{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 243750,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.006153846153846154,
      "grad_norm": 15.441011428833008,
      "learning_rate": 4.989764102564103e-05,
      "loss": 1.5402,
      "step": 500
    },
    {
      "epoch": 0.012307692307692308,
      "grad_norm": 9.19687557220459,
      "learning_rate": 4.979507692307693e-05,
      "loss": 1.2127,
      "step": 1000
    },
    {
      "epoch": 0.018461538461538463,
      "grad_norm": 11.655864715576172,
      "learning_rate": 4.969251282051282e-05,
      "loss": 1.1338,
      "step": 1500
    },
    {
      "epoch": 0.024615384615384615,
      "grad_norm": 20.931015014648438,
      "learning_rate": 4.958994871794872e-05,
      "loss": 1.0639,
      "step": 2000
    },
    {
      "epoch": 0.03076923076923077,
      "grad_norm": 12.394627571105957,
      "learning_rate": 4.948738461538462e-05,
      "loss": 1.0317,
      "step": 2500
    },
    {
      "epoch": 0.036923076923076927,
      "grad_norm": 12.100915908813477,
      "learning_rate": 4.9384820512820516e-05,
      "loss": 1.0283,
      "step": 3000
    },
    {
      "epoch": 0.043076923076923075,
      "grad_norm": 14.47819995880127,
      "learning_rate": 4.928225641025641e-05,
      "loss": 1.0115,
      "step": 3500
    },
    {
      "epoch": 0.04923076923076923,
      "grad_norm": 15.608175277709961,
      "learning_rate": 4.917969230769231e-05,
      "loss": 1.0152,
      "step": 4000
    },
    {
      "epoch": 0.055384615384615386,
      "grad_norm": 40.89103317260742,
      "learning_rate": 4.9077128205128206e-05,
      "loss": 0.9857,
      "step": 4500
    },
    {
      "epoch": 0.06153846153846154,
      "grad_norm": 5.455167293548584,
      "learning_rate": 4.89745641025641e-05,
      "loss": 1.0276,
      "step": 5000
    },
    {
      "epoch": 0.06769230769230769,
      "grad_norm": 25.55486297607422,
      "learning_rate": 4.8872e-05,
      "loss": 0.9773,
      "step": 5500
    },
    {
      "epoch": 0.07384615384615385,
      "grad_norm": 16.84791374206543,
      "learning_rate": 4.87694358974359e-05,
      "loss": 0.9584,
      "step": 6000
    },
    {
      "epoch": 0.08,
      "grad_norm": 16.22901725769043,
      "learning_rate": 4.86668717948718e-05,
      "loss": 1.0158,
      "step": 6500
    },
    {
      "epoch": 0.08615384615384615,
      "grad_norm": 22.361555099487305,
      "learning_rate": 4.856430769230769e-05,
      "loss": 0.9774,
      "step": 7000
    },
    {
      "epoch": 0.09230769230769231,
      "grad_norm": 11.770121574401855,
      "learning_rate": 4.8461743589743594e-05,
      "loss": 0.9599,
      "step": 7500
    },
    {
      "epoch": 0.09846153846153846,
      "grad_norm": 17.554147720336914,
      "learning_rate": 4.835917948717949e-05,
      "loss": 0.9438,
      "step": 8000
    },
    {
      "epoch": 0.10461538461538461,
      "grad_norm": 11.665477752685547,
      "learning_rate": 4.825661538461539e-05,
      "loss": 0.9608,
      "step": 8500
    },
    {
      "epoch": 0.11076923076923077,
      "grad_norm": 17.011043548583984,
      "learning_rate": 4.8154051282051285e-05,
      "loss": 0.9836,
      "step": 9000
    },
    {
      "epoch": 0.11692307692307692,
      "grad_norm": 12.644675254821777,
      "learning_rate": 4.805148717948718e-05,
      "loss": 0.9505,
      "step": 9500
    },
    {
      "epoch": 0.12307692307692308,
      "grad_norm": 16.7060489654541,
      "learning_rate": 4.794892307692308e-05,
      "loss": 0.9767,
      "step": 10000
    },
    {
      "epoch": 0.12923076923076923,
      "grad_norm": 15.813322067260742,
      "learning_rate": 4.784635897435898e-05,
      "loss": 0.947,
      "step": 10500
    },
    {
      "epoch": 0.13538461538461538,
      "grad_norm": 7.84370231628418,
      "learning_rate": 4.774379487179487e-05,
      "loss": 0.9553,
      "step": 11000
    },
    {
      "epoch": 0.14153846153846153,
      "grad_norm": 7.926013469696045,
      "learning_rate": 4.764123076923077e-05,
      "loss": 0.9518,
      "step": 11500
    },
    {
      "epoch": 0.1476923076923077,
      "grad_norm": 12.878284454345703,
      "learning_rate": 4.753866666666667e-05,
      "loss": 0.9307,
      "step": 12000
    },
    {
      "epoch": 0.15384615384615385,
      "grad_norm": 14.422932624816895,
      "learning_rate": 4.743610256410257e-05,
      "loss": 0.9562,
      "step": 12500
    },
    {
      "epoch": 0.16,
      "grad_norm": 8.4630708694458,
      "learning_rate": 4.733353846153846e-05,
      "loss": 0.931,
      "step": 13000
    },
    {
      "epoch": 0.16615384615384615,
      "grad_norm": 18.567655563354492,
      "learning_rate": 4.723097435897436e-05,
      "loss": 0.9471,
      "step": 13500
    },
    {
      "epoch": 0.1723076923076923,
      "grad_norm": 13.360151290893555,
      "learning_rate": 4.712841025641026e-05,
      "loss": 0.9538,
      "step": 14000
    },
    {
      "epoch": 0.17846153846153845,
      "grad_norm": 8.191686630249023,
      "learning_rate": 4.702584615384615e-05,
      "loss": 0.9378,
      "step": 14500
    },
    {
      "epoch": 0.18461538461538463,
      "grad_norm": 16.853740692138672,
      "learning_rate": 4.6923282051282054e-05,
      "loss": 0.9602,
      "step": 15000
    },
    {
      "epoch": 0.19076923076923077,
      "grad_norm": 13.758535385131836,
      "learning_rate": 4.682071794871795e-05,
      "loss": 0.9451,
      "step": 15500
    },
    {
      "epoch": 0.19692307692307692,
      "grad_norm": 11.444204330444336,
      "learning_rate": 4.671815384615385e-05,
      "loss": 0.919,
      "step": 16000
    },
    {
      "epoch": 0.20307692307692307,
      "grad_norm": 8.91698169708252,
      "learning_rate": 4.6615589743589744e-05,
      "loss": 0.9341,
      "step": 16500
    },
    {
      "epoch": 0.20923076923076922,
      "grad_norm": 47.140968322753906,
      "learning_rate": 4.6513025641025646e-05,
      "loss": 0.9455,
      "step": 17000
    },
    {
      "epoch": 0.2153846153846154,
      "grad_norm": 9.847350120544434,
      "learning_rate": 4.641046153846154e-05,
      "loss": 0.9336,
      "step": 17500
    },
    {
      "epoch": 0.22153846153846155,
      "grad_norm": 7.988911151885986,
      "learning_rate": 4.630789743589744e-05,
      "loss": 0.9189,
      "step": 18000
    },
    {
      "epoch": 0.2276923076923077,
      "grad_norm": 12.728645324707031,
      "learning_rate": 4.6205333333333336e-05,
      "loss": 0.9206,
      "step": 18500
    },
    {
      "epoch": 0.23384615384615384,
      "grad_norm": 16.50530433654785,
      "learning_rate": 4.610276923076923e-05,
      "loss": 0.9325,
      "step": 19000
    },
    {
      "epoch": 0.24,
      "grad_norm": 12.972574234008789,
      "learning_rate": 4.600020512820513e-05,
      "loss": 0.8916,
      "step": 19500
    },
    {
      "epoch": 0.24615384615384617,
      "grad_norm": 12.662789344787598,
      "learning_rate": 4.5897641025641034e-05,
      "loss": 0.9135,
      "step": 20000
    },
    {
      "epoch": 0.2523076923076923,
      "grad_norm": 6.047543525695801,
      "learning_rate": 4.579507692307692e-05,
      "loss": 0.9244,
      "step": 20500
    },
    {
      "epoch": 0.25846153846153846,
      "grad_norm": 16.157981872558594,
      "learning_rate": 4.569251282051282e-05,
      "loss": 0.9131,
      "step": 21000
    },
    {
      "epoch": 0.26461538461538464,
      "grad_norm": 5.811090469360352,
      "learning_rate": 4.5589948717948724e-05,
      "loss": 0.9179,
      "step": 21500
    },
    {
      "epoch": 0.27076923076923076,
      "grad_norm": 7.486561298370361,
      "learning_rate": 4.548738461538461e-05,
      "loss": 0.9209,
      "step": 22000
    },
    {
      "epoch": 0.27692307692307694,
      "grad_norm": 9.555835723876953,
      "learning_rate": 4.5384820512820513e-05,
      "loss": 0.9323,
      "step": 22500
    },
    {
      "epoch": 0.28307692307692306,
      "grad_norm": 7.111580848693848,
      "learning_rate": 4.5282256410256415e-05,
      "loss": 0.9645,
      "step": 23000
    },
    {
      "epoch": 0.28923076923076924,
      "grad_norm": 87.1783218383789,
      "learning_rate": 4.517969230769231e-05,
      "loss": 0.9342,
      "step": 23500
    },
    {
      "epoch": 0.2953846153846154,
      "grad_norm": 13.641851425170898,
      "learning_rate": 4.5077128205128204e-05,
      "loss": 0.9286,
      "step": 24000
    },
    {
      "epoch": 0.30153846153846153,
      "grad_norm": 13.563926696777344,
      "learning_rate": 4.4974564102564105e-05,
      "loss": 0.9342,
      "step": 24500
    },
    {
      "epoch": 0.3076923076923077,
      "grad_norm": 16.500226974487305,
      "learning_rate": 4.4872e-05,
      "loss": 0.9239,
      "step": 25000
    },
    {
      "epoch": 0.31384615384615383,
      "grad_norm": 11.090267181396484,
      "learning_rate": 4.47694358974359e-05,
      "loss": 0.9261,
      "step": 25500
    },
    {
      "epoch": 0.32,
      "grad_norm": 9.335100173950195,
      "learning_rate": 4.4666871794871796e-05,
      "loss": 0.9099,
      "step": 26000
    },
    {
      "epoch": 0.3261538461538461,
      "grad_norm": 13.137558937072754,
      "learning_rate": 4.45643076923077e-05,
      "loss": 0.9318,
      "step": 26500
    },
    {
      "epoch": 0.3323076923076923,
      "grad_norm": 8.562922477722168,
      "learning_rate": 4.446174358974359e-05,
      "loss": 0.9289,
      "step": 27000
    },
    {
      "epoch": 0.3384615384615385,
      "grad_norm": 9.399090766906738,
      "learning_rate": 4.435917948717949e-05,
      "loss": 0.9225,
      "step": 27500
    },
    {
      "epoch": 0.3446153846153846,
      "grad_norm": 12.728873252868652,
      "learning_rate": 4.425661538461539e-05,
      "loss": 0.9334,
      "step": 28000
    },
    {
      "epoch": 0.3507692307692308,
      "grad_norm": 14.883627891540527,
      "learning_rate": 4.415405128205128e-05,
      "loss": 0.9224,
      "step": 28500
    },
    {
      "epoch": 0.3569230769230769,
      "grad_norm": 7.362912654876709,
      "learning_rate": 4.4051487179487184e-05,
      "loss": 0.9306,
      "step": 29000
    },
    {
      "epoch": 0.3630769230769231,
      "grad_norm": 7.706457614898682,
      "learning_rate": 4.394892307692308e-05,
      "loss": 0.9071,
      "step": 29500
    },
    {
      "epoch": 0.36923076923076925,
      "grad_norm": 10.347256660461426,
      "learning_rate": 4.384635897435897e-05,
      "loss": 0.9291,
      "step": 30000
    },
    {
      "epoch": 0.37538461538461537,
      "grad_norm": 8.69521713256836,
      "learning_rate": 4.3743794871794874e-05,
      "loss": 0.9067,
      "step": 30500
    },
