{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 54600,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.004578796509125542,
      "grad_norm": 2.848018169403076,
      "learning_rate": 2.28021978021978e-05,
      "loss": 3.32117333984375,
      "step": 250
    },
    {
      "epoch": 0.009157593018251083,
      "grad_norm": 3.734480142593384,
      "learning_rate": 4.5695970695970695e-05,
      "loss": 1.6078341064453125,
      "step": 500
    },
    {
      "epoch": 0.013736389527376625,
      "grad_norm": 1.8585106134414673,
      "learning_rate": 6.85897435897436e-05,
      "loss": 1.2332542724609374,
      "step": 750
    },
    {
      "epoch": 0.018315186036502167,
      "grad_norm": 1.4551349878311157,
      "learning_rate": 9.148351648351648e-05,
      "loss": 1.11114892578125,
      "step": 1000
    },
    {
      "epoch": 0.022893982545627706,
      "grad_norm": 0.9371363520622253,
      "learning_rate": 0.00011437728937728938,
      "loss": 1.02870654296875,
      "step": 1250
    },
    {
      "epoch": 0.02747277905475325,
      "grad_norm": 0.7468550801277161,
      "learning_rate": 0.00013727106227106228,
      "loss": 0.9670721435546875,
      "step": 1500
    },
    {
      "epoch": 0.03205157556387879,
      "grad_norm": 0.4782555401325226,
      "learning_rate": 0.00016016483516483518,
      "loss": 0.9067027587890625,
      "step": 1750
    },
    {
      "epoch": 0.03663037207300433,
      "grad_norm": 0.36062029004096985,
      "learning_rate": 0.00018305860805860804,
      "loss": 0.8399249267578125,
      "step": 2000
    },
    {
      "epoch": 0.041209168582129876,
      "grad_norm": 0.29721319675445557,
      "learning_rate": 0.00020595238095238094,
      "loss": 0.80204296875,
      "step": 2250
    },
    {
      "epoch": 0.04578796509125541,
      "grad_norm": 0.2194763571023941,
      "learning_rate": 0.00022884615384615383,
      "loss": 0.7781054077148437,
      "step": 2500
    },
    {
      "epoch": 0.050366761600380956,
      "grad_norm": 0.28626367449760437,
      "learning_rate": 0.0002517399267399267,
      "loss": 0.759437744140625,
      "step": 2750
    },
    {
      "epoch": 0.0549455581095065,
      "grad_norm": 0.18353205919265747,
      "learning_rate": 0.00027463369963369965,
      "loss": 0.7461485595703125,
      "step": 3000
    },
    {
      "epoch": 0.059524354618632036,
      "grad_norm": 0.1829853504896164,
      "learning_rate": 0.0002975274725274725,
      "loss": 0.7340205688476562,
      "step": 3250
    },
    {
      "epoch": 0.06410315112775758,
      "grad_norm": 0.17057757079601288,
      "learning_rate": 0.00032042124542124543,
      "loss": 0.7216248168945313,
      "step": 3500
    },
    {
      "epoch": 0.06868194763688312,
      "grad_norm": 0.15916429460048676,
      "learning_rate": 0.0003433150183150183,
      "loss": 0.7127350463867187,
      "step": 3750
    },
    {
      "epoch": 0.07326074414600867,
      "grad_norm": 0.13619782030582428,
      "learning_rate": 0.00036620879120879117,
      "loss": 0.704721923828125,
      "step": 4000
    },
    {
      "epoch": 0.07783954065513421,
      "grad_norm": 0.1638839691877365,
      "learning_rate": 0.00038910256410256414,
      "loss": 0.697459228515625,
      "step": 4250
    },
    {
      "epoch": 0.08241833716425975,
      "grad_norm": 0.1422419250011444,
      "learning_rate": 0.000411996336996337,
      "loss": 0.6936741943359375,
      "step": 4500
    },
    {
      "epoch": 0.08699713367338528,
      "grad_norm": 0.11795759946107864,
      "learning_rate": 0.00043489010989010993,
      "loss": 0.6863809814453125,
      "step": 4750
    },
    {
      "epoch": 0.09157593018251083,
      "grad_norm": 0.12040770798921585,
      "learning_rate": 0.0004577838827838828,
      "loss": 0.6824598388671875,
      "step": 5000
    },
    {
      "epoch": 0.09615472669163637,
      "grad_norm": 0.10738283395767212,
      "learning_rate": 0.0004806776556776557,
      "loss": 0.677829833984375,
      "step": 5250
    },
    {
      "epoch": 0.10073352320076191,
      "grad_norm": 0.11751581728458405,
      "learning_rate": 0.0004996031746031746,
      "loss": 0.6711824951171875,
      "step": 5500
    },
    {
      "epoch": 0.10531231970988746,
      "grad_norm": 0.11239449679851532,
      "learning_rate": 0.000497059422059422,
      "loss": 0.669865966796875,
      "step": 5750
    },
    {
      "epoch": 0.109891116219013,
      "grad_norm": 0.11104018986225128,
      "learning_rate": 0.0004945156695156695,
      "loss": 0.6635555419921875,
      "step": 6000
    },
    {
      "epoch": 0.11446991272813854,
      "grad_norm": 0.10596565157175064,
      "learning_rate": 0.0004919719169719169,
      "loss": 0.65825244140625,
      "step": 6250
    },
    {
      "epoch": 0.11904870923726407,
      "grad_norm": 0.10106229782104492,
      "learning_rate": 0.0004894281644281644,
      "loss": 0.655421630859375,
      "step": 6500
    },
    {
      "epoch": 0.12362750574638962,
      "grad_norm": 0.09866887331008911,
      "learning_rate": 0.0004868844118844119,
      "loss": 0.64850390625,
      "step": 6750
    },
    {
      "epoch": 0.12820630225551516,
