{
  "best_global_step": 6000,
  "best_metric": 0.6813357472419739,
  "best_model_checkpoint": "task2file/sft_devstral_24B_v2/checkpoints/checkpoint-6000",
  "epoch": 2.5316455696202533,
  "eval_steps": 100,
  "global_step": 6000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0008438818565400844,
      "grad_norm": 1.597854733467102,
      "learning_rate": 8.787346221441124e-08,
      "loss": 1.3927901983261108,
      "step": 2
    },
    {
      "epoch": 0.0016877637130801688,
      "grad_norm": 1.6547431945800781,
      "learning_rate": 2.6362038664323375e-07,
      "loss": 1.407160758972168,
      "step": 4
    },
    {
      "epoch": 0.002531645569620253,
      "grad_norm": 1.8221601247787476,
      "learning_rate": 4.393673110720563e-07,
      "loss": 1.376656174659729,
      "step": 6
    },
    {
      "epoch": 0.0033755274261603376,
      "grad_norm": 1.4831048250198364,
      "learning_rate": 6.151142355008788e-07,
      "loss": 1.247712254524231,
      "step": 8
    },
    {
      "epoch": 0.004219409282700422,
      "grad_norm": 1.668201208114624,
      "learning_rate": 7.908611599297013e-07,
      "loss": 1.2685163021087646,
      "step": 10
    },
    {
      "epoch": 0.005063291139240506,
      "grad_norm": 1.67417311668396,
      "learning_rate": 9.666080843585237e-07,
      "loss": 1.2942761182785034,
      "step": 12
    },
    {
      "epoch": 0.00590717299578059,
      "grad_norm": 1.7154079675674438,
      "learning_rate": 1.1423550087873463e-06,
      "loss": 1.3638604879379272,
      "step": 14
    },
    {
      "epoch": 0.006751054852320675,
      "grad_norm": 1.729427456855774,
      "learning_rate": 1.3181019332161688e-06,
      "loss": 1.3476728200912476,
      "step": 16
    },
    {
      "epoch": 0.007594936708860759,
      "grad_norm": 1.3813447952270508,
      "learning_rate": 1.4938488576449913e-06,
      "loss": 1.3476393222808838,
      "step": 18
    },
    {
      "epoch": 0.008438818565400843,
      "grad_norm": 1.557220458984375,
      "learning_rate": 1.6695957820738139e-06,
      "loss": 1.2449309825897217,
      "step": 20
    },
    {
      "epoch": 0.009282700421940928,
      "grad_norm": 1.1883500814437866,
      "learning_rate": 1.8453427065026362e-06,
      "loss": 1.3125361204147339,
      "step": 22
    },
    {
      "epoch": 0.010126582278481013,
      "grad_norm": 1.7290029525756836,
      "learning_rate": 2.0210896309314587e-06,
      "loss": 1.3724769353866577,
      "step": 24
    },
    {
      "epoch": 0.010970464135021098,
      "grad_norm": 1.5627557039260864,
      "learning_rate": 2.1968365553602812e-06,
      "loss": 1.3401387929916382,
      "step": 26
    },
    {
      "epoch": 0.01181434599156118,
      "grad_norm": 1.796866774559021,
      "learning_rate": 2.3725834797891038e-06,
      "loss": 1.365437388420105,
      "step": 28
    },
    {
      "epoch": 0.012658227848101266,
      "grad_norm": 1.7030404806137085,
      "learning_rate": 2.5483304042179263e-06,
      "loss": 1.2706533670425415,
      "step": 30
    },
    {
      "epoch": 0.01350210970464135,
      "grad_norm": 1.3186293840408325,
      "learning_rate": 2.724077328646749e-06,
      "loss": 1.3084994554519653,
      "step": 32
    },
    {
      "epoch": 0.014345991561181435,
      "grad_norm": 1.5762513875961304,
      "learning_rate": 2.8998242530755714e-06,
      "loss": 1.3259696960449219,
      "step": 34
    },
    {
      "epoch": 0.015189873417721518,
      "grad_norm": 1.422295331954956,
      "learning_rate": 3.075571177504394e-06,
      "loss": 1.3205676078796387,
      "step": 36
    },
    {
      "epoch": 0.016033755274261603,
      "grad_norm": 1.495523452758789,
      "learning_rate": 3.2513181019332165e-06,
      "loss": 1.3740568161010742,
      "step": 38
    },
    {
      "epoch": 0.016877637130801686,
      "grad_norm": 1.5112254619598389,
      "learning_rate": 3.427065026362039e-06,
      "loss": 1.321828842163086,
      "step": 40
    },
    {
      "epoch": 0.017721518987341773,
      "grad_norm": 1.4667807817459106,
      "learning_rate": 3.602811950790861e-06,
      "loss": 1.3673173189163208,
      "step": 42
    },
    {
      "epoch": 0.018565400843881856,
      "grad_norm": 1.6609723567962646,
      "learning_rate": 3.7785588752196836e-06,
      "loss": 1.3968093395233154,
      "step": 44
    },
    {
      "epoch": 0.019409282700421943,
      "grad_norm": 1.59381103515625,
      "learning_rate": 3.954305799648506e-06,
      "loss": 1.4295302629470825,
      "step": 46
    },
    {
      "epoch": 0.020253164556962026,
      "grad_norm": 1.1470608711242676,
      "learning_rate": 4.130052724077329e-06,
      "loss": 1.2536572217941284,
      "step": 48
    },
    {
      "epoch": 0.02109704641350211,
      "grad_norm": 1.2014588117599487,
      "learning_rate": 4.305799648506151e-06,
      "loss": 1.242217779159546,
      "step": 50
    },
    {
      "epoch": 0.021940928270042195,
      "grad_norm": 1.2327464818954468,
      "learning_rate": 4.481546572934974e-06,
      "loss": 1.2166963815689087,
      "step": 52
    },
    {
      "epoch": 0.02278481012658228,
      "grad_norm": 1.9708983898162842,
      "learning_rate": 4.657293497363796e-06,
      "loss": 1.25709867477417,
      "step": 54
    },
    {
      "epoch": 0.02362869198312236,
      "grad_norm": 1.180569052696228,
      "learning_rate": 4.833040421792619e-06,
      "loss": 1.2886158227920532,
      "step": 56
    },
    {
      "epoch": 0.024472573839662448,
      "grad_norm": 1.5029548406600952,
      "learning_rate": 5.008787346221441e-06,
      "loss": 1.29886794090271,
      "step": 58
    },
    {
      "epoch": 0.02531645569620253,
      "grad_norm": 1.5380216836929321,
      "learning_rate": 5.184534270650264e-06,
      "loss": 1.2387628555297852,
      "step": 60
    },
    {
      "epoch": 0.026160337552742614,
      "grad_norm": 1.572144865989685,
      "learning_rate": 5.3602811950790864e-06,
      "loss": 1.2177000045776367,
      "step": 62
    },
    {
      "epoch": 0.0270042194092827,
      "grad_norm": 1.4882780313491821,
      "learning_rate": 5.536028119507909e-06,
      "loss": 1.181516170501709,
      "step": 64
    },
    {
      "epoch": 0.027848101265822784,
      "grad_norm": 1.2982488870620728,
      "learning_rate": 5.7117750439367315e-06,
      "loss": 1.2101733684539795,
      "step": 66
    },
    {
      "epoch": 0.02869198312236287,
      "grad_norm": 1.5236955881118774,
      "learning_rate": 5.887521968365554e-06,
      "loss": 1.2277681827545166,
      "step": 68
    },
    {
      "epoch": 0.029535864978902954,
      "grad_norm": 1.4521006345748901,
      "learning_rate": 6.0632688927943766e-06,
      "loss": 1.1688424348831177,
      "step": 70
    },
    {
      "epoch": 0.030379746835443037,
      "grad_norm": 1.2352311611175537,
      "learning_rate": 6.239015817223199e-06,
      "loss": 1.273059368133545,
      "step": 72
    },
    {
      "epoch": 0.031223628691983123,
      "grad_norm": 1.3438209295272827,
      "learning_rate": 6.414762741652021e-06,
      "loss": 1.1609034538269043,
      "step": 74
    },
    {
      "epoch": 0.032067510548523206,
      "grad_norm": 1.9009398221969604,
      "learning_rate": 6.590509666080843e-06,
      "loss": 1.2508260011672974,
      "step": 76
    },
    {
      "epoch": 0.03291139240506329,
      "grad_norm": 1.6718412637710571,
      "learning_rate": 6.766256590509666e-06,
      "loss": 1.2524956464767456,
      "step": 78
    },
    {
      "epoch": 0.03375527426160337,
      "grad_norm": 1.249891757965088,
      "learning_rate": 6.942003514938488e-06,
      "loss": 1.1472493410110474,
      "step": 80
    },
    {
      "epoch": 0.03459915611814346,
      "grad_norm": 1.4398653507232666,
      "learning_rate": 7.117750439367312e-06,
      "loss": 1.0845389366149902,
      "step": 82
    },
    {
      "epoch": 0.035443037974683546,
      "grad_norm": 1.3701167106628418,
      "learning_rate": 7.293497363796134e-06,
      "loss": 1.1088868379592896,
      "step": 84
    },
    {
      "epoch": 0.036286919831223625,
      "grad_norm": 1.277998924255371,
      "learning_rate": 7.469244288224957e-06,
      "loss": 1.1513772010803223,
      "step": 86
    },
    {
      "epoch": 0.03713080168776371,
      "grad_norm": 1.4970002174377441,
      "learning_rate": 7.644991212653779e-06,
      "loss": 1.1385771036148071,
      "step": 88
    },
    {
      "epoch": 0.0379746835443038,
      "grad_norm": 1.3384218215942383,
      "learning_rate": 7.820738137082601e-06,
      "loss": 1.1632680892944336,
      "step": 90
    },
    {
      "epoch": 0.038818565400843885,
      "grad_norm": 1.4317446947097778,
      "learning_rate": 7.996485061511425e-06,
      "loss": 1.2256064414978027,
      "step": 92
    },
    {
      "epoch": 0.039662447257383965,
      "grad_norm": 1.8743640184402466,
      "learning_rate": 8.172231985940246e-06,
      "loss": 1.1935789585113525,
      "step": 94
    },
    {
      "epoch": 0.04050632911392405,
      "grad_norm": 1.4789546728134155,
      "learning_rate": 8.347978910369069e-06,
      "loss": 1.1429362297058105,
      "step": 96
    },
    {
      "epoch": 0.04135021097046414,
      "grad_norm": 1.658605694770813,
      "learning_rate": 8.523725834797891e-06,
      "loss": 1.1831508874893188,
      "step": 98
    },
    {
      "epoch": 0.04219409282700422,
      "grad_norm": 1.5077892541885376,
      "learning_rate": 8.699472759226714e-06,
      "loss": 1.0539867877960205,
      "step": 100
    },
    {
      "epoch": 0.04219409282700422,
      "eval_loss": 1.138856053352356,
      "eval_runtime": 859.7128,
      "eval_samples_per_second": 2.451,
      "eval_steps_per_second": 2.451,
      "step": 100
    },
    {
      "epoch": 0.043037974683544304,
      "grad_norm": 1.4335681200027466,
      "learning_rate": 8.875219683655536e-06,
      "loss": 1.0719901323318481,
      "step": 102
    },
    {
      "epoch": 0.04388185654008439,
      "grad_norm": 1.7387681007385254,
      "learning_rate": 9.050966608084359e-06,
      "loss": 1.0654313564300537,
      "step": 104
    },
    {
      "epoch": 0.04472573839662447,
      "grad_norm": 1.6071950197219849,
      "learning_rate": 9.226713532513181e-06,
      "loss": 1.0752698183059692,
      "step": 106
    },
    {
      "epoch": 0.04556962025316456,
      "grad_norm": 1.40005362033844,
      "learning_rate": 9.402460456942004e-06,
      "loss": 1.1029763221740723,
      "step": 108
    },
    {
      "epoch": 0.046413502109704644,
      "grad_norm": 2.2338669300079346,
      "learning_rate": 9.578207381370826e-06,
      "loss": 1.1157960891723633,
      "step": 110
    },
    {
      "epoch": 0.04725738396624472,
      "grad_norm": 1.4972727298736572,
      "learning_rate": 9.753954305799649e-06,
      "loss": 1.1095420122146606,
      "step": 112
    },
    {
      "epoch": 0.04810126582278481,
      "grad_norm": 1.317979097366333,
      "learning_rate": 9.929701230228471e-06,
      "loss": 1.109113097190857,
      "step": 114
    },
    {
      "epoch": 0.048945147679324896,
      "grad_norm": 1.496346116065979,
      "learning_rate": 1.0105448154657294e-05,
      "loss": 1.1055104732513428,
      "step": 116
    },
    {
      "epoch": 0.049789029535864976,
      "grad_norm": 1.385406732559204,
      "learning_rate": 1.0281195079086117e-05,
      "loss": 1.118395209312439,
      "step": 118
    },
    {
      "epoch": 0.05063291139240506,
      "grad_norm": 1.524222731590271,
      "learning_rate": 1.0456942003514939e-05,
      "loss": 1.1008446216583252,
      "step": 120
    },
    {
      "epoch": 0.05147679324894515,
      "grad_norm": 1.6308200359344482,
      "learning_rate": 1.0632688927943762e-05,
      "loss": 1.0891425609588623,
      "step": 122
    },
    {
      "epoch": 0.05232067510548523,
      "grad_norm": 1.3681106567382812,
      "learning_rate": 1.0808435852372584e-05,
      "loss": 0.9080473184585571,
      "step": 124
    },
    {
      "epoch": 0.053164556962025315,
      "grad_norm": 1.9429908990859985,
      "learning_rate": 1.0984182776801407e-05,
      "loss": 1.0337369441986084,
      "step": 126
    },
    {
      "epoch": 0.0540084388185654,
      "grad_norm": 1.5830830335617065,
      "learning_rate": 1.115992970123023e-05,
      "loss": 1.0703333616256714,
      "step": 128
    },
    {
      "epoch": 0.05485232067510549,
      "grad_norm": 1.4792555570602417,
      "learning_rate": 1.1335676625659052e-05,
      "loss": 1.004652738571167,
      "step": 130
    },
    {
      "epoch": 0.05569620253164557,
      "grad_norm": 1.7196226119995117,
      "learning_rate": 1.1511423550087874e-05,
      "loss": 0.9798293709754944,
      "step": 132
    },
    {
      "epoch": 0.056540084388185655,
      "grad_norm": 1.8733659982681274,
      "learning_rate": 1.1687170474516697e-05,
      "loss": 1.0213249921798706,
      "step": 134
    },
    {
      "epoch": 0.05738396624472574,
      "grad_norm": 1.3431142568588257,
      "learning_rate": 1.186291739894552e-05,
      "loss": 1.0358591079711914,
      "step": 136
    },
    {
      "epoch": 0.05822784810126582,
      "grad_norm": 1.527864933013916,
      "learning_rate": 1.2038664323374342e-05,
      "loss": 0.9372249841690063,
      "step": 138
    },
    {
      "epoch": 0.05907172995780591,
      "grad_norm": 1.5495563745498657,
      "learning_rate": 1.2214411247803164e-05,
      "loss": 1.0277758836746216,
      "step": 140
    },
    {
      "epoch": 0.059915611814345994,
      "grad_norm": 1.6792418956756592,
      "learning_rate": 1.2390158172231985e-05,
      "loss": 1.0349801778793335,
      "step": 142
    },
    {
      "epoch": 0.060759493670886074,
      "grad_norm": 1.6468945741653442,
      "learning_rate": 1.256590509666081e-05,
      "loss": 0.9578297734260559,
      "step": 144
    },
    {
      "epoch": 0.06160337552742616,
      "grad_norm": 1.7243824005126953,
      "learning_rate": 1.2741652021089632e-05,
      "loss": 1.0628854036331177,
      "step": 146
    },
    {
      "epoch": 0.06244725738396625,
      "grad_norm": 1.7286981344223022,
      "learning_rate": 1.2917398945518455e-05,
      "loss": 0.9336449503898621,
      "step": 148
    },
    {
      "epoch": 0.06329113924050633,
      "grad_norm": 1.6411832571029663,
      "learning_rate": 1.3093145869947277e-05,
      "loss": 0.953730583190918,
      "step": 150
    },
    {
      "epoch": 0.06413502109704641,
      "grad_norm": 1.8297001123428345,
      "learning_rate": 1.3268892794376098e-05,
      "loss": 1.051239013671875,
      "step": 152
    },
    {
      "epoch": 0.06497890295358649,
      "grad_norm": 1.9660519361495972,
      "learning_rate": 1.3444639718804922e-05,
      "loss": 0.9955035448074341,
      "step": 154
    },
    {
      "epoch": 0.06582278481012659,
      "grad_norm": 1.8423733711242676,
      "learning_rate": 1.3620386643233743e-05,
      "loss": 0.913300096988678,
      "step": 156
    },
    {
      "epoch": 0.06666666666666667,
      "grad_norm": 1.9146347045898438,
      "learning_rate": 1.3796133567662567e-05,
      "loss": 1.0429846048355103,
      "step": 158
    },
    {
      "epoch": 0.06751054852320675,
      "grad_norm": 1.6221821308135986,
      "learning_rate": 1.3971880492091388e-05,
      "loss": 1.0360238552093506,
      "step": 160
    },
    {
      "epoch": 0.06835443037974684,
      "grad_norm": 2.173283338546753,
      "learning_rate": 1.4147627416520212e-05,
      "loss": 1.0227266550064087,
      "step": 162
    },
    {
      "epoch": 0.06919831223628692,
      "grad_norm": 1.7091665267944336,
      "learning_rate": 1.4323374340949033e-05,
      "loss": 1.0075194835662842,
      "step": 164
    },
    {
      "epoch": 0.070042194092827,
      "grad_norm": 1.7219135761260986,
      "learning_rate": 1.4499121265377857e-05,
      "loss": 1.0044782161712646,
      "step": 166
    },
    {
      "epoch": 0.07088607594936709,
      "grad_norm": 1.6558159589767456,
      "learning_rate": 1.4674868189806678e-05,
      "loss": 0.9393973350524902,
      "step": 168
    },
    {
      "epoch": 0.07172995780590717,
      "grad_norm": 1.9362739324569702,
      "learning_rate": 1.4850615114235502e-05,
      "loss": 0.9955337643623352,
      "step": 170
    },
    {
      "epoch": 0.07257383966244725,
      "grad_norm": 1.7792853116989136,
      "learning_rate": 1.5026362038664323e-05,
      "loss": 0.9659126400947571,
      "step": 172
    },
    {
      "epoch": 0.07341772151898734,
      "grad_norm": 1.7184511423110962,
      "learning_rate": 1.5202108963093147e-05,
      "loss": 0.9077855348587036,
      "step": 174
    },
    {
      "epoch": 0.07426160337552742,
      "grad_norm": 1.5701428651809692,
      "learning_rate": 1.537785588752197e-05,
      "loss": 0.9305018782615662,
      "step": 176
    },
    {
      "epoch": 0.0751054852320675,
      "grad_norm": 1.970229148864746,
      "learning_rate": 1.555360281195079e-05,
      "loss": 1.0211774110794067,
      "step": 178
    },
    {
      "epoch": 0.0759493670886076,
      "grad_norm": 1.8410269021987915,
      "learning_rate": 1.5729349736379615e-05,
      "loss": 0.9479315876960754,
      "step": 180
    },
    {
      "epoch": 0.07679324894514768,
      "grad_norm": 1.8991246223449707,
      "learning_rate": 1.5905096660808434e-05,
      "loss": 1.0629050731658936,
      "step": 182
    },
    {
      "epoch": 0.07763713080168777,
      "grad_norm": 1.8052008152008057,
      "learning_rate": 1.608084358523726e-05,
      "loss": 0.946983814239502,
      "step": 184
    },
    {
      "epoch": 0.07848101265822785,
      "grad_norm": 1.547108769416809,
      "learning_rate": 1.625659050966608e-05,
      "loss": 0.9413356184959412,
      "step": 186
    },
    {
      "epoch": 0.07932489451476793,
      "grad_norm": 1.9713538885116577,
      "learning_rate": 1.6432337434094905e-05,
      "loss": 0.9337888956069946,
      "step": 188
    },
    {
      "epoch": 0.08016877637130802,
      "grad_norm": 1.708789348602295,
      "learning_rate": 1.6608084358523728e-05,
      "loss": 0.9816337823867798,
      "step": 190
    },
    {
      "epoch": 0.0810126582278481,
      "grad_norm": 1.815292477607727,
      "learning_rate": 1.678383128295255e-05,
      "loss": 1.017122507095337,
      "step": 192
    },
    {
      "epoch": 0.08185654008438818,
      "grad_norm": 1.7950682640075684,
      "learning_rate": 1.6959578207381373e-05,
      "loss": 0.991599440574646,
      "step": 194
    },
    {
      "epoch": 0.08270042194092828,
      "grad_norm": 1.692512035369873,
      "learning_rate": 1.7135325131810195e-05,
      "loss": 0.9570834040641785,
      "step": 196
    },
    {
      "epoch": 0.08354430379746836,
      "grad_norm": 2.056089162826538,
      "learning_rate": 1.7311072056239018e-05,
      "loss": 1.035754919052124,
      "step": 198
    },
    {
      "epoch": 0.08438818565400844,
      "grad_norm": 1.7022203207015991,
      "learning_rate": 1.7486818980667837e-05,
      "loss": 1.0124205350875854,
      "step": 200
    },
    {
      "epoch": 0.08438818565400844,
      "eval_loss": 0.995743453502655,
      "eval_runtime": 846.8257,
      "eval_samples_per_second": 2.488,
      "eval_steps_per_second": 2.488,
      "step": 200
    },
    {
      "epoch": 0.08523206751054853,
      "grad_norm": 1.6088604927062988,
      "learning_rate": 1.7662565905096663e-05,
      "loss": 0.8946985006332397,
      "step": 202
    },
    {
      "epoch": 0.08607594936708861,
      "grad_norm": 2.02270770072937,
      "learning_rate": 1.7838312829525482e-05,
      "loss": 0.976133406162262,
      "step": 204
    },
    {
      "epoch": 0.08691983122362869,
      "grad_norm": 1.7832789421081543,
      "learning_rate": 1.8014059753954308e-05,
      "loss": 0.9079383611679077,
      "step": 206
    },
    {
      "epoch": 0.08776371308016878,
      "grad_norm": 1.9793545007705688,
      "learning_rate": 1.8189806678383127e-05,
      "loss": 0.8650367856025696,
      "step": 208
    },
    {
      "epoch": 0.08860759493670886,
      "grad_norm": 1.8124271631240845,
      "learning_rate": 1.8365553602811953e-05,
      "loss": 0.9327266812324524,
      "step": 210
    },
    {
      "epoch": 0.08945147679324894,
      "grad_norm": 1.8581212759017944,
      "learning_rate": 1.8541300527240772e-05,
      "loss": 0.9811079502105713,
      "step": 212
    },
    {
      "epoch": 0.09029535864978903,
      "grad_norm": 2.001699447631836,
      "learning_rate": 1.8717047451669598e-05,
      "loss": 0.9546971321105957,
      "step": 214
    },
    {
      "epoch": 0.09113924050632911,
      "grad_norm": 1.6994978189468384,
      "learning_rate": 1.8892794376098417e-05,
      "loss": 0.9611319899559021,
      "step": 216
    },
    {
      "epoch": 0.0919831223628692,
      "grad_norm": 2.1379497051239014,
      "learning_rate": 1.9068541300527243e-05,
      "loss": 0.9781531095504761,
      "step": 218
    },
    {
      "epoch": 0.09282700421940929,
      "grad_norm": 1.8961224555969238,
      "learning_rate": 1.9244288224956066e-05,
      "loss": 0.9374833106994629,
      "step": 220
    },
    {
      "epoch": 0.09367088607594937,
      "grad_norm": 1.851464033126831,
      "learning_rate": 1.9420035149384885e-05,
      "loss": 0.9681299328804016,
      "step": 222
    },
    {
      "epoch": 0.09451476793248945,
      "grad_norm": 2.0642266273498535,
      "learning_rate": 1.959578207381371e-05,
      "loss": 1.0086225271224976,
      "step": 224
    },
    {
      "epoch": 0.09535864978902954,
      "grad_norm": 1.8658756017684937,
      "learning_rate": 1.977152899824253e-05,
      "loss": 0.9190312623977661,
      "step": 226
    },
    {
      "epoch": 0.09620253164556962,
      "grad_norm": 2.4398674964904785,
      "learning_rate": 1.9947275922671356e-05,
      "loss": 0.9740874171257019,
      "step": 228
    },
    {
      "epoch": 0.0970464135021097,
      "grad_norm": 1.849183440208435,
      "learning_rate": 2.0123022847100175e-05,
      "loss": 0.884376049041748,
      "step": 230
    },
    {
      "epoch": 0.09789029535864979,
      "grad_norm": 2.027320384979248,
      "learning_rate": 2.0298769771529e-05,
      "loss": 0.9116487503051758,
      "step": 232
    },
    {
      "epoch": 0.09873417721518987,
      "grad_norm": 1.6800135374069214,
      "learning_rate": 2.047451669595782e-05,
      "loss": 0.9035115242004395,
      "step": 234
    },
    {
      "epoch": 0.09957805907172995,
      "grad_norm": 2.2362256050109863,
      "learning_rate": 2.0650263620386646e-05,
      "loss": 0.9043796062469482,
      "step": 236
    },
    {
      "epoch": 0.10042194092827005,
      "grad_norm": 1.938215970993042,
      "learning_rate": 2.0826010544815465e-05,
      "loss": 1.0888828039169312,
      "step": 238
    },
    {
      "epoch": 0.10126582278481013,
      "grad_norm": 1.890328049659729,
      "learning_rate": 2.100175746924429e-05,
      "loss": 0.9960280656814575,
      "step": 240
    },
    {
      "epoch": 0.1021097046413502,
      "grad_norm": 2.021235227584839,
      "learning_rate": 2.117750439367311e-05,
      "loss": 0.9848901629447937,
      "step": 242
    },
    {
      "epoch": 0.1029535864978903,
      "grad_norm": 2.023920774459839,
      "learning_rate": 2.1353251318101936e-05,
      "loss": 0.891694188117981,
      "step": 244
    },
    {
      "epoch": 0.10379746835443038,
      "grad_norm": 1.8061069250106812,
      "learning_rate": 2.1528998242530755e-05,
      "loss": 0.9059976935386658,
      "step": 246
    },
    {
      "epoch": 0.10464135021097046,
      "grad_norm": 2.176302194595337,
      "learning_rate": 2.1704745166959578e-05,
      "loss": 1.0056109428405762,
      "step": 248
    },
    {
      "epoch": 0.10548523206751055,
      "grad_norm": 1.9820969104766846,
      "learning_rate": 2.18804920913884e-05,
      "loss": 0.9645357728004456,
      "step": 250
    },
    {
      "epoch": 0.10632911392405063,
      "grad_norm": 1.8764572143554688,
      "learning_rate": 2.2056239015817223e-05,
      "loss": 1.0178182125091553,
      "step": 252
    },
    {
      "epoch": 0.10717299578059072,
      "grad_norm": 2.56221342086792,
      "learning_rate": 2.223198594024605e-05,
      "loss": 0.9546761512756348,
      "step": 254
    },
    {
      "epoch": 0.1080168776371308,
      "grad_norm": 2.6779074668884277,
      "learning_rate": 2.2407732864674868e-05,
      "loss": 0.9300968647003174,
      "step": 256
    },
    {
      "epoch": 0.10886075949367088,
      "grad_norm": 2.140897512435913,
      "learning_rate": 2.2583479789103694e-05,
      "loss": 0.926638662815094,
      "step": 258
    },
    {
      "epoch": 0.10970464135021098,
      "grad_norm": 2.0880508422851562,
      "learning_rate": 2.2759226713532513e-05,
      "loss": 1.0681840181350708,
      "step": 260
    },
    {
      "epoch": 0.11054852320675106,
      "grad_norm": 2.7273616790771484,
      "learning_rate": 2.293497363796134e-05,
      "loss": 1.0840941667556763,
      "step": 262
    },
    {
      "epoch": 0.11139240506329114,
      "grad_norm": 1.6723874807357788,
      "learning_rate": 2.3110720562390158e-05,
      "loss": 0.8637182116508484,
      "step": 264
    },
    {
      "epoch": 0.11223628691983123,
      "grad_norm": 1.806243896484375,
      "learning_rate": 2.3286467486818984e-05,
      "loss": 0.9554686546325684,
      "step": 266
    },
    {
      "epoch": 0.11308016877637131,
      "grad_norm": 1.9086743593215942,
      "learning_rate": 2.3462214411247803e-05,
      "loss": 0.9556593894958496,
      "step": 268
    },
    {
      "epoch": 0.11392405063291139,
      "grad_norm": 2.1822304725646973,
      "learning_rate": 2.3637961335676626e-05,
      "loss": 0.9177709817886353,
      "step": 270
    },
    {
      "epoch": 0.11476793248945148,
      "grad_norm": 2.1009039878845215,
      "learning_rate": 2.3813708260105448e-05,
      "loss": 0.9288759827613831,
      "step": 272
    },
    {
      "epoch": 0.11561181434599156,
      "grad_norm": 1.9814810752868652,
      "learning_rate": 2.398945518453427e-05,
      "loss": 0.9881691932678223,
      "step": 274
    },
    {
      "epoch": 0.11645569620253164,
      "grad_norm": 1.9946284294128418,
      "learning_rate": 2.4165202108963093e-05,
      "loss": 0.9390727281570435,
      "step": 276
    },
    {
      "epoch": 0.11729957805907174,
      "grad_norm": 2.4489169120788574,
      "learning_rate": 2.4340949033391916e-05,
      "loss": 0.9625692963600159,
      "step": 278
    },
    {
      "epoch": 0.11814345991561181,
      "grad_norm": 2.0919103622436523,
      "learning_rate": 2.451669595782074e-05,
      "loss": 0.9304702877998352,
      "step": 280
    },
    {
      "epoch": 0.1189873417721519,
      "grad_norm": 1.912914752960205,
      "learning_rate": 2.469244288224956e-05,
      "loss": 0.9313994646072388,
      "step": 282
    },
    {
      "epoch": 0.11983122362869199,
      "grad_norm": 2.1553256511688232,
      "learning_rate": 2.4868189806678387e-05,
      "loss": 1.004011869430542,
      "step": 284
    },
    {
      "epoch": 0.12067510548523207,
      "grad_norm": 2.0129058361053467,
      "learning_rate": 2.504393673110721e-05,
      "loss": 0.9092531204223633,
      "step": 286
    },
    {
      "epoch": 0.12151898734177215,
      "grad_norm": 2.1632325649261475,
      "learning_rate": 2.5219683655536032e-05,
      "loss": 0.993347704410553,
      "step": 288
    },
    {
      "epoch": 0.12236286919831224,
      "grad_norm": 2.3072738647460938,
      "learning_rate": 2.539543057996485e-05,
      "loss": 0.978348433971405,
      "step": 290
    },
    {
      "epoch": 0.12320675105485232,
      "grad_norm": 2.056560516357422,
      "learning_rate": 2.5571177504393674e-05,
      "loss": 1.0018101930618286,
      "step": 292
    },
    {
      "epoch": 0.1240506329113924,
      "grad_norm": 1.8906747102737427,
      "learning_rate": 2.5746924428822493e-05,
      "loss": 0.9607775211334229,
      "step": 294
    },
    {
      "epoch": 0.1248945147679325,
      "grad_norm": 2.1375651359558105,
      "learning_rate": 2.5922671353251322e-05,
      "loss": 0.9259153008460999,
      "step": 296
    },
    {
      "epoch": 0.1257383966244726,
      "grad_norm": 1.9994823932647705,
      "learning_rate": 2.609841827768014e-05,
      "loss": 0.8524524569511414,
      "step": 298
    },
    {
      "epoch": 0.12658227848101267,
      "grad_norm": 2.2421181201934814,
      "learning_rate": 2.6274165202108964e-05,
      "loss": 1.0047069787979126,
      "step": 300
    },
    {
      "epoch": 0.12658227848101267,
      "eval_loss": 0.9517185688018799,
      "eval_runtime": 860.0287,
      "eval_samples_per_second": 2.45,
      "eval_steps_per_second": 2.45,
      "step": 300
    },
    {
      "epoch": 0.12742616033755275,
      "grad_norm": 2.1206254959106445,
      "learning_rate": 2.6449912126537786e-05,
      "loss": 0.8475471138954163,
      "step": 302
    },
    {
      "epoch": 0.12827004219409283,
      "grad_norm": 1.885161280632019,
      "learning_rate": 2.6625659050966612e-05,
      "loss": 0.8643121123313904,
      "step": 304
    },
    {
      "epoch": 0.1291139240506329,
      "grad_norm": 3.1441781520843506,
      "learning_rate": 2.680140597539543e-05,
      "loss": 0.8804612159729004,
      "step": 306
    },
    {
      "epoch": 0.12995780590717299,
      "grad_norm": 1.953133225440979,
      "learning_rate": 2.6977152899824254e-05,
      "loss": 0.8348029255867004,
      "step": 308
    },
    {
      "epoch": 0.1308016877637131,
      "grad_norm": 2.3762667179107666,
      "learning_rate": 2.7152899824253076e-05,
      "loss": 0.8889057040214539,
      "step": 310
    },
    {
      "epoch": 0.13164556962025317,
      "grad_norm": 2.4651103019714355,
      "learning_rate": 2.7328646748681902e-05,
      "loss": 1.025565505027771,
      "step": 312
    },
    {
      "epoch": 0.13248945147679325,
      "grad_norm": 1.8522284030914307,
      "learning_rate": 2.7504393673110725e-05,
      "loss": 0.868915855884552,
      "step": 314
    },
    {
      "epoch": 0.13333333333333333,
      "grad_norm": 1.8048083782196045,
      "learning_rate": 2.7680140597539544e-05,
      "loss": 0.8821638226509094,
      "step": 316
    },
    {
      "epoch": 0.1341772151898734,
      "grad_norm": 1.9933605194091797,
      "learning_rate": 2.7855887521968367e-05,
      "loss": 0.8735360503196716,
      "step": 318
    },
    {
      "epoch": 0.1350210970464135,
      "grad_norm": 2.044337034225464,
      "learning_rate": 2.8031634446397186e-05,
      "loss": 0.8288834691047668,
      "step": 320
    },
    {
      "epoch": 0.1358649789029536,
      "grad_norm": 2.416067361831665,
      "learning_rate": 2.8207381370826015e-05,
      "loss": 0.9104969501495361,
      "step": 322
    },
    {
      "epoch": 0.13670886075949368,
      "grad_norm": 2.0731265544891357,
      "learning_rate": 2.8383128295254834e-05,
      "loss": 0.8689924478530884,
      "step": 324
    },
    {
      "epoch": 0.13755274261603376,
      "grad_norm": 2.049126386642456,
      "learning_rate": 2.8558875219683657e-05,
      "loss": 0.9312222003936768,
      "step": 326
    },
    {
      "epoch": 0.13839662447257384,
      "grad_norm": 2.131026268005371,
      "learning_rate": 2.8734622144112476e-05,
      "loss": 0.8933501839637756,
      "step": 328
    },
    {
      "epoch": 0.13924050632911392,
      "grad_norm": 1.766754150390625,
      "learning_rate": 2.8910369068541305e-05,
      "loss": 0.8998261094093323,
      "step": 330
    },
    {
      "epoch": 0.140084388185654,
      "grad_norm": 2.197706460952759,
      "learning_rate": 2.9086115992970124e-05,
      "loss": 0.8826426267623901,
      "step": 332
    },
    {
      "epoch": 0.1409282700421941,
      "grad_norm": 1.953715443611145,
      "learning_rate": 2.9261862917398947e-05,
      "loss": 0.8590307831764221,
      "step": 334
    },
    {
      "epoch": 0.14177215189873418,
      "grad_norm": 2.200929880142212,
      "learning_rate": 2.943760984182777e-05,
      "loss": 0.9317060708999634,
      "step": 336
    },
    {
      "epoch": 0.14261603375527426,
      "grad_norm": 2.1195082664489746,
      "learning_rate": 2.961335676625659e-05,
      "loss": 0.9965578317642212,
      "step": 338
    },
    {
      "epoch": 0.14345991561181434,
      "grad_norm": 2.3449771404266357,
      "learning_rate": 2.9789103690685414e-05,
      "loss": 0.8353848457336426,
      "step": 340
    },
    {
      "epoch": 0.14430379746835442,
      "grad_norm": 2.000497579574585,
      "learning_rate": 2.9964850615114237e-05,
      "loss": 0.9154735803604126,
      "step": 342
    },
    {
      "epoch": 0.1451476793248945,
      "grad_norm": 2.141890525817871,
      "learning_rate": 3.014059753954306e-05,
      "loss": 0.9530655741691589,
      "step": 344
    },
    {
      "epoch": 0.1459915611814346,
      "grad_norm": 1.7717392444610596,
      "learning_rate": 3.031634446397188e-05,
      "loss": 0.896998405456543,
      "step": 346
    },
    {
      "epoch": 0.1468354430379747,
      "grad_norm": 1.8796685934066772,
      "learning_rate": 3.0492091388400708e-05,
      "loss": 0.9084208011627197,
      "step": 348
    },
    {
      "epoch": 0.14767932489451477,
      "grad_norm": 2.0298709869384766,
      "learning_rate": 3.066783831282953e-05,
      "loss": 0.9183387756347656,
      "step": 350
    },
    {
      "epoch": 0.14852320675105485,
      "grad_norm": 1.9245645999908447,
      "learning_rate": 3.084358523725835e-05,
      "loss": 0.8624772429466248,
      "step": 352
    },
    {
      "epoch": 0.14936708860759493,
      "grad_norm": 2.325681209564209,
      "learning_rate": 3.101933216168717e-05,
      "loss": 0.9142400026321411,
      "step": 354
    },
    {
      "epoch": 0.150210970464135,
      "grad_norm": 2.1200530529022217,
      "learning_rate": 3.1195079086115995e-05,
      "loss": 0.9064018130302429,
      "step": 356
    },
    {
      "epoch": 0.15105485232067511,
      "grad_norm": 1.979314923286438,
      "learning_rate": 3.137082601054482e-05,
      "loss": 0.9199238419532776,
      "step": 358
    },
    {
      "epoch": 0.1518987341772152,
      "grad_norm": 2.1122689247131348,
      "learning_rate": 3.154657293497364e-05,
      "loss": 0.8030132055282593,
      "step": 360
    },
    {
      "epoch": 0.15274261603375527,
      "grad_norm": 2.105767250061035,
      "learning_rate": 3.172231985940246e-05,
      "loss": 0.9185854196548462,
      "step": 362
    },
    {
      "epoch": 0.15358649789029535,
      "grad_norm": 2.179471015930176,
      "learning_rate": 3.1898066783831285e-05,
      "loss": 0.9365083575248718,
      "step": 364
    },
    {
      "epoch": 0.15443037974683543,
      "grad_norm": 2.1444311141967773,
      "learning_rate": 3.207381370826011e-05,
      "loss": 0.8965140581130981,
      "step": 366
    },
    {
      "epoch": 0.15527426160337554,
      "grad_norm": 2.4171674251556396,
      "learning_rate": 3.224956063268893e-05,
      "loss": 0.8787504434585571,
      "step": 368
    },
    {
      "epoch": 0.15611814345991562,
      "grad_norm": 2.418628215789795,
      "learning_rate": 3.242530755711775e-05,
      "loss": 0.8925284147262573,
      "step": 370
    },
    {
      "epoch": 0.1569620253164557,
      "grad_norm": 2.2228314876556396,
      "learning_rate": 3.2601054481546575e-05,
      "loss": 0.876179039478302,
      "step": 372
    },
    {
      "epoch": 0.15780590717299578,
      "grad_norm": 2.324237108230591,
      "learning_rate": 3.27768014059754e-05,
      "loss": 0.8365707993507385,
      "step": 374
    },
    {
      "epoch": 0.15864978902953586,
      "grad_norm": 2.6344552040100098,
      "learning_rate": 3.295254833040422e-05,
      "loss": 0.7864399552345276,
      "step": 376
    },
    {
      "epoch": 0.15949367088607594,
      "grad_norm": 2.047536611557007,
      "learning_rate": 3.312829525483304e-05,
      "loss": 0.9271875023841858,
      "step": 378
    },
    {
      "epoch": 0.16033755274261605,
      "grad_norm": 2.120025157928467,
      "learning_rate": 3.3304042179261865e-05,
      "loss": 0.8799133896827698,
      "step": 380
    },
    {
      "epoch": 0.16118143459915613,
      "grad_norm": 2.363692045211792,
      "learning_rate": 3.347978910369069e-05,
      "loss": 0.8973530530929565,
      "step": 382
    },
    {
      "epoch": 0.1620253164556962,
      "grad_norm": 2.1796772480010986,
      "learning_rate": 3.365553602811951e-05,
      "loss": 1.0277652740478516,
      "step": 384
    },
    {
      "epoch": 0.16286919831223629,
      "grad_norm": 1.9192595481872559,
      "learning_rate": 3.383128295254833e-05,
      "loss": 0.8909643888473511,
      "step": 386
    },
    {
      "epoch": 0.16371308016877636,
      "grad_norm": 1.7874376773834229,
      "learning_rate": 3.4007029876977155e-05,
      "loss": 0.837049663066864,
      "step": 388
    },
    {
      "epoch": 0.16455696202531644,
      "grad_norm": 2.3402366638183594,
      "learning_rate": 3.4182776801405974e-05,
      "loss": 0.8625202775001526,
      "step": 390
    },
    {
      "epoch": 0.16540084388185655,
      "grad_norm": 2.1137185096740723,
      "learning_rate": 3.43585237258348e-05,
      "loss": 0.9288321137428284,
      "step": 392
    },
    {
      "epoch": 0.16624472573839663,
      "grad_norm": 2.3776895999908447,
      "learning_rate": 3.453427065026362e-05,
      "loss": 0.9328726530075073,
      "step": 394
    },
    {
      "epoch": 0.1670886075949367,
      "grad_norm": 2.34941029548645,
      "learning_rate": 3.4710017574692445e-05,
      "loss": 0.9273309707641602,
      "step": 396
    },
    {
      "epoch": 0.1679324894514768,
      "grad_norm": 2.1272573471069336,
      "learning_rate": 3.4885764499121264e-05,
      "loss": 0.8703887462615967,
      "step": 398
    },
    {
      "epoch": 0.16877637130801687,
      "grad_norm": 2.047290802001953,
      "learning_rate": 3.506151142355009e-05,
      "loss": 0.8808165788650513,
      "step": 400
    },
    {
      "epoch": 0.16877637130801687,
      "eval_loss": 0.9282881617546082,
      "eval_runtime": 869.6867,
      "eval_samples_per_second": 2.423,
      "eval_steps_per_second": 2.423,
      "step": 400
    },
    {
      "epoch": 0.16962025316455695,
      "grad_norm": 1.9874159097671509,
      "learning_rate": 3.5237258347978916e-05,
      "loss": 0.9643645286560059,
      "step": 402
    },
    {
      "epoch": 0.17046413502109706,
      "grad_norm": 1.9299919605255127,
      "learning_rate": 3.5413005272407735e-05,
      "loss": 0.9173495769500732,
      "step": 404
    },
    {
      "epoch": 0.17130801687763714,
      "grad_norm": 2.3379697799682617,
      "learning_rate": 3.5588752196836555e-05,
      "loss": 0.8998411893844604,
      "step": 406
    },
    {
      "epoch": 0.17215189873417722,
      "grad_norm": 2.241370916366577,
      "learning_rate": 3.5764499121265374e-05,
      "loss": 0.9310802221298218,
      "step": 408
    },
    {
      "epoch": 0.1729957805907173,
      "grad_norm": 2.4490108489990234,
      "learning_rate": 3.5940246045694206e-05,
      "loss": 0.9605053067207336,
      "step": 410
    },
    {
      "epoch": 0.17383966244725738,
      "grad_norm": 1.8247230052947998,
      "learning_rate": 3.6115992970123026e-05,
      "loss": 0.8485683798789978,
      "step": 412
    },
    {
      "epoch": 0.17468354430379746,
      "grad_norm": 2.4608843326568604,
      "learning_rate": 3.6291739894551845e-05,
      "loss": 0.9325968623161316,
      "step": 414
    },
    {
      "epoch": 0.17552742616033756,
      "grad_norm": 1.8923161029815674,
      "learning_rate": 3.646748681898067e-05,
      "loss": 0.9125096201896667,
      "step": 416
    },
    {
      "epoch": 0.17637130801687764,
      "grad_norm": 1.8502769470214844,
      "learning_rate": 3.6643233743409497e-05,
      "loss": 0.8852217197418213,
      "step": 418
    },
    {
      "epoch": 0.17721518987341772,
      "grad_norm": 1.9155100584030151,
      "learning_rate": 3.6818980667838316e-05,
      "loss": 0.9192792773246765,
      "step": 420
    },
    {
      "epoch": 0.1780590717299578,
      "grad_norm": 2.181476593017578,
      "learning_rate": 3.6994727592267135e-05,
      "loss": 0.8787404298782349,
      "step": 422
    },
    {
      "epoch": 0.17890295358649788,
      "grad_norm": 2.2469847202301025,
      "learning_rate": 3.717047451669596e-05,
      "loss": 0.9109582901000977,
      "step": 424
    },
    {
      "epoch": 0.17974683544303796,
      "grad_norm": 2.08145809173584,
      "learning_rate": 3.734622144112479e-05,
      "loss": 0.8560389280319214,
      "step": 426
    },
    {
      "epoch": 0.18059071729957807,
      "grad_norm": 4.121932506561279,
      "learning_rate": 3.7521968365553606e-05,
      "loss": 0.9456104040145874,
      "step": 428
    },
    {
      "epoch": 0.18143459915611815,
      "grad_norm": 2.177459478378296,
      "learning_rate": 3.7697715289982425e-05,
      "loss": 0.8421300649642944,
      "step": 430
    },
    {
      "epoch": 0.18227848101265823,
      "grad_norm": 2.324970245361328,
      "learning_rate": 3.787346221441125e-05,
      "loss": 0.9199858903884888,
      "step": 432
    },
    {
      "epoch": 0.1831223628691983,
      "grad_norm": 2.133718490600586,
      "learning_rate": 3.804920913884007e-05,
      "loss": 0.8953126668930054,
      "step": 434
    },
    {
      "epoch": 0.1839662447257384,
      "grad_norm": 1.8527995347976685,
      "learning_rate": 3.8224956063268896e-05,
      "loss": 0.8732239007949829,
      "step": 436
    },
    {
      "epoch": 0.1848101265822785,
      "grad_norm": 1.95817232131958,
      "learning_rate": 3.8400702987697715e-05,
      "loss": 0.8818746209144592,
      "step": 438
    },
    {
      "epoch": 0.18565400843881857,
      "grad_norm": 2.2107293605804443,
      "learning_rate": 3.857644991212654e-05,
      "loss": 0.9153507947921753,
      "step": 440
    },
    {
      "epoch": 0.18649789029535865,
      "grad_norm": 2.004754066467285,
      "learning_rate": 3.875219683655536e-05,
      "loss": 0.8960154056549072,
      "step": 442
    },
    {
      "epoch": 0.18734177215189873,
      "grad_norm": 2.1851706504821777,
      "learning_rate": 3.8927943760984186e-05,
      "loss": 0.909011721611023,
      "step": 444
    },
    {
      "epoch": 0.1881856540084388,
      "grad_norm": 2.4492485523223877,
      "learning_rate": 3.9103690685413005e-05,
      "loss": 0.8880158066749573,
      "step": 446
    },
    {
      "epoch": 0.1890295358649789,
      "grad_norm": 2.745453119277954,
      "learning_rate": 3.927943760984183e-05,
      "loss": 0.8500842452049255,
      "step": 448
    },
    {
      "epoch": 0.189873417721519,
      "grad_norm": 2.1924264430999756,
      "learning_rate": 3.945518453427065e-05,
      "loss": 0.9004045724868774,
      "step": 450
    },
    {
      "epoch": 0.19071729957805908,
      "grad_norm": 2.4051687717437744,
      "learning_rate": 3.9630931458699476e-05,
      "loss": 0.9020664095878601,
      "step": 452
    },
    {
      "epoch": 0.19156118143459916,
      "grad_norm": 1.8077667951583862,
      "learning_rate": 3.9806678383128295e-05,
      "loss": 0.8639500737190247,
      "step": 454
    },
    {
      "epoch": 0.19240506329113924,
      "grad_norm": 2.089043378829956,
      "learning_rate": 3.998242530755712e-05,
      "loss": 0.8642048239707947,
      "step": 456
    },
    {
      "epoch": 0.19324894514767932,
      "grad_norm": 2.029578447341919,
      "learning_rate": 4.015817223198594e-05,
      "loss": 0.9371927380561829,
      "step": 458
    },
    {
      "epoch": 0.1940928270042194,
      "grad_norm": 2.26582407951355,
      "learning_rate": 4.033391915641476e-05,
      "loss": 0.9120588302612305,
      "step": 460
    },
    {
      "epoch": 0.1949367088607595,
      "grad_norm": 1.8671411275863647,
      "learning_rate": 4.050966608084359e-05,
      "loss": 0.8758644461631775,
      "step": 462
    },
    {
      "epoch": 0.19578059071729959,
      "grad_norm": 1.9403492212295532,
      "learning_rate": 4.068541300527241e-05,
      "loss": 0.914577305316925,
      "step": 464
    },
    {
      "epoch": 0.19662447257383966,
      "grad_norm": 1.9939641952514648,
      "learning_rate": 4.086115992970123e-05,
      "loss": 0.8592531681060791,
      "step": 466
    },
    {
      "epoch": 0.19746835443037974,
      "grad_norm": 2.1511380672454834,
      "learning_rate": 4.103690685413005e-05,
      "loss": 0.9251965880393982,
      "step": 468
    },
    {
      "epoch": 0.19831223628691982,
      "grad_norm": 2.2260982990264893,
      "learning_rate": 4.121265377855888e-05,
      "loss": 0.8465172052383423,
      "step": 470
    },
    {
      "epoch": 0.1991561181434599,
      "grad_norm": 2.0510010719299316,
      "learning_rate": 4.13884007029877e-05,
      "loss": 0.8943672180175781,
      "step": 472
    },
    {
      "epoch": 0.2,
      "grad_norm": 2.2040133476257324,
      "learning_rate": 4.156414762741652e-05,
      "loss": 0.9594319462776184,
      "step": 474
    },
    {
      "epoch": 0.2008438818565401,
      "grad_norm": 2.355181932449341,
      "learning_rate": 4.173989455184534e-05,
      "loss": 0.9031813144683838,
      "step": 476
    },
    {
      "epoch": 0.20168776371308017,
      "grad_norm": 2.8434665203094482,
      "learning_rate": 4.1915641476274166e-05,
      "loss": 0.9225798845291138,
      "step": 478
    },
    {
      "epoch": 0.20253164556962025,
      "grad_norm": 2.1715340614318848,
      "learning_rate": 4.209138840070299e-05,
      "loss": 0.894163966178894,
      "step": 480
    },
    {
      "epoch": 0.20337552742616033,
      "grad_norm": 2.078916072845459,
      "learning_rate": 4.226713532513181e-05,
      "loss": 0.8424109816551208,
      "step": 482
    },
    {
      "epoch": 0.2042194092827004,
      "grad_norm": 1.9760961532592773,
      "learning_rate": 4.244288224956064e-05,
      "loss": 0.9102715849876404,
      "step": 484
    },
    {
      "epoch": 0.20506329113924052,
      "grad_norm": 1.9684507846832275,
      "learning_rate": 4.2618629173989456e-05,
      "loss": 0.8693854808807373,
      "step": 486
    },
    {
      "epoch": 0.2059071729957806,
      "grad_norm": 2.1633450984954834,
      "learning_rate": 4.279437609841828e-05,
      "loss": 0.8617543578147888,
      "step": 488
    },
    {
      "epoch": 0.20675105485232068,
      "grad_norm": 2.2695257663726807,
      "learning_rate": 4.29701230228471e-05,
      "loss": 0.9167086482048035,
      "step": 490
    },
    {
      "epoch": 0.20759493670886076,
      "grad_norm": 2.4180049896240234,
      "learning_rate": 4.314586994727593e-05,
      "loss": 0.8333520889282227,
      "step": 492
    },
    {
      "epoch": 0.20843881856540084,
      "grad_norm": 2.2942769527435303,
      "learning_rate": 4.3321616871704746e-05,
      "loss": 0.918351411819458,
      "step": 494
    },
    {
      "epoch": 0.20928270042194091,
      "grad_norm": 1.826458215713501,
      "learning_rate": 4.349736379613357e-05,
      "loss": 0.8565171957015991,
      "step": 496
    },
    {
      "epoch": 0.21012658227848102,
      "grad_norm": 1.9694055318832397,
      "learning_rate": 4.367311072056239e-05,
      "loss": 0.8684167861938477,
      "step": 498
    },
    {
      "epoch": 0.2109704641350211,
      "grad_norm": 1.892659306526184,
      "learning_rate": 4.384885764499122e-05,
      "loss": 0.7752788662910461,
      "step": 500
    },
    {
      "epoch": 0.2109704641350211,
      "eval_loss": 0.9080732464790344,
      "eval_runtime": 857.0753,
      "eval_samples_per_second": 2.458,
      "eval_steps_per_second": 2.458,
      "step": 500
    },
    {
      "epoch": 0.21181434599156118,
      "grad_norm": 1.9322253465652466,
      "learning_rate": 4.4024604569420036e-05,
      "loss": 0.948570728302002,
      "step": 502
    },
    {
      "epoch": 0.21265822784810126,
      "grad_norm": 2.0456058979034424,
      "learning_rate": 4.4200351493848855e-05,
      "loss": 0.8741024732589722,
      "step": 504
    },
    {
      "epoch": 0.21350210970464134,
      "grad_norm": 2.2406177520751953,
      "learning_rate": 4.437609841827768e-05,
      "loss": 0.9053841829299927,
      "step": 506
    },
    {
      "epoch": 0.21434599156118145,
      "grad_norm": 2.013934850692749,
      "learning_rate": 4.455184534270651e-05,
      "loss": 0.8886576294898987,
      "step": 508
    },
    {
      "epoch": 0.21518987341772153,
      "grad_norm": 1.9771125316619873,
      "learning_rate": 4.4727592267135326e-05,
      "loss": 0.8834167718887329,
      "step": 510
    },
    {
      "epoch": 0.2160337552742616,
      "grad_norm": 1.785905361175537,
      "learning_rate": 4.4903339191564146e-05,
      "loss": 0.7938863039016724,
      "step": 512
    },
    {
      "epoch": 0.2168776371308017,
      "grad_norm": 1.7946031093597412,
      "learning_rate": 4.507908611599297e-05,
      "loss": 0.8071596026420593,
      "step": 514
    },
    {
      "epoch": 0.21772151898734177,
      "grad_norm": 2.2217721939086914,
      "learning_rate": 4.52548330404218e-05,
      "loss": 0.797417163848877,
      "step": 516
    },
    {
      "epoch": 0.21856540084388185,
      "grad_norm": 1.9022471904754639,
      "learning_rate": 4.5430579964850617e-05,
      "loss": 0.8109536170959473,
      "step": 518
    },
    {
      "epoch": 0.21940928270042195,
      "grad_norm": 1.8988343477249146,
      "learning_rate": 4.5606326889279436e-05,
      "loss": 0.8647034168243408,
      "step": 520
    },
    {
      "epoch": 0.22025316455696203,
      "grad_norm": 2.6014881134033203,
      "learning_rate": 4.578207381370827e-05,
      "loss": 0.8763713240623474,
      "step": 522
    },
    {
      "epoch": 0.2210970464135021,
      "grad_norm": 1.9512032270431519,
      "learning_rate": 4.595782073813709e-05,
      "loss": 0.9525764584541321,
      "step": 524
    },
    {
      "epoch": 0.2219409282700422,
      "grad_norm": 1.9246160984039307,
      "learning_rate": 4.613356766256591e-05,
      "loss": 0.8839208483695984,
      "step": 526
    },
    {
      "epoch": 0.22278481012658227,
      "grad_norm": 1.9713703393936157,
      "learning_rate": 4.6309314586994726e-05,
      "loss": 0.8888868093490601,
      "step": 528
    },
    {
      "epoch": 0.22362869198312235,
      "grad_norm": 2.1175239086151123,
      "learning_rate": 4.648506151142355e-05,
      "loss": 0.8123540878295898,
      "step": 530
    },
    {
      "epoch": 0.22447257383966246,
      "grad_norm": 1.7656135559082031,
      "learning_rate": 4.666080843585238e-05,
      "loss": 0.7447702884674072,
      "step": 532
    },
    {
      "epoch": 0.22531645569620254,
      "grad_norm": 2.15748929977417,
      "learning_rate": 4.68365553602812e-05,
      "loss": 0.8778411746025085,
      "step": 534
    },
    {
      "epoch": 0.22616033755274262,
      "grad_norm": 2.1733345985412598,
      "learning_rate": 4.7012302284710016e-05,
      "loss": 0.8985894918441772,
      "step": 536
    },
    {
      "epoch": 0.2270042194092827,
      "grad_norm": 1.7182204723358154,
      "learning_rate": 4.718804920913884e-05,
      "loss": 0.8031114339828491,
      "step": 538
    },
    {
      "epoch": 0.22784810126582278,
      "grad_norm": 1.8586329221725464,
      "learning_rate": 4.736379613356767e-05,
      "loss": 0.9399706721305847,
      "step": 540
    },
    {
      "epoch": 0.22869198312236286,
      "grad_norm": 2.105637311935425,
      "learning_rate": 4.753954305799649e-05,
      "loss": 0.8672119975090027,
      "step": 542
    },
    {
      "epoch": 0.22953586497890296,
      "grad_norm": 1.760584831237793,
      "learning_rate": 4.771528998242531e-05,
      "loss": 0.8663905262947083,
      "step": 544
    },
    {
      "epoch": 0.23037974683544304,
      "grad_norm": 1.579990267753601,
      "learning_rate": 4.789103690685413e-05,
      "loss": 0.8575801849365234,
      "step": 546
    },
    {
      "epoch": 0.23122362869198312,
      "grad_norm": 1.9242485761642456,
      "learning_rate": 4.806678383128295e-05,
      "loss": 0.828412652015686,
      "step": 548
    },
    {
      "epoch": 0.2320675105485232,
      "grad_norm": 1.812137246131897,
      "learning_rate": 4.824253075571178e-05,
      "loss": 0.8183464407920837,
      "step": 550
    },
    {
      "epoch": 0.23291139240506328,
      "grad_norm": 1.804733395576477,
      "learning_rate": 4.84182776801406e-05,
      "loss": 0.7822491526603699,
      "step": 552
    },
    {
      "epoch": 0.23375527426160336,
      "grad_norm": 2.052257537841797,
      "learning_rate": 4.859402460456942e-05,
      "loss": 0.9050943851470947,
      "step": 554
    },
    {
      "epoch": 0.23459915611814347,
      "grad_norm": 1.9803621768951416,
      "learning_rate": 4.876977152899824e-05,
      "loss": 0.8846852779388428,
      "step": 556
    },
    {
      "epoch": 0.23544303797468355,
      "grad_norm": 1.820125937461853,
      "learning_rate": 4.894551845342707e-05,
      "loss": 0.8649531602859497,
      "step": 558
    },
    {
      "epoch": 0.23628691983122363,
      "grad_norm": 2.0963921546936035,
      "learning_rate": 4.912126537785589e-05,
      "loss": 0.9307748079299927,
      "step": 560
    },
    {
      "epoch": 0.2371308016877637,
      "grad_norm": 2.079697847366333,
      "learning_rate": 4.929701230228471e-05,
      "loss": 0.9092473387718201,
      "step": 562
    },
    {
      "epoch": 0.2379746835443038,
      "grad_norm": 2.0291287899017334,
      "learning_rate": 4.947275922671353e-05,
      "loss": 0.8976567983627319,
      "step": 564
    },
    {
      "epoch": 0.23881856540084387,
      "grad_norm": 1.9636707305908203,
      "learning_rate": 4.964850615114236e-05,
      "loss": 0.8931006193161011,
      "step": 566
    },
    {
      "epoch": 0.23966244725738398,
      "grad_norm": 1.922049880027771,
      "learning_rate": 4.982425307557118e-05,
      "loss": 0.829562246799469,
      "step": 568
    },
    {
      "epoch": 0.24050632911392406,
      "grad_norm": 2.150334596633911,
      "learning_rate": 5e-05,
      "loss": 0.8568030595779419,
      "step": 570
    },
    {
      "epoch": 0.24135021097046414,
      "grad_norm": 2.024437427520752,
      "learning_rate": 5.017574692442882e-05,
      "loss": 0.8623508810997009,
      "step": 572
    },
    {
      "epoch": 0.24219409282700421,
      "grad_norm": 1.8312673568725586,
      "learning_rate": 5.035149384885765e-05,
      "loss": 0.7853795886039734,
      "step": 574
    },
    {
      "epoch": 0.2430379746835443,
      "grad_norm": 1.9271961450576782,
      "learning_rate": 5.0527240773286467e-05,
      "loss": 0.9727587103843689,
      "step": 576
    },
    {
      "epoch": 0.2438818565400844,
      "grad_norm": 1.931249976158142,
      "learning_rate": 5.0702987697715286e-05,
      "loss": 0.8859632015228271,
      "step": 578
    },
    {
      "epoch": 0.24472573839662448,
      "grad_norm": 1.8195210695266724,
      "learning_rate": 5.087873462214412e-05,
      "loss": 0.8959492444992065,
      "step": 580
    },
    {
      "epoch": 0.24556962025316456,
      "grad_norm": 2.0018749237060547,
      "learning_rate": 5.105448154657294e-05,
      "loss": 0.8146185874938965,
      "step": 582
    },
    {
      "epoch": 0.24641350210970464,
      "grad_norm": 2.09798526763916,
      "learning_rate": 5.1230228471001764e-05,
      "loss": 0.8545317053794861,
      "step": 584
    },
    {
      "epoch": 0.24725738396624472,
      "grad_norm": 1.8063944578170776,
      "learning_rate": 5.140597539543058e-05,
      "loss": 0.8650105595588684,
      "step": 586
    },
    {
      "epoch": 0.2481012658227848,
      "grad_norm": 1.8535740375518799,
      "learning_rate": 5.15817223198594e-05,
      "loss": 0.8395693302154541,
      "step": 588
    },
    {
      "epoch": 0.2489451476793249,
      "grad_norm": 2.1443960666656494,
      "learning_rate": 5.175746924428823e-05,
      "loss": 0.8267397284507751,
      "step": 590
    },
    {
      "epoch": 0.249789029535865,
      "grad_norm": 1.9637391567230225,
      "learning_rate": 5.193321616871705e-05,
      "loss": 0.8500015139579773,
      "step": 592
    },
    {
      "epoch": 0.25063291139240507,
      "grad_norm": 1.9457582235336304,
      "learning_rate": 5.2108963093145866e-05,
      "loss": 0.887481153011322,
      "step": 594
    },
    {
      "epoch": 0.2514767932489452,
      "grad_norm": 1.7458715438842773,
      "learning_rate": 5.228471001757469e-05,
      "loss": 0.8444154858589172,
      "step": 596
    },
    {
      "epoch": 0.2523206751054852,
      "grad_norm": 1.8341439962387085,
      "learning_rate": 5.2460456942003525e-05,
      "loss": 0.8301781415939331,
      "step": 598
    },
    {
      "epoch": 0.25316455696202533,
      "grad_norm": 2.127747058868408,
      "learning_rate": 5.2636203866432344e-05,
      "loss": 0.8921551704406738,
      "step": 600
    },
    {
      "epoch": 0.25316455696202533,
      "eval_loss": 0.8903881311416626,
      "eval_runtime": 845.9969,
      "eval_samples_per_second": 2.491,
      "eval_steps_per_second": 2.491,
      "step": 600
    },
    {
      "epoch": 0.2540084388185654,
      "grad_norm": 2.421459674835205,
      "learning_rate": 5.281195079086116e-05,
      "loss": 0.8678019642829895,
      "step": 602
    },
    {
      "epoch": 0.2548523206751055,
      "grad_norm": 1.7736057043075562,
      "learning_rate": 5.298769771528999e-05,
      "loss": 0.8564275503158569,
      "step": 604
    },
    {
      "epoch": 0.25569620253164554,
      "grad_norm": 2.28430438041687,
      "learning_rate": 5.316344463971881e-05,
      "loss": 0.8529049158096313,
      "step": 606
    },
    {
      "epoch": 0.25654008438818565,
      "grad_norm": 1.8892366886138916,
      "learning_rate": 5.333919156414763e-05,
      "loss": 0.8672881126403809,
      "step": 608
    },
    {
      "epoch": 0.25738396624472576,
      "grad_norm": 1.9059702157974243,
      "learning_rate": 5.3514938488576446e-05,
      "loss": 0.9094445109367371,
      "step": 610
    },
    {
      "epoch": 0.2582278481012658,
      "grad_norm": 2.0657339096069336,
      "learning_rate": 5.369068541300527e-05,
      "loss": 0.8361946940422058,
      "step": 612
    },
    {
      "epoch": 0.2590717299578059,
      "grad_norm": 1.8987553119659424,
      "learning_rate": 5.3866432337434105e-05,
      "loss": 0.8319925665855408,
      "step": 614
    },
    {
      "epoch": 0.25991561181434597,
      "grad_norm": 2.1176226139068604,
      "learning_rate": 5.4042179261862924e-05,
      "loss": 0.9818069934844971,
      "step": 616
    },
    {
      "epoch": 0.2607594936708861,
      "grad_norm": 2.142096519470215,
      "learning_rate": 5.421792618629174e-05,
      "loss": 0.8675919771194458,
      "step": 618
    },
    {
      "epoch": 0.2616033755274262,
      "grad_norm": 1.9527089595794678,
      "learning_rate": 5.439367311072057e-05,
      "loss": 0.8845479488372803,
      "step": 620
    },
    {
      "epoch": 0.26244725738396624,
      "grad_norm": 1.7071453332901,
      "learning_rate": 5.456942003514939e-05,
      "loss": 0.809393048286438,
      "step": 622
    },
    {
      "epoch": 0.26329113924050634,
      "grad_norm": 1.9133527278900146,
      "learning_rate": 5.474516695957821e-05,
      "loss": 0.8262377977371216,
      "step": 624
    },
    {
      "epoch": 0.2641350210970464,
      "grad_norm": 2.0217554569244385,
      "learning_rate": 5.492091388400703e-05,
      "loss": 0.9006736278533936,
      "step": 626
    },
    {
      "epoch": 0.2649789029535865,
      "grad_norm": 1.773273229598999,
      "learning_rate": 5.509666080843585e-05,
      "loss": 0.8243603110313416,
      "step": 628
    },
    {
      "epoch": 0.26582278481012656,
      "grad_norm": 1.6580880880355835,
      "learning_rate": 5.527240773286467e-05,
      "loss": 0.8112778663635254,
      "step": 630
    },
    {
      "epoch": 0.26666666666666666,
      "grad_norm": 1.8342082500457764,
      "learning_rate": 5.5448154657293504e-05,
      "loss": 0.8390820622444153,
      "step": 632
    },
    {
      "epoch": 0.26751054852320677,
      "grad_norm": 1.863695502281189,
      "learning_rate": 5.5623901581722323e-05,
      "loss": 0.8264521360397339,
      "step": 634
    },
    {
      "epoch": 0.2683544303797468,
      "grad_norm": 1.9462928771972656,
      "learning_rate": 5.579964850615115e-05,
      "loss": 0.9512701630592346,
      "step": 636
    },
    {
      "epoch": 0.26919831223628693,
      "grad_norm": 1.7776058912277222,
      "learning_rate": 5.597539543057997e-05,
      "loss": 0.9422703981399536,
      "step": 638
    },
    {
      "epoch": 0.270042194092827,
      "grad_norm": 2.9457077980041504,
      "learning_rate": 5.615114235500879e-05,
      "loss": 0.7991042137145996,
      "step": 640
    },
    {
      "epoch": 0.2708860759493671,
      "grad_norm": 1.445265531539917,
      "learning_rate": 5.6326889279437614e-05,
      "loss": 0.8188099265098572,
      "step": 642
    },
    {
      "epoch": 0.2717299578059072,
      "grad_norm": 2.063850164413452,
      "learning_rate": 5.650263620386643e-05,
      "loss": 0.9799772500991821,
      "step": 644
    },
    {
      "epoch": 0.27257383966244725,
      "grad_norm": 2.0488009452819824,
      "learning_rate": 5.667838312829525e-05,
      "loss": 0.8462742567062378,
      "step": 646
    },
    {
      "epoch": 0.27341772151898736,
      "grad_norm": 1.8747851848602295,
      "learning_rate": 5.685413005272408e-05,
      "loss": 0.8226412534713745,
      "step": 648
    },
    {
      "epoch": 0.2742616033755274,
      "grad_norm": 1.849074125289917,
      "learning_rate": 5.702987697715291e-05,
      "loss": 0.9146338105201721,
      "step": 650
    },
    {
      "epoch": 0.2751054852320675,
      "grad_norm": 1.7738500833511353,
      "learning_rate": 5.720562390158173e-05,
      "loss": 0.7574424147605896,
      "step": 652
    },
    {
      "epoch": 0.2759493670886076,
      "grad_norm": 1.911102294921875,
      "learning_rate": 5.738137082601055e-05,
      "loss": 0.8930003046989441,
      "step": 654
    },
    {
      "epoch": 0.2767932489451477,
      "grad_norm": 1.5716617107391357,
      "learning_rate": 5.755711775043937e-05,
      "loss": 0.7578965425491333,
      "step": 656
    },
    {
      "epoch": 0.2776371308016878,
      "grad_norm": 1.789036512374878,
      "learning_rate": 5.7732864674868194e-05,
      "loss": 0.8149038553237915,
      "step": 658
    },
    {
      "epoch": 0.27848101265822783,
      "grad_norm": 1.68622624874115,
      "learning_rate": 5.790861159929701e-05,
      "loss": 0.8265765905380249,
      "step": 660
    },
    {
      "epoch": 0.27932489451476794,
      "grad_norm": 2.078423261642456,
      "learning_rate": 5.808435852372583e-05,
      "loss": 0.9651970267295837,
      "step": 662
    },
    {
      "epoch": 0.280168776371308,
      "grad_norm": 1.7878645658493042,
      "learning_rate": 5.826010544815466e-05,
      "loss": 0.8295148015022278,
      "step": 664
    },
    {
      "epoch": 0.2810126582278481,
      "grad_norm": 1.970838189125061,
      "learning_rate": 5.843585237258348e-05,
      "loss": 0.7778491377830505,
      "step": 666
    },
    {
      "epoch": 0.2818565400843882,
      "grad_norm": 1.943596363067627,
      "learning_rate": 5.861159929701231e-05,
      "loss": 0.9818071722984314,
      "step": 668
    },
    {
      "epoch": 0.28270042194092826,
      "grad_norm": 1.8793812990188599,
      "learning_rate": 5.878734622144113e-05,
      "loss": 0.9297797083854675,
      "step": 670
    },
    {
      "epoch": 0.28354430379746837,
      "grad_norm": 1.8813483715057373,
      "learning_rate": 5.8963093145869955e-05,
      "loss": 0.8748109936714172,
      "step": 672
    },
    {
      "epoch": 0.2843881856540084,
      "grad_norm": 1.7658562660217285,
      "learning_rate": 5.9138840070298774e-05,
      "loss": 0.8505244851112366,
      "step": 674
    },
    {
      "epoch": 0.2852320675105485,
      "grad_norm": 1.6767617464065552,
      "learning_rate": 5.931458699472759e-05,
      "loss": 0.8476597666740417,
      "step": 676
    },
    {
      "epoch": 0.28607594936708863,
      "grad_norm": 2.703104257583618,
      "learning_rate": 5.949033391915641e-05,
      "loss": 0.8775192499160767,
      "step": 678
    },
    {
      "epoch": 0.2869198312236287,
      "grad_norm": 1.9959728717803955,
      "learning_rate": 5.966608084358524e-05,
      "loss": 0.855262279510498,
      "step": 680
    },
    {
      "epoch": 0.2877637130801688,
      "grad_norm": 1.9093716144561768,
      "learning_rate": 5.984182776801406e-05,
      "loss": 0.7574936151504517,
      "step": 682
    },
    {
      "epoch": 0.28860759493670884,
      "grad_norm": 1.9829599857330322,
      "learning_rate": 6.001757469244289e-05,
      "loss": 0.8630690574645996,
      "step": 684
    },
    {
      "epoch": 0.28945147679324895,
      "grad_norm": 1.8777490854263306,
      "learning_rate": 6.019332161687171e-05,
      "loss": 0.8513249158859253,
      "step": 686
    },
    {
      "epoch": 0.290295358649789,
      "grad_norm": 1.9453173875808716,
      "learning_rate": 6.0369068541300535e-05,
      "loss": 0.9097008109092712,
      "step": 688
    },
    {
      "epoch": 0.2911392405063291,
      "grad_norm": 1.8527908325195312,
      "learning_rate": 6.0544815465729354e-05,
      "loss": 0.8291722536087036,
      "step": 690
    },
    {
      "epoch": 0.2919831223628692,
      "grad_norm": 1.9255812168121338,
      "learning_rate": 6.0720562390158174e-05,
      "loss": 0.880009651184082,
      "step": 692
    },
    {
      "epoch": 0.29282700421940927,
      "grad_norm": 1.6637977361679077,
      "learning_rate": 6.0896309314587e-05,
      "loss": 0.8791794180870056,
      "step": 694
    },
    {
      "epoch": 0.2936708860759494,
      "grad_norm": 1.825940728187561,
      "learning_rate": 6.107205623901582e-05,
      "loss": 0.8662407398223877,
      "step": 696
    },
    {
      "epoch": 0.29451476793248943,
      "grad_norm": 1.9348198175430298,
      "learning_rate": 6.124780316344464e-05,
      "loss": 0.8984515070915222,
      "step": 698
    },
    {
      "epoch": 0.29535864978902954,
      "grad_norm": 1.659345030784607,
      "learning_rate": 6.142355008787346e-05,
      "loss": 0.827385663986206,
      "step": 700
    },
    {
      "epoch": 0.29535864978902954,
      "eval_loss": 0.8730722069740295,
      "eval_runtime": 858.184,
      "eval_samples_per_second": 2.455,
      "eval_steps_per_second": 2.455,
      "step": 700
    },
    {
      "epoch": 0.29620253164556964,
      "grad_norm": 1.6531789302825928,
      "learning_rate": 6.159929701230229e-05,
      "loss": 0.9337764382362366,
      "step": 702
    },
    {
      "epoch": 0.2970464135021097,
      "grad_norm": 1.8269121646881104,
      "learning_rate": 6.177504393673111e-05,
      "loss": 0.8250943422317505,
      "step": 704
    },
    {
      "epoch": 0.2978902953586498,
      "grad_norm": 1.692808747291565,
      "learning_rate": 6.195079086115994e-05,
      "loss": 0.8657428026199341,
      "step": 706
    },
    {
      "epoch": 0.29873417721518986,
      "grad_norm": 1.6736913919448853,
      "learning_rate": 6.212653778558876e-05,
      "loss": 0.8889590501785278,
      "step": 708
    },
    {
      "epoch": 0.29957805907172996,
      "grad_norm": 1.6841140985488892,
      "learning_rate": 6.230228471001758e-05,
      "loss": 0.7822914123535156,
      "step": 710
    },
    {
      "epoch": 0.30042194092827,
      "grad_norm": 1.6644599437713623,
      "learning_rate": 6.24780316344464e-05,
      "loss": 0.8747053742408752,
      "step": 712
    },
    {
      "epoch": 0.3012658227848101,
      "grad_norm": 1.8187819719314575,
      "learning_rate": 6.265377855887522e-05,
      "loss": 0.8976446390151978,
      "step": 714
    },
    {
      "epoch": 0.30210970464135023,
      "grad_norm": 1.7845178842544556,
      "learning_rate": 6.282952548330404e-05,
      "loss": 0.9401160478591919,
      "step": 716
    },
    {
      "epoch": 0.3029535864978903,
      "grad_norm": 1.559773564338684,
      "learning_rate": 6.300527240773286e-05,
      "loss": 0.8754280209541321,
      "step": 718
    },
    {
      "epoch": 0.3037974683544304,
      "grad_norm": 1.5919631719589233,
      "learning_rate": 6.318101933216169e-05,
      "loss": 0.8278581500053406,
      "step": 720
    },
    {
      "epoch": 0.30464135021097044,
      "grad_norm": 1.8551076650619507,
      "learning_rate": 6.335676625659052e-05,
      "loss": 0.8868640065193176,
      "step": 722
    },
    {
      "epoch": 0.30548523206751055,
      "grad_norm": 1.6907769441604614,
      "learning_rate": 6.353251318101934e-05,
      "loss": 0.8631605505943298,
      "step": 724
    },
    {
      "epoch": 0.30632911392405066,
      "grad_norm": 1.820867657661438,
      "learning_rate": 6.370826010544816e-05,
      "loss": 0.9142873883247375,
      "step": 726
    },
    {
      "epoch": 0.3071729957805907,
      "grad_norm": 1.685154676437378,
      "learning_rate": 6.388400702987698e-05,
      "loss": 0.8258634805679321,
      "step": 728
    },
    {
      "epoch": 0.3080168776371308,
      "grad_norm": 1.9294627904891968,
      "learning_rate": 6.40597539543058e-05,
      "loss": 0.9545516967773438,
      "step": 730
    },
    {
      "epoch": 0.30886075949367087,
      "grad_norm": 1.6075409650802612,
      "learning_rate": 6.423550087873462e-05,
      "loss": 0.8370757699012756,
      "step": 732
    },
    {
      "epoch": 0.309704641350211,
      "grad_norm": 1.635750651359558,
      "learning_rate": 6.441124780316345e-05,
      "loss": 0.8356084823608398,
      "step": 734
    },
    {
      "epoch": 0.3105485232067511,
      "grad_norm": 1.6376131772994995,
      "learning_rate": 6.458699472759227e-05,
      "loss": 0.7579531669616699,
      "step": 736
    },
    {
      "epoch": 0.31139240506329113,
      "grad_norm": 1.7135766744613647,
      "learning_rate": 6.47627416520211e-05,
      "loss": 0.8436318039894104,
      "step": 738
    },
    {
      "epoch": 0.31223628691983124,
      "grad_norm": 1.7095093727111816,
      "learning_rate": 6.493848857644992e-05,
      "loss": 0.7998805046081543,
      "step": 740
    },
    {
      "epoch": 0.3130801687763713,
      "grad_norm": 1.782615303993225,
      "learning_rate": 6.511423550087874e-05,
      "loss": 0.915776789188385,
      "step": 742
    },
    {
      "epoch": 0.3139240506329114,
      "grad_norm": 1.8461172580718994,
      "learning_rate": 6.528998242530756e-05,
      "loss": 0.8300962448120117,
      "step": 744
    },
    {
      "epoch": 0.31476793248945145,
      "grad_norm": 1.5659871101379395,
      "learning_rate": 6.546572934973638e-05,
      "loss": 0.8239848017692566,
      "step": 746
    },
    {
      "epoch": 0.31561181434599156,
      "grad_norm": 1.9997349977493286,
      "learning_rate": 6.56414762741652e-05,
      "loss": 0.8236988186836243,
      "step": 748
    },
    {
