{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 20.0,
  "eval_steps": 500,
  "global_step": 310860,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.06433764395547835,
      "grad_norm": 1.052815556526184,
      "learning_rate": 3.125e-05,
      "loss": 6.1013,
      "step": 1000
    },
    {
      "epoch": 0.1286752879109567,
      "grad_norm": 1.0303527116775513,
      "learning_rate": 6.25e-05,
      "loss": 4.5097,
      "step": 2000
    },
    {
      "epoch": 0.19301293186643506,
      "grad_norm": 0.9319628477096558,
      "learning_rate": 9.375e-05,
      "loss": 4.1714,
      "step": 3000
    },
    {
      "epoch": 0.2573505758219134,
      "grad_norm": 0.8786389827728271,
      "learning_rate": 0.000125,
      "loss": 3.9371,
      "step": 4000
    },
    {
      "epoch": 0.32168821977739176,
      "grad_norm": 0.7859598398208618,
      "learning_rate": 0.00015625,
      "loss": 3.7854,
      "step": 5000
    },
    {
      "epoch": 0.3860258637328701,
      "grad_norm": 0.7061841487884521,
      "learning_rate": 0.0001875,
      "loss": 3.6647,
      "step": 6000
    },
    {
      "epoch": 0.45036350768834843,
      "grad_norm": 0.6750425100326538,
      "learning_rate": 0.00021875,
      "loss": 3.576,
      "step": 7000
    },
    {
      "epoch": 0.5147011516438268,
      "grad_norm": 0.6492879986763,
      "learning_rate": 0.00025,
      "loss": 3.5131,
      "step": 8000
    },
    {
      "epoch": 0.5790387955993052,
      "grad_norm": 0.639130175113678,
      "learning_rate": 0.00028125000000000003,
      "loss": 3.4541,
      "step": 9000
    },
    {
      "epoch": 0.6433764395547835,
      "grad_norm": 0.5938445329666138,
      "learning_rate": 0.00031246875000000003,
      "loss": 3.4065,
      "step": 10000
    },
    {
      "epoch": 0.7077140835102619,
      "grad_norm": 0.5456812381744385,
      "learning_rate": 0.00034371875,
      "loss": 3.3722,
      "step": 11000
    },
    {
      "epoch": 0.7720517274657402,
      "grad_norm": 0.5338535308837891,
      "learning_rate": 0.00037496875000000003,
      "loss": 3.3338,
      "step": 12000
    },
    {
      "epoch": 0.8363893714212186,
      "grad_norm": 0.5021286606788635,
      "learning_rate": 0.0004061875,
      "loss": 3.3068,
      "step": 13000
    },
    {
      "epoch": 0.9007270153766969,
      "grad_norm": 0.45191365480422974,
      "learning_rate": 0.0004374375,
      "loss": 3.2859,
      "step": 14000
    },
    {
      "epoch": 0.9650646593321752,
      "grad_norm": 0.40418481826782227,
      "learning_rate": 0.0004686875,
      "loss": 3.26,
      "step": 15000
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.38307077369468,
      "eval_loss": 3.3135979175567627,
      "eval_runtime": 113.2217,
      "eval_samples_per_second": 463.162,
      "eval_steps_per_second": 7.242,
      "step": 15543
    },
    {
      "epoch": 1.0294023032876536,
      "grad_norm": 0.39907780289649963,
      "learning_rate": 0.0004999375,
      "loss": 3.2167,
      "step": 16000
    },
    {
      "epoch": 1.093739947243132,
      "grad_norm": 0.380867600440979,
      "learning_rate": 0.00053115625,
      "loss": 3.1969,
      "step": 17000
    },
    {
      "epoch": 1.1580775911986103,
      "grad_norm": 0.3618770241737366,
      "learning_rate": 0.0005623749999999999,
      "loss": 3.1923,
      "step": 18000
    },
    {
      "epoch": 1.2224152351540887,
      "grad_norm": 0.3256314694881439,
      "learning_rate": 0.000593625,
      "loss": 3.1767,
      "step": 19000
    },
    {
      "epoch": 1.286752879109567,
      "grad_norm": 0.31108617782592773,
      "learning_rate": 0.000624875,
      "loss": 3.1643,
      "step": 20000
    },
    {
      "epoch": 1.3510905230650454,
      "grad_norm": 0.30397939682006836,
      "learning_rate": 0.000656125,
      "loss": 3.154,
      "step": 21000
    },
    {
      "epoch": 1.4154281670205238,
      "grad_norm": 0.2824731171131134,
      "learning_rate": 0.00068734375,
      "loss": 3.1451,
      "step": 22000
    },
    {
      "epoch": 1.4797658109760021,
      "grad_norm": 0.28574618697166443,
      "learning_rate": 0.00071859375,
      "loss": 3.1347,
      "step": 23000
    },
    {
      "epoch": 1.5441034549314803,
      "grad_norm": 0.26015007495880127,
      "learning_rate": 0.0007498125,
      "loss": 3.1263,
      "step": 24000
    },
    {
      "epoch": 1.6084410988869586,
      "grad_norm": 0.2511335611343384,
      "learning_rate": 0.0007810312499999999,
      "loss": 3.1217,
      "step": 25000
    },
    {
      "epoch": 1.672778742842437,
      "grad_norm": 0.2407851368188858,
      "learning_rate": 0.00081228125,
      "loss": 3.0962,
      "step": 26000
    },
    {
      "epoch": 1.7371163867979154,
      "grad_norm": 0.23978936672210693,
      "learning_rate": 0.0008435000000000001,
      "loss": 3.1022,
      "step": 27000
    },
    {
      "epoch": 1.8014540307533937,
      "grad_norm": 0.20840179920196533,
      "learning_rate": 0.0008747500000000001,
      "loss": 3.0929,
      "step": 28000
    },
    {
      "epoch": 1.865791674708872,
      "grad_norm": 0.2035219669342041,
      "learning_rate": 0.000906,
      "loss": 3.0824,
      "step": 29000
    },
    {
      "epoch": 1.9301293186643504,
      "grad_norm": 0.19087807834148407,
      "learning_rate": 0.00093721875,
      "loss": 3.0732,
      "step": 30000
    },
    {
      "epoch": 1.9944669626198288,
      "grad_norm": 0.19520416855812073,
      "learning_rate": 0.00096846875,
      "loss": 3.0629,
      "step": 31000
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.399560057432584,
      "eval_loss": 3.1529746055603027,
      "eval_runtime": 113.8443,
      "eval_samples_per_second": 460.629,
      "eval_steps_per_second": 7.203,
      "step": 31086
    },
    {
      "epoch": 2.058804606575307,
      "grad_norm": 0.205551877617836,
      "learning_rate": 0.00099971875,
      "loss": 3.0167,
      "step": 32000
    },
    {
      "epoch": 2.1231422505307855,
      "grad_norm": 0.18326979875564575,
      "learning_rate": 0.000996449831456645,
      "loss": 3.0168,
      "step": 33000
    },
    {
      "epoch": 2.187479894486264,
      "grad_norm": 0.18665415048599243,
      "learning_rate": 0.000992863802624973,
      "loss": 3.0026,
      "step": 34000
    },
    {
      "epoch": 2.2518175384417423,
      "grad_norm": 0.186910018324852,
      "learning_rate": 0.0009892813598221329,
      "loss": 3.0041,
      "step": 35000
    },
    {
      "epoch": 2.3161551823972206,
      "grad_norm": 0.17289380729198456,
      "learning_rate": 0.0009856953309904611,
      "loss": 2.9883,
      "step": 36000
    },
    {
      "epoch": 2.380492826352699,
      "grad_norm": 0.18370352685451508,
      "learning_rate": 0.000982112888187621,
      "loss": 2.9854,
      "step": 37000
    },
    {
      "epoch": 2.4448304703081774,
      "grad_norm": 0.17503343522548676,
      "learning_rate": 0.0009785268593559492,
      "loss": 2.9813,
      "step": 38000
    },
    {
      "epoch": 2.5091681142636557,
      "grad_norm": 0.19140906631946564,
      "learning_rate": 0.0009749408305242774,
      "loss": 2.9785,
      "step": 39000
    },
    {
      "epoch": 2.573505758219134,
      "grad_norm": 0.19327034056186676,
      "learning_rate": 0.0009713583877214373,
      "loss": 2.9781,
      "step": 40000
    },
    {
      "epoch": 2.6378434021746124,
      "grad_norm": 0.17439500987529755,
      "learning_rate": 0.0009677723588897654,