      "grad_norm": 0.08995469659566879,
      "learning_rate": 0.00048434065934065937,
      "loss": 0.6459603271484375,
      "step": 7000
    },
    {
      "epoch": 0.1327850987646407,
      "grad_norm": 0.08975188434123993,
      "learning_rate": 0.0004817969067969068,
      "loss": 0.642448486328125,
      "step": 7250
    },
    {
      "epoch": 0.13736389527376625,
      "grad_norm": 0.09255772829055786,
      "learning_rate": 0.00047925315425315426,
      "loss": 0.6384821166992187,
      "step": 7500
    },
    {
      "epoch": 0.1419426917828918,
      "grad_norm": 0.09045405685901642,
      "learning_rate": 0.00047670940170940176,
      "loss": 0.6367230224609375,
      "step": 7750
    },
    {
      "epoch": 0.14652148829201733,
      "grad_norm": 0.08942568302154541,
      "learning_rate": 0.0004741656491656492,
      "loss": 0.6320804443359375,
      "step": 8000
    },
    {
      "epoch": 0.15110028480114288,
      "grad_norm": 0.0798359364271164,
      "learning_rate": 0.00047162189662189666,
      "loss": 0.630943359375,
      "step": 8250
    },
    {
      "epoch": 0.15567908131026842,
      "grad_norm": 0.08887682855129242,
      "learning_rate": 0.0004690781440781441,
      "loss": 0.6291470947265625,
      "step": 8500
    },
    {
      "epoch": 0.16025787781939396,
      "grad_norm": 0.08339811861515045,
      "learning_rate": 0.00046653439153439155,
      "loss": 0.6263828735351562,
      "step": 8750
    },
    {
      "epoch": 0.1648366743285195,
      "grad_norm": 0.07981119304895401,
      "learning_rate": 0.000463990638990639,
      "loss": 0.622819091796875,
      "step": 9000
    },
    {
      "epoch": 0.16941547083764502,
      "grad_norm": 0.08066423237323761,
      "learning_rate": 0.0004614468864468865,
      "loss": 0.62232373046875,
      "step": 9250
    },
    {
      "epoch": 0.17399426734677056,
      "grad_norm": 0.08569934964179993,
      "learning_rate": 0.00045890313390313395,
      "loss": 0.6196232299804687,
      "step": 9500
    },
    {
      "epoch": 0.1785730638558961,
      "grad_norm": 0.07788624614477158,
      "learning_rate": 0.0004563593813593814,
      "loss": 0.6152354736328125,
      "step": 9750
    },
    {
      "epoch": 0.18315186036502165,
      "grad_norm": 0.08253168314695358,
      "learning_rate": 0.00045381562881562884,
      "loss": 0.61491357421875,
      "step": 10000
    },
    {
      "epoch": 0.1877306568741472,
      "grad_norm": 0.07646993547677994,
      "learning_rate": 0.0004512718762718763,
      "loss": 0.6130479125976562,
      "step": 10250
    },
    {
      "epoch": 0.19230945338327274,
      "grad_norm": 0.08307339996099472,
      "learning_rate": 0.00044872812372812374,
      "loss": 0.6101732177734375,
      "step": 10500
    },
    {
      "epoch": 0.19688824989239828,
      "grad_norm": 0.09029801934957504,
      "learning_rate": 0.00044618437118437124,
      "loss": 0.608197021484375,
      "step": 10750
    },
    {
      "epoch": 0.20146704640152382,
      "grad_norm": 0.07931365817785263,
      "learning_rate": 0.0004436406186406187,
      "loss": 0.6081648559570313,
      "step": 11000
    },
    {
      "epoch": 0.20604584291064937,
      "grad_norm": 0.08072754740715027,
      "learning_rate": 0.00044109686609686613,
      "loss": 0.604770751953125,
      "step": 11250
    },
    {
      "epoch": 0.2106246394197749,
      "grad_norm": 0.07558581978082657,
      "learning_rate": 0.0004385531135531136,
      "loss": 0.6046954956054688,
      "step": 11500
    },
    {
      "epoch": 0.21520343592890046,
      "grad_norm": 0.08121418952941895,
      "learning_rate": 0.00043600936100936103,
      "loss": 0.6042259521484376,
      "step": 11750
    },
    {
      "epoch": 0.219782232438026,
      "grad_norm": 0.07723360508680344,
      "learning_rate": 0.0004334656084656085,
      "loss": 0.6015655517578125,
      "step": 12000
    },
    {
      "epoch": 0.22436102894715154,
      "grad_norm": 0.07910860329866409,
      "learning_rate": 0.000430921855921856,
      "loss": 0.6005347900390625,
      "step": 12250
    },
    {
      "epoch": 0.22893982545627709,
      "grad_norm": 0.07615973055362701,
      "learning_rate": 0.0004283781033781034,
      "loss": 0.5994979248046876,
      "step": 12500
    },
    {
      "epoch": 0.2335186219654026,
      "grad_norm": 0.07880640774965286,
      "learning_rate": 0.00042583435083435087,
      "loss": 0.5965632934570313,
      "step": 12750
    },
    {
      "epoch": 0.23809741847452814,
      "grad_norm": 0.07529937475919724,
      "learning_rate": 0.0004232905982905983,
      "loss": 0.597671142578125,
      "step": 13000
    },
    {
      "epoch": 0.2426762149836537,
      "grad_norm": 0.08778645843267441,
      "learning_rate": 0.00042074684574684577,
      "loss": 0.5972130737304687,
      "step": 13250
    },
    {
      "epoch": 0.24725501149277923,
      "grad_norm": 0.07203567028045654,
      "learning_rate": 0.0004182030932030932,
      "loss": 0.5954554443359374,
      "step": 13500
    },
    {
      "epoch": 0.2518338080019048,
      "grad_norm": 0.07240218669176102,
      "learning_rate": 0.0004156593406593407,
      "loss": 0.59423681640625,
