learning_rate": 3.99331462214778e-06,
      "loss": 0.91272312,
      "num_input_tokens_seen": 19585330,
      "step": 919,
      "time_per_iteration": 2.6846773624420166
    },
    {
      "auxiliary_loss_clip": 0.01213326,
      "auxiliary_loss_mlp": 0.01069349,
      "balance_loss_clip": 1.05950904,
      "balance_loss_mlp": 1.04417229,
      "epoch": 0.05531339245453179,
      "flos": 28439635288320.0,
      "grad_norm": 2.246354931091656,
      "language_loss": 0.8746047,
      "learning_rate": 3.993282766969699e-06,
      "loss": 0.89743137,
      "num_input_tokens_seen": 19604970,
      "step": 920,
      "time_per_iteration": 2.6699845790863037
    },
    {
      "auxiliary_loss_clip": 0.01190424,
      "auxiliary_loss_mlp": 0.0106036,
      "balance_loss_clip": 1.06023288,
      "balance_loss_mlp": 1.03657782,
      "epoch": 0.05537351570719976,
      "flos": 37375143131520.0,
      "grad_norm": 1.975714125194334,
      "language_loss": 0.6568011,
      "learning_rate": 3.993250836206136e-06,
      "loss": 0.67930895,
      "num_input_tokens_seen": 19626235,
      "step": 921,
      "time_per_iteration": 2.833644390106201
    },
    {
      "auxiliary_loss_clip": 0.01209678,
      "auxiliary_loss_mlp": 0.01065483,
      "balance_loss_clip": 1.06060767,
      "balance_loss_mlp": 1.03874445,
      "epoch": 0.05543363895986773,
      "flos": 20084479488000.0,
      "grad_norm": 1.7242493696651606,
      "language_loss": 0.71861136,
      "learning_rate": 3.993218829858301e-06,
      "loss": 0.74136293,
      "num_input_tokens_seen": 19644305,
      "step": 922,
      "time_per_iteration": 2.6168808937072754
    },
    {
      "auxiliary_loss_clip": 0.01187138,
      "auxiliary_loss_mlp": 0.01067213,
      "balance_loss_clip": 1.05423355,
      "balance_loss_mlp": 1.04223895,
      "epoch": 0.0554937622125357,
      "flos": 24533200690560.0,
      "grad_norm": 2.6848185900705412,
      "language_loss": 0.82304025,
      "learning_rate": 3.993186747927408e-06,
      "loss": 0.8455838,
      "num_input_tokens_seen": 19662130,
      "step": 923,
      "time_per_iteration": 2.7298316955566406
    },
    {
      "auxiliary_loss_clip": 0.01202941,
      "auxiliary_loss_mlp": 0.01064106,
      "balance_loss_clip": 1.05725455,
      "balance_loss_mlp": 1.03933442,
      "epoch": 0.055553885465203665,
      "flos": 14320063013760.0,
      "grad_norm": 1.9334372940525173,
      "language_loss": 0.78759122,
      "learning_rate": 3.993154590414675e-06,
      "loss": 0.81026167,
      "num_input_tokens_seen": 19680715,
      "step": 924,
      "time_per_iteration": 2.6869630813598633
    },
    {
      "auxiliary_loss_clip": 0.0116422,
      "auxiliary_loss_mlp": 0.01053758,
      "balance_loss_clip": 1.05395627,
      "balance_loss_mlp": 1.02844954,
      "epoch": 0.05561400871787164,
      "flos": 27381132374400.0,
      "grad_norm": 2.005203138116014,
      "language_loss": 1.02005315,
      "learning_rate": 3.993122357321319e-06,
      "loss": 1.04223299,
      "num_input_tokens_seen": 19700535,
      "step": 925,
      "time_per_iteration": 2.716089963912964
    },
    {
      "auxiliary_loss_clip": 0.01163201,
      "auxiliary_loss_mlp": 0.01052104,
      "balance_loss_clip": 1.05070591,
      "balance_loss_mlp": 1.02739179,
      "epoch": 0.05567413197053961,
      "flos": 23221096778880.0,
      "grad_norm": 2.0106641835017482,
      "language_loss": 0.80939209,
      "learning_rate": 3.993090048648564e-06,
      "loss": 0.83154511,
      "num_input_tokens_seen": 19718825,
      "step": 926,
      "time_per_iteration": 2.895803451538086
    },
    {
      "auxiliary_loss_clip": 0.01207515,
      "auxiliary_loss_mlp": 0.01068168,
      "balance_loss_clip": 1.05892682,
      "balance_loss_mlp": 1.0419066,
      "epoch": 0.055734255223207574,
      "flos": 25264952559360.0,
      "grad_norm": 2.9732625845644045,
      "language_loss": 0.73220479,
      "learning_rate": 3.993057664397634e-06,
      "loss": 0.75496161,
      "num_input_tokens_seen": 19739080,
      "step": 927,
      "time_per_iteration": 2.677725076675415
    },
    {
      "auxiliary_loss_clip": 0.01101002,
      "auxiliary_loss_mlp": 0.01015011,
      "balance_loss_clip": 1.02922702,
      "balance_loss_mlp": 1.01014709,
      "epoch": 0.055794378475875546,
      "flos": 66503116702080.0,
      "grad_norm": 0.8406874373244947,
      "language_loss": 0.59841412,
      "learning_rate": 3.9930252045697585e-06,
      "loss": 0.61957431,
      "num_input_tokens_seen": 19802960,
      "step": 928,
      "time_per_iteration": 3.187382221221924
    },
    {
      "auxiliary_loss_clip": 0.01202438,
      "auxiliary_loss_mlp": 0.01065066,
      "balance_loss_clip": 1.05921853,
      "balance_loss_mlp": 1.04070008,
      "epoch": 0.05585450172854351,
      "flos": 25337635729920.0,
      "grad_norm": 2.0668361967965994,
      "language_loss": 0.95411372,
      "learning_rate": 3.992992669166168e-06,
      "loss": 0.97678876,
      "num_input_tokens_seen": 19822765,
      "step": 929,
      "time_per_iteration": 2.6930506229400635
    },
    {
      "auxiliary_loss_clip": 0.01171806,
      "auxiliary_loss_mlp": 0.01068051,
      "balance_loss_clip": 1.05343258,
      "balance_loss_mlp": 1.04101443,
      "epoch": 0.05591462498121148,
      "flos": 33911738881920.0,
      "grad_norm": 2.1442452677256627,
      "language_loss": 0.71756601,
      "learning_rate": 3.992960058188094e-06,
      "loss": 0.7399646,
      "num_input_tokens_seen": 19843590,
      "step": 930,
      "time_per_iteration": 2.803219795227051
    },
    {
      "auxiliary_loss_clip": 0.01188277,
      "auxiliary_loss_mlp": 0.01058888,
      "balance_loss_clip": 1.05783677,
      "balance_loss_mlp": 1.03377056,
      "epoch": 0.055974748233879455,
      "flos": 17930880679680.0,
      "grad_norm": 2.381261552273062,
      "language_loss": 0.85279298,
      "learning_rate": 3.992927371636776e-06,
      "loss": 0.87526459,
      "num_input_tokens_seen": 19860230,
      "step": 931,
      "time_per_iteration": 2.6215872764587402
    },
    {
      "auxiliary_loss_clip": 0.01203533,
      "auxiliary_loss_mlp": 0.00783076,
      "balance_loss_clip": 1.05677414,
      "balance_loss_mlp": 1.00025761,
      "epoch": 0.05603487148654742,
      "flos": 24021976371840.0,
      "grad_norm": 2.2861197477099973,
      "language_loss": 0.83645165,
      "learning_rate": 3.9928946095134525e-06,
      "loss": 0.85631776,
      "num_input_tokens_seen": 19880795,
      "step": 932,
      "time_per_iteration": 2.664062261581421
    },
    {
      "auxiliary_loss_clip": 0.01200637,
      "auxiliary_loss_mlp": 0.0107041,
      "balance_loss_clip": 1.05897784,
      "balance_loss_mlp": 1.04407716,
      "epoch": 0.05609499473921539,
      "flos": 17307758517120.0,
      "grad_norm": 1.8036739452122519,
      "language_loss": 0.73694205,
      "learning_rate": 3.992861771819365e-06,
      "loss": 0.7596525,
      "num_input_tokens_seen": 19897960,
      "step": 933,
      "time_per_iteration": 2.631620168685913
    },
    {
      "auxiliary_loss_clip": 0.01153445,
      "auxiliary_loss_mlp": 0.01076903,
      "balance_loss_clip": 1.04885209,
      "balance_loss_mlp": 1.05060577,
      "epoch": 0.05615511799188336,
      "flos": 20994742972800.0,
      "grad_norm": 2.385249039382274,
      "language_loss": 0.86660421,
      "learning_rate": 3.99282885855576e-06,
      "loss": 0.88890779,
      "num_input_tokens_seen": 19913315,
      "step": 934,
      "time_per_iteration": 2.7739439010620117
    },
    {
      "auxiliary_loss_clip": 0.01164295,
      "auxiliary_loss_mlp": 0.0108083,
      "balance_loss_clip": 1.05509257,
      "balance_loss_mlp": 1.0557723,
      "epoch": 0.05621524124455133,
      "flos": 17273535834240.0,
      "grad_norm": 2.2740258482680433,
      "language_loss": 0.80388415,
      "learning_rate": 3.992795869723885e-06,
      "loss": 0.82633543,
      "num_input_tokens_seen": 19928790,
      "step": 935,
      "time_per_iteration": 5.93512487411499
    },
    {
      "auxiliary_loss_clip": 0.01093927,
      "auxiliary_loss_mlp": 0.01019701,
      "balance_loss_clip": 1.02288604,
      "balance_loss_mlp": 1.01540911,
      "epoch": 0.0562753644972193,
      "flos": 58719370458240.0,
      "grad_norm": 0.820561718243334,
      "language_loss": 0.69191676,
      "learning_rate": 3.99276280532499e-06,
      "loss": 0.71305299,
      "num_input_tokens_seen": 19988785,
      "step": 936,
      "time_per_iteration": 4.862478733062744
    },
    {
      "auxiliary_loss_clip": 0.01213648,
      "auxiliary_loss_mlp": 0.01068507,
      "balance_loss_clip": 1.05806684,
      "balance_loss_mlp": 1.04429567,
      "epoch": 0.05633548774988727,
      "flos": 17457039440640.0,
      "grad_norm": 1.9573264311231433,
      "language_loss": 0.7572521,
      "learning_rate": 3.992729665360331e-06,
      "loss": 0.78007358,
      "num_input_tokens_seen": 20007685,
      "step": 937,
      "time_per_iteration": 4.219425916671753
    },
    {
      "auxiliary_loss_clip": 0.01085529,
      "auxiliary_loss_mlp": 0.01013805,
      "balance_loss_clip": 1.02476001,
      "balance_loss_mlp": 1.00944233,
      "epoch": 0.05639561100255524,
      "flos": 70654928083200.0,
      "grad_norm": 0.9053055994078011,
      "language_loss": 0.64309287,
      "learning_rate": 3.992696449831162e-06,
      "loss": 0.66408622,
      "num_input_tokens_seen": 20072750,
      "step": 938,
      "time_per_iteration": 3.1298794746398926
    },
    {
      "auxiliary_loss_clip": 0.01171203,
      "auxiliary_loss_mlp": 0.01068815,
      "balance_loss_clip": 1.05175185,
      "balance_loss_mlp": 1.0426966,
      "epoch": 0.056455734255223204,
      "flos": 20485996692480.0,
      "grad_norm": 2.7427540631348832,
      "language_loss": 0.79751205,
      "learning_rate": 3.992663158738745e-06,
      "loss": 0.8199122,
      "num_input_tokens_seen": 20089070,
      "step": 939,
      "time_per_iteration": 2.6863484382629395
    },
    {
      "auxiliary_loss_clip": 0.01175528,
      "auxiliary_loss_mlp": 0.01068297,
      "balance_loss_clip": 1.0509069,
      "balance_loss_mlp": 1.04338217,
      "epoch": 0.056515857507891176,
      "flos": 22053569109120.0,
      "grad_norm": 1.8374791395473227,
      "language_loss": 0.73919088,
      "learning_rate": 3.992629792084341e-06,
      "loss": 0.76162916,
      "num_input_tokens_seen": 20108790,
      "step": 940,
      "time_per_iteration": 2.7111120223999023
    },
    {
      "auxiliary_loss_clip": 0.01198483,
      "auxiliary_loss_mlp": 0.01058511,
      "balance_loss_clip": 1.05900669,
      "balance_loss_mlp": 1.03252339,
      "epoch": 0.05657598076055915,
      "flos": 24025316336640.0,
      "grad_norm": 2.2993716569389813,
      "language_loss": 0.70622003,
      "learning_rate": 3.992596349869216e-06,
      "loss": 0.72878999,
      "num_input_tokens_seen": 20128455,
      "step": 941,
      "time_per_iteration": 2.657594680786133
    },
    {
      "auxiliary_loss_clip": 0.01135396,
      "auxiliary_loss_mlp": 0.01059543,
      "balance_loss_clip": 1.04961574,
      "balance_loss_mlp": 1.03382993,
      "epoch": 0.05663610401322711,
      "flos": 20480609652480.0,
      "grad_norm": 2.0678542992190847,
      "language_loss": 0.80921417,
      "learning_rate": 3.992562832094637e-06,
      "loss": 0.83116359,
      "num_input_tokens_seen": 20145775,
      "step": 942,
      "time_per_iteration": 2.7379891872406006
    },
    {
      "auxiliary_loss_clip": 0.01186767,
      "auxiliary_loss_mlp": 0.01062055,
      "balance_loss_clip": 1.05228579,
      "balance_loss_mlp": 1.03554332,
      "epoch": 0.056696227265895086,
      "flos": 21069042255360.0,
      "grad_norm": 2.245249922529115,
      "language_loss": 0.88858449,
      "learning_rate": 3.9925292387618755e-06,
      "loss": 0.91107273,
      "num_input_tokens_seen": 20164315,
      "step": 943,
      "time_per_iteration": 2.6502583026885986
    },
    {
      "auxiliary_loss_clip": 0.01199122,
      "auxiliary_loss_mlp": 0.0105963,
      "balance_loss_clip": 1.05991781,
      "balance_loss_mlp": 1.03534663,
      "epoch": 0.05675635051856306,
      "flos": 17821317219840.0,
      "grad_norm": 2.5514256959015995,
      "language_loss": 0.74771839,
      "learning_rate": 3.992495569872206e-06,
      "loss": 0.77030593,
      "num_input_tokens_seen": 20182760,
      "step": 944,
      "time_per_iteration": 2.676079034805298
    },
    {
      "auxiliary_loss_clip": 0.01204502,
      "auxiliary_loss_mlp": 0.01064591,
      "balance_loss_clip": 1.05980551,
      "balance_loss_mlp": 1.04085672,
      "epoch": 0.05681647377123102,
      "flos": 23114945111040.0,
      "grad_norm": 1.5959266123312272,
      "language_loss": 0.79406166,
      "learning_rate": 3.992461825426906e-06,
      "loss": 0.81675267,
      "num_input_tokens_seen": 20203830,
      "step": 945,
      "time_per_iteration": 2.734299421310425
    },
    {
      "auxiliary_loss_clip": 0.01195984,
      "auxiliary_loss_mlp": 0.0105672,
      "balance_loss_clip": 1.05686593,
      "balance_loss_mlp": 1.03156662,
      "epoch": 0.056876597023898995,
      "flos": 16070528505600.0,
      "grad_norm": 2.5637081249861824,
      "language_loss": 0.82651746,
      "learning_rate": 3.992428005427252e-06,
      "loss": 0.84904456,
      "num_input_tokens_seen": 20220365,
      "step": 946,
      "time_per_iteration": 2.6636929512023926
    },
    {
      "auxiliary_loss_clip": 0.0122014,
      "auxiliary_loss_mlp": 0.01061449,
      "balance_loss_clip": 1.06224144,
      "balance_loss_mlp": 1.03524721,
      "epoch": 0.05693672027656696,
      "flos": 16835641130880.0,
      "grad_norm": 1.8433174156507384,
      "language_loss": 0.79031301,
      "learning_rate": 3.992394109874529e-06,
      "loss": 0.81312895,
      "num_input_tokens_seen": 20238640,
      "step": 947,
      "time_per_iteration": 2.623671293258667
    },
    {
      "auxiliary_loss_clip": 0.0117587,
      "auxiliary_loss_mlp": 0.01061489,
      "balance_loss_clip": 1.05605412,
      "balance_loss_mlp": 1.03569245,
      "epoch": 0.05699684352923493,
      "flos": 21389113370880.0,
      "grad_norm": 6.8661947111986725,
      "language_loss": 0.85425055,
      "learning_rate": 3.9923601387700225e-06,
      "loss": 0.87662411,
      "num_input_tokens_seen": 20251025,
      "step": 948,
      "time_per_iteration": 2.7410409450531006
    },
    {
      "auxiliary_loss_clip": 0.01214005,
      "auxiliary_loss_mlp": 0.01063231,
      "balance_loss_clip": 1.05969238,
      "balance_loss_mlp": 1.03598022,
      "epoch": 0.057056966781902904,
      "flos": 15560309767680.0,
      "grad_norm": 3.649211317819821,
      "language_loss": 0.87346625,
      "learning_rate": 3.992326092115019e-06,
      "loss": 0.89623863,
      "num_input_tokens_seen": 20269775,
      "step": 949,
      "time_per_iteration": 2.6893157958984375
    },
    {
      "auxiliary_loss_clip": 0.01194543,
      "auxiliary_loss_mlp": 0.0106695,
      "balance_loss_clip": 1.05799937,
      "balance_loss_mlp": 1.04266715,
      "epoch": 0.05711709003457087,
      "flos": 19937856170880.0,
      "grad_norm": 1.8324883776363103,
      "language_loss": 0.7874645,
      "learning_rate": 3.992291969910811e-06,
      "loss": 0.8100794,
      "num_input_tokens_seen": 20287715,
      "step": 950,
      "time_per_iteration": 2.623924732208252
    },
    {
      "auxiliary_loss_clip": 0.01180518,
      "auxiliary_loss_mlp": 0.01068771,
      "balance_loss_clip": 1.05322623,
      "balance_loss_mlp": 1.04384422,
      "epoch": 0.05717721328723884,
      "flos": 30332701774080.0,
      "grad_norm": 3.8045132244795816,
      "language_loss": 0.82477522,
      "learning_rate": 3.992257772158691e-06,
      "loss": 0.8472681,
      "num_input_tokens_seen": 20307070,
      "step": 951,
      "time_per_iteration": 2.697479724884033
    },
    {
      "auxiliary_loss_clip": 0.01167302,
      "auxiliary_loss_mlp": 0.01061039,
      "balance_loss_clip": 1.04906607,
      "balance_loss_mlp": 1.03375173,
      "epoch": 0.05723733653990681,
      "flos": 23654358627840.0,
      "grad_norm": 2.4180383362968634,
      "language_loss": 0.86899263,
      "learning_rate": 3.992223498859958e-06,
      "loss": 0.89127606,
      "num_input_tokens_seen": 20324945,
      "step": 952,
      "time_per_iteration": 2.707716226577759
    },
    {
      "auxiliary_loss_clip": 0.01191405,
      "auxiliary_loss_mlp": 0.01064705,
      "balance_loss_clip": 1.05511189,
      "balance_loss_mlp": 1.03630924,
      "epoch": 0.05729745979257478,
      "flos": 22055759838720.0,
      "grad_norm": 2.195434645270168,
      "language_loss": 0.79087842,
      "learning_rate": 3.9921891500159084e-06,
      "loss": 0.81343949,
      "num_input_tokens_seen": 20346135,
      "step": 953,
      "time_per_iteration": 2.671255588531494
    },
    {
      "auxiliary_loss_clip": 0.01190026,
      "auxiliary_loss_mlp": 0.01066447,
      "balance_loss_clip": 1.05984342,
      "balance_loss_mlp": 1.04056656,
      "epoch": 0.05735758304524275,
      "flos": 19604353368960.0,
      "grad_norm": 2.2066085695914466,
      "language_loss": 0.86644447,
      "learning_rate": 3.992154725627848e-06,
      "loss": 0.88900924,
      "num_input_tokens_seen": 20364450,
      "step": 954,
      "time_per_iteration": 2.671657085418701
    },
    {
      "auxiliary_loss_clip": 0.01210569,
      "auxiliary_loss_mlp": 0.01062619,
      "balance_loss_clip": 1.06119955,
      "balance_loss_mlp": 1.03723955,
      "epoch": 0.057417706297910716,
      "flos": 19099018880640.0,
      "grad_norm": 2.2872795023766113,
      "language_loss": 0.88071024,
      "learning_rate": 3.9921202256970804e-06,
      "loss": 0.90344214,
      "num_input_tokens_seen": 20383500,
      "step": 955,
      "time_per_iteration": 2.69960880279541
    },
    {
      "auxiliary_loss_clip": 0.01179864,
      "auxiliary_loss_mlp": 0.01068889,
      "balance_loss_clip": 1.0523231,
      "balance_loss_mlp": 1.04209054,
      "epoch": 0.05747782955057869,
      "flos": 16654507822080.0,
      "grad_norm": 1.9113555723128555,
      "language_loss": 0.89160776,
      "learning_rate": 3.992085650224914e-06,
      "loss": 0.91409534,
      "num_input_tokens_seen": 20400295,
      "step": 956,
      "time_per_iteration": 2.667868137359619
    },
    {
      "auxiliary_loss_clip": 0.01167867,
      "auxiliary_loss_mlp": 0.01060669,
      "balance_loss_clip": 1.05720079,
      "balance_loss_mlp": 1.03450251,
      "epoch": 0.05753795280324665,
      "flos": 14502058248960.0,
      "grad_norm": 3.2877973901728095,
      "language_loss": 0.75473189,
      "learning_rate": 3.99205099921266e-06,
      "loss": 0.77701724,
      "num_input_tokens_seen": 20419085,
      "step": 957,
      "time_per_iteration": 2.6938796043395996
    },
    {
      "auxiliary_loss_clip": 0.0117627,
      "auxiliary_loss_mlp": 0.01072849,
      "balance_loss_clip": 1.05432248,
      "balance_loss_mlp": 1.0448705,
      "epoch": 0.057598076055914625,
      "flos": 18076318848000.0,
      "grad_norm": 2.0004055711005257,
      "language_loss": 0.79582155,
      "learning_rate": 3.992016272661633e-06,
      "loss": 0.81831264,
      "num_input_tokens_seen": 20437465,
      "step": 958,
      "time_per_iteration": 2.6933834552764893
    },
    {
      "auxiliary_loss_clip": 0.01186244,
      "auxiliary_loss_mlp": 0.01059908,
      "balance_loss_clip": 1.05851364,
      "balance_loss_mlp": 1.03572011,
      "epoch": 0.0576581993085826,
      "flos": 22124600254080.0,
      "grad_norm": 2.669863855173802,
      "language_loss": 0.8840394,
      "learning_rate": 3.99198147057315e-06,
      "loss": 0.906501,
      "num_input_tokens_seen": 20456235,
      "step": 959,
      "time_per_iteration": 2.7094578742980957
    },
    {
      "auxiliary_loss_clip": 0.01169479,
      "auxiliary_loss_mlp": 0.01063656,
      "balance_loss_clip": 1.05511999,
      "balance_loss_mlp": 1.03881276,
      "epoch": 0.05771832256125056,
      "flos": 33181746779520.0,
      "grad_norm": 2.0960373333994764,
      "language_loss": 0.78850955,
      "learning_rate": 3.991946592948529e-06,
      "loss": 0.8108409,
      "num_input_tokens_seen": 20476825,
      "step": 960,
      "time_per_iteration": 2.822922945022583
    },
    {
      "auxiliary_loss_clip": 0.0113413,
      "auxiliary_loss_mlp": 0.01067189,
      "balance_loss_clip": 1.05177355,
      "balance_loss_mlp": 1.04020023,
      "epoch": 0.057778445813918534,
      "flos": 24170143973760.0,
      "grad_norm": 2.063464892179025,
      "language_loss": 0.92986894,
      "learning_rate": 3.991911639789094e-06,
      "loss": 0.95188212,
      "num_input_tokens_seen": 20496965,
      "step": 961,
      "time_per_iteration": 2.793952226638794
    },
    {
      "auxiliary_loss_clip": 0.01182535,
      "auxiliary_loss_mlp": 0.0106764,
      "balance_loss_clip": 1.0554297,
      "balance_loss_mlp": 1.04091299,
      "epoch": 0.0578385690665865,
      "flos": 29643037666560.0,
      "grad_norm": 2.0649993155313067,
      "language_loss": 0.68164188,
      "learning_rate": 3.991876611096169e-06,
      "loss": 0.70414358,
      "num_input_tokens_seen": 20518035,
      "step": 962,
      "time_per_iteration": 2.8396694660186768
    },
    {
      "auxiliary_loss_clip": 0.01159524,
      "auxiliary_loss_mlp": 0.01073851,
      "balance_loss_clip": 1.05128908,
      "balance_loss_mlp": 1.04909074,
      "epoch": 0.05789869231925447,
      "flos": 20885430908160.0,
      "grad_norm": 2.2685465488517074,
      "language_loss": 0.8848027,
      "learning_rate": 3.991841506871084e-06,
      "loss": 0.90713644,
      "num_input_tokens_seen": 20534740,
      "step": 963,
      "time_per_iteration": 2.7077019214630127
    },
    {
      "auxiliary_loss_clip": 0.01183778,
      "auxiliary_loss_mlp": 0.01061251,
      "balance_loss_clip": 1.06018209,
      "balance_loss_mlp": 1.03516829,
      "epoch": 0.057958815571922444,
      "flos": 26031106679040.0,
      "grad_norm": 2.392959969035536,
      "language_loss": 0.85288298,
      "learning_rate": 3.99180632711517e-06,
      "loss": 0.87533331,
      "num_input_tokens_seen": 20553485,
      "step": 964,
      "time_per_iteration": 2.7218217849731445
    },
    {
      "auxiliary_loss_clip": 0.01188683,
      "auxiliary_loss_mlp": 0.01069422,
      "balance_loss_clip": 1.05959499,
      "balance_loss_mlp": 1.04325557,
      "epoch": 0.05801893882459041,
      "flos": 18077683564800.0,
      "grad_norm": 3.087349735715565,
      "language_loss": 0.78159416,
      "learning_rate": 3.99177107182976e-06,
      "loss": 0.80417526,
      "num_input_tokens_seen": 20572155,
      "step": 965,
      "time_per_iteration": 2.6902661323547363
    },
    {
      "auxiliary_loss_clip": 0.01156531,
      "auxiliary_loss_mlp": 0.0107109,
      "balance_loss_clip": 1.0523715,
      "balance_loss_mlp": 1.04462528,
      "epoch": 0.05807906207725838,
      "flos": 17748885444480.0,
      "grad_norm": 1.9742288518319486,
      "language_loss": 0.81403655,
      "learning_rate": 3.99173574101619e-06,
      "loss": 0.83631277,
      "num_input_tokens_seen": 20590395,
      "step": 966,
      "time_per_iteration": 2.7423267364501953
    },
    {
      "auxiliary_loss_clip": 0.01198908,
      "auxiliary_loss_mlp": 0.01065021,
      "balance_loss_clip": 1.058887,
      "balance_loss_mlp": 1.04113197,
      "epoch": 0.058139185329926346,
      "flos": 18040372312320.0,
      "grad_norm": 1.8776530142118544,
      "language_loss": 0.76480806,
      "learning_rate": 3.9917003346758035e-06,
      "loss": 0.78744727,
      "num_input_tokens_seen": 20608435,
      "step": 967,
      "time_per_iteration": 2.642885446548462
    },
    {
      "auxiliary_loss_clip": 0.01084339,
      "auxiliary_loss_mlp": 0.0103139,
      "balance_loss_clip": 1.02675521,
      "balance_loss_mlp": 1.0269078,
      "epoch": 0.05819930858259432,
      "flos": 62363297485440.0,
      "grad_norm": 0.985564929959949,
      "language_loss": 0.57357776,
      "learning_rate": 3.991664852809939e-06,
      "loss": 0.59473509,
      "num_input_tokens_seen": 20668575,
      "step": 968,
      "time_per_iteration": 3.1017024517059326
    },
    {
      "auxiliary_loss_clip": 0.01188824,
      "auxiliary_loss_mlp": 0.01057715,
      "balance_loss_clip": 1.05784404,
      "balance_loss_mlp": 1.03147697,
      "epoch": 0.05825943183526229,
      "flos": 19135360465920.0,
      "grad_norm": 2.1276337565108485,
      "language_loss": 0.82286429,
      "learning_rate": 3.991629295419945e-06,
      "loss": 0.84532964,
      "num_input_tokens_seen": 20687355,
      "step": 969,
      "time_per_iteration": 2.669055461883545
    },
    {
      "auxiliary_loss_clip": 0.01206272,
      "auxiliary_loss_mlp": 0.00782724,
      "balance_loss_clip": 1.06255269,
      "balance_loss_mlp": 1.00024962,
      "epoch": 0.058319555087930255,
      "flos": 29022465369600.0,
      "grad_norm": 7.916507288074279,
      "language_loss": 0.7803669,
      "learning_rate": 3.991593662507167e-06,
      "loss": 0.80025685,
      "num_input_tokens_seen": 20705710,
      "step": 970,
      "time_per_iteration": 2.733030080795288
    },
    {
      "auxiliary_loss_clip": 0.01181452,
      "auxiliary_loss_mlp": 0.01064945,
      "balance_loss_clip": 1.05691695,
      "balance_loss_mlp": 1.03887415,
      "epoch": 0.05837967834059823,
      "flos": 18879999701760.0,
      "grad_norm": 3.163102883752813,
      "language_loss": 0.92229038,
      "learning_rate": 3.991557954072958e-06,
      "loss": 0.94475436,
      "num_input_tokens_seen": 20722405,
      "step": 971,
      "time_per_iteration": 2.730377435684204
    },
    {
      "auxiliary_loss_clip": 0.01180948,
      "auxiliary_loss_mlp": 0.01062613,
      "balance_loss_clip": 1.05320477,
      "balance_loss_mlp": 1.03722143,
      "epoch": 0.05843980159326619,
      "flos": 25703062744320.0,
      "grad_norm": 1.700187330091603,
      "language_loss": 0.85959208,
      "learning_rate": 3.991522170118673e-06,
      "loss": 0.88202775,
      "num_input_tokens_seen": 20741480,
      "step": 972,
      "time_per_iteration": 2.687185049057007
    },
    {
      "auxiliary_loss_clip": 0.0116993,
      "auxiliary_loss_mlp": 0.01079713,
      "balance_loss_clip": 1.05714142,
      "balance_loss_mlp": 1.05601454,
      "epoch": 0.058499924845934165,
      "flos": 25552129795200.0,
      "grad_norm": 2.00599255988541,
      "language_loss": 0.87503272,
      "learning_rate": 3.991486310645667e-06,
      "loss": 0.89752913,
      "num_input_tokens_seen": 20759685,
      "step": 973,
      "time_per_iteration": 2.7166664600372314
    },
    {
      "auxiliary_loss_clip": 0.01206524,
      "auxiliary_loss_mlp": 0.00784111,
      "balance_loss_clip": 1.06111121,
      "balance_loss_mlp": 1.00026989,
      "epoch": 0.05856004809860214,
      "flos": 16436171001600.0,
      "grad_norm": 1.879365930358842,
      "language_loss": 0.74800295,
      "learning_rate": 3.991450375655301e-06,
      "loss": 0.76790935,
      "num_input_tokens_seen": 20778180,
      "step": 974,
      "time_per_iteration": 2.713594675064087
    },
    {
      "auxiliary_loss_clip": 0.01197101,
      "auxiliary_loss_mlp": 0.00782207,
      "balance_loss_clip": 1.059551,
      "balance_loss_mlp": 1.00025892,
      "epoch": 0.0586201713512701,
      "flos": 39458824116480.0,
      "grad_norm": 1.5923993506380014,
      "language_loss": 0.76874506,
      "learning_rate": 3.991414365148936e-06,
      "loss": 0.78853816,
      "num_input_tokens_seen": 20802705,
      "step": 975,
      "time_per_iteration": 7.600914716720581
    },
    {
      "auxiliary_loss_clip": 0.01215491,
      "auxiliary_loss_mlp": 0.01069506,
      "balance_loss_clip": 1.06030774,
      "balance_loss_mlp": 1.0444721,
      "epoch": 0.058680294603938074,
      "flos": 23365170230400.0,
      "grad_norm": 3.6132976830219734,
      "language_loss": 0.76748288,
      "learning_rate": 3.99137827912794e-06,
      "loss": 0.79033279,
      "num_input_tokens_seen": 20822540,
      "step": 976,
      "time_per_iteration": 4.324799537658691
    },
    {
      "auxiliary_loss_clip": 0.01176132,
      "auxiliary_loss_mlp": 0.01077003,
      "balance_loss_clip": 1.05271626,
      "balance_loss_mlp": 1.04963279,
      "epoch": 0.05874041785660604,
      "flos": 32232017226240.0,
      "grad_norm": 1.943198757110789,
      "language_loss": 0.87343585,
      "learning_rate": 3.991342117593679e-06,
      "loss": 0.89596725,
      "num_input_tokens_seen": 20844175,
      "step": 977,
      "time_per_iteration": 2.7742488384246826
    },
    {
      "auxiliary_loss_clip": 0.01187161,
      "auxiliary_loss_mlp": 0.01067914,
      "balance_loss_clip": 1.06209528,
      "balance_loss_mlp": 1.04231977,
      "epoch": 0.05880054110927401,
      "flos": 22310043194880.0,
      "grad_norm": 1.718987046197629,
      "language_loss": 0.7969116,
      "learning_rate": 3.991305880547527e-06,
      "loss": 0.81946236,
      "num_input_tokens_seen": 20864730,
      "step": 978,
      "time_per_iteration": 2.733372926712036
    },
    {
      "auxiliary_loss_clip": 0.01136264,
      "auxiliary_loss_mlp": 0.01076585,
      "balance_loss_clip": 1.05591321,
      "balance_loss_mlp": 1.04927468,
      "epoch": 0.05886066436194198,
      "flos": 27380450016000.0,
      "grad_norm": 1.8692877257975375,
      "language_loss": 0.80665666,
      "learning_rate": 3.991269567990855e-06,
      "loss": 0.82878518,
      "num_input_tokens_seen": 20885200,
      "step": 979,
      "time_per_iteration": 3.2624220848083496
    },
    {
      "auxiliary_loss_clip": 0.01074686,
      "auxiliary_loss_mlp": 0.01029701,
      "balance_loss_clip": 1.02640033,
      "balance_loss_mlp": 1.02495658,
      "epoch": 0.05892078761460995,
      "flos": 59584493525760.0,
      "grad_norm": 0.9436493040005753,
      "language_loss": 0.59004962,
      "learning_rate": 3.9912331799250415e-06,
      "loss": 0.6110934,
      "num_input_tokens_seen": 20940325,
      "step": 980,
      "time_per_iteration": 3.4688587188720703
    },
    {
      "auxiliary_loss_clip": 0.01211665,
      "auxiliary_loss_mlp": 0.01078603,
      "balance_loss_clip": 1.06178868,
      "balance_loss_mlp": 1.05242431,
      "epoch": 0.05898091086727792,
      "flos": 15414081500160.0,
      "grad_norm": 2.2770545408130514,
      "language_loss": 0.86436182,
      "learning_rate": 3.9911967163514665e-06,
      "loss": 0.88726455,
      "num_input_tokens_seen": 20958220,
      "step": 981,
      "time_per_iteration": 2.5824644565582275
    },
    {
      "auxiliary_loss_clip": 0.01190085,
      "auxiliary_loss_mlp": 0.0106921,
      "balance_loss_clip": 1.05943286,
      "balance_loss_mlp": 1.04629803,
      "epoch": 0.059041034119945886,
      "flos": 23655328295040.0,
      "grad_norm": 2.1333982175691855,
      "language_loss": 0.79293346,
      "learning_rate": 3.991160177271513e-06,
      "loss": 0.81552643,
      "num_input_tokens_seen": 20978920,
      "step": 982,
      "time_per_iteration": 2.68428897857666
    },
    {
      "auxiliary_loss_clip": 0.01192274,
      "auxiliary_loss_mlp": 0.01068234,
      "balance_loss_clip": 1.05926657,
      "balance_loss_mlp": 1.04356933,
      "epoch": 0.05910115737261386,
      "flos": 24754087376640.0,
      "grad_norm": 2.319627739094249,
      "language_loss": 0.84413779,
      "learning_rate": 3.9911235626865654e-06,
      "loss": 0.86674285,
      "num_input_tokens_seen": 20999490,
      "step": 983,
      "time_per_iteration": 2.7006261348724365
    },
    {
      "auxiliary_loss_clip": 0.0120015,
      "auxiliary_loss_mlp": 0.01072669,
      "balance_loss_clip": 1.05969584,
      "balance_loss_mlp": 1.04799283,
      "epoch": 0.05916128062528183,
      "flos": 11728749070080.0,
      "grad_norm": 1.8014395118859294,
      "language_loss": 0.84510243,
      "learning_rate": 3.9910868725980125e-06,
      "loss": 0.86783063,
      "num_input_tokens_seen": 21017865,
      "step": 984,
      "time_per_iteration": 2.640246868133545
    },
    {
      "auxiliary_loss_clip": 0.01188594,
      "auxiliary_loss_mlp": 0.01055296,
      "balance_loss_clip": 1.05650342,
      "balance_loss_mlp": 1.03171611,
      "epoch": 0.059221403877949795,
      "flos": 21902995296000.0,
      "grad_norm": 2.473231587287368,
      "language_loss": 0.77611595,
      "learning_rate": 3.9910501070072465e-06,
      "loss": 0.7985549,
      "num_input_tokens_seen": 21035900,
      "step": 985,
      "time_per_iteration": 2.626371383666992
    },
    {
      "auxiliary_loss_clip": 0.01150113,
      "auxiliary_loss_mlp": 0.01060814,
      "balance_loss_clip": 1.05341148,
      "balance_loss_mlp": 1.03542209,
      "epoch": 0.05928152713061777,
      "flos": 20514580940160.0,
      "grad_norm": 1.9082382068459252,
      "language_loss": 0.90593231,
      "learning_rate": 3.991013265915661e-06,
      "loss": 0.92804158,
      "num_input_tokens_seen": 21053235,
      "step": 986,
      "time_per_iteration": 2.7834935188293457
    },
    {
      "auxiliary_loss_clip": 0.01200704,
      "auxiliary_loss_mlp": 0.01061312,
      "balance_loss_clip": 1.05555892,
      "balance_loss_mlp": 1.03425193,
      "epoch": 0.05934165038328574,
      "flos": 24495135252480.0,
      "grad_norm": 2.216017383423336,
      "language_loss": 0.75688565,
      "learning_rate": 3.9909763493246525e-06,
      "loss": 0.77950585,
      "num_input_tokens_seen": 21073090,
      "step": 987,
      "time_per_iteration": 2.6669981479644775
    },
    {
      "auxiliary_loss_clip": 0.01203558,
      "auxiliary_loss_mlp": 0.01057756,
      "balance_loss_clip": 1.06134868,
      "balance_loss_mlp": 1.03331852,
      "epoch": 0.059401773635953704,
      "flos": 38728041914880.0,
      "grad_norm": 2.2869993581633827,
      "language_loss": 0.71867943,
      "learning_rate": 3.990939357235621e-06,
      "loss": 0.7412926,
      "num_input_tokens_seen": 21094895,
      "step": 988,
      "time_per_iteration": 2.805851697921753
    },
    {
      "auxiliary_loss_clip": 0.0105006,
      "auxiliary_loss_mlp": 0.0101134,
      "balance_loss_clip": 1.02230322,
      "balance_loss_mlp": 1.00688171,
      "epoch": 0.059461896888621676,
      "flos": 58023565125120.0,
      "grad_norm": 0.9416454944601763,
      "language_loss": 0.7124939,
      "learning_rate": 3.99090228964997e-06,
      "loss": 0.73310792,
      "num_input_tokens_seen": 21147555,
      "step": 989,
      "time_per_iteration": 3.100306749343872
    },
    {
      "auxiliary_loss_clip": 0.0117797,
      "auxiliary_loss_mlp": 0.01072264,
      "balance_loss_clip": 1.05793095,
      "balance_loss_mlp": 1.04389191,
      "epoch": 0.05952202014128964,
      "flos": 22127760650880.0,
      "grad_norm": 2.0167260155417113,
      "language_loss": 0.78245646,
      "learning_rate": 3.990865146569105e-06,
      "loss": 0.80495882,
      "num_input_tokens_seen": 21167845,
      "step": 990,
      "time_per_iteration": 2.8133904933929443
    },
    {
      "auxiliary_loss_clip": 0.01198295,
      "auxiliary_loss_mlp": 0.01053485,
      "balance_loss_clip": 1.06166339,
      "balance_loss_mlp": 1.02761686,
      "epoch": 0.059582143393957614,
      "flos": 20445776438400.0,
      "grad_norm": 2.2411623387553727,
      "language_loss": 0.86522102,
      "learning_rate": 3.990827927994434e-06,
      "loss": 0.88773882,
      "num_input_tokens_seen": 21185085,
      "step": 991,
      "time_per_iteration": 2.6964831352233887
    },
    {
      "auxiliary_loss_clip": 0.0121783,
      "auxiliary_loss_mlp": 0.01064707,
      "balance_loss_clip": 1.0613625,
      "balance_loss_mlp": 1.03943431,
      "epoch": 0.059642266646625586,
      "flos": 20594877793920.0,
      "grad_norm": 1.8566945591898132,
      "language_loss": 0.76738375,
      "learning_rate": 3.9907906339273674e-06,
      "loss": 0.79020917,
      "num_input_tokens_seen": 21204230,
      "step": 992,
      "time_per_iteration": 2.646942377090454
    },
    {
      "auxiliary_loss_clip": 0.01146457,
      "auxiliary_loss_mlp": 0.01062309,
      "balance_loss_clip": 1.05571234,
      "balance_loss_mlp": 1.03834832,
      "epoch": 0.05970238989929355,
      "flos": 19352655792000.0,
      "grad_norm": 2.3469050968731233,
      "language_loss": 0.75117075,
      "learning_rate": 3.9907532643693215e-06,
      "loss": 0.77325845,
      "num_input_tokens_seen": 21222655,
      "step": 993,
      "time_per_iteration": 2.7642974853515625
    },
    {
      "auxiliary_loss_clip": 0.01157785,
      "auxiliary_loss_mlp": 0.01075532,
      "balance_loss_clip": 1.05397618,
      "balance_loss_mlp": 1.04774487,
      "epoch": 0.05976251315196152,
      "flos": 30264040926720.0,
      "grad_norm": 2.725207959052886,
      "language_loss": 0.79177904,
      "learning_rate": 3.990715819321712e-06,
      "loss": 0.81411219,
      "num_input_tokens_seen": 21242310,
      "step": 994,
      "time_per_iteration": 2.8414714336395264
    },
    {
      "auxiliary_loss_clip": 0.01214724,
      "auxiliary_loss_mlp": 0.01079016,
      "balance_loss_clip": 1.06264019,
      "balance_loss_mlp": 1.05361295,
      "epoch": 0.05982263640462949,
      "flos": 23185150243200.0,
      "grad_norm": 2.8097993094234983,
      "language_loss": 0.79917169,
      "learning_rate": 3.99067829878596e-06,
      "loss": 0.82210916,
      "num_input_tokens_seen": 21261410,
      "step": 995,
      "time_per_iteration": 2.6524364948272705
    },
    {
      "auxiliary_loss_clip": 0.0116696,
      "auxiliary_loss_mlp": 0.01068218,
      "balance_loss_clip": 1.05704355,
      "balance_loss_mlp": 1.04208767,
      "epoch": 0.05988275965729746,
      "flos": 27850879463040.0,
      "grad_norm": 1.902030256537741,
      "language_loss": 0.87013257,
      "learning_rate": 3.990640702763487e-06,
      "loss": 0.89248431,
      "num_input_tokens_seen": 21280080,
      "step": 996,
      "time_per_iteration": 2.7431676387786865
    },
    {
      "auxiliary_loss_clip": 0.01177854,
      "auxiliary_loss_mlp": 0.01081123,
      "balance_loss_clip": 1.05684328,
      "balance_loss_mlp": 1.05055761,
      "epoch": 0.05994288290996543,
      "flos": 24680003575680.0,
      "grad_norm": 2.971039758986745,
      "language_loss": 0.87273014,
      "learning_rate": 3.990603031255718e-06,
      "loss": 0.89531994,
      "num_input_tokens_seen": 21296765,
      "step": 997,
      "time_per_iteration": 2.748448371887207
    },
    {
      "auxiliary_loss_clip": 0.01069915,
      "auxiliary_loss_mlp": 0.01014417,
      "balance_loss_clip": 1.02303648,
      "balance_loss_mlp": 1.00972033,
      "epoch": 0.0600030061626334,
      "flos": 69929568835200.0,
      "grad_norm": 1.0091092068179202,
      "language_loss": 0.75381488,
      "learning_rate": 3.990565284264083e-06,
      "loss": 0.7746582,
      "num_input_tokens_seen": 21363345,
      "step": 998,
      "time_per_iteration": 3.2950518131256104
    },
    {
      "auxiliary_loss_clip": 0.01170062,
      "auxiliary_loss_mlp": 0.01065521,
      "balance_loss_clip": 1.05893683,
      "balance_loss_mlp": 1.03893745,
      "epoch": 0.06006312941530137,
      "flos": 26540140268160.0,
      "grad_norm": 1.8197691299520968,
      "language_loss": 0.76053095,
      "learning_rate": 3.990527461790013e-06,
      "loss": 0.7828868,
      "num_input_tokens_seen": 21385290,
      "step": 999,
      "time_per_iteration": 2.733802556991577
    },
    {
      "auxiliary_loss_clip": 0.01197834,
      "auxiliary_loss_mlp": 0.01059542,
      "balance_loss_clip": 1.05646563,
      "balance_loss_mlp": 1.03339899,
      "epoch": 0.060123252667969335,
      "flos": 27344000689920.0,
      "grad_norm": 2.5948629341774874,
      "language_loss": 0.82992184,
      "learning_rate": 3.990489563834943e-06,
      "loss": 0.85249555,
      "num_input_tokens_seen": 21407625,
      "step": 1000,
      "time_per_iteration": 2.710981845855713
    },
    {
      "auxiliary_loss_clip": 0.0118571,
      "auxiliary_loss_mlp": 0.01062188,
      "balance_loss_clip": 1.05856955,
      "balance_loss_mlp": 1.03480577,
      "epoch": 0.06018337592063731,
      "flos": 27016710940800.0,
      "grad_norm": 2.111409807940472,
      "language_loss": 0.85820085,
      "learning_rate": 3.990451590400309e-06,
      "loss": 0.88067985,
      "num_input_tokens_seen": 21426835,
      "step": 1001,
      "time_per_iteration": 2.73445463180542
    },
    {
      "auxiliary_loss_clip": 0.01191917,
      "auxiliary_loss_mlp": 0.01062059,
      "balance_loss_clip": 1.06167853,
      "balance_loss_mlp": 1.03719211,
      "epoch": 0.06024349917330528,
      "flos": 25592960580480.0,
      "grad_norm": 1.8359711451165206,
      "language_loss": 0.74128318,
      "learning_rate": 3.990413541487551e-06,
      "loss": 0.76382297,
      "num_input_tokens_seen": 21444920,
      "step": 1002,
      "time_per_iteration": 2.8861100673675537
    },
    {
      "auxiliary_loss_clip": 0.01214316,
      "auxiliary_loss_mlp": 0.01062589,
      "balance_loss_clip": 1.06316125,
      "balance_loss_mlp": 1.03737664,
      "epoch": 0.060303622425973244,
      "flos": 26133271937280.0,
      "grad_norm": 2.1835040648243997,
      "language_loss": 0.75520515,
      "learning_rate": 3.990375417098112e-06,
      "loss": 0.77797419,
      "num_input_tokens_seen": 21463555,
      "step": 1003,
      "time_per_iteration": 2.632889747619629
    },
    {
      "auxiliary_loss_clip": 0.01187709,
      "auxiliary_loss_mlp": 0.01064806,
      "balance_loss_clip": 1.05934548,
      "balance_loss_mlp": 1.03928304,
      "epoch": 0.060363745678641216,
      "flos": 20377187418240.0,
      "grad_norm": 2.3150099602993155,
      "language_loss": 0.70349169,
      "learning_rate": 3.990337217233437e-06,
      "loss": 0.72601682,
      "num_input_tokens_seen": 21481990,
      "step": 1004,
      "time_per_iteration": 2.6947617530822754
    },
    {
      "auxiliary_loss_clip": 0.01212815,
      "auxiliary_loss_mlp": 0.01077454,
      "balance_loss_clip": 1.06629324,
      "balance_loss_mlp": 1.05168116,
      "epoch": 0.06042386893130918,
      "flos": 17749172753280.0,
      "grad_norm": 2.276868338025253,
      "language_loss": 0.83444524,
      "learning_rate": 3.990298941894976e-06,
      "loss": 0.85734791,
      "num_input_tokens_seen": 21500385,
      "step": 1005,
      "time_per_iteration": 2.581683397293091
    },
    {
      "auxiliary_loss_clip": 0.01077621,
      "auxiliary_loss_mlp": 0.01004707,
      "balance_loss_clip": 1.02541244,
      "balance_loss_mlp": 1.00029612,
      "epoch": 0.06048399218397715,
      "flos": 68538496872960.0,
      "grad_norm": 0.903813421793838,
      "language_loss": 0.59018111,
      "learning_rate": 3.9902605910841794e-06,
      "loss": 0.61100447,
      "num_input_tokens_seen": 21561040,
      "step": 1006,
      "time_per_iteration": 3.222104787826538
    },
    {
      "auxiliary_loss_clip": 0.01183553,
      "auxiliary_loss_mlp": 0.01059038,
      "balance_loss_clip": 1.05334234,
      "balance_loss_mlp": 1.03284812,
      "epoch": 0.060544115436645125,
      "flos": 23258515772160.0,
      "grad_norm": 2.1584333764290853,
      "language_loss": 0.74229443,
      "learning_rate": 3.990222164802503e-06,
      "loss": 0.76472032,
      "num_input_tokens_seen": 21580655,
      "step": 1007,
      "time_per_iteration": 2.7130653858184814
    },
    {
      "auxiliary_loss_clip": 0.0119408,
      "auxiliary_loss_mlp": 0.01060431,
      "balance_loss_clip": 1.06008601,
      "balance_loss_mlp": 1.03493261,
      "epoch": 0.06060423868931309,
      "flos": 23878441624320.0,
      "grad_norm": 1.7956876298455304,
      "language_loss": 0.8081426,
      "learning_rate": 3.9901836630514006e-06,
      "loss": 0.8306877,
      "num_input_tokens_seen": 21599650,
      "step": 1008,
      "time_per_iteration": 2.7151994705200195
    },
    {
      "auxiliary_loss_clip": 0.01175291,
      "auxiliary_loss_mlp": 0.01056357,
      "balance_loss_clip": 1.05982351,
      "balance_loss_mlp": 1.0305717,
      "epoch": 0.06066436194198106,
      "flos": 18728061171840.0,
      "grad_norm": 2.3069524306559837,
      "language_loss": 0.78198558,
      "learning_rate": 3.990145085832335e-06,
      "loss": 0.8043021,
      "num_input_tokens_seen": 21617550,
      "step": 1009,
      "time_per_iteration": 2.7313599586486816
    },
    {
      "auxiliary_loss_clip": 0.01194621,
      "auxiliary_loss_mlp": 0.01061233,
      "balance_loss_clip": 1.06150866,
      "balance_loss_mlp": 1.03726041,
      "epoch": 0.06072448519464903,
      "flos": 24640465680000.0,
      "grad_norm": 1.7452257697216769,
      "language_loss": 0.93148172,
      "learning_rate": 3.990106433146769e-06,
      "loss": 0.95404023,
      "num_input_tokens_seen": 21635865,
      "step": 1010,
      "time_per_iteration": 2.7233662605285645
    },
    {
      "auxiliary_loss_clip": 0.01148246,
      "auxiliary_loss_mlp": 0.00784144,
      "balance_loss_clip": 1.05304599,
      "balance_loss_mlp": 1.00029802,
      "epoch": 0.060784608447317,
      "flos": 17378825575680.0,
      "grad_norm": 2.9999367504779517,
      "language_loss": 0.72022474,
      "learning_rate": 3.9900677049961665e-06,
      "loss": 0.73954868,
      "num_input_tokens_seen": 21653945,
      "step": 1011,
      "time_per_iteration": 2.804858446121216
    },
    {
      "auxiliary_loss_clip": 0.01194231,
      "auxiliary_loss_mlp": 0.01077344,
      "balance_loss_clip": 1.05968046,
      "balance_loss_mlp": 1.04868615,
      "epoch": 0.06084473169998497,
      "flos": 23692208584320.0,
      "grad_norm": 1.9573218215833301,
      "language_loss": 0.87526691,
      "learning_rate": 3.990028901381999e-06,
      "loss": 0.89798272,
      "num_input_tokens_seen": 21671230,
      "step": 1012,
      "time_per_iteration": 2.6466245651245117
    },
    {
      "auxiliary_loss_clip": 0.01184459,
      "auxiliary_loss_mlp": 0.01064264,
      "balance_loss_clip": 1.05652905,
      "balance_loss_mlp": 1.03838325,
      "epoch": 0.06090485495265294,
      "flos": 23546339452800.0,
      "grad_norm": 1.9062230938156723,
      "language_loss": 0.76947677,
      "learning_rate": 3.989990022305734e-06,
      "loss": 0.79196405,
      "num_input_tokens_seen": 21691155,
      "step": 1013,
      "time_per_iteration": 4.297588586807251
    },
    {
      "auxiliary_loss_clip": 0.01207383,
      "auxiliary_loss_mlp": 0.00783488,
      "balance_loss_clip": 1.06573224,
      "balance_loss_mlp": 1.00034499,
      "epoch": 0.06096497820532091,
      "flos": 20339301548160.0,
      "grad_norm": 2.441711811862119,
      "language_loss": 0.86151874,
      "learning_rate": 3.98995106776885e-06,
      "loss": 0.88142747,
      "num_input_tokens_seen": 21707405,
      "step": 1014,
      "time_per_iteration": 4.301488637924194
    },
    {
      "auxiliary_loss_clip": 0.0121503,
      "auxiliary_loss_mlp": 0.01072817,
      "balance_loss_clip": 1.06605387,
      "balance_loss_mlp": 1.04508948,
      "epoch": 0.061025101457988874,
      "flos": 26939035779840.0,
      "grad_norm": 2.4309754772209184,
      "language_loss": 0.73197287,
      "learning_rate": 3.98991203777282e-06,
      "loss": 0.75485134,
      "num_input_tokens_seen": 21728090,
      "step": 1015,
      "time_per_iteration": 4.384514808654785
    },
    {
      "auxiliary_loss_clip": 0.01187374,
      "auxiliary_loss_mlp": 0.01068593,
      "balance_loss_clip": 1.06084347,
      "balance_loss_mlp": 1.04228365,
      "epoch": 0.061085224710656846,
      "flos": 25375054723200.0,
      "grad_norm": 1.5896529502124837,
      "language_loss": 0.79109907,
      "learning_rate": 3.9898729323191275e-06,
      "loss": 0.81365877,
      "num_input_tokens_seen": 21747950,
      "step": 1016,
      "time_per_iteration": 4.3249351978302
    },
    {
      "auxiliary_loss_clip": 0.01173015,
      "auxiliary_loss_mlp": 0.0105746,
      "balance_loss_clip": 1.06036103,
      "balance_loss_mlp": 1.03249741,
      "epoch": 0.06114534796332482,
      "flos": 24824759385600.0,
      "grad_norm": 1.6772682648410928,
      "language_loss": 0.76014191,
      "learning_rate": 3.989833751409254e-06,
      "loss": 0.78244662,
      "num_input_tokens_seen": 21767900,
      "step": 1017,
      "time_per_iteration": 2.7983243465423584
    },
    {
      "auxiliary_loss_clip": 0.01188817,
      "auxiliary_loss_mlp": 0.01074603,
      "balance_loss_clip": 1.06584609,
      "balance_loss_mlp": 1.0483532,
      "epoch": 0.061205471215992784,
      "flos": 20631434860800.0,
      "grad_norm": 2.001716657382839,
      "language_loss": 0.85798436,
      "learning_rate": 3.989794495044685e-06,
      "loss": 0.88061857,
      "num_input_tokens_seen": 21787375,
      "step": 1018,
      "time_per_iteration": 2.702399253845215
    },
    {
      "auxiliary_loss_clip": 0.01174344,
      "auxiliary_loss_mlp": 0.01069438,
      "balance_loss_clip": 1.06325769,
      "balance_loss_mlp": 1.04231787,
      "epoch": 0.061265594468660756,
      "flos": 16508351381760.0,
      "grad_norm": 2.9546929267460813,
      "language_loss": 0.76985347,
      "learning_rate": 3.989755163226909e-06,
      "loss": 0.79229128,
      "num_input_tokens_seen": 21806275,
      "step": 1019,
      "time_per_iteration": 2.780104875564575
    },
    {
      "auxiliary_loss_clip": 0.01160861,
      "auxiliary_loss_mlp": 0.0106141,
      "balance_loss_clip": 1.05355084,
      "balance_loss_mlp": 1.03511262,
      "epoch": 0.06132571772132872,
      "flos": 26246211275520.0,
      "grad_norm": 2.1848809329980288,
      "language_loss": 0.84122044,
      "learning_rate": 3.989715755957418e-06,
      "loss": 0.86344314,
      "num_input_tokens_seen": 21826430,
      "step": 1020,
      "time_per_iteration": 2.785963535308838
    },
    {
      "auxiliary_loss_clip": 0.01198473,
      "auxiliary_loss_mlp": 0.01063342,
      "balance_loss_clip": 1.06365371,
      "balance_loss_mlp": 1.03604269,
      "epoch": 0.06138584097399669,
      "flos": 37414788768000.0,
      "grad_norm": 1.933053672026977,
      "language_loss": 0.79114467,
      "learning_rate": 3.989676273237705e-06,
      "loss": 0.81376278,
      "num_input_tokens_seen": 21847800,
      "step": 1021,
      "time_per_iteration": 2.7968955039978027
    },
    {
      "auxiliary_loss_clip": 0.01189659,
      "auxiliary_loss_mlp": 0.01064044,
      "balance_loss_clip": 1.06159925,
      "balance_loss_mlp": 1.04114437,
      "epoch": 0.061445964226664665,
      "flos": 17420661941760.0,
      "grad_norm": 2.089525934673828,
      "language_loss": 0.87768298,
      "learning_rate": 3.9896367150692705e-06,
      "loss": 0.90022004,
      "num_input_tokens_seen": 21863385,
      "step": 1022,
      "time_per_iteration": 2.70906138420105
    },
    {
      "auxiliary_loss_clip": 0.01198737,
      "auxiliary_loss_mlp": 0.0106635,
      "balance_loss_clip": 1.06627858,
      "balance_loss_mlp": 1.04079151,
      "epoch": 0.06150608747933263,
      "flos": 22600021691520.0,
      "grad_norm": 1.7284486983379121,
      "language_loss": 0.82892007,
      "learning_rate": 3.989597081453611e-06,
      "loss": 0.85157096,
      "num_input_tokens_seen": 21881880,
      "step": 1023,
      "time_per_iteration": 2.71539568901062
    },
    {
      "auxiliary_loss_clip": 0.01100664,
      "auxiliary_loss_mlp": 0.01010751,
      "balance_loss_clip": 1.03727341,
      "balance_loss_mlp": 1.00614953,
      "epoch": 0.0615662107320006,
      "flos": 56741482005120.0,
      "grad_norm": 0.8894752517384502,
      "language_loss": 0.6505782,
      "learning_rate": 3.989557372392231e-06,
      "loss": 0.67169237,
      "num_input_tokens_seen": 21940550,
      "step": 1024,
      "time_per_iteration": 3.175217628479004
    },
    {
      "auxiliary_loss_clip": 0.01167458,
      "auxiliary_loss_mlp": 0.01073669,
      "balance_loss_clip": 1.05906856,
      "balance_loss_mlp": 1.04553604,
      "epoch": 0.06162633398466857,
      "flos": 22564793427840.0,
      "grad_norm": 2.320347485789288,
      "language_loss": 0.88069236,
      "learning_rate": 3.989517587886636e-06,
      "loss": 0.90310359,
      "num_input_tokens_seen": 21958390,
      "step": 1025,
      "time_per_iteration": 2.690725564956665
    },
    {
      "auxiliary_loss_clip": 0.01197064,
      "auxiliary_loss_mlp": 0.01066504,
      "balance_loss_clip": 1.06452,
      "balance_loss_mlp": 1.04173219,
      "epoch": 0.06168645723733654,
      "flos": 25593104234880.0,
      "grad_norm": 2.5217294712155414,
      "language_loss": 0.84536898,
      "learning_rate": 3.989477727938335e-06,
      "loss": 0.86800468,
      "num_input_tokens_seen": 21978625,
      "step": 1026,
      "time_per_iteration": 2.7420806884765625
    },
    {
      "auxiliary_loss_clip": 0.01160797,
      "auxiliary_loss_mlp": 0.0107525,
      "balance_loss_clip": 1.05669701,
      "balance_loss_mlp": 1.04934609,
      "epoch": 0.06174658049000451,
      "flos": 15997917162240.0,
      "grad_norm": 2.354014397396182,
      "language_loss": 0.8228389,
      "learning_rate": 3.989437792548839e-06,
      "loss": 0.84519935,
      "num_input_tokens_seen": 21996035,
      "step": 1027,
      "time_per_iteration": 2.6683874130249023
    },
    {
      "auxiliary_loss_clip": 0.01160199,
      "auxiliary_loss_mlp": 0.01067253,
      "balance_loss_clip": 1.06181073,
      "balance_loss_mlp": 1.04232645,
      "epoch": 0.06180670374267248,
      "flos": 11285970117120.0,
      "grad_norm": 4.43492107605727,
      "language_loss": 0.83898664,
      "learning_rate": 3.989397781719663e-06,
      "loss": 0.86126107,
      "num_input_tokens_seen": 22011625,
      "step": 1028,
      "time_per_iteration": 2.705387592315674
    },
    {
      "auxiliary_loss_clip": 0.0106503,
      "auxiliary_loss_mlp": 0.01008074,
      "balance_loss_clip": 1.02410197,
      "balance_loss_mlp": 1.00347257,
      "epoch": 0.06186682699534045,
      "flos": 65130142216320.0,
      "grad_norm": 0.9383255649985517,
      "language_loss": 0.604738,
      "learning_rate": 3.989357695452323e-06,
      "loss": 0.62546903,
      "num_input_tokens_seen": 22066035,
      "step": 1029,
      "time_per_iteration": 3.0268616676330566
    },
    {
      "auxiliary_loss_clip": 0.01176182,
      "auxiliary_loss_mlp": 0.01074173,
      "balance_loss_clip": 1.05641246,
      "balance_loss_mlp": 1.04737473,
      "epoch": 0.061926950248008414,
      "flos": 21105742976640.0,
      "grad_norm": 4.246634693563946,
      "language_loss": 0.82589179,
      "learning_rate": 3.98931753374834e-06,
      "loss": 0.84839535,
      "num_input_tokens_seen": 22085015,
      "step": 1030,
      "time_per_iteration": 2.7035892009735107
    },
    {
      "auxiliary_loss_clip": 0.0122298,
      "auxiliary_loss_mlp": 0.01077745,
      "balance_loss_clip": 1.06850278,
      "balance_loss_mlp": 1.05185235,
      "epoch": 0.061987073500676386,
      "flos": 17748454481280.0,
      "grad_norm": 2.585240230669548,
      "language_loss": 0.79576576,
      "learning_rate": 3.989277296609237e-06,
      "loss": 0.81877303,
      "num_input_tokens_seen": 22102775,
      "step": 1031,
      "time_per_iteration": 2.60622501373291
    },
    {
      "auxiliary_loss_clip": 0.01188957,
      "auxiliary_loss_mlp": 0.01076754,
      "balance_loss_clip": 1.06396544,
      "balance_loss_mlp": 1.04982424,
      "epoch": 0.06204719675334436,
      "flos": 21836237869440.0,
      "grad_norm": 1.8815476991595563,
      "language_loss": 0.77384412,
      "learning_rate": 3.98923698403654e-06,
      "loss": 0.79650116,
      "num_input_tokens_seen": 22121680,
      "step": 1032,
      "time_per_iteration": 2.6753971576690674
    },
    {
      "auxiliary_loss_clip": 0.01198757,
      "auxiliary_loss_mlp": 0.01074736,
      "balance_loss_clip": 1.05916619,
      "balance_loss_mlp": 1.04848623,
      "epoch": 0.06210732000601232,
      "flos": 19353697286400.0,
      "grad_norm": 3.147941025479245,
      "language_loss": 0.89323574,
      "learning_rate": 3.989196596031776e-06,
      "loss": 0.91597068,
      "num_input_tokens_seen": 22138155,
      "step": 1033,
      "time_per_iteration": 2.7313079833984375
    },
    {
      "auxiliary_loss_clip": 0.01209161,
      "auxiliary_loss_mlp": 0.01066082,
      "balance_loss_clip": 1.06214237,
      "balance_loss_mlp": 1.04119134,
      "epoch": 0.062167443258680295,
      "flos": 24749382695040.0,
      "grad_norm": 2.1035343880884145,
      "language_loss": 0.8455385,
      "learning_rate": 3.989156132596479e-06,
      "loss": 0.8682909,
      "num_input_tokens_seen": 22157420,
      "step": 1034,
      "time_per_iteration": 2.7541439533233643
    },
    {
      "auxiliary_loss_clip": 0.01180042,
      "auxiliary_loss_mlp": 0.01057312,
      "balance_loss_clip": 1.05896068,
      "balance_loss_mlp": 1.03155136,
      "epoch": 0.06222756651134827,
      "flos": 34458478773120.0,
      "grad_norm": 1.8983498110529735,
      "language_loss": 0.8082794,
      "learning_rate": 3.989115593732182e-06,
      "loss": 0.83065289,
      "num_input_tokens_seen": 22178620,
      "step": 1035,
      "time_per_iteration": 2.7965424060821533
    },
    {
      "auxiliary_loss_clip": 0.01158806,
      "auxiliary_loss_mlp": 0.01072478,
      "balance_loss_clip": 1.05936599,
      "balance_loss_mlp": 1.04432034,
      "epoch": 0.06228768976401623,
      "flos": 25666469763840.0,
      "grad_norm": 2.145216314952277,
      "language_loss": 0.78365827,
      "learning_rate": 3.989074979440421e-06,
      "loss": 0.80597103,
      "num_input_tokens_seen": 22197125,
      "step": 1036,
      "time_per_iteration": 2.7858450412750244
    },
    {
      "auxiliary_loss_clip": 0.01192097,
      "auxiliary_loss_mlp": 0.01071382,
      "balance_loss_clip": 1.05977845,
      "balance_loss_mlp": 1.04663444,
      "epoch": 0.062347813016684205,
      "flos": 25295619795840.0,
      "grad_norm": 1.9535870339716077,
      "language_loss": 0.86544567,
      "learning_rate": 3.989034289722739e-06,
      "loss": 0.88808048,
      "num_input_tokens_seen": 22217575,
      "step": 1037,
      "time_per_iteration": 2.685373306274414
    },
    {
      "auxiliary_loss_clip": 0.01197778,
      "auxiliary_loss_mlp": 0.01057095,
      "balance_loss_clip": 1.06127763,
      "balance_loss_mlp": 1.02966499,
      "epoch": 0.06240793626935217,
      "flos": 26907039740160.0,
      "grad_norm": 2.697396725345887,
      "language_loss": 0.8067717,
      "learning_rate": 3.988993524580676e-06,
      "loss": 0.82932043,
      "num_input_tokens_seen": 22236840,
      "step": 1038,
      "time_per_iteration": 2.7305831909179688
    },
    {
      "auxiliary_loss_clip": 0.01145896,
      "auxiliary_loss_mlp": 0.01072721,
      "balance_loss_clip": 1.05226004,
      "balance_loss_mlp": 1.04330015,
      "epoch": 0.06246805952202014,
      "flos": 21615782146560.0,
      "grad_norm": 1.8888526922505675,
      "language_loss": 0.85465872,
      "learning_rate": 3.98895268401578e-06,
      "loss": 0.87684488,
      "num_input_tokens_seen": 22256465,
      "step": 1039,
      "time_per_iteration": 2.7351109981536865
    },
    {
      "auxiliary_loss_clip": 0.01188545,
      "auxiliary_loss_mlp": 0.01070323,
      "balance_loss_clip": 1.05834138,
      "balance_loss_mlp": 1.04472923,
      "epoch": 0.0625281827746881,
      "flos": 19311896833920.0,
      "grad_norm": 2.217895985816133,
      "language_loss": 0.81172895,
      "learning_rate": 3.9889117680296e-06,
      "loss": 0.83431756,
      "num_input_tokens_seen": 22274025,
      "step": 1040,
      "time_per_iteration": 2.6532907485961914
    },
    {
      "auxiliary_loss_clip": 0.0121654,
      "auxiliary_loss_mlp": 0.0106312,
      "balance_loss_clip": 1.06718016,
      "balance_loss_mlp": 1.03808582,
      "epoch": 0.06258830602735609,
      "flos": 27745769289600.0,
      "grad_norm": 2.1960038080149817,
      "language_loss": 0.69304991,
      "learning_rate": 3.988870776623685e-06,
      "loss": 0.71584648,
      "num_input_tokens_seen": 22292245,
      "step": 1041,
      "time_per_iteration": 2.6445486545562744
    },
    {
      "auxiliary_loss_clip": 0.01214659,
      "auxiliary_loss_mlp": 0.01057975,
      "balance_loss_clip": 1.06247008,
      "balance_loss_mlp": 1.03182077,
      "epoch": 0.06264842928002405,
      "flos": 23222605150080.0,
      "grad_norm": 2.7326158002445,
      "language_loss": 0.81187552,
      "learning_rate": 3.9888297097995905e-06,
      "loss": 0.83460188,
      "num_input_tokens_seen": 22311455,
      "step": 1042,
      "time_per_iteration": 2.6111559867858887
    },
    {
      "auxiliary_loss_clip": 0.01211653,
      "auxiliary_loss_mlp": 0.01052676,
      "balance_loss_clip": 1.06253886,
      "balance_loss_mlp": 1.02871442,
      "epoch": 0.06270855253269202,
      "flos": 38399495189760.0,
      "grad_norm": 1.7165873820424848,
      "language_loss": 0.76349056,
      "learning_rate": 3.988788567558874e-06,
      "loss": 0.78613389,
      "num_input_tokens_seen": 22333750,
      "step": 1043,
      "time_per_iteration": 2.761768341064453
    },
    {
      "auxiliary_loss_clip": 0.0118944,
      "auxiliary_loss_mlp": 0.01063189,
      "balance_loss_clip": 1.06111181,
      "balance_loss_mlp": 1.03912091,
      "epoch": 0.06276867578535998,
      "flos": 22453542028800.0,
      "grad_norm": 8.34017761542712,
      "language_loss": 0.92031956,
      "learning_rate": 3.988747349903097e-06,
      "loss": 0.94284582,
      "num_input_tokens_seen": 22351940,
      "step": 1044,
      "time_per_iteration": 2.636179208755493
    },
    {
      "auxiliary_loss_clip": 0.01192566,
      "auxiliary_loss_mlp": 0.01070128,
      "balance_loss_clip": 1.05862689,
      "balance_loss_mlp": 1.0456785,
      "epoch": 0.06282879903802796,
      "flos": 22930435923840.0,
      "grad_norm": 2.3486674311430944,
      "language_loss": 0.85913992,
      "learning_rate": 3.988706056833821e-06,
      "loss": 0.88176692,
      "num_input_tokens_seen": 22372085,
      "step": 1045,
      "time_per_iteration": 2.7749502658843994
    },
    {
      "auxiliary_loss_clip": 0.01179197,
      "auxiliary_loss_mlp": 0.01065176,
      "balance_loss_clip": 1.05804443,
      "balance_loss_mlp": 1.04053521,
      "epoch": 0.06288892229069593,
      "flos": 34819237019520.0,
      "grad_norm": 1.9846122850853416,
      "language_loss": 0.7796576,
      "learning_rate": 3.9886646883526125e-06,
      "loss": 0.80210131,
      "num_input_tokens_seen": 22392020,
      "step": 1046,
      "time_per_iteration": 2.803135871887207
    },
    {
      "auxiliary_loss_clip": 0.01197344,
      "auxiliary_loss_mlp": 0.01069269,
      "balance_loss_clip": 1.06361508,
      "balance_loss_mlp": 1.04558206,
      "epoch": 0.06294904554336389,
      "flos": 19427134642560.0,
      "grad_norm": 2.174325060947129,
      "language_loss": 0.77326387,
      "learning_rate": 3.988623244461039e-06,
      "loss": 0.79592997,
      "num_input_tokens_seen": 22411180,
      "step": 1047,
      "time_per_iteration": 2.647446632385254
    },
    {
      "auxiliary_loss_clip": 0.01200907,
      "auxiliary_loss_mlp": 0.0105799,
      "balance_loss_clip": 1.06238222,
      "balance_loss_mlp": 1.03314662,
      "epoch": 0.06300916879603187,
      "flos": 40661867358720.0,
      "grad_norm": 2.4899372640825046,
      "language_loss": 0.77190751,
      "learning_rate": 3.988581725160672e-06,
      "loss": 0.79449654,
      "num_input_tokens_seen": 22435105,
      "step": 1048,
      "time_per_iteration": 2.8167293071746826
    },
    {
      "auxiliary_loss_clip": 0.0118184,
      "auxiliary_loss_mlp": 0.01064361,
      "balance_loss_clip": 1.0613215,
      "balance_loss_mlp": 1.03914821,
      "epoch": 0.06306929204869983,
      "flos": 23804142341760.0,
      "grad_norm": 4.606540291271834,
      "language_loss": 0.77258086,
      "learning_rate": 3.988540130453087e-06,
      "loss": 0.79504287,
      "num_input_tokens_seen": 22452710,
      "step": 1049,
      "time_per_iteration": 2.6908538341522217
    },
    {
      "auxiliary_loss_clip": 0.01194538,
      "auxiliary_loss_mlp": 0.0105682,
      "balance_loss_clip": 1.06043661,
      "balance_loss_mlp": 1.03290701,
      "epoch": 0.0631294153013678,
      "flos": 18915802583040.0,
      "grad_norm": 2.515998307474139,
      "language_loss": 0.83302009,
      "learning_rate": 3.988498460339862e-06,
      "loss": 0.85553372,
      "num_input_tokens_seen": 22470175,
      "step": 1050,
      "time_per_iteration": 2.62186861038208
    },
    {
      "auxiliary_loss_clip": 0.01210654,
      "auxiliary_loss_mlp": 0.01062894,
      "balance_loss_clip": 1.06468701,
      "balance_loss_mlp": 1.04008913,
      "epoch": 0.06318953855403578,
      "flos": 24280174310400.0,
      "grad_norm": 5.5478202090132065,
      "language_loss": 0.76564771,
      "learning_rate": 3.988456714822575e-06,
      "loss": 0.78838319,
      "num_input_tokens_seen": 22490020,
      "step": 1051,
      "time_per_iteration": 2.732269525527954
    },
    {
      "auxiliary_loss_clip": 0.01188416,
      "auxiliary_loss_mlp": 0.01069443,
      "balance_loss_clip": 1.06340146,
      "balance_loss_mlp": 1.04492211,
      "epoch": 0.06324966180670374,
      "flos": 22528918719360.0,
      "grad_norm": 1.9993900469270787,
      "language_loss": 0.80410004,
      "learning_rate": 3.98841489390281e-06,
      "loss": 0.82667863,
      "num_input_tokens_seen": 22509685,
      "step": 1052,
      "time_per_iteration": 2.7683873176574707
    },
    {
      "auxiliary_loss_clip": 0.01211333,
      "auxiliary_loss_mlp": 0.01058255,
      "balance_loss_clip": 1.06324601,
      "balance_loss_mlp": 1.03468728,
      "epoch": 0.06330978505937171,
      "flos": 15778107884160.0,
      "grad_norm": 2.370007457349547,
      "language_loss": 0.77433288,
      "learning_rate": 3.988372997582155e-06,
      "loss": 0.79702866,
      "num_input_tokens_seen": 22527905,
      "step": 1053,
      "time_per_iteration": 5.757168531417847
    },
    {
      "auxiliary_loss_clip": 0.01190721,
      "auxiliary_loss_mlp": 0.00780448,
      "balance_loss_clip": 1.06378174,
      "balance_loss_mlp": 1.00028598,
      "epoch": 0.06336990831203967,
      "flos": 21471098163840.0,
      "grad_norm": 3.085258828985267,
      "language_loss": 0.84931248,
      "learning_rate": 3.988331025862195e-06,
      "loss": 0.86902416,
      "num_input_tokens_seen": 22546335,
      "step": 1054,
      "time_per_iteration": 2.7733829021453857
    },
    {
      "auxiliary_loss_clip": 0.01172281,
      "auxiliary_loss_mlp": 0.01061232,
      "balance_loss_clip": 1.05722666,
      "balance_loss_mlp": 1.03753328,
      "epoch": 0.06343003156470765,
      "flos": 18478877546880.0,
      "grad_norm": 2.0168531459993435,
      "language_loss": 0.85884213,
      "learning_rate": 3.9882889787445225e-06,
      "loss": 0.88117731,
      "num_input_tokens_seen": 22563885,
      "step": 1055,
      "time_per_iteration": 4.490305185317993
    },
    {
      "auxiliary_loss_clip": 0.01164237,
      "auxiliary_loss_mlp": 0.01069785,
      "balance_loss_clip": 1.05727792,
      "balance_loss_mlp": 1.04534709,
      "epoch": 0.06349015481737562,
      "flos": 25154886309120.0,
      "grad_norm": 2.4509218988768,
      "language_loss": 0.8113938,
      "learning_rate": 3.988246856230734e-06,
      "loss": 0.83373404,
      "num_input_tokens_seen": 22583035,
      "step": 1056,
      "time_per_iteration": 5.345282793045044
    },
    {
      "auxiliary_loss_clip": 0.01144181,
      "auxiliary_loss_mlp": 0.01061125,
      "balance_loss_clip": 1.04991364,
      "balance_loss_mlp": 1.03449368,
      "epoch": 0.06355027807004358,
      "flos": 26871775562880.0,
      "grad_norm": 2.2117272688527128,
      "language_loss": 0.81083393,
      "learning_rate": 3.988204658322426e-06,
      "loss": 0.83288693,
      "num_input_tokens_seen": 22605055,
      "step": 1057,
      "time_per_iteration": 2.866757392883301
    },
    {
      "auxiliary_loss_clip": 0.01139076,
      "auxiliary_loss_mlp": 0.01061742,
      "balance_loss_clip": 1.04970908,
      "balance_loss_mlp": 1.03918755,
      "epoch": 0.06361040132271156,
      "flos": 21396691140480.0,
      "grad_norm": 1.9636971972870172,
      "language_loss": 0.83353591,
      "learning_rate": 3.988162385021196e-06,
      "loss": 0.85554409,
      "num_input_tokens_seen": 22623760,
      "step": 1058,
      "time_per_iteration": 2.767024278640747
    },
    {
      "auxiliary_loss_clip": 0.0117752,
      "auxiliary_loss_mlp": 0.01059639,
      "balance_loss_clip": 1.0576936,
      "balance_loss_mlp": 1.03408027,
      "epoch": 0.06367052457537953,
      "flos": 25733765894400.0,
      "grad_norm": 2.137077300251244,
      "language_loss": 0.87556928,
      "learning_rate": 3.988120036328651e-06,
      "loss": 0.89794087,
      "num_input_tokens_seen": 22643000,
      "step": 1059,
      "time_per_iteration": 2.794734239578247
    },
    {
      "auxiliary_loss_clip": 0.01169658,
      "auxiliary_loss_mlp": 0.01063463,
      "balance_loss_clip": 1.06196678,
      "balance_loss_mlp": 1.0383693,
      "epoch": 0.0637306478280475,
      "flos": 17631420992640.0,
      "grad_norm": 2.543966627588717,
      "language_loss": 0.91561133,
      "learning_rate": 3.988077612246394e-06,
      "loss": 0.93794256,
      "num_input_tokens_seen": 22660460,
      "step": 1060,
      "time_per_iteration": 2.8223626613616943
    },
    {
      "auxiliary_loss_clip": 0.01173933,
      "auxiliary_loss_mlp": 0.01065151,
      "balance_loss_clip": 1.05715585,
      "balance_loss_mlp": 1.03981876,
      "epoch": 0.06379077108071547,
      "flos": 13662610427520.0,
      "grad_norm": 1.9401711052692647,
      "language_loss": 0.87242293,
      "learning_rate": 3.988035112776035e-06,
      "loss": 0.89481378,
      "num_input_tokens_seen": 22679270,
      "step": 1061,
      "time_per_iteration": 2.7783865928649902
    },
    {
      "auxiliary_loss_clip": 0.01190039,
      "auxiliary_loss_mlp": 0.01059971,
      "balance_loss_clip": 1.05976009,
      "balance_loss_mlp": 1.03388786,
      "epoch": 0.06385089433338344,
      "flos": 28478849961600.0,
      "grad_norm": 5.360593029379932,
      "language_loss": 0.77407908,
      "learning_rate": 3.987992537919185e-06,
      "loss": 0.79657912,
      "num_input_tokens_seen": 22699330,
      "step": 1062,
      "time_per_iteration": 2.872587203979492
    },
    {
      "auxiliary_loss_clip": 0.01172912,
      "auxiliary_loss_mlp": 0.01061175,
      "balance_loss_clip": 1.05884075,
      "balance_loss_mlp": 1.03798842,
      "epoch": 0.0639110175860514,
      "flos": 24311057028480.0,
      "grad_norm": 2.2658654128491436,
      "language_loss": 0.86522883,
      "learning_rate": 3.987949887677459e-06,
      "loss": 0.88756967,
      "num_input_tokens_seen": 22717945,
      "step": 1063,
      "time_per_iteration": 2.7915029525756836
    },
    {
      "auxiliary_loss_clip": 0.01207773,
      "auxiliary_loss_mlp": 0.01062698,
      "balance_loss_clip": 1.05969334,
      "balance_loss_mlp": 1.03846335,
      "epoch": 0.06397114083871938,
      "flos": 22090772620800.0,
      "grad_norm": 2.302236346678267,
      "language_loss": 0.79908657,
      "learning_rate": 3.9879071620524744e-06,
      "loss": 0.82179129,
      "num_input_tokens_seen": 22736790,
      "step": 1064,
      "time_per_iteration": 2.6880991458892822
    },
    {
      "auxiliary_loss_clip": 0.01198826,
      "auxiliary_loss_mlp": 0.01066465,
      "balance_loss_clip": 1.0603801,
      "balance_loss_mlp": 1.04149103,
      "epoch": 0.06403126409138735,
      "flos": 19572824206080.0,
      "grad_norm": 3.1552731138796215,
      "language_loss": 0.84327948,
      "learning_rate": 3.987864361045851e-06,
      "loss": 0.8659324,
      "num_input_tokens_seen": 22754745,
      "step": 1065,
      "time_per_iteration": 2.6956398487091064
    },
    {
      "auxiliary_loss_clip": 0.01168098,
      "auxiliary_loss_mlp": 0.01054905,
      "balance_loss_clip": 1.0597136,
      "balance_loss_mlp": 1.03162324,
      "epoch": 0.06409138734405531,
      "flos": 40807413267840.0,
      "grad_norm": 1.52830872536012,
      "language_loss": 0.68177885,
      "learning_rate": 3.987821484659211e-06,
      "loss": 0.70400894,
      "num_input_tokens_seen": 22776780,
      "step": 1066,
      "time_per_iteration": 2.9867773056030273
    },
    {
      "auxiliary_loss_clip": 0.01214184,
      "auxiliary_loss_mlp": 0.01070649,
      "balance_loss_clip": 1.06780005,
      "balance_loss_mlp": 1.04609215,
      "epoch": 0.06415151059672328,
      "flos": 20441610460800.0,
      "grad_norm": 1.8546001537284342,
      "language_loss": 0.90349269,
      "learning_rate": 3.987778532894181e-06,
      "loss": 0.926341,
      "num_input_tokens_seen": 22793915,
      "step": 1067,
      "time_per_iteration": 2.685896873474121
    },
    {
      "auxiliary_loss_clip": 0.01188134,
      "auxiliary_loss_mlp": 0.01063022,
      "balance_loss_clip": 1.0623709,
      "balance_loss_mlp": 1.03969264,
      "epoch": 0.06421163384939126,
      "flos": 18072045129600.0,
      "grad_norm": 2.189788428445167,
      "language_loss": 0.83437371,
      "learning_rate": 3.987735505752391e-06,
      "loss": 0.85688531,
      "num_input_tokens_seen": 22812670,
      "step": 1068,
      "time_per_iteration": 2.851602554321289
    },
    {
      "auxiliary_loss_clip": 0.01178972,
      "auxiliary_loss_mlp": 0.01057745,
      "balance_loss_clip": 1.05909026,
      "balance_loss_mlp": 1.03426039,
      "epoch": 0.06427175710205922,
      "flos": 25119442563840.0,
      "grad_norm": 3.045176948020938,
      "language_loss": 0.89311272,
      "learning_rate": 3.987692403235471e-06,
      "loss": 0.9154799,
      "num_input_tokens_seen": 22832440,
      "step": 1069,
      "time_per_iteration": 2.7825255393981934
    },
    {
      "auxiliary_loss_clip": 0.01185672,
      "auxiliary_loss_mlp": 0.01071834,
      "balance_loss_clip": 1.06158304,
      "balance_loss_mlp": 1.04663301,
      "epoch": 0.06433188035472719,
      "flos": 17380549428480.0,
      "grad_norm": 2.7038488706194808,
      "language_loss": 0.95759481,
      "learning_rate": 3.987649225345056e-06,
      "loss": 0.98016989,
      "num_input_tokens_seen": 22845495,
      "step": 1070,
      "time_per_iteration": 2.715296506881714
    },
    {
      "auxiliary_loss_clip": 0.01140792,
      "auxiliary_loss_mlp": 0.01056718,
      "balance_loss_clip": 1.05607581,
      "balance_loss_mlp": 1.03027749,
      "epoch": 0.06439200360739517,
      "flos": 23546267625600.0,
      "grad_norm": 1.630790580283393,
      "language_loss": 0.8811003,
      "learning_rate": 3.987605972082782e-06,
      "loss": 0.90307534,
      "num_input_tokens_seen": 22865390,
      "step": 1071,
      "time_per_iteration": 2.8445394039154053
    },
    {
      "auxiliary_loss_clip": 0.01155172,
      "auxiliary_loss_mlp": 0.01054986,
      "balance_loss_clip": 1.05483651,
      "balance_loss_mlp": 1.03102481,
      "epoch": 0.06445212686006313,
      "flos": 21979772616960.0,
      "grad_norm": 1.8349443396730127,
      "language_loss": 0.76116478,
      "learning_rate": 3.987562643450292e-06,
      "loss": 0.78326637,
      "num_input_tokens_seen": 22885495,
      "step": 1072,
      "time_per_iteration": 2.8330819606781006
    },
    {
      "auxiliary_loss_clip": 0.01172997,
      "auxiliary_loss_mlp": 0.01070104,
      "balance_loss_clip": 1.05975842,
      "balance_loss_mlp": 1.04362798,
      "epoch": 0.0645122501127311,
      "flos": 25921291824000.0,
      "grad_norm": 2.724283900767911,
      "language_loss": 0.80849886,
      "learning_rate": 3.987519239449226e-06,
      "loss": 0.83092993,
      "num_input_tokens_seen": 22904845,
      "step": 1073,
      "time_per_iteration": 2.748286247253418
    },
    {
      "auxiliary_loss_clip": 0.01194712,
      "auxiliary_loss_mlp": 0.01062452,
      "balance_loss_clip": 1.06345201,
      "balance_loss_mlp": 1.03825283,
      "epoch": 0.06457237336539907,
      "flos": 25626034028160.0,
      "grad_norm": 5.0538746884234245,
      "language_loss": 0.80282539,
      "learning_rate": 3.987475760081233e-06,
      "loss": 0.82539707,
      "num_input_tokens_seen": 22925940,
      "step": 1074,
      "time_per_iteration": 2.7482337951660156
    },
    {
      "auxiliary_loss_clip": 0.01173366,
      "auxiliary_loss_mlp": 0.01057774,
      "balance_loss_clip": 1.05920076,
      "balance_loss_mlp": 1.03256142,
      "epoch": 0.06463249661806704,
      "flos": 19463979018240.0,
      "grad_norm": 2.0209756517373707,
      "language_loss": 0.79249811,
      "learning_rate": 3.987432205347958e-06,
      "loss": 0.8148095,
      "num_input_tokens_seen": 22944375,
      "step": 1075,
      "time_per_iteration": 2.6937224864959717
    },
    {
      "auxiliary_loss_clip": 0.01171297,
      "auxiliary_loss_mlp": 0.01063569,
      "balance_loss_clip": 1.05735481,
      "balance_loss_mlp": 1.04025126,
      "epoch": 0.064692619870735,
      "flos": 24498044254080.0,
      "grad_norm": 2.9028991302223357,
      "language_loss": 0.88208115,
      "learning_rate": 3.987388575251055e-06,
      "loss": 0.90442967,
      "num_input_tokens_seen": 22959145,
      "step": 1076,
      "time_per_iteration": 2.878103256225586
    },
    {
      "auxiliary_loss_clip": 0.01192915,
      "auxiliary_loss_mlp": 0.01052877,
      "balance_loss_clip": 1.06164443,
      "balance_loss_mlp": 1.0288558,
      "epoch": 0.06475274312340297,
      "flos": 17018677860480.0,
      "grad_norm": 2.225760792581628,
      "language_loss": 0.80876106,
      "learning_rate": 3.98734486979218e-06,
      "loss": 0.83121902,
      "num_input_tokens_seen": 22978100,
      "step": 1077,
      "time_per_iteration": 2.7221076488494873
    },
    {
      "auxiliary_loss_clip": 0.01200466,
      "auxiliary_loss_mlp": 0.01064019,
      "balance_loss_clip": 1.0656153,
      "balance_loss_mlp": 1.03866291,
      "epoch": 0.06481286637607095,
      "flos": 24572379450240.0,
      "grad_norm": 2.256787147683815,
      "language_loss": 0.91727465,
      "learning_rate": 3.987301088972986e-06,
      "loss": 0.93991947,
      "num_input_tokens_seen": 22997285,
      "step": 1078,
      "time_per_iteration": 2.862365484237671
    },
    {
      "auxiliary_loss_clip": 0.0122435,
      "auxiliary_loss_mlp": 0.01060225,
      "balance_loss_clip": 1.06826639,
      "balance_loss_mlp": 1.03552508,
      "epoch": 0.06487298962873891,
      "flos": 21105635235840.0,
      "grad_norm": 2.080056711608912,
      "language_loss": 0.78349572,
      "learning_rate": 3.987257232795137e-06,
      "loss": 0.80634147,
      "num_input_tokens_seen": 23016285,
      "step": 1079,
      "time_per_iteration": 2.6435368061065674
    },
    {
      "auxiliary_loss_clip": 0.01156927,
      "auxiliary_loss_mlp": 0.01063794,
      "balance_loss_clip": 1.05512071,
      "balance_loss_mlp": 1.03899896,
      "epoch": 0.06493311288140688,
      "flos": 24608182331520.0,
      "grad_norm": 2.274862403364013,
      "language_loss": 0.68702769,
      "learning_rate": 3.987213301260294e-06,
      "loss": 0.70923495,
      "num_input_tokens_seen": 23036420,
      "step": 1080,
      "time_per_iteration": 2.7782626152038574
    },
    {
      "auxiliary_loss_clip": 0.01175684,
      "auxiliary_loss_mlp": 0.01062351,
      "balance_loss_clip": 1.06640029,
      "balance_loss_mlp": 1.03610086,
      "epoch": 0.06499323613407486,
      "flos": 25337994865920.0,
      "grad_norm": 1.886196453243775,
      "language_loss": 0.72291583,
      "learning_rate": 3.987169294370123e-06,
      "loss": 0.74529618,
      "num_input_tokens_seen": 23056945,
      "step": 1081,
      "time_per_iteration": 2.7983880043029785
    },
    {
      "auxiliary_loss_clip": 0.01139671,
      "auxiliary_loss_mlp": 0.01066686,
      "balance_loss_clip": 1.0504055,
      "balance_loss_mlp": 1.04076982,
      "epoch": 0.06505335938674282,
      "flos": 20375714960640.0,
      "grad_norm": 3.3093934650613566,
      "language_loss": 0.84059012,
      "learning_rate": 3.987125212126294e-06,
      "loss": 0.86265367,
      "num_input_tokens_seen": 23074940,
      "step": 1082,
      "time_per_iteration": 2.8351900577545166
    },
    {
      "auxiliary_loss_clip": 0.01204185,
      "auxiliary_loss_mlp": 0.01063692,
      "balance_loss_clip": 1.06306195,
      "balance_loss_mlp": 1.03809738,
      "epoch": 0.06511348263941079,
      "flos": 25337923038720.0,
      "grad_norm": 2.894360492506304,
      "language_loss": 0.82550305,
      "learning_rate": 3.987081054530478e-06,
      "loss": 0.84818184,
      "num_input_tokens_seen": 23093420,
      "step": 1083,
      "time_per_iteration": 2.866729974746704
    },
    {
      "auxiliary_loss_clip": 0.01168245,
      "auxiliary_loss_mlp": 0.01062938,
      "balance_loss_clip": 1.06021011,
      "balance_loss_mlp": 1.03655696,
      "epoch": 0.06517360589207877,
      "flos": 20332801186560.0,
      "grad_norm": 2.468736383036802,
      "language_loss": 0.79289383,
      "learning_rate": 3.987036821584348e-06,
      "loss": 0.81520569,
      "num_input_tokens_seen": 23111550,
      "step": 1084,
      "time_per_iteration": 2.816601276397705
    },
    {
      "auxiliary_loss_clip": 0.01174068,
      "auxiliary_loss_mlp": 0.0106167,
      "balance_loss_clip": 1.05854714,
      "balance_loss_mlp": 1.03667152,
      "epoch": 0.06523372914474673,
      "flos": 31681650061440.0,
      "grad_norm": 2.571590277205686,
      "language_loss": 0.66443276,
      "learning_rate": 3.986992513289584e-06,
      "loss": 0.68679011,
      "num_input_tokens_seen": 23130335,
      "step": 1085,
      "time_per_iteration": 2.8260092735290527
    },
    {
      "auxiliary_loss_clip": 0.01170818,
      "auxiliary_loss_mlp": 0.01062435,
      "balance_loss_clip": 1.0600934,
      "balance_loss_mlp": 1.03833067,
      "epoch": 0.0652938523974147,
      "flos": 20778165918720.0,
      "grad_norm": 2.0478791529086977,
      "language_loss": 0.76548934,
      "learning_rate": 3.9869481296478645e-06,
      "loss": 0.78782183,
      "num_input_tokens_seen": 23152380,
      "step": 1086,
      "time_per_iteration": 2.7937023639678955
    },
    {
      "auxiliary_loss_clip": 0.01198609,
      "auxiliary_loss_mlp": 0.01059288,
      "balance_loss_clip": 1.06335294,
      "balance_loss_mlp": 1.03519547,
      "epoch": 0.06535397565008266,
      "flos": 16690993061760.0,
      "grad_norm": 2.1629448601391017,
      "language_loss": 0.85109925,
      "learning_rate": 3.986903670660872e-06,
      "loss": 0.87367821,
      "num_input_tokens_seen": 23171630,
      "step": 1087,
      "time_per_iteration": 2.7510013580322266
    },
    {
      "auxiliary_loss_clip": 0.01184978,
      "auxiliary_loss_mlp": 0.01059017,
      "balance_loss_clip": 1.06293821,
      "balance_loss_mlp": 1.03510392,
      "epoch": 0.06541409890275064,
      "flos": 26868220116480.0,
      "grad_norm": 1.7886353193129139,
      "language_loss": 0.77776635,
      "learning_rate": 3.9868591363302945e-06,
      "loss": 0.80020636,
      "num_input_tokens_seen": 23192520,
      "step": 1088,
      "time_per_iteration": 2.7792751789093018
    },
    {
      "auxiliary_loss_clip": 0.01192707,
      "auxiliary_loss_mlp": 0.01067634,
      "balance_loss_clip": 1.06569457,
      "balance_loss_mlp": 1.04498422,
      "epoch": 0.06547422215541861,
      "flos": 20521620005760.0,
      "grad_norm": 3.0334087154373375,
      "language_loss": 0.71050513,
      "learning_rate": 3.9868145266578186e-06,
      "loss": 0.73310852,
      "num_input_tokens_seen": 23210710,
      "step": 1089,
      "time_per_iteration": 2.8832852840423584
    },
    {
      "auxiliary_loss_clip": 0.01173663,
      "auxiliary_loss_mlp": 0.00781529,
      "balance_loss_clip": 1.06159782,
      "balance_loss_mlp": 1.00019014,
      "epoch": 0.06553434540808657,
      "flos": 22016616992640.0,
      "grad_norm": 2.02973275746688,
      "language_loss": 0.85650897,
      "learning_rate": 3.9867698416451366e-06,
      "loss": 0.87606084,
      "num_input_tokens_seen": 23230305,
      "step": 1090,
      "time_per_iteration": 2.7933149337768555
    },
    {
      "auxiliary_loss_clip": 0.01214666,
      "auxiliary_loss_mlp": 0.0105885,
      "balance_loss_clip": 1.06735325,
      "balance_loss_mlp": 1.03460288,
      "epoch": 0.06559446866075455,
      "flos": 24608649208320.0,
      "grad_norm": 2.137212216289862,
      "language_loss": 0.71829313,
      "learning_rate": 3.9867250812939434e-06,
      "loss": 0.74102825,
      "num_input_tokens_seen": 23249015,
      "step": 1091,
      "time_per_iteration": 2.646592855453491
    },
    {
      "auxiliary_loss_clip": 0.01121055,
      "auxiliary_loss_mlp": 0.0106405,
      "balance_loss_clip": 1.05242276,
      "balance_loss_mlp": 1.03961205,
      "epoch": 0.06565459191342252,
      "flos": 24274679529600.0,
      "grad_norm": 2.2773849385721956,
      "language_loss": 0.82839823,
      "learning_rate": 3.986680245605936e-06,
      "loss": 0.85024923,
      "num_input_tokens_seen": 23265105,
      "step": 1092,
      "time_per_iteration": 4.799649715423584
    },
    {
      "auxiliary_loss_clip": 0.01215092,
      "auxiliary_loss_mlp": 0.01059151,
      "balance_loss_clip": 1.0640471,
      "balance_loss_mlp": 1.03352082,
      "epoch": 0.06571471516609048,
      "flos": 24787124910720.0,
      "grad_norm": 2.268968080418226,
      "language_loss": 0.71134168,
      "learning_rate": 3.986635334582814e-06,
      "loss": 0.73408413,
      "num_input_tokens_seen": 23283950,
      "step": 1093,
      "time_per_iteration": 5.3356239795684814
    },
    {
      "auxiliary_loss_clip": 0.01190682,
      "auxiliary_loss_mlp": 0.01064498,
      "balance_loss_clip": 1.06751943,
      "balance_loss_mlp": 1.0392611,
      "epoch": 0.06577483841875846,
      "flos": 26214071581440.0,
      "grad_norm": 3.829837904337144,
      "language_loss": 0.87996346,
      "learning_rate": 3.986590348226282e-06,
      "loss": 0.90251523,
      "num_input_tokens_seen": 23305005,
      "step": 1094,
      "time_per_iteration": 2.853489637374878
    },
    {
      "auxiliary_loss_clip": 0.01192742,
      "auxiliary_loss_mlp": 0.01065068,
      "balance_loss_clip": 1.06367433,
      "balance_loss_mlp": 1.03843689,
      "epoch": 0.06583496167142643,
      "flos": 25080802508160.0,
      "grad_norm": 1.6736216436017588,
      "language_loss": 0.81483954,
      "learning_rate": 3.986545286538044e-06,
      "loss": 0.8374176,
      "num_input_tokens_seen": 23323220,
      "step": 1095,
      "time_per_iteration": 5.1613922119140625
    },
    {
      "auxiliary_loss_clip": 0.01166049,
      "auxiliary_loss_mlp": 0.01058945,
      "balance_loss_clip": 1.06295943,
      "balance_loss_mlp": 1.03598547,
      "epoch": 0.06589508492409439,
      "flos": 25629804956160.0,
      "grad_norm": 2.0200125290673068,
      "language_loss": 0.69789279,
      "learning_rate": 3.986500149519811e-06,
      "loss": 0.72014272,
      "num_input_tokens_seen": 23342235,
      "step": 1096,
      "time_per_iteration": 2.804025173187256
    },
    {
      "auxiliary_loss_clip": 0.01201939,
      "auxiliary_loss_mlp": 0.01070786,
      "balance_loss_clip": 1.06405246,
      "balance_loss_mlp": 1.04614568,
      "epoch": 0.06595520817676236,
      "flos": 23621249266560.0,
      "grad_norm": 1.7011375462517908,
      "language_loss": 0.77430046,
      "learning_rate": 3.986454937173292e-06,
      "loss": 0.79702777,
      "num_input_tokens_seen": 23363680,
      "step": 1097,
      "time_per_iteration": 2.7658958435058594
    },
    {
      "auxiliary_loss_clip": 0.01215996,
      "auxiliary_loss_mlp": 0.01063445,
      "balance_loss_clip": 1.06707537,
      "balance_loss_mlp": 1.03959155,
      "epoch": 0.06601533142943034,
      "flos": 33801708545280.0,
      "grad_norm": 1.8316558452843608,
      "language_loss": 0.78217584,
      "learning_rate": 3.986409649500203e-06,
      "loss": 0.80497026,
      "num_input_tokens_seen": 23385590,
      "step": 1098,
      "time_per_iteration": 2.865684747695923
    },
    {
      "auxiliary_loss_clip": 0.01197349,
      "auxiliary_loss_mlp": 0.01069192,
      "balance_loss_clip": 1.06328607,
      "balance_loss_mlp": 1.04443276,
      "epoch": 0.0660754546820983,
      "flos": 20259184262400.0,
      "grad_norm": 1.9237510259783663,
      "language_loss": 0.81525648,
      "learning_rate": 3.986364286502261e-06,
      "loss": 0.83792192,
      "num_input_tokens_seen": 23402945,
      "step": 1099,
      "time_per_iteration": 2.690377950668335
    },
    {
      "auxiliary_loss_clip": 0.01179995,
      "auxiliary_loss_mlp": 0.0105819,
      "balance_loss_clip": 1.0578239,
      "balance_loss_mlp": 1.03428841,
      "epoch": 0.06613557793476627,
      "flos": 19354164163200.0,
      "grad_norm": 1.9906927310803755,
      "language_loss": 0.82793295,
      "learning_rate": 3.986318848181186e-06,
      "loss": 0.8503148,
      "num_input_tokens_seen": 23421410,
      "step": 1100,
      "time_per_iteration": 2.7613909244537354
    },
    {
      "auxiliary_loss_clip": 0.01191263,
      "auxiliary_loss_mlp": 0.0105903,
      "balance_loss_clip": 1.06985724,
      "balance_loss_mlp": 1.03529549,
      "epoch": 0.06619570118743424,
      "flos": 13772568936960.0,
      "grad_norm": 2.079994286400427,
      "language_loss": 0.73502243,
      "learning_rate": 3.986273334538702e-06,
      "loss": 0.75752538,
      "num_input_tokens_seen": 23438870,
      "step": 1101,
      "time_per_iteration": 2.7795870304107666
    },
    {
      "auxiliary_loss_clip": 0.01199256,
      "auxiliary_loss_mlp": 0.01061171,
      "balance_loss_clip": 1.06278944,
      "balance_loss_mlp": 1.03773487,
      "epoch": 0.06625582444010221,
      "flos": 17857874286720.0,
      "grad_norm": 2.875757629612747,
      "language_loss": 0.85861301,
      "learning_rate": 3.986227745576533e-06,
      "loss": 0.88121736,
      "num_input_tokens_seen": 23456975,
      "step": 1102,
      "time_per_iteration": 2.737269401550293
    },
    {
      "auxiliary_loss_clip": 0.01191982,
      "auxiliary_loss_mlp": 0.01058639,
      "balance_loss_clip": 1.06898165,
      "balance_loss_mlp": 1.03410578,
      "epoch": 0.06631594769277017,
      "flos": 11838707579520.0,
      "grad_norm": 2.8924251757501778,
      "language_loss": 0.81655926,
      "learning_rate": 3.98618208129641e-06,
      "loss": 0.83906543,
      "num_input_tokens_seen": 23473440,
      "step": 1103,
      "time_per_iteration": 2.9345293045043945
    },
    {
      "auxiliary_loss_clip": 0.01203522,
      "auxiliary_loss_mlp": 0.00780451,
      "balance_loss_clip": 1.06721628,
      "balance_loss_mlp": 1.00042021,
      "epoch": 0.06637607094543815,
      "flos": 19793351756160.0,
      "grad_norm": 5.176370819061919,
      "language_loss": 0.81749105,
      "learning_rate": 3.986136341700063e-06,
      "loss": 0.83733076,
      "num_input_tokens_seen": 23493880,
      "step": 1104,
      "time_per_iteration": 2.753657102584839
    },
    {
      "auxiliary_loss_clip": 0.0116508,
      "auxiliary_loss_mlp": 0.01050687,
      "balance_loss_clip": 1.0576005,
      "balance_loss_mlp": 1.02608228,
      "epoch": 0.06643619419810612,
      "flos": 25485659677440.0,
      "grad_norm": 1.5448539486038575,
      "language_loss": 0.80422902,
      "learning_rate": 3.986090526789227e-06,
      "loss": 0.82638663,
      "num_input_tokens_seen": 23514920,
      "step": 1105,
      "time_per_iteration": 2.8904521465301514
    },
    {
      "auxiliary_loss_clip": 0.01179397,
      "auxiliary_loss_mlp": 0.0106197,
      "balance_loss_clip": 1.06348729,
      "balance_loss_mlp": 1.0391891,
      "epoch": 0.06649631745077408,
      "flos": 16946533393920.0,
      "grad_norm": 2.7426455725749896,
      "language_loss": 0.96762037,
      "learning_rate": 3.986044636565639e-06,
      "loss": 0.99003398,
      "num_input_tokens_seen": 23531635,
      "step": 1106,
      "time_per_iteration": 2.890073299407959
    },
    {
      "auxiliary_loss_clip": 0.01198065,
      "auxiliary_loss_mlp": 0.01059975,
      "balance_loss_clip": 1.06069684,
      "balance_loss_mlp": 1.03511953,
      "epoch": 0.06655644070344206,
      "flos": 17858592558720.0,
      "grad_norm": 1.9297768479693453,
      "language_loss": 0.82528949,
      "learning_rate": 3.985998671031039e-06,
      "loss": 0.84786987,
      "num_input_tokens_seen": 23551020,
      "step": 1107,
      "time_per_iteration": 2.778857469558716
    },
    {
      "auxiliary_loss_clip": 0.01104176,
      "auxiliary_loss_mlp": 0.01010935,
      "balance_loss_clip": 1.04708242,
      "balance_loss_mlp": 1.0072155,
      "epoch": 0.06661656395611003,
      "flos": 61419350021760.0,
      "grad_norm": 0.7967940032222198,
      "language_loss": 0.56789279,
      "learning_rate": 3.9859526301871705e-06,
      "loss": 0.58904392,
      "num_input_tokens_seen": 23610675,
      "step": 1108,
      "time_per_iteration": 3.2717819213867188
    },
    {
      "auxiliary_loss_clip": 0.0118327,
      "auxiliary_loss_mlp": 0.01062625,
      "balance_loss_clip": 1.05651307,
      "balance_loss_mlp": 1.0376507,
      "epoch": 0.066676687208778,
      "flos": 20662856282880.0,
      "grad_norm": 2.682842555407744,
      "language_loss": 0.7287578,
      "learning_rate": 3.9859065140357795e-06,
      "loss": 0.75121677,
      "num_input_tokens_seen": 23628710,
      "step": 1109,
      "time_per_iteration": 2.829623222351074
    },
    {
      "auxiliary_loss_clip": 0.01148971,
      "auxiliary_loss_mlp": 0.01071895,
      "balance_loss_clip": 1.05459642,
      "balance_loss_mlp": 1.04714715,
      "epoch": 0.06673681046144596,
      "flos": 20923280864640.0,
      "grad_norm": 1.7914435942805436,
      "language_loss": 0.78140426,
      "learning_rate": 3.985860322578614e-06,
      "loss": 0.80361295,
      "num_input_tokens_seen": 23649160,
      "step": 1110,
      "time_per_iteration": 2.892786741256714
    },
    {
      "auxiliary_loss_clip": 0.01153553,
      "auxiliary_loss_mlp": 0.0106147,
      "balance_loss_clip": 1.05590594,
      "balance_loss_mlp": 1.03700781,
      "epoch": 0.06679693371411394,
      "flos": 31065818359680.0,
      "grad_norm": 2.5260725451831805,
      "language_loss": 0.71425366,
      "learning_rate": 3.985814055817427e-06,
      "loss": 0.73640382,
      "num_input_tokens_seen": 23671995,
      "step": 1111,
      "time_per_iteration": 2.9349052906036377
    },
    {
      "auxiliary_loss_clip": 0.01170538,
      "auxiliary_loss_mlp": 0.01066103,
      "balance_loss_clip": 1.05776191,
      "balance_loss_mlp": 1.04199934,
      "epoch": 0.0668570569667819,
      "flos": 21726135705600.0,
      "grad_norm": 1.8396663794990693,
      "language_loss": 0.78767776,
      "learning_rate": 3.985767713753971e-06,
      "loss": 0.81004417,
      "num_input_tokens_seen": 23690705,
      "step": 1112,
      "time_per_iteration": 2.8676345348358154
    },
    {
      "auxiliary_loss_clip": 0.01153291,
      "auxiliary_loss_mlp": 0.01065421,
      "balance_loss_clip": 1.05340791,
      "balance_loss_mlp": 1.04163861,
      "epoch": 0.06691718021944987,
      "flos": 22747255539840.0,
      "grad_norm": 2.071048188460824,
      "language_loss": 0.78481978,
      "learning_rate": 3.985721296390005e-06,
      "loss": 0.80700684,
      "num_input_tokens_seen": 23709990,
      "step": 1113,
      "time_per_iteration": 2.8688411712646484
    },
    {
      "auxiliary_loss_clip": 0.0114872,
      "auxiliary_loss_mlp": 0.01057074,
      "balance_loss_clip": 1.05157375,
      "balance_loss_mlp": 1.03376842,
      "epoch": 0.06697730347211785,
      "flos": 16545626720640.0,
      "grad_norm": 1.7560007918285245,
      "language_loss": 0.82399213,
      "learning_rate": 3.985674803727289e-06,
      "loss": 0.84605002,
      "num_input_tokens_seen": 23728485,
      "step": 1114,
      "time_per_iteration": 2.832458019256592
    },
    {
      "auxiliary_loss_clip": 0.01075626,
      "auxiliary_loss_mlp": 0.01006906,
      "balance_loss_clip": 1.04995251,
      "balance_loss_mlp": 1.00271022,
      "epoch": 0.06703742672478581,
      "flos": 59782326658560.0,
      "grad_norm": 0.8370646888074905,
      "language_loss": 0.58147323,
      "learning_rate": 3.985628235767584e-06,
      "loss": 0.60229862,
      "num_input_tokens_seen": 23786650,
      "step": 1115,
      "time_per_iteration": 3.550837755203247
    },
    {
      "auxiliary_loss_clip": 0.01177193,
      "auxiliary_loss_mlp": 0.01059174,
      "balance_loss_clip": 1.05986214,
      "balance_loss_mlp": 1.03381801,
      "epoch": 0.06709754997745378,
      "flos": 16800197385600.0,
      "grad_norm": 2.8944873563712235,
      "language_loss": 0.91280693,
      "learning_rate": 3.985581592512658e-06,
      "loss": 0.93517065,
      "num_input_tokens_seen": 23802555,
      "step": 1116,
      "time_per_iteration": 2.994608163833618
    },
    {
      "auxiliary_loss_clip": 0.01169376,
      "auxiliary_loss_mlp": 0.0078227,
      "balance_loss_clip": 1.05839634,
      "balance_loss_mlp": 1.00045347,
      "epoch": 0.06715767323012176,
      "flos": 22123917895680.0,
      "grad_norm": 1.9249158333763592,
      "language_loss": 0.87154609,
      "learning_rate": 3.985534873964279e-06,
      "loss": 0.89106256,
      "num_input_tokens_seen": 23822945,
      "step": 1117,
      "time_per_iteration": 2.794400453567505
    },
    {
      "auxiliary_loss_clip": 0.01095782,
      "auxiliary_loss_mlp": 0.01003785,
      "balance_loss_clip": 1.0387876,
      "balance_loss_mlp": 0.99963647,
      "epoch": 0.06721779648278972,
      "flos": 66618100137600.0,
      "grad_norm": 0.8644388721740246,
      "language_loss": 0.5981611,
      "learning_rate": 3.985488080124218e-06,
      "loss": 0.61915678,
      "num_input_tokens_seen": 23874075,
      "step": 1118,
      "time_per_iteration": 3.1695809364318848
    },
    {
      "auxiliary_loss_clip": 0.01178972,
      "auxiliary_loss_mlp": 0.01051993,
      "balance_loss_clip": 1.05301392,
      "balance_loss_mlp": 1.02780545,
      "epoch": 0.06727791973545769,
      "flos": 22382474970240.0,
      "grad_norm": 3.6923711141076447,
      "language_loss": 0.83045954,
      "learning_rate": 3.985441210994251e-06,
      "loss": 0.85276914,
      "num_input_tokens_seen": 23889720,
      "step": 1119,
      "time_per_iteration": 2.7538814544677734
    },
    {
      "auxiliary_loss_clip": 0.01182384,
      "auxiliary_loss_mlp": 0.01058422,
      "balance_loss_clip": 1.06102347,
      "balance_loss_mlp": 1.03566504,
      "epoch": 0.06733804298812565,
      "flos": 24280210224000.0,
      "grad_norm": 4.541743494234462,
      "language_loss": 0.8451674,
      "learning_rate": 3.9853942665761545e-06,
      "loss": 0.86757541,
      "num_input_tokens_seen": 23909385,
      "step": 1120,
      "time_per_iteration": 2.76581072807312
    },
    {
      "auxiliary_loss_clip": 0.0121565,
      "auxiliary_loss_mlp": 0.01064916,
      "balance_loss_clip": 1.06757379,
      "balance_loss_mlp": 1.04028773,
      "epoch": 0.06739816624079363,
      "flos": 15918230839680.0,
      "grad_norm": 2.503866645162978,
      "language_loss": 0.78722781,
      "learning_rate": 3.985347246871708e-06,
      "loss": 0.81003344,
      "num_input_tokens_seen": 23926830,
      "step": 1121,
      "time_per_iteration": 2.651175022125244
    },
    {
      "auxiliary_loss_clip": 0.01080914,
      "auxiliary_loss_mlp": 0.01011889,
      "balance_loss_clip": 1.03108025,
      "balance_loss_mlp": 1.00802636,
      "epoch": 0.0674582894934616,
      "flos": 71398567353600.0,
      "grad_norm": 0.7540288133642103,
      "language_loss": 0.58320796,
      "learning_rate": 3.985300151882694e-06,
      "loss": 0.60413599,
      "num_input_tokens_seen": 23992640,
      "step": 1122,
      "time_per_iteration": 3.3794541358947754
    },
    {
      "auxiliary_loss_clip": 0.01145486,
      "auxiliary_loss_mlp": 0.01066136,
      "balance_loss_clip": 1.05581403,
      "balance_loss_mlp": 1.04167438,
      "epoch": 0.06751841274612956,
      "flos": 25264952559360.0,
      "grad_norm": 2.3361170394687076,
      "language_loss": 0.71965349,
      "learning_rate": 3.985252981610901e-06,
      "loss": 0.74176967,
      "num_input_tokens_seen": 24011135,
      "step": 1123,
      "time_per_iteration": 2.8049354553222656
    },
    {
      "auxiliary_loss_clip": 0.01144994,
      "auxiliary_loss_mlp": 0.01064196,
      "balance_loss_clip": 1.05373979,
      "balance_loss_mlp": 1.03612232,
      "epoch": 0.06757853599879754,
      "flos": 23802741711360.0,
      "grad_norm": 1.7380479869896208,
      "language_loss": 0.78987843,
      "learning_rate": 3.985205736058114e-06,
      "loss": 0.81197035,
      "num_input_tokens_seen": 24030695,
      "step": 1124,
      "time_per_iteration": 2.8595056533813477
    },
    {
      "auxiliary_loss_clip": 0.01189686,
      "auxiliary_loss_mlp": 0.01055169,
      "balance_loss_clip": 1.05663013,
      "balance_loss_mlp": 1.03200674,
      "epoch": 0.0676386592514655,
      "flos": 21033742164480.0,
      "grad_norm": 3.1450673626590793,
      "language_loss": 0.70999855,
      "learning_rate": 3.985158415226128e-06,
      "loss": 0.73244709,
      "num_input_tokens_seen": 24050680,
      "step": 1125,
      "time_per_iteration": 2.726163625717163
    },
    {
      "auxiliary_loss_clip": 0.01165518,
      "auxiliary_loss_mlp": 0.01068918,
      "balance_loss_clip": 1.05826426,
      "balance_loss_mlp": 1.04290628,
      "epoch": 0.06769878250413347,
      "flos": 25556331686400.0,
      "grad_norm": 3.340323364887528,
      "language_loss": 0.81440383,
      "learning_rate": 3.985111019116736e-06,
      "loss": 0.83674812,
      "num_input_tokens_seen": 24067205,
      "step": 1126,
      "time_per_iteration": 2.7356598377227783
    },
    {
      "auxiliary_loss_clip": 0.0107201,
      "auxiliary_loss_mlp": 0.01004999,
      "balance_loss_clip": 1.0293622,
      "balance_loss_mlp": 1.00092208,
      "epoch": 0.06775890575680145,
      "flos": 70655251305600.0,
      "grad_norm": 0.77802311726495,
      "language_loss": 0.59720373,
      "learning_rate": 3.985063547731735e-06,
      "loss": 0.6179738,
      "num_input_tokens_seen": 24131320,
      "step": 1127,
      "time_per_iteration": 3.2627320289611816
    },
    {
      "auxiliary_loss_clip": 0.01206438,
      "auxiliary_loss_mlp": 0.01055509,
      "balance_loss_clip": 1.06308687,
      "balance_loss_mlp": 1.03189397,
      "epoch": 0.06781902900946941,
      "flos": 24235500769920.0,
      "grad_norm": 2.2535941175889054,
      "language_loss": 0.81097019,
      "learning_rate": 3.985016001072925e-06,
      "loss": 0.83358967,
      "num_input_tokens_seen": 24149930,
      "step": 1128,
      "time_per_iteration": 2.6652371883392334
    },
    {
      "auxiliary_loss_clip": 0.01158345,
      "auxiliary_loss_mlp": 0.01052658,
      "balance_loss_clip": 1.05360484,
      "balance_loss_mlp": 1.02804112,
      "epoch": 0.06787915226213738,
      "flos": 22417523665920.0,
      "grad_norm": 2.24200367657907,
      "language_loss": 0.75559127,
      "learning_rate": 3.984968379142109e-06,
      "loss": 0.77770138,
      "num_input_tokens_seen": 24169590,
      "step": 1129,
      "time_per_iteration": 2.7023732662200928
    },
    {
      "auxiliary_loss_clip": 0.01117595,
      "auxiliary_loss_mlp": 0.01053995,
      "balance_loss_clip": 1.04627228,
      "balance_loss_mlp": 1.03006983,
      "epoch": 0.06793927551480534,
      "flos": 37706922080640.0,
      "grad_norm": 1.890559803272908,
      "language_loss": 0.71710479,
      "learning_rate": 3.984920681941094e-06,
      "loss": 0.73882067,
      "num_input_tokens_seen": 24189965,
      "step": 1130,
      "time_per_iteration": 3.0757689476013184
    },
    {
      "auxiliary_loss_clip": 0.01158117,
      "auxiliary_loss_mlp": 0.010592,
      "balance_loss_clip": 1.05734515,
      "balance_loss_mlp": 1.03481019,
      "epoch": 0.06799939876747332,
      "flos": 20631398947200.0,
      "grad_norm": 2.24421862356218,
      "language_loss": 0.80776262,
      "learning_rate": 3.984872909471688e-06,
      "loss": 0.82993579,
      "num_input_tokens_seen": 24208045,
      "step": 1131,
      "time_per_iteration": 5.00832724571228
    },
    {
      "auxiliary_loss_clip": 0.01195331,
      "auxiliary_loss_mlp": 0.01070142,
      "balance_loss_clip": 1.06155944,
      "balance_loss_mlp": 1.04614532,
      "epoch": 0.06805952202014129,
      "flos": 14864755829760.0,
      "grad_norm": 2.0533244923502463,
      "language_loss": 0.80371779,
      "learning_rate": 3.984825061735701e-06,
      "loss": 0.8263725,
      "num_input_tokens_seen": 24223805,
      "step": 1132,
      "time_per_iteration": 4.487931251525879
    },
    {
      "auxiliary_loss_clip": 0.01170581,
      "auxiliary_loss_mlp": 0.01061867,
      "balance_loss_clip": 1.05438542,
      "balance_loss_mlp": 1.03756022,
      "epoch": 0.06811964527280925,
      "flos": 48909434947200.0,
      "grad_norm": 1.7182324226465766,
      "language_loss": 0.6341064,
      "learning_rate": 3.9847771387349495e-06,
      "loss": 0.65643084,
      "num_input_tokens_seen": 24249475,
      "step": 1133,
      "time_per_iteration": 4.48089337348938
    },
    {
      "auxiliary_loss_clip": 0.01125599,
      "auxiliary_loss_mlp": 0.01055984,
      "balance_loss_clip": 1.04700482,
      "balance_loss_mlp": 1.02973366,
      "epoch": 0.06817976852547723,
      "flos": 15377273038080.0,
      "grad_norm": 1.9264963116598819,
      "language_loss": 0.74771935,
      "learning_rate": 3.9847291404712506e-06,
      "loss": 0.76953518,
      "num_input_tokens_seen": 24267980,
      "step": 1134,
      "time_per_iteration": 5.287277936935425
    },
    {
      "auxiliary_loss_clip": 0.01169269,
      "auxiliary_loss_mlp": 0.00782536,
      "balance_loss_clip": 1.05878353,
      "balance_loss_mlp": 1.00042605,
      "epoch": 0.0682398917781452,
      "flos": 20155690200960.0,
      "grad_norm": 2.151108605399924,
      "language_loss": 0.86871451,
      "learning_rate": 3.984681066946423e-06,
      "loss": 0.88823259,
      "num_input_tokens_seen": 24286805,
      "step": 1135,
      "time_per_iteration": 2.8024110794067383
    },
    {
      "auxiliary_loss_clip": 0.0117656,
      "auxiliary_loss_mlp": 0.007818,
      "balance_loss_clip": 1.0543226,
      "balance_loss_mlp": 1.00046515,
      "epoch": 0.06830001503081316,
      "flos": 23440618748160.0,
      "grad_norm": 2.521942237810997,
      "language_loss": 0.78131735,
      "learning_rate": 3.984632918162291e-06,
      "loss": 0.80090094,
      "num_input_tokens_seen": 24305855,
      "step": 1136,
      "time_per_iteration": 2.7595040798187256
    },
    {
      "auxiliary_loss_clip": 0.01185832,
      "auxiliary_loss_mlp": 0.01063587,
      "balance_loss_clip": 1.05952621,
      "balance_loss_mlp": 1.03868449,
      "epoch": 0.06836013828348114,
      "flos": 34349813153280.0,
      "grad_norm": 2.275643110468061,
      "language_loss": 0.83968467,
      "learning_rate": 3.984584694120679e-06,
      "loss": 0.86217892,
      "num_input_tokens_seen": 24326535,
      "step": 1137,
      "time_per_iteration": 2.7738285064697266
    },
    {
      "auxiliary_loss_clip": 0.01153105,
      "auxiliary_loss_mlp": 0.01059471,
      "balance_loss_clip": 1.05239427,
      "balance_loss_mlp": 1.0348897,
      "epoch": 0.06842026153614911,
      "flos": 23148844571520.0,
      "grad_norm": 2.068206081593879,
      "language_loss": 0.788486,
      "learning_rate": 3.984536394823418e-06,
      "loss": 0.81061178,
      "num_input_tokens_seen": 24345810,
      "step": 1138,
      "time_per_iteration": 2.804537296295166
    },
    {
      "auxiliary_loss_clip": 0.01209658,
      "auxiliary_loss_mlp": 0.01058353,
      "balance_loss_clip": 1.06288362,
      "balance_loss_mlp": 1.03415346,
      "epoch": 0.06848038478881707,
      "flos": 24608972430720.0,
      "grad_norm": 2.3335265924104096,
      "language_loss": 0.85507643,
      "learning_rate": 3.984488020272336e-06,
      "loss": 0.87775654,
      "num_input_tokens_seen": 24366095,
      "step": 1139,
      "time_per_iteration": 2.746884822845459
    },
    {
      "auxiliary_loss_clip": 0.01153855,
      "auxiliary_loss_mlp": 0.01063721,
      "balance_loss_clip": 1.05325532,
      "balance_loss_mlp": 1.03679228,
      "epoch": 0.06854050804148504,
      "flos": 40880994278400.0,
      "grad_norm": 1.9254794009430078,
      "language_loss": 0.74899161,
      "learning_rate": 3.984439570469271e-06,
      "loss": 0.7711674,
      "num_input_tokens_seen": 24388665,
      "step": 1140,
      "time_per_iteration": 2.938143253326416
    },
    {
      "auxiliary_loss_clip": 0.01186218,
      "auxiliary_loss_mlp": 0.00782227,
      "balance_loss_clip": 1.06101704,
      "balance_loss_mlp": 1.00036597,
      "epoch": 0.06860063129415302,
      "flos": 31686354743040.0,
      "grad_norm": 2.1250887020504767,
      "language_loss": 0.68258876,
      "learning_rate": 3.9843910454160574e-06,
      "loss": 0.70227319,
      "num_input_tokens_seen": 24407705,
      "step": 1141,
      "time_per_iteration": 2.8180530071258545
    },
    {
      "auxiliary_loss_clip": 0.01197117,
      "auxiliary_loss_mlp": 0.01067748,
      "balance_loss_clip": 1.05978489,
      "balance_loss_mlp": 1.04266596,
      "epoch": 0.06866075454682098,
      "flos": 26542007775360.0,
      "grad_norm": 1.8460768582410394,
      "language_loss": 0.78959155,
      "learning_rate": 3.984342445114538e-06,
      "loss": 0.81224018,
      "num_input_tokens_seen": 24428390,
      "step": 1142,
      "time_per_iteration": 2.712876558303833
    },
    {
      "auxiliary_loss_clip": 0.01186915,
      "auxiliary_loss_mlp": 0.01060882,
      "balance_loss_clip": 1.06245089,
      "balance_loss_mlp": 1.03702831,
      "epoch": 0.06872087779948895,
      "flos": 29789768724480.0,
      "grad_norm": 1.7867268614306446,
      "language_loss": 0.68287402,
      "learning_rate": 3.984293769566553e-06,
      "loss": 0.70535195,
      "num_input_tokens_seen": 24450810,
      "step": 1143,
      "time_per_iteration": 2.752659320831299
    },
    {
      "auxiliary_loss_clip": 0.01177843,
      "auxiliary_loss_mlp": 0.01059894,
      "balance_loss_clip": 1.05798244,
      "balance_loss_mlp": 1.03773308,
      "epoch": 0.06878100105215693,
      "flos": 26941118768640.0,
      "grad_norm": 1.7582250309313294,
      "language_loss": 0.74307454,
      "learning_rate": 3.98424501877395e-06,
      "loss": 0.76545191,
      "num_input_tokens_seen": 24469965,
      "step": 1144,
      "time_per_iteration": 2.6448662281036377
    },
    {
      "auxiliary_loss_clip": 0.01189197,
      "auxiliary_loss_mlp": 0.0106544,
      "balance_loss_clip": 1.0565474,
      "balance_loss_mlp": 1.04039407,
      "epoch": 0.06884112430482489,
      "flos": 10670748946560.0,
      "grad_norm": 2.699041414372958,
      "language_loss": 0.91755033,
      "learning_rate": 3.984196192738577e-06,
      "loss": 0.94009674,
      "num_input_tokens_seen": 24486370,
      "step": 1145,
      "time_per_iteration": 2.6621482372283936
    },
    {
      "auxiliary_loss_clip": 0.01212189,
      "auxiliary_loss_mlp": 0.0106819,
      "balance_loss_clip": 1.06225932,
      "balance_loss_mlp": 1.04258406,
      "epoch": 0.06890124755749286,
      "flos": 20193647898240.0,
      "grad_norm": 2.2014676012481487,
      "language_loss": 0.81726635,
      "learning_rate": 3.984147291462285e-06,
      "loss": 0.84007025,
      "num_input_tokens_seen": 24503780,
      "step": 1146,
      "time_per_iteration": 2.623964548110962
    },
    {
      "auxiliary_loss_clip": 0.01204602,
      "auxiliary_loss_mlp": 0.01065301,
      "balance_loss_clip": 1.06215203,
      "balance_loss_mlp": 1.04191244,
      "epoch": 0.06896137081016084,
      "flos": 20449224144000.0,
      "grad_norm": 2.1265245828428108,
      "language_loss": 0.84968954,
      "learning_rate": 3.98409831494693e-06,
      "loss": 0.8723886,
      "num_input_tokens_seen": 24522320,
      "step": 1147,
      "time_per_iteration": 2.5898265838623047
    },
    {
      "auxiliary_loss_clip": 0.01156886,
      "auxiliary_loss_mlp": 0.01064453,
      "balance_loss_clip": 1.05563867,
      "balance_loss_mlp": 1.03949046,
      "epoch": 0.0690214940628288,
      "flos": 18368703555840.0,
      "grad_norm": 1.7557033260323716,
      "language_loss": 0.86094105,
      "learning_rate": 3.984049263194367e-06,
      "loss": 0.88315445,
      "num_input_tokens_seen": 24540445,
      "step": 1148,
      "time_per_iteration": 2.748782157897949
    },
    {
      "auxiliary_loss_clip": 0.01173365,
      "auxiliary_loss_mlp": 0.01060047,
      "balance_loss_clip": 1.05569541,
      "balance_loss_mlp": 1.03370178,
      "epoch": 0.06908161731549677,
      "flos": 20558033418240.0,
      "grad_norm": 2.322434023005448,
      "language_loss": 0.69602191,
      "learning_rate": 3.9840001362064575e-06,
      "loss": 0.71835601,
      "num_input_tokens_seen": 24557105,
      "step": 1149,
      "time_per_iteration": 2.741854429244995
    },
    {
      "auxiliary_loss_clip": 0.01207871,
      "auxiliary_loss_mlp": 0.01051245,
      "balance_loss_clip": 1.06034219,
      "balance_loss_mlp": 1.02692604,
      "epoch": 0.06914174056816474,
      "flos": 27563666313600.0,
      "grad_norm": 1.9440351937259064,
      "language_loss": 0.8374452,
      "learning_rate": 3.983950933985064e-06,
      "loss": 0.86003637,
      "num_input_tokens_seen": 24578240,
      "step": 1150,
      "time_per_iteration": 2.6919586658477783
    },
    {
      "auxiliary_loss_clip": 0.01181406,
      "auxiliary_loss_mlp": 0.01058015,
      "balance_loss_clip": 1.06063652,
      "balance_loss_mlp": 1.03380394,
      "epoch": 0.06920186382083271,
      "flos": 15304015249920.0,
      "grad_norm": 4.11905785776886,
      "language_loss": 0.81464434,
      "learning_rate": 3.983901656532052e-06,
      "loss": 0.83703858,
      "num_input_tokens_seen": 24593585,
      "step": 1151,
      "time_per_iteration": 2.7979934215545654
    },
    {
      "auxiliary_loss_clip": 0.01206831,
      "auxiliary_loss_mlp": 0.01058184,
      "balance_loss_clip": 1.06409955,
      "balance_loss_mlp": 1.03434169,
      "epoch": 0.06926198707350067,
      "flos": 25191227894400.0,
      "grad_norm": 2.0324362571668724,
      "language_loss": 0.85408235,
      "learning_rate": 3.983852303849291e-06,
      "loss": 0.87673247,
      "num_input_tokens_seen": 24613110,
      "step": 1152,
      "time_per_iteration": 2.686021089553833
    },
    {
      "auxiliary_loss_clip": 0.01190935,
      "auxiliary_loss_mlp": 0.01062076,
      "balance_loss_clip": 1.06250155,
      "balance_loss_mlp": 1.03866374,
      "epoch": 0.06932211032616864,
      "flos": 13256137146240.0,
      "grad_norm": 2.182544196511779,
      "language_loss": 0.90594423,
      "learning_rate": 3.983802875938651e-06,
      "loss": 0.92847437,
      "num_input_tokens_seen": 24628795,
      "step": 1153,
      "time_per_iteration": 2.58366060256958
    },
    {
      "auxiliary_loss_clip": 0.01169877,
      "auxiliary_loss_mlp": 0.01055253,
      "balance_loss_clip": 1.05681062,
      "balance_loss_mlp": 1.03088629,
      "epoch": 0.06938223357883662,
      "flos": 24827381078400.0,
      "grad_norm": 2.1214794624630846,
      "language_loss": 0.81526846,
      "learning_rate": 3.983753372802008e-06,
      "loss": 0.83751976,
      "num_input_tokens_seen": 24645480,
      "step": 1154,
      "time_per_iteration": 2.696794271469116
    },
    {
      "auxiliary_loss_clip": 0.01188774,
      "auxiliary_loss_mlp": 0.01066335,
      "balance_loss_clip": 1.0691216,
      "balance_loss_mlp": 1.04200506,
      "epoch": 0.06944235683150458,
      "flos": 27267977554560.0,
      "grad_norm": 2.102018399986892,
      "language_loss": 0.75022292,
      "learning_rate": 3.983703794441237e-06,
      "loss": 0.77277398,
      "num_input_tokens_seen": 24664630,
      "step": 1155,
      "time_per_iteration": 2.7718143463134766
    },
    {
      "auxiliary_loss_clip": 0.01180696,
      "auxiliary_loss_mlp": 0.00782152,
      "balance_loss_clip": 1.05586052,
      "balance_loss_mlp": 1.00041056,
      "epoch": 0.06950248008417255,
      "flos": 25808065176960.0,
      "grad_norm": 1.7459449483933205,
      "language_loss": 0.7110405,
      "learning_rate": 3.98365414085822e-06,
      "loss": 0.73066902,
      "num_input_tokens_seen": 24684210,
      "step": 1156,
      "time_per_iteration": 2.7014200687408447
    },
    {
      "auxiliary_loss_clip": 0.01179101,
      "auxiliary_loss_mlp": 0.00782674,
      "balance_loss_clip": 1.0593586,
      "balance_loss_mlp": 1.00037348,
      "epoch": 0.06956260333684053,
      "flos": 22271546793600.0,
      "grad_norm": 2.067241397655847,
      "language_loss": 0.74882817,
      "learning_rate": 3.98360441205484e-06,
      "loss": 0.76844591,
      "num_input_tokens_seen": 24702490,
      "step": 1157,
      "time_per_iteration": 2.7571897506713867
    },
    {
      "auxiliary_loss_clip": 0.01178249,
      "auxiliary_loss_mlp": 0.01061737,
      "balance_loss_clip": 1.05653787,
      "balance_loss_mlp": 1.03697729,
      "epoch": 0.0696227265895085,
      "flos": 29681390413440.0,
      "grad_norm": 1.9827644507913538,
      "language_loss": 0.7165724,
      "learning_rate": 3.983554608032982e-06,
      "loss": 0.73897225,
      "num_input_tokens_seen": 24724340,
      "step": 1158,
      "time_per_iteration": 2.839745044708252
    },
    {
      "auxiliary_loss_clip": 0.01207855,
      "auxiliary_loss_mlp": 0.01058558,
      "balance_loss_clip": 1.0605582,
      "balance_loss_mlp": 1.03370285,
      "epoch": 0.06968284984217646,
      "flos": 25523545547520.0,
      "grad_norm": 1.9692207215605615,
      "language_loss": 0.79595017,
      "learning_rate": 3.983504728794533e-06,
      "loss": 0.8186143,
      "num_input_tokens_seen": 24745550,
      "step": 1159,
      "time_per_iteration": 2.7535817623138428
    },
    {
      "auxiliary_loss_clip": 0.01212717,
      "auxiliary_loss_mlp": 0.01068535,
      "balance_loss_clip": 1.06535673,
      "balance_loss_mlp": 1.04094958,
      "epoch": 0.06974297309484444,
      "flos": 20698192287360.0,
      "grad_norm": 3.5530789367722373,
      "language_loss": 0.80517769,
      "learning_rate": 3.983454774341387e-06,
      "loss": 0.82799017,
      "num_input_tokens_seen": 24762575,
      "step": 1160,
      "time_per_iteration": 2.7455785274505615
    },
    {
      "auxiliary_loss_clip": 0.0119075,
      "auxiliary_loss_mlp": 0.01057887,
      "balance_loss_clip": 1.05680609,
      "balance_loss_mlp": 1.03294837,
      "epoch": 0.0698030963475124,
      "flos": 26505199313280.0,
      "grad_norm": 1.6303409062485206,
      "language_loss": 0.7607069,
      "learning_rate": 3.983404744675437e-06,
      "loss": 0.78319323,
      "num_input_tokens_seen": 24782605,
      "step": 1161,
      "time_per_iteration": 2.773775100708008
    },
    {
      "auxiliary_loss_clip": 0.01175787,
      "auxiliary_loss_mlp": 0.01062083,
      "balance_loss_clip": 1.05773759,
      "balance_loss_mlp": 1.03673923,
      "epoch": 0.06986321960018037,
      "flos": 23040430346880.0,
      "grad_norm": 1.6605796421434038,
      "language_loss": 0.82758528,
      "learning_rate": 3.9833546397985794e-06,
      "loss": 0.84996402,
      "num_input_tokens_seen": 24802910,
      "step": 1162,
      "time_per_iteration": 2.7426044940948486
    },
    {
      "auxiliary_loss_clip": 0.01182513,
      "auxiliary_loss_mlp": 0.01058124,
      "balance_loss_clip": 1.05717576,
      "balance_loss_mlp": 1.03092098,
      "epoch": 0.06992334285284833,
      "flos": 28584822061440.0,
      "grad_norm": 1.9523155091610094,
      "language_loss": 0.79563475,
      "learning_rate": 3.983304459712716e-06,
      "loss": 0.81804121,
      "num_input_tokens_seen": 24823305,
      "step": 1163,
      "time_per_iteration": 2.720947742462158
    },
    {
      "auxiliary_loss_clip": 0.01190519,
      "auxiliary_loss_mlp": 0.01063375,
      "balance_loss_clip": 1.05861616,
      "balance_loss_mlp": 1.03722012,
      "epoch": 0.06998346610551631,
      "flos": 20595344670720.0,
      "grad_norm": 2.213365660843382,
      "language_loss": 0.79187214,
      "learning_rate": 3.983254204419749e-06,
      "loss": 0.81441104,
      "num_input_tokens_seen": 24842155,
      "step": 1164,
      "time_per_iteration": 2.6554183959960938
    },
    {
      "auxiliary_loss_clip": 0.01143916,
      "auxiliary_loss_mlp": 0.01067459,
      "balance_loss_clip": 1.05240798,
      "balance_loss_mlp": 1.03875315,
      "epoch": 0.07004358935818428,
      "flos": 22528810978560.0,
      "grad_norm": 1.421930435008642,
      "language_loss": 0.72855628,
      "learning_rate": 3.983203873921583e-06,
      "loss": 0.75067008,
      "num_input_tokens_seen": 24862080,
      "step": 1165,
      "time_per_iteration": 2.753063440322876
    },
    {
      "auxiliary_loss_clip": 0.01183824,
      "auxiliary_loss_mlp": 0.01059612,
      "balance_loss_clip": 1.06135893,
      "balance_loss_mlp": 1.03522193,
      "epoch": 0.07010371261085224,
      "flos": 28949997680640.0,
      "grad_norm": 2.453348821242437,
      "language_loss": 0.81136239,
      "learning_rate": 3.983153468220128e-06,
      "loss": 0.83379674,
      "num_input_tokens_seen": 24886165,
      "step": 1166,
      "time_per_iteration": 2.802016496658325
    },
    {
      "auxiliary_loss_clip": 0.011718,
      "auxiliary_loss_mlp": 0.01053529,
      "balance_loss_clip": 1.05450797,
      "balance_loss_mlp": 1.02754176,
      "epoch": 0.07016383586352022,
      "flos": 23659171050240.0,
      "grad_norm": 2.457667377154448,
      "language_loss": 0.84640259,
      "learning_rate": 3.983102987317295e-06,
      "loss": 0.86865586,
      "num_input_tokens_seen": 24905775,
      "step": 1167,
      "time_per_iteration": 2.7066097259521484
    },
    {
      "auxiliary_loss_clip": 0.01193446,
      "auxiliary_loss_mlp": 0.01064209,
      "balance_loss_clip": 1.06136739,
      "balance_loss_mlp": 1.03887713,
      "epoch": 0.07022395911618819,
      "flos": 19792130693760.0,
      "grad_norm": 2.6158204436543,
      "language_loss": 0.89524722,
      "learning_rate": 3.983052431214997e-06,
      "loss": 0.91782373,
      "num_input_tokens_seen": 24924295,
      "step": 1168,
      "time_per_iteration": 2.6258392333984375
    },
    {
      "auxiliary_loss_clip": 0.01190821,
      "auxiliary_loss_mlp": 0.01065905,
      "balance_loss_clip": 1.06090224,
      "balance_loss_mlp": 1.03705645,
      "epoch": 0.07028408236885615,
      "flos": 21689147675520.0,
      "grad_norm": 2.6445150319591035,
      "language_loss": 0.89008862,
      "learning_rate": 3.983001799915153e-06,
      "loss": 0.91265589,
      "num_input_tokens_seen": 24943210,
      "step": 1169,
      "time_per_iteration": 2.6858527660369873
    },
    {
      "auxiliary_loss_clip": 0.01211063,
      "auxiliary_loss_mlp": 0.01065533,
      "balance_loss_clip": 1.06400895,
      "balance_loss_mlp": 1.03950977,
      "epoch": 0.07034420562152413,
      "flos": 25630271832960.0,
      "grad_norm": 1.9672897290124218,
      "language_loss": 0.83834457,
      "learning_rate": 3.982951093419681e-06,
      "loss": 0.86111057,
      "num_input_tokens_seen": 24960360,
      "step": 1170,
      "time_per_iteration": 2.6278069019317627
    },
    {
      "auxiliary_loss_clip": 0.01180333,
      "auxiliary_loss_mlp": 0.00782328,
      "balance_loss_clip": 1.0613637,
      "balance_loss_mlp": 1.00041986,
      "epoch": 0.0704043288741921,
      "flos": 20810449267200.0,
      "grad_norm": 1.8542795171503503,
      "language_loss": 0.75687242,
      "learning_rate": 3.982900311730506e-06,
      "loss": 0.77649903,
      "num_input_tokens_seen": 24978290,
      "step": 1171,
      "time_per_iteration": 5.806530475616455
    },
    {
      "auxiliary_loss_clip": 0.01179645,
      "auxiliary_loss_mlp": 0.0106394,
      "balance_loss_clip": 1.06133175,
      "balance_loss_mlp": 1.03919196,
      "epoch": 0.07046445212686006,
      "flos": 25593176062080.0,
      "grad_norm": 2.482864122539831,
      "language_loss": 0.88865125,
      "learning_rate": 3.9828494548495514e-06,
      "loss": 0.91108704,
      "num_input_tokens_seen": 24997055,
      "step": 1172,
      "time_per_iteration": 4.371561288833618
    },
    {
      "auxiliary_loss_clip": 0.01197698,
      "auxiliary_loss_mlp": 0.01054991,
      "balance_loss_clip": 1.06532764,
      "balance_loss_mlp": 1.02858603,
      "epoch": 0.07052457537952803,
      "flos": 25556978131200.0,
      "grad_norm": 1.6816354314161714,
      "language_loss": 0.82075119,
      "learning_rate": 3.982798522778748e-06,
      "loss": 0.84327805,
      "num_input_tokens_seen": 25017490,
      "step": 1173,
      "time_per_iteration": 4.611542463302612
    },
    {
      "auxiliary_loss_clip": 0.01200886,
      "auxiliary_loss_mlp": 0.01060851,
      "balance_loss_clip": 1.06317592,
      "balance_loss_mlp": 1.03503036,
      "epoch": 0.070584698632196,
      "flos": 17968515154560.0,
      "grad_norm": 2.007232853627583,
      "language_loss": 0.82071686,
      "learning_rate": 3.9827475155200245e-06,
      "loss": 0.8433342,
      "num_input_tokens_seen": 25035660,
      "step": 1174,
      "time_per_iteration": 2.6334969997406006
    },
    {
      "auxiliary_loss_clip": 0.01180907,
      "auxiliary_loss_mlp": 0.01059972,
      "balance_loss_clip": 1.05857778,
      "balance_loss_mlp": 1.03473568,
      "epoch": 0.07064482188486397,
      "flos": 25370888745600.0,
      "grad_norm": 2.09222115072597,
      "language_loss": 0.85013211,
      "learning_rate": 3.982696433075317e-06,
      "loss": 0.87254095,
      "num_input_tokens_seen": 25054785,
      "step": 1175,
      "time_per_iteration": 2.861591339111328
    },
    {
      "auxiliary_loss_clip": 0.01196955,
      "auxiliary_loss_mlp": 0.01069941,
      "balance_loss_clip": 1.06447482,
      "balance_loss_mlp": 1.04605186,
      "epoch": 0.07070494513753194,
      "flos": 24899848767360.0,
      "grad_norm": 1.7270820646539309,
      "language_loss": 0.83103871,
      "learning_rate": 3.982645275446563e-06,
      "loss": 0.85370767,
      "num_input_tokens_seen": 25075180,
      "step": 1176,
      "time_per_iteration": 2.754521608352661
    },
    {
      "auxiliary_loss_clip": 0.01152261,
      "auxiliary_loss_mlp": 0.01062154,
      "balance_loss_clip": 1.05370057,
      "balance_loss_mlp": 1.0352838,
      "epoch": 0.07076506839019991,
      "flos": 22338447874560.0,
      "grad_norm": 3.4939498355716996,
      "language_loss": 0.74409902,
      "learning_rate": 3.982594042635701e-06,
      "loss": 0.7662431,
      "num_input_tokens_seen": 25093035,
      "step": 1177,
      "time_per_iteration": 2.692426919937134
    },
    {
      "auxiliary_loss_clip": 0.01188551,
      "auxiliary_loss_mlp": 0.0106394,
      "balance_loss_clip": 1.06080353,
      "balance_loss_mlp": 1.03801203,
      "epoch": 0.07082519164286788,
      "flos": 18660800954880.0,
      "grad_norm": 1.8240190288677762,
      "language_loss": 0.85965598,
      "learning_rate": 3.982542734644673e-06,
      "loss": 0.88218087,
      "num_input_tokens_seen": 25112520,
      "step": 1178,
      "time_per_iteration": 2.7197048664093018
    },
    {
      "auxiliary_loss_clip": 0.01082521,
      "auxiliary_loss_mlp": 0.01013999,
      "balance_loss_clip": 1.03661168,
      "balance_loss_mlp": 1.01023197,
      "epoch": 0.07088531489553584,
      "flos": 63654107610240.0,
      "grad_norm": 0.8453670789764802,
      "language_loss": 0.63256603,
      "learning_rate": 3.982491351475427e-06,
      "loss": 0.65353125,
      "num_input_tokens_seen": 25177760,
      "step": 1179,
      "time_per_iteration": 3.3419978618621826
    },
    {
      "auxiliary_loss_clip": 0.01211274,
      "auxiliary_loss_mlp": 0.01073372,
      "balance_loss_clip": 1.06935215,
      "balance_loss_mlp": 1.04858887,
      "epoch": 0.07094543814820382,
      "flos": 21572688804480.0,
      "grad_norm": 3.2714198066984177,
      "language_loss": 0.83388901,
      "learning_rate": 3.98243989312991e-06,
      "loss": 0.85673553,
      "num_input_tokens_seen": 25195260,
      "step": 1180,
      "time_per_iteration": 2.631992816925049
    },
    {
      "auxiliary_loss_clip": 0.01182661,
      "auxiliary_loss_mlp": 0.01071326,
      "balance_loss_clip": 1.06119037,
      "balance_loss_mlp": 1.04624391,
      "epoch": 0.07100556140087179,
      "flos": 22089946608000.0,
      "grad_norm": 2.0409456536886386,
      "language_loss": 0.88649988,
      "learning_rate": 3.982388359610074e-06,
      "loss": 0.90903974,
      "num_input_tokens_seen": 25212740,
      "step": 1181,
      "time_per_iteration": 2.696789264678955
    },
    {
      "auxiliary_loss_clip": 0.01180377,
      "auxiliary_loss_mlp": 0.01070036,
      "balance_loss_clip": 1.06187141,
      "balance_loss_mlp": 1.04516935,
      "epoch": 0.07106568465353975,
      "flos": 47922286400640.0,
      "grad_norm": 1.8294049229574356,
      "language_loss": 0.83244783,
      "learning_rate": 3.9823367509178725e-06,
      "loss": 0.85495198,
      "num_input_tokens_seen": 25236420,
      "step": 1182,
      "time_per_iteration": 2.9415605068206787
    },
    {
      "auxiliary_loss_clip": 0.01193669,
      "auxiliary_loss_mlp": 0.01067019,
      "balance_loss_clip": 1.0641923,
      "balance_loss_mlp": 1.04150808,
      "epoch": 0.07112580790620772,
      "flos": 23440798316160.0,
      "grad_norm": 3.5892595189310903,
      "language_loss": 0.79067838,
      "learning_rate": 3.982285067055262e-06,
      "loss": 0.81328523,
      "num_input_tokens_seen": 25255120,
      "step": 1183,
      "time_per_iteration": 2.7284862995147705
    },
    {
      "auxiliary_loss_clip": 0.01211976,
      "auxiliary_loss_mlp": 0.01064792,
      "balance_loss_clip": 1.06126475,
      "balance_loss_mlp": 1.03866172,
      "epoch": 0.0711859311588757,
      "flos": 31868888682240.0,
      "grad_norm": 2.5463322111759354,
      "language_loss": 0.788867,
      "learning_rate": 3.982233308024204e-06,
      "loss": 0.81163466,
      "num_input_tokens_seen": 25275150,
      "step": 1184,
      "time_per_iteration": 2.7531635761260986
    },
    {
      "auxiliary_loss_clip": 0.01152059,
      "auxiliary_loss_mlp": 0.01062006,
      "balance_loss_clip": 1.05961919,
      "balance_loss_mlp": 1.03752065,
      "epoch": 0.07124605441154366,
      "flos": 19610315026560.0,
      "grad_norm": 1.904751850318294,
      "language_loss": 0.76806915,
      "learning_rate": 3.98218147382666e-06,
      "loss": 0.79020983,
      "num_input_tokens_seen": 25293680,
      "step": 1185,
      "time_per_iteration": 2.732539176940918
    },
    {
      "auxiliary_loss_clip": 0.01208288,
      "auxiliary_loss_mlp": 0.01073792,
      "balance_loss_clip": 1.06328642,
      "balance_loss_mlp": 1.04903185,
      "epoch": 0.07130617766421163,
      "flos": 14684448533760.0,
      "grad_norm": 2.1301142092644696,
      "language_loss": 0.65472758,
      "learning_rate": 3.982129564464596e-06,
      "loss": 0.67754835,
      "num_input_tokens_seen": 25310050,
      "step": 1186,
      "time_per_iteration": 2.757812261581421
    },
    {
      "auxiliary_loss_clip": 0.01195497,
      "auxiliary_loss_mlp": 0.01057322,
      "balance_loss_clip": 1.06479859,
      "balance_loss_mlp": 1.03274107,
      "epoch": 0.07136630091687961,
      "flos": 26067915141120.0,
      "grad_norm": 2.1671481434625894,
      "language_loss": 0.69743419,
      "learning_rate": 3.98207757993998e-06,
      "loss": 0.71996236,
      "num_input_tokens_seen": 25331020,
      "step": 1187,
      "time_per_iteration": 2.746615409851074
    },
    {
      "auxiliary_loss_clip": 0.01151827,
      "auxiliary_loss_mlp": 0.01067347,
      "balance_loss_clip": 1.05412316,
      "balance_loss_mlp": 1.04367232,
      "epoch": 0.07142642416954757,
      "flos": 15669190869120.0,
      "grad_norm": 2.8037131445876597,
      "language_loss": 0.7861973,
      "learning_rate": 3.9820255202547845e-06,
      "loss": 0.80838895,
      "num_input_tokens_seen": 25347875,
      "step": 1188,
      "time_per_iteration": 2.738281726837158
    },
    {
      "auxiliary_loss_clip": 0.01203626,
      "auxiliary_loss_mlp": 0.01059966,
      "balance_loss_clip": 1.06304908,
      "balance_loss_mlp": 1.03530121,
      "epoch": 0.07148654742221554,
      "flos": 19755322231680.0,
      "grad_norm": 1.8909260147246576,
      "language_loss": 0.84754103,
      "learning_rate": 3.981973385410981e-06,
      "loss": 0.87017697,
      "num_input_tokens_seen": 25366715,
      "step": 1189,
      "time_per_iteration": 2.5770246982574463
    },
    {
      "auxiliary_loss_clip": 0.01173135,
      "auxiliary_loss_mlp": 0.0078213,
      "balance_loss_clip": 1.06234396,
      "balance_loss_mlp": 1.00041807,
      "epoch": 0.07154667067488352,
      "flos": 23471824688640.0,
      "grad_norm": 5.212083930118342,
      "language_loss": 0.76932275,
      "learning_rate": 3.9819211754105494e-06,
      "loss": 0.78887534,
      "num_input_tokens_seen": 25385450,
      "step": 1190,
      "time_per_iteration": 2.7057712078094482
    },
    {
      "auxiliary_loss_clip": 0.01208346,
      "auxiliary_loss_mlp": 0.01074705,
      "balance_loss_clip": 1.06283545,
      "balance_loss_mlp": 1.04751348,
      "epoch": 0.07160679392755148,
      "flos": 18332936588160.0,
      "grad_norm": 2.5312098602102084,
      "language_loss": 0.75201792,
      "learning_rate": 3.981868890255468e-06,
      "loss": 0.7748484,
      "num_input_tokens_seen": 25403940,
      "step": 1191,
      "time_per_iteration": 2.6071674823760986
    },
    {
      "auxiliary_loss_clip": 0.01162268,
      "auxiliary_loss_mlp": 0.01063437,
      "balance_loss_clip": 1.0519917,
      "balance_loss_mlp": 1.03649545,
      "epoch": 0.07166691718021945,
      "flos": 17747017937280.0,
      "grad_norm": 2.470839013019174,
      "language_loss": 0.74334443,
      "learning_rate": 3.981816529947719e-06,
      "loss": 0.76560152,
      "num_input_tokens_seen": 25420410,
      "step": 1192,
      "time_per_iteration": 2.661078453063965
    },
    {
      "auxiliary_loss_clip": 0.01202036,
      "auxiliary_loss_mlp": 0.01054727,
      "balance_loss_clip": 1.05904579,
      "balance_loss_mlp": 1.03099298,
      "epoch": 0.07172704043288743,
      "flos": 22451925916800.0,
      "grad_norm": 2.443309122344248,
      "language_loss": 0.78010541,
      "learning_rate": 3.9817640944892896e-06,
      "loss": 0.8026731,
      "num_input_tokens_seen": 25439415,
      "step": 1193,
      "time_per_iteration": 2.5603158473968506
    },
    {
      "auxiliary_loss_clip": 0.01186747,
      "auxiliary_loss_mlp": 0.01059465,
      "balance_loss_clip": 1.06358278,
      "balance_loss_mlp": 1.03319085,
      "epoch": 0.07178716368555539,
      "flos": 23222210100480.0,
      "grad_norm": 2.1011663585924585,
      "language_loss": 0.85497916,
      "learning_rate": 3.981711583882166e-06,
      "loss": 0.87744129,
      "num_input_tokens_seen": 25458715,
      "step": 1194,
      "time_per_iteration": 2.6819851398468018
    },
    {
      "auxiliary_loss_clip": 0.01184191,
      "auxiliary_loss_mlp": 0.01067737,
      "balance_loss_clip": 1.05706751,
      "balance_loss_mlp": 1.04135609,
      "epoch": 0.07184728693822336,
      "flos": 25150828072320.0,
      "grad_norm": 2.0205668140023185,
      "language_loss": 0.8183766,
      "learning_rate": 3.981658998128341e-06,
      "loss": 0.84089589,
      "num_input_tokens_seen": 25477985,
      "step": 1195,
      "time_per_iteration": 2.6646647453308105
    },
    {
      "auxiliary_loss_clip": 0.01165951,
      "auxiliary_loss_mlp": 0.01063438,
      "balance_loss_clip": 1.0578239,
      "balance_loss_mlp": 1.03976321,
      "epoch": 0.07190741019089132,
      "flos": 22711237176960.0,
      "grad_norm": 2.161995064372768,
      "language_loss": 0.80093575,
      "learning_rate": 3.981606337229808e-06,
      "loss": 0.82322967,
      "num_input_tokens_seen": 25497110,
      "step": 1196,
      "time_per_iteration": 2.7217979431152344
    },
    {
      "auxiliary_loss_clip": 0.01176131,
      "auxiliary_loss_mlp": 0.00784114,
      "balance_loss_clip": 1.06106043,
      "balance_loss_mlp": 1.00034249,
      "epoch": 0.0719675334435593,
      "flos": 29349791032320.0,
      "grad_norm": 2.5905261146074263,
      "language_loss": 0.71339291,
      "learning_rate": 3.9815536011885655e-06,
      "loss": 0.73299539,
      "num_input_tokens_seen": 25516555,
      "step": 1197,
      "time_per_iteration": 2.7931766510009766
    },
    {
      "auxiliary_loss_clip": 0.01157444,
      "auxiliary_loss_mlp": 0.01055247,
      "balance_loss_clip": 1.06130266,
      "balance_loss_mlp": 1.03074968,
      "epoch": 0.07202765669622727,
      "flos": 17639788861440.0,
      "grad_norm": 3.074283933156949,
      "language_loss": 0.85951984,
      "learning_rate": 3.98150079000661e-06,
      "loss": 0.88164675,
      "num_input_tokens_seen": 25533895,
      "step": 1198,
      "time_per_iteration": 2.7241532802581787
    },
    {
      "auxiliary_loss_clip": 0.01160083,
      "auxiliary_loss_mlp": 0.0106501,
      "balance_loss_clip": 1.0597434,
      "balance_loss_mlp": 1.03944004,
      "epoch": 0.07208777994889523,
      "flos": 21434038306560.0,
      "grad_norm": 2.052617638295489,
      "language_loss": 0.83840948,
      "learning_rate": 3.981447903685947e-06,
      "loss": 0.86066043,
      "num_input_tokens_seen": 25554195,
      "step": 1199,
      "time_per_iteration": 2.71362566947937
    },
    {
      "auxiliary_loss_clip": 0.01212755,
      "auxiliary_loss_mlp": 0.01060557,
      "balance_loss_clip": 1.06877887,
      "balance_loss_mlp": 1.03709614,
      "epoch": 0.07214790320156321,
      "flos": 26940867373440.0,
      "grad_norm": 3.1601590133124837,
      "language_loss": 0.7623595,
      "learning_rate": 3.981394942228581e-06,
      "loss": 0.78509259,
      "num_input_tokens_seen": 25574155,
      "step": 1200,
      "time_per_iteration": 2.6913061141967773
    },
    {
      "auxiliary_loss_clip": 0.0119008,
      "auxiliary_loss_mlp": 0.010701,
      "balance_loss_clip": 1.06442261,
      "balance_loss_mlp": 1.04487491,
      "epoch": 0.07220802645423118,
      "flos": 23879949995520.0,
      "grad_norm": 2.2017873087036226,
      "language_loss": 0.83013475,
      "learning_rate": 3.98134190563652e-06,
      "loss": 0.85273659,
      "num_input_tokens_seen": 25592735,
      "step": 1201,
      "time_per_iteration": 2.6983115673065186
    },
    {
      "auxiliary_loss_clip": 0.01196941,
      "auxiliary_loss_mlp": 0.01065672,
      "balance_loss_clip": 1.06197119,
      "balance_loss_mlp": 1.03952968,
      "epoch": 0.07226814970689914,
      "flos": 19243631036160.0,
      "grad_norm": 20.835065187143087,
      "language_loss": 0.68601412,
      "learning_rate": 3.981288793911775e-06,
      "loss": 0.70864022,
      "num_input_tokens_seen": 25611510,
      "step": 1202,
      "time_per_iteration": 2.691742420196533
    },
    {
      "auxiliary_loss_clip": 0.01182684,
      "auxiliary_loss_mlp": 0.00782201,
      "balance_loss_clip": 1.06256962,
      "balance_loss_mlp": 1.00038218,
      "epoch": 0.07232827295956712,
      "flos": 19172025273600.0,
      "grad_norm": 1.9661831136137597,
      "language_loss": 0.87487721,
      "learning_rate": 3.98123560705636e-06,
      "loss": 0.89452606,
      "num_input_tokens_seen": 25629560,
      "step": 1203,
      "time_per_iteration": 2.7832019329071045
    },
    {
      "auxiliary_loss_clip": 0.01154778,
      "auxiliary_loss_mlp": 0.01065748,
      "balance_loss_clip": 1.05210066,
      "balance_loss_mlp": 1.04065442,
      "epoch": 0.07238839621223508,
      "flos": 17639752947840.0,
      "grad_norm": 1.731721557525142,
      "language_loss": 0.78053147,
      "learning_rate": 3.981182345072293e-06,
      "loss": 0.80273676,
      "num_input_tokens_seen": 25648330,
      "step": 1204,
      "time_per_iteration": 2.7754547595977783
    },
    {
      "auxiliary_loss_clip": 0.01191832,
      "auxiliary_loss_mlp": 0.01065794,
      "balance_loss_clip": 1.06211591,
      "balance_loss_mlp": 1.04084373,
      "epoch": 0.07244851946490305,
      "flos": 28292401440000.0,
      "grad_norm": 1.5043252978087258,
      "language_loss": 0.82094097,
      "learning_rate": 3.981129007961593e-06,
      "loss": 0.84351724,
      "num_input_tokens_seen": 25669470,
      "step": 1205,
      "time_per_iteration": 2.680457353591919
    },
    {
      "auxiliary_loss_clip": 0.01180244,
      "auxiliary_loss_mlp": 0.00782807,
      "balance_loss_clip": 1.06221068,
      "balance_loss_mlp": 1.00036049,
      "epoch": 0.07250864271757101,
      "flos": 22564829341440.0,
      "grad_norm": 1.6438962430217685,
      "language_loss": 0.76715982,
      "learning_rate": 3.981075595726283e-06,
      "loss": 0.78679025,
      "num_input_tokens_seen": 25690470,
      "step": 1206,
      "time_per_iteration": 2.7028439044952393
    },
    {
      "auxiliary_loss_clip": 0.01188223,
      "auxiliary_loss_mlp": 0.01059861,
      "balance_loss_clip": 1.06262684,
      "balance_loss_mlp": 1.03442228,
      "epoch": 0.072568765970239,
      "flos": 21762405463680.0,
      "grad_norm": 1.9378198243304647,
      "language_loss": 0.77272987,
      "learning_rate": 3.981022108368387e-06,
      "loss": 0.79521072,
      "num_input_tokens_seen": 25709205,
      "step": 1207,
      "time_per_iteration": 2.779289960861206
    },
    {
      "auxiliary_loss_clip": 0.01185538,
      "auxiliary_loss_mlp": 0.01053693,
      "balance_loss_clip": 1.05844951,
      "balance_loss_mlp": 1.03062558,
      "epoch": 0.07262888922290696,
      "flos": 25519702792320.0,
      "grad_norm": 1.8716528383816402,
      "language_loss": 0.79480875,
      "learning_rate": 3.9809685458899345e-06,
      "loss": 0.81720108,
      "num_input_tokens_seen": 25728485,
      "step": 1208,
      "time_per_iteration": 2.682965040206909
    },
    {
      "auxiliary_loss_clip": 0.01184899,
      "auxiliary_loss_mlp": 0.01054862,
      "balance_loss_clip": 1.05801737,
      "balance_loss_mlp": 1.03198612,
      "epoch": 0.07268901247557492,
      "flos": 21246548290560.0,
      "grad_norm": 2.5612886109689765,
      "language_loss": 0.78537548,
      "learning_rate": 3.980914908292955e-06,
      "loss": 0.80777311,
      "num_input_tokens_seen": 25747730,
      "step": 1209,
      "time_per_iteration": 2.6582658290863037
    },
    {
      "auxiliary_loss_clip": 0.01191905,
      "auxiliary_loss_mlp": 0.01067741,
      "balance_loss_clip": 1.05931175,
      "balance_loss_mlp": 1.04408956,
      "epoch": 0.0727491357282429,
      "flos": 25479302970240.0,
      "grad_norm": 2.351303434522043,
      "language_loss": 0.80920583,
      "learning_rate": 3.980861195579486e-06,
      "loss": 0.83180225,
      "num_input_tokens_seen": 25768050,
      "step": 1210,
      "time_per_iteration": 4.241993427276611
    },
    {
      "auxiliary_loss_clip": 0.0117493,
      "auxiliary_loss_mlp": 0.01063711,
      "balance_loss_clip": 1.06087565,
      "balance_loss_mlp": 1.03891551,
      "epoch": 0.07280925898091087,
      "flos": 24462169545600.0,
      "grad_norm": 1.875347829314158,
      "language_loss": 0.84302205,
      "learning_rate": 3.98080740775156e-06,
      "loss": 0.86540848,
      "num_input_tokens_seen": 25787985,
      "step": 1211,
      "time_per_iteration": 4.289919853210449
    },
    {
      "auxiliary_loss_clip": 0.01162055,
      "auxiliary_loss_mlp": 0.01060218,
      "balance_loss_clip": 1.05356658,
      "balance_loss_mlp": 1.03629231,
      "epoch": 0.07286938223357883,
      "flos": 18288191220480.0,
      "grad_norm": 2.991110515222773,
      "language_loss": 0.90684664,
      "learning_rate": 3.98075354481122e-06,
      "loss": 0.92906934,
      "num_input_tokens_seen": 25803620,
      "step": 1212,
      "time_per_iteration": 2.660780906677246
    },
    {
      "auxiliary_loss_clip": 0.01202443,
      "auxiliary_loss_mlp": 0.01058817,
      "balance_loss_clip": 1.0623759,
      "balance_loss_mlp": 1.03490353,
      "epoch": 0.07292950548624681,
      "flos": 21214803646080.0,
      "grad_norm": 1.7918815842724805,
      "language_loss": 0.72358596,
      "learning_rate": 3.9806996067605055e-06,
      "loss": 0.74619853,
      "num_input_tokens_seen": 25823315,
      "step": 1213,
      "time_per_iteration": 4.303524017333984
    },
    {
      "auxiliary_loss_clip": 0.01153662,
      "auxiliary_loss_mlp": 0.01055706,
      "balance_loss_clip": 1.05658662,
      "balance_loss_mlp": 1.03089869,
      "epoch": 0.07298962873891478,
      "flos": 24642009964800.0,
      "grad_norm": 1.8655932637344164,
      "language_loss": 0.84356117,
      "learning_rate": 3.980645593601465e-06,
      "loss": 0.86565483,
      "num_input_tokens_seen": 25842605,
      "step": 1214,
      "time_per_iteration": 2.7505569458007812
    },
    {
      "auxiliary_loss_clip": 0.01208881,
      "auxiliary_loss_mlp": 0.01062075,
      "balance_loss_clip": 1.06484771,
      "balance_loss_mlp": 1.03723145,
      "epoch": 0.07304975199158274,
      "flos": 27052765217280.0,
      "grad_norm": 2.025651344907852,
      "language_loss": 0.84113681,
      "learning_rate": 3.980591505336144e-06,
      "loss": 0.86384636,
      "num_input_tokens_seen": 25863030,
      "step": 1215,
      "time_per_iteration": 2.7235965728759766
    },
    {
      "auxiliary_loss_clip": 0.01149957,
      "auxiliary_loss_mlp": 0.01062992,
      "balance_loss_clip": 1.05138278,
      "balance_loss_mlp": 1.03744531,
      "epoch": 0.07310987524425071,
      "flos": 33549544091520.0,
      "grad_norm": 1.9312816725096997,
      "language_loss": 0.80926049,
      "learning_rate": 3.980537341966595e-06,
      "loss": 0.83139002,
      "num_input_tokens_seen": 25888015,
      "step": 1216,
      "time_per_iteration": 2.9129130840301514
    },
    {
      "auxiliary_loss_clip": 0.01167944,
      "auxiliary_loss_mlp": 0.01060276,
      "balance_loss_clip": 1.05619049,
      "balance_loss_mlp": 1.03680408,
      "epoch": 0.07316999849691869,
      "flos": 28110944908800.0,
      "grad_norm": 3.2846247291101975,
      "language_loss": 0.75949144,
      "learning_rate": 3.980483103494872e-06,
      "loss": 0.78177369,
      "num_input_tokens_seen": 25908660,
      "step": 1217,
      "time_per_iteration": 2.7106521129608154
    },
    {
      "auxiliary_loss_clip": 0.01169026,
      "auxiliary_loss_mlp": 0.01056631,
      "balance_loss_clip": 1.06182647,
      "balance_loss_mlp": 1.03477991,
      "epoch": 0.07323012174958665,
      "flos": 14392602529920.0,
      "grad_norm": 1.9658490798069863,
      "language_loss": 0.86455309,
      "learning_rate": 3.98042878992303e-06,
      "loss": 0.88680959,
      "num_input_tokens_seen": 25927215,
      "step": 1218,
      "time_per_iteration": 2.5911786556243896
    },
    {
      "auxiliary_loss_clip": 0.01192266,
      "auxiliary_loss_mlp": 0.0106258,
      "balance_loss_clip": 1.06015348,
      "balance_loss_mlp": 1.03916681,
      "epoch": 0.07329024500225462,
      "flos": 21616428591360.0,
      "grad_norm": 2.2310702082820675,
      "language_loss": 0.86782354,
      "learning_rate": 3.9803744012531305e-06,
      "loss": 0.89037204,
      "num_input_tokens_seen": 25945500,
      "step": 1219,
      "time_per_iteration": 2.608562707901001
    },
    {
      "auxiliary_loss_clip": 0.01201545,
      "auxiliary_loss_mlp": 0.01058282,
      "balance_loss_clip": 1.06024373,
      "balance_loss_mlp": 1.03539419,
      "epoch": 0.0733503682549226,
      "flos": 13224141106560.0,
      "grad_norm": 2.095886373367052,
      "language_loss": 0.84608674,
      "learning_rate": 3.980319937487235e-06,
      "loss": 0.86868501,
      "num_input_tokens_seen": 25963105,
      "step": 1220,
      "time_per_iteration": 2.469189405441284
