{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9998451612903225,
  "eval_steps": 500,
  "global_step": 4843,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.002064516129032258,
      "grad_norm": 3.8347715644338702,
      "learning_rate": 2.0618556701030925e-08,
      "loss": 0.4845,
      "step": 10
    },
    {
      "epoch": 0.004129032258064516,
      "grad_norm": 3.622781800764145,
      "learning_rate": 4.123711340206185e-08,
      "loss": 0.4768,
      "step": 20
    },
    {
      "epoch": 0.006193548387096774,
      "grad_norm": 3.2353201981730644,
      "learning_rate": 6.185567010309278e-08,
      "loss": 0.4652,
      "step": 30
    },
    {
      "epoch": 0.008258064516129033,
      "grad_norm": 2.947330357126175,
      "learning_rate": 8.24742268041237e-08,
      "loss": 0.4542,
      "step": 40
    },
    {
      "epoch": 0.01032258064516129,
      "grad_norm": 2.417908848980943,
      "learning_rate": 1.0309278350515462e-07,
      "loss": 0.4363,
      "step": 50
    },
    {
      "epoch": 0.012387096774193548,
      "grad_norm": 1.74970179232699,
      "learning_rate": 1.2371134020618556e-07,
      "loss": 0.4185,
      "step": 60
    },
    {
      "epoch": 0.014451612903225806,
      "grad_norm": 1.4949679798399427,
      "learning_rate": 1.4432989690721648e-07,
      "loss": 0.4099,
      "step": 70
    },
    {
      "epoch": 0.016516129032258065,
      "grad_norm": 1.2117967081127008,
      "learning_rate": 1.649484536082474e-07,
      "loss": 0.3901,
      "step": 80
    },
    {
      "epoch": 0.01858064516129032,
      "grad_norm": 1.0236940060259172,
      "learning_rate": 1.8556701030927835e-07,
      "loss": 0.3781,
      "step": 90
    },
    {
      "epoch": 0.02064516129032258,
      "grad_norm": 0.8673034546104332,
      "learning_rate": 2.0618556701030925e-07,
      "loss": 0.3732,
      "step": 100
    },
    {
      "epoch": 0.02270967741935484,
      "grad_norm": 0.8133872914570585,
      "learning_rate": 2.268041237113402e-07,
      "loss": 0.3728,
      "step": 110
    },
    {
      "epoch": 0.024774193548387096,
      "grad_norm": 0.8191926249288893,
      "learning_rate": 2.474226804123711e-07,
      "loss": 0.3613,
      "step": 120
    },
    {
      "epoch": 0.026838709677419356,
      "grad_norm": 0.7422414846989145,
      "learning_rate": 2.6804123711340204e-07,
      "loss": 0.359,
      "step": 130
    },
    {
      "epoch": 0.02890322580645161,
      "grad_norm": 0.7201726657809826,
      "learning_rate": 2.8865979381443296e-07,
      "loss": 0.3598,
      "step": 140
    },
    {
      "epoch": 0.03096774193548387,
      "grad_norm": 0.77452988304961,
      "learning_rate": 3.0927835051546394e-07,
      "loss": 0.3557,
      "step": 150
    },
    {
      "epoch": 0.03303225806451613,
      "grad_norm": 0.7191351812558439,
      "learning_rate": 3.298969072164948e-07,
      "loss": 0.3508,
      "step": 160
    },
    {
      "epoch": 0.03509677419354839,
      "grad_norm": 0.6938460773620077,
      "learning_rate": 3.5051546391752573e-07,
      "loss": 0.3512,
      "step": 170
    },
    {
      "epoch": 0.03716129032258064,
      "grad_norm": 0.6640697172414236,
      "learning_rate": 3.711340206185567e-07,
      "loss": 0.3498,
      "step": 180
    },
    {
      "epoch": 0.0392258064516129,
      "grad_norm": 0.6978131389871531,
      "learning_rate": 3.917525773195876e-07,
      "loss": 0.3372,
      "step": 190
    },
    {
      "epoch": 0.04129032258064516,
      "grad_norm": 0.6748840406060271,
      "learning_rate": 4.123711340206185e-07,
      "loss": 0.3473,
      "step": 200
    },
    {
      "epoch": 0.04335483870967742,
      "grad_norm": 0.6903300485581471,
      "learning_rate": 4.3298969072164947e-07,
      "loss": 0.3436,
      "step": 210
    },
    {
      "epoch": 0.04541935483870968,
      "grad_norm": 0.6900391616850294,
      "learning_rate": 4.536082474226804e-07,
      "loss": 0.3458,
      "step": 220
    },
    {
      "epoch": 0.04748387096774193,
      "grad_norm": 0.7042637181386346,
      "learning_rate": 4.742268041237113e-07,
      "loss": 0.3377,
      "step": 230
    },
    {
      "epoch": 0.04954838709677419,
      "grad_norm": 0.659813926490704,
      "learning_rate": 4.948453608247422e-07,
      "loss": 0.3387,
      "step": 240
    },
    {
      "epoch": 0.05161290322580645,
      "grad_norm": 0.6515723207892065,
      "learning_rate": 5.154639175257731e-07,
      "loss": 0.3443,
      "step": 250
    },
    {
      "epoch": 0.05367741935483871,
      "grad_norm": 0.6545766074333088,
      "learning_rate": 5.360824742268041e-07,
      "loss": 0.342,
      "step": 260
    },
    {
      "epoch": 0.05574193548387097,
      "grad_norm": 0.6637287687009573,
      "learning_rate": 5.56701030927835e-07,
      "loss": 0.3343,
      "step": 270
    },
    {
      "epoch": 0.05780645161290322,
      "grad_norm": 0.6483801784518098,
      "learning_rate": 5.773195876288659e-07,
      "loss": 0.3315,
      "step": 280
    },
    {
      "epoch": 0.05987096774193548,
      "grad_norm": 0.6709786596445975,
      "learning_rate": 5.979381443298969e-07,
      "loss": 0.3449,
      "step": 290
    },
    {
      "epoch": 0.06193548387096774,
      "grad_norm": 0.6388273578010589,
      "learning_rate": 6.185567010309279e-07,
      "loss": 0.3331,
      "step": 300
    },
    {
      "epoch": 0.064,
      "grad_norm": 0.6215237839848085,
      "learning_rate": 6.391752577319586e-07,
      "loss": 0.3364,
      "step": 310
    },
    {
      "epoch": 0.06606451612903226,
      "grad_norm": 0.6960918549923397,
      "learning_rate": 6.597938144329896e-07,
      "loss": 0.3434,
      "step": 320
    },
    {
      "epoch": 0.06812903225806452,
      "grad_norm": 0.6810690836972719,
      "learning_rate": 6.804123711340206e-07,
      "loss": 0.3368,
      "step": 330
    },
    {
      "epoch": 0.07019354838709678,
      "grad_norm": 0.6358648816123277,
      "learning_rate": 7.010309278350515e-07,
      "loss": 0.3338,
      "step": 340
    },
    {
      "epoch": 0.07225806451612904,
      "grad_norm": 0.668088207474176,
      "learning_rate": 7.216494845360824e-07,
      "loss": 0.3414,
      "step": 350
    },
    {
      "epoch": 0.07432258064516128,
      "grad_norm": 0.6478792898005766,
      "learning_rate": 7.422680412371134e-07,
      "loss": 0.3364,
      "step": 360
    },
    {
      "epoch": 0.07638709677419354,
      "grad_norm": 0.6795998111394932,
      "learning_rate": 7.628865979381443e-07,
      "loss": 0.327,
      "step": 370
    },
    {
      "epoch": 0.0784516129032258,
      "grad_norm": 0.6525046715909505,
      "learning_rate": 7.835051546391752e-07,
      "loss": 0.3348,
      "step": 380
    },
    {
      "epoch": 0.08051612903225806,
      "grad_norm": 0.685633155774819,
      "learning_rate": 8.041237113402062e-07,
      "loss": 0.3314,
      "step": 390
    },
    {
      "epoch": 0.08258064516129032,
      "grad_norm": 0.6308341522120817,
      "learning_rate": 8.24742268041237e-07,
      "loss": 0.3346,
      "step": 400
    },
    {
      "epoch": 0.08464516129032258,
      "grad_norm": 0.6777484559645547,
      "learning_rate": 8.45360824742268e-07,
      "loss": 0.332,
      "step": 410
    },
    {
      "epoch": 0.08670967741935484,
      "grad_norm": 0.6664332905967745,
      "learning_rate": 8.659793814432989e-07,
      "loss": 0.3276,
      "step": 420
    },
    {
      "epoch": 0.0887741935483871,
      "grad_norm": 0.678260164078579,
      "learning_rate": 8.865979381443298e-07,
      "loss": 0.3347,
      "step": 430
    },
    {
      "epoch": 0.09083870967741936,
      "grad_norm": 0.6639398379059098,
      "learning_rate": 9.072164948453608e-07,
      "loss": 0.3349,
      "step": 440
    },
    {
      "epoch": 0.09290322580645162,
      "grad_norm": 0.6544729326125452,
      "learning_rate": 9.278350515463918e-07,
      "loss": 0.3351,
      "step": 450
    },
    {
      "epoch": 0.09496774193548387,
      "grad_norm": 0.6736956960554942,
      "learning_rate": 9.484536082474226e-07,
      "loss": 0.3364,
      "step": 460
    },
    {
      "epoch": 0.09703225806451612,
      "grad_norm": 0.6798760919991121,
      "learning_rate": 9.690721649484535e-07,
      "loss": 0.3329,
      "step": 470
    },
    {
      "epoch": 0.09909677419354838,
      "grad_norm": 0.6860410772296518,
      "learning_rate": 9.896907216494845e-07,
      "loss": 0.3267,
      "step": 480
    },
    {
      "epoch": 0.10116129032258064,
      "grad_norm": 0.6724242514524695,
      "learning_rate": 9.999967520836107e-07,
      "loss": 0.3301,
      "step": 490
    },
    {
      "epoch": 0.1032258064516129,
      "grad_norm": 0.7146272907830368,
      "learning_rate": 9.999707690056706e-07,
      "loss": 0.3405,
      "step": 500
    },
    {
      "epoch": 0.10529032258064516,
      "grad_norm": 0.6602284351487545,
      "learning_rate": 9.999188042000401e-07,
      "loss": 0.3331,
      "step": 510
    },
    {
      "epoch": 0.10735483870967742,
      "grad_norm": 0.6500269744684952,
      "learning_rate": 9.998408603671476e-07,
      "loss": 0.3349,
      "step": 520
    },
    {
      "epoch": 0.10941935483870968,
      "grad_norm": 0.6659080270743315,
      "learning_rate": 9.997369415574612e-07,
      "loss": 0.3255,
      "step": 530
    },
    {
      "epoch": 0.11148387096774194,
      "grad_norm": 0.6616751264530504,
      "learning_rate": 9.996070531712766e-07,
      "loss": 0.3277,
      "step": 540
    },
    {
      "epoch": 0.1135483870967742,
      "grad_norm": 0.6964651744599499,
      "learning_rate": 9.99451201958438e-07,
      "loss": 0.3317,
      "step": 550
    },
    {
      "epoch": 0.11561290322580645,
      "grad_norm": 0.6322033648536264,
      "learning_rate": 9.992693960179864e-07,
      "loss": 0.3274,
      "step": 560
    },
    {
      "epoch": 0.1176774193548387,
      "grad_norm": 0.6651477736231276,
      "learning_rate": 9.99061644797739e-07,
      "loss": 0.3362,
      "step": 570
    },
    {
      "epoch": 0.11974193548387096,
      "grad_norm": 0.6508795261048231,
      "learning_rate": 9.988279590937983e-07,
      "loss": 0.3304,
      "step": 580
    },
    {
      "epoch": 0.12180645161290322,
      "grad_norm": 0.6364459903886801,
      "learning_rate": 9.985683510499907e-07,
      "loss": 0.3242,
      "step": 590
    },
    {
      "epoch": 0.12387096774193548,
      "grad_norm": 0.6619954180634886,
      "learning_rate": 9.982828341572362e-07,
      "loss": 0.329,
      "step": 600
    },
    {
      "epoch": 0.12593548387096773,