      "step": 13750
    },
    {
      "epoch": 0.2564126045110303,
      "grad_norm": 0.07828880846500397,
      "learning_rate": 0.00041311558811558816,
      "loss": 0.593840576171875,
      "step": 14000
    },
    {
      "epoch": 0.2609914010201559,
      "grad_norm": 0.08217330276966095,
      "learning_rate": 0.0004105718355718356,
      "loss": 0.589479736328125,
      "step": 14250
    },
    {
      "epoch": 0.2655701975292814,
      "grad_norm": 0.08034295588731766,
      "learning_rate": 0.00040802808302808306,
      "loss": 0.59010009765625,
      "step": 14500
    },
    {
      "epoch": 0.2701489940384069,
      "grad_norm": 0.07022342830896378,
      "learning_rate": 0.0004054843304843305,
      "loss": 0.5894483642578126,
      "step": 14750
    },
    {
      "epoch": 0.2747277905475325,
      "grad_norm": 0.072320356965065,
      "learning_rate": 0.0004029405779405779,
      "loss": 0.5865537109375,
      "step": 15000
    },
    {
      "epoch": 0.279306587056658,
      "grad_norm": 0.0711088627576828,
      "learning_rate": 0.00040039682539682545,
      "loss": 0.587281005859375,
      "step": 15250
    },
    {
      "epoch": 0.2838853835657836,
      "grad_norm": 0.07165085524320602,
      "learning_rate": 0.0003978530728530729,
      "loss": 0.5887276611328125,
      "step": 15500
    },
    {
      "epoch": 0.2884641800749091,
      "grad_norm": 0.07520093768835068,
      "learning_rate": 0.0003953093203093203,
      "loss": 0.586100341796875,
      "step": 15750
    },
    {
      "epoch": 0.29304297658403466,
      "grad_norm": 0.07864724844694138,
      "learning_rate": 0.00039276556776556774,
      "loss": 0.585071533203125,
      "step": 16000
    },
    {
      "epoch": 0.2976217730931602,
      "grad_norm": 0.07668787240982056,
      "learning_rate": 0.0003902218152218152,
      "loss": 0.5834588623046875,
      "step": 16250
    },
    {
      "epoch": 0.30220056960228575,
      "grad_norm": 0.07498002797365189,
      "learning_rate": 0.00038767806267806263,
      "loss": 0.5809573974609376,
      "step": 16500
    },
    {
      "epoch": 0.30677936611141127,
      "grad_norm": 0.06531748920679092,
      "learning_rate": 0.00038513431013431014,
      "loss": 0.5816318969726563,
      "step": 16750
    },
    {
      "epoch": 0.31135816262053684,
      "grad_norm": 0.07089975476264954,
      "learning_rate": 0.0003825905575905576,
      "loss": 0.5804892578125,
      "step": 17000
    },
    {
      "epoch": 0.31593695912966235,
      "grad_norm": 0.07220444083213806,
      "learning_rate": 0.00038004680504680503,
      "loss": 0.5800992431640625,
      "step": 17250
    },
    {
      "epoch": 0.3205157556387879,
      "grad_norm": 0.08013845980167389,
      "learning_rate": 0.0003775030525030525,
      "loss": 0.5800794677734376,
      "step": 17500
    },
    {
      "epoch": 0.32509455214791344,
      "grad_norm": 0.07268445193767548,
      "learning_rate": 0.0003749592999592999,
      "loss": 0.5797327880859375,
      "step": 17750
    },
    {
      "epoch": 0.329673348657039,
      "grad_norm": 0.06680341064929962,
      "learning_rate": 0.0003724155474155474,
      "loss": 0.5781187133789063,
      "step": 18000
    },
    {
      "epoch": 0.3342521451661645,
      "grad_norm": 0.07133681327104568,
      "learning_rate": 0.0003698717948717949,
      "loss": 0.5785706176757812,
      "step": 18250
    },
    {
      "epoch": 0.33883094167529004,
      "grad_norm": 0.06551344692707062,
      "learning_rate": 0.0003673280423280423,
      "loss": 0.5755133056640624,
      "step": 18500
    },
    {
      "epoch": 0.3434097381844156,
      "grad_norm": 0.0680370181798935,
      "learning_rate": 0.00036478428978428977,
      "loss": 0.575121826171875,
      "step": 18750
    },
    {
      "epoch": 0.34798853469354113,
      "grad_norm": 0.07072953879833221,
      "learning_rate": 0.0003622405372405372,
      "loss": 0.5754507446289062,
      "step": 19000
    },
    {
      "epoch": 0.3525673312026667,
      "grad_norm": 0.07297784835100174,
      "learning_rate": 0.00035969678469678466,
      "loss": 0.5739122924804687,
      "step": 19250
    },
    {
      "epoch": 0.3571461277117922,
      "grad_norm": 0.06893276423215866,
      "learning_rate": 0.00035715303215303216,
      "loss": 0.5741212158203125,
      "step": 19500
    },
    {
      "epoch": 0.3617249242209178,
      "grad_norm": 0.06911567598581314,
      "learning_rate": 0.0003546092796092796,
      "loss": 0.5721881713867187,
      "step": 19750
    },
    {
      "epoch": 0.3663037207300433,
      "grad_norm": 0.06189514696598053,
      "learning_rate": 0.00035206552706552706,
      "loss": 0.5718677978515625,
      "step": 20000
    },
    {
      "epoch": 0.3708825172391689,
      "grad_norm": 0.06711564213037491,
      "learning_rate": 0.0003495217745217745,
      "loss": 0.5711065063476563,
      "step": 20250
    },
    {
      "epoch": 0.3754613137482944,
      "grad_norm": 0.0697239488363266,
      "learning_rate": 0.00034697802197802195,
      "loss": 0.5702218017578125,
      "step": 20500
    },
    {
      "epoch": 0.38004011025741996,
      "grad_norm": 0.06532745063304901,
