{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 105,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.009523809523809525,
      "grad_norm": 2.447430372238159,
      "learning_rate": 0.0,
      "loss": 1.8171,
      "memory/device_mem_reserved(gib)": 36.42,
      "memory/max_mem_active(gib)": 35.17,
      "memory/max_mem_allocated(gib)": 35.17,
      "step": 1
    },
    {
      "epoch": 0.01904761904761905,
      "grad_norm": 2.280487537384033,
      "learning_rate": 2.2222222222222223e-05,
      "loss": 1.7544,
      "memory/device_mem_reserved(gib)": 36.45,
      "memory/max_mem_active(gib)": 35.23,
      "memory/max_mem_allocated(gib)": 35.23,
      "step": 2
    },
    {
      "epoch": 0.02857142857142857,
      "grad_norm": 2.248966693878174,
      "learning_rate": 4.4444444444444447e-05,
      "loss": 1.7546,
      "memory/device_mem_reserved(gib)": 47.1,
      "memory/max_mem_active(gib)": 45.45,
      "memory/max_mem_allocated(gib)": 45.45,
      "step": 3
    },
    {
      "epoch": 0.0380952380952381,
      "grad_norm": 2.2559540271759033,
      "learning_rate": 6.666666666666667e-05,
      "loss": 1.8011,
      "memory/device_mem_reserved(gib)": 47.1,
      "memory/max_mem_active(gib)": 45.45,
      "memory/max_mem_allocated(gib)": 45.45,
      "step": 4
    },
    {
      "epoch": 0.047619047619047616,
      "grad_norm": 2.149872064590454,
      "learning_rate": 8.888888888888889e-05,
      "loss": 1.6465,
      "memory/device_mem_reserved(gib)": 47.1,
      "memory/max_mem_active(gib)": 45.45,
      "memory/max_mem_allocated(gib)": 45.45,
      "step": 5
    },
    {
      "epoch": 0.05714285714285714,
      "grad_norm": 1.6111228466033936,
      "learning_rate": 0.00011111111111111112,
      "loss": 1.5935,
      "memory/device_mem_reserved(gib)": 47.1,
      "memory/max_mem_active(gib)": 45.45,
      "memory/max_mem_allocated(gib)": 45.45,
      "step": 6
    },
    {
      "epoch": 0.06666666666666667,
      "grad_norm": 1.363952875137329,
      "learning_rate": 0.00013333333333333334,
      "loss": 1.4371,
      "memory/device_mem_reserved(gib)": 47.1,
      "memory/max_mem_active(gib)": 45.45,
      "memory/max_mem_allocated(gib)": 45.45,
      "step": 7
    },
    {
      "epoch": 0.0761904761904762,
      "grad_norm": 1.1439425945281982,
      "learning_rate": 0.00015555555555555556,
      "loss": 1.2409,
      "memory/device_mem_reserved(gib)": 47.1,
      "memory/max_mem_active(gib)": 45.45,
      "memory/max_mem_allocated(gib)": 45.45,
      "step": 8
    },
    {
      "epoch": 0.08571428571428572,
      "grad_norm": 0.877252995967865,
      "learning_rate": 0.00017777777777777779,
      "loss": 1.1083,
      "memory/device_mem_reserved(gib)": 47.1,
      "memory/max_mem_active(gib)": 45.45,
      "memory/max_mem_allocated(gib)": 45.45,
      "step": 9
    },
    {
      "epoch": 0.09523809523809523,
      "grad_norm": 0.8071479797363281,
      "learning_rate": 0.0002,
      "loss": 0.9937,
      "memory/device_mem_reserved(gib)": 47.1,
      "memory/max_mem_active(gib)": 45.45,
      "memory/max_mem_allocated(gib)": 45.45,
      "step": 10
    },
    {
      "epoch": 0.10476190476190476,
      "grad_norm": 0.7772606015205383,
      "learning_rate": 0.00019999472984871732,
      "loss": 0.9235,
      "memory/device_mem_reserved(gib)": 47.1,
      "memory/max_mem_active(gib)": 45.45,
      "memory/max_mem_allocated(gib)": 45.45,
      "step": 11
    },
    {
      "epoch": 0.11428571428571428,
      "grad_norm": 0.811764657497406,
      "learning_rate": 0.00019997891995035912,
      "loss": 0.7635,
      "memory/device_mem_reserved(gib)": 47.1,
      "memory/max_mem_active(gib)": 45.45,
      "memory/max_mem_allocated(gib)": 45.45,
      "step": 12
    },
    {
      "epoch": 0.12380952380952381,
      "grad_norm": 0.8423659801483154,
      "learning_rate": 0.0001999525719713366,
      "loss": 0.6836,
      "memory/device_mem_reserved(gib)": 47.1,
      "memory/max_mem_active(gib)": 45.45,
      "memory/max_mem_allocated(gib)": 45.45,
      "step": 13
    },
    {
