{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9493463695685385,
  "eval_steps": 500,
  "global_step": 43500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "grad_norm": 0.7957815527915955,
      "learning_rate": 9.97817594552716e-06,
      "loss": 5.9209,
      "step": 100
    },
    {
      "epoch": 0.0,
      "grad_norm": 0.7227272391319275,
      "learning_rate": 9.956351891054321e-06,
      "loss": 5.8363,
      "step": 200
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.6150572896003723,
      "learning_rate": 9.93452783658148e-06,
      "loss": 5.7212,
      "step": 300
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.4721869230270386,
      "learning_rate": 9.912703782108642e-06,
      "loss": 5.5998,
      "step": 400
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.4436299204826355,
      "learning_rate": 9.890879727635801e-06,
      "loss": 5.4923,
      "step": 500
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.39484986662864685,
      "learning_rate": 9.86905567316296e-06,
      "loss": 5.3936,
      "step": 600
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.336157888174057,
      "learning_rate": 9.847231618690121e-06,
      "loss": 5.3086,
      "step": 700
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.28207510709762573,
      "learning_rate": 9.82540756421728e-06,
      "loss": 5.2389,
      "step": 800
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.21387213468551636,
      "learning_rate": 9.80358350974444e-06,
      "loss": 5.1929,
      "step": 900
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.19382569193840027,
      "learning_rate": 9.781759455271601e-06,
      "loss": 5.1608,
      "step": 1000
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.18469111621379852,
      "learning_rate": 9.75993540079876e-06,
      "loss": 5.1402,
      "step": 1100
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.16184094548225403,
      "learning_rate": 9.738111346325922e-06,
      "loss": 5.121,
      "step": 1200
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.16706742346286774,
      "learning_rate": 9.716287291853081e-06,
      "loss": 5.1087,
      "step": 1300
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.16692855954170227,
      "learning_rate": 9.694463237380242e-06,
      "loss": 5.0948,
      "step": 1400
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.15095993876457214,
      "learning_rate": 9.672639182907401e-06,
      "loss": 5.0854,
      "step": 1500
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.1381816416978836,
      "learning_rate": 9.65081512843456e-06,
      "loss": 5.0749,
      "step": 1600
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.14653028547763824,
      "learning_rate": 9.62899107396172e-06,
      "loss": 5.0654,
      "step": 1700
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.15313006937503815,
      "learning_rate": 9.607167019488881e-06,
      "loss": 5.0578,
      "step": 1800
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.14888273179531097,
      "learning_rate": 9.58534296501604e-06,
      "loss": 5.0526,
      "step": 1900
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.16502290964126587,
      "learning_rate": 9.563518910543202e-06,
      "loss": 5.0386,
      "step": 2000
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.15124522149562836,
      "learning_rate": 9.541694856070361e-06,
      "loss": 5.0313,
      "step": 2100
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.1460607498884201,
      "learning_rate": 9.519870801597522e-06,
      "loss": 5.0253,
      "step": 2200
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.15850311517715454,
      "learning_rate": 9.498046747124681e-06,
      "loss": 5.0167,
      "step": 2300
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.21040180325508118,
      "learning_rate": 9.476222692651842e-06,
      "loss": 5.0112,
      "step": 2400
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.1923857480287552,
      "learning_rate": 9.454398638179002e-06,
      "loss": 5.002,
      "step": 2500
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.1804654449224472,
      "learning_rate": 9.432574583706163e-06,
      "loss": 5.0006,
      "step": 2600
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.18370215594768524,
      "learning_rate": 9.41075052923332e-06,
      "loss": 4.9894,
      "step": 2700
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.22127391397953033,
      "learning_rate": 9.388926474760482e-06,
      "loss": 4.9866,
      "step": 2800
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.2355722337961197,
      "learning_rate": 9.367102420287641e-06,
      "loss": 4.9807,
      "step": 2900
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.2686903178691864,
      "learning_rate": 9.345278365814802e-06,
      "loss": 4.9762,
      "step": 3000
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.2140015810728073,
      "learning_rate": 9.323454311341961e-06,
      "loss": 4.9703,
      "step": 3100
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.2413332760334015,
      "learning_rate": 9.301630256869122e-06,
      "loss": 4.9677,
      "step": 3200
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.36390700936317444,
      "learning_rate": 9.279806202396282e-06,
      "loss": 4.9618,
      "step": 3300
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.20959585905075073,
      "learning_rate": 9.257982147923443e-06,
      "loss": 4.9577,
      "step": 3400
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.2816881239414215,
      "learning_rate": 9.236158093450602e-06,
      "loss": 4.9579,
      "step": 3500
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.565319299697876,
      "learning_rate": 9.214334038977763e-06,
      "loss": 4.9513,
      "step": 3600
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.6039319634437561,
      "learning_rate": 9.192509984504923e-06,
      "loss": 4.9505,
      "step": 3700
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.18052706122398376,
      "learning_rate": 9.170685930032082e-06,
      "loss": 4.9481,
      "step": 3800
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.22476325929164886,
      "learning_rate": 9.148861875559241e-06,
      "loss": 4.945,
      "step": 3900
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.21381278336048126,
      "learning_rate": 9.127037821086402e-06,
      "loss": 4.942,
      "step": 4000
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.1974494308233261,
      "learning_rate": 9.105213766613562e-06,
      "loss": 4.9382,
      "step": 4100
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.41070911288261414,
      "learning_rate": 9.083389712140723e-06,
      "loss": 4.9366,
      "step": 4200
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.1987224966287613,
      "learning_rate": 9.061565657667882e-06,
      "loss": 4.9356,
      "step": 4300
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.24896694719791412,
      "learning_rate": 9.039741603195043e-06,
      "loss": 4.934,
      "step": 4400
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.2133394479751587,
      "learning_rate": 9.017917548722203e-06,
      "loss": 4.9317,
      "step": 4500
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.24069222807884216,
      "learning_rate": 8.996093494249362e-06,
      "loss": 4.9308,
      "step": 4600
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.30186885595321655,
      "learning_rate": 8.974269439776523e-06,
      "loss": 4.9294,
      "step": 4700
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.22445006668567657,
      "learning_rate": 8.952445385303682e-06,
      "loss": 4.9279,
      "step": 4800
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.5278885960578918,
      "learning_rate": 8.930621330830842e-06,
      "loss": 4.9219,
      "step": 4900
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.34161177277565,
      "learning_rate": 8.908797276358003e-06,
      "loss": 4.9185,
      "step": 5000
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.46790817379951477,
      "learning_rate": 8.886973221885162e-06,
      "loss": 4.9219,
      "step": 5100
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.24288785457611084,
      "learning_rate": 8.865149167412323e-06,
      "loss": 4.9184,
      "step": 5200
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.0053044557571411,
      "learning_rate": 8.843325112939482e-06,
      "loss": 4.9202,
      "step": 5300
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.2430969476699829,
      "learning_rate": 8.821501058466644e-06,
      "loss": 4.9138,
      "step": 5400