      "learning_rate": 0.0003444342694342694,
      "loss": 0.57056787109375,
      "step": 20750
    },
    {
      "epoch": 0.3846189067665455,
      "grad_norm": 0.06752527505159378,
      "learning_rate": 0.0003418905168905169,
      "loss": 0.57018017578125,
      "step": 21000
    },
    {
      "epoch": 0.38919770327567105,
      "grad_norm": 0.07022079825401306,
      "learning_rate": 0.00033934676434676435,
      "loss": 0.5674702758789063,
      "step": 21250
    },
    {
      "epoch": 0.39377649978479656,
      "grad_norm": 0.07235999405384064,
      "learning_rate": 0.0003368030118030118,
      "loss": 0.5680103149414063,
      "step": 21500
    },
    {
      "epoch": 0.3983552962939221,
      "grad_norm": 0.06506355851888657,
      "learning_rate": 0.00033425925925925924,
      "loss": 0.567474365234375,
      "step": 21750
    },
    {
      "epoch": 0.40293409280304765,
      "grad_norm": 0.06570252776145935,
      "learning_rate": 0.0003317155067155067,
      "loss": 0.5681478271484375,
      "step": 22000
    },
    {
      "epoch": 0.40751288931217317,
      "grad_norm": 0.06754681468009949,
      "learning_rate": 0.00032917175417175414,
      "loss": 0.5649468994140625,
      "step": 22250
    },
    {
      "epoch": 0.41209168582129874,
      "grad_norm": 0.06729352474212646,
      "learning_rate": 0.00032662800162800164,
      "loss": 0.5659503173828125,
      "step": 22500
    },
    {
      "epoch": 0.41667048233042425,
      "grad_norm": 0.07104545831680298,
      "learning_rate": 0.0003240842490842491,
      "loss": 0.5642489013671875,
      "step": 22750
    },
    {
      "epoch": 0.4212492788395498,
      "grad_norm": 0.06511082500219345,
      "learning_rate": 0.00032154049654049653,
      "loss": 0.5645473022460937,
      "step": 23000
    },
    {
      "epoch": 0.42582807534867534,
      "grad_norm": 0.06568155437707901,
      "learning_rate": 0.000318996743996744,
      "loss": 0.5643197021484375,
      "step": 23250
    },
    {
      "epoch": 0.4304068718578009,
      "grad_norm": 0.0653456449508667,
      "learning_rate": 0.00031645299145299143,
      "loss": 0.5639840087890625,
      "step": 23500
    },
    {
      "epoch": 0.4349856683669264,
      "grad_norm": 0.07045958191156387,
      "learning_rate": 0.0003139092389092389,
      "loss": 0.564418212890625,
      "step": 23750
    },
    {
      "epoch": 0.439564464876052,
      "grad_norm": 0.0655277818441391,
      "learning_rate": 0.0003113654863654864,
      "loss": 0.561712646484375,
      "step": 24000
    },
    {
      "epoch": 0.4441432613851775,
      "grad_norm": 0.06346478313207626,
      "learning_rate": 0.0003088217338217338,
      "loss": 0.56320263671875,
      "step": 24250
    },
    {
      "epoch": 0.4487220578943031,
      "grad_norm": 0.06733353435993195,
      "learning_rate": 0.00030627798127798127,
      "loss": 0.5613779907226563,
      "step": 24500
    },
    {
      "epoch": 0.4533008544034286,
      "grad_norm": 0.06213892623782158,
      "learning_rate": 0.0003037342287342287,
      "loss": 0.5591483154296875,
      "step": 24750
    },
    {
      "epoch": 0.45787965091255417,
      "grad_norm": 0.06473983824253082,
      "learning_rate": 0.00030119047619047616,
      "loss": 0.5607928466796875,
      "step": 25000
    },
    {
      "epoch": 0.4624584474216797,
      "grad_norm": 0.06556923687458038,
      "learning_rate": 0.0002986467236467236,
      "loss": 0.5601920166015625,
      "step": 25250
    },
    {
      "epoch": 0.4670372439308052,
      "grad_norm": 0.06275345385074615,
      "learning_rate": 0.0002961029711029711,
      "loss": 0.5587906494140625,
      "step": 25500
    },
    {
      "epoch": 0.4716160404399308,
      "grad_norm": 0.06278958916664124,
      "learning_rate": 0.00029355921855921856,
      "loss": 0.5598369140625,
      "step": 25750
    },
    {
      "epoch": 0.4761948369490563,
      "grad_norm": 0.06175178289413452,
      "learning_rate": 0.000291015466015466,
      "loss": 0.5599090576171875,
      "step": 26000
    },
    {
      "epoch": 0.48077363345818186,
      "grad_norm": 0.06806890666484833,
      "learning_rate": 0.00028847171347171346,
      "loss": 0.557162109375,
      "step": 26250
    },
    {
      "epoch": 0.4853524299673074,
      "grad_norm": 0.06017167866230011,
      "learning_rate": 0.0002859279609279609,
      "loss": 0.5581229248046875,
      "step": 26500
    },
    {
      "epoch": 0.48993122647643295,
      "grad_norm": 0.06641512364149094,
      "learning_rate": 0.00028338420838420835,
      "loss": 0.5567603149414062,
      "step": 26750
    },
    {
      "epoch": 0.49451002298555846,
      "grad_norm": 0.06273702532052994,
      "learning_rate": 0.00028084045584045585,
      "loss": 0.55789404296875,
      "step": 27000
    },
    {
      "epoch": 0.49908881949468403,
      "grad_norm": 0.06796413660049438,
      "learning_rate": 0.0002782967032967033,
      "loss": 0.5564705200195312,
      "step": 27250
    },
    {
      "epoch": 0.5036676160038096,
      "grad_norm": 0.06819527596235275,
      "learning_rate": 0.00027575295075295075,
      "loss": 0.555724853515625,
      "step": 27500
    },
    {
