{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 20.837148463047743,
  "eval_steps": 500,
  "global_step": 4000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.052321778940483975,
      "grad_norm": 0.5172898769378662,
      "learning_rate": 4.9999695642048685e-05,
      "loss": 1.2268,
      "num_input_tokens_seen": 55264,
      "step": 10
    },
    {
      "epoch": 0.10464355788096795,
      "grad_norm": 0.2547127306461334,
      "learning_rate": 4.9998643550002796e-05,
      "loss": 0.163,
      "num_input_tokens_seen": 111136,
      "step": 20
    },
    {
      "epoch": 0.15696533682145192,
      "grad_norm": 0.19801495969295502,
      "learning_rate": 4.999683999797514e-05,
      "loss": 0.1584,
      "num_input_tokens_seen": 167232,
      "step": 30
    },
    {
      "epoch": 0.2092871157619359,
      "grad_norm": 0.17607145011425018,
      "learning_rate": 4.999428504018057e-05,
      "loss": 0.1571,
      "num_input_tokens_seen": 222368,
      "step": 40
    },
    {
      "epoch": 0.2616088947024199,
      "grad_norm": 0.123909592628479,
      "learning_rate": 4.999097875342117e-05,
      "loss": 0.1535,
      "num_input_tokens_seen": 280080,
      "step": 50
    },
    {
      "epoch": 0.31393067364290383,
      "grad_norm": 0.15530619025230408,
      "learning_rate": 4.998692123708403e-05,
      "loss": 0.1592,
      "num_input_tokens_seen": 336144,
      "step": 60
    },
    {
      "epoch": 0.36625245258338784,
      "grad_norm": 0.08514747023582458,
      "learning_rate": 4.998211261313822e-05,
      "loss": 0.1581,
      "num_input_tokens_seen": 392928,
      "step": 70
    },
    {
      "epoch": 0.4185742315238718,
      "grad_norm": 0.08711712062358856,
      "learning_rate": 4.997655302613111e-05,
      "loss": 0.1479,
      "num_input_tokens_seen": 448288,
      "step": 80
    },
    {
      "epoch": 0.4708960104643558,
      "grad_norm": 0.14576469361782074,
      "learning_rate": 4.997024264318406e-05,
      "loss": 0.1574,
      "num_input_tokens_seen": 504112,
      "step": 90
    },
    {
      "epoch": 0.5232177894048398,
      "grad_norm": 0.14594055712223053,
      "learning_rate": 4.9963181653987373e-05,
      "loss": 0.1532,
      "num_input_tokens_seen": 559824,
      "step": 100
    },
    {
      "epoch": 0.5755395683453237,
      "grad_norm": 1.3401381969451904,
      "learning_rate": 4.99553702707946e-05,
      "loss": 0.1527,
      "num_input_tokens_seen": 615856,
      "step": 110
    },
    {
      "epoch": 0.6278613472858077,
      "grad_norm": 0.21902425587177277,
      "learning_rate": 4.9946808728416143e-05,
      "loss": 0.1464,
      "num_input_tokens_seen": 672448,
      "step": 120
    },
    {
      "epoch": 0.6801831262262917,
      "grad_norm": 0.24769121408462524,
      "learning_rate": 4.993749728421224e-05,
      "loss": 0.1545,
      "num_input_tokens_seen": 727840,
      "step": 130
    },
    {
      "epoch": 0.7325049051667757,
      "grad_norm": 0.2848469018936157,
      "learning_rate": 4.992743621808518e-05,
      "loss": 0.1448,
      "num_input_tokens_seen": 784496,
      "step": 140
    },
    {
      "epoch": 0.7848266841072596,
      "grad_norm": 0.31721895933151245,
      "learning_rate": 4.991662583247092e-05,
      "loss": 0.1458,
      "num_input_tokens_seen": 840784,
      "step": 150
    },
    {
      "epoch": 0.8371484630477436,
      "grad_norm": 0.21474598348140717,
      "learning_rate": 4.9905066452329964e-05,
      "loss": 0.1445,
      "num_input_tokens_seen": 897920,
      "step": 160
    },
    {
      "epoch": 0.8894702419882276,
      "grad_norm": 0.5110652446746826,
      "learning_rate": 4.9892758425137643e-05,
      "loss": 0.1471,
      "num_input_tokens_seen": 953184,
      "step": 170
    },
    {
      "epoch": 0.9417920209287116,
      "grad_norm": 0.29877740144729614,
      "learning_rate": 4.987970212087363e-05,
      "loss": 0.1464,
      "num_input_tokens_seen": 1009168,
      "step": 180
    },
    {
      "epoch": 0.9941137998691956,
      "grad_norm": 0.16597020626068115,
      "learning_rate": 4.986589793201081e-05,
      "loss": 0.1407,
      "num_input_tokens_seen": 1065200,
      "step": 190
    },
    {
      "epoch": 1.0418574231523872,
      "grad_norm": 0.2527844309806824,
      "learning_rate": 4.985134627350353e-05,
      "loss": 0.1492,
      "num_input_tokens_seen": 1115736,
      "step": 200
    },
    {
      "epoch": 1.0941792020928711,
      "grad_norm": 0.19597752392292023,
      "learning_rate": 4.9836047582775084e-05,
      "loss": 0.1431,
      "num_input_tokens_seen": 1171656,
      "step": 210
    },
    {
      "epoch": 1.146500981033355,
      "grad_norm": 0.23703065514564514,
      "learning_rate": 4.9820002319704576e-05,
      "loss": 0.1434,
      "num_input_tokens_seen": 1229224,
      "step": 220
    },
    {
      "epoch": 1.198822759973839,
      "grad_norm": 0.19536039233207703,
      "learning_rate": 4.98032109666131e-05,
      "loss": 0.14,
      "num_input_tokens_seen": 1285992,
      "step": 230
    },
    {
      "epoch": 1.251144538914323,
      "grad_norm": 0.3111051321029663,
      "learning_rate": 4.978567402824924e-05,
      "loss": 0.1355,
      "num_input_tokens_seen": 1341256,
      "step": 240
    },
    {
      "epoch": 1.3034663178548072,
      "grad_norm": 0.19106437265872955,
      "learning_rate": 4.97673920317739e-05,
      "loss": 0.1425,
      "num_input_tokens_seen": 1396392,
      "step": 250
    },
    {
      "epoch": 1.355788096795291,
      "grad_norm": 0.30303388833999634,
      "learning_rate": 4.9748365526744423e-05,
      "loss": 0.1343,
      "num_input_tokens_seen": 1454088,
      "step": 260
    },
    {
      "epoch": 1.408109875735775,
      "grad_norm": 0.3765646815299988,
      "learning_rate": 4.972859508509816e-05,
      "loss": 0.1343,
      "num_input_tokens_seen": 1509656,
      "step": 270
    },
    {
      "epoch": 1.460431654676259,
      "grad_norm": 0.3608955442905426,
      "learning_rate": 4.9708081301135155e-05,
      "loss": 0.1377,
      "num_input_tokens_seen": 1565976,
      "step": 280
    },
    {
      "epoch": 1.512753433616743,
      "grad_norm": 0.19421738386154175,
      "learning_rate": 4.9686824791500396e-05,
      "loss": 0.1381,
      "num_input_tokens_seen": 1621544,
      "step": 290
    },
    {
      "epoch": 1.565075212557227,
      "grad_norm": 0.6034450531005859,
      "learning_rate": 4.96648261951652e-05,
      "loss": 0.1339,
      "num_input_tokens_seen": 1677624,
      "step": 300
    },
    {
      "epoch": 1.6173969914977109,
      "grad_norm": 0.28776681423187256,
      "learning_rate": 4.964208617340803e-05,
      "loss": 0.1351,
      "num_input_tokens_seen": 1733192,
      "step": 310
    },
    {
      "epoch": 1.669718770438195,
      "grad_norm": 0.15169182419776917,
      "learning_rate": 4.961860540979464e-05,
      "loss": 0.1406,
      "num_input_tokens_seen": 1788648,
      "step": 320
    },
    {
      "epoch": 1.7220405493786788,
      "grad_norm": 0.24536730349063873,
      "learning_rate": 4.9594384610157483e-05,
      "loss": 0.1354,
      "num_input_tokens_seen": 1844360,
      "step": 330
    },
    {
      "epoch": 1.774362328319163,
      "grad_norm": 0.4901825487613678,
      "learning_rate": 4.9569424502574544e-05,
      "loss": 0.1318,
      "num_input_tokens_seen": 1900888,
      "step": 340
    },
    {
      "epoch": 1.8266841072596467,
      "grad_norm": 0.33420756459236145,
      "learning_rate": 4.954372583734741e-05,
      "loss": 0.1324,
      "num_input_tokens_seen": 1956840,
      "step": 350
    },
    {
      "epoch": 1.8790058862001309,
      "grad_norm": 0.24992604553699493,
      "learning_rate": 4.951728938697872e-05,
      "loss": 0.1352,
      "num_input_tokens_seen": 2013624,
      "step": 360
    },
    {
      "epoch": 1.9313276651406148,
      "grad_norm": 0.3095349669456482,
      "learning_rate": 4.9490115946148985e-05,
      "loss": 0.1332,
      "num_input_tokens_seen": 2069768,
      "step": 370
    },
    {
      "epoch": 1.9836494440810988,
      "grad_norm": 0.35465294122695923,
      "learning_rate": 4.946220633169266e-05,
      "loss": 0.128,
      "num_input_tokens_seen": 2125768,
      "step": 380
    },
    {
      "epoch": 2.0313930673642906,
      "grad_norm": 0.2656111419200897,
      "learning_rate": 4.943356138257359e-05,
      "loss": 0.1356,
      "num_input_tokens_seen": 2176424,
      "step": 390
    },
    {
      "epoch": 2.0837148463047743,
      "grad_norm": 0.16828453540802002,
      "learning_rate": 4.940418195985983e-05,
      "loss": 0.1332,
      "num_input_tokens_seen": 2232200,
      "step": 400
    },
    {
      "epoch": 2.1360366252452585,
      "grad_norm": 0.22813495993614197,
      "learning_rate": 4.9374068946697695e-05,
      "loss": 0.1296,
      "num_input_tokens_seen": 2289432,
      "step": 410
    },
    {
      "epoch": 2.1883584041857422,
      "grad_norm": 0.30772924423217773,
      "learning_rate": 4.934322324828529e-05,
      "loss": 0.1288,
      "num_input_tokens_seen": 2345576,
      "step": 420
    },
    {
      "epoch": 2.2406801831262264,
      "grad_norm": 0.29362913966178894,
      "learning_rate": 4.931164579184523e-05,
      "loss": 0.1307,
      "num_input_tokens_seen": 2400696,
      "step": 430
    },
    {
      "epoch": 2.29300196206671,
      "grad_norm": 0.21813659369945526,
      "learning_rate": 4.9279337526596814e-05,
      "loss": 0.1292,
      "num_input_tokens_seen": 2456552,
      "step": 440
    },
    {
      "epoch": 2.3453237410071943,
      "grad_norm": 0.3129786550998688,
      "learning_rate": 4.924629942372748e-05,
      "loss": 0.1329,
      "num_input_tokens_seen": 2512808,
      "step": 450
    },
    {
      "epoch": 2.397645519947678,
      "grad_norm": 0.19762465357780457,
      "learning_rate": 4.9212532476363596e-05,
      "loss": 0.1261,
      "num_input_tokens_seen": 2569016,
      "step": 460
    },
    {
      "epoch": 2.4499672988881622,
      "grad_norm": 0.30003225803375244,
      "learning_rate": 4.917803769954062e-05,
      "loss": 0.124,
      "num_input_tokens_seen": 2625208,
      "step": 470
    },
    {
      "epoch": 2.502289077828646,
      "grad_norm": 0.2731872797012329,
      "learning_rate": 4.9142816130172596e-05,
      "loss": 0.1285,
      "num_input_tokens_seen": 2680824,
      "step": 480
    },
    {
      "epoch": 2.55461085676913,
      "grad_norm": 0.25849199295043945,
      "learning_rate": 4.9106868827020955e-05,
      "loss": 0.1323,
      "num_input_tokens_seen": 2737304,
      "step": 490
    },
    {
      "epoch": 2.6069326357096143,
      "grad_norm": 0.24799242615699768,
      "learning_rate": 4.907019687066271e-05,
      "loss": 0.1299,
      "num_input_tokens_seen": 2793432,
