{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.134162834730863,
  "eval_steps": 500,
  "global_step": 7890,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.002704895861509332,
      "grad_norm": 1.0350933074951172,
      "learning_rate": 9e-06,
      "loss": 0.77,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 10
    },
    {
      "epoch": 0.005409791723018664,
      "grad_norm": 0.9031785726547241,
      "learning_rate": 1.9e-05,
      "loss": 0.5401,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 20
    },
    {
      "epoch": 0.008114687584527996,
      "grad_norm": 0.9268996119499207,
      "learning_rate": 2.9e-05,
      "loss": 0.5503,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 30
    },
    {
      "epoch": 0.010819583446037328,
      "grad_norm": 0.7162524461746216,
      "learning_rate": 3.9000000000000006e-05,
      "loss": 0.5374,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 40
    },
    {
      "epoch": 0.013524479307546659,
      "grad_norm": 0.7313466668128967,
      "learning_rate": 4.9e-05,
      "loss": 0.5174,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 50
    },
    {
      "epoch": 0.016229375169055992,
      "grad_norm": 0.7597206830978394,
      "learning_rate": 5.9e-05,
      "loss": 0.5663,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 60
    },
    {
      "epoch": 0.018934271030565324,
      "grad_norm": 0.7689372301101685,
      "learning_rate": 6.9e-05,
      "loss": 0.5196,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 70
    },
    {
      "epoch": 0.021639166892074655,
      "grad_norm": 1.5285792350769043,
      "learning_rate": 7.900000000000001e-05,
      "loss": 0.5682,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 80
    },
    {
      "epoch": 0.024344062753583987,
      "grad_norm": 0.8935228586196899,
      "learning_rate": 8.900000000000001e-05,
      "loss": 0.5851,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 90
    },
    {
      "epoch": 0.027048958615093318,
      "grad_norm": 1.2526829242706299,
      "learning_rate": 9.900000000000001e-05,
      "loss": 0.6073,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 100
    },
    {
      "epoch": 0.02975385447660265,
      "grad_norm": 1.175955891609192,
      "learning_rate": 0.000109,
      "loss": 0.6028,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 110
    },
    {
      "epoch": 0.032458750338111984,
      "grad_norm": 1.601058840751648,
      "learning_rate": 0.000119,
      "loss": 0.6469,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 120
    },
    {
      "epoch": 0.03516364619962131,
      "grad_norm": 40.73721694946289,
      "learning_rate": 0.00012900000000000002,
      "loss": 0.6308,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 130
    },
    {
      "epoch": 0.03786854206113065,
      "grad_norm": 3.446838617324829,
      "learning_rate": 0.000139,
      "loss": 0.6569,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 140
    },
    {
      "epoch": 0.040573437922639975,
      "grad_norm": 1.7140140533447266,
      "learning_rate": 0.00014900000000000002,
      "loss": 0.6379,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 150
    },
    {
      "epoch": 0.04327833378414931,
      "grad_norm": 1.6273250579833984,
      "learning_rate": 0.00015900000000000002,
      "loss": 0.6607,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 160
    },
    {
      "epoch": 0.045983229645658645,
      "grad_norm": 2.926894426345825,
      "learning_rate": 0.00016900000000000002,
      "loss": 0.6907,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 170
    },
    {
      "epoch": 0.04868812550716797,
      "grad_norm": 2.496704578399658,
      "learning_rate": 0.00017900000000000001,
      "loss": 0.6903,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 180
    },
    {
      "epoch": 0.05139302136867731,
      "grad_norm": 2.472465753555298,
      "learning_rate": 0.00018899999999999999,
      "loss": 0.6701,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 190
    },
    {
      "epoch": 0.054097917230186636,
      "grad_norm": 2.492600440979004,
      "learning_rate": 0.000199,
      "loss": 0.7782,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 200
    },
    {
      "epoch": 0.05680281309169597,
      "grad_norm": 2.6532986164093018,
      "learning_rate": 0.00019999952684913351,
      "loss": 0.6921,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 210
    },
    {
      "epoch": 0.0595077089532053,
      "grad_norm": 3.1855242252349854,
      "learning_rate": 0.00019999789127410326,
      "loss": 0.7015,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 220
    },
    {
      "epoch": 0.062212604814714634,
      "grad_norm": 2.710665702819824,
      "learning_rate": 0.00019999508746083112,
      "loss": 0.7376,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 230
    },
    {
      "epoch": 0.06491750067622397,
      "grad_norm": 3.2956366539001465,
      "learning_rate": 0.00019999111545611174,
      "loss": 0.7482,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 240
    },
    {
      "epoch": 0.0676223965377333,
      "grad_norm": 3.2918241024017334,
      "learning_rate": 0.00019998597532623627,
      "loss": 0.7546,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 250
    },
    {
      "epoch": 0.07032729239924262,
      "grad_norm": 2.7929885387420654,
      "learning_rate": 0.0001999796671569916,
      "loss": 0.7333,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 260
    },
    {
      "epoch": 0.07303218826075196,
      "grad_norm": 2.963374376296997,
      "learning_rate": 0.00019997219105365868,
      "loss": 0.7796,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 270
    },
    {
      "epoch": 0.0757370841222613,
      "grad_norm": 4.98594856262207,
      "learning_rate": 0.0001999635471410108,
      "loss": 0.7522,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 280
    },
    {
      "epoch": 0.07844197998377063,
      "grad_norm": 4.028115749359131,
      "learning_rate": 0.0001999537355633116,
      "loss": 0.7184,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 290
    },
    {
      "epoch": 0.08114687584527995,
      "grad_norm": 3.6761696338653564,
      "learning_rate": 0.00019994275648431266,
      "loss": 0.7948,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 300
    },
    {
      "epoch": 0.08385177170678929,
      "grad_norm": 2.3481454849243164,
      "learning_rate": 0.00019993061008725049,
      "loss": 0.7413,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 310
    },
    {
      "epoch": 0.08655666756829862,
      "grad_norm": 4.488561630249023,
      "learning_rate": 0.00019991729657484398,
      "loss": 0.8014,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 320
    },
    {
      "epoch": 0.08926156342980796,
      "grad_norm": 3.144627809524536,
      "learning_rate": 0.0001999028161692905,
      "loss": 0.7638,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 330
    },
    {
      "epoch": 0.09196645929131729,
      "grad_norm": 3.1161437034606934,
      "learning_rate": 0.00019988716911226262,
      "loss": 0.8029,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 340
    },
    {
      "epoch": 0.09467135515282661,
      "grad_norm": 2.8825762271881104,
      "learning_rate": 0.00019987035566490372,
      "loss": 0.8064,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 350
    },
    {
      "epoch": 0.09737625101433595,
      "grad_norm": 3.697293758392334,
      "learning_rate": 0.00019985237610782388,
      "loss": 0.7795,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 360
    },
    {
      "epoch": 0.10008114687584528,
      "grad_norm": 4.7071404457092285,
      "learning_rate": 0.00019983323074109505,
      "loss": 0.8487,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 370
    },
    {
      "epoch": 0.10278604273735462,
      "grad_norm": 6.040957927703857,
      "learning_rate": 0.00019981291988424623,
      "loss": 0.8461,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 380
    },
    {
      "epoch": 0.10549093859886394,
      "grad_norm": 3.213670015335083,
      "learning_rate": 0.00019979144387625782,
      "loss": 0.8091,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 390
    },
    {
      "epoch": 0.10819583446037327,
      "grad_norm": 3.7140798568725586,
      "learning_rate": 0.00019976880307555624,
      "loss": 0.8146,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 400
    },
    {
      "epoch": 0.1109007303218826,
      "grad_norm": 4.564661502838135,
      "learning_rate": 0.00019974499786000788,
      "loss": 0.8005,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 410
    },
    {
      "epoch": 0.11360562618339194,
      "grad_norm": 3.8599958419799805,
      "learning_rate": 0.0001997200286269128,
      "loss": 0.8317,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 420
    },
    {
      "epoch": 0.11631052204490128,
      "grad_norm": 2.96380615234375,
      "learning_rate": 0.000199693895792998,
      "loss": 0.7838,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 430
    },
    {
      "epoch": 0.1190154179064106,
      "grad_norm": 3.082740545272827,
      "learning_rate": 0.00019966659979441055,
      "loss": 0.8341,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 440
    },
    {
      "epoch": 0.12172031376791993,
      "grad_norm": 3.56964111328125,
      "learning_rate": 0.00019963814108671027,
      "loss": 0.8391,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 450
    },
    {
      "epoch": 0.12442520962942927,
      "grad_norm": 4.308651447296143,
      "learning_rate": 0.00019960852014486226,
      "loss": 0.8382,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 460
    },
    {
      "epoch": 0.1271301054909386,
      "grad_norm": 3.098444938659668,
      "learning_rate": 0.00019957773746322874,
      "loss": 0.7951,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 470
    },
    {
      "epoch": 0.12983500135244794,
      "grad_norm": 3.7626776695251465,
      "learning_rate": 0.00019954579355556105,
      "loss": 0.8025,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 480
    },
    {
      "epoch": 0.13253989721395726,
      "grad_norm": 3.3552496433258057,
      "learning_rate": 0.00019951268895499085,
      "loss": 0.8294,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 490
    },
    {
      "epoch": 0.1352447930754666,
      "grad_norm": 3.391998052597046,
      "learning_rate": 0.00019947842421402134,
      "loss": 0.8243,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 500
    },
    {
      "epoch": 0.13794968893697593,
      "grad_norm": 2.992677927017212,
      "learning_rate": 0.00019944299990451808,
      "loss": 0.897,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 510
    },
    {
      "epoch": 0.14065458479848525,
      "grad_norm": 3.5256876945495605,
      "learning_rate": 0.00019940641661769933,
      "loss": 0.879,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 520
    },
    {
      "epoch": 0.1433594806599946,
      "grad_norm": 4.034213542938232,
      "learning_rate": 0.0001993686749641263,
      "loss": 0.8805,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 530
    },
    {
      "epoch": 0.14606437652150392,
      "grad_norm": 3.703495979309082,
      "learning_rate": 0.0001993297755736928,
      "loss": 0.8671,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 540
    },
    {
      "epoch": 0.14876927238301324,
      "grad_norm": 3.831580877304077,
      "learning_rate": 0.00019928971909561493,
      "loss": 0.8823,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 550
    },
    {
      "epoch": 0.1514741682445226,
      "grad_norm": 3.811540126800537,
      "learning_rate": 0.00019924850619842014,
      "loss": 0.8995,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 560
    },
    {
      "epoch": 0.1541790641060319,
      "grad_norm": 3.8973500728607178,
      "learning_rate": 0.00019920613756993597,
      "loss": 0.9378,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 570
    },
    {
      "epoch": 0.15688395996754126,
      "grad_norm": 3.6269121170043945,
      "learning_rate": 0.00019916261391727887,
      "loss": 0.9098,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 580
    },
    {
      "epoch": 0.15958885582905058,
      "grad_norm": 4.271853446960449,
      "learning_rate": 0.00019911793596684194,
      "loss": 0.9003,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 590
    },
    {
      "epoch": 0.1622937516905599,
      "grad_norm": 3.5396647453308105,
      "learning_rate": 0.0001990721044642833,
      "loss": 0.8791,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 600
    },
    {
      "epoch": 0.16499864755206925,
      "grad_norm": 3.55220103263855,
      "learning_rate": 0.00019902512017451333,
      "loss": 0.9134,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 610
    },
    {
      "epoch": 0.16770354341357857,
      "grad_norm": 3.247384786605835,
      "learning_rate": 0.000198976983881682,
      "loss": 0.9218,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 620
    },
    {
      "epoch": 0.17040843927508792,
      "grad_norm": 3.8020927906036377,
      "learning_rate": 0.00019892769638916576,
      "loss": 0.8763,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 630
    },
    {
      "epoch": 0.17311333513659724,
      "grad_norm": 4.78267240524292,
      "learning_rate": 0.00019887725851955415,
      "loss": 0.9472,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 640
    },
    {
      "epoch": 0.17581823099810656,
      "grad_norm": 3.5183682441711426,
      "learning_rate": 0.00019882567111463608,
      "loss": 0.8954,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 650
    },
    {
      "epoch": 0.1785231268596159,
      "grad_norm": 3.467961311340332,
      "learning_rate": 0.0001987729350353858,
      "loss": 0.9749,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 660
    },
    {
      "epoch": 0.18122802272112523,
      "grad_norm": 3.8429675102233887,
      "learning_rate": 0.00019871905116194838,
      "loss": 0.9785,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 670
    },
    {
      "epoch": 0.18393291858263458,
      "grad_norm": 3.941807270050049,
      "learning_rate": 0.00019866402039362525,
      "loss": 0.8846,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 680
    },
    {
      "epoch": 0.1866378144441439,
      "grad_norm": 3.5058398246765137,
      "learning_rate": 0.00019860784364885903,
      "loss": 0.897,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 690
    },
    {
      "epoch": 0.18934271030565322,
      "grad_norm": 3.5675482749938965,
      "learning_rate": 0.00019855052186521839,
      "loss": 0.973,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 700
    },
    {
      "epoch": 0.19204760616716257,
      "grad_norm": 4.911406517028809,
      "learning_rate": 0.000198492055999382,
      "loss": 0.9128,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 710
    },
    {
      "epoch": 0.1947525020286719,
      "grad_norm": 4.035984039306641,
      "learning_rate": 0.00019843244702712301,
      "loss": 0.9503,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 720
    },
    {
      "epoch": 0.19745739789018124,
      "grad_norm": 4.40349006652832,
      "learning_rate": 0.00019837169594329263,
      "loss": 0.9878,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 730
    },
    {
      "epoch": 0.20016229375169056,
      "grad_norm": 3.85548996925354,
      "learning_rate": 0.0001983098037618033,
      "loss": 0.9568,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 740
    },
    {
      "epoch": 0.20286718961319988,
      "grad_norm": 3.667692184448242,
      "learning_rate": 0.00019824677151561204,
      "loss": 0.9331,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 750
    },
    {
      "epoch": 0.20557208547470923,
      "grad_norm": 3.787501096725464,
      "learning_rate": 0.00019818260025670318,
      "loss": 0.9069,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 760
    },
    {
      "epoch": 0.20827698133621855,
      "grad_norm": 5.294687747955322,
      "learning_rate": 0.00019811729105607064,
      "loss": 0.9702,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 770
    },
    {
      "epoch": 0.21098187719772787,
      "grad_norm": 3.899674654006958,
      "learning_rate": 0.00019805084500370025,
      "loss": 0.9416,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 780
    },
    {
      "epoch": 0.21368677305923722,
      "grad_norm": 4.834954738616943,
      "learning_rate": 0.0001979832632085513,
      "loss": 0.9743,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 790
    },
    {
      "epoch": 0.21639166892074654,
      "grad_norm": 3.3451271057128906,
      "learning_rate": 0.00019791454679853843,
      "loss": 0.9088,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 800
    },
    {
      "epoch": 0.2190965647822559,
      "grad_norm": 3.8231406211853027,
      "learning_rate": 0.00019784469692051232,
      "loss": 0.9134,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 810
    },
    {
      "epoch": 0.2218014606437652,
      "grad_norm": 3.7224199771881104,
      "learning_rate": 0.00019777371474024108,
      "loss": 0.9601,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 820
    },
    {
      "epoch": 0.22450635650527453,
      "grad_norm": 4.090831279754639,
      "learning_rate": 0.00019770160144239016,
      "loss": 0.9174,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 830
    },
    {
      "epoch": 0.22721125236678388,
      "grad_norm": 3.9667510986328125,
      "learning_rate": 0.00019762835823050333,
      "loss": 0.9982,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 840
    },
    {
      "epoch": 0.2299161482282932,
      "grad_norm": 4.062752723693848,
      "learning_rate": 0.00019755398632698186,
      "loss": 0.894,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 850
    },
    {
      "epoch": 0.23262104408980255,
      "grad_norm": 3.72538161277771,
      "learning_rate": 0.00019747848697306472,
      "loss": 0.982,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 860
    },
    {
      "epoch": 0.23532593995131187,
      "grad_norm": 3.6619009971618652,
      "learning_rate": 0.00019740186142880734,
      "loss": 0.9428,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 870
    },
    {
      "epoch": 0.2380308358128212,
      "grad_norm": 4.779994010925293,
      "learning_rate": 0.0001973241109730611,
      "loss": 0.9814,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 880
    },
    {
      "epoch": 0.24073573167433054,
      "grad_norm": 4.4546332359313965,
      "learning_rate": 0.00019724523690345155,
      "loss": 0.9889,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 890
    },
    {
      "epoch": 0.24344062753583987,
      "grad_norm": 3.9327263832092285,
      "learning_rate": 0.00019716524053635693,
      "loss": 0.9673,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 900
    },
    {
      "epoch": 0.24614552339734921,
      "grad_norm": 4.1958465576171875,
      "learning_rate": 0.0001970841232068863,
      "loss": 0.9762,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 910
    },
    {
      "epoch": 0.24885041925885854,
      "grad_norm": 4.2037506103515625,
      "learning_rate": 0.00019700188626885714,
      "loss": 0.9728,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 920
    },
    {
      "epoch": 0.2515553151203679,
      "grad_norm": 4.760533332824707,
      "learning_rate": 0.00019691853109477268,
      "loss": 0.9676,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 930
    },
    {
      "epoch": 0.2542602109818772,
      "grad_norm": 3.573894500732422,
      "learning_rate": 0.00019683405907579914,
      "loss": 0.9757,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 940
    },
    {
      "epoch": 0.2569651068433865,
      "grad_norm": 3.1295063495635986,
      "learning_rate": 0.00019674847162174253,
      "loss": 0.9461,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 950
    },
    {
      "epoch": 0.2596700027048959,
      "grad_norm": 3.911470890045166,
      "learning_rate": 0.0001966617701610249,
      "loss": 0.9866,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 960
    },
    {
      "epoch": 0.26237489856640517,
      "grad_norm": 3.971096992492676,
      "learning_rate": 0.00019657395614066078,
      "loss": 0.9896,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 970
    },
    {
      "epoch": 0.2650797944279145,
      "grad_norm": 3.545895576477051,
      "learning_rate": 0.00019648503102623276,
      "loss": 0.9737,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 980
    },
    {
      "epoch": 0.26778469028942387,