      "epoch": 0.13333333333333333,
      "grad_norm": 0.7858280539512634,
      "learning_rate": 0.0001999156886888064,
      "loss": 0.6159,
      "memory/device_mem_reserved(gib)": 47.1,
      "memory/max_mem_active(gib)": 45.45,
      "memory/max_mem_allocated(gib)": 45.45,
      "step": 14
    },
    {
      "epoch": 0.14285714285714285,
      "grad_norm": 0.739560604095459,
      "learning_rate": 0.00019986827399037812,
      "loss": 0.5082,
      "memory/device_mem_reserved(gib)": 47.1,
      "memory/max_mem_active(gib)": 45.45,
      "memory/max_mem_allocated(gib)": 45.45,
      "step": 15
    },
    {
      "epoch": 0.1523809523809524,
      "grad_norm": 0.6787500381469727,
      "learning_rate": 0.00019981033287370443,
      "loss": 0.4553,
      "memory/device_mem_reserved(gib)": 47.1,
      "memory/max_mem_active(gib)": 45.45,
      "memory/max_mem_allocated(gib)": 45.45,
      "step": 16
    },
    {
      "epoch": 0.1619047619047619,
      "grad_norm": 0.6258607506752014,
      "learning_rate": 0.00019974187144595432,
      "loss": 0.3913,
      "memory/device_mem_reserved(gib)": 47.1,
      "memory/max_mem_active(gib)": 45.45,
      "memory/max_mem_allocated(gib)": 45.45,
      "step": 17
    },
    {
      "epoch": 0.17142857142857143,
      "grad_norm": 0.47939175367355347,
      "learning_rate": 0.00019966289692316944,
      "loss": 0.3048,
      "memory/device_mem_reserved(gib)": 47.1,
      "memory/max_mem_active(gib)": 45.45,
      "memory/max_mem_allocated(gib)": 45.45,
      "step": 18
    },
    {
      "epoch": 0.18095238095238095,
      "grad_norm": 0.49321815371513367,
      "learning_rate": 0.00019957341762950344,
      "loss": 0.3417,
      "memory/device_mem_reserved(gib)": 47.1,
      "memory/max_mem_active(gib)": 45.45,
      "memory/max_mem_allocated(gib)": 45.45,
      "step": 19
    },
    {
      "epoch": 0.19047619047619047,
      "grad_norm": 0.4252234101295471,
      "learning_rate": 0.00019947344299634464,
      "loss": 0.3033,
      "memory/device_mem_reserved(gib)": 47.1,
      "memory/max_mem_active(gib)": 45.45,
      "memory/max_mem_allocated(gib)": 45.45,
      "step": 20
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.46225014328956604,
      "learning_rate": 0.00019936298356132176,
      "loss": 0.2984,
      "memory/device_mem_reserved(gib)": 47.1,
      "memory/max_mem_active(gib)": 45.45,
      "memory/max_mem_allocated(gib)": 45.45,
      "step": 21
    },
    {
      "epoch": 0.20952380952380953,
      "grad_norm": 0.4106466472148895,
      "learning_rate": 0.0001992420509671936,
      "loss": 0.2764,
      "memory/device_mem_reserved(gib)": 47.12,
      "memory/max_mem_active(gib)": 45.46,
      "memory/max_mem_allocated(gib)": 45.46,
      "step": 22
    },
    {
      "epoch": 0.21904761904761905,
      "grad_norm": 0.3764067590236664,
      "learning_rate": 0.00019911065796062135,
      "loss": 0.2352,
      "memory/device_mem_reserved(gib)": 47.12,
      "memory/max_mem_active(gib)": 45.46,
      "memory/max_mem_allocated(gib)": 45.46,
      "step": 23
    },
    {
      "epoch": 0.22857142857142856,
      "grad_norm": 0.38518962264060974,
      "learning_rate": 0.00019896881839082556,
      "loss": 0.2239,
      "memory/device_mem_reserved(gib)": 47.12,
      "memory/max_mem_active(gib)": 45.46,
      "memory/max_mem_allocated(gib)": 45.46,
      "step": 24
    },
    {
      "epoch": 0.23809523809523808,
      "grad_norm": 0.39728403091430664,
      "learning_rate": 0.00019881654720812594,
      "loss": 0.2192,
      "memory/device_mem_reserved(gib)": 47.13,
      "memory/max_mem_active(gib)": 45.46,
      "memory/max_mem_allocated(gib)": 45.46,
      "step": 25
    },
    {
      "epoch": 0.24761904761904763,
      "grad_norm": 0.3667093813419342,
      "learning_rate": 0.00019865386046236596,
      "loss": 0.2834,
      "memory/device_mem_reserved(gib)": 47.13,
      "memory/max_mem_active(gib)": 45.46,
      "memory/max_mem_allocated(gib)": 45.46,
      "step": 26
    },
    {
      "epoch": 0.2571428571428571,
      "grad_norm": 0.32502660155296326,
