{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 10.0,
  "eval_steps": 500,
  "global_step": 1570,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.064,
      "grad_norm": 17.60050094292921,
      "learning_rate": 5.732484076433121e-07,
      "loss": 0.1956,
      "step": 10
    },
    {
      "epoch": 0.128,
      "grad_norm": 0.360872577474102,
      "learning_rate": 1.210191082802548e-06,
      "loss": 0.0375,
      "step": 20
    },
    {
      "epoch": 0.192,
      "grad_norm": 0.19682313468384985,
      "learning_rate": 1.8471337579617835e-06,
      "loss": 0.0033,
      "step": 30
    },
    {
      "epoch": 0.256,
      "grad_norm": 0.1063560099577119,
      "learning_rate": 2.4840764331210194e-06,
      "loss": 0.0024,
      "step": 40
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.0815568233065803,
      "learning_rate": 3.121019108280255e-06,
      "loss": 0.0023,
      "step": 50
    },
    {
      "epoch": 0.384,
      "grad_norm": 0.08743440250912729,
      "learning_rate": 3.757961783439491e-06,
      "loss": 0.0022,
      "step": 60
    },
    {
      "epoch": 0.448,
      "grad_norm": 0.08581063789526745,
      "learning_rate": 4.394904458598727e-06,
      "loss": 0.0024,
      "step": 70
    },
    {
      "epoch": 0.512,
      "grad_norm": 0.12211726984887894,
      "learning_rate": 5.031847133757962e-06,
      "loss": 0.0029,
      "step": 80
    },
    {
      "epoch": 0.576,
      "grad_norm": 0.07537531616308987,
      "learning_rate": 5.668789808917198e-06,
      "loss": 0.0021,
      "step": 90
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.1340344984825586,
      "learning_rate": 6.305732484076433e-06,
      "loss": 0.0019,
      "step": 100
    },
    {
      "epoch": 0.704,
      "grad_norm": 0.08953134246511378,
      "learning_rate": 6.942675159235669e-06,
      "loss": 0.0018,
      "step": 110
    },
    {
      "epoch": 0.768,
      "grad_norm": 0.0858578822485225,
      "learning_rate": 7.579617834394906e-06,
      "loss": 0.002,
      "step": 120
    },
    {
      "epoch": 0.832,
      "grad_norm": 0.0655815960237088,
      "learning_rate": 8.21656050955414e-06,
      "loss": 0.0022,
      "step": 130
    },
    {
      "epoch": 0.896,
      "grad_norm": 0.06259879392044645,
      "learning_rate": 8.853503184713377e-06,
      "loss": 0.0021,
      "step": 140
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.11115786959066185,
      "learning_rate": 9.490445859872613e-06,
      "loss": 0.0022,
      "step": 150
    },
    {
      "epoch": 1.0192,
      "grad_norm": 0.11253567844278162,
      "learning_rate": 9.985845718329796e-06,
      "loss": 0.0022,
      "step": 160
    },
    {
      "epoch": 1.0832,
      "grad_norm": 0.08722803511321106,
      "learning_rate": 9.91507430997877e-06,
      "loss": 0.0016,
      "step": 170
    },
    {
      "epoch": 1.1472,
      "grad_norm": 0.06678686196789735,
      "learning_rate": 9.844302901627743e-06,
      "loss": 0.0017,
      "step": 180
    },
    {
      "epoch": 1.2112,
      "grad_norm": 0.04730574861303533,
      "learning_rate": 9.773531493276717e-06,
      "loss": 0.0016,
      "step": 190
    },
    {
      "epoch": 1.2752,
      "grad_norm": 0.0531043878627717,
      "learning_rate": 9.70276008492569e-06,