      "epoch": 0.31645569620253167,
      "grad_norm": 1.9811526536941528,
      "learning_rate": 6.581722319859403e-05,
      "loss": 0.8516603112220764,
      "step": 750
    },
    {
      "epoch": 0.3172995780590717,
      "grad_norm": 1.9877923727035522,
      "learning_rate": 6.599297012302285e-05,
      "loss": 0.9037567973136902,
      "step": 752
    },
    {
      "epoch": 0.3181434599156118,
      "grad_norm": 1.6729352474212646,
      "learning_rate": 6.616871704745168e-05,
      "loss": 0.8350864052772522,
      "step": 754
    },
    {
      "epoch": 0.3189873417721519,
      "grad_norm": 1.9055802822113037,
      "learning_rate": 6.63444639718805e-05,
      "loss": 0.8246616125106812,
      "step": 756
    },
    {
      "epoch": 0.319831223628692,
      "grad_norm": 1.597999930381775,
      "learning_rate": 6.652021089630932e-05,
      "loss": 0.8014416098594666,
      "step": 758
    },
    {
      "epoch": 0.3206751054852321,
      "grad_norm": 1.7432531118392944,
      "learning_rate": 6.669595782073814e-05,
      "loss": 0.9199523329734802,
      "step": 760
    },
    {
      "epoch": 0.32151898734177214,
      "grad_norm": 1.820164442062378,
      "learning_rate": 6.687170474516696e-05,
      "loss": 0.7764829397201538,
      "step": 762
    },
    {
      "epoch": 0.32236286919831225,
      "grad_norm": 1.6408652067184448,
      "learning_rate": 6.704745166959578e-05,
      "loss": 0.8072620630264282,
      "step": 764
    },
    {
      "epoch": 0.3232067510548523,
      "grad_norm": 1.8894155025482178,
      "learning_rate": 6.722319859402461e-05,
      "loss": 0.9006885886192322,
      "step": 766
    },
    {
      "epoch": 0.3240506329113924,
      "grad_norm": 1.6903613805770874,
      "learning_rate": 6.739894551845343e-05,
      "loss": 0.7772189378738403,
      "step": 768
    },
    {
      "epoch": 0.32489451476793246,
      "grad_norm": 1.7540696859359741,
      "learning_rate": 6.757469244288225e-05,
      "loss": 0.8825590014457703,
      "step": 770
    },
    {
      "epoch": 0.32573839662447257,
      "grad_norm": 1.603008508682251,
      "learning_rate": 6.775043936731108e-05,
      "loss": 0.8376453518867493,
      "step": 772
    },
    {
      "epoch": 0.3265822784810127,
      "grad_norm": 1.5381462574005127,
      "learning_rate": 6.79261862917399e-05,
      "loss": 0.92608243227005,
      "step": 774
    },
    {
      "epoch": 0.32742616033755273,
      "grad_norm": 1.4815537929534912,
      "learning_rate": 6.810193321616872e-05,
      "loss": 0.6842183470726013,
      "step": 776
    },
    {
      "epoch": 0.32827004219409284,
      "grad_norm": 1.8543411493301392,
      "learning_rate": 6.827768014059754e-05,
      "loss": 0.8868235349655151,
      "step": 778
    },
    {
      "epoch": 0.3291139240506329,
      "grad_norm": 1.8895748853683472,
      "learning_rate": 6.845342706502637e-05,
      "loss": 0.8148112297058105,
      "step": 780
    },
    {
      "epoch": 0.329957805907173,
      "grad_norm": 1.8150591850280762,
      "learning_rate": 6.862917398945519e-05,
      "loss": 0.8760337829589844,
      "step": 782
    },
    {
      "epoch": 0.3308016877637131,
      "grad_norm": 1.6661378145217896,
      "learning_rate": 6.880492091388401e-05,
      "loss": 0.8266322612762451,
      "step": 784
    },
    {
      "epoch": 0.33164556962025316,
      "grad_norm": 2.2849128246307373,
      "learning_rate": 6.898066783831283e-05,
      "loss": 0.8599053025245667,
      "step": 786
    },
    {
      "epoch": 0.33248945147679326,
      "grad_norm": 1.7233171463012695,
      "learning_rate": 6.915641476274165e-05,
      "loss": 0.8312317132949829,
      "step": 788
    },
    {
      "epoch": 0.3333333333333333,
      "grad_norm": 1.7637618780136108,
      "learning_rate": 6.933216168717048e-05,
      "loss": 0.8379700779914856,
      "step": 790
    },
    {
      "epoch": 0.3341772151898734,
      "grad_norm": 1.7780474424362183,
      "learning_rate": 6.95079086115993e-05,
      "loss": 0.8994934558868408,
      "step": 792
    },
    {
      "epoch": 0.33502109704641353,
      "grad_norm": 1.5798883438110352,
      "learning_rate": 6.968365553602812e-05,
      "loss": 0.8021857738494873,
      "step": 794
    },
    {
      "epoch": 0.3358649789029536,
      "grad_norm": 1.7316070795059204,
      "learning_rate": 6.985940246045695e-05,
      "loss": 0.8814419507980347,
      "step": 796
    },
    {
      "epoch": 0.3367088607594937,
      "grad_norm": 1.711315631866455,
      "learning_rate": 7.003514938488577e-05,
      "loss": 0.8545029163360596,
      "step": 798
    },
    {
      "epoch": 0.33755274261603374,
      "grad_norm": 1.5023137331008911,
      "learning_rate": 7.021089630931459e-05,
      "loss": 0.8006189465522766,
      "step": 800
    },
    {
      "epoch": 0.33755274261603374,
      "eval_loss": 0.8635594248771667,
      "eval_runtime": 865.9348,
      "eval_samples_per_second": 2.433,
      "eval_steps_per_second": 2.433,
      "step": 800
    },
    {
      "epoch": 0.33839662447257385,
      "grad_norm": 1.8377124071121216,
      "learning_rate": 7.038664323374341e-05,
      "loss": 0.7625874280929565,
      "step": 802
    },
    {
      "epoch": 0.3392405063291139,
      "grad_norm": 1.5361332893371582,
      "learning_rate": 7.056239015817223e-05,
      "loss": 0.8490484356880188,
      "step": 804
    },
    {
      "epoch": 0.340084388185654,
      "grad_norm": 1.8727388381958008,
      "learning_rate": 7.073813708260105e-05,
      "loss": 0.8915753364562988,
      "step": 806
    },
    {
      "epoch": 0.3409282700421941,
      "grad_norm": 1.567700743675232,
      "learning_rate": 7.091388400702988e-05,
      "loss": 0.8902620077133179,
      "step": 808
    },
    {
      "epoch": 0.34177215189873417,
      "grad_norm": 1.5302914381027222,
      "learning_rate": 7.10896309314587e-05,
      "loss": 0.7897103428840637,
      "step": 810
    },
    {
      "epoch": 0.3426160337552743,
      "grad_norm": 1.8819153308868408,
      "learning_rate": 7.126537785588753e-05,
      "loss": 0.8648831248283386,
      "step": 812
    },
    {
      "epoch": 0.3434599156118143,
      "grad_norm": 1.5671379566192627,
      "learning_rate": 7.144112478031635e-05,
      "loss": 0.8449499607086182,
      "step": 814
    },
    {
      "epoch": 0.34430379746835443,
      "grad_norm": 1.6570971012115479,
      "learning_rate": 7.161687170474517e-05,
      "loss": 0.848559558391571,
      "step": 816
    },
    {
      "epoch": 0.34514767932489454,
      "grad_norm": 1.9108437299728394,
      "learning_rate": 7.179261862917399e-05,
      "loss": 0.8847543597221375,
      "step": 818
    },
    {
      "epoch": 0.3459915611814346,
      "grad_norm": 1.4909496307373047,
      "learning_rate": 7.196836555360281e-05,
      "loss": 0.7642563581466675,
      "step": 820
    },
    {
      "epoch": 0.3468354430379747,
      "grad_norm": 1.768518328666687,
      "learning_rate": 7.214411247803163e-05,
      "loss": 0.8714305758476257,
      "step": 822
    },
    {
      "epoch": 0.34767932489451475,
      "grad_norm": 1.715343952178955,
      "learning_rate": 7.231985940246046e-05,
      "loss": 0.7712987661361694,
      "step": 824
    },
    {
      "epoch": 0.34852320675105486,
      "grad_norm": 1.6687803268432617,
      "learning_rate": 7.24956063268893e-05,
      "loss": 0.8122798204421997,
      "step": 826
    },
    {
      "epoch": 0.3493670886075949,
      "grad_norm": 1.5160514116287231,
      "learning_rate": 7.267135325131811e-05,
      "loss": 0.793245792388916,
      "step": 828
    },
    {
      "epoch": 0.350210970464135,
      "grad_norm": 1.6449401378631592,
      "learning_rate": 7.284710017574693e-05,
      "loss": 0.8747497200965881,
      "step": 830
    },
    {
      "epoch": 0.3510548523206751,
      "grad_norm": 1.3907722234725952,
      "learning_rate": 7.302284710017575e-05,
      "loss": 0.6743978261947632,
      "step": 832
    },
    {
      "epoch": 0.3518987341772152,
      "grad_norm": 1.633555293083191,
      "learning_rate": 7.319859402460457e-05,
      "loss": 0.8524789214134216,
      "step": 834
    },
    {
      "epoch": 0.3527426160337553,
      "grad_norm": 1.5414257049560547,
      "learning_rate": 7.337434094903339e-05,
      "loss": 0.8045110702514648,
      "step": 836
    },
    {
      "epoch": 0.35358649789029534,
      "grad_norm": 1.8520616292953491,
      "learning_rate": 7.355008787346221e-05,
      "loss": 0.8319593071937561,
      "step": 838
    },
    {
      "epoch": 0.35443037974683544,
      "grad_norm": 1.6629763841629028,
      "learning_rate": 7.372583479789104e-05,
      "loss": 0.8188939094543457,
      "step": 840
    },
    {
      "epoch": 0.35527426160337555,
      "grad_norm": 1.804087519645691,
      "learning_rate": 7.390158172231987e-05,
      "loss": 0.8875360488891602,
      "step": 842
    },
    {
      "epoch": 0.3561181434599156,
      "grad_norm": 1.6031663417816162,
      "learning_rate": 7.407732864674869e-05,
      "loss": 0.8159612417221069,
      "step": 844
    },
    {
      "epoch": 0.3569620253164557,
      "grad_norm": 1.7413033246994019,
      "learning_rate": 7.425307557117751e-05,
      "loss": 0.8422684669494629,
      "step": 846
    },
    {
      "epoch": 0.35780590717299576,
      "grad_norm": 1.7699719667434692,
      "learning_rate": 7.442882249560633e-05,
      "loss": 0.9343502521514893,
      "step": 848
    },
    {
      "epoch": 0.35864978902953587,
      "grad_norm": 1.4613301753997803,
      "learning_rate": 7.460456942003515e-05,
      "loss": 0.8168979287147522,
      "step": 850
    },
    {
      "epoch": 0.3594936708860759,
      "grad_norm": 1.542431354522705,
      "learning_rate": 7.478031634446397e-05,
      "loss": 0.9014382362365723,
      "step": 852
    },
    {
      "epoch": 0.36033755274261603,
      "grad_norm": 1.6070159673690796,
      "learning_rate": 7.49560632688928e-05,
      "loss": 0.8162738084793091,
      "step": 854
    },
    {
      "epoch": 0.36118143459915614,
      "grad_norm": 1.7979451417922974,
      "learning_rate": 7.513181019332162e-05,
      "loss": 0.8354527950286865,
      "step": 856
    },
    {
      "epoch": 0.3620253164556962,
      "grad_norm": 2.327045202255249,
      "learning_rate": 7.530755711775044e-05,
      "loss": 0.8214042782783508,
      "step": 858
    },
    {
      "epoch": 0.3628691983122363,
      "grad_norm": 1.5085111856460571,
      "learning_rate": 7.548330404217927e-05,
      "loss": 0.7472147941589355,
      "step": 860
    },
    {
      "epoch": 0.36371308016877635,
      "grad_norm": 1.6006290912628174,
      "learning_rate": 7.565905096660809e-05,
      "loss": 0.7586950063705444,
      "step": 862
    },
    {
      "epoch": 0.36455696202531646,
      "grad_norm": 1.5170620679855347,
      "learning_rate": 7.583479789103691e-05,
      "loss": 0.8169914484024048,
      "step": 864
    },
    {
      "epoch": 0.36540084388185656,
      "grad_norm": 1.5848352909088135,
      "learning_rate": 7.601054481546573e-05,
      "loss": 0.8263922929763794,
      "step": 866
    },
    {
      "epoch": 0.3662447257383966,
      "grad_norm": 1.8502342700958252,
      "learning_rate": 7.618629173989455e-05,
      "loss": 0.8726240992546082,
      "step": 868
    },
    {
      "epoch": 0.3670886075949367,
      "grad_norm": 1.506847620010376,
      "learning_rate": 7.636203866432338e-05,
      "loss": 0.7220374941825867,
      "step": 870
    },
    {
      "epoch": 0.3679324894514768,
      "grad_norm": 1.5350452661514282,
      "learning_rate": 7.65377855887522e-05,
      "loss": 0.8028547167778015,
      "step": 872
    },
    {
      "epoch": 0.3687763713080169,
      "grad_norm": 1.5011043548583984,
      "learning_rate": 7.671353251318102e-05,
      "loss": 0.7659649848937988,
      "step": 874
    },
    {
      "epoch": 0.369620253164557,
      "grad_norm": 1.7019832134246826,
      "learning_rate": 7.688927943760984e-05,
      "loss": 0.8773653507232666,
      "step": 876
    },
    {
      "epoch": 0.37046413502109704,
      "grad_norm": 1.4918498992919922,
      "learning_rate": 7.706502636203867e-05,
      "loss": 0.7977569103240967,
      "step": 878
    },
    {
      "epoch": 0.37130801687763715,
      "grad_norm": 1.6422638893127441,
      "learning_rate": 7.724077328646749e-05,
      "loss": 0.7491976022720337,
      "step": 880
    },
    {
      "epoch": 0.3721518987341772,
      "grad_norm": 1.7590434551239014,
      "learning_rate": 7.741652021089631e-05,
      "loss": 0.8754181265830994,
      "step": 882
    },
    {
      "epoch": 0.3729957805907173,
      "grad_norm": 3.868894100189209,
      "learning_rate": 7.759226713532513e-05,
      "loss": 0.8482301235198975,
      "step": 884
    },
    {
      "epoch": 0.37383966244725736,
      "grad_norm": 2.111875534057617,
      "learning_rate": 7.776801405975396e-05,
      "loss": 0.8109031915664673,
      "step": 886
    },
    {
      "epoch": 0.37468354430379747,
      "grad_norm": 2.0838418006896973,
      "learning_rate": 7.794376098418278e-05,
      "loss": 0.8660775423049927,
      "step": 888
    },
    {
      "epoch": 0.3755274261603376,
      "grad_norm": 1.553022027015686,
      "learning_rate": 7.81195079086116e-05,
      "loss": 0.8418024778366089,
      "step": 890
    },
    {
      "epoch": 0.3763713080168776,
      "grad_norm": 1.334747314453125,
      "learning_rate": 7.829525483304042e-05,
      "loss": 0.7764869928359985,
      "step": 892
    },
    {
      "epoch": 0.37721518987341773,
      "grad_norm": 1.4692286252975464,
      "learning_rate": 7.847100175746925e-05,
      "loss": 0.7460401654243469,
      "step": 894
    },
    {
      "epoch": 0.3780590717299578,
      "grad_norm": 1.5374023914337158,
      "learning_rate": 7.864674868189807e-05,
      "loss": 0.7662873268127441,
      "step": 896
    },
    {
      "epoch": 0.3789029535864979,
      "grad_norm": 1.5662524700164795,
      "learning_rate": 7.882249560632689e-05,
      "loss": 0.8165306448936462,
      "step": 898
    },
    {
      "epoch": 0.379746835443038,
      "grad_norm": 4.498590469360352,
      "learning_rate": 7.899824253075572e-05,
      "loss": 0.7913232445716858,
      "step": 900
    },
    {
      "epoch": 0.379746835443038,
      "eval_loss": 0.8491304516792297,
      "eval_runtime": 852.6211,
      "eval_samples_per_second": 2.471,
      "eval_steps_per_second": 2.471,
      "step": 900
    },
    {
      "epoch": 0.38059071729957805,
      "grad_norm": 1.6320613622665405,
      "learning_rate": 7.917398945518454e-05,
      "loss": 0.8097161054611206,
      "step": 902
    },
    {
      "epoch": 0.38143459915611816,
      "grad_norm": 1.2562934160232544,
      "learning_rate": 7.934973637961336e-05,
      "loss": 0.786399781703949,
      "step": 904
    },
    {
      "epoch": 0.3822784810126582,
      "grad_norm": 1.6957594156265259,
      "learning_rate": 7.952548330404218e-05,
      "loss": 0.8385500311851501,
      "step": 906
    },
    {
      "epoch": 0.3831223628691983,
      "grad_norm": 1.6662386655807495,
      "learning_rate": 7.9701230228471e-05,
      "loss": 0.8157848715782166,
      "step": 908
    },
    {
      "epoch": 0.38396624472573837,
      "grad_norm": 1.6717777252197266,
      "learning_rate": 7.987697715289982e-05,
      "loss": 0.7937968373298645,
      "step": 910
    },
    {
      "epoch": 0.3848101265822785,
      "grad_norm": 1.399484395980835,
      "learning_rate": 8.005272407732865e-05,
      "loss": 0.7800109386444092,
      "step": 912
    },
    {
      "epoch": 0.3856540084388186,
      "grad_norm": 1.5671080350875854,
      "learning_rate": 8.022847100175747e-05,
      "loss": 0.8135939240455627,
      "step": 914
    },
    {
      "epoch": 0.38649789029535864,
      "grad_norm": 1.4427763223648071,
      "learning_rate": 8.04042179261863e-05,
      "loss": 0.7482035160064697,
      "step": 916
    },
    {
      "epoch": 0.38734177215189874,
      "grad_norm": 1.3314121961593628,
      "learning_rate": 8.057996485061512e-05,
      "loss": 0.7201873064041138,
      "step": 918
    },
    {
      "epoch": 0.3881856540084388,
      "grad_norm": 1.5695286989212036,
      "learning_rate": 8.075571177504394e-05,
      "loss": 0.7933040857315063,
      "step": 920
    },
    {
      "epoch": 0.3890295358649789,
      "grad_norm": 1.5091747045516968,
      "learning_rate": 8.093145869947276e-05,
      "loss": 0.8058338165283203,
      "step": 922
    },
    {
      "epoch": 0.389873417721519,
      "grad_norm": 1.6287630796432495,
      "learning_rate": 8.110720562390158e-05,
      "loss": 0.7617828249931335,
      "step": 924
    },
    {
      "epoch": 0.39071729957805906,
      "grad_norm": 1.6129482984542847,
      "learning_rate": 8.12829525483304e-05,
      "loss": 0.8710150122642517,
      "step": 926
    },
    {
      "epoch": 0.39156118143459917,
      "grad_norm": 1.6457173824310303,
      "learning_rate": 8.145869947275922e-05,
      "loss": 0.9122233390808105,
      "step": 928
    },
    {
      "epoch": 0.3924050632911392,
      "grad_norm": 1.6768827438354492,
      "learning_rate": 8.163444639718805e-05,
      "loss": 0.8339303731918335,
      "step": 930
    },
    {
      "epoch": 0.39324894514767933,
      "grad_norm": 1.5419740676879883,
      "learning_rate": 8.181019332161688e-05,
      "loss": 0.8220396041870117,
      "step": 932
    },
    {
      "epoch": 0.39409282700421944,
      "grad_norm": 1.4563747644424438,
      "learning_rate": 8.19859402460457e-05,
      "loss": 0.8531478047370911,
      "step": 934
    },
    {
      "epoch": 0.3949367088607595,
      "grad_norm": 1.6208328008651733,
      "learning_rate": 8.216168717047452e-05,
      "loss": 0.8330869078636169,
      "step": 936
    },
    {
      "epoch": 0.3957805907172996,
      "grad_norm": 1.6492482423782349,
      "learning_rate": 8.233743409490334e-05,
      "loss": 0.8011296987533569,
      "step": 938
    },
    {
      "epoch": 0.39662447257383965,
      "grad_norm": 2.1611905097961426,
      "learning_rate": 8.251318101933216e-05,
      "loss": 0.8111353516578674,
      "step": 940
    },
    {
      "epoch": 0.39746835443037976,
      "grad_norm": 1.7108231782913208,
      "learning_rate": 8.268892794376098e-05,
      "loss": 0.8282017111778259,
      "step": 942
    },
    {
      "epoch": 0.3983122362869198,
      "grad_norm": 1.543465495109558,
      "learning_rate": 8.286467486818981e-05,
      "loss": 0.7770059704780579,
      "step": 944
    },
    {
      "epoch": 0.3991561181434599,
      "grad_norm": 1.419969081878662,
      "learning_rate": 8.304042179261863e-05,
      "loss": 0.8646430373191833,
      "step": 946
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.5002100467681885,
      "learning_rate": 8.321616871704746e-05,
      "loss": 0.7949403524398804,
      "step": 948
    },
    {
      "epoch": 0.4008438818565401,
      "grad_norm": 1.38933265209198,
      "learning_rate": 8.339191564147628e-05,
      "loss": 0.8124079704284668,
      "step": 950
    },
    {
      "epoch": 0.4016877637130802,
      "grad_norm": 1.5948443412780762,
      "learning_rate": 8.35676625659051e-05,
      "loss": 0.8634148836135864,
      "step": 952
    },
    {
      "epoch": 0.40253164556962023,
      "grad_norm": 1.4437624216079712,
      "learning_rate": 8.374340949033392e-05,
      "loss": 0.7410681247711182,
      "step": 954
    },
    {
      "epoch": 0.40337552742616034,
      "grad_norm": 1.3457095623016357,
      "learning_rate": 8.391915641476274e-05,
      "loss": 0.7680280208587646,
      "step": 956
    },
    {
      "epoch": 0.40421940928270045,
      "grad_norm": 1.610288143157959,
      "learning_rate": 8.409490333919156e-05,
      "loss": 0.7921904921531677,
      "step": 958
    },
    {
      "epoch": 0.4050632911392405,
      "grad_norm": 1.5321530103683472,
      "learning_rate": 8.427065026362039e-05,
      "loss": 0.8320037126541138,
      "step": 960
    },
    {
      "epoch": 0.4059071729957806,
      "grad_norm": 1.699881672859192,
      "learning_rate": 8.444639718804921e-05,
      "loss": 0.8303092122077942,
      "step": 962
    },
    {
      "epoch": 0.40675105485232066,
      "grad_norm": 1.591515064239502,
      "learning_rate": 8.462214411247804e-05,
      "loss": 0.9029796719551086,
      "step": 964
    },
    {
      "epoch": 0.40759493670886077,
      "grad_norm": 1.5930429697036743,
      "learning_rate": 8.479789103690686e-05,
      "loss": 0.8165359497070312,
      "step": 966
    },
    {
      "epoch": 0.4084388185654008,
      "grad_norm": 1.509774923324585,
      "learning_rate": 8.497363796133568e-05,
      "loss": 0.8276026248931885,
      "step": 968
    },
    {
      "epoch": 0.4092827004219409,
      "grad_norm": 1.3617016077041626,
      "learning_rate": 8.51493848857645e-05,
      "loss": 0.8159419894218445,
      "step": 970
    },
    {
      "epoch": 0.41012658227848103,
      "grad_norm": 1.3580708503723145,
      "learning_rate": 8.532513181019332e-05,
      "loss": 0.7882336378097534,
      "step": 972
    },
    {
      "epoch": 0.4109704641350211,
      "grad_norm": 1.3337358236312866,
      "learning_rate": 8.550087873462214e-05,
      "loss": 0.7462319731712341,
      "step": 974
    },
    {
      "epoch": 0.4118143459915612,
      "grad_norm": 1.450363278388977,
      "learning_rate": 8.567662565905097e-05,
      "loss": 0.7500866651535034,
      "step": 976
    },
    {
      "epoch": 0.41265822784810124,
      "grad_norm": 1.5305321216583252,
      "learning_rate": 8.585237258347979e-05,
      "loss": 0.8432503342628479,
      "step": 978
    },
    {
      "epoch": 0.41350210970464135,
      "grad_norm": 1.2097326517105103,
      "learning_rate": 8.602811950790861e-05,
      "loss": 0.8330482840538025,
      "step": 980
    },
    {
      "epoch": 0.41434599156118146,
      "grad_norm": 1.3916101455688477,
      "learning_rate": 8.620386643233744e-05,
      "loss": 0.8137149810791016,
      "step": 982
    },
    {
      "epoch": 0.4151898734177215,
      "grad_norm": 1.6411453485488892,
      "learning_rate": 8.637961335676626e-05,
      "loss": 0.8273854851722717,
      "step": 984
    },
    {
      "epoch": 0.4160337552742616,
      "grad_norm": 1.6734566688537598,
      "learning_rate": 8.655536028119508e-05,
      "loss": 0.794026255607605,
      "step": 986
    },
    {
      "epoch": 0.41687763713080167,
      "grad_norm": 1.352325677871704,
      "learning_rate": 8.67311072056239e-05,
      "loss": 0.7721655368804932,
      "step": 988
    },
    {
      "epoch": 0.4177215189873418,
      "grad_norm": 1.5368729829788208,
      "learning_rate": 8.690685413005273e-05,
      "loss": 0.8123438954353333,
      "step": 990
    },
    {
      "epoch": 0.41856540084388183,
      "grad_norm": 1.4903568029403687,
      "learning_rate": 8.708260105448155e-05,
      "loss": 0.8370974659919739,
      "step": 992
    },
    {
      "epoch": 0.41940928270042194,
      "grad_norm": 1.3405622243881226,
      "learning_rate": 8.725834797891037e-05,
      "loss": 0.780426561832428,
      "step": 994
    },
    {
      "epoch": 0.42025316455696204,
      "grad_norm": 1.4761021137237549,
      "learning_rate": 8.743409490333919e-05,
      "loss": 0.8304934501647949,
      "step": 996
    },
    {
      "epoch": 0.4210970464135021,
      "grad_norm": 1.520033359527588,
      "learning_rate": 8.760984182776801e-05,
      "loss": 0.7960568070411682,
      "step": 998
    },
    {
      "epoch": 0.4219409282700422,
      "grad_norm": 1.6916255950927734,
      "learning_rate": 8.778558875219684e-05,
      "loss": 0.7884663939476013,
      "step": 1000
    },
    {
      "epoch": 0.4219409282700422,
      "eval_loss": 0.8388314247131348,
      "eval_runtime": 847.4828,
      "eval_samples_per_second": 2.486,
      "eval_steps_per_second": 2.486,
      "step": 1000
    },
    {
      "epoch": 0.42278481012658226,
      "grad_norm": 1.6796396970748901,
      "learning_rate": 8.796133567662566e-05,
      "loss": 0.7930826544761658,
      "step": 1002
    },
    {
      "epoch": 0.42362869198312236,
      "grad_norm": 1.4480048418045044,
      "learning_rate": 8.813708260105448e-05,
      "loss": 0.7138194441795349,
      "step": 1004
    },
    {
      "epoch": 0.42447257383966247,
      "grad_norm": 1.2499021291732788,
      "learning_rate": 8.831282952548331e-05,
      "loss": 0.7367453575134277,
      "step": 1006
    },
    {
      "epoch": 0.4253164556962025,
      "grad_norm": 1.6906769275665283,
      "learning_rate": 8.848857644991213e-05,
      "loss": 0.9051005244255066,
      "step": 1008
    },
    {
      "epoch": 0.42616033755274263,
      "grad_norm": 1.4196792840957642,
      "learning_rate": 8.866432337434095e-05,
      "loss": 0.7469457387924194,
      "step": 1010
    },
    {
      "epoch": 0.4270042194092827,
      "grad_norm": 1.5132776498794556,
      "learning_rate": 8.884007029876977e-05,
      "loss": 0.7443049550056458,
      "step": 1012
    },
    {
      "epoch": 0.4278481012658228,
      "grad_norm": 1.335705280303955,
      "learning_rate": 8.901581722319859e-05,
      "loss": 0.784084677696228,
      "step": 1014
    },
    {
      "epoch": 0.4286919831223629,
      "grad_norm": 1.6510252952575684,
      "learning_rate": 8.919156414762741e-05,
      "loss": 0.8603647947311401,
      "step": 1016
    },
    {
      "epoch": 0.42953586497890295,
      "grad_norm": 1.35535728931427,
      "learning_rate": 8.936731107205624e-05,
      "loss": 0.7921645641326904,
      "step": 1018
    },
    {
      "epoch": 0.43037974683544306,
      "grad_norm": 1.4952049255371094,
      "learning_rate": 8.954305799648506e-05,
      "loss": 0.799993634223938,
      "step": 1020
    },
    {
      "epoch": 0.4312236286919831,
      "grad_norm": 1.5026042461395264,
      "learning_rate": 8.97188049209139e-05,
      "loss": 0.7697094082832336,
      "step": 1022
    },
    {
      "epoch": 0.4320675105485232,
      "grad_norm": 1.5424275398254395,
      "learning_rate": 8.989455184534271e-05,
      "loss": 0.7988215684890747,
      "step": 1024
    },
    {
      "epoch": 0.43291139240506327,
      "grad_norm": 1.438716173171997,
      "learning_rate": 9.007029876977153e-05,
      "loss": 0.7841635942459106,
      "step": 1026
    },
    {
      "epoch": 0.4337552742616034,
      "grad_norm": 1.5040369033813477,
      "learning_rate": 9.024604569420035e-05,
      "loss": 0.7485025525093079,
      "step": 1028
    },
    {
      "epoch": 0.4345991561181435,
      "grad_norm": 1.4354394674301147,
      "learning_rate": 9.042179261862917e-05,
      "loss": 0.7735623121261597,
      "step": 1030
    },
    {
      "epoch": 0.43544303797468353,
      "grad_norm": 1.4841680526733398,
      "learning_rate": 9.059753954305799e-05,
      "loss": 0.8918828964233398,
      "step": 1032
    },
    {
      "epoch": 0.43628691983122364,
      "grad_norm": 1.428813099861145,
      "learning_rate": 9.077328646748682e-05,
      "loss": 0.835110068321228,
      "step": 1034
    },
    {
      "epoch": 0.4371308016877637,
      "grad_norm": 1.559020757675171,
      "learning_rate": 9.094903339191566e-05,
      "loss": 0.746295690536499,
      "step": 1036
    },
    {
      "epoch": 0.4379746835443038,
      "grad_norm": 1.6996115446090698,
      "learning_rate": 9.112478031634448e-05,
      "loss": 0.8089123368263245,
      "step": 1038
    },
    {
      "epoch": 0.4388185654008439,
      "grad_norm": 1.6615465879440308,
      "learning_rate": 9.13005272407733e-05,
      "loss": 0.8807073831558228,
      "step": 1040
    },
    {
      "epoch": 0.43966244725738396,
      "grad_norm": 1.239142894744873,
      "learning_rate": 9.147627416520211e-05,
      "loss": 0.7638427019119263,
      "step": 1042
    },
    {
      "epoch": 0.44050632911392407,
      "grad_norm": 1.1915178298950195,
      "learning_rate": 9.165202108963093e-05,
      "loss": 0.7817409634590149,
      "step": 1044
    },
    {
      "epoch": 0.4413502109704641,
      "grad_norm": 1.6276934146881104,
      "learning_rate": 9.182776801405975e-05,
      "loss": 0.8586427569389343,
      "step": 1046
    },
    {
      "epoch": 0.4421940928270042,
      "grad_norm": 1.480345606803894,
      "learning_rate": 9.200351493848857e-05,
      "loss": 0.7481811046600342,
      "step": 1048
    },
    {
      "epoch": 0.4430379746835443,
      "grad_norm": 1.308419108390808,
      "learning_rate": 9.21792618629174e-05,
      "loss": 0.8074686527252197,
      "step": 1050
    },
    {
      "epoch": 0.4438818565400844,
      "grad_norm": 1.6167182922363281,
      "learning_rate": 9.235500878734624e-05,
      "loss": 0.8455166816711426,
      "step": 1052
    },
    {
      "epoch": 0.4447257383966245,
      "grad_norm": 1.6058826446533203,
      "learning_rate": 9.253075571177506e-05,
      "loss": 0.7255295515060425,
      "step": 1054
    },
    {
      "epoch": 0.44556962025316454,
      "grad_norm": 1.6745728254318237,
      "learning_rate": 9.270650263620387e-05,
      "loss": 0.8329368233680725,
      "step": 1056
    },
    {
      "epoch": 0.44641350210970465,
      "grad_norm": 1.5657380819320679,
      "learning_rate": 9.28822495606327e-05,
      "loss": 0.8583613634109497,
      "step": 1058
    },
    {
      "epoch": 0.4472573839662447,
      "grad_norm": 1.5052601099014282,
      "learning_rate": 9.305799648506151e-05,
      "loss": 0.8546127080917358,
      "step": 1060
    },
    {
      "epoch": 0.4481012658227848,
      "grad_norm": 1.510636806488037,
      "learning_rate": 9.323374340949033e-05,
      "loss": 0.8416863679885864,
      "step": 1062
    },
    {
      "epoch": 0.4489451476793249,
      "grad_norm": 1.4446617364883423,
      "learning_rate": 9.340949033391916e-05,
      "loss": 0.830390453338623,
      "step": 1064
    },
    {
      "epoch": 0.44978902953586497,
      "grad_norm": 1.6032582521438599,
      "learning_rate": 9.358523725834798e-05,
      "loss": 0.8000447154045105,
      "step": 1066
    },
    {
      "epoch": 0.4506329113924051,
      "grad_norm": 1.5295692682266235,
      "learning_rate": 9.37609841827768e-05,
      "loss": 0.8310818672180176,
      "step": 1068
    },
    {
      "epoch": 0.45147679324894513,
      "grad_norm": 1.3161942958831787,
      "learning_rate": 9.393673110720564e-05,
      "loss": 0.8377846479415894,
      "step": 1070
    },
    {
      "epoch": 0.45232067510548524,
      "grad_norm": 1.4101601839065552,
      "learning_rate": 9.411247803163445e-05,
      "loss": 0.7852389216423035,
      "step": 1072
    },
    {
      "epoch": 0.4531645569620253,
      "grad_norm": 1.4352775812149048,
      "learning_rate": 9.428822495606327e-05,
      "loss": 0.8763723969459534,
      "step": 1074
    },
    {
      "epoch": 0.4540084388185654,
      "grad_norm": 1.4584673643112183,
      "learning_rate": 9.44639718804921e-05,
      "loss": 0.8177199363708496,
      "step": 1076
    },
    {
      "epoch": 0.4548523206751055,
      "grad_norm": 1.6470575332641602,
      "learning_rate": 9.463971880492091e-05,
      "loss": 0.8333053588867188,
      "step": 1078
    },
    {
      "epoch": 0.45569620253164556,
      "grad_norm": 1.4429512023925781,
      "learning_rate": 9.481546572934975e-05,
      "loss": 0.8546649217605591,
      "step": 1080
    },
    {
      "epoch": 0.45654008438818566,
      "grad_norm": 1.4885371923446655,
      "learning_rate": 9.499121265377856e-05,
      "loss": 0.838036298751831,
      "step": 1082
    },
    {
      "epoch": 0.4573839662447257,
      "grad_norm": 1.4601678848266602,
      "learning_rate": 9.516695957820738e-05,
      "loss": 0.7295010089874268,
      "step": 1084
    },
    {
      "epoch": 0.4582278481012658,
      "grad_norm": 1.2399365901947021,
      "learning_rate": 9.53427065026362e-05,
      "loss": 0.6990782618522644,
      "step": 1086
    },
    {
      "epoch": 0.45907172995780593,
      "grad_norm": 1.2936921119689941,
      "learning_rate": 9.551845342706504e-05,
      "loss": 0.7790928483009338,
      "step": 1088
    },
    {
      "epoch": 0.459915611814346,
      "grad_norm": 1.3408331871032715,
      "learning_rate": 9.569420035149385e-05,
      "loss": 0.8061056733131409,
      "step": 1090
    },
    {
      "epoch": 0.4607594936708861,
      "grad_norm": 1.5525178909301758,
      "learning_rate": 9.586994727592267e-05,
      "loss": 0.856796383857727,
      "step": 1092
    },
    {
      "epoch": 0.46160337552742614,
      "grad_norm": 1.2944618463516235,
      "learning_rate": 9.604569420035149e-05,
      "loss": 0.7626663446426392,
      "step": 1094
    },
    {
      "epoch": 0.46244725738396625,
      "grad_norm": 1.412204623222351,
      "learning_rate": 9.622144112478033e-05,
      "loss": 0.7524681091308594,
      "step": 1096
    },
    {
      "epoch": 0.46329113924050636,
      "grad_norm": 1.4851596355438232,
      "learning_rate": 9.639718804920914e-05,
      "loss": 0.8430375456809998,
      "step": 1098
    },
    {
      "epoch": 0.4641350210970464,
      "grad_norm": 1.831943154335022,
      "learning_rate": 9.657293497363796e-05,
      "loss": 0.8374918103218079,
      "step": 1100
    },
    {
      "epoch": 0.4641350210970464,
      "eval_loss": 0.8283821940422058,
      "eval_runtime": 861.0464,
      "eval_samples_per_second": 2.447,
      "eval_steps_per_second": 2.447,
      "step": 1100
    },
    {
      "epoch": 0.4649789029535865,
      "grad_norm": 1.4989945888519287,
      "learning_rate": 9.674868189806678e-05,
      "loss": 0.8063139915466309,
      "step": 1102
    },
    {
      "epoch": 0.46582278481012657,
      "grad_norm": 1.3772722482681274,
      "learning_rate": 9.692442882249562e-05,
      "loss": 0.8109207153320312,
      "step": 1104
    },
    {
      "epoch": 0.4666666666666667,
      "grad_norm": 1.4963124990463257,
      "learning_rate": 9.710017574692443e-05,
      "loss": 0.8667853474617004,
      "step": 1106
    },
    {
      "epoch": 0.4675105485232067,
      "grad_norm": 1.4250836372375488,
      "learning_rate": 9.727592267135325e-05,
      "loss": 0.8020523190498352,
      "step": 1108
    },
    {
      "epoch": 0.46835443037974683,
      "grad_norm": 1.475599765777588,
      "learning_rate": 9.745166959578209e-05,
      "loss": 0.8271048069000244,
      "step": 1110
    },
    {
      "epoch": 0.46919831223628694,
      "grad_norm": 1.3727436065673828,
      "learning_rate": 9.76274165202109e-05,
      "loss": 0.7615619897842407,
      "step": 1112
    },
    {
      "epoch": 0.470042194092827,
      "grad_norm": 1.2233914136886597,
      "learning_rate": 9.780316344463972e-05,
      "loss": 0.7843242883682251,
      "step": 1114
    },
    {
      "epoch": 0.4708860759493671,
      "grad_norm": 1.5734832286834717,
      "learning_rate": 9.797891036906854e-05,
      "loss": 0.834839940071106,
      "step": 1116
    },
    {
      "epoch": 0.47172995780590715,
      "grad_norm": 1.3778531551361084,
      "learning_rate": 9.815465729349736e-05,
      "loss": 0.7584373950958252,
      "step": 1118
    },
    {
      "epoch": 0.47257383966244726,
      "grad_norm": 1.5535035133361816,
      "learning_rate": 9.833040421792618e-05,
      "loss": 0.8204697370529175,
      "step": 1120
    },
    {
      "epoch": 0.47341772151898737,
      "grad_norm": 1.4743636846542358,
      "learning_rate": 9.850615114235501e-05,
      "loss": 0.9012852311134338,
      "step": 1122
    },
    {
      "epoch": 0.4742616033755274,
      "grad_norm": 1.4134864807128906,
      "learning_rate": 9.868189806678383e-05,
      "loss": 0.8392805457115173,
      "step": 1124
    },
    {
      "epoch": 0.4751054852320675,
      "grad_norm": 1.3308019638061523,
      "learning_rate": 9.885764499121267e-05,
      "loss": 0.7135441303253174,
      "step": 1126
    },
    {
      "epoch": 0.4759493670886076,
      "grad_norm": 1.5354844331741333,
      "learning_rate": 9.903339191564149e-05,
      "loss": 0.8464727401733398,
      "step": 1128
    },
    {
      "epoch": 0.4767932489451477,
      "grad_norm": 1.2730523347854614,
      "learning_rate": 9.92091388400703e-05,
      "loss": 0.7691597938537598,
      "step": 1130
    },
    {
      "epoch": 0.47763713080168774,
      "grad_norm": 1.5459758043289185,
      "learning_rate": 9.938488576449912e-05,
      "loss": 0.8068788647651672,
      "step": 1132
    },
    {
      "epoch": 0.47848101265822784,
      "grad_norm": 1.345678687095642,
      "learning_rate": 9.956063268892794e-05,
      "loss": 0.8091006278991699,
      "step": 1134
    },
    {
      "epoch": 0.47932489451476795,
      "grad_norm": 1.317076563835144,
      "learning_rate": 9.973637961335676e-05,
      "loss": 0.735533595085144,
      "step": 1136
    },
    {
      "epoch": 0.480168776371308,
      "grad_norm": 1.5011168718338013,
      "learning_rate": 9.99121265377856e-05,
      "loss": 0.7935182452201843,
      "step": 1138
    },
    {
      "epoch": 0.4810126582278481,
      "grad_norm": 1.673899531364441,
      "learning_rate": 9.999999855824502e-05,
      "loss": 0.8203520774841309,
      "step": 1140
    },
    {
      "epoch": 0.48185654008438816,
      "grad_norm": 1.344337821006775,
      "learning_rate": 9.999998702420562e-05,
      "loss": 0.7233241200447083,
      "step": 1142
    },
    {
      "epoch": 0.48270042194092827,
      "grad_norm": 1.5819076299667358,
      "learning_rate": 9.999996395612948e-05,
      "loss": 0.8795552849769592,
      "step": 1144
    },
    {
      "epoch": 0.4835443037974684,
      "grad_norm": 1.7427241802215576,
      "learning_rate": 9.999992935402192e-05,
      "loss": 0.8482733964920044,
      "step": 1146
    },
    {
      "epoch": 0.48438818565400843,
      "grad_norm": 1.2877503633499146,
      "learning_rate": 9.999988321789093e-05,
      "loss": 0.7905706167221069,
      "step": 1148
    },
    {
      "epoch": 0.48523206751054854,
      "grad_norm": 1.4887222051620483,
      "learning_rate": 9.999982554774715e-05,
      "loss": 0.8609708547592163,
      "step": 1150
    },
    {
      "epoch": 0.4860759493670886,
      "grad_norm": 1.3625136613845825,
      "learning_rate": 9.999975634360388e-05,
      "loss": 0.7890065908432007,
      "step": 1152
    },
    {
      "epoch": 0.4869198312236287,
      "grad_norm": 1.3631492853164673,
      "learning_rate": 9.999967560547708e-05,
      "loss": 0.7908958196640015,
      "step": 1154
    },
    {
      "epoch": 0.4877637130801688,
      "grad_norm": 1.5244156122207642,
      "learning_rate": 9.99995833333854e-05,
      "loss": 0.8509655594825745,
      "step": 1156
    },
    {
      "epoch": 0.48860759493670886,
      "grad_norm": 1.2513200044631958,
      "learning_rate": 9.999947952735007e-05,
      "loss": 0.7329106330871582,
      "step": 1158
    },
    {
      "epoch": 0.48945147679324896,
      "grad_norm": 1.1539413928985596,
      "learning_rate": 9.99993641873951e-05,
      "loss": 0.7237489223480225,
      "step": 1160
    },
    {
      "epoch": 0.490295358649789,
      "grad_norm": 1.3859314918518066,
      "learning_rate": 9.999923731354706e-05,
      "loss": 0.8650591373443604,
      "step": 1162
    },
    {
      "epoch": 0.4911392405063291,
      "grad_norm": 1.2910805940628052,
      "learning_rate": 9.999909890583521e-05,
      "loss": 0.7516807913780212,
      "step": 1164
    },
    {
      "epoch": 0.4919831223628692,
      "grad_norm": 1.6100077629089355,
      "learning_rate": 9.999894896429152e-05,
      "loss": 0.7082475423812866,
      "step": 1166
    },
    {
      "epoch": 0.4928270042194093,
      "grad_norm": 1.2313556671142578,
      "learning_rate": 9.999878748895053e-05,
      "loss": 0.8403750658035278,
      "step": 1168
    },
    {
      "epoch": 0.4936708860759494,
      "grad_norm": 1.3402830362319946,
      "learning_rate": 9.999861447984952e-05,
      "loss": 0.8083041906356812,
      "step": 1170
    },
    {
      "epoch": 0.49451476793248944,
      "grad_norm": 1.516775131225586,
      "learning_rate": 9.999842993702839e-05,
      "loss": 0.8339354991912842,
      "step": 1172
    },
    {
      "epoch": 0.49535864978902955,
      "grad_norm": 1.2698423862457275,
      "learning_rate": 9.999823386052971e-05,
      "loss": 0.7708724141120911,
      "step": 1174
    },
    {
      "epoch": 0.4962025316455696,
      "grad_norm": 1.339390516281128,
      "learning_rate": 9.999802625039872e-05,
      "loss": 0.7589715719223022,
      "step": 1176
    },
    {
      "epoch": 0.4970464135021097,
      "grad_norm": 1.4618452787399292,
      "learning_rate": 9.99978071066833e-05,
      "loss": 0.8523206114768982,
      "step": 1178
    },
    {
      "epoch": 0.4978902953586498,
      "grad_norm": 1.4812564849853516,
      "learning_rate": 9.9997576429434e-05,
      "loss": 0.8143196105957031,
      "step": 1180
    },
    {
      "epoch": 0.49873417721518987,
      "grad_norm": 1.5720716714859009,
      "learning_rate": 9.999733421870405e-05,
      "loss": 0.800125002861023,
      "step": 1182
    },
    {
      "epoch": 0.49957805907173,
      "grad_norm": 1.4421230554580688,
      "learning_rate": 9.99970804745493e-05,
      "loss": 0.7618259191513062,
      "step": 1184
    },
    {
      "epoch": 0.5004219409282701,
      "grad_norm": 1.5794934034347534,
      "learning_rate": 9.99968151970283e-05,
      "loss": 0.7162163853645325,
      "step": 1186
    },
    {
      "epoch": 0.5012658227848101,
      "grad_norm": 1.8590432405471802,
      "learning_rate": 9.999653838620225e-05,
      "loss": 0.8089820146560669,
      "step": 1188
    },
    {
      "epoch": 0.5021097046413502,
      "grad_norm": 1.5194507837295532,
      "learning_rate": 9.999625004213498e-05,
      "loss": 0.8011203408241272,
      "step": 1190
    },
    {
      "epoch": 0.5029535864978903,
      "grad_norm": 1.6986470222473145,
      "learning_rate": 9.999595016489303e-05,
      "loss": 0.761158287525177,
      "step": 1192
    },
    {
      "epoch": 0.5037974683544304,
      "grad_norm": 1.4413946866989136,
      "learning_rate": 9.999563875454559e-05,
      "loss": 0.7898027300834656,
      "step": 1194
    },
    {
      "epoch": 0.5046413502109705,
      "grad_norm": 1.4509994983673096,
      "learning_rate": 9.999531581116443e-05,
      "loss": 0.8018442392349243,
      "step": 1196
    },
    {
      "epoch": 0.5054852320675105,
      "grad_norm": 1.400659441947937,
      "learning_rate": 9.999498133482412e-05,
      "loss": 0.7804076075553894,
      "step": 1198
    },
    {
      "epoch": 0.5063291139240507,
      "grad_norm": 1.486840009689331,
      "learning_rate": 9.999463532560178e-05,
      "loss": 0.82496178150177,
      "step": 1200
    },
    {
      "epoch": 0.5063291139240507,
      "eval_loss": 0.8186545968055725,
      "eval_runtime": 862.1638,
      "eval_samples_per_second": 2.444,
      "eval_steps_per_second": 2.444,
      "step": 1200
    },
    {
      "epoch": 0.5071729957805907,
      "grad_norm": 1.2770357131958008,
      "learning_rate": 9.999427778357723e-05,
      "loss": 0.8037722706794739,
      "step": 1202
    },
    {
      "epoch": 0.5080168776371308,
      "grad_norm": 1.4540977478027344,
      "learning_rate": 9.999390870883297e-05,
      "loss": 0.7329373359680176,
      "step": 1204
    },
    {
      "epoch": 0.5088607594936709,
      "grad_norm": 1.4469913244247437,
      "learning_rate": 9.999352810145412e-05,
      "loss": 0.8224589824676514,
      "step": 1206
    },
    {
      "epoch": 0.509704641350211,
      "grad_norm": 1.46500563621521,
      "learning_rate": 9.999313596152847e-05,
      "loss": 0.8106292486190796,
      "step": 1208
    },
    {
      "epoch": 0.510548523206751,
      "grad_norm": 1.3526637554168701,
      "learning_rate": 9.999273228914649e-05,
      "loss": 0.747698187828064,
      "step": 1210
    },
    {
      "epoch": 0.5113924050632911,
      "grad_norm": 1.28840172290802,
      "learning_rate": 9.999231708440131e-05,
      "loss": 0.7612425684928894,
      "step": 1212
    },
    {
      "epoch": 0.5122362869198313,
      "grad_norm": 1.0283230543136597,
      "learning_rate": 9.99918903473887e-05,
      "loss": 0.6839463710784912,
      "step": 1214
    },
    {
      "epoch": 0.5130801687763713,
      "grad_norm": 1.5231431722640991,
      "learning_rate": 9.999145207820708e-05,
      "loss": 0.8539203405380249,
      "step": 1216
    },
    {
      "epoch": 0.5139240506329114,
      "grad_norm": 1.3289231061935425,
      "learning_rate": 9.999100227695758e-05,
      "loss": 0.7960102558135986,
      "step": 1218
    },
    {
      "epoch": 0.5147679324894515,
      "grad_norm": 1.3770930767059326,
      "learning_rate": 9.999054094374396e-05,
      "loss": 0.7639255523681641,
      "step": 1220
    },
    {
      "epoch": 0.5156118143459916,
      "grad_norm": 1.3028030395507812,
      "learning_rate": 9.999006807867262e-05,
      "loss": 0.7743061780929565,
      "step": 1222
    },
    {
      "epoch": 0.5164556962025316,
      "grad_norm": 1.1827034950256348,
      "learning_rate": 9.998958368185265e-05,
      "loss": 0.7922407984733582,
      "step": 1224
    },
    {
      "epoch": 0.5172995780590718,
      "grad_norm": 1.2973705530166626,
      "learning_rate": 9.99890877533958e-05,
      "loss": 0.7671286463737488,
      "step": 1226
    },
    {
      "epoch": 0.5181434599156118,
      "grad_norm": 1.5820153951644897,
      "learning_rate": 9.998858029341646e-05,
      "loss": 0.7546951174736023,
      "step": 1228
    },
    {
      "epoch": 0.5189873417721519,
      "grad_norm": 1.6140317916870117,
      "learning_rate": 9.99880613020317e-05,
      "loss": 0.8734183311462402,
      "step": 1230
    },
    {
      "epoch": 0.5198312236286919,
      "grad_norm": 1.1190184354782104,
      "learning_rate": 9.998753077936122e-05,
      "loss": 0.8410643339157104,
      "step": 1232
    },
    {
      "epoch": 0.5206751054852321,
      "grad_norm": 1.3876196146011353,
      "learning_rate": 9.998698872552744e-05,
      "loss": 0.7769841551780701,
      "step": 1234
    },
    {
      "epoch": 0.5215189873417722,
      "grad_norm": 1.699522852897644,
      "learning_rate": 9.998643514065535e-05,
      "loss": 0.8846109509468079,
      "step": 1236
    },
    {
      "epoch": 0.5223628691983122,
      "grad_norm": 1.3805134296417236,
      "learning_rate": 9.998587002487271e-05,
      "loss": 0.7664945125579834,
      "step": 1238
    },
    {
      "epoch": 0.5232067510548524,
      "grad_norm": 1.3679476976394653,
      "learning_rate": 9.998529337830984e-05,
      "loss": 0.7243514060974121,
      "step": 1240
    },
    {
      "epoch": 0.5240506329113924,
      "grad_norm": 1.399200677871704,
      "learning_rate": 9.998470520109977e-05,
      "loss": 0.8061941862106323,
      "step": 1242
    },
    {
      "epoch": 0.5248945147679325,
      "grad_norm": 1.3441044092178345,
      "learning_rate": 9.99841054933782e-05,
      "loss": 0.7741840481758118,
      "step": 1244
    },
    {
      "epoch": 0.5257383966244725,
      "grad_norm": 1.3375325202941895,
      "learning_rate": 9.998349425528344e-05,
      "loss": 0.7619491815567017,
      "step": 1246
    },
    {
      "epoch": 0.5265822784810127,
      "grad_norm": 1.5517847537994385,
      "learning_rate": 9.998287148695651e-05,
      "loss": 0.8315094113349915,
      "step": 1248
    },
    {
      "epoch": 0.5274261603375527,
      "grad_norm": 1.244997501373291,
      "learning_rate": 9.998223718854107e-05,
      "loss": 0.7536082863807678,
      "step": 1250
    },
    {
      "epoch": 0.5282700421940928,
      "grad_norm": 1.3190033435821533,
      "learning_rate": 9.998159136018344e-05,
      "loss": 0.826419472694397,
      "step": 1252
    },
    {
      "epoch": 0.529113924050633,
      "grad_norm": 1.2750061750411987,
      "learning_rate": 9.998093400203259e-05,
      "loss": 0.7866435647010803,
      "step": 1254
    },
    {
      "epoch": 0.529957805907173,
      "grad_norm": 1.422908067703247,
      "learning_rate": 9.998026511424017e-05,
      "loss": 0.7796626687049866,
      "step": 1256
    },
    {
      "epoch": 0.5308016877637131,
      "grad_norm": 1.435552954673767,
      "learning_rate": 9.997958469696048e-05,
      "loss": 0.815027117729187,
      "step": 1258
    },
    {
      "epoch": 0.5316455696202531,
      "grad_norm": 1.1950994729995728,
      "learning_rate": 9.997889275035049e-05,
      "loss": 0.6925795674324036,
      "step": 1260
    },
    {
      "epoch": 0.5324894514767933,
      "grad_norm": 1.3049622774124146,
      "learning_rate": 9.997818927456978e-05,
      "loss": 0.822464108467102,
      "step": 1262
    },
    {
      "epoch": 0.5333333333333333,
      "grad_norm": 1.2197340726852417,
      "learning_rate": 9.997747426978066e-05,
      "loss": 0.7955381274223328,
      "step": 1264
    },
    {
      "epoch": 0.5341772151898734,
      "grad_norm": 1.2463661432266235,
      "learning_rate": 9.997674773614807e-05,
      "loss": 0.8642181754112244,
      "step": 1266
    },
    {
      "epoch": 0.5350210970464135,
      "grad_norm": 1.421393871307373,
      "learning_rate": 9.99760096738396e-05,
      "loss": 0.8776891827583313,
      "step": 1268
    },
    {
      "epoch": 0.5358649789029536,
      "grad_norm": 1.4347561597824097,
      "learning_rate": 9.997526008302549e-05,
      "loss": 0.7446491122245789,
      "step": 1270
    },
    {
      "epoch": 0.5367088607594936,
      "grad_norm": 1.2056710720062256,
      "learning_rate": 9.99744989638787e-05,
      "loss": 0.8581281304359436,
      "step": 1272
    },
    {
      "epoch": 0.5375527426160338,
      "grad_norm": 1.1672608852386475,
      "learning_rate": 9.997372631657475e-05,
      "loss": 0.7386330366134644,
      "step": 1274
    },
    {
      "epoch": 0.5383966244725739,
      "grad_norm": 1.4313966035842896,
      "learning_rate": 9.997294214129191e-05,
      "loss": 0.7806804776191711,
      "step": 1276
    },
    {
      "epoch": 0.5392405063291139,
      "grad_norm": 1.1666971445083618,
      "learning_rate": 9.997214643821107e-05,
      "loss": 0.6830351948738098,
      "step": 1278
    },
    {
      "epoch": 0.540084388185654,
      "grad_norm": 1.491783857345581,
      "learning_rate": 9.997133920751578e-05,
      "loss": 0.8570694327354431,
      "step": 1280
    },
    {
      "epoch": 0.5409282700421941,
      "grad_norm": 1.1879212856292725,
      "learning_rate": 9.997052044939226e-05,
      "loss": 0.7016772031784058,
      "step": 1282
    },
    {
      "epoch": 0.5417721518987342,
      "grad_norm": 1.2692012786865234,
      "learning_rate": 9.996969016402935e-05,
      "loss": 0.7711107134819031,
      "step": 1284
    },
    {
      "epoch": 0.5426160337552742,
      "grad_norm": 1.3318448066711426,
      "learning_rate": 9.996884835161863e-05,
      "loss": 0.7807164788246155,
      "step": 1286
    },
    {
      "epoch": 0.5434599156118144,
      "grad_norm": 1.1786744594573975,
      "learning_rate": 9.996799501235425e-05,
      "loss": 0.7331319451332092,
      "step": 1288
    },
    {
      "epoch": 0.5443037974683544,
      "grad_norm": 1.4092369079589844,
      "learning_rate": 9.996713014643309e-05,
      "loss": 0.7191547155380249,
      "step": 1290
    },
    {
      "epoch": 0.5451476793248945,
      "grad_norm": 1.377099633216858,
      "learning_rate": 9.996625375405463e-05,
      "loss": 0.7233871221542358,
      "step": 1292
    },
    {
      "epoch": 0.5459915611814345,
      "grad_norm": 1.404945969581604,
      "learning_rate": 9.996536583542105e-05,
      "loss": 0.7925472855567932,
      "step": 1294
    },
    {
      "epoch": 0.5468354430379747,
      "grad_norm": 1.2555286884307861,
      "learning_rate": 9.996446639073718e-05,
      "loss": 0.7749786376953125,
      "step": 1296
    },
    {
      "epoch": 0.5476793248945148,
      "grad_norm": 1.2577459812164307,
      "learning_rate": 9.996355542021048e-05,
      "loss": 0.7647517919540405,
      "step": 1298
    },
    {
      "epoch": 0.5485232067510548,
      "grad_norm": 1.3587758541107178,
      "learning_rate": 9.996263292405113e-05,
      "loss": 0.8621891140937805,
      "step": 1300
    },
    {
      "epoch": 0.5485232067510548,
      "eval_loss": 0.808323085308075,
      "eval_runtime": 853.577,
      "eval_samples_per_second": 2.468,
      "eval_steps_per_second": 2.468,
      "step": 1300
    },
    {
      "epoch": 0.549367088607595,
      "grad_norm": 1.327125906944275,
      "learning_rate": 9.996169890247191e-05,
      "loss": 0.749254584312439,
      "step": 1302
    },
    {
      "epoch": 0.550210970464135,
      "grad_norm": 1.4620670080184937,
      "learning_rate": 9.99607533556883e-05,
      "loss": 0.7362856268882751,
      "step": 1304
    },
    {
      "epoch": 0.5510548523206751,
      "grad_norm": 1.4119454622268677,
      "learning_rate": 9.99597962839184e-05,
      "loss": 0.7918445467948914,
      "step": 1306
    },
    {
      "epoch": 0.5518987341772152,
      "grad_norm": 1.497522234916687,
      "learning_rate": 9.995882768738298e-05,
      "loss": 0.7348005175590515,
      "step": 1308
    },
    {
      "epoch": 0.5527426160337553,
      "grad_norm": 1.535741925239563,
      "learning_rate": 9.99578475663055e-05,
      "loss": 0.8310725688934326,
      "step": 1310
    },
    {
      "epoch": 0.5535864978902953,
      "grad_norm": 1.4606215953826904,
      "learning_rate": 9.995685592091204e-05,
      "loss": 0.8232766389846802,
      "step": 1312
    },
    {
      "epoch": 0.5544303797468354,
      "grad_norm": 1.2442357540130615,
      "learning_rate": 9.995585275143136e-05,
      "loss": 0.8273071050643921,
      "step": 1314
    },
    {
      "epoch": 0.5552742616033756,
      "grad_norm": 1.5128520727157593,
      "learning_rate": 9.995483805809487e-05,
      "loss": 0.7518656253814697,
      "step": 1316
    },
    {
      "epoch": 0.5561181434599156,
      "grad_norm": 1.340149998664856,
      "learning_rate": 9.995381184113664e-05,
      "loss": 0.8261662721633911,
      "step": 1318
    },
    {
      "epoch": 0.5569620253164557,
      "grad_norm": 1.1409451961517334,
      "learning_rate": 9.99527741007934e-05,
      "loss": 0.5775256156921387,
      "step": 1320
    },
    {
      "epoch": 0.5578059071729958,
      "grad_norm": 1.3489247560501099,
      "learning_rate": 9.995172483730455e-05,
      "loss": 0.7698423862457275,
      "step": 1322
    },
    {
      "epoch": 0.5586497890295359,
      "grad_norm": 1.4950530529022217,
      "learning_rate": 9.995066405091211e-05,
      "loss": 0.8053334355354309,
      "step": 1324
    },
    {
      "epoch": 0.5594936708860759,
      "grad_norm": 1.3814653158187866,
      "learning_rate": 9.994959174186078e-05,
      "loss": 0.7826266288757324,
      "step": 1326
    },
    {
      "epoch": 0.560337552742616,
      "grad_norm": 1.3383625745773315,
      "learning_rate": 9.994850791039796e-05,
      "loss": 0.7862131595611572,
      "step": 1328
    },
    {
      "epoch": 0.5611814345991561,
      "grad_norm": 1.3529670238494873,
      "learning_rate": 9.994741255677363e-05,
      "loss": 0.8428501486778259,
      "step": 1330
    },
    {
      "epoch": 0.5620253164556962,
      "grad_norm": 1.254215121269226,
      "learning_rate": 9.994630568124049e-05,
      "loss": 0.7340869307518005,
      "step": 1332
    },
    {
      "epoch": 0.5628691983122363,
      "grad_norm": 1.2869828939437866,
      "learning_rate": 9.994518728405386e-05,
      "loss": 0.7052226662635803,
      "step": 1334
    },
    {
      "epoch": 0.5637130801687764,
      "grad_norm": 1.4321808815002441,
      "learning_rate": 9.994405736547174e-05,
      "loss": 0.8297074437141418,
      "step": 1336
    },
    {
      "epoch": 0.5645569620253165,
      "grad_norm": 1.4638891220092773,
      "learning_rate": 9.994291592575478e-05,
      "loss": 0.7183220982551575,
      "step": 1338
    },
    {
      "epoch": 0.5654008438818565,
      "grad_norm": 1.4947413206100464,
      "learning_rate": 9.994176296516628e-05,
      "loss": 0.8146093487739563,
      "step": 1340
    },
    {
      "epoch": 0.5662447257383966,
      "grad_norm": 1.343862533569336,
      "learning_rate": 9.994059848397221e-05,
      "loss": 0.7583593130111694,
      "step": 1342
    },
    {
      "epoch": 0.5670886075949367,
      "grad_norm": 1.203550100326538,
      "learning_rate": 9.993942248244121e-05,
      "loss": 0.7682924270629883,
      "step": 1344
    },
    {
      "epoch": 0.5679324894514768,
      "grad_norm": 1.287660002708435,
      "learning_rate": 9.993823496084455e-05,
      "loss": 0.8139828443527222,
      "step": 1346
    },
    {
      "epoch": 0.5687763713080168,
      "grad_norm": 1.3326014280319214,
      "learning_rate": 9.993703591945616e-05,
      "loss": 0.7529099583625793,
      "step": 1348
    },
    {
      "epoch": 0.569620253164557,
      "grad_norm": 1.2441487312316895,
      "learning_rate": 9.993582535855263e-05,
      "loss": 0.6997471451759338,
      "step": 1350
    },
    {
      "epoch": 0.570464135021097,
      "grad_norm": 1.2647649049758911,
      "learning_rate": 9.993460327841325e-05,
      "loss": 0.7421218752861023,
      "step": 1352
    },
    {
      "epoch": 0.5713080168776371,
      "grad_norm": 1.146399974822998,
      "learning_rate": 9.99333696793199e-05,
      "loss": 0.7342398166656494,
      "step": 1354
    },
    {
      "epoch": 0.5721518987341773,
      "grad_norm": 1.3346691131591797,
      "learning_rate": 9.993212456155715e-05,
      "loss": 0.7175891399383545,
      "step": 1356
    },
    {
      "epoch": 0.5729957805907173,
      "grad_norm": 1.3950672149658203,
      "learning_rate": 9.993086792541222e-05,
      "loss": 0.8108891248703003,
      "step": 1358
    },
    {
      "epoch": 0.5738396624472574,
      "grad_norm": 1.339931845664978,
      "learning_rate": 9.992959977117502e-05,
      "loss": 0.6979889273643494,
      "step": 1360
    },
    {
      "epoch": 0.5746835443037974,
      "grad_norm": 1.3276840448379517,
      "learning_rate": 9.992832009913806e-05,
      "loss": 0.7635799050331116,
      "step": 1362
    },
    {
      "epoch": 0.5755274261603376,
      "grad_norm": 1.5015610456466675,
      "learning_rate": 9.992702890959653e-05,
      "loss": 0.7575043439865112,
      "step": 1364
    },
    {
      "epoch": 0.5763713080168776,
      "grad_norm": 1.4755414724349976,
      "learning_rate": 9.99257262028483e-05,
      "loss": 0.8134847283363342,
      "step": 1366
    },
    {
      "epoch": 0.5772151898734177,
      "grad_norm": 1.3788783550262451,
      "learning_rate": 9.992441197919388e-05,
      "loss": 0.7663828134536743,
      "step": 1368
    },
    {
      "epoch": 0.5780590717299579,
      "grad_norm": 1.2814711332321167,
      "learning_rate": 9.992308623893644e-05,
      "loss": 0.6711251735687256,
      "step": 1370
    },
    {
      "epoch": 0.5789029535864979,
      "grad_norm": 1.5343635082244873,
      "learning_rate": 9.99217489823818e-05,
      "loss": 0.8097200393676758,
      "step": 1372
    },
    {
      "epoch": 0.579746835443038,
      "grad_norm": 1.3029557466506958,
      "learning_rate": 9.992040020983843e-05,
      "loss": 0.8274240493774414,
      "step": 1374
    },
    {
      "epoch": 0.580590717299578,
      "grad_norm": 1.4034144878387451,
      "learning_rate": 9.991903992161746e-05,
      "loss": 0.7758964896202087,
      "step": 1376
    },
    {
      "epoch": 0.5814345991561182,
      "grad_norm": 1.2340021133422852,
      "learning_rate": 9.991766811803271e-05,
      "loss": 0.6571930050849915,
      "step": 1378
    },
    {
      "epoch": 0.5822784810126582,
      "grad_norm": 1.3082842826843262,
      "learning_rate": 9.991628479940061e-05,
      "loss": 0.7381542921066284,
      "step": 1380
    },
    {
      "epoch": 0.5831223628691983,
      "grad_norm": 1.8134801387786865,
      "learning_rate": 9.991488996604025e-05,
      "loss": 0.8081237077713013,
      "step": 1382
    },
    {
      "epoch": 0.5839662447257384,
      "grad_norm": 1.4598309993743896,
      "learning_rate": 9.991348361827343e-05,
      "loss": 0.7761610746383667,
      "step": 1384
    },
    {
      "epoch": 0.5848101265822785,
      "grad_norm": 1.2974225282669067,
      "learning_rate": 9.991206575642453e-05,
      "loss": 0.6872953176498413,
      "step": 1386
    },
    {
      "epoch": 0.5856540084388185,
      "grad_norm": 1.24009370803833,
      "learning_rate": 9.991063638082065e-05,
      "loss": 0.7601345777511597,
      "step": 1388
    },
    {
      "epoch": 0.5864978902953587,
      "grad_norm": 1.176713228225708,
      "learning_rate": 9.99091954917915e-05,
      "loss": 0.7138593792915344,
      "step": 1390
    },
    {
      "epoch": 0.5873417721518988,
      "grad_norm": 1.1056525707244873,
      "learning_rate": 9.990774308966949e-05,
      "loss": 0.7730305194854736,
      "step": 1392
    },
    {
      "epoch": 0.5881856540084388,
      "grad_norm": 1.382847547531128,
      "learning_rate": 9.990627917478962e-05,
      "loss": 0.7076689600944519,
      "step": 1394
    },
    {
      "epoch": 0.5890295358649789,
      "grad_norm": 1.2507930994033813,
      "learning_rate": 9.990480374748964e-05,
      "loss": 0.7970513105392456,
      "step": 1396
    },
    {
      "epoch": 0.589873417721519,
      "grad_norm": 1.2266724109649658,
      "learning_rate": 9.990331680810987e-05,
      "loss": 0.7906717658042908,
      "step": 1398
    },
    {
      "epoch": 0.5907172995780591,
      "grad_norm": 1.299920916557312,
      "learning_rate": 9.99018183569933e-05,
      "loss": 0.853204607963562,
      "step": 1400
    },
    {
      "epoch": 0.5907172995780591,
      "eval_loss": 0.8009664416313171,
      "eval_runtime": 851.9417,
      "eval_samples_per_second": 2.473,
      "eval_steps_per_second": 2.473,
      "step": 1400
    },
    {
      "epoch": 0.5915611814345991,
      "grad_norm": 1.2114863395690918,
      "learning_rate": 9.990030839448564e-05,
      "loss": 0.8140703439712524,
      "step": 1402
    },
    {
      "epoch": 0.5924050632911393,
      "grad_norm": 1.3301794528961182,
      "learning_rate": 9.989878692093518e-05,
      "loss": 0.7471320629119873,
      "step": 1404
    },
    {
      "epoch": 0.5932489451476793,
      "grad_norm": 1.2611899375915527,
      "learning_rate": 9.98972539366929e-05,
      "loss": 0.7307024002075195,
      "step": 1406
    },
    {
      "epoch": 0.5940928270042194,
      "grad_norm": 1.1717802286148071,
      "learning_rate": 9.989570944211244e-05,
      "loss": 0.6843112111091614,
      "step": 1408
    },
    {
      "epoch": 0.5949367088607594,
      "grad_norm": 1.3323513269424438,
      "learning_rate": 9.989415343755006e-05,
      "loss": 0.7025372385978699,
      "step": 1410
    },
    {
      "epoch": 0.5957805907172996,
      "grad_norm": 1.4225109815597534,
      "learning_rate": 9.989258592336473e-05,
      "loss": 0.7792683839797974,
      "step": 1412
    },
    {
      "epoch": 0.5966244725738397,
      "grad_norm": 1.2878522872924805,
      "learning_rate": 9.989100689991804e-05,
      "loss": 0.8328315019607544,
      "step": 1414
    },
    {
      "epoch": 0.5974683544303797,
      "grad_norm": 1.2067214250564575,
      "learning_rate": 9.988941636757421e-05,
      "loss": 0.7700617909431458,
      "step": 1416
    },
    {
      "epoch": 0.5983122362869199,
      "grad_norm": 1.1213195323944092,
      "learning_rate": 9.988781432670019e-05,
      "loss": 0.6872363090515137,
      "step": 1418
    },
    {
      "epoch": 0.5991561181434599,
      "grad_norm": 1.3211694955825806,
      "learning_rate": 9.98862007776655e-05,
      "loss": 0.7184111475944519,
      "step": 1420
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.1916998624801636,
      "learning_rate": 9.98845757208424e-05,
      "loss": 0.8120859265327454,
      "step": 1422
    },
    {
      "epoch": 0.60084388185654,
      "grad_norm": 1.2772804498672485,
      "learning_rate": 9.988293915660572e-05,
      "loss": 0.7586462497711182,
      "step": 1424
    },
    {
      "epoch": 0.6016877637130802,
      "grad_norm": 1.4139106273651123,
      "learning_rate": 9.988129108533299e-05,
      "loss": 0.8175994157791138,
      "step": 1426
    },
    {
      "epoch": 0.6025316455696202,
      "grad_norm": 1.4481157064437866,
      "learning_rate": 9.987963150740439e-05,
      "loss": 0.7662636041641235,
      "step": 1428
    },
    {
      "epoch": 0.6033755274261603,
      "grad_norm": 1.6000999212265015,
      "learning_rate": 9.987796042320277e-05,
      "loss": 0.7477837800979614,
      "step": 1430
    },
    {
      "epoch": 0.6042194092827005,
      "grad_norm": 1.26194429397583,
      "learning_rate": 9.98762778331136e-05,
      "loss": 0.7392798662185669,
      "step": 1432
    },
    {
      "epoch": 0.6050632911392405,
      "grad_norm": 1.2370645999908447,
      "learning_rate": 9.987458373752503e-05,
      "loss": 0.7795998454093933,
      "step": 1434
    },
    {
      "epoch": 0.6059071729957806,
      "grad_norm": 1.4908311367034912,
      "learning_rate": 9.987287813682784e-05,
      "loss": 0.7833777070045471,
      "step": 1436
    },
    {
      "epoch": 0.6067510548523207,
      "grad_norm": 1.2918652296066284,
      "learning_rate": 9.987116103141549e-05,
      "loss": 0.7269768118858337,
      "step": 1438
    },
    {
      "epoch": 0.6075949367088608,
      "grad_norm": 1.2170461416244507,
      "learning_rate": 9.98694324216841e-05,
      "loss": 0.7599279284477234,
      "step": 1440
    },
    {
      "epoch": 0.6084388185654008,
      "grad_norm": 1.4373505115509033,
      "learning_rate": 9.98676923080324e-05,
      "loss": 0.8256514668464661,
      "step": 1442
    },
    {
      "epoch": 0.6092827004219409,
      "grad_norm": 1.3523614406585693,
      "learning_rate": 9.986594069086181e-05,
      "loss": 0.8462428450584412,
      "step": 1444
    },
    {
      "epoch": 0.610126582278481,
      "grad_norm": 1.5131851434707642,
      "learning_rate": 9.98641775705764e-05,
      "loss": 0.8402239084243774,
      "step": 1446
    },
    {
      "epoch": 0.6109704641350211,
      "grad_norm": 1.3518229722976685,
      "learning_rate": 9.98624029475829e-05,
      "loss": 0.7585759162902832,
      "step": 1448
    },
    {
      "epoch": 0.6118143459915611,
      "grad_norm": 1.3403998613357544,
      "learning_rate": 9.986061682229064e-05,
      "loss": 0.773881733417511,
      "step": 1450
    },
    {
      "epoch": 0.6126582278481013,
      "grad_norm": 1.1835366487503052,
      "learning_rate": 9.985881919511168e-05,
      "loss": 0.6770316958427429,
      "step": 1452
    },
    {
      "epoch": 0.6135021097046414,
      "grad_norm": 1.1825730800628662,
      "learning_rate": 9.985701006646069e-05,
      "loss": 0.7081645727157593,
      "step": 1454
    },
    {
      "epoch": 0.6143459915611814,
      "grad_norm": 1.378994345664978,
      "learning_rate": 9.9855189436755e-05,
      "loss": 0.7750917673110962,
      "step": 1456
    },
    {
      "epoch": 0.6151898734177215,
      "grad_norm": 1.4208749532699585,
      "learning_rate": 9.985335730641458e-05,
      "loss": 0.7517801523208618,
      "step": 1458
    },
    {
      "epoch": 0.6160337552742616,
      "grad_norm": 1.1413639783859253,
      "learning_rate": 9.98515136758621e-05,
      "loss": 0.712832510471344,
      "step": 1460
    },
    {
      "epoch": 0.6168776371308017,
      "grad_norm": 1.3949562311172485,
      "learning_rate": 9.984965854552283e-05,
      "loss": 0.7884142994880676,
      "step": 1462
    },
    {
      "epoch": 0.6177215189873417,
      "grad_norm": 1.4057096242904663,
      "learning_rate": 9.984779191582471e-05,
      "loss": 0.796623706817627,
      "step": 1464
    },
    {
      "epoch": 0.6185654008438819,
      "grad_norm": 1.1681689023971558,
      "learning_rate": 9.984591378719834e-05,
      "loss": 0.7862933874130249,
      "step": 1466
    },
    {
      "epoch": 0.619409282700422,
      "grad_norm": 1.2585291862487793,
      "learning_rate": 9.984402416007696e-05,
      "loss": 0.7889828681945801,
      "step": 1468
    },
    {
      "epoch": 0.620253164556962,
      "grad_norm": 1.2598098516464233,
      "learning_rate": 9.984212303489649e-05,
      "loss": 0.7375997304916382,
      "step": 1470
    },
    {
      "epoch": 0.6210970464135022,
      "grad_norm": 1.4628467559814453,
      "learning_rate": 9.984021041209547e-05,
      "loss": 0.7839564085006714,
      "step": 1472
    },
    {
      "epoch": 0.6219409282700422,
      "grad_norm": 1.3606770038604736,
      "learning_rate": 9.983828629211511e-05,
      "loss": 0.7566051483154297,
      "step": 1474
    },
    {
      "epoch": 0.6227848101265823,
      "grad_norm": 1.182644248008728,
      "learning_rate": 9.983635067539927e-05,
      "loss": 0.6638457179069519,
      "step": 1476
    },
    {
      "epoch": 0.6236286919831223,
      "grad_norm": 1.5617793798446655,
      "learning_rate": 9.983440356239445e-05,
      "loss": 0.8227225542068481,
      "step": 1478
    },
    {
      "epoch": 0.6244725738396625,
      "grad_norm": 1.2290058135986328,
      "learning_rate": 9.98324449535498e-05,
      "loss": 0.7086431980133057,
      "step": 1480
    },
    {
      "epoch": 0.6253164556962025,
      "grad_norm": 1.3822678327560425,
      "learning_rate": 9.983047484931716e-05,
      "loss": 0.8076596856117249,
      "step": 1482
    },
    {
      "epoch": 0.6261603375527426,
      "grad_norm": 1.163699746131897,
      "learning_rate": 9.982849325015098e-05,
      "loss": 0.7514539361000061,
      "step": 1484
    },
    {
      "epoch": 0.6270042194092827,
      "grad_norm": 1.2635631561279297,
      "learning_rate": 9.982650015650839e-05,
      "loss": 0.7298142910003662,
      "step": 1486
    },
    {
      "epoch": 0.6278481012658228,
      "grad_norm": 1.3135387897491455,
      "learning_rate": 9.982449556884914e-05,
      "loss": 0.8092831373214722,
      "step": 1488
    },
    {
      "epoch": 0.6286919831223629,
      "grad_norm": 1.3577877283096313,
      "learning_rate": 9.982247948763567e-05,
      "loss": 0.7934147715568542,
      "step": 1490
    },
    {
      "epoch": 0.6295358649789029,
      "grad_norm": 1.1482092142105103,
      "learning_rate": 9.982045191333304e-05,
      "loss": 0.789363443851471,
      "step": 1492
    },
    {
      "epoch": 0.6303797468354431,
      "grad_norm": 1.189771056175232,
      "learning_rate": 9.981841284640895e-05,
      "loss": 0.7458413243293762,
      "step": 1494
    },
    {
      "epoch": 0.6312236286919831,
      "grad_norm": 1.2815836668014526,
      "learning_rate": 9.981636228733383e-05,
      "loss": 0.7299918532371521,
      "step": 1496
    },
    {
      "epoch": 0.6320675105485232,
      "grad_norm": 1.36761474609375,
      "learning_rate": 9.981430023658068e-05,
      "loss": 0.7545169591903687,
      "step": 1498
    },
    {
      "epoch": 0.6329113924050633,
      "grad_norm": 1.2594345808029175,
      "learning_rate": 9.981222669462513e-05,
      "loss": 0.7358481884002686,
      "step": 1500