      "loss": 2.964,
      "step": 41000
    },
    {
      "epoch": 2.702181046130091,
      "grad_norm": 0.16128753125667572,
      "learning_rate": 0.0009641899160869255,
      "loss": 2.9543,
      "step": 42000
    },
    {
      "epoch": 2.766518690085569,
      "grad_norm": 0.16443367302417755,
      "learning_rate": 0.0009606038872552536,
      "loss": 2.9606,
      "step": 43000
    },
    {
      "epoch": 2.8308563340410475,
      "grad_norm": 0.17449279129505157,
      "learning_rate": 0.0009570178584235818,
      "loss": 2.9465,
      "step": 44000
    },
    {
      "epoch": 2.895193977996526,
      "grad_norm": 0.17141327261924744,
      "learning_rate": 0.00095343182959191,
      "loss": 2.9433,
      "step": 45000
    },
    {
      "epoch": 2.9595316219520043,
      "grad_norm": 0.15691405534744263,
      "learning_rate": 0.0009498493867890699,
      "loss": 2.9355,
      "step": 46000
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.4113637995243864,
      "eval_loss": 3.042464017868042,
      "eval_runtime": 113.33,
      "eval_samples_per_second": 462.719,
      "eval_steps_per_second": 7.236,
      "step": 46629
    },
    {
      "epoch": 3.0238692659074826,
      "grad_norm": 0.1875714659690857,
      "learning_rate": 0.000946263357957398,
      "loss": 2.9059,
      "step": 47000
    },
    {
      "epoch": 3.088206909862961,
      "grad_norm": 0.18929240107536316,
      "learning_rate": 0.0009426773291257262,
      "loss": 2.8621,
      "step": 48000
    },
    {
      "epoch": 3.1525445538184393,
      "grad_norm": 0.169860377907753,
      "learning_rate": 0.0009390948863228861,
      "loss": 2.8641,
      "step": 49000
    },
    {
      "epoch": 3.2168821977739177,
      "grad_norm": 0.18511784076690674,
      "learning_rate": 0.0009355088574912143,
      "loss": 2.867,
      "step": 50000
    },
    {
      "epoch": 3.2812198417293956,
      "grad_norm": 0.2011699676513672,
      "learning_rate": 0.0009319228286595425,
      "loss": 2.864,
      "step": 51000
    },
    {
      "epoch": 3.345557485684874,
      "grad_norm": 0.17710469663143158,
      "learning_rate": 0.0009283403858567023,
      "loss": 2.8627,
      "step": 52000
    },
    {
      "epoch": 3.4098951296403524,
      "grad_norm": 0.176873579621315,
      "learning_rate": 0.0009247543570250305,
      "loss": 2.8658,
      "step": 53000
    },
    {
      "epoch": 3.4742327735958307,
      "grad_norm": 0.16512398421764374,
      "learning_rate": 0.0009211683281933588,
      "loss": 2.8703,
      "step": 54000
    },
    {
      "epoch": 3.538570417551309,
      "grad_norm": 0.15719538927078247,
      "learning_rate": 0.0009175858853905186,
      "loss": 2.8667,
      "step": 55000
    },
    {
      "epoch": 3.6029080615067874,
      "grad_norm": 0.15996688604354858,
      "learning_rate": 0.0009139998565588467,
      "loss": 2.8633,
      "step": 56000
    },
    {
      "epoch": 3.667245705462266,
      "grad_norm": 0.16842438280582428,
      "learning_rate": 0.0009104174137560066,
      "loss": 2.8584,
      "step": 57000
    },
    {
      "epoch": 3.731583349417744,
      "grad_norm": 0.16790573298931122,
      "learning_rate": 0.0009068313849243348,
      "loss": 2.8594,
      "step": 58000
    },
    {
      "epoch": 3.7959209933732225,
      "grad_norm": 0.2214391678571701,
      "learning_rate": 0.000903245356092663,
      "loss": 2.8556,
      "step": 59000
    },
    {
      "epoch": 3.860258637328701,
      "grad_norm": 0.16262733936309814,
      "learning_rate": 0.0008996593272609912,
      "loss": 2.86,
      "step": 60000
    },
    {
      "epoch": 3.9245962812841793,
      "grad_norm": 0.17010310292243958,
      "learning_rate": 0.000896076884458151,
      "loss": 2.8557,
      "step": 61000
    },
    {
      "epoch": 3.9889339252396576,
      "grad_norm": 0.1839476078748703,
      "learning_rate": 0.0008924908556264793,
      "loss": 2.8499,
      "step": 62000
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.41723433690791345,
      "eval_loss": 2.993454694747925,
      "eval_runtime": 114.1434,
      "eval_samples_per_second": 459.422,
      "eval_steps_per_second": 7.184,
      "step": 62172
    },
    {
      "epoch": 4.053271569195136,
      "grad_norm": 0.1961131989955902,
      "learning_rate": 0.0008889048267948075,
      "loss": 2.7827,
      "step": 63000
    },
    {
      "epoch": 4.117609213150614,
      "grad_norm": 0.17030374705791473,
      "learning_rate": 0.0008853187979631356,
      "loss": 2.7733,
      "step": 64000
    },
    {
      "epoch": 4.181946857106093,
      "grad_norm": 0.1887962967157364,
      "learning_rate": 0.0008817363551602955,
      "loss": 2.7735,
      "step": 65000
    },
    {
      "epoch": 4.246284501061571,
      "grad_norm": 0.20483681559562683,
      "learning_rate": 0.0008781539123574553,
      "loss": 2.7816,
      "step": 66000
    },
    {
      "epoch": 4.310622145017049,
      "grad_norm": 0.19994668662548065,
      "learning_rate": 0.0008745678835257835,
      "loss": 2.7895,
      "step": 67000
    },
    {
      "epoch": 4.374959788972528,
      "grad_norm": 0.17334817349910736,
      "learning_rate": 0.0008709818546941118,
      "loss": 2.7845,
      "step": 68000
    },
    {
      "epoch": 4.439297432928006,
      "grad_norm": 0.2339029610157013,
      "learning_rate": 0.0008673994118912716,
      "loss": 2.7915,
      "step": 69000
    },
    {
      "epoch": 4.5036350768834845,
      "grad_norm": 0.17018085718154907,
      "learning_rate": 0.0008638133830595997,
      "loss": 2.7879,
      "step": 70000
    },
    {
      "epoch": 4.567972720838963,
      "grad_norm": 0.23095539212226868,
      "learning_rate": 0.000860227354227928,
      "loss": 2.7954,
      "step": 71000
    },
    {
      "epoch": 4.632310364794441,
      "grad_norm": 0.2322332113981247,
      "learning_rate": 0.0008566413253962562,
      "loss": 2.7897,
      "step": 72000
    },
    {
      "epoch": 4.69664800874992,
      "grad_norm": 0.19111447036266327,
      "learning_rate": 0.0008530588825934161,
      "loss": 2.799,
      "step": 73000
    },
    {
      "epoch": 4.760985652705398,
      "grad_norm": 0.17069219052791595,
      "learning_rate": 0.0008494728537617442,
      "loss": 2.7923,
      "step": 74000
    },
    {
      "epoch": 4.825323296660876,
      "grad_norm": 0.20082932710647583,
      "learning_rate": 0.0008458904109589041,
      "loss": 2.7956,
      "step": 75000
    },
    {
      "epoch": 4.889660940616355,
      "grad_norm": 0.20119306445121765,
      "learning_rate": 0.000842307968156064,
      "loss": 2.7873,
      "step": 76000
    },
    {
      "epoch": 4.953998584571833,
      "grad_norm": 0.16935457289218903,
      "learning_rate": 0.0008387219393243922,
      "loss": 2.7951,
      "step": 77000
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.4206606990622336,
      "eval_loss": 2.9680817127227783,
      "eval_runtime": 113.8277,
      "eval_samples_per_second": 460.696,
      "eval_steps_per_second": 7.204,
      "step": 77715
    },
    {
      "epoch": 5.018336228527311,
      "grad_norm": 0.19914424419403076,
      "learning_rate": 0.0008351359104927204,
      "loss": 2.7629,
      "step": 78000
    },
    {
      "epoch": 5.08267387248279,
      "grad_norm": 0.19470812380313873,
      "learning_rate": 0.0008315534676898803,
      "loss": 2.7089,
      "step": 79000
    },
    {
      "epoch": 5.147011516438268,
