{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 209,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.004784688995215311,
      "grad_norm": 0.14907115697860718,
      "learning_rate": 2e-05,
      "loss": 1.2491,
      "step": 1
    },
    {
      "epoch": 0.009569377990430622,
      "grad_norm": 0.17295576632022858,
      "learning_rate": 4e-05,
      "loss": 1.4909,
      "step": 2
    },
    {
      "epoch": 0.014354066985645933,
      "grad_norm": 0.18045826256275177,
      "learning_rate": 6e-05,
      "loss": 1.5363,
      "step": 3
    },
    {
      "epoch": 0.019138755980861243,
      "grad_norm": 0.19427408277988434,
      "learning_rate": 8e-05,
      "loss": 1.6397,
      "step": 4
    },
    {
      "epoch": 0.023923444976076555,
      "grad_norm": 0.2216596156358719,
      "learning_rate": 0.0001,
      "loss": 1.7394,
      "step": 5
    },
    {
      "epoch": 0.028708133971291867,
      "grad_norm": 0.20222336053848267,
      "learning_rate": 9.999407114490384e-05,
      "loss": 1.6014,
      "step": 6
    },
    {
      "epoch": 0.03349282296650718,
      "grad_norm": 0.21694432199001312,
      "learning_rate": 9.99762859856683e-05,
      "loss": 1.5539,
      "step": 7
    },
    {
      "epoch": 0.03827751196172249,
      "grad_norm": 0.21403349936008453,
      "learning_rate": 9.994664874011863e-05,
      "loss": 1.4513,
      "step": 8
    },
    {
      "epoch": 0.0430622009569378,
      "grad_norm": 0.22947654128074646,
      "learning_rate": 9.990516643685222e-05,
      "loss": 1.4614,
      "step": 9
    },
    {
      "epoch": 0.04784688995215311,
      "grad_norm": 0.23931479454040527,
      "learning_rate": 9.985184891357164e-05,
      "loss": 1.5114,
      "step": 10
    },
    {
      "epoch": 0.05263157894736842,
      "grad_norm": 0.23815925419330597,
      "learning_rate": 9.978670881475172e-05,
      "loss": 1.5015,
      "step": 11
    },
    {
      "epoch": 0.05741626794258373,
      "grad_norm": 0.25154542922973633,
      "learning_rate": 9.970976158864073e-05,
      "loss": 1.5904,
      "step": 12
    },
    {
      "epoch": 0.06220095693779904,
      "grad_norm": 0.2565719783306122,
      "learning_rate": 9.96210254835968e-05,
      "loss": 1.4701,
      "step": 13
    },
    {
      "epoch": 0.06698564593301436,
      "grad_norm": 0.2682940363883972,
      "learning_rate": 9.952052154376026e-05,
      "loss": 1.5519,
      "step": 14
    },
    {
      "epoch": 0.07177033492822966,
      "grad_norm": 0.26406288146972656,
      "learning_rate": 9.940827360406297e-05,
      "loss": 1.405,
      "step": 15
    },
    {
      "epoch": 0.07655502392344497,
      "grad_norm": 0.29123812913894653,
      "learning_rate": 9.928430828457572e-05,
      "loss": 1.3998,
      "step": 16
    },
    {
      "epoch": 0.08133971291866028,
      "grad_norm": 0.3094870448112488,
      "learning_rate": 9.91486549841951e-05,
      "loss": 1.6672,
      "step": 17
    },
    {
      "epoch": 0.0861244019138756,
      "grad_norm": 0.2948465049266815,
      "learning_rate": 9.90013458736716e-05,
      "loss": 1.5797,
      "step": 18
    },
    {
      "epoch": 0.09090909090909091,
      "grad_norm": 0.29864171147346497,
      "learning_rate": 9.884241588798005e-05,
      "loss": 1.4558,
      "step": 19
    },
    {
      "epoch": 0.09569377990430622,
      "grad_norm": 0.31656384468078613,
      "learning_rate": 9.867190271803465e-05,
      "loss": 1.4255,
      "step": 20
    },
    {
      "epoch": 0.10047846889952153,
      "grad_norm": 0.33296871185302734,
      "learning_rate": 9.848984680175049e-05,
      "loss": 1.5504,
      "step": 21
    },
    {
      "epoch": 0.10526315789473684,
      "grad_norm": 0.3384315073490143,
      "learning_rate": 9.829629131445342e-05,
      "loss": 1.5032,
      "step": 22
    },
    {
      "epoch": 0.11004784688995216,
      "grad_norm": 0.3654746413230896,
      "learning_rate": 9.809128215864097e-05,
      "loss": 1.7055,
      "step": 23
    },
    {
      "epoch": 0.11483253588516747,
      "grad_norm": 0.3699825406074524,
      "learning_rate": 9.787486795309621e-05,
      "loss": 1.7864,
      "step": 24
    },
    {
      "epoch": 0.11961722488038277,
      "grad_norm": 0.38653433322906494,
      "learning_rate": 9.764710002135784e-05,
      "loss": 1.5302,
      "step": 25
    },
    {
      "epoch": 0.12440191387559808,
      "grad_norm": 0.39083021879196167,
      "learning_rate": 9.74080323795483e-05,
      "loss": 1.687,
      "step": 26
    },
    {
      "epoch": 0.1291866028708134,