      "step": 500
    },
    {
      "epoch": 2.659254414650098,
      "grad_norm": 0.1981252282857895,
      "learning_rate": 4.9032801363458e-05,
      "loss": 0.1281,
      "num_input_tokens_seen": 2850008,
      "step": 510
    },
    {
      "epoch": 2.711576193590582,
      "grad_norm": 0.40522608160972595,
      "learning_rate": 4.8994683429516896e-05,
      "loss": 0.1309,
      "num_input_tokens_seen": 2905304,
      "step": 520
    },
    {
      "epoch": 2.763897972531066,
      "grad_norm": 0.2624233067035675,
      "learning_rate": 4.895584421466565e-05,
      "loss": 0.1271,
      "num_input_tokens_seen": 2961112,
      "step": 530
    },
    {
      "epoch": 2.81621975147155,
      "grad_norm": 0.28190159797668457,
      "learning_rate": 4.8916284886412214e-05,
      "loss": 0.1222,
      "num_input_tokens_seen": 3017208,
      "step": 540
    },
    {
      "epoch": 2.868541530412034,
      "grad_norm": 0.21756145358085632,
      "learning_rate": 4.887600663391122e-05,
      "loss": 0.1288,
      "num_input_tokens_seen": 3074216,
      "step": 550
    },
    {
      "epoch": 2.920863309352518,
      "grad_norm": 0.33730047941207886,
      "learning_rate": 4.883501066792814e-05,
      "loss": 0.1267,
      "num_input_tokens_seen": 3129784,
      "step": 560
    },
    {
      "epoch": 2.973185088293002,
      "grad_norm": 0.24758093059062958,
      "learning_rate": 4.8793298220802963e-05,
      "loss": 0.1288,
      "num_input_tokens_seen": 3187048,
      "step": 570
    },
    {
      "epoch": 3.0209287115761936,
      "grad_norm": 0.15806905925273895,
      "learning_rate": 4.87508705464131e-05,
      "loss": 0.1316,
      "num_input_tokens_seen": 3238608,
      "step": 580
    },
    {
      "epoch": 3.0732504905166778,
      "grad_norm": 0.28546327352523804,
      "learning_rate": 4.8707728920135744e-05,
      "loss": 0.1256,
      "num_input_tokens_seen": 3294400,
      "step": 590
    },
    {
      "epoch": 3.1255722694571615,
      "grad_norm": 0.3583323359489441,
      "learning_rate": 4.866387463880947e-05,
      "loss": 0.1223,
      "num_input_tokens_seen": 3350560,
      "step": 600
    },
    {
      "epoch": 3.1778940483976457,
      "grad_norm": 0.21477927267551422,
      "learning_rate": 4.861930902069531e-05,
      "loss": 0.1192,
      "num_input_tokens_seen": 3406256,
      "step": 610
    },
    {
      "epoch": 3.2302158273381294,
      "grad_norm": 0.3112389147281647,
      "learning_rate": 4.8574033405437094e-05,
      "loss": 0.1209,
      "num_input_tokens_seen": 3461680,
      "step": 620
    },
    {
      "epoch": 3.2825376062786136,
      "grad_norm": 0.22634848952293396,
      "learning_rate": 4.8528049154021186e-05,
      "loss": 0.1318,
      "num_input_tokens_seen": 3517984,
      "step": 630
    },
    {
      "epoch": 3.3348593852190973,
      "grad_norm": 0.2349083423614502,
      "learning_rate": 4.848135764873557e-05,
      "loss": 0.1264,
      "num_input_tokens_seen": 3573376,
      "step": 640
    },
    {
      "epoch": 3.3871811641595815,
      "grad_norm": 0.29488080739974976,
      "learning_rate": 4.843396029312832e-05,
      "loss": 0.1238,
      "num_input_tokens_seen": 3630544,
      "step": 650
    },
    {
      "epoch": 3.439502943100065,
      "grad_norm": 0.28836989402770996,
      "learning_rate": 4.838585851196537e-05,
      "loss": 0.124,
      "num_input_tokens_seen": 3686432,
      "step": 660
    },
    {
      "epoch": 3.4918247220405494,
      "grad_norm": 0.2634967267513275,
      "learning_rate": 4.833705375118772e-05,
      "loss": 0.1212,
      "num_input_tokens_seen": 3741552,
      "step": 670
    },
    {
      "epoch": 3.544146500981033,
      "grad_norm": 0.30187705159187317,
      "learning_rate": 4.828754747786796e-05,
      "loss": 0.1225,
      "num_input_tokens_seen": 3797760,
      "step": 680
    },
    {
      "epoch": 3.5964682799215173,
      "grad_norm": 0.31589755415916443,
      "learning_rate": 4.823734118016616e-05,
      "loss": 0.1236,
      "num_input_tokens_seen": 3854704,
      "step": 690
    },
    {
      "epoch": 3.6487900588620015,
      "grad_norm": 0.24722936749458313,
      "learning_rate": 4.818643636728515e-05,
      "loss": 0.1154,
      "num_input_tokens_seen": 3910544,
      "step": 700
    },
    {
      "epoch": 3.701111837802485,
      "grad_norm": 0.20096950232982635,
      "learning_rate": 4.813483456942515e-05,
      "loss": 0.119,
      "num_input_tokens_seen": 3966448,
      "step": 710
    },
    {
      "epoch": 3.7534336167429694,
      "grad_norm": 0.22696824371814728,
      "learning_rate": 4.808253733773775e-05,
      "loss": 0.1277,
      "num_input_tokens_seen": 4022880,
      "step": 720
    },
    {
      "epoch": 3.805755395683453,
      "grad_norm": 0.2600644826889038,
      "learning_rate": 4.8029546244279346e-05,
      "loss": 0.1245,
      "num_input_tokens_seen": 4079280,
      "step": 730
    },
    {
      "epoch": 3.8580771746239373,
      "grad_norm": 0.3461725115776062,
      "learning_rate": 4.797586288196378e-05,
      "loss": 0.1245,
      "num_input_tokens_seen": 4136096,
      "step": 740
    },
    {
      "epoch": 3.910398953564421,
      "grad_norm": 0.20807071030139923,
      "learning_rate": 4.792148886451456e-05,
      "loss": 0.1238,
      "num_input_tokens_seen": 4192832,
      "step": 750
    },
    {
      "epoch": 3.962720732504905,
      "grad_norm": 0.25355812907218933,
      "learning_rate": 4.7866425826416316e-05,
      "loss": 0.1249,
      "num_input_tokens_seen": 4248368,
      "step": 760
    },
    {
      "epoch": 4.010464355788097,
      "grad_norm": 0.28272193670272827,
      "learning_rate": 4.781067542286561e-05,
      "loss": 0.1245,
      "num_input_tokens_seen": 4299232,
      "step": 770
    },
    {
      "epoch": 4.062786134728581,
      "grad_norm": 0.39174169301986694,
      "learning_rate": 4.7754239329721274e-05,
      "loss": 0.1216,
      "num_input_tokens_seen": 4356192,
      "step": 780
    },
    {
      "epoch": 4.115107913669065,
      "grad_norm": 0.2807229459285736,
      "learning_rate": 4.769711924345397e-05,
      "loss": 0.1195,
      "num_input_tokens_seen": 4411648,
      "step": 790
    },
    {
      "epoch": 4.167429692609549,
      "grad_norm": 0.24898973107337952,
      "learning_rate": 4.763931688109524e-05,
      "loss": 0.1174,
      "num_input_tokens_seen": 4467568,
      "step": 800
    },
    {
      "epoch": 4.219751471550032,
      "grad_norm": 0.3005022704601288,
      "learning_rate": 4.7580833980185816e-05,
      "loss": 0.1251,
      "num_input_tokens_seen": 4522624,
      "step": 810
    },
    {
      "epoch": 4.272073250490517,
      "grad_norm": 0.3487663269042969,
      "learning_rate": 4.7521672298723495e-05,
      "loss": 0.1182,
      "num_input_tokens_seen": 4578640,
      "step": 820
    },
    {
      "epoch": 4.324395029431001,
      "grad_norm": 0.19975335896015167,
      "learning_rate": 4.7461833615110194e-05,
      "loss": 0.1211,
      "num_input_tokens_seen": 4633712,
      "step": 830
    },
    {
      "epoch": 4.3767168083714845,
      "grad_norm": 0.2742094397544861,
      "learning_rate": 4.740131972809856e-05,
      "loss": 0.1208,
      "num_input_tokens_seen": 4690160,
      "step": 840
    },
    {
      "epoch": 4.429038587311968,
      "grad_norm": 0.49232161045074463,
      "learning_rate": 4.734013245673788e-05,
      "loss": 0.1213,
      "num_input_tokens_seen": 4747104,
      "step": 850
    },
    {
      "epoch": 4.481360366252453,
      "grad_norm": 0.3197735548019409,
      "learning_rate": 4.727827364031936e-05,
      "loss": 0.1137,
      "num_input_tokens_seen": 4803376,
      "step": 860
    },
    {
      "epoch": 4.5336821451929366,
      "grad_norm": 0.43819674849510193,
      "learning_rate": 4.721574513832091e-05,
      "loss": 0.1163,
      "num_input_tokens_seen": 4859840,
      "step": 870
    },
    {
      "epoch": 4.58600392413342,
      "grad_norm": 0.21390356123447418,
      "learning_rate": 4.715254883035119e-05,
      "loss": 0.121,
      "num_input_tokens_seen": 4916272,
      "step": 880
    },
    {
      "epoch": 4.638325703073905,
      "grad_norm": 0.30768075585365295,
      "learning_rate": 4.708868661609314e-05,
      "loss": 0.1194,
      "num_input_tokens_seen": 4971728,
      "step": 890
    },
    {
      "epoch": 4.690647482014389,
      "grad_norm": 0.34879791736602783,
      "learning_rate": 4.702416041524683e-05,
      "loss": 0.1223,
      "num_input_tokens_seen": 5027680,
      "step": 900
    },
    {
      "epoch": 4.742969260954872,
      "grad_norm": 0.23973870277404785,
      "learning_rate": 4.695897216747183e-05,
      "loss": 0.1225,
      "num_input_tokens_seen": 5083392,
      "step": 910
    },
    {
      "epoch": 4.795291039895356,
      "grad_norm": 0.33110499382019043,
      "learning_rate": 4.689312383232883e-05,
      "loss": 0.1248,
      "num_input_tokens_seen": 5140112,
      "step": 920
    },
    {
      "epoch": 4.847612818835841,
      "grad_norm": 0.2787615954875946,
      "learning_rate": 4.682661738922078e-05,
      "loss": 0.1204,
      "num_input_tokens_seen": 5195072,
      "step": 930
    },
    {
      "epoch": 4.8999345977763245,
      "grad_norm": 0.20854991674423218,
      "learning_rate": 4.6759454837333376e-05,
      "loss": 0.1181,
      "num_input_tokens_seen": 5251408,
      "step": 940
    },
    {
      "epoch": 4.952256376716808,
      "grad_norm": 0.3582795262336731,
      "learning_rate": 4.6691638195574963e-05,
      "loss": 0.118,
      "num_input_tokens_seen": 5307776,
      "step": 950
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.18520788848400116,
      "learning_rate": 4.662316950251584e-05,
      "loss": 0.1192,
      "num_input_tokens_seen": 5359280,
      "step": 960
    },
    {
      "epoch": 5.052321778940484,
      "grad_norm": 0.2574616074562073,
      "learning_rate": 4.655405081632699e-05,
      "loss": 0.1179,
      "num_input_tokens_seen": 5415968,
      "step": 970
    },
    {
      "epoch": 5.104643557880968,
      "grad_norm": 0.3383731544017792,
      "learning_rate": 4.648428421471822e-05,
      "loss": 0.1137,
      "num_input_tokens_seen": 5471840,
      "step": 980
    },
    {
      "epoch": 5.156965336821452,
      "grad_norm": 0.3150075376033783,
      "learning_rate": 4.641387179487569e-05,
      "loss": 0.1179,
      "num_input_tokens_seen": 5528224,
      "step": 990
    },
    {
      "epoch": 5.209287115761936,
      "grad_norm": 0.2862718999385834,
      "learning_rate": 4.634281567339885e-05,
      "loss": 0.1117,
      "num_input_tokens_seen": 5583680,
      "step": 1000
    },
    {
      "epoch": 5.2616088947024195,
      "grad_norm": 0.32703572511672974,
