{
  "best_metric": 0.3201138377189636,
  "best_model_checkpoint": "saved_model/lop_dec2024/checkpoint-5422",
  "epoch": 0.9999538936788234,
  "eval_steps": 500,
  "global_step": 5422,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "grad_norm": NaN,
      "learning_rate": 0.0,
      "loss": 72.616,
      "step": 1
    },
    {
      "epoch": 0.0,
      "grad_norm": 15.978808403015137,
      "learning_rate": 3e-06,
      "loss": 72.0656,
      "step": 10
    },
    {
      "epoch": 0.0,
      "grad_norm": 16.57687759399414,
      "learning_rate": 7.000000000000001e-06,
      "loss": 71.4774,
      "step": 20
    },
    {
      "epoch": 0.01,
      "grad_norm": 17.607620239257812,
      "learning_rate": 1.2e-05,
      "loss": 70.0161,
      "step": 30
    },
    {
      "epoch": 0.01,
      "grad_norm": 20.96339988708496,
      "learning_rate": 1.7000000000000003e-05,
      "loss": 66.9798,
      "step": 40
    },
    {
      "epoch": 0.01,
      "grad_norm": 28.76253890991211,
      "learning_rate": 2.15e-05,
      "loss": 61.5471,
      "step": 50
    },
    {
      "epoch": 0.01,
      "grad_norm": 34.237281799316406,
      "learning_rate": 2.6500000000000004e-05,
      "loss": 50.4159,
      "step": 60
    },
    {
      "epoch": 0.01,
      "grad_norm": 37.09980773925781,
      "learning_rate": 3.15e-05,
      "loss": 30.1299,
      "step": 70
    },
    {
      "epoch": 0.01,
      "grad_norm": 21.59116554260254,
      "learning_rate": 3.65e-05,
      "loss": 10.0018,
      "step": 80
    },
    {
      "epoch": 0.02,
      "grad_norm": 4.791250705718994,
      "learning_rate": 4.15e-05,
      "loss": 1.4075,
      "step": 90
    },
    {
      "epoch": 0.02,
      "grad_norm": 6.963352203369141,
      "learning_rate": 4.6500000000000005e-05,
      "loss": 0.6076,
      "step": 100
    },
    {
      "epoch": 0.02,
      "grad_norm": 8.645480155944824,
      "learning_rate": 5.1500000000000005e-05,
      "loss": 0.6529,
      "step": 110
    },
    {
      "epoch": 0.02,
      "grad_norm": 10.286620140075684,
      "learning_rate": 5.65e-05,
      "loss": 0.606,
      "step": 120
    },
    {
      "epoch": 0.02,
      "grad_norm": 4.937899589538574,
      "learning_rate": 6.15e-05,
      "loss": 0.6646,
      "step": 130
    },
    {
      "epoch": 0.03,
      "grad_norm": 5.505192279815674,
      "learning_rate": 6.65e-05,
      "loss": 0.5808,
      "step": 140
    },
    {
      "epoch": 0.03,
      "grad_norm": 2.3285677433013916,
      "learning_rate": 7.15e-05,
      "loss": 0.4994,
      "step": 150
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.4964683055877686,
      "learning_rate": 7.65e-05,
      "loss": 0.4659,
      "step": 160
    },
    {
      "epoch": 0.03,
      "grad_norm": 6.808905124664307,
      "learning_rate": 8.15e-05,
      "loss": 0.4923,
      "step": 170
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.631983757019043,
      "learning_rate": 8.65e-05,
      "loss": 0.4065,
      "step": 180
    },
    {
      "epoch": 0.04,
      "grad_norm": 4.93166971206665,
      "learning_rate": 9.15e-05,
      "loss": 0.4412,
      "step": 190
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.9249387979507446,
      "learning_rate": 9.65e-05,
      "loss": 0.3778,
      "step": 200
    },
    {
      "epoch": 0.04,
      "grad_norm": 3.0214664936065674,
      "learning_rate": 9.999444650129582e-05,
      "loss": 0.3697,
      "step": 210
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.8178681135177612,
      "learning_rate": 9.997593483894854e-05,
      "loss": 0.3376,
      "step": 220
    },
    {
      "epoch": 0.04,
      "grad_norm": 3.6914970874786377,
      "learning_rate": 9.995742317660126e-05,
      "loss": 0.3389,
      "step": 230
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.8259241580963135,
      "learning_rate": 9.993891151425399e-05,
      "loss": 0.3186,
      "step": 240
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.6828951835632324,
      "learning_rate": 9.99203998519067e-05,
      "loss": 0.3228,
      "step": 250
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.0691156387329102,
      "learning_rate": 9.990188818955942e-05,
      "loss": 0.3243,
      "step": 260
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.3478268384933472,
      "learning_rate": 9.988337652721215e-05,
      "loss": 0.3061,
      "step": 270
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.275899887084961,
      "learning_rate": 9.986486486486487e-05,
      "loss": 0.297,
      "step": 280
    },
    {
      "epoch": 0.05,
      "grad_norm": 3.873141050338745,
      "learning_rate": 9.98463532025176e-05,
      "loss": 0.3115,
      "step": 290
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.8778353929519653,
      "learning_rate": 9.98278415401703e-05,
      "loss": 0.3045,
      "step": 300
    },
    {
      "epoch": 0.06,
      "grad_norm": 2.3832242488861084,
      "learning_rate": 9.980932987782303e-05,
      "loss": 0.3058,
      "step": 310
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.6602181196212769,
      "learning_rate": 9.979081821547575e-05,
      "loss": 0.2714,
      "step": 320
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.796381950378418,
      "learning_rate": 9.977230655312848e-05,
      "loss": 0.2911,
      "step": 330
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.6985359191894531,
      "learning_rate": 9.97537948907812e-05,
      "loss": 0.2827,
      "step": 340
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.9519887566566467,
      "learning_rate": 9.973528322843392e-05,
      "loss": 0.2828,
      "step": 350
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.771391749382019,
      "learning_rate": 9.971677156608665e-05,
      "loss": 0.274,
      "step": 360
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.5956106185913086,
      "learning_rate": 9.969825990373937e-05,
      "loss": 0.2509,
      "step": 370
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.8278146982192993,
      "learning_rate": 9.967974824139208e-05,
      "loss": 0.2644,
      "step": 380
    },
    {
      "epoch": 0.07,
      "grad_norm": 3.159097671508789,
      "learning_rate": 9.96612365790448e-05,
      "loss": 0.2855,
      "step": 390
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.152933120727539,
      "learning_rate": 9.964272491669753e-05,
      "loss": 0.2755,
      "step": 400
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.9050334692001343,
      "learning_rate": 9.962421325435023e-05,
      "loss": 0.2837,
      "step": 410
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.3858613967895508,
      "learning_rate": 9.960570159200297e-05,
      "loss": 0.2584,
      "step": 420
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.629534125328064,
      "learning_rate": 9.958718992965568e-05,
      "loss": 0.2432,
      "step": 430
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.7582938075065613,
      "learning_rate": 9.956867826730841e-05,
      "loss": 0.242,
      "step": 440
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.5532256364822388,
      "learning_rate": 9.955016660496113e-05,
      "loss": 0.2828,
      "step": 450
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.8100008964538574,
      "learning_rate": 9.953165494261385e-05,
      "loss": 0.2391,
      "step": 460
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.3843342065811157,
      "learning_rate": 9.951314328026658e-05,
      "loss": 0.2513,
      "step": 470
    },
    {
      "epoch": 0.09,
      "grad_norm": 2.4204509258270264,
      "learning_rate": 9.94946316179193e-05,
      "loss": 0.2378,
      "step": 480
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.645175576210022,
      "learning_rate": 9.947611995557201e-05,
      "loss": 0.2704,
      "step": 490
    },
    {
      "epoch": 0.09,
      "grad_norm": 2.546315908432007,
      "learning_rate": 9.945760829322473e-05,
      "loss": 0.2762,
      "step": 500
    },
    {
      "epoch": 0.09,
      "grad_norm": 3.5709307193756104,
      "learning_rate": 9.943909663087746e-05,
      "loss": 0.2753,
      "step": 510
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.6997337341308594,
      "learning_rate": 9.942058496853018e-05,
      "loss": 0.2637,
      "step": 520
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.6340448260307312,
      "learning_rate": 9.94020733061829e-05,
      "loss": 0.2447,
      "step": 530
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.1261272430419922,
      "learning_rate": 9.938356164383563e-05,
      "loss": 0.2345,
      "step": 540
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.7460525631904602,
      "learning_rate": 9.936504998148834e-05,
      "loss": 0.2356,
      "step": 550
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.7596022486686707,
      "learning_rate": 9.934653831914106e-05,
      "loss": 0.2497,
      "step": 560
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.7767930030822754,
      "learning_rate": 9.932802665679378e-05,
      "loss": 0.2332,
      "step": 570
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.36362847685813904,
      "learning_rate": 9.930951499444651e-05,
      "loss": 0.234,
      "step": 580
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.1732509136199951,
      "learning_rate": 9.929100333209923e-05,
      "loss": 0.2281,
      "step": 590
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.9395415782928467,
      "learning_rate": 9.927249166975194e-05,
      "loss": 0.2262,
      "step": 600
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.7397685050964355,
      "learning_rate": 9.925398000740466e-05,
      "loss": 0.2375,
      "step": 610
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.398878812789917,
      "learning_rate": 9.923546834505739e-05,
      "loss": 0.2408,
      "step": 620
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.213331937789917,
      "learning_rate": 9.921695668271012e-05,
      "loss": 0.2402,
      "step": 630
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.7745940685272217,
      "learning_rate": 9.919844502036283e-05,
      "loss": 0.259,
      "step": 640
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.9329665303230286,
      "learning_rate": 9.917993335801556e-05,
      "loss": 0.2457,
      "step": 650
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.9290269613265991,
      "learning_rate": 9.916142169566827e-05,
      "loss": 0.2236,
      "step": 660
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.089816689491272,
      "learning_rate": 9.9142910033321e-05,
      "loss": 0.2437,
      "step": 670
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.6095532178878784,
      "learning_rate": 9.912439837097371e-05,
