{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 2423,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0004127115146512588,
      "grad_norm": 24.122843960524676,
      "learning_rate": 4.1152263374485605e-08,
      "loss": 1.4137,
      "step": 1
    },
    {
      "epoch": 0.0020635575732562937,
      "grad_norm": 23.76872795123526,
      "learning_rate": 2.05761316872428e-07,
      "loss": 1.4306,
      "step": 5
    },
    {
      "epoch": 0.004127115146512587,
      "grad_norm": 15.835819961484214,
      "learning_rate": 4.11522633744856e-07,
      "loss": 1.3895,
      "step": 10
    },
    {
      "epoch": 0.006190672719768881,
      "grad_norm": 8.711444647130131,
      "learning_rate": 6.17283950617284e-07,
      "loss": 1.2741,
      "step": 15
    },
    {
      "epoch": 0.008254230293025175,
      "grad_norm": 10.763225748048312,
      "learning_rate": 8.23045267489712e-07,
      "loss": 1.1629,
      "step": 20
    },
    {
      "epoch": 0.010317787866281469,
      "grad_norm": 5.160551789043553,
      "learning_rate": 1.02880658436214e-06,
      "loss": 1.0567,
      "step": 25
    },
    {
      "epoch": 0.012381345439537762,
      "grad_norm": 3.472599934267948,
      "learning_rate": 1.234567901234568e-06,
      "loss": 0.9904,
      "step": 30
    },
    {
      "epoch": 0.014444903012794058,
      "grad_norm": 3.5380974995691528,
      "learning_rate": 1.440329218106996e-06,
      "loss": 0.9621,
      "step": 35
    },
    {
      "epoch": 0.01650846058605035,
      "grad_norm": 3.0479724307668077,
      "learning_rate": 1.646090534979424e-06,
      "loss": 0.9573,
      "step": 40
    },
    {
      "epoch": 0.018572018159306643,
      "grad_norm": 2.97801536675108,
      "learning_rate": 1.8518518518518519e-06,
      "loss": 0.9295,
      "step": 45
    },
    {
      "epoch": 0.020635575732562937,
      "grad_norm": 3.0790095562756514,
      "learning_rate": 2.05761316872428e-06,
      "loss": 0.9106,
      "step": 50
    },
    {
      "epoch": 0.02269913330581923,
      "grad_norm": 3.0873399280284626,
      "learning_rate": 2.263374485596708e-06,
      "loss": 0.9054,
      "step": 55
    },
    {
      "epoch": 0.024762690879075525,
      "grad_norm": 3.046348701050192,
      "learning_rate": 2.469135802469136e-06,
      "loss": 0.9016,
      "step": 60
    },
    {
      "epoch": 0.026826248452331822,
      "grad_norm": 3.117521509733326,
      "learning_rate": 2.674897119341564e-06,
      "loss": 0.9005,
      "step": 65
    },
    {
      "epoch": 0.028889806025588115,
      "grad_norm": 3.046385395158289,
      "learning_rate": 2.880658436213992e-06,
      "loss": 0.8882,
      "step": 70
    },
    {
      "epoch": 0.03095336359884441,
      "grad_norm": 3.1884168387342537,
      "learning_rate": 3.08641975308642e-06,
      "loss": 0.8799,
      "step": 75
    },
    {
      "epoch": 0.0330169211721007,
      "grad_norm": 3.2618306435716176,
      "learning_rate": 3.292181069958848e-06,
      "loss": 0.8907,
      "step": 80
    },
    {
      "epoch": 0.035080478745357,
      "grad_norm": 3.322364354858461,
      "learning_rate": 3.4979423868312762e-06,
      "loss": 0.8684,
      "step": 85
    },
    {
      "epoch": 0.03714403631861329,
      "grad_norm": 2.9808270129935495,
      "learning_rate": 3.7037037037037037e-06,
      "loss": 0.8604,
      "step": 90
    },
    {
      "epoch": 0.039207593891869584,
      "grad_norm": 3.1835415746195435,
      "learning_rate": 3.909465020576132e-06,
      "loss": 0.8707,
      "step": 95
    },
    {
      "epoch": 0.041271151465125874,
      "grad_norm": 3.057661991370929,
      "learning_rate": 4.11522633744856e-06,
      "loss": 0.8703,
      "step": 100
    },
    {
      "epoch": 0.04333470903838217,
      "grad_norm": 2.9074769839799277,
      "learning_rate": 4.3209876543209875e-06,
      "loss": 0.8633,
      "step": 105
    },
    {
      "epoch": 0.04539826661163846,
      "grad_norm": 2.9699831351967405,
      "learning_rate": 4.526748971193416e-06,
      "loss": 0.8759,
      "step": 110
    },
    {
      "epoch": 0.04746182418489476,
      "grad_norm": 3.219620261483576,
      "learning_rate": 4.732510288065844e-06,
      "loss": 0.8636,
      "step": 115
    },
    {
      "epoch": 0.04952538175815105,
      "grad_norm": 3.060892913438638,
      "learning_rate": 4.938271604938272e-06,
      "loss": 0.8512,
      "step": 120
    },
    {
      "epoch": 0.051588939331407346,
      "grad_norm": 3.0812348532752543,
      "learning_rate": 5.1440329218107e-06,
      "loss": 0.8412,
      "step": 125
    },
    {
      "epoch": 0.053652496904663644,
      "grad_norm": 3.208845547096666,
      "learning_rate": 5.349794238683128e-06,
      "loss": 0.8581,
      "step": 130
    },
    {
      "epoch": 0.055716054477919934,
      "grad_norm": 3.461069159058445,
      "learning_rate": 5.555555555555557e-06,
      "loss": 0.8512,
      "step": 135
    },
    {
      "epoch": 0.05777961205117623,
      "grad_norm": 3.040452060763894,
      "learning_rate": 5.761316872427984e-06,
      "loss": 0.857,
      "step": 140
    },
    {
      "epoch": 0.05984316962443252,
      "grad_norm": 2.9241553701138763,
      "learning_rate": 5.967078189300412e-06,
      "loss": 0.8556,
      "step": 145
    },
    {
      "epoch": 0.06190672719768882,
      "grad_norm": 3.22517785299545,
      "learning_rate": 6.17283950617284e-06,
      "loss": 0.8504,
      "step": 150
    },
    {
      "epoch": 0.06397028477094512,
      "grad_norm": 3.066199171323176,
      "learning_rate": 6.3786008230452675e-06,
      "loss": 0.8455,
      "step": 155
    },
    {
      "epoch": 0.0660338423442014,
      "grad_norm": 3.1076485395976436,
      "learning_rate": 6.584362139917696e-06,
      "loss": 0.8343,
      "step": 160
    },
    {
      "epoch": 0.0680973999174577,
      "grad_norm": 3.057886812615224,
      "learning_rate": 6.790123456790124e-06,
      "loss": 0.833,
      "step": 165
    },
    {
      "epoch": 0.070160957490714,
      "grad_norm": 3.12354443289718,
      "learning_rate": 6.9958847736625525e-06,
      "loss": 0.8445,
      "step": 170
    },
    {
      "epoch": 0.07222451506397029,
      "grad_norm": 3.1566533318265515,
      "learning_rate": 7.201646090534981e-06,
      "loss": 0.8336,
      "step": 175
    },
    {
      "epoch": 0.07428807263722657,
      "grad_norm": 3.0910067969250288,
      "learning_rate": 7.4074074074074075e-06,
      "loss": 0.8335,
      "step": 180
    },
    {
      "epoch": 0.07635163021048287,
      "grad_norm": 3.2341343019009754,
      "learning_rate": 7.613168724279836e-06,
      "loss": 0.8313,
      "step": 185
    },
    {
      "epoch": 0.07841518778373917,
      "grad_norm": 2.9443616643904984,
      "learning_rate": 7.818930041152263e-06,
      "loss": 0.8271,
      "step": 190
    },
    {
      "epoch": 0.08047874535699547,
      "grad_norm": 2.955202520255723,
      "learning_rate": 8.024691358024692e-06,
      "loss": 0.833,
      "step": 195
    },
    {
      "epoch": 0.08254230293025175,
      "grad_norm": 3.0004205685291896,
      "learning_rate": 8.23045267489712e-06,
      "loss": 0.8046,
      "step": 200
    },
    {
      "epoch": 0.08460586050350805,
      "grad_norm": 3.0549793246137984,
      "learning_rate": 8.43621399176955e-06,
      "loss": 0.8155,
      "step": 205
    },
    {
      "epoch": 0.08666941807676434,
      "grad_norm": 3.0535661505541896,
      "learning_rate": 8.641975308641975e-06,
      "loss": 0.8196,
      "step": 210
    },
    {
      "epoch": 0.08873297565002064,
      "grad_norm": 3.116053820452697,
      "learning_rate": 8.847736625514404e-06,
      "loss": 0.8347,
      "step": 215
    },
    {
      "epoch": 0.09079653322327692,
      "grad_norm": 3.022529098579132,
      "learning_rate": 9.053497942386832e-06,
      "loss": 0.8081,
      "step": 220
    },
    {
      "epoch": 0.09286009079653322,
      "grad_norm": 3.0433063189548792,
      "learning_rate": 9.25925925925926e-06,
      "loss": 0.8208,
      "step": 225
    },
    {
      "epoch": 0.09492364836978952,
      "grad_norm": 3.085322202490418,
      "learning_rate": 9.465020576131688e-06,
      "loss": 0.8339,
      "step": 230
    },
    {
      "epoch": 0.09698720594304581,
      "grad_norm": 3.6047031312875584,
      "learning_rate": 9.670781893004116e-06,
      "loss": 0.8181,
      "step": 235
    },
    {
      "epoch": 0.0990507635163021,
      "grad_norm": 3.0604827472581255,
      "learning_rate": 9.876543209876543e-06,
      "loss": 0.8166,
      "step": 240
    },
    {
      "epoch": 0.1011143210895584,
      "grad_norm": 3.2691636345258277,
      "learning_rate": 9.999979232392962e-06,
      "loss": 0.799,
      "step": 245
    },
    {
      "epoch": 0.10317787866281469,
      "grad_norm": 2.9570633048095964,
      "learning_rate": 9.999745598795033e-06,
      "loss": 0.8205,
      "step": 250
    },
    {
      "epoch": 0.10524143623607099,
      "grad_norm": 3.335892639195958,
      "learning_rate": 9.999252384260794e-06,
      "loss": 0.7989,
      "step": 255
    },
    {
      "epoch": 0.10730499380932729,
      "grad_norm": 2.9248399326152725,
      "learning_rate": 9.998499614397364e-06,
      "loss": 0.8063,
      "step": 260
    },
    {
      "epoch": 0.10936855138258357,
      "grad_norm": 3.0768852625514422,
      "learning_rate": 9.997487328287675e-06,
      "loss": 0.7968,
      "step": 265
    },
    {
      "epoch": 0.11143210895583987,
      "grad_norm": 3.002775395507999,
      "learning_rate": 9.996215578488434e-06,
      "loss": 0.8062,
      "step": 270
    },
    {
      "epoch": 0.11349566652909616,
      "grad_norm": 2.9801626597554174,
      "learning_rate": 9.994684431027407e-06,
      "loss": 0.7844,
      "step": 275
    },
    {
      "epoch": 0.11555922410235246,
      "grad_norm": 3.1190858706402604,
      "learning_rate": 9.99289396539997e-06,
      "loss": 0.8089,
      "step": 280
    },
    {
      "epoch": 0.11762278167560875,
      "grad_norm": 2.785366704095957,
      "learning_rate": 9.990844274565004e-06,
      "loss": 0.7997,
      "step": 285
    },
    {
      "epoch": 0.11968633924886504,
      "grad_norm": 3.175006190343879,
      "learning_rate": 9.98853546494006e-06,
      "loss": 0.8186,
      "step": 290
    },
    {
      "epoch": 0.12174989682212134,
      "grad_norm": 2.883637899845788,
      "learning_rate": 9.985967656395823e-06,
      "loss": 0.7889,
      "step": 295
    },
    {
      "epoch": 0.12381345439537764,
      "grad_norm": 2.991597043168701,
      "learning_rate": 9.983140982249913e-06,
