{
  "best_metric": 0.0013051489368081093,
  "best_model_checkpoint": "./ViT-NIH-Chest-X-ray-dataset-small/checkpoint-2400",
  "epoch": 9.0,
  "eval_steps": 100,
  "global_step": 2439,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.03690036900369004,
      "grad_norm": 1.528047800064087,
      "learning_rate": 0.000199179991799918,
      "loss": 0.0667,
      "step": 10
    },
    {
      "epoch": 0.07380073800738007,
      "grad_norm": 0.4356657862663269,
      "learning_rate": 0.000198359983599836,
      "loss": 0.0497,
      "step": 20
    },
    {
      "epoch": 0.11070110701107011,
      "grad_norm": 0.7035173177719116,
      "learning_rate": 0.000197539975399754,
      "loss": 0.0452,
      "step": 30
    },
    {
      "epoch": 0.14760147601476015,
      "grad_norm": 0.6605591773986816,
      "learning_rate": 0.000196719967199672,
      "loss": 0.0513,
      "step": 40
    },
    {
      "epoch": 0.18450184501845018,
      "grad_norm": 0.3567582964897156,
      "learning_rate": 0.00019589995899959,
      "loss": 0.0582,
      "step": 50
    },
    {
      "epoch": 0.22140221402214022,
      "grad_norm": 0.8019290566444397,
      "learning_rate": 0.000195079950799508,
      "loss": 0.066,
      "step": 60
    },
    {
      "epoch": 0.25830258302583026,
      "grad_norm": 0.3395977020263672,
      "learning_rate": 0.000194259942599426,
      "loss": 0.0461,
      "step": 70
    },
    {
      "epoch": 0.2952029520295203,
      "grad_norm": 1.1884530782699585,
      "learning_rate": 0.000193439934399344,
      "loss": 0.055,
      "step": 80
    },
    {
      "epoch": 0.33210332103321033,
      "grad_norm": 0.3934822082519531,
      "learning_rate": 0.000192619926199262,
      "loss": 0.0341,
      "step": 90
    },
    {
      "epoch": 0.36900369003690037,
      "grad_norm": 0.19823391735553741,
      "learning_rate": 0.00019179991799918,
      "loss": 0.0271,
      "step": 100
    },
    {
      "epoch": 0.36900369003690037,
      "eval_accuracy": 0.8583815028901735,
      "eval_loss": 0.03472097963094711,
      "eval_runtime": 53.2441,
      "eval_samples_per_second": 32.492,
      "eval_steps_per_second": 4.076,
      "step": 100
    },
    {
      "epoch": 0.4059040590405904,
      "grad_norm": 0.6334025859832764,
      "learning_rate": 0.00019097990979909799,
      "loss": 0.0347,
      "step": 110
    },
    {
      "epoch": 0.44280442804428044,
      "grad_norm": 0.4480656683444977,
      "learning_rate": 0.000190159901599016,
      "loss": 0.038,
      "step": 120
    },
    {
      "epoch": 0.4797047970479705,
      "grad_norm": 0.35815778374671936,
      "learning_rate": 0.000189339893398934,
      "loss": 0.0308,
      "step": 130
    },
    {
      "epoch": 0.5166051660516605,
      "grad_norm": 0.5924233794212341,
      "learning_rate": 0.000188519885198852,
      "loss": 0.0329,
      "step": 140
    },
    {
      "epoch": 0.5535055350553506,
      "grad_norm": 0.8932539224624634,
      "learning_rate": 0.00018769987699876998,
      "loss": 0.0447,
      "step": 150
    },
    {
      "epoch": 0.5904059040590406,
      "grad_norm": 0.4765835702419281,
      "learning_rate": 0.000186879868798688,
      "loss": 0.0335,
      "step": 160
    },
    {
      "epoch": 0.6273062730627307,
      "grad_norm": 1.848928451538086,
      "learning_rate": 0.00018605986059860597,
      "loss": 0.0353,
      "step": 170
    },
    {
      "epoch": 0.6642066420664207,
      "grad_norm": 0.5017064809799194,
      "learning_rate": 0.000185239852398524,
      "loss": 0.0365,
      "step": 180
    },
    {
      "epoch": 0.7011070110701108,
      "grad_norm": 0.4620348811149597,
      "learning_rate": 0.00018441984419844198,
      "loss": 0.0284,
      "step": 190
    },
    {
      "epoch": 0.7380073800738007,
      "grad_norm": 0.716968834400177,
      "learning_rate": 0.00018359983599836,
      "loss": 0.0334,
      "step": 200
    },
    {
      "epoch": 0.7380073800738007,
      "eval_accuracy": 0.8624277456647399,
      "eval_loss": 0.02910423092544079,
      "eval_runtime": 53.0242,
      "eval_samples_per_second": 32.627,
      "eval_steps_per_second": 4.092,
      "step": 200
    },
    {
      "epoch": 0.7749077490774908,
      "grad_norm": 0.6060870289802551,
      "learning_rate": 0.00018277982779827797,
      "loss": 0.0383,
      "step": 210
    },
    {
      "epoch": 0.8118081180811808,
      "grad_norm": 1.053126335144043,
      "learning_rate": 0.000181959819598196,
      "loss": 0.0521,
      "step": 220
    },
    {
      "epoch": 0.8487084870848709,
      "grad_norm": 0.48799070715904236,
      "learning_rate": 0.00018113981139811398,
      "loss": 0.0529,
      "step": 230
    },
    {
      "epoch": 0.8856088560885609,
      "grad_norm": 0.8857760429382324,
      "learning_rate": 0.00018031980319803198,
      "loss": 0.0359,
      "step": 240
    },
    {
      "epoch": 0.922509225092251,
      "grad_norm": 0.2952769696712494,
      "learning_rate": 0.00017949979499794997,
      "loss": 0.0457,
      "step": 250
    },
    {
      "epoch": 0.959409594095941,
      "grad_norm": 0.39872637391090393,
      "learning_rate": 0.000178679786797868,
      "loss": 0.0573,
      "step": 260
    },
    {
      "epoch": 0.996309963099631,
      "grad_norm": 0.5376998782157898,
      "learning_rate": 0.00017785977859778598,
      "loss": 0.0352,
      "step": 270
    },
    {
      "epoch": 1.033210332103321,
      "grad_norm": 0.32833996415138245,
      "learning_rate": 0.00017703977039770397,
      "loss": 0.0444,
      "step": 280
    },
    {
      "epoch": 1.070110701107011,
      "grad_norm": 0.6235490441322327,
      "learning_rate": 0.00017621976219762197,
      "loss": 0.0347,
      "step": 290
    },
    {
      "epoch": 1.1070110701107012,
      "grad_norm": 0.612684428691864,
      "learning_rate": 0.00017539975399754,
      "loss": 0.0438,
      "step": 300
    },
    {
      "epoch": 1.1070110701107012,
      "eval_accuracy": 0.8606936416184972,
      "eval_loss": 0.03517782315611839,
      "eval_runtime": 53.0738,
      "eval_samples_per_second": 32.596,
      "eval_steps_per_second": 4.089,