      "grad_norm": 4.3235063552856445,
      "learning_rate": 0.00019639499630186739,
      "loss": 0.9937,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 990
    },
    {
      "epoch": 0.2704895861509332,
      "grad_norm": 4.395487308502197,
      "learning_rate": 0.00019630385347020992,
      "loss": 0.9579,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1000
    },
    {
      "epoch": 0.2731944820124425,
      "grad_norm": 4.645142555236816,
      "learning_rate": 0.00019621160405239972,
      "loss": 1.0315,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1010
    },
    {
      "epoch": 0.27589937787395186,
      "grad_norm": 3.7461893558502197,
      "learning_rate": 0.00019611824958804452,
      "loss": 0.9766,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1020
    },
    {
      "epoch": 0.2786042737354612,
      "grad_norm": 3.593327760696411,
      "learning_rate": 0.0001960237916351949,
      "loss": 1.0186,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1030
    },
    {
      "epoch": 0.2813091695969705,
      "grad_norm": 3.475508451461792,
      "learning_rate": 0.00019592823177031824,
      "loss": 0.97,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1040
    },
    {
      "epoch": 0.28401406545847985,
      "grad_norm": 3.6203982830047607,
      "learning_rate": 0.0001958315715882725,
      "loss": 0.9572,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1050
    },
    {
      "epoch": 0.2867189613199892,
      "grad_norm": 4.392990589141846,
      "learning_rate": 0.00019573381270227933,
      "loss": 0.8977,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1060
    },
    {
      "epoch": 0.2894238571814985,
      "grad_norm": 4.05267333984375,
      "learning_rate": 0.0001956349567438975,
      "loss": 1.0108,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1070
    },
    {
      "epoch": 0.29212875304300784,
      "grad_norm": 3.279749870300293,
      "learning_rate": 0.0001955350053629954,
      "loss": 0.951,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1080
    },
    {
      "epoch": 0.2948336489045172,
      "grad_norm": 3.6689045429229736,
      "learning_rate": 0.00019543396022772368,
      "loss": 0.9474,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1090
    },
    {
      "epoch": 0.2975385447660265,
      "grad_norm": 3.889159679412842,
      "learning_rate": 0.0001953318230244873,
      "loss": 0.9618,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1100
    },
    {
      "epoch": 0.30024344062753583,
      "grad_norm": 4.421080112457275,
      "learning_rate": 0.00019522859545791744,
      "loss": 0.9371,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1110
    },
    {
      "epoch": 0.3029483364890452,
      "grad_norm": 4.434532642364502,
      "learning_rate": 0.00019512427925084293,
      "loss": 0.9422,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1120
    },
    {
      "epoch": 0.3056532323505545,
      "grad_norm": 3.772916316986084,
      "learning_rate": 0.00019501887614426172,
      "loss": 1.0188,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1130
    },
    {
      "epoch": 0.3083581282120638,
      "grad_norm": 3.7862441539764404,
      "learning_rate": 0.00019491238789731176,
      "loss": 0.9497,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1140
    },
    {
      "epoch": 0.31106302407357317,
      "grad_norm": 4.361104488372803,
      "learning_rate": 0.00019480481628724133,
      "loss": 0.9602,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1150
    },
    {
      "epoch": 0.3137679199350825,
      "grad_norm": 4.093398571014404,
      "learning_rate": 0.0001946961631093799,
      "loss": 0.927,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1160
    },
    {
      "epoch": 0.3164728157965918,
      "grad_norm": 3.8494110107421875,
      "learning_rate": 0.0001945864301771077,
      "loss": 0.9768,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1170
    },
    {
      "epoch": 0.31917771165810116,
      "grad_norm": 4.4325056076049805,
      "learning_rate": 0.00019447561932182575,
      "loss": 0.948,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1180
    },
    {
      "epoch": 0.3218826075196105,
      "grad_norm": 4.117772579193115,
      "learning_rate": 0.00019436373239292514,
      "loss": 1.03,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1190
    },
    {
      "epoch": 0.3245875033811198,
      "grad_norm": 3.751755952835083,
      "learning_rate": 0.0001942507712577563,
      "loss": 1.0115,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1200
    },
    {
      "epoch": 0.32729239924262915,
      "grad_norm": 4.140499114990234,
      "learning_rate": 0.0001941367378015977,
      "loss": 0.9549,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1210
    },
    {
      "epoch": 0.3299972951041385,
      "grad_norm": 3.6646945476531982,
      "learning_rate": 0.00019402163392762446,
      "loss": 0.9667,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1220
    },
    {
      "epoch": 0.33270219096564785,
      "grad_norm": 3.5485453605651855,
      "learning_rate": 0.0001939054615568765,
      "loss": 0.9595,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1230
    },
    {
      "epoch": 0.33540708682715714,
      "grad_norm": 4.874852657318115,
      "learning_rate": 0.00019378822262822664,
      "loss": 0.9765,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1240
    },
    {
      "epoch": 0.3381119826886665,
      "grad_norm": 3.762605667114258,
      "learning_rate": 0.00019366991909834815,
      "loss": 0.9489,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1250
    },
    {
      "epoch": 0.34081687855017584,
      "grad_norm": 4.045661926269531,
      "learning_rate": 0.00019355055294168199,
      "loss": 0.9725,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1260
    },
    {
      "epoch": 0.34352177441168513,
      "grad_norm": 2.9851863384246826,
      "learning_rate": 0.00019343012615040403,
      "loss": 0.9469,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1270
    },
    {
      "epoch": 0.3462266702731945,
      "grad_norm": 5.218445777893066,
      "learning_rate": 0.00019330864073439168,
      "loss": 0.9382,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1280
    },
    {
      "epoch": 0.34893156613470383,
      "grad_norm": 4.675766944885254,
      "learning_rate": 0.00019318609872119051,
      "loss": 0.9603,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1290
    },
    {
      "epoch": 0.3516364619962131,
      "grad_norm": 3.479995012283325,
      "learning_rate": 0.00019306250215598016,
      "loss": 0.9985,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1300
    },
    {
      "epoch": 0.35434135785772247,
      "grad_norm": 4.1693196296691895,
      "learning_rate": 0.00019293785310154043,
      "loss": 0.9127,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1310
    },
    {
      "epoch": 0.3570462537192318,
      "grad_norm": 3.9952392578125,
      "learning_rate": 0.00019281215363821673,
      "loss": 1.0016,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1320
    },
    {
      "epoch": 0.3597511495807411,
      "grad_norm": 3.962355375289917,
      "learning_rate": 0.00019268540586388534,
      "loss": 0.9992,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1330
    },
    {
      "epoch": 0.36245604544225046,
      "grad_norm": 4.055516719818115,
      "learning_rate": 0.00019255761189391866,
      "loss": 0.8976,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1340
    },
    {
      "epoch": 0.3651609413037598,
      "grad_norm": 3.6345603466033936,
      "learning_rate": 0.00019242877386114934,
      "loss": 0.9189,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1350
    },
    {
      "epoch": 0.36786583716526916,
      "grad_norm": 3.9780850410461426,
      "learning_rate": 0.00019229889391583549,
      "loss": 0.9408,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1360
    },
    {
      "epoch": 0.37057073302677845,
      "grad_norm": 3.901360034942627,
      "learning_rate": 0.0001921679742256239,
      "loss": 0.9157,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1370
    },
    {
      "epoch": 0.3732756288882878,
      "grad_norm": 3.978969097137451,
      "learning_rate": 0.00019203601697551467,
      "loss": 1.0196,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1380
    },
    {
      "epoch": 0.37598052474979715,
      "grad_norm": 4.857093811035156,
      "learning_rate": 0.00019190302436782416,
      "loss": 0.9521,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1390
    },
    {
      "epoch": 0.37868542061130644,
      "grad_norm": 3.598557710647583,
      "learning_rate": 0.00019176899862214847,
      "loss": 0.9786,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1400
    },
    {
      "epoch": 0.3813903164728158,
      "grad_norm": 3.6305673122406006,
      "learning_rate": 0.00019163394197532654,
      "loss": 0.9389,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1410
    },
    {
      "epoch": 0.38409521233432514,
      "grad_norm": 4.535244464874268,
      "learning_rate": 0.00019149785668140252,
      "loss": 0.9684,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1420
    },
    {
      "epoch": 0.38680010819583444,
      "grad_norm": 3.8296279907226562,
      "learning_rate": 0.00019136074501158827,
      "loss": 0.9368,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1430
    },
    {
      "epoch": 0.3895050040573438,
      "grad_norm": 4.05790901184082,
      "learning_rate": 0.00019122260925422558,
      "loss": 0.9432,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1440
    },
    {
      "epoch": 0.39220989991885313,
      "grad_norm": 4.169580936431885,
      "learning_rate": 0.00019108345171474781,
      "loss": 1.0215,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1450
    },
    {
      "epoch": 0.3949147957803625,
      "grad_norm": 3.9768381118774414,
      "learning_rate": 0.0001909432747156415,
      "loss": 0.9545,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1460
    },
    {
      "epoch": 0.3976196916418718,
      "grad_norm": 4.347015380859375,
      "learning_rate": 0.0001908020805964075,
      "loss": 0.9757,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1470
    },
    {
      "epoch": 0.4003245875033811,
      "grad_norm": 3.8541741371154785,
      "learning_rate": 0.00019065987171352217,
      "loss": 0.9746,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1480
    },
    {
      "epoch": 0.4030294833648905,
      "grad_norm": 4.323551177978516,
      "learning_rate": 0.00019051665044039773,
      "loss": 0.9668,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1490
    },
    {
      "epoch": 0.40573437922639977,
      "grad_norm": 4.268146991729736,
      "learning_rate": 0.0001903724191673429,
      "loss": 0.9736,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1500
    },
    {
      "epoch": 0.4084392750879091,
      "grad_norm": 3.524773597717285,
      "learning_rate": 0.00019022718030152293,
      "loss": 0.9367,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1510
    },
    {
      "epoch": 0.41114417094941846,
      "grad_norm": 4.370333671569824,
      "learning_rate": 0.00019008093626691932,
      "loss": 1.0111,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1520
    },
    {
      "epoch": 0.41384906681092776,
      "grad_norm": 4.684153079986572,
      "learning_rate": 0.00018993368950428953,
      "loss": 0.9313,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1530
    },
    {
      "epoch": 0.4165539626724371,
      "grad_norm": 3.5086281299591064,
      "learning_rate": 0.00018978544247112616,
      "loss": 0.9126,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1540
    },
    {
      "epoch": 0.41925885853394645,
      "grad_norm": 4.595691204071045,
      "learning_rate": 0.00018963619764161588,
      "loss": 0.9421,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1550
    },
    {
      "epoch": 0.42196375439545575,
      "grad_norm": 4.869218826293945,
      "learning_rate": 0.00018948595750659838,
      "loss": 0.8691,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1560
    },
    {
      "epoch": 0.4246686502569651,
      "grad_norm": 4.286134719848633,
      "learning_rate": 0.00018933472457352438,
      "loss": 0.953,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1570
    },
    {
      "epoch": 0.42737354611847445,
      "grad_norm": 3.7497172355651855,
      "learning_rate": 0.00018918250136641418,
      "loss": 0.9954,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1580
    },
    {
      "epoch": 0.4300784419799838,
      "grad_norm": 3.616608142852783,
      "learning_rate": 0.00018902929042581535,
      "loss": 0.9737,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1590
    },
    {
      "epoch": 0.4327833378414931,
      "grad_norm": 4.514087200164795,
      "learning_rate": 0.00018887509430876031,
      "loss": 0.9705,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1600
    },
    {
      "epoch": 0.43548823370300244,
      "grad_norm": 3.8514137268066406,
      "learning_rate": 0.00018871991558872377,
      "loss": 0.9743,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1610
    },
    {
      "epoch": 0.4381931295645118,
      "grad_norm": 3.9302988052368164,
      "learning_rate": 0.00018856375685557972,
      "loss": 0.9343,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1620
    },
    {
      "epoch": 0.4408980254260211,
      "grad_norm": 4.061647415161133,
      "learning_rate": 0.0001884066207155581,
      "loss": 0.9115,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1630
    },
    {
      "epoch": 0.4436029212875304,
      "grad_norm": 4.214900493621826,
      "learning_rate": 0.0001882485097912014,
      "loss": 0.9684,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1640
    },
    {
      "epoch": 0.4463078171490398,
      "grad_norm": 4.488533973693848,
      "learning_rate": 0.00018808942672132107,
      "loss": 0.9074,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1650
    },
    {
      "epoch": 0.44901271301054907,
      "grad_norm": 3.918598175048828,
      "learning_rate": 0.00018792937416095313,
      "loss": 0.9515,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1660
    },
    {
      "epoch": 0.4517176088720584,
      "grad_norm": 4.252569198608398,
      "learning_rate": 0.00018776835478131405,
      "loss": 0.9649,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1670
    },
    {
      "epoch": 0.45442250473356777,
      "grad_norm": 5.241060256958008,
      "learning_rate": 0.00018760637126975614,
      "loss": 0.9627,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1680
    },
    {
      "epoch": 0.4571274005950771,
      "grad_norm": 3.8027868270874023,
      "learning_rate": 0.0001874434263297228,
      "loss": 0.967,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1690
    },
    {
      "epoch": 0.4598322964565864,
      "grad_norm": 4.088855743408203,
      "learning_rate": 0.00018727952268070324,
      "loss": 0.9693,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1700
    },
    {
      "epoch": 0.46253719231809576,
      "grad_norm": 3.044389009475708,
      "learning_rate": 0.00018711466305818716,
      "loss": 0.9721,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1710
    },
    {
      "epoch": 0.4652420881796051,
      "grad_norm": 4.544606685638428,
      "learning_rate": 0.00018694885021361906,
      "loss": 0.9702,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1720
    },
    {
      "epoch": 0.4679469840411144,
      "grad_norm": 3.730590343475342,
      "learning_rate": 0.0001867820869143524,
      "loss": 0.8969,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1730
    },
    {
      "epoch": 0.47065187990262375,
      "grad_norm": 3.8797590732574463,
      "learning_rate": 0.00018661437594360338,
      "loss": 0.9757,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1740
    },
    {
      "epoch": 0.4733567757641331,
      "grad_norm": 3.943190336227417,
      "learning_rate": 0.00018644572010040444,
      "loss": 0.9678,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1750
    },
    {
      "epoch": 0.4760616716256424,
      "grad_norm": 3.682328701019287,
      "learning_rate": 0.00018627612219955766,
      "loss": 0.9506,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1760
    },
    {
      "epoch": 0.47876656748715174,
      "grad_norm": 4.8144450187683105,
      "learning_rate": 0.00018610558507158755,
      "loss": 0.9126,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1770
    },
    {
      "epoch": 0.4814714633486611,
      "grad_norm": 4.706192493438721,
      "learning_rate": 0.00018593411156269413,
      "loss": 0.8951,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1780
    },
    {
      "epoch": 0.4841763592101704,
      "grad_norm": 3.818739414215088,
      "learning_rate": 0.00018576170453470515,
      "loss": 0.9506,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1790
    },
    {
      "epoch": 0.48688125507167973,
      "grad_norm": 4.092812538146973,
      "learning_rate": 0.0001855883668650285,
      "loss": 0.8943,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1800
    },
    {
      "epoch": 0.4895861509331891,
      "grad_norm": 4.369778633117676,
      "learning_rate": 0.00018541410144660408,
      "loss": 0.9837,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1810
    },
    {
      "epoch": 0.49229104679469843,
      "grad_norm": 4.818594932556152,
      "learning_rate": 0.00018523891118785564,
      "loss": 0.9565,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1820
    },
    {
      "epoch": 0.4949959426562077,
      "grad_norm": 3.503969430923462,
      "learning_rate": 0.00018506279901264207,
      "loss": 0.9278,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1830
    },
    {
      "epoch": 0.49770083851771707,
      "grad_norm": 3.4476447105407715,
      "learning_rate": 0.00018488576786020877,
      "loss": 0.9111,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1840
    },
    {
      "epoch": 0.5004057343792264,
      "grad_norm": 3.41809344291687,
      "learning_rate": 0.00018470782068513846,
      "loss": 0.946,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1850
    },
    {
      "epoch": 0.5031106302407358,
      "grad_norm": 4.1345367431640625,
      "learning_rate": 0.00018452896045730203,
      "loss": 0.9651,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1860
    },
    {
      "epoch": 0.5058155261022451,
      "grad_norm": 3.9654111862182617,
      "learning_rate": 0.00018434919016180876,
      "loss": 0.9694,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1870
    },
    {
      "epoch": 0.5085204219637544,
      "grad_norm": 4.497703552246094,
      "learning_rate": 0.00018416851279895666,
      "loss": 0.938,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1880
    },
    {
      "epoch": 0.5112253178252637,
      "grad_norm": 3.6628170013427734,
      "learning_rate": 0.00018398693138418234,
      "loss": 0.9124,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1890
    },
    {
      "epoch": 0.513930213686773,
      "grad_norm": 4.3850603103637695,
      "learning_rate": 0.00018380444894801072,
      "loss": 0.9353,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1900
    },
    {
      "epoch": 0.5166351095482824,
      "grad_norm": 4.043147087097168,
      "learning_rate": 0.00018362106853600442,
      "loss": 0.9037,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1910
    },
    {
      "epoch": 0.5193400054097917,
      "grad_norm": 4.230565547943115,
      "learning_rate": 0.0001834367932087129,
      "loss": 0.991,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1920
    },
    {
      "epoch": 0.5220449012713011,
      "grad_norm": 3.6754353046417236,
      "learning_rate": 0.00018325162604162145,
      "loss": 0.9589,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1930
    },
    {
      "epoch": 0.5247497971328103,
      "grad_norm": 4.023629665374756,
      "learning_rate": 0.00018306557012509985,
      "loss": 0.9183,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1940
    },
    {
      "epoch": 0.5274546929943197,
      "grad_norm": 3.9141383171081543,
      "learning_rate": 0.00018287862856435067,
      "loss": 0.958,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1950
    },
    {
      "epoch": 0.530159588855829,
      "grad_norm": 3.6754729747772217,
      "learning_rate": 0.0001826908044793577,
      "loss": 0.9775,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1960
    },
    {
      "epoch": 0.5328644847173384,
      "grad_norm": 4.166747570037842,
      "learning_rate": 0.00018250210100483356,
      "loss": 0.915,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1970