    {
      "epoch": 0.38153846153846155,
      "grad_norm": 8.357994079589844,
      "learning_rate": 4.3641230769230776e-05,
      "loss": 0.9012,
      "step": 31000
    },
    {
      "epoch": 0.38769230769230767,
      "grad_norm": 14.694459915161133,
      "learning_rate": 4.3538666666666664e-05,
      "loss": 0.9056,
      "step": 31500
    },
    {
      "epoch": 0.39384615384615385,
      "grad_norm": 12.09356689453125,
      "learning_rate": 4.3436102564102565e-05,
      "loss": 0.945,
      "step": 32000
    },
    {
      "epoch": 0.4,
      "grad_norm": 17.006885528564453,
      "learning_rate": 4.3333538461538466e-05,
      "loss": 0.9466,
      "step": 32500
    },
    {
      "epoch": 0.40615384615384614,
      "grad_norm": 10.885141372680664,
      "learning_rate": 4.323097435897436e-05,
      "loss": 0.8933,
      "step": 33000
    },
    {
      "epoch": 0.4123076923076923,
      "grad_norm": 17.807836532592773,
      "learning_rate": 4.3128410256410256e-05,
      "loss": 0.9188,
      "step": 33500
    },
    {
      "epoch": 0.41846153846153844,
      "grad_norm": 15.311736106872559,
      "learning_rate": 4.302584615384616e-05,
      "loss": 0.8805,
      "step": 34000
    },
    {
      "epoch": 0.4246153846153846,
      "grad_norm": 7.447861671447754,
      "learning_rate": 4.292328205128205e-05,
      "loss": 0.9043,
      "step": 34500
    },
    {
      "epoch": 0.4307692307692308,
      "grad_norm": 21.085975646972656,
      "learning_rate": 4.282071794871795e-05,
      "loss": 0.8851,
      "step": 35000
    },
    {
      "epoch": 0.4369230769230769,
      "grad_norm": 19.813968658447266,
      "learning_rate": 4.271815384615385e-05,
      "loss": 0.8974,
      "step": 35500
    },
    {
      "epoch": 0.4430769230769231,
      "grad_norm": 9.81804084777832,
      "learning_rate": 4.261558974358975e-05,
      "loss": 0.8886,
      "step": 36000
    },
    {
      "epoch": 0.4492307692307692,
      "grad_norm": 17.820144653320312,
      "learning_rate": 4.2513025641025643e-05,
      "loss": 0.879,
      "step": 36500
    },
    {
      "epoch": 0.4553846153846154,
      "grad_norm": 10.280338287353516,
      "learning_rate": 4.241046153846154e-05,
      "loss": 0.9004,
      "step": 37000
    },
    {
      "epoch": 0.46153846153846156,
      "grad_norm": 12.174972534179688,
      "learning_rate": 4.230789743589744e-05,
      "loss": 0.8885,
      "step": 37500
    },
    {
      "epoch": 0.4676923076923077,
      "grad_norm": 13.532380104064941,
      "learning_rate": 4.2205333333333334e-05,
      "loss": 0.893,
      "step": 38000
    },
    {
      "epoch": 0.47384615384615386,
      "grad_norm": 13.15241527557373,
      "learning_rate": 4.2102769230769235e-05,
      "loss": 0.8866,
      "step": 38500
    },
    {
      "epoch": 0.48,
      "grad_norm": 11.801546096801758,
      "learning_rate": 4.200020512820513e-05,
      "loss": 0.8872,
      "step": 39000
    },
    {
      "epoch": 0.48615384615384616,
      "grad_norm": 7.681485652923584,
      "learning_rate": 4.1897641025641025e-05,
      "loss": 0.9096,
      "step": 39500
    },
    {
      "epoch": 0.49230769230769234,
      "grad_norm": 11.140621185302734,
      "learning_rate": 4.1795076923076926e-05,
      "loss": 0.8976,
      "step": 40000
    },
    {
      "epoch": 0.49846153846153846,
      "grad_norm": 10.523269653320312,
      "learning_rate": 4.169251282051283e-05,
      "loss": 0.9242,
      "step": 40500
    },
    {
      "epoch": 0.5046153846153846,
      "grad_norm": 26.289159774780273,
      "learning_rate": 4.1589948717948715e-05,
      "loss": 0.9132,
      "step": 41000
    },
    {
      "epoch": 0.5107692307692308,
      "grad_norm": 7.7466511726379395,
      "learning_rate": 4.1487384615384617e-05,
      "loss": 0.8945,
      "step": 41500
    },
    {
      "epoch": 0.5169230769230769,
      "grad_norm": 11.930273056030273,
      "learning_rate": 4.138482051282052e-05,
      "loss": 0.8633,
      "step": 42000
    },
    {
      "epoch": 0.5230769230769231,
      "grad_norm": 10.300634384155273,
      "learning_rate": 4.128225641025641e-05,
      "loss": 0.8913,
      "step": 42500
    },
    {
      "epoch": 0.5292307692307693,
      "grad_norm": 11.300227165222168,
      "learning_rate": 4.117969230769231e-05,
      "loss": 0.887,
      "step": 43000
    },
    {
      "epoch": 0.5353846153846153,
      "grad_norm": 5.9636993408203125,
      "learning_rate": 4.107712820512821e-05,
      "loss": 0.8979,
      "step": 43500
    },
    {
      "epoch": 0.5415384615384615,
      "grad_norm": 10.92505168914795,
      "learning_rate": 4.09745641025641e-05,
      "loss": 0.8941,
      "step": 44000
    },
    {
      "epoch": 0.5476923076923077,
      "grad_norm": 8.724580764770508,
      "learning_rate": 4.0872000000000004e-05,
      "loss": 0.9351,
      "step": 44500
    },
    {
      "epoch": 0.5538461538461539,
      "grad_norm": 43.44682312011719,
      "learning_rate": 4.07694358974359e-05,
      "loss": 0.8973,
      "step": 45000
    },
    {
      "epoch": 0.56,
      "grad_norm": 5.966015338897705,
      "learning_rate": 4.06668717948718e-05,
      "loss": 0.9027,
      "step": 45500
    },
    {
      "epoch": 0.5661538461538461,
      "grad_norm": 15.022377014160156,
      "learning_rate": 4.0564307692307695e-05,
      "loss": 0.899,
      "step": 46000
    },
    {
      "epoch": 0.5723076923076923,
      "grad_norm": 17.20585823059082,
      "learning_rate": 4.046174358974359e-05,
      "loss": 0.8816,
      "step": 46500
    },
    {
      "epoch": 0.5784615384615385,
      "grad_norm": 15.906556129455566,
      "learning_rate": 4.035917948717949e-05,
      "loss": 0.9028,
      "step": 47000
    },
    {
      "epoch": 0.5846153846153846,
      "grad_norm": 9.913738250732422,
      "learning_rate": 4.0256615384615386e-05,
      "loss": 0.8919,
      "step": 47500
    },
    {
      "epoch": 0.5907692307692308,
      "grad_norm": 8.156500816345215,
      "learning_rate": 4.015405128205129e-05,
      "loss": 0.8961,
      "step": 48000
    },
    {
      "epoch": 0.5969230769230769,
      "grad_norm": 5.659274578094482,
      "learning_rate": 4.005148717948718e-05,
      "loss": 0.8822,
      "step": 48500
    },
    {
      "epoch": 0.6030769230769231,
      "grad_norm": 14.798992156982422,
      "learning_rate": 3.9948923076923076e-05,
      "loss": 0.8831,
      "step": 49000
    },
    {
      "epoch": 0.6092307692307692,
      "grad_norm": 8.390789031982422,
      "learning_rate": 3.984635897435898e-05,
      "loss": 0.8873,
      "step": 49500
    },
    {
      "epoch": 0.6153846153846154,
      "grad_norm": 16.352163314819336,
      "learning_rate": 3.974379487179488e-05,
      "loss": 0.8772,
      "step": 50000
    },
    {
      "epoch": 0.6215384615384615,
      "grad_norm": 13.056022644042969,
      "learning_rate": 3.964123076923077e-05,
      "loss": 0.8895,
      "step": 50500
    },
    {
      "epoch": 0.6276923076923077,
      "grad_norm": 9.904449462890625,
      "learning_rate": 3.953866666666667e-05,
      "loss": 0.8684,
      "step": 51000
    },
    {
      "epoch": 0.6338461538461538,
      "grad_norm": 6.988232135772705,
      "learning_rate": 3.943610256410257e-05,
      "loss": 0.8763,
      "step": 51500
    },
    {
      "epoch": 0.64,
      "grad_norm": 7.327888488769531,
      "learning_rate": 3.9333538461538464e-05,
      "loss": 0.8991,
      "step": 52000
    },
    {
      "epoch": 0.6461538461538462,
      "grad_norm": 13.140373229980469,
      "learning_rate": 3.923097435897436e-05,
      "loss": 0.8447,
      "step": 52500
    },
    {
      "epoch": 0.6523076923076923,
      "grad_norm": 8.636076927185059,
      "learning_rate": 3.912841025641026e-05,
      "loss": 0.8935,
      "step": 53000
    },
    {
      "epoch": 0.6584615384615384,
      "grad_norm": 7.659938335418701,
      "learning_rate": 3.9025846153846155e-05,
      "loss": 0.8545,
      "step": 53500
    },
    {
      "epoch": 0.6646153846153846,
      "grad_norm": 16.80904197692871,
      "learning_rate": 3.892328205128205e-05,
      "loss": 0.8742,
      "step": 54000
    },
    {
      "epoch": 0.6707692307692308,
      "grad_norm": 46.774444580078125,
      "learning_rate": 3.882071794871795e-05,
      "loss": 0.9116,
      "step": 54500
    },
    {
      "epoch": 0.676923076923077,
      "grad_norm": 13.705849647521973,
      "learning_rate": 3.8718153846153845e-05,
      "loss": 0.8751,
      "step": 55000
    },
    {
      "epoch": 0.683076923076923,
      "grad_norm": 10.969395637512207,
      "learning_rate": 3.8615589743589746e-05,
      "loss": 0.8746,
      "step": 55500
    },
    {
      "epoch": 0.6892307692307692,
      "grad_norm": 8.494479179382324,
      "learning_rate": 3.851302564102564e-05,
      "loss": 0.8965,
      "step": 56000
    },
    {
      "epoch": 0.6953846153846154,
      "grad_norm": 8.132564544677734,
      "learning_rate": 3.841046153846154e-05,
      "loss": 0.8956,
      "step": 56500
    },
    {
      "epoch": 0.7015384615384616,
      "grad_norm": 9.6904935836792,
      "learning_rate": 3.830789743589744e-05,
      "loss": 0.8873,
      "step": 57000
    },
    {
      "epoch": 0.7076923076923077,
      "grad_norm": 13.937239646911621,
      "learning_rate": 3.820533333333334e-05,
      "loss": 0.8742,
      "step": 57500
    },
    {
      "epoch": 0.7138461538461538,
      "grad_norm": 14.830101013183594,
      "learning_rate": 3.810276923076923e-05,
      "loss": 0.8829,
      "step": 58000
    },
    {
      "epoch": 0.72,
      "grad_norm": 19.47818946838379,
      "learning_rate": 3.800020512820513e-05,
      "loss": 0.8977,
      "step": 58500
    },
    {
      "epoch": 0.7261538461538461,
      "grad_norm": 7.47867488861084,
      "learning_rate": 3.789764102564103e-05,
      "loss": 0.8792,
      "step": 59000
    },
    {
      "epoch": 0.7323076923076923,
      "grad_norm": 21.68954849243164,
      "learning_rate": 3.779507692307693e-05,
      "loss": 0.8681,
      "step": 59500
    },
    {
      "epoch": 0.7384615384615385,
      "grad_norm": 10.632262229919434,
      "learning_rate": 3.769251282051282e-05,
      "loss": 0.9502,
      "step": 60000
    },
    {
      "epoch": 0.7446153846153846,
      "grad_norm": 6.99915885925293,
      "learning_rate": 3.758994871794872e-05,
      "loss": 0.9691,
      "step": 60500
    },
    {
      "epoch": 0.7507692307692307,
      "grad_norm": 8.757305145263672,
      "learning_rate": 3.748738461538462e-05,
      "loss": 0.9446,
      "step": 61000
    },
    {
      "epoch": 0.7569230769230769,
      "grad_norm": 60.0024299621582,