      "grad_norm": 0.6413848598383421,
      "learning_rate": 9.979714232528463e-07,
      "loss": 0.3229,
      "step": 610
    },
    {
      "epoch": 0.128,
      "grad_norm": 0.6424245333342599,
      "learning_rate": 9.97634134519754e-07,
      "loss": 0.3311,
      "step": 620
    },
    {
      "epoch": 0.13006451612903225,
      "grad_norm": 0.6344772952123243,
      "learning_rate": 9.972709854856716e-07,
      "loss": 0.3306,
      "step": 630
    },
    {
      "epoch": 0.13212903225806452,
      "grad_norm": 0.6790645384085968,
      "learning_rate": 9.968819950221812e-07,
      "loss": 0.3218,
      "step": 640
    },
    {
      "epoch": 0.13419354838709677,
      "grad_norm": 0.6360664660727335,
      "learning_rate": 9.964671833437533e-07,
      "loss": 0.3296,
      "step": 650
    },
    {
      "epoch": 0.13625806451612904,
      "grad_norm": 0.6708293122866612,
      "learning_rate": 9.960265720066961e-07,
      "loss": 0.3194,
      "step": 660
    },
    {
      "epoch": 0.13832258064516129,
      "grad_norm": 0.682103425250404,
      "learning_rate": 9.95560183908036e-07,
      "loss": 0.3216,
      "step": 670
    },
    {
      "epoch": 0.14038709677419356,
      "grad_norm": 0.649381758646969,
      "learning_rate": 9.950680432843267e-07,
      "loss": 0.326,
      "step": 680
    },
    {
      "epoch": 0.1424516129032258,
      "grad_norm": 0.6584415316713365,
      "learning_rate": 9.94550175710391e-07,
      "loss": 0.3268,
      "step": 690
    },
    {
      "epoch": 0.14451612903225808,
      "grad_norm": 0.6299571550013747,
      "learning_rate": 9.940066080979909e-07,
      "loss": 0.3233,
      "step": 700
    },
    {
      "epoch": 0.14658064516129032,
      "grad_norm": 0.6538409379488348,
      "learning_rate": 9.934373686944286e-07,
      "loss": 0.3259,
      "step": 710
    },
    {
      "epoch": 0.14864516129032257,
      "grad_norm": 0.6261303918046864,
      "learning_rate": 9.928424870810804e-07,
      "loss": 0.3227,
      "step": 720
    },
    {
      "epoch": 0.15070967741935484,
      "grad_norm": 0.6504421655013074,
      "learning_rate": 9.922219941718576e-07,
      "loss": 0.3245,
      "step": 730
    },
    {
      "epoch": 0.1527741935483871,
      "grad_norm": 0.6338912710677398,
      "learning_rate": 9.915759222116008e-07,
      "loss": 0.3272,
      "step": 740
    },
    {
      "epoch": 0.15483870967741936,
      "grad_norm": 0.6483630749442493,
      "learning_rate": 9.909043047744044e-07,
      "loss": 0.3322,
      "step": 750
    },
    {
      "epoch": 0.1569032258064516,
      "grad_norm": 0.6427038784083127,
      "learning_rate": 9.902071767618715e-07,
      "loss": 0.3259,
      "step": 760
    },
    {
      "epoch": 0.15896774193548388,
      "grad_norm": 0.6384424550604813,
      "learning_rate": 9.894845744013002e-07,
      "loss": 0.3299,
      "step": 770
    },
    {
      "epoch": 0.16103225806451613,
      "grad_norm": 0.6436281486654595,
      "learning_rate": 9.887365352438015e-07,
      "loss": 0.3202,
      "step": 780
    },
    {
      "epoch": 0.1630967741935484,
      "grad_norm": 0.6600853885272884,
      "learning_rate": 9.879630981623475e-07,
      "loss": 0.3316,
      "step": 790
    },
    {
      "epoch": 0.16516129032258065,
      "grad_norm": 0.7273173646706375,
      "learning_rate": 9.871643033497512e-07,
      "loss": 0.3268,
      "step": 800
    },
    {
      "epoch": 0.1672258064516129,
      "grad_norm": 0.6475774878384731,
      "learning_rate": 9.863401923165778e-07,
      "loss": 0.3279,
      "step": 810
    },
    {
      "epoch": 0.16929032258064516,
      "grad_norm": 0.6616526425163223,
      "learning_rate": 9.854908078889881e-07,
      "loss": 0.3241,
      "step": 820
    },
    {
      "epoch": 0.1713548387096774,
      "grad_norm": 0.6582121362859629,
      "learning_rate": 9.846161942065123e-07,
      "loss": 0.3256,
      "step": 830
    },
    {
      "epoch": 0.17341935483870968,
      "grad_norm": 0.6359674660372312,
      "learning_rate": 9.837163967197567e-07,
      "loss": 0.3254,
      "step": 840
    },
    {
      "epoch": 0.17548387096774193,
      "grad_norm": 0.6686054614543271,
      "learning_rate": 9.827914621880412e-07,
      "loss": 0.3271,
      "step": 850
    },
    {
      "epoch": 0.1775483870967742,
      "grad_norm": 0.64804057560635,
      "learning_rate": 9.818414386769702e-07,
      "loss": 0.3286,
      "step": 860
    },
    {
      "epoch": 0.17961290322580645,
      "grad_norm": 0.6412450081891817,
      "learning_rate": 9.808663755559345e-07,
      "loss": 0.3235,
      "step": 870
    },
    {
      "epoch": 0.18167741935483872,
      "grad_norm": 0.6233022789634588,
      "learning_rate": 9.798663234955452e-07,
      "loss": 0.325,
      "step": 880
    },
    {
      "epoch": 0.18374193548387097,
      "grad_norm": 0.6103780146802401,
      "learning_rate": 9.788413344650013e-07,
      "loss": 0.3229,
      "step": 890
    },
    {
      "epoch": 0.18580645161290324,
      "grad_norm": 0.6990894346417448,
      "learning_rate": 9.777914617293884e-07,
      "loss": 0.3167,
      "step": 900
    },
    {
      "epoch": 0.18787096774193549,
      "grad_norm": 0.668098707213347,
      "learning_rate": 9.767167598469108e-07,
      "loss": 0.3242,
      "step": 910
    },
    {
      "epoch": 0.18993548387096773,
      "grad_norm": 0.6305920111862352,
      "learning_rate": 9.756172846660576e-07,
      "loss": 0.326,
      "step": 920
    },
    {
      "epoch": 0.192,
      "grad_norm": 0.6201553079735226,
      "learning_rate": 9.744930933226977e-07,
      "loss": 0.3292,
      "step": 930
    },
    {
      "epoch": 0.19406451612903225,
      "grad_norm": 0.6555873449992747,
      "learning_rate": 9.73344244237114e-07,
      "loss": 0.3252,
      "step": 940
    },
    {
      "epoch": 0.19612903225806452,
      "grad_norm": 0.6173747729705402,
      "learning_rate": 9.721707971109645e-07,
      "loss": 0.3221,
      "step": 950
    },
    {
      "epoch": 0.19819354838709677,
      "grad_norm": 0.6409642693454491,
      "learning_rate": 9.709728129241817e-07,
      "loss": 0.3183,
      "step": 960
    },
    {
      "epoch": 0.20025806451612904,
      "grad_norm": 0.692958036040751,
      "learning_rate": 9.69750353931803e-07,
      "loss": 0.321,
      "step": 970
    },
    {
      "epoch": 0.2023225806451613,
      "grad_norm": 0.6540304743009737,
      "learning_rate": 9.685034836607358e-07,
      "loss": 0.3306,
      "step": 980
    },
    {
      "epoch": 0.20438709677419356,
      "grad_norm": 0.606153875842293,
      "learning_rate": 9.672322669064552e-07,
      "loss": 0.3267,
      "step": 990
    },
    {
      "epoch": 0.2064516129032258,
      "grad_norm": 0.6355082616879962,
      "learning_rate": 9.659367697296393e-07,
      "loss": 0.3246,
      "step": 1000
    },
    {
      "epoch": 0.20851612903225805,
      "grad_norm": 0.6540770595021805,
      "learning_rate": 9.646170594527327e-07,
      "loss": 0.3207,
      "step": 1010
    },
    {
      "epoch": 0.21058064516129033,
      "grad_norm": 0.6427115148761209,
      "learning_rate": 9.632732046564517e-07,
      "loss": 0.3341,
      "step": 1020
    },
    {
      "epoch": 0.21264516129032257,
      "grad_norm": 0.6583782369634305,
      "learning_rate": 9.619052751762172e-07,
      "loss": 0.3258,
      "step": 1030
    },
    {
      "epoch": 0.21470967741935484,
      "grad_norm": 0.6465456812217837,
      "learning_rate": 9.605133420985276e-07,
      "loss": 0.3197,
      "step": 1040
    },
    {
      "epoch": 0.2167741935483871,
      "grad_norm": 0.6392843901176241,
      "learning_rate": 9.590974777572643e-07,
      "loss": 0.3196,
      "step": 1050
    },
    {
      "epoch": 0.21883870967741936,
      "grad_norm": 0.6928785215501877,
      "learning_rate": 9.576577557299323e-07,
      "loss": 0.3238,
      "step": 1060
    },
    {
      "epoch": 0.2209032258064516,
      "grad_norm": 0.6375416863344153,
      "learning_rate": 9.561942508338368e-07,
      "loss": 0.328,
      "step": 1070
    },
    {
      "epoch": 0.22296774193548388,
      "grad_norm": 0.630092402055495,
      "learning_rate": 9.547070391221953e-07,
      "loss": 0.321,
      "step": 1080
    },
    {
      "epoch": 0.22503225806451613,
      "grad_norm": 0.6429556555744979,
      "learning_rate": 9.531961978801855e-07,
      "loss": 0.3289,
      "step": 1090
    },
    {
      "epoch": 0.2270967741935484,
      "grad_norm": 0.6477437711690731,
      "learning_rate": 9.516618056209291e-07,
      "loss": 0.3174,
      "step": 1100
    },
    {
      "epoch": 0.22916129032258065,
      "grad_norm": 0.6374229581293065,
      "learning_rate": 9.501039420814111e-07,
      "loss": 0.3236,
      "step": 1110
    },
    {
      "epoch": 0.2312258064516129,
      "grad_norm": 0.6333096872853811,
      "learning_rate": 9.485226882183372e-07,
      "loss": 0.3272,
      "step": 1120
    },
    {
      "epoch": 0.23329032258064517,
      "grad_norm": 0.6590626784400708,
      "learning_rate": 9.469181262039256e-07,
      "loss": 0.3283,
      "step": 1130
    },
    {
      "epoch": 0.2353548387096774,
      "grad_norm": 0.6597258358538342,
      "learning_rate": 9.45290339421638e-07,
      "loss": 0.3178,
      "step": 1140
    },
    {
      "epoch": 0.23741935483870968,
      "grad_norm": 0.6726623465784939,
      "learning_rate": 9.436394124618454e-07,
      "loss": 0.3322,
      "step": 1150
    },
    {
      "epoch": 0.23948387096774193,
      "grad_norm": 0.6601259805802864,
      "learning_rate": 9.419654311174329e-07,
      "loss": 0.3191,
      "step": 1160
    },
    {
      "epoch": 0.2415483870967742,
      "grad_norm": 0.6466405190915544,
      "learning_rate": 9.40268482379341e-07,
      "loss": 0.3283,
      "step": 1170
    },
    {
      "epoch": 0.24361290322580645,
      "grad_norm": 0.6411852843547549,
      "learning_rate": 9.385486544320451e-07,
      "loss": 0.3227,
      "step": 1180
    },
    {
      "epoch": 0.24567741935483872,
      "grad_norm": 0.6402118423685845,
      "learning_rate": 9.368060366489732e-07,
      "loss": 0.3178,
      "step": 1190
    },
    {
      "epoch": 0.24774193548387097,
      "grad_norm": 0.6524107490419644,
      "learning_rate": 9.350407195878607e-07,
      "loss": 0.3242,
      "step": 1200
    },
    {
      "epoch": 0.2498064516129032,
      "grad_norm": 0.6513244138766866,
      "learning_rate": 9.332527949860451e-07,
      "loss": 0.3281,
      "step": 1210
    },
    {
      "epoch": 0.25187096774193546,
