{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 731,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.006842285323297981,
      "grad_norm": 45.0,
      "learning_rate": 3.6363636363636366e-06,
      "loss": 2.809466743469238,
      "step": 5
    },
    {
      "epoch": 0.013684570646595963,
      "grad_norm": 12.25,
      "learning_rate": 8.181818181818183e-06,
      "loss": 2.548305892944336,
      "step": 10
    },
    {
      "epoch": 0.020526855969893943,
      "grad_norm": 8.1875,
      "learning_rate": 1.2727272727272728e-05,
      "loss": 2.402619743347168,
      "step": 15
    },
    {
      "epoch": 0.027369141293191925,
      "grad_norm": 8.5,
      "learning_rate": 1.7272727272727274e-05,
      "loss": 1.9256172180175781,
      "step": 20
    },
    {
      "epoch": 0.03421142661648991,
      "grad_norm": 3.953125,
      "learning_rate": 1.9999607323936722e-05,
      "loss": 2.0962652206420898,
      "step": 25
    },
    {
      "epoch": 0.041053711939787886,
      "grad_norm": 3.796875,
      "learning_rate": 1.9995190072382676e-05,
      "loss": 1.8469963073730469,
      "step": 30
    },
    {
      "epoch": 0.04789599726308587,
      "grad_norm": 4.28125,
      "learning_rate": 1.9985866899535076e-05,
      "loss": 1.8669818878173827,
      "step": 35
    },
    {
      "epoch": 0.05473828258638385,
      "grad_norm": 3.359375,
      "learning_rate": 1.9971642381470207e-05,
      "loss": 1.7637271881103516,
      "step": 40
    },
    {
      "epoch": 0.06158056790968183,
      "grad_norm": 2.359375,
      "learning_rate": 1.9952523499982866e-05,
      "loss": 1.705878448486328,
      "step": 45
    },
    {
      "epoch": 0.06842285323297982,
      "grad_norm": 2.953125,
      "learning_rate": 1.9928519639159508e-05,
      "loss": 1.6680316925048828,
      "step": 50
    },
    {
      "epoch": 0.0752651385562778,
      "grad_norm": 3.796875,
      "learning_rate": 1.9899642580772274e-05,
      "loss": 1.7056447982788085,
      "step": 55
    },
    {
      "epoch": 0.08210742387957577,
      "grad_norm": 2.328125,
      "learning_rate": 1.9865906498496164e-05,
      "loss": 1.636207389831543,
      "step": 60
    },
    {
      "epoch": 0.08894970920287376,
      "grad_norm": 3.5,
      "learning_rate": 1.9827327950952197e-05,
      "loss": 1.623453712463379,
      "step": 65
    },
    {
      "epoch": 0.09579199452617174,
      "grad_norm": 2.3125,
      "learning_rate": 1.9783925873579967e-05,
      "loss": 1.6949512481689453,
      "step": 70
    },
    {
      "epoch": 0.10263427984946973,
      "grad_norm": 2.65625,
      "learning_rate": 1.97357215693436e-05,
      "loss": 1.5341100692749023,
      "step": 75
    },
    {
      "epoch": 0.1094765651727677,
      "grad_norm": 2.109375,
      "learning_rate": 1.9682738698275663e-05,
      "loss": 1.606492805480957,
      "step": 80
    },
    {
      "epoch": 0.11631885049606569,
      "grad_norm": 2.78125,
      "learning_rate": 1.9625003265864148e-05,
      "loss": 1.4498618125915528,
      "step": 85
    },
    {
      "epoch": 0.12316113581936367,
      "grad_norm": 2.28125,
      "learning_rate": 1.956254361028823e-05,
      "loss": 1.5422866821289063,
      "step": 90
    },
    {
      "epoch": 0.13000342114266164,
      "grad_norm": 3.703125,
      "learning_rate": 1.9495390388509123e-05,
      "loss": 1.5427611351013184,
      "step": 95
    },
    {
      "epoch": 0.13684570646595964,
      "grad_norm": 3.546875,
      "learning_rate": 1.9423576561222742e-05,
      "loss": 1.564542007446289,
      "step": 100
    },
    {
      "epoch": 0.14368799178925762,
      "grad_norm": 2.359375,
      "learning_rate": 1.9347137376681692e-05,
      "loss": 1.4948160171508789,
