{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.9985842378480414,
  "eval_steps": 500,
  "global_step": 3177,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.009438414346389807,
      "grad_norm": 0.7047261682571029,
      "learning_rate": 2.0833333333333334e-06,
      "loss": 0.3368,
      "step": 10
    },
    {
      "epoch": 0.018876828692779613,
      "grad_norm": 0.44565794909772116,
      "learning_rate": 4.166666666666667e-06,
      "loss": 0.2696,
      "step": 20
    },
    {
      "epoch": 0.028315243039169418,
      "grad_norm": 0.2469001773100438,
      "learning_rate": 6.25e-06,
      "loss": 0.2345,
      "step": 30
    },
    {
      "epoch": 0.037753657385559226,
      "grad_norm": 0.17504888016598083,
      "learning_rate": 8.333333333333334e-06,
      "loss": 0.213,
      "step": 40
    },
    {
      "epoch": 0.04719207173194903,
      "grad_norm": 0.14319985578848382,
      "learning_rate": 1.0416666666666668e-05,
      "loss": 0.194,
      "step": 50
    },
    {
      "epoch": 0.056630486078338836,
      "grad_norm": 0.1594795618687463,
      "learning_rate": 1.25e-05,
      "loss": 0.1871,
      "step": 60
    },
    {
      "epoch": 0.06606890042472864,
      "grad_norm": 0.14625288121586646,
      "learning_rate": 1.4583333333333333e-05,
      "loss": 0.1791,
      "step": 70
    },
    {
      "epoch": 0.07550731477111845,
      "grad_norm": 0.44536813060583275,
      "learning_rate": 1.6666666666666667e-05,
      "loss": 0.1777,
      "step": 80
    },
    {
      "epoch": 0.08494572911750826,
      "grad_norm": 0.17858201185676476,
      "learning_rate": 1.8750000000000002e-05,
      "loss": 0.1743,
      "step": 90
    },
    {
      "epoch": 0.09438414346389806,
      "grad_norm": 0.22679600482590775,
      "learning_rate": 1.9999916822524766e-05,
      "loss": 0.173,
      "step": 100
    },
    {
      "epoch": 0.10382255781028787,
      "grad_norm": 0.2245846959065181,
      "learning_rate": 1.999898109181919e-05,
      "loss": 0.1697,
      "step": 110
    },
    {
      "epoch": 0.11326097215667767,
      "grad_norm": 0.21274961909318216,
      "learning_rate": 1.9997005756177228e-05,
      "loss": 0.1669,
      "step": 120
    },
    {
      "epoch": 0.12269938650306748,
      "grad_norm": 0.15535944242468744,
      "learning_rate": 1.999399102097668e-05,
      "loss": 0.1664,
      "step": 130
    },
    {
      "epoch": 0.13213780084945728,
      "grad_norm": 0.1677351780597522,
      "learning_rate": 1.9989937199662845e-05,
      "loss": 0.1652,
      "step": 140
    },
    {
      "epoch": 0.1415762151958471,
      "grad_norm": 0.15664753634679404,
      "learning_rate": 1.998484471371593e-05,
      "loss": 0.1619,
      "step": 150
    },
    {
      "epoch": 0.1510146295422369,
      "grad_norm": 0.1936710831336268,
      "learning_rate": 1.9978714092607234e-05,
      "loss": 0.1606,
      "step": 160
    },
    {
      "epoch": 0.16045304388862672,
      "grad_norm": 0.21595388620143963,
      "learning_rate": 1.9971545973744102e-05,
      "loss": 0.16,
      "step": 170
    },
    {
      "epoch": 0.16989145823501653,
      "grad_norm": 0.19084712938417736,
      "learning_rate": 1.9963341102403652e-05,
      "loss": 0.1582,
      "step": 180
    },
    {
      "epoch": 0.1793298725814063,
      "grad_norm": 0.19755743286804991,
      "learning_rate": 1.9954100331655265e-05,
      "loss": 0.1551,
      "step": 190
    },
    {
      "epoch": 0.18876828692779613,
      "grad_norm": 0.21785901423412252,
      "learning_rate": 1.9943824622271934e-05,
      "loss": 0.1559,
      "step": 200
    },
    {
      "epoch": 0.19820670127418594,
      "grad_norm": 0.1472275149396309,
      "learning_rate": 1.9932515042630335e-05,
      "loss": 0.1534,
      "step": 210
    },
    {
      "epoch": 0.20764511562057575,
      "grad_norm": 0.14184865805893884,
      "learning_rate": 1.9920172768599763e-05,
      "loss": 0.1545,
      "step": 220
    },
    {
      "epoch": 0.21708352996696556,
      "grad_norm": 0.1833363522186809,
      "learning_rate": 1.9906799083419865e-05,
      "loss": 0.1543,
      "step": 230
    },
    {
      "epoch": 0.22652194431335534,
      "grad_norm": 0.141212536462394,
      "learning_rate": 1.989239537756723e-05,
      "loss": 0.1538,
      "step": 240
    },
    {
      "epoch": 0.23596035865974516,
      "grad_norm": 0.1556507383209179,
      "learning_rate": 1.987696314861082e-05,
      "loss": 0.1541,
      "step": 250
    },
    {
      "epoch": 0.24539877300613497,
      "grad_norm": 0.17173283637998002,
      "learning_rate": 1.986050400105626e-05,
      "loss": 0.1518,
      "step": 260
    },
    {
      "epoch": 0.25483718735252475,
      "grad_norm": 0.150629748875022,
      "learning_rate": 1.9843019646179014e-05,
      "loss": 0.1501,
      "step": 270
    },
    {
      "epoch": 0.26427560169891456,
      "grad_norm": 0.19334006281958221,
      "learning_rate": 1.9824511901846475e-05,
      "loss": 0.1483,
      "step": 280
    },
    {
      "epoch": 0.2737140160453044,
      "grad_norm": 0.12527664802237196,
      "learning_rate": 1.9804982692328944e-05,
      "loss": 0.1514,
      "step": 290
    },
    {
      "epoch": 0.2831524303916942,
      "grad_norm": 0.19858199254198736,
      "learning_rate": 1.9784434048099565e-05,
      "loss": 0.151,
      "step": 300
    },
    {
      "epoch": 0.292590844738084,
      "grad_norm": 0.1677799134496006,
      "learning_rate": 1.976286810562323e-05,
      "loss": 0.1498,
      "step": 310
    },
    {
      "epoch": 0.3020292590844738,
      "grad_norm": 0.15743349400071904,
      "learning_rate": 1.9740287107134417e-05,
      "loss": 0.1513,
      "step": 320
    },
    {
      "epoch": 0.3114676734308636,
      "grad_norm": 0.1616061413079364,
      "learning_rate": 1.97166934004041e-05,
      "loss": 0.1489,
      "step": 330
    },
    {
      "epoch": 0.32090608777725343,
      "grad_norm": 0.1878283137943562,
      "learning_rate": 1.9692089438495622e-05,
      "loss": 0.1449,
      "step": 340
    },
    {
      "epoch": 0.33034450212364325,
      "grad_norm": 0.16166637953640253,
      "learning_rate": 1.9666477779509655e-05,
      "loss": 0.1469,
      "step": 350
    },
    {
      "epoch": 0.33978291647003306,
      "grad_norm": 0.12292532976422958,
      "learning_rate": 1.963986108631823e-05,
      "loss": 0.1468,
      "step": 360
    },
    {
      "epoch": 0.3492213308164228,
      "grad_norm": 0.16469874208354635,
      "learning_rate": 1.9612242126287876e-05,
      "loss": 0.1483,
      "step": 370
    },
    {
      "epoch": 0.3586597451628126,
      "grad_norm": 0.13369950841636608,
      "learning_rate": 1.958362377099191e-05,
      "loss": 0.1443,
      "step": 380
    },
    {
      "epoch": 0.36809815950920244,
      "grad_norm": 0.12551292002845085,
      "learning_rate": 1.9554008995911837e-05,
      "loss": 0.1463,
      "step": 390
    },
    {
      "epoch": 0.37753657385559225,
      "grad_norm": 0.14851921568961518,