    },
    {
      "epoch": 0.6329113924050633,
      "eval_loss": 0.7896141409873962,
      "eval_runtime": 865.9069,
      "eval_samples_per_second": 2.433,
      "eval_steps_per_second": 2.433,
      "step": 1500
    },
    {
      "epoch": 0.6337552742616034,
      "grad_norm": 3.6419246196746826,
      "learning_rate": 9.981014166194556e-05,
      "loss": 0.8253764510154724,
      "step": 1502
    },
    {
      "epoch": 0.6345991561181434,
      "grad_norm": 1.7333487272262573,
      "learning_rate": 9.980804513902294e-05,
      "loss": 0.8254884481430054,
      "step": 1504
    },
    {
      "epoch": 0.6354430379746835,
      "grad_norm": 1.1998231410980225,
      "learning_rate": 9.980593712634088e-05,
      "loss": 0.7833738327026367,
      "step": 1506
    },
    {
      "epoch": 0.6362869198312237,
      "grad_norm": 1.347011685371399,
      "learning_rate": 9.980381762438566e-05,
      "loss": 0.753408670425415,
      "step": 1508
    },
    {
      "epoch": 0.6371308016877637,
      "grad_norm": 1.1759053468704224,
      "learning_rate": 9.980168663364622e-05,
      "loss": 0.7867791652679443,
      "step": 1510
    },
    {
      "epoch": 0.6379746835443038,
      "grad_norm": 1.3113552331924438,
      "learning_rate": 9.979954415461412e-05,
      "loss": 0.6753612160682678,
      "step": 1512
    },
    {
      "epoch": 0.6388185654008439,
      "grad_norm": 1.3258320093154907,
      "learning_rate": 9.979739018778362e-05,
      "loss": 0.750367283821106,
      "step": 1514
    },
    {
      "epoch": 0.639662447257384,
      "grad_norm": 1.175145149230957,
      "learning_rate": 9.979522473365157e-05,
      "loss": 0.7505861520767212,
      "step": 1516
    },
    {
      "epoch": 0.640506329113924,
      "grad_norm": 1.2276148796081543,
      "learning_rate": 9.979304779271752e-05,
      "loss": 0.7429317831993103,
      "step": 1518
    },
    {
      "epoch": 0.6413502109704642,
      "grad_norm": 1.3262875080108643,
      "learning_rate": 9.979085936548362e-05,
      "loss": 0.786217212677002,
      "step": 1520
    },
    {
      "epoch": 0.6421940928270042,
      "grad_norm": 1.3067121505737305,
      "learning_rate": 9.978865945245473e-05,
      "loss": 0.6942036151885986,
      "step": 1522
    },
    {
      "epoch": 0.6430379746835443,
      "grad_norm": 1.5352400541305542,
      "learning_rate": 9.978644805413832e-05,
      "loss": 0.8281817436218262,
      "step": 1524
    },
    {
      "epoch": 0.6438818565400843,
      "grad_norm": 1.2848507165908813,
      "learning_rate": 9.97842251710445e-05,
      "loss": 0.8110972046852112,
      "step": 1526
    },
    {
      "epoch": 0.6447257383966245,
      "grad_norm": 1.352196216583252,
      "learning_rate": 9.978199080368607e-05,
      "loss": 0.7354730367660522,
      "step": 1528
    },
    {
      "epoch": 0.6455696202531646,
      "grad_norm": 1.2427687644958496,
      "learning_rate": 9.977974495257842e-05,
      "loss": 0.7915583848953247,
      "step": 1530
    },
    {
      "epoch": 0.6464135021097046,
      "grad_norm": 1.3163504600524902,
      "learning_rate": 9.977748761823967e-05,
      "loss": 0.7400109171867371,
      "step": 1532
    },
    {
      "epoch": 0.6472573839662448,
      "grad_norm": 1.2496893405914307,
      "learning_rate": 9.977521880119049e-05,
      "loss": 0.7104899287223816,
      "step": 1534
    },
    {
      "epoch": 0.6481012658227848,
      "grad_norm": 1.0907179117202759,
      "learning_rate": 9.97729385019543e-05,
      "loss": 0.8074463605880737,
      "step": 1536
    },
    {
      "epoch": 0.6489451476793249,
      "grad_norm": 1.2323429584503174,
      "learning_rate": 9.977064672105712e-05,
      "loss": 0.7770540714263916,
      "step": 1538
    },
    {
      "epoch": 0.6497890295358649,
      "grad_norm": 1.224428415298462,
      "learning_rate": 9.976834345902759e-05,
      "loss": 0.806465208530426,
      "step": 1540
    },
    {
      "epoch": 0.6506329113924051,
      "grad_norm": 1.3529564142227173,
      "learning_rate": 9.976602871639705e-05,
      "loss": 0.7306749224662781,
      "step": 1542
    },
    {
      "epoch": 0.6514767932489451,
      "grad_norm": 1.1770031452178955,
      "learning_rate": 9.976370249369946e-05,
      "loss": 0.783933699131012,
      "step": 1544
    },
    {
      "epoch": 0.6523206751054852,
      "grad_norm": 1.205283522605896,
      "learning_rate": 9.976136479147144e-05,
      "loss": 0.6937689185142517,
      "step": 1546
    },
    {
      "epoch": 0.6531645569620254,
      "grad_norm": 1.2329360246658325,
      "learning_rate": 9.975901561025223e-05,
      "loss": 0.8041763305664062,
      "step": 1548
    },
    {
      "epoch": 0.6540084388185654,
      "grad_norm": 1.499973177909851,
      "learning_rate": 9.975665495058377e-05,
      "loss": 0.750390887260437,
      "step": 1550
    },
    {
      "epoch": 0.6548523206751055,
      "grad_norm": 1.31832754611969,
      "learning_rate": 9.975428281301061e-05,
      "loss": 0.7658298015594482,
      "step": 1552
    },
    {
      "epoch": 0.6556962025316456,
      "grad_norm": 1.3998414278030396,
      "learning_rate": 9.975189919807994e-05,
      "loss": 0.8651264905929565,
      "step": 1554
    },
    {
      "epoch": 0.6565400843881857,
      "grad_norm": 1.2002551555633545,
      "learning_rate": 9.974950410634164e-05,
      "loss": 0.6776561141014099,
      "step": 1556
    },
    {
      "epoch": 0.6573839662447257,
      "grad_norm": 1.1986602544784546,
      "learning_rate": 9.97470975383482e-05,
      "loss": 0.8159130811691284,
      "step": 1558
    },
    {
      "epoch": 0.6582278481012658,
      "grad_norm": 1.3583602905273438,
      "learning_rate": 9.974467949465477e-05,
      "loss": 0.7528039216995239,
      "step": 1560
    },
    {
      "epoch": 0.6590717299578059,
      "grad_norm": 1.4176239967346191,
      "learning_rate": 9.974224997581913e-05,
      "loss": 0.6970920562744141,
      "step": 1562
    },
    {
      "epoch": 0.659915611814346,
      "grad_norm": 1.3899401426315308,
      "learning_rate": 9.973980898240177e-05,
      "loss": 0.7718377113342285,
      "step": 1564
    },
    {
      "epoch": 0.660759493670886,
      "grad_norm": 1.222413182258606,
      "learning_rate": 9.973735651496571e-05,
      "loss": 0.7346280217170715,
      "step": 1566
    },
    {
      "epoch": 0.6616033755274262,
      "grad_norm": 1.3750087022781372,
      "learning_rate": 9.973489257407676e-05,
      "loss": 0.7923588156700134,
      "step": 1568
    },
    {
      "epoch": 0.6624472573839663,
      "grad_norm": 1.24547278881073,
      "learning_rate": 9.973241716030325e-05,
      "loss": 0.8258910179138184,
      "step": 1570
    },
    {
      "epoch": 0.6632911392405063,
      "grad_norm": 1.2464141845703125,
      "learning_rate": 9.972993027421624e-05,
      "loss": 0.7869232296943665,
      "step": 1572
    },
    {
      "epoch": 0.6641350210970464,
      "grad_norm": 1.3088903427124023,
      "learning_rate": 9.972743191638939e-05,
      "loss": 0.8144775629043579,
      "step": 1574
    },
    {
      "epoch": 0.6649789029535865,
      "grad_norm": 1.2252418994903564,
      "learning_rate": 9.972492208739903e-05,
      "loss": 0.7432073950767517,
      "step": 1576
    },
    {
      "epoch": 0.6658227848101266,
      "grad_norm": 1.2303717136383057,
      "learning_rate": 9.972240078782413e-05,
      "loss": 0.7386854887008667,
      "step": 1578
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 1.0226294994354248,
      "learning_rate": 9.971986801824631e-05,
      "loss": 0.7127882838249207,
      "step": 1580
    },
    {
      "epoch": 0.6675105485232068,
      "grad_norm": 1.362332820892334,
      "learning_rate": 9.971732377924982e-05,
      "loss": 0.7557716369628906,
      "step": 1582
    },
    {
      "epoch": 0.6683544303797468,
      "grad_norm": 1.4436695575714111,
      "learning_rate": 9.971476807142158e-05,
      "loss": 0.7832611203193665,
      "step": 1584
    },
    {
      "epoch": 0.6691983122362869,
      "grad_norm": 1.276695966720581,
      "learning_rate": 9.971220089535113e-05,
      "loss": 0.8190197944641113,
      "step": 1586
    },
    {
      "epoch": 0.6700421940928271,
      "grad_norm": 1.2413527965545654,
      "learning_rate": 9.970962225163069e-05,
      "loss": 0.747222363948822,
      "step": 1588
    },
    {
      "epoch": 0.6708860759493671,
      "grad_norm": 1.3395767211914062,
      "learning_rate": 9.970703214085507e-05,
      "loss": 0.7846449017524719,
      "step": 1590
    },
    {
      "epoch": 0.6717299578059072,
      "grad_norm": 1.291327953338623,
      "learning_rate": 9.970443056362178e-05,
      "loss": 0.8160232901573181,
      "step": 1592
    },
    {
      "epoch": 0.6725738396624472,
      "grad_norm": 1.3139684200286865,
      "learning_rate": 9.970181752053097e-05,
      "loss": 0.7413806915283203,
      "step": 1594
    },
    {
      "epoch": 0.6734177215189874,
      "grad_norm": 1.3170921802520752,
      "learning_rate": 9.969919301218537e-05,
      "loss": 0.7637304067611694,
      "step": 1596
    },
    {
      "epoch": 0.6742616033755274,
      "grad_norm": 1.3349758386611938,
      "learning_rate": 9.969655703919044e-05,
      "loss": 0.7823366522789001,
      "step": 1598
    },
    {
      "epoch": 0.6751054852320675,
      "grad_norm": 1.2151578664779663,
      "learning_rate": 9.969390960215425e-05,
      "loss": 0.6587790846824646,
      "step": 1600
    },
    {
      "epoch": 0.6751054852320675,
      "eval_loss": 0.7836604714393616,
      "eval_runtime": 861.5352,
      "eval_samples_per_second": 2.446,
      "eval_steps_per_second": 2.446,
      "step": 1600
    },
    {
      "epoch": 0.6759493670886076,
      "grad_norm": 1.2541478872299194,
      "learning_rate": 9.96912507016875e-05,
      "loss": 0.7314544320106506,
      "step": 1602
    },
    {
      "epoch": 0.6767932489451477,
      "grad_norm": 1.091790795326233,
      "learning_rate": 9.968858033840357e-05,
      "loss": 0.702468752861023,
      "step": 1604
    },
    {
      "epoch": 0.6776371308016877,
      "grad_norm": 1.36745285987854,
      "learning_rate": 9.968589851291841e-05,
      "loss": 0.7691897749900818,
      "step": 1606
    },
    {
      "epoch": 0.6784810126582278,
      "grad_norm": 1.1325993537902832,
      "learning_rate": 9.968320522585072e-05,
      "loss": 0.7422228455543518,
      "step": 1608
    },
    {
      "epoch": 0.679324894514768,
      "grad_norm": 1.1015450954437256,
      "learning_rate": 9.968050047782176e-05,
      "loss": 0.677532434463501,
      "step": 1610
    },
    {
      "epoch": 0.680168776371308,
      "grad_norm": 1.2216695547103882,
      "learning_rate": 9.967778426945548e-05,
      "loss": 0.7973438501358032,
      "step": 1612
    },
    {
      "epoch": 0.6810126582278481,
      "grad_norm": 1.159395456314087,
      "learning_rate": 9.967505660137843e-05,
      "loss": 0.6742876172065735,
      "step": 1614
    },
    {
      "epoch": 0.6818565400843882,
      "grad_norm": 1.404433250427246,
      "learning_rate": 9.967231747421988e-05,
      "loss": 0.7592008709907532,
      "step": 1616
    },
    {
      "epoch": 0.6827004219409283,
      "grad_norm": 1.2489168643951416,
      "learning_rate": 9.966956688861164e-05,
      "loss": 0.7565826177597046,
      "step": 1618
    },
    {
      "epoch": 0.6835443037974683,
      "grad_norm": 1.2960615158081055,
      "learning_rate": 9.966680484518825e-05,
      "loss": 0.7694597840309143,
      "step": 1620
    },
    {
      "epoch": 0.6843881856540084,
      "grad_norm": 1.3598436117172241,
      "learning_rate": 9.966403134458685e-05,
      "loss": 0.8392959833145142,
      "step": 1622
    },
    {
      "epoch": 0.6852320675105485,
      "grad_norm": 1.258065938949585,
      "learning_rate": 9.966124638744722e-05,
      "loss": 0.8014217019081116,
      "step": 1624
    },
    {
      "epoch": 0.6860759493670886,
      "grad_norm": 1.3132309913635254,
      "learning_rate": 9.965844997441184e-05,
      "loss": 0.7029755711555481,
      "step": 1626
    },
    {
      "epoch": 0.6869198312236287,
      "grad_norm": 1.1204946041107178,
      "learning_rate": 9.965564210612575e-05,
      "loss": 0.7213528752326965,
      "step": 1628
    },
    {
      "epoch": 0.6877637130801688,
      "grad_norm": 1.037251591682434,
      "learning_rate": 9.965282278323667e-05,
      "loss": 0.6895437240600586,
      "step": 1630
    },
    {
      "epoch": 0.6886075949367089,
      "grad_norm": 1.093807578086853,
      "learning_rate": 9.964999200639498e-05,
      "loss": 0.8035063743591309,
      "step": 1632
    },
    {
      "epoch": 0.6894514767932489,
      "grad_norm": 1.367386817932129,
      "learning_rate": 9.964714977625367e-05,
      "loss": 0.6191847920417786,
      "step": 1634
    },
    {
      "epoch": 0.6902953586497891,
      "grad_norm": 1.3160961866378784,
      "learning_rate": 9.964429609346841e-05,
      "loss": 0.7469727993011475,
      "step": 1636
    },
    {
      "epoch": 0.6911392405063291,
      "grad_norm": 1.3736863136291504,
      "learning_rate": 9.964143095869748e-05,
      "loss": 0.7987836599349976,
      "step": 1638
    },
    {
      "epoch": 0.6919831223628692,
      "grad_norm": 1.323209524154663,
      "learning_rate": 9.963855437260182e-05,
      "loss": 0.7901709675788879,
      "step": 1640
    },
    {
      "epoch": 0.6928270042194092,
      "grad_norm": 1.3943440914154053,
      "learning_rate": 9.963566633584496e-05,
      "loss": 0.7889530658721924,
      "step": 1642
    },
    {
      "epoch": 0.6936708860759494,
      "grad_norm": 1.3699116706848145,
      "learning_rate": 9.963276684909317e-05,
      "loss": 0.756829559803009,
      "step": 1644
    },
    {
      "epoch": 0.6945147679324895,
      "grad_norm": 1.4216378927230835,
      "learning_rate": 9.962985591301529e-05,
      "loss": 0.7840303182601929,
      "step": 1646
    },
    {
      "epoch": 0.6953586497890295,
      "grad_norm": 1.2231985330581665,
      "learning_rate": 9.962693352828279e-05,
      "loss": 0.700393557548523,
      "step": 1648
    },
    {
      "epoch": 0.6962025316455697,
      "grad_norm": 1.3568313121795654,
      "learning_rate": 9.962399969556983e-05,
      "loss": 0.7010306715965271,
      "step": 1650
    },
    {
      "epoch": 0.6970464135021097,
      "grad_norm": 1.1662907600402832,
      "learning_rate": 9.96210544155532e-05,
      "loss": 0.6935506463050842,
      "step": 1652
    },
    {
      "epoch": 0.6978902953586498,
      "grad_norm": 1.3066680431365967,
      "learning_rate": 9.96180976889123e-05,
      "loss": 0.7913851141929626,
      "step": 1654
    },
    {
      "epoch": 0.6987341772151898,
      "grad_norm": 1.2268375158309937,
      "learning_rate": 9.961512951632918e-05,
      "loss": 0.764849066734314,
      "step": 1656
    },
    {
      "epoch": 0.69957805907173,
      "grad_norm": 1.4509469270706177,
      "learning_rate": 9.96121498984886e-05,
      "loss": 0.7544103860855103,
      "step": 1658
    },
    {
      "epoch": 0.70042194092827,
      "grad_norm": 1.200772762298584,
      "learning_rate": 9.960915883607782e-05,
      "loss": 0.7766591310501099,
      "step": 1660
    },
    {
      "epoch": 0.7012658227848101,
      "grad_norm": 1.3825311660766602,
      "learning_rate": 9.960615632978687e-05,
      "loss": 0.7433559894561768,
      "step": 1662
    },
    {
      "epoch": 0.7021097046413503,
      "grad_norm": 1.3197243213653564,
      "learning_rate": 9.960314238030836e-05,
      "loss": 0.7770103812217712,
      "step": 1664
    },
    {
      "epoch": 0.7029535864978903,
      "grad_norm": 1.515163779258728,
      "learning_rate": 9.960011698833755e-05,
      "loss": 0.8597216606140137,
      "step": 1666
    },
    {
      "epoch": 0.7037974683544304,
      "grad_norm": 1.2329891920089722,
      "learning_rate": 9.959708015457234e-05,
      "loss": 0.7630532383918762,
      "step": 1668
    },
    {
      "epoch": 0.7046413502109705,
      "grad_norm": 1.0592037439346313,
      "learning_rate": 9.959403187971327e-05,
      "loss": 0.7299806475639343,
      "step": 1670
    },
    {
      "epoch": 0.7054852320675106,
      "grad_norm": 2.2717394828796387,
      "learning_rate": 9.959097216446351e-05,
      "loss": 0.6999854445457458,
      "step": 1672
    },
    {
      "epoch": 0.7063291139240506,
      "grad_norm": 1.1552131175994873,
      "learning_rate": 9.958790100952889e-05,
      "loss": 0.8403060436248779,
      "step": 1674
    },
    {
      "epoch": 0.7071729957805907,
      "grad_norm": 1.290488839149475,
      "learning_rate": 9.958481841561787e-05,
      "loss": 0.7729134559631348,
      "step": 1676
    },
    {
      "epoch": 0.7080168776371308,
      "grad_norm": 1.1913278102874756,
      "learning_rate": 9.958172438344152e-05,
      "loss": 0.7100697755813599,
      "step": 1678
    },
    {
      "epoch": 0.7088607594936709,
      "grad_norm": 1.2355852127075195,
      "learning_rate": 9.957861891371359e-05,
      "loss": 0.7014795541763306,
      "step": 1680
    },
    {
      "epoch": 0.7097046413502109,
      "grad_norm": 1.258705496788025,
      "learning_rate": 9.957550200715044e-05,
      "loss": 0.8131424784660339,
      "step": 1682
    },
    {
      "epoch": 0.7105485232067511,
      "grad_norm": 1.1102997064590454,
      "learning_rate": 9.957237366447112e-05,
      "loss": 0.6842480301856995,
      "step": 1684
    },
    {
      "epoch": 0.7113924050632912,
      "grad_norm": 1.4466290473937988,
      "learning_rate": 9.956923388639724e-05,
      "loss": 0.6730120182037354,
      "step": 1686
    },
    {
      "epoch": 0.7122362869198312,
      "grad_norm": 1.261152982711792,
      "learning_rate": 9.956608267365311e-05,
      "loss": 0.7109374403953552,
      "step": 1688
    },
    {
      "epoch": 0.7130801687763713,
      "grad_norm": 1.4070630073547363,
      "learning_rate": 9.956292002696562e-05,
      "loss": 0.7545008063316345,
      "step": 1690
    },
    {
      "epoch": 0.7139240506329114,
      "grad_norm": 1.2532793283462524,
      "learning_rate": 9.955974594706436e-05,
      "loss": 0.7892587184906006,
      "step": 1692
    },
    {
      "epoch": 0.7147679324894515,
      "grad_norm": 1.1180293560028076,
      "learning_rate": 9.955656043468153e-05,
      "loss": 0.7348554134368896,
      "step": 1694
    },
    {
      "epoch": 0.7156118143459915,
      "grad_norm": 1.333054542541504,
      "learning_rate": 9.955336349055195e-05,
      "loss": 0.8207674026489258,
      "step": 1696
    },
    {
      "epoch": 0.7164556962025317,
      "grad_norm": 1.1373547315597534,
      "learning_rate": 9.95501551154131e-05,
      "loss": 0.7226691842079163,
      "step": 1698
    },
    {
      "epoch": 0.7172995780590717,
      "grad_norm": 1.2342052459716797,
      "learning_rate": 9.95469353100051e-05,
      "loss": 0.726982831954956,
      "step": 1700
    },
    {
      "epoch": 0.7172995780590717,
      "eval_loss": 0.7783148884773254,
      "eval_runtime": 846.1986,
      "eval_samples_per_second": 2.49,
      "eval_steps_per_second": 2.49,
      "step": 1700
    },
    {
      "epoch": 0.7181434599156118,
      "grad_norm": 1.3781483173370361,
      "learning_rate": 9.95437040750707e-05,
      "loss": 0.7623077034950256,
      "step": 1702
    },
    {
      "epoch": 0.7189873417721518,
      "grad_norm": 1.301440715789795,
      "learning_rate": 9.954046141135526e-05,
      "loss": 0.7421616315841675,
      "step": 1704
    },
    {
      "epoch": 0.719831223628692,
      "grad_norm": 1.1375854015350342,
      "learning_rate": 9.953720731960683e-05,
      "loss": 0.685523509979248,
      "step": 1706
    },
    {
      "epoch": 0.7206751054852321,
      "grad_norm": 1.2014397382736206,
      "learning_rate": 9.953394180057604e-05,
      "loss": 0.756073534488678,
      "step": 1708
    },
    {
      "epoch": 0.7215189873417721,
      "grad_norm": 1.232802152633667,
      "learning_rate": 9.95306648550162e-05,
      "loss": 0.7364522814750671,
      "step": 1710
    },
    {
      "epoch": 0.7223628691983123,
      "grad_norm": 1.4462472200393677,
      "learning_rate": 9.952737648368323e-05,
      "loss": 0.7073688507080078,
      "step": 1712
    },
    {
      "epoch": 0.7232067510548523,
      "grad_norm": 1.123523473739624,
      "learning_rate": 9.95240766873357e-05,
      "loss": 0.7147064805030823,
      "step": 1714
    },
    {
      "epoch": 0.7240506329113924,
      "grad_norm": 1.4111510515213013,
      "learning_rate": 9.95207654667348e-05,
      "loss": 0.7108398079872131,
      "step": 1716
    },
    {
      "epoch": 0.7248945147679325,
      "grad_norm": 1.2785903215408325,
      "learning_rate": 9.951744282264437e-05,
      "loss": 0.7080079317092896,
      "step": 1718
    },
    {
      "epoch": 0.7257383966244726,
      "grad_norm": 1.1361653804779053,
      "learning_rate": 9.951410875583089e-05,
      "loss": 0.7396624684333801,
      "step": 1720
    },
    {
      "epoch": 0.7265822784810126,
      "grad_norm": 1.0762585401535034,
      "learning_rate": 9.951076326706346e-05,
      "loss": 0.7724334597587585,
      "step": 1722
    },
    {
      "epoch": 0.7274261603375527,
      "grad_norm": 1.3104428052902222,
      "learning_rate": 9.950740635711379e-05,
      "loss": 0.7311923503875732,
      "step": 1724
    },
    {
      "epoch": 0.7282700421940929,
      "grad_norm": 1.1291942596435547,
      "learning_rate": 9.95040380267563e-05,
      "loss": 0.6878296732902527,
      "step": 1726
    },
    {
      "epoch": 0.7291139240506329,
      "grad_norm": 1.5171746015548706,
      "learning_rate": 9.9500658276768e-05,
      "loss": 0.7410538196563721,
      "step": 1728
    },
    {
      "epoch": 0.729957805907173,
      "grad_norm": 1.0966423749923706,
      "learning_rate": 9.949726710792848e-05,
      "loss": 0.6953532695770264,
      "step": 1730
    },
    {
      "epoch": 0.7308016877637131,
      "grad_norm": 1.2436997890472412,
      "learning_rate": 9.949386452102007e-05,
      "loss": 0.6679023504257202,
      "step": 1732
    },
    {
      "epoch": 0.7316455696202532,
      "grad_norm": 1.1364835500717163,
      "learning_rate": 9.949045051682766e-05,
      "loss": 0.8046789765357971,
      "step": 1734
    },
    {
      "epoch": 0.7324894514767932,
      "grad_norm": 1.296648383140564,
      "learning_rate": 9.948702509613878e-05,
      "loss": 0.7322937846183777,
      "step": 1736
    },
    {
      "epoch": 0.7333333333333333,
      "grad_norm": 1.2355525493621826,
      "learning_rate": 9.948358825974365e-05,
      "loss": 0.7442626357078552,
      "step": 1738
    },
    {
      "epoch": 0.7341772151898734,
      "grad_norm": 1.1634451150894165,
      "learning_rate": 9.948014000843504e-05,
      "loss": 0.7231078743934631,
      "step": 1740
    },
    {
      "epoch": 0.7350210970464135,
      "grad_norm": 1.1500129699707031,
      "learning_rate": 9.947668034300843e-05,
      "loss": 0.6436833143234253,
      "step": 1742
    },
    {
      "epoch": 0.7358649789029535,
      "grad_norm": 1.3881278038024902,
      "learning_rate": 9.947320926426189e-05,
      "loss": 0.8170580863952637,
      "step": 1744
    },
    {
      "epoch": 0.7367088607594937,
      "grad_norm": 1.3479492664337158,
      "learning_rate": 9.94697267729961e-05,
      "loss": 0.7830947041511536,
      "step": 1746
    },
    {
      "epoch": 0.7375527426160338,
      "grad_norm": 1.0187158584594727,
      "learning_rate": 9.946623287001444e-05,
      "loss": 0.7358533143997192,
      "step": 1748
    },
    {
      "epoch": 0.7383966244725738,
      "grad_norm": 1.2575689554214478,
      "learning_rate": 9.946272755612287e-05,
      "loss": 0.7279790639877319,
      "step": 1750
    },
    {
      "epoch": 0.739240506329114,
      "grad_norm": 1.2045027017593384,
      "learning_rate": 9.945921083213002e-05,
      "loss": 0.6953092217445374,
      "step": 1752
    },
    {
      "epoch": 0.740084388185654,
      "grad_norm": 1.3994466066360474,
      "learning_rate": 9.945568269884708e-05,
      "loss": 0.8094141483306885,
      "step": 1754
    },
    {
      "epoch": 0.7409282700421941,
      "grad_norm": 1.2892286777496338,
      "learning_rate": 9.945214315708797e-05,
      "loss": 0.6979201436042786,
      "step": 1756
    },
    {
      "epoch": 0.7417721518987341,
      "grad_norm": 1.2006971836090088,
      "learning_rate": 9.944859220766919e-05,
      "loss": 0.6810774803161621,
      "step": 1758
    },
    {
      "epoch": 0.7426160337552743,
      "grad_norm": 1.055793285369873,
      "learning_rate": 9.944502985140986e-05,
      "loss": 0.6796762347221375,
      "step": 1760
    },
    {
      "epoch": 0.7434599156118143,
      "grad_norm": 1.174714207649231,
      "learning_rate": 9.944145608913175e-05,
      "loss": 0.7954121828079224,
      "step": 1762
    },
    {
      "epoch": 0.7443037974683544,
      "grad_norm": 1.1638222932815552,
      "learning_rate": 9.943787092165926e-05,
      "loss": 0.6939491629600525,
      "step": 1764
    },
    {
      "epoch": 0.7451476793248946,
      "grad_norm": 1.1861820220947266,
      "learning_rate": 9.943427434981942e-05,
      "loss": 0.8112956285476685,
      "step": 1766
    },
    {
      "epoch": 0.7459915611814346,
      "grad_norm": 0.9667421579360962,
      "learning_rate": 9.943066637444189e-05,
      "loss": 0.6812481880187988,
      "step": 1768
    },
    {
      "epoch": 0.7468354430379747,
      "grad_norm": 1.2826191186904907,
      "learning_rate": 9.942704699635898e-05,
      "loss": 0.7598370313644409,
      "step": 1770
    },
    {
      "epoch": 0.7476793248945147,
      "grad_norm": 1.2257909774780273,
      "learning_rate": 9.942341621640558e-05,
      "loss": 0.7118877172470093,
      "step": 1772
    },
    {
      "epoch": 0.7485232067510549,
      "grad_norm": 1.5224615335464478,
      "learning_rate": 9.941977403541925e-05,
      "loss": 0.8037024736404419,
      "step": 1774
    },
    {
      "epoch": 0.7493670886075949,
      "grad_norm": 1.188689947128296,
      "learning_rate": 9.941612045424018e-05,
      "loss": 0.6795828938484192,
      "step": 1776
    },
    {
      "epoch": 0.750210970464135,
      "grad_norm": 1.0685369968414307,
      "learning_rate": 9.941245547371116e-05,
      "loss": 0.6934568881988525,
      "step": 1778
    },
    {
      "epoch": 0.7510548523206751,
      "grad_norm": 1.1643654108047485,
      "learning_rate": 9.940877909467767e-05,
      "loss": 0.6883851289749146,
      "step": 1780
    },
    {
      "epoch": 0.7518987341772152,
      "grad_norm": 1.15621018409729,
      "learning_rate": 9.940509131798775e-05,
      "loss": 0.8284637928009033,
      "step": 1782
    },
    {
      "epoch": 0.7527426160337553,
      "grad_norm": 1.1946302652359009,
      "learning_rate": 9.94013921444921e-05,
      "loss": 0.7108310461044312,
      "step": 1784
    },
    {
      "epoch": 0.7535864978902953,
      "grad_norm": 1.1536555290222168,
      "learning_rate": 9.939768157504404e-05,
      "loss": 0.7166154384613037,
      "step": 1786
    },
    {
      "epoch": 0.7544303797468355,
      "grad_norm": 1.3184611797332764,
      "learning_rate": 9.939395961049956e-05,
      "loss": 0.7774572372436523,
      "step": 1788
    },
    {
      "epoch": 0.7552742616033755,
      "grad_norm": 1.0782374143600464,
      "learning_rate": 9.939022625171723e-05,
      "loss": 0.7386471033096313,
      "step": 1790
    },
    {
      "epoch": 0.7561181434599156,
      "grad_norm": 1.1616696119308472,
      "learning_rate": 9.938648149955824e-05,
      "loss": 0.6495215892791748,
      "step": 1792
    },
    {
      "epoch": 0.7569620253164557,
      "grad_norm": 1.1715892553329468,
      "learning_rate": 9.938272535488647e-05,
      "loss": 0.7733646631240845,
      "step": 1794
    },
    {
      "epoch": 0.7578059071729958,
      "grad_norm": 1.203466773033142,
      "learning_rate": 9.937895781856838e-05,
      "loss": 0.7354782223701477,
      "step": 1796
    },
    {
      "epoch": 0.7586497890295358,
      "grad_norm": 1.246559977531433,
      "learning_rate": 9.937517889147305e-05,
      "loss": 0.823226273059845,
      "step": 1798
    },
    {
      "epoch": 0.759493670886076,
      "grad_norm": 0.9968833923339844,
      "learning_rate": 9.937138857447221e-05,
      "loss": 0.6221681833267212,
      "step": 1800
    },
    {
      "epoch": 0.759493670886076,
      "eval_loss": 0.7719914317131042,
      "eval_runtime": 853.1943,
      "eval_samples_per_second": 2.47,
      "eval_steps_per_second": 2.47,
      "step": 1800
    },
    {
      "epoch": 0.760337552742616,
      "grad_norm": 1.5454338788986206,
      "learning_rate": 9.936758686844024e-05,
      "loss": 0.7799059152603149,
      "step": 1802
    },
    {
      "epoch": 0.7611814345991561,
      "grad_norm": 1.1954455375671387,
      "learning_rate": 9.936377377425409e-05,
      "loss": 0.653838038444519,
      "step": 1804
    },
    {
      "epoch": 0.7620253164556962,
      "grad_norm": 1.2538350820541382,
      "learning_rate": 9.935994929279339e-05,
      "loss": 0.7046942710876465,
      "step": 1806
    },
    {
      "epoch": 0.7628691983122363,
      "grad_norm": 1.2358729839324951,
      "learning_rate": 9.935611342494035e-05,
      "loss": 0.7821131348609924,
      "step": 1808
    },
    {
      "epoch": 0.7637130801687764,
      "grad_norm": 1.2401310205459595,
      "learning_rate": 9.935226617157986e-05,
      "loss": 0.7594596147537231,
      "step": 1810
    },
    {
      "epoch": 0.7645569620253164,
      "grad_norm": 1.3197205066680908,
      "learning_rate": 9.934840753359938e-05,
      "loss": 0.7512493133544922,
      "step": 1812
    },
    {
      "epoch": 0.7654008438818566,
      "grad_norm": 1.2482305765151978,
      "learning_rate": 9.934453751188903e-05,
      "loss": 0.6953311562538147,
      "step": 1814
    },
    {
      "epoch": 0.7662447257383966,
      "grad_norm": 1.5995157957077026,
      "learning_rate": 9.934065610734157e-05,
      "loss": 0.7699819803237915,
      "step": 1816
    },
    {
      "epoch": 0.7670886075949367,
      "grad_norm": 1.2414922714233398,
      "learning_rate": 9.933676332085235e-05,
      "loss": 0.6532001495361328,
      "step": 1818
    },
    {
      "epoch": 0.7679324894514767,
      "grad_norm": 1.2274713516235352,
      "learning_rate": 9.933285915331937e-05,
      "loss": 0.7716373801231384,
      "step": 1820
    },
    {
      "epoch": 0.7687763713080169,
      "grad_norm": 1.2894618511199951,
      "learning_rate": 9.932894360564322e-05,
      "loss": 0.7002654671669006,
      "step": 1822
    },
    {
      "epoch": 0.769620253164557,
      "grad_norm": 1.10796320438385,
      "learning_rate": 9.932501667872718e-05,
      "loss": 0.7970587015151978,
      "step": 1824
    },
    {
      "epoch": 0.770464135021097,
      "grad_norm": 1.2393653392791748,
      "learning_rate": 9.932107837347708e-05,
      "loss": 0.8071644306182861,
      "step": 1826
    },
    {
      "epoch": 0.7713080168776372,
      "grad_norm": 1.1999030113220215,
      "learning_rate": 9.931712869080144e-05,
      "loss": 0.7376157641410828,
      "step": 1828
    },
    {
      "epoch": 0.7721518987341772,
      "grad_norm": 1.1166026592254639,
      "learning_rate": 9.931316763161135e-05,
      "loss": 0.7487053275108337,
      "step": 1830
    },
    {
      "epoch": 0.7729957805907173,
      "grad_norm": 1.1788052320480347,
      "learning_rate": 9.930919519682059e-05,
      "loss": 0.733161985874176,
      "step": 1832
    },
    {
      "epoch": 0.7738396624472574,
      "grad_norm": 1.309968113899231,
      "learning_rate": 9.930521138734548e-05,
      "loss": 0.7907692790031433,
      "step": 1834
    },
    {
      "epoch": 0.7746835443037975,
      "grad_norm": 1.1685889959335327,
      "learning_rate": 9.930121620410502e-05,
      "loss": 0.7192210555076599,
      "step": 1836
    },
    {
      "epoch": 0.7755274261603375,
      "grad_norm": 1.2243701219558716,
      "learning_rate": 9.929720964802085e-05,
      "loss": 0.7394438982009888,
      "step": 1838
    },
    {
      "epoch": 0.7763713080168776,
      "grad_norm": 1.2940958738327026,
      "learning_rate": 9.929319172001717e-05,
      "loss": 0.7885041832923889,
      "step": 1840
    },
    {
      "epoch": 0.7772151898734178,
      "grad_norm": 1.0952763557434082,
      "learning_rate": 9.928916242102086e-05,
      "loss": 0.6822885274887085,
      "step": 1842
    },
    {
      "epoch": 0.7780590717299578,
      "grad_norm": 1.0333503484725952,
      "learning_rate": 9.928512175196139e-05,
      "loss": 0.7070927619934082,
      "step": 1844
    },
    {
      "epoch": 0.7789029535864979,
      "grad_norm": 1.201359510421753,
      "learning_rate": 9.928106971377088e-05,
      "loss": 0.7041296362876892,
      "step": 1846
    },
    {
      "epoch": 0.779746835443038,
      "grad_norm": 1.5381278991699219,
      "learning_rate": 9.927700630738404e-05,
      "loss": 0.6630192995071411,
      "step": 1848
    },
    {
      "epoch": 0.7805907172995781,
      "grad_norm": 1.2858322858810425,
      "learning_rate": 9.927293153373823e-05,
      "loss": 0.7628101110458374,
      "step": 1850
    },
    {
      "epoch": 0.7814345991561181,
      "grad_norm": 1.3730580806732178,
      "learning_rate": 9.926884539377343e-05,
      "loss": 0.7557390928268433,
      "step": 1852
    },
    {
      "epoch": 0.7822784810126582,
      "grad_norm": 1.4954931735992432,
      "learning_rate": 9.92647478884322e-05,
      "loss": 0.8217329978942871,
      "step": 1854
    },
    {
      "epoch": 0.7831223628691983,
      "grad_norm": 1.1092652082443237,
      "learning_rate": 9.92606390186598e-05,
      "loss": 0.672879695892334,
      "step": 1856
    },
    {
      "epoch": 0.7839662447257384,
      "grad_norm": 1.2077893018722534,
      "learning_rate": 9.925651878540404e-05,
      "loss": 0.7380653619766235,
      "step": 1858
    },
    {
      "epoch": 0.7848101265822784,
      "grad_norm": 1.0789313316345215,
      "learning_rate": 9.925238718961538e-05,
      "loss": 0.6648160219192505,
      "step": 1860
    },
    {
      "epoch": 0.7856540084388186,
      "grad_norm": 1.3950812816619873,
      "learning_rate": 9.924824423224692e-05,
      "loss": 0.8316769003868103,
      "step": 1862
    },
    {
      "epoch": 0.7864978902953587,
      "grad_norm": 1.3934763669967651,
      "learning_rate": 9.924408991425433e-05,
      "loss": 0.7901778817176819,
      "step": 1864
    },
    {
      "epoch": 0.7873417721518987,
      "grad_norm": 1.2191659212112427,
      "learning_rate": 9.923992423659596e-05,
      "loss": 0.7643826007843018,
      "step": 1866
    },
    {
      "epoch": 0.7881856540084389,
      "grad_norm": 0.986673891544342,
      "learning_rate": 9.923574720023274e-05,
      "loss": 0.6314064860343933,
      "step": 1868
    },
    {
      "epoch": 0.7890295358649789,
      "grad_norm": 1.003552794456482,
      "learning_rate": 9.923155880612823e-05,
      "loss": 0.8244763016700745,
      "step": 1870
    },
    {
      "epoch": 0.789873417721519,
      "grad_norm": 1.0831382274627686,
      "learning_rate": 9.92273590552486e-05,
      "loss": 0.7398403882980347,
      "step": 1872
    },
    {
      "epoch": 0.790717299578059,
      "grad_norm": 1.1782667636871338,
      "learning_rate": 9.922314794856267e-05,
      "loss": 0.735211968421936,
      "step": 1874
    },
    {
      "epoch": 0.7915611814345992,
      "grad_norm": 2.230534076690674,
      "learning_rate": 9.921892548704186e-05,
      "loss": 0.7550510764122009,
      "step": 1876
    },
    {
      "epoch": 0.7924050632911392,
      "grad_norm": 1.0191401243209839,
      "learning_rate": 9.92146916716602e-05,
      "loss": 0.7676286697387695,
      "step": 1878
    },
    {
      "epoch": 0.7932489451476793,
      "grad_norm": 1.1347072124481201,
      "learning_rate": 9.921044650339438e-05,
      "loss": 0.7409467697143555,
      "step": 1880
    },
    {
      "epoch": 0.7940928270042195,
      "grad_norm": 1.107528567314148,
      "learning_rate": 9.920618998322364e-05,
      "loss": 0.7760165333747864,
      "step": 1882
    },
    {
      "epoch": 0.7949367088607595,
      "grad_norm": 1.1110666990280151,
      "learning_rate": 9.92019221121299e-05,
      "loss": 0.7360131740570068,
      "step": 1884
    },
    {
      "epoch": 0.7957805907172996,
      "grad_norm": 1.267580509185791,
      "learning_rate": 9.919764289109765e-05,
      "loss": 0.7784845232963562,
      "step": 1886
    },
    {
      "epoch": 0.7966244725738396,
      "grad_norm": 1.5894557237625122,
      "learning_rate": 9.919335232111407e-05,
      "loss": 0.7880831360816956,
      "step": 1888
    },
    {
      "epoch": 0.7974683544303798,
      "grad_norm": 1.1906384229660034,
      "learning_rate": 9.918905040316886e-05,
      "loss": 0.7315587997436523,
      "step": 1890
    },
    {
      "epoch": 0.7983122362869198,
      "grad_norm": 1.3626811504364014,
      "learning_rate": 9.918473713825445e-05,
      "loss": 0.7808622121810913,
      "step": 1892
    },
    {
      "epoch": 0.7991561181434599,
      "grad_norm": 1.1801300048828125,
      "learning_rate": 9.918041252736577e-05,
      "loss": 0.7055642604827881,
      "step": 1894
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.2669063806533813,
      "learning_rate": 9.917607657150046e-05,
      "loss": 0.7188893556594849,
      "step": 1896
    },
    {
      "epoch": 0.8008438818565401,
      "grad_norm": 1.1746855974197388,
      "learning_rate": 9.91717292716587e-05,
      "loss": 0.7787454128265381,
      "step": 1898
    },
    {
      "epoch": 0.8016877637130801,
      "grad_norm": 1.120012640953064,
      "learning_rate": 9.916737062884338e-05,
      "loss": 0.720715343952179,
      "step": 1900
    },
    {
      "epoch": 0.8016877637130801,
      "eval_loss": 0.7648926973342896,
      "eval_runtime": 865.9394,
      "eval_samples_per_second": 2.433,
      "eval_steps_per_second": 2.433,
      "step": 1900
    },
    {
      "epoch": 0.8025316455696202,
      "grad_norm": 1.1745549440383911,
      "learning_rate": 9.916300064405993e-05,
      "loss": 0.7544789910316467,
      "step": 1902
    },
    {
      "epoch": 0.8033755274261604,
      "grad_norm": 1.1439874172210693,
      "learning_rate": 9.915861931831643e-05,
      "loss": 0.7479203343391418,
      "step": 1904
    },
    {
      "epoch": 0.8042194092827004,
      "grad_norm": 1.3508219718933105,
      "learning_rate": 9.915422665262356e-05,
      "loss": 0.6995842456817627,
      "step": 1906
    },
    {
      "epoch": 0.8050632911392405,
      "grad_norm": 1.1519006490707397,
      "learning_rate": 9.914982264799462e-05,
      "loss": 0.7152725458145142,
      "step": 1908
    },
    {
      "epoch": 0.8059071729957806,
      "grad_norm": 1.0818005800247192,
      "learning_rate": 9.914540730544554e-05,
      "loss": 0.7105516195297241,
      "step": 1910
    },
    {
      "epoch": 0.8067510548523207,
      "grad_norm": 1.1611127853393555,
      "learning_rate": 9.914098062599485e-05,
      "loss": 0.6911059617996216,
      "step": 1912
    },
    {
      "epoch": 0.8075949367088607,
      "grad_norm": 1.1964445114135742,
      "learning_rate": 9.91365426106637e-05,
      "loss": 0.6897286772727966,
      "step": 1914
    },
    {
      "epoch": 0.8084388185654009,
      "grad_norm": 1.3873497247695923,
      "learning_rate": 9.913209326047585e-05,
      "loss": 0.7263250350952148,
      "step": 1916
    },
    {
      "epoch": 0.809282700421941,
      "grad_norm": 1.1729894876480103,
      "learning_rate": 9.91276325764577e-05,
      "loss": 0.7045295238494873,
      "step": 1918
    },
    {
      "epoch": 0.810126582278481,
      "grad_norm": 0.9089694619178772,
      "learning_rate": 9.912316055963822e-05,
      "loss": 0.587131142616272,
      "step": 1920
    },
    {
      "epoch": 0.810970464135021,
      "grad_norm": 1.2051384449005127,
      "learning_rate": 9.911867721104902e-05,
      "loss": 0.7237880229949951,
      "step": 1922
    },
    {
      "epoch": 0.8118143459915612,
      "grad_norm": 1.2152670621871948,
      "learning_rate": 9.911418253172433e-05,
      "loss": 0.6967294216156006,
      "step": 1924
    },
    {
      "epoch": 0.8126582278481013,
      "grad_norm": 1.1193642616271973,
      "learning_rate": 9.9109676522701e-05,
      "loss": 0.7636315822601318,
      "step": 1926
    },
    {
      "epoch": 0.8135021097046413,
      "grad_norm": 1.2457597255706787,
      "learning_rate": 9.910515918501843e-05,
      "loss": 0.7451969981193542,
      "step": 1928
    },
    {
      "epoch": 0.8143459915611815,
      "grad_norm": 1.057009220123291,
      "learning_rate": 9.910063051971876e-05,
      "loss": 0.6320056319236755,
      "step": 1930
    },
    {
      "epoch": 0.8151898734177215,
      "grad_norm": 1.2820258140563965,
      "learning_rate": 9.909609052784661e-05,
      "loss": 0.691004753112793,
      "step": 1932
    },
    {
      "epoch": 0.8160337552742616,
      "grad_norm": 1.331312656402588,
      "learning_rate": 9.909153921044927e-05,
      "loss": 0.7741923332214355,
      "step": 1934
    },
    {
      "epoch": 0.8168776371308016,
      "grad_norm": 1.2055360078811646,
      "learning_rate": 9.908697656857668e-05,
      "loss": 0.668049156665802,
      "step": 1936
    },
    {
      "epoch": 0.8177215189873418,
      "grad_norm": 1.2124541997909546,
      "learning_rate": 9.90824026032813e-05,
      "loss": 0.6584748029708862,
      "step": 1938
    },
    {
      "epoch": 0.8185654008438819,
      "grad_norm": 1.244288682937622,
      "learning_rate": 9.90778173156183e-05,
      "loss": 0.7081992626190186,
      "step": 1940
    },
    {
      "epoch": 0.8194092827004219,
      "grad_norm": 1.250558853149414,
      "learning_rate": 9.907322070664542e-05,
      "loss": 0.7977840900421143,
      "step": 1942
    },
    {
      "epoch": 0.8202531645569621,
      "grad_norm": 1.3892892599105835,
      "learning_rate": 9.906861277742297e-05,
      "loss": 0.7830103635787964,
      "step": 1944
    },
    {
      "epoch": 0.8210970464135021,
      "grad_norm": 1.3152644634246826,
      "learning_rate": 9.906399352901393e-05,
      "loss": 0.8451479077339172,
      "step": 1946
    },
    {
      "epoch": 0.8219409282700422,
      "grad_norm": 1.1102250814437866,
      "learning_rate": 9.905936296248388e-05,
      "loss": 0.7035528421401978,
      "step": 1948
    },
    {
      "epoch": 0.8227848101265823,
      "grad_norm": 1.0271214246749878,
      "learning_rate": 9.905472107890101e-05,
      "loss": 0.764616847038269,
      "step": 1950
    },
    {
      "epoch": 0.8236286919831224,
      "grad_norm": 1.1772255897521973,
      "learning_rate": 9.905006787933609e-05,
      "loss": 0.7699717283248901,
      "step": 1952
    },
    {
      "epoch": 0.8244725738396624,
      "grad_norm": 1.2486404180526733,
      "learning_rate": 9.904540336486252e-05,
      "loss": 0.7755605578422546,
      "step": 1954
    },
    {
      "epoch": 0.8253164556962025,
      "grad_norm": 1.070148229598999,
      "learning_rate": 9.904072753655635e-05,
      "loss": 0.688934326171875,
      "step": 1956
    },
    {
      "epoch": 0.8261603375527427,
      "grad_norm": 1.118401288986206,
      "learning_rate": 9.903604039549617e-05,
      "loss": 0.7447791695594788,
      "step": 1958
    },
    {
      "epoch": 0.8270042194092827,
      "grad_norm": 1.2209899425506592,
      "learning_rate": 9.903134194276323e-05,
      "loss": 0.7990683317184448,
      "step": 1960
    },
    {
      "epoch": 0.8278481012658228,
      "grad_norm": 1.296093225479126,
      "learning_rate": 9.902663217944137e-05,
      "loss": 0.7290873527526855,
      "step": 1962
    },
    {
      "epoch": 0.8286919831223629,
      "grad_norm": 1.2594937086105347,
      "learning_rate": 9.902191110661704e-05,
      "loss": 0.7971217036247253,
      "step": 1964
    },
    {
      "epoch": 0.829535864978903,
      "grad_norm": 1.6016536951065063,
      "learning_rate": 9.90171787253793e-05,
      "loss": 0.6728768348693848,
      "step": 1966
    },
    {
      "epoch": 0.830379746835443,
      "grad_norm": 3.3128950595855713,
      "learning_rate": 9.901243503681983e-05,
      "loss": 0.7684211730957031,
      "step": 1968
    },
    {
      "epoch": 0.8312236286919831,
      "grad_norm": 1.2970373630523682,
      "learning_rate": 9.90076800420329e-05,
      "loss": 0.756637454032898,
      "step": 1970
    },
    {
      "epoch": 0.8320675105485232,
      "grad_norm": 1.1388959884643555,
      "learning_rate": 9.900291374211538e-05,
      "loss": 0.6692084074020386,
      "step": 1972
    },
    {
      "epoch": 0.8329113924050633,
      "grad_norm": 1.050641655921936,
      "learning_rate": 9.899813613816677e-05,
      "loss": 0.7298309803009033,
      "step": 1974
    },
    {
      "epoch": 0.8337552742616033,
      "grad_norm": 1.2598577737808228,
      "learning_rate": 9.899334723128922e-05,
      "loss": 0.6886547803878784,
      "step": 1976
    },
    {
      "epoch": 0.8345991561181435,
      "grad_norm": 1.2800767421722412,
      "learning_rate": 9.898854702258735e-05,
      "loss": 0.745341420173645,
      "step": 1978
    },
    {
      "epoch": 0.8354430379746836,
      "grad_norm": 1.1923155784606934,
      "learning_rate": 9.898373551316856e-05,
      "loss": 0.7133575081825256,
      "step": 1980
    },
    {
      "epoch": 0.8362869198312236,
      "grad_norm": 1.156121015548706,
      "learning_rate": 9.897891270414272e-05,
      "loss": 0.8117790818214417,
      "step": 1982
    },
    {
      "epoch": 0.8371308016877637,
      "grad_norm": 1.0400618314743042,
      "learning_rate": 9.897407859662238e-05,
      "loss": 0.6094260215759277,
      "step": 1984
    },
    {
      "epoch": 0.8379746835443038,
      "grad_norm": 1.451953411102295,
      "learning_rate": 9.896923319172268e-05,
      "loss": 0.7680332064628601,
      "step": 1986
    },
    {
      "epoch": 0.8388185654008439,
      "grad_norm": 1.2560248374938965,
      "learning_rate": 9.896437649056134e-05,
      "loss": 0.6918784379959106,
      "step": 1988
    },
    {
      "epoch": 0.8396624472573839,
      "grad_norm": 1.2744325399398804,
      "learning_rate": 9.895950849425874e-05,
      "loss": 0.7654696106910706,
      "step": 1990
    },
    {
      "epoch": 0.8405063291139241,
      "grad_norm": 1.304439902305603,
      "learning_rate": 9.895462920393781e-05,
      "loss": 0.7585932612419128,
      "step": 1992
    },
    {
      "epoch": 0.8413502109704641,
      "grad_norm": 1.578957200050354,
      "learning_rate": 9.89497386207241e-05,
      "loss": 0.7474164962768555,
      "step": 1994
    },
    {
      "epoch": 0.8421940928270042,
      "grad_norm": 1.0358996391296387,
      "learning_rate": 9.89448367457458e-05,
      "loss": 0.663844883441925,
      "step": 1996
    },
    {
      "epoch": 0.8430379746835444,
      "grad_norm": 1.2285103797912598,
      "learning_rate": 9.893992358013366e-05,
      "loss": 0.7578557729721069,
      "step": 1998
    },
    {
      "epoch": 0.8438818565400844,
      "grad_norm": 1.2051875591278076,
      "learning_rate": 9.893499912502108e-05,
      "loss": 0.7795036435127258,
      "step": 2000
    },
    {
      "epoch": 0.8438818565400844,
      "eval_loss": 0.7587011456489563,
      "eval_runtime": 856.2276,
      "eval_samples_per_second": 2.461,
      "eval_steps_per_second": 2.461,
      "step": 2000
    },
    {
      "epoch": 0.8447257383966245,
      "grad_norm": 1.145434021949768,
      "learning_rate": 9.893006338154401e-05,
      "loss": 0.731850802898407,
      "step": 2002
    },
    {
      "epoch": 0.8455696202531645,
      "grad_norm": 1.0618077516555786,
      "learning_rate": 9.892511635084101e-05,
      "loss": 0.6711665391921997,
      "step": 2004
    },
    {
      "epoch": 0.8464135021097047,
      "grad_norm": 1.1657867431640625,
      "learning_rate": 9.892015803405331e-05,
      "loss": 0.6894803643226624,
      "step": 2006
    },
    {
      "epoch": 0.8472573839662447,
      "grad_norm": 1.080140233039856,
      "learning_rate": 9.891518843232467e-05,
      "loss": 0.628146231174469,
      "step": 2008
    },
    {
      "epoch": 0.8481012658227848,
      "grad_norm": 1.0664509534835815,
      "learning_rate": 9.891020754680151e-05,
      "loss": 0.740858793258667,
      "step": 2010
    },
    {
      "epoch": 0.8489451476793249,
      "grad_norm": 1.5567615032196045,
      "learning_rate": 9.89052153786328e-05,
      "loss": 0.7763919234275818,
      "step": 2012
    },
    {
      "epoch": 0.849789029535865,
      "grad_norm": 1.4347095489501953,
      "learning_rate": 9.890021192897016e-05,
      "loss": 0.8131396770477295,
      "step": 2014
    },
    {
      "epoch": 0.850632911392405,
      "grad_norm": 1.1787892580032349,
      "learning_rate": 9.889519719896776e-05,
      "loss": 0.6829051375389099,
      "step": 2016
    },
    {
      "epoch": 0.8514767932489451,
      "grad_norm": 1.239745855331421,
      "learning_rate": 9.889017118978241e-05,
      "loss": 0.7664558291435242,
      "step": 2018
    },
    {
      "epoch": 0.8523206751054853,
      "grad_norm": 1.1224207878112793,
      "learning_rate": 9.888513390257352e-05,
      "loss": 0.7307376861572266,
      "step": 2020
    },
    {
      "epoch": 0.8531645569620253,
      "grad_norm": 1.100536823272705,
      "learning_rate": 9.88800853385031e-05,
      "loss": 0.6786578893661499,
      "step": 2022
    },
    {
      "epoch": 0.8540084388185654,
      "grad_norm": 1.25773024559021,
      "learning_rate": 9.887502549873576e-05,
      "loss": 0.7971984148025513,
      "step": 2024
    },
    {
      "epoch": 0.8548523206751055,
      "grad_norm": 0.9980104565620422,
      "learning_rate": 9.886995438443868e-05,
      "loss": 0.6990941166877747,
      "step": 2026
    },
    {
      "epoch": 0.8556962025316456,
      "grad_norm": 1.0464621782302856,
      "learning_rate": 9.886487199678171e-05,
      "loss": 0.763938307762146,
      "step": 2028
    },
    {
      "epoch": 0.8565400843881856,
      "grad_norm": 1.2303017377853394,
      "learning_rate": 9.885977833693724e-05,
      "loss": 0.7165632247924805,
      "step": 2030
    },
    {
      "epoch": 0.8573839662447258,
      "grad_norm": 1.2203325033187866,
      "learning_rate": 9.885467340608027e-05,
      "loss": 0.7586364150047302,
      "step": 2032
    },
    {
      "epoch": 0.8582278481012658,
      "grad_norm": 1.113882064819336,
      "learning_rate": 9.884955720538843e-05,
      "loss": 0.703253984451294,
      "step": 2034
    },
    {
      "epoch": 0.8590717299578059,
      "grad_norm": 1.1731632947921753,
      "learning_rate": 9.88444297360419e-05,
      "loss": 0.8530917763710022,
      "step": 2036
    },
    {
      "epoch": 0.859915611814346,
      "grad_norm": 1.4592338800430298,
      "learning_rate": 9.883929099922349e-05,
      "loss": 0.8166638612747192,
      "step": 2038
    },
    {
      "epoch": 0.8607594936708861,
      "grad_norm": 1.1279125213623047,
      "learning_rate": 9.883414099611864e-05,
      "loss": 0.6762415170669556,
      "step": 2040
    },
    {
      "epoch": 0.8616033755274262,
      "grad_norm": 1.1587293148040771,
      "learning_rate": 9.882897972791534e-05,
      "loss": 0.6826539039611816,
      "step": 2042
    },
    {
      "epoch": 0.8624472573839662,
      "grad_norm": 1.1909502744674683,
      "learning_rate": 9.88238071958042e-05,
      "loss": 0.7372410893440247,
      "step": 2044
    },
    {
      "epoch": 0.8632911392405064,
      "grad_norm": 1.0340155363082886,
      "learning_rate": 9.881862340097841e-05,
      "loss": 0.699260950088501,
      "step": 2046
    },
    {
      "epoch": 0.8641350210970464,
      "grad_norm": 1.1745870113372803,
      "learning_rate": 9.881342834463379e-05,
      "loss": 0.7689789533615112,
      "step": 2048
    },
    {
      "epoch": 0.8649789029535865,
      "grad_norm": 1.0003606081008911,
      "learning_rate": 9.880822202796872e-05,
      "loss": 0.6877372860908508,
      "step": 2050
    },
    {
      "epoch": 0.8658227848101265,
      "grad_norm": 1.2546781301498413,
      "learning_rate": 9.88030044521842e-05,
      "loss": 0.7632413506507874,
      "step": 2052
    },
    {
      "epoch": 0.8666666666666667,
      "grad_norm": 1.1178704500198364,
      "learning_rate": 9.879777561848385e-05,
      "loss": 0.6776729822158813,
      "step": 2054
    },
    {
      "epoch": 0.8675105485232067,
      "grad_norm": 1.523606777191162,
      "learning_rate": 9.879253552807384e-05,
      "loss": 0.7592973709106445,
      "step": 2056
    },
    {
      "epoch": 0.8683544303797468,
      "grad_norm": 1.3490995168685913,
      "learning_rate": 9.878728418216296e-05,
      "loss": 0.8028839230537415,
      "step": 2058
    },
    {
      "epoch": 0.869198312236287,
      "grad_norm": 1.1851624250411987,
      "learning_rate": 9.87820215819626e-05,
      "loss": 0.7499933838844299,
      "step": 2060
    },
    {
      "epoch": 0.870042194092827,
      "grad_norm": 1.1877925395965576,
      "learning_rate": 9.877674772868672e-05,
      "loss": 0.7324717044830322,
      "step": 2062
    },
    {
      "epoch": 0.8708860759493671,
      "grad_norm": 1.2982885837554932,
      "learning_rate": 9.877146262355194e-05,
      "loss": 0.7456585168838501,
      "step": 2064
    },
    {
      "epoch": 0.8717299578059071,
      "grad_norm": 1.043912649154663,
      "learning_rate": 9.876616626777739e-05,
      "loss": 0.7552799582481384,
      "step": 2066
    },
    {
      "epoch": 0.8725738396624473,
      "grad_norm": 1.172580599784851,
      "learning_rate": 9.876085866258487e-05,
      "loss": 0.6964990496635437,
      "step": 2068
    },
    {
      "epoch": 0.8734177215189873,
      "grad_norm": 1.26815927028656,
      "learning_rate": 9.875553980919871e-05,
      "loss": 0.7368612289428711,
      "step": 2070
    },
    {
      "epoch": 0.8742616033755274,
      "grad_norm": 1.1268136501312256,
      "learning_rate": 9.875020970884587e-05,
      "loss": 0.7400802969932556,
      "step": 2072
    },
    {
      "epoch": 0.8751054852320675,
      "grad_norm": 1.0556721687316895,
      "learning_rate": 9.874486836275594e-05,
      "loss": 0.6931334137916565,
      "step": 2074
    },
    {
      "epoch": 0.8759493670886076,
      "grad_norm": 1.1967823505401611,
      "learning_rate": 9.873951577216106e-05,
      "loss": 0.7124089002609253,
      "step": 2076
    },
    {
      "epoch": 0.8767932489451477,
      "grad_norm": 1.1753164529800415,
      "learning_rate": 9.873415193829591e-05,
      "loss": 0.7462030053138733,
      "step": 2078
    },
    {
      "epoch": 0.8776371308016878,
      "grad_norm": 1.326923131942749,
      "learning_rate": 9.872877686239789e-05,
      "loss": 0.778078019618988,
      "step": 2080
    },
    {
      "epoch": 0.8784810126582279,
      "grad_norm": 1.1472662687301636,
      "learning_rate": 9.87233905457069e-05,
      "loss": 0.6592919826507568,
      "step": 2082
    },
    {
      "epoch": 0.8793248945147679,
      "grad_norm": 1.1162762641906738,
      "learning_rate": 9.871799298946544e-05,
      "loss": 0.661717414855957,
      "step": 2084
    },
    {
      "epoch": 0.880168776371308,
      "grad_norm": 1.1694408655166626,
      "learning_rate": 9.871258419491866e-05,
      "loss": 0.6203670501708984,
      "step": 2086
    },
    {
      "epoch": 0.8810126582278481,
      "grad_norm": 1.229691505432129,
      "learning_rate": 9.870716416331425e-05,
      "loss": 0.758888304233551,
      "step": 2088
    },
    {
      "epoch": 0.8818565400843882,
      "grad_norm": 1.540377140045166,
      "learning_rate": 9.870173289590251e-05,
      "loss": 0.760649561882019,
      "step": 2090
    },
    {
      "epoch": 0.8827004219409282,
      "grad_norm": 1.173628568649292,
      "learning_rate": 9.869629039393632e-05,
      "loss": 0.6981227397918701,
      "step": 2092
    },
    {
      "epoch": 0.8835443037974684,
      "grad_norm": 1.1404013633728027,
      "learning_rate": 9.869083665867116e-05,
      "loss": 0.7808336615562439,
      "step": 2094
    },
    {
      "epoch": 0.8843881856540085,
      "grad_norm": 1.1038721799850464,
      "learning_rate": 9.868537169136511e-05,
      "loss": 0.7540555596351624,
      "step": 2096
    },
    {
      "epoch": 0.8852320675105485,
      "grad_norm": 1.1510080099105835,
      "learning_rate": 9.867989549327885e-05,
      "loss": 0.6650454998016357,
      "step": 2098
    },
    {
      "epoch": 0.8860759493670886,
      "grad_norm": 1.166912317276001,
      "learning_rate": 9.867440806567561e-05,
      "loss": 0.673769474029541,
      "step": 2100
    },
    {
      "epoch": 0.8860759493670886,
      "eval_loss": 0.7559094429016113,
      "eval_runtime": 847.8311,
      "eval_samples_per_second": 2.485,
      "eval_steps_per_second": 2.485,
      "step": 2100
    },
    {
      "epoch": 0.8869198312236287,
      "grad_norm": 1.227583885192871,
      "learning_rate": 9.866890940982121e-05,
      "loss": 0.8314241766929626,
      "step": 2102
    },
    {
      "epoch": 0.8877637130801688,
      "grad_norm": 1.1813976764678955,
      "learning_rate": 9.866339952698413e-05,
      "loss": 0.6770843863487244,
      "step": 2104
    },
    {
      "epoch": 0.8886075949367088,
      "grad_norm": 1.2471063137054443,
      "learning_rate": 9.865787841843539e-05,
      "loss": 0.7142292857170105,
      "step": 2106
    },
    {
      "epoch": 0.889451476793249,
      "grad_norm": 1.1602860689163208,
      "learning_rate": 9.865234608544858e-05,
      "loss": 0.6981731653213501,
      "step": 2108
    },
    {
      "epoch": 0.890295358649789,
      "grad_norm": 1.145677089691162,
      "learning_rate": 9.864680252929992e-05,
      "loss": 0.7019379138946533,
      "step": 2110
    },
    {
      "epoch": 0.8911392405063291,
      "grad_norm": 1.2222462892532349,
      "learning_rate": 9.86412477512682e-05,
      "loss": 0.7690986394882202,
      "step": 2112
    },
    {
      "epoch": 0.8919831223628693,
      "grad_norm": 1.1288166046142578,
      "learning_rate": 9.863568175263478e-05,
      "loss": 0.7241792678833008,
      "step": 2114
    },
    {
      "epoch": 0.8928270042194093,
      "grad_norm": 1.1773978471755981,
      "learning_rate": 9.863010453468364e-05,
      "loss": 0.7392162084579468,
      "step": 2116
    },
    {
      "epoch": 0.8936708860759494,
      "grad_norm": 1.102638840675354,
      "learning_rate": 9.862451609870136e-05,
      "loss": 0.7603078484535217,
      "step": 2118
    },
    {
      "epoch": 0.8945147679324894,
      "grad_norm": 1.1325360536575317,
      "learning_rate": 9.861891644597707e-05,
      "loss": 0.6804911494255066,
      "step": 2120
    },
    {
      "epoch": 0.8953586497890296,
      "grad_norm": 1.1381969451904297,
      "learning_rate": 9.86133055778025e-05,
      "loss": 0.787288248538971,
      "step": 2122
    },
    {
      "epoch": 0.8962025316455696,
      "grad_norm": 1.2454546689987183,
      "learning_rate": 9.860768349547196e-05,
      "loss": 0.7282505035400391,
      "step": 2124
    },
    {
      "epoch": 0.8970464135021097,
      "grad_norm": 1.2568305730819702,
      "learning_rate": 9.860205020028237e-05,
      "loss": 0.7554803490638733,
      "step": 2126
    },
    {
      "epoch": 0.8978902953586498,
      "grad_norm": 1.1523523330688477,
      "learning_rate": 9.859640569353321e-05,
      "loss": 0.7126525044441223,
      "step": 2128
    },
    {
      "epoch": 0.8987341772151899,
      "grad_norm": 1.314878225326538,
      "learning_rate": 9.859074997652658e-05,
      "loss": 0.7300811409950256,
      "step": 2130
    },
    {
      "epoch": 0.8995780590717299,
      "grad_norm": 1.1272218227386475,
      "learning_rate": 9.858508305056713e-05,
      "loss": 0.7217329144477844,
      "step": 2132
    },
    {
      "epoch": 0.90042194092827,
      "grad_norm": 1.10934317111969,
      "learning_rate": 9.857940491696211e-05,
      "loss": 0.714308500289917,
      "step": 2134
    },
    {
      "epoch": 0.9012658227848102,
      "grad_norm": 1.1991039514541626,
      "learning_rate": 9.857371557702136e-05,
      "loss": 0.6613366007804871,
      "step": 2136
    },
    {
      "epoch": 0.9021097046413502,
      "grad_norm": 1.3176918029785156,
      "learning_rate": 9.85680150320573e-05,
      "loss": 0.6972863078117371,
      "step": 2138
    },
    {
      "epoch": 0.9029535864978903,
      "grad_norm": 1.1966592073440552,
      "learning_rate": 9.856230328338496e-05,
      "loss": 0.7299100160598755,
      "step": 2140
    },
    {
      "epoch": 0.9037974683544304,
      "grad_norm": 1.2889270782470703,
      "learning_rate": 9.85565803323219e-05,
      "loss": 0.7145020961761475,
      "step": 2142
    },
    {
      "epoch": 0.9046413502109705,
      "grad_norm": 1.2112789154052734,
      "learning_rate": 9.855084618018828e-05,
      "loss": 0.6717942953109741,
      "step": 2144
    },
    {
      "epoch": 0.9054852320675105,
      "grad_norm": 1.2550239562988281,
      "learning_rate": 9.85451008283069e-05,
      "loss": 0.7460196018218994,
      "step": 2146
    },
    {
      "epoch": 0.9063291139240506,
      "grad_norm": 1.2926387786865234,
      "learning_rate": 9.853934427800309e-05,
      "loss": 0.8300626873970032,
      "step": 2148
    },
    {
      "epoch": 0.9071729957805907,
      "grad_norm": 1.0690672397613525,
      "learning_rate": 9.853357653060478e-05,
      "loss": 0.715215802192688,
      "step": 2150
    },
    {
      "epoch": 0.9080168776371308,
      "grad_norm": 1.1021424531936646,
      "learning_rate": 9.852779758744245e-05,
      "loss": 0.7021427154541016,
      "step": 2152
    },
    {
      "epoch": 0.9088607594936708,
      "grad_norm": 1.0713517665863037,
      "learning_rate": 9.852200744984921e-05,
      "loss": 0.7576406598091125,
      "step": 2154
    },
    {
      "epoch": 0.909704641350211,
      "grad_norm": 1.277526617050171,
      "learning_rate": 9.851620611916075e-05,
      "loss": 0.7008846998214722,
      "step": 2156
    },
    {
      "epoch": 0.9105485232067511,
      "grad_norm": 1.2434618473052979,
      "learning_rate": 9.85103935967153e-05,
      "loss": 0.7536613345146179,
      "step": 2158
    },
    {
      "epoch": 0.9113924050632911,
      "grad_norm": 1.1654841899871826,
      "learning_rate": 9.850456988385371e-05,
      "loss": 0.7435567378997803,
      "step": 2160
    },
    {
      "epoch": 0.9122362869198313,
      "grad_norm": 1.0718246698379517,
      "learning_rate": 9.849873498191939e-05,
      "loss": 0.7725666165351868,
      "step": 2162
    },
    {
      "epoch": 0.9130801687763713,
      "grad_norm": 1.3425630331039429,
      "learning_rate": 9.849288889225835e-05,
      "loss": 0.7833593487739563,
      "step": 2164
    },
    {
      "epoch": 0.9139240506329114,
      "grad_norm": 1.1989985704421997,
      "learning_rate": 9.848703161621917e-05,
      "loss": 0.7290158867835999,
      "step": 2166
    },
    {
      "epoch": 0.9147679324894514,
      "grad_norm": 1.0549380779266357,
      "learning_rate": 9.8481163155153e-05,
      "loss": 0.6787996888160706,
      "step": 2168
    },
    {
      "epoch": 0.9156118143459916,
      "grad_norm": 1.0757017135620117,
      "learning_rate": 9.847528351041359e-05,
      "loss": 0.7645748853683472,
      "step": 2170
    },
    {
      "epoch": 0.9164556962025316,
      "grad_norm": 1.0636975765228271,
      "learning_rate": 9.846939268335726e-05,
      "loss": 0.6640698313713074,
      "step": 2172
    },
    {
      "epoch": 0.9172995780590717,
      "grad_norm": 1.2038439512252808,
      "learning_rate": 9.846349067534291e-05,
      "loss": 0.7216284275054932,
      "step": 2174
    },
    {
      "epoch": 0.9181434599156119,
      "grad_norm": 1.17854642868042,
      "learning_rate": 9.845757748773203e-05,
      "loss": 0.7244991660118103,
      "step": 2176
    },
    {
      "epoch": 0.9189873417721519,
      "grad_norm": 1.0391159057617188,
      "learning_rate": 9.845165312188864e-05,
      "loss": 0.6043152809143066,
      "step": 2178
    },
    {
      "epoch": 0.919831223628692,
      "grad_norm": 1.2382071018218994,
      "learning_rate": 9.844571757917944e-05,
      "loss": 0.7791659832000732,
      "step": 2180
    },
    {
      "epoch": 0.920675105485232,
      "grad_norm": 1.0855708122253418,
      "learning_rate": 9.84397708609736e-05,
      "loss": 0.7190433144569397,
      "step": 2182
    },
    {
      "epoch": 0.9215189873417722,
      "grad_norm": 1.103308916091919,
      "learning_rate": 9.843381296864291e-05,
      "loss": 0.6648658514022827,
      "step": 2184
    },
    {
      "epoch": 0.9223628691983122,
      "grad_norm": 1.073517918586731,
      "learning_rate": 9.842784390356178e-05,
      "loss": 0.6891760230064392,
      "step": 2186
    },
    {
      "epoch": 0.9232067510548523,
      "grad_norm": 1.0806199312210083,
      "learning_rate": 9.842186366710712e-05,
      "loss": 0.6880859136581421,
      "step": 2188
    },
    {
      "epoch": 0.9240506329113924,
      "grad_norm": 1.0631483793258667,
      "learning_rate": 9.841587226065848e-05,
      "loss": 0.6238307952880859,
      "step": 2190
    },
    {
      "epoch": 0.9248945147679325,
      "grad_norm": 1.2630863189697266,
      "learning_rate": 9.840986968559795e-05,
      "loss": 0.6905744075775146,
      "step": 2192
    },
    {
      "epoch": 0.9257383966244725,
      "grad_norm": 1.1307560205459595,
      "learning_rate": 9.840385594331022e-05,
      "loss": 0.7531564235687256,
      "step": 2194
    },
    {
      "epoch": 0.9265822784810127,
      "grad_norm": 1.0294862985610962,
      "learning_rate": 9.839783103518254e-05,
      "loss": 0.6750671863555908,
      "step": 2196
    },
    {
      "epoch": 0.9274261603375528,
      "grad_norm": 1.2446976900100708,
      "learning_rate": 9.839179496260472e-05,
      "loss": 0.7200804352760315,
      "step": 2198
    },
    {
      "epoch": 0.9282700421940928,
      "grad_norm": 1.2673420906066895,
      "learning_rate": 9.83857477269692e-05,
      "loss": 0.7002623677253723,
      "step": 2200
    },
    {
      "epoch": 0.9282700421940928,
      "eval_loss": 0.7497645616531372,
      "eval_runtime": 856.8766,
      "eval_samples_per_second": 2.459,
      "eval_steps_per_second": 2.459,
      "step": 2200
    },
    {
      "epoch": 0.9291139240506329,
      "grad_norm": 1.5114624500274658,
      "learning_rate": 9.837968932967094e-05,
      "loss": 0.7718265056610107,
      "step": 2202
    },
    {
      "epoch": 0.929957805907173,
      "grad_norm": 1.2059369087219238,
      "learning_rate": 9.837361977210751e-05,
      "loss": 0.7204271554946899,
      "step": 2204
    },
    {
      "epoch": 0.9308016877637131,
      "grad_norm": 1.2077301740646362,
      "learning_rate": 9.836753905567902e-05,
      "loss": 0.7371073961257935,
      "step": 2206
    },
    {
      "epoch": 0.9316455696202531,
      "grad_norm": 1.120097279548645,
      "learning_rate": 9.836144718178818e-05,
      "loss": 0.6601167321205139,
      "step": 2208
    },
    {
      "epoch": 0.9324894514767933,
      "grad_norm": 1.1755714416503906,
      "learning_rate": 9.835534415184029e-05,
      "loss": 0.6897423267364502,
      "step": 2210
    },
    {
      "epoch": 0.9333333333333333,
      "grad_norm": 1.3587000370025635,
      "learning_rate": 9.834922996724317e-05,
      "loss": 0.758438229560852,
      "step": 2212
    },
    {
      "epoch": 0.9341772151898734,
      "grad_norm": 1.1898177862167358,
      "learning_rate": 9.834310462940727e-05,
      "loss": 0.7489214539527893,
      "step": 2214
    },
    {
      "epoch": 0.9350210970464135,
      "grad_norm": 1.0814623832702637,
      "learning_rate": 9.833696813974558e-05,
      "loss": 0.6844488382339478,
      "step": 2216
    },
    {
      "epoch": 0.9358649789029536,
      "grad_norm": 1.1060179471969604,
      "learning_rate": 9.833082049967366e-05,
      "loss": 0.6617586016654968,
      "step": 2218
    },
    {
      "epoch": 0.9367088607594937,
      "grad_norm": 1.1780575513839722,
      "learning_rate": 9.832466171060968e-05,
      "loss": 0.7383584976196289,
      "step": 2220
    },
    {
      "epoch": 0.9375527426160337,
      "grad_norm": 1.3734618425369263,
      "learning_rate": 9.831849177397432e-05,
      "loss": 0.7764308452606201,
      "step": 2222
    },
    {
      "epoch": 0.9383966244725739,
      "grad_norm": 1.1367733478546143,
      "learning_rate": 9.831231069119089e-05,
      "loss": 0.6834397912025452,
      "step": 2224
    },
    {
      "epoch": 0.9392405063291139,
      "grad_norm": 1.1695492267608643,
      "learning_rate": 9.830611846368524e-05,
      "loss": 0.7054480910301208,
      "step": 2226
    },
    {
      "epoch": 0.940084388185654,
      "grad_norm": 1.0345736742019653,
      "learning_rate": 9.829991509288579e-05,
      "loss": 0.694448709487915,
      "step": 2228
    },
    {
      "epoch": 0.9409282700421941,
      "grad_norm": 1.298105239868164,
      "learning_rate": 9.829370058022356e-05,
      "loss": 0.6839741468429565,
      "step": 2230
    },
    {
      "epoch": 0.9417721518987342,
      "grad_norm": 1.2905502319335938,
      "learning_rate": 9.828747492713209e-05,
      "loss": 0.7886884212493896,
      "step": 2232
    },
    {
      "epoch": 0.9426160337552743,
      "grad_norm": 1.12301504611969,
      "learning_rate": 9.828123813504753e-05,
      "loss": 0.7206413149833679,
      "step": 2234
    },
    {
      "epoch": 0.9434599156118143,
      "grad_norm": 1.2644896507263184,
      "learning_rate": 9.82749902054086e-05,
      "loss": 0.7700693607330322,
      "step": 2236
    },
    {
      "epoch": 0.9443037974683545,
      "grad_norm": 1.1626365184783936,
      "learning_rate": 9.826873113965655e-05,
      "loss": 0.7199711203575134,
      "step": 2238
    },
    {
      "epoch": 0.9451476793248945,
      "grad_norm": 1.0728627443313599,
      "learning_rate": 9.826246093923528e-05,
      "loss": 0.7183539271354675,
      "step": 2240
    },
    {
      "epoch": 0.9459915611814346,
      "grad_norm": 1.1444766521453857,
      "learning_rate": 9.825617960559114e-05,
      "loss": 0.7417964935302734,
      "step": 2242
    },
    {
      "epoch": 0.9468354430379747,
      "grad_norm": 1.4059823751449585,
      "learning_rate": 9.824988714017316e-05,
      "loss": 0.7949740290641785,
      "step": 2244
    },
    {
      "epoch": 0.9476793248945148,
      "grad_norm": 1.1349766254425049,
      "learning_rate": 9.824358354443286e-05,
      "loss": 0.6433083415031433,
      "step": 2246
    },
    {
      "epoch": 0.9485232067510548,
      "grad_norm": 1.0879144668579102,
      "learning_rate": 9.823726881982438e-05,
      "loss": 0.6519861817359924,
      "step": 2248
    },
    {
      "epoch": 0.9493670886075949,
      "grad_norm": 1.2289162874221802,
      "learning_rate": 9.82309429678044e-05,