    },
    {
      "auxiliary_loss_clip": 0.01158855,
      "auxiliary_loss_mlp": 0.01063399,
      "balance_loss_clip": 1.05358922,
      "balance_loss_mlp": 1.03942597,
      "epoch": 0.07341049150759056,
      "flos": 20886975192960.0,
      "grad_norm": 2.648884311755534,
      "language_loss": 0.77114344,
      "learning_rate": 3.98026539862741e-06,
      "loss": 0.79336596,
      "num_input_tokens_seen": 25981690,
      "step": 1221,
      "time_per_iteration": 2.671762466430664
    },
    {
      "auxiliary_loss_clip": 0.01158201,
      "auxiliary_loss_mlp": 0.01064916,
      "balance_loss_clip": 1.05726743,
      "balance_loss_mlp": 1.04082406,
      "epoch": 0.07347061476025853,
      "flos": 15413614623360.0,
      "grad_norm": 2.5357389392469942,
      "language_loss": 0.91631913,
      "learning_rate": 3.980210784675722e-06,
      "loss": 0.93855029,
      "num_input_tokens_seen": 25999890,
      "step": 1222,
      "time_per_iteration": 2.6973063945770264
    },
    {
      "auxiliary_loss_clip": 0.01135907,
      "auxiliary_loss_mlp": 0.01064872,
      "balance_loss_clip": 1.05333126,
      "balance_loss_mlp": 1.04169726,
      "epoch": 0.0735307380129265,
      "flos": 11108859131520.0,
      "grad_norm": 2.8024324299253047,
      "language_loss": 0.90976465,
      "learning_rate": 3.980156095634242e-06,
      "loss": 0.93177247,
      "num_input_tokens_seen": 26016445,
      "step": 1223,
      "time_per_iteration": 2.8141093254089355
    },
    {
      "auxiliary_loss_clip": 0.01202875,
      "auxiliary_loss_mlp": 0.01077185,
      "balance_loss_clip": 1.06232905,
      "balance_loss_mlp": 1.05341494,
      "epoch": 0.07359086126559447,
      "flos": 23732392924800.0,
      "grad_norm": 1.9348534518871447,
      "language_loss": 0.82161939,
      "learning_rate": 3.980101331505045e-06,
      "loss": 0.84442002,
      "num_input_tokens_seen": 26036080,
      "step": 1224,
      "time_per_iteration": 2.640432119369507
    },
    {
      "auxiliary_loss_clip": 0.01200329,
      "auxiliary_loss_mlp": 0.01057586,
      "balance_loss_clip": 1.05987597,
      "balance_loss_mlp": 1.03229022,
      "epoch": 0.07365098451826244,
      "flos": 20993270515200.0,
      "grad_norm": 2.31744406237409,
      "language_loss": 0.83194047,
      "learning_rate": 3.9800464922902076e-06,
      "loss": 0.85451961,
      "num_input_tokens_seen": 26055805,
      "step": 1225,
      "time_per_iteration": 2.6159210205078125
    },
    {
      "auxiliary_loss_clip": 0.01170115,
      "auxiliary_loss_mlp": 0.01056068,
      "balance_loss_clip": 1.05743551,
      "balance_loss_mlp": 1.03190422,
      "epoch": 0.0737111077709304,
      "flos": 19933582452480.0,
      "grad_norm": 2.2959030425986544,
      "language_loss": 0.90388274,
      "learning_rate": 3.979991577991808e-06,
      "loss": 0.9261446,
      "num_input_tokens_seen": 26073905,
      "step": 1226,
      "time_per_iteration": 2.6527435779571533
    },
    {
      "auxiliary_loss_clip": 0.01207799,
      "auxiliary_loss_mlp": 0.0104599,
      "balance_loss_clip": 1.05913424,
      "balance_loss_mlp": 1.02080154,
      "epoch": 0.07377123102359838,
      "flos": 16581537342720.0,
      "grad_norm": 2.579592162134606,
      "language_loss": 0.76626784,
      "learning_rate": 3.97993658861193e-06,
      "loss": 0.78880572,
      "num_input_tokens_seen": 26091700,
      "step": 1227,
      "time_per_iteration": 2.596151351928711
    },
    {
      "auxiliary_loss_clip": 0.0118909,
      "auxiliary_loss_mlp": 0.01053386,
      "balance_loss_clip": 1.06296694,
      "balance_loss_mlp": 1.02954459,
      "epoch": 0.07383135427626634,
      "flos": 28328563457280.0,
      "grad_norm": 7.788838200212175,
      "language_loss": 0.8555491,
      "learning_rate": 3.9798815241526575e-06,
      "loss": 0.87797379,
      "num_input_tokens_seen": 26114105,
      "step": 1228,
      "time_per_iteration": 2.6955716609954834
    },
    {
      "auxiliary_loss_clip": 0.01191175,
      "auxiliary_loss_mlp": 0.01062669,
      "balance_loss_clip": 1.05897212,
      "balance_loss_mlp": 1.03860044,
      "epoch": 0.07389147752893431,
      "flos": 20047168235520.0,
      "grad_norm": 2.2575099517148898,
      "language_loss": 0.79598552,
      "learning_rate": 3.97982638461608e-06,
      "loss": 0.818524,
      "num_input_tokens_seen": 26131165,
      "step": 1229,
      "time_per_iteration": 2.6544861793518066
    },
    {
      "auxiliary_loss_clip": 0.01192886,
      "auxiliary_loss_mlp": 0.00782044,
      "balance_loss_clip": 1.05966699,
      "balance_loss_mlp": 1.00032902,
      "epoch": 0.07395160078160229,
      "flos": 18114132890880.0,
      "grad_norm": 2.2881874382496377,
      "language_loss": 0.78209347,
      "learning_rate": 3.979771170004287e-06,
      "loss": 0.80184281,
      "num_input_tokens_seen": 26150040,
      "step": 1230,
      "time_per_iteration": 2.6001133918762207
    },
    {
      "auxiliary_loss_clip": 0.0120142,
      "auxiliary_loss_mlp": 0.01052342,
      "balance_loss_clip": 1.06209648,
      "balance_loss_mlp": 1.02739108,
      "epoch": 0.07401172403427025,
      "flos": 23586918842880.0,
      "grad_norm": 2.038847041772147,
      "language_loss": 0.8136946,
      "learning_rate": 3.979715880319372e-06,
      "loss": 0.83623219,
      "num_input_tokens_seen": 26169380,
      "step": 1231,
      "time_per_iteration": 2.6364073753356934
    },
    {
      "auxiliary_loss_clip": 0.01179975,
      "auxiliary_loss_mlp": 0.01070917,
      "balance_loss_clip": 1.05690873,
      "balance_loss_mlp": 1.04599047,
      "epoch": 0.07407184728693822,
      "flos": 26359904799360.0,
      "grad_norm": 2.096832924731062,
      "language_loss": 0.95204866,
      "learning_rate": 3.979660515563434e-06,
      "loss": 0.97455758,
      "num_input_tokens_seen": 26189420,
      "step": 1232,
      "time_per_iteration": 2.7929203510284424
    },
    {
      "auxiliary_loss_clip": 0.01187282,
      "auxiliary_loss_mlp": 0.01059661,
      "balance_loss_clip": 1.06202245,
      "balance_loss_mlp": 1.03733301,
      "epoch": 0.0741319705396062,
      "flos": 22200443821440.0,
      "grad_norm": 1.7778448126368063,
      "language_loss": 0.80695188,
      "learning_rate": 3.979605075738569e-06,
      "loss": 0.82942128,
      "num_input_tokens_seen": 26209300,
      "step": 1233,
      "time_per_iteration": 2.7945051193237305
    },
    {
      "auxiliary_loss_clip": 0.01209245,
      "auxiliary_loss_mlp": 0.0106207,
      "balance_loss_clip": 1.06238747,
      "balance_loss_mlp": 1.03602231,
      "epoch": 0.07419209379227416,
      "flos": 39200482523520.0,
      "grad_norm": 2.136728864247421,
      "language_loss": 0.70708907,
      "learning_rate": 3.979549560846883e-06,
      "loss": 0.72980225,
      "num_input_tokens_seen": 26228110,
      "step": 1234,
      "time_per_iteration": 2.9646782875061035
    },
    {
      "auxiliary_loss_clip": 0.01167486,
      "auxiliary_loss_mlp": 0.01068879,
      "balance_loss_clip": 1.0542618,
      "balance_loss_mlp": 1.04265285,
      "epoch": 0.07425221704494213,
      "flos": 22781657790720.0,
      "grad_norm": 1.7921102377369336,
      "language_loss": 0.76852918,
      "learning_rate": 3.979493970890478e-06,
      "loss": 0.79089284,
      "num_input_tokens_seen": 26247020,
      "step": 1235,
      "time_per_iteration": 2.820577621459961
    },
    {
      "auxiliary_loss_clip": 0.01198028,
      "auxiliary_loss_mlp": 0.01055883,
      "balance_loss_clip": 1.05918813,
      "balance_loss_mlp": 1.0321244,
      "epoch": 0.0743123402976101,
      "flos": 22272983337600.0,
      "grad_norm": 2.3018318065058097,
      "language_loss": 0.82748145,
      "learning_rate": 3.979438305871464e-06,
      "loss": 0.85002053,
      "num_input_tokens_seen": 26265750,
      "step": 1236,
      "time_per_iteration": 2.6302287578582764
    },
    {
      "auxiliary_loss_clip": 0.01154783,
      "auxiliary_loss_mlp": 0.00782014,
      "balance_loss_clip": 1.05519629,
      "balance_loss_mlp": 1.00039148,
      "epoch": 0.07437246355027807,
      "flos": 29315029645440.0,
      "grad_norm": 1.7985383717833268,
      "language_loss": 0.7595011,
      "learning_rate": 3.979382565791951e-06,
      "loss": 0.77886909,
      "num_input_tokens_seen": 26287905,
      "step": 1237,
      "time_per_iteration": 2.721931219100952
    },
    {
      "auxiliary_loss_clip": 0.01135551,
      "auxiliary_loss_mlp": 0.00783311,
      "balance_loss_clip": 1.0505693,
      "balance_loss_mlp": 1.00031757,
      "epoch": 0.07443258680294604,
      "flos": 31944732249600.0,
      "grad_norm": 1.6915170784810407,
      "language_loss": 0.77458763,
      "learning_rate": 3.979326750654053e-06,
      "loss": 0.79377621,
      "num_input_tokens_seen": 26311795,
      "step": 1238,
      "time_per_iteration": 2.831620931625366
    },
    {
      "auxiliary_loss_clip": 0.01177529,
      "auxiliary_loss_mlp": 0.01057762,
      "balance_loss_clip": 1.05673254,
      "balance_loss_mlp": 1.03311002,
      "epoch": 0.074492710055614,
      "flos": 22675290641280.0,
      "grad_norm": 1.9053364150897723,
      "language_loss": 0.867737,
      "learning_rate": 3.9792708604598854e-06,
      "loss": 0.89008987,
      "num_input_tokens_seen": 26330330,
      "step": 1239,
      "time_per_iteration": 2.6697263717651367
    },
    {
      "auxiliary_loss_clip": 0.01159844,
      "auxiliary_loss_mlp": 0.01050954,
      "balance_loss_clip": 1.05222142,
      "balance_loss_mlp": 1.02532458,
      "epoch": 0.07455283330828198,
      "flos": 21284901037440.0,
      "grad_norm": 26.978042105238785,
      "language_loss": 0.89356089,
      "learning_rate": 3.979214895211569e-06,
      "loss": 0.91566885,
      "num_input_tokens_seen": 26348865,
      "step": 1240,
      "time_per_iteration": 2.846013069152832
    },
    {
      "auxiliary_loss_clip": 0.01174117,
      "auxiliary_loss_mlp": 0.01063539,
      "balance_loss_clip": 1.05857158,
      "balance_loss_mlp": 1.03713393,
      "epoch": 0.07461295656094995,
      "flos": 24388408967040.0,
      "grad_norm": 1.9346624045484253,
      "language_loss": 0.88873678,
      "learning_rate": 3.979158854911225e-06,
      "loss": 0.91111326,
      "num_input_tokens_seen": 26368210,
      "step": 1241,
      "time_per_iteration": 2.6926562786102295
    },
    {
      "auxiliary_loss_clip": 0.01079637,
      "auxiliary_loss_mlp": 0.01009562,
      "balance_loss_clip": 1.03489435,
      "balance_loss_mlp": 1.00405502,
      "epoch": 0.07467307981361791,
      "flos": 62109660574080.0,
      "grad_norm": 0.8973011136706247,
      "language_loss": 0.63067901,
      "learning_rate": 3.979102739560979e-06,
      "loss": 0.65157104,
      "num_input_tokens_seen": 26424890,
      "step": 1242,
      "time_per_iteration": 3.298609972000122
    },
    {
      "auxiliary_loss_clip": 0.01164269,
      "auxiliary_loss_mlp": 0.01068833,
      "balance_loss_clip": 1.05246222,
      "balance_loss_mlp": 1.03819644,
      "epoch": 0.07473320306628589,
      "flos": 24863148046080.0,
      "grad_norm": 3.87499965477456,
      "language_loss": 0.62926078,
      "learning_rate": 3.9790465491629595e-06,
      "loss": 0.65159178,
      "num_input_tokens_seen": 26446405,
      "step": 1243,
      "time_per_iteration": 2.7774572372436523
    },
    {
      "auxiliary_loss_clip": 0.01188864,
      "auxiliary_loss_mlp": 0.01059918,
      "balance_loss_clip": 1.05716145,
      "balance_loss_mlp": 1.03499091,
      "epoch": 0.07479332631895386,
      "flos": 24897442556160.0,
      "grad_norm": 1.6252135866538246,
      "language_loss": 0.76259589,
      "learning_rate": 3.978990283719296e-06,
      "loss": 0.78508377,
      "num_input_tokens_seen": 26466070,
      "step": 1244,
      "time_per_iteration": 2.714459180831909
    },
    {
      "auxiliary_loss_clip": 0.01184345,
      "auxiliary_loss_mlp": 0.00783076,
      "balance_loss_clip": 1.0611167,
      "balance_loss_mlp": 1.00038469,
      "epoch": 0.07485344957162182,
      "flos": 17815247821440.0,
      "grad_norm": 5.636002853507256,
      "language_loss": 0.69419599,
      "learning_rate": 3.978933943232123e-06,
      "loss": 0.71387023,
      "num_input_tokens_seen": 26479350,
      "step": 1245,
      "time_per_iteration": 2.640895366668701
    },
    {
      "auxiliary_loss_clip": 0.01203955,
      "auxiliary_loss_mlp": 0.01062684,
      "balance_loss_clip": 1.06098139,
      "balance_loss_mlp": 1.0372088,
      "epoch": 0.0749135728242898,
      "flos": 25010202326400.0,
      "grad_norm": 2.5525245798098757,
      "language_loss": 0.88635457,
      "learning_rate": 3.978877527703576e-06,
      "loss": 0.90902102,
      "num_input_tokens_seen": 26498255,
      "step": 1246,
      "time_per_iteration": 2.747765302658081
    },
    {
      "auxiliary_loss_clip": 0.01212369,
      "auxiliary_loss_mlp": 0.01077452,
      "balance_loss_clip": 1.06102896,
      "balance_loss_mlp": 1.049402,
      "epoch": 0.07497369607695777,
      "flos": 17822071405440.0,
      "grad_norm": 2.675073323546491,
      "language_loss": 0.8825295,
      "learning_rate": 3.9788210371357945e-06,
      "loss": 0.90542769,
      "num_input_tokens_seen": 26515375,
      "step": 1247,
      "time_per_iteration": 2.6810224056243896
    },
    {
      "auxiliary_loss_clip": 0.0118495,
      "auxiliary_loss_mlp": 0.01069489,
      "balance_loss_clip": 1.06058884,
      "balance_loss_mlp": 1.04383492,
      "epoch": 0.07503381932962573,
      "flos": 15121086261120.0,
      "grad_norm": 2.620559853720615,
      "language_loss": 0.64849806,
      "learning_rate": 3.978764471530921e-06,
      "loss": 0.67104244,
      "num_input_tokens_seen": 26533595,
      "step": 1248,
      "time_per_iteration": 2.706862449645996
    },
    {
      "auxiliary_loss_clip": 0.01181878,
      "auxiliary_loss_mlp": 0.00782677,
      "balance_loss_clip": 1.0575974,
      "balance_loss_mlp": 1.0004611,
      "epoch": 0.0750939425822937,
      "flos": 12816734071680.0,
      "grad_norm": 2.872208543000993,
      "language_loss": 0.74216163,
      "learning_rate": 3.978707830891102e-06,
      "loss": 0.7618072,
      "num_input_tokens_seen": 26549405,
      "step": 1249,
      "time_per_iteration": 4.309665679931641
    },
    {
      "auxiliary_loss_clip": 0.01168375,
      "auxiliary_loss_mlp": 0.01079691,
      "balance_loss_clip": 1.0579834,
      "balance_loss_mlp": 1.05296445,
      "epoch": 0.07515406583496168,
      "flos": 24206844695040.0,
      "grad_norm": 2.679176110316805,
      "language_loss": 0.82353318,
      "learning_rate": 3.978651115218482e-06,
      "loss": 0.84601378,
      "num_input_tokens_seen": 26567200,
      "step": 1250,
      "time_per_iteration": 4.367432594299316
    },
    {
      "auxiliary_loss_clip": 0.011507,
      "auxiliary_loss_mlp": 0.01064103,
      "balance_loss_clip": 1.05736125,
      "balance_loss_mlp": 1.0380677,
      "epoch": 0.07521418908762964,
      "flos": 26688164215680.0,
      "grad_norm": 2.015636709873133,
      "language_loss": 0.6679548,
      "learning_rate": 3.978594324515215e-06,
      "loss": 0.69010288,
      "num_input_tokens_seen": 26586190,
      "step": 1251,
      "time_per_iteration": 4.339111089706421
    },
    {
      "auxiliary_loss_clip": 0.01061099,
      "auxiliary_loss_mlp": 0.01007289,
      "balance_loss_clip": 1.02992618,
      "balance_loss_mlp": 1.00314093,
      "epoch": 0.0752743123402976,
      "flos": 59095140589440.0,
      "grad_norm": 0.9014655793512963,
      "language_loss": 0.7038399,
      "learning_rate": 3.9785374587834515e-06,
      "loss": 0.72452378,
      "num_input_tokens_seen": 26650710,
      "step": 1252,
      "time_per_iteration": 4.984445333480835
    },
    {
      "auxiliary_loss_clip": 0.0120348,
      "auxiliary_loss_mlp": 0.01071343,
      "balance_loss_clip": 1.06016684,
      "balance_loss_mlp": 1.04651129,
      "epoch": 0.07533443559296558,
      "flos": 23477032160640.0,
      "grad_norm": 2.2789224049077226,
      "language_loss": 0.79936707,
      "learning_rate": 3.97848051802535e-06,
      "loss": 0.82211524,
      "num_input_tokens_seen": 26669000,
      "step": 1253,
      "time_per_iteration": 2.613696575164795
    },
    {
      "auxiliary_loss_clip": 0.01165402,
      "auxiliary_loss_mlp": 0.01062493,
      "balance_loss_clip": 1.05703712,
      "balance_loss_mlp": 1.03758967,
      "epoch": 0.07539455884563355,
      "flos": 20879110114560.0,
      "grad_norm": 3.1057458778243263,
      "language_loss": 0.93360364,
      "learning_rate": 3.978423502243069e-06,
      "loss": 0.95588255,
      "num_input_tokens_seen": 26683075,
      "step": 1254,
      "time_per_iteration": 2.7332606315612793
    },
    {
      "auxiliary_loss_clip": 0.011733,
      "auxiliary_loss_mlp": 0.01064454,
      "balance_loss_clip": 1.06050682,
      "balance_loss_mlp": 1.03958726,
      "epoch": 0.07545468209830151,
      "flos": 27672906551040.0,
      "grad_norm": 2.090631066181037,
      "language_loss": 0.88087487,
      "learning_rate": 3.97836641143877e-06,
      "loss": 0.90325236,
      "num_input_tokens_seen": 26701875,
      "step": 1255,
      "time_per_iteration": 2.713636875152588
    },
    {
      "auxiliary_loss_clip": 0.01202338,
      "auxiliary_loss_mlp": 0.01071467,
      "balance_loss_clip": 1.06138325,
      "balance_loss_mlp": 1.04531264,
      "epoch": 0.0755148053509695,
      "flos": 14136990370560.0,
      "grad_norm": 1.9772348994273161,
      "language_loss": 0.79305708,
      "learning_rate": 3.978309245614618e-06,
      "loss": 0.81579506,
      "num_input_tokens_seen": 26719050,
      "step": 1256,
      "time_per_iteration": 2.688812255859375
    },
    {
      "auxiliary_loss_clip": 0.01064506,
      "auxiliary_loss_mlp": 0.01008663,
      "balance_loss_clip": 1.0281384,
      "balance_loss_mlp": 1.0043,
      "epoch": 0.07557492860363746,
      "flos": 58235257929600.0,
      "grad_norm": 0.7721513084275832,
      "language_loss": 0.58031851,
      "learning_rate": 3.9782520047727825e-06,
      "loss": 0.6010502,
      "num_input_tokens_seen": 26780650,
      "step": 1257,
      "time_per_iteration": 3.290971517562866
    },
    {
      "auxiliary_loss_clip": 0.01154091,
      "auxiliary_loss_mlp": 0.01065293,
      "balance_loss_clip": 1.06175375,
      "balance_loss_mlp": 1.04035461,
      "epoch": 0.07563505185630542,
      "flos": 24644380262400.0,
      "grad_norm": 2.5700283098608026,
      "language_loss": 0.90029764,
      "learning_rate": 3.978194688915432e-06,
      "loss": 0.92249143,
      "num_input_tokens_seen": 26798725,
      "step": 1258,
      "time_per_iteration": 2.800297975540161
    },
    {
      "auxiliary_loss_clip": 0.01169581,
      "auxiliary_loss_mlp": 0.01064585,
      "balance_loss_clip": 1.06184185,
      "balance_loss_mlp": 1.03797793,
      "epoch": 0.07569517510897339,
      "flos": 15522998515200.0,
      "grad_norm": 2.1868972302346377,
      "language_loss": 0.81404132,
      "learning_rate": 3.978137298044741e-06,
      "loss": 0.83638299,
      "num_input_tokens_seen": 26817005,
      "step": 1259,
      "time_per_iteration": 2.767717123031616
    },
    {
      "auxiliary_loss_clip": 0.01194891,
      "auxiliary_loss_mlp": 0.01062022,
      "balance_loss_clip": 1.06317782,
      "balance_loss_mlp": 1.03766739,
      "epoch": 0.07575529836164137,
      "flos": 22928532503040.0,
      "grad_norm": 1.8876128491153832,
      "language_loss": 0.7609086,
      "learning_rate": 3.978079832162885e-06,
      "loss": 0.78347778,
      "num_input_tokens_seen": 26836655,
      "step": 1260,
      "time_per_iteration": 2.859339714050293
    },
    {
      "auxiliary_loss_clip": 0.01160098,
      "auxiliary_loss_mlp": 0.01068568,
      "balance_loss_clip": 1.05432057,
      "balance_loss_mlp": 1.04222322,
      "epoch": 0.07581542161430933,
      "flos": 19500428344320.0,
      "grad_norm": 1.7028037437197219,
      "language_loss": 0.84734851,
      "learning_rate": 3.978022291272044e-06,
      "loss": 0.86963522,
      "num_input_tokens_seen": 26854925,
      "step": 1261,
      "time_per_iteration": 2.773087978363037
    },
    {
      "auxiliary_loss_clip": 0.01212087,
      "auxiliary_loss_mlp": 0.0106726,
      "balance_loss_clip": 1.06821966,
      "balance_loss_mlp": 1.04273915,
      "epoch": 0.0758755448669773,
      "flos": 24973465691520.0,
      "grad_norm": 1.8668314773439494,
      "language_loss": 0.82578814,
      "learning_rate": 3.977964675374399e-06,
      "loss": 0.84858155,
      "num_input_tokens_seen": 26876170,
      "step": 1262,
      "time_per_iteration": 2.681764841079712
    },
    {
      "auxiliary_loss_clip": 0.01206367,
      "auxiliary_loss_mlp": 0.0106285,
      "balance_loss_clip": 1.06333947,
      "balance_loss_mlp": 1.03685009,
      "epoch": 0.07593566811964528,
      "flos": 22747973811840.0,
      "grad_norm": 2.501362251414687,
      "language_loss": 0.82448232,
      "learning_rate": 3.977906984472136e-06,
      "loss": 0.84717447,
      "num_input_tokens_seen": 26895005,
      "step": 1263,
      "time_per_iteration": 2.6262786388397217
    },
    {
      "auxiliary_loss_clip": 0.01166059,
      "auxiliary_loss_mlp": 0.01068738,
      "balance_loss_clip": 1.06484997,
      "balance_loss_mlp": 1.04334641,
      "epoch": 0.07599579137231324,
      "flos": 23112395245440.0,
      "grad_norm": 2.171520639750579,
      "language_loss": 0.76149648,
      "learning_rate": 3.977849218567442e-06,
      "loss": 0.78384447,
      "num_input_tokens_seen": 26913930,
      "step": 1264,
      "time_per_iteration": 2.7735466957092285
    },
    {
      "auxiliary_loss_clip": 0.01181777,
      "auxiliary_loss_mlp": 0.01061673,
      "balance_loss_clip": 1.06183577,
      "balance_loss_mlp": 1.03704381,
      "epoch": 0.07605591462498121,
      "flos": 14502058248960.0,
      "grad_norm": 2.252731793921747,
      "language_loss": 0.80919051,
      "learning_rate": 3.977791377662507e-06,
      "loss": 0.83162498,
      "num_input_tokens_seen": 26931485,
      "step": 1265,
      "time_per_iteration": 2.6076793670654297
    },
    {
      "auxiliary_loss_clip": 0.01143593,
      "auxiliary_loss_mlp": 0.01068856,
      "balance_loss_clip": 1.05383801,
      "balance_loss_mlp": 1.0411638,
      "epoch": 0.07611603787764919,
      "flos": 23514199758720.0,
      "grad_norm": 2.117217065332582,
      "language_loss": 0.65244937,
      "learning_rate": 3.977733461759524e-06,
      "loss": 0.67457378,
      "num_input_tokens_seen": 26951670,
      "step": 1266,
      "time_per_iteration": 2.714848041534424
    },
    {
      "auxiliary_loss_clip": 0.0116364,
      "auxiliary_loss_mlp": 0.01066982,
      "balance_loss_clip": 1.05869627,
      "balance_loss_mlp": 1.04194832,
      "epoch": 0.07617616113031715,
      "flos": 21507188353920.0,
      "grad_norm": 2.0157381540709416,
      "language_loss": 0.79570109,
      "learning_rate": 3.977675470860691e-06,
      "loss": 0.81800735,
      "num_input_tokens_seen": 26970335,
      "step": 1267,
      "time_per_iteration": 2.692220687866211
    },
    {
      "auxiliary_loss_clip": 0.01186526,
      "auxiliary_loss_mlp": 0.01060572,
      "balance_loss_clip": 1.06368709,
      "balance_loss_mlp": 1.03644359,
      "epoch": 0.07623628438298512,
      "flos": 14573161221120.0,
      "grad_norm": 2.573855585409162,
      "language_loss": 0.72936547,
      "learning_rate": 3.977617404968205e-06,
      "loss": 0.75183642,
      "num_input_tokens_seen": 26986025,
      "step": 1268,
      "time_per_iteration": 2.666487216949463
    },
    {
      "auxiliary_loss_clip": 0.01189272,
      "auxiliary_loss_mlp": 0.01056943,
      "balance_loss_clip": 1.05925119,
      "balance_loss_mlp": 1.03146791,
      "epoch": 0.07629640763565308,
      "flos": 14720395069440.0,
      "grad_norm": 2.3531002902867018,
      "language_loss": 0.82087409,
      "learning_rate": 3.977559264084269e-06,
      "loss": 0.84333622,
      "num_input_tokens_seen": 27004045,
      "step": 1269,
      "time_per_iteration": 2.6196024417877197
    },
    {
      "auxiliary_loss_clip": 0.01198264,
      "auxiliary_loss_mlp": 0.01062408,
      "balance_loss_clip": 1.06528163,
      "balance_loss_mlp": 1.03656352,
      "epoch": 0.07635653088832106,
      "flos": 14902929008640.0,
      "grad_norm": 2.6660741307472424,
      "language_loss": 0.88614184,
      "learning_rate": 3.977501048211088e-06,
      "loss": 0.90874851,
      "num_input_tokens_seen": 27022070,
      "step": 1270,
      "time_per_iteration": 2.6423919200897217
    },
    {
      "auxiliary_loss_clip": 0.01195764,
      "auxiliary_loss_mlp": 0.01062092,
      "balance_loss_clip": 1.06443572,
      "balance_loss_mlp": 1.0371294,
      "epoch": 0.07641665414098903,
      "flos": 26651571235200.0,
      "grad_norm": 2.486841045046768,
      "language_loss": 0.7104162,
      "learning_rate": 3.977442757350869e-06,
      "loss": 0.73299474,
      "num_input_tokens_seen": 27041755,
      "step": 1271,
      "time_per_iteration": 2.6679437160491943
    },
    {
      "auxiliary_loss_clip": 0.01157818,
      "auxiliary_loss_mlp": 0.01068131,
      "balance_loss_clip": 1.05973268,
      "balance_loss_mlp": 1.04282308,
      "epoch": 0.07647677739365699,
      "flos": 25192808092800.0,
      "grad_norm": 1.5691807400142836,
      "language_loss": 0.82570392,
      "learning_rate": 3.977384391505823e-06,
      "loss": 0.84796339,
      "num_input_tokens_seen": 27061540,
      "step": 1272,
      "time_per_iteration": 2.7613680362701416
    },
    {
      "auxiliary_loss_clip": 0.01176176,
      "auxiliary_loss_mlp": 0.00782751,
      "balance_loss_clip": 1.05822372,
      "balance_loss_mlp": 1.00051665,
      "epoch": 0.07653690064632497,
      "flos": 20558141159040.0,
      "grad_norm": 1.811509476700225,
      "language_loss": 0.79854733,
      "learning_rate": 3.977325950678162e-06,
      "loss": 0.81813657,
      "num_input_tokens_seen": 27081395,
      "step": 1273,
      "time_per_iteration": 2.696317434310913
    },
    {
      "auxiliary_loss_clip": 0.01185133,
      "auxiliary_loss_mlp": 0.01064308,
      "balance_loss_clip": 1.06556833,
      "balance_loss_mlp": 1.03910685,
      "epoch": 0.07659702389899294,
      "flos": 22269320150400.0,
      "grad_norm": 1.7399681078894738,
      "language_loss": 0.81519866,
      "learning_rate": 3.977267434870103e-06,
      "loss": 0.83769304,
      "num_input_tokens_seen": 27101175,
      "step": 1274,
      "time_per_iteration": 2.8570950031280518
    },
    {
      "auxiliary_loss_clip": 0.0118748,
      "auxiliary_loss_mlp": 0.01078696,
      "balance_loss_clip": 1.06516898,
      "balance_loss_mlp": 1.05164731,
      "epoch": 0.0766571471516609,
      "flos": 32636120209920.0,
      "grad_norm": 2.6845981005996453,
      "language_loss": 0.73083639,
      "learning_rate": 3.977208844083865e-06,
      "loss": 0.75349814,
      "num_input_tokens_seen": 27124505,
      "step": 1275,
      "time_per_iteration": 2.75947904586792
    },
    {
      "auxiliary_loss_clip": 0.0121081,
      "auxiliary_loss_mlp": 0.01063745,
      "balance_loss_clip": 1.06740415,
      "balance_loss_mlp": 1.03694642,
      "epoch": 0.07671727040432888,
      "flos": 15267386355840.0,
      "grad_norm": 2.828157953752124,
      "language_loss": 0.79507053,
      "learning_rate": 3.9771501783216685e-06,
      "loss": 0.81781602,
      "num_input_tokens_seen": 27140960,
      "step": 1276,
      "time_per_iteration": 2.626683473587036
    },
    {
      "auxiliary_loss_clip": 0.01198279,
      "auxiliary_loss_mlp": 0.01058719,
      "balance_loss_clip": 1.06486118,
      "balance_loss_mlp": 1.03485298,
      "epoch": 0.07677739365699685,
      "flos": 28184094956160.0,
      "grad_norm": 2.406514987231471,
      "language_loss": 0.58915478,
      "learning_rate": 3.97709143758574e-06,
      "loss": 0.61172473,
      "num_input_tokens_seen": 27160985,
      "step": 1277,
      "time_per_iteration": 2.6684958934783936
    },
    {
      "auxiliary_loss_clip": 0.01201282,
      "auxiliary_loss_mlp": 0.01064396,
      "balance_loss_clip": 1.06430948,
      "balance_loss_mlp": 1.03919542,
      "epoch": 0.07683751690966481,
      "flos": 18296128126080.0,
      "grad_norm": 2.8024245322836046,
      "language_loss": 0.74957907,
      "learning_rate": 3.977032621878305e-06,
      "loss": 0.77223587,
      "num_input_tokens_seen": 27178390,
      "step": 1278,
      "time_per_iteration": 2.723675012588501
    },
    {
      "auxiliary_loss_clip": 0.01160972,
      "auxiliary_loss_mlp": 0.01063133,
      "balance_loss_clip": 1.0584681,
      "balance_loss_mlp": 1.0390408,
      "epoch": 0.07689764016233278,
      "flos": 21981101420160.0,
      "grad_norm": 5.339853944094037,
      "language_loss": 0.88594604,
      "learning_rate": 3.976973731201596e-06,
      "loss": 0.90818715,
      "num_input_tokens_seen": 27197505,
      "step": 1279,
      "time_per_iteration": 2.655036211013794
    },
    {
      "auxiliary_loss_clip": 0.01172627,
      "auxiliary_loss_mlp": 0.01066586,
      "balance_loss_clip": 1.06065845,
      "balance_loss_mlp": 1.04077685,
      "epoch": 0.07695776341500075,
      "flos": 22235995307520.0,
      "grad_norm": 2.4937131241937256,
      "language_loss": 0.8300451,
      "learning_rate": 3.976914765557845e-06,
      "loss": 0.85243726,
      "num_input_tokens_seen": 27214260,
      "step": 1280,
      "time_per_iteration": 2.7717065811157227
    },
    {
      "auxiliary_loss_clip": 0.01194022,
      "auxiliary_loss_mlp": 0.01066533,
      "balance_loss_clip": 1.06593037,
      "balance_loss_mlp": 1.04104638,
      "epoch": 0.07701788666766872,
      "flos": 16143750380160.0,
      "grad_norm": 2.044864943195716,
      "language_loss": 0.7581439,
      "learning_rate": 3.9768557249492875e-06,
      "loss": 0.78074944,
      "num_input_tokens_seen": 27232525,
      "step": 1281,
      "time_per_iteration": 2.7444865703582764
    },
    {
      "auxiliary_loss_clip": 0.01170775,
      "auxiliary_loss_mlp": 0.01062526,
      "balance_loss_clip": 1.05879402,
      "balance_loss_mlp": 1.03669322,
      "epoch": 0.07707800992033668,
      "flos": 19463045264640.0,
      "grad_norm": 1.8925477349429178,
      "language_loss": 0.75091648,
      "learning_rate": 3.9767966093781634e-06,
      "loss": 0.77324951,
      "num_input_tokens_seen": 27249800,
      "step": 1282,
      "time_per_iteration": 2.829145908355713
    },
    {
      "auxiliary_loss_clip": 0.01213222,
      "auxiliary_loss_mlp": 0.01071082,
      "balance_loss_clip": 1.07007408,
      "balance_loss_mlp": 1.04549992,
      "epoch": 0.07713813317300466,
      "flos": 18990281433600.0,
      "grad_norm": 2.1558853998977527,
      "language_loss": 0.83863324,
      "learning_rate": 3.976737418846713e-06,
      "loss": 0.8614763,
      "num_input_tokens_seen": 27268895,
      "step": 1283,
      "time_per_iteration": 2.6955173015594482
    },
    {
      "auxiliary_loss_clip": 0.0119621,
      "auxiliary_loss_mlp": 0.01066889,
      "balance_loss_clip": 1.06603825,
      "balance_loss_mlp": 1.03925657,
      "epoch": 0.07719825642567263,
      "flos": 18113953322880.0,
      "grad_norm": 2.520477290704422,
      "language_loss": 0.75147104,
      "learning_rate": 3.976678153357181e-06,
      "loss": 0.77410209,
      "num_input_tokens_seen": 27288180,
      "step": 1284,
      "time_per_iteration": 2.6589291095733643
    },
    {
      "auxiliary_loss_clip": 0.01182212,
      "auxiliary_loss_mlp": 0.01068485,
      "balance_loss_clip": 1.06304765,
      "balance_loss_mlp": 1.0438329,
      "epoch": 0.0772583796783406,
      "flos": 42194426993280.0,
      "grad_norm": 5.2953301239297295,
      "language_loss": 0.76224041,
      "learning_rate": 3.976618812911817e-06,
      "loss": 0.78474742,
      "num_input_tokens_seen": 27311815,
      "step": 1285,
      "time_per_iteration": 2.847702741622925
    },
    {
      "auxiliary_loss_clip": 0.01216302,
      "auxiliary_loss_mlp": 0.01071451,
      "balance_loss_clip": 1.07193899,
      "balance_loss_mlp": 1.04729891,
      "epoch": 0.07731850293100857,
      "flos": 24753692327040.0,
      "grad_norm": 2.0564733507641,
      "language_loss": 0.84193194,
      "learning_rate": 3.9765593975128685e-06,
      "loss": 0.86480945,
      "num_input_tokens_seen": 27331890,
      "step": 1286,
      "time_per_iteration": 2.713963270187378
    },
    {
      "auxiliary_loss_clip": 0.01180469,
      "auxiliary_loss_mlp": 0.01061062,
      "balance_loss_clip": 1.06331325,
      "balance_loss_mlp": 1.03646958,
      "epoch": 0.07737862618367654,
      "flos": 17565884628480.0,
      "grad_norm": 2.810253293244863,
      "language_loss": 0.76899689,
      "learning_rate": 3.97649990716259e-06,
      "loss": 0.79141217,
      "num_input_tokens_seen": 27348320,
      "step": 1287,
      "time_per_iteration": 2.669168472290039
    },
    {
      "auxiliary_loss_clip": 0.011763,
      "auxiliary_loss_mlp": 0.01061108,
      "balance_loss_clip": 1.05891848,
      "balance_loss_mlp": 1.03696775,
      "epoch": 0.0774387494363445,
      "flos": 25627147349760.0,
      "grad_norm": 1.6525652726351308,
      "language_loss": 0.84699571,
      "learning_rate": 3.976440341863237e-06,
      "loss": 0.86936986,
      "num_input_tokens_seen": 27367670,
      "step": 1288,
      "time_per_iteration": 2.7794599533081055
    },
    {
      "auxiliary_loss_clip": 0.01206182,
      "auxiliary_loss_mlp": 0.0106604,
      "balance_loss_clip": 1.06214797,
      "balance_loss_mlp": 1.04203176,
      "epoch": 0.07749887268901248,
      "flos": 12239865648000.0,
      "grad_norm": 2.0424090794957523,
      "language_loss": 0.85576034,
      "learning_rate": 3.976380701617068e-06,
      "loss": 0.87848258,
      "num_input_tokens_seen": 27385485,
      "step": 1289,
      "time_per_iteration": 4.232934236526489
    },
    {
      "auxiliary_loss_clip": 0.01207527,
      "auxiliary_loss_mlp": 0.01052975,
      "balance_loss_clip": 1.06487668,
      "balance_loss_mlp": 1.0291574,
      "epoch": 0.07755899594168045,
      "flos": 25081736261760.0,
      "grad_norm": 2.840721047922519,
      "language_loss": 0.85548425,
      "learning_rate": 3.976320986426344e-06,
      "loss": 0.87808931,
      "num_input_tokens_seen": 27405110,
      "step": 1290,
      "time_per_iteration": 4.218302965164185
    },
    {
      "auxiliary_loss_clip": 0.0117374,
      "auxiliary_loss_mlp": 0.01066698,
      "balance_loss_clip": 1.06411862,
      "balance_loss_mlp": 1.04041266,
      "epoch": 0.07761911919434841,
      "flos": 14246410176000.0,
      "grad_norm": 2.3756178078405976,
      "language_loss": 0.91390574,
      "learning_rate": 3.9762611962933315e-06,
      "loss": 0.93631011,
      "num_input_tokens_seen": 27422855,
      "step": 1291,
      "time_per_iteration": 4.468304395675659
    },
    {
      "auxiliary_loss_clip": 0.01081301,
      "auxiliary_loss_mlp": 0.01043026,
      "balance_loss_clip": 1.04092944,
      "balance_loss_mlp": 1.03894901,
      "epoch": 0.07767924244701638,
      "flos": 67237202954880.0,
      "grad_norm": 0.8973948861970446,
      "language_loss": 0.65065891,
      "learning_rate": 3.9762013312202955e-06,
      "loss": 0.67190224,
      "num_input_tokens_seen": 27487190,
      "step": 1292,
      "time_per_iteration": 3.3142755031585693
    },
    {
      "auxiliary_loss_clip": 0.01195822,
      "auxiliary_loss_mlp": 0.01062751,
      "balance_loss_clip": 1.06527543,
      "balance_loss_mlp": 1.03846776,
      "epoch": 0.07773936569968436,
      "flos": 28550635292160.0,
      "grad_norm": 1.7595227960044768,
      "language_loss": 0.87530363,
      "learning_rate": 3.9761413912095075e-06,
      "loss": 0.89788938,
      "num_input_tokens_seen": 27510465,
      "step": 1293,
      "time_per_iteration": 2.801603078842163
    },
    {
      "auxiliary_loss_clip": 0.01116633,
      "auxiliary_loss_mlp": 0.01078659,
      "balance_loss_clip": 1.05041039,
      "balance_loss_mlp": 1.05012059,
      "epoch": 0.07779948895235232,
      "flos": 27490264871040.0,
      "grad_norm": 2.2898991349098528,
      "language_loss": 0.84518278,
      "learning_rate": 3.976081376263239e-06,
      "loss": 0.8671357,
      "num_input_tokens_seen": 27528645,
      "step": 1294,
      "time_per_iteration": 2.898597002029419
    },
    {
      "auxiliary_loss_clip": 0.01158796,
      "auxiliary_loss_mlp": 0.01059505,
      "balance_loss_clip": 1.05967593,
      "balance_loss_mlp": 1.0342207,
      "epoch": 0.07785961220502029,
      "flos": 18223301301120.0,
      "grad_norm": 2.7292442592472073,
      "language_loss": 0.79365373,
      "learning_rate": 3.976021286383768e-06,
      "loss": 0.81583679,
      "num_input_tokens_seen": 27546165,
      "step": 1295,
      "time_per_iteration": 2.8481552600860596
    },
    {
      "auxiliary_loss_clip": 0.01155886,
      "auxiliary_loss_mlp": 0.01061351,
      "balance_loss_clip": 1.06015158,
      "balance_loss_mlp": 1.0356493,
      "epoch": 0.07791973545768827,
      "flos": 24608218245120.0,
      "grad_norm": 3.472740252224496,
      "language_loss": 0.88351864,
      "learning_rate": 3.975961121573371e-06,
      "loss": 0.90569103,
      "num_input_tokens_seen": 27566520,
      "step": 1296,
      "time_per_iteration": 2.697831392288208
    },
    {
      "auxiliary_loss_clip": 0.0120756,
      "auxiliary_loss_mlp": 0.01074146,
      "balance_loss_clip": 1.06552935,
      "balance_loss_mlp": 1.04791999,
      "epoch": 0.07797985871035623,
      "flos": 14282069402880.0,
      "grad_norm": 2.384603846473911,
      "language_loss": 0.9625901,
      "learning_rate": 3.9759008818343305e-06,
      "loss": 0.98540717,
      "num_input_tokens_seen": 27581960,
      "step": 1297,
      "time_per_iteration": 2.62660551071167
    },
    {
      "auxiliary_loss_clip": 0.01175852,
      "auxiliary_loss_mlp": 0.01069298,
      "balance_loss_clip": 1.06147313,
      "balance_loss_mlp": 1.04517019,
      "epoch": 0.0780399819630242,
      "flos": 26610453141120.0,
      "grad_norm": 2.15152040651991,
      "language_loss": 0.7600193,
      "learning_rate": 3.97584056716893e-06,
      "loss": 0.78247076,
      "num_input_tokens_seen": 27601415,
      "step": 1298,
      "time_per_iteration": 2.8040499687194824
    },
    {
      "auxiliary_loss_clip": 0.0114505,
      "auxiliary_loss_mlp": 0.00783981,
      "balance_loss_clip": 1.05864501,
      "balance_loss_mlp": 1.0006063,
      "epoch": 0.07810010521569218,
      "flos": 21834514016640.0,
      "grad_norm": 1.6697657327886877,
      "language_loss": 0.8097105,
      "learning_rate": 3.9757801775794575e-06,
      "loss": 0.82900077,
      "num_input_tokens_seen": 27621490,
      "step": 1299,
      "time_per_iteration": 2.7667653560638428
    },
    {
      "auxiliary_loss_clip": 0.01162638,
      "auxiliary_loss_mlp": 0.01064395,
      "balance_loss_clip": 1.06191885,
      "balance_loss_mlp": 1.0393368,
      "epoch": 0.07816022846836014,
      "flos": 25081233471360.0,
      "grad_norm": 1.9748762517467437,
      "language_loss": 0.86755943,
      "learning_rate": 3.975719713068202e-06,
      "loss": 0.8898297,
      "num_input_tokens_seen": 27640600,
      "step": 1300,
      "time_per_iteration": 2.7819204330444336
    },
    {
      "auxiliary_loss_clip": 0.0120807,
      "auxiliary_loss_mlp": 0.01056805,
      "balance_loss_clip": 1.06663537,
      "balance_loss_mlp": 1.03180683,
      "epoch": 0.0782203517210281,
      "flos": 40917515431680.0,
      "grad_norm": 3.040560411644486,
      "language_loss": 0.71822268,
      "learning_rate": 3.975659173637458e-06,
      "loss": 0.74087137,
      "num_input_tokens_seen": 27663070,
      "step": 1301,
      "time_per_iteration": 2.845107316970825
    },
    {
      "auxiliary_loss_clip": 0.01196566,
      "auxiliary_loss_mlp": 0.01075534,
      "balance_loss_clip": 1.06426311,
      "balance_loss_mlp": 1.05100083,
      "epoch": 0.07828047497369607,
      "flos": 41172014269440.0,
      "grad_norm": 1.6425838754876312,
      "language_loss": 0.70782864,
      "learning_rate": 3.97559855928952e-06,
      "loss": 0.73054957,
      "num_input_tokens_seen": 27686425,
      "step": 1302,
      "time_per_iteration": 2.898069381713867
    },
    {
      "auxiliary_loss_clip": 0.01162032,
      "auxiliary_loss_mlp": 0.00783256,
      "balance_loss_clip": 1.06019354,
      "balance_loss_mlp": 1.00062823,
      "epoch": 0.07834059822636405,
      "flos": 23508130360320.0,
      "grad_norm": 2.067506704059933,
      "language_loss": 0.82100385,
      "learning_rate": 3.9755378700266864e-06,
      "loss": 0.84045678,
      "num_input_tokens_seen": 27704900,
      "step": 1303,
      "time_per_iteration": 2.7862839698791504
    },
    {
      "auxiliary_loss_clip": 0.01191742,
      "auxiliary_loss_mlp": 0.01074585,
      "balance_loss_clip": 1.06583321,
      "balance_loss_mlp": 1.04908574,
      "epoch": 0.07840072147903202,
      "flos": 20193899293440.0,
      "grad_norm": 1.8830773419754625,
      "language_loss": 0.75206572,
      "learning_rate": 3.9754771058512585e-06,
      "loss": 0.77472901,
      "num_input_tokens_seen": 27724890,
      "step": 1304,
      "time_per_iteration": 2.7380170822143555
    },
    {
      "auxiliary_loss_clip": 0.01211207,
      "auxiliary_loss_mlp": 0.01074343,
      "balance_loss_clip": 1.07114935,
      "balance_loss_mlp": 1.04922605,
      "epoch": 0.07846084473169998,
      "flos": 21360816432000.0,
      "grad_norm": 1.6118444643214749,
      "language_loss": 0.76141047,
      "learning_rate": 3.975416266765542e-06,
      "loss": 0.784266,
      "num_input_tokens_seen": 27743115,
      "step": 1305,
      "time_per_iteration": 2.6788928508758545
    },
    {
      "auxiliary_loss_clip": 0.01137547,
      "auxiliary_loss_mlp": 0.01064795,
      "balance_loss_clip": 1.05611205,
      "balance_loss_mlp": 1.04021358,
      "epoch": 0.07852096798436796,
      "flos": 25410965345280.0,
      "grad_norm": 1.9541638070229452,
      "language_loss": 0.85011744,
      "learning_rate": 3.975355352771841e-06,
      "loss": 0.87214082,
      "num_input_tokens_seen": 27763570,
      "step": 1306,
      "time_per_iteration": 3.048137903213501
    },
    {
      "auxiliary_loss_clip": 0.01194779,
      "auxiliary_loss_mlp": 0.01049822,
      "balance_loss_clip": 1.06754708,
      "balance_loss_mlp": 1.02668333,
      "epoch": 0.07858109123703592,
      "flos": 24571481610240.0,
      "grad_norm": 6.108459548145404,
      "language_loss": 0.90882134,
      "learning_rate": 3.975294363872468e-06,
      "loss": 0.93126732,
      "num_input_tokens_seen": 27780030,
      "step": 1307,
      "time_per_iteration": 3.1597135066986084
    },
    {
      "auxiliary_loss_clip": 0.01145989,
      "auxiliary_loss_mlp": 0.01060478,
      "balance_loss_clip": 1.05529833,
      "balance_loss_mlp": 1.034729,
      "epoch": 0.07864121448970389,
      "flos": 20698874645760.0,
      "grad_norm": 3.4991416096159136,
      "language_loss": 0.83695096,
      "learning_rate": 3.975233300069735e-06,
      "loss": 0.85901558,
      "num_input_tokens_seen": 27796225,
      "step": 1308,
      "time_per_iteration": 2.749174118041992
    },
    {
      "auxiliary_loss_clip": 0.01151044,
      "auxiliary_loss_mlp": 0.01061966,
      "balance_loss_clip": 1.05445218,
      "balance_loss_mlp": 1.03789735,
      "epoch": 0.07870133774237187,
      "flos": 22966526113920.0,
      "grad_norm": 1.7092634116882437,
      "language_loss": 0.77521002,
      "learning_rate": 3.975172161365958e-06,
      "loss": 0.7973401,
      "num_input_tokens_seen": 27815975,
      "step": 1309,
      "time_per_iteration": 2.752854108810425
    },
    {
      "auxiliary_loss_clip": 0.01200102,
      "auxiliary_loss_mlp": 0.01070583,
      "balance_loss_clip": 1.06396675,
      "balance_loss_mlp": 1.04449987,
      "epoch": 0.07876146099503983,
      "flos": 18842832103680.0,
      "grad_norm": 1.8729662604656268,
      "language_loss": 0.80561006,
      "learning_rate": 3.975110947763453e-06,
      "loss": 0.82831693,
      "num_input_tokens_seen": 27832255,
      "step": 1310,
      "time_per_iteration": 2.6966710090637207
    },
    {
      "auxiliary_loss_clip": 0.01173381,
      "auxiliary_loss_mlp": 0.0078245,
      "balance_loss_clip": 1.06193507,
      "balance_loss_mlp": 1.00060987,
      "epoch": 0.0788215842477078,
      "flos": 23805794367360.0,
      "grad_norm": 1.796715978968241,
      "language_loss": 0.73187977,
      "learning_rate": 3.9750496592645435e-06,
      "loss": 0.75143808,
      "num_input_tokens_seen": 27852180,
      "step": 1311,
      "time_per_iteration": 2.7588090896606445
    },
    {
      "auxiliary_loss_clip": 0.01188438,
      "auxiliary_loss_mlp": 0.01078546,
      "balance_loss_clip": 1.06358969,
      "balance_loss_mlp": 1.05342865,
      "epoch": 0.07888170750037576,
      "flos": 21579907438080.0,
      "grad_norm": 1.7490617386556226,
      "language_loss": 0.86002982,
      "learning_rate": 3.974988295871553e-06,
      "loss": 0.88269973,
      "num_input_tokens_seen": 27871435,
      "step": 1312,
      "time_per_iteration": 2.6969683170318604
    },
    {
      "auxiliary_loss_clip": 0.01178338,
      "auxiliary_loss_mlp": 0.01059112,
      "balance_loss_clip": 1.06324685,
      "balance_loss_mlp": 1.03633142,
      "epoch": 0.07894183075304374,
      "flos": 19864849777920.0,
      "grad_norm": 1.825664315845032,
      "language_loss": 0.82087892,
      "learning_rate": 3.9749268575868085e-06,
      "loss": 0.84325337,
      "num_input_tokens_seen": 27890625,
      "step": 1313,
      "time_per_iteration": 2.6936304569244385
    },
    {
      "auxiliary_loss_clip": 0.01184798,
      "auxiliary_loss_mlp": 0.00783631,
      "balance_loss_clip": 1.06229842,
      "balance_loss_mlp": 1.00053823,
      "epoch": 0.07900195400571171,
      "flos": 16143463071360.0,
      "grad_norm": 2.837190319075622,
      "language_loss": 0.73569417,
      "learning_rate": 3.97486534441264e-06,
      "loss": 0.75537837,
      "num_input_tokens_seen": 27906530,
      "step": 1314,
      "time_per_iteration": 2.653505325317383
    },
    {
      "auxiliary_loss_clip": 0.01154585,
      "auxiliary_loss_mlp": 0.00782352,
      "balance_loss_clip": 1.05730104,
      "balance_loss_mlp": 1.00044668,
      "epoch": 0.07906207725837967,
      "flos": 23730417676800.0,
      "grad_norm": 1.6153694611764058,
      "language_loss": 0.79490477,
      "learning_rate": 3.974803756351379e-06,
      "loss": 0.81427419,
      "num_input_tokens_seen": 27926725,
      "step": 1315,
      "time_per_iteration": 2.797306776046753
    },
    {
      "auxiliary_loss_clip": 0.01189107,
      "auxiliary_loss_mlp": 0.01060743,
      "balance_loss_clip": 1.05841756,
      "balance_loss_mlp": 1.03487444,
      "epoch": 0.07912220051104765,
      "flos": 24315905364480.0,
      "grad_norm": 1.6362349035659796,
      "language_loss": 0.73546493,
      "learning_rate": 3.974742093405362e-06,
      "loss": 0.75796348,
      "num_input_tokens_seen": 27947875,
      "step": 1316,
      "time_per_iteration": 2.688997507095337
    },
    {
      "auxiliary_loss_clip": 0.01162651,
      "auxiliary_loss_mlp": 0.01066617,
      "balance_loss_clip": 1.05845332,
      "balance_loss_mlp": 1.0418098,
      "epoch": 0.07918232376371562,
      "flos": 18880035615360.0,
      "grad_norm": 2.157376902111077,
      "language_loss": 0.65540409,
      "learning_rate": 3.974680355576927e-06,
      "loss": 0.67769682,
      "num_input_tokens_seen": 27965040,
      "step": 1317,
      "time_per_iteration": 2.6998519897460938
    },
    {
      "auxiliary_loss_clip": 0.01177674,
      "auxiliary_loss_mlp": 0.01068635,
      "balance_loss_clip": 1.06280386,
      "balance_loss_mlp": 1.0428021,
      "epoch": 0.07924244701638358,
      "flos": 27376284038400.0,
      "grad_norm": 2.382161374765057,
      "language_loss": 0.73105192,
      "learning_rate": 3.974618542868415e-06,
      "loss": 0.75351495,
      "num_input_tokens_seen": 27985330,
      "step": 1318,
      "time_per_iteration": 2.8350789546966553
    },
    {
      "auxiliary_loss_clip": 0.01139638,
      "auxiliary_loss_mlp": 0.01058798,
      "balance_loss_clip": 1.05582452,
      "balance_loss_mlp": 1.03515935,
      "epoch": 0.07930257026905156,
      "flos": 25120340403840.0,
      "grad_norm": 2.635941883481154,
      "language_loss": 0.90381306,
      "learning_rate": 3.97455665528217e-06,
      "loss": 0.92579746,
      "num_input_tokens_seen": 28007615,
      "step": 1319,
      "time_per_iteration": 2.8553895950317383
    },
    {
      "auxiliary_loss_clip": 0.01175059,
      "auxiliary_loss_mlp": 0.01055333,
      "balance_loss_clip": 1.05662942,
      "balance_loss_mlp": 1.03122926,
      "epoch": 0.07936269352171953,
      "flos": 21834478103040.0,
      "grad_norm": 1.9449065990449943,
      "language_loss": 0.80134505,
      "learning_rate": 3.974494692820539e-06,
      "loss": 0.82364893,
      "num_input_tokens_seen": 28027765,
      "step": 1320,
      "time_per_iteration": 2.6651997566223145
    },
    {
      "auxiliary_loss_clip": 0.01181808,
      "auxiliary_loss_mlp": 0.01060151,
      "balance_loss_clip": 1.06380332,
      "balance_loss_mlp": 1.03657198,
      "epoch": 0.07942281677438749,
      "flos": 16939889377920.0,
      "grad_norm": 2.1078540484546746,
      "language_loss": 0.6901226,
      "learning_rate": 3.974432655485872e-06,
      "loss": 0.71254218,
      "num_input_tokens_seen": 28044225,
      "step": 1321,
      "time_per_iteration": 2.6500401496887207
    },
    {
      "auxiliary_loss_clip": 0.01189002,
      "auxiliary_loss_mlp": 0.01060598,
      "balance_loss_clip": 1.06469131,
      "balance_loss_mlp": 1.03688753,
      "epoch": 0.07948294002705546,
      "flos": 18986941468800.0,
      "grad_norm": 1.9310950096267907,
      "language_loss": 0.8359012,
      "learning_rate": 3.9743705432805195e-06,
      "loss": 0.85839725,
      "num_input_tokens_seen": 28062915,
      "step": 1322,
      "time_per_iteration": 2.684978723526001
    },
    {
      "auxiliary_loss_clip": 0.01202147,
      "auxiliary_loss_mlp": 0.01057117,
      "balance_loss_clip": 1.06135976,
      "balance_loss_mlp": 1.03304851,
      "epoch": 0.07954306327972344,
      "flos": 21653452535040.0,
      "grad_norm": 2.128262121046283,
      "language_loss": 0.90555447,
      "learning_rate": 3.974308356206838e-06,
      "loss": 0.92814714,
      "num_input_tokens_seen": 28082175,
      "step": 1323,
      "time_per_iteration": 2.6192240715026855
    },
    {
      "auxiliary_loss_clip": 0.01164151,
      "auxiliary_loss_mlp": 0.01062303,
      "balance_loss_clip": 1.06272292,
      "balance_loss_mlp": 1.03809166,
      "epoch": 0.0796031865323914,
      "flos": 23220270766080.0,
      "grad_norm": 1.8373443631598505,
      "language_loss": 0.82521075,
      "learning_rate": 3.974246094267187e-06,
      "loss": 0.84747529,
      "num_input_tokens_seen": 28102645,
      "step": 1324,
      "time_per_iteration": 2.8283956050872803
    },
    {
      "auxiliary_loss_clip": 0.01180787,
      "auxiliary_loss_mlp": 0.01053463,
      "balance_loss_clip": 1.06256735,
      "balance_loss_mlp": 1.02834535,
      "epoch": 0.07966330978505937,
      "flos": 23294534135040.0,
      "grad_norm": 2.119290865165494,
      "language_loss": 0.79162025,
      "learning_rate": 3.974183757463925e-06,
      "loss": 0.8139627,
      "num_input_tokens_seen": 28122805,
      "step": 1325,
      "time_per_iteration": 2.6996092796325684
    },
    {
      "auxiliary_loss_clip": 0.01119286,
      "auxiliary_loss_mlp": 0.00785175,
      "balance_loss_clip": 1.04844928,
      "balance_loss_mlp": 1.00035501,
      "epoch": 0.07972343303772735,
      "flos": 18363783392640.0,
      "grad_norm": 2.2621745256944448,
      "language_loss": 0.88038248,
      "learning_rate": 3.974121345799418e-06,
      "loss": 0.89942712,
      "num_input_tokens_seen": 28140530,
      "step": 1326,
      "time_per_iteration": 2.881410837173462
    },
    {
      "auxiliary_loss_clip": 0.012,
      "auxiliary_loss_mlp": 0.01056877,
      "balance_loss_clip": 1.06257951,
      "balance_loss_mlp": 1.03168797,
      "epoch": 0.07978355629039531,
      "flos": 21762513204480.0,
      "grad_norm": 1.8538865301137586,
      "language_loss": 0.8328709,
      "learning_rate": 3.974058859276032e-06,
      "loss": 0.85543966,
      "num_input_tokens_seen": 28159640,
      "step": 1327,
      "time_per_iteration": 2.7277982234954834
    },
    {
      "auxiliary_loss_clip": 0.01207207,
      "auxiliary_loss_mlp": 0.01056886,
      "balance_loss_clip": 1.06532371,
      "balance_loss_mlp": 1.03223395,
      "epoch": 0.07984367954306328,
      "flos": 18551309322240.0,
      "grad_norm": 2.3216818645515636,
      "language_loss": 0.78599,
      "learning_rate": 3.9739962978961354e-06,
      "loss": 0.80863088,
      "num_input_tokens_seen": 28177050,
      "step": 1328,
      "time_per_iteration": 4.2137157917022705
    },
    {
      "auxiliary_loss_clip": 0.01201442,
      "auxiliary_loss_mlp": 0.01052053,
      "balance_loss_clip": 1.06778932,
      "balance_loss_mlp": 1.02722156,
      "epoch": 0.07990380279573125,
      "flos": 16904050583040.0,
      "grad_norm": 4.209530911932697,
      "language_loss": 0.73918134,
      "learning_rate": 3.973933661662101e-06,
      "loss": 0.76171625,
      "num_input_tokens_seen": 28193245,
      "step": 1329,
      "time_per_iteration": 5.853717565536499
    },
    {
      "auxiliary_loss_clip": 0.01169795,
      "auxiliary_loss_mlp": 0.01064631,
      "balance_loss_clip": 1.06039059,
      "balance_loss_mlp": 1.04069376,
      "epoch": 0.07996392604839922,
      "flos": 24098358643200.0,
      "grad_norm": 1.6102544328312476,
      "language_loss": 0.81743932,
      "learning_rate": 3.973870950576305e-06,
      "loss": 0.83978355,
      "num_input_tokens_seen": 28213570,
      "step": 1330,
      "time_per_iteration": 4.307915687561035
    },
    {
      "auxiliary_loss_clip": 0.01205148,
      "auxiliary_loss_mlp": 0.00780735,
      "balance_loss_clip": 1.06445098,
      "balance_loss_mlp": 1.00030971,
      "epoch": 0.08002404930106718,
      "flos": 14278729438080.0,
      "grad_norm": 3.0935981151455865,
      "language_loss": 0.88962448,
      "learning_rate": 3.9738081646411255e-06,
      "loss": 0.90948325,
      "num_input_tokens_seen": 28229980,
      "step": 1331,
      "time_per_iteration": 2.645198345184326
    },
    {
      "auxiliary_loss_clip": 0.01196019,
      "auxiliary_loss_mlp": 0.00781409,
      "balance_loss_clip": 1.05950165,
      "balance_loss_mlp": 1.00032377,
      "epoch": 0.08008417255373516,
      "flos": 40406219285760.0,
      "grad_norm": 1.8933982437719925,
      "language_loss": 0.7335732,
      "learning_rate": 3.973745303858942e-06,
      "loss": 0.75334752,
      "num_input_tokens_seen": 28253840,
      "step": 1332,
      "time_per_iteration": 2.792128562927246
    },
    {
      "auxiliary_loss_clip": 0.01180359,
      "auxiliary_loss_mlp": 0.01055118,
      "balance_loss_clip": 1.06217384,
      "balance_loss_mlp": 1.03216982,
      "epoch": 0.08014429580640313,
      "flos": 18478913460480.0,
      "grad_norm": 1.7464568676953767,
      "language_loss": 0.82765031,
      "learning_rate": 3.973682368232138e-06,
      "loss": 0.85000509,
      "num_input_tokens_seen": 28271675,
      "step": 1333,
      "time_per_iteration": 2.635579824447632
    },
    {
      "auxiliary_loss_clip": 0.01160554,
      "auxiliary_loss_mlp": 0.01059025,
      "balance_loss_clip": 1.05944169,
      "balance_loss_mlp": 1.03502798,
      "epoch": 0.0802044190590711,
      "flos": 22053461368320.0,
      "grad_norm": 2.677615191761892,
      "language_loss": 0.74862051,
      "learning_rate": 3.9736193577631015e-06,
      "loss": 0.77081633,
      "num_input_tokens_seen": 28291850,
      "step": 1334,
      "time_per_iteration": 2.8150298595428467
    },
    {
      "auxiliary_loss_clip": 0.01176175,
      "auxiliary_loss_mlp": 0.01063593,
      "balance_loss_clip": 1.06460369,
      "balance_loss_mlp": 1.04010868,
      "epoch": 0.08026454231173906,
      "flos": 24572128055040.0,
      "grad_norm": 1.8723728369534094,
      "language_loss": 0.79970533,
      "learning_rate": 3.973556272454221e-06,
      "loss": 0.82210302,
      "num_input_tokens_seen": 28310780,
      "step": 1335,
      "time_per_iteration": 2.6858503818511963
    },
    {
      "auxiliary_loss_clip": 0.01068232,
      "auxiliary_loss_mlp": 0.01020395,
      "balance_loss_clip": 1.04101062,
      "balance_loss_mlp": 1.01693749,
      "epoch": 0.08032466556440704,
      "flos": 52581841459200.0,
      "grad_norm": 0.7491611763509133,
      "language_loss": 0.56056821,
      "learning_rate": 3.973493112307889e-06,
      "loss": 0.58145452,
      "num_input_tokens_seen": 28369985,
      "step": 1336,
      "time_per_iteration": 3.324230670928955
    },
    {
      "auxiliary_loss_clip": 0.01179495,
      "auxiliary_loss_mlp": 0.01064433,
      "balance_loss_clip": 1.06005239,
      "balance_loss_mlp": 1.04149771,
      "epoch": 0.080384788817075,
      "flos": 23842602829440.0,
      "grad_norm": 2.8990759307469256,
      "language_loss": 0.67587668,
      "learning_rate": 3.9734298773265005e-06,
      "loss": 0.69831598,
      "num_input_tokens_seen": 28388670,
      "step": 1337,
      "time_per_iteration": 2.755451202392578
    },
    {
      "auxiliary_loss_clip": 0.01171763,
      "auxiliary_loss_mlp": 0.0107788,
      "balance_loss_clip": 1.06270492,
      "balance_loss_mlp": 1.05304837,
      "epoch": 0.08044491206974297,
      "flos": 25300719527040.0,
      "grad_norm": 1.9421039451316542,
      "language_loss": 0.86847901,
      "learning_rate": 3.973366567512453e-06,
      "loss": 0.89097536,
      "num_input_tokens_seen": 28411845,
      "step": 1338,
      "time_per_iteration": 2.758418560028076
    },
    {
      "auxiliary_loss_clip": 0.01136344,
      "auxiliary_loss_mlp": 0.01082295,
      "balance_loss_clip": 1.04883683,
      "balance_loss_mlp": 1.05596161,
      "epoch": 0.08050503532241095,
      "flos": 22376549226240.0,
      "grad_norm": 2.4557709650828157,
      "language_loss": 0.87217385,
      "learning_rate": 3.973303182868147e-06,
      "loss": 0.89436018,
      "num_input_tokens_seen": 28427875,
      "step": 1339,
      "time_per_iteration": 2.72682785987854
    },
    {
      "auxiliary_loss_clip": 0.01188632,
      "auxiliary_loss_mlp": 0.01055953,
      "balance_loss_clip": 1.06334567,
      "balance_loss_mlp": 1.03417385,
      "epoch": 0.08056515857507891,
      "flos": 18369421827840.0,
      "grad_norm": 10.603370056653041,
      "language_loss": 0.89504963,
      "learning_rate": 3.973239723395988e-06,
      "loss": 0.91749549,
      "num_input_tokens_seen": 28446615,
      "step": 1340,
      "time_per_iteration": 2.639601469039917
    },
    {
      "auxiliary_loss_clip": 0.01080107,
      "auxiliary_loss_mlp": 0.01012224,
      "balance_loss_clip": 1.02943289,
      "balance_loss_mlp": 1.00850451,
      "epoch": 0.08062528182774688,
      "flos": 51348130980480.0,
      "grad_norm": 0.8861598592181924,
      "language_loss": 0.64834231,
      "learning_rate": 3.97317618909838e-06,
      "loss": 0.66926563,
      "num_input_tokens_seen": 28505290,
      "step": 1341,
      "time_per_iteration": 3.0625648498535156
    },
    {
      "auxiliary_loss_clip": 0.01197538,
      "auxiliary_loss_mlp": 0.01061885,
      "balance_loss_clip": 1.0628854,
      "balance_loss_mlp": 1.0364095,
      "epoch": 0.08068540508041486,
      "flos": 17599712261760.0,
      "grad_norm": 3.3156125209451286,
      "language_loss": 0.89471233,
      "learning_rate": 3.973112579977733e-06,
      "loss": 0.9173066,
      "num_input_tokens_seen": 28522735,
      "step": 1342,
      "time_per_iteration": 2.6123783588409424
    },
    {
      "auxiliary_loss_clip": 0.01177687,
      "auxiliary_loss_mlp": 0.01062063,
      "balance_loss_clip": 1.0644995,
      "balance_loss_mlp": 1.03818512,
      "epoch": 0.08074552833308282,
      "flos": 10561185486720.0,
      "grad_norm": 2.2904075751929365,
      "language_loss": 0.76354575,
      "learning_rate": 3.973048896036459e-06,
      "loss": 0.78594327,
      "num_input_tokens_seen": 28539460,
      "step": 1343,
      "time_per_iteration": 2.7564918994903564
    },
    {
      "auxiliary_loss_clip": 0.01064182,
      "auxiliary_loss_mlp": 0.01010488,
      "balance_loss_clip": 1.02542567,
      "balance_loss_mlp": 1.0066731,
      "epoch": 0.08080565158575079,
      "flos": 60840254954880.0,
      "grad_norm": 0.8071281523255156,
      "language_loss": 0.57418531,
      "learning_rate": 3.972985137276974e-06,
      "loss": 0.59493202,
      "num_input_tokens_seen": 28599855,
      "step": 1344,
      "time_per_iteration": 3.170443058013916
    },
    {
      "auxiliary_loss_clip": 0.01158029,
      "auxiliary_loss_mlp": 0.01063108,
      "balance_loss_clip": 1.05839872,
      "balance_loss_mlp": 1.03846788,
      "epoch": 0.08086577483841875,
      "flos": 18332361970560.0,
      "grad_norm": 2.5953739346171676,
      "language_loss": 0.86569476,
      "learning_rate": 3.972921303701695e-06,
      "loss": 0.88790607,
      "num_input_tokens_seen": 28617585,
      "step": 1345,
      "time_per_iteration": 2.765254497528076
    },
    {
      "auxiliary_loss_clip": 0.01203428,
      "auxiliary_loss_mlp": 0.01057879,
      "balance_loss_clip": 1.06629944,
      "balance_loss_mlp": 1.03603959,
      "epoch": 0.08092589809108673,
      "flos": 21543601766400.0,
      "grad_norm": 1.8653844332842058,
      "language_loss": 0.87646407,
      "learning_rate": 3.972857395313042e-06,
      "loss": 0.89907712,
      "num_input_tokens_seen": 28636355,
      "step": 1346,
      "time_per_iteration": 2.655611991882324
    },
    {
      "auxiliary_loss_clip": 0.01191822,
      "auxiliary_loss_mlp": 0.0105414,
      "balance_loss_clip": 1.06450033,
      "balance_loss_mlp": 1.03047693,
      "epoch": 0.0809860213437547,
      "flos": 22128012046080.0,
      "grad_norm": 1.7047476553504466,
      "language_loss": 0.9298563,
      "learning_rate": 3.972793412113439e-06,
      "loss": 0.95231593,
      "num_input_tokens_seen": 28656260,
      "step": 1347,
      "time_per_iteration": 2.718355417251587
    },
    {
      "auxiliary_loss_clip": 0.01188696,
      "auxiliary_loss_mlp": 0.01066703,
      "balance_loss_clip": 1.06260633,
      "balance_loss_mlp": 1.04144263,
      "epoch": 0.08104614459642266,
      "flos": 21725489260800.0,
      "grad_norm": 1.9307860049130865,
      "language_loss": 0.89506733,
      "learning_rate": 3.972729354105312e-06,
      "loss": 0.91762137,
      "num_input_tokens_seen": 28675865,
      "step": 1348,
      "time_per_iteration": 2.763735771179199
    },
    {
      "auxiliary_loss_clip": 0.01137961,
      "auxiliary_loss_mlp": 0.01059733,
      "balance_loss_clip": 1.06026649,
      "balance_loss_mlp": 1.03730989,
      "epoch": 0.08110626784909064,
      "flos": 23951878980480.0,
      "grad_norm": 1.6214351378274148,
      "language_loss": 0.76906884,
      "learning_rate": 3.97266522129109e-06,
      "loss": 0.79104578,
      "num_input_tokens_seen": 28696255,
      "step": 1349,
      "time_per_iteration": 2.778050661087036
    },
    {
      "auxiliary_loss_clip": 0.01202122,
      "auxiliary_loss_mlp": 0.01065092,
      "balance_loss_clip": 1.06290889,
      "balance_loss_mlp": 1.04144049,
      "epoch": 0.0811663911017586,
      "flos": 19025689265280.0,
      "grad_norm": 1.777484449358279,
      "language_loss": 0.8877703,
      "learning_rate": 3.972601013673205e-06,
      "loss": 0.91044247,
      "num_input_tokens_seen": 28713905,
      "step": 1350,
      "time_per_iteration": 2.5871450901031494
    },
    {
      "auxiliary_loss_clip": 0.01164889,
      "auxiliary_loss_mlp": 0.00780958,
      "balance_loss_clip": 1.06011164,
      "balance_loss_mlp": 1.00028801,
      "epoch": 0.08122651435442657,
      "flos": 15341290588800.0,
      "grad_norm": 2.7472756845793156,
      "language_loss": 0.82298493,
      "learning_rate": 3.972536731254092e-06,
      "loss": 0.84244347,
      "num_input_tokens_seen": 28732075,
      "step": 1351,
      "time_per_iteration": 2.840271234512329
    },
    {
      "auxiliary_loss_clip": 0.01198177,
      "auxiliary_loss_mlp": 0.01055773,
      "balance_loss_clip": 1.06010592,
      "balance_loss_mlp": 1.03090644,
      "epoch": 0.08128663760709455,
      "flos": 23221563655680.0,
      "grad_norm": 2.2808101252466724,
      "language_loss": 0.75274944,
      "learning_rate": 3.972472374036189e-06,
      "loss": 0.775289,
      "num_input_tokens_seen": 28751150,
      "step": 1352,
      "time_per_iteration": 2.733644485473633
    },
    {
      "auxiliary_loss_clip": 0.01194643,
      "auxiliary_loss_mlp": 0.00783595,
      "balance_loss_clip": 1.06613326,
      "balance_loss_mlp": 1.00036311,
      "epoch": 0.08134676085976252,
      "flos": 22965628273920.0,
      "grad_norm": 1.678520960707938,
      "language_loss": 0.82936156,
      "learning_rate": 3.972407942021935e-06,
      "loss": 0.84914398,
      "num_input_tokens_seen": 28773360,
      "step": 1353,
      "time_per_iteration": 2.742149829864502
    },
    {
      "auxiliary_loss_clip": 0.01068236,
      "auxiliary_loss_mlp": 0.01015932,
      "balance_loss_clip": 1.02440155,
      "balance_loss_mlp": 1.01242769,
      "epoch": 0.08140688411243048,
      "flos": 64322115816960.0,
      "grad_norm": 0.8516312511934722,
      "language_loss": 0.59741521,
      "learning_rate": 3.972343435213775e-06,
      "loss": 0.61825693,
      "num_input_tokens_seen": 28833390,
      "step": 1354,
      "time_per_iteration": 3.1912426948547363
    },
    {
      "auxiliary_loss_clip": 0.01150343,
      "auxiliary_loss_mlp": 0.01058874,
      "balance_loss_clip": 1.0546236,
      "balance_loss_mlp": 1.03583086,
      "epoch": 0.08146700736509845,
      "flos": 22491858862080.0,
      "grad_norm": 2.1234068486581643,
      "language_loss": 0.82310611,
      "learning_rate": 3.972278853614154e-06,
      "loss": 0.84519827,
      "num_input_tokens_seen": 28852430,
      "step": 1355,
      "time_per_iteration": 2.782442808151245
    },
    {
      "auxiliary_loss_clip": 0.01186948,
      "auxiliary_loss_mlp": 0.01062856,
      "balance_loss_clip": 1.0600667,
      "balance_loss_mlp": 1.03801262,
      "epoch": 0.08152713061776642,
      "flos": 20447823513600.0,
      "grad_norm": 1.8366299277102565,
      "language_loss": 0.7135247,
      "learning_rate": 3.972214197225521e-06,
      "loss": 0.73602271,
      "num_input_tokens_seen": 28870685,
      "step": 1356,
      "time_per_iteration": 2.7777554988861084
    },
    {
      "auxiliary_loss_clip": 0.01194666,
      "auxiliary_loss_mlp": 0.01056522,
      "balance_loss_clip": 1.06462216,
      "balance_loss_mlp": 1.03259718,
      "epoch": 0.08158725387043439,
      "flos": 23550218121600.0,
      "grad_norm": 2.050923525150184,
      "language_loss": 0.70426142,
      "learning_rate": 3.972149466050329e-06,
      "loss": 0.72677326,
      "num_input_tokens_seen": 28889860,
      "step": 1357,
      "time_per_iteration": 2.852046012878418
    },
    {
      "auxiliary_loss_clip": 0.01186996,
      "auxiliary_loss_mlp": 0.01054475,
      "balance_loss_clip": 1.06138206,
      "balance_loss_mlp": 1.03070426,
      "epoch": 0.08164737712310235,
      "flos": 22017335264640.0,
      "grad_norm": 2.634204556872777,
      "language_loss": 0.84203482,
      "learning_rate": 3.97208466009103e-06,
      "loss": 0.8644495,
      "num_input_tokens_seen": 28905865,
      "step": 1358,
      "time_per_iteration": 2.7127115726470947
    },
    {
      "auxiliary_loss_clip": 0.01176629,
      "auxiliary_loss_mlp": 0.010566,
      "balance_loss_clip": 1.06037402,
      "balance_loss_mlp": 1.03154182,
      "epoch": 0.08170750037577033,
      "flos": 23367827836800.0,
      "grad_norm": 2.1726272773281097,
      "language_loss": 1.02781308,
      "learning_rate": 3.972019779350084e-06,
      "loss": 1.05014539,
      "num_input_tokens_seen": 28925250,
      "step": 1359,
      "time_per_iteration": 2.7171826362609863
    },
    {
      "auxiliary_loss_clip": 0.01128357,
      "auxiliary_loss_mlp": 0.01056774,
      "balance_loss_clip": 1.05009234,
      "balance_loss_mlp": 1.03263426,
      "epoch": 0.0817676236284383,
      "flos": 28397978490240.0,
      "grad_norm": 2.0494617207464945,
      "language_loss": 0.8313604,
      "learning_rate": 3.971954823829951e-06,
      "loss": 0.85321164,
      "num_input_tokens_seen": 28943445,
      "step": 1360,
      "time_per_iteration": 2.9020919799804688
    },
    {
      "auxiliary_loss_clip": 0.01202956,
      "auxiliary_loss_mlp": 0.0106887,
      "balance_loss_clip": 1.06274688,
      "balance_loss_mlp": 1.04469395,
      "epoch": 0.08182774688110626,
      "flos": 19208905562880.0,
      "grad_norm": 5.2377005088202075,
      "language_loss": 0.72322488,
      "learning_rate": 3.971889793533093e-06,
      "loss": 0.74594313,
      "num_input_tokens_seen": 28962695,
      "step": 1361,
      "time_per_iteration": 2.6643178462982178
    },
    {
      "auxiliary_loss_clip": 0.01166556,
      "auxiliary_loss_mlp": 0.01056311,
      "balance_loss_clip": 1.0552367,
      "balance_loss_mlp": 1.03184962,
      "epoch": 0.08188787013377424,
      "flos": 22784099915520.0,
      "grad_norm": 28.302545492028134,
      "language_loss": 0.76657653,
      "learning_rate": 3.971824688461976e-06,
      "loss": 0.78880513,
      "num_input_tokens_seen": 28982120,
      "step": 1362,
      "time_per_iteration": 2.7439064979553223
    },
    {
      "auxiliary_loss_clip": 0.01199728,
      "auxiliary_loss_mlp": 0.01053492,
      "balance_loss_clip": 1.06350708,
      "balance_loss_mlp": 1.03104496,
      "epoch": 0.08194799338644221,
      "flos": 16468095214080.0,
      "grad_norm": 2.1850191919210338,
      "language_loss": 0.72384715,
      "learning_rate": 3.971759508619069e-06,
      "loss": 0.74637932,
      "num_input_tokens_seen": 28998100,
      "step": 1363,
      "time_per_iteration": 2.7082791328430176
    },
    {
      "auxiliary_loss_clip": 0.01202887,
      "auxiliary_loss_mlp": 0.01066374,
      "balance_loss_clip": 1.06580126,
      "balance_loss_mlp": 1.04083955,
      "epoch": 0.08200811663911017,
      "flos": 23913633974400.0,
      "grad_norm": 2.142285699657122,
      "language_loss": 0.7726444,
      "learning_rate": 3.971694254006844e-06,
      "loss": 0.79533696,
      "num_input_tokens_seen": 29017095,
      "step": 1364,
      "time_per_iteration": 2.777156114578247
    },
    {
      "auxiliary_loss_clip": 0.01135428,
      "auxiliary_loss_mlp": 0.01063854,
      "balance_loss_clip": 1.05182433,
      "balance_loss_mlp": 1.03645968,
      "epoch": 0.08206823989177814,
      "flos": 17896550256000.0,
      "grad_norm": 1.85589982882842,
      "language_loss": 0.82242119,
      "learning_rate": 3.971628924627776e-06,
      "loss": 0.844414,
      "num_input_tokens_seen": 29037240,
      "step": 1365,
      "time_per_iteration": 2.8192803859710693
    },
    {
      "auxiliary_loss_clip": 0.01196582,
      "auxiliary_loss_mlp": 0.01059945,
      "balance_loss_clip": 1.07006347,
      "balance_loss_mlp": 1.03706884,
      "epoch": 0.08212836314444612,
      "flos": 22088186841600.0,
      "grad_norm": 1.7803424706125983,
      "language_loss": 0.82062519,
      "learning_rate": 3.97156352048434e-06,
      "loss": 0.84319043,
      "num_input_tokens_seen": 29056250,
      "step": 1366,
      "time_per_iteration": 2.7482311725616455
    },
    {
      "auxiliary_loss_clip": 0.01153262,
      "auxiliary_loss_mlp": 0.0107233,
      "balance_loss_clip": 1.05320215,
      "balance_loss_mlp": 1.04779685,
      "epoch": 0.08218848639711408,
      "flos": 17597485618560.0,
      "grad_norm": 2.010209091244133,
      "language_loss": 0.81944495,
      "learning_rate": 3.97149804157902e-06,
      "loss": 0.84170091,
      "num_input_tokens_seen": 29073380,
      "step": 1367,
      "time_per_iteration": 4.352729797363281
    },
    {
      "auxiliary_loss_clip": 0.01206125,
      "auxiliary_loss_mlp": 0.01066888,
      "balance_loss_clip": 1.06541765,
      "balance_loss_mlp": 1.04241478,
      "epoch": 0.08224860964978205,
      "flos": 17857838373120.0,
      "grad_norm": 2.518996379768439,
      "language_loss": 0.8331567,
      "learning_rate": 3.9714324879142946e-06,
      "loss": 0.85588682,
      "num_input_tokens_seen": 29091330,
      "step": 1368,
      "time_per_iteration": 6.077457666397095
    },
    {
      "auxiliary_loss_clip": 0.01159992,
      "auxiliary_loss_mlp": 0.01049874,
      "balance_loss_clip": 1.06314564,
      "balance_loss_mlp": 1.02790344,
      "epoch": 0.08230873290245003,
      "flos": 25227533566080.0,
      "grad_norm": 3.198110530618569,
      "language_loss": 0.81336468,
      "learning_rate": 3.971366859492653e-06,
      "loss": 0.8354634,
      "num_input_tokens_seen": 29110375,
      "step": 1369,
      "time_per_iteration": 2.769972085952759
    },
    {
      "auxiliary_loss_clip": 0.01137456,
      "auxiliary_loss_mlp": 0.00781814,
      "balance_loss_clip": 1.05438268,
      "balance_loss_mlp": 1.00027657,
      "epoch": 0.08236885615511799,
      "flos": 31759935753600.0,
      "grad_norm": 2.610758273724768,
      "language_loss": 0.74818152,
      "learning_rate": 3.971301156316582e-06,
      "loss": 0.76737428,
      "num_input_tokens_seen": 29129395,
      "step": 1370,
      "time_per_iteration": 4.497304201126099
    },
    {
      "auxiliary_loss_clip": 0.0115498,
      "auxiliary_loss_mlp": 0.01064278,
      "balance_loss_clip": 1.06403351,
      "balance_loss_mlp": 1.03987551,
      "epoch": 0.08242897940778596,
      "flos": 23185832601600.0,
      "grad_norm": 1.5246391685186451,
      "language_loss": 0.7398203,
      "learning_rate": 3.971235378388573e-06,
      "loss": 0.76201284,
      "num_input_tokens_seen": 29148650,
      "step": 1371,
      "time_per_iteration": 2.758089065551758
    },
    {
      "auxiliary_loss_clip": 0.01097162,
      "auxiliary_loss_mlp": 0.0106614,
      "balance_loss_clip": 1.05124569,
      "balance_loss_mlp": 1.04098701,
      "epoch": 0.08248910266045394,
      "flos": 34491480393600.0,
      "grad_norm": 1.9670948823939327,
      "language_loss": 0.70851803,
      "learning_rate": 3.971169525711122e-06,
      "loss": 0.73015106,
      "num_input_tokens_seen": 29170785,
      "step": 1372,
      "time_per_iteration": 4.069301605224609
    },
    {
      "auxiliary_loss_clip": 0.01162292,
      "auxiliary_loss_mlp": 0.01056859,
      "balance_loss_clip": 1.0571332,
      "balance_loss_mlp": 1.03261209,
      "epoch": 0.0825492259131219,
      "flos": 13436228960640.0,
      "grad_norm": 2.750431245604494,
      "language_loss": 0.88363653,
      "learning_rate": 3.9711035982867246e-06,
      "loss": 0.905828,
      "num_input_tokens_seen": 29185210,
      "step": 1373,
      "time_per_iteration": 3.9346964359283447
    },
    {
      "auxiliary_loss_clip": 0.01147291,
      "auxiliary_loss_mlp": 0.01062343,
      "balance_loss_clip": 1.05334187,
      "balance_loss_mlp": 1.03878665,
      "epoch": 0.08260934916578987,
      "flos": 25812446636160.0,
      "grad_norm": 2.128923272573014,
      "language_loss": 0.82465184,
      "learning_rate": 3.971037596117882e-06,
      "loss": 0.84674811,
      "num_input_tokens_seen": 29205210,
      "step": 1374,
      "time_per_iteration": 2.933377981185913
    },
    {
      "auxiliary_loss_clip": 0.01044322,
      "auxiliary_loss_mlp": 0.01017124,
      "balance_loss_clip": 1.03154135,
      "balance_loss_mlp": 1.0135479,
      "epoch": 0.08266947241845783,
      "flos": 63460009491840.0,
      "grad_norm": 0.8272339650193923,
      "language_loss": 0.60641956,
      "learning_rate": 3.970971519207095e-06,
      "loss": 0.62703401,
      "num_input_tokens_seen": 29265350,
      "step": 1375,
      "time_per_iteration": 3.3287038803100586
    },
    {
      "auxiliary_loss_clip": 0.01060461,
      "auxiliary_loss_mlp": 0.01013653,
      "balance_loss_clip": 1.02398169,
      "balance_loss_mlp": 1.01017237,
      "epoch": 0.08272959567112581,
      "flos": 69993704568960.0,
      "grad_norm": 0.9162492148708097,
      "language_loss": 0.62171799,
      "learning_rate": 3.970905367556871e-06,
      "loss": 0.64245915,
      "num_input_tokens_seen": 29321475,
      "step": 1376,
      "time_per_iteration": 3.218834161758423
    },
    {
      "auxiliary_loss_clip": 0.01159103,
      "auxiliary_loss_mlp": 0.0106347,
      "balance_loss_clip": 1.06229186,
      "balance_loss_mlp": 1.03942561,
      "epoch": 0.08278971892379378,
      "flos": 20413205781120.0,
      "grad_norm": 1.9191670647860084,
      "language_loss": 0.82577401,
      "learning_rate": 3.970839141169718e-06,
      "loss": 0.84799975,
      "num_input_tokens_seen": 29341405,
      "step": 1377,
      "time_per_iteration": 2.8763558864593506
    },
    {
      "auxiliary_loss_clip": 0.01176967,
      "auxiliary_loss_mlp": 0.01054072,
      "balance_loss_clip": 1.06486619,
      "balance_loss_mlp": 1.03011107,
      "epoch": 0.08284984217646174,
      "flos": 26250233598720.0,
      "grad_norm": 1.915539507671093,
      "language_loss": 0.84923226,
      "learning_rate": 3.970772840048147e-06,
      "loss": 0.87154263,
      "num_input_tokens_seen": 29361955,
      "step": 1378,
      "time_per_iteration": 2.8232595920562744
    },
    {
      "auxiliary_loss_clip": 0.01185329,
      "auxiliary_loss_mlp": 0.01058999,
      "balance_loss_clip": 1.06043923,
      "balance_loss_mlp": 1.0344305,
      "epoch": 0.08290996542912972,
      "flos": 27194683852800.0,
      "grad_norm": 6.4689921779024795,
      "language_loss": 0.87319231,
      "learning_rate": 3.970706464194672e-06,
      "loss": 0.8956356,
      "num_input_tokens_seen": 29382395,
      "step": 1379,
      "time_per_iteration": 2.756082534790039
    },
    {
      "auxiliary_loss_clip": 0.01158173,
      "auxiliary_loss_mlp": 0.01061479,
      "balance_loss_clip": 1.05779433,
      "balance_loss_mlp": 1.03829277,
      "epoch": 0.08297008868179769,
      "flos": 38618191146240.0,
      "grad_norm": 2.078993196749275,
      "language_loss": 0.78545237,
      "learning_rate": 3.970640013611812e-06,
      "loss": 0.8076489,
      "num_input_tokens_seen": 29404460,
      "step": 1380,
      "time_per_iteration": 2.9525601863861084
    },
    {
      "auxiliary_loss_clip": 0.01183492,
      "auxiliary_loss_mlp": 0.01059448,
      "balance_loss_clip": 1.06308961,
      "balance_loss_mlp": 1.0344255,
      "epoch": 0.08303021193446565,
      "flos": 19974736460160.0,
      "grad_norm": 2.6608111668609697,
      "language_loss": 0.86125714,
      "learning_rate": 3.970573488302083e-06,
      "loss": 0.88368654,
      "num_input_tokens_seen": 29422675,
      "step": 1381,
      "time_per_iteration": 2.735203742980957
    },
    {
      "auxiliary_loss_clip": 0.01197152,
      "auxiliary_loss_mlp": 0.00781814,
      "balance_loss_clip": 1.06611753,
      "balance_loss_mlp": 1.00034571,
      "epoch": 0.08309033518713363,
      "flos": 13662646341120.0,
      "grad_norm": 2.9433398182948203,
      "language_loss": 0.87471211,
      "learning_rate": 3.970506888268011e-06,
      "loss": 0.89450181,
      "num_input_tokens_seen": 29439840,
      "step": 1382,
      "time_per_iteration": 2.6392617225646973
    },
    {
      "auxiliary_loss_clip": 0.0115996,
      "auxiliary_loss_mlp": 0.01055463,
      "balance_loss_clip": 1.06138313,
      "balance_loss_mlp": 1.03337312,
      "epoch": 0.0831504584398016,
      "flos": 17968551068160.0,
      "grad_norm": 1.9901989904031434,
      "language_loss": 0.77085757,
      "learning_rate": 3.970440213512121e-06,
      "loss": 0.79301178,
      "num_input_tokens_seen": 29457360,
      "step": 1383,
      "time_per_iteration": 2.756565809249878
    },
    {
      "auxiliary_loss_clip": 0.01191549,
      "auxiliary_loss_mlp": 0.01058014,
      "balance_loss_clip": 1.06211782,
      "balance_loss_mlp": 1.03395748,
      "epoch": 0.08321058169246956,
      "flos": 22601386408320.0,
      "grad_norm": 1.818236548161018,
      "language_loss": 0.82858944,
      "learning_rate": 3.97037346403694e-06,
      "loss": 0.85108507,
      "num_input_tokens_seen": 29477040,
      "step": 1384,
      "time_per_iteration": 2.7848587036132812
    },
    {
      "auxiliary_loss_clip": 0.01148661,
      "auxiliary_loss_mlp": 0.01063605,
      "balance_loss_clip": 1.05671442,
      "balance_loss_mlp": 1.03610373,
      "epoch": 0.08327070494513754,
      "flos": 22850426378880.0,
      "grad_norm": 3.9982776391866346,
      "language_loss": 0.85219657,
      "learning_rate": 3.970306639845e-06,
      "loss": 0.8743192,
      "num_input_tokens_seen": 29492010,
      "step": 1385,
      "time_per_iteration": 2.803893566131592
    },
    {
      "auxiliary_loss_clip": 0.01157001,
      "auxiliary_loss_mlp": 0.01061891,
      "balance_loss_clip": 1.05823874,
      "balance_loss_mlp": 1.03750122,
      "epoch": 0.0833308281978055,
      "flos": 22782986593920.0,
      "grad_norm": 1.7071515381676081,
      "language_loss": 0.69195282,
      "learning_rate": 3.970239740938835e-06,
      "loss": 0.71414173,
      "num_input_tokens_seen": 29511850,
      "step": 1386,
      "time_per_iteration": 3.004786252975464
    },
    {
      "auxiliary_loss_clip": 0.01172803,
      "auxiliary_loss_mlp": 0.01058809,
      "balance_loss_clip": 1.05489016,
      "balance_loss_mlp": 1.03483546,
      "epoch": 0.08339095145047347,
      "flos": 20812604083200.0,
      "grad_norm": 1.672791522425571,
      "language_loss": 0.81894958,
      "learning_rate": 3.97017276732098e-06,
      "loss": 0.84126568,
      "num_input_tokens_seen": 29531415,
      "step": 1387,
      "time_per_iteration": 2.7678542137145996
    },
    {
      "auxiliary_loss_clip": 0.01179554,
      "auxiliary_loss_mlp": 0.01074251,
      "balance_loss_clip": 1.06179345,
      "balance_loss_mlp": 1.04817975,
      "epoch": 0.08345107470314143,
      "flos": 18515326872960.0,
      "grad_norm": 2.071322011459688,
      "language_loss": 0.77205479,
      "learning_rate": 3.970105718993978e-06,
      "loss": 0.7945928,
      "num_input_tokens_seen": 29549525,
      "step": 1388,
      "time_per_iteration": 2.8246304988861084
    },
    {
      "auxiliary_loss_clip": 0.01130856,
      "auxiliary_loss_mlp": 0.01062414,
      "balance_loss_clip": 1.05684018,
      "balance_loss_mlp": 1.03742766,
      "epoch": 0.08351119795580941,
      "flos": 18807567926400.0,
      "grad_norm": 2.0255270252506636,
      "language_loss": 0.79527366,
      "learning_rate": 3.970038595960369e-06,
      "loss": 0.81720638,
      "num_input_tokens_seen": 29568705,
      "step": 1389,
      "time_per_iteration": 2.8606414794921875
    },
    {
      "auxiliary_loss_clip": 0.01172785,
      "auxiliary_loss_mlp": 0.01064077,
      "balance_loss_clip": 1.05787444,
      "balance_loss_mlp": 1.03923428,
      "epoch": 0.08357132120847738,
      "flos": 18441817689600.0,
      "grad_norm": 2.546615132743645,
      "language_loss": 0.87427586,
      "learning_rate": 3.969971398222699e-06,
      "loss": 0.89664447,
      "num_input_tokens_seen": 29585855,
      "step": 1390,
      "time_per_iteration": 2.795931577682495
    },
    {
      "auxiliary_loss_clip": 0.01160426,
      "auxiliary_loss_mlp": 0.01067723,
      "balance_loss_clip": 1.05447149,
      "balance_loss_mlp": 1.04082966,
      "epoch": 0.08363144446114534,
      "flos": 25922333318400.0,
      "grad_norm": 1.8703157168219726,
      "language_loss": 0.86833143,
      "learning_rate": 3.969904125783517e-06,
      "loss": 0.89061296,
      "num_input_tokens_seen": 29607280,
      "step": 1391,
      "time_per_iteration": 2.811598062515259
    },
    {
      "auxiliary_loss_clip": 0.01156119,
      "auxiliary_loss_mlp": 0.01076482,
      "balance_loss_clip": 1.05575848,
      "balance_loss_mlp": 1.05180562,
      "epoch": 0.08369156771381332,
      "flos": 18041306065920.0,
      "grad_norm": 3.7979396758909263,
      "language_loss": 0.87688571,
      "learning_rate": 3.969836778645371e-06,
      "loss": 0.89921176,
      "num_input_tokens_seen": 29624130,
      "step": 1392,
      "time_per_iteration": 2.776819944381714
    },
    {
      "auxiliary_loss_clip": 0.01183316,
      "auxiliary_loss_mlp": 0.01058545,
      "balance_loss_clip": 1.05830503,
      "balance_loss_mlp": 1.03500128,
      "epoch": 0.08375169096648129,
      "flos": 22675111073280.0,
      "grad_norm": 8.95243370865895,
      "language_loss": 0.80574775,
      "learning_rate": 3.969769356810819e-06,
      "loss": 0.82816637,
      "num_input_tokens_seen": 29643210,
      "step": 1393,
      "time_per_iteration": 2.735761880874634
    },
    {
      "auxiliary_loss_clip": 0.01197686,
      "auxiliary_loss_mlp": 0.01058125,
      "balance_loss_clip": 1.06329441,
      "balance_loss_mlp": 1.03466487,
      "epoch": 0.08381181421914925,
      "flos": 26103215232000.0,
      "grad_norm": 1.7485261130451684,
      "language_loss": 0.85064757,
      "learning_rate": 3.969701860282415e-06,
      "loss": 0.87320572,
      "num_input_tokens_seen": 29663920,
      "step": 1394,
      "time_per_iteration": 2.950211524963379
    },
    {
      "auxiliary_loss_clip": 0.01145594,
      "auxiliary_loss_mlp": 0.01058123,
      "balance_loss_clip": 1.05994248,
      "balance_loss_mlp": 1.03432918,
      "epoch": 0.08387193747181723,
      "flos": 20629782835200.0,
      "grad_norm": 1.782466846937859,
      "language_loss": 0.82979721,
      "learning_rate": 3.969634289062719e-06,
      "loss": 0.85183442,
      "num_input_tokens_seen": 29683825,
      "step": 1395,
      "time_per_iteration": 2.883977174758911
    },
    {
      "auxiliary_loss_clip": 0.01187279,
      "auxiliary_loss_mlp": 0.00782865,
      "balance_loss_clip": 1.06065941,
      "balance_loss_mlp": 1.00028706,
      "epoch": 0.0839320607244852,
      "flos": 13443196199040.0,
      "grad_norm": 3.330409107955743,
      "language_loss": 0.82481396,
      "learning_rate": 3.969566643154293e-06,
      "loss": 0.84451544,
      "num_input_tokens_seen": 29698775,
      "step": 1396,
      "time_per_iteration": 2.6729378700256348
    },
    {
      "auxiliary_loss_clip": 0.0118605,
      "auxiliary_loss_mlp": 0.01060468,
      "balance_loss_clip": 1.06378388,
      "balance_loss_mlp": 1.03475475,
      "epoch": 0.08399218397715316,
      "flos": 23477247642240.0,
      "grad_norm": 1.780410555630689,
      "language_loss": 0.76843297,
      "learning_rate": 3.969498922559703e-06,
      "loss": 0.79089814,
      "num_input_tokens_seen": 29719430,
      "step": 1397,
      "time_per_iteration": 2.64888334274292
    },
    {
      "auxiliary_loss_clip": 0.01153742,
      "auxiliary_loss_mlp": 0.01050759,
      "balance_loss_clip": 1.05790138,
      "balance_loss_mlp": 1.02621412,
      "epoch": 0.08405230722982113,
      "flos": 25920717206400.0,
      "grad_norm": 2.1323769932413184,
      "language_loss": 0.77941638,
      "learning_rate": 3.969431127281516e-06,
      "loss": 0.8014614,
      "num_input_tokens_seen": 29739685,
      "step": 1398,
      "time_per_iteration": 2.8302125930786133
    },
    {
      "auxiliary_loss_clip": 0.01191086,
      "auxiliary_loss_mlp": 0.01052374,
      "balance_loss_clip": 1.05962944,
      "balance_loss_mlp": 1.02943766,
      "epoch": 0.0841124304824891,
      "flos": 17967437746560.0,
      "grad_norm": 2.150764713624159,
      "language_loss": 0.94635069,
      "learning_rate": 3.969363257322304e-06,
      "loss": 0.96878529,
      "num_input_tokens_seen": 29756165,
      "step": 1399,
      "time_per_iteration": 2.650517702102661
    },
    {
      "auxiliary_loss_clip": 0.01172403,
      "auxiliary_loss_mlp": 0.0106738,
      "balance_loss_clip": 1.0562712,
      "balance_loss_mlp": 1.04168999,
      "epoch": 0.08417255373515707,
      "flos": 25629661301760.0,
      "grad_norm": 3.6141849657848137,
      "language_loss": 0.81904209,
      "learning_rate": 3.96929531268464e-06,
      "loss": 0.8414399,
      "num_input_tokens_seen": 29776425,
      "step": 1400,
      "time_per_iteration": 2.777369260787964
    },
    {
      "auxiliary_loss_clip": 0.01170173,
      "auxiliary_loss_mlp": 0.01064292,
      "balance_loss_clip": 1.05968165,
      "balance_loss_mlp": 1.03957999,
      "epoch": 0.08423267698782504,
      "flos": 26249730808320.0,
      "grad_norm": 8.998651919840762,
      "language_loss": 0.8642807,
      "learning_rate": 3.969227293371099e-06,
      "loss": 0.88662529,
      "num_input_tokens_seen": 29796440,
      "step": 1401,
      "time_per_iteration": 2.91375732421875
    },
    {
      "auxiliary_loss_clip": 0.01196,
      "auxiliary_loss_mlp": 0.01066109,
      "balance_loss_clip": 1.05935979,
      "balance_loss_mlp": 1.04053831,
      "epoch": 0.08429280024049302,
      "flos": 20119707751680.0,
      "grad_norm": 2.9792515680869114,
      "language_loss": 0.87500131,
      "learning_rate": 3.969159199384263e-06,
      "loss": 0.89762247,
      "num_input_tokens_seen": 29814755,
      "step": 1402,
      "time_per_iteration": 2.7827296257019043
    },
    {
      "auxiliary_loss_clip": 0.01144907,
      "auxiliary_loss_mlp": 0.00781428,
      "balance_loss_clip": 1.05105817,
      "balance_loss_mlp": 1.00033188,
      "epoch": 0.08435292349316098,
      "flos": 42924526836480.0,
      "grad_norm": 2.1517994230241566,
      "language_loss": 0.8905524,
      "learning_rate": 3.9690910307267125e-06,
      "loss": 0.90981579,
      "num_input_tokens_seen": 29834785,
      "step": 1403,
      "time_per_iteration": 2.931666374206543
    },
    {
      "auxiliary_loss_clip": 0.01165276,
      "auxiliary_loss_mlp": 0.01061696,
      "balance_loss_clip": 1.05570936,
      "balance_loss_mlp": 1.03715038,
      "epoch": 0.08441304674582895,
      "flos": 22857285876480.0,
      "grad_norm": 1.790271378285476,
      "language_loss": 0.80321431,
      "learning_rate": 3.969022787401033e-06,
      "loss": 0.82548404,
      "num_input_tokens_seen": 29854695,
      "step": 1404,
      "time_per_iteration": 2.7397725582122803
    },
    {
      "auxiliary_loss_clip": 0.01181709,
      "auxiliary_loss_mlp": 0.01071408,
      "balance_loss_clip": 1.06211567,
      "balance_loss_mlp": 1.04649353,
      "epoch": 0.08447316999849692,
      "flos": 18697501676160.0,
      "grad_norm": 2.0849305916509193,
      "language_loss": 0.83557045,
      "learning_rate": 3.968954469409811e-06,
      "loss": 0.85810155,
      "num_input_tokens_seen": 29872180,
      "step": 1405,
      "time_per_iteration": 2.8052847385406494
    },
    {
      "auxiliary_loss_clip": 0.0118246,
      "auxiliary_loss_mlp": 0.01058347,
      "balance_loss_clip": 1.05636072,
      "balance_loss_mlp": 1.03588748,
      "epoch": 0.08453329325116489,
      "flos": 25483971738240.0,
      "grad_norm": 1.5225846020503528,
      "language_loss": 0.7991904,
      "learning_rate": 3.968886076755639e-06,
      "loss": 0.82159847,
      "num_input_tokens_seen": 29893205,
      "step": 1406,
      "time_per_iteration": 4.301243305206299
    },
    {
      "auxiliary_loss_clip": 0.0117117,
      "auxiliary_loss_mlp": 0.01068275,
      "balance_loss_clip": 1.05790758,
      "balance_loss_mlp": 1.04406369,
      "epoch": 0.08459341650383286,
      "flos": 20920048640640.0,
      "grad_norm": 1.717770739318623,
      "language_loss": 0.79441547,
      "learning_rate": 3.96881760944111e-06,
      "loss": 0.81680995,
      "num_input_tokens_seen": 29911970,
      "step": 1407,
      "time_per_iteration": 2.6535613536834717
    },
    {
      "auxiliary_loss_clip": 0.01186501,
      "auxiliary_loss_mlp": 0.01057881,
      "balance_loss_clip": 1.05982685,
      "balance_loss_mlp": 1.03409886,
      "epoch": 0.08465353975650082,
      "flos": 13043079624960.0,
      "grad_norm": 2.191354041218588,
      "language_loss": 0.91799384,
      "learning_rate": 3.968749067468819e-06,
      "loss": 0.94043779,
      "num_input_tokens_seen": 29929925,
      "step": 1408,
      "time_per_iteration": 5.774486064910889
    },
    {
      "auxiliary_loss_clip": 0.01058217,
      "auxiliary_loss_mlp": 0.01015213,
      "balance_loss_clip": 1.0231359,
      "balance_loss_mlp": 1.01139832,
      "epoch": 0.0847136630091688,
      "flos": 60877422552960.0,
      "grad_norm": 0.9559717259642487,
      "language_loss": 0.61891782,
      "learning_rate": 3.968680450841368e-06,
      "loss": 0.63965201,
      "num_input_tokens_seen": 29985950,
      "step": 1409,
      "time_per_iteration": 4.9455225467681885
    },
    {
      "auxiliary_loss_clip": 0.01188186,
      "auxiliary_loss_mlp": 0.01061718,
      "balance_loss_clip": 1.05840743,
      "balance_loss_mlp": 1.03878236,
      "epoch": 0.08477378626183676,
      "flos": 22046530043520.0,
      "grad_norm": 1.6980375913788566,
      "language_loss": 0.86357373,
      "learning_rate": 3.968611759561355e-06,
      "loss": 0.88607281,
      "num_input_tokens_seen": 30004330,
      "step": 1410,
      "time_per_iteration": 2.640355110168457
    },
    {
      "auxiliary_loss_clip": 0.01181512,
      "auxiliary_loss_mlp": 0.01053874,
      "balance_loss_clip": 1.0583061,
      "balance_loss_mlp": 1.02870846,
      "epoch": 0.08483390951450473,
      "flos": 16690059308160.0,
      "grad_norm": 2.248971712939306,
      "language_loss": 0.74384397,
      "learning_rate": 3.968542993631388e-06,
      "loss": 0.7661978,
      "num_input_tokens_seen": 30022555,
      "step": 1411,
      "time_per_iteration": 2.6200830936431885
    },
    {
      "auxiliary_loss_clip": 0.01077929,
      "auxiliary_loss_mlp": 0.01003535,
      "balance_loss_clip": 1.02317524,
      "balance_loss_mlp": 0.99991113,
      "epoch": 0.08489403276717271,
      "flos": 51584640082560.0,
      "grad_norm": 0.9014663966204861,
      "language_loss": 0.56748837,
      "learning_rate": 3.968474153054073e-06,
      "loss": 0.58830309,
      "num_input_tokens_seen": 30077220,
      "step": 1412,
      "time_per_iteration": 3.0746512413024902
    },
    {
      "auxiliary_loss_clip": 0.01156137,
      "auxiliary_loss_mlp": 0.01067795,
      "balance_loss_clip": 1.05325568,
      "balance_loss_mlp": 1.04265356,
      "epoch": 0.08495415601984067,
      "flos": 17092330698240.0,
      "grad_norm": 2.2757293876932945,
      "language_loss": 0.88754624,
      "learning_rate": 3.96840523783202e-06,
      "loss": 0.90978551,
      "num_input_tokens_seen": 30094600,
      "step": 1413,
      "time_per_iteration": 2.7309420108795166
    },
    {
      "auxiliary_loss_clip": 0.01164895,
      "auxiliary_loss_mlp": 0.01057479,
      "balance_loss_clip": 1.05780244,
      "balance_loss_mlp": 1.03295755,
      "epoch": 0.08501427927250864,
      "flos": 23148413608320.0,
      "grad_norm": 1.9781781646219805,
      "language_loss": 0.87963474,
      "learning_rate": 3.968336247967844e-06,
      "loss": 0.90185857,
      "num_input_tokens_seen": 30114475,
      "step": 1414,
      "time_per_iteration": 2.692030668258667
    },
    {
      "auxiliary_loss_clip": 0.01168145,
      "auxiliary_loss_mlp": 0.01063751,
      "balance_loss_clip": 1.05704033,
      "balance_loss_mlp": 1.04170966,
      "epoch": 0.08507440252517662,
      "flos": 19063467394560.0,
      "grad_norm": 1.9706021333256292,
      "language_loss": 0.77636635,
      "learning_rate": 3.96826718346416e-06,
      "loss": 0.79868531,
      "num_input_tokens_seen": 30133350,
      "step": 1415,
      "time_per_iteration": 2.8435540199279785
    },
    {
      "auxiliary_loss_clip": 0.01182108,
      "auxiliary_loss_mlp": 0.01059478,
      "balance_loss_clip": 1.0588963,
      "balance_loss_mlp": 1.03701878,
      "epoch": 0.08513452577784458,
      "flos": 60182296600320.0,
      "grad_norm": 1.7170282174092708,
      "language_loss": 0.70545506,
      "learning_rate": 3.968198044323587e-06,
      "loss": 0.72787094,
      "num_input_tokens_seen": 30159005,
      "step": 1416,
      "time_per_iteration": 3.021360158920288
    },
    {
      "auxiliary_loss_clip": 0.01174166,
      "auxiliary_loss_mlp": 0.01066487,
      "balance_loss_clip": 1.05930233,
      "balance_loss_mlp": 1.04131043,
      "epoch": 0.08519464903051255,
      "flos": 27308485117440.0,
      "grad_norm": 2.8159853289102053,
      "language_loss": 0.74938154,
      "learning_rate": 3.968128830548748e-06,
      "loss": 0.771788,
      "num_input_tokens_seen": 30179450,
      "step": 1417,
      "time_per_iteration": 2.738301992416382
    },
    {
      "auxiliary_loss_clip": 0.01171292,
      "auxiliary_loss_mlp": 0.01057092,
      "balance_loss_clip": 1.05715823,
      "balance_loss_mlp": 1.03313112,
      "epoch": 0.08525477228318051,
      "flos": 20266438809600.0,
      "grad_norm": 2.4132423968154635,
      "language_loss": 0.8258723,
      "learning_rate": 3.968059542142265e-06,
      "loss": 0.84815615,
      "num_input_tokens_seen": 30197235,
      "step": 1418,
      "time_per_iteration": 2.671574831008911
    },
    {
      "auxiliary_loss_clip": 0.0104499,
      "auxiliary_loss_mlp": 0.01004818,
      "balance_loss_clip": 1.02242994,
      "balance_loss_mlp": 1.0004549,
      "epoch": 0.08531489553584849,
      "flos": 67615017183360.0,
      "grad_norm": 0.8667411864001444,
      "language_loss": 0.56638753,
      "learning_rate": 3.9679901791067685e-06,
      "loss": 0.58688557,
      "num_input_tokens_seen": 30257410,
      "step": 1419,
      "time_per_iteration": 3.199730396270752
    },
    {
      "auxiliary_loss_clip": 0.01192231,
      "auxiliary_loss_mlp": 0.01067737,
      "balance_loss_clip": 1.05757999,
      "balance_loss_mlp": 1.04369283,
      "epoch": 0.08537501878851646,
      "flos": 27526965592320.0,
      "grad_norm": 2.2357492693560466,
      "language_loss": 0.70111859,
      "learning_rate": 3.967920741444886e-06,
      "loss": 0.72371829,
      "num_input_tokens_seen": 30277865,
      "step": 1420,
      "time_per_iteration": 2.7176027297973633
    },
    {
      "auxiliary_loss_clip": 0.01155207,
      "auxiliary_loss_mlp": 0.01050755,
      "balance_loss_clip": 1.05377483,
      "balance_loss_mlp": 1.02692556,
      "epoch": 0.08543514204118442,
      "flos": 22784243569920.0,
      "grad_norm": 1.5975069204011494,
      "language_loss": 0.88011539,
      "learning_rate": 3.967851229159252e-06,
      "loss": 0.90217495,
      "num_input_tokens_seen": 30298545,
      "step": 1421,
      "time_per_iteration": 2.7552106380462646
    },
    {
      "auxiliary_loss_clip": 0.01077473,
      "auxiliary_loss_mlp": 0.01013517,
      "balance_loss_clip": 1.02364218,
      "balance_loss_mlp": 1.01020324,
      "epoch": 0.0854952652938524,
      "flos": 60990721027200.0,
      "grad_norm": 0.9142209544576306,
      "language_loss": 0.63506877,
      "learning_rate": 3.967781642252502e-06,
      "loss": 0.65597868,
      "num_input_tokens_seen": 30361725,
      "step": 1422,
      "time_per_iteration": 3.134183168411255
    },
    {
      "auxiliary_loss_clip": 0.01152948,
      "auxiliary_loss_mlp": 0.01063847,
      "balance_loss_clip": 1.05932307,
      "balance_loss_mlp": 1.0406723,
      "epoch": 0.08555538854652037,
      "flos": 28038046256640.0,
      "grad_norm": 1.8757015124159093,
      "language_loss": 0.82691669,
      "learning_rate": 3.967711980727276e-06,
      "loss": 0.84908462,
      "num_input_tokens_seen": 30382180,
      "step": 1423,
      "time_per_iteration": 2.789393424987793
    },
    {
      "auxiliary_loss_clip": 0.01153439,
      "auxiliary_loss_mlp": 0.01064169,
      "balance_loss_clip": 1.0526228,
      "balance_loss_mlp": 1.04089928,
      "epoch": 0.08561551179918833,
      "flos": 23509279595520.0,
      "grad_norm": 1.6593534429066656,
      "language_loss": 0.75424892,
      "learning_rate": 3.967642244586213e-06,
      "loss": 0.776425,
      "num_input_tokens_seen": 30402980,
      "step": 1424,
      "time_per_iteration": 2.7805826663970947
    },
    {
      "auxiliary_loss_clip": 0.01139579,
      "auxiliary_loss_mlp": 0.01060342,
      "balance_loss_clip": 1.05769765,
      "balance_loss_mlp": 1.03751373,
      "epoch": 0.08567563505185631,
      "flos": 17926930183680.0,
      "grad_norm": 1.7999307606718091,
      "language_loss": 0.75948423,
      "learning_rate": 3.96757243383196e-06,
      "loss": 0.78148341,
      "num_input_tokens_seen": 30420800,
      "step": 1425,
      "time_per_iteration": 2.677889823913574
    },
    {
      "auxiliary_loss_clip": 0.0118966,
      "auxiliary_loss_mlp": 0.01055231,
      "balance_loss_clip": 1.05982256,
      "balance_loss_mlp": 1.03230715,
      "epoch": 0.08573575830452428,
      "flos": 19719519350400.0,
      "grad_norm": 2.1792756220437743,
      "language_loss": 0.93362999,
      "learning_rate": 3.9675025484671624e-06,
      "loss": 0.95607889,
      "num_input_tokens_seen": 30439620,
      "step": 1426,
      "time_per_iteration": 2.6270906925201416
    },
    {
      "auxiliary_loss_clip": 0.01145994,
      "auxiliary_loss_mlp": 0.01066219,
      "balance_loss_clip": 1.05707717,
      "balance_loss_mlp": 1.0406251,
      "epoch": 0.08579588155719224,
      "flos": 17931563038080.0,
      "grad_norm": 2.3679064075186553,
      "language_loss": 0.75424731,
      "learning_rate": 3.967432588494471e-06,
      "loss": 0.77636945,
      "num_input_tokens_seen": 30457300,
      "step": 1427,
      "time_per_iteration": 2.84614634513855
    },
    {
      "auxiliary_loss_clip": 0.01190697,
      "auxiliary_loss_mlp": 0.01052992,
      "balance_loss_clip": 1.06006169,
      "balance_loss_mlp": 1.0305804,
      "epoch": 0.08585600480986022,
      "flos": 16033324993920.0,
      "grad_norm": 3.503048788198607,
      "language_loss": 0.82108849,
      "learning_rate": 3.96736255391654e-06,
      "loss": 0.84352541,
      "num_input_tokens_seen": 30471580,
      "step": 1428,
      "time_per_iteration": 2.5882396697998047
    },
    {
      "auxiliary_loss_clip": 0.01173688,
      "auxiliary_loss_mlp": 0.0106298,
      "balance_loss_clip": 1.05633736,
      "balance_loss_mlp": 1.03832793,
      "epoch": 0.08591612806252819,
      "flos": 28657433404800.0,
      "grad_norm": 2.088481658755078,
      "language_loss": 0.79929984,
      "learning_rate": 3.967292444736023e-06,
      "loss": 0.82166648,
      "num_input_tokens_seen": 30492720,
      "step": 1429,
      "time_per_iteration": 2.720500946044922
    },
    {
      "auxiliary_loss_clip": 0.01169119,
      "auxiliary_loss_mlp": 0.010606,
      "balance_loss_clip": 1.05971265,
      "balance_loss_mlp": 1.0379504,
      "epoch": 0.08597625131519615,
      "flos": 20959119659520.0,
      "grad_norm": 1.9029222975672677,
      "language_loss": 0.87716508,
      "learning_rate": 3.967222260955578e-06,
      "loss": 0.89946228,
      "num_input_tokens_seen": 30509535,
      "step": 1430,
      "time_per_iteration": 2.6914596557617188
    },
    {
      "auxiliary_loss_clip": 0.01144304,
      "auxiliary_loss_mlp": 0.01074633,
      "balance_loss_clip": 1.05802035,
      "balance_loss_mlp": 1.05125606,
      "epoch": 0.08603637456786412,
      "flos": 23256360956160.0,
      "grad_norm": 1.6366623508781384,
      "language_loss": 0.81859726,
      "learning_rate": 3.96715200257787e-06,
      "loss": 0.84078664,
      "num_input_tokens_seen": 30529490,
      "step": 1431,
      "time_per_iteration": 2.834402322769165
    },
    {
      "auxiliary_loss_clip": 0.01148362,
      "auxiliary_loss_mlp": 0.01054323,
      "balance_loss_clip": 1.05620182,
      "balance_loss_mlp": 1.03132737,
      "epoch": 0.0860964978205321,
      "flos": 28694170039680.0,
      "grad_norm": 1.5497375505717568,
      "language_loss": 0.78109461,
      "learning_rate": 3.967081669605559e-06,
      "loss": 0.80312145,
      "num_input_tokens_seen": 30550205,
      "step": 1432,
      "time_per_iteration": 2.767860174179077
    },
    {
      "auxiliary_loss_clip": 0.01167351,
      "auxiliary_loss_mlp": 0.0106333,
      "balance_loss_clip": 1.0540905,
      "balance_loss_mlp": 1.03914225,
      "epoch": 0.08615662107320006,
      "flos": 19318397195520.0,
      "grad_norm": 1.9631692713893694,
      "language_loss": 0.73365706,
      "learning_rate": 3.967011262041315e-06,
      "loss": 0.75596392,
      "num_input_tokens_seen": 30568830,
      "step": 1433,
      "time_per_iteration": 2.6930699348449707
    },
    {
      "auxiliary_loss_clip": 0.01150098,
      "auxiliary_loss_mlp": 0.00781967,
      "balance_loss_clip": 1.05335927,
      "balance_loss_mlp": 1.00044179,
      "epoch": 0.08621674432586802,
      "flos": 15851688894720.0,
      "grad_norm": 2.468588778716135,
      "language_loss": 0.85340321,
      "learning_rate": 3.9669407798878065e-06,
      "loss": 0.87272388,
      "num_input_tokens_seen": 30585730,
      "step": 1434,
      "time_per_iteration": 2.735690116882324
    },
    {
      "auxiliary_loss_clip": 0.01170363,
      "auxiliary_loss_mlp": 0.01057659,
      "balance_loss_clip": 1.05604434,
      "balance_loss_mlp": 1.0344249,
      "epoch": 0.086276867578536,
      "flos": 14100648785280.0,
      "grad_norm": 2.160640509122794,
      "language_loss": 0.7870298,
      "learning_rate": 3.966870223147707e-06,
      "loss": 0.80931008,
      "num_input_tokens_seen": 30603180,
      "step": 1435,
      "time_per_iteration": 2.776567220687866
    },
    {
      "auxiliary_loss_clip": 0.01047768,
      "auxiliary_loss_mlp": 0.01015597,
      "balance_loss_clip": 1.023893,
      "balance_loss_mlp": 1.01206815,
      "epoch": 0.08633699083120397,
      "flos": 70184857772160.0,
      "grad_norm": 0.8900716332014227,
      "language_loss": 0.57975936,
      "learning_rate": 3.96679959182369e-06,
      "loss": 0.60039294,
      "num_input_tokens_seen": 30668895,
      "step": 1436,
      "time_per_iteration": 3.344207763671875
    },
    {
      "auxiliary_loss_clip": 0.0117372,
      "auxiliary_loss_mlp": 0.01056829,
      "balance_loss_clip": 1.05617976,
      "balance_loss_mlp": 1.03153312,
      "epoch": 0.08639711408387193,
      "flos": 30298874140800.0,
      "grad_norm": 2.240343996649645,
      "language_loss": 0.69169062,
      "learning_rate": 3.966728885918437e-06,
      "loss": 0.71399617,
      "num_input_tokens_seen": 30688955,
      "step": 1437,
      "time_per_iteration": 2.7171547412872314
    },
    {
      "auxiliary_loss_clip": 0.01121044,
      "auxiliary_loss_mlp": 0.01055264,
      "balance_loss_clip": 1.05334914,
      "balance_loss_mlp": 1.03223276,
      "epoch": 0.08645723733653991,
      "flos": 20297680663680.0,
      "grad_norm": 2.1340571114707245,
      "language_loss": 0.72624576,
      "learning_rate": 3.966658105434627e-06,
      "loss": 0.74800885,
      "num_input_tokens_seen": 30706095,
      "step": 1438,
      "time_per_iteration": 2.7815651893615723
    },
    {
      "auxiliary_loss_clip": 0.01179626,
      "auxiliary_loss_mlp": 0.01052578,
      "balance_loss_clip": 1.06052637,
      "balance_loss_mlp": 1.02872419,
      "epoch": 0.08651736058920788,
      "flos": 32890583134080.0,
      "grad_norm": 1.5339762166114281,
      "language_loss": 0.64377135,
      "learning_rate": 3.966587250374945e-06,
      "loss": 0.66609335,
      "num_input_tokens_seen": 30729025,
      "step": 1439,
      "time_per_iteration": 2.8935797214508057
    },
    {
      "auxiliary_loss_clip": 0.01153286,
      "auxiliary_loss_mlp": 0.01056452,
      "balance_loss_clip": 1.05530453,
      "balance_loss_mlp": 1.03213322,
      "epoch": 0.08657748384187584,
      "flos": 22637368857600.0,
      "grad_norm": 5.193932354158579,
      "language_loss": 0.87521696,
      "learning_rate": 3.966516320742077e-06,
      "loss": 0.89731431,
      "num_input_tokens_seen": 30746155,
      "step": 1440,
      "time_per_iteration": 2.731531858444214
    },
    {
      "auxiliary_loss_clip": 0.01155923,
      "auxiliary_loss_mlp": 0.00782787,
      "balance_loss_clip": 1.05752945,
      "balance_loss_mlp": 1.00043201,
      "epoch": 0.08663760709454381,
      "flos": 23658380951040.0,
      "grad_norm": 2.023462963415533,
      "language_loss": 0.83434939,
      "learning_rate": 3.9664453165387124e-06,
      "loss": 0.85373652,
      "num_input_tokens_seen": 30761410,
      "step": 1441,
      "time_per_iteration": 2.7126500606536865
    },
    {
      "auxiliary_loss_clip": 0.01074667,
      "auxiliary_loss_mlp": 0.01004602,
      "balance_loss_clip": 1.0222367,
      "balance_loss_mlp": 1.00100195,
      "epoch": 0.08669773034721179,
      "flos": 62686564911360.0,
      "grad_norm": 0.8541685426878655,
      "language_loss": 0.60479522,
      "learning_rate": 3.966374237767545e-06,
      "loss": 0.62558794,
      "num_input_tokens_seen": 30823010,
      "step": 1442,
      "time_per_iteration": 3.25555157661438
    },
    {
      "auxiliary_loss_clip": 0.0116729,
      "auxiliary_loss_mlp": 0.01054262,
      "balance_loss_clip": 1.05768681,
      "balance_loss_mlp": 1.03075421,
      "epoch": 0.08675785359987975,
      "flos": 20667489137280.0,
      "grad_norm": 2.8449103562639073,
      "language_loss": 0.79304373,
      "learning_rate": 3.96630308443127e-06,
      "loss": 0.81525922,
      "num_input_tokens_seen": 30841980,
      "step": 1443,
      "time_per_iteration": 2.7314631938934326
    },
    {
      "auxiliary_loss_clip": 0.01180858,
      "auxiliary_loss_mlp": 0.01051075,
      "balance_loss_clip": 1.05780149,
      "balance_loss_mlp": 1.02755547,
      "epoch": 0.08681797685254772,
      "flos": 26941118768640.0,
      "grad_norm": 1.6739262813835734,
      "language_loss": 0.82399666,
      "learning_rate": 3.966231856532584e-06,
      "loss": 0.84631598,
      "num_input_tokens_seen": 30863280,
      "step": 1444,
      "time_per_iteration": 2.7341418266296387
    },
    {
      "auxiliary_loss_clip": 0.01196759,
      "auxiliary_loss_mlp": 0.01051473,
      "balance_loss_clip": 1.06044626,
      "balance_loss_mlp": 1.02810788,
      "epoch": 0.0868781001052157,
      "flos": 17712831168000.0,
      "grad_norm": 2.3015461969915747,
      "language_loss": 0.87354827,
      "learning_rate": 3.966160554074189e-06,
      "loss": 0.8960306,
      "num_input_tokens_seen": 30881710,
      "step": 1445,
      "time_per_iteration": 4.25179386138916
    },
    {
      "auxiliary_loss_clip": 0.01180784,
      "auxiliary_loss_mlp": 0.01055896,
      "balance_loss_clip": 1.06094933,
      "balance_loss_mlp": 1.03446186,
      "epoch": 0.08693822335788366,
      "flos": 19896522595200.0,
      "grad_norm": 1.8066650797875201,
      "language_loss": 0.81863767,
      "learning_rate": 3.96608917705879e-06,
      "loss": 0.84100449,
      "num_input_tokens_seen": 30900225,
      "step": 1446,
      "time_per_iteration": 4.197181940078735
    },
    {
      "auxiliary_loss_clip": 0.01056056,
      "auxiliary_loss_mlp": 0.01004371,
      "balance_loss_clip": 1.01782191,
      "balance_loss_mlp": 1.00031781,
      "epoch": 0.08699834661055163,
      "flos": 67023747406080.0,
      "grad_norm": 0.7255245569613363,
      "language_loss": 0.54762936,
      "learning_rate": 3.966017725489091e-06,
      "loss": 0.56823361,
      "num_input_tokens_seen": 30959580,
      "step": 1447,
      "time_per_iteration": 3.2158126831054688
    },
    {
      "auxiliary_loss_clip": 0.0114861,
      "auxiliary_loss_mlp": 0.01056824,
      "balance_loss_clip": 1.05373001,
      "balance_loss_mlp": 1.03518772,
      "epoch": 0.0870584698632196,
      "flos": 13480507451520.0,
      "grad_norm": 2.1586118179593696,
      "language_loss": 0.84592307,
      "learning_rate": 3.965946199367804e-06,
      "loss": 0.86797738,
      "num_input_tokens_seen": 30976775,
      "step": 1448,
      "time_per_iteration": 4.262767314910889
    },
    {
      "auxiliary_loss_clip": 0.01194173,
      "auxiliary_loss_mlp": 0.01050219,
      "balance_loss_clip": 1.05891991,
      "balance_loss_mlp": 1.02768826,
      "epoch": 0.08711859311588757,
      "flos": 16107013745280.0,
      "grad_norm": 3.4326906921347096,
      "language_loss": 0.80644608,
      "learning_rate": 3.965874598697638e-06,
      "loss": 0.82888997,
      "num_input_tokens_seen": 30990495,
      "step": 1449,
      "time_per_iteration": 4.553676128387451
    },
    {
      "auxiliary_loss_clip": 0.01138548,
      "auxiliary_loss_mlp": 0.01052142,
      "balance_loss_clip": 1.05437374,
      "balance_loss_mlp": 1.02946854,
      "epoch": 0.08717871636855554,
      "flos": 38472357928320.0,
      "grad_norm": 1.5251600336566102,
      "language_loss": 0.70971417,
      "learning_rate": 3.965802923481313e-06,
      "loss": 0.73162109,
      "num_input_tokens_seen": 31014080,
      "step": 1450,
      "time_per_iteration": 2.9082705974578857
    },
    {
      "auxiliary_loss_clip": 0.01124466,
      "auxiliary_loss_mlp": 0.01054883,
      "balance_loss_clip": 1.05164719,
      "balance_loss_mlp": 1.03207827,
      "epoch": 0.0872388396212235,
      "flos": 17600574188160.0,
      "grad_norm": 1.9392114767205617,
      "language_loss": 0.83684897,
      "learning_rate": 3.965731173721542e-06,
      "loss": 0.85864246,
      "num_input_tokens_seen": 31031210,
      "step": 1451,
      "time_per_iteration": 2.809880495071411
    },
    {
      "auxiliary_loss_clip": 0.01134251,
      "auxiliary_loss_mlp": 0.00780873,
      "balance_loss_clip": 1.05147851,
      "balance_loss_mlp": 1.00039482,
      "epoch": 0.08729896287389148,
      "flos": 25259385951360.0,
      "grad_norm": 2.5160845512367773,
      "language_loss": 0.74654591,
      "learning_rate": 3.965659349421049e-06,
      "loss": 0.76569718,
      "num_input_tokens_seen": 31049710,
      "step": 1452,
      "time_per_iteration": 2.88580060005188
    },
    {
      "auxiliary_loss_clip": 0.01157134,
      "auxiliary_loss_mlp": 0.01063328,
      "balance_loss_clip": 1.05607891,
      "balance_loss_mlp": 1.0388428,
      "epoch": 0.08735908612655945,
      "flos": 15632454234240.0,
      "grad_norm": 4.56941406999875,
      "language_loss": 0.80543101,
      "learning_rate": 3.965587450582556e-06,
      "loss": 0.82763565,
      "num_input_tokens_seen": 31066160,
      "step": 1453,
      "time_per_iteration": 2.733632802963257
    },
    {
      "auxiliary_loss_clip": 0.01169707,
      "auxiliary_loss_mlp": 0.01059533,
      "balance_loss_clip": 1.05905569,
      "balance_loss_mlp": 1.03625154,
      "epoch": 0.08741920937922741,
      "flos": 20339660684160.0,
      "grad_norm": 2.0102093196988102,
      "language_loss": 0.71041977,
      "learning_rate": 3.96551547720879e-06,
      "loss": 0.73271215,
      "num_input_tokens_seen": 31085270,
      "step": 1454,
      "time_per_iteration": 2.7568745613098145
    },
    {
      "auxiliary_loss_clip": 0.0106426,
      "auxiliary_loss_mlp": 0.01008112,
      "balance_loss_clip": 1.0215131,
      "balance_loss_mlp": 1.00463128,
      "epoch": 0.08747933263189539,
      "flos": 62819795433600.0,
      "grad_norm": 0.7713706503543015,
      "language_loss": 0.5859946,
      "learning_rate": 3.96544342930248e-06,
      "loss": 0.6067183,
      "num_input_tokens_seen": 31148445,
      "step": 1455,
      "time_per_iteration": 3.2372186183929443
    },
    {
      "auxiliary_loss_clip": 0.01189404,
      "auxiliary_loss_mlp": 0.01060742,
      "balance_loss_clip": 1.05742884,
      "balance_loss_mlp": 1.03688788,
      "epoch": 0.08753945588456336,
      "flos": 33035877648000.0,
      "grad_norm": 1.6485208275358016,
      "language_loss": 0.77564865,
      "learning_rate": 3.965371306866359e-06,
      "loss": 0.79815018,
      "num_input_tokens_seen": 31168770,
      "step": 1456,
      "time_per_iteration": 2.790663003921509
    },
    {
      "auxiliary_loss_clip": 0.01127959,
      "auxiliary_loss_mlp": 0.01054526,
      "balance_loss_clip": 1.04962158,
      "balance_loss_mlp": 1.03071976,
      "epoch": 0.08759957913723132,
      "flos": 35547182046720.0,
      "grad_norm": 1.83889407784057,
      "language_loss": 0.72420907,
      "learning_rate": 3.96529910990316e-06,
      "loss": 0.74603397,
      "num_input_tokens_seen": 31189270,
      "step": 1457,
      "time_per_iteration": 2.9099740982055664
    },
    {
      "auxiliary_loss_clip": 0.01176549,
      "auxiliary_loss_mlp": 0.0104866,
      "balance_loss_clip": 1.05627227,
      "balance_loss_mlp": 1.02633214,
      "epoch": 0.0876597023898993,
      "flos": 23911120022400.0,
      "grad_norm": 1.5250401870177361,
      "language_loss": 0.86412215,
      "learning_rate": 3.965226838415622e-06,
      "loss": 0.88637424,
      "num_input_tokens_seen": 31210385,
      "step": 1458,
      "time_per_iteration": 2.7517166137695312
    },
    {
      "auxiliary_loss_clip": 0.01169535,
      "auxiliary_loss_mlp": 0.01061413,
      "balance_loss_clip": 1.05884266,
      "balance_loss_mlp": 1.03825045,
      "epoch": 0.08771982564256726,
      "flos": 18114025150080.0,
      "grad_norm": 1.7412813512419094,
      "language_loss": 0.80268395,
      "learning_rate": 3.965154492406486e-06,
      "loss": 0.82499349,
      "num_input_tokens_seen": 31229745,
      "step": 1459,
      "time_per_iteration": 2.71455717086792
    },
    {
      "auxiliary_loss_clip": 0.01130491,
      "auxiliary_loss_mlp": 0.01054334,
      "balance_loss_clip": 1.05256546,
      "balance_loss_mlp": 1.03018188,
      "epoch": 0.08777994889523523,
      "flos": 17712005155200.0,
      "grad_norm": 2.1450339680450714,
      "language_loss": 0.84538847,
      "learning_rate": 3.9650820718784945e-06,
      "loss": 0.86723673,
      "num_input_tokens_seen": 31248280,
      "step": 1460,
      "time_per_iteration": 2.8737733364105225
    },
    {
      "auxiliary_loss_clip": 0.01177787,
      "auxiliary_loss_mlp": 0.01057974,
      "balance_loss_clip": 1.0572983,
      "balance_loss_mlp": 1.03640938,
      "epoch": 0.0878400721479032,
      "flos": 12819930382080.0,
      "grad_norm": 4.917361835698274,
      "language_loss": 0.79993135,
      "learning_rate": 3.965009576834394e-06,
      "loss": 0.82228899,
      "num_input_tokens_seen": 31262190,
      "step": 1461,
      "time_per_iteration": 2.8436062335968018
    },
    {
      "auxiliary_loss_clip": 0.01169165,
      "auxiliary_loss_mlp": 0.01058947,
      "balance_loss_clip": 1.05800629,
      "balance_loss_mlp": 1.03704822,
      "epoch": 0.08790019540057117,
      "flos": 26392690938240.0,
      "grad_norm": 1.566202508611165,
      "language_loss": 0.76571167,
      "learning_rate": 3.964937007276932e-06,
      "loss": 0.78799284,
      "num_input_tokens_seen": 31283690,
      "step": 1462,
      "time_per_iteration": 2.7895474433898926
    },
    {
      "auxiliary_loss_clip": 0.0117563,
      "auxiliary_loss_mlp": 0.01060064,
      "balance_loss_clip": 1.05839491,
      "balance_loss_mlp": 1.03580475,
      "epoch": 0.08796031865323914,
      "flos": 19134031662720.0,
      "grad_norm": 2.89717114041641,
      "language_loss": 0.74710488,
      "learning_rate": 3.9648643632088634e-06,
      "loss": 0.76946187,
      "num_input_tokens_seen": 31302505,
      "step": 1463,
      "time_per_iteration": 2.760404348373413
    },
    {
      "auxiliary_loss_clip": 0.01191543,
      "auxiliary_loss_mlp": 0.01061609,
      "balance_loss_clip": 1.06145048,
      "balance_loss_mlp": 1.03680158,
      "epoch": 0.0880204419059071,
      "flos": 26064287867520.0,
      "grad_norm": 2.431514195311041,
      "language_loss": 0.83797103,
      "learning_rate": 3.964791644632941e-06,
      "loss": 0.8605026,
      "num_input_tokens_seen": 31323070,
      "step": 1464,
      "time_per_iteration": 2.7417759895324707
    },
    {
      "auxiliary_loss_clip": 0.011733,
      "auxiliary_loss_mlp": 0.01063475,
      "balance_loss_clip": 1.05683231,
      "balance_loss_mlp": 1.04093289,
      "epoch": 0.08808056515857508,
      "flos": 22377842115840.0,
      "grad_norm": 2.1775753375634963,
      "language_loss": 0.78104752,
      "learning_rate": 3.964718851551923e-06,
      "loss": 0.8034153,
      "num_input_tokens_seen": 31341880,
      "step": 1465,
      "time_per_iteration": 2.6852309703826904
    },
    {
      "auxiliary_loss_clip": 0.01199489,
      "auxiliary_loss_mlp": 0.01059873,
      "balance_loss_clip": 1.0619812,
      "balance_loss_mlp": 1.03791499,
      "epoch": 0.08814068841124305,
      "flos": 23185293897600.0,
      "grad_norm": 2.412657222564686,
      "language_loss": 0.85187089,
      "learning_rate": 3.9646459839685675e-06,
      "loss": 0.87446451,
      "num_input_tokens_seen": 31361995,
      "step": 1466,
      "time_per_iteration": 2.706264019012451
    },
    {
      "auxiliary_loss_clip": 0.01120627,
      "auxiliary_loss_mlp": 0.00782645,
      "balance_loss_clip": 1.04989958,
      "balance_loss_mlp": 1.00037241,
      "epoch": 0.08820081166391101,
      "flos": 25155281358720.0,
      "grad_norm": 1.9900601596102498,
      "language_loss": 0.84168816,
      "learning_rate": 3.964573041885641e-06,
      "loss": 0.86072087,
      "num_input_tokens_seen": 31381515,
      "step": 1467,
      "time_per_iteration": 2.8636934757232666
    },
    {
      "auxiliary_loss_clip": 0.01178935,
      "auxiliary_loss_mlp": 0.01055379,
      "balance_loss_clip": 1.05910301,
      "balance_loss_mlp": 1.03219247,
      "epoch": 0.08826093491657899,
      "flos": 22231685675520.0,
      "grad_norm": 1.660218686828999,
      "language_loss": 0.75506544,
      "learning_rate": 3.964500025305907e-06,
      "loss": 0.77740854,
      "num_input_tokens_seen": 31400345,
      "step": 1468,
      "time_per_iteration": 2.661501884460449
    },
    {
      "auxiliary_loss_clip": 0.01181261,
      "auxiliary_loss_mlp": 0.01054252,
      "balance_loss_clip": 1.0629456,
      "balance_loss_mlp": 1.03266358,
      "epoch": 0.08832105816924696,
      "flos": 22126826897280.0,
      "grad_norm": 4.868504388441724,
      "language_loss": 0.80322379,
      "learning_rate": 3.9644269342321355e-06,
      "loss": 0.82557893,
      "num_input_tokens_seen": 31419620,
      "step": 1469,
      "time_per_iteration": 2.7473137378692627
    },
    {
      "auxiliary_loss_clip": 0.01198542,
      "auxiliary_loss_mlp": 0.01059353,
      "balance_loss_clip": 1.0627017,
      "balance_loss_mlp": 1.03677487,
      "epoch": 0.08838118142191492,
      "flos": 17566495159680.0,
      "grad_norm": 2.0179242193855806,
      "language_loss": 0.77437651,
      "learning_rate": 3.9643537686670974e-06,
      "loss": 0.79695547,
      "num_input_tokens_seen": 31437970,
      "step": 1470,
      "time_per_iteration": 2.7672410011291504
    },
    {
      "auxiliary_loss_clip": 0.01193825,
      "auxiliary_loss_mlp": 0.01067102,
      "balance_loss_clip": 1.06180143,
      "balance_loss_mlp": 1.04281926,
      "epoch": 0.0884413046745829,
      "flos": 20777196251520.0,
      "grad_norm": 1.6812425162011504,
      "language_loss": 0.84297001,
      "learning_rate": 3.964280528613569e-06,
      "loss": 0.86557925,
      "num_input_tokens_seen": 31457040,
      "step": 1471,
      "time_per_iteration": 2.7584216594696045
    },
    {
      "auxiliary_loss_clip": 0.01156315,
      "auxiliary_loss_mlp": 0.01054307,
      "balance_loss_clip": 1.05682266,
      "balance_loss_mlp": 1.03342199,
      "epoch": 0.08850142792725087,
      "flos": 22125462180480.0,
      "grad_norm": 1.6938350729430058,
      "language_loss": 0.83321345,
      "learning_rate": 3.964207214074324e-06,
      "loss": 0.85531968,
      "num_input_tokens_seen": 31477520,
      "step": 1472,
      "time_per_iteration": 2.7895469665527344
    },
    {
      "auxiliary_loss_clip": 0.01176151,
      "auxiliary_loss_mlp": 0.01058616,
      "balance_loss_clip": 1.06106544,
      "balance_loss_mlp": 1.03529835,
      "epoch": 0.08856155117991883,
      "flos": 22418744728320.0,
      "grad_norm": 2.3638705809965,
      "language_loss": 0.82781172,
      "learning_rate": 3.964133825052146e-06,
      "loss": 0.85015941,
      "num_input_tokens_seen": 31495575,
      "step": 1473,
      "time_per_iteration": 2.7361483573913574
    },
    {
      "auxiliary_loss_clip": 0.01129906,
      "auxiliary_loss_mlp": 0.01064148,
      "balance_loss_clip": 1.05552769,
      "balance_loss_mlp": 1.04263091,
      "epoch": 0.0886216744325868,
      "flos": 29937002572800.0,
      "grad_norm": 1.6022277785896435,
      "language_loss": 0.78712153,
      "learning_rate": 3.964060361549816e-06,
      "loss": 0.80906206,
      "num_input_tokens_seen": 31520020,
      "step": 1474,
      "time_per_iteration": 2.894319534301758
    },
    {
      "auxiliary_loss_clip": 0.01146238,
      "auxiliary_loss_mlp": 0.01068131,
      "balance_loss_clip": 1.05575764,
      "balance_loss_mlp": 1.04175043,
      "epoch": 0.08868179768525478,
      "flos": 23982833525760.0,
      "grad_norm": 1.6120869011213488,
      "language_loss": 0.79030406,
      "learning_rate": 3.963986823570121e-06,
      "loss": 0.81244779,
      "num_input_tokens_seen": 31539265,
      "step": 1475,
      "time_per_iteration": 2.8806042671203613
    },
    {
      "auxiliary_loss_clip": 0.01191986,
      "auxiliary_loss_mlp": 0.01047451,
      "balance_loss_clip": 1.05980015,
      "balance_loss_mlp": 1.02478909,
      "epoch": 0.08874192093792274,
      "flos": 43177553216640.0,
      "grad_norm": 1.4679464237421194,
      "language_loss": 0.74202317,
      "learning_rate": 3.963913211115848e-06,
      "loss": 0.76441753,
      "num_input_tokens_seen": 31563425,
      "step": 1476,
      "time_per_iteration": 2.8381049633026123
    },
    {
      "auxiliary_loss_clip": 0.01174628,
      "auxiliary_loss_mlp": 0.01059934,
      "balance_loss_clip": 1.06217527,
      "balance_loss_mlp": 1.03678358,
      "epoch": 0.0888020441905907,
      "flos": 32852445868800.0,
      "grad_norm": 1.712954575149443,
      "language_loss": 0.74220836,
      "learning_rate": 3.9638395241897895e-06,
      "loss": 0.76455402,
      "num_input_tokens_seen": 31584525,
      "step": 1477,
      "time_per_iteration": 2.8452210426330566
    },
    {
      "auxiliary_loss_clip": 0.01191865,
      "auxiliary_loss_mlp": 0.01051229,
      "balance_loss_clip": 1.06062829,
      "balance_loss_mlp": 1.0278163,
      "epoch": 0.08886216744325869,
      "flos": 23149347361920.0,
      "grad_norm": 1.95844459768748,
      "language_loss": 0.87194049,
      "learning_rate": 3.963765762794739e-06,
      "loss": 0.89437139,
      "num_input_tokens_seen": 31603325,
      "step": 1478,
      "time_per_iteration": 2.644918203353882
    },
    {
      "auxiliary_loss_clip": 0.01176299,
      "auxiliary_loss_mlp": 0.01058069,
      "balance_loss_clip": 1.0572443,
      "balance_loss_mlp": 1.03546739,
      "epoch": 0.08892229069592665,
      "flos": 23331593992320.0,
      "grad_norm": 1.6306868156426517,
      "language_loss": 0.77571511,
      "learning_rate": 3.963691926933495e-06,
      "loss": 0.79805881,
      "num_input_tokens_seen": 31624820,
      "step": 1479,
      "time_per_iteration": 2.738168954849243
    },
    {
      "auxiliary_loss_clip": 0.01164179,
      "auxiliary_loss_mlp": 0.010526,
      "balance_loss_clip": 1.05629039,
      "balance_loss_mlp": 1.02801871,
      "epoch": 0.08898241394859462,
      "flos": 26213784272640.0,
      "grad_norm": 2.199164032289915,
      "language_loss": 0.77797234,
      "learning_rate": 3.9636180166088555e-06,
      "loss": 0.80014014,
      "num_input_tokens_seen": 31646080,
      "step": 1480,
      "time_per_iteration": 2.837562322616577
    },
    {
      "auxiliary_loss_clip": 0.01180168,
      "auxiliary_loss_mlp": 0.01060894,
      "balance_loss_clip": 1.05762577,
      "balance_loss_mlp": 1.03656292,
      "epoch": 0.0890425372012626,
      "flos": 23550613171200.0,
      "grad_norm": 2.9471668635954273,
      "language_loss": 0.66437578,
      "learning_rate": 3.963544031823624e-06,
      "loss": 0.68678641,
      "num_input_tokens_seen": 31665770,
      "step": 1481,
      "time_per_iteration": 2.742422580718994
    },
    {
      "auxiliary_loss_clip": 0.01143445,
      "auxiliary_loss_mlp": 0.01055318,
      "balance_loss_clip": 1.05510306,
      "balance_loss_mlp": 1.03273988,
      "epoch": 0.08910266045393056,
      "flos": 23002795872000.0,
      "grad_norm": 2.124586862599894,
      "language_loss": 0.96630967,
      "learning_rate": 3.9634699725806065e-06,
      "loss": 0.9882974,
      "num_input_tokens_seen": 31683805,
      "step": 1482,
      "time_per_iteration": 2.8150243759155273
    },
    {
      "auxiliary_loss_clip": 0.0115336,
      "auxiliary_loss_mlp": 0.01057266,
      "balance_loss_clip": 1.05521989,
      "balance_loss_mlp": 1.03353167,
      "epoch": 0.08916278370659853,
      "flos": 31936508035200.0,
      "grad_norm": 1.7904792435575492,
      "language_loss": 0.78683239,
      "learning_rate": 3.96339583888261e-06,
      "loss": 0.80893862,
      "num_input_tokens_seen": 31704630,
      "step": 1483,
      "time_per_iteration": 2.869084119796753
    },
    {
      "auxiliary_loss_clip": 0.0116904,
      "auxiliary_loss_mlp": 0.01082082,
      "balance_loss_clip": 1.05540919,
      "balance_loss_mlp": 1.05829978,
      "epoch": 0.08922290695926649,
      "flos": 17530404969600.0,
      "grad_norm": 2.2229749189835677,
      "language_loss": 0.85424453,
      "learning_rate": 3.963321630732448e-06,
      "loss": 0.87675571,
      "num_input_tokens_seen": 31723255,
      "step": 1484,
      "time_per_iteration": 4.280332326889038
    },
    {
      "auxiliary_loss_clip": 0.01199312,
      "auxiliary_loss_mlp": 0.01060639,
      "balance_loss_clip": 1.06350458,
      "balance_loss_mlp": 1.03701186,
      "epoch": 0.08928303021193447,
      "flos": 32125075459200.0,
      "grad_norm": 1.7208139316694195,
      "language_loss": 0.80205405,
      "learning_rate": 3.963247348132932e-06,
      "loss": 0.82465357,
      "num_input_tokens_seen": 31747045,
      "step": 1485,
      "time_per_iteration": 2.761733055114746
    },
    {
      "auxiliary_loss_clip": 0.01173167,
      "auxiliary_loss_mlp": 0.01056554,
      "balance_loss_clip": 1.0563333,
      "balance_loss_mlp": 1.03228331,
      "epoch": 0.08934315346460243,
      "flos": 22125210785280.0,
      "grad_norm": 1.8969438127775513,
      "language_loss": 0.82859123,
      "learning_rate": 3.96317299108688e-06,
      "loss": 0.85088843,
      "num_input_tokens_seen": 31766615,
      "step": 1486,
      "time_per_iteration": 4.144649028778076
    },
    {
      "auxiliary_loss_clip": 0.01144509,
      "auxiliary_loss_mlp": 0.01063805,
      "balance_loss_clip": 1.05592823,
      "balance_loss_mlp": 1.04021382,
      "epoch": 0.0894032767172704,
      "flos": 22565583527040.0,
      "grad_norm": 2.1520807598980185,
      "language_loss": 0.76365155,
      "learning_rate": 3.963098559597111e-06,
      "loss": 0.78573477,
      "num_input_tokens_seen": 31785855,
      "step": 1487,
      "time_per_iteration": 4.432489395141602
    },
    {
      "auxiliary_loss_clip": 0.01157327,
      "auxiliary_loss_mlp": 0.01060261,
      "balance_loss_clip": 1.05041027,
      "balance_loss_mlp": 1.03542995,
      "epoch": 0.08946339996993838,
      "flos": 20193396503040.0,
      "grad_norm": 3.851280697857004,
      "language_loss": 0.83030224,
      "learning_rate": 3.963024053666449e-06,
      "loss": 0.85247803,
      "num_input_tokens_seen": 31804210,
      "step": 1488,
      "time_per_iteration": 2.7262001037597656
    },
    {
      "auxiliary_loss_clip": 0.01171869,
      "auxiliary_loss_mlp": 0.01051875,
      "balance_loss_clip": 1.05546355,
      "balance_loss_mlp": 1.02916527,
      "epoch": 0.08952352322260634,
      "flos": 48360181104000.0,
      "grad_norm": 1.7759111472560039,
      "language_loss": 0.71783459,
      "learning_rate": 3.962949473297718e-06,
      "loss": 0.74007201,
      "num_input_tokens_seen": 31826150,
      "step": 1489,
      "time_per_iteration": 4.562536954879761
    },
    {
      "auxiliary_loss_clip": 0.01150585,
      "auxiliary_loss_mlp": 0.01051382,
      "balance_loss_clip": 1.05190349,
      "balance_loss_mlp": 1.02830291,
      "epoch": 0.08958364647527431,
      "flos": 31793081028480.0,
      "grad_norm": 1.6999724957706692,
      "language_loss": 0.89717221,
      "learning_rate": 3.962874818493745e-06,
      "loss": 0.91919196,
      "num_input_tokens_seen": 31848060,
      "step": 1490,
      "time_per_iteration": 2.838327646255493
    },
    {
      "auxiliary_loss_clip": 0.01184278,
      "auxiliary_loss_mlp": 0.01064168,
      "balance_loss_clip": 1.05656135,
      "balance_loss_mlp": 1.04102957,
      "epoch": 0.08964376972794229,
      "flos": 23368186972800.0,
      "grad_norm": 3.9062133325383126,
      "language_loss": 0.73075998,
      "learning_rate": 3.9628000892573635e-06,
      "loss": 0.7532444,
      "num_input_tokens_seen": 31870040,
      "step": 1491,
      "time_per_iteration": 2.7007367610931396
    },
    {
      "auxiliary_loss_clip": 0.01189564,
      "auxiliary_loss_mlp": 0.00780167,
      "balance_loss_clip": 1.05968356,
      "balance_loss_mlp": 1.00023544,
      "epoch": 0.08970389298061025,
      "flos": 23294785530240.0,
      "grad_norm": 1.7021050418948058,
      "language_loss": 0.77235049,
      "learning_rate": 3.9627252855914055e-06,
      "loss": 0.79204774,
      "num_input_tokens_seen": 31890400,
      "step": 1492,
      "time_per_iteration": 2.7799623012542725
    },
    {
      "auxiliary_loss_clip": 0.01187114,
      "auxiliary_loss_mlp": 0.01057952,
      "balance_loss_clip": 1.05902028,
      "balance_loss_mlp": 1.03512359,
      "epoch": 0.08976401623327822,
      "flos": 33761703772800.0,
      "grad_norm": 1.9236790530591625,
      "language_loss": 0.71429193,
      "learning_rate": 3.962650407498707e-06,
      "loss": 0.73674262,
      "num_input_tokens_seen": 31913435,
      "step": 1493,
      "time_per_iteration": 2.8479840755462646
    },
    {
      "auxiliary_loss_clip": 0.01188796,
      "auxiliary_loss_mlp": 0.01057103,
      "balance_loss_clip": 1.05757976,
      "balance_loss_mlp": 1.03371406,
      "epoch": 0.08982413948594618,
      "flos": 23911335504000.0,
      "grad_norm": 2.6977604073852053,
      "language_loss": 0.87175488,
      "learning_rate": 3.962575454982109e-06,
      "loss": 0.8942138,
      "num_input_tokens_seen": 31932435,
      "step": 1494,
      "time_per_iteration": 2.855658769607544
    },
    {
      "auxiliary_loss_clip": 0.0108466,
      "auxiliary_loss_mlp": 0.01070478,
      "balance_loss_clip": 1.04641223,
      "balance_loss_mlp": 1.04551601,
      "epoch": 0.08988426273861416,
      "flos": 16837544551680.0,
      "grad_norm": 1.6162523894431247,
      "language_loss": 0.82929438,
      "learning_rate": 3.962500428044454e-06,
      "loss": 0.85084569,
      "num_input_tokens_seen": 31950125,
      "step": 1495,
      "time_per_iteration": 2.9265449047088623
    },
    {
      "auxiliary_loss_clip": 0.01171464,
      "auxiliary_loss_mlp": 0.01059756,
      "balance_loss_clip": 1.05779243,
      "balance_loss_mlp": 1.03682017,
      "epoch": 0.08994438599128213,
      "flos": 14793365548800.0,
      "grad_norm": 9.387255385257733,
      "language_loss": 0.70191383,
      "learning_rate": 3.962425326688585e-06,
      "loss": 0.72422606,
      "num_input_tokens_seen": 31968050,
      "step": 1496,
      "time_per_iteration": 2.773693799972534
    },
    {
      "auxiliary_loss_clip": 0.01164171,
      "auxiliary_loss_mlp": 0.01049454,
      "balance_loss_clip": 1.05397439,
      "balance_loss_mlp": 1.02888989,
      "epoch": 0.09000450924395009,
      "flos": 17384320356480.0,
      "grad_norm": 1.6327835891742186,
      "language_loss": 0.79752576,
      "learning_rate": 3.962350150917351e-06,
      "loss": 0.81966203,
      "num_input_tokens_seen": 31985675,
      "step": 1497,
      "time_per_iteration": 2.6850852966308594
    },
    {
      "auxiliary_loss_clip": 0.01129609,
      "auxiliary_loss_mlp": 0.01054903,
      "balance_loss_clip": 1.05307686,
      "balance_loss_mlp": 1.03146648,
      "epoch": 0.09006463249661807,
      "flos": 24280317964800.0,
      "grad_norm": 8.517000212139891,
      "language_loss": 0.82940567,
      "learning_rate": 3.9622749007336035e-06,
      "loss": 0.85125089,
      "num_input_tokens_seen": 32005180,
      "step": 1498,
      "time_per_iteration": 2.786205768585205
    },
    {
      "auxiliary_loss_clip": 0.01170006,
      "auxiliary_loss_mlp": 0.01059397,
      "balance_loss_clip": 1.0577898,
      "balance_loss_mlp": 1.03718853,
      "epoch": 0.09012475574928604,
      "flos": 13661928069120.0,
      "grad_norm": 2.220597323082783,
      "language_loss": 0.78609937,
      "learning_rate": 3.962199576140195e-06,
      "loss": 0.80839342,
      "num_input_tokens_seen": 32022970,
      "step": 1499,
      "time_per_iteration": 2.71785831451416
    },
    {
      "auxiliary_loss_clip": 0.01161539,
      "auxiliary_loss_mlp": 0.00780528,
      "balance_loss_clip": 1.05444527,
      "balance_loss_mlp": 1.00024021,
      "epoch": 0.090184879001954,
      "flos": 23327751237120.0,
      "grad_norm": 2.049001350461653,
      "language_loss": 0.93337607,
      "learning_rate": 3.962124177139981e-06,
      "loss": 0.95279682,
      "num_input_tokens_seen": 32043055,
      "step": 1500,
      "time_per_iteration": 2.7077536582946777
    },
    {
      "auxiliary_loss_clip": 0.01148009,
      "auxiliary_loss_mlp": 0.01055246,
      "balance_loss_clip": 1.05371249,
      "balance_loss_mlp": 1.0308435,
      "epoch": 0.09024500225462198,
      "flos": 23002688131200.0,
      "grad_norm": 3.0778515668575492,
      "language_loss": 0.74595469,
      "learning_rate": 3.962048703735822e-06,
      "loss": 0.76798725,
      "num_input_tokens_seen": 32061900,
      "step": 1501,
      "time_per_iteration": 2.7073416709899902
    },
    {
      "auxiliary_loss_clip": 0.01056535,
      "auxiliary_loss_mlp": 0.01013118,
      "balance_loss_clip": 1.03392363,
      "balance_loss_mlp": 1.00963676,
      "epoch": 0.09030512550728995,
      "flos": 62189203242240.0,
      "grad_norm": 0.7274487593473578,
      "language_loss": 0.58316052,
      "learning_rate": 3.96197315593058e-06,
      "loss": 0.60385704,
      "num_input_tokens_seen": 32122745,
      "step": 1502,
      "time_per_iteration": 3.274049997329712
    },
    {
      "auxiliary_loss_clip": 0.0114469,
      "auxiliary_loss_mlp": 0.01062533,
      "balance_loss_clip": 1.04626393,
      "balance_loss_mlp": 1.03896546,
      "epoch": 0.09036524875995791,
      "flos": 38800689171840.0,
      "grad_norm": 2.1727281711500095,
      "language_loss": 0.69501173,
      "learning_rate": 3.961897533727119e-06,
      "loss": 0.71708393,
      "num_input_tokens_seen": 32145125,
      "step": 1503,
      "time_per_iteration": 2.87554669380188
    },
    {
      "auxiliary_loss_clip": 0.01133108,
      "auxiliary_loss_mlp": 0.0105903,
      "balance_loss_clip": 1.04783726,
      "balance_loss_mlp": 1.03660655,
      "epoch": 0.09042537201262588,
      "flos": 21690081429120.0,
      "grad_norm": 2.169205134580129,
      "language_loss": 0.86124271,
      "learning_rate": 3.961821837128306e-06,
      "loss": 0.88316405,
      "num_input_tokens_seen": 32166255,
      "step": 1504,
      "time_per_iteration": 2.844688892364502
    },
    {
      "auxiliary_loss_clip": 0.01146301,
      "auxiliary_loss_mlp": 0.01069714,
      "balance_loss_clip": 1.05341232,
      "balance_loss_mlp": 1.04261804,
      "epoch": 0.09048549526529386,
      "flos": 22267021680000.0,
      "grad_norm": 2.178155372989796,
      "language_loss": 0.7233696,
      "learning_rate": 3.961746066137014e-06,
      "loss": 0.74552977,
      "num_input_tokens_seen": 32184010,
      "step": 1505,
      "time_per_iteration": 2.7992677688598633
    },
    {
      "auxiliary_loss_clip": 0.01137399,
      "auxiliary_loss_mlp": 0.01056414,
      "balance_loss_clip": 1.05097985,
      "balance_loss_mlp": 1.03302479,
      "epoch": 0.09054561851796182,
      "flos": 14610939350400.0,
      "grad_norm": 2.5107188210784526,
      "language_loss": 0.80730999,
      "learning_rate": 3.961670220756114e-06,
      "loss": 0.82924813,
      "num_input_tokens_seen": 32201635,
      "step": 1506,
      "time_per_iteration": 2.7458760738372803
    },
    {
      "auxiliary_loss_clip": 0.01140643,
      "auxiliary_loss_mlp": 0.01053315,
      "balance_loss_clip": 1.05161858,
      "balance_loss_mlp": 1.03197718,
      "epoch": 0.09060574177062979,
      "flos": 27636169916160.0,
      "grad_norm": 2.166956120197676,
      "language_loss": 0.75915337,
      "learning_rate": 3.961594300988482e-06,
      "loss": 0.78109294,
      "num_input_tokens_seen": 32221940,
      "step": 1507,
      "time_per_iteration": 2.873826742172241
    },
    {
      "auxiliary_loss_clip": 0.01051873,
      "auxiliary_loss_mlp": 0.01005715,
      "balance_loss_clip": 1.02043629,
      "balance_loss_mlp": 1.00175714,
      "epoch": 0.09066586502329776,
      "flos": 66085797513600.0,
      "grad_norm": 0.7272435825555993,
      "language_loss": 0.57699698,
      "learning_rate": 3.961518306836998e-06,
      "loss": 0.59757286,
      "num_input_tokens_seen": 32276495,
      "step": 1508,
      "time_per_iteration": 3.064926862716675
    },
    {
      "auxiliary_loss_clip": 0.01165416,
      "auxiliary_loss_mlp": 0.01054804,
      "balance_loss_clip": 1.055233,
      "balance_loss_mlp": 1.03155804,
      "epoch": 0.09072598827596573,
      "flos": 18916449027840.0,
      "grad_norm": 1.7601330807914457,
      "language_loss": 0.85090744,
      "learning_rate": 3.961442238304543e-06,
      "loss": 0.87310958,
      "num_input_tokens_seen": 32294130,
      "step": 1509,
      "time_per_iteration": 2.6664113998413086
    },
    {
      "auxiliary_loss_clip": 0.01168837,
      "auxiliary_loss_mlp": 0.01064138,
      "balance_loss_clip": 1.05745769,
      "balance_loss_mlp": 1.03949761,
      "epoch": 0.0907861115286337,
      "flos": 24821742643200.0,
      "grad_norm": 2.3794507710009203,
      "language_loss": 0.84110659,
      "learning_rate": 3.961366095394002e-06,
      "loss": 0.8634364,
      "num_input_tokens_seen": 32313555,
      "step": 1510,
      "time_per_iteration": 2.783484697341919
    },
    {
      "auxiliary_loss_clip": 0.01153141,
      "auxiliary_loss_mlp": 0.01058569,
      "balance_loss_clip": 1.05423617,
      "balance_loss_mlp": 1.03482211,
      "epoch": 0.09084623478130167,
      "flos": 21652842003840.0,
      "grad_norm": 1.8490761573484715,
      "language_loss": 0.85247588,
      "learning_rate": 3.961289878108262e-06,
      "loss": 0.87459302,
      "num_input_tokens_seen": 32331430,
      "step": 1511,
      "time_per_iteration": 2.714620351791382
    },
    {
      "auxiliary_loss_clip": 0.01145395,
      "auxiliary_loss_mlp": 0.01052919,
      "balance_loss_clip": 1.05182219,
      "balance_loss_mlp": 1.02983987,
      "epoch": 0.09090635803396964,
      "flos": 27639258485760.0,
      "grad_norm": 1.5734326837562458,
      "language_loss": 0.84977764,
      "learning_rate": 3.9612135864502135e-06,
      "loss": 0.87176073,
      "num_input_tokens_seen": 32353705,
      "step": 1512,
      "time_per_iteration": 2.75361704826355
    },
    {
      "auxiliary_loss_clip": 0.01155239,
      "auxiliary_loss_mlp": 0.01053669,
      "balance_loss_clip": 1.05740952,
      "balance_loss_mlp": 1.03185391,
      "epoch": 0.0909664812866376,
      "flos": 17669127294720.0,
      "grad_norm": 3.0235926431973654,
      "language_loss": 0.87346804,
      "learning_rate": 3.961137220422749e-06,
      "loss": 0.89555705,
      "num_input_tokens_seen": 32370520,
      "step": 1513,
      "time_per_iteration": 2.6864211559295654
    },
    {
      "auxiliary_loss_clip": 0.01168585,
      "auxiliary_loss_mlp": 0.01049408,
      "balance_loss_clip": 1.05562937,
      "balance_loss_mlp": 1.02841544,
      "epoch": 0.09102660453930557,
      "flos": 23951448017280.0,
      "grad_norm": 1.7883280971870592,
      "language_loss": 0.86802679,
      "learning_rate": 3.961060780028764e-06,
      "loss": 0.89020675,
      "num_input_tokens_seen": 32389105,
      "step": 1514,
      "time_per_iteration": 2.6788065433502197
    },
    {
      "auxiliary_loss_clip": 0.01134005,
      "auxiliary_loss_mlp": 0.01064386,
      "balance_loss_clip": 1.05571628,
      "balance_loss_mlp": 1.04252315,
      "epoch": 0.09108672779197355,
      "flos": 25812949426560.0,
      "grad_norm": 1.7666120550996132,
      "language_loss": 0.89944756,
      "learning_rate": 3.960984265271159e-06,
      "loss": 0.92143154,
      "num_input_tokens_seen": 32408065,
      "step": 1515,
      "time_per_iteration": 2.757390022277832
    },
    {
      "auxiliary_loss_clip": 0.01162518,
      "auxiliary_loss_mlp": 0.01056937,
      "balance_loss_clip": 1.05547726,
      "balance_loss_mlp": 1.03360808,
      "epoch": 0.09114685104464151,
      "flos": 29639482220160.0,
      "grad_norm": 2.1090985009837646,
      "language_loss": 0.85576892,
      "learning_rate": 3.9609076761528335e-06,
      "loss": 0.87796342,
      "num_input_tokens_seen": 32427225,
      "step": 1516,
      "time_per_iteration": 2.704784870147705
    },
    {
      "auxiliary_loss_clip": 0.01158781,
      "auxiliary_loss_mlp": 0.01057165,
      "balance_loss_clip": 1.05135357,
      "balance_loss_mlp": 1.03451526,
      "epoch": 0.09120697429730948,
      "flos": 33729635905920.0,
      "grad_norm": 2.086405156201108,
      "language_loss": 0.81167233,
      "learning_rate": 3.960831012676692e-06,
      "loss": 0.83383185,
      "num_input_tokens_seen": 32450510,
      "step": 1517,
      "time_per_iteration": 2.8586854934692383
    },
    {
      "auxiliary_loss_clip": 0.0117857,
      "auxiliary_loss_mlp": 0.01065492,
      "balance_loss_clip": 1.05741739,
      "balance_loss_mlp": 1.04280686,
      "epoch": 0.09126709754997746,
      "flos": 18401381953920.0,
      "grad_norm": 2.104468567304263,
      "language_loss": 0.78067243,
      "learning_rate": 3.960754274845642e-06,
      "loss": 0.80311304,
      "num_input_tokens_seen": 32468425,
      "step": 1518,
      "time_per_iteration": 2.7862088680267334
    },
    {
      "auxiliary_loss_clip": 0.01165395,
      "auxiliary_loss_mlp": 0.01061371,
      "balance_loss_clip": 1.05285823,
      "balance_loss_mlp": 1.03900695,
      "epoch": 0.09132722080264542,
      "flos": 22091957769600.0,
      "grad_norm": 1.6816479812467473,
      "language_loss": 0.86124098,
      "learning_rate": 3.960677462662594e-06,
      "loss": 0.88350856,
      "num_input_tokens_seen": 32487510,
      "step": 1519,
      "time_per_iteration": 2.723714828491211
    },
    {
      "auxiliary_loss_clip": 0.01163599,
      "auxiliary_loss_mlp": 0.01052792,
      "balance_loss_clip": 1.05454183,
      "balance_loss_mlp": 1.02914131,
      "epoch": 0.09138734405531339,
      "flos": 21033131633280.0,
      "grad_norm": 1.9681293960876167,
      "language_loss": 0.73279071,
      "learning_rate": 3.96060057613046e-06,
      "loss": 0.75495458,
      "num_input_tokens_seen": 32507250,
      "step": 1520,
      "time_per_iteration": 2.8098628520965576
    },
    {
      "auxiliary_loss_clip": 0.01161166,
      "auxiliary_loss_mlp": 0.01058035,
      "balance_loss_clip": 1.05696058,
      "balance_loss_mlp": 1.03469419,
      "epoch": 0.09144746730798137,
      "flos": 20083940784000.0,
      "grad_norm": 2.6988457876937066,
      "language_loss": 0.85236609,
      "learning_rate": 3.960523615252156e-06,
      "loss": 0.87455815,
      "num_input_tokens_seen": 32526045,
      "step": 1521,
      "time_per_iteration": 2.7134172916412354
    },
    {
      "auxiliary_loss_clip": 0.01120174,
      "auxiliary_loss_mlp": 0.01063979,
      "balance_loss_clip": 1.05189717,
      "balance_loss_mlp": 1.03991079,
