{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9998474446987032,
  "eval_steps": 500,
  "global_step": 3277,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0030511060259344014,
      "grad_norm": 4.5989179611206055,
      "learning_rate": 9e-06,
      "loss": 1.1046,
      "step": 10
    },
    {
      "epoch": 0.006102212051868803,
      "grad_norm": 4.813294887542725,
      "learning_rate": 1.9e-05,
      "loss": 0.8003,
      "step": 20
    },
    {
      "epoch": 0.009153318077803204,
      "grad_norm": 4.563686847686768,
      "learning_rate": 2.9e-05,
      "loss": 0.7886,
      "step": 30
    },
    {
      "epoch": 0.012204424103737605,
      "grad_norm": 5.083153247833252,
      "learning_rate": 3.9000000000000006e-05,
      "loss": 0.7638,
      "step": 40
    },
    {
      "epoch": 0.015255530129672006,
      "grad_norm": 4.403526306152344,
      "learning_rate": 4.9e-05,
      "loss": 0.7635,
      "step": 50
    },
    {
      "epoch": 0.018306636155606407,
      "grad_norm": 3.6779983043670654,
      "learning_rate": 4.986055159590952e-05,
      "loss": 0.841,
      "step": 60
    },
    {
      "epoch": 0.02135774218154081,
      "grad_norm": 3.7986652851104736,
      "learning_rate": 4.970560892469787e-05,
      "loss": 0.9439,
      "step": 70
    },
    {
      "epoch": 0.02440884820747521,
      "grad_norm": 3.810528516769409,
      "learning_rate": 4.955066625348621e-05,
      "loss": 0.843,
      "step": 80
    },
    {
      "epoch": 0.02745995423340961,
      "grad_norm": 3.7793450355529785,
      "learning_rate": 4.939572358227456e-05,
      "loss": 0.9459,
      "step": 90
    },
    {
      "epoch": 0.03051106025934401,
      "grad_norm": 4.002682685852051,
      "learning_rate": 4.924078091106291e-05,
      "loss": 0.9242,
      "step": 100
    },
    {
      "epoch": 0.033562166285278416,
      "grad_norm": 3.5819356441497803,
      "learning_rate": 4.908583823985126e-05,
      "loss": 0.8683,
      "step": 110
    },
    {
      "epoch": 0.036613272311212815,
      "grad_norm": 3.728360891342163,
      "learning_rate": 4.893089556863961e-05,
      "loss": 0.9133,
      "step": 120
    },
    {
      "epoch": 0.03966437833714721,
      "grad_norm": 4.5038981437683105,
      "learning_rate": 4.8775952897427956e-05,
      "loss": 0.8466,
      "step": 130
    },
    {
      "epoch": 0.04271548436308162,
      "grad_norm": 3.0992960929870605,
      "learning_rate": 4.8621010226216305e-05,
      "loss": 0.8424,
      "step": 140
    },
    {
      "epoch": 0.04576659038901602,
      "grad_norm": 3.688044786453247,
      "learning_rate": 4.8466067555004654e-05,
      "loss": 0.8283,
      "step": 150
    },
    {
      "epoch": 0.04881769641495042,
      "grad_norm": 4.070577621459961,
      "learning_rate": 4.8311124883792996e-05,
      "loss": 0.9207,
      "step": 160
    },
    {
      "epoch": 0.05186880244088482,
      "grad_norm": 3.9466311931610107,
      "learning_rate": 4.8156182212581345e-05,
      "loss": 0.8095,
      "step": 170
    },
    {
      "epoch": 0.05491990846681922,
      "grad_norm": 3.0422556400299072,
      "learning_rate": 4.8001239541369694e-05,
      "loss": 0.8649,
      "step": 180
    },
    {
      "epoch": 0.057971014492753624,
      "grad_norm": 3.414745330810547,
      "learning_rate": 4.784629687015804e-05,
      "loss": 0.8963,
      "step": 190
    },
    {
      "epoch": 0.06102212051868802,
      "grad_norm": 3.7459030151367188,
      "learning_rate": 4.769135419894639e-05,
      "loss": 0.8352,
      "step": 200
    },
    {
      "epoch": 0.06407322654462243,
      "grad_norm": 2.6838042736053467,
      "learning_rate": 4.753641152773474e-05,
      "loss": 0.8089,
      "step": 210
    },
    {
      "epoch": 0.06712433257055683,
      "grad_norm": 5.595691680908203,
      "learning_rate": 4.738146885652309e-05,
      "loss": 0.8701,
      "step": 220
    },
    {
      "epoch": 0.07017543859649122,
      "grad_norm": 3.1771981716156006,
      "learning_rate": 4.722652618531144e-05,
      "loss": 0.8714,
      "step": 230
    },
    {
      "epoch": 0.07322654462242563,
      "grad_norm": 2.728522539138794,
      "learning_rate": 4.707158351409979e-05,
      "loss": 0.8655,
      "step": 240
    },
    {
      "epoch": 0.07627765064836003,
      "grad_norm": 3.781033515930176,
      "learning_rate": 4.691664084288813e-05,
      "loss": 0.7953,
      "step": 250
    },
    {
      "epoch": 0.07932875667429443,
      "grad_norm": 2.9846785068511963,
      "learning_rate": 4.676169817167648e-05,
      "loss": 0.8079,
      "step": 260
    },
    {
      "epoch": 0.08237986270022883,
      "grad_norm": 3.5596115589141846,
      "learning_rate": 4.6606755500464835e-05,
      "loss": 0.8413,
      "step": 270
    },
    {
      "epoch": 0.08543096872616324,
      "grad_norm": 3.188471555709839,
      "learning_rate": 4.645181282925318e-05,
      "loss": 0.8411,
      "step": 280
    },
    {
      "epoch": 0.08848207475209764,
      "grad_norm": 4.387840270996094,
      "learning_rate": 4.629687015804153e-05,
      "loss": 0.9073,
      "step": 290
    },
    {
      "epoch": 0.09153318077803203,
      "grad_norm": 3.842625379562378,
      "learning_rate": 4.6141927486829876e-05,
      "loss": 0.8269,
      "step": 300
    },
    {
      "epoch": 0.09458428680396644,
      "grad_norm": 4.832952499389648,
      "learning_rate": 4.5986984815618225e-05,
      "loss": 0.9064,
      "step": 310
    },
    {
      "epoch": 0.09763539282990084,
      "grad_norm": 2.8753535747528076,
      "learning_rate": 4.5832042144406574e-05,
      "loss": 0.8873,
      "step": 320
    },
    {
      "epoch": 0.10068649885583524,
      "grad_norm": 2.5598573684692383,
      "learning_rate": 4.5677099473194916e-05,
      "loss": 0.8913,
      "step": 330
    },
    {
      "epoch": 0.10373760488176964,
      "grad_norm": 2.9012229442596436,
      "learning_rate": 4.5522156801983265e-05,
      "loss": 0.7988,
      "step": 340
    },
    {
      "epoch": 0.10678871090770405,
      "grad_norm": 2.888648748397827,
      "learning_rate": 4.536721413077162e-05,
      "loss": 0.8404,
      "step": 350
    },
    {
      "epoch": 0.10983981693363844,
      "grad_norm": 3.794020891189575,
      "learning_rate": 4.521227145955996e-05,
      "loss": 0.8429,
      "step": 360
    },
    {
      "epoch": 0.11289092295957284,
      "grad_norm": 3.925490617752075,
      "learning_rate": 4.505732878834831e-05,
      "loss": 0.8768,
      "step": 370
    },
    {
      "epoch": 0.11594202898550725,
      "grad_norm": 3.4535255432128906,
      "learning_rate": 4.490238611713666e-05,
      "loss": 0.7951,
      "step": 380
    },
    {
      "epoch": 0.11899313501144165,
      "grad_norm": 3.9904849529266357,
      "learning_rate": 4.474744344592501e-05,
      "loss": 0.8524,
      "step": 390
    },
    {
      "epoch": 0.12204424103737604,
      "grad_norm": 2.6464405059814453,
      "learning_rate": 4.459250077471336e-05,
      "loss": 0.8194,
      "step": 400
    },
    {
      "epoch": 0.12509534706331046,
      "grad_norm": 3.6185855865478516,