      "grad_norm": 0.4041406810283661,
      "learning_rate": 9.715772172356388e-05,
      "loss": 1.7685,
      "step": 27
    },
    {
      "epoch": 0.1339712918660287,
      "grad_norm": 0.43206948041915894,
      "learning_rate": 9.689622741562892e-05,
      "loss": 1.6886,
      "step": 28
    },
    {
      "epoch": 0.13875598086124402,
      "grad_norm": 0.4564664959907532,
      "learning_rate": 9.662361147021779e-05,
      "loss": 1.7957,
      "step": 29
    },
    {
      "epoch": 0.14354066985645933,
      "grad_norm": 0.4699450135231018,
      "learning_rate": 9.633993853934803e-05,
      "loss": 2.04,
      "step": 30
    },
    {
      "epoch": 0.14832535885167464,
      "grad_norm": 0.4722703695297241,
      "learning_rate": 9.60452758972477e-05,
      "loss": 1.9598,
      "step": 31
    },
    {
      "epoch": 0.15311004784688995,
      "grad_norm": 0.4768441319465637,
      "learning_rate": 9.573969342440106e-05,
      "loss": 1.7342,
      "step": 32
    },
    {
      "epoch": 0.15789473684210525,
      "grad_norm": 0.5047369599342346,
      "learning_rate": 9.542326359097619e-05,
      "loss": 1.7313,
      "step": 33
    },
    {
      "epoch": 0.16267942583732056,
      "grad_norm": 0.49944713711738586,
      "learning_rate": 9.509606143963832e-05,
      "loss": 1.8511,
      "step": 34
    },
    {
      "epoch": 0.1674641148325359,
      "grad_norm": 0.5391089916229248,
      "learning_rate": 9.475816456775313e-05,
      "loss": 1.9546,
      "step": 35
    },
    {
      "epoch": 0.1722488038277512,
      "grad_norm": 0.584690272808075,
      "learning_rate": 9.440965310898424e-05,
      "loss": 1.9743,
      "step": 36
    },
    {
      "epoch": 0.17703349282296652,
      "grad_norm": 0.6214085817337036,
      "learning_rate": 9.405060971428923e-05,
      "loss": 1.9413,
      "step": 37
    },
    {
      "epoch": 0.18181818181818182,
      "grad_norm": 0.5669549703598022,
      "learning_rate": 9.368111953231848e-05,
      "loss": 1.9106,
      "step": 38
    },
    {
      "epoch": 0.18660287081339713,
      "grad_norm": 0.6074210405349731,
      "learning_rate": 9.330127018922194e-05,
      "loss": 1.9595,
      "step": 39
    },
    {
      "epoch": 0.19138755980861244,
      "grad_norm": 0.6308673620223999,
      "learning_rate": 9.291115176786814e-05,
      "loss": 2.0112,
      "step": 40
    },
    {
      "epoch": 0.19617224880382775,
      "grad_norm": 0.6849383115768433,
      "learning_rate": 9.251085678648072e-05,
      "loss": 1.9215,
      "step": 41
    },
    {
      "epoch": 0.20095693779904306,
      "grad_norm": 0.8266085386276245,
      "learning_rate": 9.210048017669726e-05,
      "loss": 2.0382,
      "step": 42
    },
    {
      "epoch": 0.20574162679425836,
      "grad_norm": 0.7204662561416626,
      "learning_rate": 9.168011926105598e-05,
      "loss": 1.9057,
      "step": 43
    },
    {
      "epoch": 0.21052631578947367,
      "grad_norm": 0.8284960389137268,
      "learning_rate": 9.124987372991511e-05,
      "loss": 1.9009,
      "step": 44
    },
    {
      "epoch": 0.215311004784689,
      "grad_norm": 0.8978604674339294,
      "learning_rate": 9.08098456178111e-05,
      "loss": 2.0893,
      "step": 45
    },
    {
      "epoch": 0.22009569377990432,
      "grad_norm": 1.1396929025650024,
      "learning_rate": 9.036013927926048e-05,
      "loss": 2.2963,
      "step": 46
    },
    {
      "epoch": 0.22488038277511962,
      "grad_norm": 1.2800486087799072,
      "learning_rate": 8.9900861364012e-05,
      "loss": 2.0588,
      "step": 47
    },
    {
      "epoch": 0.22966507177033493,
      "grad_norm": 1.6371203660964966,
      "learning_rate": 8.943212079175391e-05,
      "loss": 1.8706,
      "step": 48
    },
    {
      "epoch": 0.23444976076555024,
      "grad_norm": 2.216677188873291,
      "learning_rate": 8.895402872628352e-05,
      "loss": 2.0251,
      "step": 49
    },
    {
      "epoch": 0.23923444976076555,
      "grad_norm": 3.7424046993255615,
      "learning_rate": 8.846669854914396e-05,
      "loss": 2.6344,
      "step": 50
    },
    {
      "epoch": 0.24401913875598086,
      "grad_norm": 0.3241454064846039,
      "learning_rate": 8.797024583273537e-05,
      "loss": 1.562,
      "step": 51
    },
    {
      "epoch": 0.24880382775119617,
      "grad_norm": 0.36598026752471924,
      "learning_rate": 8.746478831290648e-05,
      "loss": 1.6533,
      "step": 52
    },
    {
      "epoch": 0.2535885167464115,
      "grad_norm": 0.4164574444293976,
