{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 12500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.002,
      "grad_norm": 9.999999046325684,
      "learning_rate": 4.000000000000001e-06,
      "loss": 18.7387,
      "step": 25
    },
    {
      "epoch": 0.004,
      "grad_norm": 10.0,
      "learning_rate": 8.000000000000001e-06,
      "loss": 22.1438,
      "step": 50
    },
    {
      "epoch": 0.006,
      "grad_norm": 10.0,
      "learning_rate": 1.2e-05,
      "loss": 23.3287,
      "step": 75
    },
    {
      "epoch": 0.008,
      "grad_norm": 9.999999046325684,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 21.9887,
      "step": 100
    },
    {
      "epoch": 0.01,
      "grad_norm": 10.0,
      "learning_rate": 2e-05,
      "loss": 18.4825,
      "step": 125
    },
    {
      "epoch": 0.012,
      "grad_norm": 10.000000953674316,
      "learning_rate": 2.4e-05,
      "loss": 16.1713,
      "step": 150
    },
    {
      "epoch": 0.014,
      "grad_norm": 10.0,
      "learning_rate": 2.8000000000000003e-05,
      "loss": 15.2963,
      "step": 175
    },
    {
      "epoch": 0.016,
      "grad_norm": 9.999999046325684,
      "learning_rate": 3.2000000000000005e-05,
      "loss": 13.4162,
      "step": 200
    },
    {
      "epoch": 0.018,
      "grad_norm": 10.0,
      "learning_rate": 3.6e-05,
      "loss": 9.6975,
      "step": 225
    },
    {
      "epoch": 0.02,
      "grad_norm": 10.0,
      "learning_rate": 4e-05,
      "loss": 7.6338,
      "step": 250
    },
    {
      "epoch": 0.022,
      "grad_norm": 10.0,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 6.175,
      "step": 275
    },
    {
      "epoch": 0.024,
      "grad_norm": 9.999998092651367,
      "learning_rate": 4.8e-05,
      "loss": 4.6375,
      "step": 300
    },
    {
      "epoch": 0.026,
      "grad_norm": 9.999998092651367,
      "learning_rate": 5.2000000000000004e-05,
      "loss": 3.1523,
      "step": 325
    },
    {
      "epoch": 0.028,
      "grad_norm": 9.999998092651367,
      "learning_rate": 5.6000000000000006e-05,
      "loss": 2.9066,
      "step": 350
    },
    {
      "epoch": 0.03,
      "grad_norm": 10.0,
      "learning_rate": 6e-05,
      "loss": 1.7845,
      "step": 375
    },
    {
      "epoch": 0.032,
      "grad_norm": 9.999999046325684,
      "learning_rate": 6.400000000000001e-05,
      "loss": 1.2612,
      "step": 400
    },
    {
      "epoch": 0.034,
      "grad_norm": 9.999999046325684,
      "learning_rate": 6.800000000000001e-05,
      "loss": 1.1808,
      "step": 425
    },
    {
      "epoch": 0.036,
      "grad_norm": 9.34283447265625,
      "learning_rate": 7.2e-05,
      "loss": 1.0462,
      "step": 450
    },
    {
      "epoch": 0.038,
      "grad_norm": 8.628987312316895,
      "learning_rate": 7.6e-05,
      "loss": 0.847,
      "step": 475
    },
    {
      "epoch": 0.04,
      "grad_norm": 7.913747787475586,
      "learning_rate": 8e-05,
      "loss": 0.7295,
      "step": 500
    },
    {
      "epoch": 0.042,
      "grad_norm": 10.0,
      "learning_rate": 8.4e-05,
      "loss": 0.7154,
      "step": 525
    },
    {
      "epoch": 0.044,
      "grad_norm": 5.550658702850342,
      "learning_rate": 8.800000000000001e-05,
      "loss": 0.5505,
      "step": 550
    },
    {
      "epoch": 0.046,
      "grad_norm": 2.6023919582366943,
      "learning_rate": 9.200000000000001e-05,
      "loss": 0.6063,
      "step": 575
    },
    {
      "epoch": 0.048,
      "grad_norm": 6.640954971313477,
      "learning_rate": 9.6e-05,
      "loss": 0.6737,
      "step": 600
    },
    {
      "epoch": 0.05,
      "grad_norm": 6.298529624938965,
      "learning_rate": 0.0001,
      "loss": 0.6029,
      "step": 625
    },
    {
      "epoch": 0.052,
      "grad_norm": 4.304025173187256,
      "learning_rate": 9.978947368421054e-05,
      "loss": 0.5277,
      "step": 650
    },
    {
      "epoch": 0.054,
      "grad_norm": 4.449888229370117,
      "learning_rate": 9.957894736842106e-05,
      "loss": 0.4861,
      "step": 675
    },
    {
      "epoch": 0.056,
      "grad_norm": 2.9031755924224854,
      "learning_rate": 9.936842105263159e-05,
      "loss": 0.4413,
      "step": 700
    },
    {
      "epoch": 0.058,
      "grad_norm": 4.989285469055176,
      "learning_rate": 9.915789473684211e-05,
      "loss": 0.4143,
      "step": 725
    },
    {
      "epoch": 0.06,
      "grad_norm": 5.094226360321045,
      "learning_rate": 9.894736842105263e-05,
      "loss": 0.4132,
      "step": 750
    },
    {
      "epoch": 0.062,
      "grad_norm": 6.01052713394165,
      "learning_rate": 9.873684210526316e-05,
      "loss": 0.3645,
      "step": 775
    },
    {
      "epoch": 0.064,
      "grad_norm": 3.638715982437134,
      "learning_rate": 9.852631578947368e-05,
      "loss": 0.3468,
      "step": 800
    },
    {
      "epoch": 0.066,
      "grad_norm": 5.820480823516846,
      "learning_rate": 9.831578947368422e-05,
      "loss": 0.3396,
      "step": 825
    },
    {
      "epoch": 0.068,
      "grad_norm": 4.334630966186523,
      "learning_rate": 9.810526315789475e-05,
      "loss": 0.3035,
      "step": 850
    },
    {
      "epoch": 0.07,
      "grad_norm": 4.271489143371582,
      "learning_rate": 9.789473684210527e-05,
      "loss": 0.3495,
      "step": 875
    },
    {
      "epoch": 0.072,
      "grad_norm": 4.406955718994141,
      "learning_rate": 9.768421052631579e-05,
      "loss": 0.3061,
      "step": 900
    },
    {
      "epoch": 0.074,
      "grad_norm": 3.8621153831481934,
      "learning_rate": 9.747368421052632e-05,
      "loss": 0.2804,
      "step": 925
    },
    {
      "epoch": 0.076,
      "grad_norm": 3.116670846939087,
      "learning_rate": 9.726315789473684e-05,
      "loss": 0.2664,
      "step": 950
    },
    {
      "epoch": 0.078,
      "grad_norm": 2.6488585472106934,
      "learning_rate": 9.705263157894738e-05,
      "loss": 0.2616,
      "step": 975
    },
    {
      "epoch": 0.08,
      "grad_norm": 3.343226194381714,
      "learning_rate": 9.68421052631579e-05,
      "loss": 0.2389,
      "step": 1000
    },
    {
      "epoch": 0.082,
      "grad_norm": 2.8532614707946777,
      "learning_rate": 9.663157894736843e-05,
      "loss": 0.2532,
      "step": 1025
    },
    {
      "epoch": 0.084,
      "grad_norm": 2.423327922821045,
      "learning_rate": 9.642105263157896e-05,
      "loss": 0.2479,
      "step": 1050
    },
    {
      "epoch": 0.086,
      "grad_norm": 2.269280195236206,
      "learning_rate": 9.621052631578947e-05,
      "loss": 0.2208,
      "step": 1075
    },
    {
      "epoch": 0.088,
      "grad_norm": 2.4925410747528076,
      "learning_rate": 9.6e-05,
      "loss": 0.2013,
      "step": 1100
    },
    {
      "epoch": 0.09,
      "grad_norm": 4.284687042236328,
      "learning_rate": 9.578947368421052e-05,
      "loss": 0.1964,
      "step": 1125
    },
    {
      "epoch": 0.092,
      "grad_norm": 2.524617910385132,
      "learning_rate": 9.557894736842106e-05,
      "loss": 0.2117,
      "step": 1150
    },
    {
      "epoch": 0.094,
      "grad_norm": 3.9951181411743164,
      "learning_rate": 9.536842105263159e-05,
      "loss": 0.1844,
      "step": 1175
    },
    {
      "epoch": 0.096,
      "grad_norm": 7.369756698608398,
      "learning_rate": 9.515789473684211e-05,
      "loss": 0.1838,
      "step": 1200
    },
    {
      "epoch": 0.098,
      "grad_norm": 3.532125949859619,
      "learning_rate": 9.494736842105264e-05,
      "loss": 0.1766,
      "step": 1225
    },
    {
      "epoch": 0.1,
      "grad_norm": 2.8758790493011475,
      "learning_rate": 9.473684210526316e-05,
      "loss": 0.2019,
      "step": 1250
    },
    {
      "epoch": 0.102,
      "grad_norm": 2.530642032623291,
      "learning_rate": 9.452631578947368e-05,
      "loss": 0.1539,
      "step": 1275
    },
    {
      "epoch": 0.104,
      "grad_norm": 3.9389524459838867,
      "learning_rate": 9.431578947368421e-05,
      "loss": 0.1865,
      "step": 1300
    },
    {
      "epoch": 0.106,
      "grad_norm": 2.398953437805176,
      "learning_rate": 9.410526315789473e-05,
      "loss": 0.1365,
      "step": 1325
    },
    {
      "epoch": 0.108,
      "grad_norm": 2.8597166538238525,
      "learning_rate": 9.389473684210527e-05,
      "loss": 0.1653,
      "step": 1350
    },
    {
      "epoch": 0.11,
      "grad_norm": 3.6551222801208496,
      "learning_rate": 9.36842105263158e-05,
      "loss": 0.1531,
      "step": 1375
    },
    {
      "epoch": 0.112,
      "grad_norm": 2.472316026687622,
      "learning_rate": 9.347368421052632e-05,
      "loss": 0.1281,
      "step": 1400
    },
    {
      "epoch": 0.114,
      "grad_norm": 2.308349847793579,
      "learning_rate": 9.326315789473684e-05,
      "loss": 0.1554,
      "step": 1425
    },
    {
      "epoch": 0.116,
      "grad_norm": 3.2640528678894043,
      "learning_rate": 9.305263157894737e-05,
      "loss": 0.1693,
      "step": 1450
    },
    {
      "epoch": 0.118,
      "grad_norm": 3.1926233768463135,
      "learning_rate": 9.28421052631579e-05,
      "loss": 0.2025,
      "step": 1475
    },
    {
      "epoch": 0.12,
      "grad_norm": 6.277203559875488,
      "learning_rate": 9.263157894736843e-05,
      "loss": 0.2018,
      "step": 1500
    },
    {
      "epoch": 0.122,
      "grad_norm": 5.876811981201172,
      "learning_rate": 9.242105263157895e-05,
      "loss": 0.2255,
      "step": 1525
    },
    {
      "epoch": 0.124,
      "grad_norm": 3.3951916694641113,
      "learning_rate": 9.221052631578948e-05,
      "loss": 0.1818,
      "step": 1550
    },
    {
      "epoch": 0.126,
      "grad_norm": 2.0903260707855225,
      "learning_rate": 9.200000000000001e-05,
      "loss": 0.1694,
      "step": 1575
    },
    {
      "epoch": 0.128,
      "grad_norm": 3.4474785327911377,
      "learning_rate": 9.178947368421052e-05,
      "loss": 0.1522,
      "step": 1600
    },
    {
      "epoch": 0.13,