      "epoch": 0.09150759056064933,
      "flos": 22778210085120.0,
      "grad_norm": 1.6991603177293335,
      "language_loss": 0.83933008,
      "learning_rate": 3.960446580030599e-06,
      "loss": 0.8611716,
      "num_input_tokens_seen": 32546575,
      "step": 1522,
      "time_per_iteration": 2.93745493888855
    },
    {
      "auxiliary_loss_clip": 0.01182362,
      "auxiliary_loss_mlp": 0.01064589,
      "balance_loss_clip": 1.05630755,
      "balance_loss_mlp": 1.04153395,
      "epoch": 0.0915677138133173,
      "flos": 27564205017600.0,
      "grad_norm": 1.647915064434875,
      "language_loss": 0.81012994,
      "learning_rate": 3.960369470468711e-06,
      "loss": 0.8325994,
      "num_input_tokens_seen": 32568795,
      "step": 1523,
      "time_per_iteration": 4.378152847290039
    },
    {
      "auxiliary_loss_clip": 0.01157976,
      "auxiliary_loss_mlp": 0.00781395,
      "balance_loss_clip": 1.05422449,
      "balance_loss_mlp": 1.00037968,
      "epoch": 0.09162783706598528,
      "flos": 17674765729920.0,
      "grad_norm": 2.106497620262502,
      "language_loss": 0.7460072,
      "learning_rate": 3.960292286569418e-06,
      "loss": 0.76540089,
      "num_input_tokens_seen": 32587010,
      "step": 1524,
      "time_per_iteration": 2.7146124839782715
    },
    {
      "auxiliary_loss_clip": 0.01135228,
      "auxiliary_loss_mlp": 0.0106119,
      "balance_loss_clip": 1.05092478,
      "balance_loss_mlp": 1.03782487,
      "epoch": 0.09168796031865324,
      "flos": 18478195188480.0,
      "grad_norm": 2.0992608845945413,
      "language_loss": 0.86498803,
      "learning_rate": 3.960215028335644e-06,
      "loss": 0.88695222,
      "num_input_tokens_seen": 32602375,
      "step": 1525,
      "time_per_iteration": 4.314826965332031
    },
    {
      "auxiliary_loss_clip": 0.01164396,
      "auxiliary_loss_mlp": 0.01049506,
      "balance_loss_clip": 1.05688822,
      "balance_loss_mlp": 1.0263319,
      "epoch": 0.0917480835713212,
      "flos": 29387605075200.0,
      "grad_norm": 2.1146348399758237,
      "language_loss": 0.74512708,
      "learning_rate": 3.96013769577032e-06,
      "loss": 0.76726609,
      "num_input_tokens_seen": 32621460,
      "step": 1526,
      "time_per_iteration": 5.878855466842651
    },
    {
      "auxiliary_loss_clip": 0.01186002,
      "auxiliary_loss_mlp": 0.01055817,
      "balance_loss_clip": 1.05732703,
      "balance_loss_mlp": 1.03392982,
      "epoch": 0.09180820682398917,
      "flos": 19829262378240.0,
      "grad_norm": 2.5135282962071215,
      "language_loss": 0.77581728,
      "learning_rate": 3.960060288876378e-06,
      "loss": 0.79823542,
      "num_input_tokens_seen": 32640440,
      "step": 1527,
      "time_per_iteration": 2.693847179412842
    },
    {
      "auxiliary_loss_clip": 0.01173605,
      "auxiliary_loss_mlp": 0.01052264,
      "balance_loss_clip": 1.0534333,
      "balance_loss_mlp": 1.02868414,
      "epoch": 0.09186833007665715,
      "flos": 23841848643840.0,
      "grad_norm": 2.655631139677705,
      "language_loss": 0.78546697,
      "learning_rate": 3.959982807656753e-06,
      "loss": 0.80772561,
      "num_input_tokens_seen": 32660020,
      "step": 1528,
      "time_per_iteration": 2.774219512939453
    },
    {
      "auxiliary_loss_clip": 0.01146017,
      "auxiliary_loss_mlp": 0.01050376,
      "balance_loss_clip": 1.0499053,
      "balance_loss_mlp": 1.02827477,
      "epoch": 0.09192845332932512,
      "flos": 12932726065920.0,
      "grad_norm": 2.682547324044482,
      "language_loss": 0.76732361,
      "learning_rate": 3.959905252114384e-06,
      "loss": 0.78928751,
      "num_input_tokens_seen": 32678170,
      "step": 1529,
      "time_per_iteration": 4.603156089782715
    },
    {
      "auxiliary_loss_clip": 0.01186538,
      "auxiliary_loss_mlp": 0.00780856,
      "balance_loss_clip": 1.05415928,
      "balance_loss_mlp": 1.00045025,
      "epoch": 0.09198857658199308,
      "flos": 24568177559040.0,
      "grad_norm": 1.7410660090049153,
      "language_loss": 0.82906747,
      "learning_rate": 3.959827622252211e-06,
      "loss": 0.84874141,
      "num_input_tokens_seen": 32697540,
      "step": 1530,
      "time_per_iteration": 2.7118582725524902
    },
    {
      "auxiliary_loss_clip": 0.01130108,
      "auxiliary_loss_mlp": 0.0106509,
      "balance_loss_clip": 1.04975331,
      "balance_loss_mlp": 1.04220152,
      "epoch": 0.09204869983466106,
      "flos": 20266941600000.0,
      "grad_norm": 2.182960664479704,
      "language_loss": 0.84001881,
      "learning_rate": 3.959749918073179e-06,
      "loss": 0.86197078,
      "num_input_tokens_seen": 32716805,
      "step": 1531,
      "time_per_iteration": 2.791947603225708
    },
    {
      "auxiliary_loss_clip": 0.0113655,
      "auxiliary_loss_mlp": 0.01051554,
      "balance_loss_clip": 1.04906452,
      "balance_loss_mlp": 1.02853465,
      "epoch": 0.09210882308732903,
      "flos": 20885646389760.0,
      "grad_norm": 1.7570281394880602,
      "language_loss": 0.81253195,
      "learning_rate": 3.959672139580233e-06,
      "loss": 0.83441293,
      "num_input_tokens_seen": 32736385,
      "step": 1532,
      "time_per_iteration": 2.737739324569702
    },
    {
      "auxiliary_loss_clip": 0.01157728,
      "auxiliary_loss_mlp": 0.01056753,
      "balance_loss_clip": 1.052163,
      "balance_loss_mlp": 1.03385305,
      "epoch": 0.09216894633999699,
      "flos": 30956326727040.0,
      "grad_norm": 2.2821036564882182,
      "language_loss": 0.84194255,
      "learning_rate": 3.9595942867763235e-06,
      "loss": 0.86408734,
      "num_input_tokens_seen": 32757140,
      "step": 1533,
      "time_per_iteration": 2.7542598247528076
    },
    {
      "auxiliary_loss_clip": 0.01149262,
      "auxiliary_loss_mlp": 0.01053623,
      "balance_loss_clip": 1.05813503,
      "balance_loss_mlp": 1.03190327,
      "epoch": 0.09222906959266497,
      "flos": 13151565676800.0,
      "grad_norm": 1.9396914937933663,
      "language_loss": 0.9009546,
      "learning_rate": 3.959516359664402e-06,
      "loss": 0.92298347,
      "num_input_tokens_seen": 32774860,
      "step": 1534,
      "time_per_iteration": 2.6450984477996826
    },
    {
      "auxiliary_loss_clip": 0.01150273,
      "auxiliary_loss_mlp": 0.0106298,
      "balance_loss_clip": 1.0495038,
      "balance_loss_mlp": 1.03849435,
      "epoch": 0.09228919284533293,
      "flos": 25994477784960.0,
      "grad_norm": 5.065477266086046,
      "language_loss": 0.75779241,
      "learning_rate": 3.959438358247424e-06,
      "loss": 0.77992499,
      "num_input_tokens_seen": 32795250,
      "step": 1535,
      "time_per_iteration": 2.730915069580078
    },
    {
      "auxiliary_loss_clip": 0.01168283,
      "auxiliary_loss_mlp": 0.01045276,
      "balance_loss_clip": 1.05278873,
      "balance_loss_mlp": 1.02403271,
      "epoch": 0.0923493160980009,
      "flos": 18660800954880.0,
      "grad_norm": 1.8085584532497372,
      "language_loss": 0.81631637,
      "learning_rate": 3.959360282528346e-06,
      "loss": 0.83845198,
      "num_input_tokens_seen": 32813805,
      "step": 1536,
      "time_per_iteration": 2.7326817512512207
    },
    {
      "auxiliary_loss_clip": 0.01181977,
      "auxiliary_loss_mlp": 0.01053699,
      "balance_loss_clip": 1.05431938,
      "balance_loss_mlp": 1.03224182,
      "epoch": 0.09240943935066886,
      "flos": 21140576190720.0,
      "grad_norm": 2.0929096884707556,
      "language_loss": 0.89092755,
      "learning_rate": 3.959282132510131e-06,
      "loss": 0.9132843,
      "num_input_tokens_seen": 32830960,
      "step": 1537,
      "time_per_iteration": 2.675771713256836
    },
    {
      "auxiliary_loss_clip": 0.01157238,
      "auxiliary_loss_mlp": 0.01058647,
      "balance_loss_clip": 1.05114293,
      "balance_loss_mlp": 1.03605688,
      "epoch": 0.09246956260333684,
      "flos": 20592435669120.0,
      "grad_norm": 1.9480116987165197,
      "language_loss": 0.80702311,
      "learning_rate": 3.959203908195741e-06,
      "loss": 0.82918191,
      "num_input_tokens_seen": 32848275,
      "step": 1538,
      "time_per_iteration": 2.71618390083313
    },
    {
      "auxiliary_loss_clip": 0.01060495,
      "auxiliary_loss_mlp": 0.0101237,
      "balance_loss_clip": 1.03095436,
      "balance_loss_mlp": 1.00872231,
      "epoch": 0.09252968585600481,
      "flos": 67558710614400.0,
      "grad_norm": 0.7534074452314953,
      "language_loss": 0.57429332,
      "learning_rate": 3.959125609588142e-06,
      "loss": 0.59502202,
      "num_input_tokens_seen": 32917730,
      "step": 1539,
      "time_per_iteration": 3.3933441638946533
    },
    {
      "auxiliary_loss_clip": 0.01159831,
      "auxiliary_loss_mlp": 0.01050602,
      "balance_loss_clip": 1.05638027,
      "balance_loss_mlp": 1.02863121,
      "epoch": 0.09258980910867277,
      "flos": 17383853479680.0,
      "grad_norm": 2.849299216868502,
      "language_loss": 0.67554641,
      "learning_rate": 3.959047236690304e-06,
      "loss": 0.69765073,
      "num_input_tokens_seen": 32934910,
      "step": 1540,
      "time_per_iteration": 2.757084608078003
    },
    {
      "auxiliary_loss_clip": 0.01144239,
      "auxiliary_loss_mlp": 0.01048444,
      "balance_loss_clip": 1.04954028,
      "balance_loss_mlp": 1.026438,
      "epoch": 0.09264993236134075,
      "flos": 19865927185920.0,
      "grad_norm": 2.044335478602743,
      "language_loss": 0.83917534,
      "learning_rate": 3.958968789505198e-06,
      "loss": 0.86110216,
      "num_input_tokens_seen": 32953840,
      "step": 1541,
      "time_per_iteration": 2.8497180938720703
    },
    {
      "auxiliary_loss_clip": 0.01077839,
      "auxiliary_loss_mlp": 0.01013078,
      "balance_loss_clip": 1.02602255,
      "balance_loss_mlp": 1.0097636,
      "epoch": 0.09271005561400872,
      "flos": 62284401262080.0,
      "grad_norm": 0.8790732834061692,
      "language_loss": 0.61881655,
      "learning_rate": 3.9588902680358e-06,
      "loss": 0.63972563,
      "num_input_tokens_seen": 33011410,
      "step": 1542,
      "time_per_iteration": 3.3079330921173096
    },
    {
      "auxiliary_loss_clip": 0.01161232,
      "auxiliary_loss_mlp": 0.01059438,
      "balance_loss_clip": 1.05441117,
      "balance_loss_mlp": 1.03808808,
      "epoch": 0.09277017886667668,
      "flos": 23329870139520.0,
      "grad_norm": 1.6256118826429122,
      "language_loss": 0.82802349,
      "learning_rate": 3.958811672285086e-06,
      "loss": 0.85023022,
      "num_input_tokens_seen": 33031675,
      "step": 1543,
      "time_per_iteration": 2.7408807277679443
    },
    {
      "auxiliary_loss_clip": 0.01135873,
      "auxiliary_loss_mlp": 0.01060295,
      "balance_loss_clip": 1.04848838,
      "balance_loss_mlp": 1.03863442,
      "epoch": 0.09283030211934466,
      "flos": 54745169875200.0,
      "grad_norm": 1.706948475246468,
      "language_loss": 0.72265279,
      "learning_rate": 3.958733002256038e-06,
      "loss": 0.74461448,
      "num_input_tokens_seen": 33056355,
      "step": 1544,
      "time_per_iteration": 3.104156255722046
    },
    {
      "auxiliary_loss_clip": 0.01166071,
      "auxiliary_loss_mlp": 0.01055881,
      "balance_loss_clip": 1.05165935,
      "balance_loss_mlp": 1.03138375,
      "epoch": 0.09289042537201263,
      "flos": 30334784762880.0,
      "grad_norm": 1.7720844214030114,
      "language_loss": 0.77286768,
      "learning_rate": 3.958654257951637e-06,
      "loss": 0.79508722,
      "num_input_tokens_seen": 33079520,
      "step": 1545,
      "time_per_iteration": 2.808180570602417
    },
    {
      "auxiliary_loss_clip": 0.01140161,
      "auxiliary_loss_mlp": 0.01050495,
      "balance_loss_clip": 1.0526737,
      "balance_loss_mlp": 1.02872682,
      "epoch": 0.09295054862468059,
      "flos": 17746838369280.0,
      "grad_norm": 2.7089619481030076,
      "language_loss": 0.74396008,
      "learning_rate": 3.9585754393748706e-06,
      "loss": 0.76586664,
      "num_input_tokens_seen": 33096135,
      "step": 1546,
      "time_per_iteration": 2.7634081840515137
    },
    {
      "auxiliary_loss_clip": 0.01163775,
      "auxiliary_loss_mlp": 0.0105305,
      "balance_loss_clip": 1.05357957,
      "balance_loss_mlp": 1.02956545,
      "epoch": 0.09301067187734856,
      "flos": 23658021815040.0,
      "grad_norm": 1.9423225100503794,
      "language_loss": 0.84200966,
      "learning_rate": 3.9584965465287275e-06,
      "loss": 0.86417794,
      "num_input_tokens_seen": 33115245,
      "step": 1547,
      "time_per_iteration": 2.790003776550293
    },
    {
      "auxiliary_loss_clip": 0.01141839,
      "auxiliary_loss_mlp": 0.01053941,
      "balance_loss_clip": 1.04740989,
      "balance_loss_mlp": 1.03195918,
      "epoch": 0.09307079513001654,
      "flos": 27527719777920.0,
      "grad_norm": 2.6545433694843488,
      "language_loss": 0.67698336,
      "learning_rate": 3.958417579416199e-06,
      "loss": 0.69894123,
      "num_input_tokens_seen": 33136640,
      "step": 1548,
      "time_per_iteration": 2.8367013931274414
    },
    {
      "auxiliary_loss_clip": 0.01123899,
      "auxiliary_loss_mlp": 0.01059885,
      "balance_loss_clip": 1.04744387,
      "balance_loss_mlp": 1.03754544,
      "epoch": 0.0931309183826845,
      "flos": 20627340710400.0,
      "grad_norm": 1.6829727803454704,
      "language_loss": 0.8326273,
      "learning_rate": 3.9583385380402795e-06,
      "loss": 0.85446513,
      "num_input_tokens_seen": 33155060,
      "step": 1549,
      "time_per_iteration": 2.8462016582489014
    },
    {
      "auxiliary_loss_clip": 0.01176243,
      "auxiliary_loss_mlp": 0.0104617,
      "balance_loss_clip": 1.05815506,
      "balance_loss_mlp": 1.02473652,
      "epoch": 0.09319104163535247,
      "flos": 29020921084800.0,
      "grad_norm": 1.5528514681372962,
      "language_loss": 0.75838119,
      "learning_rate": 3.958259422403966e-06,
      "loss": 0.78060532,
      "num_input_tokens_seen": 33175420,
      "step": 1550,
      "time_per_iteration": 2.7325351238250732
    },
    {
      "auxiliary_loss_clip": 0.01150315,
      "auxiliary_loss_mlp": 0.01069257,
      "balance_loss_clip": 1.05249369,
      "balance_loss_mlp": 1.04483092,
      "epoch": 0.09325116488802045,
      "flos": 25301545539840.0,
      "grad_norm": 2.1922696027472233,
      "language_loss": 0.82828665,
      "learning_rate": 3.95818023251026e-06,
      "loss": 0.85048234,
      "num_input_tokens_seen": 33194120,
      "step": 1551,
      "time_per_iteration": 2.852602481842041
    },
    {
      "auxiliary_loss_clip": 0.01064371,
      "auxiliary_loss_mlp": 0.00760109,
      "balance_loss_clip": 1.02203059,
      "balance_loss_mlp": 0.99984246,
      "epoch": 0.09331128814068841,
      "flos": 61536203942400.0,
      "grad_norm": 0.7384225982202158,
      "language_loss": 0.61837572,
      "learning_rate": 3.958100968362163e-06,
      "loss": 0.63662052,
      "num_input_tokens_seen": 33261080,
      "step": 1552,
      "time_per_iteration": 3.3453099727630615
    },
    {
      "auxiliary_loss_clip": 0.01059175,
      "auxiliary_loss_mlp": 0.01016654,
      "balance_loss_clip": 1.02415061,
      "balance_loss_mlp": 1.01338792,
      "epoch": 0.09337141139335638,
      "flos": 53293700171520.0,
      "grad_norm": 0.8524917480784928,
      "language_loss": 0.58986926,
      "learning_rate": 3.958021629962681e-06,
      "loss": 0.61062753,
      "num_input_tokens_seen": 33330235,
      "step": 1553,
      "time_per_iteration": 3.37673282623291
    },
    {
      "auxiliary_loss_clip": 0.01146955,
      "auxiliary_loss_mlp": 0.01056683,
      "balance_loss_clip": 1.05026984,
      "balance_loss_mlp": 1.03336585,
      "epoch": 0.09343153464602436,
      "flos": 23476852592640.0,
      "grad_norm": 2.3365109182487,
      "language_loss": 0.87665397,
      "learning_rate": 3.957942217314823e-06,
      "loss": 0.8986904,
      "num_input_tokens_seen": 33349035,
      "step": 1554,
      "time_per_iteration": 2.8098127841949463
    },
    {
      "auxiliary_loss_clip": 0.01153047,
      "auxiliary_loss_mlp": 0.01057257,
      "balance_loss_clip": 1.05439448,
      "balance_loss_mlp": 1.03393972,
      "epoch": 0.09349165789869232,
      "flos": 19353481804800.0,
      "grad_norm": 4.388884220182432,
      "language_loss": 0.81678319,
      "learning_rate": 3.957862730421599e-06,
      "loss": 0.83888626,
      "num_input_tokens_seen": 33368060,
      "step": 1555,
      "time_per_iteration": 2.726207971572876
    },
    {
      "auxiliary_loss_clip": 0.01058869,
      "auxiliary_loss_mlp": 0.01003892,
      "balance_loss_clip": 1.0202632,
      "balance_loss_mlp": 1.00045919,
      "epoch": 0.09355178115136029,
      "flos": 67502580635520.0,
      "grad_norm": 0.8683826280274983,
      "language_loss": 0.59606886,
      "learning_rate": 3.957783169286024e-06,
      "loss": 0.61669648,
      "num_input_tokens_seen": 33430825,
      "step": 1556,
      "time_per_iteration": 3.209326982498169
    },
    {
      "auxiliary_loss_clip": 0.01174249,
      "auxiliary_loss_mlp": 0.01059741,
      "balance_loss_clip": 1.05518138,
      "balance_loss_mlp": 1.03727031,
      "epoch": 0.09361190440402825,
      "flos": 37341638720640.0,
      "grad_norm": 1.6803158790244075,
      "language_loss": 0.84290808,
      "learning_rate": 3.9577035339111155e-06,
      "loss": 0.86524796,
      "num_input_tokens_seen": 33454855,
      "step": 1557,
      "time_per_iteration": 2.831650733947754
    },
    {
      "auxiliary_loss_clip": 0.01110857,
      "auxiliary_loss_mlp": 0.01065156,
      "balance_loss_clip": 1.04900038,
      "balance_loss_mlp": 1.04112351,
      "epoch": 0.09367202765669623,
      "flos": 24899705112960.0,
      "grad_norm": 1.6725809358966677,
      "language_loss": 0.780913,
      "learning_rate": 3.957623824299893e-06,
      "loss": 0.8026731,
      "num_input_tokens_seen": 33476000,
      "step": 1558,
      "time_per_iteration": 3.0111780166625977
    },
    {
      "auxiliary_loss_clip": 0.01164994,
      "auxiliary_loss_mlp": 0.01051229,
      "balance_loss_clip": 1.0558666,
      "balance_loss_mlp": 1.02881753,
      "epoch": 0.0937321509093642,
      "flos": 15705568368000.0,
      "grad_norm": 2.0141986314124414,
      "language_loss": 0.80066288,
      "learning_rate": 3.957544040455379e-06,
      "loss": 0.82282507,
      "num_input_tokens_seen": 33493845,
      "step": 1559,
      "time_per_iteration": 3.024117946624756
    },
    {
      "auxiliary_loss_clip": 0.01141277,
      "auxiliary_loss_mlp": 0.01061718,
      "balance_loss_clip": 1.05060387,
      "balance_loss_mlp": 1.04012942,
      "epoch": 0.09379227416203216,
      "flos": 20483698222080.0,
      "grad_norm": 1.8358373674042003,
      "language_loss": 0.76418209,
      "learning_rate": 3.957464182380599e-06,
      "loss": 0.78621197,
      "num_input_tokens_seen": 33510850,
      "step": 1560,
      "time_per_iteration": 2.68558406829834
    },
    {
      "auxiliary_loss_clip": 0.01137939,
      "auxiliary_loss_mlp": 0.01054925,
      "balance_loss_clip": 1.05014277,
      "balance_loss_mlp": 1.03213274,
      "epoch": 0.09385239741470014,
      "flos": 24352498344960.0,
      "grad_norm": 3.575155933252121,
      "language_loss": 0.80784953,
      "learning_rate": 3.95738425007858e-06,
      "loss": 0.82977819,
      "num_input_tokens_seen": 33530430,
      "step": 1561,
      "time_per_iteration": 2.759148359298706
    },
    {
      "auxiliary_loss_clip": 0.01173652,
      "auxiliary_loss_mlp": 0.01052448,
      "balance_loss_clip": 1.05276573,
      "balance_loss_mlp": 1.02989376,
      "epoch": 0.0939125206673681,
      "flos": 33291489807360.0,
      "grad_norm": 2.448664627367939,
      "language_loss": 0.6140722,
      "learning_rate": 3.957304243552354e-06,
      "loss": 0.63633323,
      "num_input_tokens_seen": 33551975,
      "step": 1562,
      "time_per_iteration": 2.9014978408813477
    },
    {
      "auxiliary_loss_clip": 0.01162693,
      "auxiliary_loss_mlp": 0.0106374,
      "balance_loss_clip": 1.05719543,
      "balance_loss_mlp": 1.04213953,
      "epoch": 0.09397264392003607,
      "flos": 19244923925760.0,
      "grad_norm": 3.5098220300578555,
      "language_loss": 0.8496151,
      "learning_rate": 3.957224162804956e-06,
      "loss": 0.87187934,
      "num_input_tokens_seen": 33569850,
      "step": 1563,
      "time_per_iteration": 4.404061555862427
    },
    {
      "auxiliary_loss_clip": 0.01164811,
      "auxiliary_loss_mlp": 0.01047932,
      "balance_loss_clip": 1.05775142,
      "balance_loss_mlp": 1.02652228,
      "epoch": 0.09403276717270405,
      "flos": 19317930318720.0,
      "grad_norm": 1.6765528861156813,
      "language_loss": 0.76511294,
      "learning_rate": 3.9571440078394205e-06,
      "loss": 0.78724039,
      "num_input_tokens_seen": 33590510,
      "step": 1564,
      "time_per_iteration": 4.255565166473389
    },
    {
      "auxiliary_loss_clip": 0.01151297,
      "auxiliary_loss_mlp": 0.01063256,
      "balance_loss_clip": 1.05196142,
      "balance_loss_mlp": 1.04172707,
      "epoch": 0.09409289042537201,
      "flos": 23583471137280.0,
      "grad_norm": 1.9762038777899962,
      "language_loss": 0.80134326,
      "learning_rate": 3.9570637786587895e-06,
      "loss": 0.82348871,
      "num_input_tokens_seen": 33608810,
      "step": 1565,
      "time_per_iteration": 2.8548545837402344
    },
    {
      "auxiliary_loss_clip": 0.01158602,
      "auxiliary_loss_mlp": 0.01063767,
      "balance_loss_clip": 1.05420566,
      "balance_loss_mlp": 1.04233313,
      "epoch": 0.09415301367803998,
      "flos": 20078446003200.0,
      "grad_norm": 1.6810250981626251,
      "language_loss": 0.75134379,
      "learning_rate": 3.956983475266103e-06,
      "loss": 0.77356744,
      "num_input_tokens_seen": 33627265,
      "step": 1566,
      "time_per_iteration": 4.889045715332031
    },
    {
      "auxiliary_loss_clip": 0.01145856,
      "auxiliary_loss_mlp": 0.00780689,
      "balance_loss_clip": 1.05168366,
      "balance_loss_mlp": 1.00022864,
      "epoch": 0.09421313693070796,
      "flos": 21062075016960.0,
      "grad_norm": 1.6828919748843199,
      "language_loss": 0.77958012,
      "learning_rate": 3.956903097664407e-06,
      "loss": 0.79884553,
      "num_input_tokens_seen": 33644810,
      "step": 1567,
      "time_per_iteration": 4.445765972137451
    },
    {
      "auxiliary_loss_clip": 0.01156815,
      "auxiliary_loss_mlp": 0.01056228,
      "balance_loss_clip": 1.05256855,
      "balance_loss_mlp": 1.03591454,
      "epoch": 0.09427326018337592,
      "flos": 24316156759680.0,
      "grad_norm": 2.008686295040646,
      "language_loss": 0.82608044,
      "learning_rate": 3.956822645856749e-06,
      "loss": 0.84821093,
      "num_input_tokens_seen": 33665665,
      "step": 1568,
      "time_per_iteration": 2.881535768508911
    },
    {
      "auxiliary_loss_clip": 0.01187915,
      "auxiliary_loss_mlp": 0.01051731,
      "balance_loss_clip": 1.05717778,
      "balance_loss_mlp": 1.02927184,
      "epoch": 0.09433338343604389,
      "flos": 20263888944000.0,
      "grad_norm": 1.9573151026586577,
      "language_loss": 0.76943743,
      "learning_rate": 3.9567421198461814e-06,
      "loss": 0.79183388,
      "num_input_tokens_seen": 33684760,
      "step": 1569,
      "time_per_iteration": 2.6097726821899414
    },
    {
      "auxiliary_loss_clip": 0.01120191,
      "auxiliary_loss_mlp": 0.01060805,
      "balance_loss_clip": 1.04771852,
      "balance_loss_mlp": 1.03625941,
      "epoch": 0.09439350668871185,
      "flos": 12742973493120.0,
      "grad_norm": 3.3813700161908917,
      "language_loss": 0.85488856,
      "learning_rate": 3.956661519635756e-06,
      "loss": 0.87669849,
      "num_input_tokens_seen": 33700750,
      "step": 1570,
      "time_per_iteration": 2.7571377754211426
    },
    {
      "auxiliary_loss_clip": 0.01122458,
      "auxiliary_loss_mlp": 0.01055939,
      "balance_loss_clip": 1.04927301,
      "balance_loss_mlp": 1.03183508,
      "epoch": 0.09445362994137983,
      "flos": 25962266263680.0,
      "grad_norm": 1.540414635950846,
      "language_loss": 0.76415235,
      "learning_rate": 3.95658084522853e-06,
      "loss": 0.7859363,
      "num_input_tokens_seen": 33724430,
      "step": 1571,
      "time_per_iteration": 2.913569211959839
    },
    {
      "auxiliary_loss_clip": 0.01135683,
      "auxiliary_loss_mlp": 0.01057111,
      "balance_loss_clip": 1.0490278,
      "balance_loss_mlp": 1.0349735,
      "epoch": 0.0945137531940478,
      "flos": 19715353372800.0,
      "grad_norm": 1.6745378641752047,
      "language_loss": 0.79397607,
      "learning_rate": 3.956500096627561e-06,
      "loss": 0.81590402,
      "num_input_tokens_seen": 33743455,
      "step": 1572,
      "time_per_iteration": 2.813410758972168
    },
    {
      "auxiliary_loss_clip": 0.01148251,
      "auxiliary_loss_mlp": 0.0106927,
      "balance_loss_clip": 1.05619979,
      "balance_loss_mlp": 1.04524922,
      "epoch": 0.09457387644671576,
      "flos": 23617047375360.0,
      "grad_norm": 1.7559396294879055,
      "language_loss": 0.87707287,
      "learning_rate": 3.956419273835913e-06,
      "loss": 0.89924812,
      "num_input_tokens_seen": 33763435,
      "step": 1573,
      "time_per_iteration": 2.776535987854004
    },
    {
      "auxiliary_loss_clip": 0.01161183,
      "auxiliary_loss_mlp": 0.01063326,
      "balance_loss_clip": 1.05485129,
      "balance_loss_mlp": 1.03804219,
      "epoch": 0.09463399969938374,
      "flos": 26907291135360.0,
      "grad_norm": 2.9707854698090097,
      "language_loss": 0.81982428,
      "learning_rate": 3.95633837685665e-06,
      "loss": 0.84206939,
      "num_input_tokens_seen": 33784325,
      "step": 1574,
      "time_per_iteration": 2.7604806423187256
    },
    {
      "auxiliary_loss_clip": 0.01156287,
      "auxiliary_loss_mlp": 0.01055594,
      "balance_loss_clip": 1.05234718,
      "balance_loss_mlp": 1.0344342,
      "epoch": 0.0946941229520517,
      "flos": 23659566099840.0,
      "grad_norm": 1.7178511535677499,
      "language_loss": 0.80855322,
      "learning_rate": 3.95625740569284e-06,
      "loss": 0.83067203,
      "num_input_tokens_seen": 33802510,
      "step": 1575,
      "time_per_iteration": 2.713247299194336
    },
    {
      "auxiliary_loss_clip": 0.01182326,
      "auxiliary_loss_mlp": 0.01068689,
      "balance_loss_clip": 1.05578864,
      "balance_loss_mlp": 1.04581285,
      "epoch": 0.09475424620471967,
      "flos": 24134053783680.0,
      "grad_norm": 1.9110861379460222,
      "language_loss": 0.86483347,
      "learning_rate": 3.956176360347553e-06,
      "loss": 0.88734365,
      "num_input_tokens_seen": 33819980,
      "step": 1576,
      "time_per_iteration": 2.682644844055176
    },
    {
      "auxiliary_loss_clip": 0.01056441,
      "auxiliary_loss_mlp": 0.01027284,
      "balance_loss_clip": 1.0225811,
      "balance_loss_mlp": 1.02344561,
      "epoch": 0.09481436945738765,
      "flos": 68426168065920.0,
      "grad_norm": 0.9789918611127905,
      "language_loss": 0.6582402,
      "learning_rate": 3.956095240823862e-06,
      "loss": 0.67907751,
      "num_input_tokens_seen": 33878925,
      "step": 1577,
      "time_per_iteration": 3.2106685638427734
    },
    {
      "auxiliary_loss_clip": 0.01147668,
      "auxiliary_loss_mlp": 0.01051958,
      "balance_loss_clip": 1.05218005,
      "balance_loss_mlp": 1.03098869,
      "epoch": 0.09487449271005562,
      "flos": 16654076858880.0,
      "grad_norm": 1.8223175005615506,
      "language_loss": 0.79152733,
      "learning_rate": 3.956014047124844e-06,
      "loss": 0.81352365,
      "num_input_tokens_seen": 33897600,
      "step": 1578,
      "time_per_iteration": 2.820089340209961
    },
    {
      "auxiliary_loss_clip": 0.01185941,
      "auxiliary_loss_mlp": 0.01066432,
      "balance_loss_clip": 1.05838132,
      "balance_loss_mlp": 1.04437804,
      "epoch": 0.09493461596272358,
      "flos": 24275685110400.0,
      "grad_norm": 3.480730999818176,
      "language_loss": 0.78161818,
      "learning_rate": 3.955932779253578e-06,
      "loss": 0.80414188,
      "num_input_tokens_seen": 33917365,
      "step": 1579,
      "time_per_iteration": 2.6518983840942383
    },
    {
      "auxiliary_loss_clip": 0.01128319,
      "auxiliary_loss_mlp": 0.01065633,
      "balance_loss_clip": 1.04771328,
      "balance_loss_mlp": 1.04001498,
      "epoch": 0.09499473921539155,
      "flos": 21870173243520.0,
      "grad_norm": 2.0084876987684526,
      "language_loss": 0.73410392,
      "learning_rate": 3.955851437213144e-06,
      "loss": 0.75604343,
      "num_input_tokens_seen": 33936680,
      "step": 1580,
      "time_per_iteration": 2.679461717605591
    },
    {
      "auxiliary_loss_clip": 0.01157568,
      "auxiliary_loss_mlp": 0.01062628,
      "balance_loss_clip": 1.05573344,
      "balance_loss_mlp": 1.04095626,
      "epoch": 0.09505486246805953,
      "flos": 33547137880320.0,
      "grad_norm": 14.809542792179553,
      "language_loss": 0.77565914,
      "learning_rate": 3.955770021006627e-06,
      "loss": 0.7978611,
      "num_input_tokens_seen": 33960685,
      "step": 1581,
      "time_per_iteration": 2.765394449234009
    },
    {
      "auxiliary_loss_clip": 0.01144835,
      "auxiliary_loss_mlp": 0.0106468,
      "balance_loss_clip": 1.05426359,
      "balance_loss_mlp": 1.04276967,
      "epoch": 0.09511498572072749,
      "flos": 21215342350080.0,
      "grad_norm": 1.8617167187056045,
      "language_loss": 0.87230825,
      "learning_rate": 3.955688530637116e-06,
      "loss": 0.89440346,
      "num_input_tokens_seen": 33980015,
      "step": 1582,
      "time_per_iteration": 2.691364288330078
    },
    {
      "auxiliary_loss_clip": 0.01174295,
      "auxiliary_loss_mlp": 0.0106431,
      "balance_loss_clip": 1.05508888,
      "balance_loss_mlp": 1.04039705,
      "epoch": 0.09517510897339546,
      "flos": 14611262572800.0,
      "grad_norm": 1.8512060219658202,
      "language_loss": 0.67043924,
      "learning_rate": 3.955606966107699e-06,
      "loss": 0.69282532,
      "num_input_tokens_seen": 33997705,
      "step": 1583,
      "time_per_iteration": 2.6693732738494873
    },
    {
      "auxiliary_loss_clip": 0.01177751,
      "auxiliary_loss_mlp": 0.01053743,
      "balance_loss_clip": 1.0593859,
      "balance_loss_mlp": 1.03035378,
      "epoch": 0.09523523222606343,
      "flos": 27817339138560.0,
      "grad_norm": 2.144216926782962,
      "language_loss": 0.70752859,
      "learning_rate": 3.95552532742147e-06,
      "loss": 0.7298435,
      "num_input_tokens_seen": 34017465,
      "step": 1584,
      "time_per_iteration": 2.7164390087127686
    },
    {
      "auxiliary_loss_clip": 0.01138507,
      "auxiliary_loss_mlp": 0.0105762,
      "balance_loss_clip": 1.05243039,
      "balance_loss_mlp": 1.03584039,
      "epoch": 0.0952953554787314,
      "flos": 20706272847360.0,
      "grad_norm": 1.4654737580846544,
      "language_loss": 0.8080442,
      "learning_rate": 3.955443614581525e-06,
      "loss": 0.83000553,
      "num_input_tokens_seen": 34038550,
      "step": 1585,
      "time_per_iteration": 2.879831314086914
    },
    {
      "auxiliary_loss_clip": 0.01159374,
      "auxiliary_loss_mlp": 0.01057717,
      "balance_loss_clip": 1.05387473,
      "balance_loss_mlp": 1.03355336,
      "epoch": 0.09535547873139937,
      "flos": 24787627701120.0,
      "grad_norm": 1.638250735795891,
      "language_loss": 0.71921158,
      "learning_rate": 3.955361827590961e-06,
      "loss": 0.74138248,
      "num_input_tokens_seen": 34058665,
      "step": 1586,
      "time_per_iteration": 2.750436544418335
    },
    {
      "auxiliary_loss_clip": 0.01048565,
      "auxiliary_loss_mlp": 0.01003302,
      "balance_loss_clip": 1.03115988,
      "balance_loss_mlp": 0.99901009,
      "epoch": 0.09541560198406734,
      "flos": 71912194905600.0,
      "grad_norm": 0.8099482252624973,
      "language_loss": 0.55475175,
      "learning_rate": 3.955279966452883e-06,
      "loss": 0.57527041,
      "num_input_tokens_seen": 34109655,
      "step": 1587,
      "time_per_iteration": 3.0975699424743652
    },
    {
      "auxiliary_loss_clip": 0.01128884,
      "auxiliary_loss_mlp": 0.0105965,
      "balance_loss_clip": 1.04768586,
      "balance_loss_mlp": 1.03661847,
      "epoch": 0.09547572523673531,
      "flos": 28982604251520.0,
      "grad_norm": 1.708481785076906,
      "language_loss": 0.81062275,
      "learning_rate": 3.955198031170391e-06,
      "loss": 0.83250809,
      "num_input_tokens_seen": 34131115,
      "step": 1588,
      "time_per_iteration": 2.7718451023101807
    },
    {
      "auxiliary_loss_clip": 0.01131602,
      "auxiliary_loss_mlp": 0.01056117,
      "balance_loss_clip": 1.04894614,
      "balance_loss_mlp": 1.03438473,
      "epoch": 0.09553584848940327,
      "flos": 24133910129280.0,
      "grad_norm": 1.5119879232668088,
      "language_loss": 0.81481898,
      "learning_rate": 3.955116021746594e-06,
      "loss": 0.83669615,
      "num_input_tokens_seen": 34151925,
      "step": 1589,
      "time_per_iteration": 2.782468795776367
    },
    {
      "auxiliary_loss_clip": 0.0112194,
      "auxiliary_loss_mlp": 0.00780573,
      "balance_loss_clip": 1.0508883,
      "balance_loss_mlp": 1.00013089,
      "epoch": 0.09559597174207124,
      "flos": 42851376789120.0,
      "grad_norm": 1.525287399882202,
      "language_loss": 0.64882791,
      "learning_rate": 3.955033938184601e-06,
      "loss": 0.667853,
      "num_input_tokens_seen": 34175395,
      "step": 1590,
      "time_per_iteration": 3.0783450603485107
    },
    {
      "auxiliary_loss_clip": 0.01143501,
      "auxiliary_loss_mlp": 0.01058399,
      "balance_loss_clip": 1.05087948,
      "balance_loss_mlp": 1.0358206,
      "epoch": 0.09565609499473922,
      "flos": 32670845683200.0,
      "grad_norm": 2.0745314237741916,
      "language_loss": 0.83290577,
      "learning_rate": 3.954951780487526e-06,
      "loss": 0.85492468,
      "num_input_tokens_seen": 34197760,
      "step": 1591,
      "time_per_iteration": 2.8393962383270264
    },
    {
      "auxiliary_loss_clip": 0.01163486,
      "auxiliary_loss_mlp": 0.01065588,
      "balance_loss_clip": 1.0522387,
      "balance_loss_mlp": 1.04266405,
      "epoch": 0.09571621824740718,
      "flos": 18478410670080.0,
      "grad_norm": 2.825705290827541,
      "language_loss": 0.74087322,
      "learning_rate": 3.9548695486584835e-06,
      "loss": 0.76316392,
      "num_input_tokens_seen": 34215330,
      "step": 1592,
      "time_per_iteration": 2.6828882694244385
    },
    {
      "auxiliary_loss_clip": 0.01169239,
      "auxiliary_loss_mlp": 0.01055073,
      "balance_loss_clip": 1.05161428,
      "balance_loss_mlp": 1.03337741,
      "epoch": 0.09577634150007515,
      "flos": 29387497334400.0,
      "grad_norm": 2.18277080043521,
      "language_loss": 0.74483889,
      "learning_rate": 3.954787242700592e-06,
      "loss": 0.76708198,
      "num_input_tokens_seen": 34237745,
      "step": 1593,
      "time_per_iteration": 2.7193498611450195
    },
    {
      "auxiliary_loss_clip": 0.01177343,
      "auxiliary_loss_mlp": 0.01055096,
      "balance_loss_clip": 1.05910873,
      "balance_loss_mlp": 1.03307831,
      "epoch": 0.09583646475274313,
      "flos": 22747830157440.0,
      "grad_norm": 1.887493467708827,
      "language_loss": 0.69782627,
      "learning_rate": 3.954704862616971e-06,
      "loss": 0.72015071,
      "num_input_tokens_seen": 34256565,
      "step": 1594,
      "time_per_iteration": 2.635383367538452
    },
    {
      "auxiliary_loss_clip": 0.01173222,
      "auxiliary_loss_mlp": 0.01051806,
      "balance_loss_clip": 1.05618978,
      "balance_loss_mlp": 1.03037214,
      "epoch": 0.0958965880054111,
      "flos": 23218367345280.0,
      "grad_norm": 2.1411006117727682,
      "language_loss": 0.82780552,
      "learning_rate": 3.954622408410747e-06,
      "loss": 0.85005581,
      "num_input_tokens_seen": 34275970,
      "step": 1595,
      "time_per_iteration": 2.7158257961273193
    },
    {
      "auxiliary_loss_clip": 0.01153253,
      "auxiliary_loss_mlp": 0.01054246,
      "balance_loss_clip": 1.05143809,
      "balance_loss_mlp": 1.0301652,
      "epoch": 0.09595671125807906,
      "flos": 21324438933120.0,
      "grad_norm": 1.7751890788987925,
      "language_loss": 0.84513396,
      "learning_rate": 3.954539880085045e-06,
      "loss": 0.86720896,
      "num_input_tokens_seen": 34295490,
      "step": 1596,
      "time_per_iteration": 2.710228204727173
    },
    {
      "auxiliary_loss_clip": 0.01166586,
      "auxiliary_loss_mlp": 0.0105804,
      "balance_loss_clip": 1.05440903,
      "balance_loss_mlp": 1.03376901,
      "epoch": 0.09601683451074704,
      "flos": 39603472185600.0,
      "grad_norm": 1.8335529067237837,
      "language_loss": 0.69328064,
      "learning_rate": 3.9544572776429945e-06,
      "loss": 0.71552688,
      "num_input_tokens_seen": 34319990,
      "step": 1597,
      "time_per_iteration": 2.802959442138672
    },
    {
      "auxiliary_loss_clip": 0.01167235,
      "auxiliary_loss_mlp": 0.00780978,
      "balance_loss_clip": 1.0503217,
      "balance_loss_mlp": 1.00010371,
      "epoch": 0.096076957763415,
      "flos": 23732716147200.0,
      "grad_norm": 2.0491570740921885,
      "language_loss": 0.7486403,
      "learning_rate": 3.954374601087729e-06,
      "loss": 0.76812243,
      "num_input_tokens_seen": 34339225,
      "step": 1598,
      "time_per_iteration": 2.6502270698547363
    },
    {
      "auxiliary_loss_clip": 0.01176661,
      "auxiliary_loss_mlp": 0.01053936,
      "balance_loss_clip": 1.05745888,
      "balance_loss_mlp": 1.03009462,
      "epoch": 0.09613708101608297,
      "flos": 34678108483200.0,
      "grad_norm": 1.6831440826618358,
      "language_loss": 0.68804371,
      "learning_rate": 3.954291850422382e-06,
      "loss": 0.71034968,
      "num_input_tokens_seen": 34361020,
      "step": 1599,
      "time_per_iteration": 2.74243426322937
    },
    {
      "auxiliary_loss_clip": 0.01157322,
      "auxiliary_loss_mlp": 0.01059883,
      "balance_loss_clip": 1.05754852,
      "balance_loss_mlp": 1.0371263,
      "epoch": 0.09619720426875093,
      "flos": 20740028653440.0,
      "grad_norm": 2.9774251326108367,
      "language_loss": 0.83950365,
      "learning_rate": 3.954209025650093e-06,
      "loss": 0.86167574,
      "num_input_tokens_seen": 34378630,
      "step": 1600,
      "time_per_iteration": 2.702907085418701
    },
    {
      "auxiliary_loss_clip": 0.01150263,
      "auxiliary_loss_mlp": 0.01054168,
      "balance_loss_clip": 1.05129707,
      "balance_loss_mlp": 1.03093433,
      "epoch": 0.09625732752141891,
      "flos": 13042720488960.0,
      "grad_norm": 2.287254549480118,
      "language_loss": 0.80520785,
      "learning_rate": 3.954126126774001e-06,
      "loss": 0.82725215,
      "num_input_tokens_seen": 34397110,
      "step": 1601,
      "time_per_iteration": 2.693399429321289
    },
    {
      "auxiliary_loss_clip": 0.01181247,
      "auxiliary_loss_mlp": 0.01054578,
      "balance_loss_clip": 1.05711937,
      "balance_loss_mlp": 1.03133249,
      "epoch": 0.09631745077408688,
      "flos": 22273629782400.0,
      "grad_norm": 2.4356926646094954,
      "language_loss": 0.81959623,
      "learning_rate": 3.954043153797251e-06,
      "loss": 0.84195447,
      "num_input_tokens_seen": 34414165,
      "step": 1602,
      "time_per_iteration": 2.639479875564575
    },
    {
      "auxiliary_loss_clip": 0.01137855,
      "auxiliary_loss_mlp": 0.01051495,
      "balance_loss_clip": 1.05295444,
      "balance_loss_mlp": 1.02681863,
      "epoch": 0.09637757402675484,
      "flos": 24754266944640.0,
      "grad_norm": 3.099164686790191,
      "language_loss": 0.62498438,
      "learning_rate": 3.953960106722989e-06,
      "loss": 0.64687788,
      "num_input_tokens_seen": 34434445,
      "step": 1603,
      "time_per_iteration": 4.341834306716919
    },
    {
      "auxiliary_loss_clip": 0.01189954,
      "auxiliary_loss_mlp": 0.01054376,
      "balance_loss_clip": 1.05902839,
      "balance_loss_mlp": 1.02918696,
      "epoch": 0.09643769727942282,
      "flos": 22525758322560.0,
      "grad_norm": 3.121905357886113,
      "language_loss": 0.70996022,
      "learning_rate": 3.953876985554364e-06,
      "loss": 0.73240346,
      "num_input_tokens_seen": 34453095,
      "step": 1604,
      "time_per_iteration": 2.6520893573760986
    },
    {
      "auxiliary_loss_clip": 0.01176446,
      "auxiliary_loss_mlp": 0.01055314,
      "balance_loss_clip": 1.0570209,
      "balance_loss_mlp": 1.03358221,
      "epoch": 0.09649782053209079,
      "flos": 30921026636160.0,
      "grad_norm": 2.082890345500055,
      "language_loss": 0.7993719,
      "learning_rate": 3.953793790294527e-06,
      "loss": 0.82168949,
      "num_input_tokens_seen": 34473680,
      "step": 1605,
      "time_per_iteration": 4.5557661056518555
    },
    {
      "auxiliary_loss_clip": 0.01161047,
      "auxiliary_loss_mlp": 0.01047918,
      "balance_loss_clip": 1.05455577,
      "balance_loss_mlp": 1.0245893,
      "epoch": 0.09655794378475875,
      "flos": 25337635729920.0,
      "grad_norm": 1.990204665194141,
      "language_loss": 0.74550986,
      "learning_rate": 3.953710520946634e-06,
      "loss": 0.76759952,
      "num_input_tokens_seen": 34492610,
      "step": 1606,
      "time_per_iteration": 2.7172651290893555
    },
    {
      "auxiliary_loss_clip": 0.01172416,
      "auxiliary_loss_mlp": 0.01046772,
      "balance_loss_clip": 1.05834222,
      "balance_loss_mlp": 1.02378857,
      "epoch": 0.09661806703742673,
      "flos": 22346061557760.0,
      "grad_norm": 1.6403710807101601,
      "language_loss": 0.7571919,
      "learning_rate": 3.953627177513843e-06,
      "loss": 0.77938372,
      "num_input_tokens_seen": 34511855,
      "step": 1607,
      "time_per_iteration": 4.302686452865601
    },
    {
      "auxiliary_loss_clip": 0.01139491,
      "auxiliary_loss_mlp": 0.01051546,
      "balance_loss_clip": 1.04833579,
      "balance_loss_mlp": 1.0289799,
      "epoch": 0.0966781902900947,
      "flos": 17457578144640.0,
      "grad_norm": 1.975850982703557,
      "language_loss": 0.86756283,
      "learning_rate": 3.953543759999312e-06,
      "loss": 0.88947326,
      "num_input_tokens_seen": 34528905,
      "step": 1608,
      "time_per_iteration": 2.6280455589294434
    },
    {
      "auxiliary_loss_clip": 0.01126253,
      "auxiliary_loss_mlp": 0.01064704,
      "balance_loss_clip": 1.05433142,
      "balance_loss_mlp": 1.03940821,
      "epoch": 0.09673831354276266,
      "flos": 36903995412480.0,
      "grad_norm": 2.3082762386200266,
      "language_loss": 0.71363097,
      "learning_rate": 3.953460268406207e-06,
      "loss": 0.73554057,
      "num_input_tokens_seen": 34548480,
      "step": 1609,
      "time_per_iteration": 2.9116146564483643
    },
    {
      "auxiliary_loss_clip": 0.01149353,
      "auxiliary_loss_mlp": 0.01058179,
      "balance_loss_clip": 1.0546515,
      "balance_loss_mlp": 1.03606534,
      "epoch": 0.09679843679543064,
      "flos": 20701388597760.0,
      "grad_norm": 1.9988414994799784,
      "language_loss": 0.84810984,
      "learning_rate": 3.953376702737693e-06,
      "loss": 0.87018514,
      "num_input_tokens_seen": 34565410,
      "step": 1610,
      "time_per_iteration": 2.8005051612854004
    },
    {
      "auxiliary_loss_clip": 0.01161389,
      "auxiliary_loss_mlp": 0.01056267,
      "balance_loss_clip": 1.05790925,
      "balance_loss_mlp": 1.03228188,
      "epoch": 0.0968585600480986,
      "flos": 23514415240320.0,
      "grad_norm": 2.176236379770122,
      "language_loss": 0.6696198,
      "learning_rate": 3.953293062996939e-06,
      "loss": 0.69179636,
      "num_input_tokens_seen": 34584840,
      "step": 1611,
      "time_per_iteration": 2.731931447982788
    },
    {
      "auxiliary_loss_clip": 0.01125259,
      "auxiliary_loss_mlp": 0.01057116,
      "balance_loss_clip": 1.04740572,
      "balance_loss_mlp": 1.03385806,
      "epoch": 0.09691868330076657,
      "flos": 20121072468480.0,
      "grad_norm": 1.6508278294088392,
      "language_loss": 0.81067657,
      "learning_rate": 3.953209349187115e-06,
      "loss": 0.83250034,
      "num_input_tokens_seen": 34603360,
      "step": 1612,
      "time_per_iteration": 2.7998390197753906
    },
    {
      "auxiliary_loss_clip": 0.01182404,
      "auxiliary_loss_mlp": 0.01069551,
      "balance_loss_clip": 1.06046534,
      "balance_loss_mlp": 1.04600716,
      "epoch": 0.09697880655343454,
      "flos": 16544692967040.0,
      "grad_norm": 3.304939197664143,
      "language_loss": 0.80836105,
      "learning_rate": 3.953125561311398e-06,
      "loss": 0.83088064,
      "num_input_tokens_seen": 34620760,
      "step": 1613,
      "time_per_iteration": 2.624218702316284
    },
    {
      "auxiliary_loss_clip": 0.01148565,
      "auxiliary_loss_mlp": 0.01054743,
      "balance_loss_clip": 1.05542159,
      "balance_loss_mlp": 1.03047192,
      "epoch": 0.09703892980610251,
      "flos": 26104184899200.0,
      "grad_norm": 1.7164386274315457,
      "language_loss": 0.84289789,
      "learning_rate": 3.953041699372964e-06,
      "loss": 0.86493099,
      "num_input_tokens_seen": 34640695,
      "step": 1614,
      "time_per_iteration": 2.744340419769287
    },
    {
      "auxiliary_loss_clip": 0.01066618,
      "auxiliary_loss_mlp": 0.00759744,
      "balance_loss_clip": 1.02654934,
      "balance_loss_mlp": 1.00008702,
      "epoch": 0.09709905305877048,
      "flos": 60443622000000.0,
      "grad_norm": 0.7127167896900892,
      "language_loss": 0.54629624,
      "learning_rate": 3.952957763374992e-06,
      "loss": 0.56455994,
      "num_input_tokens_seen": 34702395,
      "step": 1615,
      "time_per_iteration": 3.1547679901123047
    },
    {
      "auxiliary_loss_clip": 0.01033143,
      "auxiliary_loss_mlp": 0.01017555,
      "balance_loss_clip": 1.02384067,
      "balance_loss_mlp": 1.01381195,
      "epoch": 0.09715917631143844,
      "flos": 57639932893440.0,
      "grad_norm": 0.7689373847786285,
      "language_loss": 0.58190405,
      "learning_rate": 3.952873753320666e-06,
      "loss": 0.60241103,
      "num_input_tokens_seen": 34768910,
      "step": 1616,
      "time_per_iteration": 3.3940556049346924
    },
    {
      "auxiliary_loss_clip": 0.01155533,
      "auxiliary_loss_mlp": 0.01067983,
      "balance_loss_clip": 1.05504358,
      "balance_loss_mlp": 1.04205465,
      "epoch": 0.09721929956410642,
      "flos": 20558212986240.0,
      "grad_norm": 1.8932449927934136,
      "language_loss": 0.69031835,
      "learning_rate": 3.952789669213172e-06,
      "loss": 0.7125535,
      "num_input_tokens_seen": 34787680,
      "step": 1617,
      "time_per_iteration": 2.714629888534546
    },
    {
      "auxiliary_loss_clip": 0.01152637,
      "auxiliary_loss_mlp": 0.01057882,
      "balance_loss_clip": 1.05386162,
      "balance_loss_mlp": 1.03127456,
      "epoch": 0.09727942281677439,
      "flos": 27344359825920.0,
      "grad_norm": 1.755493071880773,
      "language_loss": 0.80910909,
      "learning_rate": 3.952705511055698e-06,
      "loss": 0.83121431,
      "num_input_tokens_seen": 34808330,
      "step": 1618,
      "time_per_iteration": 2.8081507682800293
    },
    {
      "auxiliary_loss_clip": 0.01168356,
      "auxiliary_loss_mlp": 0.01058179,
      "balance_loss_clip": 1.06048679,
      "balance_loss_mlp": 1.03678131,
      "epoch": 0.09733954606944235,
      "flos": 24900028335360.0,
      "grad_norm": 1.667659488760432,
      "language_loss": 0.92901695,
      "learning_rate": 3.952621278851435e-06,
      "loss": 0.95128226,
      "num_input_tokens_seen": 34830020,
      "step": 1619,
      "time_per_iteration": 2.7752275466918945
    },
    {
      "auxiliary_loss_clip": 0.01175515,
      "auxiliary_loss_mlp": 0.01058252,
      "balance_loss_clip": 1.05952573,
      "balance_loss_mlp": 1.03512526,
      "epoch": 0.09739966932211033,
      "flos": 31503928544640.0,
      "grad_norm": 2.1973967195348902,
      "language_loss": 0.88978708,
      "learning_rate": 3.9525369726035784e-06,
      "loss": 0.91212475,
      "num_input_tokens_seen": 34850330,
      "step": 1620,
      "time_per_iteration": 2.771176338195801
    },
    {
      "auxiliary_loss_clip": 0.01153763,
      "auxiliary_loss_mlp": 0.01065329,
      "balance_loss_clip": 1.05353975,
      "balance_loss_mlp": 1.0397464,
      "epoch": 0.0974597925747783,
      "flos": 23878764846720.0,
      "grad_norm": 2.154793183838835,
      "language_loss": 0.77331412,
      "learning_rate": 3.952452592315324e-06,
      "loss": 0.79550499,
      "num_input_tokens_seen": 34871640,
      "step": 1621,
      "time_per_iteration": 2.6740832328796387
    },
    {
      "auxiliary_loss_clip": 0.01131342,
      "auxiliary_loss_mlp": 0.01082359,
      "balance_loss_clip": 1.04798269,
      "balance_loss_mlp": 1.05640674,
      "epoch": 0.09751991582744626,
      "flos": 17019575700480.0,
      "grad_norm": 1.9420195171733425,
      "language_loss": 0.77671158,
      "learning_rate": 3.952368137989871e-06,
      "loss": 0.79884863,
      "num_input_tokens_seen": 34888100,
      "step": 1622,
      "time_per_iteration": 2.7247347831726074
    },
    {
      "auxiliary_loss_clip": 0.01150185,
      "auxiliary_loss_mlp": 0.01064277,
      "balance_loss_clip": 1.05335355,
      "balance_loss_mlp": 1.04025626,
      "epoch": 0.09758003908011423,
      "flos": 28402826826240.0,
      "grad_norm": 1.8603109065807166,
      "language_loss": 0.85784447,
      "learning_rate": 3.9522836096304225e-06,
      "loss": 0.87998909,
      "num_input_tokens_seen": 34910485,
      "step": 1623,
      "time_per_iteration": 2.785388469696045
    },
    {
      "auxiliary_loss_clip": 0.0117659,
      "auxiliary_loss_mlp": 0.01064102,
      "balance_loss_clip": 1.05769634,
      "balance_loss_mlp": 1.04043913,
      "epoch": 0.09764016233278221,
      "flos": 18144297336960.0,
      "grad_norm": 2.39630116599036,
      "language_loss": 0.80534065,
      "learning_rate": 3.952199007240184e-06,
      "loss": 0.82774758,
      "num_input_tokens_seen": 34928615,
      "step": 1624,
      "time_per_iteration": 2.6818184852600098
    },
    {
      "auxiliary_loss_clip": 0.01176335,
      "auxiliary_loss_mlp": 0.01056788,
      "balance_loss_clip": 1.05616927,
      "balance_loss_mlp": 1.03465128,
      "epoch": 0.09770028558545017,
      "flos": 15265842071040.0,
      "grad_norm": 2.44379144971104,
      "language_loss": 0.85556966,
      "learning_rate": 3.952114330822364e-06,
      "loss": 0.8779009,
      "num_input_tokens_seen": 34946045,
      "step": 1625,
      "time_per_iteration": 2.6594324111938477
    },
    {
      "auxiliary_loss_clip": 0.01181411,
      "auxiliary_loss_mlp": 0.0106682,
      "balance_loss_clip": 1.06004012,
      "balance_loss_mlp": 1.04411101,
      "epoch": 0.09776040883811814,
      "flos": 23472435219840.0,
      "grad_norm": 2.058269503362464,
      "language_loss": 0.85431635,
      "learning_rate": 3.952029580380172e-06,
      "loss": 0.87679869,
      "num_input_tokens_seen": 34962865,
      "step": 1626,
      "time_per_iteration": 2.7384841442108154
    },
    {
      "auxiliary_loss_clip": 0.01165311,
      "auxiliary_loss_mlp": 0.007823,
      "balance_loss_clip": 1.05467701,
      "balance_loss_mlp": 1.000211,
      "epoch": 0.09782053209078612,
      "flos": 24499480798080.0,
      "grad_norm": 2.0701580273163036,
      "language_loss": 0.83370024,
      "learning_rate": 3.9519447559168234e-06,
      "loss": 0.85317636,
      "num_input_tokens_seen": 34983505,
      "step": 1627,
      "time_per_iteration": 2.8269948959350586
    },
    {
      "auxiliary_loss_clip": 0.01168188,
      "auxiliary_loss_mlp": 0.01065332,
      "balance_loss_clip": 1.05557203,
      "balance_loss_mlp": 1.04275417,
      "epoch": 0.09788065534345408,
      "flos": 21580158833280.0,
      "grad_norm": 1.8143281262319713,
      "language_loss": 0.84674478,
      "learning_rate": 3.951859857435534e-06,
      "loss": 0.86907995,
      "num_input_tokens_seen": 35001825,
      "step": 1628,
      "time_per_iteration": 2.6151821613311768
    },
    {
      "auxiliary_loss_clip": 0.01170257,
      "auxiliary_loss_mlp": 0.01058367,
      "balance_loss_clip": 1.05374515,
      "balance_loss_mlp": 1.03558636,
      "epoch": 0.09794077859612205,
      "flos": 23842459175040.0,
      "grad_norm": 1.5658807312485334,
      "language_loss": 0.75531614,
      "learning_rate": 3.951774884939523e-06,
      "loss": 0.77760237,
      "num_input_tokens_seen": 35023075,
      "step": 1629,
      "time_per_iteration": 2.6794557571411133
    },
    {
      "auxiliary_loss_clip": 0.01129604,
      "auxiliary_loss_mlp": 0.01056904,
      "balance_loss_clip": 1.0577755,
      "balance_loss_mlp": 1.03169131,
      "epoch": 0.09800090184879003,
      "flos": 23659889322240.0,
      "grad_norm": 1.6755762488260617,
      "language_loss": 0.78487194,
      "learning_rate": 3.951689838432013e-06,
      "loss": 0.80673707,
      "num_input_tokens_seen": 35043480,
      "step": 1630,
      "time_per_iteration": 2.7986228466033936
    },
    {
      "auxiliary_loss_clip": 0.01167766,
      "auxiliary_loss_mlp": 0.01063441,
      "balance_loss_clip": 1.05938148,
      "balance_loss_mlp": 1.03804946,
      "epoch": 0.09806102510145799,
      "flos": 17055773631360.0,
      "grad_norm": 1.8175370389297836,
      "language_loss": 0.86677933,
      "learning_rate": 3.951604717916228e-06,
      "loss": 0.88909143,
      "num_input_tokens_seen": 35061490,
      "step": 1631,
      "time_per_iteration": 2.6350157260894775
    },
    {
      "auxiliary_loss_clip": 0.01171369,
      "auxiliary_loss_mlp": 0.01058643,
      "balance_loss_clip": 1.0610745,
      "balance_loss_mlp": 1.03625536,
      "epoch": 0.09812114835412596,
      "flos": 23878477537920.0,
      "grad_norm": 2.2030333753544773,
      "language_loss": 0.82996809,
      "learning_rate": 3.9515195233953975e-06,
      "loss": 0.85226822,
      "num_input_tokens_seen": 35079670,
      "step": 1632,
      "time_per_iteration": 2.7990314960479736
    },
    {
      "auxiliary_loss_clip": 0.01148453,
      "auxiliary_loss_mlp": 0.01064004,
      "balance_loss_clip": 1.05554819,
      "balance_loss_mlp": 1.04102039,
      "epoch": 0.09818127160679392,
      "flos": 20595488325120.0,
      "grad_norm": 1.531777801288569,
      "language_loss": 0.7882973,
      "learning_rate": 3.951434254872751e-06,
      "loss": 0.81042188,
      "num_input_tokens_seen": 35099205,
      "step": 1633,
      "time_per_iteration": 2.735353708267212
    },
    {
      "auxiliary_loss_clip": 0.01170992,
      "auxiliary_loss_mlp": 0.01061681,
      "balance_loss_clip": 1.05558002,
      "balance_loss_mlp": 1.03731489,
      "epoch": 0.0982413948594619,
      "flos": 15487339288320.0,
      "grad_norm": 2.4037572513069687,
      "language_loss": 0.73209554,
      "learning_rate": 3.951348912351521e-06,
      "loss": 0.75442231,
      "num_input_tokens_seen": 35115270,
      "step": 1634,
      "time_per_iteration": 2.688596248626709
    },
    {
      "auxiliary_loss_clip": 0.01162743,
      "auxiliary_loss_mlp": 0.01071164,
      "balance_loss_clip": 1.05591321,
      "balance_loss_mlp": 1.04672611,
      "epoch": 0.09830151811212987,
      "flos": 24207958016640.0,
      "grad_norm": 3.2244021303311405,
      "language_loss": 0.72553629,
      "learning_rate": 3.951263495834947e-06,
      "loss": 0.74787533,
      "num_input_tokens_seen": 35134065,
      "step": 1635,
      "time_per_iteration": 2.720266342163086
    },
    {
      "auxiliary_loss_clip": 0.01154765,
      "auxiliary_loss_mlp": 0.01068349,
      "balance_loss_clip": 1.05526268,
      "balance_loss_mlp": 1.04177701,
      "epoch": 0.09836164136479783,
      "flos": 20594590485120.0,
      "grad_norm": 1.7699592352066487,
      "language_loss": 0.78026646,
      "learning_rate": 3.951178005326264e-06,
      "loss": 0.80249763,
      "num_input_tokens_seen": 35154870,
      "step": 1636,
      "time_per_iteration": 2.9618239402770996
    },
    {
      "auxiliary_loss_clip": 0.01162744,
      "auxiliary_loss_mlp": 0.01060716,
      "balance_loss_clip": 1.05561686,
      "balance_loss_mlp": 1.0368979,
      "epoch": 0.09842176461746581,
      "flos": 19934157070080.0,
      "grad_norm": 1.8332710343018006,
      "language_loss": 0.69524407,
      "learning_rate": 3.951092440828715e-06,
      "loss": 0.71747863,
      "num_input_tokens_seen": 35171850,
      "step": 1637,
      "time_per_iteration": 2.671178102493286
    },
    {
      "auxiliary_loss_clip": 0.01188316,
      "auxiliary_loss_mlp": 0.01058851,
      "balance_loss_clip": 1.05926394,
      "balance_loss_mlp": 1.03500926,
      "epoch": 0.09848188787013377,
      "flos": 21214659991680.0,
      "grad_norm": 2.3775286970935503,
      "language_loss": 0.77050996,
      "learning_rate": 3.951006802345545e-06,
      "loss": 0.79298162,
      "num_input_tokens_seen": 35188795,
      "step": 1638,
      "time_per_iteration": 2.62457537651062
    },
    {
      "auxiliary_loss_clip": 0.01140265,
      "auxiliary_loss_mlp": 0.01052026,
      "balance_loss_clip": 1.05538166,
      "balance_loss_mlp": 1.02941203,
      "epoch": 0.09854201112280174,
      "flos": 30154226071680.0,
      "grad_norm": 1.4014263071342075,
      "language_loss": 0.72620296,
      "learning_rate": 3.950921089880003e-06,
      "loss": 0.74812591,
      "num_input_tokens_seen": 35212100,
      "step": 1639,
      "time_per_iteration": 2.7499618530273438
    },
    {
      "auxiliary_loss_clip": 0.01173752,
      "auxiliary_loss_mlp": 0.01051382,
      "balance_loss_clip": 1.0582087,
      "balance_loss_mlp": 1.02831531,
      "epoch": 0.09860213437546972,
      "flos": 21795730306560.0,
      "grad_norm": 1.7213189449892274,
      "language_loss": 0.88679075,
      "learning_rate": 3.950835303435337e-06,
      "loss": 0.90904212,
      "num_input_tokens_seen": 35230390,
      "step": 1640,
      "time_per_iteration": 2.664133071899414
    },
    {
      "auxiliary_loss_clip": 0.01177786,
      "auxiliary_loss_mlp": 0.01044457,
      "balance_loss_clip": 1.05981517,
      "balance_loss_mlp": 1.02130616,
      "epoch": 0.09866225762813768,
      "flos": 21835555511040.0,
      "grad_norm": 2.0701766566296915,
      "language_loss": 0.80567038,
      "learning_rate": 3.950749443014801e-06,
      "loss": 0.82789278,
      "num_input_tokens_seen": 35250405,
      "step": 1641,
      "time_per_iteration": 2.645353317260742
    },
    {
      "auxiliary_loss_clip": 0.011756,
      "auxiliary_loss_mlp": 0.01062641,
      "balance_loss_clip": 1.05896795,
      "balance_loss_mlp": 1.03742838,
      "epoch": 0.09872238088080565,
      "flos": 17599855916160.0,
      "grad_norm": 2.64335263522248,
      "language_loss": 0.86117625,
      "learning_rate": 3.95066350862165e-06,
      "loss": 0.88355863,
      "num_input_tokens_seen": 35262820,
      "step": 1642,
      "time_per_iteration": 5.81004524230957
    },
    {
      "auxiliary_loss_clip": 0.01151329,
      "auxiliary_loss_mlp": 0.01056693,
      "balance_loss_clip": 1.05857074,
      "balance_loss_mlp": 1.03404331,
      "epoch": 0.09878250413347361,
      "flos": 27636134002560.0,
      "grad_norm": 2.7092208079201607,
      "language_loss": 0.8058275,
      "learning_rate": 3.950577500259144e-06,
      "loss": 0.82790768,
      "num_input_tokens_seen": 35284490,
      "step": 1643,
      "time_per_iteration": 2.7235090732574463
    },
    {
      "auxiliary_loss_clip": 0.01174075,
      "auxiliary_loss_mlp": 0.01077435,
      "balance_loss_clip": 1.05761337,
      "balance_loss_mlp": 1.05470192,
      "epoch": 0.0988426273861416,
      "flos": 16544728880640.0,
      "grad_norm": 2.0561742686210676,
      "language_loss": 0.82546467,
      "learning_rate": 3.950491417930543e-06,
      "loss": 0.84797978,
      "num_input_tokens_seen": 35302815,
      "step": 1644,
      "time_per_iteration": 4.318823575973511
    },
    {
      "auxiliary_loss_clip": 0.01163142,
      "auxiliary_loss_mlp": 0.00782463,
      "balance_loss_clip": 1.05607629,
      "balance_loss_mlp": 1.00010633,
      "epoch": 0.09890275063880956,
      "flos": 21215270522880.0,
      "grad_norm": 1.6945489721625269,
      "language_loss": 0.68219113,
      "learning_rate": 3.9504052616391124e-06,
      "loss": 0.70164716,
      "num_input_tokens_seen": 35321175,
      "step": 1645,
      "time_per_iteration": 2.6626670360565186
    },
    {
      "auxiliary_loss_clip": 0.01059795,
      "auxiliary_loss_mlp": 0.01047617,
      "balance_loss_clip": 1.02852345,
      "balance_loss_mlp": 1.04404068,
      "epoch": 0.09896287389147752,
      "flos": 59379372910080.0,
      "grad_norm": 0.8512889940087613,
      "language_loss": 0.60885167,
      "learning_rate": 3.950319031388119e-06,
      "loss": 0.62992585,
      "num_input_tokens_seen": 35381740,
      "step": 1646,
      "time_per_iteration": 4.752669095993042
    },
    {
      "auxiliary_loss_clip": 0.01147006,
      "auxiliary_loss_mlp": 0.0105976,
      "balance_loss_clip": 1.0574733,
      "balance_loss_mlp": 1.03464222,
      "epoch": 0.0990229971441455,
      "flos": 29642678530560.0,
      "grad_norm": 5.785751121573768,
      "language_loss": 0.73211443,
      "learning_rate": 3.950232727180833e-06,
      "loss": 0.7541821,
      "num_input_tokens_seen": 35403760,
      "step": 1647,
      "time_per_iteration": 2.783442974090576
    },
    {
      "auxiliary_loss_clip": 0.01161789,
      "auxiliary_loss_mlp": 0.01066314,
      "balance_loss_clip": 1.06016421,
      "balance_loss_mlp": 1.04445136,
      "epoch": 0.09908312039681347,
      "flos": 21834873152640.0,
      "grad_norm": 1.828428298130997,
      "language_loss": 0.84094375,
      "learning_rate": 3.950146349020525e-06,
      "loss": 0.86322474,
      "num_input_tokens_seen": 35424050,
      "step": 1648,
      "time_per_iteration": 2.709559679031372
    },
    {
      "auxiliary_loss_clip": 0.01065954,
      "auxiliary_loss_mlp": 0.01020799,
      "balance_loss_clip": 1.02565169,
      "balance_loss_mlp": 1.01722264,
      "epoch": 0.09914324364948143,
      "flos": 57564304807680.0,
      "grad_norm": 0.7317434537206132,
      "language_loss": 0.55672908,
      "learning_rate": 3.950059896910473e-06,
      "loss": 0.5775966,
      "num_input_tokens_seen": 35481690,
      "step": 1649,
      "time_per_iteration": 3.0944156646728516
    },
    {
      "auxiliary_loss_clip": 0.0117133,
      "auxiliary_loss_mlp": 0.01049543,
      "balance_loss_clip": 1.05603158,
      "balance_loss_mlp": 1.02723897,
      "epoch": 0.09920336690214941,
      "flos": 34123934476800.0,
      "grad_norm": 2.195431109372502,
      "language_loss": 0.8975327,
      "learning_rate": 3.949973370853954e-06,
      "loss": 0.91974139,
      "num_input_tokens_seen": 35498635,
      "step": 1650,
      "time_per_iteration": 2.7438554763793945
    },
    {
      "auxiliary_loss_clip": 0.01033978,
      "auxiliary_loss_mlp": 0.00758727,
      "balance_loss_clip": 1.02943921,
      "balance_loss_mlp": 0.9997822,
      "epoch": 0.09926349015481738,
      "flos": 71216428464000.0,
      "grad_norm": 0.8036050505402587,
      "language_loss": 0.63734978,
      "learning_rate": 3.94988677085425e-06,
      "loss": 0.65527683,
      "num_input_tokens_seen": 35565720,
      "step": 1651,
      "time_per_iteration": 3.40269136428833
    },
    {
      "auxiliary_loss_clip": 0.01170347,
      "auxiliary_loss_mlp": 0.01062486,
      "balance_loss_clip": 1.05790281,
      "balance_loss_mlp": 1.03842974,
      "epoch": 0.09932361340748534,
      "flos": 23148700917120.0,
      "grad_norm": 1.9744130417114842,
      "language_loss": 0.88115525,
      "learning_rate": 3.949800096914643e-06,
      "loss": 0.90348363,
      "num_input_tokens_seen": 35586000,
      "step": 1652,
      "time_per_iteration": 2.6695117950439453
    },
    {
      "auxiliary_loss_clip": 0.0116773,
      "auxiliary_loss_mlp": 0.01062073,
      "balance_loss_clip": 1.06095552,
      "balance_loss_mlp": 1.03895831,
      "epoch": 0.09938373666015332,
      "flos": 19828651847040.0,
      "grad_norm": 2.166773052437996,
      "language_loss": 0.81789082,
      "learning_rate": 3.949713349038422e-06,
      "loss": 0.84018886,
      "num_input_tokens_seen": 35604355,
      "step": 1653,
      "time_per_iteration": 2.7136831283569336
    },
    {
      "auxiliary_loss_clip": 0.01173152,
      "auxiliary_loss_mlp": 0.00780466,
      "balance_loss_clip": 1.05683279,
      "balance_loss_mlp": 1.00016594,
      "epoch": 0.09944385991282129,
      "flos": 22090664880000.0,
      "grad_norm": 1.662037391605293,
      "language_loss": 0.79489207,
      "learning_rate": 3.949626527228875e-06,
      "loss": 0.81442821,
      "num_input_tokens_seen": 35625495,
      "step": 1654,
      "time_per_iteration": 2.645875930786133
    },
    {
      "auxiliary_loss_clip": 0.01187918,
      "auxiliary_loss_mlp": 0.01056849,
      "balance_loss_clip": 1.06405056,
      "balance_loss_mlp": 1.03561759,
      "epoch": 0.09950398316548925,
      "flos": 19828867328640.0,
      "grad_norm": 1.7263610037420916,
      "language_loss": 0.81038272,
      "learning_rate": 3.949539631489295e-06,
      "loss": 0.83283037,
      "num_input_tokens_seen": 35645030,
      "step": 1655,
      "time_per_iteration": 2.630404233932495
    },
    {
      "auxiliary_loss_clip": 0.01181205,
      "auxiliary_loss_mlp": 0.01055977,
      "balance_loss_clip": 1.05679035,
      "balance_loss_mlp": 1.03294599,
      "epoch": 0.09956410641815722,
      "flos": 25003701964800.0,
      "grad_norm": 2.426795421082641,
      "language_loss": 0.80429518,
      "learning_rate": 3.9494526618229765e-06,
      "loss": 0.82666701,
      "num_input_tokens_seen": 35664305,
      "step": 1656,
      "time_per_iteration": 2.6283950805664062
    },
    {
      "auxiliary_loss_clip": 0.01170003,
      "auxiliary_loss_mlp": 0.01061881,
      "balance_loss_clip": 1.05787742,
      "balance_loss_mlp": 1.03870714,
      "epoch": 0.0996242296708252,
      "flos": 19317714837120.0,
      "grad_norm": 1.4960238412267362,
      "language_loss": 0.89040691,
      "learning_rate": 3.949365618233217e-06,
      "loss": 0.91272575,
      "num_input_tokens_seen": 35684060,
      "step": 1657,
      "time_per_iteration": 2.653674602508545
    },
    {
      "auxiliary_loss_clip": 0.01165842,
      "auxiliary_loss_mlp": 0.01057352,
      "balance_loss_clip": 1.05830753,
      "balance_loss_mlp": 1.0329144,
      "epoch": 0.09968435292349316,
      "flos": 21871609787520.0,
      "grad_norm": 2.1866084372248062,
      "language_loss": 0.84684521,
      "learning_rate": 3.9492785007233195e-06,
      "loss": 0.86907715,
      "num_input_tokens_seen": 35703250,
      "step": 1658,
      "time_per_iteration": 2.6897473335266113
    },
    {
      "auxiliary_loss_clip": 0.01069806,
      "auxiliary_loss_mlp": 0.01015844,
      "balance_loss_clip": 1.02042234,
      "balance_loss_mlp": 1.01292348,
      "epoch": 0.09974447617616113,
      "flos": 65384533313280.0,
      "grad_norm": 0.9123227767672076,
      "language_loss": 0.60828507,
      "learning_rate": 3.949191309296585e-06,
      "loss": 0.62914157,
      "num_input_tokens_seen": 35762165,
      "step": 1659,
      "time_per_iteration": 3.273890495300293
    },
    {
      "auxiliary_loss_clip": 0.01152432,
      "auxiliary_loss_mlp": 0.01051829,
      "balance_loss_clip": 1.05082798,
      "balance_loss_mlp": 1.02814245,
      "epoch": 0.0998045994288291,
      "flos": 23659817495040.0,
      "grad_norm": 1.9344290476513741,
      "language_loss": 0.84892076,
      "learning_rate": 3.949104043956321e-06,
      "loss": 0.87096334,
      "num_input_tokens_seen": 35781520,
      "step": 1660,
      "time_per_iteration": 2.788018226623535
    },
    {
      "auxiliary_loss_clip": 0.01149163,
      "auxiliary_loss_mlp": 0.01060092,
      "balance_loss_clip": 1.05374026,
      "balance_loss_mlp": 1.03514171,
      "epoch": 0.09986472268149707,
      "flos": 19609704495360.0,
      "grad_norm": 1.9493882663610318,
      "language_loss": 0.80024737,
      "learning_rate": 3.949016704705836e-06,
      "loss": 0.82234001,
      "num_input_tokens_seen": 35799565,
      "step": 1661,
      "time_per_iteration": 2.6537399291992188
    },
    {
      "auxiliary_loss_clip": 0.01172787,
      "auxiliary_loss_mlp": 0.01055532,
      "balance_loss_clip": 1.05715156,
      "balance_loss_mlp": 1.03153503,
      "epoch": 0.09992484593416504,
      "flos": 26213317395840.0,
      "grad_norm": 2.0152235709188377,
      "language_loss": 0.83560598,
      "learning_rate": 3.948929291548443e-06,
      "loss": 0.85788912,
      "num_input_tokens_seen": 35821085,
      "step": 1662,
      "time_per_iteration": 2.753807783126831
    },
    {
      "auxiliary_loss_clip": 0.01154838,
      "auxiliary_loss_mlp": 0.01061466,
      "balance_loss_clip": 1.05079484,
      "balance_loss_mlp": 1.03616929,
      "epoch": 0.09998496918683301,
      "flos": 17493632421120.0,
      "grad_norm": 1.9355779644050557,
      "language_loss": 0.88865256,
      "learning_rate": 3.9488418044874546e-06,
      "loss": 0.91081554,
      "num_input_tokens_seen": 35839840,
      "step": 1663,
      "time_per_iteration": 2.6829047203063965
    },
    {
      "auxiliary_loss_clip": 0.0118246,
      "auxiliary_loss_mlp": 0.01061692,
      "balance_loss_clip": 1.06228638,
      "balance_loss_mlp": 1.03825521,
      "epoch": 0.10004509243950098,
      "flos": 22784925928320.0,
      "grad_norm": 1.7925330820671084,
      "language_loss": 0.70140731,
      "learning_rate": 3.948754243526191e-06,
      "loss": 0.72384882,
      "num_input_tokens_seen": 35861545,
      "step": 1664,
      "time_per_iteration": 2.809300184249878
    },
    {
      "auxiliary_loss_clip": 0.01142878,
      "auxiliary_loss_mlp": 0.01055306,
      "balance_loss_clip": 1.05475903,
      "balance_loss_mlp": 1.03312087,
      "epoch": 0.10010521569216894,
      "flos": 16253385667200.0,
      "grad_norm": 2.4978474602303895,
      "language_loss": 0.78981555,
      "learning_rate": 3.94866660866797e-06,
      "loss": 0.81179744,
      "num_input_tokens_seen": 35878295,
      "step": 1665,
      "time_per_iteration": 2.7010488510131836
    },
    {
      "auxiliary_loss_clip": 0.01175861,
      "auxiliary_loss_mlp": 0.01070341,
      "balance_loss_clip": 1.06286561,
      "balance_loss_mlp": 1.04742861,
      "epoch": 0.10016533894483691,
      "flos": 23402589223680.0,
      "grad_norm": 3.1438625724360945,
      "language_loss": 0.70054829,
      "learning_rate": 3.9485788999161165e-06,
      "loss": 0.7230103,
      "num_input_tokens_seen": 35898990,
      "step": 1666,
      "time_per_iteration": 2.689879894256592
    },
    {
      "auxiliary_loss_clip": 0.01110848,
      "auxiliary_loss_mlp": 0.01074593,
      "balance_loss_clip": 1.05082703,
      "balance_loss_mlp": 1.04946339,
      "epoch": 0.10022546219750489,
      "flos": 19354164163200.0,
      "grad_norm": 1.7583449522195267,
      "language_loss": 0.78647351,
      "learning_rate": 3.948491117273956e-06,
      "loss": 0.80832791,
      "num_input_tokens_seen": 35916225,
      "step": 1667,
      "time_per_iteration": 2.8973352909088135
    },
    {
      "auxiliary_loss_clip": 0.01153352,
      "auxiliary_loss_mlp": 0.01062819,
      "balance_loss_clip": 1.05452693,
      "balance_loss_mlp": 1.03752255,
      "epoch": 0.10028558545017285,
      "flos": 27085766837760.0,
      "grad_norm": 2.4011089045072187,
      "language_loss": 0.77357388,
      "learning_rate": 3.948403260744817e-06,
      "loss": 0.7957356,
      "num_input_tokens_seen": 35934630,
      "step": 1668,
      "time_per_iteration": 3.2600321769714355
    },
    {
      "auxiliary_loss_clip": 0.01184879,
      "auxiliary_loss_mlp": 0.01059367,
      "balance_loss_clip": 1.05833495,
      "balance_loss_mlp": 1.03523922,
      "epoch": 0.10034570870284082,
      "flos": 25847136195840.0,
      "grad_norm": 1.7407668002390366,
      "language_loss": 0.77520061,
      "learning_rate": 3.948315330332031e-06,
      "loss": 0.79764307,
      "num_input_tokens_seen": 35953855,
      "step": 1669,
      "time_per_iteration": 2.6899471282958984
    },
    {
      "auxiliary_loss_clip": 0.0118887,
      "auxiliary_loss_mlp": 0.01067842,
      "balance_loss_clip": 1.05948365,
      "balance_loss_mlp": 1.04416728,
      "epoch": 0.1004058319555088,
      "flos": 26249587153920.0,
      "grad_norm": 5.441134829238958,
      "language_loss": 0.85160148,
      "learning_rate": 3.948227326038933e-06,
      "loss": 0.87416857,
      "num_input_tokens_seen": 35974555,
      "step": 1670,
      "time_per_iteration": 2.616867780685425
    },
    {
      "auxiliary_loss_clip": 0.011763,
      "auxiliary_loss_mlp": 0.01055607,
      "balance_loss_clip": 1.05584121,
      "balance_loss_mlp": 1.03354108,
      "epoch": 0.10046595520817676,
      "flos": 25374480105600.0,
      "grad_norm": 1.4849262119454174,
      "language_loss": 0.76836258,
      "learning_rate": 3.9481392478688586e-06,
      "loss": 0.79068166,
      "num_input_tokens_seen": 35996830,
      "step": 1671,
      "time_per_iteration": 2.658254384994507
    },
    {
      "auxiliary_loss_clip": 0.01061447,
      "auxiliary_loss_mlp": 0.01017561,
      "balance_loss_clip": 1.02178144,
      "balance_loss_mlp": 1.01454473,
      "epoch": 0.10052607846084473,
      "flos": 67461821677440.0,
      "grad_norm": 0.7781454358921105,
      "language_loss": 0.60718858,
      "learning_rate": 3.948051095825149e-06,
      "loss": 0.62797856,
      "num_input_tokens_seen": 36054465,
      "step": 1672,
      "time_per_iteration": 3.1269097328186035
    },
    {
      "auxiliary_loss_clip": 0.01143177,
      "auxiliary_loss_mlp": 0.01063346,
      "balance_loss_clip": 1.05112922,
      "balance_loss_mlp": 1.04055333,
      "epoch": 0.10058620171351271,
      "flos": 21360493209600.0,
      "grad_norm": 2.433278134910662,
      "language_loss": 0.7711426,
      "learning_rate": 3.947962869911147e-06,
      "loss": 0.79320776,
      "num_input_tokens_seen": 36073480,
      "step": 1673,
      "time_per_iteration": 2.6931638717651367
    },
    {
      "auxiliary_loss_clip": 0.01132094,
      "auxiliary_loss_mlp": 0.01056611,
      "balance_loss_clip": 1.04989302,
      "balance_loss_mlp": 1.03262639,
      "epoch": 0.10064632496618067,
      "flos": 16800125558400.0,
      "grad_norm": 2.074683072839241,
      "language_loss": 0.73173523,
      "learning_rate": 3.947874570130197e-06,
      "loss": 0.75362229,
      "num_input_tokens_seen": 36091830,
      "step": 1674,
      "time_per_iteration": 2.7188127040863037
    },
    {
      "auxiliary_loss_clip": 0.01172389,
      "auxiliary_loss_mlp": 0.00779533,
      "balance_loss_clip": 1.0556165,
      "balance_loss_mlp": 1.00024796,
      "epoch": 0.10070644821884864,
      "flos": 23624445576960.0,
      "grad_norm": 2.1982379565146872,
      "language_loss": 0.79456973,
      "learning_rate": 3.947786196485649e-06,
      "loss": 0.81408894,
      "num_input_tokens_seen": 36111400,
      "step": 1675,
      "time_per_iteration": 2.712090253829956
    },
    {
      "auxiliary_loss_clip": 0.01182659,
      "auxiliary_loss_mlp": 0.01063327,
      "balance_loss_clip": 1.05801332,
      "balance_loss_mlp": 1.04239404,
      "epoch": 0.1007665714715166,
      "flos": 24462564595200.0,
      "grad_norm": 2.408955682155161,
      "language_loss": 0.8120935,
      "learning_rate": 3.947697748980853e-06,
      "loss": 0.83455336,
      "num_input_tokens_seen": 36129345,
      "step": 1676,
      "time_per_iteration": 2.685472249984741
    },
    {
      "auxiliary_loss_clip": 0.01175397,
      "auxiliary_loss_mlp": 0.01057105,
      "balance_loss_clip": 1.05950332,
      "balance_loss_mlp": 1.03546858,
      "epoch": 0.10082669472418458,
      "flos": 16799119977600.0,
      "grad_norm": 2.008035557658629,
      "language_loss": 0.86132157,
      "learning_rate": 3.947609227619163e-06,
      "loss": 0.88364655,
      "num_input_tokens_seen": 36146255,
      "step": 1677,
      "time_per_iteration": 2.6589157581329346
    },
    {
      "auxiliary_loss_clip": 0.01162997,
      "auxiliary_loss_mlp": 0.010508,
      "balance_loss_clip": 1.05363441,
      "balance_loss_mlp": 1.02896047,
      "epoch": 0.10088681797685255,
      "flos": 13553513844480.0,
      "grad_norm": 2.160847391025828,
      "language_loss": 0.86006588,
      "learning_rate": 3.947520632403936e-06,
      "loss": 0.88220382,
      "num_input_tokens_seen": 36164050,
      "step": 1678,
      "time_per_iteration": 2.694347858428955
    },
    {
      "auxiliary_loss_clip": 0.0116292,
      "auxiliary_loss_mlp": 0.01056376,
      "balance_loss_clip": 1.0587275,
      "balance_loss_mlp": 1.03406048,
      "epoch": 0.10094694122952051,
      "flos": 25265706744960.0,
      "grad_norm": 12.700254532531051,
      "language_loss": 0.89978886,
      "learning_rate": 3.947431963338532e-06,
      "loss": 0.92198181,
      "num_input_tokens_seen": 36183530,
      "step": 1679,
      "time_per_iteration": 2.6741397380828857
    },
    {
      "auxiliary_loss_clip": 0.01071086,
      "auxiliary_loss_mlp": 0.0101685,
      "balance_loss_clip": 1.02328789,
      "balance_loss_mlp": 1.01360798,
      "epoch": 0.10100706448218849,
      "flos": 69854299885440.0,
      "grad_norm": 0.7882499243548835,
      "language_loss": 0.52985126,
      "learning_rate": 3.947343220426312e-06,
      "loss": 0.55073065,
      "num_input_tokens_seen": 36248550,
      "step": 1680,
      "time_per_iteration": 3.169893503189087
    },
    {
      "auxiliary_loss_clip": 0.01185252,
      "auxiliary_loss_mlp": 0.00779951,
      "balance_loss_clip": 1.06022644,
      "balance_loss_mlp": 1.00017488,
      "epoch": 0.10106718773485646,
      "flos": 20007163463040.0,
      "grad_norm": 1.6642182724084642,
      "language_loss": 0.76869059,
      "learning_rate": 3.947254403670641e-06,
      "loss": 0.7883426,
      "num_input_tokens_seen": 36266065,
      "step": 1681,
      "time_per_iteration": 4.146950006484985
    },
    {
      "auxiliary_loss_clip": 0.01156046,
      "auxiliary_loss_mlp": 0.01059972,
      "balance_loss_clip": 1.0539515,
      "balance_loss_mlp": 1.03469992,
      "epoch": 0.10112731098752442,
      "flos": 13479825093120.0,
      "grad_norm": 2.3884003317971225,
      "language_loss": 0.93957508,
      "learning_rate": 3.947165513074889e-06,
      "loss": 0.96173531,
      "num_input_tokens_seen": 36280960,
      "step": 1682,
      "time_per_iteration": 4.220505237579346
    },
    {
      "auxiliary_loss_clip": 0.01173183,
      "auxiliary_loss_mlp": 0.01053261,
      "balance_loss_clip": 1.05487084,
      "balance_loss_mlp": 1.03133821,
      "epoch": 0.1011874342401924,
      "flos": 18515901490560.0,
      "grad_norm": 3.5300660189263917,
      "language_loss": 0.87618893,
      "learning_rate": 3.947076548642425e-06,
      "loss": 0.89845335,
      "num_input_tokens_seen": 36299010,
      "step": 1683,
      "time_per_iteration": 2.635636329650879
    },
    {
      "auxiliary_loss_clip": 0.01128888,
      "auxiliary_loss_mlp": 0.01063089,
      "balance_loss_clip": 1.04814756,
      "balance_loss_mlp": 1.04008126,
      "epoch": 0.10124755749286037,
      "flos": 20702861055360.0,
      "grad_norm": 2.3337760241024923,
      "language_loss": 0.74566805,
      "learning_rate": 3.946987510376624e-06,
      "loss": 0.76758784,
      "num_input_tokens_seen": 36318400,
      "step": 1684,
      "time_per_iteration": 4.417364835739136
    },
    {
      "auxiliary_loss_clip": 0.01053031,
      "auxiliary_loss_mlp": 0.0101182,
      "balance_loss_clip": 1.02547038,
      "balance_loss_mlp": 1.00853014,
      "epoch": 0.10130768074552833,
      "flos": 56109456247680.0,
      "grad_norm": 0.7564631726021327,
      "language_loss": 0.61085057,
      "learning_rate": 3.9468983982808615e-06,
      "loss": 0.63149905,
      "num_input_tokens_seen": 36381815,
      "step": 1685,
      "time_per_iteration": 4.87179970741272
    },
    {
      "auxiliary_loss_clip": 0.01157045,
      "auxiliary_loss_mlp": 0.01056064,
      "balance_loss_clip": 1.05233479,
      "balance_loss_mlp": 1.0341655,
      "epoch": 0.1013678039981963,
      "flos": 33402346156800.0,
      "grad_norm": 4.297801792672815,
      "language_loss": 0.61381406,
      "learning_rate": 3.946809212358516e-06,
      "loss": 0.6359452,
      "num_input_tokens_seen": 36404320,
      "step": 1686,
      "time_per_iteration": 2.8289108276367188
    },
    {
      "auxiliary_loss_clip": 0.01144631,
      "auxiliary_loss_mlp": 0.01059888,
      "balance_loss_clip": 1.05645001,
      "balance_loss_mlp": 1.03678524,
      "epoch": 0.10142792725086427,
      "flos": 31905338008320.0,
      "grad_norm": 2.21923850158845,
      "language_loss": 0.81216162,
      "learning_rate": 3.946719952612972e-06,
      "loss": 0.83420682,
      "num_input_tokens_seen": 36427510,
      "step": 1687,
      "time_per_iteration": 2.947535276412964
    },
    {
      "auxiliary_loss_clip": 0.0117612,
      "auxiliary_loss_mlp": 0.0105614,
      "balance_loss_clip": 1.05933213,
      "balance_loss_mlp": 1.03403926,
      "epoch": 0.10148805050353224,
      "flos": 28475905046400.0,
      "grad_norm": 1.7955898786084035,
      "language_loss": 0.71943259,
      "learning_rate": 3.94663061904761e-06,
      "loss": 0.74175525,
      "num_input_tokens_seen": 36448230,
      "step": 1688,
      "time_per_iteration": 2.693249225616455
    },
    {
      "auxiliary_loss_clip": 0.01151953,
      "auxiliary_loss_mlp": 0.01063362,
      "balance_loss_clip": 1.05288756,
      "balance_loss_mlp": 1.04079556,
      "epoch": 0.1015481737562002,
      "flos": 25148888737920.0,
      "grad_norm": 2.636795901714516,
      "language_loss": 0.86876953,
      "learning_rate": 3.94654121166582e-06,
      "loss": 0.89092261,
      "num_input_tokens_seen": 36464395,
      "step": 1689,
      "time_per_iteration": 2.677992820739746
    },
    {
      "auxiliary_loss_clip": 0.01172188,
      "auxiliary_loss_mlp": 0.01057982,
      "balance_loss_clip": 1.05476904,
      "balance_loss_mlp": 1.0378834,
      "epoch": 0.10160829700886818,
      "flos": 30882781630080.0,
      "grad_norm": 2.2211105929909696,
      "language_loss": 0.88170946,
      "learning_rate": 3.946451730470993e-06,
      "loss": 0.90401113,
      "num_input_tokens_seen": 36486475,
      "step": 1690,
      "time_per_iteration": 2.707209348678589
    },
    {
      "auxiliary_loss_clip": 0.01158767,
      "auxiliary_loss_mlp": 0.01052386,
      "balance_loss_clip": 1.05507553,
      "balance_loss_mlp": 1.02973664,
      "epoch": 0.10166842026153615,
      "flos": 20412020632320.0,
      "grad_norm": 2.08291471600754,
      "language_loss": 0.83348423,
      "learning_rate": 3.946362175466521e-06,
      "loss": 0.85559577,
      "num_input_tokens_seen": 36505310,
      "step": 1691,
      "time_per_iteration": 2.6521170139312744
    },
    {
      "auxiliary_loss_clip": 0.01162159,
      "auxiliary_loss_mlp": 0.01051716,
      "balance_loss_clip": 1.05550599,
      "balance_loss_mlp": 1.03016281,
      "epoch": 0.10172854351420411,
      "flos": 33476968661760.0,
      "grad_norm": 1.704519528530946,
      "language_loss": 0.66773653,
      "learning_rate": 3.946272546655801e-06,
      "loss": 0.68987525,
      "num_input_tokens_seen": 36529820,
      "step": 1692,
      "time_per_iteration": 2.799353837966919
    },
    {
      "auxiliary_loss_clip": 0.01144502,
      "auxiliary_loss_mlp": 0.0107473,
      "balance_loss_clip": 1.05057836,
      "balance_loss_mlp": 1.05258095,
      "epoch": 0.1017886667668721,
      "flos": 23550325862400.0,
      "grad_norm": 1.8345924563029705,
      "language_loss": 0.75939322,
      "learning_rate": 3.94618284404223e-06,
      "loss": 0.78158557,
      "num_input_tokens_seen": 36549000,
      "step": 1693,
      "time_per_iteration": 2.6711113452911377
    },
    {
      "auxiliary_loss_clip": 0.01132621,
      "auxiliary_loss_mlp": 0.01057162,
      "balance_loss_clip": 1.04893303,
      "balance_loss_mlp": 1.03289056,
      "epoch": 0.10184879001954006,
      "flos": 23296078419840.0,
      "grad_norm": 1.7027745569702395,
      "language_loss": 0.87503564,
      "learning_rate": 3.9460930676292105e-06,
      "loss": 0.89693356,
      "num_input_tokens_seen": 36567515,
      "step": 1694,
      "time_per_iteration": 2.749119520187378
    },
    {
      "auxiliary_loss_clip": 0.01130673,
      "auxiliary_loss_mlp": 0.01058451,
      "balance_loss_clip": 1.04954553,
      "balance_loss_mlp": 1.033095,
      "epoch": 0.10190891327220802,
      "flos": 18333116156160.0,
      "grad_norm": 1.7649462193878245,
      "language_loss": 0.79299057,
      "learning_rate": 3.946003217420147e-06,
      "loss": 0.8148818,
      "num_input_tokens_seen": 36586190,
      "step": 1695,
      "time_per_iteration": 2.839081048965454
    },
    {
      "auxiliary_loss_clip": 0.0112732,
      "auxiliary_loss_mlp": 0.01061103,
      "balance_loss_clip": 1.04818296,
      "balance_loss_mlp": 1.03772628,
      "epoch": 0.10196903652487599,
      "flos": 26465374108800.0,
      "grad_norm": 2.7190993931598446,
      "language_loss": 0.86494684,
      "learning_rate": 3.945913293418447e-06,
      "loss": 0.88683105,
      "num_input_tokens_seen": 36607495,
      "step": 1696,
      "time_per_iteration": 2.7802348136901855
    },
    {
      "auxiliary_loss_clip": 0.01168675,
      "auxiliary_loss_mlp": 0.01054661,
      "balance_loss_clip": 1.05711746,
      "balance_loss_mlp": 1.03315568,
      "epoch": 0.10202915977754397,
      "flos": 21869526798720.0,
      "grad_norm": 1.7889048836535288,
      "language_loss": 0.82350796,
      "learning_rate": 3.945823295627519e-06,
      "loss": 0.84574133,
      "num_input_tokens_seen": 36628555,
      "step": 1697,
      "time_per_iteration": 2.667962074279785
    },
    {
      "auxiliary_loss_clip": 0.01184333,
      "auxiliary_loss_mlp": 0.01055548,
      "balance_loss_clip": 1.05680871,
      "balance_loss_mlp": 1.033149,
      "epoch": 0.10208928303021193,
      "flos": 22309755886080.0,
      "grad_norm": 2.0464291543972006,
      "language_loss": 0.81198204,
      "learning_rate": 3.9457332240507775e-06,
      "loss": 0.83438087,
      "num_input_tokens_seen": 36646250,
      "step": 1698,
      "time_per_iteration": 2.6484432220458984
    },
    {
      "auxiliary_loss_clip": 0.01150498,
      "auxiliary_loss_mlp": 0.01053546,
      "balance_loss_clip": 1.05696845,
      "balance_loss_mlp": 1.03226686,
      "epoch": 0.1021494062828799,
      "flos": 22125569921280.0,
      "grad_norm": 2.3020250981163226,
      "language_loss": 0.75612724,
      "learning_rate": 3.945643078691637e-06,
      "loss": 0.77816761,
      "num_input_tokens_seen": 36666675,
      "step": 1699,
      "time_per_iteration": 2.8040614128112793
    },
    {
      "auxiliary_loss_clip": 0.01162088,
      "auxiliary_loss_mlp": 0.01050379,
      "balance_loss_clip": 1.06041551,
      "balance_loss_mlp": 1.02827764,
      "epoch": 0.10220952953554788,
      "flos": 19646728439040.0,
      "grad_norm": 1.6839869206777538,
      "language_loss": 0.80395639,
      "learning_rate": 3.945552859553516e-06,
      "loss": 0.8260811,
      "num_input_tokens_seen": 36685225,
      "step": 1700,
      "time_per_iteration": 2.6701290607452393
    },
    {
      "auxiliary_loss_clip": 0.0117076,
      "auxiliary_loss_mlp": 0.0104804,
      "balance_loss_clip": 1.05714083,
      "balance_loss_mlp": 1.02653444,
      "epoch": 0.10226965278821584,
      "flos": 29787290686080.0,
      "grad_norm": 2.102621975458346,
      "language_loss": 0.76877582,
      "learning_rate": 3.945462566639836e-06,
      "loss": 0.79096377,
      "num_input_tokens_seen": 36705985,
      "step": 1701,
      "time_per_iteration": 2.748201847076416
    },
    {
      "auxiliary_loss_clip": 0.01182259,
      "auxiliary_loss_mlp": 0.01050364,
      "balance_loss_clip": 1.06157088,
      "balance_loss_mlp": 1.02852523,
      "epoch": 0.10232977604088381,
      "flos": 27016818681600.0,
      "grad_norm": 2.1099726588763965,
      "language_loss": 0.77922845,
      "learning_rate": 3.945372199954019e-06,
      "loss": 0.80155474,
      "num_input_tokens_seen": 36725815,
      "step": 1702,
      "time_per_iteration": 2.6703274250030518
    },
    {
      "auxiliary_loss_clip": 0.01156323,
      "auxiliary_loss_mlp": 0.01052524,
      "balance_loss_clip": 1.05596721,
      "balance_loss_mlp": 1.03126872,
      "epoch": 0.10238989929355179,
      "flos": 20777519473920.0,
      "grad_norm": 2.2326457826946293,
      "language_loss": 0.94093609,
      "learning_rate": 3.945281759499494e-06,
      "loss": 0.96302462,
      "num_input_tokens_seen": 36742345,
      "step": 1703,
      "time_per_iteration": 2.6712698936462402
    },
    {
      "auxiliary_loss_clip": 0.01034483,
      "auxiliary_loss_mlp": 0.01037784,
      "balance_loss_clip": 1.02765131,
      "balance_loss_mlp": 1.03315914,
      "epoch": 0.10245002254621975,
      "flos": 57698322451200.0,
      "grad_norm": 0.8815387598011586,
      "language_loss": 0.55096036,
      "learning_rate": 3.94519124527969e-06,
      "loss": 0.57168299,
      "num_input_tokens_seen": 36798775,
      "step": 1704,
      "time_per_iteration": 3.2863855361938477
    },
    {
      "auxiliary_loss_clip": 0.01186822,
      "auxiliary_loss_mlp": 0.01053701,
      "balance_loss_clip": 1.06026638,
      "balance_loss_mlp": 1.03088403,
      "epoch": 0.10251014579888772,
      "flos": 16800125558400.0,
      "grad_norm": 2.051901555713709,
      "language_loss": 0.84025991,
      "learning_rate": 3.945100657298039e-06,
      "loss": 0.86266518,
      "num_input_tokens_seen": 36816295,
      "step": 1705,
      "time_per_iteration": 2.8991851806640625
    },
    {
      "auxiliary_loss_clip": 0.01045354,
      "auxiliary_loss_mlp": 0.01018361,
      "balance_loss_clip": 1.02622223,
      "balance_loss_mlp": 1.01526153,
      "epoch": 0.1025702690515557,
      "flos": 68565500922240.0,
      "grad_norm": 0.7692746082941451,
      "language_loss": 0.60408181,
      "learning_rate": 3.9450099955579765e-06,
      "loss": 0.62471896,
      "num_input_tokens_seen": 36882030,
      "step": 1706,
      "time_per_iteration": 3.2174558639526367
    },
    {
      "auxiliary_loss_clip": 0.01149922,
      "auxiliary_loss_mlp": 0.01051211,
      "balance_loss_clip": 1.05388391,
      "balance_loss_mlp": 1.02812052,
      "epoch": 0.10263039230422366,
      "flos": 14866623336960.0,
      "grad_norm": 2.201796189576969,
      "language_loss": 0.85937822,
      "learning_rate": 3.94491926006294e-06,
      "loss": 0.88138962,
      "num_input_tokens_seen": 36899245,
      "step": 1707,
      "time_per_iteration": 2.689208507537842
    },
    {
      "auxiliary_loss_clip": 0.01169165,
      "auxiliary_loss_mlp": 0.0105297,
      "balance_loss_clip": 1.05941081,
      "balance_loss_mlp": 1.03114319,
      "epoch": 0.10269051555689163,
      "flos": 25337599816320.0,
      "grad_norm": 1.471109036018689,
      "language_loss": 0.73299325,
      "learning_rate": 3.944828450816369e-06,
      "loss": 0.75521457,
      "num_input_tokens_seen": 36920950,
      "step": 1708,
      "time_per_iteration": 2.679760456085205
    },
    {
      "auxiliary_loss_clip": 0.01155833,
      "auxiliary_loss_mlp": 0.00780571,
      "balance_loss_clip": 1.05718231,
      "balance_loss_mlp": 1.00042295,
      "epoch": 0.10275063880955959,
      "flos": 21068826773760.0,
      "grad_norm": 1.7051644476897239,
      "language_loss": 0.91616452,
      "learning_rate": 3.944737567821709e-06,
      "loss": 0.93552846,
      "num_input_tokens_seen": 36938900,
      "step": 1709,
      "time_per_iteration": 2.6754679679870605
    },
    {
      "auxiliary_loss_clip": 0.01124911,
      "auxiliary_loss_mlp": 0.01057008,
      "balance_loss_clip": 1.05144072,
      "balance_loss_mlp": 1.0343945,
      "epoch": 0.10281076206222757,
      "flos": 30366780802560.0,
      "grad_norm": 2.1056252966717275,
      "language_loss": 0.88004494,
      "learning_rate": 3.944646611082406e-06,
      "loss": 0.90186411,
      "num_input_tokens_seen": 36957010,
      "step": 1710,
      "time_per_iteration": 2.708723306655884
    },
    {
      "auxiliary_loss_clip": 0.01171004,
      "auxiliary_loss_mlp": 0.0105967,
      "balance_loss_clip": 1.05658317,
      "balance_loss_mlp": 1.036973,
      "epoch": 0.10287088531489554,
      "flos": 22418313765120.0,
      "grad_norm": 1.7046493271202992,
      "language_loss": 0.79370153,
      "learning_rate": 3.944555580601908e-06,
      "loss": 0.81600821,
      "num_input_tokens_seen": 36977690,
      "step": 1711,
      "time_per_iteration": 2.631908416748047
    },
    {
      "auxiliary_loss_clip": 0.01156003,
      "auxiliary_loss_mlp": 0.01055126,
      "balance_loss_clip": 1.05841637,
      "balance_loss_mlp": 1.03189242,
      "epoch": 0.1029310085675635,
      "flos": 25115994858240.0,
      "grad_norm": 3.2168061349371135,
      "language_loss": 0.73666596,
      "learning_rate": 3.944464476383668e-06,
      "loss": 0.75877726,
      "num_input_tokens_seen": 36997300,
      "step": 1712,
      "time_per_iteration": 2.7107467651367188
    },
    {
      "auxiliary_loss_clip": 0.01133407,
      "auxiliary_loss_mlp": 0.01056055,
      "balance_loss_clip": 1.05496907,
      "balance_loss_mlp": 1.03334546,
      "epoch": 0.10299113182023148,
      "flos": 19865639877120.0,
      "grad_norm": 1.974447377126898,
      "language_loss": 0.87049067,
      "learning_rate": 3.94437329843114e-06,
      "loss": 0.89238536,
      "num_input_tokens_seen": 37016110,
      "step": 1713,
      "time_per_iteration": 2.6532411575317383
    },
    {
      "auxiliary_loss_clip": 0.0116832,
      "auxiliary_loss_mlp": 0.01060237,
      "balance_loss_clip": 1.05669498,
      "balance_loss_mlp": 1.03877962,
      "epoch": 0.10305125507289944,
      "flos": 20447608032000.0,
      "grad_norm": 1.57388574383124,
      "language_loss": 0.72406238,
      "learning_rate": 3.944282046747782e-06,
      "loss": 0.74634796,
      "num_input_tokens_seen": 37036405,
      "step": 1714,
      "time_per_iteration": 2.5987610816955566
    },
    {
      "auxiliary_loss_clip": 0.01174482,
      "auxiliary_loss_mlp": 0.01063165,
      "balance_loss_clip": 1.05715692,
      "balance_loss_mlp": 1.03934693,
      "epoch": 0.10311137832556741,
      "flos": 26250772302720.0,
      "grad_norm": 2.1959530175190434,
      "language_loss": 0.91065919,
      "learning_rate": 3.944190721337053e-06,
      "loss": 0.93303567,
      "num_input_tokens_seen": 37057580,
      "step": 1715,
      "time_per_iteration": 2.743833303451538
    },
    {
      "auxiliary_loss_clip": 0.01170297,
      "auxiliary_loss_mlp": 0.01054891,
      "balance_loss_clip": 1.05448914,
      "balance_loss_mlp": 1.03305221,
      "epoch": 0.10317150157823539,
      "flos": 35298932175360.0,
      "grad_norm": 1.8741123562687005,
      "language_loss": 0.75969976,
      "learning_rate": 3.944099322202418e-06,
      "loss": 0.78195167,
      "num_input_tokens_seen": 37079120,
      "step": 1716,
      "time_per_iteration": 2.748903274536133
    },
    {
      "auxiliary_loss_clip": 0.01162664,
      "auxiliary_loss_mlp": 0.01061895,
      "balance_loss_clip": 1.05617428,
      "balance_loss_mlp": 1.03804111,
      "epoch": 0.10323162483090335,
      "flos": 25739943033600.0,
      "grad_norm": 3.178190042364093,
      "language_loss": 0.85308528,
      "learning_rate": 3.944007849347342e-06,
      "loss": 0.87533092,
      "num_input_tokens_seen": 37099710,
      "step": 1717,
      "time_per_iteration": 2.690772533416748
    },
    {
      "auxiliary_loss_clip": 0.01127019,
      "auxiliary_loss_mlp": 0.01067935,
      "balance_loss_clip": 1.05048633,
      "balance_loss_mlp": 1.04436755,
      "epoch": 0.10329174808357132,
      "flos": 16289870906880.0,
      "grad_norm": 1.8474438265561113,
      "language_loss": 0.82945001,
      "learning_rate": 3.943916302775292e-06,
      "loss": 0.85139954,
      "num_input_tokens_seen": 37117775,
      "step": 1718,
      "time_per_iteration": 2.7029476165771484
    },
    {
      "auxiliary_loss_clip": 0.01171184,
      "auxiliary_loss_mlp": 0.01049869,
      "balance_loss_clip": 1.05912328,
      "balance_loss_mlp": 1.02701616,
      "epoch": 0.10335187133623928,
      "flos": 36687166963200.0,
      "grad_norm": 1.7728224248964342,
      "language_loss": 0.73396438,
      "learning_rate": 3.943824682489742e-06,
      "loss": 0.75617492,
      "num_input_tokens_seen": 37140280,
      "step": 1719,
      "time_per_iteration": 2.7653820514678955
    },
    {
      "auxiliary_loss_clip": 0.01168859,
      "auxiliary_loss_mlp": 0.01048444,
      "balance_loss_clip": 1.05861163,
      "balance_loss_mlp": 1.02786827,
      "epoch": 0.10341199458890726,
      "flos": 14975648092800.0,
      "grad_norm": 1.7819459058763836,
      "language_loss": 0.92692196,
      "learning_rate": 3.9437329884941665e-06,
      "loss": 0.94909501,
      "num_input_tokens_seen": 37158350,
      "step": 1720,
      "time_per_iteration": 4.1962480545043945
    },
    {
      "auxiliary_loss_clip": 0.01139894,
      "auxiliary_loss_mlp": 0.01051033,
      "balance_loss_clip": 1.05092323,
      "balance_loss_mlp": 1.02827597,
      "epoch": 0.10347211784157523,
      "flos": 21031587348480.0,
      "grad_norm": 1.6861044154168399,
      "language_loss": 0.79497123,
      "learning_rate": 3.943641220792039e-06,
      "loss": 0.81688046,
      "num_input_tokens_seen": 37177120,
      "step": 1721,
      "time_per_iteration": 4.524151802062988
    },
    {
      "auxiliary_loss_clip": 0.01130482,
      "auxiliary_loss_mlp": 0.01067754,
      "balance_loss_clip": 1.05380797,
      "balance_loss_mlp": 1.04109859,
      "epoch": 0.1035322410942432,
      "flos": 19792094780160.0,
      "grad_norm": 1.951940775381607,
      "language_loss": 0.80707669,
      "learning_rate": 3.9435493793868434e-06,
      "loss": 0.829059,
      "num_input_tokens_seen": 37195895,
      "step": 1722,
      "time_per_iteration": 2.7972562313079834
    },
    {
      "auxiliary_loss_clip": 0.01059018,
      "auxiliary_loss_mlp": 0.01038991,
      "balance_loss_clip": 1.02668202,
      "balance_loss_mlp": 1.03536737,
      "epoch": 0.10359236434691117,
      "flos": 52698874947840.0,
      "grad_norm": 0.9413879826908518,
      "language_loss": 0.67161834,
      "learning_rate": 3.943457464282059e-06,
      "loss": 0.69259846,
      "num_input_tokens_seen": 37247270,
      "step": 1723,
      "time_per_iteration": 4.899553060531616
    },
    {
      "auxiliary_loss_clip": 0.01169875,
      "auxiliary_loss_mlp": 0.01062977,
      "balance_loss_clip": 1.05482125,
      "balance_loss_mlp": 1.04193664,
      "epoch": 0.10365248759957914,
      "flos": 18405404277120.0,
      "grad_norm": 2.8641520576523116,
      "language_loss": 0.77715755,
      "learning_rate": 3.9433654754811745e-06,
      "loss": 0.7994861,
      "num_input_tokens_seen": 37265595,
      "step": 1724,
      "time_per_iteration": 2.7613437175750732
    },
    {
      "auxiliary_loss_clip": 0.01151829,
      "auxiliary_loss_mlp": 0.01069246,
      "balance_loss_clip": 1.05667496,
      "balance_loss_mlp": 1.04753852,
      "epoch": 0.1037126108522471,
      "flos": 47553555335040.0,
      "grad_norm": 2.6433978354033543,
      "language_loss": 0.74533165,
      "learning_rate": 3.943273412987676e-06,
      "loss": 0.76754242,
      "num_input_tokens_seen": 37286660,
      "step": 1725,
      "time_per_iteration": 4.557274580001831
    },
    {
      "auxiliary_loss_clip": 0.01137065,
      "auxiliary_loss_mlp": 0.01081067,
      "balance_loss_clip": 1.05264461,
      "balance_loss_mlp": 1.05832207,
      "epoch": 0.10377273410491508,
      "flos": 22816670572800.0,
      "grad_norm": 2.2241153649877865,
      "language_loss": 0.75043738,
      "learning_rate": 3.943181276805054e-06,
      "loss": 0.77261865,
      "num_input_tokens_seen": 37304915,
      "step": 1726,
      "time_per_iteration": 2.7098495960235596
    },
    {
      "auxiliary_loss_clip": 0.01150932,
      "auxiliary_loss_mlp": 0.0107864,
      "balance_loss_clip": 1.05345368,
      "balance_loss_mlp": 1.05610991,
      "epoch": 0.10383285735758305,
      "flos": 26138694890880.0,
      "grad_norm": 2.783771441956431,
      "language_loss": 0.73243797,
      "learning_rate": 3.9430890669368035e-06,
      "loss": 0.75473368,
      "num_input_tokens_seen": 37325265,
      "step": 1727,
      "time_per_iteration": 2.74774169921875
    },
    {
      "auxiliary_loss_clip": 0.01157922,
      "auxiliary_loss_mlp": 0.01068007,
      "balance_loss_clip": 1.05303776,
      "balance_loss_mlp": 1.04625082,
      "epoch": 0.10389298061025101,
      "flos": 17091791994240.0,
      "grad_norm": 2.172978726198527,
      "language_loss": 0.84373868,
      "learning_rate": 3.942996783386422e-06,
      "loss": 0.86599791,
      "num_input_tokens_seen": 37341650,
      "step": 1728,
      "time_per_iteration": 2.675724744796753
    },
    {
      "auxiliary_loss_clip": 0.01154897,
      "auxiliary_loss_mlp": 0.01060505,
      "balance_loss_clip": 1.0545603,
      "balance_loss_mlp": 1.0393219,
      "epoch": 0.10395310386291898,
      "flos": 20776513893120.0,
      "grad_norm": 2.1406499008555513,
      "language_loss": 0.70776087,
      "learning_rate": 3.942904426157406e-06,
      "loss": 0.7299149,
      "num_input_tokens_seen": 37360270,
      "step": 1729,
      "time_per_iteration": 2.6885008811950684
    },
    {
      "auxiliary_loss_clip": 0.01158623,
      "auxiliary_loss_mlp": 0.01068311,
      "balance_loss_clip": 1.05437422,
      "balance_loss_mlp": 1.04520774,
      "epoch": 0.10401322711558696,
      "flos": 12820540913280.0,
      "grad_norm": 2.4133379049648283,
      "language_loss": 0.81237471,
      "learning_rate": 3.9428119952532605e-06,
      "loss": 0.83464402,
      "num_input_tokens_seen": 37375225,
      "step": 1730,
      "time_per_iteration": 2.6659536361694336
    },
    {
      "auxiliary_loss_clip": 0.01085856,
      "auxiliary_loss_mlp": 0.01063394,
      "balance_loss_clip": 1.04733562,
      "balance_loss_mlp": 1.04314065,
      "epoch": 0.10407335036825492,
      "flos": 23184683366400.0,
      "grad_norm": 1.6634499611984725,
      "language_loss": 0.75829297,
      "learning_rate": 3.942719490677489e-06,
      "loss": 0.77978551,
      "num_input_tokens_seen": 37395165,
      "step": 1731,
      "time_per_iteration": 3.043125629425049
    },
    {
      "auxiliary_loss_clip": 0.01129913,
      "auxiliary_loss_mlp": 0.01065783,
      "balance_loss_clip": 1.0526607,
      "balance_loss_mlp": 1.04604149,
      "epoch": 0.10413347362092289,
      "flos": 26104184899200.0,
      "grad_norm": 1.8280179918091173,
      "language_loss": 0.8268069,
      "learning_rate": 3.9426269124336e-06,
      "loss": 0.84876388,
      "num_input_tokens_seen": 37414845,
      "step": 1732,
      "time_per_iteration": 2.96221661567688
    },
    {
      "auxiliary_loss_clip": 0.01141505,
      "auxiliary_loss_mlp": 0.01067805,
      "balance_loss_clip": 1.05805755,
      "balance_loss_mlp": 1.04852867,
      "epoch": 0.10419359687359087,
      "flos": 12641059630080.0,
      "grad_norm": 1.9919813178368582,
      "language_loss": 0.83320522,
      "learning_rate": 3.942534260525104e-06,
      "loss": 0.85529828,
      "num_input_tokens_seen": 37432490,
      "step": 1733,
      "time_per_iteration": 2.7364420890808105
    },
    {
      "auxiliary_loss_clip": 0.01153374,
      "auxiliary_loss_mlp": 0.0106675,
      "balance_loss_clip": 1.05592012,
      "balance_loss_mlp": 1.04654372,
      "epoch": 0.10425372012625883,
      "flos": 12125094716160.0,
      "grad_norm": 2.4441875881355597,
      "language_loss": 0.76683885,
      "learning_rate": 3.942441534955514e-06,
      "loss": 0.78904009,
      "num_input_tokens_seen": 37449435,
      "step": 1734,
      "time_per_iteration": 2.669623851776123
    },
    {
      "auxiliary_loss_clip": 0.0113597,
      "auxiliary_loss_mlp": 0.01052567,
      "balance_loss_clip": 1.05042601,
      "balance_loss_mlp": 1.03255177,
      "epoch": 0.1043138433789268,
      "flos": 25337563902720.0,
      "grad_norm": 1.6775801166329647,
      "language_loss": 0.74826896,
      "learning_rate": 3.9423487357283465e-06,
      "loss": 0.7701543,
      "num_input_tokens_seen": 37469105,
      "step": 1735,
      "time_per_iteration": 2.8477160930633545
    },
    {
      "auxiliary_loss_clip": 0.01167698,
      "auxiliary_loss_mlp": 0.01055716,
      "balance_loss_clip": 1.05678105,
      "balance_loss_mlp": 1.0344727,
      "epoch": 0.10437396663159478,
      "flos": 29167149352320.0,
      "grad_norm": 1.7228393064183538,
      "language_loss": 0.78835273,
      "learning_rate": 3.94225586284712e-06,
      "loss": 0.81058681,
      "num_input_tokens_seen": 37490540,
      "step": 1736,
      "time_per_iteration": 2.690453052520752
    },
    {
      "auxiliary_loss_clip": 0.0116734,
      "auxiliary_loss_mlp": 0.01064692,
      "balance_loss_clip": 1.05800533,
      "balance_loss_mlp": 1.04357982,
      "epoch": 0.10443408988426274,
      "flos": 25080946162560.0,
      "grad_norm": 1.8549131823334455,
      "language_loss": 0.7058785,
      "learning_rate": 3.942162916315356e-06,
      "loss": 0.72819883,
      "num_input_tokens_seen": 37511905,
      "step": 1737,
      "time_per_iteration": 2.6296744346618652
    },
    {
      "auxiliary_loss_clip": 0.01150138,
      "auxiliary_loss_mlp": 0.01059407,
      "balance_loss_clip": 1.04806042,
      "balance_loss_mlp": 1.03600669,
      "epoch": 0.1044942131369307,
      "flos": 26759662237440.0,
      "grad_norm": 2.415613377802324,
      "language_loss": 0.81624997,
      "learning_rate": 3.942069896136581e-06,
      "loss": 0.83834541,
      "num_input_tokens_seen": 37533635,
      "step": 1738,
      "time_per_iteration": 2.7436723709106445
    },
    {
      "auxiliary_loss_clip": 0.01181471,
      "auxiliary_loss_mlp": 0.01062035,
      "balance_loss_clip": 1.05579174,
      "balance_loss_mlp": 1.03950453,
      "epoch": 0.10455433638959867,
      "flos": 18442571875200.0,
      "grad_norm": 2.1004590024567897,
      "language_loss": 0.75419426,
      "learning_rate": 3.9419768023143196e-06,
      "loss": 0.77662933,
      "num_input_tokens_seen": 37552035,
      "step": 1739,
      "time_per_iteration": 2.585538148880005
    },
    {
      "auxiliary_loss_clip": 0.01146716,
      "auxiliary_loss_mlp": 0.01054893,
      "balance_loss_clip": 1.05417264,
      "balance_loss_mlp": 1.03348303,
      "epoch": 0.10461445964226665,
      "flos": 23218977876480.0,
      "grad_norm": 1.586314706443492,
      "language_loss": 0.77523744,
      "learning_rate": 3.941883634852104e-06,
      "loss": 0.79725355,
      "num_input_tokens_seen": 37571540,
      "step": 1740,
      "time_per_iteration": 2.8947789669036865
    },
    {
      "auxiliary_loss_clip": 0.01152077,
      "auxiliary_loss_mlp": 0.01049503,
      "balance_loss_clip": 1.05725431,
      "balance_loss_mlp": 1.0288676,
      "epoch": 0.10467458289493461,
      "flos": 24345243797760.0,
      "grad_norm": 1.964868695493703,
      "language_loss": 0.85976374,
      "learning_rate": 3.941790393753467e-06,
      "loss": 0.88177955,
      "num_input_tokens_seen": 37588265,
      "step": 1741,
      "time_per_iteration": 2.7706260681152344
    },
    {
      "auxiliary_loss_clip": 0.01158134,
      "auxiliary_loss_mlp": 0.01056311,
      "balance_loss_clip": 1.05614483,
      "balance_loss_mlp": 1.03350592,
      "epoch": 0.10473470614760258,
      "flos": 21287953693440.0,
      "grad_norm": 5.197245251055922,
      "language_loss": 0.75592613,
      "learning_rate": 3.941697079021942e-06,
      "loss": 0.77807057,
      "num_input_tokens_seen": 37606860,
      "step": 1742,
      "time_per_iteration": 2.784748077392578
    },
    {
      "auxiliary_loss_clip": 0.0113066,
      "auxiliary_loss_mlp": 0.01057571,
      "balance_loss_clip": 1.05678856,
      "balance_loss_mlp": 1.03735304,
      "epoch": 0.10479482940027056,
      "flos": 21687208341120.0,
      "grad_norm": 2.1426857583950416,
      "language_loss": 0.87614191,
      "learning_rate": 3.94160369066107e-06,
      "loss": 0.89802414,
      "num_input_tokens_seen": 37625210,
      "step": 1743,
      "time_per_iteration": 2.819350004196167
    },
    {
      "auxiliary_loss_clip": 0.01139959,
      "auxiliary_loss_mlp": 0.01048534,
      "balance_loss_clip": 1.0552268,
      "balance_loss_mlp": 1.0254786,
      "epoch": 0.10485495265293852,
      "flos": 21573694385280.0,
      "grad_norm": 2.060686178474056,
      "language_loss": 0.75927812,
      "learning_rate": 3.941510228674391e-06,
      "loss": 0.7811631,
      "num_input_tokens_seen": 37644110,
      "step": 1744,
      "time_per_iteration": 2.7817211151123047
    },
    {
      "auxiliary_loss_clip": 0.01170232,
      "auxiliary_loss_mlp": 0.01054483,
      "balance_loss_clip": 1.05992889,
      "balance_loss_mlp": 1.03442037,
      "epoch": 0.10491507590560649,
      "flos": 37961923708800.0,
      "grad_norm": 1.9689383181633062,
      "language_loss": 0.78905094,
      "learning_rate": 3.941416693065451e-06,
      "loss": 0.81129813,
      "num_input_tokens_seen": 37665800,
      "step": 1745,
      "time_per_iteration": 2.88080096244812
    },
    {
      "auxiliary_loss_clip": 0.01180482,
      "auxiliary_loss_mlp": 0.01060479,
      "balance_loss_clip": 1.05740213,
      "balance_loss_mlp": 1.03920031,
      "epoch": 0.10497519915827447,
      "flos": 26396282298240.0,
      "grad_norm": 2.64819141351011,
      "language_loss": 0.82568693,
      "learning_rate": 3.941323083837794e-06,
      "loss": 0.84809649,
      "num_input_tokens_seen": 37685095,
      "step": 1746,
      "time_per_iteration": 2.7068004608154297
    },
    {
      "auxiliary_loss_clip": 0.01158367,
      "auxiliary_loss_mlp": 0.0105595,
      "balance_loss_clip": 1.05737162,
      "balance_loss_mlp": 1.03448033,
      "epoch": 0.10503532241094243,
      "flos": 40662190581120.0,
      "grad_norm": 1.6274602877205533,
      "language_loss": 0.70573747,
      "learning_rate": 3.941229400994971e-06,
      "loss": 0.7278806,
      "num_input_tokens_seen": 37707445,
      "step": 1747,
      "time_per_iteration": 2.8689963817596436
    },
    {
      "auxiliary_loss_clip": 0.01159389,
      "auxiliary_loss_mlp": 0.01056346,
      "balance_loss_clip": 1.06035507,
      "balance_loss_mlp": 1.03492367,
      "epoch": 0.1050954456636104,
      "flos": 29789409588480.0,
      "grad_norm": 2.386885173400054,
      "language_loss": 0.8447504,
      "learning_rate": 3.941135644540535e-06,
      "loss": 0.86690772,
      "num_input_tokens_seen": 37728325,
      "step": 1748,
      "time_per_iteration": 2.8022749423980713
    },
    {
      "auxiliary_loss_clip": 0.01175489,
      "auxiliary_loss_mlp": 0.01049407,
      "balance_loss_clip": 1.05471563,
      "balance_loss_mlp": 1.02701974,
      "epoch": 0.10515556891627838,
      "flos": 23948754497280.0,
      "grad_norm": 1.759895679837136,
      "language_loss": 0.71681082,
      "learning_rate": 3.941041814478041e-06,
      "loss": 0.73905981,
      "num_input_tokens_seen": 37748910,
      "step": 1749,
      "time_per_iteration": 2.6568849086761475
    },
    {
      "auxiliary_loss_clip": 0.01158221,
      "auxiliary_loss_mlp": 0.01058697,
      "balance_loss_clip": 1.05427456,
      "balance_loss_mlp": 1.03590393,
      "epoch": 0.10521569216894634,
      "flos": 18259606972800.0,
      "grad_norm": 2.95022560634889,
      "language_loss": 0.81510806,
      "learning_rate": 3.940947910811047e-06,
      "loss": 0.83727717,
      "num_input_tokens_seen": 37765745,
      "step": 1750,
      "time_per_iteration": 2.6282739639282227
    },
    {
      "auxiliary_loss_clip": 0.01156475,
      "auxiliary_loss_mlp": 0.01062657,
      "balance_loss_clip": 1.06022298,
      "balance_loss_mlp": 1.03973269,
      "epoch": 0.10527581542161431,
      "flos": 15630909949440.0,
      "grad_norm": 2.2218325288878953,
      "language_loss": 0.92364043,
      "learning_rate": 3.940853933543114e-06,
      "loss": 0.94583178,
      "num_input_tokens_seen": 37780520,
      "step": 1751,
      "time_per_iteration": 2.703376531600952
    },
    {
      "auxiliary_loss_clip": 0.01165779,
      "auxiliary_loss_mlp": 0.01053304,
      "balance_loss_clip": 1.0570029,
      "balance_loss_mlp": 1.03171563,
      "epoch": 0.10533593867428227,
      "flos": 18296559089280.0,
      "grad_norm": 2.0356912608722877,
      "language_loss": 0.79293752,
      "learning_rate": 3.940759882677805e-06,
      "loss": 0.81512833,
      "num_input_tokens_seen": 37799515,
      "step": 1752,
      "time_per_iteration": 2.6501150131225586
    },
    {
      "auxiliary_loss_clip": 0.01116865,
      "auxiliary_loss_mlp": 0.01055489,
      "balance_loss_clip": 1.05116987,
      "balance_loss_mlp": 1.03264856,
      "epoch": 0.10539606192695025,
      "flos": 29023219555200.0,
      "grad_norm": 2.022904639316529,
      "language_loss": 0.75978744,
      "learning_rate": 3.940665758218686e-06,
      "loss": 0.78151095,
      "num_input_tokens_seen": 37818695,
      "step": 1753,
      "time_per_iteration": 2.871335744857788
    },
    {
      "auxiliary_loss_clip": 0.01141721,
      "auxiliary_loss_mlp": 0.01057356,
      "balance_loss_clip": 1.05547547,
      "balance_loss_mlp": 1.03415775,
      "epoch": 0.10545618517961822,
      "flos": 19969313506560.0,
      "grad_norm": 2.0563919939847914,
      "language_loss": 0.83969283,
      "learning_rate": 3.940571560169328e-06,
      "loss": 0.86168355,
      "num_input_tokens_seen": 37837860,
      "step": 1754,
      "time_per_iteration": 2.685591459274292
    },
    {
      "auxiliary_loss_clip": 0.01136802,
      "auxiliary_loss_mlp": 0.01053577,
      "balance_loss_clip": 1.05587101,
      "balance_loss_mlp": 1.03034329,
      "epoch": 0.10551630843228618,
      "flos": 16143427157760.0,
      "grad_norm": 2.7567281016961087,
      "language_loss": 0.68732727,
      "learning_rate": 3.940477288533302e-06,
      "loss": 0.70923102,
      "num_input_tokens_seen": 37856260,
      "step": 1755,
      "time_per_iteration": 2.754117727279663
    },
    {
      "auxiliary_loss_clip": 0.01161626,
      "auxiliary_loss_mlp": 0.010623,
      "balance_loss_clip": 1.05367684,
      "balance_loss_mlp": 1.040187,
      "epoch": 0.10557643168495416,
      "flos": 23440115957760.0,
      "grad_norm": 2.26658946748733,
      "language_loss": 0.76382339,
      "learning_rate": 3.940382943314182e-06,
      "loss": 0.7860626,
      "num_input_tokens_seen": 37876960,
      "step": 1756,
      "time_per_iteration": 2.686790943145752
    },
    {
      "auxiliary_loss_clip": 0.01182062,
      "auxiliary_loss_mlp": 0.01062906,
      "balance_loss_clip": 1.05688286,
      "balance_loss_mlp": 1.04203284,
      "epoch": 0.10563655493762213,
      "flos": 21799034357760.0,
      "grad_norm": 1.5917029795724482,
      "language_loss": 0.79926664,
      "learning_rate": 3.940288524515547e-06,
      "loss": 0.82171631,
      "num_input_tokens_seen": 37897070,
      "step": 1757,
      "time_per_iteration": 2.6543681621551514
    },
    {
      "auxiliary_loss_clip": 0.01149304,
      "auxiliary_loss_mlp": 0.01057523,
      "balance_loss_clip": 1.0524838,
      "balance_loss_mlp": 1.03563643,
      "epoch": 0.10569667819029009,
      "flos": 53800863275520.0,
      "grad_norm": 1.6583181970862437,
      "language_loss": 0.78714895,
      "learning_rate": 3.940194032140976e-06,
      "loss": 0.80921721,
      "num_input_tokens_seen": 37923635,
      "step": 1758,
      "time_per_iteration": 3.013157367706299
    },
    {
      "auxiliary_loss_clip": 0.01165597,
      "auxiliary_loss_mlp": 0.01054919,
      "balance_loss_clip": 1.05894113,
      "balance_loss_mlp": 1.03347349,
      "epoch": 0.10575680144295807,
      "flos": 22925515760640.0,
      "grad_norm": 1.870482409236857,
      "language_loss": 0.91388202,
      "learning_rate": 3.940099466194054e-06,
      "loss": 0.93608713,
      "num_input_tokens_seen": 37942650,
      "step": 1759,
      "time_per_iteration": 4.1841137409210205
    },
    {
      "auxiliary_loss_clip": 0.0115455,
      "auxiliary_loss_mlp": 0.01056708,
      "balance_loss_clip": 1.05242109,
      "balance_loss_mlp": 1.03346229,
      "epoch": 0.10581692469562604,
      "flos": 14136667148160.0,
      "grad_norm": 2.509404173865799,
      "language_loss": 0.77406812,
      "learning_rate": 3.940004826678365e-06,
      "loss": 0.79618067,
      "num_input_tokens_seen": 37960660,
      "step": 1760,
      "time_per_iteration": 4.476959228515625
    },
    {
      "auxiliary_loss_clip": 0.01161737,
      "auxiliary_loss_mlp": 0.01064522,
      "balance_loss_clip": 1.0536418,
      "balance_loss_mlp": 1.04053712,
      "epoch": 0.105877047948294,
      "flos": 25958674903680.0,
      "grad_norm": 2.27300461956159,
      "language_loss": 0.88896096,
      "learning_rate": 3.939910113597498e-06,
      "loss": 0.91122353,
      "num_input_tokens_seen": 37978625,
      "step": 1761,
      "time_per_iteration": 2.6907520294189453
    },
    {
      "auxiliary_loss_clip": 0.01110571,
      "auxiliary_loss_mlp": 0.00782389,
      "balance_loss_clip": 1.04964042,
      "balance_loss_mlp": 1.00012767,
      "epoch": 0.10593717120096197,
      "flos": 30664768032000.0,
      "grad_norm": 2.010693315376097,
      "language_loss": 0.7809304,
      "learning_rate": 3.9398153269550464e-06,
      "loss": 0.79986,
      "num_input_tokens_seen": 38000005,
      "step": 1762,
      "time_per_iteration": 2.869051456451416
    },
    {
      "auxiliary_loss_clip": 0.01053171,
      "auxiliary_loss_mlp": 0.0105371,
      "balance_loss_clip": 1.02694225,
      "balance_loss_mlp": 1.05056334,
      "epoch": 0.10599729445362994,
      "flos": 66436682497920.0,
      "grad_norm": 0.8956567750819878,
      "language_loss": 0.60503203,
      "learning_rate": 3.939720466754602e-06,
      "loss": 0.6261009,
      "num_input_tokens_seen": 38066165,
      "step": 1763,
      "time_per_iteration": 5.049196720123291
    },
    {
      "auxiliary_loss_clip": 0.01156865,
      "auxiliary_loss_mlp": 0.01048706,
      "balance_loss_clip": 1.05424261,
      "balance_loss_mlp": 1.02708137,
      "epoch": 0.10605741770629791,
      "flos": 23948179879680.0,
      "grad_norm": 2.0510547250099633,
      "language_loss": 0.80232942,
      "learning_rate": 3.939625532999763e-06,
      "loss": 0.82438517,
      "num_input_tokens_seen": 38086150,
      "step": 1764,
      "time_per_iteration": 4.288762807846069
    },
    {
      "auxiliary_loss_clip": 0.01136032,
      "auxiliary_loss_mlp": 0.01055975,
      "balance_loss_clip": 1.04879069,
      "balance_loss_mlp": 1.03218043,
      "epoch": 0.10611754095896588,
      "flos": 19387524919680.0,
      "grad_norm": 1.693202084864273,
      "language_loss": 0.801691,
      "learning_rate": 3.9395305256941314e-06,
      "loss": 0.82361102,
      "num_input_tokens_seen": 38104205,
      "step": 1765,
      "time_per_iteration": 2.931269407272339
    },
    {
      "auxiliary_loss_clip": 0.01163261,
      "auxiliary_loss_mlp": 0.01058956,
      "balance_loss_clip": 1.05457163,
      "balance_loss_mlp": 1.0367949,
      "epoch": 0.10617766421163385,
      "flos": 22237755073920.0,
      "grad_norm": 1.7665774264343403,
      "language_loss": 0.76864165,
      "learning_rate": 3.939435444841306e-06,
      "loss": 0.79086387,
      "num_input_tokens_seen": 38122005,
      "step": 1766,
      "time_per_iteration": 2.5976176261901855
    },
    {
      "auxiliary_loss_clip": 0.01182495,
      "auxiliary_loss_mlp": 0.01059246,
      "balance_loss_clip": 1.05923963,
      "balance_loss_mlp": 1.03766894,
      "epoch": 0.10623778746430182,
      "flos": 28404407024640.0,
      "grad_norm": 1.6265727447650185,
      "language_loss": 0.77311498,
      "learning_rate": 3.939340290444895e-06,
      "loss": 0.79553241,
      "num_input_tokens_seen": 38143365,
      "step": 1767,
      "time_per_iteration": 2.6356630325317383
    },
    {
      "auxiliary_loss_clip": 0.01006515,
      "auxiliary_loss_mlp": 0.01018751,
      "balance_loss_clip": 1.03004837,
      "balance_loss_mlp": 1.0151509,
      "epoch": 0.10629791071696978,
      "flos": 64234639221120.0,
      "grad_norm": 0.9172341423433896,
      "language_loss": 0.57889944,
      "learning_rate": 3.939245062508506e-06,
      "loss": 0.59915209,
      "num_input_tokens_seen": 38210035,
      "step": 1768,
      "time_per_iteration": 3.6866471767425537
    },
    {
      "auxiliary_loss_clip": 0.01144481,
      "auxiliary_loss_mlp": 0.01047419,
      "balance_loss_clip": 1.0546546,
      "balance_loss_mlp": 1.02687907,
      "epoch": 0.10635803396963776,
      "flos": 22747578762240.0,
      "grad_norm": 1.4529696494540971,
      "language_loss": 0.86711109,
      "learning_rate": 3.939149761035749e-06,
      "loss": 0.8890301,
      "num_input_tokens_seen": 38231230,
      "step": 1769,
      "time_per_iteration": 3.936905860900879
    },
    {
      "auxiliary_loss_clip": 0.01141219,
      "auxiliary_loss_mlp": 0.00780338,
      "balance_loss_clip": 1.05321527,
      "balance_loss_mlp": 1.00008726,
      "epoch": 0.10641815722230573,
      "flos": 31395586147200.0,
      "grad_norm": 1.8275276693890916,
      "language_loss": 0.61906171,
      "learning_rate": 3.9390543860302395e-06,
      "loss": 0.63827729,
      "num_input_tokens_seen": 38253890,
      "step": 1770,
      "time_per_iteration": 2.8926138877868652
    },
    {
      "auxiliary_loss_clip": 0.01057689,
      "auxiliary_loss_mlp": 0.01010808,
      "balance_loss_clip": 1.02007711,
      "balance_loss_mlp": 1.00775671,
      "epoch": 0.1064782804749737,
      "flos": 58552527784320.0,
      "grad_norm": 0.9163874753670794,
      "language_loss": 0.57049137,
      "learning_rate": 3.9389589374955925e-06,
      "loss": 0.59117633,
      "num_input_tokens_seen": 38304290,
      "step": 1771,
      "time_per_iteration": 3.0783088207244873
    },
    {
      "auxiliary_loss_clip": 0.01146276,
      "auxiliary_loss_mlp": 0.01065918,
      "balance_loss_clip": 1.05574095,
      "balance_loss_mlp": 1.04465103,
      "epoch": 0.10653840372764166,
      "flos": 23987825516160.0,
      "grad_norm": 12.794881398939157,
      "language_loss": 0.88265753,
      "learning_rate": 3.938863415435429e-06,
      "loss": 0.90477949,
      "num_input_tokens_seen": 38324725,
      "step": 1772,
      "time_per_iteration": 2.770202159881592
    },
    {
      "auxiliary_loss_clip": 0.0118421,
      "auxiliary_loss_mlp": 0.01058161,
      "balance_loss_clip": 1.05697048,
      "balance_loss_mlp": 1.03497458,
      "epoch": 0.10659852698030964,
      "flos": 18294655668480.0,
      "grad_norm": 2.576940958490313,
      "language_loss": 0.76030588,
      "learning_rate": 3.93876781985337e-06,
      "loss": 0.78272957,
      "num_input_tokens_seen": 38340735,
      "step": 1773,
      "time_per_iteration": 2.6177070140838623
    },
    {
      "auxiliary_loss_clip": 0.01122733,
      "auxiliary_loss_mlp": 0.01067657,
      "balance_loss_clip": 1.04691553,
      "balance_loss_mlp": 1.04205084,
      "epoch": 0.1066586502329776,
      "flos": 32160591031680.0,
      "grad_norm": 1.868288871406422,
      "language_loss": 0.8330853,
      "learning_rate": 3.938672150753041e-06,
      "loss": 0.85498923,
      "num_input_tokens_seen": 38361315,
      "step": 1774,
      "time_per_iteration": 2.7396061420440674
    },
    {
      "auxiliary_loss_clip": 0.01156305,
      "auxiliary_loss_mlp": 0.00780518,
      "balance_loss_clip": 1.05627465,
      "balance_loss_mlp": 1.00011277,
      "epoch": 0.10671877348564557,
      "flos": 17785155202560.0,
      "grad_norm": 2.73383407032925,
      "language_loss": 0.76446521,
      "learning_rate": 3.9385764081380704e-06,
      "loss": 0.78383344,
      "num_input_tokens_seen": 38377425,
      "step": 1775,
      "time_per_iteration": 2.624208927154541
    },
    {
      "auxiliary_loss_clip": 0.01063199,
      "auxiliary_loss_mlp": 0.01007654,
      "balance_loss_clip": 1.01726675,
      "balance_loss_mlp": 1.00443542,
      "epoch": 0.10677889673831355,
      "flos": 63510177813120.0,
      "grad_norm": 0.8200823962511624,
      "language_loss": 0.57477289,
      "learning_rate": 3.9384805920120876e-06,
      "loss": 0.5954814,
      "num_input_tokens_seen": 38440275,
      "step": 1776,
      "time_per_iteration": 3.1782386302948
    },
    {
      "auxiliary_loss_clip": 0.01150087,
      "auxiliary_loss_mlp": 0.01066244,
      "balance_loss_clip": 1.05192852,
      "balance_loss_mlp": 1.0407691,
      "epoch": 0.10683901999098151,
      "flos": 22017694400640.0,
      "grad_norm": 1.4232532718517703,
      "language_loss": 0.83442962,
      "learning_rate": 3.938384702378727e-06,
      "loss": 0.85659301,
      "num_input_tokens_seen": 38461820,
      "step": 1777,
      "time_per_iteration": 2.7342305183410645
    },
    {
      "auxiliary_loss_clip": 0.01113855,
      "auxiliary_loss_mlp": 0.00780712,
      "balance_loss_clip": 1.04919302,
      "balance_loss_mlp": 1.00015831,
      "epoch": 0.10689914324364948,
      "flos": 25042952551680.0,
      "grad_norm": 1.8326039994575831,
      "language_loss": 0.87207437,
      "learning_rate": 3.938288739241625e-06,
      "loss": 0.89102006,
      "num_input_tokens_seen": 38482235,
      "step": 1778,
      "time_per_iteration": 2.859834671020508
    },
    {
      "auxiliary_loss_clip": 0.01152509,
      "auxiliary_loss_mlp": 0.00780436,
      "balance_loss_clip": 1.06804752,
      "balance_loss_mlp": 1.00019765,
      "epoch": 0.10695926649631746,
      "flos": 16435129507200.0,
      "grad_norm": 2.4525249429301823,
      "language_loss": 0.84165859,
      "learning_rate": 3.938192702604417e-06,
      "loss": 0.86098808,
      "num_input_tokens_seen": 38500690,
      "step": 1779,
      "time_per_iteration": 2.81423020362854
    },
    {
      "auxiliary_loss_clip": 0.01141718,
      "auxiliary_loss_mlp": 0.00779857,
      "balance_loss_clip": 1.05215359,
      "balance_loss_mlp": 1.0001775,
      "epoch": 0.10701938974898542,
      "flos": 16979211792000.0,
      "grad_norm": 1.9378348403129941,
      "language_loss": 0.66915894,
      "learning_rate": 3.9380965924707495e-06,
      "loss": 0.68837464,
      "num_input_tokens_seen": 38518405,
      "step": 1780,
      "time_per_iteration": 2.616684913635254
    },
    {
      "auxiliary_loss_clip": 0.01166288,
      "auxiliary_loss_mlp": 0.01054109,
      "balance_loss_clip": 1.05843914,
      "balance_loss_mlp": 1.03268683,
      "epoch": 0.10707951300165339,
      "flos": 15888102307200.0,
      "grad_norm": 1.9168180254288365,
      "language_loss": 0.92058647,
      "learning_rate": 3.938000408844265e-06,
      "loss": 0.94279045,
      "num_input_tokens_seen": 38535060,
      "step": 1781,
      "time_per_iteration": 2.6167802810668945
    },
    {
      "auxiliary_loss_clip": 0.0113109,
      "auxiliary_loss_mlp": 0.01064554,
      "balance_loss_clip": 1.0531441,
      "balance_loss_mlp": 1.04344225,
      "epoch": 0.10713963625432135,
      "flos": 14247164361600.0,
      "grad_norm": 1.8357670097294174,
      "language_loss": 0.79336482,
      "learning_rate": 3.9379041517286105e-06,
      "loss": 0.81532121,
      "num_input_tokens_seen": 38552855,
      "step": 1782,
      "time_per_iteration": 2.7669336795806885
    },
    {
      "auxiliary_loss_clip": 0.01158369,
      "auxiliary_loss_mlp": 0.01061646,
      "balance_loss_clip": 1.05510604,
      "balance_loss_mlp": 1.04016423,
      "epoch": 0.10719975950698933,
      "flos": 16756780821120.0,
      "grad_norm": 2.0914095256513945,
      "language_loss": 0.79086542,
      "learning_rate": 3.937807821127436e-06,
      "loss": 0.81306553,
      "num_input_tokens_seen": 38570075,
      "step": 1783,
      "time_per_iteration": 2.6349542140960693
    },
    {
      "auxiliary_loss_clip": 0.01164267,
      "auxiliary_loss_mlp": 0.01065333,
      "balance_loss_clip": 1.0570296,
      "balance_loss_mlp": 1.04299295,
      "epoch": 0.1072598827596573,
      "flos": 22710626645760.0,
      "grad_norm": 2.1874612027367806,
      "language_loss": 0.86421812,
      "learning_rate": 3.937711417044395e-06,
      "loss": 0.88651407,
      "num_input_tokens_seen": 38587970,
      "step": 1784,
      "time_per_iteration": 2.8452541828155518
    },
    {
      "auxiliary_loss_clip": 0.01153461,
      "auxiliary_loss_mlp": 0.01055605,
      "balance_loss_clip": 1.05502176,
      "balance_loss_mlp": 1.03321707,
      "epoch": 0.10732000601232526,
      "flos": 23258264376960.0,
      "grad_norm": 2.4649130783319553,
      "language_loss": 1.01192284,
      "learning_rate": 3.937614939483143e-06,
      "loss": 1.03401351,
      "num_input_tokens_seen": 38605840,
      "step": 1785,
      "time_per_iteration": 2.690018653869629
    },
    {
      "auxiliary_loss_clip": 0.01168517,
      "auxiliary_loss_mlp": 0.01060763,
      "balance_loss_clip": 1.05854678,
      "balance_loss_mlp": 1.03984189,
      "epoch": 0.10738012926499324,
      "flos": 24207060176640.0,
      "grad_norm": 1.397915549237645,
      "language_loss": 0.84951413,
      "learning_rate": 3.937518388447339e-06,
      "loss": 0.87180698,
      "num_input_tokens_seen": 38627070,
      "step": 1786,
      "time_per_iteration": 2.637430191040039
    },
    {
      "auxiliary_loss_clip": 0.01183118,
      "auxiliary_loss_mlp": 0.01059079,
      "balance_loss_clip": 1.05716729,
      "balance_loss_mlp": 1.03520155,
      "epoch": 0.1074402525176612,
      "flos": 20923065383040.0,
      "grad_norm": 1.7951357311742837,
      "language_loss": 0.78861409,
      "learning_rate": 3.937421763940642e-06,
      "loss": 0.81103605,
      "num_input_tokens_seen": 38645840,
      "step": 1787,
      "time_per_iteration": 2.54508900642395
    },
    {
      "auxiliary_loss_clip": 0.01174896,
      "auxiliary_loss_mlp": 0.01047406,
      "balance_loss_clip": 1.05971575,
      "balance_loss_mlp": 1.02528071,
      "epoch": 0.10750037577032917,
      "flos": 16946928443520.0,
      "grad_norm": 1.8536072321218278,
      "language_loss": 0.82307518,
      "learning_rate": 3.937325065966719e-06,
      "loss": 0.84529817,
      "num_input_tokens_seen": 38664770,
      "step": 1788,
      "time_per_iteration": 2.706247568130493
    },
    {
      "auxiliary_loss_clip": 0.01180896,
      "auxiliary_loss_mlp": 0.01064682,
      "balance_loss_clip": 1.05843878,
      "balance_loss_mlp": 1.04427314,
      "epoch": 0.10756049902299715,
      "flos": 20266546550400.0,
      "grad_norm": 2.110245519520894,
      "language_loss": 0.77840686,
      "learning_rate": 3.9372282945292335e-06,
      "loss": 0.80086267,
      "num_input_tokens_seen": 38683865,
      "step": 1789,
      "time_per_iteration": 2.6274654865264893
    },
    {
      "auxiliary_loss_clip": 0.01185566,
      "auxiliary_loss_mlp": 0.01065099,
      "balance_loss_clip": 1.0604099,
      "balance_loss_mlp": 1.04049408,
      "epoch": 0.10762062227566511,
      "flos": 23586523793280.0,
      "grad_norm": 2.7248977042722524,
      "language_loss": 0.74817526,
      "learning_rate": 3.937131449631859e-06,
      "loss": 0.77068192,
      "num_input_tokens_seen": 38702485,
      "step": 1790,
      "time_per_iteration": 2.624382972717285
    },
    {
      "auxiliary_loss_clip": 0.01178128,
      "auxiliary_loss_mlp": 0.00780572,
      "balance_loss_clip": 1.06110644,
      "balance_loss_mlp": 1.00021124,
      "epoch": 0.10768074552833308,
      "flos": 24310626065280.0,
      "grad_norm": 2.350797373347828,
      "language_loss": 0.78764236,
      "learning_rate": 3.9370345312782645e-06,
      "loss": 0.80722934,
      "num_input_tokens_seen": 38722475,
      "step": 1791,
      "time_per_iteration": 2.696162223815918
    },
    {
      "auxiliary_loss_clip": 0.01134133,
      "auxiliary_loss_mlp": 0.01065057,
      "balance_loss_clip": 1.05280125,
      "balance_loss_mlp": 1.04117918,
      "epoch": 0.10774086878100106,
      "flos": 25299965341440.0,
      "grad_norm": 1.5879424734455678,
      "language_loss": 0.70638013,
      "learning_rate": 3.936937539472126e-06,
      "loss": 0.7283721,
      "num_input_tokens_seen": 38743285,
      "step": 1792,
      "time_per_iteration": 2.770874261856079
    },
    {
      "auxiliary_loss_clip": 0.01149934,
      "auxiliary_loss_mlp": 0.01051019,
      "balance_loss_clip": 1.05610943,
      "balance_loss_mlp": 1.02764249,
      "epoch": 0.10780099203366902,
      "flos": 22054035985920.0,
      "grad_norm": 1.920104493539276,
      "language_loss": 0.76565266,
      "learning_rate": 3.9368404742171236e-06,
      "loss": 0.78766215,
      "num_input_tokens_seen": 38763035,
      "step": 1793,
      "time_per_iteration": 2.7218761444091797
    },
    {
      "auxiliary_loss_clip": 0.01116412,
      "auxiliary_loss_mlp": 0.01064574,
      "balance_loss_clip": 1.05029237,
      "balance_loss_mlp": 1.0414238,
      "epoch": 0.10786111528633699,
      "flos": 22747471021440.0,
      "grad_norm": 1.7475786500241859,
      "language_loss": 0.85103315,
      "learning_rate": 3.936743335516936e-06,
      "loss": 0.87284303,
      "num_input_tokens_seen": 38784900,
      "step": 1794,
      "time_per_iteration": 2.7590620517730713
    },
    {
      "auxiliary_loss_clip": 0.01115198,
      "auxiliary_loss_mlp": 0.01055294,
      "balance_loss_clip": 1.04807687,
      "balance_loss_mlp": 1.03146446,
      "epoch": 0.10792123853900495,
      "flos": 20851064570880.0,
      "grad_norm": 2.5236234593460924,
      "language_loss": 0.74585378,
      "learning_rate": 3.936646123375246e-06,
      "loss": 0.76755869,
      "num_input_tokens_seen": 38804695,
      "step": 1795,
      "time_per_iteration": 2.8500585556030273
    },
    {
      "auxiliary_loss_clip": 0.01124895,
      "auxiliary_loss_mlp": 0.01058294,
      "balance_loss_clip": 1.04831553,
      "balance_loss_mlp": 1.03479767,
      "epoch": 0.10798136179167293,
      "flos": 17748705876480.0,
      "grad_norm": 2.842374039298248,
      "language_loss": 0.81653619,
      "learning_rate": 3.936548837795741e-06,
      "loss": 0.83836806,
      "num_input_tokens_seen": 38822395,
      "step": 1796,
      "time_per_iteration": 2.7549750804901123
    },
    {
      "auxiliary_loss_clip": 0.01140492,
      "auxiliary_loss_mlp": 0.01083966,
      "balance_loss_clip": 1.05246449,
      "balance_loss_mlp": 1.05721593,
      "epoch": 0.1080414850443409,
      "flos": 13589639948160.0,
      "grad_norm": 2.59635455269928,
      "language_loss": 0.74233043,
      "learning_rate": 3.936451478782111e-06,
      "loss": 0.764575,
      "num_input_tokens_seen": 38839865,
      "step": 1797,
      "time_per_iteration": 2.6396753787994385
    },
    {
      "auxiliary_loss_clip": 0.01160286,
      "auxiliary_loss_mlp": 0.01049954,
      "balance_loss_clip": 1.05505061,
      "balance_loss_mlp": 1.02874684,
      "epoch": 0.10810160829700886,
      "flos": 16253421580800.0,
      "grad_norm": 2.0852339617015025,
      "language_loss": 0.81855786,
      "learning_rate": 3.936354046338046e-06,
      "loss": 0.84066033,
      "num_input_tokens_seen": 38857300,
      "step": 1798,
      "time_per_iteration": 2.7105324268341064
    },
    {
      "auxiliary_loss_clip": 0.01142859,
      "auxiliary_loss_mlp": 0.01054502,
      "balance_loss_clip": 1.05379176,
      "balance_loss_mlp": 1.03117299,
      "epoch": 0.10816173154967684,
      "flos": 15158002464000.0,
      "grad_norm": 2.4443000829323687,
      "language_loss": 0.85516405,
      "learning_rate": 3.936256540467242e-06,
      "loss": 0.87713766,
      "num_input_tokens_seen": 38874960,
      "step": 1799,
      "time_per_iteration": 4.159978628158569
    },
    {
      "auxiliary_loss_clip": 0.01154352,
      "auxiliary_loss_mlp": 0.01062903,
      "balance_loss_clip": 1.05493283,
      "balance_loss_mlp": 1.04114687,
      "epoch": 0.10822185480234481,
      "flos": 17785334770560.0,
      "grad_norm": 2.7405734706827825,
      "language_loss": 0.77434146,
      "learning_rate": 3.9361589611733955e-06,
      "loss": 0.79651403,
      "num_input_tokens_seen": 38893610,
      "step": 1800,
      "time_per_iteration": 4.52047872543335
    },
    {
      "auxiliary_loss_clip": 0.01178634,
      "auxiliary_loss_mlp": 0.0104758,
      "balance_loss_clip": 1.05722904,
      "balance_loss_mlp": 1.02689719,
      "epoch": 0.10828197805501277,
      "flos": 25556654908800.0,
      "grad_norm": 1.582468034859118,
      "language_loss": 0.72897375,
      "learning_rate": 3.9360613084602075e-06,
      "loss": 0.75123584,
      "num_input_tokens_seen": 38913485,
      "step": 1801,
      "time_per_iteration": 4.291400909423828
    },
    {
      "auxiliary_loss_clip": 0.01190595,
      "auxiliary_loss_mlp": 0.01056056,
      "balance_loss_clip": 1.06095624,
      "balance_loss_mlp": 1.03478956,
      "epoch": 0.10834210130768075,
      "flos": 28984435845120.0,
      "grad_norm": 1.951139287607183,
      "language_loss": 0.6634692,
      "learning_rate": 3.935963582331381e-06,
      "loss": 0.68593562,
      "num_input_tokens_seen": 38935650,
      "step": 1802,
      "time_per_iteration": 2.722628355026245
    },
    {
      "auxiliary_loss_clip": 0.01155661,
      "auxiliary_loss_mlp": 0.01059375,
      "balance_loss_clip": 1.05326533,
      "balance_loss_mlp": 1.03695142,
      "epoch": 0.10840222456034872,
      "flos": 20264212166400.0,
      "grad_norm": 2.084551157592464,
      "language_loss": 0.81612957,
      "learning_rate": 3.935865782790621e-06,
      "loss": 0.8382799,
      "num_input_tokens_seen": 38954130,
      "step": 1803,
      "time_per_iteration": 4.239379167556763
    },
    {
      "auxiliary_loss_clip": 0.01163104,
      "auxiliary_loss_mlp": 0.01061781,
      "balance_loss_clip": 1.0567112,
      "balance_loss_mlp": 1.03921473,
      "epoch": 0.10846234781301668,
      "flos": 19863054097920.0,
      "grad_norm": 1.9102934552723363,
      "language_loss": 0.91127038,
      "learning_rate": 3.9357679098416365e-06,
      "loss": 0.93351918,
      "num_input_tokens_seen": 38972905,
      "step": 1804,
      "time_per_iteration": 2.5836737155914307
    },
    {
      "auxiliary_loss_clip": 0.01136188,
      "auxiliary_loss_mlp": 0.01060133,
      "balance_loss_clip": 1.05617714,
      "balance_loss_mlp": 1.03718543,
      "epoch": 0.10852247106568465,
      "flos": 26469037296000.0,
      "grad_norm": 2.5742522317806262,
      "language_loss": 0.76198906,
      "learning_rate": 3.935669963488139e-06,
      "loss": 0.78395224,
      "num_input_tokens_seen": 38993255,
      "step": 1805,
      "time_per_iteration": 2.783137321472168
    },
    {
      "auxiliary_loss_clip": 0.01149468,
      "auxiliary_loss_mlp": 0.01050946,
      "balance_loss_clip": 1.05419612,
      "balance_loss_mlp": 1.03050184,
      "epoch": 0.10858259431835263,
      "flos": 30081506987520.0,
      "grad_norm": 1.7049574807827799,
      "language_loss": 0.85876733,
      "learning_rate": 3.935571943733843e-06,
      "loss": 0.88077152,
      "num_input_tokens_seen": 39012610,
      "step": 1806,
      "time_per_iteration": 2.8148701190948486
    },
    {
      "auxiliary_loss_clip": 0.01168733,
      "auxiliary_loss_mlp": 0.00779888,
      "balance_loss_clip": 1.05462408,
      "balance_loss_mlp": 1.00006652,
      "epoch": 0.10864271757102059,
      "flos": 19063180085760.0,
      "grad_norm": 2.554050049117878,
      "language_loss": 0.8108198,
      "learning_rate": 3.9354738505824635e-06,
      "loss": 0.83030605,
      "num_input_tokens_seen": 39030120,
      "step": 1807,
      "time_per_iteration": 2.6275649070739746
    },
    {
      "auxiliary_loss_clip": 0.01139085,
      "auxiliary_loss_mlp": 0.01055438,
      "balance_loss_clip": 1.05193985,
      "balance_loss_mlp": 1.03522038,
      "epoch": 0.10870284082368856,
      "flos": 24715052271360.0,
      "grad_norm": 1.834914777588586,
      "language_loss": 0.78910971,
      "learning_rate": 3.9353756840377225e-06,
      "loss": 0.81105494,
      "num_input_tokens_seen": 39049875,
      "step": 1808,
      "time_per_iteration": 2.722910165786743
    },
    {
      "auxiliary_loss_clip": 0.01157997,
      "auxiliary_loss_mlp": 0.01056971,
      "balance_loss_clip": 1.05918014,
      "balance_loss_mlp": 1.03548992,
      "epoch": 0.10876296407635654,
      "flos": 20627663932800.0,
      "grad_norm": 1.6201371380093192,
      "language_loss": 0.79013431,
      "learning_rate": 3.935277444103342e-06,
      "loss": 0.81228393,
      "num_input_tokens_seen": 39068935,
      "step": 1809,
      "time_per_iteration": 2.7261481285095215
    },
    {
      "auxiliary_loss_clip": 0.01180468,
      "auxiliary_loss_mlp": 0.01057915,
      "balance_loss_clip": 1.0568099,
      "balance_loss_mlp": 1.03705359,
      "epoch": 0.1088230873290245,
      "flos": 21579835610880.0,
      "grad_norm": 1.9004896030263678,
      "language_loss": 0.85129547,
      "learning_rate": 3.935179130783046e-06,
      "loss": 0.87367928,
      "num_input_tokens_seen": 39087370,
      "step": 1810,
      "time_per_iteration": 2.672696828842163
    },
    {
      "auxiliary_loss_clip": 0.01124301,
      "auxiliary_loss_mlp": 0.01057363,
      "balance_loss_clip": 1.04580724,
      "balance_loss_mlp": 1.0335803,
      "epoch": 0.10888321058169247,
      "flos": 26469037296000.0,
      "grad_norm": 1.5993643379141278,
      "language_loss": 0.63822675,
      "learning_rate": 3.935080744080564e-06,
      "loss": 0.66004336,
      "num_input_tokens_seen": 39106635,
      "step": 1811,
      "time_per_iteration": 2.7731611728668213
    },
    {
      "auxiliary_loss_clip": 0.01151891,
      "auxiliary_loss_mlp": 0.01050225,
      "balance_loss_clip": 1.05335796,
      "balance_loss_mlp": 1.02836192,
      "epoch": 0.10894333383436045,
      "flos": 25848608653440.0,
      "grad_norm": 1.9284151803363307,
      "language_loss": 0.74238706,
      "learning_rate": 3.934982283999626e-06,
      "loss": 0.76440823,
      "num_input_tokens_seen": 39126335,
      "step": 1812,
      "time_per_iteration": 2.727743625640869
    },
    {
      "auxiliary_loss_clip": 0.01142498,
      "auxiliary_loss_mlp": 0.01057826,
      "balance_loss_clip": 1.05199611,
      "balance_loss_mlp": 1.03546214,
      "epoch": 0.10900345708702841,
      "flos": 19537093152000.0,
      "grad_norm": 1.5783196636767667,
      "language_loss": 0.72746086,
      "learning_rate": 3.934883750543966e-06,
      "loss": 0.74946409,
      "num_input_tokens_seen": 39144820,
      "step": 1813,
      "time_per_iteration": 2.798297166824341
    },
    {
      "auxiliary_loss_clip": 0.0113892,
      "auxiliary_loss_mlp": 0.01056639,
      "balance_loss_clip": 1.0511452,
      "balance_loss_mlp": 1.03515792,
      "epoch": 0.10906358033969638,
      "flos": 23623296341760.0,
      "grad_norm": 1.635228619121262,
      "language_loss": 0.82981038,
      "learning_rate": 3.93478514371732e-06,
      "loss": 0.85176599,
      "num_input_tokens_seen": 39165945,
      "step": 1814,
      "time_per_iteration": 2.7120048999786377
    },
    {
      "auxiliary_loss_clip": 0.01141958,
      "auxiliary_loss_mlp": 0.01058857,
      "balance_loss_clip": 1.0537864,
      "balance_loss_mlp": 1.03787625,
      "epoch": 0.10912370359236434,
      "flos": 21214731818880.0,
      "grad_norm": 1.9556743991494996,
      "language_loss": 0.84310579,
      "learning_rate": 3.934686463523429e-06,
      "loss": 0.86511397,
      "num_input_tokens_seen": 39183520,
      "step": 1815,
      "time_per_iteration": 2.788870096206665
    },
    {
      "auxiliary_loss_clip": 0.01146878,
      "auxiliary_loss_mlp": 0.01055141,
      "balance_loss_clip": 1.05443966,
      "balance_loss_mlp": 1.03182411,
      "epoch": 0.10918382684503232,
      "flos": 13553190622080.0,
      "grad_norm": 2.5374826422013195,
      "language_loss": 0.71670222,
      "learning_rate": 3.9345877099660315e-06,
      "loss": 0.73872244,
      "num_input_tokens_seen": 39201190,
      "step": 1816,
      "time_per_iteration": 2.8424103260040283
    },
    {
      "auxiliary_loss_clip": 0.01164173,
      "auxiliary_loss_mlp": 0.01064184,
      "balance_loss_clip": 1.05216932,
      "balance_loss_mlp": 1.04052126,
      "epoch": 0.10924395009770028,
      "flos": 27964321591680.0,
      "grad_norm": 2.016899555923086,
      "language_loss": 0.72880268,
      "learning_rate": 3.9344888830488744e-06,
      "loss": 0.75108624,
      "num_input_tokens_seen": 39221210,
      "step": 1817,
      "time_per_iteration": 2.7320947647094727
    },
    {
      "auxiliary_loss_clip": 0.01116915,
      "auxiliary_loss_mlp": 0.01057856,
      "balance_loss_clip": 1.05173278,
      "balance_loss_mlp": 1.03517008,
      "epoch": 0.10930407335036825,
      "flos": 25593750679680.0,
      "grad_norm": 1.5988628345308824,
      "language_loss": 0.67275256,
      "learning_rate": 3.934389982775706e-06,
      "loss": 0.69450033,
      "num_input_tokens_seen": 39242025,
      "step": 1818,
      "time_per_iteration": 2.8700790405273438
    },
    {
      "auxiliary_loss_clip": 0.01155804,
      "auxiliary_loss_mlp": 0.01065952,
      "balance_loss_clip": 1.05673873,
      "balance_loss_mlp": 1.04313517,
      "epoch": 0.10936419660303623,
      "flos": 18406194376320.0,
      "grad_norm": 3.593580913512793,
      "language_loss": 0.73149616,
      "learning_rate": 3.934291009150275e-06,
      "loss": 0.75371373,
      "num_input_tokens_seen": 39259870,
      "step": 1819,
      "time_per_iteration": 2.7091007232666016
    },
    {
      "auxiliary_loss_clip": 0.01142955,
      "auxiliary_loss_mlp": 0.00779155,
      "balance_loss_clip": 1.05341268,
      "balance_loss_mlp": 1.00027704,
      "epoch": 0.1094243198557042,
      "flos": 23840052963840.0,
      "grad_norm": 4.531598275817935,
      "language_loss": 0.73764241,
      "learning_rate": 3.934191962176335e-06,
      "loss": 0.75686359,
      "num_input_tokens_seen": 39278500,
      "step": 1820,
      "time_per_iteration": 2.6513099670410156
    },
    {
      "auxiliary_loss_clip": 0.01179358,
      "auxiliary_loss_mlp": 0.01056073,
      "balance_loss_clip": 1.05747604,
      "balance_loss_mlp": 1.03297031,
      "epoch": 0.10948444310837216,
      "flos": 14643940970880.0,
      "grad_norm": 2.2567103978329337,
      "language_loss": 0.82532805,
      "learning_rate": 3.934092841857642e-06,
      "loss": 0.84768236,
      "num_input_tokens_seen": 39294800,
      "step": 1821,
      "time_per_iteration": 2.5348384380340576
    },
    {
      "auxiliary_loss_clip": 0.01148016,
      "auxiliary_loss_mlp": 0.01052031,
      "balance_loss_clip": 1.05133605,
      "balance_loss_mlp": 1.03077567,
      "epoch": 0.10954456636104014,
      "flos": 27818811596160.0,
      "grad_norm": 2.0770330480401578,
      "language_loss": 0.76271641,
      "learning_rate": 3.933993648197955e-06,
      "loss": 0.7847169,
      "num_input_tokens_seen": 39314625,
      "step": 1822,
      "time_per_iteration": 2.730079174041748
    },
    {
      "auxiliary_loss_clip": 0.01142446,
      "auxiliary_loss_mlp": 0.01049259,
      "balance_loss_clip": 1.04849207,
      "balance_loss_mlp": 1.02856421,
      "epoch": 0.1096046896137081,
      "flos": 33620934372480.0,
      "grad_norm": 1.734419613996414,