      "grad_norm": 0.669880198956457,
      "learning_rate": 9.314423557556986e-07,
      "loss": 0.3248,
      "step": 1220
    },
    {
      "epoch": 0.25393548387096776,
      "grad_norm": 0.6356540042907762,
      "learning_rate": 9.296094959789994e-07,
      "loss": 0.3166,
      "step": 1230
    },
    {
      "epoch": 0.256,
      "grad_norm": 0.666974189133567,
      "learning_rate": 9.277543109032433e-07,
      "loss": 0.3224,
      "step": 1240
    },
    {
      "epoch": 0.25806451612903225,
      "grad_norm": 0.6513050299994531,
      "learning_rate": 9.258768969358933e-07,
      "loss": 0.3232,
      "step": 1250
    },
    {
      "epoch": 0.2601290322580645,
      "grad_norm": 0.6481734665370465,
      "learning_rate": 9.2397735163957e-07,
      "loss": 0.3175,
      "step": 1260
    },
    {
      "epoch": 0.2621935483870968,
      "grad_norm": 0.643020157091917,
      "learning_rate": 9.220557737269816e-07,
      "loss": 0.3218,
      "step": 1270
    },
    {
      "epoch": 0.26425806451612904,
      "grad_norm": 0.6426811969692795,
      "learning_rate": 9.201122630557943e-07,
      "loss": 0.3275,
      "step": 1280
    },
    {
      "epoch": 0.2663225806451613,
      "grad_norm": 0.6947621385955748,
      "learning_rate": 9.181469206234422e-07,
      "loss": 0.3285,
      "step": 1290
    },
    {
      "epoch": 0.26838709677419353,
      "grad_norm": 0.6312730128131891,
      "learning_rate": 9.161598485618803e-07,
      "loss": 0.3236,
      "step": 1300
    },
    {
      "epoch": 0.2704516129032258,
      "grad_norm": 0.6250577395835617,
      "learning_rate": 9.141511501322758e-07,
      "loss": 0.319,
      "step": 1310
    },
    {
      "epoch": 0.2725161290322581,
      "grad_norm": 0.6761809621959772,
      "learning_rate": 9.121209297196424e-07,
      "loss": 0.324,
      "step": 1320
    },
    {
      "epoch": 0.2745806451612903,
      "grad_norm": 0.6559064362731541,
      "learning_rate": 9.100692928274158e-07,
      "loss": 0.3235,
      "step": 1330
    },
    {
      "epoch": 0.27664516129032257,
      "grad_norm": 0.6470753912100222,
      "learning_rate": 9.079963460719714e-07,
      "loss": 0.3136,
      "step": 1340
    },
    {
      "epoch": 0.2787096774193548,
      "grad_norm": 0.67716550873512,
      "learning_rate": 9.059021971770828e-07,
      "loss": 0.323,
      "step": 1350
    },
    {
      "epoch": 0.2807741935483871,
      "grad_norm": 0.6526933363094527,
      "learning_rate": 9.037869549683252e-07,
      "loss": 0.3158,
      "step": 1360
    },
    {
      "epoch": 0.28283870967741936,
      "grad_norm": 0.616298302365804,
      "learning_rate": 9.016507293674187e-07,
      "loss": 0.3248,
      "step": 1370
    },
    {
      "epoch": 0.2849032258064516,
      "grad_norm": 0.6404102199659266,
      "learning_rate": 8.994936313865171e-07,
      "loss": 0.3271,
      "step": 1380
    },
    {
      "epoch": 0.28696774193548386,
      "grad_norm": 0.6412752672717336,
      "learning_rate": 8.973157731224385e-07,
      "loss": 0.3299,
      "step": 1390
    },
    {
      "epoch": 0.28903225806451616,
      "grad_norm": 0.6297868348665913,
      "learning_rate": 8.9511726775084e-07,
      "loss": 0.3225,
      "step": 1400
    },
    {
      "epoch": 0.2910967741935484,
      "grad_norm": 0.6325507536718176,
      "learning_rate": 8.928982295203367e-07,
      "loss": 0.319,
      "step": 1410
    },
    {
      "epoch": 0.29316129032258065,
      "grad_norm": 0.6472477044345786,
      "learning_rate": 8.906587737465642e-07,
      "loss": 0.3228,
      "step": 1420
    },
    {
      "epoch": 0.2952258064516129,
      "grad_norm": 0.6783964173187546,
      "learning_rate": 8.883990168061863e-07,
      "loss": 0.3211,
      "step": 1430
    },
    {
      "epoch": 0.29729032258064514,
      "grad_norm": 0.624197836950835,
      "learning_rate": 8.861190761308472e-07,
      "loss": 0.3211,
      "step": 1440
    },
    {
      "epoch": 0.29935483870967744,
      "grad_norm": 0.6717372599198446,
      "learning_rate": 8.838190702010693e-07,
      "loss": 0.3231,
      "step": 1450
    },
    {
      "epoch": 0.3014193548387097,
      "grad_norm": 0.6769990705574879,
      "learning_rate": 8.814991185400951e-07,
      "loss": 0.3208,
      "step": 1460
    },
    {
      "epoch": 0.30348387096774193,
      "grad_norm": 0.6659941754213017,
      "learning_rate": 8.791593417076781e-07,
      "loss": 0.3164,
      "step": 1470
    },
    {
      "epoch": 0.3055483870967742,
      "grad_norm": 0.6414833851246282,
      "learning_rate": 8.767998612938152e-07,
      "loss": 0.3206,
      "step": 1480
    },
    {
      "epoch": 0.3076129032258065,
      "grad_norm": 0.6514676034582378,
      "learning_rate": 8.744207999124301e-07,
      "loss": 0.3263,
      "step": 1490
    },
    {
      "epoch": 0.3096774193548387,
      "grad_norm": 0.6489872854216381,
      "learning_rate": 8.720222811950004e-07,
      "loss": 0.3154,
      "step": 1500
    },
    {
      "epoch": 0.31174193548387097,
      "grad_norm": 0.623717284587322,
      "learning_rate": 8.696044297841334e-07,
      "loss": 0.3176,
      "step": 1510
    },
    {
      "epoch": 0.3138064516129032,
      "grad_norm": 0.6504605677148272,
      "learning_rate": 8.671673713270886e-07,
      "loss": 0.3157,
      "step": 1520
    },
    {
      "epoch": 0.31587096774193546,
      "grad_norm": 0.6566805248169392,
      "learning_rate": 8.647112324692481e-07,
      "loss": 0.3188,
      "step": 1530
    },
    {
      "epoch": 0.31793548387096776,
      "grad_norm": 0.646097457626664,
      "learning_rate": 8.622361408475361e-07,
      "loss": 0.3203,
      "step": 1540
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.6545852284748576,
      "learning_rate": 8.597422250837848e-07,
      "loss": 0.3239,
      "step": 1550
    },
    {
      "epoch": 0.32206451612903225,
      "grad_norm": 0.6572912261816161,
      "learning_rate": 8.572296147780515e-07,
      "loss": 0.3161,
      "step": 1560
    },
    {
      "epoch": 0.3241290322580645,
      "grad_norm": 0.641031543823275,
      "learning_rate": 8.546984405018834e-07,
      "loss": 0.327,
      "step": 1570
    },
    {
      "epoch": 0.3261935483870968,
      "grad_norm": 0.646834473464689,
      "learning_rate": 8.521488337915318e-07,
      "loss": 0.328,
      "step": 1580
    },
    {
      "epoch": 0.32825806451612904,
      "grad_norm": 0.7030362083327168,
      "learning_rate": 8.49580927141117e-07,
      "loss": 0.3218,
      "step": 1590
    },
    {
      "epoch": 0.3303225806451613,
      "grad_norm": 0.6315828365581327,
      "learning_rate": 8.469948539957432e-07,
      "loss": 0.3249,
      "step": 1600
    },
    {
      "epoch": 0.33238709677419354,
      "grad_norm": 0.6718128979339714,
      "learning_rate": 8.443907487445634e-07,
      "loss": 0.316,
      "step": 1610
    },
    {
      "epoch": 0.3344516129032258,
      "grad_norm": 0.6429184826992661,
      "learning_rate": 8.417687467137964e-07,
      "loss": 0.3218,
      "step": 1620
    },
    {
      "epoch": 0.3365161290322581,
      "grad_norm": 0.6374025154785593,
      "learning_rate": 8.391289841596934e-07,
      "loss": 0.3158,
      "step": 1630
    },
    {
      "epoch": 0.33858064516129033,
      "grad_norm": 0.6412004629975684,
      "learning_rate": 8.364715982614577e-07,
      "loss": 0.3227,
      "step": 1640
    },
    {
      "epoch": 0.3406451612903226,
      "grad_norm": 0.6421068921819796,
      "learning_rate": 8.337967271141164e-07,
      "loss": 0.3217,
      "step": 1650
    },
    {
      "epoch": 0.3427096774193548,
      "grad_norm": 0.6387875404347384,
      "learning_rate": 8.311045097213431e-07,
      "loss": 0.3294,
      "step": 1660
    },
    {
      "epoch": 0.3447741935483871,
      "grad_norm": 0.6400040799598268,
      "learning_rate": 8.283950859882358e-07,
      "loss": 0.3289,
      "step": 1670
    },
    {
      "epoch": 0.34683870967741937,
      "grad_norm": 0.6544339151363505,
      "learning_rate": 8.256685967140449e-07,
      "loss": 0.3189,
      "step": 1680
    },
    {
      "epoch": 0.3489032258064516,
      "grad_norm": 0.6474316864675644,
      "learning_rate": 8.229251835848573e-07,
      "loss": 0.3198,
      "step": 1690
    },
    {
      "epoch": 0.35096774193548386,
      "grad_norm": 0.6331712614233053,
      "learning_rate": 8.201649891662334e-07,
      "loss": 0.3232,
      "step": 1700
    },
    {
      "epoch": 0.3530322580645161,
      "grad_norm": 0.6798002483244694,
      "learning_rate": 8.173881568957986e-07,
      "loss": 0.3204,
      "step": 1710
    },
    {
      "epoch": 0.3550967741935484,
      "grad_norm": 0.6397649167953239,
      "learning_rate": 8.145948310757886e-07,
      "loss": 0.3188,
      "step": 1720
    },
    {
      "epoch": 0.35716129032258065,
      "grad_norm": 0.6158268086730608,
      "learning_rate": 8.117851568655517e-07,
      "loss": 0.3207,
      "step": 1730
    },
    {
      "epoch": 0.3592258064516129,
      "grad_norm": 0.6344149660303308,
      "learning_rate": 8.089592802740039e-07,
      "loss": 0.3213,
      "step": 1740
    },
    {
      "epoch": 0.36129032258064514,
      "grad_norm": 0.6596708905977811,
      "learning_rate": 8.061173481520429e-07,
      "loss": 0.3169,
      "step": 1750
    },
    {
      "epoch": 0.36335483870967744,
      "grad_norm": 0.594571278861055,
      "learning_rate": 8.032595081849154e-07,
      "loss": 0.315,
      "step": 1760
    },
    {
      "epoch": 0.3654193548387097,
      "grad_norm": 0.6200795570147979,
      "learning_rate": 8.003859088845436e-07,
      "loss": 0.3175,
      "step": 1770
    },
    {
      "epoch": 0.36748387096774193,
      "grad_norm": 0.6357933185771896,
      "learning_rate": 7.974966995818066e-07,
      "loss": 0.3291,
      "step": 1780
    },
    {
      "epoch": 0.3695483870967742,
      "grad_norm": 0.6298425334317519,
      "learning_rate": 7.945920304187806e-07,
      "loss": 0.3202,
      "step": 1790
    },
    {
      "epoch": 0.3716129032258065,
      "grad_norm": 0.6197664857789437,
      "learning_rate": 7.916720523409366e-07,
      "loss": 0.3195,
      "step": 1800
    },
    {
      "epoch": 0.3736774193548387,
      "grad_norm": 0.6638430965232455,
      "learning_rate": 7.887369170892964e-07,
      "loss": 0.3242,
      "step": 1810
    },
    {
      "epoch": 0.37574193548387097,
      "grad_norm": 0.623255717934661,
      "learning_rate": 7.857867771925468e-07,
      "loss": 0.3171,
      "step": 1820
    },
    {
      "epoch": 0.3778064516129032,