      "learning_rate": 3.738482051282051e-05,
      "loss": 0.8887,
      "step": 61500
    },
    {
      "epoch": 0.7630769230769231,
      "grad_norm": 9.167777061462402,
      "learning_rate": 3.728225641025641e-05,
      "loss": 0.8668,
      "step": 62000
    },
    {
      "epoch": 0.7692307692307693,
      "grad_norm": 14.931575775146484,
      "learning_rate": 3.717969230769231e-05,
      "loss": 0.8832,
      "step": 62500
    },
    {
      "epoch": 0.7753846153846153,
      "grad_norm": 9.063596725463867,
      "learning_rate": 3.7077128205128206e-05,
      "loss": 0.8853,
      "step": 63000
    },
    {
      "epoch": 0.7815384615384615,
      "grad_norm": 8.906396865844727,
      "learning_rate": 3.69745641025641e-05,
      "loss": 0.8824,
      "step": 63500
    },
    {
      "epoch": 0.7876923076923077,
      "grad_norm": 12.211151123046875,
      "learning_rate": 3.6872e-05,
      "loss": 0.8747,
      "step": 64000
    },
    {
      "epoch": 0.7938461538461539,
      "grad_norm": 14.395017623901367,
      "learning_rate": 3.67694358974359e-05,
      "loss": 0.8858,
      "step": 64500
    },
    {
      "epoch": 0.8,
      "grad_norm": 6.887247085571289,
      "learning_rate": 3.66668717948718e-05,
      "loss": 0.8726,
      "step": 65000
    },
    {
      "epoch": 0.8061538461538461,
      "grad_norm": 9.119629859924316,
      "learning_rate": 3.656430769230769e-05,
      "loss": 0.901,
      "step": 65500
    },
    {
      "epoch": 0.8123076923076923,
      "grad_norm": 12.743978500366211,
      "learning_rate": 3.6461743589743594e-05,
      "loss": 0.8833,
      "step": 66000
    },
    {
      "epoch": 0.8184615384615385,
      "grad_norm": 22.06646156311035,
      "learning_rate": 3.635917948717949e-05,
      "loss": 0.9003,
      "step": 66500
    },
    {
      "epoch": 0.8246153846153846,
      "grad_norm": 10.94338607788086,
      "learning_rate": 3.625661538461539e-05,
      "loss": 0.8665,
      "step": 67000
    },
    {
      "epoch": 0.8307692307692308,
      "grad_norm": 6.3582892417907715,
      "learning_rate": 3.6154051282051285e-05,
      "loss": 0.875,
      "step": 67500
    },
    {
      "epoch": 0.8369230769230769,
      "grad_norm": 9.930424690246582,
      "learning_rate": 3.605148717948718e-05,
      "loss": 0.8544,
      "step": 68000
    },
    {
      "epoch": 0.8430769230769231,
      "grad_norm": 5.821447372436523,
      "learning_rate": 3.594892307692308e-05,
      "loss": 0.89,
      "step": 68500
    },
    {
      "epoch": 0.8492307692307692,
      "grad_norm": 112.40758514404297,
      "learning_rate": 3.5846358974358975e-05,
      "loss": 0.8703,
      "step": 69000
    },
    {
      "epoch": 0.8553846153846154,
      "grad_norm": 31.642396926879883,
      "learning_rate": 3.574379487179487e-05,
      "loss": 0.8625,
      "step": 69500
    },
    {
      "epoch": 0.8615384615384616,
      "grad_norm": 9.700116157531738,
      "learning_rate": 3.564123076923077e-05,
      "loss": 0.8722,
      "step": 70000
    },
    {
      "epoch": 0.8676923076923077,
      "grad_norm": 5.7299299240112305,
      "learning_rate": 3.553866666666667e-05,
      "loss": 0.8499,
      "step": 70500
    },
    {
      "epoch": 0.8738461538461538,
      "grad_norm": 8.070367813110352,
      "learning_rate": 3.543610256410256e-05,
      "loss": 0.8658,
      "step": 71000
    },
    {
      "epoch": 0.88,
      "grad_norm": 6.736323833465576,
      "learning_rate": 3.533353846153846e-05,
      "loss": 0.8882,
      "step": 71500
    },
    {
      "epoch": 0.8861538461538462,
      "grad_norm": 13.512341499328613,
      "learning_rate": 3.523097435897436e-05,
      "loss": 0.8704,
      "step": 72000
    },
    {
      "epoch": 0.8923076923076924,
      "grad_norm": 13.418638229370117,
      "learning_rate": 3.512841025641026e-05,
      "loss": 0.8669,
      "step": 72500
    },
    {
      "epoch": 0.8984615384615384,
      "grad_norm": 11.72990894317627,
      "learning_rate": 3.502584615384615e-05,
      "loss": 0.8479,
      "step": 73000
    },
    {
      "epoch": 0.9046153846153846,
      "grad_norm": 6.2977166175842285,
      "learning_rate": 3.4923282051282054e-05,
      "loss": 0.8371,
      "step": 73500
    },
    {
      "epoch": 0.9107692307692308,
      "grad_norm": 12.015417098999023,
      "learning_rate": 3.482071794871795e-05,
      "loss": 0.9644,
      "step": 74000
    },
    {
      "epoch": 0.916923076923077,
      "grad_norm": 7.836276531219482,
      "learning_rate": 3.471815384615385e-05,
      "loss": 0.999,
      "step": 74500
    },
    {
      "epoch": 0.9230769230769231,
      "grad_norm": 12.245140075683594,
      "learning_rate": 3.4615589743589744e-05,
      "loss": 0.869,
      "step": 75000
    },
    {
      "epoch": 0.9292307692307692,
      "grad_norm": 10.9988431930542,
      "learning_rate": 3.4513025641025646e-05,
      "loss": 0.855,
      "step": 75500
    },
    {
      "epoch": 0.9353846153846154,
      "grad_norm": 15.52302360534668,
      "learning_rate": 3.441046153846154e-05,
      "loss": 0.8352,
      "step": 76000
    },
    {
      "epoch": 0.9415384615384615,
      "grad_norm": 8.803082466125488,
      "learning_rate": 3.4307897435897435e-05,
      "loss": 0.9458,
      "step": 76500
    },
    {
      "epoch": 0.9476923076923077,
      "grad_norm": 5.698943138122559,
      "learning_rate": 3.4205333333333336e-05,
      "loss": 0.8773,
      "step": 77000
    },
    {
      "epoch": 0.9538461538461539,
      "grad_norm": 11.936111450195312,
      "learning_rate": 3.410276923076923e-05,
      "loss": 0.8875,
      "step": 77500
    },
    {
      "epoch": 0.96,
      "grad_norm": 9.272172927856445,
      "learning_rate": 3.400020512820513e-05,
      "loss": 0.879,
      "step": 78000
    },
    {
      "epoch": 0.9661538461538461,
      "grad_norm": 11.030377388000488,
      "learning_rate": 3.389764102564103e-05,
      "loss": 0.9059,
      "step": 78500
    },
    {
      "epoch": 0.9723076923076923,
      "grad_norm": 11.022510528564453,
      "learning_rate": 3.379507692307692e-05,
      "loss": 0.9664,
      "step": 79000
    },
    {
      "epoch": 0.9784615384615385,
      "grad_norm": 3.991237163543701,
      "learning_rate": 3.369251282051282e-05,
      "loss": 0.9605,
      "step": 79500
    },
    {
      "epoch": 0.9846153846153847,
      "grad_norm": 9.413519859313965,
      "learning_rate": 3.3589948717948724e-05,
      "loss": 0.8966,
      "step": 80000
    },
    {
      "epoch": 0.9907692307692307,
      "grad_norm": 8.189461708068848,
      "learning_rate": 3.348738461538461e-05,
      "loss": 0.9141,
      "step": 80500
    },
    {
      "epoch": 0.9969230769230769,
      "grad_norm": 11.961719512939453,
      "learning_rate": 3.338482051282051e-05,
      "loss": 0.9069,
      "step": 81000
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.63552,
      "eval_loss": 0.8960253000259399,
      "eval_runtime": 365.8936,
      "eval_samples_per_second": 136.652,
      "eval_steps_per_second": 17.081,
      "step": 81250
    },
    {
      "epoch": 1.003076923076923,
      "grad_norm": 16.277856826782227,
      "learning_rate": 3.3282256410256415e-05,
      "loss": 0.8695,
      "step": 81500
    },
    {
      "epoch": 1.0092307692307692,
      "grad_norm": 11.194209098815918,
      "learning_rate": 3.317969230769231e-05,
      "loss": 0.8373,
      "step": 82000
    },
    {
      "epoch": 1.0153846153846153,
      "grad_norm": 39.81032943725586,
      "learning_rate": 3.3077128205128204e-05,
      "loss": 0.8235,
      "step": 82500
    },
    {
      "epoch": 1.0215384615384615,
      "grad_norm": 4.980942249298096,
      "learning_rate": 3.2974564102564105e-05,
      "loss": 0.8384,
      "step": 83000
    },
    {
      "epoch": 1.0276923076923077,
      "grad_norm": 6.843782424926758,
      "learning_rate": 3.2872e-05,
      "loss": 0.8436,
      "step": 83500
    },
    {
      "epoch": 1.0338461538461539,
      "grad_norm": 18.53227996826172,
      "learning_rate": 3.27694358974359e-05,
      "loss": 0.8283,
      "step": 84000
    },
    {
      "epoch": 1.04,
      "grad_norm": 9.532831192016602,
      "learning_rate": 3.2666871794871796e-05,
      "loss": 0.81,
      "step": 84500
    },
    {
      "epoch": 1.0461538461538462,
      "grad_norm": 11.284141540527344,
      "learning_rate": 3.25643076923077e-05,
      "loss": 0.8255,
      "step": 85000
    },
    {
      "epoch": 1.0523076923076924,
      "grad_norm": 44.83882522583008,
      "learning_rate": 3.246174358974359e-05,
      "loss": 0.8654,
      "step": 85500
    },
    {
      "epoch": 1.0584615384615386,
      "grad_norm": 9.345219612121582,
      "learning_rate": 3.2359179487179486e-05,
      "loss": 0.8342,
      "step": 86000
    },
    {
      "epoch": 1.0646153846153845,
      "grad_norm": 45.78304672241211,
      "learning_rate": 3.225661538461539e-05,
      "loss": 0.8108,
      "step": 86500
    },
    {
      "epoch": 1.0707692307692307,
      "grad_norm": 7.037914276123047,
      "learning_rate": 3.215405128205128e-05,
      "loss": 0.828,
      "step": 87000
    },
    {
      "epoch": 1.0769230769230769,
      "grad_norm": 9.096734046936035,
      "learning_rate": 3.2051487179487184e-05,
      "loss": 0.8203,
      "step": 87500
    },
    {
      "epoch": 1.083076923076923,
      "grad_norm": 11.846463203430176,
      "learning_rate": 3.194892307692308e-05,
      "loss": 0.8453,
      "step": 88000
    },
    {
      "epoch": 1.0892307692307692,
      "grad_norm": 13.351137161254883,
      "learning_rate": 3.184635897435897e-05,
      "loss": 0.8554,
      "step": 88500
    },
    {
      "epoch": 1.0953846153846154,
      "grad_norm": 11.435630798339844,
      "learning_rate": 3.1743794871794874e-05,
      "loss": 1.0251,
      "step": 89000
    },
    {
      "epoch": 1.1015384615384616,
      "grad_norm": 18.781211853027344,
      "learning_rate": 3.1641230769230775e-05,
      "loss": 1.0217,
      "step": 89500
    },
    {
      "epoch": 1.1076923076923078,
      "grad_norm": 9.165902137756348,
      "learning_rate": 3.153866666666666e-05,
      "loss": 0.9597,
      "step": 90000
    },
    {
      "epoch": 1.113846153846154,
      "grad_norm": 7.910101890563965,
      "learning_rate": 3.1436102564102565e-05,
      "loss": 1.0139,
      "step": 90500
    },
    {
      "epoch": 1.12,
      "grad_norm": 9.269392967224121,
      "learning_rate": 3.1333538461538466e-05,
      "loss": 1.0017,
      "step": 91000
    },
    {
      "epoch": 1.126153846153846,
      "grad_norm": 1578.6934814453125,
      "learning_rate": 3.123097435897436e-05,
      "loss": 0.945,
      "step": 91500
    },
    {