      "language_loss": 0.79309607,
      "learning_rate": 3.933894381201034e-06,
      "loss": 0.81501311,
      "num_input_tokens_seen": 39336465,
      "step": 1823,
      "time_per_iteration": 2.756969928741455
    },
    {
      "auxiliary_loss_clip": 0.01148165,
      "auxiliary_loss_mlp": 0.01049595,
      "balance_loss_clip": 1.05160606,
      "balance_loss_mlp": 1.02745807,
      "epoch": 0.10966481286637607,
      "flos": 26980010219520.0,
      "grad_norm": 1.4318009514182364,
      "language_loss": 0.79590744,
      "learning_rate": 3.933795040870645e-06,
      "loss": 0.81788504,
      "num_input_tokens_seen": 39357930,
      "step": 1824,
      "time_per_iteration": 2.798168182373047
    },
    {
      "auxiliary_loss_clip": 0.01142146,
      "auxiliary_loss_mlp": 0.01055513,
      "balance_loss_clip": 1.05104232,
      "balance_loss_mlp": 1.03381693,
      "epoch": 0.10972493611904403,
      "flos": 23036551678080.0,
      "grad_norm": 2.127143421089703,
      "language_loss": 0.88138539,
      "learning_rate": 3.933695627210554e-06,
      "loss": 0.90336192,
      "num_input_tokens_seen": 39376380,
      "step": 1825,
      "time_per_iteration": 2.6804513931274414
    },
    {
      "auxiliary_loss_clip": 0.01128623,
      "auxiliary_loss_mlp": 0.01056127,
      "balance_loss_clip": 1.04586983,
      "balance_loss_mlp": 1.03439498,
      "epoch": 0.10978505937171201,
      "flos": 38104632443520.0,
      "grad_norm": 1.721192594935189,
      "language_loss": 0.76441038,
      "learning_rate": 3.933596140224532e-06,
      "loss": 0.78625786,
      "num_input_tokens_seen": 39399935,
      "step": 1826,
      "time_per_iteration": 2.8315086364746094
    },
    {
      "auxiliary_loss_clip": 0.01063155,
      "auxiliary_loss_mlp": 0.01016957,
      "balance_loss_clip": 1.02709544,
      "balance_loss_mlp": 1.01409554,
      "epoch": 0.10984518262437998,
      "flos": 59849694616320.0,
      "grad_norm": 0.8518463216820418,
      "language_loss": 0.54997343,
      "learning_rate": 3.93349657991635e-06,
      "loss": 0.57077461,
      "num_input_tokens_seen": 39460685,
      "step": 1827,
      "time_per_iteration": 3.1425766944885254
    },
    {
      "auxiliary_loss_clip": 0.01072651,
      "auxiliary_loss_mlp": 0.01010167,
      "balance_loss_clip": 1.02693772,
      "balance_loss_mlp": 1.00717473,
      "epoch": 0.10990530587704794,
      "flos": 66719837410560.0,
      "grad_norm": 0.7375455878808789,
      "language_loss": 0.55382878,
      "learning_rate": 3.933396946289784e-06,
      "loss": 0.57465696,
      "num_input_tokens_seen": 39524765,
      "step": 1828,
      "time_per_iteration": 3.168165922164917
    },
    {
      "auxiliary_loss_clip": 0.01156998,
      "auxiliary_loss_mlp": 0.01059335,
      "balance_loss_clip": 1.05407059,
      "balance_loss_mlp": 1.03618491,
      "epoch": 0.10996542912971592,
      "flos": 25447199189760.0,
      "grad_norm": 2.250827401167328,
      "language_loss": 0.84010404,
      "learning_rate": 3.933297239348612e-06,
      "loss": 0.86226743,
      "num_input_tokens_seen": 39543640,
      "step": 1829,
      "time_per_iteration": 2.7341628074645996
    },
    {
      "auxiliary_loss_clip": 0.01130747,
      "auxiliary_loss_mlp": 0.01053464,
      "balance_loss_clip": 1.0547024,
      "balance_loss_mlp": 1.03036165,
      "epoch": 0.11002555238238389,
      "flos": 44018186186880.0,
      "grad_norm": 2.342204785330024,
      "language_loss": 0.88880253,
      "learning_rate": 3.933197459096614e-06,
      "loss": 0.91064465,
      "num_input_tokens_seen": 39567525,
      "step": 1830,
      "time_per_iteration": 2.9093260765075684
    },
    {
      "auxiliary_loss_clip": 0.01049643,
      "auxiliary_loss_mlp": 0.01009685,
      "balance_loss_clip": 1.02618647,
      "balance_loss_mlp": 1.00681162,
      "epoch": 0.11008567563505185,
      "flos": 54065133590400.0,
      "grad_norm": 0.6882192363357665,
      "language_loss": 0.55566543,
      "learning_rate": 3.9330976055375756e-06,
      "loss": 0.57625872,
      "num_input_tokens_seen": 39628470,
      "step": 1831,
      "time_per_iteration": 3.1713974475860596
    },
    {
      "auxiliary_loss_clip": 0.01156783,
      "auxiliary_loss_mlp": 0.01073931,
      "balance_loss_clip": 1.05708003,
      "balance_loss_mlp": 1.04965997,
      "epoch": 0.11014579888771983,
      "flos": 24243150366720.0,
      "grad_norm": 2.4937725361201495,
      "language_loss": 0.90836191,
      "learning_rate": 3.932997678675282e-06,
      "loss": 0.93066907,
      "num_input_tokens_seen": 39646670,
      "step": 1832,
      "time_per_iteration": 2.6786489486694336
    },
    {
      "auxiliary_loss_clip": 0.0106111,
      "auxiliary_loss_mlp": 0.01010664,
      "balance_loss_clip": 1.02332854,
      "balance_loss_mlp": 1.00769615,
      "epoch": 0.1102059221403878,
      "flos": 57743965658880.0,
      "grad_norm": 0.7154576595208243,
      "language_loss": 0.59911001,
      "learning_rate": 3.932897678513523e-06,
      "loss": 0.61982775,
      "num_input_tokens_seen": 39712915,
      "step": 1833,
      "time_per_iteration": 3.1802401542663574
    },
    {
      "auxiliary_loss_clip": 0.01167201,
      "auxiliary_loss_mlp": 0.0105502,
      "balance_loss_clip": 1.05312014,
      "balance_loss_mlp": 1.03285873,
      "epoch": 0.11026604539305576,
      "flos": 16795923667200.0,
      "grad_norm": 2.6772934272606923,
      "language_loss": 0.80799395,
      "learning_rate": 3.93279760505609e-06,
      "loss": 0.83021617,
      "num_input_tokens_seen": 39730650,
      "step": 1834,
      "time_per_iteration": 2.591374635696411
    },
    {
      "auxiliary_loss_clip": 0.01141662,
      "auxiliary_loss_mlp": 0.01054827,
      "balance_loss_clip": 1.05557871,
      "balance_loss_mlp": 1.03004324,
      "epoch": 0.11032616864572373,
      "flos": 23988076911360.0,
      "grad_norm": 2.4853906687508247,
      "language_loss": 0.89856094,
      "learning_rate": 3.932697458306779e-06,
      "loss": 0.92052579,
      "num_input_tokens_seen": 39751065,
      "step": 1835,
      "time_per_iteration": 2.742330312728882
    },
    {
      "auxiliary_loss_clip": 0.01131787,
      "auxiliary_loss_mlp": 0.01063812,
      "balance_loss_clip": 1.0524013,
      "balance_loss_mlp": 1.03758645,
      "epoch": 0.1103862918983917,
      "flos": 19683141851520.0,
      "grad_norm": 2.2754442269720023,
      "language_loss": 0.63256055,
      "learning_rate": 3.932597238269386e-06,
      "loss": 0.65451658,
      "num_input_tokens_seen": 39769245,
      "step": 1836,
      "time_per_iteration": 2.6935038566589355
    },
    {
      "auxiliary_loss_clip": 0.01138919,
      "auxiliary_loss_mlp": 0.01061469,
      "balance_loss_clip": 1.05021358,
      "balance_loss_mlp": 1.03954661,
      "epoch": 0.11044641515105967,
      "flos": 32160878340480.0,
      "grad_norm": 1.6726289784191204,
      "language_loss": 0.72792488,
      "learning_rate": 3.932496944947711e-06,
      "loss": 0.74992871,
      "num_input_tokens_seen": 39790830,
      "step": 1837,
      "time_per_iteration": 2.7790510654449463
    },
    {
      "auxiliary_loss_clip": 0.01165472,
      "auxiliary_loss_mlp": 0.01057035,
      "balance_loss_clip": 1.05463088,
      "balance_loss_mlp": 1.03551781,
      "epoch": 0.11050653840372764,
      "flos": 16689233295360.0,
      "grad_norm": 2.027055787194766,
      "language_loss": 0.78489268,
      "learning_rate": 3.93239657834556e-06,
      "loss": 0.8071177,
      "num_input_tokens_seen": 39809475,
      "step": 1838,
      "time_per_iteration": 4.098532438278198
    },
    {
      "auxiliary_loss_clip": 0.01154042,
      "auxiliary_loss_mlp": 0.01062407,
      "balance_loss_clip": 1.05542612,
      "balance_loss_mlp": 1.03970969,
      "epoch": 0.11056666165639562,
      "flos": 21208877902080.0,
      "grad_norm": 2.046221888979386,
      "language_loss": 0.71451718,
      "learning_rate": 3.932296138466736e-06,
      "loss": 0.7366817,
      "num_input_tokens_seen": 39826355,
      "step": 1839,
      "time_per_iteration": 4.205714464187622
    },
    {
      "auxiliary_loss_clip": 0.01187588,
      "auxiliary_loss_mlp": 0.00781104,
      "balance_loss_clip": 1.06183171,
      "balance_loss_mlp": 1.00018013,
      "epoch": 0.11062678490906358,
      "flos": 19165488998400.0,
      "grad_norm": 2.623062836625425,
      "language_loss": 0.79027873,
      "learning_rate": 3.93219562531505e-06,
      "loss": 0.80996567,
      "num_input_tokens_seen": 39845335,
      "step": 1840,
      "time_per_iteration": 2.6023378372192383
    },
    {
      "auxiliary_loss_clip": 0.01156508,
      "auxiliary_loss_mlp": 0.01052512,
      "balance_loss_clip": 1.05206251,
      "balance_loss_mlp": 1.02887261,
      "epoch": 0.11068690816173155,
      "flos": 24895287740160.0,
      "grad_norm": 1.7551987843009527,
      "language_loss": 0.88083529,
      "learning_rate": 3.932095038894311e-06,
      "loss": 0.90292549,
      "num_input_tokens_seen": 39865065,
      "step": 1841,
      "time_per_iteration": 4.3361639976501465
    },
    {
      "auxiliary_loss_clip": 0.01130203,
      "auxiliary_loss_mlp": 0.01067683,
      "balance_loss_clip": 1.05036247,
      "balance_loss_mlp": 1.04453301,
      "epoch": 0.11074703141439952,
      "flos": 16472368932480.0,
      "grad_norm": 3.1603067125494126,
      "language_loss": 0.90521991,
      "learning_rate": 3.931994379208334e-06,
      "loss": 0.92719877,
      "num_input_tokens_seen": 39882780,
      "step": 1842,
      "time_per_iteration": 2.7086760997772217
    },
    {
      "auxiliary_loss_clip": 0.01152506,
      "auxiliary_loss_mlp": 0.01061227,
      "balance_loss_clip": 1.05065131,
      "balance_loss_mlp": 1.03982854,
      "epoch": 0.11080715466706749,
      "flos": 19172420323200.0,
      "grad_norm": 2.112801816568727,
      "language_loss": 0.85845053,
      "learning_rate": 3.931893646260937e-06,
      "loss": 0.88058788,
      "num_input_tokens_seen": 39900295,
      "step": 1843,
      "time_per_iteration": 4.263117790222168
    },
    {
      "auxiliary_loss_clip": 0.01119254,
      "auxiliary_loss_mlp": 0.00783076,
      "balance_loss_clip": 1.05050898,
      "balance_loss_mlp": 1.00012159,
      "epoch": 0.11086727791973545,
      "flos": 27704687109120.0,
      "grad_norm": 1.4511349711086798,
      "language_loss": 0.74735641,
      "learning_rate": 3.931792840055941e-06,
      "loss": 0.76637971,
      "num_input_tokens_seen": 39922075,
      "step": 1844,
      "time_per_iteration": 2.7999000549316406
    },
    {
      "auxiliary_loss_clip": 0.01180395,
      "auxiliary_loss_mlp": 0.01055824,
      "balance_loss_clip": 1.05662274,
      "balance_loss_mlp": 1.03238785,
      "epoch": 0.11092740117240343,
      "flos": 18514967736960.0,
      "grad_norm": 2.017286766878137,
      "language_loss": 0.7566812,
      "learning_rate": 3.931691960597165e-06,
      "loss": 0.77904338,
      "num_input_tokens_seen": 39940115,
      "step": 1845,
      "time_per_iteration": 2.5305535793304443
    },
    {
      "auxiliary_loss_clip": 0.01153403,
      "auxiliary_loss_mlp": 0.01058911,
      "balance_loss_clip": 1.05442989,
      "balance_loss_mlp": 1.03807366,
      "epoch": 0.1109875244250714,
      "flos": 20522446018560.0,
      "grad_norm": 1.9628359583393364,
      "language_loss": 0.75953126,
      "learning_rate": 3.9315910078884375e-06,
      "loss": 0.78165436,
      "num_input_tokens_seen": 39959920,
      "step": 1846,
      "time_per_iteration": 2.719325542449951
    },
    {
      "auxiliary_loss_clip": 0.01173899,
      "auxiliary_loss_mlp": 0.01059369,
      "balance_loss_clip": 1.05823123,
      "balance_loss_mlp": 1.03717244,
      "epoch": 0.11104764767773936,
      "flos": 14098601710080.0,
      "grad_norm": 2.612459533347621,
      "language_loss": 0.8620472,
      "learning_rate": 3.931489981933584e-06,
      "loss": 0.88437986,
      "num_input_tokens_seen": 39974755,
      "step": 1847,
      "time_per_iteration": 2.7705559730529785
    },
    {
      "auxiliary_loss_clip": 0.01181158,
      "auxiliary_loss_mlp": 0.01055145,
      "balance_loss_clip": 1.05562854,
      "balance_loss_mlp": 1.0322808,
      "epoch": 0.11110777093040733,
      "flos": 20594518657920.0,
      "grad_norm": 1.8452742714770096,
      "language_loss": 0.76981926,
      "learning_rate": 3.931388882736438e-06,
      "loss": 0.79218227,
      "num_input_tokens_seen": 39993355,
      "step": 1848,
      "time_per_iteration": 2.605933666229248
    },
    {
      "auxiliary_loss_clip": 0.01172398,
      "auxiliary_loss_mlp": 0.01056349,
      "balance_loss_clip": 1.06262445,
      "balance_loss_mlp": 1.03455794,
      "epoch": 0.11116789418307531,
      "flos": 21870065502720.0,
      "grad_norm": 1.6943193134392138,
      "language_loss": 0.77621841,
      "learning_rate": 3.931287710300832e-06,
      "loss": 0.7985059,
      "num_input_tokens_seen": 40012410,
      "step": 1849,
      "time_per_iteration": 2.678415536880493
    },
    {
      "auxiliary_loss_clip": 0.01138995,
      "auxiliary_loss_mlp": 0.00781122,
      "balance_loss_clip": 1.05277848,
      "balance_loss_mlp": 1.00010324,
      "epoch": 0.11122801743574327,
      "flos": 15523106256000.0,
      "grad_norm": 3.3234972538165066,
      "language_loss": 0.72098577,
      "learning_rate": 3.931186464630601e-06,
      "loss": 0.74018693,
      "num_input_tokens_seen": 40029315,
      "step": 1850,
      "time_per_iteration": 2.7763028144836426
    },
    {
      "auxiliary_loss_clip": 0.01170569,
      "auxiliary_loss_mlp": 0.01061108,
      "balance_loss_clip": 1.05759382,
      "balance_loss_mlp": 1.03874469,
      "epoch": 0.11128814068841124,
      "flos": 14392279307520.0,
      "grad_norm": 2.0638339407107873,
      "language_loss": 0.81499028,
      "learning_rate": 3.931085145729588e-06,
      "loss": 0.83730704,
      "num_input_tokens_seen": 40045765,
      "step": 1851,
      "time_per_iteration": 2.688854694366455
    },
    {
      "auxiliary_loss_clip": 0.01164692,
      "auxiliary_loss_mlp": 0.01061301,
      "balance_loss_clip": 1.05789042,
      "balance_loss_mlp": 1.04027295,
      "epoch": 0.11134826394107922,
      "flos": 16653933204480.0,
      "grad_norm": 2.365035468310974,
      "language_loss": 0.88270009,
      "learning_rate": 3.930983753601631e-06,
      "loss": 0.90496004,
      "num_input_tokens_seen": 40061660,
      "step": 1852,
      "time_per_iteration": 2.659914493560791
    },
    {
      "auxiliary_loss_clip": 0.01166772,
      "auxiliary_loss_mlp": 0.01060698,
      "balance_loss_clip": 1.05489326,
      "balance_loss_mlp": 1.03791702,
      "epoch": 0.11140838719374718,
      "flos": 16690993061760.0,
      "grad_norm": 2.1825610274136054,
      "language_loss": 0.72492862,
      "learning_rate": 3.930882288250578e-06,
      "loss": 0.74720335,
      "num_input_tokens_seen": 40080180,
      "step": 1853,
      "time_per_iteration": 2.7840964794158936
    },
    {
      "auxiliary_loss_clip": 0.01069898,
      "auxiliary_loss_mlp": 0.01019902,
      "balance_loss_clip": 1.02549517,
      "balance_loss_mlp": 1.01701725,
      "epoch": 0.11146851044641515,
      "flos": 60976355587200.0,
      "grad_norm": 0.772231443606995,
      "language_loss": 0.53664064,
      "learning_rate": 3.930780749680273e-06,
      "loss": 0.55753863,
      "num_input_tokens_seen": 40138910,
      "step": 1854,
      "time_per_iteration": 3.089354991912842
    },
    {
      "auxiliary_loss_clip": 0.01159576,
      "auxiliary_loss_mlp": 0.0105585,
      "balance_loss_clip": 1.05390525,
      "balance_loss_mlp": 1.03184092,
      "epoch": 0.11152863369908313,
      "flos": 22193835719040.0,
      "grad_norm": 1.863523240792578,
      "language_loss": 0.8468501,
      "learning_rate": 3.9306791378945705e-06,
      "loss": 0.86900431,
      "num_input_tokens_seen": 40157745,
      "step": 1855,
      "time_per_iteration": 2.7361156940460205
    },
    {
      "auxiliary_loss_clip": 0.01147504,
      "auxiliary_loss_mlp": 0.01064479,
      "balance_loss_clip": 1.05225825,
      "balance_loss_mlp": 1.0424726,
      "epoch": 0.11158875695175109,
      "flos": 19537524115200.0,
      "grad_norm": 2.1217067547931756,
      "language_loss": 0.81187081,
      "learning_rate": 3.9305774528973205e-06,
      "loss": 0.83399057,
      "num_input_tokens_seen": 40175375,
      "step": 1856,
      "time_per_iteration": 2.7158002853393555
    },
    {
      "auxiliary_loss_clip": 0.01168288,
      "auxiliary_loss_mlp": 0.01052259,
      "balance_loss_clip": 1.05843937,
      "balance_loss_mlp": 1.02957392,
      "epoch": 0.11164888020441906,
      "flos": 25442709989760.0,
      "grad_norm": 2.0555738298465314,
      "language_loss": 0.82761133,
      "learning_rate": 3.93047569469238e-06,
      "loss": 0.8498168,
      "num_input_tokens_seen": 40195715,
      "step": 1857,
      "time_per_iteration": 2.647184133529663
    },
    {
      "auxiliary_loss_clip": 0.01144196,
      "auxiliary_loss_mlp": 0.01044915,
      "balance_loss_clip": 1.05255508,
      "balance_loss_mlp": 1.02395833,
      "epoch": 0.11170900345708702,
      "flos": 15632741543040.0,
      "grad_norm": 2.3199985887988914,
      "language_loss": 0.83131742,
      "learning_rate": 3.930373863283608e-06,
      "loss": 0.85320854,
      "num_input_tokens_seen": 40213975,
      "step": 1858,
      "time_per_iteration": 2.726905107498169
    },
    {
      "auxiliary_loss_clip": 0.01134962,
      "auxiliary_loss_mlp": 0.01067658,
      "balance_loss_clip": 1.04900265,
      "balance_loss_mlp": 1.04350638,
      "epoch": 0.111769126709755,
      "flos": 23039424766080.0,
      "grad_norm": 2.0395414997027657,
      "language_loss": 0.9133389,
      "learning_rate": 3.930271958674866e-06,
      "loss": 0.93536508,
      "num_input_tokens_seen": 40233905,
      "step": 1859,
      "time_per_iteration": 3.0006766319274902
    },
    {
      "auxiliary_loss_clip": 0.01167289,
      "auxiliary_loss_mlp": 0.01049698,
      "balance_loss_clip": 1.05445409,
      "balance_loss_mlp": 1.02751315,
      "epoch": 0.11182924996242297,
      "flos": 20850705434880.0,
      "grad_norm": 2.048197345879043,
      "language_loss": 0.81528586,
      "learning_rate": 3.930169980870018e-06,
      "loss": 0.83745575,
      "num_input_tokens_seen": 40252810,
      "step": 1860,
      "time_per_iteration": 2.7216553688049316
    },
    {
      "auxiliary_loss_clip": 0.01154007,
      "auxiliary_loss_mlp": 0.01060885,
      "balance_loss_clip": 1.05737674,
      "balance_loss_mlp": 1.03920078,
      "epoch": 0.11188937321509093,
      "flos": 17455315587840.0,
      "grad_norm": 2.00330439318394,
      "language_loss": 0.75250578,
      "learning_rate": 3.930067929872931e-06,
      "loss": 0.77465475,
      "num_input_tokens_seen": 40272000,
      "step": 1861,
      "time_per_iteration": 2.6878490447998047
    },
    {
      "auxiliary_loss_clip": 0.01177651,
      "auxiliary_loss_mlp": 0.01054452,
      "balance_loss_clip": 1.0565964,
      "balance_loss_mlp": 1.03360212,
      "epoch": 0.11194949646775891,
      "flos": 24095916518400.0,
      "grad_norm": 1.9427039767358767,
      "language_loss": 0.88888168,
      "learning_rate": 3.929965805687474e-06,
      "loss": 0.91120267,
      "num_input_tokens_seen": 40290660,
      "step": 1862,
      "time_per_iteration": 2.615057945251465
    },
    {
      "auxiliary_loss_clip": 0.01164251,
      "auxiliary_loss_mlp": 0.01062894,
      "balance_loss_clip": 1.05994737,
      "balance_loss_mlp": 1.04086459,
      "epoch": 0.11200961972042688,
      "flos": 25153880728320.0,
      "grad_norm": 2.2273555113866847,
      "language_loss": 0.87719512,
      "learning_rate": 3.92986360831752e-06,
      "loss": 0.89946657,
      "num_input_tokens_seen": 40307820,
      "step": 1863,
      "time_per_iteration": 2.6778175830841064
    },
    {
      "auxiliary_loss_clip": 0.01158667,
      "auxiliary_loss_mlp": 0.01055299,
      "balance_loss_clip": 1.05455208,
      "balance_loss_mlp": 1.03071773,
      "epoch": 0.11206974297309484,
      "flos": 21288312829440.0,
      "grad_norm": 2.8013407816012226,
      "language_loss": 0.64245486,
      "learning_rate": 3.929761337766945e-06,
      "loss": 0.66459453,
      "num_input_tokens_seen": 40327430,
      "step": 1864,
      "time_per_iteration": 2.724076509475708
    },
    {
      "auxiliary_loss_clip": 0.01110154,
      "auxiliary_loss_mlp": 0.01047933,
      "balance_loss_clip": 1.04924703,
      "balance_loss_mlp": 1.02672601,
      "epoch": 0.11212986622576282,
      "flos": 18915982151040.0,
      "grad_norm": 2.0303098144917135,
      "language_loss": 0.74043733,
      "learning_rate": 3.929658994039627e-06,
      "loss": 0.7620182,
      "num_input_tokens_seen": 40344545,
      "step": 1865,
      "time_per_iteration": 2.8119356632232666
    },
    {
      "auxiliary_loss_clip": 0.01114683,
      "auxiliary_loss_mlp": 0.01070203,
      "balance_loss_clip": 1.05348182,
      "balance_loss_mlp": 1.04483545,
      "epoch": 0.11218998947843078,
      "flos": 22054754257920.0,
      "grad_norm": 2.7389427033573375,
      "language_loss": 0.84692436,
      "learning_rate": 3.929556577139446e-06,
      "loss": 0.86877316,
      "num_input_tokens_seen": 40362300,
      "step": 1866,
      "time_per_iteration": 2.8022067546844482
    },
    {
      "auxiliary_loss_clip": 0.01092364,
      "auxiliary_loss_mlp": 0.00781014,
      "balance_loss_clip": 1.04227424,
      "balance_loss_mlp": 1.00006938,
      "epoch": 0.11225011273109875,
      "flos": 24571697091840.0,
      "grad_norm": 1.704208120094955,
      "language_loss": 0.8104012,
      "learning_rate": 3.929454087070286e-06,
      "loss": 0.82913494,
      "num_input_tokens_seen": 40384720,
      "step": 1867,
      "time_per_iteration": 2.915989875793457
    },
    {
      "auxiliary_loss_clip": 0.01179505,
      "auxiliary_loss_mlp": 0.01060529,
      "balance_loss_clip": 1.05720687,
      "balance_loss_mlp": 1.03959608,
      "epoch": 0.11231023598376672,
      "flos": 28438665621120.0,
      "grad_norm": 2.0811636681692844,
      "language_loss": 0.86840278,
      "learning_rate": 3.929351523836035e-06,
      "loss": 0.8908031,
      "num_input_tokens_seen": 40404000,
      "step": 1868,
      "time_per_iteration": 2.6855647563934326
    },
    {
      "auxiliary_loss_clip": 0.01161412,
      "auxiliary_loss_mlp": 0.00779977,
      "balance_loss_clip": 1.06005311,
      "balance_loss_mlp": 1.00010097,
      "epoch": 0.1123703592364347,
      "flos": 14426466076800.0,
      "grad_norm": 2.1491178409138376,
      "language_loss": 0.68308532,
      "learning_rate": 3.9292488874405795e-06,
      "loss": 0.70249927,
      "num_input_tokens_seen": 40418665,
      "step": 1869,
      "time_per_iteration": 2.7404487133026123
    },
    {
      "auxiliary_loss_clip": 0.01133783,
      "auxiliary_loss_mlp": 0.01066188,
      "balance_loss_clip": 1.04932964,
      "balance_loss_mlp": 1.04225063,
      "epoch": 0.11243048248910266,
      "flos": 22236282616320.0,
      "grad_norm": 1.5255545896853626,
      "language_loss": 0.76943326,
      "learning_rate": 3.929146177887814e-06,
      "loss": 0.79143298,
      "num_input_tokens_seen": 40437870,
      "step": 1870,
      "time_per_iteration": 2.809734344482422
    },
    {
      "auxiliary_loss_clip": 0.01129358,
      "auxiliary_loss_mlp": 0.01056867,
      "balance_loss_clip": 1.0509038,
      "balance_loss_mlp": 1.03300166,
      "epoch": 0.11249060574177062,
      "flos": 18584167288320.0,
      "grad_norm": 1.8186132867503446,
      "language_loss": 0.76056099,
      "learning_rate": 3.929043395181631e-06,
      "loss": 0.78242326,
      "num_input_tokens_seen": 40455570,
      "step": 1871,
      "time_per_iteration": 2.727161169052124
    },
    {
      "auxiliary_loss_clip": 0.01105662,
      "auxiliary_loss_mlp": 0.01051114,
      "balance_loss_clip": 1.04993379,
      "balance_loss_mlp": 1.03026426,
      "epoch": 0.1125507289944386,
      "flos": 22856567604480.0,
      "grad_norm": 1.9425066802508644,
      "language_loss": 0.81811988,
      "learning_rate": 3.928940539325929e-06,
      "loss": 0.83968765,
      "num_input_tokens_seen": 40473600,
      "step": 1872,
      "time_per_iteration": 2.851868152618408
    },
    {
      "auxiliary_loss_clip": 0.01179923,
      "auxiliary_loss_mlp": 0.01055722,
      "balance_loss_clip": 1.05722499,
      "balance_loss_mlp": 1.03359652,
      "epoch": 0.11261085224710657,
      "flos": 19676390094720.0,
      "grad_norm": 2.186176467187071,
      "language_loss": 0.8361913,
      "learning_rate": 3.9288376103246095e-06,
      "loss": 0.85854775,
      "num_input_tokens_seen": 40490025,
      "step": 1873,
      "time_per_iteration": 2.6668763160705566
    },
    {
      "auxiliary_loss_clip": 0.01144862,
      "auxiliary_loss_mlp": 0.01054726,
      "balance_loss_clip": 1.0525465,
      "balance_loss_mlp": 1.03196871,
      "epoch": 0.11267097549977453,
      "flos": 26063246373120.0,
      "grad_norm": 1.8822875514234196,
      "language_loss": 0.92342389,
      "learning_rate": 3.928734608181575e-06,
      "loss": 0.94541967,
      "num_input_tokens_seen": 40511580,
      "step": 1874,
      "time_per_iteration": 2.700533866882324
    },
    {
      "auxiliary_loss_clip": 0.01140327,
      "auxiliary_loss_mlp": 0.01056402,
      "balance_loss_clip": 1.05100179,
      "balance_loss_mlp": 1.03509891,
      "epoch": 0.11273109875244251,
      "flos": 21068036674560.0,
      "grad_norm": 1.6564425098873434,
      "language_loss": 0.75359404,
      "learning_rate": 3.928631532900729e-06,
      "loss": 0.77556133,
      "num_input_tokens_seen": 40530155,
      "step": 1875,
      "time_per_iteration": 2.7642719745635986
    },
    {
      "auxiliary_loss_clip": 0.01167091,
      "auxiliary_loss_mlp": 0.01055271,
      "balance_loss_clip": 1.05893159,
      "balance_loss_mlp": 1.0348264,
      "epoch": 0.11279122200511048,
      "flos": 27088999061760.0,
      "grad_norm": 2.12758140825061,
      "language_loss": 0.71578634,
      "learning_rate": 3.928528384485984e-06,
      "loss": 0.73800993,
      "num_input_tokens_seen": 40549500,
      "step": 1876,
      "time_per_iteration": 2.8505096435546875
    },
    {
      "auxiliary_loss_clip": 0.01147417,
      "auxiliary_loss_mlp": 0.01054094,
      "balance_loss_clip": 1.05223966,
      "balance_loss_mlp": 1.03200495,
      "epoch": 0.11285134525777844,
      "flos": 20187901722240.0,
      "grad_norm": 1.8103612630164048,
      "language_loss": 0.76795971,
      "learning_rate": 3.9284251629412475e-06,
      "loss": 0.78997481,
      "num_input_tokens_seen": 40567475,
      "step": 1877,
      "time_per_iteration": 2.6972849369049072
    },
    {
      "auxiliary_loss_clip": 0.01168106,
      "auxiliary_loss_mlp": 0.01063056,
      "balance_loss_clip": 1.05518627,
      "balance_loss_mlp": 1.04026341,
      "epoch": 0.11291146851044641,
      "flos": 12458453863680.0,
      "grad_norm": 2.1601834607000368,
      "language_loss": 0.87843502,
      "learning_rate": 3.928321868270436e-06,
      "loss": 0.90074658,
      "num_input_tokens_seen": 40583280,
      "step": 1878,
      "time_per_iteration": 5.6992692947387695
    },
    {
      "auxiliary_loss_clip": 0.01140682,
      "auxiliary_loss_mlp": 0.01054902,
      "balance_loss_clip": 1.05420399,
      "balance_loss_mlp": 1.03333724,
      "epoch": 0.11297159176311439,
      "flos": 23842315520640.0,
      "grad_norm": 2.151084139284284,
      "language_loss": 0.81623232,
      "learning_rate": 3.928218500477466e-06,
      "loss": 0.83818817,
      "num_input_tokens_seen": 40603080,
      "step": 1879,
      "time_per_iteration": 2.8688366413116455
    },
    {
      "auxiliary_loss_clip": 0.01155904,
      "auxiliary_loss_mlp": 0.01059079,
      "balance_loss_clip": 1.05238748,
      "balance_loss_mlp": 1.03609526,
      "epoch": 0.11303171501578235,
      "flos": 29930538124800.0,
      "grad_norm": 1.941623939252122,
      "language_loss": 0.70234305,
      "learning_rate": 3.928115059566259e-06,
      "loss": 0.72449279,
      "num_input_tokens_seen": 40623255,
      "step": 1880,
      "time_per_iteration": 5.567574739456177
    },
    {
      "auxiliary_loss_clip": 0.01155691,
      "auxiliary_loss_mlp": 0.01052309,
      "balance_loss_clip": 1.05585837,
      "balance_loss_mlp": 1.0306015,
      "epoch": 0.11309183826845032,
      "flos": 16180558842240.0,
      "grad_norm": 1.6696082535169858,
      "language_loss": 0.72690225,
      "learning_rate": 3.928011545540734e-06,
      "loss": 0.74898225,
      "num_input_tokens_seen": 40641570,
      "step": 1881,
      "time_per_iteration": 2.792428493499756
    },
    {
      "auxiliary_loss_clip": 0.011425,
      "auxiliary_loss_mlp": 0.00781179,
      "balance_loss_clip": 1.05046606,
      "balance_loss_mlp": 1.00008667,
      "epoch": 0.1131519615211183,
      "flos": 12020702814720.0,
      "grad_norm": 2.2964043184115783,
      "language_loss": 0.74205768,
      "learning_rate": 3.927907958404819e-06,
      "loss": 0.76129448,
      "num_input_tokens_seen": 40658775,
      "step": 1882,
      "time_per_iteration": 4.414916515350342
    },
    {
      "auxiliary_loss_clip": 0.01177281,
      "auxiliary_loss_mlp": 0.01054815,
      "balance_loss_clip": 1.05680335,
      "balance_loss_mlp": 1.03203452,
      "epoch": 0.11321208477378626,
      "flos": 26250125857920.0,
      "grad_norm": 2.4326158086005965,
      "language_loss": 0.7923016,
      "learning_rate": 3.92780429816244e-06,
      "loss": 0.81462252,
      "num_input_tokens_seen": 40679555,
      "step": 1883,
      "time_per_iteration": 2.762615919113159
    },
    {
      "auxiliary_loss_clip": 0.01140926,
      "auxiliary_loss_mlp": 0.01058465,
      "balance_loss_clip": 1.05226314,
      "balance_loss_mlp": 1.03520727,
      "epoch": 0.11327220802645423,
      "flos": 13626376583040.0,
      "grad_norm": 2.2898863699254974,
      "language_loss": 0.77047318,
      "learning_rate": 3.927700564817529e-06,
      "loss": 0.79246712,
      "num_input_tokens_seen": 40697295,
      "step": 1884,
      "time_per_iteration": 2.835468292236328
    },
    {
      "auxiliary_loss_clip": 0.01074478,
      "auxiliary_loss_mlp": 0.01009476,
      "balance_loss_clip": 1.03993821,
      "balance_loss_mlp": 1.00620937,
      "epoch": 0.1133323312791222,
      "flos": 57191802814080.0,
      "grad_norm": 0.8138652948403053,
      "language_loss": 0.55151373,
      "learning_rate": 3.927596758374019e-06,
      "loss": 0.5723533,
      "num_input_tokens_seen": 40758095,
      "step": 1885,
      "time_per_iteration": 3.179532289505005
    },
    {
      "auxiliary_loss_clip": 0.01083888,
      "auxiliary_loss_mlp": 0.01050751,
      "balance_loss_clip": 1.04415166,
      "balance_loss_mlp": 1.02910316,
      "epoch": 0.11339245453179017,
      "flos": 24351708245760.0,
      "grad_norm": 1.9836288003076585,
      "language_loss": 0.90384823,
      "learning_rate": 3.927492878835848e-06,
      "loss": 0.92519462,
      "num_input_tokens_seen": 40777140,
      "step": 1886,
      "time_per_iteration": 3.038928747177124
    },
    {
      "auxiliary_loss_clip": 0.01116325,
      "auxiliary_loss_mlp": 0.01057697,
      "balance_loss_clip": 1.05137897,
      "balance_loss_mlp": 1.03634632,
      "epoch": 0.11345257778445814,
      "flos": 22670693700480.0,
      "grad_norm": 2.0132756022974023,
      "language_loss": 0.84852886,
      "learning_rate": 3.927388926206953e-06,
      "loss": 0.87026906,
      "num_input_tokens_seen": 40797505,
      "step": 1887,
      "time_per_iteration": 3.178863048553467
    },
    {
      "auxiliary_loss_clip": 0.01136567,
      "auxiliary_loss_mlp": 0.01056557,
      "balance_loss_clip": 1.05091035,
      "balance_loss_mlp": 1.03549314,
      "epoch": 0.11351270103712612,
      "flos": 20988242611200.0,
      "grad_norm": 2.847610033990257,
      "language_loss": 0.75826252,
      "learning_rate": 3.927284900491277e-06,
      "loss": 0.78019381,
      "num_input_tokens_seen": 40812970,
      "step": 1888,
      "time_per_iteration": 2.7349846363067627
    },
    {
      "auxiliary_loss_clip": 0.0113463,
      "auxiliary_loss_mlp": 0.01062359,
      "balance_loss_clip": 1.05614805,
      "balance_loss_mlp": 1.03892243,
      "epoch": 0.11357282428979408,
      "flos": 37347923600640.0,
      "grad_norm": 2.0598279187313624,
      "language_loss": 0.68104899,
      "learning_rate": 3.927180801692764e-06,
      "loss": 0.7030189,
      "num_input_tokens_seen": 40837745,
      "step": 1889,
      "time_per_iteration": 3.144444465637207
    },
    {
      "auxiliary_loss_clip": 0.01177206,
      "auxiliary_loss_mlp": 0.01049162,
      "balance_loss_clip": 1.05653095,
      "balance_loss_mlp": 1.02694094,
      "epoch": 0.11363294754246205,
      "flos": 21757018423680.0,
      "grad_norm": 1.7896678692754837,
      "language_loss": 0.83947051,
      "learning_rate": 3.927076629815362e-06,
      "loss": 0.86173415,
      "num_input_tokens_seen": 40856490,
      "step": 1890,
      "time_per_iteration": 2.73126482963562
    },
    {
      "auxiliary_loss_clip": 0.01145149,
      "auxiliary_loss_mlp": 0.01056017,
      "balance_loss_clip": 1.05039728,
      "balance_loss_mlp": 1.03395164,
      "epoch": 0.11369307079513001,
      "flos": 22601637803520.0,
      "grad_norm": 2.1678723202845256,
      "language_loss": 0.64663875,
      "learning_rate": 3.926972384863022e-06,
      "loss": 0.66865045,
      "num_input_tokens_seen": 40874070,
      "step": 1891,
      "time_per_iteration": 2.7474160194396973
    },
    {
      "auxiliary_loss_clip": 0.01145505,
      "auxiliary_loss_mlp": 0.01049015,
      "balance_loss_clip": 1.05395687,
      "balance_loss_mlp": 1.02773631,
      "epoch": 0.11375319404779799,
      "flos": 21944257044480.0,
      "grad_norm": 2.126575023047711,
      "language_loss": 0.87889415,
      "learning_rate": 3.9268680668396956e-06,
      "loss": 0.90083933,
      "num_input_tokens_seen": 40892425,
      "step": 1892,
      "time_per_iteration": 2.795269250869751
    },
    {
      "auxiliary_loss_clip": 0.01119535,
      "auxiliary_loss_mlp": 0.01079586,
      "balance_loss_clip": 1.05541015,
      "balance_loss_mlp": 1.05461168,
      "epoch": 0.11381331730046595,
      "flos": 26395456285440.0,
      "grad_norm": 3.1806920305576973,
      "language_loss": 0.72902197,
      "learning_rate": 3.926763675749339e-06,
      "loss": 0.75101316,
      "num_input_tokens_seen": 40912190,
      "step": 1893,
      "time_per_iteration": 2.890289306640625
    },
    {
      "auxiliary_loss_clip": 0.01175698,
      "auxiliary_loss_mlp": 0.0106591,
      "balance_loss_clip": 1.05438137,
      "balance_loss_mlp": 1.04290223,
      "epoch": 0.11387344055313392,
      "flos": 23804716959360.0,
      "grad_norm": 1.8842571229841023,
      "language_loss": 0.79247093,
      "learning_rate": 3.92665921159591e-06,
      "loss": 0.81488699,
      "num_input_tokens_seen": 40928395,
      "step": 1894,
      "time_per_iteration": 2.6820743083953857
    },
    {
      "auxiliary_loss_clip": 0.01150233,
      "auxiliary_loss_mlp": 0.01061956,
      "balance_loss_clip": 1.05356526,
      "balance_loss_mlp": 1.03944933,
      "epoch": 0.1139335638058019,
      "flos": 34522865902080.0,
      "grad_norm": 3.429237983174195,
      "language_loss": 0.79718482,
      "learning_rate": 3.926554674383371e-06,
      "loss": 0.81930667,
      "num_input_tokens_seen": 40946555,
      "step": 1895,
      "time_per_iteration": 2.829946994781494
    },
    {
      "auxiliary_loss_clip": 0.01075529,
      "auxiliary_loss_mlp": 0.01018518,
      "balance_loss_clip": 1.03062391,
      "balance_loss_mlp": 1.0155375,
      "epoch": 0.11399368705846986,
      "flos": 70587811520640.0,
      "grad_norm": 0.8041110638842961,
      "language_loss": 0.63357508,
      "learning_rate": 3.926450064115686e-06,
      "loss": 0.65451556,
      "num_input_tokens_seen": 41004910,
      "step": 1896,
      "time_per_iteration": 3.3087315559387207
    },
    {
      "auxiliary_loss_clip": 0.01147265,
      "auxiliary_loss_mlp": 0.0106086,
      "balance_loss_clip": 1.05560398,
      "balance_loss_mlp": 1.03663635,
      "epoch": 0.11405381031113783,
      "flos": 21324259365120.0,
      "grad_norm": 1.5952307342327186,
      "language_loss": 0.85055745,
      "learning_rate": 3.926345380796821e-06,
      "loss": 0.8726387,
      "num_input_tokens_seen": 41026385,
      "step": 1897,
      "time_per_iteration": 2.8522274494171143
    },
    {
      "auxiliary_loss_clip": 0.0117836,
      "auxiliary_loss_mlp": 0.00780276,
      "balance_loss_clip": 1.05591989,
      "balance_loss_mlp": 1.0001986,
      "epoch": 0.11411393356380581,
      "flos": 19719627091200.0,
      "grad_norm": 3.3624139627125587,
      "language_loss": 0.79675245,
      "learning_rate": 3.9262406244307465e-06,
      "loss": 0.81633884,
      "num_input_tokens_seen": 41045315,
      "step": 1898,
      "time_per_iteration": 2.760057210922241
    },
    {
      "auxiliary_loss_clip": 0.01115338,
      "auxiliary_loss_mlp": 0.01064417,
      "balance_loss_clip": 1.04594529,
      "balance_loss_mlp": 1.03965724,
      "epoch": 0.11417405681647377,
      "flos": 17530440883200.0,
      "grad_norm": 2.0191769665152903,
      "language_loss": 0.73251313,
      "learning_rate": 3.926135795021435e-06,
      "loss": 0.75431061,
      "num_input_tokens_seen": 41063390,
      "step": 1899,
      "time_per_iteration": 2.7363204956054688
    },
    {
      "auxiliary_loss_clip": 0.01042449,
      "auxiliary_loss_mlp": 0.01003313,
      "balance_loss_clip": 1.03643703,
      "balance_loss_mlp": 1.0003922,
      "epoch": 0.11423418006914174,
      "flos": 59674666619520.0,
      "grad_norm": 0.9089505356695228,
      "language_loss": 0.63434029,
      "learning_rate": 3.92603089257286e-06,
      "loss": 0.65479791,
      "num_input_tokens_seen": 41124180,
      "step": 1900,
      "time_per_iteration": 3.2045955657958984
    },
    {
      "auxiliary_loss_clip": 0.01113626,
      "auxiliary_loss_mlp": 0.01066815,
      "balance_loss_clip": 1.04929233,
      "balance_loss_mlp": 1.04378414,
      "epoch": 0.1142943033218097,
      "flos": 22963114321920.0,
      "grad_norm": 1.577500478750639,
      "language_loss": 0.77943742,
      "learning_rate": 3.925925917089001e-06,
      "loss": 0.80124187,
      "num_input_tokens_seen": 41143485,
      "step": 1901,
      "time_per_iteration": 2.745089530944824
    },
    {
      "auxiliary_loss_clip": 0.01171621,
      "auxiliary_loss_mlp": 0.01057834,
      "balance_loss_clip": 1.05803061,
      "balance_loss_mlp": 1.0359118,
      "epoch": 0.11435442657447768,
      "flos": 18256267008000.0,
      "grad_norm": 2.175933638179557,
      "language_loss": 0.84158623,
      "learning_rate": 3.925820868573839e-06,
      "loss": 0.86388075,
      "num_input_tokens_seen": 41161695,
      "step": 1902,
      "time_per_iteration": 2.6433799266815186
    },
    {
      "auxiliary_loss_clip": 0.01159941,
      "auxiliary_loss_mlp": 0.01056662,
      "balance_loss_clip": 1.05280399,
      "balance_loss_mlp": 1.03122306,
      "epoch": 0.11441454982714565,
      "flos": 24061191045120.0,
      "grad_norm": 1.7702735053047673,
      "language_loss": 0.77720451,
      "learning_rate": 3.925715747031356e-06,
      "loss": 0.79937053,
      "num_input_tokens_seen": 41181715,
      "step": 1903,
      "time_per_iteration": 2.6385905742645264
    },
    {
      "auxiliary_loss_clip": 0.01145143,
      "auxiliary_loss_mlp": 0.0104196,
      "balance_loss_clip": 1.05293322,
      "balance_loss_mlp": 1.02174175,
      "epoch": 0.11447467307981361,
      "flos": 25337707557120.0,
      "grad_norm": 2.212790565732917,
      "language_loss": 0.75751555,
      "learning_rate": 3.925610552465539e-06,
      "loss": 0.77938658,
      "num_input_tokens_seen": 41201770,
      "step": 1904,
      "time_per_iteration": 2.632152557373047
    },
    {
      "auxiliary_loss_clip": 0.01149375,
      "auxiliary_loss_mlp": 0.01056532,
      "balance_loss_clip": 1.05207586,
      "balance_loss_mlp": 1.03279781,
      "epoch": 0.11453479633248159,
      "flos": 21726063878400.0,
      "grad_norm": 2.4422699353972006,
      "language_loss": 0.91853034,
      "learning_rate": 3.9255052848803764e-06,
      "loss": 0.94058943,
      "num_input_tokens_seen": 41220590,
      "step": 1905,
      "time_per_iteration": 2.7421486377716064
    },
    {
      "auxiliary_loss_clip": 0.01161686,
      "auxiliary_loss_mlp": 0.01050264,
      "balance_loss_clip": 1.04978943,
      "balance_loss_mlp": 1.02612448,
      "epoch": 0.11459491958514956,
      "flos": 12969714096000.0,
      "grad_norm": 2.5117992419356066,
      "language_loss": 0.77484202,
      "learning_rate": 3.925399944279861e-06,
      "loss": 0.79696143,
      "num_input_tokens_seen": 41237250,
      "step": 1906,
      "time_per_iteration": 2.69333553314209
    },
    {
      "auxiliary_loss_clip": 0.0117911,
      "auxiliary_loss_mlp": 0.01055129,
      "balance_loss_clip": 1.05697322,
      "balance_loss_mlp": 1.03222847,
      "epoch": 0.11465504283781752,
      "flos": 22711273090560.0,
      "grad_norm": 2.0720467666322113,
      "language_loss": 0.81739306,
      "learning_rate": 3.925294530667986e-06,
      "loss": 0.83973539,
      "num_input_tokens_seen": 41256680,
      "step": 1907,
      "time_per_iteration": 2.6531317234039307
    },
    {
      "auxiliary_loss_clip": 0.0113647,
      "auxiliary_loss_mlp": 0.01065473,
      "balance_loss_clip": 1.05235374,
      "balance_loss_mlp": 1.04227471,
      "epoch": 0.1147151660904855,
      "flos": 23398387332480.0,
      "grad_norm": 2.1769364553121293,
      "language_loss": 0.84901214,
      "learning_rate": 3.92518904404875e-06,
      "loss": 0.87103164,
      "num_input_tokens_seen": 41270955,
      "step": 1908,
      "time_per_iteration": 2.8768258094787598
    },
    {
      "auxiliary_loss_clip": 0.01029536,
      "auxiliary_loss_mlp": 0.01020856,
      "balance_loss_clip": 1.02524137,
      "balance_loss_mlp": 1.01694632,
      "epoch": 0.11477528934315347,
      "flos": 63011843498880.0,
      "grad_norm": 0.9197306473097341,
      "language_loss": 0.61072773,
      "learning_rate": 3.925083484426153e-06,
      "loss": 0.63123173,
      "num_input_tokens_seen": 41319180,
      "step": 1909,
      "time_per_iteration": 3.0845727920532227
    },
    {
      "auxiliary_loss_clip": 0.01182744,
      "auxiliary_loss_mlp": 0.01054075,
      "balance_loss_clip": 1.06014562,
      "balance_loss_mlp": 1.03219986,
      "epoch": 0.11483541259582143,
      "flos": 16325601960960.0,
      "grad_norm": 7.319166590530674,
      "language_loss": 0.79170966,
      "learning_rate": 3.924977851804197e-06,
      "loss": 0.81407785,
      "num_input_tokens_seen": 41337480,
      "step": 1910,
      "time_per_iteration": 2.708704710006714
    },
    {
      "auxiliary_loss_clip": 0.01156489,
      "auxiliary_loss_mlp": 0.01052406,
      "balance_loss_clip": 1.0580864,
      "balance_loss_mlp": 1.03029275,
      "epoch": 0.1148955358484894,
      "flos": 21580410228480.0,
      "grad_norm": 2.117911712245717,
      "language_loss": 0.7702589,
      "learning_rate": 3.9248721461868875e-06,
      "loss": 0.79234779,
      "num_input_tokens_seen": 41354650,
      "step": 1911,
      "time_per_iteration": 2.7597720623016357
    },
    {
      "auxiliary_loss_clip": 0.01159986,
      "auxiliary_loss_mlp": 0.01054599,
      "balance_loss_clip": 1.05726957,
      "balance_loss_mlp": 1.03227139,
      "epoch": 0.11495565910115738,
      "flos": 27673696650240.0,
      "grad_norm": 1.677508784227342,
      "language_loss": 0.79177421,
      "learning_rate": 3.9247663675782336e-06,
      "loss": 0.81392002,
      "num_input_tokens_seen": 41376935,
      "step": 1912,
      "time_per_iteration": 2.8143310546875
    },
    {
      "auxiliary_loss_clip": 0.01183047,
      "auxiliary_loss_mlp": 0.00779659,
      "balance_loss_clip": 1.06065917,
      "balance_loss_mlp": 1.00014925,
      "epoch": 0.11501578235382534,
      "flos": 20632368614400.0,
      "grad_norm": 2.291252405113977,
      "language_loss": 0.77942276,
      "learning_rate": 3.924660515982246e-06,
      "loss": 0.79904979,
      "num_input_tokens_seen": 41396105,
      "step": 1913,
      "time_per_iteration": 2.696430206298828
    },
    {
      "auxiliary_loss_clip": 0.01166892,
      "auxiliary_loss_mlp": 0.01052769,
      "balance_loss_clip": 1.05442226,
      "balance_loss_mlp": 1.02953506,
      "epoch": 0.1150759056064933,
      "flos": 19829046896640.0,
      "grad_norm": 1.8145547055361753,
      "language_loss": 0.7003395,
      "learning_rate": 3.924554591402939e-06,
      "loss": 0.72253609,
      "num_input_tokens_seen": 41415600,
      "step": 1914,
      "time_per_iteration": 2.739251136779785
    },
    {
      "auxiliary_loss_clip": 0.01007182,
      "auxiliary_loss_mlp": 0.01004682,
      "balance_loss_clip": 1.02677619,
      "balance_loss_mlp": 1.00191641,
      "epoch": 0.11513602885916129,
      "flos": 70045776311040.0,
      "grad_norm": 0.7558771871458172,
      "language_loss": 0.61059874,
      "learning_rate": 3.92444859384433e-06,
      "loss": 0.6307174,
      "num_input_tokens_seen": 41478760,
      "step": 1915,
      "time_per_iteration": 3.56019926071167
    },
    {
      "auxiliary_loss_clip": 0.01166434,
      "auxiliary_loss_mlp": 0.01058573,
      "balance_loss_clip": 1.05994964,
      "balance_loss_mlp": 1.03595936,
      "epoch": 0.11519615211182925,
      "flos": 15741730385280.0,
      "grad_norm": 2.437201506258279,
      "language_loss": 0.93116963,
      "learning_rate": 3.924342523310436e-06,
      "loss": 0.95341969,
      "num_input_tokens_seen": 41495720,
      "step": 1916,
      "time_per_iteration": 3.244772434234619
    },
    {
      "auxiliary_loss_clip": 0.01161132,
      "auxiliary_loss_mlp": 0.01059827,
      "balance_loss_clip": 1.05798697,
      "balance_loss_mlp": 1.03470993,
      "epoch": 0.11525627536449722,
      "flos": 20667632791680.0,
      "grad_norm": 1.8909260082350545,
      "language_loss": 0.72560197,
      "learning_rate": 3.9242363798052806e-06,
      "loss": 0.74781156,
      "num_input_tokens_seen": 41513585,
      "step": 1917,
      "time_per_iteration": 4.502236843109131
    },
    {
      "auxiliary_loss_clip": 0.01138773,
      "auxiliary_loss_mlp": 0.0104964,
      "balance_loss_clip": 1.05739903,
      "balance_loss_mlp": 1.02700245,
      "epoch": 0.1153163986171652,
      "flos": 20303283185280.0,
      "grad_norm": 9.147356795176979,
      "language_loss": 0.74213129,
      "learning_rate": 3.92413016333289e-06,
      "loss": 0.76401544,
      "num_input_tokens_seen": 41533390,
      "step": 1918,
      "time_per_iteration": 4.344711065292358
    },
    {
      "auxiliary_loss_clip": 0.0114898,
      "auxiliary_loss_mlp": 0.010469,
      "balance_loss_clip": 1.05532503,
      "balance_loss_mlp": 1.02450073,
      "epoch": 0.11537652186983316,
      "flos": 17639321984640.0,
      "grad_norm": 3.182152136597976,
      "language_loss": 0.86367452,
      "learning_rate": 3.92402387389729e-06,
      "loss": 0.88563335,
      "num_input_tokens_seen": 41551015,
      "step": 1919,
      "time_per_iteration": 4.540036201477051
    },
    {
      "auxiliary_loss_clip": 0.01134044,
      "auxiliary_loss_mlp": 0.01067867,
      "balance_loss_clip": 1.0496366,
      "balance_loss_mlp": 1.04172444,
      "epoch": 0.11543664512250112,
      "flos": 21069401391360.0,
      "grad_norm": 1.93595243799445,
      "language_loss": 0.86735415,
      "learning_rate": 3.923917511502512e-06,
      "loss": 0.8893733,
      "num_input_tokens_seen": 41568055,
      "step": 1920,
      "time_per_iteration": 2.7719242572784424
    },
    {
      "auxiliary_loss_clip": 0.011686,
      "auxiliary_loss_mlp": 0.010528,
      "balance_loss_clip": 1.0593946,
      "balance_loss_mlp": 1.0302341,
      "epoch": 0.11549676837516909,
      "flos": 22747542848640.0,
      "grad_norm": 4.512761907267092,
      "language_loss": 0.79294932,
      "learning_rate": 3.923811076152589e-06,
      "loss": 0.81516337,
      "num_input_tokens_seen": 41587435,
      "step": 1921,
      "time_per_iteration": 2.798673629760742
    },
    {
      "auxiliary_loss_clip": 0.01174604,
      "auxiliary_loss_mlp": 0.01063526,
      "balance_loss_clip": 1.05685806,
      "balance_loss_mlp": 1.04007721,
      "epoch": 0.11555689162783707,
      "flos": 19168972617600.0,
      "grad_norm": 2.4057040360661484,
      "language_loss": 0.78464305,
      "learning_rate": 3.923704567851557e-06,
      "loss": 0.80702436,
      "num_input_tokens_seen": 41604975,
      "step": 1922,
      "time_per_iteration": 4.352341651916504
    },
    {
      "auxiliary_loss_clip": 0.01092284,
      "auxiliary_loss_mlp": 0.01064602,
      "balance_loss_clip": 1.04645681,
      "balance_loss_mlp": 1.04229808,
      "epoch": 0.11561701488050503,
      "flos": 24572056227840.0,
      "grad_norm": 1.8560991769949675,
      "language_loss": 0.84293079,
      "learning_rate": 3.923597986603456e-06,
      "loss": 0.86449969,
      "num_input_tokens_seen": 41626155,
      "step": 1923,
      "time_per_iteration": 3.2956740856170654
    },
    {
      "auxiliary_loss_clip": 0.01171957,
      "auxiliary_loss_mlp": 0.01056739,
      "balance_loss_clip": 1.0600003,
      "balance_loss_mlp": 1.03317094,
      "epoch": 0.115677138133173,
      "flos": 17092546179840.0,
      "grad_norm": 1.944851076041885,
      "language_loss": 0.80890471,
      "learning_rate": 3.9234913324123264e-06,
      "loss": 0.83119166,
      "num_input_tokens_seen": 41644805,
      "step": 1924,
      "time_per_iteration": 3.0939247608184814
    },
    {
      "auxiliary_loss_clip": 0.01055916,
      "auxiliary_loss_mlp": 0.01027131,
      "balance_loss_clip": 1.03045607,
      "balance_loss_mlp": 1.02436543,
      "epoch": 0.11573726138584098,
      "flos": 62703875266560.0,
      "grad_norm": 0.8171642061509322,
      "language_loss": 0.61196578,
      "learning_rate": 3.923384605282212e-06,
      "loss": 0.63279623,
      "num_input_tokens_seen": 41709345,
      "step": 1925,
      "time_per_iteration": 3.3765265941619873
    },
    {
      "auxiliary_loss_clip": 0.01155845,
      "auxiliary_loss_mlp": 0.01079328,
      "balance_loss_clip": 1.05374098,
      "balance_loss_mlp": 1.0549382,
      "epoch": 0.11579738463850894,
      "flos": 22601135013120.0,
      "grad_norm": 1.7772533553430212,
      "language_loss": 0.74766397,
      "learning_rate": 3.923277805217161e-06,
      "loss": 0.77001572,
      "num_input_tokens_seen": 41730210,
      "step": 1926,
      "time_per_iteration": 2.754974126815796
    },
    {
      "auxiliary_loss_clip": 0.01116228,
      "auxiliary_loss_mlp": 0.00781701,
      "balance_loss_clip": 1.04683304,
      "balance_loss_mlp": 1.00016665,
      "epoch": 0.11585750789117691,
      "flos": 21726135705600.0,
      "grad_norm": 4.731879086182685,
      "language_loss": 0.71978599,
      "learning_rate": 3.923170932221222e-06,
      "loss": 0.7387653,
      "num_input_tokens_seen": 41750270,
      "step": 1927,
      "time_per_iteration": 2.9454004764556885
    },
    {
      "auxiliary_loss_clip": 0.01137955,
      "auxiliary_loss_mlp": 0.01058796,
      "balance_loss_clip": 1.05250621,
      "balance_loss_mlp": 1.03572917,
      "epoch": 0.11591763114384489,
      "flos": 26287544851200.0,
      "grad_norm": 1.5938674022456252,
      "language_loss": 0.86854041,
      "learning_rate": 3.92306398629845e-06,
      "loss": 0.89050794,
      "num_input_tokens_seen": 41772975,
      "step": 1928,
      "time_per_iteration": 2.832750082015991
    },
    {
      "auxiliary_loss_clip": 0.01129041,
      "auxiliary_loss_mlp": 0.01060836,
      "balance_loss_clip": 1.05032003,
      "balance_loss_mlp": 1.03706551,
      "epoch": 0.11597775439651285,
      "flos": 23000461488000.0,
      "grad_norm": 1.6639520350020578,
      "language_loss": 0.77450585,
      "learning_rate": 3.922956967452898e-06,
      "loss": 0.79640466,
      "num_input_tokens_seen": 41791765,
      "step": 1929,
      "time_per_iteration": 2.7876811027526855
    },
    {
      "auxiliary_loss_clip": 0.01176887,
      "auxiliary_loss_mlp": 0.01063611,
      "balance_loss_clip": 1.05667901,
      "balance_loss_mlp": 1.0424509,
      "epoch": 0.11603787764918082,
      "flos": 31941715507200.0,
      "grad_norm": 1.8085677541874856,
      "language_loss": 0.76831949,
      "learning_rate": 3.922849875688626e-06,
      "loss": 0.79072452,
      "num_input_tokens_seen": 41815615,
      "step": 1930,
      "time_per_iteration": 2.819934844970703
    },
    {
      "auxiliary_loss_clip": 0.01145781,
      "auxiliary_loss_mlp": 0.01054046,
      "balance_loss_clip": 1.05066586,
      "balance_loss_mlp": 1.03165817,
      "epoch": 0.1160980009018488,
      "flos": 22271654534400.0,
      "grad_norm": 1.9434791543130712,
      "language_loss": 0.72291863,
      "learning_rate": 3.922742711009693e-06,
      "loss": 0.74491692,
      "num_input_tokens_seen": 41834810,
      "step": 1931,
      "time_per_iteration": 2.8078088760375977
    },
    {
      "auxiliary_loss_clip": 0.01146409,
      "auxiliary_loss_mlp": 0.01061336,
      "balance_loss_clip": 1.05090261,
      "balance_loss_mlp": 1.03575325,
      "epoch": 0.11615812415451676,
      "flos": 22783633038720.0,
      "grad_norm": 1.7378937044391531,
      "language_loss": 0.8222791,
      "learning_rate": 3.922635473420164e-06,
      "loss": 0.8443566,
      "num_input_tokens_seen": 41854975,
      "step": 1932,
      "time_per_iteration": 2.7495200634002686
    },
    {
      "auxiliary_loss_clip": 0.01030493,
      "auxiliary_loss_mlp": 0.01018834,
      "balance_loss_clip": 1.02184403,
      "balance_loss_mlp": 1.01556778,
      "epoch": 0.11621824740718473,
      "flos": 67146096107520.0,
      "grad_norm": 0.7669378012870447,
      "language_loss": 0.61050332,
      "learning_rate": 3.922528162924105e-06,
      "loss": 0.63099658,
      "num_input_tokens_seen": 41911105,
      "step": 1933,
      "time_per_iteration": 3.256678581237793
    },
    {
      "auxiliary_loss_clip": 0.01108577,
      "auxiliary_loss_mlp": 0.00780156,
      "balance_loss_clip": 1.04764509,
      "balance_loss_mlp": 1.00006175,
      "epoch": 0.11627837065985269,
      "flos": 20375930442240.0,
      "grad_norm": 2.830760437639296,
      "language_loss": 0.85790741,
      "learning_rate": 3.922420779525586e-06,
      "loss": 0.8767947,
      "num_input_tokens_seen": 41931750,
      "step": 1934,
      "time_per_iteration": 2.9144253730773926
    },
    {
      "auxiliary_loss_clip": 0.01117671,
      "auxiliary_loss_mlp": 0.01059839,
      "balance_loss_clip": 1.04929256,
      "balance_loss_mlp": 1.03453088,
      "epoch": 0.11633849391252067,
      "flos": 21725812483200.0,
      "grad_norm": 2.625764216143105,
      "language_loss": 0.66222906,
      "learning_rate": 3.9223133232286776e-06,
      "loss": 0.68400419,
      "num_input_tokens_seen": 41949400,
      "step": 1935,
      "time_per_iteration": 2.867152452468872
    },
    {
      "auxiliary_loss_clip": 0.01183991,
      "auxiliary_loss_mlp": 0.01052492,
      "balance_loss_clip": 1.05868936,
      "balance_loss_mlp": 1.03111792,
      "epoch": 0.11639861716518864,
      "flos": 18805341283200.0,
      "grad_norm": 2.025938843377603,
      "language_loss": 0.75678742,
      "learning_rate": 3.922205794037456e-06,
      "loss": 0.77915227,
      "num_input_tokens_seen": 41968100,
      "step": 1936,
      "time_per_iteration": 2.7282185554504395
    },
    {
      "auxiliary_loss_clip": 0.01179718,
      "auxiliary_loss_mlp": 0.01049532,
      "balance_loss_clip": 1.05632091,
      "balance_loss_mlp": 1.02639306,
      "epoch": 0.1164587404178566,
      "flos": 21214983214080.0,
      "grad_norm": 2.0032002399718905,
      "language_loss": 0.84086847,
      "learning_rate": 3.922098191955998e-06,
      "loss": 0.86316097,
      "num_input_tokens_seen": 41986375,
      "step": 1937,
      "time_per_iteration": 2.715386152267456
    },
    {
      "auxiliary_loss_clip": 0.01152084,
      "auxiliary_loss_mlp": 0.01048961,
      "balance_loss_clip": 1.05258632,
      "balance_loss_mlp": 1.0268234,
      "epoch": 0.11651886367052458,
      "flos": 27818632028160.0,
      "grad_norm": 3.0485930101216607,
      "language_loss": 0.7617709,
      "learning_rate": 3.921990516988384e-06,
      "loss": 0.78378135,
      "num_input_tokens_seen": 42006055,
      "step": 1938,
      "time_per_iteration": 2.7624804973602295
    },
    {
      "auxiliary_loss_clip": 0.01182576,
      "auxiliary_loss_mlp": 0.01055104,
      "balance_loss_clip": 1.05742419,
      "balance_loss_mlp": 1.03250146,
      "epoch": 0.11657898692319255,
      "flos": 22889569224960.0,
      "grad_norm": 1.7682499083089231,
      "language_loss": 0.79677606,
      "learning_rate": 3.921882769138696e-06,
      "loss": 0.81915289,
      "num_input_tokens_seen": 42024995,
      "step": 1939,
      "time_per_iteration": 2.71458101272583
    },
    {
      "auxiliary_loss_clip": 0.01148291,
      "auxiliary_loss_mlp": 0.01057951,
      "balance_loss_clip": 1.05209351,
      "balance_loss_mlp": 1.03508627,
      "epoch": 0.11663911017586051,
      "flos": 24315905364480.0,
      "grad_norm": 2.2281245193552475,
      "language_loss": 0.85916591,
      "learning_rate": 3.9217749484110215e-06,
      "loss": 0.88122833,
      "num_input_tokens_seen": 42042640,
      "step": 1940,
      "time_per_iteration": 2.7322728633880615
    },
    {
      "auxiliary_loss_clip": 0.01153746,
      "auxiliary_loss_mlp": 0.01056301,
      "balance_loss_clip": 1.05659437,
      "balance_loss_mlp": 1.03548717,
      "epoch": 0.11669923342852849,
      "flos": 42340152470400.0,
      "grad_norm": 1.4952807995381137,
      "language_loss": 0.75590646,
      "learning_rate": 3.921667054809449e-06,
      "loss": 0.77800703,
      "num_input_tokens_seen": 42067005,
      "step": 1941,
      "time_per_iteration": 2.9211390018463135
    },
    {
      "auxiliary_loss_clip": 0.01149585,
      "auxiliary_loss_mlp": 0.00780203,
      "balance_loss_clip": 1.05181897,
      "balance_loss_mlp": 1.00006557,
      "epoch": 0.11675935668119646,
      "flos": 14642288945280.0,
      "grad_norm": 2.277225749463833,
      "language_loss": 0.88847101,
      "learning_rate": 3.921559088338068e-06,
      "loss": 0.90776885,
      "num_input_tokens_seen": 42082295,
      "step": 1942,
      "time_per_iteration": 2.7145469188690186
    },
    {
      "auxiliary_loss_clip": 0.01165183,
      "auxiliary_loss_mlp": 0.01056257,
      "balance_loss_clip": 1.05553317,
      "balance_loss_mlp": 1.03552663,
      "epoch": 0.11681947993386442,
      "flos": 35116470063360.0,
      "grad_norm": 1.6547450593003057,
      "language_loss": 0.67979252,
      "learning_rate": 3.921451049000975e-06,
      "loss": 0.70200694,
      "num_input_tokens_seen": 42105295,
      "step": 1943,
      "time_per_iteration": 2.789701461791992
    },
    {
      "auxiliary_loss_clip": 0.01153022,
      "auxiliary_loss_mlp": 0.01047648,
      "balance_loss_clip": 1.05515063,
      "balance_loss_mlp": 1.02591634,
      "epoch": 0.11687960318653239,
      "flos": 38983259024640.0,
      "grad_norm": 1.9817763000300312,
      "language_loss": 0.69831288,
      "learning_rate": 3.921342936802265e-06,
      "loss": 0.72031963,
      "num_input_tokens_seen": 42125520,
      "step": 1944,
      "time_per_iteration": 2.827150583267212
    },
    {
      "auxiliary_loss_clip": 0.01155915,
      "auxiliary_loss_mlp": 0.01051888,
      "balance_loss_clip": 1.05038309,
      "balance_loss_mlp": 1.03158641,
      "epoch": 0.11693972643920036,
      "flos": 25994980575360.0,
      "grad_norm": 1.4963028532298175,
      "language_loss": 0.82662582,
      "learning_rate": 3.921234751746038e-06,
      "loss": 0.84870374,
      "num_input_tokens_seen": 42146335,
      "step": 1945,
      "time_per_iteration": 2.7190194129943848
    },
    {
      "auxiliary_loss_clip": 0.01137101,
      "auxiliary_loss_mlp": 0.01062082,
      "balance_loss_clip": 1.04682803,
      "balance_loss_mlp": 1.04005265,
      "epoch": 0.11699984969186833,
      "flos": 27272107618560.0,
      "grad_norm": 2.3643045784637735,
      "language_loss": 0.76298034,
      "learning_rate": 3.9211264938363975e-06,
      "loss": 0.78497219,
      "num_input_tokens_seen": 42165320,
      "step": 1946,
      "time_per_iteration": 2.792555093765259
    },
    {
      "auxiliary_loss_clip": 0.01134728,
      "auxiliary_loss_mlp": 0.01056112,
      "balance_loss_clip": 1.0507704,
      "balance_loss_mlp": 1.03536999,
      "epoch": 0.1170599729445363,
      "flos": 15267853232640.0,
      "grad_norm": 2.058923240355934,
      "language_loss": 0.69014907,
      "learning_rate": 3.921018163077448e-06,
      "loss": 0.71205747,
      "num_input_tokens_seen": 42182955,
      "step": 1947,
      "time_per_iteration": 2.643807888031006
    },
    {
      "auxiliary_loss_clip": 0.01154759,
      "auxiliary_loss_mlp": 0.01067767,
      "balance_loss_clip": 1.05707347,
      "balance_loss_mlp": 1.04604673,
      "epoch": 0.11712009619720427,
      "flos": 17164439251200.0,
      "grad_norm": 2.0690991629011615,
      "language_loss": 0.85044622,
      "learning_rate": 3.920909759473295e-06,
      "loss": 0.87267148,
      "num_input_tokens_seen": 42200760,
      "step": 1948,
      "time_per_iteration": 2.6399292945861816
    },
    {
      "auxiliary_loss_clip": 0.01051031,
      "auxiliary_loss_mlp": 0.0075782,
      "balance_loss_clip": 1.0245688,
      "balance_loss_mlp": 0.99997467,
      "epoch": 0.11718021944987224,
      "flos": 70940991997440.0,
      "grad_norm": 0.8206821069070506,
      "language_loss": 0.65139282,
      "learning_rate": 3.920801283028054e-06,
      "loss": 0.66948134,
      "num_input_tokens_seen": 42265745,
      "step": 1949,
      "time_per_iteration": 3.3030900955200195
    },
    {
      "auxiliary_loss_clip": 0.01159399,
      "auxiliary_loss_mlp": 0.01061163,
      "balance_loss_clip": 1.05735683,
      "balance_loss_mlp": 1.04054022,
      "epoch": 0.1172403427025402,
      "flos": 27453456408960.0,
      "grad_norm": 1.512876015443777,
      "language_loss": 0.71746683,
      "learning_rate": 3.920692733745835e-06,
      "loss": 0.73967248,
      "num_input_tokens_seen": 42286245,
      "step": 1950,
      "time_per_iteration": 2.739341974258423
    },
    {
      "auxiliary_loss_clip": 0.01175731,
      "auxiliary_loss_mlp": 0.01061149,
      "balance_loss_clip": 1.06152189,
      "balance_loss_mlp": 1.03907192,
      "epoch": 0.11730046595520818,
      "flos": 15668723992320.0,
      "grad_norm": 2.1258853115079996,
      "language_loss": 0.76671386,
      "learning_rate": 3.920584111630755e-06,
      "loss": 0.78908259,
      "num_input_tokens_seen": 42302710,
      "step": 1951,
      "time_per_iteration": 2.624788999557495
    },
    {
      "auxiliary_loss_clip": 0.01129104,
      "auxiliary_loss_mlp": 0.0106562,
      "balance_loss_clip": 1.05285251,
      "balance_loss_mlp": 1.04435349,
      "epoch": 0.11736058920787615,
      "flos": 25630164092160.0,
      "grad_norm": 1.7264952730121887,
      "language_loss": 0.75964963,
      "learning_rate": 3.9204754166869325e-06,
      "loss": 0.7815969,
      "num_input_tokens_seen": 42324115,
      "step": 1952,
      "time_per_iteration": 2.824826955795288
    },
    {
      "auxiliary_loss_clip": 0.01123677,
      "auxiliary_loss_mlp": 0.01065929,
      "balance_loss_clip": 1.04589534,
      "balance_loss_mlp": 1.04451907,
      "epoch": 0.11742071246054411,
      "flos": 21434289701760.0,
      "grad_norm": 2.2111022500713453,
      "language_loss": 0.72316217,
      "learning_rate": 3.920366648918491e-06,
      "loss": 0.74505818,
      "num_input_tokens_seen": 42342505,
      "step": 1953,
      "time_per_iteration": 2.7456531524658203
    },
    {
      "auxiliary_loss_clip": 0.01149214,
      "auxiliary_loss_mlp": 0.00781136,
      "balance_loss_clip": 1.0549686,
      "balance_loss_mlp": 1.0000577,
      "epoch": 0.11748083571321208,
      "flos": 15997845335040.0,
      "grad_norm": 2.1208802652878522,
      "language_loss": 0.79780388,
      "learning_rate": 3.920257808329552e-06,
      "loss": 0.81710744,
      "num_input_tokens_seen": 42360525,
      "step": 1954,
      "time_per_iteration": 2.653949737548828
    },
    {
      "auxiliary_loss_clip": 0.01112399,
      "auxiliary_loss_mlp": 0.01059787,
      "balance_loss_clip": 1.04880822,
      "balance_loss_mlp": 1.03763783,
      "epoch": 0.11754095896588006,
      "flos": 16180056051840.0,
      "grad_norm": 1.9673692595826442,
      "language_loss": 0.8553021,
      "learning_rate": 3.920148894924246e-06,
      "loss": 0.87702394,
      "num_input_tokens_seen": 42377045,
      "step": 1955,
      "time_per_iteration": 2.7987124919891357
    },
    {
      "auxiliary_loss_clip": 0.01163172,
      "auxiliary_loss_mlp": 0.00779783,
      "balance_loss_clip": 1.05209899,
      "balance_loss_mlp": 1.00016606,
      "epoch": 0.11760108221854802,
      "flos": 13261596013440.0,
      "grad_norm": 2.12926288831445,
      "language_loss": 0.78105426,
      "learning_rate": 3.920039908706701e-06,
      "loss": 0.80048382,
      "num_input_tokens_seen": 42393960,
      "step": 1956,
      "time_per_iteration": 2.6247944831848145
    },
    {
      "auxiliary_loss_clip": 0.01158287,
      "auxiliary_loss_mlp": 0.01058454,
      "balance_loss_clip": 1.05559933,
      "balance_loss_mlp": 1.03601909,
      "epoch": 0.11766120547121599,
      "flos": 24498439303680.0,
      "grad_norm": 2.264983200322237,
      "language_loss": 0.80487299,
      "learning_rate": 3.91993084968105e-06,
      "loss": 0.82704043,
      "num_input_tokens_seen": 42413160,
      "step": 1957,
      "time_per_iteration": 5.862411260604858
    },
    {
      "auxiliary_loss_clip": 0.01168294,
      "auxiliary_loss_mlp": 0.0105259,
      "balance_loss_clip": 1.05703866,
      "balance_loss_mlp": 1.0308696,
      "epoch": 0.11772132872388397,
      "flos": 17784005967360.0,
      "grad_norm": 4.8672025609093215,
      "language_loss": 0.77955222,
      "learning_rate": 3.919821717851428e-06,
      "loss": 0.80176103,
      "num_input_tokens_seen": 42432590,
      "step": 1958,
      "time_per_iteration": 4.4218549728393555
    },
    {
      "auxiliary_loss_clip": 0.01149976,
      "auxiliary_loss_mlp": 0.0105003,
      "balance_loss_clip": 1.05451894,
      "balance_loss_mlp": 1.02680755,
      "epoch": 0.11778145197655193,
      "flos": 13217030213760.0,
      "grad_norm": 1.7537692363765556,
      "language_loss": 0.77002251,
      "learning_rate": 3.919712513221976e-06,
      "loss": 0.79202259,
      "num_input_tokens_seen": 42450135,
      "step": 1959,
      "time_per_iteration": 2.674323558807373
    },
    {
      "auxiliary_loss_clip": 0.01162585,
      "auxiliary_loss_mlp": 0.01057019,
      "balance_loss_clip": 1.05857027,
      "balance_loss_mlp": 1.03484631,
      "epoch": 0.1178415752292199,
      "flos": 20230204965120.0,
      "grad_norm": 2.2026367524708927,
      "language_loss": 0.70078689,
      "learning_rate": 3.919603235796832e-06,
      "loss": 0.722983,
      "num_input_tokens_seen": 42470050,
      "step": 1960,
      "time_per_iteration": 2.7704508304595947
    },
    {
      "auxiliary_loss_clip": 0.01161089,
      "auxiliary_loss_mlp": 0.01055224,
      "balance_loss_clip": 1.05841374,
      "balance_loss_mlp": 1.03228831,
      "epoch": 0.11790169848188788,
      "flos": 13040134709760.0,
      "grad_norm": 2.663996374773888,
      "language_loss": 0.81045067,
      "learning_rate": 3.9194938855801406e-06,
      "loss": 0.83261371,
      "num_input_tokens_seen": 42484335,
      "step": 1961,
      "time_per_iteration": 4.67006778717041
    },
    {
      "auxiliary_loss_clip": 0.01163817,
      "auxiliary_loss_mlp": 0.00779643,
      "balance_loss_clip": 1.05658793,
      "balance_loss_mlp": 1.00009537,
      "epoch": 0.11796182173455584,
      "flos": 22265728790400.0,
      "grad_norm": 1.71345119244153,
      "language_loss": 0.92273545,
      "learning_rate": 3.919384462576049e-06,
      "loss": 0.94217002,
      "num_input_tokens_seen": 42502720,
      "step": 1962,
      "time_per_iteration": 2.6559524536132812
    },
    {
      "auxiliary_loss_clip": 0.01139826,
      "auxiliary_loss_mlp": 0.01058964,
      "balance_loss_clip": 1.05222392,
      "balance_loss_mlp": 1.03704107,
      "epoch": 0.1180219449872238,
      "flos": 10635017892480.0,
      "grad_norm": 2.157203116008796,
      "language_loss": 0.87635934,
      "learning_rate": 3.919274966788707e-06,
      "loss": 0.8983472,
      "num_input_tokens_seen": 42519460,
      "step": 1963,
      "time_per_iteration": 2.710042715072632
    },
    {
      "auxiliary_loss_clip": 0.0115823,
      "auxiliary_loss_mlp": 0.00779391,
      "balance_loss_clip": 1.05600929,
      "balance_loss_mlp": 1.00011134,
      "epoch": 0.11808206823989177,
      "flos": 20923532259840.0,
      "grad_norm": 2.8331529324994333,
      "language_loss": 0.83879703,
      "learning_rate": 3.919165398222265e-06,
      "loss": 0.85817325,
      "num_input_tokens_seen": 42539420,
      "step": 1964,
      "time_per_iteration": 2.734941244125366
    },
    {
      "auxiliary_loss_clip": 0.01122529,
      "auxiliary_loss_mlp": 0.01069054,
      "balance_loss_clip": 1.05171156,
      "balance_loss_mlp": 1.04628491,
      "epoch": 0.11814219149255975,
      "flos": 20777770869120.0,
      "grad_norm": 3.9132941826799543,
      "language_loss": 0.8313272,
      "learning_rate": 3.919055756880879e-06,
      "loss": 0.85324299,
      "num_input_tokens_seen": 42558225,
      "step": 1965,
      "time_per_iteration": 2.7427306175231934
    },
    {
      "auxiliary_loss_clip": 0.01178673,
      "auxiliary_loss_mlp": 0.01053338,
      "balance_loss_clip": 1.05815279,
      "balance_loss_mlp": 1.03163004,
      "epoch": 0.11820231474522772,
      "flos": 48759938542080.0,
      "grad_norm": 1.6720023918141877,
      "language_loss": 0.74227381,
      "learning_rate": 3.918946042768707e-06,
      "loss": 0.76459396,
      "num_input_tokens_seen": 42580790,
      "step": 1966,
      "time_per_iteration": 2.8265397548675537
    },
    {
      "auxiliary_loss_clip": 0.01163407,
      "auxiliary_loss_mlp": 0.0106081,
      "balance_loss_clip": 1.06309748,
      "balance_loss_mlp": 1.03836274,
      "epoch": 0.11826243799789568,
      "flos": 16690598012160.0,
      "grad_norm": 2.5628488285375397,
      "language_loss": 0.73137337,
      "learning_rate": 3.918836255889908e-06,
      "loss": 0.7536155,
      "num_input_tokens_seen": 42597355,
      "step": 1967,
      "time_per_iteration": 2.706193685531616
    },
    {
      "auxiliary_loss_clip": 0.01167052,
      "auxiliary_loss_mlp": 0.01053471,
      "balance_loss_clip": 1.05852592,
      "balance_loss_mlp": 1.03141701,
      "epoch": 0.11832256125056366,
      "flos": 16909868586240.0,
      "grad_norm": 5.332816815546028,
      "language_loss": 0.8831054,
      "learning_rate": 3.9187263962486456e-06,
      "loss": 0.90531063,
      "num_input_tokens_seen": 42616060,
      "step": 1968,
      "time_per_iteration": 2.6308343410491943
    },
    {
      "auxiliary_loss_clip": 0.01168356,
      "auxiliary_loss_mlp": 0.01051817,
      "balance_loss_clip": 1.06406927,
      "balance_loss_mlp": 1.0294776,
      "epoch": 0.11838268450323162,
      "flos": 22820405587200.0,
      "grad_norm": 2.252087054693662,
      "language_loss": 0.67010254,
      "learning_rate": 3.918616463849087e-06,
      "loss": 0.69230425,
      "num_input_tokens_seen": 42636285,
      "step": 1969,
      "time_per_iteration": 2.662480592727661
    },
    {
      "auxiliary_loss_clip": 0.01130071,
      "auxiliary_loss_mlp": 0.0106143,
      "balance_loss_clip": 1.05177045,
      "balance_loss_mlp": 1.03774357,
      "epoch": 0.11844280775589959,
      "flos": 33545844990720.0,
      "grad_norm": 2.153814675458072,
      "language_loss": 0.80455101,
      "learning_rate": 3.918506458695399e-06,
      "loss": 0.82646602,
      "num_input_tokens_seen": 42658320,
      "step": 1970,
      "time_per_iteration": 2.798050880432129
    },
    {
      "auxiliary_loss_clip": 0.01060284,
      "auxiliary_loss_mlp": 0.01021383,
      "balance_loss_clip": 1.02553701,
      "balance_loss_mlp": 1.01892686,
      "epoch": 0.11850293100856757,
      "flos": 66350998604160.0,
      "grad_norm": 0.8165911228106061,
      "language_loss": 0.66192186,
      "learning_rate": 3.918396380791754e-06,
      "loss": 0.68273854,
      "num_input_tokens_seen": 42721500,
      "step": 1971,
      "time_per_iteration": 3.167018413543701
    },
    {
      "auxiliary_loss_clip": 0.01151504,
      "auxiliary_loss_mlp": 0.0105629,
      "balance_loss_clip": 1.05294323,
      "balance_loss_mlp": 1.03422379,
      "epoch": 0.11856305426123553,
      "flos": 24681045070080.0,
      "grad_norm": 2.1839859106137554,
      "language_loss": 0.79782552,
      "learning_rate": 3.918286230142327e-06,
      "loss": 0.81990343,
      "num_input_tokens_seen": 42739825,
      "step": 1972,
      "time_per_iteration": 2.6908793449401855
    },
    {
      "auxiliary_loss_clip": 0.01133219,
      "auxiliary_loss_mlp": 0.00778766,
      "balance_loss_clip": 1.05341005,
      "balance_loss_mlp": 1.00005877,
      "epoch": 0.1186231775139035,
      "flos": 24280102483200.0,
      "grad_norm": 2.0473813607633384,
      "language_loss": 0.72843599,
      "learning_rate": 3.918176006751292e-06,
      "loss": 0.74755585,
      "num_input_tokens_seen": 42758695,
      "step": 1973,
      "time_per_iteration": 2.7801859378814697
    },
    {
      "auxiliary_loss_clip": 0.01138022,
      "auxiliary_loss_mlp": 0.01049764,
      "balance_loss_clip": 1.05580497,
      "balance_loss_mlp": 1.02707887,
      "epoch": 0.11868330076657148,
      "flos": 21757413473280.0,
      "grad_norm": 1.6449677647733996,
      "language_loss": 0.72019619,
      "learning_rate": 3.918065710622832e-06,
      "loss": 0.74207413,
      "num_input_tokens_seen": 42778510,
      "step": 1974,
      "time_per_iteration": 2.7337663173675537
    },
    {
      "auxiliary_loss_clip": 0.01129602,
      "auxiliary_loss_mlp": 0.01043161,
      "balance_loss_clip": 1.05265522,
      "balance_loss_mlp": 1.02086854,
      "epoch": 0.11874342401923944,
      "flos": 17193274894080.0,
      "grad_norm": 2.017372400194955,
      "language_loss": 0.77409399,
      "learning_rate": 3.917955341761128e-06,
      "loss": 0.79582161,
      "num_input_tokens_seen": 42793995,
      "step": 1975,
      "time_per_iteration": 2.669546604156494
    },
    {
      "auxiliary_loss_clip": 0.01131477,
      "auxiliary_loss_mlp": 0.01059968,
      "balance_loss_clip": 1.05880177,
      "balance_loss_mlp": 1.03908277,
      "epoch": 0.11880354727190741,
      "flos": 15229572312960.0,
      "grad_norm": 2.3842578575289,
      "language_loss": 0.75110453,
      "learning_rate": 3.917844900170364e-06,
      "loss": 0.77301902,
      "num_input_tokens_seen": 42809000,
      "step": 1976,
      "time_per_iteration": 2.8439090251922607
    },
    {
      "auxiliary_loss_clip": 0.0116819,
      "auxiliary_loss_mlp": 0.01049523,
      "balance_loss_clip": 1.05999744,
      "balance_loss_mlp": 1.02835166,
      "epoch": 0.11886367052457537,
      "flos": 27309706179840.0,
      "grad_norm": 1.8674311015318124,
      "language_loss": 0.74877423,
      "learning_rate": 3.91773438585473e-06,
      "loss": 0.77095133,
      "num_input_tokens_seen": 42831585,
      "step": 1977,
      "time_per_iteration": 2.6747169494628906
    },
    {
      "auxiliary_loss_clip": 0.01182095,
      "auxiliary_loss_mlp": 0.01059621,
      "balance_loss_clip": 1.05954552,
      "balance_loss_mlp": 1.03805614,
      "epoch": 0.11892379377724335,
      "flos": 21798280172160.0,
      "grad_norm": 2.1793079873879604,
      "language_loss": 0.74207634,
      "learning_rate": 3.9176237988184165e-06,
      "loss": 0.76449353,
      "num_input_tokens_seen": 42848420,
      "step": 1978,
      "time_per_iteration": 2.631664514541626
    },
    {
      "auxiliary_loss_clip": 0.01142323,
      "auxiliary_loss_mlp": 0.01050585,
      "balance_loss_clip": 1.06037045,
      "balance_loss_mlp": 1.0289247,
      "epoch": 0.11898391702991132,
      "flos": 13991013498240.0,
      "grad_norm": 1.7170872786869797,
      "language_loss": 0.73256385,
      "learning_rate": 3.917513139065616e-06,
      "loss": 0.754493,
      "num_input_tokens_seen": 42866645,
      "step": 1979,
      "time_per_iteration": 2.7442541122436523
    },
    {
      "auxiliary_loss_clip": 0.01137516,
      "auxiliary_loss_mlp": 0.01051378,
      "balance_loss_clip": 1.0566175,
      "balance_loss_mlp": 1.02968168,
      "epoch": 0.11904404028257928,
      "flos": 32234567091840.0,
      "grad_norm": 1.876224505386343,
      "language_loss": 0.98293436,
      "learning_rate": 3.917402406600525e-06,
      "loss": 1.00482333,
      "num_input_tokens_seen": 42888515,
      "step": 1980,
      "time_per_iteration": 2.787667989730835
    },
    {
      "auxiliary_loss_clip": 0.01153629,
      "auxiliary_loss_mlp": 0.01053612,
      "balance_loss_clip": 1.05595791,
      "balance_loss_mlp": 1.03077161,
      "epoch": 0.11910416353524726,
      "flos": 23586272398080.0,
      "grad_norm": 1.7507584506289393,
      "language_loss": 0.86265099,
      "learning_rate": 3.917291601427342e-06,
      "loss": 0.88472342,
      "num_input_tokens_seen": 42909035,
      "step": 1981,
      "time_per_iteration": 2.6680359840393066
    },
    {
      "auxiliary_loss_clip": 0.01158736,
      "auxiliary_loss_mlp": 0.01064978,
      "balance_loss_clip": 1.06144083,
      "balance_loss_mlp": 1.04214907,
      "epoch": 0.11916428678791523,
      "flos": 25333038789120.0,
      "grad_norm": 1.8908045276276995,
      "language_loss": 0.85375237,
      "learning_rate": 3.91718072355027e-06,
      "loss": 0.87598956,
      "num_input_tokens_seen": 42927555,
      "step": 1982,
      "time_per_iteration": 2.732797861099243
    },
    {
      "auxiliary_loss_clip": 0.01146432,
      "auxiliary_loss_mlp": 0.01050259,
      "balance_loss_clip": 1.05539966,
      "balance_loss_mlp": 1.02843213,
      "epoch": 0.11922441004058319,
      "flos": 19788431592960.0,
      "grad_norm": 2.3856086229742877,
      "language_loss": 0.85202634,
      "learning_rate": 3.917069772973513e-06,
      "loss": 0.87399322,
      "num_input_tokens_seen": 42945300,
      "step": 1983,
      "time_per_iteration": 2.6839804649353027
    },
    {
      "auxiliary_loss_clip": 0.01126589,
      "auxiliary_loss_mlp": 0.01056051,
      "balance_loss_clip": 1.05602145,
      "balance_loss_mlp": 1.03399742,
      "epoch": 0.11928453329325117,
      "flos": 21536347219200.0,
      "grad_norm": 3.6641824085676022,
      "language_loss": 0.7693429,
      "learning_rate": 3.916958749701277e-06,
      "loss": 0.79116929,
      "num_input_tokens_seen": 42961295,
      "step": 1984,
      "time_per_iteration": 2.7008767127990723
    },
    {
      "auxiliary_loss_clip": 0.01161623,
      "auxiliary_loss_mlp": 0.01055251,
      "balance_loss_clip": 1.05752373,
      "balance_loss_mlp": 1.0334003,
      "epoch": 0.11934465654591914,
      "flos": 20815010294400.0,
      "grad_norm": 1.917528093726237,
      "language_loss": 0.83058321,
      "learning_rate": 3.9168476537377745e-06,
      "loss": 0.85275191,
      "num_input_tokens_seen": 42980330,
      "step": 1985,
      "time_per_iteration": 2.6692728996276855
    },
    {
      "auxiliary_loss_clip": 0.01151831,
      "auxiliary_loss_mlp": 0.01050086,
      "balance_loss_clip": 1.0541923,
      "balance_loss_mlp": 1.02835393,
      "epoch": 0.1194047797985871,
      "flos": 19060486565760.0,
      "grad_norm": 1.8732848573733223,
      "language_loss": 0.74398553,
      "learning_rate": 3.916736485087216e-06,
      "loss": 0.76600474,
      "num_input_tokens_seen": 42996125,
      "step": 1986,
      "time_per_iteration": 2.722013473510742
    },
    {
      "auxiliary_loss_clip": 0.01146125,
      "auxiliary_loss_mlp": 0.01059008,
      "balance_loss_clip": 1.05472732,
      "balance_loss_mlp": 1.03791952,
      "epoch": 0.11946490305125507,
      "flos": 27190805184000.0,
      "grad_norm": 2.4724436343771083,
      "language_loss": 0.72123617,
      "learning_rate": 3.916625243753819e-06,
      "loss": 0.74328756,
      "num_input_tokens_seen": 43014180,
      "step": 1987,
      "time_per_iteration": 2.814481258392334
    },
    {
      "auxiliary_loss_clip": 0.01156854,
      "auxiliary_loss_mlp": 0.01054644,
      "balance_loss_clip": 1.05747938,
      "balance_loss_mlp": 1.03138638,
      "epoch": 0.11952502630392305,
      "flos": 21140791672320.0,
      "grad_norm": 1.9246234449532542,
      "language_loss": 0.72007513,
      "learning_rate": 3.916513929741799e-06,
      "loss": 0.74219012,
      "num_input_tokens_seen": 43032120,
      "step": 1988,
      "time_per_iteration": 2.7242019176483154
    },
    {
      "auxiliary_loss_clip": 0.0116348,
      "auxiliary_loss_mlp": 0.01062102,
      "balance_loss_clip": 1.05559146,
      "balance_loss_mlp": 1.03913057,
      "epoch": 0.11958514955659101,
      "flos": 22124241118080.0,
      "grad_norm": 1.7561483239324645,
      "language_loss": 0.81144297,
      "learning_rate": 3.91640254305538e-06,
      "loss": 0.83369875,
      "num_input_tokens_seen": 43052215,
      "step": 1989,
      "time_per_iteration": 2.6259546279907227
    },
    {
      "auxiliary_loss_clip": 0.01135956,
      "auxiliary_loss_mlp": 0.01057689,
      "balance_loss_clip": 1.05254042,
      "balance_loss_mlp": 1.03325129,
      "epoch": 0.11964527280925898,
      "flos": 17421452040960.0,
      "grad_norm": 2.5516320258539795,
      "language_loss": 0.75881672,
      "learning_rate": 3.916291083698784e-06,
      "loss": 0.7807532,
      "num_input_tokens_seen": 43069720,
      "step": 1990,
      "time_per_iteration": 2.6779251098632812
    },
    {
      "auxiliary_loss_clip": 0.0105322,
      "auxiliary_loss_mlp": 0.01019112,
      "balance_loss_clip": 1.02816892,
      "balance_loss_mlp": 1.01647794,
      "epoch": 0.11970539606192696,
      "flos": 70679741402880.0,
      "grad_norm": 0.8628582727639288,
      "language_loss": 0.55184531,
      "learning_rate": 3.916179551676238e-06,
      "loss": 0.57256866,
      "num_input_tokens_seen": 43123130,
      "step": 1991,
      "time_per_iteration": 3.3713693618774414
    },
    {
      "auxiliary_loss_clip": 0.01136423,
      "auxiliary_loss_mlp": 0.01053959,
      "balance_loss_clip": 1.05748868,
      "balance_loss_mlp": 1.03326464,
      "epoch": 0.11976551931459492,
      "flos": 21215019127680.0,
      "grad_norm": 2.286300891386994,
      "language_loss": 0.78371406,
      "learning_rate": 3.916067946991971e-06,
      "loss": 0.80561793,
      "num_input_tokens_seen": 43140015,
      "step": 1992,
      "time_per_iteration": 2.6797914505004883
    },
    {
      "auxiliary_loss_clip": 0.0117949,
      "auxiliary_loss_mlp": 0.01056635,
      "balance_loss_clip": 1.05811,
      "balance_loss_mlp": 1.03453374,
      "epoch": 0.11982564256726289,
      "flos": 25989306226560.0,
      "grad_norm": 1.8481811043026504,
      "language_loss": 0.78911144,
      "learning_rate": 3.915956269650216e-06,
      "loss": 0.81147265,
      "num_input_tokens_seen": 43160105,
      "step": 1993,
      "time_per_iteration": 2.691301107406616
    },
    {
      "auxiliary_loss_clip": 0.01126423,
      "auxiliary_loss_mlp": 0.0106217,
      "balance_loss_clip": 1.05012226,
      "balance_loss_mlp": 1.04081941,
      "epoch": 0.11988576581993086,
      "flos": 21650866755840.0,
      "grad_norm": 1.644866568705103,
      "language_loss": 0.82088816,
      "learning_rate": 3.915844519655208e-06,
      "loss": 0.84277415,
      "num_input_tokens_seen": 43179835,
      "step": 1994,
      "time_per_iteration": 2.772905111312866
    },
    {
      "auxiliary_loss_clip": 0.0115068,
      "auxiliary_loss_mlp": 0.01063961,
      "balance_loss_clip": 1.05523098,
      "balance_loss_mlp": 1.0433259,
      "epoch": 0.11994588907259883,
      "flos": 17857407409920.0,
      "grad_norm": 2.0065598513575247,
      "language_loss": 0.88392794,
      "learning_rate": 3.915732697011183e-06,
      "loss": 0.9060744,
      "num_input_tokens_seen": 43197210,
      "step": 1995,
      "time_per_iteration": 4.206532716751099
    },
    {
      "auxiliary_loss_clip": 0.01153482,
      "auxiliary_loss_mlp": 0.01066415,
      "balance_loss_clip": 1.06005812,
      "balance_loss_mlp": 1.0441823,
      "epoch": 0.1200060123252668,
      "flos": 24462744163200.0,
      "grad_norm": 1.8775058007239456,
      "language_loss": 0.73949909,
      "learning_rate": 3.9156208017223825e-06,
      "loss": 0.76169801,
      "num_input_tokens_seen": 43215050,
      "step": 1996,
      "time_per_iteration": 2.7263944149017334
    },
    {
      "auxiliary_loss_clip": 0.01141484,
      "auxiliary_loss_mlp": 0.01060112,
      "balance_loss_clip": 1.05754757,
      "balance_loss_mlp": 1.03808212,
      "epoch": 0.12006613557793476,
      "flos": 18732191235840.0,
      "grad_norm": 1.976051865072764,
      "language_loss": 0.88125587,
      "learning_rate": 3.915508833793048e-06,
      "loss": 0.90327179,
      "num_input_tokens_seen": 43233900,
      "step": 1997,
      "time_per_iteration": 4.29426383972168
    },
    {
      "auxiliary_loss_clip": 0.01165634,
      "auxiliary_loss_mlp": 0.00779568,
      "balance_loss_clip": 1.05701697,
      "balance_loss_mlp": 1.00001049,
      "epoch": 0.12012625883060274,
      "flos": 22267739952000.0,
      "grad_norm": 2.1091392562336018,
      "language_loss": 0.79031086,
      "learning_rate": 3.915396793227428e-06,
      "loss": 0.80976284,
      "num_input_tokens_seen": 43252105,
      "step": 1998,
      "time_per_iteration": 4.330955266952515
    },
    {
      "auxiliary_loss_clip": 0.0116661,
      "auxiliary_loss_mlp": 0.00779642,
      "balance_loss_clip": 1.0576719,
      "balance_loss_mlp": 1.00002396,
      "epoch": 0.1201863820832707,
      "flos": 21758885930880.0,
      "grad_norm": 1.799585336659533,
      "language_loss": 0.73583078,
      "learning_rate": 3.915284680029769e-06,
      "loss": 0.75529337,
      "num_input_tokens_seen": 43270315,
      "step": 1999,
      "time_per_iteration": 2.754770040512085
    },
    {
      "auxiliary_loss_clip": 0.01178966,
      "auxiliary_loss_mlp": 0.01073097,
      "balance_loss_clip": 1.0602119,
      "balance_loss_mlp": 1.05115068,
      "epoch": 0.12024650533593867,
      "flos": 21907987286400.0,
      "grad_norm": 2.916355473014409,
      "language_loss": 0.74854898,
      "learning_rate": 3.915172494204323e-06,
      "loss": 0.77106953,
      "num_input_tokens_seen": 43289935,
      "step": 2000,
      "time_per_iteration": 4.3900322914123535
    },
    {
      "auxiliary_loss_clip": 0.01149374,
      "auxiliary_loss_mlp": 0.01069735,
      "balance_loss_clip": 1.05375695,
      "balance_loss_mlp": 1.04763341,
      "epoch": 0.12030662858860665,
      "flos": 21689219502720.0,
      "grad_norm": 1.5203973891597686,
      "language_loss": 0.8496564,
      "learning_rate": 3.915060235755344e-06,
      "loss": 0.87184751,
      "num_input_tokens_seen": 43309325,
      "step": 2001,
      "time_per_iteration": 2.6912643909454346
    },
    {
      "auxiliary_loss_clip": 0.01154057,
      "auxiliary_loss_mlp": 0.01063637,
      "balance_loss_clip": 1.05600786,
      "balance_loss_mlp": 1.04265642,
      "epoch": 0.12036675184127461,
      "flos": 12933228856320.0,
      "grad_norm": 2.932264271186656,
      "language_loss": 0.74711967,
      "learning_rate": 3.91494790468709e-06,
      "loss": 0.76929653,
      "num_input_tokens_seen": 43327010,
      "step": 2002,
      "time_per_iteration": 2.6991024017333984
    },
    {
      "auxiliary_loss_clip": 0.01129169,
      "auxiliary_loss_mlp": 0.01066705,
      "balance_loss_clip": 1.05340302,
      "balance_loss_mlp": 1.0429939,
      "epoch": 0.12042687509394258,
      "flos": 20851028657280.0,
      "grad_norm": 2.117271428042382,
      "language_loss": 0.78029454,
      "learning_rate": 3.9148355010038185e-06,
      "loss": 0.80225325,
      "num_input_tokens_seen": 43345650,
      "step": 2003,
      "time_per_iteration": 2.731381416320801
    },
    {
      "auxiliary_loss_clip": 0.01163252,
      "auxiliary_loss_mlp": 0.01062886,
      "balance_loss_clip": 1.05728662,
      "balance_loss_mlp": 1.04073668,
      "epoch": 0.12048699834661056,
      "flos": 23878513451520.0,
      "grad_norm": 1.585850552088038,
      "language_loss": 0.72205627,
      "learning_rate": 3.914723024709793e-06,
      "loss": 0.74431765,
      "num_input_tokens_seen": 43365555,
      "step": 2004,
      "time_per_iteration": 2.725092649459839
    },
    {
      "auxiliary_loss_clip": 0.01160616,
      "auxiliary_loss_mlp": 0.01069457,
      "balance_loss_clip": 1.05870187,
      "balance_loss_mlp": 1.04645014,
      "epoch": 0.12054712159927852,
      "flos": 19756363726080.0,
      "grad_norm": 1.9357732467170252,
      "language_loss": 0.78415942,
      "learning_rate": 3.914610475809279e-06,
      "loss": 0.8064602,
      "num_input_tokens_seen": 43384990,
      "step": 2005,
      "time_per_iteration": 2.7232437133789062
    },
    {
      "auxiliary_loss_clip": 0.01073016,
      "auxiliary_loss_mlp": 0.00758901,
      "balance_loss_clip": 1.02995479,
      "balance_loss_mlp": 1.00011683,
      "epoch": 0.12060724485194649,
      "flos": 51672763123200.0,
      "grad_norm": 0.9264315537536937,
      "language_loss": 0.58087146,
      "learning_rate": 3.914497854306543e-06,
      "loss": 0.59919059,
      "num_input_tokens_seen": 43436335,
      "step": 2006,
      "time_per_iteration": 2.9570157527923584
    },
    {
      "auxiliary_loss_clip": 0.01155081,
      "auxiliary_loss_mlp": 0.01053472,
      "balance_loss_clip": 1.05803597,
      "balance_loss_mlp": 1.03299201,
      "epoch": 0.12066736810461445,
      "flos": 18990425088000.0,
      "grad_norm": 1.6109316320484448,
      "language_loss": 0.76524282,
      "learning_rate": 3.9143851602058575e-06,
      "loss": 0.78732836,
      "num_input_tokens_seen": 43456495,
      "step": 2007,
      "time_per_iteration": 2.763380289077759
    },
    {
      "auxiliary_loss_clip": 0.01147254,
      "auxiliary_loss_mlp": 0.01064209,
      "balance_loss_clip": 1.05931091,
      "balance_loss_mlp": 1.04177368,
      "epoch": 0.12072749135728243,
      "flos": 16471973882880.0,
      "grad_norm": 2.449779851562752,
      "language_loss": 0.83023942,
      "learning_rate": 3.914272393511494e-06,
      "loss": 0.85235405,
      "num_input_tokens_seen": 43473085,
      "step": 2008,
      "time_per_iteration": 2.7693119049072266
    },
    {
      "auxiliary_loss_clip": 0.01176157,
      "auxiliary_loss_mlp": 0.01052894,
      "balance_loss_clip": 1.0584172,
      "balance_loss_mlp": 1.03135288,
      "epoch": 0.1207876146099504,
      "flos": 18077108947200.0,
      "grad_norm": 2.203355340521787,
      "language_loss": 0.83835697,
      "learning_rate": 3.91415955422773e-06,
      "loss": 0.86064744,
      "num_input_tokens_seen": 43491135,
      "step": 2009,
      "time_per_iteration": 2.640944242477417
    },
    {
      "auxiliary_loss_clip": 0.01180076,
      "auxiliary_loss_mlp": 0.01053549,
      "balance_loss_clip": 1.06196725,
      "balance_loss_mlp": 1.02994514,
      "epoch": 0.12084773786261836,
      "flos": 21871573873920.0,
      "grad_norm": 1.6799099601218046,
      "language_loss": 0.83870012,
      "learning_rate": 3.914046642358844e-06,
      "loss": 0.8610363,
      "num_input_tokens_seen": 43510440,
      "step": 2010,
      "time_per_iteration": 2.716127634048462
    },
    {
      "auxiliary_loss_clip": 0.01145261,
      "auxiliary_loss_mlp": 0.00780804,
      "balance_loss_clip": 1.05555713,
      "balance_loss_mlp": 1.0000627,
      "epoch": 0.12090786111528634,
      "flos": 18333044328960.0,
      "grad_norm": 1.8933604390076018,
      "language_loss": 0.84194541,
      "learning_rate": 3.9139336579091174e-06,
      "loss": 0.86120605,
      "num_input_tokens_seen": 43530145,
      "step": 2011,
      "time_per_iteration": 2.73793625831604
    },
    {
      "auxiliary_loss_clip": 0.01148418,
      "auxiliary_loss_mlp": 0.01060974,
      "balance_loss_clip": 1.05480969,
      "balance_loss_mlp": 1.03905129,
      "epoch": 0.1209679843679543,
      "flos": 21105850717440.0,
      "grad_norm": 2.0524904800028154,
      "language_loss": 0.96236968,
      "learning_rate": 3.913820600882834e-06,
      "loss": 0.98446357,
      "num_input_tokens_seen": 43549315,
      "step": 2012,
      "time_per_iteration": 2.7269980907440186
    },
    {
      "auxiliary_loss_clip": 0.01146369,
      "auxiliary_loss_mlp": 0.01051396,
      "balance_loss_clip": 1.05808425,
      "balance_loss_mlp": 1.0289607,
      "epoch": 0.12102810762062227,
      "flos": 29241053585280.0,
      "grad_norm": 1.853151366811655,
      "language_loss": 0.80903435,
      "learning_rate": 3.913707471284283e-06,
      "loss": 0.83101201,
      "num_input_tokens_seen": 43569240,
      "step": 2013,
      "time_per_iteration": 2.740489959716797
    },
    {
      "auxiliary_loss_clip": 0.01124703,
      "auxiliary_loss_mlp": 0.0105341,
      "balance_loss_clip": 1.05300117,
      "balance_loss_mlp": 1.02962804,
      "epoch": 0.12108823087329025,
      "flos": 17930701111680.0,
      "grad_norm": 5.099975898232357,
      "language_loss": 0.77255923,
      "learning_rate": 3.9135942691177515e-06,
      "loss": 0.79434031,
      "num_input_tokens_seen": 43587710,
      "step": 2014,
      "time_per_iteration": 2.7361485958099365
    },
    {
      "auxiliary_loss_clip": 0.0116607,
      "auxiliary_loss_mlp": 0.01051056,
      "balance_loss_clip": 1.05832791,
      "balance_loss_mlp": 1.02791715,
      "epoch": 0.12114835412595822,
      "flos": 22091850028800.0,
      "grad_norm": 5.8570343294144465,
      "language_loss": 0.87169874,
      "learning_rate": 3.913480994387535e-06,
      "loss": 0.89387,
      "num_input_tokens_seen": 43606000,
      "step": 2015,
      "time_per_iteration": 2.6881515979766846
    },
    {
      "auxiliary_loss_clip": 0.01170382,
      "auxiliary_loss_mlp": 0.01051162,
      "balance_loss_clip": 1.05500197,
      "balance_loss_mlp": 1.0289886,
      "epoch": 0.12120847737862618,
      "flos": 20412343854720.0,
      "grad_norm": 2.087765239068409,
      "language_loss": 0.69146478,
      "learning_rate": 3.913367647097926e-06,
      "loss": 0.71368027,
      "num_input_tokens_seen": 43624815,
      "step": 2016,
      "time_per_iteration": 2.7096211910247803
    },
    {
      "auxiliary_loss_clip": 0.01152563,
      "auxiliary_loss_mlp": 0.0104714,
      "balance_loss_clip": 1.05737591,
      "balance_loss_mlp": 1.02390599,
      "epoch": 0.12126860063129415,
      "flos": 22309037614080.0,
      "grad_norm": 2.8043603396252865,
      "language_loss": 0.79858959,
      "learning_rate": 3.913254227253225e-06,
      "loss": 0.82058656,
      "num_input_tokens_seen": 43643960,
      "step": 2017,
      "time_per_iteration": 2.7042336463928223
    },
    {
      "auxiliary_loss_clip": 0.01156022,
      "auxiliary_loss_mlp": 0.0105052,
      "balance_loss_clip": 1.05479789,
      "balance_loss_mlp": 1.02740538,
      "epoch": 0.12132872388396213,
      "flos": 13699275235200.0,
      "grad_norm": 2.8700241463026654,
      "language_loss": 0.68828821,
      "learning_rate": 3.913140734857731e-06,
      "loss": 0.71035373,
      "num_input_tokens_seen": 43662650,
      "step": 2018,
      "time_per_iteration": 2.7015058994293213
    },
    {
      "auxiliary_loss_clip": 0.01136376,
      "auxiliary_loss_mlp": 0.01050749,
      "balance_loss_clip": 1.05524123,
      "balance_loss_mlp": 1.02873111,
      "epoch": 0.12138884713663009,
      "flos": 26466954307200.0,
      "grad_norm": 1.6132330771570709,
      "language_loss": 0.72476816,
      "learning_rate": 3.91302716991575e-06,
      "loss": 0.74663943,
      "num_input_tokens_seen": 43684205,
      "step": 2019,
      "time_per_iteration": 2.8956947326660156
    },
    {
      "auxiliary_loss_clip": 0.01107167,
      "auxiliary_loss_mlp": 0.01057916,
      "balance_loss_clip": 1.05286384,
      "balance_loss_mlp": 1.03482556,
      "epoch": 0.12144897038929806,
      "flos": 26141603892480.0,
      "grad_norm": 1.853626515444831,
      "language_loss": 0.92125106,
      "learning_rate": 3.912913532431586e-06,
      "loss": 0.94290185,
      "num_input_tokens_seen": 43706320,
      "step": 2020,
      "time_per_iteration": 2.9980764389038086
    },
    {
      "auxiliary_loss_clip": 0.0114145,
      "auxiliary_loss_mlp": 0.01055455,
      "balance_loss_clip": 1.05289125,
      "balance_loss_mlp": 1.03360391,
      "epoch": 0.12150909364196603,
      "flos": 24717530309760.0,
      "grad_norm": 1.9227427415613194,
      "language_loss": 0.7772885,
      "learning_rate": 3.912799822409549e-06,
      "loss": 0.79925752,
      "num_input_tokens_seen": 43724805,
      "step": 2021,
      "time_per_iteration": 3.01798939704895
    },
    {
      "auxiliary_loss_clip": 0.0117749,
      "auxiliary_loss_mlp": 0.01049007,
      "balance_loss_clip": 1.0610733,
      "balance_loss_mlp": 1.0277164,
      "epoch": 0.121569216894634,
      "flos": 25186990089600.0,
      "grad_norm": 2.054228820960504,
      "language_loss": 0.80712306,
      "learning_rate": 3.912686039853952e-06,
      "loss": 0.82938808,
      "num_input_tokens_seen": 43742320,
      "step": 2022,
      "time_per_iteration": 2.684309244155884
    },
    {
      "auxiliary_loss_clip": 0.01144749,
      "auxiliary_loss_mlp": 0.0106163,
      "balance_loss_clip": 1.055619,
      "balance_loss_mlp": 1.03697765,
      "epoch": 0.12162934014730196,
      "flos": 13444094039040.0,
      "grad_norm": 1.734031517866852,
      "language_loss": 0.84842217,
      "learning_rate": 3.912572184769108e-06,
      "loss": 0.87048596,
      "num_input_tokens_seen": 43760665,
      "step": 2023,
      "time_per_iteration": 2.6886441707611084
    },
    {
      "auxiliary_loss_clip": 0.01139348,
      "auxiliary_loss_mlp": 0.01053043,
      "balance_loss_clip": 1.05162323,
      "balance_loss_mlp": 1.03081048,
      "epoch": 0.12168946339996994,
      "flos": 16946138344320.0,
      "grad_norm": 2.3397199529221546,
      "language_loss": 0.85514021,
      "learning_rate": 3.912458257159335e-06,
      "loss": 0.87706411,
      "num_input_tokens_seen": 43779020,
      "step": 2024,
      "time_per_iteration": 2.8043718338012695
    },
    {
      "auxiliary_loss_clip": 0.01169767,
      "auxiliary_loss_mlp": 0.01055534,
      "balance_loss_clip": 1.05277538,
      "balance_loss_mlp": 1.03389716,
      "epoch": 0.12174958665263791,
      "flos": 29821585196160.0,
      "grad_norm": 1.8432491304976684,
      "language_loss": 0.72088945,
      "learning_rate": 3.912344257028954e-06,
      "loss": 0.74314243,
      "num_input_tokens_seen": 43798850,
      "step": 2025,
      "time_per_iteration": 2.704876184463501
    },
    {
      "auxiliary_loss_clip": 0.01148564,
      "auxiliary_loss_mlp": 0.01047618,
      "balance_loss_clip": 1.05486572,
      "balance_loss_mlp": 1.02555275,
      "epoch": 0.12180970990530587,
      "flos": 24641902224000.0,
      "grad_norm": 1.4969552271445652,
      "language_loss": 0.76075011,
      "learning_rate": 3.912230184382286e-06,
      "loss": 0.78271192,
      "num_input_tokens_seen": 43820130,
      "step": 2026,
      "time_per_iteration": 2.6957921981811523
    },
    {
      "auxiliary_loss_clip": 0.01147374,
      "auxiliary_loss_mlp": 0.01046261,
      "balance_loss_clip": 1.05086374,
      "balance_loss_mlp": 1.02474427,
      "epoch": 0.12186983315797385,
      "flos": 20521691832960.0,
      "grad_norm": 2.2064263994277478,
      "language_loss": 0.88769746,
      "learning_rate": 3.912116039223659e-06,
      "loss": 0.90963376,
      "num_input_tokens_seen": 43838485,
      "step": 2027,
      "time_per_iteration": 2.6847639083862305
    },
    {
      "auxiliary_loss_clip": 0.01143778,
      "auxiliary_loss_mlp": 0.01056715,
      "balance_loss_clip": 1.05258501,
      "balance_loss_mlp": 1.03667617,
      "epoch": 0.12192995641064182,
      "flos": 27818344719360.0,
      "grad_norm": 1.5725885574076592,
      "language_loss": 0.75544459,
      "learning_rate": 3.912001821557399e-06,
      "loss": 0.77744961,
      "num_input_tokens_seen": 43859080,
      "step": 2028,
      "time_per_iteration": 2.7706027030944824
    },
    {
      "auxiliary_loss_clip": 0.01123185,
      "auxiliary_loss_mlp": 0.01057136,
      "balance_loss_clip": 1.0518471,
      "balance_loss_mlp": 1.03554714,
      "epoch": 0.12199007966330978,
      "flos": 22017119783040.0,
      "grad_norm": 2.0550419223931193,
      "language_loss": 0.76802504,
      "learning_rate": 3.911887531387839e-06,
      "loss": 0.78982824,
      "num_input_tokens_seen": 43879030,
      "step": 2029,
      "time_per_iteration": 2.732637405395508
    },
    {
      "auxiliary_loss_clip": 0.01156591,
      "auxiliary_loss_mlp": 0.01052355,
      "balance_loss_clip": 1.05253625,
      "balance_loss_mlp": 1.03107572,
      "epoch": 0.12205020291597775,
      "flos": 23295216493440.0,
      "grad_norm": 1.707195979328818,
      "language_loss": 0.79164296,
      "learning_rate": 3.911773168719313e-06,
      "loss": 0.81373239,
      "num_input_tokens_seen": 43898505,
      "step": 2030,
      "time_per_iteration": 2.7254061698913574
    },
    {
      "auxiliary_loss_clip": 0.0116997,
      "auxiliary_loss_mlp": 0.01051357,
      "balance_loss_clip": 1.05618095,
      "balance_loss_mlp": 1.02930319,
      "epoch": 0.12211032616864573,
      "flos": 26031609469440.0,
      "grad_norm": 3.038077546298312,
      "language_loss": 0.74411637,
      "learning_rate": 3.911658733556155e-06,
      "loss": 0.76632965,
      "num_input_tokens_seen": 43917945,
      "step": 2031,
      "time_per_iteration": 2.6711080074310303
    },
    {
      "auxiliary_loss_clip": 0.01174332,
      "auxiliary_loss_mlp": 0.01045812,
      "balance_loss_clip": 1.05888343,
      "balance_loss_mlp": 1.02545118,
      "epoch": 0.12217044942131369,
      "flos": 20410943224320.0,
      "grad_norm": 1.7636188348969384,
      "language_loss": 0.75230348,
      "learning_rate": 3.911544225902707e-06,
      "loss": 0.7745049,
      "num_input_tokens_seen": 43937385,
      "step": 2032,
      "time_per_iteration": 2.7134530544281006
    },
    {
      "auxiliary_loss_clip": 0.01152363,
      "auxiliary_loss_mlp": 0.01045735,
      "balance_loss_clip": 1.05129802,
      "balance_loss_mlp": 1.02538586,
      "epoch": 0.12223057267398166,
      "flos": 22857142222080.0,
      "grad_norm": 1.5809359138264147,
      "language_loss": 0.89502287,
      "learning_rate": 3.911429645763311e-06,
      "loss": 0.91700387,
      "num_input_tokens_seen": 43958130,
      "step": 2033,
      "time_per_iteration": 2.7105965614318848
    },
    {
      "auxiliary_loss_clip": 0.01155694,
      "auxiliary_loss_mlp": 0.01051169,
      "balance_loss_clip": 1.05740523,
      "balance_loss_mlp": 1.03005767,
      "epoch": 0.12229069592664964,
      "flos": 20047563285120.0,
      "grad_norm": 1.9580868921695649,
      "language_loss": 0.65195286,
      "learning_rate": 3.911314993142311e-06,
      "loss": 0.67402148,
      "num_input_tokens_seen": 43976800,
      "step": 2034,
      "time_per_iteration": 4.222668886184692
    },
    {
      "auxiliary_loss_clip": 0.01152239,
      "auxiliary_loss_mlp": 0.01055659,
      "balance_loss_clip": 1.05550218,
      "balance_loss_mlp": 1.0327704,
      "epoch": 0.1223508191793176,
      "flos": 22274240313600.0,
      "grad_norm": 1.6376942269871653,
      "language_loss": 0.76459455,
      "learning_rate": 3.911200268044055e-06,
      "loss": 0.78667355,
      "num_input_tokens_seen": 43996620,
      "step": 2035,
      "time_per_iteration": 2.7306556701660156
    },
    {
      "auxiliary_loss_clip": 0.01176703,
      "auxiliary_loss_mlp": 0.01050008,
      "balance_loss_clip": 1.0577215,
      "balance_loss_mlp": 1.02798975,
      "epoch": 0.12241094243198557,
      "flos": 21285978445440.0,
      "grad_norm": 1.8460180606974623,
      "language_loss": 0.71294892,
      "learning_rate": 3.911085470472892e-06,
      "loss": 0.73521602,
      "num_input_tokens_seen": 44016175,
      "step": 2036,
      "time_per_iteration": 2.7327258586883545
    },
    {
      "auxiliary_loss_clip": 0.01144473,
      "auxiliary_loss_mlp": 0.01058389,
      "balance_loss_clip": 1.05778408,
      "balance_loss_mlp": 1.03623962,
      "epoch": 0.12247106568465355,
      "flos": 17382381022080.0,
      "grad_norm": 1.5772021569883852,
      "language_loss": 0.83130831,
      "learning_rate": 3.910970600433178e-06,
      "loss": 0.85333693,
      "num_input_tokens_seen": 44035060,
      "step": 2037,
      "time_per_iteration": 4.248440742492676
    },
    {
      "auxiliary_loss_clip": 0.01153641,
      "auxiliary_loss_mlp": 0.01060257,
      "balance_loss_clip": 1.0556947,
      "balance_loss_mlp": 1.0366174,
      "epoch": 0.12253118893732151,
      "flos": 27045438842880.0,
      "grad_norm": 2.676780030246967,
      "language_loss": 0.79765236,
      "learning_rate": 3.910855657929267e-06,
      "loss": 0.81979132,
      "num_input_tokens_seen": 44053330,
      "step": 2038,
      "time_per_iteration": 2.7321341037750244
    },
    {
      "auxiliary_loss_clip": 0.010642,
      "auxiliary_loss_mlp": 0.00759248,
      "balance_loss_clip": 1.02961969,
      "balance_loss_mlp": 1.00006962,
      "epoch": 0.12259131218998948,
      "flos": 53861518368000.0,
      "grad_norm": 0.8248048644272604,
      "language_loss": 0.58659601,
      "learning_rate": 3.910740642965518e-06,
      "loss": 0.6048305,
      "num_input_tokens_seen": 44107575,
      "step": 2039,
      "time_per_iteration": 4.739040851593018
    },
    {
      "auxiliary_loss_clip": 0.01128,
      "auxiliary_loss_mlp": 0.01064411,
      "balance_loss_clip": 1.05292714,
      "balance_loss_mlp": 1.03912663,
      "epoch": 0.12265143544265744,
      "flos": 17891917401600.0,
      "grad_norm": 2.1548467753138136,
      "language_loss": 0.80099291,
      "learning_rate": 3.910625555546292e-06,
      "loss": 0.82291704,
      "num_input_tokens_seen": 44126075,
      "step": 2040,
      "time_per_iteration": 2.723247766494751
    },
    {
      "auxiliary_loss_clip": 0.01149343,
      "auxiliary_loss_mlp": 0.01058534,
      "balance_loss_clip": 1.05517352,
      "balance_loss_mlp": 1.03673029,
      "epoch": 0.12271155869532542,
      "flos": 21799932197760.0,
      "grad_norm": 1.8247690225218605,
      "language_loss": 0.82841176,
      "learning_rate": 3.910510395675953e-06,
      "loss": 0.85049051,
      "num_input_tokens_seen": 44145605,
      "step": 2041,
      "time_per_iteration": 2.699110984802246
    },
    {
      "auxiliary_loss_clip": 0.01136001,
      "auxiliary_loss_mlp": 0.01053451,
      "balance_loss_clip": 1.05120957,
      "balance_loss_mlp": 1.03061032,
      "epoch": 0.12277168194799339,
      "flos": 19828759587840.0,
      "grad_norm": 1.9386136063873771,
      "language_loss": 0.67272276,
      "learning_rate": 3.9103951633588694e-06,
      "loss": 0.69461727,
      "num_input_tokens_seen": 44164770,
      "step": 2042,
      "time_per_iteration": 2.7042133808135986
    },
    {
      "auxiliary_loss_clip": 0.01133115,
      "auxiliary_loss_mlp": 0.01056941,
      "balance_loss_clip": 1.05079007,
      "balance_loss_mlp": 1.03517294,
      "epoch": 0.12283180520066135,
      "flos": 23221024951680.0,
      "grad_norm": 1.912164915278887,
      "language_loss": 0.81765604,
      "learning_rate": 3.910279858599409e-06,
      "loss": 0.83955657,
      "num_input_tokens_seen": 44184025,
      "step": 2043,
      "time_per_iteration": 2.6942050457000732
    },
    {
      "auxiliary_loss_clip": 0.01146416,
      "auxiliary_loss_mlp": 0.01052365,
      "balance_loss_clip": 1.05161905,
      "balance_loss_mlp": 1.03040695,
      "epoch": 0.12289192845332933,
      "flos": 18588476920320.0,
      "grad_norm": 1.7894844734354058,
      "language_loss": 0.80192459,
      "learning_rate": 3.910164481401946e-06,
      "loss": 0.82391244,
      "num_input_tokens_seen": 44202950,
      "step": 2044,
      "time_per_iteration": 2.6227192878723145
    },
    {
      "auxiliary_loss_clip": 0.01116285,
      "auxiliary_loss_mlp": 0.01052013,
      "balance_loss_clip": 1.05284619,
      "balance_loss_mlp": 1.03055525,
      "epoch": 0.1229520517059973,
      "flos": 25769532862080.0,
      "grad_norm": 1.7152742607840916,
      "language_loss": 0.7794897,
      "learning_rate": 3.910049031770853e-06,
      "loss": 0.80117267,
      "num_input_tokens_seen": 44221115,
      "step": 2045,
      "time_per_iteration": 2.769017219543457
    },
    {
      "auxiliary_loss_clip": 0.01163545,
      "auxiliary_loss_mlp": 0.01060468,
      "balance_loss_clip": 1.05796146,
      "balance_loss_mlp": 1.03827095,
      "epoch": 0.12301217495866526,
      "flos": 20887154760960.0,
      "grad_norm": 1.852572781372854,
      "language_loss": 0.67284262,
      "learning_rate": 3.90993350971051e-06,
      "loss": 0.69508278,
      "num_input_tokens_seen": 44240575,
      "step": 2046,
      "time_per_iteration": 2.6377944946289062
    },
    {
      "auxiliary_loss_clip": 0.01173803,
      "auxiliary_loss_mlp": 0.01053755,
      "balance_loss_clip": 1.06010675,
      "balance_loss_mlp": 1.03202295,
      "epoch": 0.12307229821133324,
      "flos": 22378811783040.0,
      "grad_norm": 4.982373490718116,
      "language_loss": 0.72730684,
      "learning_rate": 3.909817915225297e-06,
      "loss": 0.74958241,
      "num_input_tokens_seen": 44257145,
      "step": 2047,
      "time_per_iteration": 2.5791239738464355
    },
    {
      "auxiliary_loss_clip": 0.01155159,
      "auxiliary_loss_mlp": 0.01060632,
      "balance_loss_clip": 1.05398846,
      "balance_loss_mlp": 1.03817296,
      "epoch": 0.1231324214640012,
      "flos": 23367396873600.0,
      "grad_norm": 1.8194194024321948,
      "language_loss": 0.76583183,
      "learning_rate": 3.909702248319597e-06,
      "loss": 0.78798974,
      "num_input_tokens_seen": 44278035,
      "step": 2048,
      "time_per_iteration": 2.6997592449188232
    },
    {
      "auxiliary_loss_clip": 0.01146796,
      "auxiliary_loss_mlp": 0.01047309,
      "balance_loss_clip": 1.05524468,
      "balance_loss_mlp": 1.02798486,
      "epoch": 0.12319254471666917,
      "flos": 23767154311680.0,
      "grad_norm": 1.8097490634569602,
      "language_loss": 0.85359102,
      "learning_rate": 3.909586508997797e-06,
      "loss": 0.87553203,
      "num_input_tokens_seen": 44296980,
      "step": 2049,
      "time_per_iteration": 2.739617109298706
    },
    {
      "auxiliary_loss_clip": 0.01120276,
      "auxiliary_loss_mlp": 0.01050145,
      "balance_loss_clip": 1.0533725,
      "balance_loss_mlp": 1.02887857,
      "epoch": 0.12325266796933713,
      "flos": 23550146294400.0,
      "grad_norm": 2.6582136339172724,
      "language_loss": 0.75563407,
      "learning_rate": 3.909470697264285e-06,
      "loss": 0.77733827,
      "num_input_tokens_seen": 44318005,
      "step": 2050,
      "time_per_iteration": 2.7814078330993652
    },
    {
      "auxiliary_loss_clip": 0.01138568,
      "auxiliary_loss_mlp": 0.01057939,
      "balance_loss_clip": 1.05428278,
      "balance_loss_mlp": 1.03608823,
      "epoch": 0.12331279122200511,
      "flos": 24423996366720.0,
      "grad_norm": 1.81408967902731,
      "language_loss": 0.81166679,
      "learning_rate": 3.909354813123452e-06,
      "loss": 0.83363187,
      "num_input_tokens_seen": 44335260,
      "step": 2051,
      "time_per_iteration": 2.7555224895477295
    },
    {
      "auxiliary_loss_clip": 0.01171646,
      "auxiliary_loss_mlp": 0.00779218,
      "balance_loss_clip": 1.05882978,
      "balance_loss_mlp": 0.99996465,
      "epoch": 0.12337291447467308,
      "flos": 25484294960640.0,
      "grad_norm": 1.8885516327307212,
      "language_loss": 0.80445349,
      "learning_rate": 3.909238856579693e-06,
      "loss": 0.82396215,
      "num_input_tokens_seen": 44355315,
      "step": 2052,
      "time_per_iteration": 2.7676405906677246
    },
    {
      "auxiliary_loss_clip": 0.01165489,
      "auxiliary_loss_mlp": 0.010569,
      "balance_loss_clip": 1.0581975,
      "balance_loss_mlp": 1.03537059,
      "epoch": 0.12343303772734104,
      "flos": 23550002640000.0,
      "grad_norm": 2.171205541070781,
      "language_loss": 0.73676848,
      "learning_rate": 3.909122827637406e-06,
      "loss": 0.75899243,
      "num_input_tokens_seen": 44373020,
      "step": 2053,
      "time_per_iteration": 2.648609161376953
    },
    {
      "auxiliary_loss_clip": 0.01168883,
      "auxiliary_loss_mlp": 0.00778478,
      "balance_loss_clip": 1.05302441,
      "balance_loss_mlp": 0.99995315,
      "epoch": 0.12349316098000902,
      "flos": 47557074867840.0,
      "grad_norm": 1.5051513438882418,
      "language_loss": 0.7413671,
      "learning_rate": 3.909006726300991e-06,
      "loss": 0.76084077,
      "num_input_tokens_seen": 44397525,
      "step": 2054,
      "time_per_iteration": 2.871469020843506
    },
    {
      "auxiliary_loss_clip": 0.01147607,
      "auxiliary_loss_mlp": 0.01044612,
      "balance_loss_clip": 1.05402803,
      "balance_loss_mlp": 1.02482307,
      "epoch": 0.12355328423267699,
      "flos": 25045969294080.0,
      "grad_norm": 4.50189877271012,
      "language_loss": 0.85417157,
      "learning_rate": 3.908890552574849e-06,
      "loss": 0.8760938,
      "num_input_tokens_seen": 44415890,
      "step": 2055,
      "time_per_iteration": 2.7136077880859375
    },
    {
      "auxiliary_loss_clip": 0.01133829,
      "auxiliary_loss_mlp": 0.01047458,
      "balance_loss_clip": 1.05999517,
      "balance_loss_mlp": 1.02802706,
      "epoch": 0.12361340748534495,
      "flos": 27709140395520.0,
      "grad_norm": 2.0629908776416688,
      "language_loss": 0.77506042,
      "learning_rate": 3.908774306463384e-06,
      "loss": 0.79687333,
      "num_input_tokens_seen": 44436625,
      "step": 2056,
      "time_per_iteration": 2.83107852935791
    },
    {
      "auxiliary_loss_clip": 0.01158234,
      "auxiliary_loss_mlp": 0.01055,
      "balance_loss_clip": 1.05444396,
      "balance_loss_mlp": 1.03405499,
      "epoch": 0.12367353073801293,
      "flos": 26140598311680.0,
      "grad_norm": 1.9893743253373262,
      "language_loss": 0.83361745,
      "learning_rate": 3.908657987971009e-06,
      "loss": 0.85574985,
      "num_input_tokens_seen": 44455265,
      "step": 2057,
      "time_per_iteration": 2.6987085342407227
    },
    {
      "auxiliary_loss_clip": 0.01141319,
      "auxiliary_loss_mlp": 0.01051708,
      "balance_loss_clip": 1.05057144,
      "balance_loss_mlp": 1.02991605,
      "epoch": 0.1237336539906809,
      "flos": 25156035544320.0,
      "grad_norm": 1.4905135493793764,
      "language_loss": 0.77818203,
      "learning_rate": 3.90854159710213e-06,
      "loss": 0.80011231,
      "num_input_tokens_seen": 44475815,
      "step": 2058,
      "time_per_iteration": 2.7149016857147217
    },
    {
      "auxiliary_loss_clip": 0.01138087,
      "auxiliary_loss_mlp": 0.01058134,
      "balance_loss_clip": 1.05117273,
      "balance_loss_mlp": 1.03482866,
      "epoch": 0.12379377724334886,
      "flos": 15304589867520.0,
      "grad_norm": 1.8387803476985631,
      "language_loss": 0.8342883,
      "learning_rate": 3.9084251338611624e-06,
      "loss": 0.85625052,
      "num_input_tokens_seen": 44494045,
      "step": 2059,
      "time_per_iteration": 2.7030091285705566
    },
    {
      "auxiliary_loss_clip": 0.01133517,
      "auxiliary_loss_mlp": 0.01057399,
      "balance_loss_clip": 1.05123472,
      "balance_loss_mlp": 1.03445077,
      "epoch": 0.12385390049601683,
      "flos": 21316717509120.0,
      "grad_norm": 2.7478129466394217,
      "language_loss": 0.81420219,
      "learning_rate": 3.908308598252523e-06,
      "loss": 0.83611137,
      "num_input_tokens_seen": 44509120,
      "step": 2060,
      "time_per_iteration": 2.738499402999878
    },
    {
      "auxiliary_loss_clip": 0.01150334,
      "auxiliary_loss_mlp": 0.01054424,
      "balance_loss_clip": 1.05367386,
      "balance_loss_mlp": 1.0315125,
      "epoch": 0.1239140237486848,
      "flos": 15116309752320.0,
      "grad_norm": 1.8699548955873522,
      "language_loss": 0.86224365,
      "learning_rate": 3.9081919902806306e-06,
      "loss": 0.88429129,
      "num_input_tokens_seen": 44525780,
      "step": 2061,
      "time_per_iteration": 2.6492960453033447
    },
    {
      "auxiliary_loss_clip": 0.0115523,
      "auxiliary_loss_mlp": 0.01050307,
      "balance_loss_clip": 1.05506253,
      "balance_loss_mlp": 1.03031528,
      "epoch": 0.12397414700135277,
      "flos": 21976791788160.0,
      "grad_norm": 2.006361909654615,
      "language_loss": 0.84949362,
      "learning_rate": 3.908075309949906e-06,
      "loss": 0.87154901,
      "num_input_tokens_seen": 44543125,
      "step": 2062,
      "time_per_iteration": 2.5925393104553223
    },
    {
      "auxiliary_loss_clip": 0.01124676,
      "auxiliary_loss_mlp": 0.01058304,
      "balance_loss_clip": 1.05198252,
      "balance_loss_mlp": 1.03498697,
      "epoch": 0.12403427025402074,
      "flos": 13400892956160.0,
      "grad_norm": 1.6181471799462952,
      "language_loss": 0.78765064,
      "learning_rate": 3.907958557264774e-06,
      "loss": 0.80948043,
      "num_input_tokens_seen": 44560275,
      "step": 2063,
      "time_per_iteration": 2.7551674842834473
    },
    {
      "auxiliary_loss_clip": 0.01124369,
      "auxiliary_loss_mlp": 0.01057465,
      "balance_loss_clip": 1.05492854,
      "balance_loss_mlp": 1.03450513,
      "epoch": 0.12409439350668872,
      "flos": 15304374385920.0,
      "grad_norm": 2.9315517002695017,
      "language_loss": 0.79452097,
      "learning_rate": 3.907841732229663e-06,
      "loss": 0.81633931,
      "num_input_tokens_seen": 44577640,
      "step": 2064,
      "time_per_iteration": 2.699711322784424
    },
    {
      "auxiliary_loss_clip": 0.01144709,
      "auxiliary_loss_mlp": 0.01058768,
      "balance_loss_clip": 1.05316699,
      "balance_loss_mlp": 1.03847849,
      "epoch": 0.12415451675935668,
      "flos": 25009376313600.0,
      "grad_norm": 2.5611248351266016,
      "language_loss": 0.92676973,
      "learning_rate": 3.907724834849002e-06,
      "loss": 0.9488045,
      "num_input_tokens_seen": 44594860,
      "step": 2065,
      "time_per_iteration": 2.7114996910095215
    },
    {
      "auxiliary_loss_clip": 0.01147841,
      "auxiliary_loss_mlp": 0.01052058,
      "balance_loss_clip": 1.05113554,
      "balance_loss_mlp": 1.02943158,
      "epoch": 0.12421464001202465,
      "flos": 23659673840640.0,
      "grad_norm": 1.7498294279318665,
      "language_loss": 0.80540735,
      "learning_rate": 3.907607865127225e-06,
      "loss": 0.82740629,
      "num_input_tokens_seen": 44614780,
      "step": 2066,
      "time_per_iteration": 2.6958389282226562
    },
    {
      "auxiliary_loss_clip": 0.01030831,
      "auxiliary_loss_mlp": 0.01051436,
      "balance_loss_clip": 1.02768898,
      "balance_loss_mlp": 1.04884958,
      "epoch": 0.12427476326469263,
      "flos": 65732904345600.0,
      "grad_norm": 0.8715885531008962,
      "language_loss": 0.63299954,
      "learning_rate": 3.907490823068766e-06,
      "loss": 0.6538223,
      "num_input_tokens_seen": 44671240,
      "step": 2067,
      "time_per_iteration": 3.200000762939453
    },
    {
      "auxiliary_loss_clip": 0.01117858,
      "auxiliary_loss_mlp": 0.01057985,
      "balance_loss_clip": 1.04878855,
      "balance_loss_mlp": 1.0344646,
      "epoch": 0.12433488651736059,
      "flos": 24535427333760.0,
      "grad_norm": 1.9218217735084064,
      "language_loss": 0.93783462,
      "learning_rate": 3.907373708678063e-06,
      "loss": 0.959593,
      "num_input_tokens_seen": 44691050,
      "step": 2068,
      "time_per_iteration": 2.7631025314331055
    },
    {
      "auxiliary_loss_clip": 0.01166393,
      "auxiliary_loss_mlp": 0.0105657,
      "balance_loss_clip": 1.05994427,
      "balance_loss_mlp": 1.03697169,
      "epoch": 0.12439500977002856,
      "flos": 21031659175680.0,
      "grad_norm": 1.8717926968048342,
      "language_loss": 0.80861229,
      "learning_rate": 3.9072565219595596e-06,
      "loss": 0.83084196,
      "num_input_tokens_seen": 44709850,
      "step": 2069,
      "time_per_iteration": 2.6630098819732666
    },
    {
      "auxiliary_loss_clip": 0.01113262,
      "auxiliary_loss_mlp": 0.01062592,
      "balance_loss_clip": 1.04863238,
      "balance_loss_mlp": 1.03963184,
      "epoch": 0.12445513302269653,
      "flos": 26830621555200.0,
      "grad_norm": 1.5649570979854035,
      "language_loss": 0.777978,
      "learning_rate": 3.907139262917696e-06,
      "loss": 0.79973656,
      "num_input_tokens_seen": 44731475,
      "step": 2070,
      "time_per_iteration": 2.7750463485717773
    },
    {
      "auxiliary_loss_clip": 0.01156875,
      "auxiliary_loss_mlp": 0.01052509,
      "balance_loss_clip": 1.05520415,
      "balance_loss_mlp": 1.03055048,
      "epoch": 0.1245152562753645,
      "flos": 18368919037440.0,
      "grad_norm": 2.2051981544638166,
      "language_loss": 0.80743957,
      "learning_rate": 3.907021931556922e-06,
      "loss": 0.8295334,
      "num_input_tokens_seen": 44749685,
      "step": 2071,
      "time_per_iteration": 2.654171943664551
    },
    {
      "auxiliary_loss_clip": 0.01154683,
      "auxiliary_loss_mlp": 0.01055767,
      "balance_loss_clip": 1.05492425,
      "balance_loss_mlp": 1.03405952,
      "epoch": 0.12457537952803246,
      "flos": 33107986200960.0,
      "grad_norm": 2.118828414072521,
      "language_loss": 0.78278041,
      "learning_rate": 3.906904527881684e-06,
      "loss": 0.80488491,
      "num_input_tokens_seen": 44772165,
      "step": 2072,
      "time_per_iteration": 2.753159284591675
    },
    {
      "auxiliary_loss_clip": 0.0114568,
      "auxiliary_loss_mlp": 0.01055287,
      "balance_loss_clip": 1.05651307,
      "balance_loss_mlp": 1.03381729,
      "epoch": 0.12463550278070043,
      "flos": 22270217990400.0,
      "grad_norm": 7.360489773093417,
      "language_loss": 0.752267,
      "learning_rate": 3.9067870518964355e-06,
      "loss": 0.77427667,
      "num_input_tokens_seen": 44790580,
      "step": 2073,
      "time_per_iteration": 2.6561899185180664
    },
    {
      "auxiliary_loss_clip": 0.01096485,
      "auxiliary_loss_mlp": 0.01053193,
      "balance_loss_clip": 1.04471385,
      "balance_loss_mlp": 1.03086543,
      "epoch": 0.12469562603336841,
      "flos": 14679025580160.0,
      "grad_norm": 1.9234955386089483,
      "language_loss": 0.90560025,
      "learning_rate": 3.906669503605631e-06,
      "loss": 0.92709696,
      "num_input_tokens_seen": 44806730,
      "step": 2074,
      "time_per_iteration": 2.7846343517303467
    },
    {
      "auxiliary_loss_clip": 0.01105332,
      "auxiliary_loss_mlp": 0.01056651,
      "balance_loss_clip": 1.04977274,
      "balance_loss_mlp": 1.03346491,
      "epoch": 0.12475574928603637,
      "flos": 24644775312000.0,
      "grad_norm": 2.8321626325497493,
      "language_loss": 0.83836985,
      "learning_rate": 3.906551883013728e-06,
      "loss": 0.8599897,
      "num_input_tokens_seen": 44825550,
      "step": 2075,
      "time_per_iteration": 4.412928342819214
    },
    {
      "auxiliary_loss_clip": 0.01107078,
      "auxiliary_loss_mlp": 0.01062819,
      "balance_loss_clip": 1.04380202,
      "balance_loss_mlp": 1.03972864,
      "epoch": 0.12481587253870434,
      "flos": 21762980081280.0,
      "grad_norm": 2.042892519020311,
      "language_loss": 0.73648787,
      "learning_rate": 3.9064341901251865e-06,
      "loss": 0.75818682,
      "num_input_tokens_seen": 44844155,
      "step": 2076,
      "time_per_iteration": 5.925223112106323
    },
    {
      "auxiliary_loss_clip": 0.01101731,
      "auxiliary_loss_mlp": 0.01048176,
      "balance_loss_clip": 1.04774427,
      "balance_loss_mlp": 1.02751708,
      "epoch": 0.12487599579137232,
      "flos": 21432529935360.0,
      "grad_norm": 1.8779339700875872,
      "language_loss": 0.7622484,
      "learning_rate": 3.906316424944469e-06,
      "loss": 0.78374755,
      "num_input_tokens_seen": 44863780,
      "step": 2077,
      "time_per_iteration": 2.70566987991333
    },
    {
      "auxiliary_loss_clip": 0.01156274,
      "auxiliary_loss_mlp": 0.01062042,
      "balance_loss_clip": 1.05365288,
      "balance_loss_mlp": 1.04001164,
      "epoch": 0.12493611904404028,
      "flos": 16107624276480.0,
      "grad_norm": 2.022280968605665,
      "language_loss": 0.82290226,
      "learning_rate": 3.906198587476043e-06,
      "loss": 0.84508544,
      "num_input_tokens_seen": 44881480,
      "step": 2078,
      "time_per_iteration": 4.302385568618774
    },
    {
      "auxiliary_loss_clip": 0.01144821,
      "auxiliary_loss_mlp": 0.01050482,
      "balance_loss_clip": 1.05281842,
      "balance_loss_mlp": 1.02855957,
      "epoch": 0.12499624229670825,
      "flos": 21580266574080.0,
      "grad_norm": 1.6413520418295044,
      "language_loss": 0.75195324,
      "learning_rate": 3.906080677724374e-06,
      "loss": 0.77390629,
      "num_input_tokens_seen": 44900390,
      "step": 2079,
      "time_per_iteration": 2.6915946006774902
    },
    {
      "auxiliary_loss_clip": 0.01166758,
      "auxiliary_loss_mlp": 0.01058474,
      "balance_loss_clip": 1.05881989,
      "balance_loss_mlp": 1.03696847,
      "epoch": 0.1250563655493762,
      "flos": 25699040421120.0,
      "grad_norm": 6.733284446627088,
      "language_loss": 0.83874094,
      "learning_rate": 3.905962695693935e-06,
      "loss": 0.86099327,
      "num_input_tokens_seen": 44920375,
      "step": 2080,
      "time_per_iteration": 2.7467572689056396
    },
    {
      "auxiliary_loss_clip": 0.01156163,
      "auxiliary_loss_mlp": 0.01059409,
      "balance_loss_clip": 1.05525088,
      "balance_loss_mlp": 1.03885686,
      "epoch": 0.12511648880204418,
      "flos": 16909509450240.0,
      "grad_norm": 1.8581885454518776,
      "language_loss": 0.84644079,
      "learning_rate": 3.9058446413892e-06,
      "loss": 0.86859655,
      "num_input_tokens_seen": 44938415,
      "step": 2081,
      "time_per_iteration": 2.685875654220581
    },
    {
      "auxiliary_loss_clip": 0.01156835,
      "auxiliary_loss_mlp": 0.01046398,
      "balance_loss_clip": 1.05375946,
      "balance_loss_mlp": 1.02594149,
      "epoch": 0.12517661205471217,
      "flos": 17567500740480.0,
      "grad_norm": 1.8191819349610059,
      "language_loss": 0.76739037,
      "learning_rate": 3.905726514814646e-06,
      "loss": 0.78942269,
      "num_input_tokens_seen": 44957135,
      "step": 2082,
      "time_per_iteration": 2.6133053302764893
    },
    {
      "auxiliary_loss_clip": 0.01152911,
      "auxiliary_loss_mlp": 0.0104632,
      "balance_loss_clip": 1.05701911,
      "balance_loss_mlp": 1.02463615,
      "epoch": 0.12523673530738014,
      "flos": 16033791870720.0,
      "grad_norm": 2.5415589476696265,
      "language_loss": 0.79044539,
      "learning_rate": 3.9056083159747495e-06,
      "loss": 0.81243765,
      "num_input_tokens_seen": 44974480,
      "step": 2083,
      "time_per_iteration": 2.6963307857513428
    },
    {
      "auxiliary_loss_clip": 0.01147874,
      "auxiliary_loss_mlp": 0.01047351,
      "balance_loss_clip": 1.05509973,
      "balance_loss_mlp": 1.02421284,
      "epoch": 0.1252968585600481,
      "flos": 18807747494400.0,
      "grad_norm": 2.1696249857299,
      "language_loss": 0.89831448,
      "learning_rate": 3.9054900448739966e-06,
      "loss": 0.92026675,
      "num_input_tokens_seen": 44990310,
      "step": 2084,
      "time_per_iteration": 2.6770403385162354
    },
    {
      "auxiliary_loss_clip": 0.01131068,
      "auxiliary_loss_mlp": 0.01048299,
      "balance_loss_clip": 1.05299771,
      "balance_loss_mlp": 1.02729464,
      "epoch": 0.12535698181271607,
      "flos": 27271568914560.0,
      "grad_norm": 1.8896331095253402,
      "language_loss": 0.80354226,
      "learning_rate": 3.905371701516869e-06,
      "loss": 0.82533598,
      "num_input_tokens_seen": 45010720,
      "step": 2085,
      "time_per_iteration": 2.749783515930176
    },
    {
      "auxiliary_loss_clip": 0.01170318,
      "auxiliary_loss_mlp": 0.01051018,
      "balance_loss_clip": 1.05725896,
      "balance_loss_mlp": 1.03001356,
      "epoch": 0.12541710506538403,
      "flos": 22054107813120.0,
      "grad_norm": 1.8300316094254767,
      "language_loss": 0.88228154,
      "learning_rate": 3.905253285907856e-06,
      "loss": 0.90449488,
      "num_input_tokens_seen": 45030360,
      "step": 2086,
      "time_per_iteration": 2.603515148162842
    },
    {
      "auxiliary_loss_clip": 0.01134598,
      "auxiliary_loss_mlp": 0.01044925,
      "balance_loss_clip": 1.05278981,
      "balance_loss_mlp": 1.02522027,
      "epoch": 0.125477228318052,
      "flos": 12603173760000.0,
      "grad_norm": 2.0471238132540344,
      "language_loss": 0.86819696,
      "learning_rate": 3.905134798051447e-06,
      "loss": 0.88999224,
      "num_input_tokens_seen": 45045085,
      "step": 2087,
      "time_per_iteration": 2.6265859603881836
    },
    {
      "auxiliary_loss_clip": 0.01146999,
      "auxiliary_loss_mlp": 0.01058875,
      "balance_loss_clip": 1.05599046,
      "balance_loss_mlp": 1.03651142,
      "epoch": 0.12553735157071996,
      "flos": 23878549365120.0,
      "grad_norm": 2.3362397674907758,
      "language_loss": 0.73027468,
      "learning_rate": 3.905016237952136e-06,
      "loss": 0.75233346,
      "num_input_tokens_seen": 45065145,
      "step": 2088,
      "time_per_iteration": 2.65324330329895
    },
    {
      "auxiliary_loss_clip": 0.01062529,
      "auxiliary_loss_mlp": 0.01013405,
      "balance_loss_clip": 1.02985716,
      "balance_loss_mlp": 1.01079392,
      "epoch": 0.12559747482338796,
      "flos": 69920841830400.0,
      "grad_norm": 0.7742255614948045,
      "language_loss": 0.61767036,
      "learning_rate": 3.904897605614418e-06,
      "loss": 0.6384297,
      "num_input_tokens_seen": 45126230,
      "step": 2089,
      "time_per_iteration": 3.1219804286956787
    },
    {
      "auxiliary_loss_clip": 0.01149606,
      "auxiliary_loss_mlp": 0.01060841,
      "balance_loss_clip": 1.05670094,
      "balance_loss_mlp": 1.0388943,
      "epoch": 0.12565759807605592,
      "flos": 24279563779200.0,
      "grad_norm": 1.817095421446176,
      "language_loss": 0.7781918,
      "learning_rate": 3.904778901042793e-06,
      "loss": 0.80029625,
      "num_input_tokens_seen": 45145545,
      "step": 2090,
      "time_per_iteration": 2.700425863265991
    },
    {
      "auxiliary_loss_clip": 0.01046946,
      "auxiliary_loss_mlp": 0.01013884,
      "balance_loss_clip": 1.03125095,
      "balance_loss_mlp": 1.01101136,
      "epoch": 0.12571772132872389,
      "flos": 56451180286080.0,
      "grad_norm": 0.760599485634597,
      "language_loss": 0.59434772,
      "learning_rate": 3.90466012424176e-06,
      "loss": 0.61495602,
      "num_input_tokens_seen": 45206845,
      "step": 2091,
      "time_per_iteration": 3.0814294815063477
    },
    {
      "auxiliary_loss_clip": 0.01159814,
      "auxiliary_loss_mlp": 0.01060546,
      "balance_loss_clip": 1.05760789,
      "balance_loss_mlp": 1.041067,
      "epoch": 0.12577784458139185,
      "flos": 41245846675200.0,
      "grad_norm": 1.6552462178493936,
      "language_loss": 0.62916517,
      "learning_rate": 3.904541275215825e-06,
      "loss": 0.6513688,
      "num_input_tokens_seen": 45228495,
      "step": 2092,
      "time_per_iteration": 2.7813880443573
    },
    {
      "auxiliary_loss_clip": 0.01147016,
      "auxiliary_loss_mlp": 0.01061963,
      "balance_loss_clip": 1.05395663,
      "balance_loss_mlp": 1.04069614,
      "epoch": 0.12583796783405982,
      "flos": 19755501799680.0,
      "grad_norm": 2.279616692029291,
      "language_loss": 0.80507946,
      "learning_rate": 3.904422353969493e-06,
      "loss": 0.82716924,
      "num_input_tokens_seen": 45245720,
      "step": 2093,
      "time_per_iteration": 2.6768014430999756
    },
    {
      "auxiliary_loss_clip": 0.01146976,
      "auxiliary_loss_mlp": 0.01075616,
      "balance_loss_clip": 1.0524025,
      "balance_loss_mlp": 1.05380058,
      "epoch": 0.12589809108672778,
      "flos": 22602104680320.0,
      "grad_norm": 1.7347385846840702,
      "language_loss": 0.76003867,
      "learning_rate": 3.904303360507276e-06,
      "loss": 0.78226459,
      "num_input_tokens_seen": 45265650,
      "step": 2094,
      "time_per_iteration": 2.6730611324310303
    },
    {
      "auxiliary_loss_clip": 0.01117887,
      "auxiliary_loss_mlp": 0.01069309,
      "balance_loss_clip": 1.0500071,
      "balance_loss_mlp": 1.04892457,
      "epoch": 0.12595821433939577,
      "flos": 45222845541120.0,
      "grad_norm": 1.5703706409155747,
      "language_loss": 0.76664734,
      "learning_rate": 3.9041842948336835e-06,
      "loss": 0.78851926,
      "num_input_tokens_seen": 45287790,
      "step": 2095,
      "time_per_iteration": 2.958367109298706
    },
    {
      "auxiliary_loss_clip": 0.01147751,
      "auxiliary_loss_mlp": 0.01058477,
      "balance_loss_clip": 1.05202031,
      "balance_loss_mlp": 1.03782988,
      "epoch": 0.12601833759206374,
      "flos": 14319811618560.0,
      "grad_norm": 2.2556524892449326,
      "language_loss": 0.83266854,
      "learning_rate": 3.904065156953232e-06,
      "loss": 0.85473078,
      "num_input_tokens_seen": 45305720,
      "step": 2096,
      "time_per_iteration": 2.7097342014312744
    },
    {
      "auxiliary_loss_clip": 0.01163652,
      "auxiliary_loss_mlp": 0.01056552,
      "balance_loss_clip": 1.05806553,
      "balance_loss_mlp": 1.03577375,
      "epoch": 0.1260784608447317,
      "flos": 21288241002240.0,
      "grad_norm": 1.7589400475615893,
      "language_loss": 0.75478256,
      "learning_rate": 3.903945946870439e-06,
      "loss": 0.77698463,
      "num_input_tokens_seen": 45325290,
      "step": 2097,
      "time_per_iteration": 2.642056703567505
    },
    {
      "auxiliary_loss_clip": 0.01156719,
      "auxiliary_loss_mlp": 0.01063976,
      "balance_loss_clip": 1.05648863,
      "balance_loss_mlp": 1.04527175,
      "epoch": 0.12613858409739967,
      "flos": 26251311006720.0,
      "grad_norm": 1.8828235460619742,
      "language_loss": 0.87110066,
      "learning_rate": 3.9038266645898246e-06,
      "loss": 0.89330757,
      "num_input_tokens_seen": 45344465,
      "step": 2098,
      "time_per_iteration": 2.63826584815979
    },
    {
      "auxiliary_loss_clip": 0.01117414,
      "auxiliary_loss_mlp": 0.01058025,
      "balance_loss_clip": 1.04983974,
      "balance_loss_mlp": 1.03475559,
      "epoch": 0.12619870735006763,
      "flos": 21579979265280.0,
      "grad_norm": 1.8855647331078333,
      "language_loss": 0.69494271,
      "learning_rate": 3.903707310115912e-06,
      "loss": 0.7166971,
      "num_input_tokens_seen": 45362465,
      "step": 2099,
      "time_per_iteration": 2.7813057899475098
    },
    {
      "auxiliary_loss_clip": 0.01142696,
      "auxiliary_loss_mlp": 0.01061431,
      "balance_loss_clip": 1.04979372,
      "balance_loss_mlp": 1.03923464,
      "epoch": 0.1262588306027356,
      "flos": 23367037737600.0,
      "grad_norm": 2.0457253500590498,
      "language_loss": 0.81949925,
      "learning_rate": 3.903587883453228e-06,
      "loss": 0.84154058,
      "num_input_tokens_seen": 45382700,
      "step": 2100,
      "time_per_iteration": 2.704871416091919
    },
    {
      "auxiliary_loss_clip": 0.01159613,
      "auxiliary_loss_mlp": 0.01055067,
      "balance_loss_clip": 1.0620985,
      "balance_loss_mlp": 1.03408623,
      "epoch": 0.12631895385540357,
      "flos": 23949185460480.0,
      "grad_norm": 1.7810176086536167,
      "language_loss": 0.80399859,
      "learning_rate": 3.903468384606302e-06,
      "loss": 0.82614541,
      "num_input_tokens_seen": 45401005,
      "step": 2101,
      "time_per_iteration": 2.7071452140808105
    },
    {
      "auxiliary_loss_clip": 0.0106985,
      "auxiliary_loss_mlp": 0.01010859,
      "balance_loss_clip": 1.02823138,
      "balance_loss_mlp": 1.00803375,
      "epoch": 0.12637907710807156,
      "flos": 70282138780800.0,
      "grad_norm": 0.7128618749962091,
      "language_loss": 0.57087427,
      "learning_rate": 3.903348813579662e-06,
      "loss": 0.59168136,
      "num_input_tokens_seen": 45466555,
      "step": 2102,
      "time_per_iteration": 3.20320987701416
    },
    {
      "auxiliary_loss_clip": 0.01140495,
      "auxiliary_loss_mlp": 0.01056574,
      "balance_loss_clip": 1.053671,
      "balance_loss_mlp": 1.03661788,
      "epoch": 0.12643920036073952,
      "flos": 18915084311040.0,
      "grad_norm": 2.0306165352193988,
      "language_loss": 0.93653679,
      "learning_rate": 3.903229170377845e-06,
      "loss": 0.95850742,
      "num_input_tokens_seen": 45485165,
      "step": 2103,
      "time_per_iteration": 2.6628894805908203
    },
    {
      "auxiliary_loss_clip": 0.01144405,
      "auxiliary_loss_mlp": 0.01040745,
      "balance_loss_clip": 1.04991472,
      "balance_loss_mlp": 1.02174282,
      "epoch": 0.1264993236134075,
      "flos": 27782470010880.0,
      "grad_norm": 1.5962316578756222,
      "language_loss": 0.7804662,
      "learning_rate": 3.903109455005387e-06,
      "loss": 0.80231774,
      "num_input_tokens_seen": 45504630,
      "step": 2104,
      "time_per_iteration": 2.6215474605560303
    },
    {
      "auxiliary_loss_clip": 0.01135927,
      "auxiliary_loss_mlp": 0.01056343,
      "balance_loss_clip": 1.05414486,
      "balance_loss_mlp": 1.03683996,
      "epoch": 0.12655944686607545,
      "flos": 24754697907840.0,
      "grad_norm": 1.7362499149688688,
      "language_loss": 0.80728614,
      "learning_rate": 3.902989667466828e-06,
      "loss": 0.82920885,
      "num_input_tokens_seen": 45524885,
      "step": 2105,
      "time_per_iteration": 2.74128794670105
    },
    {
      "auxiliary_loss_clip": 0.01162904,
      "auxiliary_loss_mlp": 0.01056367,
      "balance_loss_clip": 1.05482686,
      "balance_loss_mlp": 1.03514743,
      "epoch": 0.12661957011874342,
      "flos": 24133048202880.0,
      "grad_norm": 1.9810187943106816,
      "language_loss": 0.83402872,
      "learning_rate": 3.90286980776671e-06,
      "loss": 0.85622144,
      "num_input_tokens_seen": 45545000,
      "step": 2106,
      "time_per_iteration": 2.676694631576538
    },
    {
      "auxiliary_loss_clip": 0.01126632,
      "auxiliary_loss_mlp": 0.01052067,
      "balance_loss_clip": 1.05697966,
      "balance_loss_mlp": 1.03147984,
      "epoch": 0.12667969337141138,
      "flos": 24569614103040.0,
      "grad_norm": 1.6951691508845637,
      "language_loss": 0.73469931,
      "learning_rate": 3.902749875909578e-06,
      "loss": 0.7564863,
      "num_input_tokens_seen": 45564210,
      "step": 2107,
      "time_per_iteration": 2.7506372928619385
    },
    {
      "auxiliary_loss_clip": 0.01162931,
      "auxiliary_loss_mlp": 0.01044317,
      "balance_loss_clip": 1.05320692,
      "balance_loss_mlp": 1.02599406,
      "epoch": 0.12673981662407935,
      "flos": 22961677777920.0,
      "grad_norm": 2.0116792159666477,
      "language_loss": 0.79395336,
      "learning_rate": 3.90262987189998e-06,
      "loss": 0.81602579,
      "num_input_tokens_seen": 45583030,
      "step": 2108,
      "time_per_iteration": 2.6611146926879883
    },
    {
      "auxiliary_loss_clip": 0.01168073,
      "auxiliary_loss_mlp": 0.01049192,
      "balance_loss_clip": 1.05300844,
      "balance_loss_mlp": 1.02945089,
      "epoch": 0.12679993987674734,
      "flos": 17274864637440.0,
      "grad_norm": 1.9298328790617403,
      "language_loss": 0.7561394,
      "learning_rate": 3.902509795742467e-06,
      "loss": 0.77831209,
      "num_input_tokens_seen": 45602265,
      "step": 2109,
      "time_per_iteration": 2.5963573455810547
    },
    {
      "auxiliary_loss_clip": 0.01111025,
      "auxiliary_loss_mlp": 0.01053822,
      "balance_loss_clip": 1.04636049,
      "balance_loss_mlp": 1.0335331,
      "epoch": 0.1268600631294153,
      "flos": 17275080119040.0,
      "grad_norm": 1.6171901700648081,
      "language_loss": 0.82806516,
      "learning_rate": 3.902389647441592e-06,
      "loss": 0.84971368,
      "num_input_tokens_seen": 45620595,
      "step": 2110,
      "time_per_iteration": 2.6745550632476807
    },
    {
      "auxiliary_loss_clip": 0.01145969,
      "auxiliary_loss_mlp": 0.00778071,
      "balance_loss_clip": 1.05419564,
      "balance_loss_mlp": 0.99996144,
      "epoch": 0.12692018638208327,
      "flos": 24061047390720.0,
      "grad_norm": 1.6765217216011241,
      "language_loss": 0.78092968,
      "learning_rate": 3.90226942700191e-06,
      "loss": 0.80017006,
      "num_input_tokens_seen": 45641140,
      "step": 2111,
      "time_per_iteration": 2.65983510017395
    },
    {
      "auxiliary_loss_clip": 0.01130932,
      "auxiliary_loss_mlp": 0.01076547,
      "balance_loss_clip": 1.05490458,
      "balance_loss_mlp": 1.05352807,
      "epoch": 0.12698030963475124,
      "flos": 31831900652160.0,
      "grad_norm": 2.15738266202174,
      "language_loss": 0.77103376,
      "learning_rate": 3.902149134427982e-06,
      "loss": 0.79310858,
      "num_input_tokens_seen": 45662315,
      "step": 2112,
      "time_per_iteration": 2.870299816131592
    },
    {
      "auxiliary_loss_clip": 0.01129438,
      "auxiliary_loss_mlp": 0.01074863,
      "balance_loss_clip": 1.05213726,
      "balance_loss_mlp": 1.05427516,
      "epoch": 0.1270404328874192,
      "flos": 25187744275200.0,
      "grad_norm": 1.9191529425470424,
      "language_loss": 0.85806453,
      "learning_rate": 3.902028769724367e-06,
      "loss": 0.88010758,
      "num_input_tokens_seen": 45680335,
      "step": 2113,
      "time_per_iteration": 4.26338267326355
    },
    {
      "auxiliary_loss_clip": 0.01137468,
      "auxiliary_loss_mlp": 0.01078067,
      "balance_loss_clip": 1.05511892,
      "balance_loss_mlp": 1.05670488,
      "epoch": 0.12710055614008717,
      "flos": 15997342544640.0,
      "grad_norm": 1.9721234476704599,
      "language_loss": 0.74027002,
      "learning_rate": 3.9019083328956315e-06,
      "loss": 0.7624253,
      "num_input_tokens_seen": 45696240,
      "step": 2114,
      "time_per_iteration": 2.7573230266571045
    },
    {
      "auxiliary_loss_clip": 0.01156713,
      "auxiliary_loss_mlp": 0.01060574,
      "balance_loss_clip": 1.05770111,
      "balance_loss_mlp": 1.03924704,
      "epoch": 0.12716067939275516,
      "flos": 15085642515840.0,
      "grad_norm": 1.7921743813213327,
      "language_loss": 0.83240676,
      "learning_rate": 3.901787823946341e-06,
      "loss": 0.85457963,
      "num_input_tokens_seen": 45713695,
      "step": 2115,
      "time_per_iteration": 4.1369829177856445
    },
    {
      "auxiliary_loss_clip": 0.01154653,
      "auxiliary_loss_mlp": 0.01065557,
      "balance_loss_clip": 1.05875492,
      "balance_loss_mlp": 1.04476702,
      "epoch": 0.12722080264542313,
      "flos": 28366736636160.0,
      "grad_norm": 1.4840591347809418,
      "language_loss": 0.87010503,
      "learning_rate": 3.901667242881065e-06,
      "loss": 0.89230716,
      "num_input_tokens_seen": 45736655,
      "step": 2116,
      "time_per_iteration": 2.73896861076355
    },
    {
      "auxiliary_loss_clip": 0.01139498,
      "auxiliary_loss_mlp": 0.00777066,
      "balance_loss_clip": 1.05413389,
      "balance_loss_mlp": 0.99995339,
      "epoch": 0.1272809258980911,
      "flos": 32379897519360.0,
      "grad_norm": 1.753205985010591,
      "language_loss": 0.70374918,
      "learning_rate": 3.9015465897043775e-06,
      "loss": 0.72291481,
      "num_input_tokens_seen": 45758195,
      "step": 2117,
      "time_per_iteration": 2.783156156539917
    },
    {
      "auxiliary_loss_clip": 0.01127455,
      "auxiliary_loss_mlp": 0.0106424,
      "balance_loss_clip": 1.04978406,
      "balance_loss_mlp": 1.04068434,
      "epoch": 0.12734104915075906,
      "flos": 16034402401920.0,
      "grad_norm": 1.9957647698478755,
      "language_loss": 0.86237884,
      "learning_rate": 3.901425864420852e-06,
      "loss": 0.8842957,
      "num_input_tokens_seen": 45774280,
      "step": 2118,
      "time_per_iteration": 4.322036266326904
    },
    {
      "auxiliary_loss_clip": 0.01161417,
      "auxiliary_loss_mlp": 0.01049008,
      "balance_loss_clip": 1.05827069,
      "balance_loss_mlp": 1.02951694,
      "epoch": 0.12740117240342702,
      "flos": 18260325244800.0,
      "grad_norm": 1.705293179953873,
      "language_loss": 0.87577266,
      "learning_rate": 3.901305067035068e-06,
      "loss": 0.89787692,
      "num_input_tokens_seen": 45792760,
      "step": 2119,
      "time_per_iteration": 2.6559741497039795
    },
    {
      "auxiliary_loss_clip": 0.01145426,
      "auxiliary_loss_mlp": 0.0077754,
      "balance_loss_clip": 1.05233431,
      "balance_loss_mlp": 0.99984539,
      "epoch": 0.127461295656095,
      "flos": 12121790664960.0,
      "grad_norm": 2.05013605026053,
      "language_loss": 0.87824571,
      "learning_rate": 3.901184197551605e-06,
      "loss": 0.89747536,
      "num_input_tokens_seen": 45804300,
      "step": 2120,
      "time_per_iteration": 2.6154048442840576
    },
    {
      "auxiliary_loss_clip": 0.01170497,
      "auxiliary_loss_mlp": 0.01046075,
      "balance_loss_clip": 1.05822706,
      "balance_loss_mlp": 1.02626204,
      "epoch": 0.12752141890876295,
      "flos": 23149095966720.0,
      "grad_norm": 1.9784951602308867,
      "language_loss": 0.75584805,
      "learning_rate": 3.901063255975046e-06,
      "loss": 0.77801377,
      "num_input_tokens_seen": 45823780,
      "step": 2121,
      "time_per_iteration": 2.579265832901001
    },
    {
      "auxiliary_loss_clip": 0.0111249,
      "auxiliary_loss_mlp": 0.01047949,
      "balance_loss_clip": 1.04741263,
      "balance_loss_mlp": 1.02727842,
      "epoch": 0.12758154216143094,
      "flos": 21615997628160.0,
      "grad_norm": 2.0293629108662405,
      "language_loss": 0.82732606,
      "learning_rate": 3.900942242309978e-06,
      "loss": 0.84893048,
      "num_input_tokens_seen": 45840495,
      "step": 2122,
      "time_per_iteration": 2.793870210647583
    },
    {
      "auxiliary_loss_clip": 0.01151713,
      "auxiliary_loss_mlp": 0.01049724,
      "balance_loss_clip": 1.05901408,
      "balance_loss_mlp": 1.02983987,
      "epoch": 0.1276416654140989,
      "flos": 15924874855680.0,
      "grad_norm": 1.7660235451894624,
      "language_loss": 0.78699338,
      "learning_rate": 3.90082115656099e-06,
      "loss": 0.80900776,
      "num_input_tokens_seen": 45857735,
      "step": 2123,
      "time_per_iteration": 2.70546293258667
    },
    {
      "auxiliary_loss_clip": 0.01172823,
      "auxiliary_loss_mlp": 0.01055328,
      "balance_loss_clip": 1.05931985,
      "balance_loss_mlp": 1.03478789,
      "epoch": 0.12770178866676687,
      "flos": 22382690451840.0,
      "grad_norm": 1.5643885422181942,
      "language_loss": 0.78931451,
      "learning_rate": 3.900699998732673e-06,
      "loss": 0.81159604,
      "num_input_tokens_seen": 45876485,
