{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 1170,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.025682182985553772,
      "grad_norm": 41.89281463623047,
      "learning_rate": 5e-05,
      "loss": 41.062384033203124,
      "step": 10
    },
    {
      "epoch": 0.051364365971107544,
      "grad_norm": 19.25795555114746,
      "learning_rate": 0.00010555555555555557,
      "loss": 23.114556884765626,
      "step": 20
    },
    {
      "epoch": 0.07704654895666131,
      "grad_norm": 15.17092514038086,
      "learning_rate": 0.0001611111111111111,
      "loss": 19.892947387695312,
      "step": 30
    },
    {
      "epoch": 0.10272873194221509,
      "grad_norm": 12.6154146194458,
      "learning_rate": 0.0001999965463076377,
      "loss": 18.954251098632813,
      "step": 40
    },
    {
      "epoch": 0.12841091492776885,
      "grad_norm": 13.695039749145508,
      "learning_rate": 0.00019993515396856082,
      "loss": 18.60211944580078,
      "step": 50
    },
    {
      "epoch": 0.15409309791332262,
      "grad_norm": 13.03475570678711,
      "learning_rate": 0.00019979706714271113,
      "loss": 17.00911102294922,
      "step": 60
    },
    {
      "epoch": 0.1797752808988764,
      "grad_norm": 12.657264709472656,
      "learning_rate": 0.0001995823918037908,
      "loss": 16.737196350097655,
      "step": 70
    },
    {
      "epoch": 0.20545746388443017,
      "grad_norm": 9.450530052185059,
      "learning_rate": 0.00019929129270278366,
      "loss": 17.192404174804686,
      "step": 80
    },
    {
      "epoch": 0.23113964686998395,
      "grad_norm": 11.038394927978516,
      "learning_rate": 0.0001989239932415185,
      "loss": 17.626002502441406,
      "step": 90
    },
    {
      "epoch": 0.2568218298555377,
      "grad_norm": 10.112431526184082,
      "learning_rate": 0.00019848077530122083,
      "loss": 17.375982666015624,
      "step": 100
    },
    {
      "epoch": 0.2825040128410915,
      "grad_norm": 10.951290130615234,
      "learning_rate": 0.0001979619790261853,
      "loss": 16.52630157470703,
      "step": 110
    },
    {
      "epoch": 0.30818619582664525,
      "grad_norm": 13.097779273986816,
      "learning_rate": 0.00019736800256273457,
      "loss": 18.660000610351563,
      "step": 120
    },
    {
      "epoch": 0.33386837881219905,
      "grad_norm": 10.489601135253906,
      "learning_rate": 0.00019669930175366472,
      "loss": 18.038082885742188,
      "step": 130
    },
    {
      "epoch": 0.3595505617977528,
      "grad_norm": 12.037221908569336,
      "learning_rate": 0.0001959563897884124,
      "loss": 16.976956176757813,
      "step": 140
    },
    {
      "epoch": 0.3852327447833066,
      "grad_norm": 13.149781227111816,
      "learning_rate": 0.00019513983680921108,
      "loss": 17.55983123779297,
      "step": 150
    },
    {
      "epoch": 0.41091492776886035,
      "grad_norm": 12.796188354492188,
      "learning_rate": 0.00019425026947353992,
      "loss": 17.86356201171875,
      "step": 160
    },
    {
      "epoch": 0.43659711075441415,
      "grad_norm": 14.400433540344238,
      "learning_rate": 0.0001932883704732001,
      "loss": 17.5066162109375,
      "step": 170
    },
    {
      "epoch": 0.4622792937399679,
      "grad_norm": 16.03416633605957,
      "learning_rate": 0.00019225487801038788,
      "loss": 18.057034301757813,
      "step": 180
    },
    {
      "epoch": 0.48796147672552165,
      "grad_norm": 11.53113842010498,
      "learning_rate": 0.00019115058523116733,
      "loss": 16.775308227539064,
      "step": 190
    },
    {
      "epoch": 0.5136436597110754,
      "grad_norm": 12.079059600830078,