      "learning_rate": 0.00019848077530122083,
      "loss": 0.2166,
      "memory/device_mem_reserved(gib)": 47.13,
      "memory/max_mem_active(gib)": 45.46,
      "memory/max_mem_allocated(gib)": 45.46,
      "step": 27
    },
    {
      "epoch": 0.26666666666666666,
      "grad_norm": 0.31225672364234924,
      "learning_rate": 0.0001982973099683902,
      "loss": 0.2238,
      "memory/device_mem_reserved(gib)": 47.13,
      "memory/max_mem_active(gib)": 45.46,
      "memory/max_mem_allocated(gib)": 45.46,
      "step": 28
    },
    {
      "epoch": 0.2761904761904762,
      "grad_norm": 0.27448731660842896,
      "learning_rate": 0.00019810348380167527,
      "loss": 0.1983,
      "memory/device_mem_reserved(gib)": 47.13,
      "memory/max_mem_active(gib)": 45.46,
      "memory/max_mem_allocated(gib)": 45.46,
      "step": 29
    },
    {
      "epoch": 0.2857142857142857,
      "grad_norm": 0.32014158368110657,
      "learning_rate": 0.00019789931723094046,
      "loss": 0.2554,
      "memory/device_mem_reserved(gib)": 47.13,
      "memory/max_mem_active(gib)": 45.46,
      "memory/max_mem_allocated(gib)": 45.46,
      "step": 30
    },
    {
      "epoch": 0.29523809523809524,
      "grad_norm": 0.39480143785476685,
      "learning_rate": 0.0001976848317759601,
      "loss": 0.2438,
      "memory/device_mem_reserved(gib)": 47.13,
      "memory/max_mem_active(gib)": 45.46,
      "memory/max_mem_allocated(gib)": 45.46,
      "step": 31
    },
    {
      "epoch": 0.3047619047619048,
      "grad_norm": 0.28004321455955505,
      "learning_rate": 0.00019746005004415005,
      "loss": 0.2145,
      "memory/device_mem_reserved(gib)": 47.13,
      "memory/max_mem_active(gib)": 45.46,
      "memory/max_mem_allocated(gib)": 45.46,
      "step": 32
    },
    {
      "epoch": 0.3142857142857143,
      "grad_norm": 0.2754361033439636,
      "learning_rate": 0.00019722499572818496,
      "loss": 0.2284,
      "memory/device_mem_reserved(gib)": 47.13,
      "memory/max_mem_active(gib)": 45.46,
      "memory/max_mem_allocated(gib)": 45.46,
      "step": 33
    },
    {
      "epoch": 0.3238095238095238,
      "grad_norm": 0.3160915970802307,
      "learning_rate": 0.00019697969360350098,
      "loss": 0.2482,
      "memory/device_mem_reserved(gib)": 47.13,
      "memory/max_mem_active(gib)": 45.47,
      "memory/max_mem_allocated(gib)": 45.47,
      "step": 34
    },
    {
      "epoch": 0.3333333333333333,
      "grad_norm": 0.24699093401432037,
      "learning_rate": 0.00019672416952568416,
      "loss": 0.1916,
      "memory/device_mem_reserved(gib)": 47.13,
      "memory/max_mem_active(gib)": 45.47,
      "memory/max_mem_allocated(gib)": 45.47,
      "step": 35
    },
    {
      "epoch": 0.34285714285714286,
      "grad_norm": 0.2674485743045807,
      "learning_rate": 0.00019645845042774553,
      "loss": 0.2204,
      "memory/device_mem_reserved(gib)": 47.13,
      "memory/max_mem_active(gib)": 45.47,
      "memory/max_mem_allocated(gib)": 45.47,
      "step": 36
    },
    {
      "epoch": 0.3523809523809524,
      "grad_norm": 0.33608749508857727,
      "learning_rate": 0.00019618256431728194,
      "loss": 0.235,
      "memory/device_mem_reserved(gib)": 47.13,
      "memory/max_mem_active(gib)": 45.47,
      "memory/max_mem_allocated(gib)": 45.47,
      "step": 37
    },
    {
      "epoch": 0.3619047619047619,
      "grad_norm": 0.29581594467163086,
      "learning_rate": 0.00019589654027352414,
      "loss": 0.2175,
      "memory/device_mem_reserved(gib)": 47.13,
      "memory/max_mem_active(gib)": 45.47,
      "memory/max_mem_allocated(gib)": 45.47,
      "step": 38
    },
    {
      "epoch": 0.37142857142857144,
      "grad_norm": 0.3526640832424164,
      "learning_rate": 0.0001956004084442718,
      "loss": 0.1989,
      "memory/device_mem_reserved(gib)": 47.13,
      "memory/max_mem_active(gib)": 45.47,
      "memory/max_mem_allocated(gib)": 45.47,
      "step": 39
    },
    {
      "epoch": 0.38095238095238093,
      "grad_norm": 0.2882135808467865,
      "learning_rate": 0.00019529420004271567,