      "learning_rate": 1.9523400880128032e-05,
      "loss": 0.1471,
      "step": 400
    },
    {
      "epoch": 0.38697498820198206,
      "grad_norm": 0.13162403368451694,
      "learning_rate": 1.949180260599957e-05,
      "loss": 0.1452,
      "step": 410
    },
    {
      "epoch": 0.3964134025483719,
      "grad_norm": 0.12724887401811377,
      "learning_rate": 1.945921745883337e-05,
      "loss": 0.1455,
      "step": 420
    },
    {
      "epoch": 0.4058518168947617,
      "grad_norm": 0.11779906396951238,
      "learning_rate": 1.9425648826542618e-05,
      "loss": 0.1435,
      "step": 430
    },
    {
      "epoch": 0.4152902312411515,
      "grad_norm": 0.1610933457769652,
      "learning_rate": 1.939110019929451e-05,
      "loss": 0.1436,
      "step": 440
    },
    {
      "epoch": 0.4247286455875413,
      "grad_norm": 0.12250958203512534,
      "learning_rate": 1.935557516914739e-05,
      "loss": 0.1451,
      "step": 450
    },
    {
      "epoch": 0.4341670599339311,
      "grad_norm": 0.1380572524992858,
      "learning_rate": 1.931907742967727e-05,
      "loss": 0.1444,
      "step": 460
    },
    {
      "epoch": 0.44360547428032093,
      "grad_norm": 0.13646993698111895,
      "learning_rate": 1.92816107755938e-05,
      "loss": 0.142,
      "step": 470
    },
    {
      "epoch": 0.4530438886267107,
      "grad_norm": 0.11765542306036501,
      "learning_rate": 1.9243179102345753e-05,
      "loss": 0.1406,
      "step": 480
    },
    {
      "epoch": 0.4624823029731005,
      "grad_norm": 0.1266567901893174,
      "learning_rate": 1.9203786405715984e-05,
      "loss": 0.144,
      "step": 490
    },
    {
      "epoch": 0.4719207173194903,
      "grad_norm": 0.1113634311573256,
      "learning_rate": 1.9163436781405992e-05,
      "loss": 0.1428,
      "step": 500
    },
    {
      "epoch": 0.4813591316658801,
      "grad_norm": 0.13808836428511967,
      "learning_rate": 1.912213442461009e-05,
      "loss": 0.1399,
      "step": 510
    },
    {
      "epoch": 0.49079754601226994,
      "grad_norm": 0.1226613837593307,
      "learning_rate": 1.9079883629579224e-05,
      "loss": 0.1396,
      "step": 520
    },
    {
      "epoch": 0.5002359603586597,
      "grad_norm": 0.14272835200919645,
      "learning_rate": 1.9036688789174496e-05,
      "loss": 0.1403,
      "step": 530
    },
    {
      "epoch": 0.5096743747050495,
      "grad_norm": 0.12981510040553715,
      "learning_rate": 1.899255439441043e-05,
      "loss": 0.1399,
      "step": 540
    },
    {
      "epoch": 0.5191127890514393,
      "grad_norm": 0.1190871345092575,
      "learning_rate": 1.8947485033988034e-05,
      "loss": 0.1376,
      "step": 550
    },
    {
      "epoch": 0.5285512033978291,
      "grad_norm": 0.1271477738963388,
      "learning_rate": 1.8901485393817724e-05,
      "loss": 0.1415,
      "step": 560
    },
    {
      "epoch": 0.5379896177442189,
      "grad_norm": 0.12965211048846748,
      "learning_rate": 1.8854560256532098e-05,
      "loss": 0.1423,
      "step": 570
    },
    {
      "epoch": 0.5474280320906088,
      "grad_norm": 0.13373262160455968,
      "learning_rate": 1.880671450098871e-05,
      "loss": 0.139,
      "step": 580
    },
    {
      "epoch": 0.5568664464369986,
      "grad_norm": 0.1322939697550499,
      "learning_rate": 1.8757953101762786e-05,
      "loss": 0.1396,
      "step": 590
    },
    {
      "epoch": 0.5663048607833884,
      "grad_norm": 0.11918437239832326,
      "learning_rate": 1.8708281128630023e-05,
      "loss": 0.138,
      "step": 600
    },
    {
      "epoch": 0.5757432751297782,
      "grad_norm": 0.12338738357381479,
      "learning_rate": 1.865770374603948e-05,
      "loss": 0.1406,
      "step": 610
    },
    {
      "epoch": 0.585181689476168,
      "grad_norm": 0.11573754594906395,
      "learning_rate": 1.8606226212576612e-05,
      "loss": 0.138,
      "step": 620
    },
    {
      "epoch": 0.5946201038225578,
      "grad_norm": 0.1419588706141848,
      "learning_rate": 1.8553853880416555e-05,
      "loss": 0.1408,
      "step": 630
    },
    {
      "epoch": 0.6040585181689476,
      "grad_norm": 0.13998266637185536,
      "learning_rate": 1.8500592194767625e-05,
      "loss": 0.1394,
      "step": 640
    },
    {
      "epoch": 0.6134969325153374,
      "grad_norm": 0.11868995175822014,
      "learning_rate": 1.8446446693305194e-05,
      "loss": 0.1384,
      "step": 650
    },
    {
      "epoch": 0.6229353468617272,
      "grad_norm": 0.1328472026088287,
      "learning_rate": 1.8391423005595928e-05,
      "loss": 0.1393,
      "step": 660
    },
    {
      "epoch": 0.6323737612081171,
      "grad_norm": 0.11726921800593894,
      "learning_rate": 1.833552685251246e-05,
      "loss": 0.1398,
      "step": 670
    },
    {
      "epoch": 0.6418121755545069,
      "grad_norm": 0.11466260187649016,
      "learning_rate": 1.827876404563861e-05,
      "loss": 0.1369,
      "step": 680
    },
    {
      "epoch": 0.6512505899008967,
      "grad_norm": 0.11234281014514101,
      "learning_rate": 1.8221140486665125e-05,
      "loss": 0.1346,
      "step": 690
    },
    {
      "epoch": 0.6606890042472865,
      "grad_norm": 0.11159741277810285,
      "learning_rate": 1.8162662166776085e-05,
      "loss": 0.1357,
      "step": 700
    },
    {
      "epoch": 0.6701274185936763,
      "grad_norm": 0.12752868267859116,
      "learning_rate": 1.8103335166026002e-05,
      "loss": 0.1389,
      "step": 710
    },
    {
      "epoch": 0.6795658329400661,
      "grad_norm": 0.12084535348559353,
      "learning_rate": 1.804316565270765e-05,
      "loss": 0.1375,
      "step": 720
    },
    {
      "epoch": 0.6890042472864559,
      "grad_norm": 0.12102077085461252,
      "learning_rate": 1.798215988271075e-05,
      "loss": 0.1364,
      "step": 730
    },
    {
      "epoch": 0.6984426616328456,
      "grad_norm": 0.11713742692774234,
      "learning_rate": 1.7920324198871546e-05,
      "loss": 0.138,
      "step": 740
    },
    {
      "epoch": 0.7078810759792354,
      "grad_norm": 0.11656355822805255,
      "learning_rate": 1.785766503031332e-05,
      "loss": 0.1346,
      "step": 750
    },
    {
      "epoch": 0.7173194903256253,
      "grad_norm": 0.11377328844943654,
      "learning_rate": 1.7794188891777964e-05,
      "loss": 0.1352,
      "step": 760
    },
    {
      "epoch": 0.7267579046720151,
      "grad_norm": 0.12103799646507679,
      "learning_rate": 1.7729902382948617e-05,
      "loss": 0.1353,
      "step": 770
    },
    {
      "epoch": 0.7361963190184049,
      "grad_norm": 0.1073585292390918,
      "learning_rate": 1.76648121877635e-05,
      "loss": 0.1352,
      "step": 780
    },
    {
      "epoch": 0.7456347333647947,
      "grad_norm": 0.11214075940260533,
      "learning_rate": 1.759892507372099e-05,
      "loss": 0.1341,
      "step": 790
    },
    {
      "epoch": 0.7550731477111845,
      "grad_norm": 0.11706899066793994,