      "grad_norm": 3.1567347049713135,
      "learning_rate": 9.157894736842105e-05,
      "loss": 0.1635,
      "step": 1625
    },
    {
      "epoch": 0.132,
      "grad_norm": 3.8927764892578125,
      "learning_rate": 9.136842105263159e-05,
      "loss": 0.1639,
      "step": 1650
    },
    {
      "epoch": 0.134,
      "grad_norm": 2.9919180870056152,
      "learning_rate": 9.11578947368421e-05,
      "loss": 0.1537,
      "step": 1675
    },
    {
      "epoch": 0.136,
      "grad_norm": 3.9411697387695312,
      "learning_rate": 9.094736842105264e-05,
      "loss": 0.1543,
      "step": 1700
    },
    {
      "epoch": 0.138,
      "grad_norm": 3.9650681018829346,
      "learning_rate": 9.073684210526316e-05,
      "loss": 0.1662,
      "step": 1725
    },
    {
      "epoch": 0.14,
      "grad_norm": 3.127971649169922,
      "learning_rate": 9.052631578947369e-05,
      "loss": 0.157,
      "step": 1750
    },
    {
      "epoch": 0.142,
      "grad_norm": 2.24344539642334,
      "learning_rate": 9.031578947368423e-05,
      "loss": 0.1414,
      "step": 1775
    },
    {
      "epoch": 0.144,
      "grad_norm": 2.141857385635376,
      "learning_rate": 9.010526315789473e-05,
      "loss": 0.1265,
      "step": 1800
    },
    {
      "epoch": 0.146,
      "grad_norm": 4.17902946472168,
      "learning_rate": 8.989473684210527e-05,
      "loss": 0.1626,
      "step": 1825
    },
    {
      "epoch": 0.148,
      "grad_norm": 2.6167194843292236,
      "learning_rate": 8.96842105263158e-05,
      "loss": 0.1488,
      "step": 1850
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.361194372177124,
      "learning_rate": 8.947368421052632e-05,
      "loss": 0.1134,
      "step": 1875
    },
    {
      "epoch": 0.152,
      "grad_norm": 3.2407174110412598,
      "learning_rate": 8.926315789473685e-05,
      "loss": 0.1277,
      "step": 1900
    },
    {
      "epoch": 0.154,
      "grad_norm": 2.556647539138794,
      "learning_rate": 8.905263157894737e-05,
      "loss": 0.1205,
      "step": 1925
    },
    {
      "epoch": 0.156,
      "grad_norm": 1.8379806280136108,
      "learning_rate": 8.88421052631579e-05,
      "loss": 0.1511,
      "step": 1950
    },
    {
      "epoch": 0.158,
      "grad_norm": 1.8115049600601196,
      "learning_rate": 8.863157894736842e-05,
      "loss": 0.0999,
      "step": 1975
    },
    {
      "epoch": 0.16,
      "grad_norm": 2.2672791481018066,
      "learning_rate": 8.842105263157894e-05,
      "loss": 0.1025,
      "step": 2000
    },
    {
      "epoch": 0.162,
      "grad_norm": 3.3364503383636475,
      "learning_rate": 8.821052631578948e-05,
      "loss": 0.1158,
      "step": 2025
    },
    {
      "epoch": 0.164,
      "grad_norm": 2.527956247329712,
      "learning_rate": 8.800000000000001e-05,
      "loss": 0.1248,
      "step": 2050
    },
    {
      "epoch": 0.166,
      "grad_norm": 3.559095859527588,
      "learning_rate": 8.778947368421053e-05,
      "loss": 0.1139,
      "step": 2075
    },
    {
      "epoch": 0.168,
      "grad_norm": 3.2123589515686035,
      "learning_rate": 8.757894736842106e-05,
      "loss": 0.1646,
      "step": 2100
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.7022438049316406,
      "learning_rate": 8.736842105263158e-05,
      "loss": 0.1502,
      "step": 2125
    },
    {
      "epoch": 0.172,
      "grad_norm": 3.6390633583068848,
      "learning_rate": 8.71578947368421e-05,
      "loss": 0.1943,
      "step": 2150
    },
    {
      "epoch": 0.174,
      "grad_norm": 2.4349706172943115,
      "learning_rate": 8.694736842105264e-05,
      "loss": 0.1619,
      "step": 2175
    },
    {
      "epoch": 0.176,
      "grad_norm": 6.100596904754639,
      "learning_rate": 8.673684210526316e-05,
      "loss": 0.1419,
      "step": 2200
    },
    {
      "epoch": 0.178,
      "grad_norm": 2.8502817153930664,
      "learning_rate": 8.652631578947369e-05,
      "loss": 0.1194,
      "step": 2225
    },
    {
      "epoch": 0.18,
      "grad_norm": 3.1560254096984863,
      "learning_rate": 8.631578947368421e-05,
      "loss": 0.1253,
      "step": 2250
    },
    {
      "epoch": 0.182,
      "grad_norm": 3.613964080810547,
      "learning_rate": 8.610526315789474e-05,
      "loss": 0.1161,
      "step": 2275
    },
    {
      "epoch": 0.184,
      "grad_norm": 2.2053143978118896,
      "learning_rate": 8.589473684210528e-05,
      "loss": 0.1303,
      "step": 2300
    },
    {
      "epoch": 0.186,
      "grad_norm": 6.676514148712158,
      "learning_rate": 8.568421052631578e-05,
      "loss": 0.2192,
      "step": 2325
    },
    {
      "epoch": 0.188,
      "grad_norm": 1.9195454120635986,
      "learning_rate": 8.547368421052632e-05,
      "loss": 0.1837,
      "step": 2350
    },
    {
      "epoch": 0.19,
      "grad_norm": 3.2146458625793457,
      "learning_rate": 8.526315789473685e-05,
      "loss": 0.1669,
      "step": 2375
    },
    {
      "epoch": 0.192,
      "grad_norm": 2.7265331745147705,
      "learning_rate": 8.505263157894737e-05,
      "loss": 0.1997,
      "step": 2400
    },
    {
      "epoch": 0.194,
      "grad_norm": 5.918653964996338,
      "learning_rate": 8.48421052631579e-05,
      "loss": 0.1934,
      "step": 2425
    },
    {
      "epoch": 0.196,
      "grad_norm": 2.6463940143585205,
      "learning_rate": 8.463157894736842e-05,
      "loss": 0.1706,
      "step": 2450
    },
    {
      "epoch": 0.198,
      "grad_norm": 2.56866455078125,
      "learning_rate": 8.442105263157896e-05,
      "loss": 0.1581,
      "step": 2475
    },
    {
      "epoch": 0.2,
      "grad_norm": 2.507850170135498,
      "learning_rate": 8.421052631578948e-05,
      "loss": 0.1593,
      "step": 2500
    },
    {
      "epoch": 0.202,
      "grad_norm": 2.009465456008911,
      "learning_rate": 8.4e-05,
      "loss": 0.1474,
      "step": 2525
    },
    {
      "epoch": 0.204,
      "grad_norm": 4.39176082611084,
      "learning_rate": 8.378947368421053e-05,
      "loss": 0.1629,
      "step": 2550
    },
    {
      "epoch": 0.206,
      "grad_norm": 3.6246728897094727,
      "learning_rate": 8.357894736842106e-05,
      "loss": 0.1717,
      "step": 2575
    },
    {
      "epoch": 0.208,
      "grad_norm": 2.2975339889526367,
      "learning_rate": 8.336842105263158e-05,
      "loss": 0.1584,
      "step": 2600
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.475097894668579,
      "learning_rate": 8.315789473684212e-05,
      "loss": 0.1474,
      "step": 2625
    },
    {
      "epoch": 0.212,
      "grad_norm": 2.2008955478668213,
      "learning_rate": 8.294736842105263e-05,
      "loss": 0.1887,
      "step": 2650
    },
    {
      "epoch": 0.214,
      "grad_norm": 3.3320555686950684,
      "learning_rate": 8.273684210526315e-05,
      "loss": 0.2136,
      "step": 2675
    },
    {
      "epoch": 0.216,
      "grad_norm": 4.016301155090332,
      "learning_rate": 8.252631578947369e-05,
      "loss": 0.1777,
      "step": 2700
    },
    {
      "epoch": 0.218,
      "grad_norm": 2.3852882385253906,
      "learning_rate": 8.231578947368421e-05,
      "loss": 0.2038,
      "step": 2725
    },
    {
      "epoch": 0.22,
      "grad_norm": 4.516410827636719,
      "learning_rate": 8.210526315789474e-05,
      "loss": 0.2191,
      "step": 2750
    },
    {
      "epoch": 0.222,
      "grad_norm": 2.775944948196411,
      "learning_rate": 8.189473684210527e-05,
      "loss": 0.239,
      "step": 2775
    },
    {
      "epoch": 0.224,
      "grad_norm": 3.0027830600738525,
      "learning_rate": 8.16842105263158e-05,
      "loss": 0.2271,
      "step": 2800
    },
    {
      "epoch": 0.226,
      "grad_norm": 3.5212559700012207,
      "learning_rate": 8.147368421052633e-05,
      "loss": 0.2025,
      "step": 2825
    },
    {
      "epoch": 0.228,
      "grad_norm": 4.692851543426514,
      "learning_rate": 8.126315789473685e-05,
      "loss": 0.1637,
      "step": 2850
    },
    {
      "epoch": 0.23,
      "grad_norm": 3.9487805366516113,
      "learning_rate": 8.105263157894737e-05,
      "loss": 0.1594,
      "step": 2875
    },
    {
      "epoch": 0.232,
      "grad_norm": 3.8476414680480957,
      "learning_rate": 8.08421052631579e-05,
      "loss": 0.2109,
      "step": 2900
    },
    {
      "epoch": 0.234,
      "grad_norm": 2.902121067047119,
      "learning_rate": 8.063157894736842e-05,
      "loss": 0.1735,
      "step": 2925
    },
    {
      "epoch": 0.236,
      "grad_norm": 3.348264694213867,
      "learning_rate": 8.042105263157895e-05,
      "loss": 0.1995,
      "step": 2950
    },
    {
      "epoch": 0.238,
      "grad_norm": 3.9797656536102295,
      "learning_rate": 8.021052631578949e-05,
      "loss": 0.1916,
      "step": 2975
    },
    {
      "epoch": 0.24,
      "grad_norm": 2.963280200958252,
      "learning_rate": 8e-05,
      "loss": 0.1941,
      "step": 3000
    },
    {
      "epoch": 0.242,
      "grad_norm": 3.173311948776245,
      "learning_rate": 7.978947368421053e-05,
      "loss": 0.1578,
      "step": 3025
    },
    {
      "epoch": 0.244,
      "grad_norm": 2.4859843254089355,
      "learning_rate": 7.957894736842106e-05,
      "loss": 0.1913,
      "step": 3050
    },
    {
      "epoch": 0.246,
      "grad_norm": 2.6782002449035645,
      "learning_rate": 7.936842105263158e-05,
      "loss": 0.1526,
      "step": 3075
    },
    {
      "epoch": 0.248,
      "grad_norm": 6.26693058013916,
      "learning_rate": 7.915789473684211e-05,
      "loss": 0.1602,
      "step": 3100
    },
    {
      "epoch": 0.25,
      "grad_norm": 2.1816000938415527,
      "learning_rate": 7.894736842105263e-05,
      "loss": 0.1652,
      "step": 3125
    },
    {
      "epoch": 0.252,
      "grad_norm": 2.8874270915985107,
      "learning_rate": 7.873684210526317e-05,
      "loss": 0.1888,
      "step": 3150
    },
    {
      "epoch": 0.254,
      "grad_norm": 2.7631444931030273,
      "learning_rate": 7.852631578947369e-05,