      "loss": 0.2295,
      "step": 680
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.6958633065223694,
      "learning_rate": 9.910588670862644e-05,
      "loss": 0.231,
      "step": 690
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.5149015784263611,
      "learning_rate": 9.908737504627916e-05,
      "loss": 0.2567,
      "step": 700
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.9667665362358093,
      "learning_rate": 9.906886338393189e-05,
      "loss": 0.2324,
      "step": 710
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.8621126413345337,
      "learning_rate": 9.90503517215846e-05,
      "loss": 0.2349,
      "step": 720
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.7839102149009705,
      "learning_rate": 9.903184005923732e-05,
      "loss": 0.2395,
      "step": 730
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.9927177429199219,
      "learning_rate": 9.901332839689005e-05,
      "loss": 0.2253,
      "step": 740
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.7064800262451172,
      "learning_rate": 9.899481673454276e-05,
      "loss": 0.2178,
      "step": 750
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.4743956327438354,
      "learning_rate": 9.897630507219549e-05,
      "loss": 0.2186,
      "step": 760
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.73115873336792,
      "learning_rate": 9.89577934098482e-05,
      "loss": 0.2406,
      "step": 770
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.5700548887252808,
      "learning_rate": 9.893928174750094e-05,
      "loss": 0.2368,
      "step": 780
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.7140907049179077,
      "learning_rate": 9.892077008515365e-05,
      "loss": 0.2303,
      "step": 790
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.316851019859314,
      "learning_rate": 9.890225842280637e-05,
      "loss": 0.2271,
      "step": 800
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.3242087364196777,
      "learning_rate": 9.88837467604591e-05,
      "loss": 0.2213,
      "step": 810
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.3230711817741394,
      "learning_rate": 9.886523509811182e-05,
      "loss": 0.2366,
      "step": 820
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.613010287284851,
      "learning_rate": 9.884672343576454e-05,
      "loss": 0.2172,
      "step": 830
    },
    {
      "epoch": 0.15,
      "grad_norm": 2.3217711448669434,
      "learning_rate": 9.882821177341725e-05,
      "loss": 0.2401,
      "step": 840
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.38820984959602356,
      "learning_rate": 9.880970011106998e-05,
      "loss": 0.2376,
      "step": 850
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.5649935007095337,
      "learning_rate": 9.87911884487227e-05,
      "loss": 0.2335,
      "step": 860
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.5969343781471252,
      "learning_rate": 9.877267678637542e-05,
      "loss": 0.228,
      "step": 870
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.4941283464431763,
      "learning_rate": 9.875416512402814e-05,
      "loss": 0.2388,
      "step": 880
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.7500633597373962,
      "learning_rate": 9.873565346168087e-05,
      "loss": 0.2203,
      "step": 890
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.6874450445175171,
      "learning_rate": 9.871714179933358e-05,
      "loss": 0.2397,
      "step": 900
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.4555752277374268,
      "learning_rate": 9.86986301369863e-05,
      "loss": 0.2418,
      "step": 910
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.051475167274475,
      "learning_rate": 9.868011847463903e-05,
      "loss": 0.2411,
      "step": 920
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.377054214477539,
      "learning_rate": 9.866160681229175e-05,
      "loss": 0.2312,
      "step": 930
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.4420924186706543,
      "learning_rate": 9.864309514994447e-05,
      "loss": 0.2179,
      "step": 940
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.9395197629928589,
      "learning_rate": 9.862458348759718e-05,
      "loss": 0.2286,
      "step": 950
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.31522631645202637,
      "learning_rate": 9.860607182524991e-05,
      "loss": 0.2267,
      "step": 960
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.4355124831199646,
      "learning_rate": 9.858756016290263e-05,
      "loss": 0.228,
      "step": 970
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.5540971159934998,
      "learning_rate": 9.856904850055535e-05,
      "loss": 0.2202,
      "step": 980
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.43327707052230835,
      "learning_rate": 9.855053683820808e-05,
      "loss": 0.2316,
      "step": 990
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.765080451965332,
      "learning_rate": 9.85320251758608e-05,
      "loss": 0.2302,
      "step": 1000
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.822860598564148,
      "learning_rate": 9.851351351351353e-05,
      "loss": 0.2213,
      "step": 1010
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.31268832087516785,
      "learning_rate": 9.849500185116623e-05,
      "loss": 0.2065,
      "step": 1020
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.3371509313583374,
      "learning_rate": 9.847649018881896e-05,
      "loss": 0.2261,
      "step": 1030
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.3734799325466156,
      "learning_rate": 9.845797852647168e-05,
      "loss": 0.2228,
      "step": 1040
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.45795735716819763,
      "learning_rate": 9.843946686412441e-05,
      "loss": 0.223,
      "step": 1050
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.4650135040283203,
      "learning_rate": 9.842095520177711e-05,
      "loss": 0.2301,
      "step": 1060
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.4706295728683472,
      "learning_rate": 9.840244353942985e-05,
      "loss": 0.2222,
      "step": 1070
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.3891691565513611,
      "learning_rate": 9.838393187708256e-05,
      "loss": 0.2103,
      "step": 1080
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.0710134506225586,
      "learning_rate": 9.836542021473529e-05,
      "loss": 0.2242,
      "step": 1090
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.9032573699951172,
      "learning_rate": 9.834690855238801e-05,
      "loss": 0.2124,
      "step": 1100
    },
    {
      "epoch": 0.2,
      "grad_norm": 2.119032859802246,
      "learning_rate": 9.832839689004073e-05,
      "loss": 0.2289,
      "step": 1110
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.8338161110877991,
      "learning_rate": 9.830988522769346e-05,
      "loss": 0.2167,
      "step": 1120
    },
    {
      "epoch": 0.21,
      "grad_norm": 2.242032527923584,
      "learning_rate": 9.829137356534616e-05,
      "loss": 0.2225,
      "step": 1130
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.4238918125629425,
      "learning_rate": 9.827286190299889e-05,
      "loss": 0.2205,
      "step": 1140
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.5318244099617004,
      "learning_rate": 9.825435024065161e-05,
      "loss": 0.222,
      "step": 1150
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.75505131483078,
      "learning_rate": 9.823583857830434e-05,
      "loss": 0.2311,
      "step": 1160
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.4188501834869385,
      "learning_rate": 9.821732691595706e-05,
      "loss": 0.2149,
      "step": 1170
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.6075870990753174,
      "learning_rate": 9.819881525360978e-05,
      "loss": 0.2233,
      "step": 1180
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.4305758476257324,
      "learning_rate": 9.81803035912625e-05,
      "loss": 0.2238,
      "step": 1190
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.5145848989486694,
      "learning_rate": 9.816179192891522e-05,
      "loss": 0.2217,
      "step": 1200
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.7991466522216797,
      "learning_rate": 9.814328026656794e-05,
      "loss": 0.2118,
      "step": 1210
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.0274105072021484,
      "learning_rate": 9.812476860422066e-05,
      "loss": 0.2178,
      "step": 1220
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.4420272409915924,
      "learning_rate": 9.810625694187339e-05,
      "loss": 0.219,
      "step": 1230
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.5142613053321838,
      "learning_rate": 9.80877452795261e-05,
      "loss": 0.2123,
      "step": 1240
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.32571473717689514,
      "learning_rate": 9.806923361717882e-05,
      "loss": 0.2115,
      "step": 1250
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.3837771713733673,
      "learning_rate": 9.805072195483155e-05,
      "loss": 0.1992,
      "step": 1260
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.0336960554122925,
      "learning_rate": 9.803221029248427e-05,
      "loss": 0.2109,
      "step": 1270
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.2124392986297607,
      "learning_rate": 9.8013698630137e-05,
      "loss": 0.2422,
      "step": 1280
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.8734978437423706,
      "learning_rate": 9.79951869677897e-05,
      "loss": 0.2264,
      "step": 1290
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.44908058643341064,
      "learning_rate": 9.797667530544244e-05,
      "loss": 0.2123,
      "step": 1300
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.4297716021537781,
      "learning_rate": 9.795816364309515e-05,
      "loss": 0.2249,
      "step": 1310
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.1651443243026733,
      "learning_rate": 9.793965198074787e-05,
      "loss": 0.2222,
      "step": 1320
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.3527076542377472,
      "learning_rate": 9.792114031840059e-05,
      "loss": 0.2206,
      "step": 1330
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.29916948080062866,
      "learning_rate": 9.790262865605332e-05,
      "loss": 0.2209,
      "step": 1340
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.24127215147018433,
      "learning_rate": 9.788411699370604e-05,
      "loss": 0.2167,
      "step": 1350
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.4984634220600128,
      "learning_rate": 9.786560533135875e-05,
      "loss": 0.2493,
      "step": 1360