      "learning_rate": 0.00018997633961677582,
      "loss": 17.352998352050783,
      "step": 200
    },
    {
      "epoch": 0.5393258426966292,
      "grad_norm": 11.70594310760498,
      "learning_rate": 0.00018873304233323122,
      "loss": 18.17870635986328,
      "step": 210
    },
    {
      "epoch": 0.565008025682183,
      "grad_norm": 13.156736373901367,
      "learning_rate": 0.00018742164753973855,
      "loss": 16.766079711914063,
      "step": 220
    },
    {
      "epoch": 0.5906902086677368,
      "grad_norm": 11.801448822021484,
      "learning_rate": 0.0001860431616564278,
      "loss": 17.98457489013672,
      "step": 230
    },
    {
      "epoch": 0.6163723916532905,
      "grad_norm": 11.434338569641113,
      "learning_rate": 0.0001845986425919841,
      "loss": 17.011558532714844,
      "step": 240
    },
    {
      "epoch": 0.6420545746388443,
      "grad_norm": 11.592355728149414,
      "learning_rate": 0.00018308919893176396,
      "loss": 18.073362731933592,
      "step": 250
    },
    {
      "epoch": 0.6677367576243981,
      "grad_norm": 12.140549659729004,
      "learning_rate": 0.00018151598908701947,
      "loss": 17.267474365234374,
      "step": 260
    },
    {
      "epoch": 0.6934189406099518,
      "grad_norm": 12.207261085510254,
      "learning_rate": 0.00017988022040588416,
      "loss": 17.970367431640625,
      "step": 270
    },
    {
      "epoch": 0.7191011235955056,
      "grad_norm": 12.024649620056152,
      "learning_rate": 0.000178183148246803,
      "loss": 18.595138549804688,
      "step": 280
    },
    {
      "epoch": 0.7447833065810594,
      "grad_norm": 10.953705787658691,
      "learning_rate": 0.0001764260750151167,
      "loss": 17.661044311523437,
      "step": 290
    },
    {
      "epoch": 0.7704654895666132,
      "grad_norm": 12.161933898925781,
      "learning_rate": 0.0001746103491635407,
      "loss": 17.323040771484376,
      "step": 300
    },
    {
      "epoch": 0.7961476725521669,
      "grad_norm": 14.1478271484375,
      "learning_rate": 0.00017273736415730488,
      "loss": 17.589439392089844,
      "step": 310
    },
    {
      "epoch": 0.8218298555377207,
      "grad_norm": 11.344724655151367,
      "learning_rate": 0.0001708085574047494,
      "loss": 18.087261962890626,
      "step": 320
    },
    {
      "epoch": 0.8475120385232745,
      "grad_norm": 11.49538803100586,
      "learning_rate": 0.00016882540915419623,
      "loss": 16.709014892578125,
      "step": 330
    },
    {
      "epoch": 0.8731942215088283,
      "grad_norm": 11.975954055786133,
      "learning_rate": 0.00016678944135794374,
      "loss": 18.784584045410156,
      "step": 340
    },
    {
      "epoch": 0.898876404494382,
      "grad_norm": 10.979130744934082,
      "learning_rate": 0.00016470221650425582,
      "loss": 17.035598754882812,
      "step": 350
    },
    {
      "epoch": 0.9245585874799358,
      "grad_norm": 10.86307144165039,
      "learning_rate": 0.00016256533641824177,
      "loss": 17.912120056152343,
      "step": 360
    },
    {
      "epoch": 0.9502407704654896,
      "grad_norm": 11.18655776977539,
      "learning_rate": 0.00016038044103254775,
      "loss": 16.4998046875,
      "step": 370
    },
    {
      "epoch": 0.9759229534510433,
      "grad_norm": 11.321154594421387,
      "learning_rate": 0.00015814920712880267,
      "loss": 17.580471801757813,
      "step": 380
    },
    {
      "epoch": 1.0,
      "grad_norm": 6.271613597869873,
      "learning_rate": 0.0001558733470507847,
      "loss": 15.328689575195312,
      "step": 390
    },
    {
      "epoch": 1.0256821829855538,
      "grad_norm": 13.181764602661133,
      "learning_rate": 0.00015355460739029586,