      "loss": 0.1484,
      "step": 3175
    },
    {
      "epoch": 0.256,
      "grad_norm": 2.1312832832336426,
      "learning_rate": 7.83157894736842e-05,
      "loss": 0.2022,
      "step": 3200
    },
    {
      "epoch": 0.258,
      "grad_norm": 3.328763723373413,
      "learning_rate": 7.810526315789474e-05,
      "loss": 0.1938,
      "step": 3225
    },
    {
      "epoch": 0.26,
      "grad_norm": 3.270240306854248,
      "learning_rate": 7.789473684210526e-05,
      "loss": 0.1671,
      "step": 3250
    },
    {
      "epoch": 0.262,
      "grad_norm": 1.2207618951797485,
      "learning_rate": 7.768421052631579e-05,
      "loss": 0.1434,
      "step": 3275
    },
    {
      "epoch": 0.264,
      "grad_norm": 2.9835286140441895,
      "learning_rate": 7.747368421052633e-05,
      "loss": 0.1609,
      "step": 3300
    },
    {
      "epoch": 0.266,
      "grad_norm": 3.5259110927581787,
      "learning_rate": 7.726315789473684e-05,
      "loss": 0.145,
      "step": 3325
    },
    {
      "epoch": 0.268,
      "grad_norm": 3.514284372329712,
      "learning_rate": 7.705263157894738e-05,
      "loss": 0.1796,
      "step": 3350
    },
    {
      "epoch": 0.27,
      "grad_norm": 3.7748453617095947,
      "learning_rate": 7.68421052631579e-05,
      "loss": 0.167,
      "step": 3375
    },
    {
      "epoch": 0.272,
      "grad_norm": 4.1441779136657715,
      "learning_rate": 7.663157894736842e-05,
      "loss": 0.2027,
      "step": 3400
    },
    {
      "epoch": 0.274,
      "grad_norm": 4.4492878913879395,
      "learning_rate": 7.642105263157895e-05,
      "loss": 0.2903,
      "step": 3425
    },
    {
      "epoch": 0.276,
      "grad_norm": 4.148222923278809,
      "learning_rate": 7.621052631578947e-05,
      "loss": 0.3515,
      "step": 3450
    },
    {
      "epoch": 0.278,
      "grad_norm": 3.1954922676086426,
      "learning_rate": 7.6e-05,
      "loss": 0.2598,
      "step": 3475
    },
    {
      "epoch": 0.28,
      "grad_norm": 2.256054401397705,
      "learning_rate": 7.578947368421054e-05,
      "loss": 0.2331,
      "step": 3500
    },
    {
      "epoch": 0.282,
      "grad_norm": 1.3954627513885498,
      "learning_rate": 7.557894736842106e-05,
      "loss": 0.2147,
      "step": 3525
    },
    {
      "epoch": 0.284,
      "grad_norm": 4.935629844665527,
      "learning_rate": 7.536842105263158e-05,
      "loss": 0.3906,
      "step": 3550
    },
    {
      "epoch": 0.286,
      "grad_norm": 3.462646245956421,
      "learning_rate": 7.515789473684211e-05,
      "loss": 0.4341,
      "step": 3575
    },
    {
      "epoch": 0.288,
      "grad_norm": 1.512658953666687,
      "learning_rate": 7.494736842105263e-05,
      "loss": 0.4142,
      "step": 3600
    },
    {
      "epoch": 0.29,
      "grad_norm": 4.282130241394043,
      "learning_rate": 7.473684210526316e-05,
      "loss": 0.3227,
      "step": 3625
    },
    {
      "epoch": 0.292,
      "grad_norm": 6.799490928649902,
      "learning_rate": 7.452631578947368e-05,
      "loss": 0.2913,
      "step": 3650
    },
    {
      "epoch": 0.294,
      "grad_norm": 2.9833290576934814,
      "learning_rate": 7.431578947368422e-05,
      "loss": 0.195,
      "step": 3675
    },
    {
      "epoch": 0.296,
      "grad_norm": 3.2424564361572266,
      "learning_rate": 7.410526315789475e-05,
      "loss": 0.2175,
      "step": 3700
    },
    {
      "epoch": 0.298,
      "grad_norm": 3.1551601886749268,
      "learning_rate": 7.389473684210527e-05,
      "loss": 0.2779,
      "step": 3725
    },
    {
      "epoch": 0.3,
      "grad_norm": 5.371734142303467,
      "learning_rate": 7.368421052631579e-05,
      "loss": 0.2768,
      "step": 3750
    },
    {
      "epoch": 0.302,
      "grad_norm": 4.353482723236084,
      "learning_rate": 7.347368421052632e-05,
      "loss": 0.227,
      "step": 3775
    },
    {
      "epoch": 0.304,
      "grad_norm": 1.761165738105774,
      "learning_rate": 7.326315789473684e-05,
      "loss": 0.2184,
      "step": 3800
    },
    {
      "epoch": 0.306,
      "grad_norm": 1.952088713645935,
      "learning_rate": 7.305263157894738e-05,
      "loss": 0.279,
      "step": 3825
    },
    {
      "epoch": 0.308,
      "grad_norm": 2.8295345306396484,
      "learning_rate": 7.28421052631579e-05,
      "loss": 0.2077,
      "step": 3850
    },
    {
      "epoch": 0.31,
      "grad_norm": 3.803010940551758,
      "learning_rate": 7.263157894736843e-05,
      "loss": 0.2618,
      "step": 3875
    },
    {
      "epoch": 0.312,
      "grad_norm": 2.9954023361206055,
      "learning_rate": 7.242105263157896e-05,
      "loss": 0.278,
      "step": 3900
    },
    {
      "epoch": 0.314,
      "grad_norm": 3.2522518634796143,
      "learning_rate": 7.221052631578947e-05,
      "loss": 0.1834,
      "step": 3925
    },
    {
      "epoch": 0.316,
      "grad_norm": 2.8170676231384277,
      "learning_rate": 7.2e-05,
      "loss": 0.2125,
      "step": 3950
    },
    {
      "epoch": 0.318,
      "grad_norm": 1.8490381240844727,
      "learning_rate": 7.178947368421054e-05,
      "loss": 0.1691,
      "step": 3975
    },
    {
      "epoch": 0.32,
      "grad_norm": 5.5511579513549805,
      "learning_rate": 7.157894736842105e-05,
      "loss": 0.1718,
      "step": 4000
    },
    {
      "epoch": 0.322,
      "grad_norm": 2.619178295135498,
      "learning_rate": 7.136842105263159e-05,
      "loss": 0.1799,
      "step": 4025
    },
    {
      "epoch": 0.324,
      "grad_norm": 3.8128204345703125,
      "learning_rate": 7.115789473684211e-05,
      "loss": 0.1771,
      "step": 4050
    },
    {
      "epoch": 0.326,
      "grad_norm": 3.535412311553955,
      "learning_rate": 7.094736842105264e-05,
      "loss": 0.1798,
      "step": 4075
    },
    {
      "epoch": 0.328,
      "grad_norm": 1.8671201467514038,
      "learning_rate": 7.073684210526316e-05,
      "loss": 0.1576,
      "step": 4100
    },
    {
      "epoch": 0.33,
      "grad_norm": 2.651156187057495,
      "learning_rate": 7.052631578947368e-05,
      "loss": 0.1653,
      "step": 4125
    },
    {
      "epoch": 0.332,
      "grad_norm": 2.2884953022003174,
      "learning_rate": 7.031578947368421e-05,
      "loss": 0.155,
      "step": 4150
    },
    {
      "epoch": 0.334,
      "grad_norm": 2.7888643741607666,
      "learning_rate": 7.010526315789473e-05,
      "loss": 0.1702,
      "step": 4175
    },
    {
      "epoch": 0.336,
      "grad_norm": 2.2567873001098633,
      "learning_rate": 6.989473684210527e-05,
      "loss": 0.1466,
      "step": 4200
    },
    {
      "epoch": 0.338,
      "grad_norm": 3.6883699893951416,
      "learning_rate": 6.96842105263158e-05,
      "loss": 0.1465,
      "step": 4225
    },
    {
      "epoch": 0.34,
      "grad_norm": 2.7656190395355225,
      "learning_rate": 6.947368421052632e-05,
      "loss": 0.1496,
      "step": 4250
    },
    {
      "epoch": 0.342,
      "grad_norm": 4.047220230102539,
      "learning_rate": 6.926315789473684e-05,
      "loss": 0.1485,
      "step": 4275
    },
    {
      "epoch": 0.344,
      "grad_norm": 2.5935311317443848,
      "learning_rate": 6.905263157894737e-05,
      "loss": 0.1641,
      "step": 4300
    },
    {
      "epoch": 0.346,
      "grad_norm": 3.29296875,
      "learning_rate": 6.88421052631579e-05,
      "loss": 0.1691,
      "step": 4325
    },
    {
      "epoch": 0.348,
      "grad_norm": 1.8748677968978882,
      "learning_rate": 6.863157894736843e-05,
      "loss": 0.1564,
      "step": 4350
    },
    {
      "epoch": 0.35,
      "grad_norm": 2.8620316982269287,
      "learning_rate": 6.842105263157895e-05,
      "loss": 0.1977,
      "step": 4375
    },
    {
      "epoch": 0.352,
      "grad_norm": 3.427701711654663,
      "learning_rate": 6.821052631578948e-05,
      "loss": 0.1699,
      "step": 4400
    },
    {
      "epoch": 0.354,
      "grad_norm": 2.8630993366241455,
      "learning_rate": 6.800000000000001e-05,
      "loss": 0.1841,
      "step": 4425
    },
    {
      "epoch": 0.356,
      "grad_norm": 1.8674124479293823,
      "learning_rate": 6.778947368421052e-05,
      "loss": 0.1545,
      "step": 4450
    },
    {
      "epoch": 0.358,
      "grad_norm": 2.2906653881073,
      "learning_rate": 6.757894736842105e-05,
      "loss": 0.155,
      "step": 4475
    },
    {
      "epoch": 0.36,
      "grad_norm": 2.981401205062866,
      "learning_rate": 6.736842105263159e-05,
      "loss": 0.1604,
      "step": 4500
    },
    {
      "epoch": 0.362,
      "grad_norm": 3.040834903717041,
      "learning_rate": 6.71578947368421e-05,
      "loss": 0.153,
      "step": 4525
    },
    {
      "epoch": 0.364,
      "grad_norm": 2.600574493408203,
      "learning_rate": 6.694736842105264e-05,
      "loss": 0.1469,
      "step": 4550
    },
    {
      "epoch": 0.366,
      "grad_norm": 2.2893636226654053,
      "learning_rate": 6.673684210526316e-05,
      "loss": 0.1501,
      "step": 4575
    },
    {
      "epoch": 0.368,
      "grad_norm": 6.196401119232178,
      "learning_rate": 6.652631578947369e-05,
      "loss": 0.1561,
      "step": 4600
    },
    {
      "epoch": 0.37,
      "grad_norm": 3.3594653606414795,
      "learning_rate": 6.631578947368421e-05,
      "loss": 0.1757,
      "step": 4625
    },
    {
      "epoch": 0.372,
      "grad_norm": 2.437387704849243,
      "learning_rate": 6.610526315789473e-05,
      "loss": 0.1365,
      "step": 4650
    },
    {
      "epoch": 0.374,
      "grad_norm": 3.190577507019043,
      "learning_rate": 6.589473684210526e-05,
      "loss": 0.168,
      "step": 4675
    },
    {
      "epoch": 0.376,
      "grad_norm": 5.2853102684021,
      "learning_rate": 6.56842105263158e-05,
      "loss": 0.1422,
      "step": 4700
    },
    {
      "epoch": 0.378,
      "grad_norm": 4.163969039916992,
      "learning_rate": 6.547368421052632e-05,
      "loss": 0.1391,
      "step": 4725
    },
    {
      "epoch": 0.38,