      "loss": 0.2144,
      "memory/device_mem_reserved(gib)": 47.13,
      "memory/max_mem_active(gib)": 45.47,
      "memory/max_mem_allocated(gib)": 45.47,
      "step": 40
    },
    {
      "epoch": 0.3904761904761905,
      "grad_norm": 0.2675183117389679,
      "learning_rate": 0.0001949779473441478,
      "loss": 0.1987,
      "memory/device_mem_reserved(gib)": 47.13,
      "memory/max_mem_active(gib)": 45.47,
      "memory/max_mem_allocated(gib)": 45.47,
      "step": 41
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.2908264696598053,
      "learning_rate": 0.00019465168368255946,
      "loss": 0.1893,
      "memory/device_mem_reserved(gib)": 47.13,
      "memory/max_mem_active(gib)": 45.47,
      "memory/max_mem_allocated(gib)": 45.47,
      "step": 42
    },
    {
      "epoch": 0.4095238095238095,
      "grad_norm": 0.3211234211921692,
      "learning_rate": 0.00019431544344712776,
      "loss": 0.2008,
      "memory/device_mem_reserved(gib)": 47.13,
      "memory/max_mem_active(gib)": 45.47,
      "memory/max_mem_allocated(gib)": 45.47,
      "step": 43
    },
    {
      "epoch": 0.41904761904761906,
      "grad_norm": 0.2795293927192688,
      "learning_rate": 0.00019396926207859084,
      "loss": 0.1935,
      "memory/device_mem_reserved(gib)": 47.13,
      "memory/max_mem_active(gib)": 45.47,
      "memory/max_mem_allocated(gib)": 45.47,
      "step": 44
    },
    {
      "epoch": 0.42857142857142855,
      "grad_norm": 0.28679871559143066,
      "learning_rate": 0.00019361317606551238,
      "loss": 0.1987,
      "memory/device_mem_reserved(gib)": 47.13,
      "memory/max_mem_active(gib)": 45.47,
      "memory/max_mem_allocated(gib)": 45.47,
      "step": 45
    },
    {
      "epoch": 0.4380952380952381,
      "grad_norm": 0.3165067732334137,
      "learning_rate": 0.00019324722294043558,
      "loss": 0.2118,
      "memory/device_mem_reserved(gib)": 47.13,
      "memory/max_mem_active(gib)": 45.47,
      "memory/max_mem_allocated(gib)": 45.47,
      "step": 46
    },
    {
      "epoch": 0.44761904761904764,
      "grad_norm": 0.25720369815826416,
      "learning_rate": 0.00019287144127592704,
      "loss": 0.1962,
      "memory/device_mem_reserved(gib)": 47.13,
      "memory/max_mem_active(gib)": 45.47,
      "memory/max_mem_allocated(gib)": 45.47,
      "step": 47
    },
    {
      "epoch": 0.45714285714285713,
      "grad_norm": 0.2577202022075653,
      "learning_rate": 0.0001924858706805112,
      "loss": 0.2022,
      "memory/device_mem_reserved(gib)": 47.13,
      "memory/max_mem_active(gib)": 45.47,
      "memory/max_mem_allocated(gib)": 45.47,
      "step": 48
    },
    {
      "epoch": 0.4666666666666667,
      "grad_norm": 0.2605520188808441,
      "learning_rate": 0.0001920905517944954,
      "loss": 0.1891,
      "memory/device_mem_reserved(gib)": 47.13,
      "memory/max_mem_active(gib)": 45.47,
      "memory/max_mem_allocated(gib)": 45.47,
      "step": 49
    },
    {
      "epoch": 0.47619047619047616,
      "grad_norm": 0.2890350818634033,
      "learning_rate": 0.00019168552628568631,
      "loss": 0.17,
      "memory/device_mem_reserved(gib)": 47.13,
      "memory/max_mem_active(gib)": 45.47,
      "memory/max_mem_allocated(gib)": 45.47,
      "step": 50
    },
    {
      "epoch": 0.4857142857142857,
      "grad_norm": 0.25532475113868713,
      "learning_rate": 0.00019127083684499806,
      "loss": 0.1796,
      "memory/device_mem_reserved(gib)": 47.13,
      "memory/max_mem_active(gib)": 45.47,
      "memory/max_mem_allocated(gib)": 45.47,
      "step": 51
    },
    {
      "epoch": 0.49523809523809526,
      "grad_norm": 0.2963904142379761,
      "learning_rate": 0.00019084652718195238,
      "loss": 0.19,
      "memory/device_mem_reserved(gib)": 47.13,
      "memory/max_mem_active(gib)": 45.47,
      "memory/max_mem_allocated(gib)": 45.47,
      "step": 52
    },
    {
      "epoch": 0.5047619047619047,
      "grad_norm": 0.24455095827579498,
      "learning_rate": 0.0001904126420200716,
      "loss": 0.2017,
      "memory/device_mem_reserved(gib)": 47.13,