      "grad_norm": 0.2205093652009964,
      "learning_rate": 0.0008279674388582084,
      "loss": 2.7133,
      "step": 80000
    },
    {
      "epoch": 5.2113491603937465,
      "grad_norm": 0.2074008285999298,
      "learning_rate": 0.0008243814100265367,
      "loss": 2.723,
      "step": 81000
    },
    {
      "epoch": 5.275686804349225,
      "grad_norm": 0.3117210566997528,
      "learning_rate": 0.0008207989672236965,
      "loss": 2.7205,
      "step": 82000
    },
    {
      "epoch": 5.340024448304703,
      "grad_norm": 0.22853530943393707,
      "learning_rate": 0.0008172129383920247,
      "loss": 2.7304,
      "step": 83000
    },
    {
      "epoch": 5.404362092260182,
      "grad_norm": 0.1936911940574646,
      "learning_rate": 0.0008136304955891846,
      "loss": 2.7259,
      "step": 84000
    },
    {
      "epoch": 5.46869973621566,
      "grad_norm": 0.19292935729026794,
      "learning_rate": 0.0008100444667575127,
      "loss": 2.7283,
      "step": 85000
    },
    {
      "epoch": 5.533037380171138,
      "grad_norm": 0.19306513667106628,
      "learning_rate": 0.0008064584379258409,
      "loss": 2.7349,
      "step": 86000
    },
    {
      "epoch": 5.597375024126617,
      "grad_norm": 0.1781722456216812,
      "learning_rate": 0.0008028759951230008,
      "loss": 2.7367,
      "step": 87000
    },
    {
      "epoch": 5.661712668082095,
      "grad_norm": 0.19567584991455078,
      "learning_rate": 0.000799289966291329,
      "loss": 2.7386,
      "step": 88000
    },
    {
      "epoch": 5.726050312037573,
      "grad_norm": 0.17735910415649414,
      "learning_rate": 0.0007957075234884889,
      "loss": 2.7434,
      "step": 89000
    },
    {
      "epoch": 5.790387955993052,
      "grad_norm": 0.18065251410007477,
      "learning_rate": 0.000792121494656817,
      "loss": 2.7402,
      "step": 90000
    },
    {
      "epoch": 5.85472559994853,
      "grad_norm": 0.17205560207366943,
      "learning_rate": 0.0007885354658251452,
      "loss": 2.7437,
      "step": 91000
    },
    {
      "epoch": 5.9190632439040085,
      "grad_norm": 0.2087228149175644,
      "learning_rate": 0.0007849494369934735,
      "loss": 2.7442,
      "step": 92000
    },
    {
      "epoch": 5.983400887859487,
      "grad_norm": 0.21673254668712616,
      "learning_rate": 0.0007813669941906333,
      "loss": 2.7489,
      "step": 93000
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.42259418794214865,
      "eval_loss": 2.9527182579040527,
      "eval_runtime": 113.5442,
      "eval_samples_per_second": 461.847,
      "eval_steps_per_second": 7.222,
      "step": 93258
    },
    {
      "epoch": 6.047738531814965,
      "grad_norm": 0.22024141252040863,
      "learning_rate": 0.0007777809653589614,
      "loss": 2.6731,
      "step": 94000
    },
    {
      "epoch": 6.112076175770444,
      "grad_norm": 0.20330588519573212,
      "learning_rate": 0.0007741949365272897,
      "loss": 2.663,
      "step": 95000
    },
    {
      "epoch": 6.176413819725922,
      "grad_norm": 0.2359427958726883,
      "learning_rate": 0.0007706124937244495,
      "loss": 2.666,
      "step": 96000
    },
    {
      "epoch": 6.2407514636814,
      "grad_norm": 0.18912553787231445,
      "learning_rate": 0.0007670264648927777,
      "loss": 2.6751,
      "step": 97000
    },
    {
      "epoch": 6.305089107636879,
      "grad_norm": 0.18407639861106873,
      "learning_rate": 0.0007634440220899376,
      "loss": 2.6759,
      "step": 98000
    },
    {
      "epoch": 6.369426751592357,
      "grad_norm": 0.1985590159893036,
      "learning_rate": 0.0007598579932582657,
      "loss": 2.6759,
      "step": 99000
    },
    {
      "epoch": 6.433764395547835,
      "grad_norm": 0.18936724960803986,
      "learning_rate": 0.0007562719644265939,
      "loss": 2.685,
      "step": 100000
    },
    {
      "epoch": 6.498102039503314,
      "grad_norm": 0.23301662504673004,
      "learning_rate": 0.0007526895216237539,
      "loss": 2.6839,
      "step": 101000
    },
    {
      "epoch": 6.562439683458791,
      "grad_norm": 0.19822722673416138,
      "learning_rate": 0.000749103492792082,
      "loss": 2.6893,
      "step": 102000
    },
    {
      "epoch": 6.6267773274142705,
      "grad_norm": 0.18612362444400787,
      "learning_rate": 0.0007455174639604102,
      "loss": 2.6973,
      "step": 103000
    },
    {
      "epoch": 6.691114971369748,
      "grad_norm": 0.20918478071689606,
      "learning_rate": 0.0007419314351287384,
      "loss": 2.691,
      "step": 104000
    },
    {
      "epoch": 6.755452615325227,
      "grad_norm": 0.18845367431640625,
      "learning_rate": 0.0007383489923258983,
      "loss": 2.7032,
      "step": 105000
    },
    {
      "epoch": 6.819790259280705,
      "grad_norm": 0.19247417151927948,
      "learning_rate": 0.0007347665495230582,
      "loss": 2.6998,
      "step": 106000
    },
    {
      "epoch": 6.884127903236184,
      "grad_norm": 0.18703711032867432,
      "learning_rate": 0.0007311805206913864,
      "loss": 2.6992,
      "step": 107000
    },
    {
      "epoch": 6.948465547191661,
      "grad_norm": 0.21239759027957916,
      "learning_rate": 0.0007275944918597146,
      "loss": 2.7031,
      "step": 108000
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.42415787978044,
      "eval_loss": 2.9438040256500244,
      "eval_runtime": 113.2221,
      "eval_samples_per_second": 463.16,
      "eval_steps_per_second": 7.242,
      "step": 108801
    },
    {
      "epoch": 7.01280319114714,
      "grad_norm": 0.21643811464309692,
      "learning_rate": 0.0007240084630280428,
      "loss": 2.6826,
      "step": 109000
    },
    {
      "epoch": 7.077140835102618,
      "grad_norm": 0.21807773411273956,
      "learning_rate": 0.0007204260202252026,
      "loss": 2.6153,
      "step": 110000
    },
    {
      "epoch": 7.1414784790580965,
      "grad_norm": 0.23160989582538605,
      "learning_rate": 0.0007168399913935309,
      "loss": 2.6234,
      "step": 111000
    },
    {
      "epoch": 7.205816123013575,
      "grad_norm": 0.25499585270881653,
      "learning_rate": 0.0007132539625618591,
      "loss": 2.6296,
      "step": 112000
    },
    {
      "epoch": 7.270153766969053,
      "grad_norm": 0.20166288316249847,
      "learning_rate": 0.0007096679337301872,
      "loss": 2.6332,
      "step": 113000
    },
    {
      "epoch": 7.334491410924532,
      "grad_norm": 0.23443199694156647,
      "learning_rate": 0.0007060854909273471,
      "loss": 2.639,
      "step": 114000
    },
    {
      "epoch": 7.39882905488001,
      "grad_norm": 0.2512323260307312,
      "learning_rate": 0.0007024994620956753,
      "loss": 2.6427,
      "step": 115000
    },
    {
      "epoch": 7.463166698835488,
      "grad_norm": 0.21170075237751007,
      "learning_rate": 0.0006989134332640035,
      "loss": 2.6516,
      "step": 116000
    },
    {
      "epoch": 7.527504342790967,
      "grad_norm": 0.216237410902977,
      "learning_rate": 0.0006953274044323316,
      "loss": 2.6487,
      "step": 117000
    },
    {
      "epoch": 7.591841986746445,
      "grad_norm": 0.21823635697364807,
      "learning_rate": 0.0006917449616294915,
      "loss": 2.6499,
      "step": 118000
    },
    {
      "epoch": 7.656179630701923,
      "grad_norm": 0.23018254339694977,
      "learning_rate": 0.0006881625188266513,
      "loss": 2.6536,
      "step": 119000
    },
    {
