{
  "best_global_step": 50000,
  "best_metric": 0.5460931658744812,
  "best_model_checkpoint": "checkpoints/mla-o_baseline/checkpoint-50000",
  "epoch": 7.105300554213443,
  "eval_steps": 2000,
  "global_step": 50000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0071053005542134435,
      "grad_norm": 0.3552660346031189,
      "learning_rate": 4.9e-06,
      "loss": 10.3029,
      "step": 50
    },
    {
      "epoch": 0.014210601108426887,
      "grad_norm": 0.5037913918495178,
      "learning_rate": 9.900000000000002e-06,
      "loss": 10.2361,
      "step": 100
    },
    {
      "epoch": 0.02131590166264033,
      "grad_norm": 1.8683347702026367,
      "learning_rate": 1.49e-05,
      "loss": 9.9763,
      "step": 150
    },
    {
      "epoch": 0.028421202216853774,
      "grad_norm": 4.1887102127075195,
      "learning_rate": 1.9900000000000003e-05,
      "loss": 8.7258,
      "step": 200
    },
    {
      "epoch": 0.03552650277106722,
      "grad_norm": 1.4753533601760864,
      "learning_rate": 2.49e-05,
      "loss": 7.5735,
      "step": 250
    },
    {
      "epoch": 0.04263180332528066,
      "grad_norm": 1.4821197986602783,
      "learning_rate": 2.9900000000000002e-05,
      "loss": 7.3076,
      "step": 300
    },
    {
      "epoch": 0.0497371038794941,
      "grad_norm": 1.4782558679580688,
      "learning_rate": 3.49e-05,
      "loss": 7.2083,
      "step": 350
    },
    {
      "epoch": 0.05684240443370755,
      "grad_norm": 1.6048325300216675,
      "learning_rate": 3.99e-05,
      "loss": 7.1503,
      "step": 400
    },
    {
      "epoch": 0.06394770498792099,
      "grad_norm": 1.4144378900527954,
      "learning_rate": 4.49e-05,
      "loss": 7.1062,
      "step": 450
    },
    {
      "epoch": 0.07105300554213444,
      "grad_norm": 1.399983525276184,
      "learning_rate": 4.99e-05,
      "loss": 7.1068,
      "step": 500
    },
    {
      "epoch": 0.07815830609634787,
      "grad_norm": 1.180097222328186,
      "learning_rate": 5.49e-05,
      "loss": 7.0583,
      "step": 550
    },
    {
      "epoch": 0.08526360665056132,
      "grad_norm": 1.253781795501709,
      "learning_rate": 5.9900000000000006e-05,
      "loss": 7.04,
      "step": 600
    },
    {
      "epoch": 0.09236890720477477,
      "grad_norm": 1.1676445007324219,
      "learning_rate": 6.49e-05,
      "loss": 7.0169,
      "step": 650
    },
    {
      "epoch": 0.0994742077589882,
      "grad_norm": 1.308653473854065,
      "learning_rate": 6.99e-05,
      "loss": 6.94,
      "step": 700
    },
    {
      "epoch": 0.10657950831320165,
      "grad_norm": 1.7658427953720093,
      "learning_rate": 7.489999999999999e-05,
      "loss": 6.899,
      "step": 750
    },
    {
      "epoch": 0.1136848088674151,
      "grad_norm": 1.4796422719955444,
      "learning_rate": 7.99e-05,
      "loss": 6.8586,
      "step": 800
    },
    {
      "epoch": 0.12079010942162853,
      "grad_norm": 1.3457406759262085,
      "learning_rate": 8.49e-05,
      "loss": 6.8178,
      "step": 850
    },
    {
      "epoch": 0.12789540997584198,
      "grad_norm": 1.6725692749023438,
      "learning_rate": 8.989999999999999e-05,
      "loss": 6.7367,
      "step": 900
    },
    {
      "epoch": 0.1350007105300554,
      "grad_norm": 2.1698403358459473,
      "learning_rate": 9.49e-05,
      "loss": 6.7096,
      "step": 950
    },
    {
      "epoch": 0.14210601108426887,
      "grad_norm": 1.6723932027816772,
      "learning_rate": 9.99e-05,
      "loss": 6.6399,
      "step": 1000
    },
    {
      "epoch": 0.1492113116384823,
      "grad_norm": 1.8007622957229614,
      "learning_rate": 0.0001049,
      "loss": 6.597,
      "step": 1050
    },
    {
      "epoch": 0.15631661219269574,
      "grad_norm": 2.0344436168670654,
      "learning_rate": 0.0001099,
      "loss": 6.5591,
      "step": 1100
    },
    {
      "epoch": 0.1634219127469092,
      "grad_norm": 2.2933220863342285,
      "learning_rate": 0.0001149,
      "loss": 6.503,
      "step": 1150
    },
    {
      "epoch": 0.17052721330112264,
      "grad_norm": 2.4448184967041016,
      "learning_rate": 0.00011990000000000001,
      "loss": 6.4439,
      "step": 1200
    },
    {
      "epoch": 0.17763251385533607,
      "grad_norm": 2.009326219558716,
      "learning_rate": 0.0001249,
      "loss": 6.4007,
      "step": 1250
    },
    {
      "epoch": 0.18473781440954953,
      "grad_norm": 3.5166053771972656,
      "learning_rate": 0.00012989999999999999,
      "loss": 6.3217,
      "step": 1300
    },
    {
      "epoch": 0.19184311496376297,
      "grad_norm": 2.2721781730651855,
      "learning_rate": 0.0001349,
      "loss": 6.2818,
      "step": 1350
    },
    {
      "epoch": 0.1989484155179764,
      "grad_norm": 2.715651035308838,
      "learning_rate": 0.0001399,
      "loss": 6.192,
      "step": 1400
    },
    {
      "epoch": 0.20605371607218986,
      "grad_norm": 2.155087471008301,
      "learning_rate": 0.0001449,
      "loss": 6.1615,
      "step": 1450
    },
    {
      "epoch": 0.2131590166264033,
      "grad_norm": 2.541161060333252,
      "learning_rate": 0.0001499,
      "loss": 6.0676,
      "step": 1500
    },
    {
      "epoch": 0.22026431718061673,
      "grad_norm": 2.9829161167144775,
      "learning_rate": 0.00015490000000000002,
      "loss": 6.0268,
      "step": 1550
    },
    {
      "epoch": 0.2273696177348302,
      "grad_norm": 3.020001173019409,
      "learning_rate": 0.00015989999999999998,
      "loss": 5.9497,
      "step": 1600
    },
    {
      "epoch": 0.23447491828904363,
      "grad_norm": 2.6197469234466553,
      "learning_rate": 0.0001649,
      "loss": 5.888,
      "step": 1650
    },
    {
      "epoch": 0.24158021884325706,
      "grad_norm": 2.939798355102539,
      "learning_rate": 0.0001699,
      "loss": 5.8187,
      "step": 1700
    },
    {
      "epoch": 0.24868551939747052,
      "grad_norm": 4.240734100341797,
      "learning_rate": 0.0001749,
      "loss": 5.7487,
      "step": 1750
    },
    {
      "epoch": 0.25579081995168396,
      "grad_norm": 2.2773475646972656,
      "learning_rate": 0.0001799,
      "loss": 5.6821,
      "step": 1800
    },
    {
      "epoch": 0.2628961205058974,
      "grad_norm": 4.022791862487793,
      "learning_rate": 0.00018490000000000002,
      "loss": 5.6257,
      "step": 1850
    },
    {
      "epoch": 0.2700014210601108,
      "grad_norm": 2.742016315460205,
      "learning_rate": 0.0001899,
      "loss": 5.5804,
      "step": 1900
    },
    {
      "epoch": 0.27710672161432426,
      "grad_norm": 2.627027988433838,
      "learning_rate": 0.0001949,
      "loss": 5.5132,
      "step": 1950
    },
    {
      "epoch": 0.28421202216853775,
      "grad_norm": 2.689333200454712,
      "learning_rate": 0.0001999,
      "loss": 5.4796,
      "step": 2000
    },
    {
      "epoch": 0.28421202216853775,
      "eval_accuracy": 0.27806881070137024,
      "eval_loss": 5.384032726287842,
      "eval_runtime": 1.4127,
      "eval_samples_per_second": 2661.587,
      "eval_steps_per_second": 41.764,
      "step": 2000
    },
    {
      "epoch": 0.2913173227227512,
      "grad_norm": 2.620975971221924,
      "learning_rate": 0.0002049,
      "loss": 5.442,
      "step": 2050
    },
    {
      "epoch": 0.2984226232769646,
      "grad_norm": 3.2960896492004395,
      "learning_rate": 0.0002099,
      "loss": 5.3711,
      "step": 2100
    },
    {
      "epoch": 0.30552792383117805,
      "grad_norm": 2.6958446502685547,
      "learning_rate": 0.00021490000000000002,
      "loss": 5.3401,
      "step": 2150
    },
    {
      "epoch": 0.3126332243853915,
      "grad_norm": 2.4613704681396484,
      "learning_rate": 0.0002199,
      "loss": 5.317,
      "step": 2200
    },
    {
      "epoch": 0.3197385249396049,
      "grad_norm": 2.736056089401245,
      "learning_rate": 0.0002249,
      "loss": 5.2626,
      "step": 2250
    },
    {
      "epoch": 0.3268438254938184,
      "grad_norm": 3.220209836959839,
      "learning_rate": 0.0002299,
      "loss": 5.2429,
      "step": 2300
    },
    {
      "epoch": 0.33394912604803184,
      "grad_norm": 2.666049003601074,
      "learning_rate": 0.0002349,
      "loss": 5.2345,
      "step": 2350
    },
    {
      "epoch": 0.3410544266022453,
      "grad_norm": 2.9080348014831543,
      "learning_rate": 0.0002399,
      "loss": 5.2253,
      "step": 2400
    },
    {
      "epoch": 0.3481597271564587,
      "grad_norm": 2.500917673110962,
      "learning_rate": 0.0002449,
      "loss": 5.1528,
      "step": 2450
    },
    {
      "epoch": 0.35526502771067214,
      "grad_norm": 2.4509730339050293,
      "learning_rate": 0.0002499,
      "loss": 5.1371,
      "step": 2500
    },
    {
      "epoch": 0.3623703282648856,
      "grad_norm": 2.1525373458862305,
      "learning_rate": 0.0002549,
      "loss": 5.1283,
      "step": 2550
    },
    {
      "epoch": 0.36947562881909907,
      "grad_norm": 2.3644602298736572,
      "learning_rate": 0.00025990000000000003,
      "loss": 5.1132,
      "step": 2600
    },
    {
      "epoch": 0.3765809293733125,
      "grad_norm": 2.4550833702087402,
      "learning_rate": 0.00026490000000000004,
      "loss": 5.0745,
      "step": 2650
    },
    {
      "epoch": 0.38368622992752593,
      "grad_norm": 2.2757322788238525,
      "learning_rate": 0.0002699,
      "loss": 5.0384,
      "step": 2700
    },
    {
      "epoch": 0.39079153048173937,
      "grad_norm": 2.3816654682159424,
      "learning_rate": 0.00027489999999999996,
      "loss": 5.0265,
      "step": 2750
    },
    {
      "epoch": 0.3978968310359528,
      "grad_norm": 2.689765453338623,
      "learning_rate": 0.0002799,
      "loss": 5.0075,
      "step": 2800
    },
    {
      "epoch": 0.40500213159016624,
      "grad_norm": 2.3210747241973877,
      "learning_rate": 0.0002849,
      "loss": 4.9886,
      "step": 2850
    },
    {
      "epoch": 0.4121074321443797,
      "grad_norm": 2.2462069988250732,
      "learning_rate": 0.0002899,
      "loss": 4.9609,
      "step": 2900
    },
    {
      "epoch": 0.41921273269859316,
      "grad_norm": 2.407829999923706,
      "learning_rate": 0.0002949,
      "loss": 4.9633,
      "step": 2950
    },
    {
      "epoch": 0.4263180332528066,
      "grad_norm": 2.6128649711608887,
      "learning_rate": 0.0002999,
      "loss": 4.9396,
      "step": 3000
    },
    {
      "epoch": 0.43342333380702003,
      "grad_norm": 2.2061820030212402,
      "learning_rate": 0.0003049,
      "loss": 4.928,
      "step": 3050
    },
    {
      "epoch": 0.44052863436123346,
      "grad_norm": 2.3088581562042236,
      "learning_rate": 0.0003099,
      "loss": 4.9005,
      "step": 3100
    },
    {
      "epoch": 0.4476339349154469,
      "grad_norm": 2.3053336143493652,
      "learning_rate": 0.0003149,
      "loss": 4.8898,
      "step": 3150
    },
    {
      "epoch": 0.4547392354696604,
      "grad_norm": 2.061445951461792,
      "learning_rate": 0.0003199,
      "loss": 4.849,
      "step": 3200
    },
    {
      "epoch": 0.4618445360238738,
      "grad_norm": 2.8992912769317627,
      "learning_rate": 0.00032490000000000004,
      "loss": 4.8438,
      "step": 3250
    },
    {
      "epoch": 0.46894983657808725,
      "grad_norm": 2.3059287071228027,
      "learning_rate": 0.00032990000000000005,
      "loss": 4.7987,
      "step": 3300
    },
    {
      "epoch": 0.4760551371323007,
      "grad_norm": 2.248922824859619,
      "learning_rate": 0.0003349,
      "loss": 4.7949,
      "step": 3350
    },
    {
      "epoch": 0.4831604376865141,
      "grad_norm": 2.3223118782043457,
      "learning_rate": 0.00033989999999999997,
      "loss": 4.7798,
      "step": 3400
    },
    {
      "epoch": 0.49026573824072756,
      "grad_norm": 2.4109292030334473,
      "learning_rate": 0.0003449,
      "loss": 4.7713,
      "step": 3450
    },
    {
      "epoch": 0.49737103879494104,
      "grad_norm": 2.629187822341919,
      "learning_rate": 0.0003499,
      "loss": 4.7844,
      "step": 3500
    },
    {
      "epoch": 0.5044763393491545,
      "grad_norm": 2.707880735397339,
      "learning_rate": 0.0003549,
      "loss": 4.7225,
      "step": 3550
    },
    {
      "epoch": 0.5115816399033679,
      "grad_norm": 1.9972128868103027,
      "learning_rate": 0.0003599,
      "loss": 4.7385,
      "step": 3600
    },
    {
      "epoch": 0.5186869404575813,
      "grad_norm": 2.0002362728118896,
      "learning_rate": 0.00036490000000000003,
      "loss": 4.7187,
      "step": 3650
    },
    {
      "epoch": 0.5257922410117948,
      "grad_norm": 2.079793691635132,
      "learning_rate": 0.0003699,
      "loss": 4.6875,
      "step": 3700
    },
    {
      "epoch": 0.5328975415660082,
      "grad_norm": 2.2726001739501953,
      "learning_rate": 0.0003749,
      "loss": 4.6636,
      "step": 3750
    },
    {
      "epoch": 0.5400028421202216,
      "grad_norm": 2.250825881958008,
      "learning_rate": 0.0003799,
      "loss": 4.6815,
      "step": 3800
    },
    {
      "epoch": 0.5471081426744351,
      "grad_norm": 2.145050287246704,
      "learning_rate": 0.00038490000000000003,
      "loss": 4.669,
      "step": 3850
    },
    {
      "epoch": 0.5542134432286485,
      "grad_norm": 2.17869234085083,
      "learning_rate": 0.00038990000000000004,
      "loss": 4.651,
      "step": 3900
    },
    {
      "epoch": 0.5613187437828621,
      "grad_norm": 1.988489031791687,
      "learning_rate": 0.0003949,
      "loss": 4.6282,
      "step": 3950
    },
    {
      "epoch": 0.5684240443370755,
      "grad_norm": 2.583235025405884,
      "learning_rate": 0.00039989999999999996,
      "loss": 4.6166,
      "step": 4000
    },
    {
      "epoch": 0.5684240443370755,
      "eval_accuracy": 0.3410300612449646,
      "eval_loss": 4.496020317077637,
      "eval_runtime": 1.2758,
      "eval_samples_per_second": 2947.074,
      "eval_steps_per_second": 46.244,
      "step": 4000
    },
    {
      "epoch": 0.5755293448912889,
      "grad_norm": 2.1131043434143066,
      "learning_rate": 0.0004049,
      "loss": 4.6117,
      "step": 4050
    },
    {
      "epoch": 0.5826346454455024,
      "grad_norm": 2.024203062057495,
      "learning_rate": 0.0004099,
      "loss": 4.58,
      "step": 4100
    },
    {
      "epoch": 0.5897399459997158,
      "grad_norm": 2.1560163497924805,
      "learning_rate": 0.0004149,
      "loss": 4.5912,
      "step": 4150
    },
    {
      "epoch": 0.5968452465539292,
      "grad_norm": 1.8401826620101929,
      "learning_rate": 0.0004199,
      "loss": 4.5721,
      "step": 4200
    },
    {
      "epoch": 0.6039505471081427,
      "grad_norm": 1.9240000247955322,
      "learning_rate": 0.00042490000000000003,
      "loss": 4.5552,
      "step": 4250
    },
    {
      "epoch": 0.6110558476623561,
      "grad_norm": 1.9883395433425903,
      "learning_rate": 0.0004299,
      "loss": 4.5461,
      "step": 4300
    },
    {
      "epoch": 0.6181611482165695,
      "grad_norm": 1.7731436491012573,
      "learning_rate": 0.0004349,
      "loss": 4.5235,
      "step": 4350
    },
    {
      "epoch": 0.625266448770783,
      "grad_norm": 2.082683801651001,
      "learning_rate": 0.0004399,
      "loss": 4.4835,
      "step": 4400
    },
    {
      "epoch": 0.6323717493249964,
      "grad_norm": 1.7258926630020142,
      "learning_rate": 0.0004449,
      "loss": 4.4901,
      "step": 4450
    },
    {
      "epoch": 0.6394770498792098,
      "grad_norm": 1.8295708894729614,
      "learning_rate": 0.00044990000000000004,
      "loss": 4.4861,
      "step": 4500
    },
    {
      "epoch": 0.6465823504334234,
      "grad_norm": 1.955956220626831,
      "learning_rate": 0.00045490000000000005,
      "loss": 4.4275,
      "step": 4550
    },
    {
      "epoch": 0.6536876509876368,
      "grad_norm": 2.007366180419922,
      "learning_rate": 0.0004599,
      "loss": 4.4404,
      "step": 4600
    },
    {
      "epoch": 0.6607929515418502,
      "grad_norm": 1.8152107000350952,
      "learning_rate": 0.00046489999999999997,
      "loss": 4.4501,
      "step": 4650
    },
    {
      "epoch": 0.6678982520960637,
      "grad_norm": 1.5465844869613647,
      "learning_rate": 0.0004699,
      "loss": 4.4577,
      "step": 4700
    },
    {
      "epoch": 0.6750035526502771,
      "grad_norm": 1.9919116497039795,
      "learning_rate": 0.0004749,
      "loss": 4.4107,
      "step": 4750
    },
    {
      "epoch": 0.6821088532044906,
      "grad_norm": 1.665770173072815,
      "learning_rate": 0.0004799,
      "loss": 4.3893,
      "step": 4800
    },
    {
      "epoch": 0.689214153758704,
      "grad_norm": 1.6241706609725952,
      "learning_rate": 0.0004849,
      "loss": 4.3998,
      "step": 4850
    },
    {
      "epoch": 0.6963194543129174,
      "grad_norm": 1.9763983488082886,
      "learning_rate": 0.0004899,
      "loss": 4.3905,
      "step": 4900
    },
    {
      "epoch": 0.7034247548671309,
      "grad_norm": 1.7900429964065552,
      "learning_rate": 0.0004949,
      "loss": 4.3574,
      "step": 4950
    },
    {
      "epoch": 0.7105300554213443,
      "grad_norm": 1.7338964939117432,
      "learning_rate": 0.0004999000000000001,
      "loss": 4.3495,
      "step": 5000
    },
    {
      "epoch": 0.7176353559755577,
      "grad_norm": 1.7506004571914673,
      "learning_rate": 0.0004994555555555555,
      "loss": 4.3368,
      "step": 5050
    },
    {
      "epoch": 0.7247406565297712,
      "grad_norm": 1.6576265096664429,
      "learning_rate": 0.0004989,
      "loss": 4.3262,
      "step": 5100
    },
    {
      "epoch": 0.7318459570839847,
      "grad_norm": 1.6786417961120605,
      "learning_rate": 0.0004983444444444444,
      "loss": 4.3262,
      "step": 5150
    },
    {
      "epoch": 0.7389512576381981,
      "grad_norm": 1.7846481800079346,
      "learning_rate": 0.0004977888888888889,
      "loss": 4.3184,
      "step": 5200
    },
    {
      "epoch": 0.7460565581924116,
      "grad_norm": 1.50885009765625,
      "learning_rate": 0.0004972333333333334,
      "loss": 4.2821,
      "step": 5250
    },
    {
      "epoch": 0.753161858746625,
      "grad_norm": 1.887153148651123,
      "learning_rate": 0.0004966777777777778,
      "loss": 4.2808,
      "step": 5300
    },
    {
      "epoch": 0.7602671593008384,
      "grad_norm": 1.6502448320388794,
      "learning_rate": 0.0004961222222222223,
      "loss": 4.2995,
      "step": 5350
    },
    {
      "epoch": 0.7673724598550519,
      "grad_norm": 1.6177388429641724,
      "learning_rate": 0.0004955666666666667,
      "loss": 4.2695,
      "step": 5400
    },
    {
      "epoch": 0.7744777604092653,
      "grad_norm": 1.56751549243927,
      "learning_rate": 0.0004950111111111112,
      "loss": 4.2314,
      "step": 5450
    },
    {
      "epoch": 0.7815830609634787,
      "grad_norm": 1.7374509572982788,
      "learning_rate": 0.0004944555555555555,
      "loss": 4.2265,
      "step": 5500
    },
    {
      "epoch": 0.7886883615176922,
      "grad_norm": 1.7130149602890015,
      "learning_rate": 0.0004939,
      "loss": 4.2313,
      "step": 5550
    },
    {
      "epoch": 0.7957936620719056,
      "grad_norm": 1.669311285018921,
      "learning_rate": 0.0004933444444444444,
      "loss": 4.21,
      "step": 5600
    },
    {
      "epoch": 0.802898962626119,
      "grad_norm": 1.6400610208511353,
      "learning_rate": 0.0004927888888888889,
      "loss": 4.1947,
      "step": 5650
    },
    {
      "epoch": 0.8100042631803325,
      "grad_norm": 1.6835277080535889,
      "learning_rate": 0.0004922333333333334,
      "loss": 4.1966,
      "step": 5700
    },
    {
      "epoch": 0.817109563734546,
      "grad_norm": 1.7618376016616821,
      "learning_rate": 0.0004916777777777778,
      "loss": 4.1575,
      "step": 5750
    },
    {
      "epoch": 0.8242148642887595,
      "grad_norm": 1.6249600648880005,
      "learning_rate": 0.0004911222222222223,
      "loss": 4.1816,
      "step": 5800
    },
    {
      "epoch": 0.8313201648429729,
      "grad_norm": 1.5585474967956543,
      "learning_rate": 0.0004905666666666666,
      "loss": 4.1563,
      "step": 5850
    },
    {
      "epoch": 0.8384254653971863,
      "grad_norm": 1.5878101587295532,
      "learning_rate": 0.0004900111111111111,
      "loss": 4.1433,
      "step": 5900
    },
    {
      "epoch": 0.8455307659513998,
      "grad_norm": 1.7434005737304688,
      "learning_rate": 0.0004894555555555555,
      "loss": 4.1249,
      "step": 5950
    },
    {
      "epoch": 0.8526360665056132,
      "grad_norm": 1.6475976705551147,
      "learning_rate": 0.0004889,
      "loss": 4.1271,
      "step": 6000
    },
    {
      "epoch": 0.8526360665056132,
      "eval_accuracy": 0.3815571367740631,
      "eval_loss": 4.0217814445495605,
      "eval_runtime": 1.347,
      "eval_samples_per_second": 2791.439,
      "eval_steps_per_second": 43.802,
      "step": 6000
    },
    {
      "epoch": 0.8597413670598266,
      "grad_norm": 1.5479772090911865,
      "learning_rate": 0.0004883444444444445,
      "loss": 4.123,
      "step": 6050
    },
    {
      "epoch": 0.8668466676140401,
      "grad_norm": 1.6345648765563965,
      "learning_rate": 0.0004877888888888889,
      "loss": 4.0898,
      "step": 6100
    },
    {
      "epoch": 0.8739519681682535,
      "grad_norm": 1.6345279216766357,
      "learning_rate": 0.0004872333333333334,
      "loss": 4.1015,
      "step": 6150
    },
    {
      "epoch": 0.8810572687224669,
      "grad_norm": 1.691853642463684,
      "learning_rate": 0.00048667777777777776,
      "loss": 4.0686,
      "step": 6200
    },
    {
      "epoch": 0.8881625692766804,
      "grad_norm": 1.5769050121307373,
      "learning_rate": 0.00048612222222222225,
      "loss": 4.0704,
      "step": 6250
    },
    {
      "epoch": 0.8952678698308938,
      "grad_norm": 1.539277195930481,
      "learning_rate": 0.00048556666666666663,
      "loss": 4.0777,
      "step": 6300
    },
    {
      "epoch": 0.9023731703851073,
      "grad_norm": 1.5936651229858398,
      "learning_rate": 0.0004850111111111111,
      "loss": 4.0539,
      "step": 6350
    },
    {
      "epoch": 0.9094784709393208,
      "grad_norm": 1.5352519750595093,
      "learning_rate": 0.00048445555555555556,
      "loss": 4.0572,
      "step": 6400
    },
    {
      "epoch": 0.9165837714935342,
      "grad_norm": 1.5943176746368408,