      "learning_rate": 1.7532247891175968e-05,
      "loss": 0.1333,
      "step": 800
    },
    {
      "epoch": 0.7645115620575743,
      "grad_norm": 0.11789888505768062,
      "learning_rate": 1.746478757262761e-05,
      "loss": 0.136,
      "step": 810
    },
    {
      "epoch": 0.7739499764039641,
      "grad_norm": 0.11237848535926774,
      "learning_rate": 1.739655113199858e-05,
      "loss": 0.1336,
      "step": 820
    },
    {
      "epoch": 0.7833883907503539,
      "grad_norm": 0.10753154987431834,
      "learning_rate": 1.7327545663905813e-05,
      "loss": 0.1331,
      "step": 830
    },
    {
      "epoch": 0.7928268050967437,
      "grad_norm": 0.1441522552747225,
      "learning_rate": 1.7257778342922853e-05,
      "loss": 0.1328,
      "step": 840
    },
    {
      "epoch": 0.8022652194431336,
      "grad_norm": 0.1269707942863234,
      "learning_rate": 1.7187256422833928e-05,
      "loss": 0.1319,
      "step": 850
    },
    {
      "epoch": 0.8117036337895234,
      "grad_norm": 0.11091236494221275,
      "learning_rate": 1.711598723587975e-05,
      "loss": 0.1324,
      "step": 860
    },
    {
      "epoch": 0.8211420481359132,
      "grad_norm": 0.10854265306012167,
      "learning_rate": 1.7043978191995177e-05,
      "loss": 0.1325,
      "step": 870
    },
    {
      "epoch": 0.830580462482303,
      "grad_norm": 0.1110467712060928,
      "learning_rate": 1.6971236778038806e-05,
      "loss": 0.1315,
      "step": 880
    },
    {
      "epoch": 0.8400188768286928,
      "grad_norm": 0.12129611756408008,
      "learning_rate": 1.6897770557014535e-05,
      "loss": 0.1328,
      "step": 890
    },
    {
      "epoch": 0.8494572911750826,
      "grad_norm": 0.106781748696916,
      "learning_rate": 1.682358716728525e-05,
      "loss": 0.1351,
      "step": 900
    },
    {
      "epoch": 0.8588957055214724,
      "grad_norm": 0.11020118439519076,
      "learning_rate": 1.674869432177864e-05,
      "loss": 0.1325,
      "step": 910
    },
    {
      "epoch": 0.8683341198678622,
      "grad_norm": 0.11750342557908768,
      "learning_rate": 1.667309980718529e-05,
      "loss": 0.1312,
      "step": 920
    },
    {
      "epoch": 0.877772534214252,
      "grad_norm": 0.12853148875033116,
      "learning_rate": 1.6596811483149077e-05,
      "loss": 0.1317,
      "step": 930
    },
    {
      "epoch": 0.8872109485606419,
      "grad_norm": 0.11393786746070304,
      "learning_rate": 1.651983728145e-05,
      "loss": 0.1355,
      "step": 940
    },
    {
      "epoch": 0.8966493629070316,
      "grad_norm": 0.11013100846033319,
      "learning_rate": 1.6442185205179507e-05,
      "loss": 0.1309,
      "step": 950
    },
    {
      "epoch": 0.9060877772534214,
      "grad_norm": 0.10641286141618259,
      "learning_rate": 1.6363863327908405e-05,
      "loss": 0.1339,
      "step": 960
    },
    {
      "epoch": 0.9155261915998112,
      "grad_norm": 0.11308702339858176,
      "learning_rate": 1.6284879792847433e-05,
      "loss": 0.1299,
      "step": 970
    },
    {
      "epoch": 0.924964605946201,
      "grad_norm": 0.11315250527471539,
      "learning_rate": 1.620524281200062e-05,
      "loss": 0.1305,
      "step": 980
    },
    {
      "epoch": 0.9344030202925908,
      "grad_norm": 0.0972875949252274,
      "learning_rate": 1.6124960665311447e-05,
      "loss": 0.1322,
      "step": 990
    },
    {
      "epoch": 0.9438414346389806,
      "grad_norm": 0.10070713866086979,
      "learning_rate": 1.6044041699802005e-05,
      "loss": 0.129,
      "step": 1000
    },
    {
      "epoch": 0.9532798489853704,
      "grad_norm": 0.11109074990403733,
      "learning_rate": 1.5962494328705123e-05,
      "loss": 0.1321,
      "step": 1010
    },
    {
      "epoch": 0.9627182633317602,
      "grad_norm": 0.1199186598391774,
      "learning_rate": 1.588032703058964e-05,
      "loss": 0.1334,
      "step": 1020
    },
    {
      "epoch": 0.9721566776781501,
      "grad_norm": 0.10777396066893469,
      "learning_rate": 1.5797548348478893e-05,
      "loss": 0.1325,
      "step": 1030
    },
    {
      "epoch": 0.9815950920245399,
      "grad_norm": 0.11999882098060052,
      "learning_rate": 1.571416688896246e-05,
      "loss": 0.132,
      "step": 1040
    },
    {
      "epoch": 0.9910335063709297,
      "grad_norm": 0.10911342083469809,
      "learning_rate": 1.563019132130136e-05,
      "loss": 0.1301,
      "step": 1050
    },
    {
      "epoch": 1.0004719207173194,
      "grad_norm": 0.11482143235010223,
      "learning_rate": 1.5545630376526665e-05,
      "loss": 0.1282,
      "step": 1060
    },
    {
      "epoch": 1.0099103350637093,
      "grad_norm": 0.11699392564682201,
      "learning_rate": 1.5460492846531748e-05,
      "loss": 0.1142,
      "step": 1070
    },
    {
      "epoch": 1.019348749410099,
      "grad_norm": 0.09530170230868218,
      "learning_rate": 1.5374787583158188e-05,
      "loss": 0.1157,
      "step": 1080
    },
    {
      "epoch": 1.028787163756489,
      "grad_norm": 0.09486822390799159,
      "learning_rate": 1.5288523497275392e-05,
      "loss": 0.1143,
      "step": 1090
    },
    {
      "epoch": 1.0382255781028786,
      "grad_norm": 0.09531194088354993,
      "learning_rate": 1.5201709557854178e-05,
      "loss": 0.1128,
      "step": 1100
    },
    {
      "epoch": 1.0476639924492686,
      "grad_norm": 0.11206649112299381,
      "learning_rate": 1.5114354791034225e-05,
      "loss": 0.1161,
      "step": 1110
    },
    {
      "epoch": 1.0571024067956583,
      "grad_norm": 0.10196697892456508,
      "learning_rate": 1.5026468279185615e-05,
      "loss": 0.1159,
      "step": 1120
    },
    {
      "epoch": 1.0665408211420482,
      "grad_norm": 0.10326390731228648,
      "learning_rate": 1.4938059159964555e-05,
      "loss": 0.1161,
      "step": 1130
    },
    {
      "epoch": 1.0759792354884379,
      "grad_norm": 0.09969628777021497,
      "learning_rate": 1.4849136625363297e-05,
      "loss": 0.1141,
      "step": 1140
    },
    {
      "epoch": 1.0854176498348278,
      "grad_norm": 0.0939091288214549,
      "learning_rate": 1.4759709920754453e-05,
      "loss": 0.1125,
      "step": 1150
    },
    {
      "epoch": 1.0948560641812175,
      "grad_norm": 0.09506323252337912,
      "learning_rate": 1.4669788343929736e-05,
      "loss": 0.1141,
      "step": 1160
    },
    {
      "epoch": 1.1042944785276074,
      "grad_norm": 0.10154441851850998,
      "learning_rate": 1.4579381244133265e-05,
      "loss": 0.1128,
      "step": 1170
    },
    {
      "epoch": 1.1137328928739971,
      "grad_norm": 0.11822773479215737,
      "learning_rate": 1.4488498021089514e-05,
      "loss": 0.1137,
      "step": 1180
    },
    {
      "epoch": 1.123171307220387,
      "grad_norm": 0.1009800661484683,
      "learning_rate": 1.4397148124025997e-05,
      "loss": 0.1143,
      "step": 1190
    },
    {
      "epoch": 1.1326097215667768,
      "grad_norm": 0.10126420585679885,
      "learning_rate": 1.4305341050690845e-05,