      "step": 2124,
      "time_per_iteration": 2.661712408065796
    },
    {
      "auxiliary_loss_clip": 0.01160474,
      "auxiliary_loss_mlp": 0.00776885,
      "balance_loss_clip": 1.05457389,
      "balance_loss_mlp": 0.99987447,
      "epoch": 0.12776191191943484,
      "flos": 21652482867840.0,
      "grad_norm": 1.9695028631977674,
      "language_loss": 0.75605726,
      "learning_rate": 3.900578768829623e-06,
      "loss": 0.7754308,
      "num_input_tokens_seen": 45894645,
      "step": 2125,
      "time_per_iteration": 2.696021556854248
    },
    {
      "auxiliary_loss_clip": 0.01158163,
      "auxiliary_loss_mlp": 0.00777059,
      "balance_loss_clip": 1.05398965,
      "balance_loss_mlp": 1.00002348,
      "epoch": 0.1278220351721028,
      "flos": 25735561574400.0,
      "grad_norm": 3.019802885219414,
      "language_loss": 0.78016824,
      "learning_rate": 3.900457466856434e-06,
      "loss": 0.79952049,
      "num_input_tokens_seen": 45913755,
      "step": 2126,
      "time_per_iteration": 2.721435308456421
    },
    {
      "auxiliary_loss_clip": 0.01124637,
      "auxiliary_loss_mlp": 0.010537,
      "balance_loss_clip": 1.05406642,
      "balance_loss_mlp": 1.03504348,
      "epoch": 0.12788215842477077,
      "flos": 41243224982400.0,
      "grad_norm": 1.3825945270792501,
      "language_loss": 0.6927852,
      "learning_rate": 3.9003360928177085e-06,
      "loss": 0.71456861,
      "num_input_tokens_seen": 45936095,
      "step": 2127,
      "time_per_iteration": 2.902101993560791
    },
    {
      "auxiliary_loss_clip": 0.01030231,
      "auxiliary_loss_mlp": 0.00759051,
      "balance_loss_clip": 1.02830005,
      "balance_loss_mlp": 1.00050259,
      "epoch": 0.12794228167743876,
      "flos": 70877430881280.0,
      "grad_norm": 0.853491438999862,
      "language_loss": 0.62831402,
      "learning_rate": 3.900214646718047e-06,
      "loss": 0.64620686,
      "num_input_tokens_seen": 46004655,
      "step": 2128,
      "time_per_iteration": 3.3387396335601807
    },
    {
      "auxiliary_loss_clip": 0.01145823,
      "auxiliary_loss_mlp": 0.01047815,
      "balance_loss_clip": 1.05080712,
      "balance_loss_mlp": 1.02599955,
      "epoch": 0.12800240493010673,
      "flos": 16289727252480.0,
      "grad_norm": 2.066959353069841,
      "language_loss": 0.77626479,
      "learning_rate": 3.900093128562056e-06,
      "loss": 0.7982012,
      "num_input_tokens_seen": 46023610,
      "step": 2129,
      "time_per_iteration": 2.611309766769409
    },
    {
      "auxiliary_loss_clip": 0.01122914,
      "auxiliary_loss_mlp": 0.01052577,
      "balance_loss_clip": 1.05058527,
      "balance_loss_mlp": 1.03029668,
      "epoch": 0.1280625281827747,
      "flos": 20631542601600.0,
      "grad_norm": 2.1214737401843893,
      "language_loss": 0.79263359,
      "learning_rate": 3.899971538354343e-06,
      "loss": 0.81438851,
      "num_input_tokens_seen": 46041725,
      "step": 2130,
      "time_per_iteration": 2.753243923187256
    },
    {
      "auxiliary_loss_clip": 0.01139626,
      "auxiliary_loss_mlp": 0.01052453,
      "balance_loss_clip": 1.05133748,
      "balance_loss_mlp": 1.03147244,
      "epoch": 0.12812265143544266,
      "flos": 22638230784000.0,
      "grad_norm": 1.7780274650921335,
      "language_loss": 0.70945668,
      "learning_rate": 3.899849876099518e-06,
      "loss": 0.73137754,
      "num_input_tokens_seen": 46061095,
      "step": 2131,
      "time_per_iteration": 2.6809306144714355
    },
    {
      "auxiliary_loss_clip": 0.01102824,
      "auxiliary_loss_mlp": 0.01052393,
      "balance_loss_clip": 1.04982638,
      "balance_loss_mlp": 1.03163886,
      "epoch": 0.12818277468811062,
      "flos": 34714701463680.0,
      "grad_norm": 2.2916674504462655,
      "language_loss": 0.72298968,
      "learning_rate": 3.899728141802197e-06,
      "loss": 0.74454176,
      "num_input_tokens_seen": 46082670,
      "step": 2132,
      "time_per_iteration": 2.8769233226776123
    },
    {
      "auxiliary_loss_clip": 0.01102594,
      "auxiliary_loss_mlp": 0.01055993,
      "balance_loss_clip": 1.04384947,
      "balance_loss_mlp": 1.03348672,
      "epoch": 0.1282428979407786,
      "flos": 23112107936640.0,
      "grad_norm": 2.0316054281953155,
      "language_loss": 0.82128644,
      "learning_rate": 3.8996063354669935e-06,
      "loss": 0.84287226,
      "num_input_tokens_seen": 46102410,
      "step": 2133,
      "time_per_iteration": 2.766897678375244
    },
    {
      "auxiliary_loss_clip": 0.01163396,
      "auxiliary_loss_mlp": 0.01057069,
      "balance_loss_clip": 1.05397773,
      "balance_loss_mlp": 1.03458595,
      "epoch": 0.12830302119344655,
      "flos": 20886508316160.0,
      "grad_norm": 3.232115826630309,
      "language_loss": 0.80001891,
      "learning_rate": 3.899484457098528e-06,
      "loss": 0.82222354,
      "num_input_tokens_seen": 46121145,
      "step": 2134,
      "time_per_iteration": 2.6347672939300537
    },
    {
      "auxiliary_loss_clip": 0.01159056,
      "auxiliary_loss_mlp": 0.01046209,
      "balance_loss_clip": 1.05907345,
      "balance_loss_mlp": 1.02614641,
      "epoch": 0.12836314444611455,
      "flos": 21397768548480.0,
      "grad_norm": 1.731952504909339,
      "language_loss": 0.82657921,
      "learning_rate": 3.899362506701421e-06,
      "loss": 0.84863198,
      "num_input_tokens_seen": 46140740,
      "step": 2135,
      "time_per_iteration": 2.6393656730651855
    },
    {
      "auxiliary_loss_clip": 0.0114208,
      "auxiliary_loss_mlp": 0.0105553,
      "balance_loss_clip": 1.05345035,
      "balance_loss_mlp": 1.03411996,
      "epoch": 0.1284232676987825,
      "flos": 13662466773120.0,
      "grad_norm": 2.1083924470752278,
      "language_loss": 0.7764526,
      "learning_rate": 3.899240484280298e-06,
      "loss": 0.79842871,
      "num_input_tokens_seen": 46156805,
      "step": 2136,
      "time_per_iteration": 2.7195920944213867
    },
    {
      "auxiliary_loss_clip": 0.01020946,
      "auxiliary_loss_mlp": 0.01003991,
      "balance_loss_clip": 1.01967573,
      "balance_loss_mlp": 1.00096273,
      "epoch": 0.12848339095145048,
      "flos": 59994737735040.0,
      "grad_norm": 0.8964253308146478,
      "language_loss": 0.59152198,
      "learning_rate": 3.899118389839785e-06,
      "loss": 0.61177135,
      "num_input_tokens_seen": 46222085,
      "step": 2137,
      "time_per_iteration": 3.416015625
    },
    {
      "auxiliary_loss_clip": 0.01153694,
      "auxiliary_loss_mlp": 0.01054623,
      "balance_loss_clip": 1.05178177,
      "balance_loss_mlp": 1.03483438,
      "epoch": 0.12854351420411844,
      "flos": 13881378211200.0,
      "grad_norm": 3.244493357011547,
      "language_loss": 0.82344306,
      "learning_rate": 3.898996223384512e-06,
      "loss": 0.84552622,
      "num_input_tokens_seen": 46239970,
      "step": 2138,
      "time_per_iteration": 2.65515398979187
    },
    {
      "auxiliary_loss_clip": 0.01159586,
      "auxiliary_loss_mlp": 0.01049293,
      "balance_loss_clip": 1.05592752,
      "balance_loss_mlp": 1.02665496,
      "epoch": 0.1286036374567864,
      "flos": 22637943475200.0,
      "grad_norm": 2.5417837252920323,
      "language_loss": 0.78691363,
      "learning_rate": 3.898873984919113e-06,
      "loss": 0.8090024,
      "num_input_tokens_seen": 46257740,
      "step": 2139,
      "time_per_iteration": 2.651132345199585
    },
    {
      "auxiliary_loss_clip": 0.01136892,
      "auxiliary_loss_mlp": 0.01045928,
      "balance_loss_clip": 1.05267286,
      "balance_loss_mlp": 1.02582908,
      "epoch": 0.12866376070945437,
      "flos": 16324775948160.0,
      "grad_norm": 1.9541049485452633,
      "language_loss": 0.85289955,
      "learning_rate": 3.8987516744482215e-06,
      "loss": 0.87472773,
      "num_input_tokens_seen": 46275445,
      "step": 2140,
      "time_per_iteration": 2.730156183242798
    },
    {
      "auxiliary_loss_clip": 0.01143134,
      "auxiliary_loss_mlp": 0.01044337,
      "balance_loss_clip": 1.05203128,
      "balance_loss_mlp": 1.02482224,
      "epoch": 0.12872388396212234,
      "flos": 11874546374400.0,
      "grad_norm": 1.8185491602156885,
      "language_loss": 0.86268306,
      "learning_rate": 3.898629291976476e-06,
      "loss": 0.88455778,
      "num_input_tokens_seen": 46291710,
      "step": 2141,
      "time_per_iteration": 2.62223482131958
    },
    {
      "auxiliary_loss_clip": 0.01146971,
      "auxiliary_loss_mlp": 0.01045813,
      "balance_loss_clip": 1.0528295,
      "balance_loss_mlp": 1.02548814,
      "epoch": 0.12878400721479033,
      "flos": 28366700722560.0,
      "grad_norm": 3.1267362471736684,
      "language_loss": 0.68282312,
      "learning_rate": 3.898506837508518e-06,
      "loss": 0.70475101,
      "num_input_tokens_seen": 46311335,
      "step": 2142,
      "time_per_iteration": 2.71232271194458
    },
    {
      "auxiliary_loss_clip": 0.01165678,
      "auxiliary_loss_mlp": 0.0077895,
      "balance_loss_clip": 1.05764627,
      "balance_loss_mlp": 0.99990749,
      "epoch": 0.1288441304674583,
      "flos": 25885632597120.0,
      "grad_norm": 2.373838274123079,
      "language_loss": 0.83479214,
      "learning_rate": 3.89838431104899e-06,
      "loss": 0.85423845,
      "num_input_tokens_seen": 46330985,
      "step": 2143,
      "time_per_iteration": 2.677692174911499
    },
    {
      "auxiliary_loss_clip": 0.01175134,
      "auxiliary_loss_mlp": 0.00777405,
      "balance_loss_clip": 1.0598439,
      "balance_loss_mlp": 0.99994075,
      "epoch": 0.12890425372012626,
      "flos": 20813789232000.0,
      "grad_norm": 1.5662270309624111,
      "language_loss": 0.81703234,
      "learning_rate": 3.898261712602539e-06,
      "loss": 0.83655775,
      "num_input_tokens_seen": 46351295,
      "step": 2144,
      "time_per_iteration": 2.712620496749878
    },
    {
      "auxiliary_loss_clip": 0.01130321,
      "auxiliary_loss_mlp": 0.01053521,
      "balance_loss_clip": 1.04658103,
      "balance_loss_mlp": 1.03145528,
      "epoch": 0.12896437697279423,
      "flos": 22565870835840.0,
      "grad_norm": 1.8026346290528672,
      "language_loss": 0.78304374,
      "learning_rate": 3.898139042173813e-06,
      "loss": 0.80488217,
      "num_input_tokens_seen": 46368600,
      "step": 2145,
      "time_per_iteration": 2.6766605377197266
    },
    {
      "auxiliary_loss_clip": 0.01170585,
      "auxiliary_loss_mlp": 0.01047893,
      "balance_loss_clip": 1.0543592,
      "balance_loss_mlp": 1.02662635,
      "epoch": 0.1290245002254622,
      "flos": 17493776075520.0,
      "grad_norm": 2.147087506474235,
      "language_loss": 0.82865375,
      "learning_rate": 3.898016299767465e-06,
      "loss": 0.85083848,
      "num_input_tokens_seen": 46387370,
      "step": 2146,
      "time_per_iteration": 2.5860395431518555
    },
    {
      "auxiliary_loss_clip": 0.01141916,
      "auxiliary_loss_mlp": 0.0105138,
      "balance_loss_clip": 1.05367482,
      "balance_loss_mlp": 1.03062606,
      "epoch": 0.12908462347813016,
      "flos": 36315957859200.0,
      "grad_norm": 2.344626501147968,
      "language_loss": 0.71275079,
      "learning_rate": 3.897893485388149e-06,
      "loss": 0.73468375,
      "num_input_tokens_seen": 46409570,
      "step": 2147,
      "time_per_iteration": 2.7870359420776367
    },
    {
      "auxiliary_loss_clip": 0.01147238,
      "auxiliary_loss_mlp": 0.01052291,
      "balance_loss_clip": 1.05527067,
      "balance_loss_mlp": 1.03297925,
      "epoch": 0.12914474673079815,
      "flos": 22528703237760.0,
      "grad_norm": 2.120275205230366,
      "language_loss": 0.71432978,
      "learning_rate": 3.897770599040521e-06,
      "loss": 0.73632509,
      "num_input_tokens_seen": 46429320,
      "step": 2148,
      "time_per_iteration": 2.6865081787109375
    },
    {
      "auxiliary_loss_clip": 0.01168479,
      "auxiliary_loss_mlp": 0.01049575,
      "balance_loss_clip": 1.05762172,
      "balance_loss_mlp": 1.03016782,
      "epoch": 0.12920486998346611,
      "flos": 21471888263040.0,
      "grad_norm": 1.6388902851592406,
      "language_loss": 0.79064089,
      "learning_rate": 3.897647640729242e-06,
      "loss": 0.81282145,
      "num_input_tokens_seen": 46450155,
      "step": 2149,
      "time_per_iteration": 2.6041862964630127
    },
    {
      "auxiliary_loss_clip": 0.01159527,
      "auxiliary_loss_mlp": 0.01046069,
      "balance_loss_clip": 1.05377793,
      "balance_loss_mlp": 1.02531469,
      "epoch": 0.12926499323613408,
      "flos": 27308556944640.0,
      "grad_norm": 2.034796374339078,
      "language_loss": 0.75976646,
      "learning_rate": 3.897524610458975e-06,
      "loss": 0.78182244,
      "num_input_tokens_seen": 46470280,
      "step": 2150,
      "time_per_iteration": 2.647224187850952
    },
    {
      "auxiliary_loss_clip": 0.01155787,
      "auxiliary_loss_mlp": 0.01055192,
      "balance_loss_clip": 1.05445433,
      "balance_loss_mlp": 1.03491461,
      "epoch": 0.12932511648880204,
      "flos": 22091131756800.0,
      "grad_norm": 2.3830500835005592,
      "language_loss": 0.70986372,
      "learning_rate": 3.8974015082343835e-06,
      "loss": 0.73197353,
      "num_input_tokens_seen": 46487605,
      "step": 2151,
      "time_per_iteration": 2.7008492946624756
    },
    {
      "auxiliary_loss_clip": 0.01167835,
      "auxiliary_loss_mlp": 0.0104951,
      "balance_loss_clip": 1.05603719,
      "balance_loss_mlp": 1.03017378,
      "epoch": 0.12938523974147,
      "flos": 20302780394880.0,
      "grad_norm": 2.058334480733051,
      "language_loss": 0.83964819,
      "learning_rate": 3.897278334060137e-06,
      "loss": 0.86182165,
      "num_input_tokens_seen": 46505100,
      "step": 2152,
      "time_per_iteration": 2.6467373371124268
    },
    {
      "auxiliary_loss_clip": 0.01158553,
      "auxiliary_loss_mlp": 0.01058416,
      "balance_loss_clip": 1.05283821,
      "balance_loss_mlp": 1.03888893,
      "epoch": 0.12944536299413797,
      "flos": 19499961467520.0,
      "grad_norm": 1.5624811365269535,
      "language_loss": 0.78585124,
      "learning_rate": 3.897155087940906e-06,
      "loss": 0.80802095,
      "num_input_tokens_seen": 46524020,
      "step": 2153,
      "time_per_iteration": 4.286921262741089
    },
    {
      "auxiliary_loss_clip": 0.01113716,
      "auxiliary_loss_mlp": 0.00777812,
      "balance_loss_clip": 1.04707122,
      "balance_loss_mlp": 0.99989671,
      "epoch": 0.12950548624680594,
      "flos": 27707919333120.0,
      "grad_norm": 1.6189787343362376,
      "language_loss": 0.80253434,
      "learning_rate": 3.897031769881364e-06,
      "loss": 0.82144964,
      "num_input_tokens_seen": 46544640,
      "step": 2154,
      "time_per_iteration": 2.7602338790893555
    },
    {
      "auxiliary_loss_clip": 0.01149958,
      "auxiliary_loss_mlp": 0.0105188,
      "balance_loss_clip": 1.05262971,
      "balance_loss_mlp": 1.03099442,
      "epoch": 0.12956560949947393,
      "flos": 17565740974080.0,
      "grad_norm": 1.8080432584650143,
      "language_loss": 0.83717728,
      "learning_rate": 3.896908379886188e-06,
      "loss": 0.85919571,
      "num_input_tokens_seen": 46561395,
      "step": 2155,
      "time_per_iteration": 5.696707010269165
    },
    {
      "auxiliary_loss_clip": 0.01161999,
      "auxiliary_loss_mlp": 0.01056273,
      "balance_loss_clip": 1.05426383,
      "balance_loss_mlp": 1.03611445,
      "epoch": 0.1296257327521419,
      "flos": 20740711011840.0,
      "grad_norm": 2.4972858828122666,
      "language_loss": 0.76114857,
      "learning_rate": 3.896784917960055e-06,
      "loss": 0.78333133,
      "num_input_tokens_seen": 46579395,
      "step": 2156,
      "time_per_iteration": 2.6279313564300537
    },
    {
      "auxiliary_loss_clip": 0.01105089,
      "auxiliary_loss_mlp": 0.01056603,
      "balance_loss_clip": 1.0510118,
      "balance_loss_mlp": 1.03679013,
      "epoch": 0.12968585600480986,
      "flos": 16395735265920.0,
      "grad_norm": 1.6652476704410177,
      "language_loss": 0.86493659,
      "learning_rate": 3.896661384107648e-06,
      "loss": 0.88655347,
      "num_input_tokens_seen": 46597090,
      "step": 2157,
      "time_per_iteration": 4.4089202880859375
    },
    {
      "auxiliary_loss_clip": 0.01170107,
      "auxiliary_loss_mlp": 0.01055814,
      "balance_loss_clip": 1.05253935,
      "balance_loss_mlp": 1.0349642,
      "epoch": 0.12974597925747783,
      "flos": 28329533124480.0,
      "grad_norm": 2.5240136552338956,
      "language_loss": 0.80393612,
      "learning_rate": 3.896537778333651e-06,
      "loss": 0.8261953,
      "num_input_tokens_seen": 46617355,
      "step": 2158,
      "time_per_iteration": 2.702765703201294
    },
    {
      "auxiliary_loss_clip": 0.01177017,
      "auxiliary_loss_mlp": 0.01060365,
      "balance_loss_clip": 1.05905974,
      "balance_loss_mlp": 1.04050517,
      "epoch": 0.1298061025101458,
      "flos": 9683025782400.0,
      "grad_norm": 2.5307604694159607,
      "language_loss": 0.74881256,
      "learning_rate": 3.896414100642752e-06,
      "loss": 0.77118635,
      "num_input_tokens_seen": 46633130,
      "step": 2159,
      "time_per_iteration": 2.534163475036621
    },
    {
      "auxiliary_loss_clip": 0.01122909,
      "auxiliary_loss_mlp": 0.01058309,
      "balance_loss_clip": 1.04594469,
      "balance_loss_mlp": 1.03471708,
      "epoch": 0.12986622576281376,
      "flos": 27709535445120.0,
      "grad_norm": 1.954419432637739,
      "language_loss": 0.8259204,
      "learning_rate": 3.89629035103964e-06,
      "loss": 0.84773254,
      "num_input_tokens_seen": 46650575,
      "step": 2160,
      "time_per_iteration": 2.7358646392822266
    },
    {
      "auxiliary_loss_clip": 0.01154348,
      "auxiliary_loss_mlp": 0.01047243,
      "balance_loss_clip": 1.05873609,
      "balance_loss_mlp": 1.02732301,
      "epoch": 0.12992634901548175,
      "flos": 18802719590400.0,
      "grad_norm": 1.7252123805741888,
      "language_loss": 0.82310414,
      "learning_rate": 3.896166529529008e-06,
      "loss": 0.84512007,
      "num_input_tokens_seen": 46668780,
      "step": 2161,
      "time_per_iteration": 2.7029623985290527
    },
    {
      "auxiliary_loss_clip": 0.01145886,
      "auxiliary_loss_mlp": 0.01060381,
      "balance_loss_clip": 1.05145073,
      "balance_loss_mlp": 1.03911448,
      "epoch": 0.12998647226814972,
      "flos": 29127575543040.0,
      "grad_norm": 2.0780374068601253,
      "language_loss": 0.82668459,
      "learning_rate": 3.896042636115551e-06,
      "loss": 0.84874725,
      "num_input_tokens_seen": 46687550,
      "step": 2162,
      "time_per_iteration": 2.674825668334961
    },
    {
      "auxiliary_loss_clip": 0.0113921,
      "auxiliary_loss_mlp": 0.0105953,
      "balance_loss_clip": 1.05468941,
      "balance_loss_mlp": 1.03957474,
      "epoch": 0.13004659552081768,
      "flos": 19573686132480.0,
      "grad_norm": 3.928222506771022,
      "language_loss": 0.72579277,
      "learning_rate": 3.895918670803968e-06,
      "loss": 0.7477802,
      "num_input_tokens_seen": 46706730,
      "step": 2163,
      "time_per_iteration": 2.678394079208374
    },
    {
      "auxiliary_loss_clip": 0.01173873,
      "auxiliary_loss_mlp": 0.00778662,
      "balance_loss_clip": 1.05635965,
      "balance_loss_mlp": 0.99994016,
      "epoch": 0.13010671877348565,
      "flos": 22490709626880.0,
      "grad_norm": 2.0196348424542827,
      "language_loss": 0.81330699,
      "learning_rate": 3.895794633598958e-06,
      "loss": 0.83283234,
      "num_input_tokens_seen": 46724250,
      "step": 2164,
      "time_per_iteration": 2.6116931438446045
    },
    {
      "auxiliary_loss_clip": 0.01119834,
      "auxiliary_loss_mlp": 0.01050661,
      "balance_loss_clip": 1.04808033,
      "balance_loss_mlp": 1.03061032,
      "epoch": 0.1301668420261536,
      "flos": 23878226142720.0,
      "grad_norm": 2.274563635903502,
      "language_loss": 0.72262049,
      "learning_rate": 3.8956705245052256e-06,
      "loss": 0.74432552,
      "num_input_tokens_seen": 46744105,
      "step": 2165,
      "time_per_iteration": 2.7646515369415283
    },
    {
      "auxiliary_loss_clip": 0.01109832,
      "auxiliary_loss_mlp": 0.01048351,
      "balance_loss_clip": 1.05059505,
      "balance_loss_mlp": 1.02707219,
      "epoch": 0.13022696527882158,
      "flos": 23150065633920.0,
      "grad_norm": 2.8383873988269217,
      "language_loss": 0.74749964,
      "learning_rate": 3.8955463435274765e-06,
      "loss": 0.76908153,
      "num_input_tokens_seen": 46764250,
      "step": 2166,
      "time_per_iteration": 2.7939398288726807
    },
    {
      "auxiliary_loss_clip": 0.01170298,
      "auxiliary_loss_mlp": 0.01048037,
      "balance_loss_clip": 1.05364752,
      "balance_loss_mlp": 1.02827251,
      "epoch": 0.13028708853148954,
      "flos": 26908548111360.0,
      "grad_norm": 1.5379857106114436,
      "language_loss": 0.83098066,
      "learning_rate": 3.895422090670421e-06,
      "loss": 0.85316396,
      "num_input_tokens_seen": 46786865,
      "step": 2167,
      "time_per_iteration": 2.700505495071411
    },
    {
      "auxiliary_loss_clip": 0.01108628,
      "auxiliary_loss_mlp": 0.01059921,
      "balance_loss_clip": 1.04567361,
      "balance_loss_mlp": 1.03841531,
      "epoch": 0.13034721178415754,
      "flos": 21251468453760.0,
      "grad_norm": 1.6054044551173634,
      "language_loss": 0.83578718,
      "learning_rate": 3.89529776593877e-06,
      "loss": 0.85747266,
      "num_input_tokens_seen": 46807030,
      "step": 2168,
      "time_per_iteration": 2.839285135269165
    },
    {
      "auxiliary_loss_clip": 0.01079188,
      "auxiliary_loss_mlp": 0.01063413,
      "balance_loss_clip": 1.04247975,
      "balance_loss_mlp": 1.03861713,
      "epoch": 0.1304073350368255,
      "flos": 18767239931520.0,
      "grad_norm": 1.950315007602454,
      "language_loss": 0.79910588,
      "learning_rate": 3.8951733693372375e-06,
      "loss": 0.8205319,
      "num_input_tokens_seen": 46826280,
      "step": 2169,
      "time_per_iteration": 2.8150076866149902
    },
    {
      "auxiliary_loss_clip": 0.01174566,
      "auxiliary_loss_mlp": 0.01044893,
      "balance_loss_clip": 1.05822575,
      "balance_loss_mlp": 1.02339983,
      "epoch": 0.13046745828949347,
      "flos": 28364653647360.0,
      "grad_norm": 2.4117618540057766,
      "language_loss": 0.66804767,
      "learning_rate": 3.8950489008705406e-06,
      "loss": 0.69024229,
      "num_input_tokens_seen": 46846505,
      "step": 2170,
      "time_per_iteration": 2.722769021987915
    },
    {
      "auxiliary_loss_clip": 0.0114216,
      "auxiliary_loss_mlp": 0.01046684,
      "balance_loss_clip": 1.05424142,
      "balance_loss_mlp": 1.02637053,
      "epoch": 0.13052758154216143,
      "flos": 29605044055680.0,
      "grad_norm": 1.9089846415842238,
      "language_loss": 0.66768706,
      "learning_rate": 3.8949243605434e-06,
      "loss": 0.68957549,
      "num_input_tokens_seen": 46867380,
      "step": 2171,
      "time_per_iteration": 2.7474682331085205
    },
    {
      "auxiliary_loss_clip": 0.01157431,
      "auxiliary_loss_mlp": 0.01049079,
      "balance_loss_clip": 1.05283058,
      "balance_loss_mlp": 1.02701378,
      "epoch": 0.1305877047948294,
      "flos": 19390864884480.0,
      "grad_norm": 2.103440896006443,
      "language_loss": 0.72157478,
      "learning_rate": 3.894799748360537e-06,
      "loss": 0.74363995,
      "num_input_tokens_seen": 46886810,
      "step": 2172,
      "time_per_iteration": 2.8062691688537598
    },
    {
      "auxiliary_loss_clip": 0.01131178,
      "auxiliary_loss_mlp": 0.01045812,
      "balance_loss_clip": 1.05676126,
      "balance_loss_mlp": 1.0248909,
      "epoch": 0.13064782804749736,
      "flos": 16873527000960.0,
      "grad_norm": 1.8662964619330822,
      "language_loss": 0.75331408,
      "learning_rate": 3.894675064326678e-06,
      "loss": 0.77508402,
      "num_input_tokens_seen": 46905620,
      "step": 2173,
      "time_per_iteration": 2.749630928039551
    },
    {
      "auxiliary_loss_clip": 0.01132129,
      "auxiliary_loss_mlp": 0.01056024,
      "balance_loss_clip": 1.05241716,
      "balance_loss_mlp": 1.03388715,
      "epoch": 0.13070795130016533,
      "flos": 24499085748480.0,
      "grad_norm": 2.8034072456055426,
      "language_loss": 0.70175481,
      "learning_rate": 3.894550308446551e-06,
      "loss": 0.72363639,
      "num_input_tokens_seen": 46925120,
      "step": 2174,
      "time_per_iteration": 2.723314046859741
    },
    {
      "auxiliary_loss_clip": 0.01047643,
      "auxiliary_loss_mlp": 0.01015006,
      "balance_loss_clip": 1.02629197,
      "balance_loss_mlp": 1.01260972,
      "epoch": 0.13076807455283332,
      "flos": 71054505953280.0,
      "grad_norm": 0.7998489021914615,
      "language_loss": 0.59026134,
      "learning_rate": 3.894425480724886e-06,
      "loss": 0.61088777,
      "num_input_tokens_seen": 46988195,
      "step": 2175,
      "time_per_iteration": 3.318049192428589
    },
    {
      "auxiliary_loss_clip": 0.01159762,
      "auxiliary_loss_mlp": 0.01053929,
      "balance_loss_clip": 1.05441868,
      "balance_loss_mlp": 1.03342521,
      "epoch": 0.13082819780550128,
      "flos": 20264499475200.0,
      "grad_norm": 2.2309284705459707,
      "language_loss": 0.80365628,
      "learning_rate": 3.894300581166417e-06,
      "loss": 0.82579315,
      "num_input_tokens_seen": 47004720,
      "step": 2176,
      "time_per_iteration": 2.631732702255249
    },
    {
      "auxiliary_loss_clip": 0.01169648,
      "auxiliary_loss_mlp": 0.01047517,
      "balance_loss_clip": 1.05513525,
      "balance_loss_mlp": 1.02529645,
      "epoch": 0.13088832105816925,
      "flos": 34203441231360.0,
      "grad_norm": 1.6906214681317566,
      "language_loss": 0.74661696,
      "learning_rate": 3.894175609775881e-06,
      "loss": 0.76878858,
      "num_input_tokens_seen": 47024255,
      "step": 2177,
      "time_per_iteration": 2.701422691345215
    },
    {
      "auxiliary_loss_clip": 0.01131124,
      "auxiliary_loss_mlp": 0.0105144,
      "balance_loss_clip": 1.051373,
      "balance_loss_mlp": 1.02905297,
      "epoch": 0.13094844431083721,
      "flos": 17894970057600.0,
      "grad_norm": 1.8043513019060269,
      "language_loss": 0.82266748,
      "learning_rate": 3.894050566558015e-06,
      "loss": 0.84449303,
      "num_input_tokens_seen": 47042465,
      "step": 2178,
      "time_per_iteration": 2.6934497356414795
    },
    {
      "auxiliary_loss_clip": 0.01170524,
      "auxiliary_loss_mlp": 0.01047895,
      "balance_loss_clip": 1.05729508,
      "balance_loss_mlp": 1.02705729,
      "epoch": 0.13100856756350518,
      "flos": 17311313963520.0,
      "grad_norm": 2.9251611149508276,
      "language_loss": 0.74291968,
      "learning_rate": 3.893925451517562e-06,
      "loss": 0.76510382,
      "num_input_tokens_seen": 47060370,
      "step": 2179,
      "time_per_iteration": 2.6111502647399902
    },
    {
      "auxiliary_loss_clip": 0.01128297,
      "auxiliary_loss_mlp": 0.01052407,
      "balance_loss_clip": 1.04917574,
      "balance_loss_mlp": 1.03184354,
      "epoch": 0.13106869081617314,
      "flos": 22200551562240.0,
      "grad_norm": 1.9805514150688242,
      "language_loss": 0.84366202,
      "learning_rate": 3.893800264659266e-06,
      "loss": 0.8654691,
      "num_input_tokens_seen": 47081415,
      "step": 2180,
      "time_per_iteration": 2.731229543685913
    },
    {
      "auxiliary_loss_clip": 0.01162028,
      "auxiliary_loss_mlp": 0.0105845,
      "balance_loss_clip": 1.05875921,
      "balance_loss_mlp": 1.03757644,
      "epoch": 0.13112881406884114,
      "flos": 21763123735680.0,
      "grad_norm": 1.8389866248015785,
      "language_loss": 0.89840436,
      "learning_rate": 3.8936750059878746e-06,
      "loss": 0.92060918,
      "num_input_tokens_seen": 47099860,
      "step": 2181,
      "time_per_iteration": 2.643890380859375
    },
    {
      "auxiliary_loss_clip": 0.01153771,
      "auxiliary_loss_mlp": 0.01051982,
      "balance_loss_clip": 1.05222976,
      "balance_loss_mlp": 1.03126323,
      "epoch": 0.1311889373215091,
      "flos": 23331091201920.0,
      "grad_norm": 2.117586475019142,
      "language_loss": 0.68813586,
      "learning_rate": 3.893549675508137e-06,
      "loss": 0.7101934,
      "num_input_tokens_seen": 47118540,
      "step": 2182,
      "time_per_iteration": 2.6198863983154297
    },
    {
      "auxiliary_loss_clip": 0.01123039,
      "auxiliary_loss_mlp": 0.01051411,
      "balance_loss_clip": 1.0502702,
      "balance_loss_mlp": 1.0292381,
      "epoch": 0.13124906057417707,
      "flos": 21467363149440.0,
      "grad_norm": 1.787500136217105,
      "language_loss": 0.78694725,
      "learning_rate": 3.893424273224806e-06,
      "loss": 0.8086918,
      "num_input_tokens_seen": 47136710,
      "step": 2183,
      "time_per_iteration": 2.715517520904541
    },
    {
      "auxiliary_loss_clip": 0.01169106,
      "auxiliary_loss_mlp": 0.01047098,
      "balance_loss_clip": 1.05452895,
      "balance_loss_mlp": 1.02586675,
      "epoch": 0.13130918382684503,
      "flos": 23255319461760.0,
      "grad_norm": 26.753588494231124,
      "language_loss": 0.85792655,
      "learning_rate": 3.893298799142636e-06,
      "loss": 0.88008863,
      "num_input_tokens_seen": 47157155,
      "step": 2184,
      "time_per_iteration": 2.632539987564087
    },
    {
      "auxiliary_loss_clip": 0.01138714,
      "auxiliary_loss_mlp": 0.01054657,
      "balance_loss_clip": 1.05349112,
      "balance_loss_mlp": 1.03230524,
      "epoch": 0.131369307079513,
      "flos": 20850274471680.0,
      "grad_norm": 2.50466124454056,
      "language_loss": 0.82703435,
      "learning_rate": 3.893173253266387e-06,
      "loss": 0.84896809,
      "num_input_tokens_seen": 47176820,
      "step": 2185,
      "time_per_iteration": 2.6809136867523193
    },
    {
      "auxiliary_loss_clip": 0.01144077,
      "auxiliary_loss_mlp": 0.01054121,
      "balance_loss_clip": 1.05262399,
      "balance_loss_mlp": 1.03236496,
      "epoch": 0.13142943033218096,
      "flos": 17858341163520.0,
      "grad_norm": 1.8949462712827352,
      "language_loss": 0.72956109,
      "learning_rate": 3.893047635600818e-06,
      "loss": 0.75154305,
      "num_input_tokens_seen": 47195855,
      "step": 2186,
      "time_per_iteration": 2.628096342086792
    },
    {
      "auxiliary_loss_clip": 0.01157778,
      "auxiliary_loss_mlp": 0.01050695,
      "balance_loss_clip": 1.05436552,
      "balance_loss_mlp": 1.02783096,
      "epoch": 0.13148955358484893,
      "flos": 20996035862400.0,
      "grad_norm": 1.9822444068613732,
      "language_loss": 0.80363685,
      "learning_rate": 3.892921946150693e-06,
      "loss": 0.82572162,
      "num_input_tokens_seen": 47214535,
      "step": 2187,
      "time_per_iteration": 2.762223720550537
    },
    {
      "auxiliary_loss_clip": 0.01027324,
      "auxiliary_loss_mlp": 0.0101023,
      "balance_loss_clip": 1.02364707,
      "balance_loss_mlp": 1.00792885,
      "epoch": 0.13154967683751692,
      "flos": 70172467580160.0,
      "grad_norm": 0.8471850380496847,
      "language_loss": 0.59082437,
      "learning_rate": 3.892796184920778e-06,
      "loss": 0.61119986,
      "num_input_tokens_seen": 47270300,
      "step": 2188,
      "time_per_iteration": 3.302457571029663
    },
    {
      "auxiliary_loss_clip": 0.01095126,
      "auxiliary_loss_mlp": 0.01059346,
      "balance_loss_clip": 1.04827487,
      "balance_loss_mlp": 1.03676724,
      "epoch": 0.1316098000901849,
      "flos": 20376145923840.0,
      "grad_norm": 1.7340345041340466,
      "language_loss": 0.74211109,
      "learning_rate": 3.892670351915842e-06,
      "loss": 0.76365584,
      "num_input_tokens_seen": 47290720,
      "step": 2189,
      "time_per_iteration": 2.7990496158599854
    },
    {
      "auxiliary_loss_clip": 0.01160124,
      "auxiliary_loss_mlp": 0.01049098,
      "balance_loss_clip": 1.05551052,
      "balance_loss_mlp": 1.02799821,
      "epoch": 0.13166992334285285,
      "flos": 23221132692480.0,
      "grad_norm": 1.8160574809616576,
      "language_loss": 0.73152113,
      "learning_rate": 3.892544447140657e-06,
      "loss": 0.75361335,
      "num_input_tokens_seen": 47311820,
      "step": 2190,
      "time_per_iteration": 2.6485326290130615
    },
    {
      "auxiliary_loss_clip": 0.01160351,
      "auxiliary_loss_mlp": 0.01058461,
      "balance_loss_clip": 1.05671644,
      "balance_loss_mlp": 1.03811169,
      "epoch": 0.13173004659552082,
      "flos": 23330947547520.0,
      "grad_norm": 1.8825588242208007,
      "language_loss": 0.74617779,
      "learning_rate": 3.892418470599996e-06,
      "loss": 0.76836598,
      "num_input_tokens_seen": 47331605,
      "step": 2191,
      "time_per_iteration": 2.644484281539917
    },
    {
      "auxiliary_loss_clip": 0.0112783,
      "auxiliary_loss_mlp": 0.01054712,
      "balance_loss_clip": 1.05129039,
      "balance_loss_mlp": 1.03356445,
      "epoch": 0.13179016984818878,
      "flos": 21251504367360.0,
      "grad_norm": 1.8823393822145031,
      "language_loss": 0.79093283,
      "learning_rate": 3.892292422298637e-06,
      "loss": 0.81275827,
      "num_input_tokens_seen": 47350455,
      "step": 2192,
      "time_per_iteration": 2.735225200653076
    },
    {
      "auxiliary_loss_clip": 0.0111282,
      "auxiliary_loss_mlp": 0.01051113,
      "balance_loss_clip": 1.04457211,
      "balance_loss_mlp": 1.02936912,
      "epoch": 0.13185029310085675,
      "flos": 17778690754560.0,
      "grad_norm": 1.7242105632860862,
      "language_loss": 0.85350716,
      "learning_rate": 3.892166302241361e-06,
      "loss": 0.87514639,
      "num_input_tokens_seen": 47368225,
      "step": 2193,
      "time_per_iteration": 4.262877941131592
    },
    {
      "auxiliary_loss_clip": 0.0104173,
      "auxiliary_loss_mlp": 0.01015651,
      "balance_loss_clip": 1.02609122,
      "balance_loss_mlp": 1.01280212,
      "epoch": 0.1319104163535247,
      "flos": 69851785933440.0,
      "grad_norm": 0.7746813180799224,
      "language_loss": 0.54112649,
      "learning_rate": 3.8920401104329475e-06,
      "loss": 0.56170022,
      "num_input_tokens_seen": 47427125,
      "step": 2194,
      "time_per_iteration": 6.223008394241333
    },
    {
      "auxiliary_loss_clip": 0.01168022,
      "auxiliary_loss_mlp": 0.01048581,
      "balance_loss_clip": 1.05420566,
      "balance_loss_mlp": 1.02828002,
      "epoch": 0.1319705396061927,
      "flos": 25193095401600.0,
      "grad_norm": 2.1079865649821925,
      "language_loss": 0.72433972,
      "learning_rate": 3.891913846878185e-06,
      "loss": 0.74650574,
      "num_input_tokens_seen": 47450275,
      "step": 2195,
      "time_per_iteration": 2.6357345581054688
    },
    {
      "auxiliary_loss_clip": 0.01136503,
      "auxiliary_loss_mlp": 0.00778731,
      "balance_loss_clip": 1.05176425,
      "balance_loss_mlp": 0.99996454,
      "epoch": 0.13203066285886067,
      "flos": 20740459616640.0,
      "grad_norm": 1.5737174748369949,
      "language_loss": 0.78126895,
      "learning_rate": 3.891787511581859e-06,
      "loss": 0.8004213,
      "num_input_tokens_seen": 47469155,
      "step": 2196,
      "time_per_iteration": 2.7118594646453857
    },
    {
      "auxiliary_loss_clip": 0.01162447,
      "auxiliary_loss_mlp": 0.010526,
      "balance_loss_clip": 1.05453539,
      "balance_loss_mlp": 1.03210831,
      "epoch": 0.13209078611152864,
      "flos": 22054395121920.0,
      "grad_norm": 1.9385650447291836,
      "language_loss": 0.74632496,
      "learning_rate": 3.89166110454876e-06,
      "loss": 0.76847541,
      "num_input_tokens_seen": 47488405,
      "step": 2197,
      "time_per_iteration": 4.270530939102173
    },
    {
      "auxiliary_loss_clip": 0.01173786,
      "auxiliary_loss_mlp": 0.01050846,
      "balance_loss_clip": 1.05440533,
      "balance_loss_mlp": 1.02947164,
      "epoch": 0.1321509093641966,
      "flos": 16284950743680.0,
      "grad_norm": 1.785688190112577,
      "language_loss": 0.79566747,
      "learning_rate": 3.891534625783685e-06,
      "loss": 0.81791383,
      "num_input_tokens_seen": 47505650,
      "step": 2198,
      "time_per_iteration": 2.6145474910736084
    },
    {
      "auxiliary_loss_clip": 0.01170264,
      "auxiliary_loss_mlp": 0.01057159,
      "balance_loss_clip": 1.05536175,
      "balance_loss_mlp": 1.03647637,
      "epoch": 0.13221103261686457,
      "flos": 16983018633600.0,
      "grad_norm": 2.56313218775589,
      "language_loss": 0.82932216,
      "learning_rate": 3.891408075291425e-06,
      "loss": 0.85159647,
      "num_input_tokens_seen": 47521540,
      "step": 2199,
      "time_per_iteration": 2.5715503692626953
    },
    {
      "auxiliary_loss_clip": 0.01122554,
      "auxiliary_loss_mlp": 0.01052148,
      "balance_loss_clip": 1.05047798,
      "balance_loss_mlp": 1.03045249,
      "epoch": 0.13227115586953253,
      "flos": 34233605677440.0,
      "grad_norm": 1.8710902505917797,
      "language_loss": 0.69579422,
      "learning_rate": 3.8912814530767826e-06,
      "loss": 0.71754128,
      "num_input_tokens_seen": 47543625,
      "step": 2200,
      "time_per_iteration": 2.8001365661621094
    },
    {
      "auxiliary_loss_clip": 0.01167798,
      "auxiliary_loss_mlp": 0.01058155,
      "balance_loss_clip": 1.05345917,
      "balance_loss_mlp": 1.03618431,
      "epoch": 0.13233127912220052,
      "flos": 20704656735360.0,
      "grad_norm": 1.647659287704997,
      "language_loss": 0.84624702,
      "learning_rate": 3.891154759144557e-06,
      "loss": 0.86850655,
      "num_input_tokens_seen": 47563740,
      "step": 2201,
      "time_per_iteration": 2.6485981941223145
    },
    {
      "auxiliary_loss_clip": 0.0117188,
      "auxiliary_loss_mlp": 0.01055627,
      "balance_loss_clip": 1.05427861,
      "balance_loss_mlp": 1.03431273,
      "epoch": 0.1323914023748685,
      "flos": 25805048434560.0,
      "grad_norm": 1.7446392584198542,
      "language_loss": 0.87088037,
      "learning_rate": 3.891027993499554e-06,
      "loss": 0.8931554,
      "num_input_tokens_seen": 47582655,
      "step": 2202,
      "time_per_iteration": 2.5921456813812256
    },
    {
      "auxiliary_loss_clip": 0.01139991,
      "auxiliary_loss_mlp": 0.01053413,
      "balance_loss_clip": 1.05299544,
      "balance_loss_mlp": 1.03267026,
      "epoch": 0.13245152562753645,
      "flos": 21251540280960.0,
      "grad_norm": 2.405254380671628,
      "language_loss": 0.72801507,
      "learning_rate": 3.89090115614658e-06,
      "loss": 0.7499491,
      "num_input_tokens_seen": 47600875,
      "step": 2203,
      "time_per_iteration": 2.6257405281066895
    },
    {
      "auxiliary_loss_clip": 0.01124508,
      "auxiliary_loss_mlp": 0.0105959,
      "balance_loss_clip": 1.05080879,
      "balance_loss_mlp": 1.03916979,
      "epoch": 0.13251164888020442,
      "flos": 26610955931520.0,
      "grad_norm": 2.044348475010678,
      "language_loss": 0.73170948,
      "learning_rate": 3.890774247090444e-06,
      "loss": 0.75355047,
      "num_input_tokens_seen": 47619250,
      "step": 2204,
      "time_per_iteration": 2.753830909729004
    },
    {
      "auxiliary_loss_clip": 0.01160826,
      "auxiliary_loss_mlp": 0.01054406,
      "balance_loss_clip": 1.05474758,
      "balance_loss_mlp": 1.03225708,
      "epoch": 0.13257177213287238,
      "flos": 29826541272960.0,
      "grad_norm": 2.094172729236468,
      "language_loss": 0.78377104,
      "learning_rate": 3.89064726633596e-06,
      "loss": 0.80592328,
      "num_input_tokens_seen": 47639445,
      "step": 2205,
      "time_per_iteration": 2.730682134628296
    },
    {
      "auxiliary_loss_clip": 0.01125154,
      "auxiliary_loss_mlp": 0.01048818,
      "balance_loss_clip": 1.04975629,
      "balance_loss_mlp": 1.02782559,
      "epoch": 0.13263189538554035,
      "flos": 21288456483840.0,
      "grad_norm": 1.8609089802832188,
      "language_loss": 0.78638101,
      "learning_rate": 3.890520213887941e-06,
      "loss": 0.80812073,
      "num_input_tokens_seen": 47658740,
      "step": 2206,
      "time_per_iteration": 2.691962718963623
    },
    {
      "auxiliary_loss_clip": 0.01124965,
      "auxiliary_loss_mlp": 0.01045957,
      "balance_loss_clip": 1.04958403,
      "balance_loss_mlp": 1.02649069,
      "epoch": 0.13269201863820831,
      "flos": 16874101618560.0,
      "grad_norm": 2.2777192787220066,
      "language_loss": 0.74672282,
      "learning_rate": 3.890393089751208e-06,
      "loss": 0.76843208,
      "num_input_tokens_seen": 47676880,
      "step": 2207,
      "time_per_iteration": 2.7062454223632812
    },
    {
      "auxiliary_loss_clip": 0.01143208,
      "auxiliary_loss_mlp": 0.01047941,
      "balance_loss_clip": 1.05257845,
      "balance_loss_mlp": 1.02672219,
      "epoch": 0.1327521418908763,
      "flos": 23768914078080.0,
      "grad_norm": 1.692212064021935,
      "language_loss": 0.84061795,
      "learning_rate": 3.890265893930578e-06,
      "loss": 0.8625294,
      "num_input_tokens_seen": 47696635,
      "step": 2208,
      "time_per_iteration": 2.687717914581299
    },
    {
      "auxiliary_loss_clip": 0.01152573,
      "auxiliary_loss_mlp": 0.0105274,
      "balance_loss_clip": 1.05847478,
      "balance_loss_mlp": 1.03411973,
      "epoch": 0.13281226514354427,
      "flos": 26505594362880.0,
      "grad_norm": 1.7032258459750478,
      "language_loss": 0.85587811,
      "learning_rate": 3.890138626430876e-06,
      "loss": 0.8779313,
      "num_input_tokens_seen": 47717760,
      "step": 2209,
      "time_per_iteration": 2.646015167236328
    },
    {
      "auxiliary_loss_clip": 0.01138084,
      "auxiliary_loss_mlp": 0.00778828,
      "balance_loss_clip": 1.05316806,
      "balance_loss_mlp": 1.00002563,
      "epoch": 0.13287238839621224,
      "flos": 24498762526080.0,
      "grad_norm": 2.237247968175465,
      "language_loss": 0.81797457,
      "learning_rate": 3.890011287256929e-06,
      "loss": 0.83714366,
      "num_input_tokens_seen": 47737685,
      "step": 2210,
      "time_per_iteration": 2.676262378692627
    },
    {
      "auxiliary_loss_clip": 0.0104445,
      "auxiliary_loss_mlp": 0.00757817,
      "balance_loss_clip": 1.03801322,
      "balance_loss_mlp": 1.00007725,
      "epoch": 0.1329325116488802,
      "flos": 67694344369920.0,
      "grad_norm": 0.7515252652740232,
      "language_loss": 0.58031559,
      "learning_rate": 3.889883876413563e-06,
      "loss": 0.59833825,
      "num_input_tokens_seen": 47802415,
      "step": 2211,
      "time_per_iteration": 3.3914146423339844
    },
    {
      "auxiliary_loss_clip": 0.01064712,
      "auxiliary_loss_mlp": 0.01012978,
      "balance_loss_clip": 1.04205871,
      "balance_loss_mlp": 1.01083231,
      "epoch": 0.13299263490154817,
      "flos": 72261894741120.0,
      "grad_norm": 0.8012428422082742,
      "language_loss": 0.55299425,
      "learning_rate": 3.889756393905611e-06,
      "loss": 0.57377112,
      "num_input_tokens_seen": 47871485,
      "step": 2212,
      "time_per_iteration": 3.2910914421081543
    },
    {
      "auxiliary_loss_clip": 0.01132433,
      "auxiliary_loss_mlp": 0.01054299,
      "balance_loss_clip": 1.05107963,
      "balance_loss_mlp": 1.0331986,
      "epoch": 0.13305275815421613,
      "flos": 17931275729280.0,
      "grad_norm": 2.484635795733661,
      "language_loss": 0.74228692,
      "learning_rate": 3.889628839737908e-06,
      "loss": 0.7641542,
      "num_input_tokens_seen": 47888315,
      "step": 2213,
      "time_per_iteration": 2.755777597427368
    },
    {
      "auxiliary_loss_clip": 0.01114671,
      "auxiliary_loss_mlp": 0.01051459,
      "balance_loss_clip": 1.04682255,
      "balance_loss_mlp": 1.03231359,
      "epoch": 0.13311288140688413,
      "flos": 22340889999360.0,
      "grad_norm": 1.850943077435394,
      "language_loss": 0.79699469,
      "learning_rate": 3.889501213915291e-06,
      "loss": 0.81865597,
      "num_input_tokens_seen": 47906600,
      "step": 2214,
      "time_per_iteration": 2.702603340148926
    },
    {
      "auxiliary_loss_clip": 0.01143494,
      "auxiliary_loss_mlp": 0.01052411,
      "balance_loss_clip": 1.05555344,
      "balance_loss_mlp": 1.03171659,
      "epoch": 0.1331730046595521,
      "flos": 31868888682240.0,
      "grad_norm": 1.8782588426913054,
      "language_loss": 0.69341159,
      "learning_rate": 3.889373516442597e-06,
      "loss": 0.71537066,
      "num_input_tokens_seen": 47927630,
      "step": 2215,
      "time_per_iteration": 2.769237518310547
    },
    {
      "auxiliary_loss_clip": 0.01167307,
      "auxiliary_loss_mlp": 0.01051423,
      "balance_loss_clip": 1.06098068,
      "balance_loss_mlp": 1.03132463,
      "epoch": 0.13323312791222006,
      "flos": 22566589107840.0,
      "grad_norm": 1.884566493826098,
      "language_loss": 0.81262428,
      "learning_rate": 3.889245747324671e-06,
      "loss": 0.83481157,
      "num_input_tokens_seen": 47947935,
      "step": 2216,
      "time_per_iteration": 2.7427120208740234
    },
    {
      "auxiliary_loss_clip": 0.01163681,
      "auxiliary_loss_mlp": 0.01056545,
      "balance_loss_clip": 1.06198788,
      "balance_loss_mlp": 1.03631544,
      "epoch": 0.13329325116488802,
      "flos": 15085319293440.0,
      "grad_norm": 3.783334161704178,
      "language_loss": 0.87299347,
      "learning_rate": 3.889117906566356e-06,
      "loss": 0.89519572,
      "num_input_tokens_seen": 47965515,
      "step": 2217,
      "time_per_iteration": 2.709527015686035
    },
    {
      "auxiliary_loss_clip": 0.01152703,
      "auxiliary_loss_mlp": 0.01056364,
      "balance_loss_clip": 1.06054497,
      "balance_loss_mlp": 1.0343225,
      "epoch": 0.133353374417556,
      "flos": 27453671890560.0,
      "grad_norm": 4.412823416345162,
      "language_loss": 0.73105222,
      "learning_rate": 3.888989994172501e-06,
      "loss": 0.75314289,
      "num_input_tokens_seen": 47985675,
      "step": 2218,
      "time_per_iteration": 2.697733163833618
    },
    {
      "auxiliary_loss_clip": 0.01129106,
      "auxiliary_loss_mlp": 0.01051151,
      "balance_loss_clip": 1.0535965,
      "balance_loss_mlp": 1.02993202,
      "epoch": 0.13341349767022395,
      "flos": 24094695456000.0,
      "grad_norm": 1.7935349411013712,
      "language_loss": 0.86911142,
      "learning_rate": 3.8888620101479565e-06,
      "loss": 0.89091408,
      "num_input_tokens_seen": 48004985,
      "step": 2219,
      "time_per_iteration": 2.7641642093658447
    },
    {
      "auxiliary_loss_clip": 0.01141172,
      "auxiliary_loss_mlp": 0.0106326,
      "balance_loss_clip": 1.05751657,
      "balance_loss_mlp": 1.04406714,
      "epoch": 0.13347362092289192,
      "flos": 24133335511680.0,
      "grad_norm": 1.8604531362737113,
      "language_loss": 0.77244747,
      "learning_rate": 3.888733954497574e-06,
      "loss": 0.79449183,
      "num_input_tokens_seen": 48024965,
      "step": 2220,
      "time_per_iteration": 2.732160806655884
    },
    {
      "auxiliary_loss_clip": 0.01146487,
      "auxiliary_loss_mlp": 0.01048662,
      "balance_loss_clip": 1.05399704,
      "balance_loss_mlp": 1.03001785,
      "epoch": 0.1335337441755599,
      "flos": 18436538390400.0,
      "grad_norm": 2.3004113327688955,
      "language_loss": 0.79467338,
      "learning_rate": 3.888605827226212e-06,
      "loss": 0.81662482,
      "num_input_tokens_seen": 48040890,
      "step": 2221,
      "time_per_iteration": 2.685612440109253
    },
    {
      "auxiliary_loss_clip": 0.01062777,
      "auxiliary_loss_mlp": 0.01021711,
      "balance_loss_clip": 1.03293467,
      "balance_loss_mlp": 1.0194701,
      "epoch": 0.13359386742822787,
      "flos": 50611997652480.0,
      "grad_norm": 0.9755051104211709,
      "language_loss": 0.68938822,
      "learning_rate": 3.8884776283387275e-06,
      "loss": 0.71023309,
      "num_input_tokens_seen": 48091855,
      "step": 2222,
      "time_per_iteration": 3.0336835384368896
    },
    {
      "auxiliary_loss_clip": 0.01130152,
      "auxiliary_loss_mlp": 0.01058574,
      "balance_loss_clip": 1.05544209,
      "balance_loss_mlp": 1.03940475,
      "epoch": 0.13365399068089584,
      "flos": 22778569221120.0,
      "grad_norm": 2.1295993667823416,
      "language_loss": 0.67389107,
      "learning_rate": 3.888349357839982e-06,
      "loss": 0.69577825,
      "num_input_tokens_seen": 48111350,
      "step": 2223,
      "time_per_iteration": 2.7134146690368652
    },
    {
      "auxiliary_loss_clip": 0.01161386,
      "auxiliary_loss_mlp": 0.01060571,
      "balance_loss_clip": 1.05785358,
      "balance_loss_mlp": 1.04010296,
      "epoch": 0.1337141139335638,
      "flos": 12531603911040.0,
      "grad_norm": 4.277142483609355,
      "language_loss": 0.82505226,
      "learning_rate": 3.88822101573484e-06,
      "loss": 0.84727186,
      "num_input_tokens_seen": 48129840,
      "step": 2224,
      "time_per_iteration": 2.608372926712036
    },
    {
      "auxiliary_loss_clip": 0.01173412,
      "auxiliary_loss_mlp": 0.01050086,
      "balance_loss_clip": 1.0573926,
      "balance_loss_mlp": 1.0290221,
      "epoch": 0.13377423718623177,
      "flos": 23038957889280.0,
      "grad_norm": 1.9890294619132924,
      "language_loss": 0.66270435,
      "learning_rate": 3.888092602028167e-06,
      "loss": 0.68493932,
      "num_input_tokens_seen": 48149240,
      "step": 2225,
      "time_per_iteration": 2.6304945945739746
    },
    {
      "auxiliary_loss_clip": 0.01153626,
      "auxiliary_loss_mlp": 0.01051637,
      "balance_loss_clip": 1.05233717,
      "balance_loss_mlp": 1.03180075,
      "epoch": 0.13383436043889974,
      "flos": 16216397637120.0,
      "grad_norm": 2.2915668787246997,
      "language_loss": 0.89469218,
      "learning_rate": 3.887964116724835e-06,
      "loss": 0.91674477,
      "num_input_tokens_seen": 48166330,
      "step": 2226,
      "time_per_iteration": 2.6002328395843506
    },
    {
      "auxiliary_loss_clip": 0.01150395,
      "auxiliary_loss_mlp": 0.01054296,
      "balance_loss_clip": 1.0549798,
      "balance_loss_mlp": 1.03423262,
      "epoch": 0.1338944836915677,
      "flos": 24279671520000.0,
      "grad_norm": 1.7271512115821777,
      "language_loss": 0.73209751,
      "learning_rate": 3.887835559829712e-06,
      "loss": 0.75414443,
      "num_input_tokens_seen": 48187600,
      "step": 2227,
      "time_per_iteration": 2.706193447113037
    },
    {
      "auxiliary_loss_clip": 0.01157707,
      "auxiliary_loss_mlp": 0.01047387,
      "balance_loss_clip": 1.05518484,
      "balance_loss_mlp": 1.02683568,
      "epoch": 0.1339546069442357,
      "flos": 17598742594560.0,
      "grad_norm": 2.848999829625599,
      "language_loss": 0.85160232,
      "learning_rate": 3.8877069313476764e-06,
      "loss": 0.87365323,
      "num_input_tokens_seen": 48204400,
      "step": 2228,
      "time_per_iteration": 2.689209222793579
    },
    {
      "auxiliary_loss_clip": 0.01132803,
      "auxiliary_loss_mlp": 0.01052829,
      "balance_loss_clip": 1.04935181,
      "balance_loss_mlp": 1.03126431,
      "epoch": 0.13401473019690366,
      "flos": 18990065952000.0,
      "grad_norm": 1.909679794697233,
      "language_loss": 0.81460214,
      "learning_rate": 3.8875782312836054e-06,
      "loss": 0.83645844,
      "num_input_tokens_seen": 48222180,
      "step": 2229,
      "time_per_iteration": 2.6380228996276855
    },
    {
      "auxiliary_loss_clip": 0.0110557,
      "auxiliary_loss_mlp": 0.01052684,
      "balance_loss_clip": 1.04774594,
      "balance_loss_mlp": 1.03233457,
      "epoch": 0.13407485344957162,
      "flos": 26943812288640.0,
      "grad_norm": 1.7464076089691416,
      "language_loss": 0.73822236,
      "learning_rate": 3.887449459642378e-06,
      "loss": 0.7598049,
      "num_input_tokens_seen": 48243245,
      "step": 2230,
      "time_per_iteration": 2.7332983016967773
    },
    {
      "auxiliary_loss_clip": 0.01125236,
      "auxiliary_loss_mlp": 0.01058977,
      "balance_loss_clip": 1.05213606,
      "balance_loss_mlp": 1.03890252,
      "epoch": 0.1341349767022396,
      "flos": 20339373375360.0,
      "grad_norm": 1.6827882777998602,
      "language_loss": 0.80133682,
      "learning_rate": 3.8873206164288785e-06,
      "loss": 0.82317901,
      "num_input_tokens_seen": 48262600,
      "step": 2231,
      "time_per_iteration": 2.6759045124053955
    },
    {
      "auxiliary_loss_clip": 0.01111387,
      "auxiliary_loss_mlp": 0.01057582,
      "balance_loss_clip": 1.04997492,
      "balance_loss_mlp": 1.03499198,
      "epoch": 0.13419509995490755,
      "flos": 29862020931840.0,
      "grad_norm": 1.746756846769887,
      "language_loss": 0.72152746,
      "learning_rate": 3.887191701647992e-06,
      "loss": 0.74321723,
      "num_input_tokens_seen": 48285075,
      "step": 2232,
      "time_per_iteration": 4.391890048980713
    },
    {
      "auxiliary_loss_clip": 0.0112104,
      "auxiliary_loss_mlp": 0.01051805,
      "balance_loss_clip": 1.0481019,
      "balance_loss_mlp": 1.03039551,
      "epoch": 0.13425522320757552,
      "flos": 26942986275840.0,
      "grad_norm": 2.4719586176391686,
      "language_loss": 0.65116024,
      "learning_rate": 3.8870627153046066e-06,
      "loss": 0.67288864,
      "num_input_tokens_seen": 48301285,
      "step": 2233,
      "time_per_iteration": 4.234508037567139
    },
    {
      "auxiliary_loss_clip": 0.01167005,
      "auxiliary_loss_mlp": 0.0104461,
      "balance_loss_clip": 1.05189967,
      "balance_loss_mlp": 1.02421367,
      "epoch": 0.1343153464602435,
      "flos": 15777281871360.0,
      "grad_norm": 2.4864430088666656,
      "language_loss": 0.80878961,
      "learning_rate": 3.886933657403615e-06,
      "loss": 0.8309058,
      "num_input_tokens_seen": 48317835,
      "step": 2234,
      "time_per_iteration": 4.175215005874634
    },
    {
      "auxiliary_loss_clip": 0.01140761,
      "auxiliary_loss_mlp": 0.01054039,
      "balance_loss_clip": 1.05052733,
      "balance_loss_mlp": 1.03268874,
      "epoch": 0.13437546971291148,
      "flos": 24314756129280.0,
      "grad_norm": 2.0569321713284827,
      "language_loss": 0.82114553,
      "learning_rate": 3.886804527949909e-06,
      "loss": 0.84309351,
      "num_input_tokens_seen": 48335670,
      "step": 2235,
      "time_per_iteration": 2.6588025093078613
    },
    {
      "auxiliary_loss_clip": 0.01149093,
      "auxiliary_loss_mlp": 0.01052015,
      "balance_loss_clip": 1.05040097,
      "balance_loss_mlp": 1.02983022,
      "epoch": 0.13443559296557944,
      "flos": 26650673395200.0,
      "grad_norm": 1.6363146905087136,
      "language_loss": 0.86092007,
      "learning_rate": 3.8866753269483864e-06,
      "loss": 0.88293117,
      "num_input_tokens_seen": 48357805,
      "step": 2236,
      "time_per_iteration": 4.349383592605591
    },
    {
      "auxiliary_loss_clip": 0.01166751,
      "auxiliary_loss_mlp": 0.01047925,
      "balance_loss_clip": 1.05288053,
      "balance_loss_mlp": 1.02724242,
      "epoch": 0.1344957162182474,
      "flos": 21796197183360.0,
      "grad_norm": 1.82135056053112,
      "language_loss": 0.77258497,
      "learning_rate": 3.886546054403946e-06,
      "loss": 0.79473174,
      "num_input_tokens_seen": 48377845,
      "step": 2237,
      "time_per_iteration": 2.6398766040802
    },
    {
      "auxiliary_loss_clip": 0.01145425,
      "auxiliary_loss_mlp": 0.01051006,
      "balance_loss_clip": 1.05016851,
      "balance_loss_mlp": 1.02919102,
      "epoch": 0.13455583947091537,
      "flos": 19865568049920.0,
      "grad_norm": 2.440947698046141,
      "language_loss": 0.78772336,
      "learning_rate": 3.886416710321491e-06,
      "loss": 0.80968761,
      "num_input_tokens_seen": 48394735,
      "step": 2238,
      "time_per_iteration": 2.6556923389434814
    },
    {
      "auxiliary_loss_clip": 0.01141594,
      "auxiliary_loss_mlp": 0.01050085,
      "balance_loss_clip": 1.05123293,
      "balance_loss_mlp": 1.02878201,
      "epoch": 0.13461596272358334,
      "flos": 30846835094400.0,
      "grad_norm": 2.9136729194949735,
      "language_loss": 0.68486369,
      "learning_rate": 3.886287294705924e-06,
      "loss": 0.70678043,
      "num_input_tokens_seen": 48414200,
      "step": 2239,
      "time_per_iteration": 2.6778814792633057
    },
    {
      "auxiliary_loss_clip": 0.01147129,
      "auxiliary_loss_mlp": 0.01052633,
      "balance_loss_clip": 1.0515976,
      "balance_loss_mlp": 1.03197384,
      "epoch": 0.1346760859762513,
      "flos": 12494436312960.0,
      "grad_norm": 2.3763106012672925,
      "language_loss": 0.81277847,
      "learning_rate": 3.8861578075621555e-06,
      "loss": 0.8347761,
      "num_input_tokens_seen": 48431065,
      "step": 2240,
      "time_per_iteration": 2.5920939445495605
    },
    {
      "auxiliary_loss_clip": 0.01107793,
      "auxiliary_loss_mlp": 0.01049909,
      "balance_loss_clip": 1.04459488,
      "balance_loss_mlp": 1.02884459,
      "epoch": 0.1347362092289193,
      "flos": 21836022387840.0,
      "grad_norm": 1.7269080191231387,
      "language_loss": 0.77183759,
      "learning_rate": 3.886028248895093e-06,
      "loss": 0.79341465,
      "num_input_tokens_seen": 48450335,
      "step": 2241,
      "time_per_iteration": 2.7224419116973877
    },
    {
      "auxiliary_loss_clip": 0.0116331,
      "auxiliary_loss_mlp": 0.01041419,
      "balance_loss_clip": 1.05439126,
      "balance_loss_mlp": 1.02324009,
      "epoch": 0.13479633248158726,
      "flos": 23509459163520.0,
      "grad_norm": 2.0305903786470743,
      "language_loss": 0.83062387,
      "learning_rate": 3.88589861870965e-06,
      "loss": 0.85267115,
      "num_input_tokens_seen": 48468555,
      "step": 2242,
      "time_per_iteration": 2.5794169902801514
    },
    {
      "auxiliary_loss_clip": 0.01170048,
      "auxiliary_loss_mlp": 0.01056609,
      "balance_loss_clip": 1.05504107,
      "balance_loss_mlp": 1.03469825,
      "epoch": 0.13485645573425523,
      "flos": 29344332165120.0,
      "grad_norm": 2.465549548535016,
      "language_loss": 0.6498239,
      "learning_rate": 3.885768917010744e-06,
      "loss": 0.67209053,
      "num_input_tokens_seen": 48488515,
      "step": 2243,
      "time_per_iteration": 2.6709110736846924
    },
    {
      "auxiliary_loss_clip": 0.01125086,
      "auxiliary_loss_mlp": 0.01046786,
      "balance_loss_clip": 1.04593956,
      "balance_loss_mlp": 1.02618706,
      "epoch": 0.1349165789869232,
      "flos": 28037112503040.0,
      "grad_norm": 1.7770524512670738,
      "language_loss": 0.72633034,
      "learning_rate": 3.8856391438032895e-06,
      "loss": 0.74804902,
      "num_input_tokens_seen": 48510515,
      "step": 2244,
      "time_per_iteration": 2.713803768157959
    },
    {
      "auxiliary_loss_clip": 0.0115377,
      "auxiliary_loss_mlp": 0.0105148,
      "balance_loss_clip": 1.05312431,
      "balance_loss_mlp": 1.03209639,
      "epoch": 0.13497670223959116,
      "flos": 22853730430080.0,
      "grad_norm": 1.7564166456764931,
      "language_loss": 0.86023217,
      "learning_rate": 3.88550929909221e-06,
      "loss": 0.88228464,
      "num_input_tokens_seen": 48529940,
      "step": 2245,
      "time_per_iteration": 2.626560926437378
    },
    {
      "auxiliary_loss_clip": 0.01149467,
      "auxiliary_loss_mlp": 0.0105327,
      "balance_loss_clip": 1.05035663,
      "balance_loss_mlp": 1.03346968,
      "epoch": 0.13503682549225912,
      "flos": 16504580453760.0,
      "grad_norm": 1.7861449859595755,
      "language_loss": 0.78912753,
      "learning_rate": 3.88537938288243e-06,
      "loss": 0.8111549,
      "num_input_tokens_seen": 48548190,
      "step": 2246,
      "time_per_iteration": 2.6543703079223633
    },
    {
      "auxiliary_loss_clip": 0.010304,
      "auxiliary_loss_mlp": 0.01015407,
      "balance_loss_clip": 1.03666449,
      "balance_loss_mlp": 1.01285601,
      "epoch": 0.1350969487449271,
      "flos": 70756303242240.0,
      "grad_norm": 0.7509256694227144,
      "language_loss": 0.6054731,
      "learning_rate": 3.885249395178874e-06,
      "loss": 0.62593114,
      "num_input_tokens_seen": 48613165,
      "step": 2247,
      "time_per_iteration": 3.3349809646606445
    },
    {
      "auxiliary_loss_clip": 0.01162017,
      "auxiliary_loss_mlp": 0.01056869,
      "balance_loss_clip": 1.05492628,
      "balance_loss_mlp": 1.03470767,
      "epoch": 0.13515707199759508,
      "flos": 23075981832960.0,
      "grad_norm": 2.562042993856578,
      "language_loss": 0.80841738,
      "learning_rate": 3.885119335986473e-06,
      "loss": 0.83060622,
      "num_input_tokens_seen": 48631705,
      "step": 2248,
      "time_per_iteration": 2.6279287338256836
    },
    {
      "auxiliary_loss_clip": 0.0114073,
      "auxiliary_loss_mlp": 0.01049128,
      "balance_loss_clip": 1.05086231,
      "balance_loss_mlp": 1.03054309,
      "epoch": 0.13521719525026304,
      "flos": 23186371305600.0,
      "grad_norm": 1.9247838227480492,
      "language_loss": 0.77108699,
      "learning_rate": 3.884989205310157e-06,
      "loss": 0.79298556,
      "num_input_tokens_seen": 48649740,
      "step": 2249,
      "time_per_iteration": 2.7100210189819336
    },
    {
      "auxiliary_loss_clip": 0.0112733,
      "auxiliary_loss_mlp": 0.01057649,
      "balance_loss_clip": 1.05325472,
      "balance_loss_mlp": 1.03863478,
      "epoch": 0.135277318502931,
      "flos": 24790931752320.0,
      "grad_norm": 1.7403695434994237,
      "language_loss": 0.84457541,
      "learning_rate": 3.884859003154862e-06,
      "loss": 0.86642522,
      "num_input_tokens_seen": 48671565,
      "step": 2250,
      "time_per_iteration": 2.789350986480713
    },
    {
      "auxiliary_loss_clip": 0.01155547,
      "auxiliary_loss_mlp": 0.0105348,
      "balance_loss_clip": 1.05310512,
      "balance_loss_mlp": 1.03243995,
      "epoch": 0.13533744175559898,
      "flos": 21908525990400.0,
      "grad_norm": 3.018154510939524,
      "language_loss": 0.81796515,
      "learning_rate": 3.884728729525524e-06,
      "loss": 0.84005541,
      "num_input_tokens_seen": 48690425,
      "step": 2251,
      "time_per_iteration": 2.685617208480835
    },
    {
      "auxiliary_loss_clip": 0.01165433,
      "auxiliary_loss_mlp": 0.01060257,
      "balance_loss_clip": 1.05235004,
      "balance_loss_mlp": 1.03888273,
      "epoch": 0.13539756500826694,
      "flos": 21211643249280.0,
      "grad_norm": 1.7680273527580506,
      "language_loss": 0.86173487,
      "learning_rate": 3.884598384427084e-06,
      "loss": 0.88399172,
      "num_input_tokens_seen": 48707505,
      "step": 2252,
      "time_per_iteration": 2.597219467163086
    },
    {
      "auxiliary_loss_clip": 0.01052296,
      "auxiliary_loss_mlp": 0.01018557,
      "balance_loss_clip": 1.02446079,
      "balance_loss_mlp": 1.01632786,
      "epoch": 0.1354576882609349,
      "flos": 63242103634560.0,
      "grad_norm": 0.8028920055572067,
      "language_loss": 0.61837333,
      "learning_rate": 3.884467967864485e-06,
      "loss": 0.6390819,
      "num_input_tokens_seen": 48775895,
      "step": 2253,
      "time_per_iteration": 3.25115704536438
    },
    {
      "auxiliary_loss_clip": 0.01155107,
      "auxiliary_loss_mlp": 0.01055639,
      "balance_loss_clip": 1.0539906,
      "balance_loss_mlp": 1.03587449,
      "epoch": 0.1355178115136029,
      "flos": 25483037984640.0,
      "grad_norm": 1.6376691715964824,
      "language_loss": 0.89441288,
      "learning_rate": 3.884337479842671e-06,
      "loss": 0.91652036,
      "num_input_tokens_seen": 48798370,
      "step": 2254,
      "time_per_iteration": 2.6803932189941406
    },
    {
      "auxiliary_loss_clip": 0.01131786,
      "auxiliary_loss_mlp": 0.01063066,
      "balance_loss_clip": 1.04506016,
      "balance_loss_mlp": 1.03872383,
      "epoch": 0.13557793476627086,
      "flos": 21616967295360.0,
      "grad_norm": 2.1104776784573787,
      "language_loss": 0.84626925,
      "learning_rate": 3.884206920366591e-06,
      "loss": 0.86821771,
      "num_input_tokens_seen": 48817955,
      "step": 2255,
      "time_per_iteration": 2.7074074745178223
    },
    {
      "auxiliary_loss_clip": 0.01165481,
      "auxiliary_loss_mlp": 0.01058458,
      "balance_loss_clip": 1.05211091,
      "balance_loss_mlp": 1.03767991,
      "epoch": 0.13563805801893883,
      "flos": 24928253447040.0,
      "grad_norm": 4.791676738707355,
      "language_loss": 0.74684238,
      "learning_rate": 3.884076289441196e-06,
      "loss": 0.76908177,
      "num_input_tokens_seen": 48836330,
      "step": 2256,
      "time_per_iteration": 2.590178966522217
    },
    {
      "auxiliary_loss_clip": 0.01127027,
      "auxiliary_loss_mlp": 0.01054317,
      "balance_loss_clip": 1.04977024,
      "balance_loss_mlp": 1.03338361,
      "epoch": 0.1356981812716068,
      "flos": 14750272206720.0,
      "grad_norm": 5.890843360804152,
      "language_loss": 0.8309083,
      "learning_rate": 3.88394558707144e-06,
      "loss": 0.85272169,
      "num_input_tokens_seen": 48851890,
      "step": 2257,
      "time_per_iteration": 2.642096519470215
    },
    {
      "auxiliary_loss_clip": 0.0114984,
      "auxiliary_loss_mlp": 0.00780177,
      "balance_loss_clip": 1.05128407,
      "balance_loss_mlp": 1.00013828,
      "epoch": 0.13575830452427476,
      "flos": 11108571822720.0,
      "grad_norm": 2.1957250492246505,
      "language_loss": 0.82045269,
      "learning_rate": 3.883814813262277e-06,
      "loss": 0.83975297,
      "num_input_tokens_seen": 48865510,
      "step": 2258,
      "time_per_iteration": 2.6279473304748535
    },
    {
      "auxiliary_loss_clip": 0.01155515,
      "auxiliary_loss_mlp": 0.01054519,
      "balance_loss_clip": 1.05172098,
      "balance_loss_mlp": 1.03152323,
      "epoch": 0.13581842777694272,
      "flos": 17960290940160.0,
      "grad_norm": 2.6364031487830464,
      "language_loss": 0.82694167,
      "learning_rate": 3.883683968018669e-06,
      "loss": 0.849042,
      "num_input_tokens_seen": 48882360,
      "step": 2259,
      "time_per_iteration": 2.677804708480835
    },
    {
      "auxiliary_loss_clip": 0.01127201,
      "auxiliary_loss_mlp": 0.01054646,
      "balance_loss_clip": 1.0495683,
      "balance_loss_mlp": 1.03547728,
      "epoch": 0.1358785510296107,
      "flos": 22857142222080.0,
      "grad_norm": 2.0790748617118853,
      "language_loss": 0.73916006,
      "learning_rate": 3.8835530513455755e-06,
      "loss": 0.76097858,
      "num_input_tokens_seen": 48902700,
      "step": 2260,
      "time_per_iteration": 2.7416799068450928
    },
    {
      "auxiliary_loss_clip": 0.01144177,
      "auxiliary_loss_mlp": 0.01056881,
      "balance_loss_clip": 1.05196047,
      "balance_loss_mlp": 1.03691387,
      "epoch": 0.13593867428227868,
      "flos": 25739404329600.0,
      "grad_norm": 3.546593987683097,
      "language_loss": 0.74799728,
      "learning_rate": 3.883422063247961e-06,
      "loss": 0.77000785,
      "num_input_tokens_seen": 48922525,
      "step": 2261,
      "time_per_iteration": 2.675342559814453
    },
    {
      "auxiliary_loss_clip": 0.01170469,
      "auxiliary_loss_mlp": 0.01050986,
      "balance_loss_clip": 1.05486035,
      "balance_loss_mlp": 1.03043413,
      "epoch": 0.13599879753494665,
      "flos": 31249214225280.0,
      "grad_norm": 2.967396076139427,
      "language_loss": 0.63602281,
      "learning_rate": 3.883291003730794e-06,
      "loss": 0.65823734,
      "num_input_tokens_seen": 48942510,
      "step": 2262,
      "time_per_iteration": 2.660538911819458
    },
    {
      "auxiliary_loss_clip": 0.01148004,
      "auxiliary_loss_mlp": 0.01052118,
      "balance_loss_clip": 1.0516696,
      "balance_loss_mlp": 1.03216195,
      "epoch": 0.1360589207876146,
      "flos": 23915034604800.0,
      "grad_norm": 2.301949377353301,
      "language_loss": 0.81810403,
      "learning_rate": 3.883159872799043e-06,
      "loss": 0.84010524,
      "num_input_tokens_seen": 48962625,
      "step": 2263,
      "time_per_iteration": 2.840043783187866
    },
    {
      "auxiliary_loss_clip": 0.01098888,
      "auxiliary_loss_mlp": 0.01064302,
      "balance_loss_clip": 1.04875195,
      "balance_loss_mlp": 1.0410558,
      "epoch": 0.13611904404028258,
      "flos": 19974197756160.0,
      "grad_norm": 1.7561035968690553,
      "language_loss": 0.87737143,
      "learning_rate": 3.8830286704576815e-06,
      "loss": 0.89900339,
      "num_input_tokens_seen": 48982525,
      "step": 2264,
      "time_per_iteration": 2.784648895263672
    },
    {
      "auxiliary_loss_clip": 0.01157618,
      "auxiliary_loss_mlp": 0.01049521,
      "balance_loss_clip": 1.05161715,
      "balance_loss_mlp": 1.02709746,
      "epoch": 0.13617916729295054,
      "flos": 15340644144000.0,
      "grad_norm": 3.151792845640157,
      "language_loss": 0.7115528,
      "learning_rate": 3.882897396711683e-06,
      "loss": 0.7336241,
      "num_input_tokens_seen": 48997605,
      "step": 2265,
      "time_per_iteration": 2.6108245849609375
    },
    {
      "auxiliary_loss_clip": 0.01111831,
      "auxiliary_loss_mlp": 0.01042545,
      "balance_loss_clip": 1.05199265,
      "balance_loss_mlp": 1.02256525,
      "epoch": 0.1362392905456185,
      "flos": 27451445247360.0,
      "grad_norm": 4.918827494175735,
      "language_loss": 0.6671263,
      "learning_rate": 3.882766051566027e-06,
      "loss": 0.68867004,
      "num_input_tokens_seen": 49018535,
      "step": 2266,
      "time_per_iteration": 2.7810373306274414
    },
    {
      "auxiliary_loss_clip": 0.01127539,
      "auxiliary_loss_mlp": 0.01057589,
      "balance_loss_clip": 1.05683684,
      "balance_loss_mlp": 1.03739524,
      "epoch": 0.1362994137982865,
      "flos": 25009017177600.0,
      "grad_norm": 1.707924588861666,
      "language_loss": 0.7634865,
      "learning_rate": 3.882634635025694e-06,
      "loss": 0.78533769,
      "num_input_tokens_seen": 49038865,
      "step": 2267,
      "time_per_iteration": 2.7682721614837646
    },
    {
      "auxiliary_loss_clip": 0.01133448,
      "auxiliary_loss_mlp": 0.01048207,
      "balance_loss_clip": 1.04668903,
      "balance_loss_mlp": 1.02641535,
      "epoch": 0.13635953705095447,
      "flos": 20303031790080.0,
      "grad_norm": 2.9531688260339934,
      "language_loss": 0.81653506,
      "learning_rate": 3.882503147095667e-06,
      "loss": 0.83835161,
      "num_input_tokens_seen": 49058010,
      "step": 2268,
      "time_per_iteration": 2.645081043243408
    },
    {
      "auxiliary_loss_clip": 0.01155147,
      "auxiliary_loss_mlp": 0.01048448,
      "balance_loss_clip": 1.05424881,
      "balance_loss_mlp": 1.02738333,
      "epoch": 0.13641966030362243,
      "flos": 31358418549120.0,
      "grad_norm": 1.9923150848418427,
      "language_loss": 0.75975174,
      "learning_rate": 3.882371587780931e-06,
      "loss": 0.78178769,
      "num_input_tokens_seen": 49080330,
      "step": 2269,
      "time_per_iteration": 2.6764814853668213
    },
    {
      "auxiliary_loss_clip": 0.0113465,
      "auxiliary_loss_mlp": 0.01049702,
      "balance_loss_clip": 1.04941857,
      "balance_loss_mlp": 1.02844727,
      "epoch": 0.1364797835562904,
      "flos": 20478095700480.0,
      "grad_norm": 2.1475090354855473,
      "language_loss": 0.81328762,
      "learning_rate": 3.882239957086477e-06,
      "loss": 0.83513117,
      "num_input_tokens_seen": 49097035,
      "step": 2270,
      "time_per_iteration": 2.6801655292510986
    },
    {
      "auxiliary_loss_clip": 0.01142111,
      "auxiliary_loss_mlp": 0.010594,
      "balance_loss_clip": 1.04989171,
      "balance_loss_mlp": 1.03773928,
      "epoch": 0.13653990680895836,
      "flos": 13078343802240.0,
      "grad_norm": 3.2227070482893976,
      "language_loss": 0.75812757,
      "learning_rate": 3.882108255017295e-06,
      "loss": 0.78014266,
      "num_input_tokens_seen": 49113945,
      "step": 2271,
      "time_per_iteration": 4.197805166244507
    },
    {
      "auxiliary_loss_clip": 0.01156913,
      "auxiliary_loss_mlp": 0.01061846,
      "balance_loss_clip": 1.05097795,
      "balance_loss_mlp": 1.03921962,
      "epoch": 0.13660003006162633,
      "flos": 16946712961920.0,
      "grad_norm": 2.2800716885469754,
      "language_loss": 0.80251753,
      "learning_rate": 3.881976481578379e-06,
      "loss": 0.82470512,
      "num_input_tokens_seen": 49132855,
      "step": 2272,
      "time_per_iteration": 4.1461029052734375
    },
    {
      "auxiliary_loss_clip": 0.01055091,
      "auxiliary_loss_mlp": 0.01042701,
      "balance_loss_clip": 1.02539539,
      "balance_loss_mlp": 1.04001904,
      "epoch": 0.1366601533142943,
      "flos": 68682749892480.0,
      "grad_norm": 0.7097054685047118,
      "language_loss": 0.60739923,
      "learning_rate": 3.8818446367747255e-06,
      "loss": 0.62837708,
      "num_input_tokens_seen": 49198310,
      "step": 2273,
      "time_per_iteration": 4.731219530105591
    },
    {
      "auxiliary_loss_clip": 0.01165514,
      "auxiliary_loss_mlp": 0.00780474,
      "balance_loss_clip": 1.0523783,
      "balance_loss_mlp": 1.00008452,
      "epoch": 0.13672027656696228,
      "flos": 19244241567360.0,
      "grad_norm": 2.4844725334882583,
      "language_loss": 0.77506429,
      "learning_rate": 3.881712720611336e-06,
      "loss": 0.79452413,
      "num_input_tokens_seen": 49217250,
      "step": 2274,
      "time_per_iteration": 2.7122738361358643
    },
    {
      "auxiliary_loss_clip": 0.01154937,
      "auxiliary_loss_mlp": 0.01054542,
      "balance_loss_clip": 1.05082417,
      "balance_loss_mlp": 1.03271496,
      "epoch": 0.13678039981963025,
      "flos": 24534924543360.0,
      "grad_norm": 2.391437383339344,
      "language_loss": 0.78256011,
      "learning_rate": 3.881580733093211e-06,
      "loss": 0.8046549,
      "num_input_tokens_seen": 49236615,
      "step": 2275,
      "time_per_iteration": 2.6674444675445557
    },
    {
      "auxiliary_loss_clip": 0.01154585,
      "auxiliary_loss_mlp": 0.01044634,
      "balance_loss_clip": 1.05220842,
      "balance_loss_mlp": 1.02449977,
      "epoch": 0.13684052307229821,
      "flos": 15669334523520.0,
      "grad_norm": 2.271072834476717,
      "language_loss": 0.81682789,
      "learning_rate": 3.881448674225356e-06,
      "loss": 0.83882004,
      "num_input_tokens_seen": 49253935,
      "step": 2276,
      "time_per_iteration": 4.202202558517456
    },
    {
      "auxiliary_loss_clip": 0.01164941,
      "auxiliary_loss_mlp": 0.01060078,
      "balance_loss_clip": 1.05228245,
      "balance_loss_mlp": 1.03604531,
      "epoch": 0.13690064632496618,
      "flos": 28364689560960.0,
      "grad_norm": 5.063053962589045,
      "language_loss": 0.69948691,
      "learning_rate": 3.881316544012779e-06,
      "loss": 0.72173715,
      "num_input_tokens_seen": 49273605,
      "step": 2277,
      "time_per_iteration": 2.708591938018799
    },
    {
      "auxiliary_loss_clip": 0.01160044,
      "auxiliary_loss_mlp": 0.00780297,
      "balance_loss_clip": 1.05169702,
      "balance_loss_mlp": 1.00017083,
      "epoch": 0.13696076957763414,
      "flos": 23404779953280.0,
      "grad_norm": 2.062701620585305,
      "language_loss": 0.80197465,
      "learning_rate": 3.88118434246049e-06,
      "loss": 0.82137805,
      "num_input_tokens_seen": 49291785,
      "step": 2278,
      "time_per_iteration": 2.6916158199310303
    },
    {
      "auxiliary_loss_clip": 0.01159146,
      "auxiliary_loss_mlp": 0.01060686,
      "balance_loss_clip": 1.05954766,
      "balance_loss_mlp": 1.03925228,
      "epoch": 0.1370208928303021,
      "flos": 37196595601920.0,
      "grad_norm": 7.088344486179519,
      "language_loss": 0.75048816,
      "learning_rate": 3.881052069573502e-06,
      "loss": 0.77268648,
      "num_input_tokens_seen": 49311405,
      "step": 2279,
      "time_per_iteration": 2.7316977977752686
    },
    {
      "auxiliary_loss_clip": 0.01101952,
      "auxiliary_loss_mlp": 0.01066685,
      "balance_loss_clip": 1.04605758,
      "balance_loss_mlp": 1.04485774,
      "epoch": 0.13708101608297008,
      "flos": 26976311118720.0,
      "grad_norm": 2.5293116992138223,
      "language_loss": 0.76743513,
      "learning_rate": 3.880919725356831e-06,
      "loss": 0.78912151,
      "num_input_tokens_seen": 49331835,
      "step": 2280,
      "time_per_iteration": 2.813720941543579
    },
    {
      "auxiliary_loss_clip": 0.01108594,
      "auxiliary_loss_mlp": 0.01060805,
      "balance_loss_clip": 1.04457331,
      "balance_loss_mlp": 1.04022956,
      "epoch": 0.13714113933563807,
      "flos": 32556864850560.0,
      "grad_norm": 2.0597640944890325,
      "language_loss": 0.79657966,
      "learning_rate": 3.880787309815496e-06,
      "loss": 0.81827366,
      "num_input_tokens_seen": 49352290,
      "step": 2281,
      "time_per_iteration": 2.8325345516204834
    },
    {
      "auxiliary_loss_clip": 0.0117656,
      "auxiliary_loss_mlp": 0.0107773,
      "balance_loss_clip": 1.05715084,
      "balance_loss_mlp": 1.05671358,
      "epoch": 0.13720126258830603,
      "flos": 16101267569280.0,
      "grad_norm": 2.0769142230572877,
      "language_loss": 0.83383757,
      "learning_rate": 3.880654822954518e-06,
      "loss": 0.85638046,
      "num_input_tokens_seen": 49370285,
      "step": 2282,
      "time_per_iteration": 2.5988755226135254
    },
    {
      "auxiliary_loss_clip": 0.01142098,
      "auxiliary_loss_mlp": 0.01075909,
      "balance_loss_clip": 1.04898703,
      "balance_loss_mlp": 1.05583453,
      "epoch": 0.137261385840974,
      "flos": 18953544798720.0,
      "grad_norm": 1.5269487193470777,
      "language_loss": 0.73526621,
      "learning_rate": 3.8805222647789195e-06,
      "loss": 0.75744629,
      "num_input_tokens_seen": 49389610,
      "step": 2283,
      "time_per_iteration": 2.7099714279174805
    },
    {
      "auxiliary_loss_clip": 0.01160178,
      "auxiliary_loss_mlp": 0.01062577,
      "balance_loss_clip": 1.05577087,
      "balance_loss_mlp": 1.04173923,
      "epoch": 0.13732150909364196,
      "flos": 23295360147840.0,
      "grad_norm": 2.2306012559941455,
      "language_loss": 0.83934438,
      "learning_rate": 3.880389635293729e-06,
      "loss": 0.86157191,
      "num_input_tokens_seen": 49408390,
      "step": 2284,
      "time_per_iteration": 2.7315831184387207
    },
    {
      "auxiliary_loss_clip": 0.01151427,
      "auxiliary_loss_mlp": 0.01070288,
      "balance_loss_clip": 1.05204272,
      "balance_loss_mlp": 1.04779351,
      "epoch": 0.13738163234630993,
      "flos": 29351263489920.0,
      "grad_norm": 2.0900141273659223,
      "language_loss": 0.7557056,
      "learning_rate": 3.880256934503974e-06,
      "loss": 0.77792281,
      "num_input_tokens_seen": 49427725,
      "step": 2285,
      "time_per_iteration": 2.7257747650146484
    },
    {
      "auxiliary_loss_clip": 0.01144078,
      "auxiliary_loss_mlp": 0.01064539,
      "balance_loss_clip": 1.05233073,
      "balance_loss_mlp": 1.04392731,
      "epoch": 0.1374417555989779,
      "flos": 26651319840000.0,
      "grad_norm": 2.727019945657865,
      "language_loss": 0.74521589,
      "learning_rate": 3.880124162414689e-06,
      "loss": 0.76730204,
      "num_input_tokens_seen": 49449000,
      "step": 2286,
      "time_per_iteration": 2.742582082748413
    },
    {
      "auxiliary_loss_clip": 0.0112541,
      "auxiliary_loss_mlp": 0.01059198,
      "balance_loss_clip": 1.04906356,
      "balance_loss_mlp": 1.03659606,
      "epoch": 0.1375018788516459,
      "flos": 28403401443840.0,
      "grad_norm": 2.2168449035378357,
      "language_loss": 0.86683542,
      "learning_rate": 3.879991319030908e-06,
      "loss": 0.88868147,
      "num_input_tokens_seen": 49468360,
      "step": 2287,
      "time_per_iteration": 2.802088499069214
    },
    {
      "auxiliary_loss_clip": 0.01124712,
      "auxiliary_loss_mlp": 0.01064517,
      "balance_loss_clip": 1.04803944,
      "balance_loss_mlp": 1.04207003,
      "epoch": 0.13756200210431385,
      "flos": 37413783187200.0,
      "grad_norm": 2.0592152854463106,
      "language_loss": 0.68410838,
      "learning_rate": 3.879858404357666e-06,
      "loss": 0.70600063,
      "num_input_tokens_seen": 49493450,
      "step": 2288,
      "time_per_iteration": 2.861175537109375
    },
    {
      "auxiliary_loss_clip": 0.01112106,
      "auxiliary_loss_mlp": 0.01071262,
      "balance_loss_clip": 1.05062151,
      "balance_loss_mlp": 1.04666936,
      "epoch": 0.13762212535698182,
      "flos": 22711021695360.0,
      "grad_norm": 2.3933568244149357,
      "language_loss": 0.87090456,
      "learning_rate": 3.879725418400005e-06,
      "loss": 0.89273822,
      "num_input_tokens_seen": 49511220,
      "step": 2289,
      "time_per_iteration": 2.7185773849487305
    },
    {
      "auxiliary_loss_clip": 0.01130193,
      "auxiliary_loss_mlp": 0.00781167,
      "balance_loss_clip": 1.0480957,
      "balance_loss_mlp": 1.00019848,
      "epoch": 0.13768224860964978,
      "flos": 23952130375680.0,
      "grad_norm": 1.8106848287624444,
      "language_loss": 0.74668044,
      "learning_rate": 3.879592361162969e-06,
      "loss": 0.76579404,
      "num_input_tokens_seen": 49529820,
      "step": 2290,
      "time_per_iteration": 2.6751222610473633
    },
    {
      "auxiliary_loss_clip": 0.01039657,
      "auxiliary_loss_mlp": 0.01081332,
      "balance_loss_clip": 1.03094769,
      "balance_loss_mlp": 1.07881641,
      "epoch": 0.13774237186231775,
      "flos": 63590438753280.0,
      "grad_norm": 0.7179159366671727,
      "language_loss": 0.51597112,
      "learning_rate": 3.8794592326516015e-06,
      "loss": 0.53718102,
      "num_input_tokens_seen": 49595325,
      "step": 2291,
      "time_per_iteration": 3.2823359966278076
    },
    {
      "auxiliary_loss_clip": 0.01157406,
      "auxiliary_loss_mlp": 0.01052846,
      "balance_loss_clip": 1.05224037,
      "balance_loss_mlp": 1.03123331,
      "epoch": 0.1378024951149857,
      "flos": 24279456038400.0,
      "grad_norm": 1.9326408617769533,
      "language_loss": 0.71273667,
      "learning_rate": 3.879326032870952e-06,
      "loss": 0.7348392,
      "num_input_tokens_seen": 49615850,
      "step": 2292,
      "time_per_iteration": 2.74045729637146
    },
    {
      "auxiliary_loss_clip": 0.01156871,
      "auxiliary_loss_mlp": 0.01049315,
      "balance_loss_clip": 1.05427122,
      "balance_loss_mlp": 1.02931166,
      "epoch": 0.13786261836765368,
      "flos": 14021537080320.0,
      "grad_norm": 6.592759889378346,
      "language_loss": 0.8047784,
      "learning_rate": 3.879192761826071e-06,
      "loss": 0.82684022,
      "num_input_tokens_seen": 49631860,
      "step": 2293,
      "time_per_iteration": 2.587576389312744
    },
    {
      "auxiliary_loss_clip": 0.0115787,
      "auxiliary_loss_mlp": 0.0104972,
      "balance_loss_clip": 1.0554558,
      "balance_loss_mlp": 1.02921653,
      "epoch": 0.13792274162032167,
      "flos": 28878679226880.0,
      "grad_norm": 1.9082895606463517,
      "language_loss": 0.78440171,
      "learning_rate": 3.879059419522011e-06,
      "loss": 0.80647767,
      "num_input_tokens_seen": 49652145,
      "step": 2294,
      "time_per_iteration": 2.7152793407440186
    },
    {
      "auxiliary_loss_clip": 0.01126374,
      "auxiliary_loss_mlp": 0.01050648,
      "balance_loss_clip": 1.05281758,
      "balance_loss_mlp": 1.03104973,
      "epoch": 0.13798286487298964,
      "flos": 21141150808320.0,
      "grad_norm": 1.991103290125302,
      "language_loss": 0.80339509,
      "learning_rate": 3.878926005963831e-06,
      "loss": 0.82516527,
      "num_input_tokens_seen": 49669880,
      "step": 2295,
      "time_per_iteration": 2.7026021480560303
    },
    {
      "auxiliary_loss_clip": 0.01154693,
      "auxiliary_loss_mlp": 0.01052186,
      "balance_loss_clip": 1.05239046,
      "balance_loss_mlp": 1.03102624,
      "epoch": 0.1380429881256576,
      "flos": 22487477402880.0,
      "grad_norm": 1.7450624966187134,
      "language_loss": 0.78661883,
      "learning_rate": 3.878792521156588e-06,
      "loss": 0.80868757,
      "num_input_tokens_seen": 49687255,
      "step": 2296,
      "time_per_iteration": 2.566929340362549
    },
    {
      "auxiliary_loss_clip": 0.01153425,
      "auxiliary_loss_mlp": 0.01069343,
      "balance_loss_clip": 1.05437231,
      "balance_loss_mlp": 1.04811132,
      "epoch": 0.13810311137832557,
      "flos": 21393674398080.0,
      "grad_norm": 1.7434096141785573,
      "language_loss": 0.78663194,
      "learning_rate": 3.8786589651053446e-06,
      "loss": 0.80885959,
      "num_input_tokens_seen": 49706650,
      "step": 2297,
      "time_per_iteration": 2.6254489421844482
    },
    {
      "auxiliary_loss_clip": 0.01110905,
      "auxiliary_loss_mlp": 0.01059754,
      "balance_loss_clip": 1.05296302,
      "balance_loss_mlp": 1.03871369,
      "epoch": 0.13816323463099353,
      "flos": 25989844930560.0,
      "grad_norm": 1.929043788877404,
      "language_loss": 0.69199705,
      "learning_rate": 3.878525337815164e-06,
      "loss": 0.71370363,
      "num_input_tokens_seen": 49725715,
      "step": 2298,
      "time_per_iteration": 2.791301965713501
    },
    {
      "auxiliary_loss_clip": 0.01137772,
      "auxiliary_loss_mlp": 0.01061768,
      "balance_loss_clip": 1.0517292,
      "balance_loss_mlp": 1.04059684,
      "epoch": 0.1382233578836615,
      "flos": 19244313394560.0,
      "grad_norm": 1.7910922430646712,
      "language_loss": 0.86382294,
      "learning_rate": 3.878391639291116e-06,
      "loss": 0.88581836,
      "num_input_tokens_seen": 49744710,
      "step": 2299,
      "time_per_iteration": 2.6075453758239746
    },
    {
      "auxiliary_loss_clip": 0.01166817,
      "auxiliary_loss_mlp": 0.01054863,
      "balance_loss_clip": 1.05378175,
      "balance_loss_mlp": 1.03292871,
      "epoch": 0.1382834811363295,
      "flos": 25666290195840.0,
      "grad_norm": 2.2378660690879606,
      "language_loss": 0.75468475,
      "learning_rate": 3.878257869538267e-06,
      "loss": 0.77690154,
      "num_input_tokens_seen": 49764300,
      "step": 2300,
      "time_per_iteration": 2.663328170776367
    },
    {
      "auxiliary_loss_clip": 0.01130608,
      "auxiliary_loss_mlp": 0.01047248,
      "balance_loss_clip": 1.05274105,
      "balance_loss_mlp": 1.02664876,
      "epoch": 0.13834360438899745,
      "flos": 19784193788160.0,
      "grad_norm": 2.5571861214345963,
      "language_loss": 0.82463622,
      "learning_rate": 3.878124028561692e-06,
      "loss": 0.8464148,
      "num_input_tokens_seen": 49778380,
      "step": 2301,
      "time_per_iteration": 2.6705129146575928
    },
    {
      "auxiliary_loss_clip": 0.0113862,
      "auxiliary_loss_mlp": 0.00777879,
      "balance_loss_clip": 1.05323792,
      "balance_loss_mlp": 1.00021625,
      "epoch": 0.13840372764166542,
      "flos": 26651858544000.0,
      "grad_norm": 1.9612043619218924,
      "language_loss": 0.85957694,
      "learning_rate": 3.877990116366466e-06,
      "loss": 0.87874192,
      "num_input_tokens_seen": 49797460,
      "step": 2302,
      "time_per_iteration": 2.679797410964966
    },
    {
      "auxiliary_loss_clip": 0.01059341,
      "auxiliary_loss_mlp": 0.01025212,
      "balance_loss_clip": 1.03226125,
      "balance_loss_mlp": 1.02244604,
      "epoch": 0.13846385089433338,
      "flos": 70510998286080.0,
      "grad_norm": 0.7598813547967705,
      "language_loss": 0.65591633,
      "learning_rate": 3.877856132957667e-06,
      "loss": 0.67676187,
      "num_input_tokens_seen": 49868005,
      "step": 2303,
      "time_per_iteration": 3.3249399662017822
    },
    {
      "auxiliary_loss_clip": 0.01151443,
      "auxiliary_loss_mlp": 0.01046478,
      "balance_loss_clip": 1.05337632,
      "balance_loss_mlp": 1.02655792,
      "epoch": 0.13852397414700135,
      "flos": 17348732956800.0,
      "grad_norm": 3.141207945865242,
      "language_loss": 0.78663635,
      "learning_rate": 3.877722078340374e-06,
      "loss": 0.80861557,
      "num_input_tokens_seen": 49885825,
      "step": 2304,
      "time_per_iteration": 2.7364001274108887
    },
    {
      "auxiliary_loss_clip": 0.01157514,
      "auxiliary_loss_mlp": 0.01043253,
      "balance_loss_clip": 1.05566275,
      "balance_loss_mlp": 1.02385736,
      "epoch": 0.13858409739966931,
      "flos": 21543781334400.0,
      "grad_norm": 1.7487365854034607,
      "language_loss": 0.77559888,
      "learning_rate": 3.877587952519672e-06,
      "loss": 0.79760659,
      "num_input_tokens_seen": 49905975,
      "step": 2305,
      "time_per_iteration": 2.7814202308654785
    },
    {
      "auxiliary_loss_clip": 0.01074766,
      "auxiliary_loss_mlp": 0.01055718,
      "balance_loss_clip": 1.04160607,
      "balance_loss_mlp": 1.03473723,
      "epoch": 0.13864422065233728,
      "flos": 21579907438080.0,
      "grad_norm": 1.8207477060355044,
      "language_loss": 0.87737936,
      "learning_rate": 3.877453755500647e-06,
      "loss": 0.89868426,
      "num_input_tokens_seen": 49925800,
      "step": 2306,
      "time_per_iteration": 2.917616605758667
    },
    {
      "auxiliary_loss_clip": 0.01064826,
      "auxiliary_loss_mlp": 0.0101208,
      "balance_loss_clip": 1.02692199,
      "balance_loss_mlp": 1.0094099,
      "epoch": 0.13870434390500527,
      "flos": 53371156872960.0,
      "grad_norm": 0.8728538231155298,
      "language_loss": 0.59008431,
      "learning_rate": 3.877319487288387e-06,
      "loss": 0.61085337,
      "num_input_tokens_seen": 49977620,
      "step": 2307,
      "time_per_iteration": 3.4345149993896484
    },
    {
      "auxiliary_loss_clip": 0.01169624,
      "auxiliary_loss_mlp": 0.00778134,
      "balance_loss_clip": 1.05528641,
      "balance_loss_mlp": 1.00021303,
      "epoch": 0.13876446715767324,
      "flos": 22565906749440.0,
      "grad_norm": 1.8467673932802395,
      "language_loss": 0.79483795,
      "learning_rate": 3.877185147887984e-06,
      "loss": 0.81431556,
      "num_input_tokens_seen": 49996650,
      "step": 2308,
      "time_per_iteration": 2.7137296199798584
    },
    {
      "auxiliary_loss_clip": 0.01131024,
      "auxiliary_loss_mlp": 0.01050332,
      "balance_loss_clip": 1.05118585,
      "balance_loss_mlp": 1.03054297,
      "epoch": 0.1388245904103412,
      "flos": 20705231352960.0,
      "grad_norm": 2.352128383160346,
      "language_loss": 0.78101134,
      "learning_rate": 3.877050737304533e-06,
      "loss": 0.80282485,
      "num_input_tokens_seen": 50015640,
      "step": 2309,
      "time_per_iteration": 2.9259471893310547
    },
    {
      "auxiliary_loss_clip": 0.01128109,
      "auxiliary_loss_mlp": 0.01057348,
      "balance_loss_clip": 1.04979932,
      "balance_loss_mlp": 1.03620028,
      "epoch": 0.13888471366300917,
      "flos": 20554729367040.0,
      "grad_norm": 3.914796791761399,
      "language_loss": 0.68133545,
      "learning_rate": 3.876916255543129e-06,
      "loss": 0.70318997,
      "num_input_tokens_seen": 50033500,
      "step": 2310,
      "time_per_iteration": 4.27877140045166
    },
    {
      "auxiliary_loss_clip": 0.01164985,
      "auxiliary_loss_mlp": 0.01062516,
      "balance_loss_clip": 1.05356944,
      "balance_loss_mlp": 1.04021168,
      "epoch": 0.13894483691567713,
      "flos": 13838033473920.0,
      "grad_norm": 1.934954545600412,
      "language_loss": 0.84295756,
      "learning_rate": 3.8767817026088725e-06,
      "loss": 0.86523259,
      "num_input_tokens_seen": 50050075,
      "step": 2311,
      "time_per_iteration": 2.5612359046936035
    },
    {
      "auxiliary_loss_clip": 0.01173749,
      "auxiliary_loss_mlp": 0.01055474,
      "balance_loss_clip": 1.05752683,
      "balance_loss_mlp": 1.0350771,
      "epoch": 0.1390049601683451,
      "flos": 28031186759040.0,
      "grad_norm": 2.9213009430481143,
      "language_loss": 0.82358992,
      "learning_rate": 3.876647078506866e-06,
      "loss": 0.84588212,
      "num_input_tokens_seen": 50070080,
      "step": 2312,
      "time_per_iteration": 5.737139701843262
    },
    {
      "auxiliary_loss_clip": 0.01129781,
      "auxiliary_loss_mlp": 0.00778347,
      "balance_loss_clip": 1.05464363,
      "balance_loss_mlp": 1.00023031,
      "epoch": 0.13906508342101306,
      "flos": 26756860976640.0,
      "grad_norm": 2.109799495913242,
      "language_loss": 0.86732674,
      "learning_rate": 3.876512383242215e-06,
      "loss": 0.88640809,
      "num_input_tokens_seen": 50090040,
      "step": 2313,
      "time_per_iteration": 2.8402304649353027
    },
    {
      "auxiliary_loss_clip": 0.01168088,
      "auxiliary_loss_mlp": 0.01061738,
      "balance_loss_clip": 1.05670547,
      "balance_loss_mlp": 1.04115057,
      "epoch": 0.13912520667368106,
      "flos": 24535104111360.0,
      "grad_norm": 1.784990717237318,
      "language_loss": 0.79935932,
      "learning_rate": 3.876377616820024e-06,
      "loss": 0.8216576,
      "num_input_tokens_seen": 50110595,
      "step": 2314,
      "time_per_iteration": 2.683448076248169
    },
    {
      "auxiliary_loss_clip": 0.01124732,
      "auxiliary_loss_mlp": 0.01061041,
      "balance_loss_clip": 1.04845023,
      "balance_loss_mlp": 1.04103708,
      "epoch": 0.13918532992634902,
      "flos": 19383215287680.0,
      "grad_norm": 2.585875079553688,
      "language_loss": 0.85367405,
      "learning_rate": 3.876242779245409e-06,
      "loss": 0.87553179,
      "num_input_tokens_seen": 50125430,
      "step": 2315,
      "time_per_iteration": 4.332594394683838
    },
    {
      "auxiliary_loss_clip": 0.01156122,
      "auxiliary_loss_mlp": 0.01058532,
      "balance_loss_clip": 1.05397022,
      "balance_loss_mlp": 1.0372889,
      "epoch": 0.139245453179017,
      "flos": 21323756574720.0,
      "grad_norm": 2.333331492160627,
      "language_loss": 0.77170396,
      "learning_rate": 3.876107870523477e-06,
      "loss": 0.79385042,
      "num_input_tokens_seen": 50144120,
      "step": 2316,
      "time_per_iteration": 2.654604911804199
    },
    {
      "auxiliary_loss_clip": 0.01163967,
      "auxiliary_loss_mlp": 0.00780027,
      "balance_loss_clip": 1.05353916,
      "balance_loss_mlp": 1.00024533,
      "epoch": 0.13930557643168495,
      "flos": 19500607912320.0,
      "grad_norm": 2.1485284032262086,
      "language_loss": 0.76820493,
      "learning_rate": 3.875972890659349e-06,
      "loss": 0.78764486,
      "num_input_tokens_seen": 50162500,
      "step": 2317,
      "time_per_iteration": 2.6501235961914062
    },
    {
      "auxiliary_loss_clip": 0.01144052,
      "auxiliary_loss_mlp": 0.01061042,
      "balance_loss_clip": 1.05156648,
      "balance_loss_mlp": 1.04074025,
      "epoch": 0.13936569968435292,
      "flos": 25410821690880.0,
      "grad_norm": 1.7797832869421444,
      "language_loss": 0.80185997,
      "learning_rate": 3.875837839658139e-06,
      "loss": 0.82391089,
      "num_input_tokens_seen": 50182415,
      "step": 2318,
      "time_per_iteration": 2.7097995281219482
    },
    {
      "auxiliary_loss_clip": 0.01049096,
      "auxiliary_loss_mlp": 0.01048478,
      "balance_loss_clip": 1.03358936,
      "balance_loss_mlp": 1.04518783,
      "epoch": 0.13942582293702088,
      "flos": 70771063731840.0,
      "grad_norm": 0.854553938374386,
      "language_loss": 0.59004617,
      "learning_rate": 3.87570271752497e-06,
      "loss": 0.61102188,
      "num_input_tokens_seen": 50245160,
      "step": 2319,
      "time_per_iteration": 3.2631640434265137
    },
    {
      "auxiliary_loss_clip": 0.0111484,
      "auxiliary_loss_mlp": 0.01055367,
      "balance_loss_clip": 1.04508984,
      "balance_loss_mlp": 1.03437412,
      "epoch": 0.13948594618968888,
      "flos": 35590885920000.0,
      "grad_norm": 2.3313836691947722,
      "language_loss": 0.64993447,
      "learning_rate": 3.875567524264967e-06,
      "loss": 0.67163646,
      "num_input_tokens_seen": 50268215,
      "step": 2320,
      "time_per_iteration": 2.8668782711029053
    },
    {
      "auxiliary_loss_clip": 0.01096421,
      "auxiliary_loss_mlp": 0.01056652,
      "balance_loss_clip": 1.04400086,
      "balance_loss_mlp": 1.03521848,
      "epoch": 0.13954606944235684,
      "flos": 21105204272640.0,
      "grad_norm": 2.285151015895421,
      "language_loss": 0.70708811,
      "learning_rate": 3.875432259883256e-06,
      "loss": 0.72861886,
      "num_input_tokens_seen": 50288575,
      "step": 2321,
      "time_per_iteration": 2.8273603916168213
    },
    {
      "auxiliary_loss_clip": 0.01117698,
      "auxiliary_loss_mlp": 0.01061754,
      "balance_loss_clip": 1.04603076,
      "balance_loss_mlp": 1.03698206,
      "epoch": 0.1396061926950248,
      "flos": 25044425009280.0,
      "grad_norm": 1.7926270181208543,
      "language_loss": 0.85931206,
      "learning_rate": 3.875296924384965e-06,
      "loss": 0.88110662,
      "num_input_tokens_seen": 50308735,
      "step": 2322,
      "time_per_iteration": 2.833807945251465
    },
    {
      "auxiliary_loss_clip": 0.01120545,
      "auxiliary_loss_mlp": 0.01055036,
      "balance_loss_clip": 1.04616976,
      "balance_loss_mlp": 1.03568828,
      "epoch": 0.13966631594769277,
      "flos": 37634023428480.0,
      "grad_norm": 1.5963293576391182,
      "language_loss": 0.67159557,
      "learning_rate": 3.875161517775226e-06,
      "loss": 0.69335139,
      "num_input_tokens_seen": 50331025,
      "step": 2323,
      "time_per_iteration": 2.875265121459961
    },
    {
      "auxiliary_loss_clip": 0.01127992,
      "auxiliary_loss_mlp": 0.01055173,
      "balance_loss_clip": 1.04900301,
      "balance_loss_mlp": 1.03432369,
      "epoch": 0.13972643920036074,
      "flos": 16690993061760.0,
      "grad_norm": 2.0757452253793485,
      "language_loss": 0.88878977,
      "learning_rate": 3.875026040059175e-06,
      "loss": 0.9106214,
      "num_input_tokens_seen": 50349725,
      "step": 2324,
      "time_per_iteration": 2.6841063499450684
    },
    {
      "auxiliary_loss_clip": 0.01154799,
      "auxiliary_loss_mlp": 0.01056834,
      "balance_loss_clip": 1.05145955,
      "balance_loss_mlp": 1.03541231,
      "epoch": 0.1397865624530287,
      "flos": 23331055288320.0,
      "grad_norm": 2.8450589371660526,
      "language_loss": 0.70621002,
      "learning_rate": 3.8748904912419485e-06,
      "loss": 0.72832638,
      "num_input_tokens_seen": 50367965,
      "step": 2325,
      "time_per_iteration": 2.694218397140503
    },
    {
      "auxiliary_loss_clip": 0.01134393,
      "auxiliary_loss_mlp": 0.00778751,
      "balance_loss_clip": 1.05273592,
      "balance_loss_mlp": 1.00028229,
      "epoch": 0.13984668570569667,
      "flos": 22778317825920.0,
      "grad_norm": 2.230299294128946,
      "language_loss": 0.81657004,
      "learning_rate": 3.874754871328688e-06,
      "loss": 0.83570141,
      "num_input_tokens_seen": 50385605,
      "step": 2326,
      "time_per_iteration": 2.715306282043457
    },
    {
      "auxiliary_loss_clip": 0.01151297,
      "auxiliary_loss_mlp": 0.01045813,
      "balance_loss_clip": 1.05490732,
      "balance_loss_mlp": 1.02745473,
      "epoch": 0.13990680895836466,
      "flos": 19464553635840.0,
      "grad_norm": 1.729713540462037,
      "language_loss": 0.89241689,
      "learning_rate": 3.874619180324534e-06,
      "loss": 0.91438794,
      "num_input_tokens_seen": 50403985,
      "step": 2327,
      "time_per_iteration": 2.679626941680908
    },
    {
      "auxiliary_loss_clip": 0.01119996,
      "auxiliary_loss_mlp": 0.01057397,
      "balance_loss_clip": 1.04873121,
      "balance_loss_mlp": 1.0352242,
      "epoch": 0.13996693221103262,
      "flos": 20303283185280.0,
      "grad_norm": 2.9217951598838363,
      "language_loss": 0.84760427,
      "learning_rate": 3.874483418234632e-06,
      "loss": 0.86937821,
      "num_input_tokens_seen": 50421590,
      "step": 2328,
      "time_per_iteration": 2.7277352809906006
    },
    {
      "auxiliary_loss_clip": 0.01151775,
      "auxiliary_loss_mlp": 0.0104443,
      "balance_loss_clip": 1.05300856,
      "balance_loss_mlp": 1.02421176,
      "epoch": 0.1400270554637006,
      "flos": 26617707688320.0,
      "grad_norm": 1.6116398320348613,
      "language_loss": 0.73835862,
      "learning_rate": 3.874347585064131e-06,
      "loss": 0.76032066,
      "num_input_tokens_seen": 50443945,
      "step": 2329,
      "time_per_iteration": 2.6911025047302246
    },
    {
      "auxiliary_loss_clip": 0.01153137,
      "auxiliary_loss_mlp": 0.01046755,
      "balance_loss_clip": 1.05254042,
      "balance_loss_mlp": 1.02644169,
      "epoch": 0.14008717871636855,
      "flos": 19391475415680.0,
      "grad_norm": 2.565670250114109,
      "language_loss": 0.78373277,
      "learning_rate": 3.874211680818183e-06,
      "loss": 0.80573165,
      "num_input_tokens_seen": 50462065,
      "step": 2330,
      "time_per_iteration": 2.703225612640381
    },
    {
      "auxiliary_loss_clip": 0.01144455,
      "auxiliary_loss_mlp": 0.01046085,
      "balance_loss_clip": 1.05247569,
      "balance_loss_mlp": 1.02692819,
      "epoch": 0.14014730196903652,
      "flos": 15304266645120.0,
      "grad_norm": 2.2215524337864143,
      "language_loss": 0.72115719,
      "learning_rate": 3.87407570550194e-06,
      "loss": 0.74306256,
      "num_input_tokens_seen": 50479565,
      "step": 2331,
      "time_per_iteration": 2.7044217586517334
    },
    {
      "auxiliary_loss_clip": 0.01159691,
      "auxiliary_loss_mlp": 0.01051771,
      "balance_loss_clip": 1.0558939,
      "balance_loss_mlp": 1.03234017,
      "epoch": 0.14020742522170448,
      "flos": 14939701557120.0,
      "grad_norm": 1.5806705357110964,
      "language_loss": 0.72634697,
      "learning_rate": 3.873939659120557e-06,
      "loss": 0.7484616,
      "num_input_tokens_seen": 50497305,
      "step": 2332,
      "time_per_iteration": 2.647564649581909
    },
    {
      "auxiliary_loss_clip": 0.01063058,
      "auxiliary_loss_mlp": 0.01022564,
      "balance_loss_clip": 1.03391051,
      "balance_loss_mlp": 1.01944101,
      "epoch": 0.14026754847437245,
      "flos": 48824580044160.0,
      "grad_norm": 0.8445516092095569,
      "language_loss": 0.56185365,
      "learning_rate": 3.873803541679196e-06,
      "loss": 0.58270991,
      "num_input_tokens_seen": 50549735,
      "step": 2333,
      "time_per_iteration": 3.038390636444092
    },
    {
      "auxiliary_loss_clip": 0.01127793,
      "auxiliary_loss_mlp": 0.01045888,
      "balance_loss_clip": 1.05246043,
      "balance_loss_mlp": 1.02587318,
      "epoch": 0.14032767172704044,
      "flos": 25773267876480.0,
      "grad_norm": 1.7702774265545234,
      "language_loss": 0.82728767,
      "learning_rate": 3.873667353183016e-06,
      "loss": 0.84902453,
      "num_input_tokens_seen": 50570100,
      "step": 2334,
      "time_per_iteration": 2.7205803394317627
    },
    {
      "auxiliary_loss_clip": 0.01129244,
      "auxiliary_loss_mlp": 0.01044663,
      "balance_loss_clip": 1.05110407,
      "balance_loss_mlp": 1.02593565,
      "epoch": 0.1403877949797084,
      "flos": 21216312017280.0,
      "grad_norm": 1.7790720657464538,
      "language_loss": 0.80958998,
      "learning_rate": 3.8735310936371825e-06,
      "loss": 0.83132899,
      "num_input_tokens_seen": 50589185,
      "step": 2335,
      "time_per_iteration": 2.7844314575195312
    },
    {
      "auxiliary_loss_clip": 0.01108373,
      "auxiliary_loss_mlp": 0.0104374,
      "balance_loss_clip": 1.04802513,
      "balance_loss_mlp": 1.02160311,
      "epoch": 0.14044791823237637,
      "flos": 22747973811840.0,
      "grad_norm": 1.739505291070366,
      "language_loss": 0.81987065,
      "learning_rate": 3.873394763046862e-06,
      "loss": 0.84139174,
      "num_input_tokens_seen": 50609645,
      "step": 2336,
      "time_per_iteration": 2.7787351608276367
    },
    {
      "auxiliary_loss_clip": 0.01150445,
      "auxiliary_loss_mlp": 0.01046319,
      "balance_loss_clip": 1.05603921,
      "balance_loss_mlp": 1.02709103,
      "epoch": 0.14050804148504434,
      "flos": 22964443125120.0,
      "grad_norm": 1.7584048007565314,
      "language_loss": 0.80606967,
      "learning_rate": 3.873258361417225e-06,
      "loss": 0.82803738,
      "num_input_tokens_seen": 50628385,
      "step": 2337,
      "time_per_iteration": 2.6119275093078613
    },
    {
      "auxiliary_loss_clip": 0.01150898,
      "auxiliary_loss_mlp": 0.01051074,
      "balance_loss_clip": 1.05363941,
      "balance_loss_mlp": 1.03202438,
      "epoch": 0.1405681647377123,
      "flos": 22200336080640.0,
      "grad_norm": 2.383737065589604,
      "language_loss": 0.78994334,
      "learning_rate": 3.873121888753442e-06,
      "loss": 0.81196302,
      "num_input_tokens_seen": 50647260,
      "step": 2338,
      "time_per_iteration": 2.672427177429199
    },
    {
      "auxiliary_loss_clip": 0.01158377,
      "auxiliary_loss_mlp": 0.01050168,
      "balance_loss_clip": 1.05894089,
      "balance_loss_mlp": 1.02919865,
      "epoch": 0.14062828799038027,
      "flos": 23732787974400.0,
      "grad_norm": 2.117725014058833,
      "language_loss": 0.79766536,
      "learning_rate": 3.87298534506069e-06,
      "loss": 0.81975079,
      "num_input_tokens_seen": 50666130,
      "step": 2339,
      "time_per_iteration": 2.68635892868042
    },
    {
      "auxiliary_loss_clip": 0.01097095,
      "auxiliary_loss_mlp": 0.01065327,
      "balance_loss_clip": 1.04686952,
      "balance_loss_mlp": 1.04463232,
      "epoch": 0.14068841124304826,
      "flos": 39202493685120.0,
      "grad_norm": 2.0269377249156793,
      "language_loss": 0.65632963,
      "learning_rate": 3.872848730344146e-06,
      "loss": 0.67795384,
      "num_input_tokens_seen": 50687440,
      "step": 2340,
      "time_per_iteration": 2.9426286220550537
    },
    {
      "auxiliary_loss_clip": 0.0114865,
      "auxiliary_loss_mlp": 0.01050723,
      "balance_loss_clip": 1.05418086,
      "balance_loss_mlp": 1.0310297,
      "epoch": 0.14074853449571623,
      "flos": 20192283181440.0,
      "grad_norm": 2.8518792803213917,
      "language_loss": 0.78760445,
      "learning_rate": 3.87271204460899e-06,
      "loss": 0.80959821,
      "num_input_tokens_seen": 50704030,
      "step": 2341,
      "time_per_iteration": 2.8814899921417236
    },
    {
      "auxiliary_loss_clip": 0.01162758,
      "auxiliary_loss_mlp": 0.01057334,
      "balance_loss_clip": 1.0554986,
      "balance_loss_mlp": 1.03876162,
      "epoch": 0.1408086577483842,
      "flos": 18405871153920.0,
      "grad_norm": 2.2693198584224454,
      "language_loss": 0.80322361,
      "learning_rate": 3.8725752878604066e-06,
      "loss": 0.82542449,
      "num_input_tokens_seen": 50723305,
      "step": 2342,
      "time_per_iteration": 2.604814291000366
    },
    {
      "auxiliary_loss_clip": 0.01152048,
      "auxiliary_loss_mlp": 0.01056552,
      "balance_loss_clip": 1.05776191,
      "balance_loss_mlp": 1.03858757,
      "epoch": 0.14086878100105216,
      "flos": 25264593423360.0,
      "grad_norm": 2.4727499245104343,
      "language_loss": 0.77686632,
      "learning_rate": 3.87243846010358e-06,
      "loss": 0.79895234,
      "num_input_tokens_seen": 50743270,
      "step": 2343,
      "time_per_iteration": 2.676823854446411
    },
    {
      "auxiliary_loss_clip": 0.0105659,
      "auxiliary_loss_mlp": 0.01037584,
      "balance_loss_clip": 1.03650093,
      "balance_loss_mlp": 1.03438878,
      "epoch": 0.14092890425372012,
      "flos": 65978388869760.0,
      "grad_norm": 0.8521752699932517,
      "language_loss": 0.61553669,
      "learning_rate": 3.872301561343699e-06,
      "loss": 0.63647842,
      "num_input_tokens_seen": 50802710,
      "step": 2344,
      "time_per_iteration": 3.156792402267456
    },
    {
      "auxiliary_loss_clip": 0.01147637,
      "auxiliary_loss_mlp": 0.01049362,
      "balance_loss_clip": 1.05167484,
      "balance_loss_mlp": 1.03121877,
      "epoch": 0.1409890275063881,
      "flos": 23694973931520.0,
      "grad_norm": 1.558783678159347,
      "language_loss": 0.64331692,
      "learning_rate": 3.872164591585956e-06,
      "loss": 0.6652869,
      "num_input_tokens_seen": 50822625,
      "step": 2345,
      "time_per_iteration": 2.654100179672241
    },
    {
      "auxiliary_loss_clip": 0.01154879,
      "auxiliary_loss_mlp": 0.0104633,
      "balance_loss_clip": 1.05009735,
      "balance_loss_mlp": 1.02562308,
      "epoch": 0.14104915075905605,
      "flos": 23623152687360.0,
      "grad_norm": 2.26337760563351,
      "language_loss": 0.73892581,
      "learning_rate": 3.8720275508355435e-06,
      "loss": 0.76093793,
      "num_input_tokens_seen": 50842330,
      "step": 2346,
      "time_per_iteration": 2.7032830715179443
    },
    {
      "auxiliary_loss_clip": 0.0115447,
      "auxiliary_loss_mlp": 0.01048793,
      "balance_loss_clip": 1.0572027,
      "balance_loss_mlp": 1.02929008,
      "epoch": 0.14110927401172405,
      "flos": 20595165102720.0,
      "grad_norm": 1.7675181118684058,
      "language_loss": 0.7727294,
      "learning_rate": 3.8718904390976585e-06,
      "loss": 0.79476202,
      "num_input_tokens_seen": 50861035,
      "step": 2347,
      "time_per_iteration": 2.678647518157959
    },
    {
      "auxiliary_loss_clip": 0.01164131,
      "auxiliary_loss_mlp": 0.01052088,
      "balance_loss_clip": 1.05490732,
      "balance_loss_mlp": 1.03370619,
      "epoch": 0.141169397264392,
      "flos": 28548049512960.0,
      "grad_norm": 2.592464695784388,
      "language_loss": 0.76753062,
      "learning_rate": 3.8717532563775e-06,
      "loss": 0.78969282,
      "num_input_tokens_seen": 50880105,
      "step": 2348,
      "time_per_iteration": 2.7450597286224365
    },
    {
      "auxiliary_loss_clip": 0.01147264,
      "auxiliary_loss_mlp": 0.01042525,
      "balance_loss_clip": 1.05267334,
      "balance_loss_mlp": 1.02295136,
      "epoch": 0.14122952051705998,
      "flos": 17092258871040.0,
      "grad_norm": 1.8617784303344698,
      "language_loss": 0.86794335,
      "learning_rate": 3.871616002680272e-06,
      "loss": 0.8898412,
      "num_input_tokens_seen": 50897720,
      "step": 2349,
      "time_per_iteration": 2.662508964538574
    },
    {
      "auxiliary_loss_clip": 0.01150971,
      "auxiliary_loss_mlp": 0.01048616,
      "balance_loss_clip": 1.05632985,
      "balance_loss_mlp": 1.02897048,
      "epoch": 0.14128964376972794,
      "flos": 28946801370240.0,
      "grad_norm": 2.650060051711467,
      "language_loss": 0.88758218,
      "learning_rate": 3.871478678011177e-06,
      "loss": 0.90957808,
      "num_input_tokens_seen": 50918385,
      "step": 2350,
      "time_per_iteration": 4.1697962284088135
    },
    {
      "auxiliary_loss_clip": 0.01142704,
      "auxiliary_loss_mlp": 0.01045134,
      "balance_loss_clip": 1.05369377,
      "balance_loss_mlp": 1.02442729,
      "epoch": 0.1413497670223959,
      "flos": 18989778643200.0,
      "grad_norm": 1.801090232061166,
      "language_loss": 0.8094542,
      "learning_rate": 3.871341282375423e-06,
      "loss": 0.83133256,
      "num_input_tokens_seen": 50938270,
      "step": 2351,
      "time_per_iteration": 2.6769907474517822
    },
    {
      "auxiliary_loss_clip": 0.01149546,
      "auxiliary_loss_mlp": 0.01040141,
      "balance_loss_clip": 1.05100775,
      "balance_loss_mlp": 1.02096045,
      "epoch": 0.14140989027506387,
      "flos": 29862236413440.0,
      "grad_norm": 2.590933181784672,
      "language_loss": 0.82796198,
      "learning_rate": 3.871203815778219e-06,
      "loss": 0.84985888,
      "num_input_tokens_seen": 50958155,
      "step": 2352,
      "time_per_iteration": 5.713203430175781
    },
    {
      "auxiliary_loss_clip": 0.01063742,
      "auxiliary_loss_mlp": 0.01009803,
      "balance_loss_clip": 1.03462291,
      "balance_loss_mlp": 1.0060122,
      "epoch": 0.14147001352773186,
      "flos": 62079532041600.0,
      "grad_norm": 0.9118003008214054,
      "language_loss": 0.61876011,
      "learning_rate": 3.87106627822478e-06,
      "loss": 0.63949555,
      "num_input_tokens_seen": 51020705,
      "step": 2353,
      "time_per_iteration": 3.1698319911956787
    },
    {
      "auxiliary_loss_clip": 0.01134069,
      "auxiliary_loss_mlp": 0.01049094,
      "balance_loss_clip": 1.0536828,
      "balance_loss_mlp": 1.03039002,
      "epoch": 0.14153013678039983,
      "flos": 22017514832640.0,
      "grad_norm": 1.5909284402791886,
      "language_loss": 0.87075388,
      "learning_rate": 3.8709286697203196e-06,
      "loss": 0.89258552,
      "num_input_tokens_seen": 51039995,
      "step": 2354,
      "time_per_iteration": 2.6781272888183594
    },
    {
      "auxiliary_loss_clip": 0.01124592,
      "auxiliary_loss_mlp": 0.0104583,
      "balance_loss_clip": 1.0527302,
      "balance_loss_mlp": 1.02562428,
      "epoch": 0.1415902600330678,
      "flos": 19720093968000.0,
      "grad_norm": 2.035812967878614,
      "language_loss": 0.74701214,
      "learning_rate": 3.870790990270057e-06,
      "loss": 0.76871634,
      "num_input_tokens_seen": 51059075,
      "step": 2355,
      "time_per_iteration": 4.464852571487427
    },
    {
      "auxiliary_loss_clip": 0.01062228,
      "auxiliary_loss_mlp": 0.01003337,
      "balance_loss_clip": 1.03320074,
      "balance_loss_mlp": 0.99947417,
      "epoch": 0.14165038328573576,
      "flos": 65900929190400.0,
      "grad_norm": 0.6801443738216844,
      "language_loss": 0.51819825,
      "learning_rate": 3.870653239879212e-06,
      "loss": 0.53885388,
      "num_input_tokens_seen": 51120380,
      "step": 2356,
      "time_per_iteration": 3.094026803970337
    },
    {
      "auxiliary_loss_clip": 0.01165635,
      "auxiliary_loss_mlp": 0.01057535,
      "balance_loss_clip": 1.05662966,
      "balance_loss_mlp": 1.0379492,
      "epoch": 0.14171050653840372,
      "flos": 12130158533760.0,
      "grad_norm": 1.9928903491175036,
      "language_loss": 0.70598352,
      "learning_rate": 3.8705154185530095e-06,
      "loss": 0.72821522,
      "num_input_tokens_seen": 51136950,
      "step": 2357,
      "time_per_iteration": 2.569486141204834
    },
    {
      "auxiliary_loss_clip": 0.01117022,
      "auxiliary_loss_mlp": 0.01054948,
      "balance_loss_clip": 1.04706419,
      "balance_loss_mlp": 1.0355413,
      "epoch": 0.1417706297910717,
      "flos": 20412487509120.0,
      "grad_norm": 2.1046358800035234,
      "language_loss": 0.82020235,
      "learning_rate": 3.870377526296674e-06,
      "loss": 0.84192204,
      "num_input_tokens_seen": 51155175,
      "step": 2358,
      "time_per_iteration": 2.719344139099121
    },
    {
      "auxiliary_loss_clip": 0.01145283,
      "auxiliary_loss_mlp": 0.01050239,
      "balance_loss_clip": 1.05257189,
      "balance_loss_mlp": 1.02932954,
      "epoch": 0.14183075304373965,
      "flos": 22380607463040.0,
      "grad_norm": 2.2336131404929787,
      "language_loss": 0.71575904,
      "learning_rate": 3.870239563115436e-06,
      "loss": 0.73771417,
      "num_input_tokens_seen": 51174500,
      "step": 2359,
      "time_per_iteration": 2.6914820671081543
    },
    {
      "auxiliary_loss_clip": 0.0111529,
      "auxiliary_loss_mlp": 0.007787,
      "balance_loss_clip": 1.0526464,
      "balance_loss_mlp": 1.00033379,
      "epoch": 0.14189087629640765,
      "flos": 21580913018880.0,
      "grad_norm": 2.4314273775499906,
      "language_loss": 0.7541784,
      "learning_rate": 3.870101529014526e-06,
      "loss": 0.77311832,
      "num_input_tokens_seen": 51194270,
      "step": 2360,
      "time_per_iteration": 2.803493022918701
    },
    {
      "auxiliary_loss_clip": 0.01108644,
      "auxiliary_loss_mlp": 0.01053684,
      "balance_loss_clip": 1.0491271,
      "balance_loss_mlp": 1.03136814,
      "epoch": 0.1419509995490756,
      "flos": 20008564093440.0,
      "grad_norm": 2.374719540518049,
      "language_loss": 0.81920552,
      "learning_rate": 3.869963423999178e-06,
      "loss": 0.84082878,
      "num_input_tokens_seen": 51211850,
      "step": 2361,
      "time_per_iteration": 2.8039920330047607
    },
    {
      "auxiliary_loss_clip": 0.0115065,
      "auxiliary_loss_mlp": 0.01057946,
      "balance_loss_clip": 1.05230403,
      "balance_loss_mlp": 1.03802609,
      "epoch": 0.14201112280174358,
      "flos": 31941464112000.0,
      "grad_norm": 1.9397979109407166,
      "language_loss": 0.74081504,
      "learning_rate": 3.86982524807463e-06,
      "loss": 0.76290095,
      "num_input_tokens_seen": 51233545,
      "step": 2362,
      "time_per_iteration": 2.7272114753723145
    },
    {
      "auxiliary_loss_clip": 0.0115354,
      "auxiliary_loss_mlp": 0.01048321,
      "balance_loss_clip": 1.05355787,
      "balance_loss_mlp": 1.02861547,
      "epoch": 0.14207124605441154,
      "flos": 41464147582080.0,
      "grad_norm": 1.7489521991344694,
      "language_loss": 0.74221587,
      "learning_rate": 3.869687001246122e-06,
      "loss": 0.76423442,
      "num_input_tokens_seen": 51257615,
      "step": 2363,
      "time_per_iteration": 2.789802312850952
    },
    {
      "auxiliary_loss_clip": 0.01128802,
      "auxiliary_loss_mlp": 0.0105205,
      "balance_loss_clip": 1.04769099,
      "balance_loss_mlp": 1.03180885,
      "epoch": 0.1421313693070795,
      "flos": 31905086613120.0,
      "grad_norm": 1.7832713632097879,
      "language_loss": 0.73034167,
      "learning_rate": 3.8695486835188946e-06,
      "loss": 0.75215018,
      "num_input_tokens_seen": 51279645,
      "step": 2364,
      "time_per_iteration": 2.8508312702178955
    },
    {
      "auxiliary_loss_clip": 0.01142769,
      "auxiliary_loss_mlp": 0.01049829,
      "balance_loss_clip": 1.05160844,
      "balance_loss_mlp": 1.03207827,
      "epoch": 0.14219149255974747,
      "flos": 26871165031680.0,
      "grad_norm": 1.875477198706701,
      "language_loss": 0.90395916,
      "learning_rate": 3.869410294898195e-06,
      "loss": 0.92588514,
      "num_input_tokens_seen": 51299775,
      "step": 2365,
      "time_per_iteration": 2.6807806491851807
    },
    {
      "auxiliary_loss_clip": 0.01127252,
      "auxiliary_loss_mlp": 0.01054912,
      "balance_loss_clip": 1.04759967,
      "balance_loss_mlp": 1.03394318,
      "epoch": 0.14225161581241544,
      "flos": 27454426076160.0,
      "grad_norm": 1.719218863067841,
      "language_loss": 0.65305161,
      "learning_rate": 3.869271835389268e-06,
      "loss": 0.67487329,
      "num_input_tokens_seen": 51319430,
      "step": 2366,
      "time_per_iteration": 2.7293641567230225
    },
    {
      "auxiliary_loss_clip": 0.01143576,
      "auxiliary_loss_mlp": 0.01051629,
      "balance_loss_clip": 1.05218709,
      "balance_loss_mlp": 1.03058839,
      "epoch": 0.14231173906508343,
      "flos": 10561436881920.0,
      "grad_norm": 2.3740196514966256,
      "language_loss": 0.80331928,
      "learning_rate": 3.8691333049973665e-06,
      "loss": 0.82527137,
      "num_input_tokens_seen": 51336045,
      "step": 2367,
      "time_per_iteration": 2.67529296875
    },
    {
      "auxiliary_loss_clip": 0.01138517,
      "auxiliary_loss_mlp": 0.01062653,
      "balance_loss_clip": 1.05117869,
      "balance_loss_mlp": 1.0402534,
      "epoch": 0.1423718623177514,
      "flos": 28360882719360.0,
      "grad_norm": 2.0081973718426283,
      "language_loss": 0.82346755,
      "learning_rate": 3.868994703727742e-06,
      "loss": 0.84547925,
      "num_input_tokens_seen": 51357030,
      "step": 2368,
      "time_per_iteration": 2.7447288036346436
    },
    {
      "auxiliary_loss_clip": 0.01122755,
      "auxiliary_loss_mlp": 0.01052229,
      "balance_loss_clip": 1.05180073,
      "balance_loss_mlp": 1.03065228,
      "epoch": 0.14243198557041936,
      "flos": 19354235990400.0,
      "grad_norm": 2.6586279461428144,
      "language_loss": 0.8711772,
      "learning_rate": 3.868856031585652e-06,
      "loss": 0.89292705,
      "num_input_tokens_seen": 51374890,
      "step": 2369,
      "time_per_iteration": 2.736872673034668
    },
    {
      "auxiliary_loss_clip": 0.01127301,
      "auxiliary_loss_mlp": 0.0104182,
      "balance_loss_clip": 1.05011857,
      "balance_loss_mlp": 1.02170992,
      "epoch": 0.14249210882308733,
      "flos": 28806857982720.0,
      "grad_norm": 1.7900856007188275,
      "language_loss": 0.75828248,
      "learning_rate": 3.868717288576354e-06,
      "loss": 0.77997375,
      "num_input_tokens_seen": 51398100,
      "step": 2370,
      "time_per_iteration": 2.762603998184204
    },
    {
      "auxiliary_loss_clip": 0.01158195,
      "auxiliary_loss_mlp": 0.00781098,
      "balance_loss_clip": 1.05268764,
      "balance_loss_mlp": 1.00028419,
      "epoch": 0.1425522320757553,
      "flos": 21835016807040.0,
      "grad_norm": 1.7770434161065212,
      "language_loss": 0.82934797,
      "learning_rate": 3.868578474705109e-06,
      "loss": 0.84874088,
      "num_input_tokens_seen": 51418745,
      "step": 2371,
      "time_per_iteration": 2.6224656105041504
    },
    {
      "auxiliary_loss_clip": 0.01173447,
      "auxiliary_loss_mlp": 0.0105718,
      "balance_loss_clip": 1.05837953,
      "balance_loss_mlp": 1.03638947,
      "epoch": 0.14261235532842326,
      "flos": 17311457617920.0,
      "grad_norm": 2.0431625041319825,
      "language_loss": 0.82982123,
      "learning_rate": 3.868439589977181e-06,
      "loss": 0.85212755,
      "num_input_tokens_seen": 51437455,
      "step": 2372,
      "time_per_iteration": 2.575690269470215
    },
    {
      "auxiliary_loss_clip": 0.01172196,
      "auxiliary_loss_mlp": 0.0105022,
      "balance_loss_clip": 1.0581125,
      "balance_loss_mlp": 1.0285356,
      "epoch": 0.14267247858109125,
      "flos": 18806741913600.0,
      "grad_norm": 3.3704326167450582,
      "language_loss": 0.8438468,
      "learning_rate": 3.868300634397836e-06,
      "loss": 0.86607099,
      "num_input_tokens_seen": 51455710,
      "step": 2373,
      "time_per_iteration": 2.7160356044769287
    },
    {
      "auxiliary_loss_clip": 0.01141742,
      "auxiliary_loss_mlp": 0.01055295,
      "balance_loss_clip": 1.05160809,
      "balance_loss_mlp": 1.03598261,
      "epoch": 0.14273260183375922,
      "flos": 11358904682880.0,
      "grad_norm": 3.5035356392631836,
      "language_loss": 0.86027539,
      "learning_rate": 3.8681616079723445e-06,
      "loss": 0.88224572,
      "num_input_tokens_seen": 51471270,
      "step": 2374,
      "time_per_iteration": 2.6845595836639404
    },
    {
      "auxiliary_loss_clip": 0.01164623,
      "auxiliary_loss_mlp": 0.01061957,
      "balance_loss_clip": 1.05515146,
      "balance_loss_mlp": 1.03996301,
      "epoch": 0.14279272508642718,
      "flos": 27567688636800.0,
      "grad_norm": 1.6059368749673757,
      "language_loss": 0.79169822,
      "learning_rate": 3.868022510705977e-06,
      "loss": 0.81396401,
      "num_input_tokens_seen": 51492705,
      "step": 2375,
      "time_per_iteration": 2.738156795501709
    },
    {
      "auxiliary_loss_clip": 0.01163115,
      "auxiliary_loss_mlp": 0.01058224,
      "balance_loss_clip": 1.05641222,
      "balance_loss_mlp": 1.0368259,
      "epoch": 0.14285284833909515,
      "flos": 16252559654400.0,
      "grad_norm": 2.559097553272684,
      "language_loss": 0.76907504,
      "learning_rate": 3.867883342604009e-06,
      "loss": 0.79128844,
      "num_input_tokens_seen": 51510780,
      "step": 2376,
      "time_per_iteration": 2.751178741455078
    },
    {
      "auxiliary_loss_clip": 0.01160115,
      "auxiliary_loss_mlp": 0.0105168,
      "balance_loss_clip": 1.054515,
      "balance_loss_mlp": 1.03040111,
      "epoch": 0.1429129715917631,
      "flos": 19755609540480.0,
      "grad_norm": 2.7331999261828592,
      "language_loss": 0.92795181,
      "learning_rate": 3.867744103671717e-06,
      "loss": 0.95006979,
      "num_input_tokens_seen": 51531400,
      "step": 2377,
      "time_per_iteration": 2.6584725379943848
    },
    {
      "auxiliary_loss_clip": 0.01147246,
      "auxiliary_loss_mlp": 0.01061419,
      "balance_loss_clip": 1.05362535,
      "balance_loss_mlp": 1.03793442,
      "epoch": 0.14297309484443108,
      "flos": 21137092571520.0,
      "grad_norm": 2.9252003733204894,
      "language_loss": 0.91754365,
      "learning_rate": 3.867604793914382e-06,
      "loss": 0.93963027,
      "num_input_tokens_seen": 51548215,
      "step": 2378,
      "time_per_iteration": 2.8107075691223145
    },
    {
      "auxiliary_loss_clip": 0.01164153,
      "auxiliary_loss_mlp": 0.0105303,
      "balance_loss_clip": 1.05712187,
      "balance_loss_mlp": 1.03092849,
      "epoch": 0.14303321809709904,
      "flos": 23586667447680.0,
      "grad_norm": 2.1292902842232966,
      "language_loss": 0.73961306,
      "learning_rate": 3.8674654133372864e-06,
      "loss": 0.76178491,
      "num_input_tokens_seen": 51566820,
      "step": 2379,
      "time_per_iteration": 2.7029881477355957
    },
    {
      "auxiliary_loss_clip": 0.01137551,
      "auxiliary_loss_mlp": 0.01055012,
      "balance_loss_clip": 1.05204058,
      "balance_loss_mlp": 1.0330174,
      "epoch": 0.14309334134976703,
      "flos": 15888281875200.0,
      "grad_norm": 2.1898245228218784,
      "language_loss": 0.78818595,
      "learning_rate": 3.867325961945714e-06,
      "loss": 0.81011152,
      "num_input_tokens_seen": 51585075,
      "step": 2380,
      "time_per_iteration": 2.7213294506073
    },
    {
      "auxiliary_loss_clip": 0.01126442,
      "auxiliary_loss_mlp": 0.01057409,
      "balance_loss_clip": 1.05457354,
      "balance_loss_mlp": 1.03580785,
      "epoch": 0.143153464602435,
      "flos": 16325601960960.0,
      "grad_norm": 4.699041640805274,
      "language_loss": 0.87895483,
      "learning_rate": 3.867186439744955e-06,
      "loss": 0.90079331,
      "num_input_tokens_seen": 51603185,
      "step": 2381,
      "time_per_iteration": 2.7144110202789307
    },
    {
      "auxiliary_loss_clip": 0.01141327,
      "auxiliary_loss_mlp": 0.01052708,
      "balance_loss_clip": 1.05200005,
      "balance_loss_mlp": 1.03088117,
      "epoch": 0.14321358785510296,
      "flos": 17092079303040.0,
      "grad_norm": 2.47508592106904,
      "language_loss": 0.76396096,
      "learning_rate": 3.867046846740299e-06,
      "loss": 0.78590137,
      "num_input_tokens_seen": 51620880,
      "step": 2382,
      "time_per_iteration": 2.6185953617095947
    },
    {
      "auxiliary_loss_clip": 0.01132222,
      "auxiliary_loss_mlp": 0.01054019,
      "balance_loss_clip": 1.05162048,
      "balance_loss_mlp": 1.03319359,
      "epoch": 0.14327371110777093,
      "flos": 26322916769280.0,
      "grad_norm": 4.3017095308344375,
      "language_loss": 0.76636785,
      "learning_rate": 3.866907182937039e-06,
      "loss": 0.7882303,
      "num_input_tokens_seen": 51640170,
      "step": 2383,
      "time_per_iteration": 2.7408525943756104
    },
    {
      "auxiliary_loss_clip": 0.01139698,
      "auxiliary_loss_mlp": 0.01052888,
      "balance_loss_clip": 1.05078864,
      "balance_loss_mlp": 1.02926064,
      "epoch": 0.1433338343604389,
      "flos": 18076462502400.0,
      "grad_norm": 2.3526544982502284,
      "language_loss": 0.87649417,
      "learning_rate": 3.866767448340471e-06,
      "loss": 0.8984201,
      "num_input_tokens_seen": 51656580,
      "step": 2384,
      "time_per_iteration": 2.6798789501190186
    },
    {
      "auxiliary_loss_clip": 0.01164805,
      "auxiliary_loss_mlp": 0.01053206,
      "balance_loss_clip": 1.05644679,
      "balance_loss_mlp": 1.02985239,
      "epoch": 0.14339395761310686,
      "flos": 15522783033600.0,
      "grad_norm": 2.6134761315069284,
      "language_loss": 0.79340684,
      "learning_rate": 3.866627642955895e-06,
      "loss": 0.81558692,
      "num_input_tokens_seen": 51674645,
      "step": 2385,
      "time_per_iteration": 2.5856544971466064
    },
    {
      "auxiliary_loss_clip": 0.01156607,
      "auxiliary_loss_mlp": 0.01042784,
      "balance_loss_clip": 1.05148256,
      "balance_loss_mlp": 1.02182722,
      "epoch": 0.14345408086577485,
      "flos": 28548767784960.0,
      "grad_norm": 2.6990187663653247,
      "language_loss": 0.74960196,
      "learning_rate": 3.866487766788612e-06,
      "loss": 0.77159584,
      "num_input_tokens_seen": 51695770,
      "step": 2386,
      "time_per_iteration": 2.6670751571655273
    },
    {
      "auxiliary_loss_clip": 0.01171639,
      "auxiliary_loss_mlp": 0.01048096,
      "balance_loss_clip": 1.05699563,
      "balance_loss_mlp": 1.02733016,
      "epoch": 0.14351420411844282,
      "flos": 20230061310720.0,
      "grad_norm": 2.299870083842227,
      "language_loss": 0.78659731,
      "learning_rate": 3.866347819843925e-06,
      "loss": 0.80879462,
      "num_input_tokens_seen": 51714165,
      "step": 2387,
      "time_per_iteration": 2.5805532932281494
    },
    {
      "auxiliary_loss_clip": 0.01140581,
      "auxiliary_loss_mlp": 0.01055299,
      "balance_loss_clip": 1.05355716,
      "balance_loss_mlp": 1.03317428,
      "epoch": 0.14357432737111078,
      "flos": 19865029345920.0,
      "grad_norm": 6.554164509194222,
      "language_loss": 0.82492924,
      "learning_rate": 3.866207802127143e-06,
      "loss": 0.84688807,
      "num_input_tokens_seen": 51734440,
      "step": 2388,
      "time_per_iteration": 2.656609058380127
    },
    {
      "auxiliary_loss_clip": 0.01155007,
      "auxiliary_loss_mlp": 0.01047154,
      "balance_loss_clip": 1.0537287,
      "balance_loss_mlp": 1.02674508,
      "epoch": 0.14363445062377875,
      "flos": 28256814040320.0,
      "grad_norm": 2.5973624291758655,
      "language_loss": 0.82025754,
      "learning_rate": 3.866067713643573e-06,
      "loss": 0.84227914,
      "num_input_tokens_seen": 51753730,
      "step": 2389,
      "time_per_iteration": 4.21793794631958
    },
    {
      "auxiliary_loss_clip": 0.01145665,
      "auxiliary_loss_mlp": 0.01046852,
      "balance_loss_clip": 1.05107975,
      "balance_loss_mlp": 1.02513266,
      "epoch": 0.1436945738764467,
      "flos": 18186672407040.0,
      "grad_norm": 3.7970835440683097,
      "language_loss": 0.83056784,
      "learning_rate": 3.8659275543985285e-06,
      "loss": 0.85249299,
      "num_input_tokens_seen": 51771195,
      "step": 2390,
      "time_per_iteration": 2.6859514713287354
    },
    {
      "auxiliary_loss_clip": 0.01152608,
      "auxiliary_loss_mlp": 0.01054404,
      "balance_loss_clip": 1.05400729,
      "balance_loss_mlp": 1.0334475,
      "epoch": 0.14375469712911468,
      "flos": 27307910499840.0,
      "grad_norm": 1.8176612067028404,
      "language_loss": 0.75018179,
      "learning_rate": 3.865787324397324e-06,
      "loss": 0.77225184,
      "num_input_tokens_seen": 51792290,
      "step": 2391,
      "time_per_iteration": 5.726900577545166
    },
    {
      "auxiliary_loss_clip": 0.01045505,
      "auxiliary_loss_mlp": 0.01033342,
      "balance_loss_clip": 1.03226101,
      "balance_loss_mlp": 1.0303973,
      "epoch": 0.14381482038178264,
      "flos": 56891445287040.0,
      "grad_norm": 0.8787809928903102,
      "language_loss": 0.61848003,
      "learning_rate": 3.865647023645277e-06,
      "loss": 0.63926852,
      "num_input_tokens_seen": 51843675,
      "step": 2392,
      "time_per_iteration": 3.113558053970337
    },
    {
      "auxiliary_loss_clip": 0.01158698,
      "auxiliary_loss_mlp": 0.01058807,
      "balance_loss_clip": 1.05467868,
      "balance_loss_mlp": 1.03608608,
      "epoch": 0.14387494363445064,
      "flos": 14282177143680.0,
      "grad_norm": 2.718376715006273,
      "language_loss": 0.77346605,
      "learning_rate": 3.865506652147709e-06,
      "loss": 0.79564106,
      "num_input_tokens_seen": 51860285,
      "step": 2393,
      "time_per_iteration": 2.6578521728515625
    },
    {
      "auxiliary_loss_clip": 0.0116951,
      "auxiliary_loss_mlp": 0.01052986,
      "balance_loss_clip": 1.05671048,
      "balance_loss_mlp": 1.03287578,
      "epoch": 0.1439350668871186,
      "flos": 26761493831040.0,
      "grad_norm": 5.715284956255472,
      "language_loss": 0.76301813,
      "learning_rate": 3.865366209909941e-06,
      "loss": 0.78524309,
      "num_input_tokens_seen": 51880105,
      "step": 2394,
      "time_per_iteration": 4.345217943191528
    },
    {
      "auxiliary_loss_clip": 0.01165266,
      "auxiliary_loss_mlp": 0.01053501,
      "balance_loss_clip": 1.05325842,
      "balance_loss_mlp": 1.03365326,
      "epoch": 0.14399519013978657,
      "flos": 40700040537600.0,
      "grad_norm": 2.2496244390836893,
      "language_loss": 0.85859704,
      "learning_rate": 3.8652256969372994e-06,
      "loss": 0.88078463,
      "num_input_tokens_seen": 51905175,
      "step": 2395,
      "time_per_iteration": 2.739717483520508
    },
    {
      "auxiliary_loss_clip": 0.0112523,
      "auxiliary_loss_mlp": 0.01051092,
      "balance_loss_clip": 1.04946184,
      "balance_loss_mlp": 1.028669,
      "epoch": 0.14405531339245453,
      "flos": 20557530627840.0,
      "grad_norm": 4.117082508421602,
      "language_loss": 0.82894099,
      "learning_rate": 3.865085113235113e-06,
      "loss": 0.85070425,
      "num_input_tokens_seen": 51924490,
      "step": 2396,
      "time_per_iteration": 2.686732053756714
    },
    {
      "auxiliary_loss_clip": 0.01126754,
      "auxiliary_loss_mlp": 0.00779833,
      "balance_loss_clip": 1.04752374,
      "balance_loss_mlp": 1.00036597,
      "epoch": 0.1441154366451225,
      "flos": 19572931946880.0,
      "grad_norm": 6.956399779275871,
      "language_loss": 0.82801461,
      "learning_rate": 3.864944458808712e-06,
      "loss": 0.84708053,
      "num_input_tokens_seen": 51940490,
      "step": 2397,
      "time_per_iteration": 2.742809534072876
    },
    {
      "auxiliary_loss_clip": 0.01168871,
      "auxiliary_loss_mlp": 0.0104994,
      "balance_loss_clip": 1.05485702,
      "balance_loss_mlp": 1.02892387,
      "epoch": 0.14417555989779046,
      "flos": 18515721922560.0,
      "grad_norm": 8.355198005975433,
      "language_loss": 0.8001197,
      "learning_rate": 3.86480373366343e-06,
      "loss": 0.82230783,
      "num_input_tokens_seen": 51957910,
      "step": 2398,
      "time_per_iteration": 2.573267936706543
    },
    {
      "auxiliary_loss_clip": 0.01152449,
      "auxiliary_loss_mlp": 0.01053407,
      "balance_loss_clip": 1.05287588,
      "balance_loss_mlp": 1.03336823,
      "epoch": 0.14423568315045843,
      "flos": 26031681296640.0,
      "grad_norm": 3.294581575970509,
      "language_loss": 0.64690518,
      "learning_rate": 3.864662937804603e-06,
      "loss": 0.66896379,
      "num_input_tokens_seen": 51978010,
      "step": 2399,
      "time_per_iteration": 2.6831774711608887
    },
    {
      "auxiliary_loss_clip": 0.01134916,
      "auxiliary_loss_mlp": 0.01052493,
      "balance_loss_clip": 1.04998159,
      "balance_loss_mlp": 1.03119016,
      "epoch": 0.14429580640312642,
      "flos": 21288743792640.0,
      "grad_norm": 3.586256880371596,
      "language_loss": 0.82207137,
      "learning_rate": 3.864522071237571e-06,
      "loss": 0.84394544,
      "num_input_tokens_seen": 51998515,
      "step": 2400,
      "time_per_iteration": 2.6812663078308105
    },
    {
      "auxiliary_loss_clip": 0.01149983,
      "auxiliary_loss_mlp": 0.01051884,
      "balance_loss_clip": 1.0567503,
      "balance_loss_mlp": 1.02954376,
      "epoch": 0.14435592965579438,
      "flos": 25627865621760.0,
      "grad_norm": 2.3908005596579165,
      "language_loss": 0.74217784,
      "learning_rate": 3.864381133967676e-06,
      "loss": 0.76419652,
      "num_input_tokens_seen": 52019270,
      "step": 2401,
      "time_per_iteration": 2.773838520050049
    },
    {
      "auxiliary_loss_clip": 0.01137207,
      "auxiliary_loss_mlp": 0.01047592,
      "balance_loss_clip": 1.05065656,
      "balance_loss_mlp": 1.02671885,
      "epoch": 0.14441605290846235,
      "flos": 22965053656320.0,
      "grad_norm": 2.616063077702737,
      "language_loss": 0.80771816,
      "learning_rate": 3.86424012600026e-06,
      "loss": 0.82956612,
      "num_input_tokens_seen": 52039315,
      "step": 2402,
      "time_per_iteration": 2.786031723022461
    },
    {
      "auxiliary_loss_clip": 0.01120897,
      "auxiliary_loss_mlp": 0.01052115,
      "balance_loss_clip": 1.04718328,
      "balance_loss_mlp": 1.02988231,
      "epoch": 0.14447617616113032,
      "flos": 17347655548800.0,
      "grad_norm": 2.397935571801219,
      "language_loss": 0.84159613,
      "learning_rate": 3.864099047340673e-06,
      "loss": 0.86332625,
      "num_input_tokens_seen": 52056555,
      "step": 2403,
      "time_per_iteration": 2.8113911151885986
    },
    {
      "auxiliary_loss_clip": 0.01129082,
      "auxiliary_loss_mlp": 0.00783127,
      "balance_loss_clip": 1.04854488,
      "balance_loss_mlp": 1.00030184,
      "epoch": 0.14453629941379828,
      "flos": 24060185464320.0,
      "grad_norm": 2.224282169770823,
      "language_loss": 0.70142806,
      "learning_rate": 3.863957897994262e-06,
      "loss": 0.72055018,
      "num_input_tokens_seen": 52075800,
      "step": 2404,
      "time_per_iteration": 2.7748003005981445
    },
    {
      "auxiliary_loss_clip": 0.01144289,
      "auxiliary_loss_mlp": 0.01051404,
      "balance_loss_clip": 1.05279732,
      "balance_loss_mlp": 1.03099549,
      "epoch": 0.14459642266646625,
      "flos": 14429554646400.0,
      "grad_norm": 2.429117427076043,
      "language_loss": 0.73179376,
      "learning_rate": 3.863816677966381e-06,
      "loss": 0.75375068,
      "num_input_tokens_seen": 52092585,
      "step": 2405,
      "time_per_iteration": 2.7927868366241455
    },
    {
      "auxiliary_loss_clip": 0.01108387,
      "auxiliary_loss_mlp": 0.01054584,
      "balance_loss_clip": 1.04661417,
      "balance_loss_mlp": 1.0326612,
      "epoch": 0.14465654591913424,
      "flos": 9867032179200.0,
      "grad_norm": 7.089523066959408,
      "language_loss": 0.73039794,
      "learning_rate": 3.863675387262386e-06,
      "loss": 0.75202763,
      "num_input_tokens_seen": 52108990,
      "step": 2406,
      "time_per_iteration": 2.742253303527832
    },
    {
      "auxiliary_loss_clip": 0.01157268,
      "auxiliary_loss_mlp": 0.01054465,
      "balance_loss_clip": 1.05420268,
      "balance_loss_mlp": 1.03198171,
      "epoch": 0.1447166691718022,
      "flos": 24972926987520.0,
      "grad_norm": 5.383630788916188,
      "language_loss": 0.75570732,
      "learning_rate": 3.8635340258876325e-06,
      "loss": 0.77782464,
      "num_input_tokens_seen": 52125385,
      "step": 2407,
      "time_per_iteration": 2.654636859893799
    },
    {
      "auxiliary_loss_clip": 0.0116674,
      "auxiliary_loss_mlp": 0.01054642,
      "balance_loss_clip": 1.05440819,
      "balance_loss_mlp": 1.03392315,
      "epoch": 0.14477679242447017,
      "flos": 21908023200000.0,
      "grad_norm": 2.0240540465866146,
      "language_loss": 0.79426706,
      "learning_rate": 3.8633925938474826e-06,
      "loss": 0.81648088,
      "num_input_tokens_seen": 52144985,
      "step": 2408,
      "time_per_iteration": 2.663611650466919
    },
    {
      "auxiliary_loss_clip": 0.01155332,
      "auxiliary_loss_mlp": 0.01053557,
      "balance_loss_clip": 1.05411625,
      "balance_loss_mlp": 1.03107429,
      "epoch": 0.14483691567713813,
      "flos": 20740746925440.0,
      "grad_norm": 2.249858190268702,
      "language_loss": 0.82188261,
      "learning_rate": 3.863251091147299e-06,
      "loss": 0.84397143,
      "num_input_tokens_seen": 52163885,
      "step": 2409,
      "time_per_iteration": 2.6218342781066895
    },
    {
      "auxiliary_loss_clip": 0.01116852,
      "auxiliary_loss_mlp": 0.01065498,
      "balance_loss_clip": 1.04859877,
      "balance_loss_mlp": 1.04340839,
      "epoch": 0.1448970389298061,
      "flos": 35407705536000.0,
      "grad_norm": 3.918408886138166,
      "language_loss": 0.74477464,
      "learning_rate": 3.863109517792446e-06,
      "loss": 0.76659817,
      "num_input_tokens_seen": 52184325,
      "step": 2410,
      "time_per_iteration": 2.8525002002716064
    },
    {
      "auxiliary_loss_clip": 0.01166422,
      "auxiliary_loss_mlp": 0.0105028,
      "balance_loss_clip": 1.05447876,
      "balance_loss_mlp": 1.0300622,
      "epoch": 0.14495716218247406,
      "flos": 15414368808960.0,
      "grad_norm": 2.976325973684052,
      "language_loss": 0.81616414,
      "learning_rate": 3.8629678737882945e-06,
      "loss": 0.8383311,
      "num_input_tokens_seen": 52202740,
      "step": 2411,
      "time_per_iteration": 2.580059051513672
    },
    {
      "auxiliary_loss_clip": 0.01143671,
      "auxiliary_loss_mlp": 0.01055066,
      "balance_loss_clip": 1.05553794,
      "balance_loss_mlp": 1.03366852,
      "epoch": 0.14501728543514203,
      "flos": 33693222493440.0,
      "grad_norm": 2.049708152728223,
      "language_loss": 0.69947547,
      "learning_rate": 3.862826159140214e-06,
      "loss": 0.72146285,
      "num_input_tokens_seen": 52223100,
      "step": 2412,
      "time_per_iteration": 2.792389392852783
    },
    {
      "auxiliary_loss_clip": 0.01153861,
      "auxiliary_loss_mlp": 0.01047504,
      "balance_loss_clip": 1.05600309,
      "balance_loss_mlp": 1.02669024,
      "epoch": 0.14507740868781002,
      "flos": 15596112648960.0,
      "grad_norm": 1.9741671649406984,
      "language_loss": 0.76655865,
      "learning_rate": 3.862684373853579e-06,
      "loss": 0.78857231,
      "num_input_tokens_seen": 52239690,
      "step": 2413,
      "time_per_iteration": 2.6535370349884033
    },
    {
      "auxiliary_loss_clip": 0.01072879,
      "auxiliary_loss_mlp": 0.01028499,
      "balance_loss_clip": 1.04041791,
      "balance_loss_mlp": 1.0252564,
      "epoch": 0.145137531940478,
      "flos": 66675343438080.0,
      "grad_norm": 0.9047547971056389,
      "language_loss": 0.58883119,
      "learning_rate": 3.8625425179337656e-06,
      "loss": 0.60984492,
      "num_input_tokens_seen": 52296705,
      "step": 2414,
      "time_per_iteration": 3.1230342388153076
    },
    {
      "auxiliary_loss_clip": 0.01059489,
      "auxiliary_loss_mlp": 0.01009718,
      "balance_loss_clip": 1.03874373,
      "balance_loss_mlp": 1.00692892,
      "epoch": 0.14519765519314595,
      "flos": 67521578929920.0,
      "grad_norm": 0.8422279258983576,
      "language_loss": 0.62171185,
      "learning_rate": 3.862400591386154e-06,
      "loss": 0.64240396,
      "num_input_tokens_seen": 52361830,
      "step": 2415,
      "time_per_iteration": 3.1932270526885986
    },
    {
      "auxiliary_loss_clip": 0.01151643,
      "auxiliary_loss_mlp": 0.01046675,
      "balance_loss_clip": 1.05383611,
      "balance_loss_mlp": 1.02500319,
      "epoch": 0.14525777844581392,
      "flos": 17198913329280.0,
      "grad_norm": 2.2913061581681036,
      "language_loss": 0.71468806,
      "learning_rate": 3.8622585942161245e-06,
      "loss": 0.73667121,
      "num_input_tokens_seen": 52379420,
      "step": 2416,
      "time_per_iteration": 2.5892374515533447
    },
    {
      "auxiliary_loss_clip": 0.01050816,
      "auxiliary_loss_mlp": 0.010049,
      "balance_loss_clip": 1.03675056,
      "balance_loss_mlp": 1.00211036,
      "epoch": 0.14531790169848188,
      "flos": 65404609015680.0,
      "grad_norm": 0.7147623603004897,
      "language_loss": 0.6037569,
      "learning_rate": 3.8621165264290635e-06,
      "loss": 0.62431407,
      "num_input_tokens_seen": 52446290,
      "step": 2417,
      "time_per_iteration": 3.3065359592437744
    },
    {
      "auxiliary_loss_clip": 0.01168766,
      "auxiliary_loss_mlp": 0.01053548,
      "balance_loss_clip": 1.05357766,
      "balance_loss_mlp": 1.03275824,
      "epoch": 0.14537802495114985,
      "flos": 32562467372160.0,
      "grad_norm": 3.7032433533234346,
      "language_loss": 0.78014368,
      "learning_rate": 3.861974388030356e-06,
      "loss": 0.80236679,
      "num_input_tokens_seen": 52467295,
      "step": 2418,
      "time_per_iteration": 2.887986183166504
    },
    {
      "auxiliary_loss_clip": 0.01114137,
      "auxiliary_loss_mlp": 0.01049779,
      "balance_loss_clip": 1.04354823,
      "balance_loss_mlp": 1.02911985,
      "epoch": 0.1454381482038178,
      "flos": 20226685432320.0,
      "grad_norm": 2.096300480609688,
      "language_loss": 0.71208847,
      "learning_rate": 3.861832179025394e-06,
      "loss": 0.73372757,
      "num_input_tokens_seen": 52487295,
      "step": 2419,
      "time_per_iteration": 2.764268636703491
    },
    {
      "auxiliary_loss_clip": 0.01142427,
      "auxiliary_loss_mlp": 0.01054976,
      "balance_loss_clip": 1.05351484,
      "balance_loss_mlp": 1.03300607,
      "epoch": 0.1454982714564858,
      "flos": 22893124671360.0,
      "grad_norm": 2.414673655978061,
      "language_loss": 0.89847761,
      "learning_rate": 3.861689899419569e-06,
      "loss": 0.92045164,
      "num_input_tokens_seen": 52504220,
      "step": 2420,
      "time_per_iteration": 2.7500016689300537
    },
    {
      "auxiliary_loss_clip": 0.01155004,
      "auxiliary_loss_mlp": 0.01060929,
      "balance_loss_clip": 1.05202007,
      "balance_loss_mlp": 1.04072309,
      "epoch": 0.14555839470915377,
      "flos": 20229845829120.0,
      "grad_norm": 2.0953123539002383,
      "language_loss": 0.82278717,
      "learning_rate": 3.861547549218276e-06,
      "loss": 0.8449465,
      "num_input_tokens_seen": 52521900,
      "step": 2421,
      "time_per_iteration": 2.672722816467285
    },
    {
      "auxiliary_loss_clip": 0.01099277,
      "auxiliary_loss_mlp": 0.01056793,
      "balance_loss_clip": 1.04282439,
      "balance_loss_mlp": 1.03507352,
      "epoch": 0.14561851796182174,
      "flos": 22236282616320.0,
      "grad_norm": 1.667429152986229,
      "language_loss": 0.81741488,
      "learning_rate": 3.861405128426914e-06,
      "loss": 0.83897555,
      "num_input_tokens_seen": 52540495,
      "step": 2422,
      "time_per_iteration": 2.739992141723633
    },
    {
      "auxiliary_loss_clip": 0.01031842,
      "auxiliary_loss_mlp": 0.00760413,
      "balance_loss_clip": 1.0271318,
      "balance_loss_mlp": 1.00019872,
      "epoch": 0.1456786412144897,
      "flos": 52636786289280.0,
      "grad_norm": 0.9102961670465963,
      "language_loss": 0.63342595,
      "learning_rate": 3.861262637050883e-06,
      "loss": 0.65134847,
      "num_input_tokens_seen": 52603305,
      "step": 2423,
      "time_per_iteration": 3.2704036235809326
    },
    {
      "auxiliary_loss_clip": 0.01112855,
      "auxiliary_loss_mlp": 0.00780065,
      "balance_loss_clip": 1.05457556,
      "balance_loss_mlp": 1.00038898,
      "epoch": 0.14573876446715767,
      "flos": 23221671396480.0,
      "grad_norm": 2.2239460229896206,
      "language_loss": 0.82163274,
      "learning_rate": 3.861120075095585e-06,
      "loss": 0.84056193,
      "num_input_tokens_seen": 52623435,
      "step": 2424,
      "time_per_iteration": 2.7993249893188477
    },
    {
      "auxiliary_loss_clip": 0.01141208,
      "auxiliary_loss_mlp": 0.01069468,
      "balance_loss_clip": 1.0535512,
      "balance_loss_mlp": 1.0496788,
      "epoch": 0.14579888771982563,
      "flos": 18114384286080.0,
      "grad_norm": 2.769336045727131,
      "language_loss": 0.78602695,
      "learning_rate": 3.860977442566429e-06,
      "loss": 0.80813372,
      "num_input_tokens_seen": 52642255,
      "step": 2425,
      "time_per_iteration": 2.698594093322754
    },
    {
      "auxiliary_loss_clip": 0.01156078,
      "auxiliary_loss_mlp": 0.01062133,