      "step": 300
    },
    {
      "epoch": 1.1439114391143912,
      "grad_norm": 0.29079392552375793,
      "learning_rate": 0.00017457974579745798,
      "loss": 0.0546,
      "step": 310
    },
    {
      "epoch": 1.1808118081180812,
      "grad_norm": 0.3669208884239197,
      "learning_rate": 0.00017375973759737597,
      "loss": 0.03,
      "step": 320
    },
    {
      "epoch": 1.2177121771217712,
      "grad_norm": 0.27689820528030396,
      "learning_rate": 0.00017293972939729397,
      "loss": 0.0287,
      "step": 330
    },
    {
      "epoch": 1.2546125461254611,
      "grad_norm": 0.2193935513496399,
      "learning_rate": 0.00017211972119721199,
      "loss": 0.0336,
      "step": 340
    },
    {
      "epoch": 1.2915129151291513,
      "grad_norm": 0.1442354917526245,
      "learning_rate": 0.00017129971299712998,
      "loss": 0.0301,
      "step": 350
    },
    {
      "epoch": 1.3284132841328413,
      "grad_norm": 0.49084681272506714,
      "learning_rate": 0.00017047970479704797,
      "loss": 0.0383,
      "step": 360
    },
    {
      "epoch": 1.3653136531365313,
      "grad_norm": 0.6412920951843262,
      "learning_rate": 0.00016965969659696597,
      "loss": 0.0338,
      "step": 370
    },
    {
      "epoch": 1.4022140221402215,
      "grad_norm": 0.5825220942497253,
      "learning_rate": 0.00016883968839688399,
      "loss": 0.0344,
      "step": 380
    },
    {
      "epoch": 1.4391143911439115,
      "grad_norm": 0.5842616558074951,
      "learning_rate": 0.00016801968019680198,
      "loss": 0.0358,
      "step": 390
    },
    {
      "epoch": 1.4760147601476015,
      "grad_norm": 0.17261776328086853,
      "learning_rate": 0.00016719967199671997,
      "loss": 0.0215,
      "step": 400
    },
    {
      "epoch": 1.4760147601476015,
      "eval_accuracy": 0.8745664739884393,
      "eval_loss": 0.03190147876739502,
      "eval_runtime": 53.163,
      "eval_samples_per_second": 32.541,
      "eval_steps_per_second": 4.082,
      "step": 400
    },
    {
      "epoch": 1.5129151291512914,
      "grad_norm": 0.42338722944259644,
      "learning_rate": 0.00016637966379663796,
      "loss": 0.0343,
      "step": 410
    },
    {
      "epoch": 1.5498154981549814,
      "grad_norm": 0.6431716680526733,
      "learning_rate": 0.00016555965559655598,
      "loss": 0.0308,
      "step": 420
    },
    {
      "epoch": 1.5867158671586716,
      "grad_norm": 0.9911162257194519,
      "learning_rate": 0.00016473964739647395,
      "loss": 0.0357,
      "step": 430
    },
    {
      "epoch": 1.6236162361623616,
      "grad_norm": 0.7283934354782104,
      "learning_rate": 0.00016391963919639197,
      "loss": 0.0548,
      "step": 440
    },
    {
      "epoch": 1.6605166051660518,
      "grad_norm": 0.3784200847148895,
      "learning_rate": 0.00016309963099630996,
      "loss": 0.0411,
      "step": 450
    },
    {
      "epoch": 1.6974169741697418,
      "grad_norm": 0.4938986599445343,
      "learning_rate": 0.00016227962279622798,
      "loss": 0.0384,
      "step": 460
    },
    {
      "epoch": 1.7343173431734318,
      "grad_norm": 0.7969509363174438,
      "learning_rate": 0.00016145961459614595,
      "loss": 0.0282,
      "step": 470
    },
    {
      "epoch": 1.7712177121771218,
      "grad_norm": 0.7997636198997498,
      "learning_rate": 0.00016063960639606397,
      "loss": 0.0299,
      "step": 480
    },
    {
      "epoch": 1.8081180811808117,
      "grad_norm": 0.13115979731082916,
      "learning_rate": 0.00015981959819598196,
      "loss": 0.0361,
      "step": 490
    },
    {
      "epoch": 1.8450184501845017,
      "grad_norm": 0.9753024578094482,
      "learning_rate": 0.00015899958999589995,
      "loss": 0.0267,
      "step": 500
    },
    {
      "epoch": 1.8450184501845017,
      "eval_accuracy": 0.8797687861271676,
      "eval_loss": 0.02772846817970276,
      "eval_runtime": 52.9913,
      "eval_samples_per_second": 32.647,
      "eval_steps_per_second": 4.095,
      "step": 500
    },
    {
      "epoch": 1.881918819188192,
      "grad_norm": 0.5084354877471924,
      "learning_rate": 0.00015817958179581795,
      "loss": 0.0432,
      "step": 510
    },
    {
      "epoch": 1.918819188191882,
      "grad_norm": 0.2513779103755951,
      "learning_rate": 0.00015735957359573597,
      "loss": 0.034,
      "step": 520
    },
    {
      "epoch": 1.9557195571955721,
      "grad_norm": 0.284483402967453,
      "learning_rate": 0.00015653956539565396,
      "loss": 0.0492,
      "step": 530
    },
    {
      "epoch": 1.992619926199262,
      "grad_norm": 0.3686124086380005,
      "learning_rate": 0.00015571955719557195,
      "loss": 0.0313,
      "step": 540
    },
    {
      "epoch": 2.029520295202952,
      "grad_norm": 0.49193885922431946,
      "learning_rate": 0.00015489954899548995,
      "loss": 0.0346,
      "step": 550
    },
    {
      "epoch": 2.066420664206642,
      "grad_norm": 0.48066815733909607,
      "learning_rate": 0.00015407954079540797,
      "loss": 0.0204,
      "step": 560
    },
    {
      "epoch": 2.103321033210332,
      "grad_norm": 0.4960450530052185,
      "learning_rate": 0.00015325953259532596,
      "loss": 0.0208,
      "step": 570
    },
    {
      "epoch": 2.140221402214022,
      "grad_norm": 0.3585696220397949,
      "learning_rate": 0.00015243952439524395,
      "loss": 0.0226,
      "step": 580
    },
    {
      "epoch": 2.177121771217712,
      "grad_norm": 0.383513480424881,
      "learning_rate": 0.00015161951619516195,
      "loss": 0.024,
      "step": 590
    },
    {
      "epoch": 2.2140221402214024,
      "grad_norm": 0.45566025376319885,
      "learning_rate": 0.00015079950799507997,
      "loss": 0.0266,
      "step": 600
    },
    {
      "epoch": 2.2140221402214024,
      "eval_accuracy": 0.9115606936416185,
      "eval_loss": 0.017673172056674957,
      "eval_runtime": 53.2922,
      "eval_samples_per_second": 32.463,
      "eval_steps_per_second": 4.072,
      "step": 600
    },
    {
      "epoch": 2.2509225092250924,
      "grad_norm": 0.09787905216217041,
      "learning_rate": 0.00014997949979499796,
