{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 6.036524590163935,
  "eval_steps": 500,
  "global_step": 24000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.000546448087431694,
      "grad_norm": 5.3125,
      "learning_rate": 6.124999999999999e-06,
      "loss": 9.6445,
      "step": 50
    },
    {
      "epoch": 0.001092896174863388,
      "grad_norm": 5.40625,
      "learning_rate": 1.2375e-05,
      "loss": 8.1001,
      "step": 100
    },
    {
      "epoch": 0.001639344262295082,
      "grad_norm": 4.0,
      "learning_rate": 1.8625e-05,
      "loss": 7.5007,
      "step": 150
    },
    {
      "epoch": 0.002185792349726776,
      "grad_norm": 6.125,
      "learning_rate": 2.4874999999999998e-05,
      "loss": 7.0884,
      "step": 200
    },
    {
      "epoch": 0.00273224043715847,
      "grad_norm": 3.703125,
      "learning_rate": 3.1125e-05,
      "loss": 6.6637,
      "step": 250
    },
    {
      "epoch": 0.003278688524590164,
      "grad_norm": 3.25,
      "learning_rate": 3.7375e-05,
      "loss": 6.3275,
      "step": 300
    },
    {
      "epoch": 0.003825136612021858,
      "grad_norm": 2.578125,
      "learning_rate": 4.3624999999999997e-05,
      "loss": 6.0834,
      "step": 350
    },
    {
      "epoch": 0.004371584699453552,
      "grad_norm": 2.640625,
      "learning_rate": 4.9875e-05,
      "loss": 5.8539,
      "step": 400
    },
    {
      "epoch": 0.004918032786885246,
      "grad_norm": 2.578125,
      "learning_rate": 5.6124999999999995e-05,
      "loss": 5.7353,
      "step": 450
    },
    {
      "epoch": 0.00546448087431694,
      "grad_norm": 2.75,
      "learning_rate": 6.2375e-05,
      "loss": 5.6411,
      "step": 500
    },
    {
      "epoch": 0.006010928961748634,
      "grad_norm": 1.8359375,
      "learning_rate": 6.8625e-05,
      "loss": 5.488,
      "step": 550
    },
    {
      "epoch": 0.006557377049180328,
      "grad_norm": 2.609375,
      "learning_rate": 7.487499999999999e-05,
      "loss": 5.3441,
      "step": 600
    },
    {
      "epoch": 0.007103825136612022,
      "grad_norm": 1.765625,
      "learning_rate": 8.112500000000001e-05,
      "loss": 5.2922,
      "step": 650
    },
    {
      "epoch": 0.007650273224043716,
      "grad_norm": 1.7265625,
      "learning_rate": 8.7375e-05,
      "loss": 5.1776,
      "step": 700
    },
    {
      "epoch": 0.00819672131147541,
      "grad_norm": 2.28125,
      "learning_rate": 9.362499999999999e-05,
      "loss": 5.0849,
      "step": 750
    },
    {
      "epoch": 0.008743169398907104,
      "grad_norm": 1.171875,
      "learning_rate": 9.9875e-05,
      "loss": 4.9607,
      "step": 800
    },
    {
      "epoch": 0.009289617486338797,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00010612499999999999,
      "loss": 4.8847,
      "step": 850
    },
    {
      "epoch": 0.009836065573770493,
      "grad_norm": 1.15625,
      "learning_rate": 0.000112375,
      "loss": 4.8233,
      "step": 900
    },
    {
      "epoch": 0.010382513661202186,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000118625,
      "loss": 4.7077,
      "step": 950
    },
    {
      "epoch": 0.01092896174863388,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00012487499999999999,
      "loss": 4.6274,
      "step": 1000
    },
    {
      "epoch": 0.011475409836065573,
      "grad_norm": 1.15625,
      "learning_rate": 0.00013112499999999998,
      "loss": 4.5506,
      "step": 1050
    },
    {
      "epoch": 0.012021857923497269,
      "grad_norm": 1.03125,
      "learning_rate": 0.000137375,
      "loss": 4.4801,
      "step": 1100
    },
    {
      "epoch": 0.012568306010928962,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00014362499999999998,
      "loss": 4.425,
      "step": 1150
    },
    {
      "epoch": 0.013114754098360656,
      "grad_norm": 1.046875,
      "learning_rate": 0.000149875,
      "loss": 4.3204,
      "step": 1200
    },
    {
      "epoch": 0.01366120218579235,
      "grad_norm": 7.5,
      "learning_rate": 0.000156125,
      "loss": 4.2244,
      "step": 1250
    },
    {
      "epoch": 0.014207650273224045,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00016237499999999998,
      "loss": 4.1705,
      "step": 1300
    },
    {
      "epoch": 0.014754098360655738,
      "grad_norm": 0.83984375,
      "learning_rate": 0.000168625,
      "loss": 4.1256,
      "step": 1350
    },
    {
      "epoch": 0.015300546448087432,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00017487499999999998,
      "loss": 4.0456,
      "step": 1400
    },
    {
      "epoch": 0.015846994535519125,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000181125,
      "loss": 3.9429,
      "step": 1450
    },
    {
      "epoch": 0.01639344262295082,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000187375,
      "loss": 3.8733,
      "step": 1500
    },
    {
      "epoch": 0.016939890710382512,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00019362499999999998,
      "loss": 3.8094,
      "step": 1550
    },
    {
      "epoch": 0.017486338797814208,
      "grad_norm": 0.890625,
      "learning_rate": 0.000199875,
      "loss": 3.8117,
      "step": 1600
    },
    {
      "epoch": 0.018032786885245903,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00020612499999999998,
      "loss": 3.6919,
      "step": 1650
    },
    {
      "epoch": 0.018579234972677595,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00021237499999999997,
      "loss": 3.6148,
      "step": 1700
    },
    {
      "epoch": 0.01912568306010929,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000218625,
      "loss": 3.5881,
      "step": 1750
    },
    {
      "epoch": 0.019672131147540985,
      "grad_norm": 0.953125,
      "learning_rate": 0.000224875,
      "loss": 3.46,
      "step": 1800
    },
    {
      "epoch": 0.020218579234972677,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00023112499999999997,
      "loss": 3.4731,
      "step": 1850
    },
    {
      "epoch": 0.020765027322404372,
      "grad_norm": 0.984375,
      "learning_rate": 0.00023737499999999998,
      "loss": 3.3781,
      "step": 1900
    },
    {
      "epoch": 0.021311475409836064,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000243625,
      "loss": 3.3166,
      "step": 1950
    },
    {
      "epoch": 0.02185792349726776,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000249875,
      "loss": 3.3105,
      "step": 2000
    },
    {
      "epoch": 0.022404371584699455,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000256125,
      "loss": 3.25,
      "step": 2050
    },
    {
      "epoch": 0.022950819672131147,
      "grad_norm": 0.9375,
      "learning_rate": 0.00026237499999999997,
      "loss": 3.1414,
      "step": 2100
    },
    {
      "epoch": 0.023497267759562842,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000268625,
      "loss": 3.1565,
      "step": 2150
    },
    {
      "epoch": 0.024043715846994537,
      "grad_norm": 0.80859375,
      "learning_rate": 0.000274875,
      "loss": 3.1131,
      "step": 2200
    },
    {
      "epoch": 0.02459016393442623,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00028112499999999996,
      "loss": 3.0784,
      "step": 2250
    },
    {
      "epoch": 0.025136612021857924,
      "grad_norm": 0.80859375,
      "learning_rate": 0.000287375,
      "loss": 3.0332,
      "step": 2300
    },
    {
      "epoch": 0.025683060109289616,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000293625,
      "loss": 3.0955,
      "step": 2350
    },
    {
      "epoch": 0.02622950819672131,
      "grad_norm": 0.734375,
      "learning_rate": 0.000299875,
      "loss": 3.045,
      "step": 2400
    },
    {
      "epoch": 0.026775956284153007,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0002999997761290961,
      "loss": 2.995,
      "step": 2450
    },
    {
      "epoch": 0.0273224043715847,
      "grad_norm": 0.75390625,
      "learning_rate": 0.0002999990861486685,
      "loss": 2.9428,
      "step": 2500
    },
    {
      "epoch": 0.027868852459016394,
      "grad_norm": 0.71484375,
      "learning_rate": 0.00029999792996762107,
      "loss": 2.9131,
      "step": 2550
    },
    {
      "epoch": 0.02841530054644809,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00029999630758954706,
      "loss": 2.896,
      "step": 2600
    },
    {
      "epoch": 0.02896174863387978,
      "grad_norm": 0.73828125,
      "learning_rate": 0.000299994219019489,
      "loss": 2.8605,
      "step": 2650
    },
    {
      "epoch": 0.029508196721311476,
      "grad_norm": 0.66796875,
      "learning_rate": 0.0002999916642639382,
      "loss": 2.8407,
      "step": 2700
    },
    {
      "epoch": 0.030054644808743168,
      "grad_norm": 0.68359375,
      "learning_rate": 0.0002999886433308348,
      "loss": 2.8313,
      "step": 2750
    },
    {
      "epoch": 0.030601092896174863,
      "grad_norm": 0.69140625,
      "learning_rate": 0.00029998515622956803,
      "loss": 2.8194,
      "step": 2800
    },
    {
      "epoch": 0.03114754098360656,
      "grad_norm": 0.61328125,
      "learning_rate": 0.00029998120297097586,
      "loss": 2.7874,
      "step": 2850
    },
    {
      "epoch": 0.03169398907103825,
      "grad_norm": 0.67578125,
      "learning_rate": 0.00029997678356734504,
      "loss": 2.7631,
      "step": 2900
    },
    {
      "epoch": 0.03224043715846994,
      "grad_norm": 0.6875,
      "learning_rate": 0.0002999718980324113,
      "loss": 2.7603,
      "step": 2950
    },
    {
      "epoch": 0.03278688524590164,
      "grad_norm": 0.62890625,
      "learning_rate": 0.0002999665463813589,
      "loss": 2.7229,
      "step": 3000
    },
    {
      "epoch": 0.03333333333333333,
      "grad_norm": 0.671875,
      "learning_rate": 0.00029996072863082093,
      "loss": 2.7895,
      "step": 3050
    },
    {