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.4104827642440796,
      "learning_rate": 8.799677003993803e-06,
      "loss": 4.9119,
      "step": 5500
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.3076772093772888,
      "learning_rate": 8.777852949520962e-06,
      "loss": 4.9135,
      "step": 5600
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.2760007679462433,
      "learning_rate": 8.756028895048123e-06,
      "loss": 4.915,
      "step": 5700
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.35674023628234863,
      "learning_rate": 8.734204840575283e-06,
      "loss": 4.908,
      "step": 5800
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.370185911655426,
      "learning_rate": 8.712380786102442e-06,
      "loss": 4.9129,
      "step": 5900
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.3864850699901581,
      "learning_rate": 8.690556731629603e-06,
      "loss": 4.9084,
      "step": 6000
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.5277231335639954,
      "learning_rate": 8.668732677156762e-06,
      "loss": 4.91,
      "step": 6100
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.4505026936531067,
      "learning_rate": 8.646908622683924e-06,
      "loss": 4.9074,
      "step": 6200
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.9321696758270264,
      "learning_rate": 8.625084568211083e-06,
      "loss": 4.905,
      "step": 6300
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.7220796346664429,
      "learning_rate": 8.603260513738242e-06,
      "loss": 4.907,
      "step": 6400
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.26252031326293945,
      "learning_rate": 8.581436459265403e-06,
      "loss": 4.9033,
      "step": 6500
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.2740485966205597,
      "learning_rate": 8.559612404792563e-06,
      "loss": 4.9014,
      "step": 6600
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.5640104413032532,
      "learning_rate": 8.537788350319724e-06,
      "loss": 4.9019,
      "step": 6700
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.3341173827648163,
      "learning_rate": 8.515964295846883e-06,
      "loss": 4.8988,
      "step": 6800
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.45002481341362,
      "learning_rate": 8.494140241374044e-06,
      "loss": 4.8982,
      "step": 6900
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.3300448954105377,
      "learning_rate": 8.472316186901203e-06,
      "loss": 4.8983,
      "step": 7000
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.38954102993011475,
      "learning_rate": 8.450492132428363e-06,
      "loss": 4.8994,
      "step": 7100
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.29785066843032837,
      "learning_rate": 8.428668077955522e-06,
      "loss": 4.8946,
      "step": 7200
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.3793606758117676,
      "learning_rate": 8.406844023482683e-06,
      "loss": 4.893,
      "step": 7300
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.3125133216381073,
      "learning_rate": 8.385019969009843e-06,
      "loss": 4.893,
      "step": 7400
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.30539774894714355,
      "learning_rate": 8.363195914537004e-06,
      "loss": 4.8938,
      "step": 7500
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.36112794280052185,
      "learning_rate": 8.341371860064163e-06,
      "loss": 4.8952,
      "step": 7600
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.29357707500457764,
      "learning_rate": 8.319547805591324e-06,
      "loss": 4.8939,
      "step": 7700
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.6816242933273315,
      "learning_rate": 8.297723751118483e-06,
      "loss": 4.8941,
      "step": 7800
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.46635544300079346,
      "learning_rate": 8.275899696645645e-06,
      "loss": 4.8914,
      "step": 7900
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.4570920467376709,
      "learning_rate": 8.254075642172802e-06,
      "loss": 4.8897,
      "step": 8000
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.0053654909133911,
      "learning_rate": 8.232251587699963e-06,
      "loss": 4.8932,
      "step": 8100
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.5443705320358276,
      "learning_rate": 8.210427533227123e-06,
      "loss": 4.8904,
      "step": 8200
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.9791301488876343,
      "learning_rate": 8.188603478754284e-06,
      "loss": 4.8887,
      "step": 8300
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.41778409481048584,
      "learning_rate": 8.166779424281443e-06,
      "loss": 4.8881,
      "step": 8400
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.3111547529697418,
      "learning_rate": 8.144955369808604e-06,
      "loss": 4.888,
      "step": 8500
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.8998810052871704,
      "learning_rate": 8.123131315335763e-06,
      "loss": 4.8882,
      "step": 8600
    },
    {
      "epoch": 0.19,
      "grad_norm": 2.0568580627441406,
      "learning_rate": 8.101307260862924e-06,
      "loss": 4.8836,
      "step": 8700
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.4567314088344574,
      "learning_rate": 8.079483206390084e-06,
      "loss": 4.8872,
      "step": 8800
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.8183225393295288,
      "learning_rate": 8.057659151917245e-06,
      "loss": 4.8832,
      "step": 8900
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.41399893164634705,
      "learning_rate": 8.035835097444404e-06,
      "loss": 4.8831,
      "step": 9000
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.7039474248886108,
      "learning_rate": 8.014011042971564e-06,
      "loss": 4.8866,
      "step": 9100
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.2951236367225647,
      "learning_rate": 7.992186988498723e-06,
      "loss": 4.8823,
      "step": 9200
    },
    {
      "epoch": 0.2,
      "grad_norm": 2.30894136428833,
      "learning_rate": 7.970362934025884e-06,
      "loss": 4.8847,
      "step": 9300
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.0861660242080688,
      "learning_rate": 7.948538879553043e-06,
      "loss": 4.8818,
      "step": 9400
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.6845309138298035,
      "learning_rate": 7.926714825080204e-06,
      "loss": 4.8821,
      "step": 9500
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.36056384444236755,
      "learning_rate": 7.904890770607364e-06,
      "loss": 4.8786,
      "step": 9600
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.4470602571964264,
      "learning_rate": 7.883066716134525e-06,
      "loss": 4.8825,
      "step": 9700
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.1526983976364136,
      "learning_rate": 7.861242661661684e-06,
      "loss": 4.8832,
      "step": 9800
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.4809672832489014,
      "learning_rate": 7.839418607188845e-06,
      "loss": 4.88,
      "step": 9900
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.5240287780761719,
      "learning_rate": 7.817594552716005e-06,
      "loss": 4.8799,
      "step": 10000
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.7026857733726501,
      "learning_rate": 7.795770498243164e-06,
      "loss": 4.878,
      "step": 10100
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.1413593292236328,
      "learning_rate": 7.773946443770323e-06,
      "loss": 4.879,
      "step": 10200
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.347445011138916,
      "learning_rate": 7.752122389297484e-06,
      "loss": 4.876,
      "step": 10300
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.108039379119873,
      "learning_rate": 7.730298334824644e-06,
      "loss": 4.875,
      "step": 10400
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.1911160945892334,
      "learning_rate": 7.708474280351805e-06,
      "loss": 4.8781,
      "step": 10500
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.3719279766082764,
      "learning_rate": 7.686650225878964e-06,
      "loss": 4.8771,
      "step": 10600
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.9556494355201721,
      "learning_rate": 7.664826171406125e-06,
      "loss": 4.8763,
      "step": 10700
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.42252323031425476,
      "learning_rate": 7.643002116933285e-06,
      "loss": 4.8751,
      "step": 10800
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.9392989873886108,
      "learning_rate": 7.621178062460445e-06,
      "loss": 4.8751,
      "step": 10900