      "loss": 0.0015,
      "step": 200
    },
    {
      "epoch": 1.3392,
      "grad_norm": 0.04150433810191339,
      "learning_rate": 9.631988676574666e-06,
      "loss": 0.0016,
      "step": 210
    },
    {
      "epoch": 1.4032,
      "grad_norm": 0.048744215851851264,
      "learning_rate": 9.561217268223637e-06,
      "loss": 0.0019,
      "step": 220
    },
    {
      "epoch": 1.4672,
      "grad_norm": 0.051141936058794614,
      "learning_rate": 9.490445859872613e-06,
      "loss": 0.0015,
      "step": 230
    },
    {
      "epoch": 1.5312000000000001,
      "grad_norm": 0.04473230852030802,
      "learning_rate": 9.419674451521586e-06,
      "loss": 0.0017,
      "step": 240
    },
    {
      "epoch": 1.5952,
      "grad_norm": 0.057517928281497044,
      "learning_rate": 9.34890304317056e-06,
      "loss": 0.0021,
      "step": 250
    },
    {
      "epoch": 1.6592,
      "grad_norm": 0.04626400515314615,
      "learning_rate": 9.278131634819534e-06,
      "loss": 0.0017,
      "step": 260
    },
    {
      "epoch": 1.7231999999999998,
      "grad_norm": 0.060866311940863976,
      "learning_rate": 9.207360226468509e-06,
      "loss": 0.0018,
      "step": 270
    },
    {
      "epoch": 1.7872,
      "grad_norm": 0.047260502110355984,
      "learning_rate": 9.13658881811748e-06,
      "loss": 0.0018,
      "step": 280
    },
    {
      "epoch": 1.8512,
      "grad_norm": 0.05629426497008717,
      "learning_rate": 9.065817409766456e-06,
      "loss": 0.002,
      "step": 290
    },
    {
      "epoch": 1.9152,
      "grad_norm": 0.03894980230831118,
      "learning_rate": 8.995046001415428e-06,
      "loss": 0.0018,
      "step": 300
    },
    {
      "epoch": 1.9792,
      "grad_norm": 0.0490808359112902,
      "learning_rate": 8.924274593064403e-06,
      "loss": 0.0023,
      "step": 310
    },
    {
      "epoch": 2.0384,
      "grad_norm": 0.04147961477820841,
      "learning_rate": 8.853503184713377e-06,
      "loss": 0.0014,
      "step": 320
    },
    {
      "epoch": 2.1024,
      "grad_norm": 0.06996184627519739,
      "learning_rate": 8.78273177636235e-06,
      "loss": 0.0011,
      "step": 330
    },
    {
      "epoch": 2.1664,
      "grad_norm": 0.059454727341288253,
      "learning_rate": 8.711960368011324e-06,
      "loss": 0.0012,
      "step": 340
    },
    {
      "epoch": 2.2304,
      "grad_norm": 0.04354879962710656,
      "learning_rate": 8.641188959660298e-06,
      "loss": 0.001,
      "step": 350
    },
    {
      "epoch": 2.2944,
      "grad_norm": 0.04462335944862259,
      "learning_rate": 8.570417551309271e-06,
      "loss": 0.0011,
      "step": 360
    },
    {
      "epoch": 2.3584,
      "grad_norm": 0.03976732177932052,
      "learning_rate": 8.499646142958245e-06,
      "loss": 0.0013,
      "step": 370
    },
    {
      "epoch": 2.4224,
      "grad_norm": 0.2689176723191193,
      "learning_rate": 8.42887473460722e-06,
      "loss": 0.0013,
      "step": 380
    },
    {
      "epoch": 2.4864,
      "grad_norm": 0.06879386397720086,
      "learning_rate": 8.358103326256194e-06,
      "loss": 0.0015,
      "step": 390
    },
    {
      "epoch": 2.5504,
      "grad_norm": 0.04660712249010195,
      "learning_rate": 8.287331917905167e-06,