      "learning_rate": 8.695044586103296e-05,
      "loss": 1.5149,
      "step": 53
    },
    {
      "epoch": 0.2583732057416268,
      "grad_norm": 0.4113615155220032,
      "learning_rate": 8.642734045558952e-05,
      "loss": 1.5673,
      "step": 54
    },
    {
      "epoch": 0.2631578947368421,
      "grad_norm": 0.39313098788261414,
      "learning_rate": 8.58955961532221e-05,
      "loss": 1.6382,
      "step": 55
    },
    {
      "epoch": 0.2679425837320574,
      "grad_norm": 0.3647348880767822,
      "learning_rate": 8.535533905932738e-05,
      "loss": 1.6045,
      "step": 56
    },
    {
      "epoch": 0.2727272727272727,
      "grad_norm": 0.35326749086380005,
      "learning_rate": 8.480669729814635e-05,
      "loss": 1.3358,
      "step": 57
    },
    {
      "epoch": 0.27751196172248804,
      "grad_norm": 0.3498622477054596,
      "learning_rate": 8.424980098237903e-05,
      "loss": 1.5901,
      "step": 58
    },
    {
      "epoch": 0.2822966507177033,
      "grad_norm": 0.34479665756225586,
      "learning_rate": 8.368478218232787e-05,
      "loss": 1.5684,
      "step": 59
    },
    {
      "epoch": 0.28708133971291866,
      "grad_norm": 0.30399441719055176,
      "learning_rate": 8.311177489457652e-05,
      "loss": 1.6434,
      "step": 60
    },
    {
      "epoch": 0.291866028708134,
      "grad_norm": 0.3355162739753723,
      "learning_rate": 8.25309150102121e-05,
      "loss": 1.6636,
      "step": 61
    },
    {
      "epoch": 0.2966507177033493,
      "grad_norm": 0.2852919101715088,
      "learning_rate": 8.194234028259806e-05,
      "loss": 1.4544,
      "step": 62
    },
    {
      "epoch": 0.3014354066985646,
      "grad_norm": 0.3010583221912384,
      "learning_rate": 8.134619029470534e-05,
      "loss": 1.5694,
      "step": 63
    },
    {
      "epoch": 0.3062200956937799,
      "grad_norm": 0.3184051811695099,
      "learning_rate": 8.074260642600964e-05,
      "loss": 1.4966,
      "step": 64
    },
    {
      "epoch": 0.31100478468899523,
      "grad_norm": 0.30305036902427673,
      "learning_rate": 8.013173181896283e-05,
      "loss": 1.5011,
      "step": 65
    },
    {
      "epoch": 0.3157894736842105,
      "grad_norm": 0.307675302028656,
      "learning_rate": 7.951371134504599e-05,
      "loss": 1.3939,
      "step": 66
    },
    {
      "epoch": 0.32057416267942584,
      "grad_norm": 0.3339138627052307,
      "learning_rate": 7.888869157041257e-05,
      "loss": 1.6544,
      "step": 67
    },
    {
      "epoch": 0.3253588516746411,
      "grad_norm": 0.3280654549598694,
      "learning_rate": 7.82568207211296e-05,
      "loss": 1.4717,
      "step": 68
    },
    {
      "epoch": 0.33014354066985646,
      "grad_norm": 0.3183118999004364,
      "learning_rate": 7.76182486480253e-05,
      "loss": 1.497,
      "step": 69
    },
    {
      "epoch": 0.3349282296650718,
      "grad_norm": 0.31861793994903564,
      "learning_rate": 7.697312679115125e-05,
      "loss": 1.5132,
      "step": 70
    },
    {
      "epoch": 0.3397129186602871,
      "grad_norm": 0.346038818359375,
      "learning_rate": 7.63216081438678e-05,
      "loss": 1.4482,
      "step": 71
    },
    {
      "epoch": 0.3444976076555024,
      "grad_norm": 0.3373595178127289,
      "learning_rate": 7.566384721656104e-05,
      "loss": 1.5695,
      "step": 72
    },
    {
      "epoch": 0.3492822966507177,
      "grad_norm": 0.35197165608406067,
      "learning_rate": 7.500000000000001e-05,
      "loss": 1.5399,
      "step": 73
    },
    {
      "epoch": 0.35406698564593303,
      "grad_norm": 0.3390144109725952,
      "learning_rate": 7.433022392834282e-05,
      "loss": 1.4598,
      "step": 74
    },
    {
      "epoch": 0.3588516746411483,
      "grad_norm": 0.36730819940567017,
      "learning_rate": 7.365467784180051e-05,
      "loss": 1.6108,
      "step": 75
    },
    {
      "epoch": 0.36363636363636365,
      "grad_norm": 0.37005236744880676,
      "learning_rate": 7.297352194896739e-05,
      "loss": 1.534,
      "step": 76
    },
    {
      "epoch": 0.3684210526315789,
      "grad_norm": 0.44057509303092957,
      "learning_rate": 7.228691778882693e-05,
      "loss": 1.6591,
      "step": 77
    },
    {
      "epoch": 0.37320574162679426,
      "grad_norm": 0.35913902521133423,
      "learning_rate": 7.159502819244206e-05,
      "loss": 1.5481,
      "step": 78
    },
    {
      "epoch": 0.37799043062200954,
      "grad_norm": 0.4454534947872162,
      "learning_rate": 7.089801724433917e-05,
      "loss": 1.8177,
