{
  "best_metric": 1.1857857704162598,
  "best_model_checkpoint": "runs/deepseek-full-hard-low-lr/checkpoint-92500",
  "epoch": 2.5,
  "eval_steps": 2500,
  "global_step": 100000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "grad_norm": 24.515625,
      "learning_rate": 6e-06,
      "loss": 1.9991,
      "step": 100
    },
    {
      "epoch": 0.01,
      "grad_norm": 22.859375,
      "learning_rate": 1.2e-05,
      "loss": 1.8155,
      "step": 200
    },
    {
      "epoch": 0.01,
      "grad_norm": 8.25,
      "learning_rate": 1.8e-05,
      "loss": 1.8039,
      "step": 300
    },
    {
      "epoch": 0.01,
      "grad_norm": 8.0234375,
      "learning_rate": 2.4e-05,
      "loss": 1.8399,
      "step": 400
    },
    {
      "epoch": 0.01,
      "grad_norm": 17.375,
      "learning_rate": 3e-05,
      "loss": 1.8311,
      "step": 500
    },
    {
      "epoch": 0.01,
      "grad_norm": 11.421875,
      "learning_rate": 2.9969849246231156e-05,
      "loss": 1.8306,
      "step": 600
    },
    {
      "epoch": 0.02,
      "grad_norm": 12.40625,
      "learning_rate": 2.993969849246231e-05,
      "loss": 1.8191,
      "step": 700
    },
    {
      "epoch": 0.02,
      "grad_norm": 4.71484375,
      "learning_rate": 2.990954773869347e-05,
      "loss": 1.8153,
      "step": 800
    },
    {
      "epoch": 0.02,
      "grad_norm": 8.1796875,
      "learning_rate": 2.9879396984924625e-05,
      "loss": 1.8115,
      "step": 900
    },
    {
      "epoch": 0.03,
      "grad_norm": 12.7734375,
      "learning_rate": 2.984924623115578e-05,
      "loss": 1.8093,
      "step": 1000
    },
    {
      "epoch": 0.03,
      "grad_norm": 7.9140625,
      "learning_rate": 2.9819095477386932e-05,
      "loss": 1.8475,
      "step": 1100
    },
    {
      "epoch": 0.03,
      "grad_norm": 4.83203125,
      "learning_rate": 2.978894472361809e-05,
      "loss": 1.7814,
      "step": 1200
    },
    {
      "epoch": 0.03,
      "grad_norm": 10.2421875,
      "learning_rate": 2.9758793969849246e-05,
      "loss": 1.7792,
      "step": 1300
    },
    {
      "epoch": 0.04,
      "grad_norm": 5.38671875,
      "learning_rate": 2.97286432160804e-05,
      "loss": 1.7633,
      "step": 1400
    },
    {
      "epoch": 0.04,
      "grad_norm": 7.8515625,
      "learning_rate": 2.9698492462311557e-05,
      "loss": 1.8018,
      "step": 1500
    },
    {
      "epoch": 0.04,
      "grad_norm": 7.8671875,
      "learning_rate": 2.9668341708542715e-05,
      "loss": 1.7645,
      "step": 1600
    },
    {
      "epoch": 0.04,
      "grad_norm": 6.91015625,
      "learning_rate": 2.963819095477387e-05,
      "loss": 1.7771,
      "step": 1700
    },
    {
      "epoch": 0.04,
      "grad_norm": 4.61328125,
      "learning_rate": 2.9608040201005026e-05,
      "loss": 1.7965,
      "step": 1800
    },
    {
      "epoch": 0.05,
      "grad_norm": 8.6796875,
      "learning_rate": 2.957788944723618e-05,
      "loss": 1.767,
      "step": 1900
    },
    {
      "epoch": 0.05,
      "grad_norm": 6.11328125,
      "learning_rate": 2.954773869346734e-05,
      "loss": 1.7448,
      "step": 2000
    },
    {
      "epoch": 0.05,
      "grad_norm": 4.79296875,
      "learning_rate": 2.9517587939698495e-05,
      "loss": 1.7822,
      "step": 2100
    },
    {
      "epoch": 0.06,
      "grad_norm": 6.8046875,
      "learning_rate": 2.9487437185929647e-05,
      "loss": 1.7997,
      "step": 2200
    },
    {
      "epoch": 0.06,
      "grad_norm": 5.39453125,
      "learning_rate": 2.9457286432160802e-05,
      "loss": 1.7795,
      "step": 2300
    },
    {
      "epoch": 0.06,
      "grad_norm": 9.2578125,
      "learning_rate": 2.942713567839196e-05,
      "loss": 1.7529,
      "step": 2400
    },
    {
      "epoch": 0.06,
      "grad_norm": 8.8828125,
      "learning_rate": 2.9396984924623116e-05,
      "loss": 1.7433,
      "step": 2500
    },
    {
      "epoch": 0.06,
      "eval_loss": 1.7661924362182617,
      "eval_runtime": 58.5928,
      "eval_samples_per_second": 17.067,
      "eval_steps_per_second": 4.267,
      "step": 2500
    },
    {
      "epoch": 0.07,
      "grad_norm": 7.8984375,
      "learning_rate": 2.936683417085427e-05,
      "loss": 1.7891,
      "step": 2600
    },
    {
      "epoch": 0.07,
      "grad_norm": 9.3828125,
      "learning_rate": 2.9336683417085427e-05,
      "loss": 1.7309,
      "step": 2700
    },
    {
      "epoch": 0.07,
      "grad_norm": 6.703125,
      "learning_rate": 2.9306532663316585e-05,
      "loss": 1.6948,
      "step": 2800
    },
    {
      "epoch": 0.07,
      "grad_norm": 9.546875,
      "learning_rate": 2.927638190954774e-05,
      "loss": 1.7142,
      "step": 2900
    },
    {
      "epoch": 0.07,
      "grad_norm": 8.59375,
      "learning_rate": 2.9246231155778896e-05,
      "loss": 1.7304,
      "step": 3000
    },
    {
      "epoch": 0.08,
      "grad_norm": 5.9375,
      "learning_rate": 2.921608040201005e-05,
      "loss": 1.7201,
      "step": 3100
    },
    {
      "epoch": 0.08,
      "grad_norm": 11.7734375,
      "learning_rate": 2.9185929648241207e-05,
      "loss": 1.7399,
      "step": 3200
    },
    {
      "epoch": 0.08,
      "grad_norm": 10.2578125,
      "learning_rate": 2.9155778894472365e-05,
      "loss": 1.7489,
      "step": 3300
    },
    {
      "epoch": 0.09,
      "grad_norm": 5.3671875,
      "learning_rate": 2.9125628140703517e-05,
      "loss": 1.7593,
      "step": 3400
    },
    {
      "epoch": 0.09,
      "grad_norm": 4.7890625,
      "learning_rate": 2.9095477386934672e-05,
      "loss": 1.7298,
      "step": 3500
    },
    {
      "epoch": 0.09,
      "grad_norm": 5.6328125,
      "learning_rate": 2.9065326633165828e-05,
      "loss": 1.727,
      "step": 3600
    },
    {
      "epoch": 0.09,
      "grad_norm": 9.0546875,
      "learning_rate": 2.9035175879396986e-05,
      "loss": 1.7101,
      "step": 3700
    },
    {
      "epoch": 0.1,
      "grad_norm": 6.26171875,
      "learning_rate": 2.900502512562814e-05,
      "loss": 1.728,
      "step": 3800
    },
    {
      "epoch": 0.1,
      "grad_norm": 8.0078125,
      "learning_rate": 2.8974874371859297e-05,
      "loss": 1.7085,
      "step": 3900
    },
    {
      "epoch": 0.1,
      "grad_norm": 5.2265625,
      "learning_rate": 2.8944723618090452e-05,
      "loss": 1.7097,
      "step": 4000
    },
    {
      "epoch": 0.1,
      "grad_norm": 5.74609375,
      "learning_rate": 2.891457286432161e-05,
      "loss": 1.7237,
      "step": 4100
    },
    {
      "epoch": 0.1,
      "grad_norm": 8.3203125,
      "learning_rate": 2.8884422110552766e-05,
      "loss": 1.7338,
      "step": 4200
    },
    {
      "epoch": 0.11,
      "grad_norm": 7.015625,
      "learning_rate": 2.885427135678392e-05,
      "loss": 1.731,
      "step": 4300
    },
    {
      "epoch": 0.11,
      "grad_norm": 7.00390625,
      "learning_rate": 2.8824120603015077e-05,
      "loss": 1.7338,
      "step": 4400
    },
    {
      "epoch": 0.11,
      "grad_norm": 4.61328125,
      "learning_rate": 2.8793969849246232e-05,
      "loss": 1.7275,
      "step": 4500
    },
    {
      "epoch": 0.12,
      "grad_norm": 7.4375,
      "learning_rate": 2.8763819095477387e-05,
      "loss": 1.737,
      "step": 4600
    },
    {
      "epoch": 0.12,
      "grad_norm": 9.625,
      "learning_rate": 2.8733668341708542e-05,
      "loss": 1.6846,
      "step": 4700
    },
    {
      "epoch": 0.12,
      "grad_norm": 6.83984375,
      "learning_rate": 2.8703517587939698e-05,
      "loss": 1.7758,
      "step": 4800
    },
    {
      "epoch": 0.12,
      "grad_norm": 5.9609375,
      "learning_rate": 2.8673366834170856e-05,
      "loss": 1.6865,
      "step": 4900
    },
    {
      "epoch": 0.12,
      "grad_norm": 7.47265625,
      "learning_rate": 2.864321608040201e-05,
      "loss": 1.7066,
      "step": 5000
    },
    {
      "epoch": 0.12,
      "eval_loss": 1.6524672508239746,
      "eval_runtime": 58.7554,
      "eval_samples_per_second": 17.02,
      "eval_steps_per_second": 4.255,
      "step": 5000
    },
    {
      "epoch": 0.13,
      "grad_norm": 7.13671875,
      "learning_rate": 2.8613065326633167e-05,
      "loss": 1.6966,
      "step": 5100
    },
    {
      "epoch": 0.13,
      "grad_norm": 8.0,
      "learning_rate": 2.8582914572864322e-05,
      "loss": 1.7145,
      "step": 5200
    },
    {
      "epoch": 0.13,
      "grad_norm": 7.125,
      "learning_rate": 2.8552763819095477e-05,
      "loss": 1.7035,
      "step": 5300
    },
    {
      "epoch": 0.14,
      "grad_norm": 9.4453125,
      "learning_rate": 2.8522613065326636e-05,
      "loss": 1.7273,
      "step": 5400
    },
    {
      "epoch": 0.14,
      "grad_norm": 6.609375,
      "learning_rate": 2.849246231155779e-05,
      "loss": 1.685,
      "step": 5500
    },
    {
      "epoch": 0.14,
      "grad_norm": 4.6171875,
      "learning_rate": 2.8462311557788943e-05,
      "loss": 1.6505,
      "step": 5600
    },
    {
      "epoch": 0.14,
      "grad_norm": 7.92578125,
      "learning_rate": 2.84321608040201e-05,
      "loss": 1.7042,
      "step": 5700
    },
    {
      "epoch": 0.14,
      "grad_norm": 6.94140625,
      "learning_rate": 2.8402010050251257e-05,
      "loss": 1.6963,
      "step": 5800
    },
    {
      "epoch": 0.15,
      "grad_norm": 7.96484375,
      "learning_rate": 2.8371859296482412e-05,
      "loss": 1.7243,
      "step": 5900
    },
    {
      "epoch": 0.15,
      "grad_norm": 7.5078125,
      "learning_rate": 2.8341708542713568e-05,
      "loss": 1.6677,
      "step": 6000
    },
    {
      "epoch": 0.15,
      "grad_norm": 7.20703125,
      "learning_rate": 2.8311557788944723e-05,
      "loss": 1.679,
      "step": 6100
    },
    {
      "epoch": 0.15,
      "grad_norm": 4.51953125,
      "learning_rate": 2.828140703517588e-05,
      "loss": 1.6874,
      "step": 6200
    },
    {
      "epoch": 0.16,
      "grad_norm": 9.5546875,
      "learning_rate": 2.8251256281407037e-05,
      "loss": 1.6843,
      "step": 6300
    },
    {
      "epoch": 0.16,
      "grad_norm": 6.48046875,
      "learning_rate": 2.8221105527638192e-05,
      "loss": 1.7014,
      "step": 6400
    },
    {
      "epoch": 0.16,
      "grad_norm": 7.5546875,
      "learning_rate": 2.8190954773869347e-05,
      "loss": 1.6564,
      "step": 6500
    },
    {
      "epoch": 0.17,
      "grad_norm": 6.37890625,
      "learning_rate": 2.8160804020100506e-05,
      "loss": 1.7107,
      "step": 6600
    },
    {
      "epoch": 0.17,
      "grad_norm": 9.5,
      "learning_rate": 2.813065326633166e-05,
      "loss": 1.7179,
      "step": 6700
    },
    {
      "epoch": 0.17,
      "grad_norm": 5.59765625,
      "learning_rate": 2.8100502512562813e-05,
      "loss": 1.6851,
      "step": 6800
    },
    {
      "epoch": 0.17,
      "grad_norm": 7.33984375,
      "learning_rate": 2.807035175879397e-05,
      "loss": 1.6718,
      "step": 6900
    },
    {
      "epoch": 0.17,
      "grad_norm": 7.5,
      "learning_rate": 2.8040201005025127e-05,
      "loss": 1.6659,
      "step": 7000
    },
    {
      "epoch": 0.18,
      "grad_norm": 9.90625,
      "learning_rate": 2.8010050251256282e-05,
      "loss": 1.6703,
      "step": 7100
    },
    {
      "epoch": 0.18,
      "grad_norm": 7.30078125,
      "learning_rate": 2.7979899497487438e-05,
      "loss": 1.6585,
      "step": 7200
    },
    {
      "epoch": 0.18,
      "grad_norm": 5.578125,
      "learning_rate": 2.7949748743718593e-05,
      "loss": 1.6567,
      "step": 7300
    },
    {
      "epoch": 0.18,
      "grad_norm": 6.83203125,
      "learning_rate": 2.791959798994975e-05,
      "loss": 1.633,
      "step": 7400
    },
    {
      "epoch": 0.19,
      "grad_norm": 6.90234375,
      "learning_rate": 2.7889447236180907e-05,
      "loss": 1.6951,
      "step": 7500
    },
    {
      "epoch": 0.19,
      "eval_loss": 1.6509525775909424,
      "eval_runtime": 58.6377,
      "eval_samples_per_second": 17.054,
      "eval_steps_per_second": 4.263,
      "step": 7500
    },
    {
      "epoch": 0.19,
      "grad_norm": 11.390625,
      "learning_rate": 2.7859296482412062e-05,
      "loss": 1.6665,
      "step": 7600
    },
    {
      "epoch": 0.19,
      "grad_norm": 6.58203125,
      "learning_rate": 2.7829145728643217e-05,
      "loss": 1.6476,
      "step": 7700
    },
    {
      "epoch": 0.2,
      "grad_norm": 5.8671875,
      "learning_rate": 2.7798994974874373e-05,
      "loss": 1.6488,
      "step": 7800
    },
    {
      "epoch": 0.2,
      "grad_norm": 4.69140625,
      "learning_rate": 2.7768844221105528e-05,
      "loss": 1.618,
      "step": 7900
    },
    {
      "epoch": 0.2,
      "grad_norm": 13.1796875,
      "learning_rate": 2.7738693467336683e-05,
      "loss": 1.6614,
      "step": 8000
    },
    {
      "epoch": 0.2,
      "grad_norm": 9.1171875,
      "learning_rate": 2.770854271356784e-05,
      "loss": 1.6136,
      "step": 8100
    },
    {
      "epoch": 0.2,
      "grad_norm": 6.9140625,
      "learning_rate": 2.7678391959798994e-05,
      "loss": 1.6854,
      "step": 8200
    },
    {
      "epoch": 0.21,
      "grad_norm": 7.15625,
      "learning_rate": 2.7648241206030152e-05,
      "loss": 1.6302,
      "step": 8300
    },
    {
      "epoch": 0.21,
      "grad_norm": 5.05078125,
      "learning_rate": 2.7618090452261308e-05,
      "loss": 1.5978,
      "step": 8400
    },
    {
      "epoch": 0.21,
      "grad_norm": 9.4765625,
      "learning_rate": 2.7587939698492463e-05,
      "loss": 1.6275,
      "step": 8500
    },
    {
      "epoch": 0.21,
      "grad_norm": 4.4765625,
      "learning_rate": 2.7557788944723618e-05,
      "loss": 1.6216,
      "step": 8600
    },
    {
      "epoch": 0.22,
      "grad_norm": 6.62109375,
      "learning_rate": 2.7527638190954777e-05,
      "loss": 1.6664,
      "step": 8700
    },
    {
      "epoch": 0.22,
      "grad_norm": 5.609375,
      "learning_rate": 2.7497487437185932e-05,
      "loss": 1.6589,
      "step": 8800
    },
    {
      "epoch": 0.22,
      "grad_norm": 7.2109375,
      "learning_rate": 2.7467336683417087e-05,
      "loss": 1.654,
      "step": 8900
    },
    {
      "epoch": 0.23,
      "grad_norm": 5.9921875,
      "learning_rate": 2.743718592964824e-05,
      "loss": 1.639,
      "step": 9000
    },
    {
      "epoch": 0.23,
      "grad_norm": 9.21875,
      "learning_rate": 2.7407035175879398e-05,
      "loss": 1.6147,
      "step": 9100
    },
    {
      "epoch": 0.23,
      "grad_norm": 6.84765625,
      "learning_rate": 2.7376884422110553e-05,
      "loss": 1.6392,
      "step": 9200
    },
    {
      "epoch": 0.23,
      "grad_norm": 4.75390625,
      "learning_rate": 2.734673366834171e-05,
      "loss": 1.6202,
      "step": 9300
    },
    {
      "epoch": 0.23,
      "grad_norm": 6.0,
      "learning_rate": 2.7316582914572864e-05,
      "loss": 1.6408,
      "step": 9400
    },
    {
      "epoch": 0.24,
      "grad_norm": 4.44921875,
      "learning_rate": 2.7286432160804022e-05,
      "loss": 1.6437,
      "step": 9500
    },
    {
      "epoch": 0.24,
      "grad_norm": 3.9921875,
      "learning_rate": 2.7256281407035178e-05,
      "loss": 1.6165,
      "step": 9600
    },
    {
      "epoch": 0.24,
      "grad_norm": 7.15234375,
      "learning_rate": 2.7226130653266333e-05,
      "loss": 1.6217,
      "step": 9700
    },
    {
      "epoch": 0.24,
      "grad_norm": 5.38671875,
      "learning_rate": 2.7195979899497488e-05,
      "loss": 1.6584,
      "step": 9800
    },
    {
      "epoch": 0.25,
      "grad_norm": 6.62109375,
      "learning_rate": 2.7165829145728643e-05,
      "loss": 1.6204,
      "step": 9900
    },
    {
      "epoch": 0.25,
      "grad_norm": 8.6796875,
      "learning_rate": 2.7135678391959802e-05,
      "loss": 1.6176,
      "step": 10000
    },
    {
      "epoch": 0.25,
      "eval_loss": 1.5965248346328735,
      "eval_runtime": 58.635,
      "eval_samples_per_second": 17.055,
      "eval_steps_per_second": 4.264,
      "step": 10000
    },
    {
      "epoch": 0.25,
      "grad_norm": 4.96484375,
      "learning_rate": 2.7105527638190957e-05,
      "loss": 1.6494,
      "step": 10100
    },
    {
      "epoch": 0.26,
      "grad_norm": 5.84375,
      "learning_rate": 2.707537688442211e-05,
      "loss": 1.5987,
      "step": 10200
    },
    {
      "epoch": 0.26,
      "grad_norm": 8.953125,
      "learning_rate": 2.7045226130653264e-05,
      "loss": 1.6262,
      "step": 10300
    },
    {
      "epoch": 0.26,
      "grad_norm": 6.92578125,
      "learning_rate": 2.7015075376884423e-05,
      "loss": 1.6362,
      "step": 10400
    },
    {
      "epoch": 0.26,
      "grad_norm": 5.35546875,
      "learning_rate": 2.698492462311558e-05,
      "loss": 1.6,
      "step": 10500
    },
    {
      "epoch": 0.27,
      "grad_norm": 7.41015625,
      "learning_rate": 2.6954773869346734e-05,
      "loss": 1.693,
      "step": 10600
    },
    {
      "epoch": 0.27,
      "grad_norm": 5.64453125,
      "learning_rate": 2.692462311557789e-05,
      "loss": 1.6496,
      "step": 10700
    },
    {
      "epoch": 0.27,
      "grad_norm": 4.953125,
      "learning_rate": 2.6894472361809048e-05,
      "loss": 1.6043,
      "step": 10800
    },
    {
      "epoch": 0.27,
      "grad_norm": 5.82421875,
      "learning_rate": 2.6864321608040203e-05,
      "loss": 1.6291,
      "step": 10900
    },
    {
      "epoch": 0.28,
      "grad_norm": 6.1484375,
      "learning_rate": 2.6834170854271358e-05,
      "loss": 1.5974,
      "step": 11000
    },
    {
      "epoch": 0.28,
      "grad_norm": 5.90234375,
      "learning_rate": 2.6804020100502513e-05,
      "loss": 1.6811,
      "step": 11100
    },
    {
      "epoch": 0.28,
      "grad_norm": 8.2890625,
      "learning_rate": 2.6773869346733672e-05,
      "loss": 1.6045,
      "step": 11200
    },
    {
      "epoch": 0.28,
      "grad_norm": 5.09765625,
      "learning_rate": 2.6743718592964824e-05,
      "loss": 1.6116,
      "step": 11300
    },
    {
      "epoch": 0.28,
      "grad_norm": 4.29296875,
      "learning_rate": 2.671356783919598e-05,
      "loss": 1.5674,
      "step": 11400
    },
    {
      "epoch": 0.29,
      "grad_norm": 5.8359375,
      "learning_rate": 2.6683417085427134e-05,
      "loss": 1.6231,
      "step": 11500
    },
    {
      "epoch": 0.29,
      "grad_norm": 5.5234375,
      "learning_rate": 2.6653266331658293e-05,
      "loss": 1.618,
      "step": 11600
    },
    {
      "epoch": 0.29,
      "grad_norm": 6.93359375,
      "learning_rate": 2.6623115577889448e-05,
      "loss": 1.5943,
      "step": 11700
    },
    {
      "epoch": 0.29,
      "grad_norm": 6.33203125,
      "learning_rate": 2.6592964824120604e-05,
      "loss": 1.6126,
      "step": 11800
    },
    {
      "epoch": 0.3,
      "grad_norm": 9.625,
      "learning_rate": 2.656281407035176e-05,
      "loss": 1.5968,
      "step": 11900
    },
    {
      "epoch": 0.3,
      "grad_norm": 6.55859375,
      "learning_rate": 2.6532663316582917e-05,
      "loss": 1.6228,
      "step": 12000
    },
    {
      "epoch": 0.3,
      "grad_norm": 5.8359375,
      "learning_rate": 2.6502512562814073e-05,
      "loss": 1.6006,
      "step": 12100
    },
    {
      "epoch": 0.3,
      "grad_norm": 10.2578125,
      "learning_rate": 2.6472361809045228e-05,
      "loss": 1.5515,
      "step": 12200
    },
    {
      "epoch": 0.31,
      "grad_norm": 7.23046875,
      "learning_rate": 2.6442211055276383e-05,
      "loss": 1.6433,
      "step": 12300
    },
    {
      "epoch": 0.31,
      "grad_norm": 7.5625,
      "learning_rate": 2.6412060301507535e-05,
      "loss": 1.6273,
      "step": 12400
    },
    {
      "epoch": 0.31,
      "grad_norm": 4.7421875,
      "learning_rate": 2.6381909547738694e-05,
      "loss": 1.5738,
      "step": 12500
    },
    {
      "epoch": 0.31,
      "eval_loss": 1.6012969017028809,
      "eval_runtime": 58.7203,
      "eval_samples_per_second": 17.03,
      "eval_steps_per_second": 4.257,