      "grad_norm": 3.255009889602661,
      "learning_rate": 6.526315789473685e-05,
      "loss": 0.1851,
      "step": 4750
    },
    {
      "epoch": 0.382,
      "grad_norm": 1.9225974082946777,
      "learning_rate": 6.505263157894737e-05,
      "loss": 0.1688,
      "step": 4775
    },
    {
      "epoch": 0.384,
      "grad_norm": 2.5862815380096436,
      "learning_rate": 6.484210526315789e-05,
      "loss": 0.1883,
      "step": 4800
    },
    {
      "epoch": 0.386,
      "grad_norm": 2.3600857257843018,
      "learning_rate": 6.463157894736842e-05,
      "loss": 0.1584,
      "step": 4825
    },
    {
      "epoch": 0.388,
      "grad_norm": 4.694042682647705,
      "learning_rate": 6.442105263157894e-05,
      "loss": 0.1621,
      "step": 4850
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.5520076751708984,
      "learning_rate": 6.421052631578948e-05,
      "loss": 0.1859,
      "step": 4875
    },
    {
      "epoch": 0.392,
      "grad_norm": 2.522648572921753,
      "learning_rate": 6.400000000000001e-05,
      "loss": 0.1549,
      "step": 4900
    },
    {
      "epoch": 0.394,
      "grad_norm": 2.736837863922119,
      "learning_rate": 6.378947368421053e-05,
      "loss": 0.1444,
      "step": 4925
    },
    {
      "epoch": 0.396,
      "grad_norm": 2.710667371749878,
      "learning_rate": 6.357894736842106e-05,
      "loss": 0.1589,
      "step": 4950
    },
    {
      "epoch": 0.398,
      "grad_norm": 2.561551570892334,
      "learning_rate": 6.336842105263158e-05,
      "loss": 0.1603,
      "step": 4975
    },
    {
      "epoch": 0.4,
      "grad_norm": 2.2533857822418213,
      "learning_rate": 6.31578947368421e-05,
      "loss": 0.1547,
      "step": 5000
    },
    {
      "epoch": 0.402,
      "grad_norm": 4.293575286865234,
      "learning_rate": 6.294736842105264e-05,
      "loss": 0.1649,
      "step": 5025
    },
    {
      "epoch": 0.404,
      "grad_norm": 2.0180907249450684,
      "learning_rate": 6.273684210526316e-05,
      "loss": 0.1542,
      "step": 5050
    },
    {
      "epoch": 0.406,
      "grad_norm": 4.162642955780029,
      "learning_rate": 6.252631578947369e-05,
      "loss": 0.1474,
      "step": 5075
    },
    {
      "epoch": 0.408,
      "grad_norm": 2.2592241764068604,
      "learning_rate": 6.231578947368422e-05,
      "loss": 0.1424,
      "step": 5100
    },
    {
      "epoch": 0.41,
      "grad_norm": 2.9657979011535645,
      "learning_rate": 6.210526315789474e-05,
      "loss": 0.141,
      "step": 5125
    },
    {
      "epoch": 0.412,
      "grad_norm": 4.562716960906982,
      "learning_rate": 6.189473684210526e-05,
      "loss": 0.1207,
      "step": 5150
    },
    {
      "epoch": 0.414,
      "grad_norm": 3.5740582942962646,
      "learning_rate": 6.168421052631578e-05,
      "loss": 0.1348,
      "step": 5175
    },
    {
      "epoch": 0.416,
      "grad_norm": 2.933711290359497,
      "learning_rate": 6.147368421052632e-05,
      "loss": 0.1262,
      "step": 5200
    },
    {
      "epoch": 0.418,
      "grad_norm": 2.674757480621338,
      "learning_rate": 6.126315789473685e-05,
      "loss": 0.1639,
      "step": 5225
    },
    {
      "epoch": 0.42,
      "grad_norm": 3.187509775161743,
      "learning_rate": 6.105263157894737e-05,
      "loss": 0.1486,
      "step": 5250
    },
    {
      "epoch": 0.422,
      "grad_norm": 2.763331174850464,
      "learning_rate": 6.08421052631579e-05,
      "loss": 0.1487,
      "step": 5275
    },
    {
      "epoch": 0.424,
      "grad_norm": 2.597900867462158,
      "learning_rate": 6.063157894736843e-05,
      "loss": 0.1482,
      "step": 5300
    },
    {
      "epoch": 0.426,
      "grad_norm": 2.390775203704834,
      "learning_rate": 6.042105263157894e-05,
      "loss": 0.1405,
      "step": 5325
    },
    {
      "epoch": 0.428,
      "grad_norm": 4.166103363037109,
      "learning_rate": 6.0210526315789475e-05,
      "loss": 0.1822,
      "step": 5350
    },
    {
      "epoch": 0.43,
      "grad_norm": 2.066412925720215,
      "learning_rate": 6e-05,
      "loss": 0.1272,
      "step": 5375
    },
    {
      "epoch": 0.432,
      "grad_norm": 3.385010242462158,
      "learning_rate": 5.978947368421053e-05,
      "loss": 0.1576,
      "step": 5400
    },
    {
      "epoch": 0.434,
      "grad_norm": 3.9283525943756104,
      "learning_rate": 5.9578947368421055e-05,
      "loss": 0.1804,
      "step": 5425
    },
    {
      "epoch": 0.436,
      "grad_norm": 1.7260828018188477,
      "learning_rate": 5.936842105263158e-05,
      "loss": 0.1547,
      "step": 5450
    },
    {
      "epoch": 0.438,
      "grad_norm": 1.3694740533828735,
      "learning_rate": 5.9157894736842114e-05,
      "loss": 0.1684,
      "step": 5475
    },
    {
      "epoch": 0.44,
      "grad_norm": 3.547086000442505,
      "learning_rate": 5.894736842105263e-05,
      "loss": 0.1731,
      "step": 5500
    },
    {
      "epoch": 0.442,
      "grad_norm": 4.220787048339844,
      "learning_rate": 5.8736842105263154e-05,
      "loss": 0.1502,
      "step": 5525
    },
    {
      "epoch": 0.444,
      "grad_norm": 5.9439239501953125,
      "learning_rate": 5.852631578947369e-05,
      "loss": 0.1775,
      "step": 5550
    },
    {
      "epoch": 0.446,
      "grad_norm": 2.953634738922119,
      "learning_rate": 5.8315789473684214e-05,
      "loss": 0.1577,
      "step": 5575
    },
    {
      "epoch": 0.448,
      "grad_norm": 3.257310628890991,
      "learning_rate": 5.810526315789474e-05,
      "loss": 0.2024,
      "step": 5600
    },
    {
      "epoch": 0.45,
      "grad_norm": 2.815178394317627,
      "learning_rate": 5.789473684210527e-05,
      "loss": 0.1858,
      "step": 5625
    },
    {
      "epoch": 0.452,
      "grad_norm": 4.573605537414551,
      "learning_rate": 5.7684210526315794e-05,
      "loss": 0.2153,
      "step": 5650
    },
    {
      "epoch": 0.454,
      "grad_norm": 1.737714409828186,
      "learning_rate": 5.747368421052633e-05,
      "loss": 0.1809,
      "step": 5675
    },
    {
      "epoch": 0.456,
      "grad_norm": 2.6098313331604004,
      "learning_rate": 5.726315789473684e-05,
      "loss": 0.166,
      "step": 5700
    },
    {
      "epoch": 0.458,
      "grad_norm": 1.715842843055725,
      "learning_rate": 5.7052631578947366e-05,
      "loss": 0.1662,
      "step": 5725
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.6159685850143433,
      "learning_rate": 5.68421052631579e-05,
      "loss": 0.1443,
      "step": 5750
    },
    {
      "epoch": 0.462,
      "grad_norm": 2.3512282371520996,
      "learning_rate": 5.6631578947368426e-05,
      "loss": 0.1672,
      "step": 5775
    },
    {
      "epoch": 0.464,
      "grad_norm": 2.50100040435791,
      "learning_rate": 5.642105263157895e-05,
      "loss": 0.1596,
      "step": 5800
    },
    {
      "epoch": 0.466,
      "grad_norm": 3.3251962661743164,
      "learning_rate": 5.621052631578948e-05,
      "loss": 0.1914,
      "step": 5825
    },
    {
      "epoch": 0.468,
      "grad_norm": 2.4758493900299072,
      "learning_rate": 5.6000000000000006e-05,
      "loss": 0.1763,
      "step": 5850
    },
    {
      "epoch": 0.47,
      "grad_norm": 2.60750150680542,
      "learning_rate": 5.5789473684210526e-05,
      "loss": 0.163,
      "step": 5875
    },
    {
      "epoch": 0.472,
      "grad_norm": 2.7172982692718506,
      "learning_rate": 5.557894736842105e-05,
      "loss": 0.2162,
      "step": 5900
    },
    {
      "epoch": 0.474,
      "grad_norm": 2.4564871788024902,
      "learning_rate": 5.536842105263158e-05,
      "loss": 0.2331,
      "step": 5925
    },
    {
      "epoch": 0.476,
      "grad_norm": 4.060207843780518,
      "learning_rate": 5.5157894736842105e-05,
      "loss": 0.1686,
      "step": 5950
    },
    {
      "epoch": 0.478,
      "grad_norm": 4.7515764236450195,
      "learning_rate": 5.494736842105264e-05,
      "loss": 0.2329,
      "step": 5975
    },
    {
      "epoch": 0.48,
      "grad_norm": 4.869494915008545,
      "learning_rate": 5.4736842105263165e-05,
      "loss": 0.2139,
      "step": 6000
    },
    {
      "epoch": 0.482,
      "grad_norm": 4.088851451873779,
      "learning_rate": 5.452631578947369e-05,
      "loss": 0.2784,
      "step": 6025
    },
    {
      "epoch": 0.484,
      "grad_norm": 3.6407222747802734,
      "learning_rate": 5.431578947368421e-05,
      "loss": 0.2477,
      "step": 6050
    },
    {
      "epoch": 0.486,
      "grad_norm": 3.3714334964752197,
      "learning_rate": 5.410526315789474e-05,
      "loss": 0.2195,
      "step": 6075
    },
    {
      "epoch": 0.488,
      "grad_norm": 3.871067523956299,
      "learning_rate": 5.3894736842105265e-05,
      "loss": 0.2696,
      "step": 6100
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.533532738685608,
      "learning_rate": 5.368421052631579e-05,
      "loss": 0.2131,
      "step": 6125
    },
    {
      "epoch": 0.492,
      "grad_norm": 4.925436973571777,
      "learning_rate": 5.347368421052632e-05,
      "loss": 0.2229,
      "step": 6150
    },
    {
      "epoch": 0.494,
      "grad_norm": 3.861382484436035,
      "learning_rate": 5.326315789473685e-05,
      "loss": 0.1841,
      "step": 6175
    },
    {
      "epoch": 0.496,
      "grad_norm": 5.30626106262207,
      "learning_rate": 5.305263157894738e-05,
      "loss": 0.2253,
      "step": 6200
    },
    {
      "epoch": 0.498,
      "grad_norm": 1.9890210628509521,
      "learning_rate": 5.284210526315789e-05,
      "loss": 0.2291,
      "step": 6225
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.197603464126587,
      "learning_rate": 5.2631578947368424e-05,
      "loss": 0.1751,
      "step": 6250
    },
    {
      "epoch": 0.502,
      "grad_norm": 2.809992551803589,
      "learning_rate": 5.242105263157895e-05,
      "loss": 0.246,
      "step": 6275
    },
    {
      "epoch": 0.504,
      "grad_norm": 3.991882801055908,
      "learning_rate": 5.221052631578948e-05,