      "epoch": 7.720517274657402,
      "grad_norm": 0.26946449279785156,
      "learning_rate": 0.0006845764899949796,
      "loss": 2.6617,
      "step": 120000
    },
    {
      "epoch": 7.78485491861288,
      "grad_norm": 0.19592489302158356,
      "learning_rate": 0.0006809940471921394,
      "loss": 2.6571,
      "step": 121000
    },
    {
      "epoch": 7.8491925625683585,
      "grad_norm": 0.1997816413640976,
      "learning_rate": 0.0006774080183604677,
      "loss": 2.6551,
      "step": 122000
    },
    {
      "epoch": 7.913530206523837,
      "grad_norm": 0.21142259240150452,
      "learning_rate": 0.0006738219895287958,
      "loss": 2.6638,
      "step": 123000
    },
    {
      "epoch": 7.977867850479315,
      "grad_norm": 0.21287241578102112,
      "learning_rate": 0.000670235960697124,
      "loss": 2.6642,
      "step": 124000
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.42513700064312526,
      "eval_loss": 2.942474126815796,
      "eval_runtime": 112.6722,
      "eval_samples_per_second": 465.421,
      "eval_steps_per_second": 7.278,
      "step": 124344
    },
    {
      "epoch": 8.042205494434794,
      "grad_norm": 0.20446690917015076,
      "learning_rate": 0.0006666535178942839,
      "loss": 2.6008,
      "step": 125000
    },
    {
      "epoch": 8.106543138390272,
      "grad_norm": 0.21123293042182922,
      "learning_rate": 0.0006630710750914437,
      "loss": 2.5828,
      "step": 126000
    },
    {
      "epoch": 8.170880782345751,
      "grad_norm": 0.22513774037361145,
      "learning_rate": 0.0006594850462597719,
      "loss": 2.5876,
      "step": 127000
    },
    {
      "epoch": 8.235218426301229,
      "grad_norm": 0.2192804217338562,
      "learning_rate": 0.0006559026034569318,
      "loss": 2.5988,
      "step": 128000
    },
    {
      "epoch": 8.299556070256708,
      "grad_norm": 0.2216465324163437,
      "learning_rate": 0.0006523165746252599,
      "loss": 2.6021,
      "step": 129000
    },
    {
      "epoch": 8.363893714212185,
      "grad_norm": 0.22872082889080048,
      "learning_rate": 0.0006487305457935882,
      "loss": 2.6028,
      "step": 130000
    },
    {
      "epoch": 8.428231358167665,
      "grad_norm": 0.2199559062719345,
      "learning_rate": 0.0006451481029907481,
      "loss": 2.6089,
      "step": 131000
    },
    {
      "epoch": 8.492569002123142,
      "grad_norm": 0.22230418026447296,
      "learning_rate": 0.0006415620741590763,
      "loss": 2.6098,
      "step": 132000
    },
    {
      "epoch": 8.556906646078621,
      "grad_norm": 0.21087001264095306,
      "learning_rate": 0.0006379760453274045,
      "loss": 2.6168,
      "step": 133000
    },
    {
      "epoch": 8.621244290034099,
      "grad_norm": 0.2557260990142822,
      "learning_rate": 0.0006343936025245643,
      "loss": 2.6205,
      "step": 134000
    },
    {
      "epoch": 8.685581933989578,
      "grad_norm": 0.2882158160209656,
      "learning_rate": 0.0006308075736928925,
      "loss": 2.6236,
      "step": 135000
    },
    {
      "epoch": 8.749919577945056,
      "grad_norm": 0.21583206951618195,
      "learning_rate": 0.0006272215448612208,
      "loss": 2.6192,
      "step": 136000
    },
    {
      "epoch": 8.814257221900533,
      "grad_norm": 0.2255946397781372,
      "learning_rate": 0.0006236355160295489,
      "loss": 2.6274,
      "step": 137000
    },
    {
      "epoch": 8.878594865856012,
      "grad_norm": 0.24251054227352142,
      "learning_rate": 0.0006200530732267087,
      "loss": 2.6293,
      "step": 138000
    },
    {
      "epoch": 8.942932509811492,
      "grad_norm": 0.2000625729560852,
      "learning_rate": 0.000616467044395037,
      "loss": 2.6242,
      "step": 139000
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.42548870043822257,
      "eval_loss": 2.944140911102295,
      "eval_runtime": 112.9581,
      "eval_samples_per_second": 464.243,
      "eval_steps_per_second": 7.259,
      "step": 139887
    },
    {
      "epoch": 9.007270153766969,
      "grad_norm": 0.2379744052886963,
      "learning_rate": 0.0006128810155633652,
      "loss": 2.6206,
      "step": 140000
    },
    {
      "epoch": 9.071607797722447,
      "grad_norm": 0.30487170815467834,
      "learning_rate": 0.0006092985727605251,
      "loss": 2.5374,
      "step": 141000
    },
    {
      "epoch": 9.135945441677926,
      "grad_norm": 0.2142684906721115,
      "learning_rate": 0.0006057125439288532,
      "loss": 2.5544,
      "step": 142000
    },
    {
      "epoch": 9.200283085633403,
      "grad_norm": 0.21194174885749817,
      "learning_rate": 0.0006021265150971814,
      "loss": 2.56,
      "step": 143000
    },
    {
      "epoch": 9.264620729588882,
      "grad_norm": 0.2444310337305069,
      "learning_rate": 0.0005985440722943413,
      "loss": 2.5591,
      "step": 144000
    },
    {
      "epoch": 9.32895837354436,
      "grad_norm": 0.2151477187871933,
      "learning_rate": 0.0005949580434626695,
      "loss": 2.5717,
      "step": 145000
    },
    {
      "epoch": 9.39329601749984,
      "grad_norm": 0.20657171308994293,
      "learning_rate": 0.000591379186688661,
      "loss": 2.5773,
      "step": 146000
    },
    {
      "epoch": 9.457633661455317,
      "grad_norm": 0.23294700682163239,
      "learning_rate": 0.0005877931578569892,
      "loss": 2.5796,
      "step": 147000
    },
    {
      "epoch": 9.521971305410796,
      "grad_norm": 0.22148990631103516,
      "learning_rate": 0.0005842071290253173,
      "loss": 2.5794,
      "step": 148000
    },
    {
      "epoch": 9.586308949366273,
      "grad_norm": 0.23063485324382782,
      "learning_rate": 0.0005806211001936455,
      "loss": 2.5883,
      "step": 149000
    },
    {
      "epoch": 9.650646593321753,
      "grad_norm": 0.2963120937347412,
      "learning_rate": 0.0005770350713619738,
      "loss": 2.591,
      "step": 150000
    },
    {
      "epoch": 9.71498423727723,
      "grad_norm": 0.22468186914920807,
      "learning_rate": 0.0005734526285591336,
      "loss": 2.59,
      "step": 151000
    },
    {
      "epoch": 9.77932188123271,
      "grad_norm": 0.2075439840555191,
      "learning_rate": 0.0005698665997274618,
      "loss": 2.5922,
      "step": 152000
    },
    {
      "epoch": 9.843659525188187,
      "grad_norm": 0.26252982020378113,
      "learning_rate": 0.00056628057089579,
      "loss": 2.5897,
      "step": 153000
    },
    {
      "epoch": 9.907997169143666,
      "grad_norm": 0.23882432281970978,
      "learning_rate": 0.0005626945420641182,
      "loss": 2.6002,
      "step": 154000
    },
    {
      "epoch": 9.972334813099144,
      "grad_norm": 0.23926019668579102,
      "learning_rate": 0.0005591156852901097,
      "loss": 2.5991,
      "step": 155000
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.42581489956776,
      "eval_loss": 2.9460246562957764,
      "eval_runtime": 113.0461,
      "eval_samples_per_second": 463.882,
      "eval_steps_per_second": 7.254,
      "step": 155430
    },
    {
      "epoch": 10.036672457054623,
      "grad_norm": 0.2680639326572418,
      "learning_rate": 0.0005555296564584379,
      "loss": 2.5553,
      "step": 156000
    },
    {
      "epoch": 10.1010101010101,
      "grad_norm": 0.2323896884918213,
      "learning_rate": 0.000551943627626766,
      "loss": 2.5132,
      "step": 157000
    },
    {
      "epoch": 10.16534774496558,
      "grad_norm": 0.22753171622753143,
      "learning_rate": 0.0005483611848239259,