      "epoch": 1.1323076923076922,
      "grad_norm": 9.8006010055542,
      "learning_rate": 3.1128410256410255e-05,
      "loss": 0.8265,
      "step": 92000
    },
    {
      "epoch": 1.1384615384615384,
      "grad_norm": 6.808800220489502,
      "learning_rate": 3.102584615384616e-05,
      "loss": 0.8309,
      "step": 92500
    },
    {
      "epoch": 1.1446153846153846,
      "grad_norm": 9.62307357788086,
      "learning_rate": 3.092328205128205e-05,
      "loss": 0.8457,
      "step": 93000
    },
    {
      "epoch": 1.1507692307692308,
      "grad_norm": 11.362460136413574,
      "learning_rate": 3.0820717948717946e-05,
      "loss": 0.854,
      "step": 93500
    },
    {
      "epoch": 1.156923076923077,
      "grad_norm": 15.59352970123291,
      "learning_rate": 3.071815384615385e-05,
      "loss": 0.8236,
      "step": 94000
    },
    {
      "epoch": 1.1630769230769231,
      "grad_norm": 16.21070671081543,
      "learning_rate": 3.061558974358975e-05,
      "loss": 0.8278,
      "step": 94500
    },
    {
      "epoch": 1.1692307692307693,
      "grad_norm": 9.07491683959961,
      "learning_rate": 3.0513025641025643e-05,
      "loss": 0.8625,
      "step": 95000
    },
    {
      "epoch": 1.1753846153846155,
      "grad_norm": 11.091276168823242,
      "learning_rate": 3.0410461538461538e-05,
      "loss": 0.8767,
      "step": 95500
    },
    {
      "epoch": 1.1815384615384614,
      "grad_norm": 9.517818450927734,
      "learning_rate": 3.0307897435897436e-05,
      "loss": 1.0201,
      "step": 96000
    },
    {
      "epoch": 1.1876923076923076,
      "grad_norm": 15.169801712036133,
      "learning_rate": 3.0205333333333337e-05,
      "loss": 1.0669,
      "step": 96500
    },
    {
      "epoch": 1.1938461538461538,
      "grad_norm": 9.808341979980469,
      "learning_rate": 3.0102769230769235e-05,
      "loss": 1.0947,
      "step": 97000
    },
    {
      "epoch": 1.2,
      "grad_norm": 5.031385898590088,
      "learning_rate": 3.0000205128205126e-05,
      "loss": 1.0694,
      "step": 97500
    },
    {
      "epoch": 1.2061538461538461,
      "grad_norm": 11.511312484741211,
      "learning_rate": 2.9897641025641028e-05,
      "loss": 1.1515,
      "step": 98000
    },
    {
      "epoch": 1.2123076923076923,
      "grad_norm": 13.318807601928711,
      "learning_rate": 2.9795076923076926e-05,
      "loss": 1.202,
      "step": 98500
    },
    {
      "epoch": 1.2184615384615385,
      "grad_norm": 10.762772560119629,
      "learning_rate": 2.9692512820512824e-05,
      "loss": 1.4163,
      "step": 99000
    },
    {
      "epoch": 1.2246153846153847,
      "grad_norm": 287.3322448730469,
      "learning_rate": 2.9589948717948718e-05,
      "loss": 1.4719,
      "step": 99500
    },
    {
      "epoch": 1.2307692307692308,
      "grad_norm": 5.190405368804932,
      "learning_rate": 2.9487384615384616e-05,
      "loss": 1.4996,
      "step": 100000
    },
    {
      "epoch": 1.236923076923077,
      "grad_norm": 5.880887508392334,
      "learning_rate": 2.9384820512820514e-05,
      "loss": 1.4431,
      "step": 100500
    },
    {
      "epoch": 1.2430769230769232,
      "grad_norm": 461.6323547363281,
      "learning_rate": 2.928225641025641e-05,
      "loss": 1.3496,
      "step": 101000
    },
    {
      "epoch": 1.2492307692307691,
      "grad_norm": 9.11361312866211,
      "learning_rate": 2.9179692307692307e-05,
      "loss": 1.2812,
      "step": 101500
    },
    {
      "epoch": 1.2553846153846153,
      "grad_norm": 693.46240234375,
      "learning_rate": 2.9077128205128208e-05,
      "loss": 1.4237,
      "step": 102000
    },
    {
      "epoch": 1.2615384615384615,
      "grad_norm": 11.541461944580078,
      "learning_rate": 2.8974564102564106e-05,
      "loss": 1.3735,
      "step": 102500
    },
    {
      "epoch": 1.2676923076923077,
      "grad_norm": 28.216230392456055,
      "learning_rate": 2.8872e-05,
      "loss": 1.201,
      "step": 103000
    },
    {
      "epoch": 1.2738461538461539,
      "grad_norm": 10.394718170166016,
      "learning_rate": 2.87694358974359e-05,
      "loss": 1.0158,
      "step": 103500
    },
    {
      "epoch": 1.28,
      "grad_norm": 29.5662784576416,
      "learning_rate": 2.8666871794871797e-05,
      "loss": 0.995,
      "step": 104000
    },
    {
      "epoch": 1.2861538461538462,
      "grad_norm": 9.405082702636719,
      "learning_rate": 2.8564307692307695e-05,
      "loss": 1.0206,
      "step": 104500
    },
    {
      "epoch": 1.2923076923076924,
      "grad_norm": 5.457224369049072,
      "learning_rate": 2.846174358974359e-05,
      "loss": 1.0332,
      "step": 105000
    },
    {
      "epoch": 1.2984615384615386,
      "grad_norm": 41.30308532714844,
      "learning_rate": 2.8359179487179487e-05,
      "loss": 1.1253,
      "step": 105500
    },
    {
      "epoch": 1.3046153846153845,
      "grad_norm": 11.211895942687988,
      "learning_rate": 2.825661538461539e-05,
      "loss": 1.2573,
      "step": 106000
    },
    {
      "epoch": 1.3107692307692307,
      "grad_norm": 8.621989250183105,
      "learning_rate": 2.8154051282051287e-05,
      "loss": 1.0103,
      "step": 106500
    },
    {
      "epoch": 1.3169230769230769,
      "grad_norm": 22.292667388916016,
      "learning_rate": 2.8051487179487178e-05,
      "loss": 0.8642,
      "step": 107000
    },
    {
      "epoch": 1.323076923076923,
      "grad_norm": 9.553186416625977,
      "learning_rate": 2.794892307692308e-05,
      "loss": 0.8808,
      "step": 107500
    },
    {
      "epoch": 1.3292307692307692,
      "grad_norm": 38.40745544433594,
      "learning_rate": 2.7846358974358977e-05,
      "loss": 0.8601,
      "step": 108000
    },
    {
      "epoch": 1.3353846153846154,
      "grad_norm": 8.445296287536621,
      "learning_rate": 2.7743794871794872e-05,
      "loss": 0.8856,
      "step": 108500
    },
    {
      "epoch": 1.3415384615384616,
      "grad_norm": 9.499748229980469,
      "learning_rate": 2.764123076923077e-05,
      "loss": 0.8597,
      "step": 109000
    },
    {
      "epoch": 1.3476923076923077,
      "grad_norm": 12.908239364624023,
      "learning_rate": 2.7538666666666668e-05,
      "loss": 0.96,
      "step": 109500
    },
    {
      "epoch": 1.353846153846154,
      "grad_norm": 12.40140438079834,
      "learning_rate": 2.7436102564102566e-05,
      "loss": 0.9255,
      "step": 110000
    },
    {
      "epoch": 1.3599999999999999,
      "grad_norm": 6.697096824645996,
      "learning_rate": 2.733353846153846e-05,
      "loss": 0.8948,
      "step": 110500
    },
    {
      "epoch": 1.3661538461538463,
      "grad_norm": 35.72837829589844,
      "learning_rate": 2.7230974358974358e-05,
      "loss": 0.8818,
      "step": 111000
    },
    {
      "epoch": 1.3723076923076922,
      "grad_norm": 238.4368133544922,
      "learning_rate": 2.712841025641026e-05,
      "loss": 0.972,
      "step": 111500
    },
    {
      "epoch": 1.3784615384615384,
      "grad_norm": 7.303473949432373,
      "learning_rate": 2.7025846153846158e-05,
      "loss": 0.9542,
      "step": 112000
    },
    {
      "epoch": 1.3846153846153846,
      "grad_norm": 7.700074195861816,
      "learning_rate": 2.692328205128205e-05,
      "loss": 0.9335,
      "step": 112500
    },
    {
      "epoch": 1.3907692307692308,
      "grad_norm": 10.46871566772461,
      "learning_rate": 2.682071794871795e-05,
      "loss": 0.9423,
      "step": 113000
    },
    {
      "epoch": 1.396923076923077,
      "grad_norm": 6448.5751953125,
      "learning_rate": 2.6718153846153848e-05,
      "loss": 0.8893,
      "step": 113500
    },
    {
      "epoch": 1.403076923076923,
      "grad_norm": 12.207977294921875,
      "learning_rate": 2.6615589743589746e-05,
      "loss": 0.8791,
      "step": 114000
    },
    {
      "epoch": 1.4092307692307693,
      "grad_norm": 10.765968322753906,
      "learning_rate": 2.651302564102564e-05,
      "loss": 0.883,
      "step": 114500
    },
    {
      "epoch": 1.4153846153846155,
      "grad_norm": 46.836402893066406,
      "learning_rate": 2.641046153846154e-05,
      "loss": 0.8503,
      "step": 115000
    },
    {
      "epoch": 1.4215384615384616,
      "grad_norm": 15.547259330749512,
      "learning_rate": 2.6307897435897437e-05,
      "loss": 0.8611,
      "step": 115500
    },
    {
      "epoch": 1.4276923076923076,
      "grad_norm": 11.880488395690918,
      "learning_rate": 2.6205333333333338e-05,
      "loss": 0.828,
      "step": 116000
    },
    {
      "epoch": 1.4338461538461538,
      "grad_norm": 20.702163696289062,
      "learning_rate": 2.610276923076923e-05,
      "loss": 0.8337,
      "step": 116500
    },
    {
      "epoch": 1.44,
      "grad_norm": 16.449716567993164,
      "learning_rate": 2.600020512820513e-05,
      "loss": 0.8362,
      "step": 117000
    },
    {
      "epoch": 1.4461538461538461,
      "grad_norm": 7.47279167175293,
      "learning_rate": 2.589764102564103e-05,
      "loss": 0.8629,
      "step": 117500
    },
    {
      "epoch": 1.4523076923076923,
      "grad_norm": 9.32483196258545,
      "learning_rate": 2.5795076923076923e-05,
      "loss": 0.9098,
      "step": 118000
    },
    {
      "epoch": 1.4584615384615385,
      "grad_norm": 18.407270431518555,
      "learning_rate": 2.569251282051282e-05,
      "loss": 0.8453,
      "step": 118500
    },
    {
      "epoch": 1.4646153846153847,
      "grad_norm": 15.567869186401367,
      "learning_rate": 2.558994871794872e-05,
      "loss": 1.0106,
      "step": 119000
    },
    {
      "epoch": 1.4707692307692308,
      "grad_norm": 10.05256462097168,
      "learning_rate": 2.5487384615384617e-05,
      "loss": 0.9477,
      "step": 119500
    },
    {
      "epoch": 1.476923076923077,
      "grad_norm": 26.323793411254883,
      "learning_rate": 2.5384820512820512e-05,
      "loss": 0.929,
      "step": 120000
    },
    {
      "epoch": 1.483076923076923,
      "grad_norm": 13.73338794708252,
      "learning_rate": 2.528225641025641e-05,
      "loss": 0.9654,
      "step": 120500
    },
    {
      "epoch": 1.4892307692307694,
      "grad_norm": 10.710335731506348,
      "learning_rate": 2.517969230769231e-05,
      "loss": 1.0033,
      "step": 121000
    },
    {
      "epoch": 1.4953846153846153,
      "grad_norm": 6.321354389190674,
      "learning_rate": 2.507712820512821e-05,
      "loss": 1.0259,
      "step": 121500
    },
    {
      "epoch": 1.5015384615384615,
      "grad_norm": 23.730915069580078,
      "learning_rate": 2.4974564102564104e-05,
      "loss": 0.935,
      "step": 122000
    },
    {
      "epoch": 1.5076923076923077,
      "grad_norm": 11.908498764038086,