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.36658474802970886,
      "learning_rate": 9.784709366901149e-05,
      "loss": 0.2203,
      "step": 1370
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.2130411863327026,
      "learning_rate": 9.78285820066642e-05,
      "loss": 0.2142,
      "step": 1380
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.1353245973587036,
      "learning_rate": 9.781007034431693e-05,
      "loss": 0.2157,
      "step": 1390
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.41700485348701477,
      "learning_rate": 9.779155868196964e-05,
      "loss": 0.2141,
      "step": 1400
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.722000002861023,
      "learning_rate": 9.777304701962237e-05,
      "loss": 0.2137,
      "step": 1410
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.7909780144691467,
      "learning_rate": 9.775453535727508e-05,
      "loss": 0.2136,
      "step": 1420
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.2209943532943726,
      "learning_rate": 9.773602369492782e-05,
      "loss": 0.2075,
      "step": 1430
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.9682492613792419,
      "learning_rate": 9.771751203258053e-05,
      "loss": 0.2206,
      "step": 1440
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.36788633465766907,
      "learning_rate": 9.769900037023325e-05,
      "loss": 0.2204,
      "step": 1450
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.2191195487976074,
      "learning_rate": 9.768048870788598e-05,
      "loss": 0.2103,
      "step": 1460
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.6930559277534485,
      "learning_rate": 9.76619770455387e-05,
      "loss": 0.2157,
      "step": 1470
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.2961112260818481,
      "learning_rate": 9.764346538319142e-05,
      "loss": 0.2244,
      "step": 1480
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.7222216725349426,
      "learning_rate": 9.762495372084413e-05,
      "loss": 0.2161,
      "step": 1490
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.9623996615409851,
      "learning_rate": 9.760644205849686e-05,
      "loss": 0.212,
      "step": 1500
    },
    {
      "epoch": 0.28,
      "grad_norm": 3.655571937561035,
      "learning_rate": 9.758793039614957e-05,
      "loss": 0.2302,
      "step": 1510
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.7312277555465698,
      "learning_rate": 9.75694187338023e-05,
      "loss": 0.22,
      "step": 1520
    },
    {
      "epoch": 0.28,
      "grad_norm": 2.0226645469665527,
      "learning_rate": 9.755090707145502e-05,
      "loss": 0.2269,
      "step": 1530
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.3837519884109497,
      "learning_rate": 9.753239540910775e-05,
      "loss": 0.2106,
      "step": 1540
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.6145792007446289,
      "learning_rate": 9.751388374676046e-05,
      "loss": 0.2204,
      "step": 1550
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.6149736642837524,
      "learning_rate": 9.749537208441318e-05,
      "loss": 0.218,
      "step": 1560
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.113775610923767,
      "learning_rate": 9.747686042206591e-05,
      "loss": 0.2194,
      "step": 1570
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.2693588137626648,
      "learning_rate": 9.745834875971863e-05,
      "loss": 0.2089,
      "step": 1580
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.5227116346359253,
      "learning_rate": 9.743983709737135e-05,
      "loss": 0.2092,
      "step": 1590
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.43677178025245667,
      "learning_rate": 9.742132543502406e-05,
      "loss": 0.2061,
      "step": 1600
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.8750597834587097,
      "learning_rate": 9.74028137726768e-05,
      "loss": 0.2137,
      "step": 1610
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.8018006086349487,
      "learning_rate": 9.738430211032951e-05,
      "loss": 0.2194,
      "step": 1620
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.531554639339447,
      "learning_rate": 9.736579044798223e-05,
      "loss": 0.2077,
      "step": 1630
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.5439698696136475,
      "learning_rate": 9.734727878563496e-05,
      "loss": 0.2153,
      "step": 1640
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.296250343322754,
      "learning_rate": 9.732876712328768e-05,
      "loss": 0.1968,
      "step": 1650
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.4914539158344269,
      "learning_rate": 9.73102554609404e-05,
      "loss": 0.2164,
      "step": 1660
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.35670381784439087,
      "learning_rate": 9.729174379859311e-05,
      "loss": 0.2132,
      "step": 1670
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.6645960807800293,
      "learning_rate": 9.727323213624584e-05,
      "loss": 0.2233,
      "step": 1680
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.7455287575721741,
      "learning_rate": 9.725472047389856e-05,
      "loss": 0.2023,
      "step": 1690
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.36129245162010193,
      "learning_rate": 9.723620881155128e-05,
      "loss": 0.2219,
      "step": 1700
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.36295002698898315,
      "learning_rate": 9.721769714920401e-05,
      "loss": 0.2138,
      "step": 1710
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.34044042229652405,
      "learning_rate": 9.719918548685672e-05,
      "loss": 0.2122,
      "step": 1720
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.8785526156425476,
      "learning_rate": 9.718067382450946e-05,
      "loss": 0.2096,
      "step": 1730
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.3848158121109009,
      "learning_rate": 9.716216216216216e-05,
      "loss": 0.2172,
      "step": 1740
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.8510532975196838,
      "learning_rate": 9.714365049981489e-05,
      "loss": 0.2094,
      "step": 1750
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.8068777322769165,
      "learning_rate": 9.712513883746761e-05,
      "loss": 0.2156,
      "step": 1760
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.7489349246025085,
      "learning_rate": 9.710662717512034e-05,
      "loss": 0.2066,
      "step": 1770
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.5769368410110474,
      "learning_rate": 9.708811551277304e-05,
      "loss": 0.2133,
      "step": 1780
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.6159785985946655,
      "learning_rate": 9.706960385042577e-05,
      "loss": 0.2083,
      "step": 1790
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.0204073190689087,
      "learning_rate": 9.705109218807849e-05,
      "loss": 0.213,
      "step": 1800
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.4422318637371063,
      "learning_rate": 9.703258052573122e-05,
      "loss": 0.1996,
      "step": 1810
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.9518413543701172,
      "learning_rate": 9.701406886338394e-05,
      "loss": 0.218,
      "step": 1820
    },
    {
      "epoch": 0.34,
      "grad_norm": 2.2408201694488525,
      "learning_rate": 9.699555720103666e-05,
      "loss": 0.2189,
      "step": 1830
    },
    {
      "epoch": 0.34,
      "grad_norm": 2.9319238662719727,
      "learning_rate": 9.697704553868939e-05,
      "loss": 0.2144,
      "step": 1840
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.5640377998352051,
      "learning_rate": 9.695853387634209e-05,
      "loss": 0.2184,
      "step": 1850
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.26673319935798645,
      "learning_rate": 9.694002221399482e-05,
      "loss": 0.2161,
      "step": 1860
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.6571794152259827,
      "learning_rate": 9.692151055164754e-05,
      "loss": 0.207,
      "step": 1870
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.20037458837032318,
      "learning_rate": 9.690299888930027e-05,
      "loss": 0.1985,
      "step": 1880
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.9381949305534363,
      "learning_rate": 9.688448722695299e-05,
      "loss": 0.2118,
      "step": 1890
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.4142616987228394,
      "learning_rate": 9.68659755646057e-05,
      "loss": 0.2155,
      "step": 1900
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.7180021405220032,
      "learning_rate": 9.684746390225843e-05,
      "loss": 0.2024,
      "step": 1910
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.6328707933425903,
      "learning_rate": 9.682895223991115e-05,
      "loss": 0.2107,
      "step": 1920
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.0013030767440796,
      "learning_rate": 9.681044057756387e-05,
      "loss": 0.1973,
      "step": 1930
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.9628821611404419,
      "learning_rate": 9.679192891521659e-05,
      "loss": 0.2074,
      "step": 1940
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.5976811647415161,
      "learning_rate": 9.677341725286932e-05,
      "loss": 0.2076,
      "step": 1950
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.5646719932556152,
      "learning_rate": 9.675490559052203e-05,
      "loss": 0.2291,
      "step": 1960
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.701019287109375,
      "learning_rate": 9.673639392817475e-05,
      "loss": 0.2108,
      "step": 1970
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.8562067747116089,
      "learning_rate": 9.671788226582747e-05,
      "loss": 0.2072,
      "step": 1980
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.2539293766021729,
      "learning_rate": 9.66993706034802e-05,
      "loss": 0.215,
      "step": 1990
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.9341269731521606,
      "learning_rate": 9.668085894113292e-05,
      "loss": 0.2184,
      "step": 2000
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.4503909945487976,
      "learning_rate": 9.666234727878563e-05,
      "loss": 0.2129,
      "step": 2010
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.4750733971595764,
      "learning_rate": 9.664383561643836e-05,
      "loss": 0.1996,
      "step": 2020
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.7526803016662598,
      "learning_rate": 9.662532395409108e-05,
      "loss": 0.2152,
      "step": 2030
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.45865398645401,
      "learning_rate": 9.66068122917438e-05,
      "loss": 0.204,
      "step": 2040
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.4336209297180176,