      "learning_rate": 0.0004839,
      "loss": 4.0325,
      "step": 6450
    },
    {
      "epoch": 0.9236890720477476,
      "grad_norm": 1.6698144674301147,
      "learning_rate": 0.0004833444444444445,
      "loss": 4.021,
      "step": 6500
    },
    {
      "epoch": 0.9307943726019611,
      "grad_norm": 1.4587650299072266,
      "learning_rate": 0.00048278888888888887,
      "loss": 4.0297,
      "step": 6550
    },
    {
      "epoch": 0.9378996731561745,
      "grad_norm": 1.6008667945861816,
      "learning_rate": 0.00048223333333333336,
      "loss": 4.0161,
      "step": 6600
    },
    {
      "epoch": 0.9450049737103879,
      "grad_norm": 1.550215721130371,
      "learning_rate": 0.00048167777777777775,
      "loss": 3.974,
      "step": 6650
    },
    {
      "epoch": 0.9521102742646014,
      "grad_norm": 1.6898516416549683,
      "learning_rate": 0.00048112222222222224,
      "loss": 3.992,
      "step": 6700
    },
    {
      "epoch": 0.9592155748188148,
      "grad_norm": 1.4572670459747314,
      "learning_rate": 0.0004805666666666667,
      "loss": 3.9695,
      "step": 6750
    },
    {
      "epoch": 0.9663208753730282,
      "grad_norm": 1.482657790184021,
      "learning_rate": 0.0004800111111111111,
      "loss": 3.9757,
      "step": 6800
    },
    {
      "epoch": 0.9734261759272417,
      "grad_norm": 1.404305100440979,
      "learning_rate": 0.0004794555555555556,
      "loss": 3.9483,
      "step": 6850
    },
    {
      "epoch": 0.9805314764814551,
      "grad_norm": 1.5593451261520386,
      "learning_rate": 0.0004789,
      "loss": 3.9546,
      "step": 6900
    },
    {
      "epoch": 0.9876367770356687,
      "grad_norm": 1.6054081916809082,
      "learning_rate": 0.0004783444444444445,
      "loss": 3.9455,
      "step": 6950
    },
    {
      "epoch": 0.9947420775898821,
      "grad_norm": 1.6005113124847412,
      "learning_rate": 0.00047778888888888886,
      "loss": 3.9228,
      "step": 7000
    },
    {
      "epoch": 1.0018473781440955,
      "grad_norm": 1.5673545598983765,
      "learning_rate": 0.00047723333333333335,
      "loss": 3.9479,
      "step": 7050
    },
    {
      "epoch": 1.008952678698309,
      "grad_norm": 1.5405232906341553,
      "learning_rate": 0.0004766777777777778,
      "loss": 3.9094,
      "step": 7100
    },
    {
      "epoch": 1.0160579792525224,
      "grad_norm": 1.550139307975769,
      "learning_rate": 0.0004761222222222222,
      "loss": 3.9012,
      "step": 7150
    },
    {
      "epoch": 1.0231632798067358,
      "grad_norm": 1.513722538948059,
      "learning_rate": 0.0004755666666666667,
      "loss": 3.9027,
      "step": 7200
    },
    {
      "epoch": 1.0302685803609493,
      "grad_norm": 1.5611164569854736,
      "learning_rate": 0.0004750111111111111,
      "loss": 3.889,
      "step": 7250
    },
    {
      "epoch": 1.0373738809151627,
      "grad_norm": 1.5630569458007812,
      "learning_rate": 0.0004744555555555556,
      "loss": 3.8921,
      "step": 7300
    },
    {
      "epoch": 1.0444791814693761,
      "grad_norm": 1.5469523668289185,
      "learning_rate": 0.00047389999999999997,
      "loss": 3.8684,
      "step": 7350
    },
    {
      "epoch": 1.0515844820235896,
      "grad_norm": 1.4361380338668823,
      "learning_rate": 0.00047334444444444446,
      "loss": 3.8806,
      "step": 7400
    },
    {
      "epoch": 1.058689782577803,
      "grad_norm": 1.5062557458877563,
      "learning_rate": 0.0004727888888888889,
      "loss": 3.8982,
      "step": 7450
    },
    {
      "epoch": 1.0657950831320164,
      "grad_norm": 1.5810432434082031,
      "learning_rate": 0.00047223333333333334,
      "loss": 3.8627,
      "step": 7500
    },
    {
      "epoch": 1.0729003836862299,
      "grad_norm": 1.4955435991287231,
      "learning_rate": 0.0004716777777777778,
      "loss": 3.8626,
      "step": 7550
    },
    {
      "epoch": 1.0800056842404433,
      "grad_norm": 1.4136838912963867,
      "learning_rate": 0.0004711222222222222,
      "loss": 3.8536,
      "step": 7600
    },
    {
      "epoch": 1.0871109847946567,
      "grad_norm": 1.5749014616012573,
      "learning_rate": 0.0004705666666666667,
      "loss": 3.8523,
      "step": 7650
    },
    {
      "epoch": 1.0942162853488702,
      "grad_norm": 1.5885382890701294,
      "learning_rate": 0.0004700111111111111,
      "loss": 3.8116,
      "step": 7700
    },
    {
      "epoch": 1.1013215859030836,
      "grad_norm": 1.5716185569763184,
      "learning_rate": 0.0004694555555555556,
      "loss": 3.8193,
      "step": 7750
    },
    {
      "epoch": 1.108426886457297,
      "grad_norm": 1.4737333059310913,
      "learning_rate": 0.0004689,
      "loss": 3.8319,
      "step": 7800
    },
    {
      "epoch": 1.1155321870115107,
      "grad_norm": 1.5206536054611206,
      "learning_rate": 0.00046834444444444445,
      "loss": 3.7899,
      "step": 7850
    },
    {
      "epoch": 1.1226374875657241,
      "grad_norm": 1.584385633468628,
      "learning_rate": 0.0004677888888888889,
      "loss": 3.8117,
      "step": 7900
    },
    {
      "epoch": 1.1297427881199376,
      "grad_norm": 1.614679217338562,
      "learning_rate": 0.0004672333333333333,
      "loss": 3.8008,
      "step": 7950
    },
    {
      "epoch": 1.136848088674151,
      "grad_norm": 1.7981069087982178,
      "learning_rate": 0.0004666777777777778,
      "loss": 3.8065,
      "step": 8000
    },
    {
      "epoch": 1.136848088674151,
      "eval_accuracy": 0.4039747714996338,
      "eval_loss": 3.722554922103882,
      "eval_runtime": 1.2852,
      "eval_samples_per_second": 2925.727,
      "eval_steps_per_second": 45.909,
      "step": 8000
    },
    {
      "epoch": 1.1439533892283644,
      "grad_norm": 1.3700244426727295,
      "learning_rate": 0.0004661222222222222,
      "loss": 3.7843,
      "step": 8050
    },
    {
      "epoch": 1.1510586897825779,
      "grad_norm": 1.6801719665527344,
      "learning_rate": 0.0004655666666666667,
      "loss": 3.7919,
      "step": 8100
    },
    {
      "epoch": 1.1581639903367913,
      "grad_norm": 1.5943037271499634,
      "learning_rate": 0.0004650111111111111,
      "loss": 3.782,
      "step": 8150
    },
    {
      "epoch": 1.1652692908910047,
      "grad_norm": 1.434883713722229,
      "learning_rate": 0.00046445555555555556,
      "loss": 3.7404,
      "step": 8200
    },
    {
      "epoch": 1.1723745914452182,
      "grad_norm": 1.5571914911270142,
      "learning_rate": 0.0004639,
      "loss": 3.7607,
      "step": 8250
    },
    {
      "epoch": 1.1794798919994316,
      "grad_norm": 1.4568759202957153,
      "learning_rate": 0.00046334444444444444,
      "loss": 3.7598,
      "step": 8300
    },
    {
      "epoch": 1.186585192553645,
      "grad_norm": 1.5035616159439087,
      "learning_rate": 0.00046278888888888893,
      "loss": 3.74,
      "step": 8350
    },
    {
      "epoch": 1.1936904931078585,
      "grad_norm": 1.396409273147583,
      "learning_rate": 0.00046223333333333337,
      "loss": 3.7324,
      "step": 8400
    },
    {
      "epoch": 1.200795793662072,
      "grad_norm": 1.4661195278167725,
      "learning_rate": 0.0004616777777777778,
      "loss": 3.7349,
      "step": 8450
    },
    {
      "epoch": 1.2079010942162853,
      "grad_norm": 1.4614042043685913,
      "learning_rate": 0.00046112222222222224,
      "loss": 3.7451,
      "step": 8500
    },
    {
      "epoch": 1.2150063947704988,
      "grad_norm": 1.4750593900680542,
      "learning_rate": 0.0004605666666666667,
      "loss": 3.7319,
      "step": 8550
    },
    {
      "epoch": 1.2221116953247122,
      "grad_norm": 1.4696282148361206,
      "learning_rate": 0.0004600111111111111,
      "loss": 3.7226,
      "step": 8600
    },
    {
      "epoch": 1.2292169958789256,
      "grad_norm": 1.4502874612808228,
      "learning_rate": 0.00045945555555555555,
      "loss": 3.7188,
      "step": 8650
    },
    {
      "epoch": 1.236322296433139,
      "grad_norm": 1.5446569919586182,
      "learning_rate": 0.0004589,
      "loss": 3.7264,
      "step": 8700
    },
    {
      "epoch": 1.2434275969873525,
      "grad_norm": 1.7836312055587769,
      "learning_rate": 0.0004583444444444445,
      "loss": 3.7129,
      "step": 8750
    },
    {
      "epoch": 1.250532897541566,
      "grad_norm": 1.430017113685608,
      "learning_rate": 0.0004577888888888889,
      "loss": 3.7156,
      "step": 8800
    },
    {
      "epoch": 1.2576381980957794,
      "grad_norm": 1.4015753269195557,
      "learning_rate": 0.00045723333333333335,
      "loss": 3.7096,
      "step": 8850
    },
    {
      "epoch": 1.264743498649993,
      "grad_norm": 1.4344234466552734,
      "learning_rate": 0.0004566777777777778,
      "loss": 3.6902,
      "step": 8900
    },
    {
      "epoch": 1.2718487992042062,
      "grad_norm": 1.450966715812683,
      "learning_rate": 0.0004561222222222222,
      "loss": 3.6944,
      "step": 8950
    },
    {
      "epoch": 1.27895409975842,
      "grad_norm": 1.482603669166565,
      "learning_rate": 0.00045556666666666666,
      "loss": 3.6716,
      "step": 9000
    },
    {
      "epoch": 1.286059400312633,
      "grad_norm": 1.5204460620880127,
      "learning_rate": 0.0004550111111111111,
      "loss": 3.6782,
      "step": 9050
    },
    {
      "epoch": 1.2931647008668468,
      "grad_norm": 1.3356878757476807,
      "learning_rate": 0.0004544555555555556,
      "loss": 3.6502,
      "step": 9100
    },
    {
      "epoch": 1.30027000142106,
      "grad_norm": 1.5904278755187988,
      "learning_rate": 0.00045390000000000003,
      "loss": 3.6615,
      "step": 9150
    },
    {
      "epoch": 1.3073753019752736,
      "grad_norm": 1.4965075254440308,
      "learning_rate": 0.00045334444444444447,
      "loss": 3.6604,
      "step": 9200
    },
    {
      "epoch": 1.314480602529487,
      "grad_norm": 1.5077792406082153,
      "learning_rate": 0.0004527888888888889,
      "loss": 3.638,
      "step": 9250
    },
    {
      "epoch": 1.3215859030837005,
      "grad_norm": 1.470985770225525,
      "learning_rate": 0.00045223333333333334,
      "loss": 3.6634,
      "step": 9300
    },
    {
      "epoch": 1.328691203637914,
      "grad_norm": 1.591996192932129,
      "learning_rate": 0.0004516777777777778,
      "loss": 3.6287,
      "step": 9350
    },
    {
      "epoch": 1.3357965041921274,
      "grad_norm": 1.4681540727615356,
      "learning_rate": 0.0004511222222222222,
      "loss": 3.6118,
      "step": 9400
    },
    {
      "epoch": 1.3429018047463408,
      "grad_norm": 1.5497246980667114,
      "learning_rate": 0.0004505666666666667,
      "loss": 3.6179,
      "step": 9450
    },
    {
      "epoch": 1.3500071053005542,
      "grad_norm": 1.499262809753418,
      "learning_rate": 0.0004500111111111111,
      "loss": 3.62,
      "step": 9500
    },
    {
      "epoch": 1.3571124058547677,
      "grad_norm": 1.4057672023773193,
      "learning_rate": 0.0004494555555555556,
      "loss": 3.6019,
      "step": 9550
    },
    {
      "epoch": 1.364217706408981,
      "grad_norm": 1.5262913703918457,
      "learning_rate": 0.0004489,
      "loss": 3.588,
      "step": 9600
    },
    {
      "epoch": 1.3713230069631945,
      "grad_norm": 1.4854165315628052,
      "learning_rate": 0.00044834444444444445,
      "loss": 3.5968,
      "step": 9650
    },
    {
      "epoch": 1.378428307517408,
      "grad_norm": 1.4390678405761719,
      "learning_rate": 0.0004477888888888889,
      "loss": 3.6228,
      "step": 9700
    },
    {
      "epoch": 1.3855336080716214,
      "grad_norm": 1.527435541152954,
      "learning_rate": 0.0004472333333333333,
      "loss": 3.6099,
      "step": 9750
    },
    {
      "epoch": 1.3926389086258348,
      "grad_norm": 1.4992133378982544,
      "learning_rate": 0.0004466777777777778,
      "loss": 3.6263,
      "step": 9800
    },
    {
      "epoch": 1.3997442091800483,
      "grad_norm": 1.5344403982162476,
      "learning_rate": 0.0004461222222222222,
      "loss": 3.5873,
      "step": 9850
    },
    {
      "epoch": 1.4068495097342617,
      "grad_norm": 1.4833427667617798,
      "learning_rate": 0.0004455666666666667,
      "loss": 3.57,
      "step": 9900
    },
    {
      "epoch": 1.4139548102884751,
      "grad_norm": 1.5680700540542603,
      "learning_rate": 0.00044501111111111113,
      "loss": 3.5832,
      "step": 9950
    },
    {
      "epoch": 1.4210601108426886,
      "grad_norm": 1.560567855834961,
      "learning_rate": 0.00044445555555555557,
      "loss": 3.5744,
      "step": 10000
    },
    {
      "epoch": 1.4210601108426886,
      "eval_accuracy": 0.4289895296096802,
      "eval_loss": 3.51322603225708,
      "eval_runtime": 1.2882,
      "eval_samples_per_second": 2918.705,
      "eval_steps_per_second": 45.799,
      "step": 10000
    },
    {
      "epoch": 1.428165411396902,
      "grad_norm": 1.402680516242981,
      "learning_rate": 0.0004439,
      "loss": 3.5277,
      "step": 10050
    },
    {
      "epoch": 1.4352707119511154,
      "grad_norm": 1.4061716794967651,
      "learning_rate": 0.00044334444444444444,
      "loss": 3.5653,
      "step": 10100
    },
    {
      "epoch": 1.442376012505329,
      "grad_norm": 1.5107778310775757,
      "learning_rate": 0.00044278888888888893,
      "loss": 3.5627,
      "step": 10150
    },
    {
      "epoch": 1.4494813130595423,
      "grad_norm": 1.4344958066940308,
      "learning_rate": 0.0004422333333333333,
      "loss": 3.5403,
      "step": 10200
    },
    {
      "epoch": 1.456586613613756,
      "grad_norm": 1.4615083932876587,
      "learning_rate": 0.0004416777777777778,
      "loss": 3.5344,
      "step": 10250
    },
    {
      "epoch": 1.4636919141679692,
      "grad_norm": 1.491568684577942,
      "learning_rate": 0.00044112222222222224,
      "loss": 3.5576,
      "step": 10300
    },
    {
      "epoch": 1.4707972147221828,
      "grad_norm": 1.4227489233016968,
      "learning_rate": 0.0004405666666666667,
      "loss": 3.533,
      "step": 10350
    },
    {
      "epoch": 1.4779025152763963,
      "grad_norm": 1.5831272602081299,
      "learning_rate": 0.0004400111111111111,
      "loss": 3.54,
      "step": 10400
    },
    {
      "epoch": 1.4850078158306097,
      "grad_norm": 1.3513504266738892,
      "learning_rate": 0.00043945555555555555,
      "loss": 3.5517,
      "step": 10450
    },
    {
      "epoch": 1.4921131163848231,
      "grad_norm": 1.704670786857605,
      "learning_rate": 0.00043890000000000004,
      "loss": 3.5364,
      "step": 10500
    },
    {
      "epoch": 1.4992184169390366,
      "grad_norm": 1.4062697887420654,
      "learning_rate": 0.0004383444444444444,
      "loss": 3.5202,
      "step": 10550
    },
    {
      "epoch": 1.50632371749325,
      "grad_norm": 1.3823599815368652,
      "learning_rate": 0.0004377888888888889,
      "loss": 3.5123,
      "step": 10600
    },
    {
      "epoch": 1.5134290180474634,
      "grad_norm": 1.617918848991394,
      "learning_rate": 0.0004372333333333333,
      "loss": 3.53,
      "step": 10650
    },
    {
      "epoch": 1.5205343186016769,
      "grad_norm": 1.5281829833984375,
      "learning_rate": 0.0004366777777777778,
      "loss": 3.5191,
      "step": 10700
    },
    {
      "epoch": 1.5276396191558903,
      "grad_norm": 1.454651951789856,
      "learning_rate": 0.00043612222222222223,
      "loss": 3.5129,
      "step": 10750
    },
    {
      "epoch": 1.5347449197101037,
      "grad_norm": 1.5583391189575195,
      "learning_rate": 0.00043556666666666666,
      "loss": 3.4786,
      "step": 10800
    },
    {
      "epoch": 1.5418502202643172,
      "grad_norm": 1.4228161573410034,
      "learning_rate": 0.00043501111111111116,
      "loss": 3.5236,
      "step": 10850
    },
    {
      "epoch": 1.5489555208185306,
      "grad_norm": 1.5005617141723633,
      "learning_rate": 0.00043445555555555554,
      "loss": 3.4942,
      "step": 10900
    },
    {
      "epoch": 1.556060821372744,
      "grad_norm": 1.4787464141845703,
      "learning_rate": 0.00043390000000000003,
      "loss": 3.4913,
      "step": 10950
    },
    {
      "epoch": 1.5631661219269575,
      "grad_norm": 1.5321528911590576,
      "learning_rate": 0.0004333444444444444,
      "loss": 3.4675,
      "step": 11000
    },
    {
      "epoch": 1.570271422481171,
      "grad_norm": 1.3829543590545654,
      "learning_rate": 0.0004327888888888889,
      "loss": 3.4547,
      "step": 11050
    },
    {
      "epoch": 1.5773767230353843,
      "grad_norm": 1.7110611200332642,
      "learning_rate": 0.00043223333333333334,
      "loss": 3.4481,
      "step": 11100
    },
    {
      "epoch": 1.5844820235895978,
      "grad_norm": 1.5753222703933716,
      "learning_rate": 0.0004316777777777778,
      "loss": 3.484,
      "step": 11150
    },
    {
      "epoch": 1.5915873241438114,
      "grad_norm": 1.454259991645813,
      "learning_rate": 0.00043112222222222227,
      "loss": 3.4728,
      "step": 11200
    },
    {
      "epoch": 1.5986926246980246,
      "grad_norm": 1.4018853902816772,
      "learning_rate": 0.00043056666666666665,
      "loss": 3.4706,
      "step": 11250
    },
    {
      "epoch": 1.6057979252522383,
      "grad_norm": 1.6062161922454834,
      "learning_rate": 0.00043001111111111114,
      "loss": 3.4497,
      "step": 11300
    },
    {
      "epoch": 1.6129032258064515,
      "grad_norm": 1.585882306098938,
      "learning_rate": 0.0004294555555555555,
      "loss": 3.4331,
      "step": 11350
    },
    {
      "epoch": 1.6200085263606652,
      "grad_norm": 1.4439284801483154,
      "learning_rate": 0.0004289,
      "loss": 3.4684,
      "step": 11400
    },
    {
      "epoch": 1.6271138269148784,
      "grad_norm": 1.3603633642196655,
      "learning_rate": 0.0004283444444444445,
      "loss": 3.4541,
      "step": 11450
    },
    {
      "epoch": 1.634219127469092,
      "grad_norm": 1.4939123392105103,
      "learning_rate": 0.0004277888888888889,
      "loss": 3.46,
      "step": 11500
    },
    {
      "epoch": 1.6413244280233052,
      "grad_norm": 1.4626458883285522,
      "learning_rate": 0.0004272333333333334,
      "loss": 3.4453,
      "step": 11550
    },
    {
      "epoch": 1.648429728577519,
      "grad_norm": 1.5030425786972046,
      "learning_rate": 0.00042667777777777776,
      "loss": 3.4265,
      "step": 11600
    },
    {
      "epoch": 1.6555350291317321,
      "grad_norm": 1.4823453426361084,
      "learning_rate": 0.00042612222222222226,
      "loss": 3.4356,
      "step": 11650
    },
    {
      "epoch": 1.6626403296859458,
      "grad_norm": 1.510815978050232,
      "learning_rate": 0.00042556666666666664,
      "loss": 3.464,
      "step": 11700
    },
    {
      "epoch": 1.6697456302401592,
      "grad_norm": 1.4318405389785767,
      "learning_rate": 0.00042501111111111113,
      "loss": 3.4274,
      "step": 11750
    },
    {
      "epoch": 1.6768509307943726,
      "grad_norm": 1.4325045347213745,
      "learning_rate": 0.0004244555555555555,
      "loss": 3.4449,
      "step": 11800
    },
    {
      "epoch": 1.683956231348586,
      "grad_norm": 1.6433807611465454,
      "learning_rate": 0.0004239,
      "loss": 3.4162,
      "step": 11850
    },
    {
      "epoch": 1.6910615319027995,
      "grad_norm": 1.6231051683425903,
      "learning_rate": 0.0004233444444444445,
      "loss": 3.4107,
      "step": 11900
    },
    {
      "epoch": 1.698166832457013,
      "grad_norm": 1.468326449394226,
      "learning_rate": 0.0004227888888888889,
      "loss": 3.3951,
      "step": 11950
    },
    {
      "epoch": 1.7052721330112264,
      "grad_norm": 1.5077193975448608,
      "learning_rate": 0.00042223333333333337,
      "loss": 3.4196,
      "step": 12000
    },
    {
      "epoch": 1.7052721330112264,
      "eval_accuracy": 0.44879472255706787,
      "eval_loss": 3.3239564895629883,
      "eval_runtime": 1.3407,
      "eval_samples_per_second": 2804.532,
      "eval_steps_per_second": 44.007,
      "step": 12000
    },
    {
      "epoch": 1.7123774335654398,
      "grad_norm": 1.4623733758926392,
      "learning_rate": 0.00042167777777777775,
      "loss": 3.392,
      "step": 12050
    },
    {
      "epoch": 1.7194827341196532,
      "grad_norm": 1.561031699180603,
      "learning_rate": 0.00042112222222222224,
      "loss": 3.4159,
      "step": 12100
    },
    {
      "epoch": 1.7265880346738667,
      "grad_norm": 1.495694875717163,
      "learning_rate": 0.0004205666666666667,
      "loss": 3.411,
      "step": 12150
    },
    {
      "epoch": 1.7336933352280801,
      "grad_norm": 1.462576150894165,
      "learning_rate": 0.0004200111111111111,
      "loss": 3.4034,
      "step": 12200
    },
    {
      "epoch": 1.7407986357822935,
      "grad_norm": 1.582942247390747,
      "learning_rate": 0.0004194555555555556,
      "loss": 3.3976,
      "step": 12250
    },
    {
      "epoch": 1.747903936336507,
      "grad_norm": 1.4122976064682007,
      "learning_rate": 0.0004189,
      "loss": 3.3722,
      "step": 12300
    },
    {
      "epoch": 1.7550092368907206,
      "grad_norm": 1.7197649478912354,
      "learning_rate": 0.0004183444444444445,
      "loss": 3.3926,
      "step": 12350
    },
    {
      "epoch": 1.7621145374449338,
      "grad_norm": 1.452899694442749,
      "learning_rate": 0.00041778888888888886,
      "loss": 3.3943,
      "step": 12400
    },
    {
      "epoch": 1.7692198379991475,
      "grad_norm": 1.4426175355911255,
      "learning_rate": 0.00041723333333333336,
      "loss": 3.375,
      "step": 12450
    },
    {
      "epoch": 1.7763251385533607,
      "grad_norm": 1.4973036050796509,
      "learning_rate": 0.0004166777777777778,
      "loss": 3.3823,
      "step": 12500
    },
    {
      "epoch": 1.7834304391075744,
      "grad_norm": 1.5862528085708618,
      "learning_rate": 0.00041612222222222223,
      "loss": 3.3689,
      "step": 12550
    },
    {
      "epoch": 1.7905357396617876,
      "grad_norm": 1.5989631414413452,
      "learning_rate": 0.00041556666666666667,
      "loss": 3.3464,
      "step": 12600
    },
    {
      "epoch": 1.7976410402160012,
      "grad_norm": 1.54103422164917,
      "learning_rate": 0.0004150111111111111,
      "loss": 3.3558,
      "step": 12650
    },
    {
      "epoch": 1.8047463407702145,
      "grad_norm": 1.4894845485687256,