      "loss": 0.8136,
      "step": 300
    },
    {
      "epoch": 0.12587701196863393,
      "grad_norm": 2.967721245447412,
      "learning_rate": 9.980055589259937e-06,
      "loss": 0.7804,
      "step": 305
    },
    {
      "epoch": 0.12794056954189023,
      "grad_norm": 2.925851320281598,
      "learning_rate": 9.976711637615886e-06,
      "loss": 0.7789,
      "step": 310
    },
    {
      "epoch": 0.1300041271151465,
      "grad_norm": 3.015507414806422,
      "learning_rate": 9.973109300931813e-06,
      "loss": 0.7821,
      "step": 315
    },
    {
      "epoch": 0.1320676846884028,
      "grad_norm": 3.076125191679388,
      "learning_rate": 9.969248766236823e-06,
      "loss": 0.7668,
      "step": 320
    },
    {
      "epoch": 0.1341312422616591,
      "grad_norm": 2.9816750909346137,
      "learning_rate": 9.965130233965353e-06,
      "loss": 0.7867,
      "step": 325
    },
    {
      "epoch": 0.1361947998349154,
      "grad_norm": 3.1216143549392514,
      "learning_rate": 9.960753917946777e-06,
      "loss": 0.7827,
      "step": 330
    },
    {
      "epoch": 0.1382583574081717,
      "grad_norm": 3.125152909470307,
      "learning_rate": 9.956120045394297e-06,
      "loss": 0.7683,
      "step": 335
    },
    {
      "epoch": 0.140321914981428,
      "grad_norm": 2.8419491540297694,
      "learning_rate": 9.951228856893152e-06,
      "loss": 0.7571,
      "step": 340
    },
    {
      "epoch": 0.14238547255468428,
      "grad_norm": 2.739631493281793,
      "learning_rate": 9.946080606388115e-06,
      "loss": 0.7743,
      "step": 345
    },
    {
      "epoch": 0.14444903012794058,
      "grad_norm": 2.888837107470469,
      "learning_rate": 9.940675561170328e-06,
      "loss": 0.7795,
      "step": 350
    },
    {
      "epoch": 0.14651258770119685,
      "grad_norm": 3.0512065932946886,
      "learning_rate": 9.935014001863405e-06,
      "loss": 0.7753,
      "step": 355
    },
    {
      "epoch": 0.14857614527445315,
      "grad_norm": 3.029181890310092,
      "learning_rate": 9.92909622240888e-06,
      "loss": 0.7515,
      "step": 360
    },
    {
      "epoch": 0.15063970284770944,
      "grad_norm": 2.9755768091749015,
      "learning_rate": 9.92292253005093e-06,
      "loss": 0.7564,
      "step": 365
    },
    {
      "epoch": 0.15270326042096574,
      "grad_norm": 2.8483284554348565,
      "learning_rate": 9.916493245320428e-06,
      "loss": 0.7642,
      "step": 370
    },
    {
      "epoch": 0.15476681799422204,
      "grad_norm": 3.2241540408474,
      "learning_rate": 9.909808702018315e-06,
      "loss": 0.758,
      "step": 375
    },
    {
      "epoch": 0.15683037556747834,
      "grad_norm": 2.871363873496698,
      "learning_rate": 9.902869247198246e-06,
      "loss": 0.7595,
      "step": 380
    },
    {
      "epoch": 0.15889393314073463,
      "grad_norm": 2.9539374709865607,
      "learning_rate": 9.895675241148588e-06,
      "loss": 0.7518,
      "step": 385
    },
    {
      "epoch": 0.16095749071399093,
      "grad_norm": 2.9560555474503043,
      "learning_rate": 9.888227057373716e-06,
      "loss": 0.7493,
      "step": 390
    },
    {
      "epoch": 0.16302104828724723,
      "grad_norm": 2.7957300167557926,
      "learning_rate": 9.880525082574604e-06,
      "loss": 0.7476,
      "step": 395
    },
    {
      "epoch": 0.1650846058605035,
      "grad_norm": 2.6981062487306606,
      "learning_rate": 9.872569716628764e-06,
      "loss": 0.7273,
      "step": 400
    },
    {
      "epoch": 0.1671481634337598,
      "grad_norm": 2.7734124248259366,
      "learning_rate": 9.86436137256948e-06,
      "loss": 0.7297,
      "step": 405
    },
    {
      "epoch": 0.1692117210070161,
      "grad_norm": 2.73464344090408,
      "learning_rate": 9.855900476564365e-06,
      "loss": 0.7445,
      "step": 410
    },
    {
      "epoch": 0.1712752785802724,
      "grad_norm": 2.7258763684985303,
      "learning_rate": 9.847187467893228e-06,
      "loss": 0.7343,
      "step": 415
    },
    {
      "epoch": 0.17333883615352869,
      "grad_norm": 2.6760448573907354,
      "learning_rate": 9.83822279892528e-06,
      "loss": 0.7451,
      "step": 420
    },
    {
      "epoch": 0.17540239372678498,
      "grad_norm": 2.782701927749318,
      "learning_rate": 9.829006935095629e-06,
      "loss": 0.7375,
      "step": 425
    },
    {
      "epoch": 0.17746595130004128,
      "grad_norm": 2.72035422100488,
      "learning_rate": 9.819540354881136e-06,
      "loss": 0.7385,
      "step": 430
    },
    {
      "epoch": 0.17952950887329758,
      "grad_norm": 2.928620231770996,
      "learning_rate": 9.809823549775559e-06,
      "loss": 0.7194,
      "step": 435
    },
    {
      "epoch": 0.18159306644655385,
      "grad_norm": 3.127420500265884,
      "learning_rate": 9.79985702426404e-06,
      "loss": 0.7191,
      "step": 440
    },
    {
      "epoch": 0.18365662401981014,
      "grad_norm": 3.075733323456217,
      "learning_rate": 9.78964129579691e-06,
      "loss": 0.7327,
      "step": 445
    },
    {
      "epoch": 0.18572018159306644,
      "grad_norm": 2.9108156590411336,
      "learning_rate": 9.779176894762833e-06,
      "loss": 0.7314,
      "step": 450
    },
    {
      "epoch": 0.18778373916632274,
      "grad_norm": 2.946027511832701,
      "learning_rate": 9.768464364461248e-06,
      "loss": 0.7154,
      "step": 455
    },
    {
      "epoch": 0.18984729673957904,
      "grad_norm": 2.9516974484587863,
      "learning_rate": 9.75750426107419e-06,
      "loss": 0.6993,
      "step": 460
    },
    {
      "epoch": 0.19191085431283533,
      "grad_norm": 2.8592349880443897,
      "learning_rate": 9.746297153637386e-06,
      "loss": 0.7116,
      "step": 465
    },
    {
      "epoch": 0.19397441188609163,
      "grad_norm": 2.9554239306486596,
      "learning_rate": 9.73484362401073e-06,
      "loss": 0.6998,
      "step": 470
    },
    {
      "epoch": 0.19603796945934793,
      "grad_norm": 2.780540575615214,
      "learning_rate": 9.723144266848073e-06,
      "loss": 0.7175,
      "step": 475
    },
    {
      "epoch": 0.1981015270326042,
      "grad_norm": 2.7790855329011657,
      "learning_rate": 9.71119968956633e-06,
      "loss": 0.7061,
      "step": 480
    },
    {
      "epoch": 0.2001650846058605,
      "grad_norm": 2.811087864792907,
      "learning_rate": 9.69901051231397e-06,
      "loss": 0.7239,
      "step": 485
    },
    {
      "epoch": 0.2022286421791168,
      "grad_norm": 2.7071720679814937,
      "learning_rate": 9.686577367938802e-06,
      "loss": 0.6857,
      "step": 490
    },
    {
      "epoch": 0.2042921997523731,
      "grad_norm": 2.9052904699729654,
      "learning_rate": 9.673900901955118e-06,
      "loss": 0.7026,
      "step": 495
    },
    {
      "epoch": 0.20635575732562939,
      "grad_norm": 2.7428004689795284,
      "learning_rate": 9.66098177251019e-06,
      "loss": 0.7043,
      "step": 500
    },
    {
      "epoch": 0.20841931489888568,
      "grad_norm": 2.9542233401997797,
      "learning_rate": 9.647820650350087e-06,
      "loss": 0.7008,
      "step": 505
    },
    {
      "epoch": 0.21048287247214198,
      "grad_norm": 2.8206780344207005,
      "learning_rate": 9.634418218784856e-06,
      "loss": 0.6827,
      "step": 510
    },
    {
      "epoch": 0.21254643004539828,
      "grad_norm": 2.7820042566137917,
      "learning_rate": 9.620775173653055e-06,
      "loss": 0.6905,
      "step": 515
    },
    {
      "epoch": 0.21460998761865457,
      "grad_norm": 2.8617753014266354,
      "learning_rate": 9.606892223285604e-06,
      "loss": 0.6875,
      "step": 520
    },
    {
      "epoch": 0.21667354519191084,
      "grad_norm": 2.818801617888441,
      "learning_rate": 9.592770088469032e-06,
      "loss": 0.7143,
      "step": 525
    },
    {
      "epoch": 0.21873710276516714,
      "grad_norm": 2.803399949113071,
      "learning_rate": 9.578409502408037e-06,
      "loss": 0.6596,
      "step": 530
    },
    {
      "epoch": 0.22080066033842344,
      "grad_norm": 2.832046830498785,
      "learning_rate": 9.563811210687433e-06,
      "loss": 0.6912,
      "step": 535
    },
    {
      "epoch": 0.22286421791167974,
      "grad_norm": 2.7793622312334842,
      "learning_rate": 9.548975971233427e-06,
      "loss": 0.6945,
      "step": 540
    },
    {
      "epoch": 0.22492777548493603,
      "grad_norm": 2.8345021788595073,
      "learning_rate": 9.53390455427428e-06,
      "loss": 0.6767,
      "step": 545
    },
    {
      "epoch": 0.22699133305819233,
      "grad_norm": 2.7737306964899284,
      "learning_rate": 9.518597742300309e-06,
      "loss": 0.6816,
      "step": 550
    },
    {
      "epoch": 0.22905489063144863,
      "grad_norm": 2.7181943065638388,
      "learning_rate": 9.503056330023267e-06,
      "loss": 0.6797,
      "step": 555
    },
    {
      "epoch": 0.23111844820470492,
      "grad_norm": 2.935392215628063,
      "learning_rate": 9.48728112433508e-06,
      "loss": 0.6993,
      "step": 560
    },
    {
      "epoch": 0.2331820057779612,
      "grad_norm": 3.339231195371882,
      "learning_rate": 9.471272944265948e-06,
      "loss": 0.6565,
      "step": 565
    },
    {
      "epoch": 0.2352455633512175,
      "grad_norm": 2.588226117443037,
      "learning_rate": 9.45503262094184e-06,
      "loss": 0.6685,
      "step": 570
    },
    {
      "epoch": 0.2373091209244738,
      "grad_norm": 2.7591510109484507,
      "learning_rate": 9.438560997541319e-06,
      "loss": 0.6712,
      "step": 575
    },
    {
      "epoch": 0.23937267849773008,
      "grad_norm": 2.5942405830165125,
      "learning_rate": 9.421858929251786e-06,
      "loss": 0.6595,
      "step": 580
    },
    {
      "epoch": 0.24143623607098638,
      "grad_norm": 3.1536570604359064,
      "learning_rate": 9.404927283225064e-06,
      "loss": 0.6624,
      "step": 585
    },
    {
      "epoch": 0.24349979364424268,
      "grad_norm": 2.6761790991992402,
      "learning_rate": 9.387766938532386e-06,
      "loss": 0.6481,
      "step": 590
    },
    {
      "epoch": 0.24556335121749898,
      "grad_norm": 2.559657538042446,
      "learning_rate": 9.370378786118755e-06,
      "loss": 0.6518,
      "step": 595
    },
    {
      "epoch": 0.24762690879075527,
      "grad_norm": 2.70639753951879,
      "learning_rate": 9.352763728756677e-06,
      "loss": 0.6704,
      "step": 600
    },
    {
      "epoch": 0.24969046636401154,
      "grad_norm": 2.7368008412010703,
      "learning_rate": 9.334922680999304e-06,
      "loss": 0.6621,