      "loss": 0.0173,
      "step": 610
    },
    {
      "epoch": 2.2878228782287824,
      "grad_norm": 0.4203137159347534,
      "learning_rate": 0.00014915949159491595,
      "loss": 0.0208,
      "step": 620
    },
    {
      "epoch": 2.3247232472324724,
      "grad_norm": 0.27736109495162964,
      "learning_rate": 0.00014833948339483394,
      "loss": 0.019,
      "step": 630
    },
    {
      "epoch": 2.3616236162361623,
      "grad_norm": 0.1169864758849144,
      "learning_rate": 0.00014751947519475196,
      "loss": 0.0144,
      "step": 640
    },
    {
      "epoch": 2.3985239852398523,
      "grad_norm": 0.4632050096988678,
      "learning_rate": 0.00014669946699466993,
      "loss": 0.022,
      "step": 650
    },
    {
      "epoch": 2.4354243542435423,
      "grad_norm": 0.25868573784828186,
      "learning_rate": 0.00014587945879458795,
      "loss": 0.0225,
      "step": 660
    },
    {
      "epoch": 2.4723247232472323,
      "grad_norm": 0.7463150024414062,
      "learning_rate": 0.00014505945059450594,
      "loss": 0.0218,
      "step": 670
    },
    {
      "epoch": 2.5092250922509223,
      "grad_norm": 0.4195776581764221,
      "learning_rate": 0.00014423944239442396,
      "loss": 0.022,
      "step": 680
    },
    {
      "epoch": 2.5461254612546127,
      "grad_norm": 1.2269337177276611,
      "learning_rate": 0.00014341943419434193,
      "loss": 0.0216,
      "step": 690
    },
    {
      "epoch": 2.5830258302583027,
      "grad_norm": 0.44495096802711487,
      "learning_rate": 0.00014259942599425995,
      "loss": 0.014,
      "step": 700
    },
    {
      "epoch": 2.5830258302583027,
      "eval_accuracy": 0.9497109826589596,
      "eval_loss": 0.01273352187126875,
      "eval_runtime": 53.0795,
      "eval_samples_per_second": 32.593,
      "eval_steps_per_second": 4.088,
      "step": 700
    },
    {
      "epoch": 2.6199261992619927,
      "grad_norm": 0.16129033267498016,
      "learning_rate": 0.00014177941779417794,
      "loss": 0.0211,
      "step": 710
    },
    {
      "epoch": 2.6568265682656826,
      "grad_norm": 0.639427900314331,
      "learning_rate": 0.00014095940959409593,
      "loss": 0.0245,
      "step": 720
    },
    {
      "epoch": 2.6937269372693726,
      "grad_norm": 1.1732105016708374,
      "learning_rate": 0.00014013940139401393,
      "loss": 0.0188,
      "step": 730
    },
    {
      "epoch": 2.7306273062730626,
      "grad_norm": 0.15935984253883362,
      "learning_rate": 0.00013931939319393195,
      "loss": 0.0225,
      "step": 740
    },
    {
      "epoch": 2.767527675276753,
      "grad_norm": 0.44544655084609985,
      "learning_rate": 0.00013849938499384994,
      "loss": 0.0181,
      "step": 750
    },
    {
      "epoch": 2.804428044280443,
      "grad_norm": 0.04136790707707405,
      "learning_rate": 0.00013767937679376793,
      "loss": 0.0193,
      "step": 760
    },
    {
      "epoch": 2.841328413284133,
      "grad_norm": 0.3928181231021881,
      "learning_rate": 0.00013685936859368593,
      "loss": 0.0314,
      "step": 770
    },
    {
      "epoch": 2.878228782287823,
      "grad_norm": 0.4807884097099304,
      "learning_rate": 0.00013603936039360395,
      "loss": 0.0204,
      "step": 780
    },
    {
      "epoch": 2.915129151291513,
      "grad_norm": 0.11494464427232742,
      "learning_rate": 0.00013521935219352194,
      "loss": 0.0163,
      "step": 790
    },
    {
      "epoch": 2.952029520295203,
      "grad_norm": 0.4516555964946747,
      "learning_rate": 0.00013439934399343993,
      "loss": 0.0207,
      "step": 800
    },
    {
      "epoch": 2.952029520295203,
      "eval_accuracy": 0.9410404624277456,
      "eval_loss": 0.014368296600878239,
      "eval_runtime": 53.187,
      "eval_samples_per_second": 32.527,
      "eval_steps_per_second": 4.08,
      "step": 800
    },
    {
      "epoch": 2.988929889298893,
      "grad_norm": 0.25021785497665405,
      "learning_rate": 0.00013357933579335793,
      "loss": 0.018,
      "step": 810
    },
    {
      "epoch": 3.025830258302583,
      "grad_norm": 0.6319343447685242,
      "learning_rate": 0.00013275932759327595,
      "loss": 0.0212,
      "step": 820
    },
    {
      "epoch": 3.062730627306273,
      "grad_norm": 0.16027286648750305,
      "learning_rate": 0.00013193931939319394,
      "loss": 0.0119,
      "step": 830
    },
    {
      "epoch": 3.0996309963099633,
      "grad_norm": 0.27029338479042053,
      "learning_rate": 0.00013111931119311193,
      "loss": 0.011,
      "step": 840
    },
    {
      "epoch": 3.1365313653136533,
      "grad_norm": 0.3908526301383972,
      "learning_rate": 0.00013029930299302992,
      "loss": 0.0113,
      "step": 850
    },
    {
      "epoch": 3.1734317343173433,
      "grad_norm": 0.08731468766927719,
      "learning_rate": 0.00012947929479294794,
      "loss": 0.0088,
      "step": 860
    },
    {
      "epoch": 3.2103321033210332,
      "grad_norm": 0.19361713528633118,
      "learning_rate": 0.00012865928659286594,
      "loss": 0.0138,
      "step": 870
    },
    {
      "epoch": 3.2472324723247232,
      "grad_norm": 0.11760518699884415,
      "learning_rate": 0.00012783927839278393,
      "loss": 0.0128,
      "step": 880
    },
    {
      "epoch": 3.284132841328413,
      "grad_norm": 0.5687548518180847,
      "learning_rate": 0.00012701927019270192,
      "loss": 0.0114,
      "step": 890
    },
    {
      "epoch": 3.321033210332103,
      "grad_norm": 0.35156571865081787,
      "learning_rate": 0.00012619926199261994,
      "loss": 0.0115,
      "step": 900
    },
    {
      "epoch": 3.321033210332103,
      "eval_accuracy": 0.9653179190751445,
      "eval_loss": 0.009723745286464691,
      "eval_runtime": 53.0303,
      "eval_samples_per_second": 32.623,
      "eval_steps_per_second": 4.092,
      "step": 900
    },
    {
      "epoch": 3.357933579335793,
      "grad_norm": 0.22674980759620667,
      "learning_rate": 0.0001253792537925379,
      "loss": 0.0126,
      "step": 910
    },
    {
      "epoch": 3.3948339483394836,
      "grad_norm": 0.23142710328102112,