      "learning_rate": 4.627111798623688e-05,
      "loss": 0.12,
      "num_input_tokens_seen": 5638640,
      "step": 1010
    },
    {
      "epoch": 5.313930673642904,
      "grad_norm": 0.3088448941707611,
      "learning_rate": 4.619878088862443e-05,
      "loss": 0.1134,
      "num_input_tokens_seen": 5694208,
      "step": 1020
    },
    {
      "epoch": 5.366252452583388,
      "grad_norm": 0.3116937279701233,
      "learning_rate": 4.612580655501683e-05,
      "loss": 0.1178,
      "num_input_tokens_seen": 5749696,
      "step": 1030
    },
    {
      "epoch": 5.418574231523872,
      "grad_norm": 0.3191389739513397,
      "learning_rate": 4.605219717902476e-05,
      "loss": 0.1136,
      "num_input_tokens_seen": 5805264,
      "step": 1040
    },
    {
      "epoch": 5.470896010464356,
      "grad_norm": 0.31546375155448914,
      "learning_rate": 4.5977954973348294e-05,
      "loss": 0.1167,
      "num_input_tokens_seen": 5861616,
      "step": 1050
    },
    {
      "epoch": 5.52321778940484,
      "grad_norm": 0.2683027386665344,
      "learning_rate": 4.590308216971038e-05,
      "loss": 0.1113,
      "num_input_tokens_seen": 5918816,
      "step": 1060
    },
    {
      "epoch": 5.575539568345324,
      "grad_norm": 0.26271599531173706,
      "learning_rate": 4.582758101878977e-05,
      "loss": 0.1113,
      "num_input_tokens_seen": 5975184,
      "step": 1070
    },
    {
      "epoch": 5.6278613472858074,
      "grad_norm": 0.2612985372543335,
      "learning_rate": 4.5751453790153325e-05,
      "loss": 0.1143,
      "num_input_tokens_seen": 6030736,
      "step": 1080
    },
    {
      "epoch": 5.680183126226292,
      "grad_norm": 0.29255977272987366,
      "learning_rate": 4.567470277218786e-05,
      "loss": 0.1159,
      "num_input_tokens_seen": 6086848,
      "step": 1090
    },
    {
      "epoch": 5.732504905166776,
      "grad_norm": 0.31331729888916016,
      "learning_rate": 4.55973302720313e-05,
      "loss": 0.1114,
      "num_input_tokens_seen": 6142544,
      "step": 1100
    },
    {
      "epoch": 5.7848266841072595,
      "grad_norm": 0.24986866116523743,
      "learning_rate": 4.551933861550333e-05,
      "loss": 0.1173,
      "num_input_tokens_seen": 6199856,
      "step": 1110
    },
    {
      "epoch": 5.837148463047743,
      "grad_norm": 0.26383090019226074,
      "learning_rate": 4.5440730147035516e-05,
      "loss": 0.1166,
      "num_input_tokens_seen": 6255488,
      "step": 1120
    },
    {
      "epoch": 5.889470241988228,
      "grad_norm": 0.28010377287864685,
      "learning_rate": 4.5361507229600784e-05,
      "loss": 0.1148,
      "num_input_tokens_seen": 6311696,
      "step": 1130
    },
    {
      "epoch": 5.941792020928712,
      "grad_norm": 0.2781098783016205,
      "learning_rate": 4.528167224464245e-05,
      "loss": 0.1152,
      "num_input_tokens_seen": 6368064,
      "step": 1140
    },
    {
      "epoch": 5.994113799869195,
      "grad_norm": 0.35192057490348816,
      "learning_rate": 4.520122759200256e-05,
      "loss": 0.1087,
      "num_input_tokens_seen": 6424048,
      "step": 1150
    },
    {
      "epoch": 6.041857423152387,
      "grad_norm": 0.408779114484787,
      "learning_rate": 4.512017568984982e-05,
      "loss": 0.1094,
      "num_input_tokens_seen": 6475464,
      "step": 1160
    },
    {
      "epoch": 6.094179202092871,
      "grad_norm": 0.5394303798675537,
      "learning_rate": 4.503851897460686e-05,
      "loss": 0.1005,
      "num_input_tokens_seen": 6531112,
      "step": 1170
    },
    {
      "epoch": 6.1465009810333555,
      "grad_norm": 0.4771885871887207,
      "learning_rate": 4.4956259900877005e-05,
      "loss": 0.107,
      "num_input_tokens_seen": 6587352,
      "step": 1180
    },
    {
      "epoch": 6.198822759973839,
      "grad_norm": 0.4851493835449219,
      "learning_rate": 4.4873400941370506e-05,
      "loss": 0.1093,
      "num_input_tokens_seen": 6643608,
      "step": 1190
    },
    {
      "epoch": 6.251144538914323,
      "grad_norm": 0.47404423356056213,
      "learning_rate": 4.4789944586830196e-05,
      "loss": 0.1082,
      "num_input_tokens_seen": 6700616,
      "step": 1200
    },
    {
      "epoch": 6.303466317854807,
      "grad_norm": 0.4265010356903076,
      "learning_rate": 4.470589334595662e-05,
      "loss": 0.1088,
      "num_input_tokens_seen": 6756344,
      "step": 1210
    },
    {
      "epoch": 6.355788096795291,
      "grad_norm": 0.2776981592178345,
      "learning_rate": 4.462124974533261e-05,
      "loss": 0.1124,
      "num_input_tokens_seen": 6813144,
      "step": 1220
    },
    {
      "epoch": 6.408109875735775,
      "grad_norm": 0.3706200122833252,
      "learning_rate": 4.453601632934737e-05,
      "loss": 0.1095,
      "num_input_tokens_seen": 6868280,
      "step": 1230
    },
    {
      "epoch": 6.460431654676259,
      "grad_norm": 0.4983665943145752,
      "learning_rate": 4.4450195660119965e-05,
      "loss": 0.1114,
      "num_input_tokens_seen": 6924296,
      "step": 1240
    },
    {
      "epoch": 6.5127534336167425,
      "grad_norm": 0.38131120800971985,
      "learning_rate": 4.4363790317422314e-05,
      "loss": 0.1141,
      "num_input_tokens_seen": 6980392,
      "step": 1250
    },
    {
      "epoch": 6.565075212557227,
      "grad_norm": 0.24885649979114532,
      "learning_rate": 4.427680289860163e-05,
      "loss": 0.1128,
      "num_input_tokens_seen": 7036056,
      "step": 1260
    },
    {
      "epoch": 6.617396991497711,
      "grad_norm": 0.4561365246772766,
      "learning_rate": 4.4189236018502356e-05,
      "loss": 0.1149,
      "num_input_tokens_seen": 7092152,
      "step": 1270
    },
    {
      "epoch": 6.669718770438195,
      "grad_norm": 0.548459529876709,
      "learning_rate": 4.410109230938755e-05,
      "loss": 0.1079,
      "num_input_tokens_seen": 7147096,
      "step": 1280
    },
    {
      "epoch": 6.722040549378679,
      "grad_norm": 0.46682003140449524,
      "learning_rate": 4.4012374420859786e-05,
      "loss": 0.1061,
      "num_input_tokens_seen": 7202584,
      "step": 1290
    },
    {
      "epoch": 6.774362328319163,
      "grad_norm": 0.32285481691360474,
      "learning_rate": 4.392308501978148e-05,
      "loss": 0.1098,
      "num_input_tokens_seen": 7258552,
      "step": 1300
    },
    {
      "epoch": 6.826684107259647,
      "grad_norm": 0.3166508078575134,
      "learning_rate": 4.383322679019472e-05,
      "loss": 0.1119,
      "num_input_tokens_seen": 7315768,
      "step": 1310
    },
    {
      "epoch": 6.87900588620013,
      "grad_norm": 0.35454803705215454,
      "learning_rate": 4.3742802433240625e-05,
      "loss": 0.1107,
      "num_input_tokens_seen": 7371352,
      "step": 1320
    },
    {
      "epoch": 6.931327665140615,
      "grad_norm": 0.2988300919532776,
      "learning_rate": 4.3651814667078086e-05,
      "loss": 0.1085,
      "num_input_tokens_seen": 7427800,
      "step": 1330
    },
    {
      "epoch": 6.983649444081099,
      "grad_norm": 0.3938722014427185,
      "learning_rate": 4.35602662268021e-05,
      "loss": 0.1064,
      "num_input_tokens_seen": 7484616,
      "step": 1340
    },
    {
      "epoch": 7.031393067364291,
      "grad_norm": 0.45549729466438293,
      "learning_rate": 4.346815986436158e-05,
      "loss": 0.1041,
      "num_input_tokens_seen": 7534896,
      "step": 1350
    },
    {
      "epoch": 7.083714846304774,
      "grad_norm": 0.6130596995353699,
      "learning_rate": 4.337549834847655e-05,
      "loss": 0.0977,
      "num_input_tokens_seen": 7591328,
      "step": 1360
    },
    {
      "epoch": 7.136036625245258,
      "grad_norm": 0.6437245607376099,
      "learning_rate": 4.328228446455498e-05,
      "loss": 0.0979,
      "num_input_tokens_seen": 7647760,
      "step": 1370
    },
    {
      "epoch": 7.188358404185743,
      "grad_norm": 0.5056234002113342,
      "learning_rate": 4.3188521014609054e-05,
      "loss": 0.0994,
      "num_input_tokens_seen": 7704672,
      "step": 1380
    },
    {
      "epoch": 7.240680183126226,
      "grad_norm": 0.5103983879089355,
      "learning_rate": 4.309421081717091e-05,
      "loss": 0.1001,
      "num_input_tokens_seen": 7761104,
      "step": 1390
    },
    {
      "epoch": 7.29300196206671,
      "grad_norm": 0.47358736395835876,
      "learning_rate": 4.299935670720794e-05,
      "loss": 0.0926,
      "num_input_tokens_seen": 7818224,
      "step": 1400
    },
    {
      "epoch": 7.345323741007194,
      "grad_norm": 0.6868699789047241,
      "learning_rate": 4.290396153603755e-05,
      "loss": 0.0975,
      "num_input_tokens_seen": 7874208,
      "step": 1410
    },
    {
      "epoch": 7.3976455199476785,
      "grad_norm": 0.39971715211868286,
      "learning_rate": 4.280802817124149e-05,
      "loss": 0.1012,
      "num_input_tokens_seen": 7930496,
      "step": 1420
    },
    {
      "epoch": 7.449967298888162,
      "grad_norm": 0.6476240754127502,
      "learning_rate": 4.271155949657959e-05,
      "loss": 0.1033,
      "num_input_tokens_seen": 7985552,
      "step": 1430
    },
    {
      "epoch": 7.502289077828646,
      "grad_norm": 0.5134351253509521,
      "learning_rate": 4.261455841190314e-05,
      "loss": 0.0966,
      "num_input_tokens_seen": 8041568,
      "step": 1440
    },
    {
      "epoch": 7.554610856769131,
      "grad_norm": 0.6232761144638062,
      "learning_rate": 4.2517027833067685e-05,
      "loss": 0.1001,
      "num_input_tokens_seen": 8098656,
      "step": 1450
    },
    {
      "epoch": 7.606932635709614,
      "grad_norm": 0.4906207025051117,
      "learning_rate": 4.241897069184537e-05,
      "loss": 0.1072,
      "num_input_tokens_seen": 8154032,
      "step": 1460
    },
    {
      "epoch": 7.659254414650098,
      "grad_norm": 0.6239527463912964,
      "learning_rate": 4.2320389935836836e-05,
      "loss": 0.1006,
      "num_input_tokens_seen": 8210032,
      "step": 1470
    },
    {
      "epoch": 7.711576193590582,
      "grad_norm": 1.2038999795913696,
      "learning_rate": 4.2221288528382584e-05,
      "loss": 0.1015,
      "num_input_tokens_seen": 8265296,
      "step": 1480
    },
    {
      "epoch": 7.763897972531066,
      "grad_norm": 0.365824431180954,
      "learning_rate": 4.212166944847392e-05,
      "loss": 0.0973,
      "num_input_tokens_seen": 8321840,
      "step": 1490
    },
    {
      "epoch": 7.81621975147155,
      "grad_norm": 0.4954347610473633,
      "learning_rate": 4.2021535690663414e-05,
      "loss": 0.1015,
      "num_input_tokens_seen": 8378064,
      "step": 1500
    },
    {
      "epoch": 7.868541530412034,
      "grad_norm": 0.5742694139480591,
      "learning_rate": 4.192089026497484e-05,
      "loss": 0.0997,
      "num_input_tokens_seen": 8434384,