      "step": 79
    },
    {
      "epoch": 0.3827751196172249,
      "grad_norm": 0.40797775983810425,
      "learning_rate": 7.019605024359474e-05,
      "loss": 1.7416,
      "step": 80
    },
    {
      "epoch": 0.3875598086124402,
      "grad_norm": 0.44979557394981384,
      "learning_rate": 6.948929366463396e-05,
      "loss": 1.8288,
      "step": 81
    },
    {
      "epoch": 0.3923444976076555,
      "grad_norm": 0.47842079401016235,
      "learning_rate": 6.877791511775063e-05,
      "loss": 1.6143,
      "step": 82
    },
    {
      "epoch": 0.39712918660287083,
      "grad_norm": 0.46703702211380005,
      "learning_rate": 6.806208330935766e-05,
      "loss": 1.7399,
      "step": 83
    },
    {
      "epoch": 0.4019138755980861,
      "grad_norm": 0.5180630087852478,
      "learning_rate": 6.734196800197762e-05,
      "loss": 1.8192,
      "step": 84
    },
    {
      "epoch": 0.40669856459330145,
      "grad_norm": 0.559816837310791,
      "learning_rate": 6.661773997398298e-05,
      "loss": 2.0921,
      "step": 85
    },
    {
      "epoch": 0.41148325358851673,
      "grad_norm": 0.4981471300125122,
      "learning_rate": 6.588957097909508e-05,
      "loss": 1.877,
      "step": 86
    },
    {
      "epoch": 0.41626794258373206,
      "grad_norm": 0.5296308398246765,
      "learning_rate": 6.515763370565218e-05,
      "loss": 1.8035,
      "step": 87
    },
    {
      "epoch": 0.42105263157894735,
      "grad_norm": 0.580813467502594,
      "learning_rate": 6.442210173565561e-05,
      "loss": 2.0139,
      "step": 88
    },
    {
      "epoch": 0.4258373205741627,
      "grad_norm": 0.5659003257751465,
      "learning_rate": 6.368314950360415e-05,
      "loss": 1.9686,
      "step": 89
    },
    {
      "epoch": 0.430622009569378,
      "grad_norm": 0.6458680629730225,
      "learning_rate": 6.294095225512603e-05,
      "loss": 1.7827,
      "step": 90
    },
    {
      "epoch": 0.4354066985645933,
      "grad_norm": 0.6537598967552185,
      "learning_rate": 6.219568600541886e-05,
      "loss": 1.8706,
      "step": 91
    },
    {
      "epoch": 0.44019138755980863,
      "grad_norm": 0.7065061330795288,
      "learning_rate": 6.14475274975067e-05,
      "loss": 2.0038,
      "step": 92
    },
    {
      "epoch": 0.4449760765550239,
      "grad_norm": 0.7320758700370789,
      "learning_rate": 6.069665416032487e-05,
      "loss": 2.0265,
      "step": 93
    },
    {
      "epoch": 0.44976076555023925,
      "grad_norm": 0.8753424882888794,
      "learning_rate": 5.9943244066641834e-05,
      "loss": 2.1909,
      "step": 94
    },
    {
      "epoch": 0.45454545454545453,
      "grad_norm": 0.988920271396637,
      "learning_rate": 5.918747589082853e-05,
      "loss": 2.1724,
      "step": 95
    },
    {
      "epoch": 0.45933014354066987,
      "grad_norm": 1.0871882438659668,
      "learning_rate": 5.842952886648496e-05,
      "loss": 2.1458,
      "step": 96
    },
    {
      "epoch": 0.46411483253588515,
      "grad_norm": 1.315832257270813,
      "learning_rate": 5.7669582743934284e-05,
      "loss": 2.0194,
      "step": 97
    },
    {
      "epoch": 0.4688995215311005,
      "grad_norm": 1.2786567211151123,
      "learning_rate": 5.6907817747594116e-05,
      "loss": 1.5575,
      "step": 98
    },
    {
      "epoch": 0.47368421052631576,
      "grad_norm": 2.389514923095703,
      "learning_rate": 5.614441453323571e-05,
      "loss": 2.0557,
      "step": 99
    },
    {
      "epoch": 0.4784688995215311,
      "grad_norm": 4.508440971374512,
      "learning_rate": 5.5379554145140574e-05,
      "loss": 2.1291,
      "step": 100
    },
    {
      "epoch": 0.48325358851674644,
      "grad_norm": 0.2242479771375656,
      "learning_rate": 5.4613417973165106e-05,
      "loss": 1.4759,
      "step": 101
    },
    {
      "epoch": 0.4880382775119617,
      "grad_norm": 0.24106784164905548,
      "learning_rate": 5.38461877097232e-05,
      "loss": 1.3887,
      "step": 102
    },
    {
      "epoch": 0.49282296650717705,
      "grad_norm": 0.26973956823349,
      "learning_rate": 5.307804530669716e-05,
      "loss": 1.5937,
      "step": 103
    },
    {
      "epoch": 0.49760765550239233,
      "grad_norm": 0.30369481444358826,
      "learning_rate": 5.230917293228699e-05,
      "loss": 1.723,
      "step": 104
    },
    {
      "epoch": 0.5023923444976076,
      "grad_norm": 0.3147946000099182,
      "learning_rate": 5.153975292780853e-05,
      "loss": 1.4143,
      "step": 105
    },
    {
      "epoch": 0.507177033492823,