      "learning_rate": 9.658830062939652e-05,
      "loss": 0.1989,
      "step": 2050
    },
    {
      "epoch": 0.38,
      "grad_norm": 2.891535520553589,
      "learning_rate": 9.656978896704925e-05,
      "loss": 0.2072,
      "step": 2060
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.5239700675010681,
      "learning_rate": 9.655127730470196e-05,
      "loss": 0.2116,
      "step": 2070
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.3464105725288391,
      "learning_rate": 9.653276564235468e-05,
      "loss": 0.2038,
      "step": 2080
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.4590921401977539,
      "learning_rate": 9.651425398000741e-05,
      "loss": 0.2151,
      "step": 2090
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.5390322208404541,
      "learning_rate": 9.649574231766013e-05,
      "loss": 0.2165,
      "step": 2100
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.8248626589775085,
      "learning_rate": 9.647723065531286e-05,
      "loss": 0.199,
      "step": 2110
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.5358192324638367,
      "learning_rate": 9.645871899296556e-05,
      "loss": 0.2159,
      "step": 2120
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.0364553928375244,
      "learning_rate": 9.64402073306183e-05,
      "loss": 0.1947,
      "step": 2130
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.628521740436554,
      "learning_rate": 9.642169566827101e-05,
      "loss": 0.2212,
      "step": 2140
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.6107978224754333,
      "learning_rate": 9.640318400592374e-05,
      "loss": 0.2064,
      "step": 2150
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.7244569063186646,
      "learning_rate": 9.638467234357646e-05,
      "loss": 0.2158,
      "step": 2160
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.4745640158653259,
      "learning_rate": 9.636616068122918e-05,
      "loss": 0.2143,
      "step": 2170
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.6135119795799255,
      "learning_rate": 9.634764901888191e-05,
      "loss": 0.2166,
      "step": 2180
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.29639098048210144,
      "learning_rate": 9.632913735653463e-05,
      "loss": 0.2031,
      "step": 2190
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.7129987478256226,
      "learning_rate": 9.631062569418734e-05,
      "loss": 0.2057,
      "step": 2200
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.079046607017517,
      "learning_rate": 9.629211403184006e-05,
      "loss": 0.2012,
      "step": 2210
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.61760014295578,
      "learning_rate": 9.627360236949279e-05,
      "loss": 0.198,
      "step": 2220
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.36462926864624023,
      "learning_rate": 9.62550907071455e-05,
      "loss": 0.2003,
      "step": 2230
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.37455418705940247,
      "learning_rate": 9.623657904479823e-05,
      "loss": 0.2234,
      "step": 2240
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.3829789459705353,
      "learning_rate": 9.621806738245094e-05,
      "loss": 0.2177,
      "step": 2250
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.760238528251648,
      "learning_rate": 9.619955572010367e-05,
      "loss": 0.2109,
      "step": 2260
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.3408792316913605,
      "learning_rate": 9.618104405775639e-05,
      "loss": 0.1997,
      "step": 2270
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.38549286127090454,
      "learning_rate": 9.616253239540911e-05,
      "loss": 0.213,
      "step": 2280
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.2211161851882935,
      "learning_rate": 9.614402073306184e-05,
      "loss": 0.2104,
      "step": 2290
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.7109599113464355,
      "learning_rate": 9.612550907071456e-05,
      "loss": 0.206,
      "step": 2300
    },
    {
      "epoch": 0.43,
      "grad_norm": 2.4902968406677246,
      "learning_rate": 9.610699740836727e-05,
      "loss": 0.2169,
      "step": 2310
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.45785343647003174,
      "learning_rate": 9.608848574601999e-05,
      "loss": 0.201,
      "step": 2320
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.5752105116844177,
      "learning_rate": 9.606997408367272e-05,
      "loss": 0.1957,
      "step": 2330
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.4003032445907593,
      "learning_rate": 9.605146242132544e-05,
      "loss": 0.2131,
      "step": 2340
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.7407622337341309,
      "learning_rate": 9.603295075897816e-05,
      "loss": 0.2069,
      "step": 2350
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.6703731417655945,
      "learning_rate": 9.601443909663089e-05,
      "loss": 0.2068,
      "step": 2360
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.8653967976570129,
      "learning_rate": 9.59959274342836e-05,
      "loss": 0.2001,
      "step": 2370
    },
    {
      "epoch": 0.44,
      "grad_norm": 3.6894798278808594,
      "learning_rate": 9.597741577193634e-05,
      "loss": 0.2133,
      "step": 2380
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.5660738945007324,
      "learning_rate": 9.595890410958904e-05,
      "loss": 0.2186,
      "step": 2390
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.0196833610534668,
      "learning_rate": 9.594039244724177e-05,
      "loss": 0.2096,
      "step": 2400
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.25469130277633667,
      "learning_rate": 9.592188078489449e-05,
      "loss": 0.2086,
      "step": 2410
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.517469584941864,
      "learning_rate": 9.59033691225472e-05,
      "loss": 0.2142,
      "step": 2420
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.8714912533760071,
      "learning_rate": 9.588485746019992e-05,
      "loss": 0.1995,
      "step": 2430
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.4467742443084717,
      "learning_rate": 9.586634579785265e-05,
      "loss": 0.1988,
      "step": 2440
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.6827601790428162,
      "learning_rate": 9.584783413550537e-05,
      "loss": 0.2004,
      "step": 2450
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.5829424262046814,
      "learning_rate": 9.582932247315809e-05,
      "loss": 0.2043,
      "step": 2460
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.6504799127578735,
      "learning_rate": 9.581081081081082e-05,
      "loss": 0.2082,
      "step": 2470
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.2879631221294403,
      "learning_rate": 9.579229914846354e-05,
      "loss": 0.2064,
      "step": 2480
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.337453693151474,
      "learning_rate": 9.577378748611627e-05,
      "loss": 0.2195,
      "step": 2490
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.7913995981216431,
      "learning_rate": 9.575527582376897e-05,
      "loss": 0.2002,
      "step": 2500
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.33184346556663513,
      "learning_rate": 9.57367641614217e-05,
      "loss": 0.1974,
      "step": 2510
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.3612488508224487,
      "learning_rate": 9.571825249907442e-05,
      "loss": 0.2081,
      "step": 2520
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.3221415281295776,
      "learning_rate": 9.569974083672715e-05,
      "loss": 0.2142,
      "step": 2530
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.7433180212974548,
      "learning_rate": 9.568122917437987e-05,
      "loss": 0.2087,
      "step": 2540
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.3186396062374115,
      "learning_rate": 9.566271751203258e-05,
      "loss": 0.2161,
      "step": 2550
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.9290895462036133,
      "learning_rate": 9.564420584968531e-05,
      "loss": 0.2165,
      "step": 2560
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.9218959212303162,
      "learning_rate": 9.562569418733803e-05,
      "loss": 0.2077,
      "step": 2570
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.6397104859352112,
      "learning_rate": 9.560718252499075e-05,
      "loss": 0.205,
      "step": 2580
    },
    {
      "epoch": 0.48,
      "grad_norm": 2.004181385040283,
      "learning_rate": 9.558867086264347e-05,
      "loss": 0.2093,
      "step": 2590
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.29606181383132935,
      "learning_rate": 9.55701592002962e-05,
      "loss": 0.2145,
      "step": 2600
    },
    {
      "epoch": 0.48,
      "grad_norm": 2.1520352363586426,
      "learning_rate": 9.55516475379489e-05,
      "loss": 0.2165,
      "step": 2610
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.1662155389785767,
      "learning_rate": 9.553313587560163e-05,
      "loss": 0.1981,
      "step": 2620
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.5705534815788269,
      "learning_rate": 9.551462421325436e-05,
      "loss": 0.2004,
      "step": 2630
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.2732492685317993,
      "learning_rate": 9.549611255090708e-05,
      "loss": 0.2137,
      "step": 2640
    },
    {
      "epoch": 0.49,
      "grad_norm": 2.1860570907592773,
      "learning_rate": 9.54776008885598e-05,
      "loss": 0.2057,
      "step": 2650
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.5216954350471497,
      "learning_rate": 9.545908922621251e-05,
      "loss": 0.2175,
      "step": 2660
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.7214330434799194,
      "learning_rate": 9.544057756386524e-05,
      "loss": 0.2104,
      "step": 2670
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.1864840984344482,
      "learning_rate": 9.542206590151796e-05,
      "loss": 0.2168,
      "step": 2680
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.7450833916664124,
      "learning_rate": 9.540355423917068e-05,
      "loss": 0.2064,
      "step": 2690
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.97665274143219,
      "learning_rate": 9.53850425768234e-05,
      "loss": 0.2132,
      "step": 2700
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.8283855319023132,
      "learning_rate": 9.536653091447613e-05,
      "loss": 0.217,
      "step": 2710
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.8062860369682312,
      "learning_rate": 9.534801925212884e-05,
      "loss": 0.1985,
      "step": 2720
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.2342814207077026,