      "memory/max_mem_active(gib)": 45.47,
      "memory/max_mem_allocated(gib)": 45.47,
      "step": 53
    },
    {
      "epoch": 0.5142857142857142,
      "grad_norm": 0.2518145442008972,
      "learning_rate": 0.00018996922709216455,
      "loss": 0.1896,
      "memory/device_mem_reserved(gib)": 47.13,
      "memory/max_mem_active(gib)": 45.47,
      "memory/max_mem_allocated(gib)": 45.47,
      "step": 54
    },
    {
      "epoch": 0.5238095238095238,
      "grad_norm": 0.26225805282592773,
      "learning_rate": 0.00018951632913550626,
      "loss": 0.1893,
      "memory/device_mem_reserved(gib)": 47.13,
      "memory/max_mem_active(gib)": 45.47,
      "memory/max_mem_allocated(gib)": 45.47,
      "step": 55
    },
    {
      "epoch": 0.5333333333333333,
      "grad_norm": 0.25169169902801514,
      "learning_rate": 0.00018905399588691163,
      "loss": 0.1939,
      "memory/device_mem_reserved(gib)": 47.13,
      "memory/max_mem_active(gib)": 45.47,
      "memory/max_mem_allocated(gib)": 45.47,
      "step": 56
    },
    {
      "epoch": 0.5428571428571428,
      "grad_norm": 0.23847182095050812,
      "learning_rate": 0.00018858227607770398,
      "loss": 0.1686,
      "memory/device_mem_reserved(gib)": 47.13,
      "memory/max_mem_active(gib)": 45.47,
      "memory/max_mem_allocated(gib)": 45.47,
      "step": 57
    },
    {
      "epoch": 0.5523809523809524,
      "grad_norm": 0.30888280272483826,
      "learning_rate": 0.00018810121942857845,
      "loss": 0.1768,
      "memory/device_mem_reserved(gib)": 47.13,
      "memory/max_mem_active(gib)": 45.47,
      "memory/max_mem_allocated(gib)": 45.47,
      "step": 58
    },
    {
      "epoch": 0.5619047619047619,
      "grad_norm": 0.27275580167770386,
      "learning_rate": 0.00018761087664436138,
      "loss": 0.1941,
      "memory/device_mem_reserved(gib)": 47.13,
      "memory/max_mem_active(gib)": 45.47,
      "memory/max_mem_allocated(gib)": 45.47,
      "step": 59
    },
    {
      "epoch": 0.5714285714285714,
      "grad_norm": 0.22851359844207764,
      "learning_rate": 0.00018711129940866575,
      "loss": 0.1658,
      "memory/device_mem_reserved(gib)": 47.13,
      "memory/max_mem_active(gib)": 45.47,
      "memory/max_mem_allocated(gib)": 45.47,
      "step": 60
    },
    {
      "epoch": 0.580952380952381,
      "grad_norm": 0.25515425205230713,
      "learning_rate": 0.00018660254037844388,
      "loss": 0.1858,
      "memory/device_mem_reserved(gib)": 47.13,
      "memory/max_mem_active(gib)": 45.47,
      "memory/max_mem_allocated(gib)": 45.47,
      "step": 61
    },
    {
      "epoch": 0.5904761904761905,
      "grad_norm": 0.24301765859127045,
      "learning_rate": 0.00018608465317843678,
      "loss": 0.1892,
      "memory/device_mem_reserved(gib)": 47.13,
      "memory/max_mem_active(gib)": 45.47,
      "memory/max_mem_allocated(gib)": 45.47,
      "step": 62
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.24329794943332672,
      "learning_rate": 0.00018555769239552233,
      "loss": 0.1719,
      "memory/device_mem_reserved(gib)": 47.13,
      "memory/max_mem_active(gib)": 45.47,
      "memory/max_mem_allocated(gib)": 45.47,
      "step": 63
    },
    {
      "epoch": 0.6095238095238096,
      "grad_norm": 0.3206610381603241,
      "learning_rate": 0.00018502171357296144,
      "loss": 0.2075,
      "memory/device_mem_reserved(gib)": 47.13,
      "memory/max_mem_active(gib)": 45.47,
      "memory/max_mem_allocated(gib)": 45.47,
      "step": 64
    },
    {
      "epoch": 0.6190476190476191,
      "grad_norm": 0.3089618980884552,
      "learning_rate": 0.00018447677320454367,
      "loss": 0.185,
      "memory/device_mem_reserved(gib)": 47.13,
      "memory/max_mem_active(gib)": 45.47,
      "memory/max_mem_allocated(gib)": 45.47,
      "step": 65
    },
    {
      "epoch": 0.6285714285714286,
      "grad_norm": 0.291916161775589,
      "learning_rate": 0.00018392292872863267,
      "loss": 0.1739,
      "memory/device_mem_reserved(gib)": 47.13,
      "memory/max_mem_active(gib)": 45.47,
      "memory/max_mem_allocated(gib)": 45.47,