      "loss": 0.7280195355415344,
      "step": 2250
    },
    {
      "epoch": 0.950210970464135,
      "grad_norm": 1.1755765676498413,
      "learning_rate": 9.822460598983217e-05,
      "loss": 0.7524687647819519,
      "step": 2252
    },
    {
      "epoch": 0.9510548523206751,
      "grad_norm": 1.179807186126709,
      "learning_rate": 9.821825788736949e-05,
      "loss": 0.7543174624443054,
      "step": 2254
    },
    {
      "epoch": 0.9518987341772152,
      "grad_norm": 1.1234289407730103,
      "learning_rate": 9.821189866188079e-05,
      "loss": 0.716377854347229,
      "step": 2256
    },
    {
      "epoch": 0.9527426160337553,
      "grad_norm": 1.0324063301086426,
      "learning_rate": 9.820552831483297e-05,
      "loss": 0.6403332948684692,
      "step": 2258
    },
    {
      "epoch": 0.9535864978902954,
      "grad_norm": 1.1459579467773438,
      "learning_rate": 9.819914684769558e-05,
      "loss": 0.7406947612762451,
      "step": 2260
    },
    {
      "epoch": 0.9544303797468354,
      "grad_norm": 1.2886124849319458,
      "learning_rate": 9.819275426194072e-05,
      "loss": 0.749687671661377,
      "step": 2262
    },
    {
      "epoch": 0.9552742616033755,
      "grad_norm": 1.3349844217300415,
      "learning_rate": 9.818635055904299e-05,
      "loss": 0.778410017490387,
      "step": 2264
    },
    {
      "epoch": 0.9561181434599156,
      "grad_norm": 1.0994901657104492,
      "learning_rate": 9.81799357404796e-05,
      "loss": 0.6701914668083191,
      "step": 2266
    },
    {
      "epoch": 0.9569620253164557,
      "grad_norm": 1.1787796020507812,
      "learning_rate": 9.817350980773038e-05,
      "loss": 0.7205135226249695,
      "step": 2268
    },
    {
      "epoch": 0.9578059071729957,
      "grad_norm": 1.100813627243042,
      "learning_rate": 9.816707276227763e-05,
      "loss": 0.6897916197776794,
      "step": 2270
    },
    {
      "epoch": 0.9586497890295359,
      "grad_norm": 1.1280698776245117,
      "learning_rate": 9.816062460560627e-05,
      "loss": 0.6763570308685303,
      "step": 2272
    },
    {
      "epoch": 0.959493670886076,
      "grad_norm": 1.2322514057159424,
      "learning_rate": 9.815416533920374e-05,
      "loss": 0.6948683857917786,
      "step": 2274
    },
    {
      "epoch": 0.960337552742616,
      "grad_norm": 1.3963630199432373,
      "learning_rate": 9.814769496456008e-05,
      "loss": 0.7876828908920288,
      "step": 2276
    },
    {
      "epoch": 0.9611814345991562,
      "grad_norm": 1.2093676328659058,
      "learning_rate": 9.814121348316792e-05,
      "loss": 0.8191362619400024,
      "step": 2278
    },
    {
      "epoch": 0.9620253164556962,
      "grad_norm": 1.2223572731018066,
      "learning_rate": 9.813472089652233e-05,
      "loss": 0.7162626385688782,
      "step": 2280
    },
    {
      "epoch": 0.9628691983122363,
      "grad_norm": 1.1498078107833862,
      "learning_rate": 9.812821720612111e-05,
      "loss": 0.7183970212936401,
      "step": 2282
    },
    {
      "epoch": 0.9637130801687763,
      "grad_norm": 1.1563853025436401,
      "learning_rate": 9.812170241346449e-05,
      "loss": 0.734487771987915,
      "step": 2284
    },
    {
      "epoch": 0.9645569620253165,
      "grad_norm": 1.1823415756225586,
      "learning_rate": 9.81151765200553e-05,
      "loss": 0.7312371730804443,
      "step": 2286
    },
    {
      "epoch": 0.9654008438818565,
      "grad_norm": 1.1336151361465454,
      "learning_rate": 9.810863952739899e-05,
      "loss": 0.7668377757072449,
      "step": 2288
    },
    {
      "epoch": 0.9662447257383966,
      "grad_norm": 1.0857036113739014,
      "learning_rate": 9.810209143700347e-05,
      "loss": 0.7100399732589722,
      "step": 2290
    },
    {
      "epoch": 0.9670886075949368,
      "grad_norm": 1.1368129253387451,
      "learning_rate": 9.809553225037926e-05,
      "loss": 0.7169836163520813,
      "step": 2292
    },
    {
      "epoch": 0.9679324894514768,
      "grad_norm": 1.141107439994812,
      "learning_rate": 9.808896196903947e-05,
      "loss": 0.7709535956382751,
      "step": 2294
    },
    {
      "epoch": 0.9687763713080169,
      "grad_norm": 1.276405930519104,
      "learning_rate": 9.808238059449971e-05,
      "loss": 0.7300511002540588,
      "step": 2296
    },
    {
      "epoch": 0.9696202531645569,
      "grad_norm": 0.9817046523094177,
      "learning_rate": 9.80757881282782e-05,
      "loss": 0.6259129047393799,
      "step": 2298
    },
    {
      "epoch": 0.9704641350210971,
      "grad_norm": 1.3965257406234741,
      "learning_rate": 9.806918457189566e-05,
      "loss": 0.7361716032028198,
      "step": 2300
    },
    {
      "epoch": 0.9704641350210971,
      "eval_loss": 0.7464568614959717,
      "eval_runtime": 864.2128,
      "eval_samples_per_second": 2.438,
      "eval_steps_per_second": 2.438,
      "step": 2300
    },
    {
      "epoch": 0.9713080168776371,
      "grad_norm": 1.2168612480163574,
      "learning_rate": 9.806256992687544e-05,
      "loss": 0.805477499961853,
      "step": 2302
    },
    {
      "epoch": 0.9721518987341772,
      "grad_norm": 1.0418168306350708,
      "learning_rate": 9.80559441947434e-05,
      "loss": 0.6673368811607361,
      "step": 2304
    },
    {
      "epoch": 0.9729957805907173,
      "grad_norm": 1.223128318786621,
      "learning_rate": 9.804930737702796e-05,
      "loss": 0.7585647106170654,
      "step": 2306
    },
    {
      "epoch": 0.9738396624472574,
      "grad_norm": 1.264511227607727,
      "learning_rate": 9.804265947526011e-05,
      "loss": 0.7642034888267517,
      "step": 2308
    },
    {
      "epoch": 0.9746835443037974,
      "grad_norm": 1.076887607574463,
      "learning_rate": 9.803600049097339e-05,
      "loss": 0.7094541192054749,
      "step": 2310
    },
    {
      "epoch": 0.9755274261603376,
      "grad_norm": 1.0214987993240356,
      "learning_rate": 9.802933042570392e-05,
      "loss": 0.7370059490203857,
      "step": 2312
    },
    {
      "epoch": 0.9763713080168777,
      "grad_norm": 1.3075295686721802,
      "learning_rate": 9.802264928099035e-05,
      "loss": 0.726834237575531,
      "step": 2314
    },
    {
      "epoch": 0.9772151898734177,
      "grad_norm": 1.057386040687561,
      "learning_rate": 9.801595705837385e-05,
      "loss": 0.6742353439331055,
      "step": 2316
    },
    {
      "epoch": 0.9780590717299578,
      "grad_norm": 1.3998085260391235,
      "learning_rate": 9.800925375939825e-05,
      "loss": 0.6862425208091736,
      "step": 2318
    },
    {
      "epoch": 0.9789029535864979,
      "grad_norm": 1.080574631690979,
      "learning_rate": 9.800253938560983e-05,
      "loss": 0.6212031245231628,
      "step": 2320
    },
    {
      "epoch": 0.979746835443038,
      "grad_norm": 1.3643771409988403,
      "learning_rate": 9.799581393855748e-05,
      "loss": 0.7522522211074829,
      "step": 2322
    },
    {
      "epoch": 0.980590717299578,
      "grad_norm": 1.2455768585205078,
      "learning_rate": 9.798907741979264e-05,
      "loss": 0.7265716791152954,
      "step": 2324
    },
    {
      "epoch": 0.9814345991561182,
      "grad_norm": 1.078774333000183,
      "learning_rate": 9.798232983086927e-05,
      "loss": 0.7160419225692749,
      "step": 2326
    },
    {
      "epoch": 0.9822784810126582,
      "grad_norm": 1.3013948202133179,
      "learning_rate": 9.797557117334394e-05,
      "loss": 0.7991124391555786,
      "step": 2328
    },
    {
      "epoch": 0.9831223628691983,
      "grad_norm": 1.2216732501983643,
      "learning_rate": 9.796880144877572e-05,
      "loss": 0.7193916440010071,
      "step": 2330
    },
    {
      "epoch": 0.9839662447257383,
      "grad_norm": 1.1469542980194092,
      "learning_rate": 9.796202065872627e-05,
      "loss": 0.7184370756149292,
      "step": 2332
    },
    {
      "epoch": 0.9848101265822785,
      "grad_norm": 1.0431830883026123,
      "learning_rate": 9.795522880475979e-05,
      "loss": 0.6474619507789612,
      "step": 2334
    },
    {
      "epoch": 0.9856540084388186,
      "grad_norm": 1.1819576025009155,
      "learning_rate": 9.794842588844299e-05,
      "loss": 0.6392545700073242,
      "step": 2336
    },
    {
      "epoch": 0.9864978902953586,
      "grad_norm": 1.1984983682632446,
      "learning_rate": 9.794161191134525e-05,
      "loss": 0.7358114719390869,
      "step": 2338
    },
    {
      "epoch": 0.9873417721518988,
      "grad_norm": 1.3378512859344482,
      "learning_rate": 9.793478687503834e-05,
      "loss": 0.6762020587921143,
      "step": 2340
    },
    {
      "epoch": 0.9881856540084388,
      "grad_norm": 1.272674560546875,
      "learning_rate": 9.792795078109673e-05,
      "loss": 0.7478934526443481,
      "step": 2342
    },
    {
      "epoch": 0.9890295358649789,
      "grad_norm": 1.153746247291565,
      "learning_rate": 9.792110363109733e-05,
      "loss": 0.7316533923149109,
      "step": 2344
    },
    {
      "epoch": 0.9898734177215189,
      "grad_norm": 1.1361702680587769,
      "learning_rate": 9.791424542661967e-05,
      "loss": 0.7078539133071899,
      "step": 2346
    },
    {
      "epoch": 0.9907172995780591,
      "grad_norm": 1.3043115139007568,
      "learning_rate": 9.790737616924581e-05,
      "loss": 0.7945935130119324,
      "step": 2348
    },
    {
      "epoch": 0.9915611814345991,
      "grad_norm": 1.1913264989852905,
      "learning_rate": 9.790049586056034e-05,
      "loss": 0.8247197866439819,
      "step": 2350
    },
    {
      "epoch": 0.9924050632911392,
      "grad_norm": 1.1560171842575073,
      "learning_rate": 9.789360450215041e-05,
      "loss": 0.7099657654762268,
      "step": 2352
    },
    {
      "epoch": 0.9932489451476794,
      "grad_norm": 1.2311041355133057,
      "learning_rate": 9.788670209560575e-05,
      "loss": 0.7480318546295166,
      "step": 2354
    },
    {
      "epoch": 0.9940928270042194,
      "grad_norm": 1.1584707498550415,
      "learning_rate": 9.787978864251859e-05,
      "loss": 0.6870889067649841,
      "step": 2356
    },
    {
      "epoch": 0.9949367088607595,
      "grad_norm": 1.057478666305542,
      "learning_rate": 9.787286414448375e-05,
      "loss": 0.6114922165870667,
      "step": 2358
    },
    {
      "epoch": 0.9957805907172996,
      "grad_norm": 1.1431775093078613,
      "learning_rate": 9.786592860309856e-05,
      "loss": 0.6955118179321289,
      "step": 2360
    },
    {
      "epoch": 0.9966244725738397,
      "grad_norm": 1.232142448425293,
      "learning_rate": 9.785898201996292e-05,
      "loss": 0.735048770904541,
      "step": 2362
    },
    {
      "epoch": 0.9974683544303797,
      "grad_norm": 1.1236306428909302,
      "learning_rate": 9.785202439667928e-05,
      "loss": 0.7150241136550903,
      "step": 2364
    },
    {
      "epoch": 0.9983122362869198,
      "grad_norm": 1.0517534017562866,
      "learning_rate": 9.784505573485263e-05,
      "loss": 0.6870222687721252,
      "step": 2366
    },
    {
      "epoch": 0.99915611814346,
      "grad_norm": 1.1747480630874634,
      "learning_rate": 9.78380760360905e-05,
      "loss": 0.7521567940711975,
      "step": 2368
    },
    {
      "epoch": 1.0,
      "grad_norm": 1.2790346145629883,
      "learning_rate": 9.783108530200298e-05,
      "loss": 0.7336234450340271,
      "step": 2370
    },
    {
      "epoch": 1.0008438818565402,
      "grad_norm": 1.1216399669647217,
      "learning_rate": 9.78240835342027e-05,
      "loss": 0.6378109455108643,
      "step": 2372
    },
    {
      "epoch": 1.00168776371308,
      "grad_norm": 1.267336368560791,
      "learning_rate": 9.781707073430482e-05,
      "loss": 0.6174905300140381,
      "step": 2374
    },
    {
      "epoch": 1.0025316455696203,
      "grad_norm": 1.1342934370040894,
      "learning_rate": 9.781004690392706e-05,
      "loss": 0.6579123139381409,
      "step": 2376
    },
    {
      "epoch": 1.0033755274261604,
      "grad_norm": 1.1317468881607056,
      "learning_rate": 9.78030120446897e-05,
      "loss": 0.6679617166519165,
      "step": 2378
    },
    {
      "epoch": 1.0042194092827004,
      "grad_norm": 1.2992616891860962,
      "learning_rate": 9.779596615821552e-05,
      "loss": 0.7368149161338806,
      "step": 2380
    },
    {
      "epoch": 1.0050632911392405,
      "grad_norm": 1.1714510917663574,
      "learning_rate": 9.77889092461299e-05,
      "loss": 0.6887164115905762,
      "step": 2382
    },
    {
      "epoch": 1.0059071729957807,
      "grad_norm": 1.1670639514923096,
      "learning_rate": 9.778184131006071e-05,
      "loss": 0.681344211101532,
      "step": 2384
    },
    {
      "epoch": 1.0067510548523206,
      "grad_norm": 1.2487291097640991,
      "learning_rate": 9.77747623516384e-05,
      "loss": 0.7342769503593445,
      "step": 2386
    },
    {
      "epoch": 1.0075949367088608,
      "grad_norm": 1.2408956289291382,
      "learning_rate": 9.776767237249595e-05,
      "loss": 0.577454149723053,
      "step": 2388
    },
    {
      "epoch": 1.0084388185654007,
      "grad_norm": 1.067991852760315,
      "learning_rate": 9.776057137426889e-05,
      "loss": 0.6588307023048401,
      "step": 2390
    },
    {
      "epoch": 1.009282700421941,
      "grad_norm": 1.2821543216705322,
      "learning_rate": 9.775345935859525e-05,
      "loss": 0.7045041918754578,
      "step": 2392
    },
    {
      "epoch": 1.010126582278481,
      "grad_norm": 1.3160134553909302,
      "learning_rate": 9.774633632711569e-05,
      "loss": 0.7141479253768921,
      "step": 2394
    },
    {
      "epoch": 1.010970464135021,
      "grad_norm": 1.66774320602417,
      "learning_rate": 9.773920228147329e-05,
      "loss": 0.723293662071228,
      "step": 2396
    },
    {
      "epoch": 1.0118143459915612,
      "grad_norm": 1.027588963508606,
      "learning_rate": 9.77320572233138e-05,
      "loss": 0.5812023878097534,
      "step": 2398
    },
    {
      "epoch": 1.0126582278481013,
      "grad_norm": 1.406507968902588,
      "learning_rate": 9.77249011542854e-05,
      "loss": 0.7071458101272583,
      "step": 2400
    },
    {
      "epoch": 1.0126582278481013,
      "eval_loss": 0.7421699166297913,
      "eval_runtime": 854.2185,
      "eval_samples_per_second": 2.467,
      "eval_steps_per_second": 2.467,
      "step": 2400
    },
    {
      "epoch": 1.0135021097046413,
      "grad_norm": 1.1236240863800049,
      "learning_rate": 9.771773407603889e-05,
      "loss": 0.7049722671508789,
      "step": 2402
    },
    {
      "epoch": 1.0143459915611814,
      "grad_norm": 1.1924289464950562,
      "learning_rate": 9.771055599022756e-05,
      "loss": 0.635308027267456,
      "step": 2404
    },
    {
      "epoch": 1.0151898734177216,
      "grad_norm": 1.1744966506958008,
      "learning_rate": 9.770336689850727e-05,
      "loss": 0.7286487817764282,
      "step": 2406
    },
    {
      "epoch": 1.0160337552742615,
      "grad_norm": 1.2131173610687256,
      "learning_rate": 9.769616680253639e-05,
      "loss": 0.6828222274780273,
      "step": 2408
    },
    {
      "epoch": 1.0168776371308017,
      "grad_norm": 1.0517828464508057,
      "learning_rate": 9.768895570397585e-05,
      "loss": 0.6652156114578247,
      "step": 2410
    },
    {
      "epoch": 1.0177215189873419,
      "grad_norm": 1.1603758335113525,
      "learning_rate": 9.768173360448912e-05,
      "loss": 0.7278267741203308,
      "step": 2412
    },
    {
      "epoch": 1.0185654008438818,
      "grad_norm": 1.3167752027511597,
      "learning_rate": 9.767450050574218e-05,
      "loss": 0.6082334518432617,
      "step": 2414
    },
    {
      "epoch": 1.019409282700422,
      "grad_norm": 1.1754449605941772,
      "learning_rate": 9.766725640940358e-05,
      "loss": 0.67228102684021,
      "step": 2416
    },
    {
      "epoch": 1.0202531645569621,
      "grad_norm": 1.060952067375183,
      "learning_rate": 9.766000131714442e-05,
      "loss": 0.5984366536140442,
      "step": 2418
    },
    {
      "epoch": 1.021097046413502,
      "grad_norm": 1.0826152563095093,
      "learning_rate": 9.765273523063825e-05,
      "loss": 0.690661609172821,
      "step": 2420
    },
    {
      "epoch": 1.0219409282700422,
      "grad_norm": 1.423723816871643,
      "learning_rate": 9.764545815156125e-05,
      "loss": 0.7960668802261353,
      "step": 2422
    },
    {
      "epoch": 1.0227848101265822,
      "grad_norm": 1.0882549285888672,
      "learning_rate": 9.763817008159212e-05,
      "loss": 0.6971074342727661,
      "step": 2424
    },
    {
      "epoch": 1.0236286919831223,
      "grad_norm": 1.1053040027618408,
      "learning_rate": 9.763087102241206e-05,
      "loss": 0.6854458451271057,
      "step": 2426
    },
    {
      "epoch": 1.0244725738396625,
      "grad_norm": 1.1975224018096924,
      "learning_rate": 9.762356097570482e-05,
      "loss": 0.6724489331245422,
      "step": 2428
    },
    {
      "epoch": 1.0253164556962024,
      "grad_norm": 1.1692171096801758,
      "learning_rate": 9.76162399431567e-05,
      "loss": 0.7064506411552429,
      "step": 2430
    },
    {
      "epoch": 1.0261603375527426,
      "grad_norm": 1.1927787065505981,
      "learning_rate": 9.760890792645649e-05,
      "loss": 0.6605257391929626,
      "step": 2432
    },
    {
      "epoch": 1.0270042194092828,
      "grad_norm": 1.4147427082061768,
      "learning_rate": 9.760156492729558e-05,
      "loss": 0.6872501373291016,
      "step": 2434
    },
    {
      "epoch": 1.0278481012658227,
      "grad_norm": 1.2503126859664917,
      "learning_rate": 9.759421094736785e-05,
      "loss": 0.7117500305175781,
      "step": 2436
    },
    {
      "epoch": 1.0286919831223629,
      "grad_norm": 1.229978084564209,
      "learning_rate": 9.758684598836971e-05,
      "loss": 0.6740369200706482,
      "step": 2438
    },
    {
      "epoch": 1.029535864978903,
      "grad_norm": 1.4765945672988892,
      "learning_rate": 9.757947005200014e-05,
      "loss": 0.7215790748596191,
      "step": 2440
    },
    {
      "epoch": 1.030379746835443,
      "grad_norm": 1.282632827758789,
      "learning_rate": 9.757208313996061e-05,
      "loss": 0.6961746215820312,
      "step": 2442
    },
    {
      "epoch": 1.0312236286919831,
      "grad_norm": 1.259828805923462,
      "learning_rate": 9.756468525395512e-05,
      "loss": 0.6348349452018738,
      "step": 2444
    },
    {
      "epoch": 1.0320675105485233,
      "grad_norm": 1.0984172821044922,
      "learning_rate": 9.755727639569024e-05,
      "loss": 0.6756057739257812,
      "step": 2446
    },
    {
      "epoch": 1.0329113924050632,
      "grad_norm": 1.235835075378418,
      "learning_rate": 9.754985656687506e-05,
      "loss": 0.6968509554862976,
      "step": 2448
    },
    {
      "epoch": 1.0337552742616034,
      "grad_norm": 1.273032546043396,
      "learning_rate": 9.754242576922119e-05,
      "loss": 0.6793950796127319,
      "step": 2450
    },
    {
      "epoch": 1.0345991561181433,
      "grad_norm": 1.251996397972107,
      "learning_rate": 9.753498400444274e-05,
      "loss": 0.645270586013794,
      "step": 2452
    },
    {
      "epoch": 1.0354430379746835,
      "grad_norm": 1.4310805797576904,
      "learning_rate": 9.752753127425642e-05,
      "loss": 0.7291322350502014,
      "step": 2454
    },
    {
      "epoch": 1.0362869198312237,
      "grad_norm": 1.6582196950912476,
      "learning_rate": 9.752006758038142e-05,
      "loss": 0.7553019523620605,
      "step": 2456
    },
    {
      "epoch": 1.0371308016877636,
      "grad_norm": 1.081773042678833,
      "learning_rate": 9.751259292453947e-05,
      "loss": 0.5637331008911133,
      "step": 2458
    },
    {
      "epoch": 1.0379746835443038,
      "grad_norm": 1.1483876705169678,
      "learning_rate": 9.750510730845483e-05,
      "loss": 0.6012396216392517,
      "step": 2460
    },
    {
      "epoch": 1.038818565400844,
      "grad_norm": 1.0879185199737549,
      "learning_rate": 9.749761073385428e-05,
      "loss": 0.6795822381973267,
      "step": 2462
    },
    {
      "epoch": 1.0396624472573839,
      "grad_norm": 1.2378218173980713,
      "learning_rate": 9.749010320246714e-05,
      "loss": 0.6895145773887634,
      "step": 2464
    },
    {
      "epoch": 1.040506329113924,
      "grad_norm": 1.253233790397644,
      "learning_rate": 9.748258471602527e-05,
      "loss": 0.7124115228652954,
      "step": 2466
    },
    {
      "epoch": 1.0413502109704642,
      "grad_norm": 1.3994864225387573,
      "learning_rate": 9.747505527626302e-05,
      "loss": 0.7304861545562744,
      "step": 2468
    },
    {
      "epoch": 1.0421940928270041,
      "grad_norm": 1.2360669374465942,
      "learning_rate": 9.74675148849173e-05,
      "loss": 0.6845837831497192,
      "step": 2470
    },
    {
      "epoch": 1.0430379746835443,
      "grad_norm": 1.126849889755249,
      "learning_rate": 9.74599635437275e-05,
      "loss": 0.6780203580856323,
      "step": 2472
    },
    {
      "epoch": 1.0438818565400845,
      "grad_norm": 1.169788122177124,
      "learning_rate": 9.745240125443562e-05,
      "loss": 0.7550003528594971,
      "step": 2474
    },
    {
      "epoch": 1.0447257383966244,
      "grad_norm": 1.1311867237091064,
      "learning_rate": 9.744482801878612e-05,
      "loss": 0.6910399198532104,
      "step": 2476
    },
    {
      "epoch": 1.0455696202531646,
      "grad_norm": 1.1267731189727783,
      "learning_rate": 9.743724383852597e-05,
      "loss": 0.7164814472198486,
      "step": 2478
    },
    {
      "epoch": 1.0464135021097047,
      "grad_norm": 1.2239704132080078,
      "learning_rate": 9.742964871540472e-05,
      "loss": 0.6428439617156982,
      "step": 2480
    },
    {
      "epoch": 1.0472573839662447,
      "grad_norm": 1.1854743957519531,
      "learning_rate": 9.742204265117443e-05,
      "loss": 0.6994290351867676,
      "step": 2482
    },
    {
      "epoch": 1.0481012658227848,
      "grad_norm": 1.0695894956588745,
      "learning_rate": 9.741442564758964e-05,
      "loss": 0.6725777983665466,
      "step": 2484
    },
    {
      "epoch": 1.048945147679325,
      "grad_norm": 1.1799863576889038,
      "learning_rate": 9.740679770640748e-05,
      "loss": 0.6538674235343933,
      "step": 2486
    },
    {
      "epoch": 1.049789029535865,
      "grad_norm": 1.295546293258667,
      "learning_rate": 9.739915882938754e-05,
      "loss": 0.780756950378418,
      "step": 2488
    },
    {
      "epoch": 1.0506329113924051,
      "grad_norm": 1.2371755838394165,
      "learning_rate": 9.739150901829198e-05,
      "loss": 0.6657930612564087,
      "step": 2490
    },
    {
      "epoch": 1.051476793248945,
      "grad_norm": 1.103037714958191,
      "learning_rate": 9.738384827488547e-05,
      "loss": 0.6675208210945129,
      "step": 2492
    },
    {
      "epoch": 1.0523206751054852,
      "grad_norm": 1.1835435628890991,
      "learning_rate": 9.737617660093517e-05,
      "loss": 0.6693358421325684,
      "step": 2494
    },
    {
      "epoch": 1.0531645569620254,
      "grad_norm": 1.003771424293518,
      "learning_rate": 9.736849399821082e-05,
      "loss": 0.624502956867218,
      "step": 2496
    },
    {
      "epoch": 1.0540084388185653,
      "grad_norm": 1.1391769647598267,
      "learning_rate": 9.736080046848463e-05,
      "loss": 0.6350868344306946,
      "step": 2498
    },
    {
      "epoch": 1.0548523206751055,
      "grad_norm": 1.376518726348877,
      "learning_rate": 9.735309601353134e-05,
      "loss": 0.6721012592315674,
      "step": 2500
    },
    {
      "epoch": 1.0548523206751055,
      "eval_loss": 0.741338849067688,
      "eval_runtime": 847.7478,
      "eval_samples_per_second": 2.485,
      "eval_steps_per_second": 2.485,
      "step": 2500
    },
    {
      "epoch": 1.0556962025316456,
      "grad_norm": 1.194190502166748,
      "learning_rate": 9.734538063512824e-05,
      "loss": 0.6888233423233032,
      "step": 2502
    },
    {
      "epoch": 1.0565400843881856,
      "grad_norm": 1.378830909729004,
      "learning_rate": 9.733765433505513e-05,
      "loss": 0.7095553278923035,
      "step": 2504
    },
    {
      "epoch": 1.0573839662447257,
      "grad_norm": 1.1289541721343994,
      "learning_rate": 9.732991711509428e-05,
      "loss": 0.6734166145324707,
      "step": 2506
    },
    {
      "epoch": 1.058227848101266,
      "grad_norm": 1.1858116388320923,
      "learning_rate": 9.732216897703054e-05,
      "loss": 0.7006195187568665,
      "step": 2508
    },
    {
      "epoch": 1.0590717299578059,
      "grad_norm": 1.1365686655044556,
      "learning_rate": 9.731440992265127e-05,
      "loss": 0.6481205821037292,
      "step": 2510
    },
    {
      "epoch": 1.059915611814346,
      "grad_norm": 1.2886228561401367,
      "learning_rate": 9.730663995374632e-05,
      "loss": 0.679282546043396,
      "step": 2512
    },
    {
      "epoch": 1.0607594936708862,
      "grad_norm": 1.355322003364563,
      "learning_rate": 9.729885907210808e-05,
      "loss": 0.7656359672546387,
      "step": 2514
    },
    {
      "epoch": 1.0616033755274261,
      "grad_norm": 1.1552364826202393,
      "learning_rate": 9.729106727953142e-05,
      "loss": 0.5996183156967163,
      "step": 2516
    },
    {
      "epoch": 1.0624472573839663,
      "grad_norm": 1.1419235467910767,
      "learning_rate": 9.728326457781381e-05,
      "loss": 0.7599716782569885,
      "step": 2518
    },
    {
      "epoch": 1.0632911392405062,
      "grad_norm": 1.2240079641342163,
      "learning_rate": 9.727545096875512e-05,
      "loss": 0.7150241732597351,
      "step": 2520
    },
    {
      "epoch": 1.0641350210970464,
      "grad_norm": 1.2463440895080566,
      "learning_rate": 9.726762645415785e-05,
      "loss": 0.734352171421051,
      "step": 2522
    },
    {
      "epoch": 1.0649789029535865,
      "grad_norm": 1.1680364608764648,
      "learning_rate": 9.725979103582697e-05,
      "loss": 0.6950796842575073,
      "step": 2524
    },
    {
      "epoch": 1.0658227848101265,
      "grad_norm": 1.1680421829223633,
      "learning_rate": 9.725194471556991e-05,
      "loss": 0.7096341252326965,
      "step": 2526
    },
    {
      "epoch": 1.0666666666666667,
      "grad_norm": 1.043717861175537,
      "learning_rate": 9.724408749519671e-05,
      "loss": 0.6486304402351379,
      "step": 2528
    },
    {
      "epoch": 1.0675105485232068,
      "grad_norm": 1.1240284442901611,
      "learning_rate": 9.723621937651985e-05,
      "loss": 0.6519505381584167,
      "step": 2530
    },
    {
      "epoch": 1.0683544303797468,
      "grad_norm": 1.185223937034607,
      "learning_rate": 9.722834036135439e-05,
      "loss": 0.6724293231964111,
      "step": 2532
    },
    {
      "epoch": 1.069198312236287,
      "grad_norm": 1.3234196901321411,
      "learning_rate": 9.722045045151784e-05,
      "loss": 0.6886576414108276,
      "step": 2534
    },
    {
      "epoch": 1.070042194092827,
      "grad_norm": 1.333084225654602,
      "learning_rate": 9.721254964883024e-05,
      "loss": 0.688493549823761,
      "step": 2536
    },
    {
      "epoch": 1.070886075949367,
      "grad_norm": 1.2435462474822998,
      "learning_rate": 9.720463795511419e-05,
      "loss": 0.6527412533760071,
      "step": 2538
    },
    {
      "epoch": 1.0717299578059072,
      "grad_norm": 1.1521880626678467,
      "learning_rate": 9.719671537219472e-05,
      "loss": 0.6508163809776306,
      "step": 2540
    },
    {
      "epoch": 1.0725738396624473,
      "grad_norm": 1.015013575553894,
      "learning_rate": 9.718878190189947e-05,
      "loss": 0.6954023838043213,
      "step": 2542
    },
    {
      "epoch": 1.0734177215189873,
      "grad_norm": 1.1507678031921387,
      "learning_rate": 9.718083754605851e-05,
      "loss": 0.7201322913169861,
      "step": 2544
    },
    {
      "epoch": 1.0742616033755275,
      "grad_norm": 1.0569016933441162,
      "learning_rate": 9.717288230650444e-05,
      "loss": 0.6688649654388428,
      "step": 2546
    },
    {
      "epoch": 1.0751054852320676,
      "grad_norm": 1.2178492546081543,
      "learning_rate": 9.716491618507241e-05,
      "loss": 0.7077898979187012,
      "step": 2548
    },
    {
      "epoch": 1.0759493670886076,
      "grad_norm": 1.3587230443954468,
      "learning_rate": 9.715693918360002e-05,
      "loss": 0.7312119603157043,
      "step": 2550
    },
    {
      "epoch": 1.0767932489451477,
      "grad_norm": 1.1930122375488281,
      "learning_rate": 9.714895130392744e-05,
      "loss": 0.6910589337348938,
      "step": 2552
    },
    {
      "epoch": 1.0776371308016879,
      "grad_norm": 1.2440707683563232,
      "learning_rate": 9.71409525478973e-05,
      "loss": 0.7942836284637451,
      "step": 2554
    },
    {
      "epoch": 1.0784810126582278,
      "grad_norm": 1.3755065202713013,
      "learning_rate": 9.713294291735477e-05,
      "loss": 0.6652286052703857,
      "step": 2556
    },
    {
      "epoch": 1.079324894514768,
      "grad_norm": 1.165448784828186,
      "learning_rate": 9.71249224141475e-05,
      "loss": 0.6025735139846802,
      "step": 2558
    },
    {
      "epoch": 1.080168776371308,
      "grad_norm": 1.2981204986572266,
      "learning_rate": 9.711689104012569e-05,
      "loss": 0.7343734502792358,
      "step": 2560
    },
    {
      "epoch": 1.081012658227848,
      "grad_norm": 1.2040622234344482,
      "learning_rate": 9.710884879714202e-05,
      "loss": 0.6903306841850281,
      "step": 2562
    },
    {
      "epoch": 1.0818565400843883,
      "grad_norm": 1.1835904121398926,
      "learning_rate": 9.710079568705168e-05,
      "loss": 0.69134920835495,
      "step": 2564
    },
    {
      "epoch": 1.0827004219409282,
      "grad_norm": 1.3345229625701904,
      "learning_rate": 9.709273171171235e-05,
      "loss": 0.6471185088157654,
      "step": 2566
    },
    {
      "epoch": 1.0835443037974684,
      "grad_norm": 1.0884469747543335,
      "learning_rate": 9.708465687298425e-05,
      "loss": 0.6302382349967957,
      "step": 2568
    },
    {
      "epoch": 1.0843881856540085,
      "grad_norm": 1.1994211673736572,
      "learning_rate": 9.707657117273007e-05,
      "loss": 0.7329678535461426,
      "step": 2570
    },
    {
      "epoch": 1.0852320675105485,
      "grad_norm": 1.2609503269195557,
      "learning_rate": 9.706847461281507e-05,
      "loss": 0.719862163066864,
      "step": 2572
    },
    {
      "epoch": 1.0860759493670886,
      "grad_norm": 1.2686879634857178,
      "learning_rate": 9.706036719510694e-05,
      "loss": 0.7142901420593262,
      "step": 2574
    },
    {
      "epoch": 1.0869198312236288,
      "grad_norm": 1.2763310670852661,
      "learning_rate": 9.705224892147591e-05,
      "loss": 0.7009075284004211,
      "step": 2576
    },
    {
      "epoch": 1.0877637130801687,
      "grad_norm": 1.1704022884368896,
      "learning_rate": 9.70441197937947e-05,
      "loss": 0.6873779296875,
      "step": 2578
    },
    {
      "epoch": 1.0886075949367089,
      "grad_norm": 1.0482875108718872,
      "learning_rate": 9.703597981393856e-05,
      "loss": 0.6437726020812988,
      "step": 2580
    },
    {
      "epoch": 1.0894514767932488,
      "grad_norm": 1.28431236743927,
      "learning_rate": 9.702782898378521e-05,
      "loss": 0.6933431625366211,
      "step": 2582
    },
    {
      "epoch": 1.090295358649789,
      "grad_norm": 1.0962283611297607,
      "learning_rate": 9.701966730521491e-05,
      "loss": 0.6488757133483887,
      "step": 2584
    },
    {
      "epoch": 1.0911392405063292,
      "grad_norm": 1.2177873849868774,
      "learning_rate": 9.70114947801104e-05,
      "loss": 0.6385396122932434,
      "step": 2586
    },
    {
      "epoch": 1.091983122362869,
      "grad_norm": 1.197059988975525,
      "learning_rate": 9.70033114103569e-05,
      "loss": 0.6826614737510681,
      "step": 2588
    },
    {
      "epoch": 1.0928270042194093,
      "grad_norm": 1.1624075174331665,
      "learning_rate": 9.699511719784217e-05,
      "loss": 0.605629563331604,
      "step": 2590
    },
    {
      "epoch": 1.0936708860759494,
      "grad_norm": 1.2975167036056519,
      "learning_rate": 9.698691214445648e-05,
      "loss": 0.734926700592041,
      "step": 2592
    },
    {
      "epoch": 1.0945147679324894,
      "grad_norm": 1.215414047241211,
      "learning_rate": 9.697869625209255e-05,
      "loss": 0.7281333804130554,
      "step": 2594
    },
    {
      "epoch": 1.0953586497890295,
      "grad_norm": 1.1862860918045044,
      "learning_rate": 9.697046952264563e-05,
      "loss": 0.7388250827789307,
      "step": 2596
    },
    {
      "epoch": 1.0962025316455697,
      "grad_norm": 1.1127797365188599,
      "learning_rate": 9.696223195801348e-05,
      "loss": 0.6495320796966553,
      "step": 2598
    },
    {
      "epoch": 1.0970464135021096,
      "grad_norm": 1.0863338708877563,
      "learning_rate": 9.695398356009636e-05,
      "loss": 0.7157143950462341,
      "step": 2600
    },
    {
      "epoch": 1.0970464135021096,
      "eval_loss": 0.7377332448959351,
      "eval_runtime": 859.6612,
      "eval_samples_per_second": 2.451,
      "eval_steps_per_second": 2.451,
      "step": 2600
    },
    {
      "epoch": 1.0978902953586498,
      "grad_norm": 1.1228652000427246,
      "learning_rate": 9.694572433079699e-05,
      "loss": 0.6597335934638977,
      "step": 2602
    },
    {
      "epoch": 1.09873417721519,
      "grad_norm": 1.3077653646469116,
      "learning_rate": 9.69374542720206e-05,
      "loss": 0.6715680360794067,
      "step": 2604
    },
    {
      "epoch": 1.09957805907173,
      "grad_norm": 1.241603970527649,
      "learning_rate": 9.692917338567499e-05,
      "loss": 0.6910243034362793,
      "step": 2606
    },
    {
      "epoch": 1.10042194092827,
      "grad_norm": 1.1372551918029785,
      "learning_rate": 9.692088167367037e-05,
      "loss": 0.6519553065299988,
      "step": 2608
    },
    {
      "epoch": 1.1012658227848102,
      "grad_norm": 1.2894765138626099,
      "learning_rate": 9.691257913791949e-05,
      "loss": 0.6542758941650391,
      "step": 2610
    },
    {
      "epoch": 1.1021097046413502,
      "grad_norm": 1.0800915956497192,
      "learning_rate": 9.690426578033755e-05,
      "loss": 0.6886795163154602,
      "step": 2612
    },
    {
      "epoch": 1.1029535864978903,
      "grad_norm": 1.3394384384155273,
      "learning_rate": 9.689594160284233e-05,
      "loss": 0.7512150406837463,
      "step": 2614
    },
    {
      "epoch": 1.1037974683544305,
      "grad_norm": 1.2175323963165283,
      "learning_rate": 9.688760660735402e-05,
      "loss": 0.67207932472229,
      "step": 2616
    },
    {
      "epoch": 1.1046413502109704,
      "grad_norm": 1.2181185483932495,
      "learning_rate": 9.687926079579537e-05,
      "loss": 0.6591740846633911,
      "step": 2618
    },
    {
      "epoch": 1.1054852320675106,
      "grad_norm": 1.1740983724594116,
      "learning_rate": 9.68709041700916e-05,
      "loss": 0.6431041359901428,
      "step": 2620
    },
    {
      "epoch": 1.1063291139240505,
      "grad_norm": 1.1792434453964233,
      "learning_rate": 9.686253673217038e-05,
      "loss": 0.6573615074157715,
      "step": 2622
    },
    {
      "epoch": 1.1071729957805907,
      "grad_norm": 1.058391809463501,
      "learning_rate": 9.685415848396196e-05,
      "loss": 0.5576209425926208,
      "step": 2624
    },
    {
      "epoch": 1.1080168776371309,
      "grad_norm": 1.3203206062316895,
      "learning_rate": 9.684576942739903e-05,
      "loss": 0.668684184551239,
      "step": 2626
    },
    {
      "epoch": 1.1088607594936708,
      "grad_norm": 1.2391762733459473,
      "learning_rate": 9.68373695644168e-05,
      "loss": 0.6800089478492737,
      "step": 2628
    },
    {
      "epoch": 1.109704641350211,
      "grad_norm": 1.2323405742645264,
      "learning_rate": 9.682895889695292e-05,
      "loss": 0.6433757543563843,
      "step": 2630
    },
    {
      "epoch": 1.1105485232067511,
      "grad_norm": 1.2656551599502563,
      "learning_rate": 9.682053742694759e-05,
      "loss": 0.6628785729408264,
      "step": 2632
    },
    {
      "epoch": 1.111392405063291,
      "grad_norm": 1.2984392642974854,
      "learning_rate": 9.681210515634349e-05,
      "loss": 0.6838971972465515,
      "step": 2634
    },
    {
      "epoch": 1.1122362869198312,
      "grad_norm": 1.3200393915176392,
      "learning_rate": 9.680366208708576e-05,
      "loss": 0.7548647522926331,
      "step": 2636
    },
    {
      "epoch": 1.1130801687763714,
      "grad_norm": 1.225388526916504,
      "learning_rate": 9.679520822112208e-05,
      "loss": 0.6553335189819336,
      "step": 2638
    },
    {
      "epoch": 1.1139240506329113,
      "grad_norm": 1.2350653409957886,
      "learning_rate": 9.678674356040259e-05,
      "loss": 0.631401538848877,
      "step": 2640
    },
    {
      "epoch": 1.1147679324894515,
      "grad_norm": 1.2325507402420044,
      "learning_rate": 9.677826810687989e-05,
      "loss": 0.6459156274795532,
      "step": 2642
    },
    {
      "epoch": 1.1156118143459917,
      "grad_norm": 1.0008996725082397,
      "learning_rate": 9.676978186250915e-05,
      "loss": 0.6425284743309021,
      "step": 2644
    },
    {
      "epoch": 1.1164556962025316,
      "grad_norm": 1.3767247200012207,
      "learning_rate": 9.676128482924796e-05,
      "loss": 0.6451422572135925,
      "step": 2646
    },
    {
      "epoch": 1.1172995780590718,
      "grad_norm": 1.2070895433425903,
      "learning_rate": 9.675277700905643e-05,
      "loss": 0.6713272929191589,
      "step": 2648
    },
    {
      "epoch": 1.1181434599156117,
      "grad_norm": 1.1582069396972656,
      "learning_rate": 9.674425840389716e-05,
      "loss": 0.6285044550895691,
      "step": 2650
    },
    {
      "epoch": 1.1189873417721519,
      "grad_norm": 1.1641311645507812,
      "learning_rate": 9.67357290157352e-05,
      "loss": 0.624229907989502,
      "step": 2652
    },
    {
      "epoch": 1.119831223628692,
      "grad_norm": 1.3071147203445435,
      "learning_rate": 9.672718884653814e-05,
      "loss": 0.7214919328689575,
      "step": 2654
    },
    {
      "epoch": 1.120675105485232,
      "grad_norm": 1.2157800197601318,
      "learning_rate": 9.671863789827602e-05,
      "loss": 0.8062215447425842,
      "step": 2656
    },
    {
      "epoch": 1.1215189873417721,
      "grad_norm": 1.2843927145004272,
      "learning_rate": 9.671007617292138e-05,
      "loss": 0.6362426280975342,
      "step": 2658
    },
    {
      "epoch": 1.1223628691983123,
      "grad_norm": 1.1182712316513062,
      "learning_rate": 9.670150367244927e-05,
      "loss": 0.6181318163871765,
      "step": 2660
    },
    {
      "epoch": 1.1232067510548522,
      "grad_norm": 1.566605806350708,
      "learning_rate": 9.669292039883717e-05,
      "loss": 0.6973897218704224,
      "step": 2662
    },
    {
      "epoch": 1.1240506329113924,
      "grad_norm": 1.0726850032806396,
      "learning_rate": 9.66843263540651e-05,
      "loss": 0.6117324829101562,
      "step": 2664
    },
    {
      "epoch": 1.1248945147679326,
      "grad_norm": 1.2953020334243774,
      "learning_rate": 9.66757215401155e-05,
      "loss": 0.642676830291748,
      "step": 2666
    },
    {
      "epoch": 1.1257383966244725,
      "grad_norm": 1.1184383630752563,
      "learning_rate": 9.66671059589734e-05,
      "loss": 0.6757452487945557,
      "step": 2668
    },
    {
      "epoch": 1.1265822784810127,
      "grad_norm": 1.2732970714569092,
      "learning_rate": 9.66584796126262e-05,
      "loss": 0.6861951947212219,
      "step": 2670
    },
    {
      "epoch": 1.1274261603375528,
      "grad_norm": 1.2713000774383545,
      "learning_rate": 9.664984250306383e-05,
      "loss": 0.6727077960968018,
      "step": 2672
    },
    {
      "epoch": 1.1282700421940928,
      "grad_norm": 1.269827961921692,
      "learning_rate": 9.664119463227874e-05,
      "loss": 0.7355974912643433,
      "step": 2674
    },
    {
      "epoch": 1.129113924050633,
      "grad_norm": 1.3067172765731812,
      "learning_rate": 9.663253600226581e-05,
      "loss": 0.7121313214302063,
      "step": 2676
    },
    {
      "epoch": 1.129957805907173,
      "grad_norm": 1.2958797216415405,
      "learning_rate": 9.662386661502242e-05,
      "loss": 0.6671369075775146,
      "step": 2678
    },
    {
      "epoch": 1.130801687763713,
      "grad_norm": 1.2943401336669922,
      "learning_rate": 9.661518647254842e-05,
      "loss": 0.6153768301010132,
      "step": 2680
    },
    {
      "epoch": 1.1316455696202532,
      "grad_norm": 1.1744167804718018,
      "learning_rate": 9.660649557684616e-05,
      "loss": 0.6070778965950012,
      "step": 2682
    },
    {
      "epoch": 1.1324894514767934,
      "grad_norm": 1.159209132194519,
      "learning_rate": 9.659779392992047e-05,
      "loss": 0.676887035369873,
      "step": 2684
    },
    {
      "epoch": 1.1333333333333333,
      "grad_norm": 1.1937510967254639,
      "learning_rate": 9.658908153377866e-05,
      "loss": 0.6086745262145996,
      "step": 2686
    },
    {
      "epoch": 1.1341772151898735,
      "grad_norm": 1.1461687088012695,
      "learning_rate": 9.658035839043049e-05,
      "loss": 0.6493708491325378,
      "step": 2688
    },
    {
      "epoch": 1.1350210970464134,
      "grad_norm": 2.066361665725708,
      "learning_rate": 9.657162450188824e-05,
      "loss": 0.6813004016876221,
      "step": 2690
    },
    {
      "epoch": 1.1358649789029536,
      "grad_norm": 1.086910367012024,
      "learning_rate": 9.656287987016664e-05,
      "loss": 0.721062183380127,
      "step": 2692
    },
    {
      "epoch": 1.1367088607594937,
      "grad_norm": 1.1869292259216309,
      "learning_rate": 9.65541244972829e-05,
      "loss": 0.5975021123886108,
      "step": 2694
    },
    {
      "epoch": 1.1375527426160337,
      "grad_norm": 1.2456518411636353,
      "learning_rate": 9.654535838525674e-05,
      "loss": 0.6818324327468872,
      "step": 2696
    },
    {
      "epoch": 1.1383966244725738,
      "grad_norm": 1.5271464586257935,
      "learning_rate": 9.653658153611031e-05,
      "loss": 0.6844469308853149,
      "step": 2698
    },
    {
      "epoch": 1.139240506329114,
      "grad_norm": 1.1403794288635254,
      "learning_rate": 9.652779395186827e-05,
      "loss": 0.6388684511184692,
      "step": 2700
    },
    {
      "epoch": 1.139240506329114,
      "eval_loss": 0.7335711717605591,
      "eval_runtime": 861.9651,
      "eval_samples_per_second": 2.444,
      "eval_steps_per_second": 2.444,
      "step": 2700
    },
    {
      "epoch": 1.140084388185654,
      "grad_norm": 1.1091634035110474,
      "learning_rate": 9.651899563455775e-05,
      "loss": 0.6154619455337524,
      "step": 2702
    },
    {
      "epoch": 1.140928270042194,
      "grad_norm": 1.3280601501464844,
      "learning_rate": 9.651018658620837e-05,
      "loss": 0.629319429397583,
      "step": 2704
    },
    {
      "epoch": 1.1417721518987343,
      "grad_norm": 1.226806402206421,
      "learning_rate": 9.650136680885216e-05,
      "loss": 0.6088175773620605,
      "step": 2706
    },
    {
      "epoch": 1.1426160337552742,
      "grad_norm": 1.0593408346176147,
      "learning_rate": 9.649253630452372e-05,
      "loss": 0.6199659705162048,
      "step": 2708
    },
    {
      "epoch": 1.1434599156118144,
      "grad_norm": 1.1112475395202637,
      "learning_rate": 9.648369507526008e-05,
      "loss": 0.7233364582061768,
      "step": 2710
    },
    {
      "epoch": 1.1443037974683543,
      "grad_norm": 1.1737885475158691,
      "learning_rate": 9.647484312310068e-05,
      "loss": 0.6687955856323242,
      "step": 2712
    },
    {
      "epoch": 1.1451476793248945,
      "grad_norm": 1.194532036781311,
      "learning_rate": 9.646598045008756e-05,
      "loss": 0.6508969068527222,
      "step": 2714
    },
    {
      "epoch": 1.1459915611814346,
      "grad_norm": 1.069395899772644,
      "learning_rate": 9.645710705826517e-05,
      "loss": 0.6408317685127258,
      "step": 2716
    },
    {
      "epoch": 1.1468354430379746,
      "grad_norm": 1.2429133653640747,
      "learning_rate": 9.644822294968037e-05,
      "loss": 0.650763750076294,
      "step": 2718
    },
    {
      "epoch": 1.1476793248945147,
      "grad_norm": 1.2950133085250854,
      "learning_rate": 9.64393281263826e-05,
      "loss": 0.6952191591262817,
      "step": 2720
    },
    {
      "epoch": 1.148523206751055,
      "grad_norm": 1.1972628831863403,
      "learning_rate": 9.643042259042372e-05,
      "loss": 0.6772956252098083,
      "step": 2722
    },
    {
      "epoch": 1.1493670886075948,
      "grad_norm": 1.1670407056808472,
      "learning_rate": 9.642150634385805e-05,
      "loss": 0.6734447479248047,
      "step": 2724
    },
    {
      "epoch": 1.150210970464135,
      "grad_norm": 1.120302677154541,
      "learning_rate": 9.641257938874243e-05,
      "loss": 0.6387717127799988,
      "step": 2726
    },
    {
      "epoch": 1.1510548523206752,
      "grad_norm": 1.1241344213485718,
      "learning_rate": 9.640364172713609e-05,
      "loss": 0.6592874526977539,
      "step": 2728
    },
    {
      "epoch": 1.1518987341772151,
      "grad_norm": 1.2627261877059937,
      "learning_rate": 9.639469336110083e-05,
      "loss": 0.7257466912269592,
      "step": 2730
    },
    {
      "epoch": 1.1527426160337553,
      "grad_norm": 1.0528618097305298,
      "learning_rate": 9.638573429270083e-05,
      "loss": 0.572188138961792,
      "step": 2732
    },
    {
      "epoch": 1.1535864978902954,
      "grad_norm": 1.212536334991455,
      "learning_rate": 9.637676452400277e-05,
      "loss": 0.678981602191925,
      "step": 2734
    },
    {
      "epoch": 1.1544303797468354,
      "grad_norm": 1.152167797088623,
      "learning_rate": 9.636778405707582e-05,
      "loss": 0.6375001072883606,
      "step": 2736
    },
    {
      "epoch": 1.1552742616033755,
      "grad_norm": 1.2400429248809814,
      "learning_rate": 9.635879289399161e-05,
      "loss": 0.7602289319038391,
      "step": 2738
    },
    {
      "epoch": 1.1561181434599157,
      "grad_norm": 1.3488622903823853,
      "learning_rate": 9.634979103682421e-05,
      "loss": 0.6209543943405151,
      "step": 2740
    },
    {
      "epoch": 1.1569620253164556,
      "grad_norm": 1.1999555826187134,
      "learning_rate": 9.634077848765019e-05,
      "loss": 0.6215830445289612,
      "step": 2742
    },
    {
      "epoch": 1.1578059071729958,
      "grad_norm": 1.2008578777313232,
      "learning_rate": 9.633175524854855e-05,
      "loss": 0.6634654998779297,
      "step": 2744
    },
    {
      "epoch": 1.158649789029536,
      "grad_norm": 1.3920676708221436,
      "learning_rate": 9.63227213216008e-05,
      "loss": 0.7515161633491516,
      "step": 2746
    },
    {
      "epoch": 1.159493670886076,
      "grad_norm": 1.0551656484603882,
      "learning_rate": 9.631367670889089e-05,
      "loss": 0.724361777305603,
      "step": 2748
    },
    {
      "epoch": 1.160337552742616,
      "grad_norm": 1.2820028066635132,
      "learning_rate": 9.630462141250523e-05,
      "loss": 0.6673553586006165,
      "step": 2750
    },
    {
      "epoch": 1.1611814345991562,
      "grad_norm": 1.1452983617782593,
      "learning_rate": 9.62955554345327e-05,
      "loss": 0.7029784917831421,
      "step": 2752
    },
    {
      "epoch": 1.1620253164556962,
      "grad_norm": 1.1808624267578125,
      "learning_rate": 9.628647877706466e-05,
      "loss": 0.7355457544326782,
      "step": 2754
    },
    {
      "epoch": 1.1628691983122363,
      "grad_norm": 1.0574703216552734,
      "learning_rate": 9.627739144219492e-05,
      "loss": 0.6144933700561523,
      "step": 2756
    },
    {
      "epoch": 1.1637130801687763,
      "grad_norm": 1.215733528137207,
      "learning_rate": 9.626829343201974e-05,
      "loss": 0.6843759417533875,
      "step": 2758
    },
    {
      "epoch": 1.1645569620253164,
      "grad_norm": 1.1667706966400146,
      "learning_rate": 9.625918474863787e-05,
      "loss": 0.6197049617767334,
      "step": 2760
    },
    {
      "epoch": 1.1654008438818566,
      "grad_norm": 1.3765631914138794,
      "learning_rate": 9.62500653941505e-05,
      "loss": 0.715958297252655,
      "step": 2762
    },
    {
      "epoch": 1.1662447257383965,
      "grad_norm": 1.173715591430664,
      "learning_rate": 9.62409353706613e-05,
      "loss": 0.7433139085769653,
      "step": 2764
    },
    {
      "epoch": 1.1670886075949367,
      "grad_norm": 1.1837430000305176,
      "learning_rate": 9.623179468027637e-05,
      "loss": 0.7174371480941772,
      "step": 2766
    },
    {
      "epoch": 1.1679324894514769,
      "grad_norm": 1.1577154397964478,
      "learning_rate": 9.622264332510432e-05,
      "loss": 0.7184823751449585,
      "step": 2768
    },
    {
      "epoch": 1.1687763713080168,
      "grad_norm": 1.165246605873108,
      "learning_rate": 9.621348130725617e-05,
      "loss": 0.693343460559845,
      "step": 2770
    },
    {
      "epoch": 1.169620253164557,
      "grad_norm": 1.2853080034255981,
      "learning_rate": 9.620430862884542e-05,
      "loss": 0.6999852061271667,
      "step": 2772
    },
    {
      "epoch": 1.1704641350210971,
      "grad_norm": 1.1782865524291992,
      "learning_rate": 9.619512529198806e-05,
      "loss": 0.6034331321716309,
      "step": 2774
    },
    {
      "epoch": 1.171308016877637,
      "grad_norm": 1.4055447578430176,
      "learning_rate": 9.61859312988025e-05,
      "loss": 0.7588269710540771,
      "step": 2776
    },
    {
      "epoch": 1.1721518987341772,
      "grad_norm": 1.1148805618286133,
      "learning_rate": 9.617672665140957e-05,
      "loss": 0.6913981437683105,
      "step": 2778
    },
    {
      "epoch": 1.1729957805907172,
      "grad_norm": 1.1311042308807373,
      "learning_rate": 9.616751135193266e-05,
      "loss": 0.5976925492286682,
      "step": 2780
    },
    {
      "epoch": 1.1738396624472573,
      "grad_norm": 1.2378602027893066,
      "learning_rate": 9.615828540249754e-05,
      "loss": 0.6897050142288208,
      "step": 2782
    },
    {
      "epoch": 1.1746835443037975,
      "grad_norm": 1.3445732593536377,
      "learning_rate": 9.614904880523248e-05,
      "loss": 0.6772098541259766,
      "step": 2784
    },
    {
      "epoch": 1.1755274261603375,
      "grad_norm": 1.3380862474441528,
      "learning_rate": 9.613980156226815e-05,
      "loss": 0.6354818344116211,
      "step": 2786
    },
    {
      "epoch": 1.1763713080168776,
      "grad_norm": 1.0955157279968262,
      "learning_rate": 9.613054367573773e-05,
      "loss": 0.6541208028793335,
      "step": 2788
    },
    {
      "epoch": 1.1772151898734178,
      "grad_norm": 1.0176626443862915,
      "learning_rate": 9.612127514777686e-05,
      "loss": 0.6472887992858887,
      "step": 2790
    },
    {
      "epoch": 1.1780590717299577,
      "grad_norm": 1.2644864320755005,
      "learning_rate": 9.611199598052357e-05,
      "loss": 0.7511212229728699,
      "step": 2792
    },
    {
      "epoch": 1.1789029535864979,
      "grad_norm": 1.248197317123413,
      "learning_rate": 9.61027061761184e-05,
      "loss": 0.696236789226532,
      "step": 2794
    },
    {
      "epoch": 1.179746835443038,
      "grad_norm": 1.189935564994812,
      "learning_rate": 9.609340573670436e-05,
      "loss": 0.5962010622024536,
      "step": 2796
    },
    {
      "epoch": 1.180590717299578,
      "grad_norm": 1.1760492324829102,
      "learning_rate": 9.608409466442685e-05,
      "loss": 0.5981685519218445,
      "step": 2798
    },
    {
      "epoch": 1.1814345991561181,
      "grad_norm": 1.1820716857910156,
      "learning_rate": 9.607477296143374e-05,
      "loss": 0.6186091303825378,
      "step": 2800
    },
    {
      "epoch": 1.1814345991561181,
      "eval_loss": 0.7298192977905273,
      "eval_runtime": 849.544,
      "eval_samples_per_second": 2.48,
      "eval_steps_per_second": 2.48,
      "step": 2800
    },
    {
      "epoch": 1.1822784810126583,
      "grad_norm": 1.0353888273239136,
      "learning_rate": 9.606544062987541e-05,
      "loss": 0.5859389901161194,
      "step": 2802
    },
    {
      "epoch": 1.1831223628691983,
      "grad_norm": 1.3141933679580688,
      "learning_rate": 9.605609767190464e-05,
      "loss": 0.6573460698127747,
      "step": 2804
    },
    {
      "epoch": 1.1839662447257384,
      "grad_norm": 1.1209372282028198,
      "learning_rate": 9.604674408967664e-05,
      "loss": 0.6991921067237854,
      "step": 2806
    },
    {
      "epoch": 1.1848101265822786,
      "grad_norm": 1.2830493450164795,
      "learning_rate": 9.603737988534913e-05,
      "loss": 0.6438087821006775,
      "step": 2808
    },
    {
      "epoch": 1.1856540084388185,
      "grad_norm": 1.1427195072174072,
      "learning_rate": 9.602800506108225e-05,
      "loss": 0.6452094316482544,
      "step": 2810
    },
    {
      "epoch": 1.1864978902953587,
      "grad_norm": 1.316420078277588,
      "learning_rate": 9.601861961903857e-05,
      "loss": 0.6745601296424866,
      "step": 2812
    },
    {
      "epoch": 1.1873417721518988,
      "grad_norm": 1.1643308401107788,
      "learning_rate": 9.600922356138317e-05,
      "loss": 0.6761514544487,
      "step": 2814
    },
    {
      "epoch": 1.1881856540084388,
      "grad_norm": 1.036056399345398,
      "learning_rate": 9.59998168902835e-05,
      "loss": 0.6453908681869507,
      "step": 2816
    },
    {
      "epoch": 1.189029535864979,
      "grad_norm": 1.2211129665374756,
      "learning_rate": 9.599039960790954e-05,
      "loss": 0.6576406359672546,
      "step": 2818
    },
    {
      "epoch": 1.189873417721519,
      "grad_norm": 1.084114670753479,
      "learning_rate": 9.598097171643364e-05,
      "loss": 0.6214181780815125,
      "step": 2820
    },
    {
      "epoch": 1.190717299578059,
      "grad_norm": 1.1297314167022705,
      "learning_rate": 9.597153321803064e-05,
      "loss": 0.6381646990776062,
      "step": 2822
    },
    {
      "epoch": 1.1915611814345992,
      "grad_norm": 1.2568120956420898,
      "learning_rate": 9.596208411487784e-05,
      "loss": 0.7129076719284058,
      "step": 2824
    },
    {
      "epoch": 1.1924050632911392,
      "grad_norm": 1.07041335105896,
      "learning_rate": 9.595262440915493e-05,
      "loss": 0.7123546004295349,
      "step": 2826
    },
    {
      "epoch": 1.1932489451476793,
      "grad_norm": 1.3950074911117554,
      "learning_rate": 9.594315410304413e-05,
      "loss": 0.7263038158416748,
      "step": 2828
    },
    {
      "epoch": 1.1940928270042195,
      "grad_norm": 1.2470672130584717,
      "learning_rate": 9.593367319873002e-05,
      "loss": 0.6863036751747131,
      "step": 2830
    },
    {
      "epoch": 1.1949367088607594,
      "grad_norm": 1.2065461874008179,
      "learning_rate": 9.592418169839968e-05,
      "loss": 0.745354175567627,
      "step": 2832
    },
    {
      "epoch": 1.1957805907172996,
      "grad_norm": 1.1710152626037598,
      "learning_rate": 9.591467960424261e-05,
      "loss": 0.6401656866073608,
      "step": 2834
    },
    {
      "epoch": 1.1966244725738397,
      "grad_norm": 1.3324087858200073,
      "learning_rate": 9.590516691845077e-05,
      "loss": 0.7402615547180176,
      "step": 2836
    },
    {
      "epoch": 1.1974683544303797,
      "grad_norm": 1.0100195407867432,
      "learning_rate": 9.589564364321855e-05,
      "loss": 0.5723769068717957,
      "step": 2838
    },
    {
      "epoch": 1.1983122362869199,
      "grad_norm": 1.2706246376037598,
      "learning_rate": 9.588610978074277e-05,
      "loss": 0.6618966460227966,
      "step": 2840
    },
    {
      "epoch": 1.1991561181434598,
      "grad_norm": 1.1921758651733398,
      "learning_rate": 9.587656533322273e-05,
      "loss": 0.7090804576873779,
      "step": 2842
    },
    {
      "epoch": 1.2,
      "grad_norm": 1.36713445186615,
      "learning_rate": 9.586701030286014e-05,
      "loss": 0.6930652856826782,
      "step": 2844
    },
    {
      "epoch": 1.2008438818565401,
      "grad_norm": 1.3084295988082886,
      "learning_rate": 9.585744469185917e-05,
      "loss": 0.7386236190795898,
      "step": 2846
    },
    {
      "epoch": 1.20168776371308,
      "grad_norm": 1.198922038078308,
      "learning_rate": 9.584786850242642e-05,
      "loss": 0.6179903149604797,
      "step": 2848
    },
    {
      "epoch": 1.2025316455696202,
      "grad_norm": 1.2106369733810425,
      "learning_rate": 9.583828173677092e-05,
      "loss": 0.7027528882026672,
      "step": 2850
    },
    {
      "epoch": 1.2033755274261604,
      "grad_norm": 1.2959522008895874,
      "learning_rate": 9.582868439710418e-05,
      "loss": 0.6612945199012756,
      "step": 2852
    },
    {
      "epoch": 1.2042194092827003,
      "grad_norm": 1.1441705226898193,
      "learning_rate": 9.58190764856401e-05,
      "loss": 0.7085917592048645,
      "step": 2854
    },
    {
      "epoch": 1.2050632911392405,
      "grad_norm": 1.1586185693740845,
      "learning_rate": 9.580945800459504e-05,
      "loss": 0.7480600476264954,
      "step": 2856
    },
    {
      "epoch": 1.2059071729957807,
      "grad_norm": 1.2068266868591309,
      "learning_rate": 9.579982895618783e-05,
      "loss": 0.7185836434364319,
      "step": 2858
    },
    {
      "epoch": 1.2067510548523206,
      "grad_norm": 1.2188525199890137,
      "learning_rate": 9.579018934263966e-05,
      "loss": 0.6737306118011475,
      "step": 2860
    },
    {
      "epoch": 1.2075949367088608,
      "grad_norm": 1.1513181924819946,
      "learning_rate": 9.578053916617423e-05,
      "loss": 0.7239293456077576,
      "step": 2862
    },
    {
      "epoch": 1.208438818565401,
      "grad_norm": 1.2063703536987305,
      "learning_rate": 9.577087842901764e-05,
      "loss": 0.6416276097297668,
      "step": 2864
    },
    {
      "epoch": 1.2092827004219409,
      "grad_norm": 1.102460503578186,
      "learning_rate": 9.576120713339844e-05,
      "loss": 0.697213351726532,
      "step": 2866
    },
    {
      "epoch": 1.210126582278481,
      "grad_norm": 1.2484638690948486,
      "learning_rate": 9.575152528154763e-05,
      "loss": 0.6664742231369019,
      "step": 2868
    },
    {
      "epoch": 1.2109704641350212,
      "grad_norm": 1.4476624727249146,
      "learning_rate": 9.57418328756986e-05,
      "loss": 0.6914868354797363,
      "step": 2870
    },
    {
      "epoch": 1.2118143459915611,
      "grad_norm": 1.0130122900009155,
      "learning_rate": 9.573212991808722e-05,
      "loss": 0.662024736404419,
      "step": 2872
    },
    {
      "epoch": 1.2126582278481013,
      "grad_norm": 1.014470100402832,
      "learning_rate": 9.572241641095177e-05,
      "loss": 0.6330409646034241,
      "step": 2874
    },
    {
      "epoch": 1.2135021097046415,
      "grad_norm": 1.1803333759307861,
      "learning_rate": 9.571269235653298e-05,
      "loss": 0.6607463955879211,
      "step": 2876
    },
    {
      "epoch": 1.2143459915611814,
      "grad_norm": 1.261366844177246,
      "learning_rate": 9.570295775707398e-05,
      "loss": 0.6925629377365112,
      "step": 2878
    },
    {
      "epoch": 1.2151898734177216,
      "grad_norm": 1.226670503616333,
      "learning_rate": 9.569321261482037e-05,
      "loss": 0.7070510983467102,
      "step": 2880
    },
    {
      "epoch": 1.2160337552742617,
      "grad_norm": 1.164565920829773,
      "learning_rate": 9.568345693202016e-05,
      "loss": 0.7243561744689941,
      "step": 2882
    },
    {
      "epoch": 1.2168776371308017,
      "grad_norm": 1.060331106185913,
      "learning_rate": 9.567369071092382e-05,
      "loss": 0.6316909790039062,
      "step": 2884
    },
    {
      "epoch": 1.2177215189873418,
      "grad_norm": 1.1998693943023682,
      "learning_rate": 9.566391395378419e-05,
      "loss": 0.6139125227928162,
      "step": 2886
    },
    {
      "epoch": 1.2185654008438818,
      "grad_norm": 1.1875834465026855,
      "learning_rate": 9.565412666285661e-05,
      "loss": 0.688897430896759,
      "step": 2888
    },
    {
      "epoch": 1.219409282700422,
      "grad_norm": 1.199174404144287,
      "learning_rate": 9.564432884039882e-05,
      "loss": 0.684590756893158,
      "step": 2890
    },
    {
      "epoch": 1.220253164556962,
      "grad_norm": 1.2428219318389893,
      "learning_rate": 9.563452048867099e-05,
      "loss": 0.67433100938797,
      "step": 2892
    },
    {
      "epoch": 1.221097046413502,
      "grad_norm": 1.0826431512832642,
      "learning_rate": 9.562470160993568e-05,
      "loss": 0.6959785223007202,
      "step": 2894
    },
    {
      "epoch": 1.2219409282700422,
      "grad_norm": 1.3140246868133545,
      "learning_rate": 9.561487220645797e-05,
      "loss": 0.6443175673484802,
      "step": 2896
    },
    {
      "epoch": 1.2227848101265824,
      "grad_norm": 1.2758334875106812,
      "learning_rate": 9.560503228050529e-05,
      "loss": 0.6715332865715027,
      "step": 2898
    },
    {
      "epoch": 1.2236286919831223,
      "grad_norm": 1.3326421976089478,
      "learning_rate": 9.559518183434753e-05,
      "loss": 0.6896081566810608,
      "step": 2900
    },
    {
      "epoch": 1.2236286919831223,
      "eval_loss": 0.7281573414802551,
      "eval_runtime": 854.563,
      "eval_samples_per_second": 2.466,
      "eval_steps_per_second": 2.466,
      "step": 2900
    },
    {
      "epoch": 1.2244725738396625,
      "grad_norm": 1.3225606679916382,
      "learning_rate": 9.558532087025697e-05,
      "loss": 0.6797633171081543,
      "step": 2902
    },
    {
      "epoch": 1.2253164556962026,
      "grad_norm": 1.3058340549468994,
      "learning_rate": 9.55754493905084e-05,
      "loss": 0.6510948538780212,
      "step": 2904
    },
    {
      "epoch": 1.2261603375527426,
      "grad_norm": 1.140268087387085,
      "learning_rate": 9.556556739737892e-05,
      "loss": 0.6481176614761353,
      "step": 2906
    },
    {
      "epoch": 1.2270042194092827,
      "grad_norm": 1.465113639831543,
      "learning_rate": 9.555567489314816e-05,
      "loss": 0.7533771991729736,
      "step": 2908
    },
    {
      "epoch": 1.2278481012658227,
      "grad_norm": 1.1468979120254517,
      "learning_rate": 9.554577188009812e-05,
      "loss": 0.6924305558204651,
      "step": 2910
    },
    {
      "epoch": 1.2286919831223628,
      "grad_norm": 1.2193517684936523,
      "learning_rate": 9.553585836051321e-05,
      "loss": 0.7082820534706116,
      "step": 2912
    },
    {
      "epoch": 1.229535864978903,
      "grad_norm": 1.2015037536621094,
      "learning_rate": 9.552593433668034e-05,
      "loss": 0.6735695004463196,
      "step": 2914
    },
    {
      "epoch": 1.230379746835443,
      "grad_norm": 1.1915435791015625,
      "learning_rate": 9.551599981088874e-05,
      "loss": 0.7312048673629761,
      "step": 2916
    },
    {
      "epoch": 1.231223628691983,
      "grad_norm": 1.2849410772323608,
      "learning_rate": 9.550605478543013e-05,
      "loss": 0.6590308547019958,
      "step": 2918
    },
    {
      "epoch": 1.2320675105485233,
      "grad_norm": 1.192238688468933,
      "learning_rate": 9.549609926259866e-05,
      "loss": 0.6237715482711792,
      "step": 2920
    },
    {
      "epoch": 1.2329113924050632,
      "grad_norm": 1.141845703125,
      "learning_rate": 9.548613324469085e-05,
      "loss": 0.6546295881271362,
      "step": 2922
    },
    {
      "epoch": 1.2337552742616034,
      "grad_norm": 1.1662311553955078,
      "learning_rate": 9.547615673400566e-05,
      "loss": 0.5800934433937073,
      "step": 2924
    },
    {
      "epoch": 1.2345991561181435,
      "grad_norm": 1.120578646659851,
      "learning_rate": 9.546616973284453e-05,
      "loss": 0.6487136483192444,
      "step": 2926
    },
    {
      "epoch": 1.2354430379746835,
      "grad_norm": 1.0884860754013062,
      "learning_rate": 9.54561722435112e-05,
      "loss": 0.7515342235565186,
      "step": 2928
    },
    {
      "epoch": 1.2362869198312236,
      "grad_norm": 1.4208670854568481,
      "learning_rate": 9.544616426831196e-05,
      "loss": 0.7162003517150879,
      "step": 2930
    },
    {
      "epoch": 1.2371308016877638,
      "grad_norm": 1.083389401435852,
      "learning_rate": 9.543614580955543e-05,
      "loss": 0.708450198173523,
      "step": 2932
    },
    {
      "epoch": 1.2379746835443037,
      "grad_norm": 1.141364336013794,
      "learning_rate": 9.542611686955268e-05,
      "loss": 0.6255859732627869,
      "step": 2934
    },
    {
      "epoch": 1.238818565400844,
      "grad_norm": 1.122036099433899,
      "learning_rate": 9.54160774506172e-05,
      "loss": 0.6485402584075928,
      "step": 2936
    },
    {
      "epoch": 1.239662447257384,
      "grad_norm": 1.3514165878295898,
      "learning_rate": 9.540602755506487e-05,
      "loss": 0.6735473871231079,
      "step": 2938
    },
    {
      "epoch": 1.240506329113924,
      "grad_norm": 1.1762629747390747,
      "learning_rate": 9.539596718521403e-05,
      "loss": 0.6154970526695251,
      "step": 2940
    },
    {
      "epoch": 1.2413502109704642,
      "grad_norm": 1.1609408855438232,
      "learning_rate": 9.53858963433854e-05,
      "loss": 0.6410251259803772,
      "step": 2942
    },
    {
      "epoch": 1.2421940928270043,
      "grad_norm": 1.1750361919403076,
      "learning_rate": 9.537581503190214e-05,
      "loss": 0.6841039657592773,
      "step": 2944
    },
    {
      "epoch": 1.2430379746835443,
      "grad_norm": 1.3125680685043335,
      "learning_rate": 9.536572325308982e-05,
      "loss": 0.7293462753295898,
      "step": 2946
    },
    {
      "epoch": 1.2438818565400844,
      "grad_norm": 1.1737277507781982,
      "learning_rate": 9.53556210092764e-05,
      "loss": 0.7713663578033447,
      "step": 2948
    },
    {
      "epoch": 1.2447257383966246,
      "grad_norm": 1.1702152490615845,
      "learning_rate": 9.53455083027923e-05,
      "loss": 0.6612298488616943,
      "step": 2950
    },
    {
      "epoch": 1.2455696202531645,
      "grad_norm": 1.2594486474990845,
      "learning_rate": 9.533538513597028e-05,
      "loss": 0.6725803017616272,
      "step": 2952
    },
    {
      "epoch": 1.2464135021097047,
      "grad_norm": 1.180816411972046,
      "learning_rate": 9.532525151114562e-05,
      "loss": 0.6421069502830505,
      "step": 2954
    },
    {
      "epoch": 1.2472573839662446,
      "grad_norm": 1.25814688205719,
      "learning_rate": 9.531510743065593e-05,
      "loss": 0.7042996287345886,
      "step": 2956
    },
    {
      "epoch": 1.2481012658227848,
      "grad_norm": 1.2101783752441406,
      "learning_rate": 9.530495289684122e-05,
      "loss": 0.7359137535095215,
      "step": 2958
    },
    {
      "epoch": 1.248945147679325,
      "grad_norm": 1.1438405513763428,
      "learning_rate": 9.5294787912044e-05,
      "loss": 0.6186386346817017,
      "step": 2960
    },
    {
      "epoch": 1.249789029535865,
      "grad_norm": 1.163364291191101,
      "learning_rate": 9.52846124786091e-05,
      "loss": 0.6243056058883667,
      "step": 2962
    },
    {
      "epoch": 1.250632911392405,
      "grad_norm": 1.0695953369140625,
      "learning_rate": 9.52744265988838e-05,
      "loss": 0.6568763852119446,
      "step": 2964
    },
    {
      "epoch": 1.2514767932489452,
      "grad_norm": 1.2228879928588867,
      "learning_rate": 9.52642302752178e-05,
      "loss": 0.6486776471138,
      "step": 2966
    },
    {
      "epoch": 1.2523206751054852,
      "grad_norm": 1.2262967824935913,
      "learning_rate": 9.52540235099632e-05,
      "loss": 0.6293455958366394,
      "step": 2968
    },
    {
      "epoch": 1.2531645569620253,
      "grad_norm": 1.0862956047058105,
      "learning_rate": 9.524380630547449e-05,
      "loss": 0.6549884080886841,
      "step": 2970
    },
    {
      "epoch": 1.2540084388185653,
      "grad_norm": 1.1721880435943604,
      "learning_rate": 9.52335786641086e-05,
      "loss": 0.6126490831375122,
      "step": 2972
    },
    {
      "epoch": 1.2548523206751054,
      "grad_norm": 1.2452391386032104,
      "learning_rate": 9.522334058822483e-05,
      "loss": 0.7078590393066406,
      "step": 2974
    },
    {
      "epoch": 1.2556962025316456,
      "grad_norm": 1.2290222644805908,
      "learning_rate": 9.521309208018492e-05,
      "loss": 0.6166214942932129,
      "step": 2976
    },
    {
      "epoch": 1.2565400843881855,
      "grad_norm": 1.1823618412017822,
      "learning_rate": 9.520283314235299e-05,
      "loss": 0.666228175163269,
      "step": 2978
    },
    {
      "epoch": 1.2573839662447257,
      "grad_norm": 1.1702475547790527,
      "learning_rate": 9.51925637770956e-05,
      "loss": 0.7436795830726624,
      "step": 2980
    },
    {
      "epoch": 1.2582278481012659,
      "grad_norm": 1.0879321098327637,
      "learning_rate": 9.518228398678168e-05,
      "loss": 0.7120893001556396,
      "step": 2982
    },
    {
      "epoch": 1.2590717299578058,
      "grad_norm": 1.1608418226242065,
      "learning_rate": 9.517199377378261e-05,
      "loss": 0.6931713223457336,
      "step": 2984
    },
    {
      "epoch": 1.259915611814346,
      "grad_norm": 1.1289087533950806,
      "learning_rate": 9.51616931404721e-05,
      "loss": 0.6803538799285889,
      "step": 2986
    },
    {
      "epoch": 1.2607594936708861,
      "grad_norm": 1.1622236967086792,
      "learning_rate": 9.515138208922633e-05,
      "loss": 0.6499706506729126,
      "step": 2988
    },
    {
      "epoch": 1.261603375527426,
      "grad_norm": 1.2492594718933105,
      "learning_rate": 9.514106062242386e-05,
      "loss": 0.6132655739784241,
      "step": 2990
    },
    {
      "epoch": 1.2624472573839662,
      "grad_norm": 1.1538822650909424,
      "learning_rate": 9.513072874244567e-05,
      "loss": 0.6309265494346619,
      "step": 2992
    },
    {
      "epoch": 1.2632911392405064,
      "grad_norm": 1.0828478336334229,
      "learning_rate": 9.512038645167509e-05,
      "loss": 0.6297751665115356,
      "step": 2994
    },
    {
      "epoch": 1.2641350210970463,
      "grad_norm": 1.2440937757492065,
      "learning_rate": 9.511003375249792e-05,
      "loss": 0.6335258483886719,
      "step": 2996
    },
    {
      "epoch": 1.2649789029535865,
      "grad_norm": 1.1259970664978027,
      "learning_rate": 9.50996706473023e-05,
      "loss": 0.6513770818710327,
      "step": 2998
    },
    {
      "epoch": 1.2658227848101267,
      "grad_norm": 1.1530309915542603,
      "learning_rate": 9.508929713847884e-05,
      "loss": 0.6490892767906189,
      "step": 3000
    },
    {
      "epoch": 1.2658227848101267,
      "eval_loss": 0.72515869140625,
      "eval_runtime": 868.0515,
      "eval_samples_per_second": 2.427,