      "learning_rate": 9.532950758978156e-05,
      "loss": 0.2043,
      "step": 2730
    },
    {
      "epoch": 0.51,
      "grad_norm": 2.1351473331451416,
      "learning_rate": 9.531099592743429e-05,
      "loss": 0.2145,
      "step": 2740
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.2654520273208618,
      "learning_rate": 9.529248426508701e-05,
      "loss": 0.2067,
      "step": 2750
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.3491131067276001,
      "learning_rate": 9.527397260273974e-05,
      "loss": 0.2008,
      "step": 2760
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.2141033411026,
      "learning_rate": 9.525546094039244e-05,
      "loss": 0.1965,
      "step": 2770
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.3704434037208557,
      "learning_rate": 9.523694927804518e-05,
      "loss": 0.2041,
      "step": 2780
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.8885278701782227,
      "learning_rate": 9.521843761569789e-05,
      "loss": 0.2062,
      "step": 2790
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.622947633266449,
      "learning_rate": 9.519992595335061e-05,
      "loss": 0.2023,
      "step": 2800
    },
    {
      "epoch": 0.52,
      "grad_norm": 2.0077390670776367,
      "learning_rate": 9.518141429100334e-05,
      "loss": 0.2025,
      "step": 2810
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.7942656874656677,
      "learning_rate": 9.516290262865606e-05,
      "loss": 0.201,
      "step": 2820
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.2571896314620972,
      "learning_rate": 9.514439096630879e-05,
      "loss": 0.206,
      "step": 2830
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.8487832546234131,
      "learning_rate": 9.512587930396149e-05,
      "loss": 0.2079,
      "step": 2840
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.46226075291633606,
      "learning_rate": 9.510736764161422e-05,
      "loss": 0.2048,
      "step": 2850
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.33343935012817383,
      "learning_rate": 9.508885597926694e-05,
      "loss": 0.2082,
      "step": 2860
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.5473183393478394,
      "learning_rate": 9.507034431691967e-05,
      "loss": 0.2033,
      "step": 2870
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.767188310623169,
      "learning_rate": 9.505183265457238e-05,
      "loss": 0.2021,
      "step": 2880
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.36617210507392883,
      "learning_rate": 9.50333209922251e-05,
      "loss": 0.2008,
      "step": 2890
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.4321964979171753,
      "learning_rate": 9.501480932987782e-05,
      "loss": 0.2019,
      "step": 2900
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.39318132400512695,
      "learning_rate": 9.499629766753055e-05,
      "loss": 0.2004,
      "step": 2910
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.42340582609176636,
      "learning_rate": 9.497778600518327e-05,
      "loss": 0.1888,
      "step": 2920
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.6964154243469238,
      "learning_rate": 9.495927434283599e-05,
      "loss": 0.1941,
      "step": 2930
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.6123028993606567,
      "learning_rate": 9.494076268048872e-05,
      "loss": 0.2073,
      "step": 2940
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.6400339007377625,
      "learning_rate": 9.492225101814142e-05,
      "loss": 0.2021,
      "step": 2950
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.3889828026294708,
      "learning_rate": 9.490373935579415e-05,
      "loss": 0.197,
      "step": 2960
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.2391413450241089,
      "learning_rate": 9.488522769344687e-05,
      "loss": 0.1988,
      "step": 2970
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.3817205429077148,
      "learning_rate": 9.48667160310996e-05,
      "loss": 0.1995,
      "step": 2980
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.47163134813308716,
      "learning_rate": 9.484820436875232e-05,
      "loss": 0.1899,
      "step": 2990
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.9131165146827698,
      "learning_rate": 9.482969270640504e-05,
      "loss": 0.1944,
      "step": 3000
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.6244738101959229,
      "learning_rate": 9.481118104405777e-05,
      "loss": 0.207,
      "step": 3010
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.5568848848342896,
      "learning_rate": 9.479266938171048e-05,
      "loss": 0.1842,
      "step": 3020
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.7403110861778259,
      "learning_rate": 9.47741577193632e-05,
      "loss": 0.1981,
      "step": 3030
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.1462550163269043,
      "learning_rate": 9.475564605701592e-05,
      "loss": 0.1967,
      "step": 3040
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.9351162910461426,
      "learning_rate": 9.473713439466865e-05,
      "loss": 0.1999,
      "step": 3050
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.4262019693851471,
      "learning_rate": 9.471862273232137e-05,
      "loss": 0.1889,
      "step": 3060
    },
    {
      "epoch": 0.57,
      "grad_norm": 1.5111364126205444,
      "learning_rate": 9.470011106997408e-05,
      "loss": 0.2139,
      "step": 3070
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.6351222991943359,
      "learning_rate": 9.468159940762682e-05,
      "loss": 0.1952,
      "step": 3080
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.46668490767478943,
      "learning_rate": 9.466308774527953e-05,
      "loss": 0.1967,
      "step": 3090
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.43086475133895874,
      "learning_rate": 9.464457608293226e-05,
      "loss": 0.2037,
      "step": 3100
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.2739882171154022,
      "learning_rate": 9.462606442058497e-05,
      "loss": 0.2141,
      "step": 3110
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.6408763527870178,
      "learning_rate": 9.46075527582377e-05,
      "loss": 0.1994,
      "step": 3120
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.33041083812713623,
      "learning_rate": 9.458904109589042e-05,
      "loss": 0.1992,
      "step": 3130
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.2604489326477051,
      "learning_rate": 9.457052943354313e-05,
      "loss": 0.2044,
      "step": 3140
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.0766637325286865,
      "learning_rate": 9.455201777119585e-05,
      "loss": 0.2121,
      "step": 3150
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.4685893058776855,
      "learning_rate": 9.453350610884858e-05,
      "loss": 0.2114,
      "step": 3160
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.42889487743377686,
      "learning_rate": 9.45149944465013e-05,
      "loss": 0.2035,
      "step": 3170
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.5236283540725708,
      "learning_rate": 9.449648278415402e-05,
      "loss": 0.1984,
      "step": 3180
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.366331696510315,
      "learning_rate": 9.447797112180675e-05,
      "loss": 0.2003,
      "step": 3190
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.7277652025222778,
      "learning_rate": 9.445945945945946e-05,
      "loss": 0.1902,
      "step": 3200
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.5104700922966003,
      "learning_rate": 9.44409477971122e-05,
      "loss": 0.2041,
      "step": 3210
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.307566374540329,
      "learning_rate": 9.44224361347649e-05,
      "loss": 0.1984,
      "step": 3220
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.6505478620529175,
      "learning_rate": 9.440392447241763e-05,
      "loss": 0.2023,
      "step": 3230
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.1189345121383667,
      "learning_rate": 9.438541281007035e-05,
      "loss": 0.1801,
      "step": 3240
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.2170076370239258,
      "learning_rate": 9.436690114772308e-05,
      "loss": 0.2064,
      "step": 3250
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.3985789716243744,
      "learning_rate": 9.43483894853758e-05,
      "loss": 0.2095,
      "step": 3260
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.2721532881259918,
      "learning_rate": 9.432987782302851e-05,
      "loss": 0.2031,
      "step": 3270
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.036005973815918,
      "learning_rate": 9.431136616068124e-05,
      "loss": 0.1997,
      "step": 3280
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.7809330224990845,
      "learning_rate": 9.429285449833396e-05,
      "loss": 0.2047,
      "step": 3290
    },
    {
      "epoch": 0.61,
      "grad_norm": 1.0563770532608032,
      "learning_rate": 9.427434283598668e-05,
      "loss": 0.2095,
      "step": 3300
    },
    {
      "epoch": 0.61,
      "grad_norm": 1.421645998954773,
      "learning_rate": 9.42558311736394e-05,
      "loss": 0.2023,
      "step": 3310
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.821112334728241,
      "learning_rate": 9.423731951129212e-05,
      "loss": 0.2007,
      "step": 3320
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.28622138500213623,
      "learning_rate": 9.421880784894483e-05,
      "loss": 0.1934,
      "step": 3330
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.34855708479881287,
      "learning_rate": 9.420029618659756e-05,
      "loss": 0.1912,
      "step": 3340
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.5029919147491455,
      "learning_rate": 9.418178452425028e-05,
      "loss": 0.1977,
      "step": 3350
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.6966719031333923,
      "learning_rate": 9.416327286190301e-05,
      "loss": 0.1929,
      "step": 3360
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.6500990390777588,
      "learning_rate": 9.414476119955572e-05,
      "loss": 0.197,
      "step": 3370
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.8321138620376587,
      "learning_rate": 9.412624953720844e-05,
      "loss": 0.1937,
      "step": 3380
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.976116418838501,
      "learning_rate": 9.410773787486117e-05,
      "loss": 0.1971,
      "step": 3390
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.2449440956115723,
      "learning_rate": 9.408922621251389e-05,
      "loss": 0.1915,
      "step": 3400
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.3859243094921112,