      "grad_norm": 0.6203963007801919,
      "learning_rate": 7.828217859591133e-07,
      "loss": 0.3204,
      "step": 1830
    },
    {
      "epoch": 0.37987096774193546,
      "grad_norm": 0.662486034824777,
      "learning_rate": 7.798420974691935e-07,
      "loss": 0.3213,
      "step": 1840
    },
    {
      "epoch": 0.38193548387096776,
      "grad_norm": 0.6817838488299565,
      "learning_rate": 7.7684786656675e-07,
      "loss": 0.3213,
      "step": 1850
    },
    {
      "epoch": 0.384,
      "grad_norm": 0.616450638065689,
      "learning_rate": 7.738392488514627e-07,
      "loss": 0.3171,
      "step": 1860
    },
    {
      "epoch": 0.38606451612903225,
      "grad_norm": 0.643917095059045,
      "learning_rate": 7.70816400670645e-07,
      "loss": 0.3185,
      "step": 1870
    },
    {
      "epoch": 0.3881290322580645,
      "grad_norm": 0.6353966219397654,
      "learning_rate": 7.677794791111168e-07,
      "loss": 0.3177,
      "step": 1880
    },
    {
      "epoch": 0.3901935483870968,
      "grad_norm": 0.6691601627176933,
      "learning_rate": 7.647286419910426e-07,
      "loss": 0.3186,
      "step": 1890
    },
    {
      "epoch": 0.39225806451612905,
      "grad_norm": 0.6324600834717697,
      "learning_rate": 7.616640478517293e-07,
      "loss": 0.3191,
      "step": 1900
    },
    {
      "epoch": 0.3943225806451613,
      "grad_norm": 0.662059038474577,
      "learning_rate": 7.585858559493885e-07,
      "loss": 0.3164,
      "step": 1910
    },
    {
      "epoch": 0.39638709677419354,
      "grad_norm": 0.6444450125482107,
      "learning_rate": 7.554942262468593e-07,
      "loss": 0.3224,
      "step": 1920
    },
    {
      "epoch": 0.3984516129032258,
      "grad_norm": 0.6633456465202244,
      "learning_rate": 7.523893194052966e-07,
      "loss": 0.3147,
      "step": 1930
    },
    {
      "epoch": 0.4005161290322581,
      "grad_norm": 0.6585550440311129,
      "learning_rate": 7.49271296775821e-07,
      "loss": 0.3196,
      "step": 1940
    },
    {
      "epoch": 0.40258064516129033,
      "grad_norm": 0.628650403491387,
      "learning_rate": 7.461403203911355e-07,
      "loss": 0.3192,
      "step": 1950
    },
    {
      "epoch": 0.4046451612903226,
      "grad_norm": 0.6506895506578853,
      "learning_rate": 7.429965529571036e-07,
      "loss": 0.3141,
      "step": 1960
    },
    {
      "epoch": 0.4067096774193548,
      "grad_norm": 0.6227648165056641,
      "learning_rate": 7.398401578442952e-07,
      "loss": 0.3162,
      "step": 1970
    },
    {
      "epoch": 0.4087741935483871,
      "grad_norm": 0.6273632185846456,
      "learning_rate": 7.366712990794961e-07,
      "loss": 0.3258,
      "step": 1980
    },
    {
      "epoch": 0.41083870967741937,
      "grad_norm": 0.67490569636163,
      "learning_rate": 7.334901413371847e-07,
      "loss": 0.3235,
      "step": 1990
    },
    {
      "epoch": 0.4129032258064516,
      "grad_norm": 0.6305199896100085,
      "learning_rate": 7.302968499309737e-07,
      "loss": 0.3178,
      "step": 2000
    },
    {
      "epoch": 0.41496774193548386,
      "grad_norm": 0.6747577125957268,
      "learning_rate": 7.270915908050204e-07,
      "loss": 0.3166,
      "step": 2010
    },
    {
      "epoch": 0.4170322580645161,
      "grad_norm": 0.6369902232569418,
      "learning_rate": 7.238745305254019e-07,
      "loss": 0.3234,
      "step": 2020
    },
    {
      "epoch": 0.4190967741935484,
      "grad_norm": 0.6862442619995625,
      "learning_rate": 7.206458362714602e-07,
      "loss": 0.3082,
      "step": 2030
    },
    {
      "epoch": 0.42116129032258065,
      "grad_norm": 0.6185800663811354,
      "learning_rate": 7.17405675827114e-07,
      "loss": 0.3256,
      "step": 2040
    },
    {
      "epoch": 0.4232258064516129,
      "grad_norm": 0.6421361475096339,
      "learning_rate": 7.1415421757214e-07,
      "loss": 0.3178,
      "step": 2050
    },
    {
      "epoch": 0.42529032258064514,
      "grad_norm": 0.6103129194741619,
      "learning_rate": 7.10891630473422e-07,
      "loss": 0.319,
      "step": 2060
    },
    {
      "epoch": 0.42735483870967744,
      "grad_norm": 0.6122550249520009,
      "learning_rate": 7.076180840761714e-07,
      "loss": 0.3162,
      "step": 2070
    },
    {
      "epoch": 0.4294193548387097,
      "grad_norm": 0.6723489782935563,
      "learning_rate": 7.043337484951154e-07,
      "loss": 0.3179,
      "step": 2080
    },
    {
      "epoch": 0.43148387096774193,
      "grad_norm": 0.6553625507026416,
      "learning_rate": 7.010387944056576e-07,
      "loss": 0.3131,
      "step": 2090
    },
    {
      "epoch": 0.4335483870967742,
      "grad_norm": 0.6732724028213406,
      "learning_rate": 6.97733393035008e-07,
      "loss": 0.3166,
      "step": 2100
    },
    {
      "epoch": 0.4356129032258064,
      "grad_norm": 0.6323321242398868,
      "learning_rate": 6.944177161532851e-07,
      "loss": 0.3119,
      "step": 2110
    },
    {
      "epoch": 0.4376774193548387,
      "grad_norm": 0.6408466129224434,
      "learning_rate": 6.910919360645902e-07,
      "loss": 0.3248,
      "step": 2120
    },
    {
      "epoch": 0.43974193548387097,
      "grad_norm": 0.6457416165799166,
      "learning_rate": 6.877562255980519e-07,
      "loss": 0.3228,
      "step": 2130
    },
    {
      "epoch": 0.4418064516129032,
      "grad_norm": 0.6455332593011724,
      "learning_rate": 6.844107580988471e-07,
      "loss": 0.3202,
      "step": 2140
    },
    {
      "epoch": 0.44387096774193546,
      "grad_norm": 0.6353916997435629,
      "learning_rate": 6.810557074191899e-07,
      "loss": 0.3172,
      "step": 2150
    },
    {
      "epoch": 0.44593548387096776,
      "grad_norm": 0.6305782916264491,
      "learning_rate": 6.776912479093001e-07,
      "loss": 0.3141,
      "step": 2160
    },
    {
      "epoch": 0.448,
      "grad_norm": 0.6455331904620217,
      "learning_rate": 6.743175544083403e-07,
      "loss": 0.3168,
      "step": 2170
    },
    {
      "epoch": 0.45006451612903225,
      "grad_norm": 0.6405530821980946,
      "learning_rate": 6.709348022353318e-07,
      "loss": 0.3177,
      "step": 2180
    },
    {
      "epoch": 0.4521290322580645,
      "grad_norm": 0.6225103140072267,
      "learning_rate": 6.675431671800436e-07,
      "loss": 0.319,
      "step": 2190
    },
    {
      "epoch": 0.4541935483870968,
      "grad_norm": 0.6230005140918707,
      "learning_rate": 6.64142825493856e-07,
      "loss": 0.3169,
      "step": 2200
    },
    {
      "epoch": 0.45625806451612905,
      "grad_norm": 0.6364344705253122,
      "learning_rate": 6.607339538806034e-07,
      "loss": 0.3134,
      "step": 2210
    },
    {
      "epoch": 0.4583225806451613,
      "grad_norm": 0.6659225896183519,
      "learning_rate": 6.573167294873898e-07,
      "loss": 0.3072,
      "step": 2220
    },
    {
      "epoch": 0.46038709677419354,
      "grad_norm": 0.6259096717319926,
      "learning_rate": 6.538913298953845e-07,
      "loss": 0.318,
      "step": 2230
    },
    {
      "epoch": 0.4624516129032258,
      "grad_norm": 0.628880285691171,
      "learning_rate": 6.504579331105928e-07,
      "loss": 0.3197,
      "step": 2240
    },
    {
      "epoch": 0.4645161290322581,
      "grad_norm": 0.6321856597080695,
      "learning_rate": 6.470167175546063e-07,
      "loss": 0.3074,
      "step": 2250
    },
    {
      "epoch": 0.46658064516129033,
      "grad_norm": 0.6110714159723901,
      "learning_rate": 6.435678620553306e-07,
      "loss": 0.3155,
      "step": 2260
    },
    {
      "epoch": 0.4686451612903226,
      "grad_norm": 0.6321195923805359,
      "learning_rate": 6.401115458376924e-07,
      "loss": 0.3085,
      "step": 2270
    },
    {
      "epoch": 0.4707096774193548,
      "grad_norm": 0.6386785331141132,
      "learning_rate": 6.366479485143257e-07,
      "loss": 0.3186,
      "step": 2280
    },
    {
      "epoch": 0.4727741935483871,
      "grad_norm": 0.6043493485728505,
      "learning_rate": 6.331772500762382e-07,
      "loss": 0.3144,
      "step": 2290
    },
    {
      "epoch": 0.47483870967741937,
      "grad_norm": 0.6255984844555192,
      "learning_rate": 6.296996308834575e-07,
      "loss": 0.3159,
      "step": 2300
    },
    {
      "epoch": 0.4769032258064516,
      "grad_norm": 0.6195083185057502,
      "learning_rate": 6.262152716556586e-07,
      "loss": 0.317,
      "step": 2310
    },
    {
      "epoch": 0.47896774193548386,
      "grad_norm": 0.6300622425863397,
      "learning_rate": 6.227243534627724e-07,
      "loss": 0.3172,
      "step": 2320
    },
    {
      "epoch": 0.4810322580645161,
      "grad_norm": 0.6595039164876313,
      "learning_rate": 6.192270577155764e-07,
      "loss": 0.3211,
      "step": 2330
    },
    {
      "epoch": 0.4830967741935484,
      "grad_norm": 0.6366219906121937,
      "learning_rate": 6.157235661562672e-07,
      "loss": 0.314,
      "step": 2340
    },
    {
      "epoch": 0.48516129032258065,
      "grad_norm": 0.6527397855569752,
      "learning_rate": 6.122140608490157e-07,
      "loss": 0.3159,
      "step": 2350
    },
    {
      "epoch": 0.4872258064516129,
      "grad_norm": 0.6603354252170923,
      "learning_rate": 6.086987241705066e-07,
      "loss": 0.315,
      "step": 2360
    },
    {
      "epoch": 0.48929032258064514,
      "grad_norm": 0.5763376111349164,
      "learning_rate": 6.051777388004603e-07,
      "loss": 0.3198,
      "step": 2370
    },
    {
      "epoch": 0.49135483870967744,
      "grad_norm": 0.629217883816345,
      "learning_rate": 6.0165128771214e-07,
      "loss": 0.3199,
      "step": 2380
    },
    {
      "epoch": 0.4934193548387097,
      "grad_norm": 0.6946582376730316,
      "learning_rate": 5.981195541628431e-07,
      "loss": 0.3209,
      "step": 2390
    },
    {
      "epoch": 0.49548387096774194,
      "grad_norm": 0.6292866101195682,
      "learning_rate": 5.945827216843779e-07,
      "loss": 0.3171,
      "step": 2400
    },
    {
      "epoch": 0.4975483870967742,
      "grad_norm": 0.6472509856407982,
      "learning_rate": 5.910409740735263e-07,
      "loss": 0.3202,
      "step": 2410
    },
    {
      "epoch": 0.4996129032258064,
      "grad_norm": 0.6397136836791187,
      "learning_rate": 5.874944953824919e-07,
      "loss": 0.3104,
      "step": 2420
    },
    {
      "epoch": 0.5016774193548387,
      "grad_norm": 0.6660104783886344,
      "learning_rate": 5.83943469909337e-07,
      "loss": 0.322,
      "step": 2430
    },
    {
      "epoch": 0.5037419354838709,