      "epoch": 0.033879781420765025,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0002999544447988791,
      "loss": 2.6505,
      "step": 3100
    },
    {
      "epoch": 0.03442622950819672,
      "grad_norm": 0.60546875,
      "learning_rate": 0.0002999476949050637,
      "loss": 2.6744,
      "step": 3150
    },
    {
      "epoch": 0.034972677595628415,
      "grad_norm": 0.59375,
      "learning_rate": 0.0002999404789703535,
      "loss": 2.6869,
      "step": 3200
    },
    {
      "epoch": 0.03551912568306011,
      "grad_norm": 0.76953125,
      "learning_rate": 0.0002999327970171759,
      "loss": 2.6726,
      "step": 3250
    },
    {
      "epoch": 0.036065573770491806,
      "grad_norm": 0.66015625,
      "learning_rate": 0.0002999246490694065,
      "loss": 2.6444,
      "step": 3300
    },
    {
      "epoch": 0.0366120218579235,
      "grad_norm": 0.69921875,
      "learning_rate": 0.0002999160351523693,
      "loss": 2.6568,
      "step": 3350
    },
    {
      "epoch": 0.03715846994535519,
      "grad_norm": 0.625,
      "learning_rate": 0.00029990695529283665,
      "loss": 2.6436,
      "step": 3400
    },
    {
      "epoch": 1.0000765027322405,
      "grad_norm": 0.5390625,
      "learning_rate": 0.00029989740951902885,
      "loss": 2.6468,
      "step": 3450
    },
    {
      "epoch": 1.0006229508196722,
      "grad_norm": 0.578125,
      "learning_rate": 0.0002998873978606145,
      "loss": 2.5703,
      "step": 3500
    },
    {
      "epoch": 1.0011693989071038,
      "grad_norm": 0.6171875,
      "learning_rate": 0.0002998769203487099,
      "loss": 2.6321,
      "step": 3550
    },
    {
      "epoch": 1.0017158469945355,
      "grad_norm": 0.62109375,
      "learning_rate": 0.0002998659770158796,
      "loss": 2.5518,
      "step": 3600
    },
    {
      "epoch": 1.0022622950819673,
      "grad_norm": 0.6171875,
      "learning_rate": 0.0002998545678961356,
      "loss": 2.5255,
      "step": 3650
    },
    {
      "epoch": 1.0028087431693988,
      "grad_norm": 0.66015625,
      "learning_rate": 0.00029984269302493776,
      "loss": 2.4976,
      "step": 3700
    },
    {
      "epoch": 1.0033551912568306,
      "grad_norm": 0.58984375,
      "learning_rate": 0.0002998303524391934,
      "loss": 2.532,
      "step": 3750
    },
    {
      "epoch": 1.0039016393442624,
      "grad_norm": 0.6484375,
      "learning_rate": 0.00029981754617725747,
      "loss": 2.5321,
      "step": 3800
    },
    {
      "epoch": 1.004448087431694,
      "grad_norm": 0.53515625,
      "learning_rate": 0.0002998042742789319,
      "loss": 2.4924,
      "step": 3850
    },
    {
      "epoch": 1.0049945355191257,
      "grad_norm": 0.6015625,
      "learning_rate": 0.0002997905367854663,
      "loss": 2.492,
      "step": 3900
    },
    {
      "epoch": 1.0055409836065574,
      "grad_norm": 0.61328125,
      "learning_rate": 0.00029977633373955696,
      "loss": 2.5266,
      "step": 3950
    },
    {
      "epoch": 1.0060874316939892,
      "grad_norm": 0.58984375,
      "learning_rate": 0.00029976166518534735,
      "loss": 2.4739,
      "step": 4000
    },
    {
      "epoch": 1.0066338797814207,
      "grad_norm": 0.5546875,
      "learning_rate": 0.00029974653116842764,
      "loss": 2.4487,
      "step": 4050
    },
    {
      "epoch": 1.0071803278688525,
      "grad_norm": 0.60546875,
      "learning_rate": 0.0002997309317358347,
      "loss": 2.4674,
      "step": 4100
    },
    {
      "epoch": 1.0077267759562842,
      "grad_norm": 0.67578125,
      "learning_rate": 0.0002997148669360519,
      "loss": 2.4814,
      "step": 4150
    },
    {
      "epoch": 1.0082732240437158,
      "grad_norm": 0.53515625,
      "learning_rate": 0.00029969833681900914,
      "loss": 2.448,
      "step": 4200
    },
    {
      "epoch": 1.0088196721311475,
      "grad_norm": 0.68359375,
      "learning_rate": 0.0002996813414360822,
      "loss": 2.4299,
      "step": 4250
    },
    {
      "epoch": 1.0093661202185793,
      "grad_norm": 0.57421875,
      "learning_rate": 0.00029966388084009334,
      "loss": 2.4271,
      "step": 4300
    },
    {
      "epoch": 1.0099125683060108,
      "grad_norm": 0.515625,
      "learning_rate": 0.00029964595508531034,
      "loss": 2.4848,
      "step": 4350
    },
    {
      "epoch": 1.0104590163934426,
      "grad_norm": 0.53125,
      "learning_rate": 0.00029962756422744695,
      "loss": 2.414,
      "step": 4400
    },
    {
      "epoch": 1.0110054644808744,
      "grad_norm": 0.51171875,
      "learning_rate": 0.00029960870832366224,
      "loss": 2.3993,
      "step": 4450
    },
    {
      "epoch": 1.0115519125683061,
      "grad_norm": 0.54296875,
      "learning_rate": 0.000299589387432561,
      "loss": 2.4171,
      "step": 4500
    },
    {
      "epoch": 1.0120983606557377,
      "grad_norm": 0.490234375,
      "learning_rate": 0.00029956960161419283,
      "loss": 2.4038,
      "step": 4550
    },
    {
      "epoch": 1.0126448087431694,
      "grad_norm": 0.498046875,
      "learning_rate": 0.0002995493509300526,
      "loss": 2.4128,
      "step": 4600
    },
    {
      "epoch": 1.0131912568306012,
      "grad_norm": 0.51171875,
      "learning_rate": 0.0002995286354430799,
      "loss": 2.3721,
      "step": 4650
    },
    {
      "epoch": 1.0137377049180327,
      "grad_norm": 0.51171875,
      "learning_rate": 0.0002995074552176589,
      "loss": 2.3734,
      "step": 4700
    },
    {
      "epoch": 1.0142841530054645,
      "grad_norm": 0.515625,
      "learning_rate": 0.00029948581031961826,
      "loss": 2.3805,
      "step": 4750
    },
    {
      "epoch": 1.0148306010928962,
      "grad_norm": 0.515625,
      "learning_rate": 0.0002994637008162308,
      "loss": 2.3819,
      "step": 4800
    },
    {
      "epoch": 1.0153770491803278,
      "grad_norm": 0.53515625,
      "learning_rate": 0.00029944112677621345,
      "loss": 2.3839,
      "step": 4850
    },
    {
      "epoch": 1.0159234972677595,
      "grad_norm": 0.4921875,
      "learning_rate": 0.00029941808826972673,
      "loss": 2.336,
      "step": 4900
    },
    {
      "epoch": 1.0164699453551913,
      "grad_norm": 0.515625,
      "learning_rate": 0.0002993945853683749,
      "loss": 2.3126,
      "step": 4950
    },
    {
      "epoch": 1.0170163934426228,
      "grad_norm": 0.53515625,
      "learning_rate": 0.00029937061814520546,
      "loss": 2.3271,
      "step": 5000
    },
    {
      "epoch": 1.0175628415300546,
      "grad_norm": 0.53125,
      "learning_rate": 0.00029934618667470925,
      "loss": 2.3275,
      "step": 5050
    },
    {
      "epoch": 1.0181092896174864,
      "grad_norm": 0.51953125,
      "learning_rate": 0.0002993212910328197,
      "loss": 2.2837,
      "step": 5100
    },
    {
      "epoch": 1.0186557377049181,
      "grad_norm": 0.56640625,
      "learning_rate": 0.00029929593129691305,
      "loss": 2.2964,
      "step": 5150
    },
    {
      "epoch": 1.0192021857923497,
      "grad_norm": 0.50390625,
      "learning_rate": 0.000299270107545808,
      "loss": 2.3155,
      "step": 5200
    },
    {
      "epoch": 1.0197486338797814,
      "grad_norm": 0.55078125,
      "learning_rate": 0.00029924381985976534,
      "loss": 2.2722,
      "step": 5250
    },
    {
      "epoch": 1.0202950819672132,
      "grad_norm": 0.49609375,
      "learning_rate": 0.00029921706832048784,
      "loss": 2.3175,
      "step": 5300
    },
    {
      "epoch": 1.0208415300546447,
      "grad_norm": 0.48828125,
      "learning_rate": 0.00029918985301111985,
      "loss": 2.2834,
      "step": 5350
    },
    {
      "epoch": 1.0213879781420765,
      "grad_norm": 0.734375,
      "learning_rate": 0.00029916217401624716,
      "loss": 2.2522,
      "step": 5400
    },
    {
      "epoch": 1.0219344262295083,
      "grad_norm": 0.46484375,
      "learning_rate": 0.00029913403142189677,
      "loss": 2.2872,
      "step": 5450
    },
    {
      "epoch": 1.0224808743169398,
      "grad_norm": 0.51953125,
      "learning_rate": 0.00029910542531553656,
      "loss": 2.2793,
      "step": 5500
    },
    {
      "epoch": 1.0230273224043716,
      "grad_norm": 0.474609375,
      "learning_rate": 0.00029907635578607487,
      "loss": 2.218,
      "step": 5550
    },
    {
      "epoch": 1.0235737704918033,
      "grad_norm": 0.49609375,
      "learning_rate": 0.00029904682292386053,
      "loss": 2.2309,
      "step": 5600
    },
    {
      "epoch": 1.024120218579235,
      "grad_norm": 0.52734375,
      "learning_rate": 0.0002990168268206823,
      "loss": 2.2285,
      "step": 5650
    },
    {
      "epoch": 1.0246666666666666,
      "grad_norm": 0.48828125,
      "learning_rate": 0.00029898636756976884,
      "loss": 2.2338,
      "step": 5700
    },
    {
      "epoch": 1.0252131147540984,
      "grad_norm": 0.462890625,
      "learning_rate": 0.0002989554452657881,
      "loss": 2.2048,
      "step": 5750
    },
    {
      "epoch": 1.0257595628415301,
      "grad_norm": 0.62109375,
      "learning_rate": 0.0002989240600048475,
      "loss": 2.2716,
      "step": 5800
    },
    {
      "epoch": 1.0263060109289617,
      "grad_norm": 0.6015625,
      "learning_rate": 0.00029889221188449295,
      "loss": 2.2618,
      "step": 5850
    },
    {
      "epoch": 1.0268524590163934,
      "grad_norm": 0.47265625,
      "learning_rate": 0.0002988599010037092,
      "loss": 2.2181,
      "step": 5900
    },
    {
      "epoch": 1.0273989071038252,
      "grad_norm": 0.5234375,
      "learning_rate": 0.0002988271274629192,
      "loss": 2.2005,
      "step": 5950
    },
    {
      "epoch": 1.0279453551912567,
      "grad_norm": 0.515625,
      "learning_rate": 0.00029879389136398403,
      "loss": 2.1958,
      "step": 6000
    },
    {
      "epoch": 1.0284918032786885,
      "grad_norm": 0.52734375,
      "learning_rate": 0.00029876019281020207,
      "loss": 2.1853,
      "step": 6050
    },