      "learning_rate": 4.443755810350171e-05,
      "loss": 0.7706,
      "step": 410
    },
    {
      "epoch": 0.12814645308924486,
      "grad_norm": 4.222907543182373,
      "learning_rate": 4.428261543229005e-05,
      "loss": 0.7858,
      "step": 420
    },
    {
      "epoch": 0.13119755911517925,
      "grad_norm": 2.8912553787231445,
      "learning_rate": 4.4127672761078406e-05,
      "loss": 0.7658,
      "step": 430
    },
    {
      "epoch": 0.13424866514111367,
      "grad_norm": 2.9529941082000732,
      "learning_rate": 4.3972730089866755e-05,
      "loss": 0.8415,
      "step": 440
    },
    {
      "epoch": 0.13729977116704806,
      "grad_norm": 3.296875238418579,
      "learning_rate": 4.38177874186551e-05,
      "loss": 0.818,
      "step": 450
    },
    {
      "epoch": 0.14035087719298245,
      "grad_norm": 2.736250400543213,
      "learning_rate": 4.366284474744345e-05,
      "loss": 0.819,
      "step": 460
    },
    {
      "epoch": 0.14340198321891687,
      "grad_norm": 3.2874772548675537,
      "learning_rate": 4.3507902076231796e-05,
      "loss": 0.8381,
      "step": 470
    },
    {
      "epoch": 0.14645308924485126,
      "grad_norm": 2.9855446815490723,
      "learning_rate": 4.3352959405020145e-05,
      "loss": 0.7591,
      "step": 480
    },
    {
      "epoch": 0.14950419527078565,
      "grad_norm": 3.37127947807312,
      "learning_rate": 4.3198016733808494e-05,
      "loss": 0.7815,
      "step": 490
    },
    {
      "epoch": 0.15255530129672007,
      "grad_norm": 3.162848472595215,
      "learning_rate": 4.3043074062596836e-05,
      "loss": 0.8115,
      "step": 500
    },
    {
      "epoch": 0.15560640732265446,
      "grad_norm": 3.0527734756469727,
      "learning_rate": 4.288813139138519e-05,
      "loss": 0.8871,
      "step": 510
    },
    {
      "epoch": 0.15865751334858885,
      "grad_norm": 3.1467580795288086,
      "learning_rate": 4.273318872017354e-05,
      "loss": 0.9112,
      "step": 520
    },
    {
      "epoch": 0.16170861937452327,
      "grad_norm": 3.414907455444336,
      "learning_rate": 4.257824604896188e-05,
      "loss": 0.8089,
      "step": 530
    },
    {
      "epoch": 0.16475972540045766,
      "grad_norm": 3.9576640129089355,
      "learning_rate": 4.242330337775023e-05,
      "loss": 0.7847,
      "step": 540
    },
    {
      "epoch": 0.16781083142639205,
      "grad_norm": 3.8652496337890625,
      "learning_rate": 4.226836070653858e-05,
      "loss": 0.801,
      "step": 550
    },
    {
      "epoch": 0.17086193745232647,
      "grad_norm": 3.028764247894287,
      "learning_rate": 4.211341803532693e-05,
      "loss": 0.8323,
      "step": 560
    },
    {
      "epoch": 0.17391304347826086,
      "grad_norm": 3.1638927459716797,
      "learning_rate": 4.195847536411528e-05,
      "loss": 0.8575,
      "step": 570
    },
    {
      "epoch": 0.17696414950419528,
      "grad_norm": 3.3378586769104004,
      "learning_rate": 4.180353269290363e-05,
      "loss": 0.8048,
      "step": 580
    },
    {
      "epoch": 0.18001525553012968,
      "grad_norm": 2.5703723430633545,
      "learning_rate": 4.164859002169198e-05,
      "loss": 0.8557,
      "step": 590
    },
    {
      "epoch": 0.18306636155606407,
      "grad_norm": 4.101804733276367,
      "learning_rate": 4.1493647350480326e-05,
      "loss": 0.8358,
      "step": 600
    },
    {
      "epoch": 0.18611746758199849,
      "grad_norm": 3.845407009124756,
      "learning_rate": 4.1338704679268675e-05,
      "loss": 0.8777,
      "step": 610
    },
    {
      "epoch": 0.18916857360793288,
      "grad_norm": 3.370732545852661,
      "learning_rate": 4.118376200805702e-05,
      "loss": 0.8485,
      "step": 620
    },
    {
      "epoch": 0.19221967963386727,
      "grad_norm": 2.168769121170044,
      "learning_rate": 4.102881933684537e-05,
      "loss": 0.7416,
      "step": 630
    },
    {
      "epoch": 0.1952707856598017,
      "grad_norm": 3.819798707962036,
      "learning_rate": 4.087387666563372e-05,
      "loss": 0.9199,
      "step": 640
    },
    {
      "epoch": 0.19832189168573608,
      "grad_norm": 3.19429612159729,
      "learning_rate": 4.0718933994422065e-05,
      "loss": 0.7747,
      "step": 650
    },
    {
      "epoch": 0.20137299771167047,
      "grad_norm": 3.17229962348938,
      "learning_rate": 4.0563991323210414e-05,
      "loss": 0.7394,
      "step": 660
    },
    {
      "epoch": 0.2044241037376049,
      "grad_norm": 2.316675901412964,
      "learning_rate": 4.040904865199876e-05,
      "loss": 0.7368,
      "step": 670
    },
    {
      "epoch": 0.20747520976353928,
      "grad_norm": 3.350698471069336,
      "learning_rate": 4.025410598078711e-05,
      "loss": 0.8839,
      "step": 680
    },
    {
      "epoch": 0.21052631578947367,
      "grad_norm": 2.5842721462249756,
      "learning_rate": 4.009916330957546e-05,
      "loss": 0.7665,
      "step": 690
    },
    {
      "epoch": 0.2135774218154081,
      "grad_norm": 2.9108049869537354,
      "learning_rate": 3.99442206383638e-05,
      "loss": 0.8145,
      "step": 700
    },
    {
      "epoch": 0.21662852784134248,
      "grad_norm": 2.9678006172180176,
      "learning_rate": 3.978927796715215e-05,
      "loss": 0.7719,
      "step": 710
    },
    {
      "epoch": 0.21967963386727687,
      "grad_norm": 2.911322832107544,
      "learning_rate": 3.963433529594051e-05,
      "loss": 0.7354,
      "step": 720
    },
    {
      "epoch": 0.2227307398932113,
      "grad_norm": 3.1599249839782715,
      "learning_rate": 3.947939262472885e-05,
      "loss": 0.7873,
      "step": 730
    },
    {
      "epoch": 0.22578184591914569,
      "grad_norm": 2.6644296646118164,
      "learning_rate": 3.93244499535172e-05,
      "loss": 0.8311,
      "step": 740
    },
    {
      "epoch": 0.2288329519450801,
      "grad_norm": 3.263995885848999,
      "learning_rate": 3.916950728230555e-05,
      "loss": 0.8185,
      "step": 750
    },
    {
      "epoch": 0.2318840579710145,
      "grad_norm": 2.542343854904175,
      "learning_rate": 3.90145646110939e-05,
      "loss": 0.7769,
      "step": 760
    },
    {
      "epoch": 0.2349351639969489,
      "grad_norm": 2.4859440326690674,
      "learning_rate": 3.8859621939882246e-05,
      "loss": 0.8328,
      "step": 770
    },
    {
      "epoch": 0.2379862700228833,
      "grad_norm": 2.632722854614258,
      "learning_rate": 3.8704679268670596e-05,
      "loss": 0.7859,
      "step": 780
    },
    {
      "epoch": 0.2410373760488177,
      "grad_norm": 3.209907054901123,
      "learning_rate": 3.854973659745894e-05,
      "loss": 0.8394,
      "step": 790
    },
    {
      "epoch": 0.2440884820747521,
      "grad_norm": 2.4981114864349365,
      "learning_rate": 3.8394793926247294e-05,
      "loss": 0.8241,
      "step": 800
    },
    {
      "epoch": 0.2471395881006865,
      "grad_norm": 2.5217268466949463,
      "learning_rate": 3.823985125503564e-05,
      "loss": 0.7501,
      "step": 810
    },
    {
      "epoch": 0.2501906941266209,
      "grad_norm": 3.676504611968994,
      "learning_rate": 3.8084908583823985e-05,