      "learning_rate": 0.00012455924559245593,
      "loss": 0.007,
      "step": 920
    },
    {
      "epoch": 3.4317343173431736,
      "grad_norm": 0.11926586925983429,
      "learning_rate": 0.00012373923739237392,
      "loss": 0.0111,
      "step": 930
    },
    {
      "epoch": 3.4686346863468636,
      "grad_norm": 0.12741585075855255,
      "learning_rate": 0.00012291922919229194,
      "loss": 0.0149,
      "step": 940
    },
    {
      "epoch": 3.5055350553505535,
      "grad_norm": 0.8535834550857544,
      "learning_rate": 0.0001220992209922099,
      "loss": 0.0098,
      "step": 950
    },
    {
      "epoch": 3.5424354243542435,
      "grad_norm": 0.2683662474155426,
      "learning_rate": 0.00012127921279212793,
      "loss": 0.0107,
      "step": 960
    },
    {
      "epoch": 3.5793357933579335,
      "grad_norm": 0.2808018624782562,
      "learning_rate": 0.00012045920459204592,
      "loss": 0.0129,
      "step": 970
    },
    {
      "epoch": 3.6162361623616235,
      "grad_norm": 1.0000771284103394,
      "learning_rate": 0.00011963919639196393,
      "loss": 0.0148,
      "step": 980
    },
    {
      "epoch": 3.6531365313653135,
      "grad_norm": 0.16155269742012024,
      "learning_rate": 0.00011881918819188192,
      "loss": 0.0118,
      "step": 990
    },
    {
      "epoch": 3.6900369003690034,
      "grad_norm": 0.053899142891168594,
      "learning_rate": 0.00011799917999179993,
      "loss": 0.0113,
      "step": 1000
    },
    {
      "epoch": 3.6900369003690034,
      "eval_accuracy": 0.9710982658959537,
      "eval_loss": 0.007686651311814785,
      "eval_runtime": 52.7853,
      "eval_samples_per_second": 32.774,
      "eval_steps_per_second": 4.111,
      "step": 1000
    },
    {
      "epoch": 3.726937269372694,
      "grad_norm": 0.23413535952568054,
      "learning_rate": 0.0001171791717917179,
      "loss": 0.0118,
      "step": 1010
    },
    {
      "epoch": 3.763837638376384,
      "grad_norm": 0.19596587121486664,
      "learning_rate": 0.00011635916359163593,
      "loss": 0.0069,
      "step": 1020
    },
    {
      "epoch": 3.800738007380074,
      "grad_norm": 0.28500398993492126,
      "learning_rate": 0.0001155391553915539,
      "loss": 0.009,
      "step": 1030
    },
    {
      "epoch": 3.837638376383764,
      "grad_norm": 0.3002331554889679,
      "learning_rate": 0.00011471914719147193,
      "loss": 0.0096,
      "step": 1040
    },
    {
      "epoch": 3.874538745387454,
      "grad_norm": 0.21130523085594177,
      "learning_rate": 0.0001138991389913899,
      "loss": 0.0122,
      "step": 1050
    },
    {
      "epoch": 3.911439114391144,
      "grad_norm": 0.14506787061691284,
      "learning_rate": 0.00011307913079130793,
      "loss": 0.0095,
      "step": 1060
    },
    {
      "epoch": 3.948339483394834,
      "grad_norm": 0.3295513987541199,
      "learning_rate": 0.0001122591225912259,
      "loss": 0.0122,
      "step": 1070
    },
    {
      "epoch": 3.985239852398524,
      "grad_norm": 0.07300284504890442,
      "learning_rate": 0.00011143911439114391,
      "loss": 0.0067,
      "step": 1080
    },
    {
      "epoch": 4.022140221402214,
      "grad_norm": 0.03413034602999687,
      "learning_rate": 0.0001106191061910619,
      "loss": 0.0119,
      "step": 1090
    },
    {
      "epoch": 4.059040590405904,
      "grad_norm": 0.06467791646718979,
      "learning_rate": 0.00010979909799097991,
      "loss": 0.0054,
      "step": 1100
    },
    {
      "epoch": 4.059040590405904,
      "eval_accuracy": 0.9843930635838151,
      "eval_loss": 0.00683161336928606,
      "eval_runtime": 52.9373,
      "eval_samples_per_second": 32.68,
      "eval_steps_per_second": 4.099,
      "step": 1100
    },
    {
      "epoch": 4.095940959409594,
      "grad_norm": 0.09930551052093506,
      "learning_rate": 0.0001089790897908979,
      "loss": 0.0057,
      "step": 1110
    },
    {
      "epoch": 4.132841328413284,
      "grad_norm": 0.11587420105934143,
      "learning_rate": 0.00010815908159081591,
      "loss": 0.0065,
      "step": 1120
    },
    {
      "epoch": 4.169741697416974,
      "grad_norm": 0.042286355048418045,
      "learning_rate": 0.0001073390733907339,
      "loss": 0.0067,
      "step": 1130
    },
    {
      "epoch": 4.206642066420664,
      "grad_norm": 0.16917578876018524,
      "learning_rate": 0.00010651906519065191,
      "loss": 0.0039,
      "step": 1140
    },
    {
      "epoch": 4.243542435424354,
      "grad_norm": 0.16002944111824036,
      "learning_rate": 0.0001056990569905699,
      "loss": 0.0066,
      "step": 1150
    },
    {
      "epoch": 4.280442804428044,
      "grad_norm": 0.049849651753902435,
      "learning_rate": 0.00010487904879048791,
      "loss": 0.0049,
      "step": 1160
    },
    {
      "epoch": 4.317343173431734,
      "grad_norm": 0.5795335173606873,
      "learning_rate": 0.0001040590405904059,
      "loss": 0.007,
      "step": 1170
    },
    {
      "epoch": 4.354243542435424,
      "grad_norm": 0.01594790630042553,
      "learning_rate": 0.00010323903239032391,
      "loss": 0.0062,
      "step": 1180
    },
    {
      "epoch": 4.391143911439114,
      "grad_norm": 0.12844012677669525,
      "learning_rate": 0.0001024190241902419,
      "loss": 0.0045,
      "step": 1190
    },
    {
      "epoch": 4.428044280442805,
      "grad_norm": 0.053635355085134506,
      "learning_rate": 0.00010159901599015991,
      "loss": 0.0047,
      "step": 1200
    },
    {
      "epoch": 4.428044280442805,
      "eval_accuracy": 0.9849710982658959,
      "eval_loss": 0.004636078607290983,
      "eval_runtime": 52.7596,
      "eval_samples_per_second": 32.79,
      "eval_steps_per_second": 4.113,
      "step": 1200
    },
    {
      "epoch": 4.464944649446495,
      "grad_norm": 0.048139333724975586,
      "learning_rate": 0.0001007790077900779,
      "loss": 0.0051,
      "step": 1210
    },
    {
      "epoch": 4.501845018450185,
      "grad_norm": 0.05954507365822792,
      "learning_rate": 9.995899958999591e-05,
      "loss": 0.0076,
      "step": 1220
    },
    {
      "epoch": 4.538745387453875,