      "step": 12500
    },
    {
      "epoch": 0.32,
      "grad_norm": 8.2421875,
      "learning_rate": 2.635175879396985e-05,
      "loss": 1.5637,
      "step": 12600
    },
    {
      "epoch": 0.32,
      "grad_norm": 7.13671875,
      "learning_rate": 2.6321608040201004e-05,
      "loss": 1.6395,
      "step": 12700
    },
    {
      "epoch": 0.32,
      "grad_norm": 9.0390625,
      "learning_rate": 2.629145728643216e-05,
      "loss": 1.6293,
      "step": 12800
    },
    {
      "epoch": 0.32,
      "grad_norm": 6.546875,
      "learning_rate": 2.6261306532663318e-05,
      "loss": 1.5729,
      "step": 12900
    },
    {
      "epoch": 0.33,
      "grad_norm": 5.2734375,
      "learning_rate": 2.6231155778894474e-05,
      "loss": 1.5743,
      "step": 13000
    },
    {
      "epoch": 0.33,
      "grad_norm": 5.83984375,
      "learning_rate": 2.620100502512563e-05,
      "loss": 1.5871,
      "step": 13100
    },
    {
      "epoch": 0.33,
      "grad_norm": 4.55078125,
      "learning_rate": 2.6170854271356784e-05,
      "loss": 1.6303,
      "step": 13200
    },
    {
      "epoch": 0.33,
      "grad_norm": 7.61328125,
      "learning_rate": 2.6140703517587943e-05,
      "loss": 1.5893,
      "step": 13300
    },
    {
      "epoch": 0.34,
      "grad_norm": 9.7890625,
      "learning_rate": 2.6110552763819098e-05,
      "loss": 1.6201,
      "step": 13400
    },
    {
      "epoch": 0.34,
      "grad_norm": 7.26953125,
      "learning_rate": 2.6080402010050253e-05,
      "loss": 1.5352,
      "step": 13500
    },
    {
      "epoch": 0.34,
      "grad_norm": 5.3515625,
      "learning_rate": 2.6050251256281405e-05,
      "loss": 1.5675,
      "step": 13600
    },
    {
      "epoch": 0.34,
      "grad_norm": 6.16796875,
      "learning_rate": 2.6020100502512564e-05,
      "loss": 1.5841,
      "step": 13700
    },
    {
      "epoch": 0.34,
      "grad_norm": 4.64453125,
      "learning_rate": 2.598994974874372e-05,
      "loss": 1.5995,
      "step": 13800
    },
    {
      "epoch": 0.35,
      "grad_norm": 7.046875,
      "learning_rate": 2.5959798994974874e-05,
      "loss": 1.5882,
      "step": 13900
    },
    {
      "epoch": 0.35,
      "grad_norm": 9.5703125,
      "learning_rate": 2.592964824120603e-05,
      "loss": 1.5834,
      "step": 14000
    },
    {
      "epoch": 0.35,
      "grad_norm": 6.1640625,
      "learning_rate": 2.5899497487437188e-05,
      "loss": 1.5723,
      "step": 14100
    },
    {
      "epoch": 0.35,
      "grad_norm": 6.83984375,
      "learning_rate": 2.5869346733668344e-05,
      "loss": 1.5971,
      "step": 14200
    },
    {
      "epoch": 0.36,
      "grad_norm": 6.5625,
      "learning_rate": 2.58391959798995e-05,
      "loss": 1.5571,
      "step": 14300
    },
    {
      "epoch": 0.36,
      "grad_norm": 4.80859375,
      "learning_rate": 2.5809045226130654e-05,
      "loss": 1.6284,
      "step": 14400
    },
    {
      "epoch": 0.36,
      "grad_norm": 7.25390625,
      "learning_rate": 2.577889447236181e-05,
      "loss": 1.6099,
      "step": 14500
    },
    {
      "epoch": 0.36,
      "grad_norm": 7.44921875,
      "learning_rate": 2.5748743718592968e-05,
      "loss": 1.6083,
      "step": 14600
    },
    {
      "epoch": 0.37,
      "grad_norm": 5.109375,
      "learning_rate": 2.571859296482412e-05,
      "loss": 1.5841,
      "step": 14700
    },
    {
      "epoch": 0.37,
      "grad_norm": 6.93359375,
      "learning_rate": 2.5688442211055275e-05,
      "loss": 1.5752,
      "step": 14800
    },
    {
      "epoch": 0.37,
      "grad_norm": 7.38671875,
      "learning_rate": 2.565829145728643e-05,
      "loss": 1.5535,
      "step": 14900
    },
    {
      "epoch": 0.38,
      "grad_norm": 5.40234375,
      "learning_rate": 2.562814070351759e-05,
      "loss": 1.5804,
      "step": 15000
    },
    {
      "epoch": 0.38,
      "eval_loss": 1.5745244026184082,
      "eval_runtime": 58.6931,
      "eval_samples_per_second": 17.038,
      "eval_steps_per_second": 4.259,
      "step": 15000
    },
    {
      "epoch": 0.38,
      "grad_norm": 5.33203125,
      "learning_rate": 2.5597989949748744e-05,
      "loss": 1.527,
      "step": 15100
    },
    {
      "epoch": 0.38,
      "grad_norm": 6.8828125,
      "learning_rate": 2.55678391959799e-05,
      "loss": 1.5769,
      "step": 15200
    },
    {
      "epoch": 0.38,
      "grad_norm": 7.4140625,
      "learning_rate": 2.5537688442211055e-05,
      "loss": 1.5983,
      "step": 15300
    },
    {
      "epoch": 0.39,
      "grad_norm": 5.0390625,
      "learning_rate": 2.5507537688442214e-05,
      "loss": 1.5873,
      "step": 15400
    },
    {
      "epoch": 0.39,
      "grad_norm": 6.39453125,
      "learning_rate": 2.547738693467337e-05,
      "loss": 1.5888,
      "step": 15500
    },
    {
      "epoch": 0.39,
      "grad_norm": 7.703125,
      "learning_rate": 2.5447236180904524e-05,
      "loss": 1.559,
      "step": 15600
    },
    {
      "epoch": 0.39,
      "grad_norm": 10.5859375,
      "learning_rate": 2.541708542713568e-05,
      "loss": 1.5862,
      "step": 15700
    },
    {
      "epoch": 0.4,
      "grad_norm": 6.8984375,
      "learning_rate": 2.5386934673366835e-05,
      "loss": 1.5644,
      "step": 15800
    },
    {
      "epoch": 0.4,
      "grad_norm": 5.9453125,
      "learning_rate": 2.535678391959799e-05,
      "loss": 1.5663,
      "step": 15900
    },
    {
      "epoch": 0.4,
      "grad_norm": 6.1171875,
      "learning_rate": 2.5326633165829145e-05,
      "loss": 1.6077,
      "step": 16000
    },
    {
      "epoch": 0.4,
      "grad_norm": 5.06640625,
      "learning_rate": 2.52964824120603e-05,
      "loss": 1.52,
      "step": 16100
    },
    {
      "epoch": 0.41,
      "grad_norm": 5.140625,
      "learning_rate": 2.526633165829146e-05,
      "loss": 1.5387,
      "step": 16200
    },
    {
      "epoch": 0.41,
      "grad_norm": 5.19140625,
      "learning_rate": 2.5236180904522614e-05,
      "loss": 1.5523,
      "step": 16300
    },
    {
      "epoch": 0.41,
      "grad_norm": 6.859375,
      "learning_rate": 2.520603015075377e-05,
      "loss": 1.5684,
      "step": 16400
    },
    {
      "epoch": 0.41,
      "grad_norm": 8.9765625,
      "learning_rate": 2.5175879396984925e-05,
      "loss": 1.591,
      "step": 16500
    },
    {
      "epoch": 0.41,
      "grad_norm": 6.6640625,
      "learning_rate": 2.5145728643216084e-05,
      "loss": 1.6167,
      "step": 16600
    },
    {
      "epoch": 0.42,
      "grad_norm": 7.8203125,
      "learning_rate": 2.511557788944724e-05,
      "loss": 1.5475,
      "step": 16700
    },
    {
      "epoch": 0.42,
      "grad_norm": 6.125,
      "learning_rate": 2.5085427135678394e-05,
      "loss": 1.5604,
      "step": 16800
    },
    {
      "epoch": 0.42,
      "grad_norm": 6.7421875,
      "learning_rate": 2.5055276381909546e-05,
      "loss": 1.6347,
      "step": 16900
    },
    {
      "epoch": 0.42,
      "grad_norm": 8.8828125,
      "learning_rate": 2.50251256281407e-05,
      "loss": 1.5626,
      "step": 17000
    },
    {
      "epoch": 0.43,
      "grad_norm": 8.0078125,
      "learning_rate": 2.499497487437186e-05,
      "loss": 1.5119,
      "step": 17100
    },
    {
      "epoch": 0.43,
      "grad_norm": 6.69921875,
      "learning_rate": 2.4964824120603015e-05,
      "loss": 1.5075,
      "step": 17200
    },
    {
      "epoch": 0.43,
      "grad_norm": 7.83203125,
      "learning_rate": 2.493467336683417e-05,
      "loss": 1.5604,
      "step": 17300
    },
    {
      "epoch": 0.43,
      "grad_norm": 6.86328125,
      "learning_rate": 2.4904522613065326e-05,
      "loss": 1.5727,
      "step": 17400
    },
    {
      "epoch": 0.44,
      "grad_norm": 7.91015625,
      "learning_rate": 2.4874371859296484e-05,
      "loss": 1.538,
      "step": 17500
    },
    {
      "epoch": 0.44,
      "eval_loss": 1.5860785245895386,
      "eval_runtime": 58.5282,
      "eval_samples_per_second": 17.086,
      "eval_steps_per_second": 4.271,
      "step": 17500
    },
    {
      "epoch": 0.44,
      "grad_norm": 8.484375,
      "learning_rate": 2.484422110552764e-05,
      "loss": 1.5551,
      "step": 17600
    },
    {
      "epoch": 0.44,
      "grad_norm": 7.64453125,
      "learning_rate": 2.4814070351758795e-05,
      "loss": 1.5169,
      "step": 17700
    },
    {
      "epoch": 0.45,
      "grad_norm": 8.4375,
      "learning_rate": 2.478391959798995e-05,
      "loss": 1.5385,
      "step": 17800
    },
    {
      "epoch": 0.45,
      "grad_norm": 6.08203125,
      "learning_rate": 2.475376884422111e-05,
      "loss": 1.5492,
      "step": 17900
    },
    {
      "epoch": 0.45,
      "grad_norm": 11.84375,
      "learning_rate": 2.4723618090452264e-05,
      "loss": 1.5695,
      "step": 18000
    },
    {
      "epoch": 0.45,
      "grad_norm": 5.98046875,
      "learning_rate": 2.4693467336683416e-05,
      "loss": 1.5804,
      "step": 18100
    },
    {
      "epoch": 0.46,
      "grad_norm": 7.75,
      "learning_rate": 2.466331658291457e-05,
      "loss": 1.5567,
      "step": 18200
    },
    {
      "epoch": 0.46,
      "grad_norm": 5.51953125,
      "learning_rate": 2.463316582914573e-05,
      "loss": 1.5492,
      "step": 18300
    },
    {
      "epoch": 0.46,
      "grad_norm": 6.86328125,
      "learning_rate": 2.4603015075376885e-05,
      "loss": 1.5733,
      "step": 18400
    },
    {
      "epoch": 0.46,
      "grad_norm": 6.56640625,
      "learning_rate": 2.457286432160804e-05,
      "loss": 1.586,
      "step": 18500
    },
    {
      "epoch": 0.47,
      "grad_norm": 4.23828125,
      "learning_rate": 2.4542713567839196e-05,
      "loss": 1.5428,
      "step": 18600
    },
    {
      "epoch": 0.47,
      "grad_norm": 7.76171875,
      "learning_rate": 2.4512562814070354e-05,
      "loss": 1.5085,
      "step": 18700
    },
    {
      "epoch": 0.47,
      "grad_norm": 6.984375,
      "learning_rate": 2.448241206030151e-05,
      "loss": 1.5726,
      "step": 18800
    },
    {
      "epoch": 0.47,
      "grad_norm": 7.92578125,
      "learning_rate": 2.4452261306532665e-05,
      "loss": 1.5666,
      "step": 18900
    },
    {
      "epoch": 0.47,
      "grad_norm": 7.66796875,
      "learning_rate": 2.442211055276382e-05,
      "loss": 1.4978,
      "step": 19000
    },
    {
      "epoch": 0.48,
      "grad_norm": 6.89453125,
      "learning_rate": 2.4391959798994975e-05,
      "loss": 1.5385,
      "step": 19100
    },
    {
      "epoch": 0.48,
      "grad_norm": 10.1484375,
      "learning_rate": 2.436180904522613e-05,
      "loss": 1.5172,
      "step": 19200
    },
    {
      "epoch": 0.48,
      "grad_norm": 6.75,
      "learning_rate": 2.4331658291457286e-05,
      "loss": 1.5754,
      "step": 19300
    },
    {
      "epoch": 0.48,
      "grad_norm": 4.90625,
      "learning_rate": 2.430150753768844e-05,
      "loss": 1.4749,
      "step": 19400
    },
    {
      "epoch": 0.49,
      "grad_norm": 8.3515625,
      "learning_rate": 2.4271356783919596e-05,
      "loss": 1.5676,
      "step": 19500
    },
    {
      "epoch": 0.49,
      "grad_norm": 9.671875,
      "learning_rate": 2.4241206030150755e-05,
      "loss": 1.5169,
      "step": 19600
    },
    {
      "epoch": 0.49,
      "grad_norm": 7.02734375,
      "learning_rate": 2.421105527638191e-05,
      "loss": 1.5026,
      "step": 19700
    },
    {
      "epoch": 0.49,
      "grad_norm": 8.0390625,
      "learning_rate": 2.4180904522613066e-05,
      "loss": 1.5726,
      "step": 19800
    },
    {
      "epoch": 0.5,
      "grad_norm": 6.265625,
      "learning_rate": 2.415075376884422e-05,
      "loss": 1.4948,
      "step": 19900
    },
    {
      "epoch": 0.5,
      "grad_norm": 6.828125,
      "learning_rate": 2.412060301507538e-05,
      "loss": 1.553,
      "step": 20000
    },
    {
      "epoch": 0.5,
      "eval_loss": 1.5626012086868286,
      "eval_runtime": 58.5266,
      "eval_samples_per_second": 17.086,
      "eval_steps_per_second": 4.272,
      "step": 20000
    },
    {
      "epoch": 0.5,
      "grad_norm": 6.4453125,
      "learning_rate": 2.4090452261306535e-05,
      "loss": 1.5808,
      "step": 20100
    },
    {
      "epoch": 0.51,
      "grad_norm": 5.28515625,
      "learning_rate": 2.406030150753769e-05,
      "loss": 1.5917,
      "step": 20200
    },
    {
      "epoch": 0.51,
      "grad_norm": 12.0390625,
      "learning_rate": 2.4030150753768842e-05,
      "loss": 1.5786,
      "step": 20300
    },
    {
      "epoch": 0.51,
      "grad_norm": 8.3125,
      "learning_rate": 2.4e-05,
      "loss": 1.5467,
      "step": 20400
    },
    {
      "epoch": 0.51,
      "grad_norm": 5.46484375,
      "learning_rate": 2.3969849246231156e-05,
      "loss": 1.5572,
      "step": 20500
    },
    {
      "epoch": 0.52,
      "grad_norm": 7.2109375,
      "learning_rate": 2.393969849246231e-05,
      "loss": 1.4939,
      "step": 20600
    },
    {
      "epoch": 0.52,
      "grad_norm": 8.703125,
      "learning_rate": 2.3909547738693466e-05,
      "loss": 1.5634,
      "step": 20700
    },
    {
      "epoch": 0.52,
      "grad_norm": 5.453125,
      "learning_rate": 2.3879396984924625e-05,
      "loss": 1.5502,
      "step": 20800
    },
    {
      "epoch": 0.52,
      "grad_norm": 3.52734375,
      "learning_rate": 2.384924623115578e-05,
      "loss": 1.5421,
      "step": 20900
    },
    {
      "epoch": 0.53,
      "grad_norm": 5.87890625,
      "learning_rate": 2.3819095477386936e-05,
      "loss": 1.569,
      "step": 21000
    },
    {
      "epoch": 0.53,
      "grad_norm": 6.60546875,
      "learning_rate": 2.378894472361809e-05,
      "loss": 1.5416,
      "step": 21100
    },
    {
      "epoch": 0.53,
      "grad_norm": 7.68359375,
      "learning_rate": 2.375879396984925e-05,
      "loss": 1.5318,
      "step": 21200
    },
    {
      "epoch": 0.53,
      "grad_norm": 7.07421875,
      "learning_rate": 2.3728643216080405e-05,
      "loss": 1.5376,
      "step": 21300
    },
    {
      "epoch": 0.54,
      "grad_norm": 6.71875,
      "learning_rate": 2.369849246231156e-05,
      "loss": 1.5728,
      "step": 21400
    },
    {
      "epoch": 0.54,
      "grad_norm": 5.34375,
      "learning_rate": 2.3668341708542712e-05,
      "loss": 1.544,
      "step": 21500
    },
    {
      "epoch": 0.54,
      "grad_norm": 5.00390625,
      "learning_rate": 2.3638190954773867e-05,
      "loss": 1.4737,
      "step": 21600
    },
    {
      "epoch": 0.54,
      "grad_norm": 11.5,
      "learning_rate": 2.3608040201005026e-05,
      "loss": 1.5458,
      "step": 21700
    },
    {
      "epoch": 0.55,
      "grad_norm": 8.75,
      "learning_rate": 2.357788944723618e-05,
      "loss": 1.5746,
      "step": 21800
    },
    {
      "epoch": 0.55,
      "grad_norm": 7.98828125,
      "learning_rate": 2.3547738693467336e-05,
      "loss": 1.4738,
      "step": 21900
    },
    {
      "epoch": 0.55,
      "grad_norm": 11.9609375,
      "learning_rate": 2.351758793969849e-05,
      "loss": 1.5951,
      "step": 22000
    },
    {
      "epoch": 0.55,
      "grad_norm": 7.09375,
      "learning_rate": 2.348743718592965e-05,
      "loss": 1.4439,
      "step": 22100
    },
    {
      "epoch": 0.56,
      "grad_norm": 9.9296875,
      "learning_rate": 2.3457286432160806e-05,
      "loss": 1.5166,
      "step": 22200
    },
    {
      "epoch": 0.56,
      "grad_norm": 4.30078125,
      "learning_rate": 2.342713567839196e-05,
      "loss": 1.496,
      "step": 22300
    },
    {
      "epoch": 0.56,
      "grad_norm": 8.171875,
      "learning_rate": 2.3396984924623116e-05,
      "loss": 1.5075,
      "step": 22400
    },
    {
      "epoch": 0.56,
      "grad_norm": 7.54296875,
      "learning_rate": 2.3366834170854275e-05,
      "loss": 1.4837,
      "step": 22500
    },
    {
      "epoch": 0.56,
      "eval_loss": 1.512613296508789,
      "eval_runtime": 58.4743,
      "eval_samples_per_second": 17.102,
      "eval_steps_per_second": 4.275,
      "step": 22500
    },
    {
      "epoch": 0.56,
      "grad_norm": 10.2265625,
      "learning_rate": 2.3336683417085427e-05,
      "loss": 1.5465,
      "step": 22600
    },
    {
      "epoch": 0.57,
      "grad_norm": 10.234375,
      "learning_rate": 2.3306532663316582e-05,
      "loss": 1.533,
      "step": 22700
    },
    {
      "epoch": 0.57,
      "grad_norm": 6.2734375,
      "learning_rate": 2.3276381909547737e-05,
      "loss": 1.5222,
      "step": 22800
    },
    {
      "epoch": 0.57,
      "grad_norm": 8.125,
      "learning_rate": 2.3246231155778896e-05,
      "loss": 1.5249,
      "step": 22900
    },
    {
      "epoch": 0.57,
      "grad_norm": 9.375,
      "learning_rate": 2.321608040201005e-05,
      "loss": 1.4978,
      "step": 23000
    },
    {
      "epoch": 0.58,
      "grad_norm": 7.8515625,
      "learning_rate": 2.3185929648241206e-05,
      "loss": 1.5416,
      "step": 23100
    },
    {
      "epoch": 0.58,
      "grad_norm": 8.796875,
      "learning_rate": 2.315577889447236e-05,
      "loss": 1.5141,
      "step": 23200
    },
    {
      "epoch": 0.58,
      "grad_norm": 6.9140625,
      "learning_rate": 2.312562814070352e-05,
      "loss": 1.4859,
      "step": 23300
    },
    {
      "epoch": 0.58,
      "grad_norm": 5.29296875,
      "learning_rate": 2.3095477386934676e-05,
      "loss": 1.4955,
      "step": 23400
    },
    {
      "epoch": 0.59,
      "grad_norm": 5.82421875,
      "learning_rate": 2.306532663316583e-05,
      "loss": 1.4968,
      "step": 23500
    },
    {
      "epoch": 0.59,
      "grad_norm": 7.01953125,
      "learning_rate": 2.3035175879396986e-05,
      "loss": 1.4652,
      "step": 23600
    },
    {
      "epoch": 0.59,
      "grad_norm": 6.625,
      "learning_rate": 2.3005025125628138e-05,
      "loss": 1.5003,
      "step": 23700
    },
    {
      "epoch": 0.59,
      "grad_norm": 6.21875,
      "learning_rate": 2.2974874371859297e-05,
      "loss": 1.5142,
      "step": 23800
    },
    {
      "epoch": 0.6,
      "grad_norm": 4.73828125,
      "learning_rate": 2.2944723618090452e-05,
      "loss": 1.5327,
      "step": 23900
    },
    {
      "epoch": 0.6,
      "grad_norm": 6.79296875,
      "learning_rate": 2.2914572864321607e-05,
      "loss": 1.4745,
      "step": 24000
    },
    {
      "epoch": 0.6,
      "grad_norm": 9.515625,
      "learning_rate": 2.2884422110552762e-05,
      "loss": 1.4769,
      "step": 24100
    },
    {
      "epoch": 0.6,
      "grad_norm": 5.5859375,
      "learning_rate": 2.285427135678392e-05,
      "loss": 1.5434,
      "step": 24200
    },
    {
      "epoch": 0.61,
      "grad_norm": 8.1796875,
      "learning_rate": 2.2824120603015076e-05,
      "loss": 1.5288,
      "step": 24300
    },
    {
      "epoch": 0.61,
      "grad_norm": 4.390625,
      "learning_rate": 2.279396984924623e-05,
      "loss": 1.4798,
      "step": 24400
    },
    {
      "epoch": 0.61,
      "grad_norm": 4.91796875,
      "learning_rate": 2.2763819095477387e-05,
      "loss": 1.499,
      "step": 24500
    },
    {
      "epoch": 0.61,
      "grad_norm": 8.9921875,
      "learning_rate": 2.2733668341708546e-05,
      "loss": 1.4819,
      "step": 24600
    },
    {
      "epoch": 0.62,
      "grad_norm": 5.5078125,
      "learning_rate": 2.27035175879397e-05,
      "loss": 1.5138,
      "step": 24700
    },
    {
      "epoch": 0.62,
      "grad_norm": 8.453125,
      "learning_rate": 2.2673366834170856e-05,
      "loss": 1.4908,
      "step": 24800
    },
    {
      "epoch": 0.62,
      "grad_norm": 7.09375,
      "learning_rate": 2.2643216080402008e-05,
      "loss": 1.5043,
      "step": 24900
    },
    {
      "epoch": 0.62,
      "grad_norm": 7.32421875,
      "learning_rate": 2.2613065326633167e-05,
      "loss": 1.5215,
      "step": 25000
    },
    {
      "epoch": 0.62,
      "eval_loss": 1.5021257400512695,
      "eval_runtime": 58.4197,
      "eval_samples_per_second": 17.118,
      "eval_steps_per_second": 4.279,
      "step": 25000
    },