      "loss": 0.8782,
      "step": 820
    },
    {
      "epoch": 0.2532418001525553,
      "grad_norm": 2.651146173477173,
      "learning_rate": 3.7929965912612334e-05,
      "loss": 0.848,
      "step": 830
    },
    {
      "epoch": 0.2562929061784897,
      "grad_norm": 3.1401736736297607,
      "learning_rate": 3.777502324140068e-05,
      "loss": 0.7991,
      "step": 840
    },
    {
      "epoch": 0.2593440122044241,
      "grad_norm": 3.0251362323760986,
      "learning_rate": 3.762008057018903e-05,
      "loss": 0.8723,
      "step": 850
    },
    {
      "epoch": 0.2623951182303585,
      "grad_norm": 2.1540679931640625,
      "learning_rate": 3.746513789897738e-05,
      "loss": 0.7478,
      "step": 860
    },
    {
      "epoch": 0.2654462242562929,
      "grad_norm": 2.6065683364868164,
      "learning_rate": 3.731019522776572e-05,
      "loss": 0.7488,
      "step": 870
    },
    {
      "epoch": 0.26849733028222733,
      "grad_norm": 2.697063684463501,
      "learning_rate": 3.715525255655408e-05,
      "loss": 0.7778,
      "step": 880
    },
    {
      "epoch": 0.2715484363081617,
      "grad_norm": 3.4640157222747803,
      "learning_rate": 3.700030988534243e-05,
      "loss": 1.0205,
      "step": 890
    },
    {
      "epoch": 0.2745995423340961,
      "grad_norm": 3.5426061153411865,
      "learning_rate": 3.684536721413077e-05,
      "loss": 0.762,
      "step": 900
    },
    {
      "epoch": 0.2776506483600305,
      "grad_norm": 2.219358444213867,
      "learning_rate": 3.669042454291912e-05,
      "loss": 0.7996,
      "step": 910
    },
    {
      "epoch": 0.2807017543859649,
      "grad_norm": 2.7404627799987793,
      "learning_rate": 3.6535481871707475e-05,
      "loss": 0.7611,
      "step": 920
    },
    {
      "epoch": 0.2837528604118993,
      "grad_norm": 2.950205087661743,
      "learning_rate": 3.638053920049582e-05,
      "loss": 0.7685,
      "step": 930
    },
    {
      "epoch": 0.28680396643783374,
      "grad_norm": 2.676258087158203,
      "learning_rate": 3.6225596529284167e-05,
      "loss": 0.7864,
      "step": 940
    },
    {
      "epoch": 0.2898550724637681,
      "grad_norm": 2.5284500122070312,
      "learning_rate": 3.6070653858072516e-05,
      "loss": 0.7856,
      "step": 950
    },
    {
      "epoch": 0.2929061784897025,
      "grad_norm": 2.2837953567504883,
      "learning_rate": 3.5915711186860865e-05,
      "loss": 0.7965,
      "step": 960
    },
    {
      "epoch": 0.2959572845156369,
      "grad_norm": 3.223839044570923,
      "learning_rate": 3.5760768515649214e-05,
      "loss": 0.8123,
      "step": 970
    },
    {
      "epoch": 0.2990083905415713,
      "grad_norm": 2.8053982257843018,
      "learning_rate": 3.560582584443756e-05,
      "loss": 0.7813,
      "step": 980
    },
    {
      "epoch": 0.30205949656750575,
      "grad_norm": 2.799560785293579,
      "learning_rate": 3.5450883173225905e-05,
      "loss": 0.8418,
      "step": 990
    },
    {
      "epoch": 0.30511060259344014,
      "grad_norm": 3.838677406311035,
      "learning_rate": 3.529594050201426e-05,
      "loss": 0.7948,
      "step": 1000
    },
    {
      "epoch": 0.30816170861937453,
      "grad_norm": 2.587244749069214,
      "learning_rate": 3.51409978308026e-05,
      "loss": 0.784,
      "step": 1010
    },
    {
      "epoch": 0.3112128146453089,
      "grad_norm": 3.555494785308838,
      "learning_rate": 3.498605515959095e-05,
      "loss": 0.7854,
      "step": 1020
    },
    {
      "epoch": 0.3142639206712433,
      "grad_norm": 2.900275468826294,
      "learning_rate": 3.48311124883793e-05,
      "loss": 0.8536,
      "step": 1030
    },
    {
      "epoch": 0.3173150266971777,
      "grad_norm": 3.058786153793335,
      "learning_rate": 3.467616981716765e-05,
      "loss": 0.8163,
      "step": 1040
    },
    {
      "epoch": 0.32036613272311215,
      "grad_norm": 3.0138072967529297,
      "learning_rate": 3.4521227145956e-05,
      "loss": 0.8534,
      "step": 1050
    },
    {
      "epoch": 0.32341723874904654,
      "grad_norm": 2.71114182472229,
      "learning_rate": 3.436628447474435e-05,
      "loss": 0.8401,
      "step": 1060
    },
    {
      "epoch": 0.32646834477498093,
      "grad_norm": 2.574453592300415,
      "learning_rate": 3.421134180353269e-05,
      "loss": 0.7596,
      "step": 1070
    },
    {
      "epoch": 0.3295194508009153,
      "grad_norm": 2.7125132083892822,
      "learning_rate": 3.4056399132321046e-05,
      "loss": 0.8042,
      "step": 1080
    },
    {
      "epoch": 0.3325705568268497,
      "grad_norm": 3.0520405769348145,
      "learning_rate": 3.3901456461109395e-05,
      "loss": 0.7713,
      "step": 1090
    },
    {
      "epoch": 0.3356216628527841,
      "grad_norm": 2.5527734756469727,
      "learning_rate": 3.374651378989774e-05,
      "loss": 0.7929,
      "step": 1100
    },
    {
      "epoch": 0.33867276887871856,
      "grad_norm": 2.1792235374450684,
      "learning_rate": 3.3591571118686087e-05,
      "loss": 0.7102,
      "step": 1110
    },
    {
      "epoch": 0.34172387490465295,
      "grad_norm": 3.192811965942383,
      "learning_rate": 3.3436628447474436e-05,
      "loss": 0.7956,
      "step": 1120
    },
    {
      "epoch": 0.34477498093058734,
      "grad_norm": 3.0279996395111084,
      "learning_rate": 3.3281685776262785e-05,
      "loss": 0.7883,
      "step": 1130
    },
    {
      "epoch": 0.34782608695652173,
      "grad_norm": 3.0758237838745117,
      "learning_rate": 3.3126743105051134e-05,
      "loss": 0.7457,
      "step": 1140
    },
    {
      "epoch": 0.3508771929824561,
      "grad_norm": 2.2491872310638428,
      "learning_rate": 3.297180043383948e-05,
      "loss": 0.7712,
      "step": 1150
    },
    {
      "epoch": 0.35392829900839057,
      "grad_norm": 2.543558120727539,
      "learning_rate": 3.281685776262783e-05,
      "loss": 0.8013,
      "step": 1160
    },
    {
      "epoch": 0.35697940503432496,
      "grad_norm": 2.996823310852051,
      "learning_rate": 3.266191509141618e-05,
      "loss": 0.7769,
      "step": 1170
    },
    {
      "epoch": 0.36003051106025935,
      "grad_norm": 2.1455206871032715,
      "learning_rate": 3.250697242020452e-05,
      "loss": 0.7778,
      "step": 1180
    },
    {
      "epoch": 0.36308161708619374,
      "grad_norm": 3.1174466609954834,
      "learning_rate": 3.235202974899287e-05,
      "loss": 0.7519,
      "step": 1190
    },
    {
      "epoch": 0.36613272311212813,
      "grad_norm": 2.6810905933380127,
      "learning_rate": 3.219708707778122e-05,
      "loss": 0.8112,
      "step": 1200
    },
    {
      "epoch": 0.3691838291380625,
      "grad_norm": 2.6105597019195557,
      "learning_rate": 3.204214440656957e-05,
      "loss": 0.7344,
      "step": 1210
    },
    {
      "epoch": 0.37223493516399697,
      "grad_norm": 3.3781626224517822,
      "learning_rate": 3.188720173535792e-05,
      "loss": 0.8401,
      "step": 1220
    },
    {
      "epoch": 0.37528604118993136,
      "grad_norm": 3.3649299144744873,
      "learning_rate": 3.173225906414627e-05,
      "loss": 0.7849,