      "step": 66
    },
    {
      "epoch": 0.638095238095238,
      "grad_norm": 0.25783228874206543,
      "learning_rate": 0.00018336023852211195,
      "loss": 0.2061,
      "memory/device_mem_reserved(gib)": 47.13,
      "memory/max_mem_active(gib)": 45.47,
      "memory/max_mem_allocated(gib)": 45.47,
      "step": 67
    },
    {
      "epoch": 0.6476190476190476,
      "grad_norm": 0.310614675283432,
      "learning_rate": 0.00018278876189423179,
      "loss": 0.1984,
      "memory/device_mem_reserved(gib)": 47.13,
      "memory/max_mem_active(gib)": 45.47,
      "memory/max_mem_allocated(gib)": 45.47,
      "step": 68
    },
    {
      "epoch": 0.6571428571428571,
      "grad_norm": 0.23836325109004974,
      "learning_rate": 0.00018220855908035783,
      "loss": 0.1742,
      "memory/device_mem_reserved(gib)": 47.13,
      "memory/max_mem_active(gib)": 45.47,
      "memory/max_mem_allocated(gib)": 45.47,
      "step": 69
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 0.30577945709228516,
      "learning_rate": 0.0001816196912356222,
      "loss": 0.1954,
      "memory/device_mem_reserved(gib)": 47.13,
      "memory/max_mem_active(gib)": 45.47,
      "memory/max_mem_allocated(gib)": 45.47,
      "step": 70
    },
    {
      "epoch": 0.6761904761904762,
      "grad_norm": 0.3177741467952728,
      "learning_rate": 0.00018102222042847737,
      "loss": 0.193,
      "memory/device_mem_reserved(gib)": 47.13,
      "memory/max_mem_active(gib)": 45.47,
      "memory/max_mem_allocated(gib)": 45.47,
      "step": 71
    },
    {
      "epoch": 0.6857142857142857,
      "grad_norm": 0.3444816470146179,
      "learning_rate": 0.00018041620963415417,
      "loss": 0.1837,
      "memory/device_mem_reserved(gib)": 47.13,
      "memory/max_mem_active(gib)": 45.47,
      "memory/max_mem_allocated(gib)": 45.47,
      "step": 72
    },
    {
      "epoch": 0.6952380952380952,
      "grad_norm": 0.3139638900756836,
      "learning_rate": 0.000179801722728024,
      "loss": 0.1902,
      "memory/device_mem_reserved(gib)": 47.13,
      "memory/max_mem_active(gib)": 45.47,
      "memory/max_mem_allocated(gib)": 45.47,
      "step": 73
    },
    {
      "epoch": 0.7047619047619048,
      "grad_norm": 0.24754184484481812,
      "learning_rate": 0.00017917882447886582,
      "loss": 0.1612,
      "memory/device_mem_reserved(gib)": 47.13,
      "memory/max_mem_active(gib)": 45.47,
      "memory/max_mem_allocated(gib)": 45.47,
      "step": 74
    },
    {
      "epoch": 0.7142857142857143,
      "grad_norm": 0.23204036056995392,
      "learning_rate": 0.00017854758054203988,
      "loss": 0.1694,
      "memory/device_mem_reserved(gib)": 47.13,
      "memory/max_mem_active(gib)": 45.47,
      "memory/max_mem_allocated(gib)": 45.47,
      "step": 75
    },
    {
      "epoch": 0.7238095238095238,
      "grad_norm": 0.2823260426521301,
      "learning_rate": 0.00017790805745256704,
      "loss": 0.1739,
      "memory/device_mem_reserved(gib)": 47.13,
      "memory/max_mem_active(gib)": 45.47,
      "memory/max_mem_allocated(gib)": 45.47,
      "step": 76
    },
    {
      "epoch": 0.7333333333333333,
      "grad_norm": 0.3193572163581848,
      "learning_rate": 0.0001772603226181159,
      "loss": 0.1924,
      "memory/device_mem_reserved(gib)": 47.13,
      "memory/max_mem_active(gib)": 45.47,
      "memory/max_mem_allocated(gib)": 45.47,
      "step": 77
    },
    {
      "epoch": 0.7428571428571429,
      "grad_norm": 0.30530717968940735,
      "learning_rate": 0.0001766044443118978,
      "loss": 0.1375,
      "memory/device_mem_reserved(gib)": 47.13,
      "memory/max_mem_active(gib)": 45.47,
      "memory/max_mem_allocated(gib)": 45.47,
      "step": 78
    },
    {
      "epoch": 0.7523809523809524,
      "grad_norm": 0.2974378764629364,
      "learning_rate": 0.00017594049166547073,
      "loss": 0.1866,
      "memory/device_mem_reserved(gib)": 47.13,
      "memory/max_mem_active(gib)": 45.47,
      "memory/max_mem_allocated(gib)": 45.47,
      "step": 79
    },
    {
      "epoch": 0.7619047619047619,