      "step": 605
    },
    {
      "epoch": 0.25175402393726787,
      "grad_norm": 2.852828202511116,
      "learning_rate": 9.316856569132942e-06,
      "loss": 0.6552,
      "step": 610
    },
    {
      "epoch": 0.25381758151052414,
      "grad_norm": 2.5928866487990057,
      "learning_rate": 9.29856633112896e-06,
      "loss": 0.666,
      "step": 615
    },
    {
      "epoch": 0.25588113908378046,
      "grad_norm": 2.7913637236454574,
      "learning_rate": 9.280052916595098e-06,
      "loss": 0.6661,
      "step": 620
    },
    {
      "epoch": 0.25794469665703673,
      "grad_norm": 2.5894434588677164,
      "learning_rate": 9.261317286726157e-06,
      "loss": 0.6654,
      "step": 625
    },
    {
      "epoch": 0.260008254230293,
      "grad_norm": 2.628011843241579,
      "learning_rate": 9.242360414254098e-06,
      "loss": 0.6649,
      "step": 630
    },
    {
      "epoch": 0.2620718118035493,
      "grad_norm": 2.7172847524203636,
      "learning_rate": 9.223183283397538e-06,
      "loss": 0.6662,
      "step": 635
    },
    {
      "epoch": 0.2641353693768056,
      "grad_norm": 2.659395752307675,
      "learning_rate": 9.203786889810655e-06,
      "loss": 0.6285,
      "step": 640
    },
    {
      "epoch": 0.2661989269500619,
      "grad_norm": 2.90375149194139,
      "learning_rate": 9.18417224053149e-06,
      "loss": 0.6479,
      "step": 645
    },
    {
      "epoch": 0.2682624845233182,
      "grad_norm": 2.6966774373935927,
      "learning_rate": 9.16434035392966e-06,
      "loss": 0.6397,
      "step": 650
    },
    {
      "epoch": 0.2703260420965745,
      "grad_norm": 2.7576279699971713,
      "learning_rate": 9.144292259653493e-06,
      "loss": 0.651,
      "step": 655
    },
    {
      "epoch": 0.2723895996698308,
      "grad_norm": 2.8455714787136364,
      "learning_rate": 9.124028998576568e-06,
      "loss": 0.634,
      "step": 660
    },
    {
      "epoch": 0.2744531572430871,
      "grad_norm": 2.524900004589425,
      "learning_rate": 9.103551622743667e-06,
      "loss": 0.6393,
      "step": 665
    },
    {
      "epoch": 0.2765167148163434,
      "grad_norm": 2.672726083965731,
      "learning_rate": 9.082861195316164e-06,
      "loss": 0.6321,
      "step": 670
    },
    {
      "epoch": 0.27858027238959965,
      "grad_norm": 3.0535454448794637,
      "learning_rate": 9.061958790516821e-06,
      "loss": 0.6219,
      "step": 675
    },
    {
      "epoch": 0.280643829962856,
      "grad_norm": 2.961262579490319,
      "learning_rate": 9.040845493574016e-06,
      "loss": 0.6239,
      "step": 680
    },
    {
      "epoch": 0.28270738753611224,
      "grad_norm": 2.867116023710867,
      "learning_rate": 9.019522400665397e-06,
      "loss": 0.6183,
      "step": 685
    },
    {
      "epoch": 0.28477094510936857,
      "grad_norm": 2.70507865608555,
      "learning_rate": 8.99799061886098e-06,
      "loss": 0.6119,
      "step": 690
    },
    {
      "epoch": 0.28683450268262484,
      "grad_norm": 2.7686804191515404,
      "learning_rate": 8.976251266065663e-06,
      "loss": 0.6205,
      "step": 695
    },
    {
      "epoch": 0.28889806025588116,
      "grad_norm": 2.7663678432556926,
      "learning_rate": 8.95430547096118e-06,
      "loss": 0.6197,
      "step": 700
    },
    {
      "epoch": 0.29096161782913743,
      "grad_norm": 2.6693543634519115,
      "learning_rate": 8.932154372947512e-06,
      "loss": 0.616,
      "step": 705
    },
    {
      "epoch": 0.2930251754023937,
      "grad_norm": 2.606058194100552,
      "learning_rate": 8.90979912208373e-06,
      "loss": 0.5959,
      "step": 710
    },
    {
      "epoch": 0.29508873297565,
      "grad_norm": 2.9126080020507903,
      "learning_rate": 8.887240879028276e-06,
      "loss": 0.6109,
      "step": 715
    },
    {
      "epoch": 0.2971522905489063,
      "grad_norm": 2.712632905481665,
      "learning_rate": 8.864480814978715e-06,
      "loss": 0.6176,
      "step": 720
    },
    {
      "epoch": 0.2992158481221626,
      "grad_norm": 2.92931205166235,
      "learning_rate": 8.841520111610914e-06,
      "loss": 0.6152,
      "step": 725
    },
    {
      "epoch": 0.3012794056954189,
      "grad_norm": 2.757497057768333,
      "learning_rate": 8.818359961017705e-06,
      "loss": 0.6269,
      "step": 730
    },
    {
      "epoch": 0.3033429632686752,
      "grad_norm": 2.911039679371386,
      "learning_rate": 8.795001565646983e-06,
      "loss": 0.617,
      "step": 735
    },
    {
      "epoch": 0.3054065208419315,
      "grad_norm": 2.6754121417561922,
      "learning_rate": 8.771446138239282e-06,
      "loss": 0.6079,
      "step": 740
    },
    {
      "epoch": 0.3074700784151878,
      "grad_norm": 2.9820721402144947,
      "learning_rate": 8.747694901764807e-06,
      "loss": 0.6028,
      "step": 745
    },
    {
      "epoch": 0.3095336359884441,
      "grad_norm": 2.7199956091577353,
      "learning_rate": 8.723749089359941e-06,
      "loss": 0.613,
      "step": 750
    },
    {
      "epoch": 0.31159719356170035,
      "grad_norm": 2.587929517491648,
      "learning_rate": 8.699609944263219e-06,
      "loss": 0.6113,
      "step": 755
    },
    {
      "epoch": 0.3136607511349567,
      "grad_norm": 2.7061730588333335,
      "learning_rate": 8.675278719750788e-06,
      "loss": 0.6027,
      "step": 760
    },
    {
      "epoch": 0.31572430870821294,
      "grad_norm": 2.79901699476363,
      "learning_rate": 8.65075667907133e-06,
      "loss": 0.6025,
      "step": 765
    },
    {
      "epoch": 0.31778786628146927,
      "grad_norm": 2.7737439842011193,
      "learning_rate": 8.626045095380477e-06,
      "loss": 0.6044,
      "step": 770
    },
    {
      "epoch": 0.31985142385472554,
      "grad_norm": 2.6503735236627453,
      "learning_rate": 8.601145251674718e-06,
      "loss": 0.5873,
      "step": 775
    },
    {
      "epoch": 0.32191498142798186,
      "grad_norm": 2.8347160208255078,
      "learning_rate": 8.576058440724777e-06,
      "loss": 0.5828,
      "step": 780
    },
    {
      "epoch": 0.32397853900123813,
      "grad_norm": 2.8544698266992996,
      "learning_rate": 8.5507859650085e-06,
      "loss": 0.5939,
      "step": 785
    },
    {
      "epoch": 0.32604209657449446,
      "grad_norm": 2.686873043228608,
      "learning_rate": 8.525329136643227e-06,
      "loss": 0.5821,
      "step": 790
    },
    {
      "epoch": 0.3281056541477507,
      "grad_norm": 2.8021270796404,
      "learning_rate": 8.499689277317675e-06,
      "loss": 0.5871,
      "step": 795
    },
    {
      "epoch": 0.330169211721007,
      "grad_norm": 2.5870353213265402,
      "learning_rate": 8.473867718223317e-06,
      "loss": 0.5813,
      "step": 800
    },
    {
      "epoch": 0.3322327692942633,
      "grad_norm": 2.7498697085626844,
      "learning_rate": 8.447865799985258e-06,
      "loss": 0.5937,
      "step": 805
    },
    {
      "epoch": 0.3342963268675196,
      "grad_norm": 2.6169464644713014,
      "learning_rate": 8.421684872592643e-06,
      "loss": 0.5769,
      "step": 810
    },
    {
      "epoch": 0.3363598844407759,
      "grad_norm": 2.7371393784664804,
      "learning_rate": 8.395326295328562e-06,
      "loss": 0.5806,
      "step": 815
    },
    {
      "epoch": 0.3384234420140322,
      "grad_norm": 2.6465634165278575,
      "learning_rate": 8.368791436699482e-06,
      "loss": 0.5914,
      "step": 820
    },
    {
      "epoch": 0.3404869995872885,
      "grad_norm": 2.57951699963865,
      "learning_rate": 8.342081674364182e-06,
      "loss": 0.5711,
      "step": 825
    },
    {
      "epoch": 0.3425505571605448,
      "grad_norm": 2.726507330124962,
      "learning_rate": 8.315198395062246e-06,
      "loss": 0.5789,
      "step": 830
    },
    {
      "epoch": 0.34461411473380105,
      "grad_norm": 2.6932144440323267,
      "learning_rate": 8.28814299454205e-06,
      "loss": 0.5868,
      "step": 835
    },
    {
      "epoch": 0.34667767230705737,
      "grad_norm": 2.6545019349833052,
      "learning_rate": 8.260916877488308e-06,
      "loss": 0.585,
      "step": 840
    },
    {
      "epoch": 0.34874122988031364,
      "grad_norm": 2.6267015172396544,
      "learning_rate": 8.233521457449131e-06,
      "loss": 0.5713,
      "step": 845
    },
    {
      "epoch": 0.35080478745356997,
      "grad_norm": 2.446024579716919,
      "learning_rate": 8.205958156762647e-06,
      "loss": 0.5668,
      "step": 850
    },
    {
      "epoch": 0.35286834502682624,
      "grad_norm": 2.5506551014516208,
      "learning_rate": 8.178228406483145e-06,
      "loss": 0.5683,
      "step": 855
    },
    {
      "epoch": 0.35493190260008256,
      "grad_norm": 2.5964627157082,
      "learning_rate": 8.150333646306787e-06,
      "loss": 0.5631,
      "step": 860
    },
    {
      "epoch": 0.35699546017333883,
      "grad_norm": 2.7366272191977212,
      "learning_rate": 8.122275324496855e-06,
      "loss": 0.5854,
      "step": 865
    },
    {
      "epoch": 0.35905901774659515,
      "grad_norm": 2.470623506679956,
      "learning_rate": 8.094054897808556e-06,
      "loss": 0.565,
      "step": 870
    },
    {
      "epoch": 0.3611225753198514,
      "grad_norm": 2.6937122631536496,
      "learning_rate": 8.065673831413396e-06,
      "loss": 0.5644,
      "step": 875
    },
    {
      "epoch": 0.3631861328931077,
      "grad_norm": 3.8142387103231847,
      "learning_rate": 8.037133598823107e-06,
      "loss": 0.5799,
      "step": 880
    },
    {
      "epoch": 0.365249690466364,
      "grad_norm": 2.760510177003008,
      "learning_rate": 8.008435681813139e-06,
      "loss": 0.566,
      "step": 885
    },
    {
      "epoch": 0.3673132480396203,
      "grad_norm": 2.6376344821820354,
      "learning_rate": 7.979581570345737e-06,
      "loss": 0.572,
      "step": 890
    },
    {
      "epoch": 0.3693768056128766,
      "grad_norm": 2.5348136936661043,
      "learning_rate": 7.950572762492577e-06,
      "loss": 0.5529,
      "step": 895
    },
    {
      "epoch": 0.3714403631861329,
      "grad_norm": 2.551175872300915,
      "learning_rate": 7.92141076435699e-06,
      "loss": 0.5593,
      "step": 900
    },
    {
      "epoch": 0.3735039207593892,
      "grad_norm": 2.609535679611351,
      "learning_rate": 7.892097089995767e-06,
      "loss": 0.5491,
      "step": 905
    },
    {
      "epoch": 0.3755674783326455,
      "grad_norm": 2.670033948965714,
      "learning_rate": 7.862633261340556e-06,
      "loss": 0.5446,
      "step": 910
    },
    {
      "epoch": 0.3776310359059018,