      "step": 1230
    },
    {
      "epoch": 0.37833714721586575,
      "grad_norm": 2.6910207271575928,
      "learning_rate": 3.157731639293462e-05,
      "loss": 0.8165,
      "step": 1240
    },
    {
      "epoch": 0.38138825324180015,
      "grad_norm": 2.5760819911956787,
      "learning_rate": 3.1422373721722966e-05,
      "loss": 0.8043,
      "step": 1250
    },
    {
      "epoch": 0.38443935926773454,
      "grad_norm": 2.5498452186584473,
      "learning_rate": 3.1267431050511315e-05,
      "loss": 0.7566,
      "step": 1260
    },
    {
      "epoch": 0.38749046529366893,
      "grad_norm": 1.8285728693008423,
      "learning_rate": 3.111248837929966e-05,
      "loss": 0.7152,
      "step": 1270
    },
    {
      "epoch": 0.3905415713196034,
      "grad_norm": 2.509260654449463,
      "learning_rate": 3.0957545708088007e-05,
      "loss": 0.8489,
      "step": 1280
    },
    {
      "epoch": 0.39359267734553777,
      "grad_norm": 2.905968427658081,
      "learning_rate": 3.080260303687636e-05,
      "loss": 0.7369,
      "step": 1290
    },
    {
      "epoch": 0.39664378337147216,
      "grad_norm": 2.5083484649658203,
      "learning_rate": 3.0647660365664705e-05,
      "loss": 0.7331,
      "step": 1300
    },
    {
      "epoch": 0.39969488939740655,
      "grad_norm": 2.2442944049835205,
      "learning_rate": 3.0492717694453054e-05,
      "loss": 0.7282,
      "step": 1310
    },
    {
      "epoch": 0.40274599542334094,
      "grad_norm": 2.856735944747925,
      "learning_rate": 3.0337775023241406e-05,
      "loss": 0.7009,
      "step": 1320
    },
    {
      "epoch": 0.4057971014492754,
      "grad_norm": 2.9385952949523926,
      "learning_rate": 3.018283235202975e-05,
      "loss": 0.7225,
      "step": 1330
    },
    {
      "epoch": 0.4088482074752098,
      "grad_norm": 3.0979793071746826,
      "learning_rate": 3.00278896808181e-05,
      "loss": 0.764,
      "step": 1340
    },
    {
      "epoch": 0.41189931350114417,
      "grad_norm": 2.6863820552825928,
      "learning_rate": 2.9872947009606443e-05,
      "loss": 0.7475,
      "step": 1350
    },
    {
      "epoch": 0.41495041952707856,
      "grad_norm": 2.9140465259552,
      "learning_rate": 2.9718004338394795e-05,
      "loss": 0.7468,
      "step": 1360
    },
    {
      "epoch": 0.41800152555301295,
      "grad_norm": 2.702855348587036,
      "learning_rate": 2.9563061667183144e-05,
      "loss": 0.7444,
      "step": 1370
    },
    {
      "epoch": 0.42105263157894735,
      "grad_norm": 2.646045207977295,
      "learning_rate": 2.940811899597149e-05,
      "loss": 0.7911,
      "step": 1380
    },
    {
      "epoch": 0.4241037376048818,
      "grad_norm": 2.568760395050049,
      "learning_rate": 2.925317632475984e-05,
      "loss": 0.7815,
      "step": 1390
    },
    {
      "epoch": 0.4271548436308162,
      "grad_norm": 2.381138801574707,
      "learning_rate": 2.909823365354819e-05,
      "loss": 0.7411,
      "step": 1400
    },
    {
      "epoch": 0.4302059496567506,
      "grad_norm": 2.6686244010925293,
      "learning_rate": 2.8943290982336534e-05,
      "loss": 0.7593,
      "step": 1410
    },
    {
      "epoch": 0.43325705568268497,
      "grad_norm": 3.8091323375701904,
      "learning_rate": 2.8788348311124886e-05,
      "loss": 0.7731,
      "step": 1420
    },
    {
      "epoch": 0.43630816170861936,
      "grad_norm": 3.0860376358032227,
      "learning_rate": 2.8633405639913235e-05,
      "loss": 0.7617,
      "step": 1430
    },
    {
      "epoch": 0.43935926773455375,
      "grad_norm": 3.304323196411133,
      "learning_rate": 2.847846296870158e-05,
      "loss": 0.8343,
      "step": 1440
    },
    {
      "epoch": 0.4424103737604882,
      "grad_norm": 2.6195857524871826,
      "learning_rate": 2.832352029748993e-05,
      "loss": 0.688,
      "step": 1450
    },
    {
      "epoch": 0.4454614797864226,
      "grad_norm": 2.3896477222442627,
      "learning_rate": 2.8168577626278282e-05,
      "loss": 0.8032,
      "step": 1460
    },
    {
      "epoch": 0.448512585812357,
      "grad_norm": 2.9045519828796387,
      "learning_rate": 2.8013634955066625e-05,
      "loss": 0.8131,
      "step": 1470
    },
    {
      "epoch": 0.45156369183829137,
      "grad_norm": 2.5903823375701904,
      "learning_rate": 2.7858692283854977e-05,
      "loss": 0.7219,
      "step": 1480
    },
    {
      "epoch": 0.45461479786422576,
      "grad_norm": 2.8301548957824707,
      "learning_rate": 2.7703749612643326e-05,
      "loss": 0.7661,
      "step": 1490
    },
    {
      "epoch": 0.4576659038901602,
      "grad_norm": 2.5708298683166504,
      "learning_rate": 2.7548806941431672e-05,
      "loss": 0.7673,
      "step": 1500
    },
    {
      "epoch": 0.4607170099160946,
      "grad_norm": 2.4464125633239746,
      "learning_rate": 2.739386427022002e-05,
      "loss": 0.7328,
      "step": 1510
    },
    {
      "epoch": 0.463768115942029,
      "grad_norm": 3.007063627243042,
      "learning_rate": 2.7238921599008366e-05,
      "loss": 0.7449,
      "step": 1520
    },
    {
      "epoch": 0.4668192219679634,
      "grad_norm": 2.8109517097473145,
      "learning_rate": 2.7083978927796715e-05,
      "loss": 0.8403,
      "step": 1530
    },
    {
      "epoch": 0.4698703279938978,
      "grad_norm": 2.756274461746216,
      "learning_rate": 2.6929036256585068e-05,
      "loss": 0.786,
      "step": 1540
    },
    {
      "epoch": 0.47292143401983217,
      "grad_norm": 3.142169237136841,
      "learning_rate": 2.677409358537341e-05,
      "loss": 0.7738,
      "step": 1550
    },
    {
      "epoch": 0.4759725400457666,
      "grad_norm": 2.546050548553467,
      "learning_rate": 2.6619150914161763e-05,
      "loss": 0.7806,
      "step": 1560
    },
    {
      "epoch": 0.479023646071701,
      "grad_norm": 2.3914120197296143,
      "learning_rate": 2.646420824295011e-05,
      "loss": 0.7641,
      "step": 1570
    },
    {
      "epoch": 0.4820747520976354,
      "grad_norm": 2.156118869781494,
      "learning_rate": 2.6309265571738457e-05,
      "loss": 0.8034,
      "step": 1580
    },
    {
      "epoch": 0.4851258581235698,
      "grad_norm": 2.6585230827331543,
      "learning_rate": 2.6154322900526806e-05,
      "loss": 0.7609,
      "step": 1590
    },
    {
      "epoch": 0.4881769641495042,
      "grad_norm": 2.175424814224243,
      "learning_rate": 2.5999380229315155e-05,
      "loss": 0.6381,
      "step": 1600
    },
    {
      "epoch": 0.49122807017543857,
      "grad_norm": 3.3766794204711914,
      "learning_rate": 2.58444375581035e-05,
      "loss": 0.7187,
      "step": 1610
    },
    {
      "epoch": 0.494279176201373,
      "grad_norm": 2.665618419647217,
      "learning_rate": 2.568949488689185e-05,
      "loss": 0.7492,
      "step": 1620
    },
    {
      "epoch": 0.4973302822273074,
      "grad_norm": 2.5544931888580322,
      "learning_rate": 2.5534552215680202e-05,
      "loss": 0.7473,
      "step": 1630
    },
    {
      "epoch": 0.5003813882532419,
      "grad_norm": 2.091601848602295,
      "learning_rate": 2.5379609544468548e-05,