      "step": 1510
    },
    {
      "epoch": 7.920863309352518,
      "grad_norm": 0.48765209317207336,
      "learning_rate": 4.181973619681276e-05,
      "loss": 0.1003,
      "num_input_tokens_seen": 8490496,
      "step": 1520
    },
    {
      "epoch": 7.973185088293002,
      "grad_norm": 0.4735221266746521,
      "learning_rate": 4.171807652687151e-05,
      "loss": 0.1051,
      "num_input_tokens_seen": 8545600,
      "step": 1530
    },
    {
      "epoch": 8.020928711576193,
      "grad_norm": 0.35401248931884766,
      "learning_rate": 4.1615914311043855e-05,
      "loss": 0.0967,
      "num_input_tokens_seen": 8595848,
      "step": 1540
    },
    {
      "epoch": 8.073250490516678,
      "grad_norm": 0.7497507929801941,
      "learning_rate": 4.151325262032908e-05,
      "loss": 0.0834,
      "num_input_tokens_seen": 8652536,
      "step": 1550
    },
    {
      "epoch": 8.125572269457162,
      "grad_norm": 0.7235395908355713,
      "learning_rate": 4.1410094540740726e-05,
      "loss": 0.0804,
      "num_input_tokens_seen": 8708952,
      "step": 1560
    },
    {
      "epoch": 8.177894048397645,
      "grad_norm": 0.9055391550064087,
      "learning_rate": 4.1306443173213785e-05,
      "loss": 0.085,
      "num_input_tokens_seen": 8765688,
      "step": 1570
    },
    {
      "epoch": 8.23021582733813,
      "grad_norm": 0.8378714919090271,
      "learning_rate": 4.1202301633511506e-05,
      "loss": 0.0813,
      "num_input_tokens_seen": 8822376,
      "step": 1580
    },
    {
      "epoch": 8.282537606278613,
      "grad_norm": 0.6385327577590942,
      "learning_rate": 4.109767305213173e-05,
      "loss": 0.0831,
      "num_input_tokens_seen": 8878456,
      "step": 1590
    },
    {
      "epoch": 8.334859385219097,
      "grad_norm": 0.8356183767318726,
      "learning_rate": 4.0992560574212764e-05,
      "loss": 0.0893,
      "num_input_tokens_seen": 8934088,
      "step": 1600
    },
    {
      "epoch": 8.387181164159582,
      "grad_norm": 0.8791753649711609,
      "learning_rate": 4.0886967359438885e-05,
      "loss": 0.087,
      "num_input_tokens_seen": 8990120,
      "step": 1610
    },
    {
      "epoch": 8.439502943100065,
      "grad_norm": 0.6865050792694092,
      "learning_rate": 4.078089658194533e-05,
      "loss": 0.0873,
      "num_input_tokens_seen": 9045848,
      "step": 1620
    },
    {
      "epoch": 8.49182472204055,
      "grad_norm": 0.7075687050819397,
      "learning_rate": 4.0674351430222864e-05,
      "loss": 0.0855,
      "num_input_tokens_seen": 9102456,
      "step": 1630
    },
    {
      "epoch": 8.544146500981034,
      "grad_norm": 0.7803745269775391,
      "learning_rate": 4.0567335107021986e-05,
      "loss": 0.0887,
      "num_input_tokens_seen": 9158712,
      "step": 1640
    },
    {
      "epoch": 8.596468279921517,
      "grad_norm": 0.6807383894920349,
      "learning_rate": 4.0459850829256604e-05,
      "loss": 0.0826,
      "num_input_tokens_seen": 9215496,
      "step": 1650
    },
    {
      "epoch": 8.648790058862001,
      "grad_norm": 0.777777373790741,
      "learning_rate": 4.035190182790738e-05,
      "loss": 0.0922,
      "num_input_tokens_seen": 9269720,
      "step": 1660
    },
    {
      "epoch": 8.701111837802486,
      "grad_norm": 0.6249901056289673,
      "learning_rate": 4.024349134792453e-05,
      "loss": 0.0862,
      "num_input_tokens_seen": 9325624,
      "step": 1670
    },
    {
      "epoch": 8.753433616742969,
      "grad_norm": 0.7085604667663574,
      "learning_rate": 4.0134622648130394e-05,
      "loss": 0.0836,
      "num_input_tokens_seen": 9380984,
      "step": 1680
    },
    {
      "epoch": 8.805755395683454,
      "grad_norm": 0.9964501261711121,
      "learning_rate": 4.0025299001121365e-05,
      "loss": 0.0862,
      "num_input_tokens_seen": 9437320,
      "step": 1690
    },
    {
      "epoch": 8.858077174623936,
      "grad_norm": 1.0215933322906494,
      "learning_rate": 3.991552369316958e-05,
      "loss": 0.0926,
      "num_input_tokens_seen": 9492888,
      "step": 1700
    },
    {
      "epoch": 8.910398953564421,
      "grad_norm": 0.7546895146369934,
      "learning_rate": 3.9805300024124125e-05,
      "loss": 0.0893,
      "num_input_tokens_seen": 9550504,
      "step": 1710
    },
    {
      "epoch": 8.962720732504906,
      "grad_norm": 0.8502705693244934,
      "learning_rate": 3.969463130731183e-05,
      "loss": 0.0948,
      "num_input_tokens_seen": 9606120,
      "step": 1720
    },
    {
      "epoch": 9.010464355788097,
      "grad_norm": 0.5291112065315247,
      "learning_rate": 3.9583520869437666e-05,
      "loss": 0.0893,
      "num_input_tokens_seen": 9656472,
      "step": 1730
    },
    {
      "epoch": 9.062786134728581,
      "grad_norm": 0.6509256958961487,
      "learning_rate": 3.9471972050484764e-05,
      "loss": 0.063,
      "num_input_tokens_seen": 9712712,
      "step": 1740
    },
    {
      "epoch": 9.115107913669064,
      "grad_norm": 1.1916331052780151,
      "learning_rate": 3.9359988203614e-05,
      "loss": 0.0602,
      "num_input_tokens_seen": 9768584,
      "step": 1750
    },
    {
      "epoch": 9.167429692609549,
      "grad_norm": 0.8077856302261353,
      "learning_rate": 3.924757269506319e-05,
      "loss": 0.0643,
      "num_input_tokens_seen": 9824360,
      "step": 1760
    },
    {
      "epoch": 9.219751471550033,
      "grad_norm": 1.0554853677749634,
      "learning_rate": 3.913472890404593e-05,
      "loss": 0.0668,
      "num_input_tokens_seen": 9880328,
      "step": 1770
    },
    {
      "epoch": 9.272073250490516,
      "grad_norm": 0.8842382431030273,
      "learning_rate": 3.9021460222649986e-05,
      "loss": 0.0633,
      "num_input_tokens_seen": 9936248,
      "step": 1780
    },
    {
      "epoch": 9.324395029431,
      "grad_norm": 0.9562489986419678,
      "learning_rate": 3.890777005573537e-05,
      "loss": 0.0625,
      "num_input_tokens_seen": 9992088,
      "step": 1790
    },
    {
      "epoch": 9.376716808371485,
      "grad_norm": 0.8844836950302124,
      "learning_rate": 3.8793661820831915e-05,
      "loss": 0.0718,
      "num_input_tokens_seen": 10048520,
      "step": 1800
    },
    {
      "epoch": 9.429038587311968,
      "grad_norm": 1.2325676679611206,
      "learning_rate": 3.867913894803663e-05,
      "loss": 0.0646,
      "num_input_tokens_seen": 10104776,
      "step": 1810
    },
    {
      "epoch": 9.481360366252453,
      "grad_norm": 0.7227656245231628,
      "learning_rate": 3.8564204879910535e-05,
      "loss": 0.0657,
      "num_input_tokens_seen": 10160888,
      "step": 1820
    },
    {
      "epoch": 9.533682145192937,
      "grad_norm": 0.9736649990081787,
      "learning_rate": 3.844886307137519e-05,
      "loss": 0.0691,
      "num_input_tokens_seen": 10216920,
      "step": 1830
    },
    {
      "epoch": 9.58600392413342,
      "grad_norm": 1.0427885055541992,
      "learning_rate": 3.833311698960888e-05,
      "loss": 0.0667,
      "num_input_tokens_seen": 10273880,
      "step": 1840
    },
    {
      "epoch": 9.638325703073905,
      "grad_norm": 0.9749732613563538,
      "learning_rate": 3.8216970113942284e-05,
      "loss": 0.0699,
      "num_input_tokens_seen": 10329752,
      "step": 1850
    },
    {
      "epoch": 9.690647482014388,
      "grad_norm": 0.6216735243797302,
      "learning_rate": 3.8100425935754025e-05,
      "loss": 0.072,
      "num_input_tokens_seen": 10387128,
      "step": 1860
    },
    {
      "epoch": 9.742969260954872,
      "grad_norm": 0.8047035932540894,
      "learning_rate": 3.798348795836562e-05,
      "loss": 0.0697,
      "num_input_tokens_seen": 10443144,
      "step": 1870
    },
    {
      "epoch": 9.795291039895357,
      "grad_norm": 0.793222188949585,
      "learning_rate": 3.786615969693621e-05,
      "loss": 0.0709,
      "num_input_tokens_seen": 10497784,
      "step": 1880
    },
    {
      "epoch": 9.84761281883584,
      "grad_norm": 0.7294532656669617,
      "learning_rate": 3.7748444678356886e-05,
      "loss": 0.0716,
      "num_input_tokens_seen": 10553416,
      "step": 1890
    },
    {
      "epoch": 9.899934597776324,
      "grad_norm": 1.014732003211975,
      "learning_rate": 3.7630346441144656e-05,
      "loss": 0.0695,
      "num_input_tokens_seen": 10610168,
      "step": 1900
    },
    {
      "epoch": 9.952256376716809,
      "grad_norm": 1.22804856300354,
      "learning_rate": 3.7511868535336134e-05,
      "loss": 0.0731,
      "num_input_tokens_seen": 10665912,
      "step": 1910
    },
    {
      "epoch": 10.0,
      "grad_norm": 0.7104299068450928,
      "learning_rate": 3.7393014522380734e-05,
      "loss": 0.0652,
      "num_input_tokens_seen": 10717424,
      "step": 1920
    },
    {
      "epoch": 10.052321778940485,
      "grad_norm": 0.9654238224029541,
      "learning_rate": 3.7273787975033686e-05,
      "loss": 0.0435,
      "num_input_tokens_seen": 10772656,
      "step": 1930
    },
    {
      "epoch": 10.104643557880967,
      "grad_norm": 1.2412711381912231,
      "learning_rate": 3.7154192477248614e-05,
      "loss": 0.0425,
      "num_input_tokens_seen": 10829200,
      "step": 1940
    },
    {
      "epoch": 10.156965336821452,
      "grad_norm": 0.6960983276367188,
      "learning_rate": 3.7034231624069796e-05,
      "loss": 0.0376,
      "num_input_tokens_seen": 10884064,
      "step": 1950
    },
    {
      "epoch": 10.209287115761937,
      "grad_norm": 1.214837908744812,
      "learning_rate": 3.691390902152412e-05,
      "loss": 0.0487,
      "num_input_tokens_seen": 10939824,
      "step": 1960
    },
    {
      "epoch": 10.26160889470242,
      "grad_norm": 0.7800888419151306,
      "learning_rate": 3.679322828651265e-05,
      "loss": 0.043,
      "num_input_tokens_seen": 10996128,
      "step": 1970
    },
    {
      "epoch": 10.313930673642904,
      "grad_norm": 1.8462598323822021,
      "learning_rate": 3.667219304670193e-05,
      "loss": 0.0454,
      "num_input_tokens_seen": 11053280,
      "step": 1980
    },
    {
      "epoch": 10.366252452583387,
      "grad_norm": 1.075537085533142,
      "learning_rate": 3.655080694041495e-05,
      "loss": 0.0479,
      "num_input_tokens_seen": 11109664,
      "step": 1990
    },
    {
      "epoch": 10.418574231523872,
      "grad_norm": 0.7474583983421326,
      "learning_rate": 3.642907361652172e-05,
      "loss": 0.0458,
      "num_input_tokens_seen": 11166016,
      "step": 2000
    },
    {
      "epoch": 10.470896010464356,
      "grad_norm": 1.0032005310058594,
      "learning_rate": 3.6306996734329656e-05,
      "loss": 0.0518,
      "num_input_tokens_seen": 11221648,
      "step": 2010
    },
    {
      "epoch": 10.523217789404839,
      "grad_norm": 1.0080397129058838,