      "learning_rate": 2.4872000000000002e-05,
      "loss": 0.9452,
      "step": 122500
    },
    {
      "epoch": 1.5138461538461538,
      "grad_norm": 14.499602317810059,
      "learning_rate": 2.47694358974359e-05,
      "loss": 0.8601,
      "step": 123000
    },
    {
      "epoch": 1.52,
      "grad_norm": 11.781792640686035,
      "learning_rate": 2.4666871794871794e-05,
      "loss": 0.8146,
      "step": 123500
    },
    {
      "epoch": 1.5261538461538462,
      "grad_norm": 13.83382797241211,
      "learning_rate": 2.4564307692307696e-05,
      "loss": 0.8392,
      "step": 124000
    },
    {
      "epoch": 1.5323076923076924,
      "grad_norm": 10.906270027160645,
      "learning_rate": 2.446174358974359e-05,
      "loss": 0.8258,
      "step": 124500
    },
    {
      "epoch": 1.5384615384615383,
      "grad_norm": 17.15850830078125,
      "learning_rate": 2.4359179487179488e-05,
      "loss": 0.8196,
      "step": 125000
    },
    {
      "epoch": 1.5446153846153847,
      "grad_norm": 5.727079391479492,
      "learning_rate": 2.4256615384615386e-05,
      "loss": 0.8243,
      "step": 125500
    },
    {
      "epoch": 1.5507692307692307,
      "grad_norm": 10.334206581115723,
      "learning_rate": 2.415405128205128e-05,
      "loss": 0.823,
      "step": 126000
    },
    {
      "epoch": 1.556923076923077,
      "grad_norm": 8.225074768066406,
      "learning_rate": 2.4051487179487182e-05,
      "loss": 0.8161,
      "step": 126500
    },
    {
      "epoch": 1.563076923076923,
      "grad_norm": 14.744596481323242,
      "learning_rate": 2.3948923076923077e-05,
      "loss": 0.8077,
      "step": 127000
    },
    {
      "epoch": 1.5692307692307692,
      "grad_norm": 17.01348876953125,
      "learning_rate": 2.3846358974358975e-05,
      "loss": 0.8177,
      "step": 127500
    },
    {
      "epoch": 1.5753846153846154,
      "grad_norm": 4.435682773590088,
      "learning_rate": 2.3743794871794873e-05,
      "loss": 0.8333,
      "step": 128000
    },
    {
      "epoch": 1.5815384615384616,
      "grad_norm": 11.365696907043457,
      "learning_rate": 2.364123076923077e-05,
      "loss": 0.8203,
      "step": 128500
    },
    {
      "epoch": 1.5876923076923077,
      "grad_norm": 10.943119049072266,
      "learning_rate": 2.353866666666667e-05,
      "loss": 0.8127,
      "step": 129000
    },
    {
      "epoch": 1.5938461538461537,
      "grad_norm": 4.344493389129639,
      "learning_rate": 2.3436102564102567e-05,
      "loss": 0.8129,
      "step": 129500
    },
    {
      "epoch": 1.6,
      "grad_norm": 8.629764556884766,
      "learning_rate": 2.333353846153846e-05,
      "loss": 1.1107,
      "step": 130000
    },
    {
      "epoch": 1.606153846153846,
      "grad_norm": 9.595703125,
      "learning_rate": 2.3230974358974363e-05,
      "loss": 1.2713,
      "step": 130500
    },
    {
      "epoch": 1.6123076923076924,
      "grad_norm": 18.25543975830078,
      "learning_rate": 2.3128410256410257e-05,
      "loss": 0.8794,
      "step": 131000
    },
    {
      "epoch": 1.6184615384615384,
      "grad_norm": 7.203105926513672,
      "learning_rate": 2.3025846153846155e-05,
      "loss": 0.8804,
      "step": 131500
    },
    {
      "epoch": 1.6246153846153846,
      "grad_norm": 5.362618923187256,
      "learning_rate": 2.2923282051282053e-05,
      "loss": 0.8374,
      "step": 132000
    },
    {
      "epoch": 1.6307692307692307,
      "grad_norm": 14.795818328857422,
      "learning_rate": 2.2820717948717948e-05,
      "loss": 0.8055,
      "step": 132500
    },
    {
      "epoch": 1.636923076923077,
      "grad_norm": 12.59130859375,
      "learning_rate": 2.2718153846153846e-05,
      "loss": 0.7875,
      "step": 133000
    },
    {
      "epoch": 1.643076923076923,
      "grad_norm": 28.4409122467041,
      "learning_rate": 2.2615589743589744e-05,
      "loss": 0.8014,
      "step": 133500
    },
    {
      "epoch": 1.6492307692307693,
      "grad_norm": 13.779533386230469,
      "learning_rate": 2.2513025641025642e-05,
      "loss": 0.8017,
      "step": 134000
    },
    {
      "epoch": 1.6553846153846155,
      "grad_norm": 7.0997233390808105,
      "learning_rate": 2.241046153846154e-05,
      "loss": 0.8027,
      "step": 134500
    },
    {
      "epoch": 1.6615384615384614,
      "grad_norm": 18.23006820678711,
      "learning_rate": 2.2307897435897438e-05,
      "loss": 0.8211,
      "step": 135000
    },
    {
      "epoch": 1.6676923076923078,
      "grad_norm": 13.56887149810791,
      "learning_rate": 2.2205333333333332e-05,
      "loss": 0.7968,
      "step": 135500
    },
    {
      "epoch": 1.6738461538461538,
      "grad_norm": 8.843082427978516,
      "learning_rate": 2.2102769230769234e-05,
      "loss": 0.8178,
      "step": 136000
    },
    {
      "epoch": 1.6800000000000002,
      "grad_norm": 16.259294509887695,
      "learning_rate": 2.200020512820513e-05,
      "loss": 0.8394,
      "step": 136500
    },
    {
      "epoch": 1.6861538461538461,
      "grad_norm": 15.540386199951172,
      "learning_rate": 2.1897641025641026e-05,
      "loss": 0.8253,
      "step": 137000
    },
    {
      "epoch": 1.6923076923076923,
      "grad_norm": 8.204901695251465,
      "learning_rate": 2.1795076923076924e-05,
      "loss": 0.8088,
      "step": 137500
    },
    {
      "epoch": 1.6984615384615385,
      "grad_norm": 11.878310203552246,
      "learning_rate": 2.1692512820512822e-05,
      "loss": 0.9658,
      "step": 138000
    },
    {
      "epoch": 1.7046153846153846,
      "grad_norm": 11.583746910095215,
      "learning_rate": 2.158994871794872e-05,
      "loss": 1.0792,
      "step": 138500
    },
    {
      "epoch": 1.7107692307692308,
      "grad_norm": 8.308416366577148,
      "learning_rate": 2.1487384615384618e-05,
      "loss": 0.8619,
      "step": 139000
    },
    {
      "epoch": 1.7169230769230768,
      "grad_norm": 17.01491928100586,
      "learning_rate": 2.1384820512820513e-05,
      "loss": 0.8132,
      "step": 139500
    },
    {
      "epoch": 1.7230769230769232,
      "grad_norm": 8.138616561889648,
      "learning_rate": 2.1282256410256414e-05,
      "loss": 0.8732,
      "step": 140000
    },
    {
      "epoch": 1.7292307692307691,
      "grad_norm": 6.041689872741699,
      "learning_rate": 2.117969230769231e-05,
      "loss": 0.8033,
      "step": 140500
    },
    {
      "epoch": 1.7353846153846155,
      "grad_norm": 17.37460708618164,
      "learning_rate": 2.1077128205128203e-05,
      "loss": 0.8168,
      "step": 141000
    },
    {
      "epoch": 1.7415384615384615,
      "grad_norm": 9.945074081420898,
      "learning_rate": 2.0974564102564105e-05,
      "loss": 0.768,
      "step": 141500
    },
    {
      "epoch": 1.7476923076923077,
      "grad_norm": 7.900252342224121,
      "learning_rate": 2.0872e-05,
      "loss": 0.8142,
      "step": 142000
    },
    {
      "epoch": 1.7538461538461538,
      "grad_norm": 11.957608222961426,
      "learning_rate": 2.0769435897435897e-05,
      "loss": 0.8076,
      "step": 142500
    },
    {
      "epoch": 1.76,
      "grad_norm": 8.552706718444824,
      "learning_rate": 2.0666871794871795e-05,
      "loss": 0.7828,
      "step": 143000
    },
    {
      "epoch": 1.7661538461538462,
      "grad_norm": 10.120139122009277,
      "learning_rate": 2.0564307692307693e-05,
      "loss": 0.7922,
      "step": 143500
    },
    {
      "epoch": 1.7723076923076924,
      "grad_norm": 11.39379596710205,
      "learning_rate": 2.046174358974359e-05,
      "loss": 0.81,
      "step": 144000
    },
    {
      "epoch": 1.7784615384615385,
      "grad_norm": 14.276771545410156,
      "learning_rate": 2.035917948717949e-05,
      "loss": 0.7994,
      "step": 144500
    },
    {
      "epoch": 1.7846153846153845,
      "grad_norm": 21.5369815826416,
      "learning_rate": 2.0256615384615384e-05,
      "loss": 0.7733,
      "step": 145000
    },
    {
      "epoch": 1.790769230769231,
      "grad_norm": 5.541769504547119,
      "learning_rate": 2.0154051282051285e-05,
      "loss": 0.804,
      "step": 145500
    },
    {
      "epoch": 1.7969230769230768,
      "grad_norm": 8.328704833984375,
      "learning_rate": 2.005148717948718e-05,
      "loss": 0.801,
      "step": 146000
    },
    {
      "epoch": 1.803076923076923,
      "grad_norm": 11.290348052978516,
      "learning_rate": 1.9948923076923078e-05,
      "loss": 0.8003,
      "step": 146500
    },
    {
      "epoch": 1.8092307692307692,
      "grad_norm": 10.958141326904297,
      "learning_rate": 1.9846358974358976e-05,
      "loss": 0.8086,
      "step": 147000
    },
    {
      "epoch": 1.8153846153846154,
      "grad_norm": 16.242244720458984,
      "learning_rate": 1.9743794871794874e-05,
      "loss": 0.7838,
      "step": 147500
    },
    {
      "epoch": 1.8215384615384616,
      "grad_norm": 5.659574031829834,
      "learning_rate": 1.964123076923077e-05,
      "loss": 0.7879,
      "step": 148000
    },
    {
      "epoch": 1.8276923076923077,
      "grad_norm": 7.112459659576416,
      "learning_rate": 1.9538666666666666e-05,
      "loss": 0.7991,
      "step": 148500
    },
    {
      "epoch": 1.833846153846154,
      "grad_norm": 24.664087295532227,
      "learning_rate": 1.9436102564102564e-05,
      "loss": 0.7605,
      "step": 149000
    },
    {
      "epoch": 1.8399999999999999,
      "grad_norm": 7.420779228210449,
      "learning_rate": 1.9333538461538462e-05,
      "loss": 0.7808,
      "step": 149500
    },
    {
      "epoch": 1.8461538461538463,
      "grad_norm": 12.564159393310547,
      "learning_rate": 1.923097435897436e-05,
      "loss": 0.7922,
      "step": 150000
    },
    {
      "epoch": 1.8523076923076922,
      "grad_norm": 6.439233779907227,
      "learning_rate": 1.9128410256410255e-05,
      "loss": 0.7836,
      "step": 150500
    },
    {
      "epoch": 1.8584615384615386,
      "grad_norm": 10.705535888671875,
      "learning_rate": 1.9025846153846156e-05,
      "loss": 0.7983,
      "step": 151000
    },
    {
      "epoch": 1.8646153846153846,
      "grad_norm": 15.499585151672363,
      "learning_rate": 1.892328205128205e-05,
      "loss": 0.7711,
      "step": 151500
    },
    {
      "epoch": 1.8707692307692307,
      "grad_norm": 23.128767013549805,
      "learning_rate": 1.882071794871795e-05,
      "loss": 0.7793,
      "step": 152000
    },
    {
      "epoch": 1.876923076923077,
      "grad_norm": 5.763806343078613,
      "learning_rate": 1.8718153846153847e-05,
      "loss": 0.798,
      "step": 152500
    },
    {
      "epoch": 1.883076923076923,
      "grad_norm": 14.247387886047363,