      "grad_norm": 2.406170232856436,
      "learning_rate": 7.83302080811883e-06,
      "loss": 0.535,
      "step": 915
    },
    {
      "epoch": 0.37969459347915807,
      "grad_norm": 2.5973677086378024,
      "learning_rate": 7.803261267774484e-06,
      "loss": 0.5501,
      "step": 920
    },
    {
      "epoch": 0.38175815105241434,
      "grad_norm": 2.7913389668469977,
      "learning_rate": 7.773356185388003e-06,
      "loss": 0.5358,
      "step": 925
    },
    {
      "epoch": 0.38382170862567067,
      "grad_norm": 2.395012173415671,
      "learning_rate": 7.74330711359624e-06,
      "loss": 0.5492,
      "step": 930
    },
    {
      "epoch": 0.38588526619892694,
      "grad_norm": 2.6947690399272153,
      "learning_rate": 7.713115612511815e-06,
      "loss": 0.5481,
      "step": 935
    },
    {
      "epoch": 0.38794882377218326,
      "grad_norm": 2.711638332847109,
      "learning_rate": 7.682783249642104e-06,
      "loss": 0.5533,
      "step": 940
    },
    {
      "epoch": 0.39001238134543953,
      "grad_norm": 2.616283183023565,
      "learning_rate": 7.652311599807868e-06,
      "loss": 0.5506,
      "step": 945
    },
    {
      "epoch": 0.39207593891869585,
      "grad_norm": 2.4832951603969162,
      "learning_rate": 7.62170224506148e-06,
      "loss": 0.5316,
      "step": 950
    },
    {
      "epoch": 0.3941394964919521,
      "grad_norm": 2.6721377852076804,
      "learning_rate": 7.590956774604791e-06,
      "loss": 0.5458,
      "step": 955
    },
    {
      "epoch": 0.3962030540652084,
      "grad_norm": 2.777241124185147,
      "learning_rate": 7.5600767847066215e-06,
      "loss": 0.5527,
      "step": 960
    },
    {
      "epoch": 0.3982666116384647,
      "grad_norm": 2.5914713737840924,
      "learning_rate": 7.529063878619886e-06,
      "loss": 0.5484,
      "step": 965
    },
    {
      "epoch": 0.400330169211721,
      "grad_norm": 2.573321705871262,
      "learning_rate": 7.497919666498346e-06,
      "loss": 0.5435,
      "step": 970
    },
    {
      "epoch": 0.4023937267849773,
      "grad_norm": 2.6520337314396993,
      "learning_rate": 7.466645765313023e-06,
      "loss": 0.5358,
      "step": 975
    },
    {
      "epoch": 0.4044572843582336,
      "grad_norm": 2.7056801557526486,
      "learning_rate": 7.435243798768242e-06,
      "loss": 0.5499,
      "step": 980
    },
    {
      "epoch": 0.4065208419314899,
      "grad_norm": 2.6559172388700047,
      "learning_rate": 7.403715397217335e-06,
      "loss": 0.5241,
      "step": 985
    },
    {
      "epoch": 0.4085843995047462,
      "grad_norm": 2.5216304625498647,
      "learning_rate": 7.3720621975779815e-06,
      "loss": 0.521,
      "step": 990
    },
    {
      "epoch": 0.4106479570780025,
      "grad_norm": 2.519874224942243,
      "learning_rate": 7.3402858432472416e-06,
      "loss": 0.5489,
      "step": 995
    },
    {
      "epoch": 0.41271151465125877,
      "grad_norm": 2.518659981447909,
      "learning_rate": 7.308387984016217e-06,
      "loss": 0.5266,
      "step": 1000
    },
    {
      "epoch": 0.41477507222451504,
      "grad_norm": 2.721359929344381,
      "learning_rate": 7.276370275984408e-06,
      "loss": 0.5325,
      "step": 1005
    },
    {
      "epoch": 0.41683862979777137,
      "grad_norm": 2.575184642938111,
      "learning_rate": 7.2442343814737135e-06,
      "loss": 0.5357,
      "step": 1010
    },
    {
      "epoch": 0.41890218737102763,
      "grad_norm": 2.5077385160634904,
      "learning_rate": 7.211981968942147e-06,
      "loss": 0.524,
      "step": 1015
    },
    {
      "epoch": 0.42096574494428396,
      "grad_norm": 2.453550934446932,
      "learning_rate": 7.179614712897195e-06,
      "loss": 0.5366,
      "step": 1020
    },
    {
      "epoch": 0.42302930251754023,
      "grad_norm": 2.561635348596983,
      "learning_rate": 7.147134293808885e-06,
      "loss": 0.5117,
      "step": 1025
    },
    {
      "epoch": 0.42509286009079655,
      "grad_norm": 2.501066318480838,
      "learning_rate": 7.114542398022539e-06,
      "loss": 0.524,
      "step": 1030
    },
    {
      "epoch": 0.4271564176640528,
      "grad_norm": 2.7215445820801714,
      "learning_rate": 7.08184071767122e-06,
      "loss": 0.5111,
      "step": 1035
    },
    {
      "epoch": 0.42921997523730915,
      "grad_norm": 2.6088491234313604,
      "learning_rate": 7.049030950587874e-06,
      "loss": 0.5113,
      "step": 1040
    },
    {
      "epoch": 0.4312835328105654,
      "grad_norm": 2.551009519200941,
      "learning_rate": 7.016114800217182e-06,
      "loss": 0.5136,
      "step": 1045
    },
    {
      "epoch": 0.4333470903838217,
      "grad_norm": 2.457255947220057,
      "learning_rate": 6.983093975527128e-06,
      "loss": 0.5208,
      "step": 1050
    },
    {
      "epoch": 0.435410647957078,
      "grad_norm": 2.6994910463118913,
      "learning_rate": 6.949970190920255e-06,
      "loss": 0.5212,
      "step": 1055
    },
    {
      "epoch": 0.4374742055303343,
      "grad_norm": 2.597397073975052,
      "learning_rate": 6.916745166144671e-06,
      "loss": 0.5128,
      "step": 1060
    },
    {
      "epoch": 0.4395377631035906,
      "grad_norm": 2.605039455033872,
      "learning_rate": 6.883420626204751e-06,
      "loss": 0.5081,
      "step": 1065
    },
    {
      "epoch": 0.4416013206768469,
      "grad_norm": 2.7509624680821,
      "learning_rate": 6.849998301271585e-06,
      "loss": 0.516,
      "step": 1070
    },
    {
      "epoch": 0.4436648782501032,
      "grad_norm": 2.4731078067929197,
      "learning_rate": 6.81647992659314e-06,
      "loss": 0.5144,
      "step": 1075
    },
    {
      "epoch": 0.44572843582335947,
      "grad_norm": 2.611162217133499,
      "learning_rate": 6.782867242404181e-06,
      "loss": 0.4979,
      "step": 1080
    },
    {
      "epoch": 0.44779199339661574,
      "grad_norm": 2.5398801893032474,
      "learning_rate": 6.749161993835901e-06,
      "loss": 0.5074,
      "step": 1085
    },
    {
      "epoch": 0.44985555096987206,
      "grad_norm": 2.5951566561368775,
      "learning_rate": 6.715365930825337e-06,
      "loss": 0.4975,
      "step": 1090
    },
    {
      "epoch": 0.45191910854312833,
      "grad_norm": 2.6092895827410034,
      "learning_rate": 6.681480808024503e-06,
      "loss": 0.5124,
      "step": 1095
    },
    {
      "epoch": 0.45398266611638466,
      "grad_norm": 2.5414548197736706,
      "learning_rate": 6.6475083847092895e-06,
      "loss": 0.4957,
      "step": 1100
    },
    {
      "epoch": 0.45604622368964093,
      "grad_norm": 2.8460005959716166,
      "learning_rate": 6.613450424688131e-06,
      "loss": 0.5028,
      "step": 1105
    },
    {
      "epoch": 0.45810978126289725,
      "grad_norm": 2.5601376301451753,
      "learning_rate": 6.579308696210428e-06,
      "loss": 0.4876,
      "step": 1110
    },
    {
      "epoch": 0.4601733388361535,
      "grad_norm": 2.6807233381503237,
      "learning_rate": 6.545084971874738e-06,
      "loss": 0.5155,
      "step": 1115
    },
    {
      "epoch": 0.46223689640940985,
      "grad_norm": 2.5044297534154336,
      "learning_rate": 6.51078102853675e-06,
      "loss": 0.5012,
      "step": 1120
    },
    {
      "epoch": 0.4643004539826661,
      "grad_norm": 2.5092897248073474,
      "learning_rate": 6.47639864721703e-06,
      "loss": 0.493,
      "step": 1125
    },
    {
      "epoch": 0.4663640115559224,
      "grad_norm": 2.4954489446094392,
      "learning_rate": 6.44193961300855e-06,
      "loss": 0.4868,
      "step": 1130
    },
    {
      "epoch": 0.4684275691291787,
      "grad_norm": 2.5451475237887666,
      "learning_rate": 6.407405714984011e-06,
      "loss": 0.4866,
      "step": 1135
    },
    {
      "epoch": 0.470491126702435,
      "grad_norm": 2.6098482720462552,
      "learning_rate": 6.3727987461029565e-06,
      "loss": 0.4956,
      "step": 1140
    },
    {
      "epoch": 0.4725546842756913,
      "grad_norm": 2.5923597566181296,
      "learning_rate": 6.338120503118684e-06,
      "loss": 0.4998,
      "step": 1145
    },
    {
      "epoch": 0.4746182418489476,
      "grad_norm": 2.7721513261613393,
      "learning_rate": 6.303372786484953e-06,
      "loss": 0.4948,
      "step": 1150
    },
    {
      "epoch": 0.4766817994222039,
      "grad_norm": 2.4572497473964012,
      "learning_rate": 6.2685574002625235e-06,
      "loss": 0.4744,
      "step": 1155
    },
    {
      "epoch": 0.47874535699546017,
      "grad_norm": 2.7248530281028653,
      "learning_rate": 6.233676152025474e-06,
      "loss": 0.4789,
      "step": 1160
    },
    {
      "epoch": 0.4808089145687165,
      "grad_norm": 2.510160739180536,
      "learning_rate": 6.198730852767364e-06,
      "loss": 0.4924,
      "step": 1165
    },
    {
      "epoch": 0.48287247214197276,
      "grad_norm": 2.5611750613730275,
      "learning_rate": 6.1637233168072066e-06,
      "loss": 0.4816,
      "step": 1170
    },
    {
      "epoch": 0.48493602971522903,
      "grad_norm": 2.580939509201322,
      "learning_rate": 6.1286553616952705e-06,
      "loss": 0.4826,
      "step": 1175
    },
    {
      "epoch": 0.48699958728848536,
      "grad_norm": 2.579364116128431,
      "learning_rate": 6.09352880811872e-06,
      "loss": 0.4926,
      "step": 1180
    },
    {
      "epoch": 0.4890631448617416,
      "grad_norm": 2.5855927528887745,
      "learning_rate": 6.058345479807077e-06,
      "loss": 0.4826,
      "step": 1185
    },
    {
      "epoch": 0.49112670243499795,
      "grad_norm": 2.468639972857448,
      "learning_rate": 6.023107203437547e-06,
      "loss": 0.4822,
      "step": 1190
    },
    {
      "epoch": 0.4931902600082542,
      "grad_norm": 2.534893122365119,
      "learning_rate": 5.987815808540169e-06,
      "loss": 0.4969,
      "step": 1195
    },
    {
      "epoch": 0.49525381758151055,
      "grad_norm": 2.4035384879566095,
      "learning_rate": 5.95247312740284e-06,
      "loss": 0.4942,
      "step": 1200
    },
    {
      "epoch": 0.4973173751547668,
      "grad_norm": 2.6544116089729552,
      "learning_rate": 5.917080994976172e-06,
      "loss": 0.4678,
      "step": 1205
    },
    {
      "epoch": 0.4993809327280231,
      "grad_norm": 2.6911232973443653,
      "learning_rate": 5.881641248778235e-06,
      "loss": 0.4796,
      "step": 1210
    },
    {
      "epoch": 0.5014444903012794,
      "grad_norm": 2.9098929338992177,
      "learning_rate": 5.8461557287991455e-06,
      "loss": 0.4926,
      "step": 1215
    },
    {
      "epoch": 0.5035080478745357,