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.5632112622261047,
      "learning_rate": 7.599354007987605e-06,
      "loss": 4.873,
      "step": 11000
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.8829624056816101,
      "learning_rate": 7.577529953514765e-06,
      "loss": 4.874,
      "step": 11100
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.4162335395812988,
      "learning_rate": 7.555705899041924e-06,
      "loss": 4.8713,
      "step": 11200
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.8155901432037354,
      "learning_rate": 7.533881844569084e-06,
      "loss": 4.8728,
      "step": 11300
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.4502034187316895,
      "learning_rate": 7.512057790096244e-06,
      "loss": 4.8763,
      "step": 11400
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.4868534803390503,
      "learning_rate": 7.490233735623404e-06,
      "loss": 4.8731,
      "step": 11500
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.3518391847610474,
      "learning_rate": 7.4684096811505646e-06,
      "loss": 4.8702,
      "step": 11600
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.5818581581115723,
      "learning_rate": 7.446585626677725e-06,
      "loss": 4.8709,
      "step": 11700
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.434921383857727,
      "learning_rate": 7.424761572204885e-06,
      "loss": 4.8708,
      "step": 11800
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.8794093132019043,
      "learning_rate": 7.402937517732045e-06,
      "loss": 4.8684,
      "step": 11900
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.3370536267757416,
      "learning_rate": 7.381113463259205e-06,
      "loss": 4.8711,
      "step": 12000
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.7272424697875977,
      "learning_rate": 7.359289408786366e-06,
      "loss": 4.875,
      "step": 12100
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.3709563612937927,
      "learning_rate": 7.337465354313526e-06,
      "loss": 4.8682,
      "step": 12200
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.8434191346168518,
      "learning_rate": 7.315641299840684e-06,
      "loss": 4.8693,
      "step": 12300
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.244709849357605,
      "learning_rate": 7.2938172453678445e-06,
      "loss": 4.8681,
      "step": 12400
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.7124190926551819,
      "learning_rate": 7.271993190895005e-06,
      "loss": 4.8702,
      "step": 12500
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.8966384530067444,
      "learning_rate": 7.250169136422165e-06,
      "loss": 4.8706,
      "step": 12600
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.42346426844596863,
      "learning_rate": 7.228345081949325e-06,
      "loss": 4.8679,
      "step": 12700
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.0583099126815796,
      "learning_rate": 7.206521027476485e-06,
      "loss": 4.8669,
      "step": 12800
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.9891448020935059,
      "learning_rate": 7.1846969730036456e-06,
      "loss": 4.8688,
      "step": 12900
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.387790322303772,
      "learning_rate": 7.162872918530806e-06,
      "loss": 4.8668,
      "step": 13000
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.7412809133529663,
      "learning_rate": 7.141048864057966e-06,
      "loss": 4.8658,
      "step": 13100
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.7861388921737671,
      "learning_rate": 7.119224809585125e-06,
      "loss": 4.8651,
      "step": 13200
    },
    {
      "epoch": 0.29,
      "grad_norm": 3.033806085586548,
      "learning_rate": 7.0974007551122855e-06,
      "loss": 4.8655,
      "step": 13300
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.232498049736023,
      "learning_rate": 7.075576700639445e-06,
      "loss": 4.8633,
      "step": 13400
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.7451385259628296,
      "learning_rate": 7.053752646166605e-06,
      "loss": 4.8668,
      "step": 13500
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.5436673760414124,
      "learning_rate": 7.031928591693765e-06,
      "loss": 4.867,
      "step": 13600
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.891558289527893,
      "learning_rate": 7.0101045372209255e-06,
      "loss": 4.8652,
      "step": 13700
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.3846653401851654,
      "learning_rate": 6.988280482748086e-06,
      "loss": 4.8668,
      "step": 13800
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.1685656309127808,
      "learning_rate": 6.966456428275246e-06,
      "loss": 4.8661,
      "step": 13900
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.38460877537727356,
      "learning_rate": 6.944632373802406e-06,
      "loss": 4.8651,
      "step": 14000
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.9353694915771484,
      "learning_rate": 6.9228083193295655e-06,
      "loss": 4.8635,
      "step": 14100
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.7546364068984985,
      "learning_rate": 6.900984264856726e-06,
      "loss": 4.8639,
      "step": 14200
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.9310697317123413,
      "learning_rate": 6.879160210383886e-06,
      "loss": 4.862,
      "step": 14300
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.5370875597000122,
      "learning_rate": 6.857336155911045e-06,
      "loss": 4.8633,
      "step": 14400
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.8054449558258057,
      "learning_rate": 6.8355121014382055e-06,
      "loss": 4.866,
      "step": 14500
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.40611472725868225,
      "learning_rate": 6.813688046965366e-06,
      "loss": 4.8638,
      "step": 14600
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.6744693517684937,
      "learning_rate": 6.791863992492526e-06,
      "loss": 4.8639,
      "step": 14700
    },
    {
      "epoch": 0.32,
      "grad_norm": 2.0651392936706543,
      "learning_rate": 6.770039938019686e-06,
      "loss": 4.8626,
      "step": 14800
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.0164268016815186,
      "learning_rate": 6.7482158835468455e-06,
      "loss": 4.8646,
      "step": 14900
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.5757657289505005,
      "learning_rate": 6.726391829074006e-06,
      "loss": 4.8642,
      "step": 15000
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.9722656607627869,
      "learning_rate": 6.704567774601166e-06,
      "loss": 4.8647,
      "step": 15100
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.559166431427002,
      "learning_rate": 6.682743720128326e-06,
      "loss": 4.8623,
      "step": 15200
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.8523690700531006,
      "learning_rate": 6.660919665655486e-06,
      "loss": 4.8643,
      "step": 15300
    },
    {
      "epoch": 0.34,
      "grad_norm": 2.1013731956481934,
      "learning_rate": 6.6390956111826465e-06,
      "loss": 4.8615,
      "step": 15400
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.3530488014221191,
      "learning_rate": 6.617271556709806e-06,
      "loss": 4.8617,
      "step": 15500
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.8760623931884766,
      "learning_rate": 6.595447502236966e-06,
      "loss": 4.8604,
      "step": 15600
    },
    {
      "epoch": 0.34,
      "grad_norm": 2.709780216217041,
      "learning_rate": 6.5736234477641254e-06,
      "loss": 4.86,
      "step": 15700
    },
    {
      "epoch": 0.34,
      "grad_norm": 2.935454845428467,
      "learning_rate": 6.551799393291286e-06,
      "loss": 4.8588,
      "step": 15800
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.8519152402877808,
      "learning_rate": 6.529975338818446e-06,
      "loss": 4.86,
      "step": 15900
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.5556691288948059,
      "learning_rate": 6.508151284345606e-06,
      "loss": 4.8596,
      "step": 16000
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.9867918491363525,
      "learning_rate": 6.486327229872766e-06,
      "loss": 4.8569,
      "step": 16100
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.39979785680770874,
      "learning_rate": 6.4645031753999265e-06,
      "loss": 4.8594,
      "step": 16200
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.42257916927337646,
      "learning_rate": 6.442679120927087e-06,
      "loss": 4.8604,
      "step": 16300
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.39576584100723267,
      "learning_rate": 6.420855066454247e-06,