      "learning_rate": 3.618457996347352e-05,
      "loss": 0.0501,
      "num_input_tokens_seen": 11277952,
      "step": 2020
    },
    {
      "epoch": 10.575539568345324,
      "grad_norm": 1.130409598350525,
      "learning_rate": 3.606182698380515e-05,
      "loss": 0.0476,
      "num_input_tokens_seen": 11334272,
      "step": 2030
    },
    {
      "epoch": 10.627861347285808,
      "grad_norm": 0.9204115867614746,
      "learning_rate": 3.593874148528284e-05,
      "loss": 0.0515,
      "num_input_tokens_seen": 11389760,
      "step": 2040
    },
    {
      "epoch": 10.680183126226291,
      "grad_norm": 0.8044779896736145,
      "learning_rate": 3.58153271678604e-05,
      "loss": 0.0477,
      "num_input_tokens_seen": 11446144,
      "step": 2050
    },
    {
      "epoch": 10.732504905166776,
      "grad_norm": 1.4428260326385498,
      "learning_rate": 3.5691587741375934e-05,
      "loss": 0.0488,
      "num_input_tokens_seen": 11502320,
      "step": 2060
    },
    {
      "epoch": 10.78482668410726,
      "grad_norm": 0.9769074320793152,
      "learning_rate": 3.5567526925440353e-05,
      "loss": 0.0565,
      "num_input_tokens_seen": 11559392,
      "step": 2070
    },
    {
      "epoch": 10.837148463047743,
      "grad_norm": 0.6837288737297058,
      "learning_rate": 3.5443148449325545e-05,
      "loss": 0.051,
      "num_input_tokens_seen": 11615824,
      "step": 2080
    },
    {
      "epoch": 10.889470241988228,
      "grad_norm": 1.1239960193634033,
      "learning_rate": 3.5318456051852264e-05,
      "loss": 0.0528,
      "num_input_tokens_seen": 11671968,
      "step": 2090
    },
    {
      "epoch": 10.941792020928713,
      "grad_norm": 1.2493700981140137,
      "learning_rate": 3.519345348127775e-05,
      "loss": 0.0479,
      "num_input_tokens_seen": 11727632,
      "step": 2100
    },
    {
      "epoch": 10.994113799869195,
      "grad_norm": 1.0371160507202148,
      "learning_rate": 3.506814449518306e-05,
      "loss": 0.0446,
      "num_input_tokens_seen": 11784032,
      "step": 2110
    },
    {
      "epoch": 11.041857423152388,
      "grad_norm": 0.6289274096488953,
      "learning_rate": 3.494253286036011e-05,
      "loss": 0.0392,
      "num_input_tokens_seen": 11835760,
      "step": 2120
    },
    {
      "epoch": 11.094179202092871,
      "grad_norm": 1.0801132917404175,
      "learning_rate": 3.481662235269844e-05,
      "loss": 0.0311,
      "num_input_tokens_seen": 11891584,
      "step": 2130
    },
    {
      "epoch": 11.146500981033356,
      "grad_norm": 0.7098966836929321,
      "learning_rate": 3.469041675707173e-05,
      "loss": 0.0253,
      "num_input_tokens_seen": 11947824,
      "step": 2140
    },
    {
      "epoch": 11.198822759973838,
      "grad_norm": 0.9716883897781372,
      "learning_rate": 3.4563919867224e-05,
      "loss": 0.0313,
      "num_input_tokens_seen": 12003328,
      "step": 2150
    },
    {
      "epoch": 11.251144538914323,
      "grad_norm": 1.6546238660812378,
      "learning_rate": 3.4437135485655575e-05,
      "loss": 0.0339,
      "num_input_tokens_seen": 12060512,
      "step": 2160
    },
    {
      "epoch": 11.303466317854808,
      "grad_norm": 0.8873046040534973,
      "learning_rate": 3.4310067423508815e-05,
      "loss": 0.0344,
      "num_input_tokens_seen": 12117584,
      "step": 2170
    },
    {
      "epoch": 11.35578809679529,
      "grad_norm": 1.2070609331130981,
      "learning_rate": 3.418271950045352e-05,
      "loss": 0.0274,
      "num_input_tokens_seen": 12172512,
      "step": 2180
    },
    {
      "epoch": 11.408109875735775,
      "grad_norm": 1.0395190715789795,
      "learning_rate": 3.405509554457211e-05,
      "loss": 0.0268,
      "num_input_tokens_seen": 12227744,
      "step": 2190
    },
    {
      "epoch": 11.46043165467626,
      "grad_norm": 1.2013176679611206,
      "learning_rate": 3.392719939224453e-05,
      "loss": 0.0363,
      "num_input_tokens_seen": 12283776,
      "step": 2200
    },
    {
      "epoch": 11.512753433616743,
      "grad_norm": 1.1089166402816772,
      "learning_rate": 3.379903488803304e-05,
      "loss": 0.0339,
      "num_input_tokens_seen": 12340288,
      "step": 2210
    },
    {
      "epoch": 11.565075212557227,
      "grad_norm": 1.4544755220413208,
      "learning_rate": 3.3670605884566484e-05,
      "loss": 0.0325,
      "num_input_tokens_seen": 12396368,
      "step": 2220
    },
    {
      "epoch": 11.61739699149771,
      "grad_norm": 1.0953842401504517,
      "learning_rate": 3.3541916242424606e-05,
      "loss": 0.0333,
      "num_input_tokens_seen": 12451872,
      "step": 2230
    },
    {
      "epoch": 11.669718770438195,
      "grad_norm": 0.7939156889915466,
      "learning_rate": 3.341296983002193e-05,
      "loss": 0.0336,
      "num_input_tokens_seen": 12507776,
      "step": 2240
    },
    {
      "epoch": 11.72204054937868,
      "grad_norm": 1.114825963973999,
      "learning_rate": 3.3283770523491535e-05,
      "loss": 0.0357,
      "num_input_tokens_seen": 12564320,
      "step": 2250
    },
    {
      "epoch": 11.774362328319162,
      "grad_norm": 1.0780022144317627,
      "learning_rate": 3.3154322206568475e-05,
      "loss": 0.0356,
      "num_input_tokens_seen": 12620912,
      "step": 2260
    },
    {
      "epoch": 11.826684107259647,
      "grad_norm": 0.9889864325523376,
      "learning_rate": 3.302462877047307e-05,
      "loss": 0.0318,
      "num_input_tokens_seen": 12676464,
      "step": 2270
    },
    {
      "epoch": 11.879005886200131,
      "grad_norm": 1.8913953304290771,
      "learning_rate": 3.2894694113793935e-05,
      "loss": 0.039,
      "num_input_tokens_seen": 12731408,
      "step": 2280
    },
    {
      "epoch": 11.931327665140614,
      "grad_norm": 0.854158341884613,
      "learning_rate": 3.27645221423708e-05,
      "loss": 0.0408,
      "num_input_tokens_seen": 12787552,
      "step": 2290
    },
    {
      "epoch": 11.983649444081099,
      "grad_norm": 0.7944401502609253,
      "learning_rate": 3.263411676917704e-05,
      "loss": 0.034,
      "num_input_tokens_seen": 12843808,
      "step": 2300
    },
    {
      "epoch": 12.03139306736429,
      "grad_norm": 1.3224999904632568,
      "learning_rate": 3.250348191420214e-05,
      "loss": 0.03,
      "num_input_tokens_seen": 12895184,
      "step": 2310
    },
    {
      "epoch": 12.083714846304774,
      "grad_norm": 0.8005937933921814,
      "learning_rate": 3.237262150433379e-05,
      "loss": 0.0219,
      "num_input_tokens_seen": 12951408,
      "step": 2320
    },
    {
      "epoch": 12.136036625245259,
      "grad_norm": 1.3628501892089844,
      "learning_rate": 3.224153947323987e-05,
      "loss": 0.0181,
      "num_input_tokens_seen": 13007776,
      "step": 2330
    },
    {
      "epoch": 12.188358404185742,
      "grad_norm": 0.7954509258270264,
      "learning_rate": 3.21102397612502e-05,
      "loss": 0.0183,
      "num_input_tokens_seen": 13064144,
      "step": 2340
    },
    {
      "epoch": 12.240680183126226,
      "grad_norm": 0.8565235733985901,
      "learning_rate": 3.1978726315238094e-05,
      "loss": 0.0183,
      "num_input_tokens_seen": 13120320,
      "step": 2350
    },
    {
      "epoch": 12.293001962066711,
      "grad_norm": 0.7555674910545349,
      "learning_rate": 3.1847003088501726e-05,
      "loss": 0.017,
      "num_input_tokens_seen": 13177168,
      "step": 2360
    },
    {
      "epoch": 12.345323741007194,
      "grad_norm": 0.7122445106506348,
      "learning_rate": 3.1715074040645275e-05,
      "loss": 0.0206,
      "num_input_tokens_seen": 13232784,
      "step": 2370
    },
    {
      "epoch": 12.397645519947678,
      "grad_norm": 0.9428816437721252,
      "learning_rate": 3.158294313745992e-05,
      "loss": 0.0194,
      "num_input_tokens_seen": 13287312,
      "step": 2380
    },
    {
      "epoch": 12.449967298888161,
      "grad_norm": 1.027761459350586,
      "learning_rate": 3.145061435080461e-05,
      "loss": 0.0165,
      "num_input_tokens_seen": 13343616,
      "step": 2390
    },
    {
      "epoch": 12.502289077828646,
      "grad_norm": 0.9591146111488342,
      "learning_rate": 3.1318091658486655e-05,
      "loss": 0.0192,
      "num_input_tokens_seen": 13398656,
      "step": 2400
    },
    {
      "epoch": 12.55461085676913,
      "grad_norm": 2.0098116397857666,
      "learning_rate": 3.1185379044142225e-05,
      "loss": 0.0202,
      "num_input_tokens_seen": 13453888,
      "step": 2410
    },
    {
      "epoch": 12.606932635709613,
      "grad_norm": 1.243646502494812,
      "learning_rate": 3.105248049711651e-05,
      "loss": 0.0184,
      "num_input_tokens_seen": 13511168,
      "step": 2420
    },
    {
      "epoch": 12.659254414650098,
      "grad_norm": 0.7831906676292419,
      "learning_rate": 3.091940001234386e-05,
      "loss": 0.0215,
      "num_input_tokens_seen": 13567168,
      "step": 2430
    },
    {
      "epoch": 12.711576193590583,
      "grad_norm": 0.6232236623764038,
      "learning_rate": 3.078614159022767e-05,
      "loss": 0.0192,
      "num_input_tokens_seen": 13623200,
      "step": 2440
    },
    {
      "epoch": 12.763897972531066,
      "grad_norm": 1.3829624652862549,
      "learning_rate": 3.065270923652015e-05,
      "loss": 0.0222,
      "num_input_tokens_seen": 13678880,
      "step": 2450
    },
    {
      "epoch": 12.81621975147155,
      "grad_norm": 0.9216393232345581,
      "learning_rate": 3.051910696220188e-05,
      "loss": 0.0159,
      "num_input_tokens_seen": 13734624,
      "step": 2460
    },
    {
      "epoch": 12.868541530412035,
      "grad_norm": 1.1284723281860352,
      "learning_rate": 3.0385338783361283e-05,
      "loss": 0.0248,
      "num_input_tokens_seen": 13790576,
      "step": 2470
    },
    {
      "epoch": 12.920863309352518,
      "grad_norm": 1.4107064008712769,
      "learning_rate": 3.025140872107386e-05,
      "loss": 0.0227,
      "num_input_tokens_seen": 13845984,
      "step": 2480
    },
    {
      "epoch": 12.973185088293002,
      "grad_norm": 0.7538278102874756,
      "learning_rate": 3.0117320801281335e-05,
      "loss": 0.0265,
      "num_input_tokens_seen": 13902400,
      "step": 2490
    },
    {
      "epoch": 13.020928711576193,
      "grad_norm": 0.9580565690994263,
      "learning_rate": 2.9983079054670627e-05,
      "loss": 0.0195,
      "num_input_tokens_seen": 13953344,
      "step": 2500
    },
    {
      "epoch": 13.073250490516678,
      "grad_norm": 0.4210267961025238,
      "learning_rate": 2.9848687516552725e-05,
      "loss": 0.0107,
      "num_input_tokens_seen": 14009424,
      "step": 2510
    },
    {
      "epoch": 13.125572269457162,
      "grad_norm": 0.534055233001709,