      "loss": 2.522,
      "step": 158000
    },
    {
      "epoch": 10.229685388921057,
      "grad_norm": 0.27533870935440063,
      "learning_rate": 0.000544778742021086,
      "loss": 2.531,
      "step": 159000
    },
    {
      "epoch": 10.294023032876536,
      "grad_norm": 0.23533160984516144,
      "learning_rate": 0.0005411927131894141,
      "loss": 2.5402,
      "step": 160000
    },
    {
      "epoch": 10.358360676832014,
      "grad_norm": 0.2618332505226135,
      "learning_rate": 0.0005376066843577423,
      "loss": 2.5426,
      "step": 161000
    },
    {
      "epoch": 10.422698320787493,
      "grad_norm": 0.23586297035217285,
      "learning_rate": 0.0005340206555260705,
      "loss": 2.5444,
      "step": 162000
    },
    {
      "epoch": 10.48703596474297,
      "grad_norm": 0.24014532566070557,
      "learning_rate": 0.0005304382127232304,
      "loss": 2.5485,
      "step": 163000
    },
    {
      "epoch": 10.55137360869845,
      "grad_norm": 0.23357678949832916,
      "learning_rate": 0.0005268521838915585,
      "loss": 2.553,
      "step": 164000
    },
    {
      "epoch": 10.615711252653927,
      "grad_norm": 0.22992868721485138,
      "learning_rate": 0.0005232661550598867,
      "loss": 2.552,
      "step": 165000
    },
    {
      "epoch": 10.680048896609406,
      "grad_norm": 0.2423822432756424,
      "learning_rate": 0.000519680126228215,
      "loss": 2.5605,
      "step": 166000
    },
    {
      "epoch": 10.744386540564884,
      "grad_norm": 0.2182675153017044,
      "learning_rate": 0.0005160976834253748,
      "loss": 2.5612,
      "step": 167000
    },
    {
      "epoch": 10.808724184520363,
      "grad_norm": 0.21750390529632568,
      "learning_rate": 0.0005125116545937029,
      "loss": 2.5638,
      "step": 168000
    },
    {
      "epoch": 10.87306182847584,
      "grad_norm": 0.22003532946109772,
      "learning_rate": 0.0005089292117908628,
      "loss": 2.5725,
      "step": 169000
    },
    {
      "epoch": 10.93739947243132,
      "grad_norm": 0.26017525792121887,
      "learning_rate": 0.000505343182959191,
      "loss": 2.5731,
      "step": 170000
    },
    {
      "epoch": 11.0,
      "eval_accuracy": 0.42594449679185176,
      "eval_loss": 2.949589490890503,
      "eval_runtime": 112.9982,
      "eval_samples_per_second": 464.078,
      "eval_steps_per_second": 7.257,
      "step": 170973
    },
    {
      "epoch": 11.001737116386797,
      "grad_norm": 0.25224384665489197,
      "learning_rate": 0.0005017607401563509,
      "loss": 2.5678,
      "step": 171000
    },
    {
      "epoch": 11.066074760342277,
      "grad_norm": 0.24872563779354095,
      "learning_rate": 0.0004981747113246791,
      "loss": 2.4739,
      "step": 172000
    },
    {
      "epoch": 11.130412404297754,
      "grad_norm": 0.24027609825134277,
      "learning_rate": 0.000494592268521839,
      "loss": 2.4942,
      "step": 173000
    },
    {
      "epoch": 11.194750048253233,
      "grad_norm": 0.26473268866539,
      "learning_rate": 0.0004910098257189988,
      "loss": 2.5027,
      "step": 174000
    },
    {
      "epoch": 11.25908769220871,
      "grad_norm": 0.2208300679922104,
      "learning_rate": 0.00048742379688732697,
      "loss": 2.5059,
      "step": 175000
    },
    {
      "epoch": 11.32342533616419,
      "grad_norm": 0.26882168650627136,
      "learning_rate": 0.00048383776805565514,
      "loss": 2.5106,
      "step": 176000
    },
    {
      "epoch": 11.387762980119668,
      "grad_norm": 0.2618742287158966,
      "learning_rate": 0.00048025173922398336,
      "loss": 2.5153,
      "step": 177000
    },
    {
      "epoch": 11.452100624075147,
      "grad_norm": 0.2615552842617035,
      "learning_rate": 0.0004766657103923115,
      "loss": 2.5264,
      "step": 178000
    },
    {
      "epoch": 11.516438268030624,
      "grad_norm": 0.2439650595188141,
      "learning_rate": 0.00047307968156063974,
      "loss": 2.5245,
      "step": 179000
    },
    {
      "epoch": 11.580775911986104,
      "grad_norm": 0.24545331299304962,
      "learning_rate": 0.00046949723875779964,
      "loss": 2.5316,
      "step": 180000
    },
    {
      "epoch": 11.645113555941581,
      "grad_norm": 0.2467048317193985,
      "learning_rate": 0.0004659183819837912,
      "loss": 2.5272,
      "step": 181000
    },
    {
      "epoch": 11.70945119989706,
      "grad_norm": 0.23762260377407074,
      "learning_rate": 0.00046233235315211934,
      "loss": 2.5375,
      "step": 182000
    },
    {
      "epoch": 11.773788843852538,
      "grad_norm": 0.26846447587013245,
      "learning_rate": 0.00045874632432044756,
      "loss": 2.5393,
      "step": 183000
    },
    {
      "epoch": 11.838126487808017,
      "grad_norm": 0.2588064968585968,
      "learning_rate": 0.0004551602954887757,
      "loss": 2.5375,
      "step": 184000
    },
    {
      "epoch": 11.902464131763494,
      "grad_norm": 0.23051457107067108,
      "learning_rate": 0.00045157426665710395,
      "loss": 2.5374,
      "step": 185000
    },
    {
      "epoch": 11.966801775718974,
      "grad_norm": 0.23640917241573334,
      "learning_rate": 0.0004479882378254321,
      "loss": 2.5387,
      "step": 186000
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.425906432748538,
      "eval_loss": 2.9570860862731934,
      "eval_runtime": 112.6715,
      "eval_samples_per_second": 465.424,
      "eval_steps_per_second": 7.278,
      "step": 186516
    },
    {
      "epoch": 12.031139419674451,
      "grad_norm": 0.3415308892726898,
      "learning_rate": 0.00044440579502259196,
      "loss": 2.5002,
      "step": 187000
    },
    {
      "epoch": 12.09547706362993,
      "grad_norm": 0.2506396472454071,
      "learning_rate": 0.0004408197661909202,
      "loss": 2.4645,
      "step": 188000
    },
    {
      "epoch": 12.159814707585408,
      "grad_norm": 0.32716119289398193,
      "learning_rate": 0.0004372337373592484,
      "loss": 2.4685,
      "step": 189000
    },
    {
      "epoch": 12.224152351540887,
      "grad_norm": 0.2909821569919586,
      "learning_rate": 0.00043364770852757657,
      "loss": 2.4784,
      "step": 190000
    },
    {
      "epoch": 12.288489995496365,
      "grad_norm": 0.24312670528888702,
      "learning_rate": 0.0004300688517535681,
      "loss": 2.4878,
      "step": 191000
    },
    {
      "epoch": 12.352827639451844,
      "grad_norm": 0.24336646497249603,
      "learning_rate": 0.00042648282292189626,
      "loss": 2.4843,
      "step": 192000
    },
    {
      "epoch": 12.417165283407321,
      "grad_norm": 0.26876741647720337,
      "learning_rate": 0.0004228967940902245,
      "loss": 2.4887,
      "step": 193000
    },
    {
      "epoch": 12.4815029273628,
      "grad_norm": 0.3456116318702698,
      "learning_rate": 0.0004193143512873844,
      "loss": 2.4934,
      "step": 194000
    },
    {
      "epoch": 12.545840571318278,
      "grad_norm": 0.24459022283554077,
      "learning_rate": 0.00041573190848454423,
      "loss": 2.4995,
      "step": 195000
    },
    {
      "epoch": 12.610178215273757,
      "grad_norm": 0.236145481467247,
      "learning_rate": 0.0004121458796528724,
      "loss": 2.5034,
      "step": 196000
    },
    {
      "epoch": 12.674515859229235,
      "grad_norm": 0.26968687772750854,
      "learning_rate": 0.0004085598508212006,
      "loss": 2.5049,
      "step": 197000
    },
    {
      "epoch": 12.738853503184714,