      "loss": 4.8595,
      "step": 16400
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.216848373413086,
      "learning_rate": 6.399031011981407e-06,
      "loss": 4.8599,
      "step": 16500
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.709771990776062,
      "learning_rate": 6.377206957508566e-06,
      "loss": 4.8571,
      "step": 16600
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.4081980884075165,
      "learning_rate": 6.355382903035726e-06,
      "loss": 4.857,
      "step": 16700
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.7303671836853027,
      "learning_rate": 6.333558848562886e-06,
      "loss": 4.8584,
      "step": 16800
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.0993214845657349,
      "learning_rate": 6.311734794090046e-06,
      "loss": 4.859,
      "step": 16900
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.951707661151886,
      "learning_rate": 6.2899107396172064e-06,
      "loss": 4.856,
      "step": 17000
    },
    {
      "epoch": 0.37,
      "grad_norm": 2.3436036109924316,
      "learning_rate": 6.268086685144367e-06,
      "loss": 4.8559,
      "step": 17100
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.7691673040390015,
      "learning_rate": 6.246262630671527e-06,
      "loss": 4.8595,
      "step": 17200
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.3263872861862183,
      "learning_rate": 6.224438576198687e-06,
      "loss": 4.8562,
      "step": 17300
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.6155282258987427,
      "learning_rate": 6.202614521725847e-06,
      "loss": 4.8578,
      "step": 17400
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.459848403930664,
      "learning_rate": 6.1807904672530075e-06,
      "loss": 4.8536,
      "step": 17500
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.0362757444381714,
      "learning_rate": 6.158966412780166e-06,
      "loss": 4.8559,
      "step": 17600
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.5081543922424316,
      "learning_rate": 6.137142358307326e-06,
      "loss": 4.8576,
      "step": 17700
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.0799150466918945,
      "learning_rate": 6.115318303834486e-06,
      "loss": 4.8553,
      "step": 17800
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.4601239562034607,
      "learning_rate": 6.093494249361647e-06,
      "loss": 4.8552,
      "step": 17900
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.4647897779941559,
      "learning_rate": 6.071670194888807e-06,
      "loss": 4.8515,
      "step": 18000
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.550953209400177,
      "learning_rate": 6.049846140415967e-06,
      "loss": 4.8559,
      "step": 18100
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.748518705368042,
      "learning_rate": 6.028022085943127e-06,
      "loss": 4.8543,
      "step": 18200
    },
    {
      "epoch": 0.4,
      "grad_norm": 2.6062777042388916,
      "learning_rate": 6.0061980314702874e-06,
      "loss": 4.8571,
      "step": 18300
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.8579190373420715,
      "learning_rate": 5.984373976997448e-06,
      "loss": 4.8566,
      "step": 18400
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.8633508682250977,
      "learning_rate": 5.962549922524608e-06,
      "loss": 4.8528,
      "step": 18500
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.5768360495567322,
      "learning_rate": 5.940725868051768e-06,
      "loss": 4.8559,
      "step": 18600
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.7455744743347168,
      "learning_rate": 5.918901813578927e-06,
      "loss": 4.8577,
      "step": 18700
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.3996860682964325,
      "learning_rate": 5.897077759106087e-06,
      "loss": 4.8547,
      "step": 18800
    },
    {
      "epoch": 0.41,
      "grad_norm": 2.049177885055542,
      "learning_rate": 5.875253704633247e-06,
      "loss": 4.8549,
      "step": 18900
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.0341905355453491,
      "learning_rate": 5.853429650160407e-06,
      "loss": 4.8506,
      "step": 19000
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.4852238893508911,
      "learning_rate": 5.831605595687567e-06,
      "loss": 4.8561,
      "step": 19100
    },
    {
      "epoch": 0.42,
      "grad_norm": 2.2729053497314453,
      "learning_rate": 5.809781541214728e-06,
      "loss": 4.8552,
      "step": 19200
    },
    {
      "epoch": 0.42,
      "grad_norm": 2.259981393814087,
      "learning_rate": 5.787957486741888e-06,
      "loss": 4.8506,
      "step": 19300
    },
    {
      "epoch": 0.42,
      "grad_norm": 2.0674850940704346,
      "learning_rate": 5.766133432269048e-06,
      "loss": 4.8543,
      "step": 19400
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.3431077003479004,
      "learning_rate": 5.744309377796207e-06,
      "loss": 4.8522,
      "step": 19500
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.9442102313041687,
      "learning_rate": 5.722485323323368e-06,
      "loss": 4.8511,
      "step": 19600
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.453003466129303,
      "learning_rate": 5.700661268850528e-06,
      "loss": 4.8512,
      "step": 19700
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.39568933844566345,
      "learning_rate": 5.678837214377687e-06,
      "loss": 4.8527,
      "step": 19800
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.3336150646209717,
      "learning_rate": 5.657013159904847e-06,
      "loss": 4.852,
      "step": 19900
    },
    {
      "epoch": 0.44,
      "grad_norm": 3.6281139850616455,
      "learning_rate": 5.635189105432008e-06,
      "loss": 4.853,
      "step": 20000
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.4694347679615021,
      "learning_rate": 5.613365050959168e-06,
      "loss": 4.8529,
      "step": 20100
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.7866901159286499,
      "learning_rate": 5.591540996486328e-06,
      "loss": 4.851,
      "step": 20200
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.5806153416633606,
      "learning_rate": 5.569716942013487e-06,
      "loss": 4.8539,
      "step": 20300
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.5267885327339172,
      "learning_rate": 5.5478928875406476e-06,
      "loss": 4.8559,
      "step": 20400
    },
    {
      "epoch": 0.45,
      "grad_norm": 2.056804656982422,
      "learning_rate": 5.526068833067808e-06,
      "loss": 4.855,
      "step": 20500
    },
    {
      "epoch": 0.45,
      "grad_norm": 3.2893295288085938,
      "learning_rate": 5.504244778594968e-06,
      "loss": 4.8512,
      "step": 20600
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.9322779774665833,
      "learning_rate": 5.482420724122128e-06,
      "loss": 4.8525,
      "step": 20700
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.505246639251709,
      "learning_rate": 5.4605966696492876e-06,
      "loss": 4.8499,
      "step": 20800
    },
    {
      "epoch": 0.46,
      "grad_norm": 2.040572166442871,
      "learning_rate": 5.438772615176448e-06,
      "loss": 4.8509,
      "step": 20900
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.8444960713386536,
      "learning_rate": 5.416948560703608e-06,
      "loss": 4.8496,
      "step": 21000
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.4201112985610962,
      "learning_rate": 5.395124506230768e-06,
      "loss": 4.8505,
      "step": 21100
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.2699060440063477,
      "learning_rate": 5.3733004517579275e-06,
      "loss": 4.8498,
      "step": 21200
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.7760267853736877,
      "learning_rate": 5.351476397285088e-06,
      "loss": 4.8506,
      "step": 21300
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.49700111150741577,
      "learning_rate": 5.329652342812248e-06,
      "loss": 4.8513,
      "step": 21400
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.6455672979354858,
      "learning_rate": 5.307828288339408e-06,
      "loss": 4.8514,
      "step": 21500
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.6039340496063232,
      "learning_rate": 5.286004233866568e-06,
      "loss": 4.8486,
      "step": 21600
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.9197044968605042,
      "learning_rate": 5.2641801793937286e-06,
      "loss": 4.8515,
      "step": 21700
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.3925398290157318,
      "learning_rate": 5.242356124920889e-06,
      "loss": 4.8506,
      "step": 21800
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.0743471384048462,