      "grad_norm": 0.6286469573260441,
      "learning_rate": 5.803880821884032e-07,
      "loss": 0.3099,
      "step": 2440
    },
    {
      "epoch": 0.5058064516129033,
      "grad_norm": 0.6333782484084417,
      "learning_rate": 5.768285169807233e-07,
      "loss": 0.3149,
      "step": 2450
    },
    {
      "epoch": 0.5078709677419355,
      "grad_norm": 0.6229633085471118,
      "learning_rate": 5.73264959264419e-07,
      "loss": 0.3102,
      "step": 2460
    },
    {
      "epoch": 0.5099354838709678,
      "grad_norm": 0.6296999697703776,
      "learning_rate": 5.696975942250896e-07,
      "loss": 0.3137,
      "step": 2470
    },
    {
      "epoch": 0.512,
      "grad_norm": 0.6068961890441498,
      "learning_rate": 5.661266072461866e-07,
      "loss": 0.3149,
      "step": 2480
    },
    {
      "epoch": 0.5140645161290323,
      "grad_norm": 0.6225249709979962,
      "learning_rate": 5.625521838993814e-07,
      "loss": 0.3162,
      "step": 2490
    },
    {
      "epoch": 0.5161290322580645,
      "grad_norm": 0.6317655451714331,
      "learning_rate": 5.589745099349219e-07,
      "loss": 0.3178,
      "step": 2500
    },
    {
      "epoch": 0.5181935483870967,
      "grad_norm": 0.6636706396751287,
      "learning_rate": 5.55393771271978e-07,
      "loss": 0.3157,
      "step": 2510
    },
    {
      "epoch": 0.520258064516129,
      "grad_norm": 0.6680947882529774,
      "learning_rate": 5.518101539889828e-07,
      "loss": 0.3155,
      "step": 2520
    },
    {
      "epoch": 0.5223225806451612,
      "grad_norm": 0.6360722820260652,
      "learning_rate": 5.482238443139597e-07,
      "loss": 0.3175,
      "step": 2530
    },
    {
      "epoch": 0.5243870967741936,
      "grad_norm": 0.6313288013084732,
      "learning_rate": 5.446350286148471e-07,
      "loss": 0.3116,
      "step": 2540
    },
    {
      "epoch": 0.5264516129032258,
      "grad_norm": 0.6493083937647598,
      "learning_rate": 5.410438933898127e-07,
      "loss": 0.3128,
      "step": 2550
    },
    {
      "epoch": 0.5285161290322581,
      "grad_norm": 0.6041537743593429,
      "learning_rate": 5.374506252575612e-07,
      "loss": 0.3188,
      "step": 2560
    },
    {
      "epoch": 0.5305806451612903,
      "grad_norm": 0.6454388738856257,
      "learning_rate": 5.338554109476379e-07,
      "loss": 0.3204,
      "step": 2570
    },
    {
      "epoch": 0.5326451612903226,
      "grad_norm": 0.6440197687114312,
      "learning_rate": 5.302584372907236e-07,
      "loss": 0.3187,
      "step": 2580
    },
    {
      "epoch": 0.5347096774193548,
      "grad_norm": 0.6554355554054035,
      "learning_rate": 5.266598912089259e-07,
      "loss": 0.3186,
      "step": 2590
    },
    {
      "epoch": 0.5367741935483871,
      "grad_norm": 0.6155584050496433,
      "learning_rate": 5.230599597060666e-07,
      "loss": 0.3155,
      "step": 2600
    },
    {
      "epoch": 0.5388387096774193,
      "grad_norm": 0.6242689232949261,
      "learning_rate": 5.194588298579623e-07,
      "loss": 0.3166,
      "step": 2610
    },
    {
      "epoch": 0.5409032258064516,
      "grad_norm": 0.631216431453038,
      "learning_rate": 5.158566888027038e-07,
      "loss": 0.312,
      "step": 2620
    },
    {
      "epoch": 0.5429677419354839,
      "grad_norm": 0.6119968685734067,
      "learning_rate": 5.122537237309305e-07,
      "loss": 0.3124,
      "step": 2630
    },
    {
      "epoch": 0.5450322580645162,
      "grad_norm": 0.6172652792931007,
      "learning_rate": 5.086501218761033e-07,
      "loss": 0.3101,
      "step": 2640
    },
    {
      "epoch": 0.5470967741935484,
      "grad_norm": 0.6270520847073866,
      "learning_rate": 5.050460705047742e-07,
      "loss": 0.3091,
      "step": 2650
    },
    {
      "epoch": 0.5491612903225807,
      "grad_norm": 0.6576304505675945,
      "learning_rate": 5.014417569068554e-07,
      "loss": 0.3165,
      "step": 2660
    },
    {
      "epoch": 0.5512258064516129,
      "grad_norm": 0.6442263821147718,
      "learning_rate": 4.978373683858858e-07,
      "loss": 0.3171,
      "step": 2670
    },
    {
      "epoch": 0.5532903225806451,
      "grad_norm": 0.6108213268100124,
      "learning_rate": 4.942330922492975e-07,
      "loss": 0.3177,
      "step": 2680
    },
    {
      "epoch": 0.5553548387096774,
      "grad_norm": 0.615217814445509,
      "learning_rate": 4.906291157986834e-07,
      "loss": 0.3163,
      "step": 2690
    },
    {
      "epoch": 0.5574193548387096,
      "grad_norm": 0.6172453178635782,
      "learning_rate": 4.870256263200616e-07,
      "loss": 0.3121,
      "step": 2700
    },
    {
      "epoch": 0.5594838709677419,
      "grad_norm": 0.5996732217672011,
      "learning_rate": 4.834228110741447e-07,
      "loss": 0.3234,
      "step": 2710
    },
    {
      "epoch": 0.5615483870967742,
      "grad_norm": 0.6653609799045959,
      "learning_rate": 4.798208572866074e-07,
      "loss": 0.3193,
      "step": 2720
    },
    {
      "epoch": 0.5636129032258065,
      "grad_norm": 0.6395915777401148,
      "learning_rate": 4.762199521383575e-07,
      "loss": 0.3195,
      "step": 2730
    },
    {
      "epoch": 0.5656774193548387,
      "grad_norm": 0.6135325202391676,
      "learning_rate": 4.72620282755809e-07,
      "loss": 0.3175,
      "step": 2740
    },
    {
      "epoch": 0.567741935483871,
      "grad_norm": 0.6284093608378871,
      "learning_rate": 4.690220362011573e-07,
      "loss": 0.3145,
      "step": 2750
    },
    {
      "epoch": 0.5698064516129032,
      "grad_norm": 0.6157411851456424,
      "learning_rate": 4.6542539946265823e-07,
      "loss": 0.3135,
      "step": 2760
    },
    {
      "epoch": 0.5718709677419355,
      "grad_norm": 0.6599419221635924,
      "learning_rate": 4.618305594449113e-07,
      "loss": 0.3158,
      "step": 2770
    },
    {
      "epoch": 0.5739354838709677,
      "grad_norm": 0.6238143697793432,
      "learning_rate": 4.5823770295914706e-07,
      "loss": 0.3126,
      "step": 2780
    },
    {
      "epoch": 0.576,
      "grad_norm": 0.6312415400069866,
      "learning_rate": 4.5464701671351815e-07,
      "loss": 0.3176,
      "step": 2790
    },
    {
      "epoch": 0.5780645161290323,
      "grad_norm": 0.6170664554211615,
      "learning_rate": 4.5105868730339785e-07,
      "loss": 0.3125,
      "step": 2800
    },
    {
      "epoch": 0.5801290322580646,
      "grad_norm": 0.6352871069335873,
      "learning_rate": 4.4747290120168316e-07,
      "loss": 0.3147,
      "step": 2810
    },
    {
      "epoch": 0.5821935483870968,
      "grad_norm": 0.6205499456241415,
      "learning_rate": 4.438898447491036e-07,
      "loss": 0.3155,
      "step": 2820
    },
    {
      "epoch": 0.584258064516129,
      "grad_norm": 0.6546974655194456,
      "learning_rate": 4.403097041445395e-07,
      "loss": 0.316,
      "step": 2830
    },
    {
      "epoch": 0.5863225806451613,
      "grad_norm": 0.6331409057896509,
      "learning_rate": 4.367326654353436e-07,
      "loss": 0.3153,
      "step": 2840
    },
    {
      "epoch": 0.5883870967741935,
      "grad_norm": 0.6541263665072116,
      "learning_rate": 4.331589145076746e-07,
      "loss": 0.319,
      "step": 2850
    },
    {
      "epoch": 0.5904516129032258,
      "grad_norm": 0.6523571212271052,
      "learning_rate": 4.295886370768367e-07,
      "loss": 0.3196,
      "step": 2860
    },
    {
      "epoch": 0.592516129032258,
      "grad_norm": 0.6166147769430033,
      "learning_rate": 4.2602201867762875e-07,
      "loss": 0.3068,
      "step": 2870
    },
    {
      "epoch": 0.5945806451612903,
      "grad_norm": 0.6055985194256791,
      "learning_rate": 4.224592446547024e-07,
      "loss": 0.3106,
      "step": 2880
    },
    {
      "epoch": 0.5966451612903226,
      "grad_norm": 0.6227015026832325,
      "learning_rate": 4.1890050015293036e-07,
      "loss": 0.3178,
      "step": 2890
    },
    {
      "epoch": 0.5987096774193549,
      "grad_norm": 0.623960880562434,
      "learning_rate": 4.153459701077856e-07,
      "loss": 0.3159,
      "step": 2900
    },
    {
      "epoch": 0.6007741935483871,
      "grad_norm": 0.6756131349407722,
      "learning_rate": 4.117958392357303e-07,
      "loss": 0.3207,
      "step": 2910
    },
    {
      "epoch": 0.6028387096774194,
      "grad_norm": 0.6506140494701154,
      "learning_rate": 4.082502920246173e-07,
      "loss": 0.3096,
      "step": 2920
    },
    {
      "epoch": 0.6049032258064516,
      "grad_norm": 0.6535456937528376,
      "learning_rate": 4.047095127241026e-07,
      "loss": 0.3197,
      "step": 2930
    },
    {
      "epoch": 0.6069677419354839,
      "grad_norm": 0.6565191550520586,
      "learning_rate": 4.011736853360702e-07,
      "loss": 0.321,
      "step": 2940
    },
    {
      "epoch": 0.6090322580645161,
      "grad_norm": 0.59801785191147,
      "learning_rate": 3.976429936050709e-07,
      "loss": 0.3141,
      "step": 2950
    },
    {
      "epoch": 0.6110967741935484,
      "grad_norm": 0.628166782801977,
      "learning_rate": 3.941176210087737e-07,
      "loss": 0.3134,
      "step": 2960
    },
    {
      "epoch": 0.6131612903225806,
      "grad_norm": 0.6702663789481017,
      "learning_rate": 3.905977507484304e-07,
      "loss": 0.3168,
      "step": 2970
    },
    {
      "epoch": 0.615225806451613,
      "grad_norm": 0.6225856950934301,
      "learning_rate": 3.870835657393558e-07,
      "loss": 0.3209,
      "step": 2980
    },
    {
      "epoch": 0.6172903225806452,
      "grad_norm": 0.6505865990947035,
      "learning_rate": 3.8357524860142205e-07,
      "loss": 0.3152,
      "step": 2990
    },
    {
      "epoch": 0.6193548387096774,
      "grad_norm": 0.6074929984057879,
      "learning_rate": 3.8007298164956866e-07,
      "loss": 0.3203,
      "step": 3000
    },
    {
      "epoch": 0.6214193548387097,
      "grad_norm": 0.6272149481166996,
      "learning_rate": 3.765769468843284e-07,
      "loss": 0.3028,
      "step": 3010
    },
    {
      "epoch": 0.6234838709677419,
      "grad_norm": 0.61659281354388,
      "learning_rate": 3.7308732598236896e-07,
      "loss": 0.312,
      "step": 3020
    },
    {
      "epoch": 0.6255483870967742,
      "grad_norm": 0.6419204625089964,
      "learning_rate": 3.696043002870521e-07,
      "loss": 0.3116,
      "step": 3030
    },
    {
      "epoch": 0.6276129032258064,
      "grad_norm": 0.6441099222868198,
      "learning_rate": 3.6612805079900954e-07,
      "loss": 0.3147,
      "step": 3040
    },
    {
      "epoch": 0.6296774193548387,
