diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,31192 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 4.997894145725116,
+  "eval_steps": 500,
+  "global_step": 4450,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0011231222799382282,
+      "grad_norm": 6.350573723690268,
+      "learning_rate": 4.494382022471911e-08,
+      "loss": 0.7656,
+      "step": 1
+    },
+    {
+      "epoch": 0.0022462445598764565,
+      "grad_norm": 6.396429921824654,
+      "learning_rate": 8.988764044943822e-08,
+      "loss": 0.8046,
+      "step": 2
+    },
+    {
+      "epoch": 0.0033693668398146847,
+      "grad_norm": 6.515479866330877,
+      "learning_rate": 1.348314606741573e-07,
+      "loss": 0.7993,
+      "step": 3
+    },
+    {
+      "epoch": 0.004492489119752913,
+      "grad_norm": 6.264174843252429,
+      "learning_rate": 1.7977528089887644e-07,
+      "loss": 0.8049,
+      "step": 4
+    },
+    {
+      "epoch": 0.005615611399691142,
+      "grad_norm": 6.57185846446011,
+      "learning_rate": 2.247191011235955e-07,
+      "loss": 0.826,
+      "step": 5
+    },
+    {
+      "epoch": 0.0067387336796293695,
+      "grad_norm": 6.517583190885418,
+      "learning_rate": 2.696629213483146e-07,
+      "loss": 0.803,
+      "step": 6
+    },
+    {
+      "epoch": 0.007861855959567598,
+      "grad_norm": 6.364802286536463,
+      "learning_rate": 3.1460674157303374e-07,
+      "loss": 0.7837,
+      "step": 7
+    },
+    {
+      "epoch": 0.008984978239505826,
+      "grad_norm": 6.365458774711909,
+      "learning_rate": 3.5955056179775287e-07,
+      "loss": 0.7702,
+      "step": 8
+    },
+    {
+      "epoch": 0.010108100519444054,
+      "grad_norm": 6.254849998413217,
+      "learning_rate": 4.044943820224719e-07,
+      "loss": 0.7543,
+      "step": 9
+    },
+    {
+      "epoch": 0.011231222799382283,
+      "grad_norm": 6.296540752960891,
+      "learning_rate": 4.49438202247191e-07,
+      "loss": 0.7835,
+      "step": 10
+    },
+    {
+      "epoch": 0.012354345079320511,
+      "grad_norm": 5.708911549789346,
+      "learning_rate": 4.943820224719102e-07,
+      "loss": 0.764,
+      "step": 11
+    },
+    {
+      "epoch": 0.013477467359258739,
+      "grad_norm": 5.900819796376942,
+      "learning_rate": 5.393258426966292e-07,
+      "loss": 0.76,
+      "step": 12
+    },
+    {
+      "epoch": 0.014600589639196967,
+      "grad_norm": 5.807726787883733,
+      "learning_rate": 5.842696629213484e-07,
+      "loss": 0.7686,
+      "step": 13
+    },
+    {
+      "epoch": 0.015723711919135196,
+      "grad_norm": 5.526670510027518,
+      "learning_rate": 6.292134831460675e-07,
+      "loss": 0.7631,
+      "step": 14
+    },
+    {
+      "epoch": 0.016846834199073422,
+      "grad_norm": 4.559877051017468,
+      "learning_rate": 6.741573033707865e-07,
+      "loss": 0.684,
+      "step": 15
+    },
+    {
+      "epoch": 0.017969956479011652,
+      "grad_norm": 4.475977243208917,
+      "learning_rate": 7.191011235955057e-07,
+      "loss": 0.7226,
+      "step": 16
+    },
+    {
+      "epoch": 0.01909307875894988,
+      "grad_norm": 4.295823217269364,
+      "learning_rate": 7.640449438202248e-07,
+      "loss": 0.7131,
+      "step": 17
+    },
+    {
+      "epoch": 0.020216201038888108,
+      "grad_norm": 4.129352981067178,
+      "learning_rate": 8.089887640449438e-07,
+      "loss": 0.6764,
+      "step": 18
+    },
+    {
+      "epoch": 0.021339323318826337,
+      "grad_norm": 4.136041569248625,
+      "learning_rate": 8.53932584269663e-07,
+      "loss": 0.6996,
+      "step": 19
+    },
+    {
+      "epoch": 0.022462445598764567,
+      "grad_norm": 3.127126410742847,
+      "learning_rate": 8.98876404494382e-07,
+      "loss": 0.6585,
+      "step": 20
+    },
+    {
+      "epoch": 0.023585567878702793,
+      "grad_norm": 2.7983535971516136,
+      "learning_rate": 9.438202247191013e-07,
+      "loss": 0.6476,
+      "step": 21
+    },
+    {
+      "epoch": 0.024708690158641022,
+      "grad_norm": 2.8307801771170498,
+      "learning_rate": 9.887640449438204e-07,
+      "loss": 0.6081,
+      "step": 22
+    },
+    {
+      "epoch": 0.025831812438579252,
+      "grad_norm": 2.6693325996833703,
+      "learning_rate": 1.0337078651685394e-06,
+      "loss": 0.6015,
+      "step": 23
+    },
+    {
+      "epoch": 0.026954934718517478,
+      "grad_norm": 2.5249433529837058,
+      "learning_rate": 1.0786516853932585e-06,
+      "loss": 0.5871,
+      "step": 24
+    },
+    {
+      "epoch": 0.028078056998455708,
+      "grad_norm": 2.7833384535564893,
+      "learning_rate": 1.1235955056179777e-06,
+      "loss": 0.644,
+      "step": 25
+    },
+    {
+      "epoch": 0.029201179278393934,
+      "grad_norm": 2.5636212715816846,
+      "learning_rate": 1.1685393258426967e-06,
+      "loss": 0.6212,
+      "step": 26
+    },
+    {
+      "epoch": 0.030324301558332163,
+      "grad_norm": 2.13400734054636,
+      "learning_rate": 1.2134831460674157e-06,
+      "loss": 0.6061,
+      "step": 27
+    },
+    {
+      "epoch": 0.03144742383827039,
+      "grad_norm": 1.8020788979606925,
+      "learning_rate": 1.258426966292135e-06,
+      "loss": 0.5786,
+      "step": 28
+    },
+    {
+      "epoch": 0.03257054611820862,
+      "grad_norm": 1.7251113036312054,
+      "learning_rate": 1.303370786516854e-06,
+      "loss": 0.519,
+      "step": 29
+    },
+    {
+      "epoch": 0.033693668398146845,
+      "grad_norm": 1.7845108367392837,
+      "learning_rate": 1.348314606741573e-06,
+      "loss": 0.5688,
+      "step": 30
+    },
+    {
+      "epoch": 0.034816790678085074,
+      "grad_norm": 1.5519488204108933,
+      "learning_rate": 1.3932584269662923e-06,
+      "loss": 0.5282,
+      "step": 31
+    },
+    {
+      "epoch": 0.035939912958023304,
+      "grad_norm": 1.4564929135509108,
+      "learning_rate": 1.4382022471910115e-06,
+      "loss": 0.5378,
+      "step": 32
+    },
+    {
+      "epoch": 0.037063035237961534,
+      "grad_norm": 1.5083545600301442,
+      "learning_rate": 1.4831460674157305e-06,
+      "loss": 0.5256,
+      "step": 33
+    },
+    {
+      "epoch": 0.03818615751789976,
+      "grad_norm": 1.2973997209272992,
+      "learning_rate": 1.5280898876404495e-06,
+      "loss": 0.5157,
+      "step": 34
+    },
+    {
+      "epoch": 0.03930927979783799,
+      "grad_norm": 1.2442177307671032,
+      "learning_rate": 1.5730337078651686e-06,
+      "loss": 0.5334,
+      "step": 35
+    },
+    {
+      "epoch": 0.040432402077776215,
+      "grad_norm": 1.1214494431535786,
+      "learning_rate": 1.6179775280898876e-06,
+      "loss": 0.5457,
+      "step": 36
+    },
+    {
+      "epoch": 0.041555524357714445,
+      "grad_norm": 1.0380116256523142,
+      "learning_rate": 1.662921348314607e-06,
+      "loss": 0.5252,
+      "step": 37
+    },
+    {
+      "epoch": 0.042678646637652674,
+      "grad_norm": 0.9512314468973523,
+      "learning_rate": 1.707865168539326e-06,
+      "loss": 0.4822,
+      "step": 38
+    },
+    {
+      "epoch": 0.043801768917590904,
+      "grad_norm": 1.1749256602840057,
+      "learning_rate": 1.752808988764045e-06,
+      "loss": 0.4876,
+      "step": 39
+    },
+    {
+      "epoch": 0.04492489119752913,
+      "grad_norm": 1.1996241804899646,
+      "learning_rate": 1.797752808988764e-06,
+      "loss": 0.5222,
+      "step": 40
+    },
+    {
+      "epoch": 0.046048013477467356,
+      "grad_norm": 1.0789479072322812,
+      "learning_rate": 1.8426966292134831e-06,
+      "loss": 0.5101,
+      "step": 41
+    },
+    {
+      "epoch": 0.047171135757405586,
+      "grad_norm": 1.0512996176581257,
+      "learning_rate": 1.8876404494382026e-06,
+      "loss": 0.5055,
+      "step": 42
+    },
+    {
+      "epoch": 0.048294258037343815,
+      "grad_norm": 0.9746980947816224,
+      "learning_rate": 1.9325842696629214e-06,
+      "loss": 0.504,
+      "step": 43
+    },
+    {
+      "epoch": 0.049417380317282045,
+      "grad_norm": 1.0580329624783904,
+      "learning_rate": 1.977528089887641e-06,
+      "loss": 0.5067,
+      "step": 44
+    },
+    {
+      "epoch": 0.050540502597220274,
+      "grad_norm": 0.9449814871215534,
+      "learning_rate": 2.02247191011236e-06,
+      "loss": 0.4902,
+      "step": 45
+    },
+    {
+      "epoch": 0.051663624877158504,
+      "grad_norm": 0.8168500336470494,
+      "learning_rate": 2.067415730337079e-06,
+      "loss": 0.4573,
+      "step": 46
+    },
+    {
+      "epoch": 0.052786747157096726,
+      "grad_norm": 0.8804112383078366,
+      "learning_rate": 2.112359550561798e-06,
+      "loss": 0.506,
+      "step": 47
+    },
+    {
+      "epoch": 0.053909869437034956,
+      "grad_norm": 0.7359559115946035,
+      "learning_rate": 2.157303370786517e-06,
+      "loss": 0.4451,
+      "step": 48
+    },
+    {
+      "epoch": 0.055032991716973186,
+      "grad_norm": 0.902331110525257,
+      "learning_rate": 2.202247191011236e-06,
+      "loss": 0.4927,
+      "step": 49
+    },
+    {
+      "epoch": 0.056156113996911415,
+      "grad_norm": 0.7375774424153347,
+      "learning_rate": 2.2471910112359554e-06,
+      "loss": 0.4777,
+      "step": 50
+    },
+    {
+      "epoch": 0.057279236276849645,
+      "grad_norm": 0.7302558001904716,
+      "learning_rate": 2.2921348314606744e-06,
+      "loss": 0.4696,
+      "step": 51
+    },
+    {
+      "epoch": 0.05840235855678787,
+      "grad_norm": 0.7169095239110965,
+      "learning_rate": 2.3370786516853934e-06,
+      "loss": 0.4555,
+      "step": 52
+    },
+    {
+      "epoch": 0.0595254808367261,
+      "grad_norm": 0.6897677482032446,
+      "learning_rate": 2.3820224719101125e-06,
+      "loss": 0.4401,
+      "step": 53
+    },
+    {
+      "epoch": 0.060648603116664326,
+      "grad_norm": 0.661989923280511,
+      "learning_rate": 2.4269662921348315e-06,
+      "loss": 0.4453,
+      "step": 54
+    },
+    {
+      "epoch": 0.061771725396602556,
+      "grad_norm": 0.6769101351556205,
+      "learning_rate": 2.4719101123595505e-06,
+      "loss": 0.4402,
+      "step": 55
+    },
+    {
+      "epoch": 0.06289484767654079,
+      "grad_norm": 0.6812241363735265,
+      "learning_rate": 2.51685393258427e-06,
+      "loss": 0.4645,
+      "step": 56
+    },
+    {
+      "epoch": 0.06401796995647902,
+      "grad_norm": 0.6757451847169096,
+      "learning_rate": 2.561797752808989e-06,
+      "loss": 0.4535,
+      "step": 57
+    },
+    {
+      "epoch": 0.06514109223641724,
+      "grad_norm": 0.6326256743817255,
+      "learning_rate": 2.606741573033708e-06,
+      "loss": 0.4194,
+      "step": 58
+    },
+    {
+      "epoch": 0.06626421451635547,
+      "grad_norm": 0.6800176400190888,
+      "learning_rate": 2.6516853932584274e-06,
+      "loss": 0.4541,
+      "step": 59
+    },
+    {
+      "epoch": 0.06738733679629369,
+      "grad_norm": 0.7164423610987447,
+      "learning_rate": 2.696629213483146e-06,
+      "loss": 0.4497,
+      "step": 60
+    },
+    {
+      "epoch": 0.06851045907623192,
+      "grad_norm": 0.7413520702346671,
+      "learning_rate": 2.7415730337078655e-06,
+      "loss": 0.4753,
+      "step": 61
+    },
+    {
+      "epoch": 0.06963358135617015,
+      "grad_norm": 0.7016085047019666,
+      "learning_rate": 2.7865168539325845e-06,
+      "loss": 0.4436,
+      "step": 62
+    },
+    {
+      "epoch": 0.07075670363610838,
+      "grad_norm": 0.6265784239939073,
+      "learning_rate": 2.8314606741573035e-06,
+      "loss": 0.4552,
+      "step": 63
+    },
+    {
+      "epoch": 0.07187982591604661,
+      "grad_norm": 0.674311288104449,
+      "learning_rate": 2.876404494382023e-06,
+      "loss": 0.458,
+      "step": 64
+    },
+    {
+      "epoch": 0.07300294819598484,
+      "grad_norm": 0.6446161293508706,
+      "learning_rate": 2.9213483146067416e-06,
+      "loss": 0.4525,
+      "step": 65
+    },
+    {
+      "epoch": 0.07412607047592307,
+      "grad_norm": 0.6444476566168919,
+      "learning_rate": 2.966292134831461e-06,
+      "loss": 0.4184,
+      "step": 66
+    },
+    {
+      "epoch": 0.0752491927558613,
+      "grad_norm": 0.6233096607639449,
+      "learning_rate": 3.0112359550561796e-06,
+      "loss": 0.457,
+      "step": 67
+    },
+    {
+      "epoch": 0.07637231503579953,
+      "grad_norm": 0.6648165469874694,
+      "learning_rate": 3.056179775280899e-06,
+      "loss": 0.4466,
+      "step": 68
+    },
+    {
+      "epoch": 0.07749543731573776,
+      "grad_norm": 0.6486022510349203,
+      "learning_rate": 3.1011235955056185e-06,
+      "loss": 0.4209,
+      "step": 69
+    },
+    {
+      "epoch": 0.07861855959567599,
+      "grad_norm": 0.6229828143529706,
+      "learning_rate": 3.146067415730337e-06,
+      "loss": 0.432,
+      "step": 70
+    },
+    {
+      "epoch": 0.0797416818756142,
+      "grad_norm": 0.5873293721571403,
+      "learning_rate": 3.1910112359550566e-06,
+      "loss": 0.4333,
+      "step": 71
+    },
+    {
+      "epoch": 0.08086480415555243,
+      "grad_norm": 0.6550686084928332,
+      "learning_rate": 3.235955056179775e-06,
+      "loss": 0.4342,
+      "step": 72
+    },
+    {
+      "epoch": 0.08198792643549066,
+      "grad_norm": 0.687632103391066,
+      "learning_rate": 3.2808988764044946e-06,
+      "loss": 0.4524,
+      "step": 73
+    },
+    {
+      "epoch": 0.08311104871542889,
+      "grad_norm": 0.5997642028480324,
+      "learning_rate": 3.325842696629214e-06,
+      "loss": 0.4151,
+      "step": 74
+    },
+    {
+      "epoch": 0.08423417099536712,
+      "grad_norm": 0.6141443910145349,
+      "learning_rate": 3.3707865168539327e-06,
+      "loss": 0.4107,
+      "step": 75
+    },
+    {
+      "epoch": 0.08535729327530535,
+      "grad_norm": 0.6086934688843036,
+      "learning_rate": 3.415730337078652e-06,
+      "loss": 0.4464,
+      "step": 76
+    },
+    {
+      "epoch": 0.08648041555524358,
+      "grad_norm": 0.6547734824010637,
+      "learning_rate": 3.4606741573033707e-06,
+      "loss": 0.4325,
+      "step": 77
+    },
+    {
+      "epoch": 0.08760353783518181,
+      "grad_norm": 0.6816992889881991,
+      "learning_rate": 3.50561797752809e-06,
+      "loss": 0.4632,
+      "step": 78
+    },
+    {
+      "epoch": 0.08872666011512004,
+      "grad_norm": 0.6858188783737146,
+      "learning_rate": 3.5505617977528096e-06,
+      "loss": 0.4527,
+      "step": 79
+    },
+    {
+      "epoch": 0.08984978239505827,
+      "grad_norm": 0.709230605422352,
+      "learning_rate": 3.595505617977528e-06,
+      "loss": 0.4385,
+      "step": 80
+    },
+    {
+      "epoch": 0.0909729046749965,
+      "grad_norm": 0.6185609555383293,
+      "learning_rate": 3.6404494382022476e-06,
+      "loss": 0.4295,
+      "step": 81
+    },
+    {
+      "epoch": 0.09209602695493471,
+      "grad_norm": 0.5593705773746459,
+      "learning_rate": 3.6853932584269662e-06,
+      "loss": 0.4107,
+      "step": 82
+    },
+    {
+      "epoch": 0.09321914923487294,
+      "grad_norm": 0.6449522343921041,
+      "learning_rate": 3.7303370786516857e-06,
+      "loss": 0.4297,
+      "step": 83
+    },
+    {
+      "epoch": 0.09434227151481117,
+      "grad_norm": 0.6162415569253238,
+      "learning_rate": 3.775280898876405e-06,
+      "loss": 0.3956,
+      "step": 84
+    },
+    {
+      "epoch": 0.0954653937947494,
+      "grad_norm": 0.6285861854737811,
+      "learning_rate": 3.820224719101124e-06,
+      "loss": 0.4209,
+      "step": 85
+    },
+    {
+      "epoch": 0.09658851607468763,
+      "grad_norm": 0.6005391526643882,
+      "learning_rate": 3.865168539325843e-06,
+      "loss": 0.4344,
+      "step": 86
+    },
+    {
+      "epoch": 0.09771163835462586,
+      "grad_norm": 0.6037551054743073,
+      "learning_rate": 3.910112359550562e-06,
+      "loss": 0.4238,
+      "step": 87
+    },
+    {
+      "epoch": 0.09883476063456409,
+      "grad_norm": 0.6317561271143405,
+      "learning_rate": 3.955056179775282e-06,
+      "loss": 0.4313,
+      "step": 88
+    },
+    {
+      "epoch": 0.09995788291450232,
+      "grad_norm": 0.6740920615829088,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.4228,
+      "step": 89
+    },
+    {
+      "epoch": 0.10108100519444055,
+      "grad_norm": 0.6163555222810396,
+      "learning_rate": 4.04494382022472e-06,
+      "loss": 0.4297,
+      "step": 90
+    },
+    {
+      "epoch": 0.10220412747437878,
+      "grad_norm": 0.5924398711122163,
+      "learning_rate": 4.089887640449439e-06,
+      "loss": 0.413,
+      "step": 91
+    },
+    {
+      "epoch": 0.10332724975431701,
+      "grad_norm": 0.6450622353301865,
+      "learning_rate": 4.134831460674158e-06,
+      "loss": 0.4055,
+      "step": 92
+    },
+    {
+      "epoch": 0.10445037203425522,
+      "grad_norm": 0.5951767882437019,
+      "learning_rate": 4.179775280898877e-06,
+      "loss": 0.4151,
+      "step": 93
+    },
+    {
+      "epoch": 0.10557349431419345,
+      "grad_norm": 0.5644268389534823,
+      "learning_rate": 4.224719101123596e-06,
+      "loss": 0.3983,
+      "step": 94
+    },
+    {
+      "epoch": 0.10669661659413168,
+      "grad_norm": 0.6348084088963297,
+      "learning_rate": 4.269662921348315e-06,
+      "loss": 0.4071,
+      "step": 95
+    },
+    {
+      "epoch": 0.10781973887406991,
+      "grad_norm": 0.5862325727673693,
+      "learning_rate": 4.314606741573034e-06,
+      "loss": 0.43,
+      "step": 96
+    },
+    {
+      "epoch": 0.10894286115400814,
+      "grad_norm": 0.6202803723418159,
+      "learning_rate": 4.359550561797753e-06,
+      "loss": 0.4189,
+      "step": 97
+    },
+    {
+      "epoch": 0.11006598343394637,
+      "grad_norm": 0.5944376188372148,
+      "learning_rate": 4.404494382022472e-06,
+      "loss": 0.4106,
+      "step": 98
+    },
+    {
+      "epoch": 0.1111891057138846,
+      "grad_norm": 0.6071370637839878,
+      "learning_rate": 4.449438202247192e-06,
+      "loss": 0.4168,
+      "step": 99
+    },
+    {
+      "epoch": 0.11231222799382283,
+      "grad_norm": 0.6135945175323315,
+      "learning_rate": 4.494382022471911e-06,
+      "loss": 0.4011,
+      "step": 100
+    },
+    {
+      "epoch": 0.11343535027376106,
+      "grad_norm": 0.5901989509886602,
+      "learning_rate": 4.53932584269663e-06,
+      "loss": 0.4226,
+      "step": 101
+    },
+    {
+      "epoch": 0.11455847255369929,
+      "grad_norm": 0.5719385399517765,
+      "learning_rate": 4.584269662921349e-06,
+      "loss": 0.4207,
+      "step": 102
+    },
+    {
+      "epoch": 0.11568159483363752,
+      "grad_norm": 0.6073755395198991,
+      "learning_rate": 4.629213483146068e-06,
+      "loss": 0.4263,
+      "step": 103
+    },
+    {
+      "epoch": 0.11680471711357573,
+      "grad_norm": 0.6019653421562294,
+      "learning_rate": 4.674157303370787e-06,
+      "loss": 0.4149,
+      "step": 104
+    },
+    {
+      "epoch": 0.11792783939351396,
+      "grad_norm": 0.6153507547879732,
+      "learning_rate": 4.719101123595506e-06,
+      "loss": 0.4089,
+      "step": 105
+    },
+    {
+      "epoch": 0.1190509616734522,
+      "grad_norm": 0.5829914630184178,
+      "learning_rate": 4.764044943820225e-06,
+      "loss": 0.4172,
+      "step": 106
+    },
+    {
+      "epoch": 0.12017408395339042,
+      "grad_norm": 0.6272470387655589,
+      "learning_rate": 4.808988764044944e-06,
+      "loss": 0.4265,
+      "step": 107
+    },
+    {
+      "epoch": 0.12129720623332865,
+      "grad_norm": 0.6284052319460708,
+      "learning_rate": 4.853932584269663e-06,
+      "loss": 0.4422,
+      "step": 108
+    },
+    {
+      "epoch": 0.12242032851326688,
+      "grad_norm": 0.6197638566033578,
+      "learning_rate": 4.898876404494383e-06,
+      "loss": 0.4058,
+      "step": 109
+    },
+    {
+      "epoch": 0.12354345079320511,
+      "grad_norm": 0.6033230010642877,
+      "learning_rate": 4.943820224719101e-06,
+      "loss": 0.4245,
+      "step": 110
+    },
+    {
+      "epoch": 0.12466657307314334,
+      "grad_norm": 0.630498552024084,
+      "learning_rate": 4.988764044943821e-06,
+      "loss": 0.3942,
+      "step": 111
+    },
+    {
+      "epoch": 0.12578969535308157,
+      "grad_norm": 0.6493542531086268,
+      "learning_rate": 5.03370786516854e-06,
+      "loss": 0.4363,
+      "step": 112
+    },
+    {
+      "epoch": 0.1269128176330198,
+      "grad_norm": 0.6237798245035485,
+      "learning_rate": 5.078651685393259e-06,
+      "loss": 0.4159,
+      "step": 113
+    },
+    {
+      "epoch": 0.12803593991295803,
+      "grad_norm": 0.6597150507909345,
+      "learning_rate": 5.123595505617978e-06,
+      "loss": 0.4351,
+      "step": 114
+    },
+    {
+      "epoch": 0.12915906219289625,
+      "grad_norm": 0.5980066757551353,
+      "learning_rate": 5.168539325842698e-06,
+      "loss": 0.389,
+      "step": 115
+    },
+    {
+      "epoch": 0.1302821844728345,
+      "grad_norm": 0.6404924520081225,
+      "learning_rate": 5.213483146067416e-06,
+      "loss": 0.4229,
+      "step": 116
+    },
+    {
+      "epoch": 0.1314053067527727,
+      "grad_norm": 0.6144448252232805,
+      "learning_rate": 5.258426966292135e-06,
+      "loss": 0.4096,
+      "step": 117
+    },
+    {
+      "epoch": 0.13252842903271095,
+      "grad_norm": 0.638697930023902,
+      "learning_rate": 5.303370786516855e-06,
+      "loss": 0.4296,
+      "step": 118
+    },
+    {
+      "epoch": 0.13365155131264916,
+      "grad_norm": 0.6304024554599722,
+      "learning_rate": 5.348314606741574e-06,
+      "loss": 0.4136,
+      "step": 119
+    },
+    {
+      "epoch": 0.13477467359258738,
+      "grad_norm": 0.6253982419034039,
+      "learning_rate": 5.393258426966292e-06,
+      "loss": 0.4171,
+      "step": 120
+    },
+    {
+      "epoch": 0.13589779587252562,
+      "grad_norm": 0.5881906390342778,
+      "learning_rate": 5.438202247191011e-06,
+      "loss": 0.4079,
+      "step": 121
+    },
+    {
+      "epoch": 0.13702091815246384,
+      "grad_norm": 0.6940061067312768,
+      "learning_rate": 5.483146067415731e-06,
+      "loss": 0.42,
+      "step": 122
+    },
+    {
+      "epoch": 0.13814404043240208,
+      "grad_norm": 0.620303869723247,
+      "learning_rate": 5.52808988764045e-06,
+      "loss": 0.4058,
+      "step": 123
+    },
+    {
+      "epoch": 0.1392671627123403,
+      "grad_norm": 0.5917642817410118,
+      "learning_rate": 5.573033707865169e-06,
+      "loss": 0.4048,
+      "step": 124
+    },
+    {
+      "epoch": 0.14039028499227854,
+      "grad_norm": 0.578694471042719,
+      "learning_rate": 5.617977528089889e-06,
+      "loss": 0.3975,
+      "step": 125
+    },
+    {
+      "epoch": 0.14151340727221676,
+      "grad_norm": 0.6356057703453721,
+      "learning_rate": 5.662921348314607e-06,
+      "loss": 0.4232,
+      "step": 126
+    },
+    {
+      "epoch": 0.142636529552155,
+      "grad_norm": 0.63035695832545,
+      "learning_rate": 5.707865168539326e-06,
+      "loss": 0.4061,
+      "step": 127
+    },
+    {
+      "epoch": 0.14375965183209322,
+      "grad_norm": 0.5901449106195371,
+      "learning_rate": 5.752808988764046e-06,
+      "loss": 0.394,
+      "step": 128
+    },
+    {
+      "epoch": 0.14488277411203146,
+      "grad_norm": 0.6552219033314495,
+      "learning_rate": 5.797752808988765e-06,
+      "loss": 0.4228,
+      "step": 129
+    },
+    {
+      "epoch": 0.14600589639196968,
+      "grad_norm": 0.5897358290667347,
+      "learning_rate": 5.842696629213483e-06,
+      "loss": 0.4097,
+      "step": 130
+    },
+    {
+      "epoch": 0.1471290186719079,
+      "grad_norm": 0.6506458224613755,
+      "learning_rate": 5.887640449438202e-06,
+      "loss": 0.4193,
+      "step": 131
+    },
+    {
+      "epoch": 0.14825214095184613,
+      "grad_norm": 0.5884872839199834,
+      "learning_rate": 5.932584269662922e-06,
+      "loss": 0.3958,
+      "step": 132
+    },
+    {
+      "epoch": 0.14937526323178435,
+      "grad_norm": 0.6383694194701096,
+      "learning_rate": 5.977528089887641e-06,
+      "loss": 0.4166,
+      "step": 133
+    },
+    {
+      "epoch": 0.1504983855117226,
+      "grad_norm": 0.5988685305190284,
+      "learning_rate": 6.022471910112359e-06,
+      "loss": 0.3987,
+      "step": 134
+    },
+    {
+      "epoch": 0.1516215077916608,
+      "grad_norm": 0.6008483348458779,
+      "learning_rate": 6.06741573033708e-06,
+      "loss": 0.4054,
+      "step": 135
+    },
+    {
+      "epoch": 0.15274463007159905,
+      "grad_norm": 0.6371268436916242,
+      "learning_rate": 6.112359550561798e-06,
+      "loss": 0.415,
+      "step": 136
+    },
+    {
+      "epoch": 0.15386775235153727,
+      "grad_norm": 0.5873165979042162,
+      "learning_rate": 6.157303370786517e-06,
+      "loss": 0.3788,
+      "step": 137
+    },
+    {
+      "epoch": 0.1549908746314755,
+      "grad_norm": 0.6321649406199696,
+      "learning_rate": 6.202247191011237e-06,
+      "loss": 0.3874,
+      "step": 138
+    },
+    {
+      "epoch": 0.15611399691141373,
+      "grad_norm": 0.5905156513112094,
+      "learning_rate": 6.247191011235956e-06,
+      "loss": 0.3865,
+      "step": 139
+    },
+    {
+      "epoch": 0.15723711919135197,
+      "grad_norm": 0.6605429195870949,
+      "learning_rate": 6.292134831460674e-06,
+      "loss": 0.4218,
+      "step": 140
+    },
+    {
+      "epoch": 0.1583602414712902,
+      "grad_norm": 0.6197445457794568,
+      "learning_rate": 6.337078651685393e-06,
+      "loss": 0.3986,
+      "step": 141
+    },
+    {
+      "epoch": 0.1594833637512284,
+      "grad_norm": 0.6510054129028581,
+      "learning_rate": 6.382022471910113e-06,
+      "loss": 0.4126,
+      "step": 142
+    },
+    {
+      "epoch": 0.16060648603116665,
+      "grad_norm": 0.6637714327455877,
+      "learning_rate": 6.426966292134832e-06,
+      "loss": 0.4105,
+      "step": 143
+    },
+    {
+      "epoch": 0.16172960831110486,
+      "grad_norm": 0.6320653980077073,
+      "learning_rate": 6.47191011235955e-06,
+      "loss": 0.3876,
+      "step": 144
+    },
+    {
+      "epoch": 0.1628527305910431,
+      "grad_norm": 0.6224333021729421,
+      "learning_rate": 6.51685393258427e-06,
+      "loss": 0.3975,
+      "step": 145
+    },
+    {
+      "epoch": 0.16397585287098132,
+      "grad_norm": 0.6523728785937435,
+      "learning_rate": 6.561797752808989e-06,
+      "loss": 0.4088,
+      "step": 146
+    },
+    {
+      "epoch": 0.16509897515091956,
+      "grad_norm": 0.6399877537131693,
+      "learning_rate": 6.606741573033708e-06,
+      "loss": 0.4243,
+      "step": 147
+    },
+    {
+      "epoch": 0.16622209743085778,
+      "grad_norm": 0.6785871204181669,
+      "learning_rate": 6.651685393258428e-06,
+      "loss": 0.4118,
+      "step": 148
+    },
+    {
+      "epoch": 0.16734521971079602,
+      "grad_norm": 0.6793168925384406,
+      "learning_rate": 6.696629213483147e-06,
+      "loss": 0.3843,
+      "step": 149
+    },
+    {
+      "epoch": 0.16846834199073424,
+      "grad_norm": 0.6765524645653702,
+      "learning_rate": 6.741573033707865e-06,
+      "loss": 0.3929,
+      "step": 150
+    },
+    {
+      "epoch": 0.16959146427067248,
+      "grad_norm": 0.6500442496922796,
+      "learning_rate": 6.786516853932584e-06,
+      "loss": 0.4157,
+      "step": 151
+    },
+    {
+      "epoch": 0.1707145865506107,
+      "grad_norm": 0.6278514376034972,
+      "learning_rate": 6.831460674157304e-06,
+      "loss": 0.384,
+      "step": 152
+    },
+    {
+      "epoch": 0.1718377088305489,
+      "grad_norm": 0.6786012740240396,
+      "learning_rate": 6.876404494382023e-06,
+      "loss": 0.4213,
+      "step": 153
+    },
+    {
+      "epoch": 0.17296083111048716,
+      "grad_norm": 0.6103427664078407,
+      "learning_rate": 6.921348314606741e-06,
+      "loss": 0.3827,
+      "step": 154
+    },
+    {
+      "epoch": 0.17408395339042537,
+      "grad_norm": 0.6373236421700184,
+      "learning_rate": 6.966292134831461e-06,
+      "loss": 0.4021,
+      "step": 155
+    },
+    {
+      "epoch": 0.17520707567036362,
+      "grad_norm": 0.6527172283097773,
+      "learning_rate": 7.01123595505618e-06,
+      "loss": 0.415,
+      "step": 156
+    },
+    {
+      "epoch": 0.17633019795030183,
+      "grad_norm": 0.6299621366906155,
+      "learning_rate": 7.056179775280899e-06,
+      "loss": 0.3922,
+      "step": 157
+    },
+    {
+      "epoch": 0.17745332023024007,
+      "grad_norm": 0.6205260967234852,
+      "learning_rate": 7.101123595505619e-06,
+      "loss": 0.404,
+      "step": 158
+    },
+    {
+      "epoch": 0.1785764425101783,
+      "grad_norm": 0.6314521969215471,
+      "learning_rate": 7.146067415730338e-06,
+      "loss": 0.4,
+      "step": 159
+    },
+    {
+      "epoch": 0.17969956479011653,
+      "grad_norm": 0.5826299033616584,
+      "learning_rate": 7.191011235955056e-06,
+      "loss": 0.3936,
+      "step": 160
+    },
+    {
+      "epoch": 0.18082268707005475,
+      "grad_norm": 0.6340335724127424,
+      "learning_rate": 7.235955056179775e-06,
+      "loss": 0.4234,
+      "step": 161
+    },
+    {
+      "epoch": 0.181945809349993,
+      "grad_norm": 0.632801507097347,
+      "learning_rate": 7.280898876404495e-06,
+      "loss": 0.4098,
+      "step": 162
+    },
+    {
+      "epoch": 0.1830689316299312,
+      "grad_norm": 0.6446077682121513,
+      "learning_rate": 7.325842696629214e-06,
+      "loss": 0.3846,
+      "step": 163
+    },
+    {
+      "epoch": 0.18419205390986942,
+      "grad_norm": 0.5774497489736088,
+      "learning_rate": 7.3707865168539325e-06,
+      "loss": 0.3763,
+      "step": 164
+    },
+    {
+      "epoch": 0.18531517618980767,
+      "grad_norm": 0.6295990680542162,
+      "learning_rate": 7.415730337078652e-06,
+      "loss": 0.4073,
+      "step": 165
+    },
+    {
+      "epoch": 0.18643829846974588,
+      "grad_norm": 0.6085108426525031,
+      "learning_rate": 7.460674157303371e-06,
+      "loss": 0.4139,
+      "step": 166
+    },
+    {
+      "epoch": 0.18756142074968413,
+      "grad_norm": 0.62992988248332,
+      "learning_rate": 7.50561797752809e-06,
+      "loss": 0.4138,
+      "step": 167
+    },
+    {
+      "epoch": 0.18868454302962234,
+      "grad_norm": 0.5836300572336676,
+      "learning_rate": 7.55056179775281e-06,
+      "loss": 0.3736,
+      "step": 168
+    },
+    {
+      "epoch": 0.18980766530956059,
+      "grad_norm": 0.6265603677231923,
+      "learning_rate": 7.5955056179775284e-06,
+      "loss": 0.375,
+      "step": 169
+    },
+    {
+      "epoch": 0.1909307875894988,
+      "grad_norm": 0.5870206417930635,
+      "learning_rate": 7.640449438202247e-06,
+      "loss": 0.4102,
+      "step": 170
+    },
+    {
+      "epoch": 0.19205390986943705,
+      "grad_norm": 0.59819579045645,
+      "learning_rate": 7.685393258426966e-06,
+      "loss": 0.4121,
+      "step": 171
+    },
+    {
+      "epoch": 0.19317703214937526,
+      "grad_norm": 0.6122535374224991,
+      "learning_rate": 7.730337078651686e-06,
+      "loss": 0.3967,
+      "step": 172
+    },
+    {
+      "epoch": 0.1943001544293135,
+      "grad_norm": 0.5721725402345272,
+      "learning_rate": 7.775280898876405e-06,
+      "loss": 0.4097,
+      "step": 173
+    },
+    {
+      "epoch": 0.19542327670925172,
+      "grad_norm": 0.5792461141723496,
+      "learning_rate": 7.820224719101124e-06,
+      "loss": 0.396,
+      "step": 174
+    },
+    {
+      "epoch": 0.19654639898918994,
+      "grad_norm": 0.5566347774031598,
+      "learning_rate": 7.865168539325843e-06,
+      "loss": 0.3748,
+      "step": 175
+    },
+    {
+      "epoch": 0.19766952126912818,
+      "grad_norm": 0.5947888148668254,
+      "learning_rate": 7.910112359550563e-06,
+      "loss": 0.3984,
+      "step": 176
+    },
+    {
+      "epoch": 0.1987926435490664,
+      "grad_norm": 0.5729701631007845,
+      "learning_rate": 7.955056179775281e-06,
+      "loss": 0.3732,
+      "step": 177
+    },
+    {
+      "epoch": 0.19991576582900464,
+      "grad_norm": 0.6308612644303433,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 0.3779,
+      "step": 178
+    },
+    {
+      "epoch": 0.20103888810894285,
+      "grad_norm": 0.602666895171434,
+      "learning_rate": 8.04494382022472e-06,
+      "loss": 0.4168,
+      "step": 179
+    },
+    {
+      "epoch": 0.2021620103888811,
+      "grad_norm": 0.594878749394712,
+      "learning_rate": 8.08988764044944e-06,
+      "loss": 0.3801,
+      "step": 180
+    },
+    {
+      "epoch": 0.2032851326688193,
+      "grad_norm": 0.5819103337590406,
+      "learning_rate": 8.13483146067416e-06,
+      "loss": 0.355,
+      "step": 181
+    },
+    {
+      "epoch": 0.20440825494875756,
+      "grad_norm": 0.6795698806664618,
+      "learning_rate": 8.179775280898877e-06,
+      "loss": 0.4105,
+      "step": 182
+    },
+    {
+      "epoch": 0.20553137722869577,
+      "grad_norm": 0.6176253376211935,
+      "learning_rate": 8.224719101123596e-06,
+      "loss": 0.4052,
+      "step": 183
+    },
+    {
+      "epoch": 0.20665449950863402,
+      "grad_norm": 0.5647204982647233,
+      "learning_rate": 8.269662921348315e-06,
+      "loss": 0.4002,
+      "step": 184
+    },
+    {
+      "epoch": 0.20777762178857223,
+      "grad_norm": 0.630533784994395,
+      "learning_rate": 8.314606741573035e-06,
+      "loss": 0.3836,
+      "step": 185
+    },
+    {
+      "epoch": 0.20890074406851045,
+      "grad_norm": 0.6619715293898263,
+      "learning_rate": 8.359550561797754e-06,
+      "loss": 0.4138,
+      "step": 186
+    },
+    {
+      "epoch": 0.2100238663484487,
+      "grad_norm": 0.5966032860325108,
+      "learning_rate": 8.404494382022472e-06,
+      "loss": 0.3896,
+      "step": 187
+    },
+    {
+      "epoch": 0.2111469886283869,
+      "grad_norm": 0.5681040377904722,
+      "learning_rate": 8.449438202247192e-06,
+      "loss": 0.3812,
+      "step": 188
+    },
+    {
+      "epoch": 0.21227011090832515,
+      "grad_norm": 0.6236783202259868,
+      "learning_rate": 8.494382022471911e-06,
+      "loss": 0.3927,
+      "step": 189
+    },
+    {
+      "epoch": 0.21339323318826336,
+      "grad_norm": 0.7424858758861099,
+      "learning_rate": 8.53932584269663e-06,
+      "loss": 0.4452,
+      "step": 190
+    },
+    {
+      "epoch": 0.2145163554682016,
+      "grad_norm": 0.7125468158261177,
+      "learning_rate": 8.58426966292135e-06,
+      "loss": 0.3609,
+      "step": 191
+    },
+    {
+      "epoch": 0.21563947774813982,
+      "grad_norm": 0.6157709477352911,
+      "learning_rate": 8.629213483146068e-06,
+      "loss": 0.4078,
+      "step": 192
+    },
+    {
+      "epoch": 0.21676260002807807,
+      "grad_norm": 0.6307743638624984,
+      "learning_rate": 8.674157303370788e-06,
+      "loss": 0.3683,
+      "step": 193
+    },
+    {
+      "epoch": 0.21788572230801628,
+      "grad_norm": 0.6873653868391115,
+      "learning_rate": 8.719101123595506e-06,
+      "loss": 0.4076,
+      "step": 194
+    },
+    {
+      "epoch": 0.21900884458795453,
+      "grad_norm": 0.6311262794185428,
+      "learning_rate": 8.764044943820226e-06,
+      "loss": 0.3853,
+      "step": 195
+    },
+    {
+      "epoch": 0.22013196686789274,
+      "grad_norm": 0.6680702186561743,
+      "learning_rate": 8.808988764044944e-06,
+      "loss": 0.4135,
+      "step": 196
+    },
+    {
+      "epoch": 0.22125508914783096,
+      "grad_norm": 0.6058308110971806,
+      "learning_rate": 8.853932584269664e-06,
+      "loss": 0.4112,
+      "step": 197
+    },
+    {
+      "epoch": 0.2223782114277692,
+      "grad_norm": 0.607906340987603,
+      "learning_rate": 8.898876404494383e-06,
+      "loss": 0.3979,
+      "step": 198
+    },
+    {
+      "epoch": 0.22350133370770742,
+      "grad_norm": 0.5987320863883703,
+      "learning_rate": 8.943820224719102e-06,
+      "loss": 0.4017,
+      "step": 199
+    },
+    {
+      "epoch": 0.22462445598764566,
+      "grad_norm": 0.6240143951994302,
+      "learning_rate": 8.988764044943822e-06,
+      "loss": 0.3869,
+      "step": 200
+    },
+    {
+      "epoch": 0.22574757826758388,
+      "grad_norm": 0.6484774713204989,
+      "learning_rate": 9.033707865168541e-06,
+      "loss": 0.3682,
+      "step": 201
+    },
+    {
+      "epoch": 0.22687070054752212,
+      "grad_norm": 0.6398454864125076,
+      "learning_rate": 9.07865168539326e-06,
+      "loss": 0.4031,
+      "step": 202
+    },
+    {
+      "epoch": 0.22799382282746034,
+      "grad_norm": 0.6071698475923116,
+      "learning_rate": 9.123595505617978e-06,
+      "loss": 0.4081,
+      "step": 203
+    },
+    {
+      "epoch": 0.22911694510739858,
+      "grad_norm": 0.5841395417042217,
+      "learning_rate": 9.168539325842698e-06,
+      "loss": 0.3842,
+      "step": 204
+    },
+    {
+      "epoch": 0.2302400673873368,
+      "grad_norm": 0.6674576527002625,
+      "learning_rate": 9.213483146067417e-06,
+      "loss": 0.4123,
+      "step": 205
+    },
+    {
+      "epoch": 0.23136318966727504,
+      "grad_norm": 0.6640606414419046,
+      "learning_rate": 9.258426966292136e-06,
+      "loss": 0.3939,
+      "step": 206
+    },
+    {
+      "epoch": 0.23248631194721325,
+      "grad_norm": 0.7214937941577327,
+      "learning_rate": 9.303370786516854e-06,
+      "loss": 0.371,
+      "step": 207
+    },
+    {
+      "epoch": 0.23360943422715147,
+      "grad_norm": 0.6793273791859282,
+      "learning_rate": 9.348314606741574e-06,
+      "loss": 0.3836,
+      "step": 208
+    },
+    {
+      "epoch": 0.2347325565070897,
+      "grad_norm": 0.6871956158863509,
+      "learning_rate": 9.393258426966294e-06,
+      "loss": 0.3932,
+      "step": 209
+    },
+    {
+      "epoch": 0.23585567878702793,
+      "grad_norm": 0.7258816009143834,
+      "learning_rate": 9.438202247191012e-06,
+      "loss": 0.414,
+      "step": 210
+    },
+    {
+      "epoch": 0.23697880106696617,
+      "grad_norm": 0.6492267333090342,
+      "learning_rate": 9.483146067415732e-06,
+      "loss": 0.416,
+      "step": 211
+    },
+    {
+      "epoch": 0.2381019233469044,
+      "grad_norm": 0.6041513090496686,
+      "learning_rate": 9.52808988764045e-06,
+      "loss": 0.3934,
+      "step": 212
+    },
+    {
+      "epoch": 0.23922504562684263,
+      "grad_norm": 0.6474076777760845,
+      "learning_rate": 9.57303370786517e-06,
+      "loss": 0.3975,
+      "step": 213
+    },
+    {
+      "epoch": 0.24034816790678085,
+      "grad_norm": 0.7092379598633823,
+      "learning_rate": 9.617977528089888e-06,
+      "loss": 0.3781,
+      "step": 214
+    },
+    {
+      "epoch": 0.2414712901867191,
+      "grad_norm": 0.6438002116207439,
+      "learning_rate": 9.662921348314608e-06,
+      "loss": 0.3996,
+      "step": 215
+    },
+    {
+      "epoch": 0.2425944124666573,
+      "grad_norm": 0.6622015252030728,
+      "learning_rate": 9.707865168539326e-06,
+      "loss": 0.3832,
+      "step": 216
+    },
+    {
+      "epoch": 0.24371753474659555,
+      "grad_norm": 0.6078817480395128,
+      "learning_rate": 9.752808988764046e-06,
+      "loss": 0.3952,
+      "step": 217
+    },
+    {
+      "epoch": 0.24484065702653376,
+      "grad_norm": 0.6551657494425536,
+      "learning_rate": 9.797752808988766e-06,
+      "loss": 0.3791,
+      "step": 218
+    },
+    {
+      "epoch": 0.24596377930647198,
+      "grad_norm": 0.6433323266927976,
+      "learning_rate": 9.842696629213484e-06,
+      "loss": 0.3774,
+      "step": 219
+    },
+    {
+      "epoch": 0.24708690158641022,
+      "grad_norm": 0.6188804166540856,
+      "learning_rate": 9.887640449438202e-06,
+      "loss": 0.3712,
+      "step": 220
+    },
+    {
+      "epoch": 0.24821002386634844,
+      "grad_norm": 0.671966975214175,
+      "learning_rate": 9.932584269662922e-06,
+      "loss": 0.4239,
+      "step": 221
+    },
+    {
+      "epoch": 0.24933314614628668,
+      "grad_norm": 0.5947753032122628,
+      "learning_rate": 9.977528089887642e-06,
+      "loss": 0.3964,
+      "step": 222
+    },
+    {
+      "epoch": 0.2504562684262249,
+      "grad_norm": 0.6607682301911262,
+      "learning_rate": 1.0022471910112362e-05,
+      "loss": 0.3903,
+      "step": 223
+    },
+    {
+      "epoch": 0.25157939070616314,
+      "grad_norm": 0.6356386896335997,
+      "learning_rate": 1.006741573033708e-05,
+      "loss": 0.4251,
+      "step": 224
+    },
+    {
+      "epoch": 0.25270251298610136,
+      "grad_norm": 0.6559033997420501,
+      "learning_rate": 1.01123595505618e-05,
+      "loss": 0.4137,
+      "step": 225
+    },
+    {
+      "epoch": 0.2538256352660396,
+      "grad_norm": 0.614989946033004,
+      "learning_rate": 1.0157303370786518e-05,
+      "loss": 0.3601,
+      "step": 226
+    },
+    {
+      "epoch": 0.25494875754597784,
+      "grad_norm": 0.6410991676909698,
+      "learning_rate": 1.0202247191011236e-05,
+      "loss": 0.402,
+      "step": 227
+    },
+    {
+      "epoch": 0.25607187982591606,
+      "grad_norm": 0.642417180353421,
+      "learning_rate": 1.0247191011235956e-05,
+      "loss": 0.3877,
+      "step": 228
+    },
+    {
+      "epoch": 0.2571950021058543,
+      "grad_norm": 0.6125173108674232,
+      "learning_rate": 1.0292134831460674e-05,
+      "loss": 0.3669,
+      "step": 229
+    },
+    {
+      "epoch": 0.2583181243857925,
+      "grad_norm": 0.5685847570038455,
+      "learning_rate": 1.0337078651685396e-05,
+      "loss": 0.374,
+      "step": 230
+    },
+    {
+      "epoch": 0.2594412466657307,
+      "grad_norm": 0.6544675947013324,
+      "learning_rate": 1.0382022471910114e-05,
+      "loss": 0.3834,
+      "step": 231
+    },
+    {
+      "epoch": 0.260564368945669,
+      "grad_norm": 0.6068343322603438,
+      "learning_rate": 1.0426966292134832e-05,
+      "loss": 0.3693,
+      "step": 232
+    },
+    {
+      "epoch": 0.2616874912256072,
+      "grad_norm": 0.6453583897346943,
+      "learning_rate": 1.0471910112359552e-05,
+      "loss": 0.4048,
+      "step": 233
+    },
+    {
+      "epoch": 0.2628106135055454,
+      "grad_norm": 0.7033942331352795,
+      "learning_rate": 1.051685393258427e-05,
+      "loss": 0.3875,
+      "step": 234
+    },
+    {
+      "epoch": 0.2639337357854836,
+      "grad_norm": 0.5715008472023303,
+      "learning_rate": 1.0561797752808988e-05,
+      "loss": 0.4069,
+      "step": 235
+    },
+    {
+      "epoch": 0.2650568580654219,
+      "grad_norm": 0.6494208885445116,
+      "learning_rate": 1.060674157303371e-05,
+      "loss": 0.4093,
+      "step": 236
+    },
+    {
+      "epoch": 0.2661799803453601,
+      "grad_norm": 0.6812214256281384,
+      "learning_rate": 1.0651685393258428e-05,
+      "loss": 0.4425,
+      "step": 237
+    },
+    {
+      "epoch": 0.26730310262529833,
+      "grad_norm": 0.5924772804766913,
+      "learning_rate": 1.0696629213483148e-05,
+      "loss": 0.387,
+      "step": 238
+    },
+    {
+      "epoch": 0.26842622490523654,
+      "grad_norm": 0.5764162625896104,
+      "learning_rate": 1.0741573033707866e-05,
+      "loss": 0.3827,
+      "step": 239
+    },
+    {
+      "epoch": 0.26954934718517476,
+      "grad_norm": 0.5965307495222452,
+      "learning_rate": 1.0786516853932584e-05,
+      "loss": 0.4129,
+      "step": 240
+    },
+    {
+      "epoch": 0.27067246946511303,
+      "grad_norm": 0.6003603754111506,
+      "learning_rate": 1.0831460674157304e-05,
+      "loss": 0.3775,
+      "step": 241
+    },
+    {
+      "epoch": 0.27179559174505125,
+      "grad_norm": 0.5922972797219954,
+      "learning_rate": 1.0876404494382022e-05,
+      "loss": 0.3877,
+      "step": 242
+    },
+    {
+      "epoch": 0.27291871402498946,
+      "grad_norm": 0.5685410001707504,
+      "learning_rate": 1.0921348314606744e-05,
+      "loss": 0.377,
+      "step": 243
+    },
+    {
+      "epoch": 0.2740418363049277,
+      "grad_norm": 0.576628928037405,
+      "learning_rate": 1.0966292134831462e-05,
+      "loss": 0.3971,
+      "step": 244
+    },
+    {
+      "epoch": 0.27516495858486595,
+      "grad_norm": 0.5579453193582883,
+      "learning_rate": 1.101123595505618e-05,
+      "loss": 0.378,
+      "step": 245
+    },
+    {
+      "epoch": 0.27628808086480416,
+      "grad_norm": 0.6011765922681929,
+      "learning_rate": 1.10561797752809e-05,
+      "loss": 0.4117,
+      "step": 246
+    },
+    {
+      "epoch": 0.2774112031447424,
+      "grad_norm": 0.6091646970674486,
+      "learning_rate": 1.1101123595505618e-05,
+      "loss": 0.4426,
+      "step": 247
+    },
+    {
+      "epoch": 0.2785343254246806,
+      "grad_norm": 0.5775758129728008,
+      "learning_rate": 1.1146067415730338e-05,
+      "loss": 0.3972,
+      "step": 248
+    },
+    {
+      "epoch": 0.27965744770461887,
+      "grad_norm": 0.6604411713439484,
+      "learning_rate": 1.1191011235955056e-05,
+      "loss": 0.3965,
+      "step": 249
+    },
+    {
+      "epoch": 0.2807805699845571,
+      "grad_norm": 0.6156558831358852,
+      "learning_rate": 1.1235955056179778e-05,
+      "loss": 0.405,
+      "step": 250
+    },
+    {
+      "epoch": 0.2819036922644953,
+      "grad_norm": 0.5623612665541853,
+      "learning_rate": 1.1280898876404496e-05,
+      "loss": 0.3901,
+      "step": 251
+    },
+    {
+      "epoch": 0.2830268145444335,
+      "grad_norm": 0.6746239022438368,
+      "learning_rate": 1.1325842696629214e-05,
+      "loss": 0.3876,
+      "step": 252
+    },
+    {
+      "epoch": 0.28414993682437173,
+      "grad_norm": 0.5964994154377549,
+      "learning_rate": 1.1370786516853934e-05,
+      "loss": 0.4163,
+      "step": 253
+    },
+    {
+      "epoch": 0.28527305910431,
+      "grad_norm": 0.6082500353591788,
+      "learning_rate": 1.1415730337078652e-05,
+      "loss": 0.4286,
+      "step": 254
+    },
+    {
+      "epoch": 0.2863961813842482,
+      "grad_norm": 0.6157506350996006,
+      "learning_rate": 1.146067415730337e-05,
+      "loss": 0.3872,
+      "step": 255
+    },
+    {
+      "epoch": 0.28751930366418643,
+      "grad_norm": 0.567260212959717,
+      "learning_rate": 1.1505617977528092e-05,
+      "loss": 0.4341,
+      "step": 256
+    },
+    {
+      "epoch": 0.28864242594412465,
+      "grad_norm": 0.5791317852426999,
+      "learning_rate": 1.155056179775281e-05,
+      "loss": 0.3499,
+      "step": 257
+    },
+    {
+      "epoch": 0.2897655482240629,
+      "grad_norm": 0.5741440071104256,
+      "learning_rate": 1.159550561797753e-05,
+      "loss": 0.3744,
+      "step": 258
+    },
+    {
+      "epoch": 0.29088867050400113,
+      "grad_norm": 0.577442355673411,
+      "learning_rate": 1.1640449438202248e-05,
+      "loss": 0.3959,
+      "step": 259
+    },
+    {
+      "epoch": 0.29201179278393935,
+      "grad_norm": 0.5701366764697986,
+      "learning_rate": 1.1685393258426966e-05,
+      "loss": 0.3661,
+      "step": 260
+    },
+    {
+      "epoch": 0.29313491506387757,
+      "grad_norm": 0.5866338939450678,
+      "learning_rate": 1.1730337078651686e-05,
+      "loss": 0.3817,
+      "step": 261
+    },
+    {
+      "epoch": 0.2942580373438158,
+      "grad_norm": 0.5869871652126405,
+      "learning_rate": 1.1775280898876404e-05,
+      "loss": 0.4031,
+      "step": 262
+    },
+    {
+      "epoch": 0.29538115962375405,
+      "grad_norm": 0.6113639662622179,
+      "learning_rate": 1.1820224719101126e-05,
+      "loss": 0.3665,
+      "step": 263
+    },
+    {
+      "epoch": 0.29650428190369227,
+      "grad_norm": 0.5815782372208732,
+      "learning_rate": 1.1865168539325844e-05,
+      "loss": 0.3718,
+      "step": 264
+    },
+    {
+      "epoch": 0.2976274041836305,
+      "grad_norm": 0.5770971481710677,
+      "learning_rate": 1.1910112359550562e-05,
+      "loss": 0.3887,
+      "step": 265
+    },
+    {
+      "epoch": 0.2987505264635687,
+      "grad_norm": 0.6378543358730397,
+      "learning_rate": 1.1955056179775282e-05,
+      "loss": 0.4026,
+      "step": 266
+    },
+    {
+      "epoch": 0.29987364874350697,
+      "grad_norm": 0.6366625582188497,
+      "learning_rate": 1.2e-05,
+      "loss": 0.3996,
+      "step": 267
+    },
+    {
+      "epoch": 0.3009967710234452,
+      "grad_norm": 0.5644318306665338,
+      "learning_rate": 1.2044943820224718e-05,
+      "loss": 0.3736,
+      "step": 268
+    },
+    {
+      "epoch": 0.3021198933033834,
+      "grad_norm": 0.5760984664326271,
+      "learning_rate": 1.208988764044944e-05,
+      "loss": 0.3831,
+      "step": 269
+    },
+    {
+      "epoch": 0.3032430155833216,
+      "grad_norm": 0.6073186836248484,
+      "learning_rate": 1.213483146067416e-05,
+      "loss": 0.3767,
+      "step": 270
+    },
+    {
+      "epoch": 0.3043661378632599,
+      "grad_norm": 0.658052587041447,
+      "learning_rate": 1.2179775280898878e-05,
+      "loss": 0.376,
+      "step": 271
+    },
+    {
+      "epoch": 0.3054892601431981,
+      "grad_norm": 0.5705277159249025,
+      "learning_rate": 1.2224719101123596e-05,
+      "loss": 0.4047,
+      "step": 272
+    },
+    {
+      "epoch": 0.3066123824231363,
+      "grad_norm": 0.5943437710250703,
+      "learning_rate": 1.2269662921348316e-05,
+      "loss": 0.3938,
+      "step": 273
+    },
+    {
+      "epoch": 0.30773550470307454,
+      "grad_norm": 0.6594875627018434,
+      "learning_rate": 1.2314606741573034e-05,
+      "loss": 0.3932,
+      "step": 274
+    },
+    {
+      "epoch": 0.30885862698301275,
+      "grad_norm": 0.6227401289864777,
+      "learning_rate": 1.2359550561797752e-05,
+      "loss": 0.368,
+      "step": 275
+    },
+    {
+      "epoch": 0.309981749262951,
+      "grad_norm": 0.5836285370383634,
+      "learning_rate": 1.2404494382022474e-05,
+      "loss": 0.3896,
+      "step": 276
+    },
+    {
+      "epoch": 0.31110487154288924,
+      "grad_norm": 0.6356736965747966,
+      "learning_rate": 1.2449438202247192e-05,
+      "loss": 0.4001,
+      "step": 277
+    },
+    {
+      "epoch": 0.31222799382282745,
+      "grad_norm": 0.5961019104865444,
+      "learning_rate": 1.2494382022471912e-05,
+      "loss": 0.3898,
+      "step": 278
+    },
+    {
+      "epoch": 0.31335111610276567,
+      "grad_norm": 0.5894192052755339,
+      "learning_rate": 1.253932584269663e-05,
+      "loss": 0.3821,
+      "step": 279
+    },
+    {
+      "epoch": 0.31447423838270394,
+      "grad_norm": 0.6139515035640439,
+      "learning_rate": 1.2584269662921348e-05,
+      "loss": 0.3869,
+      "step": 280
+    },
+    {
+      "epoch": 0.31559736066264216,
+      "grad_norm": 0.5710811085875507,
+      "learning_rate": 1.2629213483146068e-05,
+      "loss": 0.3715,
+      "step": 281
+    },
+    {
+      "epoch": 0.3167204829425804,
+      "grad_norm": 0.6174376536441986,
+      "learning_rate": 1.2674157303370786e-05,
+      "loss": 0.3887,
+      "step": 282
+    },
+    {
+      "epoch": 0.3178436052225186,
+      "grad_norm": 0.5409663843795987,
+      "learning_rate": 1.2719101123595508e-05,
+      "loss": 0.3585,
+      "step": 283
+    },
+    {
+      "epoch": 0.3189667275024568,
+      "grad_norm": 0.5842025638662178,
+      "learning_rate": 1.2764044943820226e-05,
+      "loss": 0.3648,
+      "step": 284
+    },
+    {
+      "epoch": 0.3200898497823951,
+      "grad_norm": 0.6288491942213608,
+      "learning_rate": 1.2808988764044944e-05,
+      "loss": 0.3943,
+      "step": 285
+    },
+    {
+      "epoch": 0.3212129720623333,
+      "grad_norm": 0.6339928168107889,
+      "learning_rate": 1.2853932584269664e-05,
+      "loss": 0.412,
+      "step": 286
+    },
+    {
+      "epoch": 0.3223360943422715,
+      "grad_norm": 0.5805070016232142,
+      "learning_rate": 1.2898876404494382e-05,
+      "loss": 0.3936,
+      "step": 287
+    },
+    {
+      "epoch": 0.3234592166222097,
+      "grad_norm": 0.6031249399955438,
+      "learning_rate": 1.29438202247191e-05,
+      "loss": 0.3849,
+      "step": 288
+    },
+    {
+      "epoch": 0.324582338902148,
+      "grad_norm": 0.6515523772005117,
+      "learning_rate": 1.2988764044943822e-05,
+      "loss": 0.4003,
+      "step": 289
+    },
+    {
+      "epoch": 0.3257054611820862,
+      "grad_norm": 0.6020531221934884,
+      "learning_rate": 1.303370786516854e-05,
+      "loss": 0.4059,
+      "step": 290
+    },
+    {
+      "epoch": 0.3268285834620244,
+      "grad_norm": 0.5911957722542618,
+      "learning_rate": 1.307865168539326e-05,
+      "loss": 0.3936,
+      "step": 291
+    },
+    {
+      "epoch": 0.32795170574196264,
+      "grad_norm": 0.6296525120353835,
+      "learning_rate": 1.3123595505617978e-05,
+      "loss": 0.3638,
+      "step": 292
+    },
+    {
+      "epoch": 0.3290748280219009,
+      "grad_norm": 0.6035340356032662,
+      "learning_rate": 1.3168539325842698e-05,
+      "loss": 0.3633,
+      "step": 293
+    },
+    {
+      "epoch": 0.3301979503018391,
+      "grad_norm": 0.5640220762566707,
+      "learning_rate": 1.3213483146067416e-05,
+      "loss": 0.3936,
+      "step": 294
+    },
+    {
+      "epoch": 0.33132107258177734,
+      "grad_norm": 0.659950946658227,
+      "learning_rate": 1.3258426966292135e-05,
+      "loss": 0.3939,
+      "step": 295
+    },
+    {
+      "epoch": 0.33244419486171556,
+      "grad_norm": 0.6547964598471208,
+      "learning_rate": 1.3303370786516856e-05,
+      "loss": 0.3797,
+      "step": 296
+    },