      "grad_norm": 0.3020610213279724,
      "learning_rate": 5.0769967764450345e-05,
      "loss": 1.4558,
      "step": 106
    },
    {
      "epoch": 0.5119617224880383,
      "grad_norm": 0.33380070328712463,
      "learning_rate": 5e-05,
      "loss": 1.4098,
      "step": 107
    },
    {
      "epoch": 0.5167464114832536,
      "grad_norm": 0.2972123324871063,
      "learning_rate": 4.9230032235549667e-05,
      "loss": 1.5126,
      "step": 108
    },
    {
      "epoch": 0.5215311004784688,
      "grad_norm": 0.3250975012779236,
      "learning_rate": 4.8460247072191496e-05,
      "loss": 1.5187,
      "step": 109
    },
    {
      "epoch": 0.5263157894736842,
      "grad_norm": 0.3138869106769562,
      "learning_rate": 4.7690827067713035e-05,
      "loss": 1.4622,
      "step": 110
    },
    {
      "epoch": 0.5311004784688995,
      "grad_norm": 0.31304407119750977,
      "learning_rate": 4.692195469330286e-05,
      "loss": 1.5879,
      "step": 111
    },
    {
      "epoch": 0.5358851674641149,
      "grad_norm": 0.3148694932460785,
      "learning_rate": 4.6153812290276813e-05,
      "loss": 1.4633,
      "step": 112
    },
    {
      "epoch": 0.5406698564593302,
      "grad_norm": 0.31697309017181396,
      "learning_rate": 4.5386582026834906e-05,
      "loss": 1.6007,
      "step": 113
    },
    {
      "epoch": 0.5454545454545454,
      "grad_norm": 0.3104463517665863,
      "learning_rate": 4.462044585485944e-05,
      "loss": 1.5146,
      "step": 114
    },
    {
      "epoch": 0.5502392344497608,
      "grad_norm": 0.3110423684120178,
      "learning_rate": 4.38555854667643e-05,
      "loss": 1.3926,
      "step": 115
    },
    {
      "epoch": 0.5550239234449761,
      "grad_norm": 0.3278864622116089,
      "learning_rate": 4.30921822524059e-05,
      "loss": 1.5483,
      "step": 116
    },
    {
      "epoch": 0.5598086124401914,
      "grad_norm": 0.31767237186431885,
      "learning_rate": 4.233041725606572e-05,
      "loss": 1.4713,
      "step": 117
    },
    {
      "epoch": 0.5645933014354066,
      "grad_norm": 0.3416193425655365,
      "learning_rate": 4.157047113351504e-05,
      "loss": 1.4755,
      "step": 118
    },
    {
      "epoch": 0.569377990430622,
      "grad_norm": 0.34195205569267273,
      "learning_rate": 4.0812524109171476e-05,
      "loss": 1.3458,
      "step": 119
    },
    {
      "epoch": 0.5741626794258373,
      "grad_norm": 0.34631067514419556,
      "learning_rate": 4.0056755933358184e-05,
      "loss": 1.6436,
      "step": 120
    },
    {
      "epoch": 0.5789473684210527,
      "grad_norm": 0.3526613116264343,
      "learning_rate": 3.930334583967514e-05,
      "loss": 1.69,
      "step": 121
    },
    {
      "epoch": 0.583732057416268,
      "grad_norm": 0.36171069741249084,
      "learning_rate": 3.855247250249331e-05,
      "loss": 1.4521,
      "step": 122
    },
    {
      "epoch": 0.5885167464114832,
      "grad_norm": 0.3751888871192932,
      "learning_rate": 3.780431399458114e-05,
      "loss": 1.7037,
      "step": 123
    },
    {
      "epoch": 0.5933014354066986,
      "grad_norm": 0.38322314620018005,
      "learning_rate": 3.705904774487396e-05,
      "loss": 1.5647,
      "step": 124
    },
    {
      "epoch": 0.5980861244019139,
      "grad_norm": 0.3606765568256378,
      "learning_rate": 3.631685049639586e-05,
      "loss": 1.6243,
      "step": 125
    },
    {
      "epoch": 0.6028708133971292,
      "grad_norm": 0.39295509457588196,
      "learning_rate": 3.557789826434439e-05,
      "loss": 1.6738,
      "step": 126
    },
    {
      "epoch": 0.6076555023923444,
      "grad_norm": 0.38839149475097656,
      "learning_rate": 3.484236629434783e-05,
      "loss": 1.5688,
      "step": 127
    },
    {
      "epoch": 0.6124401913875598,
      "grad_norm": 0.39863213896751404,
      "learning_rate": 3.411042902090492e-05,
      "loss": 1.5511,
      "step": 128
    },
    {
      "epoch": 0.6172248803827751,
      "grad_norm": 0.40311485528945923,
      "learning_rate": 3.338226002601703e-05,
      "loss": 1.7209,
      "step": 129
    },
    {
      "epoch": 0.6220095693779905,
      "grad_norm": 0.46205297112464905,
      "learning_rate": 3.265803199802237e-05,
      "loss": 1.8635,
      "step": 130
    },
    {
      "epoch": 0.6267942583732058,
      "grad_norm": 0.48374590277671814,
      "learning_rate": 3.1937916690642356e-05,
      "loss": 1.903,
      "step": 131
    },
    {
      "epoch": 0.631578947368421,
      "grad_norm": 0.4481237232685089,