      "epoch": 0.5082464125129351,
      "grad_norm": 0.06641633808612823,
      "learning_rate": 0.0002732091982091982,
      "loss": 0.5561719970703125,
      "step": 27750
    },
    {
      "epoch": 0.5128252090220606,
      "grad_norm": 0.06096098572015762,
      "learning_rate": 0.00027066544566544564,
      "loss": 0.5545294799804688,
      "step": 28000
    },
    {
      "epoch": 0.5174040055311862,
      "grad_norm": 0.0632636547088623,
      "learning_rate": 0.0002681216931216931,
      "loss": 0.55314794921875,
      "step": 28250
    },
    {
      "epoch": 0.5219828020403118,
      "grad_norm": 0.06154729053378105,
      "learning_rate": 0.0002655779405779406,
      "loss": 0.553865966796875,
      "step": 28500
    },
    {
      "epoch": 0.5265615985494373,
      "grad_norm": 0.06788334995508194,
      "learning_rate": 0.00026303418803418804,
      "loss": 0.5544434814453125,
      "step": 28750
    },
    {
      "epoch": 0.5311403950585628,
      "grad_norm": 0.06633545458316803,
      "learning_rate": 0.0002604904354904355,
      "loss": 0.55305615234375,
      "step": 29000
    },
    {
      "epoch": 0.5357191915676883,
      "grad_norm": 0.061695124953985214,
      "learning_rate": 0.00025794668294668293,
      "loss": 0.5519857788085938,
      "step": 29250
    },
    {
      "epoch": 0.5402979880768138,
      "grad_norm": 0.06452561169862747,
      "learning_rate": 0.0002554029304029304,
      "loss": 0.5526995239257813,
      "step": 29500
    },
    {
      "epoch": 0.5448767845859395,
      "grad_norm": 0.06754140555858612,
      "learning_rate": 0.0002528591778591779,
      "loss": 0.552205078125,
      "step": 29750
    },
    {
      "epoch": 0.549455581095065,
      "grad_norm": 0.06571507453918457,
      "learning_rate": 0.0002503154253154253,
      "loss": 0.55148388671875,
      "step": 30000
    },
    {
      "epoch": 0.5540343776041905,
      "grad_norm": 0.05953441932797432,
      "learning_rate": 0.0002477716727716728,
      "loss": 0.5519866333007812,
      "step": 30250
    },
    {
      "epoch": 0.558613174113316,
      "grad_norm": 0.06737753748893738,
      "learning_rate": 0.0002452279202279202,
      "loss": 0.5516741333007813,
      "step": 30500
    },
    {
      "epoch": 0.5631919706224416,
      "grad_norm": 0.06233954802155495,
      "learning_rate": 0.00024268416768416767,
      "loss": 0.5500148315429687,
      "step": 30750
    },
    {
      "epoch": 0.5677707671315672,
      "grad_norm": 0.05968218669295311,
      "learning_rate": 0.00024014041514041514,
      "loss": 0.551145751953125,
      "step": 31000
    },
    {
      "epoch": 0.5723495636406927,
      "grad_norm": 0.06363086402416229,
      "learning_rate": 0.0002375966625966626,
      "loss": 0.5489215087890625,
      "step": 31250
    },
    {
      "epoch": 0.5769283601498182,
      "grad_norm": 0.06691507250070572,
      "learning_rate": 0.00023505291005291004,
      "loss": 0.54915771484375,
      "step": 31500
    },
    {
      "epoch": 0.5815071566589438,
      "grad_norm": 0.06056627258658409,
      "learning_rate": 0.0002325091575091575,
      "loss": 0.5485839233398437,
      "step": 31750
    },
    {
      "epoch": 0.5860859531680693,
      "grad_norm": 0.06835993379354477,
      "learning_rate": 0.00022996540496540496,
      "loss": 0.5473211669921875,
      "step": 32000
    },
    {
      "epoch": 0.5906647496771948,
      "grad_norm": 0.058730434626340866,
      "learning_rate": 0.0002274216524216524,
      "loss": 0.545982666015625,
      "step": 32250
    },
    {
      "epoch": 0.5952435461863204,
      "grad_norm": 0.06264112889766693,
      "learning_rate": 0.00022487789987789988,
      "loss": 0.547709716796875,
      "step": 32500
    },
    {
      "epoch": 0.5998223426954459,
      "grad_norm": 0.06274031847715378,
      "learning_rate": 0.00022233414733414733,
      "loss": 0.54755126953125,
      "step": 32750
    },
    {
      "epoch": 0.6044011392045715,
      "grad_norm": 0.06258225440979004,
      "learning_rate": 0.00021979039479039477,
      "loss": 0.5456241455078125,
      "step": 33000
    },
    {
      "epoch": 0.608979935713697,
      "grad_norm": 0.0640139952301979,
      "learning_rate": 0.00021724664224664225,
      "loss": 0.5486814575195312,
      "step": 33250
    },
    {
      "epoch": 0.6135587322228225,
      "grad_norm": 0.07124609500169754,
      "learning_rate": 0.0002147028897028897,
      "loss": 0.544838623046875,
      "step": 33500
    },
    {
      "epoch": 0.618137528731948,
      "grad_norm": 0.0628821849822998,
      "learning_rate": 0.00021215913715913714,
      "loss": 0.5451561889648437,
      "step": 33750
    },
    {
      "epoch": 0.6227163252410737,
      "grad_norm": 0.059289246797561646,
      "learning_rate": 0.00020961538461538462,
      "loss": 0.5440616455078126,
      "step": 34000
    },
    {
      "epoch": 0.6272951217501992,
      "grad_norm": 0.063616544008255,
      "learning_rate": 0.00020707163207163206,
      "loss": 0.5438365478515625,
      "step": 34250
    },
    {
      "epoch": 0.6318739182593247,
      "grad_norm": 0.06502054631710052,
      "learning_rate": 0.00020452787952787954,