+    {
+      "epoch": 0.3335673171416538,
+      "grad_norm": 0.5628971266321644,
+      "learning_rate": 1.3348314606741574e-05,
+      "loss": 0.3878,
+      "step": 297
+    },
+    {
+      "epoch": 0.33469043942159205,
+      "grad_norm": 0.5916043311443739,
+      "learning_rate": 1.3393258426966294e-05,
+      "loss": 0.3705,
+      "step": 298
+    },
+    {
+      "epoch": 0.33581356170153026,
+      "grad_norm": 0.642531175947998,
+      "learning_rate": 1.3438202247191012e-05,
+      "loss": 0.3841,
+      "step": 299
+    },
+    {
+      "epoch": 0.3369366839814685,
+      "grad_norm": 0.5942467702427874,
+      "learning_rate": 1.348314606741573e-05,
+      "loss": 0.3864,
+      "step": 300
+    },
+    {
+      "epoch": 0.3380598062614067,
+      "grad_norm": 0.5576069981166258,
+      "learning_rate": 1.352808988764045e-05,
+      "loss": 0.3803,
+      "step": 301
+    },
+    {
+      "epoch": 0.33918292854134496,
+      "grad_norm": 0.637796095078525,
+      "learning_rate": 1.3573033707865169e-05,
+      "loss": 0.3845,
+      "step": 302
+    },
+    {
+      "epoch": 0.3403060508212832,
+      "grad_norm": 0.5424364787066615,
+      "learning_rate": 1.361797752808989e-05,
+      "loss": 0.3885,
+      "step": 303
+    },
+    {
+      "epoch": 0.3414291731012214,
+      "grad_norm": 0.666509048066953,
+      "learning_rate": 1.3662921348314608e-05,
+      "loss": 0.3941,
+      "step": 304
+    },
+    {
+      "epoch": 0.3425522953811596,
+      "grad_norm": 0.5667187793072223,
+      "learning_rate": 1.3707865168539327e-05,
+      "loss": 0.3568,
+      "step": 305
+    },
+    {
+      "epoch": 0.3436754176610978,
+      "grad_norm": 0.6364904468384852,
+      "learning_rate": 1.3752808988764046e-05,
+      "loss": 0.3807,
+      "step": 306
+    },
+    {
+      "epoch": 0.3447985399410361,
+      "grad_norm": 0.5971120227080189,
+      "learning_rate": 1.3797752808988765e-05,
+      "loss": 0.3848,
+      "step": 307
+    },
+    {
+      "epoch": 0.3459216622209743,
+      "grad_norm": 0.5555370524108522,
+      "learning_rate": 1.3842696629213483e-05,
+      "loss": 0.3849,
+      "step": 308
+    },
+    {
+      "epoch": 0.34704478450091253,
+      "grad_norm": 0.6236731385363299,
+      "learning_rate": 1.3887640449438204e-05,
+      "loss": 0.384,
+      "step": 309
+    },
+    {
+      "epoch": 0.34816790678085074,
+      "grad_norm": 0.6149787882648424,
+      "learning_rate": 1.3932584269662923e-05,
+      "loss": 0.3959,
+      "step": 310
+    },
+    {
+      "epoch": 0.349291029060789,
+      "grad_norm": 0.6215560633514793,
+      "learning_rate": 1.3977528089887642e-05,
+      "loss": 0.3918,
+      "step": 311
+    },
+    {
+      "epoch": 0.35041415134072723,
+      "grad_norm": 0.6209310433594782,
+      "learning_rate": 1.402247191011236e-05,
+      "loss": 0.3855,
+      "step": 312
+    },
+    {
+      "epoch": 0.35153727362066545,
+      "grad_norm": 0.5994586964661478,
+      "learning_rate": 1.4067415730337079e-05,
+      "loss": 0.373,
+      "step": 313
+    },
+    {
+      "epoch": 0.35266039590060366,
+      "grad_norm": 0.6188656847491276,
+      "learning_rate": 1.4112359550561799e-05,
+      "loss": 0.3892,
+      "step": 314
+    },
+    {
+      "epoch": 0.35378351818054193,
+      "grad_norm": 0.5701426572944555,
+      "learning_rate": 1.4157303370786517e-05,
+      "loss": 0.3587,
+      "step": 315
+    },
+    {
+      "epoch": 0.35490664046048015,
+      "grad_norm": 0.5735315497169365,
+      "learning_rate": 1.4202247191011238e-05,
+      "loss": 0.3834,
+      "step": 316
+    },
+    {
+      "epoch": 0.35602976274041837,
+      "grad_norm": 0.6041588815641994,
+      "learning_rate": 1.4247191011235957e-05,
+      "loss": 0.4063,
+      "step": 317
+    },
+    {
+      "epoch": 0.3571528850203566,
+      "grad_norm": 0.6413512842290581,
+      "learning_rate": 1.4292134831460676e-05,
+      "loss": 0.3643,
+      "step": 318
+    },
+    {
+      "epoch": 0.3582760073002948,
+      "grad_norm": 0.5633456389709988,
+      "learning_rate": 1.4337078651685395e-05,
+      "loss": 0.3782,
+      "step": 319
+    },
+    {
+      "epoch": 0.35939912958023307,
+      "grad_norm": 0.6268598664045065,
+      "learning_rate": 1.4382022471910113e-05,
+      "loss": 0.3959,
+      "step": 320
+    },
+    {
+      "epoch": 0.3605222518601713,
+      "grad_norm": 0.5800501810819474,
+      "learning_rate": 1.4426966292134833e-05,
+      "loss": 0.4013,
+      "step": 321
+    },
+    {
+      "epoch": 0.3616453741401095,
+      "grad_norm": 0.5769816751771281,
+      "learning_rate": 1.447191011235955e-05,
+      "loss": 0.3828,
+      "step": 322
+    },
+    {
+      "epoch": 0.3627684964200477,
+      "grad_norm": 0.5949290584396976,
+      "learning_rate": 1.4516853932584272e-05,
+      "loss": 0.4089,
+      "step": 323
+    },
+    {
+      "epoch": 0.363891618699986,
+      "grad_norm": 0.5852043429988183,
+      "learning_rate": 1.456179775280899e-05,
+      "loss": 0.3652,
+      "step": 324
+    },
+    {
+      "epoch": 0.3650147409799242,
+      "grad_norm": 0.572491832081323,
+      "learning_rate": 1.4606741573033709e-05,
+      "loss": 0.3984,
+      "step": 325
+    },
+    {
+      "epoch": 0.3661378632598624,
+      "grad_norm": 0.5972392800723866,
+      "learning_rate": 1.4651685393258429e-05,
+      "loss": 0.39,
+      "step": 326
+    },
+    {
+      "epoch": 0.36726098553980063,
+      "grad_norm": 0.6363257306206275,
+      "learning_rate": 1.4696629213483147e-05,
+      "loss": 0.3562,
+      "step": 327
+    },
+    {
+      "epoch": 0.36838410781973885,
+      "grad_norm": 0.5907940254264683,
+      "learning_rate": 1.4741573033707865e-05,
+      "loss": 0.3911,
+      "step": 328
+    },
+    {
+      "epoch": 0.3695072300996771,
+      "grad_norm": 0.5871546010212381,
+      "learning_rate": 1.4786516853932587e-05,
+      "loss": 0.379,
+      "step": 329
+    },
+    {
+      "epoch": 0.37063035237961534,
+      "grad_norm": 0.6618710636475701,
+      "learning_rate": 1.4831460674157305e-05,
+      "loss": 0.3946,
+      "step": 330
+    },
+    {
+      "epoch": 0.37175347465955355,
+      "grad_norm": 0.5734074870533208,
+      "learning_rate": 1.4876404494382025e-05,
+      "loss": 0.4272,
+      "step": 331
+    },
+    {
+      "epoch": 0.37287659693949177,
+      "grad_norm": 0.6643286523550941,
+      "learning_rate": 1.4921348314606743e-05,
+      "loss": 0.3963,
+      "step": 332
+    },
+    {
+      "epoch": 0.37399971921943004,
+      "grad_norm": 0.6096880457625593,
+      "learning_rate": 1.4966292134831461e-05,
+      "loss": 0.4006,
+      "step": 333
+    },
+    {
+      "epoch": 0.37512284149936825,
+      "grad_norm": 0.5812451405437918,
+      "learning_rate": 1.501123595505618e-05,
+      "loss": 0.3975,
+      "step": 334
+    },
+    {
+      "epoch": 0.37624596377930647,
+      "grad_norm": 0.5757226170172344,
+      "learning_rate": 1.5056179775280899e-05,
+      "loss": 0.3851,
+      "step": 335
+    },
+    {
+      "epoch": 0.3773690860592447,
+      "grad_norm": 0.5954813467325601,
+      "learning_rate": 1.510112359550562e-05,
+      "loss": 0.3817,
+      "step": 336
+    },
+    {
+      "epoch": 0.37849220833918296,
+      "grad_norm": 0.5726297825723002,
+      "learning_rate": 1.5146067415730339e-05,
+      "loss": 0.3882,
+      "step": 337
+    },
+    {
+      "epoch": 0.37961533061912117,
+      "grad_norm": 0.575616838800246,
+      "learning_rate": 1.5191011235955057e-05,
+      "loss": 0.4013,
+      "step": 338
+    },
+    {
+      "epoch": 0.3807384528990594,
+      "grad_norm": 0.5656104756542751,
+      "learning_rate": 1.5235955056179777e-05,
+      "loss": 0.3715,
+      "step": 339
+    },
+    {
+      "epoch": 0.3818615751789976,
+      "grad_norm": 0.5700630782329689,
+      "learning_rate": 1.5280898876404495e-05,
+      "loss": 0.382,
+      "step": 340
+    },
+    {
+      "epoch": 0.3829846974589358,
+      "grad_norm": 0.5901652073128528,
+      "learning_rate": 1.5325842696629213e-05,
+      "loss": 0.3922,
+      "step": 341
+    },
+    {
+      "epoch": 0.3841078197388741,
+      "grad_norm": 0.5680443357781206,
+      "learning_rate": 1.537078651685393e-05,
+      "loss": 0.3956,
+      "step": 342
+    },
+    {
+      "epoch": 0.3852309420188123,
+      "grad_norm": 0.6094240246912197,
+      "learning_rate": 1.5415730337078653e-05,
+      "loss": 0.4001,
+      "step": 343
+    },
+    {
+      "epoch": 0.3863540642987505,
+      "grad_norm": 0.6025863182389569,
+      "learning_rate": 1.546067415730337e-05,
+      "loss": 0.3991,
+      "step": 344
+    },
+    {
+      "epoch": 0.38747718657868874,
+      "grad_norm": 0.5849785735196502,
+      "learning_rate": 1.5505617977528093e-05,
+      "loss": 0.3561,
+      "step": 345
+    },
+    {
+      "epoch": 0.388600308858627,
+      "grad_norm": 0.6155288868533991,
+      "learning_rate": 1.555056179775281e-05,
+      "loss": 0.3885,
+      "step": 346
+    },
+    {
+      "epoch": 0.3897234311385652,
+      "grad_norm": 0.528624344003196,
+      "learning_rate": 1.559550561797753e-05,
+      "loss": 0.3368,
+      "step": 347
+    },
+    {
+      "epoch": 0.39084655341850344,
+      "grad_norm": 0.5742784751968288,
+      "learning_rate": 1.5640449438202247e-05,
+      "loss": 0.3908,
+      "step": 348
+    },
+    {
+      "epoch": 0.39196967569844166,
+      "grad_norm": 0.5790550121976005,
+      "learning_rate": 1.568539325842697e-05,
+      "loss": 0.3935,
+      "step": 349
+    },
+    {
+      "epoch": 0.39309279797837987,
+      "grad_norm": 0.5837147473027602,
+      "learning_rate": 1.5730337078651687e-05,
+      "loss": 0.3882,
+      "step": 350
+    },
+    {
+      "epoch": 0.39421592025831814,
+      "grad_norm": 0.59876417374738,
+      "learning_rate": 1.5775280898876405e-05,
+      "loss": 0.3687,
+      "step": 351
+    },
+    {
+      "epoch": 0.39533904253825636,
+      "grad_norm": 0.5577383109842206,
+      "learning_rate": 1.5820224719101127e-05,
+      "loss": 0.3838,
+      "step": 352
+    },
+    {
+      "epoch": 0.3964621648181946,
+      "grad_norm": 0.6253260037735091,
+      "learning_rate": 1.5865168539325845e-05,
+      "loss": 0.3988,
+      "step": 353
+    },
+    {
+      "epoch": 0.3975852870981328,
+      "grad_norm": 0.5709668147622707,
+      "learning_rate": 1.5910112359550563e-05,
+      "loss": 0.3797,
+      "step": 354
+    },
+    {
+      "epoch": 0.39870840937807106,
+      "grad_norm": 0.5766274168346207,
+      "learning_rate": 1.595505617977528e-05,
+      "loss": 0.4017,
+      "step": 355
+    },
+    {
+      "epoch": 0.3998315316580093,
+      "grad_norm": 0.5960841144559348,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 0.3784,
+      "step": 356
+    },
+    {
+      "epoch": 0.4009546539379475,
+      "grad_norm": 0.5637097286513578,
+      "learning_rate": 1.604494382022472e-05,
+      "loss": 0.3835,
+      "step": 357
+    },
+    {
+      "epoch": 0.4020777762178857,
+      "grad_norm": 0.621385534061573,
+      "learning_rate": 1.608988764044944e-05,
+      "loss": 0.379,
+      "step": 358
+    },
+    {
+      "epoch": 0.4032008984978239,
+      "grad_norm": 0.5804215759491298,
+      "learning_rate": 1.6134831460674157e-05,
+      "loss": 0.3717,
+      "step": 359
+    },
+    {
+      "epoch": 0.4043240207777622,
+      "grad_norm": 0.6850149117508036,
+      "learning_rate": 1.617977528089888e-05,
+      "loss": 0.3929,
+      "step": 360
+    },
+    {
+      "epoch": 0.4054471430577004,
+      "grad_norm": 0.5678267887232396,
+      "learning_rate": 1.6224719101123597e-05,
+      "loss": 0.3708,
+      "step": 361
+    },
+    {
+      "epoch": 0.4065702653376386,
+      "grad_norm": 0.5846240027794225,
+      "learning_rate": 1.626966292134832e-05,
+      "loss": 0.3806,
+      "step": 362
+    },
+    {
+      "epoch": 0.40769338761757684,
+      "grad_norm": 0.641861479794963,
+      "learning_rate": 1.6314606741573037e-05,
+      "loss": 0.3751,
+      "step": 363
+    },
+    {
+      "epoch": 0.4088165098975151,
+      "grad_norm": 0.551258860522498,
+      "learning_rate": 1.6359550561797755e-05,
+      "loss": 0.3847,
+      "step": 364
+    },
+    {
+      "epoch": 0.40993963217745333,
+      "grad_norm": 0.5436695988486194,
+      "learning_rate": 1.6404494382022473e-05,
+      "loss": 0.3588,
+      "step": 365
+    },
+    {
+      "epoch": 0.41106275445739154,
+      "grad_norm": 0.6273802594115025,
+      "learning_rate": 1.644943820224719e-05,
+      "loss": 0.3893,
+      "step": 366
+    },
+    {
+      "epoch": 0.41218587673732976,
+      "grad_norm": 0.5829276428140747,
+      "learning_rate": 1.649438202247191e-05,
+      "loss": 0.3761,
+      "step": 367
+    },
+    {
+      "epoch": 0.41330899901726803,
+      "grad_norm": 0.6510879644948344,
+      "learning_rate": 1.653932584269663e-05,
+      "loss": 0.3875,
+      "step": 368
+    },
+    {
+      "epoch": 0.41443212129720625,
+      "grad_norm": 0.5895996214987,
+      "learning_rate": 1.658426966292135e-05,
+      "loss": 0.3861,
+      "step": 369
+    },
+    {
+      "epoch": 0.41555524357714446,
+      "grad_norm": 0.5696237361887272,
+      "learning_rate": 1.662921348314607e-05,
+      "loss": 0.3733,
+      "step": 370
+    },
+    {
+      "epoch": 0.4166783658570827,
+      "grad_norm": 0.6452128769900847,
+      "learning_rate": 1.667415730337079e-05,
+      "loss": 0.377,
+      "step": 371
+    },
+    {
+      "epoch": 0.4178014881370209,
+      "grad_norm": 0.5933187468692044,
+      "learning_rate": 1.6719101123595507e-05,
+      "loss": 0.3727,
+      "step": 372
+    },
+    {
+      "epoch": 0.41892461041695916,
+      "grad_norm": 0.5956891237289303,
+      "learning_rate": 1.6764044943820225e-05,
+      "loss": 0.409,
+      "step": 373
+    },
+    {
+      "epoch": 0.4200477326968974,
+      "grad_norm": 0.6421959612844145,
+      "learning_rate": 1.6808988764044943e-05,
+      "loss": 0.4176,
+      "step": 374
+    },
+    {
+      "epoch": 0.4211708549768356,
+      "grad_norm": 0.5877612385670545,
+      "learning_rate": 1.6853932584269665e-05,
+      "loss": 0.3813,
+      "step": 375
+    },
+    {
+      "epoch": 0.4222939772567738,
+      "grad_norm": 0.5522655569800495,
+      "learning_rate": 1.6898876404494383e-05,
+      "loss": 0.3635,
+      "step": 376
+    },
+    {
+      "epoch": 0.4234170995367121,
+      "grad_norm": 0.570615445713445,
+      "learning_rate": 1.6943820224719105e-05,
+      "loss": 0.3464,
+      "step": 377
+    },
+    {
+      "epoch": 0.4245402218166503,
+      "grad_norm": 0.5335527636986674,
+      "learning_rate": 1.6988764044943823e-05,
+      "loss": 0.3519,
+      "step": 378
+    },
+    {
+      "epoch": 0.4256633440965885,
+      "grad_norm": 0.5633249413573972,
+      "learning_rate": 1.703370786516854e-05,
+      "loss": 0.3671,
+      "step": 379
+    },
+    {
+      "epoch": 0.42678646637652673,
+      "grad_norm": 0.5962150121982345,
+      "learning_rate": 1.707865168539326e-05,
+      "loss": 0.3638,
+      "step": 380
+    },
+    {
+      "epoch": 0.42790958865646495,
+      "grad_norm": 0.533130081409531,
+      "learning_rate": 1.7123595505617977e-05,
+      "loss": 0.3723,
+      "step": 381
+    },
+    {
+      "epoch": 0.4290327109364032,
+      "grad_norm": 0.5232079977615192,
+      "learning_rate": 1.71685393258427e-05,
+      "loss": 0.3377,
+      "step": 382
+    },
+    {
+      "epoch": 0.43015583321634143,
+      "grad_norm": 0.6669436739174833,
+      "learning_rate": 1.7213483146067417e-05,
+      "loss": 0.3649,
+      "step": 383
+    },
+    {
+      "epoch": 0.43127895549627965,
+      "grad_norm": 0.58553539025849,
+      "learning_rate": 1.7258426966292135e-05,
+      "loss": 0.3902,
+      "step": 384
+    },
+    {
+      "epoch": 0.43240207777621786,
+      "grad_norm": 0.5805175926211644,
+      "learning_rate": 1.7303370786516857e-05,
+      "loss": 0.3988,
+      "step": 385
+    },
+    {
+      "epoch": 0.43352520005615613,
+      "grad_norm": 0.622462053904498,
+      "learning_rate": 1.7348314606741575e-05,
+      "loss": 0.3726,
+      "step": 386
+    },
+    {
+      "epoch": 0.43464832233609435,
+      "grad_norm": 0.6586460393239919,
+      "learning_rate": 1.7393258426966293e-05,
+      "loss": 0.382,
+      "step": 387
+    },
+    {
+      "epoch": 0.43577144461603257,
+      "grad_norm": 0.503597593779572,
+      "learning_rate": 1.743820224719101e-05,
+      "loss": 0.3548,
+      "step": 388
+    },
+    {
+      "epoch": 0.4368945668959708,
+      "grad_norm": 0.6269680293566149,
+      "learning_rate": 1.7483146067415733e-05,
+      "loss": 0.3894,
+      "step": 389
+    },
+    {
+      "epoch": 0.43801768917590905,
+      "grad_norm": 0.5826375120126366,
+      "learning_rate": 1.752808988764045e-05,
+      "loss": 0.3814,
+      "step": 390
+    },
+    {
+      "epoch": 0.43914081145584727,
+      "grad_norm": 0.5730693678129642,
+      "learning_rate": 1.757303370786517e-05,
+      "loss": 0.4214,
+      "step": 391
+    },
+    {
+      "epoch": 0.4402639337357855,
+      "grad_norm": 0.5808702128043581,
+      "learning_rate": 1.7617977528089887e-05,
+      "loss": 0.354,
+      "step": 392
+    },
+    {
+      "epoch": 0.4413870560157237,
+      "grad_norm": 0.5974314280889683,
+      "learning_rate": 1.766292134831461e-05,
+      "loss": 0.3776,
+      "step": 393
+    },
+    {
+      "epoch": 0.4425101782956619,
+      "grad_norm": 0.5707766947857629,
+      "learning_rate": 1.7707865168539327e-05,
+      "loss": 0.3818,
+      "step": 394
+    },
+    {
+      "epoch": 0.4436333005756002,
+      "grad_norm": 0.5675159319972002,
+      "learning_rate": 1.7752808988764045e-05,
+      "loss": 0.4048,
+      "step": 395
+    },
+    {
+      "epoch": 0.4447564228555384,
+      "grad_norm": 0.5855000642824909,
+      "learning_rate": 1.7797752808988767e-05,
+      "loss": 0.3669,
+      "step": 396
+    },
+    {
+      "epoch": 0.4458795451354766,
+      "grad_norm": 0.5890500820148369,
+      "learning_rate": 1.7842696629213485e-05,
+      "loss": 0.3529,
+      "step": 397
+    },
+    {
+      "epoch": 0.44700266741541483,
+      "grad_norm": 0.5765910274102177,
+      "learning_rate": 1.7887640449438203e-05,
+      "loss": 0.4004,
+      "step": 398
+    },
+    {
+      "epoch": 0.4481257896953531,
+      "grad_norm": 0.5850026850384389,
+      "learning_rate": 1.793258426966292e-05,
+      "loss": 0.4024,
+      "step": 399
+    },
+    {
+      "epoch": 0.4492489119752913,
+      "grad_norm": 0.60842577325865,
+      "learning_rate": 1.7977528089887643e-05,
+      "loss": 0.362,
+      "step": 400
+    },
+    {
+      "epoch": 0.45037203425522954,
+      "grad_norm": 0.598714080261855,
+      "learning_rate": 1.802247191011236e-05,
+      "loss": 0.4056,
+      "step": 401
+    },
+    {
+      "epoch": 0.45149515653516775,
+      "grad_norm": 0.5460494368263945,
+      "learning_rate": 1.8067415730337083e-05,
+      "loss": 0.3563,
+      "step": 402
+    },
+    {
+      "epoch": 0.45261827881510597,
+      "grad_norm": 0.5338120861959073,
+      "learning_rate": 1.81123595505618e-05,
+      "loss": 0.3629,
+      "step": 403
+    },
+    {
+      "epoch": 0.45374140109504424,
+      "grad_norm": 0.5715606361300019,
+      "learning_rate": 1.815730337078652e-05,
+      "loss": 0.3717,
+      "step": 404
+    },
+    {
+      "epoch": 0.45486452337498245,
+      "grad_norm": 0.541068995106628,
+      "learning_rate": 1.8202247191011237e-05,
+      "loss": 0.3779,
+      "step": 405
+    },
+    {
+      "epoch": 0.45598764565492067,
+      "grad_norm": 0.5723001753094253,
+      "learning_rate": 1.8247191011235956e-05,
+      "loss": 0.3882,
+      "step": 406
+    },
+    {
+      "epoch": 0.4571107679348589,
+      "grad_norm": 0.5405763992724918,
+      "learning_rate": 1.8292134831460674e-05,
+      "loss": 0.3762,
+      "step": 407
+    },
+    {
+      "epoch": 0.45823389021479716,
+      "grad_norm": 0.5159698191034968,
+      "learning_rate": 1.8337078651685395e-05,
+      "loss": 0.377,
+      "step": 408
+    },
+    {
+      "epoch": 0.4593570124947354,
+      "grad_norm": 0.5354052943635401,
+      "learning_rate": 1.8382022471910113e-05,
+      "loss": 0.3735,
+      "step": 409
+    },
+    {
+      "epoch": 0.4604801347746736,
+      "grad_norm": 0.5507330621386544,
+      "learning_rate": 1.8426966292134835e-05,
+      "loss": 0.3941,
+      "step": 410
+    },
+    {
+      "epoch": 0.4616032570546118,
+      "grad_norm": 0.5470411737981918,
+      "learning_rate": 1.8471910112359553e-05,
+      "loss": 0.3788,
+      "step": 411
+    },
+    {
+      "epoch": 0.4627263793345501,
+      "grad_norm": 0.6279126208387347,
+      "learning_rate": 1.851685393258427e-05,
+      "loss": 0.3766,
+      "step": 412
+    },
+    {
+      "epoch": 0.4638495016144883,
+      "grad_norm": 0.5581331082608008,
+      "learning_rate": 1.856179775280899e-05,
+      "loss": 0.3548,
+      "step": 413
+    },
+    {
+      "epoch": 0.4649726238944265,
+      "grad_norm": 0.6339597817219623,
+      "learning_rate": 1.8606741573033708e-05,
+      "loss": 0.3928,
+      "step": 414
+    },
+    {
+      "epoch": 0.4660957461743647,
+      "grad_norm": 0.5885950731841739,
+      "learning_rate": 1.8651685393258426e-05,
+      "loss": 0.3834,
+      "step": 415
+    },
+    {
+      "epoch": 0.46721886845430294,
+      "grad_norm": 0.5517474243216776,
+      "learning_rate": 1.8696629213483147e-05,
+      "loss": 0.372,
+      "step": 416
+    },
+    {
+      "epoch": 0.4683419907342412,
+      "grad_norm": 0.5371558479053717,
+      "learning_rate": 1.8741573033707866e-05,
+      "loss": 0.3781,
+      "step": 417
+    },
+    {
+      "epoch": 0.4694651130141794,
+      "grad_norm": 0.5972062873696606,
+      "learning_rate": 1.8786516853932587e-05,
+      "loss": 0.4045,
+      "step": 418
+    },
+    {
+      "epoch": 0.47058823529411764,
+      "grad_norm": 0.5289075828695805,
+      "learning_rate": 1.8831460674157305e-05,
+      "loss": 0.3604,
+      "step": 419
+    },
+    {
+      "epoch": 0.47171135757405586,
+      "grad_norm": 0.5595781567840159,
+      "learning_rate": 1.8876404494382024e-05,
+      "loss": 0.3733,
+      "step": 420
+    },
+    {
+      "epoch": 0.4728344798539941,
+      "grad_norm": 0.5283542090574247,
+      "learning_rate": 1.8921348314606742e-05,
+      "loss": 0.3817,
+      "step": 421
+    },
+    {
+      "epoch": 0.47395760213393234,
+      "grad_norm": 0.5224861754856025,
+      "learning_rate": 1.8966292134831463e-05,
+      "loss": 0.3846,
+      "step": 422
+    },
+    {
+      "epoch": 0.47508072441387056,
+      "grad_norm": 0.5693567792868641,
+      "learning_rate": 1.901123595505618e-05,
+      "loss": 0.3784,
+      "step": 423
+    },
+    {
+      "epoch": 0.4762038466938088,
+      "grad_norm": 0.586099489925582,
+      "learning_rate": 1.90561797752809e-05,
+      "loss": 0.3762,
+      "step": 424
+    },
+    {
+      "epoch": 0.477326968973747,
+      "grad_norm": 0.5172936141285605,
+      "learning_rate": 1.910112359550562e-05,
+      "loss": 0.3833,
+      "step": 425
+    },
+    {
+      "epoch": 0.47845009125368526,
+      "grad_norm": 0.5666644649128297,
+      "learning_rate": 1.914606741573034e-05,
+      "loss": 0.3839,
+      "step": 426
+    },
+    {
+      "epoch": 0.4795732135336235,
+      "grad_norm": 0.5860548355924229,
+      "learning_rate": 1.9191011235955058e-05,
+      "loss": 0.3716,
+      "step": 427
+    },
+    {
+      "epoch": 0.4806963358135617,
+      "grad_norm": 0.5520174477409845,
+      "learning_rate": 1.9235955056179776e-05,
+      "loss": 0.3696,
+      "step": 428
+    },
+    {
+      "epoch": 0.4818194580934999,
+      "grad_norm": 0.5402453132910056,
+      "learning_rate": 1.9280898876404497e-05,
+      "loss": 0.3769,
+      "step": 429
+    },
+    {
+      "epoch": 0.4829425803734382,
+      "grad_norm": 0.5670157993103069,
+      "learning_rate": 1.9325842696629215e-05,
+      "loss": 0.3546,
+      "step": 430
+    },
+    {
+      "epoch": 0.4840657026533764,
+      "grad_norm": 0.5378630874757251,
+      "learning_rate": 1.9370786516853934e-05,
+      "loss": 0.3831,
+      "step": 431
+    },
+    {
+      "epoch": 0.4851888249333146,
+      "grad_norm": 0.5178321480341342,
+      "learning_rate": 1.9415730337078652e-05,
+      "loss": 0.3796,
+      "step": 432
+    },
+    {
+      "epoch": 0.4863119472132528,
+      "grad_norm": 0.5500034972413501,
+      "learning_rate": 1.9460674157303373e-05,
+      "loss": 0.4023,
+      "step": 433
+    },
+    {
+      "epoch": 0.4874350694931911,
+      "grad_norm": 0.4790912474912375,
+      "learning_rate": 1.950561797752809e-05,
+      "loss": 0.3695,
+      "step": 434
+    },
+    {
+      "epoch": 0.4885581917731293,
+      "grad_norm": 0.520106712842702,
+      "learning_rate": 1.955056179775281e-05,
+      "loss": 0.3907,
+      "step": 435
+    },
+    {
+      "epoch": 0.48968131405306753,
+      "grad_norm": 0.5270381224070033,
+      "learning_rate": 1.959550561797753e-05,
+      "loss": 0.3746,
+      "step": 436
+    },
+    {
+      "epoch": 0.49080443633300574,
+      "grad_norm": 0.5585640386718159,
+      "learning_rate": 1.964044943820225e-05,
+      "loss": 0.3968,
+      "step": 437
+    },
+    {
+      "epoch": 0.49192755861294396,
+      "grad_norm": 0.5212838454555632,
+      "learning_rate": 1.9685393258426968e-05,
+      "loss": 0.3848,
+      "step": 438
+    },
+    {
+      "epoch": 0.49305068089288223,
+      "grad_norm": 0.5223182787955911,
+      "learning_rate": 1.9730337078651686e-05,
+      "loss": 0.3871,
+      "step": 439
+    },
+    {
+      "epoch": 0.49417380317282045,
+      "grad_norm": 0.5014928363223362,
+      "learning_rate": 1.9775280898876404e-05,
+      "loss": 0.3749,
+      "step": 440
+    },
+    {
+      "epoch": 0.49529692545275866,
+      "grad_norm": 0.5448751880848072,
+      "learning_rate": 1.9820224719101126e-05,
+      "loss": 0.4037,
+      "step": 441
+    },
+    {
+      "epoch": 0.4964200477326969,
+      "grad_norm": 0.5115396988679667,
+      "learning_rate": 1.9865168539325844e-05,
+      "loss": 0.3578,
+      "step": 442
+    },
+    {
+      "epoch": 0.49754317001263515,
+      "grad_norm": 0.5226126289636163,
+      "learning_rate": 1.9910112359550565e-05,
+      "loss": 0.3813,
+      "step": 443
+    },
+    {
+      "epoch": 0.49866629229257337,
+      "grad_norm": 0.5651234942385593,
+      "learning_rate": 1.9955056179775283e-05,
+      "loss": 0.4008,
+      "step": 444
+    },
+    {
+      "epoch": 0.4997894145725116,
+      "grad_norm": 0.5414290026509725,
+      "learning_rate": 2e-05,
+      "loss": 0.3634,
+      "step": 445
+    },
+    {
+      "epoch": 0.5009125368524499,
+      "grad_norm": 0.5062706466010933,
+      "learning_rate": 1.9999996923444976e-05,
+      "loss": 0.3679,
+      "step": 446
+    },
+    {
+      "epoch": 0.502035659132388,
+      "grad_norm": 0.6284172666370126,
+      "learning_rate": 1.9999987693781806e-05,
+      "loss": 0.39,
+      "step": 447
+    },
+    {
+      "epoch": 0.5031587814123263,
+      "grad_norm": 0.523476357886049,
+      "learning_rate": 1.9999972311016158e-05,
+      "loss": 0.3851,
+      "step": 448
+    },
+    {
+      "epoch": 0.5042819036922644,
+      "grad_norm": 0.5198224598673091,
+      "learning_rate": 1.99999507751575e-05,
+      "loss": 0.3722,
+      "step": 449
+    },
+    {
+      "epoch": 0.5054050259722027,
+      "grad_norm": 0.5990656789702519,
+      "learning_rate": 1.999992308621909e-05,
+      "loss": 0.3961,
+      "step": 450
+    },
+    {
+      "epoch": 0.506528148252141,
+      "grad_norm": 0.5337336925829737,
+      "learning_rate": 1.9999889244217957e-05,
+      "loss": 0.3661,
+      "step": 451
+    },
+    {
+      "epoch": 0.5076512705320791,
+      "grad_norm": 0.5265236086660914,
+      "learning_rate": 1.9999849249174933e-05,
+      "loss": 0.3763,
+      "step": 452
+    },
+    {
+      "epoch": 0.5087743928120174,
+      "grad_norm": 0.6072251444919325,
+      "learning_rate": 1.9999803101114622e-05,
+      "loss": 0.3702,
+      "step": 453
+    },
+    {
+      "epoch": 0.5098975150919557,
+      "grad_norm": 0.6242928399695532,
+      "learning_rate": 1.9999750800065415e-05,
+      "loss": 0.3953,
+      "step": 454
+    },
+    {
+      "epoch": 0.5110206373718938,
+      "grad_norm": 0.5179538772395289,
+      "learning_rate": 1.9999692346059504e-05,
+      "loss": 0.3873,
+      "step": 455
+    },
+    {
+      "epoch": 0.5121437596518321,
+      "grad_norm": 0.5639243942734614,
+      "learning_rate": 1.9999627739132847e-05,
+      "loss": 0.3975,
+      "step": 456
+    },
+    {
+      "epoch": 0.5132668819317703,
+      "grad_norm": 0.6222398095764557,
+      "learning_rate": 1.9999556979325203e-05,
+      "loss": 0.3826,
+      "step": 457
+    },
+    {
+      "epoch": 0.5143900042117086,
+      "grad_norm": 0.5381548600095061,
+      "learning_rate": 1.999948006668011e-05,
+      "loss": 0.4006,
+      "step": 458
+    },
+    {
+      "epoch": 0.5155131264916468,
+      "grad_norm": 0.5460529769570657,
+      "learning_rate": 1.999939700124489e-05,
+      "loss": 0.3779,
+      "step": 459
+    },
+    {
+      "epoch": 0.516636248771585,
+      "grad_norm": 0.5285009639724508,
+      "learning_rate": 1.999930778307066e-05,
+      "loss": 0.4117,
+      "step": 460
+    },
+    {
+      "epoch": 0.5177593710515233,
+      "grad_norm": 0.588582651422445,
+      "learning_rate": 1.9999212412212313e-05,
+      "loss": 0.4135,
+      "step": 461
+    },
+    {
+      "epoch": 0.5188824933314614,
+      "grad_norm": 0.5262946007341981,
+      "learning_rate": 1.9999110888728527e-05,
+      "loss": 0.3836,
+      "step": 462
+    },
+    {
+      "epoch": 0.5200056156113997,
+      "grad_norm": 0.5074018080150626,
+      "learning_rate": 1.999900321268178e-05,
+      "loss": 0.3879,
+      "step": 463
+    },
+    {
+      "epoch": 0.521128737891338,
+      "grad_norm": 0.5699895099551994,
+      "learning_rate": 1.9998889384138324e-05,
+      "loss": 0.3994,
+      "step": 464
+    },
+    {
+      "epoch": 0.5222518601712761,
+      "grad_norm": 0.5488889184993216,
+      "learning_rate": 1.9998769403168195e-05,
+      "loss": 0.3787,
+      "step": 465
+    },
+    {
+      "epoch": 0.5233749824512144,
+      "grad_norm": 0.587601256959231,
+      "learning_rate": 1.9998643269845225e-05,
+      "loss": 0.4028,
+      "step": 466
+    },
+    {
+      "epoch": 0.5244981047311527,
+      "grad_norm": 0.5470076750391482,
+      "learning_rate": 1.9998510984247015e-05,
+      "loss": 0.3869,
+      "step": 467
+    },
+    {
+      "epoch": 0.5256212270110908,
+      "grad_norm": 0.5490698566612296,
+      "learning_rate": 1.9998372546454973e-05,
+      "loss": 0.4069,
+      "step": 468
+    },
+    {
+      "epoch": 0.5267443492910291,
+      "grad_norm": 0.5377412654267874,
+      "learning_rate": 1.999822795655427e-05,
+      "loss": 0.3622,
+      "step": 469
+    },
+    {
+      "epoch": 0.5278674715709673,
+      "grad_norm": 0.5398403722418982,
+      "learning_rate": 1.9998077214633884e-05,
+      "loss": 0.3785,
+      "step": 470
+    },
+    {
+      "epoch": 0.5289905938509055,
+      "grad_norm": 0.49636950695396015,
+      "learning_rate": 1.999792032078656e-05,
+      "loss": 0.3526,
+      "step": 471
+    },
+    {
+      "epoch": 0.5301137161308438,
+      "grad_norm": 0.6168098343297038,
+      "learning_rate": 1.9997757275108847e-05,
+      "loss": 0.3947,
+      "step": 472
+    },
+    {
+      "epoch": 0.531236838410782,
+      "grad_norm": 0.5815523034498505,
+      "learning_rate": 1.9997588077701057e-05,
+      "loss": 0.3776,
+      "step": 473
+    },
+    {
+      "epoch": 0.5323599606907202,
+      "grad_norm": 0.5214146087899991,
+      "learning_rate": 1.999741272866731e-05,
+      "loss": 0.3685,
+      "step": 474
+    },
+    {
+      "epoch": 0.5334830829706584,
+      "grad_norm": 0.554167072491176,
+      "learning_rate": 1.9997231228115487e-05,
+      "loss": 0.3987,
+      "step": 475
+    },
+    {
+      "epoch": 0.5346062052505967,
+      "grad_norm": 0.5442738139350299,
+      "learning_rate": 1.999704357615728e-05,
+      "loss": 0.3824,
+      "step": 476
+    },
+    {
+      "epoch": 0.5357293275305349,
+      "grad_norm": 0.5157448841129373,
+      "learning_rate": 1.9996849772908144e-05,
+      "loss": 0.3728,
+      "step": 477
+    },
+    {
+      "epoch": 0.5368524498104731,
+      "grad_norm": 0.5376140918675025,
+      "learning_rate": 1.9996649818487336e-05,
+      "loss": 0.409,
+      "step": 478
+    },
+    {
+      "epoch": 0.5379755720904114,
+      "grad_norm": 0.549248833785917,
+      "learning_rate": 1.9996443713017883e-05,
+      "loss": 0.3852,
+      "step": 479
+    },
+    {
+      "epoch": 0.5390986943703495,
+      "grad_norm": 0.5236716683541965,
+      "learning_rate": 1.999623145662661e-05,
+      "loss": 0.3798,
+      "step": 480
+    },
+    {
+      "epoch": 0.5402218166502878,
+      "grad_norm": 0.609473300182195,
+      "learning_rate": 1.9996013049444117e-05,
+      "loss": 0.3764,
+      "step": 481
+    },
+    {
+      "epoch": 0.5413449389302261,
+      "grad_norm": 0.5355644461303796,
+      "learning_rate": 1.9995788491604797e-05,
+      "loss": 0.3574,
+      "step": 482
+    },
+    {
+      "epoch": 0.5424680612101642,
+      "grad_norm": 0.5580759406548179,
+      "learning_rate": 1.9995557783246814e-05,
+      "loss": 0.3704,
+      "step": 483
+    },
+    {
+      "epoch": 0.5435911834901025,
+      "grad_norm": 0.5531806767952361,
+      "learning_rate": 1.9995320924512135e-05,
+      "loss": 0.3939,
+      "step": 484
+    },
+    {
+      "epoch": 0.5447143057700408,
+      "grad_norm": 0.5834026851989711,
+      "learning_rate": 1.9995077915546496e-05,
+      "loss": 0.3931,
+      "step": 485
+    },
+    {
+      "epoch": 0.5458374280499789,
+      "grad_norm": 0.5591794205487529,
+      "learning_rate": 1.9994828756499423e-05,
+      "loss": 0.4071,
+      "step": 486
+    },
+    {
+      "epoch": 0.5469605503299172,
+      "grad_norm": 0.5322788134249687,
+      "learning_rate": 1.9994573447524233e-05,
+      "loss": 0.3704,
+      "step": 487
+    },
+    {
+      "epoch": 0.5480836726098554,
+      "grad_norm": 0.6160509443550839,
+      "learning_rate": 1.999431198877801e-05,
+      "loss": 0.3668,
+      "step": 488
+    },
+    {
+      "epoch": 0.5492067948897936,
+      "grad_norm": 0.4808621346087435,
+      "learning_rate": 1.9994044380421642e-05,
+      "loss": 0.359,
+      "step": 489
+    },
+    {
+      "epoch": 0.5503299171697319,
+      "grad_norm": 0.5940190325126405,
+      "learning_rate": 1.9993770622619784e-05,
+      "loss": 0.3975,
+      "step": 490
+    },
+    {
+      "epoch": 0.5514530394496701,
+      "grad_norm": 0.5314215977553695,
+      "learning_rate": 1.9993490715540888e-05,
+      "loss": 0.3619,
+      "step": 491
+    },
+    {
+      "epoch": 0.5525761617296083,
+      "grad_norm": 0.5286779855947581,
+      "learning_rate": 1.9993204659357176e-05,
+      "loss": 0.3624,
+      "step": 492
+    },
+    {
+      "epoch": 0.5536992840095465,
+      "grad_norm": 0.6157841936932821,
+      "learning_rate": 1.9992912454244677e-05,
+      "loss": 0.3856,
+      "step": 493
+    },
+    {
+      "epoch": 0.5548224062894848,
+      "grad_norm": 0.5146925583224419,
+      "learning_rate": 1.999261410038317e-05,
+      "loss": 0.3889,
+      "step": 494
+    },
+    {
+      "epoch": 0.555945528569423,
+      "grad_norm": 0.5899681817511568,
+      "learning_rate": 1.9992309597956244e-05,
+      "loss": 0.3872,
+      "step": 495
+    },
+    {
+      "epoch": 0.5570686508493612,
+      "grad_norm": 0.5514658907626057,
+      "learning_rate": 1.999199894715126e-05,
+      "loss": 0.367,
+      "step": 496
+    },
+    {
+      "epoch": 0.5581917731292995,
+      "grad_norm": 0.5727284572822289,
+      "learning_rate": 1.9991682148159372e-05,
+      "loss": 0.3855,
+      "step": 497
+    },
+    {
+      "epoch": 0.5593148954092377,
+      "grad_norm": 0.6630678363436415,
+      "learning_rate": 1.9991359201175503e-05,
+      "loss": 0.3747,
+      "step": 498
+    },
+    {
+      "epoch": 0.5604380176891759,
+      "grad_norm": 0.5168218883882756,
+      "learning_rate": 1.9991030106398367e-05,
+      "loss": 0.3665,
+      "step": 499
+    },
+    {
+      "epoch": 0.5615611399691142,
+      "grad_norm": 0.5285711442774218,
+      "learning_rate": 1.999069486403046e-05,
+      "loss": 0.3661,
+      "step": 500
+    },
+    {
+      "epoch": 0.5626842622490523,
+      "grad_norm": 0.612117315690109,
+      "learning_rate": 1.999035347427806e-05,
+      "loss": 0.4039,
+      "step": 501
+    },
+    {
+      "epoch": 0.5638073845289906,
+      "grad_norm": 0.5866977447253838,
+      "learning_rate": 1.999000593735123e-05,
+      "loss": 0.4256,
+      "step": 502
+    },
+    {
+      "epoch": 0.5649305068089289,
+      "grad_norm": 0.5670878177199548,
+      "learning_rate": 1.998965225346381e-05,
+      "loss": 0.4193,
+      "step": 503
+    },
+    {
+      "epoch": 0.566053629088867,
+      "grad_norm": 0.5585226536761726,
+      "learning_rate": 1.998929242283343e-05,
+      "loss": 0.3909,
+      "step": 504
+    },
+    {
+      "epoch": 0.5671767513688053,
+      "grad_norm": 0.5483834024577466,
+      "learning_rate": 1.9988926445681495e-05,
+      "loss": 0.3914,
+      "step": 505
+    },
+    {
+      "epoch": 0.5682998736487435,
+      "grad_norm": 0.48877542079307956,
+      "learning_rate": 1.9988554322233188e-05,
+      "loss": 0.3718,
+      "step": 506
+    },
+    {
+      "epoch": 0.5694229959286817,
+      "grad_norm": 0.5685952858138967,
+      "learning_rate": 1.9988176052717495e-05,
+      "loss": 0.3989,
+      "step": 507
+    },
+    {
+      "epoch": 0.57054611820862,
+      "grad_norm": 0.5234278520668304,
+      "learning_rate": 1.9987791637367157e-05,
+      "loss": 0.3742,
+      "step": 508
+    },
+    {
+      "epoch": 0.5716692404885582,
+      "grad_norm": 0.5460853796115502,
+      "learning_rate": 1.9987401076418717e-05,
+      "loss": 0.402,
+      "step": 509
+    },
+    {
+      "epoch": 0.5727923627684964,
+      "grad_norm": 0.5455839187416055,
+      "learning_rate": 1.9987004370112487e-05,
+      "loss": 0.3641,
+      "step": 510
+    },
+    {
+      "epoch": 0.5739154850484347,
+      "grad_norm": 0.5056945010695153,
+      "learning_rate": 1.9986601518692567e-05,
+      "loss": 0.375,
+      "step": 511
+    },
+    {
+      "epoch": 0.5750386073283729,
+      "grad_norm": 0.5208870388950549,
+      "learning_rate": 1.9986192522406835e-05,
+      "loss": 0.3714,
+      "step": 512
+    },
+    {
+      "epoch": 0.5761617296083111,
+      "grad_norm": 0.5358983435756814,
+      "learning_rate": 1.998577738150695e-05,
+      "loss": 0.3822,
+      "step": 513
+    },
+    {
+      "epoch": 0.5772848518882493,
+      "grad_norm": 0.5399478578334925,
+      "learning_rate": 1.9985356096248357e-05,
+      "loss": 0.3978,
+      "step": 514
+    },
+    {
+      "epoch": 0.5784079741681876,
+      "grad_norm": 0.5276672395265529,
+      "learning_rate": 1.998492866689027e-05,
+      "loss": 0.351,
+      "step": 515
+    },
+    {
+      "epoch": 0.5795310964481258,
+      "grad_norm": 0.5308333056433986,
+      "learning_rate": 1.9984495093695696e-05,
+      "loss": 0.4038,
+      "step": 516
+    },
+    {
+      "epoch": 0.580654218728064,
+      "grad_norm": 0.4895548996663825,
+      "learning_rate": 1.9984055376931414e-05,
+      "loss": 0.349,
+      "step": 517
+    },
+    {
+      "epoch": 0.5817773410080023,
+      "grad_norm": 0.5250328463562076,
+      "learning_rate": 1.9983609516867993e-05,
+      "loss": 0.3766,
+      "step": 518
+    },
+    {
+      "epoch": 0.5829004632879404,
+      "grad_norm": 0.5371115914639484,
+      "learning_rate": 1.9983157513779768e-05,
+      "loss": 0.3576,
+      "step": 519
+    },
+    {
+      "epoch": 0.5840235855678787,
+      "grad_norm": 0.5127425095139194,
+      "learning_rate": 1.998269936794487e-05,
+      "loss": 0.3771,
+      "step": 520
+    },
+    {
+      "epoch": 0.585146707847817,
+      "grad_norm": 0.5327318689065801,
+      "learning_rate": 1.9982235079645192e-05,
+      "loss": 0.3629,
+      "step": 521
+    },
+    {
+      "epoch": 0.5862698301277551,
+      "grad_norm": 0.5158425879035294,
+      "learning_rate": 1.998176464916642e-05,
+      "loss": 0.3513,
+      "step": 522
+    },
+    {
+      "epoch": 0.5873929524076934,
+      "grad_norm": 0.5261415497468315,
+      "learning_rate": 1.998128807679802e-05,
+      "loss": 0.3895,
+      "step": 523
+    },
+    {
+      "epoch": 0.5885160746876316,
+      "grad_norm": 0.5330293259802309,
+      "learning_rate": 1.998080536283322e-05,
+      "loss": 0.3801,
+      "step": 524
+    },
+    {
+      "epoch": 0.5896391969675698,
+      "grad_norm": 0.5328558914321503,
+      "learning_rate": 1.998031650756905e-05,
+      "loss": 0.3746,
+      "step": 525
+    },
+    {
+      "epoch": 0.5907623192475081,
+      "grad_norm": 0.5390463848256605,
+      "learning_rate": 1.9979821511306308e-05,
+      "loss": 0.3817,
+      "step": 526
+    },
+    {
+      "epoch": 0.5918854415274463,
+      "grad_norm": 0.5431335082497138,
+      "learning_rate": 1.9979320374349564e-05,
+      "loss": 0.3908,
+      "step": 527
+    },
+    {
+      "epoch": 0.5930085638073845,
+      "grad_norm": 0.5101979024125878,
+      "learning_rate": 1.997881309700717e-05,
+      "loss": 0.3777,
+      "step": 528
+    },
+    {
+      "epoch": 0.5941316860873228,
+      "grad_norm": 0.5209616551352008,
+      "learning_rate": 1.9978299679591276e-05,
+      "loss": 0.385,
+      "step": 529
+    },
+    {
+      "epoch": 0.595254808367261,
+      "grad_norm": 0.518301945363611,
+      "learning_rate": 1.9977780122417778e-05,
+      "loss": 0.3927,
+      "step": 530
+    },
+    {
+      "epoch": 0.5963779306471992,
+      "grad_norm": 0.5064656297120799,
+      "learning_rate": 1.997725442580637e-05,
+      "loss": 0.366,
+      "step": 531
+    },
+    {
+      "epoch": 0.5975010529271374,
+      "grad_norm": 0.560763233769782,
+      "learning_rate": 1.997672259008052e-05,
+      "loss": 0.3839,
+      "step": 532
+    },
+    {
+      "epoch": 0.5986241752070757,
+      "grad_norm": 0.5185441179220639,
+      "learning_rate": 1.997618461556747e-05,
+      "loss": 0.3649,
+      "step": 533
+    },
+    {
+      "epoch": 0.5997472974870139,
+      "grad_norm": 0.5116157155704527,
+      "learning_rate": 1.9975640502598243e-05,
+      "loss": 0.4155,
+      "step": 534
+    },
+    {
+      "epoch": 0.6008704197669521,
+      "grad_norm": 0.5484375904037471,
+      "learning_rate": 1.9975090251507637e-05,
+      "loss": 0.3893,
+      "step": 535
+    },
+    {
+      "epoch": 0.6019935420468904,
+      "grad_norm": 0.5194411036484662,
+      "learning_rate": 1.9974533862634234e-05,
+      "loss": 0.3708,
+      "step": 536
+    },
+    {
+      "epoch": 0.6031166643268285,
+      "grad_norm": 0.5084933191055989,
+      "learning_rate": 1.9973971336320374e-05,
+      "loss": 0.3618,
+      "step": 537
+    },
+    {
+      "epoch": 0.6042397866067668,
+      "grad_norm": 0.500283964355109,
+      "learning_rate": 1.9973402672912196e-05,
+      "loss": 0.3773,
+      "step": 538
+    },
+    {
+      "epoch": 0.6053629088867051,
+      "grad_norm": 0.5043471751080929,
+      "learning_rate": 1.9972827872759598e-05,
+      "loss": 0.3706,
+      "step": 539
+    },
+    {
+      "epoch": 0.6064860311666432,
+      "grad_norm": 0.48051148948387623,
+      "learning_rate": 1.9972246936216268e-05,
+      "loss": 0.369,
+      "step": 540
+    },
+    {
+      "epoch": 0.6076091534465815,
+      "grad_norm": 0.5037545077388655,
+      "learning_rate": 1.9971659863639657e-05,
+      "loss": 0.3777,
+      "step": 541
+    },
+    {
+      "epoch": 0.6087322757265198,
+      "grad_norm": 0.5202846079968864,
+      "learning_rate": 1.9971066655390997e-05,
+      "loss": 0.3878,
+      "step": 542
+    },
+    {
+      "epoch": 0.6098553980064579,
+      "grad_norm": 0.5196509635056363,
+      "learning_rate": 1.99704673118353e-05,
+      "loss": 0.3786,
+      "step": 543
+    },
+    {
+      "epoch": 0.6109785202863962,
+      "grad_norm": 0.5100880924606735,
+      "learning_rate": 1.996986183334134e-05,
+      "loss": 0.3923,
+      "step": 544
+    },
+    {
+      "epoch": 0.6121016425663344,
+      "grad_norm": 0.5374754446994799,
+      "learning_rate": 1.9969250220281687e-05,
+      "loss": 0.4019,
+      "step": 545
+    },
+    {
+      "epoch": 0.6132247648462726,
+      "grad_norm": 0.5405136047123903,
+      "learning_rate": 1.9968632473032663e-05,
+      "loss": 0.3903,
+      "step": 546
+    },
+    {
+      "epoch": 0.6143478871262109,
+      "grad_norm": 0.5377137297729571,
+      "learning_rate": 1.996800859197438e-05,
+      "loss": 0.3847,
+      "step": 547
+    },
+    {
+      "epoch": 0.6154710094061491,
+      "grad_norm": 0.5280413079303611,
+      "learning_rate": 1.996737857749072e-05,
+      "loss": 0.3846,
+      "step": 548
+    },
+    {
+      "epoch": 0.6165941316860873,
+      "grad_norm": 0.4698879134163189,
+      "learning_rate": 1.996674242996933e-05,
+      "loss": 0.3624,
+      "step": 549
+    },
+    {
+      "epoch": 0.6177172539660255,
+      "grad_norm": 0.5693615931340346,
+      "learning_rate": 1.9966100149801648e-05,
+      "loss": 0.3809,
+      "step": 550
+    },
+    {
+      "epoch": 0.6188403762459638,
+      "grad_norm": 0.5390167928723439,
+      "learning_rate": 1.9965451737382872e-05,
+      "loss": 0.3944,
+      "step": 551
+    },
+    {
+      "epoch": 0.619963498525902,
+      "grad_norm": 0.4687981254942727,
+      "learning_rate": 1.9964797193111973e-05,
+      "loss": 0.3486,
+      "step": 552
+    },
+    {
+      "epoch": 0.6210866208058402,
+      "grad_norm": 0.6002959050862824,
+      "learning_rate": 1.9964136517391708e-05,
+      "loss": 0.427,
+      "step": 553
+    },
+    {
+      "epoch": 0.6222097430857785,
+      "grad_norm": 0.5424155203819168,
+      "learning_rate": 1.9963469710628592e-05,
+      "loss": 0.3864,
+      "step": 554
+    },
+    {
+      "epoch": 0.6233328653657166,
+      "grad_norm": 0.4805110497258522,
+      "learning_rate": 1.996279677323292e-05,
+      "loss": 0.3431,
+      "step": 555
+    },
+    {
+      "epoch": 0.6244559876456549,
+      "grad_norm": 0.5468592773183889,
+      "learning_rate": 1.9962117705618757e-05,
+      "loss": 0.3916,
+      "step": 556
+    },
+    {
+      "epoch": 0.6255791099255932,
+      "grad_norm": 0.5974551409729048,
+      "learning_rate": 1.9961432508203938e-05,
+      "loss": 0.3703,
+      "step": 557
+    },
+    {
+      "epoch": 0.6267022322055313,
+      "grad_norm": 0.46694957713893454,
+      "learning_rate": 1.996074118141008e-05,
+      "loss": 0.36,
+      "step": 558
+    },
+    {
+      "epoch": 0.6278253544854696,
+      "grad_norm": 0.5203301666877311,
+      "learning_rate": 1.9960043725662558e-05,
+      "loss": 0.3661,
+      "step": 559
+    },
+    {
+      "epoch": 0.6289484767654079,
+      "grad_norm": 0.5315196144666886,
+      "learning_rate": 1.995934014139053e-05,
+      "loss": 0.3878,
+      "step": 560
+    },
+    {
+      "epoch": 0.630071599045346,
+      "grad_norm": 0.4726474224175523,
+      "learning_rate": 1.9958630429026912e-05,
+      "loss": 0.3549,
+      "step": 561
+    },
+    {
+      "epoch": 0.6311947213252843,
+      "grad_norm": 0.5265877427590347,
+      "learning_rate": 1.9957914589008405e-05,
+      "loss": 0.372,
+      "step": 562
+    },
+    {
+      "epoch": 0.6323178436052225,
+      "grad_norm": 0.5108540554457581,
+      "learning_rate": 1.9957192621775466e-05,
+      "loss": 0.3709,
+      "step": 563
+    },
+    {
+      "epoch": 0.6334409658851607,
+      "grad_norm": 0.5290487085943824,
+      "learning_rate": 1.9956464527772334e-05,
+      "loss": 0.3884,
+      "step": 564
+    },
+    {
+      "epoch": 0.634564088165099,
+      "grad_norm": 0.5281270652784177,
+      "learning_rate": 1.9955730307447015e-05,
+      "loss": 0.3834,
+      "step": 565
+    },
+    {
+      "epoch": 0.6356872104450372,
+      "grad_norm": 0.5085482495756416,
+      "learning_rate": 1.9954989961251276e-05,
+      "loss": 0.3732,
+      "step": 566
+    },
+    {
+      "epoch": 0.6368103327249754,
+      "grad_norm": 0.4783417140693344,
+      "learning_rate": 1.9954243489640667e-05,
+      "loss": 0.3696,
+      "step": 567
+    },
+    {
+      "epoch": 0.6379334550049136,
+      "grad_norm": 0.5374355693068502,
+      "learning_rate": 1.995349089307449e-05,
+      "loss": 0.3755,
+      "step": 568
+    },
+    {
+      "epoch": 0.6390565772848519,
+      "grad_norm": 0.5020426636675567,
+      "learning_rate": 1.995273217201584e-05,
+      "loss": 0.3667,
+      "step": 569
+    },
+    {
+      "epoch": 0.6401796995647902,
+      "grad_norm": 0.48221702122344806,
+      "learning_rate": 1.995196732693156e-05,
+      "loss": 0.3669,
+      "step": 570
+    },
+    {
+      "epoch": 0.6413028218447283,
+      "grad_norm": 0.4948631937332312,
+      "learning_rate": 1.9951196358292266e-05,
+      "loss": 0.3815,
+      "step": 571
+    },
+    {
+      "epoch": 0.6424259441246666,
+      "grad_norm": 0.5149576535985988,
+      "learning_rate": 1.9950419266572344e-05,
+      "loss": 0.3748,
+      "step": 572
+    },
+    {
+      "epoch": 0.6435490664046049,
+      "grad_norm": 0.5077161706598761,
+      "learning_rate": 1.994963605224995e-05,
+      "loss": 0.3736,
+      "step": 573
+    },
+    {
+      "epoch": 0.644672188684543,
+      "grad_norm": 0.4930605863816736,
+      "learning_rate": 1.9948846715807e-05,
+      "loss": 0.3698,
+      "step": 574
+    },
+    {
+      "epoch": 0.6457953109644813,
+      "grad_norm": 0.514614667391616,
+      "learning_rate": 1.9948051257729184e-05,
+      "loss": 0.3736,
+      "step": 575
+    },
+    {
+      "epoch": 0.6469184332444194,
+      "grad_norm": 0.5176250628419765,
+      "learning_rate": 1.9947249678505956e-05,
+      "loss": 0.3693,
+      "step": 576
+    },
+    {
+      "epoch": 0.6480415555243577,
+      "grad_norm": 0.45913174362011855,
+      "learning_rate": 1.994644197863054e-05,
+      "loss": 0.3499,
+      "step": 577
+    },
+    {
+      "epoch": 0.649164677804296,
+      "grad_norm": 0.49883541686787625,
+      "learning_rate": 1.994562815859991e-05,
+      "loss": 0.362,
+      "step": 578
+    },
+    {
+      "epoch": 0.6502878000842341,
+      "grad_norm": 0.5146434352456531,
+      "learning_rate": 1.9944808218914833e-05,
+      "loss": 0.3906,
+      "step": 579
+    },
+    {
+      "epoch": 0.6514109223641724,
+      "grad_norm": 0.5066808040702572,
+      "learning_rate": 1.9943982160079823e-05,
+      "loss": 0.3853,
+      "step": 580
+    },
+    {
+      "epoch": 0.6525340446441106,
+      "grad_norm": 0.49848710491518,
+      "learning_rate": 1.9943149982603155e-05,
+      "loss": 0.3985,
+      "step": 581
+    },
+    {
+      "epoch": 0.6536571669240488,
+      "grad_norm": 0.4878251052204487,
+      "learning_rate": 1.994231168699689e-05,
+      "loss": 0.3714,
+      "step": 582
+    },
+    {
+      "epoch": 0.6547802892039871,
+      "grad_norm": 0.4758786623666332,
+      "learning_rate": 1.9941467273776832e-05,
+      "loss": 0.3946,
+      "step": 583
+    },
+    {
+      "epoch": 0.6559034114839253,
+      "grad_norm": 0.4983367896874346,
+      "learning_rate": 1.9940616743462557e-05,
+      "loss": 0.3722,
+      "step": 584
+    },
+    {
+      "epoch": 0.6570265337638636,
+      "grad_norm": 0.5075829148442156,
+      "learning_rate": 1.9939760096577408e-05,
+      "loss": 0.3843,
+      "step": 585
+    },
+    {
+      "epoch": 0.6581496560438018,
+      "grad_norm": 0.5301054416844904,
+      "learning_rate": 1.9938897333648492e-05,
+      "loss": 0.3997,
+      "step": 586
+    },
+    {
+      "epoch": 0.65927277832374,
+      "grad_norm": 0.5148031126628784,
+      "learning_rate": 1.9938028455206674e-05,
+      "loss": 0.3737,
+      "step": 587
+    },
+    {
+      "epoch": 0.6603959006036783,
+      "grad_norm": 0.46714827561279637,
+      "learning_rate": 1.9937153461786585e-05,
+      "loss": 0.3431,
+      "step": 588
+    },
+    {
+      "epoch": 0.6615190228836164,
+      "grad_norm": 0.5177398408210456,
+      "learning_rate": 1.9936272353926616e-05,
+      "loss": 0.3666,
+      "step": 589
+    },
+    {
+      "epoch": 0.6626421451635547,
+      "grad_norm": 0.47039452637574947,
+      "learning_rate": 1.993538513216892e-05,
+      "loss": 0.3635,
+      "step": 590
+    },
+    {
+      "epoch": 0.663765267443493,
+      "grad_norm": 0.5093614259546847,
+      "learning_rate": 1.9934491797059425e-05,
+      "loss": 0.3915,
+      "step": 591
+    },
+    {
+      "epoch": 0.6648883897234311,
+      "grad_norm": 0.5366679460711811,
+      "learning_rate": 1.99335923491478e-05,
+      "loss": 0.398,
+      "step": 592
+    },
+    {
+      "epoch": 0.6660115120033694,
+      "grad_norm": 0.5082350993851118,
+      "learning_rate": 1.993268678898749e-05,
+      "loss": 0.3938,