      "loss": 0.7364,
      "step": 1640
    },
    {
      "epoch": 0.5034324942791762,
      "grad_norm": 2.7752933502197266,
      "learning_rate": 2.5224666873256897e-05,
      "loss": 0.8124,
      "step": 1650
    },
    {
      "epoch": 0.5064836003051106,
      "grad_norm": 2.3963379859924316,
      "learning_rate": 2.5069724202045246e-05,
      "loss": 0.7213,
      "step": 1660
    },
    {
      "epoch": 0.509534706331045,
      "grad_norm": 2.0520660877227783,
      "learning_rate": 2.4914781530833592e-05,
      "loss": 0.7954,
      "step": 1670
    },
    {
      "epoch": 0.5125858123569794,
      "grad_norm": 2.9152419567108154,
      "learning_rate": 2.475983885962194e-05,
      "loss": 0.7114,
      "step": 1680
    },
    {
      "epoch": 0.5156369183829138,
      "grad_norm": 3.4632716178894043,
      "learning_rate": 2.460489618841029e-05,
      "loss": 0.7557,
      "step": 1690
    },
    {
      "epoch": 0.5186880244088482,
      "grad_norm": 3.0776329040527344,
      "learning_rate": 2.4449953517198636e-05,
      "loss": 0.7867,
      "step": 1700
    },
    {
      "epoch": 0.5217391304347826,
      "grad_norm": 2.640956163406372,
      "learning_rate": 2.4295010845986985e-05,
      "loss": 0.717,
      "step": 1710
    },
    {
      "epoch": 0.524790236460717,
      "grad_norm": 3.0681939125061035,
      "learning_rate": 2.4140068174775334e-05,
      "loss": 0.779,
      "step": 1720
    },
    {
      "epoch": 0.5278413424866514,
      "grad_norm": 3.3912854194641113,
      "learning_rate": 2.3985125503563683e-05,
      "loss": 0.7602,
      "step": 1730
    },
    {
      "epoch": 0.5308924485125858,
      "grad_norm": 3.4418396949768066,
      "learning_rate": 2.3830182832352028e-05,
      "loss": 0.7719,
      "step": 1740
    },
    {
      "epoch": 0.5339435545385202,
      "grad_norm": 2.5723013877868652,
      "learning_rate": 2.367524016114038e-05,
      "loss": 0.6693,
      "step": 1750
    },
    {
      "epoch": 0.5369946605644547,
      "grad_norm": 3.5987260341644287,
      "learning_rate": 2.3520297489928726e-05,
      "loss": 0.7202,
      "step": 1760
    },
    {
      "epoch": 0.540045766590389,
      "grad_norm": 3.0256638526916504,
      "learning_rate": 2.3365354818717075e-05,
      "loss": 0.769,
      "step": 1770
    },
    {
      "epoch": 0.5430968726163234,
      "grad_norm": 2.5137994289398193,
      "learning_rate": 2.3210412147505424e-05,
      "loss": 0.7133,
      "step": 1780
    },
    {
      "epoch": 0.5461479786422578,
      "grad_norm": 2.394585132598877,
      "learning_rate": 2.3055469476293773e-05,
      "loss": 0.7089,
      "step": 1790
    },
    {
      "epoch": 0.5491990846681922,
      "grad_norm": 2.6688249111175537,
      "learning_rate": 2.290052680508212e-05,
      "loss": 0.7819,
      "step": 1800
    },
    {
      "epoch": 0.5522501906941266,
      "grad_norm": 3.030775785446167,
      "learning_rate": 2.2745584133870468e-05,
      "loss": 0.7545,
      "step": 1810
    },
    {
      "epoch": 0.555301296720061,
      "grad_norm": 3.2252886295318604,
      "learning_rate": 2.2590641462658817e-05,
      "loss": 0.7854,
      "step": 1820
    },
    {
      "epoch": 0.5583524027459954,
      "grad_norm": 1.8850343227386475,
      "learning_rate": 2.2435698791447166e-05,
      "loss": 0.708,
      "step": 1830
    },
    {
      "epoch": 0.5614035087719298,
      "grad_norm": 3.511936902999878,
      "learning_rate": 2.2280756120235512e-05,
      "loss": 0.7384,
      "step": 1840
    },
    {
      "epoch": 0.5644546147978642,
      "grad_norm": 3.171138286590576,
      "learning_rate": 2.2125813449023864e-05,
      "loss": 0.7831,
      "step": 1850
    },
    {
      "epoch": 0.5675057208237986,
      "grad_norm": 2.5248336791992188,
      "learning_rate": 2.197087077781221e-05,
      "loss": 0.7011,
      "step": 1860
    },
    {
      "epoch": 0.5705568268497331,
      "grad_norm": 3.0410759449005127,
      "learning_rate": 2.181592810660056e-05,
      "loss": 0.7252,
      "step": 1870
    },
    {
      "epoch": 0.5736079328756675,
      "grad_norm": 3.691617012023926,
      "learning_rate": 2.1660985435388905e-05,
      "loss": 0.7585,
      "step": 1880
    },
    {
      "epoch": 0.5766590389016019,
      "grad_norm": 2.653752565383911,
      "learning_rate": 2.1506042764177257e-05,
      "loss": 0.7948,
      "step": 1890
    },
    {
      "epoch": 0.5797101449275363,
      "grad_norm": 2.0451462268829346,
      "learning_rate": 2.1351100092965603e-05,
      "loss": 0.6916,
      "step": 1900
    },
    {
      "epoch": 0.5827612509534706,
      "grad_norm": 2.8858132362365723,
      "learning_rate": 2.1196157421753952e-05,
      "loss": 0.682,
      "step": 1910
    },
    {
      "epoch": 0.585812356979405,
      "grad_norm": 2.553213596343994,
      "learning_rate": 2.10412147505423e-05,
      "loss": 0.7501,
      "step": 1920
    },
    {
      "epoch": 0.5888634630053394,
      "grad_norm": 2.1616005897521973,
      "learning_rate": 2.088627207933065e-05,
      "loss": 0.7536,
      "step": 1930
    },
    {
      "epoch": 0.5919145690312738,
      "grad_norm": 3.390302896499634,
      "learning_rate": 2.0731329408118995e-05,
      "loss": 0.787,
      "step": 1940
    },
    {
      "epoch": 0.5949656750572082,
      "grad_norm": 2.695875883102417,
      "learning_rate": 2.0576386736907348e-05,
      "loss": 0.7717,
      "step": 1950
    },
    {
      "epoch": 0.5980167810831426,
      "grad_norm": 2.517509698867798,
      "learning_rate": 2.0421444065695693e-05,
      "loss": 0.7527,
      "step": 1960
    },
    {
      "epoch": 0.601067887109077,
      "grad_norm": 2.498511552810669,
      "learning_rate": 2.0266501394484043e-05,
      "loss": 0.6925,
      "step": 1970
    },
    {
      "epoch": 0.6041189931350115,
      "grad_norm": 2.409759044647217,
      "learning_rate": 2.0111558723272388e-05,
      "loss": 0.7283,
      "step": 1980
    },
    {
      "epoch": 0.6071700991609459,
      "grad_norm": 3.212092876434326,
      "learning_rate": 1.995661605206074e-05,
      "loss": 0.6655,
      "step": 1990
    },
    {
      "epoch": 0.6102212051868803,
      "grad_norm": 2.577396869659424,
      "learning_rate": 1.9801673380849086e-05,
      "loss": 0.6565,
      "step": 2000
    },
    {
      "epoch": 0.6132723112128147,
      "grad_norm": 2.715590715408325,
      "learning_rate": 1.9646730709637435e-05,
      "loss": 0.6705,
      "step": 2010
    },
    {
      "epoch": 0.6163234172387491,
      "grad_norm": 2.6704838275909424,
      "learning_rate": 1.9491788038425784e-05,
      "loss": 0.7789,
      "step": 2020
    },
    {
      "epoch": 0.6193745232646835,
      "grad_norm": 2.4763495922088623,
      "learning_rate": 1.9336845367214133e-05,
      "loss": 0.8014,
      "step": 2030
    },
    {
      "epoch": 0.6224256292906178,
      "grad_norm": 2.754671573638916,
      "learning_rate": 1.918190269600248e-05,
      "loss": 0.7272,
      "step": 2040
    },
    {
      "epoch": 0.6254767353165522,
      "grad_norm": 3.223806858062744,
      "learning_rate": 1.9026960024790828e-05,
      "loss": 0.7103,
