{
  "best_global_step": 5000,
  "best_metric": 1.1470744609832764,
  "best_model_checkpoint": "/workspace/actual_run/checkpoint-5000",
  "epoch": 0.6012506012506013,
  "eval_steps": 1000,
  "global_step": 5000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0012025012025012026,
      "grad_norm": 1.0899138450622559,
      "learning_rate": 3.6e-06,
      "loss": 2.3927,
      "step": 10
    },
    {
      "epoch": 0.002405002405002405,
      "grad_norm": 0.9620404839515686,
      "learning_rate": 7.6e-06,
      "loss": 2.2959,
      "step": 20
    },
    {
      "epoch": 0.0036075036075036075,
      "grad_norm": 0.43678194284439087,
      "learning_rate": 1.16e-05,
      "loss": 2.036,
      "step": 30
    },
    {
      "epoch": 0.00481000481000481,
      "grad_norm": 0.35978421568870544,
      "learning_rate": 1.56e-05,
      "loss": 1.898,
      "step": 40
    },
    {
      "epoch": 0.006012506012506013,
      "grad_norm": 0.2896435856819153,
      "learning_rate": 1.9600000000000002e-05,
      "loss": 1.7747,
      "step": 50
    },
    {
      "epoch": 0.007215007215007215,
      "grad_norm": 0.2754799425601959,
      "learning_rate": 2.36e-05,
      "loss": 1.6628,
      "step": 60
    },
    {
      "epoch": 0.008417508417508417,
      "grad_norm": 0.2863922119140625,
      "learning_rate": 2.7600000000000003e-05,
      "loss": 1.6091,
      "step": 70
    },
    {
      "epoch": 0.00962000962000962,
      "grad_norm": 0.29406264424324036,
      "learning_rate": 3.16e-05,
      "loss": 1.5302,
      "step": 80
    },
    {
      "epoch": 0.010822510822510822,
      "grad_norm": 0.32185423374176025,
      "learning_rate": 3.56e-05,
      "loss": 1.498,
      "step": 90
    },
    {
      "epoch": 0.012025012025012025,
      "grad_norm": 0.34314361214637756,
      "learning_rate": 3.960000000000001e-05,
      "loss": 1.4561,
      "step": 100
    },
    {
      "epoch": 0.013227513227513227,
      "grad_norm": 0.3503110706806183,
      "learning_rate": 4.36e-05,
      "loss": 1.4569,
      "step": 110
    },
    {
      "epoch": 0.01443001443001443,
      "grad_norm": 0.3761644661426544,
      "learning_rate": 4.76e-05,
      "loss": 1.4043,
      "step": 120
    },
    {
      "epoch": 0.015632515632515633,
      "grad_norm": 0.3741130530834198,
      "learning_rate": 5.16e-05,
      "loss": 1.4147,
      "step": 130
    },
    {
      "epoch": 0.016835016835016835,
      "grad_norm": 0.381028950214386,
      "learning_rate": 5.560000000000001e-05,
      "loss": 1.4226,
      "step": 140
    },
    {
      "epoch": 0.018037518037518036,
      "grad_norm": 0.3512011468410492,
      "learning_rate": 5.96e-05,
      "loss": 1.4034,
      "step": 150
    },
    {
      "epoch": 0.01924001924001924,
      "grad_norm": 0.33474254608154297,
      "learning_rate": 6.36e-05,
      "loss": 1.3853,
      "step": 160
    },
    {
      "epoch": 0.020442520442520443,
      "grad_norm": 0.335902601480484,
      "learning_rate": 6.76e-05,
      "loss": 1.3865,
      "step": 170
    },
    {
      "epoch": 0.021645021645021644,
      "grad_norm": 0.31389617919921875,
      "learning_rate": 7.16e-05,
      "loss": 1.3739,
      "step": 180
    },
    {
      "epoch": 0.02284752284752285,
      "grad_norm": 0.3155975043773651,
      "learning_rate": 7.560000000000001e-05,
      "loss": 1.3713,
      "step": 190
    },
    {
      "epoch": 0.02405002405002405,
      "grad_norm": 0.31867676973342896,
      "learning_rate": 7.960000000000001e-05,
      "loss": 1.3689,
      "step": 200
    },
    {
      "epoch": 0.025252525252525252,
      "grad_norm": 0.2963453531265259,
      "learning_rate": 8.36e-05,
      "loss": 1.3701,
      "step": 210
    },
    {
      "epoch": 0.026455026455026454,
      "grad_norm": 0.3254355788230896,
      "learning_rate": 8.76e-05,
      "loss": 1.3618,
      "step": 220
    },
    {
      "epoch": 0.02765752765752766,
      "grad_norm": 0.31009721755981445,
      "learning_rate": 9.16e-05,
      "loss": 1.3562,
      "step": 230
    },
    {
      "epoch": 0.02886002886002886,
      "grad_norm": 0.2769719958305359,
      "learning_rate": 9.56e-05,
      "loss": 1.341,
      "step": 240
    },
    {
      "epoch": 0.03006253006253006,
      "grad_norm": 0.29058948159217834,
      "learning_rate": 9.960000000000001e-05,
      "loss": 1.3509,
      "step": 250
    },
    {
      "epoch": 0.031265031265031266,
      "grad_norm": 0.2969858646392822,
      "learning_rate": 9.999969280942074e-05,
      "loss": 1.3533,
      "step": 260
    },
    {
      "epoch": 0.032467532467532464,
      "grad_norm": 0.27514412999153137,
      "learning_rate": 9.999863092090635e-05,
      "loss": 1.3382,
      "step": 270
    },
    {
      "epoch": 0.03367003367003367,
      "grad_norm": 0.2757578492164612,
      "learning_rate": 9.99968105580855e-05,
      "loss": 1.3302,
      "step": 280
    },
    {
      "epoch": 0.034872534872534874,
      "grad_norm": 0.2729010283946991,
      "learning_rate": 9.999423174857289e-05,
      "loss": 1.3325,
      "step": 290
    },
    {
      "epoch": 0.03607503607503607,
      "grad_norm": 0.26693403720855713,
      "learning_rate": 9.999089453148882e-05,
      "loss": 1.3318,
      "step": 300
    },
    {
      "epoch": 0.03727753727753728,
      "grad_norm": 0.26203256845474243,
      "learning_rate": 9.998679895745853e-05,
      "loss": 1.3308,
      "step": 310
    },
    {
      "epoch": 0.03848003848003848,
      "grad_norm": 0.24240988492965698,
      "learning_rate": 9.998194508861147e-05,
      "loss": 1.3211,
      "step": 320
    },
    {
      "epoch": 0.03968253968253968,
      "grad_norm": 0.23748330771923065,
      "learning_rate": 9.997633299858036e-05,
      "loss": 1.317,
      "step": 330
    },
    {
      "epoch": 0.040885040885040885,
      "grad_norm": 0.27641433477401733,
      "learning_rate": 9.996996277250007e-05,
      "loss": 1.3049,
      "step": 340
    },
    {
      "epoch": 0.04208754208754209,
      "grad_norm": 0.261136531829834,
      "learning_rate": 9.996283450700628e-05,
      "loss": 1.3099,
      "step": 350
    },
    {
      "epoch": 0.04329004329004329,
      "grad_norm": 0.26694369316101074,
      "learning_rate": 9.995494831023409e-05,
      "loss": 1.3064,
      "step": 360
    },
    {
      "epoch": 0.04449254449254449,
      "grad_norm": 0.2695152759552002,
      "learning_rate": 9.994630430181631e-05,
      "loss": 1.3085,
      "step": 370
    },
    {
      "epoch": 0.0456950456950457,
      "grad_norm": 0.26653772592544556,
      "learning_rate": 9.99369026128817e-05,
      "loss": 1.3184,
      "step": 380
    },
    {
      "epoch": 0.046897546897546896,
      "grad_norm": 0.25345709919929504,
      "learning_rate": 9.992674338605297e-05,
      "loss": 1.2965,
      "step": 390
    },
    {
      "epoch": 0.0481000481000481,
      "grad_norm": 0.2470959722995758,
      "learning_rate": 9.991582677544458e-05,
      "loss": 1.29,
      "step": 400
    },
    {
      "epoch": 0.0493025493025493,
      "grad_norm": 0.2398555725812912,
      "learning_rate": 9.990415294666045e-05,
      "loss": 1.2991,
      "step": 410
    },
    {
      "epoch": 0.050505050505050504,
      "grad_norm": 0.24114194512367249,
      "learning_rate": 9.98917220767914e-05,
      "loss": 1.294,
      "step": 420
    },
    {
      "epoch": 0.05170755170755171,
      "grad_norm": 0.24480991065502167,
      "learning_rate": 9.98785343544125e-05,
      "loss": 1.2827,
      "step": 430
    },
    {
      "epoch": 0.05291005291005291,
      "grad_norm": 0.2351623922586441,
      "learning_rate": 9.986458997958021e-05,
      "loss": 1.2868,
      "step": 440
    },
    {
      "epoch": 0.05411255411255411,
      "grad_norm": 0.23396646976470947,
      "learning_rate": 9.984988916382932e-05,
      "loss": 1.2894,
      "step": 450
    },
    {
      "epoch": 0.05531505531505532,
      "grad_norm": 0.24305066466331482,
      "learning_rate": 9.983443213016975e-05,
      "loss": 1.2929,
      "step": 460
    },
    {
      "epoch": 0.056517556517556515,
      "grad_norm": 0.24993565678596497,
      "learning_rate": 9.981821911308316e-05,
      "loss": 1.2809,
      "step": 470
    },
    {
      "epoch": 0.05772005772005772,
      "grad_norm": 0.24196045100688934,
      "learning_rate": 9.980125035851945e-05,
      "loss": 1.2771,
      "step": 480
    },
    {
      "epoch": 0.058922558922558925,
      "grad_norm": 0.2293768674135208,
      "learning_rate": 9.978352612389298e-05,
      "loss": 1.2899,
      "step": 490
    },
    {
      "epoch": 0.06012506012506012,
      "grad_norm": 0.22991864383220673,
      "learning_rate": 9.97650466780786e-05,
      "loss": 1.2813,
      "step": 500
    },
    {
      "epoch": 0.06132756132756133,
      "grad_norm": 0.24029232561588287,
      "learning_rate": 9.97458123014077e-05,
      "loss": 1.2816,
      "step": 510
    },
    {
      "epoch": 0.06253006253006253,
      "grad_norm": 0.2242475003004074,
      "learning_rate": 9.972582328566387e-05,
      "loss": 1.2859,
      "step": 520
    },
    {
      "epoch": 0.06373256373256374,
      "grad_norm": 0.2273736447095871,
      "learning_rate": 9.97050799340785e-05,
      "loss": 1.2743,
      "step": 530
    },
    {
      "epoch": 0.06493506493506493,
      "grad_norm": 0.23436114192008972,
      "learning_rate": 9.968358256132623e-05,
      "loss": 1.2841,
      "step": 540
    },
    {
      "epoch": 0.06613756613756613,
      "grad_norm": 0.22182011604309082,
      "learning_rate": 9.966133149352006e-05,
      "loss": 1.2933,
      "step": 550
    },
    {
      "epoch": 0.06734006734006734,
      "grad_norm": 0.23829922080039978,
      "learning_rate": 9.963832706820649e-05,
      "loss": 1.2865,
      "step": 560
    },
    {
      "epoch": 0.06854256854256854,
      "grad_norm": 0.24821604788303375,
      "learning_rate": 9.96145696343604e-05,
      "loss": 1.2835,
      "step": 570
    },
    {
      "epoch": 0.06974506974506975,
      "grad_norm": 0.22477108240127563,
      "learning_rate": 9.959005955237972e-05,
      "loss": 1.2845,
      "step": 580
    },
    {
      "epoch": 0.07094757094757095,
      "grad_norm": 0.23542803525924683,
      "learning_rate": 9.956479719407998e-05,
      "loss": 1.287,
      "step": 590
    },
    {
      "epoch": 0.07215007215007214,
      "grad_norm": 0.22595302760601044,
      "learning_rate": 9.953878294268866e-05,
      "loss": 1.2762,
      "step": 600
    },
    {
      "epoch": 0.07335257335257335,
      "grad_norm": 0.21597643196582794,
      "learning_rate": 9.95120171928394e-05,
      "loss": 1.283,
      "step": 610
    },
    {
      "epoch": 0.07455507455507455,
      "grad_norm": 0.23060384392738342,
      "learning_rate": 9.948450035056599e-05,
      "loss": 1.2847,
      "step": 620
    },
    {
      "epoch": 0.07575757575757576,
      "grad_norm": 0.21899910271167755,
      "learning_rate": 9.945623283329621e-05,
      "loss": 1.2699,
      "step": 630
    },
    {
      "epoch": 0.07696007696007696,
      "grad_norm": 0.21934176981449127,