      "step": 105
    },
    {
      "epoch": 0.1505302771125556,
      "grad_norm": 2.671875,
      "learning_rate": 1.926611035339441e-05,
      "loss": 1.539006805419922,
      "step": 110
    },
    {
      "epoch": 0.15737256243585357,
      "grad_norm": 2.859375,
      "learning_rate": 1.918053526171004e-05,
      "loss": 1.542184543609619,
      "step": 115
    },
    {
      "epoch": 0.16421484775915154,
      "grad_norm": 2.640625,
      "learning_rate": 1.9090454104298003e-05,
      "loss": 1.6334993362426757,
      "step": 120
    },
    {
      "epoch": 0.17105713308244955,
      "grad_norm": 3.71875,
      "learning_rate": 1.8995911095531932e-05,
      "loss": 1.4427576065063477,
      "step": 125
    },
    {
      "epoch": 0.17789941840574752,
      "grad_norm": 3.21875,
      "learning_rate": 1.8896952639787977e-05,
      "loss": 1.5040811538696288,
      "step": 130
    },
    {
      "epoch": 0.1847417037290455,
      "grad_norm": 2.625,
      "learning_rate": 1.879362730866825e-05,
      "loss": 1.5777132034301757,
      "step": 135
    },
    {
      "epoch": 0.19158398905234347,
      "grad_norm": 2.15625,
      "learning_rate": 1.8685985817160503e-05,
      "loss": 1.5796485900878907,
      "step": 140
    },
    {
      "epoch": 0.19842627437564148,
      "grad_norm": 4.3125,
      "learning_rate": 1.8574080998745785e-05,
      "loss": 1.5201278686523438,
      "step": 145
    },
    {
      "epoch": 0.20526855969893945,
      "grad_norm": 2.859375,
      "learning_rate": 1.845796777946627e-05,
      "loss": 1.5156174659729005,
      "step": 150
    },
    {
      "epoch": 0.21211084502223743,
      "grad_norm": 2.265625,
      "learning_rate": 1.8337703150966025e-05,
      "loss": 1.494316291809082,
      "step": 155
    },
    {
      "epoch": 0.2189531303455354,
      "grad_norm": 2.828125,
      "learning_rate": 1.8213346142517884e-05,
      "loss": 1.4805891036987304,
      "step": 160
    },
    {
      "epoch": 0.22579541566883338,
      "grad_norm": 2.125,
      "learning_rate": 1.8084957792050226e-05,
      "loss": 1.5021228790283203,
      "step": 165
    },
    {
      "epoch": 0.23263770099213138,
      "grad_norm": 2.390625,
      "learning_rate": 1.7952601116187822e-05,
      "loss": 1.4761942863464355,
      "step": 170
    },
    {
      "epoch": 0.23947998631542936,
      "grad_norm": 2.328125,
      "learning_rate": 1.781634107932153e-05,
      "loss": 1.5559394836425782,
      "step": 175
    },
    {
      "epoch": 0.24632227163872733,
      "grad_norm": 3.171875,
      "learning_rate": 1.7676244561721905e-05,
      "loss": 1.4734925270080566,
      "step": 180
    },
    {
      "epoch": 0.25316455696202533,
      "grad_norm": 3.703125,
      "learning_rate": 1.7532380326712487e-05,
      "loss": 1.4294141769409179,
      "step": 185
    },
    {
      "epoch": 0.2600068422853233,
      "grad_norm": 2.46875,
      "learning_rate": 1.738481898691884e-05,
      "loss": 1.4497101783752442,
      "step": 190
    },
    {
      "epoch": 0.2668491276086213,
      "grad_norm": 2.875,
      "learning_rate": 1.7233632969609844e-05,
      "loss": 1.4012206077575684,
      "step": 195
    },
    {
      "epoch": 0.2736914129319193,
      "grad_norm": 2.96875,
      "learning_rate": 1.707889648114839e-05,
      "loss": 1.4317466735839843,
      "step": 200
    },
    {
      "epoch": 0.28053369825521723,
      "grad_norm": 3.59375,
      "learning_rate": 1.6920685470568775e-05,
      "loss": 1.4276908874511718,
      "step": 205
    },
    {
      "epoch": 0.28737598357851524,
      "grad_norm": 2.546875,
      "learning_rate": 1.675907759229879e-05,
      "loss": 1.4698185920715332,
      "step": 210