+      "step": 593
+    },
+    {
+      "epoch": 0.6671346342833075,
+      "grad_norm": 0.5182267099894416,
+      "learning_rate": 1.993177511713569e-05,
+      "loss": 0.4121,
+      "step": 594
+    },
+    {
+      "epoch": 0.6682577565632458,
+      "grad_norm": 0.49114322650286973,
+      "learning_rate": 1.9930857334153374e-05,
+      "loss": 0.3738,
+      "step": 595
+    },
+    {
+      "epoch": 0.6693808788431841,
+      "grad_norm": 0.47709228497356176,
+      "learning_rate": 1.992993344060525e-05,
+      "loss": 0.3767,
+      "step": 596
+    },
+    {
+      "epoch": 0.6705040011231223,
+      "grad_norm": 0.5108226641321751,
+      "learning_rate": 1.992900343705981e-05,
+      "loss": 0.3901,
+      "step": 597
+    },
+    {
+      "epoch": 0.6716271234030605,
+      "grad_norm": 0.5041216553802877,
+      "learning_rate": 1.9928067324089286e-05,
+      "loss": 0.3699,
+      "step": 598
+    },
+    {
+      "epoch": 0.6727502456829987,
+      "grad_norm": 0.4923581285784332,
+      "learning_rate": 1.9927125102269687e-05,
+      "loss": 0.3764,
+      "step": 599
+    },
+    {
+      "epoch": 0.673873367962937,
+      "grad_norm": 0.5387456019298923,
+      "learning_rate": 1.992617677218077e-05,
+      "loss": 0.3994,
+      "step": 600
+    },
+    {
+      "epoch": 0.6749964902428752,
+      "grad_norm": 0.5115844130946426,
+      "learning_rate": 1.992522233440605e-05,
+      "loss": 0.4016,
+      "step": 601
+    },
+    {
+      "epoch": 0.6761196125228134,
+      "grad_norm": 0.489549840063431,
+      "learning_rate": 1.992426178953281e-05,
+      "loss": 0.3861,
+      "step": 602
+    },
+    {
+      "epoch": 0.6772427348027517,
+      "grad_norm": 0.48352200149034974,
+      "learning_rate": 1.9923295138152076e-05,
+      "loss": 0.3501,
+      "step": 603
+    },
+    {
+      "epoch": 0.6783658570826899,
+      "grad_norm": 0.5003180468080372,
+      "learning_rate": 1.992232238085864e-05,
+      "loss": 0.3927,
+      "step": 604
+    },
+    {
+      "epoch": 0.6794889793626281,
+      "grad_norm": 0.46529290603039347,
+      "learning_rate": 1.9921343518251057e-05,
+      "loss": 0.3613,
+      "step": 605
+    },
+    {
+      "epoch": 0.6806121016425664,
+      "grad_norm": 0.48009299538190453,
+      "learning_rate": 1.9920358550931627e-05,
+      "loss": 0.3692,
+      "step": 606
+    },
+    {
+      "epoch": 0.6817352239225045,
+      "grad_norm": 0.4965554167186214,
+      "learning_rate": 1.9919367479506413e-05,
+      "loss": 0.3629,
+      "step": 607
+    },
+    {
+      "epoch": 0.6828583462024428,
+      "grad_norm": 0.48758593884923196,
+      "learning_rate": 1.9918370304585228e-05,
+      "loss": 0.3583,
+      "step": 608
+    },
+    {
+      "epoch": 0.6839814684823811,
+      "grad_norm": 0.5358513727384278,
+      "learning_rate": 1.991736702678165e-05,
+      "loss": 0.3916,
+      "step": 609
+    },
+    {
+      "epoch": 0.6851045907623192,
+      "grad_norm": 0.4867798988879104,
+      "learning_rate": 1.9916357646713006e-05,
+      "loss": 0.3452,
+      "step": 610
+    },
+    {
+      "epoch": 0.6862277130422575,
+      "grad_norm": 0.5325347519332022,
+      "learning_rate": 1.9915342165000375e-05,
+      "loss": 0.403,
+      "step": 611
+    },
+    {
+      "epoch": 0.6873508353221957,
+      "grad_norm": 0.48855001583167373,
+      "learning_rate": 1.99143205822686e-05,
+      "loss": 0.3791,
+      "step": 612
+    },
+    {
+      "epoch": 0.6884739576021339,
+      "grad_norm": 0.47712559355490186,
+      "learning_rate": 1.9913292899146262e-05,
+      "loss": 0.3626,
+      "step": 613
+    },
+    {
+      "epoch": 0.6895970798820722,
+      "grad_norm": 0.537623122561854,
+      "learning_rate": 1.9912259116265718e-05,
+      "loss": 0.3769,
+      "step": 614
+    },
+    {
+      "epoch": 0.6907202021620104,
+      "grad_norm": 0.4613452971894331,
+      "learning_rate": 1.991121923426306e-05,
+      "loss": 0.3562,
+      "step": 615
+    },
+    {
+      "epoch": 0.6918433244419486,
+      "grad_norm": 0.5176397056990323,
+      "learning_rate": 1.9910173253778136e-05,
+      "loss": 0.3937,
+      "step": 616
+    },
+    {
+      "epoch": 0.6929664467218869,
+      "grad_norm": 0.502108110386917,
+      "learning_rate": 1.9909121175454553e-05,
+      "loss": 0.3852,
+      "step": 617
+    },
+    {
+      "epoch": 0.6940895690018251,
+      "grad_norm": 0.46871949558914766,
+      "learning_rate": 1.9908062999939666e-05,
+      "loss": 0.3615,
+      "step": 618
+    },
+    {
+      "epoch": 0.6952126912817633,
+      "grad_norm": 0.49710099802582,
+      "learning_rate": 1.9906998727884582e-05,
+      "loss": 0.3624,
+      "step": 619
+    },
+    {
+      "epoch": 0.6963358135617015,
+      "grad_norm": 0.5132520818715471,
+      "learning_rate": 1.990592835994416e-05,
+      "loss": 0.381,
+      "step": 620
+    },
+    {
+      "epoch": 0.6974589358416398,
+      "grad_norm": 0.47566380028639976,
+      "learning_rate": 1.990485189677701e-05,
+      "loss": 0.391,
+      "step": 621
+    },
+    {
+      "epoch": 0.698582058121578,
+      "grad_norm": 0.47811544132679007,
+      "learning_rate": 1.990376933904549e-05,
+      "loss": 0.3662,
+      "step": 622
+    },
+    {
+      "epoch": 0.6997051804015162,
+      "grad_norm": 0.5120337787106785,
+      "learning_rate": 1.9902680687415704e-05,
+      "loss": 0.3731,
+      "step": 623
+    },
+    {
+      "epoch": 0.7008283026814545,
+      "grad_norm": 0.5300963176142912,
+      "learning_rate": 1.990158594255752e-05,
+      "loss": 0.3679,
+      "step": 624
+    },
+    {
+      "epoch": 0.7019514249613926,
+      "grad_norm": 0.47924641278798885,
+      "learning_rate": 1.9900485105144544e-05,
+      "loss": 0.3589,
+      "step": 625
+    },
+    {
+      "epoch": 0.7030745472413309,
+      "grad_norm": 0.5270278691730991,
+      "learning_rate": 1.9899378175854134e-05,
+      "loss": 0.3921,
+      "step": 626
+    },
+    {
+      "epoch": 0.7041976695212692,
+      "grad_norm": 0.5179981618980336,
+      "learning_rate": 1.9898265155367394e-05,
+      "loss": 0.4171,
+      "step": 627
+    },
+    {
+      "epoch": 0.7053207918012073,
+      "grad_norm": 0.5055637943014714,
+      "learning_rate": 1.9897146044369177e-05,
+      "loss": 0.3733,
+      "step": 628
+    },
+    {
+      "epoch": 0.7064439140811456,
+      "grad_norm": 0.5349952912388607,
+      "learning_rate": 1.9896020843548086e-05,
+      "loss": 0.3912,
+      "step": 629
+    },
+    {
+      "epoch": 0.7075670363610839,
+      "grad_norm": 0.47936450848582013,
+      "learning_rate": 1.989488955359647e-05,
+      "loss": 0.3461,
+      "step": 630
+    },
+    {
+      "epoch": 0.708690158641022,
+      "grad_norm": 0.5229256593874981,
+      "learning_rate": 1.9893752175210423e-05,
+      "loss": 0.3835,
+      "step": 631
+    },
+    {
+      "epoch": 0.7098132809209603,
+      "grad_norm": 0.5517790504856993,
+      "learning_rate": 1.9892608709089788e-05,
+      "loss": 0.379,
+      "step": 632
+    },
+    {
+      "epoch": 0.7109364032008985,
+      "grad_norm": 0.5135986709269499,
+      "learning_rate": 1.9891459155938148e-05,
+      "loss": 0.3938,
+      "step": 633
+    },
+    {
+      "epoch": 0.7120595254808367,
+      "grad_norm": 0.5235231286661839,
+      "learning_rate": 1.9890303516462842e-05,
+      "loss": 0.4047,
+      "step": 634
+    },
+    {
+      "epoch": 0.713182647760775,
+      "grad_norm": 0.5230019091855481,
+      "learning_rate": 1.9889141791374942e-05,
+      "loss": 0.3772,
+      "step": 635
+    },
+    {
+      "epoch": 0.7143057700407132,
+      "grad_norm": 0.5086902065079811,
+      "learning_rate": 1.9887973981389276e-05,
+      "loss": 0.3789,
+      "step": 636
+    },
+    {
+      "epoch": 0.7154288923206514,
+      "grad_norm": 0.45834248146994216,
+      "learning_rate": 1.9886800087224404e-05,
+      "loss": 0.3511,
+      "step": 637
+    },
+    {
+      "epoch": 0.7165520146005896,
+      "grad_norm": 0.5193295159932503,
+      "learning_rate": 1.9885620109602637e-05,
+      "loss": 0.3777,
+      "step": 638
+    },
+    {
+      "epoch": 0.7176751368805279,
+      "grad_norm": 0.4751465804458114,
+      "learning_rate": 1.9884434049250038e-05,
+      "loss": 0.3747,
+      "step": 639
+    },
+    {
+      "epoch": 0.7187982591604661,
+      "grad_norm": 0.47889259165907166,
+      "learning_rate": 1.988324190689639e-05,
+      "loss": 0.3564,
+      "step": 640
+    },
+    {
+      "epoch": 0.7199213814404043,
+      "grad_norm": 0.4504694191651383,
+      "learning_rate": 1.9882043683275235e-05,
+      "loss": 0.3569,
+      "step": 641
+    },
+    {
+      "epoch": 0.7210445037203426,
+      "grad_norm": 0.48131939806970797,
+      "learning_rate": 1.9880839379123854e-05,
+      "loss": 0.3775,
+      "step": 642
+    },
+    {
+      "epoch": 0.7221676260002807,
+      "grad_norm": 0.5054764749704692,
+      "learning_rate": 1.9879628995183274e-05,
+      "loss": 0.3804,
+      "step": 643
+    },
+    {
+      "epoch": 0.723290748280219,
+      "grad_norm": 0.49166280486529856,
+      "learning_rate": 1.987841253219825e-05,
+      "loss": 0.3846,
+      "step": 644
+    },
+    {
+      "epoch": 0.7244138705601573,
+      "grad_norm": 0.4729734631200996,
+      "learning_rate": 1.987718999091729e-05,
+      "loss": 0.3721,
+      "step": 645
+    },
+    {
+      "epoch": 0.7255369928400954,
+      "grad_norm": 0.5129501870416402,
+      "learning_rate": 1.987596137209263e-05,
+      "loss": 0.3656,
+      "step": 646
+    },
+    {
+      "epoch": 0.7266601151200337,
+      "grad_norm": 0.5182252271285124,
+      "learning_rate": 1.9874726676480264e-05,
+      "loss": 0.4007,
+      "step": 647
+    },
+    {
+      "epoch": 0.727783237399972,
+      "grad_norm": 0.5183328972199854,
+      "learning_rate": 1.98734859048399e-05,
+      "loss": 0.3853,
+      "step": 648
+    },
+    {
+      "epoch": 0.7289063596799101,
+      "grad_norm": 0.48679212429312807,
+      "learning_rate": 1.9872239057935013e-05,
+      "loss": 0.3615,
+      "step": 649
+    },
+    {
+      "epoch": 0.7300294819598484,
+      "grad_norm": 0.4599807442751718,
+      "learning_rate": 1.9870986136532792e-05,
+      "loss": 0.3447,
+      "step": 650
+    },
+    {
+      "epoch": 0.7311526042397866,
+      "grad_norm": 0.48761341111368395,
+      "learning_rate": 1.986972714140418e-05,
+      "loss": 0.3487,
+      "step": 651
+    },
+    {
+      "epoch": 0.7322757265197248,
+      "grad_norm": 0.5019293541443021,
+      "learning_rate": 1.986846207332384e-05,
+      "loss": 0.378,
+      "step": 652
+    },
+    {
+      "epoch": 0.7333988487996631,
+      "grad_norm": 0.507765982179853,
+      "learning_rate": 1.986719093307019e-05,
+      "loss": 0.3653,
+      "step": 653
+    },
+    {
+      "epoch": 0.7345219710796013,
+      "grad_norm": 0.48272271090692315,
+      "learning_rate": 1.9865913721425376e-05,
+      "loss": 0.3542,
+      "step": 654
+    },
+    {
+      "epoch": 0.7356450933595395,
+      "grad_norm": 0.5175822423440689,
+      "learning_rate": 1.9864630439175282e-05,
+      "loss": 0.366,
+      "step": 655
+    },
+    {
+      "epoch": 0.7367682156394777,
+      "grad_norm": 0.5289074782303135,
+      "learning_rate": 1.986334108710952e-05,
+      "loss": 0.3817,
+      "step": 656
+    },
+    {
+      "epoch": 0.737891337919416,
+      "grad_norm": 0.4861617055976239,
+      "learning_rate": 1.9862045666021448e-05,
+      "loss": 0.3752,
+      "step": 657
+    },
+    {
+      "epoch": 0.7390144601993542,
+      "grad_norm": 0.4957975445225569,
+      "learning_rate": 1.986074417670815e-05,
+      "loss": 0.3588,
+      "step": 658
+    },
+    {
+      "epoch": 0.7401375824792924,
+      "grad_norm": 0.4948274783820183,
+      "learning_rate": 1.9859436619970448e-05,
+      "loss": 0.3854,
+      "step": 659
+    },
+    {
+      "epoch": 0.7412607047592307,
+      "grad_norm": 0.4948086029632615,
+      "learning_rate": 1.9858122996612897e-05,
+      "loss": 0.3778,
+      "step": 660
+    },
+    {
+      "epoch": 0.7423838270391689,
+      "grad_norm": 0.5091232713262371,
+      "learning_rate": 1.9856803307443782e-05,
+      "loss": 0.3943,
+      "step": 661
+    },
+    {
+      "epoch": 0.7435069493191071,
+      "grad_norm": 0.47932797892770085,
+      "learning_rate": 1.985547755327512e-05,
+      "loss": 0.3608,
+      "step": 662
+    },
+    {
+      "epoch": 0.7446300715990454,
+      "grad_norm": 0.48396806021455513,
+      "learning_rate": 1.9854145734922668e-05,
+      "loss": 0.3691,
+      "step": 663
+    },
+    {
+      "epoch": 0.7457531938789835,
+      "grad_norm": 0.48291497456525306,
+      "learning_rate": 1.9852807853205903e-05,
+      "loss": 0.3544,
+      "step": 664
+    },
+    {
+      "epoch": 0.7468763161589218,
+      "grad_norm": 0.527750629448579,
+      "learning_rate": 1.985146390894804e-05,
+      "loss": 0.3795,
+      "step": 665
+    },
+    {
+      "epoch": 0.7479994384388601,
+      "grad_norm": 0.49379054725765187,
+      "learning_rate": 1.9850113902976025e-05,
+      "loss": 0.3595,
+      "step": 666
+    },
+    {
+      "epoch": 0.7491225607187982,
+      "grad_norm": 0.5043416333342748,
+      "learning_rate": 1.9848757836120528e-05,
+      "loss": 0.3627,
+      "step": 667
+    },
+    {
+      "epoch": 0.7502456829987365,
+      "grad_norm": 0.49989404508934815,
+      "learning_rate": 1.984739570921596e-05,
+      "loss": 0.3515,
+      "step": 668
+    },
+    {
+      "epoch": 0.7513688052786747,
+      "grad_norm": 0.508418215461715,
+      "learning_rate": 1.984602752310044e-05,
+      "loss": 0.3601,
+      "step": 669
+    },
+    {
+      "epoch": 0.7524919275586129,
+      "grad_norm": 0.5189115847179923,
+      "learning_rate": 1.9844653278615836e-05,
+      "loss": 0.3865,
+      "step": 670
+    },
+    {
+      "epoch": 0.7536150498385512,
+      "grad_norm": 0.49519160344351704,
+      "learning_rate": 1.9843272976607735e-05,
+      "loss": 0.3789,
+      "step": 671
+    },
+    {
+      "epoch": 0.7547381721184894,
+      "grad_norm": 0.5125455405404472,
+      "learning_rate": 1.984188661792545e-05,
+      "loss": 0.3902,
+      "step": 672
+    },
+    {
+      "epoch": 0.7558612943984276,
+      "grad_norm": 0.46902819064744006,
+      "learning_rate": 1.9840494203422024e-05,
+      "loss": 0.3562,
+      "step": 673
+    },
+    {
+      "epoch": 0.7569844166783659,
+      "grad_norm": 0.4712682187918327,
+      "learning_rate": 1.9839095733954226e-05,
+      "loss": 0.3724,
+      "step": 674
+    },
+    {
+      "epoch": 0.7581075389583041,
+      "grad_norm": 0.5083096985128949,
+      "learning_rate": 1.9837691210382547e-05,
+      "loss": 0.3945,
+      "step": 675
+    },
+    {
+      "epoch": 0.7592306612382423,
+      "grad_norm": 0.4756841491353359,
+      "learning_rate": 1.9836280633571206e-05,
+      "loss": 0.3964,
+      "step": 676
+    },
+    {
+      "epoch": 0.7603537835181805,
+      "grad_norm": 0.5229455215501784,
+      "learning_rate": 1.9834864004388153e-05,
+      "loss": 0.3925,
+      "step": 677
+    },
+    {
+      "epoch": 0.7614769057981188,
+      "grad_norm": 0.490426080752418,
+      "learning_rate": 1.9833441323705046e-05,
+      "loss": 0.3608,
+      "step": 678
+    },
+    {
+      "epoch": 0.762600028078057,
+      "grad_norm": 0.46495132814690193,
+      "learning_rate": 1.983201259239728e-05,
+      "loss": 0.3733,
+      "step": 679
+    },
+    {
+      "epoch": 0.7637231503579952,
+      "grad_norm": 0.5047763095575648,
+      "learning_rate": 1.9830577811343973e-05,
+      "loss": 0.3618,
+      "step": 680
+    },
+    {
+      "epoch": 0.7648462726379335,
+      "grad_norm": 0.5216773678215171,
+      "learning_rate": 1.9829136981427953e-05,
+      "loss": 0.4084,
+      "step": 681
+    },
+    {
+      "epoch": 0.7659693949178716,
+      "grad_norm": 0.48638773567106564,
+      "learning_rate": 1.9827690103535787e-05,
+      "loss": 0.3539,
+      "step": 682
+    },
+    {
+      "epoch": 0.7670925171978099,
+      "grad_norm": 0.44873074839468674,
+      "learning_rate": 1.9826237178557752e-05,
+      "loss": 0.3502,
+      "step": 683
+    },
+    {
+      "epoch": 0.7682156394777482,
+      "grad_norm": 0.48611782864710706,
+      "learning_rate": 1.982477820738785e-05,
+      "loss": 0.3755,
+      "step": 684
+    },
+    {
+      "epoch": 0.7693387617576863,
+      "grad_norm": 0.5308709710691715,
+      "learning_rate": 1.9823313190923797e-05,
+      "loss": 0.3836,
+      "step": 685
+    },
+    {
+      "epoch": 0.7704618840376246,
+      "grad_norm": 0.49106310483861326,
+      "learning_rate": 1.9821842130067034e-05,
+      "loss": 0.3525,
+      "step": 686
+    },
+    {
+      "epoch": 0.7715850063175628,
+      "grad_norm": 0.49027171214223086,
+      "learning_rate": 1.982036502572273e-05,
+      "loss": 0.3929,
+      "step": 687
+    },
+    {
+      "epoch": 0.772708128597501,
+      "grad_norm": 0.5040180030072051,
+      "learning_rate": 1.9818881878799755e-05,
+      "loss": 0.3671,
+      "step": 688
+    },
+    {
+      "epoch": 0.7738312508774393,
+      "grad_norm": 0.4673216659840785,
+      "learning_rate": 1.9817392690210705e-05,
+      "loss": 0.3687,
+      "step": 689
+    },
+    {
+      "epoch": 0.7749543731573775,
+      "grad_norm": 0.5310710286528179,
+      "learning_rate": 1.98158974608719e-05,
+      "loss": 0.3694,
+      "step": 690
+    },
+    {
+      "epoch": 0.7760774954373157,
+      "grad_norm": 0.47994947051344194,
+      "learning_rate": 1.981439619170337e-05,
+      "loss": 0.3556,
+      "step": 691
+    },
+    {
+      "epoch": 0.777200617717254,
+      "grad_norm": 0.5463444744890577,
+      "learning_rate": 1.981288888362886e-05,
+      "loss": 0.3775,
+      "step": 692
+    },
+    {
+      "epoch": 0.7783237399971922,
+      "grad_norm": 0.4724849513305692,
+      "learning_rate": 1.981137553757583e-05,
+      "loss": 0.373,
+      "step": 693
+    },
+    {
+      "epoch": 0.7794468622771304,
+      "grad_norm": 0.4833836898450605,
+      "learning_rate": 1.9809856154475463e-05,
+      "loss": 0.3602,
+      "step": 694
+    },
+    {
+      "epoch": 0.7805699845570686,
+      "grad_norm": 0.44065803236999307,
+      "learning_rate": 1.9808330735262657e-05,
+      "loss": 0.3323,
+      "step": 695
+    },
+    {
+      "epoch": 0.7816931068370069,
+      "grad_norm": 0.5062777136832733,
+      "learning_rate": 1.980679928087601e-05,
+      "loss": 0.3889,
+      "step": 696
+    },
+    {
+      "epoch": 0.7828162291169452,
+      "grad_norm": 0.4390743377796556,
+      "learning_rate": 1.980526179225785e-05,
+      "loss": 0.3619,
+      "step": 697
+    },
+    {
+      "epoch": 0.7839393513968833,
+      "grad_norm": 0.5039788856880819,
+      "learning_rate": 1.98037182703542e-05,
+      "loss": 0.3932,
+      "step": 698
+    },
+    {
+      "epoch": 0.7850624736768216,
+      "grad_norm": 0.4645613871607041,
+      "learning_rate": 1.9802168716114817e-05,
+      "loss": 0.3466,
+      "step": 699
+    },
+    {
+      "epoch": 0.7861855959567597,
+      "grad_norm": 0.48869710859846344,
+      "learning_rate": 1.9800613130493158e-05,
+      "loss": 0.3882,
+      "step": 700
+    },
+    {
+      "epoch": 0.787308718236698,
+      "grad_norm": 0.46252538633492307,
+      "learning_rate": 1.9799051514446383e-05,
+      "loss": 0.3731,
+      "step": 701
+    },
+    {
+      "epoch": 0.7884318405166363,
+      "grad_norm": 0.475702182268912,
+      "learning_rate": 1.9797483868935385e-05,
+      "loss": 0.3535,
+      "step": 702
+    },
+    {
+      "epoch": 0.7895549627965744,
+      "grad_norm": 0.4469783800160413,
+      "learning_rate": 1.979591019492474e-05,
+      "loss": 0.3448,
+      "step": 703
+    },
+    {
+      "epoch": 0.7906780850765127,
+      "grad_norm": 0.4513038149897409,
+      "learning_rate": 1.9794330493382753e-05,
+      "loss": 0.3749,
+      "step": 704
+    },
+    {
+      "epoch": 0.791801207356451,
+      "grad_norm": 0.48022563808130214,
+      "learning_rate": 1.9792744765281435e-05,
+      "loss": 0.373,
+      "step": 705
+    },
+    {
+      "epoch": 0.7929243296363891,
+      "grad_norm": 0.4482413723061531,
+      "learning_rate": 1.9791153011596497e-05,
+      "loss": 0.3506,
+      "step": 706
+    },
+    {
+      "epoch": 0.7940474519163274,
+      "grad_norm": 0.4520233506526563,
+      "learning_rate": 1.9789555233307363e-05,
+      "loss": 0.359,
+      "step": 707
+    },
+    {
+      "epoch": 0.7951705741962656,
+      "grad_norm": 0.4674282940663872,
+      "learning_rate": 1.9787951431397167e-05,
+      "loss": 0.3851,
+      "step": 708
+    },
+    {
+      "epoch": 0.7962936964762038,
+      "grad_norm": 0.4584028901339973,
+      "learning_rate": 1.9786341606852743e-05,
+      "loss": 0.3501,
+      "step": 709
+    },
+    {
+      "epoch": 0.7974168187561421,
+      "grad_norm": 0.4473101801633832,
+      "learning_rate": 1.9784725760664632e-05,
+      "loss": 0.368,
+      "step": 710
+    },
+    {
+      "epoch": 0.7985399410360803,
+      "grad_norm": 0.4657336852380225,
+      "learning_rate": 1.9783103893827088e-05,
+      "loss": 0.3558,
+      "step": 711
+    },
+    {
+      "epoch": 0.7996630633160186,
+      "grad_norm": 0.45906042695952914,
+      "learning_rate": 1.9781476007338058e-05,
+      "loss": 0.3599,
+      "step": 712
+    },
+    {
+      "epoch": 0.8007861855959567,
+      "grad_norm": 0.46033366175628637,
+      "learning_rate": 1.97798421021992e-05,
+      "loss": 0.3679,
+      "step": 713
+    },
+    {
+      "epoch": 0.801909307875895,
+      "grad_norm": 0.4911730826448443,
+      "learning_rate": 1.9778202179415875e-05,
+      "loss": 0.3872,
+      "step": 714
+    },
+    {
+      "epoch": 0.8030324301558333,
+      "grad_norm": 0.463483660266614,
+      "learning_rate": 1.9776556239997146e-05,
+      "loss": 0.3758,
+      "step": 715
+    },
+    {
+      "epoch": 0.8041555524357714,
+      "grad_norm": 0.5063628630954315,
+      "learning_rate": 1.9774904284955775e-05,
+      "loss": 0.3824,
+      "step": 716
+    },
+    {
+      "epoch": 0.8052786747157097,
+      "grad_norm": 0.5154742655615026,
+      "learning_rate": 1.9773246315308226e-05,
+      "loss": 0.3527,
+      "step": 717
+    },
+    {
+      "epoch": 0.8064017969956478,
+      "grad_norm": 0.4630501893780488,
+      "learning_rate": 1.9771582332074676e-05,
+      "loss": 0.3875,
+      "step": 718
+    },
+    {
+      "epoch": 0.8075249192755861,
+      "grad_norm": 0.49263104968044724,
+      "learning_rate": 1.9769912336278983e-05,
+      "loss": 0.3934,
+      "step": 719
+    },
+    {
+      "epoch": 0.8086480415555244,
+      "grad_norm": 0.4672454867734837,
+      "learning_rate": 1.9768236328948718e-05,
+      "loss": 0.3647,
+      "step": 720
+    },
+    {
+      "epoch": 0.8097711638354625,
+      "grad_norm": 0.4920760592750608,
+      "learning_rate": 1.9766554311115143e-05,
+      "loss": 0.365,
+      "step": 721
+    },
+    {
+      "epoch": 0.8108942861154008,
+      "grad_norm": 0.4647800245178615,
+      "learning_rate": 1.9764866283813224e-05,
+      "loss": 0.3753,
+      "step": 722
+    },
+    {
+      "epoch": 0.8120174083953391,
+      "grad_norm": 0.4411909985899985,
+      "learning_rate": 1.9763172248081623e-05,
+      "loss": 0.3604,
+      "step": 723
+    },
+    {
+      "epoch": 0.8131405306752773,
+      "grad_norm": 0.4577218647324814,
+      "learning_rate": 1.97614722049627e-05,
+      "loss": 0.3489,
+      "step": 724
+    },
+    {
+      "epoch": 0.8142636529552155,
+      "grad_norm": 0.48915076154791487,
+      "learning_rate": 1.975976615550251e-05,
+      "loss": 0.3687,
+      "step": 725
+    },
+    {
+      "epoch": 0.8153867752351537,
+      "grad_norm": 0.47499359925278556,
+      "learning_rate": 1.9758054100750802e-05,
+      "loss": 0.3724,
+      "step": 726
+    },
+    {
+      "epoch": 0.816509897515092,
+      "grad_norm": 0.4869359292315476,
+      "learning_rate": 1.9756336041761023e-05,
+      "loss": 0.3708,
+      "step": 727
+    },
+    {
+      "epoch": 0.8176330197950302,
+      "grad_norm": 0.5178982061022716,
+      "learning_rate": 1.9754611979590313e-05,
+      "loss": 0.3755,
+      "step": 728
+    },
+    {
+      "epoch": 0.8187561420749684,
+      "grad_norm": 0.5664792802714854,
+      "learning_rate": 1.9752881915299507e-05,
+      "loss": 0.3925,
+      "step": 729
+    },
+    {
+      "epoch": 0.8198792643549067,
+      "grad_norm": 0.5059697408626879,
+      "learning_rate": 1.9751145849953135e-05,
+      "loss": 0.3806,
+      "step": 730
+    },
+    {
+      "epoch": 0.8210023866348448,
+      "grad_norm": 0.5546342735387513,
+      "learning_rate": 1.9749403784619415e-05,
+      "loss": 0.3911,
+      "step": 731
+    },
+    {
+      "epoch": 0.8221255089147831,
+      "grad_norm": 0.5151210704909763,
+      "learning_rate": 1.9747655720370257e-05,
+      "loss": 0.383,
+      "step": 732
+    },
+    {
+      "epoch": 0.8232486311947214,
+      "grad_norm": 0.5218588159099736,
+      "learning_rate": 1.9745901658281267e-05,
+      "loss": 0.3831,
+      "step": 733
+    },
+    {
+      "epoch": 0.8243717534746595,
+      "grad_norm": 0.5275710652132647,
+      "learning_rate": 1.9744141599431737e-05,
+      "loss": 0.3626,
+      "step": 734
+    },
+    {
+      "epoch": 0.8254948757545978,
+      "grad_norm": 0.5597253504535993,
+      "learning_rate": 1.9742375544904654e-05,
+      "loss": 0.372,
+      "step": 735
+    },
+    {
+      "epoch": 0.8266179980345361,
+      "grad_norm": 0.5603047467724314,
+      "learning_rate": 1.9740603495786687e-05,
+      "loss": 0.3842,
+      "step": 736
+    },
+    {
+      "epoch": 0.8277411203144742,
+      "grad_norm": 0.4804454455092537,
+      "learning_rate": 1.97388254531682e-05,
+      "loss": 0.371,
+      "step": 737
+    },
+    {
+      "epoch": 0.8288642425944125,
+      "grad_norm": 0.49610624738284403,
+      "learning_rate": 1.9737041418143235e-05,
+      "loss": 0.3692,
+      "step": 738
+    },
+    {
+      "epoch": 0.8299873648743507,
+      "grad_norm": 0.5248533454817134,
+      "learning_rate": 1.973525139180954e-05,
+      "loss": 0.4152,
+      "step": 739
+    },
+    {
+      "epoch": 0.8311104871542889,
+      "grad_norm": 0.470530311658482,
+      "learning_rate": 1.9733455375268532e-05,
+      "loss": 0.3856,
+      "step": 740
+    },
+    {
+      "epoch": 0.8322336094342272,
+      "grad_norm": 0.46958211569658714,
+      "learning_rate": 1.9731653369625318e-05,
+      "loss": 0.3817,
+      "step": 741
+    },
+    {
+      "epoch": 0.8333567317141654,
+      "grad_norm": 0.5020551519019959,
+      "learning_rate": 1.9729845375988694e-05,
+      "loss": 0.393,
+      "step": 742
+    },
+    {
+      "epoch": 0.8344798539941036,
+      "grad_norm": 0.4740449984336523,
+      "learning_rate": 1.9728031395471138e-05,
+      "loss": 0.3808,
+      "step": 743
+    },
+    {
+      "epoch": 0.8356029762740418,
+      "grad_norm": 0.49152487880677487,
+      "learning_rate": 1.972621142918881e-05,
+      "loss": 0.3827,
+      "step": 744
+    },
+    {
+      "epoch": 0.8367260985539801,
+      "grad_norm": 0.5170747001231172,
+      "learning_rate": 1.972438547826156e-05,
+      "loss": 0.3909,
+      "step": 745
+    },
+    {
+      "epoch": 0.8378492208339183,
+      "grad_norm": 0.48953966689610007,
+      "learning_rate": 1.972255354381291e-05,
+      "loss": 0.3816,
+      "step": 746
+    },
+    {
+      "epoch": 0.8389723431138565,
+      "grad_norm": 0.45475300416733155,
+      "learning_rate": 1.9720715626970078e-05,
+      "loss": 0.3572,
+      "step": 747
+    },
+    {
+      "epoch": 0.8400954653937948,
+      "grad_norm": 0.526527914204446,
+      "learning_rate": 1.9718871728863946e-05,
+      "loss": 0.3707,
+      "step": 748
+    },
+    {
+      "epoch": 0.841218587673733,
+      "grad_norm": 0.47359041114522704,
+      "learning_rate": 1.971702185062909e-05,
+      "loss": 0.3832,
+      "step": 749
+    },
+    {
+      "epoch": 0.8423417099536712,
+      "grad_norm": 0.5066230520321908,
+      "learning_rate": 1.9715165993403756e-05,
+      "loss": 0.3689,
+      "step": 750
+    },
+    {
+      "epoch": 0.8434648322336095,
+      "grad_norm": 0.5222356669126356,
+      "learning_rate": 1.9713304158329873e-05,
+      "loss": 0.3761,
+      "step": 751
+    },
+    {
+      "epoch": 0.8445879545135476,
+      "grad_norm": 0.47850315807515664,
+      "learning_rate": 1.9711436346553055e-05,
+      "loss": 0.3765,
+      "step": 752
+    },
+    {
+      "epoch": 0.8457110767934859,
+      "grad_norm": 0.5204791826622412,
+      "learning_rate": 1.9709562559222585e-05,
+      "loss": 0.4,
+      "step": 753
+    },
+    {
+      "epoch": 0.8468341990734242,
+      "grad_norm": 0.5207156559730746,
+      "learning_rate": 1.970768279749142e-05,
+      "loss": 0.3838,
+      "step": 754
+    },
+    {
+      "epoch": 0.8479573213533623,
+      "grad_norm": 0.48686440349858606,
+      "learning_rate": 1.9705797062516204e-05,
+      "loss": 0.3806,
+      "step": 755
+    },
+    {
+      "epoch": 0.8490804436333006,
+      "grad_norm": 0.5009212386994344,
+      "learning_rate": 1.9703905355457244e-05,
+      "loss": 0.3766,
+      "step": 756
+    },
+    {
+      "epoch": 0.8502035659132388,
+      "grad_norm": 0.4823384548317993,
+      "learning_rate": 1.9702007677478535e-05,
+      "loss": 0.3598,
+      "step": 757
+    },
+    {
+      "epoch": 0.851326688193177,
+      "grad_norm": 0.5038461477537077,
+      "learning_rate": 1.9700104029747735e-05,
+      "loss": 0.3706,
+      "step": 758
+    },
+    {
+      "epoch": 0.8524498104731153,
+      "grad_norm": 0.499367218377909,
+      "learning_rate": 1.9698194413436186e-05,
+      "loss": 0.3931,
+      "step": 759
+    },
+    {
+      "epoch": 0.8535729327530535,
+      "grad_norm": 0.4558139088156976,
+      "learning_rate": 1.9696278829718882e-05,
+      "loss": 0.3661,
+      "step": 760
+    },
+    {
+      "epoch": 0.8546960550329917,
+      "grad_norm": 0.49796966194709463,
+      "learning_rate": 1.9694357279774516e-05,
+      "loss": 0.3762,
+      "step": 761
+    },
+    {
+      "epoch": 0.8558191773129299,
+      "grad_norm": 0.47539394574011445,
+      "learning_rate": 1.9692429764785436e-05,
+      "loss": 0.3686,
+      "step": 762
+    },
+    {
+      "epoch": 0.8569422995928682,
+      "grad_norm": 0.4631634576367207,
+      "learning_rate": 1.969049628593766e-05,
+      "loss": 0.3765,
+      "step": 763
+    },
+    {
+      "epoch": 0.8580654218728064,
+      "grad_norm": 0.4622717685068893,
+      "learning_rate": 1.9688556844420877e-05,
+      "loss": 0.3819,
+      "step": 764
+    },
+    {
+      "epoch": 0.8591885441527446,
+      "grad_norm": 0.46303379442069725,
+      "learning_rate": 1.9686611441428455e-05,
+      "loss": 0.3417,
+      "step": 765
+    },
+    {
+      "epoch": 0.8603116664326829,
+      "grad_norm": 0.48471273461639347,
+      "learning_rate": 1.968466007815741e-05,
+      "loss": 0.3832,
+      "step": 766
+    },
+    {
+      "epoch": 0.8614347887126211,
+      "grad_norm": 0.4955079045703151,
+      "learning_rate": 1.968270275580845e-05,
+      "loss": 0.3973,
+      "step": 767
+    },
+    {
+      "epoch": 0.8625579109925593,
+      "grad_norm": 0.46897950183267123,
+      "learning_rate": 1.968073947558593e-05,
+      "loss": 0.3577,
+      "step": 768
+    },
+    {
+      "epoch": 0.8636810332724976,
+      "grad_norm": 0.49932189730734644,
+      "learning_rate": 1.9678770238697876e-05,
+      "loss": 0.3828,
+      "step": 769
+    },
+    {
+      "epoch": 0.8648041555524357,
+      "grad_norm": 0.4385490040536742,
+      "learning_rate": 1.9676795046355986e-05,
+      "loss": 0.3424,
+      "step": 770
+    },
+    {
+      "epoch": 0.865927277832374,
+      "grad_norm": 0.44172765516598467,
+      "learning_rate": 1.9674813899775614e-05,
+      "loss": 0.3504,
+      "step": 771
+    },
+    {
+      "epoch": 0.8670504001123123,
+      "grad_norm": 0.4925607529971466,
+      "learning_rate": 1.9672826800175786e-05,
+      "loss": 0.3712,
+      "step": 772
+    },
+    {
+      "epoch": 0.8681735223922504,
+      "grad_norm": 0.4703364651049239,
+      "learning_rate": 1.967083374877918e-05,
+      "loss": 0.3953,
+      "step": 773
+    },
+    {
+      "epoch": 0.8692966446721887,
+      "grad_norm": 0.4744439366917524,
+      "learning_rate": 1.9668834746812148e-05,
+      "loss": 0.3789,
+      "step": 774
+    },
+    {
+      "epoch": 0.8704197669521269,
+      "grad_norm": 0.4821724265365719,
+      "learning_rate": 1.9666829795504693e-05,
+      "loss": 0.3801,
+      "step": 775
+    },
+    {
+      "epoch": 0.8715428892320651,
+      "grad_norm": 0.48264988540719894,
+      "learning_rate": 1.966481889609049e-05,
+      "loss": 0.3774,
+      "step": 776
+    },
+    {
+      "epoch": 0.8726660115120034,
+      "grad_norm": 0.4694424121246419,
+      "learning_rate": 1.9662802049806862e-05,
+      "loss": 0.3574,
+      "step": 777
+    },
+    {
+      "epoch": 0.8737891337919416,
+      "grad_norm": 0.47288161574228044,
+      "learning_rate": 1.96607792578948e-05,
+      "loss": 0.3654,
+      "step": 778
+    },
+    {
+      "epoch": 0.8749122560718798,
+      "grad_norm": 0.4556516097283803,
+      "learning_rate": 1.9658750521598944e-05,
+      "loss": 0.3557,
+      "step": 779
+    },
+    {
+      "epoch": 0.8760353783518181,
+      "grad_norm": 0.42490236766792466,
+      "learning_rate": 1.9656715842167606e-05,
+      "loss": 0.3504,
+      "step": 780
+    },
+    {
+      "epoch": 0.8771585006317563,
+      "grad_norm": 0.493236263957415,
+      "learning_rate": 1.965467522085274e-05,
+      "loss": 0.377,
+      "step": 781
+    },
+    {
+      "epoch": 0.8782816229116945,
+      "grad_norm": 0.4836659716972491,
+      "learning_rate": 1.9652628658909968e-05,
+      "loss": 0.3731,
+      "step": 782
+    },
+    {
+      "epoch": 0.8794047451916327,
+      "grad_norm": 0.46192080514474665,
+      "learning_rate": 1.965057615759856e-05,
+      "loss": 0.3961,
+      "step": 783
+    },
+    {
+      "epoch": 0.880527867471571,
+      "grad_norm": 0.48934298480858224,
+      "learning_rate": 1.964851771818144e-05,
+      "loss": 0.35,
+      "step": 784
+    },
+    {
+      "epoch": 0.8816509897515092,
+      "grad_norm": 0.46465029854890294,
+      "learning_rate": 1.9646453341925195e-05,
+      "loss": 0.372,
+      "step": 785
+    },
+    {
+      "epoch": 0.8827741120314474,
+      "grad_norm": 0.4560535944639423,
+      "learning_rate": 1.9644383030100052e-05,
+      "loss": 0.3499,
+      "step": 786
+    },
+    {
+      "epoch": 0.8838972343113857,
+      "grad_norm": 0.4472553278120461,
+      "learning_rate": 1.9642306783979902e-05,
+      "loss": 0.3563,
+      "step": 787
+    },
+    {
+      "epoch": 0.8850203565913238,
+      "grad_norm": 0.4728183298345325,
+      "learning_rate": 1.964022460484227e-05,
+      "loss": 0.3915,
+      "step": 788
+    },
+    {
+      "epoch": 0.8861434788712621,
+      "grad_norm": 0.49042138391881734,
+      "learning_rate": 1.963813649396836e-05,
+      "loss": 0.4013,
+      "step": 789
+    },
+    {
+      "epoch": 0.8872666011512004,
+      "grad_norm": 0.44584377009424664,
+      "learning_rate": 1.9636042452643004e-05,
+      "loss": 0.3408,
+      "step": 790
+    },
+    {
+      "epoch": 0.8883897234311385,
+      "grad_norm": 0.4725306653731623,
+      "learning_rate": 1.9633942482154684e-05,
+      "loss": 0.3786,
+      "step": 791
+    },
+    {
+      "epoch": 0.8895128457110768,
+      "grad_norm": 0.4614739050205394,
+      "learning_rate": 1.9631836583795537e-05,
+      "loss": 0.3546,
+      "step": 792
+    },
+    {
+      "epoch": 0.8906359679910151,
+      "grad_norm": 0.5062498553591016,
+      "learning_rate": 1.962972475886135e-05,
+      "loss": 0.379,
+      "step": 793
+    },
+    {
+      "epoch": 0.8917590902709532,
+      "grad_norm": 0.44137610647370507,
+      "learning_rate": 1.9627607008651544e-05,
+      "loss": 0.3719,
+      "step": 794
+    },
+    {
+      "epoch": 0.8928822125508915,
+      "grad_norm": 0.4748874875427979,
+      "learning_rate": 1.9625483334469198e-05,
+      "loss": 0.3944,
+      "step": 795
+    },
+    {
+      "epoch": 0.8940053348308297,
+      "grad_norm": 0.4105278757488109,
+      "learning_rate": 1.9623353737621035e-05,
+      "loss": 0.3282,
+      "step": 796
+    },
+    {
+      "epoch": 0.8951284571107679,
+      "grad_norm": 0.533115865046379,
+      "learning_rate": 1.962121821941742e-05,
+      "loss": 0.3806,
+      "step": 797
+    },
+    {
+      "epoch": 0.8962515793907062,
+      "grad_norm": 0.4641263305725229,
+      "learning_rate": 1.9619076781172355e-05,
+      "loss": 0.3488,
+      "step": 798
+    },
+    {
+      "epoch": 0.8973747016706444,
+      "grad_norm": 0.4604841991723784,
+      "learning_rate": 1.9616929424203493e-05,
+      "loss": 0.3667,
+      "step": 799
+    },
+    {
+      "epoch": 0.8984978239505826,
+      "grad_norm": 0.5045896764129607,
+      "learning_rate": 1.9614776149832127e-05,
+      "loss": 0.3634,
+      "step": 800
+    },
+    {
+      "epoch": 0.8996209462305208,
+      "grad_norm": 0.5218566397092465,
+      "learning_rate": 1.961261695938319e-05,
+      "loss": 0.351,
+      "step": 801
+    },
+    {
+      "epoch": 0.9007440685104591,
+      "grad_norm": 0.4569027978245014,
+      "learning_rate": 1.9610451854185253e-05,
+      "loss": 0.3656,
+      "step": 802
+    },
+    {
+      "epoch": 0.9018671907903973,
+      "grad_norm": 0.49667871542822645,
+      "learning_rate": 1.9608280835570537e-05,
+      "loss": 0.3625,
+      "step": 803
+    },
+    {
+      "epoch": 0.9029903130703355,
+      "grad_norm": 0.4709946091327286,
+      "learning_rate": 1.9606103904874886e-05,
+      "loss": 0.3638,
+      "step": 804
+    },
+    {
+      "epoch": 0.9041134353502738,
+      "grad_norm": 0.4727139263711212,
+      "learning_rate": 1.9603921063437795e-05,
+      "loss": 0.3594,
+      "step": 805
+    },
+    {
+      "epoch": 0.9052365576302119,
+      "grad_norm": 0.5294771408585577,
+      "learning_rate": 1.9601732312602385e-05,
+      "loss": 0.3752,
+      "step": 806
+    },
+    {
+      "epoch": 0.9063596799101502,
+      "grad_norm": 0.46959657369103186,
+      "learning_rate": 1.959953765371542e-05,
+      "loss": 0.3625,
+      "step": 807
+    },
+    {
+      "epoch": 0.9074828021900885,
+      "grad_norm": 0.4594446586146184,
+      "learning_rate": 1.95973370881273e-05,
+      "loss": 0.3655,
+      "step": 808
+    },
+    {
+      "epoch": 0.9086059244700266,
+      "grad_norm": 0.4750310076759929,
+      "learning_rate": 1.959513061719205e-05,
+      "loss": 0.3653,
+      "step": 809
+    },
+    {
+      "epoch": 0.9097290467499649,
+      "grad_norm": 0.47678540184859336,
+      "learning_rate": 1.959291824226735e-05,
+      "loss": 0.3454,
+      "step": 810
+    },
+    {
+      "epoch": 0.9108521690299032,
+      "grad_norm": 0.4700082470689754,
+      "learning_rate": 1.9590699964714486e-05,
+      "loss": 0.3786,
+      "step": 811
+    },
+    {
+      "epoch": 0.9119752913098413,
+      "grad_norm": 0.4620496295067997,
+      "learning_rate": 1.9588475785898394e-05,
+      "loss": 0.3681,
+      "step": 812
+    },
+    {
+      "epoch": 0.9130984135897796,
+      "grad_norm": 0.4642629784834156,
+      "learning_rate": 1.9586245707187634e-05,
+      "loss": 0.353,
+      "step": 813
+    },
+    {
+      "epoch": 0.9142215358697178,
+      "grad_norm": 0.466408448319715,
+      "learning_rate": 1.9584009729954395e-05,
+      "loss": 0.3603,
+      "step": 814
+    },
+    {
+      "epoch": 0.915344658149656,
+      "grad_norm": 0.47105990854376034,
+      "learning_rate": 1.9581767855574508e-05,
+      "loss": 0.3636,
+      "step": 815
+    },
+    {
+      "epoch": 0.9164677804295943,
+      "grad_norm": 0.46413275069783316,
+      "learning_rate": 1.9579520085427416e-05,
+      "loss": 0.3951,
+      "step": 816
+    },
+    {
+      "epoch": 0.9175909027095325,
+      "grad_norm": 0.471808108910633,
+      "learning_rate": 1.9577266420896194e-05,
+      "loss": 0.3544,
+      "step": 817
+    },
+    {
+      "epoch": 0.9187140249894707,
+      "grad_norm": 0.46312200003026205,
+      "learning_rate": 1.9575006863367552e-05,
+      "loss": 0.3747,
+      "step": 818
+    },
+    {
+      "epoch": 0.9198371472694089,
+      "grad_norm": 0.4833523027027275,
+      "learning_rate": 1.957274141423182e-05,
+      "loss": 0.3751,
+      "step": 819
+    },
+    {
+      "epoch": 0.9209602695493472,
+      "grad_norm": 0.484335661168015,
+      "learning_rate": 1.9570470074882947e-05,
+      "loss": 0.4085,
+      "step": 820
+    },
+    {
+      "epoch": 0.9220833918292854,
+      "grad_norm": 0.47205123389289727,
+      "learning_rate": 1.9568192846718523e-05,
+      "loss": 0.3759,
+      "step": 821
+    },
+    {
+      "epoch": 0.9232065141092236,
+      "grad_norm": 0.4799100077083178,
+      "learning_rate": 1.956590973113975e-05,
+      "loss": 0.3755,
+      "step": 822
+    },
+    {
+      "epoch": 0.9243296363891619,
+      "grad_norm": 0.48768754470310466,
+      "learning_rate": 1.9563620729551448e-05,
+      "loss": 0.3579,
+      "step": 823
+    },
+    {
+      "epoch": 0.9254527586691002,
+      "grad_norm": 0.45688489471479504,
+      "learning_rate": 1.956132584336207e-05,
+      "loss": 0.351,
+      "step": 824
+    },
+    {
+      "epoch": 0.9265758809490383,
+      "grad_norm": 0.47538544133352595,
+      "learning_rate": 1.9559025073983678e-05,
+      "loss": 0.3808,
+      "step": 825
+    },
+    {
+      "epoch": 0.9276990032289766,
+      "grad_norm": 0.4879902084963068,
+      "learning_rate": 1.955671842283197e-05,
+      "loss": 0.3722,
+      "step": 826
+    },
+    {
+      "epoch": 0.9288221255089147,
+      "grad_norm": 0.4636609831884478,
+      "learning_rate": 1.955440589132625e-05,
+      "loss": 0.3591,
+      "step": 827
+    },
+    {
+      "epoch": 0.929945247788853,
+      "grad_norm": 0.46923991530374826,
+      "learning_rate": 1.955208748088944e-05,
+      "loss": 0.3825,
+      "step": 828
+    },
+    {
+      "epoch": 0.9310683700687913,
+      "grad_norm": 0.4811858751733392,
+      "learning_rate": 1.954976319294809e-05,
+      "loss": 0.3867,
+      "step": 829
+    },
+    {
+      "epoch": 0.9321914923487294,
+      "grad_norm": 0.44788406222610444,
+      "learning_rate": 1.9547433028932357e-05,
+      "loss": 0.3581,
+      "step": 830
+    },
+    {
+      "epoch": 0.9333146146286677,
+      "grad_norm": 0.4780988778135954,
+      "learning_rate": 1.9545096990276016e-05,
+      "loss": 0.3896,
+      "step": 831
+    },
+    {
+      "epoch": 0.9344377369086059,
+      "grad_norm": 0.47068420296897273,
+      "learning_rate": 1.954275507841646e-05,
+      "loss": 0.3844,
+      "step": 832
+    },
+    {
+      "epoch": 0.9355608591885441,
+      "grad_norm": 0.4884936275913018,
+      "learning_rate": 1.9540407294794685e-05,
+      "loss": 0.4031,
+      "step": 833
+    },
+    {
+      "epoch": 0.9366839814684824,
+      "grad_norm": 0.44700134287470683,
+      "learning_rate": 1.9538053640855316e-05,
+      "loss": 0.3436,
+      "step": 834
+    },
+    {
+      "epoch": 0.9378071037484206,
+      "grad_norm": 0.45061145673357333,
+      "learning_rate": 1.9535694118046584e-05,
+      "loss": 0.3457,
+      "step": 835
+    },
+    {
+      "epoch": 0.9389302260283588,
+      "grad_norm": 0.47148123153216354,
+      "learning_rate": 1.9533328727820322e-05,
+      "loss": 0.368,
+      "step": 836
+    },
+    {
+      "epoch": 0.9400533483082971,
+      "grad_norm": 0.4456166431044062,
+      "learning_rate": 1.953095747163199e-05,
+      "loss": 0.3456,
+      "step": 837
+    },
+    {
+      "epoch": 0.9411764705882353,
+      "grad_norm": 0.44236685307564755,
+      "learning_rate": 1.9528580350940634e-05,
+      "loss": 0.3431,
+      "step": 838
+    },
+    {
+      "epoch": 0.9422995928681736,
+      "grad_norm": 0.48561300894599185,
+      "learning_rate": 1.9526197367208937e-05,
+      "loss": 0.3652,
+      "step": 839
+    },
+    {
+      "epoch": 0.9434227151481117,
+      "grad_norm": 0.4720489415303487,
+      "learning_rate": 1.9523808521903165e-05,
+      "loss": 0.3572,
+      "step": 840
+    },
+    {
+      "epoch": 0.94454583742805,
+      "grad_norm": 0.44376759133578964,
+      "learning_rate": 1.9521413816493206e-05,
+      "loss": 0.3568,
+      "step": 841
+    },
+    {
+      "epoch": 0.9456689597079883,
+      "grad_norm": 0.47211026031113734,
+      "learning_rate": 1.951901325245255e-05,
+      "loss": 0.3589,
+      "step": 842
+    },
+    {
+      "epoch": 0.9467920819879264,
+      "grad_norm": 0.487190131673575,
+      "learning_rate": 1.9516606831258282e-05,
+      "loss": 0.3781,
+      "step": 843
+    },
+    {
+      "epoch": 0.9479152042678647,
+      "grad_norm": 0.5011825293136943,
+      "learning_rate": 1.951419455439111e-05,
+      "loss": 0.3773,
+      "step": 844
+    },
+    {
+      "epoch": 0.9490383265478028,
+      "grad_norm": 0.49776220503772134,
+      "learning_rate": 1.9511776423335327e-05,
+      "loss": 0.4034,
+      "step": 845
+    },
+    {
+      "epoch": 0.9501614488277411,
+      "grad_norm": 0.48886020802858143,
+      "learning_rate": 1.9509352439578843e-05,
+      "loss": 0.354,
+      "step": 846
+    },
+    {
+      "epoch": 0.9512845711076794,
+      "grad_norm": 0.48370344677831884,
+      "learning_rate": 1.950692260461315e-05,
+      "loss": 0.3524,
+      "step": 847
+    },
+    {
+      "epoch": 0.9524076933876175,
+      "grad_norm": 0.49770690316202637,
+      "learning_rate": 1.9504486919933364e-05,
+      "loss": 0.382,
+      "step": 848
+    },
+    {
+      "epoch": 0.9535308156675558,
+      "grad_norm": 0.4947415886475524,
+      "learning_rate": 1.9502045387038184e-05,
+      "loss": 0.3573,
+      "step": 849
+    },
+    {
+      "epoch": 0.954653937947494,
+      "grad_norm": 0.5081571339909152,
+      "learning_rate": 1.949959800742991e-05,
+      "loss": 0.393,
+      "step": 850
+    },
+    {
+      "epoch": 0.9557770602274323,
+      "grad_norm": 0.42096325046895394,
+      "learning_rate": 1.9497144782614445e-05,
+      "loss": 0.319,
+      "step": 851
+    },
+    {
+      "epoch": 0.9569001825073705,
+      "grad_norm": 0.506234866583935,
+      "learning_rate": 1.949468571410128e-05,
+      "loss": 0.3887,
+      "step": 852
+    },
+    {
+      "epoch": 0.9580233047873087,
+      "grad_norm": 0.5146043757907258,
+      "learning_rate": 1.9492220803403518e-05,
+      "loss": 0.3949,
+      "step": 853
+    },
+    {
+      "epoch": 0.959146427067247,
+      "grad_norm": 0.4788729384940881,
+      "learning_rate": 1.9489750052037836e-05,
+      "loss": 0.3748,
+      "step": 854
+    },
+    {
+      "epoch": 0.9602695493471852,
+      "grad_norm": 0.44234608280074666,
+      "learning_rate": 1.9487273461524517e-05,
+      "loss": 0.3599,
+      "step": 855
+    },
+    {
+      "epoch": 0.9613926716271234,
+      "grad_norm": 0.491710752366928,
+      "learning_rate": 1.9484791033387432e-05,
+      "loss": 0.3717,
+      "step": 856
+    },
+    {
+      "epoch": 0.9625157939070617,
+      "grad_norm": 0.4698138574686433,
+      "learning_rate": 1.948230276915405e-05,
+      "loss": 0.379,
+      "step": 857
+    },
+    {
+      "epoch": 0.9636389161869998,
+      "grad_norm": 0.4584810154708352,
+      "learning_rate": 1.9479808670355425e-05,
+      "loss": 0.3686,
+      "step": 858
+    },
+    {
+      "epoch": 0.9647620384669381,
+      "grad_norm": 0.46535287144204673,
+      "learning_rate": 1.9477308738526207e-05,
+      "loss": 0.3489,
+      "step": 859
+    },
+    {
+      "epoch": 0.9658851607468764,
+      "grad_norm": 0.49539431571257847,
+      "learning_rate": 1.947480297520463e-05,
+      "loss": 0.3899,
+      "step": 860
+    },
+    {
+      "epoch": 0.9670082830268145,
+      "grad_norm": 0.4870684988697284,
+      "learning_rate": 1.9472291381932515e-05,
+      "loss": 0.3608,
+      "step": 861
+    },
+    {
+      "epoch": 0.9681314053067528,
+      "grad_norm": 0.476057375325281,
+      "learning_rate": 1.9469773960255273e-05,
+      "loss": 0.3693,
+      "step": 862
+    },
+    {
+      "epoch": 0.969254527586691,
+      "grad_norm": 0.4933511086989605,
+      "learning_rate": 1.9467250711721903e-05,
+      "loss": 0.3673,
+      "step": 863
+    },
+    {
+      "epoch": 0.9703776498666292,
+      "grad_norm": 0.486703140404838,
+      "learning_rate": 1.946472163788499e-05,
+      "loss": 0.3623,
+      "step": 864
+    },
+    {
+      "epoch": 0.9715007721465675,
+      "grad_norm": 0.46311662873138065,
+      "learning_rate": 1.9462186740300697e-05,
+      "loss": 0.3751,
+      "step": 865
+    },
+    {
+      "epoch": 0.9726238944265057,
+      "grad_norm": 0.4756703180857592,
+      "learning_rate": 1.9459646020528777e-05,
+      "loss": 0.4065,
+      "step": 866
+    },
+    {
+      "epoch": 0.9737470167064439,
+      "grad_norm": 0.45742793507262375,
+      "learning_rate": 1.9457099480132563e-05,
+      "loss": 0.3783,
+      "step": 867
+    },
+    {
+      "epoch": 0.9748701389863822,
+      "grad_norm": 0.47697882778530937,
+      "learning_rate": 1.9454547120678966e-05,
+      "loss": 0.3907,
+      "step": 868
+    },
+    {
+      "epoch": 0.9759932612663204,
+      "grad_norm": 0.4664102425023445,
+      "learning_rate": 1.9451988943738485e-05,
+      "loss": 0.374,
+      "step": 869
+    },
+    {
+      "epoch": 0.9771163835462586,
+      "grad_norm": 0.41449005991424276,
+      "learning_rate": 1.9449424950885193e-05,
+      "loss": 0.3473,
+      "step": 870
+    },
+    {
+      "epoch": 0.9782395058261968,
+      "grad_norm": 0.4344496368295649,
+      "learning_rate": 1.944685514369674e-05,
+      "loss": 0.3481,
+      "step": 871
+    },
+    {
+      "epoch": 0.9793626281061351,
+      "grad_norm": 0.47436273506860405,
+      "learning_rate": 1.9444279523754358e-05,
+      "loss": 0.3799,
+      "step": 872
+    },
+    {
+      "epoch": 0.9804857503860733,
+      "grad_norm": 0.4905734178799078,
+      "learning_rate": 1.944169809264286e-05,
+      "loss": 0.3655,
+      "step": 873
+    },
+    {
+      "epoch": 0.9816088726660115,
+      "grad_norm": 0.4643421651738806,
+      "learning_rate": 1.9439110851950623e-05,
+      "loss": 0.3705,
+      "step": 874
+    },
+    {
+      "epoch": 0.9827319949459498,
+      "grad_norm": 0.4860647845700408,
+      "learning_rate": 1.9436517803269603e-05,
+      "loss": 0.3556,
+      "step": 875
+    },
+    {
+      "epoch": 0.9838551172258879,
+      "grad_norm": 0.473335344847467,
+      "learning_rate": 1.9433918948195335e-05,
+      "loss": 0.354,
+      "step": 876
+    },
+    {
+      "epoch": 0.9849782395058262,
+      "grad_norm": 0.46317370647334366,
+      "learning_rate": 1.9431314288326925e-05,
+      "loss": 0.3564,
+      "step": 877
+    },
+    {
+      "epoch": 0.9861013617857645,
+      "grad_norm": 0.48357335447638583,
+      "learning_rate": 1.9428703825267047e-05,
+      "loss": 0.3799,
+      "step": 878
+    },
+    {
+      "epoch": 0.9872244840657026,
+      "grad_norm": 0.4575106291945355,
+      "learning_rate": 1.9426087560621944e-05,
+      "loss": 0.3681,
+      "step": 879
+    },
+    {
+      "epoch": 0.9883476063456409,
+      "grad_norm": 0.46101892774024467,
+      "learning_rate": 1.942346549600144e-05,
+      "loss": 0.3683,
+      "step": 880
+    },
+    {
+      "epoch": 0.989470728625579,
+      "grad_norm": 0.47630591998449023,
+      "learning_rate": 1.942083763301891e-05,
+      "loss": 0.3642,
+      "step": 881
+    },
+    {
+      "epoch": 0.9905938509055173,
+      "grad_norm": 0.44576916696684105,
+      "learning_rate": 1.9418203973291317e-05,
+      "loss": 0.3525,
+      "step": 882
+    },
+    {
+      "epoch": 0.9917169731854556,
+      "grad_norm": 0.43611774214843185,
+      "learning_rate": 1.9415564518439173e-05,
+      "loss": 0.3376,
+      "step": 883
+    },
+    {
+      "epoch": 0.9928400954653938,
+      "grad_norm": 0.44908126004356413,
+      "learning_rate": 1.9412919270086573e-05,
+      "loss": 0.3574,
+      "step": 884
+    },
+    {
+      "epoch": 0.993963217745332,
+      "grad_norm": 0.48679555991927803,
+      "learning_rate": 1.941026822986116e-05,
+      "loss": 0.394,
+      "step": 885
+    },
+    {
+      "epoch": 0.9950863400252703,
+      "grad_norm": 0.45873412100669514,
+      "learning_rate": 1.9407611399394145e-05,
+      "loss": 0.3587,
+      "step": 886
+    },
+    {
+      "epoch": 0.9962094623052085,
+      "grad_norm": 0.45788041705605353,
+      "learning_rate": 1.9404948780320313e-05,
+      "loss": 0.3629,
+      "step": 887
+    },
+    {
+      "epoch": 0.9973325845851467,
+      "grad_norm": 0.47526745180684987,
+      "learning_rate": 1.9402280374278e-05,
+      "loss": 0.342,
+      "step": 888
+    },
+    {
+      "epoch": 0.9984557068650849,
+      "grad_norm": 0.4320159068252172,
+      "learning_rate": 1.9399606182909104e-05,
+      "loss": 0.3578,
+      "step": 889
+    },
+    {
+      "epoch": 0.9995788291450232,