      "learning_rate": 9.942721506984551e-05,
      "loss": 1.274,
      "step": 640
    },
    {
      "epoch": 0.07816257816257816,
      "grad_norm": 0.22260645031929016,
      "learning_rate": 9.939744750041052e-05,
      "loss": 1.2653,
      "step": 650
    },
    {
      "epoch": 0.07936507936507936,
      "grad_norm": 0.22758053243160248,
      "learning_rate": 9.936693057656231e-05,
      "loss": 1.2712,
      "step": 660
    },
    {
      "epoch": 0.08056758056758057,
      "grad_norm": 0.21833918988704681,
      "learning_rate": 9.933566476123964e-05,
      "loss": 1.2631,
      "step": 670
    },
    {
      "epoch": 0.08177008177008177,
      "grad_norm": 0.2229488492012024,
      "learning_rate": 9.930365052874182e-05,
      "loss": 1.2632,
      "step": 680
    },
    {
      "epoch": 0.08297258297258298,
      "grad_norm": 0.2153736650943756,
      "learning_rate": 9.927088836472163e-05,
      "loss": 1.284,
      "step": 690
    },
    {
      "epoch": 0.08417508417508418,
      "grad_norm": 0.23731300234794617,
      "learning_rate": 9.923737876617784e-05,
      "loss": 1.2629,
      "step": 700
    },
    {
      "epoch": 0.08537758537758537,
      "grad_norm": 0.22687861323356628,
      "learning_rate": 9.920312224144779e-05,
      "loss": 1.2757,
      "step": 710
    },
    {
      "epoch": 0.08658008658008658,
      "grad_norm": 0.22836917638778687,
      "learning_rate": 9.916811931019956e-05,
      "loss": 1.2589,
      "step": 720
    },
    {
      "epoch": 0.08778258778258778,
      "grad_norm": 0.2113468050956726,
      "learning_rate": 9.91323705034242e-05,
      "loss": 1.2515,
      "step": 730
    },
    {
      "epoch": 0.08898508898508899,
      "grad_norm": 0.2195415198802948,
      "learning_rate": 9.909587636342756e-05,
      "loss": 1.2544,
      "step": 740
    },
    {
      "epoch": 0.09018759018759019,
      "grad_norm": 0.2246798574924469,
      "learning_rate": 9.905863744382217e-05,
      "loss": 1.2503,
      "step": 750
    },
    {
      "epoch": 0.0913900913900914,
      "grad_norm": 0.22869926691055298,
      "learning_rate": 9.902065430951874e-05,
      "loss": 1.2626,
      "step": 760
    },
    {
      "epoch": 0.09259259259259259,
      "grad_norm": 0.2191305309534073,
      "learning_rate": 9.898192753671774e-05,
      "loss": 1.2612,
      "step": 770
    },
    {
      "epoch": 0.09379509379509379,
      "grad_norm": 0.22004824876785278,
      "learning_rate": 9.894245771290042e-05,
      "loss": 1.2734,
      "step": 780
    },
    {
      "epoch": 0.094997594997595,
      "grad_norm": 0.22499217092990875,
      "learning_rate": 9.890224543682017e-05,
      "loss": 1.244,
      "step": 790
    },
    {
      "epoch": 0.0962000962000962,
      "grad_norm": 0.22316116094589233,
      "learning_rate": 9.886129131849324e-05,
      "loss": 1.2567,
      "step": 800
    },
    {
      "epoch": 0.09740259740259741,
      "grad_norm": 0.23548166453838348,
      "learning_rate": 9.88195959791896e-05,
      "loss": 1.2649,
      "step": 810
    },
    {
      "epoch": 0.0986050986050986,
      "grad_norm": 0.22244389355182648,
      "learning_rate": 9.877716005142339e-05,
      "loss": 1.2516,
      "step": 820
    },
    {
      "epoch": 0.0998075998075998,
      "grad_norm": 0.22191107273101807,
      "learning_rate": 9.873398417894348e-05,
      "loss": 1.2503,
      "step": 830
    },
    {
      "epoch": 0.10101010101010101,
      "grad_norm": 0.22165456414222717,
      "learning_rate": 9.869006901672362e-05,
      "loss": 1.2466,
      "step": 840
    },
    {
      "epoch": 0.10221260221260221,
      "grad_norm": 0.2198014259338379,
      "learning_rate": 9.864541523095247e-05,
      "loss": 1.246,
      "step": 850
    },
    {
      "epoch": 0.10341510341510342,
      "grad_norm": 0.22250819206237793,
      "learning_rate": 9.860002349902358e-05,
      "loss": 1.2362,
      "step": 860
    },
    {
      "epoch": 0.10461760461760462,
      "grad_norm": 0.2301010638475418,
      "learning_rate": 9.855389450952503e-05,
      "loss": 1.255,
      "step": 870
    },
    {
      "epoch": 0.10582010582010581,
      "grad_norm": 0.22151581943035126,
      "learning_rate": 9.850702896222907e-05,
      "loss": 1.2555,
      "step": 880
    },
    {
      "epoch": 0.10702260702260702,
      "grad_norm": 0.2172437161207199,
      "learning_rate": 9.845942756808145e-05,
      "loss": 1.2519,
      "step": 890
    },
    {
      "epoch": 0.10822510822510822,
      "grad_norm": 0.2380341738462448,
      "learning_rate": 9.841109104919057e-05,
      "loss": 1.2565,
      "step": 900
    },
    {
      "epoch": 0.10942760942760943,
      "grad_norm": 0.2219957560300827,
      "learning_rate": 9.836202013881673e-05,
      "loss": 1.2472,
      "step": 910
    },
    {
      "epoch": 0.11063011063011063,
      "grad_norm": 0.23652082681655884,
      "learning_rate": 9.831221558136075e-05,
      "loss": 1.249,
      "step": 920
    },
    {
      "epoch": 0.11183261183261184,
      "grad_norm": 0.21968142688274384,
      "learning_rate": 9.826167813235291e-05,
      "loss": 1.2451,
      "step": 930
    },
    {
      "epoch": 0.11303511303511303,
      "grad_norm": 0.22762618958950043,
      "learning_rate": 9.82104085584413e-05,
      "loss": 1.2433,
      "step": 940
    },
    {
      "epoch": 0.11423761423761424,
      "grad_norm": 0.20969653129577637,
      "learning_rate": 9.815840763738032e-05,
      "loss": 1.2452,
      "step": 950
    },
    {
      "epoch": 0.11544011544011544,
      "grad_norm": 0.22823800146579742,
      "learning_rate": 9.810567615801883e-05,
      "loss": 1.2482,
      "step": 960
    },
    {
      "epoch": 0.11664261664261664,
      "grad_norm": 0.22033190727233887,
      "learning_rate": 9.805221492028815e-05,
      "loss": 1.2466,
      "step": 970
    },
    {
      "epoch": 0.11784511784511785,
      "grad_norm": 0.22417309880256653,
      "learning_rate": 9.799802473519e-05,
      "loss": 1.2531,
      "step": 980
    },
    {
      "epoch": 0.11904761904761904,
      "grad_norm": 0.21632517874240875,
      "learning_rate": 9.794310642478418e-05,
      "loss": 1.2479,
      "step": 990
    },
    {
      "epoch": 0.12025012025012025,
      "grad_norm": 0.21755601465702057,
      "learning_rate": 9.788746082217601e-05,
      "loss": 1.2575,
      "step": 1000
    },
    {
      "epoch": 0.12025012025012025,
      "eval_loss": 1.2297765016555786,
      "eval_runtime": 17.3877,
      "eval_samples_per_second": 14.723,
      "eval_steps_per_second": 0.46,
      "step": 1000
    },
    {
      "epoch": 0.12145262145262145,
      "grad_norm": 0.23050642013549805,
      "learning_rate": 9.783108877150384e-05,
      "loss": 1.2438,
      "step": 1010
    },
    {
      "epoch": 0.12265512265512266,
      "grad_norm": 0.19902385771274567,
      "learning_rate": 9.77739911279261e-05,
      "loss": 1.2523,
      "step": 1020
    },
    {
      "epoch": 0.12385762385762386,
      "grad_norm": 0.27783408761024475,
      "learning_rate": 9.77161687576084e-05,
      "loss": 1.2591,
      "step": 1030
    },
    {
      "epoch": 0.12506012506012507,
      "grad_norm": 0.21347227692604065,
      "learning_rate": 9.765762253771045e-05,
      "loss": 1.2519,
      "step": 1040
    },
    {
      "epoch": 0.12626262626262627,
      "grad_norm": 0.209860697388649,
      "learning_rate": 9.75983533563726e-05,
      "loss": 1.229,
      "step": 1050
    },
    {
      "epoch": 0.12746512746512748,
      "grad_norm": 0.20719142258167267,
      "learning_rate": 9.753836211270254e-05,
      "loss": 1.2559,
      "step": 1060
    },
    {
      "epoch": 0.12866762866762868,
      "grad_norm": 0.22753511369228363,
      "learning_rate": 9.747764971676146e-05,
      "loss": 1.2511,
      "step": 1070
    },
    {
      "epoch": 0.12987012987012986,
      "grad_norm": 0.22132696211338043,
      "learning_rate": 9.741621708955048e-05,
      "loss": 1.2591,
      "step": 1080
    },
    {
      "epoch": 0.13107263107263106,
      "grad_norm": 0.22907616198062897,
      "learning_rate": 9.735406516299649e-05,
      "loss": 1.2483,
      "step": 1090
    },
    {
      "epoch": 0.13227513227513227,
      "grad_norm": 0.20939530432224274,
      "learning_rate": 9.729119487993809e-05,
      "loss": 1.2426,
      "step": 1100
    },
    {
      "epoch": 0.13347763347763347,
      "grad_norm": 0.22089892625808716,
      "learning_rate": 9.722760719411128e-05,
      "loss": 1.2527,
      "step": 1110
    },
    {
      "epoch": 0.13468013468013468,
      "grad_norm": 0.2165553718805313,
      "learning_rate": 9.716330307013499e-05,
      "loss": 1.2486,
      "step": 1120
    },
    {
      "epoch": 0.13588263588263588,
      "grad_norm": 0.22434702515602112,
      "learning_rate": 9.709828348349645e-05,
      "loss": 1.244,
      "step": 1130
    },
    {
      "epoch": 0.1370851370851371,
      "grad_norm": 0.22943449020385742,
      "learning_rate": 9.703254942053641e-05,
      "loss": 1.244,
      "step": 1140
    },
    {
      "epoch": 0.1382876382876383,
      "grad_norm": 0.22439205646514893,
      "learning_rate": 9.69661018784341e-05,
      "loss": 1.2324,
      "step": 1150
    },
    {
      "epoch": 0.1394901394901395,
      "grad_norm": 0.22962580621242523,
      "learning_rate": 9.689894186519217e-05,
      "loss": 1.2437,
      "step": 1160
    },
    {
      "epoch": 0.1406926406926407,
      "grad_norm": 0.22367720305919647,
      "learning_rate": 9.683107039962144e-05,
      "loss": 1.2353,
      "step": 1170
    },
    {
      "epoch": 0.1418951418951419,
      "grad_norm": 0.23102396726608276,
      "learning_rate": 9.676248851132536e-05,
      "loss": 1.224,
      "step": 1180
    },
    {
      "epoch": 0.14309764309764308,
      "grad_norm": 0.20691992342472076,
      "learning_rate": 9.669319724068441e-05,
      "loss": 1.2448,
      "step": 1190
    },
    {
      "epoch": 0.1443001443001443,
      "grad_norm": 0.2133265882730484,
      "learning_rate": 9.662319763884035e-05,
      "loss": 1.2562,
      "step": 1200
    },
    {
      "epoch": 0.1455026455026455,
      "grad_norm": 0.21801921725273132,
      "learning_rate": 9.655249076768023e-05,
      "loss": 1.2366,
      "step": 1210
    },
    {
      "epoch": 0.1467051467051467,
      "grad_norm": 0.22324581444263458,
      "learning_rate": 9.648107769982033e-05,
      "loss": 1.2341,
      "step": 1220
    },
    {
      "epoch": 0.1479076479076479,
      "grad_norm": 0.20417769253253937,
      "learning_rate": 9.640895951858984e-05,
      "loss": 1.2339,
      "step": 1230
    },
    {
      "epoch": 0.1491101491101491,
      "grad_norm": 0.21648766100406647,
      "learning_rate": 9.63361373180145e-05,
      "loss": 1.23,
      "step": 1240
    },
    {
      "epoch": 0.15031265031265031,
      "grad_norm": 0.21471014618873596,
      "learning_rate": 9.626261220279988e-05,
      "loss": 1.2445,
      "step": 1250
    },
    {
      "epoch": 0.15151515151515152,
      "grad_norm": 0.21205437183380127,
      "learning_rate": 9.618838528831477e-05,
      "loss": 1.2434,