      "learning_rate": 0.0004144555555555556,
      "loss": 3.3531,
      "step": 12700
    },
    {
      "epoch": 1.811851641324428,
      "grad_norm": 1.3978097438812256,
      "learning_rate": 0.0004139,
      "loss": 3.3272,
      "step": 12750
    },
    {
      "epoch": 1.8189569418786413,
      "grad_norm": 1.5314109325408936,
      "learning_rate": 0.00041334444444444447,
      "loss": 3.3423,
      "step": 12800
    },
    {
      "epoch": 1.826062242432855,
      "grad_norm": 1.595335841178894,
      "learning_rate": 0.0004127888888888889,
      "loss": 3.3897,
      "step": 12850
    },
    {
      "epoch": 1.8331675429870682,
      "grad_norm": 1.457351565361023,
      "learning_rate": 0.00041223333333333334,
      "loss": 3.3346,
      "step": 12900
    },
    {
      "epoch": 1.8402728435412818,
      "grad_norm": 1.4411284923553467,
      "learning_rate": 0.0004116777777777778,
      "loss": 3.3504,
      "step": 12950
    },
    {
      "epoch": 1.8473781440954953,
      "grad_norm": 1.4269949197769165,
      "learning_rate": 0.0004111222222222222,
      "loss": 3.3533,
      "step": 13000
    },
    {
      "epoch": 1.8544834446497087,
      "grad_norm": 1.4080978631973267,
      "learning_rate": 0.0004105666666666667,
      "loss": 3.3357,
      "step": 13050
    },
    {
      "epoch": 1.8615887452039221,
      "grad_norm": 1.4291956424713135,
      "learning_rate": 0.0004100111111111111,
      "loss": 3.3473,
      "step": 13100
    },
    {
      "epoch": 1.8686940457581356,
      "grad_norm": 1.411839246749878,
      "learning_rate": 0.0004094555555555556,
      "loss": 3.3267,
      "step": 13150
    },
    {
      "epoch": 1.875799346312349,
      "grad_norm": 1.5854250192642212,
      "learning_rate": 0.0004089,
      "loss": 3.3545,
      "step": 13200
    },
    {
      "epoch": 1.8829046468665624,
      "grad_norm": 1.330222725868225,
      "learning_rate": 0.00040834444444444446,
      "loss": 3.3221,
      "step": 13250
    },
    {
      "epoch": 1.8900099474207759,
      "grad_norm": 1.4597231149673462,
      "learning_rate": 0.0004077888888888889,
      "loss": 3.333,
      "step": 13300
    },
    {
      "epoch": 1.8971152479749893,
      "grad_norm": 1.4997475147247314,
      "learning_rate": 0.00040723333333333333,
      "loss": 3.3313,
      "step": 13350
    },
    {
      "epoch": 1.9042205485292028,
      "grad_norm": 1.5232534408569336,
      "learning_rate": 0.0004066777777777778,
      "loss": 3.3063,
      "step": 13400
    },
    {
      "epoch": 1.9113258490834162,
      "grad_norm": 1.451190710067749,
      "learning_rate": 0.0004061222222222222,
      "loss": 3.3367,
      "step": 13450
    },
    {
      "epoch": 1.9184311496376296,
      "grad_norm": 1.4469232559204102,
      "learning_rate": 0.0004055666666666667,
      "loss": 3.3166,
      "step": 13500
    },
    {
      "epoch": 1.925536450191843,
      "grad_norm": 1.5176293849945068,
      "learning_rate": 0.00040501111111111113,
      "loss": 3.3052,
      "step": 13550
    },
    {
      "epoch": 1.9326417507460567,
      "grad_norm": 1.4334474802017212,
      "learning_rate": 0.00040445555555555557,
      "loss": 3.2966,
      "step": 13600
    },
    {
      "epoch": 1.93974705130027,
      "grad_norm": 1.3909268379211426,
      "learning_rate": 0.0004039,
      "loss": 3.2984,
      "step": 13650
    },
    {
      "epoch": 1.9468523518544836,
      "grad_norm": 1.4753868579864502,
      "learning_rate": 0.00040334444444444444,
      "loss": 3.2928,
      "step": 13700
    },
    {
      "epoch": 1.9539576524086968,
      "grad_norm": 1.414066195487976,
      "learning_rate": 0.0004027888888888889,
      "loss": 3.3084,
      "step": 13750
    },
    {
      "epoch": 1.9610629529629104,
      "grad_norm": 1.5064789056777954,
      "learning_rate": 0.0004022333333333333,
      "loss": 3.3257,
      "step": 13800
    },
    {
      "epoch": 1.9681682535171237,
      "grad_norm": 1.5383269786834717,
      "learning_rate": 0.0004016777777777778,
      "loss": 3.3141,
      "step": 13850
    },
    {
      "epoch": 1.9752735540713373,
      "grad_norm": 1.6095879077911377,
      "learning_rate": 0.00040112222222222224,
      "loss": 3.2907,
      "step": 13900
    },
    {
      "epoch": 1.9823788546255505,
      "grad_norm": 1.447340488433838,
      "learning_rate": 0.0004005666666666667,
      "loss": 3.26,
      "step": 13950
    },
    {
      "epoch": 1.9894841551797642,
      "grad_norm": 1.5441666841506958,
      "learning_rate": 0.0004000111111111111,
      "loss": 3.2894,
      "step": 14000
    },
    {
      "epoch": 1.9894841551797642,
      "eval_accuracy": 0.45595255494117737,
      "eval_loss": 3.245368719100952,
      "eval_runtime": 1.4692,
      "eval_samples_per_second": 2559.288,
      "eval_steps_per_second": 40.159,
      "step": 14000
    },
    {
      "epoch": 1.9965894557339774,
      "grad_norm": 1.566861629486084,
      "learning_rate": 0.00039945555555555556,
      "loss": 3.294,
      "step": 14050
    },
    {
      "epoch": 2.003694756288191,
      "grad_norm": 1.5748252868652344,
      "learning_rate": 0.0003989,
      "loss": 3.2552,
      "step": 14100
    },
    {
      "epoch": 2.0108000568424043,
      "grad_norm": 1.5415462255477905,
      "learning_rate": 0.00039834444444444443,
      "loss": 3.2492,
      "step": 14150
    },
    {
      "epoch": 2.017905357396618,
      "grad_norm": 1.5561825037002563,
      "learning_rate": 0.0003977888888888889,
      "loss": 3.2842,
      "step": 14200
    },
    {
      "epoch": 2.025010657950831,
      "grad_norm": 1.6461340188980103,
      "learning_rate": 0.00039723333333333336,
      "loss": 3.2462,
      "step": 14250
    },
    {
      "epoch": 2.032115958505045,
      "grad_norm": 1.5616611242294312,
      "learning_rate": 0.0003966777777777778,
      "loss": 3.2676,
      "step": 14300
    },
    {
      "epoch": 2.039221259059258,
      "grad_norm": 1.4216052293777466,
      "learning_rate": 0.00039612222222222223,
      "loss": 3.2811,
      "step": 14350
    },
    {
      "epoch": 2.0463265596134717,
      "grad_norm": 1.5665044784545898,
      "learning_rate": 0.00039556666666666667,
      "loss": 3.2648,
      "step": 14400
    },
    {
      "epoch": 2.0534318601676853,
      "grad_norm": 1.4758416414260864,
      "learning_rate": 0.0003950111111111111,
      "loss": 3.2494,
      "step": 14450
    },
    {
      "epoch": 2.0605371607218985,
      "grad_norm": 1.5897761583328247,
      "learning_rate": 0.00039445555555555554,
      "loss": 3.2562,
      "step": 14500
    },
    {
      "epoch": 2.067642461276112,
      "grad_norm": 1.5198686122894287,
      "learning_rate": 0.0003939,
      "loss": 3.2323,
      "step": 14550
    },
    {
      "epoch": 2.0747477618303254,
      "grad_norm": 1.5690261125564575,
      "learning_rate": 0.00039334444444444447,
      "loss": 3.2525,
      "step": 14600
    },
    {
      "epoch": 2.081853062384539,
      "grad_norm": 1.4182822704315186,
      "learning_rate": 0.0003927888888888889,
      "loss": 3.253,
      "step": 14650
    },
    {
      "epoch": 2.0889583629387523,
      "grad_norm": 1.4269342422485352,
      "learning_rate": 0.00039223333333333334,
      "loss": 3.2332,
      "step": 14700
    },
    {
      "epoch": 2.096063663492966,
      "grad_norm": 1.3798179626464844,
      "learning_rate": 0.0003916777777777778,
      "loss": 3.2357,
      "step": 14750
    },
    {
      "epoch": 2.103168964047179,
      "grad_norm": 1.5191919803619385,
      "learning_rate": 0.0003911222222222222,
      "loss": 3.2516,
      "step": 14800
    },
    {
      "epoch": 2.110274264601393,
      "grad_norm": 1.4717212915420532,
      "learning_rate": 0.00039056666666666666,
      "loss": 3.2508,
      "step": 14850
    },
    {
      "epoch": 2.117379565155606,
      "grad_norm": 1.468482494354248,
      "learning_rate": 0.0003900111111111111,
      "loss": 3.2272,
      "step": 14900
    },
    {
      "epoch": 2.1244848657098196,
      "grad_norm": 1.424468994140625,
      "learning_rate": 0.0003894555555555556,
      "loss": 3.2288,
      "step": 14950
    },
    {
      "epoch": 2.131590166264033,
      "grad_norm": 1.4754891395568848,
      "learning_rate": 0.0003889,
      "loss": 3.2092,
      "step": 15000
    },
    {
      "epoch": 2.1386954668182465,
      "grad_norm": 1.6779147386550903,
      "learning_rate": 0.00038834444444444446,
      "loss": 3.2062,
      "step": 15050
    },
    {
      "epoch": 2.1458007673724597,
      "grad_norm": 1.3882695436477661,
      "learning_rate": 0.0003877888888888889,
      "loss": 3.23,
      "step": 15100
    },
    {
      "epoch": 2.1529060679266734,
      "grad_norm": 1.6576505899429321,
      "learning_rate": 0.00038723333333333333,
      "loss": 3.2042,
      "step": 15150
    },
    {
      "epoch": 2.1600113684808866,
      "grad_norm": 1.4277151823043823,
      "learning_rate": 0.00038667777777777777,
      "loss": 3.2246,
      "step": 15200
    },
    {
      "epoch": 2.1671166690351003,
      "grad_norm": 1.588564395904541,
      "learning_rate": 0.0003861222222222222,
      "loss": 3.2095,
      "step": 15250
    },
    {
      "epoch": 2.1742219695893135,
      "grad_norm": 1.5983752012252808,
      "learning_rate": 0.0003855666666666667,
      "loss": 3.22,
      "step": 15300
    },
    {
      "epoch": 2.181327270143527,
      "grad_norm": 1.5480470657348633,
      "learning_rate": 0.00038501111111111113,
      "loss": 3.1973,
      "step": 15350
    },
    {
      "epoch": 2.1884325706977403,
      "grad_norm": 1.7334848642349243,
      "learning_rate": 0.00038445555555555557,
      "loss": 3.1972,
      "step": 15400
    },
    {
      "epoch": 2.195537871251954,
      "grad_norm": 1.4713610410690308,
      "learning_rate": 0.0003839,
      "loss": 3.206,
      "step": 15450
    },
    {
      "epoch": 2.202643171806167,
      "grad_norm": 1.4612007141113281,
      "learning_rate": 0.00038334444444444444,
      "loss": 3.1949,
      "step": 15500
    },
    {
      "epoch": 2.209748472360381,
      "grad_norm": 1.5234096050262451,
      "learning_rate": 0.00038278888888888894,
      "loss": 3.1958,
      "step": 15550
    },
    {
      "epoch": 2.216853772914594,
      "grad_norm": 1.554410457611084,
      "learning_rate": 0.0003822333333333333,
      "loss": 3.2034,
      "step": 15600
    },
    {
      "epoch": 2.2239590734688077,
      "grad_norm": 1.4644205570220947,
      "learning_rate": 0.0003816777777777778,
      "loss": 3.2038,
      "step": 15650
    },
    {
      "epoch": 2.2310643740230214,
      "grad_norm": 1.5510737895965576,
      "learning_rate": 0.0003811222222222222,
      "loss": 3.1962,
      "step": 15700
    },
    {
      "epoch": 2.2381696745772346,
      "grad_norm": 1.564414381980896,
      "learning_rate": 0.0003805666666666667,
      "loss": 3.2068,
      "step": 15750
    },
    {
      "epoch": 2.2452749751314482,
      "grad_norm": 1.5432263612747192,
      "learning_rate": 0.0003800111111111111,
      "loss": 3.1981,
      "step": 15800
    },
    {
      "epoch": 2.2523802756856615,
      "grad_norm": 1.6288096904754639,
      "learning_rate": 0.00037945555555555556,
      "loss": 3.2062,
      "step": 15850
    },
    {
      "epoch": 2.259485576239875,
      "grad_norm": 1.5214354991912842,
      "learning_rate": 0.00037890000000000005,
      "loss": 3.194,
      "step": 15900
    },
    {
      "epoch": 2.2665908767940883,
      "grad_norm": 1.4612019062042236,
      "learning_rate": 0.00037834444444444443,
      "loss": 3.2283,
      "step": 15950
    },
    {
      "epoch": 2.273696177348302,
      "grad_norm": 1.5259031057357788,
      "learning_rate": 0.0003777888888888889,
      "loss": 3.173,
      "step": 16000
    },
    {
      "epoch": 2.273696177348302,
      "eval_accuracy": 0.4649949371814728,
      "eval_loss": 3.1757428646087646,
      "eval_runtime": 1.3251,
      "eval_samples_per_second": 2837.513,
      "eval_steps_per_second": 44.525,
      "step": 16000
    },
    {
      "epoch": 2.280801477902515,
      "grad_norm": 1.536083698272705,
      "learning_rate": 0.0003772333333333333,
      "loss": 3.1828,
      "step": 16050
    },
    {
      "epoch": 2.287906778456729,
      "grad_norm": 1.5883125066757202,
      "learning_rate": 0.0003766777777777778,
      "loss": 3.1711,
      "step": 16100
    },
    {
      "epoch": 2.295012079010942,
      "grad_norm": 1.5141252279281616,
      "learning_rate": 0.00037612222222222223,
      "loss": 3.1669,
      "step": 16150
    },
    {
      "epoch": 2.3021173795651557,
      "grad_norm": 1.5442508459091187,
      "learning_rate": 0.00037556666666666667,
      "loss": 3.1835,
      "step": 16200
    },
    {
      "epoch": 2.309222680119369,
      "grad_norm": 1.5498298406600952,
      "learning_rate": 0.00037501111111111116,
      "loss": 3.1566,
      "step": 16250
    },
    {
      "epoch": 2.3163279806735826,
      "grad_norm": 1.5234136581420898,
      "learning_rate": 0.00037445555555555554,
      "loss": 3.1623,
      "step": 16300
    },
    {
      "epoch": 2.323433281227796,
      "grad_norm": 1.557966947555542,
      "learning_rate": 0.00037390000000000004,
      "loss": 3.152,
      "step": 16350
    },
    {
      "epoch": 2.3305385817820095,
      "grad_norm": 1.641692042350769,
      "learning_rate": 0.0003733444444444444,
      "loss": 3.1508,
      "step": 16400
    },
    {
      "epoch": 2.3376438823362227,
      "grad_norm": 1.5023847818374634,
      "learning_rate": 0.0003727888888888889,
      "loss": 3.1556,
      "step": 16450
    },
    {
      "epoch": 2.3447491828904363,
      "grad_norm": 1.4382113218307495,
      "learning_rate": 0.00037223333333333335,
      "loss": 3.154,
      "step": 16500
    },
    {
      "epoch": 2.3518544834446495,
      "grad_norm": 1.6160063743591309,
      "learning_rate": 0.0003716777777777778,
      "loss": 3.1567,
      "step": 16550
    },
    {
      "epoch": 2.358959783998863,
      "grad_norm": 1.6019223928451538,
      "learning_rate": 0.0003711222222222223,
      "loss": 3.1607,
      "step": 16600
    },
    {
      "epoch": 2.3660650845530764,
      "grad_norm": 1.4011200666427612,
      "learning_rate": 0.00037056666666666666,
      "loss": 3.1499,
      "step": 16650
    },
    {
      "epoch": 2.37317038510729,
      "grad_norm": 1.4653880596160889,
      "learning_rate": 0.00037001111111111115,
      "loss": 3.1396,
      "step": 16700
    },
    {
      "epoch": 2.3802756856615037,
      "grad_norm": 1.5659642219543457,
      "learning_rate": 0.00036945555555555553,
      "loss": 3.1565,
      "step": 16750
    },
    {
      "epoch": 2.387380986215717,
      "grad_norm": 1.6067794561386108,
      "learning_rate": 0.0003689,
      "loss": 3.169,
      "step": 16800
    },
    {
      "epoch": 2.39448628676993,
      "grad_norm": 1.4000523090362549,
      "learning_rate": 0.0003683444444444444,
      "loss": 3.1505,
      "step": 16850
    },
    {
      "epoch": 2.401591587324144,
      "grad_norm": 1.4956333637237549,
      "learning_rate": 0.0003677888888888889,
      "loss": 3.1618,
      "step": 16900
    },
    {
      "epoch": 2.4086968878783575,
      "grad_norm": 1.564629316329956,
      "learning_rate": 0.0003672333333333334,
      "loss": 3.1462,
      "step": 16950
    },
    {
      "epoch": 2.4158021884325707,
      "grad_norm": 1.533850073814392,
      "learning_rate": 0.00036667777777777777,
      "loss": 3.1486,
      "step": 17000
    },
    {
      "epoch": 2.4229074889867843,
      "grad_norm": 1.4568897485733032,
      "learning_rate": 0.00036612222222222226,
      "loss": 3.138,
      "step": 17050
    },
    {
      "epoch": 2.4300127895409975,
      "grad_norm": 1.513048768043518,
      "learning_rate": 0.00036556666666666664,
      "loss": 3.1468,
      "step": 17100
    },
    {
      "epoch": 2.437118090095211,
      "grad_norm": 1.400328516960144,
      "learning_rate": 0.00036501111111111114,
      "loss": 3.1431,
      "step": 17150
    },
    {
      "epoch": 2.4442233906494244,
      "grad_norm": 1.554104208946228,
      "learning_rate": 0.0003644555555555555,
      "loss": 3.1236,
      "step": 17200
    },
    {
      "epoch": 2.451328691203638,
      "grad_norm": 1.4198516607284546,
      "learning_rate": 0.0003639,
      "loss": 3.1398,
      "step": 17250
    },
    {
      "epoch": 2.4584339917578513,
      "grad_norm": 1.5592104196548462,
      "learning_rate": 0.0003633444444444445,
      "loss": 3.1229,
      "step": 17300
    },
    {
      "epoch": 2.465539292312065,
      "grad_norm": 1.5739942789077759,
      "learning_rate": 0.0003627888888888889,
      "loss": 3.1334,
      "step": 17350
    },
    {
      "epoch": 2.472644592866278,
      "grad_norm": 1.6278862953186035,
      "learning_rate": 0.0003622333333333334,
      "loss": 3.1245,
      "step": 17400
    },
    {
      "epoch": 2.479749893420492,
      "grad_norm": 1.5226835012435913,
      "learning_rate": 0.00036167777777777776,
      "loss": 3.1179,
      "step": 17450
    },
    {
      "epoch": 2.486855193974705,
      "grad_norm": 1.4963643550872803,
      "learning_rate": 0.00036112222222222225,
      "loss": 3.1252,
      "step": 17500
    },
    {
      "epoch": 2.4939604945289187,
      "grad_norm": 1.571409821510315,
      "learning_rate": 0.00036056666666666663,
      "loss": 3.1178,
      "step": 17550
    },
    {
      "epoch": 2.501065795083132,
      "grad_norm": 1.5292372703552246,
      "learning_rate": 0.0003600111111111111,
      "loss": 3.1328,
      "step": 17600
    },
    {
      "epoch": 2.5081710956373455,
      "grad_norm": 1.5522916316986084,
      "learning_rate": 0.00035945555555555556,
      "loss": 3.1346,
      "step": 17650
    },
    {
      "epoch": 2.5152763961915587,
      "grad_norm": 1.526452660560608,
      "learning_rate": 0.0003589,
      "loss": 3.1153,
      "step": 17700
    },
    {
      "epoch": 2.5223816967457724,
      "grad_norm": 1.5139750242233276,
      "learning_rate": 0.0003583444444444445,
      "loss": 3.109,
      "step": 17750
    },
    {
      "epoch": 2.529486997299986,
      "grad_norm": 1.616378903388977,
      "learning_rate": 0.00035778888888888887,
      "loss": 3.1086,
      "step": 17800
    },
    {
      "epoch": 2.5365922978541993,
      "grad_norm": 1.4770593643188477,
      "learning_rate": 0.00035723333333333336,
      "loss": 3.11,
      "step": 17850
    },
    {
      "epoch": 2.5436975984084125,
      "grad_norm": 1.4987894296646118,
      "learning_rate": 0.00035667777777777774,
      "loss": 3.1083,
      "step": 17900
    },
    {
      "epoch": 2.550802898962626,
      "grad_norm": 1.5630767345428467,
      "learning_rate": 0.00035612222222222223,
      "loss": 3.1057,
      "step": 17950
    },
    {
      "epoch": 2.55790819951684,
      "grad_norm": 1.495004415512085,
      "learning_rate": 0.00035556666666666667,
      "loss": 3.0933,
      "step": 18000
    },
    {
      "epoch": 2.55790819951684,
      "eval_accuracy": 0.480011910200119,
      "eval_loss": 3.0557689666748047,
      "eval_runtime": 1.3066,
      "eval_samples_per_second": 2877.766,
      "eval_steps_per_second": 45.156,
      "step": 18000
    },
    {
      "epoch": 2.565013500071053,
      "grad_norm": 1.4845120906829834,
      "learning_rate": 0.0003550111111111111,
      "loss": 3.1214,
      "step": 18050
    },
    {
      "epoch": 2.572118800625266,
      "grad_norm": 1.5265833139419556,
      "learning_rate": 0.0003544555555555556,
      "loss": 3.1069,
      "step": 18100
    },
    {
      "epoch": 2.57922410117948,
      "grad_norm": 1.4948880672454834,
      "learning_rate": 0.0003539,
      "loss": 3.108,
      "step": 18150
    },
    {
      "epoch": 2.5863294017336935,
      "grad_norm": 1.5253654718399048,
      "learning_rate": 0.0003533444444444445,
      "loss": 3.0819,
      "step": 18200
    },
    {
      "epoch": 2.5934347022879067,
      "grad_norm": 1.5354139804840088,
      "learning_rate": 0.00035278888888888886,
      "loss": 3.0993,
      "step": 18250
    },
    {
      "epoch": 2.60054000284212,
      "grad_norm": 1.5608172416687012,
      "learning_rate": 0.00035223333333333335,
      "loss": 3.0908,
      "step": 18300
    },
    {
      "epoch": 2.6076453033963336,
      "grad_norm": 1.6600279808044434,
      "learning_rate": 0.0003516777777777778,
      "loss": 3.0814,
      "step": 18350
    },
    {
      "epoch": 2.6147506039505473,
      "grad_norm": 1.5104360580444336,
      "learning_rate": 0.0003511222222222222,
      "loss": 3.0731,
      "step": 18400
    },
    {
      "epoch": 2.6218559045047605,
      "grad_norm": 1.682814359664917,
      "learning_rate": 0.0003505666666666667,
      "loss": 3.0993,
      "step": 18450
    },
    {
      "epoch": 2.628961205058974,
      "grad_norm": 1.5391907691955566,
      "learning_rate": 0.0003500111111111111,
      "loss": 3.0771,
      "step": 18500
    },
    {
      "epoch": 2.6360665056131873,
      "grad_norm": 1.5016599893569946,
      "learning_rate": 0.0003494555555555556,
      "loss": 3.0913,
      "step": 18550
    },
    {
      "epoch": 2.643171806167401,
      "grad_norm": 1.6000250577926636,
      "learning_rate": 0.00034889999999999997,
      "loss": 3.1168,
      "step": 18600
    },
    {
      "epoch": 2.650277106721614,
      "grad_norm": 1.5141435861587524,
      "learning_rate": 0.00034834444444444446,
      "loss": 3.0673,
      "step": 18650
    },
    {
      "epoch": 2.657382407275828,
      "grad_norm": 1.5580344200134277,
      "learning_rate": 0.0003477888888888889,
      "loss": 3.0857,
      "step": 18700
    },
    {
      "epoch": 2.664487707830041,
      "grad_norm": 1.4756735563278198,
      "learning_rate": 0.00034723333333333333,
      "loss": 3.092,
      "step": 18750
    },
    {
      "epoch": 2.6715930083842547,
      "grad_norm": 1.4011656045913696,
      "learning_rate": 0.00034667777777777777,
      "loss": 3.0678,
      "step": 18800
    },
    {
      "epoch": 2.678698308938468,
      "grad_norm": 1.4854072332382202,
      "learning_rate": 0.0003461222222222222,
      "loss": 3.0864,
      "step": 18850
    },
    {
      "epoch": 2.6858036094926816,
      "grad_norm": 1.5279541015625,
      "learning_rate": 0.0003455666666666667,
      "loss": 3.0626,
      "step": 18900
    },
    {