    },
    {
      "epoch": 0.2942182689018132,
      "grad_norm": 2.765625,
      "learning_rate": 1.659415216804471e-05,
      "loss": 1.389587688446045,
      "step": 215
    },
    {
      "epoch": 0.3010605542251112,
      "grad_norm": 2.125,
      "learning_rate": 1.6425990147858004e-05,
      "loss": 1.4244729042053224,
      "step": 220
    },
    {
      "epoch": 0.3079028395484092,
      "grad_norm": 2.171875,
      "learning_rate": 1.625467407040273e-05,
      "loss": 1.423227310180664,
      "step": 225
    },
    {
      "epoch": 0.31474512487170714,
      "grad_norm": 2.78125,
      "learning_rate": 1.6080288022443242e-05,
      "loss": 1.4495688438415528,
      "step": 230
    },
    {
      "epoch": 0.32158741019500514,
      "grad_norm": 2.9375,
      "learning_rate": 1.590291759757204e-05,
      "loss": 1.4790194511413575,
      "step": 235
    },
    {
      "epoch": 0.3284296955183031,
      "grad_norm": 3.96875,
      "learning_rate": 1.5722649854198007e-05,
      "loss": 1.3981167793273925,
      "step": 240
    },
    {
      "epoch": 0.3352719808416011,
      "grad_norm": 2.515625,
      "learning_rate": 1.5539573272815698e-05,
      "loss": 1.3656198501586914,
      "step": 245
    },
    {
      "epoch": 0.3421142661648991,
      "grad_norm": 2.65625,
      "learning_rate": 1.5353777712576613e-05,
      "loss": 1.4632407188415528,
      "step": 250
    },
    {
      "epoch": 0.34895655148819704,
      "grad_norm": 2.53125,
      "learning_rate": 1.5165354367183778e-05,
      "loss": 1.5673239707946778,
      "step": 255
    },
    {
      "epoch": 0.35579883681149505,
      "grad_norm": 2.828125,
      "learning_rate": 1.4974395720131328e-05,
      "loss": 1.4822378158569336,
      "step": 260
    },
    {
      "epoch": 0.36264112213479305,
      "grad_norm": 2.453125,
      "learning_rate": 1.4780995499310973e-05,
      "loss": 1.4012256622314454,
      "step": 265
    },
    {
      "epoch": 0.369483407458091,
      "grad_norm": 2.6875,
      "learning_rate": 1.458524863100772e-05,
      "loss": 1.3955447196960449,
      "step": 270
    },
    {
      "epoch": 0.376325692781389,
      "grad_norm": 3.625,
      "learning_rate": 1.4387251193307369e-05,
      "loss": 1.4113948822021485,
      "step": 275
    },
    {
      "epoch": 0.38316797810468695,
      "grad_norm": 2.515625,
      "learning_rate": 1.4187100368938678e-05,
      "loss": 1.4017261505126952,
      "step": 280
    },
    {
      "epoch": 0.39001026342798495,
      "grad_norm": 2.296875,
      "learning_rate": 1.3984894397573332e-05,
      "loss": 1.5206121444702148,
      "step": 285
    },
    {
      "epoch": 0.39685254875128295,
      "grad_norm": 3.796875,
      "learning_rate": 1.3780732527607157e-05,
      "loss": 1.322197723388672,
      "step": 290
    },
    {
      "epoch": 0.4036948340745809,
      "grad_norm": 2.484375,
      "learning_rate": 1.3574714967446193e-05,
      "loss": 1.3588292121887207,
      "step": 295
    },
    {
      "epoch": 0.4105371193978789,
      "grad_norm": 3.359375,
      "learning_rate": 1.3366942836321574e-05,
      "loss": 1.3987003326416017,
      "step": 300
    },
    {
      "epoch": 0.41737940472117685,
      "grad_norm": 2.890625,
      "learning_rate": 1.3157518114657382e-05,
      "loss": 1.3717717170715331,
      "step": 305
    },
    {
      "epoch": 0.42422169004447485,
      "grad_norm": 2.3125,
      "learning_rate": 1.2946543594015752e-05,
      "loss": 1.4352380752563476,
      "step": 310
    },
    {
      "epoch": 0.43106397536777286,
      "grad_norm": 2.671875,
      "learning_rate": 1.2734122826643884e-05,
      "loss": 1.46458158493042,
      "step": 315
    },
    {
      "epoch": 0.4379062606910708,