      "learning_rate": 9.407071455016661e-05,
      "loss": 0.2041,
      "step": 3410
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.1816600561141968,
      "learning_rate": 9.405220288781932e-05,
      "loss": 0.2041,
      "step": 3420
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.6951285004615784,
      "learning_rate": 9.403369122547206e-05,
      "loss": 0.1971,
      "step": 3430
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.083073377609253,
      "learning_rate": 9.401517956312477e-05,
      "loss": 0.1996,
      "step": 3440
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.4412924647331238,
      "learning_rate": 9.399666790077749e-05,
      "loss": 0.2016,
      "step": 3450
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.7372565865516663,
      "learning_rate": 9.397815623843022e-05,
      "loss": 0.1969,
      "step": 3460
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.35396844148635864,
      "learning_rate": 9.395964457608294e-05,
      "loss": 0.2012,
      "step": 3470
    },
    {
      "epoch": 0.64,
      "grad_norm": 2.25356388092041,
      "learning_rate": 9.394113291373567e-05,
      "loss": 0.2023,
      "step": 3480
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.400888442993164,
      "learning_rate": 9.392262125138837e-05,
      "loss": 0.1998,
      "step": 3490
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.43919339776039124,
      "learning_rate": 9.39041095890411e-05,
      "loss": 0.1928,
      "step": 3500
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.5835129022598267,
      "learning_rate": 9.388559792669382e-05,
      "loss": 0.2013,
      "step": 3510
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.9350130558013916,
      "learning_rate": 9.386708626434654e-05,
      "loss": 0.1965,
      "step": 3520
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.47554656863212585,
      "learning_rate": 9.384857460199926e-05,
      "loss": 0.1922,
      "step": 3530
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.0913207530975342,
      "learning_rate": 9.383006293965199e-05,
      "loss": 0.2029,
      "step": 3540
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.34091272950172424,
      "learning_rate": 9.381155127730472e-05,
      "loss": 0.1923,
      "step": 3550
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.5650482773780823,
      "learning_rate": 9.379303961495742e-05,
      "loss": 0.2056,
      "step": 3560
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.7033718228340149,
      "learning_rate": 9.377452795261015e-05,
      "loss": 0.203,
      "step": 3570
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.5366137623786926,
      "learning_rate": 9.375601629026287e-05,
      "loss": 0.1995,
      "step": 3580
    },
    {
      "epoch": 0.66,
      "grad_norm": 1.450621485710144,
      "learning_rate": 9.37375046279156e-05,
      "loss": 0.2102,
      "step": 3590
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.5831088423728943,
      "learning_rate": 9.37189929655683e-05,
      "loss": 0.2112,
      "step": 3600
    },
    {
      "epoch": 0.67,
      "grad_norm": 1.3041341304779053,
      "learning_rate": 9.370048130322103e-05,
      "loss": 0.2059,
      "step": 3610
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.39401304721832275,
      "learning_rate": 9.368196964087375e-05,
      "loss": 0.2047,
      "step": 3620
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.43465203046798706,
      "learning_rate": 9.366345797852648e-05,
      "loss": 0.1872,
      "step": 3630
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.34132349491119385,
      "learning_rate": 9.36449463161792e-05,
      "loss": 0.2011,
      "step": 3640
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.29983776807785034,
      "learning_rate": 9.362643465383192e-05,
      "loss": 0.1926,
      "step": 3650
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.3644193410873413,
      "learning_rate": 9.360792299148465e-05,
      "loss": 0.2003,
      "step": 3660
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.7472958564758301,
      "learning_rate": 9.358941132913736e-05,
      "loss": 0.2162,
      "step": 3670
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.2634315490722656,
      "learning_rate": 9.357089966679008e-05,
      "loss": 0.2097,
      "step": 3680
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.3091323971748352,
      "learning_rate": 9.35523880044428e-05,
      "loss": 0.1798,
      "step": 3690
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.33748623728752136,
      "learning_rate": 9.353387634209553e-05,
      "loss": 0.1919,
      "step": 3700
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.7219259142875671,
      "learning_rate": 9.351536467974825e-05,
      "loss": 0.2047,
      "step": 3710
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.3145447075366974,
      "learning_rate": 9.349685301740096e-05,
      "loss": 0.2014,
      "step": 3720
    },
    {
      "epoch": 0.69,
      "grad_norm": 1.3122820854187012,
      "learning_rate": 9.34783413550537e-05,
      "loss": 0.1915,
      "step": 3730
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.5373870134353638,
      "learning_rate": 9.345982969270641e-05,
      "loss": 0.1953,
      "step": 3740
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.5105752348899841,
      "learning_rate": 9.344131803035913e-05,
      "loss": 0.2075,
      "step": 3750
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.5081068277359009,
      "learning_rate": 9.342280636801185e-05,
      "loss": 0.1976,
      "step": 3760
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.26259544491767883,
      "learning_rate": 9.340429470566458e-05,
      "loss": 0.1984,
      "step": 3770
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.6756144165992737,
      "learning_rate": 9.33857830433173e-05,
      "loss": 0.1986,
      "step": 3780
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.6026384234428406,
      "learning_rate": 9.336727138097001e-05,
      "loss": 0.1957,
      "step": 3790
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.5947535037994385,
      "learning_rate": 9.334875971862273e-05,
      "loss": 0.195,
      "step": 3800
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.6168169379234314,
      "learning_rate": 9.333024805627546e-05,
      "loss": 0.1951,
      "step": 3810
    },
    {
      "epoch": 0.7,
      "grad_norm": 1.3259109258651733,
      "learning_rate": 9.331173639392818e-05,
      "loss": 0.197,
      "step": 3820
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.5684218406677246,
      "learning_rate": 9.32932247315809e-05,
      "loss": 0.198,
      "step": 3830
    },
    {
      "epoch": 0.71,
      "grad_norm": 1.8599364757537842,
      "learning_rate": 9.327471306923363e-05,
      "loss": 0.2114,
      "step": 3840
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.33460474014282227,
      "learning_rate": 9.325620140688634e-05,
      "loss": 0.1891,
      "step": 3850
    },
    {
      "epoch": 0.71,
      "grad_norm": 2.3144965171813965,
      "learning_rate": 9.323768974453907e-05,
      "loss": 0.1831,
      "step": 3860
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.7101608514785767,
      "learning_rate": 9.321917808219178e-05,
      "loss": 0.2025,
      "step": 3870
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.42260223627090454,
      "learning_rate": 9.320066641984451e-05,
      "loss": 0.2095,
      "step": 3880
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.473034143447876,
      "learning_rate": 9.318215475749723e-05,
      "loss": 0.1965,
      "step": 3890
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.4022839665412903,
      "learning_rate": 9.316364309514994e-05,
      "loss": 0.1877,
      "step": 3900
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.41458505392074585,
      "learning_rate": 9.314513143280267e-05,
      "loss": 0.1973,
      "step": 3910
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.46193191409111023,
      "learning_rate": 9.312661977045539e-05,
      "loss": 0.2133,
      "step": 3920
    },
    {
      "epoch": 0.72,
      "grad_norm": 1.0600367784500122,
      "learning_rate": 9.310810810810812e-05,
      "loss": 0.1895,
      "step": 3930
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.3262641429901123,
      "learning_rate": 9.308959644576083e-05,
      "loss": 0.1975,
      "step": 3940
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.1876827478408813,
      "learning_rate": 9.307108478341356e-05,
      "loss": 0.1916,
      "step": 3950
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.6535803079605103,
      "learning_rate": 9.305257312106627e-05,
      "loss": 0.1966,
      "step": 3960
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.4089249074459076,
      "learning_rate": 9.3034061458719e-05,
      "loss": 0.195,
      "step": 3970
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.0222320556640625,
      "learning_rate": 9.301554979637171e-05,
      "loss": 0.193,
      "step": 3980
    },
    {
      "epoch": 0.74,
      "grad_norm": 1.0429261922836304,
      "learning_rate": 9.299703813402444e-05,
      "loss": 0.2039,
      "step": 3990
    },
    {
      "epoch": 0.74,
      "grad_norm": 2.6728320121765137,
      "learning_rate": 9.297852647167716e-05,
      "loss": 0.1963,
      "step": 4000
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.4279372990131378,
      "learning_rate": 9.296001480932989e-05,
      "loss": 0.1895,
      "step": 4010
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.4307747781276703,
      "learning_rate": 9.29415031469826e-05,
      "loss": 0.1967,
      "step": 4020
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.44748997688293457,
      "learning_rate": 9.292299148463532e-05,
      "loss": 0.1945,
      "step": 4030
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.3396751284599304,
      "learning_rate": 9.290447982228805e-05,
      "loss": 0.1901,
      "step": 4040
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.2306305170059204,
      "learning_rate": 9.288596815994076e-05,
      "loss": 0.1915,
      "step": 4050
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.0502023696899414,
      "learning_rate": 9.286745649759349e-05,
      "loss": 0.2001,
      "step": 4060
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.2933097779750824,
      "learning_rate": 9.28489448352462e-05,
      "loss": 0.1911,
      "step": 4070
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.7432071566581726,
      "learning_rate": 9.283043317289894e-05,
      "loss": 0.1882,
      "step": 4080
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.2065651416778564,