    },
    {
      "epoch": 0.5355693805788477,
      "grad_norm": 4.042303562164307,
      "learning_rate": 0.00018231252129016773,
      "loss": 0.9609,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1980
    },
    {
      "epoch": 0.5382742764403571,
      "grad_norm": 3.7295937538146973,
      "learning_rate": 0.00018212206849937357,
      "loss": 1.0206,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 1990
    },
    {
      "epoch": 0.5409791723018664,
      "grad_norm": 3.640014886856079,
      "learning_rate": 0.00018193074581103607,
      "loss": 1.0334,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2000
    },
    {
      "epoch": 0.5436840681633757,
      "grad_norm": 4.122087001800537,
      "learning_rate": 0.0001817385564182581,
      "loss": 0.9506,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2010
    },
    {
      "epoch": 0.546388964024885,
      "grad_norm": 3.576752185821533,
      "learning_rate": 0.00018154550352860778,
      "loss": 0.9081,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2020
    },
    {
      "epoch": 0.5490938598863944,
      "grad_norm": 4.507325649261475,
      "learning_rate": 0.00018135159036406458,
      "loss": 0.9845,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2030
    },
    {
      "epoch": 0.5517987557479037,
      "grad_norm": 3.9277851581573486,
      "learning_rate": 0.00018115682016096564,
      "loss": 0.9105,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2040
    },
    {
      "epoch": 0.5545036516094131,
      "grad_norm": 4.253115653991699,
      "learning_rate": 0.0001809611961699517,
      "loss": 0.9145,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2050
    },
    {
      "epoch": 0.5572085474709224,
      "grad_norm": 4.563695430755615,
      "learning_rate": 0.0001807647216559129,
      "loss": 0.8965,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2060
    },
    {
      "epoch": 0.5599134433324316,
      "grad_norm": 3.6539463996887207,
      "learning_rate": 0.00018056739989793442,
      "loss": 0.9756,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2070
    },
    {
      "epoch": 0.562618339193941,
      "grad_norm": 4.442514896392822,
      "learning_rate": 0.00018036923418924148,
      "loss": 0.9717,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2080
    },
    {
      "epoch": 0.5653232350554503,
      "grad_norm": 4.4010820388793945,
      "learning_rate": 0.00018017022783714453,
      "loss": 0.93,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2090
    },
    {
      "epoch": 0.5680281309169597,
      "grad_norm": 3.412588596343994,
      "learning_rate": 0.00017997038416298406,
      "loss": 0.9096,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2100
    },
    {
      "epoch": 0.570733026778469,
      "grad_norm": 5.173451900482178,
      "learning_rate": 0.0001797697065020752,
      "loss": 0.9684,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2110
    },
    {
      "epoch": 0.5734379226399784,
      "grad_norm": 3.7624127864837646,
      "learning_rate": 0.00017956819820365188,
      "loss": 0.9272,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2120
    },
    {
      "epoch": 0.5761428185014877,
      "grad_norm": 4.000368595123291,
      "learning_rate": 0.0001793658626308112,
      "loss": 0.9029,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2130
    },
    {
      "epoch": 0.578847714362997,
      "grad_norm": 4.823195934295654,
      "learning_rate": 0.00017916270316045707,
      "loss": 0.9308,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2140
    },
    {
      "epoch": 0.5815526102245063,
      "grad_norm": 3.2766668796539307,
      "learning_rate": 0.00017895872318324392,
      "loss": 0.9648,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2150
    },
    {
      "epoch": 0.5842575060860157,
      "grad_norm": 4.70221471786499,
      "learning_rate": 0.0001787539261035202,
      "loss": 0.9736,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2160
    },
    {
      "epoch": 0.586962401947525,
      "grad_norm": 4.25441312789917,
      "learning_rate": 0.00017854831533927149,
      "loss": 0.9786,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2170
    },
    {
      "epoch": 0.5896672978090344,
      "grad_norm": 4.238402366638184,
      "learning_rate": 0.00017834189432206328,
      "loss": 0.9267,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2180
    },
    {
      "epoch": 0.5923721936705437,
      "grad_norm": 4.358950138092041,
      "learning_rate": 0.00017813466649698416,
      "loss": 0.9864,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2190
    },
    {
      "epoch": 0.595077089532053,
      "grad_norm": 4.879393100738525,
      "learning_rate": 0.00017792663532258782,
      "loss": 0.9704,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2200
    },
    {
      "epoch": 0.5977819853935623,
      "grad_norm": 5.087660312652588,
      "learning_rate": 0.00017771780427083562,
      "loss": 0.9434,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2210
    },
    {
      "epoch": 0.6004868812550717,
      "grad_norm": 3.972977876663208,
      "learning_rate": 0.00017750817682703858,
      "loss": 0.8854,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2220
    },
    {
      "epoch": 0.603191777116581,
      "grad_norm": 4.493955135345459,
      "learning_rate": 0.00017729775648979917,
      "loss": 0.8902,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2230
    },
    {
      "epoch": 0.6058966729780904,
      "grad_norm": 4.631348609924316,
      "learning_rate": 0.00017708654677095297,
      "loss": 0.9572,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2240
    },
    {
      "epoch": 0.6086015688395997,
      "grad_norm": 4.689295291900635,
      "learning_rate": 0.00017687455119551,
      "loss": 0.9386,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2250
    },
    {
      "epoch": 0.611306464701109,
      "grad_norm": 4.065158843994141,
      "learning_rate": 0.00017666177330159606,
      "loss": 0.9338,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2260
    },
    {
      "epoch": 0.6140113605626183,
      "grad_norm": 4.302713871002197,
      "learning_rate": 0.00017644821664039334,
      "loss": 0.9198,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2270
    },
    {
      "epoch": 0.6167162564241276,
      "grad_norm": 4.278186798095703,
      "learning_rate": 0.00017623388477608156,
      "loss": 0.916,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2280
    },
    {
      "epoch": 0.619421152285637,
      "grad_norm": 4.147995948791504,
      "learning_rate": 0.00017601878128577813,
      "loss": 0.9172,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2290
    },
    {
      "epoch": 0.6221260481471463,
      "grad_norm": 3.960292339324951,
      "learning_rate": 0.00017580290975947874,
      "loss": 0.9322,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2300
    },
    {
      "epoch": 0.6248309440086557,
      "grad_norm": 5.553867816925049,
      "learning_rate": 0.0001755862737999972,
      "loss": 0.9036,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2310
    },
    {
      "epoch": 0.627535839870165,
      "grad_norm": 5.008589267730713,
      "learning_rate": 0.00017536887702290552,
      "loss": 0.8956,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2320
    },
    {
      "epoch": 0.6302407357316744,
      "grad_norm": 4.052738189697266,
      "learning_rate": 0.00017515072305647329,
      "loss": 0.9053,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2330
    },
    {
      "epoch": 0.6329456315931836,
      "grad_norm": 3.691368818283081,
      "learning_rate": 0.00017493181554160752,
      "loss": 0.9348,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2340
    },
    {
      "epoch": 0.635650527454693,
      "grad_norm": 4.584995269775391,
      "learning_rate": 0.0001747121581317915,
      "loss": 0.9534,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2350
    },
    {
      "epoch": 0.6383554233162023,
      "grad_norm": 4.836928367614746,
      "learning_rate": 0.00017449175449302416,
      "loss": 0.9388,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2360
    },
    {
      "epoch": 0.6410603191777117,
      "grad_norm": 4.251411437988281,
      "learning_rate": 0.00017427060830375846,
      "loss": 0.9467,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2370
    },
    {
      "epoch": 0.643765215039221,
      "grad_norm": 3.808366298675537,
      "learning_rate": 0.0001740487232548404,
      "loss": 0.909,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2380
    },
    {
      "epoch": 0.6464701109007304,
      "grad_norm": 3.5536386966705322,
      "learning_rate": 0.00017382610304944728,
      "loss": 0.9405,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2390
    },
    {
      "epoch": 0.6491750067622396,
      "grad_norm": 4.283199787139893,
      "learning_rate": 0.0001736027514030259,
      "loss": 0.925,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2400
    },
    {
      "epoch": 0.651879902623749,
      "grad_norm": 4.239474773406982,
      "learning_rate": 0.00017337867204323036,
      "loss": 0.9387,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2410
    },
    {
      "epoch": 0.6545847984852583,
      "grad_norm": 4.415828704833984,
      "learning_rate": 0.00017315386870986018,
      "loss": 0.8791,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2420
    },
    {
      "epoch": 0.6572896943467677,
      "grad_norm": 3.615339517593384,
      "learning_rate": 0.0001729283451547977,
      "loss": 0.9229,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2430
    },
    {
      "epoch": 0.659994590208277,
      "grad_norm": 3.2011773586273193,
      "learning_rate": 0.00017270210514194542,
      "loss": 0.9149,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2440
    },
    {
      "epoch": 0.6626994860697863,
      "grad_norm": 3.553722858428955,
      "learning_rate": 0.0001724751524471634,
      "loss": 0.9084,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2450
    },
    {
      "epoch": 0.6654043819312957,
      "grad_norm": 4.793551921844482,
      "learning_rate": 0.0001722474908582059,
      "loss": 0.8596,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2460
    },
    {
      "epoch": 0.6681092777928049,
      "grad_norm": 4.133306503295898,
      "learning_rate": 0.00017201912417465852,
      "loss": 0.9791,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2470
    },
    {
      "epoch": 0.6708141736543143,
      "grad_norm": 4.8816423416137695,
      "learning_rate": 0.00017179005620787446,
      "loss": 0.9246,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2480
    },
    {
      "epoch": 0.6735190695158236,
      "grad_norm": 4.403787136077881,
      "learning_rate": 0.00017156029078091125,
      "loss": 0.9482,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2490
    },
    {
      "epoch": 0.676223965377333,
      "grad_norm": 4.04508113861084,
      "learning_rate": 0.00017132983172846658,
      "loss": 0.9313,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2500
    },
    {
      "epoch": 0.6789288612388423,
      "grad_norm": 4.3617424964904785,
      "learning_rate": 0.00017109868289681462,
      "loss": 0.8771,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2510
    },
    {
      "epoch": 0.6816337571003517,
      "grad_norm": 3.2879831790924072,
      "learning_rate": 0.00017086684814374164,
      "loss": 0.8953,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2520
    },
    {
      "epoch": 0.6843386529618609,
      "grad_norm": 4.465785503387451,
      "learning_rate": 0.00017063433133848176,
      "loss": 0.9863,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2530
    },
    {
      "epoch": 0.6870435488233703,
      "grad_norm": 3.3583545684814453,
      "learning_rate": 0.00017040113636165212,
      "loss": 0.9106,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2540
    },
    {
      "epoch": 0.6897484446848796,
      "grad_norm": 4.483101844787598,
      "learning_rate": 0.00017016726710518843,
      "loss": 0.9296,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2550
    },
    {
      "epoch": 0.692453340546389,
      "grad_norm": 4.261110782623291,
      "learning_rate": 0.00016993272747227982,
      "loss": 0.9471,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2560
    },
    {
      "epoch": 0.6951582364078983,
      "grad_norm": 4.205190181732178,
      "learning_rate": 0.0001696975213773038,
      "loss": 0.8885,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2570
    },
    {
      "epoch": 0.6978631322694077,
      "grad_norm": 3.7448229789733887,
      "learning_rate": 0.00016946165274576077,
      "loss": 0.8821,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2580
    },
    {
      "epoch": 0.700568028130917,
      "grad_norm": 3.6259069442749023,
      "learning_rate": 0.00016922512551420876,
      "loss": 0.9047,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2590
    },
    {
      "epoch": 0.7032729239924262,
      "grad_norm": 3.925834894180298,
      "learning_rate": 0.00016898794363019742,
      "loss": 0.8879,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2600
    },
    {
      "epoch": 0.7059778198539356,
      "grad_norm": 4.769133567810059,
      "learning_rate": 0.00016875011105220246,
      "loss": 0.8736,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2610
    },
    {
      "epoch": 0.7086827157154449,
      "grad_norm": 3.497141122817993,
      "learning_rate": 0.00016851163174955928,
      "loss": 0.8627,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2620
    },
    {
      "epoch": 0.7113876115769543,
      "grad_norm": 4.127130508422852,
      "learning_rate": 0.00016827250970239703,
      "loss": 0.879,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2630
    },
    {
      "epoch": 0.7140925074384636,
      "grad_norm": 4.241309642791748,
      "learning_rate": 0.00016803274890157184,
      "loss": 0.9131,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2640
    },
    {
      "epoch": 0.716797403299973,
      "grad_norm": 4.338650226593018,
      "learning_rate": 0.00016779235334860054,
      "loss": 0.8861,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2650
    },
    {
      "epoch": 0.7195022991614822,
      "grad_norm": 3.8803446292877197,
      "learning_rate": 0.00016755132705559365,
      "loss": 0.8902,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2660
    },
    {
      "epoch": 0.7222071950229916,
      "grad_norm": 3.8507978916168213,
      "learning_rate": 0.00016730967404518862,
      "loss": 0.9076,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2670
    },
    {
      "epoch": 0.7249120908845009,
      "grad_norm": 4.108922958374023,
      "learning_rate": 0.00016706739835048237,
      "loss": 0.9534,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2680
    },
    {
      "epoch": 0.7276169867460103,
      "grad_norm": 3.700160026550293,
      "learning_rate": 0.00016682450401496435,
      "loss": 0.9344,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2690
    },
    {
      "epoch": 0.7303218826075196,
      "grad_norm": 4.938565254211426,
      "learning_rate": 0.0001665809950924489,
      "loss": 0.9271,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2700
    },
    {
      "epoch": 0.733026778469029,
      "grad_norm": 3.996375799179077,
      "learning_rate": 0.00016633687564700743,
      "loss": 0.8578,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2710
    },
    {
      "epoch": 0.7357316743305383,
      "grad_norm": 4.022902488708496,
      "learning_rate": 0.00016609214975290092,
      "loss": 0.9398,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2720
    },
    {
      "epoch": 0.7384365701920476,
      "grad_norm": 4.327158451080322,
      "learning_rate": 0.00016584682149451164,
      "loss": 0.9023,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2730
    },
    {
      "epoch": 0.7411414660535569,
      "grad_norm": 3.8225085735321045,
      "learning_rate": 0.00016560089496627512,
      "loss": 0.8965,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2740
    },
    {
      "epoch": 0.7438463619150663,
      "grad_norm": 4.230526447296143,
      "learning_rate": 0.0001653543742726118,
      "loss": 0.9207,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2750
    },
    {
      "epoch": 0.7465512577765756,
      "grad_norm": 4.1174445152282715,
      "learning_rate": 0.0001651072635278585,
      "loss": 0.9202,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2760
    },
    {
      "epoch": 0.749256153638085,
      "grad_norm": 3.8350272178649902,
      "learning_rate": 0.00016485956685619976,
      "loss": 0.8586,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2770
    },
    {
      "epoch": 0.7519610494995943,
      "grad_norm": 4.7064619064331055,
      "learning_rate": 0.00016461128839159907,
      "loss": 0.9,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2780
    },
    {
      "epoch": 0.7546659453611037,
      "grad_norm": 3.764288902282715,
      "learning_rate": 0.00016436243227772973,
      "loss": 0.9292,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2790
    },
    {
      "epoch": 0.7573708412226129,
      "grad_norm": 3.576848030090332,
      "learning_rate": 0.00016411300266790595,
      "loss": 0.9101,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2800
    },
    {
      "epoch": 0.7600757370841222,
      "grad_norm": 4.435233116149902,
      "learning_rate": 0.0001638630037250131,
      "loss": 0.892,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2810
    },
    {
      "epoch": 0.7627806329456316,
      "grad_norm": 4.196038722991943,
      "learning_rate": 0.00016361243962143883,
      "loss": 0.9169,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2820
    },
    {
      "epoch": 0.7654855288071409,
      "grad_norm": 3.818711280822754,
      "learning_rate": 0.0001633613145390028,
      "loss": 0.9098,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2830
    },
    {
      "epoch": 0.7681904246686503,
      "grad_norm": 4.29703426361084,
      "learning_rate": 0.00016310963266888743,
      "loss": 0.8704,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2840
    },
    {
      "epoch": 0.7708953205301596,
      "grad_norm": 4.465641021728516,
      "learning_rate": 0.0001628573982115676,
      "loss": 0.9254,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2850
    },
    {
      "epoch": 0.7736002163916689,
      "grad_norm": 3.674140214920044,
      "learning_rate": 0.00016260461537674062,
      "loss": 0.8769,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2860
    },
    {
      "epoch": 0.7763051122531782,
      "grad_norm": 3.6968114376068115,
      "learning_rate": 0.00016235128838325614,
      "loss": 0.9283,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2870
    },
    {
      "epoch": 0.7790100081146876,
      "grad_norm": 3.758517265319824,
      "learning_rate": 0.00016209742145904555,
      "loss": 0.8623,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2880
    },
    {
      "epoch": 0.7817149039761969,
      "grad_norm": 4.8879289627075195,
      "learning_rate": 0.0001618430188410515,
      "loss": 0.9302,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2890
    },
    {
      "epoch": 0.7844197998377063,
      "grad_norm": 4.4202880859375,
      "learning_rate": 0.00016158808477515705,
      "loss": 0.8919,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2900
    },
    {
      "epoch": 0.7871246956992156,
      "grad_norm": 3.6918702125549316,
      "learning_rate": 0.00016133262351611505,
      "loss": 0.9361,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2910
    },
    {
      "epoch": 0.789829591560725,
      "grad_norm": 4.513911724090576,
      "learning_rate": 0.00016107663932747701,
      "loss": 0.9391,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2920
    },
    {
      "epoch": 0.7925344874222342,
      "grad_norm": 4.0372724533081055,
      "learning_rate": 0.00016082013648152188,
      "loss": 0.9158,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2930
    },
    {
      "epoch": 0.7952393832837436,
      "grad_norm": 4.525071144104004,
      "learning_rate": 0.00016056311925918476,
      "loss": 0.8931,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2940
    },
    {
      "epoch": 0.7979442791452529,
      "grad_norm": 4.40305233001709,
      "learning_rate": 0.0001603055919499857,
      "loss": 0.857,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2950
    },
    {
      "epoch": 0.8006491750067622,
      "grad_norm": 3.505654811859131,
      "learning_rate": 0.00016004755885195757,
