{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 9.985272459499264,
  "eval_steps": 500,
  "global_step": 3390,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.029455081001472753,
      "grad_norm": 10.109375,
      "learning_rate": 0.00019999570594853575,
      "loss": 2.6151,
      "step": 10
    },
    {
      "epoch": 0.05891016200294551,
      "grad_norm": 9.828125,
      "learning_rate": 0.00019998282416292055,
      "loss": 2.4649,
      "step": 20
    },
    {
      "epoch": 0.08836524300441827,
      "grad_norm": 8.59375,
      "learning_rate": 0.00019996135574945544,
      "loss": 2.4015,
      "step": 30
    },
    {
      "epoch": 0.11782032400589101,
      "grad_norm": 7.8359375,
      "learning_rate": 0.00019993130255186977,
      "loss": 2.4925,
      "step": 40
    },
    {
      "epoch": 0.14727540500736377,
      "grad_norm": 7.76953125,
      "learning_rate": 0.00019989266715116316,
      "loss": 2.4309,
      "step": 50
    },
    {
      "epoch": 0.17673048600883653,
      "grad_norm": 10.2265625,
      "learning_rate": 0.0001998454528653836,
      "loss": 2.3814,
      "step": 60
    },
    {
      "epoch": 0.20618556701030927,
      "grad_norm": 10.375,
      "learning_rate": 0.00019978966374934254,
      "loss": 2.3852,
      "step": 70
    },
    {
      "epoch": 0.23564064801178203,
      "grad_norm": 9.4453125,
      "learning_rate": 0.00019972530459426663,
      "loss": 2.396,
      "step": 80
    },
    {
      "epoch": 0.2650957290132548,
      "grad_norm": 9.21875,
      "learning_rate": 0.00019965238092738643,
      "loss": 2.386,
      "step": 90
    },
    {
      "epoch": 0.29455081001472755,
      "grad_norm": 10.1328125,
      "learning_rate": 0.00019957089901146148,
      "loss": 2.3984,
      "step": 100
    },
    {
      "epoch": 0.3240058910162003,
      "grad_norm": 8.484375,
      "learning_rate": 0.00019948086584424256,
      "loss": 2.3893,
      "step": 110
    },
    {
      "epoch": 0.35346097201767307,
      "grad_norm": 7.83203125,
      "learning_rate": 0.0001993822891578708,
      "loss": 2.4262,
      "step": 120
    },
    {
      "epoch": 0.38291605301914583,
      "grad_norm": 6.61328125,
      "learning_rate": 0.00019927517741821343,
      "loss": 2.3399,
      "step": 130
    },
    {
      "epoch": 0.41237113402061853,
      "grad_norm": 7.91015625,
      "learning_rate": 0.0001991595398241369,
      "loss": 2.4667,
      "step": 140
    },
    {
      "epoch": 0.4418262150220913,
      "grad_norm": 5.46484375,
      "learning_rate": 0.0001990353863067169,
      "loss": 2.4316,
      "step": 150
    },
    {
      "epoch": 0.47128129602356406,
      "grad_norm": 7.76171875,
      "learning_rate": 0.00019890272752838518,
      "loss": 2.4068,
      "step": 160
    },
    {
      "epoch": 0.5007363770250368,
      "grad_norm": 6.78515625,
      "learning_rate": 0.00019876157488201424,
      "loss": 2.4295,
      "step": 170
    },
    {
      "epoch": 0.5301914580265096,
      "grad_norm": 6.67578125,
      "learning_rate": 0.00019861194048993863,
      "loss": 2.4124,
      "step": 180
    },
    {
      "epoch": 0.5596465390279823,
      "grad_norm": 4.140625,
      "learning_rate": 0.00019845383720291392,
      "loss": 2.3651,
      "step": 190
    },
    {
      "epoch": 0.5891016200294551,
      "grad_norm": 3.337890625,
      "learning_rate": 0.00019828727859901317,
      "loss": 2.4137,
      "step": 200
    },
    {
      "epoch": 0.6185567010309279,
      "grad_norm": 4.01953125,
      "learning_rate": 0.0001981122789824607,
      "loss": 2.3851,
      "step": 210
    },
    {
      "epoch": 0.6480117820324006,
      "grad_norm": 4.94140625,
      "learning_rate": 0.00019792885338240374,
      "loss": 2.3647,
      "step": 220
    },
    {
      "epoch": 0.6774668630338734,
      "grad_norm": 4.5234375,
      "learning_rate": 0.00019773701755162158,
      "loss": 2.2891,
      "step": 230
    },
    {
      "epoch": 0.7069219440353461,
      "grad_norm": 3.619140625,
      "learning_rate": 0.00019753678796517282,
      "loss": 2.33,
      "step": 240
    },
    {
      "epoch": 0.7363770250368189,
      "grad_norm": 3.244140625,
      "learning_rate": 0.00019732818181898045,
      "loss": 2.3588,
      "step": 250
    },
    {
      "epoch": 0.7658321060382917,
      "grad_norm": 5.95703125,
      "learning_rate": 0.00019711121702835504,
      "loss": 2.4006,
      "step": 260
    },
    {
      "epoch": 0.7952871870397643,
      "grad_norm": 3.873046875,
      "learning_rate": 0.00019688591222645607,
      "loss": 2.3874,
      "step": 270
    },
    {
      "epoch": 0.8247422680412371,
      "grad_norm": 3.833984375,
      "learning_rate": 0.0001966522867626919,
      "loss": 2.3104,
      "step": 280
    },
    {
      "epoch": 0.8541973490427098,
      "grad_norm": 4.55078125,
      "learning_rate": 0.00019641036070105778,
      "loss": 2.3146,
      "step": 290
    },
    {
      "epoch": 0.8836524300441826,
      "grad_norm": 5.9765625,
      "learning_rate": 0.0001961601548184129,
      "loss": 2.3112,
      "step": 300
    },
    {
      "epoch": 0.9131075110456554,
      "grad_norm": 5.18359375,
      "learning_rate": 0.00019590169060269602,
      "loss": 2.2807,
      "step": 310
    },
    {
      "epoch": 0.9425625920471281,
      "grad_norm": 8.359375,
      "learning_rate": 0.00019563499025107998,
      "loss": 2.3536,
      "step": 320
    },
    {
      "epoch": 0.9720176730486009,
      "grad_norm": 5.60546875,
      "learning_rate": 0.00019536007666806556,
      "loss": 2.3461,
      "step": 330
    },
    {
      "epoch": 1.0014727540500736,
      "grad_norm": 4.32421875,
      "learning_rate": 0.00019507697346351414,
      "loss": 2.3032,
      "step": 340
    },
    {
      "epoch": 1.0309278350515463,
      "grad_norm": 3.591796875,
      "learning_rate": 0.00019478570495062037,
      "loss": 2.2973,
      "step": 350
    },
    {
      "epoch": 1.0603829160530192,
      "grad_norm": 5.73046875,
      "learning_rate": 0.0001944862961438239,
      "loss": 2.3204,
      "step": 360
    },
    {
      "epoch": 1.0898379970544918,
      "grad_norm": 3.8125,
      "learning_rate": 0.0001941787727566613,
      "loss": 2.3208,
      "step": 370
    },
    {
      "epoch": 1.1192930780559647,
      "grad_norm": 4.33984375,
      "learning_rate": 0.00019386316119955756,
      "loss": 2.4171,
      "step": 380
    },
    {
      "epoch": 1.1487481590574373,
      "grad_norm": 3.208984375,
      "learning_rate": 0.00019353948857755803,
      "loss": 2.2795,
      "step": 390
    },
    {
      "epoch": 1.1782032400589102,
      "grad_norm": 4.07421875,
      "learning_rate": 0.00019320778268800066,
      "loss": 2.2644,
      "step": 400
    },
    {
      "epoch": 1.2076583210603828,
      "grad_norm": 4.53125,
      "learning_rate": 0.00019286807201812867,
      "loss": 2.2787,
      "step": 410
    },
    {
      "epoch": 1.2371134020618557,
      "grad_norm": 3.912109375,
      "learning_rate": 0.00019252038574264405,
      "loss": 2.2143,
      "step": 420
    },
    {
      "epoch": 1.2665684830633284,