      "epoch": 2.692908910046895,
      "grad_norm": 1.4168810844421387,
      "learning_rate": 0.0003450111111111111,
      "loss": 3.0822,
      "step": 18950
    },
    {
      "epoch": 2.7000142106011085,
      "grad_norm": 1.5052216053009033,
      "learning_rate": 0.0003444555555555556,
      "loss": 3.1007,
      "step": 19000
    },
    {
      "epoch": 2.707119511155322,
      "grad_norm": 1.524707317352295,
      "learning_rate": 0.0003439,
      "loss": 3.0528,
      "step": 19050
    },
    {
      "epoch": 2.7142248117095353,
      "grad_norm": 1.5324409008026123,
      "learning_rate": 0.00034334444444444445,
      "loss": 3.0779,
      "step": 19100
    },
    {
      "epoch": 2.7213301122637485,
      "grad_norm": 1.5060224533081055,
      "learning_rate": 0.0003427888888888889,
      "loss": 3.0828,
      "step": 19150
    },
    {
      "epoch": 2.728435412817962,
      "grad_norm": 1.4482327699661255,
      "learning_rate": 0.0003422333333333333,
      "loss": 3.0503,
      "step": 19200
    },
    {
      "epoch": 2.735540713372176,
      "grad_norm": 1.4725353717803955,
      "learning_rate": 0.0003416777777777778,
      "loss": 3.0714,
      "step": 19250
    },
    {
      "epoch": 2.742646013926389,
      "grad_norm": 1.528380274772644,
      "learning_rate": 0.0003411222222222222,
      "loss": 3.0669,
      "step": 19300
    },
    {
      "epoch": 2.7497513144806023,
      "grad_norm": 1.517346739768982,
      "learning_rate": 0.0003405666666666667,
      "loss": 3.0601,
      "step": 19350
    },
    {
      "epoch": 2.756856615034816,
      "grad_norm": 1.5343750715255737,
      "learning_rate": 0.0003400111111111111,
      "loss": 3.0632,
      "step": 19400
    },
    {
      "epoch": 2.7639619155890296,
      "grad_norm": 1.5137101411819458,
      "learning_rate": 0.00033945555555555556,
      "loss": 3.0508,
      "step": 19450
    },
    {
      "epoch": 2.771067216143243,
      "grad_norm": 1.4996198415756226,
      "learning_rate": 0.0003389,
      "loss": 3.0681,
      "step": 19500
    },
    {
      "epoch": 2.7781725166974565,
      "grad_norm": 1.4925514459609985,
      "learning_rate": 0.00033834444444444443,
      "loss": 3.0629,
      "step": 19550
    },
    {
      "epoch": 2.7852778172516697,
      "grad_norm": 1.5612093210220337,
      "learning_rate": 0.0003377888888888889,
      "loss": 3.0824,
      "step": 19600
    },
    {
      "epoch": 2.7923831178058833,
      "grad_norm": 1.442588210105896,
      "learning_rate": 0.00033723333333333336,
      "loss": 3.054,
      "step": 19650
    },
    {
      "epoch": 2.7994884183600965,
      "grad_norm": 1.6014519929885864,
      "learning_rate": 0.0003366777777777778,
      "loss": 3.0527,
      "step": 19700
    },
    {
      "epoch": 2.80659371891431,
      "grad_norm": 1.5083961486816406,
      "learning_rate": 0.00033612222222222224,
      "loss": 3.0582,
      "step": 19750
    },
    {
      "epoch": 2.8136990194685234,
      "grad_norm": 1.504754662513733,
      "learning_rate": 0.0003355666666666667,
      "loss": 3.0809,
      "step": 19800
    },
    {
      "epoch": 2.820804320022737,
      "grad_norm": 1.5361518859863281,
      "learning_rate": 0.0003350111111111111,
      "loss": 3.0485,
      "step": 19850
    },
    {
      "epoch": 2.8279096205769503,
      "grad_norm": 1.5289889574050903,
      "learning_rate": 0.00033445555555555555,
      "loss": 3.057,
      "step": 19900
    },
    {
      "epoch": 2.835014921131164,
      "grad_norm": 1.4377233982086182,
      "learning_rate": 0.0003339,
      "loss": 3.0319,
      "step": 19950
    },
    {
      "epoch": 2.842120221685377,
      "grad_norm": 1.4774348735809326,
      "learning_rate": 0.0003333444444444445,
      "loss": 3.0391,
      "step": 20000
    },
    {
      "epoch": 2.842120221685377,
      "eval_accuracy": 0.4836238920688629,
      "eval_loss": 2.9734725952148438,
      "eval_runtime": 1.3307,
      "eval_samples_per_second": 2825.609,
      "eval_steps_per_second": 44.338,
      "step": 20000
    },
    {
      "epoch": 2.849225522239591,
      "grad_norm": 1.490995168685913,
      "learning_rate": 0.0003327888888888889,
      "loss": 3.03,
      "step": 20050
    },
    {
      "epoch": 2.856330822793804,
      "grad_norm": 1.5990046262741089,
      "learning_rate": 0.00033223333333333335,
      "loss": 3.0526,
      "step": 20100
    },
    {
      "epoch": 2.8634361233480177,
      "grad_norm": 1.4851841926574707,
      "learning_rate": 0.0003316777777777778,
      "loss": 3.038,
      "step": 20150
    },
    {
      "epoch": 2.870541423902231,
      "grad_norm": 1.617862582206726,
      "learning_rate": 0.0003311222222222222,
      "loss": 3.0544,
      "step": 20200
    },
    {
      "epoch": 2.8776467244564445,
      "grad_norm": 1.5293192863464355,
      "learning_rate": 0.00033056666666666666,
      "loss": 3.0331,
      "step": 20250
    },
    {
      "epoch": 2.884752025010658,
      "grad_norm": 1.5155032873153687,
      "learning_rate": 0.0003300111111111111,
      "loss": 3.0372,
      "step": 20300
    },
    {
      "epoch": 2.8918573255648714,
      "grad_norm": 1.4643704891204834,
      "learning_rate": 0.0003294555555555556,
      "loss": 3.0362,
      "step": 20350
    },
    {
      "epoch": 2.8989626261190846,
      "grad_norm": 1.4701436758041382,
      "learning_rate": 0.0003289,
      "loss": 3.038,
      "step": 20400
    },
    {
      "epoch": 2.9060679266732983,
      "grad_norm": 1.4922205209732056,
      "learning_rate": 0.00032834444444444446,
      "loss": 3.0214,
      "step": 20450
    },
    {
      "epoch": 2.913173227227512,
      "grad_norm": 1.5764334201812744,
      "learning_rate": 0.0003277888888888889,
      "loss": 3.0282,
      "step": 20500
    },
    {
      "epoch": 2.920278527781725,
      "grad_norm": 1.4649850130081177,
      "learning_rate": 0.00032723333333333334,
      "loss": 3.0165,
      "step": 20550
    },
    {
      "epoch": 2.9273838283359384,
      "grad_norm": 1.5053377151489258,
      "learning_rate": 0.0003266777777777778,
      "loss": 3.0112,
      "step": 20600
    },
    {
      "epoch": 2.934489128890152,
      "grad_norm": 1.5205187797546387,
      "learning_rate": 0.0003261222222222222,
      "loss": 3.0337,
      "step": 20650
    },
    {
      "epoch": 2.9415944294443657,
      "grad_norm": 1.6407885551452637,
      "learning_rate": 0.0003255666666666667,
      "loss": 3.0087,
      "step": 20700
    },
    {
      "epoch": 2.948699729998579,
      "grad_norm": 1.6242375373840332,
      "learning_rate": 0.0003250111111111111,
      "loss": 3.0223,
      "step": 20750
    },
    {
      "epoch": 2.9558050305527925,
      "grad_norm": 1.6333444118499756,
      "learning_rate": 0.0003244555555555556,
      "loss": 3.0213,
      "step": 20800
    },
    {
      "epoch": 2.9629103311070057,
      "grad_norm": 1.5015469789505005,
      "learning_rate": 0.0003239,
      "loss": 3.024,
      "step": 20850
    },
    {
      "epoch": 2.9700156316612194,
      "grad_norm": 1.5481396913528442,
      "learning_rate": 0.00032334444444444445,
      "loss": 2.997,
      "step": 20900
    },
    {
      "epoch": 2.9771209322154326,
      "grad_norm": 1.5252339839935303,
      "learning_rate": 0.0003227888888888889,
      "loss": 3.0115,
      "step": 20950
    },
    {
      "epoch": 2.9842262327696463,
      "grad_norm": 1.524375081062317,
      "learning_rate": 0.0003222333333333333,
      "loss": 3.0242,
      "step": 21000
    },
    {
      "epoch": 2.9913315333238595,
      "grad_norm": 1.471494197845459,
      "learning_rate": 0.0003216777777777778,
      "loss": 3.0049,
      "step": 21050
    },
    {
      "epoch": 2.998436833878073,
      "grad_norm": 1.4367444515228271,
      "learning_rate": 0.0003211222222222222,
      "loss": 3.0026,
      "step": 21100
    },
    {
      "epoch": 3.0055421344322863,
      "grad_norm": 1.6314510107040405,
      "learning_rate": 0.0003205666666666667,
      "loss": 2.9714,
      "step": 21150
    },
    {
      "epoch": 3.0126474349865,
      "grad_norm": 1.5607901811599731,
      "learning_rate": 0.0003200111111111111,
      "loss": 2.9827,
      "step": 21200
    },
    {
      "epoch": 3.019752735540713,
      "grad_norm": 1.5266896486282349,
      "learning_rate": 0.00031945555555555556,
      "loss": 2.9824,
      "step": 21250
    },
    {
      "epoch": 3.026858036094927,
      "grad_norm": 1.453753113746643,
      "learning_rate": 0.0003189,
      "loss": 3.0012,
      "step": 21300
    },
    {
      "epoch": 3.03396333664914,
      "grad_norm": 1.618914246559143,
      "learning_rate": 0.00031834444444444444,
      "loss": 2.996,
      "step": 21350
    },
    {
      "epoch": 3.0410686372033537,
      "grad_norm": 1.5350817441940308,
      "learning_rate": 0.00031778888888888893,
      "loss": 2.9959,
      "step": 21400
    },
    {
      "epoch": 3.048173937757567,
      "grad_norm": 1.5359801054000854,
      "learning_rate": 0.0003172333333333333,
      "loss": 3.006,
      "step": 21450
    },
    {
      "epoch": 3.0552792383117806,
      "grad_norm": 1.571722388267517,
      "learning_rate": 0.0003166777777777778,
      "loss": 2.965,
      "step": 21500
    },
    {
      "epoch": 3.062384538865994,
      "grad_norm": 1.5369782447814941,
      "learning_rate": 0.00031612222222222224,
      "loss": 2.9839,
      "step": 21550
    },
    {
      "epoch": 3.0694898394202075,
      "grad_norm": 1.4015165567398071,
      "learning_rate": 0.0003155666666666667,
      "loss": 2.9783,
      "step": 21600
    },
    {
      "epoch": 3.0765951399744207,
      "grad_norm": 1.4333746433258057,
      "learning_rate": 0.0003150111111111111,
      "loss": 2.9957,
      "step": 21650
    },
    {
      "epoch": 3.0837004405286343,
      "grad_norm": 1.555273175239563,
      "learning_rate": 0.00031445555555555555,
      "loss": 2.9952,
      "step": 21700
    },
    {
      "epoch": 3.090805741082848,
      "grad_norm": 1.4734160900115967,
      "learning_rate": 0.00031390000000000004,
      "loss": 2.958,
      "step": 21750
    },
    {
      "epoch": 3.097911041637061,
      "grad_norm": 1.5799846649169922,
      "learning_rate": 0.0003133444444444444,
      "loss": 2.979,
      "step": 21800
    },
    {
      "epoch": 3.105016342191275,
      "grad_norm": 1.608751654624939,
      "learning_rate": 0.0003127888888888889,
      "loss": 3.0088,
      "step": 21850
    },
    {
      "epoch": 3.112121642745488,
      "grad_norm": 1.4913153648376465,
      "learning_rate": 0.0003122333333333333,
      "loss": 2.9614,
      "step": 21900
    },
    {
      "epoch": 3.1192269432997017,
      "grad_norm": 1.4407751560211182,
      "learning_rate": 0.0003116777777777778,
      "loss": 2.9573,
      "step": 21950
    },
    {
      "epoch": 3.126332243853915,
      "grad_norm": 1.4962221384048462,
      "learning_rate": 0.0003111222222222222,
      "loss": 2.9946,
      "step": 22000
    },
    {
      "epoch": 3.126332243853915,
      "eval_accuracy": 0.49943476915359497,
      "eval_loss": 2.9075963497161865,
      "eval_runtime": 1.3119,
      "eval_samples_per_second": 2865.999,
      "eval_steps_per_second": 44.972,
      "step": 22000
    },
    {
      "epoch": 3.1334375444081286,
      "grad_norm": 1.5972894430160522,
      "learning_rate": 0.00031056666666666666,
      "loss": 2.9793,
      "step": 22050
    },
    {
      "epoch": 3.140542844962342,
      "grad_norm": 1.4097793102264404,
      "learning_rate": 0.00031001111111111115,
      "loss": 2.9575,
      "step": 22100
    },
    {
      "epoch": 3.1476481455165555,
      "grad_norm": 1.4834725856781006,
      "learning_rate": 0.00030945555555555554,
      "loss": 2.9657,
      "step": 22150
    },
    {
      "epoch": 3.1547534460707687,
      "grad_norm": 1.5428999662399292,
      "learning_rate": 0.00030890000000000003,
      "loss": 2.9836,
      "step": 22200
    },
    {
      "epoch": 3.1618587466249823,
      "grad_norm": 1.4695510864257812,
      "learning_rate": 0.0003083444444444444,
      "loss": 2.9748,
      "step": 22250
    },
    {
      "epoch": 3.1689640471791956,
      "grad_norm": 1.6710104942321777,
      "learning_rate": 0.0003077888888888889,
      "loss": 2.9663,
      "step": 22300
    },
    {
      "epoch": 3.176069347733409,
      "grad_norm": 1.473408818244934,
      "learning_rate": 0.00030723333333333334,
      "loss": 2.9748,
      "step": 22350
    },
    {
      "epoch": 3.1831746482876224,
      "grad_norm": 1.5751774311065674,
      "learning_rate": 0.0003066777777777778,
      "loss": 2.9596,
      "step": 22400
    },
    {
      "epoch": 3.190279948841836,
      "grad_norm": 1.4606789350509644,
      "learning_rate": 0.00030612222222222227,
      "loss": 2.972,
      "step": 22450
    },
    {
      "epoch": 3.1973852493960493,
      "grad_norm": 1.5335890054702759,
      "learning_rate": 0.00030556666666666665,
      "loss": 2.9667,
      "step": 22500
    },
    {
      "epoch": 3.204490549950263,
      "grad_norm": 1.5628565549850464,
      "learning_rate": 0.00030501111111111114,
      "loss": 2.9691,
      "step": 22550
    },
    {
      "epoch": 3.211595850504476,
      "grad_norm": 1.4582511186599731,
      "learning_rate": 0.0003044555555555555,
      "loss": 2.9518,
      "step": 22600
    },
    {
      "epoch": 3.21870115105869,
      "grad_norm": 1.4858638048171997,
      "learning_rate": 0.0003039,
      "loss": 2.9607,
      "step": 22650
    },
    {
      "epoch": 3.225806451612903,
      "grad_norm": 1.5188275575637817,
      "learning_rate": 0.0003033444444444445,
      "loss": 2.9544,
      "step": 22700
    },
    {
      "epoch": 3.2329117521671167,
      "grad_norm": 1.564578652381897,
      "learning_rate": 0.0003027888888888889,
      "loss": 2.9541,
      "step": 22750
    },
    {
      "epoch": 3.2400170527213303,
      "grad_norm": 1.453080177307129,
      "learning_rate": 0.0003022333333333334,
      "loss": 2.9404,
      "step": 22800
    },
    {
      "epoch": 3.2471223532755435,
      "grad_norm": 1.4518792629241943,
      "learning_rate": 0.00030167777777777776,
      "loss": 2.9623,
      "step": 22850
    },
    {
      "epoch": 3.2542276538297568,
      "grad_norm": 1.548088550567627,
      "learning_rate": 0.00030112222222222225,
      "loss": 2.9484,
      "step": 22900
    },
    {
      "epoch": 3.2613329543839704,
      "grad_norm": 1.5216915607452393,
      "learning_rate": 0.00030056666666666664,
      "loss": 2.9565,
      "step": 22950
    },
    {
      "epoch": 3.268438254938184,
      "grad_norm": 1.6633533239364624,
      "learning_rate": 0.00030001111111111113,
      "loss": 2.9563,
      "step": 23000
    },
    {
      "epoch": 3.2755435554923973,
      "grad_norm": 1.429800271987915,
      "learning_rate": 0.0002994555555555555,
      "loss": 2.9477,
      "step": 23050
    },
    {
      "epoch": 3.282648856046611,
      "grad_norm": 1.6086186170578003,
      "learning_rate": 0.0002989,
      "loss": 2.9605,
      "step": 23100
    },
    {
      "epoch": 3.289754156600824,
      "grad_norm": 1.481373906135559,
      "learning_rate": 0.0002983444444444445,
      "loss": 2.9466,
      "step": 23150
    },
    {
      "epoch": 3.296859457155038,
      "grad_norm": 1.5376640558242798,
      "learning_rate": 0.0002977888888888889,
      "loss": 2.9213,
      "step": 23200
    },
    {
      "epoch": 3.303964757709251,
      "grad_norm": 1.5091279745101929,
      "learning_rate": 0.00029723333333333337,
      "loss": 2.9449,
      "step": 23250
    },
    {
      "epoch": 3.3110700582634647,
      "grad_norm": 1.5239790678024292,
      "learning_rate": 0.00029667777777777775,
      "loss": 2.9451,
      "step": 23300
    },
    {
      "epoch": 3.318175358817678,
      "grad_norm": 1.6087703704833984,
      "learning_rate": 0.00029612222222222224,
      "loss": 2.9639,
      "step": 23350
    },
    {
      "epoch": 3.3252806593718915,
      "grad_norm": 1.3892955780029297,
      "learning_rate": 0.0002955666666666667,
      "loss": 2.9194,
      "step": 23400
    },
    {
      "epoch": 3.3323859599261048,
      "grad_norm": 1.447771668434143,
      "learning_rate": 0.0002950111111111111,
      "loss": 2.9282,
      "step": 23450
    },
    {
      "epoch": 3.3394912604803184,
      "grad_norm": 1.476275086402893,
      "learning_rate": 0.0002944555555555556,
      "loss": 2.9522,
      "step": 23500
    },
    {
      "epoch": 3.3465965610345316,
      "grad_norm": 1.6364308595657349,
      "learning_rate": 0.0002939,
      "loss": 2.9321,
      "step": 23550
    },
    {
      "epoch": 3.3537018615887453,
      "grad_norm": 1.6567586660385132,
      "learning_rate": 0.0002933444444444445,
      "loss": 2.9454,
      "step": 23600
    },
    {
      "epoch": 3.3608071621429585,
      "grad_norm": 1.5751633644104004,
      "learning_rate": 0.00029278888888888886,
      "loss": 2.942,
      "step": 23650
    },
    {
      "epoch": 3.367912462697172,
      "grad_norm": 1.575961947441101,
      "learning_rate": 0.00029223333333333335,
      "loss": 2.942,
      "step": 23700
    },
    {
      "epoch": 3.3750177632513854,
      "grad_norm": 1.5118695497512817,
      "learning_rate": 0.0002916777777777778,
      "loss": 2.9382,
      "step": 23750
    },
    {
      "epoch": 3.382123063805599,
      "grad_norm": 1.5508843660354614,
      "learning_rate": 0.00029112222222222223,
      "loss": 2.9573,
      "step": 23800
    },
    {
      "epoch": 3.3892283643598122,
      "grad_norm": 1.5142732858657837,
      "learning_rate": 0.00029056666666666666,
      "loss": 2.9448,
      "step": 23850
    },
    {
      "epoch": 3.396333664914026,
      "grad_norm": 1.4622957706451416,
      "learning_rate": 0.0002900111111111111,
      "loss": 2.938,
      "step": 23900
    },
    {
      "epoch": 3.403438965468239,
      "grad_norm": 1.4186699390411377,
      "learning_rate": 0.0002894555555555556,
      "loss": 2.9116,
      "step": 23950
    },
    {
      "epoch": 3.4105442660224528,
      "grad_norm": 1.5300025939941406,
      "learning_rate": 0.0002889,
      "loss": 2.9046,
      "step": 24000
    },
    {
      "epoch": 3.4105442660224528,
      "eval_accuracy": 0.4978906214237213,
      "eval_loss": 2.8925819396972656,
      "eval_runtime": 1.3351,
      "eval_samples_per_second": 2816.374,
      "eval_steps_per_second": 44.193,
      "step": 24000
    },
    {
      "epoch": 3.4176495665766664,
      "grad_norm": 1.4906315803527832,
      "learning_rate": 0.00028834444444444447,
      "loss": 2.936,
      "step": 24050
    },
    {
      "epoch": 3.4247548671308796,
      "grad_norm": 1.5187666416168213,
      "learning_rate": 0.0002877888888888889,
      "loss": 2.8953,
      "step": 24100
    },
    {
      "epoch": 3.431860167685093,
      "grad_norm": 1.4870028495788574,
      "learning_rate": 0.00028723333333333334,
      "loss": 2.923,
      "step": 24150
    },
    {
      "epoch": 3.4389654682393065,
      "grad_norm": 1.4789930582046509,
      "learning_rate": 0.0002866777777777778,
      "loss": 2.9348,
      "step": 24200
    },
    {
      "epoch": 3.44607076879352,
      "grad_norm": 1.6222258806228638,
      "learning_rate": 0.0002861222222222222,
      "loss": 2.9264,
      "step": 24250
    },
    {
      "epoch": 3.4531760693477334,
      "grad_norm": 1.4804177284240723,
      "learning_rate": 0.0002855666666666667,
      "loss": 2.9254,
      "step": 24300
    },
    {
      "epoch": 3.460281369901947,
      "grad_norm": 1.4715673923492432,
      "learning_rate": 0.0002850111111111111,
      "loss": 2.9074,
      "step": 24350
    },
    {
      "epoch": 3.4673866704561602,
      "grad_norm": 1.4322458505630493,
      "learning_rate": 0.0002844555555555556,
      "loss": 2.9132,
      "step": 24400
    },
    {
      "epoch": 3.474491971010374,
      "grad_norm": 1.4830477237701416,
      "learning_rate": 0.0002839,
      "loss": 2.9204,
      "step": 24450
    },
    {
      "epoch": 3.481597271564587,
      "grad_norm": 1.5052211284637451,
      "learning_rate": 0.00028334444444444445,
      "loss": 2.9064,
      "step": 24500
    },
    {
      "epoch": 3.4887025721188007,
      "grad_norm": 1.7150460481643677,
      "learning_rate": 0.0002827888888888889,
      "loss": 2.9238,
      "step": 24550
    },
    {
      "epoch": 3.495807872673014,
      "grad_norm": 1.4955459833145142,
      "learning_rate": 0.0002822333333333333,
      "loss": 2.9269,
      "step": 24600
    },
    {
      "epoch": 3.5029131732272276,
      "grad_norm": 1.5976444482803345,
      "learning_rate": 0.0002816777777777778,
      "loss": 2.9462,
      "step": 24650
    },
    {
      "epoch": 3.510018473781441,
      "grad_norm": 1.6254138946533203,
      "learning_rate": 0.0002811222222222222,
      "loss": 2.898,
      "step": 24700
    },
    {
      "epoch": 3.5171237743356545,
      "grad_norm": 1.4726321697235107,
      "learning_rate": 0.0002805666666666667,
      "loss": 2.9065,
      "step": 24750
    },
    {
      "epoch": 3.5242290748898677,
      "grad_norm": 1.5584523677825928,
      "learning_rate": 0.00028001111111111113,
      "loss": 2.9357,
      "step": 24800
    },
    {
      "epoch": 3.5313343754440814,
      "grad_norm": 1.4915764331817627,
      "learning_rate": 0.00027945555555555557,
      "loss": 2.9165,
      "step": 24850
    },
    {
      "epoch": 3.5384396759982946,
      "grad_norm": 1.5386276245117188,
      "learning_rate": 0.0002789,
      "loss": 2.9232,
      "step": 24900
    },
    {
      "epoch": 3.545544976552508,
      "grad_norm": 1.5589542388916016,
      "learning_rate": 0.00027834444444444444,
      "loss": 2.9032,
      "step": 24950
    },
    {
      "epoch": 3.5526502771067214,
      "grad_norm": 1.4987221956253052,
      "learning_rate": 0.0002777888888888889,
      "loss": 2.9246,
      "step": 25000
    },
    {
      "epoch": 3.559755577660935,
      "grad_norm": 1.4958397150039673,
      "learning_rate": 0.0002772333333333333,
      "loss": 2.8938,
      "step": 25050
    },
    {
      "epoch": 3.5668608782151487,
      "grad_norm": 1.5078121423721313,
      "learning_rate": 0.0002766777777777778,
      "loss": 2.9069,
      "step": 25100
    },
    {
      "epoch": 3.573966178769362,
      "grad_norm": 1.5042097568511963,
      "learning_rate": 0.00027612222222222224,
      "loss": 2.8918,
      "step": 25150