      "loss": 0.8556,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2960
    },
    {
      "epoch": 0.8033540708682716,
      "grad_norm": 4.598380088806152,
      "learning_rate": 0.00015978902427157498,
      "loss": 0.8906,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2970
    },
    {
      "epoch": 0.806058966729781,
      "grad_norm": 4.171894550323486,
      "learning_rate": 0.00015952999252368195,
      "loss": 0.8746,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2980
    },
    {
      "epoch": 0.8087638625912902,
      "grad_norm": 3.8262321949005127,
      "learning_rate": 0.00015927046793142002,
      "loss": 0.9707,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 2990
    },
    {
      "epoch": 0.8114687584527995,
      "grad_norm": 5.524593830108643,
      "learning_rate": 0.00015901045482615616,
      "loss": 0.855,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3000
    },
    {
      "epoch": 0.8141736543143089,
      "grad_norm": 4.200673580169678,
      "learning_rate": 0.00015874995754741042,
      "loss": 0.8655,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3010
    },
    {
      "epoch": 0.8168785501758182,
      "grad_norm": 4.285470962524414,
      "learning_rate": 0.00015848898044278355,
      "loss": 0.8564,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3020
    },
    {
      "epoch": 0.8195834460373276,
      "grad_norm": 4.088680267333984,
      "learning_rate": 0.0001582275278678844,
      "loss": 0.8721,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3030
    },
    {
      "epoch": 0.8222883418988369,
      "grad_norm": 4.3820881843566895,
      "learning_rate": 0.00015796560418625716,
      "loss": 0.9223,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3040
    },
    {
      "epoch": 0.8249932377603463,
      "grad_norm": 4.356407642364502,
      "learning_rate": 0.00015770321376930874,
      "loss": 0.8425,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3050
    },
    {
      "epoch": 0.8276981336218555,
      "grad_norm": 3.508406400680542,
      "learning_rate": 0.00015744036099623557,
      "loss": 0.9029,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3060
    },
    {
      "epoch": 0.8304030294833649,
      "grad_norm": 4.378997325897217,
      "learning_rate": 0.00015717705025395078,
      "loss": 0.9074,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3070
    },
    {
      "epoch": 0.8331079253448742,
      "grad_norm": 4.8483147621154785,
      "learning_rate": 0.0001569132859370106,
      "loss": 0.8682,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3080
    },
    {
      "epoch": 0.8358128212063836,
      "grad_norm": 4.2598137855529785,
      "learning_rate": 0.00015664907244754148,
      "loss": 0.9522,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3090
    },
    {
      "epoch": 0.8385177170678929,
      "grad_norm": 5.107699871063232,
      "learning_rate": 0.00015638441419516628,
      "loss": 0.8591,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3100
    },
    {
      "epoch": 0.8412226129294023,
      "grad_norm": 3.608156681060791,
      "learning_rate": 0.00015611931559693075,
      "loss": 0.8637,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3110
    },
    {
      "epoch": 0.8439275087909115,
      "grad_norm": 4.238269329071045,
      "learning_rate": 0.00015585378107722995,
      "loss": 0.8546,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3120
    },
    {
      "epoch": 0.8466324046524208,
      "grad_norm": 4.323466777801514,
      "learning_rate": 0.00015558781506773425,
      "loss": 0.8666,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3130
    },
    {
      "epoch": 0.8493373005139302,
      "grad_norm": 4.08201789855957,
      "learning_rate": 0.00015532142200731534,
      "loss": 0.8613,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3140
    },
    {
      "epoch": 0.8520421963754395,
      "grad_norm": 4.612802028656006,
      "learning_rate": 0.00015505460634197238,
      "loss": 0.8589,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3150
    },
    {
      "epoch": 0.8547470922369489,
      "grad_norm": 4.59390115737915,
      "learning_rate": 0.00015478737252475755,
      "loss": 0.9075,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3160
    },
    {
      "epoch": 0.8574519880984582,
      "grad_norm": 4.590249538421631,
      "learning_rate": 0.00015451972501570181,
      "loss": 0.8674,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3170
    },
    {
      "epoch": 0.8601568839599676,
      "grad_norm": 4.318239212036133,
      "learning_rate": 0.0001542516682817406,
      "loss": 0.9119,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3180
    },
    {
      "epoch": 0.8628617798214768,
      "grad_norm": 4.577765464782715,
      "learning_rate": 0.000153983206796639,
      "loss": 0.8367,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3190
    },
    {
      "epoch": 0.8655666756829862,
      "grad_norm": 4.629631996154785,
      "learning_rate": 0.00015371434504091739,
      "loss": 0.8939,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3200
    },
    {
      "epoch": 0.8682715715444955,
      "grad_norm": 3.954927921295166,
      "learning_rate": 0.00015344508750177632,
      "loss": 0.8676,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3210
    },
    {
      "epoch": 0.8709764674060049,
      "grad_norm": 3.5843007564544678,
      "learning_rate": 0.00015317543867302208,
      "loss": 0.877,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3220
    },
    {
      "epoch": 0.8736813632675142,
      "grad_norm": 4.194819927215576,
      "learning_rate": 0.00015290540305499124,
      "loss": 0.8487,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3230
    },
    {
      "epoch": 0.8763862591290236,
      "grad_norm": 4.4585065841674805,
      "learning_rate": 0.00015263498515447575,
      "loss": 0.9002,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3240
    },
    {
      "epoch": 0.8790911549905329,
      "grad_norm": 4.3148722648620605,
      "learning_rate": 0.00015236418948464783,
      "loss": 0.8892,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3250
    },
    {
      "epoch": 0.8817960508520422,
      "grad_norm": 3.912666082382202,
      "learning_rate": 0.00015209302056498432,
      "loss": 0.9027,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3260
    },
    {
      "epoch": 0.8845009467135515,
      "grad_norm": 3.5493319034576416,
      "learning_rate": 0.00015182148292119173,
      "loss": 0.8744,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3270
    },
    {
      "epoch": 0.8872058425750609,
      "grad_norm": 3.8175151348114014,
      "learning_rate": 0.00015154958108513022,
      "loss": 0.899,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3280
    },
    {
      "epoch": 0.8899107384365702,
      "grad_norm": 4.987168788909912,
      "learning_rate": 0.00015127731959473822,
      "loss": 0.8695,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3290
    },
    {
      "epoch": 0.8926156342980796,
      "grad_norm": 4.610702991485596,
      "learning_rate": 0.00015100470299395671,
      "loss": 0.8272,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3300
    },
    {
      "epoch": 0.8953205301595889,
      "grad_norm": 3.404474973678589,
      "learning_rate": 0.00015073173583265333,
      "loss": 0.9312,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3310
    },
    {
      "epoch": 0.8980254260210981,
      "grad_norm": 4.291271209716797,
      "learning_rate": 0.0001504584226665464,
      "loss": 0.8663,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3320
    },
    {
      "epoch": 0.9007303218826075,
      "grad_norm": 4.307567119598389,
      "learning_rate": 0.00015018476805712892,
      "loss": 0.8891,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3330
    },
    {
      "epoch": 0.9034352177441168,
      "grad_norm": 4.637225151062012,
      "learning_rate": 0.00014991077657159244,
      "loss": 0.8879,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3340
    },
    {
      "epoch": 0.9061401136056262,
      "grad_norm": 4.115970611572266,
      "learning_rate": 0.00014963645278275096,
      "loss": 0.8894,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3350
    },
    {
      "epoch": 0.9088450094671355,
      "grad_norm": 4.381041049957275,
      "learning_rate": 0.00014936180126896438,
      "loss": 0.8427,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3360
    },
    {
      "epoch": 0.9115499053286449,
      "grad_norm": 4.984541416168213,
      "learning_rate": 0.0001490868266140622,
      "loss": 0.8766,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3370
    },
    {
      "epoch": 0.9142548011901542,
      "grad_norm": 4.789724349975586,
      "learning_rate": 0.0001488115334072671,
      "loss": 0.9049,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3380
    },
    {
      "epoch": 0.9169596970516635,
      "grad_norm": 5.119272232055664,
      "learning_rate": 0.00014853592624311824,
      "loss": 0.8882,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3390
    },
    {
      "epoch": 0.9196645929131728,
      "grad_norm": 5.7028350830078125,
      "learning_rate": 0.0001482600097213946,
      "loss": 0.8511,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3400
    },
    {
      "epoch": 0.9223694887746822,
      "grad_norm": 3.9191958904266357,
      "learning_rate": 0.00014798378844703818,
      "loss": 0.8531,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3410
    },
    {
      "epoch": 0.9250743846361915,
      "grad_norm": 4.711814880371094,
      "learning_rate": 0.00014770726703007724,
      "loss": 0.8372,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3420
    },
    {
      "epoch": 0.9277792804977009,
      "grad_norm": 4.645802021026611,
      "learning_rate": 0.00014743045008554928,
      "loss": 0.8881,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3430
    },
    {
      "epoch": 0.9304841763592102,
      "grad_norm": 4.601795196533203,
      "learning_rate": 0.0001471533422334241,
      "loss": 0.909,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3440
    },
    {
      "epoch": 0.9331890722207195,
      "grad_norm": 5.0484089851379395,
      "learning_rate": 0.00014687594809852656,
      "loss": 0.8879,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3450
    },
    {
      "epoch": 0.9358939680822288,
      "grad_norm": 4.024141788482666,
      "learning_rate": 0.00014659827231045954,
      "loss": 0.8559,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3460
    },
    {
      "epoch": 0.9385988639437381,
      "grad_norm": 4.726752758026123,
      "learning_rate": 0.0001463203195035265,
      "loss": 0.8549,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3470
    },
    {
      "epoch": 0.9413037598052475,
      "grad_norm": 4.7358832359313965,
      "learning_rate": 0.00014604209431665432,
      "loss": 0.8716,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3480
    },
    {
      "epoch": 0.9440086556667568,
      "grad_norm": 4.073987007141113,
      "learning_rate": 0.00014576360139331585,
      "loss": 0.8423,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3490
    },
    {
      "epoch": 0.9467135515282662,
      "grad_norm": 3.930253505706787,
      "learning_rate": 0.0001454848453814522,
      "loss": 0.8472,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3500
    },
    {
      "epoch": 0.9494184473897755,
      "grad_norm": 4.874025344848633,
      "learning_rate": 0.00014520583093339558,
      "loss": 0.8441,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3510
    },
    {
      "epoch": 0.9521233432512848,
      "grad_norm": 4.499694347381592,
      "learning_rate": 0.00014492656270579106,
      "loss": 0.7919,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3520
    },
    {
      "epoch": 0.9548282391127941,
      "grad_norm": 5.934467792510986,
      "learning_rate": 0.0001446470453595195,
      "loss": 0.9242,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3530
    },
    {
      "epoch": 0.9575331349743035,
      "grad_norm": 4.869564056396484,
      "learning_rate": 0.00014436728355961926,
      "loss": 0.9077,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3540
    },
    {
      "epoch": 0.9602380308358128,
      "grad_norm": 4.624287128448486,
      "learning_rate": 0.0001440872819752087,
      "loss": 0.8362,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3550
    },
    {
      "epoch": 0.9629429266973222,
      "grad_norm": 4.891798973083496,
      "learning_rate": 0.00014380704527940788,
      "loss": 0.8991,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3560
    },
    {
      "epoch": 0.9656478225588315,
      "grad_norm": 5.056575775146484,
      "learning_rate": 0.00014352657814926104,
      "loss": 0.884,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3570
    },
    {
      "epoch": 0.9683527184203408,
      "grad_norm": 4.0394392013549805,
      "learning_rate": 0.0001432458852656581,
      "loss": 0.8303,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3580
    },
    {
      "epoch": 0.9710576142818501,
      "grad_norm": 3.743896245956421,
      "learning_rate": 0.00014296497131325672,
      "loss": 0.8771,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3590
    },
    {
      "epoch": 0.9737625101433595,
      "grad_norm": 4.635112285614014,
      "learning_rate": 0.00014268384098040424,
      "loss": 0.8848,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3600
    },
    {
      "epoch": 0.9764674060048688,
      "grad_norm": 3.768984079360962,
      "learning_rate": 0.00014240249895905915,
      "loss": 0.8319,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3610
    },
    {
      "epoch": 0.9791723018663782,
      "grad_norm": 3.808288335800171,
      "learning_rate": 0.0001421209499447131,
      "loss": 0.7713,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3620
    },
    {
      "epoch": 0.9818771977278875,
      "grad_norm": 4.7236328125,
      "learning_rate": 0.0001418391986363122,
      "loss": 0.8537,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3630
    },
    {
      "epoch": 0.9845820935893969,
      "grad_norm": 4.917244911193848,
      "learning_rate": 0.00014155724973617892,
      "loss": 0.8364,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3640
    },
    {
      "epoch": 0.9872869894509061,
      "grad_norm": 4.437404155731201,
      "learning_rate": 0.00014127510794993345,
      "loss": 0.8432,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3650
    },
    {
      "epoch": 0.9899918853124154,
      "grad_norm": 4.099413871765137,
      "learning_rate": 0.00014099277798641506,
      "loss": 0.8487,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3660
    },
    {
      "epoch": 0.9926967811739248,
      "grad_norm": 3.7880699634552,
      "learning_rate": 0.00014071026455760372,
      "loss": 0.8619,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3670
    },
    {
      "epoch": 0.9954016770354341,
      "grad_norm": 3.967851161956787,
      "learning_rate": 0.00014042757237854136,
      "loss": 0.8427,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3680
    },
    {
      "epoch": 0.9981065728969435,
      "grad_norm": 5.384731769561768,
      "learning_rate": 0.00014014470616725312,
      "loss": 0.8482,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3690
    },
    {
      "epoch": 1.0008114687584528,
      "grad_norm": 5.448690414428711,
      "learning_rate": 0.00013986167064466884,
      "loss": 0.8786,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3700
    },
    {
      "epoch": 1.003516364619962,
      "grad_norm": 4.6175360679626465,
      "learning_rate": 0.00013957847053454388,
      "loss": 0.8234,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3710
    },
    {
      "epoch": 1.0062212604814715,
      "grad_norm": 4.086219787597656,
      "learning_rate": 0.00013929511056338064,
      "loss": 0.7806,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3720
    },
    {
      "epoch": 1.0089261563429808,
      "grad_norm": 4.726855278015137,
      "learning_rate": 0.0001390115954603495,
      "loss": 0.8006,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3730
    },
    {
      "epoch": 1.0116310522044902,
      "grad_norm": 3.129750967025757,
      "learning_rate": 0.00013872792995720987,
      "loss": 0.8287,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3740
    },
    {
      "epoch": 1.0143359480659995,
      "grad_norm": 4.0198564529418945,
      "learning_rate": 0.00013844411878823147,
      "loss": 0.8147,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3750
    },
    {
      "epoch": 1.0170408439275087,
      "grad_norm": 3.521083354949951,
      "learning_rate": 0.00013816016669011494,
      "loss": 0.8711,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3760
    },
    {
      "epoch": 1.0197457397890182,
      "grad_norm": 4.568667888641357,
      "learning_rate": 0.00013787607840191305,
      "loss": 0.8392,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3770
    },
    {
      "epoch": 1.0224506356505274,
      "grad_norm": 3.5283639430999756,
      "learning_rate": 0.0001375918586649515,
      "loss": 0.8921,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3780
    },
    {
      "epoch": 1.0251555315120369,
      "grad_norm": 3.8672685623168945,
      "learning_rate": 0.00013730751222274992,
      "loss": 0.8141,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3790
    },
    {
      "epoch": 1.027860427373546,
      "grad_norm": 4.737693786621094,
      "learning_rate": 0.00013702304382094246,
      "loss": 0.8428,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3800
    },
    {
      "epoch": 1.0305653232350553,
      "grad_norm": 4.085853099822998,
      "learning_rate": 0.00013673845820719885,
      "loss": 0.8633,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3810
    },
    {
      "epoch": 1.0332702190965648,
      "grad_norm": 3.8755381107330322,
      "learning_rate": 0.00013645376013114483,
      "loss": 0.8553,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3820
    },
    {
      "epoch": 1.035975114958074,
      "grad_norm": 4.089893817901611,
      "learning_rate": 0.00013616895434428344,
      "loss": 0.8089,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3830
    },
    {
      "epoch": 1.0386800108195835,
      "grad_norm": 5.560532093048096,
      "learning_rate": 0.00013588404559991508,
      "loss": 0.8326,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3840
    },
    {
      "epoch": 1.0413849066810927,
      "grad_norm": 4.632630348205566,
      "learning_rate": 0.0001355990386530586,
      "loss": 0.816,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3850
    },
    {
      "epoch": 1.0440898025426022,
      "grad_norm": 3.860368251800537,
      "learning_rate": 0.00013531393826037174,
      "loss": 0.845,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3860
    },
    {
      "epoch": 1.0467946984041114,
      "grad_norm": 4.269676208496094,
      "learning_rate": 0.00013502874918007194,
      "loss": 0.8482,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3870
    },
    {
      "epoch": 1.0494995942656207,
      "grad_norm": 4.151762008666992,
      "learning_rate": 0.00013474347617185669,
      "loss": 0.7829,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3880
    },
    {
      "epoch": 1.0522044901271301,
      "grad_norm": 4.129550933837891,
      "learning_rate": 0.00013445812399682418,
      "loss": 0.8372,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3890
    },
    {
      "epoch": 1.0549093859886394,
      "grad_norm": 4.330081939697266,
      "learning_rate": 0.00013417269741739402,
      "loss": 0.8488,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3900
    },
    {
      "epoch": 1.0576142818501488,
      "grad_norm": 3.89836049079895,
      "learning_rate": 0.00013388720119722746,
      "loss": 0.8811,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3910
    },
    {
      "epoch": 1.060319177711658,
      "grad_norm": 4.1753740310668945,
      "learning_rate": 0.00013360164010114804,
      "loss": 0.8653,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3920
    },
    {
      "epoch": 1.0630240735731675,
      "grad_norm": 3.7796201705932617,
      "learning_rate": 0.0001333160188950622,
      "loss": 0.8107,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3930
    },
    {
      "epoch": 1.0657289694346768,
      "grad_norm": 4.1096930503845215,
      "learning_rate": 0.0001330303423458793,
      "loss": 0.8079,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3940
    },
    {
      "epoch": 1.068433865296186,
      "grad_norm": 4.464737892150879,