      "step": 1260
    },
    {
      "epoch": 0.15271765271765272,
      "grad_norm": 0.20593945682048798,
      "learning_rate": 9.611345770057414e-05,
      "loss": 1.2367,
      "step": 1270
    },
    {
      "epoch": 0.15392015392015393,
      "grad_norm": 0.2065548151731491,
      "learning_rate": 9.60378305762221e-05,
      "loss": 1.2386,
      "step": 1280
    },
    {
      "epoch": 0.15512265512265513,
      "grad_norm": 0.21276754140853882,
      "learning_rate": 9.596150506251469e-05,
      "loss": 1.2358,
      "step": 1290
    },
    {
      "epoch": 0.1563251563251563,
      "grad_norm": 0.22176425158977509,
      "learning_rate": 9.588448231730245e-05,
      "loss": 1.235,
      "step": 1300
    },
    {
      "epoch": 0.15752765752765752,
      "grad_norm": 0.21053512394428253,
      "learning_rate": 9.580676350901282e-05,
      "loss": 1.2195,
      "step": 1310
    },
    {
      "epoch": 0.15873015873015872,
      "grad_norm": 0.22540676593780518,
      "learning_rate": 9.572834981663246e-05,
      "loss": 1.2313,
      "step": 1320
    },
    {
      "epoch": 0.15993265993265993,
      "grad_norm": 0.20599214732646942,
      "learning_rate": 9.56492424296894e-05,
      "loss": 1.2177,
      "step": 1330
    },
    {
      "epoch": 0.16113516113516113,
      "grad_norm": 0.2064780443906784,
      "learning_rate": 9.556944254823485e-05,
      "loss": 1.2278,
      "step": 1340
    },
    {
      "epoch": 0.16233766233766234,
      "grad_norm": 0.2161046266555786,
      "learning_rate": 9.548895138282521e-05,
      "loss": 1.2369,
      "step": 1350
    },
    {
      "epoch": 0.16354016354016354,
      "grad_norm": 0.2143697589635849,
      "learning_rate": 9.540777015450349e-05,
      "loss": 1.2163,
      "step": 1360
    },
    {
      "epoch": 0.16474266474266475,
      "grad_norm": 0.20479600131511688,
      "learning_rate": 9.532590009478094e-05,
      "loss": 1.2185,
      "step": 1370
    },
    {
      "epoch": 0.16594516594516595,
      "grad_norm": 0.2071349024772644,
      "learning_rate": 9.52433424456183e-05,
      "loss": 1.2449,
      "step": 1380
    },
    {
      "epoch": 0.16714766714766716,
      "grad_norm": 0.22588619589805603,
      "learning_rate": 9.516009845940697e-05,
      "loss": 1.2395,
      "step": 1390
    },
    {
      "epoch": 0.16835016835016836,
      "grad_norm": 0.21358510851860046,
      "learning_rate": 9.507616939895005e-05,
      "loss": 1.2259,
      "step": 1400
    },
    {
      "epoch": 0.16955266955266957,
      "grad_norm": 0.21996615827083588,
      "learning_rate": 9.49915565374431e-05,
      "loss": 1.2333,
      "step": 1410
    },
    {
      "epoch": 0.17075517075517074,
      "grad_norm": 0.22476601600646973,
      "learning_rate": 9.490626115845489e-05,
      "loss": 1.2375,
      "step": 1420
    },
    {
      "epoch": 0.17195767195767195,
      "grad_norm": 0.21539901196956635,
      "learning_rate": 9.482028455590793e-05,
      "loss": 1.2351,
      "step": 1430
    },
    {
      "epoch": 0.17316017316017315,
      "grad_norm": 0.21999700367450714,
      "learning_rate": 9.47336280340588e-05,
      "loss": 1.2217,
      "step": 1440
    },
    {
      "epoch": 0.17436267436267436,
      "grad_norm": 0.20949865877628326,
      "learning_rate": 9.464629290747842e-05,
      "loss": 1.2099,
      "step": 1450
    },
    {
      "epoch": 0.17556517556517556,
      "grad_norm": 0.21222126483917236,
      "learning_rate": 9.455828050103205e-05,
      "loss": 1.226,
      "step": 1460
    },
    {
      "epoch": 0.17676767676767677,
      "grad_norm": 0.22125640511512756,
      "learning_rate": 9.446959214985922e-05,
      "loss": 1.2223,
      "step": 1470
    },
    {
      "epoch": 0.17797017797017797,
      "grad_norm": 0.21773898601531982,
      "learning_rate": 9.438022919935348e-05,
      "loss": 1.2289,
      "step": 1480
    },
    {
      "epoch": 0.17917267917267918,
      "grad_norm": 0.22156096994876862,
      "learning_rate": 9.4290193005142e-05,
      "loss": 1.2191,
      "step": 1490
    },
    {
      "epoch": 0.18037518037518038,
      "grad_norm": 0.2058996856212616,
      "learning_rate": 9.419948493306497e-05,
      "loss": 1.2111,
      "step": 1500
    },
    {
      "epoch": 0.1815776815776816,
      "grad_norm": 0.22363074123859406,
      "learning_rate": 9.410810635915491e-05,
      "loss": 1.2171,
      "step": 1510
    },
    {
      "epoch": 0.1827801827801828,
      "grad_norm": 0.20594915747642517,
      "learning_rate": 9.401605866961579e-05,
      "loss": 1.2205,
      "step": 1520
    },
    {
      "epoch": 0.18398268398268397,
      "grad_norm": 0.22017163038253784,
      "learning_rate": 9.392334326080199e-05,
      "loss": 1.2306,
      "step": 1530
    },
    {
      "epoch": 0.18518518518518517,
      "grad_norm": 0.20129500329494476,
      "learning_rate": 9.382996153919713e-05,
      "loss": 1.2044,
      "step": 1540
    },
    {
      "epoch": 0.18638768638768638,
      "grad_norm": 0.21580781042575836,
      "learning_rate": 9.373591492139273e-05,
      "loss": 1.2207,
      "step": 1550
    },
    {
      "epoch": 0.18759018759018758,
      "grad_norm": 0.20488041639328003,
      "learning_rate": 9.364120483406672e-05,
      "loss": 1.2311,
      "step": 1560
    },
    {
      "epoch": 0.1887926887926888,
      "grad_norm": 0.22156283259391785,
      "learning_rate": 9.35458327139618e-05,
      "loss": 1.2158,
      "step": 1570
    },
    {
      "epoch": 0.18999518999519,
      "grad_norm": 0.22037768363952637,
      "learning_rate": 9.344980000786363e-05,
      "loss": 1.2246,
      "step": 1580
    },
    {
      "epoch": 0.1911976911976912,
      "grad_norm": 0.19976097345352173,
      "learning_rate": 9.33531081725789e-05,
      "loss": 1.2157,
      "step": 1590
    },
    {
      "epoch": 0.1924001924001924,
      "grad_norm": 0.2248261421918869,
      "learning_rate": 9.325575867491323e-05,
      "loss": 1.2235,
      "step": 1600
    },
    {
      "epoch": 0.1936026936026936,
      "grad_norm": 0.22592051327228546,
      "learning_rate": 9.315775299164892e-05,
      "loss": 1.2271,
      "step": 1610
    },
    {
      "epoch": 0.19480519480519481,
      "grad_norm": 0.22421322762966156,
      "learning_rate": 9.305909260952254e-05,
      "loss": 1.2378,
      "step": 1620
    },
    {
      "epoch": 0.19600769600769602,
      "grad_norm": 0.2143516093492508,
      "learning_rate": 9.295977902520235e-05,
      "loss": 1.2326,
      "step": 1630
    },
    {
      "epoch": 0.1972101972101972,
      "grad_norm": 0.2073206901550293,
      "learning_rate": 9.285981374526564e-05,
      "loss": 1.2241,
      "step": 1640
    },
    {
      "epoch": 0.1984126984126984,
      "grad_norm": 0.2234421968460083,
      "learning_rate": 9.275919828617589e-05,
      "loss": 1.2277,
      "step": 1650
    },
    {
      "epoch": 0.1996151996151996,
      "grad_norm": 0.218864306807518,
      "learning_rate": 9.265793417425967e-05,
      "loss": 1.2256,
      "step": 1660
    },
    {
      "epoch": 0.2008177008177008,
      "grad_norm": 0.2139173001050949,
      "learning_rate": 9.25560229456836e-05,
      "loss": 1.2247,
      "step": 1670
    },
    {
      "epoch": 0.20202020202020202,
      "grad_norm": 0.2014576643705368,
      "learning_rate": 9.245346614643096e-05,
      "loss": 1.2215,
      "step": 1680
    },
    {
      "epoch": 0.20322270322270322,
      "grad_norm": 0.22390897572040558,
      "learning_rate": 9.235026533227833e-05,
      "loss": 1.2273,
      "step": 1690
    },
    {
      "epoch": 0.20442520442520443,
      "grad_norm": 0.2203395515680313,
      "learning_rate": 9.224642206877187e-05,
      "loss": 1.2067,
      "step": 1700
    },
    {
      "epoch": 0.20562770562770563,
      "grad_norm": 0.22504153847694397,
      "learning_rate": 9.214193793120368e-05,
      "loss": 1.2285,
      "step": 1710
    },
    {
      "epoch": 0.20683020683020684,
      "grad_norm": 0.2131216675043106,
      "learning_rate": 9.203681450458781e-05,
      "loss": 1.2239,
      "step": 1720
    },
    {
      "epoch": 0.20803270803270804,
      "grad_norm": 0.2105296105146408,
      "learning_rate": 9.19310533836363e-05,
      "loss": 1.2145,
      "step": 1730
    },
    {
      "epoch": 0.20923520923520925,
      "grad_norm": 0.2101157158613205,
      "learning_rate": 9.182465617273494e-05,
      "loss": 1.2134,
      "step": 1740
    },
    {
      "epoch": 0.21043771043771045,
      "grad_norm": 0.20046111941337585,
      "learning_rate": 9.171762448591894e-05,
      "loss": 1.2207,
      "step": 1750
    },
    {
      "epoch": 0.21164021164021163,
      "grad_norm": 0.232273668050766,
      "learning_rate": 9.160995994684845e-05,
      "loss": 1.2315,
      "step": 1760
    },
    {
      "epoch": 0.21284271284271283,
      "grad_norm": 0.21907658874988556,
      "learning_rate": 9.15016641887839e-05,
      "loss": 1.2183,
      "step": 1770
    },
    {
      "epoch": 0.21404521404521404,
      "grad_norm": 0.20053844153881073,
      "learning_rate": 9.139273885456132e-05,
      "loss": 1.2187,
      "step": 1780
    },
    {
      "epoch": 0.21524771524771524,
      "grad_norm": 0.2047508955001831,
      "learning_rate": 9.128318559656725e-05,
      "loss": 1.2022,
      "step": 1790
    },
    {
      "epoch": 0.21645021645021645,
      "grad_norm": 0.21334955096244812,
      "learning_rate": 9.117300607671386e-05,
      "loss": 1.1995,
      "step": 1800
    },
    {
      "epoch": 0.21765271765271765,
      "grad_norm": 0.20535485446453094,
      "learning_rate": 9.106220196641357e-05,
      "loss": 1.2255,
      "step": 1810
    },
    {
      "epoch": 0.21885521885521886,
      "grad_norm": 0.20729170739650726,
      "learning_rate": 9.095077494655388e-05,
      "loss": 1.2186,
      "step": 1820
    },
    {
      "epoch": 0.22005772005772006,
      "grad_norm": 0.20496122539043427,
      "learning_rate": 9.083872670747166e-05,
      "loss": 1.2024,
      "step": 1830
    },
    {
      "epoch": 0.22126022126022127,
      "grad_norm": 0.19766771793365479,
      "learning_rate": 9.07260589489277e-05,
      "loss": 1.2097,
      "step": 1840
    },
    {
      "epoch": 0.22246272246272247,
      "grad_norm": 0.21484710276126862,
      "learning_rate": 9.061277338008077e-05,
      "loss": 1.2311,
      "step": 1850
    },
    {
      "epoch": 0.22366522366522368,
      "grad_norm": 0.206639364361763,
      "learning_rate": 9.049887171946179e-05,
      "loss": 1.2167,
      "step": 1860
    },
    {
      "epoch": 0.22486772486772486,
      "grad_norm": 0.2012685239315033,
      "learning_rate": 9.038435569494773e-05,
      "loss": 1.2172,
      "step": 1870
    },
    {
      "epoch": 0.22607022607022606,
      "grad_norm": 0.2186802476644516,
      "learning_rate": 9.026922704373543e-05,
      "loss": 1.2147,
      "step": 1880
    },
    {
      "epoch": 0.22727272727272727,
      "grad_norm": 0.19656942784786224,
      "learning_rate": 9.015348751231517e-05,
      "loss": 1.2212,
      "step": 1890
    },
    {
      "epoch": 0.22847522847522847,