      "grad_norm": 0.26553574204444885,
      "learning_rate": 0.00017526853466145244,
      "loss": 0.193,
      "memory/device_mem_reserved(gib)": 47.13,
      "memory/max_mem_active(gib)": 45.47,
      "memory/max_mem_allocated(gib)": 45.47,
      "step": 80
    },
    {
      "epoch": 0.7714285714285715,
      "grad_norm": 0.32059594988822937,
      "learning_rate": 0.00017458864412614434,
      "loss": 0.1822,
      "memory/device_mem_reserved(gib)": 47.13,
      "memory/max_mem_active(gib)": 45.47,
      "memory/max_mem_allocated(gib)": 45.47,
      "step": 81
    },
    {
      "epoch": 0.780952380952381,
      "grad_norm": 0.3260822594165802,
      "learning_rate": 0.00017390089172206592,
      "loss": 0.1734,
      "memory/device_mem_reserved(gib)": 47.13,
      "memory/max_mem_active(gib)": 45.47,
      "memory/max_mem_allocated(gib)": 45.47,
      "step": 82
    },
    {
      "epoch": 0.7904761904761904,
      "grad_norm": 0.2754499912261963,
      "learning_rate": 0.00017320534994040148,
      "loss": 0.1814,
      "memory/device_mem_reserved(gib)": 47.13,
      "memory/max_mem_active(gib)": 45.47,
      "memory/max_mem_allocated(gib)": 45.47,
      "step": 83
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.2639872431755066,
      "learning_rate": 0.00017250209209335927,
      "loss": 0.1764,
      "memory/device_mem_reserved(gib)": 47.13,
      "memory/max_mem_active(gib)": 45.47,
      "memory/max_mem_allocated(gib)": 45.47,
      "step": 84
    },
    {
      "epoch": 0.8095238095238095,
      "grad_norm": 0.22905386984348297,
      "learning_rate": 0.0001717911923064442,
      "loss": 0.1365,
      "memory/device_mem_reserved(gib)": 47.13,
      "memory/max_mem_active(gib)": 45.47,
      "memory/max_mem_allocated(gib)": 45.47,
      "step": 85
    },
    {
      "epoch": 0.819047619047619,
      "grad_norm": 0.26005449891090393,
      "learning_rate": 0.00017107272551064473,
      "loss": 0.1756,
      "memory/device_mem_reserved(gib)": 47.13,
      "memory/max_mem_active(gib)": 45.47,
      "memory/max_mem_allocated(gib)": 45.47,
      "step": 86
    },
    {
      "epoch": 0.8285714285714286,
      "grad_norm": 0.25042417645454407,
      "learning_rate": 0.00017034676743453499,
      "loss": 0.15,
      "memory/device_mem_reserved(gib)": 47.13,
      "memory/max_mem_active(gib)": 45.47,
      "memory/max_mem_allocated(gib)": 45.47,
      "step": 87
    },
    {
      "epoch": 0.8380952380952381,
      "grad_norm": 0.34358277916908264,
      "learning_rate": 0.0001696133945962927,
      "loss": 0.1763,
      "memory/device_mem_reserved(gib)": 47.13,
      "memory/max_mem_active(gib)": 45.47,
      "memory/max_mem_allocated(gib)": 45.47,
      "step": 88
    },
    {
      "epoch": 0.8476190476190476,
      "grad_norm": 0.23855257034301758,
      "learning_rate": 0.0001688726842956339,
      "loss": 0.1547,
      "memory/device_mem_reserved(gib)": 47.13,
      "memory/max_mem_active(gib)": 45.47,
      "memory/max_mem_allocated(gib)": 45.47,
      "step": 89
    },
    {
      "epoch": 0.8571428571428571,
      "grad_norm": 0.2579350173473358,
      "learning_rate": 0.0001681247146056654,
      "loss": 0.1872,
      "memory/device_mem_reserved(gib)": 47.13,
      "memory/max_mem_active(gib)": 45.47,
      "memory/max_mem_allocated(gib)": 45.47,
      "step": 90
    },
    {
      "epoch": 0.8666666666666667,
      "grad_norm": 0.30034390091896057,
      "learning_rate": 0.00016736956436465573,
      "loss": 0.1558,
      "memory/device_mem_reserved(gib)": 47.13,
      "memory/max_mem_active(gib)": 45.47,
      "memory/max_mem_allocated(gib)": 45.47,
      "step": 91
    },
    {
      "epoch": 0.8761904761904762,
      "grad_norm": 0.28185054659843445,
      "learning_rate": 0.00016660731316772505,
      "loss": 0.1878,
      "memory/device_mem_reserved(gib)": 47.13,
      "memory/max_mem_active(gib)": 45.47,
      "memory/max_mem_allocated(gib)": 45.47,
      "step": 92
    },
    {
      "epoch": 0.8857142857142857,
      "grad_norm": 0.32514333724975586,
      "learning_rate": 0.0001658380413584558,