    {
      "epoch": 0.63,
      "grad_norm": 4.16796875,
      "learning_rate": 2.2582914572864322e-05,
      "loss": 1.5173,
      "step": 25100
    },
    {
      "epoch": 0.63,
      "grad_norm": 5.7265625,
      "learning_rate": 2.2552763819095477e-05,
      "loss": 1.4825,
      "step": 25200
    },
    {
      "epoch": 0.63,
      "grad_norm": 6.3984375,
      "learning_rate": 2.2522613065326632e-05,
      "loss": 1.496,
      "step": 25300
    },
    {
      "epoch": 0.64,
      "grad_norm": 5.75,
      "learning_rate": 2.249246231155779e-05,
      "loss": 1.4803,
      "step": 25400
    },
    {
      "epoch": 0.64,
      "grad_norm": 6.71484375,
      "learning_rate": 2.2462311557788946e-05,
      "loss": 1.4516,
      "step": 25500
    },
    {
      "epoch": 0.64,
      "grad_norm": 7.2265625,
      "learning_rate": 2.24321608040201e-05,
      "loss": 1.4969,
      "step": 25600
    },
    {
      "epoch": 0.64,
      "grad_norm": 5.28125,
      "learning_rate": 2.2402010050251257e-05,
      "loss": 1.5185,
      "step": 25700
    },
    {
      "epoch": 0.65,
      "grad_norm": 6.40234375,
      "learning_rate": 2.2371859296482416e-05,
      "loss": 1.4593,
      "step": 25800
    },
    {
      "epoch": 0.65,
      "grad_norm": 6.3671875,
      "learning_rate": 2.234170854271357e-05,
      "loss": 1.5272,
      "step": 25900
    },
    {
      "epoch": 0.65,
      "grad_norm": 4.5625,
      "learning_rate": 2.2311557788944723e-05,
      "loss": 1.496,
      "step": 26000
    },
    {
      "epoch": 0.65,
      "grad_norm": 5.69140625,
      "learning_rate": 2.2281407035175878e-05,
      "loss": 1.4677,
      "step": 26100
    },
    {
      "epoch": 0.66,
      "grad_norm": 5.9609375,
      "learning_rate": 2.2251256281407033e-05,
      "loss": 1.5153,
      "step": 26200
    },
    {
      "epoch": 0.66,
      "grad_norm": 4.890625,
      "learning_rate": 2.2221105527638192e-05,
      "loss": 1.5064,
      "step": 26300
    },
    {
      "epoch": 0.66,
      "grad_norm": 7.5859375,
      "learning_rate": 2.2190954773869347e-05,
      "loss": 1.4875,
      "step": 26400
    },
    {
      "epoch": 0.66,
      "grad_norm": 9.3984375,
      "learning_rate": 2.2160804020100502e-05,
      "loss": 1.5198,
      "step": 26500
    },
    {
      "epoch": 0.67,
      "grad_norm": 5.72265625,
      "learning_rate": 2.2130653266331658e-05,
      "loss": 1.5047,
      "step": 26600
    },
    {
      "epoch": 0.67,
      "grad_norm": 4.1484375,
      "learning_rate": 2.2100502512562816e-05,
      "loss": 1.4954,
      "step": 26700
    },
    {
      "epoch": 0.67,
      "grad_norm": 4.5859375,
      "learning_rate": 2.207035175879397e-05,
      "loss": 1.5074,
      "step": 26800
    },
    {
      "epoch": 0.67,
      "grad_norm": 7.98828125,
      "learning_rate": 2.2040201005025127e-05,
      "loss": 1.5158,
      "step": 26900
    },
    {
      "epoch": 0.68,
      "grad_norm": 4.3125,
      "learning_rate": 2.2010050251256282e-05,
      "loss": 1.4735,
      "step": 27000
    },
    {
      "epoch": 0.68,
      "grad_norm": 11.2421875,
      "learning_rate": 2.1979899497487437e-05,
      "loss": 1.4868,
      "step": 27100
    },
    {
      "epoch": 0.68,
      "grad_norm": 5.69921875,
      "learning_rate": 2.1949748743718593e-05,
      "loss": 1.501,
      "step": 27200
    },
    {
      "epoch": 0.68,
      "grad_norm": 8.078125,
      "learning_rate": 2.1919597989949748e-05,
      "loss": 1.4897,
      "step": 27300
    },
    {
      "epoch": 0.69,
      "grad_norm": 8.8828125,
      "learning_rate": 2.1889447236180903e-05,
      "loss": 1.5144,
      "step": 27400
    },
    {
      "epoch": 0.69,
      "grad_norm": 9.109375,
      "learning_rate": 2.1859296482412062e-05,
      "loss": 1.4581,
      "step": 27500
    },
    {
      "epoch": 0.69,
      "eval_loss": 1.4510895013809204,
      "eval_runtime": 58.4398,
      "eval_samples_per_second": 17.112,
      "eval_steps_per_second": 4.278,
      "step": 27500
    },
    {
      "epoch": 0.69,
      "grad_norm": 4.703125,
      "learning_rate": 2.1829145728643217e-05,
      "loss": 1.5377,
      "step": 27600
    },
    {
      "epoch": 0.69,
      "grad_norm": 8.09375,
      "learning_rate": 2.1798994974874372e-05,
      "loss": 1.4735,
      "step": 27700
    },
    {
      "epoch": 0.69,
      "grad_norm": 8.328125,
      "learning_rate": 2.1768844221105528e-05,
      "loss": 1.4867,
      "step": 27800
    },
    {
      "epoch": 0.7,
      "grad_norm": 6.74609375,
      "learning_rate": 2.1738693467336686e-05,
      "loss": 1.4836,
      "step": 27900
    },
    {
      "epoch": 0.7,
      "grad_norm": 11.1875,
      "learning_rate": 2.170854271356784e-05,
      "loss": 1.4511,
      "step": 28000
    },
    {
      "epoch": 0.7,
      "grad_norm": 6.3984375,
      "learning_rate": 2.1678391959798997e-05,
      "loss": 1.4891,
      "step": 28100
    },
    {
      "epoch": 0.7,
      "grad_norm": 9.015625,
      "learning_rate": 2.1648241206030152e-05,
      "loss": 1.4966,
      "step": 28200
    },
    {
      "epoch": 0.71,
      "grad_norm": 6.51171875,
      "learning_rate": 2.1618090452261304e-05,
      "loss": 1.4795,
      "step": 28300
    },
    {
      "epoch": 0.71,
      "grad_norm": 4.52734375,
      "learning_rate": 2.1587939698492463e-05,
      "loss": 1.4697,
      "step": 28400
    },
    {
      "epoch": 0.71,
      "grad_norm": 6.171875,
      "learning_rate": 2.1557788944723618e-05,
      "loss": 1.4834,
      "step": 28500
    },
    {
      "epoch": 0.71,
      "grad_norm": 7.125,
      "learning_rate": 2.1527638190954773e-05,
      "loss": 1.4917,
      "step": 28600
    },
    {
      "epoch": 0.72,
      "grad_norm": 4.58984375,
      "learning_rate": 2.149748743718593e-05,
      "loss": 1.5206,
      "step": 28700
    },
    {
      "epoch": 0.72,
      "grad_norm": 5.5,
      "learning_rate": 2.1467336683417087e-05,
      "loss": 1.4826,
      "step": 28800
    },
    {
      "epoch": 0.72,
      "grad_norm": 5.23828125,
      "learning_rate": 2.1437185929648242e-05,
      "loss": 1.4568,
      "step": 28900
    },
    {
      "epoch": 0.72,
      "grad_norm": 3.708984375,
      "learning_rate": 2.1407035175879398e-05,
      "loss": 1.5095,
      "step": 29000
    },
    {
      "epoch": 0.73,
      "grad_norm": 8.46875,
      "learning_rate": 2.1376884422110553e-05,
      "loss": 1.4935,
      "step": 29100
    },
    {
      "epoch": 0.73,
      "grad_norm": 5.75,
      "learning_rate": 2.134673366834171e-05,
      "loss": 1.4553,
      "step": 29200
    },
    {
      "epoch": 0.73,
      "grad_norm": 5.58203125,
      "learning_rate": 2.1316582914572867e-05,
      "loss": 1.4365,
      "step": 29300
    },
    {
      "epoch": 0.73,
      "grad_norm": 8.6015625,
      "learning_rate": 2.128643216080402e-05,
      "loss": 1.4688,
      "step": 29400
    },
    {
      "epoch": 0.74,
      "grad_norm": 6.953125,
      "learning_rate": 2.1256281407035174e-05,
      "loss": 1.4744,
      "step": 29500
    },
    {
      "epoch": 0.74,
      "grad_norm": 4.53515625,
      "learning_rate": 2.1226130653266333e-05,
      "loss": 1.4777,
      "step": 29600
    },
    {
      "epoch": 0.74,
      "grad_norm": 5.68359375,
      "learning_rate": 2.1195979899497488e-05,
      "loss": 1.4753,
      "step": 29700
    },
    {
      "epoch": 0.74,
      "grad_norm": 7.76953125,
      "learning_rate": 2.1165829145728643e-05,
      "loss": 1.4355,
      "step": 29800
    },
    {
      "epoch": 0.75,
      "grad_norm": 9.9453125,
      "learning_rate": 2.11356783919598e-05,
      "loss": 1.5261,
      "step": 29900
    },
    {
      "epoch": 0.75,
      "grad_norm": 5.51171875,
      "learning_rate": 2.1105527638190957e-05,
      "loss": 1.4413,
      "step": 30000
    },
    {
      "epoch": 0.75,
      "eval_loss": 1.4673473834991455,
      "eval_runtime": 58.3856,
      "eval_samples_per_second": 17.127,
      "eval_steps_per_second": 4.282,
      "step": 30000
    },
    {
      "epoch": 0.75,
      "grad_norm": 7.42578125,
      "learning_rate": 2.1075376884422112e-05,
      "loss": 1.4513,
      "step": 30100
    },
    {
      "epoch": 0.76,
      "grad_norm": 5.3046875,
      "learning_rate": 2.1045226130653268e-05,
      "loss": 1.4775,
      "step": 30200
    },
    {
      "epoch": 0.76,
      "grad_norm": 7.609375,
      "learning_rate": 2.1015075376884423e-05,
      "loss": 1.4477,
      "step": 30300
    },
    {
      "epoch": 0.76,
      "grad_norm": 10.3203125,
      "learning_rate": 2.098492462311558e-05,
      "loss": 1.4584,
      "step": 30400
    },
    {
      "epoch": 0.76,
      "grad_norm": 6.4453125,
      "learning_rate": 2.0954773869346733e-05,
      "loss": 1.4934,
      "step": 30500
    },
    {
      "epoch": 0.77,
      "grad_norm": 3.701171875,
      "learning_rate": 2.092462311557789e-05,
      "loss": 1.4539,
      "step": 30600
    },
    {
      "epoch": 0.77,
      "grad_norm": 4.0390625,
      "learning_rate": 2.0894472361809044e-05,
      "loss": 1.4547,
      "step": 30700
    },
    {
      "epoch": 0.77,
      "grad_norm": 6.56640625,
      "learning_rate": 2.08643216080402e-05,
      "loss": 1.4646,
      "step": 30800
    },
    {
      "epoch": 0.77,
      "grad_norm": 8.1875,
      "learning_rate": 2.0834170854271358e-05,
      "loss": 1.4742,
      "step": 30900
    },
    {
      "epoch": 0.78,
      "grad_norm": 6.68359375,
      "learning_rate": 2.0804020100502513e-05,
      "loss": 1.4942,
      "step": 31000
    },
    {
      "epoch": 0.78,
      "grad_norm": 6.046875,
      "learning_rate": 2.077386934673367e-05,
      "loss": 1.4995,
      "step": 31100
    },
    {
      "epoch": 0.78,
      "grad_norm": 5.5390625,
      "learning_rate": 2.0743718592964824e-05,
      "loss": 1.4689,
      "step": 31200
    },
    {
      "epoch": 0.78,
      "grad_norm": 9.515625,
      "learning_rate": 2.0713567839195982e-05,
      "loss": 1.4649,
      "step": 31300
    },
    {
      "epoch": 0.79,
      "grad_norm": 12.640625,
      "learning_rate": 2.0683417085427138e-05,
      "loss": 1.4445,
      "step": 31400
    },
    {
      "epoch": 0.79,
      "grad_norm": 8.53125,
      "learning_rate": 2.0653266331658293e-05,
      "loss": 1.4136,
      "step": 31500
    },
    {
      "epoch": 0.79,
      "grad_norm": 9.1640625,
      "learning_rate": 2.0623115577889448e-05,
      "loss": 1.4387,
      "step": 31600
    },
    {
      "epoch": 0.79,
      "grad_norm": 6.32421875,
      "learning_rate": 2.0592964824120603e-05,
      "loss": 1.4485,
      "step": 31700
    },
    {
      "epoch": 0.8,
      "grad_norm": 7.34765625,
      "learning_rate": 2.056281407035176e-05,
      "loss": 1.4799,
      "step": 31800
    },
    {
      "epoch": 0.8,
      "grad_norm": 5.546875,
      "learning_rate": 2.0532663316582914e-05,
      "loss": 1.4794,
      "step": 31900
    },
    {
      "epoch": 0.8,
      "grad_norm": 7.671875,
      "learning_rate": 2.050251256281407e-05,
      "loss": 1.4681,
      "step": 32000
    },
    {
      "epoch": 0.8,
      "grad_norm": 4.66796875,
      "learning_rate": 2.0472361809045228e-05,
      "loss": 1.4127,
      "step": 32100
    },
    {
      "epoch": 0.81,
      "grad_norm": 9.65625,
      "learning_rate": 2.0442211055276383e-05,
      "loss": 1.4809,
      "step": 32200
    },
    {
      "epoch": 0.81,
      "grad_norm": 7.37890625,
      "learning_rate": 2.041206030150754e-05,
      "loss": 1.4713,
      "step": 32300
    },
    {
      "epoch": 0.81,
      "grad_norm": 8.15625,
      "learning_rate": 2.0381909547738694e-05,
      "loss": 1.4872,
      "step": 32400
    },
    {
      "epoch": 0.81,
      "grad_norm": 8.5234375,
      "learning_rate": 2.0351758793969852e-05,
      "loss": 1.465,
      "step": 32500
    },
    {
      "epoch": 0.81,
      "eval_loss": 1.4116058349609375,
      "eval_runtime": 58.4781,
      "eval_samples_per_second": 17.1,
      "eval_steps_per_second": 4.275,
      "step": 32500
    },
    {
      "epoch": 0.81,
      "grad_norm": 8.0859375,
      "learning_rate": 2.0321608040201008e-05,
      "loss": 1.4717,
      "step": 32600
    },
    {
      "epoch": 0.82,
      "grad_norm": 8.34375,
      "learning_rate": 2.0291457286432163e-05,
      "loss": 1.4254,
      "step": 32700
    },
    {
      "epoch": 0.82,
      "grad_norm": 5.984375,
      "learning_rate": 2.0261306532663315e-05,
      "loss": 1.5016,
      "step": 32800
    },
    {
      "epoch": 0.82,
      "grad_norm": 5.50390625,
      "learning_rate": 2.023115577889447e-05,
      "loss": 1.4433,
      "step": 32900
    },
    {
      "epoch": 0.82,
      "grad_norm": 3.642578125,
      "learning_rate": 2.020100502512563e-05,
      "loss": 1.3997,
      "step": 33000
    },
    {
      "epoch": 0.83,
      "grad_norm": 7.01953125,
      "learning_rate": 2.0170854271356784e-05,
      "loss": 1.4126,
      "step": 33100
    },
    {
      "epoch": 0.83,
      "grad_norm": 6.78515625,
      "learning_rate": 2.014070351758794e-05,
      "loss": 1.4777,
      "step": 33200
    },
    {
      "epoch": 0.83,
      "grad_norm": 4.07421875,
      "learning_rate": 2.0110552763819094e-05,
      "loss": 1.4435,
      "step": 33300
    },
    {
      "epoch": 0.83,
      "grad_norm": 6.2265625,
      "learning_rate": 2.0080402010050253e-05,
      "loss": 1.4336,
      "step": 33400
    },
    {
      "epoch": 0.84,
      "grad_norm": 5.74609375,
      "learning_rate": 2.005025125628141e-05,
      "loss": 1.4649,
      "step": 33500
    },
    {
      "epoch": 0.84,
      "grad_norm": 3.865234375,
      "learning_rate": 2.0020100502512564e-05,
      "loss": 1.4996,
      "step": 33600
    },
    {
      "epoch": 0.84,
      "grad_norm": 6.6484375,
      "learning_rate": 1.998994974874372e-05,
      "loss": 1.4449,
      "step": 33700
    },
    {
      "epoch": 0.84,
      "grad_norm": 7.8046875,
      "learning_rate": 1.9959798994974878e-05,
      "loss": 1.4507,
      "step": 33800
    },
    {
      "epoch": 0.85,
      "grad_norm": 5.7421875,
      "learning_rate": 1.992964824120603e-05,
      "loss": 1.4989,
      "step": 33900
    },
    {
      "epoch": 0.85,
      "grad_norm": 5.15234375,
      "learning_rate": 1.9899497487437185e-05,
      "loss": 1.4287,
      "step": 34000
    },
    {
      "epoch": 0.85,
      "grad_norm": 4.73828125,
      "learning_rate": 1.986934673366834e-05,
      "loss": 1.4336,
      "step": 34100
    },
    {
      "epoch": 0.85,
      "grad_norm": 5.66796875,
      "learning_rate": 1.98391959798995e-05,
      "loss": 1.4686,
      "step": 34200
    },
    {
      "epoch": 0.86,
      "grad_norm": 4.45703125,
      "learning_rate": 1.9809045226130654e-05,
      "loss": 1.4165,
      "step": 34300
    },
    {
      "epoch": 0.86,
      "grad_norm": 10.9453125,
      "learning_rate": 1.977889447236181e-05,
      "loss": 1.4619,
      "step": 34400
    },
    {
      "epoch": 0.86,
      "grad_norm": 7.75,
      "learning_rate": 1.9748743718592964e-05,
      "loss": 1.4575,
      "step": 34500
    },
    {
      "epoch": 0.86,
      "grad_norm": 7.22265625,
      "learning_rate": 1.9718592964824123e-05,
      "loss": 1.5055,
      "step": 34600
    },
    {
      "epoch": 0.87,
      "grad_norm": 6.09375,
      "learning_rate": 1.968844221105528e-05,
      "loss": 1.4246,
      "step": 34700
    },
    {
      "epoch": 0.87,
      "grad_norm": 5.6328125,
      "learning_rate": 1.9658291457286434e-05,
      "loss": 1.4136,
      "step": 34800
    },
    {
      "epoch": 0.87,
      "grad_norm": 7.3359375,
      "learning_rate": 1.962814070351759e-05,
      "loss": 1.4495,
      "step": 34900
    },
    {
      "epoch": 0.88,
      "grad_norm": 5.203125,
      "learning_rate": 1.9597989949748744e-05,
      "loss": 1.4444,
      "step": 35000
    },
    {
      "epoch": 0.88,
      "eval_loss": 1.447482943534851,
      "eval_runtime": 58.4319,
      "eval_samples_per_second": 17.114,
      "eval_steps_per_second": 4.278,
      "step": 35000
    },
    {
      "epoch": 0.88,
      "grad_norm": 6.86328125,
      "learning_rate": 1.95678391959799e-05,
      "loss": 1.4882,
      "step": 35100
    },
    {
      "epoch": 0.88,
      "grad_norm": 8.125,
      "learning_rate": 1.9537688442211055e-05,
      "loss": 1.4831,
      "step": 35200
    },
    {
      "epoch": 0.88,
      "grad_norm": 7.953125,
      "learning_rate": 1.950753768844221e-05,
      "loss": 1.391,
      "step": 35300
    },
    {
      "epoch": 0.89,
      "grad_norm": 9.1875,
      "learning_rate": 1.9477386934673365e-05,
      "loss": 1.4193,
      "step": 35400
    },
    {
      "epoch": 0.89,
      "grad_norm": 5.23828125,
      "learning_rate": 1.9447236180904524e-05,
      "loss": 1.4528,
      "step": 35500
    },
    {
      "epoch": 0.89,
      "grad_norm": 7.90234375,
      "learning_rate": 1.941708542713568e-05,
      "loss": 1.4088,
      "step": 35600
    },
    {
      "epoch": 0.89,
      "grad_norm": 6.12109375,
      "learning_rate": 1.9386934673366834e-05,
      "loss": 1.4242,
      "step": 35700
    },
    {
      "epoch": 0.9,
      "grad_norm": 5.70703125,
      "learning_rate": 1.935678391959799e-05,
      "loss": 1.382,
      "step": 35800
    },
    {
      "epoch": 0.9,
      "grad_norm": 18.375,
      "learning_rate": 1.932663316582915e-05,
      "loss": 1.4321,
      "step": 35900
    },
    {
      "epoch": 0.9,
      "grad_norm": 7.921875,
      "learning_rate": 1.9296482412060304e-05,
      "loss": 1.4566,
      "step": 36000
    },
    {
      "epoch": 0.9,
      "grad_norm": 6.87109375,
      "learning_rate": 1.926633165829146e-05,
      "loss": 1.4037,
      "step": 36100
    },
    {
      "epoch": 0.91,
      "grad_norm": 10.6953125,
      "learning_rate": 1.923618090452261e-05,
      "loss": 1.4484,
      "step": 36200
    },
    {
      "epoch": 0.91,
      "grad_norm": 9.4765625,
      "learning_rate": 1.920603015075377e-05,
      "loss": 1.4357,
      "step": 36300
    },
    {
      "epoch": 0.91,
      "grad_norm": 8.421875,
      "learning_rate": 1.9175879396984925e-05,
      "loss": 1.3922,
      "step": 36400
    },
    {
      "epoch": 0.91,
      "grad_norm": 7.39453125,
      "learning_rate": 1.914572864321608e-05,
      "loss": 1.4405,
      "step": 36500
    },
    {
      "epoch": 0.92,
      "grad_norm": 11.0625,
      "learning_rate": 1.9115577889447235e-05,
      "loss": 1.4351,
      "step": 36600
    },
    {
      "epoch": 0.92,
      "grad_norm": 4.6796875,
      "learning_rate": 1.9085427135678394e-05,
      "loss": 1.4234,
      "step": 36700
    },
    {
      "epoch": 0.92,
      "grad_norm": 3.60546875,
      "learning_rate": 1.905527638190955e-05,
      "loss": 1.4558,
      "step": 36800
    },
    {
      "epoch": 0.92,
      "grad_norm": 5.05078125,
      "learning_rate": 1.9025125628140704e-05,
      "loss": 1.4191,
      "step": 36900
    },
    {
      "epoch": 0.93,
      "grad_norm": 7.02734375,
      "learning_rate": 1.899497487437186e-05,
      "loss": 1.4019,
      "step": 37000
    },
    {
      "epoch": 0.93,
      "grad_norm": 5.7421875,
      "learning_rate": 1.896482412060302e-05,
      "loss": 1.4376,
      "step": 37100
    },
    {
      "epoch": 0.93,
      "grad_norm": 9.09375,
      "learning_rate": 1.8934673366834174e-05,
      "loss": 1.448,
      "step": 37200
    },
    {
      "epoch": 0.93,
      "grad_norm": 5.546875,
      "learning_rate": 1.8904522613065325e-05,
      "loss": 1.5019,
      "step": 37300
    },
    {
      "epoch": 0.94,
      "grad_norm": 5.43359375,
      "learning_rate": 1.887437185929648e-05,
      "loss": 1.4002,
      "step": 37400
    },
    {
      "epoch": 0.94,
      "grad_norm": 6.73046875,
      "learning_rate": 1.884422110552764e-05,
      "loss": 1.4595,
      "step": 37500
    },
    {
      "epoch": 0.94,
      "eval_loss": 1.4337172508239746,
      "eval_runtime": 58.4211,
      "eval_samples_per_second": 17.117,
      "eval_steps_per_second": 4.279,
      "step": 37500
    },