    },
    {
      "epoch": 3.581071479323575,
      "grad_norm": 1.671631932258606,
      "learning_rate": 0.0002755666666666667,
      "loss": 2.9023,
      "step": 25200
    },
    {
      "epoch": 3.588176779877789,
      "grad_norm": 1.4698235988616943,
      "learning_rate": 0.0002750111111111111,
      "loss": 2.883,
      "step": 25250
    },
    {
      "epoch": 3.5952820804320025,
      "grad_norm": 1.4968258142471313,
      "learning_rate": 0.00027445555555555555,
      "loss": 2.9238,
      "step": 25300
    },
    {
      "epoch": 3.6023873809862157,
      "grad_norm": 1.4356114864349365,
      "learning_rate": 0.0002739,
      "loss": 2.9126,
      "step": 25350
    },
    {
      "epoch": 3.609492681540429,
      "grad_norm": 1.5779649019241333,
      "learning_rate": 0.0002733444444444444,
      "loss": 2.9067,
      "step": 25400
    },
    {
      "epoch": 3.6165979820946426,
      "grad_norm": 1.580443024635315,
      "learning_rate": 0.0002727888888888889,
      "loss": 2.8901,
      "step": 25450
    },
    {
      "epoch": 3.623703282648856,
      "grad_norm": 1.555375576019287,
      "learning_rate": 0.00027223333333333335,
      "loss": 2.9117,
      "step": 25500
    },
    {
      "epoch": 3.6308085832030694,
      "grad_norm": 1.5805009603500366,
      "learning_rate": 0.0002716777777777778,
      "loss": 2.8997,
      "step": 25550
    },
    {
      "epoch": 3.637913883757283,
      "grad_norm": 1.3847641944885254,
      "learning_rate": 0.00027112222222222223,
      "loss": 2.9138,
      "step": 25600
    },
    {
      "epoch": 3.6450191843114963,
      "grad_norm": 1.647619605064392,
      "learning_rate": 0.00027056666666666667,
      "loss": 2.8953,
      "step": 25650
    },
    {
      "epoch": 3.65212448486571,
      "grad_norm": 1.5242257118225098,
      "learning_rate": 0.0002700111111111111,
      "loss": 2.9007,
      "step": 25700
    },
    {
      "epoch": 3.659229785419923,
      "grad_norm": 1.59304940700531,
      "learning_rate": 0.00026945555555555554,
      "loss": 2.8855,
      "step": 25750
    },
    {
      "epoch": 3.666335085974137,
      "grad_norm": 1.4405932426452637,
      "learning_rate": 0.0002689,
      "loss": 2.8963,
      "step": 25800
    },
    {
      "epoch": 3.67344038652835,
      "grad_norm": 1.5832966566085815,
      "learning_rate": 0.00026834444444444447,
      "loss": 2.8782,
      "step": 25850
    },
    {
      "epoch": 3.6805456870825637,
      "grad_norm": 1.4739205837249756,
      "learning_rate": 0.0002677888888888889,
      "loss": 2.875,
      "step": 25900
    },
    {
      "epoch": 3.687650987636777,
      "grad_norm": 1.439565658569336,
      "learning_rate": 0.00026723333333333334,
      "loss": 2.8956,
      "step": 25950
    },
    {
      "epoch": 3.6947562881909906,
      "grad_norm": 1.4591790437698364,
      "learning_rate": 0.0002666777777777778,
      "loss": 2.88,
      "step": 26000
    },
    {
      "epoch": 3.6947562881909906,
      "eval_accuracy": 0.4965103268623352,
      "eval_loss": 2.9005048274993896,
      "eval_runtime": 1.4001,
      "eval_samples_per_second": 2685.439,
      "eval_steps_per_second": 42.139,
      "step": 26000
    },
    {
      "epoch": 3.7018615887452038,
      "grad_norm": 1.6253418922424316,
      "learning_rate": 0.0002661222222222222,
      "loss": 2.8801,
      "step": 26050
    },
    {
      "epoch": 3.7089668892994174,
      "grad_norm": 1.652553915977478,
      "learning_rate": 0.00026556666666666665,
      "loss": 2.8799,
      "step": 26100
    },
    {
      "epoch": 3.716072189853631,
      "grad_norm": 1.4921988248825073,
      "learning_rate": 0.0002650111111111111,
      "loss": 2.8733,
      "step": 26150
    },
    {
      "epoch": 3.7231774904078443,
      "grad_norm": 1.5398036241531372,
      "learning_rate": 0.0002644555555555556,
      "loss": 2.8914,
      "step": 26200
    },
    {
      "epoch": 3.7302827909620575,
      "grad_norm": 1.5585711002349854,
      "learning_rate": 0.0002639,
      "loss": 2.894,
      "step": 26250
    },
    {
      "epoch": 3.737388091516271,
      "grad_norm": 1.5460278987884521,
      "learning_rate": 0.00026334444444444445,
      "loss": 2.8801,
      "step": 26300
    },
    {
      "epoch": 3.744493392070485,
      "grad_norm": 1.5803145170211792,
      "learning_rate": 0.0002627888888888889,
      "loss": 2.8933,
      "step": 26350
    },
    {
      "epoch": 3.751598692624698,
      "grad_norm": 1.649328589439392,
      "learning_rate": 0.00026223333333333333,
      "loss": 2.8536,
      "step": 26400
    },
    {
      "epoch": 3.7587039931789112,
      "grad_norm": 1.4576084613800049,
      "learning_rate": 0.00026167777777777777,
      "loss": 2.8657,
      "step": 26450
    },
    {
      "epoch": 3.765809293733125,
      "grad_norm": 1.6992441415786743,
      "learning_rate": 0.0002611222222222222,
      "loss": 2.8739,
      "step": 26500
    },
    {
      "epoch": 3.7729145942873386,
      "grad_norm": 1.479893445968628,
      "learning_rate": 0.0002605666666666667,
      "loss": 2.865,
      "step": 26550
    },
    {
      "epoch": 3.7800198948415518,
      "grad_norm": 1.599707841873169,
      "learning_rate": 0.00026001111111111113,
      "loss": 2.8904,
      "step": 26600
    },
    {
      "epoch": 3.787125195395765,
      "grad_norm": 1.5905715227127075,
      "learning_rate": 0.00025945555555555557,
      "loss": 2.8701,
      "step": 26650
    },
    {
      "epoch": 3.7942304959499786,
      "grad_norm": 1.6773452758789062,
      "learning_rate": 0.0002589,
      "loss": 2.8836,
      "step": 26700
    },
    {
      "epoch": 3.8013357965041923,
      "grad_norm": 1.4900513887405396,
      "learning_rate": 0.00025834444444444444,
      "loss": 2.8706,
      "step": 26750
    },
    {
      "epoch": 3.8084410970584055,
      "grad_norm": 1.6035610437393188,
      "learning_rate": 0.00025778888888888893,
      "loss": 2.8387,
      "step": 26800
    },
    {
      "epoch": 3.815546397612619,
      "grad_norm": 1.5936367511749268,
      "learning_rate": 0.0002572333333333333,
      "loss": 2.8678,
      "step": 26850
    },
    {
      "epoch": 3.8226516981668324,
      "grad_norm": 1.4388103485107422,
      "learning_rate": 0.0002566777777777778,
      "loss": 2.8685,
      "step": 26900
    },
    {
      "epoch": 3.829756998721046,
      "grad_norm": 1.6475917100906372,
      "learning_rate": 0.0002561222222222222,
      "loss": 2.8578,
      "step": 26950
    },
    {
      "epoch": 3.8368622992752592,
      "grad_norm": 1.4207926988601685,
      "learning_rate": 0.0002555666666666667,
      "loss": 2.8609,
      "step": 27000
    },
    {
      "epoch": 3.843967599829473,
      "grad_norm": 1.626503586769104,
      "learning_rate": 0.0002550111111111111,
      "loss": 2.8567,
      "step": 27050
    },
    {
      "epoch": 3.851072900383686,
      "grad_norm": 1.7468626499176025,
      "learning_rate": 0.00025445555555555555,
      "loss": 2.8804,
      "step": 27100
    },
    {
      "epoch": 3.8581782009378998,
      "grad_norm": 1.525142788887024,
      "learning_rate": 0.00025390000000000005,
      "loss": 2.8795,
      "step": 27150
    },
    {
      "epoch": 3.865283501492113,
      "grad_norm": 1.6227000951766968,
      "learning_rate": 0.00025334444444444443,
      "loss": 2.872,
      "step": 27200
    },
    {
      "epoch": 3.8723888020463266,
      "grad_norm": 1.4426857233047485,
      "learning_rate": 0.0002527888888888889,
      "loss": 2.859,
      "step": 27250
    },
    {
      "epoch": 3.87949410260054,
      "grad_norm": 1.6314607858657837,
      "learning_rate": 0.0002522333333333333,
      "loss": 2.8617,
      "step": 27300
    },
    {
      "epoch": 3.8865994031547535,
      "grad_norm": 1.6732324361801147,
      "learning_rate": 0.0002516777777777778,
      "loss": 2.8533,
      "step": 27350
    },
    {
      "epoch": 3.893704703708967,
      "grad_norm": 1.671032190322876,
      "learning_rate": 0.00025112222222222223,
      "loss": 2.8521,
      "step": 27400
    },
    {
      "epoch": 3.9008100042631804,
      "grad_norm": 1.5386325120925903,
      "learning_rate": 0.00025056666666666667,
      "loss": 2.8558,
      "step": 27450
    },
    {
      "epoch": 3.9079153048173936,
      "grad_norm": 1.48641037940979,
      "learning_rate": 0.00025001111111111116,
      "loss": 2.8706,
      "step": 27500
    },
    {
      "epoch": 3.9150206053716072,
      "grad_norm": 1.487549901008606,
      "learning_rate": 0.0002494555555555556,
      "loss": 2.8681,
      "step": 27550
    },
    {
      "epoch": 3.922125905925821,
      "grad_norm": 1.5427532196044922,
      "learning_rate": 0.00024890000000000003,
      "loss": 2.8411,
      "step": 27600
    },
    {
      "epoch": 3.929231206480034,
      "grad_norm": 1.6081897020339966,
      "learning_rate": 0.00024834444444444447,
      "loss": 2.8598,
      "step": 27650
    },
    {
      "epoch": 3.9363365070342473,
      "grad_norm": 1.4773359298706055,
      "learning_rate": 0.0002477888888888889,
      "loss": 2.8975,
      "step": 27700
    },
    {
      "epoch": 3.943441807588461,
      "grad_norm": 1.6076672077178955,
      "learning_rate": 0.00024723333333333334,
      "loss": 2.8473,
      "step": 27750
    },
    {
      "epoch": 3.9505471081426746,
      "grad_norm": 1.5951896905899048,
      "learning_rate": 0.0002466777777777778,
      "loss": 2.8476,
      "step": 27800
    },
    {
      "epoch": 3.957652408696888,
      "grad_norm": 1.5349781513214111,
      "learning_rate": 0.0002461222222222222,
      "loss": 2.8614,
      "step": 27850
    },
    {
      "epoch": 3.964757709251101,
      "grad_norm": 1.531628131866455,
      "learning_rate": 0.00024556666666666665,
      "loss": 2.8488,
      "step": 27900
    },
    {
      "epoch": 3.9718630098053147,
      "grad_norm": 1.4372403621673584,
      "learning_rate": 0.00024501111111111115,
      "loss": 2.8677,
      "step": 27950
    },
    {
      "epoch": 3.9789683103595284,
      "grad_norm": 1.4930979013442993,
      "learning_rate": 0.0002444555555555556,
      "loss": 2.8408,
      "step": 28000
    },
    {
      "epoch": 3.9789683103595284,
      "eval_accuracy": 0.5038213729858398,
      "eval_loss": 2.8301947116851807,
      "eval_runtime": 1.3222,
      "eval_samples_per_second": 2843.851,
      "eval_steps_per_second": 44.624,
      "step": 28000
    },
    {
      "epoch": 3.9860736109137416,
      "grad_norm": 1.5813312530517578,
      "learning_rate": 0.00024390000000000002,
      "loss": 2.8482,
      "step": 28050
    },
    {
      "epoch": 3.9931789114679552,
      "grad_norm": 1.5875399112701416,
      "learning_rate": 0.00024334444444444446,
      "loss": 2.8717,
      "step": 28100
    },
    {
      "epoch": 4.000284212022168,
      "grad_norm": 1.5822824239730835,
      "learning_rate": 0.0002427888888888889,
      "loss": 2.8383,
      "step": 28150
    },
    {
      "epoch": 4.007389512576382,
      "grad_norm": 1.4386391639709473,
      "learning_rate": 0.00024223333333333333,
      "loss": 2.8187,
      "step": 28200
    },
    {
      "epoch": 4.014494813130596,
      "grad_norm": 1.545289158821106,
      "learning_rate": 0.00024167777777777777,
      "loss": 2.811,
      "step": 28250
    },
    {
      "epoch": 4.0216001136848085,
      "grad_norm": 1.6034023761749268,
      "learning_rate": 0.0002411222222222222,
      "loss": 2.8284,
      "step": 28300
    },
    {
      "epoch": 4.028705414239022,
      "grad_norm": 1.6022422313690186,
      "learning_rate": 0.0002405666666666667,
      "loss": 2.8463,
      "step": 28350
    },
    {
      "epoch": 4.035810714793236,
      "grad_norm": 1.7777578830718994,
      "learning_rate": 0.00024001111111111113,
      "loss": 2.8322,
      "step": 28400
    },
    {
      "epoch": 4.0429160153474495,
      "grad_norm": 1.4642590284347534,
      "learning_rate": 0.00023945555555555557,
      "loss": 2.8207,
      "step": 28450
    },
    {
      "epoch": 4.050021315901662,
      "grad_norm": 1.5625720024108887,
      "learning_rate": 0.0002389,
      "loss": 2.8108,
      "step": 28500
    },
    {
      "epoch": 4.057126616455876,
      "grad_norm": 1.5230220556259155,
      "learning_rate": 0.00023834444444444444,
      "loss": 2.8204,
      "step": 28550
    },
    {
      "epoch": 4.06423191701009,
      "grad_norm": 1.7173887491226196,
      "learning_rate": 0.00023778888888888888,
      "loss": 2.8146,
      "step": 28600
    },
    {
      "epoch": 4.071337217564303,
      "grad_norm": 1.551294207572937,
      "learning_rate": 0.00023723333333333332,
      "loss": 2.845,
      "step": 28650
    },
    {
      "epoch": 4.078442518118516,
      "grad_norm": 1.4647568464279175,
      "learning_rate": 0.00023667777777777778,
      "loss": 2.8286,
      "step": 28700
    },
    {
      "epoch": 4.08554781867273,
      "grad_norm": 1.5898845195770264,
      "learning_rate": 0.00023612222222222225,
      "loss": 2.8352,
      "step": 28750
    },
    {
      "epoch": 4.092653119226943,
      "grad_norm": 1.4554657936096191,
      "learning_rate": 0.00023556666666666668,
      "loss": 2.8263,
      "step": 28800
    },
    {
      "epoch": 4.099758419781157,
      "grad_norm": 1.603566288948059,
      "learning_rate": 0.00023501111111111112,
      "loss": 2.7992,
      "step": 28850
    },
    {
      "epoch": 4.106863720335371,
      "grad_norm": 1.590300440788269,
      "learning_rate": 0.00023445555555555556,
      "loss": 2.7964,
      "step": 28900
    },
    {
      "epoch": 4.113969020889583,
      "grad_norm": 1.7716232538223267,
      "learning_rate": 0.0002339,
      "loss": 2.8384,
      "step": 28950
    },
    {
      "epoch": 4.121074321443797,
      "grad_norm": 1.6027381420135498,
      "learning_rate": 0.00023334444444444443,
      "loss": 2.8064,
      "step": 29000
    },
    {
      "epoch": 4.128179621998011,
      "grad_norm": 1.6521331071853638,
      "learning_rate": 0.0002327888888888889,
      "loss": 2.8155,
      "step": 29050
    },
    {
      "epoch": 4.135284922552224,
      "grad_norm": 1.635908603668213,
      "learning_rate": 0.00023223333333333336,
      "loss": 2.8135,
      "step": 29100
    },
    {
      "epoch": 4.142390223106437,
      "grad_norm": 1.4520660638809204,
      "learning_rate": 0.0002316777777777778,
      "loss": 2.8221,
      "step": 29150
    },
    {
      "epoch": 4.149495523660651,
      "grad_norm": 1.57816743850708,
      "learning_rate": 0.00023112222222222223,
      "loss": 2.8149,
      "step": 29200
    },
    {
      "epoch": 4.156600824214864,
      "grad_norm": 1.5677634477615356,
      "learning_rate": 0.00023056666666666667,
      "loss": 2.8118,
      "step": 29250
    },
    {
      "epoch": 4.163706124769078,
      "grad_norm": 1.4617812633514404,
      "learning_rate": 0.0002300111111111111,
      "loss": 2.8092,
      "step": 29300
    },
    {
      "epoch": 4.170811425323291,
      "grad_norm": 1.501502513885498,
      "learning_rate": 0.00022945555555555554,
      "loss": 2.812,
      "step": 29350
    },
    {
      "epoch": 4.1779167258775045,
      "grad_norm": 1.6901543140411377,
      "learning_rate": 0.0002289,
      "loss": 2.821,
      "step": 29400
    },
    {
      "epoch": 4.185022026431718,
      "grad_norm": 1.5179857015609741,
      "learning_rate": 0.00022834444444444444,
      "loss": 2.8395,
      "step": 29450
    },
    {
      "epoch": 4.192127326985932,
      "grad_norm": 1.5568187236785889,
      "learning_rate": 0.0002277888888888889,
      "loss": 2.8194,
      "step": 29500
    },
    {
      "epoch": 4.199232627540145,
      "grad_norm": 1.584041714668274,
      "learning_rate": 0.00022723333333333335,
      "loss": 2.7991,
      "step": 29550
    },
    {
      "epoch": 4.206337928094358,
      "grad_norm": 1.479154109954834,
      "learning_rate": 0.00022667777777777778,
      "loss": 2.8233,
      "step": 29600
    },
    {
      "epoch": 4.213443228648572,
      "grad_norm": 1.4960718154907227,
      "learning_rate": 0.00022612222222222222,
      "loss": 2.7909,
      "step": 29650
    },
    {
      "epoch": 4.220548529202786,
      "grad_norm": 1.6419041156768799,
      "learning_rate": 0.00022556666666666668,
      "loss": 2.8214,
      "step": 29700
    },
    {
      "epoch": 4.227653829756998,
      "grad_norm": 1.555156946182251,
      "learning_rate": 0.00022501111111111112,
      "loss": 2.8272,
      "step": 29750
    },
    {
      "epoch": 4.234759130311212,
      "grad_norm": 1.6603648662567139,
      "learning_rate": 0.00022445555555555556,
      "loss": 2.8172,
      "step": 29800
    },
    {
      "epoch": 4.241864430865426,
      "grad_norm": 1.4937909841537476,
      "learning_rate": 0.0002239,
      "loss": 2.7813,
      "step": 29850
    },
    {
      "epoch": 4.248969731419639,
      "grad_norm": 1.5089703798294067,
      "learning_rate": 0.00022334444444444446,
      "loss": 2.8092,
      "step": 29900
    },
    {
      "epoch": 4.256075031973852,
      "grad_norm": 1.4582308530807495,
      "learning_rate": 0.0002227888888888889,
      "loss": 2.7991,
      "step": 29950
    },
    {
      "epoch": 4.263180332528066,
      "grad_norm": 1.507042646408081,
      "learning_rate": 0.00022223333333333333,
      "loss": 2.7974,
      "step": 30000
    },
    {
      "epoch": 4.263180332528066,
      "eval_accuracy": 0.5150346159934998,
      "eval_loss": 2.7642641067504883,
      "eval_runtime": 1.4549,
      "eval_samples_per_second": 2584.342,
      "eval_steps_per_second": 40.552,
      "step": 30000
    },
    {
      "epoch": 4.270285633082279,
      "grad_norm": 1.5285909175872803,
      "learning_rate": 0.0002216777777777778,
      "loss": 2.7813,
      "step": 30050
    },
    {
      "epoch": 4.277390933636493,
      "grad_norm": 1.5540330410003662,
      "learning_rate": 0.00022112222222222223,
      "loss": 2.7818,
      "step": 30100
    },
    {
      "epoch": 4.284496234190707,
      "grad_norm": 1.6137828826904297,
      "learning_rate": 0.00022056666666666667,
      "loss": 2.8132,
      "step": 30150
    },
    {
      "epoch": 4.2916015347449195,
      "grad_norm": 1.5205047130584717,
      "learning_rate": 0.0002200111111111111,
      "loss": 2.7866,
      "step": 30200
    },
    {
      "epoch": 4.298706835299133,
      "grad_norm": 1.541063904762268,
      "learning_rate": 0.00021945555555555554,
      "loss": 2.8089,
      "step": 30250
    },
    {
      "epoch": 4.305812135853347,
      "grad_norm": 1.6706923246383667,
      "learning_rate": 0.0002189,
      "loss": 2.7891,
      "step": 30300
    },
    {
      "epoch": 4.31291743640756,
      "grad_norm": 1.6725937128067017,
      "learning_rate": 0.00021834444444444445,
      "loss": 2.7965,
      "step": 30350
    },
    {
      "epoch": 4.320022736961773,
      "grad_norm": 1.731666922569275,
      "learning_rate": 0.0002177888888888889,
      "loss": 2.7659,
      "step": 30400
    },
    {
      "epoch": 4.327128037515987,
      "grad_norm": 1.5914294719696045,
      "learning_rate": 0.00021723333333333335,
      "loss": 2.7915,
      "step": 30450
    },
    {
      "epoch": 4.3342333380702005,
      "grad_norm": 1.5213390588760376,
      "learning_rate": 0.00021667777777777778,
      "loss": 2.8122,
      "step": 30500
    },
    {
      "epoch": 4.341338638624414,
      "grad_norm": 1.7755457162857056,
      "learning_rate": 0.00021612222222222222,
      "loss": 2.7972,
      "step": 30550
    },
    {
      "epoch": 4.348443939178627,
      "grad_norm": 1.6222538948059082,
      "learning_rate": 0.00021556666666666666,
      "loss": 2.7935,
      "step": 30600
    },
    {
      "epoch": 4.355549239732841,
      "grad_norm": 1.7443130016326904,
      "learning_rate": 0.0002150111111111111,
      "loss": 2.8111,
      "step": 30650
    },
    {
      "epoch": 4.362654540287054,
      "grad_norm": 1.5261310338974,
      "learning_rate": 0.00021445555555555556,
      "loss": 2.7848,
      "step": 30700
    },
    {
      "epoch": 4.369759840841268,
      "grad_norm": 1.5985714197158813,
      "learning_rate": 0.00021390000000000002,
      "loss": 2.7745,
      "step": 30750
    },
    {
      "epoch": 4.376865141395481,
      "grad_norm": 1.5596888065338135,
      "learning_rate": 0.00021334444444444446,
      "loss": 2.8015,
      "step": 30800
    },
    {
      "epoch": 4.383970441949694,
      "grad_norm": 1.4315415620803833,
      "learning_rate": 0.0002127888888888889,
      "loss": 2.7807,
      "step": 30850
    },
    {
      "epoch": 4.391075742503908,
      "grad_norm": 1.545339822769165,
      "learning_rate": 0.00021223333333333333,
      "loss": 2.7805,
      "step": 30900
    },
    {
      "epoch": 4.398181043058122,
      "grad_norm": 1.5493332147598267,
      "learning_rate": 0.00021167777777777777,
      "loss": 2.8105,
      "step": 30950
    },
    {
      "epoch": 4.405286343612334,
      "grad_norm": 1.536258339881897,
      "learning_rate": 0.0002111222222222222,
      "loss": 2.7994,
      "step": 31000
    },
    {
      "epoch": 4.412391644166548,
      "grad_norm": 1.4628187417984009,
      "learning_rate": 0.00021056666666666667,
      "loss": 2.7527,
      "step": 31050
    },
    {
      "epoch": 4.419496944720762,
      "grad_norm": 1.5581260919570923,
      "learning_rate": 0.00021001111111111114,
      "loss": 2.8013,
      "step": 31100
    },
    {
      "epoch": 4.426602245274975,
      "grad_norm": 1.6152867078781128,
      "learning_rate": 0.00020945555555555557,
      "loss": 2.8103,
      "step": 31150
    },
    {
      "epoch": 4.433707545829188,
      "grad_norm": 1.5480724573135376,
      "learning_rate": 0.0002089,
      "loss": 2.7777,
      "step": 31200
    },
    {
      "epoch": 4.440812846383402,
      "grad_norm": 1.5430995225906372,
      "learning_rate": 0.00020834444444444445,
      "loss": 2.7929,
      "step": 31250
    },
    {
      "epoch": 4.4479181469376154,
      "grad_norm": 1.5283119678497314,
      "learning_rate": 0.00020778888888888888,
      "loss": 2.779,
      "step": 31300
    },
    {
      "epoch": 4.455023447491829,
      "grad_norm": 1.5975980758666992,
      "learning_rate": 0.00020723333333333332,
      "loss": 2.8032,
      "step": 31350
    },
    {
      "epoch": 4.462128748046043,
      "grad_norm": 1.5572302341461182,
      "learning_rate": 0.00020667777777777776,