      "learning_rate": 1.8615589743589745e-05,
      "loss": 0.7735,
      "step": 153000
    },
    {
      "epoch": 1.8892307692307693,
      "grad_norm": 10.602540969848633,
      "learning_rate": 1.8513025641025643e-05,
      "loss": 0.7922,
      "step": 153500
    },
    {
      "epoch": 1.8953846153846152,
      "grad_norm": 8.66236400604248,
      "learning_rate": 1.841046153846154e-05,
      "loss": 0.7976,
      "step": 154000
    },
    {
      "epoch": 1.9015384615384616,
      "grad_norm": 10.02337646484375,
      "learning_rate": 1.8307897435897435e-05,
      "loss": 0.8002,
      "step": 154500
    },
    {
      "epoch": 1.9076923076923076,
      "grad_norm": 5.49017333984375,
      "learning_rate": 1.8205333333333337e-05,
      "loss": 0.7812,
      "step": 155000
    },
    {
      "epoch": 1.913846153846154,
      "grad_norm": 10.681382179260254,
      "learning_rate": 1.810276923076923e-05,
      "loss": 0.7858,
      "step": 155500
    },
    {
      "epoch": 1.92,
      "grad_norm": 6.697972297668457,
      "learning_rate": 1.800020512820513e-05,
      "loss": 0.7823,
      "step": 156000
    },
    {
      "epoch": 1.926153846153846,
      "grad_norm": 12.531645774841309,
      "learning_rate": 1.7897641025641027e-05,
      "loss": 0.766,
      "step": 156500
    },
    {
      "epoch": 1.9323076923076923,
      "grad_norm": 15.05484676361084,
      "learning_rate": 1.7795076923076922e-05,
      "loss": 0.8025,
      "step": 157000
    },
    {
      "epoch": 1.9384615384615385,
      "grad_norm": 7.559576511383057,
      "learning_rate": 1.769251282051282e-05,
      "loss": 0.7902,
      "step": 157500
    },
    {
      "epoch": 1.9446153846153846,
      "grad_norm": 10.720475196838379,
      "learning_rate": 1.7589948717948718e-05,
      "loss": 0.7806,
      "step": 158000
    },
    {
      "epoch": 1.9507692307692308,
      "grad_norm": 4.4791131019592285,
      "learning_rate": 1.7487384615384616e-05,
      "loss": 0.766,
      "step": 158500
    },
    {
      "epoch": 1.956923076923077,
      "grad_norm": 10.95101547241211,
      "learning_rate": 1.7384820512820514e-05,
      "loss": 0.7756,
      "step": 159000
    },
    {
      "epoch": 1.963076923076923,
      "grad_norm": 11.451761245727539,
      "learning_rate": 1.7282256410256412e-05,
      "loss": 0.8042,
      "step": 159500
    },
    {
      "epoch": 1.9692307692307693,
      "grad_norm": 4.005158424377441,
      "learning_rate": 1.7179692307692306e-05,
      "loss": 0.776,
      "step": 160000
    },
    {
      "epoch": 1.9753846153846153,
      "grad_norm": 10.658405303955078,
      "learning_rate": 1.7077128205128208e-05,
      "loss": 0.7729,
      "step": 160500
    },
    {
      "epoch": 1.9815384615384617,
      "grad_norm": 20.476329803466797,
      "learning_rate": 1.6974564102564102e-05,
      "loss": 0.7744,
      "step": 161000
    },
    {
      "epoch": 1.9876923076923076,
      "grad_norm": 6.919234275817871,
      "learning_rate": 1.6872e-05,
      "loss": 0.7602,
      "step": 161500
    },
    {
      "epoch": 1.9938461538461538,
      "grad_norm": 11.781153678894043,
      "learning_rate": 1.67694358974359e-05,
      "loss": 0.7743,
      "step": 162000
    },
    {
      "epoch": 2.0,
      "grad_norm": 20.044790267944336,
      "learning_rate": 1.6666871794871796e-05,
      "loss": 0.7397,
      "step": 162500
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.65574,
      "eval_loss": 0.8237444758415222,
      "eval_runtime": 367.5973,
      "eval_samples_per_second": 136.018,
      "eval_steps_per_second": 17.002,
      "step": 162500
    },
    {
      "epoch": 2.006153846153846,
      "grad_norm": 9.063031196594238,
      "learning_rate": 1.6564307692307694e-05,
      "loss": 0.7207,
      "step": 163000
    },
    {
      "epoch": 2.0123076923076924,
      "grad_norm": 11.945392608642578,
      "learning_rate": 1.6461743589743592e-05,
      "loss": 0.7343,
      "step": 163500
    },
    {
      "epoch": 2.0184615384615383,
      "grad_norm": 10.161272048950195,
      "learning_rate": 1.6359179487179487e-05,
      "loss": 0.7373,
      "step": 164000
    },
    {
      "epoch": 2.0246153846153847,
      "grad_norm": 12.465530395507812,
      "learning_rate": 1.6256615384615385e-05,
      "loss": 0.7249,
      "step": 164500
    },
    {
      "epoch": 2.0307692307692307,
      "grad_norm": 9.874961853027344,
      "learning_rate": 1.6154051282051283e-05,
      "loss": 0.7294,
      "step": 165000
    },
    {
      "epoch": 2.036923076923077,
      "grad_norm": 11.658003807067871,
      "learning_rate": 1.6051487179487178e-05,
      "loss": 0.7581,
      "step": 165500
    },
    {
      "epoch": 2.043076923076923,
      "grad_norm": 13.40922737121582,
      "learning_rate": 1.594892307692308e-05,
      "loss": 0.7284,
      "step": 166000
    },
    {
      "epoch": 2.0492307692307694,
      "grad_norm": 7.658026218414307,
      "learning_rate": 1.5846358974358973e-05,
      "loss": 0.7391,
      "step": 166500
    },
    {
      "epoch": 2.0553846153846154,
      "grad_norm": 9.820539474487305,
      "learning_rate": 1.574379487179487e-05,
      "loss": 0.737,
      "step": 167000
    },
    {
      "epoch": 2.0615384615384613,
      "grad_norm": 11.957843780517578,
      "learning_rate": 1.564123076923077e-05,
      "loss": 0.7257,
      "step": 167500
    },
    {
      "epoch": 2.0676923076923077,
      "grad_norm": 4.503595352172852,
      "learning_rate": 1.5538666666666667e-05,
      "loss": 0.7271,
      "step": 168000
    },
    {
      "epoch": 2.0738461538461537,
      "grad_norm": 13.069839477539062,
      "learning_rate": 1.5436102564102565e-05,
      "loss": 0.7292,
      "step": 168500
    },
    {
      "epoch": 2.08,
      "grad_norm": 27.35700225830078,
      "learning_rate": 1.5333538461538463e-05,
      "loss": 0.7381,
      "step": 169000
    },
    {
      "epoch": 2.086153846153846,
      "grad_norm": 16.696022033691406,
      "learning_rate": 1.523097435897436e-05,
      "loss": 0.7449,
      "step": 169500
    },
    {
      "epoch": 2.0923076923076924,
      "grad_norm": 15.537075996398926,
      "learning_rate": 1.5128410256410258e-05,
      "loss": 0.7271,
      "step": 170000
    },
    {
      "epoch": 2.0984615384615384,
      "grad_norm": 10.185432434082031,
      "learning_rate": 1.5025846153846154e-05,
      "loss": 0.7321,
      "step": 170500
    },
    {
      "epoch": 2.1046153846153848,
      "grad_norm": 19.049474716186523,
      "learning_rate": 1.4923282051282054e-05,
      "loss": 0.7125,
      "step": 171000
    },
    {
      "epoch": 2.1107692307692307,
      "grad_norm": 9.854257583618164,
      "learning_rate": 1.482071794871795e-05,
      "loss": 0.7453,
      "step": 171500
    },
    {
      "epoch": 2.116923076923077,
      "grad_norm": 19.2335262298584,
      "learning_rate": 1.4718153846153848e-05,
      "loss": 0.7235,
      "step": 172000
    },
    {
      "epoch": 2.123076923076923,
      "grad_norm": 13.282673835754395,
      "learning_rate": 1.4615589743589744e-05,
      "loss": 0.7597,
      "step": 172500
    },
    {
      "epoch": 2.129230769230769,
      "grad_norm": 15.75724983215332,
      "learning_rate": 1.451302564102564e-05,
      "loss": 0.7227,
      "step": 173000
    },
    {
      "epoch": 2.1353846153846154,
      "grad_norm": 7.548396587371826,
      "learning_rate": 1.441046153846154e-05,
      "loss": 0.7052,
      "step": 173500
    },
    {
      "epoch": 2.1415384615384614,
      "grad_norm": 16.264101028442383,
      "learning_rate": 1.4307897435897435e-05,
      "loss": 0.7449,
      "step": 174000
    },
    {
      "epoch": 2.147692307692308,
      "grad_norm": 16.141807556152344,
      "learning_rate": 1.4205333333333334e-05,
      "loss": 0.7478,
      "step": 174500
    },
    {
      "epoch": 2.1538461538461537,
      "grad_norm": 12.128105163574219,
      "learning_rate": 1.410276923076923e-05,
      "loss": 0.7373,
      "step": 175000
    },
    {
      "epoch": 2.16,
      "grad_norm": 12.569074630737305,
      "learning_rate": 1.4000205128205129e-05,
      "loss": 0.7279,
      "step": 175500
    },
    {
      "epoch": 2.166153846153846,
      "grad_norm": 14.476816177368164,
      "learning_rate": 1.3897641025641025e-05,
      "loss": 0.7369,
      "step": 176000
    },
    {
      "epoch": 2.1723076923076925,
      "grad_norm": 18.914968490600586,
      "learning_rate": 1.3795076923076925e-05,
      "loss": 0.7368,
      "step": 176500
    },
    {
      "epoch": 2.1784615384615384,
      "grad_norm": 7.522904396057129,
      "learning_rate": 1.3692512820512821e-05,
      "loss": 0.7548,
      "step": 177000
    },
    {
      "epoch": 2.184615384615385,
      "grad_norm": 6.408961772918701,
      "learning_rate": 1.3589948717948719e-05,
      "loss": 0.7335,
      "step": 177500
    },
    {
      "epoch": 2.190769230769231,
      "grad_norm": 3.879934310913086,
      "learning_rate": 1.3487384615384615e-05,
      "loss": 0.7447,
      "step": 178000
    },
    {
      "epoch": 2.1969230769230768,
      "grad_norm": 7.733790397644043,
      "learning_rate": 1.3384820512820515e-05,
      "loss": 0.7229,
      "step": 178500
    },
    {
      "epoch": 2.203076923076923,
      "grad_norm": 8.292427062988281,
      "learning_rate": 1.3282256410256411e-05,
      "loss": 0.7214,
      "step": 179000
    },
    {
      "epoch": 2.209230769230769,
      "grad_norm": 13.548661231994629,
      "learning_rate": 1.317969230769231e-05,
      "loss": 0.7406,
      "step": 179500
    },
    {
      "epoch": 2.2153846153846155,
      "grad_norm": 14.540371894836426,
      "learning_rate": 1.3077128205128205e-05,
      "loss": 0.732,
      "step": 180000
    },
    {
      "epoch": 2.2215384615384615,
      "grad_norm": 11.103569030761719,
      "learning_rate": 1.2974564102564102e-05,
      "loss": 0.7217,
      "step": 180500
    },
    {
      "epoch": 2.227692307692308,
      "grad_norm": 10.219368934631348,
      "learning_rate": 1.2872000000000001e-05,
      "loss": 0.7375,
      "step": 181000
    },
    {
      "epoch": 2.233846153846154,
      "grad_norm": 6.829085350036621,
      "learning_rate": 1.2769435897435896e-05,
      "loss": 0.7237,
      "step": 181500
    },
    {
      "epoch": 2.24,
      "grad_norm": 6.320499897003174,
      "learning_rate": 1.2666871794871796e-05,
      "loss": 0.7119,
      "step": 182000
    },
    {
      "epoch": 2.246153846153846,
      "grad_norm": 8.686861038208008,
      "learning_rate": 1.2564307692307692e-05,
      "loss": 0.724,
      "step": 182500
    },
    {
      "epoch": 2.252307692307692,
      "grad_norm": 6.650514125823975,