      "loss": 0.117,
      "step": 1200
    },
    {
      "epoch": 1.1420481359131667,
      "grad_norm": 0.09588874040008494,
      "learning_rate": 1.421308634636529e-05,
      "loss": 0.1137,
      "step": 1210
    },
    {
      "epoch": 1.1514865502595564,
      "grad_norm": 0.10922875430592754,
      "learning_rate": 1.412039360287126e-05,
      "loss": 0.1145,
      "step": 1220
    },
    {
      "epoch": 1.1609249646059463,
      "grad_norm": 0.11298890031757797,
      "learning_rate": 1.4027272457574082e-05,
      "loss": 0.1138,
      "step": 1230
    },
    {
      "epoch": 1.170363378952336,
      "grad_norm": 0.10857815603466196,
      "learning_rate": 1.3933732592380485e-05,
      "loss": 0.1135,
      "step": 1240
    },
    {
      "epoch": 1.1798017932987257,
      "grad_norm": 0.10213279825434321,
      "learning_rate": 1.3839783732731966e-05,
      "loss": 0.1134,
      "step": 1250
    },
    {
      "epoch": 1.1892402076451156,
      "grad_norm": 0.10027833977041692,
      "learning_rate": 1.3745435646593613e-05,
      "loss": 0.1136,
      "step": 1260
    },
    {
      "epoch": 1.1986786219915055,
      "grad_norm": 0.09590585357482817,
      "learning_rate": 1.3650698143438534e-05,
      "loss": 0.113,
      "step": 1270
    },
    {
      "epoch": 1.2081170363378952,
      "grad_norm": 0.10321073236266613,
      "learning_rate": 1.3555581073227942e-05,
      "loss": 0.1167,
      "step": 1280
    },
    {
      "epoch": 1.217555450684285,
      "grad_norm": 0.09327710523878686,
      "learning_rate": 1.346009432538705e-05,
      "loss": 0.1147,
      "step": 1290
    },
    {
      "epoch": 1.2269938650306749,
      "grad_norm": 0.0933025296287067,
      "learning_rate": 1.3364247827776854e-05,
      "loss": 0.1145,
      "step": 1300
    },
    {
      "epoch": 1.2364322793770646,
      "grad_norm": 0.09493771082819326,
      "learning_rate": 1.3268051545661937e-05,
      "loss": 0.1141,
      "step": 1310
    },
    {
      "epoch": 1.2458706937234545,
      "grad_norm": 0.10053484854502866,
      "learning_rate": 1.3171515480674342e-05,
      "loss": 0.1122,
      "step": 1320
    },
    {
      "epoch": 1.2553091080698442,
      "grad_norm": 0.1108335770631105,
      "learning_rate": 1.3074649669773716e-05,
      "loss": 0.1173,
      "step": 1330
    },
    {
      "epoch": 1.2647475224162341,
      "grad_norm": 0.10521299166726314,
      "learning_rate": 1.297746418420374e-05,
      "loss": 0.1103,
      "step": 1340
    },
    {
      "epoch": 1.2741859367626238,
      "grad_norm": 0.10478814209881943,
      "learning_rate": 1.2879969128445025e-05,
      "loss": 0.1122,
      "step": 1350
    },
    {
      "epoch": 1.2836243511090137,
      "grad_norm": 0.0969588608522638,
      "learning_rate": 1.2782174639164528e-05,
      "loss": 0.1112,
      "step": 1360
    },
    {
      "epoch": 1.2930627654554034,
      "grad_norm": 0.10783687256200783,
      "learning_rate": 1.2684090884161636e-05,
      "loss": 0.1125,
      "step": 1370
    },
    {
      "epoch": 1.3025011798017934,
      "grad_norm": 0.10076692580892369,
      "learning_rate": 1.2585728061311003e-05,
      "loss": 0.1107,
      "step": 1380
    },
    {
      "epoch": 1.311939594148183,
      "grad_norm": 0.09895358395270354,
      "learning_rate": 1.248709639750228e-05,
      "loss": 0.1122,
      "step": 1390
    },
    {
      "epoch": 1.321378008494573,
      "grad_norm": 0.10215738990006902,
      "learning_rate": 1.2388206147576796e-05,
      "loss": 0.1124,
      "step": 1400
    },
    {
      "epoch": 1.3308164228409627,
      "grad_norm": 0.09611638665301472,
      "learning_rate": 1.2289067593261358e-05,
      "loss": 0.1151,
      "step": 1410
    },
    {
      "epoch": 1.3402548371873526,
      "grad_norm": 0.09899073401009041,
      "learning_rate": 1.2189691042099265e-05,
      "loss": 0.1124,
      "step": 1420
    },
    {
      "epoch": 1.3496932515337423,
      "grad_norm": 0.1157109248340035,
      "learning_rate": 1.209008682637859e-05,
      "loss": 0.1154,
      "step": 1430
    },
    {
      "epoch": 1.359131665880132,
      "grad_norm": 0.09358448441833775,
      "learning_rate": 1.1990265302057948e-05,
      "loss": 0.1127,
      "step": 1440
    },
    {
      "epoch": 1.368570080226522,
      "grad_norm": 0.10318474117014907,
      "learning_rate": 1.1890236847689762e-05,
      "loss": 0.1134,
      "step": 1450
    },
    {
      "epoch": 1.3780084945729119,
      "grad_norm": 0.10507403584009326,
      "learning_rate": 1.1790011863341197e-05,
      "loss": 0.1145,
      "step": 1460
    },
    {
      "epoch": 1.3874469089193016,
      "grad_norm": 0.09210160678217687,
      "learning_rate": 1.1689600769512855e-05,
      "loss": 0.1128,
      "step": 1470
    },
    {
      "epoch": 1.3968853232656913,
      "grad_norm": 0.09890271495599744,
      "learning_rate": 1.1589014006055337e-05,
      "loss": 0.1158,
      "step": 1480
    },
    {
      "epoch": 1.4063237376120812,
      "grad_norm": 0.09768042621781026,
      "learning_rate": 1.1488262031083816e-05,
      "loss": 0.1107,
      "step": 1490
    },
    {
      "epoch": 1.415762151958471,
      "grad_norm": 0.09501221720590393,
      "learning_rate": 1.1387355319890685e-05,
      "loss": 0.1138,
      "step": 1500
    },
    {
      "epoch": 1.4252005663048608,
      "grad_norm": 0.09098951194154016,
      "learning_rate": 1.1286304363856418e-05,
      "loss": 0.112,
      "step": 1510
    },
    {
      "epoch": 1.4346389806512505,
      "grad_norm": 0.08856225785605727,
      "learning_rate": 1.1185119669358792e-05,
      "loss": 0.1137,
      "step": 1520
    },
    {
      "epoch": 1.4440773949976404,
      "grad_norm": 0.09111265321801558,
      "learning_rate": 1.1083811756680523e-05,
      "loss": 0.1093,
      "step": 1530
    },
    {
      "epoch": 1.4535158093440301,
      "grad_norm": 0.09329783759350743,
      "learning_rate": 1.0982391158915441e-05,
      "loss": 0.1138,
      "step": 1540
    },
    {
      "epoch": 1.46295422369042,
      "grad_norm": 0.09114905230583735,
      "learning_rate": 1.0880868420873375e-05,
      "loss": 0.1135,
      "step": 1550
    },
    {
      "epoch": 1.4723926380368098,
      "grad_norm": 0.10049302399783284,
      "learning_rate": 1.0779254097983788e-05,
      "loss": 0.1104,
      "step": 1560
    },
    {
      "epoch": 1.4818310523831997,
      "grad_norm": 0.08811790258148439,
      "learning_rate": 1.0677558755198327e-05,
      "loss": 0.114,
      "step": 1570
    },
    {
      "epoch": 1.4912694667295894,
      "grad_norm": 0.09402068008649977,
      "learning_rate": 1.0575792965892349e-05,
      "loss": 0.1112,
      "step": 1580
    },
    {
      "epoch": 1.500707881075979,
      "grad_norm": 0.09144293350756144,
      "learning_rate": 1.0473967310765629e-05,
      "loss": 0.1099,
      "step": 1590
    },
    {
      "epoch": 1.510146295422369,
      "grad_norm": 0.08887643971229772,