      "grad_norm": 0.07452525943517685,
      "learning_rate": 9.91389913899139e-05,
      "loss": 0.0041,
      "step": 1230
    },
    {
      "epoch": 4.575645756457565,
      "grad_norm": 0.008173154667019844,
      "learning_rate": 9.83189831898319e-05,
      "loss": 0.0047,
      "step": 1240
    },
    {
      "epoch": 4.612546125461255,
      "grad_norm": 0.062416139990091324,
      "learning_rate": 9.74989749897499e-05,
      "loss": 0.0058,
      "step": 1250
    },
    {
      "epoch": 4.649446494464945,
      "grad_norm": 0.3318597972393036,
      "learning_rate": 9.66789667896679e-05,
      "loss": 0.0052,
      "step": 1260
    },
    {
      "epoch": 4.686346863468635,
      "grad_norm": 0.1266857534646988,
      "learning_rate": 9.58589585895859e-05,
      "loss": 0.0079,
      "step": 1270
    },
    {
      "epoch": 4.723247232472325,
      "grad_norm": 0.061126936227083206,
      "learning_rate": 9.50389503895039e-05,
      "loss": 0.0051,
      "step": 1280
    },
    {
      "epoch": 4.760147601476015,
      "grad_norm": 0.09749143570661545,
      "learning_rate": 9.42189421894219e-05,
      "loss": 0.0078,
      "step": 1290
    },
    {
      "epoch": 4.797047970479705,
      "grad_norm": 0.34020960330963135,
      "learning_rate": 9.33989339893399e-05,
      "loss": 0.0056,
      "step": 1300
    },
    {
      "epoch": 4.797047970479705,
      "eval_accuracy": 0.9901734104046243,
      "eval_loss": 0.0040424894541502,
      "eval_runtime": 52.7044,
      "eval_samples_per_second": 32.825,
      "eval_steps_per_second": 4.117,
      "step": 1300
    },
    {
      "epoch": 4.833948339483395,
      "grad_norm": 0.19669051468372345,
      "learning_rate": 9.25789257892579e-05,
      "loss": 0.0075,
      "step": 1310
    },
    {
      "epoch": 4.870848708487085,
      "grad_norm": 0.09900686144828796,
      "learning_rate": 9.175891758917589e-05,
      "loss": 0.0077,
      "step": 1320
    },
    {
      "epoch": 4.907749077490775,
      "grad_norm": 0.1532040238380432,
      "learning_rate": 9.09389093890939e-05,
      "loss": 0.0067,
      "step": 1330
    },
    {
      "epoch": 4.944649446494465,
      "grad_norm": 0.051097236573696136,
      "learning_rate": 9.011890118901189e-05,
      "loss": 0.0039,
      "step": 1340
    },
    {
      "epoch": 4.9815498154981555,
      "grad_norm": 0.05327645316720009,
      "learning_rate": 8.92988929889299e-05,
      "loss": 0.0055,
      "step": 1350
    },
    {
      "epoch": 5.018450184501845,
      "grad_norm": 0.026003248989582062,
      "learning_rate": 8.847888478884789e-05,
      "loss": 0.0055,
      "step": 1360
    },
    {
      "epoch": 5.055350553505535,
      "grad_norm": 0.11153847724199295,
      "learning_rate": 8.76588765887659e-05,
      "loss": 0.0044,
      "step": 1370
    },
    {
      "epoch": 5.092250922509225,
      "grad_norm": 0.02700236812233925,
      "learning_rate": 8.683886838868389e-05,
      "loss": 0.0028,
      "step": 1380
    },
    {
      "epoch": 5.129151291512915,
      "grad_norm": 0.07289482653141022,
      "learning_rate": 8.60188601886019e-05,
      "loss": 0.0029,
      "step": 1390
    },
    {
      "epoch": 5.166051660516605,
      "grad_norm": 0.0497610978782177,
      "learning_rate": 8.519885198851989e-05,
      "loss": 0.0026,
      "step": 1400
    },
    {
      "epoch": 5.166051660516605,
      "eval_accuracy": 0.992485549132948,
      "eval_loss": 0.003214773256331682,
      "eval_runtime": 52.8552,
      "eval_samples_per_second": 32.731,
      "eval_steps_per_second": 4.106,
      "step": 1400
    },
    {
      "epoch": 5.202952029520295,
      "grad_norm": 0.15320487320423126,
      "learning_rate": 8.43788437884379e-05,
      "loss": 0.0041,
      "step": 1410
    },
    {
      "epoch": 5.239852398523985,
      "grad_norm": 0.018703853711485863,
      "learning_rate": 8.355883558835589e-05,
      "loss": 0.0041,
      "step": 1420
    },
    {
      "epoch": 5.276752767527675,
      "grad_norm": 0.007155160885304213,
      "learning_rate": 8.273882738827388e-05,
      "loss": 0.0028,
      "step": 1430
    },
    {
      "epoch": 5.313653136531365,
      "grad_norm": 0.026553437113761902,
      "learning_rate": 8.191881918819189e-05,
      "loss": 0.0046,
      "step": 1440
    },
    {
      "epoch": 5.350553505535055,
      "grad_norm": 0.031090274453163147,
      "learning_rate": 8.109881098810988e-05,
      "loss": 0.0044,
      "step": 1450
    },
    {
      "epoch": 5.387453874538745,
      "grad_norm": 0.1262507289648056,
      "learning_rate": 8.027880278802789e-05,
      "loss": 0.0033,
      "step": 1460
    },
    {
      "epoch": 5.424354243542435,
      "grad_norm": 0.17291103303432465,
      "learning_rate": 7.945879458794588e-05,
      "loss": 0.0043,
      "step": 1470
    },
    {
      "epoch": 5.461254612546125,
      "grad_norm": 0.036809444427490234,
      "learning_rate": 7.863878638786389e-05,
      "loss": 0.0038,
      "step": 1480
    },
    {
      "epoch": 5.498154981549815,
      "grad_norm": 0.027763400226831436,
      "learning_rate": 7.781877818778188e-05,
      "loss": 0.002,
      "step": 1490
    },
    {
      "epoch": 5.535055350553505,
      "grad_norm": 0.10087448358535767,
      "learning_rate": 7.699876998769989e-05,
      "loss": 0.0037,
      "step": 1500
    },
    {
      "epoch": 5.535055350553505,
      "eval_accuracy": 0.9936416184971099,
      "eval_loss": 0.002714467002078891,
      "eval_runtime": 52.8635,
      "eval_samples_per_second": 32.726,
      "eval_steps_per_second": 4.105,
      "step": 1500
    },
    {
      "epoch": 5.571955719557195,
      "grad_norm": 0.04997282847762108,
      "learning_rate": 7.617876178761788e-05,
      "loss": 0.0024,
      "step": 1510
    },
    {
      "epoch": 5.608856088560886,
      "grad_norm": 0.015307437628507614,
      "learning_rate": 7.535875358753589e-05,
      "loss": 0.0038,
      "step": 1520
    },
    {
      "epoch": 5.645756457564576,
      "grad_norm": 0.0023889692965894938,
      "learning_rate": 7.453874538745388e-05,
      "loss": 0.0038,
      "step": 1530
    },
    {
      "epoch": 5.682656826568266,