      "grad_norm": 0.20295904576778412,
      "learning_rate": 9.003713885644425e-05,
      "loss": 1.2021,
      "step": 1900
    },
    {
      "epoch": 0.22967772967772968,
      "grad_norm": 0.2020638883113861,
      "learning_rate": 8.99201828411203e-05,
      "loss": 1.2098,
      "step": 1910
    },
    {
      "epoch": 0.23088023088023088,
      "grad_norm": 0.2027895748615265,
      "learning_rate": 8.980262124055458e-05,
      "loss": 1.2116,
      "step": 1920
    },
    {
      "epoch": 0.23208273208273208,
      "grad_norm": 0.1999751627445221,
      "learning_rate": 8.968445583814497e-05,
      "loss": 1.2056,
      "step": 1930
    },
    {
      "epoch": 0.2332852332852333,
      "grad_norm": 0.20799754559993744,
      "learning_rate": 8.956568842644901e-05,
      "loss": 1.2088,
      "step": 1940
    },
    {
      "epoch": 0.2344877344877345,
      "grad_norm": 0.21467134356498718,
      "learning_rate": 8.944632080715662e-05,
      "loss": 1.2239,
      "step": 1950
    },
    {
      "epoch": 0.2356902356902357,
      "grad_norm": 0.2006864845752716,
      "learning_rate": 8.932635479106283e-05,
      "loss": 1.203,
      "step": 1960
    },
    {
      "epoch": 0.2368927368927369,
      "grad_norm": 0.20447371900081635,
      "learning_rate": 8.920579219804028e-05,
      "loss": 1.222,
      "step": 1970
    },
    {
      "epoch": 0.23809523809523808,
      "grad_norm": 0.21055077016353607,
      "learning_rate": 8.908463485701164e-05,
      "loss": 1.2173,
      "step": 1980
    },
    {
      "epoch": 0.2392977392977393,
      "grad_norm": 0.2071334421634674,
      "learning_rate": 8.896288460592186e-05,
      "loss": 1.2014,
      "step": 1990
    },
    {
      "epoch": 0.2405002405002405,
      "grad_norm": 0.21648217737674713,
      "learning_rate": 8.884054329171021e-05,
      "loss": 1.2085,
      "step": 2000
    },
    {
      "epoch": 0.2405002405002405,
      "eval_loss": 1.1967148780822754,
      "eval_runtime": 14.3237,
      "eval_samples_per_second": 17.872,
      "eval_steps_per_second": 0.559,
      "step": 2000
    },
    {
      "epoch": 0.2417027417027417,
      "grad_norm": 0.20826809108257294,
      "learning_rate": 8.871761277028243e-05,
      "loss": 1.2277,
      "step": 2010
    },
    {
      "epoch": 0.2429052429052429,
      "grad_norm": 0.2014724761247635,
      "learning_rate": 8.859409490648238e-05,
      "loss": 1.2081,
      "step": 2020
    },
    {
      "epoch": 0.2441077441077441,
      "grad_norm": 0.21313191950321198,
      "learning_rate": 8.84699915740639e-05,
      "loss": 1.2047,
      "step": 2030
    },
    {
      "epoch": 0.2453102453102453,
      "grad_norm": 0.21019341051578522,
      "learning_rate": 8.834530465566233e-05,
      "loss": 1.2178,
      "step": 2040
    },
    {
      "epoch": 0.24651274651274652,
      "grad_norm": 0.2147923707962036,
      "learning_rate": 8.822003604276595e-05,
      "loss": 1.2189,
      "step": 2050
    },
    {
      "epoch": 0.24771524771524772,
      "grad_norm": 0.214359313249588,
      "learning_rate": 8.809418763568725e-05,
      "loss": 1.2094,
      "step": 2060
    },
    {
      "epoch": 0.24891774891774893,
      "grad_norm": 0.21459384262561798,
      "learning_rate": 8.796776134353418e-05,
      "loss": 1.216,
      "step": 2070
    },
    {
      "epoch": 0.25012025012025013,
      "grad_norm": 0.2167990505695343,
      "learning_rate": 8.784075908418115e-05,
      "loss": 1.2014,
      "step": 2080
    },
    {
      "epoch": 0.25132275132275134,
      "grad_norm": 0.2073349505662918,
      "learning_rate": 8.771318278423992e-05,
      "loss": 1.2033,
      "step": 2090
    },
    {
      "epoch": 0.25252525252525254,
      "grad_norm": 0.20970456302165985,
      "learning_rate": 8.758503437903038e-05,
      "loss": 1.1972,
      "step": 2100
    },
    {
      "epoch": 0.25372775372775375,
      "grad_norm": 0.19900967180728912,
      "learning_rate": 8.745631581255125e-05,
      "loss": 1.1896,
      "step": 2110
    },
    {
      "epoch": 0.25493025493025495,
      "grad_norm": 0.21769461035728455,
      "learning_rate": 8.732702903745048e-05,
      "loss": 1.213,
      "step": 2120
    },
    {
      "epoch": 0.25613275613275616,
      "grad_norm": 0.21593014895915985,
      "learning_rate": 8.719717601499571e-05,
      "loss": 1.2076,
      "step": 2130
    },
    {
      "epoch": 0.25733525733525736,
      "grad_norm": 0.20195241272449493,
      "learning_rate": 8.706675871504451e-05,
      "loss": 1.2142,
      "step": 2140
    },
    {
      "epoch": 0.2585377585377585,
      "grad_norm": 0.2088335007429123,
      "learning_rate": 8.693577911601447e-05,
      "loss": 1.2198,
      "step": 2150
    },
    {
      "epoch": 0.2597402597402597,
      "grad_norm": 0.21312519907951355,
      "learning_rate": 8.680423920485321e-05,
      "loss": 1.2187,
      "step": 2160
    },
    {
      "epoch": 0.2609427609427609,
      "grad_norm": 0.20363889634609222,
      "learning_rate": 8.667214097700824e-05,
      "loss": 1.1993,
      "step": 2170
    },
    {
      "epoch": 0.2621452621452621,
      "grad_norm": 0.21147218346595764,
      "learning_rate": 8.653948643639665e-05,
      "loss": 1.2278,
      "step": 2180
    },
    {
      "epoch": 0.26334776334776333,
      "grad_norm": 0.1997493952512741,
      "learning_rate": 8.640627759537475e-05,
      "loss": 1.2036,
      "step": 2190
    },
    {
      "epoch": 0.26455026455026454,
      "grad_norm": 0.21293872594833374,
      "learning_rate": 8.627251647470756e-05,
      "loss": 1.2096,
      "step": 2200
    },
    {
      "epoch": 0.26575276575276574,
      "grad_norm": 0.19573794305324554,
      "learning_rate": 8.613820510353807e-05,
      "loss": 1.2034,
      "step": 2210
    },
    {
      "epoch": 0.26695526695526695,
      "grad_norm": 0.21020345389842987,
      "learning_rate": 8.600334551935658e-05,
      "loss": 1.2207,
      "step": 2220
    },
    {
      "epoch": 0.26815776815776815,
      "grad_norm": 0.20261149108409882,
      "learning_rate": 8.58679397679697e-05,
      "loss": 1.2128,
      "step": 2230
    },
    {
      "epoch": 0.26936026936026936,
      "grad_norm": 0.20684467256069183,
      "learning_rate": 8.573198990346933e-05,
      "loss": 1.1972,
      "step": 2240
    },
    {
      "epoch": 0.27056277056277056,
      "grad_norm": 0.19626250863075256,
      "learning_rate": 8.559549798820153e-05,
      "loss": 1.1957,
      "step": 2250
    },
    {
      "epoch": 0.27176527176527177,
      "grad_norm": 0.20869341492652893,
      "learning_rate": 8.545846609273522e-05,
      "loss": 1.2164,
      "step": 2260
    },
    {
      "epoch": 0.27296777296777297,
      "grad_norm": 0.21097266674041748,
      "learning_rate": 8.532089629583073e-05,
      "loss": 1.2131,
      "step": 2270
    },
    {
      "epoch": 0.2741702741702742,
      "grad_norm": 0.2031770646572113,
      "learning_rate": 8.518279068440835e-05,
      "loss": 1.206,
      "step": 2280
    },
    {
      "epoch": 0.2753727753727754,
      "grad_norm": 0.21206524968147278,
      "learning_rate": 8.50441513535166e-05,
      "loss": 1.2118,
      "step": 2290
    },
    {
      "epoch": 0.2765752765752766,
      "grad_norm": 0.19096438586711884,
      "learning_rate": 8.490498040630048e-05,
      "loss": 1.2009,
      "step": 2300
    },
    {
      "epoch": 0.2777777777777778,
      "grad_norm": 0.20884640514850616,
      "learning_rate": 8.476527995396953e-05,
      "loss": 1.1908,
      "step": 2310
    },
    {
      "epoch": 0.278980278980279,
      "grad_norm": 0.20409995317459106,
      "learning_rate": 8.462505211576588e-05,
      "loss": 1.1933,
      "step": 2320
    },
    {
      "epoch": 0.2801827801827802,
      "grad_norm": 0.2128736823797226,
      "learning_rate": 8.4484299018932e-05,
      "loss": 1.207,
      "step": 2330
    },
    {
      "epoch": 0.2813852813852814,
      "grad_norm": 0.2101830691099167,
      "learning_rate": 8.434302279867852e-05,
      "loss": 1.1973,
      "step": 2340
    },
    {
      "epoch": 0.2825877825877826,
      "grad_norm": 0.19666483998298645,
      "learning_rate": 8.420122559815177e-05,
      "loss": 1.1878,
      "step": 2350
    },
    {
      "epoch": 0.2837902837902838,
      "grad_norm": 0.1945558786392212,
      "learning_rate": 8.405890956840135e-05,
      "loss": 1.2162,
      "step": 2360
    },
    {
      "epoch": 0.284992784992785,
      "grad_norm": 0.2024092972278595,
      "learning_rate": 8.39160768683474e-05,
      "loss": 1.2096,
      "step": 2370
    },
    {
      "epoch": 0.28619528619528617,
      "grad_norm": 0.2207275927066803,
      "learning_rate": 8.377272966474788e-05,
      "loss": 1.2031,
      "step": 2380
    },
    {
      "epoch": 0.2873977873977874,
      "grad_norm": 0.1977287232875824,
      "learning_rate": 8.362887013216578e-05,
      "loss": 1.1946,
      "step": 2390
    },
    {
      "epoch": 0.2886002886002886,
      "grad_norm": 0.19938814640045166,
      "learning_rate": 8.348450045293604e-05,
      "loss": 1.2117,
      "step": 2400
    },
    {
      "epoch": 0.2898027898027898,
      "grad_norm": 0.21072271466255188,
      "learning_rate": 8.333962281713246e-05,
      "loss": 1.2087,
      "step": 2410
    },
    {
      "epoch": 0.291005291005291,
      "grad_norm": 0.2154613584280014,
      "learning_rate": 8.319423942253451e-05,
      "loss": 1.1915,
      "step": 2420
    },
    {
      "epoch": 0.2922077922077922,
      "grad_norm": 0.20417426526546478,
      "learning_rate": 8.304835247459397e-05,
      "loss": 1.1966,
      "step": 2430
    },
    {
      "epoch": 0.2934102934102934,
      "grad_norm": 0.21270157396793365,
      "learning_rate": 8.290196418640148e-05,
      "loss": 1.1878,
      "step": 2440
    },
    {
      "epoch": 0.2946127946127946,
      "grad_norm": 0.19687621295452118,
      "learning_rate": 8.275507677865295e-05,
      "loss": 1.2036,
      "step": 2450
    },
    {
      "epoch": 0.2958152958152958,
      "grad_norm": 0.19195468723773956,
      "learning_rate": 8.260769247961586e-05,
      "loss": 1.1901,
      "step": 2460
    },
    {
      "epoch": 0.297017797017797,
      "grad_norm": 0.2017616629600525,
      "learning_rate": 8.245981352509555e-05,
      "loss": 1.2007,
      "step": 2470
    },
    {
      "epoch": 0.2982202982202982,
      "grad_norm": 0.21277114748954773,
      "learning_rate": 8.231144215840117e-05,
      "loss": 1.1899,
      "step": 2480
    },
    {
      "epoch": 0.2994227994227994,
      "grad_norm": 0.21405024826526642,
      "learning_rate": 8.216258063031175e-05,
      "loss": 1.2029,
      "step": 2490
    },
    {
      "epoch": 0.30062530062530063,
      "grad_norm": 0.20833241939544678,
      "learning_rate": 8.201323119904201e-05,
      "loss": 1.2087,
      "step": 2500
    },
    {
      "epoch": 0.30182780182780183,
      "grad_norm": 0.20577466487884521,
      "learning_rate": 8.186339613020809e-05,
      "loss": 1.1951,
      "step": 2510
    },
    {
      "epoch": 0.30303030303030304,
      "grad_norm": 0.20994292199611664,