      "grad_norm": 0.24692216515541077,
      "learning_rate": 0.00040497382198952884,
      "loss": 2.5082,
      "step": 198000
    },
    {
      "epoch": 12.803191147140192,
      "grad_norm": 0.2576483488082886,
      "learning_rate": 0.000401387793157857,
      "loss": 2.5093,
      "step": 199000
    },
    {
      "epoch": 12.86752879109567,
      "grad_norm": 0.24602724611759186,
      "learning_rate": 0.00039780535035501685,
      "loss": 2.5098,
      "step": 200000
    },
    {
      "epoch": 12.931866435051148,
      "grad_norm": 0.22797554731369019,
      "learning_rate": 0.0003942229075521767,
      "loss": 2.5143,
      "step": 201000
    },
    {
      "epoch": 12.996204079006628,
      "grad_norm": 0.23557151854038239,
      "learning_rate": 0.0003906368787205049,
      "loss": 2.5106,
      "step": 202000
    },
    {
      "epoch": 13.0,
      "eval_accuracy": 0.42556520243490226,
      "eval_loss": 2.9726738929748535,
      "eval_runtime": 112.8886,
      "eval_samples_per_second": 464.529,
      "eval_steps_per_second": 7.264,
      "step": 202059
    },
    {
      "epoch": 13.060541722962105,
      "grad_norm": 0.2934684455394745,
      "learning_rate": 0.0003870508498888331,
      "loss": 2.438,
      "step": 203000
    },
    {
      "epoch": 13.124879366917584,
      "grad_norm": 0.266162633895874,
      "learning_rate": 0.0003834648210571613,
      "loss": 2.4391,
      "step": 204000
    },
    {
      "epoch": 13.189217010873062,
      "grad_norm": 0.2488025277853012,
      "learning_rate": 0.00037988237825432115,
      "loss": 2.4474,
      "step": 205000
    },
    {
      "epoch": 13.253554654828541,
      "grad_norm": 0.2948077917098999,
      "learning_rate": 0.0003762963494226493,
      "loss": 2.4527,
      "step": 206000
    },
    {
      "epoch": 13.317892298784018,
      "grad_norm": 0.2885175049304962,
      "learning_rate": 0.00037271032059097754,
      "loss": 2.466,
      "step": 207000
    },
    {
      "epoch": 13.382229942739498,
      "grad_norm": 0.2502969801425934,
      "learning_rate": 0.00036912787778813744,
      "loss": 2.4656,
      "step": 208000
    },
    {
      "epoch": 13.446567586694975,
      "grad_norm": 0.26107537746429443,
      "learning_rate": 0.00036554184895646566,
      "loss": 2.4676,
      "step": 209000
    },
    {
      "epoch": 13.510905230650454,
      "grad_norm": 0.2713301181793213,
      "learning_rate": 0.0003619594061536255,
      "loss": 2.4736,
      "step": 210000
    },
    {
      "epoch": 13.575242874605932,
      "grad_norm": 0.2778390347957611,
      "learning_rate": 0.0003583733773219537,
      "loss": 2.4732,
      "step": 211000
    },
    {
      "epoch": 13.63958051856141,
      "grad_norm": 0.27868494391441345,
      "learning_rate": 0.0003547873484902819,
      "loss": 2.4785,
      "step": 212000
    },
    {
      "epoch": 13.703918162516889,
      "grad_norm": 0.24166850745677948,
      "learning_rate": 0.00035120131965861006,
      "loss": 2.4829,
      "step": 213000
    },
    {
      "epoch": 13.768255806472368,
      "grad_norm": 0.26391276717185974,
      "learning_rate": 0.0003476152908269383,
      "loss": 2.4861,
      "step": 214000
    },
    {
      "epoch": 13.832593450427845,
      "grad_norm": 0.2995707392692566,
      "learning_rate": 0.00034403284802409813,
      "loss": 2.4851,
      "step": 215000
    },
    {
      "epoch": 13.896931094383323,
      "grad_norm": 0.2664172947406769,
      "learning_rate": 0.0003404468191924263,
      "loss": 2.4851,
      "step": 216000
    },
    {
      "epoch": 13.961268738338802,
      "grad_norm": 0.2712934613227844,
      "learning_rate": 0.0003368607903607545,
      "loss": 2.4846,
      "step": 217000
    },
    {
      "epoch": 14.0,
      "eval_accuracy": 0.4257042221923094,
      "eval_loss": 2.9749884605407715,
      "eval_runtime": 113.0113,
      "eval_samples_per_second": 464.024,
      "eval_steps_per_second": 7.256,
      "step": 217602
    },
    {
      "epoch": 14.02560638229428,
      "grad_norm": 0.2511956989765167,
      "learning_rate": 0.00033327834755791436,
      "loss": 2.4623,
      "step": 218000
    },
    {
      "epoch": 14.089944026249759,
      "grad_norm": 0.2726948857307434,
      "learning_rate": 0.0003296923187262426,
      "loss": 2.42,
      "step": 219000
    },
    {
      "epoch": 14.154281670205236,
      "grad_norm": 0.2592310309410095,
      "learning_rate": 0.00032610987592340243,
      "loss": 2.4247,
      "step": 220000
    },
    {
      "epoch": 14.218619314160716,
      "grad_norm": 0.24364928901195526,
      "learning_rate": 0.0003225238470917306,
      "loss": 2.4293,
      "step": 221000
    },
    {
      "epoch": 14.282956958116193,
      "grad_norm": 0.26270249485969543,
      "learning_rate": 0.0003189414042888905,
      "loss": 2.43,
      "step": 222000
    },
    {
      "epoch": 14.347294602071672,
      "grad_norm": 0.24876274168491364,
      "learning_rate": 0.0003153553754572187,
      "loss": 2.4434,
      "step": 223000
    },
    {
      "epoch": 14.41163224602715,
      "grad_norm": 0.27520838379859924,
      "learning_rate": 0.0003117693466255469,
      "loss": 2.4427,
      "step": 224000
    },
    {
      "epoch": 14.475969889982629,
      "grad_norm": 0.28508856892585754,
      "learning_rate": 0.0003081833177938751,
      "loss": 2.4431,
      "step": 225000
    },
    {
      "epoch": 14.540307533938106,
      "grad_norm": 0.26081016659736633,
      "learning_rate": 0.00030460087499103495,
      "loss": 2.4414,
      "step": 226000
    },
    {
      "epoch": 14.604645177893586,
      "grad_norm": 0.24568462371826172,
      "learning_rate": 0.0003010184321881948,
      "loss": 2.455,
      "step": 227000
    },
    {
      "epoch": 14.668982821849063,
      "grad_norm": 0.2919858694076538,
      "learning_rate": 0.000297432403356523,
      "loss": 2.4528,
      "step": 228000
    },
    {
      "epoch": 14.733320465804542,
      "grad_norm": 0.2552844285964966,
      "learning_rate": 0.00029384996055368287,
      "loss": 2.4591,
      "step": 229000
    },
    {
      "epoch": 14.79765810976002,
      "grad_norm": 0.24218083918094635,
      "learning_rate": 0.00029026393172201104,
      "loss": 2.4593,
      "step": 230000
    },
    {
      "epoch": 14.8619957537155,
      "grad_norm": 0.2795206904411316,
      "learning_rate": 0.00028667790289033926,
      "loss": 2.4562,
      "step": 231000
    },
    {
      "epoch": 14.926333397670977,
      "grad_norm": 0.24688765406608582,
      "learning_rate": 0.0002830918740586674,
      "loss": 2.4526,
      "step": 232000
    },
    {
      "epoch": 14.990671041626456,
      "grad_norm": 0.26866546273231506,
      "learning_rate": 0.00027950584522699564,
      "loss": 2.4638,
      "step": 233000
    },
    {
      "epoch": 15.0,
      "eval_accuracy": 0.4250153303121401,
      "eval_loss": 2.985236644744873,
      "eval_runtime": 112.77,
      "eval_samples_per_second": 465.017,
      "eval_steps_per_second": 7.271,
      "step": 233145
    },
    {
      "epoch": 15.055008685581933,
      "grad_norm": 0.2907277047634125,
      "learning_rate": 0.0002759198163953238,
      "loss": 2.4022,
      "step": 234000
    },
    {
      "epoch": 15.119346329537413,
      "grad_norm": 0.28493550419807434,
      "learning_rate": 0.0002723409596213154,
      "loss": 2.3966,
      "step": 235000
    },
    {
      "epoch": 15.18368397349289,