      "grad_norm": 2.724813117289689,
      "learning_rate": 5.810626277405548e-06,
      "loss": 0.4837,
      "step": 1220
    },
    {
      "epoch": 0.505571605447792,
      "grad_norm": 2.5885450608927782,
      "learning_rate": 5.775054739244948e-06,
      "loss": 0.4827,
      "step": 1225
    },
    {
      "epoch": 0.5076351630210483,
      "grad_norm": 2.4385273279955677,
      "learning_rate": 5.739442961149949e-06,
      "loss": 0.4777,
      "step": 1230
    },
    {
      "epoch": 0.5096987205943045,
      "grad_norm": 9.983765788937596,
      "learning_rate": 5.703792792042363e-06,
      "loss": 0.4657,
      "step": 1235
    },
    {
      "epoch": 0.5117622781675609,
      "grad_norm": 2.5992188623557575,
      "learning_rate": 5.668106082837226e-06,
      "loss": 0.4677,
      "step": 1240
    },
    {
      "epoch": 0.5138258357408172,
      "grad_norm": 2.5248538601870916,
      "learning_rate": 5.632384686346682e-06,
      "loss": 0.4648,
      "step": 1245
    },
    {
      "epoch": 0.5158893933140735,
      "grad_norm": 2.4528442128112653,
      "learning_rate": 5.5966304571838084e-06,
      "loss": 0.468,
      "step": 1250
    },
    {
      "epoch": 0.5179529508873297,
      "grad_norm": 2.4985756922245375,
      "learning_rate": 5.560845251666307e-06,
      "loss": 0.4691,
      "step": 1255
    },
    {
      "epoch": 0.520016508460586,
      "grad_norm": 2.588098554859599,
      "learning_rate": 5.525030927720142e-06,
      "loss": 0.4843,
      "step": 1260
    },
    {
      "epoch": 0.5220800660338424,
      "grad_norm": 2.596100924626437,
      "learning_rate": 5.489189344783072e-06,
      "loss": 0.4722,
      "step": 1265
    },
    {
      "epoch": 0.5241436236070987,
      "grad_norm": 2.4880471647467526,
      "learning_rate": 5.4533223637081075e-06,
      "loss": 0.4652,
      "step": 1270
    },
    {
      "epoch": 0.5262071811803549,
      "grad_norm": 2.5470383673418926,
      "learning_rate": 5.417431846666903e-06,
      "loss": 0.4657,
      "step": 1275
    },
    {
      "epoch": 0.5282707387536112,
      "grad_norm": 2.32837647206492,
      "learning_rate": 5.381519657053072e-06,
      "loss": 0.4542,
      "step": 1280
    },
    {
      "epoch": 0.5303342963268676,
      "grad_norm": 2.5139826780698753,
      "learning_rate": 5.3455876593854435e-06,
      "loss": 0.4665,
      "step": 1285
    },
    {
      "epoch": 0.5323978539001238,
      "grad_norm": 2.4870649815664496,
      "learning_rate": 5.309637719211256e-06,
      "loss": 0.4527,
      "step": 1290
    },
    {
      "epoch": 0.5344614114733801,
      "grad_norm": 2.509973264998628,
      "learning_rate": 5.273671703009301e-06,
      "loss": 0.4507,
      "step": 1295
    },
    {
      "epoch": 0.5365249690466364,
      "grad_norm": 2.690719386362163,
      "learning_rate": 5.2376914780930225e-06,
      "loss": 0.4576,
      "step": 1300
    },
    {
      "epoch": 0.5385885266198926,
      "grad_norm": 4.083713443737158,
      "learning_rate": 5.201698912513561e-06,
      "loss": 0.4412,
      "step": 1305
    },
    {
      "epoch": 0.540652084193149,
      "grad_norm": 2.3850290751433016,
      "learning_rate": 5.165695874962769e-06,
      "loss": 0.4633,
      "step": 1310
    },
    {
      "epoch": 0.5427156417664053,
      "grad_norm": 2.383468999919073,
      "learning_rate": 5.129684234676195e-06,
      "loss": 0.4482,
      "step": 1315
    },
    {
      "epoch": 0.5447791993396616,
      "grad_norm": 2.3602215294945568,
      "learning_rate": 5.0936658613360304e-06,
      "loss": 0.4683,
      "step": 1320
    },
    {
      "epoch": 0.5468427569129178,
      "grad_norm": 2.3769603536661035,
      "learning_rate": 5.057642624974036e-06,
      "loss": 0.4587,
      "step": 1325
    },
    {
      "epoch": 0.5489063144861742,
      "grad_norm": 2.5837187422329837,
      "learning_rate": 5.021616395874459e-06,
      "loss": 0.439,
      "step": 1330
    },
    {
      "epoch": 0.5509698720594305,
      "grad_norm": 2.587958298155525,
      "learning_rate": 4.9855890444769226e-06,
      "loss": 0.4547,
      "step": 1335
    },
    {
      "epoch": 0.5530334296326868,
      "grad_norm": 2.4489991448969795,
      "learning_rate": 4.94956244127932e-06,
      "loss": 0.4435,
      "step": 1340
    },
    {
      "epoch": 0.555096987205943,
      "grad_norm": 2.70305559316515,
      "learning_rate": 4.9135384567407e-06,
      "loss": 0.4635,
      "step": 1345
    },
    {
      "epoch": 0.5571605447791993,
      "grad_norm": 2.552605702749918,
      "learning_rate": 4.877518961184149e-06,
      "loss": 0.4395,
      "step": 1350
    },
    {
      "epoch": 0.5592241023524557,
      "grad_norm": 2.4491029644956717,
      "learning_rate": 4.841505824699697e-06,
      "loss": 0.4446,
      "step": 1355
    },
    {
      "epoch": 0.561287659925712,
      "grad_norm": 2.2903925047271003,
      "learning_rate": 4.805500917047217e-06,
      "loss": 0.4398,
      "step": 1360
    },
    {
      "epoch": 0.5633512174989682,
      "grad_norm": 2.379688312715005,
      "learning_rate": 4.769506107559344e-06,
      "loss": 0.4515,
      "step": 1365
    },
    {
      "epoch": 0.5654147750722245,
      "grad_norm": 2.881348434158553,
      "learning_rate": 4.7335232650444376e-06,
      "loss": 0.4609,
      "step": 1370
    },
    {
      "epoch": 0.5674783326454808,
      "grad_norm": 2.450954066121652,
      "learning_rate": 4.697554257689541e-06,
      "loss": 0.4556,
      "step": 1375
    },
    {
      "epoch": 0.5695418902187371,
      "grad_norm": 2.418284415029761,
      "learning_rate": 4.661600952963391e-06,
      "loss": 0.4452,
      "step": 1380
    },
    {
      "epoch": 0.5716054477919934,
      "grad_norm": 2.3480748218612457,
      "learning_rate": 4.6256652175194675e-06,
      "loss": 0.461,
      "step": 1385
    },
    {
      "epoch": 0.5736690053652497,
      "grad_norm": 2.472904201761537,
      "learning_rate": 4.589748917099067e-06,
      "loss": 0.4492,
      "step": 1390
    },
    {
      "epoch": 0.5757325629385059,
      "grad_norm": 2.468262838594691,
      "learning_rate": 4.553853916434448e-06,
      "loss": 0.4403,
      "step": 1395
    },
    {
      "epoch": 0.5777961205117623,
      "grad_norm": 2.361041942268244,
      "learning_rate": 4.517982079152009e-06,
      "loss": 0.4377,
      "step": 1400
    },
    {
      "epoch": 0.5798596780850186,
      "grad_norm": 2.5500737849526423,
      "learning_rate": 4.482135267675529e-06,
      "loss": 0.45,
      "step": 1405
    },
    {
      "epoch": 0.5819232356582749,
      "grad_norm": 2.453235877539221,
      "learning_rate": 4.446315343129484e-06,
      "loss": 0.4328,
      "step": 1410
    },
    {
      "epoch": 0.5839867932315311,
      "grad_norm": 2.3281079794605355,
      "learning_rate": 4.410524165242407e-06,
      "loss": 0.4374,
      "step": 1415
    },
    {
      "epoch": 0.5860503508047874,
      "grad_norm": 2.551505972004267,
      "learning_rate": 4.3747635922503364e-06,
      "loss": 0.4432,
      "step": 1420
    },
    {
      "epoch": 0.5881139083780438,
      "grad_norm": 2.5826844758469276,
      "learning_rate": 4.3390354808003464e-06,
      "loss": 0.4402,
      "step": 1425
    },
    {
      "epoch": 0.5901774659513,
      "grad_norm": 2.3612299259996936,
      "learning_rate": 4.30334168585414e-06,
      "loss": 0.4388,
      "step": 1430
    },
    {
      "epoch": 0.5922410235245563,
      "grad_norm": 2.4199132831049246,
      "learning_rate": 4.26768406059175e-06,
      "loss": 0.4296,
      "step": 1435
    },
    {
      "epoch": 0.5943045810978126,
      "grad_norm": 2.3293439174487274,
      "learning_rate": 4.2320644563153244e-06,
      "loss": 0.4391,
      "step": 1440
    },
    {
      "epoch": 0.596368138671069,
      "grad_norm": 2.5508830684770043,
      "learning_rate": 4.196484722352999e-06,
      "loss": 0.4341,
      "step": 1445
    },
    {
      "epoch": 0.5984316962443252,
      "grad_norm": 2.459534568261172,
      "learning_rate": 4.160946705962895e-06,
      "loss": 0.4298,
      "step": 1450
    },
    {
      "epoch": 0.6004952538175815,
      "grad_norm": 2.4607489784078806,
      "learning_rate": 4.12545225223721e-06,
      "loss": 0.4238,
      "step": 1455
    },
    {
      "epoch": 0.6025588113908378,
      "grad_norm": 2.5772914418218336,
      "learning_rate": 4.090003204006406e-06,
      "loss": 0.4165,
      "step": 1460
    },
    {
      "epoch": 0.604622368964094,
      "grad_norm": 2.273824183445231,
      "learning_rate": 4.054601401743561e-06,
      "loss": 0.4366,
      "step": 1465
    },
    {
      "epoch": 0.6066859265373504,
      "grad_norm": 2.5922865720792183,
      "learning_rate": 4.019248683468781e-06,
      "loss": 0.4372,
      "step": 1470
    },
    {
      "epoch": 0.6087494841106067,
      "grad_norm": 2.520992944241331,
      "learning_rate": 3.983946884653804e-06,
      "loss": 0.4418,
      "step": 1475
    },
    {
      "epoch": 0.610813041683863,
      "grad_norm": 2.4527707525439415,
      "learning_rate": 3.948697838126681e-06,
      "loss": 0.4186,
      "step": 1480
    },
    {
      "epoch": 0.6128765992571192,
      "grad_norm": 2.3893935033867546,
      "learning_rate": 3.9135033739766275e-06,
      "loss": 0.4332,
      "step": 1485
    },
    {
      "epoch": 0.6149401568303756,
      "grad_norm": 2.303193161104287,
      "learning_rate": 3.8783653194590055e-06,
      "loss": 0.4308,
      "step": 1490
    },
    {
      "epoch": 0.6170037144036319,
      "grad_norm": 2.568914194114212,
      "learning_rate": 3.843285498900457e-06,
      "loss": 0.4238,
      "step": 1495
    },
    {
      "epoch": 0.6190672719768882,
      "grad_norm": 2.3874626733288093,
      "learning_rate": 3.808265733604181e-06,
      "loss": 0.4211,
      "step": 1500
    },
    {
      "epoch": 0.6211308295501444,
      "grad_norm": 2.397949299188709,
      "learning_rate": 3.7733078417553835e-06,
      "loss": 0.4198,
      "step": 1505
    },
    {
      "epoch": 0.6231943871234007,
      "grad_norm": 2.4304137659983223,
      "learning_rate": 3.7384136383268673e-06,
      "loss": 0.4364,
      "step": 1510
    },
    {
      "epoch": 0.6252579446966571,
      "grad_norm": 2.4027742214175727,
      "learning_rate": 3.70358493498481e-06,
      "loss": 0.431,
      "step": 1515
    },
    {
      "epoch": 0.6273215022699133,
      "grad_norm": 2.6288095509256255,
      "learning_rate": 3.668823539994706e-06,
      "loss": 0.4252,
      "step": 1520
    },
    {
      "epoch": 0.6293850598431696,
      "grad_norm": 2.386671047852444,