      "loss": 2.7722,
      "step": 31400
    },
    {
      "epoch": 4.4692340486002555,
      "grad_norm": 1.6391266584396362,
      "learning_rate": 0.00020612222222222225,
      "loss": 2.7831,
      "step": 31450
    },
    {
      "epoch": 4.476339349154469,
      "grad_norm": 1.6277772188186646,
      "learning_rate": 0.00020556666666666669,
      "loss": 2.7939,
      "step": 31500
    },
    {
      "epoch": 4.483444649708683,
      "grad_norm": 1.6991925239562988,
      "learning_rate": 0.00020501111111111112,
      "loss": 2.7588,
      "step": 31550
    },
    {
      "epoch": 4.4905499502628965,
      "grad_norm": 1.5235546827316284,
      "learning_rate": 0.00020445555555555556,
      "loss": 2.7738,
      "step": 31600
    },
    {
      "epoch": 4.497655250817109,
      "grad_norm": 1.751791000366211,
      "learning_rate": 0.0002039,
      "loss": 2.7764,
      "step": 31650
    },
    {
      "epoch": 4.504760551371323,
      "grad_norm": 1.6844935417175293,
      "learning_rate": 0.00020334444444444443,
      "loss": 2.7975,
      "step": 31700
    },
    {
      "epoch": 4.511865851925537,
      "grad_norm": 1.4950183629989624,
      "learning_rate": 0.0002027888888888889,
      "loss": 2.7504,
      "step": 31750
    },
    {
      "epoch": 4.51897115247975,
      "grad_norm": 1.6350730657577515,
      "learning_rate": 0.00020223333333333333,
      "loss": 2.797,
      "step": 31800
    },
    {
      "epoch": 4.526076453033963,
      "grad_norm": 1.5607668161392212,
      "learning_rate": 0.0002016777777777778,
      "loss": 2.7793,
      "step": 31850
    },
    {
      "epoch": 4.533181753588177,
      "grad_norm": 1.506750226020813,
      "learning_rate": 0.00020112222222222223,
      "loss": 2.7641,
      "step": 31900
    },
    {
      "epoch": 4.54028705414239,
      "grad_norm": 1.5720961093902588,
      "learning_rate": 0.00020056666666666667,
      "loss": 2.782,
      "step": 31950
    },
    {
      "epoch": 4.547392354696604,
      "grad_norm": 1.601806640625,
      "learning_rate": 0.0002000111111111111,
      "loss": 2.7733,
      "step": 32000
    },
    {
      "epoch": 4.547392354696604,
      "eval_accuracy": 0.5182931423187256,
      "eval_loss": 2.738097906112671,
      "eval_runtime": 1.2841,
      "eval_samples_per_second": 2928.059,
      "eval_steps_per_second": 45.946,
      "step": 32000
    },
    {
      "epoch": 4.554497655250817,
      "grad_norm": 1.5717216730117798,
      "learning_rate": 0.00019945555555555555,
      "loss": 2.7728,
      "step": 32050
    },
    {
      "epoch": 4.56160295580503,
      "grad_norm": 1.6468063592910767,
      "learning_rate": 0.0001989,
      "loss": 2.7679,
      "step": 32100
    },
    {
      "epoch": 4.568708256359244,
      "grad_norm": 1.6788794994354248,
      "learning_rate": 0.00019834444444444445,
      "loss": 2.7733,
      "step": 32150
    },
    {
      "epoch": 4.575813556913458,
      "grad_norm": 1.6205172538757324,
      "learning_rate": 0.00019778888888888888,
      "loss": 2.7692,
      "step": 32200
    },
    {
      "epoch": 4.582918857467671,
      "grad_norm": 1.57520592212677,
      "learning_rate": 0.00019723333333333335,
      "loss": 2.7715,
      "step": 32250
    },
    {
      "epoch": 4.590024158021884,
      "grad_norm": 1.5816936492919922,
      "learning_rate": 0.00019667777777777778,
      "loss": 2.767,
      "step": 32300
    },
    {
      "epoch": 4.597129458576098,
      "grad_norm": 1.518502116203308,
      "learning_rate": 0.00019612222222222222,
      "loss": 2.7713,
      "step": 32350
    },
    {
      "epoch": 4.604234759130311,
      "grad_norm": 1.516652226448059,
      "learning_rate": 0.00019556666666666666,
      "loss": 2.7631,
      "step": 32400
    },
    {
      "epoch": 4.611340059684524,
      "grad_norm": 1.6790320873260498,
      "learning_rate": 0.00019501111111111112,
      "loss": 2.7556,
      "step": 32450
    },
    {
      "epoch": 4.618445360238738,
      "grad_norm": 1.5499494075775146,
      "learning_rate": 0.00019445555555555556,
      "loss": 2.7683,
      "step": 32500
    },
    {
      "epoch": 4.6255506607929515,
      "grad_norm": 1.4426767826080322,
      "learning_rate": 0.0001939,
      "loss": 2.7845,
      "step": 32550
    },
    {
      "epoch": 4.632655961347165,
      "grad_norm": 1.5415791273117065,
      "learning_rate": 0.00019334444444444446,
      "loss": 2.7534,
      "step": 32600
    },
    {
      "epoch": 4.639761261901379,
      "grad_norm": 1.4372013807296753,
      "learning_rate": 0.0001927888888888889,
      "loss": 2.7741,
      "step": 32650
    },
    {
      "epoch": 4.646866562455592,
      "grad_norm": 1.555567979812622,
      "learning_rate": 0.00019223333333333333,
      "loss": 2.7633,
      "step": 32700
    },
    {
      "epoch": 4.653971863009805,
      "grad_norm": 1.7388957738876343,
      "learning_rate": 0.00019167777777777777,
      "loss": 2.7753,
      "step": 32750
    },
    {
      "epoch": 4.661077163564019,
      "grad_norm": 1.6249463558197021,
      "learning_rate": 0.00019112222222222224,
      "loss": 2.7339,
      "step": 32800
    },
    {
      "epoch": 4.668182464118233,
      "grad_norm": 1.642773985862732,
      "learning_rate": 0.00019056666666666667,
      "loss": 2.7721,
      "step": 32850
    },
    {
      "epoch": 4.675287764672445,
      "grad_norm": 1.544217824935913,
      "learning_rate": 0.0001900111111111111,
      "loss": 2.743,
      "step": 32900
    },
    {
      "epoch": 4.682393065226659,
      "grad_norm": 1.688503384590149,
      "learning_rate": 0.00018945555555555555,
      "loss": 2.7594,
      "step": 32950
    },
    {
      "epoch": 4.689498365780873,
      "grad_norm": 1.5570727586746216,
      "learning_rate": 0.0001889,
      "loss": 2.7632,
      "step": 33000
    },
    {
      "epoch": 4.696603666335086,
      "grad_norm": 1.4844149351119995,
      "learning_rate": 0.00018834444444444445,
      "loss": 2.7619,
      "step": 33050
    },
    {
      "epoch": 4.703708966889299,
      "grad_norm": 1.5461333990097046,
      "learning_rate": 0.00018778888888888888,
      "loss": 2.7787,
      "step": 33100
    },
    {
      "epoch": 4.710814267443513,
      "grad_norm": 1.6980046033859253,
      "learning_rate": 0.00018723333333333335,
      "loss": 2.7698,
      "step": 33150
    },
    {
      "epoch": 4.717919567997726,
      "grad_norm": 1.4724769592285156,
      "learning_rate": 0.00018667777777777779,
      "loss": 2.765,
      "step": 33200
    },
    {
      "epoch": 4.72502486855194,
      "grad_norm": 1.5083281993865967,
      "learning_rate": 0.00018612222222222222,
      "loss": 2.7679,
      "step": 33250
    },
    {
      "epoch": 4.732130169106153,
      "grad_norm": 1.7240538597106934,
      "learning_rate": 0.00018556666666666666,
      "loss": 2.7522,
      "step": 33300
    },
    {
      "epoch": 4.7392354696603665,
      "grad_norm": 1.508767008781433,
      "learning_rate": 0.0001850111111111111,
      "loss": 2.7485,
      "step": 33350
    },
    {
      "epoch": 4.74634077021458,
      "grad_norm": 1.636539101600647,
      "learning_rate": 0.00018445555555555556,
      "loss": 2.742,
      "step": 33400
    },
    {
      "epoch": 4.753446070768794,
      "grad_norm": 1.4749336242675781,
      "learning_rate": 0.00018390000000000002,
      "loss": 2.7663,
      "step": 33450
    },
    {
      "epoch": 4.760551371323007,
      "grad_norm": 1.504374623298645,
      "learning_rate": 0.00018334444444444446,
      "loss": 2.7579,
      "step": 33500
    },
    {
      "epoch": 4.76765667187722,
      "grad_norm": 1.668945550918579,
      "learning_rate": 0.0001827888888888889,
      "loss": 2.7532,
      "step": 33550
    },
    {
      "epoch": 4.774761972431434,
      "grad_norm": 1.6482501029968262,
      "learning_rate": 0.00018223333333333334,
      "loss": 2.7407,
      "step": 33600
    },
    {
      "epoch": 4.7818672729856475,
      "grad_norm": 1.57821524143219,
      "learning_rate": 0.00018167777777777777,
      "loss": 2.7731,
      "step": 33650
    },
    {
      "epoch": 4.78897257353986,
      "grad_norm": 1.5972754955291748,
      "learning_rate": 0.0001811222222222222,
      "loss": 2.74,
      "step": 33700
    },
    {
      "epoch": 4.796077874094074,
      "grad_norm": 1.537202000617981,
      "learning_rate": 0.00018056666666666665,
      "loss": 2.7553,
      "step": 33750
    },
    {
      "epoch": 4.803183174648288,
      "grad_norm": 1.4743520021438599,
      "learning_rate": 0.00018001111111111114,
      "loss": 2.7268,
      "step": 33800
    },
    {
      "epoch": 4.810288475202501,
      "grad_norm": 1.5363125801086426,
      "learning_rate": 0.00017945555555555557,
      "loss": 2.7475,
      "step": 33850
    },
    {
      "epoch": 4.817393775756715,
      "grad_norm": 1.6069902181625366,
      "learning_rate": 0.0001789,
      "loss": 2.7431,
      "step": 33900
    },
    {
      "epoch": 4.824499076310928,
      "grad_norm": 1.5969340801239014,
      "learning_rate": 0.00017834444444444445,
      "loss": 2.7594,
      "step": 33950
    },
    {
      "epoch": 4.831604376865141,
      "grad_norm": 1.5152361392974854,
      "learning_rate": 0.00017778888888888889,
      "loss": 2.7431,
      "step": 34000
    },
    {
      "epoch": 4.831604376865141,
      "eval_accuracy": 0.5208772420883179,
      "eval_loss": 2.706029176712036,
      "eval_runtime": 1.4028,
      "eval_samples_per_second": 2680.365,
      "eval_steps_per_second": 42.059,
      "step": 34000
    },
    {
      "epoch": 4.838709677419355,
      "grad_norm": 1.5891512632369995,
      "learning_rate": 0.00017723333333333332,
      "loss": 2.7503,
      "step": 34050
    },
    {
      "epoch": 4.845814977973569,
      "grad_norm": 1.6110639572143555,
      "learning_rate": 0.00017667777777777776,
      "loss": 2.7449,
      "step": 34100
    },
    {
      "epoch": 4.852920278527781,
      "grad_norm": 1.7298341989517212,
      "learning_rate": 0.00017612222222222225,
      "loss": 2.7515,
      "step": 34150
    },
    {
      "epoch": 4.860025579081995,
      "grad_norm": 1.5957584381103516,
      "learning_rate": 0.0001755666666666667,
      "loss": 2.7374,
      "step": 34200
    },
    {
      "epoch": 4.867130879636209,
      "grad_norm": 1.4720990657806396,
      "learning_rate": 0.00017501111111111112,
      "loss": 2.7342,
      "step": 34250
    },
    {
      "epoch": 4.874236180190422,
      "grad_norm": 1.6834919452667236,
      "learning_rate": 0.00017445555555555556,
      "loss": 2.7247,
      "step": 34300
    },
    {
      "epoch": 4.881341480744635,
      "grad_norm": 1.521272897720337,
      "learning_rate": 0.0001739,
      "loss": 2.739,
      "step": 34350
    },
    {
      "epoch": 4.888446781298849,
      "grad_norm": 1.671993374824524,
      "learning_rate": 0.00017334444444444444,
      "loss": 2.7624,
      "step": 34400
    },
    {
      "epoch": 4.8955520818530625,
      "grad_norm": 1.6694529056549072,
      "learning_rate": 0.00017278888888888887,
      "loss": 2.7306,
      "step": 34450
    },
    {
      "epoch": 4.902657382407276,
      "grad_norm": 1.4730091094970703,
      "learning_rate": 0.00017223333333333334,
      "loss": 2.7275,
      "step": 34500
    },
    {
      "epoch": 4.909762682961489,
      "grad_norm": 1.5129313468933105,
      "learning_rate": 0.0001716777777777778,
      "loss": 2.7323,
      "step": 34550
    },
    {
      "epoch": 4.9168679835157025,
      "grad_norm": 1.5696403980255127,
      "learning_rate": 0.00017112222222222224,
      "loss": 2.7295,
      "step": 34600
    },
    {
      "epoch": 4.923973284069916,
      "grad_norm": 1.7613575458526611,
      "learning_rate": 0.00017056666666666667,
      "loss": 2.7267,
      "step": 34650
    },
    {
      "epoch": 4.93107858462413,
      "grad_norm": 1.5198171138763428,
      "learning_rate": 0.0001700111111111111,
      "loss": 2.7287,
      "step": 34700
    },
    {
      "epoch": 4.9381838851783435,
      "grad_norm": 1.5724774599075317,
      "learning_rate": 0.00016945555555555555,
      "loss": 2.7424,
      "step": 34750
    },
    {
      "epoch": 4.945289185732556,
      "grad_norm": 1.5571963787078857,
      "learning_rate": 0.00016889999999999999,
      "loss": 2.7071,
      "step": 34800
    },
    {
      "epoch": 4.95239448628677,
      "grad_norm": 1.590010404586792,
      "learning_rate": 0.00016834444444444445,
      "loss": 2.7464,
      "step": 34850
    },
    {
      "epoch": 4.959499786840984,
      "grad_norm": 1.6421610116958618,
      "learning_rate": 0.0001677888888888889,
      "loss": 2.7104,
      "step": 34900
    },
    {
      "epoch": 4.966605087395196,
      "grad_norm": 1.5812050104141235,
      "learning_rate": 0.00016723333333333335,
      "loss": 2.748,
      "step": 34950
    },
    {
      "epoch": 4.97371038794941,
      "grad_norm": 1.5282282829284668,
      "learning_rate": 0.0001666777777777778,
      "loss": 2.763,
      "step": 35000
    },
    {
      "epoch": 4.980815688503624,
      "grad_norm": 1.509498953819275,
      "learning_rate": 0.00016612222222222222,
      "loss": 2.7346,
      "step": 35050
    },
    {
      "epoch": 4.987920989057837,
      "grad_norm": 1.6109907627105713,
      "learning_rate": 0.00016556666666666666,
      "loss": 2.7215,
      "step": 35100
    },
    {
      "epoch": 4.995026289612051,
      "grad_norm": 1.5312402248382568,
      "learning_rate": 0.0001650111111111111,
      "loss": 2.7163,
      "step": 35150
    },
    {
      "epoch": 5.002131590166264,
      "grad_norm": 1.620906949043274,
      "learning_rate": 0.00016445555555555556,
      "loss": 2.7254,
      "step": 35200
    },
    {
      "epoch": 5.009236890720477,
      "grad_norm": 1.5913270711898804,
      "learning_rate": 0.0001639,
      "loss": 2.6946,
      "step": 35250
    },
    {
      "epoch": 5.016342191274691,
      "grad_norm": 1.6128920316696167,
      "learning_rate": 0.00016334444444444444,
      "loss": 2.7067,
      "step": 35300
    },
    {
      "epoch": 5.023447491828905,
      "grad_norm": 1.598669171333313,
      "learning_rate": 0.0001627888888888889,
      "loss": 2.7214,
      "step": 35350
    },
    {
      "epoch": 5.0305527923831175,
      "grad_norm": 1.576566457748413,
      "learning_rate": 0.00016223333333333334,
      "loss": 2.7142,
      "step": 35400
    },
    {
      "epoch": 5.037658092937331,
      "grad_norm": 1.5625091791152954,
      "learning_rate": 0.00016167777777777777,
      "loss": 2.7218,
      "step": 35450
    },
    {
      "epoch": 5.044763393491545,
      "grad_norm": 1.595491886138916,
      "learning_rate": 0.00016112222222222224,
      "loss": 2.7142,
      "step": 35500
    },
    {
      "epoch": 5.0518686940457584,
      "grad_norm": 1.8340257406234741,
      "learning_rate": 0.00016056666666666668,
      "loss": 2.6812,
      "step": 35550
    },
    {
      "epoch": 5.058973994599971,
      "grad_norm": 1.5786815881729126,
      "learning_rate": 0.0001600111111111111,
      "loss": 2.6866,
      "step": 35600
    },
    {
      "epoch": 5.066079295154185,
      "grad_norm": 1.6554768085479736,
      "learning_rate": 0.00015945555555555555,
      "loss": 2.7131,
      "step": 35650
    },
    {
      "epoch": 5.0731845957083985,
      "grad_norm": 1.7100026607513428,
      "learning_rate": 0.0001589,
      "loss": 2.7062,
      "step": 35700
    },
    {
      "epoch": 5.080289896262612,
      "grad_norm": 1.4170799255371094,
      "learning_rate": 0.00015834444444444445,
      "loss": 2.703,
      "step": 35750
    },
    {
      "epoch": 5.087395196816825,
      "grad_norm": 1.573917031288147,
      "learning_rate": 0.0001577888888888889,
      "loss": 2.7189,
      "step": 35800
    },
    {
      "epoch": 5.094500497371039,
      "grad_norm": 1.5848358869552612,
      "learning_rate": 0.00015723333333333335,
      "loss": 2.7044,
      "step": 35850
    },
    {
      "epoch": 5.101605797925252,
      "grad_norm": 1.602611780166626,
      "learning_rate": 0.0001566777777777778,
      "loss": 2.7033,
      "step": 35900
    },
    {
      "epoch": 5.108711098479466,
      "grad_norm": 1.728371262550354,
      "learning_rate": 0.00015612222222222223,
      "loss": 2.709,
      "step": 35950
    },
    {
      "epoch": 5.11581639903368,
      "grad_norm": 1.5747239589691162,
      "learning_rate": 0.00015556666666666666,
      "loss": 2.6966,
      "step": 36000
    },
    {
      "epoch": 5.11581639903368,
      "eval_accuracy": 0.521781861782074,
      "eval_loss": 2.6849184036254883,
      "eval_runtime": 1.4614,
      "eval_samples_per_second": 2572.842,
      "eval_steps_per_second": 40.372,
      "step": 36000
    },
    {
      "epoch": 5.122921699587892,
      "grad_norm": 1.4712780714035034,
      "learning_rate": 0.0001550111111111111,
      "loss": 2.7073,
      "step": 36050
    },
    {
      "epoch": 5.130027000142106,
      "grad_norm": 1.6234378814697266,
      "learning_rate": 0.00015445555555555556,
      "loss": 2.7039,
      "step": 36100
    },
    {
      "epoch": 5.13713230069632,
      "grad_norm": 1.6177140474319458,
      "learning_rate": 0.0001539,
      "loss": 2.7057,
      "step": 36150
    },
    {
      "epoch": 5.144237601250533,
      "grad_norm": 1.6571050882339478,
      "learning_rate": 0.00015334444444444446,
      "loss": 2.6836,
      "step": 36200
    },
    {
      "epoch": 5.151342901804746,
      "grad_norm": 1.4596971273422241,
      "learning_rate": 0.0001527888888888889,
      "loss": 2.7236,
      "step": 36250
    },
    {
      "epoch": 5.15844820235896,
      "grad_norm": 1.5042576789855957,
      "learning_rate": 0.00015223333333333334,
      "loss": 2.6988,
      "step": 36300
    },
    {
      "epoch": 5.165553502913173,
      "grad_norm": 1.5042853355407715,
      "learning_rate": 0.00015167777777777778,
      "loss": 2.7274,
      "step": 36350
    },
    {
      "epoch": 5.172658803467387,
      "grad_norm": 1.6068828105926514,
      "learning_rate": 0.0001511222222222222,
      "loss": 2.6741,
      "step": 36400
    },
    {
      "epoch": 5.1797641040216,
      "grad_norm": 1.4389551877975464,
      "learning_rate": 0.00015056666666666665,
      "loss": 2.7118,
      "step": 36450
    },
    {
      "epoch": 5.1868694045758135,
      "grad_norm": 1.5431450605392456,
      "learning_rate": 0.0001500111111111111,
      "loss": 2.7079,
      "step": 36500
    },
    {
      "epoch": 5.193974705130027,
      "grad_norm": 1.6403303146362305,
      "learning_rate": 0.00014945555555555558,
      "loss": 2.7069,
      "step": 36550
    },
    {
      "epoch": 5.201080005684241,
      "grad_norm": 1.7466474771499634,
      "learning_rate": 0.00014890000000000001,
      "loss": 2.6675,
      "step": 36600
    },
    {
      "epoch": 5.2081853062384535,
      "grad_norm": 1.5739290714263916,
      "learning_rate": 0.00014834444444444445,
      "loss": 2.7048,
      "step": 36650
    },
    {
      "epoch": 5.215290606792667,
      "grad_norm": 1.6087383031845093,
      "learning_rate": 0.0001477888888888889,
      "loss": 2.6779,
      "step": 36700
    },
    {
      "epoch": 5.222395907346881,
      "grad_norm": 1.5931442975997925,
      "learning_rate": 0.00014723333333333333,
      "loss": 2.6958,
      "step": 36750
    },
    {
      "epoch": 5.2295012079010945,
      "grad_norm": 1.8047977685928345,
      "learning_rate": 0.00014667777777777776,
      "loss": 2.6802,
      "step": 36800
    },
    {
      "epoch": 5.236606508455307,
      "grad_norm": 1.7050204277038574,
      "learning_rate": 0.0001461222222222222,
      "loss": 2.7088,
      "step": 36850
    },
    {
      "epoch": 5.243711809009521,
      "grad_norm": 1.5651583671569824,
      "learning_rate": 0.0001455666666666667,
      "loss": 2.6978,
      "step": 36900
    },
    {
      "epoch": 5.250817109563735,
      "grad_norm": 1.519932508468628,
      "learning_rate": 0.00014501111111111113,
      "loss": 2.6886,
      "step": 36950
    },
    {
      "epoch": 5.257922410117948,
      "grad_norm": 1.6694849729537964,
      "learning_rate": 0.00014445555555555556,
      "loss": 2.7073,
      "step": 37000
    },
    {
      "epoch": 5.265027710672161,
      "grad_norm": 1.5857856273651123,
      "learning_rate": 0.0001439,
      "loss": 2.7079,
      "step": 37050
    },
    {
      "epoch": 5.272133011226375,
      "grad_norm": 1.78690505027771,
      "learning_rate": 0.00014334444444444444,
      "loss": 2.6991,
      "step": 37100
    },
    {
      "epoch": 5.279238311780588,
      "grad_norm": 1.7014470100402832,
      "learning_rate": 0.00014278888888888888,
      "loss": 2.6957,
      "step": 37150
    },
    {
      "epoch": 5.286343612334802,
      "grad_norm": 1.4961591958999634,
      "learning_rate": 0.00014223333333333334,
      "loss": 2.6875,
      "step": 37200
    },
    {
      "epoch": 5.293448912889016,
      "grad_norm": 1.682266116142273,
      "learning_rate": 0.0001416777777777778,
      "loss": 2.717,
      "step": 37250
    },
    {
      "epoch": 5.300554213443228,
      "grad_norm": 1.613564372062683,
      "learning_rate": 0.00014112222222222224,
      "loss": 2.6989,
      "step": 37300
    },
    {
      "epoch": 5.307659513997442,
      "grad_norm": 1.5683411359786987,
      "learning_rate": 0.00014056666666666668,
      "loss": 2.6864,
      "step": 37350
    },
    {
      "epoch": 5.314764814551656,
      "grad_norm": 1.5730217695236206,
      "learning_rate": 0.00014001111111111111,
      "loss": 2.6997,
      "step": 37400
    },
    {
      "epoch": 5.321870115105869,
      "grad_norm": 1.8172276020050049,
      "learning_rate": 0.00013945555555555555,
      "loss": 2.6929,
      "step": 37450
    },
    {
      "epoch": 5.328975415660082,
      "grad_norm": 1.5851980447769165,
      "learning_rate": 0.0001389,
      "loss": 2.6976,
      "step": 37500
    },
    {
      "epoch": 5.336080716214296,
      "grad_norm": 1.569605827331543,
      "learning_rate": 0.00013834444444444445,
      "loss": 2.7113,
      "step": 37550
    },
    {
      "epoch": 5.3431860167685095,
      "grad_norm": 1.4252275228500366,
      "learning_rate": 0.0001377888888888889,
      "loss": 2.6696,
      "step": 37600
    },
    {
      "epoch": 5.350291317322723,
      "grad_norm": 1.5046370029449463,