      "step": 2050
    },
    {
      "epoch": 0.6285278413424866,
      "grad_norm": 2.489527702331543,
      "learning_rate": 1.8872017353579177e-05,
      "loss": 0.6492,
      "step": 2060
    },
    {
      "epoch": 0.631578947368421,
      "grad_norm": 3.088130235671997,
      "learning_rate": 1.8717074682367526e-05,
      "loss": 0.6808,
      "step": 2070
    },
    {
      "epoch": 0.6346300533943554,
      "grad_norm": 3.130558967590332,
      "learning_rate": 1.8562132011155872e-05,
      "loss": 0.7399,
      "step": 2080
    },
    {
      "epoch": 0.6376811594202898,
      "grad_norm": 2.672267436981201,
      "learning_rate": 1.840718933994422e-05,
      "loss": 0.7012,
      "step": 2090
    },
    {
      "epoch": 0.6407322654462243,
      "grad_norm": 3.445889472961426,
      "learning_rate": 1.825224666873257e-05,
      "loss": 0.7329,
      "step": 2100
    },
    {
      "epoch": 0.6437833714721587,
      "grad_norm": 2.1221354007720947,
      "learning_rate": 1.809730399752092e-05,
      "loss": 0.7105,
      "step": 2110
    },
    {
      "epoch": 0.6468344774980931,
      "grad_norm": 2.5181918144226074,
      "learning_rate": 1.7942361326309268e-05,
      "loss": 0.7719,
      "step": 2120
    },
    {
      "epoch": 0.6498855835240275,
      "grad_norm": 3.100332021713257,
      "learning_rate": 1.7787418655097614e-05,
      "loss": 0.6837,
      "step": 2130
    },
    {
      "epoch": 0.6529366895499619,
      "grad_norm": 3.2809691429138184,
      "learning_rate": 1.7632475983885963e-05,
      "loss": 0.7173,
      "step": 2140
    },
    {
      "epoch": 0.6559877955758963,
      "grad_norm": 2.191570520401001,
      "learning_rate": 1.747753331267431e-05,
      "loss": 0.6494,
      "step": 2150
    },
    {
      "epoch": 0.6590389016018307,
      "grad_norm": 2.8507699966430664,
      "learning_rate": 1.732259064146266e-05,
      "loss": 0.7094,
      "step": 2160
    },
    {
      "epoch": 0.662090007627765,
      "grad_norm": 2.7249112129211426,
      "learning_rate": 1.7167647970251006e-05,
      "loss": 0.6549,
      "step": 2170
    },
    {
      "epoch": 0.6651411136536994,
      "grad_norm": 1.7811733484268188,
      "learning_rate": 1.7012705299039355e-05,
      "loss": 0.6353,
      "step": 2180
    },
    {
      "epoch": 0.6681922196796338,
      "grad_norm": 2.451258659362793,
      "learning_rate": 1.6857762627827704e-05,
      "loss": 0.7104,
      "step": 2190
    },
    {
      "epoch": 0.6712433257055682,
      "grad_norm": 2.2157397270202637,
      "learning_rate": 1.6702819956616053e-05,
      "loss": 0.7135,
      "step": 2200
    },
    {
      "epoch": 0.6742944317315027,
      "grad_norm": 2.1415586471557617,
      "learning_rate": 1.65478772854044e-05,
      "loss": 0.6959,
      "step": 2210
    },
    {
      "epoch": 0.6773455377574371,
      "grad_norm": 2.9360268115997314,
      "learning_rate": 1.6392934614192748e-05,
      "loss": 0.7511,
      "step": 2220
    },
    {
      "epoch": 0.6803966437833715,
      "grad_norm": 2.7337160110473633,
      "learning_rate": 1.6237991942981097e-05,
      "loss": 0.7181,
      "step": 2230
    },
    {
      "epoch": 0.6834477498093059,
      "grad_norm": 3.827877998352051,
      "learning_rate": 1.6083049271769446e-05,
      "loss": 0.6735,
      "step": 2240
    },
    {
      "epoch": 0.6864988558352403,
      "grad_norm": 2.284428834915161,
      "learning_rate": 1.5928106600557792e-05,
      "loss": 0.6822,
      "step": 2250
    },
    {
      "epoch": 0.6895499618611747,
      "grad_norm": 2.9292097091674805,
      "learning_rate": 1.5773163929346144e-05,
      "loss": 0.7522,
      "step": 2260
    },
    {
      "epoch": 0.6926010678871091,
      "grad_norm": 2.607977867126465,
      "learning_rate": 1.561822125813449e-05,
      "loss": 0.656,
      "step": 2270
    },
    {
      "epoch": 0.6956521739130435,
      "grad_norm": 3.1531264781951904,
      "learning_rate": 1.546327858692284e-05,
      "loss": 0.7427,
      "step": 2280
    },
    {
      "epoch": 0.6987032799389779,
      "grad_norm": 2.421393871307373,
      "learning_rate": 1.5308335915711188e-05,
      "loss": 0.6408,
      "step": 2290
    },
    {
      "epoch": 0.7017543859649122,
      "grad_norm": 3.7150917053222656,
      "learning_rate": 1.5153393244499537e-05,
      "loss": 0.7536,
      "step": 2300
    },
    {
      "epoch": 0.7048054919908466,
      "grad_norm": 2.622455358505249,
      "learning_rate": 1.4998450573287884e-05,
      "loss": 0.7359,
      "step": 2310
    },
    {
      "epoch": 0.7078565980167811,
      "grad_norm": 2.524982452392578,
      "learning_rate": 1.4843507902076232e-05,
      "loss": 0.689,
      "step": 2320
    },
    {
      "epoch": 0.7109077040427155,
      "grad_norm": 3.156085729598999,
      "learning_rate": 1.4688565230864582e-05,
      "loss": 0.7111,
      "step": 2330
    },
    {
      "epoch": 0.7139588100686499,
      "grad_norm": 2.8505616188049316,
      "learning_rate": 1.453362255965293e-05,
      "loss": 0.6698,
      "step": 2340
    },
    {
      "epoch": 0.7170099160945843,
      "grad_norm": 2.8530240058898926,
      "learning_rate": 1.4378679888441277e-05,
      "loss": 0.7196,
      "step": 2350
    },
    {
      "epoch": 0.7200610221205187,
      "grad_norm": 2.8734195232391357,
      "learning_rate": 1.4223737217229626e-05,
      "loss": 0.7656,
      "step": 2360
    },
    {
      "epoch": 0.7231121281464531,
      "grad_norm": 2.84820556640625,
      "learning_rate": 1.4068794546017975e-05,
      "loss": 0.6777,
      "step": 2370
    },
    {
      "epoch": 0.7261632341723875,
      "grad_norm": 2.4297478199005127,
      "learning_rate": 1.3913851874806322e-05,
      "loss": 0.7145,
      "step": 2380
    },
    {
      "epoch": 0.7292143401983219,
      "grad_norm": 2.734090566635132,
      "learning_rate": 1.375890920359467e-05,
      "loss": 0.6577,
      "step": 2390
    },
    {
      "epoch": 0.7322654462242563,
      "grad_norm": 2.2257931232452393,
      "learning_rate": 1.3603966532383019e-05,
      "loss": 0.6857,
      "step": 2400
    },
    {
      "epoch": 0.7353165522501907,
      "grad_norm": 2.3681576251983643,
      "learning_rate": 1.3449023861171368e-05,
      "loss": 0.7619,
      "step": 2410
    },
    {
      "epoch": 0.738367658276125,
      "grad_norm": 2.033923864364624,
      "learning_rate": 1.3294081189959715e-05,
      "loss": 0.691,
      "step": 2420
    },
    {
      "epoch": 0.7414187643020596,
      "grad_norm": 2.131638526916504,
      "learning_rate": 1.3139138518748064e-05,
      "loss": 0.7191,
      "step": 2430
    },
    {
      "epoch": 0.7444698703279939,
      "grad_norm": 2.609546661376953,
      "learning_rate": 1.2984195847536412e-05,
      "loss": 0.6737,
      "step": 2440
    },
    {
      "epoch": 0.7475209763539283,
      "grad_norm": 2.4548799991607666,
      "learning_rate": 1.282925317632476e-05,
      "loss": 0.6301,
      "step": 2450
    },
    {
      "epoch": 0.7505720823798627,
      "grad_norm": 2.644568920135498,
      "learning_rate": 1.267431050511311e-05,
      "loss": 0.6824,