      "learning_rate": 1.0372092376742247e-05,
      "loss": 0.1109,
      "step": 1600
    },
    {
      "epoch": 1.519584709768759,
      "grad_norm": 0.09042876745354687,
      "learning_rate": 1.0270178755869861e-05,
      "loss": 0.1123,
      "step": 1610
    },
    {
      "epoch": 1.5290231241151486,
      "grad_norm": 0.08799872003450031,
      "learning_rate": 1.0168237044218452e-05,
      "loss": 0.1088,
      "step": 1620
    },
    {
      "epoch": 1.5384615384615383,
      "grad_norm": 0.08558681617619375,
      "learning_rate": 1.0066277840778626e-05,
      "loss": 0.1125,
      "step": 1630
    },
    {
      "epoch": 1.5478999528079282,
      "grad_norm": 0.08917702916102198,
      "learning_rate": 9.964311746359631e-06,
      "loss": 0.1078,
      "step": 1640
    },
    {
      "epoch": 1.5573383671543182,
      "grad_norm": 0.09365014825092945,
      "learning_rate": 9.862349362487172e-06,
      "loss": 0.108,
      "step": 1650
    },
    {
      "epoch": 1.5667767815007079,
      "grad_norm": 0.08881624424784158,
      "learning_rate": 9.760401290301164e-06,
      "loss": 0.1073,
      "step": 1660
    },
    {
      "epoch": 1.5762151958470976,
      "grad_norm": 0.09040629927788134,
      "learning_rate": 9.658478129453532e-06,
      "loss": 0.1095,
      "step": 1670
    },
    {
      "epoch": 1.5856536101934875,
      "grad_norm": 0.09668654356646131,
      "learning_rate": 9.556590477006123e-06,
      "loss": 0.109,
      "step": 1680
    },
    {
      "epoch": 1.5950920245398774,
      "grad_norm": 0.08945527030759594,
      "learning_rate": 9.454748926328962e-06,
      "loss": 0.1111,
      "step": 1690
    },
    {
      "epoch": 1.6045304388862671,
      "grad_norm": 0.09096700102455489,
      "learning_rate": 9.352964065998801e-06,
      "loss": 0.1091,
      "step": 1700
    },
    {
      "epoch": 1.6139688532326568,
      "grad_norm": 0.098227701202526,
      "learning_rate": 9.251246478698242e-06,
      "loss": 0.1124,
      "step": 1710
    },
    {
      "epoch": 1.6234072675790467,
      "grad_norm": 0.08702267003826049,
      "learning_rate": 9.149606740115444e-06,
      "loss": 0.1091,
      "step": 1720
    },
    {
      "epoch": 1.6328456819254367,
      "grad_norm": 0.0912741057496456,
      "learning_rate": 9.04805541784454e-06,
      "loss": 0.1084,
      "step": 1730
    },
    {
      "epoch": 1.6422840962718264,
      "grad_norm": 0.09473199957469115,
      "learning_rate": 8.946603070286926e-06,
      "loss": 0.1071,
      "step": 1740
    },
    {
      "epoch": 1.651722510618216,
      "grad_norm": 0.09786418318351309,
      "learning_rate": 8.845260245553493e-06,
      "loss": 0.1106,
      "step": 1750
    },
    {
      "epoch": 1.661160924964606,
      "grad_norm": 0.09376970400308367,
      "learning_rate": 8.744037480367922e-06,
      "loss": 0.1095,
      "step": 1760
    },
    {
      "epoch": 1.670599339310996,
      "grad_norm": 0.09927967174299174,
      "learning_rate": 8.642945298971168e-06,
      "loss": 0.1086,
      "step": 1770
    },
    {
      "epoch": 1.6800377536573856,
      "grad_norm": 0.08941325714107755,
      "learning_rate": 8.54199421202726e-06,
      "loss": 0.1096,
      "step": 1780
    },
    {
      "epoch": 1.6894761680037753,
      "grad_norm": 0.09076378690689199,
      "learning_rate": 8.441194715530472e-06,
      "loss": 0.111,
      "step": 1790
    },
    {
      "epoch": 1.6989145823501652,
      "grad_norm": 0.08350889813597084,
      "learning_rate": 8.340557289714055e-06,
      "loss": 0.1089,
      "step": 1800
    },
    {
      "epoch": 1.708352996696555,
      "grad_norm": 0.1102767062542333,
      "learning_rate": 8.240092397960601e-06,
      "loss": 0.1077,
      "step": 1810
    },
    {
      "epoch": 1.7177914110429446,
      "grad_norm": 0.0928092514701947,
      "learning_rate": 8.139810485714142e-06,
      "loss": 0.109,
      "step": 1820
    },
    {
      "epoch": 1.7272298253893346,
      "grad_norm": 0.08827741386544802,
      "learning_rate": 8.03972197939414e-06,
      "loss": 0.1103,
      "step": 1830
    },
    {
      "epoch": 1.7366682397357245,
      "grad_norm": 0.09074532027658296,
      "learning_rate": 7.939837285311425e-06,
      "loss": 0.106,
      "step": 1840
    },
    {
      "epoch": 1.7461066540821142,
      "grad_norm": 0.08863060213083432,
      "learning_rate": 7.840166788586244e-06,
      "loss": 0.1111,
      "step": 1850
    },
    {
      "epoch": 1.7555450684285039,
      "grad_norm": 0.0873194235737418,
      "learning_rate": 7.740720852068524e-06,
      "loss": 0.1107,
      "step": 1860
    },
    {
      "epoch": 1.7649834827748938,
      "grad_norm": 0.09463118593541094,
      "learning_rate": 7.641509815260412e-06,
      "loss": 0.1067,
      "step": 1870
    },
    {
      "epoch": 1.7744218971212837,
      "grad_norm": 0.08277999614215281,
      "learning_rate": 7.542543993241278e-06,
      "loss": 0.1092,
      "step": 1880
    },
    {
      "epoch": 1.7838603114676734,
      "grad_norm": 0.08350764579820083,
      "learning_rate": 7.443833675595254e-06,
      "loss": 0.1033,
      "step": 1890
    },
    {
      "epoch": 1.7932987258140631,
      "grad_norm": 0.0892168369121696,
      "learning_rate": 7.3453891253413935e-06,
      "loss": 0.1088,
      "step": 1900
    },
    {
      "epoch": 1.802737140160453,
      "grad_norm": 0.09068391166704463,
      "learning_rate": 7.247220577866625e-06,
      "loss": 0.1074,
      "step": 1910
    },
    {
      "epoch": 1.812175554506843,
      "grad_norm": 0.09208367707491026,
      "learning_rate": 7.149338239861579e-06,
      "loss": 0.1069,
      "step": 1920
    },
    {
      "epoch": 1.8216139688532327,
      "grad_norm": 0.09334448658561058,
      "learning_rate": 7.051752288259366e-06,
      "loss": 0.1051,
      "step": 1930
    },
    {
      "epoch": 1.8310523831996224,
      "grad_norm": 0.0867013966015152,
      "learning_rate": 6.954472869177479e-06,
      "loss": 0.1071,
      "step": 1940
    },
    {
      "epoch": 1.8404907975460123,
      "grad_norm": 0.08513824070105314,
      "learning_rate": 6.857510096862901e-06,
      "loss": 0.108,
      "step": 1950
    },
    {
      "epoch": 1.8499292118924022,
      "grad_norm": 0.08880515925379688,
      "learning_rate": 6.760874052640494e-06,
      "loss": 0.1081,
      "step": 1960
    },
    {
      "epoch": 1.859367626238792,
      "grad_norm": 0.09395118992476542,
      "learning_rate": 6.664574783864862e-06,
      "loss": 0.1079,
      "step": 1970
    },
    {
      "epoch": 1.8688060405851816,
      "grad_norm": 0.09253843263366972,
      "learning_rate": 6.568622302875682e-06,
      "loss": 0.1068,
      "step": 1980
    },
    {
      "epoch": 1.8782444549315715,
      "grad_norm": 0.09103342170778085,
      "learning_rate": 6.473026585956736e-06,
      "loss": 0.106,
      "step": 1990
    },
    {
      "epoch": 1.8876828692779613,
      "grad_norm": 0.08266259287550586,
      "learning_rate": 6.377797572298661e-06,