      "grad_norm": 4.15625,
      "learning_rate": 1.2520360074647686e-05,
      "loss": 1.4723733901977538,
      "step": 320
    },
    {
      "epoch": 0.4447485460143688,
      "grad_norm": 3.171875,
      "learning_rate": 1.2305360258816976e-05,
      "loss": 1.448530387878418,
      "step": 325
    },
    {
      "epoch": 0.45159083133766675,
      "grad_norm": 2.421875,
      "learning_rate": 1.2089228907127403e-05,
      "loss": 1.4141918182373048,
      "step": 330
    },
    {
      "epoch": 0.45843311666096476,
      "grad_norm": 3.1875,
      "learning_rate": 1.1872072102944345e-05,
      "loss": 1.4687261581420898,
      "step": 335
    },
    {
      "epoch": 0.46527540198426276,
      "grad_norm": 2.578125,
      "learning_rate": 1.1653996432954194e-05,
      "loss": 1.381917667388916,
      "step": 340
    },
    {
      "epoch": 0.4721176873075607,
      "grad_norm": 2.484375,
      "learning_rate": 1.1435108934848609e-05,
      "loss": 1.398417091369629,
      "step": 345
    },
    {
      "epoch": 0.4789599726308587,
      "grad_norm": 2.34375,
      "learning_rate": 1.1215517044787387e-05,
      "loss": 1.3287519454956054,
      "step": 350
    },
    {
      "epoch": 0.4858022579541567,
      "grad_norm": 2.5,
      "learning_rate": 1.0995328544665776e-05,
      "loss": 1.4714726448059081,
      "step": 355
    },
    {
      "epoch": 0.49264454327745466,
      "grad_norm": 3.5,
      "learning_rate": 1.077465150921207e-05,
      "loss": 1.431112003326416,
      "step": 360
    },
    {
      "epoch": 0.49948682860075266,
      "grad_norm": 3.359375,
      "learning_rate": 1.0553594252941489e-05,
      "loss": 1.4446621894836427,
      "step": 365
    },
    {
      "epoch": 0.5063291139240507,
      "grad_norm": 3.34375,
      "learning_rate": 1.0332265276992361e-05,
      "loss": 1.3496983528137207,
      "step": 370
    },
    {
      "epoch": 0.5131713992473487,
      "grad_norm": 2.53125,
      "learning_rate": 1.0110773215870695e-05,
      "loss": 1.324970531463623,
      "step": 375
    },
    {
      "epoch": 0.5200136845706466,
      "grad_norm": 3.046875,
      "learning_rate": 9.889226784129307e-06,
      "loss": 1.3443440437316894,
      "step": 380
    },
    {
      "epoch": 0.5268559698939446,
      "grad_norm": 2.40625,
      "learning_rate": 9.66773472300764e-06,
      "loss": 1.4248334884643554,
      "step": 385
    },
    {
      "epoch": 0.5336982552172426,
      "grad_norm": 2.34375,
      "learning_rate": 9.446405747058513e-06,
      "loss": 1.416274070739746,
      "step": 390
    },
    {
      "epoch": 0.5405405405405406,
      "grad_norm": 2.390625,
      "learning_rate": 9.225348490787935e-06,
      "loss": 1.4464293479919434,
      "step": 395
    },
    {
      "epoch": 0.5473828258638386,
      "grad_norm": 2.296875,
      "learning_rate": 9.004671455334227e-06,
      "loss": 1.3776466369628906,
      "step": 400
    },
    {
      "epoch": 0.5542251111871365,
      "grad_norm": 3.75,
      "learning_rate": 8.784482955212615e-06,
      "loss": 1.3454496383666992,
      "step": 405
    },
    {
      "epoch": 0.5610673965104345,
      "grad_norm": 2.796875,
      "learning_rate": 8.564891065151391e-06,
      "loss": 1.3769558906555175,
      "step": 410
    },
    {
      "epoch": 0.5679096818337325,
      "grad_norm": 2.328125,
      "learning_rate": 8.346003567045806e-06,
      "loss": 1.4825267791748047,
      "step": 415
    },
    {
      "epoch": 0.5747519671570305,
      "grad_norm": 2.5625,
      "learning_rate": 8.127927897055658e-06,
      "loss": 1.3498231887817382,
      "step": 420
    },
    {
      "epoch": 0.5815942524803285,
      "grad_norm": 2.234375,