      "grad_norm": 3.607421875,
      "learning_rate": 0.00019216475372120197,
      "loss": 2.2773,
      "step": 430
    },
    {
      "epoch": 1.2960235640648012,
      "grad_norm": 2.80859375,
      "learning_rate": 0.00019180120649584653,
      "loss": 2.2923,
      "step": 440
    },
    {
      "epoch": 1.3254786450662739,
      "grad_norm": 2.791015625,
      "learning_rate": 0.00019142977528838762,
      "loss": 2.2583,
      "step": 450
    },
    {
      "epoch": 1.3549337260677468,
      "grad_norm": 2.541015625,
      "learning_rate": 0.00019105049199771962,
      "loss": 2.2466,
      "step": 460
    },
    {
      "epoch": 1.3843888070692194,
      "grad_norm": 2.541015625,
      "learning_rate": 0.00019066338919708197,
      "loss": 2.2686,
      "step": 470
    },
    {
      "epoch": 1.413843888070692,
      "grad_norm": 3.03125,
      "learning_rate": 0.00019026850013126157,
      "loss": 2.246,
      "step": 480
    },
    {
      "epoch": 1.443298969072165,
      "grad_norm": 2.46484375,
      "learning_rate": 0.0001898658587137379,
      "loss": 2.2931,
      "step": 490
    },
    {
      "epoch": 1.4727540500736378,
      "grad_norm": 2.740234375,
      "learning_rate": 0.0001894554995237703,
      "loss": 2.255,
      "step": 500
    },
    {
      "epoch": 1.5022091310751104,
      "grad_norm": 2.658203125,
      "learning_rate": 0.00018903745780342839,
      "loss": 2.2697,
      "step": 510
    },
    {
      "epoch": 1.531664212076583,
      "grad_norm": 2.486328125,
      "learning_rate": 0.0001886117694545654,
      "loss": 2.2511,
      "step": 520
    },
    {
      "epoch": 1.561119293078056,
      "grad_norm": 3.216796875,
      "learning_rate": 0.00018817847103573486,
      "loss": 2.2199,
      "step": 530
    },
    {
      "epoch": 1.5905743740795288,
      "grad_norm": 2.8984375,
      "learning_rate": 0.00018773759975905098,
      "loss": 2.2668,
      "step": 540
    },
    {
      "epoch": 1.6200294550810015,
      "grad_norm": 3.052734375,
      "learning_rate": 0.00018728919348699283,
      "loss": 2.283,
      "step": 550
    },
    {
      "epoch": 1.6494845360824741,
      "grad_norm": 2.8515625,
      "learning_rate": 0.00018683329072915252,
      "loss": 2.2239,
      "step": 560
    },
    {
      "epoch": 1.678939617083947,
      "grad_norm": 2.607421875,
      "learning_rate": 0.0001863699306389282,
      "loss": 2.1978,
      "step": 570
    },
    {
      "epoch": 1.7083946980854199,
      "grad_norm": 2.978515625,
      "learning_rate": 0.0001858991530101613,
      "loss": 2.2468,
      "step": 580
    },
    {
      "epoch": 1.7378497790868925,
      "grad_norm": 2.78125,
      "learning_rate": 0.0001854209982737192,
      "loss": 2.2494,
      "step": 590
    },
    {
      "epoch": 1.7673048600883652,
      "grad_norm": 2.75390625,
      "learning_rate": 0.00018493550749402278,
      "loss": 2.2163,
      "step": 600
    },
    {
      "epoch": 1.7967599410898378,
      "grad_norm": 2.908203125,
      "learning_rate": 0.0001844427223655199,
      "loss": 2.2156,
      "step": 610
    },
    {
      "epoch": 1.8262150220913107,
      "grad_norm": 3.236328125,
      "learning_rate": 0.00018394268520910466,
      "loss": 2.2757,
      "step": 620
    },
    {
      "epoch": 1.8556701030927836,
      "grad_norm": 2.51953125,
      "learning_rate": 0.00018343543896848273,
      "loss": 2.1927,
      "step": 630
    },
    {
      "epoch": 1.8851251840942562,
      "grad_norm": 3.451171875,
      "learning_rate": 0.00018292102720648333,
      "loss": 2.1343,
      "step": 640
    },
    {
      "epoch": 1.9145802650957289,
      "grad_norm": 2.0625,
      "learning_rate": 0.00018239949410131802,
      "loss": 2.2521,
      "step": 650
    },
    {
      "epoch": 1.9440353460972017,
      "grad_norm": 3.177734375,
      "learning_rate": 0.00018187088444278674,
      "loss": 2.2107,
      "step": 660
    },
    {
      "epoch": 1.9734904270986746,
      "grad_norm": 2.892578125,
      "learning_rate": 0.00018133524362843104,
      "loss": 2.1933,
      "step": 670
    },
    {
      "epoch": 2.0029455081001473,
      "grad_norm": 2.4296875,
      "learning_rate": 0.00018079261765963537,
      "loss": 2.1848,
      "step": 680
    },
    {
      "epoch": 2.03240058910162,
      "grad_norm": 2.189453125,
      "learning_rate": 0.00018024305313767646,
      "loss": 2.2219,
      "step": 690
    },
    {
      "epoch": 2.0618556701030926,
      "grad_norm": 2.189453125,
      "learning_rate": 0.00017968659725972112,
      "loss": 2.1635,
      "step": 700
    },
    {
      "epoch": 2.0913107511045657,
      "grad_norm": 2.607421875,
      "learning_rate": 0.00017912329781477287,
      "loss": 2.2145,
      "step": 710
    },
    {
      "epoch": 2.1207658321060383,
      "grad_norm": 1.8662109375,
      "learning_rate": 0.00017855320317956784,
      "loss": 2.178,
      "step": 720
    },
    {
      "epoch": 2.150220913107511,
      "grad_norm": 2.13671875,
      "learning_rate": 0.00017797636231442016,
      "loss": 2.1433,
      "step": 730
    },
    {
      "epoch": 2.1796759941089836,
      "grad_norm": 2.98828125,
      "learning_rate": 0.000177392824759017,
      "loss": 2.1158,
      "step": 740
    },
    {
      "epoch": 2.2091310751104567,
      "grad_norm": 2.25,
      "learning_rate": 0.0001768026406281642,
      "loss": 2.2127,
      "step": 750
    },
    {
      "epoch": 2.2385861561119293,
      "grad_norm": 2.484375,
      "learning_rate": 0.00017620586060748252,
      "loss": 2.1268,
      "step": 760
    },
    {
      "epoch": 2.268041237113402,
      "grad_norm": 1.9482421875,
      "learning_rate": 0.00017560253594905425,
      "loss": 2.1628,
      "step": 770
    },
    {
      "epoch": 2.2974963181148746,
      "grad_norm": 1.94140625,
      "learning_rate": 0.00017499271846702213,
      "loss": 2.1489,
      "step": 780
    },
    {
      "epoch": 2.3269513991163477,
      "grad_norm": 1.6005859375,
      "learning_rate": 0.0001743764605331392,
      "loss": 2.1078,
      "step": 790
    },
    {
      "epoch": 2.3564064801178204,
      "grad_norm": 1.677734375,
      "learning_rate": 0.00017375381507227108,
      "loss": 2.1365,
      "step": 800
    },
    {
      "epoch": 2.385861561119293,
      "grad_norm": 1.537109375,
      "learning_rate": 0.00017312483555785086,
      "loss": 2.1179,
      "step": 810
    },
    {
      "epoch": 2.4153166421207657,
      "grad_norm": 1.8291015625,
      "learning_rate": 0.00017248957600728664,
      "loss": 2.0896,
      "step": 820
    },
    {
      "epoch": 2.444771723122239,
      "grad_norm": 1.9423828125,
      "learning_rate": 0.00017184809097732246,
      "loss": 2.1057,
      "step": 830
    },
    {
      "epoch": 2.4742268041237114,
      "grad_norm": 1.443359375,
      "learning_rate": 0.00017120043555935298,
      "loss": 2.1325,
      "step": 840
    },
    {
      "epoch": 2.503681885125184,
      "grad_norm": 2.630859375,
      "learning_rate": 0.00017054666537469213,
      "loss": 2.0701,
      "step": 850
    },
    {