    {
      "epoch": 1.0290382513661203,
      "grad_norm": 0.52734375,
      "learning_rate": 0.00029872603190630927,
      "loss": 2.1753,
      "step": 6100
    },
    {
      "epoch": 1.029584699453552,
      "grad_norm": 0.5546875,
      "learning_rate": 0.00029869140875847847,
      "loss": 2.1931,
      "step": 6150
    },
    {
      "epoch": 1.0301311475409836,
      "grad_norm": 0.53125,
      "learning_rate": 0.0002986563234743193,
      "loss": 2.1846,
      "step": 6200
    },
    {
      "epoch": 1.0306775956284153,
      "grad_norm": 0.51953125,
      "learning_rate": 0.0002986207761628775,
      "loss": 2.1928,
      "step": 6250
    },
    {
      "epoch": 1.031224043715847,
      "grad_norm": 0.6171875,
      "learning_rate": 0.00029858476693463506,
      "loss": 2.1942,
      "step": 6300
    },
    {
      "epoch": 1.0317704918032786,
      "grad_norm": 0.48828125,
      "learning_rate": 0.0002985482959015094,
      "loss": 2.1653,
      "step": 6350
    },
    {
      "epoch": 1.0323169398907104,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00029851136317685345,
      "loss": 2.1659,
      "step": 6400
    },
    {
      "epoch": 1.0328633879781421,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00029847396887545485,
      "loss": 2.1829,
      "step": 6450
    },
    {
      "epoch": 1.0334098360655737,
      "grad_norm": 0.54296875,
      "learning_rate": 0.00029843611311353597,
      "loss": 2.1911,
      "step": 6500
    },
    {
      "epoch": 1.0339562841530054,
      "grad_norm": 0.51953125,
      "learning_rate": 0.00029839779600875343,
      "loss": 2.1041,
      "step": 6550
    },
    {
      "epoch": 1.0345027322404372,
      "grad_norm": 0.53125,
      "learning_rate": 0.00029835901768019763,
      "loss": 2.1634,
      "step": 6600
    },
    {
      "epoch": 1.0350491803278687,
      "grad_norm": 0.515625,
      "learning_rate": 0.0002983197782483926,
      "loss": 2.1642,
      "step": 6650
    },
    {
      "epoch": 1.0355956284153005,
      "grad_norm": 0.58203125,
      "learning_rate": 0.00029828007783529533,
      "loss": 2.1621,
      "step": 6700
    },
    {
      "epoch": 1.0361420765027323,
      "grad_norm": 0.5078125,
      "learning_rate": 0.0002982399165642956,
      "loss": 2.1553,
      "step": 6750
    },
    {
      "epoch": 1.036688524590164,
      "grad_norm": 0.51171875,
      "learning_rate": 0.00029819929456021565,
      "loss": 2.1592,
      "step": 6800
    },
    {
      "epoch": 1.0372349726775956,
      "grad_norm": 0.51953125,
      "learning_rate": 0.0002981582119493095,
      "loss": 2.1527,
      "step": 6850
    },
    {
      "epoch": 2.000153005464481,
      "grad_norm": 0.5,
      "learning_rate": 0.0002981166688592629,
      "loss": 2.1674,
      "step": 6900
    },
    {
      "epoch": 2.0006994535519125,
      "grad_norm": 0.5390625,
      "learning_rate": 0.00029807466541919273,
      "loss": 2.1168,
      "step": 6950
    },
    {
      "epoch": 2.0012459016393445,
      "grad_norm": 0.53515625,
      "learning_rate": 0.00029803220175964675,
      "loss": 2.1439,
      "step": 7000
    },
    {
      "epoch": 2.001792349726776,
      "grad_norm": 0.5,
      "learning_rate": 0.0002979892780126028,
      "loss": 2.0952,
      "step": 7050
    },
    {
      "epoch": 2.0023387978142075,
      "grad_norm": 0.490234375,
      "learning_rate": 0.00029794589431146904,
      "loss": 2.0817,
      "step": 7100
    },
    {
      "epoch": 2.0028852459016395,
      "grad_norm": 0.462890625,
      "learning_rate": 0.00029790205079108294,
      "loss": 2.0643,
      "step": 7150
    },
    {
      "epoch": 2.003431693989071,
      "grad_norm": 0.55078125,
      "learning_rate": 0.00029785774758771114,
      "loss": 2.0993,
      "step": 7200
    },
    {
      "epoch": 2.0039781420765026,
      "grad_norm": 0.51953125,
      "learning_rate": 0.00029781298483904907,
      "loss": 2.1085,
      "step": 7250
    },
    {
      "epoch": 2.0045245901639346,
      "grad_norm": 0.4921875,
      "learning_rate": 0.0002977677626842204,
      "loss": 2.0645,
      "step": 7300
    },
    {
      "epoch": 2.005071038251366,
      "grad_norm": 0.50390625,
      "learning_rate": 0.0002977220812637766,
      "loss": 2.0929,
      "step": 7350
    },
    {
      "epoch": 2.0056174863387977,
      "grad_norm": 0.53515625,
      "learning_rate": 0.0002976759407196966,
      "loss": 2.0919,
      "step": 7400
    },
    {
      "epoch": 2.0061639344262296,
      "grad_norm": 0.55078125,
      "learning_rate": 0.00029762934119538623,
      "loss": 2.0903,
      "step": 7450
    },
    {
      "epoch": 2.006710382513661,
      "grad_norm": 0.5625,
      "learning_rate": 0.00029758228283567796,
      "loss": 2.0481,
      "step": 7500
    },
    {
      "epoch": 2.0072568306010927,
      "grad_norm": 0.5703125,
      "learning_rate": 0.00029753476578683023,
      "loss": 2.0737,
      "step": 7550
    },
    {
      "epoch": 2.0078032786885247,
      "grad_norm": 0.49609375,
      "learning_rate": 0.00029748679019652704,
      "loss": 2.1032,
      "step": 7600
    },
    {
      "epoch": 2.0083497267759562,
      "grad_norm": 0.5546875,
      "learning_rate": 0.00029743835621387775,
      "loss": 2.0722,
      "step": 7650
    },
    {
      "epoch": 2.008896174863388,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00029738946398941623,
      "loss": 2.057,
      "step": 7700
    },
    {
      "epoch": 2.0094426229508198,
      "grad_norm": 0.490234375,
      "learning_rate": 0.0002973401136751007,
      "loss": 2.0802,
      "step": 7750
    },
    {
      "epoch": 2.0099890710382513,
      "grad_norm": 0.5,
      "learning_rate": 0.0002972903054243129,
      "loss": 2.1094,
      "step": 7800
    },
    {
      "epoch": 2.010535519125683,
      "grad_norm": 0.515625,
      "learning_rate": 0.0002972400393918583,
      "loss": 2.0409,
      "step": 7850
    },
    {
      "epoch": 2.011081967213115,
      "grad_norm": 0.578125,
      "learning_rate": 0.0002971893157339647,
      "loss": 2.0543,
      "step": 7900
    },
    {
      "epoch": 2.0116284153005464,
      "grad_norm": 0.5234375,
      "learning_rate": 0.0002971381346082824,
      "loss": 2.0776,
      "step": 7950
    },
    {
      "epoch": 2.0121748633879784,
      "grad_norm": 0.55078125,
      "learning_rate": 0.00029708649617388356,
      "loss": 2.0629,
      "step": 8000
    },
    {
      "epoch": 2.01272131147541,
      "grad_norm": 0.515625,
      "learning_rate": 0.0002970344005912617,
      "loss": 2.0588,
      "step": 8050
    },
    {
      "epoch": 2.0132677595628414,
      "grad_norm": 0.4765625,
      "learning_rate": 0.000296981848022331,
      "loss": 2.0373,
      "step": 8100
    },
    {
      "epoch": 2.0138142076502734,
      "grad_norm": 0.44921875,
      "learning_rate": 0.000296928838630426,
      "loss": 2.0348,
      "step": 8150
    },
    {
      "epoch": 2.014360655737705,
      "grad_norm": 0.490234375,
      "learning_rate": 0.0002968753725803013,
      "loss": 2.0784,
      "step": 8200
    },
    {
      "epoch": 2.0149071038251365,
      "grad_norm": 0.5390625,
      "learning_rate": 0.0002968214500381304,
      "loss": 2.0531,
      "step": 8250
    },
    {
      "epoch": 2.0154535519125685,
      "grad_norm": 0.546875,
      "learning_rate": 0.000296767071171506,
      "loss": 2.0482,
      "step": 8300
    },
    {
      "epoch": 2.016,
      "grad_norm": 0.484375,
      "learning_rate": 0.00029671223614943874,
      "loss": 2.0193,
      "step": 8350
    },
    {
      "epoch": 2.0165464480874316,
      "grad_norm": 0.4765625,
      "learning_rate": 0.0002966569451423572,
      "loss": 2.007,
      "step": 8400
    },
    {
      "epoch": 2.0170928961748635,
      "grad_norm": 0.48828125,
      "learning_rate": 0.000296601198322107,
      "loss": 2.0325,
      "step": 8450
    },
    {
      "epoch": 2.017639344262295,
      "grad_norm": 0.4921875,
      "learning_rate": 0.0002965449958619508,
      "loss": 2.0173,
      "step": 8500
    },
    {
      "epoch": 2.0181857923497266,
      "grad_norm": 0.51953125,
      "learning_rate": 0.0002964883379365668,
      "loss": 1.9927,
      "step": 8550
    },
    {
      "epoch": 2.0187322404371586,
      "grad_norm": 0.546875,
      "learning_rate": 0.00029643122472204934,
      "loss": 2.0149,
      "step": 8600
    },
    {
      "epoch": 2.01927868852459,
      "grad_norm": 0.49609375,
      "learning_rate": 0.00029637365639590763,
      "loss": 2.0077,
      "step": 8650
    },
    {
      "epoch": 2.0198251366120217,
      "grad_norm": 0.5390625,
      "learning_rate": 0.00029631563313706525,
      "loss": 1.9926,
      "step": 8700
    },
    {
      "epoch": 2.0203715846994537,
      "grad_norm": 0.494140625,
      "learning_rate": 0.0002962571551258599,
      "loss": 2.0248,
      "step": 8750
    },
    {
      "epoch": 2.020918032786885,
      "grad_norm": 0.51953125,
      "learning_rate": 0.00029619822254404256,
      "loss": 1.998,
      "step": 8800
    },
    {
      "epoch": 2.0214644808743167,
      "grad_norm": 0.478515625,
      "learning_rate": 0.00029613883557477706,
      "loss": 1.9957,
      "step": 8850
    },
    {
      "epoch": 2.0220109289617487,
      "grad_norm": 0.49609375,
      "learning_rate": 0.00029607899440263946,
      "loss": 2.0132,
      "step": 8900
    },
    {
      "epoch": 2.0225573770491803,
      "grad_norm": 0.5390625,
      "learning_rate": 0.00029601869921361756,
      "loss": 2.0038,
      "step": 8950
    },
    {
      "epoch": 2.0231038251366122,
      "grad_norm": 0.458984375,
      "learning_rate": 0.00029595795019511005,
      "loss": 1.9447,
      "step": 9000
    },
    {
      "epoch": 2.023650273224044,
      "grad_norm": 0.5078125,
      "learning_rate": 0.00029589674753592647,
      "loss": 1.9806,
      "step": 9050
    },
    {
      "epoch": 2.0241967213114753,