      "step": 2460
    },
    {
      "epoch": 0.7536231884057971,
      "grad_norm": 2.595654249191284,
      "learning_rate": 1.2519367833901457e-05,
      "loss": 0.654,
      "step": 2470
    },
    {
      "epoch": 0.7566742944317315,
      "grad_norm": 2.7976434230804443,
      "learning_rate": 1.2364425162689804e-05,
      "loss": 0.7883,
      "step": 2480
    },
    {
      "epoch": 0.7597254004576659,
      "grad_norm": 2.874528646469116,
      "learning_rate": 1.2209482491478153e-05,
      "loss": 0.7202,
      "step": 2490
    },
    {
      "epoch": 0.7627765064836003,
      "grad_norm": 2.6641483306884766,
      "learning_rate": 1.20545398202665e-05,
      "loss": 0.6692,
      "step": 2500
    },
    {
      "epoch": 0.7658276125095347,
      "grad_norm": 2.610469102859497,
      "learning_rate": 1.189959714905485e-05,
      "loss": 0.7201,
      "step": 2510
    },
    {
      "epoch": 0.7688787185354691,
      "grad_norm": 2.2959210872650146,
      "learning_rate": 1.1744654477843199e-05,
      "loss": 0.6302,
      "step": 2520
    },
    {
      "epoch": 0.7719298245614035,
      "grad_norm": 2.454118013381958,
      "learning_rate": 1.1589711806631546e-05,
      "loss": 0.7102,
      "step": 2530
    },
    {
      "epoch": 0.7749809305873379,
      "grad_norm": 2.6348319053649902,
      "learning_rate": 1.1434769135419895e-05,
      "loss": 0.7214,
      "step": 2540
    },
    {
      "epoch": 0.7780320366132724,
      "grad_norm": 2.254340887069702,
      "learning_rate": 1.1279826464208242e-05,
      "loss": 0.6973,
      "step": 2550
    },
    {
      "epoch": 0.7810831426392068,
      "grad_norm": 2.1454594135284424,
      "learning_rate": 1.1124883792996592e-05,
      "loss": 0.7037,
      "step": 2560
    },
    {
      "epoch": 0.7841342486651411,
      "grad_norm": 2.7335195541381836,
      "learning_rate": 1.096994112178494e-05,
      "loss": 0.7059,
      "step": 2570
    },
    {
      "epoch": 0.7871853546910755,
      "grad_norm": 3.499448776245117,
      "learning_rate": 1.0814998450573288e-05,
      "loss": 0.6518,
      "step": 2580
    },
    {
      "epoch": 0.7902364607170099,
      "grad_norm": 2.517699956893921,
      "learning_rate": 1.0660055779361637e-05,
      "loss": 0.6884,
      "step": 2590
    },
    {
      "epoch": 0.7932875667429443,
      "grad_norm": 2.7374229431152344,
      "learning_rate": 1.0505113108149984e-05,
      "loss": 0.6146,
      "step": 2600
    },
    {
      "epoch": 0.7963386727688787,
      "grad_norm": 3.5942227840423584,
      "learning_rate": 1.0350170436938333e-05,
      "loss": 0.7788,
      "step": 2610
    },
    {
      "epoch": 0.7993897787948131,
      "grad_norm": 2.709524393081665,
      "learning_rate": 1.019522776572668e-05,
      "loss": 0.6843,
      "step": 2620
    },
    {
      "epoch": 0.8024408848207475,
      "grad_norm": 2.7754933834075928,
      "learning_rate": 1.004028509451503e-05,
      "loss": 0.7694,
      "step": 2630
    },
    {
      "epoch": 0.8054919908466819,
      "grad_norm": 2.025585412979126,
      "learning_rate": 9.885342423303379e-06,
      "loss": 0.6237,
      "step": 2640
    },
    {
      "epoch": 0.8085430968726163,
      "grad_norm": 2.5337514877319336,
      "learning_rate": 9.730399752091726e-06,
      "loss": 0.7284,
      "step": 2650
    },
    {
      "epoch": 0.8115942028985508,
      "grad_norm": 2.5185277462005615,
      "learning_rate": 9.575457080880075e-06,
      "loss": 0.695,
      "step": 2660
    },
    {
      "epoch": 0.8146453089244852,
      "grad_norm": 3.0368494987487793,
      "learning_rate": 9.420514409668422e-06,
      "loss": 0.7291,
      "step": 2670
    },
    {
      "epoch": 0.8176964149504196,
      "grad_norm": 2.6141366958618164,
      "learning_rate": 9.265571738456771e-06,
      "loss": 0.6768,
      "step": 2680
    },
    {
      "epoch": 0.820747520976354,
      "grad_norm": 2.2215969562530518,
      "learning_rate": 9.11062906724512e-06,
      "loss": 0.6832,
      "step": 2690
    },
    {
      "epoch": 0.8237986270022883,
      "grad_norm": 3.0464236736297607,
      "learning_rate": 8.955686396033468e-06,
      "loss": 0.7362,
      "step": 2700
    },
    {
      "epoch": 0.8268497330282227,
      "grad_norm": 2.384838104248047,
      "learning_rate": 8.800743724821817e-06,
      "loss": 0.6501,
      "step": 2710
    },
    {
      "epoch": 0.8299008390541571,
      "grad_norm": 2.2411282062530518,
      "learning_rate": 8.645801053610164e-06,
      "loss": 0.6274,
      "step": 2720
    },
    {
      "epoch": 0.8329519450800915,
      "grad_norm": 2.8559658527374268,
      "learning_rate": 8.490858382398513e-06,
      "loss": 0.6282,
      "step": 2730
    },
    {
      "epoch": 0.8360030511060259,
      "grad_norm": 2.729321002960205,
      "learning_rate": 8.335915711186862e-06,
      "loss": 0.6966,
      "step": 2740
    },
    {
      "epoch": 0.8390541571319603,
      "grad_norm": 2.846073865890503,
      "learning_rate": 8.18097303997521e-06,
      "loss": 0.6925,
      "step": 2750
    },
    {
      "epoch": 0.8421052631578947,
      "grad_norm": 2.1397573947906494,
      "learning_rate": 8.026030368763559e-06,
      "loss": 0.6435,
      "step": 2760
    },
    {
      "epoch": 0.8451563691838292,
      "grad_norm": 3.3590445518493652,
      "learning_rate": 7.871087697551906e-06,
      "loss": 0.6592,
      "step": 2770
    },
    {
      "epoch": 0.8482074752097636,
      "grad_norm": 2.7317698001861572,
      "learning_rate": 7.716145026340255e-06,
      "loss": 0.6915,
      "step": 2780
    },
    {
      "epoch": 0.851258581235698,
      "grad_norm": 2.3746912479400635,
      "learning_rate": 7.561202355128602e-06,
      "loss": 0.7039,
      "step": 2790
    },
    {
      "epoch": 0.8543096872616324,
      "grad_norm": 3.0164999961853027,
      "learning_rate": 7.4062596839169505e-06,
      "loss": 0.7024,
      "step": 2800
    },
    {
      "epoch": 0.8573607932875668,
      "grad_norm": 3.2100493907928467,
      "learning_rate": 7.2513170127053e-06,
      "loss": 0.7137,
      "step": 2810
    },
    {
      "epoch": 0.8604118993135011,
      "grad_norm": 2.1263020038604736,
      "learning_rate": 7.096374341493647e-06,
      "loss": 0.6334,
      "step": 2820
    },
    {
      "epoch": 0.8634630053394355,
      "grad_norm": 2.614156484603882,
      "learning_rate": 6.941431670281996e-06,
      "loss": 0.7299,
      "step": 2830
    },
    {
      "epoch": 0.8665141113653699,
      "grad_norm": 4.311347961425781,
      "learning_rate": 6.786488999070343e-06,
      "loss": 0.6464,
      "step": 2840
    },
    {
      "epoch": 0.8695652173913043,
      "grad_norm": 2.2194840908050537,
      "learning_rate": 6.631546327858692e-06,
      "loss": 0.7055,
      "step": 2850
    },
    {
      "epoch": 0.8726163234172387,
      "grad_norm": 2.8160476684570312,
      "learning_rate": 6.476603656647041e-06,
      "loss": 0.7233,
      "step": 2860
    },
    {
      "epoch": 0.8756674294431731,
      "grad_norm": 1.9882564544677734,
      "learning_rate": 6.321660985435389e-06,
      "loss": 0.7041,
      "step": 2870