    {
      "epoch": 0.94,
      "grad_norm": 3.904296875,
      "learning_rate": 1.8814070351758795e-05,
      "loss": 1.4425,
      "step": 37600
    },
    {
      "epoch": 0.94,
      "grad_norm": 6.24609375,
      "learning_rate": 1.878391959798995e-05,
      "loss": 1.4301,
      "step": 37700
    },
    {
      "epoch": 0.94,
      "grad_norm": 6.34375,
      "learning_rate": 1.8753768844221105e-05,
      "loss": 1.4815,
      "step": 37800
    },
    {
      "epoch": 0.95,
      "grad_norm": 7.95703125,
      "learning_rate": 1.872361809045226e-05,
      "loss": 1.4687,
      "step": 37900
    },
    {
      "epoch": 0.95,
      "grad_norm": 3.689453125,
      "learning_rate": 1.869346733668342e-05,
      "loss": 1.4096,
      "step": 38000
    },
    {
      "epoch": 0.95,
      "grad_norm": 6.2578125,
      "learning_rate": 1.8663316582914574e-05,
      "loss": 1.4627,
      "step": 38100
    },
    {
      "epoch": 0.95,
      "grad_norm": 7.89453125,
      "learning_rate": 1.863316582914573e-05,
      "loss": 1.4379,
      "step": 38200
    },
    {
      "epoch": 0.96,
      "grad_norm": 9.0546875,
      "learning_rate": 1.8603015075376885e-05,
      "loss": 1.4455,
      "step": 38300
    },
    {
      "epoch": 0.96,
      "grad_norm": 3.818359375,
      "learning_rate": 1.857286432160804e-05,
      "loss": 1.4592,
      "step": 38400
    },
    {
      "epoch": 0.96,
      "grad_norm": 6.12109375,
      "learning_rate": 1.8542713567839195e-05,
      "loss": 1.4622,
      "step": 38500
    },
    {
      "epoch": 0.96,
      "grad_norm": 6.3671875,
      "learning_rate": 1.851256281407035e-05,
      "loss": 1.4051,
      "step": 38600
    },
    {
      "epoch": 0.97,
      "grad_norm": 6.67578125,
      "learning_rate": 1.8482412060301506e-05,
      "loss": 1.4165,
      "step": 38700
    },
    {
      "epoch": 0.97,
      "grad_norm": 9.421875,
      "learning_rate": 1.8452261306532665e-05,
      "loss": 1.4038,
      "step": 38800
    },
    {
      "epoch": 0.97,
      "grad_norm": 6.25390625,
      "learning_rate": 1.842211055276382e-05,
      "loss": 1.3569,
      "step": 38900
    },
    {
      "epoch": 0.97,
      "grad_norm": 6.1953125,
      "learning_rate": 1.8391959798994975e-05,
      "loss": 1.4433,
      "step": 39000
    },
    {
      "epoch": 0.98,
      "grad_norm": 6.71875,
      "learning_rate": 1.836180904522613e-05,
      "loss": 1.4515,
      "step": 39100
    },
    {
      "epoch": 0.98,
      "grad_norm": 7.2890625,
      "learning_rate": 1.833165829145729e-05,
      "loss": 1.4557,
      "step": 39200
    },
    {
      "epoch": 0.98,
      "grad_norm": 4.93359375,
      "learning_rate": 1.8301507537688444e-05,
      "loss": 1.4292,
      "step": 39300
    },
    {
      "epoch": 0.98,
      "grad_norm": 5.2734375,
      "learning_rate": 1.82713567839196e-05,
      "loss": 1.3862,
      "step": 39400
    },
    {
      "epoch": 0.99,
      "grad_norm": 3.75,
      "learning_rate": 1.8241206030150755e-05,
      "loss": 1.3642,
      "step": 39500
    },
    {
      "epoch": 0.99,
      "grad_norm": 6.80078125,
      "learning_rate": 1.821105527638191e-05,
      "loss": 1.4442,
      "step": 39600
    },
    {
      "epoch": 0.99,
      "grad_norm": 4.80078125,
      "learning_rate": 1.8180904522613065e-05,
      "loss": 1.3977,
      "step": 39700
    },
    {
      "epoch": 0.99,
      "grad_norm": 3.52734375,
      "learning_rate": 1.815075376884422e-05,
      "loss": 1.4081,
      "step": 39800
    },
    {
      "epoch": 1.0,
      "grad_norm": 9.40625,
      "learning_rate": 1.8120603015075376e-05,
      "loss": 1.3949,
      "step": 39900
    },
    {
      "epoch": 1.0,
      "grad_norm": 10.0234375,
      "learning_rate": 1.809045226130653e-05,
      "loss": 1.3952,
      "step": 40000
    },
    {
      "epoch": 1.0,
      "eval_loss": 1.4081411361694336,
      "eval_runtime": 58.4216,
      "eval_samples_per_second": 17.117,
      "eval_steps_per_second": 4.279,
      "step": 40000
    },
    {
      "epoch": 1.0,
      "grad_norm": 7.73828125,
      "learning_rate": 1.806030150753769e-05,
      "loss": 1.3862,
      "step": 40100
    },
    {
      "epoch": 1.0,
      "grad_norm": 6.5859375,
      "learning_rate": 1.8030150753768845e-05,
      "loss": 1.4586,
      "step": 40200
    },
    {
      "epoch": 1.01,
      "grad_norm": 7.20703125,
      "learning_rate": 1.8e-05,
      "loss": 1.4018,
      "step": 40300
    },
    {
      "epoch": 1.01,
      "grad_norm": 8.2265625,
      "learning_rate": 1.7969849246231156e-05,
      "loss": 1.4112,
      "step": 40400
    },
    {
      "epoch": 1.01,
      "grad_norm": 7.18359375,
      "learning_rate": 1.7939698492462314e-05,
      "loss": 1.3652,
      "step": 40500
    },
    {
      "epoch": 1.01,
      "grad_norm": 6.2109375,
      "learning_rate": 1.790954773869347e-05,
      "loss": 1.4143,
      "step": 40600
    },
    {
      "epoch": 1.02,
      "grad_norm": 4.31640625,
      "learning_rate": 1.787939698492462e-05,
      "loss": 1.4024,
      "step": 40700
    },
    {
      "epoch": 1.02,
      "grad_norm": 7.57421875,
      "learning_rate": 1.7849246231155777e-05,
      "loss": 1.442,
      "step": 40800
    },
    {
      "epoch": 1.02,
      "grad_norm": 6.2109375,
      "learning_rate": 1.7819095477386935e-05,
      "loss": 1.3748,
      "step": 40900
    },
    {
      "epoch": 1.02,
      "grad_norm": 6.99609375,
      "learning_rate": 1.778894472361809e-05,
      "loss": 1.4098,
      "step": 41000
    },
    {
      "epoch": 1.03,
      "grad_norm": 9.4453125,
      "learning_rate": 1.7758793969849246e-05,
      "loss": 1.3765,
      "step": 41100
    },
    {
      "epoch": 1.03,
      "grad_norm": 8.390625,
      "learning_rate": 1.77286432160804e-05,
      "loss": 1.4175,
      "step": 41200
    },
    {
      "epoch": 1.03,
      "grad_norm": 6.94140625,
      "learning_rate": 1.769849246231156e-05,
      "loss": 1.3656,
      "step": 41300
    },
    {
      "epoch": 1.03,
      "grad_norm": 8.8203125,
      "learning_rate": 1.7668341708542715e-05,
      "loss": 1.319,
      "step": 41400
    },
    {
      "epoch": 1.04,
      "grad_norm": 9.3125,
      "learning_rate": 1.763819095477387e-05,
      "loss": 1.4038,
      "step": 41500
    },
    {
      "epoch": 1.04,
      "grad_norm": 4.515625,
      "learning_rate": 1.7608040201005026e-05,
      "loss": 1.4096,
      "step": 41600
    },
    {
      "epoch": 1.04,
      "grad_norm": 6.9375,
      "learning_rate": 1.7577889447236184e-05,
      "loss": 1.4093,
      "step": 41700
    },
    {
      "epoch": 1.04,
      "grad_norm": 5.62890625,
      "learning_rate": 1.7547738693467336e-05,
      "loss": 1.4099,
      "step": 41800
    },
    {
      "epoch": 1.05,
      "grad_norm": 6.1484375,
      "learning_rate": 1.751758793969849e-05,
      "loss": 1.3245,
      "step": 41900
    },
    {
      "epoch": 1.05,
      "grad_norm": 7.41796875,
      "learning_rate": 1.7487437185929647e-05,
      "loss": 1.4454,
      "step": 42000
    },
    {
      "epoch": 1.05,
      "grad_norm": 4.08203125,
      "learning_rate": 1.7457286432160805e-05,
      "loss": 1.4143,
      "step": 42100
    },
    {
      "epoch": 1.05,
      "grad_norm": 8.3046875,
      "learning_rate": 1.742713567839196e-05,
      "loss": 1.3896,
      "step": 42200
    },
    {
      "epoch": 1.06,
      "grad_norm": 6.328125,
      "learning_rate": 1.7396984924623116e-05,
      "loss": 1.4059,
      "step": 42300
    },
    {
      "epoch": 1.06,
      "grad_norm": 6.92578125,
      "learning_rate": 1.736683417085427e-05,
      "loss": 1.3806,
      "step": 42400
    },
    {
      "epoch": 1.06,
      "grad_norm": 4.88671875,
      "learning_rate": 1.7336683417085427e-05,
      "loss": 1.382,
      "step": 42500
    },
    {
      "epoch": 1.06,
      "eval_loss": 1.4033676385879517,
      "eval_runtime": 58.3505,
      "eval_samples_per_second": 17.138,
      "eval_steps_per_second": 4.284,
      "step": 42500
    },
    {
      "epoch": 1.06,
      "grad_norm": 4.0390625,
      "learning_rate": 1.7306532663316585e-05,
      "loss": 1.4022,
      "step": 42600
    },
    {
      "epoch": 1.07,
      "grad_norm": 7.046875,
      "learning_rate": 1.727638190954774e-05,
      "loss": 1.3839,
      "step": 42700
    },
    {
      "epoch": 1.07,
      "grad_norm": 6.36328125,
      "learning_rate": 1.7246231155778896e-05,
      "loss": 1.3864,
      "step": 42800
    },
    {
      "epoch": 1.07,
      "grad_norm": 7.46484375,
      "learning_rate": 1.721608040201005e-05,
      "loss": 1.3551,
      "step": 42900
    },
    {
      "epoch": 1.07,
      "grad_norm": 6.73828125,
      "learning_rate": 1.7185929648241206e-05,
      "loss": 1.4038,
      "step": 43000
    },
    {
      "epoch": 1.08,
      "grad_norm": 5.7109375,
      "learning_rate": 1.715577889447236e-05,
      "loss": 1.4333,
      "step": 43100
    },
    {
      "epoch": 1.08,
      "grad_norm": 6.84765625,
      "learning_rate": 1.7125628140703517e-05,
      "loss": 1.3559,
      "step": 43200
    },
    {
      "epoch": 1.08,
      "grad_norm": 6.77734375,
      "learning_rate": 1.7095477386934672e-05,
      "loss": 1.3974,
      "step": 43300
    },
    {
      "epoch": 1.08,
      "grad_norm": 7.0703125,
      "learning_rate": 1.706532663316583e-05,
      "loss": 1.4373,
      "step": 43400
    },
    {
      "epoch": 1.09,
      "grad_norm": 6.5390625,
      "learning_rate": 1.7035175879396986e-05,
      "loss": 1.441,
      "step": 43500
    },
    {
      "epoch": 1.09,
      "grad_norm": 5.91015625,
      "learning_rate": 1.700502512562814e-05,
      "loss": 1.4241,
      "step": 43600
    },
    {
      "epoch": 1.09,
      "grad_norm": 8.2421875,
      "learning_rate": 1.6974874371859296e-05,
      "loss": 1.3692,
      "step": 43700
    },
    {
      "epoch": 1.09,
      "grad_norm": 5.36328125,
      "learning_rate": 1.6944723618090455e-05,
      "loss": 1.4058,
      "step": 43800
    },
    {
      "epoch": 1.1,
      "grad_norm": 5.71484375,
      "learning_rate": 1.691457286432161e-05,
      "loss": 1.3616,
      "step": 43900
    },
    {
      "epoch": 1.1,
      "grad_norm": 7.41015625,
      "learning_rate": 1.6884422110552766e-05,
      "loss": 1.4298,
      "step": 44000
    },
    {
      "epoch": 1.1,
      "grad_norm": 7.79296875,
      "learning_rate": 1.6854271356783918e-05,
      "loss": 1.4036,
      "step": 44100
    },
    {
      "epoch": 1.1,
      "grad_norm": 9.09375,
      "learning_rate": 1.6824120603015076e-05,
      "loss": 1.3283,
      "step": 44200
    },
    {
      "epoch": 1.11,
      "grad_norm": 5.484375,
      "learning_rate": 1.679396984924623e-05,
      "loss": 1.3767,
      "step": 44300
    },
    {
      "epoch": 1.11,
      "grad_norm": 8.5,
      "learning_rate": 1.6763819095477387e-05,
      "loss": 1.3854,
      "step": 44400
    },
    {
      "epoch": 1.11,
      "grad_norm": 5.47265625,
      "learning_rate": 1.6733668341708542e-05,
      "loss": 1.4026,
      "step": 44500
    },
    {
      "epoch": 1.11,
      "grad_norm": 4.12109375,
      "learning_rate": 1.6703517587939697e-05,
      "loss": 1.3708,
      "step": 44600
    },
    {
      "epoch": 1.12,
      "grad_norm": 4.5625,
      "learning_rate": 1.6673366834170856e-05,
      "loss": 1.4159,
      "step": 44700
    },
    {
      "epoch": 1.12,
      "grad_norm": 7.28515625,
      "learning_rate": 1.664321608040201e-05,
      "loss": 1.3806,
      "step": 44800
    },
    {
      "epoch": 1.12,
      "grad_norm": 7.31640625,
      "learning_rate": 1.6613065326633166e-05,
      "loss": 1.3393,
      "step": 44900
    },
    {
      "epoch": 1.12,
      "grad_norm": 7.17578125,
      "learning_rate": 1.6582914572864322e-05,
      "loss": 1.34,
      "step": 45000
    },
    {
      "epoch": 1.12,
      "eval_loss": 1.4302189350128174,
      "eval_runtime": 58.4696,
      "eval_samples_per_second": 17.103,
      "eval_steps_per_second": 4.276,
      "step": 45000
    },
    {
      "epoch": 1.13,
      "grad_norm": 7.31640625,
      "learning_rate": 1.655276381909548e-05,
      "loss": 1.392,
      "step": 45100
    },
    {
      "epoch": 1.13,
      "grad_norm": 4.79296875,
      "learning_rate": 1.6522613065326632e-05,
      "loss": 1.4112,
      "step": 45200
    },
    {
      "epoch": 1.13,
      "grad_norm": 5.48828125,
      "learning_rate": 1.6492462311557788e-05,
      "loss": 1.3707,
      "step": 45300
    },
    {
      "epoch": 1.14,
      "grad_norm": 4.9921875,
      "learning_rate": 1.6462311557788943e-05,
      "loss": 1.4066,
      "step": 45400
    },
    {
      "epoch": 1.14,
      "grad_norm": 8.21875,
      "learning_rate": 1.64321608040201e-05,
      "loss": 1.4262,
      "step": 45500
    },
    {
      "epoch": 1.14,
      "grad_norm": 7.63671875,
      "learning_rate": 1.6402010050251257e-05,
      "loss": 1.3637,
      "step": 45600
    },
    {
      "epoch": 1.14,
      "grad_norm": 6.12890625,
      "learning_rate": 1.6371859296482412e-05,
      "loss": 1.3972,
      "step": 45700
    },
    {
      "epoch": 1.15,
      "grad_norm": 6.765625,
      "learning_rate": 1.6341708542713567e-05,
      "loss": 1.3538,
      "step": 45800
    },
    {
      "epoch": 1.15,
      "grad_norm": 5.44921875,
      "learning_rate": 1.6311557788944726e-05,
      "loss": 1.3902,
      "step": 45900
    },
    {
      "epoch": 1.15,
      "grad_norm": 5.734375,
      "learning_rate": 1.628140703517588e-05,
      "loss": 1.407,
      "step": 46000
    },
    {
      "epoch": 1.15,
      "grad_norm": 7.203125,
      "learning_rate": 1.6251256281407036e-05,
      "loss": 1.3354,
      "step": 46100
    },
    {
      "epoch": 1.16,
      "grad_norm": 7.87109375,
      "learning_rate": 1.6221105527638192e-05,
      "loss": 1.3464,
      "step": 46200
    },
    {
      "epoch": 1.16,
      "grad_norm": 3.75,
      "learning_rate": 1.6190954773869347e-05,
      "loss": 1.3949,
      "step": 46300
    },
    {
      "epoch": 1.16,
      "grad_norm": 6.1875,
      "learning_rate": 1.6160804020100502e-05,
      "loss": 1.4038,
      "step": 46400
    },
    {
      "epoch": 1.16,
      "grad_norm": 6.44140625,
      "learning_rate": 1.6130653266331658e-05,
      "loss": 1.3513,
      "step": 46500
    },
    {
      "epoch": 1.17,
      "grad_norm": 3.861328125,
      "learning_rate": 1.6100502512562813e-05,
      "loss": 1.349,
      "step": 46600
    },
    {
      "epoch": 1.17,
      "grad_norm": 6.92578125,
      "learning_rate": 1.607035175879397e-05,
      "loss": 1.3736,
      "step": 46700
    },
    {
      "epoch": 1.17,
      "grad_norm": 5.6328125,
      "learning_rate": 1.6040201005025127e-05,
      "loss": 1.392,
      "step": 46800
    },
    {
      "epoch": 1.17,
      "grad_norm": 8.125,
      "learning_rate": 1.6010050251256282e-05,
      "loss": 1.387,
      "step": 46900
    },
    {
      "epoch": 1.18,
      "grad_norm": 10.765625,
      "learning_rate": 1.5979899497487437e-05,
      "loss": 1.365,
      "step": 47000
    },
    {
      "epoch": 1.18,
      "grad_norm": 5.28125,
      "learning_rate": 1.5949748743718593e-05,
      "loss": 1.3596,
      "step": 47100
    },
    {
      "epoch": 1.18,
      "grad_norm": 5.96875,
      "learning_rate": 1.591959798994975e-05,
      "loss": 1.3406,
      "step": 47200
    },
    {
      "epoch": 1.18,
      "grad_norm": 4.4609375,
      "learning_rate": 1.5889447236180906e-05,
      "loss": 1.3779,
      "step": 47300
    },
    {
      "epoch": 1.19,
      "grad_norm": 4.47265625,
      "learning_rate": 1.5859296482412062e-05,
      "loss": 1.3254,
      "step": 47400
    },
    {
      "epoch": 1.19,
      "grad_norm": 5.77734375,
      "learning_rate": 1.5829145728643214e-05,
      "loss": 1.3618,
      "step": 47500
    },
    {
      "epoch": 1.19,
      "eval_loss": 1.3727421760559082,
      "eval_runtime": 58.4695,
      "eval_samples_per_second": 17.103,
      "eval_steps_per_second": 4.276,
      "step": 47500
    },
    {
      "epoch": 1.19,
      "grad_norm": 4.19140625,
      "learning_rate": 1.5798994974874372e-05,
      "loss": 1.3738,
      "step": 47600
    },
    {
      "epoch": 1.19,
      "grad_norm": 7.140625,
      "learning_rate": 1.5768844221105528e-05,
      "loss": 1.3791,
      "step": 47700
    },
    {
      "epoch": 1.2,
      "grad_norm": 7.6015625,
      "learning_rate": 1.5738693467336683e-05,
      "loss": 1.3381,
      "step": 47800
    },
    {
      "epoch": 1.2,
      "grad_norm": 6.31640625,
      "learning_rate": 1.5708542713567838e-05,
      "loss": 1.3935,
      "step": 47900
    },
    {
      "epoch": 1.2,
      "grad_norm": 5.19921875,
      "learning_rate": 1.5678391959798997e-05,
      "loss": 1.3314,
      "step": 48000
    },
    {
      "epoch": 1.2,
      "grad_norm": 8.390625,
      "learning_rate": 1.5648241206030152e-05,
      "loss": 1.3614,
      "step": 48100
    },
    {
      "epoch": 1.21,
      "grad_norm": 4.45703125,
      "learning_rate": 1.5618090452261307e-05,
      "loss": 1.3347,
      "step": 48200
    },
    {
      "epoch": 1.21,
      "grad_norm": 4.8125,
      "learning_rate": 1.5587939698492463e-05,
      "loss": 1.3963,
      "step": 48300
    },
    {
      "epoch": 1.21,
      "grad_norm": 5.7734375,
      "learning_rate": 1.555778894472362e-05,
      "loss": 1.4068,
      "step": 48400
    },
    {
      "epoch": 1.21,
      "grad_norm": 8.109375,
      "learning_rate": 1.5527638190954776e-05,
      "loss": 1.2939,
      "step": 48500
    },
    {
      "epoch": 1.22,
      "grad_norm": 5.67578125,
      "learning_rate": 1.5497487437185928e-05,
      "loss": 1.2909,
      "step": 48600
    },
    {
      "epoch": 1.22,
      "grad_norm": 5.0,
      "learning_rate": 1.5467336683417084e-05,
      "loss": 1.342,
      "step": 48700
    },
    {
      "epoch": 1.22,
      "grad_norm": 6.62109375,
      "learning_rate": 1.5437185929648242e-05,
      "loss": 1.3664,
      "step": 48800
    },
    {
      "epoch": 1.22,
      "grad_norm": 4.0,
      "learning_rate": 1.5407035175879397e-05,
      "loss": 1.3519,
      "step": 48900
    },
    {
      "epoch": 1.23,
      "grad_norm": 7.35546875,
      "learning_rate": 1.5376884422110553e-05,
      "loss": 1.3675,
      "step": 49000
    },
    {
      "epoch": 1.23,
      "grad_norm": 7.02734375,
      "learning_rate": 1.5346733668341708e-05,
      "loss": 1.3635,
      "step": 49100
    },
    {
      "epoch": 1.23,
      "grad_norm": 10.796875,
      "learning_rate": 1.5316582914572863e-05,
      "loss": 1.3527,
      "step": 49200
    },
    {
      "epoch": 1.23,
      "grad_norm": 6.140625,
      "learning_rate": 1.5286432160804022e-05,
      "loss": 1.3308,
      "step": 49300
    },
    {
      "epoch": 1.23,
      "grad_norm": 7.33984375,
      "learning_rate": 1.5256281407035177e-05,
      "loss": 1.3418,
      "step": 49400
    },
    {
      "epoch": 1.24,
      "grad_norm": 8.34375,
      "learning_rate": 1.522613065326633e-05,
      "loss": 1.3602,
      "step": 49500
    },
    {
      "epoch": 1.24,
      "grad_norm": 6.44140625,
      "learning_rate": 1.5195979899497486e-05,
      "loss": 1.4051,
      "step": 49600
    },
    {
      "epoch": 1.24,
      "grad_norm": 5.20703125,
      "learning_rate": 1.5165829145728645e-05,
      "loss": 1.3698,
      "step": 49700
    },
    {
      "epoch": 1.25,
      "grad_norm": 4.57421875,
      "learning_rate": 1.51356783919598e-05,
      "loss": 1.369,
      "step": 49800
    },
    {
      "epoch": 1.25,
      "grad_norm": 5.0078125,
      "learning_rate": 1.5105527638190955e-05,
      "loss": 1.3848,
      "step": 49900
    },
    {
      "epoch": 1.25,
      "grad_norm": 6.70703125,
      "learning_rate": 1.5075376884422109e-05,
      "loss": 1.3905,
      "step": 50000
    },
    {
      "epoch": 1.25,
      "eval_loss": 1.3214133977890015,
      "eval_runtime": 58.3319,
      "eval_samples_per_second": 17.143,
      "eval_steps_per_second": 4.286,
      "step": 50000
    },