      "grad_norm": 0.5546875,
      "learning_rate": 0.000295835091426286,
      "loss": 1.9738,
      "step": 9100
    },
    {
      "epoch": 2.0247431693989073,
      "grad_norm": 0.5078125,
      "learning_rate": 0.00029577298205781726,
      "loss": 1.9568,
      "step": 9150
    },
    {
      "epoch": 2.025289617486339,
      "grad_norm": 0.5078125,
      "learning_rate": 0.00029571041962355755,
      "loss": 1.9778,
      "step": 9200
    },
    {
      "epoch": 2.0258360655737704,
      "grad_norm": 0.5078125,
      "learning_rate": 0.0002956474043179525,
      "loss": 2.0212,
      "step": 9250
    },
    {
      "epoch": 2.0263825136612024,
      "grad_norm": 0.51953125,
      "learning_rate": 0.0002955839363368549,
      "loss": 1.9981,
      "step": 9300
    },
    {
      "epoch": 2.026928961748634,
      "grad_norm": 0.5,
      "learning_rate": 0.00029552001587752495,
      "loss": 1.9763,
      "step": 9350
    },
    {
      "epoch": 2.0274754098360654,
      "grad_norm": 0.515625,
      "learning_rate": 0.0002954556431386288,
      "loss": 1.9687,
      "step": 9400
    },
    {
      "epoch": 2.0280218579234974,
      "grad_norm": 0.5390625,
      "learning_rate": 0.00029539081832023837,
      "loss": 1.9391,
      "step": 9450
    },
    {
      "epoch": 2.028568306010929,
      "grad_norm": 0.56640625,
      "learning_rate": 0.0002953255416238308,
      "loss": 1.9614,
      "step": 9500
    },
    {
      "epoch": 2.0291147540983605,
      "grad_norm": 0.490234375,
      "learning_rate": 0.0002952598132522874,
      "loss": 1.9405,
      "step": 9550
    },
    {
      "epoch": 2.0296612021857925,
      "grad_norm": 0.546875,
      "learning_rate": 0.00029519363340989367,
      "loss": 1.9653,
      "step": 9600
    },
    {
      "epoch": 2.030207650273224,
      "grad_norm": 0.50390625,
      "learning_rate": 0.0002951270023023379,
      "loss": 1.9704,
      "step": 9650
    },
    {
      "epoch": 2.0307540983606556,
      "grad_norm": 0.53515625,
      "learning_rate": 0.00029505992013671126,
      "loss": 1.9592,
      "step": 9700
    },
    {
      "epoch": 2.0313005464480876,
      "grad_norm": 0.474609375,
      "learning_rate": 0.0002949923871215065,
      "loss": 1.9679,
      "step": 9750
    },
    {
      "epoch": 2.031846994535519,
      "grad_norm": 0.5546875,
      "learning_rate": 0.000294924403466618,
      "loss": 1.9398,
      "step": 9800
    },
    {
      "epoch": 2.0323934426229506,
      "grad_norm": 0.5625,
      "learning_rate": 0.00029485596938334037,
      "loss": 1.9469,
      "step": 9850
    },
    {
      "epoch": 2.0329398907103826,
      "grad_norm": 0.51953125,
      "learning_rate": 0.00029478708508436834,
      "loss": 1.9742,
      "step": 9900
    },
    {
      "epoch": 2.033486338797814,
      "grad_norm": 0.482421875,
      "learning_rate": 0.000294717750783796,
      "loss": 1.9619,
      "step": 9950
    },
    {
      "epoch": 2.0340327868852457,
      "grad_norm": 0.6015625,
      "learning_rate": 0.0002946479666971158,
      "loss": 1.8817,
      "step": 10000
    },
    {
      "epoch": 2.0345792349726777,
      "grad_norm": 0.50390625,
      "learning_rate": 0.0002945777330412184,
      "loss": 1.9465,
      "step": 10050
    },
    {
      "epoch": 2.035125683060109,
      "grad_norm": 0.5703125,
      "learning_rate": 0.00029450705003439156,
      "loss": 1.9628,
      "step": 10100
    },
    {
      "epoch": 2.035672131147541,
      "grad_norm": 0.52734375,
      "learning_rate": 0.0002944359178963198,
      "loss": 1.9421,
      "step": 10150
    },
    {
      "epoch": 2.0362185792349727,
      "grad_norm": 0.51171875,
      "learning_rate": 0.00029436433684808336,
      "loss": 1.953,
      "step": 10200
    },
    {
      "epoch": 2.0367650273224043,
      "grad_norm": 0.5390625,
      "learning_rate": 0.0002942923071121578,
      "loss": 1.9499,
      "step": 10250
    },
    {
      "epoch": 2.0373114754098363,
      "grad_norm": 0.53515625,
      "learning_rate": 0.0002942198289124132,
      "loss": 1.9522,
      "step": 10300
    },
    {
      "epoch": 3.000229508196721,
      "grad_norm": 0.5078125,
      "learning_rate": 0.00029414690247411346,
      "loss": 1.9537,
      "step": 10350
    },
    {
      "epoch": 3.000775956284153,
      "grad_norm": 0.5390625,
      "learning_rate": 0.0002940735280239157,
      "loss": 1.9224,
      "step": 10400
    },
    {
      "epoch": 3.0013224043715847,
      "grad_norm": 0.5234375,
      "learning_rate": 0.0002939997057898693,
      "loss": 1.9165,
      "step": 10450
    },
    {
      "epoch": 3.0018688524590162,
      "grad_norm": 0.490234375,
      "learning_rate": 0.0002939254360014156,
      "loss": 1.8814,
      "step": 10500
    },
    {
      "epoch": 3.0024153005464482,
      "grad_norm": 0.52734375,
      "learning_rate": 0.0002938507188893867,
      "loss": 1.8547,
      "step": 10550
    },
    {
      "epoch": 3.0029617486338798,
      "grad_norm": 0.8125,
      "learning_rate": 0.00029377555468600516,
      "loss": 1.9014,
      "step": 10600
    },
    {
      "epoch": 3.0035081967213113,
      "grad_norm": 0.8125,
      "learning_rate": 0.00029369994362488306,
      "loss": 1.8837,
      "step": 10650
    },
    {
      "epoch": 3.0040546448087433,
      "grad_norm": 0.51953125,
      "learning_rate": 0.0002936238859410213,
      "loss": 1.9095,
      "step": 10700
    },
    {
      "epoch": 3.004601092896175,
      "grad_norm": 0.52734375,
      "learning_rate": 0.0002935473818708089,
      "loss": 1.8654,
      "step": 10750
    },
    {
      "epoch": 3.0051475409836064,
      "grad_norm": 0.58203125,
      "learning_rate": 0.00029347043165202233,
      "loss": 1.9018,
      "step": 10800
    },
    {
      "epoch": 3.0056939890710384,
      "grad_norm": 0.51171875,
      "learning_rate": 0.0002933930355238246,
      "loss": 1.895,
      "step": 10850
    },
    {
      "epoch": 3.00624043715847,
      "grad_norm": 0.50390625,
      "learning_rate": 0.0002933151937267647,
      "loss": 1.8872,
      "step": 10900
    },
    {
      "epoch": 3.0067868852459014,
      "grad_norm": 0.53125,
      "learning_rate": 0.0002932369065027767,
      "loss": 1.8532,
      "step": 10950
    },
    {
      "epoch": 3.0073333333333334,
      "grad_norm": 0.546875,
      "learning_rate": 0.0002931581740951791,
      "loss": 1.8935,
      "step": 11000
    },
    {
      "epoch": 3.007879781420765,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00029307899674867405,
      "loss": 1.8991,
      "step": 11050
    },
    {
      "epoch": 3.008426229508197,
      "grad_norm": 0.53515625,
      "learning_rate": 0.00029299937470934656,
      "loss": 1.8784,
      "step": 11100
    },
    {
      "epoch": 3.0089726775956285,
      "grad_norm": 0.52734375,
      "learning_rate": 0.00029291930822466383,
      "loss": 1.8775,
      "step": 11150
    },
    {
      "epoch": 3.00951912568306,
      "grad_norm": 0.56640625,
      "learning_rate": 0.0002928387975434742,
      "loss": 1.8874,
      "step": 11200
    },
    {
      "epoch": 3.010065573770492,
      "grad_norm": 0.59765625,
      "learning_rate": 0.00029275784291600684,
      "loss": 1.9137,
      "step": 11250
    },
    {
      "epoch": 3.0106120218579235,
      "grad_norm": 0.53125,
      "learning_rate": 0.0002926764445938705,
      "loss": 1.8568,
      "step": 11300
    },
    {
      "epoch": 3.011158469945355,
      "grad_norm": 0.53125,
      "learning_rate": 0.0002925946028300532,
      "loss": 1.8578,
      "step": 11350
    },
    {
      "epoch": 3.011704918032787,
      "grad_norm": 0.53515625,
      "learning_rate": 0.0002925123178789209,
      "loss": 1.9092,
      "step": 11400
    },
    {
      "epoch": 3.0122513661202186,
      "grad_norm": 0.515625,
      "learning_rate": 0.00029242958999621717,
      "loss": 1.8663,
      "step": 11450
    },
    {
      "epoch": 3.01279781420765,
      "grad_norm": 0.54296875,
      "learning_rate": 0.00029234641943906223,
      "loss": 1.862,
      "step": 11500
    },
    {
      "epoch": 3.013344262295082,
      "grad_norm": 0.56640625,
      "learning_rate": 0.0002922628064659519,
      "loss": 1.8594,
      "step": 11550
    },
    {
      "epoch": 3.0138907103825137,
      "grad_norm": 0.546875,
      "learning_rate": 0.0002921787513367575,
      "loss": 1.8633,
      "step": 11600
    },
    {
      "epoch": 3.014437158469945,
      "grad_norm": 0.51953125,
      "learning_rate": 0.0002920942543127241,
      "loss": 1.8929,
      "step": 11650
    },
    {
      "epoch": 3.014983606557377,
      "grad_norm": 0.54296875,
      "learning_rate": 0.0002920093156564705,
      "loss": 1.8794,
      "step": 11700
    },
    {
      "epoch": 3.0155300546448087,
      "grad_norm": 0.5234375,
      "learning_rate": 0.0002919239356319879,
      "loss": 1.8691,
      "step": 11750
    },
    {
      "epoch": 3.0160765027322403,
      "grad_norm": 0.53515625,
      "learning_rate": 0.00029183811450463954,
      "loss": 1.8429,
      "step": 11800
    },
    {
      "epoch": 3.0166229508196722,
      "grad_norm": 0.5,
      "learning_rate": 0.00029175185254115934,
      "loss": 1.8325,
      "step": 11850
    },
    {
      "epoch": 3.017169398907104,
      "grad_norm": 0.53125,
      "learning_rate": 0.00029166515000965154,
      "loss": 1.8598,
      "step": 11900
    },
    {
      "epoch": 3.0177158469945353,
      "grad_norm": 0.55078125,
      "learning_rate": 0.0002915780071795896,
      "loss": 1.8376,
      "step": 11950
    },
    {
      "epoch": 3.0182622950819673,
      "grad_norm": 0.57421875,
      "learning_rate": 0.0002914904243218154,
      "loss": 1.8142,
      "step": 12000
    },
    {
      "epoch": 3.018808743169399,
      "grad_norm": 0.55078125,
      "learning_rate": 0.00029140240170853857,
      "loss": 1.8505,
      "step": 12050
    },
    {
      "epoch": 3.0193551912568304,