      "loss": 0.0013,
      "step": 400
    },
    {
      "epoch": 2.6144,
      "grad_norm": 0.049288480673167614,
      "learning_rate": 8.21656050955414e-06,
      "loss": 0.0012,
      "step": 410
    },
    {
      "epoch": 2.6784,
      "grad_norm": 0.0736787873870253,
      "learning_rate": 8.145789101203114e-06,
      "loss": 0.0013,
      "step": 420
    },
    {
      "epoch": 2.7424,
      "grad_norm": 0.03501443136894185,
      "learning_rate": 8.075017692852088e-06,
      "loss": 0.0014,
      "step": 430
    },
    {
      "epoch": 2.8064,
      "grad_norm": 0.05662961561934753,
      "learning_rate": 8.004246284501063e-06,
      "loss": 0.0013,
      "step": 440
    },
    {
      "epoch": 2.8704,
      "grad_norm": 0.034163817502979116,
      "learning_rate": 7.933474876150035e-06,
      "loss": 0.0013,
      "step": 450
    },
    {
      "epoch": 2.9344,
      "grad_norm": 0.03627782201165022,
      "learning_rate": 7.86270346779901e-06,
      "loss": 0.0015,
      "step": 460
    },
    {
      "epoch": 2.9984,
      "grad_norm": 0.0378327369660673,
      "learning_rate": 7.791932059447984e-06,
      "loss": 0.0014,
      "step": 470
    },
    {
      "epoch": 3.0576,
      "grad_norm": 0.05647581306266388,
      "learning_rate": 7.721160651096958e-06,
      "loss": 0.0011,
      "step": 480
    },
    {
      "epoch": 3.1216,
      "grad_norm": 0.0359488394078568,
      "learning_rate": 7.650389242745931e-06,
      "loss": 0.0009,
      "step": 490
    },
    {
      "epoch": 3.1856,
      "grad_norm": 0.03876989103654088,
      "learning_rate": 7.579617834394906e-06,
      "loss": 0.0009,
      "step": 500
    },
    {
      "epoch": 3.2496,
      "grad_norm": 0.025465697683673762,
      "learning_rate": 7.5088464260438785e-06,
      "loss": 0.0009,
      "step": 510
    },
    {
      "epoch": 3.3136,
      "grad_norm": 0.03335723703353167,
      "learning_rate": 7.438075017692853e-06,
      "loss": 0.001,
      "step": 520
    },
    {
      "epoch": 3.3776,
      "grad_norm": 0.023575631149543714,
      "learning_rate": 7.367303609341826e-06,
      "loss": 0.0009,
      "step": 530
    },
    {
      "epoch": 3.4416,
      "grad_norm": 0.030623516564894863,
      "learning_rate": 7.2965322009908e-06,
      "loss": 0.0009,
      "step": 540
    },
    {
      "epoch": 3.5056000000000003,
      "grad_norm": 0.024108954363881076,
      "learning_rate": 7.2257607926397746e-06,
      "loss": 0.0012,
      "step": 550
    },
    {
      "epoch": 3.5696,
      "grad_norm": 0.05329738469797095,
      "learning_rate": 7.154989384288747e-06,
      "loss": 0.0011,
      "step": 560
    },
    {
      "epoch": 3.6336,
      "grad_norm": 0.05843959664676447,
      "learning_rate": 7.084217975937722e-06,
      "loss": 0.001,
      "step": 570
    },
    {
      "epoch": 3.6976,
      "grad_norm": 0.03573169633656267,
      "learning_rate": 7.013446567586696e-06,
      "loss": 0.001,
      "step": 580
    },
    {
      "epoch": 3.7616,
      "grad_norm": 0.03098931841718372,
      "learning_rate": 6.942675159235669e-06,
      "loss": 0.0011,
      "step": 590
    },
    {
      "epoch": 3.8256,
      "grad_norm": 0.029839006708716795,
      "learning_rate": 6.871903750884643e-06,