      "learning_rate": 3.1222084882249375e-05,
      "loss": 1.6987,
      "step": 132
    },
    {
      "epoch": 0.6363636363636364,
      "grad_norm": 0.4824537932872772,
      "learning_rate": 3.0510706335366035e-05,
      "loss": 1.8271,
      "step": 133
    },
    {
      "epoch": 0.6411483253588517,
      "grad_norm": 0.4872766435146332,
      "learning_rate": 2.980394975640526e-05,
      "loss": 1.7831,
      "step": 134
    },
    {
      "epoch": 0.645933014354067,
      "grad_norm": 0.5188060402870178,
      "learning_rate": 2.910198275566085e-05,
      "loss": 1.7229,
      "step": 135
    },
    {
      "epoch": 0.6507177033492823,
      "grad_norm": 0.5814325213432312,
      "learning_rate": 2.8404971807557957e-05,
      "loss": 1.6868,
      "step": 136
    },
    {
      "epoch": 0.6555023923444976,
      "grad_norm": 0.5563578605651855,
      "learning_rate": 2.771308221117309e-05,
      "loss": 1.9105,
      "step": 137
    },
    {
      "epoch": 0.6602870813397129,
      "grad_norm": 0.7155702114105225,
      "learning_rate": 2.7026478051032623e-05,
      "loss": 2.0886,
      "step": 138
    },
    {
      "epoch": 0.6650717703349283,
      "grad_norm": 0.6362230181694031,
      "learning_rate": 2.6345322158199503e-05,
      "loss": 1.8675,
      "step": 139
    },
    {
      "epoch": 0.6698564593301436,
      "grad_norm": 0.7237346172332764,
      "learning_rate": 2.5669776071657192e-05,
      "loss": 1.9425,
      "step": 140
    },
    {
      "epoch": 0.6746411483253588,
      "grad_norm": 0.6796498894691467,
      "learning_rate": 2.500000000000001e-05,
      "loss": 1.7823,
      "step": 141
    },
    {
      "epoch": 0.6794258373205742,
      "grad_norm": 0.9227808117866516,
      "learning_rate": 2.4336152783438982e-05,
      "loss": 2.08,
      "step": 142
    },
    {
      "epoch": 0.6842105263157895,
      "grad_norm": 0.8645951747894287,
      "learning_rate": 2.3678391856132204e-05,
      "loss": 2.072,
      "step": 143
    },
    {
      "epoch": 0.6889952153110048,
      "grad_norm": 0.9071637988090515,
      "learning_rate": 2.302687320884876e-05,
      "loss": 1.9359,
      "step": 144
    },
    {
      "epoch": 0.69377990430622,
      "grad_norm": 0.9696687459945679,
      "learning_rate": 2.238175135197471e-05,
      "loss": 1.9997,
      "step": 145
    },
    {
      "epoch": 0.6985645933014354,
      "grad_norm": 1.3215420246124268,
      "learning_rate": 2.1743179278870407e-05,
      "loss": 2.4706,
      "step": 146
    },
    {
      "epoch": 0.7033492822966507,
      "grad_norm": 1.3039065599441528,
      "learning_rate": 2.1111308429587444e-05,
      "loss": 1.8514,
      "step": 147
    },
    {
      "epoch": 0.7081339712918661,
      "grad_norm": 1.5614275932312012,
      "learning_rate": 2.0486288654954028e-05,
      "loss": 1.6851,
      "step": 148
    },
    {
      "epoch": 0.7129186602870813,
      "grad_norm": 1.524907112121582,
      "learning_rate": 1.9868268181037185e-05,
      "loss": 1.4203,
      "step": 149
    },
    {
      "epoch": 0.7177033492822966,
      "grad_norm": 2.6234681606292725,
      "learning_rate": 1.925739357399038e-05,
      "loss": 1.613,
      "step": 150
    },
    {
      "epoch": 0.722488038277512,
      "grad_norm": 0.20270580053329468,
      "learning_rate": 1.8653809705294688e-05,
      "loss": 1.5107,
      "step": 151
    },
    {
      "epoch": 0.7272727272727273,
      "grad_norm": 0.2348601371049881,
      "learning_rate": 1.8057659717401947e-05,
      "loss": 1.4045,
      "step": 152
    },
    {
      "epoch": 0.7320574162679426,
      "grad_norm": 0.22996756434440613,
      "learning_rate": 1.746908498978791e-05,
      "loss": 1.4323,
      "step": 153
    },
    {
      "epoch": 0.7368421052631579,
      "grad_norm": 0.22488181293010712,
      "learning_rate": 1.6888225105423507e-05,
      "loss": 1.3006,
      "step": 154
    },
    {
      "epoch": 0.7416267942583732,
      "grad_norm": 0.2689021825790405,
      "learning_rate": 1.631521781767214e-05,
      "loss": 1.5859,
      "step": 155
    },
    {
      "epoch": 0.7464114832535885,
      "grad_norm": 0.24600256979465485,
      "learning_rate": 1.575019901762097e-05,
      "loss": 1.3297,
      "step": 156
    },
    {
      "epoch": 0.7511961722488039,
      "grad_norm": 0.2761135995388031,
      "learning_rate": 1.5193302701853673e-05,
      "loss": 1.3502,
      "step": 157
    },
    {
      "epoch": 0.7559808612440191,
      "grad_norm": 0.2814018130302429,
      "learning_rate": 1.4644660940672627e-05,