      "loss": 0.54453271484375,
      "step": 34500
    },
    {
      "epoch": 0.6364527147684502,
      "grad_norm": 0.06169392541050911,
      "learning_rate": 0.00020198412698412699,
      "loss": 0.5431817016601562,
      "step": 34750
    },
    {
      "epoch": 0.6410315112775758,
      "grad_norm": 0.061637215316295624,
      "learning_rate": 0.00019944037444037443,
      "loss": 0.5426368408203125,
      "step": 35000
    },
    {
      "epoch": 0.6456103077867014,
      "grad_norm": 0.06308504194021225,
      "learning_rate": 0.0001968966218966219,
      "loss": 0.5428656005859375,
      "step": 35250
    },
    {
      "epoch": 0.6501891042958269,
      "grad_norm": 0.06042623519897461,
      "learning_rate": 0.00019435286935286935,
      "loss": 0.5427608642578124,
      "step": 35500
    },
    {
      "epoch": 0.6547679008049524,
      "grad_norm": 0.06185432896018028,
      "learning_rate": 0.0001918091168091168,
      "loss": 0.542866943359375,
      "step": 35750
    },
    {
      "epoch": 0.659346697314078,
      "grad_norm": 0.06367085129022598,
      "learning_rate": 0.00018926536426536428,
      "loss": 0.5427926025390625,
      "step": 36000
    },
    {
      "epoch": 0.6639254938232035,
      "grad_norm": 0.0623827800154686,
      "learning_rate": 0.00018672161172161172,
      "loss": 0.5416766357421875,
      "step": 36250
    },
    {
      "epoch": 0.668504290332329,
      "grad_norm": 0.06463074684143066,
      "learning_rate": 0.00018417785917785917,
      "loss": 0.5395673828125,
      "step": 36500
    },
    {
      "epoch": 0.6730830868414546,
      "grad_norm": 0.06329692900180817,
      "learning_rate": 0.00018163410663410664,
      "loss": 0.5405562744140625,
      "step": 36750
    },
    {
      "epoch": 0.6776618833505801,
      "grad_norm": 0.0642353892326355,
      "learning_rate": 0.0001790903540903541,
      "loss": 0.5405491943359375,
      "step": 37000
    },
    {
      "epoch": 0.6822406798597057,
      "grad_norm": 0.07013506442308426,
      "learning_rate": 0.00017654660154660154,
      "loss": 0.5401270141601563,
      "step": 37250
    },
    {
      "epoch": 0.6868194763688312,
      "grad_norm": 0.06612392514944077,
      "learning_rate": 0.000174002849002849,
      "loss": 0.5385986328125,
      "step": 37500
    },
    {
      "epoch": 0.6913982728779567,
      "grad_norm": 0.06670144200325012,
      "learning_rate": 0.00017145909645909646,
      "loss": 0.5392545166015625,
      "step": 37750
    },
    {
      "epoch": 0.6959770693870823,
      "grad_norm": 0.05942818522453308,
      "learning_rate": 0.0001689153439153439,
      "loss": 0.536975341796875,
      "step": 38000
    },
    {
      "epoch": 0.7005558658962079,
      "grad_norm": 0.06282128393650055,
      "learning_rate": 0.00016637159137159138,
      "loss": 0.538390625,
      "step": 38250
    },
    {
      "epoch": 0.7051346624053334,
      "grad_norm": 0.06333746016025543,
      "learning_rate": 0.00016382783882783883,
      "loss": 0.541278564453125,
      "step": 38500
    },
    {
      "epoch": 0.7097134589144589,
      "grad_norm": 0.06994464248418808,
      "learning_rate": 0.00016128408628408628,
      "loss": 0.5388087158203125,
      "step": 38750
    },
    {
      "epoch": 0.7142922554235844,
      "grad_norm": 0.06431885808706284,
      "learning_rate": 0.00015874033374033375,
      "loss": 0.5385784912109375,
      "step": 39000
    },
    {
      "epoch": 0.7188710519327101,
      "grad_norm": 0.05879560485482216,
      "learning_rate": 0.0001561965811965812,
      "loss": 0.5362100219726562,
      "step": 39250
    },
    {
      "epoch": 0.7234498484418356,
      "grad_norm": 0.06899562478065491,
      "learning_rate": 0.00015365282865282865,
      "loss": 0.536516845703125,
      "step": 39500
    },
    {
      "epoch": 0.7280286449509611,
      "grad_norm": 0.05971398949623108,
      "learning_rate": 0.00015110907610907612,
      "loss": 0.53708984375,
      "step": 39750
    },
    {
      "epoch": 0.7326074414600866,
      "grad_norm": 0.064848393201828,
      "learning_rate": 0.00014856532356532357,
      "loss": 0.5373258056640625,
      "step": 40000
    },
    {
      "epoch": 0.7371862379692121,
      "grad_norm": 0.0600309744477272,
      "learning_rate": 0.00014602157102157101,
      "loss": 0.53594677734375,
      "step": 40250
    },
    {
      "epoch": 0.7417650344783377,
      "grad_norm": 0.06224419176578522,
      "learning_rate": 0.0001434778184778185,
      "loss": 0.5362578125,
      "step": 40500
    },
    {
      "epoch": 0.7463438309874633,
      "grad_norm": 0.05738287419080734,
      "learning_rate": 0.00014093406593406594,
      "loss": 0.5352327880859375,
      "step": 40750
    },
    {
      "epoch": 0.7509226274965888,
      "grad_norm": 0.06629467755556107,
      "learning_rate": 0.00013839031339031338,
      "loss": 0.5343209838867188,
      "step": 41000
    },
    {
      "epoch": 0.7555014240057143,
      "grad_norm": 0.0607992485165596,
      "learning_rate": 0.00013584656084656086,
      "loss": 0.536358154296875,
      "step": 41250
    },
    {
      "epoch": 0.7600802205148399,
      "grad_norm": 0.05663478001952171,