      "epoch": 2.5331369661266567,
      "grad_norm": 1.95703125,
      "learning_rate": 0.00016988683656979624,
      "loss": 2.1342,
      "step": 860
    },
    {
      "epoch": 2.5625920471281294,
      "grad_norm": 1.50390625,
      "learning_rate": 0.00016922100581144228,
      "loss": 2.1223,
      "step": 870
    },
    {
      "epoch": 2.5920471281296025,
      "grad_norm": 2.083984375,
      "learning_rate": 0.00016854923028186111,
      "loss": 2.1597,
      "step": 880
    },
    {
      "epoch": 2.621502209131075,
      "grad_norm": 1.7548828125,
      "learning_rate": 0.00016787156767382659,
      "loss": 2.1915,
      "step": 890
    },
    {
      "epoch": 2.6509572901325478,
      "grad_norm": 1.4912109375,
      "learning_rate": 0.00016718807618570106,
      "loss": 2.1678,
      "step": 900
    },
    {
      "epoch": 2.680412371134021,
      "grad_norm": 3.720703125,
      "learning_rate": 0.00016649881451643705,
      "loss": 2.1261,
      "step": 910
    },
    {
      "epoch": 2.7098674521354935,
      "grad_norm": 1.833984375,
      "learning_rate": 0.0001658038418605361,
      "loss": 2.0961,
      "step": 920
    },
    {
      "epoch": 2.739322533136966,
      "grad_norm": 1.80859375,
      "learning_rate": 0.00016510321790296525,
      "loss": 2.106,
      "step": 930
    },
    {
      "epoch": 2.768777614138439,
      "grad_norm": 1.5380859375,
      "learning_rate": 0.00016439700281403114,
      "loss": 2.112,
      "step": 940
    },
    {
      "epoch": 2.7982326951399115,
      "grad_norm": 1.341796875,
      "learning_rate": 0.00016368525724421248,
      "loss": 2.1424,
      "step": 950
    },
    {
      "epoch": 2.827687776141384,
      "grad_norm": 1.8505859375,
      "learning_rate": 0.00016296804231895142,
      "loss": 2.0889,
      "step": 960
    },
    {
      "epoch": 2.857142857142857,
      "grad_norm": 1.58203125,
      "learning_rate": 0.00016224541963340391,
      "loss": 2.0933,
      "step": 970
    },
    {
      "epoch": 2.88659793814433,
      "grad_norm": 1.447265625,
      "learning_rate": 0.00016151745124715002,
      "loss": 2.0716,
      "step": 980
    },
    {
      "epoch": 2.9160530191458025,
      "grad_norm": 1.2646484375,
      "learning_rate": 0.00016078419967886402,
      "loss": 2.0821,
      "step": 990
    },
    {
      "epoch": 2.9455081001472756,
      "grad_norm": 1.1318359375,
      "learning_rate": 0.00016004572790094535,
      "loss": 2.1024,
      "step": 1000
    },
    {
      "epoch": 2.9749631811487482,
      "grad_norm": 1.330078125,
      "learning_rate": 0.00015930209933411036,
      "loss": 2.0347,
      "step": 1010
    },
    {
      "epoch": 3.004418262150221,
      "grad_norm": 1.2509765625,
      "learning_rate": 0.00015855337784194577,
      "loss": 2.0021,
      "step": 1020
    },
    {
      "epoch": 3.0338733431516935,
      "grad_norm": 1.4970703125,
      "learning_rate": 0.00015779962772542402,
      "loss": 2.0147,
      "step": 1030
    },
    {
      "epoch": 3.063328424153166,
      "grad_norm": 1.4833984375,
      "learning_rate": 0.0001570409137173809,
      "loss": 2.078,
      "step": 1040
    },
    {
      "epoch": 3.0927835051546393,
      "grad_norm": 1.5166015625,
      "learning_rate": 0.00015627730097695638,
      "loss": 2.0246,
      "step": 1050
    },
    {
      "epoch": 3.122238586156112,
      "grad_norm": 1.5107421875,
      "learning_rate": 0.00015550885508399856,
      "loss": 2.0394,
      "step": 1060
    },
    {
      "epoch": 3.1516936671575846,
      "grad_norm": 1.375,
      "learning_rate": 0.00015473564203343174,
      "loss": 2.0159,
      "step": 1070
    },
    {
      "epoch": 3.1811487481590572,
      "grad_norm": 1.2431640625,
      "learning_rate": 0.00015395772822958845,
      "loss": 2.0241,
      "step": 1080
    },
    {
      "epoch": 3.2106038291605303,
      "grad_norm": 1.3994140625,
      "learning_rate": 0.00015317518048050697,
      "loss": 1.9786,
      "step": 1090
    },
    {
      "epoch": 3.240058910162003,
      "grad_norm": 1.34375,
      "learning_rate": 0.00015238806599219336,
      "loss": 2.0276,
      "step": 1100
    },
    {
      "epoch": 3.2695139911634756,
      "grad_norm": 1.251953125,
      "learning_rate": 0.0001515964523628501,
      "loss": 1.9723,
      "step": 1110
    },
    {
      "epoch": 3.2989690721649483,
      "grad_norm": 1.392578125,
      "learning_rate": 0.00015080040757707046,
      "loss": 1.951,
      "step": 1120
    },
    {
      "epoch": 3.3284241531664214,
      "grad_norm": 1.380859375,
      "learning_rate": 0.00015000000000000001,
      "loss": 1.989,
      "step": 1130
    },
    {
      "epoch": 3.357879234167894,
      "grad_norm": 1.5498046875,
      "learning_rate": 0.00014919529837146528,
      "loss": 2.0103,
      "step": 1140
    },
    {
      "epoch": 3.3873343151693667,
      "grad_norm": 1.5712890625,
      "learning_rate": 0.00014838637180007047,
      "loss": 1.9914,
      "step": 1150
    },
    {
      "epoch": 3.4167893961708393,
      "grad_norm": 1.8037109375,
      "learning_rate": 0.00014757328975726207,
      "loss": 1.9981,
      "step": 1160
    },
    {
      "epoch": 3.4462444771723124,
      "grad_norm": 1.3818359375,
      "learning_rate": 0.0001467561220713628,
      "loss": 2.0569,
      "step": 1170
    },
    {
      "epoch": 3.475699558173785,
      "grad_norm": 1.263671875,
      "learning_rate": 0.00014593493892157473,
      "loss": 2.0572,
      "step": 1180
    },
    {
      "epoch": 3.5051546391752577,
      "grad_norm": 1.322265625,
      "learning_rate": 0.00014510981083195188,
      "loss": 2.0181,
      "step": 1190
    },
    {
      "epoch": 3.5346097201767304,
      "grad_norm": 1.2841796875,
      "learning_rate": 0.00014428080866534396,
      "loss": 1.9768,
      "step": 1200
    },
    {
      "epoch": 3.564064801178203,
      "grad_norm": 1.0927734375,
      "learning_rate": 0.00014344800361731027,
      "loss": 1.9746,
      "step": 1210
    },
    {
      "epoch": 3.593519882179676,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00014261146721000553,
      "loss": 1.9891,
      "step": 1220
    },
    {
      "epoch": 3.6229749631811488,
      "grad_norm": 1.3505859375,
      "learning_rate": 0.00014177127128603745,
      "loss": 1.9369,
      "step": 1230
    },
    {
      "epoch": 3.6524300441826214,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00014092748800229683,
      "loss": 1.963,
      "step": 1240
    },
    {
      "epoch": 3.6818851251840945,
      "grad_norm": 1.123046875,
      "learning_rate": 0.00014008018982376044,
      "loss": 1.9741,
      "step": 1250
    },
    {
      "epoch": 3.711340206185567,
      "grad_norm": 1.1748046875,
      "learning_rate": 0.0001392294495172681,
      "loss": 1.9275,
      "step": 1260
    },
    {
      "epoch": 3.74079528718704,
      "grad_norm": 1.369140625,
      "learning_rate": 0.0001383753401452729,
      "loss": 1.9831,
      "step": 1270
    },
    {
      "epoch": 3.7702503681885124,
      "grad_norm": 1.2236328125,
      "learning_rate": 0.0001375179350595669,