      "loss": 15.265965270996094,
      "step": 400
    },
    {
      "epoch": 1.0513643659711076,
      "grad_norm": 11.904574394226074,
      "learning_rate": 0.00015119476764675305,
      "loss": 14.833596801757812,
      "step": 410
    },
    {
      "epoch": 1.0770465489566614,
      "grad_norm": 12.352519989013672,
      "learning_rate": 0.0001487956388615247,
      "loss": 13.833314514160156,
      "step": 420
    },
    {
      "epoch": 1.102728731942215,
      "grad_norm": 15.478001594543457,
      "learning_rate": 0.00014635906222806058,
      "loss": 15.767561340332032,
      "step": 430
    },
    {
      "epoch": 1.1284109149277688,
      "grad_norm": 15.335821151733398,
      "learning_rate": 0.00014388690767888154,
      "loss": 14.462684631347656,
      "step": 440
    },
    {
      "epoch": 1.1540930979133226,
      "grad_norm": 16.043373107910156,
      "learning_rate": 0.00014138107245051392,
      "loss": 15.908651733398438,
      "step": 450
    },
    {
      "epoch": 1.1797752808988764,
      "grad_norm": 12.852351188659668,
      "learning_rate": 0.00013884347962746948,
      "loss": 15.758128356933593,
      "step": 460
    },
    {
      "epoch": 1.2054574638844302,
      "grad_norm": 15.320487022399902,
      "learning_rate": 0.00013627607666638858,
      "loss": 15.051063537597656,
      "step": 470
    },
    {
      "epoch": 1.231139646869984,
      "grad_norm": 16.30264663696289,
      "learning_rate": 0.00013368083390147913,
      "loss": 14.599794006347656,
      "step": 480
    },
    {
      "epoch": 1.2568218298555376,
      "grad_norm": 13.734286308288574,
      "learning_rate": 0.00013105974303239838,
      "loss": 14.052903747558593,
      "step": 490
    },
    {
      "epoch": 1.2825040128410916,
      "grad_norm": 15.704442024230957,
      "learning_rate": 0.0001284148155957379,
      "loss": 13.530386352539063,
      "step": 500
    },
    {
      "epoch": 1.3081861958266452,
      "grad_norm": 14.808961868286133,
      "learning_rate": 0.00012574808142128477,
      "loss": 14.240873718261719,
      "step": 510
    },
    {
      "epoch": 1.333868378812199,
      "grad_norm": 13.389018058776855,
      "learning_rate": 0.00012306158707424403,
      "loss": 14.467668151855468,
      "step": 520
    },
    {
      "epoch": 1.3595505617977528,
      "grad_norm": 16.077634811401367,
      "learning_rate": 0.00012035739428461739,
      "loss": 13.67303466796875,
      "step": 530
    },
    {
      "epoch": 1.3852327447833066,
      "grad_norm": 14.300647735595703,
      "learning_rate": 0.00011763757836494403,
      "loss": 15.072747802734375,
      "step": 540
    },
    {
      "epoch": 1.4109149277688604,
      "grad_norm": 12.59499454498291,
      "learning_rate": 0.00011490422661761744,
      "loss": 14.030448913574219,
      "step": 550
    },
    {
      "epoch": 1.4365971107544142,
      "grad_norm": 17.819976806640625,
      "learning_rate": 0.00011215943673300093,
      "loss": 13.556326293945313,
      "step": 560
    },
    {
      "epoch": 1.462279293739968,
      "grad_norm": 15.307595252990723,
      "learning_rate": 0.00010940531517957073,
      "loss": 15.464706420898438,
      "step": 570
    },
    {
      "epoch": 1.4879614767255216,
      "grad_norm": 14.918625831604004,
      "learning_rate": 0.00010664397558732244,
      "loss": 14.040945434570313,
      "step": 580
    },
    {
      "epoch": 1.5136436597110754,
      "grad_norm": 13.040802955627441,
      "learning_rate": 0.0001038775371256817,
      "loss": 14.671842956542969,
      "step": 590
    },
    {
      "epoch": 1.5393258426966292,
      "grad_norm": 14.143183708190918,