    {
      "epoch": 1.25,
      "grad_norm": 7.6484375,
      "learning_rate": 1.5045226130653267e-05,
      "loss": 1.3656,
      "step": 50100
    },
    {
      "epoch": 1.25,
      "grad_norm": 5.04296875,
      "learning_rate": 1.5015075376884423e-05,
      "loss": 1.3529,
      "step": 50200
    },
    {
      "epoch": 1.26,
      "grad_norm": 5.41796875,
      "learning_rate": 1.4984924623115578e-05,
      "loss": 1.348,
      "step": 50300
    },
    {
      "epoch": 1.26,
      "grad_norm": 4.77734375,
      "learning_rate": 1.4954773869346735e-05,
      "loss": 1.3442,
      "step": 50400
    },
    {
      "epoch": 1.26,
      "grad_norm": 5.02734375,
      "learning_rate": 1.492462311557789e-05,
      "loss": 1.34,
      "step": 50500
    },
    {
      "epoch": 1.27,
      "grad_norm": 7.89453125,
      "learning_rate": 1.4894472361809046e-05,
      "loss": 1.3631,
      "step": 50600
    },
    {
      "epoch": 1.27,
      "grad_norm": 17.5625,
      "learning_rate": 1.48643216080402e-05,
      "loss": 1.3492,
      "step": 50700
    },
    {
      "epoch": 1.27,
      "grad_norm": 6.22265625,
      "learning_rate": 1.4834170854271358e-05,
      "loss": 1.3604,
      "step": 50800
    },
    {
      "epoch": 1.27,
      "grad_norm": 5.5234375,
      "learning_rate": 1.4804020100502513e-05,
      "loss": 1.3412,
      "step": 50900
    },
    {
      "epoch": 1.27,
      "grad_norm": 5.93359375,
      "learning_rate": 1.477386934673367e-05,
      "loss": 1.3628,
      "step": 51000
    },
    {
      "epoch": 1.28,
      "grad_norm": 8.46875,
      "learning_rate": 1.4743718592964824e-05,
      "loss": 1.3496,
      "step": 51100
    },
    {
      "epoch": 1.28,
      "grad_norm": 6.32421875,
      "learning_rate": 1.471356783919598e-05,
      "loss": 1.3683,
      "step": 51200
    },
    {
      "epoch": 1.28,
      "grad_norm": 5.64453125,
      "learning_rate": 1.4683417085427136e-05,
      "loss": 1.3648,
      "step": 51300
    },
    {
      "epoch": 1.28,
      "grad_norm": 5.328125,
      "learning_rate": 1.4653266331658293e-05,
      "loss": 1.365,
      "step": 51400
    },
    {
      "epoch": 1.29,
      "grad_norm": 5.7578125,
      "learning_rate": 1.4623115577889448e-05,
      "loss": 1.3176,
      "step": 51500
    },
    {
      "epoch": 1.29,
      "grad_norm": 6.6875,
      "learning_rate": 1.4592964824120603e-05,
      "loss": 1.339,
      "step": 51600
    },
    {
      "epoch": 1.29,
      "grad_norm": 4.8671875,
      "learning_rate": 1.4562814070351759e-05,
      "loss": 1.3725,
      "step": 51700
    },
    {
      "epoch": 1.29,
      "grad_norm": 7.10546875,
      "learning_rate": 1.4532663316582914e-05,
      "loss": 1.3552,
      "step": 51800
    },
    {
      "epoch": 1.3,
      "grad_norm": 6.890625,
      "learning_rate": 1.450251256281407e-05,
      "loss": 1.2784,
      "step": 51900
    },
    {
      "epoch": 1.3,
      "grad_norm": 7.17578125,
      "learning_rate": 1.4472361809045226e-05,
      "loss": 1.3438,
      "step": 52000
    },
    {
      "epoch": 1.3,
      "grad_norm": 10.40625,
      "learning_rate": 1.4442211055276383e-05,
      "loss": 1.2733,
      "step": 52100
    },
    {
      "epoch": 1.3,
      "grad_norm": 7.32421875,
      "learning_rate": 1.4412060301507538e-05,
      "loss": 1.3616,
      "step": 52200
    },
    {
      "epoch": 1.31,
      "grad_norm": 3.615234375,
      "learning_rate": 1.4381909547738694e-05,
      "loss": 1.3346,
      "step": 52300
    },
    {
      "epoch": 1.31,
      "grad_norm": 6.54296875,
      "learning_rate": 1.4351758793969849e-05,
      "loss": 1.3893,
      "step": 52400
    },
    {
      "epoch": 1.31,
      "grad_norm": 7.2578125,
      "learning_rate": 1.4321608040201006e-05,
      "loss": 1.313,
      "step": 52500
    },
    {
      "epoch": 1.31,
      "eval_loss": 1.3575935363769531,
      "eval_runtime": 58.4472,
      "eval_samples_per_second": 17.109,
      "eval_steps_per_second": 4.277,
      "step": 52500
    },
    {
      "epoch": 1.31,
      "grad_norm": 7.12109375,
      "learning_rate": 1.4291457286432161e-05,
      "loss": 1.3669,
      "step": 52600
    },
    {
      "epoch": 1.32,
      "grad_norm": 5.38671875,
      "learning_rate": 1.4261306532663318e-05,
      "loss": 1.3356,
      "step": 52700
    },
    {
      "epoch": 1.32,
      "grad_norm": 6.97265625,
      "learning_rate": 1.4231155778894472e-05,
      "loss": 1.326,
      "step": 52800
    },
    {
      "epoch": 1.32,
      "grad_norm": 6.5078125,
      "learning_rate": 1.4201005025125629e-05,
      "loss": 1.4087,
      "step": 52900
    },
    {
      "epoch": 1.32,
      "grad_norm": 7.66015625,
      "learning_rate": 1.4170854271356784e-05,
      "loss": 1.3237,
      "step": 53000
    },
    {
      "epoch": 1.33,
      "grad_norm": 4.0,
      "learning_rate": 1.414070351758794e-05,
      "loss": 1.3516,
      "step": 53100
    },
    {
      "epoch": 1.33,
      "grad_norm": 6.90625,
      "learning_rate": 1.4110552763819096e-05,
      "loss": 1.3495,
      "step": 53200
    },
    {
      "epoch": 1.33,
      "grad_norm": 8.5625,
      "learning_rate": 1.4080402010050253e-05,
      "loss": 1.3308,
      "step": 53300
    },
    {
      "epoch": 1.33,
      "grad_norm": 8.078125,
      "learning_rate": 1.4050251256281407e-05,
      "loss": 1.3762,
      "step": 53400
    },
    {
      "epoch": 1.34,
      "grad_norm": 4.453125,
      "learning_rate": 1.4020100502512564e-05,
      "loss": 1.3929,
      "step": 53500
    },
    {
      "epoch": 1.34,
      "grad_norm": 8.0234375,
      "learning_rate": 1.3989949748743719e-05,
      "loss": 1.2895,
      "step": 53600
    },
    {
      "epoch": 1.34,
      "grad_norm": 8.125,
      "learning_rate": 1.3959798994974876e-05,
      "loss": 1.3772,
      "step": 53700
    },
    {
      "epoch": 1.34,
      "grad_norm": 5.9609375,
      "learning_rate": 1.3929648241206031e-05,
      "loss": 1.3427,
      "step": 53800
    },
    {
      "epoch": 1.35,
      "grad_norm": 6.0625,
      "learning_rate": 1.3899497487437186e-05,
      "loss": 1.3433,
      "step": 53900
    },
    {
      "epoch": 1.35,
      "grad_norm": 5.52734375,
      "learning_rate": 1.3869346733668342e-05,
      "loss": 1.3341,
      "step": 54000
    },
    {
      "epoch": 1.35,
      "grad_norm": 8.1328125,
      "learning_rate": 1.3839195979899497e-05,
      "loss": 1.377,
      "step": 54100
    },
    {
      "epoch": 1.35,
      "grad_norm": 5.265625,
      "learning_rate": 1.3809045226130654e-05,
      "loss": 1.3653,
      "step": 54200
    },
    {
      "epoch": 1.36,
      "grad_norm": 6.80859375,
      "learning_rate": 1.3778894472361809e-05,
      "loss": 1.3282,
      "step": 54300
    },
    {
      "epoch": 1.36,
      "grad_norm": 6.078125,
      "learning_rate": 1.3748743718592966e-05,
      "loss": 1.3511,
      "step": 54400
    },
    {
      "epoch": 1.36,
      "grad_norm": 7.08984375,
      "learning_rate": 1.371859296482412e-05,
      "loss": 1.3379,
      "step": 54500
    },
    {
      "epoch": 1.36,
      "grad_norm": 4.79296875,
      "learning_rate": 1.3688442211055277e-05,
      "loss": 1.3352,
      "step": 54600
    },
    {
      "epoch": 1.37,
      "grad_norm": 3.83984375,
      "learning_rate": 1.3658291457286432e-05,
      "loss": 1.3652,
      "step": 54700
    },
    {
      "epoch": 1.37,
      "grad_norm": 6.2734375,
      "learning_rate": 1.3628140703517589e-05,
      "loss": 1.3457,
      "step": 54800
    },
    {
      "epoch": 1.37,
      "grad_norm": 8.8203125,
      "learning_rate": 1.3597989949748744e-05,
      "loss": 1.3861,
      "step": 54900
    },
    {
      "epoch": 1.38,
      "grad_norm": 7.9921875,
      "learning_rate": 1.3567839195979901e-05,
      "loss": 1.3345,
      "step": 55000
    },
    {
      "epoch": 1.38,
      "eval_loss": 1.352059006690979,
      "eval_runtime": 58.4301,
      "eval_samples_per_second": 17.114,
      "eval_steps_per_second": 4.279,
      "step": 55000
    },
    {
      "epoch": 1.38,
      "grad_norm": 4.3046875,
      "learning_rate": 1.3537688442211055e-05,
      "loss": 1.3387,
      "step": 55100
    },
    {
      "epoch": 1.38,
      "grad_norm": 8.2734375,
      "learning_rate": 1.3507537688442212e-05,
      "loss": 1.3813,
      "step": 55200
    },
    {
      "epoch": 1.38,
      "grad_norm": 5.5078125,
      "learning_rate": 1.3477386934673367e-05,
      "loss": 1.3275,
      "step": 55300
    },
    {
      "epoch": 1.39,
      "grad_norm": 6.3125,
      "learning_rate": 1.3447236180904524e-05,
      "loss": 1.3102,
      "step": 55400
    },
    {
      "epoch": 1.39,
      "grad_norm": 6.84375,
      "learning_rate": 1.3417085427135679e-05,
      "loss": 1.3661,
      "step": 55500
    },
    {
      "epoch": 1.39,
      "grad_norm": 5.234375,
      "learning_rate": 1.3386934673366836e-05,
      "loss": 1.3706,
      "step": 55600
    },
    {
      "epoch": 1.39,
      "grad_norm": 6.6328125,
      "learning_rate": 1.335678391959799e-05,
      "loss": 1.3108,
      "step": 55700
    },
    {
      "epoch": 1.4,
      "grad_norm": 4.51171875,
      "learning_rate": 1.3326633165829147e-05,
      "loss": 1.3616,
      "step": 55800
    },
    {
      "epoch": 1.4,
      "grad_norm": 7.078125,
      "learning_rate": 1.3296482412060302e-05,
      "loss": 1.3603,
      "step": 55900
    },
    {
      "epoch": 1.4,
      "grad_norm": 5.26953125,
      "learning_rate": 1.3266331658291459e-05,
      "loss": 1.344,
      "step": 56000
    },
    {
      "epoch": 1.4,
      "grad_norm": 10.1328125,
      "learning_rate": 1.3236180904522614e-05,
      "loss": 1.3127,
      "step": 56100
    },
    {
      "epoch": 1.41,
      "grad_norm": 6.19140625,
      "learning_rate": 1.3206030150753768e-05,
      "loss": 1.3231,
      "step": 56200
    },
    {
      "epoch": 1.41,
      "grad_norm": 7.38671875,
      "learning_rate": 1.3175879396984925e-05,
      "loss": 1.3684,
      "step": 56300
    },
    {
      "epoch": 1.41,
      "grad_norm": 11.8203125,
      "learning_rate": 1.314572864321608e-05,
      "loss": 1.3543,
      "step": 56400
    },
    {
      "epoch": 1.41,
      "grad_norm": 6.22265625,
      "learning_rate": 1.3115577889447237e-05,
      "loss": 1.314,
      "step": 56500
    },
    {
      "epoch": 1.42,
      "grad_norm": 6.4609375,
      "learning_rate": 1.3085427135678392e-05,
      "loss": 1.3074,
      "step": 56600
    },
    {
      "epoch": 1.42,
      "grad_norm": 7.19921875,
      "learning_rate": 1.3055276381909549e-05,
      "loss": 1.3272,
      "step": 56700
    },
    {
      "epoch": 1.42,
      "grad_norm": 3.60546875,
      "learning_rate": 1.3025125628140703e-05,
      "loss": 1.2822,
      "step": 56800
    },
    {
      "epoch": 1.42,
      "grad_norm": 5.015625,
      "learning_rate": 1.299497487437186e-05,
      "loss": 1.3366,
      "step": 56900
    },
    {
      "epoch": 1.43,
      "grad_norm": 7.0859375,
      "learning_rate": 1.2964824120603015e-05,
      "loss": 1.3094,
      "step": 57000
    },
    {
      "epoch": 1.43,
      "grad_norm": 6.21484375,
      "learning_rate": 1.2934673366834172e-05,
      "loss": 1.313,
      "step": 57100
    },
    {
      "epoch": 1.43,
      "grad_norm": 8.9140625,
      "learning_rate": 1.2904522613065327e-05,
      "loss": 1.2882,
      "step": 57200
    },
    {
      "epoch": 1.43,
      "grad_norm": 4.85546875,
      "learning_rate": 1.2874371859296484e-05,
      "loss": 1.3425,
      "step": 57300
    },
    {
      "epoch": 1.44,
      "grad_norm": 7.71875,
      "learning_rate": 1.2844221105527638e-05,
      "loss": 1.3341,
      "step": 57400
    },
    {
      "epoch": 1.44,
      "grad_norm": 6.60546875,
      "learning_rate": 1.2814070351758795e-05,
      "loss": 1.2881,
      "step": 57500
    },
    {
      "epoch": 1.44,
      "eval_loss": 1.3498035669326782,
      "eval_runtime": 58.3991,
      "eval_samples_per_second": 17.124,
      "eval_steps_per_second": 4.281,
      "step": 57500
    },
    {
      "epoch": 1.44,
      "grad_norm": 5.59375,
      "learning_rate": 1.278391959798995e-05,
      "loss": 1.3218,
      "step": 57600
    },
    {
      "epoch": 1.44,
      "grad_norm": 7.36328125,
      "learning_rate": 1.2753768844221107e-05,
      "loss": 1.3059,
      "step": 57700
    },
    {
      "epoch": 1.45,
      "grad_norm": 6.1171875,
      "learning_rate": 1.2723618090452262e-05,
      "loss": 1.2796,
      "step": 57800
    },
    {
      "epoch": 1.45,
      "grad_norm": 5.33984375,
      "learning_rate": 1.2693467336683417e-05,
      "loss": 1.2952,
      "step": 57900
    },
    {
      "epoch": 1.45,
      "grad_norm": 7.80078125,
      "learning_rate": 1.2663316582914573e-05,
      "loss": 1.3225,
      "step": 58000
    },
    {
      "epoch": 1.45,
      "grad_norm": 6.1796875,
      "learning_rate": 1.263316582914573e-05,
      "loss": 1.332,
      "step": 58100
    },
    {
      "epoch": 1.46,
      "grad_norm": 6.26171875,
      "learning_rate": 1.2603015075376885e-05,
      "loss": 1.3215,
      "step": 58200
    },
    {
      "epoch": 1.46,
      "grad_norm": 7.08984375,
      "learning_rate": 1.2572864321608042e-05,
      "loss": 1.3355,
      "step": 58300
    },
    {
      "epoch": 1.46,
      "grad_norm": 5.46484375,
      "learning_rate": 1.2542713567839197e-05,
      "loss": 1.3136,
      "step": 58400
    },
    {
      "epoch": 1.46,
      "grad_norm": 7.5234375,
      "learning_rate": 1.251256281407035e-05,
      "loss": 1.3429,
      "step": 58500
    },
    {
      "epoch": 1.47,
      "grad_norm": 5.74609375,
      "learning_rate": 1.2482412060301508e-05,
      "loss": 1.3638,
      "step": 58600
    },
    {
      "epoch": 1.47,
      "grad_norm": 6.1328125,
      "learning_rate": 1.2452261306532663e-05,
      "loss": 1.2987,
      "step": 58700
    },
    {
      "epoch": 1.47,
      "grad_norm": 4.96484375,
      "learning_rate": 1.242211055276382e-05,
      "loss": 1.3281,
      "step": 58800
    },
    {
      "epoch": 1.47,
      "grad_norm": 4.68359375,
      "learning_rate": 1.2391959798994975e-05,
      "loss": 1.2599,
      "step": 58900
    },
    {
      "epoch": 1.48,
      "grad_norm": 7.87890625,
      "learning_rate": 1.2361809045226132e-05,
      "loss": 1.334,
      "step": 59000
    },
    {
      "epoch": 1.48,
      "grad_norm": 5.8125,
      "learning_rate": 1.2331658291457286e-05,
      "loss": 1.32,
      "step": 59100
    },
    {
      "epoch": 1.48,
      "grad_norm": 6.26953125,
      "learning_rate": 1.2301507537688443e-05,
      "loss": 1.3368,
      "step": 59200
    },
    {
      "epoch": 1.48,
      "grad_norm": 4.5703125,
      "learning_rate": 1.2271356783919598e-05,
      "loss": 1.2395,
      "step": 59300
    },
    {
      "epoch": 1.48,
      "grad_norm": 7.453125,
      "learning_rate": 1.2241206030150755e-05,
      "loss": 1.3287,
      "step": 59400
    },
    {
      "epoch": 1.49,
      "grad_norm": 6.453125,
      "learning_rate": 1.221105527638191e-05,
      "loss": 1.3069,
      "step": 59500
    },
    {
      "epoch": 1.49,
      "grad_norm": 5.75,
      "learning_rate": 1.2180904522613065e-05,
      "loss": 1.2809,
      "step": 59600
    },
    {
      "epoch": 1.49,
      "grad_norm": 6.00390625,
      "learning_rate": 1.215075376884422e-05,
      "loss": 1.3378,
      "step": 59700
    },
    {
      "epoch": 1.5,
      "grad_norm": 7.78125,
      "learning_rate": 1.2120603015075378e-05,
      "loss": 1.3107,
      "step": 59800
    },
    {
      "epoch": 1.5,
      "grad_norm": 5.94140625,
      "learning_rate": 1.2090452261306533e-05,
      "loss": 1.2776,
      "step": 59900
    },
    {
      "epoch": 1.5,
      "grad_norm": 5.78125,
      "learning_rate": 1.206030150753769e-05,
      "loss": 1.3065,
      "step": 60000
    },
    {
      "epoch": 1.5,
      "eval_loss": 1.3107746839523315,
      "eval_runtime": 58.384,
      "eval_samples_per_second": 17.128,
      "eval_steps_per_second": 4.282,
      "step": 60000
    },
    {
      "epoch": 1.5,
      "grad_norm": 8.734375,
      "learning_rate": 1.2030150753768845e-05,
      "loss": 1.3049,
      "step": 60100
    },
    {
      "epoch": 1.5,
      "grad_norm": 5.95703125,
      "learning_rate": 1.2e-05,
      "loss": 1.319,
      "step": 60200
    },
    {
      "epoch": 1.51,
      "grad_norm": 5.3515625,
      "learning_rate": 1.1969849246231156e-05,
      "loss": 1.2789,
      "step": 60300
    },
    {
      "epoch": 1.51,
      "grad_norm": 14.1171875,
      "learning_rate": 1.1939698492462313e-05,
      "loss": 1.3548,
      "step": 60400
    },
    {
      "epoch": 1.51,
      "grad_norm": 5.328125,
      "learning_rate": 1.1909547738693468e-05,
      "loss": 1.3843,
      "step": 60500
    },
    {
      "epoch": 1.52,
      "grad_norm": 9.5,
      "learning_rate": 1.1879396984924625e-05,
      "loss": 1.2877,
      "step": 60600
    },
    {
      "epoch": 1.52,
      "grad_norm": 3.416015625,
      "learning_rate": 1.184924623115578e-05,
      "loss": 1.3448,
      "step": 60700
    },
    {
      "epoch": 1.52,
      "grad_norm": 7.03515625,
      "learning_rate": 1.1819095477386934e-05,
      "loss": 1.3308,
      "step": 60800
    },
    {
      "epoch": 1.52,
      "grad_norm": 4.8671875,
      "learning_rate": 1.178894472361809e-05,
      "loss": 1.2599,
      "step": 60900
    },
    {
      "epoch": 1.52,
      "grad_norm": 6.1171875,
      "learning_rate": 1.1758793969849246e-05,
      "loss": 1.2777,
      "step": 61000
    },
    {
      "epoch": 1.53,
      "grad_norm": 5.17578125,
      "learning_rate": 1.1728643216080403e-05,
      "loss": 1.3046,
      "step": 61100
    },
    {
      "epoch": 1.53,
      "grad_norm": 6.5859375,
      "learning_rate": 1.1698492462311558e-05,
      "loss": 1.3241,
      "step": 61200
    },
    {
      "epoch": 1.53,
      "grad_norm": 7.91796875,
      "learning_rate": 1.1668341708542713e-05,
      "loss": 1.3214,
      "step": 61300
    },
    {
      "epoch": 1.54,
      "grad_norm": 4.66796875,
      "learning_rate": 1.1638190954773869e-05,
      "loss": 1.3132,
      "step": 61400
    },
    {
      "epoch": 1.54,
      "grad_norm": 6.0546875,
      "learning_rate": 1.1608040201005026e-05,
      "loss": 1.3265,
      "step": 61500
    },
    {
      "epoch": 1.54,
      "grad_norm": 5.17578125,
      "learning_rate": 1.157788944723618e-05,
      "loss": 1.261,
      "step": 61600
    },
    {
      "epoch": 1.54,
      "grad_norm": 4.625,
      "learning_rate": 1.1547738693467338e-05,
      "loss": 1.3262,
      "step": 61700
    },
    {
      "epoch": 1.54,
      "grad_norm": 4.87890625,
      "learning_rate": 1.1517587939698493e-05,
      "loss": 1.2664,
      "step": 61800
    },
    {
      "epoch": 1.55,
      "grad_norm": 4.79296875,
      "learning_rate": 1.1487437185929648e-05,
      "loss": 1.3297,
      "step": 61900
    },
    {
      "epoch": 1.55,
      "grad_norm": 8.34375,
      "learning_rate": 1.1457286432160804e-05,
      "loss": 1.28,
      "step": 62000
    },
    {
      "epoch": 1.55,
      "grad_norm": 8.3203125,
      "learning_rate": 1.142713567839196e-05,
      "loss": 1.266,
      "step": 62100
    },
    {
      "epoch": 1.56,
      "grad_norm": 7.94921875,
      "learning_rate": 1.1396984924623116e-05,
      "loss": 1.2977,
      "step": 62200
    },
    {
      "epoch": 1.56,
      "grad_norm": 10.5234375,
      "learning_rate": 1.1366834170854273e-05,
      "loss": 1.3281,
      "step": 62300
    },
    {
      "epoch": 1.56,
      "grad_norm": 6.87890625,
      "learning_rate": 1.1336683417085428e-05,
      "loss": 1.2998,
      "step": 62400
    },
    {
      "epoch": 1.56,
      "grad_norm": 5.69921875,
      "learning_rate": 1.1306532663316583e-05,
      "loss": 1.2484,
      "step": 62500
    },
    {
      "epoch": 1.56,
      "eval_loss": 1.3006246089935303,
      "eval_runtime": 58.2951,
      "eval_samples_per_second": 17.154,
      "eval_steps_per_second": 4.289,
      "step": 62500