      "loss": 1.9445,
      "step": 1280
    },
    {
      "epoch": 3.799705449189985,
      "grad_norm": 1.341796875,
      "learning_rate": 0.0001366573078949813,
      "loss": 2.014,
      "step": 1290
    },
    {
      "epoch": 3.829160530191458,
      "grad_norm": 1.1904296875,
      "learning_rate": 0.00013579353256306287,
      "loss": 2.006,
      "step": 1300
    },
    {
      "epoch": 3.858615611192931,
      "grad_norm": 1.2060546875,
      "learning_rate": 0.00013492668324572614,
      "loss": 2.007,
      "step": 1310
    },
    {
      "epoch": 3.8880706921944035,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00013405683438888282,
      "loss": 1.9583,
      "step": 1320
    },
    {
      "epoch": 3.917525773195876,
      "grad_norm": 1.216796875,
      "learning_rate": 0.00013318406069604794,
      "loss": 2.0087,
      "step": 1330
    },
    {
      "epoch": 3.9469808541973492,
      "grad_norm": 1.2822265625,
      "learning_rate": 0.00013230843712192463,
      "loss": 1.993,
      "step": 1340
    },
    {
      "epoch": 3.976435935198822,
      "grad_norm": 1.3349609375,
      "learning_rate": 0.00013143003886596669,
      "loss": 1.9554,
      "step": 1350
    },
    {
      "epoch": 4.0058910162002945,
      "grad_norm": 1.1884765625,
      "learning_rate": 0.00013054894136592052,
      "loss": 1.8819,
      "step": 1360
    },
    {
      "epoch": 4.035346097201767,
      "grad_norm": 1.3740234375,
      "learning_rate": 0.00012966522029134623,
      "loss": 1.8809,
      "step": 1370
    },
    {
      "epoch": 4.06480117820324,
      "grad_norm": 1.3447265625,
      "learning_rate": 0.00012877895153711935,
      "loss": 1.8892,
      "step": 1380
    },
    {
      "epoch": 4.0942562592047125,
      "grad_norm": 1.462890625,
      "learning_rate": 0.00012789021121691274,
      "loss": 1.892,
      "step": 1390
    },
    {
      "epoch": 4.123711340206185,
      "grad_norm": 1.5107421875,
      "learning_rate": 0.00012699907565665982,
      "loss": 1.8828,
      "step": 1400
    },
    {
      "epoch": 4.153166421207659,
      "grad_norm": 1.376953125,
      "learning_rate": 0.00012610562138799978,
      "loss": 1.9302,
      "step": 1410
    },
    {
      "epoch": 4.182621502209131,
      "grad_norm": 1.4072265625,
      "learning_rate": 0.0001252099251417048,
      "loss": 1.8978,
      "step": 1420
    },
    {
      "epoch": 4.212076583210604,
      "grad_norm": 1.3017578125,
      "learning_rate": 0.00012431206384109044,
      "loss": 1.8759,
      "step": 1430
    },
    {
      "epoch": 4.241531664212077,
      "grad_norm": 1.302734375,
      "learning_rate": 0.0001234121145954094,
      "loss": 1.83,
      "step": 1440
    },
    {
      "epoch": 4.270986745213549,
      "grad_norm": 1.2998046875,
      "learning_rate": 0.00012251015469322916,
      "loss": 1.9297,
      "step": 1450
    },
    {
      "epoch": 4.300441826215022,
      "grad_norm": 1.1630859375,
      "learning_rate": 0.00012160626159579447,
      "loss": 1.8802,
      "step": 1460
    },
    {
      "epoch": 4.329896907216495,
      "grad_norm": 1.2919921875,
      "learning_rate": 0.00012070051293037492,
      "loss": 1.8425,
      "step": 1470
    },
    {
      "epoch": 4.359351988217967,
      "grad_norm": 1.2509765625,
      "learning_rate": 0.00011979298648359823,
      "loss": 1.9012,
      "step": 1480
    },
    {
      "epoch": 4.388807069219441,
      "grad_norm": 1.1875,
      "learning_rate": 0.00011888376019476966,
      "loss": 1.8615,
      "step": 1490
    },
    {
      "epoch": 4.418262150220913,
      "grad_norm": 1.423828125,
      "learning_rate": 0.00011797291214917881,
      "loss": 1.8457,
      "step": 1500
    },
    {
      "epoch": 4.447717231222386,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00011706052057139335,
      "loss": 1.8759,
      "step": 1510
    },
    {
      "epoch": 4.477172312223859,
      "grad_norm": 1.2333984375,
      "learning_rate": 0.00011614666381854107,
      "loss": 1.8354,
      "step": 1520
    },
    {
      "epoch": 4.506627393225331,
      "grad_norm": 1.4306640625,
      "learning_rate": 0.0001152314203735805,
      "loss": 1.9068,
      "step": 1530
    },
    {
      "epoch": 4.536082474226804,
      "grad_norm": 1.1806640625,
      "learning_rate": 0.00011431486883856082,
      "loss": 1.8969,
      "step": 1540
    },
    {
      "epoch": 4.565537555228277,
      "grad_norm": 1.11328125,
      "learning_rate": 0.00011339708792787119,
      "loss": 1.8408,
      "step": 1550
    },
    {
      "epoch": 4.594992636229749,
      "grad_norm": 1.2548828125,
      "learning_rate": 0.00011247815646148087,
      "loss": 1.9036,
      "step": 1560
    },
    {
      "epoch": 4.624447717231222,
      "grad_norm": 1.2119140625,
      "learning_rate": 0.00011155815335817011,
      "loss": 1.8163,
      "step": 1570
    },
    {
      "epoch": 4.6539027982326955,
      "grad_norm": 1.2255859375,
      "learning_rate": 0.00011063715762875225,
      "loss": 1.8382,
      "step": 1580
    },
    {
      "epoch": 4.683357879234168,
      "grad_norm": 1.3173828125,
      "learning_rate": 0.0001097152483692886,
      "loss": 1.8436,
      "step": 1590
    },
    {
      "epoch": 4.712812960235641,
      "grad_norm": 1.17578125,
      "learning_rate": 0.00010879250475429523,
      "loss": 1.8737,
      "step": 1600
    },
    {
      "epoch": 4.742268041237113,
      "grad_norm": 1.134765625,
      "learning_rate": 0.00010786900602994359,
      "loss": 1.7844,
      "step": 1610
    },
    {
      "epoch": 4.771723122238586,
      "grad_norm": 1.3564453125,
      "learning_rate": 0.00010694483150725458,
      "loss": 1.8593,
      "step": 1620
    },
    {
      "epoch": 4.801178203240059,
      "grad_norm": 1.0625,
      "learning_rate": 0.0001060200605552876,
      "loss": 1.9182,
      "step": 1630
    },
    {
      "epoch": 4.830633284241531,
      "grad_norm": 1.2099609375,
      "learning_rate": 0.00010509477259432372,
      "loss": 1.8436,
      "step": 1640
    },
    {
      "epoch": 4.860088365243005,
      "grad_norm": 1.2333984375,
      "learning_rate": 0.00010416904708904548,
      "loss": 1.8418,
      "step": 1650
    },
    {
      "epoch": 4.889543446244478,
      "grad_norm": 1.240234375,
      "learning_rate": 0.00010324296354171207,
      "loss": 1.8142,
      "step": 1660
    },
    {
      "epoch": 4.91899852724595,
      "grad_norm": 1.169921875,
      "learning_rate": 0.00010231660148533183,
      "loss": 1.8852,
      "step": 1670
    },
    {
      "epoch": 4.948453608247423,
      "grad_norm": 1.142578125,
      "learning_rate": 0.00010139004047683151,
      "loss": 1.8006,
      "step": 1680
    },
    {
      "epoch": 4.9779086892488955,
      "grad_norm": 1.1513671875,
      "learning_rate": 0.00010046336009022435,
      "loss": 1.8859,
      "step": 1690
    },
    {
      "epoch": 5.007363770250368,
      "grad_norm": 1.189453125,
      "learning_rate": 9.953663990977568e-05,
      "loss": 1.8146,
      "step": 1700
    },
    {
      "epoch": 5.036818851251841,
      "grad_norm": 1.2392578125,