      "grad_norm": 0.2639086842536926,
      "learning_rate": 0.00026875493078964356,
      "loss": 2.4034,
      "step": 236000
    },
    {
      "epoch": 15.24802161744837,
      "grad_norm": 0.2657439708709717,
      "learning_rate": 0.0002651689019579717,
      "loss": 2.4088,
      "step": 237000
    },
    {
      "epoch": 15.312359261403847,
      "grad_norm": 0.2811715602874756,
      "learning_rate": 0.0002615828731262999,
      "loss": 2.416,
      "step": 238000
    },
    {
      "epoch": 15.376696905359326,
      "grad_norm": 0.2565077245235443,
      "learning_rate": 0.00025800043032345985,
      "loss": 2.4167,
      "step": 239000
    },
    {
      "epoch": 15.441034549314804,
      "grad_norm": 0.2618282735347748,
      "learning_rate": 0.000254414401491788,
      "loss": 2.4197,
      "step": 240000
    },
    {
      "epoch": 15.505372193270283,
      "grad_norm": 0.33380481600761414,
      "learning_rate": 0.00025083195868894786,
      "loss": 2.4218,
      "step": 241000
    },
    {
      "epoch": 15.56970983722576,
      "grad_norm": 0.28201743960380554,
      "learning_rate": 0.0002472459298572761,
      "loss": 2.4304,
      "step": 242000
    },
    {
      "epoch": 15.63404748118124,
      "grad_norm": 0.24747717380523682,
      "learning_rate": 0.00024366348705443593,
      "loss": 2.4276,
      "step": 243000
    },
    {
      "epoch": 15.698385125136717,
      "grad_norm": 0.31805720925331116,
      "learning_rate": 0.00024007745822276412,
      "loss": 2.4288,
      "step": 244000
    },
    {
      "epoch": 15.762722769092196,
      "grad_norm": 0.28187525272369385,
      "learning_rate": 0.0002364914293910923,
      "loss": 2.439,
      "step": 245000
    },
    {
      "epoch": 15.827060413047674,
      "grad_norm": 0.26413655281066895,
      "learning_rate": 0.0002329089865882522,
      "loss": 2.4329,
      "step": 246000
    },
    {
      "epoch": 15.891398057003153,
      "grad_norm": 0.28774192929267883,
      "learning_rate": 0.00022932295775658038,
      "loss": 2.436,
      "step": 247000
    },
    {
      "epoch": 15.95573570095863,
      "grad_norm": 0.3739126920700073,
      "learning_rate": 0.00022574051495374023,
      "loss": 2.4351,
      "step": 248000
    },
    {
      "epoch": 16.0,
      "eval_accuracy": 0.42470191890638787,
      "eval_loss": 2.996845006942749,
      "eval_runtime": 112.9222,
      "eval_samples_per_second": 464.39,
      "eval_steps_per_second": 7.262,
      "step": 248688
    },
    {
      "epoch": 16.02007334491411,
      "grad_norm": 0.2598513662815094,
      "learning_rate": 0.00022215448612206842,
      "loss": 2.4157,
      "step": 249000
    },
    {
      "epoch": 16.08441098886959,
      "grad_norm": 0.2784959673881531,
      "learning_rate": 0.00021856845729039662,
      "loss": 2.3753,
      "step": 250000
    },
    {
      "epoch": 16.148748632825065,
      "grad_norm": 0.26741138100624084,
      "learning_rate": 0.00021498601448755646,
      "loss": 2.383,
      "step": 251000
    },
    {
      "epoch": 16.213086276780544,
      "grad_norm": 0.2786239981651306,
      "learning_rate": 0.00021139998565588466,
      "loss": 2.3848,
      "step": 252000
    },
    {
      "epoch": 16.277423920736023,
      "grad_norm": 0.2728557586669922,
      "learning_rate": 0.00020781395682421288,
      "loss": 2.3863,
      "step": 253000
    },
    {
      "epoch": 16.341761564691502,
      "grad_norm": 0.2834782898426056,
      "learning_rate": 0.00020422792799254107,
      "loss": 2.3946,
      "step": 254000
    },
    {
      "epoch": 16.406099208646978,
      "grad_norm": 0.2955917418003082,
      "learning_rate": 0.0002006490712185326,
      "loss": 2.3936,
      "step": 255000
    },
    {
      "epoch": 16.470436852602457,
      "grad_norm": 0.27384132146835327,
      "learning_rate": 0.0001970630423868608,
      "loss": 2.3991,
      "step": 256000
    },
    {
      "epoch": 16.534774496557937,
      "grad_norm": 0.2943706512451172,
      "learning_rate": 0.00019347701355518899,
      "loss": 2.4013,
      "step": 257000
    },
    {
      "epoch": 16.599112140513416,
      "grad_norm": 0.263576477766037,
      "learning_rate": 0.00018989457075234886,
      "loss": 2.4069,
      "step": 258000
    },
    {
      "epoch": 16.66344978446889,
      "grad_norm": 0.2836902141571045,
      "learning_rate": 0.00018630854192067703,
      "loss": 2.4024,
      "step": 259000
    },
    {
      "epoch": 16.72778742842437,
      "grad_norm": 0.30342864990234375,
      "learning_rate": 0.00018272251308900522,
      "loss": 2.4049,
      "step": 260000
    },
    {
      "epoch": 16.79212507237985,
      "grad_norm": 0.28382593393325806,
      "learning_rate": 0.0001791400702861651,
      "loss": 2.4104,
      "step": 261000
    },
    {
      "epoch": 16.85646271633533,
      "grad_norm": 0.30956143140792847,
      "learning_rate": 0.00017555404145449331,
      "loss": 2.4126,
      "step": 262000
    },
    {
      "epoch": 16.920800360290805,
      "grad_norm": 0.29086652398109436,
      "learning_rate": 0.0001719680126228215,
      "loss": 2.4128,
      "step": 263000
    },
    {
      "epoch": 16.985138004246284,
      "grad_norm": 0.29363515973091125,
      "learning_rate": 0.0001683819837911497,
      "loss": 2.4119,
      "step": 264000
    },
    {
      "epoch": 17.0,
      "eval_accuracy": 0.42441498033233127,
      "eval_loss": 3.0043680667877197,
      "eval_runtime": 112.7018,
      "eval_samples_per_second": 465.299,
      "eval_steps_per_second": 7.276,
      "step": 264231
    },
    {
      "epoch": 17.049475648201764,
      "grad_norm": 0.2627195119857788,
      "learning_rate": 0.00016479954098830955,
      "loss": 2.3717,
      "step": 265000
    },
    {
      "epoch": 17.113813292157243,
      "grad_norm": 0.31472471356391907,
      "learning_rate": 0.00016121709818546942,
      "loss": 2.3624,
      "step": 266000
    },
    {
      "epoch": 17.17815093611272,
      "grad_norm": 0.30606958270072937,
      "learning_rate": 0.00015763106935379762,
      "loss": 2.3688,
      "step": 267000
    },
    {
      "epoch": 17.242488580068198,
      "grad_norm": 0.3092518448829651,
      "learning_rate": 0.00015404862655095746,
      "loss": 2.3707,
      "step": 268000
    },
    {
      "epoch": 17.306826224023677,
      "grad_norm": 0.29389309883117676,
      "learning_rate": 0.00015046259771928566,
      "loss": 2.3684,
      "step": 269000
    },
    {
      "epoch": 17.371163867979156,
      "grad_norm": 0.26194649934768677,
      "learning_rate": 0.00014687656888761385,
      "loss": 2.3751,
      "step": 270000
    },
    {
      "epoch": 17.435501511934632,
      "grad_norm": 0.32368776202201843,
      "learning_rate": 0.00014329054005594204,
      "loss": 2.3746,
      "step": 271000
    },
    {
      "epoch": 17.49983915589011,
      "grad_norm": 0.2756269872188568,
      "learning_rate": 0.00013970809725310192,
      "loss": 2.376,
      "step": 272000
    },
    {
      "epoch": 17.56417679984559,
      "grad_norm": 0.32635217905044556,
      "learning_rate": 0.0001361220684214301,
      "loss": 2.3872,
      "step": 273000
    },
    {
      "epoch": 17.62851444380107,
      "grad_norm": 0.31273502111434937,
      "learning_rate": 0.00013253962561859,
      "loss": 2.3812,
      "step": 274000
    },
    {
      "epoch": 17.692852087756545,
      "grad_norm": 0.28867122530937195,
      "learning_rate": 0.00012895359678691818,