      "loss": 1.5272,
      "step": 158
    },
    {
      "epoch": 0.7607655502392344,
      "grad_norm": 0.29007115960121155,
      "learning_rate": 1.4104403846777909e-05,
      "loss": 1.4903,
      "step": 159
    },
    {
      "epoch": 0.7655502392344498,
      "grad_norm": 0.2952757477760315,
      "learning_rate": 1.3572659544410494e-05,
      "loss": 1.5378,
      "step": 160
    },
    {
      "epoch": 0.7703349282296651,
      "grad_norm": 0.311739057302475,
      "learning_rate": 1.3049554138967051e-05,
      "loss": 1.5632,
      "step": 161
    },
    {
      "epoch": 0.7751196172248804,
      "grad_norm": 0.32148003578186035,
      "learning_rate": 1.2535211687093535e-05,
      "loss": 1.5378,
      "step": 162
    },
    {
      "epoch": 0.7799043062200957,
      "grad_norm": 0.326665461063385,
      "learning_rate": 1.202975416726464e-05,
      "loss": 1.5604,
      "step": 163
    },
    {
      "epoch": 0.784688995215311,
      "grad_norm": 0.33070483803749084,
      "learning_rate": 1.1533301450856054e-05,
      "loss": 1.4631,
      "step": 164
    },
    {
      "epoch": 0.7894736842105263,
      "grad_norm": 0.34155699610710144,
      "learning_rate": 1.1045971273716477e-05,
      "loss": 1.53,
      "step": 165
    },
    {
      "epoch": 0.7942583732057417,
      "grad_norm": 0.3466791808605194,
      "learning_rate": 1.0567879208246084e-05,
      "loss": 1.5503,
      "step": 166
    },
    {
      "epoch": 0.7990430622009569,
      "grad_norm": 0.3450649082660675,
      "learning_rate": 1.0099138635988026e-05,
      "loss": 1.5562,
      "step": 167
    },
    {
      "epoch": 0.8038277511961722,
      "grad_norm": 0.3394903242588043,
      "learning_rate": 9.639860720739525e-06,
      "loss": 1.3966,
      "step": 168
    },
    {
      "epoch": 0.8086124401913876,
      "grad_norm": 0.34013882279396057,
      "learning_rate": 9.190154382188921e-06,
      "loss": 1.4578,
      "step": 169
    },
    {
      "epoch": 0.8133971291866029,
      "grad_norm": 0.40502288937568665,
      "learning_rate": 8.75012627008489e-06,
      "loss": 1.4135,
      "step": 170
    },
    {
      "epoch": 0.8181818181818182,
      "grad_norm": 0.3667527437210083,
      "learning_rate": 8.31988073894403e-06,
      "loss": 1.539,
      "step": 171
    },
    {
      "epoch": 0.8229665071770335,
      "grad_norm": 0.39868226647377014,
      "learning_rate": 7.899519823302743e-06,
      "loss": 1.7387,
      "step": 172
    },
    {
      "epoch": 0.8277511961722488,
      "grad_norm": 0.3658325672149658,
      "learning_rate": 7.489143213519301e-06,
      "loss": 1.5761,
      "step": 173
    },
    {
      "epoch": 0.8325358851674641,
      "grad_norm": 0.3974853456020355,
      "learning_rate": 7.088848232131861e-06,
      "loss": 1.5148,
      "step": 174
    },
    {
      "epoch": 0.8373205741626795,
      "grad_norm": 0.4094543755054474,
      "learning_rate": 6.698729810778065e-06,
      "loss": 1.5331,
      "step": 175
    },
    {
      "epoch": 0.8421052631578947,
      "grad_norm": 0.4123484790325165,
      "learning_rate": 6.318880467681526e-06,
      "loss": 1.7296,
      "step": 176
    },
    {
      "epoch": 0.84688995215311,
      "grad_norm": 0.44035616517066956,
      "learning_rate": 5.949390285710776e-06,
      "loss": 1.6646,
      "step": 177
    },
    {
      "epoch": 0.8516746411483254,
      "grad_norm": 0.40625545382499695,
      "learning_rate": 5.590346891015758e-06,
      "loss": 1.5258,
      "step": 178
    },
    {
      "epoch": 0.8564593301435407,
      "grad_norm": 0.5013072490692139,
      "learning_rate": 5.241835432246889e-06,
      "loss": 1.8122,
      "step": 179
    },
    {
      "epoch": 0.861244019138756,
      "grad_norm": 0.4660516083240509,
      "learning_rate": 4.903938560361698e-06,
      "loss": 1.5325,
      "step": 180
    },
    {
      "epoch": 0.8660287081339713,
      "grad_norm": 0.48506811261177063,
      "learning_rate": 4.576736409023813e-06,
      "loss": 1.7429,
      "step": 181
    },
    {
      "epoch": 0.8708133971291866,
      "grad_norm": 0.5250177979469299,
      "learning_rate": 4.260306575598949e-06,
      "loss": 1.6214,
      "step": 182
    },
    {
      "epoch": 0.8755980861244019,
      "grad_norm": 0.5322927236557007,
      "learning_rate": 3.954724102752316e-06,
      "loss": 1.732,
      "step": 183
    },
    {
      "epoch": 0.8803827751196173,
      "grad_norm": 0.5320829153060913,
      "learning_rate": 3.660061460651981e-06,
      "loss": 1.8157,