      "loss": 0.001,
      "step": 600
    },
    {
      "epoch": 3.8895999999999997,
      "grad_norm": 0.034632444699442086,
      "learning_rate": 6.801132342533617e-06,
      "loss": 0.0011,
      "step": 610
    },
    {
      "epoch": 3.9536,
      "grad_norm": 0.025514404182222564,
      "learning_rate": 6.730360934182591e-06,
      "loss": 0.0011,
      "step": 620
    },
    {
      "epoch": 4.0128,
      "grad_norm": 0.018837993447164435,
      "learning_rate": 6.659589525831564e-06,
      "loss": 0.0009,
      "step": 630
    },
    {
      "epoch": 4.0768,
      "grad_norm": 0.030434272679411154,
      "learning_rate": 6.588818117480539e-06,
      "loss": 0.0007,
      "step": 640
    },
    {
      "epoch": 4.1408,
      "grad_norm": 0.025073707651371353,
      "learning_rate": 6.518046709129512e-06,
      "loss": 0.0008,
      "step": 650
    },
    {
      "epoch": 4.2048,
      "grad_norm": 0.03899789663804376,
      "learning_rate": 6.447275300778486e-06,
      "loss": 0.0007,
      "step": 660
    },
    {
      "epoch": 4.2688,
      "grad_norm": 0.01589697038233231,
      "learning_rate": 6.37650389242746e-06,
      "loss": 0.0008,
      "step": 670
    },
    {
      "epoch": 4.3328,
      "grad_norm": 0.024560187315444824,
      "learning_rate": 6.305732484076433e-06,
      "loss": 0.0007,
      "step": 680
    },
    {
      "epoch": 4.3968,
      "grad_norm": 0.02518563194507897,
      "learning_rate": 6.2349610757254074e-06,
      "loss": 0.0008,
      "step": 690
    },
    {
      "epoch": 4.4608,
      "grad_norm": 0.02855642906111709,
      "learning_rate": 6.164189667374382e-06,
      "loss": 0.0008,
      "step": 700
    },
    {
      "epoch": 4.5248,
      "grad_norm": 0.03178443031646811,
      "learning_rate": 6.093418259023355e-06,
      "loss": 0.0009,
      "step": 710
    },
    {
      "epoch": 4.5888,
      "grad_norm": 0.021081075903526782,
      "learning_rate": 6.022646850672329e-06,
      "loss": 0.0008,
      "step": 720
    },
    {
      "epoch": 4.6528,
      "grad_norm": 0.018073305283650833,
      "learning_rate": 5.9518754423213035e-06,
      "loss": 0.0008,
      "step": 730
    },
    {
      "epoch": 4.7168,
      "grad_norm": 0.04811056774674395,
      "learning_rate": 5.881104033970276e-06,
      "loss": 0.0008,
      "step": 740
    },
    {
      "epoch": 4.7808,
      "grad_norm": 0.018757359542175032,
      "learning_rate": 5.810332625619251e-06,
      "loss": 0.0009,
      "step": 750
    },
    {
      "epoch": 4.8448,
      "grad_norm": 0.021213357394449274,
      "learning_rate": 5.7395612172682235e-06,
      "loss": 0.001,
      "step": 760
    },
    {
      "epoch": 4.9088,
      "grad_norm": 0.03275271950493163,
      "learning_rate": 5.668789808917198e-06,
      "loss": 0.0009,
      "step": 770
    },
    {
      "epoch": 4.9728,
      "grad_norm": 0.023316003525068154,
      "learning_rate": 5.598018400566172e-06,
      "loss": 0.0009,
      "step": 780
    },
    {
      "epoch": 5.032,
      "grad_norm": 0.014334840854956359,
      "learning_rate": 5.527246992215145e-06,
      "loss": 0.0007,
      "step": 790
    },
    {
      "epoch": 5.096,
      "grad_norm": 0.05422632736773663,
      "learning_rate": 5.4564755838641195e-06,