      "grad_norm": 0.5546875,
      "learning_rate": 0.0002913139396133353,
      "loss": 1.8315,
      "step": 12100
    },
    {
      "epoch": 3.0199016393442624,
      "grad_norm": 0.53515625,
      "learning_rate": 0.0002912250383111479,
      "loss": 1.8337,
      "step": 12150
    },
    {
      "epoch": 3.020448087431694,
      "grad_norm": 0.5234375,
      "learning_rate": 0.0002911356980782837,
      "loss": 1.8647,
      "step": 12200
    },
    {
      "epoch": 3.020994535519126,
      "grad_norm": 0.498046875,
      "learning_rate": 0.0002910459191924141,
      "loss": 1.8303,
      "step": 12250
    },
    {
      "epoch": 3.0215409836065574,
      "grad_norm": 0.515625,
      "learning_rate": 0.00029095570193257405,
      "loss": 1.8347,
      "step": 12300
    },
    {
      "epoch": 3.022087431693989,
      "grad_norm": 0.515625,
      "learning_rate": 0.0002908650465791608,
      "loss": 1.842,
      "step": 12350
    },
    {
      "epoch": 3.022633879781421,
      "grad_norm": 0.48046875,
      "learning_rate": 0.00029077395341393334,
      "loss": 1.8282,
      "step": 12400
    },
    {
      "epoch": 3.0231803278688525,
      "grad_norm": 0.546875,
      "learning_rate": 0.00029068242272001135,
      "loss": 1.7943,
      "step": 12450
    },
    {
      "epoch": 3.023726775956284,
      "grad_norm": 0.5078125,
      "learning_rate": 0.00029059045478187424,
      "loss": 1.8147,
      "step": 12500
    },
    {
      "epoch": 3.024273224043716,
      "grad_norm": 0.51953125,
      "learning_rate": 0.00029049804988536053,
      "loss": 1.8135,
      "step": 12550
    },
    {
      "epoch": 3.0248196721311476,
      "grad_norm": 0.52734375,
      "learning_rate": 0.00029040520831766676,
      "loss": 1.8067,
      "step": 12600
    },
    {
      "epoch": 3.025366120218579,
      "grad_norm": 0.515625,
      "learning_rate": 0.00029031193036734666,
      "loss": 1.8333,
      "step": 12650
    },
    {
      "epoch": 3.025912568306011,
      "grad_norm": 0.5546875,
      "learning_rate": 0.0002902182163243103,
      "loss": 1.8624,
      "step": 12700
    },
    {
      "epoch": 3.0264590163934426,
      "grad_norm": 0.57421875,
      "learning_rate": 0.00029012406647982306,
      "loss": 1.8277,
      "step": 12750
    },
    {
      "epoch": 3.027005464480874,
      "grad_norm": 0.51953125,
      "learning_rate": 0.0002900294811265048,
      "loss": 1.8209,
      "step": 12800
    },
    {
      "epoch": 3.027551912568306,
      "grad_norm": 0.5078125,
      "learning_rate": 0.0002899344605583291,
      "loss": 1.8295,
      "step": 12850
    },
    {
      "epoch": 3.0280983606557377,
      "grad_norm": 0.4921875,
      "learning_rate": 0.0002898390050706219,
      "loss": 1.7926,
      "step": 12900
    },
    {
      "epoch": 3.028644808743169,
      "grad_norm": 0.515625,
      "learning_rate": 0.0002897431149600612,
      "loss": 1.8064,
      "step": 12950
    },
    {
      "epoch": 3.029191256830601,
      "grad_norm": 0.51953125,
      "learning_rate": 0.0002896467905246755,
      "loss": 1.7923,
      "step": 13000
    },
    {
      "epoch": 3.0297377049180327,
      "grad_norm": 0.5625,
      "learning_rate": 0.00028955003206384357,
      "loss": 1.8346,
      "step": 13050
    },
    {
      "epoch": 3.0302841530054643,
      "grad_norm": 0.5078125,
      "learning_rate": 0.0002894528398782929,
      "loss": 1.8187,
      "step": 13100
    },
    {
      "epoch": 3.0308306010928963,
      "grad_norm": 0.55078125,
      "learning_rate": 0.0002893552142700989,
      "loss": 1.8035,
      "step": 13150
    },
    {
      "epoch": 3.031377049180328,
      "grad_norm": 0.55078125,
      "learning_rate": 0.0002892571555426843,
      "loss": 1.8248,
      "step": 13200
    },
    {
      "epoch": 3.0319234972677593,
      "grad_norm": 0.51953125,
      "learning_rate": 0.00028915866400081795,
      "loss": 1.8066,
      "step": 13250
    },
    {
      "epoch": 3.0324699453551913,
      "grad_norm": 0.50390625,
      "learning_rate": 0.00028905973995061373,
      "loss": 1.8087,
      "step": 13300
    },
    {
      "epoch": 3.033016393442623,
      "grad_norm": 0.53515625,
      "learning_rate": 0.00028896038369953,
      "loss": 1.8208,
      "step": 13350
    },
    {
      "epoch": 3.033562841530055,
      "grad_norm": 0.53125,
      "learning_rate": 0.00028886059555636816,
      "loss": 1.8282,
      "step": 13400
    },
    {
      "epoch": 3.0341092896174864,
      "grad_norm": 0.5546875,
      "learning_rate": 0.00028876037583127213,
      "loss": 1.7288,
      "step": 13450
    },
    {
      "epoch": 3.034655737704918,
      "grad_norm": 0.58203125,
      "learning_rate": 0.000288659724835727,
      "loss": 1.841,
      "step": 13500
    },
    {
      "epoch": 3.03520218579235,
      "grad_norm": 0.51171875,
      "learning_rate": 0.00028855864288255856,
      "loss": 1.8044,
      "step": 13550
    },
    {
      "epoch": 3.0357486338797814,
      "grad_norm": 0.52734375,
      "learning_rate": 0.00028845713028593183,
      "loss": 1.8101,
      "step": 13600
    },
    {
      "epoch": 3.036295081967213,
      "grad_norm": 0.53515625,
      "learning_rate": 0.00028835518736135013,
      "loss": 1.8193,
      "step": 13650
    },
    {
      "epoch": 3.036841530054645,
      "grad_norm": 0.5625,
      "learning_rate": 0.0002882528144256546,
      "loss": 1.8219,
      "step": 13700
    },
    {
      "epoch": 3.0373879781420765,
      "grad_norm": 0.53125,
      "learning_rate": 0.00028815001179702265,
      "loss": 1.8044,
      "step": 13750
    },
    {
      "epoch": 4.000306010928962,
      "grad_norm": 0.546875,
      "learning_rate": 0.0002880467797949671,
      "loss": 1.8068,
      "step": 13800
    },
    {
      "epoch": 4.000852459016394,
      "grad_norm": 0.57421875,
      "learning_rate": 0.00028794311874033563,
      "loss": 1.7919,
      "step": 13850
    },
    {
      "epoch": 4.001398907103825,
      "grad_norm": 0.53515625,
      "learning_rate": 0.00028783902895530893,
      "loss": 1.7501,
      "step": 13900
    },
    {
      "epoch": 4.001945355191257,
      "grad_norm": 0.6171875,
      "learning_rate": 0.00028773451076340064,
      "loss": 1.7494,
      "step": 13950
    },
    {
      "epoch": 4.002491803278689,
      "grad_norm": 0.546875,
      "learning_rate": 0.00028762956448945563,
      "loss": 1.6976,
      "step": 14000
    },
    {
      "epoch": 4.00303825136612,
      "grad_norm": 0.53125,
      "learning_rate": 0.00028752419045964935,
      "loss": 1.7673,
      "step": 14050
    },
    {
      "epoch": 4.003584699453552,
      "grad_norm": 0.6328125,
      "learning_rate": 0.0002874183890014867,
      "loss": 1.7385,
      "step": 14100
    },
    {
      "epoch": 4.004131147540984,
      "grad_norm": 0.5703125,
      "learning_rate": 0.0002873121604438011,
      "loss": 1.7526,
      "step": 14150
    },
    {
      "epoch": 4.004677595628415,
      "grad_norm": 0.53125,
      "learning_rate": 0.0002872055051167533,
      "loss": 1.7323,
      "step": 14200
    },
    {
      "epoch": 4.005224043715847,
      "grad_norm": 0.578125,
      "learning_rate": 0.0002870984233518306,
      "loss": 1.7666,
      "step": 14250
    },
    {
      "epoch": 4.005770491803279,
      "grad_norm": 0.5546875,
      "learning_rate": 0.0002869909154818455,
      "loss": 1.7382,
      "step": 14300
    },
    {
      "epoch": 4.00631693989071,
      "grad_norm": 0.55859375,
      "learning_rate": 0.00028688298184093497,
      "loss": 1.7366,
      "step": 14350
    },
    {
      "epoch": 4.006863387978142,
      "grad_norm": 0.5703125,
      "learning_rate": 0.0002867746227645593,
      "loss": 1.7206,
      "step": 14400
    },
    {
      "epoch": 4.007409836065574,
      "grad_norm": 0.5703125,
      "learning_rate": 0.000286665838589501,
      "loss": 1.7655,
      "step": 14450
    },
    {
      "epoch": 4.007956284153005,
      "grad_norm": 0.55859375,
      "learning_rate": 0.0002865566296538637,
      "loss": 1.7599,
      "step": 14500
    },
    {
      "epoch": 4.008502732240437,
      "grad_norm": 0.62890625,
      "learning_rate": 0.00028644699629707136,
      "loss": 1.7298,
      "step": 14550
    },
    {
      "epoch": 4.009049180327869,
      "grad_norm": 0.671875,
      "learning_rate": 0.00028633693885986696,
      "loss": 1.7392,
      "step": 14600
    },
    {
      "epoch": 4.0095956284153,
      "grad_norm": 0.5703125,
      "learning_rate": 0.0002862264576843116,
      "loss": 1.7556,
      "step": 14650
    },
    {
      "epoch": 4.010142076502732,
      "grad_norm": 0.57421875,
      "learning_rate": 0.0002861155531137833,
      "loss": 1.7677,
      "step": 14700
    },
    {
      "epoch": 4.010688524590164,
      "grad_norm": 0.62890625,
      "learning_rate": 0.00028600422549297604,
      "loss": 1.7283,
      "step": 14750
    },
    {
      "epoch": 4.011234972677595,
      "grad_norm": 0.51953125,
      "learning_rate": 0.00028589247516789856,
      "loss": 1.7389,
      "step": 14800
    },
    {
      "epoch": 4.011781420765027,
      "grad_norm": 0.5703125,
      "learning_rate": 0.0002857803024858735,
      "loss": 1.7769,
      "step": 14850
    },
    {
      "epoch": 4.012327868852459,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00028566770779553613,
      "loss": 1.7306,
      "step": 14900
    },
    {
      "epoch": 4.01287431693989,
      "grad_norm": 0.54296875,
      "learning_rate": 0.00028555469144683337,
      "loss": 1.7341,
      "step": 14950
    },
    {
      "epoch": 4.013420765027322,
      "grad_norm": 0.53515625,
      "learning_rate": 0.00028544125379102264,
      "loss": 1.7364,
      "step": 15000
    },
    {
      "epoch": 4.013967213114754,
      "grad_norm": 0.52734375,
      "learning_rate": 0.0002853273951806708,
      "loss": 1.7203,
      "step": 15050
    },
    {
      "epoch": 4.0145136612021854,