      "learning_rate": 2.9714150226741312e-05,
      "loss": 0.013,
      "num_input_tokens_seen": 14064880,
      "step": 2520
    },
    {
      "epoch": 13.177894048397645,
      "grad_norm": 0.8242612481117249,
      "learning_rate": 2.9579471229431394e-05,
      "loss": 0.0095,
      "num_input_tokens_seen": 14120896,
      "step": 2530
    },
    {
      "epoch": 13.23021582733813,
      "grad_norm": 1.1644961833953857,
      "learning_rate": 2.944465457307771e-05,
      "loss": 0.0125,
      "num_input_tokens_seen": 14176512,
      "step": 2540
    },
    {
      "epoch": 13.282537606278613,
      "grad_norm": 0.682830274105072,
      "learning_rate": 2.930970431027304e-05,
      "loss": 0.0106,
      "num_input_tokens_seen": 14232608,
      "step": 2550
    },
    {
      "epoch": 13.334859385219097,
      "grad_norm": 0.7904958724975586,
      "learning_rate": 2.9174624497626353e-05,
      "loss": 0.012,
      "num_input_tokens_seen": 14289360,
      "step": 2560
    },
    {
      "epoch": 13.387181164159582,
      "grad_norm": 0.8092711567878723,
      "learning_rate": 2.903941919564091e-05,
      "loss": 0.0124,
      "num_input_tokens_seen": 14346096,
      "step": 2570
    },
    {
      "epoch": 13.439502943100065,
      "grad_norm": 0.4784017503261566,
      "learning_rate": 2.8904092468592187e-05,
      "loss": 0.0132,
      "num_input_tokens_seen": 14401872,
      "step": 2580
    },
    {
      "epoch": 13.49182472204055,
      "grad_norm": 0.5194114446640015,
      "learning_rate": 2.8768648384405695e-05,
      "loss": 0.0101,
      "num_input_tokens_seen": 14458864,
      "step": 2590
    },
    {
      "epoch": 13.544146500981034,
      "grad_norm": 0.6601864099502563,
      "learning_rate": 2.863309101453469e-05,
      "loss": 0.0135,
      "num_input_tokens_seen": 14515664,
      "step": 2600
    },
    {
      "epoch": 13.596468279921517,
      "grad_norm": 0.9567685723304749,
      "learning_rate": 2.8497424433837833e-05,
      "loss": 0.0138,
      "num_input_tokens_seen": 14572256,
      "step": 2610
    },
    {
      "epoch": 13.648790058862001,
      "grad_norm": 0.5563291311264038,
      "learning_rate": 2.836165272045663e-05,
      "loss": 0.0132,
      "num_input_tokens_seen": 14627248,
      "step": 2620
    },
    {
      "epoch": 13.701111837802486,
      "grad_norm": 0.9716143608093262,
      "learning_rate": 2.8225779955692905e-05,
      "loss": 0.0134,
      "num_input_tokens_seen": 14683728,
      "step": 2630
    },
    {
      "epoch": 13.753433616742969,
      "grad_norm": 0.9606854915618896,
      "learning_rate": 2.8089810223886076e-05,
      "loss": 0.0154,
      "num_input_tokens_seen": 14740864,
      "step": 2640
    },
    {
      "epoch": 13.805755395683454,
      "grad_norm": 1.01091730594635,
      "learning_rate": 2.79537476122904e-05,
      "loss": 0.0121,
      "num_input_tokens_seen": 14796176,
      "step": 2650
    },
    {
      "epoch": 13.858077174623936,
      "grad_norm": 0.6134788990020752,
      "learning_rate": 2.781759621095209e-05,
      "loss": 0.0119,
      "num_input_tokens_seen": 14852304,
      "step": 2660
    },
    {
      "epoch": 13.910398953564421,
      "grad_norm": 1.1731514930725098,
      "learning_rate": 2.7681360112586403e-05,
      "loss": 0.0188,
      "num_input_tokens_seen": 14908624,
      "step": 2670
    },
    {
      "epoch": 13.962720732504906,
      "grad_norm": 0.7572017908096313,
      "learning_rate": 2.7545043412454568e-05,
      "loss": 0.0153,
      "num_input_tokens_seen": 14964784,
      "step": 2680
    },
    {
      "epoch": 14.010464355788097,
      "grad_norm": 0.2877230942249298,
      "learning_rate": 2.7408650208240733e-05,
      "loss": 0.0093,
      "num_input_tokens_seen": 15016112,
      "step": 2690
    },
    {
      "epoch": 14.062786134728581,
      "grad_norm": 1.2057104110717773,
      "learning_rate": 2.7272184599928723e-05,
      "loss": 0.006,
      "num_input_tokens_seen": 15072240,
      "step": 2700
    },
    {
      "epoch": 14.115107913669064,
      "grad_norm": 0.2993405759334564,
      "learning_rate": 2.7135650689678873e-05,
      "loss": 0.0082,
      "num_input_tokens_seen": 15128432,
      "step": 2710
    },
    {
      "epoch": 14.167429692609549,
      "grad_norm": 0.4322413504123688,
      "learning_rate": 2.6999052581704643e-05,
      "loss": 0.0052,
      "num_input_tokens_seen": 15185232,
      "step": 2720
    },
    {
      "epoch": 14.219751471550033,
      "grad_norm": 0.4944108724594116,
      "learning_rate": 2.6862394382149308e-05,
      "loss": 0.0066,
      "num_input_tokens_seen": 15241040,
      "step": 2730
    },
    {
      "epoch": 14.272073250490516,
      "grad_norm": 0.6533095836639404,
      "learning_rate": 2.672568019896248e-05,
      "loss": 0.0088,
      "num_input_tokens_seen": 15297904,
      "step": 2740
    },
    {
      "epoch": 14.324395029431,
      "grad_norm": 0.316057026386261,
      "learning_rate": 2.6588914141776626e-05,
      "loss": 0.0056,
      "num_input_tokens_seen": 15355584,
      "step": 2750
    },
    {
      "epoch": 14.376716808371485,
      "grad_norm": 0.502487063407898,
      "learning_rate": 2.6452100321783585e-05,
      "loss": 0.0029,
      "num_input_tokens_seen": 15410592,
      "step": 2760
    },
    {
      "epoch": 14.429038587311968,
      "grad_norm": 0.5012995004653931,
      "learning_rate": 2.6315242851610923e-05,
      "loss": 0.0109,
      "num_input_tokens_seen": 15466448,
      "step": 2770
    },
    {
      "epoch": 14.481360366252453,
      "grad_norm": 0.8451622128486633,
      "learning_rate": 2.6178345845198328e-05,
      "loss": 0.0057,
      "num_input_tokens_seen": 15522816,
      "step": 2780
    },
    {
      "epoch": 14.533682145192937,
      "grad_norm": 0.33364975452423096,
      "learning_rate": 2.6041413417673966e-05,
      "loss": 0.009,
      "num_input_tokens_seen": 15578672,
      "step": 2790
    },
    {
      "epoch": 14.58600392413342,
      "grad_norm": 0.3638412058353424,
      "learning_rate": 2.590444968523074e-05,
      "loss": 0.0089,
      "num_input_tokens_seen": 15635408,
      "step": 2800
    },
    {
      "epoch": 14.638325703073905,
      "grad_norm": 0.5961637496948242,
      "learning_rate": 2.5767458765002606e-05,
      "loss": 0.008,
      "num_input_tokens_seen": 15691648,
      "step": 2810
    },
    {
      "epoch": 14.690647482014388,
      "grad_norm": 0.7401494979858398,
      "learning_rate": 2.5630444774940765e-05,
      "loss": 0.0081,
      "num_input_tokens_seen": 15748032,
      "step": 2820
    },
    {
      "epoch": 14.742969260954872,
      "grad_norm": 0.18349328637123108,
      "learning_rate": 2.5493411833689907e-05,
      "loss": 0.0071,
      "num_input_tokens_seen": 15803232,
      "step": 2830
    },
    {
      "epoch": 14.795291039895357,
      "grad_norm": 1.5877436399459839,
      "learning_rate": 2.5356364060464398e-05,
      "loss": 0.0078,
      "num_input_tokens_seen": 15859120,
      "step": 2840
    },
    {
      "epoch": 14.84761281883584,
      "grad_norm": 1.041685938835144,
      "learning_rate": 2.521930557492444e-05,
      "loss": 0.0089,
      "num_input_tokens_seen": 15915872,
      "step": 2850
    },
    {
      "epoch": 14.899934597776324,
      "grad_norm": 0.6710309982299805,
      "learning_rate": 2.5082240497052267e-05,
      "loss": 0.0088,
      "num_input_tokens_seen": 15973472,
      "step": 2860
    },
    {
      "epoch": 14.952256376716809,
      "grad_norm": 0.9839669466018677,
      "learning_rate": 2.494517294702826e-05,
      "loss": 0.0069,
      "num_input_tokens_seen": 16029920,
      "step": 2870
    },
    {
      "epoch": 15.0,
      "grad_norm": 1.3188862800598145,
      "learning_rate": 2.4808107045107123e-05,
      "loss": 0.0098,
      "num_input_tokens_seen": 16080272,
      "step": 2880
    },
    {
      "epoch": 15.052321778940485,
      "grad_norm": 0.41307705640792847,
      "learning_rate": 2.4671046911494025e-05,
      "loss": 0.0037,
      "num_input_tokens_seen": 16136752,
      "step": 2890
    },
    {
      "epoch": 15.104643557880967,
      "grad_norm": 0.8025826811790466,
      "learning_rate": 2.453399666622072e-05,
      "loss": 0.0032,
      "num_input_tokens_seen": 16191920,
      "step": 2900
    },
    {
      "epoch": 15.156965336821452,
      "grad_norm": 0.2138717621564865,
      "learning_rate": 2.4396960429021738e-05,
      "loss": 0.0028,
      "num_input_tokens_seen": 16246912,
      "step": 2910
    },
    {
      "epoch": 15.209287115761937,
      "grad_norm": 0.16525974869728088,
      "learning_rate": 2.4259942319210498e-05,
      "loss": 0.0058,
      "num_input_tokens_seen": 16303520,
      "step": 2920
    },
    {
      "epoch": 15.26160889470242,
      "grad_norm": 0.22977286577224731,
      "learning_rate": 2.412294645555555e-05,
      "loss": 0.005,
      "num_input_tokens_seen": 16359888,
      "step": 2930
    },
    {
      "epoch": 15.313930673642904,
      "grad_norm": 1.8685427904129028,
      "learning_rate": 2.39859769561567e-05,
      "loss": 0.0048,
      "num_input_tokens_seen": 16416512,
      "step": 2940
    },
    {
      "epoch": 15.366252452583387,
      "grad_norm": 0.7653654217720032,
      "learning_rate": 2.3849037938321235e-05,
      "loss": 0.0056,
      "num_input_tokens_seen": 16473664,
      "step": 2950
    },
    {
      "epoch": 15.418574231523872,
      "grad_norm": 0.21836823225021362,
      "learning_rate": 2.3712133518440176e-05,
      "loss": 0.0072,
      "num_input_tokens_seen": 16529312,
      "step": 2960
    },
    {
      "epoch": 15.470896010464356,
      "grad_norm": 0.6065989136695862,
      "learning_rate": 2.3575267811864543e-05,
      "loss": 0.0074,
      "num_input_tokens_seen": 16586048,
      "step": 2970
    },
    {
      "epoch": 15.523217789404839,
      "grad_norm": 0.21767759323120117,
      "learning_rate": 2.34384449327816e-05,
      "loss": 0.0037,
      "num_input_tokens_seen": 16642560,
      "step": 2980
    },
    {
      "epoch": 15.575539568345324,
      "grad_norm": 0.35699793696403503,
      "learning_rate": 2.330166899409124e-05,
      "loss": 0.0039,
      "num_input_tokens_seen": 16699248,
      "step": 2990
    },
    {
      "epoch": 15.627861347285808,
      "grad_norm": 0.46648091077804565,
      "learning_rate": 2.3164944107282333e-05,
      "loss": 0.0067,
      "num_input_tokens_seen": 16755952,
      "step": 3000
    },
    {
      "epoch": 15.680183126226291,
      "grad_norm": 0.8464080691337585,
      "learning_rate": 2.3028274382309097e-05,
      "loss": 0.0061,
      "num_input_tokens_seen": 16811536,
      "step": 3010
    },
    {
      "epoch": 15.732504905166776,
      "grad_norm": 0.30818283557891846,