      "loss": 0.1705,
      "memory/device_mem_reserved(gib)": 47.13,
      "memory/max_mem_active(gib)": 45.47,
      "memory/max_mem_allocated(gib)": 45.47,
      "step": 93
    },
    {
      "epoch": 0.8952380952380953,
      "grad_norm": 0.30903860926628113,
      "learning_rate": 0.0001650618300204242,
      "loss": 0.1708,
      "memory/device_mem_reserved(gib)": 47.13,
      "memory/max_mem_active(gib)": 45.47,
      "memory/max_mem_allocated(gib)": 45.47,
      "step": 94
    },
    {
      "epoch": 0.9047619047619048,
      "grad_norm": 0.24723806977272034,
      "learning_rate": 0.00016427876096865394,
      "loss": 0.1571,
      "memory/device_mem_reserved(gib)": 47.13,
      "memory/max_mem_active(gib)": 45.47,
      "memory/max_mem_allocated(gib)": 45.47,
      "step": 95
    },
    {
      "epoch": 0.9142857142857143,
      "grad_norm": 0.2727813422679901,
      "learning_rate": 0.0001634889167409923,
      "loss": 0.1779,
      "memory/device_mem_reserved(gib)": 47.13,
      "memory/max_mem_active(gib)": 45.47,
      "memory/max_mem_allocated(gib)": 45.47,
      "step": 96
    },
    {
      "epoch": 0.9238095238095239,
      "grad_norm": 0.2521083354949951,
      "learning_rate": 0.0001626923805894107,
      "loss": 0.1451,
      "memory/device_mem_reserved(gib)": 47.13,
      "memory/max_mem_active(gib)": 45.47,
      "memory/max_mem_allocated(gib)": 45.47,
      "step": 97
    },
    {
      "epoch": 0.9333333333333333,
      "grad_norm": 0.25517475605010986,
      "learning_rate": 0.00016188923647122947,
      "loss": 0.1789,
      "memory/device_mem_reserved(gib)": 47.13,
      "memory/max_mem_active(gib)": 45.47,
      "memory/max_mem_allocated(gib)": 45.47,
      "step": 98
    },
    {
      "epoch": 0.9428571428571428,
      "grad_norm": 0.26679396629333496,
      "learning_rate": 0.0001610795690402688,
      "loss": 0.1835,
      "memory/device_mem_reserved(gib)": 47.13,
      "memory/max_mem_active(gib)": 45.47,
      "memory/max_mem_allocated(gib)": 45.47,
      "step": 99
    },
    {
      "epoch": 0.9523809523809523,
      "grad_norm": 0.22741852700710297,
      "learning_rate": 0.00016026346363792567,
      "loss": 0.1599,
      "memory/device_mem_reserved(gib)": 47.13,
      "memory/max_mem_active(gib)": 45.47,
      "memory/max_mem_allocated(gib)": 45.47,
      "step": 100
    },
    {
      "epoch": 0.9619047619047619,
      "grad_norm": 0.28691986203193665,
      "learning_rate": 0.00015944100628417868,
      "loss": 0.1654,
      "memory/device_mem_reserved(gib)": 47.13,
      "memory/max_mem_active(gib)": 45.47,
      "memory/max_mem_allocated(gib)": 45.47,
      "step": 101
    },
    {
      "epoch": 0.9714285714285714,
      "grad_norm": 0.24840302765369415,
      "learning_rate": 0.00015861228366852148,
      "loss": 0.1584,
      "memory/device_mem_reserved(gib)": 47.13,
      "memory/max_mem_active(gib)": 45.47,
      "memory/max_mem_allocated(gib)": 45.47,
      "step": 102
    },
    {
      "epoch": 0.9809523809523809,
      "grad_norm": 0.24804599583148956,
      "learning_rate": 0.00015777738314082514,
      "loss": 0.1391,
      "memory/device_mem_reserved(gib)": 47.13,
      "memory/max_mem_active(gib)": 45.47,
      "memory/max_mem_allocated(gib)": 45.47,
      "step": 103
    },
    {
      "epoch": 0.9904761904761905,
      "grad_norm": 0.2463735193014145,
      "learning_rate": 0.00015693639270213136,
      "loss": 0.1408,
      "memory/device_mem_reserved(gib)": 47.13,
      "memory/max_mem_active(gib)": 45.47,
      "memory/max_mem_allocated(gib)": 45.47,
      "step": 104
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.29896217584609985,
      "learning_rate": 0.000156089400995377,
      "loss": 0.1666,
      "memory/device_mem_reserved(gib)": 47.13,
      "memory/max_mem_active(gib)": 45.47,
      "memory/max_mem_allocated(gib)": 45.47,
      "step": 105
    }
  ],
  "logging_steps": 1,
  "max_steps": 315,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.6079273462857728e+16,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}