      "learning_rate": 0.00013274461522143268,
      "loss": 0.7937,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3950
    },
    {
      "epoch": 1.0711387611576955,
      "grad_norm": 3.9034583568573,
      "learning_rate": 0.00013245884229039958,
      "loss": 0.8434,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3960
    },
    {
      "epoch": 1.0738436570192047,
      "grad_norm": 3.5338995456695557,
      "learning_rate": 0.0001321730283222217,
      "loss": 0.8705,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3970
    },
    {
      "epoch": 1.0765485528807142,
      "grad_norm": 5.958913326263428,
      "learning_rate": 0.00013188717808702583,
      "loss": 0.8676,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3980
    },
    {
      "epoch": 1.0792534487422234,
      "grad_norm": 5.412989139556885,
      "learning_rate": 0.00013160129635554383,
      "loss": 0.8631,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 3990
    },
    {
      "epoch": 1.0819583446037329,
      "grad_norm": 4.116124629974365,
      "learning_rate": 0.00013131538789903333,
      "loss": 0.8524,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4000
    },
    {
      "epoch": 1.084663240465242,
      "grad_norm": 3.657663106918335,
      "learning_rate": 0.00013102945748919798,
      "loss": 0.7878,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4010
    },
    {
      "epoch": 1.0873681363267513,
      "grad_norm": 4.144266605377197,
      "learning_rate": 0.0001307435098981078,
      "loss": 0.8192,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4020
    },
    {
      "epoch": 1.0900730321882608,
      "grad_norm": 3.4914867877960205,
      "learning_rate": 0.00013045754989811953,
      "loss": 0.757,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4030
    },
    {
      "epoch": 1.09277792804977,
      "grad_norm": 3.444493055343628,
      "learning_rate": 0.00013017158226179715,
      "loss": 0.7971,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4040
    },
    {
      "epoch": 1.0954828239112795,
      "grad_norm": 4.933722972869873,
      "learning_rate": 0.00012988561176183194,
      "loss": 0.8009,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4050
    },
    {
      "epoch": 1.0981877197727887,
      "grad_norm": 3.731696844100952,
      "learning_rate": 0.00012959964317096298,
      "loss": 0.8421,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4060
    },
    {
      "epoch": 1.1008926156342982,
      "grad_norm": 4.1172590255737305,
      "learning_rate": 0.00012931368126189758,
      "loss": 0.837,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4070
    },
    {
      "epoch": 1.1035975114958074,
      "grad_norm": 4.204513072967529,
      "learning_rate": 0.0001290277308072315,
      "loss": 0.8117,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4080
    },
    {
      "epoch": 1.1063024073573167,
      "grad_norm": 3.9016475677490234,
      "learning_rate": 0.00012874179657936923,
      "loss": 0.7951,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4090
    },
    {
      "epoch": 1.1090073032188261,
      "grad_norm": 7.60538387298584,
      "learning_rate": 0.0001284558833504446,
      "loss": 0.8207,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4100
    },
    {
      "epoch": 1.1117121990803354,
      "grad_norm": 3.7334892749786377,
      "learning_rate": 0.00012816999589224085,
      "loss": 0.7448,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4110
    },
    {
      "epoch": 1.1144170949418448,
      "grad_norm": 3.5912861824035645,
      "learning_rate": 0.00012788413897611123,
      "loss": 0.8491,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4120
    },
    {
      "epoch": 1.117121990803354,
      "grad_norm": 4.438766002655029,
      "learning_rate": 0.00012759831737289913,
      "loss": 0.8295,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4130
    },
    {
      "epoch": 1.1198268866648635,
      "grad_norm": 3.2297072410583496,
      "learning_rate": 0.00012731253585285867,
      "loss": 0.8082,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4140
    },
    {
      "epoch": 1.1225317825263728,
      "grad_norm": 4.35548734664917,
      "learning_rate": 0.00012702679918557495,
      "loss": 0.7688,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4150
    },
    {
      "epoch": 1.125236678387882,
      "grad_norm": 3.8470194339752197,
      "learning_rate": 0.00012674111213988452,
      "loss": 0.7657,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4160
    },
    {
      "epoch": 1.1279415742493915,
      "grad_norm": 4.626344203948975,
      "learning_rate": 0.00012645547948379577,
      "loss": 0.8122,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4170
    },
    {
      "epoch": 1.1306464701109007,
      "grad_norm": 5.0677809715271,
      "learning_rate": 0.00012616990598440935,
      "loss": 0.7846,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4180
    },
    {
      "epoch": 1.1333513659724102,
      "grad_norm": 3.8272931575775146,
      "learning_rate": 0.0001258843964078385,
      "loss": 0.8713,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4190
    },
    {
      "epoch": 1.1360562618339194,
      "grad_norm": 3.5707645416259766,
      "learning_rate": 0.00012559895551912984,
      "loss": 0.7931,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4200
    },
    {
      "epoch": 1.1387611576954288,
      "grad_norm": 4.14451265335083,
      "learning_rate": 0.00012531358808218337,
      "loss": 0.7637,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4210
    },
    {
      "epoch": 1.141466053556938,
      "grad_norm": 4.060018062591553,
      "learning_rate": 0.00012502829885967337,
      "loss": 0.7824,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4220
    },
    {
      "epoch": 1.1441709494184473,
      "grad_norm": 4.8129754066467285,
      "learning_rate": 0.0001247430926129687,
      "loss": 0.8092,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4230
    },
    {
      "epoch": 1.1468758452799568,
      "grad_norm": 4.509297847747803,
      "learning_rate": 0.00012445797410205338,
      "loss": 0.7734,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4240
    },
    {
      "epoch": 1.149580741141466,
      "grad_norm": 4.392861843109131,
      "learning_rate": 0.00012417294808544715,
      "loss": 0.809,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4250
    },
    {
      "epoch": 1.1522856370029755,
      "grad_norm": 4.265031814575195,
      "learning_rate": 0.00012388801932012604,
      "loss": 0.7832,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4260
    },
    {
      "epoch": 1.1549905328644847,
      "grad_norm": 4.353422164916992,
      "learning_rate": 0.00012360319256144307,
      "loss": 0.7955,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4270
    },
    {
      "epoch": 1.157695428725994,
      "grad_norm": 4.444879531860352,
      "learning_rate": 0.00012331847256304872,
      "loss": 0.7722,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4280
    },
    {
      "epoch": 1.1604003245875034,
      "grad_norm": 3.9174537658691406,
      "learning_rate": 0.00012303386407681174,
      "loss": 0.8425,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4290
    },
    {
      "epoch": 1.1631052204490127,
      "grad_norm": 4.4876861572265625,
      "learning_rate": 0.00012274937185273972,
      "loss": 0.7779,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4300
    },
    {
      "epoch": 1.1658101163105221,
      "grad_norm": 4.164872646331787,
      "learning_rate": 0.00012246500063889992,
      "loss": 0.8054,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4310
    },
    {
      "epoch": 1.1685150121720314,
      "grad_norm": 3.824904680252075,
      "learning_rate": 0.00012218075518134001,
      "loss": 0.8083,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4320
    },
    {
      "epoch": 1.1712199080335406,
      "grad_norm": 4.3014068603515625,
      "learning_rate": 0.00012189664022400882,
      "loss": 0.9044,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4330
    },
    {
      "epoch": 1.17392480389505,
      "grad_norm": 4.1722893714904785,
      "learning_rate": 0.00012161266050867712,
      "loss": 0.8897,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4340
    },
    {
      "epoch": 1.1766296997565593,
      "grad_norm": 2.8309264183044434,
      "learning_rate": 0.00012132882077485863,
      "loss": 0.7553,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4350
    },
    {
      "epoch": 1.1793345956180687,
      "grad_norm": 3.842503786087036,
      "learning_rate": 0.00012104512575973077,
      "loss": 0.8046,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4360
    },
    {
      "epoch": 1.182039491479578,
      "grad_norm": 3.749830961227417,
      "learning_rate": 0.00012076158019805573,
      "loss": 0.8079,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4370
    },
    {
      "epoch": 1.1847443873410874,
      "grad_norm": 3.7389776706695557,
      "learning_rate": 0.00012047818882210125,
      "loss": 0.7722,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4380
    },
    {
      "epoch": 1.1874492832025967,
      "grad_norm": 3.657909631729126,
      "learning_rate": 0.00012019495636156195,
      "loss": 0.7895,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4390
    },
    {
      "epoch": 1.190154179064106,
      "grad_norm": 3.7227067947387695,
      "learning_rate": 0.00011991188754348002,
      "loss": 0.812,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4400
    },
    {
      "epoch": 1.1928590749256154,
      "grad_norm": 3.645612955093384,
      "learning_rate": 0.00011962898709216666,
      "loss": 0.8027,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4410
    },
    {
      "epoch": 1.1955639707871246,
      "grad_norm": 11.574708938598633,
      "learning_rate": 0.00011934625972912301,
      "loss": 0.8095,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4420
    },
    {
      "epoch": 1.198268866648634,
      "grad_norm": 4.083512783050537,
      "learning_rate": 0.00011906371017296154,
      "loss": 0.8416,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4430
    },
    {
      "epoch": 1.2009737625101433,
      "grad_norm": 3.764439821243286,
      "learning_rate": 0.00011878134313932705,
      "loss": 0.8318,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4440
    },
    {
      "epoch": 1.2036786583716528,
      "grad_norm": 3.559797525405884,
      "learning_rate": 0.00011849916334081821,
      "loss": 0.8062,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4450
    },
    {
      "epoch": 1.206383554233162,
      "grad_norm": 4.464412212371826,
      "learning_rate": 0.00011821717548690872,
      "loss": 0.7876,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4460
    },
    {
      "epoch": 1.2090884500946713,
      "grad_norm": 4.090634822845459,
      "learning_rate": 0.00011793538428386894,
      "loss": 0.8011,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4470
    },
    {
      "epoch": 1.2117933459561807,
      "grad_norm": 4.831105709075928,
      "learning_rate": 0.00011765379443468703,
      "loss": 0.8164,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4480
    },
    {
      "epoch": 1.21449824181769,
      "grad_norm": 3.587251663208008,
      "learning_rate": 0.00011737241063899076,
      "loss": 0.7698,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4490
    },
    {
      "epoch": 1.2172031376791994,
      "grad_norm": 4.748007774353027,
      "learning_rate": 0.00011709123759296887,
      "loss": 0.7756,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4500
    },
    {
      "epoch": 1.2199080335407086,
      "grad_norm": 4.16516637802124,
      "learning_rate": 0.00011681027998929279,
      "loss": 0.7928,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4510
    },
    {
      "epoch": 1.222612929402218,
      "grad_norm": 4.2038187980651855,
      "learning_rate": 0.00011652954251703834,
      "loss": 0.7508,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4520
    },
    {
      "epoch": 1.2253178252637273,
      "grad_norm": 3.931687355041504,
      "learning_rate": 0.00011624902986160737,
      "loss": 0.7971,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4530
    },
    {
      "epoch": 1.2280227211252366,
      "grad_norm": 4.040994644165039,
      "learning_rate": 0.00011596874670464968,
      "loss": 0.7923,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4540
    },
    {
      "epoch": 1.230727616986746,
      "grad_norm": 3.236663341522217,
      "learning_rate": 0.00011568869772398474,
      "loss": 0.7999,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4550
    },
    {
      "epoch": 1.2334325128482553,
      "grad_norm": 4.613912105560303,
      "learning_rate": 0.00011540888759352379,
      "loss": 0.7877,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4560
    },
    {
      "epoch": 1.2361374087097647,
      "grad_norm": 3.4476230144500732,
      "learning_rate": 0.00011512932098319176,
      "loss": 0.7944,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4570
    },
    {
      "epoch": 1.238842304571274,
      "grad_norm": 4.322662353515625,
      "learning_rate": 0.00011485000255884921,
      "loss": 0.7343,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4580
    },
    {
      "epoch": 1.2415472004327834,
      "grad_norm": 3.7558603286743164,
      "learning_rate": 0.0001145709369822147,
      "loss": 0.8047,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4590
    },
    {
      "epoch": 1.2442520962942927,
      "grad_norm": 3.8777272701263428,
      "learning_rate": 0.00011429212891078677,
      "loss": 0.7745,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4600
    },
    {
      "epoch": 1.246956992155802,
      "grad_norm": 3.810472011566162,
      "learning_rate": 0.00011401358299776635,
      "loss": 0.7425,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4610
    },
    {
      "epoch": 1.2496618880173114,
      "grad_norm": 3.5155868530273438,
      "learning_rate": 0.000113735303891979,
      "loss": 0.7774,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4620
    },
    {
      "epoch": 1.2523667838788206,
      "grad_norm": 4.224593639373779,
      "learning_rate": 0.00011345729623779742,
      "loss": 0.7554,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4630
    },
    {
      "epoch": 1.25507167974033,
      "grad_norm": 3.860917091369629,
      "learning_rate": 0.00011317956467506381,
      "loss": 0.7789,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4640
    },
    {
      "epoch": 1.2577765756018393,
      "grad_norm": 3.948108196258545,
      "learning_rate": 0.00011290211383901261,
      "loss": 0.765,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4650
    },
    {
      "epoch": 1.2604814714633488,
      "grad_norm": 3.9504477977752686,
      "learning_rate": 0.00011262494836019295,
      "loss": 0.7663,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4660
    },
    {
      "epoch": 1.263186367324858,
      "grad_norm": 3.798619270324707,
      "learning_rate": 0.00011234807286439151,
      "loss": 0.7916,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4670
    },
    {
      "epoch": 1.2658912631863672,
      "grad_norm": 3.08143949508667,
      "learning_rate": 0.00011207149197255522,
      "loss": 0.7941,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4680
    },
    {
      "epoch": 1.2685961590478767,
      "grad_norm": 3.4709136486053467,
      "learning_rate": 0.00011179521030071423,
      "loss": 0.8227,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4690
    },
    {
      "epoch": 1.271301054909386,
      "grad_norm": 3.6086714267730713,
      "learning_rate": 0.00011151923245990481,
      "loss": 0.7296,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4700
    },
    {
      "epoch": 1.2740059507708954,
      "grad_norm": 3.868987798690796,
      "learning_rate": 0.00011124356305609232,
      "loss": 0.7837,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4710
    },
    {
      "epoch": 1.2767108466324046,
      "grad_norm": 3.548504114151001,
      "learning_rate": 0.00011096820669009455,
      "loss": 0.7689,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4720
    },
    {
      "epoch": 1.279415742493914,
      "grad_norm": 4.151698589324951,
      "learning_rate": 0.00011069316795750478,
      "loss": 0.7503,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4730
    },
    {
      "epoch": 1.2821206383554233,
      "grad_norm": 3.465074300765991,
      "learning_rate": 0.00011041845144861502,
      "loss": 0.7531,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4740
    },
    {
      "epoch": 1.2848255342169326,
      "grad_norm": 3.779144525527954,
      "learning_rate": 0.0001101440617483396,
      "loss": 0.7879,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4750
    },
    {
      "epoch": 1.287530430078442,
      "grad_norm": 4.2339863777160645,
      "learning_rate": 0.00010987000343613842,
      "loss": 0.7544,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4760
    },
    {
      "epoch": 1.2902353259399513,
      "grad_norm": 4.341187953948975,
      "learning_rate": 0.00010959628108594082,
      "loss": 0.7995,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4770
    },
    {
      "epoch": 1.2929402218014605,
      "grad_norm": 4.022242546081543,
      "learning_rate": 0.00010932289926606887,
      "loss": 0.7456,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4780
    },
    {
      "epoch": 1.29564511766297,
      "grad_norm": 3.7069079875946045,
      "learning_rate": 0.00010904986253916142,
      "loss": 0.7697,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4790
    },
    {
      "epoch": 1.2983500135244794,
      "grad_norm": 3.4390454292297363,
      "learning_rate": 0.00010877717546209782,
      "loss": 0.7546,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4800
    },
    {
      "epoch": 1.3010549093859887,
      "grad_norm": 4.179173946380615,
      "learning_rate": 0.00010850484258592191,
      "loss": 0.7375,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4810
    },
    {
      "epoch": 1.303759805247498,
      "grad_norm": 4.064438343048096,
      "learning_rate": 0.00010823286845576603,
      "loss": 0.7406,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4820
    },
    {
      "epoch": 1.3064647011090074,
      "grad_norm": 3.992492914199829,
      "learning_rate": 0.00010796125761077525,
      "loss": 0.7868,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4830
    },
    {
      "epoch": 1.3091695969705166,
      "grad_norm": 3.318981409072876,
      "learning_rate": 0.00010769001458403143,
      "loss": 0.7627,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4840
    },
    {
      "epoch": 1.3118744928320258,
      "grad_norm": 3.757812261581421,
      "learning_rate": 0.00010741914390247778,
      "loss": 0.7701,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4850
    },
    {
      "epoch": 1.3145793886935353,
      "grad_norm": 3.936673164367676,
      "learning_rate": 0.00010714865008684319,
      "loss": 0.7668,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4860
    },
    {
      "epoch": 1.3172842845550448,
      "grad_norm": 4.599918842315674,
      "learning_rate": 0.00010687853765156681,
      "loss": 0.7774,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4870
    },
    {
      "epoch": 1.319989180416554,
      "grad_norm": 3.4092042446136475,
      "learning_rate": 0.00010660881110472267,
      "loss": 0.7125,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4880
    },
    {
      "epoch": 1.3226940762780632,
      "grad_norm": 3.48287296295166,
      "learning_rate": 0.00010633947494794446,
      "loss": 0.7219,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4890
    },
    {
      "epoch": 1.3253989721395727,
      "grad_norm": 3.6531386375427246,
      "learning_rate": 0.00010607053367635045,
      "loss": 0.8138,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4900
    },
    {
      "epoch": 1.328103868001082,
      "grad_norm": 3.259798526763916,
      "learning_rate": 0.00010580199177846835,
      "loss": 0.7558,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4910
    },
    {
      "epoch": 1.3308087638625912,
      "grad_norm": 3.502213478088379,
      "learning_rate": 0.00010553385373616058,
      "loss": 0.7647,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4920
    },
    {
      "epoch": 1.3335136597241006,
      "grad_norm": 4.745849132537842,
      "learning_rate": 0.00010526612402454929,
      "loss": 0.7654,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4930
    },
    {
      "epoch": 1.33621855558561,