+      "grad_norm": 0.43875671365459706,
+      "learning_rate": 1.9396926207859085e-05,
+      "loss": 0.3598,
+      "step": 890
+    },
+    {
+      "epoch": 1.0007019514249613,
+      "grad_norm": 0.8018480221556775,
+      "learning_rate": 1.9394240450776963e-05,
+      "loss": 0.5421,
+      "step": 891
+    },
+    {
+      "epoch": 1.0018250737048997,
+      "grad_norm": 0.4696264632124667,
+      "learning_rate": 1.9391548913315312e-05,
+      "loss": 0.3137,
+      "step": 892
+    },
+    {
+      "epoch": 1.0029481959848379,
+      "grad_norm": 0.4583140254324182,
+      "learning_rate": 1.9388851597130268e-05,
+      "loss": 0.3087,
+      "step": 893
+    },
+    {
+      "epoch": 1.004071318264776,
+      "grad_norm": 0.4690268898240805,
+      "learning_rate": 1.9386148503881515e-05,
+      "loss": 0.2877,
+      "step": 894
+    },
+    {
+      "epoch": 1.0051944405447144,
+      "grad_norm": 0.5122329490697226,
+      "learning_rate": 1.9383439635232296e-05,
+      "loss": 0.2904,
+      "step": 895
+    },
+    {
+      "epoch": 1.0063175628246526,
+      "grad_norm": 0.5371156002795432,
+      "learning_rate": 1.938072499284941e-05,
+      "loss": 0.3008,
+      "step": 896
+    },
+    {
+      "epoch": 1.0074406851045907,
+      "grad_norm": 0.4902755957708443,
+      "learning_rate": 1.9378004578403208e-05,
+      "loss": 0.2819,
+      "step": 897
+    },
+    {
+      "epoch": 1.0085638073845289,
+      "grad_norm": 0.4741699815198155,
+      "learning_rate": 1.9375278393567588e-05,
+      "loss": 0.2627,
+      "step": 898
+    },
+    {
+      "epoch": 1.0096869296644673,
+      "grad_norm": 0.48986713359299766,
+      "learning_rate": 1.9372546440020003e-05,
+      "loss": 0.2914,
+      "step": 899
+    },
+    {
+      "epoch": 1.0108100519444054,
+      "grad_norm": 0.5014813747394071,
+      "learning_rate": 1.9369808719441448e-05,
+      "loss": 0.2912,
+      "step": 900
+    },
+    {
+      "epoch": 1.0119331742243436,
+      "grad_norm": 0.5393899996183346,
+      "learning_rate": 1.936706523351648e-05,
+      "loss": 0.3034,
+      "step": 901
+    },
+    {
+      "epoch": 1.013056296504282,
+      "grad_norm": 0.48629973546736344,
+      "learning_rate": 1.93643159839332e-05,
+      "loss": 0.315,
+      "step": 902
+    },
+    {
+      "epoch": 1.0141794187842201,
+      "grad_norm": 0.47646907262990806,
+      "learning_rate": 1.9361560972383237e-05,
+      "loss": 0.2829,
+      "step": 903
+    },
+    {
+      "epoch": 1.0153025410641583,
+      "grad_norm": 0.4784634253911548,
+      "learning_rate": 1.935880020056179e-05,
+      "loss": 0.2634,
+      "step": 904
+    },
+    {
+      "epoch": 1.0164256633440967,
+      "grad_norm": 0.4991566494669612,
+      "learning_rate": 1.9356033670167595e-05,
+      "loss": 0.2836,
+      "step": 905
+    },
+    {
+      "epoch": 1.0175487856240348,
+      "grad_norm": 0.47357930015207117,
+      "learning_rate": 1.935326138290292e-05,
+      "loss": 0.2845,
+      "step": 906
+    },
+    {
+      "epoch": 1.018671907903973,
+      "grad_norm": 0.4650462704839138,
+      "learning_rate": 1.9350483340473595e-05,
+      "loss": 0.2772,
+      "step": 907
+    },
+    {
+      "epoch": 1.0197950301839114,
+      "grad_norm": 0.5050485921674276,
+      "learning_rate": 1.9347699544588966e-05,
+      "loss": 0.2964,
+      "step": 908
+    },
+    {
+      "epoch": 1.0209181524638495,
+      "grad_norm": 0.441877998501999,
+      "learning_rate": 1.9344909996961943e-05,
+      "loss": 0.2783,
+      "step": 909
+    },
+    {
+      "epoch": 1.0220412747437877,
+      "grad_norm": 0.4689605191500745,
+      "learning_rate": 1.9342114699308962e-05,
+      "loss": 0.2759,
+      "step": 910
+    },
+    {
+      "epoch": 1.0231643970237259,
+      "grad_norm": 0.4806645271199599,
+      "learning_rate": 1.933931365335e-05,
+      "loss": 0.2724,
+      "step": 911
+    },
+    {
+      "epoch": 1.0242875193036642,
+      "grad_norm": 0.48206510864685853,
+      "learning_rate": 1.9336506860808576e-05,
+      "loss": 0.2845,
+      "step": 912
+    },
+    {
+      "epoch": 1.0254106415836024,
+      "grad_norm": 0.473977653318953,
+      "learning_rate": 1.9333694323411732e-05,
+      "loss": 0.2691,
+      "step": 913
+    },
+    {
+      "epoch": 1.0265337638635406,
+      "grad_norm": 0.4959577608373437,
+      "learning_rate": 1.9330876042890065e-05,
+      "loss": 0.2744,
+      "step": 914
+    },
+    {
+      "epoch": 1.027656886143479,
+      "grad_norm": 0.4962252725676315,
+      "learning_rate": 1.932805202097768e-05,
+      "loss": 0.2855,
+      "step": 915
+    },
+    {
+      "epoch": 1.028780008423417,
+      "grad_norm": 0.5128348503516724,
+      "learning_rate": 1.9325222259412242e-05,
+      "loss": 0.2795,
+      "step": 916
+    },
+    {
+      "epoch": 1.0299031307033553,
+      "grad_norm": 0.4632243182786546,
+      "learning_rate": 1.9322386759934923e-05,
+      "loss": 0.2643,
+      "step": 917
+    },
+    {
+      "epoch": 1.0310262529832936,
+      "grad_norm": 0.4730264051539773,
+      "learning_rate": 1.9319545524290447e-05,
+      "loss": 0.2654,
+      "step": 918
+    },
+    {
+      "epoch": 1.0321493752632318,
+      "grad_norm": 0.5126131024471591,
+      "learning_rate": 1.931669855422705e-05,
+      "loss": 0.2902,
+      "step": 919
+    },
+    {
+      "epoch": 1.03327249754317,
+      "grad_norm": 0.4743950568309829,
+      "learning_rate": 1.9313845851496507e-05,
+      "loss": 0.2665,
+      "step": 920
+    },
+    {
+      "epoch": 1.0343956198231083,
+      "grad_norm": 0.5013671474523442,
+      "learning_rate": 1.931098741785412e-05,
+      "loss": 0.2954,
+      "step": 921
+    },
+    {
+      "epoch": 1.0355187421030465,
+      "grad_norm": 0.5096143882338355,
+      "learning_rate": 1.930812325505871e-05,
+      "loss": 0.2737,
+      "step": 922
+    },
+    {
+      "epoch": 1.0366418643829847,
+      "grad_norm": 0.48645730923494584,
+      "learning_rate": 1.930525336487263e-05,
+      "loss": 0.2871,
+      "step": 923
+    },
+    {
+      "epoch": 1.0377649866629228,
+      "grad_norm": 0.48648949738495667,
+      "learning_rate": 1.9302377749061753e-05,
+      "loss": 0.2871,
+      "step": 924
+    },
+    {
+      "epoch": 1.0388881089428612,
+      "grad_norm": 0.5445641193684013,
+      "learning_rate": 1.9299496409395482e-05,
+      "loss": 0.2788,
+      "step": 925
+    },
+    {
+      "epoch": 1.0400112312227994,
+      "grad_norm": 0.4832715223137077,
+      "learning_rate": 1.9296609347646732e-05,
+      "loss": 0.2705,
+      "step": 926
+    },
+    {
+      "epoch": 1.0411343535027375,
+      "grad_norm": 0.4610567782289876,
+      "learning_rate": 1.9293716565591948e-05,
+      "loss": 0.266,
+      "step": 927
+    },
+    {
+      "epoch": 1.042257475782676,
+      "grad_norm": 0.476608249508321,
+      "learning_rate": 1.9290818065011084e-05,
+      "loss": 0.2806,
+      "step": 928
+    },
+    {
+      "epoch": 1.043380598062614,
+      "grad_norm": 0.4820198451645742,
+      "learning_rate": 1.9287913847687627e-05,
+      "loss": 0.2745,
+      "step": 929
+    },
+    {
+      "epoch": 1.0445037203425522,
+      "grad_norm": 0.5092363784952347,
+      "learning_rate": 1.9285003915408575e-05,
+      "loss": 0.29,
+      "step": 930
+    },
+    {
+      "epoch": 1.0456268426224906,
+      "grad_norm": 0.4745997969575985,
+      "learning_rate": 1.928208826996443e-05,
+      "loss": 0.2679,
+      "step": 931
+    },
+    {
+      "epoch": 1.0467499649024288,
+      "grad_norm": 0.48451499822152283,
+      "learning_rate": 1.927916691314923e-05,
+      "loss": 0.2979,
+      "step": 932
+    },
+    {
+      "epoch": 1.047873087182367,
+      "grad_norm": 0.4559532527795186,
+      "learning_rate": 1.9276239846760515e-05,
+      "loss": 0.2745,
+      "step": 933
+    },
+    {
+      "epoch": 1.048996209462305,
+      "grad_norm": 0.4942046444197993,
+      "learning_rate": 1.9273307072599343e-05,
+      "loss": 0.2733,
+      "step": 934
+    },
+    {
+      "epoch": 1.0501193317422435,
+      "grad_norm": 0.4716142833431438,
+      "learning_rate": 1.927036859247028e-05,
+      "loss": 0.2848,
+      "step": 935
+    },
+    {
+      "epoch": 1.0512424540221816,
+      "grad_norm": 0.6276556942610505,
+      "learning_rate": 1.9267424408181406e-05,
+      "loss": 0.2928,
+      "step": 936
+    },
+    {
+      "epoch": 1.0523655763021198,
+      "grad_norm": 0.4840738368552477,
+      "learning_rate": 1.9264474521544315e-05,
+      "loss": 0.2831,
+      "step": 937
+    },
+    {
+      "epoch": 1.0534886985820582,
+      "grad_norm": 0.513332959686835,
+      "learning_rate": 1.9261518934374093e-05,
+      "loss": 0.3035,
+      "step": 938
+    },
+    {
+      "epoch": 1.0546118208619963,
+      "grad_norm": 0.4556522892250145,
+      "learning_rate": 1.9258557648489357e-05,
+      "loss": 0.27,
+      "step": 939
+    },
+    {
+      "epoch": 1.0557349431419345,
+      "grad_norm": 0.5074845207806349,
+      "learning_rate": 1.9255590665712214e-05,
+      "loss": 0.283,
+      "step": 940
+    },
+    {
+      "epoch": 1.0568580654218729,
+      "grad_norm": 0.499752040349772,
+      "learning_rate": 1.9252617987868278e-05,
+      "loss": 0.2857,
+      "step": 941
+    },
+    {
+      "epoch": 1.057981187701811,
+      "grad_norm": 0.5396255130756957,
+      "learning_rate": 1.9249639616786674e-05,
+      "loss": 0.3088,
+      "step": 942
+    },
+    {
+      "epoch": 1.0591043099817492,
+      "grad_norm": 0.49212073826967684,
+      "learning_rate": 1.9246655554300028e-05,
+      "loss": 0.2748,
+      "step": 943
+    },
+    {
+      "epoch": 1.0602274322616876,
+      "grad_norm": 0.4968561055146951,
+      "learning_rate": 1.9243665802244465e-05,
+      "loss": 0.2749,
+      "step": 944
+    },
+    {
+      "epoch": 1.0613505545416257,
+      "grad_norm": 0.49427378612848244,
+      "learning_rate": 1.924067036245961e-05,
+      "loss": 0.2953,
+      "step": 945
+    },
+    {
+      "epoch": 1.062473676821564,
+      "grad_norm": 0.48543259955055773,
+      "learning_rate": 1.9237669236788595e-05,
+      "loss": 0.282,
+      "step": 946
+    },
+    {
+      "epoch": 1.0635967991015023,
+      "grad_norm": 0.5219319492699731,
+      "learning_rate": 1.923466242707804e-05,
+      "loss": 0.2971,
+      "step": 947
+    },
+    {
+      "epoch": 1.0647199213814404,
+      "grad_norm": 0.4692640333578925,
+      "learning_rate": 1.923164993517807e-05,
+      "loss": 0.2753,
+      "step": 948
+    },
+    {
+      "epoch": 1.0658430436613786,
+      "grad_norm": 0.47927624646872385,
+      "learning_rate": 1.9228631762942307e-05,
+      "loss": 0.2738,
+      "step": 949
+    },
+    {
+      "epoch": 1.0669661659413168,
+      "grad_norm": 0.5057928867460986,
+      "learning_rate": 1.9225607912227864e-05,
+      "loss": 0.2734,
+      "step": 950
+    },
+    {
+      "epoch": 1.0680892882212552,
+      "grad_norm": 0.46267709140380153,
+      "learning_rate": 1.922257838489535e-05,
+      "loss": 0.2751,
+      "step": 951
+    },
+    {
+      "epoch": 1.0692124105011933,
+      "grad_norm": 0.5332480394868323,
+      "learning_rate": 1.9219543182808862e-05,
+      "loss": 0.2973,
+      "step": 952
+    },
+    {
+      "epoch": 1.0703355327811315,
+      "grad_norm": 0.4839164977782545,
+      "learning_rate": 1.9216502307836002e-05,
+      "loss": 0.2835,
+      "step": 953
+    },
+    {
+      "epoch": 1.0714586550610699,
+      "grad_norm": 0.4999219278635034,
+      "learning_rate": 1.9213455761847845e-05,
+      "loss": 0.2927,
+      "step": 954
+    },
+    {
+      "epoch": 1.072581777341008,
+      "grad_norm": 0.5232385009052505,
+      "learning_rate": 1.921040354671897e-05,
+      "loss": 0.3031,
+      "step": 955
+    },
+    {
+      "epoch": 1.0737048996209462,
+      "grad_norm": 0.4541280942778829,
+      "learning_rate": 1.9207345664327434e-05,
+      "loss": 0.2697,
+      "step": 956
+    },
+    {
+      "epoch": 1.0748280219008846,
+      "grad_norm": 0.5031936868920343,
+      "learning_rate": 1.9204282116554792e-05,
+      "loss": 0.2867,
+      "step": 957
+    },
+    {
+      "epoch": 1.0759511441808227,
+      "grad_norm": 0.4713104216421202,
+      "learning_rate": 1.9201212905286074e-05,
+      "loss": 0.2788,
+      "step": 958
+    },
+    {
+      "epoch": 1.0770742664607609,
+      "grad_norm": 0.45894312520199465,
+      "learning_rate": 1.91981380324098e-05,
+      "loss": 0.2848,
+      "step": 959
+    },
+    {
+      "epoch": 1.078197388740699,
+      "grad_norm": 0.5223017626157912,
+      "learning_rate": 1.919505749981798e-05,
+      "loss": 0.293,
+      "step": 960
+    },
+    {
+      "epoch": 1.0793205110206374,
+      "grad_norm": 0.47157901979531547,
+      "learning_rate": 1.9191971309406085e-05,
+      "loss": 0.2774,
+      "step": 961
+    },
+    {
+      "epoch": 1.0804436333005756,
+      "grad_norm": 0.4988305143165187,
+      "learning_rate": 1.9188879463073093e-05,
+      "loss": 0.2967,
+      "step": 962
+    },
+    {
+      "epoch": 1.0815667555805137,
+      "grad_norm": 0.47998629427503187,
+      "learning_rate": 1.918578196272145e-05,
+      "loss": 0.2739,
+      "step": 963
+    },
+    {
+      "epoch": 1.0826898778604521,
+      "grad_norm": 0.46801323046606513,
+      "learning_rate": 1.918267881025708e-05,
+      "loss": 0.2689,
+      "step": 964
+    },
+    {
+      "epoch": 1.0838130001403903,
+      "grad_norm": 0.42494477612916176,
+      "learning_rate": 1.9179570007589384e-05,
+      "loss": 0.2707,
+      "step": 965
+    },
+    {
+      "epoch": 1.0849361224203284,
+      "grad_norm": 0.44609307554632605,
+      "learning_rate": 1.9176455556631247e-05,
+      "loss": 0.2626,
+      "step": 966
+    },
+    {
+      "epoch": 1.0860592447002668,
+      "grad_norm": 0.5054138308738797,
+      "learning_rate": 1.9173335459299025e-05,
+      "loss": 0.2902,
+      "step": 967
+    },
+    {
+      "epoch": 1.087182366980205,
+      "grad_norm": 0.44616997151039844,
+      "learning_rate": 1.9170209717512546e-05,
+      "loss": 0.2687,
+      "step": 968
+    },
+    {
+      "epoch": 1.0883054892601431,
+      "grad_norm": 0.497359281811381,
+      "learning_rate": 1.9167078333195116e-05,
+      "loss": 0.3009,
+      "step": 969
+    },
+    {
+      "epoch": 1.0894286115400815,
+      "grad_norm": 0.4643079791688917,
+      "learning_rate": 1.9163941308273504e-05,
+      "loss": 0.2792,
+      "step": 970
+    },
+    {
+      "epoch": 1.0905517338200197,
+      "grad_norm": 0.47811323587663396,
+      "learning_rate": 1.916079864467796e-05,
+      "loss": 0.2976,
+      "step": 971
+    },
+    {
+      "epoch": 1.0916748560999578,
+      "grad_norm": 0.4360155405235776,
+      "learning_rate": 1.9157650344342205e-05,
+      "loss": 0.2621,
+      "step": 972
+    },
+    {
+      "epoch": 1.0927979783798962,
+      "grad_norm": 0.45820803802250853,
+      "learning_rate": 1.9154496409203416e-05,
+      "loss": 0.2901,
+      "step": 973
+    },
+    {
+      "epoch": 1.0939211006598344,
+      "grad_norm": 0.4370878083723664,
+      "learning_rate": 1.9151336841202246e-05,
+      "loss": 0.2744,
+      "step": 974
+    },
+    {
+      "epoch": 1.0950442229397725,
+      "grad_norm": 0.450814956559242,
+      "learning_rate": 1.9148171642282812e-05,
+      "loss": 0.2691,
+      "step": 975
+    },
+    {
+      "epoch": 1.0961673452197107,
+      "grad_norm": 0.4621925080101941,
+      "learning_rate": 1.9145000814392696e-05,
+      "loss": 0.2737,
+      "step": 976
+    },
+    {
+      "epoch": 1.097290467499649,
+      "grad_norm": 0.4737134489184813,
+      "learning_rate": 1.914182435948294e-05,
+      "loss": 0.2937,
+      "step": 977
+    },
+    {
+      "epoch": 1.0984135897795873,
+      "grad_norm": 0.4510174026564208,
+      "learning_rate": 1.9138642279508054e-05,
+      "loss": 0.2743,
+      "step": 978
+    },
+    {
+      "epoch": 1.0995367120595254,
+      "grad_norm": 0.48310076248938627,
+      "learning_rate": 1.913545457642601e-05,
+      "loss": 0.2821,
+      "step": 979
+    },
+    {
+      "epoch": 1.1006598343394638,
+      "grad_norm": 0.45975463127730243,
+      "learning_rate": 1.9132261252198236e-05,
+      "loss": 0.2743,
+      "step": 980
+    },
+    {
+      "epoch": 1.101782956619402,
+      "grad_norm": 0.47566940351291187,
+      "learning_rate": 1.912906230878961e-05,
+      "loss": 0.2901,
+      "step": 981
+    },
+    {
+      "epoch": 1.1029060788993401,
+      "grad_norm": 0.49639829721895823,
+      "learning_rate": 1.912585774816849e-05,
+      "loss": 0.3062,
+      "step": 982
+    },
+    {
+      "epoch": 1.1040292011792785,
+      "grad_norm": 0.47301791959429074,
+      "learning_rate": 1.912264757230667e-05,
+      "loss": 0.2869,
+      "step": 983
+    },
+    {
+      "epoch": 1.1051523234592167,
+      "grad_norm": 0.43846903068715937,
+      "learning_rate": 1.9119431783179413e-05,
+      "loss": 0.2585,
+      "step": 984
+    },
+    {
+      "epoch": 1.1062754457391548,
+      "grad_norm": 0.5192008280625632,
+      "learning_rate": 1.911621038276542e-05,
+      "loss": 0.3036,
+      "step": 985
+    },
+    {
+      "epoch": 1.107398568019093,
+      "grad_norm": 0.5011419662554821,
+      "learning_rate": 1.911298337304686e-05,
+      "loss": 0.2813,
+      "step": 986
+    },
+    {
+      "epoch": 1.1085216902990314,
+      "grad_norm": 0.48085556410924124,
+      "learning_rate": 1.9109750756009348e-05,
+      "loss": 0.2857,
+      "step": 987
+    },
+    {
+      "epoch": 1.1096448125789695,
+      "grad_norm": 0.5116521172198443,
+      "learning_rate": 1.9106512533641948e-05,
+      "loss": 0.2867,
+      "step": 988
+    },
+    {
+      "epoch": 1.1107679348589077,
+      "grad_norm": 0.4753703314504095,
+      "learning_rate": 1.9103268707937174e-05,
+      "loss": 0.2723,
+      "step": 989
+    },
+    {
+      "epoch": 1.111891057138846,
+      "grad_norm": 0.48728987587657996,
+      "learning_rate": 1.9100019280890984e-05,
+      "loss": 0.2938,
+      "step": 990
+    },
+    {
+      "epoch": 1.1130141794187842,
+      "grad_norm": 0.46500967954346223,
+      "learning_rate": 1.909676425450279e-05,
+      "loss": 0.277,
+      "step": 991
+    },
+    {
+      "epoch": 1.1141373016987224,
+      "grad_norm": 0.4311368054737046,
+      "learning_rate": 1.9093503630775445e-05,
+      "loss": 0.2589,
+      "step": 992
+    },
+    {
+      "epoch": 1.1152604239786608,
+      "grad_norm": 0.4806685121172132,
+      "learning_rate": 1.9090237411715248e-05,
+      "loss": 0.2828,
+      "step": 993
+    },
+    {
+      "epoch": 1.116383546258599,
+      "grad_norm": 0.47413677573033675,
+      "learning_rate": 1.9086965599331938e-05,
+      "loss": 0.2922,
+      "step": 994
+    },
+    {
+      "epoch": 1.117506668538537,
+      "grad_norm": 0.463329362201204,
+      "learning_rate": 1.9083688195638694e-05,
+      "loss": 0.2991,
+      "step": 995
+    },
+    {
+      "epoch": 1.1186297908184755,
+      "grad_norm": 0.424765333204886,
+      "learning_rate": 1.9080405202652143e-05,
+      "loss": 0.2589,
+      "step": 996
+    },
+    {
+      "epoch": 1.1197529130984136,
+      "grad_norm": 0.4664315339207049,
+      "learning_rate": 1.9077116622392347e-05,
+      "loss": 0.278,
+      "step": 997
+    },
+    {
+      "epoch": 1.1208760353783518,
+      "grad_norm": 0.4580593579853824,
+      "learning_rate": 1.9073822456882806e-05,
+      "loss": 0.2781,
+      "step": 998
+    },
+    {
+      "epoch": 1.12199915765829,
+      "grad_norm": 0.4428162338559992,
+      "learning_rate": 1.907052270815045e-05,
+      "loss": 0.2788,
+      "step": 999
+    },
+    {
+      "epoch": 1.1231222799382283,
+      "grad_norm": 0.4317189895111999,
+      "learning_rate": 1.9067217378225655e-05,
+      "loss": 0.273,
+      "step": 1000
+    },
+    {
+      "epoch": 1.1242454022181665,
+      "grad_norm": 0.46512373441877797,
+      "learning_rate": 1.906390646914223e-05,
+      "loss": 0.2987,
+      "step": 1001
+    },
+    {
+      "epoch": 1.1253685244981046,
+      "grad_norm": 0.48722030698963165,
+      "learning_rate": 1.906058998293741e-05,
+      "loss": 0.3105,
+      "step": 1002
+    },
+    {
+      "epoch": 1.126491646778043,
+      "grad_norm": 0.4758813346060563,
+      "learning_rate": 1.9057267921651865e-05,
+      "loss": 0.3104,
+      "step": 1003
+    },
+    {
+      "epoch": 1.1276147690579812,
+      "grad_norm": 0.4603226186566387,
+      "learning_rate": 1.9053940287329696e-05,
+      "loss": 0.2714,
+      "step": 1004
+    },
+    {
+      "epoch": 1.1287378913379194,
+      "grad_norm": 0.5016859736501262,
+      "learning_rate": 1.9050607082018437e-05,
+      "loss": 0.3054,
+      "step": 1005
+    },
+    {
+      "epoch": 1.1298610136178577,
+      "grad_norm": 0.45587589839460446,
+      "learning_rate": 1.9047268307769044e-05,
+      "loss": 0.295,
+      "step": 1006
+    },
+    {
+      "epoch": 1.130984135897796,
+      "grad_norm": 0.4681404958373164,
+      "learning_rate": 1.90439239666359e-05,
+      "loss": 0.2888,
+      "step": 1007
+    },
+    {
+      "epoch": 1.132107258177734,
+      "grad_norm": 0.472944576189367,
+      "learning_rate": 1.9040574060676813e-05,
+      "loss": 0.2875,
+      "step": 1008
+    },
+    {
+      "epoch": 1.1332303804576722,
+      "grad_norm": 0.4496881501850784,
+      "learning_rate": 1.903721859195302e-05,
+      "loss": 0.28,
+      "step": 1009
+    },
+    {
+      "epoch": 1.1343535027376106,
+      "grad_norm": 0.469620980155489,
+      "learning_rate": 1.9033857562529176e-05,
+      "loss": 0.2731,
+      "step": 1010
+    },
+    {
+      "epoch": 1.1354766250175488,
+      "grad_norm": 0.45917004853908955,
+      "learning_rate": 1.9030490974473363e-05,
+      "loss": 0.283,
+      "step": 1011
+    },
+    {
+      "epoch": 1.136599747297487,
+      "grad_norm": 0.47268979034926534,
+      "learning_rate": 1.902711882985708e-05,
+      "loss": 0.2866,
+      "step": 1012
+    },
+    {
+      "epoch": 1.1377228695774253,
+      "grad_norm": 0.4383225568741319,
+      "learning_rate": 1.9023741130755237e-05,
+      "loss": 0.2756,
+      "step": 1013
+    },
+    {
+      "epoch": 1.1388459918573635,
+      "grad_norm": 0.46958663656332716,
+      "learning_rate": 1.9020357879246173e-05,
+      "loss": 0.2826,
+      "step": 1014
+    },
+    {
+      "epoch": 1.1399691141373016,
+      "grad_norm": 0.45079995860583716,
+      "learning_rate": 1.9016969077411645e-05,
+      "loss": 0.2872,
+      "step": 1015
+    },
+    {
+      "epoch": 1.14109223641724,
+      "grad_norm": 0.47973270092319154,
+      "learning_rate": 1.9013574727336817e-05,
+      "loss": 0.3096,
+      "step": 1016
+    },
+    {
+      "epoch": 1.1422153586971782,
+      "grad_norm": 0.501462362289904,
+      "learning_rate": 1.9010174831110268e-05,
+      "loss": 0.3117,
+      "step": 1017
+    },
+    {
+      "epoch": 1.1433384809771163,
+      "grad_norm": 0.4513154018538136,
+      "learning_rate": 1.9006769390823994e-05,
+      "loss": 0.252,
+      "step": 1018
+    },
+    {
+      "epoch": 1.1444616032570547,
+      "grad_norm": 0.4500736376781029,
+      "learning_rate": 1.9003358408573396e-05,
+      "loss": 0.2735,
+      "step": 1019
+    },
+    {
+      "epoch": 1.1455847255369929,
+      "grad_norm": 0.5023208553675916,
+      "learning_rate": 1.8999941886457292e-05,
+      "loss": 0.2987,
+      "step": 1020
+    },
+    {
+      "epoch": 1.146707847816931,
+      "grad_norm": 0.45282147717459975,
+      "learning_rate": 1.8996519826577907e-05,
+      "loss": 0.2651,
+      "step": 1021
+    },
+    {
+      "epoch": 1.1478309700968694,
+      "grad_norm": 0.4648324226943947,
+      "learning_rate": 1.899309223104087e-05,
+      "loss": 0.2985,
+      "step": 1022
+    },
+    {
+      "epoch": 1.1489540923768076,
+      "grad_norm": 0.4849118937452435,
+      "learning_rate": 1.898965910195522e-05,
+      "loss": 0.2966,
+      "step": 1023
+    },
+    {
+      "epoch": 1.1500772146567457,
+      "grad_norm": 0.4845634318497336,
+      "learning_rate": 1.89862204414334e-05,
+      "loss": 0.2997,
+      "step": 1024
+    },
+    {
+      "epoch": 1.1512003369366839,
+      "grad_norm": 0.4849781830896908,
+      "learning_rate": 1.8982776251591247e-05,
+      "loss": 0.301,
+      "step": 1025
+    },
+    {
+      "epoch": 1.1523234592166223,
+      "grad_norm": 0.46264538140977657,
+      "learning_rate": 1.8979326534548023e-05,
+      "loss": 0.2892,
+      "step": 1026
+    },
+    {
+      "epoch": 1.1534465814965604,
+      "grad_norm": 0.4751680003078539,
+      "learning_rate": 1.8975871292426365e-05,
+      "loss": 0.2943,
+      "step": 1027
+    },
+    {
+      "epoch": 1.1545697037764986,
+      "grad_norm": 0.47000462603028687,
+      "learning_rate": 1.8972410527352324e-05,
+      "loss": 0.2944,
+      "step": 1028
+    },
+    {
+      "epoch": 1.155692826056437,
+      "grad_norm": 0.46720471177522077,
+      "learning_rate": 1.8968944241455352e-05,
+      "loss": 0.2754,
+      "step": 1029
+    },
+    {
+      "epoch": 1.1568159483363751,
+      "grad_norm": 0.4682045577743917,
+      "learning_rate": 1.8965472436868288e-05,
+      "loss": 0.2945,
+      "step": 1030
+    },
+    {
+      "epoch": 1.1579390706163133,
+      "grad_norm": 0.4681264807044535,
+      "learning_rate": 1.8961995115727373e-05,
+      "loss": 0.2855,
+      "step": 1031
+    },
+    {
+      "epoch": 1.1590621928962517,
+      "grad_norm": 0.46711223820824194,
+      "learning_rate": 1.895851228017224e-05,
+      "loss": 0.2828,
+      "step": 1032
+    },
+    {
+      "epoch": 1.1601853151761898,
+      "grad_norm": 0.4305801770746027,
+      "learning_rate": 1.8955023932345916e-05,
+      "loss": 0.2572,
+      "step": 1033
+    },
+    {
+      "epoch": 1.161308437456128,
+      "grad_norm": 0.4783787183648787,
+      "learning_rate": 1.8951530074394828e-05,
+      "loss": 0.3005,
+      "step": 1034
+    },
+    {
+      "epoch": 1.1624315597360662,
+      "grad_norm": 0.4580608030880634,
+      "learning_rate": 1.894803070846877e-05,
+      "loss": 0.292,
+      "step": 1035
+    },
+    {
+      "epoch": 1.1635546820160045,
+      "grad_norm": 0.44482733683564024,
+      "learning_rate": 1.894452583672095e-05,
+      "loss": 0.2814,
+      "step": 1036
+    },
+    {
+      "epoch": 1.1646778042959427,
+      "grad_norm": 0.46029498474111563,
+      "learning_rate": 1.8941015461307955e-05,
+      "loss": 0.2893,
+      "step": 1037
+    },
+    {
+      "epoch": 1.1658009265758809,
+      "grad_norm": 0.4615488233008817,
+      "learning_rate": 1.8937499584389755e-05,
+      "loss": 0.2932,
+      "step": 1038
+    },
+    {
+      "epoch": 1.1669240488558192,
+      "grad_norm": 0.42956178577189436,
+      "learning_rate": 1.8933978208129705e-05,
+      "loss": 0.2739,
+      "step": 1039
+    },
+    {
+      "epoch": 1.1680471711357574,
+      "grad_norm": 0.4802415097521688,
+      "learning_rate": 1.893045133469455e-05,
+      "loss": 0.2883,
+      "step": 1040
+    },
+    {
+      "epoch": 1.1691702934156956,
+      "grad_norm": 0.4480867146233576,
+      "learning_rate": 1.8926918966254416e-05,
+      "loss": 0.2729,
+      "step": 1041
+    },
+    {
+      "epoch": 1.170293415695634,
+      "grad_norm": 0.49140396781805756,
+      "learning_rate": 1.8923381104982806e-05,
+      "loss": 0.28,
+      "step": 1042
+    },
+    {
+      "epoch": 1.171416537975572,
+      "grad_norm": 0.46490086726476276,
+      "learning_rate": 1.8919837753056606e-05,
+      "loss": 0.2657,
+      "step": 1043
+    },
+    {
+      "epoch": 1.1725396602555103,
+      "grad_norm": 0.4657704225340458,
+      "learning_rate": 1.8916288912656077e-05,
+      "loss": 0.3002,
+      "step": 1044
+    },
+    {
+      "epoch": 1.1736627825354486,
+      "grad_norm": 0.45315444310300473,
+      "learning_rate": 1.891273458596486e-05,
+      "loss": 0.2671,
+      "step": 1045
+    },
+    {
+      "epoch": 1.1747859048153868,
+      "grad_norm": 0.5014981783064109,
+      "learning_rate": 1.8909174775169968e-05,
+      "loss": 0.2791,
+      "step": 1046
+    },
+    {
+      "epoch": 1.175909027095325,
+      "grad_norm": 0.5179938024360514,
+      "learning_rate": 1.89056094824618e-05,
+      "loss": 0.3352,
+      "step": 1047
+    },
+    {
+      "epoch": 1.1770321493752633,
+      "grad_norm": 0.4232454478701444,
+      "learning_rate": 1.8902038710034113e-05,
+      "loss": 0.2557,
+      "step": 1048
+    },
+    {
+      "epoch": 1.1781552716552015,
+      "grad_norm": 0.5139754763017151,
+      "learning_rate": 1.889846246008405e-05,
+      "loss": 0.2959,
+      "step": 1049
+    },
+    {
+      "epoch": 1.1792783939351397,
+      "grad_norm": 0.48647478582657605,
+      "learning_rate": 1.8894880734812106e-05,
+      "loss": 0.3023,
+      "step": 1050
+    },
+    {
+      "epoch": 1.1804015162150778,
+      "grad_norm": 0.46616236114663656,
+      "learning_rate": 1.8891293536422165e-05,
+      "loss": 0.2845,
+      "step": 1051
+    },
+    {
+      "epoch": 1.1815246384950162,
+      "grad_norm": 0.4287393409593001,
+      "learning_rate": 1.888770086712147e-05,
+      "loss": 0.2696,
+      "step": 1052
+    },
+    {
+      "epoch": 1.1826477607749544,
+      "grad_norm": 0.4949520898537966,
+      "learning_rate": 1.8884102729120624e-05,
+      "loss": 0.3007,
+      "step": 1053
+    },
+    {
+      "epoch": 1.1837708830548925,
+      "grad_norm": 0.446833141730547,
+      "learning_rate": 1.88804991246336e-05,
+      "loss": 0.2692,
+      "step": 1054
+    },
+    {
+      "epoch": 1.184894005334831,
+      "grad_norm": 0.48608855458145783,
+      "learning_rate": 1.8876890055877745e-05,
+      "loss": 0.318,
+      "step": 1055
+    },
+    {
+      "epoch": 1.186017127614769,
+      "grad_norm": 0.457882561215241,
+      "learning_rate": 1.887327552507375e-05,
+      "loss": 0.2802,
+      "step": 1056
+    },
+    {
+      "epoch": 1.1871402498947072,
+      "grad_norm": 0.45242524194849565,
+      "learning_rate": 1.886965553444568e-05,
+      "loss": 0.264,
+      "step": 1057
+    },
+    {
+      "epoch": 1.1882633721746454,
+      "grad_norm": 0.48868541054842407,
+      "learning_rate": 1.886603008622095e-05,
+      "loss": 0.3076,
+      "step": 1058
+    },
+    {
+      "epoch": 1.1893864944545838,
+      "grad_norm": 0.44249560468632987,
+      "learning_rate": 1.8862399182630347e-05,
+      "loss": 0.2661,
+      "step": 1059
+    },
+    {
+      "epoch": 1.190509616734522,
+      "grad_norm": 0.4588142894528934,
+      "learning_rate": 1.8858762825908e-05,
+      "loss": 0.2865,
+      "step": 1060
+    },
+    {
+      "epoch": 1.19163273901446,
+      "grad_norm": 0.46946406735526547,
+      "learning_rate": 1.8855121018291394e-05,
+      "loss": 0.3046,
+      "step": 1061
+    },
+    {
+      "epoch": 1.1927558612943985,
+      "grad_norm": 0.45375091401248635,
+      "learning_rate": 1.8851473762021384e-05,
+      "loss": 0.2732,
+      "step": 1062
+    },
+    {
+      "epoch": 1.1938789835743366,
+      "grad_norm": 0.4676425499504646,
+      "learning_rate": 1.8847821059342163e-05,
+      "loss": 0.2932,
+      "step": 1063
+    },
+    {
+      "epoch": 1.1950021058542748,
+      "grad_norm": 0.46249196792362884,
+      "learning_rate": 1.8844162912501277e-05,
+      "loss": 0.302,
+      "step": 1064
+    },
+    {
+      "epoch": 1.1961252281342132,
+      "grad_norm": 0.4812578734002899,
+      "learning_rate": 1.8840499323749624e-05,
+      "loss": 0.303,
+      "step": 1065
+    },
+    {
+      "epoch": 1.1972483504141513,
+      "grad_norm": 0.48405098483559544,
+      "learning_rate": 1.883683029534145e-05,
+      "loss": 0.2887,
+      "step": 1066
+    },
+    {
+      "epoch": 1.1983714726940895,
+      "grad_norm": 0.4608928377281792,
+      "learning_rate": 1.8833155829534356e-05,
+      "loss": 0.2958,
+      "step": 1067
+    },
+    {
+      "epoch": 1.1994945949740279,
+      "grad_norm": 0.43401836542504524,
+      "learning_rate": 1.8829475928589272e-05,
+      "loss": 0.2603,
+      "step": 1068
+    },
+    {
+      "epoch": 1.200617717253966,
+      "grad_norm": 0.4780260992243674,
+      "learning_rate": 1.8825790594770487e-05,
+      "loss": 0.2809,
+      "step": 1069
+    },
+    {
+      "epoch": 1.2017408395339042,
+      "grad_norm": 0.44275206432188047,
+      "learning_rate": 1.882209983034562e-05,
+      "loss": 0.2773,
+      "step": 1070
+    },
+    {
+      "epoch": 1.2028639618138426,
+      "grad_norm": 0.47460553242353476,
+      "learning_rate": 1.881840363758565e-05,
+      "loss": 0.3128,
+      "step": 1071
+    },
+    {
+      "epoch": 1.2039870840937807,
+      "grad_norm": 0.46237921124263054,
+      "learning_rate": 1.881470201876488e-05,
+      "loss": 0.2952,
+      "step": 1072
+    },
+    {
+      "epoch": 1.205110206373719,
+      "grad_norm": 0.48654497979875483,
+      "learning_rate": 1.8810994976160955e-05,
+      "loss": 0.3147,
+      "step": 1073
+    },
+    {
+      "epoch": 1.2062333286536573,
+      "grad_norm": 0.4921863140740264,
+      "learning_rate": 1.880728251205486e-05,
+      "loss": 0.3093,
+      "step": 1074
+    },
+    {
+      "epoch": 1.2073564509335954,
+      "grad_norm": 0.471423938361344,
+      "learning_rate": 1.8803564628730916e-05,
+      "loss": 0.2903,
+      "step": 1075
+    },
+    {
+      "epoch": 1.2084795732135336,
+      "grad_norm": 0.46874454339863675,
+      "learning_rate": 1.8799841328476776e-05,
+      "loss": 0.2889,
+      "step": 1076
+    },
+    {
+      "epoch": 1.2096026954934718,
+      "grad_norm": 0.45281746346669155,
+      "learning_rate": 1.8796112613583427e-05,
+      "loss": 0.2895,
+      "step": 1077
+    },
+    {
+      "epoch": 1.2107258177734102,
+      "grad_norm": 0.45353972532299486,
+      "learning_rate": 1.8792378486345196e-05,
+      "loss": 0.2789,
+      "step": 1078
+    },
+    {
+      "epoch": 1.2118489400533483,
+      "grad_norm": 0.4725888875307044,
+      "learning_rate": 1.878863894905972e-05,
+      "loss": 0.2847,
+      "step": 1079
+    },
+    {
+      "epoch": 1.2129720623332865,
+      "grad_norm": 0.442084650431914,
+      "learning_rate": 1.878489400402799e-05,
+      "loss": 0.274,
+      "step": 1080
+    },
+    {
+      "epoch": 1.2140951846132249,
+      "grad_norm": 0.49254544284536805,
+      "learning_rate": 1.8781143653554305e-05,
+      "loss": 0.3088,
+      "step": 1081
+    },
+    {
+      "epoch": 1.215218306893163,
+      "grad_norm": 0.4617957372006431,
+      "learning_rate": 1.8777387899946294e-05,
+      "loss": 0.2879,
+      "step": 1082
+    },
+    {
+      "epoch": 1.2163414291731012,
+      "grad_norm": 0.4646226561763834,
+      "learning_rate": 1.877362674551492e-05,
+      "loss": 0.2929,
+      "step": 1083
+    },
+    {
+      "epoch": 1.2174645514530393,
+      "grad_norm": 0.5009688641472497,
+      "learning_rate": 1.876986019257446e-05,
+      "loss": 0.314,
+      "step": 1084
+    },
+    {
+      "epoch": 1.2185876737329777,
+      "grad_norm": 0.49964534711190756,
+      "learning_rate": 1.8766088243442514e-05,
+      "loss": 0.32,
+      "step": 1085
+    },
+    {
+      "epoch": 1.2197107960129159,
+      "grad_norm": 0.47363249608950986,
+      "learning_rate": 1.8762310900440007e-05,
+      "loss": 0.2895,
+      "step": 1086
+    },
+    {
+      "epoch": 1.220833918292854,
+      "grad_norm": 0.4781168700712531,
+      "learning_rate": 1.875852816589118e-05,
+      "loss": 0.2826,
+      "step": 1087
+    },
+    {
+      "epoch": 1.2219570405727924,
+      "grad_norm": 0.45871143064600073,
+      "learning_rate": 1.8754740042123583e-05,
+      "loss": 0.2762,
+      "step": 1088
+    },
+    {
+      "epoch": 1.2230801628527306,
+      "grad_norm": 0.5147809750511391,
+      "learning_rate": 1.8750946531468098e-05,
+      "loss": 0.3179,
+      "step": 1089
+    },
+    {
+      "epoch": 1.2242032851326687,
+      "grad_norm": 0.45406534949791066,
+      "learning_rate": 1.874714763625892e-05,
+      "loss": 0.2656,
+      "step": 1090
+    },
+    {
+      "epoch": 1.2253264074126071,
+      "grad_norm": 0.46488456668844064,
+      "learning_rate": 1.8743343358833536e-05,
+      "loss": 0.2867,
+      "step": 1091
+    },
+    {
+      "epoch": 1.2264495296925453,
+      "grad_norm": 0.46081222446789055,
+      "learning_rate": 1.8739533701532768e-05,
+      "loss": 0.2891,
+      "step": 1092
+    },
+    {
+      "epoch": 1.2275726519724834,
+      "grad_norm": 0.47759019663041424,
+      "learning_rate": 1.873571866670074e-05,
+      "loss": 0.2884,
+      "step": 1093
+    },
+    {
+      "epoch": 1.2286957742524218,
+      "grad_norm": 0.4825230030891384,
+      "learning_rate": 1.8731898256684885e-05,
+      "loss": 0.2886,
+      "step": 1094
+    },
+    {
+      "epoch": 1.22981889653236,
+      "grad_norm": 0.4568440667981614,
+      "learning_rate": 1.8728072473835944e-05,
+      "loss": 0.2801,
+      "step": 1095
+    },
+    {
+      "epoch": 1.2309420188122981,
+      "grad_norm": 0.4517593846376601,
+      "learning_rate": 1.8724241320507958e-05,
+      "loss": 0.2687,
+      "step": 1096
+    },
+    {
+      "epoch": 1.2320651410922365,
+      "grad_norm": 0.473458572983134,
+      "learning_rate": 1.8720404799058284e-05,
+      "loss": 0.2811,
+      "step": 1097
+    },
+    {
+      "epoch": 1.2331882633721747,
+      "grad_norm": 0.47853878049484405,
+      "learning_rate": 1.8716562911847572e-05,
+      "loss": 0.3026,
+      "step": 1098
+    },
+    {
+      "epoch": 1.2343113856521128,
+      "grad_norm": 0.4732508123997025,
+      "learning_rate": 1.8712715661239783e-05,
+      "loss": 0.2858,
+      "step": 1099
+    },
+    {
+      "epoch": 1.235434507932051,
+      "grad_norm": 0.4943536151259743,
+      "learning_rate": 1.8708863049602163e-05,
+      "loss": 0.3155,
+      "step": 1100
+    },
+    {
+      "epoch": 1.2365576302119894,
+      "grad_norm": 0.45750032528758056,
+      "learning_rate": 1.8705005079305274e-05,
+      "loss": 0.2851,
+      "step": 1101
+    },
+    {
+      "epoch": 1.2376807524919275,
+      "grad_norm": 0.4521267906219757,
+      "learning_rate": 1.8701141752722966e-05,
+      "loss": 0.2747,
+      "step": 1102
+    },
+    {
+      "epoch": 1.2388038747718657,
+      "grad_norm": 0.46594419380408797,
+      "learning_rate": 1.8697273072232385e-05,
+      "loss": 0.3064,
+      "step": 1103
+    },
+    {
+      "epoch": 1.239926997051804,
+      "grad_norm": 0.43383827589550866,
+      "learning_rate": 1.8693399040213974e-05,
+      "loss": 0.2676,
+      "step": 1104
+    },
+    {
+      "epoch": 1.2410501193317423,
+      "grad_norm": 0.4474050960888805,
+      "learning_rate": 1.8689519659051467e-05,
+      "loss": 0.2741,
+      "step": 1105
+    },
+    {
+      "epoch": 1.2421732416116804,
+      "grad_norm": 0.45523790295881006,
+      "learning_rate": 1.868563493113189e-05,
+      "loss": 0.2881,
+      "step": 1106
+    },
+    {
+      "epoch": 1.2432963638916188,
+      "grad_norm": 0.4584723123279694,
+      "learning_rate": 1.8681744858845555e-05,
+      "loss": 0.284,
+      "step": 1107
+    },
+    {
+      "epoch": 1.244419486171557,
+      "grad_norm": 0.46457228328541583,
+      "learning_rate": 1.8677849444586073e-05,
+      "loss": 0.2799,
+      "step": 1108
+    },
+    {
+      "epoch": 1.2455426084514951,
+      "grad_norm": 0.45539666514036803,
+      "learning_rate": 1.8673948690750333e-05,
+      "loss": 0.2887,
+      "step": 1109
+    },
+    {
+      "epoch": 1.2466657307314333,
+      "grad_norm": 0.4677915933786495,
+      "learning_rate": 1.867004259973851e-05,
+      "loss": 0.292,
+      "step": 1110
+    },
+    {
+      "epoch": 1.2477888530113717,
+      "grad_norm": 0.4667047750396251,
+      "learning_rate": 1.866613117395407e-05,
+      "loss": 0.2889,
+      "step": 1111
+    },
+    {
+      "epoch": 1.2489119752913098,
+      "grad_norm": 0.48997059381677216,
+      "learning_rate": 1.8662214415803748e-05,
+      "loss": 0.2968,
+      "step": 1112
+    },
+    {
+      "epoch": 1.250035097571248,
+      "grad_norm": 0.4384615139812607,
+      "learning_rate": 1.8658292327697574e-05,
+      "loss": 0.2715,
+      "step": 1113
+    },
+    {
+      "epoch": 1.2511582198511864,
+      "grad_norm": 0.4547012939945187,
+      "learning_rate": 1.865436491204885e-05,
+      "loss": 0.2747,
+      "step": 1114
+    },
+    {
+      "epoch": 1.2522813421311245,
+      "grad_norm": 0.5098417390678678,
+      "learning_rate": 1.865043217127416e-05,
+      "loss": 0.3327,
+      "step": 1115
+    },
+    {
+      "epoch": 1.2534044644110627,
+      "grad_norm": 0.47158047623028704,
+      "learning_rate": 1.864649410779336e-05,
+      "loss": 0.2977,
+      "step": 1116
+    },
+    {
+      "epoch": 1.254527586691001,
+      "grad_norm": 0.47071696073688374,
+      "learning_rate": 1.8642550724029584e-05,
+      "loss": 0.3032,
+      "step": 1117
+    },
+    {
+      "epoch": 1.2556507089709392,
+      "grad_norm": 0.4756046517133648,
+      "learning_rate": 1.863860202240924e-05,
+      "loss": 0.3077,
+      "step": 1118
+    },
+    {
+      "epoch": 1.2567738312508774,
+      "grad_norm": 0.47511377917679826,
+      "learning_rate": 1.863464800536201e-05,
+      "loss": 0.2969,
+      "step": 1119
+    },
+    {
+      "epoch": 1.2578969535308158,
+      "grad_norm": 0.47570563192173576,
+      "learning_rate": 1.8630688675320844e-05,
+      "loss": 0.3081,
+      "step": 1120
+    },
+    {
+      "epoch": 1.259020075810754,
+      "grad_norm": 0.4905178570778228,
+      "learning_rate": 1.8626724034721955e-05,
+      "loss": 0.2967,
+      "step": 1121
+    },
+    {
+      "epoch": 1.260143198090692,
+      "grad_norm": 0.4603739199538433,
+      "learning_rate": 1.8622754086004837e-05,
+      "loss": 0.2852,
+      "step": 1122
+    },
+    {
+      "epoch": 1.2612663203706305,
+      "grad_norm": 0.48083734404310896,
+      "learning_rate": 1.8618778831612243e-05,
+      "loss": 0.2836,
+      "step": 1123
+    },
+    {
+      "epoch": 1.2623894426505686,
+      "grad_norm": 0.46026086440947167,
+      "learning_rate": 1.8614798273990186e-05,
+      "loss": 0.2785,
+      "step": 1124
+    },
+    {
+      "epoch": 1.2635125649305068,
+      "grad_norm": 0.49105292345069845,
+      "learning_rate": 1.8610812415587948e-05,
+      "loss": 0.2841,
+      "step": 1125
+    },
+    {
+      "epoch": 1.2646356872104452,
+      "grad_norm": 0.45003283326379967,
+      "learning_rate": 1.860682125885808e-05,
+      "loss": 0.2817,
+      "step": 1126
+    },
+    {
+      "epoch": 1.2657588094903833,
+      "grad_norm": 0.49522134976029414,
+      "learning_rate": 1.860282480625637e-05,
+      "loss": 0.3027,
+      "step": 1127
+    },
+    {
+      "epoch": 1.2668819317703215,
+      "grad_norm": 0.504015437884683,
+      "learning_rate": 1.859882306024189e-05,
+      "loss": 0.3062,
+      "step": 1128
+    },
+    {
+      "epoch": 1.2680050540502597,
+      "grad_norm": 0.4935114092813406,
+      "learning_rate": 1.8594816023276954e-05,
+      "loss": 0.3145,
+      "step": 1129
+    },
+    {
+      "epoch": 1.269128176330198,
+      "grad_norm": 0.45349496634172853,
+      "learning_rate": 1.8590803697827138e-05,
+      "loss": 0.2756,
+      "step": 1130
+    },
+    {
+      "epoch": 1.2702512986101362,
+      "grad_norm": 0.4923320856662289,
+      "learning_rate": 1.8586786086361268e-05,
+      "loss": 0.293,
+      "step": 1131
+    },
+    {
+      "epoch": 1.2713744208900744,
+      "grad_norm": 0.48349777935390137,
+      "learning_rate": 1.8582763191351427e-05,
+      "loss": 0.2881,
+      "step": 1132
+    },
+    {
+      "epoch": 1.2724975431700125,
+      "grad_norm": 0.45227425022413825,
+      "learning_rate": 1.8578735015272947e-05,
+      "loss": 0.2738,
+      "step": 1133
+    },
+    {
+      "epoch": 1.273620665449951,
+      "grad_norm": 0.4796542700286378,
+      "learning_rate": 1.8574701560604405e-05,
+      "loss": 0.2974,
+      "step": 1134
+    },
+    {
+      "epoch": 1.274743787729889,
+      "grad_norm": 0.45414376951278596,
+      "learning_rate": 1.8570662829827632e-05,
+      "loss": 0.3021,
+      "step": 1135
+    },
+    {
+      "epoch": 1.2758669100098272,
+      "grad_norm": 0.4329561559626584,
+      "learning_rate": 1.8566618825427704e-05,
+      "loss": 0.2636,
+      "step": 1136
+    },
+    {
+      "epoch": 1.2769900322897656,
+      "grad_norm": 0.49933859807478426,
+      "learning_rate": 1.8562569549892945e-05,
+      "loss": 0.3034,
+      "step": 1137
+    },
+    {
+      "epoch": 1.2781131545697038,
+      "grad_norm": 0.4531437341854577,
+      "learning_rate": 1.855851500571491e-05,
+      "loss": 0.281,
+      "step": 1138
+    },
+    {
+      "epoch": 1.279236276849642,
+      "grad_norm": 0.4567172782028271,
+      "learning_rate": 1.8554455195388414e-05,
+      "loss": 0.2751,
+      "step": 1139
+    },
+    {
+      "epoch": 1.2803593991295803,
+      "grad_norm": 0.5083738307195607,
+      "learning_rate": 1.8550390121411497e-05,
+      "loss": 0.2964,
+      "step": 1140
+    },
+    {
+      "epoch": 1.2814825214095185,
+      "grad_norm": 0.4599915752167777,
+      "learning_rate": 1.8546319786285443e-05,
+      "loss": 0.2804,
+      "step": 1141
+    },
+    {
+      "epoch": 1.2826056436894566,
+      "grad_norm": 0.47423327374698915,
+      "learning_rate": 1.854224419251478e-05,
+      "loss": 0.3044,
+      "step": 1142
+    },
+    {
+      "epoch": 1.283728765969395,
+      "grad_norm": 0.47354584223797824,
+      "learning_rate": 1.853816334260726e-05,
+      "loss": 0.3022,
+      "step": 1143
+    },
+    {
+      "epoch": 1.2848518882493332,
+      "grad_norm": 0.48012669773887895,
+      "learning_rate": 1.8534077239073877e-05,
+      "loss": 0.291,
+      "step": 1144
+    },
+    {
+      "epoch": 1.2859750105292713,
+      "grad_norm": 0.4835082015975579,
+      "learning_rate": 1.8529985884428855e-05,
+      "loss": 0.2957,
+      "step": 1145
+    },
+    {
+      "epoch": 1.2870981328092097,
+      "grad_norm": 0.44167167547950387,
+      "learning_rate": 1.8525889281189654e-05,
+      "loss": 0.2927,
+      "step": 1146
+    },
+    {
+      "epoch": 1.2882212550891479,
+      "grad_norm": 0.4526647766195691,
+      "learning_rate": 1.8521787431876954e-05,
+      "loss": 0.2801,
+      "step": 1147
+    },
+    {
+      "epoch": 1.289344377369086,
+      "grad_norm": 0.4568245323185659,
+      "learning_rate": 1.8517680339014667e-05,
+      "loss": 0.2648,
+      "step": 1148
+    },
+    {
+      "epoch": 1.2904674996490244,
+      "grad_norm": 0.4883019800865997,
+      "learning_rate": 1.8513568005129937e-05,
+      "loss": 0.2865,
+      "step": 1149
+    },
+    {
+      "epoch": 1.2915906219289626,
+      "grad_norm": 0.47025667774471347,
+      "learning_rate": 1.8509450432753123e-05,
+      "loss": 0.2938,
+      "step": 1150
+    },
+    {
+      "epoch": 1.2927137442089007,
+      "grad_norm": 0.4428247847588197,
+      "learning_rate": 1.8505327624417816e-05,
+      "loss": 0.2732,
+      "step": 1151
+    },
+    {
+      "epoch": 1.293836866488839,
+      "grad_norm": 0.4774238303086379,
+      "learning_rate": 1.8501199582660824e-05,
+      "loss": 0.3058,
+      "step": 1152
+    },
+    {
+      "epoch": 1.2949599887687773,
+      "grad_norm": 0.4948410577106829,
+      "learning_rate": 1.849706631002218e-05,
+      "loss": 0.3044,
+      "step": 1153
+    },
+    {
+      "epoch": 1.2960831110487154,
+      "grad_norm": 0.48963021776942295,
+      "learning_rate": 1.849292780904513e-05,
+      "loss": 0.3007,
+      "step": 1154
+    },
+    {
+      "epoch": 1.2972062333286536,
+      "grad_norm": 0.4741585592952024,
+      "learning_rate": 1.8488784082276137e-05,
+      "loss": 0.2818,
+      "step": 1155
+    },
+    {
+      "epoch": 1.2983293556085918,
+      "grad_norm": 0.4604185684567528,
+      "learning_rate": 1.848463513226488e-05,
+      "loss": 0.2471,
+      "step": 1156
+    },
+    {
+      "epoch": 1.2994524778885301,
+      "grad_norm": 0.4877301769198903,
+      "learning_rate": 1.848048096156426e-05,
+      "loss": 0.2926,
+      "step": 1157
+    },
+    {
+      "epoch": 1.3005756001684683,
+      "grad_norm": 0.4742673297417078,
+      "learning_rate": 1.8476321572730382e-05,
+      "loss": 0.2907,
+      "step": 1158
+    },
+    {
+      "epoch": 1.3016987224484065,
+      "grad_norm": 0.4748985003791319,
+      "learning_rate": 1.847215696832256e-05,
+      "loss": 0.2857,
+      "step": 1159
+    },
+    {
+      "epoch": 1.3028218447283448,
+      "grad_norm": 0.4785822244074492,
+      "learning_rate": 1.8467987150903325e-05,
+      "loss": 0.2896,
+      "step": 1160
+    },
+    {
+      "epoch": 1.303944967008283,
+      "grad_norm": 0.4852477184928872,
+      "learning_rate": 1.846381212303841e-05,
+      "loss": 0.3077,
+      "step": 1161
+    },
+    {
+      "epoch": 1.3050680892882212,
+      "grad_norm": 0.46752327714434433,
+      "learning_rate": 1.8459631887296757e-05,
+      "loss": 0.3059,
+      "step": 1162
+    },
+    {
+      "epoch": 1.3061912115681595,
+      "grad_norm": 0.491665240404064,
+      "learning_rate": 1.8455446446250508e-05,
+      "loss": 0.2982,
+      "step": 1163
+    },
+    {
+      "epoch": 1.3073143338480977,
+      "grad_norm": 0.4481191351234828,
+      "learning_rate": 1.8451255802475014e-05,
+      "loss": 0.2835,
+      "step": 1164
+    },
+    {
+      "epoch": 1.3084374561280359,
+      "grad_norm": 0.4419919772694303,
+      "learning_rate": 1.8447059958548822e-05,
+      "loss": 0.2744,
+      "step": 1165
+    },
+    {
+      "epoch": 1.3095605784079742,
+      "grad_norm": 0.46574040350278373,
+      "learning_rate": 1.8442858917053682e-05,
+      "loss": 0.291,
+      "step": 1166
+    },
+    {
+      "epoch": 1.3106837006879124,
+      "grad_norm": 0.43091837390173826,
+      "learning_rate": 1.843865268057454e-05,
+      "loss": 0.2657,
+      "step": 1167
+    },
+    {
+      "epoch": 1.3118068229678506,
+      "grad_norm": 0.4642199123609139,
+      "learning_rate": 1.843444125169954e-05,
+      "loss": 0.2876,
+      "step": 1168
+    },
+    {
+      "epoch": 1.312929945247789,
+      "grad_norm": 0.457887008950542,
+      "learning_rate": 1.843022463302002e-05,
+      "loss": 0.2727,
+      "step": 1169
+    },
+    {
+      "epoch": 1.314053067527727,
+      "grad_norm": 0.44022972541172534,
+      "learning_rate": 1.8426002827130517e-05,
+      "loss": 0.2892,
+      "step": 1170
+    },
+    {
+      "epoch": 1.3151761898076653,
+      "grad_norm": 0.4621964331966014,
+      "learning_rate": 1.842177583662875e-05,
+      "loss": 0.3,
+      "step": 1171
+    },
+    {
+      "epoch": 1.3162993120876036,
+      "grad_norm": 0.44527517838102726,
+      "learning_rate": 1.8417543664115632e-05,
+      "loss": 0.286,
+      "step": 1172
+    },
+    {
+      "epoch": 1.3174224343675418,
+      "grad_norm": 0.455981271490681,
+      "learning_rate": 1.8413306312195265e-05,
+      "loss": 0.2785,
+      "step": 1173
+    },
+    {
+      "epoch": 1.31854555664748,
+      "grad_norm": 0.47535325172610315,
+      "learning_rate": 1.840906378347494e-05,
+      "loss": 0.3106,
+      "step": 1174
+    },
+    {
+      "epoch": 1.3196686789274183,
+      "grad_norm": 0.4559106843629929,
+      "learning_rate": 1.8404816080565133e-05,
+      "loss": 0.2955,
+      "step": 1175
+    },
+    {
+      "epoch": 1.3207918012073565,
+      "grad_norm": 0.4772186174990515,
+      "learning_rate": 1.84005632060795e-05,
+      "loss": 0.2859,
+      "step": 1176
+    },
+    {
+      "epoch": 1.3219149234872947,
+      "grad_norm": 0.5126527615592691,
+      "learning_rate": 1.8396305162634885e-05,
+      "loss": 0.3171,
+      "step": 1177
+    },
+    {
+      "epoch": 1.3230380457672328,
+      "grad_norm": 0.4863596644425538,
+      "learning_rate": 1.83920419528513e-05,
+      "loss": 0.3076,
+      "step": 1178
+    },
+    {
+      "epoch": 1.3241611680471712,
+      "grad_norm": 0.4329722715025753,
+      "learning_rate": 1.838777357935196e-05,
+      "loss": 0.2594,
+      "step": 1179
+    },
+    {
+      "epoch": 1.3252842903271094,
+      "grad_norm": 0.4735307567938337,
+      "learning_rate": 1.8383500044763226e-05,
+      "loss": 0.2752,
+      "step": 1180
+    },
+    {
+      "epoch": 1.3264074126070475,
+      "grad_norm": 0.4708270647203179,
+      "learning_rate": 1.837922135171466e-05,
+      "loss": 0.3007,
+      "step": 1181
+    },
+    {
+      "epoch": 1.3275305348869857,
+      "grad_norm": 0.4328720336288202,
+      "learning_rate": 1.837493750283899e-05,
+      "loss": 0.2743,
+      "step": 1182
+    },
+    {
+      "epoch": 1.328653657166924,
+      "grad_norm": 0.45831796385684703,
+      "learning_rate": 1.8370648500772107e-05,
+      "loss": 0.2961,
+      "step": 1183
+    },
+    {
+      "epoch": 1.3297767794468622,
+      "grad_norm": 0.4507244345775207,
+      "learning_rate": 1.836635434815309e-05,
+      "loss": 0.2854,
+      "step": 1184
+    },
+    {
+      "epoch": 1.3308999017268004,
+      "grad_norm": 0.454897865730844,
+      "learning_rate": 1.8362055047624175e-05,
+      "loss": 0.2685,
+      "step": 1185
+    },
+    {