      "learning_rate": 8.171307769679326e-05,
      "loss": 1.2139,
      "step": 2520
    },
    {
      "epoch": 0.30423280423280424,
      "grad_norm": 0.2054295688867569,
      "learning_rate": 8.156227817911333e-05,
      "loss": 1.2085,
      "step": 2530
    },
    {
      "epoch": 0.30543530543530545,
      "grad_norm": 0.1991153508424759,
      "learning_rate": 8.141099986478212e-05,
      "loss": 1.1936,
      "step": 2540
    },
    {
      "epoch": 0.30663780663780665,
      "grad_norm": 0.20854881405830383,
      "learning_rate": 8.125924504867679e-05,
      "loss": 1.204,
      "step": 2550
    },
    {
      "epoch": 0.30784030784030786,
      "grad_norm": 0.20640797913074493,
      "learning_rate": 8.110701603290292e-05,
      "loss": 1.1792,
      "step": 2560
    },
    {
      "epoch": 0.30904280904280906,
      "grad_norm": 0.20775394141674042,
      "learning_rate": 8.09543151267597e-05,
      "loss": 1.1927,
      "step": 2570
    },
    {
      "epoch": 0.31024531024531027,
      "grad_norm": 0.21143817901611328,
      "learning_rate": 8.080114464670485e-05,
      "loss": 1.1949,
      "step": 2580
    },
    {
      "epoch": 0.3114478114478115,
      "grad_norm": 0.2006131410598755,
      "learning_rate": 8.064750691631944e-05,
      "loss": 1.192,
      "step": 2590
    },
    {
      "epoch": 0.3126503126503126,
      "grad_norm": 0.2078796625137329,
      "learning_rate": 8.049340426627274e-05,
      "loss": 1.2008,
      "step": 2600
    },
    {
      "epoch": 0.31385281385281383,
      "grad_norm": 0.20483560860157013,
      "learning_rate": 8.033883903428674e-05,
      "loss": 1.2157,
      "step": 2610
    },
    {
      "epoch": 0.31505531505531503,
      "grad_norm": 0.2086118459701538,
      "learning_rate": 8.018381356510079e-05,
      "loss": 1.1891,
      "step": 2620
    },
    {
      "epoch": 0.31625781625781624,
      "grad_norm": 0.20505578815937042,
      "learning_rate": 8.0028330210436e-05,
      "loss": 1.2005,
      "step": 2630
    },
    {
      "epoch": 0.31746031746031744,
      "grad_norm": 0.2115529179573059,
      "learning_rate": 7.987239132895954e-05,
      "loss": 1.2056,
      "step": 2640
    },
    {
      "epoch": 0.31866281866281865,
      "grad_norm": 0.19756314158439636,
      "learning_rate": 7.97159992862489e-05,
      "loss": 1.2031,
      "step": 2650
    },
    {
      "epoch": 0.31986531986531985,
      "grad_norm": 0.20881183445453644,
      "learning_rate": 7.955915645475593e-05,
      "loss": 1.1852,
      "step": 2660
    },
    {
      "epoch": 0.32106782106782106,
      "grad_norm": 0.21202848851680756,
      "learning_rate": 7.940186521377097e-05,
      "loss": 1.1975,
      "step": 2670
    },
    {
      "epoch": 0.32227032227032226,
      "grad_norm": 0.20970021188259125,
      "learning_rate": 7.924412794938664e-05,
      "loss": 1.1986,
      "step": 2680
    },
    {
      "epoch": 0.32347282347282347,
      "grad_norm": 0.2039441466331482,
      "learning_rate": 7.90859470544617e-05,
      "loss": 1.2239,
      "step": 2690
    },
    {
      "epoch": 0.3246753246753247,
      "grad_norm": 0.2211122065782547,
      "learning_rate": 7.892732492858474e-05,
      "loss": 1.1958,
      "step": 2700
    },
    {
      "epoch": 0.3258778258778259,
      "grad_norm": 0.20219026505947113,
      "learning_rate": 7.876826397803781e-05,
      "loss": 1.2029,
      "step": 2710
    },
    {
      "epoch": 0.3270803270803271,
      "grad_norm": 0.197218120098114,
      "learning_rate": 7.860876661575985e-05,
      "loss": 1.1929,
      "step": 2720
    },
    {
      "epoch": 0.3282828282828283,
      "grad_norm": 0.20465494692325592,
      "learning_rate": 7.844883526131013e-05,
      "loss": 1.198,
      "step": 2730
    },
    {
      "epoch": 0.3294853294853295,
      "grad_norm": 0.22161713242530823,
      "learning_rate": 7.828847234083153e-05,
      "loss": 1.1981,
      "step": 2740
    },
    {
      "epoch": 0.3306878306878307,
      "grad_norm": 0.20785829424858093,
      "learning_rate": 7.81276802870138e-05,
      "loss": 1.1915,
      "step": 2750
    },
    {
      "epoch": 0.3318903318903319,
      "grad_norm": 0.20283977687358856,
      "learning_rate": 7.796646153905653e-05,
      "loss": 1.1856,
      "step": 2760
    },
    {
      "epoch": 0.3330928330928331,
      "grad_norm": 0.20320013165473938,
      "learning_rate": 7.780481854263225e-05,
      "loss": 1.1918,
      "step": 2770
    },
    {
      "epoch": 0.3342953342953343,
      "grad_norm": 0.20670950412750244,
      "learning_rate": 7.76427537498493e-05,
      "loss": 1.1965,
      "step": 2780
    },
    {
      "epoch": 0.3354978354978355,
      "grad_norm": 0.199970081448555,
      "learning_rate": 7.748026961921465e-05,
      "loss": 1.2006,
      "step": 2790
    },
    {
      "epoch": 0.3367003367003367,
      "grad_norm": 0.2103297859430313,
      "learning_rate": 7.731736861559651e-05,
      "loss": 1.1927,
      "step": 2800
    },
    {
      "epoch": 0.3379028379028379,
      "grad_norm": 0.2028837651014328,
      "learning_rate": 7.715405321018707e-05,
      "loss": 1.2001,
      "step": 2810
    },
    {
      "epoch": 0.33910533910533913,
      "grad_norm": 0.20478375256061554,
      "learning_rate": 7.699032588046498e-05,
      "loss": 1.1841,
      "step": 2820
    },
    {
      "epoch": 0.3403078403078403,
      "grad_norm": 0.20195071399211884,
      "learning_rate": 7.682618911015767e-05,
      "loss": 1.1666,
      "step": 2830
    },
    {
      "epoch": 0.3415103415103415,
      "grad_norm": 0.21081767976284027,
      "learning_rate": 7.666164538920378e-05,
      "loss": 1.1951,
      "step": 2840
    },
    {
      "epoch": 0.3427128427128427,
      "grad_norm": 0.2054494321346283,
      "learning_rate": 7.649669721371537e-05,
      "loss": 1.2028,
      "step": 2850
    },
    {
      "epoch": 0.3439153439153439,
      "grad_norm": 0.20955149829387665,
      "learning_rate": 7.633134708594001e-05,
      "loss": 1.2108,
      "step": 2860
    },
    {
      "epoch": 0.3451178451178451,
      "grad_norm": 0.2094833254814148,
      "learning_rate": 7.616559751422285e-05,
      "loss": 1.2022,
      "step": 2870
    },
    {
      "epoch": 0.3463203463203463,
      "grad_norm": 0.20287872850894928,
      "learning_rate": 7.599945101296856e-05,
      "loss": 1.202,
      "step": 2880
    },
    {
      "epoch": 0.3475228475228475,
      "grad_norm": 0.2033422887325287,
      "learning_rate": 7.583291010260321e-05,
      "loss": 1.1895,
      "step": 2890
    },
    {
      "epoch": 0.3487253487253487,
      "grad_norm": 0.20638012886047363,
      "learning_rate": 7.566597730953602e-05,
      "loss": 1.1981,
      "step": 2900
    },
    {
      "epoch": 0.3499278499278499,
      "grad_norm": 0.19885151088237762,
      "learning_rate": 7.549865516612098e-05,
      "loss": 1.1895,
      "step": 2910
    },
    {
      "epoch": 0.3511303511303511,
      "grad_norm": 0.20345020294189453,
      "learning_rate": 7.533094621061854e-05,
      "loss": 1.1855,
      "step": 2920
    },
    {
      "epoch": 0.35233285233285233,
      "grad_norm": 0.19838982820510864,
      "learning_rate": 7.516285298715706e-05,
      "loss": 1.1892,
      "step": 2930
    },
    {
      "epoch": 0.35353535353535354,
      "grad_norm": 0.1952482908964157,
      "learning_rate": 7.499437804569416e-05,
      "loss": 1.174,
      "step": 2940
    },
    {
      "epoch": 0.35473785473785474,
      "grad_norm": 0.20667259395122528,
      "learning_rate": 7.482552394197812e-05,
      "loss": 1.1913,
      "step": 2950
    },
    {
      "epoch": 0.35594035594035595,
      "grad_norm": 0.2147095650434494,
      "learning_rate": 7.465629323750905e-05,
      "loss": 1.1997,
      "step": 2960
    },
    {
      "epoch": 0.35714285714285715,
      "grad_norm": 0.21083101630210876,
      "learning_rate": 7.448668849950008e-05,
      "loss": 1.2023,
      "step": 2970
    },
    {
      "epoch": 0.35834535834535836,
      "grad_norm": 0.2125108540058136,
      "learning_rate": 7.431671230083836e-05,
      "loss": 1.1925,
      "step": 2980
    },
    {
      "epoch": 0.35954785954785956,
      "grad_norm": 0.20733962953090668,
      "learning_rate": 7.414636722004614e-05,
      "loss": 1.1871,
      "step": 2990
    },
    {
      "epoch": 0.36075036075036077,
      "grad_norm": 0.20381608605384827,
      "learning_rate": 7.397565584124144e-05,
      "loss": 1.1913,
      "step": 3000
    },
    {
      "epoch": 0.36075036075036077,
      "eval_loss": 1.1786988973617554,
      "eval_runtime": 14.3861,
      "eval_samples_per_second": 17.795,
      "eval_steps_per_second": 0.556,
      "step": 3000
    },
    {
      "epoch": 0.36195286195286197,
      "grad_norm": 0.19937773048877716,
      "learning_rate": 7.380458075409912e-05,
      "loss": 1.1869,
      "step": 3010
    },
    {
      "epoch": 0.3631553631553632,
      "grad_norm": 0.20516780018806458,
      "learning_rate": 7.363314455381136e-05,
      "loss": 1.175,
      "step": 3020
    },
    {
      "epoch": 0.3643578643578644,
      "grad_norm": 0.20108605921268463,
      "learning_rate": 7.346134984104846e-05,
      "loss": 1.1788,
      "step": 3030
    },
    {
      "epoch": 0.3655603655603656,
      "grad_norm": 0.20104803144931793,
      "learning_rate": 7.328919922191928e-05,
      "loss": 1.1832,
      "step": 3040
    },
    {
      "epoch": 0.3667628667628668,
      "grad_norm": 0.21416722238063812,
      "learning_rate": 7.311669530793176e-05,
      "loss": 1.1949,
      "step": 3050
    },
    {
      "epoch": 0.36796536796536794,
      "grad_norm": 0.19773589074611664,
      "learning_rate": 7.29438407159533e-05,
      "loss": 1.1887,
      "step": 3060
    },
    {
      "epoch": 0.36916786916786914,
      "grad_norm": 0.2011365294456482,
      "learning_rate": 7.277063806817102e-05,
      "loss": 1.1924,
      "step": 3070
    },
    {
      "epoch": 0.37037037037037035,
      "grad_norm": 0.19436123967170715,
      "learning_rate": 7.259708999205203e-05,
      "loss": 1.1853,
      "step": 3080
    },
    {
      "epoch": 0.37157287157287155,
      "grad_norm": 0.20969006419181824,
      "learning_rate": 7.242319912030356e-05,
      "loss": 1.1869,
      "step": 3090
    },
    {
      "epoch": 0.37277537277537276,
      "grad_norm": 0.19676007330417633,
      "learning_rate": 7.224896809083297e-05,
      "loss": 1.194,
      "step": 3100
    },
    {
      "epoch": 0.37397787397787396,
      "grad_norm": 0.20664817094802856,
      "learning_rate": 7.207439954670784e-05,
      "loss": 1.1919,
      "step": 3110
    },
    {
      "epoch": 0.37518037518037517,
      "grad_norm": 0.19718950986862183,
      "learning_rate": 7.189949613611582e-05,
      "loss": 1.1814,
      "step": 3120
    },
    {
      "epoch": 0.3763828763828764,
      "grad_norm": 0.19568046927452087,
      "learning_rate": 7.172426051232437e-05,
      "loss": 1.1958,
      "step": 3130
    },
    {
      "epoch": 0.3775853775853776,
      "grad_norm": 0.21798531711101532,