      "learning_rate": 0.00010110812287716327,
      "loss": 14.198591613769532,
      "step": 600
    },
    {
      "epoch": 1.565008025682183,
      "grad_norm": 16.373567581176758,
      "learning_rate": 9.833785820802739e-05,
      "loss": 14.670704650878907,
      "step": 610
    },
    {
      "epoch": 1.5906902086677368,
      "grad_norm": 17.14202117919922,
      "learning_rate": 9.556886913718317e-05,
      "loss": 14.713813781738281,
      "step": 620
    },
    {
      "epoch": 1.6163723916532904,
      "grad_norm": 13.428468704223633,
      "learning_rate": 9.280328070459135e-05,
      "loss": 14.256681823730469,
      "step": 630
    },
    {
      "epoch": 1.6420545746388444,
      "grad_norm": 14.759570121765137,
      "learning_rate": 9.004321534041835e-05,
      "loss": 14.875436401367187,
      "step": 640
    },
    {
      "epoch": 1.667736757624398,
      "grad_norm": 14.605072021484375,
      "learning_rate": 8.729079123619286e-05,
      "loss": 14.151382446289062,
      "step": 650
    },
    {
      "epoch": 1.6934189406099518,
      "grad_norm": 15.300275802612305,
      "learning_rate": 8.454812071921596e-05,
      "loss": 14.209205627441406,
      "step": 660
    },
    {
      "epoch": 1.7191011235955056,
      "grad_norm": 15.565445899963379,
      "learning_rate": 8.181730863147093e-05,
      "loss": 15.246949768066406,
      "step": 670
    },
    {
      "epoch": 1.7447833065810594,
      "grad_norm": 13.463878631591797,
      "learning_rate": 7.910045071427829e-05,
      "loss": 14.081675720214843,
      "step": 680
    },
    {
      "epoch": 1.7704654895666132,
      "grad_norm": 15.799697875976562,
      "learning_rate": 7.63996319999347e-05,
      "loss": 14.659947204589844,
      "step": 690
    },
    {
      "epoch": 1.7961476725521668,
      "grad_norm": 15.421786308288574,
      "learning_rate": 7.371692521157048e-05,
      "loss": 13.642781066894532,
      "step": 700
    },
    {
      "epoch": 1.8218298555377208,
      "grad_norm": 16.400951385498047,
      "learning_rate": 7.10543891724537e-05,
      "loss": 14.9028076171875,
      "step": 710
    },
    {
      "epoch": 1.8475120385232744,
      "grad_norm": 16.549213409423828,
      "learning_rate": 6.841406722596191e-05,
      "loss": 14.747923278808594,
      "step": 720
    },
    {
      "epoch": 1.8731942215088284,
      "grad_norm": 14.868678092956543,
      "learning_rate": 6.579798566743314e-05,
      "loss": 14.379522705078125,
      "step": 730
    },
    {
      "epoch": 1.898876404494382,
      "grad_norm": 17.223217010498047,
      "learning_rate": 6.320815218910101e-05,
      "loss": 14.513031005859375,
      "step": 740
    },
    {
      "epoch": 1.9245585874799358,
      "grad_norm": 15.178607940673828,
      "learning_rate": 6.064655433930624e-05,
      "loss": 14.75238494873047,
      "step": 750
    },
    {
      "epoch": 1.9502407704654896,
      "grad_norm": 14.607568740844727,
      "learning_rate": 5.8115157997167536e-05,
      "loss": 14.215359497070313,
      "step": 760
    },
    {
      "epoch": 1.9759229534510432,
      "grad_norm": 15.849374771118164,
      "learning_rate": 5.561590586388221e-05,
      "loss": 13.844842529296875,
      "step": 770
    },
    {
      "epoch": 2.0,
      "grad_norm": 8.945368766784668,
      "learning_rate": 5.315071597181504e-05,
      "loss": 13.500308227539062,
      "step": 780
    },
    {
      "epoch": 2.0256821829855536,
      "grad_norm": 14.872233390808105,
      "learning_rate": 5.072148021251821e-05,
      "loss": 10.96926498413086,
      "step": 790
    },
    {
      "epoch": 2.0513643659711076,
      "grad_norm": 16.96340560913086,