      "loss": 0.1961,
      "step": 6300
    },
    {
      "epoch": 0.506,
      "grad_norm": 4.238216876983643,
      "learning_rate": 5.2000000000000004e-05,
      "loss": 0.1843,
      "step": 6325
    },
    {
      "epoch": 0.508,
      "grad_norm": 1.8953478336334229,
      "learning_rate": 5.178947368421053e-05,
      "loss": 0.2101,
      "step": 6350
    },
    {
      "epoch": 0.51,
      "grad_norm": 2.997884750366211,
      "learning_rate": 5.157894736842106e-05,
      "loss": 0.2066,
      "step": 6375
    },
    {
      "epoch": 0.512,
      "grad_norm": 3.8943450450897217,
      "learning_rate": 5.1368421052631576e-05,
      "loss": 0.2258,
      "step": 6400
    },
    {
      "epoch": 0.514,
      "grad_norm": 3.5741677284240723,
      "learning_rate": 5.11578947368421e-05,
      "loss": 0.2189,
      "step": 6425
    },
    {
      "epoch": 0.516,
      "grad_norm": 1.6271083354949951,
      "learning_rate": 5.094736842105263e-05,
      "loss": 0.1745,
      "step": 6450
    },
    {
      "epoch": 0.518,
      "grad_norm": 1.796264886856079,
      "learning_rate": 5.073684210526316e-05,
      "loss": 0.1925,
      "step": 6475
    },
    {
      "epoch": 0.52,
      "grad_norm": 3.7582969665527344,
      "learning_rate": 5.052631578947369e-05,
      "loss": 0.2079,
      "step": 6500
    },
    {
      "epoch": 0.522,
      "grad_norm": 1.788479208946228,
      "learning_rate": 5.0315789473684216e-05,
      "loss": 0.1729,
      "step": 6525
    },
    {
      "epoch": 0.524,
      "grad_norm": 1.6168099641799927,
      "learning_rate": 5.010526315789474e-05,
      "loss": 0.1717,
      "step": 6550
    },
    {
      "epoch": 0.526,
      "grad_norm": 2.9334585666656494,
      "learning_rate": 4.989473684210527e-05,
      "loss": 0.1446,
      "step": 6575
    },
    {
      "epoch": 0.528,
      "grad_norm": 2.1925525665283203,
      "learning_rate": 4.9684210526315796e-05,
      "loss": 0.1478,
      "step": 6600
    },
    {
      "epoch": 0.53,
      "grad_norm": 2.6039555072784424,
      "learning_rate": 4.9473684210526315e-05,
      "loss": 0.1736,
      "step": 6625
    },
    {
      "epoch": 0.532,
      "grad_norm": 2.0845723152160645,
      "learning_rate": 4.926315789473684e-05,
      "loss": 0.1874,
      "step": 6650
    },
    {
      "epoch": 0.534,
      "grad_norm": 3.34181547164917,
      "learning_rate": 4.9052631578947375e-05,
      "loss": 0.1644,
      "step": 6675
    },
    {
      "epoch": 0.536,
      "grad_norm": 2.063713550567627,
      "learning_rate": 4.8842105263157895e-05,
      "loss": 0.1764,
      "step": 6700
    },
    {
      "epoch": 0.538,
      "grad_norm": 2.799060583114624,
      "learning_rate": 4.863157894736842e-05,
      "loss": 0.1868,
      "step": 6725
    },
    {
      "epoch": 0.54,
      "grad_norm": 2.2900466918945312,
      "learning_rate": 4.842105263157895e-05,
      "loss": 0.2066,
      "step": 6750
    },
    {
      "epoch": 0.542,
      "grad_norm": 3.7162959575653076,
      "learning_rate": 4.821052631578948e-05,
      "loss": 0.1997,
      "step": 6775
    },
    {
      "epoch": 0.544,
      "grad_norm": 3.615978956222534,
      "learning_rate": 4.8e-05,
      "loss": 0.2407,
      "step": 6800
    },
    {
      "epoch": 0.546,
      "grad_norm": 2.926126003265381,
      "learning_rate": 4.778947368421053e-05,
      "loss": 0.1995,
      "step": 6825
    },
    {
      "epoch": 0.548,
      "grad_norm": 3.1077892780303955,
      "learning_rate": 4.7578947368421054e-05,
      "loss": 0.1562,
      "step": 6850
    },
    {
      "epoch": 0.55,
      "grad_norm": 2.667742967605591,
      "learning_rate": 4.736842105263158e-05,
      "loss": 0.195,
      "step": 6875
    },
    {
      "epoch": 0.552,
      "grad_norm": 3.075545310974121,
      "learning_rate": 4.715789473684211e-05,
      "loss": 0.1747,
      "step": 6900
    },
    {
      "epoch": 0.554,
      "grad_norm": 5.871755123138428,
      "learning_rate": 4.6947368421052634e-05,
      "loss": 0.2339,
      "step": 6925
    },
    {
      "epoch": 0.556,
      "grad_norm": 1.96011221408844,
      "learning_rate": 4.673684210526316e-05,
      "loss": 0.1731,
      "step": 6950
    },
    {
      "epoch": 0.558,
      "grad_norm": 2.2573795318603516,
      "learning_rate": 4.652631578947369e-05,
      "loss": 0.1752,
      "step": 6975
    },
    {
      "epoch": 0.56,
      "grad_norm": 4.410872936248779,
      "learning_rate": 4.6315789473684214e-05,
      "loss": 0.1833,
      "step": 7000
    },
    {
      "epoch": 0.562,
      "grad_norm": 2.1633858680725098,
      "learning_rate": 4.610526315789474e-05,
      "loss": 0.1678,
      "step": 7025
    },
    {
      "epoch": 0.564,
      "grad_norm": 2.8176636695861816,
      "learning_rate": 4.589473684210526e-05,
      "loss": 0.2261,
      "step": 7050
    },
    {
      "epoch": 0.566,
      "grad_norm": 4.914159297943115,
      "learning_rate": 4.568421052631579e-05,
      "loss": 0.2307,
      "step": 7075
    },
    {
      "epoch": 0.568,
      "grad_norm": 3.0818915367126465,
      "learning_rate": 4.547368421052632e-05,
      "loss": 0.3216,
      "step": 7100
    },
    {
      "epoch": 0.57,
      "grad_norm": 2.862865447998047,
      "learning_rate": 4.5263157894736846e-05,
      "loss": 0.2771,
      "step": 7125
    },
    {
      "epoch": 0.572,
      "grad_norm": 3.212938070297241,
      "learning_rate": 4.5052631578947366e-05,
      "loss": 0.2252,
      "step": 7150
    },
    {
      "epoch": 0.574,
      "grad_norm": 5.601623058319092,
      "learning_rate": 4.48421052631579e-05,
      "loss": 0.2069,
      "step": 7175
    },
    {
      "epoch": 0.576,
      "grad_norm": 2.006655693054199,
      "learning_rate": 4.4631578947368426e-05,
      "loss": 0.2076,
      "step": 7200
    },
    {
      "epoch": 0.578,
      "grad_norm": 5.423392295837402,
      "learning_rate": 4.442105263157895e-05,
      "loss": 0.1807,
      "step": 7225
    },
    {
      "epoch": 0.58,
      "grad_norm": 2.1953580379486084,
      "learning_rate": 4.421052631578947e-05,
      "loss": 0.1854,
      "step": 7250
    },
    {
      "epoch": 0.582,
      "grad_norm": 2.8728268146514893,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.2083,
      "step": 7275
    },
    {
      "epoch": 0.584,
      "grad_norm": 1.9604209661483765,
      "learning_rate": 4.378947368421053e-05,
      "loss": 0.1715,
      "step": 7300
    },
    {
      "epoch": 0.586,
      "grad_norm": 2.668545961380005,
      "learning_rate": 4.357894736842105e-05,
      "loss": 0.1471,
      "step": 7325
    },
    {
      "epoch": 0.588,
      "grad_norm": 1.3440651893615723,
      "learning_rate": 4.336842105263158e-05,
      "loss": 0.1729,
      "step": 7350
    },
    {
      "epoch": 0.59,
      "grad_norm": 2.6942169666290283,
      "learning_rate": 4.3157894736842105e-05,
      "loss": 0.2032,
      "step": 7375
    },
    {
      "epoch": 0.592,
      "grad_norm": 2.4484572410583496,
      "learning_rate": 4.294736842105264e-05,
      "loss": 0.2064,
      "step": 7400
    },
    {
      "epoch": 0.594,
      "grad_norm": 3.1453819274902344,
      "learning_rate": 4.273684210526316e-05,
      "loss": 0.2224,
      "step": 7425
    },
    {
      "epoch": 0.596,
      "grad_norm": 2.246695041656494,
      "learning_rate": 4.2526315789473685e-05,
      "loss": 0.1716,
      "step": 7450
    },
    {
      "epoch": 0.598,
      "grad_norm": 3.6021037101745605,
      "learning_rate": 4.231578947368421e-05,
      "loss": 0.1618,
      "step": 7475
    },
    {
      "epoch": 0.6,
      "grad_norm": 2.7792000770568848,
      "learning_rate": 4.210526315789474e-05,
      "loss": 0.1398,
      "step": 7500
    },
    {
      "epoch": 0.602,
      "grad_norm": 3.792266845703125,
      "learning_rate": 4.1894736842105264e-05,
      "loss": 0.1491,
      "step": 7525
    },
    {
      "epoch": 0.604,
      "grad_norm": 3.832718849182129,
      "learning_rate": 4.168421052631579e-05,
      "loss": 0.1734,
      "step": 7550
    },
    {
      "epoch": 0.606,
      "grad_norm": 5.681286334991455,
      "learning_rate": 4.147368421052632e-05,
      "loss": 0.2279,
      "step": 7575
    },
    {
      "epoch": 0.608,
      "grad_norm": 3.5109989643096924,
      "learning_rate": 4.1263157894736844e-05,
      "loss": 0.2148,
      "step": 7600
    },
    {
      "epoch": 0.61,
      "grad_norm": 2.3420827388763428,
      "learning_rate": 4.105263157894737e-05,
      "loss": 0.2353,
      "step": 7625
    },
    {
      "epoch": 0.612,
      "grad_norm": 4.0047407150268555,
      "learning_rate": 4.08421052631579e-05,
      "loss": 0.2745,
      "step": 7650
    },
    {
      "epoch": 0.614,
      "grad_norm": 2.9700982570648193,
      "learning_rate": 4.0631578947368424e-05,
      "loss": 0.2061,
      "step": 7675
    },
    {
      "epoch": 0.616,
      "grad_norm": 1.1286704540252686,
      "learning_rate": 4.042105263157895e-05,
      "loss": 0.2002,
      "step": 7700
    },
    {
      "epoch": 0.618,
      "grad_norm": 3.2874717712402344,
      "learning_rate": 4.021052631578948e-05,
      "loss": 0.1782,
      "step": 7725
    },
    {
      "epoch": 0.62,
      "grad_norm": 2.102379560470581,
      "learning_rate": 4e-05,
      "loss": 0.1683,
      "step": 7750
    },
    {
      "epoch": 0.622,
      "grad_norm": 2.282027244567871,
      "learning_rate": 3.978947368421053e-05,
      "loss": 0.16,
      "step": 7775
    },
    {
      "epoch": 0.624,
      "grad_norm": 3.09382963180542,
      "learning_rate": 3.9578947368421056e-05,
      "loss": 0.183,
      "step": 7800
    },
    {
      "epoch": 0.626,
      "grad_norm": 2.2616071701049805,
      "learning_rate": 3.936842105263158e-05,
      "loss": 0.1696,
      "step": 7825
    },
    {
      "epoch": 0.628,
      "grad_norm": 2.725660562515259,
      "learning_rate": 3.91578947368421e-05,
      "loss": 0.1798,
      "step": 7850
    },
    {