      "grad_norm": 0.040821630507707596,
      "learning_rate": 7.371873718737188e-05,
      "loss": 0.0041,
      "step": 1540
    },
    {
      "epoch": 5.719557195571956,
      "grad_norm": 0.09734688699245453,
      "learning_rate": 7.289872898728988e-05,
      "loss": 0.0038,
      "step": 1550
    },
    {
      "epoch": 5.756457564575646,
      "grad_norm": 0.016573524102568626,
      "learning_rate": 7.207872078720787e-05,
      "loss": 0.0025,
      "step": 1560
    },
    {
      "epoch": 5.793357933579336,
      "grad_norm": 0.009829455986618996,
      "learning_rate": 7.125871258712588e-05,
      "loss": 0.0041,
      "step": 1570
    },
    {
      "epoch": 5.830258302583026,
      "grad_norm": 0.07733350247144699,
      "learning_rate": 7.043870438704387e-05,
      "loss": 0.0024,
      "step": 1580
    },
    {
      "epoch": 5.867158671586716,
      "grad_norm": 0.04763353243470192,
      "learning_rate": 6.961869618696188e-05,
      "loss": 0.0025,
      "step": 1590
    },
    {
      "epoch": 5.904059040590406,
      "grad_norm": 0.009967833757400513,
      "learning_rate": 6.879868798687987e-05,
      "loss": 0.0039,
      "step": 1600
    },
    {
      "epoch": 5.904059040590406,
      "eval_accuracy": 0.9976878612716763,
      "eval_loss": 0.0022707099560648203,
      "eval_runtime": 53.0061,
      "eval_samples_per_second": 32.638,
      "eval_steps_per_second": 4.094,
      "step": 1600
    },
    {
      "epoch": 5.940959409594096,
      "grad_norm": 0.020813655108213425,
      "learning_rate": 6.797867978679788e-05,
      "loss": 0.0029,
      "step": 1610
    },
    {
      "epoch": 5.977859778597786,
      "grad_norm": 0.062338754534721375,
      "learning_rate": 6.715867158671587e-05,
      "loss": 0.0023,
      "step": 1620
    },
    {
      "epoch": 6.014760147601476,
      "grad_norm": 0.018345272168517113,
      "learning_rate": 6.633866338663388e-05,
      "loss": 0.0029,
      "step": 1630
    },
    {
      "epoch": 6.051660516605166,
      "grad_norm": 0.023753851652145386,
      "learning_rate": 6.551865518655187e-05,
      "loss": 0.0018,
      "step": 1640
    },
    {
      "epoch": 6.088560885608856,
      "grad_norm": 0.016533376649022102,
      "learning_rate": 6.469864698646987e-05,
      "loss": 0.0025,
      "step": 1650
    },
    {
      "epoch": 6.125461254612546,
      "grad_norm": 0.008262280374765396,
      "learning_rate": 6.387863878638787e-05,
      "loss": 0.0025,
      "step": 1660
    },
    {
      "epoch": 6.162361623616236,
      "grad_norm": 0.010828894563019276,
      "learning_rate": 6.305863058630586e-05,
      "loss": 0.0018,
      "step": 1670
    },
    {
      "epoch": 6.199261992619927,
      "grad_norm": 0.01505562849342823,
      "learning_rate": 6.223862238622387e-05,
      "loss": 0.0017,
      "step": 1680
    },
    {
      "epoch": 6.236162361623617,
      "grad_norm": 0.0033724599052220583,
      "learning_rate": 6.141861418614186e-05,
      "loss": 0.0018,
      "step": 1690
    },
    {
      "epoch": 6.273062730627307,
      "grad_norm": 0.026863617822527885,
      "learning_rate": 6.0598605986059866e-05,
      "loss": 0.0019,
      "step": 1700
    },
    {
      "epoch": 6.273062730627307,
      "eval_accuracy": 0.9971098265895953,
      "eval_loss": 0.0018508523935452104,
      "eval_runtime": 53.4671,
      "eval_samples_per_second": 32.356,
      "eval_steps_per_second": 4.059,
      "step": 1700
    },
    {
      "epoch": 6.3099630996309966,
      "grad_norm": 0.011449555866420269,
      "learning_rate": 5.9778597785977866e-05,
      "loss": 0.0019,
      "step": 1710
    },
    {
      "epoch": 6.3468634686346865,
      "grad_norm": 0.04869832843542099,
      "learning_rate": 5.8958589585895866e-05,
      "loss": 0.0022,
      "step": 1720
    },
    {
      "epoch": 6.3837638376383765,
      "grad_norm": 0.2211851179599762,
      "learning_rate": 5.813858138581386e-05,
      "loss": 0.0021,
      "step": 1730
    },
    {
      "epoch": 6.4206642066420665,
      "grad_norm": 0.021321425214409828,
      "learning_rate": 5.731857318573186e-05,
      "loss": 0.0017,
      "step": 1740
    },
    {
      "epoch": 6.4575645756457565,
      "grad_norm": 0.014219704084098339,
      "learning_rate": 5.649856498564986e-05,
      "loss": 0.0026,
      "step": 1750
    },
    {
      "epoch": 6.4944649446494465,
      "grad_norm": 0.011448672972619534,
      "learning_rate": 5.567855678556786e-05,
      "loss": 0.0026,
      "step": 1760
    },
    {
      "epoch": 6.531365313653136,
      "grad_norm": 0.008601737208664417,
      "learning_rate": 5.485854858548586e-05,
      "loss": 0.0016,
      "step": 1770
    },
    {
      "epoch": 6.568265682656826,
      "grad_norm": 0.013017091900110245,
      "learning_rate": 5.403854038540386e-05,
      "loss": 0.0023,
      "step": 1780
    },
    {
      "epoch": 6.605166051660516,
      "grad_norm": 0.008623339235782623,
      "learning_rate": 5.321853218532186e-05,
      "loss": 0.0019,
      "step": 1790
    },
    {
      "epoch": 6.642066420664206,
      "grad_norm": 0.027804844081401825,
      "learning_rate": 5.239852398523986e-05,
      "loss": 0.0019,
      "step": 1800
    },
    {
      "epoch": 6.642066420664206,
      "eval_accuracy": 0.9988439306358381,
      "eval_loss": 0.0016573916655033827,
      "eval_runtime": 53.3851,
      "eval_samples_per_second": 32.406,
      "eval_steps_per_second": 4.065,
      "step": 1800
    },
    {
      "epoch": 6.678966789667896,
      "grad_norm": 0.027201993390917778,
      "learning_rate": 5.1578515785157857e-05,
      "loss": 0.0021,
      "step": 1810
    },
    {
      "epoch": 6.715867158671586,
      "grad_norm": 0.07681945711374283,
      "learning_rate": 5.0758507585075856e-05,
      "loss": 0.0022,
      "step": 1820
    },
    {
      "epoch": 6.752767527675276,
      "grad_norm": 0.010648882016539574,
      "learning_rate": 4.9938499384993856e-05,
      "loss": 0.0017,
      "step": 1830
    },
    {
      "epoch": 6.789667896678967,
      "grad_norm": 0.01609162986278534,
      "learning_rate": 4.9118491184911856e-05,