      "grad_norm": 2.9877102375030518,
      "learning_rate": 0.00010499880711194182,
      "loss": 0.8245,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4940
    },
    {
      "epoch": 1.3389234514471193,
      "grad_norm": 4.280433177947998,
      "learning_rate": 0.0001047319074597559,
      "loss": 0.7213,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4950
    },
    {
      "epoch": 1.3416283473086286,
      "grad_norm": 3.988755702972412,
      "learning_rate": 0.0001044654295224455,
      "loss": 0.8199,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4960
    },
    {
      "epoch": 1.344333243170138,
      "grad_norm": 3.135068416595459,
      "learning_rate": 0.00010419937774742623,
      "loss": 0.7676,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4970
    },
    {
      "epoch": 1.3470381390316473,
      "grad_norm": 4.093833923339844,
      "learning_rate": 0.00010393375657500133,
      "loss": 0.767,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4980
    },
    {
      "epoch": 1.3497430348931565,
      "grad_norm": 4.11683464050293,
      "learning_rate": 0.00010366857043828729,
      "loss": 0.7226,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 4990
    },
    {
      "epoch": 1.352447930754666,
      "grad_norm": 3.3312768936157227,
      "learning_rate": 0.00010340382376314012,
      "loss": 0.7996,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5000
    },
    {
      "epoch": 1.3551528266161752,
      "grad_norm": 4.228743076324463,
      "learning_rate": 0.00010313952096808135,
      "loss": 0.7633,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5010
    },
    {
      "epoch": 1.3578577224776847,
      "grad_norm": 3.3424339294433594,
      "learning_rate": 0.00010287566646422435,
      "loss": 0.7454,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5020
    },
    {
      "epoch": 1.360562618339194,
      "grad_norm": 3.903388500213623,
      "learning_rate": 0.00010261226465520067,
      "loss": 0.7486,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5030
    },
    {
      "epoch": 1.3632675142007034,
      "grad_norm": 3.0429282188415527,
      "learning_rate": 0.0001023493199370865,
      "loss": 0.7415,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5040
    },
    {
      "epoch": 1.3659724100622126,
      "grad_norm": 3.4914298057556152,
      "learning_rate": 0.0001020868366983294,
      "loss": 0.725,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5050
    },
    {
      "epoch": 1.3686773059237218,
      "grad_norm": 4.062960624694824,
      "learning_rate": 0.00010182481931967507,
      "loss": 0.7253,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5060
    },
    {
      "epoch": 1.3713822017852313,
      "grad_norm": 3.146116018295288,
      "learning_rate": 0.0001015632721740941,
      "loss": 0.7403,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5070
    },
    {
      "epoch": 1.3740870976467405,
      "grad_norm": 3.4970767498016357,
      "learning_rate": 0.00010130219962670908,
      "loss": 0.7202,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5080
    },
    {
      "epoch": 1.37679199350825,
      "grad_norm": 4.373836040496826,
      "learning_rate": 0.00010104160603472177,
      "loss": 0.7626,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5090
    },
    {
      "epoch": 1.3794968893697592,
      "grad_norm": 4.679487705230713,
      "learning_rate": 0.00010078149574734031,
      "loss": 0.7513,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5100
    },
    {
      "epoch": 1.3822017852312687,
      "grad_norm": 4.024169921875,
      "learning_rate": 0.00010052187310570669,
      "loss": 0.7712,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5110
    },
    {
      "epoch": 1.384906681092778,
      "grad_norm": 3.360781669616699,
      "learning_rate": 0.0001002627424428243,
      "loss": 0.763,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5120
    },
    {
      "epoch": 1.3876115769542872,
      "grad_norm": 3.365064859390259,
      "learning_rate": 0.00010000410808348552,
      "loss": 0.7431,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5130
    },
    {
      "epoch": 1.3903164728157966,
      "grad_norm": 4.349980354309082,
      "learning_rate": 9.974597434419971e-05,
      "loss": 0.7362,
      "memory/device_mem_reserved(gib)": 33.61,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5140
    },
    {
      "epoch": 1.3930213686773059,
      "grad_norm": 2.9434914588928223,
      "learning_rate": 9.9488345533121e-05,
      "loss": 0.7293,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5150
    },
    {
      "epoch": 1.3957262645388153,
      "grad_norm": 3.9609878063201904,
      "learning_rate": 9.923122594997645e-05,
      "loss": 0.7321,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5160
    },
    {
      "epoch": 1.3984311604003246,
      "grad_norm": 3.407517194747925,
      "learning_rate": 9.897461988599435e-05,
      "loss": 0.687,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5170
    },
    {
      "epoch": 1.401136056261834,
      "grad_norm": 3.0949556827545166,
      "learning_rate": 9.871853162383256e-05,
      "loss": 0.7438,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5180
    },
    {
      "epoch": 1.4038409521233433,
      "grad_norm": 3.5743606090545654,
      "learning_rate": 9.846296543750698e-05,
      "loss": 0.8028,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5190
    },
    {
      "epoch": 1.4065458479848525,
      "grad_norm": 4.036952972412109,
      "learning_rate": 9.820792559232021e-05,
      "loss": 0.749,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5200
    },
    {
      "epoch": 1.409250743846362,
      "grad_norm": 3.4962141513824463,
      "learning_rate": 9.795341634479051e-05,
      "loss": 0.7106,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5210
    },
    {
      "epoch": 1.4119556397078712,
      "grad_norm": 4.316569805145264,
      "learning_rate": 9.769944194258073e-05,
      "loss": 0.7658,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5220
    },
    {
      "epoch": 1.4146605355693807,
      "grad_norm": 3.947174310684204,
      "learning_rate": 9.744600662442716e-05,
      "loss": 0.725,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5230
    },
    {
      "epoch": 1.4173654314308899,
      "grad_norm": 3.8810696601867676,
      "learning_rate": 9.719311462006912e-05,
      "loss": 0.7443,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5240
    },
    {
      "epoch": 1.4200703272923993,
      "grad_norm": 3.816995859146118,
      "learning_rate": 9.694077015017817e-05,
      "loss": 0.7612,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5250
    },
    {
      "epoch": 1.4227752231539086,
      "grad_norm": 4.156252861022949,
      "learning_rate": 9.668897742628775e-05,
      "loss": 0.7436,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5260
    },
    {
      "epoch": 1.4254801190154178,
      "grad_norm": 3.965712785720825,
      "learning_rate": 9.643774065072293e-05,
      "loss": 0.7395,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5270
    },
    {
      "epoch": 1.4281850148769273,
      "grad_norm": 4.437236785888672,
      "learning_rate": 9.618706401653009e-05,
      "loss": 0.7532,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5280
    },
    {
      "epoch": 1.4308899107384365,
      "grad_norm": 3.1023387908935547,
      "learning_rate": 9.593695170740713e-05,
      "loss": 0.7367,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5290
    },
    {
      "epoch": 1.433594806599946,
      "grad_norm": 3.9593067169189453,
      "learning_rate": 9.56874078976336e-05,
      "loss": 0.7445,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5300
    },
    {
      "epoch": 1.4362997024614552,
      "grad_norm": 3.3441007137298584,
      "learning_rate": 9.543843675200097e-05,
      "loss": 0.7305,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5310
    },
    {
      "epoch": 1.4390045983229647,
      "grad_norm": 3.6775174140930176,
      "learning_rate": 9.519004242574319e-05,
      "loss": 0.7308,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5320
    },
    {
      "epoch": 1.441709494184474,
      "grad_norm": 3.3923141956329346,
      "learning_rate": 9.494222906446726e-05,
      "loss": 0.7656,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5330
    },
    {
      "epoch": 1.4444143900459832,
      "grad_norm": 3.3706727027893066,
      "learning_rate": 9.469500080408422e-05,
      "loss": 0.7358,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5340
    },
    {
      "epoch": 1.4471192859074926,
      "grad_norm": 4.048244476318359,
      "learning_rate": 9.444836177073984e-05,
      "loss": 0.7043,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5350
    },
    {
      "epoch": 1.4498241817690019,
      "grad_norm": 4.9177703857421875,
      "learning_rate": 9.420231608074601e-05,
      "loss": 0.7271,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5360
    },
    {
      "epoch": 1.452529077630511,
      "grad_norm": 3.962493419647217,
      "learning_rate": 9.395686784051192e-05,
      "loss": 0.7301,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5370
    },
    {
      "epoch": 1.4552339734920205,
      "grad_norm": 4.14715576171875,
      "learning_rate": 9.371202114647559e-05,
      "loss": 0.7518,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5380
    },
    {
      "epoch": 1.45793886935353,
      "grad_norm": 3.6769769191741943,
      "learning_rate": 9.34677800850354e-05,
      "loss": 0.7236,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5390
    },
    {
      "epoch": 1.4606437652150392,
      "grad_norm": 3.5924251079559326,
      "learning_rate": 9.322414873248195e-05,
      "loss": 0.7166,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5400
    },
    {
      "epoch": 1.4633486610765485,
      "grad_norm": 3.1487386226654053,
      "learning_rate": 9.29811311549301e-05,
      "loss": 0.7222,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5410
    },
    {
      "epoch": 1.466053556938058,
      "grad_norm": 3.417743682861328,
      "learning_rate": 9.2738731408251e-05,
      "loss": 0.7095,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5420
    },
    {
      "epoch": 1.4687584527995672,
      "grad_norm": 3.2829761505126953,
      "learning_rate": 9.249695353800442e-05,
      "loss": 0.734,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5430
    },
    {
      "epoch": 1.4714633486610764,
      "grad_norm": 3.6236085891723633,
      "learning_rate": 9.225580157937123e-05,
      "loss": 0.6949,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5440
    },
    {
      "epoch": 1.4741682445225859,
      "grad_norm": 2.9395477771759033,
      "learning_rate": 9.201527955708616e-05,
      "loss": 0.7281,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5450
    },
    {
      "epoch": 1.4768731403840953,
      "grad_norm": 3.4516067504882812,
      "learning_rate": 9.177539148537043e-05,
      "loss": 0.7309,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5460
    },
    {
      "epoch": 1.4795780362456046,
      "grad_norm": 3.7450966835021973,
      "learning_rate": 9.153614136786499e-05,
      "loss": 0.735,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5470
    },
    {
      "epoch": 1.4822829321071138,
      "grad_norm": 4.362335681915283,
      "learning_rate": 9.129753319756344e-05,
      "loss": 0.7133,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5480
    },
    {
      "epoch": 1.4849878279686233,
      "grad_norm": 3.6823859214782715,
      "learning_rate": 9.105957095674565e-05,
      "loss": 0.7392,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5490
    },
    {
      "epoch": 1.4876927238301325,
      "grad_norm": 3.737245798110962,
      "learning_rate": 9.08222586169111e-05,
      "loss": 0.7429,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5500
    },
    {
      "epoch": 1.4903976196916418,
      "grad_norm": 2.9608211517333984,
      "learning_rate": 9.058560013871262e-05,
      "loss": 0.7187,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5510
    },
    {
      "epoch": 1.4931025155531512,
      "grad_norm": 3.4424057006835938,
      "learning_rate": 9.034959947189047e-05,
      "loss": 0.7328,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5520
    },
    {
      "epoch": 1.4958074114146607,
      "grad_norm": 3.8036484718322754,
      "learning_rate": 9.011426055520623e-05,
      "loss": 0.718,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5530
    },
    {
      "epoch": 1.49851230727617,
      "grad_norm": 3.4891293048858643,
      "learning_rate": 8.987958731637709e-05,
      "loss": 0.7947,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5540
    },
    {
      "epoch": 1.5012172031376791,
      "grad_norm": 3.1536202430725098,
      "learning_rate": 8.96455836720103e-05,
      "loss": 0.7712,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5550
    },
    {
      "epoch": 1.5039220989991886,
      "grad_norm": 3.4930453300476074,
      "learning_rate": 8.941225352753794e-05,
      "loss": 0.755,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5560
    },
    {
      "epoch": 1.5066269948606978,
      "grad_norm": 4.301180839538574,
      "learning_rate": 8.91796007771516e-05,
      "loss": 0.7075,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5570
    },
    {
      "epoch": 1.509331890722207,
      "grad_norm": 3.503283739089966,
      "learning_rate": 8.894762930373735e-05,
      "loss": 0.7169,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5580
    },
    {
      "epoch": 1.5120367865837165,
      "grad_norm": 3.062073230743408,
      "learning_rate": 8.871634297881102e-05,
      "loss": 0.7131,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5590
    },
    {
      "epoch": 1.514741682445226,
      "grad_norm": 4.109286785125732,
      "learning_rate": 8.848574566245366e-05,
      "loss": 0.7074,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5600
    },
    {
      "epoch": 1.5174465783067352,
      "grad_norm": 3.173372507095337,
      "learning_rate": 8.825584120324693e-05,
      "loss": 0.7571,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5610
    },
    {
      "epoch": 1.5201514741682445,
      "grad_norm": 3.356884479522705,
      "learning_rate": 8.802663343820906e-05,
      "loss": 0.6887,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5620
    },
    {
      "epoch": 1.522856370029754,
      "grad_norm": 3.24686598777771,
      "learning_rate": 8.77981261927306e-05,
      "loss": 0.6868,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5630
    },
    {
      "epoch": 1.5255612658912632,
      "grad_norm": 4.0800461769104,
      "learning_rate": 8.757032328051083e-05,
      "loss": 0.753,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5640
    },
    {
      "epoch": 1.5282661617527724,
      "grad_norm": 3.529125928878784,
      "learning_rate": 8.734322850349386e-05,
      "loss": 0.7125,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5650
    },
    {
      "epoch": 1.5309710576142819,
      "grad_norm": 3.4760022163391113,
      "learning_rate": 8.71168456518053e-05,
      "loss": 0.7595,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5660
    },
    {
      "epoch": 1.5336759534757913,
      "grad_norm": 3.0900919437408447,
      "learning_rate": 8.689117850368913e-05,
      "loss": 0.6614,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5670
    },
    {
      "epoch": 1.5363808493373006,
      "grad_norm": 3.913034200668335,
      "learning_rate": 8.666623082544437e-05,
      "loss": 0.6745,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5680
    },
    {
      "epoch": 1.5390857451988098,
      "grad_norm": 3.2551372051239014,
      "learning_rate": 8.644200637136239e-05,
      "loss": 0.7229,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5690
    },
    {
      "epoch": 1.5417906410603193,
      "grad_norm": 3.5529627799987793,
      "learning_rate": 8.621850888366427e-05,
      "loss": 0.6859,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5700
    },
    {
      "epoch": 1.5444955369218285,
      "grad_norm": 2.924013614654541,
      "learning_rate": 8.599574209243819e-05,
      "loss": 0.7,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5710
    },
    {
      "epoch": 1.5472004327833377,
      "grad_norm": 3.1644272804260254,
      "learning_rate": 8.577370971557745e-05,
      "loss": 0.6983,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5720
    },
    {
      "epoch": 1.5499053286448472,
      "grad_norm": 4.470064640045166,
      "learning_rate": 8.555241545871812e-05,
      "loss": 0.6841,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5730
    },
    {
      "epoch": 1.5526102245063567,
      "grad_norm": 2.8502070903778076,
      "learning_rate": 8.533186301517739e-05,
      "loss": 0.7052,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5740
    },
    {
      "epoch": 1.5553151203678657,
      "grad_norm": 3.082613468170166,
      "learning_rate": 8.511205606589179e-05,
      "loss": 0.698,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5750
    },
    {
      "epoch": 1.5580200162293751,
      "grad_norm": 3.370327949523926,
      "learning_rate": 8.489299827935592e-05,
      "loss": 0.6434,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5760
    },
    {
      "epoch": 1.5607249120908846,
      "grad_norm": 3.3076109886169434,
      "learning_rate": 8.467469331156115e-05,
      "loss": 0.7073,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5770
    },
    {
      "epoch": 1.5634298079523938,
      "grad_norm": 3.948122024536133,
      "learning_rate": 8.445714480593448e-05,
      "loss": 0.7137,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5780
    },
    {
      "epoch": 1.566134703813903,
      "grad_norm": 2.9354052543640137,
      "learning_rate": 8.424035639327795e-05,
      "loss": 0.7037,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5790
    },
    {
      "epoch": 1.5688395996754125,
      "grad_norm": 3.1925742626190186,
      "learning_rate": 8.40243316917079e-05,
      "loss": 0.65,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5800
    },
    {
      "epoch": 1.571544495536922,
      "grad_norm": 3.0358006954193115,
      "learning_rate": 8.380907430659462e-05,
      "loss": 0.7453,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5810
    },
    {
      "epoch": 1.574249391398431,
      "grad_norm": 3.5184242725372314,
      "learning_rate": 8.359458783050209e-05,
      "loss": 0.7408,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5820
    },
    {
      "epoch": 1.5769542872599405,
      "grad_norm": 3.407390832901001,
      "learning_rate": 8.338087584312835e-05,
      "loss": 0.7238,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5830
    },
    {
      "epoch": 1.57965918312145,
      "grad_norm": 3.332489013671875,
      "learning_rate": 8.31679419112453e-05,
      "loss": 0.7374,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5840
    },
    {
      "epoch": 1.5823640789829592,
      "grad_norm": 3.500960111618042,
      "learning_rate": 8.295578958863945e-05,
      "loss": 0.6831,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5850
    },
    {
      "epoch": 1.5850689748444684,
      "grad_norm": 3.145552635192871,
      "learning_rate": 8.274442241605251e-05,
      "loss": 0.6849,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5860
    },
    {
      "epoch": 1.5877738707059779,
      "grad_norm": 3.061462640762329,
      "learning_rate": 8.253384392112246e-05,
      "loss": 0.7097,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5870
    },
    {
      "epoch": 1.590478766567487,
      "grad_norm": 2.6818971633911133,
      "learning_rate": 8.232405761832445e-05,
      "loss": 0.731,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5880
    },
    {
      "epoch": 1.5931836624289963,
      "grad_norm": 3.427539110183716,
      "learning_rate": 8.211506700891225e-05,
      "loss": 0.6835,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5890
    },
    {
      "epoch": 1.5958885582905058,
      "grad_norm": 3.392904281616211,
      "learning_rate": 8.190687558085978e-05,
      "loss": 0.7034,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5900
    },
    {
      "epoch": 1.5985934541520153,
      "grad_norm": 2.950566291809082,
      "learning_rate": 8.169948680880304e-05,
      "loss": 0.7186,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5910
    },
    {
      "epoch": 1.6012983500135245,
      "grad_norm": 2.5075855255126953,
      "learning_rate": 8.149290415398188e-05,
      "loss": 0.7002,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5920
    },
    {