      "learning_rate": 9.281192151055165e-05,
      "loss": 0.2026,
      "step": 4090
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.331061452627182,
      "learning_rate": 9.279340984820437e-05,
      "loss": 0.1941,
      "step": 4100
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.819383978843689,
      "learning_rate": 9.27748981858571e-05,
      "loss": 0.1951,
      "step": 4110
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.4251770079135895,
      "learning_rate": 9.275638652350982e-05,
      "loss": 0.1976,
      "step": 4120
    },
    {
      "epoch": 0.76,
      "grad_norm": 1.0525916814804077,
      "learning_rate": 9.273787486116254e-05,
      "loss": 0.1998,
      "step": 4130
    },
    {
      "epoch": 0.76,
      "grad_norm": 1.4899557828903198,
      "learning_rate": 9.271936319881525e-05,
      "loss": 0.1933,
      "step": 4140
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.8893213272094727,
      "learning_rate": 9.270085153646798e-05,
      "loss": 0.1943,
      "step": 4150
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.6278595924377441,
      "learning_rate": 9.26823398741207e-05,
      "loss": 0.1973,
      "step": 4160
    },
    {
      "epoch": 0.77,
      "grad_norm": 1.8130468130111694,
      "learning_rate": 9.266382821177342e-05,
      "loss": 0.1954,
      "step": 4170
    },
    {
      "epoch": 0.77,
      "grad_norm": 1.0211946964263916,
      "learning_rate": 9.264531654942615e-05,
      "loss": 0.1937,
      "step": 4180
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.3028915226459503,
      "learning_rate": 9.262680488707887e-05,
      "loss": 0.208,
      "step": 4190
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.33269938826560974,
      "learning_rate": 9.26082932247316e-05,
      "loss": 0.2032,
      "step": 4200
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.9252434372901917,
      "learning_rate": 9.25897815623843e-05,
      "loss": 0.2018,
      "step": 4210
    },
    {
      "epoch": 0.78,
      "grad_norm": 1.3774528503417969,
      "learning_rate": 9.257126990003703e-05,
      "loss": 0.1895,
      "step": 4220
    },
    {
      "epoch": 0.78,
      "grad_norm": 1.9812060594558716,
      "learning_rate": 9.255275823768975e-05,
      "loss": 0.1951,
      "step": 4230
    },
    {
      "epoch": 0.78,
      "grad_norm": 1.7997092008590698,
      "learning_rate": 9.253424657534247e-05,
      "loss": 0.2014,
      "step": 4240
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.35307908058166504,
      "learning_rate": 9.251573491299518e-05,
      "loss": 0.2026,
      "step": 4250
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.906162440776825,
      "learning_rate": 9.249722325064791e-05,
      "loss": 0.1892,
      "step": 4260
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.8749983906745911,
      "learning_rate": 9.247871158830063e-05,
      "loss": 0.2037,
      "step": 4270
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.1504167318344116,
      "learning_rate": 9.246019992595335e-05,
      "loss": 0.1942,
      "step": 4280
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.6674988865852356,
      "learning_rate": 9.244168826360608e-05,
      "loss": 0.1974,
      "step": 4290
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.0109151601791382,
      "learning_rate": 9.24231766012588e-05,
      "loss": 0.1937,
      "step": 4300
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.5439345836639404,
      "learning_rate": 9.240466493891153e-05,
      "loss": 0.1825,
      "step": 4310
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.3471992015838623,
      "learning_rate": 9.238615327656423e-05,
      "loss": 0.1932,
      "step": 4320
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.5230699181556702,
      "learning_rate": 9.236764161421696e-05,
      "loss": 0.1869,
      "step": 4330
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.1102708578109741,
      "learning_rate": 9.234912995186968e-05,
      "loss": 0.1934,
      "step": 4340
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.2299565076828003,
      "learning_rate": 9.233061828952241e-05,
      "loss": 0.1946,
      "step": 4350
    },
    {
      "epoch": 0.8,
      "grad_norm": 2.019484281539917,
      "learning_rate": 9.231210662717513e-05,
      "loss": 0.2005,
      "step": 4360
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.49182599782943726,
      "learning_rate": 9.229359496482784e-05,
      "loss": 0.1862,
      "step": 4370
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.1424809694290161,
      "learning_rate": 9.227508330248058e-05,
      "loss": 0.1857,
      "step": 4380
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.5882079601287842,
      "learning_rate": 9.225657164013329e-05,
      "loss": 0.1866,
      "step": 4390
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.8041399717330933,
      "learning_rate": 9.223805997778601e-05,
      "loss": 0.1882,
      "step": 4400
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.2162193059921265,
      "learning_rate": 9.221954831543873e-05,
      "loss": 0.1893,
      "step": 4410
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.516737163066864,
      "learning_rate": 9.220103665309146e-05,
      "loss": 0.1954,
      "step": 4420
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.78972327709198,
      "learning_rate": 9.218252499074416e-05,
      "loss": 0.1959,
      "step": 4430
    },
    {
      "epoch": 0.82,
      "grad_norm": 1.6012400388717651,
      "learning_rate": 9.216401332839689e-05,
      "loss": 0.2035,
      "step": 4440
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.8954785466194153,
      "learning_rate": 9.214550166604961e-05,
      "loss": 0.1922,
      "step": 4450
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.6251717805862427,
      "learning_rate": 9.212699000370234e-05,
      "loss": 0.1898,
      "step": 4460
    },
    {
      "epoch": 0.82,
      "grad_norm": 1.7891993522644043,
      "learning_rate": 9.210847834135506e-05,
      "loss": 0.2006,
      "step": 4470
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.3996617794036865,
      "learning_rate": 9.208996667900777e-05,
      "loss": 0.2078,
      "step": 4480
    },
    {
      "epoch": 0.83,
      "grad_norm": 1.112613320350647,
      "learning_rate": 9.20714550166605e-05,
      "loss": 0.2025,
      "step": 4490
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.3055773079395294,
      "learning_rate": 9.205294335431322e-05,
      "loss": 0.193,
      "step": 4500
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.9582087993621826,
      "learning_rate": 9.203443169196594e-05,
      "loss": 0.1899,
      "step": 4510
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.7985404133796692,
      "learning_rate": 9.201592002961866e-05,
      "loss": 0.1868,
      "step": 4520
    },
    {
      "epoch": 0.84,
      "grad_norm": 1.6095155477523804,
      "learning_rate": 9.199740836727139e-05,
      "loss": 0.1899,
      "step": 4530
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.6676468849182129,
      "learning_rate": 9.19788967049241e-05,
      "loss": 0.1969,
      "step": 4540
    },
    {
      "epoch": 0.84,
      "grad_norm": 1.045982003211975,
      "learning_rate": 9.196038504257682e-05,
      "loss": 0.1814,
      "step": 4550
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.7359940409660339,
      "learning_rate": 9.194187338022955e-05,
      "loss": 0.1865,
      "step": 4560
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.9083600044250488,
      "learning_rate": 9.192336171788227e-05,
      "loss": 0.2012,
      "step": 4570
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.340991348028183,
      "learning_rate": 9.1904850055535e-05,
      "loss": 0.1892,
      "step": 4580
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.2833189368247986,
      "learning_rate": 9.18863383931877e-05,
      "loss": 0.1888,
      "step": 4590
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.6355910301208496,
      "learning_rate": 9.186782673084044e-05,
      "loss": 0.2041,
      "step": 4600
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.7121712565422058,
      "learning_rate": 9.184931506849315e-05,
      "loss": 0.1783,
      "step": 4610
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.8881193399429321,
      "learning_rate": 9.183080340614587e-05,
      "loss": 0.1981,
      "step": 4620
    },
    {
      "epoch": 0.85,
      "grad_norm": 1.2488733530044556,
      "learning_rate": 9.18122917437986e-05,
      "loss": 0.2016,
      "step": 4630
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.283454567193985,
      "learning_rate": 9.179378008145132e-05,
      "loss": 0.1966,
      "step": 4640
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.306062787771225,
      "learning_rate": 9.177526841910405e-05,
      "loss": 0.193,
      "step": 4650
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.45514628291130066,
      "learning_rate": 9.175675675675675e-05,
      "loss": 0.192,
      "step": 4660
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.6448143124580383,
      "learning_rate": 9.173824509440948e-05,
      "loss": 0.1996,
      "step": 4670
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.8043811321258545,
      "learning_rate": 9.17197334320622e-05,
      "loss": 0.1973,
      "step": 4680
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.3627445101737976,
      "learning_rate": 9.170122176971493e-05,
      "loss": 0.1864,
      "step": 4690
    },
    {
      "epoch": 0.87,
      "grad_norm": 1.8266276121139526,
      "learning_rate": 9.168271010736764e-05,
      "loss": 0.1939,
      "step": 4700
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.8964821100234985,
      "learning_rate": 9.166419844502037e-05,
      "loss": 0.21,
      "step": 4710
    },
    {
      "epoch": 0.87,
      "grad_norm": 2.066237449645996,
      "learning_rate": 9.164568678267308e-05,
      "loss": 0.2042,
      "step": 4720
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.5604624152183533,
      "learning_rate": 9.162717512032582e-05,
      "loss": 0.195,
      "step": 4730
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.3185977041721344,
      "learning_rate": 9.160866345797853e-05,
      "loss": 0.1913,
      "step": 4740
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.7183963060379028,
      "learning_rate": 9.159015179563125e-05,
      "loss": 0.1893,
      "step": 4750
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.30005547404289246,
      "learning_rate": 9.157164013328398e-05,
      "loss": 0.2004,
      "step": 4760
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.1535359621047974,
      "learning_rate": 9.15531284709367e-05,