      "loss": 0.0016,
      "step": 1840
    },
    {
      "epoch": 6.826568265682657,
      "grad_norm": 0.013306355103850365,
      "learning_rate": 4.829848298482985e-05,
      "loss": 0.0016,
      "step": 1850
    },
    {
      "epoch": 6.863468634686347,
      "grad_norm": 0.015135644935071468,
      "learning_rate": 4.747847478474785e-05,
      "loss": 0.002,
      "step": 1860
    },
    {
      "epoch": 6.900369003690037,
      "grad_norm": 0.019170399755239487,
      "learning_rate": 4.665846658466585e-05,
      "loss": 0.0015,
      "step": 1870
    },
    {
      "epoch": 6.937269372693727,
      "grad_norm": 0.02915087901055813,
      "learning_rate": 4.583845838458385e-05,
      "loss": 0.0012,
      "step": 1880
    },
    {
      "epoch": 6.974169741697417,
      "grad_norm": 0.023285195231437683,
      "learning_rate": 4.501845018450185e-05,
      "loss": 0.0019,
      "step": 1890
    },
    {
      "epoch": 7.011070110701107,
      "grad_norm": 0.013183875940740108,
      "learning_rate": 4.419844198441985e-05,
      "loss": 0.0016,
      "step": 1900
    },
    {
      "epoch": 7.011070110701107,
      "eval_accuracy": 1.0,
      "eval_loss": 0.0015279006911441684,
      "eval_runtime": 53.4443,
      "eval_samples_per_second": 32.37,
      "eval_steps_per_second": 4.06,
      "step": 1900
    },
    {
      "epoch": 7.047970479704797,
      "grad_norm": 0.008983751758933067,
      "learning_rate": 4.337843378433785e-05,
      "loss": 0.0015,
      "step": 1910
    },
    {
      "epoch": 7.084870848708487,
      "grad_norm": 0.003578983247280121,
      "learning_rate": 4.2558425584255847e-05,
      "loss": 0.0013,
      "step": 1920
    },
    {
      "epoch": 7.121771217712177,
      "grad_norm": 0.009858568198978901,
      "learning_rate": 4.1738417384173846e-05,
      "loss": 0.0017,
      "step": 1930
    },
    {
      "epoch": 7.158671586715867,
      "grad_norm": 0.01723896525800228,
      "learning_rate": 4.0918409184091846e-05,
      "loss": 0.0011,
      "step": 1940
    },
    {
      "epoch": 7.195571955719557,
      "grad_norm": 0.007069133687764406,
      "learning_rate": 4.0098400984009846e-05,
      "loss": 0.0012,
      "step": 1950
    },
    {
      "epoch": 7.232472324723247,
      "grad_norm": 0.006838188972324133,
      "learning_rate": 3.927839278392784e-05,
      "loss": 0.0021,
      "step": 1960
    },
    {
      "epoch": 7.269372693726937,
      "grad_norm": 0.011245744302868843,
      "learning_rate": 3.845838458384584e-05,
      "loss": 0.0018,
      "step": 1970
    },
    {
      "epoch": 7.306273062730627,
      "grad_norm": 0.002784464741125703,
      "learning_rate": 3.763837638376384e-05,
      "loss": 0.0013,
      "step": 1980
    },
    {
      "epoch": 7.343173431734318,
      "grad_norm": 0.007723368238657713,
      "learning_rate": 3.681836818368184e-05,
      "loss": 0.0017,
      "step": 1990
    },
    {
      "epoch": 7.380073800738008,
      "grad_norm": 0.016839824616909027,
      "learning_rate": 3.599835998359984e-05,
      "loss": 0.002,
      "step": 2000
    },
    {
      "epoch": 7.380073800738008,
      "eval_accuracy": 1.0,
      "eval_loss": 0.0014427828136831522,
      "eval_runtime": 53.204,
      "eval_samples_per_second": 32.516,
      "eval_steps_per_second": 4.079,
      "step": 2000
    },
    {
      "epoch": 7.416974169741698,
      "grad_norm": 0.015248976647853851,
      "learning_rate": 3.517835178351784e-05,
      "loss": 0.0018,
      "step": 2010
    },
    {
      "epoch": 7.453874538745388,
      "grad_norm": 0.013244880363345146,
      "learning_rate": 3.435834358343584e-05,
      "loss": 0.0022,
      "step": 2020
    },
    {
      "epoch": 7.490774907749078,
      "grad_norm": 0.009770199656486511,
      "learning_rate": 3.353833538335384e-05,
      "loss": 0.0017,
      "step": 2030
    },
    {
      "epoch": 7.527675276752768,
      "grad_norm": 0.010310194455087185,
      "learning_rate": 3.2718327183271836e-05,
      "loss": 0.0018,
      "step": 2040
    },
    {
      "epoch": 7.564575645756458,
      "grad_norm": 0.0034243548288941383,
      "learning_rate": 3.1898318983189836e-05,
      "loss": 0.0014,
      "step": 2050
    },
    {
      "epoch": 7.601476014760148,
      "grad_norm": 0.005529376212507486,
      "learning_rate": 3.1078310783107836e-05,
      "loss": 0.0013,
      "step": 2060
    },
    {
      "epoch": 7.638376383763838,
      "grad_norm": 0.005842520389705896,
      "learning_rate": 3.0258302583025832e-05,
      "loss": 0.0017,
      "step": 2070
    },
    {
      "epoch": 7.675276752767528,
      "grad_norm": 0.00041994985076598823,
      "learning_rate": 2.9438294382943832e-05,
      "loss": 0.0016,
      "step": 2080
    },
    {
      "epoch": 7.712177121771218,
      "grad_norm": 0.005925205536186695,
      "learning_rate": 2.861828618286183e-05,
      "loss": 0.0013,
      "step": 2090
    },
    {
      "epoch": 7.749077490774908,
      "grad_norm": 0.00783997867256403,
      "learning_rate": 2.779827798277983e-05,
      "loss": 0.0013,
      "step": 2100
    },
    {
      "epoch": 7.749077490774908,
      "eval_accuracy": 1.0,
      "eval_loss": 0.0013818982988595963,
      "eval_runtime": 53.3908,
      "eval_samples_per_second": 32.403,
      "eval_steps_per_second": 4.064,
      "step": 2100
    },
    {
      "epoch": 7.785977859778598,
      "grad_norm": 0.0075693195685744286,
      "learning_rate": 2.6978269782697828e-05,
      "loss": 0.0018,
      "step": 2110
    },
    {
      "epoch": 7.822878228782288,
      "grad_norm": 0.00795042049139738,
      "learning_rate": 2.6158261582615827e-05,
      "loss": 0.0011,
      "step": 2120
    },
    {
      "epoch": 7.8597785977859775,
      "grad_norm": 0.00859429594129324,
      "learning_rate": 2.5338253382533827e-05,
      "loss": 0.0015,
      "step": 2130
    },
    {
      "epoch": 7.8966789667896675,
      "grad_norm": 0.005215910263359547,
      "learning_rate": 2.4518245182451827e-05,
      "loss": 0.0016,
      "step": 2140
    },
    {
      "epoch": 7.9335793357933575,
      "grad_norm": 0.0038353356067091227,
      "learning_rate": 2.3698236982369827e-05,