      "learning_rate": 7.154869533364067e-05,
      "loss": 1.1898,
      "step": 3140
    },
    {
      "epoch": 0.3787878787878788,
      "grad_norm": 0.20783917605876923,
      "learning_rate": 7.13728032633712e-05,
      "loss": 1.1806,
      "step": 3150
    },
    {
      "epoch": 0.37999037999038,
      "grad_norm": 0.20036672055721283,
      "learning_rate": 7.119658696978132e-05,
      "loss": 1.1902,
      "step": 3160
    },
    {
      "epoch": 0.3811928811928812,
      "grad_norm": 0.20404507219791412,
      "learning_rate": 7.102004912605485e-05,
      "loss": 1.1825,
      "step": 3170
    },
    {
      "epoch": 0.3823953823953824,
      "grad_norm": 0.21140117943286896,
      "learning_rate": 7.084319241025351e-05,
      "loss": 1.1956,
      "step": 3180
    },
    {
      "epoch": 0.3835978835978836,
      "grad_norm": 0.1950325220823288,
      "learning_rate": 7.066601950527625e-05,
      "loss": 1.1872,
      "step": 3190
    },
    {
      "epoch": 0.3848003848003848,
      "grad_norm": 0.20742125809192657,
      "learning_rate": 7.048853309881858e-05,
      "loss": 1.1929,
      "step": 3200
    },
    {
      "epoch": 0.386002886002886,
      "grad_norm": 0.21259550750255585,
      "learning_rate": 7.031073588333179e-05,
      "loss": 1.1935,
      "step": 3210
    },
    {
      "epoch": 0.3872053872053872,
      "grad_norm": 0.2075534164905548,
      "learning_rate": 7.013263055598216e-05,
      "loss": 1.1821,
      "step": 3220
    },
    {
      "epoch": 0.3884078884078884,
      "grad_norm": 0.19778668880462646,
      "learning_rate": 6.995421981860994e-05,
      "loss": 1.178,
      "step": 3230
    },
    {
      "epoch": 0.38961038961038963,
      "grad_norm": 0.19642433524131775,
      "learning_rate": 6.977550637768845e-05,
      "loss": 1.1993,
      "step": 3240
    },
    {
      "epoch": 0.39081289081289083,
      "grad_norm": 0.2102692574262619,
      "learning_rate": 6.959649294428301e-05,
      "loss": 1.192,
      "step": 3250
    },
    {
      "epoch": 0.39201539201539204,
      "grad_norm": 0.20500397682189941,
      "learning_rate": 6.941718223400975e-05,
      "loss": 1.1994,
      "step": 3260
    },
    {
      "epoch": 0.39321789321789324,
      "grad_norm": 0.21970514953136444,
      "learning_rate": 6.923757696699446e-05,
      "loss": 1.1792,
      "step": 3270
    },
    {
      "epoch": 0.3944203944203944,
      "grad_norm": 0.19493255019187927,
      "learning_rate": 6.90576798678314e-05,
      "loss": 1.1938,
      "step": 3280
    },
    {
      "epoch": 0.3956228956228956,
      "grad_norm": 0.19722020626068115,
      "learning_rate": 6.887749366554182e-05,
      "loss": 1.1904,
      "step": 3290
    },
    {
      "epoch": 0.3968253968253968,
      "grad_norm": 0.21302542090415955,
      "learning_rate": 6.869702109353264e-05,
      "loss": 1.1888,
      "step": 3300
    },
    {
      "epoch": 0.398027898027898,
      "grad_norm": 0.20728740096092224,
      "learning_rate": 6.851626488955505e-05,
      "loss": 1.2069,
      "step": 3310
    },
    {
      "epoch": 0.3992303992303992,
      "grad_norm": 0.2085309624671936,
      "learning_rate": 6.833522779566281e-05,
      "loss": 1.184,
      "step": 3320
    },
    {
      "epoch": 0.4004329004329004,
      "grad_norm": 0.2124137580394745,
      "learning_rate": 6.815391255817086e-05,
      "loss": 1.1877,
      "step": 3330
    },
    {
      "epoch": 0.4016354016354016,
      "grad_norm": 0.2014608383178711,
      "learning_rate": 6.797232192761345e-05,
      "loss": 1.1776,
      "step": 3340
    },
    {
      "epoch": 0.40283790283790283,
      "grad_norm": 0.19843342900276184,
      "learning_rate": 6.779045865870256e-05,
      "loss": 1.1966,
      "step": 3350
    },
    {
      "epoch": 0.40404040404040403,
      "grad_norm": 0.20755374431610107,
      "learning_rate": 6.760832551028609e-05,
      "loss": 1.1876,
      "step": 3360
    },
    {
      "epoch": 0.40524290524290524,
      "grad_norm": 0.19694215059280396,
      "learning_rate": 6.742592524530595e-05,
      "loss": 1.1788,
      "step": 3370
    },
    {
      "epoch": 0.40644540644540644,
      "grad_norm": 0.20904555916786194,
      "learning_rate": 6.72432606307562e-05,
      "loss": 1.1818,
      "step": 3380
    },
    {
      "epoch": 0.40764790764790765,
      "grad_norm": 0.21066156029701233,
      "learning_rate": 6.706033443764104e-05,
      "loss": 1.1899,
      "step": 3390
    },
    {
      "epoch": 0.40885040885040885,
      "grad_norm": 0.20113535225391388,
      "learning_rate": 6.687714944093282e-05,
      "loss": 1.1867,
      "step": 3400
    },
    {
      "epoch": 0.41005291005291006,
      "grad_norm": 0.198592409491539,
      "learning_rate": 6.669370841952987e-05,
      "loss": 1.186,
      "step": 3410
    },
    {
      "epoch": 0.41125541125541126,
      "grad_norm": 0.2045023888349533,
      "learning_rate": 6.651001415621446e-05,
      "loss": 1.1737,
      "step": 3420
    },
    {
      "epoch": 0.41245791245791247,
      "grad_norm": 0.20362548530101776,
      "learning_rate": 6.632606943761042e-05,
      "loss": 1.1602,
      "step": 3430
    },
    {
      "epoch": 0.4136604136604137,
      "grad_norm": 0.20577895641326904,
      "learning_rate": 6.614187705414105e-05,
      "loss": 1.1885,
      "step": 3440
    },
    {
      "epoch": 0.4148629148629149,
      "grad_norm": 0.202180415391922,
      "learning_rate": 6.595743979998668e-05,
      "loss": 1.1863,
      "step": 3450
    },
    {
      "epoch": 0.4160654160654161,
      "grad_norm": 0.19655773043632507,
      "learning_rate": 6.577276047304222e-05,
      "loss": 1.1767,
      "step": 3460
    },
    {
      "epoch": 0.4172679172679173,
      "grad_norm": 0.1869806945323944,
      "learning_rate": 6.558784187487494e-05,
      "loss": 1.1911,
      "step": 3470
    },
    {
      "epoch": 0.4184704184704185,
      "grad_norm": 0.19990621507167816,
      "learning_rate": 6.540268681068174e-05,
      "loss": 1.1714,
      "step": 3480
    },
    {
      "epoch": 0.4196729196729197,
      "grad_norm": 0.2027176469564438,
      "learning_rate": 6.521729808924668e-05,
      "loss": 1.1753,
      "step": 3490
    },
    {
      "epoch": 0.4208754208754209,
      "grad_norm": 0.19918115437030792,
      "learning_rate": 6.503167852289844e-05,
      "loss": 1.1862,
      "step": 3500
    },
    {
      "epoch": 0.42207792207792205,
      "grad_norm": 0.20005355775356293,
      "learning_rate": 6.484583092746753e-05,
      "loss": 1.1848,
      "step": 3510
    },
    {
      "epoch": 0.42328042328042326,
      "grad_norm": 0.19497598707675934,
      "learning_rate": 6.46597581222437e-05,
      "loss": 1.1677,
      "step": 3520
    },
    {
      "epoch": 0.42448292448292446,
      "grad_norm": 0.21364623308181763,
      "learning_rate": 6.447346292993307e-05,
      "loss": 1.1889,
      "step": 3530
    },
    {
      "epoch": 0.42568542568542567,
      "grad_norm": 0.1999025195837021,
      "learning_rate": 6.428694817661531e-05,
      "loss": 1.1833,
      "step": 3540
    },
    {
      "epoch": 0.42688792688792687,
      "grad_norm": 0.20487244427204132,
      "learning_rate": 6.410021669170093e-05,
      "loss": 1.189,
      "step": 3550
    },
    {
      "epoch": 0.4280904280904281,
      "grad_norm": 0.20337916910648346,
      "learning_rate": 6.391327130788813e-05,
      "loss": 1.1849,
      "step": 3560
    },
    {
      "epoch": 0.4292929292929293,
      "grad_norm": 0.20541474223136902,
      "learning_rate": 6.372611486111994e-05,
      "loss": 1.1879,
      "step": 3570
    },
    {
      "epoch": 0.4304954304954305,
      "grad_norm": 0.1998613476753235,
      "learning_rate": 6.353875019054128e-05,
      "loss": 1.1795,
      "step": 3580
    },
    {
      "epoch": 0.4316979316979317,
      "grad_norm": 0.19440466165542603,
      "learning_rate": 6.335118013845572e-05,
      "loss": 1.1865,
      "step": 3590
    },
    {
      "epoch": 0.4329004329004329,
      "grad_norm": 0.2010820358991623,
      "learning_rate": 6.316340755028247e-05,
      "loss": 1.184,
      "step": 3600
    },
    {
      "epoch": 0.4341029341029341,
      "grad_norm": 0.20916162431240082,
      "learning_rate": 6.297543527451322e-05,
      "loss": 1.1758,
      "step": 3610
    },
    {
      "epoch": 0.4353054353054353,
      "grad_norm": 0.20128333568572998,
      "learning_rate": 6.278726616266884e-05,
      "loss": 1.1753,
      "step": 3620
    },
    {
      "epoch": 0.4365079365079365,
      "grad_norm": 0.20334778726100922,
      "learning_rate": 6.259890306925627e-05,
      "loss": 1.1811,
      "step": 3630
    },
    {
      "epoch": 0.4377104377104377,
      "grad_norm": 0.21417754888534546,
      "learning_rate": 6.241034885172505e-05,
      "loss": 1.1873,
      "step": 3640
    },
    {
      "epoch": 0.4389129389129389,
      "grad_norm": 0.2282254844903946,
      "learning_rate": 6.22216063704241e-05,
      "loss": 1.1857,
      "step": 3650
    },
    {
      "epoch": 0.4401154401154401,
      "grad_norm": 0.1929093897342682,
      "learning_rate": 6.203267848855828e-05,
      "loss": 1.1864,
      "step": 3660
    },
    {
      "epoch": 0.44131794131794133,
      "grad_norm": 0.19937632977962494,
      "learning_rate": 6.184356807214493e-05,
      "loss": 1.1752,
      "step": 3670
    },
    {
      "epoch": 0.44252044252044254,
      "grad_norm": 0.20052993297576904,
      "learning_rate": 6.165427798997046e-05,
      "loss": 1.1702,
      "step": 3680
    },
    {
      "epoch": 0.44372294372294374,
      "grad_norm": 0.20837515592575073,
      "learning_rate": 6.146481111354676e-05,
      "loss": 1.1869,
      "step": 3690
    },
    {
      "epoch": 0.44492544492544495,
      "grad_norm": 0.19809991121292114,
      "learning_rate": 6.127517031706769e-05,
      "loss": 1.1922,
      "step": 3700
    },
    {
      "epoch": 0.44612794612794615,
      "grad_norm": 0.21065044403076172,
      "learning_rate": 6.108535847736546e-05,
      "loss": 1.1839,
      "step": 3710
    },
    {
      "epoch": 0.44733044733044736,
      "grad_norm": 0.20427262783050537,
      "learning_rate": 6.089537847386696e-05,
      "loss": 1.1687,
      "step": 3720
    },
    {
      "epoch": 0.4485329485329485,
      "grad_norm": 0.21668106317520142,
      "learning_rate": 6.070523318855015e-05,
      "loss": 1.1811,
      "step": 3730
    },
    {
      "epoch": 0.4497354497354497,
      "grad_norm": 0.20621725916862488,
      "learning_rate": 6.051492550590029e-05,
      "loss": 1.1824,
      "step": 3740
    },
    {
      "epoch": 0.4509379509379509,
      "grad_norm": 0.21493861079216003,
      "learning_rate": 6.032445831286615e-05,
      "loss": 1.1799,
      "step": 3750
    },
    {
      "epoch": 0.4521404521404521,
      "grad_norm": 0.20860455930233002,
      "learning_rate": 6.013383449881631e-05,
      "loss": 1.1793,
      "step": 3760
    },
    {
      "epoch": 0.4533429533429533,
      "grad_norm": 0.19611907005310059,
      "learning_rate": 5.994305695549526e-05,
      "loss": 1.1845,
      "step": 3770