    },
    {
      "epoch": 1.56,
      "grad_norm": 6.99609375,
      "learning_rate": 1.1276381909547739e-05,
      "loss": 1.2923,
      "step": 62600
    },
    {
      "epoch": 1.57,
      "grad_norm": 7.94140625,
      "learning_rate": 1.1246231155778896e-05,
      "loss": 1.2975,
      "step": 62700
    },
    {
      "epoch": 1.57,
      "grad_norm": 6.96484375,
      "learning_rate": 1.121608040201005e-05,
      "loss": 1.2703,
      "step": 62800
    },
    {
      "epoch": 1.57,
      "grad_norm": 6.7578125,
      "learning_rate": 1.1185929648241208e-05,
      "loss": 1.3221,
      "step": 62900
    },
    {
      "epoch": 1.57,
      "grad_norm": 6.00390625,
      "learning_rate": 1.1155778894472361e-05,
      "loss": 1.3186,
      "step": 63000
    },
    {
      "epoch": 1.58,
      "grad_norm": 4.48828125,
      "learning_rate": 1.1125628140703517e-05,
      "loss": 1.2878,
      "step": 63100
    },
    {
      "epoch": 1.58,
      "grad_norm": 5.16015625,
      "learning_rate": 1.1095477386934674e-05,
      "loss": 1.302,
      "step": 63200
    },
    {
      "epoch": 1.58,
      "grad_norm": 10.0546875,
      "learning_rate": 1.1065326633165829e-05,
      "loss": 1.2607,
      "step": 63300
    },
    {
      "epoch": 1.58,
      "grad_norm": 4.77734375,
      "learning_rate": 1.1035175879396986e-05,
      "loss": 1.3128,
      "step": 63400
    },
    {
      "epoch": 1.59,
      "grad_norm": 4.2578125,
      "learning_rate": 1.1005025125628141e-05,
      "loss": 1.2557,
      "step": 63500
    },
    {
      "epoch": 1.59,
      "grad_norm": 11.1015625,
      "learning_rate": 1.0974874371859296e-05,
      "loss": 1.3114,
      "step": 63600
    },
    {
      "epoch": 1.59,
      "grad_norm": 2.9296875,
      "learning_rate": 1.0944723618090452e-05,
      "loss": 1.2991,
      "step": 63700
    },
    {
      "epoch": 1.59,
      "grad_norm": 7.55859375,
      "learning_rate": 1.0914572864321609e-05,
      "loss": 1.3479,
      "step": 63800
    },
    {
      "epoch": 1.6,
      "grad_norm": 4.71875,
      "learning_rate": 1.0884422110552764e-05,
      "loss": 1.2951,
      "step": 63900
    },
    {
      "epoch": 1.6,
      "grad_norm": 5.328125,
      "learning_rate": 1.085427135678392e-05,
      "loss": 1.3348,
      "step": 64000
    },
    {
      "epoch": 1.6,
      "grad_norm": 6.984375,
      "learning_rate": 1.0824120603015076e-05,
      "loss": 1.2897,
      "step": 64100
    },
    {
      "epoch": 1.6,
      "grad_norm": 3.748046875,
      "learning_rate": 1.0793969849246231e-05,
      "loss": 1.3096,
      "step": 64200
    },
    {
      "epoch": 1.61,
      "grad_norm": 7.83984375,
      "learning_rate": 1.0763819095477387e-05,
      "loss": 1.2854,
      "step": 64300
    },
    {
      "epoch": 1.61,
      "grad_norm": 4.6640625,
      "learning_rate": 1.0733668341708544e-05,
      "loss": 1.2507,
      "step": 64400
    },
    {
      "epoch": 1.61,
      "grad_norm": 7.77734375,
      "learning_rate": 1.0703517587939699e-05,
      "loss": 1.271,
      "step": 64500
    },
    {
      "epoch": 1.61,
      "grad_norm": 6.4765625,
      "learning_rate": 1.0673366834170856e-05,
      "loss": 1.2176,
      "step": 64600
    },
    {
      "epoch": 1.62,
      "grad_norm": 6.984375,
      "learning_rate": 1.064321608040201e-05,
      "loss": 1.301,
      "step": 64700
    },
    {
      "epoch": 1.62,
      "grad_norm": 5.8203125,
      "learning_rate": 1.0613065326633166e-05,
      "loss": 1.3127,
      "step": 64800
    },
    {
      "epoch": 1.62,
      "grad_norm": 6.3671875,
      "learning_rate": 1.0582914572864322e-05,
      "loss": 1.2781,
      "step": 64900
    },
    {
      "epoch": 1.62,
      "grad_norm": 7.41015625,
      "learning_rate": 1.0552763819095479e-05,
      "loss": 1.3218,
      "step": 65000
    },
    {
      "epoch": 1.62,
      "eval_loss": 1.321094274520874,
      "eval_runtime": 58.4017,
      "eval_samples_per_second": 17.123,
      "eval_steps_per_second": 4.281,
      "step": 65000
    },
    {
      "epoch": 1.63,
      "grad_norm": 7.20703125,
      "learning_rate": 1.0522613065326634e-05,
      "loss": 1.2238,
      "step": 65100
    },
    {
      "epoch": 1.63,
      "grad_norm": 7.1640625,
      "learning_rate": 1.049246231155779e-05,
      "loss": 1.2775,
      "step": 65200
    },
    {
      "epoch": 1.63,
      "grad_norm": 4.03125,
      "learning_rate": 1.0462311557788944e-05,
      "loss": 1.3113,
      "step": 65300
    },
    {
      "epoch": 1.64,
      "grad_norm": 10.484375,
      "learning_rate": 1.04321608040201e-05,
      "loss": 1.2948,
      "step": 65400
    },
    {
      "epoch": 1.64,
      "grad_norm": 5.609375,
      "learning_rate": 1.0402010050251257e-05,
      "loss": 1.3212,
      "step": 65500
    },
    {
      "epoch": 1.64,
      "grad_norm": 3.39453125,
      "learning_rate": 1.0371859296482412e-05,
      "loss": 1.2958,
      "step": 65600
    },
    {
      "epoch": 1.64,
      "grad_norm": 5.3828125,
      "learning_rate": 1.0341708542713569e-05,
      "loss": 1.2665,
      "step": 65700
    },
    {
      "epoch": 1.65,
      "grad_norm": 6.5390625,
      "learning_rate": 1.0311557788944724e-05,
      "loss": 1.2708,
      "step": 65800
    },
    {
      "epoch": 1.65,
      "grad_norm": 9.9296875,
      "learning_rate": 1.028140703517588e-05,
      "loss": 1.3221,
      "step": 65900
    },
    {
      "epoch": 1.65,
      "grad_norm": 6.21484375,
      "learning_rate": 1.0251256281407035e-05,
      "loss": 1.2757,
      "step": 66000
    },
    {
      "epoch": 1.65,
      "grad_norm": 5.140625,
      "learning_rate": 1.0221105527638192e-05,
      "loss": 1.2619,
      "step": 66100
    },
    {
      "epoch": 1.66,
      "grad_norm": 4.375,
      "learning_rate": 1.0190954773869347e-05,
      "loss": 1.3567,
      "step": 66200
    },
    {
      "epoch": 1.66,
      "grad_norm": 5.93359375,
      "learning_rate": 1.0160804020100504e-05,
      "loss": 1.2923,
      "step": 66300
    },
    {
      "epoch": 1.66,
      "grad_norm": 6.234375,
      "learning_rate": 1.0130653266331657e-05,
      "loss": 1.2919,
      "step": 66400
    },
    {
      "epoch": 1.66,
      "grad_norm": 6.51171875,
      "learning_rate": 1.0100502512562814e-05,
      "loss": 1.2798,
      "step": 66500
    },
    {
      "epoch": 1.67,
      "grad_norm": 7.75,
      "learning_rate": 1.007035175879397e-05,
      "loss": 1.2884,
      "step": 66600
    },
    {
      "epoch": 1.67,
      "grad_norm": 5.21875,
      "learning_rate": 1.0040201005025127e-05,
      "loss": 1.2824,
      "step": 66700
    },
    {
      "epoch": 1.67,
      "grad_norm": 5.703125,
      "learning_rate": 1.0010050251256282e-05,
      "loss": 1.272,
      "step": 66800
    },
    {
      "epoch": 1.67,
      "grad_norm": 6.50390625,
      "learning_rate": 9.979899497487439e-06,
      "loss": 1.3051,
      "step": 66900
    },
    {
      "epoch": 1.68,
      "grad_norm": 5.8046875,
      "learning_rate": 9.949748743718592e-06,
      "loss": 1.2931,
      "step": 67000
    },
    {
      "epoch": 1.68,
      "grad_norm": 6.51171875,
      "learning_rate": 9.91959798994975e-06,
      "loss": 1.2462,
      "step": 67100
    },
    {
      "epoch": 1.68,
      "grad_norm": 5.41015625,
      "learning_rate": 9.889447236180905e-06,
      "loss": 1.262,
      "step": 67200
    },
    {
      "epoch": 1.68,
      "grad_norm": 5.578125,
      "learning_rate": 9.859296482412062e-06,
      "loss": 1.2758,
      "step": 67300
    },
    {
      "epoch": 1.69,
      "grad_norm": 7.79296875,
      "learning_rate": 9.829145728643217e-06,
      "loss": 1.2548,
      "step": 67400
    },
    {
      "epoch": 1.69,
      "grad_norm": 7.046875,
      "learning_rate": 9.798994974874372e-06,
      "loss": 1.3057,
      "step": 67500
    },
    {
      "epoch": 1.69,
      "eval_loss": 1.3317267894744873,
      "eval_runtime": 58.3095,
      "eval_samples_per_second": 17.15,
      "eval_steps_per_second": 4.287,
      "step": 67500
    },
    {
      "epoch": 1.69,
      "grad_norm": 9.78125,
      "learning_rate": 9.768844221105527e-06,
      "loss": 1.2899,
      "step": 67600
    },
    {
      "epoch": 1.69,
      "grad_norm": 5.09765625,
      "learning_rate": 9.738693467336683e-06,
      "loss": 1.2609,
      "step": 67700
    },
    {
      "epoch": 1.69,
      "grad_norm": 8.0859375,
      "learning_rate": 9.70854271356784e-06,
      "loss": 1.3006,
      "step": 67800
    },
    {
      "epoch": 1.7,
      "grad_norm": 5.0234375,
      "learning_rate": 9.678391959798995e-06,
      "loss": 1.275,
      "step": 67900
    },
    {
      "epoch": 1.7,
      "grad_norm": 5.44921875,
      "learning_rate": 9.648241206030152e-06,
      "loss": 1.2872,
      "step": 68000
    },
    {
      "epoch": 1.7,
      "grad_norm": 5.50390625,
      "learning_rate": 9.618090452261305e-06,
      "loss": 1.316,
      "step": 68100
    },
    {
      "epoch": 1.71,
      "grad_norm": 3.345703125,
      "learning_rate": 9.587939698492462e-06,
      "loss": 1.295,
      "step": 68200
    },
    {
      "epoch": 1.71,
      "grad_norm": 9.9140625,
      "learning_rate": 9.557788944723618e-06,
      "loss": 1.2607,
      "step": 68300
    },
    {
      "epoch": 1.71,
      "grad_norm": 6.43359375,
      "learning_rate": 9.527638190954775e-06,
      "loss": 1.2936,
      "step": 68400
    },
    {
      "epoch": 1.71,
      "grad_norm": 7.26953125,
      "learning_rate": 9.49748743718593e-06,
      "loss": 1.299,
      "step": 68500
    },
    {
      "epoch": 1.71,
      "grad_norm": 5.04296875,
      "learning_rate": 9.467336683417087e-06,
      "loss": 1.2669,
      "step": 68600
    },
    {
      "epoch": 1.72,
      "grad_norm": 4.80859375,
      "learning_rate": 9.43718592964824e-06,
      "loss": 1.2794,
      "step": 68700
    },
    {
      "epoch": 1.72,
      "grad_norm": 9.578125,
      "learning_rate": 9.407035175879397e-06,
      "loss": 1.2817,
      "step": 68800
    },
    {
      "epoch": 1.72,
      "grad_norm": 11.0546875,
      "learning_rate": 9.376884422110553e-06,
      "loss": 1.215,
      "step": 68900
    },
    {
      "epoch": 1.73,
      "grad_norm": 6.125,
      "learning_rate": 9.34673366834171e-06,
      "loss": 1.1854,
      "step": 69000
    },
    {
      "epoch": 1.73,
      "grad_norm": 6.05859375,
      "learning_rate": 9.316582914572865e-06,
      "loss": 1.2695,
      "step": 69100
    },
    {
      "epoch": 1.73,
      "grad_norm": 9.546875,
      "learning_rate": 9.28643216080402e-06,
      "loss": 1.2513,
      "step": 69200
    },
    {
      "epoch": 1.73,
      "grad_norm": 6.578125,
      "learning_rate": 9.256281407035175e-06,
      "loss": 1.2677,
      "step": 69300
    },
    {
      "epoch": 1.73,
      "grad_norm": 7.69921875,
      "learning_rate": 9.226130653266332e-06,
      "loss": 1.3259,
      "step": 69400
    },
    {
      "epoch": 1.74,
      "grad_norm": 4.6171875,
      "learning_rate": 9.195979899497488e-06,
      "loss": 1.2217,
      "step": 69500
    },
    {
      "epoch": 1.74,
      "grad_norm": 6.1484375,
      "learning_rate": 9.165829145728645e-06,
      "loss": 1.254,
      "step": 69600
    },
    {
      "epoch": 1.74,
      "grad_norm": 6.375,
      "learning_rate": 9.1356783919598e-06,
      "loss": 1.2685,
      "step": 69700
    },
    {
      "epoch": 1.75,
      "grad_norm": 5.44921875,
      "learning_rate": 9.105527638190955e-06,
      "loss": 1.1982,
      "step": 69800
    },
    {
      "epoch": 1.75,
      "grad_norm": 5.57421875,
      "learning_rate": 9.07537688442211e-06,
      "loss": 1.2908,
      "step": 69900
    },
    {
      "epoch": 1.75,
      "grad_norm": 6.02734375,
      "learning_rate": 9.045226130653266e-06,
      "loss": 1.2438,
      "step": 70000
    },
    {
      "epoch": 1.75,
      "eval_loss": 1.229135513305664,
      "eval_runtime": 58.2748,
      "eval_samples_per_second": 17.16,
      "eval_steps_per_second": 4.29,
      "step": 70000
    },
    {
      "epoch": 1.75,
      "grad_norm": 6.41796875,
      "learning_rate": 9.015075376884423e-06,
      "loss": 1.2803,
      "step": 70100
    },
    {
      "epoch": 1.75,
      "grad_norm": 7.65625,
      "learning_rate": 8.984924623115578e-06,
      "loss": 1.2674,
      "step": 70200
    },
    {
      "epoch": 1.76,
      "grad_norm": 11.1796875,
      "learning_rate": 8.954773869346735e-06,
      "loss": 1.2702,
      "step": 70300
    },
    {
      "epoch": 1.76,
      "grad_norm": 11.6015625,
      "learning_rate": 8.924623115577888e-06,
      "loss": 1.3009,
      "step": 70400
    },
    {
      "epoch": 1.76,
      "grad_norm": 6.42578125,
      "learning_rate": 8.894472361809045e-06,
      "loss": 1.2585,
      "step": 70500
    },
    {
      "epoch": 1.77,
      "grad_norm": 8.9921875,
      "learning_rate": 8.8643216080402e-06,
      "loss": 1.2714,
      "step": 70600
    },
    {
      "epoch": 1.77,
      "grad_norm": 6.359375,
      "learning_rate": 8.834170854271358e-06,
      "loss": 1.2657,
      "step": 70700
    },
    {
      "epoch": 1.77,
      "grad_norm": 4.984375,
      "learning_rate": 8.804020100502513e-06,
      "loss": 1.2992,
      "step": 70800
    },
    {
      "epoch": 1.77,
      "grad_norm": 8.53125,
      "learning_rate": 8.773869346733668e-06,
      "loss": 1.2252,
      "step": 70900
    },
    {
      "epoch": 1.77,
      "grad_norm": 5.296875,
      "learning_rate": 8.743718592964823e-06,
      "loss": 1.2618,
      "step": 71000
    },
    {
      "epoch": 1.78,
      "grad_norm": 8.515625,
      "learning_rate": 8.71356783919598e-06,
      "loss": 1.2973,
      "step": 71100
    },
    {
      "epoch": 1.78,
      "grad_norm": 9.8828125,
      "learning_rate": 8.683417085427136e-06,
      "loss": 1.2785,
      "step": 71200
    },
    {
      "epoch": 1.78,
      "grad_norm": 7.08203125,
      "learning_rate": 8.653266331658293e-06,
      "loss": 1.2496,
      "step": 71300
    },
    {
      "epoch": 1.79,
      "grad_norm": 7.35546875,
      "learning_rate": 8.623115577889448e-06,
      "loss": 1.2757,
      "step": 71400
    },
    {
      "epoch": 1.79,
      "grad_norm": 7.109375,
      "learning_rate": 8.592964824120603e-06,
      "loss": 1.283,
      "step": 71500
    },
    {
      "epoch": 1.79,
      "grad_norm": 8.9765625,
      "learning_rate": 8.562814070351758e-06,
      "loss": 1.2705,
      "step": 71600
    },
    {
      "epoch": 1.79,
      "grad_norm": 5.4921875,
      "learning_rate": 8.532663316582915e-06,
      "loss": 1.2593,
      "step": 71700
    },
    {
      "epoch": 1.79,
      "grad_norm": 6.984375,
      "learning_rate": 8.50251256281407e-06,
      "loss": 1.2839,
      "step": 71800
    },
    {
      "epoch": 1.8,
      "grad_norm": 4.89453125,
      "learning_rate": 8.472361809045228e-06,
      "loss": 1.2628,
      "step": 71900
    },
    {
      "epoch": 1.8,
      "grad_norm": 6.7265625,
      "learning_rate": 8.442211055276383e-06,
      "loss": 1.3171,
      "step": 72000
    },
    {
      "epoch": 1.8,
      "grad_norm": 6.06640625,
      "learning_rate": 8.412060301507538e-06,
      "loss": 1.2768,
      "step": 72100
    },
    {
      "epoch": 1.81,
      "grad_norm": 8.0546875,
      "learning_rate": 8.381909547738693e-06,
      "loss": 1.3066,
      "step": 72200
    },
    {
      "epoch": 1.81,
      "grad_norm": 5.91796875,
      "learning_rate": 8.351758793969849e-06,
      "loss": 1.2368,
      "step": 72300
    },
    {
      "epoch": 1.81,
      "grad_norm": 8.5625,
      "learning_rate": 8.321608040201006e-06,
      "loss": 1.2342,
      "step": 72400
    },
    {
      "epoch": 1.81,
      "grad_norm": 7.94921875,
      "learning_rate": 8.291457286432161e-06,
      "loss": 1.2615,
      "step": 72500
    },
    {
      "epoch": 1.81,
      "eval_loss": 1.2854846715927124,
      "eval_runtime": 58.3319,
      "eval_samples_per_second": 17.143,
      "eval_steps_per_second": 4.286,
      "step": 72500
    },
    {
      "epoch": 1.81,
      "grad_norm": 13.90625,
      "learning_rate": 8.261306532663316e-06,
      "loss": 1.2859,
      "step": 72600
    },
    {
      "epoch": 1.82,
      "grad_norm": 7.37890625,
      "learning_rate": 8.231155778894471e-06,
      "loss": 1.2373,
      "step": 72700
    },
    {
      "epoch": 1.82,
      "grad_norm": 6.78515625,
      "learning_rate": 8.201005025125628e-06,
      "loss": 1.2577,
      "step": 72800
    },
    {
      "epoch": 1.82,
      "grad_norm": 8.546875,
      "learning_rate": 8.170854271356784e-06,
      "loss": 1.2793,
      "step": 72900
    },
    {
      "epoch": 1.82,
      "grad_norm": 6.0625,
      "learning_rate": 8.14070351758794e-06,
      "loss": 1.3146,
      "step": 73000
    },
    {
      "epoch": 1.83,
      "grad_norm": 4.19140625,
      "learning_rate": 8.110552763819096e-06,
      "loss": 1.2822,
      "step": 73100
    },
    {
      "epoch": 1.83,
      "grad_norm": 5.15234375,
      "learning_rate": 8.080402010050251e-06,
      "loss": 1.2345,
      "step": 73200
    },
    {
      "epoch": 1.83,
      "grad_norm": 7.6640625,
      "learning_rate": 8.050251256281406e-06,
      "loss": 1.2923,
      "step": 73300
    },
    {
      "epoch": 1.83,
      "grad_norm": 8.1953125,
      "learning_rate": 8.020100502512563e-06,
      "loss": 1.2266,
      "step": 73400
    },
    {
      "epoch": 1.84,
      "grad_norm": 5.65234375,
      "learning_rate": 7.989949748743719e-06,
      "loss": 1.27,
      "step": 73500
    },
    {
      "epoch": 1.84,
      "grad_norm": 6.40625,
      "learning_rate": 7.959798994974876e-06,
      "loss": 1.2393,
      "step": 73600
    },
    {
      "epoch": 1.84,
      "grad_norm": 21.375,
      "learning_rate": 7.929648241206031e-06,
      "loss": 1.2509,
      "step": 73700
    },
    {
      "epoch": 1.84,
      "grad_norm": 6.984375,
      "learning_rate": 7.899497487437186e-06,
      "loss": 1.263,
      "step": 73800
    },
    {
      "epoch": 1.85,
      "grad_norm": 7.5234375,
      "learning_rate": 7.869346733668341e-06,
      "loss": 1.2373,
      "step": 73900
    },
    {
      "epoch": 1.85,
      "grad_norm": 5.546875,
      "learning_rate": 7.839195979899498e-06,
      "loss": 1.3024,
      "step": 74000
    },
    {
      "epoch": 1.85,
      "grad_norm": 6.68359375,
      "learning_rate": 7.809045226130654e-06,
      "loss": 1.2432,
      "step": 74100
    },
    {
      "epoch": 1.85,
      "grad_norm": 9.8046875,
      "learning_rate": 7.77889447236181e-06,
      "loss": 1.2423,
      "step": 74200
    },
    {
      "epoch": 1.86,
      "grad_norm": 7.74609375,
      "learning_rate": 7.748743718592964e-06,
      "loss": 1.2678,
      "step": 74300
    },
    {
      "epoch": 1.86,
      "grad_norm": 8.390625,
      "learning_rate": 7.718592964824121e-06,
      "loss": 1.2765,
      "step": 74400
    },
    {
      "epoch": 1.86,
      "grad_norm": 7.6328125,
      "learning_rate": 7.688442211055276e-06,
      "loss": 1.2353,
      "step": 74500
    },
    {
      "epoch": 1.86,
      "grad_norm": 5.83984375,
      "learning_rate": 7.658291457286432e-06,
      "loss": 1.2307,
      "step": 74600
    },
    {
      "epoch": 1.87,
      "grad_norm": 7.83203125,
      "learning_rate": 7.628140703517589e-06,
      "loss": 1.2646,
      "step": 74700
    },
    {
      "epoch": 1.87,
      "grad_norm": 7.10546875,
      "learning_rate": 7.597989949748743e-06,
      "loss": 1.249,
      "step": 74800
    },
    {
      "epoch": 1.87,
      "grad_norm": 7.12890625,
      "learning_rate": 7.5678391959799e-06,
      "loss": 1.1872,
      "step": 74900
    },
    {
      "epoch": 1.88,
      "grad_norm": 8.1640625,
      "learning_rate": 7.537688442211054e-06,
      "loss": 1.2345,
      "step": 75000
    },
    {
      "epoch": 1.88,
      "eval_loss": 1.2996472120285034,
      "eval_runtime": 58.363,
      "eval_samples_per_second": 17.134,
      "eval_steps_per_second": 4.284,
      "step": 75000
    },
    {
      "epoch": 1.88,