      "learning_rate": 9.860995952316851e-05,
      "loss": 1.7467,
      "step": 1710
    },
    {
      "epoch": 5.0662739322533135,
      "grad_norm": 1.302734375,
      "learning_rate": 9.768339851466818e-05,
      "loss": 1.8023,
      "step": 1720
    },
    {
      "epoch": 5.095729013254786,
      "grad_norm": 1.3076171875,
      "learning_rate": 9.675703645828794e-05,
      "loss": 1.746,
      "step": 1730
    },
    {
      "epoch": 5.125184094256259,
      "grad_norm": 1.2001953125,
      "learning_rate": 9.583095291095453e-05,
      "loss": 1.7308,
      "step": 1740
    },
    {
      "epoch": 5.154639175257732,
      "grad_norm": 1.3076171875,
      "learning_rate": 9.490522740567633e-05,
      "loss": 1.7452,
      "step": 1750
    },
    {
      "epoch": 5.184094256259205,
      "grad_norm": 1.201171875,
      "learning_rate": 9.397993944471244e-05,
      "loss": 1.7385,
      "step": 1760
    },
    {
      "epoch": 5.213549337260678,
      "grad_norm": 1.3212890625,
      "learning_rate": 9.305516849274541e-05,
      "loss": 1.7481,
      "step": 1770
    },
    {
      "epoch": 5.24300441826215,
      "grad_norm": 1.4599609375,
      "learning_rate": 9.213099397005646e-05,
      "loss": 1.7526,
      "step": 1780
    },
    {
      "epoch": 5.272459499263623,
      "grad_norm": 1.2783203125,
      "learning_rate": 9.12074952457048e-05,
      "loss": 1.7547,
      "step": 1790
    },
    {
      "epoch": 5.3019145802650955,
      "grad_norm": 1.203125,
      "learning_rate": 9.028475163071141e-05,
      "loss": 1.7631,
      "step": 1800
    },
    {
      "epoch": 5.331369661266568,
      "grad_norm": 1.4521484375,
      "learning_rate": 8.936284237124778e-05,
      "loss": 1.8386,
      "step": 1810
    },
    {
      "epoch": 5.360824742268041,
      "grad_norm": 1.1865234375,
      "learning_rate": 8.844184664182993e-05,
      "loss": 1.8157,
      "step": 1820
    },
    {
      "epoch": 5.390279823269514,
      "grad_norm": 1.2294921875,
      "learning_rate": 8.752184353851916e-05,
      "loss": 1.7475,
      "step": 1830
    },
    {
      "epoch": 5.419734904270987,
      "grad_norm": 1.2275390625,
      "learning_rate": 8.660291207212882e-05,
      "loss": 1.7475,
      "step": 1840
    },
    {
      "epoch": 5.44918998527246,
      "grad_norm": 1.263671875,
      "learning_rate": 8.568513116143919e-05,
      "loss": 1.7088,
      "step": 1850
    },
    {
      "epoch": 5.478645066273932,
      "grad_norm": 1.23046875,
      "learning_rate": 8.47685796264195e-05,
      "loss": 1.7893,
      "step": 1860
    },
    {
      "epoch": 5.508100147275405,
      "grad_norm": 1.427734375,
      "learning_rate": 8.385333618145896e-05,
      "loss": 1.7206,
      "step": 1870
    },
    {
      "epoch": 5.537555228276878,
      "grad_norm": 1.1630859375,
      "learning_rate": 8.293947942860666e-05,
      "loss": 1.7444,
      "step": 1880
    },
    {
      "epoch": 5.56701030927835,
      "grad_norm": 1.1826171875,
      "learning_rate": 8.202708785082121e-05,
      "loss": 1.7685,
      "step": 1890
    },
    {
      "epoch": 5.596465390279823,
      "grad_norm": 1.2373046875,
      "learning_rate": 8.111623980523035e-05,
      "loss": 1.7494,
      "step": 1900
    },
    {
      "epoch": 5.625920471281296,
      "grad_norm": 1.248046875,
      "learning_rate": 8.020701351640182e-05,
      "loss": 1.6737,
      "step": 1910
    },
    {
      "epoch": 5.655375552282769,
      "grad_norm": 1.361328125,
      "learning_rate": 7.929948706962508e-05,
      "loss": 1.7643,
      "step": 1920
    },
    {
      "epoch": 5.684830633284242,
      "grad_norm": 1.1435546875,
      "learning_rate": 7.839373840420554e-05,
      "loss": 1.7467,
      "step": 1930
    },
    {
      "epoch": 5.714285714285714,
      "grad_norm": 1.111328125,
      "learning_rate": 7.748984530677089e-05,
      "loss": 1.7421,
      "step": 1940
    },
    {
      "epoch": 5.743740795287187,
      "grad_norm": 1.3056640625,
      "learning_rate": 7.658788540459062e-05,
      "loss": 1.7451,
      "step": 1950
    },
    {
      "epoch": 5.77319587628866,
      "grad_norm": 1.18359375,
      "learning_rate": 7.568793615890954e-05,
      "loss": 1.7545,
      "step": 1960
    },
    {
      "epoch": 5.802650957290132,
      "grad_norm": 1.115234375,
      "learning_rate": 7.479007485829523e-05,
      "loss": 1.7657,
      "step": 1970
    },
    {
      "epoch": 5.832106038291605,
      "grad_norm": 1.2001953125,
      "learning_rate": 7.389437861200024e-05,
      "loss": 1.8164,
      "step": 1980
    },
    {
      "epoch": 5.8615611192930785,
      "grad_norm": 1.1943359375,
      "learning_rate": 7.30009243433402e-05,
      "loss": 1.748,
      "step": 1990
    },
    {
      "epoch": 5.891016200294551,
      "grad_norm": 1.2333984375,
      "learning_rate": 7.210978878308729e-05,
      "loss": 1.7395,
      "step": 2000
    },
    {
      "epoch": 5.920471281296024,
      "grad_norm": 1.2294921875,
      "learning_rate": 7.122104846288064e-05,
      "loss": 1.7774,
      "step": 2010
    },
    {
      "epoch": 5.9499263622974965,
      "grad_norm": 1.2802734375,
      "learning_rate": 7.033477970865381e-05,
      "loss": 1.7505,
      "step": 2020
    },
    {
      "epoch": 5.979381443298969,
      "grad_norm": 1.2568359375,
      "learning_rate": 6.945105863407951e-05,
      "loss": 1.7314,
      "step": 2030
    },
    {
      "epoch": 6.008836524300442,
      "grad_norm": 1.240234375,
      "learning_rate": 6.85699611340333e-05,
      "loss": 1.7157,
      "step": 2040
    },
    {
      "epoch": 6.0382916053019144,
      "grad_norm": 1.3017578125,
      "learning_rate": 6.76915628780754e-05,
      "loss": 1.6333,
      "step": 2050
    },
    {
      "epoch": 6.067746686303387,
      "grad_norm": 1.3681640625,
      "learning_rate": 6.681593930395209e-05,
      "loss": 1.678,
      "step": 2060
    },
    {
      "epoch": 6.09720176730486,
      "grad_norm": 1.2626953125,
      "learning_rate": 6.594316561111724e-05,
      "loss": 1.7078,
      "step": 2070
    },
    {
      "epoch": 6.126656848306332,
      "grad_norm": 1.228515625,
      "learning_rate": 6.507331675427387e-05,
      "loss": 1.7065,
      "step": 2080
    },
    {
      "epoch": 6.156111929307806,
      "grad_norm": 1.267578125,
      "learning_rate": 6.420646743693714e-05,
      "loss": 1.6533,
      "step": 2090
    },
    {
      "epoch": 6.185567010309279,
      "grad_norm": 1.2919921875,
      "learning_rate": 6.334269210501875e-05,
      "loss": 1.6894,
      "step": 2100
    },
    {
      "epoch": 6.215022091310751,
      "grad_norm": 1.2294921875,
      "learning_rate": 6.248206494043313e-05,
      "loss": 1.6709,
      "step": 2110
    },
    {
      "epoch": 6.244477172312224,
      "grad_norm": 1.2490234375,
      "learning_rate": 6.16246598547271e-05,
      "loss": 1.6944,
      "step": 2120
    },
    {
      "epoch": 6.2739322533136965,
      "grad_norm": 1.373046875,
      "learning_rate": 6.0770550482731924e-05,
      "loss": 1.6468,
      "step": 2130
    },
    {
      "epoch": 6.303387334315169,