      "learning_rate": 5.220532070448048e-06,
      "loss": 4.853,
      "step": 21900
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.4220006763935089,
      "learning_rate": 5.1987080159752075e-06,
      "loss": 4.8493,
      "step": 22000
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.6162962913513184,
      "learning_rate": 5.176883961502368e-06,
      "loss": 4.8518,
      "step": 22100
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.6012248992919922,
      "learning_rate": 5.155059907029528e-06,
      "loss": 4.8468,
      "step": 22200
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.838961124420166,
      "learning_rate": 5.133235852556688e-06,
      "loss": 4.8511,
      "step": 22300
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.9484496116638184,
      "learning_rate": 5.111411798083848e-06,
      "loss": 4.8523,
      "step": 22400
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.7750368118286133,
      "learning_rate": 5.0895877436110085e-06,
      "loss": 4.8467,
      "step": 22500
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.7409512996673584,
      "learning_rate": 5.067763689138169e-06,
      "loss": 4.8465,
      "step": 22600
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.6893154382705688,
      "learning_rate": 5.045939634665329e-06,
      "loss": 4.8507,
      "step": 22700
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.4250375032424927,
      "learning_rate": 5.024115580192489e-06,
      "loss": 4.8471,
      "step": 22800
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.5553233623504639,
      "learning_rate": 5.002291525719649e-06,
      "loss": 4.847,
      "step": 22900
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.42157289385795593,
      "learning_rate": 4.980467471246809e-06,
      "loss": 4.8465,
      "step": 23000
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.3181480169296265,
      "learning_rate": 4.958643416773969e-06,
      "loss": 4.8471,
      "step": 23100
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.3415127992630005,
      "learning_rate": 4.936819362301128e-06,
      "loss": 4.8476,
      "step": 23200
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.485499382019043,
      "learning_rate": 4.9149953078282885e-06,
      "loss": 4.8487,
      "step": 23300
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.6069031953811646,
      "learning_rate": 4.893171253355449e-06,
      "loss": 4.8501,
      "step": 23400
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.8160085678100586,
      "learning_rate": 4.871347198882609e-06,
      "loss": 4.8474,
      "step": 23500
    },
    {
      "epoch": 0.52,
      "grad_norm": 2.411562204360962,
      "learning_rate": 4.849523144409769e-06,
      "loss": 4.8437,
      "step": 23600
    },
    {
      "epoch": 0.52,
      "grad_norm": 2.40973162651062,
      "learning_rate": 4.827699089936929e-06,
      "loss": 4.8459,
      "step": 23700
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.4108150005340576,
      "learning_rate": 4.805875035464089e-06,
      "loss": 4.8472,
      "step": 23800
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.5337820649147034,
      "learning_rate": 4.784050980991249e-06,
      "loss": 4.8472,
      "step": 23900
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.6120834350585938,
      "learning_rate": 4.762226926518409e-06,
      "loss": 4.8471,
      "step": 24000
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.2425861358642578,
      "learning_rate": 4.740402872045569e-06,
      "loss": 4.8476,
      "step": 24100
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.7485963106155396,
      "learning_rate": 4.7185788175727295e-06,
      "loss": 4.8519,
      "step": 24200
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.4650036692619324,
      "learning_rate": 4.696754763099889e-06,
      "loss": 4.8474,
      "step": 24300
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.4887036681175232,
      "learning_rate": 4.674930708627049e-06,
      "loss": 4.8464,
      "step": 24400
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.113826036453247,
      "learning_rate": 4.653106654154209e-06,
      "loss": 4.8492,
      "step": 24500
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.9913270473480225,
      "learning_rate": 4.6312825996813695e-06,
      "loss": 4.8456,
      "step": 24600
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.8120296001434326,
      "learning_rate": 4.60945854520853e-06,
      "loss": 4.8465,
      "step": 24700
    },
    {
      "epoch": 0.54,
      "grad_norm": 2.4769647121429443,
      "learning_rate": 4.587634490735689e-06,
      "loss": 4.8434,
      "step": 24800
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.9758397340774536,
      "learning_rate": 4.565810436262849e-06,
      "loss": 4.8425,
      "step": 24900
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.4331335723400116,
      "learning_rate": 4.5439863817900095e-06,
      "loss": 4.8479,
      "step": 25000
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.6201393604278564,
      "learning_rate": 4.52216232731717e-06,
      "loss": 4.848,
      "step": 25100
    },
    {
      "epoch": 0.55,
      "grad_norm": 3.273059129714966,
      "learning_rate": 4.50033827284433e-06,
      "loss": 4.846,
      "step": 25200
    },
    {
      "epoch": 0.55,
      "grad_norm": 2.6748781204223633,
      "learning_rate": 4.47851421837149e-06,
      "loss": 4.8451,
      "step": 25300
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.6629786491394043,
      "learning_rate": 4.4566901638986495e-06,
      "loss": 4.8424,
      "step": 25400
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.5216906070709229,
      "learning_rate": 4.43486610942581e-06,
      "loss": 4.8483,
      "step": 25500
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.6084491014480591,
      "learning_rate": 4.41304205495297e-06,
      "loss": 4.8445,
      "step": 25600
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.6803710460662842,
      "learning_rate": 4.39121800048013e-06,
      "loss": 4.8436,
      "step": 25700
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.8140966892242432,
      "learning_rate": 4.3693939460072895e-06,
      "loss": 4.8457,
      "step": 25800
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.5089240670204163,
      "learning_rate": 4.34756989153445e-06,
      "loss": 4.848,
      "step": 25900
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.5802351236343384,
      "learning_rate": 4.32574583706161e-06,
      "loss": 4.8451,
      "step": 26000
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.4696178734302521,
      "learning_rate": 4.30392178258877e-06,
      "loss": 4.8445,
      "step": 26100
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.432464063167572,
      "learning_rate": 4.2820977281159294e-06,
      "loss": 4.8469,
      "step": 26200
    },
    {
      "epoch": 0.57,
      "grad_norm": 1.313357949256897,
      "learning_rate": 4.26027367364309e-06,
      "loss": 4.8451,
      "step": 26300
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.1689280271530151,
      "learning_rate": 4.23844961917025e-06,
      "loss": 4.8441,
      "step": 26400
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.7955917119979858,
      "learning_rate": 4.21662556469741e-06,
      "loss": 4.8446,
      "step": 26500
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.5141814947128296,
      "learning_rate": 4.1948015102245694e-06,
      "loss": 4.848,
      "step": 26600
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.512915849685669,
      "learning_rate": 4.17297745575173e-06,
      "loss": 4.8441,
      "step": 26700
    },
    {
      "epoch": 0.58,
      "grad_norm": 3.6091654300689697,
      "learning_rate": 4.15115340127889e-06,
      "loss": 4.8445,
      "step": 26800
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.6325381994247437,
      "learning_rate": 4.12932934680605e-06,
      "loss": 4.8447,
      "step": 26900
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.8864593505859375,
      "learning_rate": 4.107505292333209e-06,
      "loss": 4.8451,
      "step": 27000
    },
    {
      "epoch": 0.59,
      "grad_norm": 2.2388365268707275,
      "learning_rate": 4.08568123786037e-06,
      "loss": 4.8419,
      "step": 27100
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.5139906406402588,
      "learning_rate": 4.06385718338753e-06,
      "loss": 4.8436,
      "step": 27200
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.44464704394340515,
      "learning_rate": 4.04203312891469e-06,
      "loss": 4.8455,
      "step": 27300