      "grad_norm": 0.6426841231722075,
      "learning_rate": 3.6265875816673784e-07,
      "loss": 0.3146,
      "step": 3050
    },
    {
      "epoch": 0.6317419354838709,
      "grad_norm": 0.6358867549576873,
      "learning_rate": 3.591966026772096e-07,
      "loss": 0.3109,
      "step": 3060
    },
    {
      "epoch": 0.6338064516129033,
      "grad_norm": 0.6217305627910245,
      "learning_rate": 3.557417642465055e-07,
      "loss": 0.3161,
      "step": 3070
    },
    {
      "epoch": 0.6358709677419355,
      "grad_norm": 0.6172913230548381,
      "learning_rate": 3.522944224104643e-07,
      "loss": 0.3182,
      "step": 3080
    },
    {
      "epoch": 0.6379354838709678,
      "grad_norm": 0.6312545282019393,
      "learning_rate": 3.4885475631535253e-07,
      "loss": 0.3234,
      "step": 3090
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.6822374675039082,
      "learning_rate": 3.4542294470855626e-07,
      "loss": 0.3122,
      "step": 3100
    },
    {
      "epoch": 0.6420645161290323,
      "grad_norm": 0.6238435801987837,
      "learning_rate": 3.4199916592929055e-07,
      "loss": 0.3131,
      "step": 3110
    },
    {
      "epoch": 0.6441290322580645,
      "grad_norm": 0.6362639512522529,
      "learning_rate": 3.3858359789933313e-07,
      "loss": 0.3157,
      "step": 3120
    },
    {
      "epoch": 0.6461935483870968,
      "grad_norm": 0.6133147083090063,
      "learning_rate": 3.351764181137771e-07,
      "loss": 0.3176,
      "step": 3130
    },
    {
      "epoch": 0.648258064516129,
      "grad_norm": 0.6129689059138792,
      "learning_rate": 3.3177780363180923e-07,
      "loss": 0.3154,
      "step": 3140
    },
    {
      "epoch": 0.6503225806451612,
      "grad_norm": 0.6496128305000797,
      "learning_rate": 3.2838793106750625e-07,
      "loss": 0.3137,
      "step": 3150
    },
    {
      "epoch": 0.6523870967741936,
      "grad_norm": 0.6396028987013508,
      "learning_rate": 3.250069765806586e-07,
      "loss": 0.3141,
      "step": 3160
    },
    {
      "epoch": 0.6544516129032258,
      "grad_norm": 0.6530804143564158,
      "learning_rate": 3.2163511586761566e-07,
      "loss": 0.3157,
      "step": 3170
    },
    {
      "epoch": 0.6565161290322581,
      "grad_norm": 0.612114348974523,
      "learning_rate": 3.1827252415215454e-07,
      "loss": 0.3182,
      "step": 3180
    },
    {
      "epoch": 0.6585806451612903,
      "grad_norm": 0.6191950050314984,
      "learning_rate": 3.149193761763764e-07,
      "loss": 0.3131,
      "step": 3190
    },
    {
      "epoch": 0.6606451612903226,
      "grad_norm": 0.6335743458257347,
      "learning_rate": 3.115758461916229e-07,
      "loss": 0.3122,
      "step": 3200
    },
    {
      "epoch": 0.6627096774193548,
      "grad_norm": 0.6081224963128786,
      "learning_rate": 3.0824210794942333e-07,
      "loss": 0.3143,
      "step": 3210
    },
    {
      "epoch": 0.6647741935483871,
      "grad_norm": 0.6257725158537328,
      "learning_rate": 3.049183346924643e-07,
      "loss": 0.3154,
      "step": 3220
    },
    {
      "epoch": 0.6668387096774193,
      "grad_norm": 0.6652587310323215,
      "learning_rate": 3.016046991455865e-07,
      "loss": 0.3087,
      "step": 3230
    },
    {
      "epoch": 0.6689032258064516,
      "grad_norm": 0.6178973896870249,
      "learning_rate": 2.983013735068105e-07,
      "loss": 0.3179,
      "step": 3240
    },
    {
      "epoch": 0.6709677419354839,
      "grad_norm": 0.6224226055440618,
      "learning_rate": 2.9500852943838603e-07,
      "loss": 0.3091,
      "step": 3250
    },
    {
      "epoch": 0.6730322580645162,
      "grad_norm": 0.6736812063475199,
      "learning_rate": 2.9172633805787293e-07,
      "loss": 0.319,
      "step": 3260
    },
    {
      "epoch": 0.6750967741935484,
      "grad_norm": 0.625725567804461,
      "learning_rate": 2.88454969929248e-07,
      "loss": 0.3085,
      "step": 3270
    },
    {
      "epoch": 0.6771612903225807,
      "grad_norm": 0.6332936271071712,
      "learning_rate": 2.851945950540419e-07,
      "loss": 0.3163,
      "step": 3280
    },
    {
      "epoch": 0.6792258064516129,
      "grad_norm": 0.6468060996640993,
      "learning_rate": 2.819453828625038e-07,
      "loss": 0.3181,
      "step": 3290
    },
    {
      "epoch": 0.6812903225806451,
      "grad_norm": 0.6444594141879042,
      "learning_rate": 2.787075022047981e-07,
      "loss": 0.3166,
      "step": 3300
    },
    {
      "epoch": 0.6833548387096774,
      "grad_norm": 0.618461769009295,
      "learning_rate": 2.754811213422288e-07,
      "loss": 0.3173,
      "step": 3310
    },
    {
      "epoch": 0.6854193548387096,
      "grad_norm": 0.6041385090076459,
      "learning_rate": 2.722664079384954e-07,
      "loss": 0.3114,
      "step": 3320
    },
    {
      "epoch": 0.6874838709677419,
      "grad_norm": 0.6119130627331196,
      "learning_rate": 2.6906352905098116e-07,
      "loss": 0.314,
      "step": 3330
    },
    {
      "epoch": 0.6895483870967742,
      "grad_norm": 0.6480908972345034,
      "learning_rate": 2.658726511220712e-07,
      "loss": 0.3199,
      "step": 3340
    },
    {
      "epoch": 0.6916129032258065,
      "grad_norm": 0.6353300382456004,
      "learning_rate": 2.6269393997050205e-07,
      "loss": 0.3147,
      "step": 3350
    },
    {
      "epoch": 0.6936774193548387,
      "grad_norm": 0.6192059686376392,
      "learning_rate": 2.595275607827457e-07,
      "loss": 0.3129,
      "step": 3360
    },
    {
      "epoch": 0.695741935483871,
      "grad_norm": 0.6089960449636171,
      "learning_rate": 2.563736781044262e-07,
      "loss": 0.3092,
      "step": 3370
    },
    {
      "epoch": 0.6978064516129032,
      "grad_norm": 0.6266402041271057,
      "learning_rate": 2.5323245583176667e-07,
      "loss": 0.313,
      "step": 3380
    },
    {
      "epoch": 0.6998709677419355,
      "grad_norm": 0.5940796344475668,
      "learning_rate": 2.5010405720307355e-07,
      "loss": 0.3119,
      "step": 3390
    },
    {
      "epoch": 0.7019354838709677,
      "grad_norm": 0.6353842389907374,
      "learning_rate": 2.469886447902541e-07,
      "loss": 0.3152,
      "step": 3400
    },
    {
      "epoch": 0.704,
      "grad_norm": 0.6271834587115767,
      "learning_rate": 2.438863804903666e-07,
      "loss": 0.3159,
      "step": 3410
    },
    {
      "epoch": 0.7060645161290322,
      "grad_norm": 0.6335854782909629,
      "learning_rate": 2.4079742551720887e-07,
      "loss": 0.3103,
      "step": 3420
    },
    {
      "epoch": 0.7081290322580646,
      "grad_norm": 0.6645732860198468,
      "learning_rate": 2.3772194039293896e-07,
      "loss": 0.3158,
      "step": 3430
    },
    {
      "epoch": 0.7101935483870968,
      "grad_norm": 0.6288245840777608,
      "learning_rate": 2.3466008493973477e-07,
      "loss": 0.3148,
      "step": 3440
    },
    {
      "epoch": 0.712258064516129,
      "grad_norm": 0.6387607278411881,
      "learning_rate": 2.3161201827148725e-07,
      "loss": 0.3202,
      "step": 3450
    },
    {
      "epoch": 0.7143225806451613,
      "grad_norm": 0.6231283283369418,
      "learning_rate": 2.2857789878553309e-07,
      "loss": 0.3166,
      "step": 3460
    },
    {
      "epoch": 0.7163870967741935,
      "grad_norm": 0.5952558586464339,
      "learning_rate": 2.2555788415442288e-07,
      "loss": 0.3119,
      "step": 3470
    },
    {
      "epoch": 0.7184516129032258,
      "grad_norm": 0.6660428513391485,
      "learning_rate": 2.22552131317727e-07,
      "loss": 0.3133,
      "step": 3480
    },
    {
      "epoch": 0.720516129032258,
      "grad_norm": 0.6378788981156114,
      "learning_rate": 2.1956079647388025e-07,
      "loss": 0.3057,
      "step": 3490
    },
    {
      "epoch": 0.7225806451612903,
      "grad_norm": 0.6333103274834864,
      "learning_rate": 2.165840350720655e-07,
      "loss": 0.3172,
      "step": 3500
    },
    {
      "epoch": 0.7246451612903225,
      "grad_norm": 0.6111454843884092,
      "learning_rate": 2.1362200180413481e-07,
      "loss": 0.3135,
      "step": 3510
    },
    {
      "epoch": 0.7267096774193549,
      "grad_norm": 0.6205327964139117,
      "learning_rate": 2.1067485059657032e-07,
      "loss": 0.3185,
      "step": 3520
    },
    {
      "epoch": 0.7287741935483871,
      "grad_norm": 0.6699284667407549,
      "learning_rate": 2.0774273460248577e-07,
      "loss": 0.3204,
      "step": 3530
    },
    {
      "epoch": 0.7308387096774194,
      "grad_norm": 0.6179110137346477,
      "learning_rate": 2.0482580619366796e-07,
      "loss": 0.3112,
      "step": 3540
    },
    {
      "epoch": 0.7329032258064516,
      "grad_norm": 0.6074917560720133,
      "learning_rate": 2.019242169526581e-07,
      "loss": 0.3139,
      "step": 3550
    },
    {
      "epoch": 0.7349677419354839,
      "grad_norm": 0.6310306529351029,
      "learning_rate": 1.9903811766487426e-07,
      "loss": 0.3119,
      "step": 3560
    },
    {
      "epoch": 0.7370322580645161,
      "grad_norm": 0.664083184766849,
      "learning_rate": 1.9616765831077603e-07,
      "loss": 0.3216,
      "step": 3570
    },
    {
      "epoch": 0.7390967741935484,
      "grad_norm": 0.6364296679991407,
      "learning_rate": 1.9331298805807095e-07,
      "loss": 0.3201,
      "step": 3580
    },
    {
      "epoch": 0.7411612903225806,
      "grad_norm": 0.6286717380476362,
      "learning_rate": 1.9047425525396161e-07,
      "loss": 0.3112,
      "step": 3590
    },
    {
      "epoch": 0.743225806451613,
      "grad_norm": 0.6346887009435065,
      "learning_rate": 1.876516074174379e-07,
      "loss": 0.3106,
      "step": 3600
    },
    {
      "epoch": 0.7452903225806452,
      "grad_norm": 0.6085474387411121,
      "learning_rate": 1.848451912316103e-07,
      "loss": 0.3152,
      "step": 3610
    },
    {
      "epoch": 0.7473548387096774,
      "grad_norm": 0.6250309615399917,
      "learning_rate": 1.8205515253608688e-07,
      "loss": 0.3118,
      "step": 3620
    },
    {
      "epoch": 0.7494193548387097,
      "grad_norm": 0.6329400757803024,
      "learning_rate": 1.792816363193952e-07,
      "loss": 0.316,
      "step": 3630
    },
    {
      "epoch": 0.7514838709677419,
      "grad_norm": 0.6580450917336371,
      "learning_rate": 1.7652478671144755e-07,
      "loss": 0.3131,
      "step": 3640
    },
    {
      "epoch": 0.7535483870967742,
      "grad_norm": 0.6388637187577192,
      "learning_rate": 1.7378474697605128e-07,
      "loss": 0.3177,
      "step": 3650
    },