+      "epoch": 1.3320230240067388,
+      "grad_norm": 0.46791986097149424,
+      "learning_rate": 1.835775060183077e-05,
+      "loss": 0.2824,
+      "step": 1186
+    },
+    {
+      "epoch": 1.333146146286677,
+      "grad_norm": 0.47734391363275325,
+      "learning_rate": 1.8353441013421445e-05,
+      "loss": 0.2919,
+      "step": 1187
+    },
+    {
+      "epoch": 1.334269268566615,
+      "grad_norm": 0.4525116496155753,
+      "learning_rate": 1.8349126285047937e-05,
+      "loss": 0.2897,
+      "step": 1188
+    },
+    {
+      "epoch": 1.3353923908465535,
+      "grad_norm": 0.46057547898872325,
+      "learning_rate": 1.8344806419365152e-05,
+      "loss": 0.2959,
+      "step": 1189
+    },
+    {
+      "epoch": 1.3365155131264916,
+      "grad_norm": 0.43961943911433504,
+      "learning_rate": 1.8340481419031146e-05,
+      "loss": 0.2717,
+      "step": 1190
+    },
+    {
+      "epoch": 1.3376386354064298,
+      "grad_norm": 0.46908924741585806,
+      "learning_rate": 1.833615128670714e-05,
+      "loss": 0.2921,
+      "step": 1191
+    },
+    {
+      "epoch": 1.3387617576863682,
+      "grad_norm": 0.4619903044568609,
+      "learning_rate": 1.8331816025057508e-05,
+      "loss": 0.2898,
+      "step": 1192
+    },
+    {
+      "epoch": 1.3398848799663063,
+      "grad_norm": 0.44532555047844774,
+      "learning_rate": 1.8327475636749793e-05,
+      "loss": 0.2824,
+      "step": 1193
+    },
+    {
+      "epoch": 1.3410080022462445,
+      "grad_norm": 0.4770456698496007,
+      "learning_rate": 1.8323130124454676e-05,
+      "loss": 0.2934,
+      "step": 1194
+    },
+    {
+      "epoch": 1.3421311245261829,
+      "grad_norm": 0.5013999364371049,
+      "learning_rate": 1.8318779490846005e-05,
+      "loss": 0.3106,
+      "step": 1195
+    },
+    {
+      "epoch": 1.343254246806121,
+      "grad_norm": 0.4462481861250692,
+      "learning_rate": 1.8314423738600765e-05,
+      "loss": 0.2916,
+      "step": 1196
+    },
+    {
+      "epoch": 1.3443773690860592,
+      "grad_norm": 0.4544363504817102,
+      "learning_rate": 1.8310062870399105e-05,
+      "loss": 0.2768,
+      "step": 1197
+    },
+    {
+      "epoch": 1.3455004913659976,
+      "grad_norm": 0.4739036937639808,
+      "learning_rate": 1.8305696888924312e-05,
+      "loss": 0.2818,
+      "step": 1198
+    },
+    {
+      "epoch": 1.3466236136459357,
+      "grad_norm": 0.4251780111412686,
+      "learning_rate": 1.8301325796862825e-05,
+      "loss": 0.2618,
+      "step": 1199
+    },
+    {
+      "epoch": 1.347746735925874,
+      "grad_norm": 0.4905216280674908,
+      "learning_rate": 1.829694959690422e-05,
+      "loss": 0.2881,
+      "step": 1200
+    },
+    {
+      "epoch": 1.3488698582058123,
+      "grad_norm": 0.47365961695255354,
+      "learning_rate": 1.8292568291741228e-05,
+      "loss": 0.2954,
+      "step": 1201
+    },
+    {
+      "epoch": 1.3499929804857504,
+      "grad_norm": 0.46299602153786523,
+      "learning_rate": 1.8288181884069707e-05,
+      "loss": 0.284,
+      "step": 1202
+    },
+    {
+      "epoch": 1.3511161027656886,
+      "grad_norm": 0.44021839710368676,
+      "learning_rate": 1.828379037658867e-05,
+      "loss": 0.2685,
+      "step": 1203
+    },
+    {
+      "epoch": 1.3522392250456268,
+      "grad_norm": 0.5131025179704897,
+      "learning_rate": 1.827939377200025e-05,
+      "loss": 0.3053,
+      "step": 1204
+    },
+    {
+      "epoch": 1.3533623473255652,
+      "grad_norm": 0.44547118449274636,
+      "learning_rate": 1.8274992073009736e-05,
+      "loss": 0.289,
+      "step": 1205
+    },
+    {
+      "epoch": 1.3544854696055033,
+      "grad_norm": 0.44524428838589575,
+      "learning_rate": 1.827058528232553e-05,
+      "loss": 0.2849,
+      "step": 1206
+    },
+    {
+      "epoch": 1.3556085918854415,
+      "grad_norm": 0.4649075563603362,
+      "learning_rate": 1.8266173402659193e-05,
+      "loss": 0.2807,
+      "step": 1207
+    },
+    {
+      "epoch": 1.3567317141653796,
+      "grad_norm": 0.48613198722949613,
+      "learning_rate": 1.826175643672539e-05,
+      "loss": 0.2949,
+      "step": 1208
+    },
+    {
+      "epoch": 1.357854836445318,
+      "grad_norm": 0.43266914602095147,
+      "learning_rate": 1.8257334387241944e-05,
+      "loss": 0.2712,
+      "step": 1209
+    },
+    {
+      "epoch": 1.3589779587252562,
+      "grad_norm": 0.4612767862308836,
+      "learning_rate": 1.8252907256929777e-05,
+      "loss": 0.2802,
+      "step": 1210
+    },
+    {
+      "epoch": 1.3601010810051943,
+      "grad_norm": 0.48733069495425535,
+      "learning_rate": 1.8248475048512956e-05,
+      "loss": 0.2995,
+      "step": 1211
+    },
+    {
+      "epoch": 1.3612242032851327,
+      "grad_norm": 0.4633919892138498,
+      "learning_rate": 1.8244037764718666e-05,
+      "loss": 0.2945,
+      "step": 1212
+    },
+    {
+      "epoch": 1.3623473255650709,
+      "grad_norm": 0.48010248333965416,
+      "learning_rate": 1.8239595408277216e-05,
+      "loss": 0.2702,
+      "step": 1213
+    },
+    {
+      "epoch": 1.363470447845009,
+      "grad_norm": 0.44254247447818046,
+      "learning_rate": 1.8235147981922042e-05,
+      "loss": 0.2797,
+      "step": 1214
+    },
+    {
+      "epoch": 1.3645935701249474,
+      "grad_norm": 0.4685652855739763,
+      "learning_rate": 1.8230695488389688e-05,
+      "loss": 0.2994,
+      "step": 1215
+    },
+    {
+      "epoch": 1.3657166924048856,
+      "grad_norm": 0.46157477459451246,
+      "learning_rate": 1.822623793041983e-05,
+      "loss": 0.2945,
+      "step": 1216
+    },
+    {
+      "epoch": 1.3668398146848237,
+      "grad_norm": 0.4509218965365652,
+      "learning_rate": 1.8221775310755247e-05,
+      "loss": 0.286,
+      "step": 1217
+    },
+    {
+      "epoch": 1.3679629369647621,
+      "grad_norm": 0.4584479264400474,
+      "learning_rate": 1.8217307632141835e-05,
+      "loss": 0.2933,
+      "step": 1218
+    },
+    {
+      "epoch": 1.3690860592447003,
+      "grad_norm": 0.4985130465087838,
+      "learning_rate": 1.8212834897328614e-05,
+      "loss": 0.3045,
+      "step": 1219
+    },
+    {
+      "epoch": 1.3702091815246384,
+      "grad_norm": 0.47337122837553836,
+      "learning_rate": 1.82083571090677e-05,
+      "loss": 0.2836,
+      "step": 1220
+    },
+    {
+      "epoch": 1.3713323038045768,
+      "grad_norm": 0.45448456689959993,
+      "learning_rate": 1.8203874270114327e-05,
+      "loss": 0.2708,
+      "step": 1221
+    },
+    {
+      "epoch": 1.372455426084515,
+      "grad_norm": 0.4893145984532352,
+      "learning_rate": 1.8199386383226835e-05,
+      "loss": 0.2826,
+      "step": 1222
+    },
+    {
+      "epoch": 1.3735785483644531,
+      "grad_norm": 0.44417991047867217,
+      "learning_rate": 1.8194893451166673e-05,
+      "loss": 0.2821,
+      "step": 1223
+    },
+    {
+      "epoch": 1.3747016706443915,
+      "grad_norm": 0.44598812540160476,
+      "learning_rate": 1.819039547669839e-05,
+      "loss": 0.2774,
+      "step": 1224
+    },
+    {
+      "epoch": 1.3758247929243297,
+      "grad_norm": 0.4776707667120244,
+      "learning_rate": 1.818589246258964e-05,
+      "loss": 0.3122,
+      "step": 1225
+    },
+    {
+      "epoch": 1.3769479152042678,
+      "grad_norm": 0.5082949999474378,
+      "learning_rate": 1.8181384411611173e-05,
+      "loss": 0.2948,
+      "step": 1226
+    },
+    {
+      "epoch": 1.3780710374842062,
+      "grad_norm": 0.4424043135503394,
+      "learning_rate": 1.817687132653685e-05,
+      "loss": 0.2743,
+      "step": 1227
+    },
+    {
+      "epoch": 1.3791941597641444,
+      "grad_norm": 0.48246812847596726,
+      "learning_rate": 1.8172353210143613e-05,
+      "loss": 0.3154,
+      "step": 1228
+    },
+    {
+      "epoch": 1.3803172820440825,
+      "grad_norm": 0.4637624706766306,
+      "learning_rate": 1.8167830065211513e-05,
+      "loss": 0.3047,
+      "step": 1229
+    },
+    {
+      "epoch": 1.3814404043240207,
+      "grad_norm": 0.47224752107030954,
+      "learning_rate": 1.8163301894523695e-05,
+      "loss": 0.3023,
+      "step": 1230
+    },
+    {
+      "epoch": 1.3825635266039589,
+      "grad_norm": 0.4698090761389382,
+      "learning_rate": 1.8158768700866386e-05,
+      "loss": 0.2635,
+      "step": 1231
+    },
+    {
+      "epoch": 1.3836866488838973,
+      "grad_norm": 0.4957909790614304,
+      "learning_rate": 1.8154230487028913e-05,
+      "loss": 0.2928,
+      "step": 1232
+    },
+    {
+      "epoch": 1.3848097711638354,
+      "grad_norm": 0.44768649725073345,
+      "learning_rate": 1.8149687255803687e-05,
+      "loss": 0.2874,
+      "step": 1233
+    },
+    {
+      "epoch": 1.3859328934437736,
+      "grad_norm": 0.4758485891594923,
+      "learning_rate": 1.814513900998621e-05,
+      "loss": 0.2996,
+      "step": 1234
+    },
+    {
+      "epoch": 1.387056015723712,
+      "grad_norm": 0.47272282269055016,
+      "learning_rate": 1.8140585752375063e-05,
+      "loss": 0.2884,
+      "step": 1235
+    },
+    {
+      "epoch": 1.3881791380036501,
+      "grad_norm": 0.45230047857816796,
+      "learning_rate": 1.8136027485771926e-05,
+      "loss": 0.2767,
+      "step": 1236
+    },
+    {
+      "epoch": 1.3893022602835883,
+      "grad_norm": 0.4507150818799199,
+      "learning_rate": 1.813146421298154e-05,
+      "loss": 0.2884,
+      "step": 1237
+    },
+    {
+      "epoch": 1.3904253825635267,
+      "grad_norm": 0.45879614204982117,
+      "learning_rate": 1.8126895936811745e-05,
+      "loss": 0.2708,
+      "step": 1238
+    },
+    {
+      "epoch": 1.3915485048434648,
+      "grad_norm": 0.45661246563867913,
+      "learning_rate": 1.812232266007344e-05,
+      "loss": 0.2737,
+      "step": 1239
+    },
+    {
+      "epoch": 1.392671627123403,
+      "grad_norm": 0.4601276790260715,
+      "learning_rate": 1.8117744385580627e-05,
+      "loss": 0.2938,
+      "step": 1240
+    },
+    {
+      "epoch": 1.3937947494033414,
+      "grad_norm": 0.48293515299373707,
+      "learning_rate": 1.8113161116150356e-05,
+      "loss": 0.2847,
+      "step": 1241
+    },
+    {
+      "epoch": 1.3949178716832795,
+      "grad_norm": 0.5135873286633734,
+      "learning_rate": 1.8108572854602774e-05,
+      "loss": 0.3047,
+      "step": 1242
+    },
+    {
+      "epoch": 1.3960409939632177,
+      "grad_norm": 0.49511291405284924,
+      "learning_rate": 1.8103979603761084e-05,
+      "loss": 0.296,
+      "step": 1243
+    },
+    {
+      "epoch": 1.397164116243156,
+      "grad_norm": 0.4915844024660979,
+      "learning_rate": 1.8099381366451562e-05,
+      "loss": 0.3025,
+      "step": 1244
+    },
+    {
+      "epoch": 1.3982872385230942,
+      "grad_norm": 0.457360486981577,
+      "learning_rate": 1.8094778145503555e-05,
+      "loss": 0.2809,
+      "step": 1245
+    },
+    {
+      "epoch": 1.3994103608030324,
+      "grad_norm": 0.45194447137874866,
+      "learning_rate": 1.8090169943749477e-05,
+      "loss": 0.2815,
+      "step": 1246
+    },
+    {
+      "epoch": 1.4005334830829708,
+      "grad_norm": 0.469717199113566,
+      "learning_rate": 1.8085556764024804e-05,
+      "loss": 0.2844,
+      "step": 1247
+    },
+    {
+      "epoch": 1.401656605362909,
+      "grad_norm": 0.461335245541827,
+      "learning_rate": 1.8080938609168073e-05,
+      "loss": 0.2825,
+      "step": 1248
+    },
+    {
+      "epoch": 1.402779727642847,
+      "grad_norm": 0.4710994100251086,
+      "learning_rate": 1.8076315482020893e-05,
+      "loss": 0.2925,
+      "step": 1249
+    },
+    {
+      "epoch": 1.4039028499227855,
+      "grad_norm": 0.46856919190627966,
+      "learning_rate": 1.8071687385427922e-05,
+      "loss": 0.2925,
+      "step": 1250
+    },
+    {
+      "epoch": 1.4050259722027236,
+      "grad_norm": 0.463686961189385,
+      "learning_rate": 1.8067054322236876e-05,
+      "loss": 0.2982,
+      "step": 1251
+    },
+    {
+      "epoch": 1.4061490944826618,
+      "grad_norm": 0.4233668907762543,
+      "learning_rate": 1.806241629529853e-05,
+      "loss": 0.279,
+      "step": 1252
+    },
+    {
+      "epoch": 1.4072722167626,
+      "grad_norm": 0.4446703109276029,
+      "learning_rate": 1.8057773307466717e-05,
+      "loss": 0.2688,
+      "step": 1253
+    },
+    {
+      "epoch": 1.4083953390425383,
+      "grad_norm": 0.4485358287975487,
+      "learning_rate": 1.8053125361598314e-05,
+      "loss": 0.2947,
+      "step": 1254
+    },
+    {
+      "epoch": 1.4095184613224765,
+      "grad_norm": 0.4452525960821962,
+      "learning_rate": 1.804847246055326e-05,
+      "loss": 0.2916,
+      "step": 1255
+    },
+    {
+      "epoch": 1.4106415836024147,
+      "grad_norm": 0.44434990069489066,
+      "learning_rate": 1.8043814607194528e-05,
+      "loss": 0.2957,
+      "step": 1256
+    },
+    {
+      "epoch": 1.4117647058823528,
+      "grad_norm": 0.45782505420166125,
+      "learning_rate": 1.803915180438815e-05,
+      "loss": 0.3081,
+      "step": 1257
+    },
+    {
+      "epoch": 1.4128878281622912,
+      "grad_norm": 0.44812612511388755,
+      "learning_rate": 1.80344840550032e-05,
+      "loss": 0.2925,
+      "step": 1258
+    },
+    {
+      "epoch": 1.4140109504422294,
+      "grad_norm": 0.4653220459251436,
+      "learning_rate": 1.8029811361911796e-05,
+      "loss": 0.284,
+      "step": 1259
+    },
+    {
+      "epoch": 1.4151340727221675,
+      "grad_norm": 0.4294806313427792,
+      "learning_rate": 1.8025133727989095e-05,
+      "loss": 0.2615,
+      "step": 1260
+    },
+    {
+      "epoch": 1.416257195002106,
+      "grad_norm": 0.4722245406377723,
+      "learning_rate": 1.8020451156113302e-05,
+      "loss": 0.2932,
+      "step": 1261
+    },
+    {
+      "epoch": 1.417380317282044,
+      "grad_norm": 0.4407259836445125,
+      "learning_rate": 1.801576364916565e-05,
+      "loss": 0.2783,
+      "step": 1262
+    },
+    {
+      "epoch": 1.4185034395619822,
+      "grad_norm": 0.4554354849309488,
+      "learning_rate": 1.8011071210030417e-05,
+      "loss": 0.2934,
+      "step": 1263
+    },
+    {
+      "epoch": 1.4196265618419206,
+      "grad_norm": 0.4797437708219925,
+      "learning_rate": 1.8006373841594905e-05,
+      "loss": 0.2914,
+      "step": 1264
+    },
+    {
+      "epoch": 1.4207496841218588,
+      "grad_norm": 0.4453109721657388,
+      "learning_rate": 1.8001671546749466e-05,
+      "loss": 0.273,
+      "step": 1265
+    },
+    {
+      "epoch": 1.421872806401797,
+      "grad_norm": 0.455616539511479,
+      "learning_rate": 1.7996964328387473e-05,
+      "loss": 0.2787,
+      "step": 1266
+    },
+    {
+      "epoch": 1.4229959286817353,
+      "grad_norm": 0.4660177496804102,
+      "learning_rate": 1.7992252189405318e-05,
+      "loss": 0.288,
+      "step": 1267
+    },
+    {
+      "epoch": 1.4241190509616735,
+      "grad_norm": 0.4868499349359341,
+      "learning_rate": 1.798753513270245e-05,
+      "loss": 0.3011,
+      "step": 1268
+    },
+    {
+      "epoch": 1.4252421732416116,
+      "grad_norm": 0.42379910255068853,
+      "learning_rate": 1.798281316118131e-05,
+      "loss": 0.2688,
+      "step": 1269
+    },
+    {
+      "epoch": 1.42636529552155,
+      "grad_norm": 0.44137517215745187,
+      "learning_rate": 1.797808627774738e-05,
+      "loss": 0.2779,
+      "step": 1270
+    },
+    {
+      "epoch": 1.4274884178014882,
+      "grad_norm": 0.4454397017855999,
+      "learning_rate": 1.7973354485309178e-05,
+      "loss": 0.2885,
+      "step": 1271
+    },
+    {
+      "epoch": 1.4286115400814263,
+      "grad_norm": 0.41513228116253365,
+      "learning_rate": 1.7968617786778214e-05,
+      "loss": 0.258,
+      "step": 1272
+    },
+    {
+      "epoch": 1.4297346623613647,
+      "grad_norm": 0.45824072542709143,
+      "learning_rate": 1.7963876185069032e-05,
+      "loss": 0.2982,
+      "step": 1273
+    },
+    {
+      "epoch": 1.4308577846413029,
+      "grad_norm": 0.45152285012254323,
+      "learning_rate": 1.7959129683099202e-05,
+      "loss": 0.2926,
+      "step": 1274
+    },
+    {
+      "epoch": 1.431980906921241,
+      "grad_norm": 0.4382850255611188,
+      "learning_rate": 1.795437828378929e-05,
+      "loss": 0.2748,
+      "step": 1275
+    },
+    {
+      "epoch": 1.4331040292011794,
+      "grad_norm": 0.45046526339966464,
+      "learning_rate": 1.7949621990062882e-05,
+      "loss": 0.2834,
+      "step": 1276
+    },
+    {
+      "epoch": 1.4342271514811176,
+      "grad_norm": 0.45318087991408496,
+      "learning_rate": 1.7944860804846585e-05,
+      "loss": 0.2921,
+      "step": 1277
+    },
+    {
+      "epoch": 1.4353502737610557,
+      "grad_norm": 0.4389584828107915,
+      "learning_rate": 1.7940094731070005e-05,
+      "loss": 0.2711,
+      "step": 1278
+    },
+    {
+      "epoch": 1.4364733960409939,
+      "grad_norm": 0.4551598275028435,
+      "learning_rate": 1.793532377166576e-05,
+      "loss": 0.2788,
+      "step": 1279
+    },
+    {
+      "epoch": 1.4375965183209323,
+      "grad_norm": 0.46450024527966405,
+      "learning_rate": 1.793054792956947e-05,
+      "loss": 0.2937,
+      "step": 1280
+    },
+    {
+      "epoch": 1.4387196406008704,
+      "grad_norm": 0.4631044556099893,
+      "learning_rate": 1.7925767207719774e-05,
+      "loss": 0.2961,
+      "step": 1281
+    },
+    {
+      "epoch": 1.4398427628808086,
+      "grad_norm": 0.4292656254224395,
+      "learning_rate": 1.792098160905829e-05,
+      "loss": 0.2646,
+      "step": 1282
+    },
+    {
+      "epoch": 1.4409658851607468,
+      "grad_norm": 0.47873898790679503,
+      "learning_rate": 1.791619113652966e-05,
+      "loss": 0.2891,
+      "step": 1283
+    },
+    {
+      "epoch": 1.4420890074406851,
+      "grad_norm": 0.46176874055866274,
+      "learning_rate": 1.7911395793081508e-05,
+      "loss": 0.3053,
+      "step": 1284
+    },
+    {
+      "epoch": 1.4432121297206233,
+      "grad_norm": 0.43168104217807524,
+      "learning_rate": 1.7906595581664462e-05,
+      "loss": 0.2757,
+      "step": 1285
+    },
+    {
+      "epoch": 1.4443352520005615,
+      "grad_norm": 0.4526156341532352,
+      "learning_rate": 1.790179050523215e-05,
+      "loss": 0.2906,
+      "step": 1286
+    },
+    {
+      "epoch": 1.4454583742804998,
+      "grad_norm": 0.4576287167509249,
+      "learning_rate": 1.7896980566741183e-05,
+      "loss": 0.3055,
+      "step": 1287
+    },
+    {
+      "epoch": 1.446581496560438,
+      "grad_norm": 0.4291349675446234,
+      "learning_rate": 1.7892165769151174e-05,
+      "loss": 0.2797,
+      "step": 1288
+    },
+    {
+      "epoch": 1.4477046188403762,
+      "grad_norm": 0.4330178308786235,
+      "learning_rate": 1.7887346115424712e-05,
+      "loss": 0.2879,
+      "step": 1289
+    },
+    {
+      "epoch": 1.4488277411203145,
+      "grad_norm": 0.4409857238545325,
+      "learning_rate": 1.7882521608527393e-05,
+      "loss": 0.277,
+      "step": 1290
+    },
+    {
+      "epoch": 1.4499508634002527,
+      "grad_norm": 0.46396778078552625,
+      "learning_rate": 1.7877692251427783e-05,
+      "loss": 0.292,
+      "step": 1291
+    },
+    {
+      "epoch": 1.4510739856801909,
+      "grad_norm": 0.4583722733176541,
+      "learning_rate": 1.7872858047097442e-05,
+      "loss": 0.287,
+      "step": 1292
+    },
+    {
+      "epoch": 1.4521971079601292,
+      "grad_norm": 0.4617901289674729,
+      "learning_rate": 1.7868018998510907e-05,
+      "loss": 0.2931,
+      "step": 1293
+    },
+    {
+      "epoch": 1.4533202302400674,
+      "grad_norm": 0.4688919751696024,
+      "learning_rate": 1.7863175108645698e-05,
+      "loss": 0.2795,
+      "step": 1294
+    },
+    {
+      "epoch": 1.4544433525200056,
+      "grad_norm": 0.473513507634666,
+      "learning_rate": 1.7858326380482313e-05,
+      "loss": 0.3136,
+      "step": 1295
+    },
+    {
+      "epoch": 1.455566474799944,
+      "grad_norm": 0.46308815674849557,
+      "learning_rate": 1.7853472817004235e-05,
+      "loss": 0.2871,
+      "step": 1296
+    },
+    {
+      "epoch": 1.456689597079882,
+      "grad_norm": 0.5179959894108842,
+      "learning_rate": 1.7848614421197903e-05,
+      "loss": 0.3007,
+      "step": 1297
+    },
+    {
+      "epoch": 1.4578127193598203,
+      "grad_norm": 0.4531340269397126,
+      "learning_rate": 1.784375119605275e-05,
+      "loss": 0.2959,
+      "step": 1298
+    },
+    {
+      "epoch": 1.4589358416397586,
+      "grad_norm": 0.4582919369256917,
+      "learning_rate": 1.783888314456117e-05,
+      "loss": 0.2894,
+      "step": 1299
+    },
+    {
+      "epoch": 1.4600589639196968,
+      "grad_norm": 0.4469673584516157,
+      "learning_rate": 1.7834010269718526e-05,
+      "loss": 0.28,
+      "step": 1300
+    },
+    {
+      "epoch": 1.461182086199635,
+      "grad_norm": 0.46123285741882053,
+      "learning_rate": 1.7829132574523155e-05,
+      "loss": 0.2832,
+      "step": 1301
+    },
+    {
+      "epoch": 1.4623052084795733,
+      "grad_norm": 0.45607801195534514,
+      "learning_rate": 1.7824250061976355e-05,
+      "loss": 0.287,
+      "step": 1302
+    },
+    {
+      "epoch": 1.4634283307595115,
+      "grad_norm": 0.49530717522953166,
+      "learning_rate": 1.7819362735082392e-05,
+      "loss": 0.2993,
+      "step": 1303
+    },
+    {
+      "epoch": 1.4645514530394497,
+      "grad_norm": 0.4809186057767541,
+      "learning_rate": 1.7814470596848486e-05,
+      "loss": 0.3111,
+      "step": 1304
+    },
+    {
+      "epoch": 1.4656745753193878,
+      "grad_norm": 0.4609939280068489,
+      "learning_rate": 1.780957365028483e-05,
+      "loss": 0.3016,
+      "step": 1305
+    },
+    {
+      "epoch": 1.466797697599326,
+      "grad_norm": 0.5159740391114742,
+      "learning_rate": 1.7804671898404567e-05,
+      "loss": 0.303,
+      "step": 1306
+    },
+    {
+      "epoch": 1.4679208198792644,
+      "grad_norm": 0.45316092784328266,
+      "learning_rate": 1.7799765344223798e-05,
+      "loss": 0.2806,
+      "step": 1307
+    },
+    {
+      "epoch": 1.4690439421592025,
+      "grad_norm": 0.4929344112281057,
+      "learning_rate": 1.7794853990761576e-05,
+      "loss": 0.3244,
+      "step": 1308
+    },
+    {
+      "epoch": 1.4701670644391407,
+      "grad_norm": 0.4240215914027117,
+      "learning_rate": 1.778993784103992e-05,
+      "loss": 0.2649,
+      "step": 1309
+    },
+    {
+      "epoch": 1.471290186719079,
+      "grad_norm": 0.45336019964413365,
+      "learning_rate": 1.7785016898083786e-05,
+      "loss": 0.281,
+      "step": 1310
+    },
+    {
+      "epoch": 1.4724133089990172,
+      "grad_norm": 0.4709195975090023,
+      "learning_rate": 1.778009116492108e-05,
+      "loss": 0.3032,
+      "step": 1311
+    },
+    {
+      "epoch": 1.4735364312789554,
+      "grad_norm": 0.4378426156461344,
+      "learning_rate": 1.7775160644582667e-05,
+      "loss": 0.2773,
+      "step": 1312
+    },
+    {
+      "epoch": 1.4746595535588938,
+      "grad_norm": 0.4413617070461194,
+      "learning_rate": 1.777022534010235e-05,
+      "loss": 0.295,
+      "step": 1313
+    },
+    {
+      "epoch": 1.475782675838832,
+      "grad_norm": 0.45700233268388313,
+      "learning_rate": 1.776528525451687e-05,
+      "loss": 0.2909,
+      "step": 1314
+    },
+    {
+      "epoch": 1.47690579811877,
+      "grad_norm": 0.45427285254943417,
+      "learning_rate": 1.776034039086592e-05,
+      "loss": 0.3058,
+      "step": 1315
+    },
+    {
+      "epoch": 1.4780289203987085,
+      "grad_norm": 0.45339655669768186,
+      "learning_rate": 1.775539075219213e-05,
+      "loss": 0.2792,
+      "step": 1316
+    },
+    {
+      "epoch": 1.4791520426786466,
+      "grad_norm": 0.47160522784360365,
+      "learning_rate": 1.7750436341541066e-05,
+      "loss": 0.2998,
+      "step": 1317
+    },
+    {
+      "epoch": 1.4802751649585848,
+      "grad_norm": 0.45954229553978265,
+      "learning_rate": 1.774547716196123e-05,
+      "loss": 0.2871,
+      "step": 1318
+    },
+    {
+      "epoch": 1.4813982872385232,
+      "grad_norm": 0.46825968988664846,
+      "learning_rate": 1.7740513216504064e-05,
+      "loss": 0.3125,
+      "step": 1319
+    },
+    {
+      "epoch": 1.4825214095184613,
+      "grad_norm": 0.4514724212464635,
+      "learning_rate": 1.7735544508223933e-05,
+      "loss": 0.2851,
+      "step": 1320
+    },
+    {
+      "epoch": 1.4836445317983995,
+      "grad_norm": 0.45647208589810134,
+      "learning_rate": 1.773057104017814e-05,
+      "loss": 0.2871,
+      "step": 1321
+    },
+    {
+      "epoch": 1.4847676540783379,
+      "grad_norm": 0.45834495190642427,
+      "learning_rate": 1.772559281542692e-05,
+      "loss": 0.2929,
+      "step": 1322
+    },
+    {
+      "epoch": 1.485890776358276,
+      "grad_norm": 0.4614837916910236,
+      "learning_rate": 1.7720609837033417e-05,
+      "loss": 0.2978,
+      "step": 1323
+    },
+    {
+      "epoch": 1.4870138986382142,
+      "grad_norm": 0.44184030726899937,
+      "learning_rate": 1.7715622108063725e-05,
+      "loss": 0.2663,
+      "step": 1324
+    },
+    {
+      "epoch": 1.4881370209181526,
+      "grad_norm": 0.44711212739545164,
+      "learning_rate": 1.771062963158684e-05,
+      "loss": 0.2968,
+      "step": 1325
+    },
+    {
+      "epoch": 1.4892601431980907,
+      "grad_norm": 0.44221975823221676,
+      "learning_rate": 1.770563241067469e-05,
+      "loss": 0.2952,
+      "step": 1326
+    },
+    {
+      "epoch": 1.490383265478029,
+      "grad_norm": 0.5126809383713152,
+      "learning_rate": 1.7700630448402125e-05,
+      "loss": 0.3161,
+      "step": 1327
+    },
+    {
+      "epoch": 1.491506387757967,
+      "grad_norm": 0.4410073711767202,
+      "learning_rate": 1.76956237478469e-05,
+      "loss": 0.2738,
+      "step": 1328
+    },
+    {
+      "epoch": 1.4926295100379054,
+      "grad_norm": 0.45627309597281607,
+      "learning_rate": 1.7690612312089702e-05,
+      "loss": 0.2853,
+      "step": 1329
+    },
+    {
+      "epoch": 1.4937526323178436,
+      "grad_norm": 0.5028458766464338,
+      "learning_rate": 1.768559614421411e-05,
+      "loss": 0.3025,
+      "step": 1330
+    },
+    {
+      "epoch": 1.4948757545977818,
+      "grad_norm": 0.45766426852181796,
+      "learning_rate": 1.768057524730664e-05,
+      "loss": 0.293,
+      "step": 1331
+    },
+    {
+      "epoch": 1.49599887687772,
+      "grad_norm": 0.45597657632110705,
+      "learning_rate": 1.7675549624456695e-05,
+      "loss": 0.3038,
+      "step": 1332
+    },
+    {
+      "epoch": 1.4971219991576583,
+      "grad_norm": 0.4499374531210488,
+      "learning_rate": 1.7670519278756603e-05,
+      "loss": 0.2815,
+      "step": 1333
+    },
+    {
+      "epoch": 1.4982451214375965,
+      "grad_norm": 0.4707335152594058,
+      "learning_rate": 1.7665484213301587e-05,
+      "loss": 0.303,
+      "step": 1334
+    },
+    {
+      "epoch": 1.4993682437175346,
+      "grad_norm": 0.4781982752780198,
+      "learning_rate": 1.766044443118978e-05,
+      "loss": 0.2965,
+      "step": 1335
+    },
+    {
+      "epoch": 1.500491365997473,
+      "grad_norm": 0.47356503106211406,
+      "learning_rate": 1.7655399935522216e-05,
+      "loss": 0.2957,
+      "step": 1336
+    },
+    {
+      "epoch": 1.5016144882774112,
+      "grad_norm": 0.46119467696838445,
+      "learning_rate": 1.765035072940283e-05,
+      "loss": 0.2902,
+      "step": 1337
+    },
+    {
+      "epoch": 1.5027376105573493,
+      "grad_norm": 0.44367007330882113,
+      "learning_rate": 1.764529681593845e-05,
+      "loss": 0.2745,
+      "step": 1338
+    },
+    {
+      "epoch": 1.5038607328372877,
+      "grad_norm": 0.46264595602282504,
+      "learning_rate": 1.7640238198238803e-05,
+      "loss": 0.2712,
+      "step": 1339
+    },
+    {
+      "epoch": 1.5049838551172259,
+      "grad_norm": 0.44037392967410266,
+      "learning_rate": 1.763517487941652e-05,
+      "loss": 0.2778,
+      "step": 1340
+    },
+    {
+      "epoch": 1.506106977397164,
+      "grad_norm": 0.4361452264395535,
+      "learning_rate": 1.763010686258711e-05,
+      "loss": 0.2773,
+      "step": 1341
+    },
+    {
+      "epoch": 1.5072300996771024,
+      "grad_norm": 0.5842833735748904,
+      "learning_rate": 1.7625034150868983e-05,
+      "loss": 0.3086,
+      "step": 1342
+    },
+    {
+      "epoch": 1.5083532219570406,
+      "grad_norm": 0.43427680840538785,
+      "learning_rate": 1.7619956747383435e-05,
+      "loss": 0.2765,
+      "step": 1343
+    },
+    {
+      "epoch": 1.5094763442369787,
+      "grad_norm": 0.4179475986040606,
+      "learning_rate": 1.7614874655254644e-05,
+      "loss": 0.2706,
+      "step": 1344
+    },
+    {
+      "epoch": 1.5105994665169171,
+      "grad_norm": 0.44696851307851265,
+      "learning_rate": 1.7609787877609678e-05,
+      "loss": 0.2932,
+      "step": 1345
+    },
+    {
+      "epoch": 1.5117225887968553,
+      "grad_norm": 0.43430140397443595,
+      "learning_rate": 1.760469641757849e-05,
+      "loss": 0.2748,
+      "step": 1346
+    },
+    {
+      "epoch": 1.5128457110767934,
+      "grad_norm": 0.4164245825608252,
+      "learning_rate": 1.7599600278293915e-05,
+      "loss": 0.2765,
+      "step": 1347
+    },
+    {
+      "epoch": 1.5139688333567318,
+      "grad_norm": 0.4411217535356976,
+      "learning_rate": 1.7594499462891654e-05,
+      "loss": 0.2671,
+      "step": 1348
+    },
+    {
+      "epoch": 1.51509195563667,
+      "grad_norm": 0.47168354723847505,
+      "learning_rate": 1.7589393974510304e-05,
+      "loss": 0.2663,
+      "step": 1349
+    },
+    {
+      "epoch": 1.5162150779166081,
+      "grad_norm": 0.45298191218773637,
+      "learning_rate": 1.758428381629132e-05,
+      "loss": 0.2977,
+      "step": 1350
+    },
+    {
+      "epoch": 1.5173382001965465,
+      "grad_norm": 0.48015367678036425,
+      "learning_rate": 1.7579168991379042e-05,
+      "loss": 0.3073,
+      "step": 1351
+    },
+    {
+      "epoch": 1.5184613224764845,
+      "grad_norm": 0.450921374315479,
+      "learning_rate": 1.757404950292068e-05,
+      "loss": 0.2981,
+      "step": 1352
+    },
+    {
+      "epoch": 1.5195844447564228,
+      "grad_norm": 0.4511950623917902,
+      "learning_rate": 1.7568925354066313e-05,
+      "loss": 0.2715,
+      "step": 1353
+    },
+    {
+      "epoch": 1.5207075670363612,
+      "grad_norm": 0.4645619454896311,
+      "learning_rate": 1.756379654796888e-05,
+      "loss": 0.2906,
+      "step": 1354
+    },
+    {
+      "epoch": 1.5218306893162992,
+      "grad_norm": 0.4597487903047916,
+      "learning_rate": 1.7558663087784195e-05,
+      "loss": 0.2923,
+      "step": 1355
+    },
+    {
+      "epoch": 1.5229538115962375,
+      "grad_norm": 0.46125985328422053,
+      "learning_rate": 1.7553524976670936e-05,
+      "loss": 0.3157,
+      "step": 1356
+    },
+    {
+      "epoch": 1.524076933876176,
+      "grad_norm": 0.4627880390771236,
+      "learning_rate": 1.7548382217790633e-05,
+      "loss": 0.2844,
+      "step": 1357
+    },
+    {
+      "epoch": 1.5252000561561139,
+      "grad_norm": 0.48404911382660587,
+      "learning_rate": 1.7543234814307685e-05,
+      "loss": 0.2848,
+      "step": 1358
+    },
+    {
+      "epoch": 1.5263231784360523,
+      "grad_norm": 0.4732923841586989,
+      "learning_rate": 1.753808276938935e-05,
+      "loss": 0.3009,
+      "step": 1359
+    },
+    {
+      "epoch": 1.5274463007159904,
+      "grad_norm": 0.44748155272300266,
+      "learning_rate": 1.753292608620573e-05,
+      "loss": 0.2806,
+      "step": 1360
+    },
+    {
+      "epoch": 1.5285694229959286,
+      "grad_norm": 0.4671484752305363,
+      "learning_rate": 1.7527764767929794e-05,
+      "loss": 0.2746,
+      "step": 1361
+    },
+    {
+      "epoch": 1.529692545275867,
+      "grad_norm": 0.45312930393716716,
+      "learning_rate": 1.7522598817737356e-05,
+      "loss": 0.2817,
+      "step": 1362
+    },
+    {
+      "epoch": 1.5308156675558051,
+      "grad_norm": 0.48439112821445,
+      "learning_rate": 1.7517428238807085e-05,
+      "loss": 0.2988,
+      "step": 1363
+    },
+    {
+      "epoch": 1.5319387898357433,
+      "grad_norm": 0.4258685474786322,
+      "learning_rate": 1.751225303432049e-05,
+      "loss": 0.2713,
+      "step": 1364
+    },
+    {
+      "epoch": 1.5330619121156817,
+      "grad_norm": 0.4404628717066417,
+      "learning_rate": 1.750707320746194e-05,
+      "loss": 0.2817,
+      "step": 1365
+    },
+    {
+      "epoch": 1.5341850343956198,
+      "grad_norm": 0.4477719006580123,
+      "learning_rate": 1.750188876141863e-05,
+      "loss": 0.2856,
+      "step": 1366
+    },
+    {
+      "epoch": 1.535308156675558,
+      "grad_norm": 0.4482867400976784,
+      "learning_rate": 1.7496699699380612e-05,
+      "loss": 0.2715,
+      "step": 1367
+    },
+    {
+      "epoch": 1.5364312789554964,
+      "grad_norm": 0.43210614417809273,
+      "learning_rate": 1.749150602454077e-05,
+      "loss": 0.2858,
+      "step": 1368
+    },
+    {
+      "epoch": 1.5375544012354345,
+      "grad_norm": 0.4851645221119642,
+      "learning_rate": 1.7486307740094832e-05,
+      "loss": 0.3126,
+      "step": 1369
+    },
+    {
+      "epoch": 1.5386775235153727,
+      "grad_norm": 0.43465394038491617,
+      "learning_rate": 1.7481104849241357e-05,
+      "loss": 0.2834,
+      "step": 1370
+    },
+    {
+      "epoch": 1.539800645795311,
+      "grad_norm": 0.4343631643379705,
+      "learning_rate": 1.7475897355181747e-05,
+      "loss": 0.29,
+      "step": 1371
+    },
+    {
+      "epoch": 1.5409237680752492,
+      "grad_norm": 0.43113460770410267,
+      "learning_rate": 1.747068526112022e-05,
+      "loss": 0.2712,
+      "step": 1372
+    },
+    {
+      "epoch": 1.5420468903551874,
+      "grad_norm": 0.47375551490134643,
+      "learning_rate": 1.7465468570263844e-05,
+      "loss": 0.2852,
+      "step": 1373
+    },
+    {
+      "epoch": 1.5431700126351258,
+      "grad_norm": 0.4567963890276265,
+      "learning_rate": 1.7460247285822504e-05,
+      "loss": 0.2882,
+      "step": 1374
+    },
+    {
+      "epoch": 1.544293134915064,
+      "grad_norm": 0.4689696386236284,
+      "learning_rate": 1.7455021411008906e-05,
+      "loss": 0.294,
+      "step": 1375
+    },
+    {
+      "epoch": 1.545416257195002,
+      "grad_norm": 0.4468934735798268,
+      "learning_rate": 1.7449790949038604e-05,
+      "loss": 0.2934,
+      "step": 1376
+    },
+    {
+      "epoch": 1.5465393794749405,
+      "grad_norm": 0.4590404219399577,
+      "learning_rate": 1.7444555903129943e-05,
+      "loss": 0.2996,
+      "step": 1377
+    },
+    {
+      "epoch": 1.5476625017548784,
+      "grad_norm": 0.4320176095607792,
+      "learning_rate": 1.7439316276504112e-05,
+      "loss": 0.2958,
+      "step": 1378
+    },
+    {
+      "epoch": 1.5487856240348168,
+      "grad_norm": 0.4724597686393324,
+      "learning_rate": 1.7434072072385115e-05,
+      "loss": 0.299,
+      "step": 1379
+    },
+    {
+      "epoch": 1.5499087463147552,
+      "grad_norm": 0.4551694704885351,
+      "learning_rate": 1.7428823293999757e-05,
+      "loss": 0.2898,
+      "step": 1380
+    },
+    {
+      "epoch": 1.551031868594693,
+      "grad_norm": 0.4655624233086197,
+      "learning_rate": 1.7423569944577677e-05,
+      "loss": 0.3064,
+      "step": 1381
+    },
+    {
+      "epoch": 1.5521549908746315,
+      "grad_norm": 0.4495621812588542,
+      "learning_rate": 1.7418312027351322e-05,
+      "loss": 0.277,
+      "step": 1382
+    },
+    {
+      "epoch": 1.5532781131545697,
+      "grad_norm": 0.46734262431819734,
+      "learning_rate": 1.741304954555594e-05,
+      "loss": 0.2723,
+      "step": 1383
+    },
+    {
+      "epoch": 1.5544012354345078,
+      "grad_norm": 0.4530212621436398,
+      "learning_rate": 1.7407782502429594e-05,
+      "loss": 0.2966,
+      "step": 1384
+    },
+    {
+      "epoch": 1.5555243577144462,
+      "grad_norm": 0.48906035382170715,
+      "learning_rate": 1.7402510901213158e-05,
+      "loss": 0.3166,
+      "step": 1385
+    },
+    {
+      "epoch": 1.5566474799943844,
+      "grad_norm": 0.4588087448330385,
+      "learning_rate": 1.73972347451503e-05,
+      "loss": 0.2911,
+      "step": 1386
+    },
+    {
+      "epoch": 1.5577706022743225,
+      "grad_norm": 0.45890894077015215,
+      "learning_rate": 1.7391954037487503e-05,
+      "loss": 0.2792,
+      "step": 1387
+    },
+    {
+      "epoch": 1.558893724554261,
+      "grad_norm": 0.43100684891614244,
+      "learning_rate": 1.738666878147404e-05,
+      "loss": 0.2705,
+      "step": 1388
+    },
+    {
+      "epoch": 1.560016846834199,
+      "grad_norm": 0.4533496037726458,
+      "learning_rate": 1.738137898036199e-05,
+      "loss": 0.2984,
+      "step": 1389
+    },
+    {
+      "epoch": 1.5611399691141372,
+      "grad_norm": 0.44398777502713194,
+      "learning_rate": 1.7376084637406222e-05,
+      "loss": 0.284,
+      "step": 1390
+    },
+    {
+      "epoch": 1.5622630913940756,
+      "grad_norm": 0.4398899733954843,
+      "learning_rate": 1.737078575586441e-05,
+      "loss": 0.2851,
+      "step": 1391
+    },
+    {
+      "epoch": 1.5633862136740138,
+      "grad_norm": 0.4743848381771007,
+      "learning_rate": 1.736548233899701e-05,
+      "loss": 0.2954,
+      "step": 1392
+    },
+    {
+      "epoch": 1.564509335953952,
+      "grad_norm": 0.4645458661781724,
+      "learning_rate": 1.7360174390067274e-05,
+      "loss": 0.2912,
+      "step": 1393
+    },
+    {
+      "epoch": 1.5656324582338903,
+      "grad_norm": 0.44409894402600525,
+      "learning_rate": 1.735486191234124e-05,
+      "loss": 0.2816,
+      "step": 1394
+    },
+    {
+      "epoch": 1.5667555805138285,
+      "grad_norm": 0.43988326731406074,
+      "learning_rate": 1.7349544909087737e-05,
+      "loss": 0.2777,
+      "step": 1395
+    },
+    {
+      "epoch": 1.5678787027937666,
+      "grad_norm": 0.461134202797876,
+      "learning_rate": 1.734422338357837e-05,
+      "loss": 0.2963,
+      "step": 1396
+    },
+    {
+      "epoch": 1.569001825073705,
+      "grad_norm": 0.43219608525830144,
+      "learning_rate": 1.7338897339087536e-05,
+      "loss": 0.2743,
+      "step": 1397
+    },
+    {
+      "epoch": 1.5701249473536432,
+      "grad_norm": 0.4522334985486791,
+      "learning_rate": 1.733356677889241e-05,
+      "loss": 0.293,
+      "step": 1398
+    },
+    {
+      "epoch": 1.5712480696335813,
+      "grad_norm": 0.4697839824886229,
+      "learning_rate": 1.732823170627294e-05,
+      "loss": 0.3225,
+      "step": 1399
+    },
+    {
+      "epoch": 1.5723711919135197,
+      "grad_norm": 0.4379752450384441,
+      "learning_rate": 1.7322892124511862e-05,
+      "loss": 0.2934,
+      "step": 1400
+    },
+    {
+      "epoch": 1.5734943141934579,
+      "grad_norm": 0.4403768513512486,
+      "learning_rate": 1.731754803689467e-05,
+      "loss": 0.2808,
+      "step": 1401
+    },
+    {
+      "epoch": 1.574617436473396,
+      "grad_norm": 0.4526960513869645,
+      "learning_rate": 1.731219944670965e-05,
+      "loss": 0.3075,
+      "step": 1402
+    },
+    {
+      "epoch": 1.5757405587533344,
+      "grad_norm": 0.4391813185175165,
+      "learning_rate": 1.730684635724784e-05,
+      "loss": 0.2818,
+      "step": 1403
+    },
+    {
+      "epoch": 1.5768636810332723,
+      "grad_norm": 0.45969828645408684,
+      "learning_rate": 1.7301488771803056e-05,
+      "loss": 0.3036,
+      "step": 1404
+    },
+    {
+      "epoch": 1.5779868033132107,
+      "grad_norm": 0.43457708703765435,
+      "learning_rate": 1.7296126693671886e-05,
+      "loss": 0.2746,
+      "step": 1405
+    },
+    {
+      "epoch": 1.579109925593149,
+      "grad_norm": 0.4331075300392934,
+      "learning_rate": 1.7290760126153666e-05,
+      "loss": 0.2821,
+      "step": 1406
+    },
+    {
+      "epoch": 1.580233047873087,
+      "grad_norm": 0.4208492939766656,
+      "learning_rate": 1.7285389072550515e-05,
+      "loss": 0.2795,
+      "step": 1407
+    },
+    {
+      "epoch": 1.5813561701530254,
+      "grad_norm": 0.4634956044262763,
+      "learning_rate": 1.728001353616729e-05,
+      "loss": 0.3025,
+      "step": 1408
+    },
+    {
+      "epoch": 1.5824792924329636,
+      "grad_norm": 0.46128626849790916,
+      "learning_rate": 1.727463352031163e-05,
+      "loss": 0.302,
+      "step": 1409
+    },
+    {
+      "epoch": 1.5836024147129018,
+      "grad_norm": 0.4275363771455048,
+      "learning_rate": 1.7269249028293907e-05,
+      "loss": 0.2809,
+      "step": 1410
+    },
+    {
+      "epoch": 1.5847255369928401,
+      "grad_norm": 0.4508186393663234,
+      "learning_rate": 1.7263860063427263e-05,
+      "loss": 0.2928,
+      "step": 1411
+    },
+    {
+      "epoch": 1.5858486592727783,
+      "grad_norm": 0.4641518740610698,
+      "learning_rate": 1.7258466629027586e-05,
+      "loss": 0.2966,
+      "step": 1412
+    },
+    {
+      "epoch": 1.5869717815527165,
+      "grad_norm": 0.4350653148056457,
+      "learning_rate": 1.7253068728413517e-05,
+      "loss": 0.2871,
+      "step": 1413
+    },
+    {
+      "epoch": 1.5880949038326548,
+      "grad_norm": 0.4574921332811425,
+      "learning_rate": 1.7247666364906443e-05,
+      "loss": 0.305,
+      "step": 1414
+    },
+    {
+      "epoch": 1.589218026112593,
+      "grad_norm": 0.4423661730830507,
+      "learning_rate": 1.7242259541830497e-05,
+      "loss": 0.2835,
+      "step": 1415
+    },
+    {
+      "epoch": 1.5903411483925312,
+      "grad_norm": 0.4415125572569024,
+      "learning_rate": 1.723684826251256e-05,
+      "loss": 0.287,
+      "step": 1416
+    },
+    {
+      "epoch": 1.5914642706724695,
+      "grad_norm": 0.45878372415584956,
+      "learning_rate": 1.7231432530282246e-05,
+      "loss": 0.291,
+      "step": 1417
+    },
+    {
+      "epoch": 1.5925873929524077,
+      "grad_norm": 0.4536230530206705,
+      "learning_rate": 1.722601234847192e-05,
+      "loss": 0.2676,
+      "step": 1418
+    },
+    {
+      "epoch": 1.5937105152323459,
+      "grad_norm": 0.42510000179340557,
+      "learning_rate": 1.7220587720416677e-05,
+      "loss": 0.2625,
+      "step": 1419
+    },
+    {
+      "epoch": 1.5948336375122842,
+      "grad_norm": 0.4617059875851874,
+      "learning_rate": 1.721515864945435e-05,
+      "loss": 0.3081,
+      "step": 1420
+    },
+    {
+      "epoch": 1.5959567597922224,
+      "grad_norm": 0.4325597054894661,
+      "learning_rate": 1.7209725138925506e-05,
+      "loss": 0.2713,
+      "step": 1421
+    },
+    {
+      "epoch": 1.5970798820721606,
+      "grad_norm": 0.43609135158582557,
+      "learning_rate": 1.7204287192173444e-05,
+      "loss": 0.2722,
+      "step": 1422
+    },
+    {
+      "epoch": 1.598203004352099,
+      "grad_norm": 0.43997208529177845,
+      "learning_rate": 1.7198844812544194e-05,
+      "loss": 0.2737,
+      "step": 1423
+    },
+    {
+      "epoch": 1.599326126632037,
+      "grad_norm": 0.45811922865343463,
+      "learning_rate": 1.7193398003386514e-05,
+      "loss": 0.3069,
+      "step": 1424
+    },
+    {
+      "epoch": 1.6004492489119753,
+      "grad_norm": 0.441765765575073,
+      "learning_rate": 1.718794676805188e-05,
+      "loss": 0.2878,
+      "step": 1425
+    },
+    {
+      "epoch": 1.6015723711919136,
+      "grad_norm": 0.44938030928999945,
+      "learning_rate": 1.71824911098945e-05,
+      "loss": 0.2803,
+      "step": 1426
+    },
+    {
+      "epoch": 1.6026954934718516,
+      "grad_norm": 0.4711803532185502,
+      "learning_rate": 1.7177031032271298e-05,
+      "loss": 0.2898,
+      "step": 1427
+    },
+    {
+      "epoch": 1.60381861575179,
+      "grad_norm": 0.48742752034450476,
+      "learning_rate": 1.7171566538541925e-05,
+      "loss": 0.2931,
+      "step": 1428
+    },
+    {
+      "epoch": 1.6049417380317283,
+      "grad_norm": 0.455468095932038,
+      "learning_rate": 1.7166097632068745e-05,
+      "loss": 0.3055,
+      "step": 1429
+    },
+    {
+      "epoch": 1.6060648603116663,
+      "grad_norm": 0.4372160767962161,
+      "learning_rate": 1.7160624316216825e-05,
+      "loss": 0.2829,
+      "step": 1430
+    },
+    {
+      "epoch": 1.6071879825916047,
+      "grad_norm": 0.47650809154388873,
+      "learning_rate": 1.715514659435397e-05,
+      "loss": 0.283,
+      "step": 1431
+    },
+    {
+      "epoch": 1.608311104871543,
+      "grad_norm": 0.430482793568945,
+      "learning_rate": 1.7149664469850674e-05,
+      "loss": 0.277,
+      "step": 1432
+    },
+    {
+      "epoch": 1.609434227151481,
+      "grad_norm": 0.4516995855142114,
+      "learning_rate": 1.714417794608015e-05,
+      "loss": 0.2934,
+      "step": 1433
+    },
+    {
+      "epoch": 1.6105573494314194,
+      "grad_norm": 0.4110806797442354,
+      "learning_rate": 1.713868702641832e-05,
+      "loss": 0.2593,
+      "step": 1434
+    },
+    {
+      "epoch": 1.6116804717113575,
+      "grad_norm": 0.4616915231229448,
+      "learning_rate": 1.7133191714243805e-05,
+      "loss": 0.284,
+      "step": 1435
+    },
+    {
+      "epoch": 1.6128035939912957,
+      "grad_norm": 0.41123495094639895,
+      "learning_rate": 1.712769201293793e-05,
+      "loss": 0.258,
+      "step": 1436
+    },
+    {
+      "epoch": 1.613926716271234,
+      "grad_norm": 0.4770233615190129,
+      "learning_rate": 1.7122187925884723e-05,
+      "loss": 0.3075,
+      "step": 1437
+    },
+    {
+      "epoch": 1.6150498385511722,
+      "grad_norm": 0.45932911711426877,
+      "learning_rate": 1.7116679456470908e-05,
+      "loss": 0.3073,
+      "step": 1438
+    },
+    {
+      "epoch": 1.6161729608311104,
+      "grad_norm": 0.43310100468943585,
+      "learning_rate": 1.711116660808591e-05,
+      "loss": 0.2803,
+      "step": 1439
+    },
+    {
+      "epoch": 1.6172960831110488,
+      "grad_norm": 0.4053611622253437,
+      "learning_rate": 1.710564938412184e-05,
+      "loss": 0.2552,
+      "step": 1440
+    },
+    {
+      "epoch": 1.618419205390987,
+      "grad_norm": 0.4776776817414243,
+      "learning_rate": 1.710012778797351e-05,
+      "loss": 0.3047,
+      "step": 1441
+    },
+    {
+      "epoch": 1.619542327670925,
+      "grad_norm": 0.46553831799548406,
+      "learning_rate": 1.7094601823038425e-05,
+      "loss": 0.2946,
+      "step": 1442
+    },
+    {
+      "epoch": 1.6206654499508635,
+      "grad_norm": 0.466347563259381,
+      "learning_rate": 1.7089071492716758e-05,
+      "loss": 0.2886,
+      "step": 1443
+    },
+    {
+      "epoch": 1.6217885722308016,
+      "grad_norm": 0.44584944747972316,
+      "learning_rate": 1.7083536800411392e-05,
+      "loss": 0.2731,
+      "step": 1444
+    },
+    {
+      "epoch": 1.6229116945107398,
+      "grad_norm": 0.49761550260791193,
+      "learning_rate": 1.7077997749527884e-05,
+      "loss": 0.3217,
+      "step": 1445
+    },
+    {
+      "epoch": 1.6240348167906782,
+      "grad_norm": 0.42865534739277844,
+      "learning_rate": 1.707245434347447e-05,
+      "loss": 0.2829,
+      "step": 1446
+    },
+    {
+      "epoch": 1.6251579390706163,
+      "grad_norm": 0.4159767822652411,
+      "learning_rate": 1.706690658566207e-05,
+      "loss": 0.2645,
+      "step": 1447
+    },
+    {
+      "epoch": 1.6262810613505545,
+      "grad_norm": 0.4766099212680193,
+      "learning_rate": 1.7061354479504277e-05,
+      "loss": 0.2893,
+      "step": 1448
+    },
+    {
+      "epoch": 1.6274041836304929,
+      "grad_norm": 0.4532911553458812,
+      "learning_rate": 1.705579802841737e-05,
+      "loss": 0.2789,
+      "step": 1449
+    },
+    {
+      "epoch": 1.628527305910431,
+      "grad_norm": 0.45741701789824535,
+      "learning_rate": 1.7050237235820287e-05,
+      "loss": 0.289,
+      "step": 1450
+    },
+    {
+      "epoch": 1.6296504281903692,
+      "grad_norm": 0.4846189139759324,
+      "learning_rate": 1.704467210513465e-05,
+      "loss": 0.3282,
+      "step": 1451
+    },
+    {
+      "epoch": 1.6307735504703076,
+      "grad_norm": 0.4261188062643428,
+      "learning_rate": 1.7039102639784747e-05,
+      "loss": 0.2712,
+      "step": 1452
+    },
+    {
+      "epoch": 1.6318966727502455,
+      "grad_norm": 0.4409212737827275,
+      "learning_rate": 1.7033528843197523e-05,
+      "loss": 0.2847,
+      "step": 1453
+    },
+    {
+      "epoch": 1.633019795030184,
+      "grad_norm": 0.4458915183695067,
+      "learning_rate": 1.7027950718802605e-05,
+      "loss": 0.2778,
+      "step": 1454
+    },
+    {
+      "epoch": 1.6341429173101223,
+      "grad_norm": 0.47098011281228286,
+      "learning_rate": 1.7022368270032268e-05,
+      "loss": 0.2913,
+      "step": 1455
+    },
+    {
+      "epoch": 1.6352660395900602,
+      "grad_norm": 0.4436299643362799,
+      "learning_rate": 1.7016781500321458e-05,
+      "loss": 0.2686,
+      "step": 1456
+    },
+    {
+      "epoch": 1.6363891618699986,
+      "grad_norm": 0.508079183867471,
+      "learning_rate": 1.7011190413107774e-05,
+      "loss": 0.3568,
+      "step": 1457
+    },
+    {
+      "epoch": 1.6375122841499368,
+      "grad_norm": 0.4447742608378096,
+      "learning_rate": 1.7005595011831473e-05,
+      "loss": 0.3054,
+      "step": 1458
+    },
+    {
+      "epoch": 1.638635406429875,
+      "grad_norm": 0.45902598871228406,
+      "learning_rate": 1.699999529993547e-05,
+      "loss": 0.2874,
+      "step": 1459
+    },
+    {
+      "epoch": 1.6397585287098133,
+      "grad_norm": 0.43790349463956285,
+      "learning_rate": 1.6994391280865327e-05,
+      "loss": 0.2753,
+      "step": 1460
+    },
+    {
+      "epoch": 1.6408816509897515,
+      "grad_norm": 0.43987143870781154,
+      "learning_rate": 1.698878295806926e-05,
+      "loss": 0.2812,
+      "step": 1461
+    },
+    {
+      "epoch": 1.6420047732696896,
+      "grad_norm": 0.45636973748880527,
+      "learning_rate": 1.698317033499813e-05,
+      "loss": 0.294,
+      "step": 1462
+    },
+    {
+      "epoch": 1.643127895549628,
+      "grad_norm": 0.4473056852255817,
+      "learning_rate": 1.6977553415105446e-05,
+      "loss": 0.2922,
+      "step": 1463
+    },
+    {
+      "epoch": 1.6442510178295662,
+      "grad_norm": 0.44711605854202396,
+      "learning_rate": 1.6971932201847362e-05,
+      "loss": 0.3037,
+      "step": 1464
+    },
+    {
+      "epoch": 1.6453741401095043,
+      "grad_norm": 0.47322347350994115,
+      "learning_rate": 1.6966306698682672e-05,
+      "loss": 0.3091,
+      "step": 1465
+    },
+    {
+      "epoch": 1.6464972623894427,
+      "grad_norm": 0.47589867797246205,
+      "learning_rate": 1.6960676909072808e-05,
+      "loss": 0.3044,
+      "step": 1466
+    },
+    {
+      "epoch": 1.6476203846693809,
+      "grad_norm": 0.45462083845828744,
+      "learning_rate": 1.6955042836481842e-05,
+      "loss": 0.2901,
+      "step": 1467
+    },
+    {
+      "epoch": 1.648743506949319,
+      "grad_norm": 0.43342182003648877,
+      "learning_rate": 1.6949404484376484e-05,
+      "loss": 0.2826,
+      "step": 1468
+    },
+    {
+      "epoch": 1.6498666292292574,
+      "grad_norm": 0.4420101471539028,
+      "learning_rate": 1.6943761856226072e-05,
+      "loss": 0.2823,
+      "step": 1469
+    },
+    {
+      "epoch": 1.6509897515091956,
+      "grad_norm": 0.45834897544206743,
+      "learning_rate": 1.693811495550258e-05,
+      "loss": 0.2998,
+      "step": 1470
+    },
+    {
+      "epoch": 1.6521128737891337,
+      "grad_norm": 0.4376763092414877,
+      "learning_rate": 1.69324637856806e-05,
+      "loss": 0.2906,
+      "step": 1471
+    },
+    {
+      "epoch": 1.6532359960690721,
+      "grad_norm": 0.4546259417086349,
+      "learning_rate": 1.6926808350237367e-05,
+      "loss": 0.293,
+      "step": 1472
+    },
+    {
+      "epoch": 1.6543591183490103,
+      "grad_norm": 0.4372010366201924,
+      "learning_rate": 1.692114865265273e-05,
+      "loss": 0.2712,
+      "step": 1473
+    },
+    {
+      "epoch": 1.6554822406289484,
+      "grad_norm": 0.4778527561849003,
+      "learning_rate": 1.691548469640916e-05,
+      "loss": 0.3141,
+      "step": 1474
+    },
+    {
+      "epoch": 1.6566053629088868,
+      "grad_norm": 0.4373422181074543,
+      "learning_rate": 1.690981648499176e-05,
+      "loss": 0.2776,
+      "step": 1475
+    },
+    {
+      "epoch": 1.657728485188825,
+      "grad_norm": 0.45910849316443425,
+      "learning_rate": 1.6904144021888236e-05,
+      "loss": 0.2914,
+      "step": 1476
+    },
+    {
+      "epoch": 1.6588516074687631,
+      "grad_norm": 0.4377368116005779,
+      "learning_rate": 1.6898467310588917e-05,
+      "loss": 0.2832,
+      "step": 1477
+    },
+    {
+      "epoch": 1.6599747297487015,
+      "grad_norm": 0.4343991004179873,
+      "learning_rate": 1.689278635458675e-05,
+      "loss": 0.2697,
+      "step": 1478
+    },
+    {
+      "epoch": 1.6610978520286395,
+      "grad_norm": 0.4584326718106985,
+      "learning_rate": 1.6887101157377284e-05,
+      "loss": 0.3061,
+      "step": 1479
+    },
+    {
+      "epoch": 1.6622209743085778,
+      "grad_norm": 0.48367472603690564,
+      "learning_rate": 1.6881411722458688e-05,
+      "loss": 0.2898,
+      "step": 1480
+    },