    },
    {
      "epoch": 0.8787185354691075,
      "grad_norm": 2.381114959716797,
      "learning_rate": 6.166718314223738e-06,
      "loss": 0.6916,
      "step": 2880
    },
    {
      "epoch": 0.881769641495042,
      "grad_norm": 2.911763906478882,
      "learning_rate": 6.011775643012086e-06,
      "loss": 0.6493,
      "step": 2890
    },
    {
      "epoch": 0.8848207475209764,
      "grad_norm": 2.830087900161743,
      "learning_rate": 5.856832971800434e-06,
      "loss": 0.6497,
      "step": 2900
    },
    {
      "epoch": 0.8878718535469108,
      "grad_norm": 2.609445333480835,
      "learning_rate": 5.701890300588782e-06,
      "loss": 0.6923,
      "step": 2910
    },
    {
      "epoch": 0.8909229595728452,
      "grad_norm": 2.1896908283233643,
      "learning_rate": 5.5469476293771305e-06,
      "loss": 0.6325,
      "step": 2920
    },
    {
      "epoch": 0.8939740655987796,
      "grad_norm": 3.2182352542877197,
      "learning_rate": 5.392004958165479e-06,
      "loss": 0.6613,
      "step": 2930
    },
    {
      "epoch": 0.897025171624714,
      "grad_norm": 2.7768335342407227,
      "learning_rate": 5.237062286953828e-06,
      "loss": 0.6944,
      "step": 2940
    },
    {
      "epoch": 0.9000762776506483,
      "grad_norm": 2.458332061767578,
      "learning_rate": 5.082119615742176e-06,
      "loss": 0.6689,
      "step": 2950
    },
    {
      "epoch": 0.9031273836765827,
      "grad_norm": 2.2689878940582275,
      "learning_rate": 4.927176944530524e-06,
      "loss": 0.6729,
      "step": 2960
    },
    {
      "epoch": 0.9061784897025171,
      "grad_norm": 3.256464719772339,
      "learning_rate": 4.772234273318872e-06,
      "loss": 0.6158,
      "step": 2970
    },
    {
      "epoch": 0.9092295957284515,
      "grad_norm": 3.0589096546173096,
      "learning_rate": 4.6172916021072205e-06,
      "loss": 0.7551,
      "step": 2980
    },
    {
      "epoch": 0.9122807017543859,
      "grad_norm": 2.630323648452759,
      "learning_rate": 4.4623489308955695e-06,
      "loss": 0.6235,
      "step": 2990
    },
    {
      "epoch": 0.9153318077803204,
      "grad_norm": 2.285128593444824,
      "learning_rate": 4.307406259683918e-06,
      "loss": 0.6948,
      "step": 3000
    },
    {
      "epoch": 0.9183829138062548,
      "grad_norm": 2.4126858711242676,
      "learning_rate": 4.152463588472266e-06,
      "loss": 0.6249,
      "step": 3010
    },
    {
      "epoch": 0.9214340198321892,
      "grad_norm": 2.4043056964874268,
      "learning_rate": 3.997520917260614e-06,
      "loss": 0.6822,
      "step": 3020
    },
    {
      "epoch": 0.9244851258581236,
      "grad_norm": 2.084949254989624,
      "learning_rate": 3.842578246048962e-06,
      "loss": 0.7345,
      "step": 3030
    },
    {
      "epoch": 0.927536231884058,
      "grad_norm": 2.251657247543335,
      "learning_rate": 3.68763557483731e-06,
      "loss": 0.6048,
      "step": 3040
    },
    {
      "epoch": 0.9305873379099924,
      "grad_norm": 2.20843505859375,
      "learning_rate": 3.532692903625659e-06,
      "loss": 0.6687,
      "step": 3050
    },
    {
      "epoch": 0.9336384439359268,
      "grad_norm": 2.692772150039673,
      "learning_rate": 3.3777502324140072e-06,
      "loss": 0.651,
      "step": 3060
    },
    {
      "epoch": 0.9366895499618612,
      "grad_norm": 2.778233766555786,
      "learning_rate": 3.2228075612023554e-06,
      "loss": 0.7709,
      "step": 3070
    },
    {
      "epoch": 0.9397406559877955,
      "grad_norm": 2.2121570110321045,
      "learning_rate": 3.0678648899907036e-06,
      "loss": 0.6741,
      "step": 3080
    },
    {
      "epoch": 0.9427917620137299,
      "grad_norm": 2.814641237258911,
      "learning_rate": 2.9129222187790522e-06,
      "loss": 0.6409,
      "step": 3090
    },
    {
      "epoch": 0.9458428680396643,
      "grad_norm": 2.183281660079956,
      "learning_rate": 2.7579795475674004e-06,
      "loss": 0.6192,
      "step": 3100
    },
    {
      "epoch": 0.9488939740655988,
      "grad_norm": 1.998976230621338,
      "learning_rate": 2.6030368763557486e-06,
      "loss": 0.5924,
      "step": 3110
    },
    {
      "epoch": 0.9519450800915332,
      "grad_norm": 2.7774434089660645,
      "learning_rate": 2.4480942051440968e-06,
      "loss": 0.7145,
      "step": 3120
    },
    {
      "epoch": 0.9549961861174676,
      "grad_norm": 2.510991334915161,
      "learning_rate": 2.293151533932445e-06,
      "loss": 0.6985,
      "step": 3130
    },
    {
      "epoch": 0.958047292143402,
      "grad_norm": 2.820533037185669,
      "learning_rate": 2.138208862720793e-06,
      "loss": 0.673,
      "step": 3140
    },
    {
      "epoch": 0.9610983981693364,
      "grad_norm": 3.7463274002075195,
      "learning_rate": 1.9832661915091418e-06,
      "loss": 0.6753,
      "step": 3150
    },
    {
      "epoch": 0.9641495041952708,
      "grad_norm": 2.6098906993865967,
      "learning_rate": 1.82832352029749e-06,
      "loss": 0.6077,
      "step": 3160
    },
    {
      "epoch": 0.9672006102212052,
      "grad_norm": 3.7235538959503174,
      "learning_rate": 1.6733808490858381e-06,
      "loss": 0.6528,
      "step": 3170
    },
    {
      "epoch": 0.9702517162471396,
      "grad_norm": 2.551973342895508,
      "learning_rate": 1.5184381778741865e-06,
      "loss": 0.7123,
      "step": 3180
    },
    {
      "epoch": 0.973302822273074,
      "grad_norm": 2.60445237159729,
      "learning_rate": 1.363495506662535e-06,
      "loss": 0.6368,
      "step": 3190
    },
    {
      "epoch": 0.9763539282990084,
      "grad_norm": 2.7576358318328857,
      "learning_rate": 1.2085528354508833e-06,
      "loss": 0.6172,
      "step": 3200
    },
    {
      "epoch": 0.9794050343249427,
      "grad_norm": 2.23207426071167,
      "learning_rate": 1.0536101642392315e-06,
      "loss": 0.6227,
      "step": 3210
    },
    {
      "epoch": 0.9824561403508771,
      "grad_norm": 3.0700480937957764,
      "learning_rate": 8.986674930275798e-07,
      "loss": 0.7168,
      "step": 3220
    },
    {
      "epoch": 0.9855072463768116,
      "grad_norm": 2.6852805614471436,
      "learning_rate": 7.437248218159281e-07,
      "loss": 0.6322,
      "step": 3230
    },
    {
      "epoch": 0.988558352402746,
      "grad_norm": 2.6967718601226807,
      "learning_rate": 5.887821506042764e-07,
      "loss": 0.6357,
      "step": 3240
    },
    {
      "epoch": 0.9916094584286804,
      "grad_norm": 3.2857532501220703,
      "learning_rate": 4.3383947939262475e-07,
      "loss": 0.7225,
      "step": 3250
    },
    {
      "epoch": 0.9946605644546148,
      "grad_norm": 2.565335750579834,
      "learning_rate": 2.788968081809731e-07,
      "loss": 0.6709,
      "step": 3260
    },
    {
      "epoch": 0.9977116704805492,
      "grad_norm": 3.099731922149658,
      "learning_rate": 1.2395413696932136e-07,
      "loss": 0.7549,
      "step": 3270
    }
  ],
  "logging_steps": 10,
  "max_steps": 3277,
  "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": 1.2252825778126848e+16,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}