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.5341111421585083,
      "learning_rate": 4.02020907444185e-06,
      "loss": 4.842,
      "step": 27400
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.6707934141159058,
      "learning_rate": 3.99838501996901e-06,
      "loss": 4.8473,
      "step": 27500
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.842087984085083,
      "learning_rate": 3.97656096549617e-06,
      "loss": 4.845,
      "step": 27600
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.575089693069458,
      "learning_rate": 3.95473691102333e-06,
      "loss": 4.8432,
      "step": 27700
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.7443206310272217,
      "learning_rate": 3.93291285655049e-06,
      "loss": 4.8426,
      "step": 27800
    },
    {
      "epoch": 0.61,
      "grad_norm": 3.0586435794830322,
      "learning_rate": 3.9110888020776504e-06,
      "loss": 4.8458,
      "step": 27900
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.6179745197296143,
      "learning_rate": 3.88926474760481e-06,
      "loss": 4.8441,
      "step": 28000
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.7903715372085571,
      "learning_rate": 3.86744069313197e-06,
      "loss": 4.8412,
      "step": 28100
    },
    {
      "epoch": 0.62,
      "grad_norm": 2.2073774337768555,
      "learning_rate": 3.84561663865913e-06,
      "loss": 4.8416,
      "step": 28200
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.7796311974525452,
      "learning_rate": 3.82379258418629e-06,
      "loss": 4.8427,
      "step": 28300
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.6196668148040771,
      "learning_rate": 3.8019685297134506e-06,
      "loss": 4.8421,
      "step": 28400
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.8816477656364441,
      "learning_rate": 3.7801444752406104e-06,
      "loss": 4.8424,
      "step": 28500
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.104554295539856,
      "learning_rate": 3.75832042076777e-06,
      "loss": 4.8429,
      "step": 28600
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.6221787333488464,
      "learning_rate": 3.7364963662949304e-06,
      "loss": 4.8439,
      "step": 28700
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.6664981842041016,
      "learning_rate": 3.7146723118220906e-06,
      "loss": 4.8443,
      "step": 28800
    },
    {
      "epoch": 0.63,
      "grad_norm": 2.061434268951416,
      "learning_rate": 3.692848257349251e-06,
      "loss": 4.8422,
      "step": 28900
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.919090747833252,
      "learning_rate": 3.671024202876411e-06,
      "loss": 4.841,
      "step": 29000
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.7256101369857788,
      "learning_rate": 3.6492001484035704e-06,
      "loss": 4.8413,
      "step": 29100
    },
    {
      "epoch": 0.64,
      "grad_norm": 2.418320894241333,
      "learning_rate": 3.6273760939307306e-06,
      "loss": 4.8434,
      "step": 29200
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.3488261699676514,
      "learning_rate": 3.605552039457891e-06,
      "loss": 4.8436,
      "step": 29300
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.9698038697242737,
      "learning_rate": 3.583727984985051e-06,
      "loss": 4.8434,
      "step": 29400
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.685869812965393,
      "learning_rate": 3.561903930512211e-06,
      "loss": 4.8398,
      "step": 29500
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.8971433639526367,
      "learning_rate": 3.5400798760393706e-06,
      "loss": 4.841,
      "step": 29600
    },
    {
      "epoch": 0.65,
      "grad_norm": 2.2868356704711914,
      "learning_rate": 3.5182558215665308e-06,
      "loss": 4.8409,
      "step": 29700
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.8014007210731506,
      "learning_rate": 3.496431767093691e-06,
      "loss": 4.841,
      "step": 29800
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.7061727643013,
      "learning_rate": 3.474607712620851e-06,
      "loss": 4.8433,
      "step": 29900
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.7472312450408936,
      "learning_rate": 3.4527836581480114e-06,
      "loss": 4.8423,
      "step": 30000
    },
    {
      "epoch": 0.66,
      "grad_norm": 2.389944076538086,
      "learning_rate": 3.4309596036751708e-06,
      "loss": 4.8404,
      "step": 30100
    },
    {
      "epoch": 0.66,
      "grad_norm": 2.251760959625244,
      "learning_rate": 3.409135549202331e-06,
      "loss": 4.8452,
      "step": 30200
    },
    {
      "epoch": 0.66,
      "grad_norm": 1.1334567070007324,
      "learning_rate": 3.387311494729491e-06,
      "loss": 4.8425,
      "step": 30300
    },
    {
      "epoch": 0.66,
      "grad_norm": 1.1134765148162842,
      "learning_rate": 3.3654874402566514e-06,
      "loss": 4.8393,
      "step": 30400
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.4619927704334259,
      "learning_rate": 3.343663385783811e-06,
      "loss": 4.8415,
      "step": 30500
    },
    {
      "epoch": 0.67,
      "grad_norm": 2.933467388153076,
      "learning_rate": 3.3218393313109714e-06,
      "loss": 4.8399,
      "step": 30600
    },
    {
      "epoch": 0.67,
      "grad_norm": 1.7002770900726318,
      "learning_rate": 3.300015276838131e-06,
      "loss": 4.8417,
      "step": 30700
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.8919426798820496,
      "learning_rate": 3.2781912223652914e-06,
      "loss": 4.8388,
      "step": 30800
    },
    {
      "epoch": 0.67,
      "grad_norm": 1.6836445331573486,
      "learning_rate": 3.256367167892451e-06,
      "loss": 4.8386,
      "step": 30900
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.4594995975494385,
      "learning_rate": 3.2345431134196114e-06,
      "loss": 4.8399,
      "step": 31000
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.7272505164146423,
      "learning_rate": 3.2127190589467716e-06,
      "loss": 4.8397,
      "step": 31100
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.1747781038284302,
      "learning_rate": 3.1908950044739313e-06,
      "loss": 4.8405,
      "step": 31200
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.6341750621795654,
      "learning_rate": 3.169070950001091e-06,
      "loss": 4.8415,
      "step": 31300
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.8783094882965088,
      "learning_rate": 3.1472468955282513e-06,
      "loss": 4.8351,
      "step": 31400
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.6762621402740479,
      "learning_rate": 3.1254228410554115e-06,
      "loss": 4.8446,
      "step": 31500
    },
    {
      "epoch": 0.69,
      "grad_norm": 1.1475865840911865,
      "learning_rate": 3.1035987865825718e-06,
      "loss": 4.8397,
      "step": 31600
    },
    {
      "epoch": 0.69,
      "grad_norm": 1.4153202772140503,
      "learning_rate": 3.0817747321097315e-06,
      "loss": 4.84,
      "step": 31700
    },
    {
      "epoch": 0.69,
      "grad_norm": 3.4596126079559326,
      "learning_rate": 3.0599506776368913e-06,
      "loss": 4.8424,
      "step": 31800
    },
    {
      "epoch": 0.7,
      "grad_norm": 1.4196875095367432,
      "learning_rate": 3.0381266231640515e-06,
      "loss": 4.8418,
      "step": 31900
    },
    {
      "epoch": 0.7,
      "grad_norm": 3.9644806385040283,
      "learning_rate": 3.0163025686912117e-06,
      "loss": 4.8414,
      "step": 32000
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.6097564101219177,
      "learning_rate": 2.994478514218372e-06,
      "loss": 4.8401,
      "step": 32100
    },
    {
      "epoch": 0.7,
      "grad_norm": 1.8151260614395142,
      "learning_rate": 2.972654459745532e-06,
      "loss": 4.8427,
      "step": 32200
    },
    {
      "epoch": 0.7,
      "grad_norm": 2.938615083694458,
      "learning_rate": 2.9508304052726915e-06,
      "loss": 4.8408,
      "step": 32300
    },
    {
      "epoch": 0.71,
      "grad_norm": 3.4757254123687744,
      "learning_rate": 2.9290063507998517e-06,
      "loss": 4.8429,
      "step": 32400
    },
    {
      "epoch": 0.71,
      "grad_norm": 1.0553146600723267,
      "learning_rate": 2.907182296327012e-06,
      "loss": 4.8412,
      "step": 32500
    },
    {
      "epoch": 0.71,
      "grad_norm": 2.4443321228027344,
      "learning_rate": 2.885358241854172e-06,
      "loss": 4.8405,
      "step": 32600
    },
    {
      "epoch": 0.71,
      "grad_norm": 1.0173979997634888,
      "learning_rate": 2.8635341873813323e-06,
      "loss": 4.8367,
      "step": 32700
    },
    {
      "epoch": 0.72,
      "grad_norm": 3.2443580627441406,