      "learning_rate": 0.0001333028083028083,
      "loss": 0.5357347412109374,
      "step": 41500
    },
    {
      "epoch": 0.7646590170239654,
      "grad_norm": 0.062044378370046616,
      "learning_rate": 0.00013075905575905575,
      "loss": 0.5350208740234375,
      "step": 41750
    },
    {
      "epoch": 0.769237813533091,
      "grad_norm": 0.06680475920438766,
      "learning_rate": 0.00012821530321530323,
      "loss": 0.535318115234375,
      "step": 42000
    },
    {
      "epoch": 0.7738166100422165,
      "grad_norm": 0.0646413043141365,
      "learning_rate": 0.00012567155067155067,
      "loss": 0.5335006103515625,
      "step": 42250
    },
    {
      "epoch": 0.7783954065513421,
      "grad_norm": 0.06554868817329407,
      "learning_rate": 0.00012312779812779812,
      "loss": 0.53502490234375,
      "step": 42500
    },
    {
      "epoch": 0.7829742030604676,
      "grad_norm": 0.06611758470535278,
      "learning_rate": 0.0001205840455840456,
      "loss": 0.5335166015625,
      "step": 42750
    },
    {
      "epoch": 0.7875529995695931,
      "grad_norm": 0.06753136217594147,
      "learning_rate": 0.00011804029304029304,
      "loss": 0.533784912109375,
      "step": 43000
    },
    {
      "epoch": 0.7921317960787186,
      "grad_norm": 0.0629805251955986,
      "learning_rate": 0.0001154965404965405,
      "loss": 0.5315120239257812,
      "step": 43250
    },
    {
      "epoch": 0.7967105925878442,
      "grad_norm": 0.062131430953741074,
      "learning_rate": 0.00011295278795278796,
      "loss": 0.531296142578125,
      "step": 43500
    },
    {
      "epoch": 0.8012893890969698,
      "grad_norm": 0.06105881184339523,
      "learning_rate": 0.00011040903540903541,
      "loss": 0.5311162109375,
      "step": 43750
    },
    {
      "epoch": 0.8058681856060953,
      "grad_norm": 0.06286618113517761,
      "learning_rate": 0.00010786528286528287,
      "loss": 0.5331693115234375,
      "step": 44000
    },
    {
      "epoch": 0.8104469821152208,
      "grad_norm": 0.06307170540094376,
      "learning_rate": 0.00010532153032153033,
      "loss": 0.5322763061523438,
      "step": 44250
    },
    {
      "epoch": 0.8150257786243463,
      "grad_norm": 0.059519022703170776,
      "learning_rate": 0.00010277777777777778,
      "loss": 0.5308363037109375,
      "step": 44500
    },
    {
      "epoch": 0.819604575133472,
      "grad_norm": 0.06570710241794586,
      "learning_rate": 0.00010023402523402524,
      "loss": 0.5305360107421875,
      "step": 44750
    },
    {
      "epoch": 0.8241833716425975,
      "grad_norm": 0.06340634822845459,
      "learning_rate": 9.76902726902727e-05,
      "loss": 0.5312403564453125,
      "step": 45000
    },
    {
      "epoch": 0.828762168151723,
      "grad_norm": 0.06262257695198059,
      "learning_rate": 9.514652014652015e-05,
      "loss": 0.5294259033203125,
      "step": 45250
    },
    {
      "epoch": 0.8333409646608485,
      "grad_norm": 0.059705134481191635,
      "learning_rate": 9.260276760276761e-05,
      "loss": 0.52991943359375,
      "step": 45500
    },
    {
      "epoch": 0.8379197611699741,
      "grad_norm": 0.06161255016922951,
      "learning_rate": 9.005901505901507e-05,
      "loss": 0.5287122802734375,
      "step": 45750
    },
    {
      "epoch": 0.8424985576790996,
      "grad_norm": 0.06219637393951416,
      "learning_rate": 8.751526251526252e-05,
      "loss": 0.5296033935546876,
      "step": 46000
    },
    {
      "epoch": 0.8470773541882252,
      "grad_norm": 0.060764264315366745,
      "learning_rate": 8.497150997150998e-05,
      "loss": 0.5302567749023438,
      "step": 46250
    },
    {
      "epoch": 0.8516561506973507,
      "grad_norm": 0.06355715543031693,
      "learning_rate": 8.242775742775744e-05,
      "loss": 0.5289213256835937,
      "step": 46500
    },
    {
      "epoch": 0.8562349472064763,
      "grad_norm": 0.061392754316329956,
      "learning_rate": 7.988400488400489e-05,
      "loss": 0.5266444091796875,
      "step": 46750
    },
    {
      "epoch": 0.8608137437156018,
      "grad_norm": 0.06268171966075897,
      "learning_rate": 7.734025234025235e-05,
      "loss": 0.5273516845703125,
      "step": 47000
    },
    {
      "epoch": 0.8653925402247273,
      "grad_norm": 0.06518919765949249,
      "learning_rate": 7.479649979649981e-05,
      "loss": 0.5287893676757812,
      "step": 47250
    },
    {
      "epoch": 0.8699713367338529,
      "grad_norm": 0.06212367117404938,
      "learning_rate": 7.225274725274724e-05,
      "loss": 0.5263849487304687,
      "step": 47500
    },
    {
      "epoch": 0.8745501332429784,
      "grad_norm": 0.0620686411857605,
      "learning_rate": 6.97089947089947e-05,
      "loss": 0.526240478515625,
      "step": 47750
    },
    {
      "epoch": 0.879128929752104,
      "grad_norm": 0.062115322798490524,
      "learning_rate": 6.716524216524216e-05,
      "loss": 0.52626611328125,
      "step": 48000
    },
    {
      "epoch": 0.8837077262612295,
      "grad_norm": 0.06891973316669464,
      "learning_rate": 6.462148962148962e-05,
      "loss": 0.5267440185546876,
      "step": 48250
    },
    {