      "loss": 0.0015,
      "step": 2150
    },
    {
      "epoch": 7.970479704797048,
      "grad_norm": 0.00975708942860365,
      "learning_rate": 2.2878228782287826e-05,
      "loss": 0.0015,
      "step": 2160
    },
    {
      "epoch": 8.007380073800737,
      "grad_norm": 0.006485125049948692,
      "learning_rate": 2.2058220582205823e-05,
      "loss": 0.0012,
      "step": 2170
    },
    {
      "epoch": 8.044280442804428,
      "grad_norm": 0.013848377391695976,
      "learning_rate": 2.1238212382123822e-05,
      "loss": 0.0014,
      "step": 2180
    },
    {
      "epoch": 8.081180811808117,
      "grad_norm": 0.012979693710803986,
      "learning_rate": 2.0418204182041822e-05,
      "loss": 0.0015,
      "step": 2190
    },
    {
      "epoch": 8.118081180811808,
      "grad_norm": 0.005968283861875534,
      "learning_rate": 1.9598195981959822e-05,
      "loss": 0.0015,
      "step": 2200
    },
    {
      "epoch": 8.118081180811808,
      "eval_accuracy": 1.0,
      "eval_loss": 0.0013393920380622149,
      "eval_runtime": 53.2752,
      "eval_samples_per_second": 32.473,
      "eval_steps_per_second": 4.073,
      "step": 2200
    },
    {
      "epoch": 8.154981549815497,
      "grad_norm": 0.010278232395648956,
      "learning_rate": 1.877818778187782e-05,
      "loss": 0.0009,
      "step": 2210
    },
    {
      "epoch": 8.191881918819188,
      "grad_norm": 0.007073582150042057,
      "learning_rate": 1.795817958179582e-05,
      "loss": 0.0016,
      "step": 2220
    },
    {
      "epoch": 8.228782287822877,
      "grad_norm": 0.0020103720016777515,
      "learning_rate": 1.7138171381713817e-05,
      "loss": 0.0014,
      "step": 2230
    },
    {
      "epoch": 8.265682656826568,
      "grad_norm": 0.006469924468547106,
      "learning_rate": 1.6318163181631817e-05,
      "loss": 0.0014,
      "step": 2240
    },
    {
      "epoch": 8.302583025830259,
      "grad_norm": 0.006924558896571398,
      "learning_rate": 1.5498154981549817e-05,
      "loss": 0.0013,
      "step": 2250
    },
    {
      "epoch": 8.339483394833948,
      "grad_norm": 0.010701497085392475,
      "learning_rate": 1.4678146781467817e-05,
      "loss": 0.0013,
      "step": 2260
    },
    {
      "epoch": 8.376383763837639,
      "grad_norm": 0.012286316603422165,
      "learning_rate": 1.3858138581385816e-05,
      "loss": 0.0016,
      "step": 2270
    },
    {
      "epoch": 8.413284132841328,
      "grad_norm": 0.010458541102707386,
      "learning_rate": 1.3038130381303814e-05,
      "loss": 0.0018,
      "step": 2280
    },
    {
      "epoch": 8.450184501845019,
      "grad_norm": 0.010043789632618427,
      "learning_rate": 1.2218122181221812e-05,
      "loss": 0.0012,
      "step": 2290
    },
    {
      "epoch": 8.487084870848708,
      "grad_norm": 0.00660703144967556,
      "learning_rate": 1.1398113981139812e-05,
      "loss": 0.0011,
      "step": 2300
    },
    {
      "epoch": 8.487084870848708,
      "eval_accuracy": 1.0,
      "eval_loss": 0.0013166143326088786,
      "eval_runtime": 53.3458,
      "eval_samples_per_second": 32.43,
      "eval_steps_per_second": 4.068,
      "step": 2300
    },
    {
      "epoch": 8.523985239852399,
      "grad_norm": 0.006171096581965685,
      "learning_rate": 1.057810578105781e-05,
      "loss": 0.0014,
      "step": 2310
    },
    {
      "epoch": 8.560885608856088,
      "grad_norm": 0.005463199689984322,
      "learning_rate": 9.75809758097581e-06,
      "loss": 0.0013,
      "step": 2320
    },
    {
      "epoch": 8.597785977859779,
      "grad_norm": 0.008069201372563839,
      "learning_rate": 8.93808938089381e-06,
      "loss": 0.0015,
      "step": 2330
    },
    {
      "epoch": 8.634686346863468,
      "grad_norm": 0.007250032387673855,
      "learning_rate": 8.118081180811808e-06,
      "loss": 0.0017,
      "step": 2340
    },
    {
      "epoch": 8.671586715867159,
      "grad_norm": 0.011325874365866184,
      "learning_rate": 7.298072980729807e-06,
      "loss": 0.0014,
      "step": 2350
    },
    {
      "epoch": 8.708487084870848,
      "grad_norm": 0.019912155345082283,
      "learning_rate": 6.478064780647806e-06,
      "loss": 0.0018,
      "step": 2360
    },
    {
      "epoch": 8.745387453874539,
      "grad_norm": 0.0008021511603146791,
      "learning_rate": 5.658056580565806e-06,
      "loss": 0.0012,
      "step": 2370
    },
    {
      "epoch": 8.782287822878228,
      "grad_norm": 0.01182828564196825,
      "learning_rate": 4.838048380483805e-06,
      "loss": 0.0016,
      "step": 2380
    },
    {
      "epoch": 8.819188191881919,
      "grad_norm": 0.004345746710896492,
      "learning_rate": 4.018040180401804e-06,
      "loss": 0.001,
      "step": 2390
    },
    {
      "epoch": 8.85608856088561,
      "grad_norm": 0.0077620395459234715,
      "learning_rate": 3.198031980319803e-06,
      "loss": 0.0013,
      "step": 2400
    },
    {
      "epoch": 8.85608856088561,
      "eval_accuracy": 1.0,
      "eval_loss": 0.0013051489368081093,
      "eval_runtime": 52.9588,
      "eval_samples_per_second": 32.667,
      "eval_steps_per_second": 4.098,
      "step": 2400
    },
    {
      "epoch": 8.892988929889299,
      "grad_norm": 0.01009498443454504,
      "learning_rate": 2.3780237802378026e-06,
      "loss": 0.0015,
      "step": 2410
    },
    {
      "epoch": 8.92988929889299,
      "grad_norm": 0.00728220259770751,
      "learning_rate": 1.5580155801558017e-06,
      "loss": 0.0015,
      "step": 2420
    },
    {
      "epoch": 8.966789667896679,
      "grad_norm": 0.014378752559423447,
      "learning_rate": 7.380073800738008e-07,
      "loss": 0.0014,
      "step": 2430
    },
    {
      "epoch": 9.0,
      "step": 2439,
      "total_flos": 3.017424657591816e+18,
      "train_loss": 0.01402124272053813,
      "train_runtime": 3085.6034,
      "train_samples_per_second": 12.618,
      "train_steps_per_second": 0.79
    }
  ],
  "logging_steps": 10,
  "max_steps": 2439,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 3.017424657591816e+18,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}