      "loss": 0.0005,
      "step": 800
    },
    {
      "epoch": 5.16,
      "grad_norm": 0.018449386679577014,
      "learning_rate": 5.385704175513093e-06,
      "loss": 0.0006,
      "step": 810
    },
    {
      "epoch": 5.224,
      "grad_norm": 0.02919022100324906,
      "learning_rate": 5.314932767162067e-06,
      "loss": 0.0006,
      "step": 820
    },
    {
      "epoch": 5.288,
      "grad_norm": 0.0305959616492891,
      "learning_rate": 5.24416135881104e-06,
      "loss": 0.0006,
      "step": 830
    },
    {
      "epoch": 5.352,
      "grad_norm": 0.022695467463977156,
      "learning_rate": 5.173389950460015e-06,
      "loss": 0.0006,
      "step": 840
    },
    {
      "epoch": 5.416,
      "grad_norm": 0.02483139320478547,
      "learning_rate": 5.102618542108988e-06,
      "loss": 0.0007,
      "step": 850
    },
    {
      "epoch": 5.48,
      "grad_norm": 0.01866673577908835,
      "learning_rate": 5.031847133757962e-06,
      "loss": 0.0007,
      "step": 860
    },
    {
      "epoch": 5.5440000000000005,
      "grad_norm": 0.02475309749140289,
      "learning_rate": 4.9610757254069355e-06,
      "loss": 0.0007,
      "step": 870
    },
    {
      "epoch": 5.608,
      "grad_norm": 0.02329874201328524,
      "learning_rate": 4.89030431705591e-06,
      "loss": 0.0007,
      "step": 880
    },
    {
      "epoch": 5.672,
      "grad_norm": 0.017899965001059195,
      "learning_rate": 4.819532908704884e-06,
      "loss": 0.0007,
      "step": 890
    },
    {
      "epoch": 5.736,
      "grad_norm": 0.018801411907769315,
      "learning_rate": 4.748761500353857e-06,
      "loss": 0.0007,
      "step": 900
    },
    {
      "epoch": 5.8,
      "grad_norm": 0.015720546212916624,
      "learning_rate": 4.677990092002832e-06,
      "loss": 0.0007,
      "step": 910
    },
    {
      "epoch": 5.864,
      "grad_norm": 0.03502870587760174,
      "learning_rate": 4.607218683651805e-06,
      "loss": 0.0008,
      "step": 920
    },
    {
      "epoch": 5.928,
      "grad_norm": 0.03386452230790969,
      "learning_rate": 4.536447275300779e-06,
      "loss": 0.0007,
      "step": 930
    },
    {
      "epoch": 5.992,
      "grad_norm": 0.03351146213175313,
      "learning_rate": 4.465675866949752e-06,
      "loss": 0.0007,
      "step": 940
    },
    {
      "epoch": 6.0512,
      "grad_norm": 0.011861931780566704,
      "learning_rate": 4.394904458598727e-06,
      "loss": 0.0006,
      "step": 950
    },
    {
      "epoch": 6.1152,
      "grad_norm": 0.02438378480663504,
      "learning_rate": 4.3241330502477004e-06,
      "loss": 0.0005,
      "step": 960
    },
    {
      "epoch": 6.1792,
      "grad_norm": 0.01716531081341717,
      "learning_rate": 4.253361641896674e-06,
      "loss": 0.0006,
      "step": 970
    },
    {
      "epoch": 6.2432,
      "grad_norm": 0.01938921457239796,
      "learning_rate": 4.1825902335456485e-06,
      "loss": 0.0006,
      "step": 980
    },
    {
      "epoch": 6.3072,
      "grad_norm": 0.01373221841561274,
      "learning_rate": 4.111818825194622e-06,
      "loss": 0.0005,
      "step": 990
    },
    {
      "epoch": 6.3712,
      "grad_norm": 0.0199597209044529,
      "learning_rate": 4.041047416843596e-06,
      "loss": 0.0006,
      "step": 1000