      "epoch": 0.63,
      "grad_norm": 3.027191400527954,
      "learning_rate": 3.894736842105263e-05,
      "loss": 0.1923,
      "step": 7875
    },
    {
      "epoch": 0.632,
      "grad_norm": 1.982648253440857,
      "learning_rate": 3.873684210526316e-05,
      "loss": 0.2375,
      "step": 7900
    },
    {
      "epoch": 0.634,
      "grad_norm": 1.9510787725448608,
      "learning_rate": 3.852631578947369e-05,
      "loss": 0.2464,
      "step": 7925
    },
    {
      "epoch": 0.636,
      "grad_norm": 2.4144234657287598,
      "learning_rate": 3.831578947368421e-05,
      "loss": 0.2646,
      "step": 7950
    },
    {
      "epoch": 0.638,
      "grad_norm": 2.5756216049194336,
      "learning_rate": 3.8105263157894735e-05,
      "loss": 0.2361,
      "step": 7975
    },
    {
      "epoch": 0.64,
      "grad_norm": 2.4998459815979004,
      "learning_rate": 3.789473684210527e-05,
      "loss": 0.2325,
      "step": 8000
    },
    {
      "epoch": 0.642,
      "grad_norm": 2.4527933597564697,
      "learning_rate": 3.768421052631579e-05,
      "loss": 0.2023,
      "step": 8025
    },
    {
      "epoch": 0.644,
      "grad_norm": 3.2145981788635254,
      "learning_rate": 3.7473684210526315e-05,
      "loss": 0.1794,
      "step": 8050
    },
    {
      "epoch": 0.646,
      "grad_norm": 3.4155004024505615,
      "learning_rate": 3.726315789473684e-05,
      "loss": 0.1793,
      "step": 8075
    },
    {
      "epoch": 0.648,
      "grad_norm": 1.3707998991012573,
      "learning_rate": 3.7052631578947375e-05,
      "loss": 0.145,
      "step": 8100
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.991523265838623,
      "learning_rate": 3.6842105263157895e-05,
      "loss": 0.1993,
      "step": 8125
    },
    {
      "epoch": 0.652,
      "grad_norm": 1.9854768514633179,
      "learning_rate": 3.663157894736842e-05,
      "loss": 0.1834,
      "step": 8150
    },
    {
      "epoch": 0.654,
      "grad_norm": 2.564042806625366,
      "learning_rate": 3.642105263157895e-05,
      "loss": 0.1703,
      "step": 8175
    },
    {
      "epoch": 0.656,
      "grad_norm": 2.2101786136627197,
      "learning_rate": 3.621052631578948e-05,
      "loss": 0.148,
      "step": 8200
    },
    {
      "epoch": 0.658,
      "grad_norm": 4.551767826080322,
      "learning_rate": 3.6e-05,
      "loss": 0.1797,
      "step": 8225
    },
    {
      "epoch": 0.66,
      "grad_norm": 2.3506393432617188,
      "learning_rate": 3.578947368421053e-05,
      "loss": 0.2005,
      "step": 8250
    },
    {
      "epoch": 0.662,
      "grad_norm": 2.509979486465454,
      "learning_rate": 3.5578947368421054e-05,
      "loss": 0.1593,
      "step": 8275
    },
    {
      "epoch": 0.664,
      "grad_norm": 2.1514201164245605,
      "learning_rate": 3.536842105263158e-05,
      "loss": 0.1605,
      "step": 8300
    },
    {
      "epoch": 0.666,
      "grad_norm": 1.8447412252426147,
      "learning_rate": 3.515789473684211e-05,
      "loss": 0.1631,
      "step": 8325
    },
    {
      "epoch": 0.668,
      "grad_norm": 2.2108616828918457,
      "learning_rate": 3.4947368421052634e-05,
      "loss": 0.1812,
      "step": 8350
    },
    {
      "epoch": 0.67,
      "grad_norm": 3.138340473175049,
      "learning_rate": 3.473684210526316e-05,
      "loss": 0.1535,
      "step": 8375
    },
    {
      "epoch": 0.672,
      "grad_norm": 2.089402437210083,
      "learning_rate": 3.452631578947369e-05,
      "loss": 0.1882,
      "step": 8400
    },
    {
      "epoch": 0.674,
      "grad_norm": 2.022799253463745,
      "learning_rate": 3.431578947368421e-05,
      "loss": 0.1534,
      "step": 8425
    },
    {
      "epoch": 0.676,
      "grad_norm": 4.402505874633789,
      "learning_rate": 3.410526315789474e-05,
      "loss": 0.211,
      "step": 8450
    },
    {
      "epoch": 0.678,
      "grad_norm": 2.2280025482177734,
      "learning_rate": 3.389473684210526e-05,
      "loss": 0.2052,
      "step": 8475
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.9944862127304077,
      "learning_rate": 3.368421052631579e-05,
      "loss": 0.1431,
      "step": 8500
    },
    {
      "epoch": 0.682,
      "grad_norm": 3.152775526046753,
      "learning_rate": 3.347368421052632e-05,
      "loss": 0.1697,
      "step": 8525
    },
    {
      "epoch": 0.684,
      "grad_norm": 4.06758975982666,
      "learning_rate": 3.3263157894736846e-05,
      "loss": 0.1894,
      "step": 8550
    },
    {
      "epoch": 0.686,
      "grad_norm": 4.115979194641113,
      "learning_rate": 3.3052631578947366e-05,
      "loss": 0.1746,
      "step": 8575
    },
    {
      "epoch": 0.688,
      "grad_norm": 1.844355821609497,
      "learning_rate": 3.28421052631579e-05,
      "loss": 0.1786,
      "step": 8600
    },
    {
      "epoch": 0.69,
      "grad_norm": 2.8369998931884766,
      "learning_rate": 3.2631578947368426e-05,
      "loss": 0.1718,
      "step": 8625
    },
    {
      "epoch": 0.692,
      "grad_norm": 4.434097766876221,
      "learning_rate": 3.2421052631578945e-05,
      "loss": 0.1707,
      "step": 8650
    },
    {
      "epoch": 0.694,
      "grad_norm": 3.360795497894287,
      "learning_rate": 3.221052631578947e-05,
      "loss": 0.1865,
      "step": 8675
    },
    {
      "epoch": 0.696,
      "grad_norm": 2.3571205139160156,
      "learning_rate": 3.2000000000000005e-05,
      "loss": 0.1922,
      "step": 8700
    },
    {
      "epoch": 0.698,
      "grad_norm": 2.478071451187134,
      "learning_rate": 3.178947368421053e-05,
      "loss": 0.2102,
      "step": 8725
    },
    {
      "epoch": 0.7,
      "grad_norm": 5.430673599243164,
      "learning_rate": 3.157894736842105e-05,
      "loss": 0.2257,
      "step": 8750
    },
    {
      "epoch": 0.702,
      "grad_norm": 3.069105625152588,
      "learning_rate": 3.136842105263158e-05,
      "loss": 0.2041,
      "step": 8775
    },
    {
      "epoch": 0.704,
      "grad_norm": 5.354055404663086,
      "learning_rate": 3.115789473684211e-05,
      "loss": 0.3051,
      "step": 8800
    },
    {
      "epoch": 0.706,
      "grad_norm": 2.466470956802368,
      "learning_rate": 3.094736842105263e-05,
      "loss": 0.3592,
      "step": 8825
    },
    {
      "epoch": 0.708,
      "grad_norm": 3.638484477996826,
      "learning_rate": 3.073684210526316e-05,
      "loss": 0.321,
      "step": 8850
    },
    {
      "epoch": 0.71,
      "grad_norm": 4.691796779632568,
      "learning_rate": 3.0526315789473684e-05,
      "loss": 0.3434,
      "step": 8875
    },
    {
      "epoch": 0.712,
      "grad_norm": 2.808640718460083,
      "learning_rate": 3.0315789473684214e-05,
      "loss": 0.2302,
      "step": 8900
    },
    {
      "epoch": 0.714,
      "grad_norm": 2.936375617980957,
      "learning_rate": 3.0105263157894737e-05,
      "loss": 0.2645,
      "step": 8925
    },
    {
      "epoch": 0.716,
      "grad_norm": 4.677466869354248,
      "learning_rate": 2.9894736842105264e-05,
      "loss": 0.3248,
      "step": 8950
    },
    {
      "epoch": 0.718,
      "grad_norm": 7.0504584312438965,
      "learning_rate": 2.968421052631579e-05,
      "loss": 0.3511,
      "step": 8975
    },
    {
      "epoch": 0.72,
      "grad_norm": 3.50658917427063,
      "learning_rate": 2.9473684210526314e-05,
      "loss": 0.3516,
      "step": 9000
    },
    {
      "epoch": 0.722,
      "grad_norm": 3.8559470176696777,
      "learning_rate": 2.9263157894736844e-05,
      "loss": 0.3557,
      "step": 9025
    },
    {
      "epoch": 0.724,
      "grad_norm": 9.091622352600098,
      "learning_rate": 2.905263157894737e-05,
      "loss": 0.3588,
      "step": 9050
    },
    {
      "epoch": 0.726,
      "grad_norm": 4.84503173828125,
      "learning_rate": 2.8842105263157897e-05,
      "loss": 0.3264,
      "step": 9075
    },
    {
      "epoch": 0.728,
      "grad_norm": 2.985128164291382,
      "learning_rate": 2.863157894736842e-05,
      "loss": 0.3308,
      "step": 9100
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.7857965230941772,
      "learning_rate": 2.842105263157895e-05,
      "loss": 0.3014,
      "step": 9125
    },
    {
      "epoch": 0.732,
      "grad_norm": 3.808424472808838,
      "learning_rate": 2.8210526315789476e-05,
      "loss": 0.3639,
      "step": 9150
    },
    {
      "epoch": 0.734,
      "grad_norm": 5.9339070320129395,
      "learning_rate": 2.8000000000000003e-05,
      "loss": 0.3825,
      "step": 9175
    },
    {
      "epoch": 0.736,
      "grad_norm": 3.0576109886169434,
      "learning_rate": 2.7789473684210526e-05,
      "loss": 0.3629,
      "step": 9200
    },
    {
      "epoch": 0.738,
      "grad_norm": 3.745070219039917,
      "learning_rate": 2.7578947368421053e-05,
      "loss": 0.343,
      "step": 9225
    },
    {
      "epoch": 0.74,
      "grad_norm": 4.298427104949951,
      "learning_rate": 2.7368421052631583e-05,
      "loss": 0.3701,
      "step": 9250
    },
    {
      "epoch": 0.742,
      "grad_norm": 4.324175834655762,
      "learning_rate": 2.7157894736842106e-05,
      "loss": 0.3727,
      "step": 9275
    },
    {
      "epoch": 0.744,
      "grad_norm": 3.3111040592193604,
      "learning_rate": 2.6947368421052632e-05,
      "loss": 0.4209,
      "step": 9300
    },
    {
      "epoch": 0.746,
      "grad_norm": 5.95839786529541,
      "learning_rate": 2.673684210526316e-05,
      "loss": 0.3929,
      "step": 9325
    },
    {
      "epoch": 0.748,
      "grad_norm": 3.3262205123901367,
      "learning_rate": 2.652631578947369e-05,
      "loss": 0.3867,
      "step": 9350
    },
    {
      "epoch": 0.75,
      "grad_norm": 7.750339508056641,
      "learning_rate": 2.6315789473684212e-05,
      "loss": 0.4084,
      "step": 9375
    },
    {
      "epoch": 0.752,
      "grad_norm": 6.184704303741455,
      "learning_rate": 2.610526315789474e-05,
      "loss": 0.3948,
      "step": 9400
    },
    {
      "epoch": 0.754,
      "grad_norm": 4.458641052246094,