      "learning_rate": 3.634131258127467e-06,
      "loss": 0.428,
      "step": 1525
    },
    {
      "epoch": 0.6314486174164259,
      "grad_norm": 2.3713874902931615,
      "learning_rate": 3.5995098905657457e-06,
      "loss": 0.4286,
      "step": 1530
    },
    {
      "epoch": 0.6335121749896823,
      "grad_norm": 2.300141894247011,
      "learning_rate": 3.564961234810399e-06,
      "loss": 0.4162,
      "step": 1535
    },
    {
      "epoch": 0.6355757325629385,
      "grad_norm": 2.4635471100860933,
      "learning_rate": 3.530487084587173e-06,
      "loss": 0.4102,
      "step": 1540
    },
    {
      "epoch": 0.6376392901361948,
      "grad_norm": 2.374004467594742,
      "learning_rate": 3.4960892297535765e-06,
      "loss": 0.4144,
      "step": 1545
    },
    {
      "epoch": 0.6397028477094511,
      "grad_norm": 2.3078125248362324,
      "learning_rate": 3.4617694562059465e-06,
      "loss": 0.4076,
      "step": 1550
    },
    {
      "epoch": 0.6417664052827073,
      "grad_norm": 2.513382809869699,
      "learning_rate": 3.427529545786736e-06,
      "loss": 0.4103,
      "step": 1555
    },
    {
      "epoch": 0.6438299628559637,
      "grad_norm": 2.421843916725022,
      "learning_rate": 3.393371276191997e-06,
      "loss": 0.4151,
      "step": 1560
    },
    {
      "epoch": 0.64589352042922,
      "grad_norm": 2.469381907143848,
      "learning_rate": 3.3592964208790808e-06,
      "loss": 0.4164,
      "step": 1565
    },
    {
      "epoch": 0.6479570780024763,
      "grad_norm": 2.531906105074784,
      "learning_rate": 3.3253067489745716e-06,
      "loss": 0.4071,
      "step": 1570
    },
    {
      "epoch": 0.6500206355757325,
      "grad_norm": 2.3846085481545747,
      "learning_rate": 3.291404025182432e-06,
      "loss": 0.4075,
      "step": 1575
    },
    {
      "epoch": 0.6520841931489889,
      "grad_norm": 2.433814917027714,
      "learning_rate": 3.2575900096923753e-06,
      "loss": 0.4087,
      "step": 1580
    },
    {
      "epoch": 0.6541477507222452,
      "grad_norm": 2.4424178299468506,
      "learning_rate": 3.2238664580884877e-06,
      "loss": 0.4043,
      "step": 1585
    },
    {
      "epoch": 0.6562113082955014,
      "grad_norm": 2.3585518358327833,
      "learning_rate": 3.1902351212580734e-06,
      "loss": 0.4116,
      "step": 1590
    },
    {
      "epoch": 0.6582748658687577,
      "grad_norm": 2.4272402253853174,
      "learning_rate": 3.1566977453007564e-06,
      "loss": 0.4045,
      "step": 1595
    },
    {
      "epoch": 0.660338423442014,
      "grad_norm": 2.2735139196354184,
      "learning_rate": 3.1232560714378214e-06,
      "loss": 0.3998,
      "step": 1600
    },
    {
      "epoch": 0.6624019810152704,
      "grad_norm": 2.496149939265061,
      "learning_rate": 3.0899118359218104e-06,
      "loss": 0.3986,
      "step": 1605
    },
    {
      "epoch": 0.6644655385885266,
      "grad_norm": 2.577482137677985,
      "learning_rate": 3.056666769946382e-06,
      "loss": 0.3922,
      "step": 1610
    },
    {
      "epoch": 0.6665290961617829,
      "grad_norm": 2.432120191923341,
      "learning_rate": 3.0235225995564323e-06,
      "loss": 0.3959,
      "step": 1615
    },
    {
      "epoch": 0.6685926537350392,
      "grad_norm": 2.5315892714201507,
      "learning_rate": 2.9904810455584664e-06,
      "loss": 0.396,
      "step": 1620
    },
    {
      "epoch": 0.6706562113082954,
      "grad_norm": 2.4126722223283643,
      "learning_rate": 2.9575438234312807e-06,
      "loss": 0.4034,
      "step": 1625
    },
    {
      "epoch": 0.6727197688815518,
      "grad_norm": 2.3922710783717682,
      "learning_rate": 2.9247126432368685e-06,
      "loss": 0.3969,
      "step": 1630
    },
    {
      "epoch": 0.6747833264548081,
      "grad_norm": 2.542933247382595,
      "learning_rate": 2.8919892095316616e-06,
      "loss": 0.3914,
      "step": 1635
    },
    {
      "epoch": 0.6768468840280644,
      "grad_norm": 2.503873056742514,
      "learning_rate": 2.859375221278014e-06,
      "loss": 0.4053,
      "step": 1640
    },
    {
      "epoch": 0.6789104416013206,
      "grad_norm": 2.4704924711278275,
      "learning_rate": 2.826872371756e-06,
      "loss": 0.4171,
      "step": 1645
    },
    {
      "epoch": 0.680973999174577,
      "grad_norm": 2.2126828966128347,
      "learning_rate": 2.7944823484755053e-06,
      "loss": 0.3979,
      "step": 1650
    },
    {
      "epoch": 0.6830375567478333,
      "grad_norm": 2.452244039159894,
      "learning_rate": 2.762206833088608e-06,
      "loss": 0.3992,
      "step": 1655
    },
    {
      "epoch": 0.6851011143210896,
      "grad_norm": 2.4737732563891197,
      "learning_rate": 2.7300475013022666e-06,
      "loss": 0.4013,
      "step": 1660
    },
    {
      "epoch": 0.6871646718943458,
      "grad_norm": 2.32695849246749,
      "learning_rate": 2.6980060227913273e-06,
      "loss": 0.3984,
      "step": 1665
    },
    {
      "epoch": 0.6892282294676021,
      "grad_norm": 2.5306901436703497,
      "learning_rate": 2.666084061111828e-06,
      "loss": 0.3942,
      "step": 1670
    },
    {
      "epoch": 0.6912917870408585,
      "grad_norm": 2.40430761373431,
      "learning_rate": 2.6342832736146403e-06,
      "loss": 0.3943,
      "step": 1675
    },
    {
      "epoch": 0.6933553446141147,
      "grad_norm": 2.3692497196583955,
      "learning_rate": 2.6026053113594056e-06,
      "loss": 0.383,
      "step": 1680
    },
    {
      "epoch": 0.695418902187371,
      "grad_norm": 2.45758426227379,
      "learning_rate": 2.571051819028826e-06,
      "loss": 0.4056,
      "step": 1685
    },
    {
      "epoch": 0.6974824597606273,
      "grad_norm": 2.490132984626703,
      "learning_rate": 2.539624434843272e-06,
      "loss": 0.408,
      "step": 1690
    },
    {
      "epoch": 0.6995460173338837,
      "grad_norm": 2.316910192052694,
      "learning_rate": 2.508324790475731e-06,
      "loss": 0.3908,
      "step": 1695
    },
    {
      "epoch": 0.7016095749071399,
      "grad_norm": 2.5678586000455215,
      "learning_rate": 2.4771545109670746e-06,
      "loss": 0.3886,
      "step": 1700
    },
    {
      "epoch": 0.7036731324803962,
      "grad_norm": 2.333143658853672,
      "learning_rate": 2.446115214641719e-06,
      "loss": 0.3825,
      "step": 1705
    },
    {
      "epoch": 0.7057366900536525,
      "grad_norm": 2.421284956930726,
      "learning_rate": 2.415208513023574e-06,
      "loss": 0.378,
      "step": 1710
    },
    {
      "epoch": 0.7078002476269087,
      "grad_norm": 2.348370007461818,
      "learning_rate": 2.3844360107523973e-06,
      "loss": 0.3745,
      "step": 1715
    },
    {
      "epoch": 0.7098638052001651,
      "grad_norm": 2.2561855639412047,
      "learning_rate": 2.353799305500467e-06,
      "loss": 0.3768,
      "step": 1720
    },
    {
      "epoch": 0.7119273627734214,
      "grad_norm": 2.4144762778450364,
      "learning_rate": 2.3232999878896407e-06,
      "loss": 0.3941,
      "step": 1725
    },
    {
      "epoch": 0.7139909203466777,
      "grad_norm": 2.510836282225756,
      "learning_rate": 2.2929396414087737e-06,
      "loss": 0.4012,
      "step": 1730
    },
    {
      "epoch": 0.7160544779199339,
      "grad_norm": 2.4017944766607906,
      "learning_rate": 2.2627198423314988e-06,
      "loss": 0.3955,
      "step": 1735
    },
    {
      "epoch": 0.7181180354931903,
      "grad_norm": 2.2838737017346356,
      "learning_rate": 2.2326421596343923e-06,
      "loss": 0.3833,
      "step": 1740
    },
    {
      "epoch": 0.7201815930664466,
      "grad_norm": 2.590581157502952,
      "learning_rate": 2.2027081549155188e-06,
      "loss": 0.3912,
      "step": 1745
    },
    {
      "epoch": 0.7222451506397028,
      "grad_norm": 2.3748009695139776,
      "learning_rate": 2.172919382313347e-06,
      "loss": 0.3808,
      "step": 1750
    },
    {
      "epoch": 0.7243087082129591,
      "grad_norm": 2.4065528450546445,
      "learning_rate": 2.1432773884260627e-06,
      "loss": 0.3705,
      "step": 1755
    },
    {
      "epoch": 0.7263722657862154,
      "grad_norm": 2.4049387116398564,
      "learning_rate": 2.1137837122312787e-06,
      "loss": 0.3765,
      "step": 1760
    },
    {
      "epoch": 0.7284358233594718,
      "grad_norm": 2.5287323783127174,
      "learning_rate": 2.0844398850061186e-06,
      "loss": 0.3886,
      "step": 1765
    },
    {
      "epoch": 0.730499380932728,
      "grad_norm": 2.467020632677239,
      "learning_rate": 2.0552474302477327e-06,
      "loss": 0.3875,
      "step": 1770
    },
    {
      "epoch": 0.7325629385059843,
      "grad_norm": 2.313861208086701,
      "learning_rate": 2.0262078635941818e-06,
      "loss": 0.3889,
      "step": 1775
    },
    {
      "epoch": 0.7346264960792406,
      "grad_norm": 2.3695175969299296,
      "learning_rate": 1.9973226927457567e-06,
      "loss": 0.3763,
      "step": 1780
    },
    {
      "epoch": 0.736690053652497,
      "grad_norm": 2.4901046009056884,
      "learning_rate": 1.968593417386705e-06,
      "loss": 0.3846,
      "step": 1785
    },
    {
      "epoch": 0.7387536112257532,
      "grad_norm": 2.3483856649423585,
      "learning_rate": 1.940021529107356e-06,
      "loss": 0.3856,
      "step": 1790
    },
    {
      "epoch": 0.7408171687990095,
      "grad_norm": 2.3336460330815205,
      "learning_rate": 1.911608511326688e-06,
      "loss": 0.3796,
      "step": 1795
    },
    {
      "epoch": 0.7428807263722658,
      "grad_norm": 2.3516145493656344,
      "learning_rate": 1.8833558392153118e-06,
      "loss": 0.3857,
      "step": 1800
    },
    {
      "epoch": 0.744944283945522,
      "grad_norm": 2.3634530645626164,
      "learning_rate": 1.8552649796188765e-06,
      "loss": 0.3891,
      "step": 1805
    },
    {
      "epoch": 0.7470078415187784,
      "grad_norm": 2.249211977645891,
      "learning_rate": 1.8273373909819132e-06,
      "loss": 0.3702,
      "step": 1810
    },
    {
      "epoch": 0.7490713990920347,
      "grad_norm": 2.582157058457191,
      "learning_rate": 1.7995745232721207e-06,
      "loss": 0.3849,
      "step": 1815
    },
    {
      "epoch": 0.751134956665291,
      "grad_norm": 2.3911264718615444,
      "learning_rate": 1.771977817905075e-06,
      "loss": 0.3662,
      "step": 1820
    },
    {
      "epoch": 0.7531985142385472,
      "grad_norm": 2.279432612650543,
      "learning_rate": 1.744548707669404e-06,
      "loss": 0.3697,
      "step": 1825
    },
    {
      "epoch": 0.7552620718118036,