+    {
+      "epoch": 1.6633440965885162,
+      "grad_norm": 0.4259604088571098,
+      "learning_rate": 1.6875718053331736e-05,
+      "loss": 0.2642,
+      "step": 1481
+    },
+    {
+      "epoch": 1.6644672188684542,
+      "grad_norm": 0.4427453550152015,
+      "learning_rate": 1.6870020153499796e-05,
+      "loss": 0.2742,
+      "step": 1482
+    },
+    {
+      "epoch": 1.6655903411483926,
+      "grad_norm": 0.4640041414969516,
+      "learning_rate": 1.686431802646886e-05,
+      "loss": 0.2913,
+      "step": 1483
+    },
+    {
+      "epoch": 1.6667134634283307,
+      "grad_norm": 0.46659553606141885,
+      "learning_rate": 1.68586116757475e-05,
+      "loss": 0.2979,
+      "step": 1484
+    },
+    {
+      "epoch": 1.6678365857082689,
+      "grad_norm": 0.44118444234300963,
+      "learning_rate": 1.6852901104846902e-05,
+      "loss": 0.2798,
+      "step": 1485
+    },
+    {
+      "epoch": 1.6689597079882073,
+      "grad_norm": 0.4232934978922537,
+      "learning_rate": 1.6847186317280844e-05,
+      "loss": 0.2664,
+      "step": 1486
+    },
+    {
+      "epoch": 1.6700828302681454,
+      "grad_norm": 0.4549678955070991,
+      "learning_rate": 1.684146731656569e-05,
+      "loss": 0.2691,
+      "step": 1487
+    },
+    {
+      "epoch": 1.6712059525480836,
+      "grad_norm": 0.4462793748347985,
+      "learning_rate": 1.683574410622041e-05,
+      "loss": 0.2756,
+      "step": 1488
+    },
+    {
+      "epoch": 1.672329074828022,
+      "grad_norm": 0.46943909006116125,
+      "learning_rate": 1.683001668976656e-05,
+      "loss": 0.3007,
+      "step": 1489
+    },
+    {
+      "epoch": 1.6734521971079601,
+      "grad_norm": 0.47419859082322885,
+      "learning_rate": 1.6824285070728278e-05,
+      "loss": 0.2873,
+      "step": 1490
+    },
+    {
+      "epoch": 1.6745753193878983,
+      "grad_norm": 0.47147067796892533,
+      "learning_rate": 1.6818549252632295e-05,
+      "loss": 0.2839,
+      "step": 1491
+    },
+    {
+      "epoch": 1.6756984416678367,
+      "grad_norm": 0.456348027524806,
+      "learning_rate": 1.6812809239007924e-05,
+      "loss": 0.3,
+      "step": 1492
+    },
+    {
+      "epoch": 1.6768215639477748,
+      "grad_norm": 0.40777119029954567,
+      "learning_rate": 1.6807065033387052e-05,
+      "loss": 0.2596,
+      "step": 1493
+    },
+    {
+      "epoch": 1.677944686227713,
+      "grad_norm": 0.40000140473687745,
+      "learning_rate": 1.6801316639304163e-05,
+      "loss": 0.2658,
+      "step": 1494
+    },
+    {
+      "epoch": 1.6790678085076514,
+      "grad_norm": 0.45210384075210974,
+      "learning_rate": 1.6795564060296295e-05,
+      "loss": 0.2966,
+      "step": 1495
+    },
+    {
+      "epoch": 1.6801909307875895,
+      "grad_norm": 0.40822664899597394,
+      "learning_rate": 1.678980729990308e-05,
+      "loss": 0.2535,
+      "step": 1496
+    },
+    {
+      "epoch": 1.6813140530675277,
+      "grad_norm": 0.4783878749550028,
+      "learning_rate": 1.6784046361666714e-05,
+      "loss": 0.3179,
+      "step": 1497
+    },
+    {
+      "epoch": 1.682437175347466,
+      "grad_norm": 0.4205086583335779,
+      "learning_rate": 1.6778281249131973e-05,
+      "loss": 0.2624,
+      "step": 1498
+    },
+    {
+      "epoch": 1.6835602976274042,
+      "grad_norm": 0.4448670745068746,
+      "learning_rate": 1.677251196584618e-05,
+      "loss": 0.2726,
+      "step": 1499
+    },
+    {
+      "epoch": 1.6846834199073424,
+      "grad_norm": 0.4612087079191544,
+      "learning_rate": 1.676673851535925e-05,
+      "loss": 0.2863,
+      "step": 1500
+    },
+    {
+      "epoch": 1.6858065421872808,
+      "grad_norm": 0.44897350104308953,
+      "learning_rate": 1.6760960901223647e-05,
+      "loss": 0.2777,
+      "step": 1501
+    },
+    {
+      "epoch": 1.6869296644672187,
+      "grad_norm": 0.4650716938100982,
+      "learning_rate": 1.6755179126994397e-05,
+      "loss": 0.3114,
+      "step": 1502
+    },
+    {
+      "epoch": 1.688052786747157,
+      "grad_norm": 0.4408040310278629,
+      "learning_rate": 1.6749393196229097e-05,
+      "loss": 0.2802,
+      "step": 1503
+    },
+    {
+      "epoch": 1.6891759090270955,
+      "grad_norm": 0.4323634729178194,
+      "learning_rate": 1.6743603112487888e-05,
+      "loss": 0.272,
+      "step": 1504
+    },
+    {
+      "epoch": 1.6902990313070334,
+      "grad_norm": 0.4532517237446688,
+      "learning_rate": 1.6737808879333477e-05,
+      "loss": 0.2994,
+      "step": 1505
+    },
+    {
+      "epoch": 1.6914221535869718,
+      "grad_norm": 0.41707219861308187,
+      "learning_rate": 1.6732010500331112e-05,
+      "loss": 0.2786,
+      "step": 1506
+    },
+    {
+      "epoch": 1.6925452758669102,
+      "grad_norm": 0.4790732992448698,
+      "learning_rate": 1.6726207979048604e-05,
+      "loss": 0.3217,
+      "step": 1507
+    },
+    {
+      "epoch": 1.693668398146848,
+      "grad_norm": 0.4426207341478685,
+      "learning_rate": 1.672040131905631e-05,
+      "loss": 0.2957,
+      "step": 1508
+    },
+    {
+      "epoch": 1.6947915204267865,
+      "grad_norm": 0.4491550141407061,
+      "learning_rate": 1.6714590523927127e-05,
+      "loss": 0.2963,
+      "step": 1509
+    },
+    {
+      "epoch": 1.6959146427067247,
+      "grad_norm": 0.44425919435030004,
+      "learning_rate": 1.6708775597236507e-05,
+      "loss": 0.2888,
+      "step": 1510
+    },
+    {
+      "epoch": 1.6970377649866628,
+      "grad_norm": 0.4428624792544438,
+      "learning_rate": 1.6702956542562433e-05,
+      "loss": 0.2919,
+      "step": 1511
+    },
+    {
+      "epoch": 1.6981608872666012,
+      "grad_norm": 0.4645304954497035,
+      "learning_rate": 1.669713336348544e-05,
+      "loss": 0.3044,
+      "step": 1512
+    },
+    {
+      "epoch": 1.6992840095465394,
+      "grad_norm": 0.442526153745366,
+      "learning_rate": 1.6691306063588583e-05,
+      "loss": 0.2733,
+      "step": 1513
+    },
+    {
+      "epoch": 1.7004071318264775,
+      "grad_norm": 0.44682174416662607,
+      "learning_rate": 1.6685474646457477e-05,
+      "loss": 0.288,
+      "step": 1514
+    },
+    {
+      "epoch": 1.701530254106416,
+      "grad_norm": 0.4460477682537591,
+      "learning_rate": 1.6679639115680247e-05,
+      "loss": 0.2892,
+      "step": 1515
+    },
+    {
+      "epoch": 1.702653376386354,
+      "grad_norm": 0.42381417287482237,
+      "learning_rate": 1.667379947484756e-05,
+      "loss": 0.2753,
+      "step": 1516
+    },
+    {
+      "epoch": 1.7037764986662922,
+      "grad_norm": 0.45442357297793295,
+      "learning_rate": 1.666795572755262e-05,
+      "loss": 0.2963,
+      "step": 1517
+    },
+    {
+      "epoch": 1.7048996209462306,
+      "grad_norm": 0.4467561695261649,
+      "learning_rate": 1.666210787739114e-05,
+      "loss": 0.294,
+      "step": 1518
+    },
+    {
+      "epoch": 1.7060227432261688,
+      "grad_norm": 0.4514819485972295,
+      "learning_rate": 1.665625592796137e-05,
+      "loss": 0.2916,
+      "step": 1519
+    },
+    {
+      "epoch": 1.707145865506107,
+      "grad_norm": 0.47046147688947726,
+      "learning_rate": 1.665039988286408e-05,
+      "loss": 0.3059,
+      "step": 1520
+    },
+    {
+      "epoch": 1.7082689877860453,
+      "grad_norm": 0.4664118838351173,
+      "learning_rate": 1.6644539745702558e-05,
+      "loss": 0.2977,
+      "step": 1521
+    },
+    {
+      "epoch": 1.7093921100659835,
+      "grad_norm": 0.4388011550954156,
+      "learning_rate": 1.6638675520082613e-05,
+      "loss": 0.2998,
+      "step": 1522
+    },
+    {
+      "epoch": 1.7105152323459216,
+      "grad_norm": 0.45082022297345586,
+      "learning_rate": 1.663280720961256e-05,
+      "loss": 0.2966,
+      "step": 1523
+    },
+    {
+      "epoch": 1.71163835462586,
+      "grad_norm": 0.4497822398613249,
+      "learning_rate": 1.662693481790324e-05,
+      "loss": 0.2827,
+      "step": 1524
+    },
+    {
+      "epoch": 1.7127614769057982,
+      "grad_norm": 0.46875113264717555,
+      "learning_rate": 1.6621058348568008e-05,
+      "loss": 0.3053,
+      "step": 1525
+    },
+    {
+      "epoch": 1.7138845991857363,
+      "grad_norm": 0.4547313413961331,
+      "learning_rate": 1.6615177805222703e-05,
+      "loss": 0.2875,
+      "step": 1526
+    },
+    {
+      "epoch": 1.7150077214656747,
+      "grad_norm": 0.44523115177994116,
+      "learning_rate": 1.6609293191485704e-05,
+      "loss": 0.2924,
+      "step": 1527
+    },
+    {
+      "epoch": 1.7161308437456126,
+      "grad_norm": 0.4348408552400898,
+      "learning_rate": 1.660340451097787e-05,
+      "loss": 0.2842,
+      "step": 1528
+    },
+    {
+      "epoch": 1.717253966025551,
+      "grad_norm": 0.4274690190508088,
+      "learning_rate": 1.6597511767322575e-05,
+      "loss": 0.2754,
+      "step": 1529
+    },
+    {
+      "epoch": 1.7183770883054894,
+      "grad_norm": 0.45654747366119725,
+      "learning_rate": 1.6591614964145685e-05,
+      "loss": 0.3052,
+      "step": 1530
+    },
+    {
+      "epoch": 1.7195002105854273,
+      "grad_norm": 0.4633059182553845,
+      "learning_rate": 1.6585714105075573e-05,
+      "loss": 0.3026,
+      "step": 1531
+    },
+    {
+      "epoch": 1.7206233328653657,
+      "grad_norm": 0.45324442029714246,
+      "learning_rate": 1.6579809193743094e-05,
+      "loss": 0.2861,
+      "step": 1532
+    },
+    {
+      "epoch": 1.7217464551453041,
+      "grad_norm": 0.4599466934645262,
+      "learning_rate": 1.6573900233781616e-05,
+      "loss": 0.2858,
+      "step": 1533
+    },
+    {
+      "epoch": 1.722869577425242,
+      "grad_norm": 0.45346376009896494,
+      "learning_rate": 1.656798722882698e-05,
+      "loss": 0.2902,
+      "step": 1534
+    },
+    {
+      "epoch": 1.7239926997051804,
+      "grad_norm": 0.4406609173508741,
+      "learning_rate": 1.6562070182517524e-05,
+      "loss": 0.2978,
+      "step": 1535
+    },
+    {
+      "epoch": 1.7251158219851186,
+      "grad_norm": 0.46672133569112967,
+      "learning_rate": 1.6556149098494075e-05,
+      "loss": 0.2761,
+      "step": 1536
+    },
+    {
+      "epoch": 1.7262389442650568,
+      "grad_norm": 0.48979063606476814,
+      "learning_rate": 1.6550223980399934e-05,
+      "loss": 0.2984,
+      "step": 1537
+    },
+    {
+      "epoch": 1.7273620665449951,
+      "grad_norm": 0.4511837735588295,
+      "learning_rate": 1.65442948318809e-05,
+      "loss": 0.285,
+      "step": 1538
+    },
+    {
+      "epoch": 1.7284851888249333,
+      "grad_norm": 0.45689758535181885,
+      "learning_rate": 1.6538361656585237e-05,
+      "loss": 0.3051,
+      "step": 1539
+    },
+    {
+      "epoch": 1.7296083111048715,
+      "grad_norm": 0.42911681373204696,
+      "learning_rate": 1.6532424458163692e-05,
+      "loss": 0.2855,
+      "step": 1540
+    },
+    {
+      "epoch": 1.7307314333848098,
+      "grad_norm": 0.4599079422199036,
+      "learning_rate": 1.6526483240269497e-05,
+      "loss": 0.2929,
+      "step": 1541
+    },
+    {
+      "epoch": 1.731854555664748,
+      "grad_norm": 0.4304194735001978,
+      "learning_rate": 1.6520538006558345e-05,
+      "loss": 0.2708,
+      "step": 1542
+    },
+    {
+      "epoch": 1.7329776779446862,
+      "grad_norm": 0.44178440467725394,
+      "learning_rate": 1.6514588760688397e-05,
+      "loss": 0.3007,
+      "step": 1543
+    },
+    {
+      "epoch": 1.7341008002246245,
+      "grad_norm": 0.4423330619472906,
+      "learning_rate": 1.65086355063203e-05,
+      "loss": 0.2845,
+      "step": 1544
+    },
+    {
+      "epoch": 1.7352239225045627,
+      "grad_norm": 0.433601638401388,
+      "learning_rate": 1.6502678247117146e-05,
+      "loss": 0.2796,
+      "step": 1545
+    },
+    {
+      "epoch": 1.7363470447845009,
+      "grad_norm": 0.44925607846317117,
+      "learning_rate": 1.649671698674451e-05,
+      "loss": 0.2786,
+      "step": 1546
+    },
+    {
+      "epoch": 1.7374701670644392,
+      "grad_norm": 0.4146082239786263,
+      "learning_rate": 1.6490751728870422e-05,
+      "loss": 0.2596,
+      "step": 1547
+    },
+    {
+      "epoch": 1.7385932893443774,
+      "grad_norm": 0.4338894018059736,
+      "learning_rate": 1.6484782477165365e-05,
+      "loss": 0.276,
+      "step": 1548
+    },
+    {
+      "epoch": 1.7397164116243156,
+      "grad_norm": 0.4650911120148975,
+      "learning_rate": 1.6478809235302287e-05,
+      "loss": 0.2873,
+      "step": 1549
+    },
+    {
+      "epoch": 1.740839533904254,
+      "grad_norm": 0.42522285905199153,
+      "learning_rate": 1.647283200695659e-05,
+      "loss": 0.2689,
+      "step": 1550
+    },
+    {
+      "epoch": 1.741962656184192,
+      "grad_norm": 0.44208521414529733,
+      "learning_rate": 1.6466850795806136e-05,
+      "loss": 0.3013,
+      "step": 1551
+    },
+    {
+      "epoch": 1.7430857784641303,
+      "grad_norm": 0.4517383587426911,
+      "learning_rate": 1.6460865605531214e-05,
+      "loss": 0.2905,
+      "step": 1552
+    },
+    {
+      "epoch": 1.7442089007440686,
+      "grad_norm": 0.42667270058823686,
+      "learning_rate": 1.6454876439814592e-05,
+      "loss": 0.2778,
+      "step": 1553
+    },
+    {
+      "epoch": 1.7453320230240066,
+      "grad_norm": 0.4596306665470625,
+      "learning_rate": 1.644888330234146e-05,
+      "loss": 0.2989,
+      "step": 1554
+    },
+    {
+      "epoch": 1.746455145303945,
+      "grad_norm": 0.43733172758764655,
+      "learning_rate": 1.6442886196799465e-05,
+      "loss": 0.2934,
+      "step": 1555
+    },
+    {
+      "epoch": 1.7475782675838833,
+      "grad_norm": 0.44887986338826835,
+      "learning_rate": 1.6436885126878696e-05,
+      "loss": 0.3157,
+      "step": 1556
+    },
+    {
+      "epoch": 1.7487013898638213,
+      "grad_norm": 0.451897402012675,
+      "learning_rate": 1.6430880096271672e-05,
+      "loss": 0.2903,
+      "step": 1557
+    },
+    {
+      "epoch": 1.7498245121437597,
+      "grad_norm": 0.41101933804424956,
+      "learning_rate": 1.6424871108673355e-05,
+      "loss": 0.2747,
+      "step": 1558
+    },
+    {
+      "epoch": 1.7509476344236978,
+      "grad_norm": 0.49671129631427824,
+      "learning_rate": 1.6418858167781145e-05,
+      "loss": 0.3052,
+      "step": 1559
+    },
+    {
+      "epoch": 1.752070756703636,
+      "grad_norm": 0.4181368637873254,
+      "learning_rate": 1.6412841277294865e-05,
+      "loss": 0.2834,
+      "step": 1560
+    },
+    {
+      "epoch": 1.7531938789835744,
+      "grad_norm": 0.4215454138779954,
+      "learning_rate": 1.6406820440916778e-05,
+      "loss": 0.267,
+      "step": 1561
+    },
+    {
+      "epoch": 1.7543170012635125,
+      "grad_norm": 0.4712786813972758,
+      "learning_rate": 1.6400795662351572e-05,
+      "loss": 0.3015,
+      "step": 1562
+    },
+    {
+      "epoch": 1.7554401235434507,
+      "grad_norm": 0.456622799753585,
+      "learning_rate": 1.639476694530635e-05,
+      "loss": 0.2769,
+      "step": 1563
+    },
+    {
+      "epoch": 1.756563245823389,
+      "grad_norm": 0.4518709261822103,
+      "learning_rate": 1.6388734293490666e-05,
+      "loss": 0.2828,
+      "step": 1564
+    },
+    {
+      "epoch": 1.7576863681033272,
+      "grad_norm": 0.4190624104153518,
+      "learning_rate": 1.6382697710616458e-05,
+      "loss": 0.2702,
+      "step": 1565
+    },
+    {
+      "epoch": 1.7588094903832654,
+      "grad_norm": 0.49660564296640625,
+      "learning_rate": 1.6376657200398117e-05,
+      "loss": 0.3019,
+      "step": 1566
+    },
+    {
+      "epoch": 1.7599326126632038,
+      "grad_norm": 0.43644278562274014,
+      "learning_rate": 1.6370612766552422e-05,
+      "loss": 0.2656,
+      "step": 1567
+    },
+    {
+      "epoch": 1.761055734943142,
+      "grad_norm": 0.45767759479301445,
+      "learning_rate": 1.636456441279859e-05,
+      "loss": 0.2947,
+      "step": 1568
+    },
+    {
+      "epoch": 1.76217885722308,
+      "grad_norm": 0.4594875195235337,
+      "learning_rate": 1.6358512142858234e-05,
+      "loss": 0.2993,
+      "step": 1569
+    },
+    {
+      "epoch": 1.7633019795030185,
+      "grad_norm": 0.48736759713026184,
+      "learning_rate": 1.6352455960455385e-05,
+      "loss": 0.289,
+      "step": 1570
+    },
+    {
+      "epoch": 1.7644251017829566,
+      "grad_norm": 0.4544717098398387,
+      "learning_rate": 1.634639586931648e-05,
+      "loss": 0.2781,
+      "step": 1571
+    },
+    {
+      "epoch": 1.7655482240628948,
+      "grad_norm": 0.4587177062509317,
+      "learning_rate": 1.6340331873170356e-05,
+      "loss": 0.2834,
+      "step": 1572
+    },
+    {
+      "epoch": 1.7666713463428332,
+      "grad_norm": 0.48310847080548047,
+      "learning_rate": 1.6334263975748263e-05,
+      "loss": 0.2921,
+      "step": 1573
+    },
+    {
+      "epoch": 1.7677944686227713,
+      "grad_norm": 0.46119532443782013,
+      "learning_rate": 1.632819218078383e-05,
+      "loss": 0.29,
+      "step": 1574
+    },
+    {
+      "epoch": 1.7689175909027095,
+      "grad_norm": 0.46068898687141085,
+      "learning_rate": 1.6322116492013116e-05,
+      "loss": 0.299,
+      "step": 1575
+    },
+    {
+      "epoch": 1.7700407131826479,
+      "grad_norm": 0.4439847667371739,
+      "learning_rate": 1.6316036913174555e-05,
+      "loss": 0.2855,
+      "step": 1576
+    },
+    {
+      "epoch": 1.7711638354625858,
+      "grad_norm": 0.4441579095125552,
+      "learning_rate": 1.630995344800897e-05,
+      "loss": 0.3003,
+      "step": 1577
+    },
+    {
+      "epoch": 1.7722869577425242,
+      "grad_norm": 0.4527313840660522,
+      "learning_rate": 1.6303866100259595e-05,
+      "loss": 0.2885,
+      "step": 1578
+    },
+    {
+      "epoch": 1.7734100800224626,
+      "grad_norm": 0.4295498490008525,
+      "learning_rate": 1.6297774873672036e-05,
+      "loss": 0.2913,
+      "step": 1579
+    },
+    {
+      "epoch": 1.7745332023024005,
+      "grad_norm": 0.4506045912150568,
+      "learning_rate": 1.6291679771994293e-05,
+      "loss": 0.2919,
+      "step": 1580
+    },
+    {
+      "epoch": 1.775656324582339,
+      "grad_norm": 0.43120215422811353,
+      "learning_rate": 1.6285580798976754e-05,
+      "loss": 0.288,
+      "step": 1581
+    },
+    {
+      "epoch": 1.7767794468622773,
+      "grad_norm": 0.45220082498818925,
+      "learning_rate": 1.6279477958372175e-05,
+      "loss": 0.2902,
+      "step": 1582
+    },
+    {
+      "epoch": 1.7779025691422152,
+      "grad_norm": 0.4292832161883602,
+      "learning_rate": 1.6273371253935707e-05,
+      "loss": 0.2807,
+      "step": 1583
+    },
+    {
+      "epoch": 1.7790256914221536,
+      "grad_norm": 0.43823143313617324,
+      "learning_rate": 1.626726068942487e-05,
+      "loss": 0.2899,
+      "step": 1584
+    },
+    {
+      "epoch": 1.7801488137020918,
+      "grad_norm": 0.43017131250609614,
+      "learning_rate": 1.6261146268599564e-05,
+      "loss": 0.2717,
+      "step": 1585
+    },
+    {
+      "epoch": 1.78127193598203,
+      "grad_norm": 0.45356581546068614,
+      "learning_rate": 1.6255027995222056e-05,
+      "loss": 0.2887,
+      "step": 1586
+    },
+    {
+      "epoch": 1.7823950582619683,
+      "grad_norm": 0.44928986844226365,
+      "learning_rate": 1.624890587305699e-05,
+      "loss": 0.2861,
+      "step": 1587
+    },
+    {
+      "epoch": 1.7835181805419065,
+      "grad_norm": 0.4524148407584021,
+      "learning_rate": 1.6242779905871375e-05,
+      "loss": 0.2751,
+      "step": 1588
+    },
+    {
+      "epoch": 1.7846413028218446,
+      "grad_norm": 0.43075367830451183,
+      "learning_rate": 1.6236650097434586e-05,
+      "loss": 0.2807,
+      "step": 1589
+    },
+    {
+      "epoch": 1.785764425101783,
+      "grad_norm": 0.46285463495616647,
+      "learning_rate": 1.623051645151836e-05,
+      "loss": 0.3121,
+      "step": 1590
+    },
+    {
+      "epoch": 1.7868875473817212,
+      "grad_norm": 0.43530135150806504,
+      "learning_rate": 1.6224378971896798e-05,
+      "loss": 0.2819,
+      "step": 1591
+    },
+    {
+      "epoch": 1.7880106696616593,
+      "grad_norm": 0.490996011158021,
+      "learning_rate": 1.6218237662346356e-05,
+      "loss": 0.2893,
+      "step": 1592
+    },
+    {
+      "epoch": 1.7891337919415977,
+      "grad_norm": 0.4647817119466746,
+      "learning_rate": 1.6212092526645854e-05,
+      "loss": 0.2843,
+      "step": 1593
+    },
+    {
+      "epoch": 1.7902569142215359,
+      "grad_norm": 0.4413295898097072,
+      "learning_rate": 1.6205943568576457e-05,
+      "loss": 0.2799,
+      "step": 1594
+    },
+    {
+      "epoch": 1.791380036501474,
+      "grad_norm": 0.43007143575326345,
+      "learning_rate": 1.6199790791921693e-05,
+      "loss": 0.2593,
+      "step": 1595
+    },
+    {
+      "epoch": 1.7925031587814124,
+      "grad_norm": 0.4881130643190996,
+      "learning_rate": 1.6193634200467426e-05,
+      "loss": 0.2961,
+      "step": 1596
+    },
+    {
+      "epoch": 1.7936262810613506,
+      "grad_norm": 0.4528464937764476,
+      "learning_rate": 1.618747379800188e-05,
+      "loss": 0.2854,
+      "step": 1597
+    },
+    {
+      "epoch": 1.7947494033412887,
+      "grad_norm": 0.45653808613442953,
+      "learning_rate": 1.6181309588315616e-05,
+      "loss": 0.2919,
+      "step": 1598
+    },
+    {
+      "epoch": 1.7958725256212271,
+      "grad_norm": 0.4327325119499413,
+      "learning_rate": 1.6175141575201537e-05,
+      "loss": 0.2709,
+      "step": 1599
+    },
+    {
+      "epoch": 1.7969956479011653,
+      "grad_norm": 0.4377952749193301,
+      "learning_rate": 1.6168969762454897e-05,
+      "loss": 0.2761,
+      "step": 1600
+    },
+    {
+      "epoch": 1.7981187701811034,
+      "grad_norm": 0.4878789221305916,
+      "learning_rate": 1.616279415387327e-05,
+      "loss": 0.312,
+      "step": 1601
+    },
+    {
+      "epoch": 1.7992418924610418,
+      "grad_norm": 0.4072202956012937,
+      "learning_rate": 1.6156614753256583e-05,
+      "loss": 0.2731,
+      "step": 1602
+    },
+    {
+      "epoch": 1.8003650147409798,
+      "grad_norm": 0.425387058556243,
+      "learning_rate": 1.615043156440709e-05,
+      "loss": 0.2647,
+      "step": 1603
+    },
+    {
+      "epoch": 1.8014881370209181,
+      "grad_norm": 0.44570656361437866,
+      "learning_rate": 1.6144244591129373e-05,
+      "loss": 0.2856,
+      "step": 1604
+    },
+    {
+      "epoch": 1.8026112593008565,
+      "grad_norm": 0.5020676279175268,
+      "learning_rate": 1.6138053837230345e-05,
+      "loss": 0.3087,
+      "step": 1605
+    },
+    {
+      "epoch": 1.8037343815807945,
+      "grad_norm": 0.43825994146175534,
+      "learning_rate": 1.6131859306519243e-05,
+      "loss": 0.2899,
+      "step": 1606
+    },
+    {
+      "epoch": 1.8048575038607328,
+      "grad_norm": 0.4446328374851358,
+      "learning_rate": 1.612566100280763e-05,
+      "loss": 0.28,
+      "step": 1607
+    },
+    {
+      "epoch": 1.8059806261406712,
+      "grad_norm": 0.4774225518830962,
+      "learning_rate": 1.6119458929909394e-05,
+      "loss": 0.3031,
+      "step": 1608
+    },
+    {
+      "epoch": 1.8071037484206092,
+      "grad_norm": 0.426328321425137,
+      "learning_rate": 1.611325309164074e-05,
+      "loss": 0.2766,
+      "step": 1609
+    },
+    {
+      "epoch": 1.8082268707005476,
+      "grad_norm": 0.43586684491416294,
+      "learning_rate": 1.610704349182018e-05,
+      "loss": 0.2928,
+      "step": 1610
+    },
+    {
+      "epoch": 1.8093499929804857,
+      "grad_norm": 0.4567133760129506,
+      "learning_rate": 1.6100830134268558e-05,
+      "loss": 0.2722,
+      "step": 1611
+    },
+    {
+      "epoch": 1.8104731152604239,
+      "grad_norm": 0.45026824414792077,
+      "learning_rate": 1.6094613022809017e-05,
+      "loss": 0.2876,
+      "step": 1612
+    },
+    {
+      "epoch": 1.8115962375403623,
+      "grad_norm": 0.45757626469045953,
+      "learning_rate": 1.6088392161267018e-05,
+      "loss": 0.2961,
+      "step": 1613
+    },
+    {
+      "epoch": 1.8127193598203004,
+      "grad_norm": 0.4502992493729303,
+      "learning_rate": 1.6082167553470318e-05,
+      "loss": 0.3049,
+      "step": 1614
+    },
+    {
+      "epoch": 1.8138424821002386,
+      "grad_norm": 0.4454769781082224,
+      "learning_rate": 1.607593920324899e-05,
+      "loss": 0.2817,
+      "step": 1615
+    },
+    {
+      "epoch": 1.814965604380177,
+      "grad_norm": 0.4435355751540588,
+      "learning_rate": 1.606970711443541e-05,
+      "loss": 0.2704,
+      "step": 1616
+    },
+    {
+      "epoch": 1.8160887266601151,
+      "grad_norm": 0.4434669970265584,
+      "learning_rate": 1.606347129086425e-05,
+      "loss": 0.2893,
+      "step": 1617
+    },
+    {
+      "epoch": 1.8172118489400533,
+      "grad_norm": 0.4472789556626577,
+      "learning_rate": 1.6057231736372478e-05,
+      "loss": 0.2784,
+      "step": 1618
+    },
+    {
+      "epoch": 1.8183349712199917,
+      "grad_norm": 0.45291111323420824,
+      "learning_rate": 1.605098845479936e-05,
+      "loss": 0.2933,
+      "step": 1619
+    },
+    {
+      "epoch": 1.8194580934999298,
+      "grad_norm": 0.4599287948047513,
+      "learning_rate": 1.6044741449986458e-05,
+      "loss": 0.2984,
+      "step": 1620
+    },
+    {
+      "epoch": 1.820581215779868,
+      "grad_norm": 0.4584871742864635,
+      "learning_rate": 1.6038490725777624e-05,
+      "loss": 0.3051,
+      "step": 1621
+    },
+    {
+      "epoch": 1.8217043380598064,
+      "grad_norm": 0.4696965226478849,
+      "learning_rate": 1.6032236286018995e-05,
+      "loss": 0.3184,
+      "step": 1622
+    },
+    {
+      "epoch": 1.8228274603397445,
+      "grad_norm": 0.4456532072721999,
+      "learning_rate": 1.6025978134559e-05,
+      "loss": 0.2814,
+      "step": 1623
+    },
+    {
+      "epoch": 1.8239505826196827,
+      "grad_norm": 0.45088143754678195,
+      "learning_rate": 1.6019716275248342e-05,
+      "loss": 0.3047,
+      "step": 1624
+    },
+    {
+      "epoch": 1.825073704899621,
+      "grad_norm": 0.43175491411132033,
+      "learning_rate": 1.6013450711940017e-05,
+      "loss": 0.274,
+      "step": 1625
+    },
+    {
+      "epoch": 1.8261968271795592,
+      "grad_norm": 0.46280692541168383,
+      "learning_rate": 1.600718144848929e-05,
+      "loss": 0.2808,
+      "step": 1626
+    },
+    {
+      "epoch": 1.8273199494594974,
+      "grad_norm": 0.45335668599330486,
+      "learning_rate": 1.600090848875372e-05,
+      "loss": 0.2749,
+      "step": 1627
+    },
+    {
+      "epoch": 1.8284430717394358,
+      "grad_norm": 0.4903202775753055,
+      "learning_rate": 1.5994631836593116e-05,
+      "loss": 0.3107,
+      "step": 1628
+    },
+    {
+      "epoch": 1.8295661940193737,
+      "grad_norm": 0.4715442977840691,
+      "learning_rate": 1.5988351495869574e-05,
+      "loss": 0.3007,
+      "step": 1629
+    },
+    {
+      "epoch": 1.830689316299312,
+      "grad_norm": 0.4535990104723046,
+      "learning_rate": 1.598206747044746e-05,
+      "loss": 0.2845,
+      "step": 1630
+    },
+    {
+      "epoch": 1.8318124385792505,
+      "grad_norm": 0.45369244026805666,
+      "learning_rate": 1.59757797641934e-05,
+      "loss": 0.2893,
+      "step": 1631
+    },
+    {
+      "epoch": 1.8329355608591884,
+      "grad_norm": 0.4406335157305204,
+      "learning_rate": 1.596948838097629e-05,
+      "loss": 0.2882,
+      "step": 1632
+    },
+    {
+      "epoch": 1.8340586831391268,
+      "grad_norm": 0.4471966455869977,
+      "learning_rate": 1.596319332466729e-05,
+      "loss": 0.275,
+      "step": 1633
+    },
+    {
+      "epoch": 1.835181805419065,
+      "grad_norm": 0.43360372640049477,
+      "learning_rate": 1.5956894599139814e-05,
+      "loss": 0.2819,
+      "step": 1634
+    },
+    {
+      "epoch": 1.836304927699003,
+      "grad_norm": 0.43343847003699065,
+      "learning_rate": 1.5950592208269536e-05,
+      "loss": 0.2875,
+      "step": 1635
+    },
+    {
+      "epoch": 1.8374280499789415,
+      "grad_norm": 0.4165129497318286,
+      "learning_rate": 1.5944286155934396e-05,
+      "loss": 0.2636,
+      "step": 1636
+    },
+    {
+      "epoch": 1.8385511722588797,
+      "grad_norm": 0.45238285001736483,
+      "learning_rate": 1.5937976446014563e-05,
+      "loss": 0.263,
+      "step": 1637
+    },
+    {
+      "epoch": 1.8396742945388178,
+      "grad_norm": 0.4466843595442531,
+      "learning_rate": 1.593166308239248e-05,
+      "loss": 0.276,
+      "step": 1638
+    },
+    {
+      "epoch": 1.8407974168187562,
+      "grad_norm": 0.4358078236250983,
+      "learning_rate": 1.5925346068952833e-05,
+      "loss": 0.2951,
+      "step": 1639
+    },
+    {
+      "epoch": 1.8419205390986944,
+      "grad_norm": 0.44824760657818036,
+      "learning_rate": 1.5919025409582537e-05,
+      "loss": 0.3078,
+      "step": 1640
+    },
+    {
+      "epoch": 1.8430436613786325,
+      "grad_norm": 0.43735777546518884,
+      "learning_rate": 1.5912701108170777e-05,
+      "loss": 0.2877,
+      "step": 1641
+    },
+    {
+      "epoch": 1.844166783658571,
+      "grad_norm": 0.4578607904759152,
+      "learning_rate": 1.5906373168608952e-05,
+      "loss": 0.2865,
+      "step": 1642
+    },
+    {
+      "epoch": 1.845289905938509,
+      "grad_norm": 0.41061673520301367,
+      "learning_rate": 1.5900041594790722e-05,
+      "loss": 0.2744,
+      "step": 1643
+    },
+    {
+      "epoch": 1.8464130282184472,
+      "grad_norm": 0.4765113577474518,
+      "learning_rate": 1.5893706390611978e-05,
+      "loss": 0.3107,
+      "step": 1644
+    },
+    {
+      "epoch": 1.8475361504983856,
+      "grad_norm": 0.435712372199893,
+      "learning_rate": 1.5887367559970825e-05,
+      "loss": 0.2664,
+      "step": 1645
+    },
+    {
+      "epoch": 1.8486592727783238,
+      "grad_norm": 0.4753377958035371,
+      "learning_rate": 1.588102510676763e-05,
+      "loss": 0.2915,
+      "step": 1646
+    },
+    {
+      "epoch": 1.849782395058262,
+      "grad_norm": 0.45798588409720853,
+      "learning_rate": 1.5874679034904966e-05,
+      "loss": 0.2794,
+      "step": 1647
+    },
+    {
+      "epoch": 1.8509055173382003,
+      "grad_norm": 0.42891102168652906,
+      "learning_rate": 1.5868329348287647e-05,
+      "loss": 0.2712,
+      "step": 1648
+    },
+    {
+      "epoch": 1.8520286396181385,
+      "grad_norm": 0.45380196793915967,
+      "learning_rate": 1.58619760508227e-05,
+      "loss": 0.2872,
+      "step": 1649
+    },
+    {
+      "epoch": 1.8531517618980766,
+      "grad_norm": 0.4721744608194615,
+      "learning_rate": 1.5855619146419382e-05,
+      "loss": 0.3042,
+      "step": 1650
+    },
+    {
+      "epoch": 1.854274884178015,
+      "grad_norm": 0.4426415279566046,
+      "learning_rate": 1.5849258638989166e-05,
+      "loss": 0.2886,
+      "step": 1651
+    },
+    {
+      "epoch": 1.8553980064579532,
+      "grad_norm": 0.4403624280134429,
+      "learning_rate": 1.5842894532445738e-05,
+      "loss": 0.29,
+      "step": 1652
+    },
+    {
+      "epoch": 1.8565211287378913,
+      "grad_norm": 0.44346698610518404,
+      "learning_rate": 1.583652683070501e-05,
+      "loss": 0.286,
+      "step": 1653
+    },
+    {
+      "epoch": 1.8576442510178297,
+      "grad_norm": 0.43077248160418,
+      "learning_rate": 1.5830155537685093e-05,
+      "loss": 0.2822,
+      "step": 1654
+    },
+    {
+      "epoch": 1.8587673732977676,
+      "grad_norm": 0.4347083314214242,
+      "learning_rate": 1.5823780657306313e-05,
+      "loss": 0.2749,
+      "step": 1655
+    },
+    {
+      "epoch": 1.859890495577706,
+      "grad_norm": 0.4278133742132193,
+      "learning_rate": 1.581740219349121e-05,
+      "loss": 0.2656,
+      "step": 1656
+    },
+    {
+      "epoch": 1.8610136178576444,
+      "grad_norm": 0.418571552787694,
+      "learning_rate": 1.5811020150164518e-05,
+      "loss": 0.2563,
+      "step": 1657
+    },
+    {
+      "epoch": 1.8621367401375823,
+      "grad_norm": 0.4692428605964284,
+      "learning_rate": 1.5804634531253184e-05,
+      "loss": 0.2982,
+      "step": 1658
+    },
+    {
+      "epoch": 1.8632598624175207,
+      "grad_norm": 0.4763812793557251,
+      "learning_rate": 1.5798245340686342e-05,
+      "loss": 0.2839,
+      "step": 1659
+    },
+    {
+      "epoch": 1.864382984697459,
+      "grad_norm": 0.45442752765196487,
+      "learning_rate": 1.5791852582395334e-05,
+      "loss": 0.3055,
+      "step": 1660
+    },
+    {
+      "epoch": 1.865506106977397,
+      "grad_norm": 0.4210856174208563,
+      "learning_rate": 1.5785456260313702e-05,
+      "loss": 0.2787,
+      "step": 1661
+    },
+    {
+      "epoch": 1.8666292292573354,
+      "grad_norm": 0.4701122518568725,
+      "learning_rate": 1.577905637837716e-05,
+      "loss": 0.297,
+      "step": 1662
+    },
+    {
+      "epoch": 1.8677523515372736,
+      "grad_norm": 0.45300215556011086,
+      "learning_rate": 1.5772652940523637e-05,
+      "loss": 0.287,
+      "step": 1663
+    },
+    {
+      "epoch": 1.8688754738172118,
+      "grad_norm": 0.4483737642832065,
+      "learning_rate": 1.576624595069323e-05,
+      "loss": 0.2889,
+      "step": 1664
+    },
+    {
+      "epoch": 1.8699985960971501,
+      "grad_norm": 0.44444518527475235,
+      "learning_rate": 1.575983541282824e-05,
+      "loss": 0.278,
+      "step": 1665
+    },
+    {
+      "epoch": 1.8711217183770883,
+      "grad_norm": 0.4736568759863554,
+      "learning_rate": 1.5753421330873134e-05,
+      "loss": 0.2886,
+      "step": 1666
+    },
+    {
+      "epoch": 1.8722448406570265,
+      "grad_norm": 0.47085233017222466,
+      "learning_rate": 1.574700370877457e-05,
+      "loss": 0.2988,
+      "step": 1667
+    },
+    {
+      "epoch": 1.8733679629369648,
+      "grad_norm": 0.4503283744113668,
+      "learning_rate": 1.574058255048138e-05,
+      "loss": 0.2842,
+      "step": 1668
+    },
+    {
+      "epoch": 1.874491085216903,
+      "grad_norm": 0.45383270715255947,
+      "learning_rate": 1.5734157859944574e-05,
+      "loss": 0.2956,
+      "step": 1669
+    },
+    {
+      "epoch": 1.8756142074968412,
+      "grad_norm": 0.458441923267404,
+      "learning_rate": 1.5727729641117338e-05,
+      "loss": 0.3,
+      "step": 1670
+    },
+    {
+      "epoch": 1.8767373297767795,
+      "grad_norm": 0.45050041121130363,
+      "learning_rate": 1.5721297897955023e-05,
+      "loss": 0.2843,
+      "step": 1671
+    },
+    {
+      "epoch": 1.8778604520567177,
+      "grad_norm": 0.42269479225207923,
+      "learning_rate": 1.5714862634415145e-05,
+      "loss": 0.2755,
+      "step": 1672
+    },
+    {
+      "epoch": 1.8789835743366559,
+      "grad_norm": 0.4118289888578974,
+      "learning_rate": 1.5708423854457408e-05,
+      "loss": 0.2683,
+      "step": 1673
+    },
+    {
+      "epoch": 1.8801066966165942,
+      "grad_norm": 0.4519087067087661,
+      "learning_rate": 1.5701981562043648e-05,
+      "loss": 0.2941,
+      "step": 1674
+    },
+    {
+      "epoch": 1.8812298188965324,
+      "grad_norm": 0.45169859494837117,
+      "learning_rate": 1.569553576113789e-05,
+      "loss": 0.3081,
+      "step": 1675
+    },
+    {
+      "epoch": 1.8823529411764706,
+      "grad_norm": 0.45818548968025397,
+      "learning_rate": 1.56890864557063e-05,
+      "loss": 0.2906,
+      "step": 1676
+    },
+    {
+      "epoch": 1.883476063456409,
+      "grad_norm": 0.41903925729828695,
+      "learning_rate": 1.5682633649717206e-05,
+      "loss": 0.256,
+      "step": 1677
+    },
+    {
+      "epoch": 1.8845991857363469,
+      "grad_norm": 0.4494067572971225,
+      "learning_rate": 1.5676177347141096e-05,
+      "loss": 0.2999,
+      "step": 1678
+    },
+    {
+      "epoch": 1.8857223080162853,
+      "grad_norm": 0.430053388913964,
+      "learning_rate": 1.5669717551950595e-05,
+      "loss": 0.2778,
+      "step": 1679
+    },
+    {
+      "epoch": 1.8868454302962236,
+      "grad_norm": 0.45114702164180737,
+      "learning_rate": 1.5663254268120497e-05,
+      "loss": 0.2782,
+      "step": 1680
+    },
+    {
+      "epoch": 1.8879685525761616,
+      "grad_norm": 0.4767521278617856,
+      "learning_rate": 1.5656787499627727e-05,
+      "loss": 0.3075,
+      "step": 1681
+    },
+    {
+      "epoch": 1.8890916748561,
+      "grad_norm": 0.43432528267843223,
+      "learning_rate": 1.5650317250451357e-05,
+      "loss": 0.2925,
+      "step": 1682
+    },
+    {
+      "epoch": 1.8902147971360383,
+      "grad_norm": 0.43280983326124206,
+      "learning_rate": 1.5643843524572605e-05,
+      "loss": 0.2833,
+      "step": 1683
+    },
+    {
+      "epoch": 1.8913379194159763,
+      "grad_norm": 0.4460244169687265,
+      "learning_rate": 1.5637366325974823e-05,
+      "loss": 0.3001,
+      "step": 1684
+    },
+    {
+      "epoch": 1.8924610416959147,
+      "grad_norm": 0.4748835442825777,
+      "learning_rate": 1.5630885658643508e-05,
+      "loss": 0.305,
+      "step": 1685
+    },
+    {
+      "epoch": 1.8935841639758528,
+      "grad_norm": 0.4145939481842465,
+      "learning_rate": 1.5624401526566277e-05,
+      "loss": 0.2629,
+      "step": 1686
+    },
+    {
+      "epoch": 1.894707286255791,
+      "grad_norm": 0.452591025755552,
+      "learning_rate": 1.5617913933732892e-05,
+      "loss": 0.3203,
+      "step": 1687
+    },
+    {
+      "epoch": 1.8958304085357294,
+      "grad_norm": 0.42763483263940355,
+      "learning_rate": 1.5611422884135245e-05,
+      "loss": 0.2815,
+      "step": 1688
+    },
+    {
+      "epoch": 1.8969535308156675,
+      "grad_norm": 0.43158239697123274,
+      "learning_rate": 1.5604928381767345e-05,
+      "loss": 0.2793,
+      "step": 1689
+    },
+    {
+      "epoch": 1.8980766530956057,
+      "grad_norm": 0.4293853340809833,
+      "learning_rate": 1.5598430430625335e-05,
+      "loss": 0.2809,
+      "step": 1690
+    },
+    {
+      "epoch": 1.899199775375544,
+      "grad_norm": 0.46535273456706855,
+      "learning_rate": 1.5591929034707468e-05,
+      "loss": 0.3093,
+      "step": 1691
+    },
+    {
+      "epoch": 1.9003228976554822,
+      "grad_norm": 0.43293986112214783,
+      "learning_rate": 1.5585424198014135e-05,
+      "loss": 0.2704,
+      "step": 1692
+    },
+    {
+      "epoch": 1.9014460199354204,
+      "grad_norm": 0.47308351663116804,
+      "learning_rate": 1.5578915924547824e-05,
+      "loss": 0.3087,
+      "step": 1693
+    },
+    {
+      "epoch": 1.9025691422153588,
+      "grad_norm": 0.45499586114643786,
+      "learning_rate": 1.557240421831315e-05,
+      "loss": 0.2979,
+      "step": 1694
+    },
+    {
+      "epoch": 1.903692264495297,
+      "grad_norm": 0.43179402389716814,
+      "learning_rate": 1.5565889083316847e-05,
+      "loss": 0.2795,
+      "step": 1695
+    },
+    {
+      "epoch": 1.904815386775235,
+      "grad_norm": 0.4337080321597782,
+      "learning_rate": 1.5559370523567734e-05,
+      "loss": 0.2884,
+      "step": 1696
+    },
+    {
+      "epoch": 1.9059385090551735,
+      "grad_norm": 0.4416737410605399,
+      "learning_rate": 1.5552848543076762e-05,
+      "loss": 0.2916,
+      "step": 1697
+    },
+    {
+      "epoch": 1.9070616313351116,
+      "grad_norm": 0.45327771862806987,
+      "learning_rate": 1.5546323145856976e-05,
+      "loss": 0.283,
+      "step": 1698
+    },
+    {
+      "epoch": 1.9081847536150498,
+      "grad_norm": 0.43897115726266495,
+      "learning_rate": 1.5539794335923523e-05,
+      "loss": 0.2757,
+      "step": 1699
+    },
+    {
+      "epoch": 1.9093078758949882,
+      "grad_norm": 0.45535826828398407,
+      "learning_rate": 1.553326211729365e-05,
+      "loss": 0.2846,
+      "step": 1700
+    },
+    {
+      "epoch": 1.9104309981749263,
+      "grad_norm": 0.44632754329033314,
+      "learning_rate": 1.5526726493986707e-05,
+      "loss": 0.2854,
+      "step": 1701
+    },
+    {
+      "epoch": 1.9115541204548645,
+      "grad_norm": 0.44264267367794924,
+      "learning_rate": 1.5520187470024138e-05,
+      "loss": 0.291,
+      "step": 1702
+    },
+    {
+      "epoch": 1.9126772427348029,
+      "grad_norm": 0.4261191699485715,
+      "learning_rate": 1.5513645049429468e-05,
+      "loss": 0.2744,
+      "step": 1703
+    },
+    {
+      "epoch": 1.9138003650147408,
+      "grad_norm": 0.4718910711066927,
+      "learning_rate": 1.550709923622832e-05,
+      "loss": 0.3094,
+      "step": 1704
+    },
+    {
+      "epoch": 1.9149234872946792,
+      "grad_norm": 0.443144418899658,
+      "learning_rate": 1.5500550034448415e-05,
+      "loss": 0.265,
+      "step": 1705
+    },
+    {
+      "epoch": 1.9160466095746176,
+      "grad_norm": 0.4680353665068334,
+      "learning_rate": 1.549399744811954e-05,
+      "loss": 0.2984,
+      "step": 1706
+    },
+    {
+      "epoch": 1.9171697318545555,
+      "grad_norm": 0.451871229122914,
+      "learning_rate": 1.5487441481273576e-05,
+      "loss": 0.2856,
+      "step": 1707
+    },
+    {
+      "epoch": 1.918292854134494,
+      "grad_norm": 0.45525795558547877,
+      "learning_rate": 1.5480882137944483e-05,
+      "loss": 0.2972,
+      "step": 1708
+    },
+    {
+      "epoch": 1.919415976414432,
+      "grad_norm": 0.457501569499871,
+      "learning_rate": 1.54743194221683e-05,
+      "loss": 0.2842,
+      "step": 1709
+    },
+    {
+      "epoch": 1.9205390986943702,
+      "grad_norm": 0.40429630388545335,
+      "learning_rate": 1.546775333798313e-05,
+      "loss": 0.2635,
+      "step": 1710
+    },
+    {
+      "epoch": 1.9216622209743086,
+      "grad_norm": 0.43949600121406834,
+      "learning_rate": 1.5461183889429163e-05,
+      "loss": 0.287,
+      "step": 1711
+    },
+    {
+      "epoch": 1.9227853432542468,
+      "grad_norm": 0.44623475635576615,
+      "learning_rate": 1.545461108054865e-05,
+      "loss": 0.295,
+      "step": 1712
+    },
+    {
+      "epoch": 1.923908465534185,
+      "grad_norm": 0.4332274620156582,
+      "learning_rate": 1.5448034915385912e-05,
+      "loss": 0.2764,
+      "step": 1713
+    },
+    {
+      "epoch": 1.9250315878141233,
+      "grad_norm": 0.43976301663859074,
+      "learning_rate": 1.5441455397987342e-05,
+      "loss": 0.2796,
+      "step": 1714
+    },
+    {
+      "epoch": 1.9261547100940615,
+      "grad_norm": 0.48004372769154424,
+      "learning_rate": 1.543487253240138e-05,
+      "loss": 0.3315,
+      "step": 1715
+    },
+    {
+      "epoch": 1.9272778323739996,
+      "grad_norm": 0.44346770478396347,
+      "learning_rate": 1.5428286322678544e-05,
+      "loss": 0.3029,
+      "step": 1716
+    },
+    {
+      "epoch": 1.928400954653938,
+      "grad_norm": 0.452216885627314,
+      "learning_rate": 1.54216967728714e-05,
+      "loss": 0.2889,
+      "step": 1717
+    },
+    {
+      "epoch": 1.9295240769338762,
+      "grad_norm": 0.43338802482350897,
+      "learning_rate": 1.5415103887034565e-05,
+      "loss": 0.2847,
+      "step": 1718
+    },
+    {
+      "epoch": 1.9306471992138143,
+      "grad_norm": 0.42192888016892804,
+      "learning_rate": 1.540850766922472e-05,
+      "loss": 0.283,
+      "step": 1719
+    },
+    {
+      "epoch": 1.9317703214937527,
+      "grad_norm": 0.45160195890660065,
+      "learning_rate": 1.540190812350059e-05,
+      "loss": 0.2745,
+      "step": 1720
+    },
+    {
+      "epoch": 1.9328934437736909,
+      "grad_norm": 0.45447959773978414,
+      "learning_rate": 1.539530525392294e-05,
+      "loss": 0.3008,
+      "step": 1721
+    },
+    {
+      "epoch": 1.934016566053629,
+      "grad_norm": 0.4371017493232612,
+      "learning_rate": 1.53886990645546e-05,
+      "loss": 0.2948,
+      "step": 1722
+    },
+    {
+      "epoch": 1.9351396883335674,
+      "grad_norm": 0.4334080460875823,
+      "learning_rate": 1.5382089559460423e-05,
+      "loss": 0.2818,
+      "step": 1723
+    },
+    {
+      "epoch": 1.9362628106135056,
+      "grad_norm": 0.4731746980122572,
+      "learning_rate": 1.5375476742707314e-05,
+      "loss": 0.3038,
+      "step": 1724
+    },
+    {
+      "epoch": 1.9373859328934437,
+      "grad_norm": 0.454174378429582,
+      "learning_rate": 1.536886061836421e-05,
+      "loss": 0.2948,
+      "step": 1725
+    },
+    {
+      "epoch": 1.9385090551733821,
+      "grad_norm": 0.4432630231842709,
+      "learning_rate": 1.5362241190502086e-05,
+      "loss": 0.2852,
+      "step": 1726
+    },
+    {
+      "epoch": 1.9396321774533203,
+      "grad_norm": 0.4598840425874724,
+      "learning_rate": 1.5355618463193945e-05,
+      "loss": 0.2815,
+      "step": 1727
+    },
+    {
+      "epoch": 1.9407552997332584,
+      "grad_norm": 0.451831626436298,
+      "learning_rate": 1.5348992440514832e-05,
+      "loss": 0.2729,
+      "step": 1728
+    },
+    {
+      "epoch": 1.9418784220131968,
+      "grad_norm": 0.46544809796158676,
+      "learning_rate": 1.534236312654181e-05,
+      "loss": 0.3169,
+      "step": 1729
+    },
+    {
+      "epoch": 1.9430015442931348,
+      "grad_norm": 0.44645885927284146,
+      "learning_rate": 1.5335730525353962e-05,
+      "loss": 0.2785,
+      "step": 1730
+    },
+    {
+      "epoch": 1.9441246665730731,
+      "grad_norm": 0.469136480578403,
+      "learning_rate": 1.5329094641032406e-05,
+      "loss": 0.3056,
+      "step": 1731
+    },
+    {
+      "epoch": 1.9452477888530115,
+      "grad_norm": 0.43983149032385477,
+      "learning_rate": 1.5322455477660274e-05,
+      "loss": 0.2836,
+      "step": 1732
+    },
+    {
+      "epoch": 1.9463709111329495,
+      "grad_norm": 0.44719810124182485,
+      "learning_rate": 1.5315813039322714e-05,
+      "loss": 0.2801,
+      "step": 1733
+    },
+    {
+      "epoch": 1.9474940334128878,
+      "grad_norm": 0.4290827977526451,
+      "learning_rate": 1.5309167330106895e-05,
+      "loss": 0.2704,
+      "step": 1734
+    },
+    {
+      "epoch": 1.948617155692826,
+      "grad_norm": 0.4497567848165034,
+      "learning_rate": 1.5302518354101992e-05,
+      "loss": 0.2967,
+      "step": 1735
+    },
+    {
+      "epoch": 1.9497402779727642,
+      "grad_norm": 0.468862605371959,
+      "learning_rate": 1.5295866115399193e-05,
+      "loss": 0.3041,
+      "step": 1736
+    },
+    {
+      "epoch": 1.9508634002527026,
+      "grad_norm": 0.4888366595716739,
+      "learning_rate": 1.5289210618091695e-05,
+      "loss": 0.3024,
+      "step": 1737
+    },
+    {
+      "epoch": 1.9519865225326407,
+      "grad_norm": 0.476112693226597,
+      "learning_rate": 1.52825518662747e-05,
+      "loss": 0.2621,
+      "step": 1738
+    },
+    {
+      "epoch": 1.9531096448125789,
+      "grad_norm": 0.44772753456589154,
+      "learning_rate": 1.527588986404541e-05,
+      "loss": 0.2796,
+      "step": 1739
+    },
+    {
+      "epoch": 1.9542327670925173,
+      "grad_norm": 0.5107778899997112,
+      "learning_rate": 1.5269224615503025e-05,
+      "loss": 0.3308,
+      "step": 1740
+    },
+    {
+      "epoch": 1.9553558893724554,
+      "grad_norm": 0.4674299014245734,
+      "learning_rate": 1.5262556124748754e-05,
+      "loss": 0.2826,
+      "step": 1741
+    },
+    {
+      "epoch": 1.9564790116523936,
+      "grad_norm": 0.46960805580213283,
+      "learning_rate": 1.5255884395885785e-05,
+      "loss": 0.2939,
+      "step": 1742
+    },
+    {
+      "epoch": 1.957602133932332,
+      "grad_norm": 0.4209722750164603,
+      "learning_rate": 1.5249209433019307e-05,
+      "loss": 0.2722,
+      "step": 1743
+    },
+    {
+      "epoch": 1.9587252562122701,
+      "grad_norm": 0.4568155973367222,
+      "learning_rate": 1.5242531240256501e-05,
+      "loss": 0.2811,
+      "step": 1744
+    },
+    {
+      "epoch": 1.9598483784922083,
+      "grad_norm": 0.4307082739584054,
+      "learning_rate": 1.5235849821706531e-05,
+      "loss": 0.28,
+      "step": 1745
+    },
+    {
+      "epoch": 1.9609715007721467,
+      "grad_norm": 0.4234189998569159,
+      "learning_rate": 1.5229165181480552e-05,
+      "loss": 0.2671,
+      "step": 1746
+    },
+    {
+      "epoch": 1.9620946230520848,
+      "grad_norm": 0.4739925558985982,
+      "learning_rate": 1.5222477323691687e-05,
+      "loss": 0.307,
+      "step": 1747
+    },
+    {
+      "epoch": 1.963217745332023,
+      "grad_norm": 0.44810995148794674,
+      "learning_rate": 1.5215786252455056e-05,
+      "loss": 0.2925,
+      "step": 1748
+    },
+    {
+      "epoch": 1.9643408676119614,
+      "grad_norm": 0.43430736966909755,
+      "learning_rate": 1.5209091971887747e-05,
+      "loss": 0.281,
+      "step": 1749
+    },
+    {
+      "epoch": 1.9654639898918995,
+      "grad_norm": 0.44148058197118256,
+      "learning_rate": 1.5202394486108823e-05,
+      "loss": 0.2981,
+      "step": 1750
+    },
+    {
+      "epoch": 1.9665871121718377,
+      "grad_norm": 0.4108291724424686,
+      "learning_rate": 1.5195693799239322e-05,
+      "loss": 0.2675,
+      "step": 1751
+    },
+    {
+      "epoch": 1.967710234451776,
+      "grad_norm": 0.4412806106090446,
+      "learning_rate": 1.5188989915402253e-05,
+      "loss": 0.2827,
+      "step": 1752
+    },
+    {
+      "epoch": 1.968833356731714,
+      "grad_norm": 0.4529168112870875,
+      "learning_rate": 1.5182282838722584e-05,
+      "loss": 0.2939,
+      "step": 1753
+    },
+    {
+      "epoch": 1.9699564790116524,
+      "grad_norm": 0.41285809272736673,
+      "learning_rate": 1.5175572573327257e-05,
+      "loss": 0.2731,
+      "step": 1754
+    },
+    {
+      "epoch": 1.9710796012915908,
+      "grad_norm": 0.43190618430843086,
+      "learning_rate": 1.5168859123345172e-05,
+      "loss": 0.2774,
+      "step": 1755
+    },
+    {
+      "epoch": 1.9722027235715287,
+      "grad_norm": 0.4552856993615373,
+      "learning_rate": 1.5162142492907186e-05,
+      "loss": 0.2862,
+      "step": 1756
+    },
+    {
+      "epoch": 1.973325845851467,
+      "grad_norm": 0.4514244233597325,
+      "learning_rate": 1.5155422686146118e-05,
+      "loss": 0.2727,
+      "step": 1757
+    },
+    {
+      "epoch": 1.9744489681314055,
+      "grad_norm": 0.43333285646026426,
+      "learning_rate": 1.5148699707196739e-05,
+      "loss": 0.2774,
+      "step": 1758
+    },
+    {
+      "epoch": 1.9755720904113434,
+      "grad_norm": 0.4508656102546517,
+      "learning_rate": 1.5141973560195768e-05,
+      "loss": 0.2944,
+      "step": 1759
+    },
+    {
+      "epoch": 1.9766952126912818,
+      "grad_norm": 0.4743163650426743,
+      "learning_rate": 1.5135244249281884e-05,
+      "loss": 0.323,
+      "step": 1760
+    },
+    {
+      "epoch": 1.97781833497122,
+      "grad_norm": 0.4564248851686335,
+      "learning_rate": 1.5128511778595703e-05,
+      "loss": 0.2717,
+      "step": 1761
+    },
+    {
+      "epoch": 1.978941457251158,
+      "grad_norm": 0.470303991029109,
+      "learning_rate": 1.5121776152279786e-05,
+      "loss": 0.3205,
+      "step": 1762
+    },
+    {
+      "epoch": 1.9800645795310965,
+      "grad_norm": 0.4224772001099396,
+      "learning_rate": 1.5115037374478641e-05,
+      "loss": 0.2678,
+      "step": 1763
+    },
+    {
+      "epoch": 1.9811877018110347,
+      "grad_norm": 0.4457682890054829,
+      "learning_rate": 1.510829544933871e-05,
+      "loss": 0.297,
+      "step": 1764
+    },
+    {
+      "epoch": 1.9823108240909728,
+      "grad_norm": 0.4371918445392385,
+      "learning_rate": 1.5101550381008377e-05,
+      "loss": 0.272,
+      "step": 1765
+    },
+    {
+      "epoch": 1.9834339463709112,
+      "grad_norm": 0.4565606365981388,
+      "learning_rate": 1.5094802173637953e-05,
+      "loss": 0.3024,
+      "step": 1766
+    },
+    {
+      "epoch": 1.9845570686508494,
+      "grad_norm": 0.4712494126543963,
+      "learning_rate": 1.5088050831379684e-05,
+      "loss": 0.2958,
+      "step": 1767
+    },
+    {
+      "epoch": 1.9856801909307875,
+      "grad_norm": 0.4813432876456791,
+      "learning_rate": 1.508129635838775e-05,
+      "loss": 0.2996,
+      "step": 1768
+    },
+    {
+      "epoch": 1.986803313210726,
+      "grad_norm": 0.42456961820085154,
+      "learning_rate": 1.5074538758818247e-05,
+      "loss": 0.2701,
+      "step": 1769
+    },
+    {
+      "epoch": 1.987926435490664,
+      "grad_norm": 0.4466264321919789,
+      "learning_rate": 1.5067778036829204e-05,
+      "loss": 0.2844,
+      "step": 1770
+    },
+    {
+      "epoch": 1.9890495577706022,
+      "grad_norm": 0.45361513546811316,
+      "learning_rate": 1.5061014196580565e-05,
+      "loss": 0.279,
+      "step": 1771
+    },
+    {
+      "epoch": 1.9901726800505406,
+      "grad_norm": 0.4418593232740048,
+      "learning_rate": 1.5054247242234197e-05,
+      "loss": 0.2736,
+      "step": 1772
+    },
+    {
+      "epoch": 1.9912958023304788,
+      "grad_norm": 0.4860660602595377,
+      "learning_rate": 1.5047477177953887e-05,
+      "loss": 0.313,
+      "step": 1773
+    },
+    {
+      "epoch": 1.992418924610417,
+      "grad_norm": 0.43959539799018116,
+      "learning_rate": 1.5040704007905319e-05,
+      "loss": 0.2737,
+      "step": 1774
+    },
+    {
+      "epoch": 1.9935420468903553,
+      "grad_norm": 0.49294497786554403,
+      "learning_rate": 1.5033927736256107e-05,
+      "loss": 0.3041,
+      "step": 1775
+    },