      "loss": 0.1076,
      "step": 2000
    },
    {
      "epoch": 1.897121283624351,
      "grad_norm": 0.08360823987901486,
      "learning_rate": 6.282945162965548e-06,
      "loss": 0.1079,
      "step": 2010
    },
    {
      "epoch": 1.9065596979707409,
      "grad_norm": 0.09004863363551058,
      "learning_rate": 6.188479219865529e-06,
      "loss": 0.1064,
      "step": 2020
    },
    {
      "epoch": 1.9159981123171308,
      "grad_norm": 0.0931263680297109,
      "learning_rate": 6.094409564725435e-06,
      "loss": 0.1054,
      "step": 2030
    },
    {
      "epoch": 1.9254365266635205,
      "grad_norm": 0.09051504504706874,
      "learning_rate": 6.0007459780695885e-06,
      "loss": 0.1082,
      "step": 2040
    },
    {
      "epoch": 1.9348749410099102,
      "grad_norm": 0.0904194328886728,
      "learning_rate": 5.907498198202939e-06,
      "loss": 0.1081,
      "step": 2050
    },
    {
      "epoch": 1.9443133553563001,
      "grad_norm": 0.08711490496456058,
      "learning_rate": 5.8146759201985525e-06,
      "loss": 0.1069,
      "step": 2060
    },
    {
      "epoch": 1.95375176970269,
      "grad_norm": 0.08686808156213838,
      "learning_rate": 5.722288794889603e-06,
      "loss": 0.1064,
      "step": 2070
    },
    {
      "epoch": 1.9631901840490797,
      "grad_norm": 0.08817123970600604,
      "learning_rate": 5.630346427865965e-06,
      "loss": 0.1045,
      "step": 2080
    },
    {
      "epoch": 1.9726285983954694,
      "grad_norm": 0.08339231796976605,
      "learning_rate": 5.538858378475508e-06,
      "loss": 0.1066,
      "step": 2090
    },
    {
      "epoch": 1.9820670127418594,
      "grad_norm": 0.08824324960180577,
      "learning_rate": 5.447834158830202e-06,
      "loss": 0.1037,
      "step": 2100
    },
    {
      "epoch": 1.9915054270882493,
      "grad_norm": 0.09311929590923752,
      "learning_rate": 5.357283232817147e-06,
      "loss": 0.1054,
      "step": 2110
    },
    {
      "epoch": 2.0009438414346388,
      "grad_norm": 0.1334431719993605,
      "learning_rate": 5.267215015114574e-06,
      "loss": 0.1031,
      "step": 2120
    },
    {
      "epoch": 2.0103822557810287,
      "grad_norm": 0.10361861757914052,
      "learning_rate": 5.177638870213008e-06,
      "loss": 0.0868,
      "step": 2130
    },
    {
      "epoch": 2.0198206701274186,
      "grad_norm": 0.084172782289755,
      "learning_rate": 5.088564111441645e-06,
      "loss": 0.0834,
      "step": 2140
    },
    {
      "epoch": 2.0292590844738085,
      "grad_norm": 0.08727410339549913,
      "learning_rate": 5.000000000000003e-06,
      "loss": 0.0852,
      "step": 2150
    },
    {
      "epoch": 2.038697498820198,
      "grad_norm": 0.08994662846331379,
      "learning_rate": 4.911955743995042e-06,
      "loss": 0.0845,
      "step": 2160
    },
    {
      "epoch": 2.048135913166588,
      "grad_norm": 0.08715835430152602,
      "learning_rate": 4.824440497483802e-06,
      "loss": 0.0847,
      "step": 2170
    },
    {
      "epoch": 2.057574327512978,
      "grad_norm": 0.0943214139106939,
      "learning_rate": 4.737463359521618e-06,
      "loss": 0.0845,
      "step": 2180
    },
    {
      "epoch": 2.067012741859368,
      "grad_norm": 0.09049982951899538,
      "learning_rate": 4.6510333732160915e-06,
      "loss": 0.085,
      "step": 2190
    },
    {
      "epoch": 2.0764511562057573,
      "grad_norm": 0.08823142440331375,
      "learning_rate": 4.565159524786888e-06,
      "loss": 0.0867,
      "step": 2200
    },
    {
      "epoch": 2.085889570552147,
      "grad_norm": 0.08727162151071796,
      "learning_rate": 4.479850742631396e-06,
      "loss": 0.0834,
      "step": 2210
    },
    {
      "epoch": 2.095327984898537,
      "grad_norm": 0.08243182067038303,
      "learning_rate": 4.395115896396457e-06,
      "loss": 0.0849,
      "step": 2220
    },
    {
      "epoch": 2.104766399244927,
      "grad_norm": 0.08720723034547441,
      "learning_rate": 4.310963796056168e-06,
      "loss": 0.084,
      "step": 2230
    },
    {
      "epoch": 2.1142048135913165,
      "grad_norm": 0.08588933137845103,
      "learning_rate": 4.227403190995901e-06,
      "loss": 0.0875,
      "step": 2240
    },
    {
      "epoch": 2.1236432279377064,
      "grad_norm": 0.09200761679022347,
      "learning_rate": 4.14444276910263e-06,
      "loss": 0.0853,
      "step": 2250
    },
    {
      "epoch": 2.1330816422840964,
      "grad_norm": 0.08831298949051568,
      "learning_rate": 4.06209115586162e-06,
      "loss": 0.0867,
      "step": 2260
    },
    {
      "epoch": 2.1425200566304863,
      "grad_norm": 0.0893828115241757,
      "learning_rate": 3.980356913459642e-06,
      "loss": 0.0865,
      "step": 2270
    },
    {
      "epoch": 2.1519584709768758,
      "grad_norm": 0.09119946740323005,
      "learning_rate": 3.899248539894756e-06,
      "loss": 0.0848,
      "step": 2280
    },
    {
      "epoch": 2.1613968853232657,
      "grad_norm": 0.08882008929472095,
      "learning_rate": 3.818774468092754e-06,
      "loss": 0.0843,
      "step": 2290
    },
    {
      "epoch": 2.1708352996696556,
      "grad_norm": 0.08922739717614447,
      "learning_rate": 3.738943065030376e-06,
      "loss": 0.0811,
      "step": 2300
    },
    {
      "epoch": 2.180273714016045,
      "grad_norm": 0.08622476744102522,
      "learning_rate": 3.659762630865411e-06,
      "loss": 0.083,
      "step": 2310
    },
    {
      "epoch": 2.189712128362435,
      "grad_norm": 0.08399251697806781,
      "learning_rate": 3.5812413980736916e-06,
      "loss": 0.0827,
      "step": 2320
    },
    {
      "epoch": 2.199150542708825,
      "grad_norm": 0.09006770579644241,
      "learning_rate": 3.5033875305931662e-06,
      "loss": 0.0849,
      "step": 2330
    },
    {
      "epoch": 2.208588957055215,
      "grad_norm": 0.08747795506814363,
      "learning_rate": 3.4262091229750973e-06,
      "loss": 0.0822,
      "step": 2340
    },
    {
      "epoch": 2.2180273714016043,
      "grad_norm": 0.08840099962821243,
      "learning_rate": 3.3497141995424397e-06,
      "loss": 0.0835,
      "step": 2350
    },
    {
      "epoch": 2.2274657857479943,
      "grad_norm": 0.08986511740506226,
      "learning_rate": 3.2739107135555603e-06,
      "loss": 0.0841,
      "step": 2360
    },
    {
      "epoch": 2.236904200094384,
      "grad_norm": 0.08924913381765429,
      "learning_rate": 3.1988065463853204e-06,
      "loss": 0.0849,
      "step": 2370
    },
    {
      "epoch": 2.246342614440774,
      "grad_norm": 0.08622955784811655,
      "learning_rate": 3.1244095066936396e-06,
      "loss": 0.0848,
      "step": 2380
    },
    {
      "epoch": 2.2557810287871636,
      "grad_norm": 0.08819044440789944,
      "learning_rate": 3.050727329621637e-06,
      "loss": 0.0835,
      "step": 2390
    },
    {
      "epoch": 2.2652194431335535,
      "grad_norm": 0.08900963097417651,
      "learning_rate": 2.977767675985377e-06,