    {
      "epoch": 0.7556129032258064,
      "grad_norm": 0.6143703605590203,
      "learning_rate": 1.7106165950346318e-07,
      "loss": 0.3191,
      "step": 3660
    },
    {
      "epoch": 0.7576774193548387,
      "grad_norm": 0.6273664608270475,
      "learning_rate": 1.683556658029903e-07,
      "loss": 0.3125,
      "step": 3670
    },
    {
      "epoch": 0.7597419354838709,
      "grad_norm": 0.6586710436884331,
      "learning_rate": 1.656669064956368e-07,
      "loss": 0.3105,
      "step": 3680
    },
    {
      "epoch": 0.7618064516129033,
      "grad_norm": 0.6461005253964683,
      "learning_rate": 1.6299552130679578e-07,
      "loss": 0.3175,
      "step": 3690
    },
    {
      "epoch": 0.7638709677419355,
      "grad_norm": 0.6473704793930669,
      "learning_rate": 1.6034164905898768e-07,
      "loss": 0.3158,
      "step": 3700
    },
    {
      "epoch": 0.7659354838709678,
      "grad_norm": 0.6887318889016925,
      "learning_rate": 1.5770542766464773e-07,
      "loss": 0.3158,
      "step": 3710
    },
    {
      "epoch": 0.768,
      "grad_norm": 0.6145074649843777,
      "learning_rate": 1.5508699411895738e-07,
      "loss": 0.3189,
      "step": 3720
    },
    {
      "epoch": 0.7700645161290323,
      "grad_norm": 0.6117699994497662,
      "learning_rate": 1.524864844927266e-07,
      "loss": 0.3192,
      "step": 3730
    },
    {
      "epoch": 0.7721290322580645,
      "grad_norm": 0.6442807746229446,
      "learning_rate": 1.4990403392532159e-07,
      "loss": 0.3091,
      "step": 3740
    },
    {
      "epoch": 0.7741935483870968,
      "grad_norm": 0.6232325765752856,
      "learning_rate": 1.473397766176431e-07,
      "loss": 0.3174,
      "step": 3750
    },
    {
      "epoch": 0.776258064516129,
      "grad_norm": 0.6579855922539816,
      "learning_rate": 1.4479384582515153e-07,
      "loss": 0.3183,
      "step": 3760
    },
    {
      "epoch": 0.7783225806451612,
      "grad_norm": 0.6400471085265736,
      "learning_rate": 1.4226637385094247e-07,
      "loss": 0.315,
      "step": 3770
    },
    {
      "epoch": 0.7803870967741936,
      "grad_norm": 0.6339992409795476,
      "learning_rate": 1.3975749203887228e-07,
      "loss": 0.318,
      "step": 3780
    },
    {
      "epoch": 0.7824516129032258,
      "grad_norm": 0.6614920296989713,
      "learning_rate": 1.3726733076673085e-07,
      "loss": 0.3155,
      "step": 3790
    },
    {
      "epoch": 0.7845161290322581,
      "grad_norm": 0.6149620463556447,
      "learning_rate": 1.3479601943946761e-07,
      "loss": 0.3093,
      "step": 3800
    },
    {
      "epoch": 0.7865806451612903,
      "grad_norm": 0.642745944566185,
      "learning_rate": 1.323436864824664e-07,
      "loss": 0.3171,
      "step": 3810
    },
    {
      "epoch": 0.7886451612903226,
      "grad_norm": 0.6152220466779936,
      "learning_rate": 1.299104593348721e-07,
      "loss": 0.3207,
      "step": 3820
    },
    {
      "epoch": 0.7907096774193548,
      "grad_norm": 0.6429579966758033,
      "learning_rate": 1.2749646444296703e-07,
      "loss": 0.316,
      "step": 3830
    },
    {
      "epoch": 0.7927741935483871,
      "grad_norm": 0.6338893602645223,
      "learning_rate": 1.2510182725360086e-07,
      "loss": 0.3077,
      "step": 3840
    },
    {
      "epoch": 0.7948387096774193,
      "grad_norm": 0.6092406969812145,
      "learning_rate": 1.2272667220767158e-07,
      "loss": 0.32,
      "step": 3850
    },
    {
      "epoch": 0.7969032258064516,
      "grad_norm": 0.609726533631423,
      "learning_rate": 1.2037112273365818e-07,
      "loss": 0.3127,
      "step": 3860
    },
    {
      "epoch": 0.7989677419354839,
      "grad_norm": 0.6419258740410181,
      "learning_rate": 1.1803530124120714e-07,
      "loss": 0.3141,
      "step": 3870
    },
    {
      "epoch": 0.8010322580645162,
      "grad_norm": 0.623356524162269,
      "learning_rate": 1.157193291147705e-07,
      "loss": 0.3133,
      "step": 3880
    },
    {
      "epoch": 0.8030967741935484,
      "grad_norm": 0.6205282254055249,
      "learning_rate": 1.1342332670729882e-07,
      "loss": 0.3099,
      "step": 3890
    },
    {
      "epoch": 0.8051612903225807,
      "grad_norm": 0.6371775065392936,
      "learning_rate": 1.1114741333398592e-07,
      "loss": 0.3152,
      "step": 3900
    },
    {
      "epoch": 0.8072258064516129,
      "grad_norm": 0.6100575734172149,
      "learning_rate": 1.0889170726606933e-07,
      "loss": 0.3106,
      "step": 3910
    },
    {
      "epoch": 0.8092903225806451,
      "grad_norm": 0.632559732600231,
      "learning_rate": 1.066563257246838e-07,
      "loss": 0.3088,
      "step": 3920
    },
    {
      "epoch": 0.8113548387096774,
      "grad_norm": 0.6181373045584015,
      "learning_rate": 1.0444138487476944e-07,
      "loss": 0.3138,
      "step": 3930
    },
    {
      "epoch": 0.8134193548387096,
      "grad_norm": 0.6259151100943298,
      "learning_rate": 1.0224699981903517e-07,
      "loss": 0.3177,
      "step": 3940
    },
    {
      "epoch": 0.8154838709677419,
      "grad_norm": 0.6117914186807475,
      "learning_rate": 1.0007328459197778e-07,
      "loss": 0.3162,
      "step": 3950
    },
    {
      "epoch": 0.8175483870967742,
      "grad_norm": 0.6387066615699482,
      "learning_rate": 9.792035215395556e-08,
      "loss": 0.3156,
      "step": 3960
    },
    {
      "epoch": 0.8196129032258065,
      "grad_norm": 0.6289156638043008,
      "learning_rate": 9.578831438531776e-08,
      "loss": 0.3216,
      "step": 3970
    },
    {
      "epoch": 0.8216774193548387,
      "grad_norm": 0.6494192268070564,
      "learning_rate": 9.3677282080591e-08,
      "loss": 0.3155,
      "step": 3980
    },
    {
      "epoch": 0.823741935483871,
      "grad_norm": 0.6152684331534928,
      "learning_rate": 9.158736494272179e-08,
      "loss": 0.3142,
      "step": 3990
    },
    {
      "epoch": 0.8258064516129032,
      "grad_norm": 0.6252970551532324,
      "learning_rate": 8.951867157737558e-08,
      "loss": 0.3206,
      "step": 4000
    },
    {
      "epoch": 0.8278709677419355,
      "grad_norm": 0.6124835286826313,
      "learning_rate": 8.747130948729226e-08,
      "loss": 0.3118,
      "step": 4010
    },
    {
      "epoch": 0.8299354838709677,
      "grad_norm": 0.6416414008426811,
      "learning_rate": 8.544538506670074e-08,
      "loss": 0.3101,
      "step": 4020
    },
    {
      "epoch": 0.832,
      "grad_norm": 0.6334705227262246,
      "learning_rate": 8.344100359578904e-08,
      "loss": 0.311,
      "step": 4030
    },
    {
      "epoch": 0.8340645161290322,
      "grad_norm": 0.603175358881232,
      "learning_rate": 8.145826923523358e-08,
      "loss": 0.3051,
      "step": 4040
    },
    {
      "epoch": 0.8361290322580646,
      "grad_norm": 0.6298247075384997,
      "learning_rate": 7.949728502078668e-08,
      "loss": 0.3103,
      "step": 4050
    },
    {
      "epoch": 0.8381935483870968,
      "grad_norm": 0.6579316907441995,
      "learning_rate": 7.755815285792172e-08,
      "loss": 0.3227,
      "step": 4060
    },
    {
      "epoch": 0.840258064516129,
      "grad_norm": 0.6066666948849365,
      "learning_rate": 7.564097351653742e-08,
      "loss": 0.3249,
      "step": 4070
    },
    {
      "epoch": 0.8423225806451613,
      "grad_norm": 0.6180551794332899,
      "learning_rate": 7.374584662572142e-08,
      "loss": 0.319,
      "step": 4080
    },
    {
      "epoch": 0.8443870967741935,
      "grad_norm": 0.5883783638702929,
      "learning_rate": 7.187287066857289e-08,
      "loss": 0.3085,
      "step": 4090
    },
    {
      "epoch": 0.8464516129032258,
      "grad_norm": 0.6473801259147952,
      "learning_rate": 7.002214297708481e-08,
      "loss": 0.3087,
      "step": 4100
    },
    {
      "epoch": 0.848516129032258,
      "grad_norm": 0.6148667467318439,
      "learning_rate": 6.819375972708536e-08,
      "loss": 0.3092,
      "step": 4110
    },
    {
      "epoch": 0.8505806451612903,
      "grad_norm": 0.6348250486724527,
      "learning_rate": 6.6387815933241e-08,
      "loss": 0.3143,
      "step": 4120
    },
    {
      "epoch": 0.8526451612903225,
      "grad_norm": 0.6315802775439991,
      "learning_rate": 6.460440544411777e-08,
      "loss": 0.3143,
      "step": 4130
    },
    {
      "epoch": 0.8547096774193549,
      "grad_norm": 0.6076346417593953,
      "learning_rate": 6.284362093730545e-08,
      "loss": 0.3105,
      "step": 4140
    },
    {
      "epoch": 0.8567741935483871,
      "grad_norm": 0.5962317858372677,
      "learning_rate": 6.110555391460026e-08,
      "loss": 0.3152,
      "step": 4150
    },
    {
      "epoch": 0.8588387096774194,
      "grad_norm": 0.5927251015186508,
      "learning_rate": 5.9390294697251045e-08,
      "loss": 0.3108,
      "step": 4160
    },
    {
      "epoch": 0.8609032258064516,
      "grad_norm": 0.6066991888644118,
      "learning_rate": 5.7697932421264415e-08,
      "loss": 0.3037,
      "step": 4170
    },
    {
      "epoch": 0.8629677419354839,
      "grad_norm": 0.6003227795214134,
      "learning_rate": 5.602855503277376e-08,
      "loss": 0.3126,
      "step": 4180
    },
    {
      "epoch": 0.8650322580645161,
      "grad_norm": 0.60428884775973,
      "learning_rate": 5.438224928346791e-08,
      "loss": 0.3138,
      "step": 4190
    },
    {
      "epoch": 0.8670967741935484,
      "grad_norm": 0.62060666748443,
      "learning_rate": 5.275910072608408e-08,
      "loss": 0.3187,
      "step": 4200
    },
    {
      "epoch": 0.8691612903225806,
      "grad_norm": 0.6290563694560117,
      "learning_rate": 5.115919370996097e-08,
      "loss": 0.3181,
      "step": 4210
    },
    {
      "epoch": 0.8712258064516128,
      "grad_norm": 0.5932518096375957,
      "learning_rate": 4.9582611376655924e-08,
      "loss": 0.3231,
      "step": 4220
    },
    {
      "epoch": 0.8732903225806452,
      "grad_norm": 0.6352198420824648,
      "learning_rate": 4.8029435655624785e-08,
      "loss": 0.309,
      "step": 4230
    },
    {
      "epoch": 0.8753548387096775,
      "grad_norm": 0.6188377103818619,
      "learning_rate": 4.6499747259963254e-08,
      "loss": 0.3138,
      "step": 4240
    },
    {
      "epoch": 0.8774193548387097,
      "grad_norm": 0.6387211687173309,
      "learning_rate": 4.499362568221327e-08,
      "loss": 0.3072,
      "step": 4250
    },
    {
      "epoch": 0.8794838709677419,
      "grad_norm": 0.6469413697906533,
      "learning_rate": 4.351114919023197e-08,
      "loss": 0.3174,
      "step": 4260