+    {
+      "epoch": 1.9946651691702935,
+      "grad_norm": 0.42461221519723574,
+      "learning_rate": 1.5027148367175759e-05,
+      "loss": 0.2616,
+      "step": 1776
+    },
+    {
+      "epoch": 1.9957882914502316,
+      "grad_norm": 0.4402615651724432,
+      "learning_rate": 1.50203659048357e-05,
+      "loss": 0.2697,
+      "step": 1777
+    },
+    {
+      "epoch": 1.99691141373017,
+      "grad_norm": 0.46310993944775436,
+      "learning_rate": 1.5013580353409259e-05,
+      "loss": 0.2917,
+      "step": 1778
+    },
+    {
+      "epoch": 1.998034536010108,
+      "grad_norm": 0.46147694875445594,
+      "learning_rate": 1.500679171707165e-05,
+      "loss": 0.2791,
+      "step": 1779
+    },
+    {
+      "epoch": 1.9991576582900463,
+      "grad_norm": 0.4680035095821885,
+      "learning_rate": 1.5000000000000002e-05,
+      "loss": 0.2826,
+      "step": 1780
+    },
+    {
+      "epoch": 2.0002807805699847,
+      "grad_norm": 0.8085198551259174,
+      "learning_rate": 1.499320520637333e-05,
+      "loss": 0.4237,
+      "step": 1781
+    },
+    {
+      "epoch": 2.0014039028499226,
+      "grad_norm": 0.541462070507787,
+      "learning_rate": 1.4986407340372546e-05,
+      "loss": 0.177,
+      "step": 1782
+    },
+    {
+      "epoch": 2.002527025129861,
+      "grad_norm": 0.514819434140705,
+      "learning_rate": 1.4979606406180456e-05,
+      "loss": 0.1881,
+      "step": 1783
+    },
+    {
+      "epoch": 2.0036501474097994,
+      "grad_norm": 0.45400673841814165,
+      "learning_rate": 1.4972802407981744e-05,
+      "loss": 0.174,
+      "step": 1784
+    },
+    {
+      "epoch": 2.0047732696897373,
+      "grad_norm": 0.5894341185038876,
+      "learning_rate": 1.4965995349962987e-05,
+      "loss": 0.1669,
+      "step": 1785
+    },
+    {
+      "epoch": 2.0058963919696757,
+      "grad_norm": 0.780454484731567,
+      "learning_rate": 1.4959185236312642e-05,
+      "loss": 0.1779,
+      "step": 1786
+    },
+    {
+      "epoch": 2.007019514249614,
+      "grad_norm": 0.5115911523467761,
+      "learning_rate": 1.495237207122105e-05,
+      "loss": 0.1719,
+      "step": 1787
+    },
+    {
+      "epoch": 2.008142636529552,
+      "grad_norm": 0.47746167967424985,
+      "learning_rate": 1.4945555858880422e-05,
+      "loss": 0.1658,
+      "step": 1788
+    },
+    {
+      "epoch": 2.0092657588094904,
+      "grad_norm": 0.49235675543550705,
+      "learning_rate": 1.493873660348485e-05,
+      "loss": 0.1862,
+      "step": 1789
+    },
+    {
+      "epoch": 2.010388881089429,
+      "grad_norm": 0.4566009363822769,
+      "learning_rate": 1.49319143092303e-05,
+      "loss": 0.1768,
+      "step": 1790
+    },
+    {
+      "epoch": 2.0115120033693668,
+      "grad_norm": 0.4648626440058988,
+      "learning_rate": 1.4925088980314604e-05,
+      "loss": 0.1751,
+      "step": 1791
+    },
+    {
+      "epoch": 2.012635125649305,
+      "grad_norm": 0.45375792194533593,
+      "learning_rate": 1.4918260620937458e-05,
+      "loss": 0.1647,
+      "step": 1792
+    },
+    {
+      "epoch": 2.0137582479292435,
+      "grad_norm": 0.42584485799570987,
+      "learning_rate": 1.4911429235300425e-05,
+      "loss": 0.156,
+      "step": 1793
+    },
+    {
+      "epoch": 2.0148813702091815,
+      "grad_norm": 0.460137628374996,
+      "learning_rate": 1.490459482760694e-05,
+      "loss": 0.1708,
+      "step": 1794
+    },
+    {
+      "epoch": 2.01600449248912,
+      "grad_norm": 0.45344627339365856,
+      "learning_rate": 1.4897757402062285e-05,
+      "loss": 0.1567,
+      "step": 1795
+    },
+    {
+      "epoch": 2.0171276147690578,
+      "grad_norm": 0.4953474815052228,
+      "learning_rate": 1.48909169628736e-05,
+      "loss": 0.1707,
+      "step": 1796
+    },
+    {
+      "epoch": 2.018250737048996,
+      "grad_norm": 0.4753046189955386,
+      "learning_rate": 1.488407351424989e-05,
+      "loss": 0.1641,
+      "step": 1797
+    },
+    {
+      "epoch": 2.0193738593289345,
+      "grad_norm": 0.4519314500216831,
+      "learning_rate": 1.4877227060401997e-05,
+      "loss": 0.1668,
+      "step": 1798
+    },
+    {
+      "epoch": 2.0204969816088725,
+      "grad_norm": 0.45671522761655353,
+      "learning_rate": 1.4870377605542624e-05,
+      "loss": 0.1601,
+      "step": 1799
+    },
+    {
+      "epoch": 2.021620103888811,
+      "grad_norm": 0.456319516767008,
+      "learning_rate": 1.4863525153886314e-05,
+      "loss": 0.1641,
+      "step": 1800
+    },
+    {
+      "epoch": 2.0227432261687492,
+      "grad_norm": 0.45491434648308354,
+      "learning_rate": 1.4856669709649455e-05,
+      "loss": 0.1707,
+      "step": 1801
+    },
+    {
+      "epoch": 2.023866348448687,
+      "grad_norm": 0.4492812134103008,
+      "learning_rate": 1.4849811277050279e-05,
+      "loss": 0.1667,
+      "step": 1802
+    },
+    {
+      "epoch": 2.0249894707286256,
+      "grad_norm": 0.47837793383271804,
+      "learning_rate": 1.4842949860308854e-05,
+      "loss": 0.1762,
+      "step": 1803
+    },
+    {
+      "epoch": 2.026112593008564,
+      "grad_norm": 0.4366394160857708,
+      "learning_rate": 1.4836085463647088e-05,
+      "loss": 0.1627,
+      "step": 1804
+    },
+    {
+      "epoch": 2.027235715288502,
+      "grad_norm": 0.4819217078029733,
+      "learning_rate": 1.4829218091288713e-05,
+      "loss": 0.1833,
+      "step": 1805
+    },
+    {
+      "epoch": 2.0283588375684403,
+      "grad_norm": 0.4592333930190147,
+      "learning_rate": 1.4822347747459307e-05,
+      "loss": 0.1668,
+      "step": 1806
+    },
+    {
+      "epoch": 2.0294819598483786,
+      "grad_norm": 0.45422769074810193,
+      "learning_rate": 1.4815474436386263e-05,
+      "loss": 0.1683,
+      "step": 1807
+    },
+    {
+      "epoch": 2.0306050821283166,
+      "grad_norm": 0.4600329925419827,
+      "learning_rate": 1.4808598162298806e-05,
+      "loss": 0.1747,
+      "step": 1808
+    },
+    {
+      "epoch": 2.031728204408255,
+      "grad_norm": 0.4583134137959598,
+      "learning_rate": 1.4801718929427986e-05,
+      "loss": 0.1649,
+      "step": 1809
+    },
+    {
+      "epoch": 2.0328513266881933,
+      "grad_norm": 0.49051047861023445,
+      "learning_rate": 1.4794836742006667e-05,
+      "loss": 0.1714,
+      "step": 1810
+    },
+    {
+      "epoch": 2.0339744489681313,
+      "grad_norm": 0.45752788677919076,
+      "learning_rate": 1.4787951604269533e-05,
+      "loss": 0.17,
+      "step": 1811
+    },
+    {
+      "epoch": 2.0350975712480697,
+      "grad_norm": 0.46474367178946024,
+      "learning_rate": 1.478106352045309e-05,
+      "loss": 0.1784,
+      "step": 1812
+    },
+    {
+      "epoch": 2.036220693528008,
+      "grad_norm": 0.44888182231043505,
+      "learning_rate": 1.4774172494795651e-05,
+      "loss": 0.1655,
+      "step": 1813
+    },
+    {
+      "epoch": 2.037343815807946,
+      "grad_norm": 0.4874435144426795,
+      "learning_rate": 1.4767278531537335e-05,
+      "loss": 0.1746,
+      "step": 1814
+    },
+    {
+      "epoch": 2.0384669380878844,
+      "grad_norm": 0.45309578620813684,
+      "learning_rate": 1.476038163492008e-05,
+      "loss": 0.1598,
+      "step": 1815
+    },
+    {
+      "epoch": 2.0395900603678228,
+      "grad_norm": 0.47320504356622956,
+      "learning_rate": 1.4753481809187617e-05,
+      "loss": 0.1834,
+      "step": 1816
+    },
+    {
+      "epoch": 2.0407131826477607,
+      "grad_norm": 0.4720649193376181,
+      "learning_rate": 1.474657905858549e-05,
+      "loss": 0.1739,
+      "step": 1817
+    },
+    {
+      "epoch": 2.041836304927699,
+      "grad_norm": 0.4728143038805621,
+      "learning_rate": 1.4739673387361033e-05,
+      "loss": 0.1764,
+      "step": 1818
+    },
+    {
+      "epoch": 2.042959427207637,
+      "grad_norm": 0.44374918226300764,
+      "learning_rate": 1.4732764799763383e-05,
+      "loss": 0.1656,
+      "step": 1819
+    },
+    {
+      "epoch": 2.0440825494875754,
+      "grad_norm": 0.4923042826192989,
+      "learning_rate": 1.4725853300043472e-05,
+      "loss": 0.1987,
+      "step": 1820
+    },
+    {
+      "epoch": 2.0452056717675138,
+      "grad_norm": 0.46751435491816906,
+      "learning_rate": 1.4718938892454018e-05,
+      "loss": 0.1572,
+      "step": 1821
+    },
+    {
+      "epoch": 2.0463287940474517,
+      "grad_norm": 0.47339333449870347,
+      "learning_rate": 1.4712021581249534e-05,
+      "loss": 0.1776,
+      "step": 1822
+    },
+    {
+      "epoch": 2.04745191632739,
+      "grad_norm": 0.46020101583349715,
+      "learning_rate": 1.4705101370686316e-05,
+      "loss": 0.1604,
+      "step": 1823
+    },
+    {
+      "epoch": 2.0485750386073285,
+      "grad_norm": 0.49263307013323115,
+      "learning_rate": 1.469817826502245e-05,
+      "loss": 0.1668,
+      "step": 1824
+    },
+    {
+      "epoch": 2.0496981608872664,
+      "grad_norm": 0.4710764801250717,
+      "learning_rate": 1.4691252268517794e-05,
+      "loss": 0.1775,
+      "step": 1825
+    },
+    {
+      "epoch": 2.050821283167205,
+      "grad_norm": 0.49880730266155526,
+      "learning_rate": 1.4684323385433997e-05,
+      "loss": 0.1822,
+      "step": 1826
+    },
+    {
+      "epoch": 2.051944405447143,
+      "grad_norm": 0.46625550510782054,
+      "learning_rate": 1.4677391620034467e-05,
+      "loss": 0.174,
+      "step": 1827
+    },
+    {
+      "epoch": 2.053067527727081,
+      "grad_norm": 0.4872551100589393,
+      "learning_rate": 1.4670456976584401e-05,
+      "loss": 0.1656,
+      "step": 1828
+    },
+    {
+      "epoch": 2.0541906500070195,
+      "grad_norm": 0.43451500585845937,
+      "learning_rate": 1.4663519459350763e-05,
+      "loss": 0.1589,
+      "step": 1829
+    },
+    {
+      "epoch": 2.055313772286958,
+      "grad_norm": 0.428654270229592,
+      "learning_rate": 1.4656579072602281e-05,
+      "loss": 0.1596,
+      "step": 1830
+    },
+    {
+      "epoch": 2.056436894566896,
+      "grad_norm": 0.48053898356788094,
+      "learning_rate": 1.4649635820609457e-05,
+      "loss": 0.1831,
+      "step": 1831
+    },
+    {
+      "epoch": 2.057560016846834,
+      "grad_norm": 0.4472515086562773,
+      "learning_rate": 1.464268970764454e-05,
+      "loss": 0.1618,
+      "step": 1832
+    },
+    {
+      "epoch": 2.0586831391267726,
+      "grad_norm": 0.49366586224960823,
+      "learning_rate": 1.4635740737981557e-05,
+      "loss": 0.1771,
+      "step": 1833
+    },
+    {
+      "epoch": 2.0598062614067105,
+      "grad_norm": 0.455074161806074,
+      "learning_rate": 1.4628788915896282e-05,
+      "loss": 0.1661,
+      "step": 1834
+    },
+    {
+      "epoch": 2.060929383686649,
+      "grad_norm": 0.44819907357299266,
+      "learning_rate": 1.4621834245666254e-05,
+      "loss": 0.1563,
+      "step": 1835
+    },
+    {
+      "epoch": 2.0620525059665873,
+      "grad_norm": 0.4826290613150395,
+      "learning_rate": 1.4614876731570751e-05,
+      "loss": 0.1823,
+      "step": 1836
+    },
+    {
+      "epoch": 2.0631756282465252,
+      "grad_norm": 0.4657177556230326,
+      "learning_rate": 1.4607916377890807e-05,
+      "loss": 0.1703,
+      "step": 1837
+    },
+    {
+      "epoch": 2.0642987505264636,
+      "grad_norm": 0.47786499964299944,
+      "learning_rate": 1.4600953188909214e-05,
+      "loss": 0.1727,
+      "step": 1838
+    },
+    {
+      "epoch": 2.065421872806402,
+      "grad_norm": 0.4642252898495588,
+      "learning_rate": 1.4593987168910491e-05,
+      "loss": 0.1743,
+      "step": 1839
+    },
+    {
+      "epoch": 2.06654499508634,
+      "grad_norm": 0.4695641601793918,
+      "learning_rate": 1.4587018322180906e-05,
+      "loss": 0.1747,
+      "step": 1840
+    },
+    {
+      "epoch": 2.0676681173662783,
+      "grad_norm": 0.4662822283785743,
+      "learning_rate": 1.4580046653008474e-05,
+      "loss": 0.1692,
+      "step": 1841
+    },
+    {
+      "epoch": 2.0687912396462167,
+      "grad_norm": 0.4576878116997234,
+      "learning_rate": 1.457307216568293e-05,
+      "loss": 0.1643,
+      "step": 1842
+    },
+    {
+      "epoch": 2.0699143619261546,
+      "grad_norm": 0.44945289175219166,
+      "learning_rate": 1.4566094864495761e-05,
+      "loss": 0.1669,
+      "step": 1843
+    },
+    {
+      "epoch": 2.071037484206093,
+      "grad_norm": 0.4812629802641515,
+      "learning_rate": 1.4559114753740174e-05,
+      "loss": 0.1838,
+      "step": 1844
+    },
+    {
+      "epoch": 2.072160606486031,
+      "grad_norm": 0.4255497433471811,
+      "learning_rate": 1.4552131837711108e-05,
+      "loss": 0.1515,
+      "step": 1845
+    },
+    {
+      "epoch": 2.0732837287659693,
+      "grad_norm": 0.4489807287129697,
+      "learning_rate": 1.4545146120705229e-05,
+      "loss": 0.1668,
+      "step": 1846
+    },
+    {
+      "epoch": 2.0744068510459077,
+      "grad_norm": 0.4451775619304231,
+      "learning_rate": 1.4538157607020923e-05,
+      "loss": 0.1671,
+      "step": 1847
+    },
+    {
+      "epoch": 2.0755299733258457,
+      "grad_norm": 0.4736713116971495,
+      "learning_rate": 1.4531166300958303e-05,
+      "loss": 0.1715,
+      "step": 1848
+    },
+    {
+      "epoch": 2.076653095605784,
+      "grad_norm": 0.44599651030040116,
+      "learning_rate": 1.4524172206819195e-05,
+      "loss": 0.1695,
+      "step": 1849
+    },
+    {
+      "epoch": 2.0777762178857224,
+      "grad_norm": 0.4925862342950297,
+      "learning_rate": 1.4517175328907141e-05,
+      "loss": 0.177,
+      "step": 1850
+    },
+    {
+      "epoch": 2.0788993401656604,
+      "grad_norm": 0.4918233012105012,
+      "learning_rate": 1.4510175671527397e-05,
+      "loss": 0.1756,
+      "step": 1851
+    },
+    {
+      "epoch": 2.0800224624455987,
+      "grad_norm": 0.46620816928480197,
+      "learning_rate": 1.4503173238986932e-05,
+      "loss": 0.1762,
+      "step": 1852
+    },
+    {
+      "epoch": 2.081145584725537,
+      "grad_norm": 0.45469289458689577,
+      "learning_rate": 1.4496168035594418e-05,
+      "loss": 0.1696,
+      "step": 1853
+    },
+    {
+      "epoch": 2.082268707005475,
+      "grad_norm": 0.5073180847345866,
+      "learning_rate": 1.4489160065660231e-05,
+      "loss": 0.1748,
+      "step": 1854
+    },
+    {
+      "epoch": 2.0833918292854134,
+      "grad_norm": 0.49912895873161844,
+      "learning_rate": 1.4482149333496455e-05,
+      "loss": 0.1711,
+      "step": 1855
+    },
+    {
+      "epoch": 2.084514951565352,
+      "grad_norm": 0.4713637049357059,
+      "learning_rate": 1.4475135843416873e-05,
+      "loss": 0.1696,
+      "step": 1856
+    },
+    {
+      "epoch": 2.0856380738452898,
+      "grad_norm": 0.4886436819964026,
+      "learning_rate": 1.4468119599736957e-05,
+      "loss": 0.1719,
+      "step": 1857
+    },
+    {
+      "epoch": 2.086761196125228,
+      "grad_norm": 0.47437394171967345,
+      "learning_rate": 1.4461100606773884e-05,
+      "loss": 0.1583,
+      "step": 1858
+    },
+    {
+      "epoch": 2.0878843184051665,
+      "grad_norm": 0.4738862023625834,
+      "learning_rate": 1.4454078868846513e-05,
+      "loss": 0.1758,
+      "step": 1859
+    },
+    {
+      "epoch": 2.0890074406851045,
+      "grad_norm": 0.47210633789898465,
+      "learning_rate": 1.4447054390275401e-05,
+      "loss": 0.1707,
+      "step": 1860
+    },
+    {
+      "epoch": 2.090130562965043,
+      "grad_norm": 0.47219676032402647,
+      "learning_rate": 1.4440027175382784e-05,
+      "loss": 0.1538,
+      "step": 1861
+    },
+    {
+      "epoch": 2.0912536852449812,
+      "grad_norm": 0.45850818588081294,
+      "learning_rate": 1.4432997228492586e-05,
+      "loss": 0.1704,
+      "step": 1862
+    },
+    {
+      "epoch": 2.092376807524919,
+      "grad_norm": 0.445147848398028,
+      "learning_rate": 1.4425964553930412e-05,
+      "loss": 0.1633,
+      "step": 1863
+    },
+    {
+      "epoch": 2.0934999298048576,
+      "grad_norm": 0.462368673003725,
+      "learning_rate": 1.4418929156023543e-05,
+      "loss": 0.164,
+      "step": 1864
+    },
+    {
+      "epoch": 2.094623052084796,
+      "grad_norm": 0.477089375485279,
+      "learning_rate": 1.4411891039100934e-05,
+      "loss": 0.1747,
+      "step": 1865
+    },
+    {
+      "epoch": 2.095746174364734,
+      "grad_norm": 0.4633909335409037,
+      "learning_rate": 1.4404850207493217e-05,
+      "loss": 0.1706,
+      "step": 1866
+    },
+    {
+      "epoch": 2.0968692966446723,
+      "grad_norm": 0.4913608818406286,
+      "learning_rate": 1.4397806665532693e-05,
+      "loss": 0.1802,
+      "step": 1867
+    },
+    {
+      "epoch": 2.09799241892461,
+      "grad_norm": 0.4843264536525825,
+      "learning_rate": 1.4390760417553338e-05,
+      "loss": 0.1752,
+      "step": 1868
+    },
+    {
+      "epoch": 2.0991155412045486,
+      "grad_norm": 0.4786559227982498,
+      "learning_rate": 1.4383711467890776e-05,
+      "loss": 0.1733,
+      "step": 1869
+    },
+    {
+      "epoch": 2.100238663484487,
+      "grad_norm": 0.47724859281601445,
+      "learning_rate": 1.4376659820882308e-05,
+      "loss": 0.1706,
+      "step": 1870
+    },
+    {
+      "epoch": 2.101361785764425,
+      "grad_norm": 0.4781288580611537,
+      "learning_rate": 1.4369605480866888e-05,
+      "loss": 0.1682,
+      "step": 1871
+    },
+    {
+      "epoch": 2.1024849080443633,
+      "grad_norm": 0.47404027812905025,
+      "learning_rate": 1.436254845218513e-05,
+      "loss": 0.1605,
+      "step": 1872
+    },
+    {
+      "epoch": 2.1036080303243017,
+      "grad_norm": 0.4693877600646867,
+      "learning_rate": 1.4355488739179304e-05,
+      "loss": 0.1713,
+      "step": 1873
+    },
+    {
+      "epoch": 2.1047311526042396,
+      "grad_norm": 0.4465220690930916,
+      "learning_rate": 1.4348426346193325e-05,
+      "loss": 0.1616,
+      "step": 1874
+    },
+    {
+      "epoch": 2.105854274884178,
+      "grad_norm": 0.4533942365528121,
+      "learning_rate": 1.4341361277572766e-05,
+      "loss": 0.1597,
+      "step": 1875
+    },
+    {
+      "epoch": 2.1069773971641164,
+      "grad_norm": 0.4752746801317046,
+      "learning_rate": 1.4334293537664836e-05,
+      "loss": 0.1898,
+      "step": 1876
+    },
+    {
+      "epoch": 2.1081005194440543,
+      "grad_norm": 0.44524813953865283,
+      "learning_rate": 1.4327223130818393e-05,
+      "loss": 0.1629,
+      "step": 1877
+    },
+    {
+      "epoch": 2.1092236417239927,
+      "grad_norm": 0.4740454300839204,
+      "learning_rate": 1.4320150061383941e-05,
+      "loss": 0.1712,
+      "step": 1878
+    },
+    {
+      "epoch": 2.110346764003931,
+      "grad_norm": 0.46263291757398317,
+      "learning_rate": 1.4313074333713614e-05,
+      "loss": 0.172,
+      "step": 1879
+    },
+    {
+      "epoch": 2.111469886283869,
+      "grad_norm": 0.4605970497488254,
+      "learning_rate": 1.4305995952161189e-05,
+      "loss": 0.1718,
+      "step": 1880
+    },
+    {
+      "epoch": 2.1125930085638074,
+      "grad_norm": 0.47336811725490663,
+      "learning_rate": 1.4298914921082068e-05,
+      "loss": 0.1689,
+      "step": 1881
+    },
+    {
+      "epoch": 2.1137161308437458,
+      "grad_norm": 0.4671378312674368,
+      "learning_rate": 1.429183124483329e-05,
+      "loss": 0.1637,
+      "step": 1882
+    },
+    {
+      "epoch": 2.1148392531236837,
+      "grad_norm": 0.5100884705433952,
+      "learning_rate": 1.4284744927773515e-05,
+      "loss": 0.1888,
+      "step": 1883
+    },
+    {
+      "epoch": 2.115962375403622,
+      "grad_norm": 0.48122328157905797,
+      "learning_rate": 1.4277655974263035e-05,
+      "loss": 0.1717,
+      "step": 1884
+    },
+    {
+      "epoch": 2.1170854976835605,
+      "grad_norm": 0.46648639584107954,
+      "learning_rate": 1.4270564388663761e-05,
+      "loss": 0.1791,
+      "step": 1885
+    },
+    {
+      "epoch": 2.1182086199634984,
+      "grad_norm": 0.4897115936255588,
+      "learning_rate": 1.4263470175339223e-05,
+      "loss": 0.1784,
+      "step": 1886
+    },
+    {
+      "epoch": 2.119331742243437,
+      "grad_norm": 0.4695573343126349,
+      "learning_rate": 1.425637333865457e-05,
+      "loss": 0.179,
+      "step": 1887
+    },
+    {
+      "epoch": 2.120454864523375,
+      "grad_norm": 0.4472253157982482,
+      "learning_rate": 1.424927388297656e-05,
+      "loss": 0.1593,
+      "step": 1888
+    },
+    {
+      "epoch": 2.121577986803313,
+      "grad_norm": 0.4699369871519553,
+      "learning_rate": 1.4242171812673569e-05,
+      "loss": 0.1685,
+      "step": 1889
+    },
+    {
+      "epoch": 2.1227011090832515,
+      "grad_norm": 0.4762283130843815,
+      "learning_rate": 1.4235067132115581e-05,
+      "loss": 0.1714,
+      "step": 1890
+    },
+    {
+      "epoch": 2.12382423136319,
+      "grad_norm": 0.4835690093328314,
+      "learning_rate": 1.4227959845674182e-05,
+      "loss": 0.1901,
+      "step": 1891
+    },
+    {
+      "epoch": 2.124947353643128,
+      "grad_norm": 0.4787055121638117,
+      "learning_rate": 1.4220849957722562e-05,
+      "loss": 0.1807,
+      "step": 1892
+    },
+    {
+      "epoch": 2.126070475923066,
+      "grad_norm": 0.4737650279943973,
+      "learning_rate": 1.4213737472635513e-05,
+      "loss": 0.1776,
+      "step": 1893
+    },
+    {
+      "epoch": 2.1271935982030046,
+      "grad_norm": 0.4686012980137878,
+      "learning_rate": 1.4206622394789432e-05,
+      "loss": 0.1816,
+      "step": 1894
+    },
+    {
+      "epoch": 2.1283167204829425,
+      "grad_norm": 0.43340996309070035,
+      "learning_rate": 1.4199504728562294e-05,
+      "loss": 0.1585,
+      "step": 1895
+    },
+    {
+      "epoch": 2.129439842762881,
+      "grad_norm": 0.45042187104895964,
+      "learning_rate": 1.4192384478333686e-05,
+      "loss": 0.1626,
+      "step": 1896
+    },
+    {
+      "epoch": 2.130562965042819,
+      "grad_norm": 0.46788318772366466,
+      "learning_rate": 1.4185261648484772e-05,
+      "loss": 0.1641,
+      "step": 1897
+    },
+    {
+      "epoch": 2.131686087322757,
+      "grad_norm": 0.45501111518763515,
+      "learning_rate": 1.4178136243398308e-05,
+      "loss": 0.1704,
+      "step": 1898
+    },
+    {
+      "epoch": 2.1328092096026956,
+      "grad_norm": 0.458503838473636,
+      "learning_rate": 1.4171008267458636e-05,
+      "loss": 0.1649,
+      "step": 1899
+    },
+    {
+      "epoch": 2.1339323318826335,
+      "grad_norm": 0.4467967887329821,
+      "learning_rate": 1.4163877725051677e-05,
+      "loss": 0.1692,
+      "step": 1900
+    },
+    {
+      "epoch": 2.135055454162572,
+      "grad_norm": 0.46421473127298934,
+      "learning_rate": 1.4156744620564933e-05,
+      "loss": 0.1764,
+      "step": 1901
+    },
+    {
+      "epoch": 2.1361785764425103,
+      "grad_norm": 0.4604460398633165,
+      "learning_rate": 1.4149608958387484e-05,
+      "loss": 0.1638,
+      "step": 1902
+    },
+    {
+      "epoch": 2.1373016987224482,
+      "grad_norm": 0.4634460358657593,
+      "learning_rate": 1.4142470742909976e-05,
+      "loss": 0.1757,
+      "step": 1903
+    },
+    {
+      "epoch": 2.1384248210023866,
+      "grad_norm": 0.470255942011479,
+      "learning_rate": 1.4135329978524634e-05,
+      "loss": 0.1709,
+      "step": 1904
+    },
+    {
+      "epoch": 2.139547943282325,
+      "grad_norm": 0.4547022434396764,
+      "learning_rate": 1.4128186669625247e-05,
+      "loss": 0.1753,
+      "step": 1905
+    },
+    {
+      "epoch": 2.140671065562263,
+      "grad_norm": 0.4739510061930601,
+      "learning_rate": 1.4121040820607175e-05,
+      "loss": 0.1746,
+      "step": 1906
+    },
+    {
+      "epoch": 2.1417941878422013,
+      "grad_norm": 0.4708878833969285,
+      "learning_rate": 1.4113892435867337e-05,
+      "loss": 0.1739,
+      "step": 1907
+    },
+    {
+      "epoch": 2.1429173101221397,
+      "grad_norm": 0.44335049166860324,
+      "learning_rate": 1.410674151980421e-05,
+      "loss": 0.1633,
+      "step": 1908
+    },
+    {
+      "epoch": 2.1440404324020776,
+      "grad_norm": 0.45141039205299494,
+      "learning_rate": 1.4099588076817837e-05,
+      "loss": 0.1638,
+      "step": 1909
+    },
+    {
+      "epoch": 2.145163554682016,
+      "grad_norm": 0.4443166702543141,
+      "learning_rate": 1.4092432111309804e-05,
+      "loss": 0.1636,
+      "step": 1910
+    },
+    {
+      "epoch": 2.1462866769619544,
+      "grad_norm": 0.44643125071661977,
+      "learning_rate": 1.4085273627683257e-05,
+      "loss": 0.1651,
+      "step": 1911
+    },
+    {
+      "epoch": 2.1474097992418923,
+      "grad_norm": 0.48330445218374524,
+      "learning_rate": 1.4078112630342891e-05,
+      "loss": 0.1741,
+      "step": 1912
+    },
+    {
+      "epoch": 2.1485329215218307,
+      "grad_norm": 0.48560309798250084,
+      "learning_rate": 1.4070949123694945e-05,
+      "loss": 0.1882,
+      "step": 1913
+    },
+    {
+      "epoch": 2.149656043801769,
+      "grad_norm": 0.46382488718871623,
+      "learning_rate": 1.4063783112147207e-05,
+      "loss": 0.1794,
+      "step": 1914
+    },
+    {
+      "epoch": 2.150779166081707,
+      "grad_norm": 0.46801002995391355,
+      "learning_rate": 1.4056614600108998e-05,
+      "loss": 0.1691,
+      "step": 1915
+    },
+    {
+      "epoch": 2.1519022883616454,
+      "grad_norm": 0.4753045332485001,
+      "learning_rate": 1.4049443591991185e-05,
+      "loss": 0.1746,
+      "step": 1916
+    },
+    {
+      "epoch": 2.1530254106415834,
+      "grad_norm": 0.4526989009991009,
+      "learning_rate": 1.404227009220617e-05,
+      "loss": 0.173,
+      "step": 1917
+    },
+    {
+      "epoch": 2.1541485329215218,
+      "grad_norm": 0.46338849838297363,
+      "learning_rate": 1.403509410516788e-05,
+      "loss": 0.1613,
+      "step": 1918
+    },
+    {
+      "epoch": 2.15527165520146,
+      "grad_norm": 0.45937563487016014,
+      "learning_rate": 1.4027915635291786e-05,
+      "loss": 0.1774,
+      "step": 1919
+    },
+    {
+      "epoch": 2.156394777481398,
+      "grad_norm": 0.4603844415907502,
+      "learning_rate": 1.4020734686994875e-05,
+      "loss": 0.1709,
+      "step": 1920
+    },
+    {
+      "epoch": 2.1575178997613365,
+      "grad_norm": 0.47586340827626805,
+      "learning_rate": 1.4013551264695663e-05,
+      "loss": 0.1739,
+      "step": 1921
+    },
+    {
+      "epoch": 2.158641022041275,
+      "grad_norm": 0.4698577828683483,
+      "learning_rate": 1.4006365372814192e-05,
+      "loss": 0.1729,
+      "step": 1922
+    },
+    {
+      "epoch": 2.1597641443212128,
+      "grad_norm": 0.44550557016391734,
+      "learning_rate": 1.3999177015772021e-05,
+      "loss": 0.1602,
+      "step": 1923
+    },
+    {
+      "epoch": 2.160887266601151,
+      "grad_norm": 0.47621681660889276,
+      "learning_rate": 1.3991986197992223e-05,
+      "loss": 0.1775,
+      "step": 1924
+    },
+    {
+      "epoch": 2.1620103888810895,
+      "grad_norm": 0.45703981645014513,
+      "learning_rate": 1.3984792923899387e-05,
+      "loss": 0.1654,
+      "step": 1925
+    },
+    {
+      "epoch": 2.1631335111610275,
+      "grad_norm": 0.5176298847902593,
+      "learning_rate": 1.3977597197919614e-05,
+      "loss": 0.1895,
+      "step": 1926
+    },
+    {
+      "epoch": 2.164256633440966,
+      "grad_norm": 0.45890734681710893,
+      "learning_rate": 1.3970399024480512e-05,
+      "loss": 0.1733,
+      "step": 1927
+    },
+    {
+      "epoch": 2.1653797557209042,
+      "grad_norm": 0.4531704744476037,
+      "learning_rate": 1.39631984080112e-05,
+      "loss": 0.1692,
+      "step": 1928
+    },
+    {
+      "epoch": 2.166502878000842,
+      "grad_norm": 0.4631173822015671,
+      "learning_rate": 1.3955995352942296e-05,
+      "loss": 0.1727,
+      "step": 1929
+    },
+    {
+      "epoch": 2.1676260002807806,
+      "grad_norm": 0.4510321349529347,
+      "learning_rate": 1.3948789863705914e-05,
+      "loss": 0.1671,
+      "step": 1930
+    },
+    {
+      "epoch": 2.168749122560719,
+      "grad_norm": 0.449699047258487,
+      "learning_rate": 1.3941581944735675e-05,
+      "loss": 0.173,
+      "step": 1931
+    },
+    {
+      "epoch": 2.169872244840657,
+      "grad_norm": 0.4602286914189392,
+      "learning_rate": 1.3934371600466692e-05,
+      "loss": 0.1777,
+      "step": 1932
+    },
+    {
+      "epoch": 2.1709953671205953,
+      "grad_norm": 0.45736864355992796,
+      "learning_rate": 1.3927158835335567e-05,
+      "loss": 0.1796,
+      "step": 1933
+    },
+    {
+      "epoch": 2.1721184894005336,
+      "grad_norm": 0.4811354606513004,
+      "learning_rate": 1.3919943653780395e-05,
+      "loss": 0.1726,
+      "step": 1934
+    },
+    {
+      "epoch": 2.1732416116804716,
+      "grad_norm": 0.4670363828225833,
+      "learning_rate": 1.3912726060240754e-05,
+      "loss": 0.179,
+      "step": 1935
+    },
+    {
+      "epoch": 2.17436473396041,
+      "grad_norm": 0.48423978492534575,
+      "learning_rate": 1.3905506059157712e-05,
+      "loss": 0.1633,
+      "step": 1936
+    },
+    {
+      "epoch": 2.1754878562403483,
+      "grad_norm": 0.48037570474597213,
+      "learning_rate": 1.3898283654973812e-05,
+      "loss": 0.1854,
+      "step": 1937
+    },
+    {
+      "epoch": 2.1766109785202863,
+      "grad_norm": 0.458965202094496,
+      "learning_rate": 1.3891058852133083e-05,
+      "loss": 0.162,
+      "step": 1938
+    },
+    {
+      "epoch": 2.1777341008002247,
+      "grad_norm": 0.48209417167054247,
+      "learning_rate": 1.388383165508102e-05,
+      "loss": 0.1829,
+      "step": 1939
+    },
+    {
+      "epoch": 2.178857223080163,
+      "grad_norm": 0.4561793614583002,
+      "learning_rate": 1.38766020682646e-05,
+      "loss": 0.1699,
+      "step": 1940
+    },
+    {
+      "epoch": 2.179980345360101,
+      "grad_norm": 0.47026614355709756,
+      "learning_rate": 1.3869370096132269e-05,
+      "loss": 0.1817,
+      "step": 1941
+    },
+    {
+      "epoch": 2.1811034676400394,
+      "grad_norm": 0.4724639589057093,
+      "learning_rate": 1.3862135743133937e-05,
+      "loss": 0.1668,
+      "step": 1942
+    },
+    {
+      "epoch": 2.1822265899199778,
+      "grad_norm": 0.4717566607972279,
+      "learning_rate": 1.3854899013720982e-05,
+      "loss": 0.1874,
+      "step": 1943
+    },
+    {
+      "epoch": 2.1833497121999157,
+      "grad_norm": 0.47264122089233784,
+      "learning_rate": 1.384765991234624e-05,
+      "loss": 0.1785,
+      "step": 1944
+    },
+    {
+      "epoch": 2.184472834479854,
+      "grad_norm": 0.4458357059113397,
+      "learning_rate": 1.3840418443464015e-05,
+      "loss": 0.1662,
+      "step": 1945
+    },
+    {
+      "epoch": 2.1855959567597925,
+      "grad_norm": 0.4670183468434894,
+      "learning_rate": 1.383317461153006e-05,
+      "loss": 0.1689,
+      "step": 1946
+    },
+    {
+      "epoch": 2.1867190790397304,
+      "grad_norm": 0.4802944051246216,
+      "learning_rate": 1.3825928421001583e-05,
+      "loss": 0.1751,
+      "step": 1947
+    },
+    {
+      "epoch": 2.1878422013196688,
+      "grad_norm": 0.48140552348423143,
+      "learning_rate": 1.381867987633725e-05,
+      "loss": 0.1789,
+      "step": 1948
+    },
+    {
+      "epoch": 2.1889653235996067,
+      "grad_norm": 0.47327163615370027,
+      "learning_rate": 1.3811428981997159e-05,
+      "loss": 0.1742,
+      "step": 1949
+    },
+    {
+      "epoch": 2.190088445879545,
+      "grad_norm": 0.4580890578922541,
+      "learning_rate": 1.3804175742442878e-05,
+      "loss": 0.1655,
+      "step": 1950
+    },
+    {
+      "epoch": 2.1912115681594835,
+      "grad_norm": 0.47350507091879795,
+      "learning_rate": 1.3796920162137396e-05,
+      "loss": 0.1705,
+      "step": 1951
+    },
+    {
+      "epoch": 2.1923346904394214,
+      "grad_norm": 0.46014818205305075,
+      "learning_rate": 1.3789662245545158e-05,
+      "loss": 0.1685,
+      "step": 1952
+    },
+    {
+      "epoch": 2.19345781271936,
+      "grad_norm": 0.47613768985331506,
+      "learning_rate": 1.3782401997132037e-05,
+      "loss": 0.1769,
+      "step": 1953
+    },
+    {
+      "epoch": 2.194580934999298,
+      "grad_norm": 0.46754997862422304,
+      "learning_rate": 1.3775139421365342e-05,
+      "loss": 0.1766,
+      "step": 1954
+    },
+    {
+      "epoch": 2.195704057279236,
+      "grad_norm": 0.4517613096656541,
+      "learning_rate": 1.376787452271382e-05,
+      "loss": 0.1689,
+      "step": 1955
+    },
+    {
+      "epoch": 2.1968271795591745,
+      "grad_norm": 0.4808075598405213,
+      "learning_rate": 1.3760607305647637e-05,
+      "loss": 0.1859,
+      "step": 1956
+    },
+    {
+      "epoch": 2.197950301839113,
+      "grad_norm": 0.4640306895626649,
+      "learning_rate": 1.3753337774638397e-05,
+      "loss": 0.1692,
+      "step": 1957
+    },
+    {
+      "epoch": 2.199073424119051,
+      "grad_norm": 0.48015518700390997,
+      "learning_rate": 1.3746065934159123e-05,
+      "loss": 0.1809,
+      "step": 1958
+    },
+    {
+      "epoch": 2.200196546398989,
+      "grad_norm": 0.45119869946429864,
+      "learning_rate": 1.3738791788684254e-05,
+      "loss": 0.1646,
+      "step": 1959
+    },
+    {
+      "epoch": 2.2013196686789276,
+      "grad_norm": 0.46414622545090645,
+      "learning_rate": 1.3731515342689654e-05,
+      "loss": 0.1773,
+      "step": 1960
+    },
+    {
+      "epoch": 2.2024427909588655,
+      "grad_norm": 0.4431677408444893,
+      "learning_rate": 1.3724236600652598e-05,
+      "loss": 0.1594,
+      "step": 1961
+    },
+    {
+      "epoch": 2.203565913238804,
+      "grad_norm": 0.4780229496799192,
+      "learning_rate": 1.371695556705178e-05,
+      "loss": 0.1644,
+      "step": 1962
+    },
+    {
+      "epoch": 2.2046890355187423,
+      "grad_norm": 0.49218595733042086,
+      "learning_rate": 1.3709672246367299e-05,
+      "loss": 0.1762,
+      "step": 1963
+    },
+    {
+      "epoch": 2.2058121577986802,
+      "grad_norm": 0.47496532240919137,
+      "learning_rate": 1.370238664308066e-05,
+      "loss": 0.168,
+      "step": 1964
+    },
+    {
+      "epoch": 2.2069352800786186,
+      "grad_norm": 0.48178589082299556,
+      "learning_rate": 1.3695098761674779e-05,
+      "loss": 0.1811,
+      "step": 1965
+    },
+    {
+      "epoch": 2.208058402358557,
+      "grad_norm": 0.46520249661248453,
+      "learning_rate": 1.3687808606633965e-05,
+      "loss": 0.1762,
+      "step": 1966
+    },
+    {
+      "epoch": 2.209181524638495,
+      "grad_norm": 0.470425727912398,
+      "learning_rate": 1.3680516182443935e-05,
+      "loss": 0.1723,
+      "step": 1967
+    },
+    {
+      "epoch": 2.2103046469184333,
+      "grad_norm": 0.4840533890242917,
+      "learning_rate": 1.3673221493591795e-05,
+      "loss": 0.1861,
+      "step": 1968
+    },
+    {
+      "epoch": 2.2114277691983713,
+      "grad_norm": 0.4675135965762372,
+      "learning_rate": 1.3665924544566047e-05,
+      "loss": 0.1717,
+      "step": 1969
+    },
+    {
+      "epoch": 2.2125508914783096,
+      "grad_norm": 0.4556393300346666,
+      "learning_rate": 1.3658625339856586e-05,
+      "loss": 0.1725,
+      "step": 1970
+    },
+    {
+      "epoch": 2.213674013758248,
+      "grad_norm": 0.4703624670439313,
+      "learning_rate": 1.365132388395469e-05,
+      "loss": 0.1659,
+      "step": 1971
+    },
+    {
+      "epoch": 2.214797136038186,
+      "grad_norm": 0.4657783144698897,
+      "learning_rate": 1.364402018135303e-05,
+      "loss": 0.1808,
+      "step": 1972
+    },
+    {
+      "epoch": 2.2159202583181243,
+      "grad_norm": 0.462982700227875,
+      "learning_rate": 1.3636714236545649e-05,
+      "loss": 0.1728,
+      "step": 1973
+    },
+    {
+      "epoch": 2.2170433805980627,
+      "grad_norm": 0.46495072208063093,
+      "learning_rate": 1.362940605402798e-05,
+      "loss": 0.1684,
+      "step": 1974
+    },
+    {
+      "epoch": 2.2181665028780007,
+      "grad_norm": 0.4676792084934364,
+      "learning_rate": 1.3622095638296827e-05,
+      "loss": 0.1675,
+      "step": 1975
+    },
+    {
+      "epoch": 2.219289625157939,
+      "grad_norm": 0.498799056127904,
+      "learning_rate": 1.3614782993850367e-05,
+      "loss": 0.1753,
+      "step": 1976
+    },
+    {
+      "epoch": 2.2204127474378774,
+      "grad_norm": 0.47991466108630665,
+      "learning_rate": 1.3607468125188153e-05,
+      "loss": 0.1722,
+      "step": 1977
+    },
+    {
+      "epoch": 2.2215358697178154,
+      "grad_norm": 0.48676775722149834,
+      "learning_rate": 1.3600151036811101e-05,
+      "loss": 0.1686,
+      "step": 1978
+    },
+    {
+      "epoch": 2.2226589919977537,
+      "grad_norm": 0.46122036954736967,
+      "learning_rate": 1.3592831733221499e-05,
+      "loss": 0.1666,
+      "step": 1979
+    },
+    {
+      "epoch": 2.223782114277692,
+      "grad_norm": 0.4622532176427233,
+      "learning_rate": 1.3585510218922997e-05,
+      "loss": 0.1744,
+      "step": 1980
+    },
+    {
+      "epoch": 2.22490523655763,
+      "grad_norm": 0.4404281659461841,
+      "learning_rate": 1.3578186498420598e-05,
+      "loss": 0.162,
+      "step": 1981
+    },
+    {
+      "epoch": 2.2260283588375684,
+      "grad_norm": 0.4647477463673942,
+      "learning_rate": 1.357086057622067e-05,
+      "loss": 0.1687,
+      "step": 1982
+    },
+    {
+      "epoch": 2.227151481117507,
+      "grad_norm": 0.45022268319323117,
+      "learning_rate": 1.3563532456830934e-05,
+      "loss": 0.1633,
+      "step": 1983
+    },
+    {
+      "epoch": 2.2282746033974448,
+      "grad_norm": 0.4900592975701646,
+      "learning_rate": 1.3556202144760461e-05,
+      "loss": 0.1939,
+      "step": 1984
+    },
+    {
+      "epoch": 2.229397725677383,
+      "grad_norm": 0.4450602192202563,
+      "learning_rate": 1.3548869644519677e-05,
+      "loss": 0.1622,
+      "step": 1985
+    },
+    {
+      "epoch": 2.2305208479573215,
+      "grad_norm": 0.45292148908157803,
+      "learning_rate": 1.3541534960620349e-05,
+      "loss": 0.1772,
+      "step": 1986
+    },
+    {
+      "epoch": 2.2316439702372595,
+      "grad_norm": 0.4495070935381503,
+      "learning_rate": 1.3534198097575581e-05,
+      "loss": 0.1793,
+      "step": 1987
+    },
+    {
+      "epoch": 2.232767092517198,
+      "grad_norm": 0.46669693575362725,
+      "learning_rate": 1.3526859059899834e-05,
+      "loss": 0.1724,
+      "step": 1988
+    },
+    {
+      "epoch": 2.2338902147971362,
+      "grad_norm": 0.4485177100335402,
+      "learning_rate": 1.3519517852108899e-05,
+      "loss": 0.1679,
+      "step": 1989
+    },
+    {
+      "epoch": 2.235013337077074,
+      "grad_norm": 0.4765654300275507,
+      "learning_rate": 1.3512174478719896e-05,
+      "loss": 0.1741,
+      "step": 1990
+    },
+    {
+      "epoch": 2.2361364593570126,
+      "grad_norm": 0.46917420600760523,
+      "learning_rate": 1.3504828944251287e-05,
+      "loss": 0.1852,
+      "step": 1991
+    },
+    {
+      "epoch": 2.237259581636951,
+      "grad_norm": 0.4599691780330107,
+      "learning_rate": 1.349748125322286e-05,
+      "loss": 0.1678,
+      "step": 1992
+    },
+    {
+      "epoch": 2.238382703916889,
+      "grad_norm": 0.47810945057013843,
+      "learning_rate": 1.349013141015573e-05,
+      "loss": 0.1642,
+      "step": 1993
+    },
+    {
+      "epoch": 2.2395058261968273,
+      "grad_norm": 0.462210138404077,
+      "learning_rate": 1.3482779419572336e-05,
+      "loss": 0.1722,
+      "step": 1994
+    },
+    {
+      "epoch": 2.2406289484767656,
+      "grad_norm": 0.4877043211428916,
+      "learning_rate": 1.3475425285996438e-05,
+      "loss": 0.1763,
+      "step": 1995
+    },
+    {
+      "epoch": 2.2417520707567036,
+      "grad_norm": 0.4713028550897873,
+      "learning_rate": 1.3468069013953115e-05,
+      "loss": 0.1672,
+      "step": 1996
+    },
+    {
+      "epoch": 2.242875193036642,
+      "grad_norm": 0.4543701026881802,
+      "learning_rate": 1.3460710607968767e-05,
+      "loss": 0.1761,
+      "step": 1997
+    },
+    {
+      "epoch": 2.24399831531658,
+      "grad_norm": 0.46559021955454394,
+      "learning_rate": 1.3453350072571097e-05,
+      "loss": 0.1729,
+      "step": 1998
+    },
+    {
+      "epoch": 2.2451214375965183,
+      "grad_norm": 0.45021396422109927,
+      "learning_rate": 1.3445987412289126e-05,
+      "loss": 0.1757,
+      "step": 1999
+    },
+    {
+      "epoch": 2.2462445598764567,
+      "grad_norm": 0.4335318154822215,
+      "learning_rate": 1.3438622631653178e-05,
+      "loss": 0.1572,
+      "step": 2000
+    },
+    {
+      "epoch": 2.2473676821563946,
+      "grad_norm": 0.49286805287821256,
+      "learning_rate": 1.343125573519488e-05,
+      "loss": 0.1827,
+      "step": 2001
+    },
+    {
+      "epoch": 2.248490804436333,
+      "grad_norm": 0.48479831173676324,
+      "learning_rate": 1.3423886727447176e-05,
+      "loss": 0.1793,
+      "step": 2002
+    },
+    {
+      "epoch": 2.2496139267162714,
+      "grad_norm": 0.46128111314066944,
+      "learning_rate": 1.3416515612944288e-05,
+      "loss": 0.1713,
+      "step": 2003
+    },
+    {
+      "epoch": 2.2507370489962093,
+      "grad_norm": 0.47588445133183477,
+      "learning_rate": 1.3409142396221747e-05,
+      "loss": 0.1838,
+      "step": 2004
+    },
+    {
+      "epoch": 2.2518601712761477,
+      "grad_norm": 0.4630902693099745,
+      "learning_rate": 1.340176708181637e-05,
+      "loss": 0.1734,
+      "step": 2005
+    },
+    {
+      "epoch": 2.252983293556086,
+      "grad_norm": 0.4733521097328874,
+      "learning_rate": 1.3394389674266275e-05,
+      "loss": 0.1702,
+      "step": 2006
+    },
+    {
+      "epoch": 2.254106415836024,
+      "grad_norm": 0.4638411258433599,
+      "learning_rate": 1.3387010178110859e-05,
+      "loss": 0.1825,
+      "step": 2007
+    },
+    {
+      "epoch": 2.2552295381159624,
+      "grad_norm": 0.4636059997009433,
+      "learning_rate": 1.3379628597890808e-05,
+      "loss": 0.1723,
+      "step": 2008
+    },
+    {
+      "epoch": 2.2563526603959008,
+      "grad_norm": 0.4634064303623194,
+      "learning_rate": 1.337224493814809e-05,
+      "loss": 0.1668,
+      "step": 2009
+    },
+    {
+      "epoch": 2.2574757826758387,
+      "grad_norm": 0.4582783319228768,
+      "learning_rate": 1.3364859203425953e-05,
+      "loss": 0.1731,
+      "step": 2010
+    },
+    {
+      "epoch": 2.258598904955777,
+      "grad_norm": 0.5057417740276605,
+      "learning_rate": 1.335747139826892e-05,
+      "loss": 0.1802,
+      "step": 2011
+    },
+    {
+      "epoch": 2.2597220272357155,
+      "grad_norm": 0.4966593306531945,
+      "learning_rate": 1.3350081527222787e-05,
+      "loss": 0.1856,
+      "step": 2012
+    },
+    {
+      "epoch": 2.2608451495156534,
+      "grad_norm": 0.47721378699943995,
+      "learning_rate": 1.3342689594834623e-05,
+      "loss": 0.1778,
+      "step": 2013
+    },
+    {
+      "epoch": 2.261968271795592,
+      "grad_norm": 0.499924567330496,
+      "learning_rate": 1.333529560565277e-05,
+      "loss": 0.1782,
+      "step": 2014
+    },
+    {
+      "epoch": 2.2630913940755297,
+      "grad_norm": 0.4761749575886551,
+      "learning_rate": 1.3327899564226826e-05,
+      "loss": 0.1776,
+      "step": 2015
+    },
+    {
+      "epoch": 2.264214516355468,
+      "grad_norm": 0.47595190829952716,
+      "learning_rate": 1.332050147510766e-05,
+      "loss": 0.1799,
+      "step": 2016
+    },
+    {
+      "epoch": 2.2653376386354065,
+      "grad_norm": 0.4819227222937039,
+      "learning_rate": 1.3313101342847393e-05,
+      "loss": 0.1777,
+      "step": 2017
+    },
+    {
+      "epoch": 2.2664607609153444,
+      "grad_norm": 0.47028267836611976,
+      "learning_rate": 1.3305699171999409e-05,
+      "loss": 0.1783,
+      "step": 2018
+    },
+    {
+      "epoch": 2.267583883195283,
+      "grad_norm": 0.4694525543038002,
+      "learning_rate": 1.3298294967118351e-05,
+      "loss": 0.1744,
+      "step": 2019
+    },
+    {
+      "epoch": 2.268707005475221,
+      "grad_norm": 0.5046329050414854,
+      "learning_rate": 1.32908887327601e-05,
+      "loss": 0.1857,
+      "step": 2020
+    },
+    {
+      "epoch": 2.269830127755159,
+      "grad_norm": 0.46102708624099675,
+      "learning_rate": 1.3283480473481803e-05,
+      "loss": 0.1789,
+      "step": 2021
+    },
+    {
+      "epoch": 2.2709532500350975,
+      "grad_norm": 0.47172572667433565,
+      "learning_rate": 1.3276070193841833e-05,
+      "loss": 0.1731,
+      "step": 2022
+    },
+    {
+      "epoch": 2.272076372315036,
+      "grad_norm": 0.49124489515157277,
+      "learning_rate": 1.3268657898399822e-05,
+      "loss": 0.1819,
+      "step": 2023
+    },
+    {
+      "epoch": 2.273199494594974,
+      "grad_norm": 0.4720309863370387,
+      "learning_rate": 1.3261243591716634e-05,
+      "loss": 0.1686,
+      "step": 2024
+    },
+    {
+      "epoch": 2.274322616874912,
+      "grad_norm": 0.47523894040627884,
+      "learning_rate": 1.3253827278354378e-05,
+      "loss": 0.1744,
+      "step": 2025
+    },
+    {
+      "epoch": 2.2754457391548506,
+      "grad_norm": 0.4923383132196056,
+      "learning_rate": 1.3246408962876391e-05,
+      "loss": 0.1847,
+      "step": 2026
+    },
+    {
+      "epoch": 2.2765688614347885,
+      "grad_norm": 0.4588643937702198,
+      "learning_rate": 1.3238988649847243e-05,
+      "loss": 0.1699,
+      "step": 2027
+    },
+    {
+      "epoch": 2.277691983714727,
+      "grad_norm": 0.4754365804448629,
+      "learning_rate": 1.3231566343832736e-05,
+      "loss": 0.1763,
+      "step": 2028
+    },
+    {
+      "epoch": 2.2788151059946653,
+      "grad_norm": 0.4708694394595169,
+      "learning_rate": 1.3224142049399896e-05,
+      "loss": 0.1806,
+      "step": 2029
+    },
+    {
+      "epoch": 2.2799382282746032,
+      "grad_norm": 0.4532504916608009,
+      "learning_rate": 1.321671577111697e-05,
+      "loss": 0.1726,
+      "step": 2030
+    },
+    {
+      "epoch": 2.2810613505545416,
+      "grad_norm": 0.45695642633903705,
+      "learning_rate": 1.3209287513553437e-05,
+      "loss": 0.1709,
+      "step": 2031
+    },
+    {
+      "epoch": 2.28218447283448,
+      "grad_norm": 0.4436999178216967,
+      "learning_rate": 1.3201857281279978e-05,
+      "loss": 0.1662,
+      "step": 2032
+    },
+    {
+      "epoch": 2.283307595114418,
+      "grad_norm": 0.4361709759275738,
+      "learning_rate": 1.3194425078868498e-05,
+      "loss": 0.171,
+      "step": 2033
+    },
+    {
+      "epoch": 2.2844307173943563,
+      "grad_norm": 0.4685768299567099,
+      "learning_rate": 1.3186990910892115e-05,
+      "loss": 0.1845,
+      "step": 2034
+    },
+    {
+      "epoch": 2.2855538396742947,
+      "grad_norm": 0.4716954023929674,
+      "learning_rate": 1.317955478192515e-05,
+      "loss": 0.171,
+      "step": 2035
+    },
+    {
+      "epoch": 2.2866769619542326,
+      "grad_norm": 0.4508227597245992,
+      "learning_rate": 1.3172116696543142e-05,
+      "loss": 0.1686,
+      "step": 2036
+    },
+    {
+      "epoch": 2.287800084234171,
+      "grad_norm": 0.4768566069887853,
+      "learning_rate": 1.3164676659322823e-05,
+      "loss": 0.1771,
+      "step": 2037
+    },
+    {
+      "epoch": 2.2889232065141094,
+      "grad_norm": 0.4935990854907148,
+      "learning_rate": 1.315723467484213e-05,
+      "loss": 0.1806,
+      "step": 2038
+    },
+    {
+      "epoch": 2.2900463287940473,
+      "grad_norm": 0.46467316460978925,
+      "learning_rate": 1.3149790747680196e-05,
+      "loss": 0.1721,
+      "step": 2039
+    },
+    {
+      "epoch": 2.2911694510739857,
+      "grad_norm": 0.4496552803361268,
+      "learning_rate": 1.3142344882417355e-05,
+      "loss": 0.1681,
+      "step": 2040
+    },
+    {
+      "epoch": 2.292292573353924,
+      "grad_norm": 0.46285417434016424,
+      "learning_rate": 1.3134897083635126e-05,
+      "loss": 0.177,
+      "step": 2041
+    },
+    {
+      "epoch": 2.293415695633862,
+      "grad_norm": 0.45431363080930554,
+      "learning_rate": 1.3127447355916223e-05,
+      "loss": 0.182,
+      "step": 2042
+    },
+    {
+      "epoch": 2.2945388179138004,
+      "grad_norm": 0.4507422204052235,
+      "learning_rate": 1.3119995703844551e-05,
+      "loss": 0.174,
+      "step": 2043
+    },
+    {
+      "epoch": 2.295661940193739,
+      "grad_norm": 0.45097987664427835,
+      "learning_rate": 1.3112542132005182e-05,
+      "loss": 0.1726,
+      "step": 2044
+    },
+    {
+      "epoch": 2.2967850624736768,
+      "grad_norm": 0.4644989156686283,
+      "learning_rate": 1.310508664498439e-05,
+      "loss": 0.1855,
+      "step": 2045
+    },
+    {
+      "epoch": 2.297908184753615,
+      "grad_norm": 0.4417912412022252,
+      "learning_rate": 1.3097629247369613e-05,
+      "loss": 0.1663,
+      "step": 2046
+    },
+    {
+      "epoch": 2.2990313070335535,
+      "grad_norm": 0.4635140434651788,
+      "learning_rate": 1.3090169943749475e-05,
+      "loss": 0.1789,
+      "step": 2047
+    },
+    {
+      "epoch": 2.3001544293134915,
+      "grad_norm": 0.4734702085500251,
+      "learning_rate": 1.3082708738713765e-05,
+      "loss": 0.1699,
+      "step": 2048
+    },
+    {
+      "epoch": 2.30127755159343,
+      "grad_norm": 0.47850190337364856,
+      "learning_rate": 1.3075245636853444e-05,
+      "loss": 0.1715,
+      "step": 2049
+    },
+    {
+      "epoch": 2.3024006738733678,
+      "grad_norm": 0.466268874412243,
+      "learning_rate": 1.306778064276064e-05,
+      "loss": 0.1735,
+      "step": 2050
+    },
+    {
+      "epoch": 2.303523796153306,
+      "grad_norm": 0.4479379317186277,
+      "learning_rate": 1.3060313761028647e-05,
+      "loss": 0.1727,
+      "step": 2051
+    },
+    {
+      "epoch": 2.3046469184332445,
+      "grad_norm": 0.44726243257194265,
+      "learning_rate": 1.305284499625192e-05,
+      "loss": 0.1632,
+      "step": 2052
+    },
+    {
+      "epoch": 2.3057700407131825,
+      "grad_norm": 0.49557743765194856,
+      "learning_rate": 1.3045374353026073e-05,
+      "loss": 0.1819,
+      "step": 2053
+    },
+    {
+      "epoch": 2.306893162993121,
+      "grad_norm": 0.46796170967007167,
+      "learning_rate": 1.3037901835947873e-05,
+      "loss": 0.1758,
+      "step": 2054
+    },
+    {
+      "epoch": 2.3080162852730592,
+      "grad_norm": 0.4805637353304701,
+      "learning_rate": 1.3030427449615241e-05,
+      "loss": 0.1809,
+      "step": 2055
+    },
+    {
+      "epoch": 2.309139407552997,
+      "grad_norm": 0.4846347936669786,
+      "learning_rate": 1.3022951198627254e-05,
+      "loss": 0.1808,
+      "step": 2056
+    },
+    {
+      "epoch": 2.3102625298329356,
+      "grad_norm": 0.4567235403598188,
+      "learning_rate": 1.3015473087584127e-05,
+      "loss": 0.1717,
+      "step": 2057
+    },
+    {
+      "epoch": 2.311385652112874,
+      "grad_norm": 0.46361930650452676,
+      "learning_rate": 1.3007993121087226e-05,
+      "loss": 0.1724,
+      "step": 2058
+    },
+    {
+      "epoch": 2.312508774392812,
+      "grad_norm": 0.4757333926502388,
+      "learning_rate": 1.3000511303739054e-05,
+      "loss": 0.1764,
+      "step": 2059
+    },
+    {
+      "epoch": 2.3136318966727503,
+      "grad_norm": 0.4447398100189297,
+      "learning_rate": 1.299302764014326e-05,
+      "loss": 0.1643,
+      "step": 2060
+    },
+    {
+      "epoch": 2.3147550189526886,
+      "grad_norm": 0.45522814478101575,
+      "learning_rate": 1.2985542134904621e-05,
+      "loss": 0.1725,
+      "step": 2061
+    },
+    {
+      "epoch": 2.3158781412326266,
+      "grad_norm": 0.47650848146899405,
+      "learning_rate": 1.2978054792629054e-05,
+      "loss": 0.1765,
+      "step": 2062
+    },
+    {
+      "epoch": 2.317001263512565,
+      "grad_norm": 0.4501821499764709,
+      "learning_rate": 1.2970565617923598e-05,
+      "loss": 0.1751,
+      "step": 2063
+    },
+    {
+      "epoch": 2.3181243857925034,
+      "grad_norm": 0.45950502464250575,
+      "learning_rate": 1.2963074615396428e-05,
+      "loss": 0.1801,
+      "step": 2064
+    },
+    {
+      "epoch": 2.3192475080724413,
+      "grad_norm": 0.4613132581972528,
+      "learning_rate": 1.2955581789656844e-05,
+      "loss": 0.1822,
+      "step": 2065
+    },
+    {
+      "epoch": 2.3203706303523797,
+      "grad_norm": 0.4619765649138744,
+      "learning_rate": 1.2948087145315256e-05,
+      "loss": 0.1801,
+      "step": 2066
+    },
+    {
+      "epoch": 2.3214937526323176,
+      "grad_norm": 0.45361708539848494,
+      "learning_rate": 1.2940590686983208e-05,
+      "loss": 0.1696,
+      "step": 2067
+    },
+    {
+      "epoch": 2.322616874912256,
+      "grad_norm": 0.47042240961247933,
+      "learning_rate": 1.2933092419273348e-05,
+      "loss": 0.1822,
+      "step": 2068
+    },
+    {
+      "epoch": 2.3237399971921944,
+      "grad_norm": 0.44936652373893554,
+      "learning_rate": 1.2925592346799444e-05,
+      "loss": 0.1715,
+      "step": 2069
+    },
+    {
+      "epoch": 2.3248631194721323,
+      "grad_norm": 0.45944253764029824,
+      "learning_rate": 1.2918090474176378e-05,
+      "loss": 0.1728,
+      "step": 2070
+    },