      "balance_loss_clip": 1.05603778,
      "balance_loss_mlp": 1.04148602,
      "epoch": 0.14585901097249362,
      "flos": 23001107932800.0,
      "grad_norm": 50.77412231982301,
      "language_loss": 0.83184898,
      "learning_rate": 3.860834739468821e-06,
      "loss": 0.85403109,
      "num_input_tokens_seen": 52658700,
      "step": 2426,
      "time_per_iteration": 2.6948676109313965
    },
    {
      "auxiliary_loss_clip": 0.01166642,
      "auxiliary_loss_mlp": 0.01060596,
      "balance_loss_clip": 1.05706,
      "balance_loss_mlp": 1.04040194,
      "epoch": 0.1459191342251616,
      "flos": 21908669644800.0,
      "grad_norm": 3.7420612082917475,
      "language_loss": 0.87215799,
      "learning_rate": 3.860691965808173e-06,
      "loss": 0.8944304,
      "num_input_tokens_seen": 52678140,
      "step": 2427,
      "time_per_iteration": 2.6479666233062744
    },
    {
      "auxiliary_loss_clip": 0.01128634,
      "auxiliary_loss_mlp": 0.01064346,
      "balance_loss_clip": 1.04835391,
      "balance_loss_mlp": 1.0405997,
      "epoch": 0.14597925747782955,
      "flos": 14975504438400.0,
      "grad_norm": 1.9221483903926033,
      "language_loss": 0.66815829,
      "learning_rate": 3.8605491215899e-06,
      "loss": 0.69008809,
      "num_input_tokens_seen": 52696825,
      "step": 2428,
      "time_per_iteration": 2.6971306800842285
    },
    {
      "auxiliary_loss_clip": 0.01155557,
      "auxiliary_loss_mlp": 0.01059343,
      "balance_loss_clip": 1.05335426,
      "balance_loss_mlp": 1.03842235,
      "epoch": 0.14603938073049752,
      "flos": 21068898600960.0,
      "grad_norm": 2.0918238083564242,
      "language_loss": 0.83231717,
      "learning_rate": 3.860406206819417e-06,
      "loss": 0.8544662,
      "num_input_tokens_seen": 52715125,
      "step": 2429,
      "time_per_iteration": 4.283279895782471
    },
    {
      "auxiliary_loss_clip": 0.01120809,
      "auxiliary_loss_mlp": 0.01053505,
      "balance_loss_clip": 1.04625869,
      "balance_loss_mlp": 1.03446746,
      "epoch": 0.14609950398316549,
      "flos": 19864777950720.0,
      "grad_norm": 2.4559042296603746,
      "language_loss": 0.79087842,
      "learning_rate": 3.860263221502145e-06,
      "loss": 0.81262159,
      "num_input_tokens_seen": 52734015,
      "step": 2430,
      "time_per_iteration": 4.197890758514404
    },
    {
      "auxiliary_loss_clip": 0.01170782,
      "auxiliary_loss_mlp": 0.01061965,
      "balance_loss_clip": 1.05820751,
      "balance_loss_mlp": 1.04179525,
      "epoch": 0.14615962723583345,
      "flos": 22418852469120.0,
      "grad_norm": 2.4376691278662506,
      "language_loss": 0.82910693,
      "learning_rate": 3.860120165643504e-06,
      "loss": 0.85143435,
      "num_input_tokens_seen": 52753025,
      "step": 2431,
      "time_per_iteration": 4.162708282470703
    },
    {
      "auxiliary_loss_clip": 0.011607,
      "auxiliary_loss_mlp": 0.01060112,
      "balance_loss_clip": 1.05553937,
      "balance_loss_mlp": 1.03853524,
      "epoch": 0.14621975048850142,
      "flos": 22346241125760.0,
      "grad_norm": 2.881661839068268,
      "language_loss": 0.78330141,
      "learning_rate": 3.859977039248921e-06,
      "loss": 0.80550951,
      "num_input_tokens_seen": 52773420,
      "step": 2432,
      "time_per_iteration": 2.6907777786254883
    },
    {
      "auxiliary_loss_clip": 0.01165399,
      "auxiliary_loss_mlp": 0.00782861,
      "balance_loss_clip": 1.05517077,
      "balance_loss_mlp": 1.00040507,
      "epoch": 0.1462798737411694,
      "flos": 24389163152640.0,
      "grad_norm": 2.3488382544651887,
      "language_loss": 0.79515982,
      "learning_rate": 3.859833842323822e-06,
      "loss": 0.81464243,
      "num_input_tokens_seen": 52792870,
      "step": 2433,
      "time_per_iteration": 2.719841241836548
    },
    {
      "auxiliary_loss_clip": 0.01124303,
      "auxiliary_loss_mlp": 0.01055776,
      "balance_loss_clip": 1.05385411,
      "balance_loss_mlp": 1.03484273,
      "epoch": 0.14633999699383737,
      "flos": 19244672530560.0,
      "grad_norm": 2.0782880949269926,
      "language_loss": 0.77905983,
      "learning_rate": 3.859690574873638e-06,
      "loss": 0.80086064,
      "num_input_tokens_seen": 52811615,
      "step": 2434,
      "time_per_iteration": 4.371506929397583
    },
    {
      "auxiliary_loss_clip": 0.01066282,
      "auxiliary_loss_mlp": 0.01033141,
      "balance_loss_clip": 1.05327988,
      "balance_loss_mlp": 1.03022039,
      "epoch": 0.14640012024650534,
      "flos": 62660638270080.0,
      "grad_norm": 0.8566726319617045,
      "language_loss": 0.58453119,
      "learning_rate": 3.8595472369038e-06,
      "loss": 0.60552537,
      "num_input_tokens_seen": 52873230,
      "step": 2435,
      "time_per_iteration": 3.229882001876831
    },
    {
      "auxiliary_loss_clip": 0.01160087,
      "auxiliary_loss_mlp": 0.01045043,
      "balance_loss_clip": 1.05263698,
      "balance_loss_mlp": 1.0257076,
      "epoch": 0.1464602434991733,
      "flos": 12276243146880.0,
      "grad_norm": 3.775553645712452,
      "language_loss": 0.88436592,
      "learning_rate": 3.859403828419744e-06,
      "loss": 0.90641725,
      "num_input_tokens_seen": 52889325,
      "step": 2436,
      "time_per_iteration": 2.568624973297119
    },
    {
      "auxiliary_loss_clip": 0.011561,
      "auxiliary_loss_mlp": 0.00780257,
      "balance_loss_clip": 1.05587268,
      "balance_loss_mlp": 1.00041819,
      "epoch": 0.14652036675184127,
      "flos": 20922311197440.0,
      "grad_norm": 2.028718201913856,
      "language_loss": 0.74904168,
      "learning_rate": 3.85926034942691e-06,
      "loss": 0.7684052,
      "num_input_tokens_seen": 52909705,
      "step": 2437,
      "time_per_iteration": 2.6361188888549805
    },
    {
      "auxiliary_loss_clip": 0.01165187,
      "auxiliary_loss_mlp": 0.01050068,
      "balance_loss_clip": 1.05295086,
      "balance_loss_mlp": 1.02729869,
      "epoch": 0.14658049000450923,
      "flos": 27703681528320.0,
      "grad_norm": 3.0822234004311033,
      "language_loss": 0.73914421,
      "learning_rate": 3.859116799930736e-06,
      "loss": 0.76129669,
      "num_input_tokens_seen": 52930300,
      "step": 2438,
      "time_per_iteration": 2.7590928077697754
    },
    {
      "auxiliary_loss_clip": 0.01154571,
      "auxiliary_loss_mlp": 0.01046509,
      "balance_loss_clip": 1.05747688,
      "balance_loss_mlp": 1.02708936,
      "epoch": 0.14664061325717723,
      "flos": 24936513575040.0,
      "grad_norm": 4.476318678757457,
      "language_loss": 0.74410725,
      "learning_rate": 3.858973179936668e-06,
      "loss": 0.76611805,
      "num_input_tokens_seen": 52949955,
      "step": 2439,
      "time_per_iteration": 2.627037763595581
    },
    {
      "auxiliary_loss_clip": 0.01152452,
      "auxiliary_loss_mlp": 0.01051294,
      "balance_loss_clip": 1.05477583,
      "balance_loss_mlp": 1.0309453,
      "epoch": 0.1467007365098452,
      "flos": 40297661406720.0,
      "grad_norm": 2.1583973700525343,
      "language_loss": 0.74123728,
      "learning_rate": 3.85882948945015e-06,
      "loss": 0.76327467,
      "num_input_tokens_seen": 52972905,
      "step": 2440,
      "time_per_iteration": 2.79715633392334
    },
    {
      "auxiliary_loss_clip": 0.01160843,
      "auxiliary_loss_mlp": 0.01044034,
      "balance_loss_clip": 1.05471611,
      "balance_loss_mlp": 1.02493691,
      "epoch": 0.14676085976251316,
      "flos": 26541074021760.0,
      "grad_norm": 1.9756103236146798,
      "language_loss": 0.82730794,
      "learning_rate": 3.85868572847663e-06,
      "loss": 0.84935671,
      "num_input_tokens_seen": 52994850,
      "step": 2441,
      "time_per_iteration": 2.6505653858184814
    },
    {
      "auxiliary_loss_clip": 0.01152605,
      "auxiliary_loss_mlp": 0.01049175,
      "balance_loss_clip": 1.05408478,
      "balance_loss_mlp": 1.02796757,
      "epoch": 0.14682098301518112,
      "flos": 23550110380800.0,
      "grad_norm": 2.582118236216862,
      "language_loss": 0.71455544,
      "learning_rate": 3.858541897021563e-06,
      "loss": 0.73657322,
      "num_input_tokens_seen": 53014740,
      "step": 2442,
      "time_per_iteration": 2.772648572921753
    },
    {
      "auxiliary_loss_clip": 0.0113053,
      "auxiliary_loss_mlp": 0.0104246,
      "balance_loss_clip": 1.05283213,
      "balance_loss_mlp": 1.02224207,
      "epoch": 0.1468811062678491,
      "flos": 11651073909120.0,
      "grad_norm": 3.6780587187273155,
      "language_loss": 0.81992352,
      "learning_rate": 3.8583979950904e-06,
      "loss": 0.84165335,
      "num_input_tokens_seen": 53029780,
      "step": 2443,
      "time_per_iteration": 2.6979780197143555
    },
    {
      "auxiliary_loss_clip": 0.01147138,
      "auxiliary_loss_mlp": 0.0105693,
      "balance_loss_clip": 1.05402422,
      "balance_loss_mlp": 1.03474557,
      "epoch": 0.14694122952051705,
      "flos": 23002616304000.0,
      "grad_norm": 3.190851099873364,
      "language_loss": 0.83093917,
      "learning_rate": 3.858254022688599e-06,
      "loss": 0.85297978,
      "num_input_tokens_seen": 53048620,
      "step": 2444,
      "time_per_iteration": 2.7177255153656006
    },
    {
      "auxiliary_loss_clip": 0.01134628,
      "auxiliary_loss_mlp": 0.01051986,
      "balance_loss_clip": 1.05385137,
      "balance_loss_mlp": 1.03213811,
      "epoch": 0.14700135277318502,
      "flos": 26502972670080.0,
      "grad_norm": 3.1425569240832414,
      "language_loss": 0.71183646,
      "learning_rate": 3.85810997982162e-06,
      "loss": 0.7337026,
      "num_input_tokens_seen": 53070055,
      "step": 2445,
      "time_per_iteration": 2.735361099243164
    },
    {
      "auxiliary_loss_clip": 0.01095177,
      "auxiliary_loss_mlp": 0.01023118,
      "balance_loss_clip": 1.05335557,
      "balance_loss_mlp": 1.01999438,
      "epoch": 0.147061476025853,
      "flos": 59449434387840.0,
      "grad_norm": 0.824990401786658,
      "language_loss": 0.63083708,
      "learning_rate": 3.857965866494923e-06,
      "loss": 0.65202004,
      "num_input_tokens_seen": 53126945,
      "step": 2446,
      "time_per_iteration": 3.0853025913238525
    },
    {
      "auxiliary_loss_clip": 0.01120664,
      "auxiliary_loss_mlp": 0.01045249,
      "balance_loss_clip": 1.05621576,
      "balance_loss_mlp": 1.02491164,
      "epoch": 0.14712159927852098,
      "flos": 28330897841280.0,
      "grad_norm": 2.813052009295296,
      "language_loss": 0.74895924,
      "learning_rate": 3.857821682713975e-06,
      "loss": 0.77061838,
      "num_input_tokens_seen": 53149130,
      "step": 2447,
      "time_per_iteration": 2.858643054962158
    },
    {
      "auxiliary_loss_clip": 0.01168929,
      "auxiliary_loss_mlp": 0.01042907,
      "balance_loss_clip": 1.0604012,
      "balance_loss_mlp": 1.02383327,
      "epoch": 0.14718172253118894,
      "flos": 27089825074560.0,
      "grad_norm": 2.2427639286159367,
      "language_loss": 0.8528471,
      "learning_rate": 3.857677428484242e-06,
      "loss": 0.87496543,
      "num_input_tokens_seen": 53167120,
      "step": 2448,
      "time_per_iteration": 2.699781894683838
    },
    {
      "auxiliary_loss_clip": 0.01092169,
      "auxiliary_loss_mlp": 0.01019616,
      "balance_loss_clip": 1.05051064,
      "balance_loss_mlp": 1.01654005,
      "epoch": 0.1472418457838569,
      "flos": 66706764860160.0,
      "grad_norm": 0.7683837313264128,
      "language_loss": 0.56829578,
      "learning_rate": 3.857533103811195e-06,
      "loss": 0.58941364,
      "num_input_tokens_seen": 53227945,
      "step": 2449,
      "time_per_iteration": 3.1478211879730225
    },
    {
      "auxiliary_loss_clip": 0.01135016,
      "auxiliary_loss_mlp": 0.01050801,
      "balance_loss_clip": 1.05464292,
      "balance_loss_mlp": 1.03023791,
      "epoch": 0.14730196903652487,
      "flos": 19573578391680.0,
      "grad_norm": 1.9048653074507311,
      "language_loss": 0.85067344,
      "learning_rate": 3.857388708700307e-06,
      "loss": 0.87253165,
      "num_input_tokens_seen": 53244615,
      "step": 2450,
      "time_per_iteration": 2.726008653640747
    },
    {
      "auxiliary_loss_clip": 0.01158708,
      "auxiliary_loss_mlp": 0.01049735,
      "balance_loss_clip": 1.05984712,
      "balance_loss_mlp": 1.02994645,
      "epoch": 0.14736209228919284,
      "flos": 16071031296000.0,
      "grad_norm": 2.306043539040143,
      "language_loss": 0.74523091,
      "learning_rate": 3.857244243157052e-06,
      "loss": 0.76731533,
      "num_input_tokens_seen": 53262205,
      "step": 2451,
      "time_per_iteration": 2.641082286834717
    },
    {
      "auxiliary_loss_clip": 0.01133915,
      "auxiliary_loss_mlp": 0.01038458,
      "balance_loss_clip": 1.05399728,
      "balance_loss_mlp": 1.02031422,
      "epoch": 0.1474222155418608,
      "flos": 23039460679680.0,
      "grad_norm": 1.8026547738986978,
      "language_loss": 0.82384264,
      "learning_rate": 3.85709970718691e-06,
      "loss": 0.84556639,
      "num_input_tokens_seen": 53282445,
      "step": 2452,
      "time_per_iteration": 2.7810096740722656
    },
    {
      "auxiliary_loss_clip": 0.01101553,
      "auxiliary_loss_mlp": 0.01041864,
      "balance_loss_clip": 1.05924153,
      "balance_loss_mlp": 1.0238874,
      "epoch": 0.1474823387945288,
      "flos": 17018641946880.0,
      "grad_norm": 1.6675065143572472,
      "language_loss": 0.74075705,
      "learning_rate": 3.856955100795361e-06,
      "loss": 0.76219124,
      "num_input_tokens_seen": 53299060,
      "step": 2453,
      "time_per_iteration": 2.7913167476654053
    },
    {
      "auxiliary_loss_clip": 0.01141798,
      "auxiliary_loss_mlp": 0.0104607,
      "balance_loss_clip": 1.05557632,
      "balance_loss_mlp": 1.026353,
      "epoch": 0.14754246204719676,
      "flos": 17895041884800.0,
      "grad_norm": 1.9958141581621542,
      "language_loss": 0.7558704,
      "learning_rate": 3.856810423987889e-06,
      "loss": 0.77774906,
      "num_input_tokens_seen": 53315970,
      "step": 2454,
      "time_per_iteration": 2.7199089527130127
    },
    {
      "auxiliary_loss_clip": 0.01147348,
      "auxiliary_loss_mlp": 0.01038134,
      "balance_loss_clip": 1.05733335,
      "balance_loss_mlp": 1.01864362,
      "epoch": 0.14760258529986472,
      "flos": 13079097987840.0,
      "grad_norm": 2.0858167958418674,
      "language_loss": 0.83077228,
      "learning_rate": 3.856665676769979e-06,
      "loss": 0.85262716,
      "num_input_tokens_seen": 53332940,
      "step": 2455,
      "time_per_iteration": 2.75616192817688
    },
    {
      "auxiliary_loss_clip": 0.01130504,
      "auxiliary_loss_mlp": 0.01042951,
      "balance_loss_clip": 1.05704689,
      "balance_loss_mlp": 1.02452159,
      "epoch": 0.1476627085525327,
      "flos": 30806399358720.0,
      "grad_norm": 2.3702229998953976,
      "language_loss": 0.83881497,
      "learning_rate": 3.85652085914712e-06,
      "loss": 0.86054951,
      "num_input_tokens_seen": 53353295,
      "step": 2456,
      "time_per_iteration": 2.7914254665374756
    },
    {
      "auxiliary_loss_clip": 0.01154014,
      "auxiliary_loss_mlp": 0.01043715,
      "balance_loss_clip": 1.05863023,
      "balance_loss_mlp": 1.02514231,
      "epoch": 0.14772283180520066,
      "flos": 21689434984320.0,
      "grad_norm": 2.4172359629848996,
      "language_loss": 0.84154665,
      "learning_rate": 3.856375971124805e-06,
      "loss": 0.86352402,
      "num_input_tokens_seen": 53373410,
      "step": 2457,
      "time_per_iteration": 2.688265323638916
    },
    {
      "auxiliary_loss_clip": 0.01155788,
      "auxiliary_loss_mlp": 0.01042903,
      "balance_loss_clip": 1.06250155,
      "balance_loss_mlp": 1.02529585,
      "epoch": 0.14778295505786862,
      "flos": 18770400328320.0,
      "grad_norm": 6.310680797376285,
      "language_loss": 0.75692672,
      "learning_rate": 3.856231012708527e-06,
      "loss": 0.77891362,
      "num_input_tokens_seen": 53391430,
      "step": 2458,
      "time_per_iteration": 2.698697805404663
    },
    {
      "auxiliary_loss_clip": 0.01117404,
      "auxiliary_loss_mlp": 0.01047753,
      "balance_loss_clip": 1.05451179,
      "balance_loss_mlp": 1.02718902,
      "epoch": 0.1478430783105366,
      "flos": 22893555634560.0,
      "grad_norm": 3.1268711361266393,
      "language_loss": 0.83348328,
      "learning_rate": 3.856085983903782e-06,
      "loss": 0.85513484,
      "num_input_tokens_seen": 53409960,
      "step": 2459,
      "time_per_iteration": 2.790552854537964
    },
    {
      "auxiliary_loss_clip": 0.01126767,
      "auxiliary_loss_mlp": 0.01042293,
      "balance_loss_clip": 1.05070424,
      "balance_loss_mlp": 1.02435231,
      "epoch": 0.14790320156320458,
      "flos": 15085319293440.0,
      "grad_norm": 3.1203941208753534,
      "language_loss": 0.7554391,
      "learning_rate": 3.855940884716071e-06,
      "loss": 0.77712965,
      "num_input_tokens_seen": 53426160,
      "step": 2460,
      "time_per_iteration": 2.815455675125122
    },
    {
      "auxiliary_loss_clip": 0.01134117,
      "auxiliary_loss_mlp": 0.01056838,
      "balance_loss_clip": 1.05845904,
      "balance_loss_mlp": 1.03770471,
      "epoch": 0.14796332481587254,
      "flos": 26504768350080.0,
      "grad_norm": 3.59241393994,
      "language_loss": 0.81227219,
      "learning_rate": 3.855795715150896e-06,
      "loss": 0.83418173,
      "num_input_tokens_seen": 53448530,
      "step": 2461,
      "time_per_iteration": 2.785569190979004
    },
    {
      "auxiliary_loss_clip": 0.01156748,
      "auxiliary_loss_mlp": 0.01051178,
      "balance_loss_clip": 1.05812359,
      "balance_loss_mlp": 1.03044713,
      "epoch": 0.1480234480685405,
      "flos": 17563191108480.0,
      "grad_norm": 3.2910626990147183,
      "language_loss": 0.66117477,
      "learning_rate": 3.855650475213761e-06,
      "loss": 0.683254,
      "num_input_tokens_seen": 53465915,
      "step": 2462,
      "time_per_iteration": 2.7222983837127686
    },
    {
      "auxiliary_loss_clip": 0.01136035,
      "auxiliary_loss_mlp": 0.01049537,
      "balance_loss_clip": 1.05622339,
      "balance_loss_mlp": 1.02965331,
      "epoch": 0.14808357132120847,
      "flos": 53582203232640.0,
      "grad_norm": 1.8120706772856114,
      "language_loss": 0.67226064,
      "learning_rate": 3.8555051649101745e-06,
      "loss": 0.69411635,
      "num_input_tokens_seen": 53496055,
      "step": 2463,
      "time_per_iteration": 3.0344398021698
    },
    {
      "auxiliary_loss_clip": 0.01153077,
      "auxiliary_loss_mlp": 0.01050435,
      "balance_loss_clip": 1.05550933,
      "balance_loss_mlp": 1.0307889,
      "epoch": 0.14814369457387644,
      "flos": 19829190551040.0,
      "grad_norm": 1.9881580745750587,
      "language_loss": 0.76870739,
      "learning_rate": 3.855359784245646e-06,
      "loss": 0.79074258,
      "num_input_tokens_seen": 53513790,
      "step": 2464,
      "time_per_iteration": 2.69480037689209
    },
    {
      "auxiliary_loss_clip": 0.01133748,
      "auxiliary_loss_mlp": 0.01057139,
      "balance_loss_clip": 1.05392432,
      "balance_loss_mlp": 1.03769565,
      "epoch": 0.1482038178265444,
      "flos": 23914962777600.0,
      "grad_norm": 1.8401367705559406,
      "language_loss": 0.79628456,
      "learning_rate": 3.855214333225688e-06,
      "loss": 0.81819344,
      "num_input_tokens_seen": 53533410,
      "step": 2465,
      "time_per_iteration": 2.6989939212799072
    },
    {
      "auxiliary_loss_clip": 0.01170385,
      "auxiliary_loss_mlp": 0.01054925,
      "balance_loss_clip": 1.06119514,
      "balance_loss_mlp": 1.03568494,
      "epoch": 0.1482639410792124,
      "flos": 24170503109760.0,
      "grad_norm": 2.005541134809237,
      "language_loss": 0.76272273,
      "learning_rate": 3.855068811855817e-06,
      "loss": 0.78497583,
      "num_input_tokens_seen": 53554775,
      "step": 2466,
      "time_per_iteration": 2.646245002746582
    },
    {
      "auxiliary_loss_clip": 0.01018939,
      "auxiliary_loss_mlp": 0.0114331,
      "balance_loss_clip": 1.03313899,
      "balance_loss_mlp": 1.14004362,
      "epoch": 0.14832406433188036,
      "flos": 66191051341440.0,
      "grad_norm": 0.8320983618395327,
      "language_loss": 0.6004858,
      "learning_rate": 3.854923220141551e-06,
      "loss": 0.62210834,
      "num_input_tokens_seen": 53609675,
      "step": 2467,
      "time_per_iteration": 3.33776593208313
    },
    {
      "auxiliary_loss_clip": 0.01141854,
      "auxiliary_loss_mlp": 0.01044026,
      "balance_loss_clip": 1.05437851,
      "balance_loss_mlp": 1.02509522,
      "epoch": 0.14838418758454833,
      "flos": 25411252654080.0,
      "grad_norm": 2.92694776694492,
      "language_loss": 0.87666196,
      "learning_rate": 3.85477755808841e-06,
      "loss": 0.89852077,
      "num_input_tokens_seen": 53626950,
      "step": 2468,
      "time_per_iteration": 4.266207456588745
    },
    {
      "auxiliary_loss_clip": 0.01130189,
      "auxiliary_loss_mlp": 0.01048186,
      "balance_loss_clip": 1.05255163,
      "balance_loss_mlp": 1.02782488,
      "epoch": 0.1484443108372163,
      "flos": 23289901280640.0,
      "grad_norm": 2.2284173124426223,
      "language_loss": 0.7598694,
      "learning_rate": 3.854631825701919e-06,
      "loss": 0.78165317,
      "num_input_tokens_seen": 53644200,
      "step": 2469,
      "time_per_iteration": 4.217481851577759
    },
    {
      "auxiliary_loss_clip": 0.01126269,
      "auxiliary_loss_mlp": 0.0104139,
      "balance_loss_clip": 1.05208421,
      "balance_loss_mlp": 1.02251911,
      "epoch": 0.14850443408988426,
      "flos": 14647675985280.0,
      "grad_norm": 6.591244267451795,
      "language_loss": 0.75895017,
      "learning_rate": 3.854486022987603e-06,
      "loss": 0.78062677,
      "num_input_tokens_seen": 53659650,
      "step": 2470,
      "time_per_iteration": 2.7157187461853027
    },
    {
      "auxiliary_loss_clip": 0.01161157,
      "auxiliary_loss_mlp": 0.01044729,
      "balance_loss_clip": 1.05831027,
      "balance_loss_mlp": 1.02571499,
      "epoch": 0.14856455734255222,
      "flos": 23548314700800.0,
      "grad_norm": 1.8610043660805562,
      "language_loss": 0.7215873,
      "learning_rate": 3.8543401499509905e-06,
      "loss": 0.74364614,
      "num_input_tokens_seen": 53680275,
      "step": 2471,
      "time_per_iteration": 4.162387132644653
    },
    {
      "auxiliary_loss_clip": 0.01135244,
      "auxiliary_loss_mlp": 0.01047611,
      "balance_loss_clip": 1.05438995,
      "balance_loss_mlp": 1.02717888,
      "epoch": 0.1486246805952202,
      "flos": 18077288515200.0,
      "grad_norm": 1.979025280241548,
      "language_loss": 0.89558828,
      "learning_rate": 3.854194206597615e-06,
      "loss": 0.91741687,
      "num_input_tokens_seen": 53698270,
      "step": 2472,
      "time_per_iteration": 2.739457607269287
    },
    {
      "auxiliary_loss_clip": 0.01134625,
      "auxiliary_loss_mlp": 0.01049109,
      "balance_loss_clip": 1.06334805,
      "balance_loss_mlp": 1.02964163,
      "epoch": 0.14868480384788818,
      "flos": 19353625459200.0,
      "grad_norm": 2.6029609251362764,
      "language_loss": 0.80801564,
      "learning_rate": 3.854048192933008e-06,
      "loss": 0.82985294,
      "num_input_tokens_seen": 53716845,
      "step": 2473,
      "time_per_iteration": 4.412883758544922
    },
    {
      "auxiliary_loss_clip": 0.01161034,
      "auxiliary_loss_mlp": 0.01051306,
      "balance_loss_clip": 1.0626657,
      "balance_loss_mlp": 1.03267312,
      "epoch": 0.14874492710055615,
      "flos": 22200192426240.0,
      "grad_norm": 3.426519274325147,
      "language_loss": 0.77372944,
      "learning_rate": 3.853902108962709e-06,
      "loss": 0.79585278,
      "num_input_tokens_seen": 53734970,
      "step": 2474,
      "time_per_iteration": 2.6879520416259766
    },
    {
      "auxiliary_loss_clip": 0.01124216,
      "auxiliary_loss_mlp": 0.01059785,
      "balance_loss_clip": 1.05597806,
      "balance_loss_mlp": 1.04041362,
      "epoch": 0.1488050503532241,
      "flos": 21103444506240.0,
      "grad_norm": 2.4771626433268734,
      "language_loss": 0.82151824,
      "learning_rate": 3.853755954692255e-06,
      "loss": 0.84335828,
      "num_input_tokens_seen": 53753415,
      "step": 2475,
      "time_per_iteration": 2.7828469276428223
    },
    {
      "auxiliary_loss_clip": 0.01115855,
      "auxiliary_loss_mlp": 0.01052322,
      "balance_loss_clip": 1.0614953,
      "balance_loss_mlp": 1.03341544,
      "epoch": 0.14886517360589208,
      "flos": 12786569625600.0,
      "grad_norm": 1.9349243252831771,
      "language_loss": 0.80917645,
      "learning_rate": 3.85360973012719e-06,
      "loss": 0.83085823,
      "num_input_tokens_seen": 53770305,
      "step": 2476,
      "time_per_iteration": 2.7227590084075928
    },
    {
      "auxiliary_loss_clip": 0.01156019,
      "auxiliary_loss_mlp": 0.0105036,
      "balance_loss_clip": 1.06338036,
      "balance_loss_mlp": 1.03216898,
      "epoch": 0.14892529685856004,
      "flos": 29022860419200.0,
      "grad_norm": 2.0032169897498346,
      "language_loss": 0.77659523,
      "learning_rate": 3.853463435273058e-06,
      "loss": 0.79865897,
      "num_input_tokens_seen": 53788895,
      "step": 2477,
      "time_per_iteration": 2.740241765975952
    },
    {
      "auxiliary_loss_clip": 0.0110234,
      "auxiliary_loss_mlp": 0.01092005,
      "balance_loss_clip": 1.07879949,
      "balance_loss_mlp": 1.08730817,
      "epoch": 0.148985420111228,
      "flos": 61926121054080.0,
      "grad_norm": 0.8188153224748298,
      "language_loss": 0.60153681,
      "learning_rate": 3.853317070135407e-06,
      "loss": 0.62348026,
      "num_input_tokens_seen": 53850260,
      "step": 2478,
      "time_per_iteration": 3.2467947006225586
    },
    {
      "auxiliary_loss_clip": 0.01107417,
      "auxiliary_loss_mlp": 0.01048452,
      "balance_loss_clip": 1.0516423,
      "balance_loss_mlp": 1.03041577,
      "epoch": 0.149045543363896,
      "flos": 23915106432000.0,
      "grad_norm": 2.666109649137694,
      "language_loss": 0.7139731,
      "learning_rate": 3.853170634719787e-06,
      "loss": 0.73553181,
      "num_input_tokens_seen": 53867520,
      "step": 2479,
      "time_per_iteration": 2.7973475456237793
    },
    {
      "auxiliary_loss_clip": 0.01140551,
      "auxiliary_loss_mlp": 0.01043104,
      "balance_loss_clip": 1.05563831,
      "balance_loss_mlp": 1.02407789,
      "epoch": 0.14910566661656396,
      "flos": 23654394541440.0,
      "grad_norm": 1.7687137634424535,
      "language_loss": 0.80758464,
      "learning_rate": 3.853024129031751e-06,
      "loss": 0.82942122,
      "num_input_tokens_seen": 53886620,
      "step": 2480,
      "time_per_iteration": 2.7238829135894775
    },
    {
      "auxiliary_loss_clip": 0.01138106,
      "auxiliary_loss_mlp": 0.0104537,
      "balance_loss_clip": 1.0584991,
      "balance_loss_mlp": 1.02627277,
      "epoch": 0.14916578986923193,
      "flos": 20515299212160.0,
      "grad_norm": 4.65741826395702,
      "language_loss": 0.84375542,
      "learning_rate": 3.852877553076854e-06,
      "loss": 0.86559021,
      "num_input_tokens_seen": 53902230,
      "step": 2481,
      "time_per_iteration": 2.791550874710083
    },
    {
      "auxiliary_loss_clip": 0.01149484,
      "auxiliary_loss_mlp": 0.01050268,
      "balance_loss_clip": 1.05772805,
      "balance_loss_mlp": 1.02948999,
      "epoch": 0.1492259131218999,
      "flos": 22491822948480.0,
      "grad_norm": 8.035113387353048,
      "language_loss": 0.77703977,
      "learning_rate": 3.8527309068606546e-06,
      "loss": 0.79903734,
      "num_input_tokens_seen": 53919475,
      "step": 2482,
      "time_per_iteration": 2.7310593128204346
    },
    {
      "auxiliary_loss_clip": 0.01133163,
      "auxiliary_loss_mlp": 0.01040426,
      "balance_loss_clip": 1.05452228,
      "balance_loss_mlp": 1.02032781,
      "epoch": 0.14928603637456786,
      "flos": 23185868515200.0,
      "grad_norm": 2.207731010812049,
      "language_loss": 0.78967929,
      "learning_rate": 3.852584190388713e-06,
      "loss": 0.81141514,
      "num_input_tokens_seen": 53939150,
      "step": 2483,
      "time_per_iteration": 2.749671220779419
    },
    {
      "auxiliary_loss_clip": 0.01154122,
      "auxiliary_loss_mlp": 0.00776708,
      "balance_loss_clip": 1.06144214,
      "balance_loss_mlp": 1.00029397,
      "epoch": 0.14934615962723582,
      "flos": 21653237053440.0,
      "grad_norm": 2.020127706544282,
      "language_loss": 0.70361555,
      "learning_rate": 3.852437403666595e-06,
      "loss": 0.72292387,
      "num_input_tokens_seen": 53958735,
      "step": 2484,
      "time_per_iteration": 2.737781524658203
    },
    {
      "auxiliary_loss_clip": 0.01141919,
      "auxiliary_loss_mlp": 0.00778215,
      "balance_loss_clip": 1.05718136,
      "balance_loss_mlp": 1.00030363,
      "epoch": 0.1494062828799038,
      "flos": 27010066924800.0,
      "grad_norm": 2.165877689982274,
      "language_loss": 0.84666765,
      "learning_rate": 3.852290546699863e-06,
      "loss": 0.86586899,
      "num_input_tokens_seen": 53975065,
      "step": 2485,
      "time_per_iteration": 2.697976589202881
    },
    {
      "auxiliary_loss_clip": 0.01145272,
      "auxiliary_loss_mlp": 0.0104224,
      "balance_loss_clip": 1.05639958,
      "balance_loss_mlp": 1.02257001,
      "epoch": 0.14946640613257178,
      "flos": 21214947300480.0,
      "grad_norm": 2.5229241908443023,
      "language_loss": 0.8476423,
      "learning_rate": 3.8521436194940894e-06,
      "loss": 0.86951739,
      "num_input_tokens_seen": 53993330,
      "step": 2486,
      "time_per_iteration": 2.6799628734588623
    },
    {
      "auxiliary_loss_clip": 0.01149031,
      "auxiliary_loss_mlp": 0.01039312,
      "balance_loss_clip": 1.05667424,
      "balance_loss_mlp": 1.0230875,
      "epoch": 0.14952652938523975,
      "flos": 13370872164480.0,
      "grad_norm": 2.1822908802725203,
      "language_loss": 0.74762607,
      "learning_rate": 3.851996622054842e-06,
      "loss": 0.76950949,
      "num_input_tokens_seen": 54010515,
      "step": 2487,
      "time_per_iteration": 2.8037290573120117
    },
    {
      "auxiliary_loss_clip": 0.01153097,
      "auxiliary_loss_mlp": 0.01044274,
      "balance_loss_clip": 1.05934322,
      "balance_loss_mlp": 1.02611899,
      "epoch": 0.1495866526379077,
      "flos": 35517699959040.0,
      "grad_norm": 16.320028017118723,
      "language_loss": 0.72210175,
      "learning_rate": 3.8518495543877e-06,
      "loss": 0.74407548,
      "num_input_tokens_seen": 54031315,
      "step": 2488,
      "time_per_iteration": 2.8031094074249268
    },
    {
      "auxiliary_loss_clip": 0.01137536,
      "auxiliary_loss_mlp": 0.01054916,
      "balance_loss_clip": 1.05569518,
      "balance_loss_mlp": 1.03636682,
      "epoch": 0.14964677589057568,
      "flos": 17632749795840.0,
      "grad_norm": 3.2458980886023143,
      "language_loss": 0.71352434,
      "learning_rate": 3.851702416498235e-06,
      "loss": 0.73544884,
      "num_input_tokens_seen": 54045965,
      "step": 2489,
      "time_per_iteration": 2.648883819580078
    },
    {
      "auxiliary_loss_clip": 0.0113767,
      "auxiliary_loss_mlp": 0.01052603,
      "balance_loss_clip": 1.05376494,
      "balance_loss_mlp": 1.03357768,
      "epoch": 0.14970689914324364,
      "flos": 20185280029440.0,
      "grad_norm": 3.893198448080141,
      "language_loss": 0.81559736,
      "learning_rate": 3.8515552083920295e-06,
      "loss": 0.8375001,
      "num_input_tokens_seen": 54059960,
      "step": 2490,
      "time_per_iteration": 2.702808380126953
    },
    {
      "auxiliary_loss_clip": 0.01125097,
      "auxiliary_loss_mlp": 0.01055928,
      "balance_loss_clip": 1.05606139,
      "balance_loss_mlp": 1.03803492,
      "epoch": 0.1497670223959116,
      "flos": 37228699382400.0,
      "grad_norm": 1.9071281232744548,
      "language_loss": 0.80057055,
      "learning_rate": 3.851407930074666e-06,
      "loss": 0.82238084,
      "num_input_tokens_seen": 54079330,
      "step": 2491,
      "time_per_iteration": 2.833272933959961
    },
    {
      "auxiliary_loss_clip": 0.01143407,
      "auxiliary_loss_mlp": 0.01052558,
      "balance_loss_clip": 1.05301452,
      "balance_loss_mlp": 1.03195894,
      "epoch": 0.1498271456485796,
      "flos": 24455848752000.0,
      "grad_norm": 2.3105790695512294,
      "language_loss": 0.90820229,
      "learning_rate": 3.851260581551727e-06,
      "loss": 0.93016195,
      "num_input_tokens_seen": 54097555,
      "step": 2492,
      "time_per_iteration": 2.684178352355957
    },
    {
      "auxiliary_loss_clip": 0.01152331,
      "auxiliary_loss_mlp": 0.01063543,
      "balance_loss_clip": 1.05835843,
      "balance_loss_mlp": 1.04508913,
      "epoch": 0.14988726890124757,
      "flos": 16253601148800.0,
      "grad_norm": 6.881290297472923,
      "language_loss": 0.79406559,
      "learning_rate": 3.851113162828802e-06,
      "loss": 0.81622434,
      "num_input_tokens_seen": 54115600,
      "step": 2493,
      "time_per_iteration": 2.6558918952941895
    },
    {
      "auxiliary_loss_clip": 0.0114858,
      "auxiliary_loss_mlp": 0.01052018,
      "balance_loss_clip": 1.05345511,
      "balance_loss_mlp": 1.03258693,
      "epoch": 0.14994739215391553,
      "flos": 20666555383680.0,
      "grad_norm": 2.3431247769189967,
      "language_loss": 0.79894584,
      "learning_rate": 3.85096567391148e-06,
      "loss": 0.82095182,
      "num_input_tokens_seen": 54135220,
      "step": 2494,
      "time_per_iteration": 2.6774168014526367
    },
    {
      "auxiliary_loss_clip": 0.01137216,
      "auxiliary_loss_mlp": 0.01050857,
      "balance_loss_clip": 1.05474579,
      "balance_loss_mlp": 1.03212965,
      "epoch": 0.1500075154065835,
      "flos": 70652375239680.0,
      "grad_norm": 1.928941284350508,
      "language_loss": 0.66480517,
      "learning_rate": 3.850818114805354e-06,
      "loss": 0.68668592,
      "num_input_tokens_seen": 54161065,
      "step": 2495,
      "time_per_iteration": 3.1090729236602783
    },
    {
      "auxiliary_loss_clip": 0.01103374,
      "auxiliary_loss_mlp": 0.01038654,
      "balance_loss_clip": 1.06896818,
      "balance_loss_mlp": 1.03560257,
      "epoch": 0.15006763865925146,
      "flos": 68011937447040.0,
      "grad_norm": 0.9030283421527312,
      "language_loss": 0.59524739,
      "learning_rate": 3.850670485516019e-06,
      "loss": 0.61666763,
      "num_input_tokens_seen": 54225095,
      "step": 2496,
      "time_per_iteration": 3.2250726222991943
    },
    {
      "auxiliary_loss_clip": 0.01163934,
      "auxiliary_loss_mlp": 0.01055725,
      "balance_loss_clip": 1.05690169,
      "balance_loss_mlp": 1.0360074,
      "epoch": 0.15012776191191943,
      "flos": 18916269459840.0,
      "grad_norm": 3.063784198565679,
      "language_loss": 0.65276247,
      "learning_rate": 3.850522786049075e-06,
      "loss": 0.67495906,
      "num_input_tokens_seen": 54243750,
      "step": 2497,
      "time_per_iteration": 2.619946002960205
    },
    {
      "auxiliary_loss_clip": 0.01125657,
      "auxiliary_loss_mlp": 0.01054091,
      "balance_loss_clip": 1.05308235,
      "balance_loss_mlp": 1.03316998,
      "epoch": 0.1501878851645874,
      "flos": 23701330638720.0,
      "grad_norm": 1.5552670947231086,
      "language_loss": 0.75182658,
      "learning_rate": 3.850375016410121e-06,
      "loss": 0.77362406,
      "num_input_tokens_seen": 54266185,
      "step": 2498,
      "time_per_iteration": 2.778163433074951
    },
    {
      "auxiliary_loss_clip": 0.01132738,
      "auxiliary_loss_mlp": 0.01046919,
      "balance_loss_clip": 1.05919099,
      "balance_loss_mlp": 1.02701163,
      "epoch": 0.15024800841725539,
      "flos": 20412523422720.0,
      "grad_norm": 3.357364003851319,
      "language_loss": 0.71821117,
      "learning_rate": 3.850227176604761e-06,
      "loss": 0.74000776,
      "num_input_tokens_seen": 54283940,
      "step": 2499,
      "time_per_iteration": 2.6929259300231934
    },
    {
      "auxiliary_loss_clip": 0.01134239,
      "auxiliary_loss_mlp": 0.01051817,
      "balance_loss_clip": 1.0547812,
      "balance_loss_mlp": 1.03236222,
      "epoch": 0.15030813166992335,
      "flos": 31831002812160.0,
      "grad_norm": 2.1406696998963652,
      "language_loss": 0.7206136,
      "learning_rate": 3.850079266638601e-06,
      "loss": 0.7424742,
      "num_input_tokens_seen": 54304830,
      "step": 2500,
      "time_per_iteration": 2.769988536834717
    },
    {
      "auxiliary_loss_clip": 0.01134021,
      "auxiliary_loss_mlp": 0.0105021,
      "balance_loss_clip": 1.06063724,
      "balance_loss_mlp": 1.03181624,
      "epoch": 0.15036825492259132,
      "flos": 35657822914560.0,
      "grad_norm": 2.0251881980439306,
      "language_loss": 0.65127194,
      "learning_rate": 3.849931286517249e-06,
      "loss": 0.6731143,
      "num_input_tokens_seen": 54325595,
      "step": 2501,
      "time_per_iteration": 2.810945510864258
    },
    {
      "auxiliary_loss_clip": 0.01137877,
      "auxiliary_loss_mlp": 0.01055223,
      "balance_loss_clip": 1.0541079,
      "balance_loss_mlp": 1.03511274,
      "epoch": 0.15042837817525928,
      "flos": 18838163335680.0,
      "grad_norm": 2.209666371186328,
      "language_loss": 0.83401144,
      "learning_rate": 3.849783236246318e-06,
      "loss": 0.85594243,
      "num_input_tokens_seen": 54342180,
      "step": 2502,
      "time_per_iteration": 2.6780545711517334
    },
    {
      "auxiliary_loss_clip": 0.01122961,
      "auxiliary_loss_mlp": 0.01049887,
      "balance_loss_clip": 1.05318308,
      "balance_loss_mlp": 1.0323875,
      "epoch": 0.15048850142792725,
      "flos": 19535548867200.0,
      "grad_norm": 2.0319272128830947,
      "language_loss": 0.77134645,
      "learning_rate": 3.849635115831421e-06,
      "loss": 0.79307491,
      "num_input_tokens_seen": 54360255,
      "step": 2503,
      "time_per_iteration": 2.7579123973846436
    },
    {
      "auxiliary_loss_clip": 0.01159116,
      "auxiliary_loss_mlp": 0.01044094,
      "balance_loss_clip": 1.05766046,
      "balance_loss_mlp": 1.02692807,
      "epoch": 0.1505486246805952,
      "flos": 22017550746240.0,
      "grad_norm": 1.9852139459946199,
      "language_loss": 0.85514295,
      "learning_rate": 3.849486925278176e-06,
      "loss": 0.87717503,
      "num_input_tokens_seen": 54378260,
      "step": 2504,
      "time_per_iteration": 2.631882905960083
    },
    {
      "auxiliary_loss_clip": 0.01146113,
      "auxiliary_loss_mlp": 0.01048035,
      "balance_loss_clip": 1.05622697,
      "balance_loss_mlp": 1.03098798,
      "epoch": 0.15060874793326318,
      "flos": 20743153136640.0,
      "grad_norm": 1.8222645508164372,
      "language_loss": 0.83178544,
      "learning_rate": 3.8493386645922e-06,
      "loss": 0.85372692,
      "num_input_tokens_seen": 54399745,
      "step": 2505,
      "time_per_iteration": 2.7706007957458496
    },
    {
      "auxiliary_loss_clip": 0.01125699,
      "auxiliary_loss_mlp": 0.01053819,
      "balance_loss_clip": 1.05586648,
      "balance_loss_mlp": 1.03590202,
      "epoch": 0.15066887118593117,
      "flos": 16471902055680.0,
      "grad_norm": 2.0148067518000445,
      "language_loss": 0.76044405,
      "learning_rate": 3.849190333779117e-06,
      "loss": 0.7822392,
      "num_input_tokens_seen": 54417105,
      "step": 2506,
      "time_per_iteration": 2.70989990234375
    },
    {
      "auxiliary_loss_clip": 0.01165314,
      "auxiliary_loss_mlp": 0.01041911,
      "balance_loss_clip": 1.05785728,
      "balance_loss_mlp": 1.02305174,
      "epoch": 0.15072899443859913,
      "flos": 19859319083520.0,
      "grad_norm": 2.823460856599666,
      "language_loss": 0.76220375,
      "learning_rate": 3.849041932844552e-06,
      "loss": 0.78427601,
      "num_input_tokens_seen": 54433920,
      "step": 2507,
      "time_per_iteration": 2.5367634296417236
    },
    {
      "auxiliary_loss_clip": 0.01144479,
      "auxiliary_loss_mlp": 0.01041094,
      "balance_loss_clip": 1.05261898,
      "balance_loss_mlp": 1.02306986,
      "epoch": 0.1507891176912671,
      "flos": 20776226584320.0,
      "grad_norm": 2.5197772895304906,
      "language_loss": 0.68633789,
      "learning_rate": 3.848893461794131e-06,
      "loss": 0.70819366,
      "num_input_tokens_seen": 54451540,
      "step": 2508,
      "time_per_iteration": 4.303388833999634
    },
    {
      "auxiliary_loss_clip": 0.01130299,
      "auxiliary_loss_mlp": 0.01046507,
      "balance_loss_clip": 1.05477214,
      "balance_loss_mlp": 1.02835178,
      "epoch": 0.15084924094393506,
      "flos": 23586631534080.0,
      "grad_norm": 2.840517748098311,
      "language_loss": 0.77994299,
      "learning_rate": 3.8487449206334845e-06,
      "loss": 0.80171108,
      "num_input_tokens_seen": 54470800,
      "step": 2509,
      "time_per_iteration": 4.380200147628784
    },
    {
      "auxiliary_loss_clip": 0.01141335,
      "auxiliary_loss_mlp": 0.00776843,
      "balance_loss_clip": 1.05463386,
      "balance_loss_mlp": 1.00027037,
      "epoch": 0.15090936419660303,
      "flos": 18911313383040.0,
      "grad_norm": 2.53406994590866,
      "language_loss": 0.79959804,
      "learning_rate": 3.848596309368246e-06,
      "loss": 0.81877983,
      "num_input_tokens_seen": 54486525,
      "step": 2510,
      "time_per_iteration": 4.219487428665161
    },
    {
      "auxiliary_loss_clip": 0.01150641,
      "auxiliary_loss_mlp": 0.01047345,
      "balance_loss_clip": 1.05529225,
      "balance_loss_mlp": 1.02794981,
      "epoch": 0.150969487449271,
      "flos": 17928223073280.0,
      "grad_norm": 1.8628702139594306,
      "language_loss": 0.73398602,
      "learning_rate": 3.8484476280040495e-06,
      "loss": 0.75596589,
      "num_input_tokens_seen": 54503795,
      "step": 2511,
      "time_per_iteration": 2.62237811088562
    },
    {
      "auxiliary_loss_clip": 0.01094269,
      "auxiliary_loss_mlp": 0.0104236,
      "balance_loss_clip": 1.04747009,
      "balance_loss_mlp": 1.02365553,
      "epoch": 0.151029610701939,
      "flos": 24243078539520.0,
      "grad_norm": 2.20399257021602,
      "language_loss": 0.68716824,
      "learning_rate": 3.848298876546534e-06,
      "loss": 0.70853454,
      "num_input_tokens_seen": 54523025,
      "step": 2512,
      "time_per_iteration": 2.823359489440918
    },
    {
      "auxiliary_loss_clip": 0.01149398,
      "auxiliary_loss_mlp": 0.01043296,
      "balance_loss_clip": 1.05574036,
      "balance_loss_mlp": 1.02615356,
      "epoch": 0.15108973395460695,
      "flos": 30262496641920.0,
      "grad_norm": 2.6278607305338877,
      "language_loss": 0.73833561,
      "learning_rate": 3.84815005500134e-06,
      "loss": 0.76026255,
      "num_input_tokens_seen": 54545025,
      "step": 2513,
      "time_per_iteration": 4.386258602142334
    },
    {
      "auxiliary_loss_clip": 0.01059691,
      "auxiliary_loss_mlp": 0.01109321,
      "balance_loss_clip": 1.0685482,
      "balance_loss_mlp": 1.10529137,
      "epoch": 0.15114985720727492,
      "flos": 60437624428800.0,
      "grad_norm": 0.9017688875456507,
      "language_loss": 0.64720047,
      "learning_rate": 3.84800116337411e-06,
      "loss": 0.6688906,
      "num_input_tokens_seen": 54604545,
      "step": 2514,
      "time_per_iteration": 3.254983425140381
    },
    {
      "auxiliary_loss_clip": 0.01146323,
      "auxiliary_loss_mlp": 0.0104352,
      "balance_loss_clip": 1.05674648,
      "balance_loss_mlp": 1.02584124,
      "epoch": 0.15120998045994288,
      "flos": 20521691832960.0,
      "grad_norm": 3.178381755435586,
      "language_loss": 0.72995645,
      "learning_rate": 3.8478522016704916e-06,
      "loss": 0.7518549,
      "num_input_tokens_seen": 54620590,
      "step": 2515,
      "time_per_iteration": 2.67921781539917
    },
    {
      "auxiliary_loss_clip": 0.01133382,
      "auxiliary_loss_mlp": 0.01040315,
      "balance_loss_clip": 1.05675673,
      "balance_loss_mlp": 1.02120531,
      "epoch": 0.15127010371261085,
      "flos": 21178893024000.0,
      "grad_norm": 2.0712989062813243,
      "language_loss": 0.7773214,
      "learning_rate": 3.8477031698961325e-06,
      "loss": 0.79905832,
      "num_input_tokens_seen": 54640410,
      "step": 2516,
      "time_per_iteration": 2.763467788696289
    },
    {
      "auxiliary_loss_clip": 0.01087601,
      "auxiliary_loss_mlp": 0.01004779,
      "balance_loss_clip": 1.05344796,
      "balance_loss_mlp": 1.00160813,
      "epoch": 0.1513302269652788,
      "flos": 65320648974720.0,
      "grad_norm": 0.7270407819118658,
      "language_loss": 0.54622567,
      "learning_rate": 3.8475540680566835e-06,
      "loss": 0.56714946,
      "num_input_tokens_seen": 54701430,
      "step": 2517,
      "time_per_iteration": 3.2293660640716553
    },
    {
      "auxiliary_loss_clip": 0.01110142,
      "auxiliary_loss_mlp": 0.0104362,
      "balance_loss_clip": 1.04499209,
      "balance_loss_mlp": 1.02427244,
      "epoch": 0.15139035021794678,
      "flos": 19135827342720.0,
      "grad_norm": 3.035771526476276,
      "language_loss": 0.78264821,
      "learning_rate": 3.8474048961577995e-06,
      "loss": 0.80418587,
      "num_input_tokens_seen": 54720845,
      "step": 2518,
      "time_per_iteration": 2.8154754638671875
    },
    {
      "auxiliary_loss_clip": 0.01147342,
      "auxiliary_loss_mlp": 0.01056368,
      "balance_loss_clip": 1.05279088,
      "balance_loss_mlp": 1.03681803,
      "epoch": 0.15145047347061477,
      "flos": 26578564842240.0,
      "grad_norm": 2.1881526177791097,
      "language_loss": 0.70480245,
      "learning_rate": 3.847255654205137e-06,
      "loss": 0.72683954,
      "num_input_tokens_seen": 54740495,
      "step": 2519,
      "time_per_iteration": 2.7098515033721924
    },
    {
      "auxiliary_loss_clip": 0.01152463,
      "auxiliary_loss_mlp": 0.01056975,
      "balance_loss_clip": 1.05683672,
      "balance_loss_mlp": 1.03802037,
      "epoch": 0.15151059672328274,
      "flos": 20302959962880.0,
      "grad_norm": 1.9048594994100874,
      "language_loss": 0.78681207,
      "learning_rate": 3.847106342204354e-06,
      "loss": 0.80890644,
      "num_input_tokens_seen": 54758415,
      "step": 2520,
      "time_per_iteration": 2.664187431335449
    },
    {
      "auxiliary_loss_clip": 0.01140573,
      "auxiliary_loss_mlp": 0.01071607,
      "balance_loss_clip": 1.05435348,
      "balance_loss_mlp": 1.05244994,
      "epoch": 0.1515707199759507,
      "flos": 27228367831680.0,
      "grad_norm": 3.950911503454746,
      "language_loss": 0.74849677,
      "learning_rate": 3.846956960161114e-06,
      "loss": 0.77061862,
      "num_input_tokens_seen": 54779355,
      "step": 2521,
      "time_per_iteration": 2.7900772094726562
    },
    {
      "auxiliary_loss_clip": 0.01132038,
      "auxiliary_loss_mlp": 0.01055874,
      "balance_loss_clip": 1.05052209,
      "balance_loss_mlp": 1.0360136,
      "epoch": 0.15163084322861867,
      "flos": 23587349806080.0,
      "grad_norm": 4.620979243079986,
      "language_loss": 0.8253814,
      "learning_rate": 3.84680750808108e-06,
      "loss": 0.84726053,
      "num_input_tokens_seen": 54799465,
      "step": 2522,
      "time_per_iteration": 2.7216525077819824
    },
    {
      "auxiliary_loss_clip": 0.01051858,
      "auxiliary_loss_mlp": 0.01048797,
      "balance_loss_clip": 1.05645704,
      "balance_loss_mlp": 1.04595995,
      "epoch": 0.15169096648128663,
      "flos": 66889622021760.0,
      "grad_norm": 0.8362305181264502,
      "language_loss": 0.57885599,
      "learning_rate": 3.846657985969922e-06,
      "loss": 0.59986252,
      "num_input_tokens_seen": 54857665,
      "step": 2523,
      "time_per_iteration": 3.2375056743621826
    },
    {
      "auxiliary_loss_clip": 0.0114147,
      "auxiliary_loss_mlp": 0.01057964,
      "balance_loss_clip": 1.05213499,
      "balance_loss_mlp": 1.0368042,
      "epoch": 0.1517510897339546,
      "flos": 29095435848960.0,
      "grad_norm": 1.8054087157705183,
      "language_loss": 0.74795163,
      "learning_rate": 3.8465083938333066e-06,
      "loss": 0.76994598,
      "num_input_tokens_seen": 54879895,
      "step": 2524,
      "time_per_iteration": 2.711557388305664
    },
    {
      "auxiliary_loss_clip": 0.01138185,
      "auxiliary_loss_mlp": 0.01057236,
      "balance_loss_clip": 1.05304718,
      "balance_loss_mlp": 1.03865099,
      "epoch": 0.1518112129866226,
      "flos": 18406553512320.0,
      "grad_norm": 1.8255227790100423,
      "language_loss": 0.74631184,
      "learning_rate": 3.8463587316769085e-06,
      "loss": 0.76826608,
      "num_input_tokens_seen": 54898245,
      "step": 2525,
      "time_per_iteration": 2.6936984062194824
    },
    {
      "auxiliary_loss_clip": 0.01144047,
      "auxiliary_loss_mlp": 0.01057009,
      "balance_loss_clip": 1.05403006,
      "balance_loss_mlp": 1.03747034,
      "epoch": 0.15187133623929056,
      "flos": 19425410789760.0,
      "grad_norm": 1.8907352833287865,
      "language_loss": 0.79600316,
      "learning_rate": 3.846208999506402e-06,
      "loss": 0.81801373,
      "num_input_tokens_seen": 54917060,
      "step": 2526,
      "time_per_iteration": 2.651494264602661
    },
    {
      "auxiliary_loss_clip": 0.01135228,
      "auxiliary_loss_mlp": 0.01047798,
      "balance_loss_clip": 1.05538774,
      "balance_loss_mlp": 1.03056002,
      "epoch": 0.15193145949195852,
      "flos": 17566207850880.0,
      "grad_norm": 1.7677336965262924,
      "language_loss": 0.8443349,
      "learning_rate": 3.846059197327466e-06,
      "loss": 0.86616516,
      "num_input_tokens_seen": 54936365,
      "step": 2527,
      "time_per_iteration": 2.702683448791504
    },
    {
      "auxiliary_loss_clip": 0.01124925,
      "auxiliary_loss_mlp": 0.01049207,
      "balance_loss_clip": 1.04976487,
      "balance_loss_mlp": 1.02985954,
      "epoch": 0.15199158274462649,
      "flos": 36176265866880.0,
      "grad_norm": 1.85678489681458,
      "language_loss": 0.69361663,
      "learning_rate": 3.845909325145779e-06,
      "loss": 0.7153579,
      "num_input_tokens_seen": 54961365,
      "step": 2528,
      "time_per_iteration": 2.9250690937042236
    },
    {
      "auxiliary_loss_clip": 0.01134092,
      "auxiliary_loss_mlp": 0.01055056,
      "balance_loss_clip": 1.05266535,
      "balance_loss_mlp": 1.03587484,
      "epoch": 0.15205170599729445,
      "flos": 23074042498560.0,
      "grad_norm": 2.004144148858156,
      "language_loss": 0.86482549,
      "learning_rate": 3.845759382967026e-06,
      "loss": 0.88671696,
      "num_input_tokens_seen": 54980750,
      "step": 2529,
      "time_per_iteration": 2.7277863025665283
    },
    {
      "auxiliary_loss_clip": 0.01124798,
      "auxiliary_loss_mlp": 0.01041651,
      "balance_loss_clip": 1.05046487,
      "balance_loss_mlp": 1.02297091,
      "epoch": 0.15211182924996242,
      "flos": 21908382336000.0,
      "grad_norm": 2.544775548600603,
      "language_loss": 0.83399373,
      "learning_rate": 3.845609370796893e-06,
      "loss": 0.85565823,
      "num_input_tokens_seen": 54999675,
      "step": 2530,
      "time_per_iteration": 2.8717291355133057
    },
    {
      "auxiliary_loss_clip": 0.01125761,
      "auxiliary_loss_mlp": 0.01048121,
      "balance_loss_clip": 1.05035281,
      "balance_loss_mlp": 1.02940559,
      "epoch": 0.15217195250263038,
      "flos": 13881521865600.0,
      "grad_norm": 2.1410437006568723,
      "language_loss": 0.80404246,
      "learning_rate": 3.845459288641066e-06,
      "loss": 0.82578129,
      "num_input_tokens_seen": 55018295,
      "step": 2531,
      "time_per_iteration": 2.8444995880126953
    },
    {
      "auxiliary_loss_clip": 0.01143114,
      "auxiliary_loss_mlp": 0.01043494,
      "balance_loss_clip": 1.05216551,
      "balance_loss_mlp": 1.02613723,
      "epoch": 0.15223207575529837,
      "flos": 24535319592960.0,
      "grad_norm": 1.7922494378130023,
      "language_loss": 0.78874445,
      "learning_rate": 3.8453091365052394e-06,
      "loss": 0.81061059,
      "num_input_tokens_seen": 55037975,
      "step": 2532,
      "time_per_iteration": 2.9122390747070312
    },
    {
      "auxiliary_loss_clip": 0.01149502,
      "auxiliary_loss_mlp": 0.0104596,
      "balance_loss_clip": 1.05737543,
      "balance_loss_mlp": 1.02676702,
      "epoch": 0.15229219900796634,
      "flos": 25556798563200.0,
      "grad_norm": 1.9533698136575197,
      "language_loss": 0.87679356,
      "learning_rate": 3.845158914395105e-06,
      "loss": 0.89874816,
      "num_input_tokens_seen": 55057135,
      "step": 2533,
      "time_per_iteration": 2.7987985610961914
    },
    {
      "auxiliary_loss_clip": 0.01117955,
      "auxiliary_loss_mlp": 0.01048672,
      "balance_loss_clip": 1.05235386,
      "balance_loss_mlp": 1.02983665,
      "epoch": 0.1523523222606343,
      "flos": 18217806520320.0,
      "grad_norm": 2.391026063452041,
      "language_loss": 0.78886449,
      "learning_rate": 3.84500862231636e-06,
      "loss": 0.81053078,
      "num_input_tokens_seen": 55075525,
      "step": 2534,
      "time_per_iteration": 2.7587406635284424
    },
    {
      "auxiliary_loss_clip": 0.01164218,
      "auxiliary_loss_mlp": 0.0104722,
      "balance_loss_clip": 1.05609345,
      "balance_loss_mlp": 1.0270381,
      "epoch": 0.15241244551330227,
      "flos": 13260087642240.0,
      "grad_norm": 2.689732363294508,
      "language_loss": 0.76809752,
      "learning_rate": 3.844858260274702e-06,
      "loss": 0.79021192,
      "num_input_tokens_seen": 55090845,
      "step": 2535,
      "time_per_iteration": 2.7494406700134277
    },
    {
      "auxiliary_loss_clip": 0.01142628,
      "auxiliary_loss_mlp": 0.01042905,
      "balance_loss_clip": 1.05345285,
      "balance_loss_mlp": 1.02401042,
      "epoch": 0.15247256876597023,
      "flos": 19715568854400.0,
      "grad_norm": 2.2235871255319446,
      "language_loss": 0.78301942,
      "learning_rate": 3.844707828275835e-06,
      "loss": 0.80487478,
      "num_input_tokens_seen": 55108750,
      "step": 2536,
      "time_per_iteration": 2.738638401031494
    },
    {
      "auxiliary_loss_clip": 0.01128919,
      "auxiliary_loss_mlp": 0.0105368,
      "balance_loss_clip": 1.05349088,
      "balance_loss_mlp": 1.03497589,
      "epoch": 0.1525326920186382,
      "flos": 20375858615040.0,
      "grad_norm": 2.311649941233105,
      "language_loss": 0.75824189,
      "learning_rate": 3.844557326325461e-06,
      "loss": 0.78006792,
      "num_input_tokens_seen": 55126750,
      "step": 2537,
      "time_per_iteration": 2.632373809814453
    },
    {
      "auxiliary_loss_clip": 0.0114911,
      "auxiliary_loss_mlp": 0.01041421,
      "balance_loss_clip": 1.05675745,
      "balance_loss_mlp": 1.02331281,
      "epoch": 0.15259281527130616,
      "flos": 13589963170560.0,
      "grad_norm": 2.193148723631548,
      "language_loss": 0.77737647,
      "learning_rate": 3.8444067544292896e-06,
      "loss": 0.79928178,
      "num_input_tokens_seen": 55144690,
      "step": 2538,
      "time_per_iteration": 2.6835639476776123
    },
    {
      "auxiliary_loss_clip": 0.01109367,
      "auxiliary_loss_mlp": 0.01042256,
      "balance_loss_clip": 1.05477905,
      "balance_loss_mlp": 1.02480412,
      "epoch": 0.15265293852397416,
      "flos": 22860374446080.0,
      "grad_norm": 2.951423477379744,
      "language_loss": 0.89502335,
      "learning_rate": 3.844256112593029e-06,
      "loss": 0.91653961,
      "num_input_tokens_seen": 55166055,
      "step": 2539,
      "time_per_iteration": 2.7825794219970703
    },
    {
      "auxiliary_loss_clip": 0.01142581,
      "auxiliary_loss_mlp": 0.01045856,
      "balance_loss_clip": 1.05367279,
      "balance_loss_mlp": 1.02721143,
      "epoch": 0.15271306177664212,
      "flos": 29238108670080.0,
      "grad_norm": 2.1073423273657044,
      "language_loss": 0.93423879,
      "learning_rate": 3.844105400822391e-06,
      "loss": 0.95612311,
      "num_input_tokens_seen": 55186285,
      "step": 2540,
      "time_per_iteration": 2.717541456222534
    },
    {
      "auxiliary_loss_clip": 0.01131603,
      "auxiliary_loss_mlp": 0.01041863,
      "balance_loss_clip": 1.05122495,
      "balance_loss_mlp": 1.0240885,
      "epoch": 0.1527731850293101,
      "flos": 31246269310080.0,
      "grad_norm": 2.084754505375857,
      "language_loss": 0.75217843,
      "learning_rate": 3.843954619123092e-06,
      "loss": 0.77391309,
      "num_input_tokens_seen": 55207915,
      "step": 2541,
      "time_per_iteration": 2.8376123905181885
    },
    {
      "auxiliary_loss_clip": 0.01116303,
      "auxiliary_loss_mlp": 0.01045227,
      "balance_loss_clip": 1.04877007,
      "balance_loss_mlp": 1.0268805,
      "epoch": 0.15283330828197805,
      "flos": 22382079920640.0,
      "grad_norm": 2.037290364787748,
      "language_loss": 0.80996066,
      "learning_rate": 3.84380376750085e-06,
      "loss": 0.83157599,
      "num_input_tokens_seen": 55227860,
      "step": 2542,
      "time_per_iteration": 2.7110376358032227
    },
    {
      "auxiliary_loss_clip": 0.01160331,
      "auxiliary_loss_mlp": 0.01048661,
      "balance_loss_clip": 1.0566076,
      "balance_loss_mlp": 1.02992105,
      "epoch": 0.15289343153464602,
      "flos": 25520133755520.0,
      "grad_norm": 3.2152362880248857,
      "language_loss": 0.77796149,
      "learning_rate": 3.843652845961383e-06,
      "loss": 0.80005145,
      "num_input_tokens_seen": 55247330,
      "step": 2543,
      "time_per_iteration": 2.674131155014038
    },
    {
      "auxiliary_loss_clip": 0.01145565,
      "auxiliary_loss_mlp": 0.01042133,
      "balance_loss_clip": 1.05380869,
      "balance_loss_mlp": 1.02388239,
      "epoch": 0.15295355478731398,
      "flos": 22710016114560.0,
      "grad_norm": 2.4890924021550918,
      "language_loss": 0.85898137,
      "learning_rate": 3.843501854510416e-06,
      "loss": 0.88085836,
      "num_input_tokens_seen": 55266195,
      "step": 2544,
      "time_per_iteration": 2.685840606689453
    },
    {
      "auxiliary_loss_clip": 0.01149904,
      "auxiliary_loss_mlp": 0.01051141,
      "balance_loss_clip": 1.05162692,
      "balance_loss_mlp": 1.03061318,
      "epoch": 0.15301367803998198,
      "flos": 23251907669760.0,
      "grad_norm": 1.9817931887295275,
      "language_loss": 0.83159137,
      "learning_rate": 3.843350793153673e-06,
      "loss": 0.85360181,
      "num_input_tokens_seen": 55283305,
      "step": 2545,
      "time_per_iteration": 2.7415812015533447
    },
    {
      "auxiliary_loss_clip": 0.01158976,
      "auxiliary_loss_mlp": 0.01040888,
      "balance_loss_clip": 1.05556524,
      "balance_loss_mlp": 1.02257705,
      "epoch": 0.15307380129264994,
      "flos": 25886279041920.0,
      "grad_norm": 6.0131413628182,
      "language_loss": 0.71669161,
      "learning_rate": 3.843199661896884e-06,
      "loss": 0.73869026,
      "num_input_tokens_seen": 55303035,
      "step": 2546,
      "time_per_iteration": 2.6626265048980713
    },
    {
      "auxiliary_loss_clip": 0.01130357,
      "auxiliary_loss_mlp": 0.01047635,
      "balance_loss_clip": 1.05013335,
      "balance_loss_mlp": 1.02688098,
      "epoch": 0.1531339245453179,
      "flos": 46973239205760.0,
      "grad_norm": 1.6563553629779504,
      "language_loss": 0.77438712,
      "learning_rate": 3.843048460745779e-06,
      "loss": 0.79616702,
      "num_input_tokens_seen": 55327570,
      "step": 2547,
      "time_per_iteration": 4.451423168182373
    },
    {
      "auxiliary_loss_clip": 0.01107553,
      "auxiliary_loss_mlp": 0.01044692,
      "balance_loss_clip": 1.04845536,
      "balance_loss_mlp": 1.02517736,
      "epoch": 0.15319404779798587,
      "flos": 35882049565440.0,
      "grad_norm": 2.3544675813743834,
      "language_loss": 0.74357474,
      "learning_rate": 3.842897189706092e-06,
      "loss": 0.7650972,
      "num_input_tokens_seen": 55351090,
      "step": 2548,
      "time_per_iteration": 2.846991539001465
    },
    {
      "auxiliary_loss_clip": 0.01138346,
      "auxiliary_loss_mlp": 0.0105294,
      "balance_loss_clip": 1.05340147,
      "balance_loss_mlp": 1.03304434,
      "epoch": 0.15325417105065384,
      "flos": 25664638170240.0,
      "grad_norm": 1.446042531021912,
      "language_loss": 0.80296385,
      "learning_rate": 3.842745848783558e-06,
      "loss": 0.82487667,
      "num_input_tokens_seen": 55371050,
      "step": 2549,
      "time_per_iteration": 5.8849101066589355
    },
    {
      "auxiliary_loss_clip": 0.01144858,
      "auxiliary_loss_mlp": 0.01041292,
      "balance_loss_clip": 1.05108786,
      "balance_loss_mlp": 1.02255249,
      "epoch": 0.1533142943033218,
      "flos": 18770831291520.0,
      "grad_norm": 1.6149920159034452,
      "language_loss": 0.74602014,
      "learning_rate": 3.842594437983917e-06,
      "loss": 0.76788169,
      "num_input_tokens_seen": 55390375,
      "step": 2550,
      "time_per_iteration": 2.684868812561035
    },
    {
      "auxiliary_loss_clip": 0.01149823,
      "auxiliary_loss_mlp": 0.01040743,
      "balance_loss_clip": 1.05212283,
      "balance_loss_mlp": 1.02129996,
      "epoch": 0.15337441755598977,
      "flos": 23107367341440.0,
      "grad_norm": 2.33086854575276,
      "language_loss": 0.76910275,
      "learning_rate": 3.8424429573129115e-06,
      "loss": 0.79100841,
      "num_input_tokens_seen": 55408890,
      "step": 2551,
      "time_per_iteration": 4.415414333343506
    },
    {
      "auxiliary_loss_clip": 0.01086721,
      "auxiliary_loss_mlp": 0.01054065,
      "balance_loss_clip": 1.05333817,
      "balance_loss_mlp": 1.05116868,
      "epoch": 0.15343454080865776,
      "flos": 59861079227520.0,
      "grad_norm": 0.9493148205555214,
      "language_loss": 0.5665558,
      "learning_rate": 3.842291406776283e-06,
      "loss": 0.5879637,
      "num_input_tokens_seen": 55463815,
      "step": 2552,
      "time_per_iteration": 3.1105730533599854
    },
    {
      "auxiliary_loss_clip": 0.011128,
      "auxiliary_loss_mlp": 0.01039619,
      "balance_loss_clip": 1.05131924,
      "balance_loss_mlp": 1.0204618,
      "epoch": 0.15349466406132573,
      "flos": 11910887959680.0,
      "grad_norm": 2.183188616823757,
      "language_loss": 0.88550794,
      "learning_rate": 3.84213978637978e-06,
      "loss": 0.90703207,
      "num_input_tokens_seen": 55481050,
      "step": 2553,
      "time_per_iteration": 2.748298406600952
    },
    {
      "auxiliary_loss_clip": 0.01147024,
      "auxiliary_loss_mlp": 0.01042929,
      "balance_loss_clip": 1.05247378,
      "balance_loss_mlp": 1.0232954,
      "epoch": 0.1535547873139937,
      "flos": 24096922099200.0,
      "grad_norm": 1.8094820084348213,
      "language_loss": 0.7800495,
      "learning_rate": 3.841988096129152e-06,
      "loss": 0.80194902,
      "num_input_tokens_seen": 55500050,
      "step": 2554,
      "time_per_iteration": 2.6555569171905518
    },
    {
      "auxiliary_loss_clip": 0.01094445,
      "auxiliary_loss_mlp": 0.01053684,
      "balance_loss_clip": 1.04876757,
      "balance_loss_mlp": 1.03291798,
      "epoch": 0.15361491056666166,
      "flos": 17566459246080.0,
      "grad_norm": 2.372022486587551,
      "language_loss": 0.77472258,
      "learning_rate": 3.841836336030151e-06,
      "loss": 0.79620385,
      "num_input_tokens_seen": 55518125,
      "step": 2555,
      "time_per_iteration": 2.7507212162017822
    },
    {
      "auxiliary_loss_clip": 0.01129555,
      "auxiliary_loss_mlp": 0.01046723,
      "balance_loss_clip": 1.05400753,
      "balance_loss_mlp": 1.02873409,
      "epoch": 0.15367503381932962,
      "flos": 25046041121280.0,
      "grad_norm": 1.5517643759455655,
      "language_loss": 0.77453947,
      "learning_rate": 3.8416845060885305e-06,
      "loss": 0.79630232,
      "num_input_tokens_seen": 55540960,
      "step": 2556,
      "time_per_iteration": 2.7947654724121094
    },
    {
      "auxiliary_loss_clip": 0.01140725,
      "auxiliary_loss_mlp": 0.0077646,
      "balance_loss_clip": 1.05336452,
      "balance_loss_mlp": 1.00054574,
      "epoch": 0.15373515707199759,
      "flos": 21507332008320.0,
      "grad_norm": 1.8786460244833383,
      "language_loss": 0.90098578,
      "learning_rate": 3.84153260631005e-06,
      "loss": 0.92015761,
      "num_input_tokens_seen": 55559210,
      "step": 2557,
      "time_per_iteration": 2.702029228210449
    },
    {
      "auxiliary_loss_clip": 0.01137441,
      "auxiliary_loss_mlp": 0.01048546,
      "balance_loss_clip": 1.05146766,
      "balance_loss_mlp": 1.02862656,
      "epoch": 0.15379528032466555,
      "flos": 25994729180160.0,
      "grad_norm": 2.4046585493240102,
      "language_loss": 0.7092281,
      "learning_rate": 3.841380636700468e-06,
      "loss": 0.73108798,
      "num_input_tokens_seen": 55578925,
      "step": 2558,
      "time_per_iteration": 2.815653085708618
    },
    {
      "auxiliary_loss_clip": 0.01131603,
      "auxiliary_loss_mlp": 0.01045983,
      "balance_loss_clip": 1.04937947,
      "balance_loss_mlp": 1.02659965,
      "epoch": 0.15385540357733354,
      "flos": 19277315015040.0,
      "grad_norm": 2.1050139676488535,
      "language_loss": 0.92165422,
      "learning_rate": 3.841228597265548e-06,
      "loss": 0.94343007,
      "num_input_tokens_seen": 55597255,
      "step": 2559,
      "time_per_iteration": 2.7363967895507812
    },
    {
      "auxiliary_loss_clip": 0.011375,
      "auxiliary_loss_mlp": 0.01057878,
      "balance_loss_clip": 1.05492043,
      "balance_loss_mlp": 1.03711152,
      "epoch": 0.1539155268300015,
      "flos": 28549126920960.0,
      "grad_norm": 2.149412909113977,
      "language_loss": 0.63330692,
      "learning_rate": 3.841076488011055e-06,
      "loss": 0.65526068,
      "num_input_tokens_seen": 55619515,
      "step": 2560,
      "time_per_iteration": 2.811800003051758
    },
    {
      "auxiliary_loss_clip": 0.01132154,
      "auxiliary_loss_mlp": 0.01043974,
      "balance_loss_clip": 1.04914606,
      "balance_loss_mlp": 1.02416182,
      "epoch": 0.15397565008266947,
      "flos": 23547883737600.0,
      "grad_norm": 2.066473237183783,
      "language_loss": 0.88155699,
      "learning_rate": 3.8409243089427574e-06,
      "loss": 0.90331829,
      "num_input_tokens_seen": 55640050,
      "step": 2561,
      "time_per_iteration": 2.7991089820861816
    },
    {
      "auxiliary_loss_clip": 0.0114054,
      "auxiliary_loss_mlp": 0.01041879,
      "balance_loss_clip": 1.05085099,
      "balance_loss_mlp": 1.02380693,
      "epoch": 0.15403577333533744,
      "flos": 17129821518720.0,
      "grad_norm": 1.906051405357337,
      "language_loss": 0.83117974,
      "learning_rate": 3.840772060066425e-06,
      "loss": 0.85300398,
      "num_input_tokens_seen": 55658695,
      "step": 2562,
      "time_per_iteration": 2.6410810947418213
    },
    {
      "auxiliary_loss_clip": 0.01128756,
      "auxiliary_loss_mlp": 0.00778205,
      "balance_loss_clip": 1.04988563,
      "balance_loss_mlp": 1.00058532,
      "epoch": 0.1540958965880054,
      "flos": 17894503180800.0,
      "grad_norm": 2.3547297997270906,
      "language_loss": 0.74647415,
      "learning_rate": 3.840619741387832e-06,
      "loss": 0.76554382,
      "num_input_tokens_seen": 55676340,
      "step": 2563,
      "time_per_iteration": 2.6813745498657227
    },
    {
      "auxiliary_loss_clip": 0.01116857,
      "auxiliary_loss_mlp": 0.0104411,
      "balance_loss_clip": 1.05126941,
      "balance_loss_mlp": 1.02444029,
      "epoch": 0.15415601984067337,
      "flos": 32161057908480.0,
      "grad_norm": 2.842824767177756,
      "language_loss": 0.7609179,
      "learning_rate": 3.8404673529127534e-06,
      "loss": 0.78252757,
      "num_input_tokens_seen": 55698890,
      "step": 2564,
      "time_per_iteration": 2.832885265350342
    },
    {
      "auxiliary_loss_clip": 0.01133461,
      "auxiliary_loss_mlp": 0.01052887,
      "balance_loss_clip": 1.05174518,
      "balance_loss_mlp": 1.03443313,
      "epoch": 0.15421614309334136,
      "flos": 24024418496640.0,
      "grad_norm": 2.0125869911748575,
      "language_loss": 0.70960921,
      "learning_rate": 3.840314894646969e-06,
      "loss": 0.73147273,
      "num_input_tokens_seen": 55718535,
      "step": 2565,
      "time_per_iteration": 2.7352514266967773
    },
    {
      "auxiliary_loss_clip": 0.01137766,
      "auxiliary_loss_mlp": 0.01046908,
      "balance_loss_clip": 1.04731965,
      "balance_loss_mlp": 1.02787066,
      "epoch": 0.15427626634600933,
      "flos": 24386290064640.0,
      "grad_norm": 2.1021891280826965,
      "language_loss": 0.71605748,
      "learning_rate": 3.840162366596259e-06,
      "loss": 0.73790431,
      "num_input_tokens_seen": 55738970,
      "step": 2566,
      "time_per_iteration": 2.681710720062256
    },
    {
      "auxiliary_loss_clip": 0.01150619,
      "auxiliary_loss_mlp": 0.01040725,
      "balance_loss_clip": 1.04834008,
      "balance_loss_mlp": 1.02271223,
      "epoch": 0.1543363895986773,
      "flos": 23331522165120.0,
      "grad_norm": 1.7167104030167524,
      "language_loss": 0.84746087,
      "learning_rate": 3.840009768766408e-06,
      "loss": 0.86937428,
      "num_input_tokens_seen": 55759585,
      "step": 2567,
      "time_per_iteration": 2.6413686275482178
    },
    {
      "auxiliary_loss_clip": 0.01104646,
      "auxiliary_loss_mlp": 0.01050344,
      "balance_loss_clip": 1.04447246,
      "balance_loss_mlp": 1.03164053,
      "epoch": 0.15439651285134526,
      "flos": 24274284480000.0,
      "grad_norm": 2.9101336164483014,
      "language_loss": 0.78074998,
      "learning_rate": 3.839857101163202e-06,
      "loss": 0.80229992,
      "num_input_tokens_seen": 55779250,
      "step": 2568,
      "time_per_iteration": 2.7385261058807373
    },
    {
      "auxiliary_loss_clip": 0.01121993,
      "auxiliary_loss_mlp": 0.01037084,
      "balance_loss_clip": 1.04715753,
      "balance_loss_mlp": 1.01684201,
      "epoch": 0.15445663610401322,
      "flos": 22456163721600.0,
      "grad_norm": 1.852436867559063,
      "language_loss": 0.6991998,
      "learning_rate": 3.83970436379243e-06,
      "loss": 0.72079051,
      "num_input_tokens_seen": 55800470,
      "step": 2569,
      "time_per_iteration": 2.746974229812622
    },
    {
      "auxiliary_loss_clip": 0.01124209,
      "auxiliary_loss_mlp": 0.01040299,
      "balance_loss_clip": 1.04695952,
      "balance_loss_mlp": 1.02178574,
      "epoch": 0.1545167593566812,
      "flos": 22049510872320.0,
      "grad_norm": 1.7212875994527412,
      "language_loss": 0.76482332,
      "learning_rate": 3.839551556659884e-06,
      "loss": 0.78646845,
      "num_input_tokens_seen": 55817795,
      "step": 2570,
      "time_per_iteration": 2.7470619678497314
    },
    {
      "auxiliary_loss_clip": 0.01137702,
      "auxiliary_loss_mlp": 0.01038561,
      "balance_loss_clip": 1.04993737,
      "balance_loss_mlp": 1.0192852,
      "epoch": 0.15457688260934915,
      "flos": 19318253541120.0,
      "grad_norm": 2.5033166184578066,
      "language_loss": 0.77997506,
      "learning_rate": 3.839398679771359e-06,
      "loss": 0.80173767,
      "num_input_tokens_seen": 55836125,
      "step": 2571,
      "time_per_iteration": 2.692863702774048
    },
    {
      "auxiliary_loss_clip": 0.0113208,
      "auxiliary_loss_mlp": 0.0104519,
      "balance_loss_clip": 1.0498451,
      "balance_loss_mlp": 1.02704597,
      "epoch": 0.15463700586201715,
      "flos": 24133981956480.0,
      "grad_norm": 4.3242380509309015,
      "language_loss": 0.82932413,
      "learning_rate": 3.839245733132652e-06,
      "loss": 0.85109681,
      "num_input_tokens_seen": 55855280,
      "step": 2572,
      "time_per_iteration": 2.8341822624206543
    },
    {
      "auxiliary_loss_clip": 0.01156188,
      "auxiliary_loss_mlp": 0.01042592,
      "balance_loss_clip": 1.05181205,
      "balance_loss_mlp": 1.02383995,
      "epoch": 0.1546971291146851,
      "flos": 22420935457920.0,
      "grad_norm": 1.5874704718869805,
      "language_loss": 0.90373385,
      "learning_rate": 3.839092716749563e-06,
      "loss": 0.92572165,
      "num_input_tokens_seen": 55875695,
      "step": 2573,
      "time_per_iteration": 2.740121364593506
    },
    {
      "auxiliary_loss_clip": 0.01088424,
      "auxiliary_loss_mlp": 0.01049893,
      "balance_loss_clip": 1.04328668,
      "balance_loss_mlp": 1.03003311,
      "epoch": 0.15475725236735308,
      "flos": 17530225401600.0,
      "grad_norm": 1.596795561637076,
      "language_loss": 0.70298707,
      "learning_rate": 3.838939630627893e-06,
      "loss": 0.72437024,
      "num_input_tokens_seen": 55894575,
      "step": 2574,
      "time_per_iteration": 2.7629144191741943
    },
    {
      "auxiliary_loss_clip": 0.01127537,
      "auxiliary_loss_mlp": 0.01045732,
      "balance_loss_clip": 1.04714394,
      "balance_loss_mlp": 1.02509642,
      "epoch": 0.15481737562002104,
      "flos": 22561740771840.0,
      "grad_norm": 6.018921028505516,
      "language_loss": 0.82426423,
      "learning_rate": 3.838786474773448e-06,
      "loss": 0.84599686,
      "num_input_tokens_seen": 55912855,
      "step": 2575,
      "time_per_iteration": 2.656783103942871
    },
    {
      "auxiliary_loss_clip": 0.01127415,
      "auxiliary_loss_mlp": 0.01043354,
      "balance_loss_clip": 1.04681587,
      "balance_loss_mlp": 1.02584219,
      "epoch": 0.154877498872689,
      "flos": 24900567039360.0,
      "grad_norm": 1.8376318938002576,
      "language_loss": 0.85038638,
      "learning_rate": 3.838633249192036e-06,
      "loss": 0.87209404,
      "num_input_tokens_seen": 55932375,
      "step": 2576,
      "time_per_iteration": 2.648484230041504
    },
    {
      "auxiliary_loss_clip": 0.01152547,
      "auxiliary_loss_mlp": 0.01043401,
      "balance_loss_clip": 1.04872847,
      "balance_loss_mlp": 1.02499545,
      "epoch": 0.15493762212535697,
      "flos": 28147501975680.0,
      "grad_norm": 1.8027999188827728,
      "language_loss": 0.82271254,
      "learning_rate": 3.838479953889465e-06,
      "loss": 0.84467208,
      "num_input_tokens_seen": 55953970,
      "step": 2577,
      "time_per_iteration": 2.6355643272399902
    },
    {
      "auxiliary_loss_clip": 0.01126009,
      "auxiliary_loss_mlp": 0.01049018,
      "balance_loss_clip": 1.05147958,
      "balance_loss_mlp": 1.02984881,
      "epoch": 0.15499774537802496,
      "flos": 25411073086080.0,
      "grad_norm": 2.1677069711314463,
      "language_loss": 0.76556361,
      "learning_rate": 3.8383265888715525e-06,
      "loss": 0.78731394,
      "num_input_tokens_seen": 55973120,
      "step": 2578,
      "time_per_iteration": 2.649043560028076
    },
    {
      "auxiliary_loss_clip": 0.01123677,
      "auxiliary_loss_mlp": 0.01044461,
      "balance_loss_clip": 1.05155993,
      "balance_loss_mlp": 1.0253042,
      "epoch": 0.15505786863069293,
      "flos": 22091562720000.0,
      "grad_norm": 1.9614380224881987,
      "language_loss": 0.82443559,
      "learning_rate": 3.83817315414411e-06,
      "loss": 0.8461169,
      "num_input_tokens_seen": 55993260,
      "step": 2579,
      "time_per_iteration": 2.62631893157959
    },
    {
      "auxiliary_loss_clip": 0.01143904,
      "auxiliary_loss_mlp": 0.01044324,
      "balance_loss_clip": 1.05856657,
      "balance_loss_mlp": 1.02556014,
      "epoch": 0.1551179918833609,
      "flos": 18917131386240.0,
      "grad_norm": 2.610374735790095,
      "language_loss": 0.80465376,
      "learning_rate": 3.838019649712958e-06,
      "loss": 0.82653606,
      "num_input_tokens_seen": 56012130,
      "step": 2580,
      "time_per_iteration": 2.6512253284454346
    },
    {
      "auxiliary_loss_clip": 0.0107737,
      "auxiliary_loss_mlp": 0.01006304,
      "balance_loss_clip": 1.04551053,
      "balance_loss_mlp": 1.00360954,
      "epoch": 0.15517811513602886,
      "flos": 66239172587520.0,
      "grad_norm": 0.842131683094019,
      "language_loss": 0.58823448,
      "learning_rate": 3.8378660755839166e-06,
      "loss": 0.60907125,
      "num_input_tokens_seen": 56079045,
      "step": 2581,
      "time_per_iteration": 3.357855796813965
    },
    {
      "auxiliary_loss_clip": 0.01108206,
      "auxiliary_loss_mlp": 0.01047031,
      "balance_loss_clip": 1.04392648,
      "balance_loss_mlp": 1.0249418,
      "epoch": 0.15523823838869683,
      "flos": 24021078531840.0,
      "grad_norm": 1.9584677228939371,
      "language_loss": 0.84773678,
      "learning_rate": 3.8377124317628095e-06,
      "loss": 0.86928916,
      "num_input_tokens_seen": 56098745,
      "step": 2582,
      "time_per_iteration": 2.727062702178955
    },
    {
      "auxiliary_loss_clip": 0.01144131,
      "auxiliary_loss_mlp": 0.01051911,
      "balance_loss_clip": 1.05233002,
      "balance_loss_mlp": 1.03175235,
      "epoch": 0.1552983616413648,
      "flos": 20485062938880.0,
      "grad_norm": 2.466663791870015,
      "language_loss": 0.79050052,
      "learning_rate": 3.8375587182554625e-06,
      "loss": 0.81246096,
      "num_input_tokens_seen": 56117655,
      "step": 2583,
      "time_per_iteration": 2.664794683456421
    },
    {
      "auxiliary_loss_clip": 0.01139818,
      "auxiliary_loss_mlp": 0.01054771,
      "balance_loss_clip": 1.04957032,
      "balance_loss_mlp": 1.03252697,
      "epoch": 0.15535848489403276,
      "flos": 32123710742400.0,
      "grad_norm": 1.8743170599575527,
      "language_loss": 0.76320136,
      "learning_rate": 3.837404935067705e-06,
      "loss": 0.78514719,
      "num_input_tokens_seen": 56141960,
      "step": 2584,
      "time_per_iteration": 2.757392168045044
    },
    {
      "auxiliary_loss_clip": 0.01137324,
      "auxiliary_loss_mlp": 0.01042496,
      "balance_loss_clip": 1.04884958,
      "balance_loss_mlp": 1.02302885,
      "epoch": 0.15541860814670075,
      "flos": 19098444263040.0,
      "grad_norm": 1.6493041410587026,
      "language_loss": 0.75269651,
      "learning_rate": 3.837251082205368e-06,
      "loss": 0.77449471,
      "num_input_tokens_seen": 56161430,
      "step": 2585,
      "time_per_iteration": 2.6497461795806885
    },
    {
      "auxiliary_loss_clip": 0.01116144,
      "auxiliary_loss_mlp": 0.01042356,
      "balance_loss_clip": 1.04862189,
      "balance_loss_mlp": 1.02321053,
      "epoch": 0.1554787313993687,
      "flos": 19172097100800.0,
      "grad_norm": 2.068989677221064,
      "language_loss": 0.61187196,
      "learning_rate": 3.837097159674286e-06,
      "loss": 0.63345695,
      "num_input_tokens_seen": 56179390,
      "step": 2586,
      "time_per_iteration": 2.697852373123169
    },
    {
      "auxiliary_loss_clip": 0.01129408,
      "auxiliary_loss_mlp": 0.01042187,
      "balance_loss_clip": 1.04842281,
      "balance_loss_mlp": 1.02341127,
      "epoch": 0.15553885465203668,
      "flos": 16143822207360.0,
      "grad_norm": 1.8484108176722505,
      "language_loss": 0.81318939,
      "learning_rate": 3.836943167480296e-06,
      "loss": 0.83490539,
      "num_input_tokens_seen": 56198020,
      "step": 2587,
      "time_per_iteration": 4.212551593780518
    },
    {
      "auxiliary_loss_clip": 0.01160891,
      "auxiliary_loss_mlp": 0.01054822,
      "balance_loss_clip": 1.05309868,
      "balance_loss_mlp": 1.03325701,
      "epoch": 0.15559897790470464,
      "flos": 25337779384320.0,
      "grad_norm": 1.866779523391448,
      "language_loss": 0.88716942,
      "learning_rate": 3.836789105629236e-06,
      "loss": 0.90932655,
      "num_input_tokens_seen": 56218165,
      "step": 2588,
      "time_per_iteration": 4.192267894744873
    },
    {
      "auxiliary_loss_clip": 0.01094981,
      "auxiliary_loss_mlp": 0.01052123,
      "balance_loss_clip": 1.04558384,
      "balance_loss_mlp": 1.03164268,
      "epoch": 0.1556591011573726,
      "flos": 23148772744320.0,
      "grad_norm": 2.018423224363699,
      "language_loss": 0.64624381,
      "learning_rate": 3.83663497412695e-06,
      "loss": 0.66771483,
      "num_input_tokens_seen": 56237160,
      "step": 2589,
      "time_per_iteration": 4.303871154785156
    },
    {
      "auxiliary_loss_clip": 0.01104407,
      "auxiliary_loss_mlp": 0.01041976,
      "balance_loss_clip": 1.04520249,
      "balance_loss_mlp": 1.02123344,
      "epoch": 0.15571922441004057,
      "flos": 25370888745600.0,
      "grad_norm": 1.784618480549341,
      "language_loss": 0.82832813,
      "learning_rate": 3.836480772979281e-06,
      "loss": 0.84979194,
      "num_input_tokens_seen": 56257610,
      "step": 2590,
      "time_per_iteration": 4.460350751876831
    },
    {
      "auxiliary_loss_clip": 0.011248,
      "auxiliary_loss_mlp": 0.01047287,
      "balance_loss_clip": 1.05032134,
      "balance_loss_mlp": 1.02694952,
      "epoch": 0.15577934766270854,
      "flos": 14501375890560.0,
      "grad_norm": 2.6687659077907484,
      "language_loss": 0.78766,
      "learning_rate": 3.836326502192077e-06,
      "loss": 0.80938083,
      "num_input_tokens_seen": 56275215,
      "step": 2591,
      "time_per_iteration": 2.73305606842041
    },
    {
      "auxiliary_loss_clip": 0.01143879,
      "auxiliary_loss_mlp": 0.01049015,
      "balance_loss_clip": 1.05174232,
      "balance_loss_mlp": 1.03137255,
      "epoch": 0.15583947091537653,
      "flos": 37414537372800.0,
      "grad_norm": 2.0331558547393054,
      "language_loss": 0.65025747,
      "learning_rate": 3.836172161771189e-06,
      "loss": 0.67218637,
      "num_input_tokens_seen": 56297130,
      "step": 2592,
      "time_per_iteration": 2.8582632541656494
    },
    {
      "auxiliary_loss_clip": 0.01136043,
      "auxiliary_loss_mlp": 0.01052096,
      "balance_loss_clip": 1.05417228,
      "balance_loss_mlp": 1.0322001,
      "epoch": 0.1558995941680445,
      "flos": 21834729498240.0,
      "grad_norm": 2.311634250072179,
      "language_loss": 0.82506329,
      "learning_rate": 3.836017751722467e-06,
      "loss": 0.84694475,
      "num_input_tokens_seen": 56314995,
      "step": 2593,
      "time_per_iteration": 2.7230453491210938
    },
    {
      "auxiliary_loss_clip": 0.01142565,
      "auxiliary_loss_mlp": 0.01046037,
      "balance_loss_clip": 1.05237365,
      "balance_loss_mlp": 1.02676034,
      "epoch": 0.15595971742071246,
      "flos": 19792633484160.0,
      "grad_norm": 2.778410683125911,
      "language_loss": 0.73220694,
      "learning_rate": 3.8358632720517695e-06,
      "loss": 0.75409293,
      "num_input_tokens_seen": 56334005,
      "step": 2594,
      "time_per_iteration": 2.708063840866089
    },
    {
      "auxiliary_loss_clip": 0.01117989,
      "auxiliary_loss_mlp": 0.01040106,
      "balance_loss_clip": 1.0453043,
      "balance_loss_mlp": 1.02077007,
      "epoch": 0.15601984067338043,
      "flos": 26722135503360.0,
      "grad_norm": 2.1444704922101105,
      "language_loss": 0.81569934,
      "learning_rate": 3.835708722764952e-06,
      "loss": 0.83728027,
      "num_input_tokens_seen": 56353795,
      "step": 2595,
      "time_per_iteration": 2.716334581375122
    },
    {
      "auxiliary_loss_clip": 0.01155359,
      "auxiliary_loss_mlp": 0.01043269,
      "balance_loss_clip": 1.05093551,
      "balance_loss_mlp": 1.0238502,
      "epoch": 0.1560799639260484,
      "flos": 18369278173440.0,
      "grad_norm": 1.8943501893042642,
      "language_loss": 0.86674929,
      "learning_rate": 3.835554103867876e-06,
      "loss": 0.88873553,
      "num_input_tokens_seen": 56373195,
      "step": 2596,
      "time_per_iteration": 2.5947446823120117
    },
    {
      "auxiliary_loss_clip": 0.01144729,
      "auxiliary_loss_mlp": 0.01042109,
      "balance_loss_clip": 1.05225515,
      "balance_loss_mlp": 1.02360725,
      "epoch": 0.15614008717871636,
      "flos": 22598980197120.0,
      "grad_norm": 1.8059460934517404,
      "language_loss": 0.68772388,
      "learning_rate": 3.835399415366404e-06,
      "loss": 0.70959222,
      "num_input_tokens_seen": 56391525,
      "step": 2597,
      "time_per_iteration": 2.8101041316986084
    },
    {
      "auxiliary_loss_clip": 0.01130069,
      "auxiliary_loss_mlp": 0.01050835,
      "balance_loss_clip": 1.05409336,
      "balance_loss_mlp": 1.03165436,
      "epoch": 0.15620021043138435,
      "flos": 22746860490240.0,
      "grad_norm": 1.9103744906429732,
      "language_loss": 0.79860938,
      "learning_rate": 3.8352446572664035e-06,
      "loss": 0.82041842,
      "num_input_tokens_seen": 56410715,
      "step": 2598,
      "time_per_iteration": 2.695117950439453
    },
    {
      "auxiliary_loss_clip": 0.0112861,
      "auxiliary_loss_mlp": 0.00776118,
      "balance_loss_clip": 1.04750216,
      "balance_loss_mlp": 1.0006249,
      "epoch": 0.15626033368405232,
      "flos": 13114936782720.0,
      "grad_norm": 3.1104681024188827,
      "language_loss": 0.83092594,
      "learning_rate": 3.8350898295737405e-06,
      "loss": 0.84997326,
      "num_input_tokens_seen": 56429170,
      "step": 2599,
      "time_per_iteration": 2.665703773498535
    },
    {
      "auxiliary_loss_clip": 0.01160593,
      "auxiliary_loss_mlp": 0.0105002,
      "balance_loss_clip": 1.05274248,
      "balance_loss_mlp": 1.02924192,
      "epoch": 0.15632045693672028,
      "flos": 16472297105280.0,
      "grad_norm": 2.2910683048406266,
      "language_loss": 0.81530893,
      "learning_rate": 3.834934932294287e-06,
      "loss": 0.83741504,
      "num_input_tokens_seen": 56445685,
      "step": 2600,
      "time_per_iteration": 2.615651845932007
    },
    {
      "auxiliary_loss_clip": 0.01161023,
      "auxiliary_loss_mlp": 0.00776671,
      "balance_loss_clip": 1.05562234,
      "balance_loss_mlp": 1.00063944,
      "epoch": 0.15638058018938825,
      "flos": 20850346298880.0,
      "grad_norm": 1.7832591469657297,
      "language_loss": 0.88511437,
      "learning_rate": 3.834779965433917e-06,
      "loss": 0.90449131,
      "num_input_tokens_seen": 56465900,
      "step": 2601,
      "time_per_iteration": 2.6833529472351074
    },
    {
      "auxiliary_loss_clip": 0.0116257,
      "auxiliary_loss_mlp": 0.0106307,
      "balance_loss_clip": 1.05569744,
      "balance_loss_mlp": 1.04120743,
      "epoch": 0.1564407034420562,
      "flos": 21872220318720.0,
      "grad_norm": 1.9421054688538308,
      "language_loss": 0.78707534,
      "learning_rate": 3.834624928998508e-06,
      "loss": 0.80933177,
      "num_input_tokens_seen": 56485020,
      "step": 2602,
      "time_per_iteration": 2.6296608448028564
    },
    {
      "auxiliary_loss_clip": 0.01126653,
      "auxiliary_loss_mlp": 0.01043676,
      "balance_loss_clip": 1.05035329,
      "balance_loss_mlp": 1.02419758,
      "epoch": 0.15650082669472418,
      "flos": 21834549930240.0,
      "grad_norm": 1.8230718276715763,
      "language_loss": 0.74029547,
      "learning_rate": 3.8344698229939376e-06,
      "loss": 0.76199877,
      "num_input_tokens_seen": 56505205,
      "step": 2603,
      "time_per_iteration": 2.744508743286133
    },
    {
      "auxiliary_loss_clip": 0.01143305,
      "auxiliary_loss_mlp": 0.01051047,
      "balance_loss_clip": 1.04820418,
      "balance_loss_mlp": 1.03112721,
      "epoch": 0.15656094994739214,
      "flos": 13800542653440.0,
      "grad_norm": 4.041164356714064,
      "language_loss": 0.87723601,
      "learning_rate": 3.8343146474260865e-06,
      "loss": 0.89917958,
      "num_input_tokens_seen": 56521495,
      "step": 2604,
      "time_per_iteration": 2.682457447052002
    },
    {
      "auxiliary_loss_clip": 0.01145351,
      "auxiliary_loss_mlp": 0.01044759,
      "balance_loss_clip": 1.04976749,
      "balance_loss_mlp": 1.0256021,
      "epoch": 0.15662107320006013,
      "flos": 27308197808640.0,
      "grad_norm": 2.260429022209425,
      "language_loss": 0.8573193,
      "learning_rate": 3.834159402300841e-06,
      "loss": 0.87922043,
      "num_input_tokens_seen": 56540665,
      "step": 2605,
      "time_per_iteration": 2.7724974155426025
    },
    {
      "auxiliary_loss_clip": 0.0115108,
      "auxiliary_loss_mlp": 0.01047256,
      "balance_loss_clip": 1.05181313,
      "balance_loss_mlp": 1.02676356,
      "epoch": 0.1566811964527281,
      "flos": 26685075646080.0,
      "grad_norm": 1.7309636492693905,
      "language_loss": 0.73101914,
      "learning_rate": 3.834004087624087e-06,
      "loss": 0.75300246,
      "num_input_tokens_seen": 56560805,
      "step": 2606,
      "time_per_iteration": 2.7490081787109375
    },
    {
      "auxiliary_loss_clip": 0.01158388,
      "auxiliary_loss_mlp": 0.01049752,
      "balance_loss_clip": 1.0552665,
      "balance_loss_mlp": 1.03165627,
      "epoch": 0.15674131970539606,
      "flos": 16103422385280.0,
      "grad_norm": 2.968092109370304,
      "language_loss": 0.76497948,
      "learning_rate": 3.8338487034017145e-06,
      "loss": 0.78706092,
      "num_input_tokens_seen": 56576335,
      "step": 2607,
      "time_per_iteration": 2.6597230434417725
    },
    {
      "auxiliary_loss_clip": 0.01120645,
      "auxiliary_loss_mlp": 0.01047174,
      "balance_loss_clip": 1.05131412,
      "balance_loss_mlp": 1.0284934,
      "epoch": 0.15680144295806403,
      "flos": 19169690889600.0,
      "grad_norm": 1.7981763092074996,
      "language_loss": 0.82107675,
      "learning_rate": 3.833693249639615e-06,
      "loss": 0.84275496,
      "num_input_tokens_seen": 56595880,
      "step": 2608,
      "time_per_iteration": 2.7072103023529053
    },
    {
      "auxiliary_loss_clip": 0.0112834,
      "auxiliary_loss_mlp": 0.01045106,
      "balance_loss_clip": 1.04685056,
      "balance_loss_mlp": 1.02436399,
      "epoch": 0.156861566210732,
      "flos": 20813430096000.0,
      "grad_norm": 1.6817301031159713,
      "language_loss": 0.72335941,
      "learning_rate": 3.833537726343684e-06,
      "loss": 0.74509382,
      "num_input_tokens_seen": 56615130,
      "step": 2609,
      "time_per_iteration": 2.690690755844116
    },
    {
      "auxiliary_loss_clip": 0.01143972,
      "auxiliary_loss_mlp": 0.01036718,
      "balance_loss_clip": 1.04901087,
      "balance_loss_mlp": 1.01756072,
      "epoch": 0.15692168946339996,
      "flos": 20047922421120.0,
      "grad_norm": 5.132438477880424,
      "language_loss": 0.72317064,
      "learning_rate": 3.833382133519818e-06,
      "loss": 0.74497753,
      "num_input_tokens_seen": 56634005,
      "step": 2610,
      "time_per_iteration": 2.6515614986419678
    },
    {
      "auxiliary_loss_clip": 0.01159588,
      "auxiliary_loss_mlp": 0.01051513,
      "balance_loss_clip": 1.05216432,
      "balance_loss_mlp": 1.03063977,
      "epoch": 0.15698181271606793,
      "flos": 21398019943680.0,
      "grad_norm": 2.0600295188113935,
      "language_loss": 0.72915608,
      "learning_rate": 3.833226471173919e-06,
      "loss": 0.75126708,
      "num_input_tokens_seen": 56653480,
      "step": 2611,
      "time_per_iteration": 2.630988359451294
    },
    {
      "auxiliary_loss_clip": 0.01141924,
      "auxiliary_loss_mlp": 0.01042538,
      "balance_loss_clip": 1.04917872,
      "balance_loss_mlp": 1.0231905,
      "epoch": 0.15704193596873592,
      "flos": 20845785271680.0,
      "grad_norm": 2.0339762399532186,
      "language_loss": 0.70766544,
      "learning_rate": 3.833070739311887e-06,
      "loss": 0.72951007,
      "num_input_tokens_seen": 56672270,
      "step": 2612,
      "time_per_iteration": 2.6569461822509766
    },
    {
      "auxiliary_loss_clip": 0.01116284,
      "auxiliary_loss_mlp": 0.01051299,
      "balance_loss_clip": 1.04844582,
      "balance_loss_mlp": 1.03221321,
      "epoch": 0.15710205922140388,
      "flos": 21762908254080.0,
      "grad_norm": 1.9704781930994688,
      "language_loss": 0.76294881,
      "learning_rate": 3.83291493793963e-06,
      "loss": 0.78462464,
      "num_input_tokens_seen": 56691510,
      "step": 2613,
      "time_per_iteration": 2.7188539505004883
    },
    {
      "auxiliary_loss_clip": 0.01115155,
      "auxiliary_loss_mlp": 0.01049301,
      "balance_loss_clip": 1.04504919,
      "balance_loss_mlp": 1.02956033,
      "epoch": 0.15716218247407185,
      "flos": 25007760201600.0,
      "grad_norm": 2.137998057111896,
      "language_loss": 0.65944499,
      "learning_rate": 3.832759067063055e-06,
      "loss": 0.68108952,
      "num_input_tokens_seen": 56712230,
      "step": 2614,
      "time_per_iteration": 2.7550084590911865
    },
    {
      "auxiliary_loss_clip": 0.01151987,
      "auxiliary_loss_mlp": 0.01044173,
      "balance_loss_clip": 1.05387104,
      "balance_loss_mlp": 1.02374101,
      "epoch": 0.1572223057267398,
      "flos": 20191780391040.0,
      "grad_norm": 2.2755662506820915,
      "language_loss": 0.75204211,
      "learning_rate": 3.832603126688072e-06,
      "loss": 0.77400374,
      "num_input_tokens_seen": 56727490,
      "step": 2615,
      "time_per_iteration": 2.683225154876709
    },
    {
      "auxiliary_loss_clip": 0.01138545,
      "auxiliary_loss_mlp": 0.01050891,
      "balance_loss_clip": 1.05209839,
      "balance_loss_mlp": 1.03078008,
      "epoch": 0.15728242897940778,
      "flos": 20959514709120.0,
      "grad_norm": 2.581872009488739,
      "language_loss": 0.73064095,
      "learning_rate": 3.832447116820594e-06,
      "loss": 0.75253528,
      "num_input_tokens_seen": 56747385,
      "step": 2616,
      "time_per_iteration": 2.6660919189453125
    },
    {
      "auxiliary_loss_clip": 0.01130717,
      "auxiliary_loss_mlp": 0.01047511,
      "balance_loss_clip": 1.04999971,
      "balance_loss_mlp": 1.02794933,
      "epoch": 0.15734255223207574,
      "flos": 23038275530880.0,
      "grad_norm": 2.813587490853999,
      "language_loss": 0.72425079,
      "learning_rate": 3.832291037466539e-06,
      "loss": 0.74603307,
      "num_input_tokens_seen": 56768055,
      "step": 2617,
      "time_per_iteration": 2.768561363220215
    },
    {
      "auxiliary_loss_clip": 0.01138315,
      "auxiliary_loss_mlp": 0.0104637,
      "balance_loss_clip": 1.04947805,
      "balance_loss_mlp": 1.02548432,
      "epoch": 0.15740267548474374,
      "flos": 20551281661440.0,
      "grad_norm": 2.3222819484870016,
      "language_loss": 0.74358094,
      "learning_rate": 3.8321348886318235e-06,
      "loss": 0.76542777,
      "num_input_tokens_seen": 56785110,
      "step": 2618,
      "time_per_iteration": 2.66121768951416
    },
    {
      "auxiliary_loss_clip": 0.01162954,
      "auxiliary_loss_mlp": 0.01046178,
      "balance_loss_clip": 1.05417252,
      "balance_loss_mlp": 1.02526867,
      "epoch": 0.1574627987374117,
      "flos": 22666922772480.0,
      "grad_norm": 1.8808629075569874,
      "language_loss": 0.78896272,
      "learning_rate": 3.8319786703223695e-06,
      "loss": 0.81105405,
      "num_input_tokens_seen": 56804975,
      "step": 2619,
      "time_per_iteration": 2.6743338108062744
    },
    {
      "auxiliary_loss_clip": 0.01126081,
      "auxiliary_loss_mlp": 0.01055551,
      "balance_loss_clip": 1.05046356,
      "balance_loss_mlp": 1.03576207,
      "epoch": 0.15752292199007967,
      "flos": 16800664262400.0,
      "grad_norm": 1.9082963728737496,
      "language_loss": 0.76517296,
      "learning_rate": 3.831822382544101e-06,
      "loss": 0.78698927,
      "num_input_tokens_seen": 56822470,
      "step": 2620,
      "time_per_iteration": 2.6481080055236816
    },
    {
      "auxiliary_loss_clip": 0.01136128,
      "auxiliary_loss_mlp": 0.0104575,
      "balance_loss_clip": 1.05097985,
      "balance_loss_mlp": 1.02488887,
      "epoch": 0.15758304524274763,
      "flos": 29826002568960.0,
      "grad_norm": 1.6603432400664486,
      "language_loss": 0.7136035,
      "learning_rate": 3.831666025302944e-06,
      "loss": 0.73542225,
      "num_input_tokens_seen": 56842100,
      "step": 2621,
      "time_per_iteration": 2.70985746383667
    },
    {
      "auxiliary_loss_clip": 0.01103274,
      "auxiliary_loss_mlp": 0.01052522,
      "balance_loss_clip": 1.04624665,
      "balance_loss_mlp": 1.02921629,
      "epoch": 0.1576431684954156,
      "flos": 53577426723840.0,
      "grad_norm": 2.1843515622778624,
      "language_loss": 0.72136736,
      "learning_rate": 3.831509598604828e-06,
      "loss": 0.74292529,
      "num_input_tokens_seen": 56865920,
      "step": 2622,
      "time_per_iteration": 3.024561643600464
    },
    {
      "auxiliary_loss_clip": 0.01095163,
      "auxiliary_loss_mlp": 0.01043948,
      "balance_loss_clip": 1.04474711,
      "balance_loss_mlp": 1.02464843,
      "epoch": 0.15770329174808356,
      "flos": 20813609664000.0,
      "grad_norm": 1.6586715789846178,
      "language_loss": 0.87637675,
      "learning_rate": 3.831353102455684e-06,
      "loss": 0.8977679,
      "num_input_tokens_seen": 56885265,
      "step": 2623,
      "time_per_iteration": 2.9600114822387695
    },
    {
      "auxiliary_loss_clip": 0.01158714,
      "auxiliary_loss_mlp": 0.01044337,
      "balance_loss_clip": 1.05476475,
      "balance_loss_mlp": 1.02564478,
      "epoch": 0.15776341500075153,
      "flos": 24974004395520.0,
      "grad_norm": 1.6915331173398198,
      "language_loss": 0.81600082,
      "learning_rate": 3.831196536861448e-06,
      "loss": 0.83803129,
      "num_input_tokens_seen": 56906710,
      "step": 2624,
      "time_per_iteration": 2.6621103286743164
    },
    {
      "auxiliary_loss_clip": 0.01122344,
      "auxiliary_loss_mlp": 0.01049423,
      "balance_loss_clip": 1.04776418,
      "balance_loss_mlp": 1.02990842,
      "epoch": 0.15782353825341952,
      "flos": 21907915459200.0,
      "grad_norm": 2.879465237309773,
      "language_loss": 0.79977828,
      "learning_rate": 3.831039901828054e-06,
      "loss": 0.82149595,
      "num_input_tokens_seen": 56924275,
      "step": 2625,
      "time_per_iteration": 2.7291064262390137
    },
    {
      "auxiliary_loss_clip": 0.01157938,
      "auxiliary_loss_mlp": 0.01046203,
      "balance_loss_clip": 1.05403268,
      "balance_loss_mlp": 1.02857196,
      "epoch": 0.15788366150608749,
      "flos": 26177191292160.0,
      "grad_norm": 2.133783972400447,
      "language_loss": 0.80332482,
      "learning_rate": 3.830883197361445e-06,
      "loss": 0.8253662,
      "num_input_tokens_seen": 56941525,
      "step": 2626,
      "time_per_iteration": 4.252760171890259
    },
    {
      "auxiliary_loss_clip": 0.01102762,
      "auxiliary_loss_mlp": 0.01057658,
      "balance_loss_clip": 1.05214024,
      "balance_loss_mlp": 1.03512752,
      "epoch": 0.15794378475875545,
      "flos": 27709822753920.0,
      "grad_norm": 3.9802810067864045,
      "language_loss": 0.73636395,
      "learning_rate": 3.830726423467561e-06,
      "loss": 0.75796819,
      "num_input_tokens_seen": 56962145,
      "step": 2627,
      "time_per_iteration": 4.328871250152588
    },
    {
      "auxiliary_loss_clip": 0.01117433,
      "auxiliary_loss_mlp": 0.01055032,
      "balance_loss_clip": 1.0503006,
      "balance_loss_mlp": 1.0351001,
      "epoch": 0.15800390801142342,
      "flos": 12130158533760.0,
      "grad_norm": 2.0211273696228216,
      "language_loss": 0.84589541,
      "learning_rate": 3.830569580152348e-06,
      "loss": 0.86762005,
      "num_input_tokens_seen": 56977505,
      "step": 2628,
      "time_per_iteration": 2.6785013675689697
    },
    {
      "auxiliary_loss_clip": 0.01129476,
      "auxiliary_loss_mlp": 0.01040858,
      "balance_loss_clip": 1.05065978,
      "balance_loss_mlp": 1.02308416,
      "epoch": 0.15806403126409138,
      "flos": 20704728562560.0,
      "grad_norm": 1.897214582222077,
      "language_loss": 0.76437485,
      "learning_rate": 3.830412667421752e-06,
      "loss": 0.78607821,
      "num_input_tokens_seen": 56996770,
      "step": 2629,
      "time_per_iteration": 4.2878499031066895
    },
    {
      "auxiliary_loss_clip": 0.01143973,
      "auxiliary_loss_mlp": 0.01046449,
      "balance_loss_clip": 1.0529623,
      "balance_loss_mlp": 1.02675569,
      "epoch": 0.15812415451675935,
      "flos": 17821712269440.0,
      "grad_norm": 2.252423233454998,
      "language_loss": 0.73337436,
      "learning_rate": 3.8302556852817245e-06,
      "loss": 0.75527859,
      "num_input_tokens_seen": 57014970,
      "step": 2630,
      "time_per_iteration": 4.253108263015747
    },
    {
      "auxiliary_loss_clip": 0.01156261,
      "auxiliary_loss_mlp": 0.01045602,
      "balance_loss_clip": 1.05644512,
      "balance_loss_mlp": 1.02615929,
      "epoch": 0.15818427776942734,
      "flos": 20084048524800.0,
      "grad_norm": 2.390369083551665,
      "language_loss": 0.83678091,
      "learning_rate": 3.8300986337382184e-06,
      "loss": 0.85879952,
      "num_input_tokens_seen": 57034045,
      "step": 2631,
      "time_per_iteration": 2.6145882606506348
    },
    {
      "auxiliary_loss_clip": 0.01159092,
      "auxiliary_loss_mlp": 0.01045772,
      "balance_loss_clip": 1.05313432,
      "balance_loss_mlp": 1.02746117,
      "epoch": 0.1582444010220953,
      "flos": 21214911386880.0,
      "grad_norm": 1.8755653224160422,
      "language_loss": 0.78415525,
      "learning_rate": 3.8299415127971895e-06,
      "loss": 0.80620384,
      "num_input_tokens_seen": 57053695,
      "step": 2632,
      "time_per_iteration": 2.656691551208496
    },
    {
      "auxiliary_loss_clip": 0.01151481,
      "auxiliary_loss_mlp": 0.01057283,
      "balance_loss_clip": 1.05574381,
      "balance_loss_mlp": 1.03769732,
      "epoch": 0.15830452427476327,
      "flos": 17858341163520.0,
      "grad_norm": 2.079450153413421,
      "language_loss": 0.8301838,
      "learning_rate": 3.829784322464594e-06,
      "loss": 0.85227144,
      "num_input_tokens_seen": 57071290,
      "step": 2633,
      "time_per_iteration": 2.622725248336792
    },
    {
      "auxiliary_loss_clip": 0.01165069,
      "auxiliary_loss_mlp": 0.01041545,
      "balance_loss_clip": 1.05761647,
      "balance_loss_mlp": 1.02223265,
      "epoch": 0.15836464752743123,
      "flos": 24534960456960.0,
      "grad_norm": 2.1719104392782813,
      "language_loss": 0.77448404,
      "learning_rate": 3.829627062746394e-06,
      "loss": 0.79655015,
      "num_input_tokens_seen": 57091465,
      "step": 2634,
      "time_per_iteration": 2.6383235454559326
    },
    {
      "auxiliary_loss_clip": 0.01127407,
      "auxiliary_loss_mlp": 0.00777775,
      "balance_loss_clip": 1.05277348,
      "balance_loss_mlp": 1.00136137,
      "epoch": 0.1584247707800992,
      "flos": 20120821073280.0,
      "grad_norm": 3.5133527254089087,
      "language_loss": 0.88479185,
      "learning_rate": 3.829469733648552e-06,
      "loss": 0.90384364,
      "num_input_tokens_seen": 57110075,
      "step": 2635,
      "time_per_iteration": 2.725924491882324
    },
    {
      "auxiliary_loss_clip": 0.01096223,
      "auxiliary_loss_mlp": 0.01058885,
      "balance_loss_clip": 1.04816198,
      "balance_loss_mlp": 1.03847599,
      "epoch": 0.15848489403276717,
      "flos": 20375966355840.0,
      "grad_norm": 2.8627721083207627,
      "language_loss": 0.75762677,
      "learning_rate": 3.829312335177034e-06,
      "loss": 0.77917778,
      "num_input_tokens_seen": 57128945,
      "step": 2636,
      "time_per_iteration": 2.775310516357422
    },
    {
      "auxiliary_loss_clip": 0.01120174,
      "auxiliary_loss_mlp": 0.01043834,
      "balance_loss_clip": 1.05117822,
      "balance_loss_mlp": 1.02350879,
      "epoch": 0.15854501728543513,
      "flos": 39346890359040.0,
      "grad_norm": 2.388418559522659,
      "language_loss": 0.71977961,
      "learning_rate": 3.82915486733781e-06,
      "loss": 0.74141967,
      "num_input_tokens_seen": 57152385,
      "step": 2637,
      "time_per_iteration": 2.8375279903411865
    },
    {
      "auxiliary_loss_clip": 0.0115052,
      "auxiliary_loss_mlp": 0.01044842,
      "balance_loss_clip": 1.05661607,
      "balance_loss_mlp": 1.02640057,
      "epoch": 0.15860514053810312,
      "flos": 24864225454080.0,
      "grad_norm": 2.1640345554565057,
      "language_loss": 0.78352648,
      "learning_rate": 3.82899733013685e-06,
      "loss": 0.80548006,
      "num_input_tokens_seen": 57172620,
      "step": 2638,
      "time_per_iteration": 2.7298176288604736
    },
    {
      "auxiliary_loss_clip": 0.01129706,
      "auxiliary_loss_mlp": 0.01057375,
      "balance_loss_clip": 1.05311394,
      "balance_loss_mlp": 1.03715718,
      "epoch": 0.1586652637907711,
      "flos": 26177694082560.0,
      "grad_norm": 2.325769963269074,
      "language_loss": 0.75845039,
      "learning_rate": 3.828839723580128e-06,
      "loss": 0.78032124,
      "num_input_tokens_seen": 57194680,
      "step": 2639,
      "time_per_iteration": 2.7731449604034424
    },
    {
      "auxiliary_loss_clip": 0.01104856,
      "auxiliary_loss_mlp": 0.01057283,
      "balance_loss_clip": 1.05350864,
      "balance_loss_mlp": 1.03772068,
      "epoch": 0.15872538704343905,
      "flos": 19792058866560.0,
      "grad_norm": 2.173238447343554,
      "language_loss": 0.81319505,
      "learning_rate": 3.82868204767362e-06,
      "loss": 0.83481646,
      "num_input_tokens_seen": 57214675,
      "step": 2640,
      "time_per_iteration": 2.8024139404296875
    },
    {
      "auxiliary_loss_clip": 0.01135166,
      "auxiliary_loss_mlp": 0.01054673,
      "balance_loss_clip": 1.05492401,
      "balance_loss_mlp": 1.03426492,
      "epoch": 0.15878551029610702,
      "flos": 28475366342400.0,
      "grad_norm": 2.013499020988034,
      "language_loss": 0.66893363,
      "learning_rate": 3.828524302423306e-06,
      "loss": 0.69083202,
      "num_input_tokens_seen": 57235830,
      "step": 2641,
      "time_per_iteration": 2.7519116401672363
    },
    {
      "auxiliary_loss_clip": 0.01149448,
      "auxiliary_loss_mlp": 0.01051949,
      "balance_loss_clip": 1.05758858,
      "balance_loss_mlp": 1.0326376,
      "epoch": 0.15884563354877498,
      "flos": 24206701040640.0,
      "grad_norm": 2.139760259286454,
      "language_loss": 0.7552591,
      "learning_rate": 3.828366487835167e-06,
      "loss": 0.77727306,
      "num_input_tokens_seen": 57255970,
      "step": 2642,
      "time_per_iteration": 2.706136465072632
    },
    {
      "auxiliary_loss_clip": 0.01156917,
      "auxiliary_loss_mlp": 0.01042142,
      "balance_loss_clip": 1.06263423,
      "balance_loss_mlp": 1.02323556,
      "epoch": 0.15890575680144295,
      "flos": 23949795991680.0,
      "grad_norm": 1.9419610036505286,
      "language_loss": 0.70564604,
      "learning_rate": 3.828208603915186e-06,
      "loss": 0.72763658,
      "num_input_tokens_seen": 57274435,
      "step": 2643,
      "time_per_iteration": 2.682015895843506
    },
    {
      "auxiliary_loss_clip": 0.01161783,
      "auxiliary_loss_mlp": 0.01041643,
      "balance_loss_clip": 1.05891204,
      "balance_loss_mlp": 1.02389312,
      "epoch": 0.15896588005411091,
      "flos": 21215019127680.0,
      "grad_norm": 1.846517711414915,
      "language_loss": 0.78057045,
      "learning_rate": 3.828050650669353e-06,
      "loss": 0.80260473,
      "num_input_tokens_seen": 57293115,
      "step": 2644,
      "time_per_iteration": 2.683790922164917
    },
    {
      "auxiliary_loss_clip": 0.01151239,
      "auxiliary_loss_mlp": 0.01050105,
      "balance_loss_clip": 1.05701637,
      "balance_loss_mlp": 1.03154373,
      "epoch": 0.1590260033067789,
      "flos": 24352390604160.0,
      "grad_norm": 3.757920662841351,
      "language_loss": 0.81961924,
      "learning_rate": 3.827892628103657e-06,
      "loss": 0.84163266,
      "num_input_tokens_seen": 57312565,
      "step": 2645,
      "time_per_iteration": 2.698085069656372
    },
    {
      "auxiliary_loss_clip": 0.01162748,
      "auxiliary_loss_mlp": 0.01048492,
      "balance_loss_clip": 1.05487716,
      "balance_loss_mlp": 1.02854836,
      "epoch": 0.15908612655944687,
      "flos": 32048944583040.0,
      "grad_norm": 2.056693785790565,
      "language_loss": 0.69412929,
      "learning_rate": 3.827734536224087e-06,
      "loss": 0.71624172,
      "num_input_tokens_seen": 57333360,
      "step": 2646,
      "time_per_iteration": 2.7166528701782227
    },
    {
      "auxiliary_loss_clip": 0.01135067,
      "auxiliary_loss_mlp": 0.01040314,
      "balance_loss_clip": 1.05435526,
      "balance_loss_mlp": 1.02223015,
      "epoch": 0.15914624981211484,
      "flos": 17785370684160.0,
      "grad_norm": 2.5975497323405055,
      "language_loss": 0.62932581,
      "learning_rate": 3.827576375036642e-06,
      "loss": 0.65107965,
      "num_input_tokens_seen": 57350575,
      "step": 2647,
      "time_per_iteration": 2.7405354976654053
    },
    {
      "auxiliary_loss_clip": 0.01160144,
      "auxiliary_loss_mlp": 0.01047955,
      "balance_loss_clip": 1.05654776,
      "balance_loss_mlp": 1.02896523,
      "epoch": 0.1592063730647828,
      "flos": 17712507945600.0,
      "grad_norm": 2.2161421076431025,
      "language_loss": 0.89490473,
      "learning_rate": 3.827418144547318e-06,
      "loss": 0.91698575,
      "num_input_tokens_seen": 57367570,
      "step": 2648,
      "time_per_iteration": 2.6193346977233887
    },
    {
      "auxiliary_loss_clip": 0.01158791,
      "auxiliary_loss_mlp": 0.01048086,
      "balance_loss_clip": 1.05630398,
      "balance_loss_mlp": 1.03072906,
      "epoch": 0.15926649631745077,
      "flos": 18803545603200.0,
      "grad_norm": 1.9960039108301237,
      "language_loss": 0.91307199,
      "learning_rate": 3.827259844762114e-06,
      "loss": 0.93514073,
      "num_input_tokens_seen": 57383980,
      "step": 2649,
      "time_per_iteration": 2.6137378215789795
    },
    {
      "auxiliary_loss_clip": 0.01099661,
      "auxiliary_loss_mlp": 0.01044384,
      "balance_loss_clip": 1.05474401,
      "balance_loss_mlp": 1.02439272,
      "epoch": 0.15932661957011873,
      "flos": 17566243764480.0,
      "grad_norm": 2.3504548368335767,
      "language_loss": 0.71782613,
      "learning_rate": 3.827101475687033e-06,
      "loss": 0.73926663,
      "num_input_tokens_seen": 57400840,
      "step": 2650,
      "time_per_iteration": 2.8883376121520996
    },
    {
      "auxiliary_loss_clip": 0.01146809,
      "auxiliary_loss_mlp": 0.01041815,
      "balance_loss_clip": 1.05386841,
      "balance_loss_mlp": 1.02476835,
      "epoch": 0.15938674282278673,
      "flos": 13334351011200.0,
      "grad_norm": 1.8238326955956992,
      "language_loss": 0.71427429,
      "learning_rate": 3.826943037328082e-06,
      "loss": 0.73616046,
      "num_input_tokens_seen": 57419230,
      "step": 2651,
      "time_per_iteration": 2.607879638671875
    },
    {
      "auxiliary_loss_clip": 0.01118842,
      "auxiliary_loss_mlp": 0.00777496,
      "balance_loss_clip": 1.05154157,
      "balance_loss_mlp": 1.00132799,
      "epoch": 0.1594468660754547,
      "flos": 22488842119680.0,
      "grad_norm": 1.8928974850955373,
      "language_loss": 0.80185902,
      "learning_rate": 3.8267845296912674e-06,
      "loss": 0.82082248,
      "num_input_tokens_seen": 57439315,
      "step": 2652,
      "time_per_iteration": 2.718695640563965
    },
    {
      "auxiliary_loss_clip": 0.01138048,
      "auxiliary_loss_mlp": 0.00775,
      "balance_loss_clip": 1.0567826,
      "balance_loss_mlp": 1.00124729,
      "epoch": 0.15950698932812266,
      "flos": 15007320910080.0,
      "grad_norm": 2.6116065834427387,
      "language_loss": 0.69539076,
      "learning_rate": 3.826625952782601e-06,
      "loss": 0.71452117,
      "num_input_tokens_seen": 57454635,
      "step": 2653,
      "time_per_iteration": 2.7088639736175537
    },
    {
      "auxiliary_loss_clip": 0.01144826,
      "auxiliary_loss_mlp": 0.01038735,
      "balance_loss_clip": 1.05257821,
      "balance_loss_mlp": 1.02050805,
      "epoch": 0.15956711258079062,
      "flos": 30155052084480.0,
      "grad_norm": 2.1937273620657307,
      "language_loss": 0.76670635,
      "learning_rate": 3.826467306608095e-06,
      "loss": 0.78854191,
      "num_input_tokens_seen": 57476805,
      "step": 2654,
      "time_per_iteration": 2.79425048828125
    },
    {
      "auxiliary_loss_clip": 0.01114313,
      "auxiliary_loss_mlp": 0.01041134,
      "balance_loss_clip": 1.04714727,
      "balance_loss_mlp": 1.02248931,
      "epoch": 0.1596272358334586,
      "flos": 21032700670080.0,
      "grad_norm": 2.0572535633716247,
      "language_loss": 0.81873977,
      "learning_rate": 3.826308591173765e-06,
      "loss": 0.84029424,
      "num_input_tokens_seen": 57496400,
      "step": 2655,
      "time_per_iteration": 2.6990878582000732
    },
    {
      "auxiliary_loss_clip": 0.01112525,
      "auxiliary_loss_mlp": 0.01046346,
      "balance_loss_clip": 1.04670715,
      "balance_loss_mlp": 1.02849984,
      "epoch": 0.15968735908612655,
      "flos": 15268032800640.0,
      "grad_norm": 2.0964800101687486,
      "language_loss": 0.73768878,
      "learning_rate": 3.826149806485631e-06,
      "loss": 0.75927746,
      "num_input_tokens_seen": 57513700,
      "step": 2656,
      "time_per_iteration": 2.7409873008728027
    },
    {
      "auxiliary_loss_clip": 0.01111218,
      "auxiliary_loss_mlp": 0.01039948,
      "balance_loss_clip": 1.04749918,
      "balance_loss_mlp": 1.02220988,
      "epoch": 0.15974748233879452,
      "flos": 52665726695040.0,
      "grad_norm": 2.516351978408242,
      "language_loss": 0.77637637,
      "learning_rate": 3.825990952549713e-06,
      "loss": 0.79788804,
      "num_input_tokens_seen": 57536180,
      "step": 2657,
      "time_per_iteration": 2.984161376953125
    },
    {
      "auxiliary_loss_clip": 0.01142397,
      "auxiliary_loss_mlp": 0.01048058,
      "balance_loss_clip": 1.05276513,
      "balance_loss_mlp": 1.02984321,
      "epoch": 0.1598076055914625,
      "flos": 18733232730240.0,
      "grad_norm": 2.1741432296797303,
      "language_loss": 0.74654955,
      "learning_rate": 3.825832029372035e-06,
      "loss": 0.76845407,
      "num_input_tokens_seen": 57555025,
      "step": 2658,
      "time_per_iteration": 2.6795172691345215
    },
    {
      "auxiliary_loss_clip": 0.01137294,
      "auxiliary_loss_mlp": 0.01047097,
      "balance_loss_clip": 1.05887127,
      "balance_loss_mlp": 1.02581763,
      "epoch": 0.15986772884413047,
      "flos": 34349238535680.0,
      "grad_norm": 2.2676743120149916,
      "language_loss": 0.75164986,
      "learning_rate": 3.825673036958624e-06,
      "loss": 0.77349377,
      "num_input_tokens_seen": 57577660,
      "step": 2659,
      "time_per_iteration": 2.885744094848633
    },
    {
      "auxiliary_loss_clip": 0.01122752,
      "auxiliary_loss_mlp": 0.0105323,
      "balance_loss_clip": 1.0512991,
      "balance_loss_mlp": 1.0334295,
      "epoch": 0.15992785209679844,
      "flos": 22054969739520.0,
      "grad_norm": 2.181311046841435,
      "language_loss": 0.90998709,
      "learning_rate": 3.825513975315508e-06,
      "loss": 0.93174696,
      "num_input_tokens_seen": 57596335,
      "step": 2660,
      "time_per_iteration": 2.7562267780303955
    },
    {
      "auxiliary_loss_clip": 0.01114547,
      "auxiliary_loss_mlp": 0.01058378,
      "balance_loss_clip": 1.05538487,
      "balance_loss_mlp": 1.03590751,
      "epoch": 0.1599879753494664,
      "flos": 33066652625280.0,
      "grad_norm": 1.746468400789071,
      "language_loss": 0.77724659,
      "learning_rate": 3.82535484444872e-06,
      "loss": 0.79897583,
      "num_input_tokens_seen": 57616830,
      "step": 2661,
      "time_per_iteration": 2.9896914958953857
    },
    {
      "auxiliary_loss_clip": 0.0113781,
      "auxiliary_loss_mlp": 0.00777461,
      "balance_loss_clip": 1.05382478,
      "balance_loss_mlp": 1.00132632,
      "epoch": 0.16004809860213437,
      "flos": 28038010343040.0,
      "grad_norm": 2.0483033922540086,
      "language_loss": 0.74442393,
      "learning_rate": 3.825195644364292e-06,
      "loss": 0.76357663,
      "num_input_tokens_seen": 57635515,
      "step": 2662,
      "time_per_iteration": 2.7993714809417725
    },
    {
      "auxiliary_loss_clip": 0.01135674,
      "auxiliary_loss_mlp": 0.00780783,
      "balance_loss_clip": 1.05392313,
      "balance_loss_mlp": 1.0016191,
      "epoch": 0.16010822185480234,
      "flos": 22780113505920.0,
      "grad_norm": 2.9903694104875984,
      "language_loss": 0.82515085,
      "learning_rate": 3.825036375068263e-06,
      "loss": 0.84431541,
      "num_input_tokens_seen": 57654250,
      "step": 2663,
      "time_per_iteration": 2.678490161895752
    },
    {
      "auxiliary_loss_clip": 0.01112205,
      "auxiliary_loss_mlp": 0.01044917,
      "balance_loss_clip": 1.05182636,
      "balance_loss_mlp": 1.02574801,
      "epoch": 0.16016834510747033,
      "flos": 20084012611200.0,
      "grad_norm": 2.06786422122115,
      "language_loss": 0.7951405,
      "learning_rate": 3.824877036566672e-06,
      "loss": 0.81671166,
      "num_input_tokens_seen": 57672645,
      "step": 2664,
      "time_per_iteration": 2.819880962371826
    },
    {
      "auxiliary_loss_clip": 0.01151449,
      "auxiliary_loss_mlp": 0.01048023,
      "balance_loss_clip": 1.05374622,
      "balance_loss_mlp": 1.02886605,
      "epoch": 0.1602284683601383,
      "flos": 21173829206400.0,
      "grad_norm": 1.6697703441146605,
      "language_loss": 0.93748474,
      "learning_rate": 3.824717628865561e-06,
      "loss": 0.95947945,
      "num_input_tokens_seen": 57691055,
      "step": 2665,
      "time_per_iteration": 2.697660446166992
    },
    {
      "auxiliary_loss_clip": 0.01127607,
      "auxiliary_loss_mlp": 0.01047415,
      "balance_loss_clip": 1.05185676,
      "balance_loss_mlp": 1.02774525,
      "epoch": 0.16028859161280626,
      "flos": 14647568244480.0,
      "grad_norm": 2.9655602739253095,
      "language_loss": 0.85237324,
      "learning_rate": 3.824558151970974e-06,
      "loss": 0.87412339,
      "num_input_tokens_seen": 57707235,
      "step": 2666,
      "time_per_iteration": 4.282273530960083
    },
    {
      "auxiliary_loss_clip": 0.01129818,
      "auxiliary_loss_mlp": 0.00777125,
      "balance_loss_clip": 1.05257225,
      "balance_loss_mlp": 1.00145936,
      "epoch": 0.16034871486547422,
      "flos": 20990325600000.0,
      "grad_norm": 1.8366839898970433,
      "language_loss": 0.81284773,
      "learning_rate": 3.8243986058889595e-06,
      "loss": 0.83191717,
      "num_input_tokens_seen": 57724190,
      "step": 2667,
      "time_per_iteration": 2.69508695602417
    },
    {
      "auxiliary_loss_clip": 0.0116556,
      "auxiliary_loss_mlp": 0.01046526,
      "balance_loss_clip": 1.06089485,
      "balance_loss_mlp": 1.02643883,
      "epoch": 0.1604088381181422,
      "flos": 21397732634880.0,
      "grad_norm": 1.958935842080623,
      "language_loss": 0.74031079,
      "learning_rate": 3.824238990625567e-06,
      "loss": 0.76243162,
      "num_input_tokens_seen": 57743620,
      "step": 2668,
      "time_per_iteration": 4.2559425830841064
    },
    {
      "auxiliary_loss_clip": 0.01148853,
      "auxiliary_loss_mlp": 0.01051992,
      "balance_loss_clip": 1.05547619,
      "balance_loss_mlp": 1.03240585,
      "epoch": 0.16046896137081015,
      "flos": 23877040993920.0,
      "grad_norm": 1.7737626564305047,
      "language_loss": 0.77495629,
      "learning_rate": 3.824079306186848e-06,
      "loss": 0.7969647,
      "num_input_tokens_seen": 57764810,
      "step": 2669,
      "time_per_iteration": 2.6424050331115723
    },
    {
      "auxiliary_loss_clip": 0.01097339,
      "auxiliary_loss_mlp": 0.01012737,
      "balance_loss_clip": 1.06351233,
      "balance_loss_mlp": 1.00986385,
      "epoch": 0.16052908462347812,
      "flos": 59806709015040.0,
      "grad_norm": 0.8041290684345284,
      "language_loss": 0.5549804,
      "learning_rate": 3.823919552578861e-06,
      "loss": 0.57608116,
      "num_input_tokens_seen": 57824390,
      "step": 2670,
      "time_per_iteration": 4.765664100646973
    },
    {
      "auxiliary_loss_clip": 0.01149639,
      "auxiliary_loss_mlp": 0.01043383,
      "balance_loss_clip": 1.05322218,
      "balance_loss_mlp": 1.02430916,
      "epoch": 0.1605892078761461,
      "flos": 18296559089280.0,
      "grad_norm": 2.6306224128650464,
      "language_loss": 0.77778888,
      "learning_rate": 3.82375972980766e-06,
      "loss": 0.7997191,
      "num_input_tokens_seen": 57843665,
      "step": 2671,
      "time_per_iteration": 2.6876416206359863
    },
    {
      "auxiliary_loss_clip": 0.01151164,
      "auxiliary_loss_mlp": 0.01043962,
      "balance_loss_clip": 1.05529547,
      "balance_loss_mlp": 1.02503204,
      "epoch": 0.16064933112881408,
      "flos": 32160734686080.0,
      "grad_norm": 1.9167251889277674,
      "language_loss": 0.64766788,
      "learning_rate": 3.8235998378793086e-06,
      "loss": 0.66961908,
      "num_input_tokens_seen": 57863305,
      "step": 2672,
      "time_per_iteration": 2.7102553844451904
    },
    {
      "auxiliary_loss_clip": 0.01150206,
      "auxiliary_loss_mlp": 0.01046785,
      "balance_loss_clip": 1.05674481,
      "balance_loss_mlp": 1.02554154,
      "epoch": 0.16070945438148204,
      "flos": 19828795501440.0,
      "grad_norm": 2.045175098484539,
      "language_loss": 0.85708207,
      "learning_rate": 3.8234398767998675e-06,
      "loss": 0.87905198,
      "num_input_tokens_seen": 57883025,
      "step": 2673,
      "time_per_iteration": 2.656360626220703
    },
    {
      "auxiliary_loss_clip": 0.01125542,
      "auxiliary_loss_mlp": 0.01055838,
      "balance_loss_clip": 1.05366015,
      "balance_loss_mlp": 1.03716969,
      "epoch": 0.16076957763415,
      "flos": 18913144976640.0,
      "grad_norm": 2.339006860757087,
      "language_loss": 0.7289716,
      "learning_rate": 3.823279846575403e-06,
      "loss": 0.75078535,
      "num_input_tokens_seen": 57901430,
      "step": 2674,
      "time_per_iteration": 2.7122414112091064
    },
    {
      "auxiliary_loss_clip": 0.01150063,
      "auxiliary_loss_mlp": 0.01045468,
      "balance_loss_clip": 1.05416465,
      "balance_loss_mlp": 1.02464211,
      "epoch": 0.16082970088681797,
      "flos": 16764358590720.0,
      "grad_norm": 1.9341682597436423,
      "language_loss": 0.84438515,
      "learning_rate": 3.823119747211986e-06,
      "loss": 0.86634052,
      "num_input_tokens_seen": 57919550,
      "step": 2675,
      "time_per_iteration": 2.6646435260772705
    },
    {
      "auxiliary_loss_clip": 0.01116221,
      "auxiliary_loss_mlp": 0.01049343,
      "balance_loss_clip": 1.05220723,
      "balance_loss_mlp": 1.02823126,
      "epoch": 0.16088982413948594,
      "flos": 35150261783040.0,
      "grad_norm": 1.871909119220515,
      "language_loss": 0.82216591,
      "learning_rate": 3.822959578715685e-06,
      "loss": 0.84382153,
      "num_input_tokens_seen": 57939890,
      "step": 2676,
      "time_per_iteration": 2.8457534313201904
    },
    {
      "auxiliary_loss_clip": 0.01151157,
      "auxiliary_loss_mlp": 0.01049874,
      "balance_loss_clip": 1.05746996,
      "balance_loss_mlp": 1.03162253,
      "epoch": 0.1609499473921539,
      "flos": 18625105814400.0,
      "grad_norm": 2.1166154816193923,
      "language_loss": 0.73485494,
      "learning_rate": 3.822799341092573e-06,
      "loss": 0.75686526,
      "num_input_tokens_seen": 57957410,
      "step": 2677,
      "time_per_iteration": 2.65387225151062
    },
    {
      "auxiliary_loss_clip": 0.01138188,
      "auxiliary_loss_mlp": 0.01044363,
      "balance_loss_clip": 1.05438483,
      "balance_loss_mlp": 1.02537322,
      "epoch": 0.1610100706448219,
      "flos": 33145728416640.0,
      "grad_norm": 3.229282061984371,
      "language_loss": 0.76305777,
      "learning_rate": 3.822639034348728e-06,
      "loss": 0.78488332,
      "num_input_tokens_seen": 57977900,
      "step": 2678,
      "time_per_iteration": 2.836071014404297
    },
    {
      "auxiliary_loss_clip": 0.01148252,
      "auxiliary_loss_mlp": 0.01047887,
      "balance_loss_clip": 1.05379987,
      "balance_loss_mlp": 1.02789569,
      "epoch": 0.16107019389748986,
      "flos": 34676707852800.0,
      "grad_norm": 8.295814069484678,
      "language_loss": 0.70340431,
      "learning_rate": 3.822478658490228e-06,
      "loss": 0.7253657,
      "num_input_tokens_seen": 57998210,
      "step": 2679,
      "time_per_iteration": 2.771185874938965
    },
    {
      "auxiliary_loss_clip": 0.01059502,
      "auxiliary_loss_mlp": 0.00758644,
      "balance_loss_clip": 1.04695845,
      "balance_loss_mlp": 1.00150955,
      "epoch": 0.16113031715015783,
      "flos": 65713403260800.0,
      "grad_norm": 0.7819629653273137,
      "language_loss": 0.51843339,
      "learning_rate": 3.822318213523154e-06,
      "loss": 0.53661484,
      "num_input_tokens_seen": 58059420,
      "step": 2680,
      "time_per_iteration": 3.3107378482818604
    },
    {
      "auxiliary_loss_clip": 0.01144342,
      "auxiliary_loss_mlp": 0.01047358,
      "balance_loss_clip": 1.05360317,
      "balance_loss_mlp": 1.02632904,
      "epoch": 0.1611904404028258,
      "flos": 20810413353600.0,
      "grad_norm": 1.6718368455031125,
      "language_loss": 0.8028667,
      "learning_rate": 3.8221576994535925e-06,
      "loss": 0.82478368,
      "num_input_tokens_seen": 58078370,
      "step": 2681,
      "time_per_iteration": 2.6986513137817383
    },
    {
      "auxiliary_loss_clip": 0.01139192,
      "auxiliary_loss_mlp": 0.01055518,
      "balance_loss_clip": 1.05603266,
      "balance_loss_mlp": 1.03602743,
      "epoch": 0.16125056365549376,
      "flos": 27013335062400.0,
      "grad_norm": 2.154781054673542,
      "language_loss": 0.68957973,
      "learning_rate": 3.821997116287627e-06,
      "loss": 0.71152687,
      "num_input_tokens_seen": 58097395,
      "step": 2682,
      "time_per_iteration": 2.794686794281006
    },
    {
      "auxiliary_loss_clip": 0.01139216,
      "auxiliary_loss_mlp": 0.01052349,
      "balance_loss_clip": 1.05670619,
      "balance_loss_mlp": 1.03195262,
      "epoch": 0.16131068690816172,
      "flos": 19276524915840.0,
      "grad_norm": 1.9802191055590168,
      "language_loss": 0.87362224,
      "learning_rate": 3.821836464031348e-06,
      "loss": 0.89553785,
      "num_input_tokens_seen": 58115630,
      "step": 2683,
      "time_per_iteration": 2.703634262084961
    },
    {
      "auxiliary_loss_clip": 0.01165497,
      "auxiliary_loss_mlp": 0.0105575,
      "balance_loss_clip": 1.05714059,
      "balance_loss_mlp": 1.03491259,
      "epoch": 0.16137081016082971,
      "flos": 35337931367040.0,
      "grad_norm": 1.939499216066865,
      "language_loss": 0.74143028,
      "learning_rate": 3.821675742690849e-06,
      "loss": 0.76364273,
      "num_input_tokens_seen": 58138655,
      "step": 2684,
      "time_per_iteration": 2.7890264987945557
    },
    {
      "auxiliary_loss_clip": 0.01136683,
      "auxiliary_loss_mlp": 0.00778989,
      "balance_loss_clip": 1.05435085,
      "balance_loss_mlp": 1.00176883,
      "epoch": 0.16143093341349768,
      "flos": 34235257703040.0,
      "grad_norm": 1.9009911635557044,
      "language_loss": 0.70506597,
      "learning_rate": 3.821514952272223e-06,
      "loss": 0.72422272,
      "num_input_tokens_seen": 58157440,
      "step": 2685,
      "time_per_iteration": 2.803942918777466
    },
    {
      "auxiliary_loss_clip": 0.01116315,
      "auxiliary_loss_mlp": 0.01059092,
      "balance_loss_clip": 1.05291295,
      "balance_loss_mlp": 1.03757524,
      "epoch": 0.16149105666616564,
      "flos": 27999262546560.0,
      "grad_norm": 2.295686008167468,
      "language_loss": 0.72060591,
      "learning_rate": 3.821354092781567e-06,
      "loss": 0.74236,
      "num_input_tokens_seen": 58176660,
      "step": 2686,
      "time_per_iteration": 2.850309133529663
    },
    {
      "auxiliary_loss_clip": 0.01153803,
      "auxiliary_loss_mlp": 0.01048887,
      "balance_loss_clip": 1.05603862,
      "balance_loss_mlp": 1.02922952,
      "epoch": 0.1615511799188336,
      "flos": 19422214479360.0,
      "grad_norm": 2.056921120199424,
      "language_loss": 0.81720114,
      "learning_rate": 3.821193164224981e-06,
      "loss": 0.83922803,
      "num_input_tokens_seen": 58195085,
      "step": 2687,
      "time_per_iteration": 2.7077832221984863
    },
    {
      "auxiliary_loss_clip": 0.01154388,
      "auxiliary_loss_mlp": 0.01050682,
      "balance_loss_clip": 1.05335689,
      "balance_loss_mlp": 1.02910483,
      "epoch": 0.16161130317150157,
      "flos": 22854915578880.0,
      "grad_norm": 1.6747986106054085,
      "language_loss": 0.71680355,
      "learning_rate": 3.821032166608568e-06,
      "loss": 0.73885429,
      "num_input_tokens_seen": 58213540,
      "step": 2688,
      "time_per_iteration": 2.700073480606079
    },
    {
      "auxiliary_loss_clip": 0.0112226,
      "auxiliary_loss_mlp": 0.0105252,
      "balance_loss_clip": 1.0517168,
      "balance_loss_mlp": 1.03330338,
      "epoch": 0.16167142642416954,
      "flos": 26110577520000.0,
      "grad_norm": 2.2887064413695253,
      "language_loss": 0.76168394,
      "learning_rate": 3.8208710999384325e-06,
      "loss": 0.78343177,
      "num_input_tokens_seen": 58236995,
      "step": 2689,
      "time_per_iteration": 2.846964120864868
    },
    {
      "auxiliary_loss_clip": 0.01166324,
      "auxiliary_loss_mlp": 0.01052979,
      "balance_loss_clip": 1.05979431,
      "balance_loss_mlp": 1.03308284,
      "epoch": 0.1617315496768375,
      "flos": 22779646629120.0,
      "grad_norm": 2.045037041298705,
      "language_loss": 0.87211925,
      "learning_rate": 3.820709964220683e-06,
      "loss": 0.89431226,
      "num_input_tokens_seen": 58257230,
      "step": 2690,
      "time_per_iteration": 2.704497814178467
    },
    {
      "auxiliary_loss_clip": 0.01143898,
      "auxiliary_loss_mlp": 0.01046571,
      "balance_loss_clip": 1.05318451,
      "balance_loss_mlp": 1.02890396,
      "epoch": 0.1617916729295055,
      "flos": 22017299351040.0,
      "grad_norm": 1.7518031225399346,
      "language_loss": 0.87899524,
      "learning_rate": 3.8205487594614284e-06,
      "loss": 0.90089989,
      "num_input_tokens_seen": 58277080,
      "step": 2691,
      "time_per_iteration": 2.6763153076171875
    },
    {
      "auxiliary_loss_clip": 0.01150265,
      "auxiliary_loss_mlp": 0.01053114,
      "balance_loss_clip": 1.05237532,
      "balance_loss_mlp": 1.03142977,
      "epoch": 0.16185179618217346,
      "flos": 23438248450560.0,
      "grad_norm": 2.1723450057475313,
      "language_loss": 0.81989783,
      "learning_rate": 3.820387485666784e-06,
      "loss": 0.84193164,
      "num_input_tokens_seen": 58294815,
      "step": 2692,
      "time_per_iteration": 2.6381001472473145
    },
    {
      "auxiliary_loss_clip": 0.01167881,
      "auxiliary_loss_mlp": 0.0104606,
      "balance_loss_clip": 1.05555534,
      "balance_loss_mlp": 1.02499604,
      "epoch": 0.16191191943484143,
      "flos": 25666110627840.0,
      "grad_norm": 2.194958172554253,
      "language_loss": 0.81381011,
      "learning_rate": 3.820226142842862e-06,
      "loss": 0.83594954,
      "num_input_tokens_seen": 58313215,
      "step": 2693,
      "time_per_iteration": 2.6366944313049316
    },
    {
      "auxiliary_loss_clip": 0.01164466,
      "auxiliary_loss_mlp": 0.01058298,
      "balance_loss_clip": 1.0587461,
      "balance_loss_mlp": 1.03991616,
      "epoch": 0.1619720426875094,
      "flos": 23477355383040.0,
      "grad_norm": 2.778189532536263,
      "language_loss": 0.83837044,
      "learning_rate": 3.820064730995783e-06,
      "loss": 0.86059809,
      "num_input_tokens_seen": 58333215,
      "step": 2694,
      "time_per_iteration": 2.7802140712738037
    },
    {
      "auxiliary_loss_clip": 0.01116209,
      "auxiliary_loss_mlp": 0.0105764,
      "balance_loss_clip": 1.04927421,
      "balance_loss_mlp": 1.0366354,
      "epoch": 0.16203216594017736,
      "flos": 24133658734080.0,
      "grad_norm": 1.8201511645490482,
      "language_loss": 0.69709098,
      "learning_rate": 3.819903250131667e-06,
      "loss": 0.71882945,
      "num_input_tokens_seen": 58351160,
      "step": 2695,
      "time_per_iteration": 2.756904125213623
    },
    {
      "auxiliary_loss_clip": 0.01155526,
      "auxiliary_loss_mlp": 0.01050837,
      "balance_loss_clip": 1.05799723,
      "balance_loss_mlp": 1.03026128,
      "epoch": 0.16209228919284532,
      "flos": 22340889999360.0,
      "grad_norm": 2.1550523064219487,
      "language_loss": 0.82986331,
      "learning_rate": 3.819741700256637e-06,
      "loss": 0.85192692,
      "num_input_tokens_seen": 58368505,
      "step": 2696,
      "time_per_iteration": 2.651510238647461
    },
    {
      "auxiliary_loss_clip": 0.01174193,
      "auxiliary_loss_mlp": 0.01052819,
      "balance_loss_clip": 1.05826569,
      "balance_loss_mlp": 1.03095615,
      "epoch": 0.1621524124455133,
      "flos": 15815131827840.0,
      "grad_norm": 2.9267990143146503,
      "language_loss": 0.8862049,
      "learning_rate": 3.8195800813768194e-06,
      "loss": 0.90847504,
      "num_input_tokens_seen": 58385085,
      "step": 2697,
      "time_per_iteration": 2.5935380458831787
    },
    {
      "auxiliary_loss_clip": 0.01158945,
      "auxiliary_loss_mlp": 0.01045471,
      "balance_loss_clip": 1.0552485,
      "balance_loss_mlp": 1.02719641,
      "epoch": 0.16221253569818128,
      "flos": 30186688988160.0,
      "grad_norm": 1.7480298293719791,
      "language_loss": 0.80844599,
      "learning_rate": 3.819418393498343e-06,
      "loss": 0.83049017,
      "num_input_tokens_seen": 58406985,
      "step": 2698,
      "time_per_iteration": 2.6685965061187744
    },
    {
      "auxiliary_loss_clip": 0.01151678,
      "auxiliary_loss_mlp": 0.01050084,
      "balance_loss_clip": 1.05785704,
      "balance_loss_mlp": 1.03060579,
      "epoch": 0.16227265895084925,
      "flos": 24605991601920.0,
      "grad_norm": 1.590231062064763,
      "language_loss": 0.77499473,
      "learning_rate": 3.819256636627339e-06,
      "loss": 0.79701245,
      "num_input_tokens_seen": 58426205,
      "step": 2699,
      "time_per_iteration": 2.7206287384033203
    },
    {
      "auxiliary_loss_clip": 0.01134482,
      "auxiliary_loss_mlp": 0.01043888,
      "balance_loss_clip": 1.0504272,
      "balance_loss_mlp": 1.02510071,
      "epoch": 0.1623327822035172,
      "flos": 19573326996480.0,
      "grad_norm": 2.299083669251571,
      "language_loss": 0.85903585,
      "learning_rate": 3.81909481076994e-06,
      "loss": 0.88081944,
      "num_input_tokens_seen": 58443830,
      "step": 2700,
      "time_per_iteration": 2.6440224647521973
    },
    {
      "auxiliary_loss_clip": 0.01150266,
      "auxiliary_loss_mlp": 0.00778348,
      "balance_loss_clip": 1.05360484,
      "balance_loss_mlp": 1.00180686,
      "epoch": 0.16239290545618518,
      "flos": 26468462678400.0,
      "grad_norm": 1.7679372116400307,
      "language_loss": 0.80424523,
      "learning_rate": 3.818932915932284e-06,
      "loss": 0.82353133,
      "num_input_tokens_seen": 58464405,
      "step": 2701,
      "time_per_iteration": 2.6943976879119873
    },
    {
      "auxiliary_loss_clip": 0.01144477,
      "auxiliary_loss_mlp": 0.01046291,
      "balance_loss_clip": 1.05771017,
      "balance_loss_mlp": 1.02664542,
      "epoch": 0.16245302870885314,
      "flos": 15851940289920.0,
      "grad_norm": 1.6539412057050027,
      "language_loss": 0.72777367,
      "learning_rate": 3.818770952120511e-06,
      "loss": 0.74968135,
      "num_input_tokens_seen": 58483295,
      "step": 2702,
      "time_per_iteration": 2.6914141178131104
    },
    {
      "auxiliary_loss_clip": 0.01156069,
      "auxiliary_loss_mlp": 0.01050141,
      "balance_loss_clip": 1.05802381,
      "balance_loss_mlp": 1.02896905,
      "epoch": 0.1625131519615211,
      "flos": 14756521173120.0,
      "grad_norm": 1.8265391375227176,
      "language_loss": 0.7273894,
      "learning_rate": 3.81860891934076e-06,
      "loss": 0.74945152,
      "num_input_tokens_seen": 58501205,
      "step": 2703,
      "time_per_iteration": 2.6301820278167725
    },
    {
      "auxiliary_loss_clip": 0.01165642,
      "auxiliary_loss_mlp": 0.01050857,
      "balance_loss_clip": 1.0553968,
      "balance_loss_mlp": 1.02942359,
      "epoch": 0.1625732752141891,
      "flos": 28220508368640.0,
      "grad_norm": 3.0329584489902666,
      "language_loss": 0.70018482,
      "learning_rate": 3.818446817599176e-06,
      "loss": 0.72234988,
      "num_input_tokens_seen": 58522315,
      "step": 2704,
      "time_per_iteration": 2.6667227745056152
    },
    {
      "auxiliary_loss_clip": 0.01034679,
      "auxiliary_loss_mlp": 0.01001657,
      "balance_loss_clip": 1.03343439,
      "balance_loss_mlp": 0.99865305,
      "epoch": 0.16263339846685707,
      "flos": 67327947688320.0,
      "grad_norm": 0.7801109588151329,
      "language_loss": 0.5336051,
      "learning_rate": 3.818284646901907e-06,
      "loss": 0.55396849,
      "num_input_tokens_seen": 58586695,
      "step": 2705,
      "time_per_iteration": 4.808594465255737
    },
    {
      "auxiliary_loss_clip": 0.01138628,
      "auxiliary_loss_mlp": 0.00781324,
      "balance_loss_clip": 1.0539608,
      "balance_loss_mlp": 1.00171995,
      "epoch": 0.16269352171952503,
      "flos": 14319165173760.0,
      "grad_norm": 2.3827832530074455,
      "language_loss": 0.7536028,
      "learning_rate": 3.818122407255102e-06,
      "loss": 0.77280229,
      "num_input_tokens_seen": 58602435,
      "step": 2706,
      "time_per_iteration": 4.126614570617676
    },
    {
      "auxiliary_loss_clip": 0.01130684,
      "auxiliary_loss_mlp": 0.01047489,
      "balance_loss_clip": 1.0523324,
      "balance_loss_mlp": 1.02859437,
      "epoch": 0.162753644972193,
      "flos": 28361205941760.0,
      "grad_norm": 2.2272392184651038,
      "language_loss": 0.72203928,
      "learning_rate": 3.817960098664914e-06,
      "loss": 0.74382102,
      "num_input_tokens_seen": 58621275,
      "step": 2707,
      "time_per_iteration": 4.2739410400390625
    },
    {
      "auxiliary_loss_clip": 0.01142142,
      "auxiliary_loss_mlp": 0.01047652,
      "balance_loss_clip": 1.05433679,
      "balance_loss_mlp": 1.02898431,
      "epoch": 0.16281376822486096,
      "flos": 19937856170880.0,
      "grad_norm": 3.192481802987827,
      "language_loss": 0.83481139,
      "learning_rate": 3.817797721137495e-06,
      "loss": 0.85670936,
      "num_input_tokens_seen": 58637550,
      "step": 2708,
      "time_per_iteration": 2.7163965702056885
    },
    {
      "auxiliary_loss_clip": 0.01101561,
      "auxiliary_loss_mlp": 0.00781217,
      "balance_loss_clip": 1.04896522,
      "balance_loss_mlp": 1.00177419,
      "epoch": 0.16287389147752893,
      "flos": 21251719848960.0,
      "grad_norm": 2.2850459718507654,
      "language_loss": 0.86162847,
      "learning_rate": 3.817635274679006e-06,
      "loss": 0.88045627,
      "num_input_tokens_seen": 58654135,
      "step": 2709,
      "time_per_iteration": 4.474989652633667
    },
    {
      "auxiliary_loss_clip": 0.0114031,
      "auxiliary_loss_mlp": 0.00777602,
      "balance_loss_clip": 1.05267572,
      "balance_loss_mlp": 1.00172114,
      "epoch": 0.1629340147301969,
      "flos": 19244672530560.0,
      "grad_norm": 2.581053296112052,
      "language_loss": 0.91410124,
      "learning_rate": 3.817472759295605e-06,
      "loss": 0.93328035,
      "num_input_tokens_seen": 58674320,
      "step": 2710,
      "time_per_iteration": 2.6951892375946045
    },
    {
      "auxiliary_loss_clip": 0.01118597,
      "auxiliary_loss_mlp": 0.01054854,
      "balance_loss_clip": 1.05254805,
      "balance_loss_mlp": 1.03451669,
      "epoch": 0.16299413798286488,
      "flos": 21249816428160.0,
      "grad_norm": 2.4322540773438437,
      "language_loss": 0.81690979,
      "learning_rate": 3.817310174993453e-06,
      "loss": 0.83864427,
      "num_input_tokens_seen": 58691000,
      "step": 2711,
      "time_per_iteration": 2.7854437828063965
    },
    {
      "auxiliary_loss_clip": 0.01146056,
      "auxiliary_loss_mlp": 0.01040648,
      "balance_loss_clip": 1.04954815,
      "balance_loss_mlp": 1.02107334,
      "epoch": 0.16305426123553285,
      "flos": 18770579896320.0,
      "grad_norm": 3.73256798888747,
      "language_loss": 0.8091476,
      "learning_rate": 3.817147521778719e-06,
      "loss": 0.83101463,
      "num_input_tokens_seen": 58710230,
      "step": 2712,
      "time_per_iteration": 2.834291458129883
    },
    {
      "auxiliary_loss_clip": 0.01171211,
      "auxiliary_loss_mlp": 0.01053015,
      "balance_loss_clip": 1.0590024,
      "balance_loss_mlp": 1.03273714,
      "epoch": 0.16311438448820081,
      "flos": 22087648137600.0,
      "grad_norm": 2.3460895846171996,
      "language_loss": 0.7681579,
      "learning_rate": 3.816984799657568e-06,
      "loss": 0.79040015,
      "num_input_tokens_seen": 58728610,
      "step": 2713,
      "time_per_iteration": 2.6188278198242188
    },
    {
      "auxiliary_loss_clip": 0.01156539,
      "auxiliary_loss_mlp": 0.0105792,
      "balance_loss_clip": 1.06240916,
      "balance_loss_mlp": 1.03832221,
      "epoch": 0.16317450774086878,
      "flos": 16467700164480.0,
      "grad_norm": 2.543173325075216,
      "language_loss": 0.79012156,
      "learning_rate": 3.8168220086361715e-06,
      "loss": 0.81226611,
      "num_input_tokens_seen": 58744385,
      "step": 2714,
      "time_per_iteration": 2.6534018516540527
    },
    {
      "auxiliary_loss_clip": 0.01149567,
      "auxiliary_loss_mlp": 0.01056152,
      "balance_loss_clip": 1.05467987,
      "balance_loss_mlp": 1.03724504,
      "epoch": 0.16323463099353674,
      "flos": 24352929308160.0,
      "grad_norm": 1.614702766215493,
      "language_loss": 0.77693665,
      "learning_rate": 3.816659148720702e-06,
      "loss": 0.79899377,
      "num_input_tokens_seen": 58763905,
      "step": 2715,
      "time_per_iteration": 2.856006383895874
    },
    {
      "auxiliary_loss_clip": 0.01129437,
      "auxiliary_loss_mlp": 0.01044046,
      "balance_loss_clip": 1.04810584,
      "balance_loss_mlp": 1.02525854,
      "epoch": 0.1632947542462047,
      "flos": 24900782520960.0,
      "grad_norm": 2.374975046722651,
      "language_loss": 0.81513858,
      "learning_rate": 3.816496219917336e-06,
      "loss": 0.83687335,
      "num_input_tokens_seen": 58785580,
      "step": 2716,
      "time_per_iteration": 2.6750845909118652
    },
    {
      "auxiliary_loss_clip": 0.01144393,
      "auxiliary_loss_mlp": 0.01055927,
      "balance_loss_clip": 1.05851114,
      "balance_loss_mlp": 1.03703237,
      "epoch": 0.1633548774988727,
      "flos": 24900279730560.0,
      "grad_norm": 1.8186679286330678,
      "language_loss": 0.86522418,
      "learning_rate": 3.816333222232251e-06,
      "loss": 0.88722742,
      "num_input_tokens_seen": 58806075,
      "step": 2717,
      "time_per_iteration": 2.761622428894043
    },
    {
      "auxiliary_loss_clip": 0.01135377,
      "auxiliary_loss_mlp": 0.01045964,
      "balance_loss_clip": 1.05334044,
      "balance_loss_mlp": 1.0274632,
      "epoch": 0.16341500075154067,
      "flos": 30441798357120.0,
      "grad_norm": 1.8799656187942837,
      "language_loss": 0.76924133,
      "learning_rate": 3.816170155671629e-06,
      "loss": 0.79105473,
      "num_input_tokens_seen": 58827405,
      "step": 2718,
      "time_per_iteration": 2.7946770191192627
    },
    {
      "auxiliary_loss_clip": 0.01145146,
      "auxiliary_loss_mlp": 0.01043682,
      "balance_loss_clip": 1.05553615,
      "balance_loss_mlp": 1.02566922,
      "epoch": 0.16347512400420863,
      "flos": 22784530878720.0,
      "grad_norm": 2.2449478392049906,
      "language_loss": 0.73827291,
      "learning_rate": 3.816007020241652e-06,
      "loss": 0.76016116,
      "num_input_tokens_seen": 58847205,
      "step": 2719,
      "time_per_iteration": 2.719980478286743
    },
    {
      "auxiliary_loss_clip": 0.01128361,
      "auxiliary_loss_mlp": 0.01045887,
      "balance_loss_clip": 1.04900515,
      "balance_loss_mlp": 1.02732563,
      "epoch": 0.1635352472568766,
      "flos": 22633274707200.0,
      "grad_norm": 1.7092252575708884,
      "language_loss": 0.72267497,
      "learning_rate": 3.815843815948507e-06,
      "loss": 0.74441749,
      "num_input_tokens_seen": 58866865,
      "step": 2720,
      "time_per_iteration": 2.8737292289733887
    },
    {
      "auxiliary_loss_clip": 0.01109456,
      "auxiliary_loss_mlp": 0.01049703,
      "balance_loss_clip": 1.05004287,
      "balance_loss_mlp": 1.02840054,
      "epoch": 0.16359537050954456,
      "flos": 15522998515200.0,
      "grad_norm": 2.1621365878543153,
      "language_loss": 0.75120997,
      "learning_rate": 3.8156805427983824e-06,
      "loss": 0.77280164,
      "num_input_tokens_seen": 58885200,
      "step": 2721,
      "time_per_iteration": 2.785296678543091
    },
    {
      "auxiliary_loss_clip": 0.01110342,
      "auxiliary_loss_mlp": 0.01059955,
      "balance_loss_clip": 1.04597676,
      "balance_loss_mlp": 1.03734064,
      "epoch": 0.16365549376221253,
      "flos": 22090162089600.0,
      "grad_norm": 1.9032438792006017,
      "language_loss": 0.79073942,
      "learning_rate": 3.8155172007974695e-06,
      "loss": 0.81244236,
      "num_input_tokens_seen": 58906385,
      "step": 2722,
      "time_per_iteration": 2.7850708961486816
    },
    {
      "auxiliary_loss_clip": 0.01149809,
      "auxiliary_loss_mlp": 0.00778798,
      "balance_loss_clip": 1.05395257,
      "balance_loss_mlp": 1.00171757,
      "epoch": 0.1637156170148805,
      "flos": 24060400945920.0,
      "grad_norm": 2.3019049903761215,
      "language_loss": 0.84954333,
      "learning_rate": 3.8153537899519624e-06,
      "loss": 0.86882937,
      "num_input_tokens_seen": 58925040,
      "step": 2723,
      "time_per_iteration": 2.7268764972686768
    },
    {
      "auxiliary_loss_clip": 0.01108328,
      "auxiliary_loss_mlp": 0.01044851,
      "balance_loss_clip": 1.04805517,
      "balance_loss_mlp": 1.02493143,
      "epoch": 0.1637757402675485,
      "flos": 26685362954880.0,
      "grad_norm": 1.8985615531712963,
      "language_loss": 0.71018666,
      "learning_rate": 3.815190310268058e-06,
      "loss": 0.73171842,
      "num_input_tokens_seen": 58944790,
      "step": 2724,
      "time_per_iteration": 2.7691783905029297
    },
    {
      "auxiliary_loss_clip": 0.01118053,
      "auxiliary_loss_mlp": 0.01041883,
      "balance_loss_clip": 1.05226958,
      "balance_loss_mlp": 1.02364373,
      "epoch": 0.16383586352021645,
      "flos": 16106941918080.0,
      "grad_norm": 2.1059770262776136,
      "language_loss": 0.70552838,
      "learning_rate": 3.815026761751955e-06,
      "loss": 0.72712779,
      "num_input_tokens_seen": 58962500,
      "step": 2725,
      "time_per_iteration": 2.6936957836151123
    },
    {
      "auxiliary_loss_clip": 0.01112368,
      "auxiliary_loss_mlp": 0.01046594,
      "balance_loss_clip": 1.04912174,
      "balance_loss_mlp": 1.028391,
      "epoch": 0.16389598677288442,
      "flos": 19165991788800.0,
      "grad_norm": 2.27810298992254,
      "language_loss": 0.88491893,
      "learning_rate": 3.814863144409855e-06,
      "loss": 0.90650856,
      "num_input_tokens_seen": 58980355,
      "step": 2726,
      "time_per_iteration": 2.7967143058776855
    },
    {
      "auxiliary_loss_clip": 0.01157668,
      "auxiliary_loss_mlp": 0.0105068,
      "balance_loss_clip": 1.06062055,
      "balance_loss_mlp": 1.03099847,
      "epoch": 0.16395611002555238,
      "flos": 21507008785920.0,
      "grad_norm": 2.0584475237926303,