      "grad_norm": 1.3154296875,
      "learning_rate": 5.991981017623955e-05,
      "loss": 1.6622,
      "step": 2140
    },
    {
      "epoch": 6.332842415316642,
      "grad_norm": 1.2998046875,
      "learning_rate": 5.9072511997703226e-05,
      "loss": 1.7171,
      "step": 2150
    },
    {
      "epoch": 6.3622974963181145,
      "grad_norm": 1.1591796875,
      "learning_rate": 5.8228728713962543e-05,
      "loss": 1.6501,
      "step": 2160
    },
    {
      "epoch": 6.391752577319588,
      "grad_norm": 1.2763671875,
      "learning_rate": 5.7388532789994476e-05,
      "loss": 1.5946,
      "step": 2170
    },
    {
      "epoch": 6.421207658321061,
      "grad_norm": 1.1728515625,
      "learning_rate": 5.6551996382689776e-05,
      "loss": 1.6424,
      "step": 2180
    },
    {
      "epoch": 6.450662739322533,
      "grad_norm": 1.279296875,
      "learning_rate": 5.571919133465605e-05,
      "loss": 1.6173,
      "step": 2190
    },
    {
      "epoch": 6.480117820324006,
      "grad_norm": 1.26953125,
      "learning_rate": 5.489018916804813e-05,
      "loss": 1.6357,
      "step": 2200
    },
    {
      "epoch": 6.509572901325479,
      "grad_norm": 3.232421875,
      "learning_rate": 5.4065061078425315e-05,
      "loss": 1.6616,
      "step": 2210
    },
    {
      "epoch": 6.539027982326951,
      "grad_norm": 1.2939453125,
      "learning_rate": 5.324387792863719e-05,
      "loss": 1.7015,
      "step": 2220
    },
    {
      "epoch": 6.568483063328424,
      "grad_norm": 1.271484375,
      "learning_rate": 5.242671024273798e-05,
      "loss": 1.7161,
      "step": 2230
    },
    {
      "epoch": 6.597938144329897,
      "grad_norm": 1.2939453125,
      "learning_rate": 5.1613628199929544e-05,
      "loss": 1.6494,
      "step": 2240
    },
    {
      "epoch": 6.627393225331369,
      "grad_norm": 1.21484375,
      "learning_rate": 5.080470162853472e-05,
      "loss": 1.6409,
      "step": 2250
    },
    {
      "epoch": 6.656848306332843,
      "grad_norm": 1.359375,
      "learning_rate": 5.000000000000002e-05,
      "loss": 1.648,
      "step": 2260
    },
    {
      "epoch": 6.686303387334315,
      "grad_norm": 2.451171875,
      "learning_rate": 4.919959242292954e-05,
      "loss": 1.7565,
      "step": 2270
    },
    {
      "epoch": 6.715758468335788,
      "grad_norm": 1.3623046875,
      "learning_rate": 4.840354763714991e-05,
      "loss": 1.6658,
      "step": 2280
    },
    {
      "epoch": 6.745213549337261,
      "grad_norm": 1.388671875,
      "learning_rate": 4.7611934007806666e-05,
      "loss": 1.5883,
      "step": 2290
    },
    {
      "epoch": 6.774668630338733,
      "grad_norm": 1.3837890625,
      "learning_rate": 4.6824819519493057e-05,
      "loss": 1.6502,
      "step": 2300
    },
    {
      "epoch": 6.804123711340206,
      "grad_norm": 1.333984375,
      "learning_rate": 4.604227177041156e-05,
      "loss": 1.632,
      "step": 2310
    },
    {
      "epoch": 6.833578792341679,
      "grad_norm": 1.3525390625,
      "learning_rate": 4.5264357966568306e-05,
      "loss": 1.6804,
      "step": 2320
    },
    {
      "epoch": 6.863033873343152,
      "grad_norm": 1.2587890625,
      "learning_rate": 4.4491144916001425e-05,
      "loss": 1.6897,
      "step": 2330
    },
    {
      "epoch": 6.892488954344625,
      "grad_norm": 1.2490234375,
      "learning_rate": 4.372269902304363e-05,
      "loss": 1.6649,
      "step": 2340
    },
    {
      "epoch": 6.9219440353460975,
      "grad_norm": 1.2705078125,
      "learning_rate": 4.29590862826191e-05,
      "loss": 1.7002,
      "step": 2350
    },
    {
      "epoch": 6.95139911634757,
      "grad_norm": 1.240234375,
      "learning_rate": 4.2200372274576e-05,
      "loss": 1.6725,
      "step": 2360
    },
    {
      "epoch": 6.980854197349043,
      "grad_norm": 1.26953125,
      "learning_rate": 4.144662215805426e-05,
      "loss": 1.6438,
      "step": 2370
    },
    {
      "epoch": 7.010309278350515,
      "grad_norm": 1.30859375,
      "learning_rate": 4.069790066588967e-05,
      "loss": 1.6172,
      "step": 2380
    },
    {
      "epoch": 7.039764359351988,
      "grad_norm": 1.58203125,
      "learning_rate": 3.995427209905469e-05,
      "loss": 1.6116,
      "step": 2390
    },
    {
      "epoch": 7.069219440353461,
      "grad_norm": 1.31640625,
      "learning_rate": 3.921580032113602e-05,
      "loss": 1.5673,
      "step": 2400
    },
    {
      "epoch": 7.098674521354933,
      "grad_norm": 1.275390625,
      "learning_rate": 3.848254875285e-05,
      "loss": 1.5971,
      "step": 2410
    },
    {
      "epoch": 7.128129602356406,
      "grad_norm": 1.521484375,
      "learning_rate": 3.7754580366596115e-05,
      "loss": 1.6331,
      "step": 2420
    },
    {
      "epoch": 7.15758468335788,
      "grad_norm": 1.4345703125,
      "learning_rate": 3.7031957681048604e-05,
      "loss": 1.5961,
      "step": 2430
    },
    {
      "epoch": 7.187039764359352,
      "grad_norm": 1.3564453125,
      "learning_rate": 3.631474275578754e-05,
      "loss": 1.6064,
      "step": 2440
    },
    {
      "epoch": 7.216494845360825,
      "grad_norm": 1.3017578125,
      "learning_rate": 3.560299718596889e-05,
      "loss": 1.5493,
      "step": 2450
    },
    {
      "epoch": 7.2459499263622975,
      "grad_norm": 1.2578125,
      "learning_rate": 3.489678209703475e-05,
      "loss": 1.6051,
      "step": 2460
    },
    {
      "epoch": 7.27540500736377,
      "grad_norm": 1.3056640625,
      "learning_rate": 3.4196158139463915e-05,
      "loss": 1.6425,
      "step": 2470
    },
    {
      "epoch": 7.304860088365243,
      "grad_norm": 1.423828125,
      "learning_rate": 3.3501185483562994e-05,
      "loss": 1.68,
      "step": 2480
    },
    {
      "epoch": 7.3343151693667155,
      "grad_norm": 1.1806640625,
      "learning_rate": 3.281192381429894e-05,
      "loss": 1.6,
      "step": 2490
    },
    {
      "epoch": 7.363770250368188,
      "grad_norm": 1.3212890625,
      "learning_rate": 3.212843232617343e-05,
      "loss": 1.5919,
      "step": 2500
    },
    {
      "epoch": 7.393225331369662,
      "grad_norm": 1.2099609375,
      "learning_rate": 3.145076971813891e-05,
      "loss": 1.6371,
      "step": 2510
    },
    {
      "epoch": 7.422680412371134,
      "grad_norm": 1.6533203125,
      "learning_rate": 3.077899418855772e-05,
      "loss": 1.6582,
      "step": 2520
    },
    {
      "epoch": 7.452135493372607,
      "grad_norm": 1.3759765625,
      "learning_rate": 3.0113163430203772e-05,
      "loss": 1.642,
      "step": 2530
    },
    {
      "epoch": 7.48159057437408,
      "grad_norm": 1.2509765625,
      "learning_rate": 2.945333462530788e-05,
      "loss": 1.5679,
      "step": 2540
    },
    {
      "epoch": 7.511045655375552,
      "grad_norm": 1.3740234375,
      "learning_rate": 2.879956444064703e-05,
      "loss": 1.5998,
      "step": 2550
    },
    {
      "epoch": 7.540500736377025,
      "grad_norm": 1.3916015625,
      "learning_rate": 2.815190902267757e-05,
      "loss": 1.5767,
      "step": 2560