    },
    {
      "epoch": 6.4352,
      "grad_norm": 0.02424156176996985,
      "learning_rate": 3.970276008492569e-06,
      "loss": 0.0006,
      "step": 1010
    },
    {
      "epoch": 6.4992,
      "grad_norm": 0.025315655480968886,
      "learning_rate": 3.899504600141543e-06,
      "loss": 0.0006,
      "step": 1020
    },
    {
      "epoch": 6.5632,
      "grad_norm": 0.02920235438042742,
      "learning_rate": 3.8287331917905165e-06,
      "loss": 0.0006,
      "step": 1030
    },
    {
      "epoch": 6.6272,
      "grad_norm": 0.016555192151659806,
      "learning_rate": 3.757961783439491e-06,
      "loss": 0.0006,
      "step": 1040
    },
    {
      "epoch": 6.6912,
      "grad_norm": 0.013049019694846958,
      "learning_rate": 3.6871903750884645e-06,
      "loss": 0.0006,
      "step": 1050
    },
    {
      "epoch": 6.7552,
      "grad_norm": 0.02447313938635393,
      "learning_rate": 3.6164189667374385e-06,
      "loss": 0.0006,
      "step": 1060
    },
    {
      "epoch": 6.8192,
      "grad_norm": 0.011543889477958485,
      "learning_rate": 3.545647558386412e-06,
      "loss": 0.0006,
      "step": 1070
    },
    {
      "epoch": 6.8832,
      "grad_norm": 0.020945153083698518,
      "learning_rate": 3.474876150035386e-06,
      "loss": 0.0006,
      "step": 1080
    },
    {
      "epoch": 6.9472000000000005,
      "grad_norm": 0.052172806158250935,
      "learning_rate": 3.4041047416843597e-06,
      "loss": 0.0007,
      "step": 1090
    },
    {
      "epoch": 7.0064,
      "grad_norm": 0.04186589672688367,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 0.0006,
      "step": 1100
    },
    {
      "epoch": 7.0704,
      "grad_norm": 0.023917480677921623,
      "learning_rate": 3.2625619249823078e-06,
      "loss": 0.0006,
      "step": 1110
    },
    {
      "epoch": 7.1344,
      "grad_norm": 0.021509274576004597,
      "learning_rate": 3.1917905166312814e-06,
      "loss": 0.0005,
      "step": 1120
    },
    {
      "epoch": 7.1984,
      "grad_norm": 0.014051823896192348,
      "learning_rate": 3.121019108280255e-06,
      "loss": 0.0005,
      "step": 1130
    },
    {
      "epoch": 7.2624,
      "grad_norm": 0.042071708786109156,
      "learning_rate": 3.050247699929229e-06,
      "loss": 0.0005,
      "step": 1140
    },
    {
      "epoch": 7.3264,
      "grad_norm": 0.015899068217884903,
      "learning_rate": 2.9794762915782026e-06,
      "loss": 0.0005,
      "step": 1150
    },
    {
      "epoch": 7.3904,
      "grad_norm": 0.01397192533675872,
      "learning_rate": 2.908704883227176e-06,
      "loss": 0.0005,
      "step": 1160
    },
    {
      "epoch": 7.4544,
      "grad_norm": 0.026963623776955922,
      "learning_rate": 2.83793347487615e-06,
      "loss": 0.0005,
      "step": 1170
    },
    {
      "epoch": 7.5184,
      "grad_norm": 0.039247352934989634,
      "learning_rate": 2.767162066525124e-06,
      "loss": 0.0005,
      "step": 1180
    },
    {
      "epoch": 7.5824,
      "grad_norm": 0.014656470038010723,
      "learning_rate": 2.696390658174098e-06,
      "loss": 0.0005,
      "step": 1190
    },
    {
      "epoch": 7.6464,
      "grad_norm": 0.01623246283065862,
      "learning_rate": 2.6256192498230714e-06,
      "loss": 0.0006,