      "learning_rate": 2.2891663927467604e-05,
      "loss": 0.0046,
      "num_input_tokens_seen": 16867824,
      "step": 3020
    },
    {
      "epoch": 15.78482668410726,
      "grad_norm": 0.5573921799659729,
      "learning_rate": 2.2755116849272274e-05,
      "loss": 0.0041,
      "num_input_tokens_seen": 16924080,
      "step": 3030
    },
    {
      "epoch": 15.837148463047743,
      "grad_norm": 0.5058010816574097,
      "learning_rate": 2.2618637252332398e-05,
      "loss": 0.0065,
      "num_input_tokens_seen": 16979728,
      "step": 3040
    },
    {
      "epoch": 15.889470241988228,
      "grad_norm": 0.4849563241004944,
      "learning_rate": 2.2482229239228785e-05,
      "loss": 0.0047,
      "num_input_tokens_seen": 17035488,
      "step": 3050
    },
    {
      "epoch": 15.941792020928713,
      "grad_norm": 0.10410932451486588,
      "learning_rate": 2.234589691039046e-05,
      "loss": 0.0054,
      "num_input_tokens_seen": 17091072,
      "step": 3060
    },
    {
      "epoch": 15.994113799869195,
      "grad_norm": 0.31193605065345764,
      "learning_rate": 2.2209644363971337e-05,
      "loss": 0.0043,
      "num_input_tokens_seen": 17147328,
      "step": 3070
    },
    {
      "epoch": 16.041857423152386,
      "grad_norm": 0.15345972776412964,
      "learning_rate": 2.2073475695727096e-05,
      "loss": 0.0045,
      "num_input_tokens_seen": 17198200,
      "step": 3080
    },
    {
      "epoch": 16.09417920209287,
      "grad_norm": 0.8098782896995544,
      "learning_rate": 2.193739499889201e-05,
      "loss": 0.0042,
      "num_input_tokens_seen": 17254408,
      "step": 3090
    },
    {
      "epoch": 16.146500981033356,
      "grad_norm": 0.6010912656784058,
      "learning_rate": 2.1801406364055958e-05,
      "loss": 0.0049,
      "num_input_tokens_seen": 17311304,
      "step": 3100
    },
    {
      "epoch": 16.19882275997384,
      "grad_norm": 0.0812903568148613,
      "learning_rate": 2.1665513879041418e-05,
      "loss": 0.001,
      "num_input_tokens_seen": 17368152,
      "step": 3110
    },
    {
      "epoch": 16.251144538914325,
      "grad_norm": 0.08344978094100952,
      "learning_rate": 2.1529721628780593e-05,
      "loss": 0.0037,
      "num_input_tokens_seen": 17423480,
      "step": 3120
    },
    {
      "epoch": 16.303466317854806,
      "grad_norm": 0.3543494641780853,
      "learning_rate": 2.1394033695192645e-05,
      "loss": 0.0016,
      "num_input_tokens_seen": 17478984,
      "step": 3130
    },
    {
      "epoch": 16.35578809679529,
      "grad_norm": 0.683672308921814,
      "learning_rate": 2.125845415706097e-05,
      "loss": 0.0019,
      "num_input_tokens_seen": 17535592,
      "step": 3140
    },
    {
      "epoch": 16.408109875735775,
      "grad_norm": 0.07661443203687668,
      "learning_rate": 2.1122987089910577e-05,
      "loss": 0.0012,
      "num_input_tokens_seen": 17591960,
      "step": 3150
    },
    {
      "epoch": 16.46043165467626,
      "grad_norm": 0.21092914044857025,
      "learning_rate": 2.0987636565885606e-05,
      "loss": 0.004,
      "num_input_tokens_seen": 17648504,
      "step": 3160
    },
    {
      "epoch": 16.512753433616744,
      "grad_norm": 0.8990269303321838,
      "learning_rate": 2.0852406653626916e-05,
      "loss": 0.003,
      "num_input_tokens_seen": 17705240,
      "step": 3170
    },
    {
      "epoch": 16.565075212557225,
      "grad_norm": 0.22914479672908783,
      "learning_rate": 2.0717301418149742e-05,
      "loss": 0.0028,
      "num_input_tokens_seen": 17760392,
      "step": 3180
    },
    {
      "epoch": 16.61739699149771,
      "grad_norm": 1.7169655561447144,
      "learning_rate": 2.058232492072157e-05,
      "loss": 0.0033,
      "num_input_tokens_seen": 17816744,
      "step": 3190
    },
    {
      "epoch": 16.669718770438195,
      "grad_norm": 0.07557539641857147,
      "learning_rate": 2.044748121874e-05,
      "loss": 0.0021,
      "num_input_tokens_seen": 17872760,
      "step": 3200
    },
    {
      "epoch": 16.72204054937868,
      "grad_norm": 0.1961260586977005,
      "learning_rate": 2.0312774365610783e-05,
      "loss": 0.0017,
      "num_input_tokens_seen": 17928696,
      "step": 3210
    },
    {
      "epoch": 16.774362328319164,
      "grad_norm": 0.6838825941085815,
      "learning_rate": 2.0178208410626006e-05,
      "loss": 0.0011,
      "num_input_tokens_seen": 17984232,
      "step": 3220
    },
    {
      "epoch": 16.82668410725965,
      "grad_norm": 0.12144844979047775,
      "learning_rate": 2.0043787398842347e-05,
      "loss": 0.0022,
      "num_input_tokens_seen": 18040712,
      "step": 3230
    },
    {
      "epoch": 16.87900588620013,
      "grad_norm": 0.3987827003002167,
      "learning_rate": 1.9909515370959493e-05,
      "loss": 0.0029,
      "num_input_tokens_seen": 18097016,
      "step": 3240
    },
    {
      "epoch": 16.931327665140614,
      "grad_norm": 0.10082973539829254,
      "learning_rate": 1.9775396363198654e-05,
      "loss": 0.0015,
      "num_input_tokens_seen": 18152776,
      "step": 3250
    },
    {
      "epoch": 16.9836494440811,
      "grad_norm": 0.020283468067646027,
      "learning_rate": 1.9641434407181285e-05,
      "loss": 0.002,
      "num_input_tokens_seen": 18208456,
      "step": 3260
    },
    {
      "epoch": 17.03139306736429,
      "grad_norm": 0.23080122470855713,
      "learning_rate": 1.950763352980782e-05,
      "loss": 0.0006,
      "num_input_tokens_seen": 18259784,
      "step": 3270
    },
    {
      "epoch": 17.083714846304776,
      "grad_norm": 0.00946098379790783,
      "learning_rate": 1.9373997753136695e-05,
      "loss": 0.0003,
      "num_input_tokens_seen": 18316008,
      "step": 3280
    },
    {
      "epoch": 17.136036625245257,
      "grad_norm": 0.033656761050224304,
      "learning_rate": 1.9240531094263388e-05,
      "loss": 0.0008,
      "num_input_tokens_seen": 18372696,
      "step": 3290
    },
    {
      "epoch": 17.188358404185742,
      "grad_norm": 0.03647719696164131,
      "learning_rate": 1.9107237565199716e-05,
      "loss": 0.0009,
      "num_input_tokens_seen": 18428488,
      "step": 3300
    },
    {
      "epoch": 17.240680183126226,
      "grad_norm": 0.030296266078948975,
      "learning_rate": 1.8974121172753192e-05,
      "loss": 0.0004,
      "num_input_tokens_seen": 18484120,
      "step": 3310
    },
    {
      "epoch": 17.29300196206671,
      "grad_norm": 0.043923936784267426,
      "learning_rate": 1.8841185918406594e-05,
      "loss": 0.0004,
      "num_input_tokens_seen": 18539976,
      "step": 3320
    },
    {
      "epoch": 17.345323741007196,
      "grad_norm": 0.038960348814725876,
      "learning_rate": 1.870843579819771e-05,
      "loss": 0.0008,
      "num_input_tokens_seen": 18596792,
      "step": 3330
    },
    {
      "epoch": 17.397645519947677,
      "grad_norm": 0.010129265487194061,
      "learning_rate": 1.8575874802599162e-05,
      "loss": 0.001,
      "num_input_tokens_seen": 18652776,
      "step": 3340
    },
    {
      "epoch": 17.44996729888816,
      "grad_norm": 0.04958868771791458,
      "learning_rate": 1.8443506916398485e-05,
      "loss": 0.0004,
      "num_input_tokens_seen": 18709320,
      "step": 3350
    },
    {
      "epoch": 17.502289077828646,
      "grad_norm": 2.036484956741333,
      "learning_rate": 1.8311336118578355e-05,
      "loss": 0.0018,
      "num_input_tokens_seen": 18766376,
      "step": 3360
    },
    {
      "epoch": 17.55461085676913,
      "grad_norm": 0.10162217170000076,
      "learning_rate": 1.8179366382196944e-05,
      "loss": 0.0015,
      "num_input_tokens_seen": 18822440,
      "step": 3370
    },
    {
      "epoch": 17.606932635709615,
      "grad_norm": 0.41068536043167114,
      "learning_rate": 1.8047601674268522e-05,
      "loss": 0.0011,
      "num_input_tokens_seen": 18877976,
      "step": 3380
    },
    {
      "epoch": 17.6592544146501,
      "grad_norm": 0.35680681467056274,
      "learning_rate": 1.7916045955644207e-05,
      "loss": 0.0015,
      "num_input_tokens_seen": 18934728,
      "step": 3390
    },
    {
      "epoch": 17.71157619359058,
      "grad_norm": 0.16506262123584747,
      "learning_rate": 1.7784703180892882e-05,
      "loss": 0.0004,
      "num_input_tokens_seen": 18990088,
      "step": 3400
    },
    {
      "epoch": 17.763897972531066,
      "grad_norm": 0.05834071710705757,
      "learning_rate": 1.7653577298182327e-05,
      "loss": 0.0004,
      "num_input_tokens_seen": 19046728,
      "step": 3410
    },
    {
      "epoch": 17.81621975147155,
      "grad_norm": 0.14426672458648682,
      "learning_rate": 1.752267224916055e-05,
      "loss": 0.0022,
      "num_input_tokens_seen": 19101672,
      "step": 3420
    },
    {
      "epoch": 17.868541530412035,
      "grad_norm": 0.2213674634695053,
      "learning_rate": 1.7391991968837272e-05,
      "loss": 0.0007,
      "num_input_tokens_seen": 19159128,
      "step": 3430
    },
    {
      "epoch": 17.92086330935252,
      "grad_norm": 0.6867311000823975,
      "learning_rate": 1.726154038546569e-05,
      "loss": 0.0005,
      "num_input_tokens_seen": 19215448,
      "step": 3440
    },
    {
      "epoch": 17.973185088293,
      "grad_norm": 0.022834990173578262,
      "learning_rate": 1.713132142042434e-05,
      "loss": 0.0013,
      "num_input_tokens_seen": 19270328,
      "step": 3450
    },
    {
      "epoch": 18.020928711576193,
      "grad_norm": 0.017958860844373703,
      "learning_rate": 1.7001338988099264e-05,
      "loss": 0.0004,
      "num_input_tokens_seen": 19321096,
      "step": 3460
    },
    {
      "epoch": 18.073250490516678,
      "grad_norm": 0.013346249237656593,
      "learning_rate": 1.68715969957663e-05,
      "loss": 0.0001,
      "num_input_tokens_seen": 19377144,
      "step": 3470
    },
    {
      "epoch": 18.125572269457162,
      "grad_norm": 0.016560234129428864,
      "learning_rate": 1.6742099343473674e-05,
      "loss": 0.0009,
      "num_input_tokens_seen": 19433080,
      "step": 3480
    },
    {
      "epoch": 18.177894048397647,
      "grad_norm": 0.3799145519733429,
      "learning_rate": 1.6612849923924723e-05,
      "loss": 0.0003,
      "num_input_tokens_seen": 19489176,
      "step": 3490
    },
    {
      "epoch": 18.230215827338128,
      "grad_norm": 0.9833922982215881,
      "learning_rate": 1.6483852622360923e-05,
      "loss": 0.0003,
      "num_input_tokens_seen": 19544920,
      "step": 3500
    },
    {
      "epoch": 18.282537606278613,
      "grad_norm": 0.017415842041373253,
      "learning_rate": 1.635511131644505e-05,
      "loss": 0.0008,
      "num_input_tokens_seen": 19600888,
      "step": 3510
    },
    {
      "epoch": 18.334859385219097,
      "grad_norm": 0.05391722172498703,