      "learning_rate": 7.910771092872598e-06,
      "loss": 1.4994674682617188,
      "step": 425
    },
    {
      "epoch": 0.5884365378036264,
      "grad_norm": 2.5625,
      "learning_rate": 7.694639741183027e-06,
      "loss": 1.406801700592041,
      "step": 430
    },
    {
      "epoch": 0.5952788231269244,
      "grad_norm": 2.40625,
      "learning_rate": 7.479639925352318e-06,
      "loss": 1.57269344329834,
      "step": 435
    },
    {
      "epoch": 0.6021211084502224,
      "grad_norm": 3.390625,
      "learning_rate": 7.2658771733561175e-06,
      "loss": 1.3171756744384766,
      "step": 440
    },
    {
      "epoch": 0.6089633937735204,
      "grad_norm": 2.421875,
      "learning_rate": 7.053456405984254e-06,
      "loss": 1.4658644676208497,
      "step": 445
    },
    {
      "epoch": 0.6158056790968184,
      "grad_norm": 2.328125,
      "learning_rate": 6.842481885342624e-06,
      "loss": 1.4110689163208008,
      "step": 450
    },
    {
      "epoch": 0.6226479644201163,
      "grad_norm": 2.53125,
      "learning_rate": 6.633057163678427e-06,
      "loss": 1.3724404335021974,
      "step": 455
    },
    {
      "epoch": 0.6294902497434143,
      "grad_norm": 2.15625,
      "learning_rate": 6.42528503255381e-06,
      "loss": 1.3259340286254884,
      "step": 460
    },
    {
      "epoch": 0.6363325350667123,
      "grad_norm": 2.359375,
      "learning_rate": 6.2192674723928425e-06,
      "loss": 1.4074788093566895,
      "step": 465
    },
    {
      "epoch": 0.6431748203900103,
      "grad_norm": 2.734375,
      "learning_rate": 6.01510560242667e-06,
      "loss": 1.4261545181274413,
      "step": 470
    },
    {
      "epoch": 0.6500171057133083,
      "grad_norm": 3.03125,
      "learning_rate": 5.812899631061328e-06,
      "loss": 1.411759376525879,
      "step": 475
    },
    {
      "epoch": 0.6568593910366062,
      "grad_norm": 2.25,
      "learning_rate": 5.612748806692631e-06,
      "loss": 1.4686535835266112,
      "step": 480
    },
    {
      "epoch": 0.6637016763599042,
      "grad_norm": 2.875,
      "learning_rate": 5.414751368992281e-06,
      "loss": 1.3639310836791991,
      "step": 485
    },
    {
      "epoch": 0.6705439616832022,
      "grad_norm": 2.546875,
      "learning_rate": 5.219004500689031e-06,
      "loss": 1.370129108428955,
      "step": 490
    },
    {
      "epoch": 0.6773862470065002,
      "grad_norm": 4.40625,
      "learning_rate": 5.025604279868677e-06,
      "loss": 1.3223713874816894,
      "step": 495
    },
    {
      "epoch": 0.6842285323297982,
      "grad_norm": 2.59375,
      "learning_rate": 4.834645632816227e-06,
      "loss": 1.4085097312927246,
      "step": 500
    },
    {
      "epoch": 0.6910708176530961,
      "grad_norm": 3.265625,
      "learning_rate": 4.646222287423391e-06,
      "loss": 1.3346193313598633,
      "step": 505
    },
    {
      "epoch": 0.6979131029763941,
      "grad_norm": 3.53125,
      "learning_rate": 4.460426727184305e-06,
      "loss": 1.3574554443359375,
      "step": 510
    },
    {
      "epoch": 0.7047553882996921,
      "grad_norm": 3.546875,
      "learning_rate": 4.277350145801994e-06,
      "loss": 1.3625916481018066,
      "step": 515
    },
    {
      "epoch": 0.7115976736229901,
      "grad_norm": 2.5,
      "learning_rate": 4.097082402427962e-06,
      "loss": 1.4155165672302246,
      "step": 520
    },
    {
      "epoch": 0.7184399589462881,
      "grad_norm": 2.640625,
      "learning_rate": 3.9197119775567595e-06,
      "loss": 1.4300453186035156,
      "step": 525
    },
    {
      "epoch": 0.7252822442695861,
      "grad_norm": 2.375,
      "learning_rate": 3.745325929597272e-06,
      "loss": 1.369569969177246,