      "learning_rate": 1.246174358974359e-05,
      "loss": 0.7153,
      "step": 183000
    },
    {
      "epoch": 2.2584615384615385,
      "grad_norm": 5.943713188171387,
      "learning_rate": 1.2359179487179488e-05,
      "loss": 0.7315,
      "step": 183500
    },
    {
      "epoch": 2.2646153846153845,
      "grad_norm": 11.34069538116455,
      "learning_rate": 1.2256615384615386e-05,
      "loss": 0.7062,
      "step": 184000
    },
    {
      "epoch": 2.270769230769231,
      "grad_norm": 180.521240234375,
      "learning_rate": 1.2154051282051282e-05,
      "loss": 0.7384,
      "step": 184500
    },
    {
      "epoch": 2.276923076923077,
      "grad_norm": 8.282730102539062,
      "learning_rate": 1.205148717948718e-05,
      "loss": 0.7173,
      "step": 185000
    },
    {
      "epoch": 2.2830769230769232,
      "grad_norm": 12.209334373474121,
      "learning_rate": 1.1948923076923077e-05,
      "loss": 0.7165,
      "step": 185500
    },
    {
      "epoch": 2.289230769230769,
      "grad_norm": 9.707290649414062,
      "learning_rate": 1.1846358974358975e-05,
      "loss": 0.7426,
      "step": 186000
    },
    {
      "epoch": 2.2953846153846156,
      "grad_norm": 26.727867126464844,
      "learning_rate": 1.1743794871794872e-05,
      "loss": 0.7393,
      "step": 186500
    },
    {
      "epoch": 2.3015384615384615,
      "grad_norm": 11.103117942810059,
      "learning_rate": 1.1641230769230769e-05,
      "loss": 0.7199,
      "step": 187000
    },
    {
      "epoch": 2.3076923076923075,
      "grad_norm": 15.823190689086914,
      "learning_rate": 1.1538666666666667e-05,
      "loss": 0.7199,
      "step": 187500
    },
    {
      "epoch": 2.313846153846154,
      "grad_norm": 5.7931671142578125,
      "learning_rate": 1.1436102564102565e-05,
      "loss": 0.7331,
      "step": 188000
    },
    {
      "epoch": 2.32,
      "grad_norm": 3.72764253616333,
      "learning_rate": 1.1333538461538463e-05,
      "loss": 0.722,
      "step": 188500
    },
    {
      "epoch": 2.3261538461538462,
      "grad_norm": 7.738234996795654,
      "learning_rate": 1.1230974358974359e-05,
      "loss": 0.7356,
      "step": 189000
    },
    {
      "epoch": 2.332307692307692,
      "grad_norm": 12.795975685119629,
      "learning_rate": 1.1128410256410257e-05,
      "loss": 0.7152,
      "step": 189500
    },
    {
      "epoch": 2.3384615384615386,
      "grad_norm": 13.906537055969238,
      "learning_rate": 1.1025846153846155e-05,
      "loss": 0.728,
      "step": 190000
    },
    {
      "epoch": 2.3446153846153845,
      "grad_norm": 17.184505462646484,
      "learning_rate": 1.0923282051282053e-05,
      "loss": 0.7054,
      "step": 190500
    },
    {
      "epoch": 2.350769230769231,
      "grad_norm": 9.99722671508789,
      "learning_rate": 1.082071794871795e-05,
      "loss": 0.7325,
      "step": 191000
    },
    {
      "epoch": 2.356923076923077,
      "grad_norm": 8.681159973144531,
      "learning_rate": 1.0718153846153847e-05,
      "loss": 0.7369,
      "step": 191500
    },
    {
      "epoch": 2.363076923076923,
      "grad_norm": 8.222824096679688,
      "learning_rate": 1.0615589743589745e-05,
      "loss": 0.7494,
      "step": 192000
    },
    {
      "epoch": 2.3692307692307693,
      "grad_norm": 9.75273609161377,
      "learning_rate": 1.0513025641025642e-05,
      "loss": 0.7219,
      "step": 192500
    },
    {
      "epoch": 2.375384615384615,
      "grad_norm": 20.542802810668945,
      "learning_rate": 1.0410461538461538e-05,
      "loss": 0.7268,
      "step": 193000
    },
    {
      "epoch": 2.3815384615384616,
      "grad_norm": 16.02739906311035,
      "learning_rate": 1.0307897435897436e-05,
      "loss": 0.7155,
      "step": 193500
    },
    {
      "epoch": 2.3876923076923076,
      "grad_norm": 11.635347366333008,
      "learning_rate": 1.0205333333333334e-05,
      "loss": 0.7333,
      "step": 194000
    },
    {
      "epoch": 2.393846153846154,
      "grad_norm": 16.570072174072266,
      "learning_rate": 1.0102769230769232e-05,
      "loss": 0.7217,
      "step": 194500
    },
    {
      "epoch": 2.4,
      "grad_norm": 14.55761432647705,
      "learning_rate": 1.0000205128205128e-05,
      "loss": 0.7223,
      "step": 195000
    },
    {
      "epoch": 2.4061538461538463,
      "grad_norm": 10.24792194366455,
      "learning_rate": 9.897641025641026e-06,
      "loss": 0.7337,
      "step": 195500
    },
    {
      "epoch": 2.4123076923076923,
      "grad_norm": 11.387216567993164,
      "learning_rate": 9.795076923076924e-06,
      "loss": 0.7156,
      "step": 196000
    },
    {
      "epoch": 2.418461538461538,
      "grad_norm": 6.960866928100586,
      "learning_rate": 9.69251282051282e-06,
      "loss": 0.7199,
      "step": 196500
    },
    {
      "epoch": 2.4246153846153846,
      "grad_norm": 25.213428497314453,
      "learning_rate": 9.589948717948718e-06,
      "loss": 0.7166,
      "step": 197000
    },
    {
      "epoch": 2.430769230769231,
      "grad_norm": 16.338205337524414,
      "learning_rate": 9.487384615384616e-06,
      "loss": 0.7164,
      "step": 197500
    },
    {
      "epoch": 2.436923076923077,
      "grad_norm": 9.887063980102539,
      "learning_rate": 9.384820512820514e-06,
      "loss": 0.7242,
      "step": 198000
    },
    {
      "epoch": 2.443076923076923,
      "grad_norm": 15.347149848937988,
      "learning_rate": 9.28225641025641e-06,
      "loss": 0.7354,
      "step": 198500
    },
    {
      "epoch": 2.4492307692307693,
      "grad_norm": 11.793184280395508,
      "learning_rate": 9.179692307692309e-06,
      "loss": 0.7176,
      "step": 199000
    },
    {
      "epoch": 2.4553846153846153,
      "grad_norm": 6.337645053863525,
      "learning_rate": 9.077128205128207e-06,
      "loss": 0.713,
      "step": 199500
    },
    {
      "epoch": 2.4615384615384617,
      "grad_norm": 8.685958862304688,
      "learning_rate": 8.974564102564103e-06,
      "loss": 0.7392,
      "step": 200000
    },
    {
      "epoch": 2.4676923076923076,
      "grad_norm": 8.590500831604004,
      "learning_rate": 8.871999999999999e-06,
      "loss": 0.7217,
      "step": 200500
    },
    {
      "epoch": 2.473846153846154,
      "grad_norm": 16.368959426879883,
      "learning_rate": 8.769435897435897e-06,
      "loss": 0.7245,
      "step": 201000
    },
    {
      "epoch": 2.48,
      "grad_norm": 6.3712544441223145,
      "learning_rate": 8.666871794871795e-06,
      "loss": 0.7305,
      "step": 201500
    },
    {
      "epoch": 2.4861538461538464,
      "grad_norm": 13.746201515197754,
      "learning_rate": 8.564307692307693e-06,
      "loss": 0.7163,
      "step": 202000
    },
    {
      "epoch": 2.4923076923076923,
      "grad_norm": 7.711772441864014,
      "learning_rate": 8.46174358974359e-06,
      "loss": 0.6961,
      "step": 202500
    },
    {
      "epoch": 2.4984615384615383,
      "grad_norm": 12.77273941040039,
      "learning_rate": 8.359179487179487e-06,
      "loss": 0.7267,
      "step": 203000
    },
    {
      "epoch": 2.5046153846153847,
      "grad_norm": 10.931082725524902,
      "learning_rate": 8.256615384615385e-06,
      "loss": 0.7078,
      "step": 203500
    },
    {
      "epoch": 2.5107692307692306,
      "grad_norm": 16.658597946166992,
      "learning_rate": 8.154051282051282e-06,
      "loss": 0.7072,
      "step": 204000
    },
    {
      "epoch": 2.516923076923077,
      "grad_norm": 8.216838836669922,
      "learning_rate": 8.05148717948718e-06,
      "loss": 0.7244,
      "step": 204500
    },
    {
      "epoch": 2.523076923076923,
      "grad_norm": 15.0646333694458,
      "learning_rate": 7.948923076923078e-06,
      "loss": 0.7186,
      "step": 205000
    },
    {
      "epoch": 2.5292307692307694,
      "grad_norm": 6.978567600250244,
      "learning_rate": 7.846358974358976e-06,
      "loss": 0.7263,
      "step": 205500
    },
    {
      "epoch": 2.5353846153846153,
      "grad_norm": 12.0398530960083,
      "learning_rate": 7.743794871794872e-06,
      "loss": 0.715,
      "step": 206000
    },
    {
      "epoch": 2.5415384615384617,
      "grad_norm": 54.27073669433594,
      "learning_rate": 7.64123076923077e-06,
      "loss": 0.7139,
      "step": 206500
    },
    {
      "epoch": 2.5476923076923077,
      "grad_norm": 10.282230377197266,
      "learning_rate": 7.538666666666668e-06,
      "loss": 0.7304,
      "step": 207000
    },
    {
      "epoch": 2.5538461538461537,
      "grad_norm": 10.705265045166016,
      "learning_rate": 7.436102564102565e-06,
      "loss": 0.737,
      "step": 207500
    },
    {
      "epoch": 2.56,
      "grad_norm": 19.464488983154297,
      "learning_rate": 7.333538461538463e-06,
      "loss": 0.696,
      "step": 208000
    },
    {
      "epoch": 2.566153846153846,
      "grad_norm": 15.14621639251709,
      "learning_rate": 7.230974358974358e-06,
      "loss": 0.7315,
      "step": 208500
    },
    {
      "epoch": 2.5723076923076924,
      "grad_norm": 15.028178215026855,
      "learning_rate": 7.128410256410256e-06,
      "loss": 0.7099,
      "step": 209000
    },
    {
      "epoch": 2.5784615384615384,
      "grad_norm": 11.935062408447266,
      "learning_rate": 7.0258461538461535e-06,
      "loss": 0.7224,
      "step": 209500
    },
    {
      "epoch": 2.5846153846153848,
      "grad_norm": 9.3062105178833,
      "learning_rate": 6.9232820512820515e-06,
      "loss": 0.7366,
      "step": 210000
    },
    {
      "epoch": 2.5907692307692307,
      "grad_norm": 15.54145622253418,
      "learning_rate": 6.820717948717949e-06,
      "loss": 0.7005,
      "step": 210500
    },
    {
      "epoch": 2.596923076923077,
      "grad_norm": 20.382858276367188,
      "learning_rate": 6.7181538461538466e-06,
      "loss": 0.719,
      "step": 211000
    },
    {
      "epoch": 2.603076923076923,
      "grad_norm": 12.054417610168457,
      "learning_rate": 6.615589743589744e-06,
      "loss": 0.6902,
      "step": 211500
    },
    {
      "epoch": 2.609230769230769,
      "grad_norm": 18.24418067932129,
      "learning_rate": 6.513025641025642e-06,
      "loss": 0.714,
      "step": 212000
    },
    {
      "epoch": 2.6153846153846154,
      "grad_norm": 7.818421363830566,
      "learning_rate": 6.410461538461539e-06,
      "loss": 0.7108,
      "step": 212500
    },
    {
      "epoch": 2.6215384615384614,
      "grad_norm": 11.045573234558105,
      "learning_rate": 6.307897435897436e-06,
      "loss": 0.7189,
      "step": 213000
    },
    {
      "epoch": 2.6276923076923078,
      "grad_norm": 3.8956141471862793,
      "learning_rate": 6.205333333333334e-06,
      "loss": 0.7306,