      "eval_steps_per_second": 2.427,
      "step": 3000
    },
    {
      "epoch": 1.2666666666666666,
      "grad_norm": 1.2257169485092163,
      "learning_rate": 9.507891322842048e-05,
      "loss": 0.6936060786247253,
      "step": 3002
    },
    {
      "epoch": 1.2675105485232068,
      "grad_norm": 1.0380109548568726,
      "learning_rate": 9.506851891952259e-05,
      "loss": 0.5941951870918274,
      "step": 3004
    },
    {
      "epoch": 1.268354430379747,
      "grad_norm": 1.2830222845077515,
      "learning_rate": 9.505811421418296e-05,
      "loss": 0.648429811000824,
      "step": 3006
    },
    {
      "epoch": 1.2691983122362869,
      "grad_norm": 1.2212986946105957,
      "learning_rate": 9.504769911480171e-05,
      "loss": 0.6868565678596497,
      "step": 3008
    },
    {
      "epoch": 1.270042194092827,
      "grad_norm": 1.104656457901001,
      "learning_rate": 9.503727362378145e-05,
      "loss": 0.6777986288070679,
      "step": 3010
    },
    {
      "epoch": 1.2708860759493672,
      "grad_norm": 1.1449005603790283,
      "learning_rate": 9.502683774352713e-05,
      "loss": 0.6581128239631653,
      "step": 3012
    },
    {
      "epoch": 1.2717299578059071,
      "grad_norm": 1.2753362655639648,
      "learning_rate": 9.501639147644608e-05,
      "loss": 0.689930260181427,
      "step": 3014
    },
    {
      "epoch": 1.2725738396624473,
      "grad_norm": 1.3367106914520264,
      "learning_rate": 9.500593482494809e-05,
      "loss": 0.7549214363098145,
      "step": 3016
    },
    {
      "epoch": 1.2734177215189875,
      "grad_norm": 1.2309048175811768,
      "learning_rate": 9.499546779144528e-05,
      "loss": 0.6713513135910034,
      "step": 3018
    },
    {
      "epoch": 1.2742616033755274,
      "grad_norm": 1.3833240270614624,
      "learning_rate": 9.49849903783522e-05,
      "loss": 0.7045458555221558,
      "step": 3020
    },
    {
      "epoch": 1.2751054852320676,
      "grad_norm": 1.1402570009231567,
      "learning_rate": 9.49745025880858e-05,
      "loss": 0.708249568939209,
      "step": 3022
    },
    {
      "epoch": 1.2759493670886077,
      "grad_norm": 1.0476267337799072,
      "learning_rate": 9.496400442306541e-05,
      "loss": 0.616210401058197,
      "step": 3024
    },
    {
      "epoch": 1.2767932489451477,
      "grad_norm": 1.1045979261398315,
      "learning_rate": 9.495349588571274e-05,
      "loss": 0.6691827178001404,
      "step": 3026
    },
    {
      "epoch": 1.2776371308016878,
      "grad_norm": 1.1760368347167969,
      "learning_rate": 9.494297697845194e-05,
      "loss": 0.6198306083679199,
      "step": 3028
    },
    {
      "epoch": 1.2784810126582278,
      "grad_norm": 1.0015549659729004,
      "learning_rate": 9.493244770370946e-05,
      "loss": 0.5756480097770691,
      "step": 3030
    },
    {
      "epoch": 1.279324894514768,
      "grad_norm": 1.2190428972244263,
      "learning_rate": 9.492190806391427e-05,
      "loss": 0.6794419884681702,
      "step": 3032
    },
    {
      "epoch": 1.2801687763713079,
      "grad_norm": 1.0210410356521606,
      "learning_rate": 9.491135806149762e-05,
      "loss": 0.5847988724708557,
      "step": 3034
    },
    {
      "epoch": 1.281012658227848,
      "grad_norm": 1.0678503513336182,
      "learning_rate": 9.490079769889319e-05,
      "loss": 0.6760231256484985,
      "step": 3036
    },
    {
      "epoch": 1.2818565400843882,
      "grad_norm": 1.1811012029647827,
      "learning_rate": 9.489022697853709e-05,
      "loss": 0.7188448309898376,
      "step": 3038
    },
    {
      "epoch": 1.2827004219409281,
      "grad_norm": 1.1134302616119385,
      "learning_rate": 9.487964590286776e-05,
      "loss": 0.674904465675354,
      "step": 3040
    },
    {
      "epoch": 1.2835443037974683,
      "grad_norm": 1.1868232488632202,
      "learning_rate": 9.486905447432603e-05,
      "loss": 0.6016344428062439,
      "step": 3042
    },
    {
      "epoch": 1.2843881856540085,
      "grad_norm": 1.1586613655090332,
      "learning_rate": 9.485845269535517e-05,
      "loss": 0.6965603828430176,
      "step": 3044
    },
    {
      "epoch": 1.2852320675105484,
      "grad_norm": 1.149837613105774,
      "learning_rate": 9.48478405684008e-05,
      "loss": 0.656144380569458,
      "step": 3046
    },
    {
      "epoch": 1.2860759493670886,
      "grad_norm": 1.228752613067627,
      "learning_rate": 9.48372180959109e-05,
      "loss": 0.6388653516769409,
      "step": 3048
    },
    {
      "epoch": 1.2869198312236287,
      "grad_norm": 1.2403100728988647,
      "learning_rate": 9.482658528033595e-05,
      "loss": 0.6255465745925903,
      "step": 3050
    },
    {
      "epoch": 1.2877637130801687,
      "grad_norm": 1.2483839988708496,
      "learning_rate": 9.481594212412865e-05,
      "loss": 0.6828253269195557,
      "step": 3052
    },
    {
      "epoch": 1.2886075949367088,
      "grad_norm": 1.4161021709442139,
      "learning_rate": 9.480528862974422e-05,
      "loss": 0.7072080373764038,
      "step": 3054
    },
    {
      "epoch": 1.289451476793249,
      "grad_norm": 1.1500437259674072,
      "learning_rate": 9.479462479964021e-05,
      "loss": 0.6082415580749512,
      "step": 3056
    },
    {
      "epoch": 1.290295358649789,
      "grad_norm": 1.196595549583435,
      "learning_rate": 9.478395063627654e-05,
      "loss": 0.6653015613555908,
      "step": 3058
    },
    {
      "epoch": 1.2911392405063291,
      "grad_norm": 1.2832285165786743,
      "learning_rate": 9.477326614211557e-05,
      "loss": 0.7095832824707031,
      "step": 3060
    },
    {
      "epoch": 1.2919831223628693,
      "grad_norm": 1.2234288454055786,
      "learning_rate": 9.476257131962198e-05,
      "loss": 0.7183426022529602,
      "step": 3062
    },
    {
      "epoch": 1.2928270042194092,
      "grad_norm": 1.2350459098815918,
      "learning_rate": 9.475186617126286e-05,
      "loss": 0.713284432888031,
      "step": 3064
    },
    {
      "epoch": 1.2936708860759494,
      "grad_norm": 1.2079555988311768,
      "learning_rate": 9.47411506995077e-05,
      "loss": 0.6580002307891846,
      "step": 3066
    },
    {
      "epoch": 1.2945147679324895,
      "grad_norm": 1.129796028137207,
      "learning_rate": 9.473042490682835e-05,
      "loss": 0.5967763662338257,
      "step": 3068
    },
    {
      "epoch": 1.2953586497890295,
      "grad_norm": 1.1706618070602417,
      "learning_rate": 9.471968879569901e-05,
      "loss": 0.6724388003349304,
      "step": 3070
    },
    {
      "epoch": 1.2962025316455696,
      "grad_norm": 1.0336005687713623,
      "learning_rate": 9.470894236859635e-05,
      "loss": 0.6527577638626099,
      "step": 3072
    },
    {
      "epoch": 1.2970464135021098,
      "grad_norm": 1.1124558448791504,
      "learning_rate": 9.469818562799932e-05,
      "loss": 0.677132785320282,
      "step": 3074
    },
    {
      "epoch": 1.2978902953586497,
      "grad_norm": 1.158069372177124,
      "learning_rate": 9.468741857638933e-05,
      "loss": 0.649718165397644,
      "step": 3076
    },
    {
      "epoch": 1.29873417721519,
      "grad_norm": 1.092926263809204,
      "learning_rate": 9.46766412162501e-05,
      "loss": 0.6872133612632751,
      "step": 3078
    },
    {
      "epoch": 1.29957805907173,
      "grad_norm": 1.1324822902679443,
      "learning_rate": 9.466585355006777e-05,
      "loss": 0.6495246291160583,
      "step": 3080
    },
    {
      "epoch": 1.30042194092827,
      "grad_norm": 1.5882837772369385,
      "learning_rate": 9.465505558033086e-05,
      "loss": 0.6730570197105408,
      "step": 3082
    },
    {
      "epoch": 1.3012658227848102,
      "grad_norm": 0.9866069555282593,
      "learning_rate": 9.464424730953023e-05,
      "loss": 0.5677527785301208,
      "step": 3084
    },
    {
      "epoch": 1.3021097046413503,
      "grad_norm": 1.1560224294662476,
      "learning_rate": 9.463342874015917e-05,
      "loss": 0.6247856020927429,
      "step": 3086
    },
    {
      "epoch": 1.3029535864978903,
      "grad_norm": 1.135939359664917,
      "learning_rate": 9.462259987471329e-05,
      "loss": 0.6889358758926392,
      "step": 3088
    },
    {
      "epoch": 1.3037974683544304,
      "grad_norm": 1.3935760259628296,
      "learning_rate": 9.461176071569063e-05,
      "loss": 0.7097522020339966,
      "step": 3090
    },
    {
      "epoch": 1.3046413502109704,
      "grad_norm": 1.153518795967102,
      "learning_rate": 9.460091126559155e-05,
      "loss": 0.7044580578804016,
      "step": 3092
    },
    {
      "epoch": 1.3054852320675105,
      "grad_norm": 1.2112717628479004,
      "learning_rate": 9.45900515269188e-05,
      "loss": 0.6119300723075867,
      "step": 3094
    },
    {
      "epoch": 1.3063291139240507,
      "grad_norm": 1.295591115951538,
      "learning_rate": 9.457918150217754e-05,
      "loss": 0.7150222063064575,
      "step": 3096
    },
    {
      "epoch": 1.3071729957805907,
      "grad_norm": 1.1175775527954102,
      "learning_rate": 9.456830119387527e-05,
      "loss": 0.6043334007263184,
      "step": 3098
    },
    {
      "epoch": 1.3080168776371308,
      "grad_norm": 1.4022588729858398,
      "learning_rate": 9.455741060452186e-05,
      "loss": 0.6354425549507141,
      "step": 3100
    },
    {
      "epoch": 1.3080168776371308,
      "eval_loss": 0.7225774526596069,
      "eval_runtime": 862.4006,
      "eval_samples_per_second": 2.443,
      "eval_steps_per_second": 2.443,
      "step": 3100
    },
    {
      "epoch": 1.3088607594936708,
      "grad_norm": 1.1657692193984985,
      "learning_rate": 9.454650973662957e-05,
      "loss": 0.7281571626663208,
      "step": 3102
    },
    {
      "epoch": 1.309704641350211,
      "grad_norm": 1.6169127225875854,
      "learning_rate": 9.453559859271301e-05,
      "loss": 0.8038214445114136,
      "step": 3104
    },
    {
      "epoch": 1.310548523206751,
      "grad_norm": 1.1256520748138428,
      "learning_rate": 9.452467717528918e-05,
      "loss": 0.6488606333732605,
      "step": 3106
    },
    {
      "epoch": 1.311392405063291,
      "grad_norm": 1.1224530935287476,
      "learning_rate": 9.451374548687745e-05,
      "loss": 0.6897066235542297,
      "step": 3108
    },
    {
      "epoch": 1.3122362869198312,
      "grad_norm": 1.1123055219650269,
      "learning_rate": 9.450280352999952e-05,
      "loss": 0.6332913041114807,
      "step": 3110
    },
    {
      "epoch": 1.3130801687763713,
      "grad_norm": 1.1688940525054932,
      "learning_rate": 9.449185130717952e-05,
      "loss": 0.7426630854606628,
      "step": 3112
    },
    {
      "epoch": 1.3139240506329113,
      "grad_norm": 1.1898044347763062,
      "learning_rate": 9.44808888209439e-05,
      "loss": 0.7156099677085876,
      "step": 3114
    },
    {
      "epoch": 1.3147679324894515,
      "grad_norm": 1.3030686378479004,
      "learning_rate": 9.44699160738215e-05,
      "loss": 0.7150979042053223,
      "step": 3116
    },
    {
      "epoch": 1.3156118143459916,
      "grad_norm": 1.1539074182510376,
      "learning_rate": 9.445893306834352e-05,
      "loss": 0.6687285900115967,
      "step": 3118
    },
    {
      "epoch": 1.3164556962025316,
      "grad_norm": 1.311808466911316,
      "learning_rate": 9.444793980704355e-05,
      "loss": 0.7340983152389526,
      "step": 3120
    },
    {
      "epoch": 1.3172995780590717,
      "grad_norm": 1.3325430154800415,
      "learning_rate": 9.44369362924575e-05,
      "loss": 0.6620677709579468,
      "step": 3122
    },
    {
      "epoch": 1.3181434599156119,
      "grad_norm": 1.201518177986145,
      "learning_rate": 9.442592252712365e-05,
      "loss": 0.6169955134391785,
      "step": 3124
    },
    {
      "epoch": 1.3189873417721518,
      "grad_norm": 1.2124013900756836,
      "learning_rate": 9.441489851358272e-05,
      "loss": 0.6696792840957642,
      "step": 3126
    },
    {
      "epoch": 1.319831223628692,
      "grad_norm": 1.2186850309371948,
      "learning_rate": 9.440386425437768e-05,
      "loss": 0.7303428649902344,
      "step": 3128
    },
    {
      "epoch": 1.3206751054852321,
      "grad_norm": 1.3780523538589478,
      "learning_rate": 9.439281975205396e-05,
      "loss": 0.7093026638031006,
      "step": 3130
    },
    {
      "epoch": 1.321518987341772,
      "grad_norm": 1.233353614807129,
      "learning_rate": 9.438176500915932e-05,
      "loss": 0.6821767687797546,
      "step": 3132
    },
    {
      "epoch": 1.3223628691983123,
      "grad_norm": 1.2425329685211182,
      "learning_rate": 9.437070002824385e-05,
      "loss": 0.700680136680603,
      "step": 3134
    },
    {
      "epoch": 1.3232067510548524,
      "grad_norm": 1.1600432395935059,
      "learning_rate": 9.435962481186003e-05,
      "loss": 0.6173145771026611,
      "step": 3136
    },
    {
      "epoch": 1.3240506329113924,
      "grad_norm": 1.279336929321289,
      "learning_rate": 9.434853936256272e-05,
      "loss": 0.6597106456756592,
      "step": 3138
    },
    {
      "epoch": 1.3248945147679325,
      "grad_norm": 1.1787258386611938,
      "learning_rate": 9.433744368290909e-05,
      "loss": 0.6655287742614746,
      "step": 3140
    },
    {
      "epoch": 1.3257383966244727,
      "grad_norm": 1.3658509254455566,
      "learning_rate": 9.432633777545874e-05,
      "loss": 0.6312944889068604,
      "step": 3142
    },
    {
      "epoch": 1.3265822784810126,
      "grad_norm": 1.1220000982284546,
      "learning_rate": 9.431522164277356e-05,
      "loss": 0.6696156859397888,
      "step": 3144
    },
    {
      "epoch": 1.3274261603375528,
      "grad_norm": 1.224761724472046,
      "learning_rate": 9.430409528741783e-05,
      "loss": 0.6586571335792542,
      "step": 3146
    },
    {
      "epoch": 1.328270042194093,
      "grad_norm": 1.227510929107666,
      "learning_rate": 9.429295871195821e-05,
      "loss": 0.64905846118927,
      "step": 3148
    },
    {
      "epoch": 1.3291139240506329,
      "grad_norm": 1.1359103918075562,
      "learning_rate": 9.428181191896366e-05,
      "loss": 0.6407933831214905,
      "step": 3150
    },
    {
      "epoch": 1.329957805907173,
      "grad_norm": 1.2729473114013672,
      "learning_rate": 9.427065491100556e-05,
      "loss": 0.7004884481430054,
      "step": 3152
    },
    {
      "epoch": 1.3308016877637132,
      "grad_norm": 1.1182841062545776,
      "learning_rate": 9.42594876906576e-05,
      "loss": 0.6835907101631165,
      "step": 3154
    },
    {
      "epoch": 1.3316455696202532,
      "grad_norm": 1.2309781312942505,
      "learning_rate": 9.424831026049585e-05,
      "loss": 0.7476315498352051,
      "step": 3156
    },
    {
      "epoch": 1.3324894514767933,
      "grad_norm": 1.0857728719711304,
      "learning_rate": 9.423712262309873e-05,
      "loss": 0.6811426281929016,
      "step": 3158
    },
    {
      "epoch": 1.3333333333333333,
      "grad_norm": 1.299680233001709,
      "learning_rate": 9.4225924781047e-05,
      "loss": 0.6403942108154297,
      "step": 3160
    },
    {
      "epoch": 1.3341772151898734,
      "grad_norm": 1.226472020149231,
      "learning_rate": 9.421471673692382e-05,
      "loss": 0.6758930683135986,
      "step": 3162
    },
    {
      "epoch": 1.3350210970464136,
      "grad_norm": 1.1403205394744873,
      "learning_rate": 9.420349849331463e-05,
      "loss": 0.7119444608688354,
      "step": 3164
    },
    {
      "epoch": 1.3358649789029535,
      "grad_norm": 1.2888442277908325,
      "learning_rate": 9.419227005280729e-05,
      "loss": 0.7411463260650635,
      "step": 3166
    },
    {
      "epoch": 1.3367088607594937,
      "grad_norm": 1.1929190158843994,
      "learning_rate": 9.418103141799197e-05,
      "loss": 0.5992606282234192,
      "step": 3168
    },
    {
      "epoch": 1.3375527426160336,
      "grad_norm": 1.2574355602264404,
      "learning_rate": 9.416978259146122e-05,
      "loss": 0.6728890538215637,
      "step": 3170
    },
    {
      "epoch": 1.3383966244725738,
      "grad_norm": 0.9653727412223816,
      "learning_rate": 9.415852357580992e-05,
      "loss": 0.6294883489608765,
      "step": 3172
    },
    {
      "epoch": 1.339240506329114,
      "grad_norm": 1.2107670307159424,
      "learning_rate": 9.414725437363532e-05,
      "loss": 0.6816665530204773,
      "step": 3174
    },
    {
      "epoch": 1.340084388185654,
      "grad_norm": 1.024849534034729,
      "learning_rate": 9.4135974987537e-05,
      "loss": 0.6186381578445435,
      "step": 3176
    },
    {
      "epoch": 1.340928270042194,
      "grad_norm": 1.1556614637374878,
      "learning_rate": 9.41246854201169e-05,
      "loss": 0.6071005463600159,
      "step": 3178
    },
    {
      "epoch": 1.3417721518987342,
      "grad_norm": 1.2382808923721313,
      "learning_rate": 9.41133856739793e-05,
      "loss": 0.7871434092521667,
      "step": 3180
    },
    {
      "epoch": 1.3426160337552742,
      "grad_norm": 1.0499578714370728,
      "learning_rate": 9.410207575173082e-05,
      "loss": 0.6578201651573181,
      "step": 3182
    },
    {
      "epoch": 1.3434599156118143,
      "grad_norm": 1.2048250436782837,
      "learning_rate": 9.409075565598049e-05,
      "loss": 0.6271620392799377,
      "step": 3184
    },
    {
      "epoch": 1.3443037974683545,
      "grad_norm": 1.0287591218948364,
      "learning_rate": 9.407942538933958e-05,
      "loss": 0.5773864388465881,
      "step": 3186
    },
    {
      "epoch": 1.3451476793248944,
      "grad_norm": 1.1125097274780273,
      "learning_rate": 9.406808495442181e-05,
      "loss": 0.6745175719261169,
      "step": 3188
    },
    {
      "epoch": 1.3459915611814346,
      "grad_norm": 1.036125898361206,
      "learning_rate": 9.405673435384319e-05,
      "loss": 0.6001214385032654,
      "step": 3190
    },
    {
      "epoch": 1.3468354430379748,
      "grad_norm": 1.2771985530853271,
      "learning_rate": 9.404537359022207e-05,
      "loss": 0.6703945994377136,
      "step": 3192
    },
    {
      "epoch": 1.3476793248945147,
      "grad_norm": 1.0891097784042358,
      "learning_rate": 9.403400266617918e-05,
      "loss": 0.6159096360206604,
      "step": 3194
    },
    {
      "epoch": 1.3485232067510549,
      "grad_norm": 1.1926233768463135,
      "learning_rate": 9.402262158433755e-05,
      "loss": 0.6439315676689148,
      "step": 3196
    },
    {
      "epoch": 1.349367088607595,
      "grad_norm": 1.272557020187378,
      "learning_rate": 9.40112303473226e-05,
      "loss": 0.7125352025032043,
      "step": 3198
    },
    {
      "epoch": 1.350210970464135,
      "grad_norm": 1.052037239074707,
      "learning_rate": 9.399982895776207e-05,
      "loss": 0.594719648361206,
      "step": 3200
    },
    {
      "epoch": 1.350210970464135,
      "eval_loss": 0.7200453281402588,
      "eval_runtime": 846.2953,
      "eval_samples_per_second": 2.49,
      "eval_steps_per_second": 2.49,
      "step": 3200
    },
    {
      "epoch": 1.3510548523206751,
      "grad_norm": 1.204728126525879,
      "learning_rate": 9.398841741828601e-05,
      "loss": 0.6390520334243774,
      "step": 3202
    },
    {
      "epoch": 1.3518987341772153,
      "grad_norm": 1.0873899459838867,
      "learning_rate": 9.397699573152689e-05,
      "loss": 0.6010531187057495,
      "step": 3204
    },
    {
      "epoch": 1.3527426160337552,
      "grad_norm": 1.3124359846115112,
      "learning_rate": 9.396556390011944e-05,
      "loss": 0.724280834197998,
      "step": 3206
    },
    {
      "epoch": 1.3535864978902954,
      "grad_norm": 1.2179948091506958,
      "learning_rate": 9.395412192670075e-05,
      "loss": 0.6430405378341675,
      "step": 3208
    },
    {
      "epoch": 1.3544303797468356,
      "grad_norm": 1.2617219686508179,
      "learning_rate": 9.394266981391031e-05,
      "loss": 0.7188641428947449,
      "step": 3210
    },
    {
      "epoch": 1.3552742616033755,
      "grad_norm": 1.2151501178741455,
      "learning_rate": 9.393120756438988e-05,
      "loss": 0.6724364757537842,
      "step": 3212
    },
    {
      "epoch": 1.3561181434599157,
      "grad_norm": 1.221528172492981,
      "learning_rate": 9.391973518078357e-05,
      "loss": 0.6340664625167847,
      "step": 3214
    },
    {
      "epoch": 1.3569620253164558,
      "grad_norm": 1.3180092573165894,
      "learning_rate": 9.390825266573786e-05,
      "loss": 0.6914255023002625,
      "step": 3216
    },
    {
      "epoch": 1.3578059071729958,
      "grad_norm": 1.103994369506836,
      "learning_rate": 9.38967600219015e-05,
      "loss": 0.6137136220932007,
      "step": 3218
    },
    {
      "epoch": 1.358649789029536,
      "grad_norm": 1.33389413356781,
      "learning_rate": 9.38852572519257e-05,
      "loss": 0.7173700332641602,
      "step": 3220
    },
    {
      "epoch": 1.3594936708860759,
      "grad_norm": 1.1074159145355225,
      "learning_rate": 9.387374435846386e-05,
      "loss": 0.5942243933677673,
      "step": 3222
    },
    {
      "epoch": 1.360337552742616,
      "grad_norm": 1.1157063245773315,
      "learning_rate": 9.386222134417182e-05,
      "loss": 0.6362866163253784,
      "step": 3224
    },
    {
      "epoch": 1.3611814345991562,
      "grad_norm": 1.1717792749404907,
      "learning_rate": 9.38506882117077e-05,
      "loss": 0.6784523129463196,
      "step": 3226
    },
    {
      "epoch": 1.3620253164556961,
      "grad_norm": 1.0946043729782104,
      "learning_rate": 9.383914496373197e-05,
      "loss": 0.6647377014160156,
      "step": 3228
    },
    {
      "epoch": 1.3628691983122363,
      "grad_norm": 1.1519699096679688,
      "learning_rate": 9.382759160290746e-05,
      "loss": 0.6302075982093811,
      "step": 3230
    },
    {
      "epoch": 1.3637130801687762,
      "grad_norm": 0.9928684830665588,
      "learning_rate": 9.381602813189929e-05,
      "loss": 0.5979090332984924,
      "step": 3232
    },
    {
      "epoch": 1.3645569620253164,
      "grad_norm": 1.2488124370574951,
      "learning_rate": 9.380445455337492e-05,
      "loss": 0.6949353218078613,
      "step": 3234
    },
    {
      "epoch": 1.3654008438818566,
      "grad_norm": 1.3884797096252441,
      "learning_rate": 9.379287087000416e-05,
      "loss": 0.7225558161735535,
      "step": 3236
    },
    {
      "epoch": 1.3662447257383965,
      "grad_norm": 1.2981176376342773,
      "learning_rate": 9.378127708445917e-05,
      "loss": 0.6993390917778015,
      "step": 3238
    },
    {
      "epoch": 1.3670886075949367,
      "grad_norm": 0.9884640574455261,
      "learning_rate": 9.376967319941438e-05,
      "loss": 0.6983805894851685,
      "step": 3240
    },
    {
      "epoch": 1.3679324894514768,
      "grad_norm": 1.2051894664764404,
      "learning_rate": 9.375805921754659e-05,
      "loss": 0.7062534689903259,
      "step": 3242
    },
    {
      "epoch": 1.3687763713080168,
      "grad_norm": 1.1943434476852417,
      "learning_rate": 9.374643514153494e-05,
      "loss": 0.6405107378959656,
      "step": 3244
    },
    {
      "epoch": 1.369620253164557,
      "grad_norm": 1.249214768409729,
      "learning_rate": 9.373480097406086e-05,
      "loss": 0.6844781637191772,
      "step": 3246
    },
    {
      "epoch": 1.370464135021097,
      "grad_norm": 1.1847131252288818,
      "learning_rate": 9.372315671780813e-05,
      "loss": 0.6048306226730347,
      "step": 3248
    },
    {
      "epoch": 1.371308016877637,
      "grad_norm": 1.125545859336853,
      "learning_rate": 9.37115023754629e-05,
      "loss": 0.6772685050964355,
      "step": 3250
    },
    {
      "epoch": 1.3721518987341772,
      "grad_norm": 1.466615915298462,
      "learning_rate": 9.369983794971354e-05,
      "loss": 0.7536272406578064,
      "step": 3252
    },
    {
      "epoch": 1.3729957805907174,
      "grad_norm": 1.066699504852295,
      "learning_rate": 9.368816344325084e-05,
      "loss": 0.6640655398368835,
      "step": 3254
    },
    {
      "epoch": 1.3738396624472573,
      "grad_norm": 1.4793988466262817,
      "learning_rate": 9.367647885876787e-05,
      "loss": 0.7029458284378052,
      "step": 3256
    },
    {
      "epoch": 1.3746835443037975,
      "grad_norm": 1.258540153503418,
      "learning_rate": 9.366478419896006e-05,
      "loss": 0.7231863737106323,
      "step": 3258
    },
    {
      "epoch": 1.3755274261603376,
      "grad_norm": 1.176106333732605,
      "learning_rate": 9.365307946652512e-05,
      "loss": 0.6679144501686096,
      "step": 3260
    },
    {
      "epoch": 1.3763713080168776,
      "grad_norm": 1.3301753997802734,
      "learning_rate": 9.364136466416316e-05,
      "loss": 0.6282188296318054,
      "step": 3262
    },
    {
      "epoch": 1.3772151898734177,
      "grad_norm": 1.3616732358932495,
      "learning_rate": 9.362963979457648e-05,
      "loss": 0.6870840191841125,
      "step": 3264
    },
    {
      "epoch": 1.378059071729958,
      "grad_norm": 1.1982418298721313,
      "learning_rate": 9.361790486046985e-05,
      "loss": 0.6823731660842896,
      "step": 3266
    },
    {
      "epoch": 1.3789029535864978,
      "grad_norm": 1.1869033575057983,
      "learning_rate": 9.360615986455024e-05,
      "loss": 0.6582897305488586,
      "step": 3268
    },
    {
      "epoch": 1.379746835443038,
      "grad_norm": 1.1192975044250488,
      "learning_rate": 9.359440480952703e-05,
      "loss": 0.716654360294342,
      "step": 3270
    },
    {
      "epoch": 1.3805907172995782,
      "grad_norm": 1.2210016250610352,
      "learning_rate": 9.358263969811189e-05,
      "loss": 0.6880061626434326,
      "step": 3272
    },
    {
      "epoch": 1.381434599156118,
      "grad_norm": 1.0358284711837769,
      "learning_rate": 9.357086453301878e-05,
      "loss": 0.666864812374115,
      "step": 3274
    },
    {
      "epoch": 1.3822784810126583,
      "grad_norm": 1.2790803909301758,
      "learning_rate": 9.355907931696401e-05,
      "loss": 0.6872087121009827,
      "step": 3276
    },
    {
      "epoch": 1.3831223628691984,
      "grad_norm": 1.182991623878479,
      "learning_rate": 9.354728405266623e-05,
      "loss": 0.5929665565490723,
      "step": 3278
    },
    {
      "epoch": 1.3839662447257384,
      "grad_norm": 1.1071184873580933,
      "learning_rate": 9.353547874284634e-05,
      "loss": 0.5928181409835815,
      "step": 3280
    },
    {
      "epoch": 1.3848101265822785,
      "grad_norm": 1.3139623403549194,
      "learning_rate": 9.352366339022763e-05,
      "loss": 0.6783652901649475,
      "step": 3282
    },
    {
      "epoch": 1.3856540084388187,
      "grad_norm": 1.2534632682800293,
      "learning_rate": 9.351183799753567e-05,
      "loss": 0.7652941346168518,
      "step": 3284
    },
    {
      "epoch": 1.3864978902953586,
      "grad_norm": 1.4487930536270142,
      "learning_rate": 9.350000256749833e-05,
      "loss": 0.7430433630943298,
      "step": 3286
    },
    {
      "epoch": 1.3873417721518988,
      "grad_norm": 1.0786021947860718,
      "learning_rate": 9.348815710284584e-05,
      "loss": 0.5854598879814148,
      "step": 3288
    },
    {
      "epoch": 1.3881856540084387,
      "grad_norm": 1.0544480085372925,
      "learning_rate": 9.347630160631071e-05,
      "loss": 0.6365222334861755,
      "step": 3290
    },
    {
      "epoch": 1.389029535864979,
      "grad_norm": 0.9989988207817078,
      "learning_rate": 9.346443608062778e-05,
      "loss": 0.6485803127288818,
      "step": 3292
    },
    {
      "epoch": 1.389873417721519,
      "grad_norm": 1.100951910018921,
      "learning_rate": 9.345256052853419e-05,
      "loss": 0.6417753100395203,
      "step": 3294
    },
    {
      "epoch": 1.390717299578059,
      "grad_norm": 1.1398471593856812,
      "learning_rate": 9.344067495276942e-05,
      "loss": 0.6333693861961365,
      "step": 3296
    },
    {
      "epoch": 1.3915611814345992,
      "grad_norm": 1.1745941638946533,
      "learning_rate": 9.342877935607521e-05,
      "loss": 0.677288293838501,
      "step": 3298
    },
    {
      "epoch": 1.3924050632911391,
      "grad_norm": 1.2651115655899048,
      "learning_rate": 9.34168737411957e-05,
      "loss": 0.7408396005630493,
      "step": 3300
    },
    {
      "epoch": 1.3924050632911391,
      "eval_loss": 0.7173135876655579,
      "eval_runtime": 853.5344,
      "eval_samples_per_second": 2.469,
      "eval_steps_per_second": 2.469,
      "step": 3300
    },
    {
      "epoch": 1.3932489451476793,
      "grad_norm": 1.0747730731964111,
      "learning_rate": 9.340495811087723e-05,
      "loss": 0.6810371279716492,
      "step": 3302
    },
    {
      "epoch": 1.3940928270042194,
      "grad_norm": 1.2857651710510254,
      "learning_rate": 9.339303246786854e-05,
      "loss": 0.6693953275680542,
      "step": 3304
    },
    {
      "epoch": 1.3949367088607594,
      "grad_norm": 1.4544212818145752,
      "learning_rate": 9.338109681492063e-05,
      "loss": 0.7019274234771729,
      "step": 3306
    },
    {
      "epoch": 1.3957805907172995,
      "grad_norm": 1.687755823135376,
      "learning_rate": 9.336915115478685e-05,
      "loss": 0.6074224710464478,
      "step": 3308
    },
    {
      "epoch": 1.3966244725738397,
      "grad_norm": 1.1645431518554688,
      "learning_rate": 9.33571954902228e-05,
      "loss": 0.6981383562088013,
      "step": 3310
    },
    {
      "epoch": 1.3974683544303796,
      "grad_norm": 1.6173527240753174,
      "learning_rate": 9.334522982398646e-05,
      "loss": 0.7282926440238953,
      "step": 3312
    },
    {
      "epoch": 1.3983122362869198,
      "grad_norm": 1.3132909536361694,
      "learning_rate": 9.333325415883804e-05,
      "loss": 0.6574883460998535,
      "step": 3314
    },
    {
      "epoch": 1.39915611814346,
      "grad_norm": 1.1629762649536133,
      "learning_rate": 9.332126849754014e-05,
      "loss": 0.6559937596321106,
      "step": 3316
    },
    {
      "epoch": 1.4,
      "grad_norm": 1.1666897535324097,
      "learning_rate": 9.33092728428576e-05,
      "loss": 0.683718740940094,
      "step": 3318
    },
    {
      "epoch": 1.40084388185654,
      "grad_norm": 1.2269554138183594,
      "learning_rate": 9.329726719755756e-05,
      "loss": 0.6909779906272888,
      "step": 3320
    },
    {
      "epoch": 1.4016877637130802,
      "grad_norm": 1.1010066270828247,
      "learning_rate": 9.328525156440952e-05,
      "loss": 0.6051948666572571,
      "step": 3322
    },
    {
      "epoch": 1.4025316455696202,
      "grad_norm": 1.127143144607544,
      "learning_rate": 9.327322594618528e-05,
      "loss": 0.6266679763793945,
      "step": 3324
    },
    {
      "epoch": 1.4033755274261603,
      "grad_norm": 1.2160708904266357,
      "learning_rate": 9.326119034565887e-05,
      "loss": 0.6587526202201843,
      "step": 3326
    },
    {
      "epoch": 1.4042194092827005,
      "grad_norm": 1.0853947401046753,
      "learning_rate": 9.32491447656067e-05,
      "loss": 0.5916946530342102,
      "step": 3328
    },
    {
      "epoch": 1.4050632911392404,
      "grad_norm": 1.2205027341842651,
      "learning_rate": 9.323708920880744e-05,
      "loss": 0.6032452583312988,
      "step": 3330
    },
    {
      "epoch": 1.4059071729957806,
      "grad_norm": 1.1964668035507202,
      "learning_rate": 9.32250236780421e-05,
      "loss": 0.6649114489555359,
      "step": 3332
    },
    {
      "epoch": 1.4067510548523208,
      "grad_norm": 1.2507994174957275,
      "learning_rate": 9.321294817609394e-05,
      "loss": 0.7142994403839111,
      "step": 3334
    },
    {
      "epoch": 1.4075949367088607,
      "grad_norm": 1.1310259103775024,
      "learning_rate": 9.320086270574854e-05,
      "loss": 0.709568977355957,
      "step": 3336
    },
    {
      "epoch": 1.4084388185654009,
      "grad_norm": 1.2454090118408203,
      "learning_rate": 9.318876726979385e-05,
      "loss": 0.7800853848457336,
      "step": 3338
    },
    {
      "epoch": 1.409282700421941,
      "grad_norm": 1.1168389320373535,
      "learning_rate": 9.317666187101996e-05,
      "loss": 0.6187908053398132,
      "step": 3340
    },
    {
      "epoch": 1.410126582278481,
      "grad_norm": 1.6696287393569946,
      "learning_rate": 9.316454651221942e-05,
      "loss": 0.6222613453865051,
      "step": 3342
    },
    {
      "epoch": 1.4109704641350211,
      "grad_norm": 0.9500295519828796,
      "learning_rate": 9.315242119618698e-05,
      "loss": 0.6116594672203064,
      "step": 3344
    },
    {
      "epoch": 1.4118143459915613,
      "grad_norm": 1.186358094215393,
      "learning_rate": 9.314028592571973e-05,
      "loss": 0.633224368095398,
      "step": 3346
    },
    {
      "epoch": 1.4126582278481012,
      "grad_norm": 1.1855978965759277,
      "learning_rate": 9.312814070361705e-05,
      "loss": 0.6675921082496643,
      "step": 3348
    },
    {
      "epoch": 1.4135021097046414,
      "grad_norm": 1.2465872764587402,
      "learning_rate": 9.311598553268059e-05,
      "loss": 0.7268879413604736,
      "step": 3350
    },
    {
      "epoch": 1.4143459915611816,
      "grad_norm": 1.151274561882019,
      "learning_rate": 9.310382041571435e-05,
      "loss": 0.6147416830062866,
      "step": 3352
    },
    {
      "epoch": 1.4151898734177215,
      "grad_norm": 1.1226807832717896,
      "learning_rate": 9.309164535552453e-05,
      "loss": 0.6678543090820312,
      "step": 3354
    },
    {
      "epoch": 1.4160337552742617,
      "grad_norm": 1.375842571258545,
      "learning_rate": 9.307946035491975e-05,
      "loss": 0.6334129571914673,
      "step": 3356
    },
    {
      "epoch": 1.4168776371308016,
      "grad_norm": 1.058353066444397,
      "learning_rate": 9.306726541671081e-05,
      "loss": 0.6582583785057068,
      "step": 3358
    },
    {
      "epoch": 1.4177215189873418,
      "grad_norm": 1.0511330366134644,
      "learning_rate": 9.305506054371084e-05,
      "loss": 0.5877419114112854,
      "step": 3360
    },
    {
      "epoch": 1.4185654008438817,
      "grad_norm": 1.2246462106704712,
      "learning_rate": 9.304284573873532e-05,
      "loss": 0.711665689945221,
      "step": 3362
    },
    {
      "epoch": 1.4194092827004219,
      "grad_norm": 1.0242294073104858,
      "learning_rate": 9.303062100460193e-05,
      "loss": 0.6743642687797546,
      "step": 3364
    },
    {
      "epoch": 1.420253164556962,
      "grad_norm": 1.1432100534439087,
      "learning_rate": 9.301838634413069e-05,
      "loss": 0.6825576424598694,
      "step": 3366
    },
    {
      "epoch": 1.421097046413502,
      "grad_norm": 1.0128604173660278,
      "learning_rate": 9.30061417601439e-05,
      "loss": 0.624455988407135,
      "step": 3368
    },
    {
      "epoch": 1.4219409282700421,
      "grad_norm": 1.2738330364227295,
      "learning_rate": 9.299388725546617e-05,
      "loss": 0.7029586434364319,
      "step": 3370
    },
    {
      "epoch": 1.4227848101265823,
      "grad_norm": 1.0857324600219727,
      "learning_rate": 9.298162283292435e-05,
      "loss": 0.5994319915771484,
      "step": 3372
    },
    {
      "epoch": 1.4236286919831223,
      "grad_norm": 1.0811917781829834,
      "learning_rate": 9.296934849534763e-05,
      "loss": 0.6537772417068481,
      "step": 3374
    },
    {
      "epoch": 1.4244725738396624,
      "grad_norm": 1.006913185119629,
      "learning_rate": 9.295706424556745e-05,
      "loss": 0.5775008201599121,
      "step": 3376
    },
    {
      "epoch": 1.4253164556962026,
      "grad_norm": 1.2306486368179321,
      "learning_rate": 9.294477008641755e-05,
      "loss": 0.7445536255836487,
      "step": 3378
    },
    {
      "epoch": 1.4261603375527425,
      "grad_norm": 1.223608374595642,
      "learning_rate": 9.293246602073398e-05,
      "loss": 0.6081538796424866,
      "step": 3380
    },
    {
      "epoch": 1.4270042194092827,
      "grad_norm": 1.0933321714401245,
      "learning_rate": 9.2920152051355e-05,
      "loss": 0.6134634613990784,
      "step": 3382
    },
    {
      "epoch": 1.4278481012658228,
      "grad_norm": 1.1738401651382446,
      "learning_rate": 9.290782818112127e-05,
      "loss": 0.5961087346076965,
      "step": 3384
    },
    {
      "epoch": 1.4286919831223628,
      "grad_norm": 1.1493438482284546,
      "learning_rate": 9.289549441287561e-05,
      "loss": 0.6284122467041016,
      "step": 3386
    },
    {
      "epoch": 1.429535864978903,
      "grad_norm": 1.1907998323440552,
      "learning_rate": 9.288315074946324e-05,
      "loss": 0.6654639840126038,
      "step": 3388
    },
    {
      "epoch": 1.4303797468354431,
      "grad_norm": 1.3423025608062744,
      "learning_rate": 9.287079719373157e-05,
      "loss": 0.652850329875946,
      "step": 3390
    },
    {
      "epoch": 1.431223628691983,
      "grad_norm": 1.3932039737701416,
      "learning_rate": 9.285843374853034e-05,
      "loss": 0.703445315361023,
      "step": 3392
    },
    {
      "epoch": 1.4320675105485232,
      "grad_norm": 5.349400043487549,
      "learning_rate": 9.284606041671155e-05,
      "loss": 0.693265438079834,
      "step": 3394
    },
    {
      "epoch": 1.4329113924050634,
      "grad_norm": 1.0921961069107056,
      "learning_rate": 9.28336772011295e-05,
      "loss": 0.6578536033630371,
      "step": 3396
    },
    {
      "epoch": 1.4337552742616033,
      "grad_norm": 1.184157133102417,
      "learning_rate": 9.282128410464074e-05,
      "loss": 0.7092277407646179,
      "step": 3398
    },
    {
      "epoch": 1.4345991561181435,
      "grad_norm": 1.0923491716384888,
      "learning_rate": 9.280888113010415e-05,
      "loss": 0.6866328120231628,
      "step": 3400
    },
    {
      "epoch": 1.4345991561181435,
      "eval_loss": 0.715917706489563,
      "eval_runtime": 868.51,
      "eval_samples_per_second": 2.426,
      "eval_steps_per_second": 2.426,
      "step": 3400
    },
    {
      "epoch": 1.4354430379746836,
      "grad_norm": 1.2515597343444824,
      "learning_rate": 9.279646828038083e-05,
      "loss": 0.6617444157600403,
      "step": 3402
    },
    {
      "epoch": 1.4362869198312236,
      "grad_norm": 1.2122540473937988,
      "learning_rate": 9.278404555833422e-05,
      "loss": 0.6373176574707031,
      "step": 3404
    },
    {
      "epoch": 1.4371308016877637,
      "grad_norm": 1.191904902458191,
      "learning_rate": 9.277161296682997e-05,
      "loss": 0.6506488919258118,
      "step": 3406
    },
    {
      "epoch": 1.437974683544304,
      "grad_norm": 1.2492214441299438,
      "learning_rate": 9.275917050873606e-05,
      "loss": 0.7172291874885559,
      "step": 3408
    },
    {
      "epoch": 1.4388185654008439,
      "grad_norm": 1.0518640279769897,
      "learning_rate": 9.274671818692272e-05,
      "loss": 0.6180248260498047,
      "step": 3410
    },
    {
      "epoch": 1.439662447257384,
      "grad_norm": 1.150563359260559,
      "learning_rate": 9.273425600426245e-05,
      "loss": 0.6828892827033997,
      "step": 3412
    },
    {
      "epoch": 1.4405063291139242,
      "grad_norm": 1.76945960521698,
      "learning_rate": 9.272178396363005e-05,
      "loss": 0.6585919857025146,
      "step": 3414
    },
    {
      "epoch": 1.4413502109704641,
      "grad_norm": 1.2367758750915527,
      "learning_rate": 9.270930206790257e-05,
      "loss": 0.7548692226409912,
      "step": 3416
    },
    {
      "epoch": 1.4421940928270043,
      "grad_norm": 1.2292778491973877,
      "learning_rate": 9.269681031995936e-05,
      "loss": 0.7017102837562561,
      "step": 3418
    },
    {
      "epoch": 1.4430379746835442,
      "grad_norm": 1.2193396091461182,
      "learning_rate": 9.268430872268202e-05,
      "loss": 0.6657648682594299,
      "step": 3420
    },
    {
      "epoch": 1.4438818565400844,
      "grad_norm": 1.0505954027175903,
      "learning_rate": 9.267179727895443e-05,
      "loss": 0.6950910091400146,
      "step": 3422
    },
    {
      "epoch": 1.4447257383966245,
      "grad_norm": 1.1560698747634888,
      "learning_rate": 9.265927599166272e-05,
      "loss": 0.689308226108551,
      "step": 3424
    },
    {
      "epoch": 1.4455696202531645,
      "grad_norm": 1.189336895942688,
      "learning_rate": 9.264674486369533e-05,
      "loss": 0.6481659412384033,
      "step": 3426
    },
    {
      "epoch": 1.4464135021097047,
      "grad_norm": 1.3527976274490356,
      "learning_rate": 9.263420389794294e-05,
      "loss": 0.6626612544059753,
      "step": 3428
    },
    {
      "epoch": 1.4472573839662446,
      "grad_norm": 1.096303105354309,
      "learning_rate": 9.262165309729854e-05,
      "loss": 0.690841794013977,
      "step": 3430
    },
    {
      "epoch": 1.4481012658227848,
      "grad_norm": 1.2131421566009521,
      "learning_rate": 9.260909246465732e-05,
      "loss": 0.6497649550437927,
      "step": 3432
    },
    {
      "epoch": 1.448945147679325,
      "grad_norm": 1.1831032037734985,
      "learning_rate": 9.259652200291678e-05,
      "loss": 0.6236130595207214,
      "step": 3434
    },
    {
      "epoch": 1.4497890295358649,
      "grad_norm": 0.9745979309082031,
      "learning_rate": 9.25839417149767e-05,
      "loss": 0.5223423838615417,
      "step": 3436
    },
    {
      "epoch": 1.450632911392405,
      "grad_norm": 1.372460126876831,
      "learning_rate": 9.257135160373912e-05,
      "loss": 0.6642022728919983,
      "step": 3438
    },
    {
      "epoch": 1.4514767932489452,
      "grad_norm": 1.421044111251831,
      "learning_rate": 9.255875167210832e-05,
      "loss": 0.5426992774009705,
      "step": 3440
    },
    {
      "epoch": 1.4523206751054851,
      "grad_norm": 1.1694250106811523,
      "learning_rate": 9.254614192299086e-05,
      "loss": 0.6260567307472229,
      "step": 3442
    },
    {
      "epoch": 1.4531645569620253,
      "grad_norm": 1.0892298221588135,
      "learning_rate": 9.253352235929558e-05,
      "loss": 0.5776100158691406,
      "step": 3444
    },
    {
      "epoch": 1.4540084388185655,
      "grad_norm": 1.1841259002685547,
      "learning_rate": 9.252089298393356e-05,
      "loss": 0.6495202779769897,
      "step": 3446
    },
    {
      "epoch": 1.4548523206751054,
      "grad_norm": 1.1133549213409424,
      "learning_rate": 9.250825379981815e-05,
      "loss": 0.6570594906806946,
      "step": 3448
    },
    {
      "epoch": 1.4556962025316456,
      "grad_norm": 1.197100281715393,
      "learning_rate": 9.249560480986498e-05,
      "loss": 0.6496587991714478,
      "step": 3450
    },
    {
      "epoch": 1.4565400843881857,
      "grad_norm": 1.1661107540130615,
      "learning_rate": 9.248294601699193e-05,
      "loss": 0.6644704341888428,
      "step": 3452
    },
    {
      "epoch": 1.4573839662447257,
      "grad_norm": 1.2257879972457886,
      "learning_rate": 9.247027742411912e-05,
      "loss": 0.6451231241226196,
      "step": 3454
    },
    {
      "epoch": 1.4582278481012658,
      "grad_norm": 1.3634982109069824,
      "learning_rate": 9.245759903416897e-05,
      "loss": 0.6108601093292236,
      "step": 3456
    },
    {
      "epoch": 1.459071729957806,
      "grad_norm": 1.1802605390548706,
      "learning_rate": 9.244491085006615e-05,
      "loss": 0.6080004572868347,
      "step": 3458
    },
    {
      "epoch": 1.459915611814346,
      "grad_norm": 1.280831217765808,
      "learning_rate": 9.243221287473756e-05,
      "loss": 0.6406423449516296,
      "step": 3460
    },
    {
      "epoch": 1.460759493670886,
      "grad_norm": 1.3127192258834839,
      "learning_rate": 9.241950511111237e-05,
      "loss": 0.7320113778114319,
      "step": 3462
    },
    {
      "epoch": 1.4616033755274263,
      "grad_norm": 1.1711835861206055,
      "learning_rate": 9.240678756212204e-05,
      "loss": 0.572110652923584,
      "step": 3464
    },
    {
      "epoch": 1.4624472573839662,
      "grad_norm": 1.347143292427063,
      "learning_rate": 9.239406023070028e-05,
      "loss": 0.7446795105934143,
      "step": 3466
    },
    {
      "epoch": 1.4632911392405064,
      "grad_norm": 1.4953652620315552,
      "learning_rate": 9.238132311978299e-05,
      "loss": 0.6709978580474854,
      "step": 3468
    },
    {
      "epoch": 1.4641350210970465,
      "grad_norm": 1.2199387550354004,
      "learning_rate": 9.236857623230842e-05,
      "loss": 0.6691445112228394,
      "step": 3470
    },
    {
      "epoch": 1.4649789029535865,
      "grad_norm": 1.0959199666976929,
      "learning_rate": 9.235581957121702e-05,
      "loss": 0.6964292526245117,
      "step": 3472
    },
    {
      "epoch": 1.4658227848101266,
      "grad_norm": 1.455505609512329,
      "learning_rate": 9.234305313945149e-05,
      "loss": 0.6880454421043396,
      "step": 3474
    },
    {
      "epoch": 1.4666666666666668,
      "grad_norm": 1.2820862531661987,
      "learning_rate": 9.233027693995681e-05,
      "loss": 0.6737138032913208,
      "step": 3476
    },
    {
      "epoch": 1.4675105485232067,
      "grad_norm": 1.3459213972091675,
      "learning_rate": 9.231749097568023e-05,
      "loss": 0.6874006390571594,
      "step": 3478
    },
    {
      "epoch": 1.4683544303797469,
      "grad_norm": 1.2815442085266113,
      "learning_rate": 9.230469524957119e-05,
      "loss": 0.7179469466209412,
      "step": 3480
    },
    {
      "epoch": 1.469198312236287,
      "grad_norm": 1.6181597709655762,
      "learning_rate": 9.229188976458145e-05,
      "loss": 0.7525522112846375,
      "step": 3482
    },
    {
      "epoch": 1.470042194092827,
      "grad_norm": 1.0633227825164795,
      "learning_rate": 9.227907452366495e-05,
      "loss": 0.5918128490447998,
      "step": 3484
    },
    {
      "epoch": 1.4708860759493672,
      "grad_norm": 1.2055985927581787,
      "learning_rate": 9.226624952977796e-05,
      "loss": 0.6686186194419861,
      "step": 3486
    },
    {
      "epoch": 1.471729957805907,
      "grad_norm": 1.2495088577270508,
      "learning_rate": 9.225341478587893e-05,
      "loss": 0.764410674571991,
      "step": 3488
    },
    {
      "epoch": 1.4725738396624473,
      "grad_norm": 1.174229383468628,
      "learning_rate": 9.22405702949286e-05,
      "loss": 0.7066780924797058,
      "step": 3490
    },
    {
      "epoch": 1.4734177215189874,
      "grad_norm": 1.0970302820205688,
      "learning_rate": 9.222771605988995e-05,
      "loss": 0.6740228533744812,
      "step": 3492
    },
    {
      "epoch": 1.4742616033755274,
      "grad_norm": 1.2470436096191406,
      "learning_rate": 9.221485208372822e-05,
      "loss": 0.698371410369873,
      "step": 3494
    },
    {
      "epoch": 1.4751054852320675,
      "grad_norm": 1.0750112533569336,
      "learning_rate": 9.220197836941084e-05,
      "loss": 0.6354188919067383,
      "step": 3496
    },
    {
      "epoch": 1.4759493670886075,
      "grad_norm": 1.2656232118606567,
      "learning_rate": 9.218909491990757e-05,
      "loss": 0.7268608212471008,
      "step": 3498
    },
    {
      "epoch": 1.4767932489451476,
      "grad_norm": 1.2389028072357178,
      "learning_rate": 9.217620173819037e-05,
      "loss": 0.6652966141700745,
      "step": 3500
    },
    {
      "epoch": 1.4767932489451476,
      "eval_loss": 0.7155047059059143,
      "eval_runtime": 855.8428,
      "eval_samples_per_second": 2.462,
      "eval_steps_per_second": 2.462,
      "step": 3500
    },
    {
      "epoch": 1.4776371308016878,
      "grad_norm": 1.218304991722107,
      "learning_rate": 9.216329882723343e-05,
      "loss": 0.6845020651817322,
      "step": 3502
    },
    {
      "epoch": 1.4784810126582277,
      "grad_norm": 1.123903512954712,
      "learning_rate": 9.21503861900132e-05,
      "loss": 0.6972519755363464,
      "step": 3504
    },
    {
      "epoch": 1.479324894514768,
      "grad_norm": 1.1827739477157593,
      "learning_rate": 9.213746382950839e-05,
      "loss": 0.6699702739715576,
      "step": 3506
    },
    {
      "epoch": 1.480168776371308,
      "grad_norm": 0.9934872984886169,
      "learning_rate": 9.212453174869995e-05,
      "loss": 0.5623225569725037,
      "step": 3508
    },
    {
      "epoch": 1.481012658227848,
      "grad_norm": 1.221093773841858,
      "learning_rate": 9.211158995057105e-05,
      "loss": 0.6527173519134521,
      "step": 3510
    },
    {
      "epoch": 1.4818565400843882,
      "grad_norm": 1.4569166898727417,
      "learning_rate": 9.209863843810711e-05,
      "loss": 0.7015712261199951,
      "step": 3512
    },
    {
      "epoch": 1.4827004219409283,
      "grad_norm": 1.0764813423156738,
      "learning_rate": 9.208567721429581e-05,
      "loss": 0.6442505717277527,
      "step": 3514
    },
    {
      "epoch": 1.4835443037974683,
      "grad_norm": 2.1307506561279297,
      "learning_rate": 9.207270628212704e-05,
      "loss": 0.666451096534729,
      "step": 3516
    },
    {
      "epoch": 1.4843881856540084,
      "grad_norm": 1.180590271949768,
      "learning_rate": 9.205972564459296e-05,
      "loss": 0.6354807019233704,
      "step": 3518
    },
    {
      "epoch": 1.4852320675105486,
      "grad_norm": 1.2999447584152222,
      "learning_rate": 9.204673530468795e-05,
      "loss": 0.6080324053764343,
      "step": 3520
    },
    {
      "epoch": 1.4860759493670885,
      "grad_norm": 1.1680655479431152,
      "learning_rate": 9.203373526540862e-05,
      "loss": 0.6411244869232178,
      "step": 3522
    },
    {
      "epoch": 1.4869198312236287,
      "grad_norm": 1.0565013885498047,
      "learning_rate": 9.202072552975383e-05,
      "loss": 0.6498287916183472,
      "step": 3524
    },
    {
      "epoch": 1.4877637130801689,
      "grad_norm": 1.246267318725586,
      "learning_rate": 9.20077061007247e-05,
      "loss": 0.633613109588623,
      "step": 3526
    },
    {
      "epoch": 1.4886075949367088,
      "grad_norm": 1.0626300573349,
      "learning_rate": 9.199467698132453e-05,
      "loss": 0.6102107167243958,
      "step": 3528
    },
    {
      "epoch": 1.489451476793249,
      "grad_norm": 1.256600260734558,
      "learning_rate": 9.198163817455892e-05,
      "loss": 0.669352114200592,
      "step": 3530
    },
    {
      "epoch": 1.4902953586497891,
      "grad_norm": 1.143188238143921,
      "learning_rate": 9.196858968343565e-05,
      "loss": 0.6305804252624512,
      "step": 3532
    },
    {
      "epoch": 1.491139240506329,
      "grad_norm": 1.1471205949783325,
      "learning_rate": 9.195553151096475e-05,
      "loss": 0.6256994605064392,
      "step": 3534
    },
    {
      "epoch": 1.4919831223628692,
      "grad_norm": 1.1771589517593384,
      "learning_rate": 9.194246366015851e-05,
      "loss": 0.6395107507705688,
      "step": 3536
    },
    {
      "epoch": 1.4928270042194094,
      "grad_norm": 1.1997097730636597,
      "learning_rate": 9.192938613403144e-05,
      "loss": 0.6875160932540894,
      "step": 3538
    },
    {
      "epoch": 1.4936708860759493,
      "grad_norm": 1.3962169885635376,
      "learning_rate": 9.191629893560024e-05,
      "loss": 0.7216510772705078,
      "step": 3540
    },
    {
      "epoch": 1.4945147679324895,
      "grad_norm": 1.1835654973983765,
      "learning_rate": 9.19032020678839e-05,
      "loss": 0.6870693564414978,
      "step": 3542
    },
    {
      "epoch": 1.4953586497890297,
      "grad_norm": 1.112331509590149,
      "learning_rate": 9.18900955339036e-05,
      "loss": 0.6266092658042908,
      "step": 3544
    },
    {
      "epoch": 1.4962025316455696,
      "grad_norm": 1.0298354625701904,
      "learning_rate": 9.187697933668278e-05,
      "loss": 0.5906343460083008,
      "step": 3546
    },
    {
      "epoch": 1.4970464135021098,
      "grad_norm": 1.2650012969970703,
      "learning_rate": 9.186385347924709e-05,
      "loss": 0.6203610897064209,
      "step": 3548
    },
    {
      "epoch": 1.49789029535865,
      "grad_norm": 1.1208417415618896,
      "learning_rate": 9.185071796462441e-05,
      "loss": 0.6841281652450562,
      "step": 3550
    },
    {
      "epoch": 1.4987341772151899,
      "grad_norm": 1.1319488286972046,
      "learning_rate": 9.183757279584486e-05,
      "loss": 0.7089514136314392,
      "step": 3552
    },
    {
      "epoch": 1.49957805907173,
      "grad_norm": 1.1104235649108887,
      "learning_rate": 9.182441797594076e-05,
      "loss": 0.6663861870765686,
      "step": 3554
    },
    {
      "epoch": 1.5004219409282702,
      "grad_norm": 1.161412000656128,
      "learning_rate": 9.18112535079467e-05,
      "loss": 0.6713237762451172,
      "step": 3556
    },
    {
      "epoch": 1.5012658227848101,
      "grad_norm": 1.2925246953964233,
      "learning_rate": 9.179807939489945e-05,
      "loss": 0.6665274500846863,
      "step": 3558
    },
    {
      "epoch": 1.50210970464135,
      "grad_norm": 1.0968270301818848,
      "learning_rate": 9.178489563983802e-05,
      "loss": 0.6881593465805054,
      "step": 3560
    },
    {
      "epoch": 1.5029535864978905,
      "grad_norm": 1.111439824104309,
      "learning_rate": 9.177170224580368e-05,
      "loss": 0.631568431854248,
      "step": 3562
    },
    {
      "epoch": 1.5037974683544304,
      "grad_norm": 1.6731075048446655,
      "learning_rate": 9.175849921583986e-05,
      "loss": 0.6896167397499084,
      "step": 3564
    },
    {
      "epoch": 1.5046413502109703,
      "grad_norm": 1.226739525794983,
      "learning_rate": 9.174528655299226e-05,
      "loss": 0.6285277605056763,
      "step": 3566
    },
    {
      "epoch": 1.5054852320675105,
      "grad_norm": 1.2030941247940063,
      "learning_rate": 9.17320642603088e-05,
      "loss": 0.6256678700447083,
      "step": 3568
    },
    {
      "epoch": 1.5063291139240507,
      "grad_norm": 1.1980781555175781,
      "learning_rate": 9.171883234083958e-05,
      "loss": 0.6895992159843445,
      "step": 3570
    },
    {
      "epoch": 1.5071729957805906,
      "grad_norm": 1.2083429098129272,
      "learning_rate": 9.170559079763696e-05,
      "loss": 0.6642275452613831,
      "step": 3572
    },
    {
      "epoch": 1.5080168776371308,
      "grad_norm": 1.134020209312439,
      "learning_rate": 9.169233963375552e-05,
      "loss": 0.7441924214363098,
      "step": 3574
    },
    {
      "epoch": 1.508860759493671,
      "grad_norm": 1.8178621530532837,
      "learning_rate": 9.167907885225204e-05,
      "loss": 0.6435995101928711,
      "step": 3576
    },
    {
      "epoch": 1.5097046413502109,
      "grad_norm": 1.3850326538085938,
      "learning_rate": 9.166580845618553e-05,
      "loss": 0.6933603882789612,
      "step": 3578
    },
    {
      "epoch": 1.510548523206751,
      "grad_norm": 1.2500641345977783,
      "learning_rate": 9.165252844861723e-05,
      "loss": 0.6686714887619019,
      "step": 3580
    },
    {
      "epoch": 1.5113924050632912,
      "grad_norm": 1.0226643085479736,
      "learning_rate": 9.163923883261056e-05,
      "loss": 0.607890248298645,
      "step": 3582
    },
    {
      "epoch": 1.5122362869198311,
      "grad_norm": 1.233402132987976,
      "learning_rate": 9.162593961123118e-05,
      "loss": 0.6604583859443665,
      "step": 3584
    },
    {
      "epoch": 1.5130801687763713,
      "grad_norm": 1.2609056234359741,
      "learning_rate": 9.161263078754698e-05,
      "loss": 0.6756428480148315,
      "step": 3586
    },
    {
      "epoch": 1.5139240506329115,
      "grad_norm": 1.22673761844635,
      "learning_rate": 9.159931236462805e-05,
      "loss": 0.6990940570831299,
      "step": 3588
    },
    {
      "epoch": 1.5147679324894514,
      "grad_norm": 1.1386182308197021,
      "learning_rate": 9.158598434554668e-05,
      "loss": 0.6436648964881897,
      "step": 3590
    },
    {
      "epoch": 1.5156118143459916,
      "grad_norm": 1.1136831045150757,
      "learning_rate": 9.157264673337739e-05,
      "loss": 0.6420145034790039,
      "step": 3592
    },
    {
      "epoch": 1.5164556962025317,
      "grad_norm": 1.1957908868789673,
      "learning_rate": 9.155929953119693e-05,
      "loss": 0.6518592834472656,
      "step": 3594
    },
    {
      "epoch": 1.5172995780590717,
      "grad_norm": 1.1049647331237793,
      "learning_rate": 9.154594274208422e-05,
      "loss": 0.6891129612922668,
      "step": 3596
    },
    {
      "epoch": 1.5181434599156118,
      "grad_norm": 1.243675947189331,
      "learning_rate": 9.153257636912043e-05,
      "loss": 0.6945107579231262,
      "step": 3598
    },
    {
      "epoch": 1.518987341772152,
      "grad_norm": 1.2633713483810425,
      "learning_rate": 9.15192004153889e-05,
      "loss": 0.7011660933494568,
      "step": 3600
    },
    {
      "epoch": 1.518987341772152,
      "eval_loss": 0.7118256688117981,
      "eval_runtime": 851.3079,
      "eval_samples_per_second": 2.475,
      "eval_steps_per_second": 2.475,
      "step": 3600
    },
    {
      "epoch": 1.519831223628692,
      "grad_norm": 1.2995525598526,
      "learning_rate": 9.150581488397525e-05,
      "loss": 0.6843758821487427,
      "step": 3602
    },
    {
      "epoch": 1.520675105485232,
      "grad_norm": 1.3140910863876343,
      "learning_rate": 9.149241977796723e-05,
      "loss": 0.6699353456497192,
      "step": 3604
    },
    {
      "epoch": 1.5215189873417723,
      "grad_norm": 1.2674909830093384,
      "learning_rate": 9.147901510045485e-05,
      "loss": 0.7269271612167358,
      "step": 3606
    },
    {
      "epoch": 1.5223628691983122,
      "grad_norm": 1.0232038497924805,
      "learning_rate": 9.146560085453031e-05,
      "loss": 0.5556837916374207,
      "step": 3608
    },
    {
      "epoch": 1.5232067510548524,
      "grad_norm": 1.2598992586135864,
      "learning_rate": 9.1452177043288e-05,
      "loss": 0.7273092269897461,
      "step": 3610
    },
    {
      "epoch": 1.5240506329113925,
      "grad_norm": 1.2002917528152466,
      "learning_rate": 9.143874366982455e-05,
      "loss": 0.6897470355033875,
      "step": 3612
    },
    {
      "epoch": 1.5248945147679325,
      "grad_norm": 1.0959099531173706,
      "learning_rate": 9.142530073723878e-05,
      "loss": 0.6060715913772583,
      "step": 3614
    },
    {
      "epoch": 1.5257383966244724,
      "grad_norm": 1.9890750646591187,
      "learning_rate": 9.141184824863173e-05,
      "loss": 0.6585046052932739,
      "step": 3616
    },
    {
      "epoch": 1.5265822784810128,
      "grad_norm": 1.1460137367248535,
      "learning_rate": 9.139838620710663e-05,
      "loss": 0.6022046804428101,
      "step": 3618
    },
    {
      "epoch": 1.5274261603375527,
      "grad_norm": 1.193206548690796,
      "learning_rate": 9.138491461576888e-05,
      "loss": 0.6332581639289856,
      "step": 3620
    },
    {
      "epoch": 1.5282700421940927,
      "grad_norm": 1.2813689708709717,
      "learning_rate": 9.137143347772614e-05,
      "loss": 0.6690208315849304,
      "step": 3622
    },
    {
      "epoch": 1.529113924050633,
      "grad_norm": 1.0950052738189697,
      "learning_rate": 9.135794279608827e-05,
      "loss": 0.6034293174743652,
      "step": 3624
    },
    {
      "epoch": 1.529957805907173,
      "grad_norm": 1.208884358406067,
      "learning_rate": 9.134444257396729e-05,
      "loss": 0.7077960968017578,
      "step": 3626
    },
    {
      "epoch": 1.530801687763713,
      "grad_norm": 1.093759298324585,
      "learning_rate": 9.133093281447742e-05,
      "loss": 0.6741147637367249,
      "step": 3628
    },
    {
      "epoch": 1.5316455696202531,
      "grad_norm": 1.1280012130737305,
      "learning_rate": 9.131741352073514e-05,
      "loss": 0.6816818118095398,
      "step": 3630
    },
    {
      "epoch": 1.5324894514767933,
      "grad_norm": 1.2868385314941406,
      "learning_rate": 9.130388469585907e-05,
      "loss": 0.7149180769920349,
      "step": 3632
    },
    {
      "epoch": 1.5333333333333332,
      "grad_norm": 0.9654553532600403,
      "learning_rate": 9.129034634297007e-05,
      "loss": 0.613467812538147,
      "step": 3634
    },
    {
      "epoch": 1.5341772151898734,
      "grad_norm": 1.8958736658096313,
      "learning_rate": 9.127679846519115e-05,
      "loss": 0.7034116387367249,
      "step": 3636
    },
    {
      "epoch": 1.5350210970464135,
      "grad_norm": 1.305284857749939,
      "learning_rate": 9.126324106564757e-05,
      "loss": 0.7076106667518616,
      "step": 3638
    },
    {
      "epoch": 1.5358649789029535,
      "grad_norm": 1.1843762397766113,
      "learning_rate": 9.124967414746675e-05,
      "loss": 0.6671180725097656,
      "step": 3640
    },
    {
      "epoch": 1.5367088607594936,
      "grad_norm": 1.0460047721862793,
      "learning_rate": 9.123609771377832e-05,
      "loss": 0.667533814907074,
      "step": 3642
    },
    {
      "epoch": 1.5375527426160338,
      "grad_norm": 1.0441135168075562,
      "learning_rate": 9.122251176771409e-05,
      "loss": 0.6454499959945679,
      "step": 3644
    },
    {
      "epoch": 1.5383966244725737,
      "grad_norm": 1.5647634267807007,
      "learning_rate": 9.120891631240811e-05,
      "loss": 0.677007794380188,
      "step": 3646
    },
    {
      "epoch": 1.539240506329114,
      "grad_norm": 1.0650273561477661,
      "learning_rate": 9.119531135099655e-05,
      "loss": 0.7017449736595154,
      "step": 3648
    },
    {
      "epoch": 1.540084388185654,
      "grad_norm": 1.2904767990112305,
      "learning_rate": 9.118169688661784e-05,
      "loss": 0.683830738067627,
      "step": 3650
    },
    {
      "epoch": 1.540928270042194,
      "grad_norm": 1.1278672218322754,
      "learning_rate": 9.116807292241257e-05,
      "loss": 0.5923286080360413,
      "step": 3652
    },
    {
      "epoch": 1.5417721518987342,
      "grad_norm": 1.1107184886932373,
      "learning_rate": 9.115443946152352e-05,
      "loss": 0.6595140099525452,
      "step": 3654
    },
    {
      "epoch": 1.5426160337552743,
      "grad_norm": 1.0917898416519165,
      "learning_rate": 9.114079650709566e-05,
      "loss": 0.655241072177887,
      "step": 3656
    },
    {
      "epoch": 1.5434599156118143,
      "grad_norm": 1.1922433376312256,
      "learning_rate": 9.11271440622762e-05,
      "loss": 0.5987096428871155,
      "step": 3658
    },
    {
      "epoch": 1.5443037974683544,
      "grad_norm": 0.9974617958068848,
      "learning_rate": 9.111348213021445e-05,
      "loss": 0.5710145235061646,
      "step": 3660
    },
    {
      "epoch": 1.5451476793248946,
      "grad_norm": 1.133683443069458,
      "learning_rate": 9.109981071406197e-05,
      "loss": 0.6067734360694885,
      "step": 3662
    },
    {
      "epoch": 1.5459915611814345,
      "grad_norm": 1.1958736181259155,
      "learning_rate": 9.108612981697248e-05,
      "loss": 0.622981071472168,
      "step": 3664
    },
    {
      "epoch": 1.5468354430379747,
      "grad_norm": 1.234328031539917,
      "learning_rate": 9.107243944210194e-05,
      "loss": 0.6520710587501526,
      "step": 3666
    },
    {
      "epoch": 1.5476793248945149,
      "grad_norm": 1.0374714136123657,
      "learning_rate": 9.105873959260842e-05,
      "loss": 0.5993341207504272,
      "step": 3668
    },
    {
      "epoch": 1.5485232067510548,
      "grad_norm": 0.9987428784370422,
      "learning_rate": 9.104503027165223e-05,
      "loss": 0.6564813852310181,
      "step": 3670
    },
    {
      "epoch": 1.549367088607595,
      "grad_norm": 1.0823339223861694,
      "learning_rate": 9.103131148239584e-05,
      "loss": 0.61710524559021,
      "step": 3672
    },
    {
      "epoch": 1.5502109704641351,
      "grad_norm": 1.3481065034866333,
      "learning_rate": 9.101758322800391e-05,
      "loss": 0.687752366065979,
      "step": 3674
    },
    {
      "epoch": 1.551054852320675,
      "grad_norm": 1.2243965864181519,
      "learning_rate": 9.10038455116433e-05,
      "loss": 0.5981095433235168,
      "step": 3676
    },
    {
      "epoch": 1.5518987341772152,
      "grad_norm": 1.1384631395339966,
      "learning_rate": 9.0990098336483e-05,
      "loss": 0.7181004285812378,
      "step": 3678
    },
    {
      "epoch": 1.5527426160337554,
      "grad_norm": 1.042925477027893,
      "learning_rate": 9.097634170569426e-05,
      "loss": 0.6137188076972961,
      "step": 3680
    },
    {
      "epoch": 1.5535864978902953,
      "grad_norm": 1.372023105621338,
      "learning_rate": 9.096257562245045e-05,
      "loss": 0.6761168241500854,
      "step": 3682
    },
    {
      "epoch": 1.5544303797468353,
      "grad_norm": 1.0574673414230347,
      "learning_rate": 9.094880008992714e-05,
      "loss": 0.614276647567749,
      "step": 3684
    },
    {
      "epoch": 1.5552742616033757,
      "grad_norm": 1.2894645929336548,
      "learning_rate": 9.093501511130208e-05,
      "loss": 0.668122410774231,
      "step": 3686
    },
    {
      "epoch": 1.5561181434599156,
      "grad_norm": 1.2241230010986328,
      "learning_rate": 9.092122068975523e-05,
      "loss": 0.6305631399154663,
      "step": 3688
    },
    {
      "epoch": 1.5569620253164556,
      "grad_norm": 1.1316208839416504,
      "learning_rate": 9.090741682846866e-05,
      "loss": 0.633276641368866,
      "step": 3690
    },
    {
      "epoch": 1.557805907172996,
      "grad_norm": 1.2857953310012817,
      "learning_rate": 9.089360353062666e-05,
      "loss": 0.6657599806785583,
      "step": 3692
    },
    {
      "epoch": 1.5586497890295359,
      "grad_norm": 1.2325671911239624,
      "learning_rate": 9.087978079941573e-05,
      "loss": 0.6379332542419434,
      "step": 3694
    },
    {
      "epoch": 1.5594936708860758,
      "grad_norm": 1.3286080360412598,
      "learning_rate": 9.086594863802445e-05,
      "loss": 0.6841909885406494,
      "step": 3696
    },
    {
      "epoch": 1.560337552742616,
      "grad_norm": 1.261890172958374,
      "learning_rate": 9.085210704964368e-05,
      "loss": 0.6735964417457581,
      "step": 3698
    },
    {
      "epoch": 1.5611814345991561,
      "grad_norm": 1.0922305583953857,
      "learning_rate": 9.083825603746639e-05,
      "loss": 0.6602351665496826,
      "step": 3700
    },
    {
      "epoch": 1.5611814345991561,
      "eval_loss": 0.7099412679672241,
      "eval_runtime": 857.2273,
      "eval_samples_per_second": 2.458,
      "eval_steps_per_second": 2.458,
      "step": 3700
    },
    {
      "epoch": 1.562025316455696,
      "grad_norm": 1.1113468408584595,
      "learning_rate": 9.082439560468774e-05,
      "loss": 0.6590834259986877,
      "step": 3702
    },
    {
      "epoch": 1.5628691983122363,
      "grad_norm": 1.1476659774780273,
      "learning_rate": 9.081052575450508e-05,
      "loss": 0.6397460103034973,
      "step": 3704
    },
    {
      "epoch": 1.5637130801687764,
      "grad_norm": 1.2270452976226807,
      "learning_rate": 9.07966464901179e-05,
      "loss": 0.6337460279464722,
      "step": 3706
    },
    {
      "epoch": 1.5645569620253164,
      "grad_norm": 1.233667016029358,
      "learning_rate": 9.07827578147279e-05,
      "loss": 0.680374801158905,
      "step": 3708
    },
    {
      "epoch": 1.5654008438818565,
      "grad_norm": 1.0761466026306152,
      "learning_rate": 9.076885973153891e-05,
      "loss": 0.6234241724014282,
      "step": 3710
    },
    {
      "epoch": 1.5662447257383967,
      "grad_norm": 0.9219012260437012,
      "learning_rate": 9.075495224375697e-05,
      "loss": 0.6096800565719604,
      "step": 3712
    },
    {
      "epoch": 1.5670886075949366,
      "grad_norm": 1.151168942451477,
      "learning_rate": 9.074103535459026e-05,
      "loss": 0.649919867515564,
      "step": 3714
    },
    {
      "epoch": 1.5679324894514768,
      "grad_norm": 1.1380470991134644,
      "learning_rate": 9.072710906724914e-05,
      "loss": 0.6704574227333069,
      "step": 3716
    },
    {
      "epoch": 1.568776371308017,
      "grad_norm": 1.2184447050094604,
      "learning_rate": 9.071317338494614e-05,
      "loss": 0.6619362831115723,
      "step": 3718
    },
    {
      "epoch": 1.5696202531645569,
      "grad_norm": 1.131170630455017,
      "learning_rate": 9.069922831089594e-05,
      "loss": 0.6179121732711792,
      "step": 3720
    },
    {
      "epoch": 1.570464135021097,
      "grad_norm": 1.2668405771255493,
      "learning_rate": 9.06852738483154e-05,
      "loss": 0.594958484172821,
      "step": 3722
    },
    {
      "epoch": 1.5713080168776372,
      "grad_norm": 1.1624782085418701,
      "learning_rate": 9.067131000042359e-05,
      "loss": 0.6323778629302979,
      "step": 3724
    },
    {
      "epoch": 1.5721518987341772,
      "grad_norm": 1.2936128377914429,
      "learning_rate": 9.065733677044166e-05,
      "loss": 0.628058910369873,
      "step": 3726
    },
    {
      "epoch": 1.5729957805907173,
      "grad_norm": 1.1847784519195557,
      "learning_rate": 9.064335416159296e-05,
      "loss": 0.6472614407539368,
      "step": 3728
    },
    {
      "epoch": 1.5738396624472575,
      "grad_norm": 1.8903449773788452,
      "learning_rate": 9.062936217710305e-05,
      "loss": 0.6395491361618042,
      "step": 3730
    },
    {
      "epoch": 1.5746835443037974,
      "grad_norm": 1.1150785684585571,
      "learning_rate": 9.061536082019956e-05,
      "loss": 0.6911961436271667,
      "step": 3732
    },
    {
      "epoch": 1.5755274261603376,
      "grad_norm": 1.1206107139587402,
      "learning_rate": 9.060135009411239e-05,
      "loss": 0.7051874399185181,
      "step": 3734
    },
    {
      "epoch": 1.5763713080168777,
      "grad_norm": 1.27924382686615,
      "learning_rate": 9.05873300020735e-05,
      "loss": 0.7012752890586853,
      "step": 3736
    },
    {
      "epoch": 1.5772151898734177,
      "grad_norm": 1.3970832824707031,
      "learning_rate": 9.057330054731707e-05,
      "loss": 0.7185142040252686,
      "step": 3738
    },
    {
      "epoch": 1.5780590717299579,
      "grad_norm": 0.9732457995414734,
      "learning_rate": 9.055926173307945e-05,
      "loss": 0.6298858523368835,
      "step": 3740
    },
    {
      "epoch": 1.578902953586498,
      "grad_norm": 1.230928897857666,
      "learning_rate": 9.054521356259909e-05,
      "loss": 0.7142943739891052,
      "step": 3742
    },
    {
      "epoch": 1.579746835443038,
      "grad_norm": 1.1297426223754883,
      "learning_rate": 9.053115603911664e-05,
      "loss": 0.6535376310348511,
      "step": 3744
    },
    {
      "epoch": 1.580590717299578,
      "grad_norm": 1.2132076025009155,
      "learning_rate": 9.051708916587491e-05,
      "loss": 0.6236510872840881,
      "step": 3746
    },
    {
      "epoch": 1.5814345991561183,
      "grad_norm": 1.201319932937622,
      "learning_rate": 9.050301294611885e-05,
      "loss": 0.6752219200134277,
      "step": 3748
    },
    {
      "epoch": 1.5822784810126582,
      "grad_norm": 1.2969163656234741,
      "learning_rate": 9.048892738309559e-05,
      "loss": 0.7248554825782776,
      "step": 3750
    },
    {
      "epoch": 1.5831223628691982,
      "grad_norm": 1.0721957683563232,
      "learning_rate": 9.047483248005439e-05,