      "learning_rate": 4.833006288481371e-05,
      "loss": 10.601210021972657,
      "step": 800
    },
    {
      "epoch": 2.077046548956661,
      "grad_norm": 17.987470626831055,
      "learning_rate": 4.597829926405075e-05,
      "loss": 11.348848724365235,
      "step": 810
    },
    {
      "epoch": 2.102728731942215,
      "grad_norm": 18.53151512145996,
      "learning_rate": 4.3667994193637796e-05,
      "loss": 10.507981109619141,
      "step": 820
    },
    {
      "epoch": 2.128410914927769,
      "grad_norm": 23.948802947998047,
      "learning_rate": 4.140092069992867e-05,
      "loss": 11.215933227539063,
      "step": 830
    },
    {
      "epoch": 2.154093097913323,
      "grad_norm": 21.3605899810791,
      "learning_rate": 3.91788186315269e-05,
      "loss": 10.393006896972656,
      "step": 840
    },
    {
      "epoch": 2.1797752808988764,
      "grad_norm": 26.548303604125977,
      "learning_rate": 3.7003393324051874e-05,
      "loss": 10.384098052978516,
      "step": 850
    },
    {
      "epoch": 2.20545746388443,
      "grad_norm": 23.979907989501953,
      "learning_rate": 3.487631429139183e-05,
      "loss": 10.094139862060548,
      "step": 860
    },
    {
      "epoch": 2.231139646869984,
      "grad_norm": 18.00309944152832,
      "learning_rate": 3.279921394444776e-05,
      "loss": 10.624467468261718,
      "step": 870
    },
    {
      "epoch": 2.2568218298555376,
      "grad_norm": 24.657569885253906,
      "learning_rate": 3.077368633835205e-05,
      "loss": 10.711078643798828,
      "step": 880
    },
    {
      "epoch": 2.2825040128410916,
      "grad_norm": 22.551204681396484,
      "learning_rate": 2.8801285949122593e-05,
      "loss": 10.329103088378906,
      "step": 890
    },
    {
      "epoch": 2.308186195826645,
      "grad_norm": 26.180330276489258,
      "learning_rate": 2.688352648069198e-05,
      "loss": 9.944695281982423,
      "step": 900
    },
    {
      "epoch": 2.333868378812199,
      "grad_norm": 34.10860061645508,
      "learning_rate": 2.502187970322657e-05,
      "loss": 10.465196990966797,
      "step": 910
    },
    {
      "epoch": 2.359550561797753,
      "grad_norm": 25.76052474975586,
      "learning_rate": 2.321777432362764e-05,
      "loss": 10.892754364013673,
      "step": 920
    },
    {
      "epoch": 2.3852327447833064,
      "grad_norm": 25.387516021728516,
      "learning_rate": 2.1472594889080756e-05,
      "loss": 10.628679656982422,
      "step": 930
    },
    {
      "epoch": 2.4109149277688604,
      "grad_norm": 22.750202178955078,
      "learning_rate": 1.9787680724495617e-05,
      "loss": 10.46801986694336,
      "step": 940
    },
    {
      "epoch": 2.436597110754414,
      "grad_norm": 29.39733123779297,
      "learning_rate": 1.8164324904650965e-05,
      "loss": 11.32564697265625,
      "step": 950
    },
    {
      "epoch": 2.462279293739968,
      "grad_norm": 23.36044692993164,
      "learning_rate": 1.660377326183412e-05,
      "loss": 10.110736083984374,
      "step": 960
    },
    {
      "epoch": 2.4879614767255216,
      "grad_norm": 22.447628021240234,
      "learning_rate": 1.5107223429736272e-05,
      "loss": 10.582487487792969,
      "step": 970
    },
    {
      "epoch": 2.513643659711075,
      "grad_norm": 21.37103271484375,
      "learning_rate": 1.3675823924337506e-05,
      "loss": 9.974002838134766,
      "step": 980
    },
    {
      "epoch": 2.539325842696629,
      "grad_norm": 23.942567825317383,
      "learning_rate": 1.2310673262486705e-05,
      "loss": 10.401480102539063,
      "step": 990
    },
    {
      "epoch": 2.5650080256821832,
      "grad_norm": 23.232219696044922,
      "learning_rate": 1.1012819118853147e-05,