      "step": 213500
    },
    {
      "epoch": 2.6338461538461537,
      "grad_norm": 6.8342061042785645,
      "learning_rate": 6.102769230769231e-06,
      "loss": 0.705,
      "step": 214000
    },
    {
      "epoch": 2.64,
      "grad_norm": 3.917092800140381,
      "learning_rate": 6.000205128205128e-06,
      "loss": 0.7128,
      "step": 214500
    },
    {
      "epoch": 2.646153846153846,
      "grad_norm": 10.602785110473633,
      "learning_rate": 5.897641025641025e-06,
      "loss": 0.7178,
      "step": 215000
    },
    {
      "epoch": 2.6523076923076925,
      "grad_norm": 44.85240173339844,
      "learning_rate": 5.795076923076923e-06,
      "loss": 0.7124,
      "step": 215500
    },
    {
      "epoch": 2.6584615384615384,
      "grad_norm": 6.232956886291504,
      "learning_rate": 5.6925128205128205e-06,
      "loss": 0.7235,
      "step": 216000
    },
    {
      "epoch": 2.6646153846153844,
      "grad_norm": 15.6179780960083,
      "learning_rate": 5.5899487179487185e-06,
      "loss": 0.7367,
      "step": 216500
    },
    {
      "epoch": 2.670769230769231,
      "grad_norm": 28.499755859375,
      "learning_rate": 5.487384615384616e-06,
      "loss": 0.7122,
      "step": 217000
    },
    {
      "epoch": 2.676923076923077,
      "grad_norm": 7.36679220199585,
      "learning_rate": 5.384820512820514e-06,
      "loss": 0.7083,
      "step": 217500
    },
    {
      "epoch": 2.683076923076923,
      "grad_norm": 7.019028663635254,
      "learning_rate": 5.282256410256411e-06,
      "loss": 0.7167,
      "step": 218000
    },
    {
      "epoch": 2.689230769230769,
      "grad_norm": 13.332535743713379,
      "learning_rate": 5.179692307692308e-06,
      "loss": 0.6813,
      "step": 218500
    },
    {
      "epoch": 2.6953846153846155,
      "grad_norm": 6.092926025390625,
      "learning_rate": 5.077128205128205e-06,
      "loss": 0.7095,
      "step": 219000
    },
    {
      "epoch": 2.7015384615384614,
      "grad_norm": 5.190014362335205,
      "learning_rate": 4.974564102564103e-06,
      "loss": 0.6941,
      "step": 219500
    },
    {
      "epoch": 2.707692307692308,
      "grad_norm": 6.392735004425049,
      "learning_rate": 4.872e-06,
      "loss": 0.7114,
      "step": 220000
    },
    {
      "epoch": 2.713846153846154,
      "grad_norm": 18.07638931274414,
      "learning_rate": 4.769435897435898e-06,
      "loss": 0.7115,
      "step": 220500
    },
    {
      "epoch": 2.7199999999999998,
      "grad_norm": 6.718528747558594,
      "learning_rate": 4.666871794871795e-06,
      "loss": 0.7065,
      "step": 221000
    },
    {
      "epoch": 2.726153846153846,
      "grad_norm": 17.323354721069336,
      "learning_rate": 4.564307692307692e-06,
      "loss": 0.701,
      "step": 221500
    },
    {
      "epoch": 2.7323076923076925,
      "grad_norm": 11.844677925109863,
      "learning_rate": 4.46174358974359e-06,
      "loss": 0.7183,
      "step": 222000
    },
    {
      "epoch": 2.7384615384615385,
      "grad_norm": 17.19860076904297,
      "learning_rate": 4.3591794871794875e-06,
      "loss": 0.6908,
      "step": 222500
    },
    {
      "epoch": 2.7446153846153845,
      "grad_norm": 14.058614730834961,
      "learning_rate": 4.256615384615385e-06,
      "loss": 0.721,
      "step": 223000
    },
    {
      "epoch": 2.750769230769231,
      "grad_norm": 15.834685325622559,
      "learning_rate": 4.154051282051282e-06,
      "loss": 0.7267,
      "step": 223500
    },
    {
      "epoch": 2.756923076923077,
      "grad_norm": 7.217898845672607,
      "learning_rate": 4.05148717948718e-06,
      "loss": 0.7047,
      "step": 224000
    },
    {
      "epoch": 2.763076923076923,
      "grad_norm": 8.565190315246582,
      "learning_rate": 3.948923076923077e-06,
      "loss": 0.6915,
      "step": 224500
    },
    {
      "epoch": 2.769230769230769,
      "grad_norm": 6.846717357635498,
      "learning_rate": 3.846358974358975e-06,
      "loss": 0.7139,
      "step": 225000
    },
    {
      "epoch": 2.775384615384615,
      "grad_norm": 10.877586364746094,
      "learning_rate": 3.743794871794872e-06,
      "loss": 0.7226,
      "step": 225500
    },
    {
      "epoch": 2.7815384615384615,
      "grad_norm": 7.898690700531006,
      "learning_rate": 3.6412307692307696e-06,
      "loss": 0.7001,
      "step": 226000
    },
    {
      "epoch": 2.787692307692308,
      "grad_norm": 11.410025596618652,
      "learning_rate": 3.5386666666666667e-06,
      "loss": 0.6996,
      "step": 226500
    },
    {
      "epoch": 2.793846153846154,
      "grad_norm": 19.853111267089844,
      "learning_rate": 3.436102564102564e-06,
      "loss": 0.7054,
      "step": 227000
    },
    {
      "epoch": 2.8,
      "grad_norm": 11.761764526367188,
      "learning_rate": 3.3335384615384614e-06,
      "loss": 0.721,
      "step": 227500
    },
    {
      "epoch": 2.806153846153846,
      "grad_norm": 18.024303436279297,
      "learning_rate": 3.230974358974359e-06,
      "loss": 0.7255,
      "step": 228000
    },
    {
      "epoch": 2.812307692307692,
      "grad_norm": 15.604559898376465,
      "learning_rate": 3.1284102564102565e-06,
      "loss": 0.7013,
      "step": 228500
    },
    {
      "epoch": 2.8184615384615386,
      "grad_norm": 11.40152359008789,
      "learning_rate": 3.025846153846154e-06,
      "loss": 0.7042,
      "step": 229000
    },
    {
      "epoch": 2.8246153846153845,
      "grad_norm": 21.253713607788086,
      "learning_rate": 2.923282051282051e-06,
      "loss": 0.7187,
      "step": 229500
    },
    {
      "epoch": 2.830769230769231,
      "grad_norm": 10.321839332580566,
      "learning_rate": 2.8207179487179488e-06,
      "loss": 0.7111,
      "step": 230000
    },
    {
      "epoch": 2.836923076923077,
      "grad_norm": 9.331689834594727,
      "learning_rate": 2.7181538461538463e-06,
      "loss": 0.6974,
      "step": 230500
    },
    {
      "epoch": 2.8430769230769233,
      "grad_norm": 14.935019493103027,
      "learning_rate": 2.615589743589744e-06,
      "loss": 0.6993,
      "step": 231000
    },
    {
      "epoch": 2.8492307692307692,
      "grad_norm": 5.596281051635742,
      "learning_rate": 2.513025641025641e-06,
      "loss": 0.7184,
      "step": 231500
    },
    {
      "epoch": 2.855384615384615,
      "grad_norm": 8.525458335876465,
      "learning_rate": 2.4104615384615386e-06,
      "loss": 0.7073,
      "step": 232000
    },
    {
      "epoch": 2.8615384615384616,
      "grad_norm": 8.31210994720459,
      "learning_rate": 2.307897435897436e-06,
      "loss": 0.7043,
      "step": 232500
    },
    {
      "epoch": 2.8676923076923075,
      "grad_norm": 5.1119608879089355,
      "learning_rate": 2.2053333333333337e-06,
      "loss": 0.6754,
      "step": 233000
    },
    {
      "epoch": 2.873846153846154,
      "grad_norm": 11.781253814697266,
      "learning_rate": 2.102769230769231e-06,
      "loss": 0.7187,
      "step": 233500
    },
    {
      "epoch": 2.88,
      "grad_norm": 6.548630237579346,
      "learning_rate": 2.0002051282051284e-06,
      "loss": 0.7199,
      "step": 234000
    },
    {
      "epoch": 2.8861538461538463,
      "grad_norm": 12.736579895019531,
      "learning_rate": 1.8976410256410257e-06,
      "loss": 0.7146,
      "step": 234500
    },
    {
      "epoch": 2.8923076923076922,
      "grad_norm": 7.0301690101623535,
      "learning_rate": 1.7950769230769233e-06,
      "loss": 0.7299,
      "step": 235000
    },
    {
      "epoch": 2.8984615384615386,
      "grad_norm": 14.406909942626953,
      "learning_rate": 1.6925128205128204e-06,
      "loss": 0.7021,
      "step": 235500
    },
    {
      "epoch": 2.9046153846153846,
      "grad_norm": 4.011630535125732,
      "learning_rate": 1.589948717948718e-06,
      "loss": 0.7041,
      "step": 236000
    },
    {
      "epoch": 2.9107692307692306,
      "grad_norm": 3.6570491790771484,
      "learning_rate": 1.4873846153846156e-06,
      "loss": 0.7047,
      "step": 236500
    },
    {
      "epoch": 2.916923076923077,
      "grad_norm": 10.598403930664062,
      "learning_rate": 1.384820512820513e-06,
      "loss": 0.7058,
      "step": 237000
    },
    {
      "epoch": 2.9230769230769234,
      "grad_norm": 17.667306900024414,
      "learning_rate": 1.2822564102564103e-06,
      "loss": 0.6842,
      "step": 237500
    },
    {
      "epoch": 2.9292307692307693,
      "grad_norm": 4.180660724639893,
      "learning_rate": 1.1796923076923076e-06,
      "loss": 0.7185,
      "step": 238000
    },
    {
      "epoch": 2.9353846153846153,
      "grad_norm": 9.529131889343262,
      "learning_rate": 1.0771282051282052e-06,
      "loss": 0.7073,
      "step": 238500
    },
    {
      "epoch": 2.9415384615384617,
      "grad_norm": 8.090576171875,
      "learning_rate": 9.745641025641025e-07,
      "loss": 0.6974,
      "step": 239000
    },
    {
      "epoch": 2.9476923076923076,
      "grad_norm": 14.463215827941895,
      "learning_rate": 8.720000000000001e-07,
      "loss": 0.6877,
      "step": 239500
    },
    {
      "epoch": 2.953846153846154,
      "grad_norm": 9.106619834899902,
      "learning_rate": 7.694358974358975e-07,
      "loss": 0.6888,
      "step": 240000
    },
    {
      "epoch": 2.96,
      "grad_norm": 10.116360664367676,
      "learning_rate": 6.66871794871795e-07,
      "loss": 0.6956,
      "step": 240500
    },
    {
      "epoch": 2.966153846153846,
      "grad_norm": 6.25929594039917,
      "learning_rate": 5.643076923076923e-07,
      "loss": 0.734,
      "step": 241000
    },
    {
      "epoch": 2.9723076923076923,
      "grad_norm": 6.0323309898376465,
      "learning_rate": 4.617435897435898e-07,
      "loss": 0.7066,
      "step": 241500
    },
    {
      "epoch": 2.9784615384615387,
      "grad_norm": 16.856470108032227,
      "learning_rate": 3.591794871794872e-07,
      "loss": 0.7091,
      "step": 242000
    },
    {
      "epoch": 2.9846153846153847,
      "grad_norm": 10.947349548339844,
      "learning_rate": 2.566153846153846e-07,
      "loss": 0.6909,
      "step": 242500
    },
    {
      "epoch": 2.9907692307692306,
      "grad_norm": 5.442905426025391,
      "learning_rate": 1.5405128205128205e-07,
      "loss": 0.7271,
      "step": 243000
    },
    {
      "epoch": 2.996923076923077,
      "grad_norm": 7.611256122589111,
      "learning_rate": 5.148717948717949e-08,
      "loss": 0.7146,
      "step": 243500
    }
  ],
  "logging_steps": 500,
  "max_steps": 243750,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 5000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 5.130803778048e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}