      "grad_norm": 4.2578125,
      "learning_rate": 7.507537688442211e-06,
      "loss": 1.2281,
      "step": 75100
    },
    {
      "epoch": 1.88,
      "grad_norm": 6.125,
      "learning_rate": 7.4773869346733675e-06,
      "loss": 1.2433,
      "step": 75200
    },
    {
      "epoch": 1.88,
      "grad_norm": 5.61328125,
      "learning_rate": 7.447236180904523e-06,
      "loss": 1.2314,
      "step": 75300
    },
    {
      "epoch": 1.89,
      "grad_norm": 8.28125,
      "learning_rate": 7.417085427135679e-06,
      "loss": 1.2677,
      "step": 75400
    },
    {
      "epoch": 1.89,
      "grad_norm": 7.12109375,
      "learning_rate": 7.386934673366835e-06,
      "loss": 1.2832,
      "step": 75500
    },
    {
      "epoch": 1.89,
      "grad_norm": 7.1796875,
      "learning_rate": 7.35678391959799e-06,
      "loss": 1.2156,
      "step": 75600
    },
    {
      "epoch": 1.89,
      "grad_norm": 6.73828125,
      "learning_rate": 7.326633165829146e-06,
      "loss": 1.248,
      "step": 75700
    },
    {
      "epoch": 1.9,
      "grad_norm": 12.0703125,
      "learning_rate": 7.296482412060302e-06,
      "loss": 1.2283,
      "step": 75800
    },
    {
      "epoch": 1.9,
      "grad_norm": 6.546875,
      "learning_rate": 7.266331658291457e-06,
      "loss": 1.2495,
      "step": 75900
    },
    {
      "epoch": 1.9,
      "grad_norm": 7.31640625,
      "learning_rate": 7.236180904522613e-06,
      "loss": 1.1779,
      "step": 76000
    },
    {
      "epoch": 1.9,
      "grad_norm": 5.79296875,
      "learning_rate": 7.206030150753769e-06,
      "loss": 1.2431,
      "step": 76100
    },
    {
      "epoch": 1.91,
      "grad_norm": 6.67578125,
      "learning_rate": 7.175879396984924e-06,
      "loss": 1.1931,
      "step": 76200
    },
    {
      "epoch": 1.91,
      "grad_norm": 8.65625,
      "learning_rate": 7.1457286432160805e-06,
      "loss": 1.2692,
      "step": 76300
    },
    {
      "epoch": 1.91,
      "grad_norm": 7.60546875,
      "learning_rate": 7.115577889447236e-06,
      "loss": 1.2612,
      "step": 76400
    },
    {
      "epoch": 1.91,
      "grad_norm": 6.47265625,
      "learning_rate": 7.085427135678392e-06,
      "loss": 1.247,
      "step": 76500
    },
    {
      "epoch": 1.92,
      "grad_norm": 8.421875,
      "learning_rate": 7.055276381909548e-06,
      "loss": 1.2481,
      "step": 76600
    },
    {
      "epoch": 1.92,
      "grad_norm": 7.8125,
      "learning_rate": 7.025125628140703e-06,
      "loss": 1.1965,
      "step": 76700
    },
    {
      "epoch": 1.92,
      "grad_norm": 9.15625,
      "learning_rate": 6.994974874371859e-06,
      "loss": 1.2125,
      "step": 76800
    },
    {
      "epoch": 1.92,
      "grad_norm": 9.0,
      "learning_rate": 6.9648241206030155e-06,
      "loss": 1.228,
      "step": 76900
    },
    {
      "epoch": 1.93,
      "grad_norm": 4.01171875,
      "learning_rate": 6.934673366834171e-06,
      "loss": 1.2254,
      "step": 77000
    },
    {
      "epoch": 1.93,
      "grad_norm": 4.84375,
      "learning_rate": 6.904522613065327e-06,
      "loss": 1.2786,
      "step": 77100
    },
    {
      "epoch": 1.93,
      "grad_norm": 7.03125,
      "learning_rate": 6.874371859296483e-06,
      "loss": 1.2677,
      "step": 77200
    },
    {
      "epoch": 1.93,
      "grad_norm": 8.40625,
      "learning_rate": 6.844221105527638e-06,
      "loss": 1.2667,
      "step": 77300
    },
    {
      "epoch": 1.94,
      "grad_norm": 7.8359375,
      "learning_rate": 6.814070351758794e-06,
      "loss": 1.2749,
      "step": 77400
    },
    {
      "epoch": 1.94,
      "grad_norm": 8.234375,
      "learning_rate": 6.7839195979899505e-06,
      "loss": 1.252,
      "step": 77500
    },
    {
      "epoch": 1.94,
      "eval_loss": 1.2479995489120483,
      "eval_runtime": 58.4081,
      "eval_samples_per_second": 17.121,
      "eval_steps_per_second": 4.28,
      "step": 77500
    },
    {
      "epoch": 1.94,
      "grad_norm": 6.81640625,
      "learning_rate": 6.753768844221106e-06,
      "loss": 1.1964,
      "step": 77600
    },
    {
      "epoch": 1.94,
      "grad_norm": 5.234375,
      "learning_rate": 6.723618090452262e-06,
      "loss": 1.2277,
      "step": 77700
    },
    {
      "epoch": 1.94,
      "grad_norm": 5.734375,
      "learning_rate": 6.693467336683418e-06,
      "loss": 1.1941,
      "step": 77800
    },
    {
      "epoch": 1.95,
      "grad_norm": 7.47265625,
      "learning_rate": 6.663316582914573e-06,
      "loss": 1.2202,
      "step": 77900
    },
    {
      "epoch": 1.95,
      "grad_norm": 7.37109375,
      "learning_rate": 6.633165829145729e-06,
      "loss": 1.1958,
      "step": 78000
    },
    {
      "epoch": 1.95,
      "grad_norm": 5.9609375,
      "learning_rate": 6.603015075376884e-06,
      "loss": 1.2519,
      "step": 78100
    },
    {
      "epoch": 1.96,
      "grad_norm": 7.921875,
      "learning_rate": 6.57286432160804e-06,
      "loss": 1.2012,
      "step": 78200
    },
    {
      "epoch": 1.96,
      "grad_norm": 6.5,
      "learning_rate": 6.542713567839196e-06,
      "loss": 1.2375,
      "step": 78300
    },
    {
      "epoch": 1.96,
      "grad_norm": 5.31640625,
      "learning_rate": 6.512562814070351e-06,
      "loss": 1.2287,
      "step": 78400
    },
    {
      "epoch": 1.96,
      "grad_norm": 8.6796875,
      "learning_rate": 6.482412060301507e-06,
      "loss": 1.2034,
      "step": 78500
    },
    {
      "epoch": 1.96,
      "grad_norm": 6.12890625,
      "learning_rate": 6.4522613065326635e-06,
      "loss": 1.2434,
      "step": 78600
    },
    {
      "epoch": 1.97,
      "grad_norm": 5.67578125,
      "learning_rate": 6.422110552763819e-06,
      "loss": 1.2209,
      "step": 78700
    },
    {
      "epoch": 1.97,
      "grad_norm": 5.1640625,
      "learning_rate": 6.391959798994975e-06,
      "loss": 1.2706,
      "step": 78800
    },
    {
      "epoch": 1.97,
      "grad_norm": 10.0703125,
      "learning_rate": 6.361809045226131e-06,
      "loss": 1.2471,
      "step": 78900
    },
    {
      "epoch": 1.98,
      "grad_norm": 5.3359375,
      "learning_rate": 6.331658291457286e-06,
      "loss": 1.2074,
      "step": 79000
    },
    {
      "epoch": 1.98,
      "grad_norm": 8.0234375,
      "learning_rate": 6.301507537688442e-06,
      "loss": 1.2271,
      "step": 79100
    },
    {
      "epoch": 1.98,
      "grad_norm": 12.015625,
      "learning_rate": 6.2713567839195985e-06,
      "loss": 1.2533,
      "step": 79200
    },
    {
      "epoch": 1.98,
      "grad_norm": 7.5078125,
      "learning_rate": 6.241206030150754e-06,
      "loss": 1.2112,
      "step": 79300
    },
    {
      "epoch": 1.98,
      "grad_norm": 6.16796875,
      "learning_rate": 6.21105527638191e-06,
      "loss": 1.2349,
      "step": 79400
    },
    {
      "epoch": 1.99,
      "grad_norm": 6.37109375,
      "learning_rate": 6.180904522613066e-06,
      "loss": 1.2972,
      "step": 79500
    },
    {
      "epoch": 1.99,
      "grad_norm": 5.58984375,
      "learning_rate": 6.150753768844221e-06,
      "loss": 1.2197,
      "step": 79600
    },
    {
      "epoch": 1.99,
      "grad_norm": 5.625,
      "learning_rate": 6.120603015075377e-06,
      "loss": 1.2149,
      "step": 79700
    },
    {
      "epoch": 2.0,
      "grad_norm": 7.828125,
      "learning_rate": 6.090452261306533e-06,
      "loss": 1.2589,
      "step": 79800
    },
    {
      "epoch": 2.0,
      "grad_norm": 8.46875,
      "learning_rate": 6.060301507537689e-06,
      "loss": 1.2533,
      "step": 79900
    },
    {
      "epoch": 2.0,
      "grad_norm": 6.96484375,
      "learning_rate": 6.030150753768845e-06,
      "loss": 1.148,
      "step": 80000
    },
    {
      "epoch": 2.0,
      "eval_loss": 1.2425081729888916,
      "eval_runtime": 58.3641,
      "eval_samples_per_second": 17.134,
      "eval_steps_per_second": 4.283,
      "step": 80000
    },
    {
      "epoch": 2.0,
      "grad_norm": 7.7265625,
      "learning_rate": 6e-06,
      "loss": 1.2136,
      "step": 80100
    },
    {
      "epoch": 2.0,
      "grad_norm": 14.2734375,
      "learning_rate": 5.969849246231156e-06,
      "loss": 1.1365,
      "step": 80200
    },
    {
      "epoch": 2.01,
      "grad_norm": 7.55078125,
      "learning_rate": 5.939698492462312e-06,
      "loss": 1.2502,
      "step": 80300
    },
    {
      "epoch": 2.01,
      "grad_norm": 9.328125,
      "learning_rate": 5.909547738693467e-06,
      "loss": 1.2155,
      "step": 80400
    },
    {
      "epoch": 2.01,
      "grad_norm": 6.2109375,
      "learning_rate": 5.879396984924623e-06,
      "loss": 1.1532,
      "step": 80500
    },
    {
      "epoch": 2.02,
      "grad_norm": 9.2421875,
      "learning_rate": 5.849246231155779e-06,
      "loss": 1.2078,
      "step": 80600
    },
    {
      "epoch": 2.02,
      "grad_norm": 7.6484375,
      "learning_rate": 5.819095477386934e-06,
      "loss": 1.2296,
      "step": 80700
    },
    {
      "epoch": 2.02,
      "grad_norm": 6.55078125,
      "learning_rate": 5.78894472361809e-06,
      "loss": 1.2095,
      "step": 80800
    },
    {
      "epoch": 2.02,
      "grad_norm": 8.6640625,
      "learning_rate": 5.7587939698492465e-06,
      "loss": 1.1459,
      "step": 80900
    },
    {
      "epoch": 2.02,
      "grad_norm": 4.33203125,
      "learning_rate": 5.728643216080402e-06,
      "loss": 1.1374,
      "step": 81000
    },
    {
      "epoch": 2.03,
      "grad_norm": 5.7109375,
      "learning_rate": 5.698492462311558e-06,
      "loss": 1.2141,
      "step": 81100
    },
    {
      "epoch": 2.03,
      "grad_norm": 7.6484375,
      "learning_rate": 5.668341708542714e-06,
      "loss": 1.1913,
      "step": 81200
    },
    {
      "epoch": 2.03,
      "grad_norm": 6.828125,
      "learning_rate": 5.638190954773869e-06,
      "loss": 1.2124,
      "step": 81300
    },
    {
      "epoch": 2.04,
      "grad_norm": 6.52734375,
      "learning_rate": 5.608040201005025e-06,
      "loss": 1.2935,
      "step": 81400
    },
    {
      "epoch": 2.04,
      "grad_norm": 8.8203125,
      "learning_rate": 5.577889447236181e-06,
      "loss": 1.217,
      "step": 81500
    },
    {
      "epoch": 2.04,
      "grad_norm": 7.5078125,
      "learning_rate": 5.547738693467337e-06,
      "loss": 1.2339,
      "step": 81600
    },
    {
      "epoch": 2.04,
      "grad_norm": 8.1171875,
      "learning_rate": 5.517587939698493e-06,
      "loss": 1.19,
      "step": 81700
    },
    {
      "epoch": 2.04,
      "grad_norm": 7.6015625,
      "learning_rate": 5.487437185929648e-06,
      "loss": 1.1577,
      "step": 81800
    },
    {
      "epoch": 2.05,
      "grad_norm": 7.97265625,
      "learning_rate": 5.457286432160804e-06,
      "loss": 1.1531,
      "step": 81900
    },
    {
      "epoch": 2.05,
      "grad_norm": 9.7265625,
      "learning_rate": 5.42713567839196e-06,
      "loss": 1.2609,
      "step": 82000
    },
    {
      "epoch": 2.05,
      "grad_norm": 10.296875,
      "learning_rate": 5.396984924623116e-06,
      "loss": 1.2294,
      "step": 82100
    },
    {
      "epoch": 2.06,
      "grad_norm": 14.6875,
      "learning_rate": 5.366834170854272e-06,
      "loss": 1.1506,
      "step": 82200
    },
    {
      "epoch": 2.06,
      "grad_norm": 10.59375,
      "learning_rate": 5.336683417085428e-06,
      "loss": 1.207,
      "step": 82300
    },
    {
      "epoch": 2.06,
      "grad_norm": 6.87109375,
      "learning_rate": 5.306532663316583e-06,
      "loss": 1.1686,
      "step": 82400
    },
    {
      "epoch": 2.06,
      "grad_norm": 10.375,
      "learning_rate": 5.276381909547739e-06,
      "loss": 1.2497,
      "step": 82500
    },
    {
      "epoch": 2.06,
      "eval_loss": 1.2355402708053589,
      "eval_runtime": 58.4521,
      "eval_samples_per_second": 17.108,
      "eval_steps_per_second": 4.277,
      "step": 82500
    },
    {
      "epoch": 2.06,
      "grad_norm": 6.8125,
      "learning_rate": 5.246231155778895e-06,
      "loss": 1.2069,
      "step": 82600
    },
    {
      "epoch": 2.07,
      "grad_norm": 7.6953125,
      "learning_rate": 5.21608040201005e-06,
      "loss": 1.2088,
      "step": 82700
    },
    {
      "epoch": 2.07,
      "grad_norm": 7.3671875,
      "learning_rate": 5.185929648241206e-06,
      "loss": 1.193,
      "step": 82800
    },
    {
      "epoch": 2.07,
      "grad_norm": 5.32421875,
      "learning_rate": 5.155778894472362e-06,
      "loss": 1.2365,
      "step": 82900
    },
    {
      "epoch": 2.08,
      "grad_norm": 5.71484375,
      "learning_rate": 5.125628140703517e-06,
      "loss": 1.1839,
      "step": 83000
    },
    {
      "epoch": 2.08,
      "grad_norm": 7.9140625,
      "learning_rate": 5.095477386934673e-06,
      "loss": 1.2174,
      "step": 83100
    },
    {
      "epoch": 2.08,
      "grad_norm": 8.7265625,
      "learning_rate": 5.065326633165829e-06,
      "loss": 1.1731,
      "step": 83200
    },
    {
      "epoch": 2.08,
      "grad_norm": 9.109375,
      "learning_rate": 5.035175879396985e-06,
      "loss": 1.1784,
      "step": 83300
    },
    {
      "epoch": 2.08,
      "grad_norm": 13.3125,
      "learning_rate": 5.005025125628141e-06,
      "loss": 1.1334,
      "step": 83400
    },
    {
      "epoch": 2.09,
      "grad_norm": 8.953125,
      "learning_rate": 4.974874371859296e-06,
      "loss": 1.1308,
      "step": 83500
    },
    {
      "epoch": 2.09,
      "grad_norm": 5.03515625,
      "learning_rate": 4.944723618090452e-06,
      "loss": 1.1502,
      "step": 83600
    },
    {
      "epoch": 2.09,
      "grad_norm": 6.67578125,
      "learning_rate": 4.914572864321608e-06,
      "loss": 1.1708,
      "step": 83700
    },
    {
      "epoch": 2.1,
      "grad_norm": 6.328125,
      "learning_rate": 4.884422110552764e-06,
      "loss": 1.152,
      "step": 83800
    },
    {
      "epoch": 2.1,
      "grad_norm": 6.3203125,
      "learning_rate": 4.85427135678392e-06,
      "loss": 1.1794,
      "step": 83900
    },
    {
      "epoch": 2.1,
      "grad_norm": 5.12890625,
      "learning_rate": 4.824120603015076e-06,
      "loss": 1.1732,
      "step": 84000
    },
    {
      "epoch": 2.1,
      "grad_norm": 7.3125,
      "learning_rate": 4.793969849246231e-06,
      "loss": 1.1665,
      "step": 84100
    },
    {
      "epoch": 2.1,
      "grad_norm": 15.8359375,
      "learning_rate": 4.763819095477387e-06,
      "loss": 1.2151,
      "step": 84200
    },
    {
      "epoch": 2.11,
      "grad_norm": 8.4453125,
      "learning_rate": 4.733668341708543e-06,
      "loss": 1.208,
      "step": 84300
    },
    {
      "epoch": 2.11,
      "grad_norm": 9.8828125,
      "learning_rate": 4.703517587939699e-06,
      "loss": 1.1756,
      "step": 84400
    },
    {
      "epoch": 2.11,
      "grad_norm": 6.1171875,
      "learning_rate": 4.673366834170855e-06,
      "loss": 1.1569,
      "step": 84500
    },
    {
      "epoch": 2.12,
      "grad_norm": 7.6953125,
      "learning_rate": 4.64321608040201e-06,
      "loss": 1.1985,
      "step": 84600
    },
    {
      "epoch": 2.12,
      "grad_norm": 9.2734375,
      "learning_rate": 4.613065326633166e-06,
      "loss": 1.1403,
      "step": 84700
    },
    {
      "epoch": 2.12,
      "grad_norm": 7.54296875,
      "learning_rate": 4.582914572864322e-06,
      "loss": 1.1365,
      "step": 84800
    },
    {
      "epoch": 2.12,
      "grad_norm": 10.5,
      "learning_rate": 4.5527638190954775e-06,
      "loss": 1.2374,
      "step": 84900
    },
    {
      "epoch": 2.12,
      "grad_norm": 7.4609375,
      "learning_rate": 4.522613065326633e-06,
      "loss": 1.1565,
      "step": 85000
    },
    {
      "epoch": 2.12,
      "eval_loss": 1.2156389951705933,
      "eval_runtime": 58.3222,
      "eval_samples_per_second": 17.146,
      "eval_steps_per_second": 4.287,
      "step": 85000
    },
    {
      "epoch": 2.13,
      "grad_norm": 5.97265625,
      "learning_rate": 4.492462311557789e-06,
      "loss": 1.2103,
      "step": 85100
    },
    {
      "epoch": 2.13,
      "grad_norm": 4.7109375,
      "learning_rate": 4.462311557788944e-06,
      "loss": 1.2058,
      "step": 85200
    },
    {
      "epoch": 2.13,
      "grad_norm": 9.140625,
      "learning_rate": 4.4321608040201e-06,
      "loss": 1.1758,
      "step": 85300
    },
    {
      "epoch": 2.13,
      "grad_norm": 6.52734375,
      "learning_rate": 4.4020100502512564e-06,
      "loss": 1.197,
      "step": 85400
    },
    {
      "epoch": 2.14,
      "grad_norm": 6.92578125,
      "learning_rate": 4.371859296482412e-06,
      "loss": 1.176,
      "step": 85500
    },
    {
      "epoch": 2.14,
      "grad_norm": 9.8984375,
      "learning_rate": 4.341708542713568e-06,
      "loss": 1.2176,
      "step": 85600
    },
    {
      "epoch": 2.14,
      "grad_norm": 6.609375,
      "learning_rate": 4.311557788944724e-06,
      "loss": 1.1897,
      "step": 85700
    },
    {
      "epoch": 2.15,
      "grad_norm": 5.81640625,
      "learning_rate": 4.281407035175879e-06,
      "loss": 1.1756,
      "step": 85800
    },
    {
      "epoch": 2.15,
      "grad_norm": 7.0703125,
      "learning_rate": 4.251256281407035e-06,
      "loss": 1.1818,
      "step": 85900
    },
    {
      "epoch": 2.15,
      "grad_norm": 7.296875,
      "learning_rate": 4.221105527638191e-06,
      "loss": 1.1527,
      "step": 86000
    },
    {
      "epoch": 2.15,
      "grad_norm": 4.29296875,
      "learning_rate": 4.190954773869347e-06,
      "loss": 1.1846,
      "step": 86100
    },
    {
      "epoch": 2.15,
      "grad_norm": 13.1875,
      "learning_rate": 4.160804020100503e-06,
      "loss": 1.2005,
      "step": 86200
    },
    {
      "epoch": 2.16,
      "grad_norm": 7.1328125,
      "learning_rate": 4.130653266331658e-06,
      "loss": 1.183,
      "step": 86300
    },
    {
      "epoch": 2.16,
      "grad_norm": 4.86328125,
      "learning_rate": 4.100502512562814e-06,
      "loss": 1.1729,
      "step": 86400
    },
    {
      "epoch": 2.16,
      "grad_norm": 6.046875,
      "learning_rate": 4.07035175879397e-06,
      "loss": 1.1052,
      "step": 86500
    },
    {
      "epoch": 2.17,
      "grad_norm": 5.03125,
      "learning_rate": 4.0402010050251256e-06,
      "loss": 1.181,
      "step": 86600
    },
    {
      "epoch": 2.17,
      "grad_norm": 9.9609375,
      "learning_rate": 4.010050251256282e-06,
      "loss": 1.2041,
      "step": 86700
    },
    {
      "epoch": 2.17,
      "grad_norm": 7.140625,
      "learning_rate": 3.979899497487438e-06,
      "loss": 1.1897,
      "step": 86800
    },
    {
      "epoch": 2.17,
      "grad_norm": 5.45703125,
      "learning_rate": 3.949748743718593e-06,
      "loss": 1.2078,
      "step": 86900
    },
    {
      "epoch": 2.17,
      "grad_norm": 7.04296875,
      "learning_rate": 3.919597989949749e-06,
      "loss": 1.0977,
      "step": 87000
    },
    {
      "epoch": 2.18,
      "grad_norm": 6.44140625,
      "learning_rate": 3.889447236180905e-06,
      "loss": 1.209,
      "step": 87100
    },
    {
      "epoch": 2.18,
      "grad_norm": 6.3046875,
      "learning_rate": 3.8592964824120606e-06,
      "loss": 1.1484,
      "step": 87200
    },
    {
      "epoch": 2.18,
      "grad_norm": 15.6484375,
      "learning_rate": 3.829145728643216e-06,
      "loss": 1.2224,
      "step": 87300
    },
    {
      "epoch": 2.19,
      "grad_norm": 6.73046875,
      "learning_rate": 3.7989949748743715e-06,
      "loss": 1.1861,
      "step": 87400
    },
    {
      "epoch": 2.19,
      "grad_norm": 5.4921875,
      "learning_rate": 3.768844221105527e-06,
      "loss": 1.1503,
      "step": 87500
    },
    {
      "epoch": 2.19,
      "eval_loss": 1.2061641216278076,
      "eval_runtime": 58.4078,
      "eval_samples_per_second": 17.121,
      "eval_steps_per_second": 4.28,
      "step": 87500
    },
    {
      "epoch": 2.19,
      "grad_norm": 9.4921875,
      "learning_rate": 3.7386934673366837e-06,
      "loss": 1.2169,
      "step": 87600