      "loss": 0.6488997340202332,
      "step": 3752
    },
    {
      "epoch": 1.5839662447257385,
      "grad_norm": 0.9988508820533752,
      "learning_rate": 9.046072824024667e-05,
      "loss": 0.6191130876541138,
      "step": 3754
    },
    {
      "epoch": 1.5848101265822785,
      "grad_norm": 1.260183572769165,
      "learning_rate": 9.0446614666926e-05,
      "loss": 0.6681985259056091,
      "step": 3756
    },
    {
      "epoch": 1.5856540084388184,
      "grad_norm": 1.1288834810256958,
      "learning_rate": 9.043249176334812e-05,
      "loss": 0.662024736404419,
      "step": 3758
    },
    {
      "epoch": 1.5864978902953588,
      "grad_norm": 1.4384263753890991,
      "learning_rate": 9.04183595327709e-05,
      "loss": 0.609916627407074,
      "step": 3760
    },
    {
      "epoch": 1.5873417721518988,
      "grad_norm": 1.1109941005706787,
      "learning_rate": 9.04042179784544e-05,
      "loss": 0.6532528400421143,
      "step": 3762
    },
    {
      "epoch": 1.5881856540084387,
      "grad_norm": 1.0959233045578003,
      "learning_rate": 9.039006710366078e-05,
      "loss": 0.7136290669441223,
      "step": 3764
    },
    {
      "epoch": 1.5890295358649789,
      "grad_norm": 1.2313964366912842,
      "learning_rate": 9.037590691165439e-05,
      "loss": 0.6907190084457397,
      "step": 3766
    },
    {
      "epoch": 1.589873417721519,
      "grad_norm": 1.3127682209014893,
      "learning_rate": 9.036173740570172e-05,
      "loss": 0.7114790678024292,
      "step": 3768
    },
    {
      "epoch": 1.590717299578059,
      "grad_norm": 1.0038903951644897,
      "learning_rate": 9.034755858907138e-05,
      "loss": 0.6257581114768982,
      "step": 3770
    },
    {
      "epoch": 1.5915611814345991,
      "grad_norm": 1.1058061122894287,
      "learning_rate": 9.033337046503416e-05,
      "loss": 0.578145444393158,
      "step": 3772
    },
    {
      "epoch": 1.5924050632911393,
      "grad_norm": 1.0893515348434448,
      "learning_rate": 9.0319173036863e-05,
      "loss": 0.6312620043754578,
      "step": 3774
    },
    {
      "epoch": 1.5932489451476792,
      "grad_norm": 1.1091047525405884,
      "learning_rate": 9.030496630783297e-05,
      "loss": 0.6799508333206177,
      "step": 3776
    },
    {
      "epoch": 1.5940928270042194,
      "grad_norm": 1.1103609800338745,
      "learning_rate": 9.029075028122127e-05,
      "loss": 0.678726315498352,
      "step": 3778
    },
    {
      "epoch": 1.5949367088607596,
      "grad_norm": 1.1918376684188843,
      "learning_rate": 9.027652496030728e-05,
      "loss": 0.7357890009880066,
      "step": 3780
    },
    {
      "epoch": 1.5957805907172995,
      "grad_norm": 1.0541924238204956,
      "learning_rate": 9.026229034837253e-05,
      "loss": 0.6079391241073608,
      "step": 3782
    },
    {
      "epoch": 1.5966244725738397,
      "grad_norm": 1.195845603942871,
      "learning_rate": 9.024804644870062e-05,
      "loss": 0.7173702120780945,
      "step": 3784
    },
    {
      "epoch": 1.5974683544303798,
      "grad_norm": 1.1362866163253784,
      "learning_rate": 9.023379326457737e-05,
      "loss": 0.6431670188903809,
      "step": 3786
    },
    {
      "epoch": 1.5983122362869198,
      "grad_norm": 1.2327499389648438,
      "learning_rate": 9.021953079929074e-05,
      "loss": 0.6346777677536011,
      "step": 3788
    },
    {
      "epoch": 1.59915611814346,
      "grad_norm": 1.1623177528381348,
      "learning_rate": 9.020525905613078e-05,
      "loss": 0.6852784156799316,
      "step": 3790
    },
    {
      "epoch": 1.6,
      "grad_norm": 1.0258424282073975,
      "learning_rate": 9.019097803838971e-05,
      "loss": 0.6357095241546631,
      "step": 3792
    },
    {
      "epoch": 1.60084388185654,
      "grad_norm": 1.0825177431106567,
      "learning_rate": 9.017668774936188e-05,
      "loss": 0.6663659811019897,
      "step": 3794
    },
    {
      "epoch": 1.6016877637130802,
      "grad_norm": 1.1190401315689087,
      "learning_rate": 9.016238819234381e-05,
      "loss": 0.6009758710861206,
      "step": 3796
    },
    {
      "epoch": 1.6025316455696204,
      "grad_norm": 1.09871244430542,
      "learning_rate": 9.01480793706341e-05,
      "loss": 0.6907890439033508,
      "step": 3798
    },
    {
      "epoch": 1.6033755274261603,
      "grad_norm": 1.2046958208084106,
      "learning_rate": 9.013376128753354e-05,
      "loss": 0.6709389090538025,
      "step": 3800
    },
    {
      "epoch": 1.6033755274261603,
      "eval_loss": 0.7080941200256348,
      "eval_runtime": 865.6774,
      "eval_samples_per_second": 2.434,
      "eval_steps_per_second": 2.434,
      "step": 3800
    },
    {
      "epoch": 1.6042194092827005,
      "grad_norm": 1.0671489238739014,
      "learning_rate": 9.011943394634505e-05,
      "loss": 0.653937041759491,
      "step": 3802
    },
    {
      "epoch": 1.6050632911392406,
      "grad_norm": 1.4205375909805298,
      "learning_rate": 9.010509735037364e-05,
      "loss": 0.6647229194641113,
      "step": 3804
    },
    {
      "epoch": 1.6059071729957806,
      "grad_norm": 1.3793799877166748,
      "learning_rate": 9.009075150292652e-05,
      "loss": 0.6981267929077148,
      "step": 3806
    },
    {
      "epoch": 1.6067510548523207,
      "grad_norm": 1.0534380674362183,
      "learning_rate": 9.007639640731298e-05,
      "loss": 0.6151314973831177,
      "step": 3808
    },
    {
      "epoch": 1.6075949367088609,
      "grad_norm": 1.1359853744506836,
      "learning_rate": 9.006203206684447e-05,
      "loss": 0.6671237349510193,
      "step": 3810
    },
    {
      "epoch": 1.6084388185654008,
      "grad_norm": 1.2385475635528564,
      "learning_rate": 9.004765848483456e-05,
      "loss": 0.7145646810531616,
      "step": 3812
    },
    {
      "epoch": 1.6092827004219408,
      "grad_norm": 1.1323930025100708,
      "learning_rate": 9.003327566459899e-05,
      "loss": 0.6524789929389954,
      "step": 3814
    },
    {
      "epoch": 1.6101265822784812,
      "grad_norm": 1.1863508224487305,
      "learning_rate": 9.001888360945555e-05,
      "loss": 0.7574670314788818,
      "step": 3816
    },
    {
      "epoch": 1.610970464135021,
      "grad_norm": 1.0288994312286377,
      "learning_rate": 9.000448232272425e-05,
      "loss": 0.5858811736106873,
      "step": 3818
    },
    {
      "epoch": 1.611814345991561,
      "grad_norm": 1.2674148082733154,
      "learning_rate": 8.999007180772719e-05,
      "loss": 0.6834250688552856,
      "step": 3820
    },
    {
      "epoch": 1.6126582278481014,
      "grad_norm": 1.2014318704605103,
      "learning_rate": 8.997565206778856e-05,
      "loss": 0.6435309052467346,
      "step": 3822
    },
    {
      "epoch": 1.6135021097046414,
      "grad_norm": 1.205741286277771,
      "learning_rate": 8.996122310623476e-05,
      "loss": 0.6212471127510071,
      "step": 3824
    },
    {
      "epoch": 1.6143459915611813,
      "grad_norm": 1.0866186618804932,
      "learning_rate": 8.994678492639426e-05,
      "loss": 0.6832143664360046,
      "step": 3826
    },
    {
      "epoch": 1.6151898734177215,
      "grad_norm": 1.0786924362182617,
      "learning_rate": 8.993233753159768e-05,
      "loss": 0.6129988431930542,
      "step": 3828
    },
    {
      "epoch": 1.6160337552742616,
      "grad_norm": 1.176597237586975,
      "learning_rate": 8.991788092517775e-05,
      "loss": 0.6376019716262817,
      "step": 3830
    },
    {
      "epoch": 1.6168776371308016,
      "grad_norm": 1.149990200996399,
      "learning_rate": 8.99034151104693e-05,
      "loss": 0.7300569415092468,
      "step": 3832
    },
    {
      "epoch": 1.6177215189873417,
      "grad_norm": 1.0655301809310913,
      "learning_rate": 8.988894009080936e-05,
      "loss": 0.6163336634635925,
      "step": 3834
    },
    {
      "epoch": 1.618565400843882,
      "grad_norm": 1.1596909761428833,
      "learning_rate": 8.987445586953703e-05,
      "loss": 0.6459008455276489,
      "step": 3836
    },
    {
      "epoch": 1.6194092827004218,
      "grad_norm": 1.201897382736206,
      "learning_rate": 8.985996244999352e-05,
      "loss": 0.6166399121284485,
      "step": 3838
    },
    {
      "epoch": 1.620253164556962,
      "grad_norm": 1.1000950336456299,
      "learning_rate": 8.984545983552219e-05,
      "loss": 0.6438087224960327,
      "step": 3840
    },
    {
      "epoch": 1.6210970464135022,
      "grad_norm": 0.9962409734725952,
      "learning_rate": 8.983094802946854e-05,
      "loss": 0.6238043308258057,
      "step": 3842
    },
    {
      "epoch": 1.621940928270042,
      "grad_norm": 1.2501682043075562,
      "learning_rate": 8.981642703518015e-05,
      "loss": 0.6445946097373962,
      "step": 3844
    },
    {
      "epoch": 1.6227848101265823,
      "grad_norm": 1.2027913331985474,
      "learning_rate": 8.980189685600673e-05,
      "loss": 0.7147613167762756,
      "step": 3846
    },
    {
      "epoch": 1.6236286919831224,
      "grad_norm": 1.1382197141647339,
      "learning_rate": 8.97873574953001e-05,
      "loss": 0.6531714200973511,
      "step": 3848
    },
    {
      "epoch": 1.6244725738396624,
      "grad_norm": 1.2600723505020142,
      "learning_rate": 8.977280895641425e-05,
      "loss": 0.6811055541038513,
      "step": 3850
    },
    {
      "epoch": 1.6253164556962025,
      "grad_norm": 0.9908071160316467,
      "learning_rate": 8.97582512427052e-05,
      "loss": 0.6142261624336243,
      "step": 3852
    },
    {
      "epoch": 1.6261603375527427,
      "grad_norm": 1.171557068824768,
      "learning_rate": 8.974368435753117e-05,
      "loss": 0.6408987045288086,
      "step": 3854
    },
    {
      "epoch": 1.6270042194092826,
      "grad_norm": 1.1839419603347778,
      "learning_rate": 8.972910830425247e-05,
      "loss": 0.7352069616317749,
      "step": 3856
    },
    {
      "epoch": 1.6278481012658228,
      "grad_norm": 1.233730673789978,
      "learning_rate": 8.971452308623148e-05,
      "loss": 0.7663040161132812,
      "step": 3858
    },
    {
      "epoch": 1.628691983122363,
      "grad_norm": 1.3636224269866943,
      "learning_rate": 8.969992870683273e-05,
      "loss": 0.6496971249580383,
      "step": 3860
    },
    {
      "epoch": 1.629535864978903,
      "grad_norm": 1.2819573879241943,
      "learning_rate": 8.96853251694229e-05,
      "loss": 0.6079609394073486,
      "step": 3862
    },
    {
      "epoch": 1.630379746835443,
      "grad_norm": 1.087265968322754,
      "learning_rate": 8.967071247737071e-05,
      "loss": 0.6299422979354858,
      "step": 3864
    },
    {
      "epoch": 1.6312236286919832,
      "grad_norm": 1.24200439453125,
      "learning_rate": 8.965609063404706e-05,
      "loss": 0.6691840291023254,
      "step": 3866
    },
    {
      "epoch": 1.6320675105485232,
      "grad_norm": 1.0771806240081787,
      "learning_rate": 8.96414596428249e-05,
      "loss": 0.6623613238334656,
      "step": 3868
    },
    {
      "epoch": 1.6329113924050633,
      "grad_norm": 1.1830974817276,
      "learning_rate": 8.962681950707932e-05,
      "loss": 0.6663276553153992,
      "step": 3870
    },
    {
      "epoch": 1.6337552742616035,
      "grad_norm": 1.1107177734375,
      "learning_rate": 8.961217023018754e-05,
      "loss": 0.6426810622215271,
      "step": 3872
    },
    {
      "epoch": 1.6345991561181434,
      "grad_norm": 1.2528507709503174,
      "learning_rate": 8.959751181552886e-05,
      "loss": 0.7113696336746216,
      "step": 3874
    },
    {
      "epoch": 1.6354430379746834,
      "grad_norm": 1.0656070709228516,
      "learning_rate": 8.958284426648467e-05,
      "loss": 0.6211581230163574,
      "step": 3876
    },
    {
      "epoch": 1.6362869198312238,
      "grad_norm": 1.0627381801605225,
      "learning_rate": 8.956816758643852e-05,
      "loss": 0.5950066447257996,
      "step": 3878
    },
    {
      "epoch": 1.6371308016877637,
      "grad_norm": 0.9812912344932556,
      "learning_rate": 8.955348177877603e-05,
      "loss": 0.6519815325737,
      "step": 3880
    },
    {
      "epoch": 1.6379746835443036,
      "grad_norm": 1.1843842267990112,
      "learning_rate": 8.953878684688493e-05,
      "loss": 0.6830767393112183,
      "step": 3882
    },
    {
      "epoch": 1.638818565400844,
      "grad_norm": 1.0393236875534058,
      "learning_rate": 8.952408279415507e-05,
      "loss": 0.5920302271842957,
      "step": 3884
    },
    {
      "epoch": 1.639662447257384,
      "grad_norm": 0.9931944608688354,
      "learning_rate": 8.950936962397838e-05,
      "loss": 0.6269177198410034,
      "step": 3886
    },
    {
      "epoch": 1.640506329113924,
      "grad_norm": 1.1461358070373535,
      "learning_rate": 8.949464733974891e-05,
      "loss": 0.7021532654762268,
      "step": 3888
    },
    {
      "epoch": 1.6413502109704643,
      "grad_norm": 1.2654093503952026,
      "learning_rate": 8.947991594486279e-05,
      "loss": 0.7331246733665466,
      "step": 3890
    },
    {
      "epoch": 1.6421940928270042,
      "grad_norm": 1.1487081050872803,
      "learning_rate": 8.946517544271831e-05,
      "loss": 0.6438513994216919,
      "step": 3892
    },
    {
      "epoch": 1.6430379746835442,
      "grad_norm": 1.0876784324645996,
      "learning_rate": 8.945042583671579e-05,
      "loss": 0.6779276728630066,
      "step": 3894
    },
    {
      "epoch": 1.6438818565400843,
      "grad_norm": 1.2382020950317383,
      "learning_rate": 8.943566713025768e-05,
      "loss": 0.7255419492721558,
      "step": 3896
    },
    {
      "epoch": 1.6447257383966245,
      "grad_norm": 1.3502718210220337,
      "learning_rate": 8.942089932674855e-05,
      "loss": 0.7068934440612793,
      "step": 3898
    },
    {
      "epoch": 1.6455696202531644,
      "grad_norm": 1.050878643989563,
      "learning_rate": 8.940612242959503e-05,
      "loss": 0.608700156211853,
      "step": 3900
    },
    {
      "epoch": 1.6455696202531644,
      "eval_loss": 0.7049403786659241,
      "eval_runtime": 854.9866,
      "eval_samples_per_second": 2.464,
      "eval_steps_per_second": 2.464,
      "step": 3900
    },
    {
      "epoch": 1.6464135021097046,
      "grad_norm": 1.0536954402923584,
      "learning_rate": 8.939133644220588e-05,
      "loss": 0.6257222890853882,
      "step": 3902
    },
    {
      "epoch": 1.6472573839662448,
      "grad_norm": 1.1903947591781616,
      "learning_rate": 8.937654136799195e-05,
      "loss": 0.6823404431343079,
      "step": 3904
    },
    {
      "epoch": 1.6481012658227847,
      "grad_norm": 1.225679874420166,
      "learning_rate": 8.936173721036616e-05,
      "loss": 0.6596478819847107,
      "step": 3906
    },
    {
      "epoch": 1.6489451476793249,
      "grad_norm": 1.0071430206298828,
      "learning_rate": 8.934692397274354e-05,
      "loss": 0.5638422966003418,
      "step": 3908
    },
    {
      "epoch": 1.649789029535865,
      "grad_norm": 1.0146223306655884,
      "learning_rate": 8.933210165854125e-05,
      "loss": 0.5743419528007507,
      "step": 3910
    },
    {
      "epoch": 1.650632911392405,
      "grad_norm": 1.122976541519165,
      "learning_rate": 8.931727027117848e-05,
      "loss": 0.6775169372558594,
      "step": 3912
    },
    {
      "epoch": 1.6514767932489451,
      "grad_norm": 0.9223271012306213,
      "learning_rate": 8.930242981407656e-05,
      "loss": 0.5984215140342712,
      "step": 3914
    },
    {
      "epoch": 1.6523206751054853,
      "grad_norm": 1.1599735021591187,
      "learning_rate": 8.928758029065891e-05,
      "loss": 0.6342158913612366,
      "step": 3916
    },
    {
      "epoch": 1.6531645569620252,
      "grad_norm": 1.2680121660232544,
      "learning_rate": 8.927272170435101e-05,
      "loss": 0.678507924079895,
      "step": 3918
    },
    {
      "epoch": 1.6540084388185654,
      "grad_norm": 1.3628549575805664,
      "learning_rate": 8.925785405858047e-05,
      "loss": 0.6739710569381714,
      "step": 3920
    },
    {
      "epoch": 1.6548523206751056,
      "grad_norm": 1.163482427597046,
      "learning_rate": 8.924297735677694e-05,
      "loss": 0.7050020098686218,
      "step": 3922
    },
    {
      "epoch": 1.6556962025316455,
      "grad_norm": 1.2057000398635864,
      "learning_rate": 8.922809160237222e-05,
      "loss": 0.6847540140151978,
      "step": 3924
    },
    {
      "epoch": 1.6565400843881857,
      "grad_norm": 1.2784082889556885,
      "learning_rate": 8.921319679880016e-05,
      "loss": 0.7079069018363953,
      "step": 3926
    },
    {
      "epoch": 1.6573839662447258,
      "grad_norm": 1.1701157093048096,
      "learning_rate": 8.919829294949671e-05,
      "loss": 0.665060818195343,
      "step": 3928
    },
    {
      "epoch": 1.6582278481012658,
      "grad_norm": 1.3886606693267822,
      "learning_rate": 8.918338005789988e-05,
      "loss": 0.7547550201416016,
      "step": 3930
    },
    {
      "epoch": 1.659071729957806,
      "grad_norm": 0.9504727721214294,
      "learning_rate": 8.91684581274498e-05,
      "loss": 0.5718522667884827,
      "step": 3932
    },
    {
      "epoch": 1.659915611814346,
      "grad_norm": 1.1185030937194824,
      "learning_rate": 8.915352716158869e-05,
      "loss": 0.5984254479408264,
      "step": 3934
    },
    {
      "epoch": 1.660759493670886,
      "grad_norm": 1.1489602327346802,
      "learning_rate": 8.913858716376081e-05,
      "loss": 0.6749780774116516,
      "step": 3936
    },
    {
      "epoch": 1.6616033755274262,
      "grad_norm": 1.389431118965149,
      "learning_rate": 8.912363813741255e-05,
      "loss": 0.6537864804267883,
      "step": 3938
    },
    {
      "epoch": 1.6624472573839664,
      "grad_norm": 1.0958757400512695,
      "learning_rate": 8.910868008599235e-05,
      "loss": 0.6033569574356079,
      "step": 3940
    },
    {
      "epoch": 1.6632911392405063,
      "grad_norm": 1.2735344171524048,
      "learning_rate": 8.909371301295075e-05,
      "loss": 0.7404987215995789,
      "step": 3942
    },
    {
      "epoch": 1.6641350210970463,
      "grad_norm": 1.123336911201477,
      "learning_rate": 8.907873692174038e-05,
      "loss": 0.6265006065368652,
      "step": 3944
    },
    {
      "epoch": 1.6649789029535866,
      "grad_norm": 1.259470820426941,
      "learning_rate": 8.90637518158159e-05,
      "loss": 0.650705099105835,
      "step": 3946
    },
    {
      "epoch": 1.6658227848101266,
      "grad_norm": 1.4020485877990723,
      "learning_rate": 8.904875769863412e-05,
      "loss": 0.7813970446586609,
      "step": 3948
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 1.1709671020507812,
      "learning_rate": 8.903375457365389e-05,
      "loss": 0.6499447822570801,
      "step": 3950
    },
    {
      "epoch": 1.667510548523207,
      "grad_norm": 1.085585355758667,
      "learning_rate": 8.901874244433612e-05,
      "loss": 0.6141875386238098,
      "step": 3952
    },
    {
      "epoch": 1.6683544303797468,
      "grad_norm": 1.2340166568756104,
      "learning_rate": 8.900372131414386e-05,
      "loss": 0.7080221176147461,
      "step": 3954
    },
    {
      "epoch": 1.6691983122362868,
      "grad_norm": 1.148576259613037,
      "learning_rate": 8.898869118654216e-05,
      "loss": 0.6340513229370117,
      "step": 3956
    },
    {
      "epoch": 1.6700421940928272,
      "grad_norm": 1.2231999635696411,
      "learning_rate": 8.89736520649982e-05,
      "loss": 0.6999116539955139,
      "step": 3958
    },
    {
      "epoch": 1.6708860759493671,
      "grad_norm": 1.1600396633148193,
      "learning_rate": 8.895860395298121e-05,
      "loss": 0.7177759408950806,
      "step": 3960
    },
    {
      "epoch": 1.671729957805907,
      "grad_norm": 1.3019158840179443,
      "learning_rate": 8.894354685396251e-05,
      "loss": 0.6485702395439148,
      "step": 3962
    },
    {
      "epoch": 1.6725738396624472,
      "grad_norm": 1.0153226852416992,
      "learning_rate": 8.892848077141546e-05,
      "loss": 0.6189450025558472,
      "step": 3964
    },
    {
      "epoch": 1.6734177215189874,
      "grad_norm": 1.1953094005584717,
      "learning_rate": 8.891340570881555e-05,
      "loss": 0.6756728291511536,
      "step": 3966
    },
    {
      "epoch": 1.6742616033755273,
      "grad_norm": 1.3376187086105347,
      "learning_rate": 8.889832166964027e-05,
      "loss": 0.6851167678833008,
      "step": 3968
    },
    {
      "epoch": 1.6751054852320675,
      "grad_norm": 1.0045926570892334,
      "learning_rate": 8.888322865736924e-05,
      "loss": 0.5991915464401245,
      "step": 3970
    },
    {
      "epoch": 1.6759493670886076,
      "grad_norm": 1.2115750312805176,
      "learning_rate": 8.886812667548414e-05,
      "loss": 0.713362455368042,
      "step": 3972
    },
    {
      "epoch": 1.6767932489451476,
      "grad_norm": 1.1887929439544678,
      "learning_rate": 8.88530157274687e-05,
      "loss": 0.7058883309364319,
      "step": 3974
    },
    {
      "epoch": 1.6776371308016877,
      "grad_norm": 1.1465295553207397,
      "learning_rate": 8.883789581680868e-05,
      "loss": 0.6501380801200867,
      "step": 3976
    },
    {
      "epoch": 1.678481012658228,
      "grad_norm": 1.184693694114685,
      "learning_rate": 8.882276694699204e-05,
      "loss": 0.6109840273857117,
      "step": 3978
    },
    {
      "epoch": 1.6793248945147679,
      "grad_norm": 1.2034777402877808,
      "learning_rate": 8.880762912150862e-05,
      "loss": 0.6815584897994995,
      "step": 3980
    },
    {
      "epoch": 1.680168776371308,
      "grad_norm": 1.1312000751495361,
      "learning_rate": 8.879248234385052e-05,
      "loss": 0.6859248876571655,
      "step": 3982
    },
    {
      "epoch": 1.6810126582278482,
      "grad_norm": 1.2273681163787842,
      "learning_rate": 8.877732661751173e-05,
      "loss": 0.6426702737808228,
      "step": 3984
    },
    {
      "epoch": 1.6818565400843881,
      "grad_norm": 1.2550326585769653,
      "learning_rate": 8.876216194598844e-05,
      "loss": 0.6462456583976746,
      "step": 3986
    },
    {
      "epoch": 1.6827004219409283,
      "grad_norm": 1.3111321926116943,
      "learning_rate": 8.874698833277884e-05,
      "loss": 0.6293925046920776,
      "step": 3988
    },
    {
      "epoch": 1.6835443037974684,
      "grad_norm": 1.037883996963501,
      "learning_rate": 8.873180578138316e-05,
      "loss": 0.59798264503479,
      "step": 3990
    },
    {
      "epoch": 1.6843881856540084,
      "grad_norm": 1.2411901950836182,
      "learning_rate": 8.871661429530376e-05,
      "loss": 0.6741529703140259,
      "step": 3992
    },
    {
      "epoch": 1.6852320675105485,
      "grad_norm": 1.206354022026062,
      "learning_rate": 8.8701413878045e-05,
      "loss": 0.5972680449485779,
      "step": 3994
    },
    {
      "epoch": 1.6860759493670887,
      "grad_norm": 1.1922144889831543,
      "learning_rate": 8.868620453311334e-05,
      "loss": 0.5879245400428772,
      "step": 3996
    },
    {
      "epoch": 1.6869198312236287,
      "grad_norm": 1.3499996662139893,
      "learning_rate": 8.867098626401729e-05,
      "loss": 0.7381167411804199,
      "step": 3998
    },
    {
      "epoch": 1.6877637130801688,
      "grad_norm": 1.3601514101028442,
      "learning_rate": 8.865575907426737e-05,
      "loss": 0.6590276956558228,
      "step": 4000
    },
    {
      "epoch": 1.6877637130801688,
      "eval_loss": 0.7027890682220459,
      "eval_runtime": 848.7529,
      "eval_samples_per_second": 2.482,
      "eval_steps_per_second": 2.482,
      "step": 4000
    },
    {
      "epoch": 1.688607594936709,
      "grad_norm": 1.1060529947280884,
      "learning_rate": 8.864052296737624e-05,
      "loss": 0.5958077907562256,
      "step": 4002
    },
    {
      "epoch": 1.689451476793249,
      "grad_norm": 1.2067371606826782,
      "learning_rate": 8.862527794685858e-05,
      "loss": 0.6802279353141785,
      "step": 4004
    },
    {
      "epoch": 1.690295358649789,
      "grad_norm": 1.0094636678695679,
      "learning_rate": 8.86100240162311e-05,
      "loss": 0.5701603889465332,
      "step": 4006
    },
    {
      "epoch": 1.6911392405063292,
      "grad_norm": 1.0976500511169434,
      "learning_rate": 8.85947611790126e-05,
      "loss": 0.6580625176429749,
      "step": 4008
    },
    {
      "epoch": 1.6919831223628692,
      "grad_norm": 0.9448981285095215,
      "learning_rate": 8.857948943872392e-05,
      "loss": 0.5947542190551758,
      "step": 4010
    },
    {
      "epoch": 1.6928270042194091,
      "grad_norm": 1.219609260559082,
      "learning_rate": 8.856420879888796e-05,
      "loss": 0.6361464262008667,
      "step": 4012
    },
    {
      "epoch": 1.6936708860759495,
      "grad_norm": 1.2395503520965576,
      "learning_rate": 8.854891926302966e-05,
      "loss": 0.608664333820343,
      "step": 4014
    },
    {
      "epoch": 1.6945147679324895,
      "grad_norm": 1.1300057172775269,
      "learning_rate": 8.853362083467604e-05,
      "loss": 0.6932460069656372,
      "step": 4016
    },
    {
      "epoch": 1.6953586497890294,
      "grad_norm": 1.2300254106521606,
      "learning_rate": 8.851831351735616e-05,
      "loss": 0.646004855632782,
      "step": 4018
    },
    {
      "epoch": 1.6962025316455698,
      "grad_norm": 1.2328956127166748,
      "learning_rate": 8.85029973146011e-05,
      "loss": 0.6760826110839844,
      "step": 4020
    },
    {
      "epoch": 1.6970464135021097,
      "grad_norm": 1.1252286434173584,
      "learning_rate": 8.848767222994401e-05,
      "loss": 0.5943224430084229,
      "step": 4022
    },
    {
      "epoch": 1.6978902953586497,
      "grad_norm": 1.1587592363357544,
      "learning_rate": 8.847233826692012e-05,
      "loss": 0.7535276412963867,
      "step": 4024
    },
    {
      "epoch": 1.6987341772151898,
      "grad_norm": 1.0294606685638428,
      "learning_rate": 8.845699542906667e-05,
      "loss": 0.5903090834617615,
      "step": 4026
    },
    {
      "epoch": 1.69957805907173,
      "grad_norm": 1.1940597295761108,
      "learning_rate": 8.844164371992295e-05,
      "loss": 0.6031379699707031,
      "step": 4028
    },
    {
      "epoch": 1.70042194092827,
      "grad_norm": 1.0416409969329834,
      "learning_rate": 8.842628314303031e-05,
      "loss": 0.6185168623924255,
      "step": 4030
    },
    {
      "epoch": 1.70126582278481,
      "grad_norm": 1.8715689182281494,
      "learning_rate": 8.841091370193214e-05,
      "loss": 0.6325570344924927,
      "step": 4032
    },
    {
      "epoch": 1.7021097046413503,
      "grad_norm": 1.230658769607544,
      "learning_rate": 8.839553540017387e-05,
      "loss": 0.7413952350616455,
      "step": 4034
    },
    {
      "epoch": 1.7029535864978902,
      "grad_norm": 1.298003077507019,
      "learning_rate": 8.838014824130299e-05,
      "loss": 0.6973189115524292,
      "step": 4036
    },
    {
      "epoch": 1.7037974683544304,
      "grad_norm": 1.0246652364730835,
      "learning_rate": 8.836475222886902e-05,
      "loss": 0.6582493185997009,
      "step": 4038
    },
    {
      "epoch": 1.7046413502109705,
      "grad_norm": 1.3652594089508057,
      "learning_rate": 8.834934736642351e-05,
      "loss": 0.6934399008750916,
      "step": 4040
    },
    {
      "epoch": 1.7054852320675105,
      "grad_norm": 1.029778242111206,
      "learning_rate": 8.833393365752007e-05,
      "loss": 0.6437561511993408,
      "step": 4042
    },
    {
      "epoch": 1.7063291139240506,
      "grad_norm": 1.1993004083633423,
      "learning_rate": 8.831851110571437e-05,
      "loss": 0.605059027671814,
      "step": 4044
    },
    {
      "epoch": 1.7071729957805908,
      "grad_norm": 1.286389946937561,
      "learning_rate": 8.830307971456406e-05,
      "loss": 0.7035017609596252,
      "step": 4046
    },
    {
      "epoch": 1.7080168776371307,
      "grad_norm": 1.1211459636688232,
      "learning_rate": 8.82876394876289e-05,
      "loss": 0.6429924964904785,
      "step": 4048
    },
    {
      "epoch": 1.7088607594936709,
      "grad_norm": 1.1284868717193604,
      "learning_rate": 8.827219042847064e-05,
      "loss": 0.6454769968986511,
      "step": 4050
    },
    {
      "epoch": 1.709704641350211,
      "grad_norm": 1.1934884786605835,
      "learning_rate": 8.825673254065306e-05,
      "loss": 0.707233190536499,
      "step": 4052
    },
    {
      "epoch": 1.710548523206751,
      "grad_norm": 1.1560680866241455,
      "learning_rate": 8.824126582774203e-05,
      "loss": 0.6790444254875183,
      "step": 4054
    },
    {
      "epoch": 1.7113924050632912,
      "grad_norm": 1.1924364566802979,
      "learning_rate": 8.822579029330541e-05,
      "loss": 0.6115295886993408,
      "step": 4056
    },
    {
      "epoch": 1.7122362869198313,
      "grad_norm": 1.107370138168335,
      "learning_rate": 8.82103059409131e-05,
      "loss": 0.7039182186126709,
      "step": 4058
    },
    {
      "epoch": 1.7130801687763713,
      "grad_norm": 1.2554657459259033,
      "learning_rate": 8.819481277413707e-05,
      "loss": 0.6580052971839905,
      "step": 4060
    },
    {
      "epoch": 1.7139240506329114,
      "grad_norm": 1.2873135805130005,
      "learning_rate": 8.817931079655127e-05,
      "loss": 0.6042479276657104,
      "step": 4062
    },
    {
      "epoch": 1.7147679324894516,
      "grad_norm": 1.027056097984314,
      "learning_rate": 8.816380001173172e-05,
      "loss": 0.5992372632026672,
      "step": 4064
    },
    {
      "epoch": 1.7156118143459915,
      "grad_norm": 1.0694721937179565,
      "learning_rate": 8.814828042325644e-05,
      "loss": 0.7078655362129211,
      "step": 4066
    },
    {
      "epoch": 1.7164556962025317,
      "grad_norm": 1.194984793663025,
      "learning_rate": 8.813275203470555e-05,
      "loss": 0.6618752479553223,
      "step": 4068
    },
    {
      "epoch": 1.7172995780590719,
      "grad_norm": 1.1713165044784546,
      "learning_rate": 8.811721484966109e-05,
      "loss": 0.6328625679016113,
      "step": 4070
    },
    {
      "epoch": 1.7181434599156118,
      "grad_norm": 0.9993656277656555,
      "learning_rate": 8.810166887170724e-05,
      "loss": 0.5916416645050049,
      "step": 4072
    },
    {
      "epoch": 1.7189873417721517,
      "grad_norm": 1.172642707824707,
      "learning_rate": 8.808611410443011e-05,
      "loss": 0.6490002274513245,
      "step": 4074
    },
    {
      "epoch": 1.7198312236286921,
      "grad_norm": 1.1404821872711182,
      "learning_rate": 8.807055055141793e-05,
      "loss": 0.6571791172027588,
      "step": 4076
    },
    {
      "epoch": 1.720675105485232,
      "grad_norm": 1.2104214429855347,
      "learning_rate": 8.80549782162609e-05,
      "loss": 0.6233854293823242,
      "step": 4078
    },
    {
      "epoch": 1.721518987341772,
      "grad_norm": 1.1691396236419678,
      "learning_rate": 8.803939710255126e-05,
      "loss": 0.6331531405448914,
      "step": 4080
    },
    {
      "epoch": 1.7223628691983124,
      "grad_norm": 1.263174057006836,
      "learning_rate": 8.802380721388325e-05,
      "loss": 0.6321156620979309,
      "step": 4082
    },
    {
      "epoch": 1.7232067510548523,
      "grad_norm": 1.0685606002807617,
      "learning_rate": 8.80082085538532e-05,
      "loss": 0.644904613494873,
      "step": 4084
    },
    {
      "epoch": 1.7240506329113923,
      "grad_norm": 1.2289735078811646,
      "learning_rate": 8.799260112605938e-05,
      "loss": 0.6743831634521484,
      "step": 4086
    },
    {
      "epoch": 1.7248945147679327,
      "grad_norm": 1.0661355257034302,
      "learning_rate": 8.797698493410216e-05,
      "loss": 0.6866999268531799,
      "step": 4088
    },
    {
      "epoch": 1.7257383966244726,
      "grad_norm": 1.1001228094100952,
      "learning_rate": 8.796135998158386e-05,
      "loss": 0.691387414932251,
      "step": 4090
    },
    {
      "epoch": 1.7265822784810125,
      "grad_norm": 1.1078115701675415,
      "learning_rate": 8.794572627210887e-05,
      "loss": 0.5882864594459534,
      "step": 4092
    },
    {
      "epoch": 1.7274261603375527,
      "grad_norm": 1.0483999252319336,
      "learning_rate": 8.79300838092836e-05,
      "loss": 0.6192089319229126,
      "step": 4094
    },
    {
      "epoch": 1.7282700421940929,
      "grad_norm": 1.1194913387298584,
      "learning_rate": 8.791443259671645e-05,
      "loss": 0.603322446346283,
      "step": 4096
    },
    {
      "epoch": 1.7291139240506328,
      "grad_norm": 1.1800397634506226,
      "learning_rate": 8.789877263801787e-05,
      "loss": 0.6141818165779114,
      "step": 4098
    },
    {
      "epoch": 1.729957805907173,
      "grad_norm": 1.261768102645874,
      "learning_rate": 8.78831039368003e-05,
      "loss": 0.6707983016967773,
      "step": 4100
    },
    {
      "epoch": 1.729957805907173,
      "eval_loss": 0.7022181153297424,
      "eval_runtime": 844.6405,
      "eval_samples_per_second": 2.495,
      "eval_steps_per_second": 2.495,
      "step": 4100
    },
    {
      "epoch": 1.7308016877637131,
      "grad_norm": 1.2505232095718384,
      "learning_rate": 8.786742649667822e-05,
      "loss": 0.6440353989601135,
      "step": 4102
    },
    {
      "epoch": 1.731645569620253,
      "grad_norm": 1.2631809711456299,
      "learning_rate": 8.78517403212681e-05,
      "loss": 0.6712808012962341,
      "step": 4104
    },
    {
      "epoch": 1.7324894514767932,
      "grad_norm": 1.2781071662902832,
      "learning_rate": 8.783604541418845e-05,
      "loss": 0.6854958534240723,
      "step": 4106
    },
    {
      "epoch": 1.7333333333333334,
      "grad_norm": 1.1065936088562012,
      "learning_rate": 8.782034177905976e-05,
      "loss": 0.6281477808952332,
      "step": 4108
    },
    {
      "epoch": 1.7341772151898733,
      "grad_norm": 1.010961890220642,
      "learning_rate": 8.780462941950457e-05,
      "loss": 0.6835165619850159,
      "step": 4110
    },
    {
      "epoch": 1.7350210970464135,
      "grad_norm": 1.1467366218566895,
      "learning_rate": 8.778890833914744e-05,
      "loss": 0.6674962639808655,
      "step": 4112
    },
    {
      "epoch": 1.7358649789029537,
      "grad_norm": 1.0221859216690063,
      "learning_rate": 8.77731785416149e-05,
      "loss": 0.5967551469802856,
      "step": 4114
    },
    {
      "epoch": 1.7367088607594936,
      "grad_norm": 1.347937822341919,
      "learning_rate": 8.775744003053552e-05,
      "loss": 0.7356855869293213,
      "step": 4116
    },
    {
      "epoch": 1.7375527426160338,
      "grad_norm": 1.2952557802200317,
      "learning_rate": 8.774169280953988e-05,
      "loss": 0.6932644844055176,
      "step": 4118
    },
    {
      "epoch": 1.738396624472574,
      "grad_norm": 1.0157089233398438,
      "learning_rate": 8.772593688226052e-05,
      "loss": 0.5917407870292664,
      "step": 4120
    },
    {
      "epoch": 1.7392405063291139,
      "grad_norm": 1.1537878513336182,
      "learning_rate": 8.77101722523321e-05,
      "loss": 0.6335760354995728,
      "step": 4122
    },
    {
      "epoch": 1.740084388185654,
      "grad_norm": 1.0989667177200317,
      "learning_rate": 8.769439892339115e-05,
      "loss": 0.6892110109329224,
      "step": 4124
    },
    {
      "epoch": 1.7409282700421942,
      "grad_norm": 1.1293572187423706,
      "learning_rate": 8.767861689907633e-05,
      "loss": 0.5966230630874634,
      "step": 4126
    },
    {
      "epoch": 1.7417721518987341,
      "grad_norm": 1.1167775392532349,
      "learning_rate": 8.76628261830282e-05,
      "loss": 0.5981804728507996,
      "step": 4128
    },
    {
      "epoch": 1.7426160337552743,
      "grad_norm": 1.0572419166564941,
      "learning_rate": 8.76470267788894e-05,
      "loss": 0.5539529919624329,
      "step": 4130
    },
    {
      "epoch": 1.7434599156118145,
      "grad_norm": 0.937256932258606,
      "learning_rate": 8.763121869030456e-05,
      "loss": 0.6238219141960144,
      "step": 4132
    },
    {
      "epoch": 1.7443037974683544,
      "grad_norm": 1.082932472229004,
      "learning_rate": 8.761540192092029e-05,
      "loss": 0.6033329963684082,
      "step": 4134
    },
    {
      "epoch": 1.7451476793248946,
      "grad_norm": 1.0495184659957886,
      "learning_rate": 8.75995764743852e-05,
      "loss": 0.5567626357078552,
      "step": 4136
    },
    {
      "epoch": 1.7459915611814347,
      "grad_norm": 1.3143779039382935,
      "learning_rate": 8.758374235434994e-05,
      "loss": 0.6759346127510071,
      "step": 4138
    },
    {
      "epoch": 1.7468354430379747,
      "grad_norm": 1.2385786771774292,
      "learning_rate": 8.756789956446713e-05,
      "loss": 0.6439400315284729,
      "step": 4140
    },
    {
      "epoch": 1.7476793248945146,
      "grad_norm": 1.0453747510910034,
      "learning_rate": 8.75520481083914e-05,
      "loss": 0.627493679523468,
      "step": 4142
    },
    {
      "epoch": 1.748523206751055,
      "grad_norm": 1.09946608543396,
      "learning_rate": 8.753618798977935e-05,
      "loss": 0.677209198474884,
      "step": 4144
    },
    {
      "epoch": 1.749367088607595,
      "grad_norm": 1.2207063436508179,
      "learning_rate": 8.752031921228965e-05,
      "loss": 0.6874014735221863,
      "step": 4146
    },
    {
      "epoch": 1.7502109704641349,
      "grad_norm": 1.2520697116851807,
      "learning_rate": 8.750444177958288e-05,
      "loss": 0.6332831382751465,
      "step": 4148
    },
    {
      "epoch": 1.7510548523206753,
      "grad_norm": 1.2463186979293823,
      "learning_rate": 8.748855569532168e-05,
      "loss": 0.682744562625885,
      "step": 4150
    },
    {
      "epoch": 1.7518987341772152,
      "grad_norm": 1.1895235776901245,
      "learning_rate": 8.747266096317069e-05,
      "loss": 0.7006803750991821,
      "step": 4152
    },
    {
      "epoch": 1.7527426160337551,
      "grad_norm": 1.1627185344696045,
      "learning_rate": 8.745675758679646e-05,
      "loss": 0.6751191020011902,
      "step": 4154
    },
    {
      "epoch": 1.7535864978902953,
      "grad_norm": 1.324127197265625,
      "learning_rate": 8.744084556986764e-05,
      "loss": 0.661848247051239,
      "step": 4156
    },
    {
      "epoch": 1.7544303797468355,
      "grad_norm": 1.226809024810791,
      "learning_rate": 8.74249249160548e-05,
      "loss": 0.7057217955589294,
      "step": 4158
    },
    {
      "epoch": 1.7552742616033754,
      "grad_norm": 1.2341214418411255,
      "learning_rate": 8.740899562903056e-05,
      "loss": 0.6856105923652649,
      "step": 4160
    },
    {
      "epoch": 1.7561181434599156,
      "grad_norm": 1.3907564878463745,
      "learning_rate": 8.739305771246946e-05,
      "loss": 0.6616930365562439,
      "step": 4162
    },
    {
      "epoch": 1.7569620253164557,
      "grad_norm": 1.2756825685501099,
      "learning_rate": 8.737711117004812e-05,
      "loss": 0.5791551470756531,
      "step": 4164
    },
    {
      "epoch": 1.7578059071729957,
      "grad_norm": 1.2861095666885376,
      "learning_rate": 8.736115600544506e-05,
      "loss": 0.7074756622314453,
      "step": 4166
    },
    {
      "epoch": 1.7586497890295358,
      "grad_norm": 1.2198424339294434,
      "learning_rate": 8.734519222234083e-05,
      "loss": 0.6494167447090149,
      "step": 4168
    },
    {
      "epoch": 1.759493670886076,
      "grad_norm": 1.19169020652771,
      "learning_rate": 8.732921982441799e-05,
      "loss": 0.6546841859817505,
      "step": 4170
    },
    {
      "epoch": 1.760337552742616,
      "grad_norm": 1.11533784866333,
      "learning_rate": 8.731323881536108e-05,
      "loss": 0.6701815724372864,
      "step": 4172
    },
    {
      "epoch": 1.761181434599156,
      "grad_norm": 1.2148140668869019,
      "learning_rate": 8.729724919885657e-05,
      "loss": 0.6678179502487183,
      "step": 4174
    },
    {
      "epoch": 1.7620253164556963,
      "grad_norm": 1.1968709230422974,
      "learning_rate": 8.728125097859298e-05,
      "loss": 0.6505144834518433,
      "step": 4176
    },
    {
      "epoch": 1.7628691983122362,
      "grad_norm": 1.0954766273498535,
      "learning_rate": 8.726524415826079e-05,
      "loss": 0.6531696915626526,
      "step": 4178
    },
    {
      "epoch": 1.7637130801687764,
      "grad_norm": 1.5149537324905396,
      "learning_rate": 8.724922874155246e-05,
      "loss": 0.710014283657074,
      "step": 4180
    },
    {
      "epoch": 1.7645569620253165,
      "grad_norm": 1.145113229751587,
      "learning_rate": 8.723320473216245e-05,
      "loss": 0.714016318321228,
      "step": 4182
    },
    {
      "epoch": 1.7654008438818565,
      "grad_norm": 0.9454524517059326,
      "learning_rate": 8.721717213378719e-05,
      "loss": 0.6775414347648621,
      "step": 4184
    },
    {
      "epoch": 1.7662447257383966,
      "grad_norm": 1.1414754390716553,
      "learning_rate": 8.720113095012507e-05,
      "loss": 0.6279728412628174,
      "step": 4186
    },
    {
      "epoch": 1.7670886075949368,
      "grad_norm": 1.212802767753601,
      "learning_rate": 8.718508118487652e-05,
      "loss": 0.5894309282302856,
      "step": 4188
    },
    {
      "epoch": 1.7679324894514767,
      "grad_norm": 1.5213478803634644,
      "learning_rate": 8.716902284174388e-05,
      "loss": 0.6124046444892883,
      "step": 4190
    },
    {
      "epoch": 1.768776371308017,
      "grad_norm": 0.9973840713500977,
      "learning_rate": 8.715295592443154e-05,
      "loss": 0.5990801453590393,
      "step": 4192
    },
    {
      "epoch": 1.769620253164557,
      "grad_norm": 1.1084294319152832,
      "learning_rate": 8.713688043664579e-05,
      "loss": 0.6485559344291687,
      "step": 4194
    },
    {
      "epoch": 1.770464135021097,
      "grad_norm": 1.1401913166046143,
      "learning_rate": 8.712079638209493e-05,
      "loss": 0.7083099484443665,
      "step": 4196
    },
    {
      "epoch": 1.7713080168776372,
      "grad_norm": 1.278105616569519,
      "learning_rate": 8.71047037644893e-05,
      "loss": 0.7237915992736816,
      "step": 4198
    },
    {
      "epoch": 1.7721518987341773,
      "grad_norm": 1.2407530546188354,
      "learning_rate": 8.708860258754108e-05,
      "loss": 0.6259870529174805,
      "step": 4200
    },
    {
      "epoch": 1.7721518987341773,
      "eval_loss": 0.6993561387062073,
      "eval_runtime": 542.0281,
      "eval_samples_per_second": 3.887,
      "eval_steps_per_second": 3.887,
      "step": 4200
    },
    {
      "epoch": 1.7729957805907173,
      "grad_norm": 1.102859616279602,
      "learning_rate": 8.707249285496457e-05,
      "loss": 0.6604248285293579,
      "step": 4202
    },
    {
      "epoch": 1.7738396624472574,
      "grad_norm": 1.2478244304656982,
      "learning_rate": 8.705637457047594e-05,
      "loss": 0.6799775958061218,
      "step": 4204
    },
    {
      "epoch": 1.7746835443037976,
      "grad_norm": 1.1178022623062134,
      "learning_rate": 8.704024773779338e-05,
      "loss": 0.6136477589607239,
      "step": 4206
    },
    {
      "epoch": 1.7755274261603375,
      "grad_norm": 1.904076337814331,
      "learning_rate": 8.702411236063703e-05,
      "loss": 0.6568390130996704,
      "step": 4208
    },
    {
      "epoch": 1.7763713080168775,
      "grad_norm": 1.0902835130691528,
      "learning_rate": 8.700796844272903e-05,
      "loss": 0.6404406428337097,
      "step": 4210
    },
    {
      "epoch": 1.7772151898734179,
      "grad_norm": 1.1858288049697876,
      "learning_rate": 8.699181598779347e-05,
      "loss": 0.6924911737442017,
      "step": 4212
    },
    {
      "epoch": 1.7780590717299578,
      "grad_norm": 1.0015727281570435,
      "learning_rate": 8.69756549995564e-05,
      "loss": 0.572692334651947,
      "step": 4214
    },
    {
      "epoch": 1.7789029535864977,
      "grad_norm": 1.440079689025879,
      "learning_rate": 8.695948548174583e-05,
      "loss": 0.7196018695831299,
      "step": 4216
    },
    {
      "epoch": 1.7797468354430381,
      "grad_norm": 1.1320992708206177,
      "learning_rate": 8.69433074380918e-05,
      "loss": 0.5870906710624695,
      "step": 4218
    },
    {
      "epoch": 1.780590717299578,
      "grad_norm": 1.3156964778900146,
      "learning_rate": 8.692712087232626e-05,
      "loss": 0.6501539349555969,
      "step": 4220
    },
    {
      "epoch": 1.781434599156118,
      "grad_norm": 1.1869803667068481,
      "learning_rate": 8.691092578818311e-05,
      "loss": 0.7017278075218201,
      "step": 4222
    },
    {
      "epoch": 1.7822784810126582,
      "grad_norm": 0.9708380699157715,
      "learning_rate": 8.689472218939829e-05,
      "loss": 0.5954802632331848,
      "step": 4224
    },
    {
      "epoch": 1.7831223628691983,
      "grad_norm": 1.0753228664398193,
      "learning_rate": 8.687851007970962e-05,
      "loss": 0.6494144797325134,
      "step": 4226
    },
    {
      "epoch": 1.7839662447257383,
      "grad_norm": 1.1038413047790527,
      "learning_rate": 8.686228946285695e-05,
      "loss": 0.7247282862663269,
      "step": 4228
    },
    {
      "epoch": 1.7848101265822784,
      "grad_norm": 0.9666786789894104,
      "learning_rate": 8.684606034258206e-05,
      "loss": 0.5673812627792358,
      "step": 4230
    },
    {
      "epoch": 1.7856540084388186,
      "grad_norm": 1.1972676515579224,
      "learning_rate": 8.682982272262869e-05,
      "loss": 0.5950504541397095,
      "step": 4232
    },
    {
      "epoch": 1.7864978902953585,
      "grad_norm": 1.23736572265625,
      "learning_rate": 8.681357660674255e-05,
      "loss": 0.6477514505386353,
      "step": 4234
    },
    {
      "epoch": 1.7873417721518987,
      "grad_norm": 1.0238158702850342,
      "learning_rate": 8.679732199867127e-05,
      "loss": 0.6180200576782227,
      "step": 4236
    },
    {
      "epoch": 1.7881856540084389,
      "grad_norm": 1.0333375930786133,
      "learning_rate": 8.678105890216455e-05,
      "loss": 0.5771099328994751,
      "step": 4238
    },
    {
      "epoch": 1.7890295358649788,
      "grad_norm": 1.30390202999115,
      "learning_rate": 8.676478732097393e-05,
      "loss": 0.6592516899108887,
      "step": 4240
    },
    {
      "epoch": 1.789873417721519,
      "grad_norm": 1.115160346031189,
      "learning_rate": 8.674850725885294e-05,
      "loss": 0.6662757396697998,
      "step": 4242
    },
    {
      "epoch": 1.7907172995780591,
      "grad_norm": 1.2130142450332642,
      "learning_rate": 8.67322187195571e-05,
      "loss": 0.6673333048820496,
      "step": 4244
    },
    {
      "epoch": 1.791561181434599,
      "grad_norm": 1.1505554914474487,
      "learning_rate": 8.671592170684386e-05,
      "loss": 0.6698325872421265,
      "step": 4246
    },
    {
      "epoch": 1.7924050632911392,
      "grad_norm": 1.0758062601089478,
      "learning_rate": 8.669961622447262e-05,
      "loss": 0.6216199398040771,
      "step": 4248
    },
    {
      "epoch": 1.7932489451476794,
      "grad_norm": 0.9300920367240906,
      "learning_rate": 8.668330227620475e-05,
      "loss": 0.6460495591163635,
      "step": 4250
    },
    {
      "epoch": 1.7940928270042193,
      "grad_norm": 1.3860046863555908,
      "learning_rate": 8.666697986580357e-05,
      "loss": 0.6949506998062134,
      "step": 4252
    },
    {
      "epoch": 1.7949367088607595,
      "grad_norm": 1.2287555932998657,
      "learning_rate": 8.665064899703433e-05,
      "loss": 0.6320405602455139,
      "step": 4254
    },
    {
      "epoch": 1.7957805907172997,
      "grad_norm": 1.1585466861724854,
      "learning_rate": 8.663430967366426e-05,
      "loss": 0.6635019779205322,
      "step": 4256
    },
    {
      "epoch": 1.7966244725738396,
      "grad_norm": 1.1007941961288452,
      "learning_rate": 8.661796189946252e-05,
      "loss": 0.645052969455719,
      "step": 4258
    },
    {
      "epoch": 1.7974683544303798,
      "grad_norm": 1.2059847116470337,
      "learning_rate": 8.660160567820023e-05,
      "loss": 0.70420902967453,
      "step": 4260
    },
    {
      "epoch": 1.79831223628692,
      "grad_norm": 1.0648717880249023,
      "learning_rate": 8.658524101365044e-05,
      "loss": 0.6263765096664429,
      "step": 4262
    },
    {
      "epoch": 1.7991561181434599,
      "grad_norm": 1.017052412033081,
      "learning_rate": 8.656886790958821e-05,
      "loss": 0.6199937462806702,
      "step": 4264
    },
    {
      "epoch": 1.8,
      "grad_norm": 1.1153450012207031,
      "learning_rate": 8.655248636979045e-05,
      "loss": 0.5891271233558655,
      "step": 4266
    },
    {
      "epoch": 1.8008438818565402,
      "grad_norm": 1.0661747455596924,
      "learning_rate": 8.65360963980361e-05,
      "loss": 0.5442121028900146,
      "step": 4268
    },
    {
      "epoch": 1.8016877637130801,
      "grad_norm": 1.3049758672714233,
      "learning_rate": 8.6519697998106e-05,
      "loss": 0.6988245248794556,
      "step": 4270
    },
    {
      "epoch": 1.80253164556962,
      "grad_norm": 1.2679938077926636,
      "learning_rate": 8.650329117378294e-05,
      "loss": 0.7260398864746094,
      "step": 4272
    },
    {
      "epoch": 1.8033755274261605,
      "grad_norm": 1.0899536609649658,
      "learning_rate": 8.648687592885168e-05,
      "loss": 0.5757678151130676,
      "step": 4274
    },
    {
      "epoch": 1.8042194092827004,
      "grad_norm": 1.4088575839996338,
      "learning_rate": 8.647045226709887e-05,
      "loss": 0.7042108178138733,
      "step": 4276
    },
    {
      "epoch": 1.8050632911392404,
      "grad_norm": 1.2143783569335938,
      "learning_rate": 8.645402019231316e-05,
      "loss": 0.641275942325592,
      "step": 4278
    },
    {
      "epoch": 1.8059071729957807,
      "grad_norm": 1.4072896242141724,
      "learning_rate": 8.64375797082851e-05,
      "loss": 0.7657124996185303,
      "step": 4280
    },
    {
      "epoch": 1.8067510548523207,
      "grad_norm": 1.2563380002975464,
      "learning_rate": 8.642113081880718e-05,
      "loss": 0.713768720626831,
      "step": 4282
    },
    {
      "epoch": 1.8075949367088606,
      "grad_norm": 1.1195416450500488,
      "learning_rate": 8.64046735276739e-05,
      "loss": 0.6276429295539856,
      "step": 4284
    },
    {
      "epoch": 1.808438818565401,
      "grad_norm": 1.2472422122955322,
      "learning_rate": 8.638820783868158e-05,
      "loss": 0.5641238689422607,
      "step": 4286
    },
    {
      "epoch": 1.809282700421941,
      "grad_norm": 1.1974313259124756,
      "learning_rate": 8.637173375562855e-05,
      "loss": 0.6312015056610107,
      "step": 4288
    },
    {
      "epoch": 1.810126582278481,
      "grad_norm": 1.1673604249954224,
      "learning_rate": 8.63552512823151e-05,
      "loss": 0.6674410104751587,
      "step": 4290
    },
    {
      "epoch": 1.810970464135021,
      "grad_norm": 1.199095368385315,
      "learning_rate": 8.633876042254337e-05,
      "loss": 0.6772016286849976,
      "step": 4292
    },
    {
      "epoch": 1.8118143459915612,
      "grad_norm": 1.2302746772766113,
      "learning_rate": 8.632226118011752e-05,
      "loss": 0.6621671915054321,
      "step": 4294
    },
    {
      "epoch": 1.8126582278481012,
      "grad_norm": 1.304010033607483,
      "learning_rate": 8.63057535588436e-05,
      "loss": 0.6965363621711731,
      "step": 4296
    },
    {
      "epoch": 1.8135021097046413,
      "grad_norm": 1.223366618156433,
      "learning_rate": 8.62892375625296e-05,
      "loss": 0.6300807595252991,
      "step": 4298
    },
    {
      "epoch": 1.8143459915611815,
      "grad_norm": 1.028496265411377,
      "learning_rate": 8.627271319498544e-05,
      "loss": 0.5610660910606384,
      "step": 4300
    },
    {
      "epoch": 1.8143459915611815,
      "eval_loss": 0.6981000900268555,
      "eval_runtime": 514.4659,
      "eval_samples_per_second": 4.096,
      "eval_steps_per_second": 4.096,
      "step": 4300
    },
    {
      "epoch": 1.8151898734177214,
      "grad_norm": 1.2050007581710815,
      "learning_rate": 8.625618046002298e-05,
      "loss": 0.6666551232337952,
      "step": 4302
    },
    {
      "epoch": 1.8160337552742616,
      "grad_norm": 1.1233220100402832,
      "learning_rate": 8.6239639361456e-05,
      "loss": 0.6631835103034973,
      "step": 4304
    },
    {
      "epoch": 1.8168776371308017,
      "grad_norm": 1.1262956857681274,
      "learning_rate": 8.622308990310021e-05,
      "loss": 0.6395270228385925,
      "step": 4306
    },
    {
      "epoch": 1.8177215189873417,
      "grad_norm": 1.0448222160339355,
      "learning_rate": 8.620653208877328e-05,
      "loss": 0.6165015697479248,
      "step": 4308
    },
    {
      "epoch": 1.8185654008438819,
      "grad_norm": 1.1555759906768799,
      "learning_rate": 8.618996592229473e-05,
      "loss": 0.5915844440460205,
      "step": 4310
    },
    {
      "epoch": 1.819409282700422,
      "grad_norm": 1.5407506227493286,
      "learning_rate": 8.617339140748608e-05,
      "loss": 0.6491456627845764,
      "step": 4312
    },
    {
      "epoch": 1.820253164556962,
      "grad_norm": 1.3690788745880127,
      "learning_rate": 8.615680854817077e-05,
      "loss": 0.6053901314735413,
      "step": 4314
    },
    {
      "epoch": 1.8210970464135021,
      "grad_norm": 1.052583932876587,
      "learning_rate": 8.614021734817413e-05,
      "loss": 0.5821644067764282,
      "step": 4316
    },
    {
      "epoch": 1.8219409282700423,
      "grad_norm": 1.090567708015442,
      "learning_rate": 8.612361781132344e-05,
      "loss": 0.645878255367279,
      "step": 4318
    },
    {
      "epoch": 1.8227848101265822,
      "grad_norm": 1.122719645500183,
      "learning_rate": 8.610700994144787e-05,
      "loss": 0.6883123517036438,
      "step": 4320
    },
    {
      "epoch": 1.8236286919831224,
      "grad_norm": 1.3273001909255981,
      "learning_rate": 8.609039374237856e-05,
      "loss": 0.6918330788612366,
      "step": 4322
    },
    {
      "epoch": 1.8244725738396625,
      "grad_norm": 1.0628443956375122,
      "learning_rate": 8.607376921794855e-05,
      "loss": 0.6292204856872559,
      "step": 4324
    },
    {
      "epoch": 1.8253164556962025,
      "grad_norm": 1.287466287612915,
      "learning_rate": 8.605713637199279e-05,
      "loss": 0.6136105060577393,
      "step": 4326
    },
    {
      "epoch": 1.8261603375527427,
      "grad_norm": 1.1399345397949219,
      "learning_rate": 8.604049520834816e-05,
      "loss": 0.6099681854248047,
      "step": 4328
    },
    {
      "epoch": 1.8270042194092828,
      "grad_norm": 1.1131435632705688,
      "learning_rate": 8.602384573085345e-05,
      "loss": 0.6267056465148926,
      "step": 4330
    },
    {
      "epoch": 1.8278481012658228,
      "grad_norm": 1.1312925815582275,
      "learning_rate": 8.600718794334939e-05,
      "loss": 0.609437882900238,
      "step": 4332
    },
    {
      "epoch": 1.828691983122363,
      "grad_norm": 1.3711494207382202,
      "learning_rate": 8.599052184967859e-05,
      "loss": 0.727881669998169,
      "step": 4334
    },
    {
      "epoch": 1.829535864978903,
      "grad_norm": 1.1403605937957764,
      "learning_rate": 8.597384745368562e-05,
      "loss": 0.6771696209907532,
      "step": 4336
    },
    {
      "epoch": 1.830379746835443,
      "grad_norm": 1.2769951820373535,
      "learning_rate": 8.595716475921693e-05,
      "loss": 0.6812924742698669,
      "step": 4338
    },
    {
      "epoch": 1.831223628691983,
      "grad_norm": 1.055721402168274,
      "learning_rate": 8.59404737701209e-05,
      "loss": 0.6403515338897705,
      "step": 4340
    },
    {
      "epoch": 1.8320675105485233,
      "grad_norm": 1.1047639846801758,
      "learning_rate": 8.592377449024784e-05,
      "loss": 0.663240373134613,
      "step": 4342
    },
    {
      "epoch": 1.8329113924050633,
      "grad_norm": 1.0808883905410767,
      "learning_rate": 8.590706692344991e-05,
      "loss": 0.6398993134498596,
      "step": 4344
    },
    {
      "epoch": 1.8337552742616032,
      "grad_norm": 1.2433407306671143,
      "learning_rate": 8.589035107358125e-05,
      "loss": 0.6838348507881165,
      "step": 4346
    },
    {
      "epoch": 1.8345991561181436,
      "grad_norm": 1.031216025352478,
      "learning_rate": 8.58736269444979e-05,
      "loss": 0.640884280204773,
      "step": 4348
    },
    {
      "epoch": 1.8354430379746836,
      "grad_norm": 1.1417057514190674,
      "learning_rate": 8.585689454005776e-05,
      "loss": 0.6346741914749146,
      "step": 4350
    },
    {
      "epoch": 1.8362869198312235,
      "grad_norm": 1.210988998413086,
      "learning_rate": 8.584015386412072e-05,
      "loss": 0.6209521889686584,
      "step": 4352
    },
    {
      "epoch": 1.8371308016877637,
      "grad_norm": 1.2120760679244995,
      "learning_rate": 8.582340492054847e-05,
      "loss": 0.6699252128601074,
      "step": 4354
    },
    {
      "epoch": 1.8379746835443038,
      "grad_norm": 1.1768114566802979,
      "learning_rate": 8.580664771320475e-05,
      "loss": 0.6472980380058289,
      "step": 4356
    },
    {
      "epoch": 1.8388185654008438,
      "grad_norm": 1.060070276260376,
      "learning_rate": 8.578988224595506e-05,
      "loss": 0.6440452933311462,
      "step": 4358
    },
    {
      "epoch": 1.839662447257384,
      "grad_norm": 1.1366443634033203,
      "learning_rate": 8.57731085226669e-05,
      "loss": 0.5894474387168884,
      "step": 4360
    },
    {
      "epoch": 1.840506329113924,
      "grad_norm": 1.1571751832962036,
      "learning_rate": 8.575632654720963e-05,
      "loss": 0.5868900418281555,
      "step": 4362
    },
    {
      "epoch": 1.841350210970464,
      "grad_norm": 1.1983840465545654,
      "learning_rate": 8.573953632345453e-05,
      "loss": 0.5841533541679382,
      "step": 4364
    },
    {
      "epoch": 1.8421940928270042,
      "grad_norm": 1.101806640625,
      "learning_rate": 8.572273785527481e-05,
      "loss": 0.5503215193748474,
      "step": 4366
    },
    {
      "epoch": 1.8430379746835444,
      "grad_norm": 1.0327471494674683,
      "learning_rate": 8.570593114654552e-05,
      "loss": 0.6131128072738647,
      "step": 4368
    },
    {
      "epoch": 1.8438818565400843,
      "grad_norm": 1.1421098709106445,
      "learning_rate": 8.568911620114368e-05,
      "loss": 0.6614060401916504,
      "step": 4370
    },
    {
      "epoch": 1.8447257383966245,
      "grad_norm": 1.1707026958465576,
      "learning_rate": 8.567229302294814e-05,
      "loss": 0.6392307877540588,
      "step": 4372
    },
    {
      "epoch": 1.8455696202531646,
      "grad_norm": 1.1704418659210205,
      "learning_rate": 8.565546161583969e-05,
      "loss": 0.6560825109481812,
      "step": 4374
    },
    {
      "epoch": 1.8464135021097046,
      "grad_norm": 1.3618037700653076,
      "learning_rate": 8.563862198370103e-05,
      "loss": 0.6996290683746338,
      "step": 4376
    },
    {
      "epoch": 1.8472573839662447,
      "grad_norm": 1.116645097732544,
      "learning_rate": 8.562177413041674e-05,
      "loss": 0.6776535511016846,
      "step": 4378
    },
    {
      "epoch": 1.8481012658227849,
      "grad_norm": 1.1669151782989502,
      "learning_rate": 8.560491805987327e-05,
      "loss": 0.6390423774719238,
      "step": 4380
    },
    {
      "epoch": 1.8489451476793248,
      "grad_norm": 1.2188117504119873,
      "learning_rate": 8.558805377595904e-05,
      "loss": 0.6554020047187805,
      "step": 4382
    },
    {
      "epoch": 1.849789029535865,
      "grad_norm": 1.216829776763916,
      "learning_rate": 8.557118128256425e-05,
      "loss": 0.6291787624359131,
      "step": 4384
    },
    {
      "epoch": 1.8506329113924052,
      "grad_norm": 1.0431596040725708,
      "learning_rate": 8.555430058358111e-05,
      "loss": 0.6484442949295044,
      "step": 4386
    },
    {
      "epoch": 1.851476793248945,
      "grad_norm": 1.3015289306640625,
      "learning_rate": 8.553741168290367e-05,
      "loss": 0.7034047842025757,
      "step": 4388
    },
    {
      "epoch": 1.8523206751054853,
      "grad_norm": 1.2062040567398071,
      "learning_rate": 8.552051458442785e-05,
      "loss": 0.644135594367981,
      "step": 4390
    },
    {
      "epoch": 1.8531645569620254,
      "grad_norm": 1.238461971282959,
      "learning_rate": 8.55036092920515e-05,
      "loss": 0.6767282485961914,
      "step": 4392
    },
    {
      "epoch": 1.8540084388185654,
      "grad_norm": 1.2978830337524414,
      "learning_rate": 8.548669580967435e-05,
      "loss": 0.7292267680168152,
      "step": 4394
    },
    {
      "epoch": 1.8548523206751055,
      "grad_norm": 1.1448328495025635,
      "learning_rate": 8.546977414119801e-05,
      "loss": 0.6788421273231506,
      "step": 4396
    },
    {
      "epoch": 1.8556962025316457,
      "grad_norm": 1.0685368776321411,
      "learning_rate": 8.5452844290526e-05,
      "loss": 0.6745942234992981,
      "step": 4398
    },
    {
      "epoch": 1.8565400843881856,
      "grad_norm": 1.125707983970642,
      "learning_rate": 8.543590626156368e-05,
      "loss": 0.6351125836372375,
      "step": 4400
    },
    {
      "epoch": 1.8565400843881856,
      "eval_loss": 0.6961485147476196,
      "eval_runtime": 513.5724,
      "eval_samples_per_second": 4.103,
      "eval_steps_per_second": 4.103,
      "step": 4400
    },
    {
      "epoch": 1.8573839662447258,
      "grad_norm": 1.072179913520813,
      "learning_rate": 8.541896005821835e-05,
      "loss": 0.5840762257575989,
      "step": 4402
    },
    {
      "epoch": 1.858227848101266,
      "grad_norm": 1.2572803497314453,
      "learning_rate": 8.540200568439915e-05,
      "loss": 0.6431074738502502,
      "step": 4404
    },
    {
      "epoch": 1.859071729957806,
      "grad_norm": 1.3294413089752197,
      "learning_rate": 8.538504314401718e-05,
      "loss": 0.708808183670044,
      "step": 4406
    },
    {
      "epoch": 1.8599156118143458,
      "grad_norm": 1.1775587797164917,
      "learning_rate": 8.536807244098533e-05,
      "loss": 0.6580085754394531,
      "step": 4408
    },
    {
      "epoch": 1.8607594936708862,
      "grad_norm": 1.1880089044570923,
      "learning_rate": 8.53510935792184e-05,
      "loss": 0.6500136256217957,
      "step": 4410
    },
    {
      "epoch": 1.8616033755274262,
      "grad_norm": 1.2166204452514648,
      "learning_rate": 8.533410656263313e-05,
      "loss": 0.6922352313995361,
      "step": 4412
    },
    {
      "epoch": 1.862447257383966,
      "grad_norm": 1.0405415296554565,
      "learning_rate": 8.531711139514808e-05,
      "loss": 0.6761626601219177,
      "step": 4414
    },
    {
      "epoch": 1.8632911392405065,
      "grad_norm": 1.0674270391464233,
      "learning_rate": 8.530010808068371e-05,
      "loss": 0.672576904296875,
      "step": 4416
    },
    {
      "epoch": 1.8641350210970464,
      "grad_norm": 1.0584741830825806,
      "learning_rate": 8.528309662316236e-05,
      "loss": 0.5521218180656433,
      "step": 4418
    },
    {
      "epoch": 1.8649789029535864,
      "grad_norm": 1.3619039058685303,
      "learning_rate": 8.526607702650824e-05,
      "loss": 0.6546680927276611,
      "step": 4420
    },
    {
      "epoch": 1.8658227848101265,
      "grad_norm": 0.9904745221138,
      "learning_rate": 8.524904929464745e-05,
      "loss": 0.6043933629989624,
      "step": 4422
    },
    {
      "epoch": 1.8666666666666667,
      "grad_norm": 1.3046703338623047,
      "learning_rate": 8.523201343150795e-05,
      "loss": 0.7106801271438599,
      "step": 4424
    },
    {
      "epoch": 1.8675105485232066,
      "grad_norm": 1.1166832447052002,
      "learning_rate": 8.52149694410196e-05,
      "loss": 0.6456703543663025,
      "step": 4426
    },
    {
      "epoch": 1.8683544303797468,
      "grad_norm": 1.1260632276535034,
      "learning_rate": 8.519791732711412e-05,
      "loss": 0.5963318347930908,
      "step": 4428
    },
    {
      "epoch": 1.869198312236287,
      "grad_norm": 1.0990599393844604,
      "learning_rate": 8.51808570937251e-05,
      "loss": 0.6295356750488281,
      "step": 4430
    },
    {
      "epoch": 1.870042194092827,
      "grad_norm": 1.3689274787902832,
      "learning_rate": 8.516378874478801e-05,
      "loss": 0.6984617114067078,
      "step": 4432
    },
    {
      "epoch": 1.870886075949367,
      "grad_norm": 1.0986580848693848,
      "learning_rate": 8.514671228424018e-05,
      "loss": 0.5598900318145752,
      "step": 4434
    },
    {
      "epoch": 1.8717299578059072,
      "grad_norm": 0.9570761322975159,
      "learning_rate": 8.512962771602085e-05,
      "loss": 0.6286435723304749,
      "step": 4436
    },
    {
      "epoch": 1.8725738396624472,
      "grad_norm": 1.1480669975280762,
      "learning_rate": 8.511253504407107e-05,
      "loss": 0.5956313014030457,
      "step": 4438
    },
    {
      "epoch": 1.8734177215189873,
      "grad_norm": 1.1132479906082153,
      "learning_rate": 8.50954342723338e-05,
      "loss": 0.6523844599723816,
      "step": 4440
    },
    {
      "epoch": 1.8742616033755275,
      "grad_norm": 1.1569167375564575,
      "learning_rate": 8.507832540475387e-05,
      "loss": 0.6231355667114258,
      "step": 4442
    },
    {
      "epoch": 1.8751054852320674,
      "grad_norm": 1.1327043771743774,
      "learning_rate": 8.506120844527796e-05,
      "loss": 0.660773754119873,
      "step": 4444
    },
    {
      "epoch": 1.8759493670886076,
      "grad_norm": 0.8939630389213562,
      "learning_rate": 8.504408339785463e-05,
      "loss": 0.6319235563278198,
      "step": 4446
    },
    {
      "epoch": 1.8767932489451478,
      "grad_norm": 1.1910638809204102,
      "learning_rate": 8.50269502664343e-05,
      "loss": 0.6753001809120178,
      "step": 4448
    },
    {
      "epoch": 1.8776371308016877,
      "grad_norm": 1.1502408981323242,
      "learning_rate": 8.500980905496923e-05,
      "loss": 0.6300671696662903,
      "step": 4450
    },
    {
      "epoch": 1.8784810126582279,
      "grad_norm": 1.0639009475708008,
      "learning_rate": 8.49926597674136e-05,
      "loss": 0.6196691989898682,
      "step": 4452
    },
    {
      "epoch": 1.879324894514768,
      "grad_norm": 1.1072754859924316,
      "learning_rate": 8.497550240772341e-05,
      "loss": 0.7029181122779846,
      "step": 4454
    },
    {
      "epoch": 1.880168776371308,
      "grad_norm": 1.0440188646316528,
      "learning_rate": 8.495833697985652e-05,
      "loss": 0.65432208776474,
      "step": 4456
    },
    {
      "epoch": 1.8810126582278481,
      "grad_norm": 1.0646617412567139,
      "learning_rate": 8.494116348777269e-05,
      "loss": 0.6446614861488342,
      "step": 4458
    },
    {
      "epoch": 1.8818565400843883,
      "grad_norm": 1.2163805961608887,
      "learning_rate": 8.492398193543349e-05,
      "loss": 0.6430497765541077,
      "step": 4460
    },
    {
      "epoch": 1.8827004219409282,
      "grad_norm": 1.2715297937393188,
      "learning_rate": 8.490679232680241e-05,
      "loss": 0.6609845161437988,
      "step": 4462
    },
    {
      "epoch": 1.8835443037974684,
      "grad_norm": 1.0435588359832764,
      "learning_rate": 8.488959466584469e-05,
      "loss": 0.5791062712669373,
      "step": 4464
    },
    {
      "epoch": 1.8843881856540086,
      "grad_norm": 1.229202151298523,
      "learning_rate": 8.487238895652759e-05,
      "loss": 0.6312171220779419,
      "step": 4466
    },
    {
      "epoch": 1.8852320675105485,
      "grad_norm": 1.0713022947311401,
      "learning_rate": 8.485517520282008e-05,
      "loss": 0.6698815226554871,
      "step": 4468
    },
    {
      "epoch": 1.8860759493670884,
      "grad_norm": 1.0172312259674072,
      "learning_rate": 8.483795340869305e-05,
      "loss": 0.6283810138702393,
      "step": 4470
    },
    {
      "epoch": 1.8869198312236288,
      "grad_norm": 1.2880207300186157,
      "learning_rate": 8.482072357811926e-05,
      "loss": 0.6659437417984009,
      "step": 4472
    },
    {
      "epoch": 1.8877637130801688,
      "grad_norm": 1.0840508937835693,
      "learning_rate": 8.480348571507329e-05,
      "loss": 0.6190289258956909,
      "step": 4474
    },
    {
      "epoch": 1.8886075949367087,
      "grad_norm": 1.1101994514465332,
      "learning_rate": 8.478623982353156e-05,
      "loss": 0.5760066509246826,
      "step": 4476
    },
    {
      "epoch": 1.889451476793249,
      "grad_norm": 1.2388770580291748,
      "learning_rate": 8.476898590747237e-05,
      "loss": 0.6151811480522156,
      "step": 4478
    },
    {
      "epoch": 1.890295358649789,
      "grad_norm": 0.9986408948898315,
      "learning_rate": 8.475172397087591e-05,
      "loss": 0.5991593599319458,
      "step": 4480
    },
    {
      "epoch": 1.891139240506329,
      "grad_norm": 1.1380778551101685,
      "learning_rate": 8.473445401772415e-05,
      "loss": 0.7262179255485535,
      "step": 4482
    },
    {
      "epoch": 1.8919831223628694,
      "grad_norm": 1.3933676481246948,
      "learning_rate": 8.471717605200092e-05,
      "loss": 0.5806916356086731,
      "step": 4484
    },
    {
      "epoch": 1.8928270042194093,
      "grad_norm": 1.0242944955825806,
      "learning_rate": 8.469989007769194e-05,
      "loss": 0.617904782295227,
      "step": 4486
    },
    {
      "epoch": 1.8936708860759492,
      "grad_norm": 1.0909028053283691,
      "learning_rate": 8.468259609878475e-05,
      "loss": 0.6488202810287476,
      "step": 4488
    },
    {
      "epoch": 1.8945147679324894,
      "grad_norm": 1.042611002922058,
      "learning_rate": 8.466529411926874e-05,
      "loss": 0.6015118956565857,
      "step": 4490
    },
    {
      "epoch": 1.8953586497890296,
      "grad_norm": 1.3965784311294556,
      "learning_rate": 8.46479841431351e-05,
      "loss": 0.7035272717475891,
      "step": 4492
    },
    {
      "epoch": 1.8962025316455695,
      "grad_norm": 1.1486462354660034,
      "learning_rate": 8.463066617437698e-05,
      "loss": 0.6611229777336121,
      "step": 4494
    },
    {
      "epoch": 1.8970464135021097,
      "grad_norm": 1.0845859050750732,
      "learning_rate": 8.461334021698925e-05,
      "loss": 0.6378056406974792,
      "step": 4496
    },
    {
      "epoch": 1.8978902953586498,
      "grad_norm": 0.936612069606781,
      "learning_rate": 8.459600627496869e-05,
      "loss": 0.642429769039154,
      "step": 4498
    },
    {
      "epoch": 1.8987341772151898,
      "grad_norm": 1.1905454397201538,
      "learning_rate": 8.457866435231391e-05,
      "loss": 0.6341768503189087,
      "step": 4500
    },
    {
      "epoch": 1.8987341772151898,
      "eval_loss": 0.6938078999519348,
      "eval_runtime": 513.615,
      "eval_samples_per_second": 4.102,
      "eval_steps_per_second": 4.102,
      "step": 4500
    },
    {
      "epoch": 1.89957805907173,
      "grad_norm": 0.9778118133544922,
      "learning_rate": 8.456131445302538e-05,