      "learning_rate": 2.8417101329084917e-06,
      "loss": 4.8393,
      "step": 32800
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.9010906219482422,
      "learning_rate": 2.819886078435652e-06,
      "loss": 4.8391,
      "step": 32900
    },
    {
      "epoch": 0.72,
      "grad_norm": 3.2947003841400146,
      "learning_rate": 2.798062023962812e-06,
      "loss": 4.8401,
      "step": 33000
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.9409279227256775,
      "learning_rate": 2.7762379694899723e-06,
      "loss": 4.8389,
      "step": 33100
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.5245996713638306,
      "learning_rate": 2.754413915017132e-06,
      "loss": 4.8392,
      "step": 33200
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.031678318977356,
      "learning_rate": 2.732589860544292e-06,
      "loss": 4.8406,
      "step": 33300
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.3470673561096191,
      "learning_rate": 2.710765806071452e-06,
      "loss": 4.8404,
      "step": 33400
    },
    {
      "epoch": 0.73,
      "grad_norm": 2.0202999114990234,
      "learning_rate": 2.6889417515986123e-06,
      "loss": 4.8409,
      "step": 33500
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.7307863235473633,
      "learning_rate": 2.667117697125772e-06,
      "loss": 4.8378,
      "step": 33600
    },
    {
      "epoch": 0.74,
      "grad_norm": 1.679755687713623,
      "learning_rate": 2.6452936426529323e-06,
      "loss": 4.8388,
      "step": 33700
    },
    {
      "epoch": 0.74,
      "grad_norm": 2.951936721801758,
      "learning_rate": 2.6234695881800925e-06,
      "loss": 4.8403,
      "step": 33800
    },
    {
      "epoch": 0.74,
      "grad_norm": 1.8046462535858154,
      "learning_rate": 2.6016455337072523e-06,
      "loss": 4.838,
      "step": 33900
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.8763813972473145,
      "learning_rate": 2.579821479234412e-06,
      "loss": 4.8393,
      "step": 34000
    },
    {
      "epoch": 0.74,
      "grad_norm": 1.4633421897888184,
      "learning_rate": 2.5579974247615723e-06,
      "loss": 4.8403,
      "step": 34100
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.8257120847702026,
      "learning_rate": 2.5361733702887325e-06,
      "loss": 4.8364,
      "step": 34200
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.8002654314041138,
      "learning_rate": 2.5143493158158927e-06,
      "loss": 4.8422,
      "step": 34300
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.5233999490737915,
      "learning_rate": 2.4925252613430525e-06,
      "loss": 4.838,
      "step": 34400
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.0047553777694702,
      "learning_rate": 2.4707012068702123e-06,
      "loss": 4.8406,
      "step": 34500
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.5430443286895752,
      "learning_rate": 2.4488771523973725e-06,
      "loss": 4.8385,
      "step": 34600
    },
    {
      "epoch": 0.76,
      "grad_norm": 1.5413551330566406,
      "learning_rate": 2.4270530979245327e-06,
      "loss": 4.8377,
      "step": 34700
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.4709758162498474,
      "learning_rate": 2.4052290434516925e-06,
      "loss": 4.8364,
      "step": 34800
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.9328405857086182,
      "learning_rate": 2.3834049889788527e-06,
      "loss": 4.8389,
      "step": 34900
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.6147871017456055,
      "learning_rate": 2.361580934506013e-06,
      "loss": 4.8382,
      "step": 35000
    },
    {
      "epoch": 0.77,
      "grad_norm": 2.017819404602051,
      "learning_rate": 2.3397568800331727e-06,
      "loss": 4.838,
      "step": 35100
    },
    {
      "epoch": 0.77,
      "grad_norm": 2.1005585193634033,
      "learning_rate": 2.317932825560333e-06,
      "loss": 4.8383,
      "step": 35200
    },
    {
      "epoch": 0.77,
      "grad_norm": 1.5119142532348633,
      "learning_rate": 2.2961087710874927e-06,
      "loss": 4.8354,
      "step": 35300
    },
    {
      "epoch": 0.77,
      "grad_norm": 1.8534014225006104,
      "learning_rate": 2.274284716614653e-06,
      "loss": 4.8387,
      "step": 35400
    },
    {
      "epoch": 0.77,
      "grad_norm": 1.2567205429077148,
      "learning_rate": 2.252460662141813e-06,
      "loss": 4.8392,
      "step": 35500
    },
    {
      "epoch": 0.78,
      "grad_norm": 2.021512508392334,
      "learning_rate": 2.230636607668973e-06,
      "loss": 4.8365,
      "step": 35600
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.7630836367607117,
      "learning_rate": 2.208812553196133e-06,
      "loss": 4.8394,
      "step": 35700
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.742551326751709,
      "learning_rate": 2.1869884987232933e-06,
      "loss": 4.8398,
      "step": 35800
    },
    {
      "epoch": 0.78,
      "grad_norm": 3.5478837490081787,
      "learning_rate": 2.165164444250453e-06,
      "loss": 4.8369,
      "step": 35900
    },
    {
      "epoch": 0.79,
      "grad_norm": 3.5186381340026855,
      "learning_rate": 2.1433403897776133e-06,
      "loss": 4.8407,
      "step": 36000
    },
    {
      "epoch": 0.79,
      "grad_norm": 2.9799013137817383,
      "learning_rate": 2.121516335304773e-06,
      "loss": 4.8404,
      "step": 36100
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.6745467782020569,
      "learning_rate": 2.0996922808319333e-06,
      "loss": 4.8361,
      "step": 36200
    },
    {
      "epoch": 0.79,
      "grad_norm": 2.257988452911377,
      "learning_rate": 2.077868226359093e-06,
      "loss": 4.8391,
      "step": 36300
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.193894863128662,
      "learning_rate": 2.0560441718862532e-06,
      "loss": 4.8372,
      "step": 36400
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.0707862377166748,
      "learning_rate": 2.0342201174134134e-06,
      "loss": 4.8386,
      "step": 36500
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.415719747543335,
      "learning_rate": 2.0123960629405732e-06,
      "loss": 4.8397,
      "step": 36600
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.7595218420028687,
      "learning_rate": 1.9905720084677334e-06,
      "loss": 4.8362,
      "step": 36700
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.733424186706543,
      "learning_rate": 1.9687479539948932e-06,
      "loss": 4.8387,
      "step": 36800
    },
    {
      "epoch": 0.81,
      "grad_norm": 2.846148729324341,
      "learning_rate": 1.9469238995220534e-06,
      "loss": 4.8351,
      "step": 36900
    },
    {
      "epoch": 0.81,
      "grad_norm": 3.0567848682403564,
      "learning_rate": 1.9250998450492132e-06,
      "loss": 4.8384,
      "step": 37000
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.4598780870437622,
      "learning_rate": 1.9032757905763734e-06,
      "loss": 4.8391,
      "step": 37100
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.080013394355774,
      "learning_rate": 1.8814517361035334e-06,
      "loss": 4.839,
      "step": 37200
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.5742350816726685,
      "learning_rate": 1.8596276816306934e-06,
      "loss": 4.8415,
      "step": 37300
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.509831428527832,
      "learning_rate": 1.8378036271578536e-06,
      "loss": 4.8383,
      "step": 37400
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.7902163863182068,
      "learning_rate": 1.8159795726850136e-06,
      "loss": 4.8383,
      "step": 37500
    },
    {
      "epoch": 0.82,
      "grad_norm": 1.2146787643432617,
      "learning_rate": 1.7941555182121736e-06,
      "loss": 4.839,
      "step": 37600
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.5992825031280518,
      "learning_rate": 1.7723314637393336e-06,
      "loss": 4.8342,
      "step": 37700
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.7323722839355469,
      "learning_rate": 1.7505074092664936e-06,
      "loss": 4.8384,
      "step": 37800
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.4978944659233093,
      "learning_rate": 1.7286833547936538e-06,
      "loss": 4.838,
      "step": 37900
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.597199022769928,
      "learning_rate": 1.7068593003208136e-06,
      "loss": 4.8345,
      "step": 38000
    },
    {
      "epoch": 0.83,
      "grad_norm": 2.2732481956481934,
      "learning_rate": 1.6850352458479738e-06,
      "loss": 4.8382,
      "step": 38100
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.6661024689674377,
      "learning_rate": 1.663211191375134e-06,
      "loss": 4.8395,