      "learning_rate": 0.00013723333333333335,
      "loss": 2.6994,
      "step": 37650
    },
    {
      "epoch": 5.357396617876936,
      "grad_norm": 1.4966368675231934,
      "learning_rate": 0.0001366777777777778,
      "loss": 2.6858,
      "step": 37700
    },
    {
      "epoch": 5.3645019184311495,
      "grad_norm": 1.635502576828003,
      "learning_rate": 0.00013612222222222223,
      "loss": 2.6776,
      "step": 37750
    },
    {
      "epoch": 5.371607218985363,
      "grad_norm": 1.703403115272522,
      "learning_rate": 0.00013556666666666666,
      "loss": 2.6876,
      "step": 37800
    },
    {
      "epoch": 5.378712519539577,
      "grad_norm": 1.569338321685791,
      "learning_rate": 0.0001350111111111111,
      "loss": 2.6779,
      "step": 37850
    },
    {
      "epoch": 5.38581782009379,
      "grad_norm": 1.4831880331039429,
      "learning_rate": 0.00013445555555555557,
      "loss": 2.6771,
      "step": 37900
    },
    {
      "epoch": 5.392923120648003,
      "grad_norm": 1.5764096975326538,
      "learning_rate": 0.0001339,
      "loss": 2.6808,
      "step": 37950
    },
    {
      "epoch": 5.400028421202217,
      "grad_norm": 1.4974082708358765,
      "learning_rate": 0.00013334444444444444,
      "loss": 2.6965,
      "step": 38000
    },
    {
      "epoch": 5.400028421202217,
      "eval_accuracy": 0.5244875550270081,
      "eval_loss": 2.6772782802581787,
      "eval_runtime": 1.2877,
      "eval_samples_per_second": 2919.822,
      "eval_steps_per_second": 45.816,
      "step": 38000
    },
    {
      "epoch": 5.407133721756431,
      "grad_norm": 1.7094296216964722,
      "learning_rate": 0.0001327888888888889,
      "loss": 2.6857,
      "step": 38050
    },
    {
      "epoch": 5.414239022310643,
      "grad_norm": 1.6493133306503296,
      "learning_rate": 0.00013223333333333334,
      "loss": 2.6767,
      "step": 38100
    },
    {
      "epoch": 5.421344322864857,
      "grad_norm": 1.6794978380203247,
      "learning_rate": 0.00013167777777777778,
      "loss": 2.6659,
      "step": 38150
    },
    {
      "epoch": 5.428449623419071,
      "grad_norm": 1.5671613216400146,
      "learning_rate": 0.00013112222222222221,
      "loss": 2.6793,
      "step": 38200
    },
    {
      "epoch": 5.435554923973284,
      "grad_norm": 1.6433796882629395,
      "learning_rate": 0.00013056666666666668,
      "loss": 2.6739,
      "step": 38250
    },
    {
      "epoch": 5.442660224527497,
      "grad_norm": 1.6001204252243042,
      "learning_rate": 0.00013001111111111112,
      "loss": 2.6673,
      "step": 38300
    },
    {
      "epoch": 5.449765525081711,
      "grad_norm": 1.6381980180740356,
      "learning_rate": 0.00012945555555555555,
      "loss": 2.6867,
      "step": 38350
    },
    {
      "epoch": 5.456870825635924,
      "grad_norm": 1.443001627922058,
      "learning_rate": 0.0001289,
      "loss": 2.6864,
      "step": 38400
    },
    {
      "epoch": 5.463976126190138,
      "grad_norm": 1.6225110292434692,
      "learning_rate": 0.00012834444444444445,
      "loss": 2.6858,
      "step": 38450
    },
    {
      "epoch": 5.471081426744352,
      "grad_norm": 1.5502654314041138,
      "learning_rate": 0.0001277888888888889,
      "loss": 2.6823,
      "step": 38500
    },
    {
      "epoch": 5.4781867272985645,
      "grad_norm": 1.579957127571106,
      "learning_rate": 0.00012723333333333333,
      "loss": 2.6662,
      "step": 38550
    },
    {
      "epoch": 5.485292027852778,
      "grad_norm": 1.6774084568023682,
      "learning_rate": 0.0001266777777777778,
      "loss": 2.677,
      "step": 38600
    },
    {
      "epoch": 5.492397328406992,
      "grad_norm": 1.554765224456787,
      "learning_rate": 0.00012612222222222223,
      "loss": 2.6806,
      "step": 38650
    },
    {
      "epoch": 5.4995026289612055,
      "grad_norm": 1.5972063541412354,
      "learning_rate": 0.00012556666666666666,
      "loss": 2.6703,
      "step": 38700
    },
    {
      "epoch": 5.506607929515418,
      "grad_norm": 1.596091866493225,
      "learning_rate": 0.0001250111111111111,
      "loss": 2.6884,
      "step": 38750
    },
    {
      "epoch": 5.513713230069632,
      "grad_norm": 1.5787779092788696,
      "learning_rate": 0.00012445555555555557,
      "loss": 2.6518,
      "step": 38800
    },
    {
      "epoch": 5.5208185306238455,
      "grad_norm": 1.4552741050720215,
      "learning_rate": 0.0001239,
      "loss": 2.674,
      "step": 38850
    },
    {
      "epoch": 5.527923831178059,
      "grad_norm": 1.5038365125656128,
      "learning_rate": 0.00012334444444444447,
      "loss": 2.6618,
      "step": 38900
    },
    {
      "epoch": 5.535029131732272,
      "grad_norm": 1.5640016794204712,
      "learning_rate": 0.0001227888888888889,
      "loss": 2.6654,
      "step": 38950
    },
    {
      "epoch": 5.542134432286486,
      "grad_norm": 1.6808786392211914,
      "learning_rate": 0.00012223333333333334,
      "loss": 2.654,
      "step": 39000
    },
    {
      "epoch": 5.549239732840699,
      "grad_norm": 1.6088745594024658,
      "learning_rate": 0.00012167777777777778,
      "loss": 2.6694,
      "step": 39050
    },
    {
      "epoch": 5.556345033394913,
      "grad_norm": 1.5467897653579712,
      "learning_rate": 0.00012112222222222223,
      "loss": 2.6646,
      "step": 39100
    },
    {
      "epoch": 5.563450333949126,
      "grad_norm": 1.5007407665252686,
      "learning_rate": 0.00012056666666666667,
      "loss": 2.681,
      "step": 39150
    },
    {
      "epoch": 5.570555634503339,
      "grad_norm": 1.5999568700790405,
      "learning_rate": 0.00012001111111111112,
      "loss": 2.6965,
      "step": 39200
    },
    {
      "epoch": 5.577660935057553,
      "grad_norm": 1.7177321910858154,
      "learning_rate": 0.00011945555555555555,
      "loss": 2.6713,
      "step": 39250
    },
    {
      "epoch": 5.584766235611767,
      "grad_norm": 1.7111984491348267,
      "learning_rate": 0.0001189,
      "loss": 2.6562,
      "step": 39300
    },
    {
      "epoch": 5.59187153616598,
      "grad_norm": 1.4950600862503052,
      "learning_rate": 0.00011834444444444445,
      "loss": 2.646,
      "step": 39350
    },
    {
      "epoch": 5.598976836720193,
      "grad_norm": 1.7324713468551636,
      "learning_rate": 0.00011778888888888889,
      "loss": 2.6701,
      "step": 39400
    },
    {
      "epoch": 5.606082137274407,
      "grad_norm": 1.685196042060852,
      "learning_rate": 0.00011723333333333333,
      "loss": 2.6571,
      "step": 39450
    },
    {
      "epoch": 5.61318743782862,
      "grad_norm": 1.4760277271270752,
      "learning_rate": 0.00011667777777777779,
      "loss": 2.6617,
      "step": 39500
    },
    {
      "epoch": 5.620292738382833,
      "grad_norm": 1.6101094484329224,
      "learning_rate": 0.00011612222222222223,
      "loss": 2.6783,
      "step": 39550
    },
    {
      "epoch": 5.627398038937047,
      "grad_norm": 1.649571418762207,
      "learning_rate": 0.00011556666666666667,
      "loss": 2.6717,
      "step": 39600
    },
    {
      "epoch": 5.6345033394912605,
      "grad_norm": 1.6487045288085938,
      "learning_rate": 0.0001150111111111111,
      "loss": 2.6572,
      "step": 39650
    },
    {
      "epoch": 5.641608640045474,
      "grad_norm": 1.711259365081787,
      "learning_rate": 0.00011445555555555557,
      "loss": 2.6626,
      "step": 39700
    },
    {
      "epoch": 5.648713940599688,
      "grad_norm": 1.5313870906829834,
      "learning_rate": 0.0001139,
      "loss": 2.677,
      "step": 39750
    },
    {
      "epoch": 5.655819241153901,
      "grad_norm": 1.729377031326294,
      "learning_rate": 0.00011334444444444444,
      "loss": 2.674,
      "step": 39800
    },
    {
      "epoch": 5.662924541708114,
      "grad_norm": 1.6128586530685425,
      "learning_rate": 0.00011278888888888889,
      "loss": 2.6651,
      "step": 39850
    },
    {
      "epoch": 5.670029842262328,
      "grad_norm": 1.6688061952590942,
      "learning_rate": 0.00011223333333333334,
      "loss": 2.6423,
      "step": 39900
    },
    {
      "epoch": 5.6771351428165415,
      "grad_norm": 1.5778543949127197,
      "learning_rate": 0.00011167777777777778,
      "loss": 2.6659,
      "step": 39950
    },
    {
      "epoch": 5.684240443370754,
      "grad_norm": 1.7078243494033813,
      "learning_rate": 0.00011112222222222222,
      "loss": 2.6466,
      "step": 40000
    },
    {
      "epoch": 5.684240443370754,
      "eval_accuracy": 0.5260953307151794,
      "eval_loss": 2.670067310333252,
      "eval_runtime": 1.4203,
      "eval_samples_per_second": 2647.336,
      "eval_steps_per_second": 41.541,
      "step": 40000
    },
    {
      "epoch": 5.691345743924968,
      "grad_norm": 1.573383092880249,
      "learning_rate": 0.00011056666666666667,
      "loss": 2.6674,
      "step": 40050
    },
    {
      "epoch": 5.698451044479182,
      "grad_norm": 1.6435350179672241,
      "learning_rate": 0.00011001111111111112,
      "loss": 2.6686,
      "step": 40100
    },
    {
      "epoch": 5.705556345033395,
      "grad_norm": 1.5594240427017212,
      "learning_rate": 0.00010945555555555555,
      "loss": 2.6544,
      "step": 40150
    },
    {
      "epoch": 5.712661645587608,
      "grad_norm": 1.7234139442443848,
      "learning_rate": 0.0001089,
      "loss": 2.6475,
      "step": 40200
    },
    {
      "epoch": 5.719766946141822,
      "grad_norm": 1.4562207460403442,
      "learning_rate": 0.00010834444444444444,
      "loss": 2.6673,
      "step": 40250
    },
    {
      "epoch": 5.726872246696035,
      "grad_norm": 1.5317538976669312,
      "learning_rate": 0.00010778888888888889,
      "loss": 2.6356,
      "step": 40300
    },
    {
      "epoch": 5.733977547250249,
      "grad_norm": 1.5807552337646484,
      "learning_rate": 0.00010723333333333334,
      "loss": 2.6816,
      "step": 40350
    },
    {
      "epoch": 5.741082847804462,
      "grad_norm": 1.5644679069519043,
      "learning_rate": 0.00010667777777777778,
      "loss": 2.663,
      "step": 40400
    },
    {
      "epoch": 5.748188148358675,
      "grad_norm": 1.5229665040969849,
      "learning_rate": 0.00010612222222222223,
      "loss": 2.6588,
      "step": 40450
    },
    {
      "epoch": 5.755293448912889,
      "grad_norm": 1.532139778137207,
      "learning_rate": 0.00010556666666666667,
      "loss": 2.6606,
      "step": 40500
    },
    {
      "epoch": 5.762398749467103,
      "grad_norm": 1.6157749891281128,
      "learning_rate": 0.00010501111111111112,
      "loss": 2.6476,
      "step": 40550
    },
    {
      "epoch": 5.769504050021316,
      "grad_norm": 1.5783989429473877,
      "learning_rate": 0.00010445555555555555,
      "loss": 2.6586,
      "step": 40600
    },
    {
      "epoch": 5.776609350575529,
      "grad_norm": 1.6066045761108398,
      "learning_rate": 0.0001039,
      "loss": 2.6452,
      "step": 40650
    },
    {
      "epoch": 5.783714651129743,
      "grad_norm": 1.563295841217041,
      "learning_rate": 0.00010334444444444446,
      "loss": 2.6814,
      "step": 40700
    },
    {
      "epoch": 5.7908199516839565,
      "grad_norm": 1.7197033166885376,
      "learning_rate": 0.00010278888888888889,
      "loss": 2.6531,
      "step": 40750
    },
    {
      "epoch": 5.797925252238169,
      "grad_norm": 1.6908972263336182,
      "learning_rate": 0.00010223333333333333,
      "loss": 2.6506,
      "step": 40800
    },
    {
      "epoch": 5.805030552792383,
      "grad_norm": 1.5352091789245605,
      "learning_rate": 0.00010167777777777778,
      "loss": 2.6505,
      "step": 40850
    },
    {
      "epoch": 5.8121358533465965,
      "grad_norm": 1.5693176984786987,
      "learning_rate": 0.00010112222222222223,
      "loss": 2.6653,
      "step": 40900
    },
    {
      "epoch": 5.81924115390081,
      "grad_norm": 1.5169340372085571,
      "learning_rate": 0.00010056666666666667,
      "loss": 2.6262,
      "step": 40950
    },
    {
      "epoch": 5.826346454455024,
      "grad_norm": 1.6942225694656372,
      "learning_rate": 0.0001000111111111111,
      "loss": 2.6643,
      "step": 41000
    },
    {
      "epoch": 5.833451755009237,
      "grad_norm": 1.6022897958755493,
      "learning_rate": 9.945555555555557e-05,
      "loss": 2.6171,
      "step": 41050
    },
    {
      "epoch": 5.84055705556345,
      "grad_norm": 1.5821208953857422,
      "learning_rate": 9.89e-05,
      "loss": 2.6481,
      "step": 41100
    },
    {
      "epoch": 5.847662356117664,
      "grad_norm": 1.5729005336761475,
      "learning_rate": 9.834444444444444e-05,
      "loss": 2.6347,
      "step": 41150
    },
    {
      "epoch": 5.854767656671878,
      "grad_norm": 1.587731957435608,
      "learning_rate": 9.778888888888888e-05,
      "loss": 2.6503,
      "step": 41200
    },
    {
      "epoch": 5.86187295722609,
      "grad_norm": 1.5719395875930786,
      "learning_rate": 9.723333333333334e-05,
      "loss": 2.662,
      "step": 41250
    },
    {
      "epoch": 5.868978257780304,
      "grad_norm": 1.6065324544906616,
      "learning_rate": 9.667777777777778e-05,
      "loss": 2.6593,
      "step": 41300
    },
    {
      "epoch": 5.876083558334518,
      "grad_norm": 1.6367747783660889,
      "learning_rate": 9.612222222222222e-05,
      "loss": 2.647,
      "step": 41350
    },
    {
      "epoch": 5.883188858888731,
      "grad_norm": 1.6349313259124756,
      "learning_rate": 9.556666666666667e-05,
      "loss": 2.6427,
      "step": 41400
    },
    {
      "epoch": 5.890294159442944,
      "grad_norm": 1.597489356994629,
      "learning_rate": 9.501111111111112e-05,
      "loss": 2.6612,
      "step": 41450
    },
    {
      "epoch": 5.897399459997158,
      "grad_norm": 1.696363091468811,
      "learning_rate": 9.445555555555556e-05,
      "loss": 2.6384,
      "step": 41500
    },
    {
      "epoch": 5.904504760551371,
      "grad_norm": 1.6225945949554443,
      "learning_rate": 9.39e-05,
      "loss": 2.6575,
      "step": 41550
    },
    {
      "epoch": 5.911610061105585,
      "grad_norm": 1.6240595579147339,
      "learning_rate": 9.334444444444444e-05,
      "loss": 2.6414,
      "step": 41600
    },
    {
      "epoch": 5.918715361659798,
      "grad_norm": 1.6538727283477783,
      "learning_rate": 9.278888888888889e-05,
      "loss": 2.6411,
      "step": 41650
    },
    {
      "epoch": 5.9258206622140115,
      "grad_norm": 1.543818473815918,
      "learning_rate": 9.223333333333333e-05,
      "loss": 2.637,
      "step": 41700
    },
    {
      "epoch": 5.932925962768225,
      "grad_norm": 1.6847704648971558,
      "learning_rate": 9.167777777777778e-05,
      "loss": 2.6321,
      "step": 41750
    },
    {
      "epoch": 5.940031263322439,
      "grad_norm": 1.536786675453186,
      "learning_rate": 9.112222222222222e-05,
      "loss": 2.6215,
      "step": 41800
    },
    {
      "epoch": 5.9471365638766525,
      "grad_norm": 1.6496859788894653,
      "learning_rate": 9.056666666666667e-05,
      "loss": 2.6411,
      "step": 41850
    },
    {
      "epoch": 5.954241864430865,
      "grad_norm": 1.5533984899520874,
      "learning_rate": 9.001111111111112e-05,
      "loss": 2.6462,
      "step": 41900
    },
    {
      "epoch": 5.961347164985079,
      "grad_norm": 1.6307629346847534,
      "learning_rate": 8.945555555555556e-05,
      "loss": 2.6413,
      "step": 41950
    },
    {
      "epoch": 5.9684524655392925,
      "grad_norm": 1.693447470664978,
      "learning_rate": 8.89e-05,
      "loss": 2.6365,
      "step": 42000
    },
    {
      "epoch": 5.9684524655392925,
      "eval_accuracy": 0.5310431718826294,
      "eval_loss": 2.6503593921661377,
      "eval_runtime": 1.5353,
      "eval_samples_per_second": 2448.955,
      "eval_steps_per_second": 38.428,
      "step": 42000
    },
    {
      "epoch": 5.975557766093505,
      "grad_norm": 1.7850193977355957,
      "learning_rate": 8.834444444444444e-05,
      "loss": 2.6257,
      "step": 42050
    },
    {
      "epoch": 5.982663066647719,
      "grad_norm": 1.711273431777954,
      "learning_rate": 8.77888888888889e-05,
      "loss": 2.6191,
      "step": 42100
    },
    {
      "epoch": 5.989768367201933,
      "grad_norm": 1.7241146564483643,
      "learning_rate": 8.723333333333333e-05,
      "loss": 2.6228,
      "step": 42150
    },
    {
      "epoch": 5.996873667756146,
      "grad_norm": 1.6184428930282593,
      "learning_rate": 8.667777777777778e-05,
      "loss": 2.62,
      "step": 42200
    },
    {
      "epoch": 6.00397896831036,
      "grad_norm": 1.6427233219146729,
      "learning_rate": 8.612222222222223e-05,
      "loss": 2.6255,
      "step": 42250
    },
    {
      "epoch": 6.011084268864573,
      "grad_norm": 1.6040061712265015,
      "learning_rate": 8.556666666666667e-05,
      "loss": 2.6212,
      "step": 42300
    },
    {
      "epoch": 6.018189569418786,
      "grad_norm": 1.539577603340149,
      "learning_rate": 8.50111111111111e-05,
      "loss": 2.6142,
      "step": 42350
    },
    {
      "epoch": 6.025294869973,
      "grad_norm": 1.5189242362976074,
      "learning_rate": 8.445555555555557e-05,
      "loss": 2.6396,
      "step": 42400
    },
    {
      "epoch": 6.032400170527214,
      "grad_norm": 1.6133103370666504,
      "learning_rate": 8.39e-05,
      "loss": 2.6211,
      "step": 42450
    },
    {
      "epoch": 6.039505471081426,
      "grad_norm": 1.625777006149292,
      "learning_rate": 8.334444444444444e-05,
      "loss": 2.6138,
      "step": 42500
    },
    {
      "epoch": 6.04661077163564,
      "grad_norm": 1.6675585508346558,
      "learning_rate": 8.278888888888888e-05,
      "loss": 2.6357,
      "step": 42550
    },
    {
      "epoch": 6.053716072189854,
      "grad_norm": 1.5562056303024292,
      "learning_rate": 8.223333333333334e-05,
      "loss": 2.6409,
      "step": 42600
    },
    {
      "epoch": 6.060821372744067,
      "grad_norm": 1.4880784749984741,
      "learning_rate": 8.167777777777778e-05,
      "loss": 2.6206,
      "step": 42650
    },
    {
      "epoch": 6.06792667329828,
      "grad_norm": 1.6204643249511719,
      "learning_rate": 8.112222222222222e-05,
      "loss": 2.6226,
      "step": 42700
    },
    {
      "epoch": 6.075031973852494,
      "grad_norm": 1.6648606061935425,
      "learning_rate": 8.056666666666667e-05,
      "loss": 2.6139,
      "step": 42750
    },
    {
      "epoch": 6.0821372744067075,
      "grad_norm": 1.6851210594177246,
      "learning_rate": 8.001111111111112e-05,
      "loss": 2.6204,
      "step": 42800
    },
    {
      "epoch": 6.089242574960921,
      "grad_norm": 1.7008609771728516,
      "learning_rate": 7.945555555555556e-05,
      "loss": 2.6229,
      "step": 42850
    },
    {
      "epoch": 6.096347875515134,
      "grad_norm": 1.6615877151489258,
      "learning_rate": 7.89e-05,
      "loss": 2.6106,
      "step": 42900
    },
    {
      "epoch": 6.103453176069348,
      "grad_norm": 1.7349096536636353,
      "learning_rate": 7.834444444444444e-05,
      "loss": 2.6035,
      "step": 42950
    },
    {
      "epoch": 6.110558476623561,
      "grad_norm": 1.6119391918182373,
      "learning_rate": 7.77888888888889e-05,
      "loss": 2.6262,
      "step": 43000
    },
    {
      "epoch": 6.117663777177775,
      "grad_norm": 1.651769995689392,
      "learning_rate": 7.723333333333333e-05,
      "loss": 2.6131,
      "step": 43050
    },
    {
      "epoch": 6.124769077731988,
      "grad_norm": 1.5095596313476562,
      "learning_rate": 7.667777777777778e-05,
      "loss": 2.6036,
      "step": 43100
    },
    {
      "epoch": 6.131874378286201,
      "grad_norm": 1.5748904943466187,
      "learning_rate": 7.612222222222222e-05,
      "loss": 2.6016,
      "step": 43150
    },
    {
      "epoch": 6.138979678840415,
      "grad_norm": 1.5882796049118042,
      "learning_rate": 7.556666666666667e-05,
      "loss": 2.611,
      "step": 43200
    },
    {
      "epoch": 6.146084979394629,
      "grad_norm": 1.6430201530456543,
      "learning_rate": 7.501111111111112e-05,
      "loss": 2.6039,
      "step": 43250
    },
    {
      "epoch": 6.153190279948841,
      "grad_norm": 1.5699527263641357,
      "learning_rate": 7.445555555555556e-05,
      "loss": 2.6238,
      "step": 43300
    },
    {
      "epoch": 6.160295580503055,
      "grad_norm": 1.5793179273605347,
      "learning_rate": 7.39e-05,
      "loss": 2.6269,
      "step": 43350
    },
    {
      "epoch": 6.167400881057269,
      "grad_norm": 1.6893434524536133,
      "learning_rate": 7.334444444444444e-05,
      "loss": 2.6256,
      "step": 43400
    },
    {
      "epoch": 6.174506181611482,
      "grad_norm": 1.6231344938278198,
      "learning_rate": 7.27888888888889e-05,
      "loss": 2.6153,
      "step": 43450
    },
    {
      "epoch": 6.181611482165696,
      "grad_norm": 1.6697924137115479,
      "learning_rate": 7.223333333333333e-05,
      "loss": 2.5972,
      "step": 43500
    },
    {
      "epoch": 6.188716782719909,
      "grad_norm": 1.5698109865188599,
      "learning_rate": 7.167777777777778e-05,
      "loss": 2.5996,
      "step": 43550
    },
    {
      "epoch": 6.195822083274122,
      "grad_norm": 1.6216611862182617,
      "learning_rate": 7.112222222222223e-05,
      "loss": 2.5944,
      "step": 43600
    },
    {
      "epoch": 6.202927383828336,
      "grad_norm": 1.549578309059143,
      "learning_rate": 7.056666666666667e-05,
      "loss": 2.6129,
      "step": 43650
    },
    {
      "epoch": 6.21003268438255,
      "grad_norm": 1.5020294189453125,
      "learning_rate": 7.00111111111111e-05,
      "loss": 2.6016,
      "step": 43700
    },
    {
      "epoch": 6.2171379849367625,
      "grad_norm": 1.6891061067581177,
      "learning_rate": 6.945555555555556e-05,
      "loss": 2.5895,
      "step": 43750
    },
    {
      "epoch": 6.224243285490976,
      "grad_norm": 1.6669466495513916,
      "learning_rate": 6.890000000000001e-05,
      "loss": 2.6106,
      "step": 43800
    },
    {
      "epoch": 6.23134858604519,
      "grad_norm": 1.6679954528808594,
      "learning_rate": 6.834444444444444e-05,
      "loss": 2.6225,
      "step": 43850
    },
    {
      "epoch": 6.2384538865994035,
      "grad_norm": 1.5900484323501587,
      "learning_rate": 6.778888888888888e-05,
      "loss": 2.6264,