      "step": 184
    },
    {
      "epoch": 0.8851674641148325,
      "grad_norm": 0.5766329169273376,
      "learning_rate": 3.376388529782215e-06,
      "loss": 1.9319,
      "step": 185
    },
    {
      "epoch": 0.8899521531100478,
      "grad_norm": 0.5621324777603149,
      "learning_rate": 3.1037725843711062e-06,
      "loss": 1.7481,
      "step": 186
    },
    {
      "epoch": 0.8947368421052632,
      "grad_norm": 0.654734194278717,
      "learning_rate": 2.842278276436128e-06,
      "loss": 2.079,
      "step": 187
    },
    {
      "epoch": 0.8995215311004785,
      "grad_norm": 0.6522572636604309,
      "learning_rate": 2.591967620451707e-06,
      "loss": 1.8302,
      "step": 188
    },
    {
      "epoch": 0.9043062200956937,
      "grad_norm": 0.7326610088348389,
      "learning_rate": 2.3528999786421756e-06,
      "loss": 1.9044,
      "step": 189
    },
    {
      "epoch": 0.9090909090909091,
      "grad_norm": 0.7005221247673035,
      "learning_rate": 2.1251320469037827e-06,
      "loss": 1.9704,
      "step": 190
    },
    {
      "epoch": 0.9138755980861244,
      "grad_norm": 0.6769828200340271,
      "learning_rate": 1.908717841359048e-06,
      "loss": 1.7863,
      "step": 191
    },
    {
      "epoch": 0.9186602870813397,
      "grad_norm": 0.7701701521873474,
      "learning_rate": 1.70370868554659e-06,
      "loss": 1.8417,
      "step": 192
    },
    {
      "epoch": 0.9234449760765551,
      "grad_norm": 0.781731903553009,
      "learning_rate": 1.5101531982495308e-06,
      "loss": 1.7861,
      "step": 193
    },
    {
      "epoch": 0.9282296650717703,
      "grad_norm": 0.9508692026138306,
      "learning_rate": 1.328097281965357e-06,
      "loss": 1.975,
      "step": 194
    },
    {
      "epoch": 0.9330143540669856,
      "grad_norm": 0.9699769616127014,
      "learning_rate": 1.157584112019966e-06,
      "loss": 2.0208,
      "step": 195
    },
    {
      "epoch": 0.937799043062201,
      "grad_norm": 1.1134250164031982,
      "learning_rate": 9.986541263284077e-07,
      "loss": 1.7791,
      "step": 196
    },
    {
      "epoch": 0.9425837320574163,
      "grad_norm": 1.3487857580184937,
      "learning_rate": 8.513450158049108e-07,
      "loss": 1.9681,
      "step": 197
    },
    {
      "epoch": 0.9473684210526315,
      "grad_norm": 1.709845781326294,
      "learning_rate": 7.156917154243048e-07,
      "loss": 1.983,
      "step": 198
    },
    {
      "epoch": 0.9521531100478469,
      "grad_norm": 2.0014631748199463,
      "learning_rate": 5.917263959370312e-07,
      "loss": 2.0637,
      "step": 199
    },
    {
      "epoch": 0.9569377990430622,
      "grad_norm": 2.2905828952789307,
      "learning_rate": 4.794784562397458e-07,
      "loss": 1.5265,
      "step": 200
    },
    {
      "epoch": 0.9617224880382775,
      "grad_norm": 0.2522173821926117,
      "learning_rate": 3.7897451640321323e-07,
      "loss": 1.4622,
      "step": 201
    },
    {
      "epoch": 0.9665071770334929,
      "grad_norm": 0.2915673851966858,
      "learning_rate": 2.902384113592782e-07,
      "loss": 1.7226,
      "step": 202
    },
    {
      "epoch": 0.9712918660287081,
      "grad_norm": 0.31829988956451416,
      "learning_rate": 2.1329118524827662e-07,
      "loss": 1.444,
      "step": 203
    },
    {
      "epoch": 0.9760765550239234,
      "grad_norm": 0.35566747188568115,
      "learning_rate": 1.481510864283553e-07,
      "loss": 1.6455,
      "step": 204
    },
    {
      "epoch": 0.9808612440191388,
      "grad_norm": 0.43001261353492737,
      "learning_rate": 9.483356314779479e-08,
      "loss": 1.7939,
      "step": 205
    },
    {
      "epoch": 0.9856459330143541,
      "grad_norm": 0.5118626356124878,
      "learning_rate": 5.3351259881379014e-08,
      "loss": 1.823,
      "step": 206
    },
    {
      "epoch": 0.9904306220095693,
      "grad_norm": 0.6911094784736633,
      "learning_rate": 2.371401433170495e-08,
      "loss": 2.0949,
      "step": 207
    },
    {
      "epoch": 0.9952153110047847,
      "grad_norm": 0.847305953502655,
      "learning_rate": 5.928855096154484e-09,
      "loss": 1.5913,
      "step": 208
    },
    {
      "epoch": 1.0,
      "grad_norm": 1.5929991006851196,
      "learning_rate": 0.0,
      "loss": 1.6612,
      "step": 209
    }
  ],
  "logging_steps": 1,
  "max_steps": 209,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 53,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2.769507127708877e+16,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}