      "step": 38200
    },
    {
      "epoch": 0.84,
      "grad_norm": 1.767288088798523,
      "learning_rate": 1.6413871369022938e-06,
      "loss": 4.838,
      "step": 38300
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.47860872745513916,
      "learning_rate": 1.619563082429454e-06,
      "loss": 4.839,
      "step": 38400
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.5148721933364868,
      "learning_rate": 1.5977390279566138e-06,
      "loss": 4.839,
      "step": 38500
    },
    {
      "epoch": 0.84,
      "grad_norm": 1.728582739830017,
      "learning_rate": 1.575914973483774e-06,
      "loss": 4.8347,
      "step": 38600
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.9544780254364014,
      "learning_rate": 1.554090919010934e-06,
      "loss": 4.8369,
      "step": 38700
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.7922559976577759,
      "learning_rate": 1.532266864538094e-06,
      "loss": 4.8385,
      "step": 38800
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.432121217250824,
      "learning_rate": 1.510442810065254e-06,
      "loss": 4.8376,
      "step": 38900
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.711795449256897,
      "learning_rate": 1.4886187555924142e-06,
      "loss": 4.8391,
      "step": 39000
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.8019421696662903,
      "learning_rate": 1.466794701119574e-06,
      "loss": 4.8375,
      "step": 39100
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.8278266191482544,
      "learning_rate": 1.4449706466467342e-06,
      "loss": 4.8342,
      "step": 39200
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.6380483508110046,
      "learning_rate": 1.423146592173894e-06,
      "loss": 4.838,
      "step": 39300
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.9113682508468628,
      "learning_rate": 1.4013225377010542e-06,
      "loss": 4.8362,
      "step": 39400
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.5136518478393555,
      "learning_rate": 1.3794984832282144e-06,
      "loss": 4.838,
      "step": 39500
    },
    {
      "epoch": 0.86,
      "grad_norm": 1.0487759113311768,
      "learning_rate": 1.3576744287553742e-06,
      "loss": 4.8389,
      "step": 39600
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.9537123441696167,
      "learning_rate": 1.3358503742825344e-06,
      "loss": 4.8338,
      "step": 39700
    },
    {
      "epoch": 0.87,
      "grad_norm": 1.4224497079849243,
      "learning_rate": 1.3140263198096944e-06,
      "loss": 4.8358,
      "step": 39800
    },
    {
      "epoch": 0.87,
      "grad_norm": 1.1593801975250244,
      "learning_rate": 1.2922022653368544e-06,
      "loss": 4.8331,
      "step": 39900
    },
    {
      "epoch": 0.87,
      "grad_norm": 1.5574233531951904,
      "learning_rate": 1.2703782108640146e-06,
      "loss": 4.8351,
      "step": 40000
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.2431851625442505,
      "learning_rate": 1.2485541563911746e-06,
      "loss": 4.8391,
      "step": 40100
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.5098065137863159,
      "learning_rate": 1.2267301019183346e-06,
      "loss": 4.8389,
      "step": 40200
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.690344512462616,
      "learning_rate": 1.2049060474454946e-06,
      "loss": 4.8387,
      "step": 40300
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.661180019378662,
      "learning_rate": 1.1830819929726546e-06,
      "loss": 4.8373,
      "step": 40400
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.4016964435577393,
      "learning_rate": 1.1612579384998145e-06,
      "loss": 4.8373,
      "step": 40500
    },
    {
      "epoch": 0.89,
      "grad_norm": 1.7542119026184082,
      "learning_rate": 1.1394338840269745e-06,
      "loss": 4.8359,
      "step": 40600
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.7012521624565125,
      "learning_rate": 1.1176098295541345e-06,
      "loss": 4.8367,
      "step": 40700
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.8252130746841431,
      "learning_rate": 1.0957857750812947e-06,
      "loss": 4.8377,
      "step": 40800
    },
    {
      "epoch": 0.89,
      "grad_norm": 1.0341030359268188,
      "learning_rate": 1.0739617206084547e-06,
      "loss": 4.8374,
      "step": 40900
    },
    {
      "epoch": 0.89,
      "grad_norm": 1.5905879735946655,
      "learning_rate": 1.0521376661356147e-06,
      "loss": 4.84,
      "step": 41000
    },
    {
      "epoch": 0.9,
      "grad_norm": 1.8270330429077148,
      "learning_rate": 1.0303136116627747e-06,
      "loss": 4.8343,
      "step": 41100
    },
    {
      "epoch": 0.9,
      "grad_norm": 1.5131927728652954,
      "learning_rate": 1.008489557189935e-06,
      "loss": 4.8367,
      "step": 41200
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.9526300430297852,
      "learning_rate": 9.86665502717095e-07,
      "loss": 4.8365,
      "step": 41300
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.43970614671707153,
      "learning_rate": 9.64841448244255e-07,
      "loss": 4.8353,
      "step": 41400
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.5485687255859375,
      "learning_rate": 9.430173937714149e-07,
      "loss": 4.8375,
      "step": 41500
    },
    {
      "epoch": 0.91,
      "grad_norm": 1.4453191757202148,
      "learning_rate": 9.21193339298575e-07,
      "loss": 4.8366,
      "step": 41600
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.5337852835655212,
      "learning_rate": 8.99369284825735e-07,
      "loss": 4.8376,
      "step": 41700
    },
    {
      "epoch": 0.91,
      "grad_norm": 1.0073989629745483,
      "learning_rate": 8.77545230352895e-07,
      "loss": 4.8382,
      "step": 41800
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.8360026478767395,
      "learning_rate": 8.55721175880055e-07,
      "loss": 4.838,
      "step": 41900
    },
    {
      "epoch": 0.92,
      "grad_norm": 2.3463079929351807,
      "learning_rate": 8.338971214072151e-07,
      "loss": 4.8379,
      "step": 42000
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.6060824990272522,
      "learning_rate": 8.120730669343751e-07,
      "loss": 4.8386,
      "step": 42100
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.7597670555114746,
      "learning_rate": 7.902490124615351e-07,
      "loss": 4.8373,
      "step": 42200
    },
    {
      "epoch": 0.92,
      "grad_norm": 1.566273808479309,
      "learning_rate": 7.684249579886951e-07,
      "loss": 4.8366,
      "step": 42300
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.5708392262458801,
      "learning_rate": 7.466009035158553e-07,
      "loss": 4.8363,
      "step": 42400
    },
    {
      "epoch": 0.93,
      "grad_norm": 1.2547303438186646,
      "learning_rate": 7.247768490430153e-07,
      "loss": 4.8382,
      "step": 42500
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.6009268760681152,
      "learning_rate": 7.029527945701753e-07,
      "loss": 4.8361,
      "step": 42600
    },
    {
      "epoch": 0.93,
      "grad_norm": 1.8468096256256104,
      "learning_rate": 6.811287400973353e-07,
      "loss": 4.8372,
      "step": 42700
    },
    {
      "epoch": 0.93,
      "grad_norm": 2.3321032524108887,
      "learning_rate": 6.593046856244954e-07,
      "loss": 4.8369,
      "step": 42800
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.5124238729476929,
      "learning_rate": 6.374806311516554e-07,
      "loss": 4.8382,
      "step": 42900
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.6034313440322876,
      "learning_rate": 6.156565766788154e-07,
      "loss": 4.8393,
      "step": 43000
    },
    {
      "epoch": 0.94,
      "grad_norm": 3.0424869060516357,
      "learning_rate": 5.938325222059755e-07,
      "loss": 4.8362,
      "step": 43100
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.9956300854682922,
      "learning_rate": 5.720084677331355e-07,
      "loss": 4.8358,
      "step": 43200
    },
    {
      "epoch": 0.94,
      "grad_norm": 1.2779983282089233,
      "learning_rate": 5.501844132602956e-07,
      "loss": 4.8366,
      "step": 43300
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.522408664226532,
      "learning_rate": 5.283603587874556e-07,
      "loss": 4.8345,
      "step": 43400
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.5073395371437073,
      "learning_rate": 5.065363043146156e-07,
      "loss": 4.8362,
      "step": 43500
    }
  ],
  "logging_steps": 100,
  "max_steps": 45821,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 500,
  "total_flos": 4.0839050005128806e+17,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}