      "grad_norm": 0.57421875,
      "learning_rate": 0.00028521311596965297,
      "loss": 1.7735,
      "step": 15100
    },
    {
      "epoch": 4.015060109289617,
      "grad_norm": 0.5546875,
      "learning_rate": 0.00028509841651315156,
      "loss": 1.7457,
      "step": 15150
    },
    {
      "epoch": 4.015606557377049,
      "grad_norm": 0.53125,
      "learning_rate": 0.0002849832971676553,
      "loss": 1.7317,
      "step": 15200
    },
    {
      "epoch": 4.0161530054644805,
      "grad_norm": 0.5625,
      "learning_rate": 0.0002848677582909576,
      "loss": 1.7168,
      "step": 15250
    },
    {
      "epoch": 4.0166994535519125,
      "grad_norm": 0.5625,
      "learning_rate": 0.000284751800242156,
      "loss": 1.714,
      "step": 15300
    },
    {
      "epoch": 4.0172459016393445,
      "grad_norm": 0.53125,
      "learning_rate": 0.0002846354233816508,
      "loss": 1.7342,
      "step": 15350
    },
    {
      "epoch": 4.017792349726776,
      "grad_norm": 0.58203125,
      "learning_rate": 0.00028451862807114396,
      "loss": 1.7201,
      "step": 15400
    },
    {
      "epoch": 4.0183387978142076,
      "grad_norm": 0.55078125,
      "learning_rate": 0.00028440141467363803,
      "loss": 1.6881,
      "step": 15450
    },
    {
      "epoch": 4.0188852459016395,
      "grad_norm": 0.54296875,
      "learning_rate": 0.00028428378355343495,
      "loss": 1.7184,
      "step": 15500
    },
    {
      "epoch": 4.019431693989071,
      "grad_norm": 0.54296875,
      "learning_rate": 0.00028416573507613485,
      "loss": 1.7055,
      "step": 15550
    },
    {
      "epoch": 4.019978142076503,
      "grad_norm": 0.58203125,
      "learning_rate": 0.0002840472696086353,
      "loss": 1.72,
      "step": 15600
    },
    {
      "epoch": 4.020524590163935,
      "grad_norm": 0.53125,
      "learning_rate": 0.0002839283875191295,
      "loss": 1.7308,
      "step": 15650
    },
    {
      "epoch": 4.021071038251366,
      "grad_norm": 0.5625,
      "learning_rate": 0.0002838090891771059,
      "loss": 1.702,
      "step": 15700
    },
    {
      "epoch": 4.021617486338798,
      "grad_norm": 0.59375,
      "learning_rate": 0.0002836893749533465,
      "loss": 1.7191,
      "step": 15750
    },
    {
      "epoch": 4.02216393442623,
      "grad_norm": 0.54296875,
      "learning_rate": 0.0002835692452199257,
      "loss": 1.7215,
      "step": 15800
    },
    {
      "epoch": 4.022710382513662,
      "grad_norm": 0.54296875,
      "learning_rate": 0.00028344870035020963,
      "loss": 1.6984,
      "step": 15850
    },
    {
      "epoch": 4.023256830601093,
      "grad_norm": 0.53125,
      "learning_rate": 0.0002833277407188545,
      "loss": 1.6847,
      "step": 15900
    },
    {
      "epoch": 4.023803278688525,
      "grad_norm": 0.53125,
      "learning_rate": 0.00028320636670180557,
      "loss": 1.6739,
      "step": 15950
    },
    {
      "epoch": 4.024349726775957,
      "grad_norm": 0.52734375,
      "learning_rate": 0.0002830845786762962,
      "loss": 1.7125,
      "step": 16000
    },
    {
      "epoch": 4.024896174863388,
      "grad_norm": 0.6796875,
      "learning_rate": 0.0002829623770208463,
      "loss": 1.6678,
      "step": 16050
    },
    {
      "epoch": 4.02544262295082,
      "grad_norm": 0.55859375,
      "learning_rate": 0.00028283976211526137,
      "loss": 1.7396,
      "step": 16100
    },
    {
      "epoch": 4.025989071038252,
      "grad_norm": 0.578125,
      "learning_rate": 0.0002827167343406315,
      "loss": 1.752,
      "step": 16150
    },
    {
      "epoch": 4.026535519125683,
      "grad_norm": 0.56640625,
      "learning_rate": 0.0002825932940793298,
      "loss": 1.6994,
      "step": 16200
    },
    {
      "epoch": 4.027081967213115,
      "grad_norm": 0.6015625,
      "learning_rate": 0.00028246944171501145,
      "loss": 1.7189,
      "step": 16250
    },
    {
      "epoch": 4.027628415300547,
      "grad_norm": 0.5625,
      "learning_rate": 0.00028234517763261243,
      "loss": 1.7158,
      "step": 16300
    },
    {
      "epoch": 4.028174863387978,
      "grad_norm": 0.5625,
      "learning_rate": 0.00028222050221834847,
      "loss": 1.6885,
      "step": 16350
    },
    {
      "epoch": 4.02872131147541,
      "grad_norm": 0.5078125,
      "learning_rate": 0.0002820954158597134,
      "loss": 1.6865,
      "step": 16400
    },
    {
      "epoch": 4.029267759562842,
      "grad_norm": 0.52734375,
      "learning_rate": 0.0002819699189454788,
      "loss": 1.6946,
      "step": 16450
    },
    {
      "epoch": 4.029814207650273,
      "grad_norm": 0.5703125,
      "learning_rate": 0.0002818440118656918,
      "loss": 1.7194,
      "step": 16500
    },
    {
      "epoch": 4.030360655737705,
      "grad_norm": 0.52734375,
      "learning_rate": 0.0002817176950116746,
      "loss": 1.7022,
      "step": 16550
    },
    {
      "epoch": 4.030907103825137,
      "grad_norm": 0.61328125,
      "learning_rate": 0.00028159096877602275,
      "loss": 1.7159,
      "step": 16600
    },
    {
      "epoch": 4.031453551912568,
      "grad_norm": 0.5625,
      "learning_rate": 0.00028146383355260446,
      "loss": 1.7063,
      "step": 16650
    },
    {
      "epoch": 4.032,
      "grad_norm": 0.58984375,
      "learning_rate": 0.00028133628973655894,
      "loss": 1.7064,
      "step": 16700
    },
    {
      "epoch": 4.032546448087432,
      "grad_norm": 0.55859375,
      "learning_rate": 0.00028120833772429517,
      "loss": 1.6992,
      "step": 16750
    },
    {
      "epoch": 4.033092896174863,
      "grad_norm": 0.5546875,
      "learning_rate": 0.0002810799779134911,
      "loss": 1.7132,
      "step": 16800
    },
    {
      "epoch": 4.033639344262295,
      "grad_norm": 0.5390625,
      "learning_rate": 0.0002809512107030919,
      "loss": 1.7125,
      "step": 16850
    },
    {
      "epoch": 4.034185792349727,
      "grad_norm": 0.57421875,
      "learning_rate": 0.0002808220364933091,
      "loss": 1.6373,
      "step": 16900
    },
    {
      "epoch": 4.034732240437158,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00028069245568561904,
      "loss": 1.7379,
      "step": 16950
    },
    {
      "epoch": 4.03527868852459,
      "grad_norm": 0.6015625,
      "learning_rate": 0.00028056246868276186,
      "loss": 1.699,
      "step": 17000
    },
    {
      "epoch": 4.035825136612022,
      "grad_norm": 0.5546875,
      "learning_rate": 0.0002804320758887403,
      "loss": 1.6939,
      "step": 17050
    },
    {
      "epoch": 4.036371584699453,
      "grad_norm": 0.498046875,
      "learning_rate": 0.000280301277708818,
      "loss": 1.7146,
      "step": 17100
    },
    {
      "epoch": 4.036918032786885,
      "grad_norm": 0.5859375,
      "learning_rate": 0.00028017007454951884,
      "loss": 1.7363,
      "step": 17150
    },
    {
      "epoch": 4.037464480874317,
      "grad_norm": 0.55078125,
      "learning_rate": 0.00028003846681862524,
      "loss": 1.6965,
      "step": 17200
    },
    {
      "epoch": 5.000382513661203,
      "grad_norm": 0.5703125,
      "learning_rate": 0.00027990645492517697,
      "loss": 1.6887,
      "step": 17250
    },
    {
      "epoch": 5.000928961748634,
      "grad_norm": 0.62109375,
      "learning_rate": 0.0002797740392794702,
      "loss": 1.6829,
      "step": 17300
    },
    {
      "epoch": 5.001475409836066,
      "grad_norm": 0.58984375,
      "learning_rate": 0.0002796412202930557,
      "loss": 1.6235,
      "step": 17350
    },
    {
      "epoch": 5.002021857923498,
      "grad_norm": 0.59375,
      "learning_rate": 0.00027950799837873794,
      "loss": 1.634,
      "step": 17400
    },
    {
      "epoch": 5.002568306010929,
      "grad_norm": 0.55859375,
      "learning_rate": 0.0002793743739505738,
      "loss": 1.5863,
      "step": 17450
    },
    {
      "epoch": 5.003114754098361,
      "grad_norm": 0.5546875,
      "learning_rate": 0.0002792403474238709,
      "loss": 1.6513,
      "step": 17500
    },
    {
      "epoch": 5.003661202185793,
      "grad_norm": 0.59375,
      "learning_rate": 0.000279105919215187,
      "loss": 1.6388,
      "step": 17550
    },
    {
      "epoch": 5.004207650273224,
      "grad_norm": 0.59375,
      "learning_rate": 0.00027897108974232797,
      "loss": 1.6302,
      "step": 17600
    },
    {
      "epoch": 5.004754098360656,
      "grad_norm": 1.1875,
      "learning_rate": 0.0002788358594243469,
      "loss": 1.6336,
      "step": 17650
    },
    {
      "epoch": 5.005300546448088,
      "grad_norm": 0.58984375,
      "learning_rate": 0.0002787002286815428,
      "loss": 1.6513,
      "step": 17700
    },
    {
      "epoch": 5.005846994535519,
      "grad_norm": 0.6015625,
      "learning_rate": 0.000278564197935459,
      "loss": 1.6185,
      "step": 17750
    },
    {
      "epoch": 5.006393442622951,
      "grad_norm": 0.58984375,
      "learning_rate": 0.00027842776760888236,
      "loss": 1.6366,
      "step": 17800
    },
    {
      "epoch": 5.006939890710383,
      "grad_norm": 0.578125,
      "learning_rate": 0.00027829093812584143,
      "loss": 1.6162,
      "step": 17850
    },
    {
      "epoch": 5.007486338797814,
      "grad_norm": 0.56640625,
      "learning_rate": 0.0002781537099116054,
      "loss": 1.6665,
      "step": 17900
    },
    {
      "epoch": 5.008032786885246,
      "grad_norm": 0.55859375,
      "learning_rate": 0.00027801608339268275,
      "loss": 1.6505,
      "step": 17950
    },
    {
      "epoch": 5.008579234972678,
      "grad_norm": 0.578125,
      "learning_rate": 0.00027787805899681976,
      "loss": 1.6079,
      "step": 18000
    },
    {
      "epoch": 5.009125683060109,
      "grad_norm": 0.640625,
      "learning_rate": 0.00027773963715299957,
      "loss": 1.6429,
      "step": 18050
    },
    {
      "epoch": 5.009672131147541,
      "grad_norm": 0.640625,