      "language_loss": 0.7469939,
      "learning_rate": 3.814699458247963e-06,
      "loss": 0.7690773,
      "num_input_tokens_seen": 58999505,
      "step": 2727,
      "time_per_iteration": 2.6818623542785645
    },
    {
      "auxiliary_loss_clip": 0.01150971,
      "auxiliary_loss_mlp": 0.01052077,
      "balance_loss_clip": 1.0570507,
      "balance_loss_mlp": 1.03527999,
      "epoch": 0.16401623327822035,
      "flos": 21470918595840.0,
      "grad_norm": 1.6112579442237729,
      "language_loss": 0.83097756,
      "learning_rate": 3.8145357032724855e-06,
      "loss": 0.85300803,
      "num_input_tokens_seen": 59017930,
      "step": 2728,
      "time_per_iteration": 2.675360918045044
    },
    {
      "auxiliary_loss_clip": 0.01156153,
      "auxiliary_loss_mlp": 0.01045609,
      "balance_loss_clip": 1.05826735,
      "balance_loss_mlp": 1.02602315,
      "epoch": 0.1640763565308883,
      "flos": 13626232928640.0,
      "grad_norm": 2.5738755626941106,
      "language_loss": 0.84892929,
      "learning_rate": 3.814371879489633e-06,
      "loss": 0.87094688,
      "num_input_tokens_seen": 59035130,
      "step": 2729,
      "time_per_iteration": 2.7004599571228027
    },
    {
      "auxiliary_loss_clip": 0.01167293,
      "auxiliary_loss_mlp": 0.01048461,
      "balance_loss_clip": 1.0591594,
      "balance_loss_mlp": 1.03053224,
      "epoch": 0.16413647978355628,
      "flos": 15451464579840.0,
      "grad_norm": 1.9897225699042427,
      "language_loss": 0.72895479,
      "learning_rate": 3.814207986905616e-06,
      "loss": 0.75111228,
      "num_input_tokens_seen": 59053080,
      "step": 2730,
      "time_per_iteration": 2.593179702758789
    },
    {
      "auxiliary_loss_clip": 0.01142509,
      "auxiliary_loss_mlp": 0.01050071,
      "balance_loss_clip": 1.05208349,
      "balance_loss_mlp": 1.02908981,
      "epoch": 0.16419660303622427,
      "flos": 45878682015360.0,
      "grad_norm": 1.6754501336017709,
      "language_loss": 0.74384654,
      "learning_rate": 3.814044025526651e-06,
      "loss": 0.76577234,
      "num_input_tokens_seen": 59075610,
      "step": 2731,
      "time_per_iteration": 2.8702962398529053
    },
    {
      "auxiliary_loss_clip": 0.01122791,
      "auxiliary_loss_mlp": 0.01047176,
      "balance_loss_clip": 1.05006754,
      "balance_loss_mlp": 1.02650499,
      "epoch": 0.16425672628889224,
      "flos": 18952826526720.0,
      "grad_norm": 2.031351475505915,
      "language_loss": 0.79190683,
      "learning_rate": 3.8138799953589548e-06,
      "loss": 0.8136065,
      "num_input_tokens_seen": 59094555,
      "step": 2732,
      "time_per_iteration": 2.734529972076416
    },
    {
      "auxiliary_loss_clip": 0.01141118,
      "auxiliary_loss_mlp": 0.01047385,
      "balance_loss_clip": 1.05340672,
      "balance_loss_mlp": 1.02796555,
      "epoch": 0.1643168495415602,
      "flos": 24312996362880.0,
      "grad_norm": 2.250003976384769,
      "language_loss": 0.69526887,
      "learning_rate": 3.8137158964087473e-06,
      "loss": 0.71715385,
      "num_input_tokens_seen": 59113515,
      "step": 2733,
      "time_per_iteration": 2.672377109527588
    },
    {
      "auxiliary_loss_clip": 0.01143332,
      "auxiliary_loss_mlp": 0.01053232,
      "balance_loss_clip": 1.05603123,
      "balance_loss_mlp": 1.0325135,
      "epoch": 0.16437697279422817,
      "flos": 26428421992320.0,
      "grad_norm": 2.000873580428856,
      "language_loss": 0.80976766,
      "learning_rate": 3.8135517286822508e-06,
      "loss": 0.83173329,
      "num_input_tokens_seen": 59133275,
      "step": 2734,
      "time_per_iteration": 2.710293769836426
    },
    {
      "auxiliary_loss_clip": 0.01135758,
      "auxiliary_loss_mlp": 0.01056722,
      "balance_loss_clip": 1.05488348,
      "balance_loss_mlp": 1.03470409,
      "epoch": 0.16443709604689613,
      "flos": 34532239351680.0,
      "grad_norm": 2.100664117201308,
      "language_loss": 0.81810421,
      "learning_rate": 3.8133874921856914e-06,
      "loss": 0.840029,
      "num_input_tokens_seen": 59154095,
      "step": 2735,
      "time_per_iteration": 2.8074140548706055
    },
    {
      "auxiliary_loss_clip": 0.01070875,
      "auxiliary_loss_mlp": 0.01044313,
      "balance_loss_clip": 1.04323888,
      "balance_loss_mlp": 1.02508426,
      "epoch": 0.1644972192995641,
      "flos": 23258048895360.0,
      "grad_norm": 2.405088987017839,
      "language_loss": 0.78515649,
      "learning_rate": 3.813223186925296e-06,
      "loss": 0.80630839,
      "num_input_tokens_seen": 59173795,
      "step": 2736,
      "time_per_iteration": 2.839087963104248
    },
    {
      "auxiliary_loss_clip": 0.01147998,
      "auxiliary_loss_mlp": 0.01054659,
      "balance_loss_clip": 1.05859447,
      "balance_loss_mlp": 1.03513288,
      "epoch": 0.1645573425522321,
      "flos": 26979543342720.0,
      "grad_norm": 1.9462182296456145,
      "language_loss": 0.81052899,
      "learning_rate": 3.8130588129072964e-06,
      "loss": 0.83255553,
      "num_input_tokens_seen": 59191610,
      "step": 2737,
      "time_per_iteration": 2.7328996658325195
    },
    {
      "auxiliary_loss_clip": 0.01150424,
      "auxiliary_loss_mlp": 0.01052207,
      "balance_loss_clip": 1.0559026,
      "balance_loss_mlp": 1.03065443,
      "epoch": 0.16461746580490005,
      "flos": 28731768600960.0,
      "grad_norm": 1.8596348168124566,
      "language_loss": 0.87449318,
      "learning_rate": 3.8128943701379246e-06,
      "loss": 0.89651948,
      "num_input_tokens_seen": 59213000,
      "step": 2738,
      "time_per_iteration": 2.7345526218414307
    },
    {
      "auxiliary_loss_clip": 0.01139154,
      "auxiliary_loss_mlp": 0.0106055,
      "balance_loss_clip": 1.05534518,
      "balance_loss_mlp": 1.04079759,
      "epoch": 0.16467758905756802,
      "flos": 24930156867840.0,
      "grad_norm": 1.728421510231393,
      "language_loss": 0.71997833,
      "learning_rate": 3.8127298586234167e-06,
      "loss": 0.74197543,
      "num_input_tokens_seen": 59232340,
      "step": 2739,
      "time_per_iteration": 2.7091422080993652
    },
    {
      "auxiliary_loss_clip": 0.01154419,
      "auxiliary_loss_mlp": 0.0105106,
      "balance_loss_clip": 1.05673754,
      "balance_loss_mlp": 1.0312835,
      "epoch": 0.16473771231023598,
      "flos": 24826519152000.0,
      "grad_norm": 1.8559436932352185,
      "language_loss": 0.81645715,
      "learning_rate": 3.8125652783700104e-06,
      "loss": 0.83851194,
      "num_input_tokens_seen": 59253950,
      "step": 2740,
      "time_per_iteration": 2.712658166885376
    },
    {
      "auxiliary_loss_clip": 0.01114061,
      "auxiliary_loss_mlp": 0.01068725,
      "balance_loss_clip": 1.04991829,
      "balance_loss_mlp": 1.04307163,
      "epoch": 0.16479783556290395,
      "flos": 39896072375040.0,
      "grad_norm": 2.0528021789830837,
      "language_loss": 0.69467485,
      "learning_rate": 3.8124006293839475e-06,
      "loss": 0.71650267,
      "num_input_tokens_seen": 59275545,
      "step": 2741,
      "time_per_iteration": 2.8629493713378906
    },
    {
      "auxiliary_loss_clip": 0.01167543,
      "auxiliary_loss_mlp": 0.01048721,
      "balance_loss_clip": 1.05907226,
      "balance_loss_mlp": 1.02906334,
      "epoch": 0.16485795881557191,
      "flos": 19897061299200.0,
      "grad_norm": 1.7765193730452222,
      "language_loss": 0.79811072,
      "learning_rate": 3.812235911671472e-06,
      "loss": 0.8202734,
      "num_input_tokens_seen": 59293480,
      "step": 2742,
      "time_per_iteration": 2.626775026321411
    },
    {
      "auxiliary_loss_clip": 0.01141681,
      "auxiliary_loss_mlp": 0.01055663,
      "balance_loss_clip": 1.05664062,
      "balance_loss_mlp": 1.03477716,
      "epoch": 0.16491808206823988,
      "flos": 20556129997440.0,
      "grad_norm": 1.91797408289014,
      "language_loss": 0.8499459,
      "learning_rate": 3.8120711252388274e-06,
      "loss": 0.87191939,
      "num_input_tokens_seen": 59313435,
      "step": 2743,
      "time_per_iteration": 2.8218302726745605
    },
    {
      "auxiliary_loss_clip": 0.01162447,
      "auxiliary_loss_mlp": 0.01051969,
      "balance_loss_clip": 1.05743837,
      "balance_loss_mlp": 1.03196514,
      "epoch": 0.16497820532090787,
      "flos": 23800802376960.0,
      "grad_norm": 1.4425200129075006,
      "language_loss": 0.85558498,
      "learning_rate": 3.811906270092265e-06,
      "loss": 0.87772918,
      "num_input_tokens_seen": 59331535,
      "step": 2744,
      "time_per_iteration": 4.206263542175293
    },
    {
      "auxiliary_loss_clip": 0.01131671,
      "auxiliary_loss_mlp": 0.0104676,
      "balance_loss_clip": 1.05206287,
      "balance_loss_mlp": 1.02812767,
      "epoch": 0.16503832857357584,
      "flos": 25482642935040.0,
      "grad_norm": 1.6285200980820358,
      "language_loss": 0.82770813,
      "learning_rate": 3.811741346238036e-06,
      "loss": 0.84949243,
      "num_input_tokens_seen": 59350680,
      "step": 2745,
      "time_per_iteration": 4.331594467163086
    },
    {
      "auxiliary_loss_clip": 0.011344,
      "auxiliary_loss_mlp": 0.01057242,
      "balance_loss_clip": 1.05874014,
      "balance_loss_mlp": 1.03825223,
      "epoch": 0.1650984518262438,
      "flos": 17676058619520.0,
      "grad_norm": 6.766690288332402,
      "language_loss": 0.76811314,
      "learning_rate": 3.8115763536823923e-06,
      "loss": 0.79002959,
      "num_input_tokens_seen": 59367020,
      "step": 2746,
      "time_per_iteration": 4.225586414337158
    },
    {
      "auxiliary_loss_clip": 0.01164296,
      "auxiliary_loss_mlp": 0.01055636,
      "balance_loss_clip": 1.05781221,
      "balance_loss_mlp": 1.03533494,
      "epoch": 0.16515857507891177,
      "flos": 18698327688960.0,
      "grad_norm": 1.9760186874049024,
      "language_loss": 0.80818808,
      "learning_rate": 3.811411292431592e-06,
      "loss": 0.83038735,
      "num_input_tokens_seen": 59386075,
      "step": 2747,
      "time_per_iteration": 2.6862480640411377
    },
    {
      "auxiliary_loss_clip": 0.01157975,
      "auxiliary_loss_mlp": 0.0104673,
      "balance_loss_clip": 1.05990267,
      "balance_loss_mlp": 1.02664328,
      "epoch": 0.16521869833157973,
      "flos": 15010481306880.0,
      "grad_norm": 2.0608482379031337,
      "language_loss": 0.69433749,
      "learning_rate": 3.8112461624918945e-06,
      "loss": 0.71638453,
      "num_input_tokens_seen": 59402690,
      "step": 2748,
      "time_per_iteration": 2.6520986557006836
    },
    {
      "auxiliary_loss_clip": 0.01169692,
      "auxiliary_loss_mlp": 0.00778195,
      "balance_loss_clip": 1.06237423,
      "balance_loss_mlp": 1.00173104,
      "epoch": 0.1652788215842477,
      "flos": 22121152548480.0,
      "grad_norm": 2.259215537482641,
      "language_loss": 0.88012803,
      "learning_rate": 3.811080963869561e-06,
      "loss": 0.89960694,
      "num_input_tokens_seen": 59421130,
      "step": 2749,
      "time_per_iteration": 4.260679244995117
    },
    {
      "auxiliary_loss_clip": 0.01154179,
      "auxiliary_loss_mlp": 0.01045617,
      "balance_loss_clip": 1.05586052,
      "balance_loss_mlp": 1.02542281,
      "epoch": 0.16533894483691566,
      "flos": 18333080242560.0,
      "grad_norm": 2.0880864906339864,
      "language_loss": 0.79240286,
      "learning_rate": 3.8109156965708557e-06,
      "loss": 0.81440079,
      "num_input_tokens_seen": 59438970,
      "step": 2750,
      "time_per_iteration": 2.6335251331329346
    },
    {
      "auxiliary_loss_clip": 0.01153343,
      "auxiliary_loss_mlp": 0.0104591,
      "balance_loss_clip": 1.0579437,
      "balance_loss_mlp": 1.02602625,
      "epoch": 0.16539906808958366,
      "flos": 22382115834240.0,
      "grad_norm": 1.6952801391084946,
      "language_loss": 0.94854712,
      "learning_rate": 3.8107503606020455e-06,
      "loss": 0.97053963,
      "num_input_tokens_seen": 59458510,
      "step": 2751,
      "time_per_iteration": 2.697174310684204
    },
    {
      "auxiliary_loss_clip": 0.0106803,
      "auxiliary_loss_mlp": 0.0105236,
      "balance_loss_clip": 1.04625726,
      "balance_loss_mlp": 1.03247619,
      "epoch": 0.16545919134225162,
      "flos": 22711093522560.0,
      "grad_norm": 2.614588592950962,
      "language_loss": 0.71231711,
      "learning_rate": 3.8105849559693997e-06,
      "loss": 0.73352098,
      "num_input_tokens_seen": 59477110,
      "step": 2752,
      "time_per_iteration": 2.7780745029449463
    },
    {
      "auxiliary_loss_clip": 0.01090521,
      "auxiliary_loss_mlp": 0.01022104,
      "balance_loss_clip": 1.05741131,
      "balance_loss_mlp": 1.01941013,
      "epoch": 0.1655193145949196,
      "flos": 67802974076160.0,
      "grad_norm": 0.7721529651221379,
      "language_loss": 0.54058975,
      "learning_rate": 3.810419482679192e-06,
      "loss": 0.56171602,
      "num_input_tokens_seen": 59541155,
      "step": 2753,
      "time_per_iteration": 3.3371469974517822
    },
    {
      "auxiliary_loss_clip": 0.01163808,
      "auxiliary_loss_mlp": 0.00778536,
      "balance_loss_clip": 1.05587018,
      "balance_loss_mlp": 1.00172091,
      "epoch": 0.16557943784758755,
      "flos": 24280389792000.0,
      "grad_norm": 1.6411537728312637,
      "language_loss": 0.75436741,
      "learning_rate": 3.8102539407376954e-06,
      "loss": 0.7737909,
      "num_input_tokens_seen": 59561155,
      "step": 2754,
      "time_per_iteration": 2.6382133960723877
    },
    {
      "auxiliary_loss_clip": 0.01139421,
      "auxiliary_loss_mlp": 0.01060584,
      "balance_loss_clip": 1.05406713,
      "balance_loss_mlp": 1.03768396,
      "epoch": 0.16563956110025552,
      "flos": 20083617561600.0,
      "grad_norm": 2.4067479946694137,
      "language_loss": 0.86654639,
      "learning_rate": 3.810088330151188e-06,
      "loss": 0.88854647,
      "num_input_tokens_seen": 59580460,
      "step": 2755,
      "time_per_iteration": 2.6590075492858887
    },
    {
      "auxiliary_loss_clip": 0.01122817,
      "auxiliary_loss_mlp": 0.01053169,
      "balance_loss_clip": 1.04948378,
      "balance_loss_mlp": 1.03293943,
      "epoch": 0.16569968435292348,
      "flos": 28034454896640.0,
      "grad_norm": 1.7268487777137649,
      "language_loss": 0.73350251,
      "learning_rate": 3.80992265092595e-06,
      "loss": 0.75526237,
      "num_input_tokens_seen": 59600025,
      "step": 2756,
      "time_per_iteration": 2.771820545196533
    },
    {
      "auxiliary_loss_clip": 0.01128662,
      "auxiliary_loss_mlp": 0.01049666,
      "balance_loss_clip": 1.05550277,
      "balance_loss_mlp": 1.02969813,
      "epoch": 0.16575980760559147,
      "flos": 26250233598720.0,
      "grad_norm": 1.5540667033085804,
      "language_loss": 0.75308084,
      "learning_rate": 3.8097569030682636e-06,
      "loss": 0.77486414,
      "num_input_tokens_seen": 59620600,
      "step": 2757,
      "time_per_iteration": 2.8106157779693604
    },
    {
      "auxiliary_loss_clip": 0.01143608,
      "auxiliary_loss_mlp": 0.01054064,
      "balance_loss_clip": 1.057634,
      "balance_loss_mlp": 1.03390563,
      "epoch": 0.16581993085825944,
      "flos": 26943955943040.0,
      "grad_norm": 1.8675154897424497,
      "language_loss": 0.84604371,
      "learning_rate": 3.8095910865844137e-06,
      "loss": 0.86802036,
      "num_input_tokens_seen": 59641385,
      "step": 2758,
      "time_per_iteration": 2.8663368225097656
    },
    {
      "auxiliary_loss_clip": 0.01168186,
      "auxiliary_loss_mlp": 0.01058337,
      "balance_loss_clip": 1.06166434,
      "balance_loss_mlp": 1.03952527,
      "epoch": 0.1658800541109274,
      "flos": 21653632103040.0,
      "grad_norm": 2.0824774555850243,
      "language_loss": 0.78848934,
      "learning_rate": 3.809425201480689e-06,
      "loss": 0.81075454,
      "num_input_tokens_seen": 59659865,
      "step": 2759,
      "time_per_iteration": 2.655371904373169
    },
    {
      "auxiliary_loss_clip": 0.01098973,
      "auxiliary_loss_mlp": 0.0104879,
      "balance_loss_clip": 1.0491066,
      "balance_loss_mlp": 1.02846527,
      "epoch": 0.16594017736359537,
      "flos": 16435488643200.0,
      "grad_norm": 2.4005603702739613,
      "language_loss": 0.75130272,
      "learning_rate": 3.8092592477633793e-06,
      "loss": 0.77278036,
      "num_input_tokens_seen": 59678780,
      "step": 2760,
      "time_per_iteration": 2.767866611480713
    },
    {
      "auxiliary_loss_clip": 0.01117278,
      "auxiliary_loss_mlp": 0.0104823,
      "balance_loss_clip": 1.05129814,
      "balance_loss_mlp": 1.02867997,
      "epoch": 0.16600030061626334,
      "flos": 22637297030400.0,
      "grad_norm": 1.5792623632565632,
      "language_loss": 0.73425764,
      "learning_rate": 3.8090932254387774e-06,
      "loss": 0.75591272,
      "num_input_tokens_seen": 59698795,
      "step": 2761,
      "time_per_iteration": 2.762836456298828
    },
    {
      "auxiliary_loss_clip": 0.0113507,
      "auxiliary_loss_mlp": 0.01050415,
      "balance_loss_clip": 1.05250192,
      "balance_loss_mlp": 1.03018475,
      "epoch": 0.1660604238689313,
      "flos": 26396569607040.0,
      "grad_norm": 2.9515424803015033,
      "language_loss": 0.88832974,
      "learning_rate": 3.8089271345131788e-06,
      "loss": 0.91018462,
      "num_input_tokens_seen": 59718795,
      "step": 2762,
      "time_per_iteration": 2.766324281692505
    },
    {
      "auxiliary_loss_clip": 0.01115163,
      "auxiliary_loss_mlp": 0.01050144,
      "balance_loss_clip": 1.05208707,
      "balance_loss_mlp": 1.03080845,
      "epoch": 0.16612054712159927,
      "flos": 23039999383680.0,
      "grad_norm": 1.84507980271118,
      "language_loss": 0.87992418,
      "learning_rate": 3.8087609749928822e-06,
      "loss": 0.90157735,
      "num_input_tokens_seen": 59737555,
      "step": 2763,
      "time_per_iteration": 2.7734055519104004
    },
    {
      "auxiliary_loss_clip": 0.01086152,
      "auxiliary_loss_mlp": 0.01013622,
      "balance_loss_clip": 1.0448606,
      "balance_loss_mlp": 1.01065338,
      "epoch": 0.16618067037426726,
      "flos": 59241225202560.0,
      "grad_norm": 0.7790832079967882,
      "language_loss": 0.59799927,
      "learning_rate": 3.8085947468841885e-06,
      "loss": 0.61899698,
      "num_input_tokens_seen": 59800915,
      "step": 2764,
      "time_per_iteration": 3.1728692054748535
    },
    {
      "auxiliary_loss_clip": 0.01152232,
      "auxiliary_loss_mlp": 0.01053607,
      "balance_loss_clip": 1.05467176,
      "balance_loss_mlp": 1.03254318,
      "epoch": 0.16624079362693522,
      "flos": 27198813916800.0,
      "grad_norm": 1.7436496772383425,
      "language_loss": 0.82260036,
      "learning_rate": 3.808428450193401e-06,
      "loss": 0.84465873,
      "num_input_tokens_seen": 59822910,
      "step": 2765,
      "time_per_iteration": 2.72440767288208
    },
    {
      "auxiliary_loss_clip": 0.01171844,
      "auxiliary_loss_mlp": 0.01049085,
      "balance_loss_clip": 1.05882454,
      "balance_loss_mlp": 1.02746069,
      "epoch": 0.1663009168796032,
      "flos": 10925068216320.0,
      "grad_norm": 2.128015994498251,
      "language_loss": 0.69980019,
      "learning_rate": 3.8082620849268244e-06,
      "loss": 0.72200948,
      "num_input_tokens_seen": 59838805,
      "step": 2766,
      "time_per_iteration": 2.5810647010803223
    },
    {
      "auxiliary_loss_clip": 0.0115036,
      "auxiliary_loss_mlp": 0.01047665,
      "balance_loss_clip": 1.05772817,
      "balance_loss_mlp": 1.02792454,
      "epoch": 0.16636104013227115,
      "flos": 17894431353600.0,
      "grad_norm": 2.107381123394178,
      "language_loss": 0.8845337,
      "learning_rate": 3.808095651090769e-06,
      "loss": 0.90651393,
      "num_input_tokens_seen": 59855345,
      "step": 2767,
      "time_per_iteration": 2.659240245819092
    },
    {
      "auxiliary_loss_clip": 0.01077283,
      "auxiliary_loss_mlp": 0.01002999,
      "balance_loss_clip": 1.046556,
      "balance_loss_mlp": 1.00020981,
      "epoch": 0.16642116338493912,
      "flos": 66726050463360.0,
      "grad_norm": 0.6403612433239105,
      "language_loss": 0.5289067,
      "learning_rate": 3.8079291486915447e-06,
      "loss": 0.54970956,
      "num_input_tokens_seen": 59917710,
      "step": 2768,
      "time_per_iteration": 3.28488826751709
    },
    {
      "auxiliary_loss_clip": 0.01137637,
      "auxiliary_loss_mlp": 0.01051692,
      "balance_loss_clip": 1.05451822,
      "balance_loss_mlp": 1.03034163,
      "epoch": 0.16648128663760708,
      "flos": 19026048401280.0,
      "grad_norm": 2.4342686570828267,
      "language_loss": 0.84962058,
      "learning_rate": 3.8077625777354667e-06,
      "loss": 0.87151396,
      "num_input_tokens_seen": 59935105,
      "step": 2769,
      "time_per_iteration": 2.753257989883423
    },
    {
      "auxiliary_loss_clip": 0.01068987,
      "auxiliary_loss_mlp": 0.0100573,
      "balance_loss_clip": 1.04678345,
      "balance_loss_mlp": 1.00316668,
      "epoch": 0.16654140989027508,
      "flos": 70134976759680.0,
      "grad_norm": 0.8107434108728753,
      "language_loss": 0.57455683,
      "learning_rate": 3.80759593822885e-06,
      "loss": 0.59530401,
      "num_input_tokens_seen": 59984085,
      "step": 2770,
      "time_per_iteration": 3.2202906608581543
    },
    {
      "auxiliary_loss_clip": 0.01054548,
      "auxiliary_loss_mlp": 0.01003676,
      "balance_loss_clip": 1.04637623,
      "balance_loss_mlp": 1.00086308,
      "epoch": 0.16660153314294304,
      "flos": 70272406195200.0,
      "grad_norm": 0.8940719168038874,
      "language_loss": 0.56241393,
      "learning_rate": 3.807429230178015e-06,
      "loss": 0.58299619,
      "num_input_tokens_seen": 60043470,
      "step": 2771,
      "time_per_iteration": 3.3302085399627686
    },
    {
      "auxiliary_loss_clip": 0.01110714,
      "auxiliary_loss_mlp": 0.01053994,
      "balance_loss_clip": 1.04819679,
      "balance_loss_mlp": 1.03316772,
      "epoch": 0.166661656395611,
      "flos": 23075048079360.0,
      "grad_norm": 2.9137693497887778,
      "language_loss": 0.70419657,
      "learning_rate": 3.8072624535892817e-06,
      "loss": 0.72584367,
      "num_input_tokens_seen": 60063045,
      "step": 2772,
      "time_per_iteration": 2.845414161682129
    },
    {
      "auxiliary_loss_clip": 0.0114592,
      "auxiliary_loss_mlp": 0.01049708,
      "balance_loss_clip": 1.05082583,
      "balance_loss_mlp": 1.02923954,
      "epoch": 0.16672177964827897,
      "flos": 28366341586560.0,
      "grad_norm": 2.20945076195277,
      "language_loss": 0.86324167,
      "learning_rate": 3.807095608468975e-06,
      "loss": 0.88519788,
      "num_input_tokens_seen": 60081945,
      "step": 2773,
      "time_per_iteration": 2.669412851333618
    },
    {
      "auxiliary_loss_clip": 0.01095425,
      "auxiliary_loss_mlp": 0.01049097,
      "balance_loss_clip": 1.04436934,
      "balance_loss_mlp": 1.0300827,
      "epoch": 0.16678190290094694,
      "flos": 19091010147840.0,
      "grad_norm": 2.0211952616678937,
      "language_loss": 0.82141376,
      "learning_rate": 3.8069286948234224e-06,
      "loss": 0.84285897,
      "num_input_tokens_seen": 60096820,
      "step": 2774,
      "time_per_iteration": 2.7111308574676514
    },
    {
      "auxiliary_loss_clip": 0.01123493,
      "auxiliary_loss_mlp": 0.01045144,
      "balance_loss_clip": 1.05252421,
      "balance_loss_mlp": 1.02446127,
      "epoch": 0.1668420261536149,
      "flos": 21799106184960.0,
      "grad_norm": 3.3781068524499,
      "language_loss": 0.8298822,
      "learning_rate": 3.806761712658952e-06,
      "loss": 0.85156858,
      "num_input_tokens_seen": 60116140,
      "step": 2775,
      "time_per_iteration": 2.7367632389068604
    },
    {
      "auxiliary_loss_clip": 0.01150495,
      "auxiliary_loss_mlp": 0.01051475,
      "balance_loss_clip": 1.05761933,
      "balance_loss_mlp": 1.03264022,
      "epoch": 0.16690214940628287,
      "flos": 19062533640960.0,
      "grad_norm": 1.8115651629444076,
      "language_loss": 0.80919641,
      "learning_rate": 3.806594661981897e-06,
      "loss": 0.8312161,
      "num_input_tokens_seen": 60134235,
      "step": 2776,
      "time_per_iteration": 2.651723623275757
    },
    {
      "auxiliary_loss_clip": 0.0113775,
      "auxiliary_loss_mlp": 0.01054199,
      "balance_loss_clip": 1.05518723,
      "balance_loss_mlp": 1.0346483,
      "epoch": 0.16696227265895086,
      "flos": 18588548747520.0,
      "grad_norm": 2.7510345221850336,
      "language_loss": 0.80203485,
      "learning_rate": 3.8064275427985906e-06,
      "loss": 0.82395434,
      "num_input_tokens_seen": 60153275,
      "step": 2777,
      "time_per_iteration": 2.6380929946899414
    },
    {
      "auxiliary_loss_clip": 0.01147967,
      "auxiliary_loss_mlp": 0.01045166,
      "balance_loss_clip": 1.05270481,
      "balance_loss_mlp": 1.02640271,
      "epoch": 0.16702239591161883,
      "flos": 23294139085440.0,
      "grad_norm": 1.6179722336290305,
      "language_loss": 0.85384095,
      "learning_rate": 3.806260355115371e-06,
      "loss": 0.87577224,
      "num_input_tokens_seen": 60173215,
      "step": 2778,
      "time_per_iteration": 2.754652500152588
    },
    {
      "auxiliary_loss_clip": 0.01136802,
      "auxiliary_loss_mlp": 0.01040643,
      "balance_loss_clip": 1.0531714,
      "balance_loss_mlp": 1.02148652,
      "epoch": 0.1670825191642868,
      "flos": 24425648392320.0,
      "grad_norm": 3.2091470007324414,
      "language_loss": 0.74180603,
      "learning_rate": 3.8060930989385778e-06,
      "loss": 0.76358056,
      "num_input_tokens_seen": 60190515,
      "step": 2779,
      "time_per_iteration": 2.777193784713745
    },
    {
      "auxiliary_loss_clip": 0.01112683,
      "auxiliary_loss_mlp": 0.00777451,
      "balance_loss_clip": 1.04981184,
      "balance_loss_mlp": 1.0015173,
      "epoch": 0.16714264241695476,
      "flos": 26797512193920.0,
      "grad_norm": 2.127789274190337,
      "language_loss": 0.6557346,
      "learning_rate": 3.805925774274554e-06,
      "loss": 0.67463589,
      "num_input_tokens_seen": 60211655,
      "step": 2780,
      "time_per_iteration": 2.896976947784424
    },
    {
      "auxiliary_loss_clip": 0.01120921,
      "auxiliary_loss_mlp": 0.01045506,
      "balance_loss_clip": 1.04843462,
      "balance_loss_mlp": 1.02547836,
      "epoch": 0.16720276566962272,
      "flos": 21835304115840.0,
      "grad_norm": 2.46647860258999,
      "language_loss": 0.78422606,
      "learning_rate": 3.805758381129643e-06,
      "loss": 0.80589032,
      "num_input_tokens_seen": 60230860,
      "step": 2781,
      "time_per_iteration": 2.725782632827759
    },
    {
      "auxiliary_loss_clip": 0.01094692,
      "auxiliary_loss_mlp": 0.01050104,
      "balance_loss_clip": 1.04439843,
      "balance_loss_mlp": 1.03056526,
      "epoch": 0.1672628889222907,
      "flos": 21470415805440.0,
      "grad_norm": 26.23767952829368,
      "language_loss": 0.75119764,
      "learning_rate": 3.805590919510193e-06,
      "loss": 0.77264553,
      "num_input_tokens_seen": 60250535,
      "step": 2782,
      "time_per_iteration": 2.7064197063446045
    },
    {
      "auxiliary_loss_clip": 0.01129162,
      "auxiliary_loss_mlp": 0.01047612,
      "balance_loss_clip": 1.05152631,
      "balance_loss_mlp": 1.02764392,
      "epoch": 0.16732301217495865,
      "flos": 30774008269440.0,
      "grad_norm": 2.116531296279042,
      "language_loss": 0.67398441,
      "learning_rate": 3.8054233894225547e-06,
      "loss": 0.69575214,
      "num_input_tokens_seen": 60269530,
      "step": 2783,
      "time_per_iteration": 2.7901556491851807
    },
    {
      "auxiliary_loss_clip": 0.01158882,
      "auxiliary_loss_mlp": 0.0105166,
      "balance_loss_clip": 1.05460215,
      "balance_loss_mlp": 1.03271747,
      "epoch": 0.16738313542762664,
      "flos": 23474625949440.0,
      "grad_norm": 1.7768362036873409,
      "language_loss": 0.69919086,
      "learning_rate": 3.805255790873081e-06,
      "loss": 0.72129631,
      "num_input_tokens_seen": 60289900,
      "step": 2784,
      "time_per_iteration": 5.714844226837158
    },
    {
      "auxiliary_loss_clip": 0.01137618,
      "auxiliary_loss_mlp": 0.01056022,
      "balance_loss_clip": 1.05217624,
      "balance_loss_mlp": 1.03539932,
      "epoch": 0.1674432586802946,
      "flos": 29789086366080.0,
      "grad_norm": 4.741795209709136,
      "language_loss": 0.60970068,
      "learning_rate": 3.805088123868126e-06,
      "loss": 0.6316371,
      "num_input_tokens_seen": 60310025,
      "step": 2785,
      "time_per_iteration": 4.219547510147095
    },
    {
      "auxiliary_loss_clip": 0.01057886,
      "auxiliary_loss_mlp": 0.0100398,
      "balance_loss_clip": 1.03758883,
      "balance_loss_mlp": 1.00141752,
      "epoch": 0.16750338193296258,
      "flos": 66136073575680.0,
      "grad_norm": 0.773077721474628,
      "language_loss": 0.58780885,
      "learning_rate": 3.8049203884140492e-06,
      "loss": 0.60842752,
      "num_input_tokens_seen": 60377800,
      "step": 2786,
      "time_per_iteration": 3.2306320667266846
    },
    {
      "auxiliary_loss_clip": 0.0113927,
      "auxiliary_loss_mlp": 0.01044966,
      "balance_loss_clip": 1.0496738,
      "balance_loss_mlp": 1.02589226,
      "epoch": 0.16756350518563054,
      "flos": 25696777864320.0,
      "grad_norm": 1.7333132735183339,
      "language_loss": 0.76308596,
      "learning_rate": 3.80475258451721e-06,
      "loss": 0.78492826,
      "num_input_tokens_seen": 60398215,
      "step": 2787,
      "time_per_iteration": 2.6434125900268555
    },
    {
      "auxiliary_loss_clip": 0.01146924,
      "auxiliary_loss_mlp": 0.01043386,
      "balance_loss_clip": 1.0529089,
      "balance_loss_mlp": 1.02544546,
      "epoch": 0.1676236284382985,
      "flos": 23836102467840.0,
      "grad_norm": 1.7210472408736244,
      "language_loss": 0.7717936,
      "learning_rate": 3.804584712183972e-06,
      "loss": 0.79369676,
      "num_input_tokens_seen": 60416910,
      "step": 2788,
      "time_per_iteration": 4.359618425369263
    },
    {
      "auxiliary_loss_clip": 0.01054629,
      "auxiliary_loss_mlp": 0.00999991,
      "balance_loss_clip": 1.03482509,
      "balance_loss_mlp": 0.99746382,
      "epoch": 0.16768375169096647,
      "flos": 59874902985600.0,
      "grad_norm": 0.8596744797543817,
      "language_loss": 0.59331679,
      "learning_rate": 3.8044167714207013e-06,
      "loss": 0.61386299,
      "num_input_tokens_seen": 60468660,
      "step": 2789,
      "time_per_iteration": 3.0742650032043457
    },
    {
      "auxiliary_loss_clip": 0.01148272,
      "auxiliary_loss_mlp": 0.01053856,
      "balance_loss_clip": 1.05450928,
      "balance_loss_mlp": 1.03428209,
      "epoch": 0.16774387494363446,
      "flos": 38435657207040.0,
      "grad_norm": 1.689036486923415,
      "language_loss": 0.7012763,
      "learning_rate": 3.804248762233765e-06,
      "loss": 0.7232976,
      "num_input_tokens_seen": 60492370,
      "step": 2790,
      "time_per_iteration": 2.872232437133789
    },
    {
      "auxiliary_loss_clip": 0.0112492,
      "auxiliary_loss_mlp": 0.01051622,
      "balance_loss_clip": 1.0497216,
      "balance_loss_mlp": 1.0334661,
      "epoch": 0.16780399819630243,
      "flos": 22637620252800.0,
      "grad_norm": 1.864386369112868,
      "language_loss": 0.79464513,
      "learning_rate": 3.8040806846295356e-06,
      "loss": 0.81641054,
      "num_input_tokens_seen": 60512655,
      "step": 2791,
      "time_per_iteration": 2.7180140018463135
    },
    {
      "auxiliary_loss_clip": 0.01122456,
      "auxiliary_loss_mlp": 0.01050939,
      "balance_loss_clip": 1.04977369,
      "balance_loss_mlp": 1.03106701,
      "epoch": 0.1678641214489704,
      "flos": 32891516887680.0,
      "grad_norm": 1.705849915566178,
      "language_loss": 0.71547955,
      "learning_rate": 3.8039125386143853e-06,
      "loss": 0.73721349,
      "num_input_tokens_seen": 60533090,
      "step": 2792,
      "time_per_iteration": 2.9221818447113037
    },
    {
      "auxiliary_loss_clip": 0.01131469,
      "auxiliary_loss_mlp": 0.01044061,
      "balance_loss_clip": 1.05479562,
      "balance_loss_mlp": 1.02551246,
      "epoch": 0.16792424470163836,
      "flos": 19974916028160.0,
      "grad_norm": 1.9301593564774673,
      "language_loss": 0.71581644,
      "learning_rate": 3.803744324194691e-06,
      "loss": 0.73757172,
      "num_input_tokens_seen": 60553190,
      "step": 2793,
      "time_per_iteration": 2.75104022026062
    },
    {
      "auxiliary_loss_clip": 0.01143072,
      "auxiliary_loss_mlp": 0.01053231,
      "balance_loss_clip": 1.05276942,
      "balance_loss_mlp": 1.03452659,
      "epoch": 0.16798436795430632,
      "flos": 19719878486400.0,
      "grad_norm": 2.3859650274226833,
      "language_loss": 0.7717455,
      "learning_rate": 3.803576041376831e-06,
      "loss": 0.79370856,
      "num_input_tokens_seen": 60571995,
      "step": 2794,
      "time_per_iteration": 2.6007745265960693
    },
    {
      "auxiliary_loss_clip": 0.01137828,
      "auxiliary_loss_mlp": 0.0104987,
      "balance_loss_clip": 1.05250025,
      "balance_loss_mlp": 1.03010476,
      "epoch": 0.1680444912069743,
      "flos": 28104839596800.0,
      "grad_norm": 2.7692472240964747,
      "language_loss": 0.71609265,
      "learning_rate": 3.803407690167187e-06,
      "loss": 0.73796958,
      "num_input_tokens_seen": 60591275,
      "step": 2795,
      "time_per_iteration": 2.693826198577881
    },
    {
      "auxiliary_loss_clip": 0.01131865,
      "auxiliary_loss_mlp": 0.01041012,
      "balance_loss_clip": 1.04973865,
      "balance_loss_mlp": 1.02302384,
      "epoch": 0.16810461445964225,
      "flos": 18075205526400.0,
      "grad_norm": 1.990096863808903,
      "language_loss": 0.84230494,
      "learning_rate": 3.803239270572142e-06,
      "loss": 0.8640337,
      "num_input_tokens_seen": 60609235,
      "step": 2796,
      "time_per_iteration": 2.697253465652466
    },
    {
      "auxiliary_loss_clip": 0.01101634,
      "auxiliary_loss_mlp": 0.01045196,
      "balance_loss_clip": 1.04877055,
      "balance_loss_mlp": 1.0262773,
      "epoch": 0.16816473771231025,
      "flos": 23878657105920.0,
      "grad_norm": 1.9272276676322646,
      "language_loss": 0.81609607,
      "learning_rate": 3.8030707825980838e-06,
      "loss": 0.83756441,
      "num_input_tokens_seen": 60629880,
      "step": 2797,
      "time_per_iteration": 2.8784244060516357
    },
    {
      "auxiliary_loss_clip": 0.0114057,
      "auxiliary_loss_mlp": 0.01041282,
      "balance_loss_clip": 1.05136061,
      "balance_loss_mlp": 1.02448523,
      "epoch": 0.1682248609649782,
      "flos": 22783597125120.0,
      "grad_norm": 1.7015769336052518,
      "language_loss": 0.74811113,
      "learning_rate": 3.802902226251401e-06,
      "loss": 0.76992965,
      "num_input_tokens_seen": 60651175,
      "step": 2798,
      "time_per_iteration": 2.700727939605713
    },
    {
      "auxiliary_loss_clip": 0.01161342,
      "auxiliary_loss_mlp": 0.01048462,
      "balance_loss_clip": 1.05728281,
      "balance_loss_mlp": 1.03075945,
      "epoch": 0.16828498421764618,
      "flos": 20705123612160.0,
      "grad_norm": 1.5964091182578661,
      "language_loss": 0.79693568,
      "learning_rate": 3.8027336015384845e-06,
      "loss": 0.81903368,
      "num_input_tokens_seen": 60670210,
      "step": 2799,
      "time_per_iteration": 2.6582021713256836
    },
    {
      "auxiliary_loss_clip": 0.01077177,
      "auxiliary_loss_mlp": 0.01045216,
      "balance_loss_clip": 1.04514158,
      "balance_loss_mlp": 1.02374637,
      "epoch": 0.16834510747031414,
      "flos": 29420606695680.0,
      "grad_norm": 4.227726163531211,
      "language_loss": 0.70963746,
      "learning_rate": 3.8025649084657296e-06,
      "loss": 0.73086143,
      "num_input_tokens_seen": 60690895,
      "step": 2800,
      "time_per_iteration": 2.8856699466705322
    },
    {
      "auxiliary_loss_clip": 0.01108822,
      "auxiliary_loss_mlp": 0.00777078,
      "balance_loss_clip": 1.04776788,
      "balance_loss_mlp": 1.00161195,
      "epoch": 0.1684052307229821,
      "flos": 18145374744960.0,
      "grad_norm": 1.9902029671619985,
      "language_loss": 0.83663505,
      "learning_rate": 3.8023961470395326e-06,
      "loss": 0.85549408,
      "num_input_tokens_seen": 60708280,
      "step": 2801,
      "time_per_iteration": 2.6917035579681396
    },
    {
      "auxiliary_loss_clip": 0.01128148,
      "auxiliary_loss_mlp": 0.01049324,
      "balance_loss_clip": 1.05011535,
      "balance_loss_mlp": 1.03084683,
      "epoch": 0.16846535397565007,
      "flos": 16574929240320.0,
      "grad_norm": 2.4052305427948735,
      "language_loss": 0.82509923,
      "learning_rate": 3.8022273172662933e-06,
      "loss": 0.84687394,
      "num_input_tokens_seen": 60724150,
      "step": 2802,
      "time_per_iteration": 2.882611036300659
    },
    {
      "auxiliary_loss_clip": 0.01150156,
      "auxiliary_loss_mlp": 0.01048717,
      "balance_loss_clip": 1.05517435,
      "balance_loss_mlp": 1.02885723,
      "epoch": 0.16852547722831807,
      "flos": 30408868563840.0,
      "grad_norm": 3.107584498439891,
      "language_loss": 0.80643189,
      "learning_rate": 3.802058419152413e-06,
      "loss": 0.8284207,
      "num_input_tokens_seen": 60746485,
      "step": 2803,
      "time_per_iteration": 2.7886922359466553
    },
    {
      "auxiliary_loss_clip": 0.01148107,
      "auxiliary_loss_mlp": 0.01047852,
      "balance_loss_clip": 1.0556829,
      "balance_loss_mlp": 1.02918339,
      "epoch": 0.16858560048098603,
      "flos": 33507420416640.0,
      "grad_norm": 2.2127389669880713,
      "language_loss": 0.76168799,
      "learning_rate": 3.801889452704297e-06,
      "loss": 0.7836476,
      "num_input_tokens_seen": 60762875,
      "step": 2804,
      "time_per_iteration": 2.7588601112365723
    },
    {
      "auxiliary_loss_clip": 0.01045171,
      "auxiliary_loss_mlp": 0.01013955,
      "balance_loss_clip": 1.03581083,
      "balance_loss_mlp": 1.01078367,
      "epoch": 0.168645723733654,
      "flos": 67370502326400.0,
      "grad_norm": 0.8536034833258724,
      "language_loss": 0.55464876,
      "learning_rate": 3.8017204179283526e-06,
      "loss": 0.57524002,
      "num_input_tokens_seen": 60825510,
      "step": 2805,
      "time_per_iteration": 3.2089412212371826
    },
    {
      "auxiliary_loss_clip": 0.01138275,
      "auxiliary_loss_mlp": 0.0103974,
      "balance_loss_clip": 1.05013156,
      "balance_loss_mlp": 1.02239537,
      "epoch": 0.16870584698632196,
      "flos": 21324618501120.0,
      "grad_norm": 2.2836767274778427,
      "language_loss": 0.73090243,
      "learning_rate": 3.8015513148309892e-06,
      "loss": 0.75268269,
      "num_input_tokens_seen": 60844440,
      "step": 2806,
      "time_per_iteration": 2.643596649169922
    },
    {
      "auxiliary_loss_clip": 0.01117063,
      "auxiliary_loss_mlp": 0.01045402,
      "balance_loss_clip": 1.05330753,
      "balance_loss_mlp": 1.02766335,
      "epoch": 0.16876597023898993,
      "flos": 20740746925440.0,
      "grad_norm": 1.8406859431587912,
      "language_loss": 0.69773197,
      "learning_rate": 3.80138214341862e-06,
      "loss": 0.71935666,
      "num_input_tokens_seen": 60863210,
      "step": 2807,
      "time_per_iteration": 2.6946568489074707
    },
    {
      "auxiliary_loss_clip": 0.01130702,
      "auxiliary_loss_mlp": 0.01047199,
      "balance_loss_clip": 1.04842246,
      "balance_loss_mlp": 1.02794707,
      "epoch": 0.1688260934916579,
      "flos": 20303498666880.0,
      "grad_norm": 3.042021842274248,
      "language_loss": 0.70280695,
      "learning_rate": 3.8012129036976587e-06,
      "loss": 0.72458601,
      "num_input_tokens_seen": 60882510,
      "step": 2808,
      "time_per_iteration": 2.6656088829040527
    },
    {
      "auxiliary_loss_clip": 0.01119025,
      "auxiliary_loss_mlp": 0.01041739,
      "balance_loss_clip": 1.05019665,
      "balance_loss_mlp": 1.02164018,
      "epoch": 0.16888621674432586,
      "flos": 20340702178560.0,
      "grad_norm": 2.0835789337145965,
      "language_loss": 0.79903001,
      "learning_rate": 3.8010435956745236e-06,
      "loss": 0.8206377,
      "num_input_tokens_seen": 60901105,
      "step": 2809,
      "time_per_iteration": 2.7665679454803467
    },
    {
      "auxiliary_loss_clip": 0.01155146,
      "auxiliary_loss_mlp": 0.01042018,
      "balance_loss_clip": 1.0557605,
      "balance_loss_mlp": 1.02252758,
      "epoch": 0.16894633999699385,
      "flos": 16244802316800.0,
      "grad_norm": 2.0672093223845245,
      "language_loss": 0.88076419,
      "learning_rate": 3.8008742193556358e-06,
      "loss": 0.90273583,
      "num_input_tokens_seen": 60915340,
      "step": 2810,
      "time_per_iteration": 2.6186363697052
    },
    {
      "auxiliary_loss_clip": 0.01149997,
      "auxiliary_loss_mlp": 0.0104631,
      "balance_loss_clip": 1.05503082,
      "balance_loss_mlp": 1.02715337,
      "epoch": 0.16900646324966181,
      "flos": 19610171372160.0,
      "grad_norm": 1.8921026809528976,
      "language_loss": 0.92376304,
      "learning_rate": 3.800704774747416e-06,
      "loss": 0.9457261,
      "num_input_tokens_seen": 60933735,
      "step": 2811,
      "time_per_iteration": 2.6567442417144775
    },
    {
      "auxiliary_loss_clip": 0.01140053,
      "auxiliary_loss_mlp": 0.01049063,
      "balance_loss_clip": 1.05383325,
      "balance_loss_mlp": 1.03039432,
      "epoch": 0.16906658650232978,
      "flos": 22018089450240.0,
      "grad_norm": 2.116573413654177,
      "language_loss": 0.78582352,
      "learning_rate": 3.800535261856291e-06,
      "loss": 0.8077147,
      "num_input_tokens_seen": 60953105,
      "step": 2812,
      "time_per_iteration": 2.6796023845672607
    },
    {
      "auxiliary_loss_clip": 0.01147895,
      "auxiliary_loss_mlp": 0.01043917,
      "balance_loss_clip": 1.05772316,
      "balance_loss_mlp": 1.02653646,
      "epoch": 0.16912670975499774,
      "flos": 11763690024960.0,
      "grad_norm": 2.5483899062625093,
      "language_loss": 0.75195068,
      "learning_rate": 3.8003656806886887e-06,
      "loss": 0.7738688,
      "num_input_tokens_seen": 60969150,
      "step": 2813,
      "time_per_iteration": 2.621772050857544
    },
    {
      "auxiliary_loss_clip": 0.01136313,
      "auxiliary_loss_mlp": 0.01045037,
      "balance_loss_clip": 1.05311871,
      "balance_loss_mlp": 1.02599943,
      "epoch": 0.1691868330076657,
      "flos": 17161386595200.0,
      "grad_norm": 3.0041182480764554,
      "language_loss": 0.69118392,
      "learning_rate": 3.8001960312510396e-06,
      "loss": 0.7129975,
      "num_input_tokens_seen": 60982825,
      "step": 2814,
      "time_per_iteration": 2.837264060974121
    },
    {
      "auxiliary_loss_clip": 0.01163835,
      "auxiliary_loss_mlp": 0.01039837,
      "balance_loss_clip": 1.05900145,
      "balance_loss_mlp": 1.02134776,
      "epoch": 0.16924695626033368,
      "flos": 22416553998720.0,
      "grad_norm": 3.1079956206415833,
      "language_loss": 0.61439502,
      "learning_rate": 3.800026313549776e-06,
      "loss": 0.63643175,
      "num_input_tokens_seen": 61000875,
      "step": 2815,
      "time_per_iteration": 2.6967194080352783
    },
    {
      "auxiliary_loss_clip": 0.01129827,
      "auxiliary_loss_mlp": 0.01042692,
      "balance_loss_clip": 1.05139673,
      "balance_loss_mlp": 1.02382088,
      "epoch": 0.16930707951300164,
      "flos": 25739655724800.0,
      "grad_norm": 1.7930623183302479,
      "language_loss": 0.82490849,
      "learning_rate": 3.7998565275913342e-06,
      "loss": 0.84663367,
      "num_input_tokens_seen": 61021940,
      "step": 2816,
      "time_per_iteration": 2.7227163314819336
    },
    {
      "auxiliary_loss_clip": 0.01133129,
      "auxiliary_loss_mlp": 0.01047914,
      "balance_loss_clip": 1.05375743,
      "balance_loss_mlp": 1.02853012,
      "epoch": 0.16936720276566963,
      "flos": 22747040058240.0,
      "grad_norm": 3.083808689594852,
      "language_loss": 0.87322289,
      "learning_rate": 3.799686673382153e-06,
      "loss": 0.89503324,
      "num_input_tokens_seen": 61040285,
      "step": 2817,
      "time_per_iteration": 2.733180522918701
    },
    {
      "auxiliary_loss_clip": 0.01141455,
      "auxiliary_loss_mlp": 0.01052753,
      "balance_loss_clip": 1.05800366,
      "balance_loss_mlp": 1.03352427,
      "epoch": 0.1694273260183376,
      "flos": 19573973441280.0,
      "grad_norm": 1.8594303503608436,
      "language_loss": 0.81247765,
      "learning_rate": 3.799516750928672e-06,
      "loss": 0.83441973,
      "num_input_tokens_seen": 61059020,
      "step": 2818,
      "time_per_iteration": 2.7384097576141357
    },
    {
      "auxiliary_loss_clip": 0.01160132,
      "auxiliary_loss_mlp": 0.01044196,
      "balance_loss_clip": 1.05699944,
      "balance_loss_mlp": 1.02496791,
      "epoch": 0.16948744927100556,
      "flos": 12457843332480.0,
      "grad_norm": 2.739998367204505,
      "language_loss": 0.80788404,
      "learning_rate": 3.799346760237336e-06,
      "loss": 0.82992733,
      "num_input_tokens_seen": 61074245,
      "step": 2819,
      "time_per_iteration": 2.609870672225952
    },
    {
      "auxiliary_loss_clip": 0.01069019,
      "auxiliary_loss_mlp": 0.01015301,
      "balance_loss_clip": 1.0485003,
      "balance_loss_mlp": 1.0125947,
      "epoch": 0.16954757252367353,
      "flos": 71291694435840.0,
      "grad_norm": 0.9309223426502673,
      "language_loss": 0.61031163,
      "learning_rate": 3.7991767013145902e-06,
      "loss": 0.63115478,
      "num_input_tokens_seen": 61127080,
      "step": 2820,
      "time_per_iteration": 3.161051034927368
    },
    {
      "auxiliary_loss_clip": 0.01125604,
      "auxiliary_loss_mlp": 0.0105036,
      "balance_loss_clip": 1.05106986,
      "balance_loss_mlp": 1.03207326,
      "epoch": 0.1696076957763415,
      "flos": 29606516513280.0,
      "grad_norm": 1.8682266790688726,
      "language_loss": 0.78265435,
      "learning_rate": 3.7990065741668844e-06,
      "loss": 0.80441403,
      "num_input_tokens_seen": 61146955,
      "step": 2821,
      "time_per_iteration": 2.838730573654175
    },
    {
      "auxiliary_loss_clip": 0.0113863,
      "auxiliary_loss_mlp": 0.01055528,
      "balance_loss_clip": 1.05282724,
      "balance_loss_mlp": 1.03494084,
      "epoch": 0.16966781902900946,
      "flos": 24388588535040.0,
      "grad_norm": 2.1667405259997516,
      "language_loss": 0.78521514,
      "learning_rate": 3.7988363788006685e-06,
      "loss": 0.80715668,
      "num_input_tokens_seen": 61166605,
      "step": 2822,
      "time_per_iteration": 2.783385753631592
    },
    {
      "auxiliary_loss_clip": 0.01143597,
      "auxiliary_loss_mlp": 0.00777154,
      "balance_loss_clip": 1.05367076,
      "balance_loss_mlp": 1.00129986,
      "epoch": 0.16972794228167745,
      "flos": 23038814234880.0,
      "grad_norm": 1.8038457392731222,
      "language_loss": 0.74939907,
      "learning_rate": 3.7986661152223967e-06,
      "loss": 0.76860654,
      "num_input_tokens_seen": 61186535,
      "step": 2823,
      "time_per_iteration": 4.329328298568726
    },
    {
      "auxiliary_loss_clip": 0.01129469,
      "auxiliary_loss_mlp": 0.0105385,
      "balance_loss_clip": 1.05166912,
      "balance_loss_mlp": 1.03496754,
      "epoch": 0.16978806553434542,
      "flos": 35228691129600.0,
      "grad_norm": 3.336653609493179,
      "language_loss": 0.60266119,
      "learning_rate": 3.7984957834385257e-06,
      "loss": 0.62449437,
      "num_input_tokens_seen": 61208965,
      "step": 2824,
      "time_per_iteration": 5.892346620559692
    },
    {
      "auxiliary_loss_clip": 0.01138249,
      "auxiliary_loss_mlp": 0.01042322,
      "balance_loss_clip": 1.05565047,
      "balance_loss_mlp": 1.02287912,
      "epoch": 0.16984818878701338,
      "flos": 32014290936960.0,
      "grad_norm": 2.152838804074104,
      "language_loss": 0.73322558,
      "learning_rate": 3.7983253834555144e-06,
      "loss": 0.75503135,
      "num_input_tokens_seen": 61230670,
      "step": 2825,
      "time_per_iteration": 2.834482431411743
    },
    {
      "auxiliary_loss_clip": 0.01161467,
      "auxiliary_loss_mlp": 0.01047701,
      "balance_loss_clip": 1.05502653,
      "balance_loss_mlp": 1.02762675,
      "epoch": 0.16990831203968135,
      "flos": 22818609907200.0,
      "grad_norm": 2.05671259677731,
      "language_loss": 0.85638934,
      "learning_rate": 3.7981549152798245e-06,
      "loss": 0.87848103,
      "num_input_tokens_seen": 61249510,
      "step": 2826,
      "time_per_iteration": 2.6443135738372803
    },
    {
      "auxiliary_loss_clip": 0.01139368,
      "auxiliary_loss_mlp": 0.01047749,
      "balance_loss_clip": 1.05266595,
      "balance_loss_mlp": 1.02856779,
      "epoch": 0.1699684352923493,
      "flos": 23039604334080.0,
      "grad_norm": 1.9562557148441426,
      "language_loss": 0.82465482,
      "learning_rate": 3.7979843789179196e-06,
      "loss": 0.84652597,
      "num_input_tokens_seen": 61269440,
      "step": 2827,
      "time_per_iteration": 2.7683157920837402
    },
    {
      "auxiliary_loss_clip": 0.01131885,
      "auxiliary_loss_mlp": 0.0104561,
      "balance_loss_clip": 1.05320346,
      "balance_loss_mlp": 1.02536786,
      "epoch": 0.17002855854501728,
      "flos": 21434110133760.0,
      "grad_norm": 1.7386401818136152,
      "language_loss": 0.73704529,
      "learning_rate": 3.797813774376267e-06,
      "loss": 0.75882024,
      "num_input_tokens_seen": 61288195,
      "step": 2828,
      "time_per_iteration": 4.465311288833618
    },
    {
      "auxiliary_loss_clip": 0.01061458,
      "auxiliary_loss_mlp": 0.01009538,
      "balance_loss_clip": 1.04764342,
      "balance_loss_mlp": 1.00620067,
      "epoch": 0.17008868179768524,
      "flos": 71453509205760.0,
      "grad_norm": 0.7670168832041738,
      "language_loss": 0.56426483,
      "learning_rate": 3.797643101661336e-06,
      "loss": 0.58497471,
      "num_input_tokens_seen": 61350850,
      "step": 2829,
      "time_per_iteration": 3.3114631175994873
    },
    {
      "auxiliary_loss_clip": 0.01111753,
      "auxiliary_loss_mlp": 0.01051557,
      "balance_loss_clip": 1.04527223,
      "balance_loss_mlp": 1.03088641,
      "epoch": 0.17014880505035324,
      "flos": 24900315644160.0,
      "grad_norm": 1.7961285206560338,
      "language_loss": 0.83465374,
      "learning_rate": 3.7974723607795983e-06,
      "loss": 0.85628688,
      "num_input_tokens_seen": 61370765,
      "step": 2830,
      "time_per_iteration": 2.795253038406372
    },
    {
      "auxiliary_loss_clip": 0.01121533,
      "auxiliary_loss_mlp": 0.0104408,
      "balance_loss_clip": 1.04901659,
      "balance_loss_mlp": 1.02442193,
      "epoch": 0.1702089283030212,
      "flos": 29862415981440.0,
      "grad_norm": 2.4873654173451727,
      "language_loss": 0.78360993,
      "learning_rate": 3.797301551737529e-06,
      "loss": 0.80526608,
      "num_input_tokens_seen": 61388935,
      "step": 2831,
      "time_per_iteration": 2.7864232063293457
    },
    {
      "auxiliary_loss_clip": 0.01123612,
      "auxiliary_loss_mlp": 0.01051154,
      "balance_loss_clip": 1.05275893,
      "balance_loss_mlp": 1.0311985,
      "epoch": 0.17026905155568917,
      "flos": 17744180762880.0,
      "grad_norm": 2.532473263441992,
      "language_loss": 0.79668158,
      "learning_rate": 3.7971306745416044e-06,
      "loss": 0.81842923,
      "num_input_tokens_seen": 61407350,
      "step": 2832,
      "time_per_iteration": 2.842217206954956
    },
    {
      "auxiliary_loss_clip": 0.01127135,
      "auxiliary_loss_mlp": 0.01048966,
      "balance_loss_clip": 1.05029321,
      "balance_loss_mlp": 1.02984488,
      "epoch": 0.17032917480835713,
      "flos": 23148665003520.0,
      "grad_norm": 1.8387196201649116,
      "language_loss": 0.88638175,
      "learning_rate": 3.7969597291983046e-06,
      "loss": 0.90814275,
      "num_input_tokens_seen": 61429010,
      "step": 2833,
      "time_per_iteration": 2.75942325592041
    },
    {
      "auxiliary_loss_clip": 0.01158799,
      "auxiliary_loss_mlp": 0.01046883,
      "balance_loss_clip": 1.05633831,
      "balance_loss_mlp": 1.02842951,
      "epoch": 0.1703892980610251,
      "flos": 39202565512320.0,
      "grad_norm": 2.49094605220443,
      "language_loss": 0.71924698,
      "learning_rate": 3.7967887157141115e-06,
      "loss": 0.74130386,
      "num_input_tokens_seen": 61450040,
      "step": 2834,
      "time_per_iteration": 2.9035184383392334
    },
    {
      "auxiliary_loss_clip": 0.01119873,
      "auxiliary_loss_mlp": 0.01052215,
      "balance_loss_clip": 1.05165124,
      "balance_loss_mlp": 1.03428626,
      "epoch": 0.17044942131369306,
      "flos": 23039101543680.0,
      "grad_norm": 1.9093816511111852,
      "language_loss": 0.86831236,
      "learning_rate": 3.7966176340955106e-06,
      "loss": 0.89003325,
      "num_input_tokens_seen": 61468585,
      "step": 2835,
      "time_per_iteration": 2.7627484798431396
    },
    {
      "auxiliary_loss_clip": 0.01149332,
      "auxiliary_loss_mlp": 0.01049844,
      "balance_loss_clip": 1.0536654,
      "balance_loss_mlp": 1.02887547,
      "epoch": 0.17050954456636103,
      "flos": 17054983532160.0,
      "grad_norm": 2.1227367002258153,
      "language_loss": 0.74483943,
      "learning_rate": 3.796446484348989e-06,
      "loss": 0.76683116,
      "num_input_tokens_seen": 61486330,
      "step": 2836,
      "time_per_iteration": 2.6748619079589844
    },
    {
      "auxiliary_loss_clip": 0.01102249,
      "auxiliary_loss_mlp": 0.01049533,
      "balance_loss_clip": 1.04775679,
      "balance_loss_mlp": 1.02790809,
      "epoch": 0.17056966781902902,
      "flos": 16836969934080.0,
      "grad_norm": 2.1718385109372824,
      "language_loss": 0.79959226,
      "learning_rate": 3.796275266481036e-06,
      "loss": 0.82111007,
      "num_input_tokens_seen": 61503950,
      "step": 2837,
      "time_per_iteration": 2.757340908050537
    },
    {
      "auxiliary_loss_clip": 0.01144378,
      "auxiliary_loss_mlp": 0.01044803,
      "balance_loss_clip": 1.05493581,
      "balance_loss_mlp": 1.02644491,
      "epoch": 0.17062979107169698,
      "flos": 17712543859200.0,
      "grad_norm": 1.6825251002952497,
      "language_loss": 0.83258498,
      "learning_rate": 3.7961039804981456e-06,
      "loss": 0.85447681,
      "num_input_tokens_seen": 61523550,
      "step": 2838,
      "time_per_iteration": 2.705357551574707
    },
    {
      "auxiliary_loss_clip": 0.0110604,
      "auxiliary_loss_mlp": 0.01044889,
      "balance_loss_clip": 1.05217135,
      "balance_loss_mlp": 1.02685261,
      "epoch": 0.17068991432436495,
      "flos": 22525040050560.0,
      "grad_norm": 1.7789799751303759,
      "language_loss": 0.93788463,
      "learning_rate": 3.795932626406812e-06,
      "loss": 0.95939398,
      "num_input_tokens_seen": 61542720,
      "step": 2839,
      "time_per_iteration": 2.7881791591644287
    },
    {
      "auxiliary_loss_clip": 0.01126465,
      "auxiliary_loss_mlp": 0.01045617,
      "balance_loss_clip": 1.05183244,
      "balance_loss_mlp": 1.0250175,
      "epoch": 0.17075003757703291,
      "flos": 25882939077120.0,
      "grad_norm": 2.3337760403585435,
      "language_loss": 0.83974946,
      "learning_rate": 3.7957612042135336e-06,
      "loss": 0.86147022,
      "num_input_tokens_seen": 61563040,
      "step": 2840,
      "time_per_iteration": 2.7564892768859863
    },
    {
      "auxiliary_loss_clip": 0.01151834,
      "auxiliary_loss_mlp": 0.01044417,
      "balance_loss_clip": 1.05555129,
      "balance_loss_mlp": 1.02449679,
      "epoch": 0.17081016082970088,
      "flos": 20120713332480.0,
      "grad_norm": 1.9037435592597944,
      "language_loss": 0.76307738,
      "learning_rate": 3.79558971392481e-06,
      "loss": 0.7850399,
      "num_input_tokens_seen": 61581890,
      "step": 2841,
      "time_per_iteration": 2.695525646209717
    },
    {
      "auxiliary_loss_clip": 0.01136217,
      "auxiliary_loss_mlp": 0.01045847,
      "balance_loss_clip": 1.0527097,
      "balance_loss_mlp": 1.02744126,
      "epoch": 0.17087028408236885,
      "flos": 24936477661440.0,
      "grad_norm": 1.7844240011089845,
      "language_loss": 0.77076876,
      "learning_rate": 3.7954181555471443e-06,
      "loss": 0.79258937,
      "num_input_tokens_seen": 61602095,
      "step": 2842,
      "time_per_iteration": 2.773792266845703
    },
    {
      "auxiliary_loss_clip": 0.01155915,
      "auxiliary_loss_mlp": 0.01043896,
      "balance_loss_clip": 1.05616069,
      "balance_loss_mlp": 1.02503705,
      "epoch": 0.17093040733503684,
      "flos": 19057864872960.0,
      "grad_norm": 1.8430349199993477,
      "language_loss": 0.85694385,
      "learning_rate": 3.795246529087043e-06,
      "loss": 0.87894201,
      "num_input_tokens_seen": 61620400,
      "step": 2843,
      "time_per_iteration": 2.5860671997070312
    },
    {
      "auxiliary_loss_clip": 0.01154742,
      "auxiliary_loss_mlp": 0.01044059,
      "balance_loss_clip": 1.05549574,
      "balance_loss_mlp": 1.02608204,
      "epoch": 0.1709905305877048,
      "flos": 13078954333440.0,
      "grad_norm": 2.0353470349004485,
      "language_loss": 0.68646181,
      "learning_rate": 3.7950748345510126e-06,
      "loss": 0.70844984,
      "num_input_tokens_seen": 61637680,
      "step": 2844,
      "time_per_iteration": 2.5961523056030273
    },
    {
      "auxiliary_loss_clip": 0.01133396,
      "auxiliary_loss_mlp": 0.00778162,
      "balance_loss_clip": 1.05117011,
      "balance_loss_mlp": 1.00112617,
      "epoch": 0.17105065384037277,
      "flos": 19209336526080.0,
      "grad_norm": 2.027694794878894,
      "language_loss": 0.78771943,
      "learning_rate": 3.7949030719455646e-06,
      "loss": 0.806835,
      "num_input_tokens_seen": 61655630,
      "step": 2845,
      "time_per_iteration": 2.720193386077881
    },
    {
      "auxiliary_loss_clip": 0.01145033,
      "auxiliary_loss_mlp": 0.01047407,
      "balance_loss_clip": 1.05443549,
      "balance_loss_mlp": 1.02914453,
      "epoch": 0.17111077709304073,
      "flos": 18515183218560.0,
      "grad_norm": 2.2586144454646306,
      "language_loss": 0.7811147,
      "learning_rate": 3.7947312412772127e-06,
      "loss": 0.80303913,
      "num_input_tokens_seen": 61673475,
      "step": 2846,
      "time_per_iteration": 2.691033363342285
    },
    {
      "auxiliary_loss_clip": 0.01143809,
      "auxiliary_loss_mlp": 0.0104645,
      "balance_loss_clip": 1.05425262,
      "balance_loss_mlp": 1.02865243,
      "epoch": 0.1711709003457087,
      "flos": 25082670015360.0,
      "grad_norm": 2.2208975060456426,
      "language_loss": 0.79762948,
      "learning_rate": 3.794559342552472e-06,
      "loss": 0.8195321,
      "num_input_tokens_seen": 61693370,
      "step": 2847,
      "time_per_iteration": 2.7504522800445557
    },
    {
      "auxiliary_loss_clip": 0.01142651,
      "auxiliary_loss_mlp": 0.01045695,
      "balance_loss_clip": 1.05101562,
      "balance_loss_mlp": 1.02668071,
      "epoch": 0.17123102359837666,
      "flos": 17566387418880.0,
      "grad_norm": 2.4457083156230017,
      "language_loss": 0.8665086,
      "learning_rate": 3.7943873757778614e-06,
      "loss": 0.88839209,
      "num_input_tokens_seen": 61710820,
      "step": 2848,
      "time_per_iteration": 2.642946720123291
    },
    {
      "auxiliary_loss_clip": 0.0111167,
      "auxiliary_loss_mlp": 0.01044479,
      "balance_loss_clip": 1.04839015,
      "balance_loss_mlp": 1.02559662,
      "epoch": 0.17129114685104463,
      "flos": 26173635845760.0,
      "grad_norm": 3.6033710399461856,
      "language_loss": 0.75238276,
      "learning_rate": 3.794215340959902e-06,
      "loss": 0.77394426,
      "num_input_tokens_seen": 61729855,
      "step": 2849,
      "time_per_iteration": 2.7511017322540283
    },
    {
      "auxiliary_loss_clip": 0.0103263,
      "auxiliary_loss_mlp": 0.01006833,
      "balance_loss_clip": 1.02775574,
      "balance_loss_mlp": 1.00413883,
      "epoch": 0.17135127010371262,
      "flos": 69269710037760.0,
      "grad_norm": 0.7881928427119427,
      "language_loss": 0.57514679,
      "learning_rate": 3.7940432381051163e-06,
      "loss": 0.59554148,
      "num_input_tokens_seen": 61790290,
      "step": 2850,
      "time_per_iteration": 3.234609603881836
    },
    {
      "auxiliary_loss_clip": 0.01115021,
      "auxiliary_loss_mlp": 0.01044381,
      "balance_loss_clip": 1.05049884,
      "balance_loss_mlp": 1.02661848,
      "epoch": 0.1714113933563806,
      "flos": 23550110380800.0,
      "grad_norm": 2.962731712990184,
      "language_loss": 0.81328994,
      "learning_rate": 3.793871067220031e-06,
      "loss": 0.83488399,
      "num_input_tokens_seen": 61809265,
      "step": 2851,
      "time_per_iteration": 2.78957200050354
    },
    {
      "auxiliary_loss_clip": 0.01114419,
      "auxiliary_loss_mlp": 0.01043587,
      "balance_loss_clip": 1.05193233,
      "balance_loss_mlp": 1.02592039,
      "epoch": 0.17147151660904855,
      "flos": 21142443697920.0,
      "grad_norm": 2.049906502724323,
      "language_loss": 0.93085313,
      "learning_rate": 3.7936988283111764e-06,
      "loss": 0.95243311,
      "num_input_tokens_seen": 61828980,
      "step": 2852,
      "time_per_iteration": 2.8247029781341553
    },
    {
      "auxiliary_loss_clip": 0.01123258,
      "auxiliary_loss_mlp": 0.01048953,
      "balance_loss_clip": 1.04961288,
      "balance_loss_mlp": 1.03045225,
      "epoch": 0.17153163986171652,
      "flos": 18624890332800.0,
      "grad_norm": 1.8770741979814063,
      "language_loss": 0.69465554,
      "learning_rate": 3.7935265213850817e-06,
      "loss": 0.71637762,
      "num_input_tokens_seen": 61847915,
      "step": 2853,
      "time_per_iteration": 2.814162492752075
    },
    {
      "auxiliary_loss_clip": 0.01120856,
      "auxiliary_loss_mlp": 0.0104692,
      "balance_loss_clip": 1.05593121,
      "balance_loss_mlp": 1.02899122,
      "epoch": 0.17159176311438448,
      "flos": 18223265387520.0,
      "grad_norm": 2.5884803351111705,
      "language_loss": 0.66611075,
      "learning_rate": 3.7933541464482815e-06,
      "loss": 0.68778855,
      "num_input_tokens_seen": 61865570,
      "step": 2854,
      "time_per_iteration": 2.7968995571136475
    },
    {
      "auxiliary_loss_clip": 0.01120742,
      "auxiliary_loss_mlp": 0.01052217,
      "balance_loss_clip": 1.04853106,
      "balance_loss_mlp": 1.0349679,
      "epoch": 0.17165188636705245,
      "flos": 20738987159040.0,
      "grad_norm": 1.705510390491261,
      "language_loss": 0.8929621,
      "learning_rate": 3.7931817035073124e-06,
      "loss": 0.91469175,
      "num_input_tokens_seen": 61883340,
      "step": 2855,
      "time_per_iteration": 2.7045016288757324
    },
    {
      "auxiliary_loss_clip": 0.01157319,
      "auxiliary_loss_mlp": 0.01043813,
      "balance_loss_clip": 1.05505848,
      "balance_loss_mlp": 1.02662265,
      "epoch": 0.17171200961972044,
      "flos": 24899884680960.0,
      "grad_norm": 2.117219134143716,
      "language_loss": 0.83963835,
      "learning_rate": 3.7930091925687134e-06,
      "loss": 0.86164963,
      "num_input_tokens_seen": 61900610,
      "step": 2856,
      "time_per_iteration": 2.7349936962127686
    },
    {
      "auxiliary_loss_clip": 0.01150108,
      "auxiliary_loss_mlp": 0.0104615,
      "balance_loss_clip": 1.05812418,
      "balance_loss_mlp": 1.02783966,
      "epoch": 0.1717721328723884,
      "flos": 20157234485760.0,
      "grad_norm": 2.234025867710235,
      "language_loss": 0.86309886,
      "learning_rate": 3.792836613639026e-06,
      "loss": 0.88506144,
      "num_input_tokens_seen": 61916795,
      "step": 2857,
      "time_per_iteration": 2.749356746673584
    },
    {
      "auxiliary_loss_clip": 0.01144467,
      "auxiliary_loss_mlp": 0.0105057,
      "balance_loss_clip": 1.05469525,
      "balance_loss_mlp": 1.0324626,
      "epoch": 0.17183225612505637,
      "flos": 23361650697600.0,
      "grad_norm": 2.069122070501307,
      "language_loss": 0.78334701,
      "learning_rate": 3.7926639667247947e-06,
      "loss": 0.80529737,
      "num_input_tokens_seen": 61936665,
      "step": 2858,
      "time_per_iteration": 2.6673583984375
    },
    {
      "auxiliary_loss_clip": 0.01147374,
      "auxiliary_loss_mlp": 0.0105371,
      "balance_loss_clip": 1.05591416,
      "balance_loss_mlp": 1.03263378,
      "epoch": 0.17189237937772434,
      "flos": 18114240631680.0,
      "grad_norm": 2.1629422323642453,
      "language_loss": 0.77565676,
      "learning_rate": 3.7924912518325663e-06,
      "loss": 0.79766762,
      "num_input_tokens_seen": 61954415,
      "step": 2859,
      "time_per_iteration": 2.646648645401001
    },
    {
      "auxiliary_loss_clip": 0.0110879,
      "auxiliary_loss_mlp": 0.01047481,
      "balance_loss_clip": 1.05317724,
      "balance_loss_mlp": 1.02887201,
      "epoch": 0.1719525026303923,
      "flos": 23258408031360.0,
      "grad_norm": 2.088627069497316,
      "language_loss": 0.77088714,
      "learning_rate": 3.7923184689688902e-06,
      "loss": 0.79244983,
      "num_input_tokens_seen": 61973940,
      "step": 2860,
      "time_per_iteration": 2.7671573162078857
    },
    {
      "auxiliary_loss_clip": 0.01145562,
      "auxiliary_loss_mlp": 0.01042048,
      "balance_loss_clip": 1.05316472,
      "balance_loss_mlp": 1.02416611,
      "epoch": 0.17201262588306027,
      "flos": 20810413353600.0,
      "grad_norm": 2.1608688480628304,
      "language_loss": 0.81384242,
      "learning_rate": 3.792145618140317e-06,
      "loss": 0.83571851,
      "num_input_tokens_seen": 61991845,
      "step": 2861,
      "time_per_iteration": 2.6492061614990234
    },
    {
      "auxiliary_loss_clip": 0.011306,
      "auxiliary_loss_mlp": 0.01051558,
      "balance_loss_clip": 1.05280077,
      "balance_loss_mlp": 1.0335927,
      "epoch": 0.17207274913572823,
      "flos": 20375858615040.0,
      "grad_norm": 2.0128324416816192,
      "language_loss": 0.85691392,
      "learning_rate": 3.7919726993534038e-06,
      "loss": 0.87873554,
      "num_input_tokens_seen": 62009395,
      "step": 2862,
      "time_per_iteration": 4.290126323699951
    },
    {
      "auxiliary_loss_clip": 0.01116765,
      "auxiliary_loss_mlp": 0.01043444,
      "balance_loss_clip": 1.05126834,
      "balance_loss_mlp": 1.02655208,
      "epoch": 0.17213287238839622,
      "flos": 26797727675520.0,
      "grad_norm": 3.7047120479299993,
      "language_loss": 0.78047049,
      "learning_rate": 3.7917997126147054e-06,
      "loss": 0.80207253,
      "num_input_tokens_seen": 62029005,
      "step": 2863,
      "time_per_iteration": 4.275500774383545
    },
    {
      "auxiliary_loss_clip": 0.01122315,
      "auxiliary_loss_mlp": 0.00776596,
      "balance_loss_clip": 1.05132961,
      "balance_loss_mlp": 1.00090909,
      "epoch": 0.1721929956410642,
      "flos": 26030819370240.0,
      "grad_norm": 1.7350128683820358,
      "language_loss": 0.72135127,
      "learning_rate": 3.7916266579307823e-06,
      "loss": 0.74034035,
      "num_input_tokens_seen": 62048730,
      "step": 2864,
      "time_per_iteration": 4.414710998535156
    },
    {
      "auxiliary_loss_clip": 0.01121488,
      "auxiliary_loss_mlp": 0.01049611,
      "balance_loss_clip": 1.05114079,
      "balance_loss_mlp": 1.03099,
      "epoch": 0.17225311889373215,
      "flos": 22273091078400.0,
      "grad_norm": 1.9270646210248614,
      "language_loss": 0.73002023,
      "learning_rate": 3.7914535353081973e-06,
      "loss": 0.75173128,
      "num_input_tokens_seen": 62069000,
      "step": 2865,
      "time_per_iteration": 2.7463715076446533
    },
    {
      "auxiliary_loss_clip": 0.01145037,
      "auxiliary_loss_mlp": 0.0077644,
      "balance_loss_clip": 1.05669165,
      "balance_loss_mlp": 1.00120521,
      "epoch": 0.17231324214640012,
      "flos": 21287774125440.0,
      "grad_norm": 2.669585642962841,
      "language_loss": 0.78357804,
      "learning_rate": 3.7912803447535145e-06,
      "loss": 0.80279285,
      "num_input_tokens_seen": 62086750,
      "step": 2866,
      "time_per_iteration": 2.785146713256836
    },
    {
      "auxiliary_loss_clip": 0.01157272,
      "auxiliary_loss_mlp": 0.01044358,
      "balance_loss_clip": 1.05600274,
      "balance_loss_mlp": 1.02536821,
      "epoch": 0.17237336539906808,
      "flos": 19680735640320.0,
      "grad_norm": 2.551277931358127,
      "language_loss": 0.79755104,
      "learning_rate": 3.7911070862733016e-06,
      "loss": 0.81956732,
      "num_input_tokens_seen": 62106240,
      "step": 2867,
      "time_per_iteration": 4.3145318031311035
    },
    {
      "auxiliary_loss_clip": 0.01132297,
      "auxiliary_loss_mlp": 0.01041396,
      "balance_loss_clip": 1.0529356,
      "balance_loss_mlp": 1.02274013,
      "epoch": 0.17243348865173605,
      "flos": 17529650784000.0,
      "grad_norm": 1.8689780270661371,
      "language_loss": 0.79206991,
      "learning_rate": 3.7909337598741276e-06,
      "loss": 0.81380683,
      "num_input_tokens_seen": 62124895,
      "step": 2868,
      "time_per_iteration": 2.7683827877044678
    },
    {
      "auxiliary_loss_clip": 0.01111702,
      "auxiliary_loss_mlp": 0.01041717,
      "balance_loss_clip": 1.05331647,
      "balance_loss_mlp": 1.02427697,
      "epoch": 0.17249361190440402,
      "flos": 18259858368000.0,
      "grad_norm": 2.0344588273772923,
      "language_loss": 0.84221756,
      "learning_rate": 3.7907603655625674e-06,
      "loss": 0.86375177,
      "num_input_tokens_seen": 62143510,
      "step": 2869,
      "time_per_iteration": 2.729156970977783
    },
    {
      "auxiliary_loss_clip": 0.01132999,
      "auxiliary_loss_mlp": 0.01048405,
      "balance_loss_clip": 1.0535363,
      "balance_loss_mlp": 1.02955842,
      "epoch": 0.172553735157072,
      "flos": 21174367910400.0,
      "grad_norm": 1.8935704627114847,
      "language_loss": 0.77299273,
      "learning_rate": 3.7905869033451932e-06,
      "loss": 0.79480684,
      "num_input_tokens_seen": 62162285,
      "step": 2870,
      "time_per_iteration": 2.752739191055298
    },
    {
      "auxiliary_loss_clip": 0.0115398,
      "auxiliary_loss_mlp": 0.01037809,
      "balance_loss_clip": 1.05671024,
      "balance_loss_mlp": 1.02110744,
      "epoch": 0.17261385840973997,
      "flos": 22273270646400.0,
      "grad_norm": 2.0115587398764396,
      "language_loss": 0.77409238,
      "learning_rate": 3.7904133732285857e-06,
      "loss": 0.79601026,
      "num_input_tokens_seen": 62180970,
      "step": 2871,
      "time_per_iteration": 2.660627603530884
    },
    {
      "auxiliary_loss_clip": 0.01132474,
      "auxiliary_loss_mlp": 0.01041073,
      "balance_loss_clip": 1.05313993,
      "balance_loss_mlp": 1.0222379,
      "epoch": 0.17267398166240794,
      "flos": 27922233830400.0,
      "grad_norm": 2.203011669690562,
      "language_loss": 0.74197829,
      "learning_rate": 3.7902397752193228e-06,
      "loss": 0.76371384,
      "num_input_tokens_seen": 62198965,
      "step": 2872,
      "time_per_iteration": 2.6959900856018066
    },
    {
      "auxiliary_loss_clip": 0.01150773,
      "auxiliary_loss_mlp": 0.01041508,
      "balance_loss_clip": 1.05359554,
      "balance_loss_mlp": 1.02362645,
      "epoch": 0.1727341049150759,
      "flos": 21945118970880.0,
      "grad_norm": 1.7914171074077658,
      "language_loss": 0.82336062,
      "learning_rate": 3.790066109323988e-06,
      "loss": 0.84528345,
      "num_input_tokens_seen": 62219890,
      "step": 2873,
      "time_per_iteration": 2.603564977645874
    },
    {
      "auxiliary_loss_clip": 0.01108819,
      "auxiliary_loss_mlp": 0.01044995,
      "balance_loss_clip": 1.04744792,
      "balance_loss_mlp": 1.02522969,
      "epoch": 0.17279422816774387,
      "flos": 18107883924480.0,
      "grad_norm": 3.7341652608759297,
      "language_loss": 0.75355422,
      "learning_rate": 3.7898923755491678e-06,
      "loss": 0.77509236,
      "num_input_tokens_seen": 62237140,
      "step": 2874,
      "time_per_iteration": 2.8438260555267334
    },
    {
      "auxiliary_loss_clip": 0.01159322,
      "auxiliary_loss_mlp": 0.01044415,
      "balance_loss_clip": 1.05658269,
      "balance_loss_mlp": 1.02404249,
      "epoch": 0.17285435142041183,
      "flos": 21835447770240.0,
      "grad_norm": 2.7053876793207037,
      "language_loss": 0.80239916,
      "learning_rate": 3.7897185739014487e-06,
      "loss": 0.82443655,
      "num_input_tokens_seen": 62255405,
      "step": 2875,
      "time_per_iteration": 2.625183343887329
    },
    {
      "auxiliary_loss_clip": 0.01135727,
      "auxiliary_loss_mlp": 0.0105273,
      "balance_loss_clip": 1.0535475,
      "balance_loss_mlp": 1.03297722,
      "epoch": 0.17291447467307983,
      "flos": 18368452160640.0,
      "grad_norm": 3.840653645811056,
      "language_loss": 0.87621164,
      "learning_rate": 3.7895447043874217e-06,
      "loss": 0.8980962,
      "num_input_tokens_seen": 62271280,
      "step": 2876,
      "time_per_iteration": 2.6782751083374023
    },
    {
      "auxiliary_loss_clip": 0.01136898,
      "auxiliary_loss_mlp": 0.01044228,
      "balance_loss_clip": 1.05730534,
      "balance_loss_mlp": 1.02559566,
      "epoch": 0.1729745979257478,
      "flos": 18624638937600.0,
      "grad_norm": 1.8931416121171032,
      "language_loss": 0.84386718,
      "learning_rate": 3.789370767013681e-06,
      "loss": 0.86567843,
      "num_input_tokens_seen": 62289140,
      "step": 2877,
      "time_per_iteration": 2.681131362915039
    },
    {
      "auxiliary_loss_clip": 0.01120759,
      "auxiliary_loss_mlp": 0.01043962,
      "balance_loss_clip": 1.05222571,
      "balance_loss_mlp": 1.02499604,
      "epoch": 0.17303472117841576,
      "flos": 22998234844800.0,
      "grad_norm": 2.106635210245156,
      "language_loss": 0.79660022,
      "learning_rate": 3.7891967617868204e-06,
      "loss": 0.81824744,
      "num_input_tokens_seen": 62307490,
      "step": 2878,
      "time_per_iteration": 2.8118834495544434
    },
    {
      "auxiliary_loss_clip": 0.01136112,
      "auxiliary_loss_mlp": 0.01047222,
      "balance_loss_clip": 1.05593777,
      "balance_loss_mlp": 1.02953172,
      "epoch": 0.17309484443108372,
      "flos": 25664386775040.0,
      "grad_norm": 1.9675557254753375,
      "language_loss": 0.70236337,
      "learning_rate": 3.78902268871344e-06,
      "loss": 0.72419673,
      "num_input_tokens_seen": 62328570,
      "step": 2879,
      "time_per_iteration": 2.7998502254486084
    },
    {
      "auxiliary_loss_clip": 0.01130517,
      "auxiliary_loss_mlp": 0.01051722,
      "balance_loss_clip": 1.05183411,
      "balance_loss_mlp": 1.03337598,
      "epoch": 0.1731549676837517,
      "flos": 13552903313280.0,
      "grad_norm": 2.0545155253910163,
      "language_loss": 0.82884222,
      "learning_rate": 3.78884854780014e-06,
      "loss": 0.85066462,
      "num_input_tokens_seen": 62345735,
      "step": 2880,
      "time_per_iteration": 2.6707684993743896
    },
    {
      "auxiliary_loss_clip": 0.01110706,
      "auxiliary_loss_mlp": 0.01054327,
      "balance_loss_clip": 1.05214918,
      "balance_loss_mlp": 1.03303647,
      "epoch": 0.17321509093641965,
      "flos": 22857070394880.0,
      "grad_norm": 1.9029231217608267,
      "language_loss": 0.80879176,
      "learning_rate": 3.7886743390535236e-06,
      "loss": 0.83044201,
      "num_input_tokens_seen": 62365525,
      "step": 2881,
      "time_per_iteration": 2.7851576805114746
    },
    {
      "auxiliary_loss_clip": 0.01135983,
      "auxiliary_loss_mlp": 0.01046895,
      "balance_loss_clip": 1.05544055,
      "balance_loss_mlp": 1.02921653,
      "epoch": 0.17327521418908762,
      "flos": 24352785653760.0,
      "grad_norm": 2.753231520615002,
      "language_loss": 0.77268815,
      "learning_rate": 3.788500062480197e-06,
      "loss": 0.79451692,
      "num_input_tokens_seen": 62385160,
      "step": 2882,
      "time_per_iteration": 2.7785212993621826
    },
    {
      "auxiliary_loss_clip": 0.01124099,
      "auxiliary_loss_mlp": 0.01047516,
      "balance_loss_clip": 1.0633558,
      "balance_loss_mlp": 1.02947998,
      "epoch": 0.1733353374417556,
      "flos": 33105651816960.0,
      "grad_norm": 2.096311926604511,
      "language_loss": 0.76714236,
      "learning_rate": 3.788325718086769e-06,
      "loss": 0.78885853,
      "num_input_tokens_seen": 62405280,
      "step": 2883,
      "time_per_iteration": 2.838848352432251
    },
    {
      "auxiliary_loss_clip": 0.01110924,
      "auxiliary_loss_mlp": 0.0104619,
      "balance_loss_clip": 1.04929209,
      "balance_loss_mlp": 1.02821302,
      "epoch": 0.17339546069442358,
      "flos": 24388947671040.0,
      "grad_norm": 2.1194201700326873,
      "language_loss": 0.8555252,
      "learning_rate": 3.7881513058798503e-06,
      "loss": 0.87709635,
      "num_input_tokens_seen": 62423665,
      "step": 2884,
      "time_per_iteration": 2.829376220703125
    },
    {
      "auxiliary_loss_clip": 0.01133962,
      "auxiliary_loss_mlp": 0.00775817,
      "balance_loss_clip": 1.05472779,
      "balance_loss_mlp": 1.00088096,
      "epoch": 0.17345558394709154,
      "flos": 27454174680960.0,
      "grad_norm": 1.7131036779262108,
      "language_loss": 0.74756771,
      "learning_rate": 3.787976825866055e-06,
      "loss": 0.76666546,
      "num_input_tokens_seen": 62445170,
      "step": 2885,
      "time_per_iteration": 2.8710989952087402
    },
    {
      "auxiliary_loss_clip": 0.01128977,
      "auxiliary_loss_mlp": 0.01044901,
      "balance_loss_clip": 1.05498922,
      "balance_loss_mlp": 1.0280925,
      "epoch": 0.1735157071997595,
      "flos": 24682158391680.0,
      "grad_norm": 2.374438581614022,
      "language_loss": 0.7107017,
      "learning_rate": 3.7878022780519998e-06,
      "loss": 0.73244053,
      "num_input_tokens_seen": 62466135,
      "step": 2886,
      "time_per_iteration": 2.726621150970459
    },
    {
      "auxiliary_loss_clip": 0.01142411,
      "auxiliary_loss_mlp": 0.01041857,
      "balance_loss_clip": 1.05233932,
      "balance_loss_mlp": 1.02408338,
      "epoch": 0.17357583045242747,
      "flos": 21688932193920.0,
      "grad_norm": 2.0566537172661747,
      "language_loss": 0.69906294,
      "learning_rate": 3.7876276624443024e-06,
      "loss": 0.72090566,
      "num_input_tokens_seen": 62483910,
      "step": 2887,
      "time_per_iteration": 2.7066688537597656
    },
    {
      "auxiliary_loss_clip": 0.01116425,
      "auxiliary_loss_mlp": 0.01045383,
      "balance_loss_clip": 1.05328536,
      "balance_loss_mlp": 1.02728677,
      "epoch": 0.17363595370509544,
      "flos": 15375728753280.0,
      "grad_norm": 2.038016964464323,
      "language_loss": 0.85257947,
      "learning_rate": 3.787452979049585e-06,
      "loss": 0.87419748,
      "num_input_tokens_seen": 62501530,
      "step": 2888,
      "time_per_iteration": 2.7514970302581787
    },
    {
      "auxiliary_loss_clip": 0.01095063,
      "auxiliary_loss_mlp": 0.01049413,
      "balance_loss_clip": 1.05020595,
      "balance_loss_mlp": 1.02822983,
      "epoch": 0.1736960769577634,
      "flos": 23440941970560.0,
      "grad_norm": 2.196318077733749,
      "language_loss": 0.78491282,
      "learning_rate": 3.7872782278744718e-06,
      "loss": 0.80635762,
      "num_input_tokens_seen": 62521295,
      "step": 2889,
      "time_per_iteration": 2.8221559524536133
    },
    {
      "auxiliary_loss_clip": 0.01112139,
      "auxiliary_loss_mlp": 0.0077601,
      "balance_loss_clip": 1.05236733,
      "balance_loss_mlp": 1.00114667,
      "epoch": 0.1737562002104314,
      "flos": 18587830475520.0,
      "grad_norm": 2.333227367674716,
      "language_loss": 0.84076989,
      "learning_rate": 3.7871034089255883e-06,
      "loss": 0.85965133,
      "num_input_tokens_seen": 62539615,
      "step": 2890,
      "time_per_iteration": 2.7213382720947266
    },
    {
      "auxiliary_loss_clip": 0.01142218,
      "auxiliary_loss_mlp": 0.01054918,
      "balance_loss_clip": 1.05530691,
      "balance_loss_mlp": 1.03752589,
      "epoch": 0.17381632346309936,
      "flos": 15998060816640.0,
      "grad_norm": 2.7278091568285596,
      "language_loss": 0.82205319,
      "learning_rate": 3.7869285222095653e-06,
      "loss": 0.84402454,
      "num_input_tokens_seen": 62556820,
      "step": 2891,
      "time_per_iteration": 2.625162363052368
    },
    {
      "auxiliary_loss_clip": 0.01097361,
      "auxiliary_loss_mlp": 0.01050012,
      "balance_loss_clip": 1.04281187,
      "balance_loss_mlp": 1.02876878,
      "epoch": 0.17387644671576732,
      "flos": 13369830670080.0,
      "grad_norm": 1.9017653264876209,
      "language_loss": 0.81200826,
      "learning_rate": 3.7867535677330334e-06,
      "loss": 0.83348203,
      "num_input_tokens_seen": 62572450,
      "step": 2892,
      "time_per_iteration": 2.7682459354400635
    },
    {
      "auxiliary_loss_clip": 0.01148834,
      "auxiliary_loss_mlp": 0.0105551,
      "balance_loss_clip": 1.05707812,
      "balance_loss_mlp": 1.03631687,
      "epoch": 0.1739365699684353,
      "flos": 26615516958720.0,
      "grad_norm": 2.0056711213447436,
      "language_loss": 0.73950225,
      "learning_rate": 3.786578545502627e-06,
      "loss": 0.76154572,
      "num_input_tokens_seen": 62592580,
      "step": 2893,
      "time_per_iteration": 2.8463022708892822
    },
    {
      "auxiliary_loss_clip": 0.01132474,
      "auxiliary_loss_mlp": 0.01043509,
      "balance_loss_clip": 1.05198765,
      "balance_loss_mlp": 1.02443516,
      "epoch": 0.17399669322110325,
      "flos": 23367971491200.0,
      "grad_norm": 4.010773627073901,
      "language_loss": 0.82507658,
      "learning_rate": 3.7864034555249828e-06,
      "loss": 0.84683645,
      "num_input_tokens_seen": 62611220,
      "step": 2894,
      "time_per_iteration": 2.719564914703369
    },
    {
      "auxiliary_loss_clip": 0.01113951,
      "auxiliary_loss_mlp": 0.01046249,
      "balance_loss_clip": 1.0506922,
      "balance_loss_mlp": 1.02463603,
      "epoch": 0.17405681647377122,
      "flos": 22054107813120.0,
      "grad_norm": 2.3322053123967574,
      "language_loss": 0.73826683,
      "learning_rate": 3.786228297806741e-06,
      "loss": 0.7598688,
      "num_input_tokens_seen": 62629185,
      "step": 2895,
      "time_per_iteration": 2.743992805480957
    },
    {
      "auxiliary_loss_clip": 0.01037578,
      "auxiliary_loss_mlp": 0.01011099,
      "balance_loss_clip": 1.0404408,
      "balance_loss_mlp": 1.00788069,
      "epoch": 0.1741169397264392,
      "flos": 61457559114240.0,
      "grad_norm": 0.8765647158253519,
      "language_loss": 0.62754023,
      "learning_rate": 3.7860530723545435e-06,
      "loss": 0.64802706,
      "num_input_tokens_seen": 62691895,
      "step": 2896,
      "time_per_iteration": 3.345099687576294
    },
    {
      "auxiliary_loss_clip": 0.0113101,
      "auxiliary_loss_mlp": 0.00776588,
      "balance_loss_clip": 1.05246758,
      "balance_loss_mlp": 1.00102258,
      "epoch": 0.17417706297910718,
      "flos": 27017680608000.0,
      "grad_norm": 1.7338863964520728,
      "language_loss": 0.75822324,
      "learning_rate": 3.785877779175034e-06,
      "loss": 0.77729923,
      "num_input_tokens_seen": 62713790,
      "step": 2897,
      "time_per_iteration": 2.772292137145996
    },
    {
      "auxiliary_loss_clip": 0.01141357,
      "auxiliary_loss_mlp": 0.01042983,
      "balance_loss_clip": 1.0545547,
      "balance_loss_mlp": 1.02512598,
      "epoch": 0.17423718623177514,
      "flos": 33508856960640.0,
      "grad_norm": 1.944569306659421,
      "language_loss": 0.6883949,
      "learning_rate": 3.7857024182748606e-06,
      "loss": 0.71023834,
      "num_input_tokens_seen": 62736285,
      "step": 2898,
      "time_per_iteration": 2.7278554439544678
    },
    {
      "auxiliary_loss_clip": 0.01128715,
      "auxiliary_loss_mlp": 0.01044216,
      "balance_loss_clip": 1.05251193,
      "balance_loss_mlp": 1.02504694,
      "epoch": 0.1742973094844431,
      "flos": 27198634348800.0,
      "grad_norm": 2.99011081330885,
      "language_loss": 0.76445562,
      "learning_rate": 3.7855269896606717e-06,
      "loss": 0.78618491,
      "num_input_tokens_seen": 62756240,
      "step": 2899,
      "time_per_iteration": 2.8052010536193848
    },
    {
      "auxiliary_loss_clip": 0.01095069,
      "auxiliary_loss_mlp": 0.01045896,
      "balance_loss_clip": 1.04680347,
      "balance_loss_mlp": 1.02632213,
      "epoch": 0.17435743273711107,
      "flos": 22710734386560.0,
      "grad_norm": 3.2965812335226357,
      "language_loss": 0.72860038,
      "learning_rate": 3.785351493339121e-06,
      "loss": 0.75001007,
      "num_input_tokens_seen": 62775910,
      "step": 2900,
      "time_per_iteration": 2.868218421936035
    },
    {
      "auxiliary_loss_clip": 0.01110522,
      "auxiliary_loss_mlp": 0.00776698,
      "balance_loss_clip": 1.05202782,
      "balance_loss_mlp": 1.000983,
      "epoch": 0.17441755598977904,
      "flos": 41646466039680.0,
      "grad_norm": 1.5488662608930523,
      "language_loss": 0.69946706,
      "learning_rate": 3.785175929316863e-06,
      "loss": 0.71833932,
      "num_input_tokens_seen": 62799385,
      "step": 2901,
      "time_per_iteration": 4.407040596008301
    },
    {
      "auxiliary_loss_clip": 0.01129098,
      "auxiliary_loss_mlp": 0.01045525,
      "balance_loss_clip": 1.05246592,
      "balance_loss_mlp": 1.02764344,
      "epoch": 0.174477679242447,
      "flos": 26287077974400.0,
      "grad_norm": 2.1785959913748965,
      "language_loss": 0.76588804,
      "learning_rate": 3.7850002976005543e-06,
      "loss": 0.78763425,
      "num_input_tokens_seen": 62819380,
      "step": 2902,
      "time_per_iteration": 4.2244462966918945
    },
    {
      "auxiliary_loss_clip": 0.01145685,
      "auxiliary_loss_mlp": 0.0104382,
      "balance_loss_clip": 1.0531354,
      "balance_loss_mlp": 1.02567625,
      "epoch": 0.174537802495115,
      "flos": 17858412990720.0,
      "grad_norm": 2.2508699895191073,
      "language_loss": 0.81588745,
      "learning_rate": 3.7848245981968558e-06,
      "loss": 0.83778256,
      "num_input_tokens_seen": 62836205,
      "step": 2903,
      "time_per_iteration": 4.132925271987915
    },
    {
      "auxiliary_loss_clip": 0.01126443,
      "auxiliary_loss_mlp": 0.0103942,
      "balance_loss_clip": 1.05449986,
      "balance_loss_mlp": 1.02135992,
      "epoch": 0.17459792574778296,
      "flos": 16940715390720.0,
      "grad_norm": 2.4085694554154187,
      "language_loss": 0.73316491,
      "learning_rate": 3.784648831112429e-06,
      "loss": 0.75482351,
      "num_input_tokens_seen": 62854045,
      "step": 2904,
      "time_per_iteration": 2.7033374309539795
    },
    {
      "auxiliary_loss_clip": 0.01105192,
      "auxiliary_loss_mlp": 0.0104577,
      "balance_loss_clip": 1.05250716,
      "balance_loss_mlp": 1.02822256,
      "epoch": 0.17465804900045093,
      "flos": 25520026014720.0,
      "grad_norm": 1.8783326609306377,
      "language_loss": 0.64233291,
      "learning_rate": 3.7844729963539406e-06,
      "loss": 0.66384256,
      "num_input_tokens_seen": 62873075,
      "step": 2905,
      "time_per_iteration": 2.8325791358947754
    },
    {
      "auxiliary_loss_clip": 0.01135256,
      "auxiliary_loss_mlp": 0.01053006,
      "balance_loss_clip": 1.05869055,
      "balance_loss_mlp": 1.03370619,
      "epoch": 0.1747181722531189,
      "flos": 24129708238080.0,
      "grad_norm": 2.820817719352069,
      "language_loss": 0.79504299,
      "learning_rate": 3.7842970939280566e-06,
      "loss": 0.81692564,
      "num_input_tokens_seen": 62892675,
      "step": 2906,
      "time_per_iteration": 4.491498231887817
    },
    {
      "auxiliary_loss_clip": 0.01146195,
      "auxiliary_loss_mlp": 0.01050729,
      "balance_loss_clip": 1.05623174,
      "balance_loss_mlp": 1.03258538,
      "epoch": 0.17477829550578686,
      "flos": 17748813617280.0,
      "grad_norm": 2.262709441571415,
      "language_loss": 0.81318873,
      "learning_rate": 3.784121123841449e-06,
      "loss": 0.83515799,
      "num_input_tokens_seen": 62910675,
      "step": 2907,
      "time_per_iteration": 2.6855854988098145
    },
    {
      "auxiliary_loss_clip": 0.01143202,
      "auxiliary_loss_mlp": 0.01043315,
      "balance_loss_clip": 1.05374384,
      "balance_loss_mlp": 1.0253861,
      "epoch": 0.17483841875845482,
      "flos": 15377344865280.0,
      "grad_norm": 2.068635027461873,
      "language_loss": 0.81342787,
      "learning_rate": 3.7839450861007886e-06,
      "loss": 0.83529305,
      "num_input_tokens_seen": 62928130,
      "step": 2908,
      "time_per_iteration": 2.6449570655822754
    },
    {
      "auxiliary_loss_clip": 0.01127136,
      "auxiliary_loss_mlp": 0.01050925,
      "balance_loss_clip": 1.05178046,
      "balance_loss_mlp": 1.03163743,
      "epoch": 0.17489854201112282,
      "flos": 17163254102400.0,
      "grad_norm": 3.147433356867123,
      "language_loss": 0.80020624,
      "learning_rate": 3.7837689807127518e-06,
      "loss": 0.82198691,
      "num_input_tokens_seen": 62944290,
      "step": 2909,
      "time_per_iteration": 2.6820569038391113
    },
    {
      "auxiliary_loss_clip": 0.0109059,
      "auxiliary_loss_mlp": 0.01052625,
      "balance_loss_clip": 1.05020881,
      "balance_loss_mlp": 1.0310595,
      "epoch": 0.17495866526379078,
      "flos": 19755286318080.0,
      "grad_norm": 1.6978440546881337,
      "language_loss": 0.76742244,
      "learning_rate": 3.783592807684017e-06,
      "loss": 0.7888546,
      "num_input_tokens_seen": 62963505,
      "step": 2910,
      "time_per_iteration": 2.6980416774749756
    },
    {
      "auxiliary_loss_clip": 0.01158552,
      "auxiliary_loss_mlp": 0.01049407,
      "balance_loss_clip": 1.05618358,
      "balance_loss_mlp": 1.03059566,
      "epoch": 0.17501878851645875,
      "flos": 28511133310080.0,
      "grad_norm": 1.9812610358315632,
      "language_loss": 0.8698765,
      "learning_rate": 3.7834165670212645e-06,
      "loss": 0.89195609,
      "num_input_tokens_seen": 62985020,
      "step": 2911,
      "time_per_iteration": 2.692662477493286
    },
    {
      "auxiliary_loss_clip": 0.01154744,
      "auxiliary_loss_mlp": 0.00777232,
      "balance_loss_clip": 1.05323184,
      "balance_loss_mlp": 1.00110698,
      "epoch": 0.1750789117691267,
      "flos": 17931203902080.0,
      "grad_norm": 3.030740090796483,
      "language_loss": 0.89883876,
      "learning_rate": 3.7832402587311764e-06,
      "loss": 0.91815847,
      "num_input_tokens_seen": 63001745,
      "step": 2912,
      "time_per_iteration": 2.600738763809204
    },
    {
      "auxiliary_loss_clip": 0.01146165,
      "auxiliary_loss_mlp": 0.01045616,
      "balance_loss_clip": 1.0538094,
      "balance_loss_mlp": 1.02655411,
      "epoch": 0.17513903502179468,
      "flos": 18259427404800.0,
      "grad_norm": 2.03479884577424,
      "language_loss": 0.72818935,
      "learning_rate": 3.783063882820439e-06,
      "loss": 0.75010711,
      "num_input_tokens_seen": 63019750,
      "step": 2913,
      "time_per_iteration": 2.623342275619507
    },
    {
      "auxiliary_loss_clip": 0.01140074,
      "auxiliary_loss_mlp": 0.01043928,
      "balance_loss_clip": 1.05781865,
      "balance_loss_mlp": 1.02557003,
      "epoch": 0.17519915827446264,
      "flos": 20704728562560.0,
      "grad_norm": 2.137073079496124,
      "language_loss": 0.6891731,
      "learning_rate": 3.782887439295741e-06,
      "loss": 0.71101314,
      "num_input_tokens_seen": 63039500,
      "step": 2914,
      "time_per_iteration": 2.7065770626068115
    },
    {
      "auxiliary_loss_clip": 0.01142434,
      "auxiliary_loss_mlp": 0.01045043,
      "balance_loss_clip": 1.05532789,
      "balance_loss_mlp": 1.02649403,
      "epoch": 0.1752592815271306,
      "flos": 20523415685760.0,
      "grad_norm": 2.051329837479214,
      "language_loss": 0.93125081,
      "learning_rate": 3.782710928163772e-06,
      "loss": 0.9531256,
      "num_input_tokens_seen": 63059785,
      "step": 2915,
      "time_per_iteration": 2.659029245376587
    },
    {
      "auxiliary_loss_clip": 0.01114731,
      "auxiliary_loss_mlp": 0.01040999,
      "balance_loss_clip": 1.04957223,
      "balance_loss_mlp": 1.02243853,
      "epoch": 0.1753194047797986,
      "flos": 21799178012160.0,
      "grad_norm": 1.604344576738792,
      "language_loss": 0.81092978,
      "learning_rate": 3.782534349431226e-06,
      "loss": 0.83248705,
      "num_input_tokens_seen": 63079385,
      "step": 2916,
      "time_per_iteration": 2.7099549770355225
    },
    {
      "auxiliary_loss_clip": 0.0114211,
      "auxiliary_loss_mlp": 0.01046221,
      "balance_loss_clip": 1.05090034,
      "balance_loss_mlp": 1.02780342,
      "epoch": 0.17537952803246656,
      "flos": 20668351063680.0,
      "grad_norm": 3.7582760939418716,
      "language_loss": 0.73829222,
      "learning_rate": 3.782357703104799e-06,
      "loss": 0.76017547,
      "num_input_tokens_seen": 63098970,
      "step": 2917,
      "time_per_iteration": 2.666717767715454
    },
    {
      "auxiliary_loss_clip": 0.01133449,
      "auxiliary_loss_mlp": 0.01047353,
      "balance_loss_clip": 1.05319786,
      "balance_loss_mlp": 1.02821994,
      "epoch": 0.17543965128513453,
      "flos": 23295072839040.0,
      "grad_norm": 1.813699779869167,
      "language_loss": 0.76739681,
      "learning_rate": 3.7821809891911897e-06,
      "loss": 0.78920484,
      "num_input_tokens_seen": 63118750,
      "step": 2918,
      "time_per_iteration": 2.647634744644165
    },
    {
      "auxiliary_loss_clip": 0.01093958,
      "auxiliary_loss_mlp": 0.01045643,
      "balance_loss_clip": 1.0476644,
      "balance_loss_mlp": 1.02425694,
      "epoch": 0.1754997745378025,
      "flos": 29095615416960.0,
      "grad_norm": 2.436739755969174,
      "language_loss": 0.73624814,
      "learning_rate": 3.782004207697098e-06,
      "loss": 0.75764406,
      "num_input_tokens_seen": 63136865,
      "step": 2919,
      "time_per_iteration": 2.7904632091522217
    },
    {
      "auxiliary_loss_clip": 0.0112465,
      "auxiliary_loss_mlp": 0.01046524,
      "balance_loss_clip": 1.04938293,
      "balance_loss_mlp": 1.02805829,
      "epoch": 0.17555989779047046,
      "flos": 30371844620160.0,
      "grad_norm": 2.5113730227003814,
      "language_loss": 0.74840331,
      "learning_rate": 3.781827358629228e-06,
      "loss": 0.77011508,
      "num_input_tokens_seen": 63158325,
      "step": 2920,
      "time_per_iteration": 2.727890968322754
    },
    {
      "auxiliary_loss_clip": 0.01117257,
      "auxiliary_loss_mlp": 0.01042893,
      "balance_loss_clip": 1.0462867,
      "balance_loss_mlp": 1.02371216,
      "epoch": 0.17562002104313842,
      "flos": 23287746464640.0,
      "grad_norm": 3.6617213109535536,
      "language_loss": 0.79731411,
      "learning_rate": 3.7816504419942873e-06,
      "loss": 0.81891561,
      "num_input_tokens_seen": 63173115,
      "step": 2921,
      "time_per_iteration": 2.753817558288574
    },
    {
      "auxiliary_loss_clip": 0.01121718,
      "auxiliary_loss_mlp": 0.01046234,
      "balance_loss_clip": 1.05232286,
      "balance_loss_mlp": 1.02679133,
      "epoch": 0.1756801442958064,
      "flos": 24790500789120.0,
      "grad_norm": 2.6301689129577546,
      "language_loss": 0.87826073,
      "learning_rate": 3.7814734577989823e-06,
      "loss": 0.89994025,
      "num_input_tokens_seen": 63192880,
      "step": 2922,
      "time_per_iteration": 2.7411837577819824
    },
    {
      "auxiliary_loss_clip": 0.01144004,
      "auxiliary_loss_mlp": 0.01047403,
      "balance_loss_clip": 1.05196273,
      "balance_loss_mlp": 1.02778149,
      "epoch": 0.17574026754847438,
      "flos": 25771651764480.0,
      "grad_norm": 4.4893841411537085,
      "language_loss": 0.62347209,
      "learning_rate": 3.7812964060500253e-06,
      "loss": 0.64538622,
      "num_input_tokens_seen": 63214395,
      "step": 2923,
      "time_per_iteration": 2.7666683197021484
    },
    {
      "auxiliary_loss_clip": 0.01134872,
      "auxiliary_loss_mlp": 0.01048692,
      "balance_loss_clip": 1.05887377,
      "balance_loss_mlp": 1.02847457,
      "epoch": 0.17580039080114235,
      "flos": 17456608477440.0,
      "grad_norm": 2.8552131957437914,
      "language_loss": 0.80392253,
      "learning_rate": 3.78111928675413e-06,
      "loss": 0.82575822,
      "num_input_tokens_seen": 63231020,
      "step": 2924,
      "time_per_iteration": 2.729403257369995
    },
    {
      "auxiliary_loss_clip": 0.01132783,
      "auxiliary_loss_mlp": 0.01051456,
      "balance_loss_clip": 1.05193377,
      "balance_loss_mlp": 1.03082108,
      "epoch": 0.1758605140538103,
      "flos": 14864648088960.0,
      "grad_norm": 5.080042666316876,
      "language_loss": 0.71374178,
      "learning_rate": 3.7809420999180126e-06,
      "loss": 0.73558426,
      "num_input_tokens_seen": 63246245,
      "step": 2925,
      "time_per_iteration": 2.9538233280181885
    },
    {
      "auxiliary_loss_clip": 0.01117196,
      "auxiliary_loss_mlp": 0.01045706,
      "balance_loss_clip": 1.05052948,
      "balance_loss_mlp": 1.02744341,
      "epoch": 0.17592063730647828,
      "flos": 23004268329600.0,
      "grad_norm": 1.6620026542608322,
      "language_loss": 0.71931666,
      "learning_rate": 3.7807648455483934e-06,
      "loss": 0.74094564,
      "num_input_tokens_seen": 63267790,
      "step": 2926,
      "time_per_iteration": 2.7738964557647705
    },
    {
      "auxiliary_loss_clip": 0.01105944,
      "auxiliary_loss_mlp": 0.01045732,
      "balance_loss_clip": 1.04915071,
      "balance_loss_mlp": 1.02253425,
      "epoch": 0.17598076055914624,
      "flos": 20741501111040.0,
      "grad_norm": 2.6318732447225837,
      "language_loss": 0.84724289,
      "learning_rate": 3.7805875236519918e-06,
      "loss": 0.86875963,
      "num_input_tokens_seen": 63286830,
      "step": 2927,
      "time_per_iteration": 2.704437494277954
    },
    {
      "auxiliary_loss_clip": 0.01100437,
      "auxiliary_loss_mlp": 0.01046684,
      "balance_loss_clip": 1.05039644,
      "balance_loss_mlp": 1.02887452,
      "epoch": 0.1760408838118142,
      "flos": 34092441227520.0,
      "grad_norm": 1.9547597089289632,
      "language_loss": 0.72147644,
      "learning_rate": 3.7804101342355336e-06,
      "loss": 0.74294758,
      "num_input_tokens_seen": 63308870,
      "step": 2928,
      "time_per_iteration": 2.793802261352539
    },
    {
      "auxiliary_loss_clip": 0.01120251,
      "auxiliary_loss_mlp": 0.01045623,
      "balance_loss_clip": 1.0516876,
      "balance_loss_mlp": 1.02679992,
      "epoch": 0.1761010070644822,
      "flos": 24168384207360.0,
      "grad_norm": 1.8474008440192304,
      "language_loss": 0.83097279,
      "learning_rate": 3.780232677305744e-06,
      "loss": 0.85263157,
      "num_input_tokens_seen": 63329005,
      "step": 2929,
      "time_per_iteration": 2.733339786529541
    },
    {
      "auxiliary_loss_clip": 0.01124127,
      "auxiliary_loss_mlp": 0.01042521,
      "balance_loss_clip": 1.04853475,
      "balance_loss_mlp": 1.02479422,
      "epoch": 0.17616113031715017,
      "flos": 26576697335040.0,
      "grad_norm": 2.4427170552109163,
      "language_loss": 0.79211783,
      "learning_rate": 3.7800551528693535e-06,
      "loss": 0.81378424,
      "num_input_tokens_seen": 63349390,
      "step": 2930,
      "time_per_iteration": 2.748080015182495
    },
    {
      "auxiliary_loss_clip": 0.01160654,
      "auxiliary_loss_mlp": 0.01047281,
      "balance_loss_clip": 1.05925918,
      "balance_loss_mlp": 1.02758813,
      "epoch": 0.17622125356981813,
      "flos": 25666685245440.0,
      "grad_norm": 2.504124366499191,
      "language_loss": 0.76502466,
      "learning_rate": 3.7798775609330927e-06,
      "loss": 0.78710401,
      "num_input_tokens_seen": 63368835,
      "step": 2931,
      "time_per_iteration": 2.6691603660583496
    },
    {
      "auxiliary_loss_clip": 0.01076453,
      "auxiliary_loss_mlp": 0.01043586,
      "balance_loss_clip": 1.04577017,
      "balance_loss_mlp": 1.02478647,
      "epoch": 0.1762813768224861,
      "flos": 16508530949760.0,
      "grad_norm": 2.941321746162514,
      "language_loss": 0.76070881,
      "learning_rate": 3.779699901503696e-06,
      "loss": 0.78190923,
      "num_input_tokens_seen": 63385220,
      "step": 2932,
      "time_per_iteration": 2.809630870819092
    },
    {
      "auxiliary_loss_clip": 0.01148627,
      "auxiliary_loss_mlp": 0.01043149,
      "balance_loss_clip": 1.05284405,
      "balance_loss_mlp": 1.0229789,
      "epoch": 0.17634150007515406,
      "flos": 11211850402560.0,
      "grad_norm": 5.168612276821382,
      "language_loss": 0.90027422,
      "learning_rate": 3.7795221745879016e-06,
      "loss": 0.92219198,
      "num_input_tokens_seen": 63400865,
      "step": 2933,
      "time_per_iteration": 2.6665337085723877
    },
    {
      "auxiliary_loss_clip": 0.01154114,
      "auxiliary_loss_mlp": 0.01055985,
      "balance_loss_clip": 1.05539656,
      "balance_loss_mlp": 1.03766203,
      "epoch": 0.17640162332782203,
      "flos": 23659925235840.0,
      "grad_norm": 2.009210784374188,
      "language_loss": 0.88323247,
      "learning_rate": 3.779344380192448e-06,
      "loss": 0.90533352,
      "num_input_tokens_seen": 63421390,
      "step": 2934,
      "time_per_iteration": 2.6649580001831055
    },
    {
      "auxiliary_loss_clip": 0.01128495,
      "auxiliary_loss_mlp": 0.01048067,
      "balance_loss_clip": 1.05581188,
      "balance_loss_mlp": 1.03028131,
      "epoch": 0.17646174658049,
      "flos": 53796984606720.0,
      "grad_norm": 1.6302121247923247,
      "language_loss": 0.70403945,
      "learning_rate": 3.779166518324077e-06,
      "loss": 0.72580504,
      "num_input_tokens_seen": 63444715,
      "step": 2935,
      "time_per_iteration": 3.006019115447998
    },
    {
      "auxiliary_loss_clip": 0.01126189,
      "auxiliary_loss_mlp": 0.01040034,
      "balance_loss_clip": 1.05360174,
      "balance_loss_mlp": 1.02135396,
      "epoch": 0.17652186983315798,
      "flos": 24243868638720.0,
      "grad_norm": 2.5931578566124807,
      "language_loss": 0.69721985,
      "learning_rate": 3.7789885889895325e-06,
      "loss": 0.71888208,
      "num_input_tokens_seen": 63465525,
      "step": 2936,
      "time_per_iteration": 2.7517428398132324
    },
    {
      "auxiliary_loss_clip": 0.01105644,
      "auxiliary_loss_mlp": 0.01045896,
      "balance_loss_clip": 1.05023837,
      "balance_loss_mlp": 1.02737129,
      "epoch": 0.17658199308582595,
      "flos": 27454282421760.0,
      "grad_norm": 1.9170676229980566,
      "language_loss": 0.71288073,
      "learning_rate": 3.7788105921955634e-06,
      "loss": 0.73439616,
      "num_input_tokens_seen": 63485815,
      "step": 2937,
      "time_per_iteration": 2.837181329727173
    },
    {
      "auxiliary_loss_clip": 0.01141008,
      "auxiliary_loss_mlp": 0.01046843,
      "balance_loss_clip": 1.05945122,
      "balance_loss_mlp": 1.02674472,
      "epoch": 0.17664211633849392,
      "flos": 22418672901120.0,
      "grad_norm": 2.267148270780071,
      "language_loss": 0.75439745,
      "learning_rate": 3.7786325279489184e-06,
      "loss": 0.77627593,
      "num_input_tokens_seen": 63503905,
      "step": 2938,
      "time_per_iteration": 2.883162021636963
    },
    {
      "auxiliary_loss_clip": 0.01147345,
      "auxiliary_loss_mlp": 0.01043976,
      "balance_loss_clip": 1.05576169,
      "balance_loss_mlp": 1.02553487,
      "epoch": 0.17670223959116188,
      "flos": 24715124098560.0,
      "grad_norm": 2.921726967662053,
      "language_loss": 0.71015209,
      "learning_rate": 3.7784543962563495e-06,
      "loss": 0.73206532,
      "num_input_tokens_seen": 63521985,
      "step": 2939,
      "time_per_iteration": 2.6938419342041016
    },
    {
      "auxiliary_loss_clip": 0.01160437,
      "auxiliary_loss_mlp": 0.01046921,
      "balance_loss_clip": 1.05818558,
      "balance_loss_mlp": 1.02794337,
      "epoch": 0.17676236284382985,
      "flos": 22527051212160.0,
      "grad_norm": 3.114901170192376,
      "language_loss": 0.73513985,
      "learning_rate": 3.7782761971246115e-06,
      "loss": 0.75721341,
      "num_input_tokens_seen": 63539830,
      "step": 2940,
      "time_per_iteration": 4.145469665527344
    },
    {
      "auxiliary_loss_clip": 0.0112582,
      "auxiliary_loss_mlp": 0.01046611,
      "balance_loss_clip": 1.05631542,
      "balance_loss_mlp": 1.02731109,
      "epoch": 0.1768224860964978,
      "flos": 12385160161920.0,
      "grad_norm": 3.071469776016301,
      "language_loss": 0.85375023,
      "learning_rate": 3.7780979305604616e-06,
      "loss": 0.87547457,
      "num_input_tokens_seen": 63555495,
      "step": 2941,
      "time_per_iteration": 4.279599666595459
    },
    {
      "auxiliary_loss_clip": 0.01161068,
      "auxiliary_loss_mlp": 0.01045254,
      "balance_loss_clip": 1.05717027,
      "balance_loss_mlp": 1.0257628,
      "epoch": 0.1768826093491658,
      "flos": 24353360271360.0,
      "grad_norm": 2.434766510066968,
      "language_loss": 0.76885259,
      "learning_rate": 3.7779195965706607e-06,
      "loss": 0.79091585,
      "num_input_tokens_seen": 63575290,
      "step": 2942,
      "time_per_iteration": 4.2280871868133545
    },
    {
      "auxiliary_loss_clip": 0.01106234,
      "auxiliary_loss_mlp": 0.00780676,
      "balance_loss_clip": 1.04992843,
      "balance_loss_mlp": 1.00087166,
      "epoch": 0.17694273260183377,
      "flos": 23587062497280.0,
      "grad_norm": 3.301743041114179,
      "language_loss": 0.8024286,
      "learning_rate": 3.77774119516197e-06,
      "loss": 0.82129776,
      "num_input_tokens_seen": 63594670,
      "step": 2943,
      "time_per_iteration": 2.8921029567718506
    },
    {
      "auxiliary_loss_clip": 0.01132848,
      "auxiliary_loss_mlp": 0.01052225,
      "balance_loss_clip": 1.05352235,
      "balance_loss_mlp": 1.03124392,
      "epoch": 0.17700285585450173,
      "flos": 26760991040640.0,
      "grad_norm": 5.7613375603973465,
      "language_loss": 0.80809408,
      "learning_rate": 3.777562726341155e-06,
      "loss": 0.82994485,
      "num_input_tokens_seen": 63614780,
      "step": 2944,
      "time_per_iteration": 2.692831039428711
    },
    {
      "auxiliary_loss_clip": 0.01161854,
      "auxiliary_loss_mlp": 0.01056825,
      "balance_loss_clip": 1.05807233,
      "balance_loss_mlp": 1.03796625,
      "epoch": 0.1770629791071697,
      "flos": 42776323320960.0,
      "grad_norm": 2.4257754996125227,
      "language_loss": 0.73812854,
      "learning_rate": 3.7773841901149835e-06,
      "loss": 0.7603153,
      "num_input_tokens_seen": 63637190,
      "step": 2945,
      "time_per_iteration": 2.782910108566284
    },
    {
      "auxiliary_loss_clip": 0.011481,
      "auxiliary_loss_mlp": 0.01047361,
      "balance_loss_clip": 1.05756998,
      "balance_loss_mlp": 1.02862108,
      "epoch": 0.17712310235983766,
      "flos": 17345572560000.0,
      "grad_norm": 2.8106797532110637,
      "language_loss": 0.7793628,
      "learning_rate": 3.7772055864902256e-06,
      "loss": 0.80131739,
      "num_input_tokens_seen": 63652140,
      "step": 2946,
      "time_per_iteration": 4.278741121292114
    },
    {
      "auxiliary_loss_clip": 0.01109059,
      "auxiliary_loss_mlp": 0.01052842,
      "balance_loss_clip": 1.04997015,
      "balance_loss_mlp": 1.03341079,
      "epoch": 0.17718322561250563,
      "flos": 23878477537920.0,
      "grad_norm": 2.172386857191393,
      "language_loss": 0.76068008,
      "learning_rate": 3.7770269154736535e-06,
      "loss": 0.7822991,
      "num_input_tokens_seen": 63671700,
      "step": 2947,
      "time_per_iteration": 2.7949914932250977
    },
    {
      "auxiliary_loss_clip": 0.0114934,
      "auxiliary_loss_mlp": 0.01044342,
      "balance_loss_clip": 1.05480659,
      "balance_loss_mlp": 1.025388,
      "epoch": 0.1772433488651736,
      "flos": 36466352104320.0,
      "grad_norm": 2.6793588646204745,
      "language_loss": 0.72557831,
      "learning_rate": 3.7768481770720424e-06,
      "loss": 0.74751514,
      "num_input_tokens_seen": 63691685,
      "step": 2948,
      "time_per_iteration": 2.901662826538086
    },
    {
      "auxiliary_loss_clip": 0.01151572,
      "auxiliary_loss_mlp": 0.01050692,
      "balance_loss_clip": 1.05921662,
      "balance_loss_mlp": 1.03236949,
      "epoch": 0.1773034721178416,
      "flos": 26684716510080.0,
      "grad_norm": 1.8296543316983853,
      "language_loss": 0.81782824,
      "learning_rate": 3.776669371292171e-06,
      "loss": 0.8398509,
      "num_input_tokens_seen": 63711720,
      "step": 2949,
      "time_per_iteration": 2.7284891605377197
    },
    {
      "auxiliary_loss_clip": 0.01080853,
      "auxiliary_loss_mlp": 0.0100651,
      "balance_loss_clip": 1.04975748,
      "balance_loss_mlp": 1.00226629,
      "epoch": 0.17736359537050955,
      "flos": 57117467617920.0,
      "grad_norm": 0.768126622018234,
      "language_loss": 0.64989161,
      "learning_rate": 3.7764904981408186e-06,
      "loss": 0.67076528,
      "num_input_tokens_seen": 63776280,
      "step": 2950,
      "time_per_iteration": 3.2761552333831787
    },
    {
      "auxiliary_loss_clip": 0.01121454,
      "auxiliary_loss_mlp": 0.01045861,
      "balance_loss_clip": 1.05373287,
      "balance_loss_mlp": 1.02743077,
      "epoch": 0.17742371862317752,
      "flos": 27198203385600.0,
      "grad_norm": 2.9882590699755927,
      "language_loss": 0.83619881,
      "learning_rate": 3.7763115576247686e-06,
      "loss": 0.85787189,
      "num_input_tokens_seen": 63797535,
      "step": 2951,
      "time_per_iteration": 2.7637627124786377
    },
    {
      "auxiliary_loss_clip": 0.01125929,
      "auxiliary_loss_mlp": 0.01046039,
      "balance_loss_clip": 1.05109882,
      "balance_loss_mlp": 1.02682269,
      "epoch": 0.17748384187584548,
      "flos": 20959694277120.0,
      "grad_norm": 2.3133151959471796,
      "language_loss": 0.80395055,
      "learning_rate": 3.776132549750806e-06,
      "loss": 0.82567012,
      "num_input_tokens_seen": 63817045,
      "step": 2952,
      "time_per_iteration": 2.7605957984924316
    },
    {
      "auxiliary_loss_clip": 0.01162679,
      "auxiliary_loss_mlp": 0.01044862,
      "balance_loss_clip": 1.05858529,
      "balance_loss_mlp": 1.02513337,
      "epoch": 0.17754396512851345,
      "flos": 25009986844800.0,
      "grad_norm": 2.8185319653472116,
      "language_loss": 0.79273909,
      "learning_rate": 3.7759534745257194e-06,
      "loss": 0.81481451,
      "num_input_tokens_seen": 63837665,
      "step": 2953,
      "time_per_iteration": 2.798912525177002
    },
    {
      "auxiliary_loss_clip": 0.0112399,
      "auxiliary_loss_mlp": 0.01043314,
      "balance_loss_clip": 1.05482125,
      "balance_loss_mlp": 1.02470589,
      "epoch": 0.1776040883811814,
      "flos": 32051566275840.0,
      "grad_norm": 2.017710353628998,
      "language_loss": 0.87963271,
      "learning_rate": 3.7757743319562994e-06,
      "loss": 0.90130568,
      "num_input_tokens_seen": 63858455,
      "step": 2954,
      "time_per_iteration": 2.838931083679199
    },
    {
      "auxiliary_loss_clip": 0.01144028,
      "auxiliary_loss_mlp": 0.01052958,
      "balance_loss_clip": 1.06043494,
      "balance_loss_mlp": 1.03296697,
      "epoch": 0.17766421163384938,
      "flos": 21574125348480.0,
      "grad_norm": 1.9130853947826985,
      "language_loss": 0.85313326,
      "learning_rate": 3.7755951220493386e-06,
      "loss": 0.87510312,
      "num_input_tokens_seen": 63876935,
      "step": 2955,
      "time_per_iteration": 2.7965714931488037
    },
    {
      "auxiliary_loss_clip": 0.01127677,
      "auxiliary_loss_mlp": 0.01047004,
      "balance_loss_clip": 1.05093336,
      "balance_loss_mlp": 1.02660692,
      "epoch": 0.17772433488651737,
      "flos": 22419319345920.0,
      "grad_norm": 18.24238703278013,
      "language_loss": 0.71152055,
      "learning_rate": 3.7754158448116327e-06,
      "loss": 0.73326737,
      "num_input_tokens_seen": 63896815,
      "step": 2956,
      "time_per_iteration": 2.8358442783355713
    },
    {
      "auxiliary_loss_clip": 0.01150063,
      "auxiliary_loss_mlp": 0.010506,
      "balance_loss_clip": 1.05813813,
      "balance_loss_mlp": 1.03156281,
      "epoch": 0.17778445813918534,
      "flos": 25629445820160.0,
      "grad_norm": 2.981126112172262,
      "language_loss": 0.82881534,
      "learning_rate": 3.7752365002499795e-06,
      "loss": 0.85082197,
      "num_input_tokens_seen": 63916140,
      "step": 2957,
      "time_per_iteration": 2.7034976482391357
    },
    {
      "auxiliary_loss_clip": 0.01100452,
      "auxiliary_loss_mlp": 0.01047239,
      "balance_loss_clip": 1.04976833,
      "balance_loss_mlp": 1.02789164,
      "epoch": 0.1778445813918533,
      "flos": 25628871202560.0,
      "grad_norm": 2.7180995933425622,
      "language_loss": 0.75164193,
      "learning_rate": 3.7750570883711807e-06,
      "loss": 0.77311885,
      "num_input_tokens_seen": 63935220,
      "step": 2958,
      "time_per_iteration": 2.8312718868255615
    },
    {
      "auxiliary_loss_clip": 0.01146025,
      "auxiliary_loss_mlp": 0.01043359,
      "balance_loss_clip": 1.06117964,
      "balance_loss_mlp": 1.02502513,
      "epoch": 0.17790470464452127,
      "flos": 22345522853760.0,
      "grad_norm": 9.439636088267013,
      "language_loss": 0.80363399,
      "learning_rate": 3.7748776091820397e-06,
      "loss": 0.82552785,
      "num_input_tokens_seen": 63954550,
      "step": 2959,
      "time_per_iteration": 2.722102642059326
    },
    {
      "auxiliary_loss_clip": 0.01164621,
      "auxiliary_loss_mlp": 0.01049069,
      "balance_loss_clip": 1.05812871,
      "balance_loss_mlp": 1.02938771,
      "epoch": 0.17796482789718923,
      "flos": 18765875214720.0,
      "grad_norm": 2.62580469975692,
      "language_loss": 0.51511085,
      "learning_rate": 3.774698062689362e-06,
      "loss": 0.53724772,
      "num_input_tokens_seen": 63972425,
      "step": 2960,
      "time_per_iteration": 2.6222047805786133
    },
    {
      "auxiliary_loss_clip": 0.01111843,
      "auxiliary_loss_mlp": 0.01052801,
      "balance_loss_clip": 1.05275989,
      "balance_loss_mlp": 1.03228474,
      "epoch": 0.1780249511498572,
      "flos": 23440941970560.0,
      "grad_norm": 1.7626913000215665,
      "language_loss": 0.88908094,
      "learning_rate": 3.7745184488999548e-06,
      "loss": 0.91072738,
      "num_input_tokens_seen": 63992165,
      "step": 2961,
      "time_per_iteration": 2.8088786602020264
    },
    {
      "auxiliary_loss_clip": 0.01116231,
      "auxiliary_loss_mlp": 0.01054867,
      "balance_loss_clip": 1.05181062,
      "balance_loss_mlp": 1.03385067,
      "epoch": 0.1780850744025252,
      "flos": 23367468700800.0,
      "grad_norm": 1.716412227369414,
      "language_loss": 0.79170465,
      "learning_rate": 3.774338767820631e-06,
      "loss": 0.81341565,
      "num_input_tokens_seen": 64013470,
      "step": 2962,
      "time_per_iteration": 2.7546913623809814
    },
    {
      "auxiliary_loss_clip": 0.01145526,
      "auxiliary_loss_mlp": 0.01052794,
      "balance_loss_clip": 1.05649889,
      "balance_loss_mlp": 1.03104997,
      "epoch": 0.17814519765519315,
      "flos": 13771994319360.0,
      "grad_norm": 2.3241756501763446,
      "language_loss": 0.74910223,
      "learning_rate": 3.774159019458203e-06,
      "loss": 0.77108544,
      "num_input_tokens_seen": 64030975,
      "step": 2963,
      "time_per_iteration": 2.680356979370117
    },
    {
      "auxiliary_loss_clip": 0.01140656,
      "auxiliary_loss_mlp": 0.01043225,
      "balance_loss_clip": 1.05769885,
      "balance_loss_mlp": 1.02347231,
      "epoch": 0.17820532090786112,
      "flos": 21976396738560.0,
      "grad_norm": 1.747536927551571,
      "language_loss": 0.78837025,
      "learning_rate": 3.7739792038194877e-06,
      "loss": 0.81020904,
      "num_input_tokens_seen": 64050075,
      "step": 2964,
      "time_per_iteration": 2.748398780822754
    },
    {
      "auxiliary_loss_clip": 0.01151685,
      "auxiliary_loss_mlp": 0.00776982,
      "balance_loss_clip": 1.05950594,
      "balance_loss_mlp": 1.00098181,
      "epoch": 0.17826544416052909,
      "flos": 24790752184320.0,
      "grad_norm": 3.046027397796258,
      "language_loss": 0.81160808,
      "learning_rate": 3.7737993209113027e-06,
      "loss": 0.83089471,
      "num_input_tokens_seen": 64071920,
      "step": 2965,
      "time_per_iteration": 2.8090012073516846
    },
    {
      "auxiliary_loss_clip": 0.01151658,
      "auxiliary_loss_mlp": 0.01047086,
      "balance_loss_clip": 1.06002402,
      "balance_loss_mlp": 1.02916884,
      "epoch": 0.17832556741319705,
      "flos": 13879582531200.0,
      "grad_norm": 2.554359630612449,
      "language_loss": 0.95307338,
      "learning_rate": 3.7736193707404698e-06,
      "loss": 0.97506082,
      "num_input_tokens_seen": 64086835,
      "step": 2966,
      "time_per_iteration": 2.7159550189971924
    },
    {
      "auxiliary_loss_clip": 0.01112928,
      "auxiliary_loss_mlp": 0.00777395,
      "balance_loss_clip": 1.05336046,
      "balance_loss_mlp": 1.00083637,
      "epoch": 0.17838569066586502,
      "flos": 36641703323520.0,
      "grad_norm": 7.5683867487642065,
      "language_loss": 0.72833109,
      "learning_rate": 3.7734393533138127e-06,
      "loss": 0.74723434,
      "num_input_tokens_seen": 64107360,
      "step": 2967,
      "time_per_iteration": 2.9540669918060303
    },
    {
      "auxiliary_loss_clip": 0.01129124,
      "auxiliary_loss_mlp": 0.01046817,
      "balance_loss_clip": 1.05574143,
      "balance_loss_mlp": 1.02775562,
      "epoch": 0.17844581391853298,
      "flos": 18727271072640.0,
      "grad_norm": 2.1617023205672523,
      "language_loss": 0.76897681,
      "learning_rate": 3.773259268638157e-06,
      "loss": 0.7907362,
      "num_input_tokens_seen": 64124690,
      "step": 2968,
      "time_per_iteration": 2.752717971801758
    },
    {
      "auxiliary_loss_clip": 0.01085006,
      "auxiliary_loss_mlp": 0.01044958,
      "balance_loss_clip": 1.04640651,
      "balance_loss_mlp": 1.02559829,
      "epoch": 0.17850593717120097,
      "flos": 27378259286400.0,
      "grad_norm": 2.039560504387258,
      "language_loss": 0.75839806,
      "learning_rate": 3.7730791167203333e-06,
      "loss": 0.77969772,
      "num_input_tokens_seen": 64146315,
      "step": 2969,
      "time_per_iteration": 2.9161994457244873
    },
    {
      "auxiliary_loss_clip": 0.01075271,
      "auxiliary_loss_mlp": 0.01013071,
      "balance_loss_clip": 1.06177902,
      "balance_loss_mlp": 1.00932813,
      "epoch": 0.17856606042386894,
      "flos": 66996025084800.0,
      "grad_norm": 0.8520394227890811,
      "language_loss": 0.69012916,
      "learning_rate": 3.772898897567171e-06,
      "loss": 0.7110126,
      "num_input_tokens_seen": 64210875,
      "step": 2970,
      "time_per_iteration": 3.3269262313842773
    },
    {
      "auxiliary_loss_clip": 0.011313,
      "auxiliary_loss_mlp": 0.01044166,
      "balance_loss_clip": 1.05561864,
      "balance_loss_mlp": 1.02493763,
      "epoch": 0.1786261836765369,
      "flos": 36977001805440.0,
      "grad_norm": 1.9951166568015506,
      "language_loss": 0.67617297,
      "learning_rate": 3.772718611185505e-06,
      "loss": 0.69792765,
      "num_input_tokens_seen": 64230740,
      "step": 2971,
      "time_per_iteration": 2.8691961765289307
    },
    {
      "auxiliary_loss_clip": 0.01110831,
      "auxiliary_loss_mlp": 0.01052779,
      "balance_loss_clip": 1.05309939,
      "balance_loss_mlp": 1.03266823,
      "epoch": 0.17868630692920487,
      "flos": 24825441744000.0,
      "grad_norm": 1.5664358375440484,
      "language_loss": 0.8971802,
      "learning_rate": 3.7725382575821717e-06,
      "loss": 0.91881633,
      "num_input_tokens_seen": 64252300,
      "step": 2972,
      "time_per_iteration": 2.893923759460449
    },
    {
      "auxiliary_loss_clip": 0.01124705,
      "auxiliary_loss_mlp": 0.01055871,
      "balance_loss_clip": 1.05635929,
      "balance_loss_mlp": 1.03466403,
      "epoch": 0.17874643018187283,
      "flos": 16981977139200.0,
      "grad_norm": 2.4611679901229153,
      "language_loss": 0.88593906,
      "learning_rate": 3.77235783676401e-06,
      "loss": 0.90774482,
      "num_input_tokens_seen": 64270105,
      "step": 2973,
      "time_per_iteration": 2.7340333461761475
    },
    {
      "auxiliary_loss_clip": 0.01164127,
      "auxiliary_loss_mlp": 0.01047073,
      "balance_loss_clip": 1.06285155,
      "balance_loss_mlp": 1.0283215,
      "epoch": 0.1788065534345408,
      "flos": 21032233793280.0,
      "grad_norm": 3.4039298885336557,
      "language_loss": 0.7668556,
      "learning_rate": 3.7721773487378615e-06,
      "loss": 0.78896761,
      "num_input_tokens_seen": 64287250,
      "step": 2974,
      "time_per_iteration": 2.632495403289795
    },
    {
      "auxiliary_loss_clip": 0.0114187,
      "auxiliary_loss_mlp": 0.01053,
      "balance_loss_clip": 1.06101942,
      "balance_loss_mlp": 1.03390288,
      "epoch": 0.17886667668720876,
      "flos": 23987717775360.0,
      "grad_norm": 2.484949778027245,
      "language_loss": 0.74701655,
      "learning_rate": 3.7719967935105705e-06,
      "loss": 0.76896524,
      "num_input_tokens_seen": 64307140,
      "step": 2975,
      "time_per_iteration": 2.704012870788574
    },
    {
      "auxiliary_loss_clip": 0.01149026,
      "auxiliary_loss_mlp": 0.01048788,
      "balance_loss_clip": 1.05678535,
      "balance_loss_mlp": 1.03004813,
      "epoch": 0.17892679993987676,
      "flos": 25739476156800.0,
      "grad_norm": 1.518747487377626,
      "language_loss": 0.73032069,
      "learning_rate": 3.7718161710889833e-06,
      "loss": 0.75229883,
      "num_input_tokens_seen": 64328760,
      "step": 2976,
      "time_per_iteration": 2.7357017993927
    },
    {
      "auxiliary_loss_clip": 0.01150398,
      "auxiliary_loss_mlp": 0.01038685,
      "balance_loss_clip": 1.06239033,
      "balance_loss_mlp": 1.0229373,
      "epoch": 0.17898692319254472,
      "flos": 25699686865920.0,
      "grad_norm": 1.4579507247258654,
      "language_loss": 0.770594,
      "learning_rate": 3.7716354814799495e-06,
      "loss": 0.79248488,
      "num_input_tokens_seen": 64348800,
      "step": 2977,
      "time_per_iteration": 2.727318286895752
    },
    {
      "auxiliary_loss_clip": 0.01131521,
      "auxiliary_loss_mlp": 0.01045834,
      "balance_loss_clip": 1.06618452,
      "balance_loss_mlp": 1.02841735,
      "epoch": 0.1790470464452127,
      "flos": 19317786664320.0,
      "grad_norm": 2.7286854986191282,
      "language_loss": 0.80235189,
      "learning_rate": 3.7714547246903203e-06,
      "loss": 0.82412547,
      "num_input_tokens_seen": 64367955,
      "step": 2978,
      "time_per_iteration": 2.8178791999816895
    },
    {
      "auxiliary_loss_clip": 0.0114307,
      "auxiliary_loss_mlp": 0.01052978,
      "balance_loss_clip": 1.05818772,
      "balance_loss_mlp": 1.03330874,
      "epoch": 0.17910716969788065,
      "flos": 30044267562240.0,
      "grad_norm": 1.4967765935497133,
      "language_loss": 0.76192784,
      "learning_rate": 3.7712739007269508e-06,
      "loss": 0.7838884,
      "num_input_tokens_seen": 64389805,
      "step": 2979,
      "time_per_iteration": 4.241487741470337
    },
    {
      "auxiliary_loss_clip": 0.01122958,
      "auxiliary_loss_mlp": 0.0104457,
      "balance_loss_clip": 1.0590893,
      "balance_loss_mlp": 1.02660525,
      "epoch": 0.17916729295054862,
      "flos": 19427709260160.0,
      "grad_norm": 1.9491816848203256,
      "language_loss": 0.68945503,
      "learning_rate": 3.7710930095966976e-06,
      "loss": 0.71113026,
      "num_input_tokens_seen": 64408220,
      "step": 2980,
      "time_per_iteration": 2.6817352771759033
    },
    {
      "auxiliary_loss_clip": 0.01152986,
      "auxiliary_loss_mlp": 0.0104519,
      "balance_loss_clip": 1.0588038,
      "balance_loss_mlp": 1.02497244,
      "epoch": 0.17922741620321658,
      "flos": 14611549881600.0,
      "grad_norm": 1.9134992191513662,
      "language_loss": 0.70793843,
      "learning_rate": 3.7709120513064196e-06,
      "loss": 0.72992027,
      "num_input_tokens_seen": 64426380,
      "step": 2981,
      "time_per_iteration": 4.310532331466675
    },
    {
      "auxiliary_loss_clip": 0.01137747,
      "auxiliary_loss_mlp": 0.01056086,
      "balance_loss_clip": 1.06083858,
      "balance_loss_mlp": 1.03686976,
      "epoch": 0.17928753945588458,
      "flos": 17165301177600.0,
      "grad_norm": 2.529665562311581,
      "language_loss": 0.8190546,
      "learning_rate": 3.7707310258629796e-06,
      "loss": 0.84099293,
      "num_input_tokens_seen": 64444355,
      "step": 2982,
      "time_per_iteration": 2.710726261138916
    },
    {
      "auxiliary_loss_clip": 0.01162978,
      "auxiliary_loss_mlp": 0.01041014,
      "balance_loss_clip": 1.06181359,
      "balance_loss_mlp": 1.02306128,
      "epoch": 0.17934766270855254,
      "flos": 31395622060800.0,
      "grad_norm": 1.6440716861921114,
      "language_loss": 0.83123535,
      "learning_rate": 3.7705499332732413e-06,
      "loss": 0.85327524,
      "num_input_tokens_seen": 64467800,
      "step": 2983,
      "time_per_iteration": 2.700378656387329
    },
    {
      "auxiliary_loss_clip": 0.01153001,
      "auxiliary_loss_mlp": 0.01048341,
      "balance_loss_clip": 1.05694914,
      "balance_loss_mlp": 1.02932739,
      "epoch": 0.1794077859612205,
      "flos": 20814184281600.0,
      "grad_norm": 1.6703280507743268,
      "language_loss": 0.85149562,
      "learning_rate": 3.7703687735440718e-06,
      "loss": 0.87350899,
      "num_input_tokens_seen": 64487230,
      "step": 2984,
      "time_per_iteration": 2.6529407501220703
    },
    {
      "auxiliary_loss_clip": 0.01126981,
      "auxiliary_loss_mlp": 0.01043442,
      "balance_loss_clip": 1.05520201,
      "balance_loss_mlp": 1.02424896,
      "epoch": 0.17946790921388847,
      "flos": 28986447006720.0,
      "grad_norm": 2.4609160562432053,
      "language_loss": 0.8935222,
      "learning_rate": 3.7701875466823416e-06,
      "loss": 0.9152264,
      "num_input_tokens_seen": 64509165,
      "step": 2985,
      "time_per_iteration": 4.528426170349121
    },
    {
      "auxiliary_loss_clip": 0.01160091,
      "auxiliary_loss_mlp": 0.01040749,
      "balance_loss_clip": 1.06142831,
      "balance_loss_mlp": 1.02434587,
      "epoch": 0.17952803246655644,
      "flos": 20737406960640.0,
      "grad_norm": 2.095497349072142,
      "language_loss": 0.69538593,
      "learning_rate": 3.770006252694922e-06,
      "loss": 0.71739429,
      "num_input_tokens_seen": 64527940,
      "step": 2986,
      "time_per_iteration": 2.6890172958374023
    },
    {
      "auxiliary_loss_clip": 0.01158556,
      "auxiliary_loss_mlp": 0.00776,
      "balance_loss_clip": 1.05752599,
      "balance_loss_mlp": 1.00081134,
      "epoch": 0.1795881557192244,
      "flos": 28255988027520.0,
      "grad_norm": 2.4599229747435123,
      "language_loss": 0.77855188,
      "learning_rate": 3.769824891588688e-06,
      "loss": 0.79789746,
      "num_input_tokens_seen": 64545230,
      "step": 2987,
      "time_per_iteration": 2.650761842727661
    },
    {
      "auxiliary_loss_clip": 0.0116216,
      "auxiliary_loss_mlp": 0.01043775,
      "balance_loss_clip": 1.05775642,
      "balance_loss_mlp": 1.02441502,
      "epoch": 0.17964827897189237,
      "flos": 18552027594240.0,
      "grad_norm": 2.0190394876224467,
      "language_loss": 0.77958816,
      "learning_rate": 3.7696434633705164e-06,
      "loss": 0.80164748,
      "num_input_tokens_seen": 64563820,
      "step": 2988,
      "time_per_iteration": 2.6151437759399414
    },
    {
      "auxiliary_loss_clip": 0.01059513,
      "auxiliary_loss_mlp": 0.00756906,
      "balance_loss_clip": 1.07071137,
      "balance_loss_mlp": 1.00131369,
      "epoch": 0.17970840222456036,
      "flos": 58165088711040.0,
      "grad_norm": 0.7650122273387262,
      "language_loss": 0.62709254,
      "learning_rate": 3.7694619680472875e-06,
      "loss": 0.64525676,
      "num_input_tokens_seen": 64621315,
      "step": 2989,
      "time_per_iteration": 3.1990275382995605
    },
    {
      "auxiliary_loss_clip": 0.01137168,
      "auxiliary_loss_mlp": 0.01038826,
      "balance_loss_clip": 1.05553865,
      "balance_loss_mlp": 1.02128983,
      "epoch": 0.17976852547722832,
      "flos": 20300805146880.0,
      "grad_norm": 2.3566032567209483,
      "language_loss": 0.71070904,
      "learning_rate": 3.7692804056258837e-06,
      "loss": 0.73246896,
      "num_input_tokens_seen": 64639885,
      "step": 2990,
      "time_per_iteration": 2.7275335788726807
    },
    {
      "auxiliary_loss_clip": 0.01135847,
      "auxiliary_loss_mlp": 0.01044966,
      "balance_loss_clip": 1.05398035,
      "balance_loss_mlp": 1.02639365,
      "epoch": 0.1798286487298963,
      "flos": 39669367685760.0,
      "grad_norm": 1.8035266350414116,
      "language_loss": 0.68888462,
      "learning_rate": 3.7690987761131893e-06,
      "loss": 0.7106927,
      "num_input_tokens_seen": 64661220,
      "step": 2991,
      "time_per_iteration": 2.8237311840057373
    },
    {
      "auxiliary_loss_clip": 0.01104375,
      "auxiliary_loss_mlp": 0.01046061,
      "balance_loss_clip": 1.05156851,
      "balance_loss_mlp": 1.02663028,
      "epoch": 0.17988877198256426,
      "flos": 25520313323520.0,
      "grad_norm": 1.6063564491400402,
      "language_loss": 0.82933879,
      "learning_rate": 3.7689170795160924e-06,
      "loss": 0.85084313,
      "num_input_tokens_seen": 64682530,
      "step": 2992,
      "time_per_iteration": 2.8303778171539307
    },
    {
      "auxiliary_loss_clip": 0.01140805,
      "auxiliary_loss_mlp": 0.01035603,
      "balance_loss_clip": 1.05302262,
      "balance_loss_mlp": 1.0187583,
      "epoch": 0.17994889523523222,
      "flos": 18807496099200.0,
      "grad_norm": 2.076285453641059,
      "language_loss": 0.82228035,
      "learning_rate": 3.7687353158414822e-06,
      "loss": 0.84404445,
      "num_input_tokens_seen": 64701025,
      "step": 2993,
      "time_per_iteration": 2.710369110107422
    },
    {
      "auxiliary_loss_clip": 0.01135151,
      "auxiliary_loss_mlp": 0.01040493,
      "balance_loss_clip": 1.05135202,
      "balance_loss_mlp": 1.02236176,
      "epoch": 0.18000901848790019,
      "flos": 21104450087040.0,
      "grad_norm": 1.7027458997386926,
      "language_loss": 0.78129464,
      "learning_rate": 3.7685534850962517e-06,
      "loss": 0.80305111,
      "num_input_tokens_seen": 64719570,
      "step": 2994,
      "time_per_iteration": 2.6666738986968994
    },
    {
      "auxiliary_loss_clip": 0.01158877,
      "auxiliary_loss_mlp": 0.01045455,
      "balance_loss_clip": 1.05657315,
      "balance_loss_mlp": 1.02819359,
      "epoch": 0.18006914174056818,
      "flos": 19646441130240.0,
      "grad_norm": 2.4198973911698434,
      "language_loss": 0.81139499,
      "learning_rate": 3.768371587287296e-06,
      "loss": 0.83343828,
      "num_input_tokens_seen": 64738110,
      "step": 2995,
      "time_per_iteration": 2.699521541595459
    },
    {
      "auxiliary_loss_clip": 0.01142902,
      "auxiliary_loss_mlp": 0.01047606,
      "balance_loss_clip": 1.05350447,
      "balance_loss_mlp": 1.0310601,
      "epoch": 0.18012926499323614,
      "flos": 19499889640320.0,
      "grad_norm": 1.8607496799697536,
      "language_loss": 0.84162772,
      "learning_rate": 3.768189622421512e-06,
      "loss": 0.86353278,
      "num_input_tokens_seen": 64756345,
      "step": 2996,
      "time_per_iteration": 2.696723461151123
    },
    {
      "auxiliary_loss_clip": 0.01127214,
      "auxiliary_loss_mlp": 0.01039953,
      "balance_loss_clip": 1.06094205,
      "balance_loss_mlp": 1.02273917,
      "epoch": 0.1801893882459041,
      "flos": 19464553635840.0,
      "grad_norm": 2.1291201116421283,
      "language_loss": 0.88189137,
      "learning_rate": 3.7680075905058006e-06,
      "loss": 0.90356302,
      "num_input_tokens_seen": 64776375,
      "step": 2997,
      "time_per_iteration": 2.785522699356079
    },
    {
      "auxiliary_loss_clip": 0.01134376,
      "auxiliary_loss_mlp": 0.01045962,
      "balance_loss_clip": 1.04949927,
      "balance_loss_mlp": 1.02753246,
      "epoch": 0.18024951149857207,
      "flos": 26870590414080.0,
      "grad_norm": 1.7579499924576911,
      "language_loss": 0.85068727,
      "learning_rate": 3.7678254915470643e-06,
      "loss": 0.87249064,
      "num_input_tokens_seen": 64796210,
      "step": 2998,
      "time_per_iteration": 2.6912384033203125
    },
    {
      "auxiliary_loss_clip": 0.01159537,
      "auxiliary_loss_mlp": 0.01044427,
      "balance_loss_clip": 1.06019807,
      "balance_loss_mlp": 1.02641416,
      "epoch": 0.18030963475124004,
      "flos": 30226621933440.0,
      "grad_norm": 1.8075624565441775,
      "language_loss": 0.84176779,
      "learning_rate": 3.7676433255522084e-06,
      "loss": 0.86380744,
      "num_input_tokens_seen": 64818590,
      "step": 2999,
      "time_per_iteration": 2.722447395324707
    },
    {
      "auxiliary_loss_clip": 0.01143605,
      "auxiliary_loss_mlp": 0.01047321,
      "balance_loss_clip": 1.05324686,
      "balance_loss_mlp": 1.02870023,
      "epoch": 0.180369758003908,
      "flos": 22307493329280.0,
      "grad_norm": 1.8789697336390492,
      "language_loss": 0.75206578,
      "learning_rate": 3.76746109252814e-06,
      "loss": 0.77397501,
      "num_input_tokens_seen": 64838350,
      "step": 3000,
      "time_per_iteration": 2.669875144958496
    },
    {
      "auxiliary_loss_clip": 0.01130052,
      "auxiliary_loss_mlp": 0.00775745,
      "balance_loss_clip": 1.0526886,
      "balance_loss_mlp": 1.00060582,
      "epoch": 0.18042988125657597,
      "flos": 23732033788800.0,
      "grad_norm": 2.1714361871851704,
      "language_loss": 0.71088028,
      "learning_rate": 3.76727879248177e-06,
      "loss": 0.72993821,
      "num_input_tokens_seen": 64858065,
      "step": 3001,
      "time_per_iteration": 2.7207603454589844
    },
    {
      "auxiliary_loss_clip": 0.01150091,
      "auxiliary_loss_mlp": 0.01044695,
      "balance_loss_clip": 1.05701649,
      "balance_loss_mlp": 1.02605033,
      "epoch": 0.18049000450924396,
      "flos": 24093582134400.0,
      "grad_norm": 2.218812983953599,
      "language_loss": 0.8849982,
      "learning_rate": 3.767096425420011e-06,
      "loss": 0.90694606,
      "num_input_tokens_seen": 64877305,
      "step": 3002,
      "time_per_iteration": 2.6577625274658203
    },
    {
      "auxiliary_loss_clip": 0.01157827,
      "auxiliary_loss_mlp": 0.01048268,
      "balance_loss_clip": 1.05624068,
      "balance_loss_mlp": 1.03076851,
      "epoch": 0.18055012776191193,
      "flos": 22163168482560.0,
      "grad_norm": 1.6287780165264572,
      "language_loss": 0.80328667,
      "learning_rate": 3.7669139913497788e-06,
      "loss": 0.8253476,
      "num_input_tokens_seen": 64896955,
      "step": 3003,
      "time_per_iteration": 2.6274783611297607
    },
    {
      "auxiliary_loss_clip": 0.01158367,
      "auxiliary_loss_mlp": 0.01043654,
      "balance_loss_clip": 1.05622995,
      "balance_loss_mlp": 1.02596307,
      "epoch": 0.1806102510145799,
      "flos": 28913512440960.0,
      "grad_norm": 2.3308952017896956,
      "language_loss": 0.67250973,
      "learning_rate": 3.7667314902779907e-06,
      "loss": 0.69452989,
      "num_input_tokens_seen": 64917080,
      "step": 3004,
      "time_per_iteration": 2.6652631759643555
    },
    {
      "auxiliary_loss_clip": 0.01147517,
      "auxiliary_loss_mlp": 0.01054518,
      "balance_loss_clip": 1.05606318,
      "balance_loss_mlp": 1.03528929,
      "epoch": 0.18067037426724786,
      "flos": 19025689265280.0,
      "grad_norm": 2.592432277036083,
      "language_loss": 0.85111535,
      "learning_rate": 3.7665489222115677e-06,
      "loss": 0.87313569,
      "num_input_tokens_seen": 64935215,
      "step": 3005,
      "time_per_iteration": 2.654977560043335
    },
    {
      "auxiliary_loss_clip": 0.0114499,
      "auxiliary_loss_mlp": 0.01041993,
      "balance_loss_clip": 1.05690646,
      "balance_loss_mlp": 1.02489829,
      "epoch": 0.18073049751991582,
      "flos": 27453635976960.0,
      "grad_norm": 1.5217876402754629,
      "language_loss": 0.83215338,
      "learning_rate": 3.766366287157432e-06,
      "loss": 0.85402322,
      "num_input_tokens_seen": 64956275,
      "step": 3006,
      "time_per_iteration": 2.7118306159973145
    },
    {
      "auxiliary_loss_clip": 0.01127168,
      "auxiliary_loss_mlp": 0.01050084,
      "balance_loss_clip": 1.05063033,
      "balance_loss_mlp": 1.03105807,
      "epoch": 0.1807906207725838,
      "flos": 28729039167360.0,
      "grad_norm": 1.6327495611050657,
      "language_loss": 0.77377248,
      "learning_rate": 3.7661835851225103e-06,
      "loss": 0.79554498,
      "num_input_tokens_seen": 64979390,
      "step": 3007,
      "time_per_iteration": 2.7996537685394287
    },
    {
      "auxiliary_loss_clip": 0.01070026,
      "auxiliary_loss_mlp": 0.01030441,
      "balance_loss_clip": 1.04936945,
      "balance_loss_mlp": 1.02712655,
      "epoch": 0.18085074402525175,
      "flos": 64466515468800.0,
      "grad_norm": 0.801982400183398,
      "language_loss": 0.56987137,
      "learning_rate": 3.7660008161137294e-06,
      "loss": 0.5908761,
      "num_input_tokens_seen": 65043135,
      "step": 3008,
      "time_per_iteration": 3.4269092082977295
    },
    {
      "auxiliary_loss_clip": 0.01130838,
      "auxiliary_loss_mlp": 0.01047085,
      "balance_loss_clip": 1.05308366,
      "balance_loss_mlp": 1.02686691,
      "epoch": 0.18091086727791975,
      "flos": 23476960333440.0,
      "grad_norm": 1.8424126412451678,
      "language_loss": 0.67248082,
      "learning_rate": 3.765817980138021e-06,
      "loss": 0.69426012,
      "num_input_tokens_seen": 65062845,
      "step": 3009,
      "time_per_iteration": 2.7875866889953613
    },
    {
      "auxiliary_loss_clip": 0.01161719,
      "auxiliary_loss_mlp": 0.01044187,
      "balance_loss_clip": 1.0595516,
      "balance_loss_mlp": 1.02673507,
      "epoch": 0.1809709905305877,
      "flos": 24170467196160.0,
      "grad_norm": 2.4429360498363986,
      "language_loss": 0.75690198,
      "learning_rate": 3.7656350772023177e-06,
      "loss": 0.778961,
      "num_input_tokens_seen": 65082110,
      "step": 3010,
      "time_per_iteration": 2.6060268878936768
    },
    {
      "auxiliary_loss_clip": 0.01127916,
      "auxiliary_loss_mlp": 0.01037817,
      "balance_loss_clip": 1.05715132,
      "balance_loss_mlp": 1.02063942,
      "epoch": 0.18103111378325568,
      "flos": 21650902669440.0,
      "grad_norm": 1.6324915654296899,
      "language_loss": 0.67356348,
      "learning_rate": 3.7654521073135553e-06,
      "loss": 0.69522083,
      "num_input_tokens_seen": 65101985,
      "step": 3011,
      "time_per_iteration": 2.763596534729004
    },
    {
      "auxiliary_loss_clip": 0.01105034,
      "auxiliary_loss_mlp": 0.00777475,
      "balance_loss_clip": 1.04540467,
      "balance_loss_mlp": 1.00078559,
      "epoch": 0.18109123703592364,
      "flos": 53686918356480.0,
      "grad_norm": 1.551526807882757,
      "language_loss": 0.71288514,
      "learning_rate": 3.7652690704786723e-06,
      "loss": 0.73171026,
      "num_input_tokens_seen": 65129295,
      "step": 3012,
      "time_per_iteration": 3.037775993347168
    },
    {
      "auxiliary_loss_clip": 0.01132189,
      "auxiliary_loss_mlp": 0.01052085,
      "balance_loss_clip": 1.05564284,
      "balance_loss_mlp": 1.03348863,
      "epoch": 0.1811513602885916,
      "flos": 35845564325760.0,
      "grad_norm": 2.095737131475866,
      "language_loss": 0.62309992,
      "learning_rate": 3.765085966704609e-06,
      "loss": 0.64494264,
      "num_input_tokens_seen": 65150625,
      "step": 3013,
      "time_per_iteration": 2.7692227363586426
    },
    {
      "auxiliary_loss_clip": 0.01131323,
      "auxiliary_loss_mlp": 0.0105253,
      "balance_loss_clip": 1.05343401,
      "balance_loss_mlp": 1.03486276,
      "epoch": 0.18121148354125957,
      "flos": 23732572492800.0,
      "grad_norm": 1.6679267545988328,
      "language_loss": 0.76147234,
      "learning_rate": 3.764902795998309e-06,
      "loss": 0.78331089,
      "num_input_tokens_seen": 65170880,
      "step": 3014,
      "time_per_iteration": 2.7296786308288574
    },
    {
      "auxiliary_loss_clip": 0.01163543,
      "auxiliary_loss_mlp": 0.01050053,
      "balance_loss_clip": 1.05964816,
      "balance_loss_mlp": 1.02987087,
      "epoch": 0.18127160679392756,
      "flos": 28728320895360.0,
      "grad_norm": 2.1234423596691796,
      "language_loss": 0.66310829,
      "learning_rate": 3.7647195583667184e-06,
      "loss": 0.6852442,
      "num_input_tokens_seen": 65192530,
      "step": 3015,
      "time_per_iteration": 2.7575571537017822
    },
    {
      "auxiliary_loss_clip": 0.0113004,
      "auxiliary_loss_mlp": 0.00776613,
      "balance_loss_clip": 1.05429327,
      "balance_loss_mlp": 1.00067461,
      "epoch": 0.18133173004659553,
      "flos": 20485062938880.0,
      "grad_norm": 1.7837261279259933,
      "language_loss": 0.78152305,
      "learning_rate": 3.764536253816785e-06,
      "loss": 0.80058956,
      "num_input_tokens_seen": 65211675,
      "step": 3016,
      "time_per_iteration": 2.6718828678131104
    },
    {
      "auxiliary_loss_clip": 0.01145073,
      "auxiliary_loss_mlp": 0.01049504,
      "balance_loss_clip": 1.05684161,
      "balance_loss_mlp": 1.03068125,
      "epoch": 0.1813918532992635,
      "flos": 22852078404480.0,
      "grad_norm": 1.7248072345223011,
      "language_loss": 0.8351965,
      "learning_rate": 3.7643528823554602e-06,
      "loss": 0.85714233,
      "num_input_tokens_seen": 65231185,
      "step": 3017,
      "time_per_iteration": 2.6879045963287354
    },
    {
      "auxiliary_loss_clip": 0.0114091,
      "auxiliary_loss_mlp": 0.01042994,
      "balance_loss_clip": 1.05404854,
      "balance_loss_mlp": 1.02539897,
      "epoch": 0.18145197655193146,
      "flos": 36065122208640.0,
      "grad_norm": 2.2664795482488924,
      "language_loss": 0.6769017,
      "learning_rate": 3.764169443989697e-06,
      "loss": 0.69874066,
      "num_input_tokens_seen": 65251645,
      "step": 3018,
      "time_per_iteration": 4.31333327293396
    },
    {
      "auxiliary_loss_clip": 0.01147629,
      "auxiliary_loss_mlp": 0.00776661,
      "balance_loss_clip": 1.05706179,
      "balance_loss_mlp": 1.00074184,
      "epoch": 0.18151209980459942,
      "flos": 24023951619840.0,
      "grad_norm": 1.8935259017451227,
      "language_loss": 0.76396847,
      "learning_rate": 3.7639859387264518e-06,
      "loss": 0.78321135,
      "num_input_tokens_seen": 65271125,
      "step": 3019,
      "time_per_iteration": 2.7667160034179688
    },
    {
      "auxiliary_loss_clip": 0.01121465,
      "auxiliary_loss_mlp": 0.01046742,
      "balance_loss_clip": 1.05550635,
      "balance_loss_mlp": 1.02722728,
      "epoch": 0.1815722230572674,
      "flos": 23951627585280.0,
      "grad_norm": 2.042490471678265,
      "language_loss": 0.81550395,
      "learning_rate": 3.7638023665726834e-06,
      "loss": 0.83718598,
      "num_input_tokens_seen": 65290600,
      "step": 3020,
      "time_per_iteration": 4.3900346755981445
    },
    {
      "auxiliary_loss_clip": 0.01136424,
      "auxiliary_loss_mlp": 0.01046217,
      "balance_loss_clip": 1.05758023,
      "balance_loss_mlp": 1.02567708,
      "epoch": 0.18163234630993536,
      "flos": 24386469632640.0,
      "grad_norm": 1.9628186536024828,
      "language_loss": 0.7757082,
      "learning_rate": 3.763618727535352e-06,
      "loss": 0.79753458,
      "num_input_tokens_seen": 65311040,
      "step": 3021,
      "time_per_iteration": 4.3029396533966064
    },
    {
      "auxiliary_loss_clip": 0.01143245,
      "auxiliary_loss_mlp": 0.01047278,
      "balance_loss_clip": 1.05453348,
      "balance_loss_mlp": 1.02907431,
      "epoch": 0.18169246956260335,
      "flos": 24681332378880.0,
      "grad_norm": 1.725306643191844,
      "language_loss": 0.84863859,
      "learning_rate": 3.763435021621422e-06,
      "loss": 0.87054378,
      "num_input_tokens_seen": 65332115,
      "step": 3022,
      "time_per_iteration": 2.7353312969207764
    },
    {
      "auxiliary_loss_clip": 0.01132435,
      "auxiliary_loss_mlp": 0.01042747,
      "balance_loss_clip": 1.05769348,
      "balance_loss_mlp": 1.0235188,
      "epoch": 0.1817525928152713,
      "flos": 24243294021120.0,
      "grad_norm": 2.230341519134859,
      "language_loss": 0.69367266,
      "learning_rate": 3.763251248837859e-06,
      "loss": 0.71542448,
      "num_input_tokens_seen": 65352210,
      "step": 3023,
      "time_per_iteration": 2.775200605392456
    },
    {
      "auxiliary_loss_clip": 0.01127605,
      "auxiliary_loss_mlp": 0.01043947,
      "balance_loss_clip": 1.04900002,
      "balance_loss_mlp": 1.02556491,
      "epoch": 0.18181271606793928,
      "flos": 16472081623680.0,
      "grad_norm": 2.150764188548567,
      "language_loss": 0.74107385,
      "learning_rate": 3.7630674091916317e-06,
      "loss": 0.76278937,
      "num_input_tokens_seen": 65370600,
      "step": 3024,
      "time_per_iteration": 2.7364041805267334
    },
    {
      "auxiliary_loss_clip": 0.01145205,
      "auxiliary_loss_mlp": 0.01046837,
      "balance_loss_clip": 1.05719447,
      "balance_loss_mlp": 1.02900314,
      "epoch": 0.18187283932060724,
      "flos": 18581042805120.0,
      "grad_norm": 2.148591016046099,
      "language_loss": 0.8835662,
      "learning_rate": 3.7628835026897123e-06,
      "loss": 0.90548658,
      "num_input_tokens_seen": 65387270,
      "step": 3025,
      "time_per_iteration": 4.274658679962158
    },
    {
      "auxiliary_loss_clip": 0.01133667,
      "auxiliary_loss_mlp": 0.01050575,
      "balance_loss_clip": 1.05470932,
      "balance_loss_mlp": 1.03137028,
      "epoch": 0.1819329625732752,
      "flos": 20266833859200.0,
      "grad_norm": 3.6399614210311206,
      "language_loss": 0.79041791,
      "learning_rate": 3.7626995293390735e-06,
      "loss": 0.81226033,
      "num_input_tokens_seen": 65406550,
      "step": 3026,
      "time_per_iteration": 2.7589778900146484
    },
    {
      "auxiliary_loss_clip": 0.01132736,
      "auxiliary_loss_mlp": 0.01055367,
      "balance_loss_clip": 1.05774415,
      "balance_loss_mlp": 1.03679442,
      "epoch": 0.18199308582594317,
      "flos": 25915186512000.0,
      "grad_norm": 1.6980721374313217,
      "language_loss": 0.759978,
      "learning_rate": 3.762515489146692e-06,
      "loss": 0.78185904,
      "num_input_tokens_seen": 65425955,
      "step": 3027,
      "time_per_iteration": 2.7347826957702637
    },
    {
      "auxiliary_loss_clip": 0.01163558,
      "auxiliary_loss_mlp": 0.01053369,
      "balance_loss_clip": 1.05835891,
      "balance_loss_mlp": 1.03378284,
      "epoch": 0.18205320907861114,
      "flos": 15377524433280.0,
      "grad_norm": 2.2893837743041368,