      "grad_norm": 2.433098384561743,
      "learning_rate": 1.717288616652386e-06,
      "loss": 0.3714,
      "step": 1830
    },
    {
      "epoch": 0.7573256293850599,
      "grad_norm": 2.3966273952677764,
      "learning_rate": 1.6901989601660224e-06,
      "loss": 0.3701,
      "step": 1835
    },
    {
      "epoch": 0.7593891869583161,
      "grad_norm": 2.3793685219467697,
      "learning_rate": 1.6632811446735553e-06,
      "loss": 0.3749,
      "step": 1840
    },
    {
      "epoch": 0.7614527445315724,
      "grad_norm": 2.4235058401467167,
      "learning_rate": 1.6365365677164413e-06,
      "loss": 0.3798,
      "step": 1845
    },
    {
      "epoch": 0.7635163021048287,
      "grad_norm": 2.425747353853211,
      "learning_rate": 1.6099666178417927e-06,
      "loss": 0.3754,
      "step": 1850
    },
    {
      "epoch": 0.7655798596780851,
      "grad_norm": 2.378675323045764,
      "learning_rate": 1.5835726745302953e-06,
      "loss": 0.3792,
      "step": 1855
    },
    {
      "epoch": 0.7676434172513413,
      "grad_norm": 2.372493229885198,
      "learning_rate": 1.5573561081245724e-06,
      "loss": 0.3835,
      "step": 1860
    },
    {
      "epoch": 0.7697069748245976,
      "grad_norm": 2.5336093680740026,
      "learning_rate": 1.531318279758055e-06,
      "loss": 0.3619,
      "step": 1865
    },
    {
      "epoch": 0.7717705323978539,
      "grad_norm": 2.4693265387678442,
      "learning_rate": 1.5054605412842915e-06,
      "loss": 0.3736,
      "step": 1870
    },
    {
      "epoch": 0.7738340899711101,
      "grad_norm": 2.303876812670922,
      "learning_rate": 1.479784235206786e-06,
      "loss": 0.3583,
      "step": 1875
    },
    {
      "epoch": 0.7758976475443665,
      "grad_norm": 2.590842168629549,
      "learning_rate": 1.4542906946092817e-06,
      "loss": 0.3736,
      "step": 1880
    },
    {
      "epoch": 0.7779612051176228,
      "grad_norm": 2.500911682432463,
      "learning_rate": 1.428981243086549e-06,
      "loss": 0.367,
      "step": 1885
    },
    {
      "epoch": 0.7800247626908791,
      "grad_norm": 2.436965476930112,
      "learning_rate": 1.4038571946756724e-06,
      "loss": 0.3742,
      "step": 1890
    },
    {
      "epoch": 0.7820883202641353,
      "grad_norm": 2.5258796423012333,
      "learning_rate": 1.3789198537878202e-06,
      "loss": 0.3642,
      "step": 1895
    },
    {
      "epoch": 0.7841518778373917,
      "grad_norm": 2.369376183964329,
      "learning_rate": 1.3541705151405304e-06,
      "loss": 0.3746,
      "step": 1900
    },
    {
      "epoch": 0.786215435410648,
      "grad_norm": 2.3567197797039015,
      "learning_rate": 1.3296104636904806e-06,
      "loss": 0.3658,
      "step": 1905
    },
    {
      "epoch": 0.7882789929839042,
      "grad_norm": 2.4167682681302645,
      "learning_rate": 1.3052409745667783e-06,
      "loss": 0.3704,
      "step": 1910
    },
    {
      "epoch": 0.7903425505571605,
      "grad_norm": 2.3250328086103234,
      "learning_rate": 1.281063313004761e-06,
      "loss": 0.3677,
      "step": 1915
    },
    {
      "epoch": 0.7924061081304168,
      "grad_norm": 2.4369843316317,
      "learning_rate": 1.2570787342803064e-06,
      "loss": 0.3709,
      "step": 1920
    },
    {
      "epoch": 0.7944696657036732,
      "grad_norm": 2.5193296767489537,
      "learning_rate": 1.2332884836446473e-06,
      "loss": 0.3719,
      "step": 1925
    },
    {
      "epoch": 0.7965332232769294,
      "grad_norm": 2.5002330071663055,
      "learning_rate": 1.2096937962597399e-06,
      "loss": 0.3677,
      "step": 1930
    },
    {
      "epoch": 0.7985967808501857,
      "grad_norm": 2.601424944526519,
      "learning_rate": 1.1862958971341199e-06,
      "loss": 0.3686,
      "step": 1935
    },
    {
      "epoch": 0.800660338423442,
      "grad_norm": 2.3874506853275363,
      "learning_rate": 1.16309600105931e-06,
      "loss": 0.3619,
      "step": 1940
    },
    {
      "epoch": 0.8027238959966984,
      "grad_norm": 2.360432193713698,
      "learning_rate": 1.1400953125467452e-06,
      "loss": 0.3598,
      "step": 1945
    },
    {
      "epoch": 0.8047874535699546,
      "grad_norm": 2.5865175614762053,
      "learning_rate": 1.117295025765235e-06,
      "loss": 0.3678,
      "step": 1950
    },
    {
      "epoch": 0.8068510111432109,
      "grad_norm": 2.386447339664475,
      "learning_rate": 1.09469632447897e-06,
      "loss": 0.3527,
      "step": 1955
    },
    {
      "epoch": 0.8089145687164672,
      "grad_norm": 2.2243473128718607,
      "learning_rate": 1.0723003819860555e-06,
      "loss": 0.3499,
      "step": 1960
    },
    {
      "epoch": 0.8109781262897234,
      "grad_norm": 2.4765043887476,
      "learning_rate": 1.0501083610575958e-06,
      "loss": 0.358,
      "step": 1965
    },
    {
      "epoch": 0.8130416838629798,
      "grad_norm": 2.4123151679318093,
      "learning_rate": 1.0281214138773315e-06,
      "loss": 0.3411,
      "step": 1970
    },
    {
      "epoch": 0.8151052414362361,
      "grad_norm": 2.4030618342504892,
      "learning_rate": 1.0063406819818106e-06,
      "loss": 0.3492,
      "step": 1975
    },
    {
      "epoch": 0.8171687990094924,
      "grad_norm": 2.3438148800354988,
      "learning_rate": 9.847672962011233e-07,
      "loss": 0.3554,
      "step": 1980
    },
    {
      "epoch": 0.8192323565827486,
      "grad_norm": 2.328120737157089,
      "learning_rate": 9.634023766001981e-07,
      "loss": 0.3562,
      "step": 1985
    },
    {
      "epoch": 0.821295914156005,
      "grad_norm": 2.5107579831834816,
      "learning_rate": 9.422470324206385e-07,
      "loss": 0.3575,
      "step": 1990
    },
    {
      "epoch": 0.8233594717292613,
      "grad_norm": 2.455991687646852,
      "learning_rate": 9.213023620231404e-07,
      "loss": 0.3508,
      "step": 1995
    },
    {
      "epoch": 0.8254230293025175,
      "grad_norm": 2.3501569587377693,
      "learning_rate": 9.005694528304621e-07,
      "loss": 0.3529,
      "step": 2000
    },
    {
      "epoch": 0.8274865868757738,
      "grad_norm": 2.3892952569776527,
      "learning_rate": 8.800493812709677e-07,
      "loss": 0.3605,
      "step": 2005
    },
    {
      "epoch": 0.8295501444490301,
      "grad_norm": 2.387038888457062,
      "learning_rate": 8.597432127227434e-07,
      "loss": 0.3593,
      "step": 2010
    },
    {
      "epoch": 0.8316137020222865,
      "grad_norm": 2.3731590066808526,
      "learning_rate": 8.396520014582798e-07,
      "loss": 0.3487,
      "step": 2015
    },
    {
      "epoch": 0.8336772595955427,
      "grad_norm": 2.576723904790806,
      "learning_rate": 8.197767905897341e-07,
      "loss": 0.3541,
      "step": 2020
    },
    {
      "epoch": 0.835740817168799,
      "grad_norm": 2.2751284706261306,
      "learning_rate": 8.001186120147825e-07,
      "loss": 0.3463,
      "step": 2025
    },
    {
      "epoch": 0.8378043747420553,
      "grad_norm": 2.9241025316172804,
      "learning_rate": 7.806784863630324e-07,
      "loss": 0.3504,
      "step": 2030
    },
    {
      "epoch": 0.8398679323153116,
      "grad_norm": 2.4559935395470034,
      "learning_rate": 7.614574229430432e-07,
      "loss": 0.3538,
      "step": 2035
    },
    {
      "epoch": 0.8419314898885679,
      "grad_norm": 2.3573162952481037,
      "learning_rate": 7.424564196899175e-07,
      "loss": 0.3472,
      "step": 2040
    },
    {
      "epoch": 0.8439950474618242,
      "grad_norm": 2.3152148511829287,
      "learning_rate": 7.236764631134907e-07,
      "loss": 0.3446,
      "step": 2045
    },
    {
      "epoch": 0.8460586050350805,
      "grad_norm": 2.328674279050691,
      "learning_rate": 7.051185282471173e-07,
      "loss": 0.3494,
      "step": 2050
    },
    {
      "epoch": 0.8481221626083367,
      "grad_norm": 2.3835140443527982,
      "learning_rate": 6.867835785970417e-07,
      "loss": 0.3443,
      "step": 2055
    },
    {
      "epoch": 0.8501857201815931,
      "grad_norm": 2.366551971108674,
      "learning_rate": 6.68672566092376e-07,
      "loss": 0.3434,
      "step": 2060
    },
    {
      "epoch": 0.8522492777548494,
      "grad_norm": 2.4362523327596257,
      "learning_rate": 6.507864310356826e-07,
      "loss": 0.3361,
      "step": 2065
    },
    {
      "epoch": 0.8543128353281056,
      "grad_norm": 2.341036835199248,
      "learning_rate": 6.331261020541462e-07,
      "loss": 0.3658,
      "step": 2070
    },
    {
      "epoch": 0.8563763929013619,
      "grad_norm": 2.5075276094389753,
      "learning_rate": 6.156924960513638e-07,
      "loss": 0.3598,
      "step": 2075
    },
    {
      "epoch": 0.8584399504746183,
      "grad_norm": 2.304765407228246,
      "learning_rate": 5.984865181597454e-07,
      "loss": 0.3528,
      "step": 2080
    },
    {
      "epoch": 0.8605035080478746,
      "grad_norm": 2.334045444451915,
      "learning_rate": 5.815090616935115e-07,
      "loss": 0.3506,
      "step": 2085
    },
    {
      "epoch": 0.8625670656211308,
      "grad_norm": 2.434017607980088,
      "learning_rate": 5.647610081023214e-07,
      "loss": 0.3524,
      "step": 2090
    },
    {
      "epoch": 0.8646306231943871,
      "grad_norm": 2.400891346368746,
      "learning_rate": 5.482432269255011e-07,
      "loss": 0.3431,
      "step": 2095
    },
    {
      "epoch": 0.8666941807676434,
      "grad_norm": 2.4286183102500027,
      "learning_rate": 5.31956575746907e-07,
      "loss": 0.3464,
      "step": 2100
    },
    {
      "epoch": 0.8687577383408998,
      "grad_norm": 2.380179963058147,
      "learning_rate": 5.159019001503962e-07,
      "loss": 0.3499,
      "step": 2105
    },
    {
      "epoch": 0.870821295914156,
      "grad_norm": 2.306679882811888,
      "learning_rate": 5.000800336759232e-07,
      "loss": 0.3437,
      "step": 2110
    },
    {
      "epoch": 0.8728848534874123,
      "grad_norm": 2.289306218382388,
      "learning_rate": 4.844917977762653e-07,
      "loss": 0.3542,
      "step": 2115
    },
    {
      "epoch": 0.8749484110606686,
      "grad_norm": 2.401006920203725,
      "learning_rate": 4.691380017743769e-07,
      "loss": 0.3508,
      "step": 2120
    },
    {
      "epoch": 0.8770119686339248,
      "grad_norm": 2.294060885815769,
      "learning_rate": 4.5401944282136454e-07,
      "loss": 0.3592,
      "step": 2125
    },
    {
      "epoch": 0.8790755262071812,
      "grad_norm": 2.5092731081270654,
      "learning_rate": 4.3913690585510427e-07,
      "loss": 0.3343,
      "step": 2130
    },
    {
      "epoch": 0.8811390837804375,