    },
    {
      "epoch": 0.45454545454545453,
      "grad_norm": 0.20090104639530182,
      "learning_rate": 5.9752128576979514e-05,
      "loss": 1.1689,
      "step": 3780
    },
    {
      "epoch": 0.45574795574795574,
      "grad_norm": 0.22553130984306335,
      "learning_rate": 5.9561052259633796e-05,
      "loss": 1.191,
      "step": 3790
    },
    {
      "epoch": 0.45695045695045694,
      "grad_norm": 0.20412489771842957,
      "learning_rate": 5.936983090206701e-05,
      "loss": 1.1689,
      "step": 3800
    },
    {
      "epoch": 0.45815295815295815,
      "grad_norm": 0.20666392147541046,
      "learning_rate": 5.917846740508832e-05,
      "loss": 1.175,
      "step": 3810
    },
    {
      "epoch": 0.45935545935545935,
      "grad_norm": 0.20357099175453186,
      "learning_rate": 5.8986964671663115e-05,
      "loss": 1.1599,
      "step": 3820
    },
    {
      "epoch": 0.46055796055796056,
      "grad_norm": 0.20838573575019836,
      "learning_rate": 5.8795325606869025e-05,
      "loss": 1.1801,
      "step": 3830
    },
    {
      "epoch": 0.46176046176046176,
      "grad_norm": 0.21168790757656097,
      "learning_rate": 5.8603553117851753e-05,
      "loss": 1.1941,
      "step": 3840
    },
    {
      "epoch": 0.46296296296296297,
      "grad_norm": 0.20583811402320862,
      "learning_rate": 5.8411650113781094e-05,
      "loss": 1.1712,
      "step": 3850
    },
    {
      "epoch": 0.46416546416546417,
      "grad_norm": 0.20696823298931122,
      "learning_rate": 5.82196195058067e-05,
      "loss": 1.1683,
      "step": 3860
    },
    {
      "epoch": 0.4653679653679654,
      "grad_norm": 0.19953173398971558,
      "learning_rate": 5.802746420701399e-05,
      "loss": 1.1861,
      "step": 3870
    },
    {
      "epoch": 0.4665704665704666,
      "grad_norm": 0.20347267389297485,
      "learning_rate": 5.783518713237993e-05,
      "loss": 1.1963,
      "step": 3880
    },
    {
      "epoch": 0.4677729677729678,
      "grad_norm": 0.20178109407424927,
      "learning_rate": 5.7642791198728795e-05,
      "loss": 1.165,
      "step": 3890
    },
    {
      "epoch": 0.468975468975469,
      "grad_norm": 0.20376752316951752,
      "learning_rate": 5.745027932468794e-05,
      "loss": 1.1659,
      "step": 3900
    },
    {
      "epoch": 0.4701779701779702,
      "grad_norm": 0.2073439061641693,
      "learning_rate": 5.725765443064355e-05,
      "loss": 1.1892,
      "step": 3910
    },
    {
      "epoch": 0.4713804713804714,
      "grad_norm": 0.19912444055080414,
      "learning_rate": 5.7064919438696275e-05,
      "loss": 1.1858,
      "step": 3920
    },
    {
      "epoch": 0.4725829725829726,
      "grad_norm": 0.20428648591041565,
      "learning_rate": 5.687207727261699e-05,
      "loss": 1.1786,
      "step": 3930
    },
    {
      "epoch": 0.4737854737854738,
      "grad_norm": 0.21749798953533173,
      "learning_rate": 5.667913085780232e-05,
      "loss": 1.1752,
      "step": 3940
    },
    {
      "epoch": 0.474987974987975,
      "grad_norm": 0.2052384912967682,
      "learning_rate": 5.648608312123039e-05,
      "loss": 1.1752,
      "step": 3950
    },
    {
      "epoch": 0.47619047619047616,
      "grad_norm": 0.20432449877262115,
      "learning_rate": 5.629293699141638e-05,
      "loss": 1.1871,
      "step": 3960
    },
    {
      "epoch": 0.47739297739297737,
      "grad_norm": 0.19696982204914093,
      "learning_rate": 5.609969539836801e-05,
      "loss": 1.1618,
      "step": 3970
    },
    {
      "epoch": 0.4785954785954786,
      "grad_norm": 0.21920330822467804,
      "learning_rate": 5.5906361273541244e-05,
      "loss": 1.1877,
      "step": 3980
    },
    {
      "epoch": 0.4797979797979798,
      "grad_norm": 0.21005605161190033,
      "learning_rate": 5.5712937549795704e-05,
      "loss": 1.1934,
      "step": 3990
    },
    {
      "epoch": 0.481000481000481,
      "grad_norm": 0.20428964495658875,
      "learning_rate": 5.551942716135021e-05,
      "loss": 1.1762,
      "step": 4000
    },
    {
      "epoch": 0.481000481000481,
      "eval_loss": 1.1611641645431519,
      "eval_runtime": 14.3743,
      "eval_samples_per_second": 17.81,
      "eval_steps_per_second": 0.557,
      "step": 4000
    },
    {
      "epoch": 0.4822029822029822,
      "grad_norm": 0.20652806758880615,
      "learning_rate": 5.532583304373833e-05,
      "loss": 1.1826,
      "step": 4010
    },
    {
      "epoch": 0.4834054834054834,
      "grad_norm": 0.2061479687690735,
      "learning_rate": 5.513215813376373e-05,
      "loss": 1.1844,
      "step": 4020
    },
    {
      "epoch": 0.4846079846079846,
      "grad_norm": 0.2116565853357315,
      "learning_rate": 5.493840536945574e-05,
      "loss": 1.172,
      "step": 4030
    },
    {
      "epoch": 0.4858104858104858,
      "grad_norm": 0.20410141348838806,
      "learning_rate": 5.47445776900247e-05,
      "loss": 1.1754,
      "step": 4040
    },
    {
      "epoch": 0.487012987012987,
      "grad_norm": 0.20021560788154602,
      "learning_rate": 5.4550678035817404e-05,
      "loss": 1.1713,
      "step": 4050
    },
    {
      "epoch": 0.4882154882154882,
      "grad_norm": 0.20905666053295135,
      "learning_rate": 5.435670934827253e-05,
      "loss": 1.1746,
      "step": 4060
    },
    {
      "epoch": 0.4894179894179894,
      "grad_norm": 0.2209169864654541,
      "learning_rate": 5.416267456987596e-05,
      "loss": 1.1763,
      "step": 4070
    },
    {
      "epoch": 0.4906204906204906,
      "grad_norm": 0.19423747062683105,
      "learning_rate": 5.3968576644116153e-05,
      "loss": 1.1691,
      "step": 4080
    },
    {
      "epoch": 0.49182299182299183,
      "grad_norm": 0.19502019882202148,
      "learning_rate": 5.377441851543954e-05,
      "loss": 1.1726,
      "step": 4090
    },
    {
      "epoch": 0.49302549302549303,
      "grad_norm": 0.20400142669677734,
      "learning_rate": 5.358020312920582e-05,
      "loss": 1.1736,
      "step": 4100
    },
    {
      "epoch": 0.49422799422799424,
      "grad_norm": 0.2074715495109558,
      "learning_rate": 5.338593343164324e-05,
      "loss": 1.1811,
      "step": 4110
    },
    {
      "epoch": 0.49543049543049544,
      "grad_norm": 0.20175635814666748,
      "learning_rate": 5.319161236980402e-05,
      "loss": 1.1653,
      "step": 4120
    },
    {
      "epoch": 0.49663299663299665,
      "grad_norm": 0.20830634236335754,
      "learning_rate": 5.299724289151949e-05,
      "loss": 1.1735,
      "step": 4130
    },
    {
      "epoch": 0.49783549783549785,
      "grad_norm": 0.19762635231018066,
      "learning_rate": 5.280282794535552e-05,
      "loss": 1.1528,
      "step": 4140
    },
    {
      "epoch": 0.49903799903799906,
      "grad_norm": 0.22000741958618164,
      "learning_rate": 5.260837048056768e-05,
      "loss": 1.1677,
      "step": 4150
    },
    {
      "epoch": 0.5002405002405003,
      "grad_norm": 0.19429528713226318,
      "learning_rate": 5.241387344705655e-05,
      "loss": 1.1617,
      "step": 4160
    },
    {
      "epoch": 0.5014430014430015,
      "grad_norm": 0.20485764741897583,
      "learning_rate": 5.2219339795322985e-05,
      "loss": 1.1745,
      "step": 4170
    },
    {
      "epoch": 0.5026455026455027,
      "grad_norm": 0.20905248820781708,
      "learning_rate": 5.2024772476423324e-05,
      "loss": 1.1609,
      "step": 4180
    },
    {
      "epoch": 0.5038480038480039,
      "grad_norm": 0.2026577889919281,
      "learning_rate": 5.183017444192462e-05,
      "loss": 1.1693,
      "step": 4190
    },
    {
      "epoch": 0.5050505050505051,
      "grad_norm": 0.20207685232162476,
      "learning_rate": 5.16355486438599e-05,
      "loss": 1.1686,
      "step": 4200
    },
    {
      "epoch": 0.5062530062530063,
      "grad_norm": 0.1890622228384018,
      "learning_rate": 5.144089803468333e-05,
      "loss": 1.1672,
      "step": 4210
    },
    {
      "epoch": 0.5074555074555075,
      "grad_norm": 0.19660572707653046,
      "learning_rate": 5.1246225567225506e-05,
      "loss": 1.1672,
      "step": 4220
    },
    {
      "epoch": 0.5086580086580087,
      "grad_norm": 0.20353861153125763,
      "learning_rate": 5.1051534194648575e-05,
      "loss": 1.1759,
      "step": 4230
    },
    {
      "epoch": 0.5098605098605099,
      "grad_norm": 0.20563137531280518,
      "learning_rate": 5.085682687040148e-05,
      "loss": 1.1745,
      "step": 4240
    },
    {
      "epoch": 0.5110630110630111,
      "grad_norm": 0.20082563161849976,
      "learning_rate": 5.066210654817514e-05,
      "loss": 1.1726,
      "step": 4250
    },
    {
      "epoch": 0.5122655122655123,
      "grad_norm": 0.20972830057144165,
      "learning_rate": 5.0467376181857694e-05,
      "loss": 1.1588,
      "step": 4260
    },
    {
      "epoch": 0.5134680134680135,
      "grad_norm": 0.21431012451648712,
      "learning_rate": 5.0272638725489595e-05,
      "loss": 1.159,
      "step": 4270
    },
    {
      "epoch": 0.5146705146705147,
      "grad_norm": 0.20237547159194946,
      "learning_rate": 5.0077897133218884e-05,
      "loss": 1.1649,
      "step": 4280
    },
    {
      "epoch": 0.5158730158730159,
      "grad_norm": 0.19756945967674255,
      "learning_rate": 4.988315435925636e-05,
      "loss": 1.1692,
      "step": 4290
    },
    {
      "epoch": 0.517075517075517,
      "grad_norm": 0.2220674604177475,
      "learning_rate": 4.968841335783066e-05,
      "loss": 1.1713,
      "step": 4300
    },
    {
      "epoch": 0.5182780182780182,
      "grad_norm": 0.20141322910785675,
      "learning_rate": 4.949367708314365e-05,
      "loss": 1.1604,
      "step": 4310
    },
    {
      "epoch": 0.5194805194805194,
      "grad_norm": 0.2010202258825302,
      "learning_rate": 4.929894848932541e-05,
      "loss": 1.1758,
      "step": 4320
    },
    {
      "epoch": 0.5206830206830206,
      "grad_norm": 0.20987989008426666,
      "learning_rate": 4.910423053038955e-05,
      "loss": 1.1618,
      "step": 4330
    },
    {
      "epoch": 0.5218855218855218,
      "grad_norm": 0.20549160242080688,
      "learning_rate": 4.890952616018831e-05,
      "loss": 1.1745,
      "step": 4340
    },
    {
      "epoch": 0.523088023088023,
      "grad_norm": 0.20881348848342896,
      "learning_rate": 4.871483833236782e-05,
      "loss": 1.1615,
      "step": 4350
    },
    {
      "epoch": 0.5242905242905243,
      "grad_norm": 0.19820332527160645,
      "learning_rate": 4.852017000032322e-05,
      "loss": 1.1626,
      "step": 4360
    },
    {
      "epoch": 0.5254930254930255,
      "grad_norm": 0.20698867738246918,
      "learning_rate": 4.8325524117154003e-05,
      "loss": 1.1647,
      "step": 4370
    },
    {
      "epoch": 0.5266955266955267,
      "grad_norm": 0.1959027200937271,
      "learning_rate": 4.813090363561896e-05,
      "loss": 1.1777,
      "step": 4380
    },
    {
      "epoch": 0.5278980278980279,
      "grad_norm": 0.19776858389377594,
      "learning_rate": 4.79363115080917e-05,
      "loss": 1.1574,
      "step": 4390
    },
    {