      "loss": 0.5973100662231445,
      "step": 4502
    },
    {
      "epoch": 1.90042194092827,
      "grad_norm": 0.9587083458900452,
      "learning_rate": 8.454395658110536e-05,
      "loss": 0.5982911586761475,
      "step": 4504
    },
    {
      "epoch": 1.90126582278481,
      "grad_norm": 1.327643871307373,
      "learning_rate": 8.452659074055798e-05,
      "loss": 0.6858586668968201,
      "step": 4506
    },
    {
      "epoch": 1.9021097046413502,
      "grad_norm": 1.0740257501602173,
      "learning_rate": 8.450921693538922e-05,
      "loss": 0.6172328591346741,
      "step": 4508
    },
    {
      "epoch": 1.9029535864978904,
      "grad_norm": 1.0705101490020752,
      "learning_rate": 8.449183516960685e-05,
      "loss": 0.5349634289741516,
      "step": 4510
    },
    {
      "epoch": 1.9037974683544303,
      "grad_norm": 0.9151237607002258,
      "learning_rate": 8.447444544722058e-05,
      "loss": 0.5769277811050415,
      "step": 4512
    },
    {
      "epoch": 1.9046413502109705,
      "grad_norm": 1.139900803565979,
      "learning_rate": 8.44570477722418e-05,
      "loss": 0.6579093933105469,
      "step": 4514
    },
    {
      "epoch": 1.9054852320675106,
      "grad_norm": 1.2481658458709717,
      "learning_rate": 8.443964214868387e-05,
      "loss": 0.6748929619789124,
      "step": 4516
    },
    {
      "epoch": 1.9063291139240506,
      "grad_norm": 1.1661686897277832,
      "learning_rate": 8.442222858056193e-05,
      "loss": 0.6492021083831787,
      "step": 4518
    },
    {
      "epoch": 1.9071729957805907,
      "grad_norm": 1.241477370262146,
      "learning_rate": 8.440480707189295e-05,
      "loss": 0.635409951210022,
      "step": 4520
    },
    {
      "epoch": 1.908016877637131,
      "grad_norm": 1.1102054119110107,
      "learning_rate": 8.438737762669573e-05,
      "loss": 0.631928026676178,
      "step": 4522
    },
    {
      "epoch": 1.9088607594936708,
      "grad_norm": 1.0638107061386108,
      "learning_rate": 8.43699402489909e-05,
      "loss": 0.604518473148346,
      "step": 4524
    },
    {
      "epoch": 1.909704641350211,
      "grad_norm": 1.0270655155181885,
      "learning_rate": 8.435249494280096e-05,
      "loss": 0.61314457654953,
      "step": 4526
    },
    {
      "epoch": 1.9105485232067512,
      "grad_norm": 1.1840111017227173,
      "learning_rate": 8.433504171215018e-05,
      "loss": 0.661663293838501,
      "step": 4528
    },
    {
      "epoch": 1.9113924050632911,
      "grad_norm": 1.1404399871826172,
      "learning_rate": 8.43175805610647e-05,
      "loss": 0.7026967406272888,
      "step": 4530
    },
    {
      "epoch": 1.9122362869198313,
      "grad_norm": 1.2371265888214111,
      "learning_rate": 8.430011149357246e-05,
      "loss": 0.6599440574645996,
      "step": 4532
    },
    {
      "epoch": 1.9130801687763714,
      "grad_norm": 1.0042651891708374,
      "learning_rate": 8.428263451370326e-05,
      "loss": 0.5728344321250916,
      "step": 4534
    },
    {
      "epoch": 1.9139240506329114,
      "grad_norm": 1.04367196559906,
      "learning_rate": 8.426514962548866e-05,
      "loss": 0.6495450735092163,
      "step": 4536
    },
    {
      "epoch": 1.9147679324894513,
      "grad_norm": 1.0867135524749756,
      "learning_rate": 8.424765683296215e-05,
      "loss": 0.6406553387641907,
      "step": 4538
    },
    {
      "epoch": 1.9156118143459917,
      "grad_norm": 1.0751310586929321,
      "learning_rate": 8.423015614015892e-05,
      "loss": 0.6692186594009399,
      "step": 4540
    },
    {
      "epoch": 1.9164556962025316,
      "grad_norm": 1.13556969165802,
      "learning_rate": 8.421264755111607e-05,
      "loss": 0.6029785871505737,
      "step": 4542
    },
    {
      "epoch": 1.9172995780590716,
      "grad_norm": 1.1560977697372437,
      "learning_rate": 8.419513106987251e-05,
      "loss": 0.6457844972610474,
      "step": 4544
    },
    {
      "epoch": 1.918143459915612,
      "grad_norm": 1.2192902565002441,
      "learning_rate": 8.417760670046893e-05,
      "loss": 0.7082147598266602,
      "step": 4546
    },
    {
      "epoch": 1.918987341772152,
      "grad_norm": 1.1170696020126343,
      "learning_rate": 8.41600744469479e-05,
      "loss": 0.6919234991073608,
      "step": 4548
    },
    {
      "epoch": 1.9198312236286919,
      "grad_norm": 1.061253547668457,
      "learning_rate": 8.414253431335373e-05,
      "loss": 0.6310052871704102,
      "step": 4550
    },
    {
      "epoch": 1.920675105485232,
      "grad_norm": 1.0671885013580322,
      "learning_rate": 8.412498630373263e-05,
      "loss": 0.6330236792564392,
      "step": 4552
    },
    {
      "epoch": 1.9215189873417722,
      "grad_norm": 1.2085163593292236,
      "learning_rate": 8.410743042213256e-05,
      "loss": 0.7031015157699585,
      "step": 4554
    },
    {
      "epoch": 1.9223628691983121,
      "grad_norm": 1.2682013511657715,
      "learning_rate": 8.408986667260334e-05,
      "loss": 0.7078304290771484,
      "step": 4556
    },
    {
      "epoch": 1.9232067510548523,
      "grad_norm": 1.2966876029968262,
      "learning_rate": 8.407229505919658e-05,
      "loss": 0.6542860865592957,
      "step": 4558
    },
    {
      "epoch": 1.9240506329113924,
      "grad_norm": 1.1086169481277466,
      "learning_rate": 8.405471558596573e-05,
      "loss": 0.5856828093528748,
      "step": 4560
    },
    {
      "epoch": 1.9248945147679324,
      "grad_norm": 1.3175504207611084,
      "learning_rate": 8.403712825696604e-05,
      "loss": 0.7382104992866516,
      "step": 4562
    },
    {
      "epoch": 1.9257383966244725,
      "grad_norm": 1.163164496421814,
      "learning_rate": 8.401953307625454e-05,
      "loss": 0.6862360239028931,
      "step": 4564
    },
    {
      "epoch": 1.9265822784810127,
      "grad_norm": 1.207650899887085,
      "learning_rate": 8.400193004789013e-05,
      "loss": 0.7442302703857422,
      "step": 4566
    },
    {
      "epoch": 1.9274261603375527,
      "grad_norm": 1.1570589542388916,
      "learning_rate": 8.398431917593345e-05,
      "loss": 0.595226526260376,
      "step": 4568
    },
    {
      "epoch": 1.9282700421940928,
      "grad_norm": 1.091927170753479,
      "learning_rate": 8.396670046444704e-05,
      "loss": 0.6360410451889038,
      "step": 4570
    },
    {
      "epoch": 1.929113924050633,
      "grad_norm": 1.149559497833252,
      "learning_rate": 8.394907391749516e-05,
      "loss": 0.6343122124671936,
      "step": 4572
    },
    {
      "epoch": 1.929957805907173,
      "grad_norm": 1.0585254430770874,
      "learning_rate": 8.393143953914395e-05,
      "loss": 0.7394745349884033,
      "step": 4574
    },
    {
      "epoch": 1.930801687763713,
      "grad_norm": 1.1648521423339844,
      "learning_rate": 8.391379733346128e-05,
      "loss": 0.6489678025245667,
      "step": 4576
    },
    {
      "epoch": 1.9316455696202532,
      "grad_norm": 1.1756316423416138,
      "learning_rate": 8.389614730451692e-05,
      "loss": 0.6687861084938049,
      "step": 4578
    },
    {
      "epoch": 1.9324894514767932,
      "grad_norm": 0.9857237339019775,
      "learning_rate": 8.387848945638235e-05,
      "loss": 0.523727536201477,
      "step": 4580
    },
    {
      "epoch": 1.9333333333333333,
      "grad_norm": 1.1038693189620972,
      "learning_rate": 8.386082379313092e-05,
      "loss": 0.6545047760009766,
      "step": 4582
    },
    {
      "epoch": 1.9341772151898735,
      "grad_norm": 1.0780832767486572,
      "learning_rate": 8.384315031883774e-05,
      "loss": 0.6067036390304565,
      "step": 4584
    },
    {
      "epoch": 1.9350210970464135,
      "grad_norm": 1.2915070056915283,
      "learning_rate": 8.382546903757975e-05,
      "loss": 0.6880824565887451,
      "step": 4586
    },
    {
      "epoch": 1.9358649789029536,
      "grad_norm": 1.1243441104888916,
      "learning_rate": 8.380777995343568e-05,
      "loss": 0.7319117188453674,
      "step": 4588
    },
    {
      "epoch": 1.9367088607594938,
      "grad_norm": 1.1143072843551636,
      "learning_rate": 8.379008307048609e-05,
      "loss": 0.6845395565032959,
      "step": 4590
    },
    {
      "epoch": 1.9375527426160337,
      "grad_norm": 1.039494276046753,
      "learning_rate": 8.377237839281327e-05,
      "loss": 0.6653600335121155,
      "step": 4592
    },
    {
      "epoch": 1.9383966244725739,
      "grad_norm": 1.299617886543274,
      "learning_rate": 8.375466592450136e-05,
      "loss": 0.6352495551109314,
      "step": 4594
    },
    {
      "epoch": 1.939240506329114,
      "grad_norm": 0.9918657541275024,
      "learning_rate": 8.373694566963631e-05,
      "loss": 0.5660957098007202,
      "step": 4596
    },
    {
      "epoch": 1.940084388185654,
      "grad_norm": 1.0540478229522705,
      "learning_rate": 8.371921763230579e-05,
      "loss": 0.6296496987342834,
      "step": 4598
    },
    {
      "epoch": 1.9409282700421941,
      "grad_norm": 1.1309545040130615,
      "learning_rate": 8.370148181659939e-05,
      "loss": 0.6672025918960571,
      "step": 4600
    },
    {
      "epoch": 1.9409282700421941,
      "eval_loss": 0.6930755376815796,
      "eval_runtime": 617.8927,
      "eval_samples_per_second": 3.41,
      "eval_steps_per_second": 3.41,
      "step": 4600
    },
    {
      "epoch": 1.9417721518987343,
      "grad_norm": 1.2338588237762451,
      "learning_rate": 8.368373822660836e-05,
      "loss": 0.6200884580612183,
      "step": 4602
    },
    {
      "epoch": 1.9426160337552743,
      "grad_norm": 1.1756945848464966,
      "learning_rate": 8.366598686642582e-05,
      "loss": 0.653294026851654,
      "step": 4604
    },
    {
      "epoch": 1.9434599156118142,
      "grad_norm": 1.032018780708313,
      "learning_rate": 8.364822774014671e-05,
      "loss": 0.5670395493507385,
      "step": 4606
    },
    {
      "epoch": 1.9443037974683546,
      "grad_norm": 1.045280933380127,
      "learning_rate": 8.363046085186766e-05,
      "loss": 0.6819197535514832,
      "step": 4608
    },
    {
      "epoch": 1.9451476793248945,
      "grad_norm": 1.3223930597305298,
      "learning_rate": 8.36126862056872e-05,
      "loss": 0.6952820420265198,
      "step": 4610
    },
    {
      "epoch": 1.9459915611814345,
      "grad_norm": 1.0048432350158691,
      "learning_rate": 8.359490380570556e-05,
      "loss": 0.5291440486907959,
      "step": 4612
    },
    {
      "epoch": 1.9468354430379748,
      "grad_norm": 1.1477346420288086,
      "learning_rate": 8.357711365602483e-05,
      "loss": 0.6857813000679016,
      "step": 4614
    },
    {
      "epoch": 1.9476793248945148,
      "grad_norm": 0.959985077381134,
      "learning_rate": 8.355931576074882e-05,
      "loss": 0.5581508278846741,
      "step": 4616
    },
    {
      "epoch": 1.9485232067510547,
      "grad_norm": 1.1104289293289185,
      "learning_rate": 8.35415101239832e-05,
      "loss": 0.6536211371421814,
      "step": 4618
    },
    {
      "epoch": 1.9493670886075949,
      "grad_norm": 1.2344517707824707,
      "learning_rate": 8.352369674983535e-05,
      "loss": 0.6570560336112976,
      "step": 4620
    },
    {
      "epoch": 1.950210970464135,
      "grad_norm": 1.3411606550216675,
      "learning_rate": 8.350587564241451e-05,
      "loss": 0.6070495247840881,
      "step": 4622
    },
    {
      "epoch": 1.951054852320675,
      "grad_norm": 1.1713159084320068,
      "learning_rate": 8.348804680583166e-05,
      "loss": 0.6444135904312134,
      "step": 4624
    },
    {
      "epoch": 1.9518987341772152,
      "grad_norm": 1.127242922782898,
      "learning_rate": 8.347021024419954e-05,
      "loss": 0.6517419815063477,
      "step": 4626
    },
    {
      "epoch": 1.9527426160337553,
      "grad_norm": 1.0733028650283813,
      "learning_rate": 8.345236596163274e-05,
      "loss": 0.6174065470695496,
      "step": 4628
    },
    {
      "epoch": 1.9535864978902953,
      "grad_norm": 1.1114680767059326,
      "learning_rate": 8.343451396224757e-05,
      "loss": 0.7163593769073486,
      "step": 4630
    },
    {
      "epoch": 1.9544303797468354,
      "grad_norm": 1.0839568376541138,
      "learning_rate": 8.341665425016216e-05,
      "loss": 0.698553204536438,
      "step": 4632
    },
    {
      "epoch": 1.9552742616033756,
      "grad_norm": 1.17001211643219,
      "learning_rate": 8.339878682949638e-05,
      "loss": 0.6224857568740845,
      "step": 4634
    },
    {
      "epoch": 1.9561181434599155,
      "grad_norm": 3.483793020248413,
      "learning_rate": 8.338091170437193e-05,
      "loss": 0.5931200981140137,
      "step": 4636
    },
    {
      "epoch": 1.9569620253164557,
      "grad_norm": 1.1575394868850708,
      "learning_rate": 8.336302887891224e-05,
      "loss": 0.6031442284584045,
      "step": 4638
    },
    {
      "epoch": 1.9578059071729959,
      "grad_norm": 1.1494992971420288,
      "learning_rate": 8.334513835724252e-05,
      "loss": 0.6101768016815186,
      "step": 4640
    },
    {
      "epoch": 1.9586497890295358,
      "grad_norm": 1.3858197927474976,
      "learning_rate": 8.332724014348981e-05,
      "loss": 0.6571711301803589,
      "step": 4642
    },
    {
      "epoch": 1.959493670886076,
      "grad_norm": 1.1094943284988403,
      "learning_rate": 8.330933424178284e-05,
      "loss": 0.6391071677207947,
      "step": 4644
    },
    {
      "epoch": 1.9603375527426161,
      "grad_norm": 1.1640198230743408,
      "learning_rate": 8.329142065625218e-05,
      "loss": 0.6542805433273315,
      "step": 4646
    },
    {
      "epoch": 1.961181434599156,
      "grad_norm": 1.1080211400985718,
      "learning_rate": 8.327349939103016e-05,
      "loss": 0.6053075194358826,
      "step": 4648
    },
    {
      "epoch": 1.9620253164556962,
      "grad_norm": 1.0137052536010742,
      "learning_rate": 8.325557045025085e-05,
      "loss": 0.6009573340415955,
      "step": 4650
    },
    {
      "epoch": 1.9628691983122364,
      "grad_norm": 1.0867283344268799,
      "learning_rate": 8.323763383805012e-05,
      "loss": 0.5993483066558838,
      "step": 4652
    },
    {
      "epoch": 1.9637130801687763,
      "grad_norm": 1.0577161312103271,
      "learning_rate": 8.321968955856562e-05,
      "loss": 0.6788463592529297,
      "step": 4654
    },
    {
      "epoch": 1.9645569620253165,
      "grad_norm": 1.2002183198928833,
      "learning_rate": 8.320173761593672e-05,
      "loss": 0.5786917209625244,
      "step": 4656
    },
    {
      "epoch": 1.9654008438818567,
      "grad_norm": 1.2266993522644043,
      "learning_rate": 8.318377801430461e-05,
      "loss": 0.7437994480133057,
      "step": 4658
    },
    {
      "epoch": 1.9662447257383966,
      "grad_norm": 1.007582187652588,
      "learning_rate": 8.316581075781223e-05,
      "loss": 0.6763550639152527,
      "step": 4660
    },
    {
      "epoch": 1.9670886075949368,
      "grad_norm": 1.2374811172485352,
      "learning_rate": 8.314783585060425e-05,
      "loss": 0.6953140497207642,
      "step": 4662
    },
    {
      "epoch": 1.967932489451477,
      "grad_norm": 1.1791057586669922,
      "learning_rate": 8.312985329682717e-05,
      "loss": 0.6867341995239258,
      "step": 4664
    },
    {
      "epoch": 1.9687763713080169,
      "grad_norm": 1.1903331279754639,
      "learning_rate": 8.31118631006292e-05,
      "loss": 0.6445001363754272,
      "step": 4666
    },
    {
      "epoch": 1.9696202531645568,
      "grad_norm": 1.1731067895889282,
      "learning_rate": 8.309386526616034e-05,
      "loss": 0.6500589847564697,
      "step": 4668
    },
    {
      "epoch": 1.9704641350210972,
      "grad_norm": 0.9470233917236328,
      "learning_rate": 8.307585979757233e-05,
      "loss": 0.6215718984603882,
      "step": 4670
    },
    {
      "epoch": 1.9713080168776371,
      "grad_norm": 1.2900800704956055,
      "learning_rate": 8.305784669901872e-05,
      "loss": 0.6396787762641907,
      "step": 4672
    },
    {
      "epoch": 1.972151898734177,
      "grad_norm": 1.1729133129119873,
      "learning_rate": 8.303982597465474e-05,
      "loss": 0.6581959128379822,
      "step": 4674
    },
    {
      "epoch": 1.9729957805907175,
      "grad_norm": 1.1450555324554443,
      "learning_rate": 8.302179762863746e-05,
      "loss": 0.7013490796089172,
      "step": 4676
    },
    {
      "epoch": 1.9738396624472574,
      "grad_norm": 1.1506338119506836,
      "learning_rate": 8.300376166512567e-05,
      "loss": 0.6796102523803711,
      "step": 4678
    },
    {
      "epoch": 1.9746835443037973,
      "grad_norm": 1.149979591369629,
      "learning_rate": 8.298571808827991e-05,
      "loss": 0.6960519552230835,
      "step": 4680
    },
    {
      "epoch": 1.9755274261603377,
      "grad_norm": 1.1078912019729614,
      "learning_rate": 8.296766690226249e-05,
      "loss": 0.6789507865905762,
      "step": 4682
    },
    {
      "epoch": 1.9763713080168777,
      "grad_norm": 1.0199202299118042,
      "learning_rate": 8.294960811123747e-05,
      "loss": 0.5962659120559692,
      "step": 4684
    },
    {
      "epoch": 1.9772151898734176,
      "grad_norm": 1.2226134538650513,
      "learning_rate": 8.293154171937068e-05,
      "loss": 0.6483094692230225,
      "step": 4686
    },
    {
      "epoch": 1.9780590717299578,
      "grad_norm": 1.184095025062561,
      "learning_rate": 8.291346773082965e-05,
      "loss": 0.6750242710113525,
      "step": 4688
    },
    {
      "epoch": 1.978902953586498,
      "grad_norm": 1.1018693447113037,
      "learning_rate": 8.289538614978375e-05,
      "loss": 0.7094066739082336,
      "step": 4690
    },
    {
      "epoch": 1.9797468354430379,
      "grad_norm": 1.0342390537261963,
      "learning_rate": 8.287729698040403e-05,
      "loss": 0.6554126739501953,
      "step": 4692
    },
    {
      "epoch": 1.980590717299578,
      "grad_norm": 1.0603563785552979,
      "learning_rate": 8.285920022686332e-05,
      "loss": 0.5493529438972473,
      "step": 4694
    },
    {
      "epoch": 1.9814345991561182,
      "grad_norm": 1.139609932899475,
      "learning_rate": 8.284109589333617e-05,
      "loss": 0.6824741363525391,
      "step": 4696
    },
    {
      "epoch": 1.9822784810126581,
      "grad_norm": 1.2167822122573853,
      "learning_rate": 8.282298398399895e-05,
      "loss": 0.7121000289916992,
      "step": 4698
    },
    {
      "epoch": 1.9831223628691983,
      "grad_norm": 1.109857201576233,
      "learning_rate": 8.280486450302968e-05,
      "loss": 0.6711249351501465,
      "step": 4700
    },
    {
      "epoch": 1.9831223628691983,
      "eval_loss": 0.6923081278800964,
      "eval_runtime": 514.7729,
      "eval_samples_per_second": 4.093,
      "eval_steps_per_second": 4.093,
      "step": 4700
    },
    {
      "epoch": 1.9839662447257385,
      "grad_norm": 1.1387107372283936,
      "learning_rate": 8.27867374546082e-05,
      "loss": 0.581635594367981,
      "step": 4702
    },
    {
      "epoch": 1.9848101265822784,
      "grad_norm": 1.2519257068634033,
      "learning_rate": 8.27686028429161e-05,
      "loss": 0.6867302060127258,
      "step": 4704
    },
    {
      "epoch": 1.9856540084388186,
      "grad_norm": 1.0927205085754395,
      "learning_rate": 8.275046067213663e-05,
      "loss": 0.6494556665420532,
      "step": 4706
    },
    {
      "epoch": 1.9864978902953587,
      "grad_norm": 1.042035698890686,
      "learning_rate": 8.273231094645487e-05,
      "loss": 0.6949493288993835,
      "step": 4708
    },
    {
      "epoch": 1.9873417721518987,
      "grad_norm": 1.0220824480056763,
      "learning_rate": 8.271415367005762e-05,
      "loss": 0.6535884737968445,
      "step": 4710
    },
    {
      "epoch": 1.9881856540084388,
      "grad_norm": 1.3023611307144165,
      "learning_rate": 8.269598884713339e-05,
      "loss": 0.6635278463363647,
      "step": 4712
    },
    {
      "epoch": 1.989029535864979,
      "grad_norm": 1.2526965141296387,
      "learning_rate": 8.267781648187248e-05,
      "loss": 0.7194697856903076,
      "step": 4714
    },
    {
      "epoch": 1.989873417721519,
      "grad_norm": 1.0388038158416748,
      "learning_rate": 8.265963657846691e-05,
      "loss": 0.6355333924293518,
      "step": 4716
    },
    {
      "epoch": 1.990717299578059,
      "grad_norm": 1.0852965116500854,
      "learning_rate": 8.264144914111041e-05,
      "loss": 0.6898305416107178,
      "step": 4718
    },
    {
      "epoch": 1.9915611814345993,
      "grad_norm": 1.0714049339294434,
      "learning_rate": 8.262325417399847e-05,
      "loss": 0.6202836036682129,
      "step": 4720
    },
    {
      "epoch": 1.9924050632911392,
      "grad_norm": 1.0767238140106201,
      "learning_rate": 8.260505168132835e-05,
      "loss": 0.6160458326339722,
      "step": 4722
    },
    {
      "epoch": 1.9932489451476794,
      "grad_norm": 0.9605211615562439,
      "learning_rate": 8.258684166729899e-05,
      "loss": 0.6049920916557312,
      "step": 4724
    },
    {
      "epoch": 1.9940928270042195,
      "grad_norm": 1.0580185651779175,
      "learning_rate": 8.256862413611113e-05,
      "loss": 0.5622014999389648,
      "step": 4726
    },
    {
      "epoch": 1.9949367088607595,
      "grad_norm": 1.1039034128189087,
      "learning_rate": 8.255039909196713e-05,
      "loss": 0.6678924560546875,
      "step": 4728
    },
    {
      "epoch": 1.9957805907172996,
      "grad_norm": 1.1482586860656738,
      "learning_rate": 8.253216653907123e-05,
      "loss": 0.658260703086853,
      "step": 4730
    },
    {
      "epoch": 1.9966244725738398,
      "grad_norm": 1.135349988937378,
      "learning_rate": 8.251392648162929e-05,
      "loss": 0.6461613178253174,
      "step": 4732
    },
    {
      "epoch": 1.9974683544303797,
      "grad_norm": 1.0155420303344727,
      "learning_rate": 8.249567892384895e-05,
      "loss": 0.6837426424026489,
      "step": 4734
    },
    {
      "epoch": 1.9983122362869197,
      "grad_norm": 1.3392970561981201,
      "learning_rate": 8.247742386993958e-05,
      "loss": 0.6091697812080383,
      "step": 4736
    },
    {
      "epoch": 1.99915611814346,
      "grad_norm": 1.0509974956512451,
      "learning_rate": 8.245916132411226e-05,
      "loss": 0.6539653539657593,
      "step": 4738
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.9777396321296692,
      "learning_rate": 8.244089129057982e-05,
      "loss": 0.5630147457122803,
      "step": 4740
    },
    {
      "epoch": 2.00084388185654,
      "grad_norm": 1.1639164686203003,
      "learning_rate": 8.24226137735568e-05,
      "loss": 0.6190353631973267,
      "step": 4742
    },
    {
      "epoch": 2.0016877637130803,
      "grad_norm": 1.119614839553833,
      "learning_rate": 8.240432877725947e-05,
      "loss": 0.6282529234886169,
      "step": 4744
    },
    {
      "epoch": 2.0025316455696203,
      "grad_norm": 1.114739179611206,
      "learning_rate": 8.238603630590581e-05,
      "loss": 0.6176725625991821,
      "step": 4746
    },
    {
      "epoch": 2.00337552742616,
      "grad_norm": 1.0543076992034912,
      "learning_rate": 8.236773636371557e-05,
      "loss": 0.5182007551193237,
      "step": 4748
    },
    {
      "epoch": 2.0042194092827006,
      "grad_norm": 1.060389518737793,
      "learning_rate": 8.234942895491019e-05,
      "loss": 0.532536506652832,
      "step": 4750
    },
    {
      "epoch": 2.0050632911392405,
      "grad_norm": 1.0824412107467651,
      "learning_rate": 8.233111408371282e-05,
      "loss": 0.5474061369895935,
      "step": 4752
    },
    {
      "epoch": 2.0059071729957805,
      "grad_norm": 1.1450858116149902,
      "learning_rate": 8.231279175434838e-05,
      "loss": 0.586384654045105,
      "step": 4754
    },
    {
      "epoch": 2.006751054852321,
      "grad_norm": 1.1225577592849731,
      "learning_rate": 8.229446197104345e-05,
      "loss": 0.6469444036483765,
      "step": 4756
    },
    {
      "epoch": 2.007594936708861,
      "grad_norm": 1.7292449474334717,
      "learning_rate": 8.227612473802637e-05,
      "loss": 0.5371572971343994,
      "step": 4758
    },
    {
      "epoch": 2.0084388185654007,
      "grad_norm": 1.1743781566619873,
      "learning_rate": 8.22577800595272e-05,
      "loss": 0.558707058429718,
      "step": 4760
    },
    {
      "epoch": 2.009282700421941,
      "grad_norm": 1.0385273694992065,
      "learning_rate": 8.223942793977769e-05,
      "loss": 0.5943514108657837,
      "step": 4762
    },
    {
      "epoch": 2.010126582278481,
      "grad_norm": 1.1302000284194946,
      "learning_rate": 8.222106838301131e-05,
      "loss": 0.5630753636360168,
      "step": 4764
    },
    {
      "epoch": 2.010970464135021,
      "grad_norm": 1.140005111694336,
      "learning_rate": 8.220270139346327e-05,
      "loss": 0.527510404586792,
      "step": 4766
    },
    {
      "epoch": 2.0118143459915614,
      "grad_norm": 1.1979734897613525,
      "learning_rate": 8.21843269753705e-05,
      "loss": 0.6315013766288757,
      "step": 4768
    },
    {
      "epoch": 2.0126582278481013,
      "grad_norm": 1.3759459257125854,
      "learning_rate": 8.21659451329716e-05,
      "loss": 0.6225199699401855,
      "step": 4770
    },
    {
      "epoch": 2.0135021097046413,
      "grad_norm": 1.330600380897522,
      "learning_rate": 8.21475558705069e-05,
      "loss": 0.6838938593864441,
      "step": 4772
    },
    {
      "epoch": 2.014345991561181,
      "grad_norm": 1.2365351915359497,
      "learning_rate": 8.21291591922185e-05,
      "loss": 0.606302797794342,
      "step": 4774
    },
    {
      "epoch": 2.0151898734177216,
      "grad_norm": 1.1886142492294312,
      "learning_rate": 8.211075510235011e-05,
      "loss": 0.6194182634353638,
      "step": 4776
    },
    {
      "epoch": 2.0160337552742615,
      "grad_norm": 1.1414743661880493,
      "learning_rate": 8.209234360514721e-05,
      "loss": 0.639540433883667,
      "step": 4778
    },
    {
      "epoch": 2.0168776371308015,
      "grad_norm": 1.2877455949783325,
      "learning_rate": 8.2073924704857e-05,
      "loss": 0.6350902318954468,
      "step": 4780
    },
    {
      "epoch": 2.017721518987342,
      "grad_norm": 1.095578908920288,
      "learning_rate": 8.205549840572834e-05,
      "loss": 0.5152000784873962,
      "step": 4782
    },
    {
      "epoch": 2.018565400843882,
      "grad_norm": 1.0043798685073853,
      "learning_rate": 8.203706471201183e-05,
      "loss": 0.46245837211608887,
      "step": 4784
    },
    {
      "epoch": 2.0194092827004217,
      "grad_norm": 1.2133857011795044,
      "learning_rate": 8.201862362795979e-05,
      "loss": 0.6471722722053528,
      "step": 4786
    },
    {
      "epoch": 2.020253164556962,
      "grad_norm": 1.0835390090942383,
      "learning_rate": 8.200017515782619e-05,
      "loss": 0.5790625214576721,
      "step": 4788
    },
    {
      "epoch": 2.021097046413502,
      "grad_norm": 1.0176091194152832,
      "learning_rate": 8.198171930586678e-05,
      "loss": 0.5826238989830017,
      "step": 4790
    },
    {
      "epoch": 2.021940928270042,
      "grad_norm": 1.1581370830535889,
      "learning_rate": 8.196325607633893e-05,
      "loss": 0.5781272649765015,
      "step": 4792
    },
    {
      "epoch": 2.0227848101265824,
      "grad_norm": 1.243381142616272,
      "learning_rate": 8.194478547350178e-05,
      "loss": 0.6600401997566223,
      "step": 4794
    },
    {
      "epoch": 2.0236286919831223,
      "grad_norm": 1.0718560218811035,
      "learning_rate": 8.192630750161612e-05,
      "loss": 0.5291268825531006,
      "step": 4796
    },
    {
      "epoch": 2.0244725738396623,
      "grad_norm": 1.2338320016860962,
      "learning_rate": 8.190782216494448e-05,
      "loss": 0.6564924120903015,
      "step": 4798
    },
    {
      "epoch": 2.0253164556962027,
      "grad_norm": 0.978547990322113,
      "learning_rate": 8.188932946775107e-05,
      "loss": 0.5471183657646179,
      "step": 4800
    },
    {
      "epoch": 2.0253164556962027,
      "eval_loss": 0.6924457550048828,
      "eval_runtime": 514.0427,
      "eval_samples_per_second": 4.099,
      "eval_steps_per_second": 4.099,
      "step": 4800
    },
    {
      "epoch": 2.0261603375527426,
      "grad_norm": 1.1782792806625366,
      "learning_rate": 8.18708294143018e-05,
      "loss": 0.567442774772644,
      "step": 4802
    },
    {
      "epoch": 2.0270042194092825,
      "grad_norm": 1.0768574476242065,
      "learning_rate": 8.185232200886426e-05,
      "loss": 0.6005180478096008,
      "step": 4804
    },
    {
      "epoch": 2.027848101265823,
      "grad_norm": 1.3096717596054077,
      "learning_rate": 8.18338072557078e-05,
      "loss": 0.616436779499054,
      "step": 4806
    },
    {
      "epoch": 2.028691983122363,
      "grad_norm": 1.0233508348464966,
      "learning_rate": 8.181528515910336e-05,
      "loss": 0.49587416648864746,
      "step": 4808
    },
    {
      "epoch": 2.029535864978903,
      "grad_norm": 1.0800065994262695,
      "learning_rate": 8.179675572332366e-05,
      "loss": 0.5758571624755859,
      "step": 4810
    },
    {
      "epoch": 2.030379746835443,
      "grad_norm": 1.09299898147583,
      "learning_rate": 8.177821895264309e-05,
      "loss": 0.561736524105072,
      "step": 4812
    },
    {
      "epoch": 2.031223628691983,
      "grad_norm": 1.1439210176467896,
      "learning_rate": 8.175967485133771e-05,
      "loss": 0.5249468088150024,
      "step": 4814
    },
    {
      "epoch": 2.032067510548523,
      "grad_norm": 1.15841805934906,
      "learning_rate": 8.174112342368532e-05,
      "loss": 0.6429001688957214,
      "step": 4816
    },
    {
      "epoch": 2.0329113924050635,
      "grad_norm": 1.1720670461654663,
      "learning_rate": 8.172256467396533e-05,
      "loss": 0.60152667760849,
      "step": 4818
    },
    {
      "epoch": 2.0337552742616034,
      "grad_norm": 1.2652091979980469,
      "learning_rate": 8.170399860645892e-05,
      "loss": 0.5553541779518127,
      "step": 4820
    },
    {
      "epoch": 2.0345991561181433,
      "grad_norm": 1.0768507719039917,
      "learning_rate": 8.168542522544893e-05,
      "loss": 0.5369323492050171,
      "step": 4822
    },
    {
      "epoch": 2.0354430379746837,
      "grad_norm": 0.9906469583511353,
      "learning_rate": 8.166684453521986e-05,
      "loss": 0.5468952655792236,
      "step": 4824
    },
    {
      "epoch": 2.0362869198312237,
      "grad_norm": 1.3448988199234009,
      "learning_rate": 8.164825654005792e-05,
      "loss": 0.5795659422874451,
      "step": 4826
    },
    {
      "epoch": 2.0371308016877636,
      "grad_norm": 1.2502341270446777,
      "learning_rate": 8.162966124425103e-05,
      "loss": 0.6465779542922974,
      "step": 4828
    },
    {
      "epoch": 2.037974683544304,
      "grad_norm": 1.1512303352355957,
      "learning_rate": 8.161105865208875e-05,
      "loss": 0.5509394407272339,
      "step": 4830
    },
    {
      "epoch": 2.038818565400844,
      "grad_norm": 1.2513408660888672,
      "learning_rate": 8.159244876786232e-05,
      "loss": 0.5515735745429993,
      "step": 4832
    },
    {
      "epoch": 2.039662447257384,
      "grad_norm": 1.3035682439804077,
      "learning_rate": 8.157383159586473e-05,
      "loss": 0.757799506187439,
      "step": 4834
    },
    {
      "epoch": 2.0405063291139243,
      "grad_norm": 1.1136540174484253,
      "learning_rate": 8.155520714039056e-05,
      "loss": 0.607295036315918,
      "step": 4836
    },
    {
      "epoch": 2.041350210970464,
      "grad_norm": 1.220146656036377,
      "learning_rate": 8.153657540573613e-05,
      "loss": 0.5769712328910828,
      "step": 4838
    },
    {
      "epoch": 2.042194092827004,
      "grad_norm": 1.2104195356369019,
      "learning_rate": 8.151793639619944e-05,
      "loss": 0.5746933817863464,
      "step": 4840
    },
    {
      "epoch": 2.043037974683544,
      "grad_norm": 1.241708517074585,
      "learning_rate": 8.149929011608014e-05,
      "loss": 0.5932332277297974,
      "step": 4842
    },
    {
      "epoch": 2.0438818565400845,
      "grad_norm": 1.1172713041305542,
      "learning_rate": 8.148063656967955e-05,
      "loss": 0.583284318447113,
      "step": 4844
    },
    {
      "epoch": 2.0447257383966244,
      "grad_norm": 1.0867618322372437,
      "learning_rate": 8.14619757613007e-05,
      "loss": 0.5589476823806763,
      "step": 4846
    },
    {
      "epoch": 2.0455696202531644,
      "grad_norm": 1.2470483779907227,
      "learning_rate": 8.14433076952483e-05,
      "loss": 0.6118156313896179,
      "step": 4848
    },
    {
      "epoch": 2.0464135021097047,
      "grad_norm": 1.0908832550048828,
      "learning_rate": 8.142463237582868e-05,
      "loss": 0.5815895795822144,
      "step": 4850
    },
    {
      "epoch": 2.0472573839662447,
      "grad_norm": 1.2589281797409058,
      "learning_rate": 8.140594980734989e-05,
      "loss": 0.6232373714447021,
      "step": 4852
    },
    {
      "epoch": 2.0481012658227846,
      "grad_norm": 1.234152913093567,
      "learning_rate": 8.138725999412165e-05,
      "loss": 0.5992053151130676,
      "step": 4854
    },
    {
      "epoch": 2.048945147679325,
      "grad_norm": 1.3304446935653687,
      "learning_rate": 8.136856294045533e-05,
      "loss": 0.6494496464729309,
      "step": 4856
    },
    {
      "epoch": 2.049789029535865,
      "grad_norm": 1.1871088743209839,
      "learning_rate": 8.134985865066398e-05,
      "loss": 0.6263431906700134,
      "step": 4858
    },
    {
      "epoch": 2.050632911392405,
      "grad_norm": 1.1454699039459229,
      "learning_rate": 8.133114712906234e-05,
      "loss": 0.6036502122879028,
      "step": 4860
    },
    {
      "epoch": 2.0514767932489453,
      "grad_norm": 1.2953420877456665,
      "learning_rate": 8.131242837996675e-05,
      "loss": 0.5674451589584351,
      "step": 4862
    },
    {
      "epoch": 2.052320675105485,
      "grad_norm": 1.1874405145645142,
      "learning_rate": 8.129370240769534e-05,
      "loss": 0.5616317987442017,
      "step": 4864
    },
    {
      "epoch": 2.053164556962025,
      "grad_norm": 1.2936227321624756,
      "learning_rate": 8.127496921656777e-05,
      "loss": 0.6495023369789124,
      "step": 4866
    },
    {
      "epoch": 2.0540084388185655,
      "grad_norm": 1.1935228109359741,
      "learning_rate": 8.125622881090544e-05,
      "loss": 0.6028099060058594,
      "step": 4868
    },
    {
      "epoch": 2.0548523206751055,
      "grad_norm": 0.9932331442832947,
      "learning_rate": 8.123748119503143e-05,
      "loss": 0.476296067237854,
      "step": 4870
    },
    {
      "epoch": 2.0556962025316454,
      "grad_norm": 1.3878839015960693,
      "learning_rate": 8.121872637327042e-05,
      "loss": 0.6191902756690979,
      "step": 4872
    },
    {
      "epoch": 2.056540084388186,
      "grad_norm": 1.1185581684112549,
      "learning_rate": 8.11999643499488e-05,
      "loss": 0.566487729549408,
      "step": 4874
    },
    {
      "epoch": 2.0573839662447257,
      "grad_norm": 1.3729257583618164,
      "learning_rate": 8.118119512939464e-05,
      "loss": 0.5970078706741333,
      "step": 4876
    },
    {
      "epoch": 2.0582278481012657,
      "grad_norm": 1.1332688331604004,
      "learning_rate": 8.11624187159376e-05,
      "loss": 0.570341944694519,
      "step": 4878
    },
    {
      "epoch": 2.059071729957806,
      "grad_norm": 1.2648937702178955,
      "learning_rate": 8.114363511390903e-05,
      "loss": 0.6302897334098816,
      "step": 4880
    },
    {
      "epoch": 2.059915611814346,
      "grad_norm": 1.250616192817688,
      "learning_rate": 8.112484432764197e-05,
      "loss": 0.5619142651557922,
      "step": 4882
    },
    {
      "epoch": 2.060759493670886,
      "grad_norm": 0.9710861444473267,
      "learning_rate": 8.110604636147109e-05,
      "loss": 0.5426228642463684,
      "step": 4884
    },
    {
      "epoch": 2.0616033755274263,
      "grad_norm": 1.1979506015777588,
      "learning_rate": 8.108724121973271e-05,
      "loss": 0.5498107671737671,
      "step": 4886
    },
    {
      "epoch": 2.0624472573839663,
      "grad_norm": 1.0936485528945923,
      "learning_rate": 8.106842890676483e-05,
      "loss": 0.5695134401321411,
      "step": 4888
    },
    {
      "epoch": 2.0632911392405062,
      "grad_norm": 1.1246092319488525,
      "learning_rate": 8.10496094269071e-05,
      "loss": 0.5998331308364868,
      "step": 4890
    },
    {
      "epoch": 2.0641350210970466,
      "grad_norm": 1.244438648223877,
      "learning_rate": 8.103078278450075e-05,
      "loss": 0.5702623128890991,
      "step": 4892
    },
    {
      "epoch": 2.0649789029535865,
      "grad_norm": 1.1585633754730225,
      "learning_rate": 8.101194898388881e-05,
      "loss": 0.5392299890518188,
      "step": 4894
    },
    {
      "epoch": 2.0658227848101265,
      "grad_norm": 1.3044285774230957,
      "learning_rate": 8.099310802941582e-05,
      "loss": 0.5640127658843994,
      "step": 4896
    },
    {
      "epoch": 2.066666666666667,
      "grad_norm": 1.2483032941818237,
      "learning_rate": 8.097425992542804e-05,
      "loss": 0.6103175282478333,
      "step": 4898
    },
    {
      "epoch": 2.067510548523207,
      "grad_norm": 1.0845462083816528,
      "learning_rate": 8.095540467627337e-05,
      "loss": 0.5041166543960571,
      "step": 4900
    },
    {
      "epoch": 2.067510548523207,
      "eval_loss": 0.6941288113594055,
      "eval_runtime": 513.4497,
      "eval_samples_per_second": 4.104,
      "eval_steps_per_second": 4.104,
      "step": 4900
    },
    {
      "epoch": 2.0683544303797468,
      "grad_norm": 1.2493232488632202,
      "learning_rate": 8.093654228630134e-05,
      "loss": 0.6253946423530579,
      "step": 4902
    },
    {
      "epoch": 2.0691983122362867,
      "grad_norm": 1.1668756008148193,
      "learning_rate": 8.091767275986317e-05,
      "loss": 0.523486852645874,
      "step": 4904
    },
    {
      "epoch": 2.070042194092827,
      "grad_norm": 1.1709638833999634,
      "learning_rate": 8.089879610131167e-05,
      "loss": 0.5569989681243896,
      "step": 4906
    },
    {
      "epoch": 2.070886075949367,
      "grad_norm": 1.1044740676879883,
      "learning_rate": 8.087991231500133e-05,
      "loss": 0.642728865146637,
      "step": 4908
    },
    {
      "epoch": 2.071729957805907,
      "grad_norm": 1.1032549142837524,
      "learning_rate": 8.086102140528828e-05,
      "loss": 0.5998259782791138,
      "step": 4910
    },
    {
      "epoch": 2.0725738396624473,
      "grad_norm": 0.9980027079582214,
      "learning_rate": 8.08421233765303e-05,
      "loss": 0.5460172891616821,
      "step": 4912
    },
    {
      "epoch": 2.0734177215189873,
      "grad_norm": 1.0866090059280396,
      "learning_rate": 8.082321823308679e-05,
      "loss": 0.5643284916877747,
      "step": 4914
    },
    {
      "epoch": 2.0742616033755272,
      "grad_norm": 1.1942687034606934,
      "learning_rate": 8.080430597931878e-05,
      "loss": 0.554400622844696,
      "step": 4916
    },
    {
      "epoch": 2.0751054852320676,
      "grad_norm": 1.0680599212646484,
      "learning_rate": 8.078538661958901e-05,
      "loss": 0.5955621004104614,
      "step": 4918
    },
    {
      "epoch": 2.0759493670886076,
      "grad_norm": 1.20845627784729,
      "learning_rate": 8.076646015826179e-05,
      "loss": 0.5970203280448914,
      "step": 4920
    },
    {
      "epoch": 2.0767932489451475,
      "grad_norm": 1.8368924856185913,
      "learning_rate": 8.074752659970308e-05,
      "loss": 0.6467664837837219,
      "step": 4922
    },
    {
      "epoch": 2.077637130801688,
      "grad_norm": 1.3291922807693481,
      "learning_rate": 8.072858594828053e-05,
      "loss": 0.630719006061554,
      "step": 4924
    },
    {
      "epoch": 2.078481012658228,
      "grad_norm": 1.1496083736419678,
      "learning_rate": 8.070963820836333e-05,
      "loss": 0.601140022277832,
      "step": 4926
    },
    {
      "epoch": 2.0793248945147678,
      "grad_norm": 1.1562724113464355,
      "learning_rate": 8.069068338432239e-05,
      "loss": 0.6096881031990051,
      "step": 4928
    },
    {
      "epoch": 2.080168776371308,
      "grad_norm": 1.0115300416946411,
      "learning_rate": 8.067172148053021e-05,
      "loss": 0.5085908770561218,
      "step": 4930
    },
    {
      "epoch": 2.081012658227848,
      "grad_norm": 1.2181830406188965,
      "learning_rate": 8.065275250136097e-05,
      "loss": 0.5268720984458923,
      "step": 4932
    },
    {
      "epoch": 2.081856540084388,
      "grad_norm": 1.1249788999557495,
      "learning_rate": 8.06337764511904e-05,
      "loss": 0.6075665950775146,
      "step": 4934
    },
    {
      "epoch": 2.0827004219409284,
      "grad_norm": 1.1143964529037476,
      "learning_rate": 8.061479333439595e-05,
      "loss": 0.59170001745224,
      "step": 4936
    },
    {
      "epoch": 2.0835443037974684,
      "grad_norm": 1.4773131608963013,
      "learning_rate": 8.059580315535664e-05,
      "loss": 0.6689745187759399,
      "step": 4938
    },
    {
      "epoch": 2.0843881856540083,
      "grad_norm": 1.143965244293213,
      "learning_rate": 8.057680591845316e-05,
      "loss": 0.5409777760505676,
      "step": 4940
    },
    {
      "epoch": 2.0852320675105487,
      "grad_norm": 1.0384942293167114,
      "learning_rate": 8.055780162806777e-05,
      "loss": 0.5778636336326599,
      "step": 4942
    },
    {
      "epoch": 2.0860759493670886,
      "grad_norm": 1.0102177858352661,
      "learning_rate": 8.053879028858442e-05,
      "loss": 0.5576038360595703,
      "step": 4944
    },
    {
      "epoch": 2.0869198312236286,
      "grad_norm": 1.3792158365249634,
      "learning_rate": 8.051977190438868e-05,
      "loss": 0.5873376131057739,
      "step": 4946
    },
    {
      "epoch": 2.087763713080169,
      "grad_norm": 1.4402949810028076,
      "learning_rate": 8.050074647986768e-05,
      "loss": 0.6067743301391602,
      "step": 4948
    },
    {
      "epoch": 2.088607594936709,
      "grad_norm": 1.2719058990478516,
      "learning_rate": 8.048171401941027e-05,
      "loss": 0.604671835899353,
      "step": 4950
    },
    {
      "epoch": 2.089451476793249,
      "grad_norm": 1.1054867506027222,
      "learning_rate": 8.046267452740683e-05,
      "loss": 0.5743544697761536,
      "step": 4952
    },
    {
      "epoch": 2.090295358649789,
      "grad_norm": 1.0521535873413086,
      "learning_rate": 8.044362800824944e-05,
      "loss": 0.576278567314148,
      "step": 4954
    },
    {
      "epoch": 2.091139240506329,
      "grad_norm": 1.2665088176727295,
      "learning_rate": 8.042457446633174e-05,
      "loss": 0.5903641581535339,
      "step": 4956
    },
    {
      "epoch": 2.091983122362869,
      "grad_norm": 1.1283398866653442,
      "learning_rate": 8.040551390604902e-05,
      "loss": 0.5854214429855347,
      "step": 4958
    },
    {
      "epoch": 2.0928270042194095,
      "grad_norm": 1.1194316148757935,
      "learning_rate": 8.03864463317982e-05,
      "loss": 0.5843619108200073,
      "step": 4960
    },
    {
      "epoch": 2.0936708860759494,
      "grad_norm": 1.3581651449203491,
      "learning_rate": 8.036737174797778e-05,
      "loss": 0.6115096211433411,
      "step": 4962
    },
    {
      "epoch": 2.0945147679324894,
      "grad_norm": 1.341748595237732,
      "learning_rate": 8.034829015898793e-05,
      "loss": 0.5998795032501221,
      "step": 4964
    },
    {
      "epoch": 2.0953586497890297,
      "grad_norm": 1.2212611436843872,
      "learning_rate": 8.032920156923038e-05,
      "loss": 0.628372311592102,
      "step": 4966
    },
    {
      "epoch": 2.0962025316455697,
      "grad_norm": 1.1348317861557007,
      "learning_rate": 8.031010598310851e-05,
      "loss": 0.5668916702270508,
      "step": 4968
    },
    {
      "epoch": 2.0970464135021096,
      "grad_norm": 1.1106547117233276,
      "learning_rate": 8.029100340502731e-05,
      "loss": 0.5253881216049194,
      "step": 4970
    },
    {
      "epoch": 2.09789029535865,
      "grad_norm": 1.2471354007720947,
      "learning_rate": 8.027189383939339e-05,
      "loss": 0.5790762901306152,
      "step": 4972
    },
    {
      "epoch": 2.09873417721519,
      "grad_norm": 1.2477394342422485,
      "learning_rate": 8.025277729061492e-05,
      "loss": 0.6382888555526733,
      "step": 4974
    },
    {
      "epoch": 2.09957805907173,
      "grad_norm": 1.2716054916381836,
      "learning_rate": 8.023365376310176e-05,
      "loss": 0.5962072610855103,
      "step": 4976
    },
    {
      "epoch": 2.10042194092827,
      "grad_norm": 1.257820725440979,
      "learning_rate": 8.021452326126532e-05,
      "loss": 0.5882940292358398,
      "step": 4978
    },
    {
      "epoch": 2.1012658227848102,
      "grad_norm": 1.0924186706542969,
      "learning_rate": 8.019538578951864e-05,
      "loss": 0.5640701055526733,
      "step": 4980
    },
    {
      "epoch": 2.10210970464135,
      "grad_norm": 1.1250383853912354,
      "learning_rate": 8.017624135227637e-05,
      "loss": 0.5746428966522217,
      "step": 4982
    },
    {
      "epoch": 2.10295358649789,
      "grad_norm": 1.131323218345642,
      "learning_rate": 8.015708995395477e-05,
      "loss": 0.5611346960067749,
      "step": 4984
    },
    {
      "epoch": 2.1037974683544305,
      "grad_norm": 1.4267152547836304,
      "learning_rate": 8.013793159897171e-05,
      "loss": 0.6173797249794006,
      "step": 4986
    },
    {
      "epoch": 2.1046413502109704,
      "grad_norm": 1.41414213180542,
      "learning_rate": 8.011876629174662e-05,
      "loss": 0.64865642786026,
      "step": 4988
    },
    {
      "epoch": 2.1054852320675104,
      "grad_norm": 1.1498184204101562,
      "learning_rate": 8.00995940367006e-05,
      "loss": 0.6125827431678772,
      "step": 4990
    },
    {
      "epoch": 2.1063291139240508,
      "grad_norm": 1.2327708005905151,
      "learning_rate": 8.00804148382563e-05,
      "loss": 0.670495867729187,
      "step": 4992
    },
    {
      "epoch": 2.1071729957805907,
      "grad_norm": 1.2797311544418335,
      "learning_rate": 8.0061228700838e-05,
      "loss": 0.6020209193229675,
      "step": 4994
    },
    {
      "epoch": 2.1080168776371306,
      "grad_norm": 1.079584002494812,
      "learning_rate": 8.004203562887157e-05,
      "loss": 0.5974310636520386,
      "step": 4996
    },
    {
      "epoch": 2.108860759493671,
      "grad_norm": 1.4352604150772095,
      "learning_rate": 8.002283562678452e-05,
      "loss": 0.6424587368965149,
      "step": 4998
    },
    {
      "epoch": 2.109704641350211,
      "grad_norm": 1.0876719951629639,
      "learning_rate": 8.000362869900586e-05,
      "loss": 0.6185846328735352,
      "step": 5000
    },
    {
      "epoch": 2.109704641350211,
      "eval_loss": 0.6908889412879944,
      "eval_runtime": 675.8398,
      "eval_samples_per_second": 3.118,
      "eval_steps_per_second": 3.118,
      "step": 5000
    },
    {
      "epoch": 2.110548523206751,
      "grad_norm": 1.0125762224197388,
      "learning_rate": 7.998441484996631e-05,
      "loss": 0.6127280592918396,
      "step": 5002
    },
    {
      "epoch": 2.1113924050632913,
      "grad_norm": 1.0253753662109375,
      "learning_rate": 7.99651940840981e-05,
      "loss": 0.5495694875717163,
      "step": 5004
    },
    {
      "epoch": 2.1122362869198312,
      "grad_norm": 1.5620673894882202,
      "learning_rate": 7.994596640583511e-05,
      "loss": 0.6199497580528259,
      "step": 5006
    },
    {
      "epoch": 2.113080168776371,
      "grad_norm": 1.3032969236373901,
      "learning_rate": 7.992673181961281e-05,
      "loss": 0.5896390676498413,
      "step": 5008
    },
    {
      "epoch": 2.1139240506329116,
      "grad_norm": 1.0933046340942383,
      "learning_rate": 7.990749032986821e-05,
      "loss": 0.6332341432571411,
      "step": 5010
    },
    {
      "epoch": 2.1147679324894515,
      "grad_norm": 1.3115314245224,
      "learning_rate": 7.988824194104e-05,
      "loss": 0.5964323282241821,
      "step": 5012
    },
    {
      "epoch": 2.1156118143459914,
      "grad_norm": 1.229978084564209,
      "learning_rate": 7.986898665756837e-05,
      "loss": 0.5938325524330139,
      "step": 5014
    },
    {
      "epoch": 2.116455696202532,
      "grad_norm": 1.1779940128326416,
      "learning_rate": 7.984972448389517e-05,
      "loss": 0.5761791467666626,
      "step": 5016
    },
    {
      "epoch": 2.1172995780590718,
      "grad_norm": 1.063490629196167,
      "learning_rate": 7.98304554244638e-05,
      "loss": 0.6073653101921082,
      "step": 5018
    },
    {
      "epoch": 2.1181434599156117,
      "grad_norm": 1.2390391826629639,
      "learning_rate": 7.981117948371927e-05,
      "loss": 0.6126761436462402,
      "step": 5020
    },
    {
      "epoch": 2.118987341772152,
      "grad_norm": 1.1946247816085815,
      "learning_rate": 7.979189666610818e-05,
      "loss": 0.614434003829956,
      "step": 5022
    },
    {
      "epoch": 2.119831223628692,
      "grad_norm": 1.1008374691009521,
      "learning_rate": 7.977260697607867e-05,
      "loss": 0.5947603583335876,
      "step": 5024
    },
    {
      "epoch": 2.120675105485232,
      "grad_norm": 1.14899480342865,
      "learning_rate": 7.975331041808054e-05,
      "loss": 0.583965539932251,
      "step": 5026
    },
    {
      "epoch": 2.1215189873417724,
      "grad_norm": 1.1627864837646484,
      "learning_rate": 7.973400699656512e-05,
      "loss": 0.615121603012085,
      "step": 5028
    },
    {
      "epoch": 2.1223628691983123,
      "grad_norm": 1.3622617721557617,
      "learning_rate": 7.971469671598532e-05,
      "loss": 0.6268601417541504,
      "step": 5030
    },
    {
      "epoch": 2.1232067510548522,
      "grad_norm": 1.1735879182815552,
      "learning_rate": 7.96953795807957e-05,
      "loss": 0.6021270155906677,
      "step": 5032
    },
    {
      "epoch": 2.124050632911392,
      "grad_norm": 1.3856201171875,
      "learning_rate": 7.96760555954523e-05,
      "loss": 0.636816680431366,
      "step": 5034
    },
    {
      "epoch": 2.1248945147679326,
      "grad_norm": 1.1410126686096191,
      "learning_rate": 7.965672476441282e-05,
      "loss": 0.5324423313140869,
      "step": 5036
    },
    {
      "epoch": 2.1257383966244725,
      "grad_norm": 1.446070909500122,
      "learning_rate": 7.963738709213651e-05,
      "loss": 0.7433624267578125,
      "step": 5038
    },
    {
      "epoch": 2.1265822784810124,
      "grad_norm": 1.3041753768920898,
      "learning_rate": 7.961804258308419e-05,
      "loss": 0.6359145641326904,
      "step": 5040
    },
    {
      "epoch": 2.127426160337553,
      "grad_norm": 1.2043813467025757,
      "learning_rate": 7.959869124171826e-05,
      "loss": 0.6164234280586243,
      "step": 5042
    },
    {
      "epoch": 2.1282700421940928,
      "grad_norm": 1.2375630140304565,
      "learning_rate": 7.957933307250273e-05,
      "loss": 0.6437279582023621,
      "step": 5044
    },
    {
      "epoch": 2.1291139240506327,
      "grad_norm": 1.210644245147705,
      "learning_rate": 7.955996807990314e-05,
      "loss": 0.585924506187439,
      "step": 5046
    },
    {
      "epoch": 2.129957805907173,
      "grad_norm": 1.2011489868164062,
      "learning_rate": 7.954059626838661e-05,
      "loss": 0.6081803441047668,
      "step": 5048
    },
    {
      "epoch": 2.130801687763713,
      "grad_norm": 1.0365782976150513,
      "learning_rate": 7.952121764242187e-05,
      "loss": 0.5609047412872314,
      "step": 5050
    },
    {
      "epoch": 2.131645569620253,
      "grad_norm": 1.7950767278671265,
      "learning_rate": 7.950183220647918e-05,
      "loss": 0.5612874031066895,
      "step": 5052
    },
    {
      "epoch": 2.1324894514767934,
      "grad_norm": 1.2933409214019775,
      "learning_rate": 7.94824399650304e-05,
      "loss": 0.6554630994796753,
      "step": 5054
    },
    {
      "epoch": 2.1333333333333333,
      "grad_norm": 1.129828929901123,
      "learning_rate": 7.946304092254894e-05,
      "loss": 0.5623239278793335,
      "step": 5056
    },
    {
      "epoch": 2.1341772151898732,
      "grad_norm": 1.1060296297073364,
      "learning_rate": 7.944363508350978e-05,
      "loss": 0.5036910772323608,
      "step": 5058
    },
    {
      "epoch": 2.1350210970464136,
      "grad_norm": 1.2622627019882202,
      "learning_rate": 7.94242224523895e-05,
      "loss": 0.5840913653373718,
      "step": 5060
    },
    {
      "epoch": 2.1358649789029536,
      "grad_norm": 1.3803153038024902,
      "learning_rate": 7.940480303366618e-05,
      "loss": 0.6365578770637512,
      "step": 5062
    },
    {
      "epoch": 2.1367088607594935,
      "grad_norm": 1.2524651288986206,
      "learning_rate": 7.938537683181955e-05,
      "loss": 0.6167916655540466,
      "step": 5064
    },
    {
      "epoch": 2.137552742616034,
      "grad_norm": 1.3320350646972656,
      "learning_rate": 7.936594385133083e-05,
      "loss": 0.6356930732727051,
      "step": 5066
    },
    {
      "epoch": 2.138396624472574,
      "grad_norm": 1.3180949687957764,
      "learning_rate": 7.934650409668285e-05,
      "loss": 0.5888242721557617,
      "step": 5068
    },
    {
      "epoch": 2.1392405063291138,
      "grad_norm": 1.1376243829727173,
      "learning_rate": 7.932705757235999e-05,
      "loss": 0.608725905418396,
      "step": 5070
    },
    {
      "epoch": 2.140084388185654,
      "grad_norm": 1.1734369993209839,
      "learning_rate": 7.930760428284817e-05,
      "loss": 0.5824158787727356,
      "step": 5072
    },
    {
      "epoch": 2.140928270042194,
      "grad_norm": 1.1038579940795898,
      "learning_rate": 7.928814423263493e-05,
      "loss": 0.5629416704177856,
      "step": 5074
    },
    {
      "epoch": 2.141772151898734,
      "grad_norm": 1.269780158996582,
      "learning_rate": 7.926867742620929e-05,
      "loss": 0.5994445085525513,
      "step": 5076
    },
    {
      "epoch": 2.1426160337552744,
      "grad_norm": 1.2274279594421387,
      "learning_rate": 7.924920386806188e-05,
      "loss": 0.5845475792884827,
      "step": 5078
    },
    {
      "epoch": 2.1434599156118144,
      "grad_norm": 1.168766975402832,
      "learning_rate": 7.922972356268488e-05,
      "loss": 0.621201753616333,
      "step": 5080
    },
    {
      "epoch": 2.1443037974683543,
      "grad_norm": 1.0057638883590698,
      "learning_rate": 7.921023651457203e-05,
      "loss": 0.5282597541809082,
      "step": 5082
    },
    {
      "epoch": 2.1451476793248947,
      "grad_norm": 1.432309865951538,
      "learning_rate": 7.91907427282186e-05,
      "loss": 0.632583737373352,
      "step": 5084
    },
    {
      "epoch": 2.1459915611814346,
      "grad_norm": 1.3939776420593262,
      "learning_rate": 7.917124220812144e-05,
      "loss": 0.6239289045333862,
      "step": 5086
    },
    {
      "epoch": 2.1468354430379746,
      "grad_norm": 1.3741775751113892,
      "learning_rate": 7.915173495877895e-05,
      "loss": 0.5749062895774841,
      "step": 5088
    },
    {
      "epoch": 2.147679324894515,
      "grad_norm": 1.3123528957366943,
      "learning_rate": 7.913222098469109e-05,
      "loss": 0.6011738181114197,
      "step": 5090
    },
    {
      "epoch": 2.148523206751055,
      "grad_norm": 1.3473498821258545,
      "learning_rate": 7.911270029035932e-05,
      "loss": 0.5804699659347534,
      "step": 5092
    },
    {
      "epoch": 2.149367088607595,
      "grad_norm": 1.0873067378997803,
      "learning_rate": 7.909317288028673e-05,
      "loss": 0.6446103453636169,
      "step": 5094
    },
    {
      "epoch": 2.1502109704641352,
      "grad_norm": 1.1374083757400513,
      "learning_rate": 7.907363875897789e-05,
      "loss": 0.6136524677276611,
      "step": 5096
    },
    {
      "epoch": 2.151054852320675,
      "grad_norm": 1.1356533765792847,
      "learning_rate": 7.905409793093896e-05,
      "loss": 0.5107976794242859,
      "step": 5098
    },
    {
      "epoch": 2.151898734177215,
      "grad_norm": 1.2579567432403564,
      "learning_rate": 7.903455040067763e-05,
      "loss": 0.6073099374771118,
      "step": 5100
    },
    {
      "epoch": 2.151898734177215,
      "eval_loss": 0.6902023553848267,
      "eval_runtime": 733.915,
      "eval_samples_per_second": 2.871,
      "eval_steps_per_second": 2.871,
      "step": 5100
    },
    {
      "epoch": 2.1527426160337555,
      "grad_norm": 1.2401398420333862,
      "learning_rate": 7.901499617270315e-05,
      "loss": 0.5562406182289124,
      "step": 5102
    },
    {
      "epoch": 2.1535864978902954,
      "grad_norm": 1.086590051651001,
      "learning_rate": 7.899543525152628e-05,
      "loss": 0.5749467015266418,
      "step": 5104
    },
    {
      "epoch": 2.1544303797468354,
      "grad_norm": 1.206458568572998,
      "learning_rate": 7.897586764165939e-05,
      "loss": 0.6326877474784851,
      "step": 5106
    },
    {
      "epoch": 2.1552742616033758,
      "grad_norm": 1.030740737915039,
      "learning_rate": 7.895629334761632e-05,
      "loss": 0.5616445541381836,
      "step": 5108
    },
    {
      "epoch": 2.1561181434599157,
      "grad_norm": 1.3338581323623657,
      "learning_rate": 7.89367123739125e-05,
      "loss": 0.6307384371757507,
      "step": 5110
    },
    {
      "epoch": 2.1569620253164556,
      "grad_norm": 1.2684671878814697,
      "learning_rate": 7.891712472506485e-05,
      "loss": 0.6087653636932373,
      "step": 5112
    },
    {
      "epoch": 2.1578059071729956,
      "grad_norm": 1.1610581874847412,
      "learning_rate": 7.889753040559188e-05,
      "loss": 0.5747998952865601,
      "step": 5114
    },
    {
      "epoch": 2.158649789029536,
      "grad_norm": 1.4069275856018066,
      "learning_rate": 7.887792942001366e-05,
      "loss": 0.6143770217895508,
      "step": 5116
    },
    {
      "epoch": 2.159493670886076,
      "grad_norm": 1.0858227014541626,
      "learning_rate": 7.885832177285173e-05,
      "loss": 0.552534282207489,
      "step": 5118
    },
    {
      "epoch": 2.160337552742616,
      "grad_norm": 1.067070722579956,
      "learning_rate": 7.88387074686292e-05,
      "loss": 0.5781989693641663,
      "step": 5120
    },
    {
      "epoch": 2.1611814345991562,
      "grad_norm": 1.139981746673584,
      "learning_rate": 7.881908651187072e-05,
      "loss": 0.5521422624588013,
      "step": 5122
    },
    {
      "epoch": 2.162025316455696,
      "grad_norm": 1.0987457036972046,
      "learning_rate": 7.879945890710245e-05,
      "loss": 0.5755025744438171,
      "step": 5124
    },
    {
      "epoch": 2.162869198312236,
      "grad_norm": 1.1530758142471313,
      "learning_rate": 7.877982465885214e-05,
      "loss": 0.5783509612083435,
      "step": 5126
    },
    {
      "epoch": 2.1637130801687765,
      "grad_norm": 1.2285696268081665,
      "learning_rate": 7.876018377164899e-05,
      "loss": 0.5942281484603882,
      "step": 5128
    },
    {
      "epoch": 2.1645569620253164,
      "grad_norm": 1.1283711194992065,
      "learning_rate": 7.874053625002378e-05,
      "loss": 0.5539707541465759,
      "step": 5130
    },
    {
      "epoch": 2.1654008438818564,
      "grad_norm": 1.3213335275650024,
      "learning_rate": 7.872088209850885e-05,
      "loss": 0.5955292582511902,
      "step": 5132
    },
    {
      "epoch": 2.1662447257383968,
      "grad_norm": 1.1748592853546143,
      "learning_rate": 7.8701221321638e-05,
      "loss": 0.5422899723052979,
      "step": 5134
    },
    {
      "epoch": 2.1670886075949367,
      "grad_norm": 1.0752148628234863,
      "learning_rate": 7.868155392394662e-05,
      "loss": 0.5547205209732056,
      "step": 5136
    },
    {
      "epoch": 2.1679324894514767,
      "grad_norm": 1.1814554929733276,
      "learning_rate": 7.86618799099716e-05,
      "loss": 0.5938948392868042,
      "step": 5138
    },
    {
      "epoch": 2.168776371308017,
      "grad_norm": 1.3455278873443604,
      "learning_rate": 7.864219928425132e-05,
      "loss": 0.6468925476074219,
      "step": 5140
    },
    {
      "epoch": 2.169620253164557,
      "grad_norm": 1.2695354223251343,
      "learning_rate": 7.862251205132576e-05,
      "loss": 0.5704391002655029,
      "step": 5142
    },
    {
      "epoch": 2.170464135021097,
      "grad_norm": 1.1529468297958374,
      "learning_rate": 7.860281821573638e-05,
      "loss": 0.6057283878326416,
      "step": 5144
    },
    {
      "epoch": 2.1713080168776373,
      "grad_norm": 1.3461004495620728,
      "learning_rate": 7.858311778202616e-05,
      "loss": 0.6135527491569519,
      "step": 5146
    },
    {
      "epoch": 2.1721518987341772,
      "grad_norm": 1.1258536577224731,
      "learning_rate": 7.856341075473962e-05,
      "loss": 0.5585638880729675,
      "step": 5148
    },
    {
      "epoch": 2.172995780590717,
      "grad_norm": 1.254898190498352,
      "learning_rate": 7.854369713842279e-05,
      "loss": 0.5780918002128601,
      "step": 5150
    },
    {
      "epoch": 2.1738396624472576,
      "grad_norm": 1.2730201482772827,
      "learning_rate": 7.852397693762321e-05,
      "loss": 0.595267117023468,
      "step": 5152
    },
    {
      "epoch": 2.1746835443037975,
      "grad_norm": 1.1875078678131104,
      "learning_rate": 7.850425015688999e-05,
      "loss": 0.5636162161827087,
      "step": 5154
    },
    {
      "epoch": 2.1755274261603375,
      "grad_norm": 1.0930945873260498,
      "learning_rate": 7.848451680077366e-05,
      "loss": 0.6362089514732361,
      "step": 5156
    },
    {
      "epoch": 2.176371308016878,
      "grad_norm": 1.2274452447891235,
      "learning_rate": 7.846477687382639e-05,
      "loss": 0.6268675327301025,
      "step": 5158
    },
    {
      "epoch": 2.1772151898734178,
      "grad_norm": 1.2023133039474487,
      "learning_rate": 7.844503038060176e-05,
      "loss": 0.6014906167984009,
      "step": 5160
    },
    {
      "epoch": 2.1780590717299577,
      "grad_norm": 1.2616889476776123,
      "learning_rate": 7.842527732565491e-05,
      "loss": 0.6180019974708557,
      "step": 5162
    },
    {
      "epoch": 2.1789029535864977,
      "grad_norm": 1.1046907901763916,
      "learning_rate": 7.84055177135425e-05,
      "loss": 0.5400100946426392,
      "step": 5164
    },
    {
      "epoch": 2.179746835443038,
      "grad_norm": 1.1664032936096191,
      "learning_rate": 7.83857515488227e-05,
      "loss": 0.5713199973106384,
      "step": 5166
    },
    {
      "epoch": 2.180590717299578,
      "grad_norm": 1.2526558637619019,
      "learning_rate": 7.836597883605519e-05,
      "loss": 0.5741307735443115,
      "step": 5168
    },
    {
      "epoch": 2.181434599156118,
      "grad_norm": 1.0457103252410889,
      "learning_rate": 7.834619957980112e-05,
      "loss": 0.47188031673431396,
      "step": 5170
    },
    {
      "epoch": 2.1822784810126583,
      "grad_norm": 1.1978110074996948,
      "learning_rate": 7.832641378462319e-05,
      "loss": 0.6149471998214722,
      "step": 5172
    },
    {
      "epoch": 2.1831223628691983,
      "grad_norm": 1.2231460809707642,
      "learning_rate": 7.830662145508567e-05,
      "loss": 0.5520018339157104,
      "step": 5174
    },
    {
      "epoch": 2.183966244725738,
      "grad_norm": 1.4367618560791016,
      "learning_rate": 7.828682259575417e-05,
      "loss": 0.6536548733711243,
      "step": 5176
    },
    {
      "epoch": 2.1848101265822786,
      "grad_norm": 1.0891374349594116,
      "learning_rate": 7.826701721119598e-05,
      "loss": 0.5324372053146362,
      "step": 5178
    },
    {
      "epoch": 2.1856540084388185,
      "grad_norm": 1.118695616722107,
      "learning_rate": 7.82472053059798e-05,
      "loss": 0.6127952337265015,
      "step": 5180
    },
    {
      "epoch": 2.1864978902953585,
      "grad_norm": 1.1116070747375488,
      "learning_rate": 7.822738688467585e-05,
      "loss": 0.505962610244751,
      "step": 5182
    },
    {
      "epoch": 2.187341772151899,
      "grad_norm": 1.2140545845031738,
      "learning_rate": 7.820756195185586e-05,
      "loss": 0.6210073232650757,
      "step": 5184
    },
    {
      "epoch": 2.188185654008439,
      "grad_norm": 1.2135601043701172,
      "learning_rate": 7.818773051209307e-05,
      "loss": 0.6517674326896667,
      "step": 5186
    },
    {
      "epoch": 2.1890295358649787,
      "grad_norm": 1.3875514268875122,
      "learning_rate": 7.816789256996218e-05,
      "loss": 0.5577492117881775,
      "step": 5188
    },
    {
      "epoch": 2.189873417721519,
      "grad_norm": 1.181325912475586,
      "learning_rate": 7.814804813003949e-05,
      "loss": 0.6010199189186096,
      "step": 5190
    },
    {
      "epoch": 2.190717299578059,
      "grad_norm": 1.102044701576233,
      "learning_rate": 7.812819719690265e-05,
      "loss": 0.5635302662849426,
      "step": 5192
    },
    {
      "epoch": 2.191561181434599,
      "grad_norm": 1.4227958917617798,
      "learning_rate": 7.810833977513094e-05,
      "loss": 0.5804321765899658,
      "step": 5194
    },
    {
      "epoch": 2.1924050632911394,
      "grad_norm": 1.2573446035385132,
      "learning_rate": 7.80884758693051e-05,
      "loss": 0.6005555987358093,
      "step": 5196
    },
    {
      "epoch": 2.1932489451476793,
      "grad_norm": 1.3534085750579834,
      "learning_rate": 7.80686054840073e-05,
      "loss": 0.6263643503189087,
      "step": 5198
    },
    {
      "epoch": 2.1940928270042193,
      "grad_norm": 1.6895852088928223,
      "learning_rate": 7.804872862382131e-05,
      "loss": 0.6235764622688293,
      "step": 5200
    },
    {
      "epoch": 2.1940928270042193,
      "eval_loss": 0.6915348172187805,
      "eval_runtime": 1167.9782,
      "eval_samples_per_second": 1.804,
      "eval_steps_per_second": 1.804,
      "step": 5200
    },
    {
      "epoch": 2.1949367088607596,
      "grad_norm": 1.138973593711853,
      "learning_rate": 7.802884529333227e-05,
      "loss": 0.5586035847663879,
      "step": 5202
    },
    {
      "epoch": 2.1957805907172996,
      "grad_norm": 1.3664026260375977,
      "learning_rate": 7.800895549712697e-05,
      "loss": 0.5768917202949524,
      "step": 5204
    },
    {
      "epoch": 2.1966244725738395,
      "grad_norm": 1.2182449102401733,
      "learning_rate": 7.798905923979353e-05,
      "loss": 0.6046215891838074,
      "step": 5206
    },
    {
      "epoch": 2.19746835443038,
      "grad_norm": 1.2692211866378784,
      "learning_rate": 7.796915652592167e-05,
      "loss": 0.5412904024124146,
      "step": 5208
    },
    {
      "epoch": 2.19831223628692,
      "grad_norm": 1.200822114944458,
      "learning_rate": 7.794924736010256e-05,
      "loss": 0.5328584909439087,
      "step": 5210
    },
    {
      "epoch": 2.19915611814346,
      "grad_norm": 1.1093779802322388,
      "learning_rate": 7.792933174692886e-05,
      "loss": 0.5497913360595703,
      "step": 5212
    },
    {
      "epoch": 2.2,
      "grad_norm": 1.3838921785354614,
      "learning_rate": 7.790940969099471e-05,
      "loss": 0.5908066034317017,
      "step": 5214
    },
    {
      "epoch": 2.20084388185654,
      "grad_norm": 1.1411913633346558,
      "learning_rate": 7.788948119689576e-05,
      "loss": 0.6117307543754578,
      "step": 5216
    },
    {
      "epoch": 2.20168776371308,
      "grad_norm": 1.5668916702270508,
      "learning_rate": 7.786954626922913e-05,
      "loss": 0.5788605809211731,
      "step": 5218
    },
    {
      "epoch": 2.2025316455696204,
      "grad_norm": 1.195027232170105,
      "learning_rate": 7.784960491259344e-05,
      "loss": 0.5948591828346252,
      "step": 5220
    },
    {
      "epoch": 2.2033755274261604,
      "grad_norm": 1.2665271759033203,
      "learning_rate": 7.782965713158872e-05,
      "loss": 0.6321669220924377,
      "step": 5222
    },
    {
      "epoch": 2.2042194092827003,
      "grad_norm": 1.123711109161377,
      "learning_rate": 7.78097029308166e-05,
      "loss": 0.5853859186172485,
      "step": 5224
    },
    {
      "epoch": 2.2050632911392407,
      "grad_norm": 1.9381071329116821,
      "learning_rate": 7.77897423148801e-05,
      "loss": 0.6485977172851562,
      "step": 5226
    },
    {
      "epoch": 2.2059071729957807,
      "grad_norm": 1.4062265157699585,
      "learning_rate": 7.776977528838376e-05,
      "loss": 0.6243517398834229,
      "step": 5228
    },
    {
      "epoch": 2.2067510548523206,
      "grad_norm": 1.2127182483673096,
      "learning_rate": 7.774980185593358e-05,
      "loss": 0.5770578980445862,
      "step": 5230
    },
    {
      "epoch": 2.207594936708861,
      "grad_norm": 1.250847578048706,
      "learning_rate": 7.772982202213709e-05,
      "loss": 0.6521194577217102,
      "step": 5232
    },
    {
      "epoch": 2.208438818565401,
      "grad_norm": 1.2568131685256958,
      "learning_rate": 7.77098357916032e-05,
      "loss": 0.5755271911621094,
      "step": 5234
    },
    {
      "epoch": 2.209282700421941,
      "grad_norm": 1.2422975301742554,
      "learning_rate": 7.768984316894236e-05,
      "loss": 0.5486469864845276,
      "step": 5236
    },
    {
      "epoch": 2.2101265822784812,
      "grad_norm": 1.1018635034561157,
      "learning_rate": 7.766984415876652e-05,
      "loss": 0.5512928366661072,
      "step": 5238
    },
    {
      "epoch": 2.210970464135021,
      "grad_norm": 1.2261123657226562,
      "learning_rate": 7.764983876568903e-05,
      "loss": 0.5753499269485474,
      "step": 5240
    },
    {
      "epoch": 2.211814345991561,
      "grad_norm": 1.2222342491149902,
      "learning_rate": 7.762982699432474e-05,
      "loss": 0.5404848456382751,
      "step": 5242
    },
    {
      "epoch": 2.212658227848101,
      "grad_norm": 1.231494426727295,
      "learning_rate": 7.760980884929004e-05,
      "loss": 0.5999218821525574,
      "step": 5244
    },
    {
      "epoch": 2.2135021097046415,
      "grad_norm": 1.1530078649520874,
      "learning_rate": 7.758978433520268e-05,
      "loss": 0.6123101115226746,
      "step": 5246
    },
    {
      "epoch": 2.2143459915611814,
      "grad_norm": 1.182706594467163,
      "learning_rate": 7.756975345668194e-05,
      "loss": 0.5945886969566345,
      "step": 5248
    },
    {
      "epoch": 2.2151898734177213,
      "grad_norm": 1.0788652896881104,
      "learning_rate": 7.754971621834857e-05,
      "loss": 0.5698213577270508,
      "step": 5250
    },
    {
      "epoch": 2.2160337552742617,
      "grad_norm": 1.2243359088897705,
      "learning_rate": 7.752967262482477e-05,
      "loss": 0.5959678888320923,
      "step": 5252
    },
    {
      "epoch": 2.2168776371308017,
      "grad_norm": 1.4292869567871094,
      "learning_rate": 7.750962268073421e-05,
      "loss": 0.586794376373291,
      "step": 5254
    },
    {