      "step": 1200
    },
    {
      "epoch": 7.7104,
      "grad_norm": 0.01364971630536226,
      "learning_rate": 2.554847841472046e-06,
      "loss": 0.0005,
      "step": 1210
    },
    {
      "epoch": 7.7744,
      "grad_norm": 0.01843596725314251,
      "learning_rate": 2.4840764331210194e-06,
      "loss": 0.0006,
      "step": 1220
    },
    {
      "epoch": 7.8384,
      "grad_norm": 0.01324833994986271,
      "learning_rate": 2.413305024769993e-06,
      "loss": 0.0005,
      "step": 1230
    },
    {
      "epoch": 7.9024,
      "grad_norm": 0.010850155510887709,
      "learning_rate": 2.342533616418967e-06,
      "loss": 0.0005,
      "step": 1240
    },
    {
      "epoch": 7.9664,
      "grad_norm": 0.01975586454883896,
      "learning_rate": 2.2717622080679406e-06,
      "loss": 0.0006,
      "step": 1250
    },
    {
      "epoch": 8.0256,
      "grad_norm": 0.022062764393436206,
      "learning_rate": 2.2009907997169142e-06,
      "loss": 0.0006,
      "step": 1260
    },
    {
      "epoch": 8.0896,
      "grad_norm": 0.011242064989508685,
      "learning_rate": 2.1302193913658883e-06,
      "loss": 0.0004,
      "step": 1270
    },
    {
      "epoch": 8.1536,
      "grad_norm": 0.01149746346411442,
      "learning_rate": 2.0594479830148623e-06,
      "loss": 0.0004,
      "step": 1280
    },
    {
      "epoch": 8.2176,
      "grad_norm": 0.026515020603223683,
      "learning_rate": 1.988676574663836e-06,
      "loss": 0.0005,
      "step": 1290
    },
    {
      "epoch": 8.2816,
      "grad_norm": 0.013807929815403902,
      "learning_rate": 1.91790516631281e-06,
      "loss": 0.0005,
      "step": 1300
    },
    {
      "epoch": 8.3456,
      "grad_norm": 0.013074536666619924,
      "learning_rate": 1.8471337579617835e-06,
      "loss": 0.0005,
      "step": 1310
    },
    {
      "epoch": 8.4096,
      "grad_norm": 0.01259520749871116,
      "learning_rate": 1.7763623496107573e-06,
      "loss": 0.0005,
      "step": 1320
    },
    {
      "epoch": 8.4736,
      "grad_norm": 0.01040755614337679,
      "learning_rate": 1.7055909412597313e-06,
      "loss": 0.0005,
      "step": 1330
    },
    {
      "epoch": 8.5376,
      "grad_norm": 0.019423750818007783,
      "learning_rate": 1.634819532908705e-06,
      "loss": 0.0005,
      "step": 1340
    },
    {
      "epoch": 8.6016,
      "grad_norm": 0.02075041086847078,
      "learning_rate": 1.564048124557679e-06,
      "loss": 0.0006,
      "step": 1350
    },
    {
      "epoch": 8.6656,
      "grad_norm": 0.012939220630062887,
      "learning_rate": 1.4932767162066527e-06,
      "loss": 0.0005,
      "step": 1360
    },
    {
      "epoch": 8.7296,
      "grad_norm": 0.013042241021270056,
      "learning_rate": 1.4225053078556263e-06,
      "loss": 0.0005,
      "step": 1370
    },
    {
      "epoch": 8.7936,
      "grad_norm": 0.014428179836796827,
      "learning_rate": 1.3517338995046003e-06,
      "loss": 0.0005,
      "step": 1380
    },
    {
      "epoch": 8.8576,
      "grad_norm": 0.02246660360791134,
      "learning_rate": 1.280962491153574e-06,
      "loss": 0.0005,
      "step": 1390
    },
    {
      "epoch": 8.9216,
      "grad_norm": 0.024316300351936923,
      "learning_rate": 1.210191082802548e-06,