      "grad_norm": 2.4672551742312523,
      "learning_rate": 4.244911635594856e-07,
      "loss": 0.3421,
      "step": 2135
    },
    {
      "epoch": 0.8832026413536938,
      "grad_norm": 2.4349291953653234,
      "learning_rate": 4.100829763242975e-07,
      "loss": 0.3482,
      "step": 2140
    },
    {
      "epoch": 0.88526619892695,
      "grad_norm": 2.532171457134297,
      "learning_rate": 3.959130922057497e-07,
      "loss": 0.3365,
      "step": 2145
    },
    {
      "epoch": 0.8873297565002064,
      "grad_norm": 2.4156983042588647,
      "learning_rate": 3.819822468876283e-07,
      "loss": 0.3424,
      "step": 2150
    },
    {
      "epoch": 0.8893933140734627,
      "grad_norm": 2.3670827733250346,
      "learning_rate": 3.6829116364310914e-07,
      "loss": 0.3489,
      "step": 2155
    },
    {
      "epoch": 0.8914568716467189,
      "grad_norm": 2.471779388721616,
      "learning_rate": 3.548405532972016e-07,
      "loss": 0.35,
      "step": 2160
    },
    {
      "epoch": 0.8935204292199752,
      "grad_norm": 2.5818181780490184,
      "learning_rate": 3.416311141898421e-07,
      "loss": 0.3468,
      "step": 2165
    },
    {
      "epoch": 0.8955839867932315,
      "grad_norm": 2.561889277902207,
      "learning_rate": 3.286635321396392e-07,
      "loss": 0.3415,
      "step": 2170
    },
    {
      "epoch": 0.8976475443664879,
      "grad_norm": 2.3816913973065654,
      "learning_rate": 3.159384804082666e-07,
      "loss": 0.335,
      "step": 2175
    },
    {
      "epoch": 0.8997111019397441,
      "grad_norm": 2.3401913676557116,
      "learning_rate": 3.034566196655092e-07,
      "loss": 0.3438,
      "step": 2180
    },
    {
      "epoch": 0.9017746595130004,
      "grad_norm": 2.3557441169663598,
      "learning_rate": 2.912185979549587e-07,
      "loss": 0.3415,
      "step": 2185
    },
    {
      "epoch": 0.9038382170862567,
      "grad_norm": 2.5287391128485464,
      "learning_rate": 2.792250506603694e-07,
      "loss": 0.3433,
      "step": 2190
    },
    {
      "epoch": 0.905901774659513,
      "grad_norm": 2.454585929498319,
      "learning_rate": 2.67476600472672e-07,
      "loss": 0.3451,
      "step": 2195
    },
    {
      "epoch": 0.9079653322327693,
      "grad_norm": 2.596601343506189,
      "learning_rate": 2.5597385735764256e-07,
      "loss": 0.3583,
      "step": 2200
    },
    {
      "epoch": 0.9100288898060256,
      "grad_norm": 2.19805805166686,
      "learning_rate": 2.447174185242324e-07,
      "loss": 0.3461,
      "step": 2205
    },
    {
      "epoch": 0.9120924473792819,
      "grad_norm": 2.4174079650578624,
      "learning_rate": 2.3370786839356286e-07,
      "loss": 0.3396,
      "step": 2210
    },
    {
      "epoch": 0.9141560049525381,
      "grad_norm": 2.36584151995226,
      "learning_rate": 2.2294577856858236e-07,
      "loss": 0.3401,
      "step": 2215
    },
    {
      "epoch": 0.9162195625257945,
      "grad_norm": 2.434108457884742,
      "learning_rate": 2.124317078043936e-07,
      "loss": 0.3464,
      "step": 2220
    },
    {
      "epoch": 0.9182831200990508,
      "grad_norm": 2.2550641701066523,
      "learning_rate": 2.0216620197923642e-07,
      "loss": 0.3403,
      "step": 2225
    },
    {
      "epoch": 0.920346677672307,
      "grad_norm": 2.479423722525491,
      "learning_rate": 1.9214979406615097e-07,
      "loss": 0.34,
      "step": 2230
    },
    {
      "epoch": 0.9224102352455633,
      "grad_norm": 2.3709597200527965,
      "learning_rate": 1.823830041053065e-07,
      "loss": 0.3386,
      "step": 2235
    },
    {
      "epoch": 0.9244737928188197,
      "grad_norm": 2.4793227986038193,
      "learning_rate": 1.7286633917700134e-07,
      "loss": 0.3538,
      "step": 2240
    },
    {
      "epoch": 0.926537350392076,
      "grad_norm": 2.4372071315808896,
      "learning_rate": 1.6360029337533056e-07,
      "loss": 0.3399,
      "step": 2245
    },
    {
      "epoch": 0.9286009079653322,
      "grad_norm": 2.350437257669817,
      "learning_rate": 1.5458534778254319e-07,
      "loss": 0.3456,
      "step": 2250
    },
    {
      "epoch": 0.9306644655385885,
      "grad_norm": 2.3873927009091025,
      "learning_rate": 1.4582197044405556e-07,
      "loss": 0.3476,
      "step": 2255
    },
    {
      "epoch": 0.9327280231118448,
      "grad_norm": 2.3893579636725715,
      "learning_rate": 1.3731061634415787e-07,
      "loss": 0.3397,
      "step": 2260
    },
    {
      "epoch": 0.9347915806851012,
      "grad_norm": 2.384752010782048,
      "learning_rate": 1.2905172738238603e-07,
      "loss": 0.3487,
      "step": 2265
    },
    {
      "epoch": 0.9368551382583574,
      "grad_norm": 2.3497503513681552,
      "learning_rate": 1.2104573235058314e-07,
      "loss": 0.3383,
      "step": 2270
    },
    {
      "epoch": 0.9389186958316137,
      "grad_norm": 2.6245838898899687,
      "learning_rate": 1.1329304691063692e-07,
      "loss": 0.3414,
      "step": 2275
    },
    {
      "epoch": 0.94098225340487,
      "grad_norm": 2.3201420801643358,
      "learning_rate": 1.0579407357289461e-07,
      "loss": 0.3357,
      "step": 2280
    },
    {
      "epoch": 0.9430458109781263,
      "grad_norm": 2.390636102614023,
      "learning_rate": 9.854920167526983e-08,
      "loss": 0.3358,
      "step": 2285
    },
    {
      "epoch": 0.9451093685513826,
      "grad_norm": 2.6935442949745774,
      "learning_rate": 9.155880736302857e-08,
      "loss": 0.3516,
      "step": 2290
    },
    {
      "epoch": 0.9471729261246389,
      "grad_norm": 2.33174770965001,
      "learning_rate": 8.482325356925614e-08,
      "loss": 0.3383,
      "step": 2295
    },
    {
      "epoch": 0.9492364836978952,
      "grad_norm": 2.382213172213323,
      "learning_rate": 7.834288999601703e-08,
      "loss": 0.362,
      "step": 2300
    },
    {
      "epoch": 0.9513000412711514,
      "grad_norm": 2.3191616911646724,
      "learning_rate": 7.211805309620013e-08,
      "loss": 0.328,
      "step": 2305
    },
    {
      "epoch": 0.9533635988444078,
      "grad_norm": 2.469419059741572,
      "learning_rate": 6.614906605604598e-08,
      "loss": 0.3331,
      "step": 2310
    },
    {
      "epoch": 0.9554271564176641,
      "grad_norm": 2.4225879136955997,
      "learning_rate": 6.043623877837301e-08,
      "loss": 0.347,
      "step": 2315
    },
    {
      "epoch": 0.9574907139909203,
      "grad_norm": 2.3375518129030786,
      "learning_rate": 5.4979867866482015e-08,
      "loss": 0.339,
      "step": 2320
    },
    {
      "epoch": 0.9595542715641766,
      "grad_norm": 2.460240591927137,
      "learning_rate": 4.978023660876075e-08,
      "loss": 0.3409,
      "step": 2325
    },
    {
      "epoch": 0.961617829137433,
      "grad_norm": 2.5514455417806388,
      "learning_rate": 4.4837614963976205e-08,
      "loss": 0.3549,
      "step": 2330
    },
    {
      "epoch": 0.9636813867106893,
      "grad_norm": 2.4229916264085594,
      "learning_rate": 4.015225954725421e-08,
      "loss": 0.3462,
      "step": 2335
    },
    {
      "epoch": 0.9657449442839455,
      "grad_norm": 2.392552783694151,
      "learning_rate": 3.572441361676171e-08,
      "loss": 0.3423,
      "step": 2340
    },
    {
      "epoch": 0.9678085018572018,
      "grad_norm": 2.414641292682689,
      "learning_rate": 3.155430706107465e-08,
      "loss": 0.3416,
      "step": 2345
    },
    {
      "epoch": 0.9698720594304581,
      "grad_norm": 2.5378189638300164,
      "learning_rate": 2.7642156387240882e-08,
      "loss": 0.3419,
      "step": 2350
    },
    {
      "epoch": 0.9719356170037144,
      "grad_norm": 2.4175815882083502,
      "learning_rate": 2.3988164709542462e-08,
      "loss": 0.3321,
      "step": 2355
    },
    {
      "epoch": 0.9739991745769707,
      "grad_norm": 2.2974926251848227,
      "learning_rate": 2.0592521738947434e-08,
      "loss": 0.3383,
      "step": 2360
    },
    {
      "epoch": 0.976062732150227,
      "grad_norm": 2.3929562307787338,
      "learning_rate": 1.7455403773262713e-08,
      "loss": 0.3402,
      "step": 2365
    },
    {
      "epoch": 0.9781262897234833,
      "grad_norm": 2.4697653358488103,
      "learning_rate": 1.4576973687979168e-08,
      "loss": 0.3245,
      "step": 2370
    },
    {
      "epoch": 0.9801898472967395,
      "grad_norm": 2.4583519170902677,
      "learning_rate": 1.1957380927816176e-08,
      "loss": 0.3449,
      "step": 2375
    },
    {
      "epoch": 0.9822534048699959,
      "grad_norm": 2.53558630491575,
      "learning_rate": 9.596761498961715e-09,
      "loss": 0.3378,
      "step": 2380
    },
    {
      "epoch": 0.9843169624432522,
      "grad_norm": 2.4388182873283983,
      "learning_rate": 7.495237962011902e-09,
      "loss": 0.3376,
      "step": 2385
    },
    {
      "epoch": 0.9863805200165084,
      "grad_norm": 2.5307806748315764,
      "learning_rate": 5.65291942560775e-09,
      "loss": 0.3437,
      "step": 2390
    },
    {
      "epoch": 0.9884440775897647,
      "grad_norm": 2.494932764858625,
      "learning_rate": 4.0699015407702495e-09,
      "loss": 0.3466,
      "step": 2395
    },
    {
      "epoch": 0.9905076351630211,
      "grad_norm": 2.386648335012219,
      "learning_rate": 2.7462664959343465e-09,
      "loss": 0.3384,
      "step": 2400
    },
    {
      "epoch": 0.9925711927362774,
      "grad_norm": 2.7213817604169748,
      "learning_rate": 1.682083012681246e-09,
      "loss": 0.3332,
      "step": 2405
    },
    {
      "epoch": 0.9946347503095336,
      "grad_norm": 2.559632513812426,
      "learning_rate": 8.774063421712608e-10,
      "loss": 0.3494,
      "step": 2410
    },
    {
      "epoch": 0.9966983078827899,
      "grad_norm": 2.307214644489273,
      "learning_rate": 3.322782622738885e-10,
      "loss": 0.3417,
      "step": 2415
    },
    {
      "epoch": 0.9987618654560462,
      "grad_norm": 2.3274612533228907,
      "learning_rate": 4.672707540231969e-11,
      "loss": 0.3382,
      "step": 2420
    },
    {
      "epoch": 1.0,
      "eval_runtime": 3.3225,
      "eval_samples_per_second": 3.01,
      "eval_steps_per_second": 0.903,
      "step": 2423
    },
    {
      "epoch": 1.0,
      "step": 2423,
      "total_flos": 253663452856320.0,
      "train_loss": 0.5332340838443727,
      "train_runtime": 21510.4539,
      "train_samples_per_second": 1.802,
      "train_steps_per_second": 0.113
    }
  ],
  "logging_steps": 5,
  "max_steps": 2423,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 253663452856320.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}