    },
    {
      "epoch": 7.5699558173784975,
      "grad_norm": 1.3115234375,
      "learning_rate": 2.7510423992713374e-05,
      "loss": 1.597,
      "step": 2570
    },
    {
      "epoch": 7.59941089837997,
      "grad_norm": 1.3876953125,
      "learning_rate": 2.6875164442149147e-05,
      "loss": 1.559,
      "step": 2580
    },
    {
      "epoch": 7.628865979381443,
      "grad_norm": 1.330078125,
      "learning_rate": 2.624618492772891e-05,
      "loss": 1.6197,
      "step": 2590
    },
    {
      "epoch": 7.658321060382916,
      "grad_norm": 1.478515625,
      "learning_rate": 2.5623539466860813e-05,
      "loss": 1.6207,
      "step": 2600
    },
    {
      "epoch": 7.687776141384389,
      "grad_norm": 1.421875,
      "learning_rate": 2.500728153297788e-05,
      "loss": 1.6783,
      "step": 2610
    },
    {
      "epoch": 7.717231222385862,
      "grad_norm": 1.3701171875,
      "learning_rate": 2.439746405094575e-05,
      "loss": 1.6265,
      "step": 2620
    },
    {
      "epoch": 7.746686303387334,
      "grad_norm": 1.3115234375,
      "learning_rate": 2.379413939251751e-05,
      "loss": 1.6028,
      "step": 2630
    },
    {
      "epoch": 7.776141384388807,
      "grad_norm": 1.70703125,
      "learning_rate": 2.3197359371835802e-05,
      "loss": 1.6263,
      "step": 2640
    },
    {
      "epoch": 7.80559646539028,
      "grad_norm": 1.318359375,
      "learning_rate": 2.2607175240983026e-05,
      "loss": 1.6417,
      "step": 2650
    },
    {
      "epoch": 7.835051546391752,
      "grad_norm": 1.3759765625,
      "learning_rate": 2.2023637685579856e-05,
      "loss": 1.5317,
      "step": 2660
    },
    {
      "epoch": 7.864506627393226,
      "grad_norm": 1.453125,
      "learning_rate": 2.1446796820432167e-05,
      "loss": 1.5853,
      "step": 2670
    },
    {
      "epoch": 7.8939617083946985,
      "grad_norm": 1.3369140625,
      "learning_rate": 2.0876702185227137e-05,
      "loss": 1.5672,
      "step": 2680
    },
    {
      "epoch": 7.923416789396171,
      "grad_norm": 1.3466796875,
      "learning_rate": 2.0313402740278908e-05,
      "loss": 1.5869,
      "step": 2690
    },
    {
      "epoch": 7.952871870397644,
      "grad_norm": 1.3984375,
      "learning_rate": 1.9756946862323535e-05,
      "loss": 1.5456,
      "step": 2700
    },
    {
      "epoch": 7.982326951399116,
      "grad_norm": 1.310546875,
      "learning_rate": 1.9207382340364634e-05,
      "loss": 1.5585,
      "step": 2710
    },
    {
      "epoch": 8.011782032400589,
      "grad_norm": 1.47265625,
      "learning_rate": 1.866475637156898e-05,
      "loss": 1.6146,
      "step": 2720
    },
    {
      "epoch": 8.041237113402062,
      "grad_norm": 1.3291015625,
      "learning_rate": 1.8129115557213262e-05,
      "loss": 1.5909,
      "step": 2730
    },
    {
      "epoch": 8.070692194403534,
      "grad_norm": 1.4443359375,
      "learning_rate": 1.7600505898681997e-05,
      "loss": 1.5275,
      "step": 2740
    },
    {
      "epoch": 8.100147275405007,
      "grad_norm": 1.3486328125,
      "learning_rate": 1.707897279351671e-05,
      "loss": 1.4641,
      "step": 2750
    },
    {
      "epoch": 8.12960235640648,
      "grad_norm": 1.3115234375,
      "learning_rate": 1.656456103151728e-05,
      "loss": 1.61,
      "step": 2760
    },
    {
      "epoch": 8.159057437407952,
      "grad_norm": 1.3779296875,
      "learning_rate": 1.605731479089534e-05,
      "loss": 1.5912,
      "step": 2770
    },
    {
      "epoch": 8.188512518409425,
      "grad_norm": 1.3515625,
      "learning_rate": 1.5557277634480083e-05,
      "loss": 1.5664,
      "step": 2780
    },
    {
      "epoch": 8.217967599410898,
      "grad_norm": 1.458984375,
      "learning_rate": 1.5064492505977234e-05,
      "loss": 1.5658,
      "step": 2790
    },
    {
      "epoch": 8.24742268041237,
      "grad_norm": 1.3330078125,
      "learning_rate": 1.4579001726280828e-05,
      "loss": 1.6019,
      "step": 2800
    },
    {
      "epoch": 8.276877761413845,
      "grad_norm": 1.4013671875,
      "learning_rate": 1.41008469898387e-05,
      "loss": 1.6153,
      "step": 2810
    },
    {
      "epoch": 8.306332842415317,
      "grad_norm": 1.6611328125,
      "learning_rate": 1.363006936107183e-05,
      "loss": 1.6229,
      "step": 2820
    },
    {
      "epoch": 8.33578792341679,
      "grad_norm": 1.3984375,
      "learning_rate": 1.3166709270847511e-05,
      "loss": 1.5794,
      "step": 2830
    },
    {
      "epoch": 8.365243004418263,
      "grad_norm": 1.4560546875,
      "learning_rate": 1.271080651300719e-05,
      "loss": 1.5704,
      "step": 2840
    },
    {
      "epoch": 8.394698085419735,
      "grad_norm": 1.3349609375,
      "learning_rate": 1.2262400240949023e-05,
      "loss": 1.5124,
      "step": 2850
    },
    {
      "epoch": 8.424153166421208,
      "grad_norm": 1.5546875,
      "learning_rate": 1.182152896426515e-05,
      "loss": 1.5907,
      "step": 2860
    },
    {
      "epoch": 8.45360824742268,
      "grad_norm": 1.2841796875,
      "learning_rate": 1.1388230545434653e-05,
      "loss": 1.5517,
      "step": 2870
    },
    {
      "epoch": 8.483063328424153,
      "grad_norm": 1.3115234375,
      "learning_rate": 1.0962542196571634e-05,
      "loss": 1.567,
      "step": 2880
    },
    {
      "epoch": 8.512518409425626,
      "grad_norm": 1.41796875,
      "learning_rate": 1.0544500476229713e-05,
      "loss": 1.4715,
      "step": 2890
    },
    {
      "epoch": 8.541973490427099,
      "grad_norm": 1.5263671875,
      "learning_rate": 1.013414128626211e-05,
      "loss": 1.5595,
      "step": 2900
    },
    {
      "epoch": 8.571428571428571,
      "grad_norm": 1.44140625,
      "learning_rate": 9.731499868738447e-06,
      "loss": 1.5258,
      "step": 2910
    },
    {
      "epoch": 8.600883652430044,
      "grad_norm": 1.2353515625,
      "learning_rate": 9.336610802918044e-06,
      "loss": 1.5395,
      "step": 2920
    },
    {
      "epoch": 8.630338733431516,
      "grad_norm": 1.3125,
      "learning_rate": 8.949508002280382e-06,
      "loss": 1.571,
      "step": 2930
    },
    {
      "epoch": 8.65979381443299,
      "grad_norm": 1.3330078125,
      "learning_rate": 8.570224711612385e-06,
      "loss": 1.6215,
      "step": 2940
    },
    {
      "epoch": 8.689248895434462,
      "grad_norm": 1.404296875,
      "learning_rate": 8.19879350415349e-06,
      "loss": 1.5996,
      "step": 2950
    },
    {
      "epoch": 8.718703976435934,
      "grad_norm": 1.5576171875,
      "learning_rate": 7.835246278798037e-06,
      "loss": 1.5238,
      "step": 2960
    },
    {
      "epoch": 8.748159057437409,
      "grad_norm": 1.501953125,
      "learning_rate": 7.479614257355971e-06,
      "loss": 1.593,
      "step": 2970
    },
    {
      "epoch": 8.777614138438881,
      "grad_norm": 1.5166015625,
      "learning_rate": 7.1319279818713445e-06,
      "loss": 1.5612,
      "step": 2980
    },
    {
      "epoch": 8.807069219440354,
      "grad_norm": 1.7763671875,
      "learning_rate": 6.7922173119993606e-06,