      "epoch": 2.2177215189873416,
      "grad_norm": 1.1809570789337158,
      "learning_rate": 7.748956639070204e-05,
      "loss": 0.5513298511505127,
      "step": 5256
    },
    {
      "epoch": 2.218565400843882,
      "grad_norm": 1.485813856124878,
      "learning_rate": 7.746950375935484e-05,
      "loss": 0.6402831673622131,
      "step": 5258
    },
    {
      "epoch": 2.219409282700422,
      "grad_norm": 1.0851374864578247,
      "learning_rate": 7.744943479132069e-05,
      "loss": 0.5729117393493652,
      "step": 5260
    },
    {
      "epoch": 2.220253164556962,
      "grad_norm": 1.4308949708938599,
      "learning_rate": 7.742935949122911e-05,
      "loss": 0.6239725947380066,
      "step": 5262
    },
    {
      "epoch": 2.2210970464135023,
      "grad_norm": 1.379258155822754,
      "learning_rate": 7.740927786371107e-05,
      "loss": 0.6260181069374084,
      "step": 5264
    },
    {
      "epoch": 2.221940928270042,
      "grad_norm": 1.1661925315856934,
      "learning_rate": 7.738918991339905e-05,
      "loss": 0.6074157357215881,
      "step": 5266
    },
    {
      "epoch": 2.222784810126582,
      "grad_norm": 1.168901801109314,
      "learning_rate": 7.736909564492694e-05,
      "loss": 0.6119515895843506,
      "step": 5268
    },
    {
      "epoch": 2.2236286919831225,
      "grad_norm": 1.1451057195663452,
      "learning_rate": 7.734899506293008e-05,
      "loss": 0.5505842566490173,
      "step": 5270
    },
    {
      "epoch": 2.2244725738396625,
      "grad_norm": 1.2303991317749023,
      "learning_rate": 7.732888817204533e-05,
      "loss": 0.6117991805076599,
      "step": 5272
    },
    {
      "epoch": 2.2253164556962024,
      "grad_norm": 1.04572331905365,
      "learning_rate": 7.730877497691092e-05,
      "loss": 0.5589770078659058,
      "step": 5274
    },
    {
      "epoch": 2.226160337552743,
      "grad_norm": 1.2047234773635864,
      "learning_rate": 7.72886554821666e-05,
      "loss": 0.6288654208183289,
      "step": 5276
    },
    {
      "epoch": 2.2270042194092827,
      "grad_norm": 1.2036652565002441,
      "learning_rate": 7.726852969245355e-05,
      "loss": 0.6174501776695251,
      "step": 5278
    },
    {
      "epoch": 2.2278481012658227,
      "grad_norm": 1.1740167140960693,
      "learning_rate": 7.72483976124144e-05,
      "loss": 0.6027677655220032,
      "step": 5280
    },
    {
      "epoch": 2.228691983122363,
      "grad_norm": 1.0600008964538574,
      "learning_rate": 7.722825924669326e-05,
      "loss": 0.6016151309013367,
      "step": 5282
    },
    {
      "epoch": 2.229535864978903,
      "grad_norm": 1.2631008625030518,
      "learning_rate": 7.720811459993562e-05,
      "loss": 0.5905849933624268,
      "step": 5284
    },
    {
      "epoch": 2.230379746835443,
      "grad_norm": 1.1024738550186157,
      "learning_rate": 7.718796367678848e-05,
      "loss": 0.5129587054252625,
      "step": 5286
    },
    {
      "epoch": 2.2312236286919833,
      "grad_norm": 1.23116934299469,
      "learning_rate": 7.716780648190028e-05,
      "loss": 0.5709586143493652,
      "step": 5288
    },
    {
      "epoch": 2.2320675105485233,
      "grad_norm": 1.2739102840423584,
      "learning_rate": 7.714764301992088e-05,
      "loss": 0.5454761385917664,
      "step": 5290
    },
    {
      "epoch": 2.232911392405063,
      "grad_norm": 1.303963303565979,
      "learning_rate": 7.712747329550162e-05,
      "loss": 0.537248969078064,
      "step": 5292
    },
    {
      "epoch": 2.233755274261603,
      "grad_norm": 1.2454309463500977,
      "learning_rate": 7.710729731329529e-05,
      "loss": 0.6364415884017944,
      "step": 5294
    },
    {
      "epoch": 2.2345991561181435,
      "grad_norm": 1.2401882410049438,
      "learning_rate": 7.708711507795605e-05,
      "loss": 0.5640100240707397,
      "step": 5296
    },
    {
      "epoch": 2.2354430379746835,
      "grad_norm": 1.197432041168213,
      "learning_rate": 7.706692659413959e-05,
      "loss": 0.5919729471206665,
      "step": 5298
    },
    {
      "epoch": 2.2362869198312234,
      "grad_norm": 1.1779764890670776,
      "learning_rate": 7.704673186650298e-05,
      "loss": 0.5569849014282227,
      "step": 5300
    },
    {
      "epoch": 2.2362869198312234,
      "eval_loss": 0.6898328065872192,
      "eval_runtime": 739.3794,
      "eval_samples_per_second": 2.85,
      "eval_steps_per_second": 2.85,
      "step": 5300
    },
    {
      "epoch": 2.237130801687764,
      "grad_norm": 1.1371463537216187,
      "learning_rate": 7.702653089970479e-05,
      "loss": 0.5823061466217041,
      "step": 5302
    },
    {
      "epoch": 2.2379746835443037,
      "grad_norm": 1.1877846717834473,
      "learning_rate": 7.700632369840497e-05,
      "loss": 0.5556252002716064,
      "step": 5304
    },
    {
      "epoch": 2.2388185654008437,
      "grad_norm": 1.1580896377563477,
      "learning_rate": 7.698611026726492e-05,
      "loss": 0.5794119834899902,
      "step": 5306
    },
    {
      "epoch": 2.239662447257384,
      "grad_norm": 1.29141366481781,
      "learning_rate": 7.696589061094755e-05,
      "loss": 0.5828680396080017,
      "step": 5308
    },
    {
      "epoch": 2.240506329113924,
      "grad_norm": 1.1286728382110596,
      "learning_rate": 7.694566473411706e-05,
      "loss": 0.6161736845970154,
      "step": 5310
    },
    {
      "epoch": 2.241350210970464,
      "grad_norm": 1.0969985723495483,
      "learning_rate": 7.692543264143925e-05,
      "loss": 0.570767879486084,
      "step": 5312
    },
    {
      "epoch": 2.2421940928270043,
      "grad_norm": 1.2902227640151978,
      "learning_rate": 7.690519433758123e-05,
      "loss": 0.631476104259491,
      "step": 5314
    },
    {
      "epoch": 2.2430379746835443,
      "grad_norm": 1.432735800743103,
      "learning_rate": 7.68849498272116e-05,
      "loss": 0.6142309904098511,
      "step": 5316
    },
    {
      "epoch": 2.243881856540084,
      "grad_norm": 1.0824161767959595,
      "learning_rate": 7.686469911500038e-05,
      "loss": 0.5871514081954956,
      "step": 5318
    },
    {
      "epoch": 2.2447257383966246,
      "grad_norm": 1.1694978475570679,
      "learning_rate": 7.684444220561902e-05,
      "loss": 0.6144557595252991,
      "step": 5320
    },
    {
      "epoch": 2.2455696202531645,
      "grad_norm": 1.2981040477752686,
      "learning_rate": 7.68241791037404e-05,
      "loss": 0.6049425601959229,
      "step": 5322
    },
    {
      "epoch": 2.2464135021097045,
      "grad_norm": 1.132128357887268,
      "learning_rate": 7.680390981403885e-05,
      "loss": 0.5571867823600769,
      "step": 5324
    },
    {
      "epoch": 2.247257383966245,
      "grad_norm": 1.1760079860687256,
      "learning_rate": 7.678363434119005e-05,
      "loss": 0.5710517168045044,
      "step": 5326
    },
    {
      "epoch": 2.248101265822785,
      "grad_norm": 1.1918572187423706,
      "learning_rate": 7.67633526898712e-05,
      "loss": 0.5508866906166077,
      "step": 5328
    },
    {
      "epoch": 2.2489451476793247,
      "grad_norm": 1.1837294101715088,
      "learning_rate": 7.674306486476091e-05,
      "loss": 0.6242696046829224,
      "step": 5330
    },
    {
      "epoch": 2.249789029535865,
      "grad_norm": 1.384918212890625,
      "learning_rate": 7.672277087053914e-05,
      "loss": 0.5821678042411804,
      "step": 5332
    },
    {
      "epoch": 2.250632911392405,
      "grad_norm": 1.1248877048492432,
      "learning_rate": 7.670247071188738e-05,
      "loss": 0.5415928363800049,
      "step": 5334
    },
    {
      "epoch": 2.251476793248945,
      "grad_norm": 1.228140950202942,
      "learning_rate": 7.668216439348843e-05,
      "loss": 0.5475174188613892,
      "step": 5336
    },
    {
      "epoch": 2.2523206751054854,
      "grad_norm": 1.3816046714782715,
      "learning_rate": 7.666185192002662e-05,
      "loss": 0.5793306231498718,
      "step": 5338
    },
    {
      "epoch": 2.2531645569620253,
      "grad_norm": 1.2446565628051758,
      "learning_rate": 7.664153329618759e-05,
      "loss": 0.6221131682395935,
      "step": 5340
    },
    {
      "epoch": 2.2540084388185653,
      "grad_norm": 1.1677669286727905,
      "learning_rate": 7.662120852665852e-05,
      "loss": 0.5403847694396973,
      "step": 5342
    },
    {
      "epoch": 2.2548523206751057,
      "grad_norm": 1.2485873699188232,
      "learning_rate": 7.66008776161279e-05,
      "loss": 0.620201587677002,
      "step": 5344
    },
    {
      "epoch": 2.2556962025316456,
      "grad_norm": 1.2486802339553833,
      "learning_rate": 7.658054056928568e-05,
      "loss": 0.5969216227531433,
      "step": 5346
    },
    {
      "epoch": 2.2565400843881855,
      "grad_norm": 1.2621372938156128,
      "learning_rate": 7.656019739082326e-05,
      "loss": 0.6376339793205261,
      "step": 5348
    },
    {
      "epoch": 2.257383966244726,
      "grad_norm": 1.238633155822754,
      "learning_rate": 7.65398480854334e-05,
      "loss": 0.6374872326850891,
      "step": 5350
    },
    {
      "epoch": 2.258227848101266,
      "grad_norm": 1.3031803369522095,
      "learning_rate": 7.651949265781029e-05,
      "loss": 0.6348551511764526,
      "step": 5352
    },
    {
      "epoch": 2.259071729957806,
      "grad_norm": 1.3735158443450928,
      "learning_rate": 7.649913111264952e-05,
      "loss": 0.6267750859260559,
      "step": 5354
    },
    {
      "epoch": 2.259915611814346,
      "grad_norm": 1.1227772235870361,
      "learning_rate": 7.647876345464817e-05,
      "loss": 0.623030960559845,
      "step": 5356
    },
    {
      "epoch": 2.260759493670886,
      "grad_norm": 1.4555678367614746,
      "learning_rate": 7.645838968850459e-05,
      "loss": 0.5810713171958923,
      "step": 5358
    },
    {
      "epoch": 2.261603375527426,
      "grad_norm": 1.227725863456726,
      "learning_rate": 7.643800981891867e-05,
      "loss": 0.6150093078613281,
      "step": 5360
    },
    {
      "epoch": 2.2624472573839665,
      "grad_norm": 1.0648300647735596,
      "learning_rate": 7.641762385059161e-05,
      "loss": 0.5350445508956909,
      "step": 5362
    },
    {
      "epoch": 2.2632911392405064,
      "grad_norm": 1.179452896118164,
      "learning_rate": 7.639723178822613e-05,
      "loss": 0.6253421306610107,
      "step": 5364
    },
    {
      "epoch": 2.2641350210970463,
      "grad_norm": 1.0983240604400635,
      "learning_rate": 7.637683363652621e-05,
      "loss": 0.5512562990188599,
      "step": 5366
    },
    {
      "epoch": 2.2649789029535867,
      "grad_norm": 1.1825451850891113,
      "learning_rate": 7.635642940019736e-05,
      "loss": 0.5584151148796082,
      "step": 5368
    },
    {
      "epoch": 2.2658227848101267,
      "grad_norm": 1.1022000312805176,
      "learning_rate": 7.633601908394643e-05,
      "loss": 0.5881790518760681,
      "step": 5370
    },
    {
      "epoch": 2.2666666666666666,
      "grad_norm": 1.1935697793960571,
      "learning_rate": 7.631560269248169e-05,
      "loss": 0.6060683131217957,
      "step": 5372
    },
    {
      "epoch": 2.267510548523207,
      "grad_norm": 1.1174103021621704,
      "learning_rate": 7.62951802305128e-05,
      "loss": 0.5877062678337097,
      "step": 5374
    },
    {
      "epoch": 2.268354430379747,
      "grad_norm": 1.3934977054595947,
      "learning_rate": 7.627475170275086e-05,
      "loss": 0.5145504474639893,
      "step": 5376
    },
    {
      "epoch": 2.269198312236287,
      "grad_norm": 1.2637842893600464,
      "learning_rate": 7.625431711390831e-05,
      "loss": 0.6194025874137878,
      "step": 5378
    },
    {
      "epoch": 2.270042194092827,
      "grad_norm": 1.2034388780593872,
      "learning_rate": 7.623387646869902e-05,
      "loss": 0.6205627918243408,
      "step": 5380
    },
    {
      "epoch": 2.270886075949367,
      "grad_norm": 0.953880250453949,
      "learning_rate": 7.621342977183826e-05,
      "loss": 0.5609696507453918,
      "step": 5382
    },
    {
      "epoch": 2.271729957805907,
      "grad_norm": 1.2841949462890625,
      "learning_rate": 7.619297702804272e-05,
      "loss": 0.6044906377792358,
      "step": 5384
    },
    {
      "epoch": 2.272573839662447,
      "grad_norm": 1.146804690361023,
      "learning_rate": 7.617251824203037e-05,
      "loss": 0.5420435667037964,
      "step": 5386
    },
    {
      "epoch": 2.2734177215189875,
      "grad_norm": 1.2225698232650757,
      "learning_rate": 7.615205341852076e-05,
      "loss": 0.6230710744857788,
      "step": 5388
    },
    {
      "epoch": 2.2742616033755274,
      "grad_norm": 1.3423371315002441,
      "learning_rate": 7.613158256223467e-05,
      "loss": 0.6486349701881409,
      "step": 5390
    },
    {
      "epoch": 2.2751054852320673,
      "grad_norm": 1.0840023756027222,
      "learning_rate": 7.611110567789435e-05,
      "loss": 0.6527825593948364,
      "step": 5392
    },
    {
      "epoch": 2.2759493670886077,
      "grad_norm": 1.342466950416565,
      "learning_rate": 7.609062277022341e-05,
      "loss": 0.6859483122825623,
      "step": 5394
    },
    {
      "epoch": 2.2767932489451477,
      "grad_norm": 1.0406129360198975,
      "learning_rate": 7.607013384394691e-05,
      "loss": 0.5536003708839417,
      "step": 5396
    },
    {
      "epoch": 2.2776371308016876,
      "grad_norm": 1.0853544473648071,
      "learning_rate": 7.604963890379118e-05,
      "loss": 0.5488654971122742,
      "step": 5398
    },
    {
      "epoch": 2.278481012658228,
      "grad_norm": 1.0330145359039307,
      "learning_rate": 7.602913795448407e-05,
      "loss": 0.6072142720222473,
      "step": 5400
    },
    {
      "epoch": 2.278481012658228,
      "eval_loss": 0.6875645518302917,
      "eval_runtime": 861.3558,
      "eval_samples_per_second": 2.446,
      "eval_steps_per_second": 2.446,
      "step": 5400
    },
    {
      "epoch": 2.279324894514768,
      "grad_norm": 1.1858742237091064,
      "learning_rate": 7.600863100075472e-05,
      "loss": 0.5420109033584595,
      "step": 5402
    },
    {
      "epoch": 2.280168776371308,
      "grad_norm": 1.2126039266586304,
      "learning_rate": 7.598811804733373e-05,
      "loss": 0.6109243631362915,
      "step": 5404
    },
    {
      "epoch": 2.2810126582278483,
      "grad_norm": 1.1290241479873657,
      "learning_rate": 7.5967599098953e-05,
      "loss": 0.5889696478843689,
      "step": 5406
    },
    {
      "epoch": 2.281856540084388,
      "grad_norm": 1.320263147354126,
      "learning_rate": 7.594707416034586e-05,
      "loss": 0.6548630595207214,
      "step": 5408
    },
    {
      "epoch": 2.282700421940928,
      "grad_norm": 1.346169114112854,
      "learning_rate": 7.592654323624703e-05,
      "loss": 0.6556787490844727,
      "step": 5410
    },
    {
      "epoch": 2.2835443037974685,
      "grad_norm": 1.2104716300964355,
      "learning_rate": 7.590600633139265e-05,
      "loss": 0.5631673336029053,
      "step": 5412
    },
    {
      "epoch": 2.2843881856540085,
      "grad_norm": 1.3298237323760986,
      "learning_rate": 7.58854634505201e-05,
      "loss": 0.5931088328361511,
      "step": 5414
    },
    {
      "epoch": 2.2852320675105484,
      "grad_norm": 1.4201204776763916,
      "learning_rate": 7.586491459836829e-05,
      "loss": 0.6966755986213684,
      "step": 5416
    },
    {
      "epoch": 2.286075949367089,
      "grad_norm": 1.253135323524475,
      "learning_rate": 7.584435977967743e-05,
      "loss": 0.6172569394111633,
      "step": 5418
    },
    {
      "epoch": 2.2869198312236287,
      "grad_norm": 1.133144736289978,
      "learning_rate": 7.582379899918911e-05,
      "loss": 0.5376655459403992,
      "step": 5420
    },
    {
      "epoch": 2.2877637130801687,
      "grad_norm": 1.1103745698928833,
      "learning_rate": 7.580323226164632e-05,
      "loss": 0.6138498187065125,
      "step": 5422
    },
    {
      "epoch": 2.2886075949367086,
      "grad_norm": 1.091636300086975,
      "learning_rate": 7.57826595717934e-05,
      "loss": 0.5049096345901489,
      "step": 5424
    },
    {
      "epoch": 2.289451476793249,
      "grad_norm": 1.2486571073532104,
      "learning_rate": 7.57620809343761e-05,
      "loss": 0.5666115283966064,
      "step": 5426
    },
    {
      "epoch": 2.290295358649789,
      "grad_norm": 1.510684847831726,
      "learning_rate": 7.57414963541415e-05,
      "loss": 0.49512919783592224,
      "step": 5428
    },
    {
      "epoch": 2.291139240506329,
      "grad_norm": 1.1142191886901855,
      "learning_rate": 7.572090583583805e-05,
      "loss": 0.558807373046875,
      "step": 5430
    },
    {
      "epoch": 2.2919831223628693,
      "grad_norm": 1.1162657737731934,
      "learning_rate": 7.57003093842156e-05,
      "loss": 0.6245265603065491,
      "step": 5432
    },
    {
      "epoch": 2.292827004219409,
      "grad_norm": 1.2784614562988281,
      "learning_rate": 7.567970700402537e-05,
      "loss": 0.5505527853965759,
      "step": 5434
    },
    {
      "epoch": 2.293670886075949,
      "grad_norm": 1.3142638206481934,
      "learning_rate": 7.565909870001992e-05,
      "loss": 0.6137702465057373,
      "step": 5436
    },
    {
      "epoch": 2.2945147679324895,
      "grad_norm": 1.072805404663086,
      "learning_rate": 7.563848447695318e-05,
      "loss": 0.540766716003418,
      "step": 5438
    },
    {
      "epoch": 2.2953586497890295,
      "grad_norm": 1.2861377000808716,
      "learning_rate": 7.561786433958048e-05,
      "loss": 0.6806555986404419,
      "step": 5440
    },
    {
      "epoch": 2.2962025316455694,
      "grad_norm": 1.3193045854568481,
      "learning_rate": 7.559723829265847e-05,
      "loss": 0.6191258430480957,
      "step": 5442
    },
    {
      "epoch": 2.29704641350211,
      "grad_norm": 1.1969127655029297,
      "learning_rate": 7.55766063409452e-05,
      "loss": 0.6067718863487244,
      "step": 5444
    },
    {
      "epoch": 2.2978902953586497,
      "grad_norm": 1.2129666805267334,
      "learning_rate": 7.555596848920006e-05,
      "loss": 0.5673627257347107,
      "step": 5446
    },
    {
      "epoch": 2.2987341772151897,
      "grad_norm": 1.1639961004257202,
      "learning_rate": 7.553532474218379e-05,
      "loss": 0.61825031042099,
      "step": 5448
    },
    {
      "epoch": 2.29957805907173,
      "grad_norm": 1.3893283605575562,
      "learning_rate": 7.551467510465852e-05,
      "loss": 0.6096790432929993,
      "step": 5450
    },
    {
      "epoch": 2.30042194092827,
      "grad_norm": 1.0708417892456055,
      "learning_rate": 7.549401958138772e-05,
      "loss": 0.6121414303779602,
      "step": 5452
    },
    {
      "epoch": 2.30126582278481,
      "grad_norm": 1.3299298286437988,
      "learning_rate": 7.547335817713624e-05,
      "loss": 0.6504668593406677,
      "step": 5454
    },
    {
      "epoch": 2.3021097046413503,
      "grad_norm": 1.3594682216644287,
      "learning_rate": 7.545269089667022e-05,
      "loss": 0.5761144161224365,
      "step": 5456
    },
    {
      "epoch": 2.3029535864978903,
      "grad_norm": 1.1089586019515991,
      "learning_rate": 7.543201774475726e-05,
      "loss": 0.5457773804664612,
      "step": 5458
    },
    {
      "epoch": 2.3037974683544302,
      "grad_norm": 1.3472918272018433,
      "learning_rate": 7.541133872616624e-05,
      "loss": 0.6014775037765503,
      "step": 5460
    },
    {
      "epoch": 2.3046413502109706,
      "grad_norm": 1.2757689952850342,
      "learning_rate": 7.53906538456674e-05,
      "loss": 0.6246467232704163,
      "step": 5462
    },
    {
      "epoch": 2.3054852320675105,
      "grad_norm": 1.4598166942596436,
      "learning_rate": 7.536996310803236e-05,
      "loss": 0.6583935022354126,
      "step": 5464
    },
    {
      "epoch": 2.3063291139240505,
      "grad_norm": 1.2861602306365967,
      "learning_rate": 7.534926651803407e-05,
      "loss": 0.562523603439331,
      "step": 5466
    },
    {
      "epoch": 2.307172995780591,
      "grad_norm": 1.0953221321105957,
      "learning_rate": 7.532856408044684e-05,
      "loss": 0.6093505620956421,
      "step": 5468
    },
    {
      "epoch": 2.308016877637131,
      "grad_norm": 1.0982829332351685,
      "learning_rate": 7.530785580004631e-05,
      "loss": 0.6196447014808655,
      "step": 5470
    },
    {
      "epoch": 2.3088607594936708,
      "grad_norm": 1.2224280834197998,
      "learning_rate": 7.52871416816095e-05,
      "loss": 0.6360989212989807,
      "step": 5472
    },
    {
      "epoch": 2.309704641350211,
      "grad_norm": 1.244486927986145,
      "learning_rate": 7.526642172991476e-05,
      "loss": 0.6189543008804321,
      "step": 5474
    },
    {
      "epoch": 2.310548523206751,
      "grad_norm": 1.2408053874969482,
      "learning_rate": 7.524569594974178e-05,
      "loss": 0.6137582659721375,
      "step": 5476
    },
    {
      "epoch": 2.311392405063291,
      "grad_norm": 1.3323272466659546,
      "learning_rate": 7.522496434587157e-05,
      "loss": 0.6462169289588928,
      "step": 5478
    },
    {
      "epoch": 2.3122362869198314,
      "grad_norm": 1.1076425313949585,
      "learning_rate": 7.520422692308657e-05,
      "loss": 0.5495362877845764,
      "step": 5480
    },
    {
      "epoch": 2.3130801687763713,
      "grad_norm": 1.3298509120941162,
      "learning_rate": 7.518348368617046e-05,
      "loss": 0.5560636520385742,
      "step": 5482
    },
    {
      "epoch": 2.3139240506329113,
      "grad_norm": 1.0740195512771606,
      "learning_rate": 7.516273463990832e-05,
      "loss": 0.5763371586799622,
      "step": 5484
    },
    {
      "epoch": 2.3147679324894517,
      "grad_norm": 1.0748567581176758,
      "learning_rate": 7.514197978908657e-05,
      "loss": 0.5111498832702637,
      "step": 5486
    },
    {
      "epoch": 2.3156118143459916,
      "grad_norm": 1.2047218084335327,
      "learning_rate": 7.512121913849294e-05,
      "loss": 0.6599951982498169,
      "step": 5488
    },
    {
      "epoch": 2.3164556962025316,
      "grad_norm": 1.2956700325012207,
      "learning_rate": 7.510045269291651e-05,
      "loss": 0.6409770846366882,
      "step": 5490
    },
    {
      "epoch": 2.317299578059072,
      "grad_norm": 1.241860032081604,
      "learning_rate": 7.50796804571477e-05,
      "loss": 0.5967662334442139,
      "step": 5492
    },
    {
      "epoch": 2.318143459915612,
      "grad_norm": 1.1612682342529297,
      "learning_rate": 7.50589024359783e-05,
      "loss": 0.5856342315673828,
      "step": 5494
    },
    {
      "epoch": 2.318987341772152,
      "grad_norm": 1.0895500183105469,
      "learning_rate": 7.503811863420135e-05,
      "loss": 0.5652023553848267,
      "step": 5496
    },
    {
      "epoch": 2.319831223628692,
      "grad_norm": 1.3374481201171875,
      "learning_rate": 7.50173290566113e-05,
      "loss": 0.6777268648147583,
      "step": 5498
    },
    {
      "epoch": 2.320675105485232,
      "grad_norm": 1.192614197731018,
      "learning_rate": 7.499653370800391e-05,
      "loss": 0.6052314043045044,
      "step": 5500
    },
    {
      "epoch": 2.320675105485232,
      "eval_loss": 0.6867148876190186,
      "eval_runtime": 941.3545,
      "eval_samples_per_second": 2.238,
      "eval_steps_per_second": 2.238,
      "step": 5500
    },
    {
      "epoch": 2.321518987341772,
      "grad_norm": 1.1008832454681396,
      "learning_rate": 7.497573259317625e-05,
      "loss": 0.5208253860473633,
      "step": 5502
    },
    {
      "epoch": 2.3223628691983125,
      "grad_norm": 1.2141541242599487,
      "learning_rate": 7.495492571692677e-05,
      "loss": 0.6352296471595764,
      "step": 5504
    },
    {
      "epoch": 2.3232067510548524,
      "grad_norm": 1.2588802576065063,
      "learning_rate": 7.493411308405517e-05,
      "loss": 0.6132256388664246,
      "step": 5506
    },
    {
      "epoch": 2.3240506329113924,
      "grad_norm": 1.348765254020691,
      "learning_rate": 7.491329469936258e-05,
      "loss": 0.571265697479248,
      "step": 5508
    },
    {
      "epoch": 2.3248945147679323,
      "grad_norm": 1.266377329826355,
      "learning_rate": 7.489247056765135e-05,
      "loss": 0.5433708429336548,
      "step": 5510
    },
    {
      "epoch": 2.3257383966244727,
      "grad_norm": 1.2920128107070923,
      "learning_rate": 7.487164069372523e-05,
      "loss": 0.6193158030509949,
      "step": 5512
    },
    {
      "epoch": 2.3265822784810126,
      "grad_norm": 1.068169116973877,
      "learning_rate": 7.485080508238928e-05,
      "loss": 0.5817977786064148,
      "step": 5514
    },
    {
      "epoch": 2.3274261603375526,
      "grad_norm": 1.2941710948944092,
      "learning_rate": 7.482996373844985e-05,
      "loss": 0.6558082103729248,
      "step": 5516
    },
    {
      "epoch": 2.328270042194093,
      "grad_norm": 1.2143336534500122,
      "learning_rate": 7.480911666671467e-05,
      "loss": 0.5569961667060852,
      "step": 5518
    },
    {
      "epoch": 2.329113924050633,
      "grad_norm": 1.3364789485931396,
      "learning_rate": 7.478826387199274e-05,
      "loss": 0.6497300863265991,
      "step": 5520
    },
    {
      "epoch": 2.329957805907173,
      "grad_norm": 1.057530403137207,
      "learning_rate": 7.47674053590944e-05,
      "loss": 0.5793087482452393,
      "step": 5522
    },
    {
      "epoch": 2.330801687763713,
      "grad_norm": 1.1543176174163818,
      "learning_rate": 7.47465411328313e-05,
      "loss": 0.5583140850067139,
      "step": 5524
    },
    {
      "epoch": 2.331645569620253,
      "grad_norm": 1.3409180641174316,
      "learning_rate": 7.472567119801645e-05,
      "loss": 0.6318784952163696,
      "step": 5526
    },
    {
      "epoch": 2.332489451476793,
      "grad_norm": 1.2899413108825684,
      "learning_rate": 7.47047955594641e-05,
      "loss": 0.5950855612754822,
      "step": 5528
    },
    {
      "epoch": 2.3333333333333335,
      "grad_norm": 1.329220175743103,
      "learning_rate": 7.468391422198989e-05,
      "loss": 0.6181023716926575,
      "step": 5530
    },
    {
      "epoch": 2.3341772151898734,
      "grad_norm": 1.202129602432251,
      "learning_rate": 7.466302719041073e-05,
      "loss": 0.6384578943252563,
      "step": 5532
    },
    {
      "epoch": 2.3350210970464134,
      "grad_norm": 1.1890549659729004,
      "learning_rate": 7.464213446954487e-05,
      "loss": 0.6059293746948242,
      "step": 5534
    },
    {
      "epoch": 2.3358649789029537,
      "grad_norm": 1.2041429281234741,
      "learning_rate": 7.462123606421183e-05,
      "loss": 0.6432797908782959,
      "step": 5536
    },
    {
      "epoch": 2.3367088607594937,
      "grad_norm": 1.3827080726623535,
      "learning_rate": 7.460033197923249e-05,
      "loss": 0.6796717047691345,
      "step": 5538
    },
    {
      "epoch": 2.3375527426160336,
      "grad_norm": 1.2323482036590576,
      "learning_rate": 7.457942221942903e-05,
      "loss": 0.5772476196289062,
      "step": 5540
    },
    {
      "epoch": 2.338396624472574,
      "grad_norm": 1.2011388540267944,
      "learning_rate": 7.455850678962493e-05,
      "loss": 0.5964269042015076,
      "step": 5542
    },
    {
      "epoch": 2.339240506329114,
      "grad_norm": 1.1133569478988647,
      "learning_rate": 7.453758569464495e-05,
      "loss": 0.6416608095169067,
      "step": 5544
    },
    {
      "epoch": 2.340084388185654,
      "grad_norm": 1.1257679462432861,
      "learning_rate": 7.451665893931521e-05,
      "loss": 0.5668829679489136,
      "step": 5546
    },
    {
      "epoch": 2.3409282700421943,
      "grad_norm": 1.3494724035263062,
      "learning_rate": 7.449572652846311e-05,
      "loss": 0.6029916405677795,
      "step": 5548
    },
    {
      "epoch": 2.3417721518987342,
      "grad_norm": 1.2199759483337402,
      "learning_rate": 7.447478846691735e-05,
      "loss": 0.6336984634399414,
      "step": 5550
    },
    {
      "epoch": 2.342616033755274,
      "grad_norm": 1.2806570529937744,
      "learning_rate": 7.445384475950792e-05,
      "loss": 0.579140305519104,
      "step": 5552
    },
    {
      "epoch": 2.343459915611814,
      "grad_norm": 0.9874221086502075,
      "learning_rate": 7.443289541106616e-05,
      "loss": 0.6061640381813049,
      "step": 5554
    },
    {
      "epoch": 2.3443037974683545,
      "grad_norm": 1.2271486520767212,
      "learning_rate": 7.441194042642467e-05,
      "loss": 0.5502339601516724,
      "step": 5556
    },
    {
      "epoch": 2.3451476793248944,
      "grad_norm": 1.2522462606430054,
      "learning_rate": 7.439097981041738e-05,
      "loss": 0.5774438381195068,
      "step": 5558
    },
    {
      "epoch": 2.3459915611814344,
      "grad_norm": 1.267204761505127,
      "learning_rate": 7.437001356787945e-05,
      "loss": 0.6091527342796326,
      "step": 5560
    },
    {
      "epoch": 2.3468354430379748,
      "grad_norm": 1.1711935997009277,
      "learning_rate": 7.434904170364747e-05,
      "loss": 0.5443631410598755,
      "step": 5562
    },
    {
      "epoch": 2.3476793248945147,
      "grad_norm": 1.085097074508667,
      "learning_rate": 7.432806422255918e-05,
      "loss": 0.5255029201507568,
      "step": 5564
    },
    {
      "epoch": 2.3485232067510546,
      "grad_norm": 1.3244949579238892,
      "learning_rate": 7.430708112945369e-05,
      "loss": 0.5197238922119141,
      "step": 5566
    },
    {
      "epoch": 2.349367088607595,
      "grad_norm": 1.3646879196166992,
      "learning_rate": 7.428609242917141e-05,
      "loss": 0.5576170682907104,
      "step": 5568
    },
    {
      "epoch": 2.350210970464135,
      "grad_norm": 1.339190125465393,
      "learning_rate": 7.426509812655406e-05,
      "loss": 0.6254662275314331,
      "step": 5570
    },
    {
      "epoch": 2.351054852320675,
      "grad_norm": 1.4624155759811401,
      "learning_rate": 7.424409822644457e-05,
      "loss": 0.6593500375747681,
      "step": 5572
    },
    {
      "epoch": 2.3518987341772153,
      "grad_norm": 1.1931114196777344,
      "learning_rate": 7.422309273368722e-05,
      "loss": 0.6102238297462463,
      "step": 5574
    },
    {
      "epoch": 2.3527426160337552,
      "grad_norm": 1.789340615272522,
      "learning_rate": 7.420208165312762e-05,
      "loss": 0.6695854067802429,
      "step": 5576
    },
    {
      "epoch": 2.353586497890295,
      "grad_norm": 1.2364262342453003,
      "learning_rate": 7.418106498961258e-05,
      "loss": 0.578844428062439,
      "step": 5578
    },
    {
      "epoch": 2.3544303797468356,
      "grad_norm": 1.1568509340286255,
      "learning_rate": 7.416004274799027e-05,
      "loss": 0.5717503428459167,
      "step": 5580
    },
    {
      "epoch": 2.3552742616033755,
      "grad_norm": 1.1744630336761475,
      "learning_rate": 7.413901493311009e-05,
      "loss": 0.6170201897621155,
      "step": 5582
    },
    {
      "epoch": 2.3561181434599154,
      "grad_norm": 1.0684332847595215,
      "learning_rate": 7.411798154982275e-05,
      "loss": 0.6482691764831543,
      "step": 5584
    },
    {
      "epoch": 2.356962025316456,
      "grad_norm": 1.046196460723877,
      "learning_rate": 7.409694260298025e-05,
      "loss": 0.572839617729187,
      "step": 5586
    },
    {
      "epoch": 2.3578059071729958,
      "grad_norm": 1.0110210180282593,
      "learning_rate": 7.407589809743591e-05,
      "loss": 0.5645976662635803,
      "step": 5588
    },
    {
      "epoch": 2.3586497890295357,
      "grad_norm": 1.0801016092300415,
      "learning_rate": 7.405484803804425e-05,
      "loss": 0.5653133392333984,
      "step": 5590
    },
    {
      "epoch": 2.359493670886076,
      "grad_norm": 1.0934380292892456,
      "learning_rate": 7.403379242966116e-05,
      "loss": 0.5972150564193726,
      "step": 5592
    },
    {
      "epoch": 2.360337552742616,
      "grad_norm": 1.3722410202026367,
      "learning_rate": 7.40127312771437e-05,
      "loss": 0.5927542448043823,
      "step": 5594
    },
    {
      "epoch": 2.361181434599156,
      "grad_norm": 1.1567236185073853,
      "learning_rate": 7.399166458535032e-05,
      "loss": 0.547027051448822,
      "step": 5596
    },
    {
      "epoch": 2.3620253164556964,
      "grad_norm": 1.2254211902618408,
      "learning_rate": 7.397059235914067e-05,
      "loss": 0.5356617569923401,
      "step": 5598
    },
    {
      "epoch": 2.3628691983122363,
      "grad_norm": 1.1529103517532349,
      "learning_rate": 7.394951460337575e-05,
      "loss": 0.5424175262451172,
      "step": 5600
    },
    {
      "epoch": 2.3628691983122363,
      "eval_loss": 0.6851074695587158,
      "eval_runtime": 938.5536,
      "eval_samples_per_second": 2.245,
      "eval_steps_per_second": 2.245,
      "step": 5600
    },
    {
      "epoch": 2.3637130801687762,
      "grad_norm": 1.2050299644470215,
      "learning_rate": 7.392843132291777e-05,
      "loss": 0.5834107398986816,
      "step": 5602
    },
    {
      "epoch": 2.3645569620253166,
      "grad_norm": 1.264567494392395,
      "learning_rate": 7.390734252263024e-05,
      "loss": 0.5445035099983215,
      "step": 5604
    },
    {
      "epoch": 2.3654008438818566,
      "grad_norm": 1.357791781425476,
      "learning_rate": 7.388624820737791e-05,
      "loss": 0.6207653880119324,
      "step": 5606
    },
    {
      "epoch": 2.3662447257383965,
      "grad_norm": 1.2246928215026855,
      "learning_rate": 7.386514838202689e-05,
      "loss": 0.6628696322441101,
      "step": 5608
    },
    {
      "epoch": 2.367088607594937,
      "grad_norm": 1.1455399990081787,
      "learning_rate": 7.384404305144447e-05,
      "loss": 0.5870704054832458,
      "step": 5610
    },
    {
      "epoch": 2.367932489451477,
      "grad_norm": 1.2338638305664062,
      "learning_rate": 7.382293222049925e-05,
      "loss": 0.6160538792610168,
      "step": 5612
    },
    {
      "epoch": 2.3687763713080168,
      "grad_norm": 1.231271505355835,
      "learning_rate": 7.38018158940611e-05,
      "loss": 0.6274036765098572,
      "step": 5614
    },
    {
      "epoch": 2.369620253164557,
      "grad_norm": 1.022050380706787,
      "learning_rate": 7.378069407700114e-05,
      "loss": 0.5623515248298645,
      "step": 5616
    },
    {
      "epoch": 2.370464135021097,
      "grad_norm": 1.2040951251983643,
      "learning_rate": 7.375956677419178e-05,
      "loss": 0.5505564212799072,
      "step": 5618
    },
    {
      "epoch": 2.371308016877637,
      "grad_norm": 1.1754523515701294,
      "learning_rate": 7.373843399050668e-05,
      "loss": 0.6537002921104431,
      "step": 5620
    },
    {
      "epoch": 2.3721518987341774,
      "grad_norm": 1.1710485219955444,
      "learning_rate": 7.371729573082073e-05,
      "loss": 0.6224458813667297,
      "step": 5622
    },
    {
      "epoch": 2.3729957805907174,
      "grad_norm": 1.1629483699798584,
      "learning_rate": 7.36961520000102e-05,
      "loss": 0.6297177076339722,
      "step": 5624
    },
    {
      "epoch": 2.3738396624472573,
      "grad_norm": 1.1069440841674805,
      "learning_rate": 7.367500280295248e-05,
      "loss": 0.5202008485794067,
      "step": 5626
    },
    {
      "epoch": 2.3746835443037977,
      "grad_norm": 1.0068297386169434,
      "learning_rate": 7.36538481445263e-05,
      "loss": 0.5256102681159973,
      "step": 5628
    },
    {
      "epoch": 2.3755274261603376,
      "grad_norm": 1.1103417873382568,
      "learning_rate": 7.363268802961161e-05,
      "loss": 0.5460903644561768,
      "step": 5630
    },
    {
      "epoch": 2.3763713080168776,
      "grad_norm": 1.2885268926620483,
      "learning_rate": 7.361152246308969e-05,
      "loss": 0.5817124247550964,
      "step": 5632
    },
    {
      "epoch": 2.377215189873418,
      "grad_norm": 1.233831524848938,
      "learning_rate": 7.359035144984302e-05,
      "loss": 0.5415143966674805,
      "step": 5634
    },
    {
      "epoch": 2.378059071729958,
      "grad_norm": 1.3451908826828003,
      "learning_rate": 7.35691749947553e-05,
      "loss": 0.6837685108184814,
      "step": 5636
    },
    {
      "epoch": 2.378902953586498,
      "grad_norm": 1.1320621967315674,
      "learning_rate": 7.354799310271159e-05,
      "loss": 0.5966196656227112,
      "step": 5638
    },
    {
      "epoch": 2.379746835443038,
      "grad_norm": 1.1884461641311646,
      "learning_rate": 7.35268057785981e-05,
      "loss": 0.5607479214668274,
      "step": 5640
    },
    {
      "epoch": 2.380590717299578,
      "grad_norm": 1.2710856199264526,
      "learning_rate": 7.350561302730236e-05,
      "loss": 0.595242977142334,
      "step": 5642
    },
    {
      "epoch": 2.381434599156118,
      "grad_norm": 1.3110458850860596,
      "learning_rate": 7.348441485371314e-05,
      "loss": 0.6208752393722534,
      "step": 5644
    },
    {
      "epoch": 2.382278481012658,
      "grad_norm": 1.1734380722045898,
      "learning_rate": 7.346321126272044e-05,
      "loss": 0.6173125505447388,
      "step": 5646
    },
    {
      "epoch": 2.3831223628691984,
      "grad_norm": 1.2024762630462646,
      "learning_rate": 7.34420022592155e-05,
      "loss": 0.6013050675392151,
      "step": 5648
    },
    {
      "epoch": 2.3839662447257384,
      "grad_norm": 1.1305288076400757,
      "learning_rate": 7.342078784809086e-05,
      "loss": 0.5919594764709473,
      "step": 5650
    },
    {
      "epoch": 2.3848101265822783,
      "grad_norm": 1.075323462486267,
      "learning_rate": 7.339956803424028e-05,
      "loss": 0.5399283766746521,
      "step": 5652
    },
    {
      "epoch": 2.3856540084388187,
      "grad_norm": 1.2035599946975708,
      "learning_rate": 7.337834282255873e-05,
      "loss": 0.6253576874732971,
      "step": 5654
    },
    {
      "epoch": 2.3864978902953586,
      "grad_norm": 1.0572105646133423,
      "learning_rate": 7.335711221794251e-05,
      "loss": 0.5247007608413696,
      "step": 5656
    },
    {
      "epoch": 2.3873417721518986,
      "grad_norm": 1.2701191902160645,
      "learning_rate": 7.333587622528906e-05,
      "loss": 0.5800243020057678,
      "step": 5658
    },
    {
      "epoch": 2.388185654008439,
      "grad_norm": 1.1772741079330444,
      "learning_rate": 7.331463484949716e-05,
      "loss": 0.589645504951477,
      "step": 5660
    },
    {
      "epoch": 2.389029535864979,
      "grad_norm": 1.0562703609466553,
      "learning_rate": 7.329338809546674e-05,
      "loss": 0.5820419192314148,
      "step": 5662
    },
    {
      "epoch": 2.389873417721519,
      "grad_norm": 1.1634355783462524,
      "learning_rate": 7.327213596809906e-05,
      "loss": 0.591435432434082,
      "step": 5664
    },
    {
      "epoch": 2.3907172995780592,
      "grad_norm": 1.2220302820205688,
      "learning_rate": 7.325087847229655e-05,
      "loss": 0.5630883574485779,
      "step": 5666
    },
    {
      "epoch": 2.391561181434599,
      "grad_norm": 1.4087659120559692,
      "learning_rate": 7.322961561296294e-05,
      "loss": 0.6050130128860474,
      "step": 5668
    },
    {
      "epoch": 2.392405063291139,
      "grad_norm": 1.1126172542572021,
      "learning_rate": 7.320834739500313e-05,
      "loss": 0.56146240234375,
      "step": 5670
    },
    {
      "epoch": 2.3932489451476795,
      "grad_norm": 0.99373859167099,
      "learning_rate": 7.31870738233233e-05,
      "loss": 0.5507852435112,
      "step": 5672
    },
    {
      "epoch": 2.3940928270042194,
      "grad_norm": 1.14408540725708,
      "learning_rate": 7.316579490283085e-05,
      "loss": 0.5895347595214844,
      "step": 5674
    },
    {
      "epoch": 2.3949367088607594,
      "grad_norm": 1.1728581190109253,
      "learning_rate": 7.314451063843443e-05,
      "loss": 0.5304404497146606,
      "step": 5676
    },
    {
      "epoch": 2.3957805907172998,
      "grad_norm": 1.1721378564834595,
      "learning_rate": 7.31232210350439e-05,
      "loss": 0.5805793404579163,
      "step": 5678
    },
    {
      "epoch": 2.3966244725738397,
      "grad_norm": 1.0499866008758545,
      "learning_rate": 7.310192609757038e-05,
      "loss": 0.5671767592430115,
      "step": 5680
    },
    {
      "epoch": 2.3974683544303796,
      "grad_norm": 1.0959177017211914,
      "learning_rate": 7.308062583092617e-05,
      "loss": 0.6335723400115967,
      "step": 5682
    },
    {
      "epoch": 2.3983122362869196,
      "grad_norm": 1.31142258644104,
      "learning_rate": 7.305932024002487e-05,
      "loss": 0.6032374501228333,
      "step": 5684
    },
    {
      "epoch": 2.39915611814346,
      "grad_norm": 0.9212818741798401,
      "learning_rate": 7.303800932978124e-05,
      "loss": 0.5492936372756958,
      "step": 5686
    },
    {
      "epoch": 2.4,
      "grad_norm": 1.1956428289413452,
      "learning_rate": 7.301669310511132e-05,
      "loss": 0.5533297061920166,
      "step": 5688
    },
    {
      "epoch": 2.40084388185654,
      "grad_norm": 1.4048634767532349,
      "learning_rate": 7.299537157093232e-05,
      "loss": 0.5859368443489075,
      "step": 5690
    },
    {
      "epoch": 2.4016877637130802,
      "grad_norm": 1.0580679178237915,
      "learning_rate": 7.297404473216277e-05,
      "loss": 0.5099439024925232,
      "step": 5692
    },
    {
      "epoch": 2.40253164556962,
      "grad_norm": 1.2450575828552246,
      "learning_rate": 7.29527125937223e-05,
      "loss": 0.5631486177444458,
      "step": 5694
    },
    {
      "epoch": 2.40337552742616,
      "grad_norm": 1.338466763496399,
      "learning_rate": 7.293137516053187e-05,
      "loss": 0.6045404672622681,
      "step": 5696
    },
    {
      "epoch": 2.4042194092827005,
      "grad_norm": 1.198588252067566,
      "learning_rate": 7.291003243751358e-05,
      "loss": 0.6063475608825684,
      "step": 5698
    },
    {
      "epoch": 2.4050632911392404,
      "grad_norm": 1.2315080165863037,
      "learning_rate": 7.288868442959081e-05,
      "loss": 0.5734809041023254,
      "step": 5700
    },
    {
      "epoch": 2.4050632911392404,
      "eval_loss": 0.6841402053833008,
      "eval_runtime": 941.6641,
      "eval_samples_per_second": 2.238,
      "eval_steps_per_second": 2.238,
      "step": 5700
    },
    {
      "epoch": 2.4059071729957804,
      "grad_norm": 1.1494885683059692,
      "learning_rate": 7.286733114168812e-05,
      "loss": 0.5744594931602478,
      "step": 5702
    },
    {
      "epoch": 2.4067510548523208,
      "grad_norm": 1.3769505023956299,
      "learning_rate": 7.284597257873132e-05,
      "loss": 0.611789882183075,
      "step": 5704
    },
    {
      "epoch": 2.4075949367088607,
      "grad_norm": 1.2326449155807495,
      "learning_rate": 7.28246087456474e-05,
      "loss": 0.6091431975364685,
      "step": 5706
    },
    {
      "epoch": 2.4084388185654007,
      "grad_norm": 1.1960830688476562,
      "learning_rate": 7.28032396473646e-05,
      "loss": 0.49431973695755005,
      "step": 5708
    },
    {
      "epoch": 2.409282700421941,
      "grad_norm": 1.1672827005386353,
      "learning_rate": 7.278186528881237e-05,
      "loss": 0.5344718098640442,
      "step": 5710
    },
    {
      "epoch": 2.410126582278481,
      "grad_norm": 1.1923719644546509,
      "learning_rate": 7.276048567492136e-05,
      "loss": 0.6011165380477905,
      "step": 5712
    },
    {
      "epoch": 2.410970464135021,
      "grad_norm": 1.2314990758895874,
      "learning_rate": 7.273910081062341e-05,
      "loss": 0.6300925016403198,
      "step": 5714
    },
    {
      "epoch": 2.4118143459915613,
      "grad_norm": 0.8976680040359497,
      "learning_rate": 7.27177107008516e-05,
      "loss": 0.56329345703125,
      "step": 5716
    },
    {
      "epoch": 2.4126582278481012,
      "grad_norm": 1.2954038381576538,
      "learning_rate": 7.269631535054026e-05,
      "loss": 0.6266427040100098,
      "step": 5718
    },
    {
      "epoch": 2.413502109704641,
      "grad_norm": 1.3357585668563843,
      "learning_rate": 7.267491476462485e-05,
      "loss": 0.6234018802642822,
      "step": 5720
    },
    {
      "epoch": 2.4143459915611816,
      "grad_norm": 1.1913645267486572,
      "learning_rate": 7.265350894804209e-05,
      "loss": 0.5909059047698975,
      "step": 5722
    },
    {
      "epoch": 2.4151898734177215,
      "grad_norm": 1.3425955772399902,
      "learning_rate": 7.263209790572986e-05,
      "loss": 0.5708479285240173,
      "step": 5724
    },
    {
      "epoch": 2.4160337552742615,
      "grad_norm": 1.2258507013320923,
      "learning_rate": 7.261068164262734e-05,
      "loss": 0.5810034871101379,
      "step": 5726
    },
    {
      "epoch": 2.416877637130802,
      "grad_norm": 1.348794937133789,
      "learning_rate": 7.258926016367479e-05,
      "loss": 0.5939235687255859,
      "step": 5728
    },
    {
      "epoch": 2.4177215189873418,
      "grad_norm": 1.0896574258804321,
      "learning_rate": 7.256783347381375e-05,
      "loss": 0.6298259496688843,
      "step": 5730
    },
    {
      "epoch": 2.4185654008438817,
      "grad_norm": 1.164866328239441,
      "learning_rate": 7.254640157798696e-05,
      "loss": 0.5277430415153503,
      "step": 5732
    },
    {
      "epoch": 2.419409282700422,
      "grad_norm": 1.1215453147888184,
      "learning_rate": 7.252496448113833e-05,
      "loss": 0.5724055767059326,
      "step": 5734
    },
    {
      "epoch": 2.420253164556962,
      "grad_norm": 1.0640764236450195,
      "learning_rate": 7.2503522188213e-05,
      "loss": 0.5439977645874023,
      "step": 5736
    },
    {
      "epoch": 2.421097046413502,
      "grad_norm": 1.4874604940414429,
      "learning_rate": 7.248207470415729e-05,
      "loss": 0.7568614482879639,
      "step": 5738
    },
    {
      "epoch": 2.4219409282700424,
      "grad_norm": 1.2611099481582642,
      "learning_rate": 7.246062203391873e-05,
      "loss": 0.6389632225036621,
      "step": 5740
    },
    {
      "epoch": 2.4227848101265823,
      "grad_norm": 1.185644507408142,
      "learning_rate": 7.243916418244602e-05,
      "loss": 0.6180628538131714,
      "step": 5742
    },
    {
      "epoch": 2.4236286919831223,
      "grad_norm": 1.1648430824279785,
      "learning_rate": 7.241770115468909e-05,
      "loss": 0.619799017906189,
      "step": 5744
    },
    {
      "epoch": 2.4244725738396626,
      "grad_norm": 1.1974445581436157,
      "learning_rate": 7.239623295559903e-05,
      "loss": 0.6446201205253601,
      "step": 5746
    },
    {
      "epoch": 2.4253164556962026,
      "grad_norm": 1.140477180480957,
      "learning_rate": 7.237475959012818e-05,
      "loss": 0.5839580297470093,
      "step": 5748
    },
    {
      "epoch": 2.4261603375527425,
      "grad_norm": 1.1374423503875732,
      "learning_rate": 7.235328106322998e-05,
      "loss": 0.48815420269966125,
      "step": 5750
    },
    {
      "epoch": 2.427004219409283,
      "grad_norm": 1.411432147026062,
      "learning_rate": 7.233179737985916e-05,
      "loss": 0.638519287109375,
      "step": 5752
    },
    {
      "epoch": 2.427848101265823,
      "grad_norm": 1.1232497692108154,
      "learning_rate": 7.231030854497157e-05,
      "loss": 0.5776677131652832,
      "step": 5754
    },
    {
      "epoch": 2.428691983122363,
      "grad_norm": 1.0815738439559937,
      "learning_rate": 7.228881456352428e-05,
      "loss": 0.5297027230262756,
      "step": 5756
    },
    {
      "epoch": 2.429535864978903,
      "grad_norm": 1.2230733633041382,
      "learning_rate": 7.226731544047553e-05,
      "loss": 0.5630011558532715,
      "step": 5758
    },
    {
      "epoch": 2.430379746835443,
      "grad_norm": 1.2033147811889648,
      "learning_rate": 7.224581118078476e-05,
      "loss": 0.5772101283073425,
      "step": 5760
    },
    {
      "epoch": 2.431223628691983,
      "grad_norm": 1.2150053977966309,
      "learning_rate": 7.22243017894126e-05,
      "loss": 0.5412847399711609,
      "step": 5762
    },
    {
      "epoch": 2.4320675105485234,
      "grad_norm": 1.0494824647903442,
      "learning_rate": 7.220278727132083e-05,
      "loss": 0.5568405389785767,
      "step": 5764
    },
    {
      "epoch": 2.4329113924050634,
      "grad_norm": 1.2803306579589844,
      "learning_rate": 7.218126763147244e-05,
      "loss": 0.6022217869758606,
      "step": 5766
    },
    {
      "epoch": 2.4337552742616033,
      "grad_norm": 1.0832798480987549,
      "learning_rate": 7.215974287483163e-05,
      "loss": 0.5568796396255493,
      "step": 5768
    },
    {
      "epoch": 2.4345991561181437,
      "grad_norm": 1.1829264163970947,
      "learning_rate": 7.213821300636372e-05,
      "loss": 0.5607990026473999,
      "step": 5770
    },
    {
      "epoch": 2.4354430379746836,
      "grad_norm": 2.3017473220825195,
      "learning_rate": 7.211667803103523e-05,
      "loss": 0.6382274031639099,
      "step": 5772
    },
    {
      "epoch": 2.4362869198312236,
      "grad_norm": 1.1701387166976929,
      "learning_rate": 7.209513795381388e-05,
      "loss": 0.5748776793479919,
      "step": 5774
    },
    {
      "epoch": 2.4371308016877635,
      "grad_norm": 1.0480856895446777,
      "learning_rate": 7.207359277966856e-05,
      "loss": 0.5760934352874756,
      "step": 5776
    },
    {
      "epoch": 2.437974683544304,
      "grad_norm": 1.2263693809509277,
      "learning_rate": 7.20520425135693e-05,
      "loss": 0.6387208104133606,
      "step": 5778
    },
    {
      "epoch": 2.438818565400844,
      "grad_norm": 1.219246506690979,
      "learning_rate": 7.203048716048737e-05,
      "loss": 0.6078037619590759,
      "step": 5780
    },
    {
      "epoch": 2.439662447257384,
      "grad_norm": 1.2452640533447266,
      "learning_rate": 7.200892672539515e-05,
      "loss": 0.606924831867218,
      "step": 5782
    },
    {
      "epoch": 2.440506329113924,
      "grad_norm": 1.3469732999801636,
      "learning_rate": 7.198736121326621e-05,
      "loss": 0.585297703742981,
      "step": 5784
    },
    {
      "epoch": 2.441350210970464,
      "grad_norm": 1.151127576828003,
      "learning_rate": 7.196579062907533e-05,
      "loss": 0.5849902033805847,
      "step": 5786
    },
    {
      "epoch": 2.442194092827004,
      "grad_norm": 1.0669564008712769,
      "learning_rate": 7.19442149777984e-05,
      "loss": 0.6150397062301636,
      "step": 5788
    },
    {
      "epoch": 2.4430379746835444,
      "grad_norm": 1.1700209379196167,
      "learning_rate": 7.192263426441252e-05,
      "loss": 0.6324567794799805,
      "step": 5790
    },
    {
      "epoch": 2.4438818565400844,
      "grad_norm": 1.2832094430923462,
      "learning_rate": 7.190104849389597e-05,
      "loss": 0.6202381253242493,
      "step": 5792
    },
    {
      "epoch": 2.4447257383966243,
      "grad_norm": 1.2046177387237549,
      "learning_rate": 7.187945767122813e-05,
      "loss": 0.6156684756278992,
      "step": 5794
    },
    {
      "epoch": 2.4455696202531647,
      "grad_norm": 1.031133770942688,
      "learning_rate": 7.185786180138961e-05,
      "loss": 0.5763497352600098,
      "step": 5796
    },
    {
      "epoch": 2.4464135021097047,
      "grad_norm": 1.2803475856781006,
      "learning_rate": 7.183626088936216e-05,
      "loss": 0.5419677495956421,
      "step": 5798
    },
    {
      "epoch": 2.4472573839662446,
      "grad_norm": 1.2407588958740234,
      "learning_rate": 7.181465494012869e-05,
      "loss": 0.629108190536499,
      "step": 5800
    },
    {
      "epoch": 2.4472573839662446,
      "eval_loss": 0.6835155487060547,
      "eval_runtime": 758.407,
      "eval_samples_per_second": 2.778,
      "eval_steps_per_second": 2.778,
      "step": 5800
    },
    {
      "epoch": 2.448101265822785,
      "grad_norm": 1.3525878190994263,
      "learning_rate": 7.17930439586733e-05,
      "loss": 0.6146516799926758,
      "step": 5802
    },
    {
      "epoch": 2.448945147679325,
      "grad_norm": 1.255921721458435,
      "learning_rate": 7.177142794998121e-05,
      "loss": 0.5796315670013428,
      "step": 5804
    },
    {
      "epoch": 2.449789029535865,
      "grad_norm": 1.2135448455810547,
      "learning_rate": 7.174980691903881e-05,
      "loss": 0.5978766679763794,
      "step": 5806
    },
    {
      "epoch": 2.4506329113924052,
      "grad_norm": 1.117942214012146,
      "learning_rate": 7.172818087083367e-05,
      "loss": 0.5941054821014404,
      "step": 5808
    },
    {
      "epoch": 2.451476793248945,
      "grad_norm": 1.2917672395706177,
      "learning_rate": 7.17065498103545e-05,
      "loss": 0.6213865876197815,
      "step": 5810
    },
    {
      "epoch": 2.452320675105485,
      "grad_norm": 1.2287952899932861,
      "learning_rate": 7.168491374259118e-05,
      "loss": 0.627090573310852,
      "step": 5812
    },
    {
      "epoch": 2.453164556962025,
      "grad_norm": 1.2427480220794678,
      "learning_rate": 7.16632726725347e-05,
      "loss": 0.605871319770813,
      "step": 5814
    },
    {
      "epoch": 2.4540084388185655,
      "grad_norm": 1.2568929195404053,
      "learning_rate": 7.16416266051773e-05,
      "loss": 0.5961518883705139,
      "step": 5816
    },
    {
      "epoch": 2.4548523206751054,
      "grad_norm": 1.2202998399734497,
      "learning_rate": 7.161997554551226e-05,
      "loss": 0.585054874420166,
      "step": 5818
    },
    {
      "epoch": 2.4556962025316453,
      "grad_norm": 1.2326043844223022,
      "learning_rate": 7.159831949853409e-05,
      "loss": 0.6219096779823303,
      "step": 5820
    },
    {
      "epoch": 2.4565400843881857,
      "grad_norm": 1.2161623239517212,
      "learning_rate": 7.15766584692384e-05,
      "loss": 0.641189455986023,
      "step": 5822
    },
    {
      "epoch": 2.4573839662447257,
      "grad_norm": 1.2391023635864258,
      "learning_rate": 7.1554992462622e-05,
      "loss": 0.577190101146698,
      "step": 5824
    },
    {
      "epoch": 2.4582278481012656,
      "grad_norm": 1.0883333683013916,
      "learning_rate": 7.153332148368281e-05,
      "loss": 0.5264694690704346,
      "step": 5826
    },
    {
      "epoch": 2.459071729957806,
      "grad_norm": 1.2129524946212769,
      "learning_rate": 7.15116455374199e-05,
      "loss": 0.631437361240387,
      "step": 5828
    },
    {
      "epoch": 2.459915611814346,
      "grad_norm": 1.0476374626159668,
      "learning_rate": 7.148996462883352e-05,
      "loss": 0.5025489926338196,
      "step": 5830
    },
    {
      "epoch": 2.460759493670886,
      "grad_norm": 1.1389570236206055,
      "learning_rate": 7.146827876292502e-05,
      "loss": 0.5903586745262146,
      "step": 5832
    },
    {
      "epoch": 2.4616033755274263,
      "grad_norm": 1.4385539293289185,
      "learning_rate": 7.14465879446969e-05,
      "loss": 0.633786141872406,
      "step": 5834
    },
    {
      "epoch": 2.462447257383966,
      "grad_norm": 1.1184585094451904,
      "learning_rate": 7.142489217915283e-05,
      "loss": 0.5889136791229248,
      "step": 5836
    },
    {
      "epoch": 2.463291139240506,
      "grad_norm": 1.2257685661315918,
      "learning_rate": 7.140319147129763e-05,
      "loss": 0.5774597525596619,
      "step": 5838
    },
    {
      "epoch": 2.4641350210970465,
      "grad_norm": 0.9524238109588623,
      "learning_rate": 7.13814858261372e-05,
      "loss": 0.5220611095428467,
      "step": 5840
    },
    {
      "epoch": 2.4649789029535865,
      "grad_norm": 1.2814422845840454,
      "learning_rate": 7.135977524867861e-05,
      "loss": 0.5724858641624451,
      "step": 5842
    },
    {
      "epoch": 2.4658227848101264,
      "grad_norm": 1.0978140830993652,
      "learning_rate": 7.133805974393013e-05,
      "loss": 0.5469759702682495,
      "step": 5844
    },
    {
      "epoch": 2.466666666666667,
      "grad_norm": 1.310279130935669,
      "learning_rate": 7.131633931690104e-05,
      "loss": 0.6554312705993652,
      "step": 5846
    },
    {
      "epoch": 2.4675105485232067,
      "grad_norm": 1.286189317703247,
      "learning_rate": 7.129461397260187e-05,
      "loss": 0.6166019439697266,
      "step": 5848
    },
    {
      "epoch": 2.4683544303797467,
      "grad_norm": 1.1586377620697021,
      "learning_rate": 7.127288371604424e-05,
      "loss": 0.6301121711730957,
      "step": 5850
    },
    {
      "epoch": 2.469198312236287,
      "grad_norm": 1.1684564352035522,
      "learning_rate": 7.125114855224087e-05,
      "loss": 0.6022663712501526,
      "step": 5852
    },
    {
      "epoch": 2.470042194092827,
      "grad_norm": 1.182511329650879,
      "learning_rate": 7.122940848620567e-05,
      "loss": 0.5959302186965942,
      "step": 5854
    },
    {
      "epoch": 2.470886075949367,
      "grad_norm": 1.2383002042770386,
      "learning_rate": 7.120766352295366e-05,
      "loss": 0.6251413822174072,
      "step": 5856
    },
    {
      "epoch": 2.4717299578059073,
      "grad_norm": 1.2001979351043701,
      "learning_rate": 7.118591366750097e-05,
      "loss": 0.6332544088363647,
      "step": 5858
    },
    {
      "epoch": 2.4725738396624473,
      "grad_norm": 1.2166392803192139,
      "learning_rate": 7.116415892486488e-05,
      "loss": 0.5797795057296753,
      "step": 5860
    },
    {
      "epoch": 2.473417721518987,
      "grad_norm": 1.2235382795333862,
      "learning_rate": 7.114239930006379e-05,
      "loss": 0.5335313081741333,
      "step": 5862
    },
    {
      "epoch": 2.4742616033755276,
      "grad_norm": 1.2405973672866821,
      "learning_rate": 7.112063479811724e-05,
      "loss": 0.5536905527114868,
      "step": 5864
    },
    {
      "epoch": 2.4751054852320675,
      "grad_norm": 1.116328477859497,
      "learning_rate": 7.109886542404585e-05,
      "loss": 0.554654061794281,
      "step": 5866
    },
    {
      "epoch": 2.4759493670886075,
      "grad_norm": 1.2757837772369385,
      "learning_rate": 7.107709118287143e-05,
      "loss": 0.6017873287200928,
      "step": 5868
    },
    {
      "epoch": 2.476793248945148,
      "grad_norm": 1.3445937633514404,
      "learning_rate": 7.105531207961686e-05,
      "loss": 0.6479908227920532,
      "step": 5870
    },
    {
      "epoch": 2.477637130801688,
      "grad_norm": 1.1464542150497437,
      "learning_rate": 7.103352811930619e-05,
      "loss": 0.5829157829284668,
      "step": 5872
    },
    {
      "epoch": 2.4784810126582277,
      "grad_norm": 1.3275130987167358,
      "learning_rate": 7.101173930696453e-05,
      "loss": 0.54380863904953,
      "step": 5874
    },
    {
      "epoch": 2.479324894514768,
      "grad_norm": 1.006990909576416,
      "learning_rate": 7.098994564761813e-05,
      "loss": 0.6313910484313965,
      "step": 5876
    },
    {
      "epoch": 2.480168776371308,
      "grad_norm": 1.1358299255371094,
      "learning_rate": 7.09681471462944e-05,
      "loss": 0.5343483090400696,
      "step": 5878
    },
    {
      "epoch": 2.481012658227848,
      "grad_norm": 1.1456117630004883,
      "learning_rate": 7.094634380802184e-05,
      "loss": 0.49450409412384033,
      "step": 5880
    },
    {
      "epoch": 2.4818565400843884,
      "grad_norm": 1.2961846590042114,
      "learning_rate": 7.092453563783003e-05,
      "loss": 0.6378757357597351,
      "step": 5882
    },
    {
      "epoch": 2.4827004219409283,
      "grad_norm": 0.983889102935791,
      "learning_rate": 7.090272264074972e-05,
      "loss": 0.5937124490737915,
      "step": 5884
    },
    {
      "epoch": 2.4835443037974683,
      "grad_norm": 1.0205817222595215,
      "learning_rate": 7.088090482181273e-05,
      "loss": 0.5301283597946167,
      "step": 5886
    },
    {
      "epoch": 2.4843881856540087,
      "grad_norm": 1.1721397638320923,
      "learning_rate": 7.085908218605204e-05,
      "loss": 0.6191756129264832,
      "step": 5888
    },
    {
      "epoch": 2.4852320675105486,
      "grad_norm": 1.2432814836502075,
      "learning_rate": 7.083725473850168e-05,
      "loss": 0.5928890109062195,
      "step": 5890
    },
    {
      "epoch": 2.4860759493670885,
      "grad_norm": 1.252125859260559,
      "learning_rate": 7.081542248419686e-05,
      "loss": 0.6136764287948608,
      "step": 5892
    },
    {
      "epoch": 2.486919831223629,
      "grad_norm": 1.3686699867248535,
      "learning_rate": 7.079358542817382e-05,
      "loss": 0.6084910035133362,
      "step": 5894
    },
    {
      "epoch": 2.487763713080169,
      "grad_norm": 1.0877282619476318,
      "learning_rate": 7.077174357546996e-05,
      "loss": 0.5862250924110413,
      "step": 5896
    },
    {
      "epoch": 2.488607594936709,
      "grad_norm": 1.164095401763916,
      "learning_rate": 7.074989693112381e-05,
      "loss": 0.6300894021987915,
      "step": 5898
    },
    {
      "epoch": 2.489451476793249,
      "grad_norm": 1.1169507503509521,
      "learning_rate": 7.072804550017493e-05,
      "loss": 0.5508570075035095,
      "step": 5900
    },
    {
      "epoch": 2.489451476793249,
      "eval_loss": 0.6820966005325317,
      "eval_runtime": 513.3515,
      "eval_samples_per_second": 4.104,
      "eval_steps_per_second": 4.104,
      "step": 5900
    },
    {
      "epoch": 2.490295358649789,
      "grad_norm": 1.1718615293502808,
      "learning_rate": 7.070618928766406e-05,
      "loss": 0.550847589969635,
      "step": 5902
    },
    {
      "epoch": 2.491139240506329,
      "grad_norm": 1.4725650548934937,
      "learning_rate": 7.068432829863298e-05,
      "loss": 0.5663347840309143,
      "step": 5904
    },
    {
      "epoch": 2.491983122362869,
      "grad_norm": 1.042083978652954,
      "learning_rate": 7.066246253812462e-05,
      "loss": 0.5506191849708557,
      "step": 5906
    },
    {
      "epoch": 2.4928270042194094,
      "grad_norm": 1.2020974159240723,
      "learning_rate": 7.064059201118297e-05,
      "loss": 0.5656929612159729,
      "step": 5908
    },
    {
      "epoch": 2.4936708860759493,
      "grad_norm": 1.1040663719177246,
      "learning_rate": 7.061871672285317e-05,
      "loss": 0.5159370303153992,
      "step": 5910
    },
    {
      "epoch": 2.4945147679324893,
      "grad_norm": 1.3681589365005493,
      "learning_rate": 7.05968366781814e-05,
      "loss": 0.6161949634552002,
      "step": 5912
    },
    {
      "epoch": 2.4953586497890297,
      "grad_norm": 1.26628839969635,
      "learning_rate": 7.057495188221498e-05,
      "loss": 0.6357758641242981,
      "step": 5914
    },
    {
      "epoch": 2.4962025316455696,
      "grad_norm": 1.2714020013809204,
      "learning_rate": 7.05530623400023e-05,
      "loss": 0.5467366576194763,
      "step": 5916
    },
    {
      "epoch": 2.4970464135021095,
      "grad_norm": 1.2255018949508667,
      "learning_rate": 7.053116805659287e-05,
      "loss": 0.592526376247406,
      "step": 5918
    },
    {
      "epoch": 2.49789029535865,
      "grad_norm": 1.2816206216812134,
      "learning_rate": 7.050926903703729e-05,
      "loss": 0.5819981694221497,
      "step": 5920
    },
    {
      "epoch": 2.49873417721519,
      "grad_norm": 1.1938221454620361,
      "learning_rate": 7.048736528638722e-05,
      "loss": 0.6037712693214417,
      "step": 5922
    },
    {
      "epoch": 2.49957805907173,
      "grad_norm": 1.1330323219299316,
      "learning_rate": 7.046545680969545e-05,
      "loss": 0.5567215085029602,
      "step": 5924
    },
    {
      "epoch": 2.50042194092827,
      "grad_norm": 1.233564019203186,
      "learning_rate": 7.044354361201585e-05,
      "loss": 0.5626974105834961,
      "step": 5926
    },
    {
      "epoch": 2.50126582278481,
      "grad_norm": 1.1913540363311768,
      "learning_rate": 7.042162569840336e-05,
      "loss": 0.5672739744186401,
      "step": 5928
    },
    {
      "epoch": 2.50210970464135,
      "grad_norm": 1.060952067375183,
      "learning_rate": 7.039970307391402e-05,
      "loss": 0.5965602993965149,
      "step": 5930
    },
    {
      "epoch": 2.5029535864978905,
      "grad_norm": 1.2003182172775269,
      "learning_rate": 7.037777574360497e-05,
      "loss": 0.590932309627533,
      "step": 5932
    },
    {
      "epoch": 2.5037974683544304,
      "grad_norm": 1.073434829711914,
      "learning_rate": 7.035584371253441e-05,
      "loss": 0.5736868381500244,
      "step": 5934
    },
    {
      "epoch": 2.5046413502109703,
      "grad_norm": 1.2641130685806274,
      "learning_rate": 7.033390698576166e-05,
      "loss": 0.614703357219696,
      "step": 5936
    },
    {
      "epoch": 2.5054852320675103,
      "grad_norm": 1.2406511306762695,
      "learning_rate": 7.031196556834708e-05,
      "loss": 0.5866397023200989,
      "step": 5938
    },
    {
      "epoch": 2.5063291139240507,
      "grad_norm": 1.231619119644165,
      "learning_rate": 7.029001946535215e-05,
      "loss": 0.5792667865753174,
      "step": 5940
    },
    {
      "epoch": 2.5071729957805906,
      "grad_norm": 1.419447660446167,
      "learning_rate": 7.026806868183939e-05,
      "loss": 0.5686604976654053,
      "step": 5942
    },
    {
      "epoch": 2.5080168776371305,
      "grad_norm": 1.139244556427002,
      "learning_rate": 7.024611322287245e-05,
      "loss": 0.5860661268234253,
      "step": 5944
    },
    {
      "epoch": 2.508860759493671,
      "grad_norm": 1.070517897605896,
      "learning_rate": 7.022415309351602e-05,
      "loss": 0.5823250412940979,
      "step": 5946
    },
    {
      "epoch": 2.509704641350211,
      "grad_norm": 1.0775398015975952,
      "learning_rate": 7.020218829883589e-05,
      "loss": 0.5291389226913452,
      "step": 5948
    },
    {
      "epoch": 2.510548523206751,
      "grad_norm": 1.339716911315918,
      "learning_rate": 7.018021884389892e-05,
      "loss": 0.6215447783470154,
      "step": 5950
    },
    {
      "epoch": 2.511392405063291,
      "grad_norm": 1.3589707612991333,
      "learning_rate": 7.0158244733773e-05,
      "loss": 0.5419909358024597,
      "step": 5952
    },
    {
      "epoch": 2.512236286919831,
      "grad_norm": 1.1664098501205444,
      "learning_rate": 7.01362659735272e-05,
      "loss": 0.5476977229118347,
      "step": 5954
    },
    {
      "epoch": 2.513080168776371,
      "grad_norm": 1.1184223890304565,
      "learning_rate": 7.011428256823154e-05,
      "loss": 0.5896323919296265,
      "step": 5956
    },
    {
      "epoch": 2.5139240506329115,
      "grad_norm": 1.4071170091629028,
      "learning_rate": 7.00922945229572e-05,
      "loss": 0.6353691220283508,
      "step": 5958
    },
    {
      "epoch": 2.5147679324894514,
      "grad_norm": 1.3740885257720947,
      "learning_rate": 7.007030184277641e-05,
      "loss": 0.6605582237243652,
      "step": 5960
    },
    {
      "epoch": 2.5156118143459913,
      "grad_norm": 1.071395754814148,
      "learning_rate": 7.004830453276241e-05,
      "loss": 0.6399887800216675,
      "step": 5962
    },
    {
      "epoch": 2.5164556962025317,
      "grad_norm": 1.2292311191558838,
      "learning_rate": 7.002630259798962e-05,
      "loss": 0.5992775559425354,
      "step": 5964
    },
    {
      "epoch": 2.5172995780590717,
      "grad_norm": 1.0133391618728638,
      "learning_rate": 7.000429604353341e-05,
      "loss": 0.5716721415519714,
      "step": 5966
    },
    {
      "epoch": 2.5181434599156116,
      "grad_norm": 1.2669343948364258,
      "learning_rate": 6.998228487447032e-05,
      "loss": 0.5455520749092102,
      "step": 5968
    },
    {
      "epoch": 2.518987341772152,
      "grad_norm": 1.2026386260986328,
      "learning_rate": 6.996026909587785e-05,
      "loss": 0.6411572694778442,
      "step": 5970
    },
    {
      "epoch": 2.519831223628692,
      "grad_norm": 1.359923243522644,
      "learning_rate": 6.993824871283465e-05,
      "loss": 0.6687750220298767,
      "step": 5972
    },
    {
      "epoch": 2.520675105485232,
      "grad_norm": 1.1265650987625122,
      "learning_rate": 6.99162237304204e-05,
      "loss": 0.6271382570266724,
      "step": 5974
    },
    {
      "epoch": 2.5215189873417723,
      "grad_norm": 1.197667121887207,
      "learning_rate": 6.989419415371583e-05,
      "loss": 0.6191279888153076,
      "step": 5976
    },
    {
      "epoch": 2.522362869198312,
      "grad_norm": 1.169992446899414,
      "learning_rate": 6.987215998780275e-05,
      "loss": 0.6313687562942505,
      "step": 5978
    },
    {
      "epoch": 2.523206751054852,
      "grad_norm": 1.2706433534622192,
      "learning_rate": 6.9850121237764e-05,
      "loss": 0.6058336496353149,
      "step": 5980
    },
    {
      "epoch": 2.5240506329113925,
      "grad_norm": 1.322376012802124,
      "learning_rate": 6.982807790868352e-05,
      "loss": 0.6466464400291443,
      "step": 5982
    },
    {
      "epoch": 2.5248945147679325,
      "grad_norm": 1.2398571968078613,
      "learning_rate": 6.980603000564626e-05,
      "loss": 0.5730098485946655,
      "step": 5984
    },
    {
      "epoch": 2.5257383966244724,
      "grad_norm": 1.2035216093063354,
      "learning_rate": 6.978397753373826e-05,
      "loss": 0.5305635333061218,
      "step": 5986
    },
    {
      "epoch": 2.526582278481013,
      "grad_norm": 1.1951299905776978,
      "learning_rate": 6.976192049804661e-05,
      "loss": 0.5601096153259277,
      "step": 5988
    },
    {
      "epoch": 2.5274261603375527,
      "grad_norm": 0.9950459599494934,
      "learning_rate": 6.973985890365945e-05,
      "loss": 0.5049516558647156,
      "step": 5990
    },
    {
      "epoch": 2.5282700421940927,
      "grad_norm": 1.2581008672714233,
      "learning_rate": 6.971779275566593e-05,
      "loss": 0.5456960797309875,
      "step": 5992
    },
    {
      "epoch": 2.529113924050633,
      "grad_norm": 1.2196903228759766,
      "learning_rate": 6.969572205915632e-05,
      "loss": 0.6026827096939087,
      "step": 5994
    },
    {
      "epoch": 2.529957805907173,
      "grad_norm": 1.3109357357025146,
      "learning_rate": 6.967364681922189e-05,
      "loss": 0.597453236579895,
      "step": 5996
    },
    {
      "epoch": 2.530801687763713,
      "grad_norm": 1.016904354095459,
      "learning_rate": 6.965156704095498e-05,
      "loss": 0.5304323434829712,
      "step": 5998
    },
    {
      "epoch": 2.5316455696202533,
      "grad_norm": 1.2363858222961426,
      "learning_rate": 6.962948272944896e-05,
      "loss": 0.5748253464698792,
      "step": 6000
    },
    {
      "epoch": 2.5316455696202533,
      "eval_loss": 0.6813357472419739,
      "eval_runtime": 513.5491,
      "eval_samples_per_second": 4.103,
      "eval_steps_per_second": 4.103,
      "step": 6000
    }
  ],
  "logging_steps": 2,
  "max_steps": 14220,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 6,
  "save_steps": 500,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 5,
        "early_stopping_threshold": 0.001
      },
      "attributes": {
        "early_stopping_patience_counter": 1
      }
    },
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 6.231982336554848e+18,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}