      "loss": 10.403594970703125,
      "step": 1000
    },
    {
      "epoch": 2.590690208667737,
      "grad_norm": 26.392332077026367,
      "learning_rate": 9.783257521896227e-06,
      "loss": 11.170610046386718,
      "step": 1010
    },
    {
      "epoch": 2.6163723916532904,
      "grad_norm": 22.941421508789062,
      "learning_rate": 8.62293208947107e-06,
      "loss": 9.754792785644531,
      "step": 1020
    },
    {
      "epoch": 2.6420545746388444,
      "grad_norm": 31.97374153137207,
      "learning_rate": 7.532733304655848e-06,
      "loss": 10.895748138427734,
      "step": 1030
    },
    {
      "epoch": 2.667736757624398,
      "grad_norm": 22.15962791442871,
      "learning_rate": 6.5134978323574066e-06,
      "loss": 10.086806488037109,
      "step": 1040
    },
    {
      "epoch": 2.693418940609952,
      "grad_norm": 23.121976852416992,
      "learning_rate": 5.566007877218882e-06,
      "loss": 9.806757354736328,
      "step": 1050
    },
    {
      "epoch": 2.7191011235955056,
      "grad_norm": 22.89173698425293,
      "learning_rate": 4.6909905833226966e-06,
      "loss": 9.998442077636719,
      "step": 1060
    },
    {
      "epoch": 2.744783306581059,
      "grad_norm": 28.151565551757812,
      "learning_rate": 3.8891174761491735e-06,
      "loss": 10.232617950439453,
      "step": 1070
    },
    {
      "epoch": 2.770465489566613,
      "grad_norm": 27.67608070373535,
      "learning_rate": 3.161003947219421e-06,
      "loss": 11.071966552734375,
      "step": 1080
    },
    {
      "epoch": 2.796147672552167,
      "grad_norm": 26.361303329467773,
      "learning_rate": 2.5072087818176382e-06,
      "loss": 11.048786926269532,
      "step": 1090
    },
    {
      "epoch": 2.821829855537721,
      "grad_norm": 24.121335983276367,
      "learning_rate": 1.928233730155604e-06,
      "loss": 10.665110015869141,
      "step": 1100
    },
    {
      "epoch": 2.8475120385232744,
      "grad_norm": 26.22313117980957,
      "learning_rate": 1.4245231223081301e-06,
      "loss": 11.044110107421876,
      "step": 1110
    },
    {
      "epoch": 2.8731942215088284,
      "grad_norm": 26.499591827392578,
      "learning_rate": 9.964635272153633e-07,
      "loss": 10.147935485839843,
      "step": 1120
    },
    {
      "epoch": 2.898876404494382,
      "grad_norm": 23.90386390686035,
      "learning_rate": 6.443834560132534e-07,
      "loss": 10.296646881103516,
      "step": 1130
    },
    {
      "epoch": 2.924558587479936,
      "grad_norm": 24.210269927978516,
      "learning_rate": 3.685531099202111e-07,
      "loss": 10.592522430419923,
      "step": 1140
    },
    {
      "epoch": 2.9502407704654896,
      "grad_norm": 28.532909393310547,
      "learning_rate": 1.6918417287318245e-07,
      "loss": 9.853338623046875,
      "step": 1150
    },
    {
      "epoch": 2.975922953451043,
      "grad_norm": 27.218006134033203,
      "learning_rate": 4.642964907235481e-08,
      "loss": 10.533452606201172,
      "step": 1160
    },
    {
      "epoch": 3.0,
      "grad_norm": 14.907074928283691,
      "learning_rate": 3.837455592847761e-10,
      "loss": 9.5769775390625,
      "step": 1170
    },
    {
      "epoch": 3.0,
      "step": 1170,
      "total_flos": 2.709259627140219e+17,
      "train_loss": 14.446504472259782,
      "train_runtime": 5026.5595,
      "train_samples_per_second": 1.859,
      "train_steps_per_second": 0.233
    }
  ],
  "logging_steps": 10,
  "max_steps": 1170,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 200,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2.709259627140219e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}