      "loss": 0.0005,
      "step": 1400
    },
    {
      "epoch": 8.9856,
      "grad_norm": 0.017404048203407522,
      "learning_rate": 1.1394196744515216e-06,
      "loss": 0.0005,
      "step": 1410
    },
    {
      "epoch": 9.0448,
      "grad_norm": 0.012844194995988387,
      "learning_rate": 1.0686482661004954e-06,
      "loss": 0.0005,
      "step": 1420
    },
    {
      "epoch": 9.1088,
      "grad_norm": 0.020036814618503134,
      "learning_rate": 9.978768577494694e-07,
      "loss": 0.0005,
      "step": 1430
    },
    {
      "epoch": 9.1728,
      "grad_norm": 0.00765325259778791,
      "learning_rate": 9.271054493984431e-07,
      "loss": 0.0004,
      "step": 1440
    },
    {
      "epoch": 9.2368,
      "grad_norm": 0.01507156630584454,
      "learning_rate": 8.563340410474169e-07,
      "loss": 0.0005,
      "step": 1450
    },
    {
      "epoch": 9.3008,
      "grad_norm": 0.012047267490532228,
      "learning_rate": 7.855626326963907e-07,
      "loss": 0.0005,
      "step": 1460
    },
    {
      "epoch": 9.3648,
      "grad_norm": 0.016105089807538026,
      "learning_rate": 7.147912243453645e-07,
      "loss": 0.0005,
      "step": 1470
    },
    {
      "epoch": 9.4288,
      "grad_norm": 0.014283181711916166,
      "learning_rate": 6.440198159943384e-07,
      "loss": 0.0005,
      "step": 1480
    },
    {
      "epoch": 9.4928,
      "grad_norm": 0.009125720569816429,
      "learning_rate": 5.732484076433121e-07,
      "loss": 0.0005,
      "step": 1490
    },
    {
      "epoch": 9.556799999999999,
      "grad_norm": 0.011829314376872695,
      "learning_rate": 5.024769992922859e-07,
      "loss": 0.0005,
      "step": 1500
    },
    {
      "epoch": 9.6208,
      "grad_norm": 0.012106395568977475,
      "learning_rate": 4.317055909412598e-07,
      "loss": 0.0004,
      "step": 1510
    },
    {
      "epoch": 9.6848,
      "grad_norm": 0.011903236971213026,
      "learning_rate": 3.6093418259023354e-07,
      "loss": 0.0005,
      "step": 1520
    },
    {
      "epoch": 9.7488,
      "grad_norm": 0.013644455361646967,
      "learning_rate": 2.901627742392074e-07,
      "loss": 0.0005,
      "step": 1530
    },
    {
      "epoch": 9.8128,
      "grad_norm": 0.014986000780857893,
      "learning_rate": 2.1939136588818118e-07,
      "loss": 0.0005,
      "step": 1540
    },
    {
      "epoch": 9.8768,
      "grad_norm": 0.017539322056519816,
      "learning_rate": 1.4861995753715502e-07,
      "loss": 0.0005,
      "step": 1550
    },
    {
      "epoch": 9.9408,
      "grad_norm": 0.016233742810560718,
      "learning_rate": 7.784854918612881e-08,
      "loss": 0.0004,
      "step": 1560
    },
    {
      "epoch": 10.0,
      "grad_norm": 0.0117183501383805,
      "learning_rate": 7.077140835102619e-09,
      "loss": 0.0005,
      "step": 1570
    },
    {
      "epoch": 10.0,
      "step": 1570,
      "total_flos": 538322966937600.0,
      "train_loss": 0.0024459090680002595,
      "train_runtime": 33809.8122,
      "train_samples_per_second": 5.915,
      "train_steps_per_second": 0.046
    }
  ],
  "logging_steps": 10,
  "max_steps": 1570,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 538322966937600.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}