    },
    {
      "epoch": 2.19,
      "grad_norm": 14.7265625,
      "learning_rate": 3.7085427135678394e-06,
      "loss": 1.1826,
      "step": 87700
    },
    {
      "epoch": 2.19,
      "grad_norm": 7.20703125,
      "learning_rate": 3.678391959798995e-06,
      "loss": 1.2399,
      "step": 87800
    },
    {
      "epoch": 2.2,
      "grad_norm": 5.81640625,
      "learning_rate": 3.648241206030151e-06,
      "loss": 1.1242,
      "step": 87900
    },
    {
      "epoch": 2.2,
      "grad_norm": 8.4375,
      "learning_rate": 3.6180904522613065e-06,
      "loss": 1.1827,
      "step": 88000
    },
    {
      "epoch": 2.2,
      "grad_norm": 4.69921875,
      "learning_rate": 3.587939698492462e-06,
      "loss": 1.2243,
      "step": 88100
    },
    {
      "epoch": 2.21,
      "grad_norm": 7.3671875,
      "learning_rate": 3.557788944723618e-06,
      "loss": 1.2145,
      "step": 88200
    },
    {
      "epoch": 2.21,
      "grad_norm": 7.70703125,
      "learning_rate": 3.527638190954774e-06,
      "loss": 1.2503,
      "step": 88300
    },
    {
      "epoch": 2.21,
      "grad_norm": 5.734375,
      "learning_rate": 3.4974874371859297e-06,
      "loss": 1.2107,
      "step": 88400
    },
    {
      "epoch": 2.21,
      "grad_norm": 5.58984375,
      "learning_rate": 3.4673366834170854e-06,
      "loss": 1.1683,
      "step": 88500
    },
    {
      "epoch": 2.21,
      "grad_norm": 7.44921875,
      "learning_rate": 3.4371859296482415e-06,
      "loss": 1.1781,
      "step": 88600
    },
    {
      "epoch": 2.22,
      "grad_norm": 8.6953125,
      "learning_rate": 3.407035175879397e-06,
      "loss": 1.175,
      "step": 88700
    },
    {
      "epoch": 2.22,
      "grad_norm": 5.56640625,
      "learning_rate": 3.376884422110553e-06,
      "loss": 1.1351,
      "step": 88800
    },
    {
      "epoch": 2.22,
      "grad_norm": 7.08984375,
      "learning_rate": 3.346733668341709e-06,
      "loss": 1.1804,
      "step": 88900
    },
    {
      "epoch": 2.23,
      "grad_norm": 7.60546875,
      "learning_rate": 3.3165829145728647e-06,
      "loss": 1.2199,
      "step": 89000
    },
    {
      "epoch": 2.23,
      "grad_norm": 7.09765625,
      "learning_rate": 3.28643216080402e-06,
      "loss": 1.1729,
      "step": 89100
    },
    {
      "epoch": 2.23,
      "grad_norm": 6.39453125,
      "learning_rate": 3.2562814070351756e-06,
      "loss": 1.1395,
      "step": 89200
    },
    {
      "epoch": 2.23,
      "grad_norm": 8.7734375,
      "learning_rate": 3.2261306532663318e-06,
      "loss": 1.2157,
      "step": 89300
    },
    {
      "epoch": 2.23,
      "grad_norm": 8.4765625,
      "learning_rate": 3.1959798994974875e-06,
      "loss": 1.2263,
      "step": 89400
    },
    {
      "epoch": 2.24,
      "grad_norm": 8.765625,
      "learning_rate": 3.165829145728643e-06,
      "loss": 1.175,
      "step": 89500
    },
    {
      "epoch": 2.24,
      "grad_norm": 8.390625,
      "learning_rate": 3.1356783919597993e-06,
      "loss": 1.224,
      "step": 89600
    },
    {
      "epoch": 2.24,
      "grad_norm": 7.52734375,
      "learning_rate": 3.105527638190955e-06,
      "loss": 1.1259,
      "step": 89700
    },
    {
      "epoch": 2.25,
      "grad_norm": 4.87890625,
      "learning_rate": 3.0753768844221106e-06,
      "loss": 1.1547,
      "step": 89800
    },
    {
      "epoch": 2.25,
      "grad_norm": 6.99609375,
      "learning_rate": 3.0452261306532663e-06,
      "loss": 1.1514,
      "step": 89900
    },
    {
      "epoch": 2.25,
      "grad_norm": 10.671875,
      "learning_rate": 3.0150753768844224e-06,
      "loss": 1.1635,
      "step": 90000
    },
    {
      "epoch": 2.25,
      "eval_loss": 1.2134754657745361,
      "eval_runtime": 58.4436,
      "eval_samples_per_second": 17.111,
      "eval_steps_per_second": 4.278,
      "step": 90000
    },
    {
      "epoch": 2.25,
      "grad_norm": 7.99609375,
      "learning_rate": 2.984924623115578e-06,
      "loss": 1.1521,
      "step": 90100
    },
    {
      "epoch": 2.25,
      "grad_norm": 5.59375,
      "learning_rate": 2.9547738693467334e-06,
      "loss": 1.1633,
      "step": 90200
    },
    {
      "epoch": 2.26,
      "grad_norm": 6.43359375,
      "learning_rate": 2.9246231155778895e-06,
      "loss": 1.1356,
      "step": 90300
    },
    {
      "epoch": 2.26,
      "grad_norm": 9.6875,
      "learning_rate": 2.894472361809045e-06,
      "loss": 1.1495,
      "step": 90400
    },
    {
      "epoch": 2.26,
      "grad_norm": 11.3359375,
      "learning_rate": 2.864321608040201e-06,
      "loss": 1.1602,
      "step": 90500
    },
    {
      "epoch": 2.27,
      "grad_norm": 5.234375,
      "learning_rate": 2.834170854271357e-06,
      "loss": 1.1457,
      "step": 90600
    },
    {
      "epoch": 2.27,
      "grad_norm": 7.78515625,
      "learning_rate": 2.8040201005025127e-06,
      "loss": 1.1323,
      "step": 90700
    },
    {
      "epoch": 2.27,
      "grad_norm": 8.3203125,
      "learning_rate": 2.7738693467336684e-06,
      "loss": 1.166,
      "step": 90800
    },
    {
      "epoch": 2.27,
      "grad_norm": 7.22265625,
      "learning_rate": 2.743718592964824e-06,
      "loss": 1.2096,
      "step": 90900
    },
    {
      "epoch": 2.27,
      "grad_norm": 6.0,
      "learning_rate": 2.71356783919598e-06,
      "loss": 1.1357,
      "step": 91000
    },
    {
      "epoch": 2.28,
      "grad_norm": 6.1875,
      "learning_rate": 2.683417085427136e-06,
      "loss": 1.1882,
      "step": 91100
    },
    {
      "epoch": 2.28,
      "grad_norm": 14.40625,
      "learning_rate": 2.6532663316582916e-06,
      "loss": 1.1675,
      "step": 91200
    },
    {
      "epoch": 2.28,
      "grad_norm": 5.90234375,
      "learning_rate": 2.6231155778894477e-06,
      "loss": 1.1491,
      "step": 91300
    },
    {
      "epoch": 2.29,
      "grad_norm": 8.15625,
      "learning_rate": 2.592964824120603e-06,
      "loss": 1.2387,
      "step": 91400
    },
    {
      "epoch": 2.29,
      "grad_norm": 7.91015625,
      "learning_rate": 2.5628140703517587e-06,
      "loss": 1.1936,
      "step": 91500
    },
    {
      "epoch": 2.29,
      "grad_norm": 8.0859375,
      "learning_rate": 2.5326633165829143e-06,
      "loss": 1.1737,
      "step": 91600
    },
    {
      "epoch": 2.29,
      "grad_norm": 8.1796875,
      "learning_rate": 2.5025125628140705e-06,
      "loss": 1.1918,
      "step": 91700
    },
    {
      "epoch": 2.29,
      "grad_norm": 7.984375,
      "learning_rate": 2.472361809045226e-06,
      "loss": 1.1782,
      "step": 91800
    },
    {
      "epoch": 2.3,
      "grad_norm": 10.5078125,
      "learning_rate": 2.442211055276382e-06,
      "loss": 1.2198,
      "step": 91900
    },
    {
      "epoch": 2.3,
      "grad_norm": 8.1015625,
      "learning_rate": 2.412060301507538e-06,
      "loss": 1.1851,
      "step": 92000
    },
    {
      "epoch": 2.3,
      "grad_norm": 8.953125,
      "learning_rate": 2.3819095477386936e-06,
      "loss": 1.1886,
      "step": 92100
    },
    {
      "epoch": 2.31,
      "grad_norm": 7.05859375,
      "learning_rate": 2.3517587939698493e-06,
      "loss": 1.2223,
      "step": 92200
    },
    {
      "epoch": 2.31,
      "grad_norm": 5.34765625,
      "learning_rate": 2.321608040201005e-06,
      "loss": 1.1676,
      "step": 92300
    },
    {
      "epoch": 2.31,
      "grad_norm": 13.9921875,
      "learning_rate": 2.291457286432161e-06,
      "loss": 1.1743,
      "step": 92400
    },
    {
      "epoch": 2.31,
      "grad_norm": 15.0703125,
      "learning_rate": 2.2613065326633164e-06,
      "loss": 1.1538,
      "step": 92500
    },
    {
      "epoch": 2.31,
      "eval_loss": 1.1857857704162598,
      "eval_runtime": 58.4052,
      "eval_samples_per_second": 17.122,
      "eval_steps_per_second": 4.28,
      "step": 92500
    },
    {
      "epoch": 2.31,
      "grad_norm": 8.4140625,
      "learning_rate": 2.231155778894472e-06,
      "loss": 1.0991,
      "step": 92600
    },
    {
      "epoch": 2.32,
      "grad_norm": 6.92578125,
      "learning_rate": 2.2010050251256282e-06,
      "loss": 1.156,
      "step": 92700
    },
    {
      "epoch": 2.32,
      "grad_norm": 7.78515625,
      "learning_rate": 2.170854271356784e-06,
      "loss": 1.1821,
      "step": 92800
    },
    {
      "epoch": 2.32,
      "grad_norm": 6.5078125,
      "learning_rate": 2.1407035175879396e-06,
      "loss": 1.2011,
      "step": 92900
    },
    {
      "epoch": 2.33,
      "grad_norm": 8.96875,
      "learning_rate": 2.1105527638190957e-06,
      "loss": 1.1835,
      "step": 93000
    },
    {
      "epoch": 2.33,
      "grad_norm": 8.3828125,
      "learning_rate": 2.0804020100502514e-06,
      "loss": 1.1827,
      "step": 93100
    },
    {
      "epoch": 2.33,
      "grad_norm": 9.6328125,
      "learning_rate": 2.050251256281407e-06,
      "loss": 1.1546,
      "step": 93200
    },
    {
      "epoch": 2.33,
      "grad_norm": 9.4765625,
      "learning_rate": 2.0201005025125628e-06,
      "loss": 1.1461,
      "step": 93300
    },
    {
      "epoch": 2.33,
      "grad_norm": 7.52734375,
      "learning_rate": 1.989949748743719e-06,
      "loss": 1.1271,
      "step": 93400
    },
    {
      "epoch": 2.34,
      "grad_norm": 9.96875,
      "learning_rate": 1.9597989949748746e-06,
      "loss": 1.1457,
      "step": 93500
    },
    {
      "epoch": 2.34,
      "grad_norm": 8.34375,
      "learning_rate": 1.9296482412060303e-06,
      "loss": 1.2159,
      "step": 93600
    },
    {
      "epoch": 2.34,
      "grad_norm": 7.27734375,
      "learning_rate": 1.8994974874371858e-06,
      "loss": 1.1968,
      "step": 93700
    },
    {
      "epoch": 2.34,
      "grad_norm": 4.19140625,
      "learning_rate": 1.8693467336683419e-06,
      "loss": 1.0567,
      "step": 93800
    },
    {
      "epoch": 2.35,
      "grad_norm": 4.45703125,
      "learning_rate": 1.8391959798994976e-06,
      "loss": 1.1876,
      "step": 93900
    },
    {
      "epoch": 2.35,
      "grad_norm": 10.453125,
      "learning_rate": 1.8090452261306533e-06,
      "loss": 1.1298,
      "step": 94000
    },
    {
      "epoch": 2.35,
      "grad_norm": 8.3125,
      "learning_rate": 1.778894472361809e-06,
      "loss": 1.1558,
      "step": 94100
    },
    {
      "epoch": 2.35,
      "grad_norm": 5.66796875,
      "learning_rate": 1.7487437185929648e-06,
      "loss": 1.2117,
      "step": 94200
    },
    {
      "epoch": 2.36,
      "grad_norm": 9.484375,
      "learning_rate": 1.7185929648241207e-06,
      "loss": 1.1362,
      "step": 94300
    },
    {
      "epoch": 2.36,
      "grad_norm": 7.44140625,
      "learning_rate": 1.6884422110552764e-06,
      "loss": 1.2152,
      "step": 94400
    },
    {
      "epoch": 2.36,
      "grad_norm": 19.65625,
      "learning_rate": 1.6582914572864323e-06,
      "loss": 1.1914,
      "step": 94500
    },
    {
      "epoch": 2.37,
      "grad_norm": 6.1015625,
      "learning_rate": 1.6281407035175878e-06,
      "loss": 1.1662,
      "step": 94600
    },
    {
      "epoch": 2.37,
      "grad_norm": 6.44921875,
      "learning_rate": 1.5979899497487437e-06,
      "loss": 1.1681,
      "step": 94700
    },
    {
      "epoch": 2.37,
      "grad_norm": 8.515625,
      "learning_rate": 1.5678391959798996e-06,
      "loss": 1.1554,
      "step": 94800
    },
    {
      "epoch": 2.37,
      "grad_norm": 10.2734375,
      "learning_rate": 1.5376884422110553e-06,
      "loss": 1.1776,
      "step": 94900
    },
    {
      "epoch": 2.38,
      "grad_norm": 10.0703125,
      "learning_rate": 1.5075376884422112e-06,
      "loss": 1.1715,
      "step": 95000
    },
    {
      "epoch": 2.38,
      "eval_loss": 1.2057503461837769,
      "eval_runtime": 58.3696,
      "eval_samples_per_second": 17.132,
      "eval_steps_per_second": 4.283,
      "step": 95000
    },
    {
      "epoch": 2.38,
      "grad_norm": 11.359375,
      "learning_rate": 1.4773869346733667e-06,
      "loss": 1.1669,
      "step": 95100
    },
    {
      "epoch": 2.38,
      "grad_norm": 7.42578125,
      "learning_rate": 1.4472361809045226e-06,
      "loss": 1.1658,
      "step": 95200
    },
    {
      "epoch": 2.38,
      "grad_norm": 4.3203125,
      "learning_rate": 1.4170854271356785e-06,
      "loss": 1.1943,
      "step": 95300
    },
    {
      "epoch": 2.38,
      "grad_norm": 8.9765625,
      "learning_rate": 1.3869346733668342e-06,
      "loss": 1.1445,
      "step": 95400
    },
    {
      "epoch": 2.39,
      "grad_norm": 5.4765625,
      "learning_rate": 1.35678391959799e-06,
      "loss": 1.2335,
      "step": 95500
    },
    {
      "epoch": 2.39,
      "grad_norm": 6.5859375,
      "learning_rate": 1.3266331658291458e-06,
      "loss": 1.2079,
      "step": 95600
    },
    {
      "epoch": 2.39,
      "grad_norm": 7.40625,
      "learning_rate": 1.2964824120603015e-06,
      "loss": 1.1193,
      "step": 95700
    },
    {
      "epoch": 2.4,
      "grad_norm": 8.328125,
      "learning_rate": 1.2663316582914572e-06,
      "loss": 1.2107,
      "step": 95800
    },
    {
      "epoch": 2.4,
      "grad_norm": 6.96875,
      "learning_rate": 1.236180904522613e-06,
      "loss": 1.17,
      "step": 95900
    },
    {
      "epoch": 2.4,
      "grad_norm": 8.6328125,
      "learning_rate": 1.206030150753769e-06,
      "loss": 1.1605,
      "step": 96000
    },
    {
      "epoch": 2.4,
      "grad_norm": 5.61328125,
      "learning_rate": 1.1758793969849247e-06,
      "loss": 1.274,
      "step": 96100
    },
    {
      "epoch": 2.41,
      "grad_norm": 6.5,
      "learning_rate": 1.1457286432160806e-06,
      "loss": 1.195,
      "step": 96200
    },
    {
      "epoch": 2.41,
      "grad_norm": 4.8125,
      "learning_rate": 1.115577889447236e-06,
      "loss": 1.1551,
      "step": 96300
    },
    {
      "epoch": 2.41,
      "grad_norm": 5.99609375,
      "learning_rate": 1.085427135678392e-06,
      "loss": 1.16,
      "step": 96400
    },
    {
      "epoch": 2.41,
      "grad_norm": 6.90625,
      "learning_rate": 1.0552763819095479e-06,
      "loss": 1.1629,
      "step": 96500
    },
    {
      "epoch": 2.42,
      "grad_norm": 4.7109375,
      "learning_rate": 1.0251256281407035e-06,
      "loss": 1.1291,
      "step": 96600
    },
    {
      "epoch": 2.42,
      "grad_norm": 15.6796875,
      "learning_rate": 9.949748743718594e-07,
      "loss": 1.1471,
      "step": 96700
    },
    {
      "epoch": 2.42,
      "grad_norm": 6.796875,
      "learning_rate": 9.648241206030151e-07,
      "loss": 1.163,
      "step": 96800
    },
    {
      "epoch": 2.42,
      "grad_norm": 7.49609375,
      "learning_rate": 9.346733668341709e-07,
      "loss": 1.1476,
      "step": 96900
    },
    {
      "epoch": 2.42,
      "grad_norm": 6.5546875,
      "learning_rate": 9.045226130653266e-07,
      "loss": 1.233,
      "step": 97000
    },
    {
      "epoch": 2.43,
      "grad_norm": 5.06640625,
      "learning_rate": 8.743718592964824e-07,
      "loss": 1.1224,
      "step": 97100
    },
    {
      "epoch": 2.43,
      "grad_norm": 8.3828125,
      "learning_rate": 8.442211055276382e-07,
      "loss": 1.1806,
      "step": 97200
    },
    {
      "epoch": 2.43,
      "grad_norm": 8.953125,
      "learning_rate": 8.140703517587939e-07,
      "loss": 1.1643,
      "step": 97300
    },
    {
      "epoch": 2.44,
      "grad_norm": 7.90234375,
      "learning_rate": 7.839195979899498e-07,
      "loss": 1.1284,
      "step": 97400
    },
    {
      "epoch": 2.44,
      "grad_norm": 6.64453125,
      "learning_rate": 7.537688442211056e-07,
      "loss": 1.1856,
      "step": 97500
    },
    {
      "epoch": 2.44,
      "eval_loss": 1.1979784965515137,
      "eval_runtime": 58.3179,
      "eval_samples_per_second": 17.147,
      "eval_steps_per_second": 4.287,
      "step": 97500
    },
    {
      "epoch": 2.44,
      "grad_norm": 10.0625,
      "learning_rate": 7.236180904522613e-07,
      "loss": 1.1275,
      "step": 97600
    },
    {
      "epoch": 2.44,
      "grad_norm": 8.5703125,
      "learning_rate": 6.934673366834171e-07,
      "loss": 1.1365,
      "step": 97700
    },
    {
      "epoch": 2.44,
      "grad_norm": 14.328125,
      "learning_rate": 6.633165829145729e-07,
      "loss": 1.1769,
      "step": 97800
    },
    {
      "epoch": 2.45,
      "grad_norm": 6.0625,
      "learning_rate": 6.331658291457286e-07,
      "loss": 1.1797,
      "step": 97900
    },
    {
      "epoch": 2.45,
      "grad_norm": 4.9140625,
      "learning_rate": 6.030150753768845e-07,
      "loss": 1.1849,
      "step": 98000
    },
    {
      "epoch": 2.45,
      "grad_norm": 5.08984375,
      "learning_rate": 5.728643216080403e-07,
      "loss": 1.1552,
      "step": 98100
    },
    {
      "epoch": 2.46,
      "grad_norm": 6.0,
      "learning_rate": 5.42713567839196e-07,
      "loss": 1.2071,
      "step": 98200
    },
    {
      "epoch": 2.46,
      "grad_norm": 7.3125,
      "learning_rate": 5.125628140703518e-07,
      "loss": 1.1287,
      "step": 98300
    },
    {
      "epoch": 2.46,
      "grad_norm": 9.8671875,
      "learning_rate": 4.824120603015076e-07,
      "loss": 1.1739,
      "step": 98400
    },
    {
      "epoch": 2.46,
      "grad_norm": 7.5390625,
      "learning_rate": 4.522613065326633e-07,
      "loss": 1.1723,
      "step": 98500
    },
    {
      "epoch": 2.46,
      "grad_norm": 13.0,
      "learning_rate": 4.221105527638191e-07,
      "loss": 1.187,
      "step": 98600
    },
    {
      "epoch": 2.47,
      "grad_norm": 9.7890625,
      "learning_rate": 3.919597989949749e-07,
      "loss": 1.1765,
      "step": 98700
    },
    {
      "epoch": 2.47,
      "grad_norm": 4.5625,
      "learning_rate": 3.6180904522613065e-07,
      "loss": 1.2047,
      "step": 98800
    },
    {
      "epoch": 2.47,
      "grad_norm": 11.3984375,
      "learning_rate": 3.3165829145728645e-07,
      "loss": 1.2022,
      "step": 98900
    },
    {
      "epoch": 2.48,
      "grad_norm": 6.21875,
      "learning_rate": 3.0150753768844224e-07,
      "loss": 1.1459,
      "step": 99000
    },
    {
      "epoch": 2.48,
      "grad_norm": 7.47265625,
      "learning_rate": 2.71356783919598e-07,
      "loss": 1.1626,
      "step": 99100
    },
    {
      "epoch": 2.48,
      "grad_norm": 3.32421875,
      "learning_rate": 2.412060301507538e-07,
      "loss": 1.1997,
      "step": 99200
    },
    {
      "epoch": 2.48,
      "grad_norm": 4.859375,
      "learning_rate": 2.1105527638190956e-07,
      "loss": 1.2174,
      "step": 99300
    },
    {
      "epoch": 2.48,
      "grad_norm": 7.328125,
      "learning_rate": 1.8090452261306533e-07,
      "loss": 1.2133,
      "step": 99400
    },
    {
      "epoch": 2.49,
      "grad_norm": 11.375,
      "learning_rate": 1.5075376884422112e-07,
      "loss": 1.1436,
      "step": 99500
    },
    {
      "epoch": 2.49,
      "grad_norm": 6.7578125,
      "learning_rate": 1.206030150753769e-07,
      "loss": 1.149,
      "step": 99600
    },
    {
      "epoch": 2.49,
      "grad_norm": 11.4765625,
      "learning_rate": 9.045226130653266e-08,
      "loss": 1.1612,
      "step": 99700
    },
    {
      "epoch": 2.5,
      "grad_norm": 4.63671875,
      "learning_rate": 6.030150753768845e-08,
      "loss": 1.1397,
      "step": 99800
    },
    {
      "epoch": 2.5,
      "grad_norm": 11.5859375,
      "learning_rate": 3.015075376884422e-08,
      "loss": 1.1792,
      "step": 99900
    },
    {
      "epoch": 2.5,
      "grad_norm": 5.5703125,
      "learning_rate": 0.0,
      "loss": 1.1628,
      "step": 100000
    },
    {
      "epoch": 2.5,
      "eval_loss": 1.200670599937439,
      "eval_runtime": 58.2769,
      "eval_samples_per_second": 17.159,
      "eval_steps_per_second": 4.29,
      "step": 100000
    }
  ],
  "logging_steps": 100,
  "max_steps": 100000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 2500,
  "total_flos": 1.5733698330624e+18,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}