      "learning_rate": 1.6226629876144657e-05,
      "loss": 0.0002,
      "num_input_tokens_seen": 19656168,
      "step": 3520
    },
    {
      "epoch": 18.387181164159582,
      "grad_norm": 0.32104507088661194,
      "learning_rate": 1.609841216361574e-05,
      "loss": 0.0005,
      "num_input_tokens_seen": 19711224,
      "step": 3530
    },
    {
      "epoch": 18.439502943100067,
      "grad_norm": 0.019494058564305305,
      "learning_rate": 1.597046203308662e-05,
      "loss": 0.0004,
      "num_input_tokens_seen": 19767768,
      "step": 3540
    },
    {
      "epoch": 18.491824722040548,
      "grad_norm": 0.03266040235757828,
      "learning_rate": 1.584278333074208e-05,
      "loss": 0.0001,
      "num_input_tokens_seen": 19824616,
      "step": 3550
    },
    {
      "epoch": 18.544146500981032,
      "grad_norm": 0.024637416005134583,
      "learning_rate": 1.571537989460779e-05,
      "loss": 0.0007,
      "num_input_tokens_seen": 19880024,
      "step": 3560
    },
    {
      "epoch": 18.596468279921517,
      "grad_norm": 0.03031347133219242,
      "learning_rate": 1.5588255554434883e-05,
      "loss": 0.0003,
      "num_input_tokens_seen": 19936504,
      "step": 3570
    },
    {
      "epoch": 18.648790058862,
      "grad_norm": 0.005317925941199064,
      "learning_rate": 1.5461414131584873e-05,
      "loss": 0.0006,
      "num_input_tokens_seen": 19992136,
      "step": 3580
    },
    {
      "epoch": 18.701111837802486,
      "grad_norm": 0.024280209094285965,
      "learning_rate": 1.533485943891478e-05,
      "loss": 0.0035,
      "num_input_tokens_seen": 20049128,
      "step": 3590
    },
    {
      "epoch": 18.75343361674297,
      "grad_norm": 0.9302756190299988,
      "learning_rate": 1.5208595280662497e-05,
      "loss": 0.0002,
      "num_input_tokens_seen": 20106488,
      "step": 3600
    },
    {
      "epoch": 18.805755395683452,
      "grad_norm": 0.013046924024820328,
      "learning_rate": 1.5082625452332433e-05,
      "loss": 0.0003,
      "num_input_tokens_seen": 20162536,
      "step": 3610
    },
    {
      "epoch": 18.858077174623936,
      "grad_norm": 0.01642036624252796,
      "learning_rate": 1.4956953740581454e-05,
      "loss": 0.0003,
      "num_input_tokens_seen": 20219032,
      "step": 3620
    },
    {
      "epoch": 18.91039895356442,
      "grad_norm": 0.01480098720639944,
      "learning_rate": 1.4831583923104999e-05,
      "loss": 0.0003,
      "num_input_tokens_seen": 20275880,
      "step": 3630
    },
    {
      "epoch": 18.962720732504906,
      "grad_norm": 0.05232972651720047,
      "learning_rate": 1.4706519768523597e-05,
      "loss": 0.0006,
      "num_input_tokens_seen": 20332264,
      "step": 3640
    },
    {
      "epoch": 19.0104643557881,
      "grad_norm": 0.008723029866814613,
      "learning_rate": 1.458176503626949e-05,
      "loss": 0.0001,
      "num_input_tokens_seen": 20382464,
      "step": 3650
    },
    {
      "epoch": 19.06278613472858,
      "grad_norm": 0.35439035296440125,
      "learning_rate": 1.4457323476473738e-05,
      "loss": 0.0005,
      "num_input_tokens_seen": 20438720,
      "step": 3660
    },
    {
      "epoch": 19.115107913669064,
      "grad_norm": 0.02343558706343174,
      "learning_rate": 1.4333198829853394e-05,
      "loss": 0.0001,
      "num_input_tokens_seen": 20493616,
      "step": 3670
    },
    {
      "epoch": 19.16742969260955,
      "grad_norm": 0.014349430799484253,
      "learning_rate": 1.420939482759907e-05,
      "loss": 0.0003,
      "num_input_tokens_seen": 20550000,
      "step": 3680
    },
    {
      "epoch": 19.219751471550033,
      "grad_norm": 0.044981323182582855,
      "learning_rate": 1.4085915191262832e-05,
      "loss": 0.0001,
      "num_input_tokens_seen": 20606144,
      "step": 3690
    },
    {
      "epoch": 19.272073250490518,
      "grad_norm": 0.023957155644893646,
      "learning_rate": 1.396276363264629e-05,
      "loss": 0.0001,
      "num_input_tokens_seen": 20662720,
      "step": 3700
    },
    {
      "epoch": 19.324395029431,
      "grad_norm": 0.015040691941976547,
      "learning_rate": 1.3839943853689024e-05,
      "loss": 0.0002,
      "num_input_tokens_seen": 20718992,
      "step": 3710
    },
    {
      "epoch": 19.376716808371484,
      "grad_norm": 0.011137389577925205,
      "learning_rate": 1.3717459546357284e-05,
      "loss": 0.0001,
      "num_input_tokens_seen": 20776096,
      "step": 3720
    },
    {
      "epoch": 19.429038587311968,
      "grad_norm": 0.059972431510686874,
      "learning_rate": 1.3595314392533083e-05,
      "loss": 0.0003,
      "num_input_tokens_seen": 20831584,
      "step": 3730
    },
    {
      "epoch": 19.481360366252453,
      "grad_norm": 0.003549647517502308,
      "learning_rate": 1.3473512063903432e-05,
      "loss": 0.0006,
      "num_input_tokens_seen": 20887408,
      "step": 3740
    },
    {
      "epoch": 19.533682145192937,
      "grad_norm": 0.01864522323012352,
      "learning_rate": 1.335205622185003e-05,
      "loss": 0.0003,
      "num_input_tokens_seen": 20944080,
      "step": 3750
    },
    {
      "epoch": 19.586003924133422,
      "grad_norm": 0.013788777403533459,
      "learning_rate": 1.3230950517339141e-05,
      "loss": 0.0001,
      "num_input_tokens_seen": 21000576,
      "step": 3760
    },
    {
      "epoch": 19.638325703073903,
      "grad_norm": 0.007708332501351833,
      "learning_rate": 1.3110198590811918e-05,
      "loss": 0.0001,
      "num_input_tokens_seen": 21056608,
      "step": 3770
    },
    {
      "epoch": 19.690647482014388,
      "grad_norm": 0.008117050863802433,
      "learning_rate": 1.2989804072074918e-05,
      "loss": 0.0001,
      "num_input_tokens_seen": 21112528,
      "step": 3780
    },
    {
      "epoch": 19.742969260954872,
      "grad_norm": 0.005804878659546375,
      "learning_rate": 1.2869770580191051e-05,
      "loss": 0.0001,
      "num_input_tokens_seen": 21169104,
      "step": 3790
    },
    {
      "epoch": 19.795291039895357,
      "grad_norm": 0.021145416423678398,
      "learning_rate": 1.2750101723370683e-05,
      "loss": 0.0021,
      "num_input_tokens_seen": 21225440,
      "step": 3800
    },
    {
      "epoch": 19.84761281883584,
      "grad_norm": 0.031260546296834946,
      "learning_rate": 1.2630801098863284e-05,
      "loss": 0.0014,
      "num_input_tokens_seen": 21281952,
      "step": 3810
    },
    {
      "epoch": 19.899934597776323,
      "grad_norm": 0.024749331176280975,
      "learning_rate": 1.2511872292849236e-05,
      "loss": 0.0001,
      "num_input_tokens_seen": 21338448,
      "step": 3820
    },
    {
      "epoch": 19.952256376716807,
      "grad_norm": 0.11673085391521454,
      "learning_rate": 1.2393318880332062e-05,
      "loss": 0.0001,
      "num_input_tokens_seen": 21394640,
      "step": 3830
    },
    {
      "epoch": 20.0,
      "grad_norm": 0.07176396250724792,
      "learning_rate": 1.2275144425030902e-05,
      "loss": 0.0003,
      "num_input_tokens_seen": 21445504,
      "step": 3840
    },
    {
      "epoch": 20.052321778940485,
      "grad_norm": 0.024757077917456627,
      "learning_rate": 1.2157352479273465e-05,
      "loss": 0.0004,
      "num_input_tokens_seen": 21503072,
      "step": 3850
    },
    {
      "epoch": 20.10464355788097,
      "grad_norm": 0.016589034348726273,
      "learning_rate": 1.2039946583889225e-05,
      "loss": 0.0001,
      "num_input_tokens_seen": 21559312,
      "step": 3860
    },
    {
      "epoch": 20.15696533682145,
      "grad_norm": 0.004230449441820383,
      "learning_rate": 1.1922930268102949e-05,
      "loss": 0.0001,
      "num_input_tokens_seen": 21616032,
      "step": 3870
    },
    {
      "epoch": 20.209287115761935,
      "grad_norm": 0.007515770383179188,
      "learning_rate": 1.1806307049428616e-05,
      "loss": 0.0003,
      "num_input_tokens_seen": 21671872,
      "step": 3880
    },
    {
      "epoch": 20.26160889470242,
      "grad_norm": 0.0041547054424881935,
      "learning_rate": 1.1690080433563716e-05,
      "loss": 0.0001,
      "num_input_tokens_seen": 21727616,
      "step": 3890
    },
    {
      "epoch": 20.313930673642904,
      "grad_norm": 0.0042037139646708965,
      "learning_rate": 1.157425391428384e-05,
      "loss": 0.0,
      "num_input_tokens_seen": 21784400,
      "step": 3900
    },
    {
      "epoch": 20.36625245258339,
      "grad_norm": 0.024101046845316887,
      "learning_rate": 1.145883097333767e-05,
      "loss": 0.0001,
      "num_input_tokens_seen": 21841584,
      "step": 3910
    },
    {
      "epoch": 20.418574231523873,
      "grad_norm": 0.01927500218153,
      "learning_rate": 1.1343815080342279e-05,
      "loss": 0.0,
      "num_input_tokens_seen": 21897120,
      "step": 3920
    },
    {
      "epoch": 20.470896010464354,
      "grad_norm": 0.005595839582383633,
      "learning_rate": 1.1229209692678921e-05,
      "loss": 0.0001,
      "num_input_tokens_seen": 21952320,
      "step": 3930
    },
    {
      "epoch": 20.52321778940484,
      "grad_norm": 0.00780284171923995,
      "learning_rate": 1.1115018255389006e-05,
      "loss": 0.0001,
      "num_input_tokens_seen": 22008432,
      "step": 3940
    },
    {
      "epoch": 20.575539568345324,
      "grad_norm": 0.0051256874576210976,
      "learning_rate": 1.1001244201070606e-05,
      "loss": 0.0001,
      "num_input_tokens_seen": 22063664,
      "step": 3950
    },
    {
      "epoch": 20.62786134728581,
      "grad_norm": 0.0060654510743916035,
      "learning_rate": 1.088789094977522e-05,
      "loss": 0.0,
      "num_input_tokens_seen": 22119488,
      "step": 3960
    },
    {
      "epoch": 20.680183126226293,
      "grad_norm": 0.019679056480526924,
      "learning_rate": 1.077496190890502e-05,
      "loss": 0.0,
      "num_input_tokens_seen": 22175568,
      "step": 3970
    },
    {
      "epoch": 20.732504905166774,
      "grad_norm": 0.02322162687778473,
      "learning_rate": 1.0662460473110384e-05,
      "loss": 0.0,
      "num_input_tokens_seen": 22231472,
      "step": 3980
    },
    {
      "epoch": 20.78482668410726,
      "grad_norm": 0.001980294706299901,
      "learning_rate": 1.0550390024187906e-05,
      "loss": 0.0007,
      "num_input_tokens_seen": 22287120,
      "step": 3990
    },
    {
      "epoch": 20.837148463047743,
      "grad_norm": 0.3102468252182007,
      "learning_rate": 1.0438753930978643e-05,
      "loss": 0.0001,
      "num_input_tokens_seen": 22342736,
      "step": 4000
    }
  ],
  "logging_steps": 10,
  "max_steps": 5730,
  "num_input_tokens_seen": 22342736,
  "num_train_epochs": 30,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 9.588523420509798e+17,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}