      "epoch": 0.888286522770355,
      "grad_norm": 0.06106738746166229,
      "learning_rate": 6.207773707773707e-05,
      "loss": 0.5261217041015624,
      "step": 48500
    },
    {
      "epoch": 0.8928653192794805,
      "grad_norm": 0.06405936181545258,
      "learning_rate": 5.953398453398454e-05,
      "loss": 0.525679931640625,
      "step": 48750
    },
    {
      "epoch": 0.8974441157886062,
      "grad_norm": 0.062343280762434006,
      "learning_rate": 5.699023199023199e-05,
      "loss": 0.5243419799804687,
      "step": 49000
    },
    {
      "epoch": 0.9020229122977317,
      "grad_norm": 0.062424950301647186,
      "learning_rate": 5.444647944647945e-05,
      "loss": 0.5248175187544389,
      "step": 49250
    },
    {
      "epoch": 0.9066017088068572,
      "grad_norm": 0.060387108474969864,
      "learning_rate": 5.1902726902726907e-05,
      "loss": 0.5265853271484375,
      "step": 49500
    },
    {
      "epoch": 0.9111805053159827,
      "grad_norm": 0.06357312202453613,
      "learning_rate": 4.935897435897436e-05,
      "loss": 0.5228431396484375,
      "step": 49750
    },
    {
      "epoch": 0.9157593018251083,
      "grad_norm": 0.06350237876176834,
      "learning_rate": 4.681522181522182e-05,
      "loss": 0.5256588134765625,
      "step": 50000
    },
    {
      "epoch": 0.9203380983342339,
      "grad_norm": 0.06372404843568802,
      "learning_rate": 4.4271469271469275e-05,
      "loss": 0.5237154541015625,
      "step": 50250
    },
    {
      "epoch": 0.9249168948433594,
      "grad_norm": 0.06382907927036285,
      "learning_rate": 4.172771672771673e-05,
      "loss": 0.5238780517578125,
      "step": 50500
    },
    {
      "epoch": 0.9294956913524849,
      "grad_norm": 0.06671983003616333,
      "learning_rate": 3.918396418396418e-05,
      "loss": 0.5238001708984374,
      "step": 50750
    },
    {
      "epoch": 0.9340744878616104,
      "grad_norm": 0.06559401750564575,
      "learning_rate": 3.664021164021164e-05,
      "loss": 0.52430419921875,
      "step": 51000
    },
    {
      "epoch": 0.938653284370736,
      "grad_norm": 0.06468282639980316,
      "learning_rate": 3.409645909645909e-05,
      "loss": 0.523361083984375,
      "step": 51250
    },
    {
      "epoch": 0.9432320808798615,
      "grad_norm": 0.06369815021753311,
      "learning_rate": 3.155270655270655e-05,
      "loss": 0.5240621337890625,
      "step": 51500
    },
    {
      "epoch": 0.9478108773889871,
      "grad_norm": 0.062487825751304626,
      "learning_rate": 2.900895400895401e-05,
      "loss": 0.5241049194335937,
      "step": 51750
    },
    {
      "epoch": 0.9523896738981126,
      "grad_norm": 0.06275194883346558,
      "learning_rate": 2.6465201465201467e-05,
      "loss": 0.5225496215820312,
      "step": 52000
    },
    {
      "epoch": 0.9569684704072382,
      "grad_norm": 0.06413143128156662,
      "learning_rate": 2.3921448921448924e-05,
      "loss": 0.5205796508789062,
      "step": 52250
    },
    {
      "epoch": 0.9615472669163637,
      "grad_norm": 0.06309211254119873,
      "learning_rate": 2.1377696377696378e-05,
      "loss": 0.5227193603515625,
      "step": 52500
    },
    {
      "epoch": 0.9661260634254892,
      "grad_norm": 0.06116398051381111,
      "learning_rate": 1.8833943833943832e-05,
      "loss": 0.52264306640625,
      "step": 52750
    },
    {
      "epoch": 0.9707048599346147,
      "grad_norm": 0.06321769952774048,
      "learning_rate": 1.629019129019129e-05,
      "loss": 0.5229866943359375,
      "step": 53000
    },
    {
      "epoch": 0.9752836564437404,
      "grad_norm": 0.06261686980724335,
      "learning_rate": 1.3746438746438745e-05,
      "loss": 0.5215400390625,
      "step": 53250
    },
    {
      "epoch": 0.9798624529528659,
      "grad_norm": 0.061849433928728104,
      "learning_rate": 1.1202686202686203e-05,
      "loss": 0.5211468505859375,
      "step": 53500
    },
    {
      "epoch": 0.9844412494619914,
      "grad_norm": 0.06358622014522552,
      "learning_rate": 8.658933658933658e-06,
      "loss": 0.5221812744140625,
      "step": 53750
    },
    {
      "epoch": 0.9890200459711169,
      "grad_norm": 0.06303670257329941,
      "learning_rate": 6.115181115181116e-06,
      "loss": 0.521659912109375,
      "step": 54000
    },
    {
      "epoch": 0.9935988424802424,
      "grad_norm": 0.06596051901578903,
      "learning_rate": 3.5714285714285714e-06,
      "loss": 0.5219719848632812,
      "step": 54250
    },
    {
      "epoch": 0.9981776389893681,
      "grad_norm": 0.06199328973889351,
      "learning_rate": 1.0276760276760277e-06,
      "loss": 0.51972021484375,
      "step": 54500
    },
    {
      "epoch": 1.0,
      "step": 54600,
      "total_flos": 9.748592669944185e+18,
      "train_loss": 0.052300934774098376,
      "train_runtime": 7208.7947,
      "train_samples_per_second": 484.736,
      "train_steps_per_second": 7.574
    }
  ],
  "logging_steps": 250,
  "max_steps": 54600,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 5460,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 9.748592669944185e+18,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}