      "epoch": 1.6040032458750337,
      "grad_norm": 3.078691244125366,
      "learning_rate": 8.128713106418244e-05,
      "loss": 0.7103,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5930
    },
    {
      "epoch": 1.6067081417365432,
      "grad_norm": 3.341064691543579,
      "learning_rate": 8.108217097367949e-05,
      "loss": 0.7197,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5940
    },
    {
      "epoch": 1.6094130375980524,
      "grad_norm": 3.4273488521575928,
      "learning_rate": 8.087802730317916e-05,
      "loss": 0.671,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5950
    },
    {
      "epoch": 1.6121179334595617,
      "grad_norm": 3.0635130405426025,
      "learning_rate": 8.067470345976187e-05,
      "loss": 0.715,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5960
    },
    {
      "epoch": 1.6148228293210711,
      "grad_norm": 3.202666997909546,
      "learning_rate": 8.047220283682534e-05,
      "loss": 0.6486,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5970
    },
    {
      "epoch": 1.6175277251825806,
      "grad_norm": 3.3150434494018555,
      "learning_rate": 8.027052881402816e-05,
      "loss": 0.693,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5980
    },
    {
      "epoch": 1.6202326210440898,
      "grad_norm": 3.6459527015686035,
      "learning_rate": 8.006968475723325e-05,
      "loss": 0.7597,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 5990
    },
    {
      "epoch": 1.622937516905599,
      "grad_norm": 3.103414297103882,
      "learning_rate": 7.98696740184517e-05,
      "loss": 0.7051,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6000
    },
    {
      "epoch": 1.6256424127671085,
      "grad_norm": 3.593626022338867,
      "learning_rate": 7.967049993578681e-05,
      "loss": 0.7071,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6010
    },
    {
      "epoch": 1.6283473086286178,
      "grad_norm": 3.4493701457977295,
      "learning_rate": 7.947216583337844e-05,
      "loss": 0.6802,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6020
    },
    {
      "epoch": 1.631052204490127,
      "grad_norm": 3.9632515907287598,
      "learning_rate": 7.927467502134753e-05,
      "loss": 0.6503,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6030
    },
    {
      "epoch": 1.6337571003516365,
      "grad_norm": 2.9177703857421875,
      "learning_rate": 7.907803079574076e-05,
      "loss": 0.6864,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6040
    },
    {
      "epoch": 1.636461996213146,
      "grad_norm": 3.0437581539154053,
      "learning_rate": 7.88822364384756e-05,
      "loss": 0.6833,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6050
    },
    {
      "epoch": 1.6391668920746552,
      "grad_norm": 2.7047359943389893,
      "learning_rate": 7.868729521728564e-05,
      "loss": 0.713,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6060
    },
    {
      "epoch": 1.6418717879361644,
      "grad_norm": 3.962380886077881,
      "learning_rate": 7.849321038566581e-05,
      "loss": 0.7073,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6070
    },
    {
      "epoch": 1.6445766837976739,
      "grad_norm": 3.2231719493865967,
      "learning_rate": 7.829998518281834e-05,
      "loss": 0.6968,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6080
    },
    {
      "epoch": 1.647281579659183,
      "grad_norm": 3.2594354152679443,
      "learning_rate": 7.810762283359854e-05,
      "loss": 0.7421,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6090
    },
    {
      "epoch": 1.6499864755206923,
      "grad_norm": 2.738050937652588,
      "learning_rate": 7.7916126548461e-05,
      "loss": 0.6442,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6100
    },
    {
      "epoch": 1.6526913713822018,
      "grad_norm": 3.0600812435150146,
      "learning_rate": 7.772549952340608e-05,
      "loss": 0.6459,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6110
    },
    {
      "epoch": 1.6553962672437113,
      "grad_norm": 3.2444002628326416,
      "learning_rate": 7.753574493992643e-05,
      "loss": 0.6379,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6120
    },
    {
      "epoch": 1.6581011631052205,
      "grad_norm": 2.9742543697357178,
      "learning_rate": 7.734686596495408e-05,
      "loss": 0.6753,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6130
    },
    {
      "epoch": 1.6608060589667297,
      "grad_norm": 3.655578136444092,
      "learning_rate": 7.715886575080747e-05,
      "loss": 0.6628,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6140
    },
    {
      "epoch": 1.6635109548282392,
      "grad_norm": 3.0293922424316406,
      "learning_rate": 7.697174743513877e-05,
      "loss": 0.7293,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6150
    },
    {
      "epoch": 1.6662158506897484,
      "grad_norm": 3.036137819290161,
      "learning_rate": 7.678551414088169e-05,
      "loss": 0.6722,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6160
    },
    {
      "epoch": 1.6689207465512577,
      "grad_norm": 3.269745111465454,
      "learning_rate": 7.660016897619922e-05,
      "loss": 0.6983,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6170
    },
    {
      "epoch": 1.6716256424127671,
      "grad_norm": 3.1036510467529297,
      "learning_rate": 7.641571503443186e-05,
      "loss": 0.6859,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6180
    },
    {
      "epoch": 1.6743305382742766,
      "grad_norm": 3.266829490661621,
      "learning_rate": 7.623215539404587e-05,
      "loss": 0.6839,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6190
    },
    {
      "epoch": 1.6770354341357858,
      "grad_norm": 3.1859540939331055,
      "learning_rate": 7.604949311858198e-05,
      "loss": 0.6848,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6200
    },
    {
      "epoch": 1.679740329997295,
      "grad_norm": 3.4950735569000244,
      "learning_rate": 7.586773125660428e-05,
      "loss": 0.6437,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6210
    },
    {
      "epoch": 1.6824452258588045,
      "grad_norm": 2.416506290435791,
      "learning_rate": 7.568687284164924e-05,
      "loss": 0.6722,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6220
    },
    {
      "epoch": 1.6851501217203138,
      "grad_norm": 3.472022294998169,
      "learning_rate": 7.55069208921752e-05,
      "loss": 0.6887,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6230
    },
    {
      "epoch": 1.687855017581823,
      "grad_norm": 3.4345269203186035,
      "learning_rate": 7.532787841151188e-05,
      "loss": 0.7043,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6240
    },
    {
      "epoch": 1.6905599134433325,
      "grad_norm": 3.3186779022216797,
      "learning_rate": 7.514974838781035e-05,
      "loss": 0.667,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6250
    },
    {
      "epoch": 1.693264809304842,
      "grad_norm": 3.0585949420928955,
      "learning_rate": 7.497253379399311e-05,
      "loss": 0.6504,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6260
    },
    {
      "epoch": 1.6959697051663511,
      "grad_norm": 3.4880311489105225,
      "learning_rate": 7.479623758770445e-05,
      "loss": 0.6877,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6270
    },
    {
      "epoch": 1.6986746010278604,
      "grad_norm": 2.947932004928589,
      "learning_rate": 7.46208627112612e-05,
      "loss": 0.6516,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6280
    },
    {
      "epoch": 1.7013794968893698,
      "grad_norm": 2.9375832080841064,
      "learning_rate": 7.444641209160347e-05,
      "loss": 0.6771,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6290
    },
    {
      "epoch": 1.704084392750879,
      "grad_norm": 3.301255226135254,
      "learning_rate": 7.427288864024591e-05,
      "loss": 0.7194,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6300
    },
    {
      "epoch": 1.7067892886123883,
      "grad_norm": 3.0097708702087402,
      "learning_rate": 7.410029525322905e-05,
      "loss": 0.6771,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6310
    },
    {
      "epoch": 1.7094941844738978,
      "grad_norm": 2.9872400760650635,
      "learning_rate": 7.392863481107099e-05,
      "loss": 0.6718,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6320
    },
    {
      "epoch": 1.7121990803354072,
      "grad_norm": 3.524113893508911,
      "learning_rate": 7.375791017871947e-05,
      "loss": 0.668,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6330
    },
    {
      "epoch": 1.7149039761969163,
      "grad_norm": 2.582218885421753,
      "learning_rate": 7.358812420550375e-05,
      "loss": 0.6685,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6340
    },
    {
      "epoch": 1.7176088720584257,
      "grad_norm": 2.9638619422912598,
      "learning_rate": 7.341927972508731e-05,
      "loss": 0.6884,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6350
    },
    {
      "epoch": 1.7203137679199352,
      "grad_norm": 3.10848331451416,
      "learning_rate": 7.32513795554204e-05,
      "loss": 0.6652,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6360
    },
    {
      "epoch": 1.7230186637814444,
      "grad_norm": 3.0611279010772705,
      "learning_rate": 7.308442649869318e-05,
      "loss": 0.6659,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6370
    },
    {
      "epoch": 1.7257235596429537,
      "grad_norm": 3.327214241027832,
      "learning_rate": 7.291842334128883e-05,
      "loss": 0.6587,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6380
    },
    {
      "epoch": 1.7284284555044631,
      "grad_norm": 2.9344241619110107,
      "learning_rate": 7.275337285373702e-05,
      "loss": 0.6395,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6390
    },
    {
      "epoch": 1.7311333513659726,
      "grad_norm": 3.1315062046051025,
      "learning_rate": 7.258927779066775e-05,
      "loss": 0.6631,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6400
    },
    {
      "epoch": 1.7338382472274816,
      "grad_norm": 3.0949370861053467,
      "learning_rate": 7.242614089076538e-05,
      "loss": 0.6613,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6410
    },
    {
      "epoch": 1.736543143088991,
      "grad_norm": 3.2582595348358154,
      "learning_rate": 7.226396487672281e-05,
      "loss": 0.67,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6420
    },
    {
      "epoch": 1.7392480389505005,
      "grad_norm": 3.3499040603637695,
      "learning_rate": 7.210275245519622e-05,
      "loss": 0.6542,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6430
    },
    {
      "epoch": 1.7419529348120097,
      "grad_norm": 3.04258131980896,
      "learning_rate": 7.194250631675971e-05,
      "loss": 0.6743,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6440
    },
    {
      "epoch": 1.744657830673519,
      "grad_norm": 2.8085639476776123,
      "learning_rate": 7.178322913586057e-05,
      "loss": 0.6496,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6450
    },
    {
      "epoch": 1.7473627265350284,
      "grad_norm": 4.0755767822265625,
      "learning_rate": 7.162492357077445e-05,
      "loss": 0.6524,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6460
    },
    {
      "epoch": 1.7500676223965377,
      "grad_norm": 2.913227081298828,
      "learning_rate": 7.146759226356115e-05,
      "loss": 0.7211,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6470
    },
    {
      "epoch": 1.752772518258047,
      "grad_norm": 3.0723695755004883,
      "learning_rate": 7.131123784002053e-05,
      "loss": 0.7291,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6480
    },
    {
      "epoch": 1.7554774141195564,
      "grad_norm": 3.0220112800598145,
      "learning_rate": 7.115586290964857e-05,
      "loss": 0.6978,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6490
    },
    {
      "epoch": 1.7581823099810658,
      "grad_norm": 3.322822332382202,
      "learning_rate": 7.100147006559388e-05,
      "loss": 0.6769,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6500
    },
    {
      "epoch": 1.760887205842575,
      "grad_norm": 2.499387741088867,
      "learning_rate": 7.084806188461444e-05,
      "loss": 0.6696,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6510
    },
    {
      "epoch": 1.7635921017040843,
      "grad_norm": 3.1069934368133545,
      "learning_rate": 7.069564092703456e-05,
      "loss": 0.6496,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6520
    },
    {
      "epoch": 1.7662969975655938,
      "grad_norm": 3.254661798477173,
      "learning_rate": 7.054420973670223e-05,
      "loss": 0.5978,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6530
    },
    {
      "epoch": 1.769001893427103,
      "grad_norm": 3.3960485458374023,
      "learning_rate": 7.039377084094651e-05,
      "loss": 0.6136,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6540
    },
    {
      "epoch": 1.7717067892886122,
      "grad_norm": 3.1243672370910645,
      "learning_rate": 7.024432675053547e-05,
      "loss": 0.6787,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6550
    },
    {
      "epoch": 1.7744116851501217,
      "grad_norm": 3.014849901199341,
      "learning_rate": 7.00958799596343e-05,
      "loss": 0.637,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6560
    },
    {
      "epoch": 1.7771165810116312,
      "grad_norm": 2.9643986225128174,
      "learning_rate": 6.994843294576358e-05,
      "loss": 0.659,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6570
    },
    {
      "epoch": 1.7798214768731404,
      "grad_norm": 3.0430381298065186,
      "learning_rate": 6.980198816975802e-05,
      "loss": 0.6868,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6580
    },
    {
      "epoch": 1.7825263727346496,
      "grad_norm": 2.7200093269348145,
      "learning_rate": 6.965654807572537e-05,
      "loss": 0.6266,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6590
    },
    {
      "epoch": 1.785231268596159,
      "grad_norm": 3.626512050628662,
      "learning_rate": 6.951211509100561e-05,
      "loss": 0.6803,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6600
    },
    {
      "epoch": 1.7879361644576683,
      "grad_norm": 3.02445387840271,
      "learning_rate": 6.936869162613043e-05,
      "loss": 0.6272,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6610
    },
    {
      "epoch": 1.7906410603191776,
      "grad_norm": 2.7647345066070557,
      "learning_rate": 6.922628007478302e-05,
      "loss": 0.6608,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6620
    },
    {
      "epoch": 1.793345956180687,
      "grad_norm": 4.758910179138184,
      "learning_rate": 6.908488281375815e-05,
      "loss": 0.7089,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6630
    },
    {
      "epoch": 1.7960508520421965,
      "grad_norm": 3.3609447479248047,
      "learning_rate": 6.894450220292243e-05,
      "loss": 0.6386,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6640
    },
    {
      "epoch": 1.7987557479037057,
      "grad_norm": 2.3680367469787598,
      "learning_rate": 6.880514058517501e-05,
      "loss": 0.6156,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6650
    },
    {
      "epoch": 1.801460643765215,
      "grad_norm": 2.79038143157959,
      "learning_rate": 6.866680028640839e-05,
      "loss": 0.6519,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6660
    },
    {
      "epoch": 1.8041655396267244,
      "grad_norm": 2.9118330478668213,
      "learning_rate": 6.852948361546963e-05,
      "loss": 0.6654,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6670
    },
    {
      "epoch": 1.8068704354882337,
      "grad_norm": 3.048211097717285,
      "learning_rate": 6.839319286412191e-05,
      "loss": 0.6495,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6680
    },
    {
      "epoch": 1.809575331349743,
      "grad_norm": 2.6171252727508545,
      "learning_rate": 6.825793030700611e-05,
      "loss": 0.6308,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6690
    },
    {
      "epoch": 1.8122802272112524,
      "grad_norm": 2.7780067920684814,
      "learning_rate": 6.812369820160298e-05,
      "loss": 0.6652,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6700
    },
    {
      "epoch": 1.8149851230727618,
      "grad_norm": 3.2068605422973633,
      "learning_rate": 6.799049878819544e-05,
      "loss": 0.6829,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6710
    },
    {
      "epoch": 1.817690018934271,
      "grad_norm": 2.928981065750122,
      "learning_rate": 6.785833428983113e-05,
      "loss": 0.6692,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6720
    },
    {
      "epoch": 1.8203949147957803,
      "grad_norm": 3.014019250869751,
      "learning_rate": 6.772720691228538e-05,
      "loss": 0.6354,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6730
    },
    {
      "epoch": 1.8230998106572898,
      "grad_norm": 3.248959541320801,
      "learning_rate": 6.759711884402438e-05,
      "loss": 0.6317,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6740
    },
    {
      "epoch": 1.825804706518799,
      "grad_norm": 3.127004623413086,
      "learning_rate": 6.746807225616858e-05,
      "loss": 0.6288,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6750
    },
    {
      "epoch": 1.8285096023803082,
      "grad_norm": 2.981550931930542,
      "learning_rate": 6.734006930245661e-05,
      "loss": 0.6107,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6760
    },
    {
      "epoch": 1.8312144982418177,
      "grad_norm": 2.9153382778167725,
      "learning_rate": 6.721311211920913e-05,
      "loss": 0.656,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6770
    },
    {
      "epoch": 1.8339193941033272,
      "grad_norm": 2.825021505355835,
      "learning_rate": 6.70872028252934e-05,
      "loss": 0.6578,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6780
    },
    {
      "epoch": 1.8366242899648364,
      "grad_norm": 2.9255638122558594,
      "learning_rate": 6.696234352208776e-05,
      "loss": 0.6726,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6790
    },
    {
      "epoch": 1.8393291858263456,
      "grad_norm": 3.4403741359710693,
      "learning_rate": 6.683853629344655e-05,
      "loss": 0.5961,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6800
    },
    {
      "epoch": 1.842034081687855,
      "grad_norm": 2.990248441696167,
      "learning_rate": 6.671578320566546e-05,
      "loss": 0.6545,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6810
    },
    {
      "epoch": 1.8447389775493643,
      "grad_norm": 2.9776017665863037,
      "learning_rate": 6.659408630744694e-05,
      "loss": 0.6193,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6820
    },
    {
      "epoch": 1.8474438734108736,
      "grad_norm": 3.074422836303711,
      "learning_rate": 6.647344762986606e-05,
      "loss": 0.6856,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6830
    },
    {
      "epoch": 1.850148769272383,
      "grad_norm": 2.562652826309204,
      "learning_rate": 6.635386918633653e-05,
      "loss": 0.6359,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6840
    },
    {
      "epoch": 1.8528536651338925,
      "grad_norm": 4.123425006866455,
      "learning_rate": 6.623535297257721e-05,
      "loss": 0.6607,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6850
    },
    {
      "epoch": 1.8555585609954017,
      "grad_norm": 3.134983539581299,
      "learning_rate": 6.611790096657875e-05,
      "loss": 0.6726,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6860
    },
    {
      "epoch": 1.858263456856911,
      "grad_norm": 3.012802839279175,
      "learning_rate": 6.600151512857052e-05,
      "loss": 0.6524,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6870
    },
    {
      "epoch": 1.8609683527184204,
      "grad_norm": 3.057731866836548,
      "learning_rate": 6.588619740098801e-05,
      "loss": 0.7094,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6880
    },
    {
      "epoch": 1.8636732485799297,
      "grad_norm": 3.1754140853881836,
      "learning_rate": 6.577194970844033e-05,
      "loss": 0.6107,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6890
    },
    {
      "epoch": 1.866378144441439,
      "grad_norm": 2.532036781311035,
      "learning_rate": 6.565877395767813e-05,
      "loss": 0.617,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6900
    },
    {
      "epoch": 1.8690830403029484,
      "grad_norm": 3.0537824630737305,
      "learning_rate": 6.554667203756171e-05,
      "loss": 0.7119,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6910
    },
    {
      "epoch": 1.8717879361644578,