      "loss": 1.5534,
      "step": 2990
    },
    {
      "epoch": 8.836524300441827,
      "grad_norm": 1.388671875,
      "learning_rate": 6.460511422441984e-06,
      "loss": 1.5691,
      "step": 3000
    },
    {
      "epoch": 8.8659793814433,
      "grad_norm": 1.412109375,
      "learning_rate": 6.136838800442457e-06,
      "loss": 1.6044,
      "step": 3010
    },
    {
      "epoch": 8.895434462444772,
      "grad_norm": 1.36328125,
      "learning_rate": 5.821227243338712e-06,
      "loss": 1.6178,
      "step": 3020
    },
    {
      "epoch": 8.924889543446245,
      "grad_norm": 1.41796875,
      "learning_rate": 5.5137038561761115e-06,
      "loss": 1.6223,
      "step": 3030
    },
    {
      "epoch": 8.954344624447717,
      "grad_norm": 1.44140625,
      "learning_rate": 5.214295049379658e-06,
      "loss": 1.5837,
      "step": 3040
    },
    {
      "epoch": 8.98379970544919,
      "grad_norm": 1.3671875,
      "learning_rate": 4.923026536485875e-06,
      "loss": 1.523,
      "step": 3050
    },
    {
      "epoch": 9.013254786450663,
      "grad_norm": 1.33203125,
      "learning_rate": 4.639923331934471e-06,
      "loss": 1.5245,
      "step": 3060
    },
    {
      "epoch": 9.042709867452135,
      "grad_norm": 1.40625,
      "learning_rate": 4.365009748920012e-06,
      "loss": 1.5816,
      "step": 3070
    },
    {
      "epoch": 9.072164948453608,
      "grad_norm": 1.255859375,
      "learning_rate": 4.098309397303978e-06,
      "loss": 1.5324,
      "step": 3080
    },
    {
      "epoch": 9.10162002945508,
      "grad_norm": 1.5517578125,
      "learning_rate": 3.839845181587098e-06,
      "loss": 1.6157,
      "step": 3090
    },
    {
      "epoch": 9.131075110456553,
      "grad_norm": 1.2978515625,
      "learning_rate": 3.5896392989422377e-06,
      "loss": 1.5071,
      "step": 3100
    },
    {
      "epoch": 9.160530191458026,
      "grad_norm": 1.537109375,
      "learning_rate": 3.3477132373081254e-06,
      "loss": 1.646,
      "step": 3110
    },
    {
      "epoch": 9.189985272459499,
      "grad_norm": 1.373046875,
      "learning_rate": 3.1140877735439387e-06,
      "loss": 1.5886,
      "step": 3120
    },
    {
      "epoch": 9.219440353460971,
      "grad_norm": 1.314453125,
      "learning_rate": 2.8887829716449876e-06,
      "loss": 1.5618,
      "step": 3130
    },
    {
      "epoch": 9.248895434462444,
      "grad_norm": 1.326171875,
      "learning_rate": 2.6718181810195696e-06,
      "loss": 1.5741,
      "step": 3140
    },
    {
      "epoch": 9.278350515463918,
      "grad_norm": 1.3994140625,
      "learning_rate": 2.4632120348272003e-06,
      "loss": 1.5673,
      "step": 3150
    },
    {
      "epoch": 9.307805596465391,
      "grad_norm": 1.4306640625,
      "learning_rate": 2.2629824483784366e-06,
      "loss": 1.5312,
      "step": 3160
    },
    {
      "epoch": 9.337260677466864,
      "grad_norm": 1.4052734375,
      "learning_rate": 2.0711466175962756e-06,
      "loss": 1.5329,
      "step": 3170
    },
    {
      "epoch": 9.366715758468336,
      "grad_norm": 1.3193359375,
      "learning_rate": 1.88772101753929e-06,
      "loss": 1.5319,
      "step": 3180
    },
    {
      "epoch": 9.396170839469809,
      "grad_norm": 1.3486328125,
      "learning_rate": 1.7127214009868385e-06,
      "loss": 1.4908,
      "step": 3190
    },
    {
      "epoch": 9.425625920471282,
      "grad_norm": 1.4267578125,
      "learning_rate": 1.5461627970860814e-06,
      "loss": 1.6365,
      "step": 3200
    },
    {
      "epoch": 9.455081001472754,
      "grad_norm": 1.4658203125,
      "learning_rate": 1.3880595100613792e-06,
      "loss": 1.5504,
      "step": 3210
    },
    {
      "epoch": 9.484536082474227,
      "grad_norm": 1.685546875,
      "learning_rate": 1.2384251179857643e-06,
      "loss": 1.5451,
      "step": 3220
    },
    {
      "epoch": 9.5139911634757,
      "grad_norm": 1.4287109375,
      "learning_rate": 1.0972724716148187e-06,
      "loss": 1.5691,
      "step": 3230
    },
    {
      "epoch": 9.543446244477172,
      "grad_norm": 1.345703125,
      "learning_rate": 9.64613693283123e-07,
      "loss": 1.5342,
      "step": 3240
    },
    {
      "epoch": 9.572901325478645,
      "grad_norm": 1.3359375,
      "learning_rate": 8.404601758630892e-07,
      "loss": 1.5575,
      "step": 3250
    },
    {
      "epoch": 9.602356406480117,
      "grad_norm": 1.3681640625,
      "learning_rate": 7.248225817865884e-07,
      "loss": 1.5327,
      "step": 3260
    },
    {
      "epoch": 9.63181148748159,
      "grad_norm": 1.37890625,
      "learning_rate": 6.177108421292266e-07,
      "loss": 1.5548,
      "step": 3270
    },
    {
      "epoch": 9.661266568483063,
      "grad_norm": 1.353515625,
      "learning_rate": 5.191341557574392e-07,
      "loss": 1.6446,
      "step": 3280
    },
    {
      "epoch": 9.690721649484535,
      "grad_norm": 1.3310546875,
      "learning_rate": 4.291009885385333e-07,
      "loss": 1.4653,
      "step": 3290
    },
    {
      "epoch": 9.720176730486008,
      "grad_norm": 1.380859375,
      "learning_rate": 3.4761907261356976e-07,
      "loss": 1.5371,
      "step": 3300
    },
    {
      "epoch": 9.749631811487482,
      "grad_norm": 1.4384765625,
      "learning_rate": 2.746954057333606e-07,
      "loss": 1.5145,
      "step": 3310
    },
    {
      "epoch": 9.779086892488955,
      "grad_norm": 1.3583984375,
      "learning_rate": 2.1033625065747242e-07,
      "loss": 1.6238,
      "step": 3320
    },
    {
      "epoch": 9.808541973490428,
      "grad_norm": 1.3671875,
      "learning_rate": 1.545471346164007e-07,
      "loss": 1.5669,
      "step": 3330
    },
    {
      "epoch": 9.8379970544919,
      "grad_norm": 1.34765625,
      "learning_rate": 1.0733284883682749e-07,
      "loss": 1.5792,
      "step": 3340
    },
    {
      "epoch": 9.867452135493373,
      "grad_norm": 1.4501953125,
      "learning_rate": 6.869744813023937e-08,
      "loss": 1.6,
      "step": 3350
    },
    {
      "epoch": 9.896907216494846,
      "grad_norm": 1.443359375,
      "learning_rate": 3.8644250544594975e-08,
      "loss": 1.5093,
      "step": 3360
    },
    {
      "epoch": 9.926362297496318,
      "grad_norm": 1.51171875,
      "learning_rate": 1.7175837079452804e-08,
      "loss": 1.5322,
      "step": 3370
    },
    {
      "epoch": 9.955817378497791,
      "grad_norm": 1.3876953125,
      "learning_rate": 4.2940514642597626e-09,
      "loss": 1.5623,
      "step": 3380
    },
    {
      "epoch": 9.985272459499264,
      "grad_norm": 1.30078125,
      "learning_rate": 0.0,
      "loss": 1.5123,
      "step": 3390
    },
    {
      "epoch": 9.985272459499264,
      "step": 3390,
      "total_flos": 2.22628436508672e+17,
      "train_loss": 1.877983266273431,
      "train_runtime": 3392.4525,
      "train_samples_per_second": 4.003,
      "train_steps_per_second": 0.999
    }
  ],
  "logging_steps": 10,
  "max_steps": 3390,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 500,
  "total_flos": 2.22628436508672e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}