      "loss": 0.0805,
      "step": 2400
    },
    {
      "epoch": 2.2746578574799434,
      "grad_norm": 0.0889820260869723,
      "learning_rate": 2.905538131479376e-06,
      "loss": 0.0844,
      "step": 2410
    },
    {
      "epoch": 2.2840962718263333,
      "grad_norm": 0.08508956855605507,
      "learning_rate": 2.8340462058879214e-06,
      "loss": 0.082,
      "step": 2420
    },
    {
      "epoch": 2.293534686172723,
      "grad_norm": 0.0859290434178059,
      "learning_rate": 2.76329933230425e-06,
      "loss": 0.0819,
      "step": 2430
    },
    {
      "epoch": 2.3029731005191127,
      "grad_norm": 0.08310233373376713,
      "learning_rate": 2.6933048663577297e-06,
      "loss": 0.0811,
      "step": 2440
    },
    {
      "epoch": 2.3124115148655027,
      "grad_norm": 0.08745596999299074,
      "learning_rate": 2.6240700854490988e-06,
      "loss": 0.0824,
      "step": 2450
    },
    {
      "epoch": 2.3218499292118926,
      "grad_norm": 0.08451577327209396,
      "learning_rate": 2.5556021879938074e-06,
      "loss": 0.0828,
      "step": 2460
    },
    {
      "epoch": 2.331288343558282,
      "grad_norm": 0.09129945645731877,
      "learning_rate": 2.4879082926735974e-06,
      "loss": 0.0837,
      "step": 2470
    },
    {
      "epoch": 2.340726757904672,
      "grad_norm": 0.083910957600724,
      "learning_rate": 2.4209954376963797e-06,
      "loss": 0.0816,
      "step": 2480
    },
    {
      "epoch": 2.350165172251062,
      "grad_norm": 0.08447673558716574,
      "learning_rate": 2.354870580064439e-06,
      "loss": 0.0808,
      "step": 2490
    },
    {
      "epoch": 2.3596035865974514,
      "grad_norm": 0.08671014168934867,
      "learning_rate": 2.289540594851122e-06,
      "loss": 0.0814,
      "step": 2500
    },
    {
      "epoch": 2.3690420009438413,
      "grad_norm": 0.08155772855747594,
      "learning_rate": 2.225012274486028e-06,
      "loss": 0.0791,
      "step": 2510
    },
    {
      "epoch": 2.3784804152902312,
      "grad_norm": 0.08555794455636312,
      "learning_rate": 2.1612923280487883e-06,
      "loss": 0.0843,
      "step": 2520
    },
    {
      "epoch": 2.387918829636621,
      "grad_norm": 0.08566290155470521,
      "learning_rate": 2.0983873805715216e-06,
      "loss": 0.0837,
      "step": 2530
    },
    {
      "epoch": 2.397357243983011,
      "grad_norm": 0.08531077534490512,
      "learning_rate": 2.0363039723500155e-06,
      "loss": 0.0838,
      "step": 2540
    },
    {
      "epoch": 2.4067956583294006,
      "grad_norm": 0.08375539552341336,
      "learning_rate": 1.9750485582637245e-06,
      "loss": 0.0822,
      "step": 2550
    },
    {
      "epoch": 2.4162340726757905,
      "grad_norm": 0.08503293134727295,
      "learning_rate": 1.9146275071046626e-06,
      "loss": 0.0849,
      "step": 2560
    },
    {
      "epoch": 2.4256724870221804,
      "grad_norm": 0.08165485153496227,
      "learning_rate": 1.8550471009152138e-06,
      "loss": 0.0803,
      "step": 2570
    },
    {
      "epoch": 2.43511090136857,
      "grad_norm": 0.08309504205099255,
      "learning_rate": 1.7963135343349914e-06,
      "loss": 0.0789,
      "step": 2580
    },
    {
      "epoch": 2.44454931571496,
      "grad_norm": 0.08294260063335909,
      "learning_rate": 1.73843291395678e-06,
      "loss": 0.0823,
      "step": 2590
    },
    {
      "epoch": 2.4539877300613497,
      "grad_norm": 0.08537750016844735,
      "learning_rate": 1.6814112576916142e-06,
      "loss": 0.0825,
      "step": 2600
    },
    {
      "epoch": 2.4634261444077397,
      "grad_norm": 0.08433197560779272,
      "learning_rate": 1.6252544941430982e-06,
      "loss": 0.0813,
      "step": 2610
    },
    {
      "epoch": 2.472864558754129,
      "grad_norm": 0.08321923421106348,
      "learning_rate": 1.5699684619909983e-06,
      "loss": 0.0826,
      "step": 2620
    },
    {
      "epoch": 2.482302973100519,
      "grad_norm": 0.08373732680479463,
      "learning_rate": 1.5155589093841939e-06,
      "loss": 0.0802,
      "step": 2630
    },
    {
      "epoch": 2.491741387446909,
      "grad_norm": 0.09112397953039225,
      "learning_rate": 1.4620314933430269e-06,
      "loss": 0.081,
      "step": 2640
    },
    {
      "epoch": 2.501179801793299,
      "grad_norm": 0.08246005822696881,
      "learning_rate": 1.4093917791711497e-06,
      "loss": 0.0808,
      "step": 2650
    },
    {
      "epoch": 2.5106182161396884,
      "grad_norm": 0.08399111524479627,
      "learning_rate": 1.357645239876879e-06,
      "loss": 0.08,
      "step": 2660
    },
    {
      "epoch": 2.5200566304860783,
      "grad_norm": 0.08520663335129833,
      "learning_rate": 1.3067972556041753e-06,
      "loss": 0.0818,
      "step": 2670
    },
    {
      "epoch": 2.5294950448324682,
      "grad_norm": 0.08299784168003543,
      "learning_rate": 1.2568531130732498e-06,
      "loss": 0.0807,
      "step": 2680
    },
    {
      "epoch": 2.5389334591788577,
      "grad_norm": 0.08594453806774255,
      "learning_rate": 1.207818005030904e-06,
      "loss": 0.0802,
      "step": 2690
    },
    {
      "epoch": 2.5483718735252476,
      "grad_norm": 0.08039958287229476,
      "learning_rate": 1.1596970297106458e-06,
      "loss": 0.0818,
      "step": 2700
    },
    {
      "epoch": 2.5578102878716376,
      "grad_norm": 0.08085895400758615,
      "learning_rate": 1.1124951903025981e-06,
      "loss": 0.0806,
      "step": 2710
    },
    {
      "epoch": 2.5672487022180275,
      "grad_norm": 0.08458215660632902,
      "learning_rate": 1.0662173944333288e-06,
      "loss": 0.081,
      "step": 2720
    },
    {
      "epoch": 2.5766871165644174,
      "grad_norm": 0.0826446041832577,
      "learning_rate": 1.0208684536555968e-06,
      "loss": 0.081,
      "step": 2730
    },
    {
      "epoch": 2.586125530910807,
      "grad_norm": 0.08442284791430928,
      "learning_rate": 9.764530829480822e-07,
      "loss": 0.0832,
      "step": 2740
    },
    {
      "epoch": 2.595563945257197,
      "grad_norm": 0.08336636617446731,
      "learning_rate": 9.329759002251726e-07,
      "loss": 0.0802,
      "step": 2750
    },
    {
      "epoch": 2.6050023596035867,
      "grad_norm": 0.0846992268529094,
      "learning_rate": 8.904414258568306e-07,
      "loss": 0.0799,
      "step": 2760
    },
    {
      "epoch": 2.614440773949976,
      "grad_norm": 0.08300701320734614,
      "learning_rate": 8.488540821986035e-07,
      "loss": 0.0827,
      "step": 2770
    },
    {
      "epoch": 2.623879188296366,
      "grad_norm": 0.08443465909555346,
      "learning_rate": 8.082181931318311e-07,
      "loss": 0.0792,
      "step": 2780
    },
    {
      "epoch": 2.633317602642756,
      "grad_norm": 0.08476186997139065,
      "learning_rate": 7.685379836140872e-07,
      "loss": 0.079,
      "step": 2790
    },
    {
      "epoch": 2.642756016989146,
      "grad_norm": 0.08472149270611445,
      "learning_rate": 7.298175792398976e-07,