      "learning_rate": 2.5894736842105265e-05,
      "loss": 0.3993,
      "step": 9425
    },
    {
      "epoch": 0.756,
      "grad_norm": 2.420055389404297,
      "learning_rate": 2.5684210526315788e-05,
      "loss": 0.4213,
      "step": 9450
    },
    {
      "epoch": 0.758,
      "grad_norm": 3.379028081893921,
      "learning_rate": 2.5473684210526315e-05,
      "loss": 0.4065,
      "step": 9475
    },
    {
      "epoch": 0.76,
      "grad_norm": 3.8290674686431885,
      "learning_rate": 2.5263157894736845e-05,
      "loss": 0.3892,
      "step": 9500
    },
    {
      "epoch": 0.762,
      "grad_norm": 2.3934147357940674,
      "learning_rate": 2.505263157894737e-05,
      "loss": 0.363,
      "step": 9525
    },
    {
      "epoch": 0.764,
      "grad_norm": 2.6929566860198975,
      "learning_rate": 2.4842105263157898e-05,
      "loss": 0.4124,
      "step": 9550
    },
    {
      "epoch": 0.766,
      "grad_norm": 4.540988445281982,
      "learning_rate": 2.463157894736842e-05,
      "loss": 0.372,
      "step": 9575
    },
    {
      "epoch": 0.768,
      "grad_norm": 7.274412631988525,
      "learning_rate": 2.4421052631578948e-05,
      "loss": 0.3798,
      "step": 9600
    },
    {
      "epoch": 0.77,
      "grad_norm": 2.838831663131714,
      "learning_rate": 2.4210526315789474e-05,
      "loss": 0.4108,
      "step": 9625
    },
    {
      "epoch": 0.772,
      "grad_norm": 3.1290194988250732,
      "learning_rate": 2.4e-05,
      "loss": 0.3777,
      "step": 9650
    },
    {
      "epoch": 0.774,
      "grad_norm": 3.401431083679199,
      "learning_rate": 2.3789473684210527e-05,
      "loss": 0.3466,
      "step": 9675
    },
    {
      "epoch": 0.776,
      "grad_norm": 3.971301555633545,
      "learning_rate": 2.3578947368421054e-05,
      "loss": 0.3284,
      "step": 9700
    },
    {
      "epoch": 0.778,
      "grad_norm": 3.7027299404144287,
      "learning_rate": 2.336842105263158e-05,
      "loss": 0.3268,
      "step": 9725
    },
    {
      "epoch": 0.78,
      "grad_norm": 3.9830739498138428,
      "learning_rate": 2.3157894736842107e-05,
      "loss": 0.3523,
      "step": 9750
    },
    {
      "epoch": 0.782,
      "grad_norm": 3.219536304473877,
      "learning_rate": 2.294736842105263e-05,
      "loss": 0.3613,
      "step": 9775
    },
    {
      "epoch": 0.784,
      "grad_norm": 5.3083672523498535,
      "learning_rate": 2.273684210526316e-05,
      "loss": 0.3975,
      "step": 9800
    },
    {
      "epoch": 0.786,
      "grad_norm": 7.210485935211182,
      "learning_rate": 2.2526315789473683e-05,
      "loss": 0.3661,
      "step": 9825
    },
    {
      "epoch": 0.788,
      "grad_norm": 8.692153930664062,
      "learning_rate": 2.2315789473684213e-05,
      "loss": 0.3267,
      "step": 9850
    },
    {
      "epoch": 0.79,
      "grad_norm": 3.8115341663360596,
      "learning_rate": 2.2105263157894736e-05,
      "loss": 0.3664,
      "step": 9875
    },
    {
      "epoch": 0.792,
      "grad_norm": 3.92474627494812,
      "learning_rate": 2.1894736842105266e-05,
      "loss": 0.3677,
      "step": 9900
    },
    {
      "epoch": 0.794,
      "grad_norm": 3.509307622909546,
      "learning_rate": 2.168421052631579e-05,
      "loss": 0.4078,
      "step": 9925
    },
    {
      "epoch": 0.796,
      "grad_norm": 5.251331329345703,
      "learning_rate": 2.147368421052632e-05,
      "loss": 0.3243,
      "step": 9950
    },
    {
      "epoch": 0.798,
      "grad_norm": 5.842615127563477,
      "learning_rate": 2.1263157894736842e-05,
      "loss": 0.3681,
      "step": 9975
    },
    {
      "epoch": 0.8,
      "grad_norm": 2.8434536457061768,
      "learning_rate": 2.105263157894737e-05,
      "loss": 0.3559,
      "step": 10000
    },
    {
      "epoch": 0.802,
      "grad_norm": 6.5307793617248535,
      "learning_rate": 2.0842105263157895e-05,
      "loss": 0.3839,
      "step": 10025
    },
    {
      "epoch": 0.804,
      "grad_norm": 3.4104504585266113,
      "learning_rate": 2.0631578947368422e-05,
      "loss": 0.3887,
      "step": 10050
    },
    {
      "epoch": 0.806,
      "grad_norm": 3.9366180896759033,
      "learning_rate": 2.042105263157895e-05,
      "loss": 0.3695,
      "step": 10075
    },
    {
      "epoch": 0.808,
      "grad_norm": 3.8400652408599854,
      "learning_rate": 2.0210526315789475e-05,
      "loss": 0.3821,
      "step": 10100
    },
    {
      "epoch": 0.81,
      "grad_norm": 4.9908447265625,
      "learning_rate": 2e-05,
      "loss": 0.3639,
      "step": 10125
    },
    {
      "epoch": 0.812,
      "grad_norm": 4.293838024139404,
      "learning_rate": 1.9789473684210528e-05,
      "loss": 0.371,
      "step": 10150
    },
    {
      "epoch": 0.814,
      "grad_norm": 4.710839748382568,
      "learning_rate": 1.957894736842105e-05,
      "loss": 0.3739,
      "step": 10175
    },
    {
      "epoch": 0.816,
      "grad_norm": 4.468384265899658,
      "learning_rate": 1.936842105263158e-05,
      "loss": 0.3414,
      "step": 10200
    },
    {
      "epoch": 0.818,
      "grad_norm": 4.2299957275390625,
      "learning_rate": 1.9157894736842104e-05,
      "loss": 0.3293,
      "step": 10225
    },
    {
      "epoch": 0.82,
      "grad_norm": 2.4495487213134766,
      "learning_rate": 1.8947368421052634e-05,
      "loss": 0.414,
      "step": 10250
    },
    {
      "epoch": 0.822,
      "grad_norm": 4.561651706695557,
      "learning_rate": 1.8736842105263158e-05,
      "loss": 0.4099,
      "step": 10275
    },
    {
      "epoch": 0.824,
      "grad_norm": 2.8751111030578613,
      "learning_rate": 1.8526315789473687e-05,
      "loss": 0.3289,
      "step": 10300
    },
    {
      "epoch": 0.826,
      "grad_norm": 2.919625997543335,
      "learning_rate": 1.831578947368421e-05,
      "loss": 0.3553,
      "step": 10325
    },
    {
      "epoch": 0.828,
      "grad_norm": 1.8310528993606567,
      "learning_rate": 1.810526315789474e-05,
      "loss": 0.3746,
      "step": 10350
    },
    {
      "epoch": 0.83,
      "grad_norm": 2.163372755050659,
      "learning_rate": 1.7894736842105264e-05,
      "loss": 0.4245,
      "step": 10375
    },
    {
      "epoch": 0.832,
      "grad_norm": 4.151187419891357,
      "learning_rate": 1.768421052631579e-05,
      "loss": 0.3887,
      "step": 10400
    },
    {
      "epoch": 0.834,
      "grad_norm": 4.552877426147461,
      "learning_rate": 1.7473684210526317e-05,
      "loss": 0.3552,
      "step": 10425
    },
    {
      "epoch": 0.836,
      "grad_norm": 4.916106224060059,
      "learning_rate": 1.7263157894736843e-05,
      "loss": 0.4124,
      "step": 10450
    },
    {
      "epoch": 0.838,
      "grad_norm": 4.415599822998047,
      "learning_rate": 1.705263157894737e-05,
      "loss": 0.4095,
      "step": 10475
    },
    {
      "epoch": 0.84,
      "grad_norm": 5.485673904418945,
      "learning_rate": 1.6842105263157896e-05,
      "loss": 0.3779,
      "step": 10500
    },
    {
      "epoch": 0.842,
      "grad_norm": 3.694333791732788,
      "learning_rate": 1.6631578947368423e-05,
      "loss": 0.3618,
      "step": 10525
    },
    {
      "epoch": 0.844,
      "grad_norm": 3.96889066696167,
      "learning_rate": 1.642105263157895e-05,
      "loss": 0.3842,
      "step": 10550
    },
    {
      "epoch": 0.846,
      "grad_norm": 6.318436145782471,
      "learning_rate": 1.6210526315789473e-05,
      "loss": 0.3783,
      "step": 10575
    },
    {
      "epoch": 0.848,
      "grad_norm": 4.1109232902526855,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 0.3276,
      "step": 10600
    },
    {
      "epoch": 0.85,
      "grad_norm": 4.43826961517334,
      "learning_rate": 1.5789473684210526e-05,
      "loss": 0.4221,
      "step": 10625
    },
    {
      "epoch": 0.852,
      "grad_norm": 3.70735239982605,
      "learning_rate": 1.5578947368421056e-05,
      "loss": 0.3948,
      "step": 10650
    },
    {
      "epoch": 0.854,
      "grad_norm": 3.354746103286743,
      "learning_rate": 1.536842105263158e-05,
      "loss": 0.4079,
      "step": 10675
    },
    {
      "epoch": 0.856,
      "grad_norm": 5.025430679321289,
      "learning_rate": 1.5157894736842107e-05,
      "loss": 0.3651,
      "step": 10700
    },
    {
      "epoch": 0.858,
      "grad_norm": 3.7044286727905273,
      "learning_rate": 1.4947368421052632e-05,
      "loss": 0.4036,
      "step": 10725
    },
    {
      "epoch": 0.86,
      "grad_norm": 5.7737884521484375,
      "learning_rate": 1.4736842105263157e-05,
      "loss": 0.426,
      "step": 10750
    },
    {
      "epoch": 0.862,
      "grad_norm": 4.619137763977051,
      "learning_rate": 1.4526315789473685e-05,
      "loss": 0.394,
      "step": 10775
    },
    {
      "epoch": 0.864,
      "grad_norm": 3.5880117416381836,
      "learning_rate": 1.431578947368421e-05,
      "loss": 0.3663,
      "step": 10800
    },
    {
      "epoch": 0.866,
      "grad_norm": 5.826897144317627,
      "learning_rate": 1.4105263157894738e-05,
      "loss": 0.329,
      "step": 10825
    },
    {
      "epoch": 0.868,
      "grad_norm": 5.386528968811035,
      "learning_rate": 1.3894736842105263e-05,
      "loss": 0.3184,
      "step": 10850
    },
    {
      "epoch": 0.87,
      "grad_norm": 5.732900142669678,
      "learning_rate": 1.3684210526315791e-05,
      "loss": 0.3611,
      "step": 10875
    },
    {
      "epoch": 0.872,
      "grad_norm": 3.784217596054077,
      "learning_rate": 1.3473684210526316e-05,
      "loss": 0.383,
      "step": 10900
    },
    {
      "epoch": 0.874,
      "grad_norm": 4.212706089019775,
      "learning_rate": 1.3263157894736844e-05,
      "loss": 0.3435,
      "step": 10925
    },
    {
      "epoch": 0.876,
      "grad_norm": 6.505759239196777,
      "learning_rate": 1.305263157894737e-05,
      "loss": 0.3759,
      "step": 10950
    },
    {
      "epoch": 0.878,
      "grad_norm": 3.2388670444488525,