      "grad_norm": 3.008969783782959,
      "learning_rate": 6.543564581902959e-05,
      "loss": 0.6259,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6920
    },
    {
      "epoch": 1.8744928320259668,
      "grad_norm": 2.6460862159729004,
      "learning_rate": 6.532569715506723e-05,
      "loss": 0.5995,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6930
    },
    {
      "epoch": 1.8771977278874763,
      "grad_norm": 2.848828077316284,
      "learning_rate": 6.521682788067609e-05,
      "loss": 0.6923,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6940
    },
    {
      "epoch": 1.8799026237489858,
      "grad_norm": 2.922083616256714,
      "learning_rate": 6.510903981284306e-05,
      "loss": 0.5953,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6950
    },
    {
      "epoch": 1.882607519610495,
      "grad_norm": 2.930002450942993,
      "learning_rate": 6.500233475051009e-05,
      "loss": 0.6329,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6960
    },
    {
      "epoch": 1.8853124154720042,
      "grad_norm": 3.4451231956481934,
      "learning_rate": 6.489671447454415e-05,
      "loss": 0.6095,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6970
    },
    {
      "epoch": 1.8880173113335137,
      "grad_norm": 2.9342470169067383,
      "learning_rate": 6.479218074770761e-05,
      "loss": 0.6555,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6980
    },
    {
      "epoch": 1.8907222071950232,
      "grad_norm": 2.879666328430176,
      "learning_rate": 6.468873531462867e-05,
      "loss": 0.615,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 6990
    },
    {
      "epoch": 1.8934271030565322,
      "grad_norm": 2.5449421405792236,
      "learning_rate": 6.458637990177236e-05,
      "loss": 0.616,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7000
    },
    {
      "epoch": 1.8961319989180416,
      "grad_norm": 2.606356143951416,
      "learning_rate": 6.448511621741166e-05,
      "loss": 0.6276,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7010
    },
    {
      "epoch": 1.898836894779551,
      "grad_norm": 2.701772451400757,
      "learning_rate": 6.438494595159905e-05,
      "loss": 0.608,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7020
    },
    {
      "epoch": 1.9015417906410603,
      "grad_norm": 3.159738302230835,
      "learning_rate": 6.428587077613826e-05,
      "loss": 0.6752,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7030
    },
    {
      "epoch": 1.9042466865025696,
      "grad_norm": 2.644399642944336,
      "learning_rate": 6.418789234455636e-05,
      "loss": 0.6607,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7040
    },
    {
      "epoch": 1.906951582364079,
      "grad_norm": 2.847825765609741,
      "learning_rate": 6.409101229207619e-05,
      "loss": 0.6191,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7050
    },
    {
      "epoch": 1.9096564782255883,
      "grad_norm": 2.8219966888427734,
      "learning_rate": 6.399523223558908e-05,
      "loss": 0.6322,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7060
    },
    {
      "epoch": 1.9123613740870975,
      "grad_norm": 2.882685422897339,
      "learning_rate": 6.390055377362778e-05,
      "loss": 0.65,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7070
    },
    {
      "epoch": 1.915066269948607,
      "grad_norm": 3.004551410675049,
      "learning_rate": 6.380697848633993e-05,
      "loss": 0.6572,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7080
    },
    {
      "epoch": 1.9177711658101164,
      "grad_norm": 3.075967788696289,
      "learning_rate": 6.371450793546155e-05,
      "loss": 0.6092,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7090
    },
    {
      "epoch": 1.9204760616716257,
      "grad_norm": 2.8562681674957275,
      "learning_rate": 6.362314366429106e-05,
      "loss": 0.6654,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7100
    },
    {
      "epoch": 1.923180957533135,
      "grad_norm": 2.728839874267578,
      "learning_rate": 6.353288719766345e-05,
      "loss": 0.6261,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7110
    },
    {
      "epoch": 1.9258858533946444,
      "grad_norm": 2.3934872150421143,
      "learning_rate": 6.344374004192491e-05,
      "loss": 0.6314,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7120
    },
    {
      "epoch": 1.9285907492561536,
      "grad_norm": 3.402822732925415,
      "learning_rate": 6.335570368490764e-05,
      "loss": 0.6217,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7130
    },
    {
      "epoch": 1.9312956451176628,
      "grad_norm": 3.1024372577667236,
      "learning_rate": 6.326877959590502e-05,
      "loss": 0.6413,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7140
    },
    {
      "epoch": 1.9340005409791723,
      "grad_norm": 3.0662124156951904,
      "learning_rate": 6.318296922564712e-05,
      "loss": 0.6359,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7150
    },
    {
      "epoch": 1.9367054368406817,
      "grad_norm": 3.1964662075042725,
      "learning_rate": 6.309827400627642e-05,
      "loss": 0.6282,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7160
    },
    {
      "epoch": 1.939410332702191,
      "grad_norm": 3.04164719581604,
      "learning_rate": 6.301469535132405e-05,
      "loss": 0.5915,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7170
    },
    {
      "epoch": 1.9421152285637002,
      "grad_norm": 2.8097140789031982,
      "learning_rate": 6.293223465568601e-05,
      "loss": 0.6611,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7180
    },
    {
      "epoch": 1.9448201244252097,
      "grad_norm": 2.4967353343963623,
      "learning_rate": 6.285089329560002e-05,
      "loss": 0.5996,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7190
    },
    {
      "epoch": 1.947525020286719,
      "grad_norm": 2.8207314014434814,
      "learning_rate": 6.27706726286225e-05,
      "loss": 0.6595,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7200
    },
    {
      "epoch": 1.9502299161482282,
      "grad_norm": 2.516108512878418,
      "learning_rate": 6.269157399360596e-05,
      "loss": 0.6009,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7210
    },
    {
      "epoch": 1.9529348120097376,
      "grad_norm": 2.3894989490509033,
      "learning_rate": 6.261359871067658e-05,
      "loss": 0.6514,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7220
    },
    {
      "epoch": 1.955639707871247,
      "grad_norm": 3.204094409942627,
      "learning_rate": 6.253674808121228e-05,
      "loss": 0.6691,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7230
    },
    {
      "epoch": 1.9583446037327563,
      "grad_norm": 2.8803718090057373,
      "learning_rate": 6.246102338782084e-05,
      "loss": 0.62,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7240
    },
    {
      "epoch": 1.9610494995942656,
      "grad_norm": 2.9162309169769287,
      "learning_rate": 6.238642589431872e-05,
      "loss": 0.6001,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7250
    },
    {
      "epoch": 1.963754395455775,
      "grad_norm": 2.6662826538085938,
      "learning_rate": 6.231295684570974e-05,
      "loss": 0.579,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7260
    },
    {
      "epoch": 1.9664592913172843,
      "grad_norm": 3.128024101257324,
      "learning_rate": 6.224061746816444e-05,
      "loss": 0.6087,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7270
    },
    {
      "epoch": 1.9691641871787935,
      "grad_norm": 2.685338258743286,
      "learning_rate": 6.21694089689996e-05,
      "loss": 0.5965,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7280
    },
    {
      "epoch": 1.971869083040303,
      "grad_norm": 3.0504117012023926,
      "learning_rate": 6.209933253665804e-05,
      "loss": 0.6255,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7290
    },
    {
      "epoch": 1.9745739789018124,
      "grad_norm": 2.6502695083618164,
      "learning_rate": 6.203038934068881e-05,
      "loss": 0.5826,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7300
    },
    {
      "epoch": 1.9772788747633216,
      "grad_norm": 2.6905455589294434,
      "learning_rate": 6.19625805317277e-05,
      "loss": 0.7135,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7310
    },
    {
      "epoch": 1.9799837706248309,
      "grad_norm": 2.8525450229644775,
      "learning_rate": 6.189590724147798e-05,
      "loss": 0.6711,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7320
    },
    {
      "epoch": 1.9826886664863403,
      "grad_norm": 3.001568078994751,
      "learning_rate": 6.18303705826916e-05,
      "loss": 0.622,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7330
    },
    {
      "epoch": 1.9853935623478496,
      "grad_norm": 3.155000925064087,
      "learning_rate": 6.17659716491505e-05,
      "loss": 0.6766,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7340
    },
    {
      "epoch": 1.9880984582093588,
      "grad_norm": 3.0717968940734863,
      "learning_rate": 6.170271151564844e-05,
      "loss": 0.6082,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7350
    },
    {
      "epoch": 1.9908033540708683,
      "grad_norm": 3.055043935775757,
      "learning_rate": 6.164059123797304e-05,
      "loss": 0.6425,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7360
    },
    {
      "epoch": 1.9935082499323777,
      "grad_norm": 2.858274221420288,
      "learning_rate": 6.157961185288817e-05,
      "loss": 0.622,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7370
    },
    {
      "epoch": 1.996213145793887,
      "grad_norm": 3.485201358795166,
      "learning_rate": 6.151977437811662e-05,
      "loss": 0.6347,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7380
    },
    {
      "epoch": 1.9989180416553962,
      "grad_norm": 3.1326327323913574,
      "learning_rate": 6.146107981232313e-05,
      "loss": 0.5973,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7390
    },
    {
      "epoch": 2.0016229375169057,
      "grad_norm": 2.283562183380127,
      "learning_rate": 6.140352913509777e-05,
      "loss": 0.6019,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7400
    },
    {
      "epoch": 2.004327833378415,
      "grad_norm": 2.8729043006896973,
      "learning_rate": 6.134712330693944e-05,
      "loss": 0.5534,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7410
    },
    {
      "epoch": 2.007032729239924,
      "grad_norm": 2.65059757232666,
      "learning_rate": 6.129186326924005e-05,
      "loss": 0.556,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7420
    },
    {
      "epoch": 2.0097376251014336,
      "grad_norm": 2.711198091506958,
      "learning_rate": 6.123774994426864e-05,
      "loss": 0.5734,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7430
    },
    {
      "epoch": 2.012442520962943,
      "grad_norm": 2.73869252204895,
      "learning_rate": 6.118478423515608e-05,
      "loss": 0.5915,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7440
    },
    {
      "epoch": 2.015147416824452,
      "grad_norm": 2.6567530632019043,
      "learning_rate": 6.113296702587994e-05,
      "loss": 0.6016,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7450
    },
    {
      "epoch": 2.0178523126859615,
      "grad_norm": 3.0002238750457764,
      "learning_rate": 6.10822991812498e-05,
      "loss": 0.6058,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7460
    },
    {
      "epoch": 2.020557208547471,
      "grad_norm": 2.5208771228790283,
      "learning_rate": 6.103278154689273e-05,
      "loss": 0.5692,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7470
    },
    {
      "epoch": 2.0232621044089805,
      "grad_norm": 3.0592586994171143,
      "learning_rate": 6.098441494923928e-05,
      "loss": 0.5433,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7480
    },
    {
      "epoch": 2.0259670002704895,
      "grad_norm": 2.595139741897583,
      "learning_rate": 6.0937200195509614e-05,
      "loss": 0.53,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7490
    },
    {
      "epoch": 2.028671896131999,
      "grad_norm": 2.523493528366089,
      "learning_rate": 6.0891138073700026e-05,
      "loss": 0.5418,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7500
    },
    {
      "epoch": 2.0313767919935084,
      "grad_norm": 2.6988117694854736,
      "learning_rate": 6.084622935256988e-05,
      "loss": 0.583,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7510
    },
    {
      "epoch": 2.0340816878550174,
      "grad_norm": 2.9173007011413574,
      "learning_rate": 6.080247478162867e-05,
      "loss": 0.5798,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7520
    },
    {
      "epoch": 2.036786583716527,
      "grad_norm": 2.569444417953491,
      "learning_rate": 6.075987509112359e-05,
      "loss": 0.5888,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7530
    },
    {
      "epoch": 2.0394914795780363,
      "grad_norm": 2.5385310649871826,
      "learning_rate": 6.0718430992027295e-05,
      "loss": 0.5667,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7540
    },
    {
      "epoch": 2.042196375439546,
      "grad_norm": 2.933454990386963,
      "learning_rate": 6.067814317602609e-05,
      "loss": 0.6213,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7550
    },
    {
      "epoch": 2.044901271301055,
      "grad_norm": 2.752481698989868,
      "learning_rate": 6.063901231550833e-05,
      "loss": 0.634,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7560
    },
    {
      "epoch": 2.0476061671625643,
      "grad_norm": 3.026728868484497,
      "learning_rate": 6.060103906355321e-05,
      "loss": 0.5486,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7570
    },
    {
      "epoch": 2.0503110630240737,
      "grad_norm": 2.7333011627197266,
      "learning_rate": 6.0564224053919906e-05,
      "loss": 0.5959,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7580
    },
    {
      "epoch": 2.0530159588855827,
      "grad_norm": 2.766077995300293,
      "learning_rate": 6.052856790103697e-05,
      "loss": 0.6351,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7590
    },
    {
      "epoch": 2.055720854747092,
      "grad_norm": 3.1666741371154785,
      "learning_rate": 6.049407119999204e-05,
      "loss": 0.5552,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7600
    },
    {
      "epoch": 2.0584257506086017,
      "grad_norm": 2.675058364868164,
      "learning_rate": 6.0460734526522e-05,
      "loss": 0.6068,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7610
    },
    {
      "epoch": 2.0611306464701107,
      "grad_norm": 3.008115530014038,
      "learning_rate": 6.0428558437003235e-05,
      "loss": 0.579,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7620
    },
    {
      "epoch": 2.06383554233162,
      "grad_norm": 2.6898791790008545,
      "learning_rate": 6.0397543468442495e-05,
      "loss": 0.5352,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7630
    },
    {
      "epoch": 2.0665404381931296,
      "grad_norm": 2.971761465072632,
      "learning_rate": 6.036769013846784e-05,
      "loss": 0.5966,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7640
    },
    {
      "epoch": 2.069245334054639,
      "grad_norm": 2.54276967048645,
      "learning_rate": 6.033899894532e-05,
      "loss": 0.5858,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7650
    },
    {
      "epoch": 2.071950229916148,
      "grad_norm": 2.581988573074341,
      "learning_rate": 6.03114703678441e-05,
      "loss": 0.5876,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7660
    },
    {
      "epoch": 2.0746551257776575,
      "grad_norm": 2.739187717437744,
      "learning_rate": 6.028510486548168e-05,
      "loss": 0.5498,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7670
    },
    {
      "epoch": 2.077360021639167,
      "grad_norm": 2.447493314743042,
      "learning_rate": 6.0259902878262945e-05,
      "loss": 0.5677,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7680
    },
    {
      "epoch": 2.080064917500676,
      "grad_norm": 2.7747323513031006,
      "learning_rate": 6.0235864826799504e-05,
      "loss": 0.5657,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7690
    },
    {
      "epoch": 2.0827698133621855,
      "grad_norm": 3.108409881591797,
      "learning_rate": 6.0212991112277303e-05,
      "loss": 0.5659,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7700
    },
    {
      "epoch": 2.085474709223695,
      "grad_norm": 2.5329716205596924,
      "learning_rate": 6.019128211644998e-05,
      "loss": 0.5548,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7710
    },
    {
      "epoch": 2.0881796050852044,
      "grad_norm": 2.94100022315979,
      "learning_rate": 6.0170738201632406e-05,
      "loss": 0.5723,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7720
    },
    {
      "epoch": 2.0908845009467134,
      "grad_norm": 2.791166067123413,
      "learning_rate": 6.015135971069476e-05,
      "loss": 0.6089,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7730
    },
    {
      "epoch": 2.093589396808223,
      "grad_norm": 2.786120891571045,
      "learning_rate": 6.013314696705665e-05,
      "loss": 0.6132,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7740
    },
    {
      "epoch": 2.0962942926697323,
      "grad_norm": 2.6543073654174805,
      "learning_rate": 6.011610027468191e-05,
      "loss": 0.553,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7750
    },
    {
      "epoch": 2.0989991885312413,
      "grad_norm": 3.043158769607544,
      "learning_rate": 6.010021991807335e-05,
      "loss": 0.569,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7760
    },
    {
      "epoch": 2.101704084392751,
      "grad_norm": 2.7419517040252686,
      "learning_rate": 6.0085506162268076e-05,
      "loss": 0.5773,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7770
    },
    {
      "epoch": 2.1044089802542603,
      "grad_norm": 2.516381025314331,
      "learning_rate": 6.007195925283311e-05,
      "loss": 0.568,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7780
    },
    {
      "epoch": 2.1071138761157697,
      "grad_norm": 2.6597719192504883,
      "learning_rate": 6.005957941586122e-05,
      "loss": 0.6094,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7790
    },
    {
      "epoch": 2.1098187719772787,
      "grad_norm": 3.2095508575439453,
      "learning_rate": 6.00483668579672e-05,
      "loss": 0.5843,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7800
    },
    {
      "epoch": 2.112523667838788,
      "grad_norm": 3.1087212562561035,
      "learning_rate": 6.003832176628437e-05,
      "loss": 0.582,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7810
    },
    {
      "epoch": 2.1152285637002977,
      "grad_norm": 2.821138381958008,
      "learning_rate": 6.002944430846149e-05,
      "loss": 0.5834,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7820
    },
    {
      "epoch": 2.1179334595618067,
      "grad_norm": 2.795369863510132,
      "learning_rate": 6.002173463265999e-05,
      "loss": 0.587,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7830
    },
    {
      "epoch": 2.120638355423316,
      "grad_norm": 3.1482019424438477,
      "learning_rate": 6.001519286755142e-05,
      "loss": 0.5861,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7840
    },
    {
      "epoch": 2.1233432512848256,
      "grad_norm": 3.4291820526123047,
      "learning_rate": 6.000981912231535e-05,
      "loss": 0.5813,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7850
    },
    {
      "epoch": 2.126048147146335,
      "grad_norm": 2.541135311126709,
      "learning_rate": 6.000561348663755e-05,
      "loss": 0.5964,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7860
    },
    {
      "epoch": 2.128753043007844,
      "grad_norm": 2.603184461593628,
      "learning_rate": 6.00025760307085e-05,
      "loss": 0.5766,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7870
    },
    {
      "epoch": 2.1314579388693535,
      "grad_norm": 2.97566556930542,
      "learning_rate": 6.000070680522216e-05,
      "loss": 0.6151,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7880
    },
    {
      "epoch": 2.134162834730863,
      "grad_norm": 2.6513805389404297,
      "learning_rate": 6.0000005841375216e-05,
      "loss": 0.5447,
      "memory/device_mem_reserved(gib)": 33.64,
      "memory/max_mem_active(gib)": 31.19,
      "memory/max_mem_allocated(gib)": 31.19,
      "step": 7890
    }
  ],
  "logging_steps": 10,
  "max_steps": 7890,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 9.758329788086354e+18,
  "train_batch_size": 28,
  "trial_name": null,
  "trial_params": null
}