    },
    {
      "epoch": 0.8815483870967742,
      "grad_norm": 0.609278472353098,
      "learning_rate": 4.205239482312445e-08,
      "loss": 0.3127,
      "step": 4270
    },
    {
      "epoch": 0.8836129032258064,
      "grad_norm": 0.6106614475011715,
      "learning_rate": 4.0617438387239746e-08,
      "loss": 0.3169,
      "step": 4280
    },
    {
      "epoch": 0.8856774193548387,
      "grad_norm": 0.5880264333058545,
      "learning_rate": 3.9206354452232135e-08,
      "loss": 0.3103,
      "step": 4290
    },
    {
      "epoch": 0.8877419354838709,
      "grad_norm": 0.6345936372283331,
      "learning_rate": 3.7819216347185715e-08,
      "loss": 0.3161,
      "step": 4300
    },
    {
      "epoch": 0.8898064516129032,
      "grad_norm": 0.6267523098231387,
      "learning_rate": 3.6456096156803616e-08,
      "loss": 0.3112,
      "step": 4310
    },
    {
      "epoch": 0.8918709677419355,
      "grad_norm": 0.5986450915786209,
      "learning_rate": 3.5117064717662406e-08,
      "loss": 0.3096,
      "step": 4320
    },
    {
      "epoch": 0.8939354838709678,
      "grad_norm": 0.620818430567561,
      "learning_rate": 3.380219161453063e-08,
      "loss": 0.3191,
      "step": 4330
    },
    {
      "epoch": 0.896,
      "grad_norm": 0.6536206656078076,
      "learning_rate": 3.251154517675264e-08,
      "loss": 0.3132,
      "step": 4340
    },
    {
      "epoch": 0.8980645161290323,
      "grad_norm": 0.6620246651441947,
      "learning_rate": 3.124519247469814e-08,
      "loss": 0.3135,
      "step": 4350
    },
    {
      "epoch": 0.9001290322580645,
      "grad_norm": 0.6432332552038761,
      "learning_rate": 3.0003199316276486e-08,
      "loss": 0.3086,
      "step": 4360
    },
    {
      "epoch": 0.9021935483870968,
      "grad_norm": 0.6157763583488753,
      "learning_rate": 2.8785630243517156e-08,
      "loss": 0.3074,
      "step": 4370
    },
    {
      "epoch": 0.904258064516129,
      "grad_norm": 0.6365850518212122,
      "learning_rate": 2.759254852921522e-08,
      "loss": 0.3137,
      "step": 4380
    },
    {
      "epoch": 0.9063225806451612,
      "grad_norm": 0.6315485223419045,
      "learning_rate": 2.6424016173643816e-08,
      "loss": 0.3132,
      "step": 4390
    },
    {
      "epoch": 0.9083870967741936,
      "grad_norm": 0.6373095133174064,
      "learning_rate": 2.5280093901331957e-08,
      "loss": 0.3072,
      "step": 4400
    },
    {
      "epoch": 0.9104516129032258,
      "grad_norm": 0.6017307890815197,
      "learning_rate": 2.4160841157908894e-08,
      "loss": 0.3162,
      "step": 4410
    },
    {
      "epoch": 0.9125161290322581,
      "grad_norm": 0.6465531889379581,
      "learning_rate": 2.3066316107014984e-08,
      "loss": 0.3212,
      "step": 4420
    },
    {
      "epoch": 0.9145806451612903,
      "grad_norm": 0.6029856939651695,
      "learning_rate": 2.199657562727919e-08,
      "loss": 0.3171,
      "step": 4430
    },
    {
      "epoch": 0.9166451612903226,
      "grad_norm": 0.644409396257745,
      "learning_rate": 2.0951675309363038e-08,
      "loss": 0.3187,
      "step": 4440
    },
    {
      "epoch": 0.9187096774193548,
      "grad_norm": 0.6526376809487803,
      "learning_rate": 1.9931669453072064e-08,
      "loss": 0.3228,
      "step": 4450
    },
    {
      "epoch": 0.9207741935483871,
      "grad_norm": 0.6345017488836956,
      "learning_rate": 1.893661106453387e-08,
      "loss": 0.3179,
      "step": 4460
    },
    {
      "epoch": 0.9228387096774193,
      "grad_norm": 0.6421680406292284,
      "learning_rate": 1.7966551853443813e-08,
      "loss": 0.311,
      "step": 4470
    },
    {
      "epoch": 0.9249032258064516,
      "grad_norm": 0.6535141196988801,
      "learning_rate": 1.7021542230377495e-08,
      "loss": 0.3117,
      "step": 4480
    },
    {
      "epoch": 0.9269677419354839,
      "grad_norm": 0.6516997261163685,
      "learning_rate": 1.610163130417119e-08,
      "loss": 0.3164,
      "step": 4490
    },
    {
      "epoch": 0.9290322580645162,
      "grad_norm": 0.6139278478128292,
      "learning_rate": 1.520686687937006e-08,
      "loss": 0.3119,
      "step": 4500
    },
    {
      "epoch": 0.9310967741935484,
      "grad_norm": 0.6280704101718017,
      "learning_rate": 1.4337295453743848e-08,
      "loss": 0.3146,
      "step": 4510
    },
    {
      "epoch": 0.9331612903225807,
      "grad_norm": 0.6177747730934665,
      "learning_rate": 1.3492962215870208e-08,
      "loss": 0.3151,
      "step": 4520
    },
    {
      "epoch": 0.9352258064516129,
      "grad_norm": 0.6311802153022037,
      "learning_rate": 1.2673911042786812e-08,
      "loss": 0.3147,
      "step": 4530
    },
    {
      "epoch": 0.9372903225806452,
      "grad_norm": 0.6298140193496882,
      "learning_rate": 1.188018449771111e-08,
      "loss": 0.3114,
      "step": 4540
    },
    {
      "epoch": 0.9393548387096774,
      "grad_norm": 0.6033792966514346,
      "learning_rate": 1.1111823827828438e-08,
      "loss": 0.3158,
      "step": 4550
    },
    {
      "epoch": 0.9414193548387096,
      "grad_norm": 0.5961880904551417,
      "learning_rate": 1.0368868962148446e-08,
      "loss": 0.3117,
      "step": 4560
    },
    {
      "epoch": 0.9434838709677419,
      "grad_norm": 0.6557939191645362,
      "learning_rate": 9.651358509430385e-09,
      "loss": 0.3194,
      "step": 4570
    },
    {
      "epoch": 0.9455483870967742,
      "grad_norm": 0.628397602387223,
      "learning_rate": 8.959329756176359e-09,
      "loss": 0.3122,
      "step": 4580
    },
    {
      "epoch": 0.9476129032258065,
      "grad_norm": 0.6294158507698498,
      "learning_rate": 8.292818664694223e-09,
      "loss": 0.3097,
      "step": 4590
    },
    {
      "epoch": 0.9496774193548387,
      "grad_norm": 0.6329054623458709,
      "learning_rate": 7.651859871228072e-09,
      "loss": 0.3171,
      "step": 4600
    },
    {
      "epoch": 0.951741935483871,
      "grad_norm": 0.6192025568076179,
      "learning_rate": 7.0364866841589045e-09,
      "loss": 0.3169,
      "step": 4610
    },
    {
      "epoch": 0.9538064516129032,
      "grad_norm": 0.6510490608752881,
      "learning_rate": 6.446731082273449e-09,
      "loss": 0.3152,
      "step": 4620
    },
    {
      "epoch": 0.9558709677419355,
      "grad_norm": 0.6509717448947699,
      "learning_rate": 5.8826237131022196e-09,
      "loss": 0.3161,
      "step": 4630
    },
    {
      "epoch": 0.9579354838709677,
      "grad_norm": 0.6267847738737716,
      "learning_rate": 5.344193891327286e-09,
      "loss": 0.3135,
      "step": 4640
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.6072626235316539,
      "learning_rate": 4.831469597258331e-09,
      "loss": 0.3115,
      "step": 4650
    },
    {
      "epoch": 0.9620645161290322,
      "grad_norm": 0.62913760686652,
      "learning_rate": 4.344477475379027e-09,
      "loss": 0.3158,
      "step": 4660
    },
    {
      "epoch": 0.9641290322580646,
      "grad_norm": 0.6237025476138797,
      "learning_rate": 3.883242832962319e-09,
      "loss": 0.3185,
      "step": 4670
    },
    {
      "epoch": 0.9661935483870968,
      "grad_norm": 0.6386110886092738,
      "learning_rate": 3.4477896387552497e-09,
      "loss": 0.3158,
      "step": 4680
    },
    {
      "epoch": 0.9682580645161291,
      "grad_norm": 0.6128452175027437,
      "learning_rate": 3.0381405217333457e-09,
      "loss": 0.3117,
      "step": 4690
    },
    {
      "epoch": 0.9703225806451613,
      "grad_norm": 0.6373113791110392,
      "learning_rate": 2.6543167699247248e-09,
      "loss": 0.3191,
      "step": 4700
    },
    {
      "epoch": 0.9723870967741935,
      "grad_norm": 0.6687919646494422,
      "learning_rate": 2.2963383293039264e-09,
      "loss": 0.3213,
      "step": 4710
    },
    {
      "epoch": 0.9744516129032258,
      "grad_norm": 0.6488120079489716,
      "learning_rate": 1.964223802755238e-09,
      "loss": 0.3175,
      "step": 4720
    },
    {
      "epoch": 0.976516129032258,
      "grad_norm": 0.637435741828532,
      "learning_rate": 1.6579904491059726e-09,
      "loss": 0.3084,
      "step": 4730
    },
    {
      "epoch": 0.9785806451612903,
      "grad_norm": 0.5999360130685788,
      "learning_rate": 1.3776541822297926e-09,
      "loss": 0.3189,
      "step": 4740
    },
    {
      "epoch": 0.9806451612903225,
      "grad_norm": 0.619868587243893,
      "learning_rate": 1.1232295702193751e-09,
      "loss": 0.3165,
      "step": 4750
    },
    {
      "epoch": 0.9827096774193549,
      "grad_norm": 0.6601674936337429,
      "learning_rate": 8.947298346296816e-10,
      "loss": 0.3232,
      "step": 4760
    },
    {
      "epoch": 0.9847741935483871,
      "grad_norm": 0.6254003815023664,
      "learning_rate": 6.921668497907873e-10,
      "loss": 0.3146,
      "step": 4770
    },
    {
      "epoch": 0.9868387096774194,
      "grad_norm": 0.6500214007015074,
      "learning_rate": 5.155511421906511e-10,
      "loss": 0.3148,
      "step": 4780
    },
    {
      "epoch": 0.9889032258064516,
      "grad_norm": 0.6187202682538294,
      "learning_rate": 3.64891889928276e-10,
      "loss": 0.3151,
      "step": 4790
    },
    {
      "epoch": 0.9909677419354839,
      "grad_norm": 0.6323548620934831,
      "learning_rate": 2.4019692223675727e-10,
      "loss": 0.3114,
      "step": 4800
    },
    {
      "epoch": 0.9930322580645161,
      "grad_norm": 0.6355942016945179,
      "learning_rate": 1.4147271907621883e-10,
      "loss": 0.3128,
      "step": 4810
    },
    {
      "epoch": 0.9950967741935484,
      "grad_norm": 0.6569052859865055,
      "learning_rate": 6.87244107974716e-11,
      "loss": 0.312,
      "step": 4820
    },
    {
      "epoch": 0.9971612903225806,
      "grad_norm": 0.6459570341315254,
      "learning_rate": 2.1955777874838045e-11,
      "loss": 0.317,
      "step": 4830
    },
    {
      "epoch": 0.9992258064516129,
      "grad_norm": 0.6173828325733423,
      "learning_rate": 1.1692507103089334e-12,
      "loss": 0.3087,
      "step": 4840
    },
    {
      "epoch": 0.9998451612903225,
      "step": 4843,
      "total_flos": 409696561004544.0,
      "train_loss": 0.32205133897914096,
      "train_runtime": 161287.2429,
      "train_samples_per_second": 7.688,
      "train_steps_per_second": 0.03
    }
  ],
  "logging_steps": 10,
  "max_steps": 4843,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 1000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 409696561004544.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}