      "learning_rate": 0.00027760081829144044,
      "loss": 1.6671,
      "step": 18100
    },
    {
      "epoch": 5.010218579234973,
      "grad_norm": 0.6484375,
      "learning_rate": 0.0002774616028435946,
      "loss": 1.6619,
      "step": 18150
    },
    {
      "epoch": 5.010765027322404,
      "grad_norm": 0.546875,
      "learning_rate": 0.00027732199124214676,
      "loss": 1.6221,
      "step": 18200
    },
    {
      "epoch": 5.011311475409836,
      "grad_norm": 0.63671875,
      "learning_rate": 0.0002771819839210131,
      "loss": 1.6311,
      "step": 18250
    },
    {
      "epoch": 5.011857923497268,
      "grad_norm": 0.54296875,
      "learning_rate": 0.0002770415813153396,
      "loss": 1.669,
      "step": 18300
    },
    {
      "epoch": 5.012404371584699,
      "grad_norm": 0.609375,
      "learning_rate": 0.00027690078386150084,
      "loss": 1.6177,
      "step": 18350
    },
    {
      "epoch": 5.012950819672131,
      "grad_norm": 0.59375,
      "learning_rate": 0.0002767595919970984,
      "loss": 1.622,
      "step": 18400
    },
    {
      "epoch": 5.013497267759563,
      "grad_norm": 0.609375,
      "learning_rate": 0.00027661800616096,
      "loss": 1.6329,
      "step": 18450
    },
    {
      "epoch": 5.014043715846994,
      "grad_norm": 0.58203125,
      "learning_rate": 0.00027647602679313764,
      "loss": 1.6369,
      "step": 18500
    },
    {
      "epoch": 5.014590163934426,
      "grad_norm": 0.55078125,
      "learning_rate": 0.0002763336543349065,
      "loss": 1.6624,
      "step": 18550
    },
    {
      "epoch": 5.015136612021858,
      "grad_norm": 0.5859375,
      "learning_rate": 0.0002761908892287633,
      "loss": 1.6422,
      "step": 18600
    },
    {
      "epoch": 5.015683060109289,
      "grad_norm": 0.5703125,
      "learning_rate": 0.0002760477319184255,
      "loss": 1.6101,
      "step": 18650
    },
    {
      "epoch": 5.016229508196721,
      "grad_norm": 0.57421875,
      "learning_rate": 0.0002759041828488292,
      "loss": 1.6355,
      "step": 18700
    },
    {
      "epoch": 5.016775956284153,
      "grad_norm": 0.58984375,
      "learning_rate": 0.0002757602424661283,
      "loss": 1.6114,
      "step": 18750
    },
    {
      "epoch": 5.017322404371584,
      "grad_norm": 0.59375,
      "learning_rate": 0.00027561591121769277,
      "loss": 1.6274,
      "step": 18800
    },
    {
      "epoch": 5.017868852459016,
      "grad_norm": 0.57421875,
      "learning_rate": 0.00027547118955210747,
      "loss": 1.6139,
      "step": 18850
    },
    {
      "epoch": 5.018415300546448,
      "grad_norm": 0.55859375,
      "learning_rate": 0.0002753260779191706,
      "loss": 1.6027,
      "step": 18900
    },
    {
      "epoch": 5.018961748633879,
      "grad_norm": 0.62109375,
      "learning_rate": 0.0002751805767698927,
      "loss": 1.6036,
      "step": 18950
    },
    {
      "epoch": 5.019508196721311,
      "grad_norm": 0.59375,
      "learning_rate": 0.0002750346865564944,
      "loss": 1.6073,
      "step": 19000
    },
    {
      "epoch": 5.020054644808743,
      "grad_norm": 0.54296875,
      "learning_rate": 0.0002748884077324061,
      "loss": 1.6285,
      "step": 19050
    },
    {
      "epoch": 5.020601092896175,
      "grad_norm": 0.59765625,
      "learning_rate": 0.0002747417407522656,
      "loss": 1.6254,
      "step": 19100
    },
    {
      "epoch": 5.021147540983606,
      "grad_norm": 0.5546875,
      "learning_rate": 0.00027459468607191723,
      "loss": 1.6058,
      "step": 19150
    },
    {
      "epoch": 5.021693989071038,
      "grad_norm": 0.60546875,
      "learning_rate": 0.00027444724414841046,
      "loss": 1.623,
      "step": 19200
    },
    {
      "epoch": 5.02224043715847,
      "grad_norm": 0.59765625,
      "learning_rate": 0.00027429941543999814,
      "loss": 1.623,
      "step": 19250
    },
    {
      "epoch": 5.0227868852459014,
      "grad_norm": 0.5390625,
      "learning_rate": 0.0002741512004061353,
      "loss": 1.601,
      "step": 19300
    },
    {
      "epoch": 5.023333333333333,
      "grad_norm": 0.56640625,
      "learning_rate": 0.0002740025995074777,
      "loss": 1.5935,
      "step": 19350
    },
    {
      "epoch": 5.023879781420765,
      "grad_norm": 0.578125,
      "learning_rate": 0.00027385361320588034,
      "loss": 1.5831,
      "step": 19400
    },
    {
      "epoch": 5.0244262295081965,
      "grad_norm": 0.59765625,
      "learning_rate": 0.0002737042419643961,
      "loss": 1.5974,
      "step": 19450
    },
    {
      "epoch": 5.0249726775956285,
      "grad_norm": 0.61328125,
      "learning_rate": 0.0002735544862472742,
      "loss": 1.5629,
      "step": 19500
    },
    {
      "epoch": 5.0255191256830605,
      "grad_norm": 0.640625,
      "learning_rate": 0.00027340434651995887,
      "loss": 1.6589,
      "step": 19550
    },
    {
      "epoch": 5.026065573770492,
      "grad_norm": 0.59375,
      "learning_rate": 0.0002732538232490879,
      "loss": 1.6576,
      "step": 19600
    },
    {
      "epoch": 5.0266120218579236,
      "grad_norm": 0.5859375,
      "learning_rate": 0.0002731029169024911,
      "loss": 1.6112,
      "step": 19650
    },
    {
      "epoch": 5.0271584699453555,
      "grad_norm": 0.609375,
      "learning_rate": 0.00027295162794918875,
      "loss": 1.6232,
      "step": 19700
    },
    {
      "epoch": 5.027704918032787,
      "grad_norm": 0.6328125,
      "learning_rate": 0.00027279995685939055,
      "loss": 1.6242,
      "step": 19750
    },
    {
      "epoch": 5.028251366120219,
      "grad_norm": 0.546875,
      "learning_rate": 0.00027264790410449363,
      "loss": 1.6029,
      "step": 19800
    },
    {
      "epoch": 5.028797814207651,
      "grad_norm": 0.6875,
      "learning_rate": 0.0002724954701570816,
      "loss": 1.5935,
      "step": 19850
    },
    {
      "epoch": 5.029344262295082,
      "grad_norm": 0.6640625,
      "learning_rate": 0.00027234265549092257,
      "loss": 1.6099,
      "step": 19900
    },
    {
      "epoch": 5.029890710382514,
      "grad_norm": 0.6015625,
      "learning_rate": 0.00027218946058096805,
      "loss": 1.6366,
      "step": 19950
    },
    {
      "epoch": 5.030437158469946,
      "grad_norm": 0.57421875,
      "learning_rate": 0.0002720358859033514,
      "loss": 1.6249,
      "step": 20000
    },
    {
      "epoch": 5.030983606557377,
      "grad_norm": 0.59375,
      "learning_rate": 0.00027188193193538625,
      "loss": 1.6114,
      "step": 20050
    },
    {
      "epoch": 5.031530054644809,
      "grad_norm": 0.671875,
      "learning_rate": 0.00027172759915556504,
      "loss": 1.6081,
      "step": 20100
    },
    {
      "epoch": 5.032076502732241,
      "grad_norm": 0.62890625,
      "learning_rate": 0.0002715728880435577,
      "loss": 1.627,
      "step": 20150
    },
    {
      "epoch": 5.032622950819672,
      "grad_norm": 0.62890625,
      "learning_rate": 0.00027141779908020986,
      "loss": 1.5912,
      "step": 20200
    },
    {
      "epoch": 5.033169398907104,
      "grad_norm": 0.65234375,
      "learning_rate": 0.00027126233274754163,
      "loss": 1.6476,
      "step": 20250
    },
    {
      "epoch": 5.033715846994536,
      "grad_norm": 0.59375,
      "learning_rate": 0.00027110648952874595,
      "loss": 1.6051,
      "step": 20300
    },
    {
      "epoch": 5.034262295081967,
      "grad_norm": 0.65234375,
      "learning_rate": 0.0002709502699081871,
      "loss": 1.5705,
      "step": 20350
    },
    {
      "epoch": 5.034808743169399,
      "grad_norm": 0.578125,
      "learning_rate": 0.00027079367437139935,
      "loss": 1.6533,
      "step": 20400
    },
    {
      "epoch": 5.035355191256831,
      "grad_norm": 0.59765625,
      "learning_rate": 0.00027063670340508514,
      "loss": 1.6099,
      "step": 20450
    },
    {
      "epoch": 5.035901639344262,
      "grad_norm": 0.5703125,
      "learning_rate": 0.00027047935749711395,
      "loss": 1.6018,
      "step": 20500
    },
    {
      "epoch": 5.036448087431694,
      "grad_norm": 0.59375,
      "learning_rate": 0.0002703216371365204,
      "loss": 1.637,
      "step": 20550
    },
    {
      "epoch": 5.036994535519126,
      "grad_norm": 0.56640625,
      "learning_rate": 0.00027016354281350315,
      "loss": 1.6394,
      "step": 20600
    },
    {
      "epoch": 5.037540983606557,
      "grad_norm": 0.6015625,
      "learning_rate": 0.00027000507501942283,
      "loss": 1.6253,
      "step": 20650
    },
    {
      "epoch": 6.000459016393442,
      "grad_norm": 0.57421875,
      "learning_rate": 0.0002698462342468011,
      "loss": 1.5897,
      "step": 20700
    },
    {
      "epoch": 6.001005464480874,
      "grad_norm": 0.56640625,
      "learning_rate": 0.0002696870209893187,
      "loss": 1.569,
      "step": 20750
    },
    {
      "epoch": 6.001551912568306,
      "grad_norm": 0.59765625,
      "learning_rate": 0.00026952743574181414,
      "loss": 1.5482,
      "step": 20800
    },
    {
      "epoch": 6.002098360655737,
      "grad_norm": 0.59765625,
      "learning_rate": 0.00026936747900028205,
      "loss": 1.5458,
      "step": 20850
    },
    {
      "epoch": 6.002644808743169,
      "grad_norm": 0.734375,
      "learning_rate": 0.00026920715126187167,
      "loss": 1.4787,
      "step": 20900
    },
    {
      "epoch": 6.003191256830601,
      "grad_norm": 0.7109375,
      "learning_rate": 0.0002690464530248853,
      "loss": 1.5565,
      "step": 20950
    },
    {
      "epoch": 6.0037377049180325,
      "grad_norm": 0.64453125,
      "learning_rate": 0.00026888538478877675,
      "loss": 1.5588,
      "step": 21000
    },
    {
      "epoch": 6.0042841530054645,
      "grad_norm": 0.6953125,
      "learning_rate": 0.0002687239470541498,
      "loss": 1.5347,
      "step": 21050
    },
    {
      "epoch": 6.0048306010928965,
      "grad_norm": 0.58984375,