      "loss": 0.0818,
      "step": 2800
    },
    {
      "epoch": 2.652194431335536,
      "grad_norm": 0.08032205169166948,
      "learning_rate": 6.920610058118105e-07,
      "loss": 0.0804,
      "step": 2810
    },
    {
      "epoch": 2.6616328456819254,
      "grad_norm": 0.08367697987102687,
      "learning_rate": 6.552721889218194e-07,
      "loss": 0.0816,
      "step": 2820
    },
    {
      "epoch": 2.6710712600283153,
      "grad_norm": 0.08668064800964889,
      "learning_rate": 6.194549535432137e-07,
      "loss": 0.08,
      "step": 2830
    },
    {
      "epoch": 2.680509674374705,
      "grad_norm": 0.0828804628433115,
      "learning_rate": 5.846130236329073e-07,
      "loss": 0.0823,
      "step": 2840
    },
    {
      "epoch": 2.6899480887210947,
      "grad_norm": 0.08378018124842795,
      "learning_rate": 5.507500217442341e-07,
      "loss": 0.0809,
      "step": 2850
    },
    {
      "epoch": 2.6993865030674846,
      "grad_norm": 0.08220414182971268,
      "learning_rate": 5.178694686503205e-07,
      "loss": 0.0784,
      "step": 2860
    },
    {
      "epoch": 2.7088249174138745,
      "grad_norm": 0.08076213129882256,
      "learning_rate": 4.85974782978027e-07,
      "loss": 0.081,
      "step": 2870
    },
    {
      "epoch": 2.718263331760264,
      "grad_norm": 0.08189865863600006,
      "learning_rate": 4.5506928085250033e-07,
      "loss": 0.0778,
      "step": 2880
    },
    {
      "epoch": 2.727701746106654,
      "grad_norm": 0.08777777733586507,
      "learning_rate": 4.251561755524036e-07,
      "loss": 0.0832,
      "step": 2890
    },
    {
      "epoch": 2.737140160453044,
      "grad_norm": 0.08438414185840341,
      "learning_rate": 3.9623857717581813e-07,
      "loss": 0.0808,
      "step": 2900
    },
    {
      "epoch": 2.746578574799434,
      "grad_norm": 0.08245920785308986,
      "learning_rate": 3.6831949231689203e-07,
      "loss": 0.081,
      "step": 2910
    },
    {
      "epoch": 2.7560169891458237,
      "grad_norm": 0.08356304284398074,
      "learning_rate": 3.414018237532335e-07,
      "loss": 0.0821,
      "step": 2920
    },
    {
      "epoch": 2.765455403492213,
      "grad_norm": 0.08633702580648989,
      "learning_rate": 3.154883701441136e-07,
      "loss": 0.08,
      "step": 2930
    },
    {
      "epoch": 2.774893817838603,
      "grad_norm": 0.0821243443006684,
      "learning_rate": 2.905818257394799e-07,
      "loss": 0.0798,
      "step": 2940
    },
    {
      "epoch": 2.784332232184993,
      "grad_norm": 0.08233971677127812,
      "learning_rate": 2.666847800998362e-07,
      "loss": 0.0819,
      "step": 2950
    },
    {
      "epoch": 2.7937706465313825,
      "grad_norm": 0.08176593022884952,
      "learning_rate": 2.437997178270035e-07,
      "loss": 0.0807,
      "step": 2960
    },
    {
      "epoch": 2.8032090608777724,
      "grad_norm": 0.08485236840414098,
      "learning_rate": 2.219290183057865e-07,
      "loss": 0.0806,
      "step": 2970
    },
    {
      "epoch": 2.8126474752241624,
      "grad_norm": 0.08134044819035506,
      "learning_rate": 2.0107495545659829e-07,
      "loss": 0.0778,
      "step": 2980
    },
    {
      "epoch": 2.8220858895705523,
      "grad_norm": 0.08200978384685277,
      "learning_rate": 1.8123969749902714e-07,
      "loss": 0.0777,
      "step": 2990
    },
    {
      "epoch": 2.831524303916942,
      "grad_norm": 0.08409107175117113,
      "learning_rate": 1.6242530672641143e-07,
      "loss": 0.0813,
      "step": 3000
    },
    {
      "epoch": 2.8409627182633317,
      "grad_norm": 0.08288246905003016,
      "learning_rate": 1.4463373929141766e-07,
      "loss": 0.0788,
      "step": 3010
    },
    {
      "epoch": 2.8504011326097216,
      "grad_norm": 0.08212980978941326,
      "learning_rate": 1.2786684500265546e-07,
      "loss": 0.0819,
      "step": 3020
    },
    {
      "epoch": 2.8598395469561115,
      "grad_norm": 0.0832626354470357,
      "learning_rate": 1.1212636713235581e-07,
      "loss": 0.0794,
      "step": 3030
    },
    {
      "epoch": 2.869277961302501,
      "grad_norm": 0.0836378792472762,
      "learning_rate": 9.741394223512057e-08,
      "loss": 0.0814,
      "step": 3040
    },
    {
      "epoch": 2.878716375648891,
      "grad_norm": 0.08017963643030881,
      "learning_rate": 8.373109997776185e-08,
      "loss": 0.0804,
      "step": 3050
    },
    {
      "epoch": 2.888154789995281,
      "grad_norm": 0.08105963755024889,
      "learning_rate": 7.10792629802659e-08,
      "loss": 0.08,
      "step": 3060
    },
    {
      "epoch": 2.8975932043416703,
      "grad_norm": 0.08196798527897163,
      "learning_rate": 5.945974666788479e-08,
      "loss": 0.0809,
      "step": 3070
    },
    {
      "epoch": 2.9070316186880603,
      "grad_norm": 0.08445010118063752,
      "learning_rate": 4.887375913436132e-08,
      "loss": 0.0817,
      "step": 3080
    },
    {
      "epoch": 2.91647003303445,
      "grad_norm": 0.0830340739185833,
      "learning_rate": 3.932240101633178e-08,
      "loss": 0.0823,
      "step": 3090
    },
    {
      "epoch": 2.92590844738084,
      "grad_norm": 0.08107816362561618,
      "learning_rate": 3.0806665378884106e-08,
      "loss": 0.0764,
      "step": 3100
    },
    {
      "epoch": 2.93534686172723,
      "grad_norm": 0.08498347611137054,
      "learning_rate": 2.33274376123116e-08,
      "loss": 0.0784,
      "step": 3110
    },
    {
      "epoch": 2.9447852760736195,
      "grad_norm": 0.08214101571060746,
      "learning_rate": 1.68854953400599e-08,
      "loss": 0.0795,
      "step": 3120
    },
    {
      "epoch": 2.9542236904200094,
      "grad_norm": 0.0849230883706386,
      "learning_rate": 1.1481508337869429e-08,
      "loss": 0.0801,
      "step": 3130
    },
    {
      "epoch": 2.9636621047663994,
      "grad_norm": 0.09022687172778407,
      "learning_rate": 7.1160384641455475e-09,
      "loss": 0.0828,
      "step": 3140
    },
    {
      "epoch": 2.973100519112789,
      "grad_norm": 0.08399540827662202,
      "learning_rate": 3.7895396015374955e-09,
      "loss": 0.0808,
      "step": 3150
    },
    {
      "epoch": 2.9825389334591788,
      "grad_norm": 0.08027494442622467,
      "learning_rate": 1.502357609749483e-09,
      "loss": 0.0803,
      "step": 3160
    },
    {
      "epoch": 2.9919773478055687,
      "grad_norm": 0.08480380742956374,
      "learning_rate": 2.5473028957945234e-10,
      "loss": 0.0813,
      "step": 3170
    },
    {
      "epoch": 2.9985842378480414,
      "step": 3177,
      "total_flos": 5.875342281120154e+16,
      "train_loss": 0.11385704865424108,
      "train_runtime": 162519.6807,
      "train_samples_per_second": 3.441,
      "train_steps_per_second": 0.02
    }
  ],
  "logging_steps": 10,
  "max_steps": 3177,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 50000000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 5.875342281120154e+16,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}