      "step": 43900
    },
    {
      "epoch": 6.245559187153616,
      "grad_norm": 1.8170428276062012,
      "learning_rate": 6.723333333333335e-05,
      "loss": 2.6022,
      "step": 43950
    },
    {
      "epoch": 6.25266448770783,
      "grad_norm": 1.6536431312561035,
      "learning_rate": 6.667777777777778e-05,
      "loss": 2.6184,
      "step": 44000
    },
    {
      "epoch": 6.25266448770783,
      "eval_accuracy": 0.5310567021369934,
      "eval_loss": 2.6401333808898926,
      "eval_runtime": 1.3799,
      "eval_samples_per_second": 2724.796,
      "eval_steps_per_second": 42.756,
      "step": 44000
    },
    {
      "epoch": 6.2597697882620436,
      "grad_norm": 1.6533645391464233,
      "learning_rate": 6.612222222222222e-05,
      "loss": 2.5819,
      "step": 44050
    },
    {
      "epoch": 6.266875088816257,
      "grad_norm": 1.6523014307022095,
      "learning_rate": 6.556666666666666e-05,
      "loss": 2.6304,
      "step": 44100
    },
    {
      "epoch": 6.27398038937047,
      "grad_norm": 1.6646928787231445,
      "learning_rate": 6.501111111111112e-05,
      "loss": 2.6044,
      "step": 44150
    },
    {
      "epoch": 6.281085689924684,
      "grad_norm": 1.5884506702423096,
      "learning_rate": 6.445555555555556e-05,
      "loss": 2.5986,
      "step": 44200
    },
    {
      "epoch": 6.288190990478897,
      "grad_norm": 1.666008472442627,
      "learning_rate": 6.39e-05,
      "loss": 2.6306,
      "step": 44250
    },
    {
      "epoch": 6.295296291033111,
      "grad_norm": 1.7450133562088013,
      "learning_rate": 6.334444444444445e-05,
      "loss": 2.6172,
      "step": 44300
    },
    {
      "epoch": 6.302401591587325,
      "grad_norm": 1.6662380695343018,
      "learning_rate": 6.27888888888889e-05,
      "loss": 2.5978,
      "step": 44350
    },
    {
      "epoch": 6.309506892141537,
      "grad_norm": 1.7020577192306519,
      "learning_rate": 6.223333333333333e-05,
      "loss": 2.6189,
      "step": 44400
    },
    {
      "epoch": 6.316612192695751,
      "grad_norm": 1.6353223323822021,
      "learning_rate": 6.167777777777778e-05,
      "loss": 2.6044,
      "step": 44450
    },
    {
      "epoch": 6.323717493249965,
      "grad_norm": 1.668189525604248,
      "learning_rate": 6.112222222222222e-05,
      "loss": 2.5963,
      "step": 44500
    },
    {
      "epoch": 6.3308227938041775,
      "grad_norm": 1.6153579950332642,
      "learning_rate": 6.0566666666666664e-05,
      "loss": 2.5861,
      "step": 44550
    },
    {
      "epoch": 6.337928094358391,
      "grad_norm": 1.5848743915557861,
      "learning_rate": 6.0011111111111114e-05,
      "loss": 2.5926,
      "step": 44600
    },
    {
      "epoch": 6.345033394912605,
      "grad_norm": 1.603839635848999,
      "learning_rate": 5.945555555555555e-05,
      "loss": 2.593,
      "step": 44650
    },
    {
      "epoch": 6.352138695466818,
      "grad_norm": 1.665961503982544,
      "learning_rate": 5.89e-05,
      "loss": 2.5935,
      "step": 44700
    },
    {
      "epoch": 6.359243996021032,
      "grad_norm": 1.5604987144470215,
      "learning_rate": 5.8344444444444446e-05,
      "loss": 2.6095,
      "step": 44750
    },
    {
      "epoch": 6.366349296575245,
      "grad_norm": 1.6461862325668335,
      "learning_rate": 5.778888888888889e-05,
      "loss": 2.6373,
      "step": 44800
    },
    {
      "epoch": 6.3734545971294585,
      "grad_norm": 1.5700501203536987,
      "learning_rate": 5.723333333333333e-05,
      "loss": 2.6208,
      "step": 44850
    },
    {
      "epoch": 6.380559897683672,
      "grad_norm": 1.5315876007080078,
      "learning_rate": 5.667777777777778e-05,
      "loss": 2.609,
      "step": 44900
    },
    {
      "epoch": 6.387665198237886,
      "grad_norm": 1.4918439388275146,
      "learning_rate": 5.612222222222222e-05,
      "loss": 2.6031,
      "step": 44950
    },
    {
      "epoch": 6.394770498792099,
      "grad_norm": 1.6920193433761597,
      "learning_rate": 5.556666666666667e-05,
      "loss": 2.6149,
      "step": 45000
    },
    {
      "epoch": 6.401875799346312,
      "grad_norm": 1.6514570713043213,
      "learning_rate": 5.501111111111111e-05,
      "loss": 2.6164,
      "step": 45050
    },
    {
      "epoch": 6.408981099900526,
      "grad_norm": 1.57295823097229,
      "learning_rate": 5.445555555555556e-05,
      "loss": 2.599,
      "step": 45100
    },
    {
      "epoch": 6.4160864004547395,
      "grad_norm": 1.6922849416732788,
      "learning_rate": 5.39e-05,
      "loss": 2.614,
      "step": 45150
    },
    {
      "epoch": 6.423191701008952,
      "grad_norm": 1.7391736507415771,
      "learning_rate": 5.3344444444444446e-05,
      "loss": 2.5909,
      "step": 45200
    },
    {
      "epoch": 6.430297001563166,
      "grad_norm": 1.584605097770691,
      "learning_rate": 5.2788888888888897e-05,
      "loss": 2.5942,
      "step": 45250
    },
    {
      "epoch": 6.43740230211738,
      "grad_norm": 1.7213895320892334,
      "learning_rate": 5.2233333333333334e-05,
      "loss": 2.5806,
      "step": 45300
    },
    {
      "epoch": 6.444507602671593,
      "grad_norm": 1.710915207862854,
      "learning_rate": 5.1677777777777784e-05,
      "loss": 2.5869,
      "step": 45350
    },
    {
      "epoch": 6.451612903225806,
      "grad_norm": 1.5835086107254028,
      "learning_rate": 5.112222222222222e-05,
      "loss": 2.6056,
      "step": 45400
    },
    {
      "epoch": 6.45871820378002,
      "grad_norm": 1.8420978784561157,
      "learning_rate": 5.056666666666667e-05,
      "loss": 2.6039,
      "step": 45450
    },
    {
      "epoch": 6.465823504334233,
      "grad_norm": 1.7443053722381592,
      "learning_rate": 5.001111111111111e-05,
      "loss": 2.5945,
      "step": 45500
    },
    {
      "epoch": 6.472928804888447,
      "grad_norm": 1.6628175973892212,
      "learning_rate": 4.945555555555556e-05,
      "loss": 2.5674,
      "step": 45550
    },
    {
      "epoch": 6.480034105442661,
      "grad_norm": 1.5999755859375,
      "learning_rate": 4.89e-05,
      "loss": 2.6151,
      "step": 45600
    },
    {
      "epoch": 6.487139405996873,
      "grad_norm": 1.5596177577972412,
      "learning_rate": 4.8344444444444447e-05,
      "loss": 2.5888,
      "step": 45650
    },
    {
      "epoch": 6.494244706551087,
      "grad_norm": 1.651270866394043,
      "learning_rate": 4.778888888888889e-05,
      "loss": 2.6043,
      "step": 45700
    },
    {
      "epoch": 6.501350007105301,
      "grad_norm": 1.840889573097229,
      "learning_rate": 4.7233333333333334e-05,
      "loss": 2.5918,
      "step": 45750
    },
    {
      "epoch": 6.5084553076595135,
      "grad_norm": 1.56648588180542,
      "learning_rate": 4.667777777777778e-05,
      "loss": 2.5825,
      "step": 45800
    },
    {
      "epoch": 6.515560608213727,
      "grad_norm": 1.5391209125518799,
      "learning_rate": 4.612222222222223e-05,
      "loss": 2.5998,
      "step": 45850
    },
    {
      "epoch": 6.522665908767941,
      "grad_norm": 1.6329506635665894,
      "learning_rate": 4.5566666666666665e-05,
      "loss": 2.5932,
      "step": 45900
    },
    {
      "epoch": 6.5297712093221545,
      "grad_norm": 1.5591578483581543,
      "learning_rate": 4.5011111111111116e-05,
      "loss": 2.6017,
      "step": 45950
    },
    {
      "epoch": 6.536876509876368,
      "grad_norm": 1.5110472440719604,
      "learning_rate": 4.445555555555555e-05,
      "loss": 2.5993,
      "step": 46000
    },
    {
      "epoch": 6.536876509876368,
      "eval_accuracy": 0.5387080907821655,
      "eval_loss": 2.6144747734069824,
      "eval_runtime": 1.4386,
      "eval_samples_per_second": 2613.657,
      "eval_steps_per_second": 41.012,
      "step": 46000
    },
    {
      "epoch": 6.543981810430581,
      "grad_norm": 1.7392023801803589,
      "learning_rate": 4.39e-05,
      "loss": 2.5975,
      "step": 46050
    },
    {
      "epoch": 6.551087110984795,
      "grad_norm": 1.5147992372512817,
      "learning_rate": 4.334444444444444e-05,
      "loss": 2.6104,
      "step": 46100
    },
    {
      "epoch": 6.558192411539008,
      "grad_norm": 1.6592369079589844,
      "learning_rate": 4.278888888888889e-05,
      "loss": 2.5908,
      "step": 46150
    },
    {
      "epoch": 6.565297712093222,
      "grad_norm": 1.6960110664367676,
      "learning_rate": 4.2233333333333334e-05,
      "loss": 2.5899,
      "step": 46200
    },
    {
      "epoch": 6.572403012647435,
      "grad_norm": 1.6082763671875,
      "learning_rate": 4.167777777777778e-05,
      "loss": 2.591,
      "step": 46250
    },
    {
      "epoch": 6.579508313201648,
      "grad_norm": 1.7125688791275024,
      "learning_rate": 4.112222222222222e-05,
      "loss": 2.6009,
      "step": 46300
    },
    {
      "epoch": 6.586613613755862,
      "grad_norm": 1.5538341999053955,
      "learning_rate": 4.0566666666666666e-05,
      "loss": 2.5956,
      "step": 46350
    },
    {
      "epoch": 6.593718914310076,
      "grad_norm": 1.5610955953598022,
      "learning_rate": 4.001111111111111e-05,
      "loss": 2.5926,
      "step": 46400
    },
    {
      "epoch": 6.600824214864288,
      "grad_norm": 1.7150379419326782,
      "learning_rate": 3.945555555555556e-05,
      "loss": 2.5879,
      "step": 46450
    },
    {
      "epoch": 6.607929515418502,
      "grad_norm": 1.703142523765564,
      "learning_rate": 3.89e-05,
      "loss": 2.5992,
      "step": 46500
    },
    {
      "epoch": 6.615034815972716,
      "grad_norm": 1.6176077127456665,
      "learning_rate": 3.834444444444445e-05,
      "loss": 2.5836,
      "step": 46550
    },
    {
      "epoch": 6.622140116526929,
      "grad_norm": 1.619447112083435,
      "learning_rate": 3.7788888888888884e-05,
      "loss": 2.5961,
      "step": 46600
    },
    {
      "epoch": 6.629245417081142,
      "grad_norm": 1.7121543884277344,
      "learning_rate": 3.7233333333333335e-05,
      "loss": 2.5927,
      "step": 46650
    },
    {
      "epoch": 6.636350717635356,
      "grad_norm": 1.5830241441726685,
      "learning_rate": 3.667777777777777e-05,
      "loss": 2.5861,
      "step": 46700
    },
    {
      "epoch": 6.643456018189569,
      "grad_norm": 1.7162401676177979,
      "learning_rate": 3.612222222222222e-05,
      "loss": 2.5839,
      "step": 46750
    },
    {
      "epoch": 6.650561318743783,
      "grad_norm": 1.6397333145141602,
      "learning_rate": 3.556666666666667e-05,
      "loss": 2.5975,
      "step": 46800
    },
    {
      "epoch": 6.657666619297997,
      "grad_norm": 1.5965735912322998,
      "learning_rate": 3.501111111111111e-05,
      "loss": 2.5734,
      "step": 46850
    },
    {
      "epoch": 6.6647719198522095,
      "grad_norm": 1.6135231256484985,
      "learning_rate": 3.445555555555556e-05,
      "loss": 2.5634,
      "step": 46900
    },
    {
      "epoch": 6.671877220406423,
      "grad_norm": 1.681933045387268,
      "learning_rate": 3.39e-05,
      "loss": 2.5795,
      "step": 46950
    },
    {
      "epoch": 6.678982520960637,
      "grad_norm": 1.7126643657684326,
      "learning_rate": 3.334444444444445e-05,
      "loss": 2.5771,
      "step": 47000
    },
    {
      "epoch": 6.68608782151485,
      "grad_norm": 1.7277473211288452,
      "learning_rate": 3.278888888888889e-05,
      "loss": 2.579,
      "step": 47050
    },
    {
      "epoch": 6.693193122069063,
      "grad_norm": 1.5491448640823364,
      "learning_rate": 3.2233333333333335e-05,
      "loss": 2.5959,
      "step": 47100
    },
    {
      "epoch": 6.700298422623277,
      "grad_norm": 1.6598379611968994,
      "learning_rate": 3.167777777777778e-05,
      "loss": 2.5826,
      "step": 47150
    },
    {
      "epoch": 6.707403723177491,
      "grad_norm": 1.711911678314209,
      "learning_rate": 3.112222222222222e-05,
      "loss": 2.5639,
      "step": 47200
    },
    {
      "epoch": 6.714509023731704,
      "grad_norm": 1.6216599941253662,
      "learning_rate": 3.0566666666666667e-05,
      "loss": 2.5626,
      "step": 47250
    },
    {
      "epoch": 6.721614324285917,
      "grad_norm": 1.6412239074707031,
      "learning_rate": 3.001111111111111e-05,
      "loss": 2.5635,
      "step": 47300
    },
    {
      "epoch": 6.728719624840131,
      "grad_norm": 1.738433599472046,
      "learning_rate": 2.9455555555555554e-05,
      "loss": 2.5778,
      "step": 47350
    },
    {
      "epoch": 6.735824925394344,
      "grad_norm": 1.73790442943573,
      "learning_rate": 2.8899999999999998e-05,
      "loss": 2.599,
      "step": 47400
    },
    {
      "epoch": 6.742930225948558,
      "grad_norm": 1.5688190460205078,
      "learning_rate": 2.8344444444444445e-05,
      "loss": 2.5767,
      "step": 47450
    },
    {
      "epoch": 6.750035526502771,
      "grad_norm": 1.561957597732544,
      "learning_rate": 2.778888888888889e-05,
      "loss": 2.5636,
      "step": 47500
    },
    {
      "epoch": 6.757140827056984,
      "grad_norm": 1.6916581392288208,
      "learning_rate": 2.7233333333333332e-05,
      "loss": 2.5993,
      "step": 47550
    },
    {
      "epoch": 6.764246127611198,
      "grad_norm": 1.5806390047073364,
      "learning_rate": 2.667777777777778e-05,
      "loss": 2.5508,
      "step": 47600
    },
    {
      "epoch": 6.771351428165412,
      "grad_norm": 1.6202126741409302,
      "learning_rate": 2.6122222222222223e-05,
      "loss": 2.5829,
      "step": 47650
    },
    {
      "epoch": 6.7784567287196245,
      "grad_norm": 1.7637797594070435,
      "learning_rate": 2.5566666666666667e-05,
      "loss": 2.5919,
      "step": 47700
    },
    {
      "epoch": 6.785562029273838,
      "grad_norm": 1.7567532062530518,
      "learning_rate": 2.5011111111111114e-05,
      "loss": 2.5543,
      "step": 47750
    },
    {
      "epoch": 6.792667329828052,
      "grad_norm": 1.568477988243103,
      "learning_rate": 2.4455555555555558e-05,
      "loss": 2.571,
      "step": 47800
    },
    {
      "epoch": 6.799772630382265,
      "grad_norm": 1.640987515449524,
      "learning_rate": 2.39e-05,
      "loss": 2.5779,
      "step": 47850
    },
    {
      "epoch": 6.806877930936478,
      "grad_norm": 1.6859703063964844,
      "learning_rate": 2.3344444444444445e-05,
      "loss": 2.5774,
      "step": 47900
    },
    {
      "epoch": 6.813983231490692,
      "grad_norm": 1.6906059980392456,
      "learning_rate": 2.278888888888889e-05,
      "loss": 2.5953,
      "step": 47950
    },
    {
      "epoch": 6.8210885320449055,
      "grad_norm": 1.5696355104446411,
      "learning_rate": 2.2233333333333336e-05,
      "loss": 2.6005,
      "step": 48000
    },
    {
      "epoch": 6.8210885320449055,
      "eval_accuracy": 0.540194034576416,
      "eval_loss": 2.564614772796631,
      "eval_runtime": 1.3856,
      "eval_samples_per_second": 2713.711,
      "eval_steps_per_second": 42.582,
      "step": 48000
    },
    {
      "epoch": 6.828193832599119,
      "grad_norm": 1.6884602308273315,
      "learning_rate": 2.167777777777778e-05,
      "loss": 2.5969,
      "step": 48050
    },
    {
      "epoch": 6.835299133153333,
      "grad_norm": 1.663227915763855,
      "learning_rate": 2.1122222222222224e-05,
      "loss": 2.5542,
      "step": 48100
    },
    {
      "epoch": 6.842404433707546,
      "grad_norm": 1.5696767568588257,
      "learning_rate": 2.0566666666666667e-05,
      "loss": 2.5988,
      "step": 48150
    },
    {
      "epoch": 6.849509734261759,
      "grad_norm": 1.564605951309204,
      "learning_rate": 2.001111111111111e-05,
      "loss": 2.5946,
      "step": 48200
    },
    {
      "epoch": 6.856615034815973,
      "grad_norm": 1.5115107297897339,
      "learning_rate": 1.9455555555555555e-05,
      "loss": 2.5823,
      "step": 48250
    },
    {
      "epoch": 6.863720335370186,
      "grad_norm": 1.765989065170288,
      "learning_rate": 1.8900000000000002e-05,
      "loss": 2.5893,
      "step": 48300
    },
    {
      "epoch": 6.870825635924399,
      "grad_norm": 1.6937189102172852,
      "learning_rate": 1.8344444444444446e-05,
      "loss": 2.5764,
      "step": 48350
    },
    {
      "epoch": 6.877930936478613,
      "grad_norm": 1.6565457582473755,
      "learning_rate": 1.778888888888889e-05,
      "loss": 2.5779,
      "step": 48400
    },
    {
      "epoch": 6.885036237032827,
      "grad_norm": 1.5109210014343262,
      "learning_rate": 1.7233333333333333e-05,
      "loss": 2.5582,
      "step": 48450
    },
    {
      "epoch": 6.89214153758704,
      "grad_norm": 1.562355399131775,
      "learning_rate": 1.6677777777777777e-05,
      "loss": 2.5819,
      "step": 48500
    },
    {
      "epoch": 6.899246838141253,
      "grad_norm": 1.5795271396636963,
      "learning_rate": 1.612222222222222e-05,
      "loss": 2.5704,
      "step": 48550
    },
    {
      "epoch": 6.906352138695467,
      "grad_norm": 1.5818406343460083,
      "learning_rate": 1.5566666666666668e-05,
      "loss": 2.595,
      "step": 48600
    },
    {
      "epoch": 6.91345743924968,
      "grad_norm": 1.639119029045105,
      "learning_rate": 1.5011111111111112e-05,
      "loss": 2.5691,
      "step": 48650
    },
    {
      "epoch": 6.920562739803894,
      "grad_norm": 1.5755335092544556,
      "learning_rate": 1.4455555555555555e-05,
      "loss": 2.5333,
      "step": 48700
    },
    {
      "epoch": 6.927668040358107,
      "grad_norm": 1.711458683013916,
      "learning_rate": 1.3899999999999999e-05,
      "loss": 2.5762,
      "step": 48750
    },
    {
      "epoch": 6.9347733409123204,
      "grad_norm": 1.4931989908218384,
      "learning_rate": 1.3344444444444446e-05,
      "loss": 2.5961,
      "step": 48800
    },
    {
      "epoch": 6.941878641466534,
      "grad_norm": 1.5901076793670654,
      "learning_rate": 1.278888888888889e-05,
      "loss": 2.5502,
      "step": 48850
    },
    {
      "epoch": 6.948983942020748,
      "grad_norm": 1.7123291492462158,
      "learning_rate": 1.2233333333333334e-05,
      "loss": 2.5433,
      "step": 48900
    },
    {
      "epoch": 6.956089242574961,
      "grad_norm": 1.555472493171692,
      "learning_rate": 1.1677777777777779e-05,
      "loss": 2.5781,
      "step": 48950
    },
    {
      "epoch": 6.963194543129174,
      "grad_norm": 1.6610956192016602,
      "learning_rate": 1.1122222222222223e-05,
      "loss": 2.5722,
      "step": 49000
    },
    {
      "epoch": 6.970299843683388,
      "grad_norm": 1.657886028289795,
      "learning_rate": 1.0566666666666667e-05,
      "loss": 2.555,
      "step": 49050
    },
    {
      "epoch": 6.9774051442376015,
      "grad_norm": 1.5435242652893066,
      "learning_rate": 1.0011111111111112e-05,
      "loss": 2.5819,
      "step": 49100
    },
    {
      "epoch": 6.984510444791814,
      "grad_norm": 1.6257867813110352,
      "learning_rate": 9.455555555555556e-06,
      "loss": 2.5871,
      "step": 49150
    },
    {
      "epoch": 6.991615745346028,
      "grad_norm": 1.5763481855392456,
      "learning_rate": 8.9e-06,
      "loss": 2.5863,
      "step": 49200
    },
    {
      "epoch": 6.998721045900242,
      "grad_norm": 1.6068406105041504,
      "learning_rate": 8.344444444444445e-06,
      "loss": 2.5862,
      "step": 49250
    },
    {
      "epoch": 7.005826346454455,
      "grad_norm": 1.6784113645553589,
      "learning_rate": 7.788888888888889e-06,
      "loss": 2.5952,
      "step": 49300
    },
    {
      "epoch": 7.012931647008669,
      "grad_norm": 1.6384174823760986,
      "learning_rate": 7.233333333333333e-06,
      "loss": 2.5681,
      "step": 49350
    },
    {
      "epoch": 7.020036947562882,
      "grad_norm": 1.6811999082565308,
      "learning_rate": 6.677777777777779e-06,
      "loss": 2.5668,
      "step": 49400
    },
    {
      "epoch": 7.027142248117095,
      "grad_norm": 1.7817960977554321,
      "learning_rate": 6.1222222222222224e-06,
      "loss": 2.5302,
      "step": 49450
    },
    {
      "epoch": 7.034247548671309,
      "grad_norm": 1.679125428199768,
      "learning_rate": 5.566666666666667e-06,
      "loss": 2.5674,
      "step": 49500
    },
    {
      "epoch": 7.041352849225523,
      "grad_norm": 1.8110630512237549,
      "learning_rate": 5.011111111111112e-06,
      "loss": 2.5557,
      "step": 49550
    },
    {
      "epoch": 7.048458149779735,
      "grad_norm": 1.6951816082000732,
      "learning_rate": 4.455555555555555e-06,
      "loss": 2.5737,
      "step": 49600
    },
    {
      "epoch": 7.055563450333949,
      "grad_norm": 1.5674768686294556,
      "learning_rate": 3.9e-06,
      "loss": 2.5628,
      "step": 49650
    },
    {
      "epoch": 7.062668750888163,
      "grad_norm": 1.8455121517181396,
      "learning_rate": 3.3444444444444445e-06,
      "loss": 2.577,
      "step": 49700
    },
    {
      "epoch": 7.069774051442376,
      "grad_norm": 1.6875885725021362,
      "learning_rate": 2.788888888888889e-06,
      "loss": 2.5594,
      "step": 49750
    },
    {
      "epoch": 7.076879351996589,
      "grad_norm": 1.582588791847229,
      "learning_rate": 2.2333333333333333e-06,
      "loss": 2.5725,
      "step": 49800
    },
    {
      "epoch": 7.083984652550803,
      "grad_norm": 1.682963490486145,
      "learning_rate": 1.6777777777777779e-06,
      "loss": 2.556,
      "step": 49850
    },
    {
      "epoch": 7.091089953105016,
      "grad_norm": 1.7682604789733887,
      "learning_rate": 1.1222222222222222e-06,
      "loss": 2.5602,
      "step": 49900
    },
    {
      "epoch": 7.09819525365923,
      "grad_norm": 1.5831440687179565,
      "learning_rate": 5.666666666666667e-07,
      "loss": 2.5521,
      "step": 49950
    },
    {
      "epoch": 7.105300554213443,
      "grad_norm": 1.5679473876953125,
      "learning_rate": 1.1111111111111112e-08,
      "loss": 2.5724,
      "step": 50000
    },
    {
      "epoch": 7.105300554213443,
      "eval_accuracy": 0.5460931658744812,
      "eval_loss": 2.5473690032958984,
      "eval_runtime": 1.4123,
      "eval_samples_per_second": 2662.272,
      "eval_steps_per_second": 41.775,
      "step": 50000
    }
  ],
  "logging_steps": 50,
  "max_steps": 50000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 8,
  "save_steps": 2000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 5.186896471100621e+16,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}