      "loss": 2.3852,
      "step": 275000
    },
    {
      "epoch": 17.757189731712025,
      "grad_norm": 0.2799566984176636,
      "learning_rate": 0.00012536756795524637,
      "loss": 2.3825,
      "step": 276000
    },
    {
      "epoch": 17.821527375667504,
      "grad_norm": 0.2807822525501251,
      "learning_rate": 0.00012178153912357455,
      "loss": 2.3795,
      "step": 277000
    },
    {
      "epoch": 17.885865019622983,
      "grad_norm": 0.2853219211101532,
      "learning_rate": 0.00011819909632073441,
      "loss": 2.3846,
      "step": 278000
    },
    {
      "epoch": 17.95020266357846,
      "grad_norm": 0.27752065658569336,
      "learning_rate": 0.0001146130674890626,
      "loss": 2.3894,
      "step": 279000
    },
    {
      "epoch": 18.0,
      "eval_accuracy": 0.4241119636260301,
      "eval_loss": 3.013193130493164,
      "eval_runtime": 112.7432,
      "eval_samples_per_second": 465.128,
      "eval_steps_per_second": 7.273,
      "step": 279774
    },
    {
      "epoch": 18.014540307533938,
      "grad_norm": 0.30354997515678406,
      "learning_rate": 0.00011102703865739081,
      "loss": 2.3766,
      "step": 280000
    },
    {
      "epoch": 18.078877951489417,
      "grad_norm": 0.29635682702064514,
      "learning_rate": 0.00010744459585455067,
      "loss": 2.3432,
      "step": 281000
    },
    {
      "epoch": 18.143215595444893,
      "grad_norm": 0.36005303263664246,
      "learning_rate": 0.00010385856702287887,
      "loss": 2.3509,
      "step": 282000
    },
    {
      "epoch": 18.207553239400372,
      "grad_norm": 0.2774285674095154,
      "learning_rate": 0.00010027253819120706,
      "loss": 2.3499,
      "step": 283000
    },
    {
      "epoch": 18.27189088335585,
      "grad_norm": 0.29461967945098877,
      "learning_rate": 9.669009538836692e-05,
      "loss": 2.3504,
      "step": 284000
    },
    {
      "epoch": 18.33622852731133,
      "grad_norm": 0.29665911197662354,
      "learning_rate": 9.310406655669513e-05,
      "loss": 2.3559,
      "step": 285000
    },
    {
      "epoch": 18.400566171266806,
      "grad_norm": 0.30571022629737854,
      "learning_rate": 8.952162375385499e-05,
      "loss": 2.3548,
      "step": 286000
    },
    {
      "epoch": 18.464903815222286,
      "grad_norm": 0.2917598783969879,
      "learning_rate": 8.593559492218318e-05,
      "loss": 2.3527,
      "step": 287000
    },
    {
      "epoch": 18.529241459177765,
      "grad_norm": 0.2724420130252838,
      "learning_rate": 8.235315211934303e-05,
      "loss": 2.3615,
      "step": 288000
    },
    {
      "epoch": 18.593579103133244,
      "grad_norm": 0.2798316776752472,
      "learning_rate": 7.876712328767124e-05,
      "loss": 2.3656,
      "step": 289000
    },
    {
      "epoch": 18.65791674708872,
      "grad_norm": 0.31976842880249023,
      "learning_rate": 7.518109445599943e-05,
      "loss": 2.3525,
      "step": 290000
    },
    {
      "epoch": 18.7222543910442,
      "grad_norm": 0.28387773036956787,
      "learning_rate": 7.159865165315929e-05,
      "loss": 2.3598,
      "step": 291000
    },
    {
      "epoch": 18.78659203499968,
      "grad_norm": 0.28037527203559875,
      "learning_rate": 6.801262282148748e-05,
      "loss": 2.3658,
      "step": 292000
    },
    {
      "epoch": 18.850929678955158,
      "grad_norm": 0.2663221061229706,
      "learning_rate": 6.443018001864735e-05,
      "loss": 2.3658,
      "step": 293000
    },
    {
      "epoch": 18.915267322910633,
      "grad_norm": 0.270049124956131,
      "learning_rate": 6.0844151186975546e-05,
      "loss": 2.3656,
      "step": 294000
    },
    {
      "epoch": 18.979604966866113,
      "grad_norm": 0.26364874839782715,
      "learning_rate": 5.725812235530374e-05,
      "loss": 2.3569,
      "step": 295000
    },
    {
      "epoch": 19.0,
      "eval_accuracy": 0.4237278832204125,
      "eval_loss": 3.027001142501831,
      "eval_runtime": 112.7341,
      "eval_samples_per_second": 465.165,
      "eval_steps_per_second": 7.274,
      "step": 295317
    },
    {
      "epoch": 19.043942610821592,
      "grad_norm": 0.29353857040405273,
      "learning_rate": 5.367209352363193e-05,
      "loss": 2.3409,
      "step": 296000
    },
    {
      "epoch": 19.10828025477707,
      "grad_norm": 0.2566602826118469,
      "learning_rate": 5.008965072079179e-05,
      "loss": 2.3371,
      "step": 297000
    },
    {
      "epoch": 19.172617898732547,
      "grad_norm": 0.2797909677028656,
      "learning_rate": 4.650362188911999e-05,
      "loss": 2.3365,
      "step": 298000
    },
    {
      "epoch": 19.236955542688026,
      "grad_norm": 0.32513269782066345,
      "learning_rate": 4.2921179086279854e-05,
      "loss": 2.3378,
      "step": 299000
    },
    {
      "epoch": 19.301293186643505,
      "grad_norm": 0.3171839416027069,
      "learning_rate": 3.933873628343972e-05,
      "loss": 2.3325,
      "step": 300000
    },
    {
      "epoch": 19.365630830598985,
      "grad_norm": 0.3085026443004608,
      "learning_rate": 3.575270745176791e-05,
      "loss": 2.3356,
      "step": 301000
    },
    {
      "epoch": 19.42996847455446,
      "grad_norm": 0.32660767436027527,
      "learning_rate": 3.21666786200961e-05,
      "loss": 2.3407,
      "step": 302000
    },
    {
      "epoch": 19.49430611850994,
      "grad_norm": 0.3034733235836029,
      "learning_rate": 2.8580649788424298e-05,
      "loss": 2.3404,
      "step": 303000
    },
    {
      "epoch": 19.55864376246542,
      "grad_norm": 0.277503103017807,
      "learning_rate": 2.4998206985584166e-05,
      "loss": 2.3407,
      "step": 304000
    },
    {
      "epoch": 19.622981406420898,
      "grad_norm": 0.3372737467288971,
      "learning_rate": 2.141217815391236e-05,
      "loss": 2.339,
      "step": 305000
    },
    {
      "epoch": 19.687319050376374,
      "grad_norm": 0.3079584240913391,
      "learning_rate": 1.7826149322240552e-05,
      "loss": 2.3399,
      "step": 306000
    },
    {
      "epoch": 19.751656694331853,
      "grad_norm": 0.2789517939090729,
      "learning_rate": 1.4243706519400415e-05,
      "loss": 2.3396,
      "step": 307000
    },
    {
      "epoch": 19.815994338287332,
      "grad_norm": 0.2706036865711212,
      "learning_rate": 1.065767768772861e-05,
      "loss": 2.34,
      "step": 308000
    },
    {
      "epoch": 19.88033198224281,
      "grad_norm": 0.2812829911708832,
      "learning_rate": 7.075234884888475e-06,
      "loss": 2.3404,
      "step": 309000
    },
    {
      "epoch": 19.944669626198287,
      "grad_norm": 0.2975556254386902,
      "learning_rate": 3.489206053216668e-06,
      "loss": 2.3375,
      "step": 310000
    },
    {
      "epoch": 20.0,
      "eval_accuracy": 0.4232865945768086,
      "eval_loss": 3.037959337234497,
      "eval_runtime": 112.7925,
      "eval_samples_per_second": 464.925,
      "eval_steps_per_second": 7.27,
      "step": 310860
    },
    {
      "epoch": 20.0,
      "step": 310860,
      "total_flos": 1.29957250203648e+18,
      "train_loss": 2.656636161733025,
      "train_runtime": 44071.5037,
      "train_samples_per_second": 225.708,
      "train_steps_per_second": 7.054
    }
  ],
  "logging_steps": 1000,
  "max_steps": 310860,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 20,
  "save_steps": 5000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.29957250203648e+18,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}