      "step": 530
    },
    {
      "epoch": 0.732124529592884,
      "grad_norm": 2.484375,
      "learning_rate": 3.5740098521419985e-06,
      "loss": 1.4022644996643066,
      "step": 535
    },
    {
      "epoch": 0.738966814916182,
      "grad_norm": 3.15625,
      "learning_rate": 3.4058478319552935e-06,
      "loss": 1.4104683876037598,
      "step": 540
    },
    {
      "epoch": 0.74580910023948,
      "grad_norm": 2.234375,
      "learning_rate": 3.2409224077012135e-06,
      "loss": 1.4028725624084473,
      "step": 545
    },
    {
      "epoch": 0.752651385562778,
      "grad_norm": 2.34375,
      "learning_rate": 3.0793145294312253e-06,
      "loss": 1.4290534019470216,
      "step": 550
    },
    {
      "epoch": 0.759493670886076,
      "grad_norm": 2.484375,
      "learning_rate": 2.921103518851609e-06,
      "loss": 1.4063751220703125,
      "step": 555
    },
    {
      "epoch": 0.7663359562093739,
      "grad_norm": 3.0,
      "learning_rate": 2.766367030390157e-06,
      "loss": 1.3523286819458007,
      "step": 560
    },
    {
      "epoch": 0.7731782415326719,
      "grad_norm": 3.78125,
      "learning_rate": 2.615181013081164e-06,
      "loss": 1.397650909423828,
      "step": 565
    },
    {
      "epoch": 0.7800205268559699,
      "grad_norm": 2.59375,
      "learning_rate": 2.4676196732875147e-06,
      "loss": 1.4020781517028809,
      "step": 570
    },
    {
      "epoch": 0.7868628121792679,
      "grad_norm": 2.0625,
      "learning_rate": 2.3237554382781004e-06,
      "loss": 1.4961904525756835,
      "step": 575
    },
    {
      "epoch": 0.7937050975025659,
      "grad_norm": 2.78125,
      "learning_rate": 2.1836589206784742e-06,
      "loss": 1.451511764526367,
      "step": 580
    },
    {
      "epoch": 0.8005473828258638,
      "grad_norm": 2.203125,
      "learning_rate": 2.0473988838121783e-06,
      "loss": 1.3788551330566405,
      "step": 585
    },
    {
      "epoch": 0.8073896681491618,
      "grad_norm": 2.4375,
      "learning_rate": 1.9150422079497787e-06,
      "loss": 1.410994529724121,
      "step": 590
    },
    {
      "epoch": 0.8142319534724598,
      "grad_norm": 3.6875,
      "learning_rate": 1.786653857482118e-06,
      "loss": 1.3450016975402832,
      "step": 595
    },
    {
      "epoch": 0.8210742387957578,
      "grad_norm": 2.953125,
      "learning_rate": 1.6622968490339775e-06,
      "loss": 1.4508660316467286,
      "step": 600
    },
    {
      "epoch": 0.8279165241190558,
      "grad_norm": 2.640625,
      "learning_rate": 1.5420322205337335e-06,
      "loss": 1.3347270965576172,
      "step": 605
    },
    {
      "epoch": 0.8347588094423537,
      "grad_norm": 2.578125,
      "learning_rate": 1.425919001254219e-06,
      "loss": 1.2560535430908204,
      "step": 610
    },
    {
      "epoch": 0.8416010947656517,
      "grad_norm": 3.75,
      "learning_rate": 1.3140141828394993e-06,
      "loss": 1.355188751220703,
      "step": 615
    },
    {
      "epoch": 0.8484433800889497,
      "grad_norm": 2.609375,
      "learning_rate": 1.2063726913317508e-06,
      "loss": 1.397140884399414,
      "step": 620
    },
    {
      "epoch": 0.8552856654122477,
      "grad_norm": 2.765625,
      "learning_rate": 1.103047360212024e-06,
      "loss": 1.3479557037353516,
      "step": 625
    },
    {
      "epoch": 0.8621279507355457,
      "grad_norm": 2.515625,
      "learning_rate": 1.0040889044680702e-06,
      "loss": 1.3707120895385743,
      "step": 630
    },
    {
      "epoch": 0.8689702360588436,
      "grad_norm": 2.359375,
      "learning_rate": 9.095458957019986e-07,
      "loss": 1.3732416152954101,
      "step": 635
    },
    {
      "epoch": 0.8758125213821416,