      "learning_rate": 1.2842105263157894e-05,
      "loss": 0.3323,
      "step": 10975
    },
    {
      "epoch": 0.88,
      "grad_norm": 2.824962615966797,
      "learning_rate": 1.2631578947368422e-05,
      "loss": 0.3221,
      "step": 11000
    },
    {
      "epoch": 0.882,
      "grad_norm": 2.974475622177124,
      "learning_rate": 1.2421052631578949e-05,
      "loss": 0.2973,
      "step": 11025
    },
    {
      "epoch": 0.884,
      "grad_norm": 4.150785446166992,
      "learning_rate": 1.2210526315789474e-05,
      "loss": 0.3371,
      "step": 11050
    },
    {
      "epoch": 0.886,
      "grad_norm": 4.3901591300964355,
      "learning_rate": 1.2e-05,
      "loss": 0.2587,
      "step": 11075
    },
    {
      "epoch": 0.888,
      "grad_norm": 2.3966376781463623,
      "learning_rate": 1.1789473684210527e-05,
      "loss": 0.2976,
      "step": 11100
    },
    {
      "epoch": 0.89,
      "grad_norm": 3.4761784076690674,
      "learning_rate": 1.1578947368421053e-05,
      "loss": 0.307,
      "step": 11125
    },
    {
      "epoch": 0.892,
      "grad_norm": 2.4897382259368896,
      "learning_rate": 1.136842105263158e-05,
      "loss": 0.3121,
      "step": 11150
    },
    {
      "epoch": 0.894,
      "grad_norm": 2.342910051345825,
      "learning_rate": 1.1157894736842106e-05,
      "loss": 0.2486,
      "step": 11175
    },
    {
      "epoch": 0.896,
      "grad_norm": 3.4446234703063965,
      "learning_rate": 1.0947368421052633e-05,
      "loss": 0.3367,
      "step": 11200
    },
    {
      "epoch": 0.898,
      "grad_norm": 6.869594097137451,
      "learning_rate": 1.073684210526316e-05,
      "loss": 0.2779,
      "step": 11225
    },
    {
      "epoch": 0.9,
      "grad_norm": 5.409977912902832,
      "learning_rate": 1.0526315789473684e-05,
      "loss": 0.3466,
      "step": 11250
    },
    {
      "epoch": 0.902,
      "grad_norm": 3.0744941234588623,
      "learning_rate": 1.0315789473684211e-05,
      "loss": 0.3442,
      "step": 11275
    },
    {
      "epoch": 0.904,
      "grad_norm": 6.333566188812256,
      "learning_rate": 1.0105263157894738e-05,
      "loss": 0.2866,
      "step": 11300
    },
    {
      "epoch": 0.906,
      "grad_norm": 2.865696907043457,
      "learning_rate": 9.894736842105264e-06,
      "loss": 0.2836,
      "step": 11325
    },
    {
      "epoch": 0.908,
      "grad_norm": 10.0,
      "learning_rate": 9.68421052631579e-06,
      "loss": 0.2817,
      "step": 11350
    },
    {
      "epoch": 0.91,
      "grad_norm": 4.6826605796813965,
      "learning_rate": 9.473684210526317e-06,
      "loss": 0.3117,
      "step": 11375
    },
    {
      "epoch": 0.912,
      "grad_norm": 9.460478782653809,
      "learning_rate": 9.263157894736844e-06,
      "loss": 0.265,
      "step": 11400
    },
    {
      "epoch": 0.914,
      "grad_norm": 8.582769393920898,
      "learning_rate": 9.05263157894737e-06,
      "loss": 0.2552,
      "step": 11425
    },
    {
      "epoch": 0.916,
      "grad_norm": 3.5668044090270996,
      "learning_rate": 8.842105263157895e-06,
      "loss": 0.2983,
      "step": 11450
    },
    {
      "epoch": 0.918,
      "grad_norm": 4.328312873840332,
      "learning_rate": 8.631578947368422e-06,
      "loss": 0.2573,
      "step": 11475
    },
    {
      "epoch": 0.92,
      "grad_norm": 3.874371290206909,
      "learning_rate": 8.421052631578948e-06,
      "loss": 0.2822,
      "step": 11500
    },
    {
      "epoch": 0.922,
      "grad_norm": 3.0442512035369873,
      "learning_rate": 8.210526315789475e-06,
      "loss": 0.2974,
      "step": 11525
    },
    {
      "epoch": 0.924,
      "grad_norm": 4.014752388000488,
      "learning_rate": 8.000000000000001e-06,
      "loss": 0.2627,
      "step": 11550
    },
    {
      "epoch": 0.926,
      "grad_norm": 3.2828469276428223,
      "learning_rate": 7.789473684210528e-06,
      "loss": 0.302,
      "step": 11575
    },
    {
      "epoch": 0.928,
      "grad_norm": 5.778573989868164,
      "learning_rate": 7.578947368421054e-06,
      "loss": 0.2732,
      "step": 11600
    },
    {
      "epoch": 0.93,
      "grad_norm": 4.556640625,
      "learning_rate": 7.3684210526315784e-06,
      "loss": 0.2592,
      "step": 11625
    },
    {
      "epoch": 0.932,
      "grad_norm": 2.93839693069458,
      "learning_rate": 7.157894736842105e-06,
      "loss": 0.2058,
      "step": 11650
    },
    {
      "epoch": 0.934,
      "grad_norm": 2.7247869968414307,
      "learning_rate": 6.9473684210526315e-06,
      "loss": 0.2098,
      "step": 11675
    },
    {
      "epoch": 0.936,
      "grad_norm": 2.1764931678771973,
      "learning_rate": 6.736842105263158e-06,
      "loss": 0.279,
      "step": 11700
    },
    {
      "epoch": 0.938,
      "grad_norm": 1.9565917253494263,
      "learning_rate": 6.526315789473685e-06,
      "loss": 0.2394,
      "step": 11725
    },
    {
      "epoch": 0.94,
      "grad_norm": 2.1610960960388184,
      "learning_rate": 6.315789473684211e-06,
      "loss": 0.2181,
      "step": 11750
    },
    {
      "epoch": 0.942,
      "grad_norm": 4.215695381164551,
      "learning_rate": 6.105263157894737e-06,
      "loss": 0.237,
      "step": 11775
    },
    {
      "epoch": 0.944,
      "grad_norm": 2.702558994293213,
      "learning_rate": 5.8947368421052634e-06,
      "loss": 0.2401,
      "step": 11800
    },
    {
      "epoch": 0.946,
      "grad_norm": 3.1164896488189697,
      "learning_rate": 5.68421052631579e-06,
      "loss": 0.2844,
      "step": 11825
    },
    {
      "epoch": 0.948,
      "grad_norm": 3.159911632537842,
      "learning_rate": 5.4736842105263165e-06,
      "loss": 0.2637,
      "step": 11850
    },
    {
      "epoch": 0.95,
      "grad_norm": 2.3097712993621826,
      "learning_rate": 5.263157894736842e-06,
      "loss": 0.2123,
      "step": 11875
    },
    {
      "epoch": 0.952,
      "grad_norm": 2.6236393451690674,
      "learning_rate": 5.052631578947369e-06,
      "loss": 0.2124,
      "step": 11900
    },
    {
      "epoch": 0.954,
      "grad_norm": 2.240006923675537,
      "learning_rate": 4.842105263157895e-06,
      "loss": 0.2488,
      "step": 11925
    },
    {
      "epoch": 0.956,
      "grad_norm": 2.2833738327026367,
      "learning_rate": 4.631578947368422e-06,
      "loss": 0.2393,
      "step": 11950
    },
    {
      "epoch": 0.958,
      "grad_norm": 2.0853137969970703,
      "learning_rate": 4.4210526315789476e-06,
      "loss": 0.1891,
      "step": 11975
    },
    {
      "epoch": 0.96,
      "grad_norm": 5.040877342224121,
      "learning_rate": 4.210526315789474e-06,
      "loss": 0.2594,
      "step": 12000
    },
    {
      "epoch": 0.962,
      "grad_norm": 3.583343744277954,
      "learning_rate": 4.000000000000001e-06,
      "loss": 0.2463,
      "step": 12025
    },
    {
      "epoch": 0.964,
      "grad_norm": 3.5989181995391846,
      "learning_rate": 3.789473684210527e-06,
      "loss": 0.2231,
      "step": 12050
    },
    {
      "epoch": 0.966,
      "grad_norm": 3.2131879329681396,
      "learning_rate": 3.5789473684210525e-06,
      "loss": 0.2302,
      "step": 12075
    },
    {
      "epoch": 0.968,
      "grad_norm": 2.3687148094177246,
      "learning_rate": 3.368421052631579e-06,
      "loss": 0.2426,
      "step": 12100
    },
    {
      "epoch": 0.97,
      "grad_norm": 2.4739716053009033,
      "learning_rate": 3.1578947368421056e-06,
      "loss": 0.233,
      "step": 12125
    },
    {
      "epoch": 0.972,
      "grad_norm": 4.973099231719971,
      "learning_rate": 2.9473684210526317e-06,
      "loss": 0.2395,
      "step": 12150
    },
    {
      "epoch": 0.974,
      "grad_norm": 1.5512763261795044,
      "learning_rate": 2.7368421052631583e-06,
      "loss": 0.2437,
      "step": 12175
    },
    {
      "epoch": 0.976,
      "grad_norm": 3.5859758853912354,
      "learning_rate": 2.5263157894736844e-06,
      "loss": 0.2519,
      "step": 12200
    },
    {
      "epoch": 0.978,
      "grad_norm": 2.6558282375335693,
      "learning_rate": 2.315789473684211e-06,
      "loss": 0.2029,
      "step": 12225
    },
    {
      "epoch": 0.98,
      "grad_norm": 6.469668388366699,
      "learning_rate": 2.105263157894737e-06,
      "loss": 0.2182,
      "step": 12250
    },
    {
      "epoch": 0.982,
      "grad_norm": 4.018293857574463,
      "learning_rate": 1.8947368421052634e-06,
      "loss": 0.1946,
      "step": 12275
    },
    {
      "epoch": 0.984,
      "grad_norm": 2.706835985183716,
      "learning_rate": 1.6842105263157895e-06,
      "loss": 0.2391,
      "step": 12300
    },
    {
      "epoch": 0.986,
      "grad_norm": 5.652164936065674,
      "learning_rate": 1.4736842105263159e-06,
      "loss": 0.2884,
      "step": 12325
    },
    {
      "epoch": 0.988,
      "grad_norm": 2.2838077545166016,
      "learning_rate": 1.2631578947368422e-06,
      "loss": 0.2257,
      "step": 12350
    },
    {
      "epoch": 0.99,
      "grad_norm": 4.01392936706543,
      "learning_rate": 1.0526315789473685e-06,
      "loss": 0.2852,
      "step": 12375
    },
    {
      "epoch": 0.992,
      "grad_norm": 5.119837760925293,
      "learning_rate": 8.421052631578948e-07,
      "loss": 0.2021,
      "step": 12400
    },
    {
      "epoch": 0.994,
      "grad_norm": 3.6381170749664307,
      "learning_rate": 6.315789473684211e-07,
      "loss": 0.2518,
      "step": 12425
    },
    {
      "epoch": 0.996,
      "grad_norm": 3.1016297340393066,
      "learning_rate": 4.210526315789474e-07,
      "loss": 0.2371,
      "step": 12450
    },
    {
      "epoch": 0.998,
      "grad_norm": 4.031352519989014,
      "learning_rate": 2.105263157894737e-07,
      "loss": 0.2753,
      "step": 12475
    },
    {
      "epoch": 1.0,
      "grad_norm": 2.030440330505371,
      "learning_rate": 0.0,
      "loss": 0.2449,
      "step": 12500
    }
  ],
  "logging_steps": 25,
  "max_steps": 12500,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}