      "language_loss": 0.85592651,
      "learning_rate": 3.762331382119546e-06,
      "loss": 0.87809575,
      "num_input_tokens_seen": 65442820,
      "step": 3028,
      "time_per_iteration": 2.598905563354492
    },
    {
      "auxiliary_loss_clip": 0.01156921,
      "auxiliary_loss_mlp": 0.0104449,
      "balance_loss_clip": 1.0578618,
      "balance_loss_mlp": 1.0260129,
      "epoch": 0.18211333233127913,
      "flos": 25624310175360.0,
      "grad_norm": 1.8897570500397638,
      "language_loss": 0.82807779,
      "learning_rate": 3.7621472082646183e-06,
      "loss": 0.85009193,
      "num_input_tokens_seen": 65461825,
      "step": 3029,
      "time_per_iteration": 2.677332639694214
    },
    {
      "auxiliary_loss_clip": 0.01114993,
      "auxiliary_loss_mlp": 0.01050232,
      "balance_loss_clip": 1.05223596,
      "balance_loss_mlp": 1.02931094,
      "epoch": 0.1821734555839471,
      "flos": 14976007228800.0,
      "grad_norm": 10.840079090220346,
      "language_loss": 0.78091359,
      "learning_rate": 3.761962967588891e-06,
      "loss": 0.80256593,
      "num_input_tokens_seen": 65479480,
      "step": 3030,
      "time_per_iteration": 2.6865499019622803
    },
    {
      "auxiliary_loss_clip": 0.01139676,
      "auxiliary_loss_mlp": 0.01043273,
      "balance_loss_clip": 1.05401075,
      "balance_loss_mlp": 1.0240562,
      "epoch": 0.18223357883661506,
      "flos": 20194007034240.0,
      "grad_norm": 2.05958060196279,
      "language_loss": 0.85162055,
      "learning_rate": 3.761778660099352e-06,
      "loss": 0.87345004,
      "num_input_tokens_seen": 65497775,
      "step": 3031,
      "time_per_iteration": 2.6336488723754883
    },
    {
      "auxiliary_loss_clip": 0.01116657,
      "auxiliary_loss_mlp": 0.00776186,
      "balance_loss_clip": 1.0497843,
      "balance_loss_mlp": 1.00052071,
      "epoch": 0.18229370208928303,
      "flos": 15231978524160.0,
      "grad_norm": 1.83501853384953,
      "language_loss": 0.79992211,
      "learning_rate": 3.76159428580299e-06,
      "loss": 0.81885058,
      "num_input_tokens_seen": 65516505,
      "step": 3032,
      "time_per_iteration": 2.6879780292510986
    },
    {
      "auxiliary_loss_clip": 0.01166412,
      "auxiliary_loss_mlp": 0.01048902,
      "balance_loss_clip": 1.06163025,
      "balance_loss_mlp": 1.03038836,
      "epoch": 0.182353825341951,
      "flos": 23840483927040.0,
      "grad_norm": 1.8132660189598853,
      "language_loss": 0.81316388,
      "learning_rate": 3.761409844706795e-06,
      "loss": 0.83531702,
      "num_input_tokens_seen": 65536160,
      "step": 3033,
      "time_per_iteration": 2.628100872039795
    },
    {
      "auxiliary_loss_clip": 0.01048591,
      "auxiliary_loss_mlp": 0.0100128,
      "balance_loss_clip": 1.05392861,
      "balance_loss_mlp": 0.99850291,
      "epoch": 0.18241394859461896,
      "flos": 61190957393280.0,
      "grad_norm": 0.8825814513625035,
      "language_loss": 0.63439631,
      "learning_rate": 3.7612253368177625e-06,
      "loss": 0.65489495,
      "num_input_tokens_seen": 65589375,
      "step": 3034,
      "time_per_iteration": 3.2329187393188477
    },
    {
      "auxiliary_loss_clip": 0.0112853,
      "auxiliary_loss_mlp": 0.01041043,
      "balance_loss_clip": 1.05698252,
      "balance_loss_mlp": 1.02384114,
      "epoch": 0.18247407184728695,
      "flos": 18471694826880.0,
      "grad_norm": 3.107937736318082,
      "language_loss": 0.79893476,
      "learning_rate": 3.7610407621428893e-06,
      "loss": 0.82063049,
      "num_input_tokens_seen": 65606720,
      "step": 3035,
      "time_per_iteration": 2.7644357681274414
    },
    {
      "auxiliary_loss_clip": 0.01134115,
      "auxiliary_loss_mlp": 0.01046396,
      "balance_loss_clip": 1.05675578,
      "balance_loss_mlp": 1.02906322,
      "epoch": 0.18253419509995492,
      "flos": 21795191602560.0,
      "grad_norm": 1.870086430131469,
      "language_loss": 0.85076666,
      "learning_rate": 3.7608561206891735e-06,
      "loss": 0.87257177,
      "num_input_tokens_seen": 65625495,
      "step": 3036,
      "time_per_iteration": 2.7102303504943848
    },
    {
      "auxiliary_loss_clip": 0.01140083,
      "auxiliary_loss_mlp": 0.01039078,
      "balance_loss_clip": 1.05572963,
      "balance_loss_mlp": 1.02192414,
      "epoch": 0.18259431835262288,
      "flos": 20149764456960.0,
      "grad_norm": 2.1821496235124727,
      "language_loss": 0.80254716,
      "learning_rate": 3.760671412463617e-06,
      "loss": 0.82433879,
      "num_input_tokens_seen": 65643515,
      "step": 3037,
      "time_per_iteration": 2.6703832149505615
    },
    {
      "auxiliary_loss_clip": 0.01139652,
      "auxiliary_loss_mlp": 0.00776941,
      "balance_loss_clip": 1.05986989,
      "balance_loss_mlp": 1.00062871,
      "epoch": 0.18265444160529085,
      "flos": 16981653916800.0,
      "grad_norm": 3.0764011293768023,
      "language_loss": 0.7950514,
      "learning_rate": 3.7604866374732246e-06,
      "loss": 0.81421733,
      "num_input_tokens_seen": 65658155,
      "step": 3038,
      "time_per_iteration": 2.7410895824432373
    },
    {
      "auxiliary_loss_clip": 0.01125628,
      "auxiliary_loss_mlp": 0.01044597,
      "balance_loss_clip": 1.05254972,
      "balance_loss_mlp": 1.02551126,
      "epoch": 0.1827145648579588,
      "flos": 34423250509440.0,
      "grad_norm": 1.9524772610579864,
      "language_loss": 0.67722493,
      "learning_rate": 3.7603017957250023e-06,
      "loss": 0.69892722,
      "num_input_tokens_seen": 65679310,
      "step": 3039,
      "time_per_iteration": 2.756833076477051
    },
    {
      "auxiliary_loss_clip": 0.0113051,
      "auxiliary_loss_mlp": 0.01051065,
      "balance_loss_clip": 1.053087,
      "balance_loss_mlp": 1.03304029,
      "epoch": 0.18277468811062678,
      "flos": 53287017264000.0,
      "grad_norm": 1.8757227718998248,
      "language_loss": 0.73394251,
      "learning_rate": 3.7601168872259593e-06,
      "loss": 0.75575823,
      "num_input_tokens_seen": 65705235,
      "step": 3040,
      "time_per_iteration": 3.026679039001465
    },
    {
      "auxiliary_loss_clip": 0.01143558,
      "auxiliary_loss_mlp": 0.01042261,
      "balance_loss_clip": 1.05585194,
      "balance_loss_mlp": 1.02373624,
      "epoch": 0.18283481136329474,
      "flos": 31650659602560.0,
      "grad_norm": 2.017308993436446,
      "language_loss": 0.60348576,
      "learning_rate": 3.7599319119831075e-06,
      "loss": 0.62534392,
      "num_input_tokens_seen": 65727575,
      "step": 3041,
      "time_per_iteration": 2.738554000854492
    },
    {
      "auxiliary_loss_clip": 0.01116972,
      "auxiliary_loss_mlp": 0.01053827,
      "balance_loss_clip": 1.05058599,
      "balance_loss_mlp": 1.03544497,
      "epoch": 0.18289493461596273,
      "flos": 53137664513280.0,
      "grad_norm": 2.3558133433802104,
      "language_loss": 0.59825706,
      "learning_rate": 3.7597468700034616e-06,
      "loss": 0.61996508,
      "num_input_tokens_seen": 65751370,
      "step": 3042,
      "time_per_iteration": 3.0009193420410156
    },
    {
      "auxiliary_loss_clip": 0.0112422,
      "auxiliary_loss_mlp": 0.01046569,
      "balance_loss_clip": 1.05319464,
      "balance_loss_mlp": 1.02917695,
      "epoch": 0.1829550578686307,
      "flos": 25589369220480.0,
      "grad_norm": 1.5313119565207096,
      "language_loss": 0.8757726,
      "learning_rate": 3.7595617612940374e-06,
      "loss": 0.89748049,
      "num_input_tokens_seen": 65771040,
      "step": 3043,
      "time_per_iteration": 2.7406487464904785
    },
    {
      "auxiliary_loss_clip": 0.01056788,
      "auxiliary_loss_mlp": 0.01056357,
      "balance_loss_clip": 1.04592645,
      "balance_loss_mlp": 1.03712869,
      "epoch": 0.18301518112129866,
      "flos": 22601422321920.0,
      "grad_norm": 2.144378235575635,
      "language_loss": 0.70980251,
      "learning_rate": 3.7593765858618552e-06,
      "loss": 0.73093396,
      "num_input_tokens_seen": 65789345,
      "step": 3044,
      "time_per_iteration": 2.785931348800659
    },
    {
      "auxiliary_loss_clip": 0.01105073,
      "auxiliary_loss_mlp": 0.01059118,
      "balance_loss_clip": 1.05111921,
      "balance_loss_mlp": 1.0381608,
      "epoch": 0.18307530437396663,
      "flos": 34020799551360.0,
      "grad_norm": 3.097061979225562,
      "language_loss": 0.64460731,
      "learning_rate": 3.7591913437139365e-06,
      "loss": 0.66624922,
      "num_input_tokens_seen": 65810990,
      "step": 3045,
      "time_per_iteration": 2.8085720539093018
    },
    {
      "auxiliary_loss_clip": 0.01155246,
      "auxiliary_loss_mlp": 0.01044973,
      "balance_loss_clip": 1.05604315,
      "balance_loss_mlp": 1.02780676,
      "epoch": 0.1831354276266346,
      "flos": 21279765392640.0,
      "grad_norm": 11.455833434854163,
      "language_loss": 0.78461385,
      "learning_rate": 3.7590060348573066e-06,
      "loss": 0.80661607,
      "num_input_tokens_seen": 65827230,
      "step": 3046,
      "time_per_iteration": 2.603299140930176
    },
    {
      "auxiliary_loss_clip": 0.01118725,
      "auxiliary_loss_mlp": 0.01042864,
      "balance_loss_clip": 1.04837,
      "balance_loss_mlp": 1.0240643,
      "epoch": 0.18319555087930256,
      "flos": 21032952065280.0,
      "grad_norm": 1.9889932097770582,
      "language_loss": 0.78733194,
      "learning_rate": 3.7588206592989903e-06,
      "loss": 0.8089478,
      "num_input_tokens_seen": 65845900,
      "step": 3047,
      "time_per_iteration": 2.7109453678131104
    },
    {
      "auxiliary_loss_clip": 0.01144516,
      "auxiliary_loss_mlp": 0.01042422,
      "balance_loss_clip": 1.05723858,
      "balance_loss_mlp": 1.0254705,
      "epoch": 0.18325567413197055,
      "flos": 34382958428160.0,
      "grad_norm": 1.5191744259185578,
      "language_loss": 0.80704039,
      "learning_rate": 3.7586352170460194e-06,
      "loss": 0.82890975,
      "num_input_tokens_seen": 65868730,
      "step": 3048,
      "time_per_iteration": 2.7485053539276123
    },
    {
      "auxiliary_loss_clip": 0.01139433,
      "auxiliary_loss_mlp": 0.01046004,
      "balance_loss_clip": 1.05405188,
      "balance_loss_mlp": 1.02552414,
      "epoch": 0.18331579738463852,
      "flos": 20558464381440.0,
      "grad_norm": 2.1437824577601354,
      "language_loss": 0.86579728,
      "learning_rate": 3.758449708105424e-06,
      "loss": 0.88765168,
      "num_input_tokens_seen": 65888420,
      "step": 3049,
      "time_per_iteration": 2.6876962184906006
    },
    {
      "auxiliary_loss_clip": 0.01143881,
      "auxiliary_loss_mlp": 0.01045208,
      "balance_loss_clip": 1.05379057,
      "balance_loss_mlp": 1.02544308,
      "epoch": 0.18337592063730648,
      "flos": 19607872901760.0,
      "grad_norm": 2.616661567020713,
      "language_loss": 0.77827966,
      "learning_rate": 3.75826413248424e-06,
      "loss": 0.80017054,
      "num_input_tokens_seen": 65905840,
      "step": 3050,
      "time_per_iteration": 2.5814058780670166
    },
    {
      "auxiliary_loss_clip": 0.01126116,
      "auxiliary_loss_mlp": 0.01041302,
      "balance_loss_clip": 1.04954183,
      "balance_loss_mlp": 1.0238502,
      "epoch": 0.18343604388997445,
      "flos": 20850885002880.0,
      "grad_norm": 2.3686375880611656,
      "language_loss": 0.99064422,
      "learning_rate": 3.7580784901895035e-06,
      "loss": 1.01231837,
      "num_input_tokens_seen": 65922845,
      "step": 3051,
      "time_per_iteration": 2.701848268508911
    },
    {
      "auxiliary_loss_clip": 0.01125492,
      "auxiliary_loss_mlp": 0.010397,
      "balance_loss_clip": 1.05189931,
      "balance_loss_mlp": 1.02078128,
      "epoch": 0.1834961671426424,
      "flos": 24394370624640.0,
      "grad_norm": 2.0338529701436237,
      "language_loss": 0.8607648,
      "learning_rate": 3.7578927812282542e-06,
      "loss": 0.88241673,
      "num_input_tokens_seen": 65945555,
      "step": 3052,
      "time_per_iteration": 2.7252042293548584
    },
    {
      "auxiliary_loss_clip": 0.01152967,
      "auxiliary_loss_mlp": 0.01044648,
      "balance_loss_clip": 1.05449986,
      "balance_loss_mlp": 1.02737474,
      "epoch": 0.18355629039531038,
      "flos": 21251612108160.0,
      "grad_norm": 1.8649432496703628,
      "language_loss": 0.73393309,
      "learning_rate": 3.7577070056075356e-06,
      "loss": 0.7559092,
      "num_input_tokens_seen": 65963965,
      "step": 3053,
      "time_per_iteration": 2.6331369876861572
    },
    {
      "auxiliary_loss_clip": 0.01158728,
      "auxiliary_loss_mlp": 0.01044052,
      "balance_loss_clip": 1.05783379,
      "balance_loss_mlp": 1.02565801,
      "epoch": 0.18361641364797834,
      "flos": 28656499651200.0,
      "grad_norm": 1.5358769917973574,
      "language_loss": 0.61891186,
      "learning_rate": 3.7575211633343902e-06,
      "loss": 0.64093965,
      "num_input_tokens_seen": 65985965,
      "step": 3054,
      "time_per_iteration": 2.6792421340942383
    },
    {
      "auxiliary_loss_clip": 0.01108826,
      "auxiliary_loss_mlp": 0.01042654,
      "balance_loss_clip": 1.05558836,
      "balance_loss_mlp": 1.02502322,
      "epoch": 0.18367653690064634,
      "flos": 20918827578240.0,
      "grad_norm": 2.2474279661883667,
      "language_loss": 0.78218341,
      "learning_rate": 3.7573352544158663e-06,
      "loss": 0.80369824,
      "num_input_tokens_seen": 66005645,
      "step": 3055,
      "time_per_iteration": 2.778691053390503
    },
    {
      "auxiliary_loss_clip": 0.01096638,
      "auxiliary_loss_mlp": 0.01050677,
      "balance_loss_clip": 1.05003095,
      "balance_loss_mlp": 1.03211594,
      "epoch": 0.1837366601533143,
      "flos": 28765596234240.0,
      "grad_norm": 1.8043720478204575,
      "language_loss": 0.7022509,
      "learning_rate": 3.757149278859014e-06,
      "loss": 0.72372401,
      "num_input_tokens_seen": 66025675,
      "step": 3056,
      "time_per_iteration": 2.794254779815674
    },
    {
      "auxiliary_loss_clip": 0.01140367,
      "auxiliary_loss_mlp": 0.01038358,
      "balance_loss_clip": 1.05211461,
      "balance_loss_mlp": 1.02181149,
      "epoch": 0.18379678340598227,
      "flos": 21251432540160.0,
      "grad_norm": 1.8709784760841586,
      "language_loss": 0.80357504,
      "learning_rate": 3.7569632366708842e-06,
      "loss": 0.82536227,
      "num_input_tokens_seen": 66046125,
      "step": 3057,
      "time_per_iteration": 2.644728899002075
    },
    {
      "auxiliary_loss_clip": 0.01150041,
      "auxiliary_loss_mlp": 0.01043781,
      "balance_loss_clip": 1.05482352,
      "balance_loss_mlp": 1.02332497,
      "epoch": 0.18385690665865023,
      "flos": 20449619193600.0,
      "grad_norm": 7.225766788646501,
      "language_loss": 0.82570755,
      "learning_rate": 3.756777127858533e-06,
      "loss": 0.84764576,
      "num_input_tokens_seen": 66064375,
      "step": 3058,
      "time_per_iteration": 4.136845588684082
    },
    {
      "auxiliary_loss_clip": 0.01119139,
      "auxiliary_loss_mlp": 0.00776668,
      "balance_loss_clip": 1.04992914,
      "balance_loss_mlp": 1.00066566,
      "epoch": 0.1839170299113182,
      "flos": 26140562398080.0,
      "grad_norm": 2.277694088171661,
      "language_loss": 0.85071868,
      "learning_rate": 3.756590952429017e-06,
      "loss": 0.86967677,
      "num_input_tokens_seen": 66084590,
      "step": 3059,
      "time_per_iteration": 2.745020866394043
    },
    {
      "auxiliary_loss_clip": 0.01151831,
      "auxiliary_loss_mlp": 0.00775088,
      "balance_loss_clip": 1.05359423,
      "balance_loss_mlp": 1.00077426,
      "epoch": 0.18397715316398616,
      "flos": 31758032332800.0,
      "grad_norm": 2.3540516696336216,
      "language_loss": 0.72983348,
      "learning_rate": 3.756404710389396e-06,
      "loss": 0.74910271,
      "num_input_tokens_seen": 66107105,
      "step": 3060,
      "time_per_iteration": 5.792214393615723
    },
    {
      "auxiliary_loss_clip": 0.01149482,
      "auxiliary_loss_mlp": 0.01041417,
      "balance_loss_clip": 1.05812132,
      "balance_loss_mlp": 1.02266574,
      "epoch": 0.18403727641665413,
      "flos": 24611989173120.0,
      "grad_norm": 1.5810457302838978,
      "language_loss": 0.73126459,
      "learning_rate": 3.7562184017467323e-06,
      "loss": 0.75317359,
      "num_input_tokens_seen": 66129295,
      "step": 3061,
      "time_per_iteration": 2.754167318344116
    },
    {
      "auxiliary_loss_clip": 0.01138281,
      "auxiliary_loss_mlp": 0.01043599,
      "balance_loss_clip": 1.05435956,
      "balance_loss_mlp": 1.02379823,
      "epoch": 0.18409739966932212,
      "flos": 23439900476160.0,
      "grad_norm": 1.8413104246803462,
      "language_loss": 0.81937188,
      "learning_rate": 3.7560320265080906e-06,
      "loss": 0.8411907,
      "num_input_tokens_seen": 66146910,
      "step": 3062,
      "time_per_iteration": 2.7545394897460938
    },
    {
      "auxiliary_loss_clip": 0.01144664,
      "auxiliary_loss_mlp": 0.01040639,
      "balance_loss_clip": 1.05668104,
      "balance_loss_mlp": 1.02259111,
      "epoch": 0.18415752292199009,
      "flos": 21872112577920.0,
      "grad_norm": 2.011374259171591,
      "language_loss": 0.72994816,
      "learning_rate": 3.7558455846805383e-06,
      "loss": 0.75180125,
      "num_input_tokens_seen": 66165370,
      "step": 3063,
      "time_per_iteration": 2.738293170928955
    },
    {
      "auxiliary_loss_clip": 0.01133824,
      "auxiliary_loss_mlp": 0.01040987,
      "balance_loss_clip": 1.05164194,
      "balance_loss_mlp": 1.02490544,
      "epoch": 0.18421764617465805,
      "flos": 25410678036480.0,
      "grad_norm": 2.2975785147287953,
      "language_loss": 0.65614092,
      "learning_rate": 3.7556590762711463e-06,
      "loss": 0.67788899,
      "num_input_tokens_seen": 66186210,
      "step": 3064,
      "time_per_iteration": 4.404583930969238
    },
    {
      "auxiliary_loss_clip": 0.01141547,
      "auxiliary_loss_mlp": 0.01042996,
      "balance_loss_clip": 1.05395937,
      "balance_loss_mlp": 1.02498376,
      "epoch": 0.18427776942732602,
      "flos": 27198131558400.0,
      "grad_norm": 2.1874829734431898,
      "language_loss": 0.68347883,
      "learning_rate": 3.7554725012869853e-06,
      "loss": 0.70532429,
      "num_input_tokens_seen": 66204800,
      "step": 3065,
      "time_per_iteration": 2.7149577140808105
    },
    {
      "auxiliary_loss_clip": 0.01136969,
      "auxiliary_loss_mlp": 0.01045319,
      "balance_loss_clip": 1.05518305,
      "balance_loss_mlp": 1.02674615,
      "epoch": 0.18433789267999398,
      "flos": 27852351920640.0,
      "grad_norm": 2.2758854533642925,
      "language_loss": 0.73142231,
      "learning_rate": 3.7552858597351318e-06,
      "loss": 0.75324523,
      "num_input_tokens_seen": 66222195,
      "step": 3066,
      "time_per_iteration": 2.672675609588623
    },
    {
      "auxiliary_loss_clip": 0.01125186,
      "auxiliary_loss_mlp": 0.01043389,
      "balance_loss_clip": 1.04947495,
      "balance_loss_mlp": 1.0256983,
      "epoch": 0.18439801593266195,
      "flos": 17856940533120.0,
      "grad_norm": 2.1067167513095444,
      "language_loss": 0.82191038,
      "learning_rate": 3.7550991516226622e-06,
      "loss": 0.8435961,
      "num_input_tokens_seen": 66239505,
      "step": 3067,
      "time_per_iteration": 2.697768211364746
    },
    {
      "auxiliary_loss_clip": 0.01082345,
      "auxiliary_loss_mlp": 0.00756782,
      "balance_loss_clip": 1.04466891,
      "balance_loss_mlp": 1.00113225,
      "epoch": 0.18445813918532994,
      "flos": 56389522590720.0,
      "grad_norm": 0.7960107429271657,
      "language_loss": 0.59750569,
      "learning_rate": 3.754912376956657e-06,
      "loss": 0.61589694,
      "num_input_tokens_seen": 66295695,
      "step": 3068,
      "time_per_iteration": 3.0305213928222656
    },
    {
      "auxiliary_loss_clip": 0.01127048,
      "auxiliary_loss_mlp": 0.01041294,
      "balance_loss_clip": 1.05452299,
      "balance_loss_mlp": 1.02356791,
      "epoch": 0.1845182624379979,
      "flos": 20957180325120.0,
      "grad_norm": 3.7299324256794244,
      "language_loss": 0.76434112,
      "learning_rate": 3.7547255357441987e-06,
      "loss": 0.78602457,
      "num_input_tokens_seen": 66315315,
      "step": 3069,
      "time_per_iteration": 2.6757962703704834
    },
    {
      "auxiliary_loss_clip": 0.01146412,
      "auxiliary_loss_mlp": 0.010456,
      "balance_loss_clip": 1.05468106,
      "balance_loss_mlp": 1.02798057,
      "epoch": 0.18457838569066587,
      "flos": 20485170679680.0,
      "grad_norm": 1.9225240149566294,
      "language_loss": 0.8491416,
      "learning_rate": 3.7545386279923718e-06,
      "loss": 0.87106168,
      "num_input_tokens_seen": 66333675,
      "step": 3070,
      "time_per_iteration": 2.617023229598999
    },
    {
      "auxiliary_loss_clip": 0.01127789,
      "auxiliary_loss_mlp": 0.01043452,
      "balance_loss_clip": 1.0553112,
      "balance_loss_mlp": 1.02510571,
      "epoch": 0.18463850894333383,
      "flos": 25010022758400.0,
      "grad_norm": 6.700503585098448,
      "language_loss": 0.77807182,
      "learning_rate": 3.754351653708265e-06,
      "loss": 0.79978424,
      "num_input_tokens_seen": 66354075,
      "step": 3071,
      "time_per_iteration": 2.847329616546631
    },
    {
      "auxiliary_loss_clip": 0.01109458,
      "auxiliary_loss_mlp": 0.01049978,
      "balance_loss_clip": 1.05054557,
      "balance_loss_mlp": 1.03154778,
      "epoch": 0.1846986321960018,
      "flos": 16800628348800.0,
      "grad_norm": 2.0836336776071565,
      "language_loss": 0.77414191,
      "learning_rate": 3.7541646128989674e-06,
      "loss": 0.79573631,
      "num_input_tokens_seen": 66372520,
      "step": 3072,
      "time_per_iteration": 2.780921220779419
    },
    {
      "auxiliary_loss_clip": 0.01138997,
      "auxiliary_loss_mlp": 0.01043594,
      "balance_loss_clip": 1.05106127,
      "balance_loss_mlp": 1.02465141,
      "epoch": 0.18475875544866976,
      "flos": 20814327936000.0,
      "grad_norm": 4.959080593148226,
      "language_loss": 0.86546457,
      "learning_rate": 3.7539775055715715e-06,
      "loss": 0.88729048,
      "num_input_tokens_seen": 66390745,
      "step": 3073,
      "time_per_iteration": 2.631913661956787
    },
    {
      "auxiliary_loss_clip": 0.01158717,
      "auxiliary_loss_mlp": 0.0104013,
      "balance_loss_clip": 1.05862749,
      "balance_loss_mlp": 1.02366686,
      "epoch": 0.18481887870133773,
      "flos": 22601422321920.0,
      "grad_norm": 2.162700927804164,
      "language_loss": 0.91831195,
      "learning_rate": 3.7537903317331732e-06,
      "loss": 0.94030046,
      "num_input_tokens_seen": 66410525,
      "step": 3074,
      "time_per_iteration": 2.6152567863464355
    },
    {
      "auxiliary_loss_clip": 0.01104968,
      "auxiliary_loss_mlp": 0.01047718,
      "balance_loss_clip": 1.04757643,
      "balance_loss_mlp": 1.02763104,
      "epoch": 0.18487900195400572,
      "flos": 29458815788160.0,
      "grad_norm": 1.9967983521568784,
      "language_loss": 0.64783108,
      "learning_rate": 3.75360309139087e-06,
      "loss": 0.66935796,
      "num_input_tokens_seen": 66432535,
      "step": 3075,
      "time_per_iteration": 2.763559103012085
    },
    {
      "auxiliary_loss_clip": 0.01135247,
      "auxiliary_loss_mlp": 0.01046601,
      "balance_loss_clip": 1.05689573,
      "balance_loss_mlp": 1.02913702,
      "epoch": 0.1849391252066737,
      "flos": 20628777254400.0,
      "grad_norm": 1.8996898495981898,
      "language_loss": 0.72803432,
      "learning_rate": 3.753415784551761e-06,
      "loss": 0.74985278,
      "num_input_tokens_seen": 66450620,
      "step": 3076,
      "time_per_iteration": 2.76629376411438
    },
    {
      "auxiliary_loss_clip": 0.01124833,
      "auxiliary_loss_mlp": 0.01042344,
      "balance_loss_clip": 1.0584389,
      "balance_loss_mlp": 1.0249157,
      "epoch": 0.18499924845934165,
      "flos": 14428549065600.0,
      "grad_norm": 2.4862024108169556,
      "language_loss": 0.80772626,
      "learning_rate": 3.7532284112229507e-06,
      "loss": 0.82939804,
      "num_input_tokens_seen": 66467865,
      "step": 3077,
      "time_per_iteration": 2.7296142578125
    },
    {
      "auxiliary_loss_clip": 0.01128471,
      "auxiliary_loss_mlp": 0.01041495,
      "balance_loss_clip": 1.05401397,
      "balance_loss_mlp": 1.02428079,
      "epoch": 0.18505937171200962,
      "flos": 23727652329600.0,
      "grad_norm": 1.8214336253769514,
      "language_loss": 0.78693211,
      "learning_rate": 3.7530409714115424e-06,
      "loss": 0.80863178,
      "num_input_tokens_seen": 66486245,
      "step": 3078,
      "time_per_iteration": 2.715838670730591
    },
    {
      "auxiliary_loss_clip": 0.01154963,
      "auxiliary_loss_mlp": 0.01043373,
      "balance_loss_clip": 1.05546641,
      "balance_loss_mlp": 1.02655268,
      "epoch": 0.18511949496467758,
      "flos": 25957489754880.0,
      "grad_norm": 1.7455066055145632,
      "language_loss": 0.77326959,
      "learning_rate": 3.7528534651246453e-06,
      "loss": 0.79525292,
      "num_input_tokens_seen": 66506510,
      "step": 3079,
      "time_per_iteration": 2.674128770828247
    },
    {
      "auxiliary_loss_clip": 0.01119079,
      "auxiliary_loss_mlp": 0.01041512,
      "balance_loss_clip": 1.04717147,
      "balance_loss_mlp": 1.02328515,
      "epoch": 0.18517961821734555,
      "flos": 42413553912960.0,
      "grad_norm": 1.885086933557342,
      "language_loss": 0.82143807,
      "learning_rate": 3.752665892369369e-06,
      "loss": 0.84304404,
      "num_input_tokens_seen": 66530960,
      "step": 3080,
      "time_per_iteration": 2.906940460205078
    },
    {
      "auxiliary_loss_clip": 0.01123637,
      "auxiliary_loss_mlp": 0.01044031,
      "balance_loss_clip": 1.05894399,
      "balance_loss_mlp": 1.02563691,
      "epoch": 0.18523974147001354,
      "flos": 24097568544000.0,
      "grad_norm": 2.065822240576764,
      "language_loss": 0.73973286,
      "learning_rate": 3.7524782531528266e-06,
      "loss": 0.76140958,
      "num_input_tokens_seen": 66550275,
      "step": 3081,
      "time_per_iteration": 2.7960739135742188
    },
    {
      "auxiliary_loss_clip": 0.01126977,
      "auxiliary_loss_mlp": 0.01051674,
      "balance_loss_clip": 1.05360913,
      "balance_loss_mlp": 1.03286242,
      "epoch": 0.1852998647226815,
      "flos": 27375278457600.0,
      "grad_norm": 1.9854893879184425,
      "language_loss": 0.71991849,
      "learning_rate": 3.7522905474821334e-06,
      "loss": 0.74170506,
      "num_input_tokens_seen": 66569040,
      "step": 3082,
      "time_per_iteration": 2.6965079307556152
    },
    {
      "auxiliary_loss_clip": 0.01124933,
      "auxiliary_loss_mlp": 0.01046296,
      "balance_loss_clip": 1.05649543,
      "balance_loss_mlp": 1.02694798,
      "epoch": 0.18535998797534947,
      "flos": 18332757020160.0,
      "grad_norm": 2.0424653419479886,
      "language_loss": 0.69580144,
      "learning_rate": 3.752102775364407e-06,
      "loss": 0.71751374,
      "num_input_tokens_seen": 66587775,
      "step": 3083,
      "time_per_iteration": 2.727252721786499
    },
    {
      "auxiliary_loss_clip": 0.01122388,
      "auxiliary_loss_mlp": 0.01046999,
      "balance_loss_clip": 1.05204451,
      "balance_loss_mlp": 1.02964258,
      "epoch": 0.18542011122801744,
      "flos": 37845859887360.0,
      "grad_norm": 2.185713468975319,
      "language_loss": 0.68965334,
      "learning_rate": 3.751914936806767e-06,
      "loss": 0.71134722,
      "num_input_tokens_seen": 66610800,
      "step": 3084,
      "time_per_iteration": 2.95849871635437
    },
    {
      "auxiliary_loss_clip": 0.01155184,
      "auxiliary_loss_mlp": 0.01043029,
      "balance_loss_clip": 1.05578482,
      "balance_loss_mlp": 1.0257436,
      "epoch": 0.1854802344806854,
      "flos": 25186128163200.0,
      "grad_norm": 1.6859724806626923,
      "language_loss": 0.77390355,
      "learning_rate": 3.7517270318163377e-06,
      "loss": 0.79588568,
      "num_input_tokens_seen": 66630960,
      "step": 3085,
      "time_per_iteration": 2.68961501121521
    },
    {
      "auxiliary_loss_clip": 0.01152089,
      "auxiliary_loss_mlp": 0.01049004,
      "balance_loss_clip": 1.05316019,
      "balance_loss_mlp": 1.03142118,
      "epoch": 0.18554035773335337,
      "flos": 26684788337280.0,
      "grad_norm": 1.993169596996871,
      "language_loss": 0.73752379,
      "learning_rate": 3.751539060400244e-06,
      "loss": 0.75953472,
      "num_input_tokens_seen": 66650585,
      "step": 3086,
      "time_per_iteration": 2.652475595474243
    },
    {
      "auxiliary_loss_clip": 0.01142754,
      "auxiliary_loss_mlp": 0.01049865,
      "balance_loss_clip": 1.05530787,
      "balance_loss_mlp": 1.03134012,
      "epoch": 0.18560048098602133,
      "flos": 22346887570560.0,
      "grad_norm": 7.927127736744579,
      "language_loss": 0.69762361,
      "learning_rate": 3.7513510225656132e-06,
      "loss": 0.71954978,
      "num_input_tokens_seen": 66670045,
      "step": 3087,
      "time_per_iteration": 2.668849229812622
    },
    {
      "auxiliary_loss_clip": 0.01119022,
      "auxiliary_loss_mlp": 0.01055302,
      "balance_loss_clip": 1.05543649,
      "balance_loss_mlp": 1.03546548,
      "epoch": 0.18566060423868933,
      "flos": 17748526308480.0,
      "grad_norm": 2.1117122734340263,
      "language_loss": 0.72513628,
      "learning_rate": 3.7511629183195764e-06,
      "loss": 0.74687952,
      "num_input_tokens_seen": 66688790,
      "step": 3088,
      "time_per_iteration": 2.7150719165802
    },
    {
      "auxiliary_loss_clip": 0.0112638,
      "auxiliary_loss_mlp": 0.01044188,
      "balance_loss_clip": 1.04933047,
      "balance_loss_mlp": 1.02616334,
      "epoch": 0.1857207274913573,
      "flos": 24677274142080.0,
      "grad_norm": 2.112009927874319,
      "language_loss": 0.91859758,
      "learning_rate": 3.7509747476692663e-06,
      "loss": 0.94030321,
      "num_input_tokens_seen": 66708090,
      "step": 3089,
      "time_per_iteration": 2.7239248752593994
    },
    {
      "auxiliary_loss_clip": 0.01104754,
      "auxiliary_loss_mlp": 0.01046981,
      "balance_loss_clip": 1.0494597,
      "balance_loss_mlp": 1.02919531,
      "epoch": 0.18578085074402526,
      "flos": 28147825198080.0,
      "grad_norm": 2.490831087537115,
      "language_loss": 0.57275403,
      "learning_rate": 3.7507865106218176e-06,
      "loss": 0.59427136,
      "num_input_tokens_seen": 66727320,
      "step": 3090,
      "time_per_iteration": 2.8263309001922607
    },
    {
      "auxiliary_loss_clip": 0.01125877,
      "auxiliary_loss_mlp": 0.0104478,
      "balance_loss_clip": 1.04981184,
      "balance_loss_mlp": 1.02636242,
      "epoch": 0.18584097399669322,
      "flos": 23951878980480.0,
      "grad_norm": 1.7797305478565062,
      "language_loss": 0.81704801,
      "learning_rate": 3.7505982071843695e-06,
      "loss": 0.83875453,
      "num_input_tokens_seen": 66747505,
      "step": 3091,
      "time_per_iteration": 2.697525978088379
    },
    {
      "auxiliary_loss_clip": 0.01101743,
      "auxiliary_loss_mlp": 0.01050837,
      "balance_loss_clip": 1.04999971,
      "balance_loss_mlp": 1.03277707,
      "epoch": 0.18590109724936119,
      "flos": 17201678676480.0,
      "grad_norm": 2.0826959244757832,
      "language_loss": 0.83704746,
      "learning_rate": 3.7504098373640617e-06,
      "loss": 0.8585732,
      "num_input_tokens_seen": 66766425,
      "step": 3092,
      "time_per_iteration": 2.8379435539245605
    },
    {
      "auxiliary_loss_clip": 0.01136846,
      "auxiliary_loss_mlp": 0.01048758,
      "balance_loss_clip": 1.05389428,
      "balance_loss_mlp": 1.03036356,
      "epoch": 0.18596122050202915,
      "flos": 17234644383360.0,
      "grad_norm": 5.439917179387958,
      "language_loss": 0.93443698,
      "learning_rate": 3.750221401168038e-06,
      "loss": 0.95629299,
      "num_input_tokens_seen": 66781130,
      "step": 3093,
      "time_per_iteration": 2.8053483963012695
    },
    {
      "auxiliary_loss_clip": 0.01130362,
      "auxiliary_loss_mlp": 0.01042367,
      "balance_loss_clip": 1.05440521,
      "balance_loss_mlp": 1.02464092,
      "epoch": 0.18602134375469712,
      "flos": 19020733188480.0,
      "grad_norm": 1.7318887555782294,
      "language_loss": 0.77516603,
      "learning_rate": 3.750032898603443e-06,
      "loss": 0.7968933,
      "num_input_tokens_seen": 66797535,
      "step": 3094,
      "time_per_iteration": 2.7402310371398926
    },
    {
      "auxiliary_loss_clip": 0.0109741,
      "auxiliary_loss_mlp": 0.01049219,
      "balance_loss_clip": 1.0519228,
      "balance_loss_mlp": 1.0323391,
      "epoch": 0.1860814670073651,
      "flos": 50950094417280.0,
      "grad_norm": 1.7033453736007413,
      "language_loss": 0.69854707,
      "learning_rate": 3.749844329677425e-06,
      "loss": 0.72001338,
      "num_input_tokens_seen": 66821720,
      "step": 3095,
      "time_per_iteration": 3.133192777633667
    },
    {
      "auxiliary_loss_clip": 0.01113224,
      "auxiliary_loss_mlp": 0.010546,
      "balance_loss_clip": 1.0511899,
      "balance_loss_mlp": 1.03415525,
      "epoch": 0.18614159026003307,
      "flos": 19390972625280.0,
      "grad_norm": 2.2828801406167307,
      "language_loss": 0.81214821,
      "learning_rate": 3.749655694397135e-06,
      "loss": 0.83382642,
      "num_input_tokens_seen": 66839060,
      "step": 3096,
      "time_per_iteration": 2.7599101066589355
    },
    {
      "auxiliary_loss_clip": 0.01147399,
      "auxiliary_loss_mlp": 0.0104683,
      "balance_loss_clip": 1.05678356,
      "balance_loss_mlp": 1.02810192,
      "epoch": 0.18620171351270104,
      "flos": 21798782962560.0,
      "grad_norm": 2.430947734084612,
      "language_loss": 0.75326216,
      "learning_rate": 3.7494669927697255e-06,
      "loss": 0.77520448,
      "num_input_tokens_seen": 66857760,
      "step": 3097,
      "time_per_iteration": 4.255983114242554
    },
    {
      "auxiliary_loss_clip": 0.01133757,
      "auxiliary_loss_mlp": 0.01050365,
      "balance_loss_clip": 1.05756521,
      "balance_loss_mlp": 1.03228104,
      "epoch": 0.186261836765369,
      "flos": 16362877299840.0,
      "grad_norm": 2.553895603581972,
      "language_loss": 0.66602015,
      "learning_rate": 3.749278224802352e-06,
      "loss": 0.68786132,
      "num_input_tokens_seen": 66876460,
      "step": 3098,
      "time_per_iteration": 2.723567247390747
    },
    {
      "auxiliary_loss_clip": 0.01163461,
      "auxiliary_loss_mlp": 0.01052357,
      "balance_loss_clip": 1.05991709,
      "balance_loss_mlp": 1.03212702,
      "epoch": 0.18632196001803697,
      "flos": 23370054480000.0,
      "grad_norm": 1.6168121451860142,
      "language_loss": 0.69838905,
      "learning_rate": 3.7490893905021733e-06,
      "loss": 0.7205472,
      "num_input_tokens_seen": 66897960,
      "step": 3099,
      "time_per_iteration": 5.687380075454712
    },
    {
      "auxiliary_loss_clip": 0.01148363,
      "auxiliary_loss_mlp": 0.01051556,
      "balance_loss_clip": 1.05713868,
      "balance_loss_mlp": 1.03243458,
      "epoch": 0.18638208327070493,
      "flos": 22492002516480.0,
      "grad_norm": 1.7060244708994476,
      "language_loss": 0.71840072,
      "learning_rate": 3.7489004898763494e-06,
      "loss": 0.74039996,
      "num_input_tokens_seen": 66917675,
      "step": 3100,
      "time_per_iteration": 2.6711015701293945
    },
    {
      "auxiliary_loss_clip": 0.01138377,
      "auxiliary_loss_mlp": 0.01050667,
      "balance_loss_clip": 1.05749035,
      "balance_loss_mlp": 1.03133154,
      "epoch": 0.18644220652337293,
      "flos": 29165245931520.0,
      "grad_norm": 1.9639279354826686,
      "language_loss": 0.80343997,
      "learning_rate": 3.7487115229320444e-06,
      "loss": 0.82533038,
      "num_input_tokens_seen": 66936000,
      "step": 3101,
      "time_per_iteration": 2.6996583938598633
    },
    {
      "auxiliary_loss_clip": 0.01112778,
      "auxiliary_loss_mlp": 0.01042097,
      "balance_loss_clip": 1.05307627,
      "balance_loss_mlp": 1.02478826,
      "epoch": 0.1865023297760409,
      "flos": 24243796811520.0,
      "grad_norm": 1.8804860702941575,
      "language_loss": 0.77053607,
      "learning_rate": 3.7485224896764222e-06,
      "loss": 0.79208481,
      "num_input_tokens_seen": 66955700,
      "step": 3102,
      "time_per_iteration": 2.726146936416626
    },
    {
      "auxiliary_loss_clip": 0.01150817,
      "auxiliary_loss_mlp": 0.01039303,
      "balance_loss_clip": 1.057688,
      "balance_loss_mlp": 1.0213027,
      "epoch": 0.18656245302870886,
      "flos": 19128716449920.0,
      "grad_norm": 2.314682178811096,
      "language_loss": 0.76689744,
      "learning_rate": 3.7483333901166525e-06,
      "loss": 0.78879869,
      "num_input_tokens_seen": 66972815,
      "step": 3103,
      "time_per_iteration": 4.374122619628906
    },
    {
      "auxiliary_loss_clip": 0.01132531,
      "auxiliary_loss_mlp": 0.0104481,
      "balance_loss_clip": 1.05477643,
      "balance_loss_mlp": 1.02671361,
      "epoch": 0.18662257628137682,
      "flos": 17786088956160.0,
      "grad_norm": 1.6956506235876265,
      "language_loss": 0.79252636,
      "learning_rate": 3.7481442242599054e-06,
      "loss": 0.8142997,
      "num_input_tokens_seen": 66992280,
      "step": 3104,
      "time_per_iteration": 2.695012092590332
    },
    {
      "auxiliary_loss_clip": 0.01106786,
      "auxiliary_loss_mlp": 0.01050273,
      "balance_loss_clip": 1.05117702,
      "balance_loss_mlp": 1.03096056,
      "epoch": 0.1866826995340448,
      "flos": 24024382583040.0,
      "grad_norm": 2.065624302338532,
      "language_loss": 0.8496474,
      "learning_rate": 3.747954992113354e-06,
      "loss": 0.87121809,
      "num_input_tokens_seen": 67012220,
      "step": 3105,
      "time_per_iteration": 2.761521816253662
    },
    {
      "auxiliary_loss_clip": 0.0112324,
      "auxiliary_loss_mlp": 0.01043689,
      "balance_loss_clip": 1.05166531,
      "balance_loss_mlp": 1.02407932,
      "epoch": 0.18674282278671275,
      "flos": 26141244756480.0,
      "grad_norm": 1.8352441384571676,
      "language_loss": 0.86880243,
      "learning_rate": 3.7477656936841742e-06,
      "loss": 0.8904717,
      "num_input_tokens_seen": 67032030,
      "step": 3106,
      "time_per_iteration": 2.785738706588745
    },
    {
      "auxiliary_loss_clip": 0.01150222,
      "auxiliary_loss_mlp": 0.01040973,
      "balance_loss_clip": 1.0566026,
      "balance_loss_mlp": 1.02281737,
      "epoch": 0.18680294603938072,
      "flos": 19201938324480.0,
      "grad_norm": 2.128833658771433,
      "language_loss": 0.78226906,
      "learning_rate": 3.7475763289795445e-06,
      "loss": 0.80418098,
      "num_input_tokens_seen": 67048920,
      "step": 3107,
      "time_per_iteration": 2.693995237350464
    },
    {
      "auxiliary_loss_clip": 0.01153763,
      "auxiliary_loss_mlp": 0.01053056,
      "balance_loss_clip": 1.05873394,
      "balance_loss_mlp": 1.03341043,
      "epoch": 0.1868630692920487,
      "flos": 28544889116160.0,
      "grad_norm": 3.0927798335187506,
      "language_loss": 0.74159014,
      "learning_rate": 3.7473868980066446e-06,
      "loss": 0.7636584,
      "num_input_tokens_seen": 67068645,
      "step": 3108,
      "time_per_iteration": 2.795715570449829
    },
    {
      "auxiliary_loss_clip": 0.01107582,
      "auxiliary_loss_mlp": 0.01042714,
      "balance_loss_clip": 1.05207491,
      "balance_loss_mlp": 1.02451098,
      "epoch": 0.18692319254471668,
      "flos": 17238020261760.0,
      "grad_norm": 1.6837485322309411,
      "language_loss": 0.74348569,
      "learning_rate": 3.747197400772658e-06,
      "loss": 0.76498872,
      "num_input_tokens_seen": 67087075,
      "step": 3109,
      "time_per_iteration": 2.7627830505371094
    },
    {
      "auxiliary_loss_clip": 0.01145572,
      "auxiliary_loss_mlp": 0.01044117,
      "balance_loss_clip": 1.05631042,
      "balance_loss_mlp": 1.02526462,
      "epoch": 0.18698331579738464,
      "flos": 23185186156800.0,
      "grad_norm": 1.499459601293056,
      "language_loss": 0.84250218,
      "learning_rate": 3.747007837284772e-06,
      "loss": 0.86439908,
      "num_input_tokens_seen": 67108040,
      "step": 3110,
      "time_per_iteration": 2.7665328979492188
    },
    {
      "auxiliary_loss_clip": 0.01147578,
      "auxiliary_loss_mlp": 0.01042389,
      "balance_loss_clip": 1.05929494,
      "balance_loss_mlp": 1.02381575,
      "epoch": 0.1870434390500526,
      "flos": 25516721963520.0,
      "grad_norm": 1.9108380391903876,
      "language_loss": 0.84738445,
      "learning_rate": 3.7468182075501737e-06,
      "loss": 0.86928415,
      "num_input_tokens_seen": 67127605,
      "step": 3111,
      "time_per_iteration": 2.729233741760254
    },
    {
      "auxiliary_loss_clip": 0.01128,
      "auxiliary_loss_mlp": 0.01044544,
      "balance_loss_clip": 1.05348754,
      "balance_loss_mlp": 1.02635229,
      "epoch": 0.18710356230272057,
      "flos": 19500823393920.0,
      "grad_norm": 1.8704338434966796,
      "language_loss": 0.76875687,
      "learning_rate": 3.7466285115760536e-06,
      "loss": 0.79048228,
      "num_input_tokens_seen": 67145785,
      "step": 3112,
      "time_per_iteration": 2.7392494678497314
    },
    {
      "auxiliary_loss_clip": 0.0114846,
      "auxiliary_loss_mlp": 0.0104709,
      "balance_loss_clip": 1.05636978,
      "balance_loss_mlp": 1.02913654,
      "epoch": 0.18716368555538854,
      "flos": 26760847386240.0,
      "grad_norm": 1.8996972204761096,
      "language_loss": 0.64466536,
      "learning_rate": 3.7464387493696046e-06,
      "loss": 0.66662085,
      "num_input_tokens_seen": 67165930,
      "step": 3113,
      "time_per_iteration": 2.7393765449523926
    },
    {
      "auxiliary_loss_clip": 0.01153807,
      "auxiliary_loss_mlp": 0.01048748,
      "balance_loss_clip": 1.05685568,
      "balance_loss_mlp": 1.02900672,
      "epoch": 0.1872238088080565,
      "flos": 25189827264000.0,
      "grad_norm": 6.483287708452815,
      "language_loss": 0.817972,
      "learning_rate": 3.746248920938024e-06,
      "loss": 0.83999759,
      "num_input_tokens_seen": 67185830,
      "step": 3114,
      "time_per_iteration": 2.740229368209839
    },
    {
      "auxiliary_loss_clip": 0.01104278,
      "auxiliary_loss_mlp": 0.01050738,
      "balance_loss_clip": 1.04921412,
      "balance_loss_mlp": 1.03024614,
      "epoch": 0.1872839320607245,
      "flos": 24134305178880.0,
      "grad_norm": 2.3064843449079175,
      "language_loss": 0.57413173,
      "learning_rate": 3.74605902628851e-06,
      "loss": 0.59568191,
      "num_input_tokens_seen": 67206930,
      "step": 3115,
      "time_per_iteration": 2.811549663543701
    },
    {
      "auxiliary_loss_clip": 0.01123025,
      "auxiliary_loss_mlp": 0.01052226,
      "balance_loss_clip": 1.05446446,
      "balance_loss_mlp": 1.03241396,
      "epoch": 0.18734405531339246,
      "flos": 21173793292800.0,
      "grad_norm": 2.577640519639585,
      "language_loss": 0.70842528,
      "learning_rate": 3.745869065428261e-06,
      "loss": 0.73017788,
      "num_input_tokens_seen": 67226290,
      "step": 3116,
      "time_per_iteration": 2.8053951263427734
    },
    {
      "auxiliary_loss_clip": 0.0115042,
      "auxiliary_loss_mlp": 0.01035569,
      "balance_loss_clip": 1.05196476,
      "balance_loss_mlp": 1.01787841,
      "epoch": 0.18740417856606043,
      "flos": 17237697039360.0,
      "grad_norm": 3.010261965906642,
      "language_loss": 0.78994375,
      "learning_rate": 3.7456790383644833e-06,
      "loss": 0.81180358,
      "num_input_tokens_seen": 67244410,
      "step": 3117,
      "time_per_iteration": 2.819415330886841
    },
    {
      "auxiliary_loss_clip": 0.01132901,
      "auxiliary_loss_mlp": 0.01049724,
      "balance_loss_clip": 1.05260777,
      "balance_loss_mlp": 1.03047204,
      "epoch": 0.1874643018187284,
      "flos": 32558049999360.0,
      "grad_norm": 2.2828109389679865,
      "language_loss": 0.83903432,
      "learning_rate": 3.745488945104381e-06,
      "loss": 0.86086059,
      "num_input_tokens_seen": 67264470,
      "step": 3118,
      "time_per_iteration": 2.783804416656494
    },
    {
      "auxiliary_loss_clip": 0.01144867,
      "auxiliary_loss_mlp": 0.0104452,
      "balance_loss_clip": 1.05412436,
      "balance_loss_mlp": 1.02688873,
      "epoch": 0.18752442507139636,
      "flos": 23258156636160.0,
      "grad_norm": 3.566737352043019,
      "language_loss": 0.76283264,
      "learning_rate": 3.7452987856551636e-06,
      "loss": 0.78472656,
      "num_input_tokens_seen": 67284315,
      "step": 3119,
      "time_per_iteration": 2.6872506141662598
    },
    {
      "auxiliary_loss_clip": 0.01156835,
      "auxiliary_loss_mlp": 0.01046653,
      "balance_loss_clip": 1.05519438,
      "balance_loss_mlp": 1.02899814,
      "epoch": 0.18758454832406432,
      "flos": 21760933006080.0,
      "grad_norm": 1.7224942549361077,
      "language_loss": 0.82017547,
      "learning_rate": 3.7451085600240406e-06,
      "loss": 0.84221041,
      "num_input_tokens_seen": 67302780,
      "step": 3120,
      "time_per_iteration": 2.637505292892456
    },
    {
      "auxiliary_loss_clip": 0.0113033,
      "auxiliary_loss_mlp": 0.01035538,
      "balance_loss_clip": 1.05060756,
      "balance_loss_mlp": 1.01828837,
      "epoch": 0.1876446715767323,
      "flos": 29570210841600.0,
      "grad_norm": 2.5027223446471982,
      "language_loss": 0.84992659,
      "learning_rate": 3.7449182682182263e-06,
      "loss": 0.87158525,
      "num_input_tokens_seen": 67323405,
      "step": 3121,
      "time_per_iteration": 2.788353681564331
    },
    {
      "auxiliary_loss_clip": 0.01096681,
      "auxiliary_loss_mlp": 0.0104429,
      "balance_loss_clip": 1.045645,
      "balance_loss_mlp": 1.02599168,
      "epoch": 0.18770479482940028,
      "flos": 30339992234880.0,
      "grad_norm": 2.1738591443482362,
      "language_loss": 0.70032287,
      "learning_rate": 3.744727910244937e-06,
      "loss": 0.72173256,
      "num_input_tokens_seen": 67345800,
      "step": 3122,
      "time_per_iteration": 3.0225250720977783
    },
    {
      "auxiliary_loss_clip": 0.01153439,
      "auxiliary_loss_mlp": 0.01042355,
      "balance_loss_clip": 1.05445123,
      "balance_loss_mlp": 1.02288795,
      "epoch": 0.18776491808206824,
      "flos": 14465357527680.0,
      "grad_norm": 4.839579375412361,
      "language_loss": 0.70661515,
      "learning_rate": 3.7445374861113905e-06,
      "loss": 0.72857308,
      "num_input_tokens_seen": 67363575,
      "step": 3123,
      "time_per_iteration": 2.779904365539551
    },
    {
      "auxiliary_loss_clip": 0.01142265,
      "auxiliary_loss_mlp": 0.01041425,
      "balance_loss_clip": 1.05286181,
      "balance_loss_mlp": 1.02454507,
      "epoch": 0.1878250413347362,
      "flos": 24498547044480.0,
      "grad_norm": 2.057520579072589,
      "language_loss": 0.74103826,
      "learning_rate": 3.7443469958248066e-06,
      "loss": 0.76287514,
      "num_input_tokens_seen": 67381765,
      "step": 3124,
      "time_per_iteration": 2.6336071491241455
    },
    {
      "auxiliary_loss_clip": 0.01157579,
      "auxiliary_loss_mlp": 0.01052509,
      "balance_loss_clip": 1.05653572,
      "balance_loss_mlp": 1.03333998,
      "epoch": 0.18788516458740417,
      "flos": 39786185692800.0,
      "grad_norm": 3.0670363966795096,
      "language_loss": 0.80654436,
      "learning_rate": 3.7441564393924106e-06,
      "loss": 0.82864523,
      "num_input_tokens_seen": 67405000,
      "step": 3125,
      "time_per_iteration": 2.7224199771881104
    },
    {
      "auxiliary_loss_clip": 0.01046615,
      "auxiliary_loss_mlp": 0.01006504,
      "balance_loss_clip": 1.04444218,
      "balance_loss_mlp": 1.00435853,
      "epoch": 0.18794528784007214,
      "flos": 64699250664960.0,
      "grad_norm": 0.9424570711133922,
      "language_loss": 0.63647306,
      "learning_rate": 3.7439658168214273e-06,
      "loss": 0.65700436,
      "num_input_tokens_seen": 67467140,
      "step": 3126,
      "time_per_iteration": 3.313321113586426
    },
    {
      "auxiliary_loss_clip": 0.01128308,
      "auxiliary_loss_mlp": 0.01040458,
      "balance_loss_clip": 1.05377257,
      "balance_loss_mlp": 1.02236164,
      "epoch": 0.1880054110927401,
      "flos": 28622061486720.0,
      "grad_norm": 1.8734163453478039,
      "language_loss": 0.81308508,
      "learning_rate": 3.7437751281190857e-06,
      "loss": 0.83477271,
      "num_input_tokens_seen": 67487980,
      "step": 3127,
      "time_per_iteration": 2.7137866020202637
    },
    {
      "auxiliary_loss_clip": 0.01088267,
      "auxiliary_loss_mlp": 0.0101138,
      "balance_loss_clip": 1.04814553,
      "balance_loss_mlp": 1.00912714,
      "epoch": 0.1880655343454081,
      "flos": 64488958490880.0,
      "grad_norm": 0.7699217277386954,
      "language_loss": 0.61922526,
      "learning_rate": 3.7435843732926164e-06,
      "loss": 0.64022171,
      "num_input_tokens_seen": 67552500,
      "step": 3128,
      "time_per_iteration": 3.264270782470703
    },
    {
      "auxiliary_loss_clip": 0.01108205,
      "auxiliary_loss_mlp": 0.01049422,
      "balance_loss_clip": 1.04763842,
      "balance_loss_mlp": 1.02907288,
      "epoch": 0.18812565759807606,
      "flos": 32124464928000.0,
      "grad_norm": 2.4867495334212175,
      "language_loss": 0.70985162,
      "learning_rate": 3.7433935523492536e-06,
      "loss": 0.73142785,
      "num_input_tokens_seen": 67573295,
      "step": 3129,
      "time_per_iteration": 2.79929256439209
    },
    {
      "auxiliary_loss_clip": 0.01158485,
      "auxiliary_loss_mlp": 0.01050611,
      "balance_loss_clip": 1.05767536,
      "balance_loss_mlp": 1.03109634,
      "epoch": 0.18818578085074403,
      "flos": 20624539449600.0,
      "grad_norm": 2.4831518001798676,
      "language_loss": 0.85035253,
      "learning_rate": 3.7432026652962314e-06,
      "loss": 0.87244344,
      "num_input_tokens_seen": 67590010,
      "step": 3130,
      "time_per_iteration": 2.60624361038208
    },
    {
      "auxiliary_loss_clip": 0.01107202,
      "auxiliary_loss_mlp": 0.01049966,
      "balance_loss_clip": 1.04649067,
      "balance_loss_mlp": 1.03023696,
      "epoch": 0.188245904103412,
      "flos": 28840506048000.0,
      "grad_norm": 9.096753382647533,
      "language_loss": 0.7643525,
      "learning_rate": 3.7430117121407897e-06,
      "loss": 0.7859242,
      "num_input_tokens_seen": 67611110,
      "step": 3131,
      "time_per_iteration": 2.759230136871338
    },
    {
      "auxiliary_loss_clip": 0.0112329,
      "auxiliary_loss_mlp": 0.01049221,
      "balance_loss_clip": 1.05344164,
      "balance_loss_mlp": 1.03014708,
      "epoch": 0.18830602735607996,
      "flos": 29420319386880.0,
      "grad_norm": 2.109252219381847,
      "language_loss": 0.80713749,
      "learning_rate": 3.74282069289017e-06,
      "loss": 0.82886261,
      "num_input_tokens_seen": 67631990,
      "step": 3132,
      "time_per_iteration": 2.773817777633667
    },
    {
      "auxiliary_loss_clip": 0.01093588,
      "auxiliary_loss_mlp": 0.00779094,
      "balance_loss_clip": 1.04652429,
      "balance_loss_mlp": 1.00091529,
      "epoch": 0.18836615060874792,
      "flos": 28872933050880.0,
      "grad_norm": 2.092242478448591,
      "language_loss": 0.79653811,
      "learning_rate": 3.742629607551614e-06,
      "loss": 0.81526494,
      "num_input_tokens_seen": 67650490,
      "step": 3133,
      "time_per_iteration": 2.7873754501342773
    },
    {
      "auxiliary_loss_clip": 0.01119878,
      "auxiliary_loss_mlp": 0.01059381,
      "balance_loss_clip": 1.05341148,
      "balance_loss_mlp": 1.03921056,
      "epoch": 0.18842627386141592,
      "flos": 22601673717120.0,
      "grad_norm": 1.9069857551930867,
      "language_loss": 0.83001804,
      "learning_rate": 3.7424384561323698e-06,
      "loss": 0.85181063,
      "num_input_tokens_seen": 67668860,
      "step": 3134,
      "time_per_iteration": 2.9284298419952393
    },
    {
      "auxiliary_loss_clip": 0.01131578,
      "auxiliary_loss_mlp": 0.01046681,
      "balance_loss_clip": 1.05168402,
      "balance_loss_mlp": 1.02802503,
      "epoch": 0.18848639711408388,
      "flos": 24573600512640.0,
      "grad_norm": 2.0376543711114152,
      "language_loss": 0.82859468,
      "learning_rate": 3.742247238639684e-06,
      "loss": 0.85037726,
      "num_input_tokens_seen": 67690220,
      "step": 3135,
      "time_per_iteration": 2.8006811141967773
    },
    {
      "auxiliary_loss_clip": 0.01143148,
      "auxiliary_loss_mlp": 0.01050197,
      "balance_loss_clip": 1.05505157,
      "balance_loss_mlp": 1.03146911,
      "epoch": 0.18854652036675185,
      "flos": 34166920078080.0,
      "grad_norm": 1.9728388324049713,
      "language_loss": 0.78658557,
      "learning_rate": 3.7420559550808083e-06,
      "loss": 0.80851901,
      "num_input_tokens_seen": 67709820,
      "step": 3136,
      "time_per_iteration": 4.256143569946289
    },
    {
      "auxiliary_loss_clip": 0.01135545,
      "auxiliary_loss_mlp": 0.01048618,
      "balance_loss_clip": 1.05388892,
      "balance_loss_mlp": 1.03006911,
      "epoch": 0.1886066436194198,
      "flos": 24200236592640.0,
      "grad_norm": 1.7483697887361769,
      "language_loss": 0.80820233,
      "learning_rate": 3.741864605462996e-06,
      "loss": 0.83004391,
      "num_input_tokens_seen": 67729490,
      "step": 3137,
      "time_per_iteration": 2.7538130283355713
    },
    {
      "auxiliary_loss_clip": 0.01159054,
      "auxiliary_loss_mlp": 0.01048373,
      "balance_loss_clip": 1.05827475,
      "balance_loss_mlp": 1.03107548,
      "epoch": 0.18866676687208778,
      "flos": 21251109317760.0,
      "grad_norm": 1.9799764624272802,
      "language_loss": 0.81274408,
      "learning_rate": 3.741673189793504e-06,
      "loss": 0.83481836,
      "num_input_tokens_seen": 67749665,
      "step": 3138,
      "time_per_iteration": 4.143909931182861
    },
    {
      "auxiliary_loss_clip": 0.01150082,
      "auxiliary_loss_mlp": 0.01056444,
      "balance_loss_clip": 1.05626798,
      "balance_loss_mlp": 1.03713167,
      "epoch": 0.18872689012475574,
      "flos": 37308673013760.0,
      "grad_norm": 2.326218248348143,
      "language_loss": 0.63655496,
      "learning_rate": 3.7414817080795896e-06,
      "loss": 0.65862024,
      "num_input_tokens_seen": 67776230,
      "step": 3139,
      "time_per_iteration": 4.30991268157959
    },
    {
      "auxiliary_loss_clip": 0.0115289,
      "auxiliary_loss_mlp": 0.01043021,
      "balance_loss_clip": 1.05286491,
      "balance_loss_mlp": 1.02356625,
      "epoch": 0.1887870133774237,
      "flos": 21652303299840.0,
      "grad_norm": 2.1185902638296525,
      "language_loss": 0.7148211,
      "learning_rate": 3.741290160328514e-06,
      "loss": 0.73678017,
      "num_input_tokens_seen": 67795080,
      "step": 3140,
      "time_per_iteration": 2.6880578994750977
    },
    {
      "auxiliary_loss_clip": 0.01154738,
      "auxiliary_loss_mlp": 0.01043099,
      "balance_loss_clip": 1.05349982,
      "balance_loss_mlp": 1.02382278,
      "epoch": 0.1888471366300917,
      "flos": 15924659374080.0,
      "grad_norm": 2.6250212982316574,
      "language_loss": 0.87069929,
      "learning_rate": 3.7410985465475412e-06,
      "loss": 0.89267766,
      "num_input_tokens_seen": 67813110,
      "step": 3141,
      "time_per_iteration": 2.6677181720733643
    },
    {
      "auxiliary_loss_clip": 0.01130655,
      "auxiliary_loss_mlp": 0.01052882,
      "balance_loss_clip": 1.0507834,
      "balance_loss_mlp": 1.03243756,
      "epoch": 0.18890725988275966,
      "flos": 18551955767040.0,
      "grad_norm": 1.873404502116747,
      "language_loss": 0.7744689,
      "learning_rate": 3.7409068667439378e-06,
      "loss": 0.79630429,
      "num_input_tokens_seen": 67831070,
      "step": 3142,
      "time_per_iteration": 2.63077449798584
    },
    {
      "auxiliary_loss_clip": 0.01128192,
      "auxiliary_loss_mlp": 0.01038074,
      "balance_loss_clip": 1.05298221,
      "balance_loss_mlp": 1.02132463,
      "epoch": 0.18896738313542763,
      "flos": 28840865184000.0,
      "grad_norm": 1.6611052928231447,
      "language_loss": 0.78867507,
      "learning_rate": 3.740715120924971e-06,
      "loss": 0.81033778,
      "num_input_tokens_seen": 67852170,
      "step": 3143,
      "time_per_iteration": 4.417406797409058
    },
    {
      "auxiliary_loss_clip": 0.0111986,
      "auxiliary_loss_mlp": 0.01048019,
      "balance_loss_clip": 1.05024099,
      "balance_loss_mlp": 1.02821851,
      "epoch": 0.1890275063880956,
      "flos": 22412747157120.0,
      "grad_norm": 2.855732191409361,
      "language_loss": 0.71476078,
      "learning_rate": 3.740523309097912e-06,
      "loss": 0.73643959,
      "num_input_tokens_seen": 67869945,
      "step": 3144,
      "time_per_iteration": 2.8104894161224365
    },
    {
      "auxiliary_loss_clip": 0.01125398,
      "auxiliary_loss_mlp": 0.01044816,
      "balance_loss_clip": 1.05102479,
      "balance_loss_mlp": 1.02492023,
      "epoch": 0.18908762964076356,
      "flos": 24243904552320.0,
      "grad_norm": 2.5973078221757144,
      "language_loss": 0.73390597,
      "learning_rate": 3.7403314312700356e-06,
      "loss": 0.75560808,
      "num_input_tokens_seen": 67890240,
      "step": 3145,
      "time_per_iteration": 2.715609312057495
    },
    {
      "auxiliary_loss_clip": 0.01110308,
      "auxiliary_loss_mlp": 0.01042542,
      "balance_loss_clip": 1.04543984,
      "balance_loss_mlp": 1.02446938,
      "epoch": 0.18914775289343153,
      "flos": 16982910892800.0,
      "grad_norm": 2.915733862437625,
      "language_loss": 0.76263785,
      "learning_rate": 3.740139487448616e-06,
      "loss": 0.78416634,
      "num_input_tokens_seen": 67907825,
      "step": 3146,
      "time_per_iteration": 2.777221202850342
    },
    {
      "auxiliary_loss_clip": 0.01092807,
      "auxiliary_loss_mlp": 0.01049336,
      "balance_loss_clip": 1.04319823,
      "balance_loss_mlp": 1.02829611,
      "epoch": 0.1892078761460995,
      "flos": 21543781334400.0,
      "grad_norm": 1.988128972125699,
      "language_loss": 0.7837925,
      "learning_rate": 3.7399474776409326e-06,
      "loss": 0.80521393,
      "num_input_tokens_seen": 67926670,
      "step": 3147,
      "time_per_iteration": 2.8039205074310303
    },
    {
      "auxiliary_loss_clip": 0.01143577,
      "auxiliary_loss_mlp": 0.01042953,
      "balance_loss_clip": 1.0548687,
      "balance_loss_mlp": 1.02454758,
      "epoch": 0.18926799939876748,
      "flos": 23001538896000.0,
      "grad_norm": 3.932544798883504,
      "language_loss": 0.67477876,
      "learning_rate": 3.739755401854267e-06,
      "loss": 0.69664401,
      "num_input_tokens_seen": 67943645,
      "step": 3148,
      "time_per_iteration": 2.7273359298706055
    },
    {
      "auxiliary_loss_clip": 0.01112331,
      "auxiliary_loss_mlp": 0.01039139,
      "balance_loss_clip": 1.04617155,
      "balance_loss_mlp": 1.02014899,
      "epoch": 0.18932812265143545,
      "flos": 22273019251200.0,
      "grad_norm": 2.9848849244070315,
      "language_loss": 0.76207471,
      "learning_rate": 3.739563260095902e-06,
      "loss": 0.78358936,
      "num_input_tokens_seen": 67962345,
      "step": 3149,
      "time_per_iteration": 2.8031978607177734
    },
    {
      "auxiliary_loss_clip": 0.01130375,
      "auxiliary_loss_mlp": 0.01045773,
      "balance_loss_clip": 1.05438852,
      "balance_loss_mlp": 1.02797484,
      "epoch": 0.1893882459041034,
      "flos": 18624423456000.0,
      "grad_norm": 2.3661599820320136,
      "language_loss": 0.80378366,
      "learning_rate": 3.7393710523731245e-06,
      "loss": 0.82554519,
      "num_input_tokens_seen": 67979760,
      "step": 3150,
      "time_per_iteration": 2.7836129665374756
    },
    {
      "auxiliary_loss_clip": 0.01137112,
      "auxiliary_loss_mlp": 0.0104876,
      "balance_loss_clip": 1.0528239,
      "balance_loss_mlp": 1.03019929,
      "epoch": 0.18944836915677138,
      "flos": 22892981016960.0,
      "grad_norm": 2.0711129864945956,
      "language_loss": 0.85251844,
      "learning_rate": 3.7391787786932215e-06,
      "loss": 0.87437713,
      "num_input_tokens_seen": 67996895,
      "step": 3151,
      "time_per_iteration": 2.7782201766967773
    },
    {
      "auxiliary_loss_clip": 0.01121267,
      "auxiliary_loss_mlp": 0.01046776,
      "balance_loss_clip": 1.05223882,
      "balance_loss_mlp": 1.02839363,
      "epoch": 0.18950849240943934,
      "flos": 26796542526720.0,
      "grad_norm": 2.1337439707996673,
      "language_loss": 0.74114192,
      "learning_rate": 3.7389864390634857e-06,
      "loss": 0.76282233,
      "num_input_tokens_seen": 68018365,
      "step": 3152,
      "time_per_iteration": 2.8767755031585693
    },
    {
      "auxiliary_loss_clip": 0.01120312,
      "auxiliary_loss_mlp": 0.0104438,
      "balance_loss_clip": 1.05119991,
      "balance_loss_mlp": 1.02463925,
      "epoch": 0.1895686156621073,
      "flos": 24971239048320.0,
      "grad_norm": 1.9471461777193173,
      "language_loss": 0.75520492,
      "learning_rate": 3.738794033491209e-06,
      "loss": 0.77685189,
      "num_input_tokens_seen": 68037985,
      "step": 3153,
      "time_per_iteration": 2.7722980976104736
    },
    {
      "auxiliary_loss_clip": 0.01158287,
      "auxiliary_loss_mlp": 0.01049678,
      "balance_loss_clip": 1.0559293,
      "balance_loss_mlp": 1.03102183,
      "epoch": 0.1896287389147753,
      "flos": 21944544353280.0,
      "grad_norm": 2.099749434473157,
      "language_loss": 0.79984629,
      "learning_rate": 3.7386015619836887e-06,
      "loss": 0.82192594,
      "num_input_tokens_seen": 68057975,
      "step": 3154,
      "time_per_iteration": 2.6530587673187256
    },
    {
      "auxiliary_loss_clip": 0.01117992,
      "auxiliary_loss_mlp": 0.01056707,
      "balance_loss_clip": 1.04851115,
      "balance_loss_mlp": 1.03536844,
      "epoch": 0.18968886216744327,
      "flos": 18179058723840.0,
      "grad_norm": 3.210440214164498,
      "language_loss": 0.73046303,
      "learning_rate": 3.738409024548223e-06,
      "loss": 0.75220996,
      "num_input_tokens_seen": 68074175,
      "step": 3155,
      "time_per_iteration": 2.729832410812378
    },
    {
      "auxiliary_loss_clip": 0.01126019,
      "auxiliary_loss_mlp": 0.01045659,
      "balance_loss_clip": 1.05104291,
      "balance_loss_mlp": 1.02626419,
      "epoch": 0.18974898542011123,
      "flos": 20412487509120.0,
      "grad_norm": 1.8299076145086866,
      "language_loss": 0.73869717,
      "learning_rate": 3.7382164211921136e-06,
      "loss": 0.76041389,
      "num_input_tokens_seen": 68095230,
      "step": 3156,
      "time_per_iteration": 2.6747231483459473
    },
    {
      "auxiliary_loss_clip": 0.01156549,
      "auxiliary_loss_mlp": 0.0104418,
      "balance_loss_clip": 1.05489409,
      "balance_loss_mlp": 1.02645326,
      "epoch": 0.1898091086727792,
      "flos": 23985024255360.0,
      "grad_norm": 1.9629652277148564,
      "language_loss": 0.68053937,
      "learning_rate": 3.7380237519226623e-06,
      "loss": 0.70254672,
      "num_input_tokens_seen": 68113805,
      "step": 3157,
      "time_per_iteration": 2.7092478275299072
    },
    {
      "auxiliary_loss_clip": 0.01114914,
      "auxiliary_loss_mlp": 0.01044181,
      "balance_loss_clip": 1.04805827,
      "balance_loss_mlp": 1.02533436,
      "epoch": 0.18986923192544716,
      "flos": 27637067756160.0,
      "grad_norm": 1.7829025355963362,
      "language_loss": 0.79893303,
      "learning_rate": 3.737831016747176e-06,
      "loss": 0.82052404,
      "num_input_tokens_seen": 68133190,
      "step": 3158,
      "time_per_iteration": 2.7921364307403564
    },
    {
      "auxiliary_loss_clip": 0.01163231,
      "auxiliary_loss_mlp": 0.01049502,
      "balance_loss_clip": 1.05787683,
      "balance_loss_mlp": 1.02923679,
      "epoch": 0.18992935517811513,
      "flos": 25484151306240.0,
      "grad_norm": 1.856283461980025,
      "language_loss": 0.72348613,
      "learning_rate": 3.737638215672964e-06,
      "loss": 0.74561346,
      "num_input_tokens_seen": 68152330,
      "step": 3159,
      "time_per_iteration": 2.6111273765563965
    },
    {
      "auxiliary_loss_clip": 0.01149613,
      "auxiliary_loss_mlp": 0.01053808,
      "balance_loss_clip": 1.05840325,
      "balance_loss_mlp": 1.03386414,
      "epoch": 0.1899894784307831,
      "flos": 17420805596160.0,
      "grad_norm": 2.2573250756933647,
      "language_loss": 0.84977192,
      "learning_rate": 3.7374453487073366e-06,
      "loss": 0.87180614,
      "num_input_tokens_seen": 68170185,
      "step": 3160,
      "time_per_iteration": 2.659259796142578
    },
    {
      "auxiliary_loss_clip": 0.01129342,
      "auxiliary_loss_mlp": 0.01049909,
      "balance_loss_clip": 1.05297387,
      "balance_loss_mlp": 1.03289795,
      "epoch": 0.19004960168345109,
      "flos": 27492240119040.0,
      "grad_norm": 2.752358611011079,
      "language_loss": 0.73407793,
      "learning_rate": 3.7372524158576074e-06,
      "loss": 0.7558704,
      "num_input_tokens_seen": 68191665,
      "step": 3161,
      "time_per_iteration": 2.784040689468384
    },
    {
      "auxiliary_loss_clip": 0.01139858,
      "auxiliary_loss_mlp": 0.0105519,
      "balance_loss_clip": 1.05456805,
      "balance_loss_mlp": 1.03476942,
      "epoch": 0.19010972493611905,
      "flos": 38654676385920.0,
      "grad_norm": 1.6629026055958476,
      "language_loss": 0.8115741,
      "learning_rate": 3.7370594171310926e-06,
      "loss": 0.83352458,
      "num_input_tokens_seen": 68214635,
      "step": 3162,
      "time_per_iteration": 2.9375386238098145
    },
    {
      "auxiliary_loss_clip": 0.01157449,
      "auxiliary_loss_mlp": 0.01040035,
      "balance_loss_clip": 1.05625844,
      "balance_loss_mlp": 1.02062798,
      "epoch": 0.19016984818878702,
      "flos": 19244744357760.0,
      "grad_norm": 2.448016750033594,
      "language_loss": 0.75615001,
      "learning_rate": 3.73686635253511e-06,
      "loss": 0.77812481,
      "num_input_tokens_seen": 68232150,
      "step": 3163,
      "time_per_iteration": 2.7344541549682617
    },
    {
      "auxiliary_loss_clip": 0.0110099,
      "auxiliary_loss_mlp": 0.01050093,
      "balance_loss_clip": 1.050578,
      "balance_loss_mlp": 1.02880192,
      "epoch": 0.19022997144145498,
      "flos": 37596891744000.0,
      "grad_norm": 2.2644227245470514,
      "language_loss": 0.74093997,
      "learning_rate": 3.736673222076982e-06,
      "loss": 0.76245081,
      "num_input_tokens_seen": 68253370,
      "step": 3164,
      "time_per_iteration": 2.9165730476379395
    },
    {
      "auxiliary_loss_clip": 0.01141317,
      "auxiliary_loss_mlp": 0.01038043,
      "balance_loss_clip": 1.05518687,
      "balance_loss_mlp": 1.0195303,
      "epoch": 0.19029009469412295,
      "flos": 61530921665280.0,
      "grad_norm": 1.5484522746055986,
      "language_loss": 0.66844344,
      "learning_rate": 3.7364800257640313e-06,
      "loss": 0.69023699,
      "num_input_tokens_seen": 68278895,
      "step": 3165,
      "time_per_iteration": 3.006096124649048
    },
    {
      "auxiliary_loss_clip": 0.01146225,
      "auxiliary_loss_mlp": 0.0104856,
      "balance_loss_clip": 1.05512285,
      "balance_loss_mlp": 1.02848506,
      "epoch": 0.1903502179467909,
      "flos": 13954851480960.0,
      "grad_norm": 2.8598536292657144,
      "language_loss": 0.74239767,
      "learning_rate": 3.7362867636035835e-06,
      "loss": 0.76434553,
      "num_input_tokens_seen": 68294880,
      "step": 3166,
      "time_per_iteration": 2.678844928741455
    },
    {
      "auxiliary_loss_clip": 0.01050093,
      "auxiliary_loss_mlp": 0.01014959,
      "balance_loss_clip": 1.04342103,
      "balance_loss_mlp": 1.01201403,
      "epoch": 0.1904103411994589,
      "flos": 66899641916160.0,
      "grad_norm": 0.7754190343967906,
      "language_loss": 0.50311053,
      "learning_rate": 3.736093435602968e-06,
      "loss": 0.52376103,
      "num_input_tokens_seen": 68359665,
      "step": 3167,
      "time_per_iteration": 3.277529239654541
    },
    {
      "auxiliary_loss_clip": 0.01138483,
      "auxiliary_loss_mlp": 0.01051348,
      "balance_loss_clip": 1.05485487,
      "balance_loss_mlp": 1.03293037,
      "epoch": 0.19047046445212687,
      "flos": 21908741472000.0,
      "grad_norm": 2.3487387451986192,
      "language_loss": 0.74504036,
      "learning_rate": 3.7359000417695156e-06,
      "loss": 0.76693863,
      "num_input_tokens_seen": 68378950,
      "step": 3168,
      "time_per_iteration": 2.690995216369629
    },
    {
      "auxiliary_loss_clip": 0.01040165,
      "auxiliary_loss_mlp": 0.01023518,
      "balance_loss_clip": 1.03869283,
      "balance_loss_mlp": 1.02085996,
      "epoch": 0.19053058770479483,
      "flos": 59255156701440.0,
      "grad_norm": 0.8605055473788603,
      "language_loss": 0.60079956,
      "learning_rate": 3.73570658211056e-06,
      "loss": 0.62143636,
      "num_input_tokens_seen": 68434235,
      "step": 3169,
      "time_per_iteration": 3.2108101844787598
    },
    {
      "auxiliary_loss_clip": 0.01103792,
      "auxiliary_loss_mlp": 0.01056606,
      "balance_loss_clip": 1.05267787,
      "balance_loss_mlp": 1.03741288,
      "epoch": 0.1905907109574628,
      "flos": 23951304362880.0,
      "grad_norm": 1.5575975614891868,
      "language_loss": 0.78179795,
      "learning_rate": 3.735513056633436e-06,
      "loss": 0.80340189,
      "num_input_tokens_seen": 68453830,
      "step": 3170,
      "time_per_iteration": 2.832043409347534
    },
    {
      "auxiliary_loss_clip": 0.01142047,
      "auxiliary_loss_mlp": 0.01045041,
      "balance_loss_clip": 1.05325115,
      "balance_loss_mlp": 1.02605128,
      "epoch": 0.19065083421013077,
      "flos": 20812316774400.0,
      "grad_norm": 1.7671932984988854,
      "language_loss": 0.78177166,
      "learning_rate": 3.7353194653454834e-06,
      "loss": 0.80364257,
      "num_input_tokens_seen": 68473005,
      "step": 3171,
      "time_per_iteration": 2.7823612689971924
    },
    {
      "auxiliary_loss_clip": 0.01158227,
      "auxiliary_loss_mlp": 0.01047345,
      "balance_loss_clip": 1.05499291,
      "balance_loss_mlp": 1.0285697,
      "epoch": 0.19071095746279873,
      "flos": 31284981192960.0,
      "grad_norm": 2.1976685633770905,
      "language_loss": 0.77953529,
      "learning_rate": 3.7351258082540426e-06,
      "loss": 0.80159104,
      "num_input_tokens_seen": 68493470,
      "step": 3172,
      "time_per_iteration": 2.746279001235962
    },
    {
      "auxiliary_loss_clip": 0.01145112,
      "auxiliary_loss_mlp": 0.01055334,
      "balance_loss_clip": 1.05438328,
      "balance_loss_mlp": 1.03703523,
      "epoch": 0.1907710807154667,
      "flos": 14356117290240.0,
      "grad_norm": 1.5258786569967644,
      "language_loss": 0.80223799,
      "learning_rate": 3.7349320853664576e-06,
      "loss": 0.82424247,
      "num_input_tokens_seen": 68511290,
      "step": 3173,
      "time_per_iteration": 2.7396810054779053
    },
    {
      "auxiliary_loss_clip": 0.01113266,
      "auxiliary_loss_mlp": 0.00778142,
      "balance_loss_clip": 1.04967713,
      "balance_loss_mlp": 1.00094676,
      "epoch": 0.1908312039681347,
      "flos": 26907039740160.0,
      "grad_norm": 1.5341307852526682,
      "language_loss": 0.78495061,
      "learning_rate": 3.7347382966900735e-06,
      "loss": 0.80386466,
      "num_input_tokens_seen": 68532575,
      "step": 3174,
      "time_per_iteration": 2.8579304218292236
    },
    {
      "auxiliary_loss_clip": 0.01106714,
      "auxiliary_loss_mlp": 0.01047557,
      "balance_loss_clip": 1.04928994,
      "balance_loss_mlp": 1.02838778,
      "epoch": 0.19089132722080265,
      "flos": 14494695960960.0,
      "grad_norm": 1.8075853216546063,
      "language_loss": 0.81067109,
      "learning_rate": 3.7345444422322395e-06,
      "loss": 0.83221382,
      "num_input_tokens_seen": 68548760,
      "step": 3175,
      "time_per_iteration": 2.718254804611206
    },
    {
      "auxiliary_loss_clip": 0.01080497,
      "auxiliary_loss_mlp": 0.01053652,
      "balance_loss_clip": 1.04361629,
      "balance_loss_mlp": 1.0342685,
      "epoch": 0.19095145047347062,
      "flos": 13952876232960.0,
      "grad_norm": 2.2545261224105873,
      "language_loss": 0.85529047,
      "learning_rate": 3.7343505220003067e-06,
      "loss": 0.87663192,
      "num_input_tokens_seen": 68563100,
      "step": 3176,
      "time_per_iteration": 4.2962729930877686
    },
    {
      "auxiliary_loss_clip": 0.0113361,
      "auxiliary_loss_mlp": 0.01059849,
      "balance_loss_clip": 1.05418086,
      "balance_loss_mlp": 1.03928506,
      "epoch": 0.19101157372613858,
      "flos": 25301832848640.0,
      "grad_norm": 2.0896270593066832,
      "language_loss": 0.813025,
      "learning_rate": 3.7341565360016285e-06,
      "loss": 0.83495957,
      "num_input_tokens_seen": 68581650,
      "step": 3177,
      "time_per_iteration": 2.815127372741699
    },
    {
      "auxiliary_loss_clip": 0.01122377,
      "auxiliary_loss_mlp": 0.01044946,
      "balance_loss_clip": 1.0482533,
      "balance_loss_mlp": 1.0265398,
      "epoch": 0.19107169697880655,
      "flos": 20558212986240.0,
      "grad_norm": 2.67963335978105,
      "language_loss": 0.7530241,
      "learning_rate": 3.73396248424356e-06,
      "loss": 0.7746973,
      "num_input_tokens_seen": 68600360,
      "step": 3178,
      "time_per_iteration": 4.351228475570679
    },
    {
      "auxiliary_loss_clip": 0.01146729,
      "auxiliary_loss_mlp": 0.01042476,
      "balance_loss_clip": 1.05574143,
      "balance_loss_mlp": 1.02458286,
      "epoch": 0.19113182023147451,
      "flos": 22163204396160.0,
      "grad_norm": 4.753014277211421,
      "language_loss": 0.81381619,
      "learning_rate": 3.7337683667334606e-06,
      "loss": 0.83570826,
      "num_input_tokens_seen": 68617885,
      "step": 3179,
      "time_per_iteration": 4.259284019470215
    },
    {
      "auxiliary_loss_clip": 0.01147837,
      "auxiliary_loss_mlp": 0.01048144,
      "balance_loss_clip": 1.05645823,
      "balance_loss_mlp": 1.0291661,
      "epoch": 0.19119194348414248,
      "flos": 18581796990720.0,
      "grad_norm": 2.753081884541086,
      "language_loss": 0.79384613,
      "learning_rate": 3.733574183478691e-06,
      "loss": 0.81580591,
      "num_input_tokens_seen": 68634550,
      "step": 3180,
      "time_per_iteration": 2.6609203815460205
    },
    {
      "auxiliary_loss_clip": 0.01129361,
      "auxiliary_loss_mlp": 0.0105402,
      "balance_loss_clip": 1.05249727,
      "balance_loss_mlp": 1.03445804,
      "epoch": 0.19125206673681047,
      "flos": 19026623018880.0,
      "grad_norm": 2.660238694189741,
      "language_loss": 0.79517245,
      "learning_rate": 3.733379934486615e-06,
      "loss": 0.81700623,
      "num_input_tokens_seen": 68651895,
      "step": 3181,
      "time_per_iteration": 2.6877176761627197
    },
    {
      "auxiliary_loss_clip": 0.0114301,
      "auxiliary_loss_mlp": 0.01053621,
      "balance_loss_clip": 1.05339336,
      "balance_loss_mlp": 1.03527462,
      "epoch": 0.19131218998947844,
      "flos": 21690153256320.0,
      "grad_norm": 2.2179888965480243,
      "language_loss": 0.74570775,
      "learning_rate": 3.7331856197645973e-06,
      "loss": 0.76767409,
      "num_input_tokens_seen": 68671500,
      "step": 3182,
      "time_per_iteration": 4.2829508781433105
    },
    {
      "auxiliary_loss_clip": 0.01128679,
      "auxiliary_loss_mlp": 0.01044063,
      "balance_loss_clip": 1.05578041,
      "balance_loss_mlp": 1.02575254,
      "epoch": 0.1913723132421464,
      "flos": 18442500048000.0,
      "grad_norm": 1.7534728284311585,
      "language_loss": 0.64618582,
      "learning_rate": 3.7329912393200084e-06,
      "loss": 0.66791326,
      "num_input_tokens_seen": 68690570,
      "step": 3183,
      "time_per_iteration": 2.7652854919433594
    },
    {
      "auxiliary_loss_clip": 0.01132257,
      "auxiliary_loss_mlp": 0.01050867,
      "balance_loss_clip": 1.0512805,
      "balance_loss_mlp": 1.0311259,
      "epoch": 0.19143243649481437,
      "flos": 27160102033920.0,
      "grad_norm": 1.555926798692704,
      "language_loss": 0.73459226,
      "learning_rate": 3.7327967931602173e-06,
      "loss": 0.75642347,
      "num_input_tokens_seen": 68709735,
      "step": 3184,
      "time_per_iteration": 2.6929056644439697
    },
    {
      "auxiliary_loss_clip": 0.01122578,
      "auxiliary_loss_mlp": 0.01054123,
      "balance_loss_clip": 1.05015373,
      "balance_loss_mlp": 1.03347623,
      "epoch": 0.19149255974748233,
      "flos": 21718952985600.0,
      "grad_norm": 2.0989643169058514,
      "language_loss": 0.87983418,
      "learning_rate": 3.732602281292598e-06,
      "loss": 0.9016012,
      "num_input_tokens_seen": 68727565,
      "step": 3185,
      "time_per_iteration": 2.6859230995178223
    },
    {
      "auxiliary_loss_clip": 0.01153787,
      "auxiliary_loss_mlp": 0.01044436,
      "balance_loss_clip": 1.05334914,
      "balance_loss_mlp": 1.02505302,
      "epoch": 0.1915526830001503,
      "flos": 22963293889920.0,
      "grad_norm": 2.4520480945942587,
      "language_loss": 0.73240852,
      "learning_rate": 3.7324077037245267e-06,
      "loss": 0.75439072,
      "num_input_tokens_seen": 68748110,
      "step": 3186,
      "time_per_iteration": 2.6398978233337402
    },
    {
      "auxiliary_loss_clip": 0.01132874,
      "auxiliary_loss_mlp": 0.01044989,
      "balance_loss_clip": 1.05609488,
      "balance_loss_mlp": 1.02379346,
      "epoch": 0.1916128062528183,
      "flos": 26140741966080.0,
      "grad_norm": 2.739457234253781,
      "language_loss": 0.83550584,
      "learning_rate": 3.7322130604633825e-06,
      "loss": 0.85728443,
      "num_input_tokens_seen": 68769765,
      "step": 3187,
      "time_per_iteration": 2.7476372718811035
    },
    {
      "auxiliary_loss_clip": 0.01076264,
      "auxiliary_loss_mlp": 0.01021317,
      "balance_loss_clip": 1.04604995,
      "balance_loss_mlp": 1.01892138,
      "epoch": 0.19167292950548626,
      "flos": 54925767457920.0,
      "grad_norm": 0.8659386797819415,
      "language_loss": 0.55824959,
      "learning_rate": 3.732018351516544e-06,
      "loss": 0.57922542,
      "num_input_tokens_seen": 68826815,
      "step": 3188,
      "time_per_iteration": 3.2144031524658203
    },
    {
      "auxiliary_loss_clip": 0.01139007,
      "auxiliary_loss_mlp": 0.01054399,
      "balance_loss_clip": 1.054564,
      "balance_loss_mlp": 1.03537333,
      "epoch": 0.19173305275815422,
      "flos": 29935601942400.0,
      "grad_norm": 2.2897904709915573,
      "language_loss": 0.69839454,
      "learning_rate": 3.731823576891397e-06,
      "loss": 0.72032857,
      "num_input_tokens_seen": 68847585,
      "step": 3189,
      "time_per_iteration": 2.7998950481414795
    },
    {
      "auxiliary_loss_clip": 0.01118438,
      "auxiliary_loss_mlp": 0.01038566,
      "balance_loss_clip": 1.04930174,
      "balance_loss_mlp": 1.02116132,
      "epoch": 0.1917931760108222,
      "flos": 24752471264640.0,
      "grad_norm": 2.362312815249866,
      "language_loss": 0.74320328,
      "learning_rate": 3.7316287365953266e-06,
      "loss": 0.76477331,
      "num_input_tokens_seen": 68866620,
      "step": 3190,
      "time_per_iteration": 2.7386670112609863
    },
    {
      "auxiliary_loss_clip": 0.01111071,
      "auxiliary_loss_mlp": 0.0106718,
      "balance_loss_clip": 1.04946983,
      "balance_loss_mlp": 1.04702199,
      "epoch": 0.19185329926349015,
      "flos": 18843550375680.0,
      "grad_norm": 3.545467698458187,
      "language_loss": 0.8444041,
      "learning_rate": 3.73143383063572e-06,
      "loss": 0.8661865,
      "num_input_tokens_seen": 68885515,
      "step": 3191,
      "time_per_iteration": 2.7025794982910156
    },
    {
      "auxiliary_loss_clip": 0.01127894,
      "auxiliary_loss_mlp": 0.01039849,
      "balance_loss_clip": 1.05251908,
      "balance_loss_mlp": 1.02231336,
      "epoch": 0.19191342251615812,
      "flos": 22086858038400.0,
      "grad_norm": 2.0663841109071526,
      "language_loss": 0.89985192,
      "learning_rate": 3.73123885901997e-06,
      "loss": 0.92152941,
      "num_input_tokens_seen": 68903225,
      "step": 3192,
      "time_per_iteration": 2.802852153778076
    },
    {
      "auxiliary_loss_clip": 0.01130336,
      "auxiliary_loss_mlp": 0.01054766,
      "balance_loss_clip": 1.05716372,
      "balance_loss_mlp": 1.03509688,
      "epoch": 0.19197354576882608,
      "flos": 22199115018240.0,
      "grad_norm": 2.3467564445058775,
      "language_loss": 0.75159264,
      "learning_rate": 3.7310438217554687e-06,
      "loss": 0.77344358,
      "num_input_tokens_seen": 68922860,
      "step": 3193,
      "time_per_iteration": 2.7680914402008057
    },
    {
      "auxiliary_loss_clip": 0.01128303,
      "auxiliary_loss_mlp": 0.00777332,
      "balance_loss_clip": 1.05222785,
      "balance_loss_mlp": 1.00071752,
      "epoch": 0.19203366902149407,
      "flos": 24896185580160.0,
      "grad_norm": 2.078743387775855,
      "language_loss": 0.75189757,
      "learning_rate": 3.730848718849612e-06,
      "loss": 0.77095383,
      "num_input_tokens_seen": 68943000,
      "step": 3194,
      "time_per_iteration": 2.7537553310394287
    },
    {
      "auxiliary_loss_clip": 0.01068142,
      "auxiliary_loss_mlp": 0.01004387,
      "balance_loss_clip": 1.03910232,
      "balance_loss_mlp": 1.00182378,
      "epoch": 0.19209379227416204,
      "flos": 68416722789120.0,
      "grad_norm": 0.7955224937316553,
      "language_loss": 0.68507159,
      "learning_rate": 3.7306535503097985e-06,
      "loss": 0.70579696,
      "num_input_tokens_seen": 69000255,
      "step": 3195,
      "time_per_iteration": 3.117191791534424
    },
    {
      "auxiliary_loss_clip": 0.01116081,
      "auxiliary_loss_mlp": 0.01052392,
      "balance_loss_clip": 1.05205238,
      "balance_loss_mlp": 1.0320189,
      "epoch": 0.19215391552683,
      "flos": 22055185221120.0,
      "grad_norm": 2.6559439291645757,
      "language_loss": 0.73141015,
      "learning_rate": 3.730458316143429e-06,
      "loss": 0.75309479,
      "num_input_tokens_seen": 69019665,
      "step": 3196,
      "time_per_iteration": 2.7234303951263428
    },
    {
      "auxiliary_loss_clip": 0.01139018,
      "auxiliary_loss_mlp": 0.01044947,
      "balance_loss_clip": 1.06151462,
      "balance_loss_mlp": 1.02596927,
      "epoch": 0.19221403877949797,
      "flos": 20302959962880.0,
      "grad_norm": 3.0997718824135734,
      "language_loss": 0.83654135,
      "learning_rate": 3.7302630163579068e-06,
      "loss": 0.85838103,
      "num_input_tokens_seen": 69039055,
      "step": 3197,
      "time_per_iteration": 2.72575306892395
    },
    {
      "auxiliary_loss_clip": 0.01086216,
      "auxiliary_loss_mlp": 0.01055059,
      "balance_loss_clip": 1.04615641,
      "balance_loss_mlp": 1.03320754,
      "epoch": 0.19227416203216594,
      "flos": 23185329811200.0,
      "grad_norm": 2.2465298420006383,
      "language_loss": 0.80656433,
      "learning_rate": 3.7300676509606373e-06,
      "loss": 0.82797706,
      "num_input_tokens_seen": 69056370,
      "step": 3198,
      "time_per_iteration": 2.741678237915039
    },
    {
      "auxiliary_loss_clip": 0.01135487,
      "auxiliary_loss_mlp": 0.01056572,
      "balance_loss_clip": 1.05502987,
      "balance_loss_mlp": 1.03655636,
      "epoch": 0.1923342852848339,
      "flos": 25776607841280.0,
      "grad_norm": 1.9205907836873994,
      "language_loss": 0.78993976,
      "learning_rate": 3.729872219959029e-06,
      "loss": 0.81186032,
      "num_input_tokens_seen": 69075915,
      "step": 3199,
      "time_per_iteration": 2.7821297645568848
    },
    {
      "auxiliary_loss_clip": 0.01116808,
      "auxiliary_loss_mlp": 0.01056964,
      "balance_loss_clip": 1.05010581,
      "balance_loss_mlp": 1.036412,
      "epoch": 0.19239440853750187,
      "flos": 17128349061120.0,
      "grad_norm": 3.662083840248298,
      "language_loss": 0.83574522,
      "learning_rate": 3.7296767233604934e-06,
      "loss": 0.85748297,
      "num_input_tokens_seen": 69094145,
      "step": 3200,
      "time_per_iteration": 2.7095022201538086
    },
    {
      "auxiliary_loss_clip": 0.01159025,
      "auxiliary_loss_mlp": 0.01048823,
      "balance_loss_clip": 1.05997193,
      "balance_loss_mlp": 1.03060746,
      "epoch": 0.19245453179016986,
      "flos": 16435093593600.0,
      "grad_norm": 1.9278966392289572,
      "language_loss": 0.79092836,
      "learning_rate": 3.729481161172443e-06,
      "loss": 0.81300688,
      "num_input_tokens_seen": 69111110,
      "step": 3201,
      "time_per_iteration": 2.684979200363159
    },
    {
      "auxiliary_loss_clip": 0.01103349,
      "auxiliary_loss_mlp": 0.01053366,
      "balance_loss_clip": 1.04825675,
      "balance_loss_mlp": 1.03418541,
      "epoch": 0.19251465504283782,
      "flos": 20230276792320.0,
      "grad_norm": 2.4062417134527645,
      "language_loss": 0.69276404,
      "learning_rate": 3.7292855334022927e-06,
      "loss": 0.71433127,
      "num_input_tokens_seen": 69130280,
      "step": 3202,
      "time_per_iteration": 2.8284943103790283
    },
    {
      "auxiliary_loss_clip": 0.01132334,
      "auxiliary_loss_mlp": 0.01041011,
      "balance_loss_clip": 1.05389905,
      "balance_loss_mlp": 1.02256894,
      "epoch": 0.1925747782955058,
      "flos": 19464374067840.0,
      "grad_norm": 1.9491265782204168,
      "language_loss": 0.91396749,
      "learning_rate": 3.7290898400574627e-06,
      "loss": 0.93570089,
      "num_input_tokens_seen": 69149570,
      "step": 3203,
      "time_per_iteration": 2.802433729171753
    },
    {
      "auxiliary_loss_clip": 0.0114953,
      "auxiliary_loss_mlp": 0.01049732,
      "balance_loss_clip": 1.05674863,
      "balance_loss_mlp": 1.02959776,
      "epoch": 0.19263490154817375,
      "flos": 17785586165760.0,
      "grad_norm": 5.05881669068558,
      "language_loss": 0.81689429,
      "learning_rate": 3.7288940811453725e-06,
      "loss": 0.83888692,
      "num_input_tokens_seen": 69168190,
      "step": 3204,
      "time_per_iteration": 2.671285629272461
    },
    {
      "auxiliary_loss_clip": 0.01116988,
      "auxiliary_loss_mlp": 0.01048941,
      "balance_loss_clip": 1.04950142,
      "balance_loss_mlp": 1.0298202,
      "epoch": 0.19269502480084172,
      "flos": 17457075354240.0,
      "grad_norm": 2.296941025186916,
      "language_loss": 0.76167846,
      "learning_rate": 3.7286982566734454e-06,
      "loss": 0.78333771,
      "num_input_tokens_seen": 69186950,
      "step": 3205,
      "time_per_iteration": 2.8654470443725586
    },
    {
      "auxiliary_loss_clip": 0.01140852,
      "auxiliary_loss_mlp": 0.01046651,
      "balance_loss_clip": 1.05839586,
      "balance_loss_mlp": 1.02749407,
      "epoch": 0.19275514805350968,
      "flos": 21506901045120.0,
      "grad_norm": 3.761768843322395,
      "language_loss": 0.83394569,
      "learning_rate": 3.728502366649107e-06,
      "loss": 0.85582072,
      "num_input_tokens_seen": 69204850,
      "step": 3206,
      "time_per_iteration": 2.8610613346099854
    },
    {
      "auxiliary_loss_clip": 0.0105715,
      "auxiliary_loss_mlp": 0.01004055,
      "balance_loss_clip": 1.03779244,
      "balance_loss_mlp": 1.00174224,
      "epoch": 0.19281527130617768,
      "flos": 47695979738880.0,
      "grad_norm": 0.8644529519848262,
      "language_loss": 0.60561717,
      "learning_rate": 3.728306411079786e-06,
      "loss": 0.62622917,
      "num_input_tokens_seen": 69259200,
      "step": 3207,
      "time_per_iteration": 3.126537322998047
    },
    {
      "auxiliary_loss_clip": 0.01120285,
      "auxiliary_loss_mlp": 0.01045527,
      "balance_loss_clip": 1.05201781,
      "balance_loss_mlp": 1.02678764,
      "epoch": 0.19287539455884564,
      "flos": 11801252672640.0,
      "grad_norm": 2.296187182186814,
      "language_loss": 0.75463599,
      "learning_rate": 3.7281103899729125e-06,
      "loss": 0.77629405,
      "num_input_tokens_seen": 69275835,
      "step": 3208,
      "time_per_iteration": 2.6978750228881836
    },
    {
      "auxiliary_loss_clip": 0.01150534,
      "auxiliary_loss_mlp": 0.00777875,
      "balance_loss_clip": 1.05520236,
      "balance_loss_mlp": 1.00063884,
      "epoch": 0.1929355178115136,
      "flos": 20631434860800.0,
      "grad_norm": 1.9483983315924505,
      "language_loss": 0.60869855,
      "learning_rate": 3.7279143033359195e-06,
      "loss": 0.62798262,
      "num_input_tokens_seen": 69294810,
      "step": 3209,
      "time_per_iteration": 2.699798107147217
    },
    {
      "auxiliary_loss_clip": 0.01158758,
      "auxiliary_loss_mlp": 0.01053815,
      "balance_loss_clip": 1.05472994,
      "balance_loss_mlp": 1.03261995,
      "epoch": 0.19299564106418157,
      "flos": 40807916058240.0,
      "grad_norm": 1.9992177661428934,
      "language_loss": 0.80025005,
      "learning_rate": 3.727718151176243e-06,
      "loss": 0.82237577,
      "num_input_tokens_seen": 69316065,
      "step": 3210,
      "time_per_iteration": 2.832665205001831
    },
    {
      "auxiliary_loss_clip": 0.01118997,
      "auxiliary_loss_mlp": 0.01047494,
      "balance_loss_clip": 1.05044246,
      "balance_loss_mlp": 1.02920699,
      "epoch": 0.19305576431684954,
      "flos": 11361418634880.0,
      "grad_norm": 2.515510367397107,
      "language_loss": 0.82571948,
      "learning_rate": 3.7275219335013217e-06,
      "loss": 0.84738445,
      "num_input_tokens_seen": 69332900,
      "step": 3211,
      "time_per_iteration": 2.7664191722869873
    },
    {
      "auxiliary_loss_clip": 0.01073663,
      "auxiliary_loss_mlp": 0.01002544,
      "balance_loss_clip": 1.03501034,
      "balance_loss_mlp": 1.00021982,
      "epoch": 0.1931158875695175,
      "flos": 54511895975040.0,
      "grad_norm": 0.9633495631759209,
      "language_loss": 0.63641912,
      "learning_rate": 3.7273256503185953e-06,
      "loss": 0.6571812,
      "num_input_tokens_seen": 69382535,
      "step": 3212,
      "time_per_iteration": 2.974940299987793
    },
    {
      "auxiliary_loss_clip": 0.01131742,
      "auxiliary_loss_mlp": 0.01044059,
      "balance_loss_clip": 1.05586314,
      "balance_loss_mlp": 1.02565336,
      "epoch": 0.19317601082218547,
      "flos": 19828436365440.0,
      "grad_norm": 1.7209148950717332,
      "language_loss": 0.76375663,
      "learning_rate": 3.7271293016355074e-06,
      "loss": 0.78551459,
      "num_input_tokens_seen": 69400600,
      "step": 3213,
      "time_per_iteration": 2.7898454666137695
    },
    {
      "auxiliary_loss_clip": 0.01123196,
      "auxiliary_loss_mlp": 0.0105066,
      "balance_loss_clip": 1.05261111,
      "balance_loss_mlp": 1.03116894,
      "epoch": 0.19323613407485346,
      "flos": 13152068467200.0,
      "grad_norm": 2.349758973823363,
      "language_loss": 0.70871878,
      "learning_rate": 3.726932887459503e-06,
      "loss": 0.73045731,
      "num_input_tokens_seen": 69417350,
      "step": 3214,
      "time_per_iteration": 2.8155152797698975
    },
    {
      "auxiliary_loss_clip": 0.01155585,
      "auxiliary_loss_mlp": 0.01047831,
      "balance_loss_clip": 1.05412841,
      "balance_loss_mlp": 1.02807808,
      "epoch": 0.19329625732752143,
      "flos": 14027247342720.0,
      "grad_norm": 2.190607045917922,
      "language_loss": 0.75067955,
      "learning_rate": 3.72673640779803e-06,
      "loss": 0.77271378,
      "num_input_tokens_seen": 69431845,
      "step": 3215,
      "time_per_iteration": 4.111938238143921
    },
    {
      "auxiliary_loss_clip": 0.01112217,
      "auxiliary_loss_mlp": 0.01049964,
      "balance_loss_clip": 1.04928339,
      "balance_loss_mlp": 1.0323447,
      "epoch": 0.1933563805801894,
      "flos": 23441732069760.0,
      "grad_norm": 1.7842520268521305,
      "language_loss": 0.88426638,
      "learning_rate": 3.72653986265854e-06,
      "loss": 0.9058882,
      "num_input_tokens_seen": 69453275,
      "step": 3216,
      "time_per_iteration": 2.7699615955352783
    },
    {
      "auxiliary_loss_clip": 0.01153806,
      "auxiliary_loss_mlp": 0.01052131,
      "balance_loss_clip": 1.05435801,
      "balance_loss_mlp": 1.03442836,
      "epoch": 0.19341650383285736,
      "flos": 20485314334080.0,
      "grad_norm": 1.6996051239972392,
      "language_loss": 0.7974773,
      "learning_rate": 3.726343252048485e-06,
      "loss": 0.81953669,
      "num_input_tokens_seen": 69471830,
      "step": 3217,
      "time_per_iteration": 2.6788718700408936
    },
    {
      "auxiliary_loss_clip": 0.01143281,
      "auxiliary_loss_mlp": 0.0104914,
      "balance_loss_clip": 1.05695105,
      "balance_loss_mlp": 1.02864754,
      "epoch": 0.19347662708552532,
      "flos": 17858484817920.0,
      "grad_norm": 4.708784796317305,
      "language_loss": 0.6161437,
      "learning_rate": 3.7261465759753206e-06,
      "loss": 0.6380679,
      "num_input_tokens_seen": 69489320,
      "step": 3218,
      "time_per_iteration": 4.352849960327148
    },
    {
      "auxiliary_loss_clip": 0.01157355,
      "auxiliary_loss_mlp": 0.01047211,
      "balance_loss_clip": 1.05723107,
      "balance_loss_mlp": 1.02873373,
      "epoch": 0.1935367503381933,
      "flos": 18187247024640.0,
      "grad_norm": 1.9724785552136583,
      "language_loss": 0.80345452,
      "learning_rate": 3.7259498344465053e-06,
      "loss": 0.82550013,
      "num_input_tokens_seen": 69506665,
      "step": 3219,
      "time_per_iteration": 4.1739161014556885
    },
    {
      "auxiliary_loss_clip": 0.01104687,
      "auxiliary_loss_mlp": 0.01047672,
      "balance_loss_clip": 1.05145359,
      "balance_loss_mlp": 1.02819324,
      "epoch": 0.19359687359086128,
      "flos": 15957122290560.0,
      "grad_norm": 2.7508533279024077,
      "language_loss": 0.85693008,
      "learning_rate": 3.7257530274694993e-06,
      "loss": 0.87845367,
      "num_input_tokens_seen": 69523835,
      "step": 3220,
      "time_per_iteration": 2.777284622192383
    },
    {
      "auxiliary_loss_clip": 0.01149581,
      "auxiliary_loss_mlp": 0.01041747,
      "balance_loss_clip": 1.05441856,
      "balance_loss_mlp": 1.02511764,
      "epoch": 0.19365699684352924,
      "flos": 21215198695680.0,
      "grad_norm": 2.05545450883527,
      "language_loss": 0.84637755,
      "learning_rate": 3.725556155051766e-06,
      "loss": 0.86829084,
      "num_input_tokens_seen": 69542620,
      "step": 3221,
      "time_per_iteration": 4.224115371704102
    },
    {
      "auxiliary_loss_clip": 0.01143661,
      "auxiliary_loss_mlp": 0.01044558,
      "balance_loss_clip": 1.05466259,
      "balance_loss_mlp": 1.02730846,
      "epoch": 0.1937171200961972,
      "flos": 17311098481920.0,
      "grad_norm": 2.658004231066563,
      "language_loss": 0.86087942,
      "learning_rate": 3.7253592172007702e-06,
      "loss": 0.8827616,
      "num_input_tokens_seen": 69561130,
      "step": 3222,
      "time_per_iteration": 2.6400530338287354
    },
    {
      "auxiliary_loss_clip": 0.01069453,
      "auxiliary_loss_mlp": 0.01045281,
      "balance_loss_clip": 1.04206085,
      "balance_loss_mlp": 1.02599275,
      "epoch": 0.19377724334886517,
      "flos": 22635968227200.0,
      "grad_norm": 1.8604116943694204,
      "language_loss": 0.78510809,
      "learning_rate": 3.72516221392398e-06,
      "loss": 0.8062554,
      "num_input_tokens_seen": 69580425,
      "step": 3223,
      "time_per_iteration": 2.9685652256011963
    },
    {
      "auxiliary_loss_clip": 0.01146062,
      "auxiliary_loss_mlp": 0.01046815,
      "balance_loss_clip": 1.05697751,
      "balance_loss_mlp": 1.02819431,
      "epoch": 0.19383736660153314,
      "flos": 15077813351040.0,
      "grad_norm": 1.8958208586464897,
      "language_loss": 0.75391948,
      "learning_rate": 3.7249651452288653e-06,
      "loss": 0.77584827,
      "num_input_tokens_seen": 69597085,
      "step": 3224,
      "time_per_iteration": 2.665294885635376
    },
    {
      "auxiliary_loss_clip": 0.01102293,
      "auxiliary_loss_mlp": 0.01050181,
      "balance_loss_clip": 1.04728186,
      "balance_loss_mlp": 1.02927208,
      "epoch": 0.1938974898542011,
      "flos": 47119934350080.0,
      "grad_norm": 3.358076005999295,
      "language_loss": 0.71180636,
      "learning_rate": 3.7247680111229e-06,
      "loss": 0.73333108,
      "num_input_tokens_seen": 69618885,
      "step": 3225,
      "time_per_iteration": 2.997511863708496
    },
    {
      "auxiliary_loss_clip": 0.0112035,
      "auxiliary_loss_mlp": 0.01053167,
      "balance_loss_clip": 1.0519309,
      "balance_loss_mlp": 1.03480864,
      "epoch": 0.19395761310686907,
      "flos": 25812554376960.0,
      "grad_norm": 2.42331686427639,
      "language_loss": 0.69379079,
      "learning_rate": 3.7245708116135585e-06,
      "loss": 0.71552593,
      "num_input_tokens_seen": 69638200,
      "step": 3226,
      "time_per_iteration": 2.746338129043579
    },
    {
      "auxiliary_loss_clip": 0.01126783,
      "auxiliary_loss_mlp": 0.01042276,
      "balance_loss_clip": 1.05692983,
      "balance_loss_mlp": 1.02264214,
      "epoch": 0.19401773635953706,
      "flos": 23039604334080.0,
      "grad_norm": 2.1006513764454864,
      "language_loss": 0.76236808,
      "learning_rate": 3.7243735467083193e-06,
      "loss": 0.78405869,
      "num_input_tokens_seen": 69657550,
      "step": 3227,
      "time_per_iteration": 2.760087728500366
    },
    {
      "auxiliary_loss_clip": 0.01117794,
      "auxiliary_loss_mlp": 0.010438,
      "balance_loss_clip": 1.05304587,
      "balance_loss_mlp": 1.0256561,
      "epoch": 0.19407785961220503,
      "flos": 15920780705280.0,
      "grad_norm": 2.8268368707906397,
      "language_loss": 0.69577461,
      "learning_rate": 3.724176216414662e-06,
      "loss": 0.71739054,
      "num_input_tokens_seen": 69675005,
      "step": 3228,
      "time_per_iteration": 2.6779348850250244
    },
    {
      "auxiliary_loss_clip": 0.01148199,
      "auxiliary_loss_mlp": 0.01042315,
      "balance_loss_clip": 1.05775642,
      "balance_loss_mlp": 1.02445757,
      "epoch": 0.194137982864873,
      "flos": 25921722787200.0,
      "grad_norm": 1.7694943420266864,
      "language_loss": 0.74160898,
      "learning_rate": 3.72397882074007e-06,
      "loss": 0.76351416,
      "num_input_tokens_seen": 69696455,
      "step": 3229,
      "time_per_iteration": 2.7229623794555664
    },
    {
      "auxiliary_loss_clip": 0.01119678,
      "auxiliary_loss_mlp": 0.01044155,
      "balance_loss_clip": 1.05435359,
      "balance_loss_mlp": 1.0262022,
      "epoch": 0.19419810611754096,
      "flos": 13261344618240.0,
      "grad_norm": 1.9766126324167548,
      "language_loss": 0.65722096,
      "learning_rate": 3.7237813596920285e-06,
      "loss": 0.67885935,
      "num_input_tokens_seen": 69714245,
      "step": 3230,
      "time_per_iteration": 2.740324020385742
    },
    {
      "auxiliary_loss_clip": 0.01124671,
      "auxiliary_loss_mlp": 0.00776003,
      "balance_loss_clip": 1.05223823,
      "balance_loss_mlp": 1.00081468,
      "epoch": 0.19425822937020892,
      "flos": 15705568368000.0,
      "grad_norm": 1.9307338208311895,
      "language_loss": 0.82042694,
      "learning_rate": 3.7235838332780254e-06,
      "loss": 0.83943367,
      "num_input_tokens_seen": 69731515,
      "step": 3231,
      "time_per_iteration": 2.7453513145446777
    },
    {
      "auxiliary_loss_clip": 0.0113141,
      "auxiliary_loss_mlp": 0.01042332,
      "balance_loss_clip": 1.05393946,
      "balance_loss_mlp": 1.02220988,
      "epoch": 0.1943183526228769,
      "flos": 23105392093440.0,
      "grad_norm": 10.866686758212083,
      "language_loss": 0.87038374,
      "learning_rate": 3.72338624150555e-06,
      "loss": 0.89212114,
      "num_input_tokens_seen": 69748885,
      "step": 3232,
      "time_per_iteration": 2.7575178146362305
    },
    {
      "auxiliary_loss_clip": 0.01100451,
      "auxiliary_loss_mlp": 0.01050878,
      "balance_loss_clip": 1.05029583,
      "balance_loss_mlp": 1.03102958,
      "epoch": 0.19437847587554485,
      "flos": 24712610146560.0,
      "grad_norm": 2.531838729905544,
      "language_loss": 0.85189134,
      "learning_rate": 3.723188584382096e-06,
      "loss": 0.87340462,
      "num_input_tokens_seen": 69767540,
      "step": 3233,
      "time_per_iteration": 2.8617444038391113
    },
    {
      "auxiliary_loss_clip": 0.01149478,
      "auxiliary_loss_mlp": 0.01054519,
      "balance_loss_clip": 1.0574832,
      "balance_loss_mlp": 1.0357672,
      "epoch": 0.19443859912821285,
      "flos": 23116130259840.0,
      "grad_norm": 1.7408859410354203,
      "language_loss": 0.89099532,
      "learning_rate": 3.722990861915158e-06,
      "loss": 0.91303527,
      "num_input_tokens_seen": 69789340,
      "step": 3234,
      "time_per_iteration": 2.7648239135742188
    },
    {
      "auxiliary_loss_clip": 0.01135157,
      "auxiliary_loss_mlp": 0.01044708,
      "balance_loss_clip": 1.05003643,
      "balance_loss_mlp": 1.02544403,
      "epoch": 0.1944987223808808,
      "flos": 15084385539840.0,
      "grad_norm": 2.4074482975555926,
      "language_loss": 0.78673434,
      "learning_rate": 3.722793074112234e-06,
      "loss": 0.80853301,
      "num_input_tokens_seen": 69806470,
      "step": 3235,
      "time_per_iteration": 2.76930832862854
    },
    {
      "auxiliary_loss_clip": 0.01136497,
      "auxiliary_loss_mlp": 0.01046749,
      "balance_loss_clip": 1.0580672,
      "balance_loss_mlp": 1.0293448,
      "epoch": 0.19455884563354878,
      "flos": 17126876603520.0,
      "grad_norm": 2.2511193258734354,
      "language_loss": 0.79391634,
      "learning_rate": 3.7225952209808233e-06,
      "loss": 0.81574875,
      "num_input_tokens_seen": 69822655,
      "step": 3236,
      "time_per_iteration": 2.7060179710388184
    },
    {
      "auxiliary_loss_clip": 0.01156991,
      "auxiliary_loss_mlp": 0.01044638,
      "balance_loss_clip": 1.05862045,
      "balance_loss_mlp": 1.02482522,
      "epoch": 0.19461896888621674,
      "flos": 20193396503040.0,
      "grad_norm": 2.1553329609131713,
      "language_loss": 0.76224017,
      "learning_rate": 3.72239730252843e-06,
      "loss": 0.78425646,
      "num_input_tokens_seen": 69841895,
      "step": 3237,
      "time_per_iteration": 2.642235040664673
    },
    {
      "auxiliary_loss_clip": 0.01158804,
      "auxiliary_loss_mlp": 0.01051059,
      "balance_loss_clip": 1.05648041,
      "balance_loss_mlp": 1.03289127,
      "epoch": 0.1946790921388847,
      "flos": 25301365971840.0,
      "grad_norm": 1.5204653275468003,
      "language_loss": 0.74828202,
      "learning_rate": 3.7221993187625583e-06,
      "loss": 0.77038062,
      "num_input_tokens_seen": 69862220,
      "step": 3238,
      "time_per_iteration": 2.6618688106536865
    },
    {
      "auxiliary_loss_clip": 0.01108331,
      "auxiliary_loss_mlp": 0.01046572,
      "balance_loss_clip": 1.04992437,
      "balance_loss_mlp": 1.02791595,
      "epoch": 0.19473921539155267,
      "flos": 20193396503040.0,
      "grad_norm": 3.1324225641798518,
      "language_loss": 0.734164,
      "learning_rate": 3.7220012696907155e-06,
      "loss": 0.75571299,
      "num_input_tokens_seen": 69881830,
      "step": 3239,
      "time_per_iteration": 2.7637152671813965
    },
    {
      "auxiliary_loss_clip": 0.01132567,
      "auxiliary_loss_mlp": 0.01047988,
      "balance_loss_clip": 1.05458641,
      "balance_loss_mlp": 1.02947509,
      "epoch": 0.19479933864422067,
      "flos": 20887549810560.0,
      "grad_norm": 2.155392951393246,
      "language_loss": 0.73291272,
      "learning_rate": 3.721803155320412e-06,
      "loss": 0.7547183,
      "num_input_tokens_seen": 69900515,
      "step": 3240,
      "time_per_iteration": 2.6980888843536377
    },
    {
      "auxiliary_loss_clip": 0.01131601,
      "auxiliary_loss_mlp": 0.0103943,
      "balance_loss_clip": 1.05846488,
      "balance_loss_mlp": 1.02208555,
      "epoch": 0.19485946189688863,
      "flos": 23295072839040.0,
      "grad_norm": 5.847648280625993,
      "language_loss": 0.65809447,
      "learning_rate": 3.7216049756591606e-06,
      "loss": 0.6798048,
      "num_input_tokens_seen": 69920060,
      "step": 3241,
      "time_per_iteration": 2.659707546234131
    },
    {
      "auxiliary_loss_clip": 0.01128971,
      "auxiliary_loss_mlp": 0.01048707,
      "balance_loss_clip": 1.05226684,
      "balance_loss_mlp": 1.03039646,
      "epoch": 0.1949195851495566,
      "flos": 23295036925440.0,
      "grad_norm": 1.4408225707306088,
      "language_loss": 0.82747853,
      "learning_rate": 3.7214067307144754e-06,
      "loss": 0.84925532,
      "num_input_tokens_seen": 69939820,
      "step": 3242,
      "time_per_iteration": 2.7137632369995117
    },
    {
      "auxiliary_loss_clip": 0.01077632,
      "auxiliary_loss_mlp": 0.01014225,
      "balance_loss_clip": 1.04083347,
      "balance_loss_mlp": 1.01131678,
      "epoch": 0.19497970840222456,
      "flos": 64962871557120.0,
      "grad_norm": 0.853263603243422,
      "language_loss": 0.57500821,
      "learning_rate": 3.721208420493875e-06,
      "loss": 0.59592682,
      "num_input_tokens_seen": 70002145,
      "step": 3243,
      "time_per_iteration": 3.1446309089660645
    },
    {
      "auxiliary_loss_clip": 0.01138548,
      "auxiliary_loss_mlp": 0.01050428,
      "balance_loss_clip": 1.05331421,
      "balance_loss_mlp": 1.02988815,
      "epoch": 0.19503983165489253,
      "flos": 19644717277440.0,
      "grad_norm": 7.2345723863132,
      "language_loss": 0.83789021,
      "learning_rate": 3.7210100450048784e-06,
      "loss": 0.85977995,
      "num_input_tokens_seen": 70020510,
      "step": 3244,
      "time_per_iteration": 2.6194229125976562
    },
    {
      "auxiliary_loss_clip": 0.01143261,
      "auxiliary_loss_mlp": 0.01046223,
      "balance_loss_clip": 1.05732584,
      "balance_loss_mlp": 1.02869976,
      "epoch": 0.1950999549075605,
      "flos": 21141976821120.0,
      "grad_norm": 2.0710390949438837,
      "language_loss": 0.7739507,
      "learning_rate": 3.7208116042550088e-06,
      "loss": 0.79584551,
      "num_input_tokens_seen": 70040760,
      "step": 3245,
      "time_per_iteration": 2.6684374809265137
    },
    {
      "auxiliary_loss_clip": 0.01142874,
      "auxiliary_loss_mlp": 0.01043114,
      "balance_loss_clip": 1.05566645,
      "balance_loss_mlp": 1.02431464,
      "epoch": 0.19516007816022846,
      "flos": 20884820376960.0,
      "grad_norm": 2.1010289547443133,
      "language_loss": 0.83988321,
      "learning_rate": 3.7206130982517906e-06,
      "loss": 0.86174309,
      "num_input_tokens_seen": 70058720,
      "step": 3246,
      "time_per_iteration": 2.6595354080200195
    },
    {
      "auxiliary_loss_clip": 0.0114599,
      "auxiliary_loss_mlp": 0.00776442,
      "balance_loss_clip": 1.05517101,
      "balance_loss_mlp": 1.00080454,
      "epoch": 0.19522020141289645,
      "flos": 16910515031040.0,
      "grad_norm": 3.3581015873305438,
      "language_loss": 0.76840878,
      "learning_rate": 3.7204145270027514e-06,
      "loss": 0.78763306,
      "num_input_tokens_seen": 70076470,
      "step": 3247,
      "time_per_iteration": 2.7777793407440186
    },
    {
      "auxiliary_loss_clip": 0.01121778,
      "auxiliary_loss_mlp": 0.01043977,
      "balance_loss_clip": 1.05689096,
      "balance_loss_mlp": 1.02651262,
      "epoch": 0.19528032466556441,
      "flos": 26724829023360.0,
      "grad_norm": 1.8981807103962522,
      "language_loss": 0.75459039,
      "learning_rate": 3.720215890515421e-06,
      "loss": 0.77624786,
      "num_input_tokens_seen": 70096220,
      "step": 3248,
      "time_per_iteration": 2.8088901042938232
    },
    {
      "auxiliary_loss_clip": 0.01156017,
      "auxiliary_loss_mlp": 0.01048303,
      "balance_loss_clip": 1.05548215,
      "balance_loss_mlp": 1.03008783,
      "epoch": 0.19534044791823238,
      "flos": 21032808410880.0,
      "grad_norm": 2.7209722336942135,
      "language_loss": 0.77774823,
      "learning_rate": 3.7200171887973316e-06,
      "loss": 0.79979146,
      "num_input_tokens_seen": 70114800,
      "step": 3249,
      "time_per_iteration": 2.610877752304077
    },
    {
      "auxiliary_loss_clip": 0.01148434,
      "auxiliary_loss_mlp": 0.01050332,
      "balance_loss_clip": 1.05689144,
      "balance_loss_mlp": 1.03299928,
      "epoch": 0.19540057117090034,
      "flos": 22344050396160.0,
      "grad_norm": 1.5551573885822045,
      "language_loss": 0.73118901,
      "learning_rate": 3.7198184218560176e-06,
      "loss": 0.75317669,
      "num_input_tokens_seen": 70134930,
      "step": 3250,
      "time_per_iteration": 2.5901567935943604
    },
    {
      "auxiliary_loss_clip": 0.01101628,
      "auxiliary_loss_mlp": 0.01046467,
      "balance_loss_clip": 1.05080378,
      "balance_loss_mlp": 1.02876413,
      "epoch": 0.1954606944235683,
      "flos": 20301631159680.0,
      "grad_norm": 2.030501302548557,
      "language_loss": 0.79203367,
      "learning_rate": 3.719619589699017e-06,
      "loss": 0.81351459,
      "num_input_tokens_seen": 70152045,
      "step": 3251,
      "time_per_iteration": 2.6619749069213867
    },
    {
      "auxiliary_loss_clip": 0.0115825,
      "auxiliary_loss_mlp": 0.01044132,
      "balance_loss_clip": 1.05741858,
      "balance_loss_mlp": 1.02606022,
      "epoch": 0.19552081767623627,
      "flos": 17346865449600.0,
      "grad_norm": 7.451515078679223,
      "language_loss": 0.83871722,
      "learning_rate": 3.7194206923338695e-06,
      "loss": 0.86074108,
      "num_input_tokens_seen": 70169240,
      "step": 3252,
      "time_per_iteration": 2.5029656887054443
    },
    {
      "auxiliary_loss_clip": 0.01142752,
      "auxiliary_loss_mlp": 0.01057294,
      "balance_loss_clip": 1.05278862,
      "balance_loss_mlp": 1.03518057,
      "epoch": 0.19558094092890424,
      "flos": 31977626129280.0,
      "grad_norm": 1.7140417843701068,
      "language_loss": 0.73995864,
      "learning_rate": 3.719221729768117e-06,
      "loss": 0.76195908,
      "num_input_tokens_seen": 70192690,
      "step": 3253,
      "time_per_iteration": 2.609117269515991
    },
    {
      "auxiliary_loss_clip": 0.01102675,
      "auxiliary_loss_mlp": 0.01046707,
      "balance_loss_clip": 1.04759037,
      "balance_loss_mlp": 1.02782381,
      "epoch": 0.19564106418157223,
      "flos": 22268889187200.0,
      "grad_norm": 2.1302159220485675,
      "language_loss": 0.76167047,
      "learning_rate": 3.7190227020093037e-06,
      "loss": 0.78316426,
      "num_input_tokens_seen": 70209685,
      "step": 3254,
      "time_per_iteration": 4.174965858459473
    },
    {
      "auxiliary_loss_clip": 0.01043127,
      "auxiliary_loss_mlp": 0.01006966,
      "balance_loss_clip": 1.04737842,
      "balance_loss_mlp": 1.0036757,
      "epoch": 0.1957011874342402,
      "flos": 54364554385920.0,
      "grad_norm": 0.84452007287803,
      "language_loss": 0.55275303,
      "learning_rate": 3.7188236090649774e-06,
      "loss": 0.57325399,
      "num_input_tokens_seen": 70265050,
      "step": 3255,
      "time_per_iteration": 3.2241716384887695
    },
    {
      "auxiliary_loss_clip": 0.01133721,
      "auxiliary_loss_mlp": 0.01041696,
      "balance_loss_clip": 1.0557251,
      "balance_loss_mlp": 1.02349281,
      "epoch": 0.19576131068690816,
      "flos": 16506699356160.0,
      "grad_norm": 2.6103802859468392,
      "language_loss": 0.70870697,
      "learning_rate": 3.718624450942688e-06,
      "loss": 0.73046112,
      "num_input_tokens_seen": 70281830,
      "step": 3256,
      "time_per_iteration": 2.641296148300171
    },
    {
      "auxiliary_loss_clip": 0.01152768,
      "auxiliary_loss_mlp": 0.01042867,
      "balance_loss_clip": 1.0544858,
      "balance_loss_mlp": 1.02523613,
      "epoch": 0.19582143393957613,
      "flos": 14719676797440.0,
      "grad_norm": 2.649319646209249,
      "language_loss": 0.80722409,
      "learning_rate": 3.718425227649987e-06,
      "loss": 0.82918048,
      "num_input_tokens_seen": 70297420,
      "step": 3257,
      "time_per_iteration": 4.258259057998657
    },
    {
      "auxiliary_loss_clip": 0.01106644,
      "auxiliary_loss_mlp": 0.01043385,
      "balance_loss_clip": 1.05470431,
      "balance_loss_mlp": 1.02601588,
      "epoch": 0.1958815571922441,
      "flos": 24425504737920.0,
      "grad_norm": 6.015808523610408,
      "language_loss": 0.75124931,
      "learning_rate": 3.7182259391944292e-06,
      "loss": 0.77274966,
      "num_input_tokens_seen": 70319210,
      "step": 3258,
      "time_per_iteration": 4.386433362960815
    },
    {
      "auxiliary_loss_clip": 0.01082287,
      "auxiliary_loss_mlp": 0.01044148,
      "balance_loss_clip": 1.04533339,
      "balance_loss_mlp": 1.0237875,
      "epoch": 0.19594168044491206,
      "flos": 24900279730560.0,
      "grad_norm": 1.8034996675319444,
      "language_loss": 0.73872411,
      "learning_rate": 3.7180265855835714e-06,
      "loss": 0.75998843,
      "num_input_tokens_seen": 70339045,
      "step": 3259,
      "time_per_iteration": 2.815469264984131
    },
    {
      "auxiliary_loss_clip": 0.01131793,
      "auxiliary_loss_mlp": 0.01043364,
      "balance_loss_clip": 1.05167735,
      "balance_loss_mlp": 1.02392125,
      "epoch": 0.19600180369758005,
      "flos": 12057008486400.0,
      "grad_norm": 2.2096667980592,
      "language_loss": 0.77053022,
      "learning_rate": 3.7178271668249735e-06,
      "loss": 0.79228187,
      "num_input_tokens_seen": 70356505,
      "step": 3260,
      "time_per_iteration": 4.2817702293396
    },
    {
      "auxiliary_loss_clip": 0.01148118,
      "auxiliary_loss_mlp": 0.01043761,
      "balance_loss_clip": 1.0551343,
      "balance_loss_mlp": 1.0248661,
      "epoch": 0.19606192695024802,
      "flos": 20850202644480.0,
      "grad_norm": 5.605178759176999,
      "language_loss": 0.82261205,
      "learning_rate": 3.7176276829261975e-06,
      "loss": 0.84453082,
      "num_input_tokens_seen": 70375410,
      "step": 3261,
      "time_per_iteration": 2.673092842102051
    },
    {
      "auxiliary_loss_clip": 0.01121379,
      "auxiliary_loss_mlp": 0.01044043,
      "balance_loss_clip": 1.0550617,
      "balance_loss_mlp": 1.02488637,
      "epoch": 0.19612205020291598,
      "flos": 28475509996800.0,
      "grad_norm": 1.8492209450679535,
      "language_loss": 0.76671481,
      "learning_rate": 3.717428133894807e-06,
      "loss": 0.78836906,
      "num_input_tokens_seen": 70396315,
      "step": 3262,
      "time_per_iteration": 2.803938150405884
    },
    {
      "auxiliary_loss_clip": 0.01148893,
      "auxiliary_loss_mlp": 0.01047259,
      "balance_loss_clip": 1.05960584,
      "balance_loss_mlp": 1.02950907,
      "epoch": 0.19618217345558395,
      "flos": 25556618995200.0,
      "grad_norm": 1.7278621785184562,
      "language_loss": 0.8668195,
      "learning_rate": 3.71722851973837e-06,
      "loss": 0.88878107,
      "num_input_tokens_seen": 70417945,
      "step": 3263,
      "time_per_iteration": 2.6677918434143066
    },
    {
      "auxiliary_loss_clip": 0.0113123,
      "auxiliary_loss_mlp": 0.01042546,
      "balance_loss_clip": 1.05328059,
      "balance_loss_mlp": 1.02505815,
      "epoch": 0.1962422967082519,
      "flos": 25264413855360.0,
      "grad_norm": 3.447639973868791,
      "language_loss": 0.73775035,
      "learning_rate": 3.717028840464455e-06,
      "loss": 0.75948811,
      "num_input_tokens_seen": 70438690,
      "step": 3264,
      "time_per_iteration": 2.6973094940185547
    },
    {
      "auxiliary_loss_clip": 0.01144053,
      "auxiliary_loss_mlp": 0.01049918,
      "balance_loss_clip": 1.05736756,
      "balance_loss_mlp": 1.03223944,
      "epoch": 0.19630241996091988,
      "flos": 18807352444800.0,
      "grad_norm": 2.4424358562200927,
      "language_loss": 0.78513813,
      "learning_rate": 3.7168290960806344e-06,
      "loss": 0.80707777,
      "num_input_tokens_seen": 70455385,
      "step": 3265,
      "time_per_iteration": 2.625739336013794
    },
    {
      "auxiliary_loss_clip": 0.01031434,
      "auxiliary_loss_mlp": 0.01002481,
      "balance_loss_clip": 1.03386986,
      "balance_loss_mlp": 0.99983466,
      "epoch": 0.19636254321358784,
      "flos": 62321137896960.0,
      "grad_norm": 0.7932330660809486,
      "language_loss": 0.53389955,
      "learning_rate": 3.716629286594483e-06,
      "loss": 0.55423868,
      "num_input_tokens_seen": 70514280,
      "step": 3266,
      "time_per_iteration": 3.2586586475372314
    },
    {
      "auxiliary_loss_clip": 0.01124628,
      "auxiliary_loss_mlp": 0.00776501,
      "balance_loss_clip": 1.04957044,
      "balance_loss_mlp": 1.00080895,
      "epoch": 0.19642266646625584,
      "flos": 21069329564160.0,
      "grad_norm": 2.0008611208986133,
      "language_loss": 0.80109024,
      "learning_rate": 3.7164294120135767e-06,
      "loss": 0.8201015,
      "num_input_tokens_seen": 70531800,
      "step": 3267,
      "time_per_iteration": 2.678537368774414
    },
    {
      "auxiliary_loss_clip": 0.01130982,
      "auxiliary_loss_mlp": 0.01043983,
      "balance_loss_clip": 1.05263019,
      "balance_loss_mlp": 1.02660179,
      "epoch": 0.1964827897189238,
      "flos": 14538651229440.0,
      "grad_norm": 1.9909459598185588,
      "language_loss": 0.86758262,
      "learning_rate": 3.7162294723454953e-06,
      "loss": 0.88933229,
      "num_input_tokens_seen": 70550615,
      "step": 3268,
      "time_per_iteration": 2.6949849128723145
    },
    {
      "auxiliary_loss_clip": 0.01099432,
      "auxiliary_loss_mlp": 0.01041621,
      "balance_loss_clip": 1.04954004,
      "balance_loss_mlp": 1.02408528,
      "epoch": 0.19654291297159177,
      "flos": 19244636616960.0,
      "grad_norm": 2.2632495429204127,
      "language_loss": 0.68785441,
      "learning_rate": 3.7160294675978197e-06,
      "loss": 0.70926493,
      "num_input_tokens_seen": 70568690,
      "step": 3269,
      "time_per_iteration": 2.770078182220459
    },
    {
      "auxiliary_loss_clip": 0.01116538,
      "auxiliary_loss_mlp": 0.01052319,
      "balance_loss_clip": 1.05113554,
      "balance_loss_mlp": 1.03330541,
      "epoch": 0.19660303622425973,
      "flos": 25775710001280.0,
      "grad_norm": 7.1863103423452355,
      "language_loss": 0.80241841,
      "learning_rate": 3.715829397778135e-06,
      "loss": 0.82410699,
      "num_input_tokens_seen": 70588665,
      "step": 3270,
      "time_per_iteration": 2.7294864654541016
    },
    {
      "auxiliary_loss_clip": 0.01139501,
      "auxiliary_loss_mlp": 0.01045694,
      "balance_loss_clip": 1.05189824,
      "balance_loss_mlp": 1.02833724,
      "epoch": 0.1966631594769277,
      "flos": 20595093275520.0,
      "grad_norm": 1.9668649321541274,
      "language_loss": 0.83912349,
      "learning_rate": 3.715629262894028e-06,
      "loss": 0.86097538,
      "num_input_tokens_seen": 70606900,
      "step": 3271,
      "time_per_iteration": 2.640235662460327
    },
    {
      "auxiliary_loss_clip": 0.01139368,
      "auxiliary_loss_mlp": 0.01051303,
      "balance_loss_clip": 1.05468225,
      "balance_loss_mlp": 1.0332067,
      "epoch": 0.19672328272959566,
      "flos": 23623188600960.0,
      "grad_norm": 1.9968416702279483,
      "language_loss": 0.79902714,
      "learning_rate": 3.715429062953087e-06,
      "loss": 0.82093388,
      "num_input_tokens_seen": 70625955,
      "step": 3272,
      "time_per_iteration": 2.636629343032837
    },
    {
      "auxiliary_loss_clip": 0.01124328,
      "auxiliary_loss_mlp": 0.01058493,
      "balance_loss_clip": 1.05192566,
      "balance_loss_mlp": 1.03715479,
      "epoch": 0.19678340598226365,
      "flos": 23110922787840.0,
      "grad_norm": 1.7302013075823783,
      "language_loss": 0.80942369,
      "learning_rate": 3.7152287979629043e-06,
      "loss": 0.83125186,
      "num_input_tokens_seen": 70646090,
      "step": 3273,
      "time_per_iteration": 2.6967809200286865
    },
    {
      "auxiliary_loss_clip": 0.01144024,
      "auxiliary_loss_mlp": 0.01054564,
      "balance_loss_clip": 1.05456042,
      "balance_loss_mlp": 1.03655195,
      "epoch": 0.19684352923493162,
      "flos": 24534852716160.0,
      "grad_norm": 2.225126358921887,
      "language_loss": 0.77984649,
      "learning_rate": 3.7150284679310735e-06,
      "loss": 0.80183232,
      "num_input_tokens_seen": 70666065,
      "step": 3274,
      "time_per_iteration": 2.6808643341064453
    },
    {
      "auxiliary_loss_clip": 0.01141267,
      "auxiliary_loss_mlp": 0.01046445,
      "balance_loss_clip": 1.05480242,
      "balance_loss_mlp": 1.02840877,
      "epoch": 0.19690365248759958,
      "flos": 21796448578560.0,
      "grad_norm": 2.318697297640889,
      "language_loss": 0.81433225,
      "learning_rate": 3.7148280728651914e-06,
      "loss": 0.8362093,
      "num_input_tokens_seen": 70681580,
      "step": 3275,
      "time_per_iteration": 2.672672986984253
    },
    {
      "auxiliary_loss_clip": 0.01115756,
      "auxiliary_loss_mlp": 0.01045314,
      "balance_loss_clip": 1.05148947,
      "balance_loss_mlp": 1.02686024,
      "epoch": 0.19696377574026755,
      "flos": 19056643810560.0,
      "grad_norm": 2.4665004531377166,
      "language_loss": 0.80909657,
      "learning_rate": 3.7146276127728563e-06,
      "loss": 0.83070731,
      "num_input_tokens_seen": 70697745,
      "step": 3276,
      "time_per_iteration": 2.726970672607422
    },
    {
      "auxiliary_loss_clip": 0.01142619,
      "auxiliary_loss_mlp": 0.01043042,
      "balance_loss_clip": 1.05443609,
      "balance_loss_mlp": 1.02491045,
      "epoch": 0.19702389899293551,
      "flos": 22820656982400.0,
      "grad_norm": 2.17541075016206,
      "language_loss": 0.89113599,
      "learning_rate": 3.7144270876616713e-06,
      "loss": 0.9129926,
      "num_input_tokens_seen": 70715110,
      "step": 3277,
      "time_per_iteration": 2.6738827228546143
    },
    {
      "auxiliary_loss_clip": 0.01103709,
      "auxiliary_loss_mlp": 0.01048433,
      "balance_loss_clip": 1.04638815,
      "balance_loss_mlp": 1.02864444,
      "epoch": 0.19708402224560348,
      "flos": 22894237992960.0,
      "grad_norm": 2.640727897616601,
      "language_loss": 0.62070847,
      "learning_rate": 3.714226497539239e-06,
      "loss": 0.64222991,
      "num_input_tokens_seen": 70734715,
      "step": 3278,
      "time_per_iteration": 2.7382938861846924
    },
    {
      "auxiliary_loss_clip": 0.01115303,
      "auxiliary_loss_mlp": 0.0105759,
      "balance_loss_clip": 1.05033016,
      "balance_loss_mlp": 1.03793263,
      "epoch": 0.19714414549827144,
      "flos": 25662519267840.0,
      "grad_norm": 1.930104581155035,
      "language_loss": 0.73606467,
      "learning_rate": 3.714025842413166e-06,
      "loss": 0.75779366,
      "num_input_tokens_seen": 70752650,
      "step": 3279,
      "time_per_iteration": 2.8123648166656494
    },
    {
      "auxiliary_loss_clip": 0.0114648,
      "auxiliary_loss_mlp": 0.01042853,
      "balance_loss_clip": 1.05422091,
      "balance_loss_mlp": 1.02567458,
      "epoch": 0.19720426875093944,
      "flos": 23915824704000.0,
      "grad_norm": 1.7034036878345749,
      "language_loss": 0.82685816,
      "learning_rate": 3.713825122291061e-06,
      "loss": 0.84875143,
      "num_input_tokens_seen": 70772365,
      "step": 3280,
      "time_per_iteration": 2.7000861167907715
    },
    {
      "auxiliary_loss_clip": 0.01106655,
      "auxiliary_loss_mlp": 0.01048884,
      "balance_loss_clip": 1.04887283,
      "balance_loss_mlp": 1.03071654,
      "epoch": 0.1972643920036074,
      "flos": 13881952828800.0,
      "grad_norm": 2.435959864664923,
      "language_loss": 0.78173983,
      "learning_rate": 3.713624337180536e-06,
      "loss": 0.80329525,
      "num_input_tokens_seen": 70790340,
      "step": 3281,
      "time_per_iteration": 2.7017247676849365
    },
    {
      "auxiliary_loss_clip": 0.01125353,
      "auxiliary_loss_mlp": 0.0104135,
      "balance_loss_clip": 1.05461836,
      "balance_loss_mlp": 1.02519727,
      "epoch": 0.19732451525627537,
      "flos": 19863592801920.0,
      "grad_norm": 1.7390973872526612,
      "language_loss": 0.79777479,
      "learning_rate": 3.7134234870892045e-06,
      "loss": 0.8194418,
      "num_input_tokens_seen": 70809295,
      "step": 3282,
      "time_per_iteration": 2.7064146995544434
    },
    {
      "auxiliary_loss_clip": 0.01112073,
      "auxiliary_loss_mlp": 0.01043047,
      "balance_loss_clip": 1.05485284,
      "balance_loss_mlp": 1.02538049,
      "epoch": 0.19738463850894333,
      "flos": 24973429777920.0,
      "grad_norm": 2.512566515566025,
      "language_loss": 0.7192747,
      "learning_rate": 3.7132225720246826e-06,
      "loss": 0.74082589,
      "num_input_tokens_seen": 70828765,
      "step": 3283,
      "time_per_iteration": 2.775297164916992
    },
    {
      "auxiliary_loss_clip": 0.01137498,
      "auxiliary_loss_mlp": 0.01043438,
      "balance_loss_clip": 1.05320621,
      "balance_loss_mlp": 1.02665281,
      "epoch": 0.1974447617616113,
      "flos": 18368883123840.0,
      "grad_norm": 1.8864815757917637,
      "language_loss": 0.78981179,
      "learning_rate": 3.7130215919945886e-06,
      "loss": 0.81162113,
      "num_input_tokens_seen": 70846805,
      "step": 3284,
      "time_per_iteration": 2.6344916820526123
    },
    {
      "auxiliary_loss_clip": 0.01126512,
      "auxiliary_loss_mlp": 0.00776821,
      "balance_loss_clip": 1.05065584,
      "balance_loss_mlp": 1.00114048,
      "epoch": 0.19750488501427926,
      "flos": 22892945103360.0,
      "grad_norm": 2.1903874509936982,
      "language_loss": 0.86317503,
      "learning_rate": 3.7128205470065445e-06,
      "loss": 0.88220835,
      "num_input_tokens_seen": 70863805,
      "step": 3285,
      "time_per_iteration": 2.725186586380005
    },
    {
      "auxiliary_loss_clip": 0.01115791,
      "auxiliary_loss_mlp": 0.01044707,
      "balance_loss_clip": 1.05167055,
      "balance_loss_mlp": 1.02658761,
      "epoch": 0.19756500826694723,
      "flos": 21871502046720.0,
      "grad_norm": 2.208260347555195,
      "language_loss": 0.88770825,
      "learning_rate": 3.712619437068174e-06,
      "loss": 0.90931326,
      "num_input_tokens_seen": 70882660,
      "step": 3286,
      "time_per_iteration": 2.6819698810577393
    },
    {
      "auxiliary_loss_clip": 0.01118742,
      "auxiliary_loss_mlp": 0.01052526,
      "balance_loss_clip": 1.05227792,
      "balance_loss_mlp": 1.03016233,
      "epoch": 0.19762513151961522,
      "flos": 15158972131200.0,
      "grad_norm": 2.0768117117784874,
      "language_loss": 0.77941382,
      "learning_rate": 3.712418262187102e-06,
      "loss": 0.80112648,
      "num_input_tokens_seen": 70898765,
      "step": 3287,
      "time_per_iteration": 2.641193389892578
    },
    {
      "auxiliary_loss_clip": 0.01127955,
      "auxiliary_loss_mlp": 0.01047337,
      "balance_loss_clip": 1.0526104,
      "balance_loss_mlp": 1.02849019,
      "epoch": 0.1976852547722832,
      "flos": 16979175878400.0,
      "grad_norm": 2.061421898899755,
      "language_loss": 0.80853081,
      "learning_rate": 3.7122170223709584e-06,
      "loss": 0.83028376,
      "num_input_tokens_seen": 70916370,
      "step": 3288,
      "time_per_iteration": 2.625068426132202
    },
    {
      "auxiliary_loss_clip": 0.01132408,
      "auxiliary_loss_mlp": 0.01048194,
      "balance_loss_clip": 1.05143857,
      "balance_loss_mlp": 1.03045535,
      "epoch": 0.19774537802495115,
      "flos": 20302924049280.0,
      "grad_norm": 2.345717890688315,
      "language_loss": 0.7317158,
      "learning_rate": 3.712015717627374e-06,
      "loss": 0.75352174,
      "num_input_tokens_seen": 70934870,
      "step": 3289,
      "time_per_iteration": 2.6319406032562256
    },
    {
      "auxiliary_loss_clip": 0.01133413,
      "auxiliary_loss_mlp": 0.01045224,
      "balance_loss_clip": 1.05575252,
      "balance_loss_mlp": 1.02678204,
      "epoch": 0.19780550127761912,
      "flos": 27235478724480.0,
      "grad_norm": 1.9087552003653308,
      "language_loss": 0.79608113,
      "learning_rate": 3.7118143479639813e-06,
      "loss": 0.81786746,
      "num_input_tokens_seen": 70955140,
      "step": 3290,
      "time_per_iteration": 2.706570863723755
    },
    {
      "auxiliary_loss_clip": 0.01049926,
      "auxiliary_loss_mlp": 0.0101105,
      "balance_loss_clip": 1.0327636,
      "balance_loss_mlp": 1.00853467,
      "epoch": 0.19786562453028708,
      "flos": 63550972684800.0,
      "grad_norm": 0.8952067644857119,
      "language_loss": 0.60318571,
      "learning_rate": 3.711612913388418e-06,
      "loss": 0.62379545,
      "num_input_tokens_seen": 71012005,
      "step": 3291,
      "time_per_iteration": 3.2849009037017822
    },
    {
      "auxiliary_loss_clip": 0.01158891,
      "auxiliary_loss_mlp": 0.01040785,
      "balance_loss_clip": 1.05417156,
      "balance_loss_mlp": 1.02088892,
      "epoch": 0.19792574778295505,
      "flos": 26286647011200.0,
      "grad_norm": 1.932789926440358,
      "language_loss": 0.81595641,
      "learning_rate": 3.7114114139083204e-06,
      "loss": 0.83795315,
      "num_input_tokens_seen": 71031140,
      "step": 3292,
      "time_per_iteration": 2.6751551628112793
    },
    {
      "auxiliary_loss_clip": 0.01119797,
      "auxiliary_loss_mlp": 0.00778082,
      "balance_loss_clip": 1.05296063,
      "balance_loss_mlp": 1.00086236,
      "epoch": 0.19798587103562304,
      "flos": 19938107566080.0,
      "grad_norm": 2.409042629875397,
      "language_loss": 0.81013,
      "learning_rate": 3.7112098495313313e-06,
      "loss": 0.82910883,
      "num_input_tokens_seen": 71050250,
      "step": 3293,
      "time_per_iteration": 4.3039703369140625
    },
    {
      "auxiliary_loss_clip": 0.01137316,
      "auxiliary_loss_mlp": 0.01052434,
      "balance_loss_clip": 1.05370128,
      "balance_loss_mlp": 1.03277683,
      "epoch": 0.198045994288291,
      "flos": 20120282369280.0,
      "grad_norm": 1.8764131105986912,
      "language_loss": 0.61480314,
      "learning_rate": 3.711008220265093e-06,
      "loss": 0.63670063,
      "num_input_tokens_seen": 71068665,
      "step": 3294,
      "time_per_iteration": 2.671241044998169
    },
    {
      "auxiliary_loss_clip": 0.01132208,
      "auxiliary_loss_mlp": 0.01039978,
      "balance_loss_clip": 1.05456376,
      "balance_loss_mlp": 1.02201271,
      "epoch": 0.19810611754095897,
      "flos": 17967653228160.0,
      "grad_norm": 2.0334748560156393,
      "language_loss": 0.87313825,
      "learning_rate": 3.710806526117251e-06,
      "loss": 0.89486015,
      "num_input_tokens_seen": 71085320,
      "step": 3295,
      "time_per_iteration": 2.659680128097534
    },
    {
      "auxiliary_loss_clip": 0.01113106,
      "auxiliary_loss_mlp": 0.01050184,
      "balance_loss_clip": 1.05079484,
      "balance_loss_mlp": 1.03256536,
      "epoch": 0.19816624079362694,
      "flos": 15084996071040.0,
      "grad_norm": 2.5215255479345067,
      "language_loss": 0.80839241,
      "learning_rate": 3.7106047670954544e-06,
      "loss": 0.83002532,
      "num_input_tokens_seen": 71102020,
      "step": 3296,
      "time_per_iteration": 4.299339294433594
    },
    {
      "auxiliary_loss_clip": 0.01123906,
      "auxiliary_loss_mlp": 0.01045438,
      "balance_loss_clip": 1.05233586,
      "balance_loss_mlp": 1.02522039,
      "epoch": 0.1982263640462949,
      "flos": 24900315644160.0,
      "grad_norm": 2.528943220563754,
      "language_loss": 0.68126047,
      "learning_rate": 3.710402943207354e-06,
      "loss": 0.70295388,
      "num_input_tokens_seen": 71123390,
      "step": 3297,
      "time_per_iteration": 4.258284091949463
    },
    {
      "auxiliary_loss_clip": 0.01153129,
      "auxiliary_loss_mlp": 0.01037574,
      "balance_loss_clip": 1.05660713,
      "balance_loss_mlp": 1.02031219,
      "epoch": 0.19828648729896287,
      "flos": 20376181837440.0,
      "grad_norm": 1.9083451106828888,
      "language_loss": 0.81310993,
      "learning_rate": 3.7102010544606016e-06,
      "loss": 0.83501697,
      "num_input_tokens_seen": 71141800,
      "step": 3298,
      "time_per_iteration": 2.6156656742095947
    },
    {
      "auxiliary_loss_clip": 0.01137409,
      "auxiliary_loss_mlp": 0.01042227,
      "balance_loss_clip": 1.0573976,
      "balance_loss_mlp": 1.02159238,
      "epoch": 0.19834661055163083,
      "flos": 18880035615360.0,
      "grad_norm": 1.8996943203321497,
      "language_loss": 0.85154539,
      "learning_rate": 3.7099991008628544e-06,
      "loss": 0.87334174,
      "num_input_tokens_seen": 71159505,
      "step": 3299,
      "time_per_iteration": 2.6749041080474854
    },
    {
      "auxiliary_loss_clip": 0.01036953,
      "auxiliary_loss_mlp": 0.01013935,
      "balance_loss_clip": 1.02875936,
      "balance_loss_mlp": 1.01106215,
      "epoch": 0.19840673380429882,
      "flos": 60259184640000.0,
      "grad_norm": 0.82907550606663,
      "language_loss": 0.53206414,
      "learning_rate": 3.7097970824217706e-06,
      "loss": 0.55257303,
      "num_input_tokens_seen": 71223265,
      "step": 3300,
      "time_per_iteration": 4.83857798576355
    },
    {
      "auxiliary_loss_clip": 0.01105122,
      "auxiliary_loss_mlp": 0.01064471,
      "balance_loss_clip": 1.04748702,
      "balance_loss_mlp": 1.0410459,
      "epoch": 0.1984668570569668,
      "flos": 19902017376000.0,
      "grad_norm": 316.1702389408657,
      "language_loss": 0.73014295,
      "learning_rate": 3.7095949991450093e-06,
      "loss": 0.75183886,
      "num_input_tokens_seen": 71242385,
      "step": 3301,
      "time_per_iteration": 2.700654983520508
    },
    {
      "auxiliary_loss_clip": 0.01118926,
      "auxiliary_loss_mlp": 0.01044315,
      "balance_loss_clip": 1.05295372,
      "balance_loss_mlp": 1.02619529,
      "epoch": 0.19852698030963475,
      "flos": 15630766295040.0,
      "grad_norm": 2.410718710355122,
      "language_loss": 0.88264418,
      "learning_rate": 3.709392851040235e-06,
      "loss": 0.90427655,
      "num_input_tokens_seen": 71258990,
      "step": 3302,
      "time_per_iteration": 2.7190146446228027
    },
    {
      "auxiliary_loss_clip": 0.01118067,
      "auxiliary_loss_mlp": 0.01045078,
      "balance_loss_clip": 1.05155802,
      "balance_loss_mlp": 1.02661204,
      "epoch": 0.19858710356230272,
      "flos": 43143007311360.0,
      "grad_norm": 2.210364764996701,
      "language_loss": 0.73592931,
      "learning_rate": 3.709190638115111e-06,
      "loss": 0.75756073,
      "num_input_tokens_seen": 71282770,
      "step": 3303,
      "time_per_iteration": 2.9379186630249023
    },
    {
      "auxiliary_loss_clip": 0.01143275,
      "auxiliary_loss_mlp": 0.01048515,
      "balance_loss_clip": 1.05491257,
      "balance_loss_mlp": 1.03002524,
      "epoch": 0.19864722681497068,
      "flos": 35144084643840.0,
      "grad_norm": 1.9482807590384623,
      "language_loss": 0.75103521,
      "learning_rate": 3.7089883603773084e-06,
      "loss": 0.77295315,
      "num_input_tokens_seen": 71301410,
      "step": 3304,
      "time_per_iteration": 2.743474245071411
    },
    {
      "auxiliary_loss_clip": 0.01133571,
      "auxiliary_loss_mlp": 0.01034983,
      "balance_loss_clip": 1.05309725,
      "balance_loss_mlp": 1.01710188,
      "epoch": 0.19870735006763865,
      "flos": 19426200888960.0,
      "grad_norm": 1.8722016114425952,
      "language_loss": 0.8628391,
      "learning_rate": 3.7087860178344955e-06,
      "loss": 0.8845247,
      "num_input_tokens_seen": 71319670,
      "step": 3305,
      "time_per_iteration": 2.7129390239715576
    },
    {
      "auxiliary_loss_clip": 0.01128329,
      "auxiliary_loss_mlp": 0.01044081,
      "balance_loss_clip": 1.04770195,
      "balance_loss_mlp": 1.02603281,
      "epoch": 0.19876747332030664,
      "flos": 23547380947200.0,
      "grad_norm": 2.9829227362861106,
      "language_loss": 0.68476367,
      "learning_rate": 3.7085836104943445e-06,
      "loss": 0.70648777,
      "num_input_tokens_seen": 71339850,
      "step": 3306,
      "time_per_iteration": 2.7083208560943604
    },
    {
      "auxiliary_loss_clip": 0.01119386,
      "auxiliary_loss_mlp": 0.01038782,
      "balance_loss_clip": 1.04822719,
      "balance_loss_mlp": 1.02168787,
      "epoch": 0.1988275965729746,
      "flos": 19829406032640.0,
      "grad_norm": 1.683647244561179,
      "language_loss": 0.76433122,
      "learning_rate": 3.7083811383645332e-06,
      "loss": 0.78591287,
      "num_input_tokens_seen": 71359795,
      "step": 3307,
      "time_per_iteration": 2.728661298751831
    },
    {
      "auxiliary_loss_clip": 0.01157548,
      "auxiliary_loss_mlp": 0.01044665,
      "balance_loss_clip": 1.05895782,
      "balance_loss_mlp": 1.02714145,
      "epoch": 0.19888771982564257,
      "flos": 23513625141120.0,
      "grad_norm": 2.438172575069382,
      "language_loss": 0.75991976,
      "learning_rate": 3.708178601452737e-06,
      "loss": 0.78194201,
      "num_input_tokens_seen": 71378885,
      "step": 3308,
      "time_per_iteration": 2.6580557823181152
    },
    {
      "auxiliary_loss_clip": 0.01107283,
      "auxiliary_loss_mlp": 0.01041656,
      "balance_loss_clip": 1.05453563,
      "balance_loss_mlp": 1.02307141,
      "epoch": 0.19894784307831054,
      "flos": 18150510389760.0,
      "grad_norm": 1.928689575161362,
      "language_loss": 0.76043576,
      "learning_rate": 3.7079759997666374e-06,
      "loss": 0.7819252,
      "num_input_tokens_seen": 71397285,
      "step": 3309,
      "time_per_iteration": 2.77226185798645
    },
    {
      "auxiliary_loss_clip": 0.0114115,
      "auxiliary_loss_mlp": 0.01045061,
      "balance_loss_clip": 1.05222607,
      "balance_loss_mlp": 1.02592754,
      "epoch": 0.1990079663309785,
      "flos": 24276044246400.0,
      "grad_norm": 75.17312936609292,
      "language_loss": 0.87855697,
      "learning_rate": 3.707773333313917e-06,
      "loss": 0.90041906,
      "num_input_tokens_seen": 71415775,
      "step": 3310,
      "time_per_iteration": 2.6789662837982178
    },
    {
      "auxiliary_loss_clip": 0.01153037,
      "auxiliary_loss_mlp": 0.01039864,
      "balance_loss_clip": 1.05415869,
      "balance_loss_mlp": 1.02139854,
      "epoch": 0.19906808958364647,
      "flos": 34897666366080.0,
      "grad_norm": 2.3155756588664342,
      "language_loss": 0.63650048,
      "learning_rate": 3.70757060210226e-06,
      "loss": 0.6584295,
      "num_input_tokens_seen": 71437315,
      "step": 3311,
      "time_per_iteration": 2.7604620456695557
    },
    {
      "auxiliary_loss_clip": 0.01115133,
      "auxiliary_loss_mlp": 0.01043871,
      "balance_loss_clip": 1.04763019,
      "balance_loss_mlp": 1.02501202,
      "epoch": 0.19912821283631443,
      "flos": 24024885373440.0,
      "grad_norm": 3.8064295514597717,
      "language_loss": 0.74542546,
      "learning_rate": 3.707367806139355e-06,
      "loss": 0.76701546,
      "num_input_tokens_seen": 71456320,
      "step": 3312,
      "time_per_iteration": 2.796475410461426
    },
    {
      "auxiliary_loss_clip": 0.01141587,
      "auxiliary_loss_mlp": 0.01037435,
      "balance_loss_clip": 1.05358124,
      "balance_loss_mlp": 1.02017355,
      "epoch": 0.19918833608898243,
      "flos": 19859031774720.0,
      "grad_norm": 2.2312990164825943,
      "language_loss": 0.84033173,
      "learning_rate": 3.7071649454328915e-06,
      "loss": 0.86212194,
      "num_input_tokens_seen": 71475360,
      "step": 3313,
      "time_per_iteration": 2.6044952869415283
    },
    {
      "auxiliary_loss_clip": 0.01146797,
      "auxiliary_loss_mlp": 0.01042166,
      "balance_loss_clip": 1.05695391,
      "balance_loss_mlp": 1.02422476,
      "epoch": 0.1992484593416504,
      "flos": 29095794984960.0,
      "grad_norm": 3.856678450124864,
      "language_loss": 0.810305,
      "learning_rate": 3.7069620199905625e-06,
      "loss": 0.83219463,
      "num_input_tokens_seen": 71496155,
      "step": 3314,
      "time_per_iteration": 2.68841814994812
    },
    {
      "auxiliary_loss_clip": 0.01112846,
      "auxiliary_loss_mlp": 0.01043677,
      "balance_loss_clip": 1.04617178,
      "balance_loss_mlp": 1.02643955,
      "epoch": 0.19930858259431836,
      "flos": 23295001011840.0,
      "grad_norm": 1.4822079401394097,
      "language_loss": 0.87391549,
      "learning_rate": 3.7067590298200627e-06,
      "loss": 0.89548075,
      "num_input_tokens_seen": 71517295,
      "step": 3315,
      "time_per_iteration": 2.720093011856079
    },
    {
      "auxiliary_loss_clip": 0.0111589,
      "auxiliary_loss_mlp": 0.00777002,
      "balance_loss_clip": 1.04992676,
      "balance_loss_mlp": 1.00093687,
      "epoch": 0.19936870584698632,
      "flos": 25378825651200.0,
      "grad_norm": 1.7805516248937883,
      "language_loss": 0.70957202,
      "learning_rate": 3.7065559749290892e-06,
      "loss": 0.72850096,
      "num_input_tokens_seen": 71540000,
      "step": 3316,
      "time_per_iteration": 2.850100517272949
    },
    {
      "auxiliary_loss_clip": 0.01019745,
      "auxiliary_loss_mlp": 0.01012504,
      "balance_loss_clip": 1.03032303,
      "balance_loss_mlp": 1.01003671,
      "epoch": 0.1994288290996543,
      "flos": 62168053109760.0,
      "grad_norm": 0.8326978726055106,
      "language_loss": 0.66287398,
      "learning_rate": 3.706352855325342e-06,
      "loss": 0.68319643,
      "num_input_tokens_seen": 71607880,
      "step": 3317,
      "time_per_iteration": 3.425114870071411
    },
    {
      "auxiliary_loss_clip": 0.01148059,
      "auxiliary_loss_mlp": 0.01048913,
      "balance_loss_clip": 1.05397809,
      "balance_loss_mlp": 1.02964854,
      "epoch": 0.19948895235232225,
      "flos": 19025832919680.0,
      "grad_norm": 2.282515690517884,
      "language_loss": 0.74494618,
      "learning_rate": 3.7061496710165233e-06,
      "loss": 0.76691592,
      "num_input_tokens_seen": 71625695,
      "step": 3318,
      "time_per_iteration": 2.6815896034240723
    },
    {
      "auxiliary_loss_clip": 0.01114942,
      "auxiliary_loss_mlp": 0.01044681,
      "balance_loss_clip": 1.04767084,
      "balance_loss_mlp": 1.02786088,
      "epoch": 0.19954907560499022,
      "flos": 37815803182080.0,
      "grad_norm": 1.8966456913695608,
      "language_loss": 0.78894758,
      "learning_rate": 3.7059464220103385e-06,
      "loss": 0.81054389,
      "num_input_tokens_seen": 71648520,
      "step": 3319,
      "time_per_iteration": 2.847911834716797
    },
    {
      "auxiliary_loss_clip": 0.01134557,
      "auxiliary_loss_mlp": 0.01042988,
      "balance_loss_clip": 1.05354095,
      "balance_loss_mlp": 1.02312756,
      "epoch": 0.1996091988576582,
      "flos": 49565199594240.0,
      "grad_norm": 2.1348540211051197,
      "language_loss": 0.76006937,
      "learning_rate": 3.7057431083144945e-06,
      "loss": 0.78184479,
      "num_input_tokens_seen": 71672185,
      "step": 3320,
      "time_per_iteration": 2.9324615001678467
    },
    {
      "auxiliary_loss_clip": 0.01120226,
      "auxiliary_loss_mlp": 0.01042998,
      "balance_loss_clip": 1.05083311,
      "balance_loss_mlp": 1.02496171,
      "epoch": 0.19966932211032618,
      "flos": 22635788659200.0,
      "grad_norm": 2.2436863685702546,
      "language_loss": 0.80077857,
      "learning_rate": 3.705539729936701e-06,
      "loss": 0.82241082,
      "num_input_tokens_seen": 71692890,
      "step": 3321,
      "time_per_iteration": 2.7534186840057373
    },
    {
      "auxiliary_loss_clip": 0.01033096,
      "auxiliary_loss_mlp": 0.01011167,
      "balance_loss_clip": 1.02391553,
      "balance_loss_mlp": 1.00828266,
      "epoch": 0.19972944536299414,
      "flos": 54082117745280.0,
      "grad_norm": 0.874673110280983,
      "language_loss": 0.65145189,
      "learning_rate": 3.7053362868846696e-06,
      "loss": 0.67189455,
      "num_input_tokens_seen": 71745815,
      "step": 3322,
      "time_per_iteration": 3.0398683547973633
    },
    {
      "auxiliary_loss_clip": 0.01039999,
      "auxiliary_loss_mlp": 0.01007775,
      "balance_loss_clip": 1.02971482,
      "balance_loss_mlp": 1.00479472,
      "epoch": 0.1997895686156621,
      "flos": 69355031817600.0,
      "grad_norm": 0.7915334307535052,
      "language_loss": 0.56919783,
      "learning_rate": 3.7051327791661153e-06,
      "loss": 0.58967561,
      "num_input_tokens_seen": 71806915,
      "step": 3323,
      "time_per_iteration": 3.2814581394195557
    },
    {
      "auxiliary_loss_clip": 0.01131487,
      "auxiliary_loss_mlp": 0.00776139,
      "balance_loss_clip": 1.05244064,
      "balance_loss_mlp": 1.00085235,
      "epoch": 0.19984969186833007,
      "flos": 18552063507840.0,
      "grad_norm": 1.8766856730809967,
      "language_loss": 0.80573648,
      "learning_rate": 3.7049292067887555e-06,
      "loss": 0.82481277,
      "num_input_tokens_seen": 71824645,
      "step": 3324,
      "time_per_iteration": 2.66456937789917
    },
    {
      "auxiliary_loss_clip": 0.01132572,
      "auxiliary_loss_mlp": 0.01050254,
      "balance_loss_clip": 1.04625165,
      "balance_loss_mlp": 1.03027487,
      "epoch": 0.19990981512099804,
      "flos": 26429678968320.0,
      "grad_norm": 2.4535669107623486,
      "language_loss": 0.53931105,
      "learning_rate": 3.7047255697603092e-06,
      "loss": 0.56113935,
      "num_input_tokens_seen": 71845125,
      "step": 3325,
      "time_per_iteration": 2.696556329727173
    },
    {
      "auxiliary_loss_clip": 0.01130165,
      "auxiliary_loss_mlp": 0.01050725,
      "balance_loss_clip": 1.05065942,
      "balance_loss_mlp": 1.03328443,
      "epoch": 0.19996993837366603,
      "flos": 16325997010560.0,
      "grad_norm": 2.1570763946475187,
      "language_loss": 0.86074936,
      "learning_rate": 3.7045218680884984e-06,
      "loss": 0.88255823,
      "num_input_tokens_seen": 71863500,
      "step": 3326,
      "time_per_iteration": 2.7167885303497314
    },
    {
      "auxiliary_loss_clip": 0.0115173,
      "auxiliary_loss_mlp": 0.01042065,
      "balance_loss_clip": 1.05427039,
      "balance_loss_mlp": 1.02511311,
      "epoch": 0.200030061626334,
      "flos": 20844169159680.0,
      "grad_norm": 2.0419576492150395,
      "language_loss": 0.71793801,
      "learning_rate": 3.7043181017810476e-06,
      "loss": 0.73987597,
      "num_input_tokens_seen": 71881845,
      "step": 3327,
      "time_per_iteration": 2.6097662448883057
    },
    {
      "auxiliary_loss_clip": 0.01131035,
      "auxiliary_loss_mlp": 0.01052756,
      "balance_loss_clip": 1.05146813,
      "balance_loss_mlp": 1.03290796,
      "epoch": 0.20009018487900196,
      "flos": 23762629198080.0,
      "grad_norm": 1.8948781463857982,
      "language_loss": 0.7668376,
      "learning_rate": 3.7041142708456833e-06,
      "loss": 0.78867549,
      "num_input_tokens_seen": 71900940,
      "step": 3328,
      "time_per_iteration": 2.6869349479675293
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 16632,
  "num_input_tokens_seen": 71900940,
  "num_train_epochs": 1,
  "save_steps": 3328,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 2.8067294895459533e+17,
  "train_batch_size": 5,
  "trial_name": null,
  "trial_params": null
}