      "learning_rate": 0.00026856214032275675,
      "loss": 1.5341,
      "step": 21100
    },
    {
      "epoch": 6.0053770491803276,
      "grad_norm": 0.73046875,
      "learning_rate": 0.00026839996509749655,
      "loss": 1.5441,
      "step": 21150
    },
    {
      "epoch": 6.0059234972677595,
      "grad_norm": 0.58984375,
      "learning_rate": 0.00026823742188241366,
      "loss": 1.5405,
      "step": 21200
    },
    {
      "epoch": 6.0064699453551915,
      "grad_norm": 0.64453125,
      "learning_rate": 0.000268074511182696,
      "loss": 1.5327,
      "step": 21250
    },
    {
      "epoch": 6.007016393442623,
      "grad_norm": 0.61328125,
      "learning_rate": 0.00026791123350467384,
      "loss": 1.5338,
      "step": 21300
    },
    {
      "epoch": 6.007562841530055,
      "grad_norm": 0.58984375,
      "learning_rate": 0.000267747589355818,
      "loss": 1.5663,
      "step": 21350
    },
    {
      "epoch": 6.008109289617487,
      "grad_norm": 0.67578125,
      "learning_rate": 0.0002675835792447382,
      "loss": 1.5519,
      "step": 21400
    },
    {
      "epoch": 6.008655737704918,
      "grad_norm": 0.66015625,
      "learning_rate": 0.0002674192036811818,
      "loss": 1.5256,
      "step": 21450
    },
    {
      "epoch": 6.00920218579235,
      "grad_norm": 0.578125,
      "learning_rate": 0.0002672544631760317,
      "loss": 1.5428,
      "step": 21500
    },
    {
      "epoch": 6.009748633879782,
      "grad_norm": 0.75,
      "learning_rate": 0.00026708935824130514,
      "loss": 1.5806,
      "step": 21550
    },
    {
      "epoch": 6.010295081967213,
      "grad_norm": 0.57421875,
      "learning_rate": 0.00026692388939015226,
      "loss": 1.5628,
      "step": 21600
    },
    {
      "epoch": 6.010841530054645,
      "grad_norm": 0.61328125,
      "learning_rate": 0.00026675805713685387,
      "loss": 1.5275,
      "step": 21650
    },
    {
      "epoch": 6.011387978142077,
      "grad_norm": 0.61328125,
      "learning_rate": 0.0002665918619968206,
      "loss": 1.5547,
      "step": 21700
    },
    {
      "epoch": 6.011934426229508,
      "grad_norm": 0.57421875,
      "learning_rate": 0.0002664253044865907,
      "loss": 1.575,
      "step": 21750
    },
    {
      "epoch": 6.01248087431694,
      "grad_norm": 0.59765625,
      "learning_rate": 0.0002662583851238287,
      "loss": 1.5386,
      "step": 21800
    },
    {
      "epoch": 6.013027322404372,
      "grad_norm": 0.6484375,
      "learning_rate": 0.000266091104427324,
      "loss": 1.5107,
      "step": 21850
    },
    {
      "epoch": 6.013573770491803,
      "grad_norm": 0.625,
      "learning_rate": 0.00026592346291698864,
      "loss": 1.5516,
      "step": 21900
    },
    {
      "epoch": 6.014120218579235,
      "grad_norm": 0.5703125,
      "learning_rate": 0.00026575546111385647,
      "loss": 1.5431,
      "step": 21950
    },
    {
      "epoch": 6.014666666666667,
      "grad_norm": 0.6015625,
      "learning_rate": 0.00026558709954008095,
      "loss": 1.566,
      "step": 22000
    },
    {
      "epoch": 6.015213114754098,
      "grad_norm": 0.62890625,
      "learning_rate": 0.00026541837871893367,
      "loss": 1.5595,
      "step": 22050
    },
    {
      "epoch": 6.01575956284153,
      "grad_norm": 0.5625,
      "learning_rate": 0.0002652492991748029,
      "loss": 1.5206,
      "step": 22100
    },
    {
      "epoch": 6.016306010928962,
      "grad_norm": 0.640625,
      "learning_rate": 0.00026507986143319164,
      "loss": 1.5374,
      "step": 22150
    },
    {
      "epoch": 6.016852459016394,
      "grad_norm": 0.703125,
      "learning_rate": 0.0002649100660207164,
      "loss": 1.522,
      "step": 22200
    },
    {
      "epoch": 6.017398907103825,
      "grad_norm": 0.65625,
      "learning_rate": 0.0002647399134651053,
      "loss": 1.5532,
      "step": 22250
    },
    {
      "epoch": 6.017945355191257,
      "grad_norm": 0.63671875,
      "learning_rate": 0.0002645694042951963,
      "loss": 1.5274,
      "step": 22300
    },
    {
      "epoch": 6.018491803278689,
      "grad_norm": 0.65625,
      "learning_rate": 0.00026439853904093586,
      "loss": 1.517,
      "step": 22350
    },
    {
      "epoch": 6.01903825136612,
      "grad_norm": 0.6875,
      "learning_rate": 0.00026422731823337717,
      "loss": 1.5197,
      "step": 22400
    },
    {
      "epoch": 6.019584699453552,
      "grad_norm": 0.59765625,
      "learning_rate": 0.0002640557424046784,
      "loss": 1.5081,
      "step": 22450
    },
    {
      "epoch": 6.020131147540984,
      "grad_norm": 0.65234375,
      "learning_rate": 0.0002638838120881012,
      "loss": 1.5526,
      "step": 22500
    },
    {
      "epoch": 6.020677595628415,
      "grad_norm": 0.62109375,
      "learning_rate": 0.000263711527818009,
      "loss": 1.534,
      "step": 22550
    },
    {
      "epoch": 6.021224043715847,
      "grad_norm": 0.60546875,
      "learning_rate": 0.0002635388901298652,
      "loss": 1.5156,
      "step": 22600
    },
    {
      "epoch": 6.021770491803279,
      "grad_norm": 0.71875,
      "learning_rate": 0.0002633658995602318,
      "loss": 1.5402,
      "step": 22650
    },
    {
      "epoch": 6.02231693989071,
      "grad_norm": 0.68359375,
      "learning_rate": 0.0002631925566467674,
      "loss": 1.5367,
      "step": 22700
    },
    {
      "epoch": 6.022863387978142,
      "grad_norm": 0.62890625,
      "learning_rate": 0.00026301886192822585,
      "loss": 1.5126,
      "step": 22750
    },
    {
      "epoch": 6.023409836065574,
      "grad_norm": 0.625,
      "learning_rate": 0.00026284481594445434,
      "loss": 1.5097,
      "step": 22800
    },
    {
      "epoch": 6.023956284153005,
      "grad_norm": 0.60546875,
      "learning_rate": 0.00026267041923639175,
      "loss": 1.5058,
      "step": 22850
    },
    {
      "epoch": 6.024502732240437,
      "grad_norm": 0.703125,
      "learning_rate": 0.00026249567234606707,
      "loss": 1.5004,
      "step": 22900
    },
    {
      "epoch": 6.025049180327869,
      "grad_norm": 0.60546875,
      "learning_rate": 0.00026232057581659777,
      "loss": 1.4884,
      "step": 22950
    },
    {
      "epoch": 6.0255956284153,
      "grad_norm": 0.65234375,
      "learning_rate": 0.0002621451301921878,
      "loss": 1.5884,
      "step": 23000
    },
    {
      "epoch": 6.026142076502732,
      "grad_norm": 0.6484375,
      "learning_rate": 0.00026196933601812616,
      "loss": 1.565,
      "step": 23050
    },
    {
      "epoch": 6.026688524590164,
      "grad_norm": 0.69140625,
      "learning_rate": 0.00026179319384078535,
      "loss": 1.5399,
      "step": 23100
    },
    {
      "epoch": 6.027234972677595,
      "grad_norm": 0.57421875,
      "learning_rate": 0.0002616167042076192,
      "loss": 1.5319,
      "step": 23150
    },
    {
      "epoch": 6.027781420765027,
      "grad_norm": 0.62890625,
      "learning_rate": 0.0002614398676671616,
      "loss": 1.5379,
      "step": 23200
    },
    {
      "epoch": 6.028327868852459,
      "grad_norm": 0.69921875,
      "learning_rate": 0.0002612626847690247,
      "loss": 1.5344,
      "step": 23250
    },
    {
      "epoch": 6.02887431693989,
      "grad_norm": 0.59765625,
      "learning_rate": 0.0002610851560638968,
      "loss": 1.5054,
      "step": 23300
    },
    {
      "epoch": 6.029420765027322,
      "grad_norm": 0.6484375,
      "learning_rate": 0.0002609072821035415,
      "loss": 1.5421,
      "step": 23350
    },
    {
      "epoch": 6.029967213114754,
      "grad_norm": 0.69921875,
      "learning_rate": 0.00026072906344079484,
      "loss": 1.5625,
      "step": 23400
    },
    {
      "epoch": 6.0305136612021855,
      "grad_norm": 0.59765625,
      "learning_rate": 0.0002605505006295648,
      "loss": 1.5476,
      "step": 23450
    },
    {
      "epoch": 6.031060109289617,
      "grad_norm": 0.609375,
      "learning_rate": 0.00026037159422482865,
      "loss": 1.537,
      "step": 23500
    },
    {
      "epoch": 6.031606557377049,
      "grad_norm": 0.60546875,
      "learning_rate": 0.00026019234478263155,
      "loss": 1.5204,
      "step": 23550
    },
    {
      "epoch": 6.0321530054644805,
      "grad_norm": 0.69921875,
      "learning_rate": 0.000260012752860085,
      "loss": 1.5479,
      "step": 23600
    },
    {
      "epoch": 6.0326994535519125,
      "grad_norm": 0.6171875,
      "learning_rate": 0.00025983281901536474,
      "loss": 1.5304,
      "step": 23650
    },
    {
      "epoch": 6.0332459016393445,
      "grad_norm": 0.6171875,
      "learning_rate": 0.00025965254380770945,
      "loss": 1.5738,
      "step": 23700
    },
    {
      "epoch": 6.033792349726776,
      "grad_norm": 0.62109375,
      "learning_rate": 0.0002594719277974185,
      "loss": 1.5168,
      "step": 23750
    },
    {
      "epoch": 6.034338797814208,
      "grad_norm": 0.6328125,
      "learning_rate": 0.0002592909715458506,
      "loss": 1.4984,
      "step": 23800
    },
    {
      "epoch": 6.0348852459016395,
      "grad_norm": 0.57421875,
      "learning_rate": 0.0002591096756154221,
      "loss": 1.5721,
      "step": 23850
    },
    {
      "epoch": 6.035431693989071,
      "grad_norm": 0.6015625,
      "learning_rate": 0.0002589280405696048,
      "loss": 1.5369,
      "step": 23900
    },
    {
      "epoch": 6.035978142076503,
      "grad_norm": 0.61328125,
      "learning_rate": 0.00025874606697292473,
      "loss": 1.5236,
      "step": 23950
    },
    {
      "epoch": 6.036524590163935,
      "grad_norm": 0.6953125,
      "learning_rate": 0.00025856375539095986,
      "loss": 1.577,
      "step": 24000
    }
  ],
  "logging_steps": 50,
  "max_steps": 91500,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9223372036854775807,
  "save_steps": 2000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.2834709186604433e+19,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}