      "loss": 0.1859,
      "step": 4770
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.46643924713134766,
      "learning_rate": 9.153461680858941e-05,
      "loss": 0.198,
      "step": 4780
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.45236846804618835,
      "learning_rate": 9.151610514624213e-05,
      "loss": 0.1911,
      "step": 4790
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.8338435888290405,
      "learning_rate": 9.149759348389486e-05,
      "loss": 0.1944,
      "step": 4800
    },
    {
      "epoch": 0.89,
      "grad_norm": 1.7582203149795532,
      "learning_rate": 9.147908182154758e-05,
      "loss": 0.1973,
      "step": 4810
    },
    {
      "epoch": 0.89,
      "grad_norm": 1.3786252737045288,
      "learning_rate": 9.14605701592003e-05,
      "loss": 0.1867,
      "step": 4820
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.376365602016449,
      "learning_rate": 9.144205849685303e-05,
      "loss": 0.1831,
      "step": 4830
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.7689598798751831,
      "learning_rate": 9.142354683450575e-05,
      "loss": 0.1911,
      "step": 4840
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.9973986744880676,
      "learning_rate": 9.140503517215846e-05,
      "loss": 0.1984,
      "step": 4850
    },
    {
      "epoch": 0.9,
      "grad_norm": 2.4425342082977295,
      "learning_rate": 9.138652350981118e-05,
      "loss": 0.1927,
      "step": 4860
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.4806678295135498,
      "learning_rate": 9.136801184746391e-05,
      "loss": 0.1857,
      "step": 4870
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.6703113913536072,
      "learning_rate": 9.134950018511663e-05,
      "loss": 0.1924,
      "step": 4880
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.6484790444374084,
      "learning_rate": 9.133098852276935e-05,
      "loss": 0.1774,
      "step": 4890
    },
    {
      "epoch": 0.9,
      "grad_norm": 1.3424222469329834,
      "learning_rate": 9.131247686042206e-05,
      "loss": 0.1834,
      "step": 4900
    },
    {
      "epoch": 0.91,
      "grad_norm": 1.3998688459396362,
      "learning_rate": 9.12939651980748e-05,
      "loss": 0.1856,
      "step": 4910
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.37422481179237366,
      "learning_rate": 9.127545353572751e-05,
      "loss": 0.1875,
      "step": 4920
    },
    {
      "epoch": 0.91,
      "grad_norm": 1.2003308534622192,
      "learning_rate": 9.125694187338023e-05,
      "loss": 0.1886,
      "step": 4930
    },
    {
      "epoch": 0.91,
      "grad_norm": 1.248398780822754,
      "learning_rate": 9.123843021103296e-05,
      "loss": 0.1918,
      "step": 4940
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.6824427247047424,
      "learning_rate": 9.121991854868568e-05,
      "loss": 0.1795,
      "step": 4950
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.5873004794120789,
      "learning_rate": 9.120140688633841e-05,
      "loss": 0.1958,
      "step": 4960
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.7316002249717712,
      "learning_rate": 9.118289522399111e-05,
      "loss": 0.1794,
      "step": 4970
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.40432649850845337,
      "learning_rate": 9.116438356164384e-05,
      "loss": 0.1845,
      "step": 4980
    },
    {
      "epoch": 0.92,
      "grad_norm": 1.2551237344741821,
      "learning_rate": 9.114587189929656e-05,
      "loss": 0.1847,
      "step": 4990
    },
    {
      "epoch": 0.92,
      "grad_norm": 1.0834324359893799,
      "learning_rate": 9.112736023694928e-05,
      "loss": 0.1942,
      "step": 5000
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.5848122835159302,
      "learning_rate": 9.110884857460201e-05,
      "loss": 0.1909,
      "step": 5010
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.34502893686294556,
      "learning_rate": 9.109033691225472e-05,
      "loss": 0.1954,
      "step": 5020
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.7165638208389282,
      "learning_rate": 9.107182524990745e-05,
      "loss": 0.1856,
      "step": 5030
    },
    {
      "epoch": 0.93,
      "grad_norm": 1.3096377849578857,
      "learning_rate": 9.105331358756016e-05,
      "loss": 0.1865,
      "step": 5040
    },
    {
      "epoch": 0.93,
      "grad_norm": 2.7340123653411865,
      "learning_rate": 9.103480192521289e-05,
      "loss": 0.1957,
      "step": 5050
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.3616611361503601,
      "learning_rate": 9.101629026286561e-05,
      "loss": 0.1967,
      "step": 5060
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.4186195731163025,
      "learning_rate": 9.099777860051834e-05,
      "loss": 0.186,
      "step": 5070
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.2687035799026489,
      "learning_rate": 9.097926693817105e-05,
      "loss": 0.1873,
      "step": 5080
    },
    {
      "epoch": 0.94,
      "grad_norm": 1.4722837209701538,
      "learning_rate": 9.096075527582377e-05,
      "loss": 0.1881,
      "step": 5090
    },
    {
      "epoch": 0.94,
      "grad_norm": 1.3331010341644287,
      "learning_rate": 9.09422436134765e-05,
      "loss": 0.1979,
      "step": 5100
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.7769159078598022,
      "learning_rate": 9.092373195112922e-05,
      "loss": 0.1957,
      "step": 5110
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.9742639064788818,
      "learning_rate": 9.090522028878194e-05,
      "loss": 0.1883,
      "step": 5120
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.4907156527042389,
      "learning_rate": 9.088670862643465e-05,
      "loss": 0.1885,
      "step": 5130
    },
    {
      "epoch": 0.95,
      "grad_norm": 1.0384258031845093,
      "learning_rate": 9.086819696408739e-05,
      "loss": 0.183,
      "step": 5140
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.6687269806861877,
      "learning_rate": 9.084968530174009e-05,
      "loss": 0.1986,
      "step": 5150
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.49845072627067566,
      "learning_rate": 9.083117363939282e-05,
      "loss": 0.1897,
      "step": 5160
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.3812813460826874,
      "learning_rate": 9.081266197704554e-05,
      "loss": 0.1959,
      "step": 5170
    },
    {
      "epoch": 0.96,
      "grad_norm": 1.7052239179611206,
      "learning_rate": 9.079415031469827e-05,
      "loss": 0.1955,
      "step": 5180
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.8822137713432312,
      "learning_rate": 9.077563865235099e-05,
      "loss": 0.1809,
      "step": 5190
    },
    {
      "epoch": 0.96,
      "grad_norm": 1.1971176862716675,
      "learning_rate": 9.07571269900037e-05,
      "loss": 0.192,
      "step": 5200
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.6374858021736145,
      "learning_rate": 9.073861532765643e-05,
      "loss": 0.1938,
      "step": 5210
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.6527014970779419,
      "learning_rate": 9.072010366530915e-05,
      "loss": 0.1933,
      "step": 5220
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.628654956817627,
      "learning_rate": 9.070159200296187e-05,
      "loss": 0.1794,
      "step": 5230
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.7301254272460938,
      "learning_rate": 9.068308034061459e-05,
      "loss": 0.2051,
      "step": 5240
    },
    {
      "epoch": 0.97,
      "grad_norm": 1.1322051286697388,
      "learning_rate": 9.066456867826732e-05,
      "loss": 0.1811,
      "step": 5250
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.5501244068145752,
      "learning_rate": 9.064605701592003e-05,
      "loss": 0.1868,
      "step": 5260
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.9664620161056519,
      "learning_rate": 9.062754535357275e-05,
      "loss": 0.1907,
      "step": 5270
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.3740132749080658,
      "learning_rate": 9.060903369122548e-05,
      "loss": 0.1796,
      "step": 5280
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.9004091024398804,
      "learning_rate": 9.05905220288782e-05,
      "loss": 0.1814,
      "step": 5290
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.5110592246055603,
      "learning_rate": 9.057201036653093e-05,
      "loss": 0.1897,
      "step": 5300
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.6021327972412109,
      "learning_rate": 9.055349870418363e-05,
      "loss": 0.1848,
      "step": 5310
    },
    {
      "epoch": 0.98,
      "grad_norm": 1.157943844795227,
      "learning_rate": 9.053498704183636e-05,
      "loss": 0.175,
      "step": 5320
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.6326828598976135,
      "learning_rate": 9.051647537948908e-05,
      "loss": 0.1884,
      "step": 5330
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.37102219462394714,
      "learning_rate": 9.04979637171418e-05,
      "loss": 0.194,
      "step": 5340
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.9475340247154236,
      "learning_rate": 9.047945205479452e-05,
      "loss": 0.1971,
      "step": 5350
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.4836327135562897,
      "learning_rate": 9.046094039244725e-05,
      "loss": 0.1909,
      "step": 5360
    },
    {
      "epoch": 0.99,
      "grad_norm": 1.3864704370498657,
      "learning_rate": 9.044242873009996e-05,
      "loss": 0.1874,
      "step": 5370
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.7823747992515564,
      "learning_rate": 9.042391706775268e-05,
      "loss": 0.1911,
      "step": 5380
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.9502007365226746,
      "learning_rate": 9.040540540540541e-05,
      "loss": 0.1815,
      "step": 5390
    },
    {
      "epoch": 1.0,
      "grad_norm": 1.5369421243667603,
      "learning_rate": 9.038689374305813e-05,
      "loss": 0.1865,
      "step": 5400
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.3976035416126251,
      "learning_rate": 9.036838208071086e-05,
      "loss": 0.1835,
      "step": 5410
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.5696306824684143,
      "learning_rate": 9.034987041836356e-05,
      "loss": 0.1761,
      "step": 5420
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.3201138377189636,
      "eval_runtime": 3918.2947,
      "eval_samples_per_second": 262.734,
      "eval_steps_per_second": 4.105,
      "step": 5422
    }
  ],
  "logging_steps": 10,
  "max_steps": 54220,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 500,
  "total_flos": 2.3125260009084027e+18,
  "train_batch_size": 6,
  "trial_name": null,
  "trial_params": null
}