      "grad_norm": 2.453125,
      "learning_rate": 8.194647382899657e-07,
      "loss": 1.3615639686584473,
      "step": 640
    },
    {
      "epoch": 0.8826548067054396,
      "grad_norm": 2.25,
      "learning_rate": 7.338896466055934e-07,
      "loss": 1.4400001525878907,
      "step": 645
    },
    {
      "epoch": 0.8894970920287376,
      "grad_norm": 2.96875,
      "learning_rate": 6.52862623318311e-07,
      "loss": 1.3915739059448242,
      "step": 650
    },
    {
      "epoch": 0.8963393773520356,
      "grad_norm": 4.21875,
      "learning_rate": 5.764234387772593e-07,
      "loss": 1.3268877029418946,
      "step": 655
    },
    {
      "epoch": 0.9031816626753335,
      "grad_norm": 2.828125,
      "learning_rate": 5.046096114908783e-07,
      "loss": 1.3429359436035155,
      "step": 660
    },
    {
      "epoch": 0.9100239479986315,
      "grad_norm": 1.9765625,
      "learning_rate": 4.374563897117701e-07,
      "loss": 1.4380492210388183,
      "step": 665
    },
    {
      "epoch": 0.9168662333219295,
      "grad_norm": 2.359375,
      "learning_rate": 3.7499673413585516e-07,
      "loss": 1.3397421836853027,
      "step": 670
    },
    {
      "epoch": 0.9237085186452275,
      "grad_norm": 2.328125,
      "learning_rate": 3.1726130172433756e-07,
      "loss": 1.4501882553100587,
      "step": 675
    },
    {
      "epoch": 0.9305508039685255,
      "grad_norm": 4.21875,
      "learning_rate": 2.642784306564017e-07,
      "loss": 1.3810162544250488,
      "step": 680
    },
    {
      "epoch": 0.9373930892918235,
      "grad_norm": 2.78125,
      "learning_rate": 2.160741264200361e-07,
      "loss": 1.3060113906860351,
      "step": 685
    },
    {
      "epoch": 0.9442353746151214,
      "grad_norm": 2.28125,
      "learning_rate": 1.7267204904780621e-07,
      "loss": 1.4677888870239257,
      "step": 690
    },
    {
      "epoch": 0.9510776599384194,
      "grad_norm": 2.421875,
      "learning_rate": 1.3409350150383804e-07,
      "loss": 1.3917272567749024,
      "step": 695
    },
    {
      "epoch": 0.9579199452617174,
      "grad_norm": 2.3125,
      "learning_rate": 1.0035741922772902e-07,
      "loss": 1.4361152648925781,
      "step": 700
    },
    {
      "epoch": 0.9647622305850154,
      "grad_norm": 2.578125,
      "learning_rate": 7.14803608404957e-08,
      "loss": 1.452861213684082,
      "step": 705
    },
    {
      "epoch": 0.9716045159083134,
      "grad_norm": 2.09375,
      "learning_rate": 4.747650001713622e-08,
      "loss": 1.3411725044250489,
      "step": 710
    },
    {
      "epoch": 0.9784468012316113,
      "grad_norm": 2.515625,
      "learning_rate": 2.835761852979291e-08,
      "loss": 1.3774123191833496,
      "step": 715
    },
    {
      "epoch": 0.9852890865549093,
      "grad_norm": 3.359375,
      "learning_rate": 1.4133100464922556e-08,
      "loss": 1.3300366401672363,
      "step": 720
    },
    {
      "epoch": 0.9921313718782073,
      "grad_norm": 2.234375,
      "learning_rate": 4.809927617324128e-09,
      "loss": 1.295484733581543,
      "step": 725
    },
    {
      "epoch": 0.9989736572015053,
      "grad_norm": 3.359375,
      "learning_rate": 3.9267606327775266e-10,
      "loss": 1.3570124626159668,
      "step": 730
    },
    {
      "epoch": 1.0,
      "step": 731,
      "total_flos": 3.5281465280347546e+17,
      "train_loss": 1.4733744319910553,
      "train_runtime": 6290.4242,
      "train_samples_per_second": 0.465,
      "train_steps_per_second": 0.116
    }
  ],
  "logging_steps": 5,
  "max_steps": 731,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 3.5281465280347546e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}