      "epoch": 0.5291005291005291,
      "grad_norm": 0.20591264963150024,
      "learning_rate": 4.774175068651558e-05,
      "loss": 1.1642,
      "step": 4400
    },
    {
      "epoch": 0.5303030303030303,
      "grad_norm": 0.20426321029663086,
      "learning_rate": 4.7547224122359144e-05,
      "loss": 1.1537,
      "step": 4410
    },
    {
      "epoch": 0.5315055315055315,
      "grad_norm": 0.1917085498571396,
      "learning_rate": 4.735273476657116e-05,
      "loss": 1.1725,
      "step": 4420
    },
    {
      "epoch": 0.5327080327080327,
      "grad_norm": 0.21214643120765686,
      "learning_rate": 4.715828556953605e-05,
      "loss": 1.1736,
      "step": 4430
    },
    {
      "epoch": 0.5339105339105339,
      "grad_norm": 0.20908360183238983,
      "learning_rate": 4.696387948102893e-05,
      "loss": 1.1641,
      "step": 4440
    },
    {
      "epoch": 0.5351130351130351,
      "grad_norm": 0.2076905071735382,
      "learning_rate": 4.6769519450171074e-05,
      "loss": 1.1695,
      "step": 4450
    },
    {
      "epoch": 0.5363155363155363,
      "grad_norm": 0.212940976023674,
      "learning_rate": 4.657520842538494e-05,
      "loss": 1.1578,
      "step": 4460
    },
    {
      "epoch": 0.5375180375180375,
      "grad_norm": 0.20305852591991425,
      "learning_rate": 4.63809493543497e-05,
      "loss": 1.1774,
      "step": 4470
    },
    {
      "epoch": 0.5387205387205387,
      "grad_norm": 0.20637470483779907,
      "learning_rate": 4.618674518395628e-05,
      "loss": 1.1579,
      "step": 4480
    },
    {
      "epoch": 0.5399230399230399,
      "grad_norm": 0.20714348554611206,
      "learning_rate": 4.599259886026284e-05,
      "loss": 1.1549,
      "step": 4490
    },
    {
      "epoch": 0.5411255411255411,
      "grad_norm": 0.20754988491535187,
      "learning_rate": 4.5798513328449976e-05,
      "loss": 1.1563,
      "step": 4500
    },
    {
      "epoch": 0.5423280423280423,
      "grad_norm": 0.20212216675281525,
      "learning_rate": 4.560449153277611e-05,
      "loss": 1.1645,
      "step": 4510
    },
    {
      "epoch": 0.5435305435305435,
      "grad_norm": 0.1956162005662918,
      "learning_rate": 4.541053641653274e-05,
      "loss": 1.1674,
      "step": 4520
    },
    {
      "epoch": 0.5447330447330447,
      "grad_norm": 0.19848354160785675,
      "learning_rate": 4.521665092199991e-05,
      "loss": 1.1677,
      "step": 4530
    },
    {
      "epoch": 0.5459355459355459,
      "grad_norm": 0.19816353917121887,
      "learning_rate": 4.502283799040145e-05,
      "loss": 1.1626,
      "step": 4540
    },
    {
      "epoch": 0.5471380471380471,
      "grad_norm": 0.20949548482894897,
      "learning_rate": 4.482910056186047e-05,
      "loss": 1.1752,
      "step": 4550
    },
    {
      "epoch": 0.5483405483405484,
      "grad_norm": 0.20675860345363617,
      "learning_rate": 4.463544157535464e-05,
      "loss": 1.1584,
      "step": 4560
    },
    {
      "epoch": 0.5495430495430496,
      "grad_norm": 0.2039814293384552,
      "learning_rate": 4.444186396867175e-05,
      "loss": 1.1671,
      "step": 4570
    },
    {
      "epoch": 0.5507455507455508,
      "grad_norm": 0.2231522798538208,
      "learning_rate": 4.4248370678365e-05,
      "loss": 1.1648,
      "step": 4580
    },
    {
      "epoch": 0.551948051948052,
      "grad_norm": 0.2013349086046219,
      "learning_rate": 4.4054964639708554e-05,
      "loss": 1.1764,
      "step": 4590
    },
    {
      "epoch": 0.5531505531505532,
      "grad_norm": 0.20852096378803253,
      "learning_rate": 4.386164878665297e-05,
      "loss": 1.1717,
      "step": 4600
    },
    {
      "epoch": 0.5543530543530544,
      "grad_norm": 0.21214091777801514,
      "learning_rate": 4.3668426051780704e-05,
      "loss": 1.1644,
      "step": 4610
    },
    {
      "epoch": 0.5555555555555556,
      "grad_norm": 0.215733602643013,
      "learning_rate": 4.3475299366261604e-05,
      "loss": 1.1631,
      "step": 4620
    },
    {
      "epoch": 0.5567580567580568,
      "grad_norm": 0.20751333236694336,
      "learning_rate": 4.328227165980847e-05,
      "loss": 1.1545,
      "step": 4630
    },
    {
      "epoch": 0.557960557960558,
      "grad_norm": 0.1947864145040512,
      "learning_rate": 4.3089345860632594e-05,
      "loss": 1.1601,
      "step": 4640
    },
    {
      "epoch": 0.5591630591630592,
      "grad_norm": 0.20173963904380798,
      "learning_rate": 4.289652489539937e-05,
      "loss": 1.1521,
      "step": 4650
    },
    {
      "epoch": 0.5603655603655604,
      "grad_norm": 0.19809898734092712,
      "learning_rate": 4.2703811689183797e-05,
      "loss": 1.1608,
      "step": 4660
    },
    {
      "epoch": 0.5615680615680616,
      "grad_norm": 0.2071085423231125,
      "learning_rate": 4.2511209165426306e-05,
      "loss": 1.1748,
      "step": 4670
    },
    {
      "epoch": 0.5627705627705628,
      "grad_norm": 0.1951216459274292,
      "learning_rate": 4.2318720245888135e-05,
      "loss": 1.1685,
      "step": 4680
    },
    {
      "epoch": 0.563973063973064,
      "grad_norm": 0.1975964456796646,
      "learning_rate": 4.212634785060728e-05,
      "loss": 1.1637,
      "step": 4690
    },
    {
      "epoch": 0.5651755651755652,
      "grad_norm": 0.2012360543012619,
      "learning_rate": 4.193409489785403e-05,
      "loss": 1.1811,
      "step": 4700
    },
    {
      "epoch": 0.5663780663780664,
      "grad_norm": 0.20509780943393707,
      "learning_rate": 4.174196430408673e-05,
      "loss": 1.1719,
      "step": 4710
    },
    {
      "epoch": 0.5675805675805676,
      "grad_norm": 0.19332098960876465,
      "learning_rate": 4.1549958983907555e-05,
      "loss": 1.1605,
      "step": 4720
    },
    {
      "epoch": 0.5687830687830688,
      "grad_norm": 0.20153947174549103,
      "learning_rate": 4.1358081850018326e-05,
      "loss": 1.1546,
      "step": 4730
    },
    {
      "epoch": 0.56998556998557,
      "grad_norm": 0.2031155526638031,
      "learning_rate": 4.116633581317623e-05,
      "loss": 1.1708,
      "step": 4740
    },
    {
      "epoch": 0.5711880711880711,
      "grad_norm": 0.1971137523651123,
      "learning_rate": 4.097472378214981e-05,
      "loss": 1.1527,
      "step": 4750
    },
    {
      "epoch": 0.5723905723905723,
      "grad_norm": 0.20589309930801392,
      "learning_rate": 4.0783248663674645e-05,
      "loss": 1.1551,
      "step": 4760
    },
    {
      "epoch": 0.5735930735930735,
      "grad_norm": 0.21493607759475708,
      "learning_rate": 4.059191336240947e-05,
      "loss": 1.166,
      "step": 4770
    },
    {
      "epoch": 0.5747955747955747,
      "grad_norm": 0.20907354354858398,
      "learning_rate": 4.0400720780891935e-05,
      "loss": 1.1576,
      "step": 4780
    },
    {
      "epoch": 0.575998075998076,
      "grad_norm": 0.20002475380897522,
      "learning_rate": 4.0209673819494695e-05,
      "loss": 1.1631,
      "step": 4790
    },
    {
      "epoch": 0.5772005772005772,
      "grad_norm": 0.1972057819366455,
      "learning_rate": 4.00187753763813e-05,
      "loss": 1.1631,
      "step": 4800
    },
    {
      "epoch": 0.5784030784030784,
      "grad_norm": 0.2116405963897705,
      "learning_rate": 3.982802834746236e-05,
      "loss": 1.1727,
      "step": 4810
    },
    {
      "epoch": 0.5796055796055796,
      "grad_norm": 0.19388829171657562,
      "learning_rate": 3.963743562635149e-05,
      "loss": 1.1569,
      "step": 4820
    },
    {
      "epoch": 0.5808080808080808,
      "grad_norm": 0.2148236185312271,
      "learning_rate": 3.9447000104321544e-05,
      "loss": 1.1676,
      "step": 4830
    },
    {
      "epoch": 0.582010582010582,
      "grad_norm": 0.20140714943408966,
      "learning_rate": 3.925672467026057e-05,
      "loss": 1.1517,
      "step": 4840
    },
    {
      "epoch": 0.5832130832130832,
      "grad_norm": 0.20167213678359985,
      "learning_rate": 3.9066612210628216e-05,
      "loss": 1.1651,
      "step": 4850
    },
    {
      "epoch": 0.5844155844155844,
      "grad_norm": 0.20786848664283752,
      "learning_rate": 3.887666560941174e-05,
      "loss": 1.1512,
      "step": 4860
    },
    {
      "epoch": 0.5856180856180856,
      "grad_norm": 0.19768615067005157,
      "learning_rate": 3.868688774808239e-05,
      "loss": 1.1676,
      "step": 4870
    },
    {
      "epoch": 0.5868205868205868,
      "grad_norm": 0.21151971817016602,
      "learning_rate": 3.849728150555161e-05,
      "loss": 1.1466,
      "step": 4880
    },
    {
      "epoch": 0.588023088023088,
      "grad_norm": 0.21175932884216309,
      "learning_rate": 3.8307849758127454e-05,
      "loss": 1.1508,
      "step": 4890
    },
    {
      "epoch": 0.5892255892255892,
      "grad_norm": 0.20844349265098572,
      "learning_rate": 3.8118595379470826e-05,
      "loss": 1.1757,
      "step": 4900
    },
    {
      "epoch": 0.5904280904280904,
      "grad_norm": 0.19040937721729279,
      "learning_rate": 3.7929521240552066e-05,
      "loss": 1.1708,
      "step": 4910
    },
    {
      "epoch": 0.5916305916305916,
      "grad_norm": 0.20293647050857544,
      "learning_rate": 3.77406302096072e-05,
      "loss": 1.1553,
      "step": 4920
    },
    {
      "epoch": 0.5928330928330928,
      "grad_norm": 0.21380431950092316,
      "learning_rate": 3.755192515209458e-05,
      "loss": 1.1618,
      "step": 4930
    },
    {
      "epoch": 0.594035594035594,
      "grad_norm": 0.2078305035829544,
      "learning_rate": 3.7363408930651326e-05,
      "loss": 1.1498,
      "step": 4940
    },
    {
      "epoch": 0.5952380952380952,
      "grad_norm": 0.19942502677440643,
      "learning_rate": 3.7175084405049975e-05,
      "loss": 1.1555,
      "step": 4950
    },
    {
      "epoch": 0.5964405964405964,
      "grad_norm": 0.2114267349243164,
      "learning_rate": 3.6986954432155e-05,
      "loss": 1.1674,
      "step": 4960
    },
    {
      "epoch": 0.5976430976430976,
      "grad_norm": 0.19917641580104828,
      "learning_rate": 3.679902186587959e-05,
      "loss": 1.1687,
      "step": 4970
    },
    {
      "epoch": 0.5988455988455988,
      "grad_norm": 0.2174818068742752,
      "learning_rate": 3.6611289557142226e-05,
      "loss": 1.1681,
      "step": 4980
    },
    {
      "epoch": 0.6000481000481,
      "grad_norm": 0.193511962890625,
      "learning_rate": 3.642376035382359e-05,
      "loss": 1.1653,
      "step": 4990
    },
    {
      "epoch": 0.6012506012506013,
      "grad_norm": 0.2069484144449234,
      "learning_rate": 3.6236437100723194e-05,
      "loss": 1.1723,
      "step": 5000
    },
    {
      "epoch": 0.6012506012506013,
      "eval_loss": 1.1470744609832764,
      "eval_runtime": 14.362,
      "eval_samples_per_second": 17.825,
      "eval_steps_per_second": 0.557,
      "step": 5000
    }
  ],
  "logging_steps": 10,
  "max_steps": 8316,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 1000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 4.858009161721446e+19,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}