+    {
+      "epoch": 2.3259862417520707,
+      "grad_norm": 0.4670499870780179,
+      "learning_rate": 1.2910586806020128e-05,
+      "loss": 0.1778,
+      "step": 2071
+    },
+    {
+      "epoch": 2.327109364032009,
+      "grad_norm": 0.47065143463451475,
+      "learning_rate": 1.2903081346947788e-05,
+      "loss": 0.1781,
+      "step": 2072
+    },
+    {
+      "epoch": 2.328232486311947,
+      "grad_norm": 0.484564209356825,
+      "learning_rate": 1.2895574101577548e-05,
+      "loss": 0.1867,
+      "step": 2073
+    },
+    {
+      "epoch": 2.3293556085918854,
+      "grad_norm": 0.4461711655494198,
+      "learning_rate": 1.28880650745287e-05,
+      "loss": 0.1749,
+      "step": 2074
+    },
+    {
+      "epoch": 2.330478730871824,
+      "grad_norm": 0.4523170048395053,
+      "learning_rate": 1.288055427042163e-05,
+      "loss": 0.1712,
+      "step": 2075
+    },
+    {
+      "epoch": 2.3316018531517617,
+      "grad_norm": 0.45907136438347945,
+      "learning_rate": 1.2873041693877817e-05,
+      "loss": 0.169,
+      "step": 2076
+    },
+    {
+      "epoch": 2.3327249754317,
+      "grad_norm": 0.4596979480607732,
+      "learning_rate": 1.2865527349519836e-05,
+      "loss": 0.1757,
+      "step": 2077
+    },
+    {
+      "epoch": 2.3338480977116385,
+      "grad_norm": 0.46814268818136334,
+      "learning_rate": 1.285801124197134e-05,
+      "loss": 0.1747,
+      "step": 2078
+    },
+    {
+      "epoch": 2.3349712199915764,
+      "grad_norm": 0.45641847730925356,
+      "learning_rate": 1.2850493375857078e-05,
+      "loss": 0.1662,
+      "step": 2079
+    },
+    {
+      "epoch": 2.336094342271515,
+      "grad_norm": 0.4573287262646522,
+      "learning_rate": 1.2842973755802872e-05,
+      "loss": 0.1666,
+      "step": 2080
+    },
+    {
+      "epoch": 2.337217464551453,
+      "grad_norm": 0.5012334972800347,
+      "learning_rate": 1.2835452386435629e-05,
+      "loss": 0.1849,
+      "step": 2081
+    },
+    {
+      "epoch": 2.338340586831391,
+      "grad_norm": 0.4826799580775581,
+      "learning_rate": 1.282792927238333e-05,
+      "loss": 0.168,
+      "step": 2082
+    },
+    {
+      "epoch": 2.3394637091113295,
+      "grad_norm": 0.4765836277272539,
+      "learning_rate": 1.282040441827503e-05,
+      "loss": 0.1885,
+      "step": 2083
+    },
+    {
+      "epoch": 2.340586831391268,
+      "grad_norm": 0.45652229568485014,
+      "learning_rate": 1.2812877828740855e-05,
+      "loss": 0.1623,
+      "step": 2084
+    },
+    {
+      "epoch": 2.341709953671206,
+      "grad_norm": 0.512094793634767,
+      "learning_rate": 1.2805349508411996e-05,
+      "loss": 0.196,
+      "step": 2085
+    },
+    {
+      "epoch": 2.342833075951144,
+      "grad_norm": 0.4764092713379784,
+      "learning_rate": 1.2797819461920714e-05,
+      "loss": 0.1817,
+      "step": 2086
+    },
+    {
+      "epoch": 2.3439561982310826,
+      "grad_norm": 0.4704742691821975,
+      "learning_rate": 1.279028769390033e-05,
+      "loss": 0.1687,
+      "step": 2087
+    },
+    {
+      "epoch": 2.3450793205110205,
+      "grad_norm": 0.4951306244097187,
+      "learning_rate": 1.2782754208985217e-05,
+      "loss": 0.1728,
+      "step": 2088
+    },
+    {
+      "epoch": 2.346202442790959,
+      "grad_norm": 0.47760873675544463,
+      "learning_rate": 1.2775219011810822e-05,
+      "loss": 0.1913,
+      "step": 2089
+    },
+    {
+      "epoch": 2.3473255650708973,
+      "grad_norm": 0.4832646096522936,
+      "learning_rate": 1.2767682107013626e-05,
+      "loss": 0.1859,
+      "step": 2090
+    },
+    {
+      "epoch": 2.3484486873508352,
+      "grad_norm": 0.4426474698768579,
+      "learning_rate": 1.2760143499231173e-05,
+      "loss": 0.162,
+      "step": 2091
+    },
+    {
+      "epoch": 2.3495718096307736,
+      "grad_norm": 0.47174290073665726,
+      "learning_rate": 1.275260319310205e-05,
+      "loss": 0.182,
+      "step": 2092
+    },
+    {
+      "epoch": 2.350694931910712,
+      "grad_norm": 0.46639059398090704,
+      "learning_rate": 1.2745061193265896e-05,
+      "loss": 0.1681,
+      "step": 2093
+    },
+    {
+      "epoch": 2.35181805419065,
+      "grad_norm": 0.4790429406794705,
+      "learning_rate": 1.2737517504363378e-05,
+      "loss": 0.1779,
+      "step": 2094
+    },
+    {
+      "epoch": 2.3529411764705883,
+      "grad_norm": 0.47495388207027717,
+      "learning_rate": 1.2729972131036212e-05,
+      "loss": 0.1805,
+      "step": 2095
+    },
+    {
+      "epoch": 2.3540642987505267,
+      "grad_norm": 0.4730593538483904,
+      "learning_rate": 1.2722425077927157e-05,
+      "loss": 0.1753,
+      "step": 2096
+    },
+    {
+      "epoch": 2.3551874210304646,
+      "grad_norm": 0.44971031148328955,
+      "learning_rate": 1.271487634967999e-05,
+      "loss": 0.1576,
+      "step": 2097
+    },
+    {
+      "epoch": 2.356310543310403,
+      "grad_norm": 0.45993224590682336,
+      "learning_rate": 1.2707325950939529e-05,
+      "loss": 0.1705,
+      "step": 2098
+    },
+    {
+      "epoch": 2.357433665590341,
+      "grad_norm": 0.4502585923789524,
+      "learning_rate": 1.2699773886351618e-05,
+      "loss": 0.1716,
+      "step": 2099
+    },
+    {
+      "epoch": 2.3585567878702793,
+      "grad_norm": 0.4639139646897428,
+      "learning_rate": 1.2692220160563125e-05,
+      "loss": 0.1761,
+      "step": 2100
+    },
+    {
+      "epoch": 2.3596799101502177,
+      "grad_norm": 0.4619505035347143,
+      "learning_rate": 1.2684664778221943e-05,
+      "loss": 0.1817,
+      "step": 2101
+    },
+    {
+      "epoch": 2.3608030324301557,
+      "grad_norm": 0.4465823236801084,
+      "learning_rate": 1.2677107743976975e-05,
+      "loss": 0.1622,
+      "step": 2102
+    },
+    {
+      "epoch": 2.361926154710094,
+      "grad_norm": 0.4679911842787963,
+      "learning_rate": 1.2669549062478155e-05,
+      "loss": 0.1893,
+      "step": 2103
+    },
+    {
+      "epoch": 2.3630492769900324,
+      "grad_norm": 0.4801272473499729,
+      "learning_rate": 1.266198873837642e-05,
+      "loss": 0.1756,
+      "step": 2104
+    },
+    {
+      "epoch": 2.3641723992699704,
+      "grad_norm": 0.4649417211528937,
+      "learning_rate": 1.2654426776323719e-05,
+      "loss": 0.1742,
+      "step": 2105
+    },
+    {
+      "epoch": 2.3652955215499087,
+      "grad_norm": 0.4612836290072487,
+      "learning_rate": 1.2646863180973012e-05,
+      "loss": 0.1726,
+      "step": 2106
+    },
+    {
+      "epoch": 2.366418643829847,
+      "grad_norm": 0.46999430959875044,
+      "learning_rate": 1.2639297956978262e-05,
+      "loss": 0.1722,
+      "step": 2107
+    },
+    {
+      "epoch": 2.367541766109785,
+      "grad_norm": 0.482099500823595,
+      "learning_rate": 1.2631731108994436e-05,
+      "loss": 0.1782,
+      "step": 2108
+    },
+    {
+      "epoch": 2.3686648883897234,
+      "grad_norm": 0.4723654096317358,
+      "learning_rate": 1.2624162641677498e-05,
+      "loss": 0.1757,
+      "step": 2109
+    },
+    {
+      "epoch": 2.369788010669662,
+      "grad_norm": 0.47531394992362225,
+      "learning_rate": 1.2616592559684408e-05,
+      "loss": 0.1862,
+      "step": 2110
+    },
+    {
+      "epoch": 2.3709111329495998,
+      "grad_norm": 0.45674297876291126,
+      "learning_rate": 1.2609020867673123e-05,
+      "loss": 0.1663,
+      "step": 2111
+    },
+    {
+      "epoch": 2.372034255229538,
+      "grad_norm": 0.458215236610029,
+      "learning_rate": 1.2601447570302585e-05,
+      "loss": 0.181,
+      "step": 2112
+    },
+    {
+      "epoch": 2.3731573775094765,
+      "grad_norm": 0.43796983940777867,
+      "learning_rate": 1.259387267223273e-05,
+      "loss": 0.1655,
+      "step": 2113
+    },
+    {
+      "epoch": 2.3742804997894145,
+      "grad_norm": 0.4363126086769737,
+      "learning_rate": 1.2586296178124475e-05,
+      "loss": 0.1621,
+      "step": 2114
+    },
+    {
+      "epoch": 2.375403622069353,
+      "grad_norm": 0.44964027723907635,
+      "learning_rate": 1.2578718092639724e-05,
+      "loss": 0.1595,
+      "step": 2115
+    },
+    {
+      "epoch": 2.376526744349291,
+      "grad_norm": 0.4599641014987599,
+      "learning_rate": 1.2571138420441349e-05,
+      "loss": 0.1626,
+      "step": 2116
+    },
+    {
+      "epoch": 2.377649866629229,
+      "grad_norm": 0.4675544983326374,
+      "learning_rate": 1.2563557166193213e-05,
+      "loss": 0.1774,
+      "step": 2117
+    },
+    {
+      "epoch": 2.3787729889091676,
+      "grad_norm": 0.4717489195434664,
+      "learning_rate": 1.2555974334560142e-05,
+      "loss": 0.1802,
+      "step": 2118
+    },
+    {
+      "epoch": 2.3798961111891055,
+      "grad_norm": 0.4739492033223087,
+      "learning_rate": 1.2548389930207932e-05,
+      "loss": 0.176,
+      "step": 2119
+    },
+    {
+      "epoch": 2.381019233469044,
+      "grad_norm": 0.46312844512973367,
+      "learning_rate": 1.2540803957803356e-05,
+      "loss": 0.1677,
+      "step": 2120
+    },
+    {
+      "epoch": 2.3821423557489823,
+      "grad_norm": 0.49417837944594656,
+      "learning_rate": 1.2533216422014145e-05,
+      "loss": 0.184,
+      "step": 2121
+    },
+    {
+      "epoch": 2.38326547802892,
+      "grad_norm": 0.484024720454384,
+      "learning_rate": 1.2525627327508994e-05,
+      "loss": 0.1734,
+      "step": 2122
+    },
+    {
+      "epoch": 2.3843886003088586,
+      "grad_norm": 0.48942672284743194,
+      "learning_rate": 1.2518036678957554e-05,
+      "loss": 0.1844,
+      "step": 2123
+    },
+    {
+      "epoch": 2.385511722588797,
+      "grad_norm": 0.46895877443085426,
+      "learning_rate": 1.2510444481030434e-05,
+      "loss": 0.1728,
+      "step": 2124
+    },
+    {
+      "epoch": 2.386634844868735,
+      "grad_norm": 0.45324010534175113,
+      "learning_rate": 1.25028507383992e-05,
+      "loss": 0.1735,
+      "step": 2125
+    },
+    {
+      "epoch": 2.3877579671486733,
+      "grad_norm": 0.4500103538917005,
+      "learning_rate": 1.2495255455736366e-05,
+      "loss": 0.1663,
+      "step": 2126
+    },
+    {
+      "epoch": 2.3888810894286117,
+      "grad_norm": 0.46447167465312617,
+      "learning_rate": 1.2487658637715388e-05,
+      "loss": 0.1851,
+      "step": 2127
+    },
+    {
+      "epoch": 2.3900042117085496,
+      "grad_norm": 0.4689612937423477,
+      "learning_rate": 1.2480060289010677e-05,
+      "loss": 0.1812,
+      "step": 2128
+    },
+    {
+      "epoch": 2.391127333988488,
+      "grad_norm": 0.44830442220436945,
+      "learning_rate": 1.2472460414297576e-05,
+      "loss": 0.1693,
+      "step": 2129
+    },
+    {
+      "epoch": 2.3922504562684264,
+      "grad_norm": 0.47648002812678625,
+      "learning_rate": 1.2464859018252377e-05,
+      "loss": 0.1897,
+      "step": 2130
+    },
+    {
+      "epoch": 2.3933735785483643,
+      "grad_norm": 0.4776674804606051,
+      "learning_rate": 1.2457256105552297e-05,
+      "loss": 0.1719,
+      "step": 2131
+    },
+    {
+      "epoch": 2.3944967008283027,
+      "grad_norm": 0.45626152225385935,
+      "learning_rate": 1.2449651680875495e-05,
+      "loss": 0.174,
+      "step": 2132
+    },
+    {
+      "epoch": 2.395619823108241,
+      "grad_norm": 0.4641074642628347,
+      "learning_rate": 1.2442045748901057e-05,
+      "loss": 0.1765,
+      "step": 2133
+    },
+    {
+      "epoch": 2.396742945388179,
+      "grad_norm": 0.45963165730817895,
+      "learning_rate": 1.2434438314308997e-05,
+      "loss": 0.175,
+      "step": 2134
+    },
+    {
+      "epoch": 2.3978660676681174,
+      "grad_norm": 0.4559745351834007,
+      "learning_rate": 1.242682938178025e-05,
+      "loss": 0.1607,
+      "step": 2135
+    },
+    {
+      "epoch": 2.3989891899480558,
+      "grad_norm": 0.4592128892747727,
+      "learning_rate": 1.2419218955996677e-05,
+      "loss": 0.1752,
+      "step": 2136
+    },
+    {
+      "epoch": 2.4001123122279937,
+      "grad_norm": 0.47314594000118837,
+      "learning_rate": 1.2411607041641062e-05,
+      "loss": 0.1715,
+      "step": 2137
+    },
+    {
+      "epoch": 2.401235434507932,
+      "grad_norm": 0.46568965161641296,
+      "learning_rate": 1.2403993643397095e-05,
+      "loss": 0.1847,
+      "step": 2138
+    },
+    {
+      "epoch": 2.4023585567878705,
+      "grad_norm": 0.45802226050759454,
+      "learning_rate": 1.2396378765949382e-05,
+      "loss": 0.1732,
+      "step": 2139
+    },
+    {
+      "epoch": 2.4034816790678084,
+      "grad_norm": 0.4657203849841537,
+      "learning_rate": 1.2388762413983447e-05,
+      "loss": 0.1756,
+      "step": 2140
+    },
+    {
+      "epoch": 2.404604801347747,
+      "grad_norm": 0.44869629379918413,
+      "learning_rate": 1.238114459218571e-05,
+      "loss": 0.1697,
+      "step": 2141
+    },
+    {
+      "epoch": 2.405727923627685,
+      "grad_norm": 0.46164059934933155,
+      "learning_rate": 1.2373525305243499e-05,
+      "loss": 0.1774,
+      "step": 2142
+    },
+    {
+      "epoch": 2.406851045907623,
+      "grad_norm": 0.47884498767707456,
+      "learning_rate": 1.2365904557845054e-05,
+      "loss": 0.1859,
+      "step": 2143
+    },
+    {
+      "epoch": 2.4079741681875615,
+      "grad_norm": 0.4576461996017312,
+      "learning_rate": 1.2358282354679494e-05,
+      "loss": 0.1684,
+      "step": 2144
+    },
+    {
+      "epoch": 2.4090972904675,
+      "grad_norm": 0.4679384692538624,
+      "learning_rate": 1.2350658700436852e-05,
+      "loss": 0.1738,
+      "step": 2145
+    },
+    {
+      "epoch": 2.410220412747438,
+      "grad_norm": 0.46313097703304307,
+      "learning_rate": 1.2343033599808044e-05,
+      "loss": 0.1772,
+      "step": 2146
+    },
+    {
+      "epoch": 2.411343535027376,
+      "grad_norm": 0.4625911631962477,
+      "learning_rate": 1.2335407057484877e-05,
+      "loss": 0.1811,
+      "step": 2147
+    },
+    {
+      "epoch": 2.4124666573073146,
+      "grad_norm": 0.44513248630919333,
+      "learning_rate": 1.232777907816005e-05,
+      "loss": 0.1749,
+      "step": 2148
+    },
+    {
+      "epoch": 2.4135897795872525,
+      "grad_norm": 0.467659286026053,
+      "learning_rate": 1.2320149666527134e-05,
+      "loss": 0.182,
+      "step": 2149
+    },
+    {
+      "epoch": 2.414712901867191,
+      "grad_norm": 0.4596330741642717,
+      "learning_rate": 1.2312518827280603e-05,
+      "loss": 0.1725,
+      "step": 2150
+    },
+    {
+      "epoch": 2.415836024147129,
+      "grad_norm": 0.48688351873304575,
+      "learning_rate": 1.2304886565115786e-05,
+      "loss": 0.1859,
+      "step": 2151
+    },
+    {
+      "epoch": 2.416959146427067,
+      "grad_norm": 0.46822427826794677,
+      "learning_rate": 1.2297252884728904e-05,
+      "loss": 0.1832,
+      "step": 2152
+    },
+    {
+      "epoch": 2.4180822687070056,
+      "grad_norm": 0.4729830338617269,
+      "learning_rate": 1.2289617790817039e-05,
+      "loss": 0.1682,
+      "step": 2153
+    },
+    {
+      "epoch": 2.4192053909869435,
+      "grad_norm": 0.46719333183624456,
+      "learning_rate": 1.228198128807815e-05,
+      "loss": 0.1741,
+      "step": 2154
+    },
+    {
+      "epoch": 2.420328513266882,
+      "grad_norm": 0.4773732399141985,
+      "learning_rate": 1.2274343381211067e-05,
+      "loss": 0.1737,
+      "step": 2155
+    },
+    {
+      "epoch": 2.4214516355468203,
+      "grad_norm": 0.48795163089354604,
+      "learning_rate": 1.226670407491547e-05,
+      "loss": 0.1875,
+      "step": 2156
+    },
+    {
+      "epoch": 2.4225747578267582,
+      "grad_norm": 0.4579527625487055,
+      "learning_rate": 1.2259063373891911e-05,
+      "loss": 0.1772,
+      "step": 2157
+    },
+    {
+      "epoch": 2.4236978801066966,
+      "grad_norm": 0.4533126876310765,
+      "learning_rate": 1.22514212828418e-05,
+      "loss": 0.1705,
+      "step": 2158
+    },
+    {
+      "epoch": 2.424821002386635,
+      "grad_norm": 0.43848733476677665,
+      "learning_rate": 1.2243777806467396e-05,
+      "loss": 0.1676,
+      "step": 2159
+    },
+    {
+      "epoch": 2.425944124666573,
+      "grad_norm": 0.4719399363071484,
+      "learning_rate": 1.223613294947182e-05,
+      "loss": 0.1811,
+      "step": 2160
+    },
+    {
+      "epoch": 2.4270672469465113,
+      "grad_norm": 0.453477191494824,
+      "learning_rate": 1.222848671655903e-05,
+      "loss": 0.1685,
+      "step": 2161
+    },
+    {
+      "epoch": 2.4281903692264497,
+      "grad_norm": 0.4658487988412452,
+      "learning_rate": 1.222083911243384e-05,
+      "loss": 0.1773,
+      "step": 2162
+    },
+    {
+      "epoch": 2.4293134915063876,
+      "grad_norm": 0.459422798554054,
+      "learning_rate": 1.2213190141801906e-05,
+      "loss": 0.1733,
+      "step": 2163
+    },
+    {
+      "epoch": 2.430436613786326,
+      "grad_norm": 0.4887095631557401,
+      "learning_rate": 1.2205539809369719e-05,
+      "loss": 0.191,
+      "step": 2164
+    },
+    {
+      "epoch": 2.431559736066264,
+      "grad_norm": 0.48937602650227535,
+      "learning_rate": 1.2197888119844623e-05,
+      "loss": 0.1846,
+      "step": 2165
+    },
+    {
+      "epoch": 2.4326828583462023,
+      "grad_norm": 0.46406522024065605,
+      "learning_rate": 1.2190235077934776e-05,
+      "loss": 0.1747,
+      "step": 2166
+    },
+    {
+      "epoch": 2.4338059806261407,
+      "grad_norm": 0.4621944971046579,
+      "learning_rate": 1.2182580688349185e-05,
+      "loss": 0.1685,
+      "step": 2167
+    },
+    {
+      "epoch": 2.4349291029060787,
+      "grad_norm": 0.4862208277682074,
+      "learning_rate": 1.2174924955797676e-05,
+      "loss": 0.1838,
+      "step": 2168
+    },
+    {
+      "epoch": 2.436052225186017,
+      "grad_norm": 0.4488446985907642,
+      "learning_rate": 1.216726788499091e-05,
+      "loss": 0.1735,
+      "step": 2169
+    },
+    {
+      "epoch": 2.4371753474659554,
+      "grad_norm": 0.45799186763818456,
+      "learning_rate": 1.2159609480640361e-05,
+      "loss": 0.1615,
+      "step": 2170
+    },
+    {
+      "epoch": 2.4382984697458934,
+      "grad_norm": 0.4700112723112886,
+      "learning_rate": 1.2151949747458336e-05,
+      "loss": 0.1729,
+      "step": 2171
+    },
+    {
+      "epoch": 2.4394215920258318,
+      "grad_norm": 0.4613922790363946,
+      "learning_rate": 1.214428869015795e-05,
+      "loss": 0.1612,
+      "step": 2172
+    },
+    {
+      "epoch": 2.44054471430577,
+      "grad_norm": 0.4730141720396465,
+      "learning_rate": 1.2136626313453136e-05,
+      "loss": 0.1717,
+      "step": 2173
+    },
+    {
+      "epoch": 2.441667836585708,
+      "grad_norm": 0.48115765163733054,
+      "learning_rate": 1.212896262205864e-05,
+      "loss": 0.1759,
+      "step": 2174
+    },
+    {
+      "epoch": 2.4427909588656465,
+      "grad_norm": 0.4773927024472624,
+      "learning_rate": 1.2121297620690011e-05,
+      "loss": 0.1703,
+      "step": 2175
+    },
+    {
+      "epoch": 2.443914081145585,
+      "grad_norm": 0.42953251161909245,
+      "learning_rate": 1.2113631314063615e-05,
+      "loss": 0.1625,
+      "step": 2176
+    },
+    {
+      "epoch": 2.4450372034255228,
+      "grad_norm": 0.45411136690423287,
+      "learning_rate": 1.210596370689661e-05,
+      "loss": 0.1695,
+      "step": 2177
+    },
+    {
+      "epoch": 2.446160325705461,
+      "grad_norm": 0.48313926964400616,
+      "learning_rate": 1.2098294803906962e-05,
+      "loss": 0.1831,
+      "step": 2178
+    },
+    {
+      "epoch": 2.4472834479853995,
+      "grad_norm": 0.4816531131405757,
+      "learning_rate": 1.209062460981343e-05,
+      "loss": 0.1758,
+      "step": 2179
+    },
+    {
+      "epoch": 2.4484065702653375,
+      "grad_norm": 0.4399375257912147,
+      "learning_rate": 1.208295312933557e-05,
+      "loss": 0.1685,
+      "step": 2180
+    },
+    {
+      "epoch": 2.449529692545276,
+      "grad_norm": 0.467400595438487,
+      "learning_rate": 1.2075280367193727e-05,
+      "loss": 0.1851,
+      "step": 2181
+    },
+    {
+      "epoch": 2.4506528148252142,
+      "grad_norm": 0.4658665632807449,
+      "learning_rate": 1.2067606328109038e-05,
+      "loss": 0.1645,
+      "step": 2182
+    },
+    {
+      "epoch": 2.451775937105152,
+      "grad_norm": 0.46990058745256763,
+      "learning_rate": 1.2059931016803422e-05,
+      "loss": 0.168,
+      "step": 2183
+    },
+    {
+      "epoch": 2.4528990593850906,
+      "grad_norm": 0.4614529742174627,
+      "learning_rate": 1.2052254437999582e-05,
+      "loss": 0.1686,
+      "step": 2184
+    },
+    {
+      "epoch": 2.454022181665029,
+      "grad_norm": 0.46087967841279076,
+      "learning_rate": 1.2044576596421003e-05,
+      "loss": 0.182,
+      "step": 2185
+    },
+    {
+      "epoch": 2.455145303944967,
+      "grad_norm": 0.4686611799132139,
+      "learning_rate": 1.2036897496791945e-05,
+      "loss": 0.1705,
+      "step": 2186
+    },
+    {
+      "epoch": 2.4562684262249053,
+      "grad_norm": 0.45135028559868196,
+      "learning_rate": 1.2029217143837441e-05,
+      "loss": 0.1749,
+      "step": 2187
+    },
+    {
+      "epoch": 2.4573915485048436,
+      "grad_norm": 0.45774486543179027,
+      "learning_rate": 1.2021535542283297e-05,
+      "loss": 0.1786,
+      "step": 2188
+    },
+    {
+      "epoch": 2.4585146707847816,
+      "grad_norm": 0.4370453888107022,
+      "learning_rate": 1.2013852696856092e-05,
+      "loss": 0.1616,
+      "step": 2189
+    },
+    {
+      "epoch": 2.45963779306472,
+      "grad_norm": 0.4941697271173176,
+      "learning_rate": 1.2006168612283158e-05,
+      "loss": 0.1888,
+      "step": 2190
+    },
+    {
+      "epoch": 2.4607609153446584,
+      "grad_norm": 0.45220984927599905,
+      "learning_rate": 1.1998483293292602e-05,
+      "loss": 0.1719,
+      "step": 2191
+    },
+    {
+      "epoch": 2.4618840376245963,
+      "grad_norm": 0.45516650401879916,
+      "learning_rate": 1.199079674461328e-05,
+      "loss": 0.1718,
+      "step": 2192
+    },
+    {
+      "epoch": 2.4630071599045347,
+      "grad_norm": 0.47742118331301076,
+      "learning_rate": 1.1983108970974815e-05,
+      "loss": 0.1848,
+      "step": 2193
+    },
+    {
+      "epoch": 2.464130282184473,
+      "grad_norm": 0.46090370293756416,
+      "learning_rate": 1.1975419977107578e-05,
+      "loss": 0.1672,
+      "step": 2194
+    },
+    {
+      "epoch": 2.465253404464411,
+      "grad_norm": 0.46359443330017996,
+      "learning_rate": 1.1967729767742688e-05,
+      "loss": 0.1769,
+      "step": 2195
+    },
+    {
+      "epoch": 2.4663765267443494,
+      "grad_norm": 0.5022583283184264,
+      "learning_rate": 1.1960038347612021e-05,
+      "loss": 0.181,
+      "step": 2196
+    },
+    {
+      "epoch": 2.4674996490242878,
+      "grad_norm": 0.46125009756094304,
+      "learning_rate": 1.1952345721448189e-05,
+      "loss": 0.1784,
+      "step": 2197
+    },
+    {
+      "epoch": 2.4686227713042257,
+      "grad_norm": 0.4588215867147784,
+      "learning_rate": 1.1944651893984546e-05,
+      "loss": 0.1676,
+      "step": 2198
+    },
+    {
+      "epoch": 2.469745893584164,
+      "grad_norm": 0.44150474530630224,
+      "learning_rate": 1.1936956869955198e-05,
+      "loss": 0.1581,
+      "step": 2199
+    },
+    {
+      "epoch": 2.470869015864102,
+      "grad_norm": 0.4648006055553846,
+      "learning_rate": 1.192926065409497e-05,
+      "loss": 0.1756,
+      "step": 2200
+    },
+    {
+      "epoch": 2.4719921381440404,
+      "grad_norm": 0.4783515973358098,
+      "learning_rate": 1.1921563251139433e-05,
+      "loss": 0.177,
+      "step": 2201
+    },
+    {
+      "epoch": 2.473115260423979,
+      "grad_norm": 0.46380023099280016,
+      "learning_rate": 1.1913864665824878e-05,
+      "loss": 0.1738,
+      "step": 2202
+    },
+    {
+      "epoch": 2.4742383827039167,
+      "grad_norm": 0.4494157069182162,
+      "learning_rate": 1.1906164902888336e-05,
+      "loss": 0.1669,
+      "step": 2203
+    },
+    {
+      "epoch": 2.475361504983855,
+      "grad_norm": 0.45612118971443005,
+      "learning_rate": 1.189846396706755e-05,
+      "loss": 0.178,
+      "step": 2204
+    },
+    {
+      "epoch": 2.4764846272637935,
+      "grad_norm": 0.45627124504496713,
+      "learning_rate": 1.1890761863100994e-05,
+      "loss": 0.1838,
+      "step": 2205
+    },
+    {
+      "epoch": 2.4776077495437314,
+      "grad_norm": 0.448762160212113,
+      "learning_rate": 1.1883058595727862e-05,
+      "loss": 0.1715,
+      "step": 2206
+    },
+    {
+      "epoch": 2.47873087182367,
+      "grad_norm": 0.4478139255344506,
+      "learning_rate": 1.1875354169688049e-05,
+      "loss": 0.178,
+      "step": 2207
+    },
+    {
+      "epoch": 2.479853994103608,
+      "grad_norm": 0.45613528585444935,
+      "learning_rate": 1.186764858972218e-05,
+      "loss": 0.17,
+      "step": 2208
+    },
+    {
+      "epoch": 2.480977116383546,
+      "grad_norm": 0.4788049854189734,
+      "learning_rate": 1.185994186057158e-05,
+      "loss": 0.1785,
+      "step": 2209
+    },
+    {
+      "epoch": 2.4821002386634845,
+      "grad_norm": 0.4449516828341209,
+      "learning_rate": 1.1852233986978286e-05,
+      "loss": 0.1638,
+      "step": 2210
+    },
+    {
+      "epoch": 2.483223360943423,
+      "grad_norm": 0.4533109754634573,
+      "learning_rate": 1.1844524973685036e-05,
+      "loss": 0.17,
+      "step": 2211
+    },
+    {
+      "epoch": 2.484346483223361,
+      "grad_norm": 0.4615156080814565,
+      "learning_rate": 1.1836814825435272e-05,
+      "loss": 0.1753,
+      "step": 2212
+    },
+    {
+      "epoch": 2.485469605503299,
+      "grad_norm": 0.4617037737156908,
+      "learning_rate": 1.1829103546973135e-05,
+      "loss": 0.1752,
+      "step": 2213
+    },
+    {
+      "epoch": 2.4865927277832376,
+      "grad_norm": 0.4800696915193034,
+      "learning_rate": 1.1821391143043455e-05,
+      "loss": 0.1821,
+      "step": 2214
+    },
+    {
+      "epoch": 2.4877158500631755,
+      "grad_norm": 0.48034446713417756,
+      "learning_rate": 1.1813677618391759e-05,
+      "loss": 0.177,
+      "step": 2215
+    },
+    {
+      "epoch": 2.488838972343114,
+      "grad_norm": 0.46921558168312416,
+      "learning_rate": 1.1805962977764271e-05,
+      "loss": 0.1676,
+      "step": 2216
+    },
+    {
+      "epoch": 2.489962094623052,
+      "grad_norm": 0.45754664779814874,
+      "learning_rate": 1.1798247225907883e-05,
+      "loss": 0.1659,
+      "step": 2217
+    },
+    {
+      "epoch": 2.4910852169029902,
+      "grad_norm": 0.4445261764071967,
+      "learning_rate": 1.1790530367570194e-05,
+      "loss": 0.1697,
+      "step": 2218
+    },
+    {
+      "epoch": 2.4922083391829286,
+      "grad_norm": 0.44402068318245963,
+      "learning_rate": 1.1782812407499461e-05,
+      "loss": 0.1706,
+      "step": 2219
+    },
+    {
+      "epoch": 2.4933314614628665,
+      "grad_norm": 0.4853740651601103,
+      "learning_rate": 1.1775093350444638e-05,
+      "loss": 0.1787,
+      "step": 2220
+    },
+    {
+      "epoch": 2.494454583742805,
+      "grad_norm": 0.47514363972154094,
+      "learning_rate": 1.1767373201155344e-05,
+      "loss": 0.1767,
+      "step": 2221
+    },
+    {
+      "epoch": 2.4955777060227433,
+      "grad_norm": 0.4573804651093508,
+      "learning_rate": 1.1759651964381864e-05,
+      "loss": 0.1643,
+      "step": 2222
+    },
+    {
+      "epoch": 2.4967008283026813,
+      "grad_norm": 0.46823456746959147,
+      "learning_rate": 1.1751929644875171e-05,
+      "loss": 0.1801,
+      "step": 2223
+    },
+    {
+      "epoch": 2.4978239505826196,
+      "grad_norm": 0.4702347722306607,
+      "learning_rate": 1.1744206247386885e-05,
+      "loss": 0.1756,
+      "step": 2224
+    },
+    {
+      "epoch": 2.498947072862558,
+      "grad_norm": 0.4779388858348045,
+      "learning_rate": 1.1736481776669307e-05,
+      "loss": 0.1854,
+      "step": 2225
+    },
+    {
+      "epoch": 2.500070195142496,
+      "grad_norm": 0.45908097833882,
+      "learning_rate": 1.1728756237475377e-05,
+      "loss": 0.1724,
+      "step": 2226
+    },
+    {
+      "epoch": 2.5011933174224343,
+      "grad_norm": 0.47347497107975,
+      "learning_rate": 1.172102963455871e-05,
+      "loss": 0.1856,
+      "step": 2227
+    },
+    {
+      "epoch": 2.5023164397023727,
+      "grad_norm": 0.440317578484131,
+      "learning_rate": 1.1713301972673574e-05,
+      "loss": 0.1677,
+      "step": 2228
+    },
+    {
+      "epoch": 2.5034395619823107,
+      "grad_norm": 0.4698698238973603,
+      "learning_rate": 1.1705573256574875e-05,
+      "loss": 0.172,
+      "step": 2229
+    },
+    {
+      "epoch": 2.504562684262249,
+      "grad_norm": 0.4666802074158573,
+      "learning_rate": 1.1697843491018189e-05,
+      "loss": 0.1778,
+      "step": 2230
+    },
+    {
+      "epoch": 2.5056858065421874,
+      "grad_norm": 0.4675479550389423,
+      "learning_rate": 1.1690112680759714e-05,
+      "loss": 0.1765,
+      "step": 2231
+    },
+    {
+      "epoch": 2.5068089288221254,
+      "grad_norm": 0.470321624448753,
+      "learning_rate": 1.1682380830556305e-05,
+      "loss": 0.1741,
+      "step": 2232
+    },
+    {
+      "epoch": 2.5079320511020637,
+      "grad_norm": 0.4621451552720319,
+      "learning_rate": 1.1674647945165463e-05,
+      "loss": 0.1725,
+      "step": 2233
+    },
+    {
+      "epoch": 2.509055173382002,
+      "grad_norm": 0.47835096191087073,
+      "learning_rate": 1.1666914029345309e-05,
+      "loss": 0.1735,
+      "step": 2234
+    },
+    {
+      "epoch": 2.51017829566194,
+      "grad_norm": 0.44571066889134053,
+      "learning_rate": 1.165917908785461e-05,
+      "loss": 0.1658,
+      "step": 2235
+    },
+    {
+      "epoch": 2.5113014179418784,
+      "grad_norm": 0.48065746817643185,
+      "learning_rate": 1.165144312545276e-05,
+      "loss": 0.1789,
+      "step": 2236
+    },
+    {
+      "epoch": 2.512424540221817,
+      "grad_norm": 0.4865730345149686,
+      "learning_rate": 1.164370614689978e-05,
+      "loss": 0.19,
+      "step": 2237
+    },
+    {
+      "epoch": 2.5135476625017548,
+      "grad_norm": 0.46563389532044447,
+      "learning_rate": 1.1635968156956322e-05,
+      "loss": 0.1788,
+      "step": 2238
+    },
+    {
+      "epoch": 2.514670784781693,
+      "grad_norm": 0.44428883561842164,
+      "learning_rate": 1.1628229160383653e-05,
+      "loss": 0.1645,
+      "step": 2239
+    },
+    {
+      "epoch": 2.5157939070616315,
+      "grad_norm": 0.4699429571414019,
+      "learning_rate": 1.1620489161943665e-05,
+      "loss": 0.1734,
+      "step": 2240
+    },
+    {
+      "epoch": 2.5169170293415695,
+      "grad_norm": 0.4645753328504612,
+      "learning_rate": 1.161274816639886e-05,
+      "loss": 0.1695,
+      "step": 2241
+    },
+    {
+      "epoch": 2.518040151621508,
+      "grad_norm": 0.4715174058816328,
+      "learning_rate": 1.1605006178512361e-05,
+      "loss": 0.1767,
+      "step": 2242
+    },
+    {
+      "epoch": 2.5191632739014462,
+      "grad_norm": 0.4739964569319054,
+      "learning_rate": 1.15972632030479e-05,
+      "loss": 0.1778,
+      "step": 2243
+    },
+    {
+      "epoch": 2.520286396181384,
+      "grad_norm": 0.4864994655693669,
+      "learning_rate": 1.1589519244769813e-05,
+      "loss": 0.1907,
+      "step": 2244
+    },
+    {
+      "epoch": 2.5214095184613226,
+      "grad_norm": 0.4554291488338845,
+      "learning_rate": 1.1581774308443042e-05,
+      "loss": 0.1726,
+      "step": 2245
+    },
+    {
+      "epoch": 2.522532640741261,
+      "grad_norm": 0.4490036340445551,
+      "learning_rate": 1.157402839883313e-05,
+      "loss": 0.1737,
+      "step": 2246
+    },
+    {
+      "epoch": 2.523655763021199,
+      "grad_norm": 0.4560497521668537,
+      "learning_rate": 1.1566281520706228e-05,
+      "loss": 0.1698,
+      "step": 2247
+    },
+    {
+      "epoch": 2.5247788853011373,
+      "grad_norm": 0.4848150687683439,
+      "learning_rate": 1.1558533678829066e-05,
+      "loss": 0.1779,
+      "step": 2248
+    },
+    {
+      "epoch": 2.5259020075810756,
+      "grad_norm": 0.4557588727056414,
+      "learning_rate": 1.1550784877968982e-05,
+      "loss": 0.1792,
+      "step": 2249
+    },
+    {
+      "epoch": 2.5270251298610136,
+      "grad_norm": 0.46640445243584955,
+      "learning_rate": 1.1543035122893898e-05,
+      "loss": 0.1821,
+      "step": 2250
+    },
+    {
+      "epoch": 2.528148252140952,
+      "grad_norm": 0.46062983958648196,
+      "learning_rate": 1.1535284418372321e-05,
+      "loss": 0.1731,
+      "step": 2251
+    },
+    {
+      "epoch": 2.5292713744208903,
+      "grad_norm": 0.4533115380906245,
+      "learning_rate": 1.1527532769173349e-05,
+      "loss": 0.1852,
+      "step": 2252
+    },
+    {
+      "epoch": 2.5303944967008283,
+      "grad_norm": 0.45093605161163985,
+      "learning_rate": 1.1519780180066651e-05,
+      "loss": 0.1725,
+      "step": 2253
+    },
+    {
+      "epoch": 2.5315176189807667,
+      "grad_norm": 0.4739553972885986,
+      "learning_rate": 1.1512026655822483e-05,
+      "loss": 0.185,
+      "step": 2254
+    },
+    {
+      "epoch": 2.5326407412607046,
+      "grad_norm": 0.455497428773056,
+      "learning_rate": 1.150427220121168e-05,
+      "loss": 0.1667,
+      "step": 2255
+    },
+    {
+      "epoch": 2.533763863540643,
+      "grad_norm": 0.46427594398131145,
+      "learning_rate": 1.1496516821005632e-05,
+      "loss": 0.1707,
+      "step": 2256
+    },
+    {
+      "epoch": 2.5348869858205814,
+      "grad_norm": 0.4726499893846905,
+      "learning_rate": 1.1488760519976321e-05,
+      "loss": 0.1693,
+      "step": 2257
+    },
+    {
+      "epoch": 2.5360101081005193,
+      "grad_norm": 0.4770180942704747,
+      "learning_rate": 1.1481003302896274e-05,
+      "loss": 0.1791,
+      "step": 2258
+    },
+    {
+      "epoch": 2.5371332303804577,
+      "grad_norm": 0.4824404668774312,
+      "learning_rate": 1.1473245174538601e-05,
+      "loss": 0.1632,
+      "step": 2259
+    },
+    {
+      "epoch": 2.538256352660396,
+      "grad_norm": 0.46093177693875936,
+      "learning_rate": 1.1465486139676955e-05,
+      "loss": 0.171,
+      "step": 2260
+    },
+    {
+      "epoch": 2.539379474940334,
+      "grad_norm": 0.47811684922648134,
+      "learning_rate": 1.1457726203085565e-05,
+      "loss": 0.1795,
+      "step": 2261
+    },
+    {
+      "epoch": 2.5405025972202724,
+      "grad_norm": 0.4605303787580689,
+      "learning_rate": 1.14499653695392e-05,
+      "loss": 0.1717,
+      "step": 2262
+    },
+    {
+      "epoch": 2.5416257195002103,
+      "grad_norm": 0.45780123414736634,
+      "learning_rate": 1.1442203643813184e-05,
+      "loss": 0.165,
+      "step": 2263
+    },
+    {
+      "epoch": 2.5427488417801487,
+      "grad_norm": 0.4484627901652153,
+      "learning_rate": 1.1434441030683396e-05,
+      "loss": 0.1723,
+      "step": 2264
+    },
+    {
+      "epoch": 2.543871964060087,
+      "grad_norm": 0.45899357848479316,
+      "learning_rate": 1.1426677534926259e-05,
+      "loss": 0.178,
+      "step": 2265
+    },
+    {
+      "epoch": 2.544995086340025,
+      "grad_norm": 0.4520762026871274,
+      "learning_rate": 1.1418913161318735e-05,
+      "loss": 0.1718,
+      "step": 2266
+    },
+    {
+      "epoch": 2.5461182086199634,
+      "grad_norm": 0.48593149453254475,
+      "learning_rate": 1.1411147914638323e-05,
+      "loss": 0.1858,
+      "step": 2267
+    },
+    {
+      "epoch": 2.547241330899902,
+      "grad_norm": 0.4699419541357399,
+      "learning_rate": 1.1403381799663073e-05,
+      "loss": 0.1868,
+      "step": 2268
+    },
+    {
+      "epoch": 2.5483644531798397,
+      "grad_norm": 0.43072277665816994,
+      "learning_rate": 1.139561482117156e-05,
+      "loss": 0.1635,
+      "step": 2269
+    },
+    {
+      "epoch": 2.549487575459778,
+      "grad_norm": 0.48228875953990946,
+      "learning_rate": 1.138784698394289e-05,
+      "loss": 0.1838,
+      "step": 2270
+    },
+    {
+      "epoch": 2.5506106977397165,
+      "grad_norm": 0.4589878064473334,
+      "learning_rate": 1.1380078292756695e-05,
+      "loss": 0.1779,
+      "step": 2271
+    },
+    {
+      "epoch": 2.5517338200196544,
+      "grad_norm": 0.47422067140567364,
+      "learning_rate": 1.1372308752393144e-05,
+      "loss": 0.1871,
+      "step": 2272
+    },
+    {
+      "epoch": 2.552856942299593,
+      "grad_norm": 0.4841535195365073,
+      "learning_rate": 1.136453836763291e-05,
+      "loss": 0.1839,
+      "step": 2273
+    },
+    {
+      "epoch": 2.553980064579531,
+      "grad_norm": 0.46931683097362253,
+      "learning_rate": 1.1356767143257208e-05,
+      "loss": 0.1774,
+      "step": 2274
+    },
+    {
+      "epoch": 2.555103186859469,
+      "grad_norm": 0.45852223438824385,
+      "learning_rate": 1.134899508404775e-05,
+      "loss": 0.1668,
+      "step": 2275
+    },
+    {
+      "epoch": 2.5562263091394075,
+      "grad_norm": 0.4747264084612995,
+      "learning_rate": 1.1341222194786772e-05,
+      "loss": 0.1817,
+      "step": 2276
+    },
+    {
+      "epoch": 2.557349431419346,
+      "grad_norm": 0.5309294296915062,
+      "learning_rate": 1.1333448480257019e-05,
+      "loss": 0.1774,
+      "step": 2277
+    },
+    {
+      "epoch": 2.558472553699284,
+      "grad_norm": 0.44494999424664955,
+      "learning_rate": 1.132567394524174e-05,
+      "loss": 0.1652,
+      "step": 2278
+    },
+    {
+      "epoch": 2.559595675979222,
+      "grad_norm": 0.47565188072125314,
+      "learning_rate": 1.1317898594524694e-05,
+      "loss": 0.1739,
+      "step": 2279
+    },
+    {
+      "epoch": 2.5607187982591606,
+      "grad_norm": 0.4724833833160546,
+      "learning_rate": 1.131012243289014e-05,
+      "loss": 0.1756,
+      "step": 2280
+    },
+    {
+      "epoch": 2.5618419205390985,
+      "grad_norm": 0.4961383451401611,
+      "learning_rate": 1.1302345465122839e-05,
+      "loss": 0.1751,
+      "step": 2281
+    },
+    {
+      "epoch": 2.562965042819037,
+      "grad_norm": 0.45743388018733416,
+      "learning_rate": 1.1294567696008038e-05,
+      "loss": 0.1682,
+      "step": 2282
+    },
+    {
+      "epoch": 2.5640881650989753,
+      "grad_norm": 0.46320800217506247,
+      "learning_rate": 1.1286789130331487e-05,
+      "loss": 0.1765,
+      "step": 2283
+    },
+    {
+      "epoch": 2.5652112873789132,
+      "grad_norm": 0.46728239868320987,
+      "learning_rate": 1.1279009772879427e-05,
+      "loss": 0.1715,
+      "step": 2284
+    },
+    {
+      "epoch": 2.5663344096588516,
+      "grad_norm": 0.455957997479803,
+      "learning_rate": 1.1271229628438578e-05,
+      "loss": 0.174,
+      "step": 2285
+    },
+    {
+      "epoch": 2.56745753193879,
+      "grad_norm": 0.46937629962755417,
+      "learning_rate": 1.1263448701796149e-05,
+      "loss": 0.1785,
+      "step": 2286
+    },
+    {
+      "epoch": 2.568580654218728,
+      "grad_norm": 0.46560086475544704,
+      "learning_rate": 1.125566699773983e-05,
+      "loss": 0.1782,
+      "step": 2287
+    },
+    {
+      "epoch": 2.5697037764986663,
+      "grad_norm": 0.4611166282806535,
+      "learning_rate": 1.1247884521057788e-05,
+      "loss": 0.1604,
+      "step": 2288
+    },
+    {
+      "epoch": 2.5708268987786047,
+      "grad_norm": 0.4800102109761718,
+      "learning_rate": 1.1240101276538668e-05,
+      "loss": 0.1798,
+      "step": 2289
+    },
+    {
+      "epoch": 2.5719500210585426,
+      "grad_norm": 0.44962924941537197,
+      "learning_rate": 1.1232317268971586e-05,
+      "loss": 0.1645,
+      "step": 2290
+    },
+    {
+      "epoch": 2.573073143338481,
+      "grad_norm": 0.45604502325176743,
+      "learning_rate": 1.122453250314613e-05,
+      "loss": 0.1762,
+      "step": 2291
+    },
+    {
+      "epoch": 2.5741962656184194,
+      "grad_norm": 0.44941318778717404,
+      "learning_rate": 1.121674698385235e-05,
+      "loss": 0.1723,
+      "step": 2292
+    },
+    {
+      "epoch": 2.5753193878983573,
+      "grad_norm": 0.4570440838697352,
+      "learning_rate": 1.1208960715880759e-05,
+      "loss": 0.1692,
+      "step": 2293
+    },
+    {
+      "epoch": 2.5764425101782957,
+      "grad_norm": 0.4458264791275645,
+      "learning_rate": 1.1201173704022335e-05,
+      "loss": 0.1714,
+      "step": 2294
+    },
+    {
+      "epoch": 2.577565632458234,
+      "grad_norm": 0.46409333426984856,
+      "learning_rate": 1.1193385953068512e-05,
+      "loss": 0.188,
+      "step": 2295
+    },
+    {
+      "epoch": 2.578688754738172,
+      "grad_norm": 0.42425996455837706,
+      "learning_rate": 1.118559746781118e-05,
+      "loss": 0.1591,
+      "step": 2296
+    },
+    {
+      "epoch": 2.5798118770181104,
+      "grad_norm": 0.447955738326864,
+      "learning_rate": 1.1177808253042679e-05,
+      "loss": 0.1712,
+      "step": 2297
+    },
+    {
+      "epoch": 2.580934999298049,
+      "grad_norm": 0.4650895841202455,
+      "learning_rate": 1.1170018313555802e-05,
+      "loss": 0.1808,
+      "step": 2298
+    },
+    {
+      "epoch": 2.5820581215779868,
+      "grad_norm": 0.4701322334085429,
+      "learning_rate": 1.1162227654143777e-05,
+      "loss": 0.1893,
+      "step": 2299
+    },
+    {
+      "epoch": 2.583181243857925,
+      "grad_norm": 0.4531359125712725,
+      "learning_rate": 1.1154436279600287e-05,
+      "loss": 0.169,
+      "step": 2300
+    },
+    {
+      "epoch": 2.5843043661378635,
+      "grad_norm": 0.47496785745939757,
+      "learning_rate": 1.1146644194719454e-05,
+      "loss": 0.171,
+      "step": 2301
+    },
+    {
+      "epoch": 2.5854274884178015,
+      "grad_norm": 0.4612778970229022,
+      "learning_rate": 1.1138851404295826e-05,
+      "loss": 0.1671,
+      "step": 2302
+    },
+    {
+      "epoch": 2.58655061069774,
+      "grad_norm": 0.4963115959404571,
+      "learning_rate": 1.1131057913124399e-05,
+      "loss": 0.1823,
+      "step": 2303
+    },
+    {
+      "epoch": 2.587673732977678,
+      "grad_norm": 0.4776956369356712,
+      "learning_rate": 1.1123263726000588e-05,
+      "loss": 0.1766,
+      "step": 2304
+    },
+    {
+      "epoch": 2.588796855257616,
+      "grad_norm": 0.48648831354650407,
+      "learning_rate": 1.1115468847720245e-05,
+      "loss": 0.1757,
+      "step": 2305
+    },
+    {
+      "epoch": 2.5899199775375545,
+      "grad_norm": 0.4686638661781645,
+      "learning_rate": 1.110767328307965e-05,
+      "loss": 0.1821,
+      "step": 2306
+    },
+    {
+      "epoch": 2.5910430998174925,
+      "grad_norm": 0.44047248695541774,
+      "learning_rate": 1.109987703687549e-05,
+      "loss": 0.1657,
+      "step": 2307
+    },
+    {
+      "epoch": 2.592166222097431,
+      "grad_norm": 0.43543299715008577,
+      "learning_rate": 1.1092080113904886e-05,
+      "loss": 0.1599,
+      "step": 2308
+    },
+    {
+      "epoch": 2.5932893443773692,
+      "grad_norm": 0.47199067367590014,
+      "learning_rate": 1.1084282518965373e-05,
+      "loss": 0.1791,
+      "step": 2309
+    },
+    {
+      "epoch": 2.594412466657307,
+      "grad_norm": 0.4532641018961933,
+      "learning_rate": 1.1076484256854889e-05,
+      "loss": 0.1678,
+      "step": 2310
+    },
+    {
+      "epoch": 2.5955355889372456,
+      "grad_norm": 0.45735825060841906,
+      "learning_rate": 1.1068685332371802e-05,
+      "loss": 0.1838,
+      "step": 2311
+    },
+    {
+      "epoch": 2.5966587112171835,
+      "grad_norm": 0.4591137341726451,
+      "learning_rate": 1.1060885750314865e-05,
+      "loss": 0.1765,
+      "step": 2312
+    },
+    {
+      "epoch": 2.597781833497122,
+      "grad_norm": 0.4679142193330497,
+      "learning_rate": 1.1053085515483255e-05,
+      "loss": 0.179,
+      "step": 2313
+    },
+    {
+      "epoch": 2.5989049557770603,
+      "grad_norm": 0.47518636171561446,
+      "learning_rate": 1.1045284632676535e-05,
+      "loss": 0.1698,
+      "step": 2314
+    },
+    {
+      "epoch": 2.600028078056998,
+      "grad_norm": 0.4565346667945619,
+      "learning_rate": 1.1037483106694681e-05,
+      "loss": 0.1706,
+      "step": 2315
+    },
+    {
+      "epoch": 2.6011512003369366,
+      "grad_norm": 0.47006822600452486,
+      "learning_rate": 1.1029680942338053e-05,
+      "loss": 0.1694,
+      "step": 2316
+    },
+    {
+      "epoch": 2.602274322616875,
+      "grad_norm": 0.4563277326138092,
+      "learning_rate": 1.1021878144407408e-05,
+      "loss": 0.1672,
+      "step": 2317
+    },
+    {
+      "epoch": 2.603397444896813,
+      "grad_norm": 0.47107089066828634,
+      "learning_rate": 1.1014074717703897e-05,
+      "loss": 0.1749,
+      "step": 2318
+    },
+    {
+      "epoch": 2.6045205671767513,
+      "grad_norm": 0.4467643283567714,
+      "learning_rate": 1.1006270667029054e-05,
+      "loss": 0.1679,
+      "step": 2319
+    },
+    {
+      "epoch": 2.6056436894566897,
+      "grad_norm": 0.4587123434597116,
+      "learning_rate": 1.0998465997184798e-05,
+      "loss": 0.1777,
+      "step": 2320
+    },
+    {
+      "epoch": 2.6067668117366276,
+      "grad_norm": 0.4983006072884289,
+      "learning_rate": 1.099066071297342e-05,
+      "loss": 0.1874,
+      "step": 2321
+    },
+    {
+      "epoch": 2.607889934016566,
+      "grad_norm": 0.476575665366494,
+      "learning_rate": 1.0982854819197609e-05,
+      "loss": 0.1747,
+      "step": 2322
+    },
+    {
+      "epoch": 2.6090130562965044,
+      "grad_norm": 0.45400404446255577,
+      "learning_rate": 1.0975048320660408e-05,
+      "loss": 0.1801,
+      "step": 2323
+    },
+    {
+      "epoch": 2.6101361785764423,
+      "grad_norm": 0.4332841064106607,
+      "learning_rate": 1.0967241222165247e-05,
+      "loss": 0.1632,
+      "step": 2324
+    },
+    {
+      "epoch": 2.6112593008563807,
+      "grad_norm": 0.4510328447149446,
+      "learning_rate": 1.0959433528515922e-05,
+      "loss": 0.1741,
+      "step": 2325
+    },
+    {
+      "epoch": 2.612382423136319,
+      "grad_norm": 0.454925383272742,
+      "learning_rate": 1.0951625244516584e-05,
+      "loss": 0.1727,
+      "step": 2326
+    },
+    {
+      "epoch": 2.613505545416257,
+      "grad_norm": 0.4371439911786275,
+      "learning_rate": 1.094381637497176e-05,
+      "loss": 0.1673,
+      "step": 2327
+    },
+    {
+      "epoch": 2.6146286676961954,
+      "grad_norm": 0.47013253731307947,
+      "learning_rate": 1.0936006924686337e-05,
+      "loss": 0.1708,
+      "step": 2328
+    },
+    {
+      "epoch": 2.615751789976134,
+      "grad_norm": 0.46539714593918086,
+      "learning_rate": 1.0928196898465552e-05,
+      "loss": 0.1801,
+      "step": 2329
+    },
+    {
+      "epoch": 2.6168749122560717,
+      "grad_norm": 0.45190767645037216,
+      "learning_rate": 1.0920386301115e-05,
+      "loss": 0.1744,
+      "step": 2330
+    },
+    {
+      "epoch": 2.61799803453601,
+      "grad_norm": 0.4538448950274808,
+      "learning_rate": 1.091257513744063e-05,
+      "loss": 0.1681,
+      "step": 2331
+    },
+    {
+      "epoch": 2.6191211568159485,
+      "grad_norm": 0.488178449583916,
+      "learning_rate": 1.0904763412248736e-05,
+      "loss": 0.1845,
+      "step": 2332
+    },
+    {
+      "epoch": 2.6202442790958864,
+      "grad_norm": 0.46938216478064526,
+      "learning_rate": 1.0896951130345957e-05,
+      "loss": 0.1767,
+      "step": 2333
+    },
+    {
+      "epoch": 2.621367401375825,
+      "grad_norm": 0.44361139483028894,
+      "learning_rate": 1.0889138296539277e-05,
+      "loss": 0.1719,
+      "step": 2334
+    },
+    {
+      "epoch": 2.622490523655763,
+      "grad_norm": 0.4571118704858418,
+      "learning_rate": 1.088132491563602e-05,
+      "loss": 0.1693,
+      "step": 2335
+    },
+    {
+      "epoch": 2.623613645935701,
+      "grad_norm": 0.4600308129329723,
+      "learning_rate": 1.0873510992443841e-05,
+      "loss": 0.1666,
+      "step": 2336
+    },
+    {
+      "epoch": 2.6247367682156395,
+      "grad_norm": 0.4609852591114243,
+      "learning_rate": 1.086569653177074e-05,
+      "loss": 0.1714,
+      "step": 2337
+    },
+    {
+      "epoch": 2.625859890495578,
+      "grad_norm": 0.47717459396403794,
+      "learning_rate": 1.0857881538425032e-05,
+      "loss": 0.1872,
+      "step": 2338
+    },
+    {
+      "epoch": 2.626983012775516,
+      "grad_norm": 0.4617597071752641,
+      "learning_rate": 1.0850066017215375e-05,
+      "loss": 0.1745,
+      "step": 2339
+    },
+    {
+      "epoch": 2.628106135055454,
+      "grad_norm": 0.4654760240823605,
+      "learning_rate": 1.0842249972950743e-05,
+      "loss": 0.1713,
+      "step": 2340
+    },
+    {
+      "epoch": 2.6292292573353926,
+      "grad_norm": 0.4698177808893506,
+      "learning_rate": 1.0834433410440432e-05,
+      "loss": 0.1716,
+      "step": 2341
+    },
+    {
+      "epoch": 2.6303523796153305,
+      "grad_norm": 0.47185791755913564,
+      "learning_rate": 1.0826616334494068e-05,
+      "loss": 0.177,
+      "step": 2342
+    },
+    {
+      "epoch": 2.631475501895269,
+      "grad_norm": 0.46380423420344924,
+      "learning_rate": 1.0818798749921569e-05,
+      "loss": 0.172,
+      "step": 2343
+    },
+    {
+      "epoch": 2.6325986241752073,
+      "grad_norm": 0.4790054584186303,
+      "learning_rate": 1.081098066153319e-05,
+      "loss": 0.1726,
+      "step": 2344
+    },
+    {
+      "epoch": 2.6337217464551452,
+      "grad_norm": 0.46882773567793207,
+      "learning_rate": 1.0803162074139489e-05,
+      "loss": 0.1699,
+      "step": 2345
+    },
+    {
+      "epoch": 2.6348448687350836,
+      "grad_norm": 0.45878850668533416,
+      "learning_rate": 1.0795342992551323e-05,
+      "loss": 0.1749,
+      "step": 2346
+    },
+    {
+      "epoch": 2.635967991015022,
+      "grad_norm": 0.46305430397540676,
+      "learning_rate": 1.0787523421579862e-05,
+      "loss": 0.1797,
+      "step": 2347
+    },
+    {
+      "epoch": 2.63709111329496,
+      "grad_norm": 0.4663636966697608,
+      "learning_rate": 1.0779703366036573e-05,
+      "loss": 0.1733,
+      "step": 2348
+    },
+    {
+      "epoch": 2.6382142355748983,
+      "grad_norm": 0.48294894032218394,
+      "learning_rate": 1.0771882830733223e-05,
+      "loss": 0.1887,
+      "step": 2349
+    },
+    {
+      "epoch": 2.6393373578548367,
+      "grad_norm": 0.45323268388669674,
+      "learning_rate": 1.0764061820481872e-05,
+      "loss": 0.1717,
+      "step": 2350
+    },
+    {
+      "epoch": 2.6404604801347746,
+      "grad_norm": 0.4532472744181654,
+      "learning_rate": 1.0756240340094877e-05,
+      "loss": 0.1732,
+      "step": 2351
+    },
+    {
+      "epoch": 2.641583602414713,
+      "grad_norm": 0.4383917984038164,
+      "learning_rate": 1.0748418394384876e-05,
+      "loss": 0.1658,
+      "step": 2352
+    },
+    {
+      "epoch": 2.6427067246946514,
+      "grad_norm": 0.4661107283107099,
+      "learning_rate": 1.07405959881648e-05,
+      "loss": 0.1863,
+      "step": 2353
+    },
+    {
+      "epoch": 2.6438298469745893,
+      "grad_norm": 0.4737303429465663,
+      "learning_rate": 1.0732773126247867e-05,
+      "loss": 0.1771,
+      "step": 2354
+    },
+    {
+      "epoch": 2.6449529692545277,
+      "grad_norm": 0.47017332443037857,
+      "learning_rate": 1.0724949813447563e-05,
+      "loss": 0.1724,
+      "step": 2355
+    },
+    {
+      "epoch": 2.6460760915344657,
+      "grad_norm": 0.49757478615437767,
+      "learning_rate": 1.071712605457766e-05,
+      "loss": 0.1824,
+      "step": 2356
+    },
+    {
+      "epoch": 2.647199213814404,
+      "grad_norm": 0.47195996519974315,
+      "learning_rate": 1.0709301854452207e-05,
+      "loss": 0.1778,
+      "step": 2357
+    },
+    {
+      "epoch": 2.6483223360943424,
+      "grad_norm": 0.48072524454750337,
+      "learning_rate": 1.0701477217885517e-05,
+      "loss": 0.1854,
+      "step": 2358
+    },
+    {
+      "epoch": 2.6494454583742804,
+      "grad_norm": 0.48656148421443146,
+      "learning_rate": 1.0693652149692175e-05,
+      "loss": 0.1838,
+      "step": 2359
+    },
+    {
+      "epoch": 2.6505685806542187,
+      "grad_norm": 0.4565743939469196,
+      "learning_rate": 1.068582665468703e-05,
+      "loss": 0.1767,
+      "step": 2360
+    },
+    {
+      "epoch": 2.6516917029341567,
+      "grad_norm": 0.4785810820290836,
+      "learning_rate": 1.0678000737685197e-05,
+      "loss": 0.1738,
+      "step": 2361
+    },
+    {
+      "epoch": 2.652814825214095,
+      "grad_norm": 0.4574908421572849,
+      "learning_rate": 1.0670174403502051e-05,
+      "loss": 0.1765,
+      "step": 2362
+    },
+    {
+      "epoch": 2.6539379474940334,
+      "grad_norm": 0.465593556602867,
+      "learning_rate": 1.0662347656953221e-05,
+      "loss": 0.1679,
+      "step": 2363
+    },
+    {
+      "epoch": 2.6550610697739714,
+      "grad_norm": 0.46131939499983016,
+      "learning_rate": 1.0654520502854588e-05,
+      "loss": 0.1777,
+      "step": 2364
+    },
+    {
+      "epoch": 2.6561841920539098,
+      "grad_norm": 0.48218612435761826,
+      "learning_rate": 1.0646692946022285e-05,
+      "loss": 0.1927,
+      "step": 2365
+    },
+    {
+      "epoch": 2.657307314333848,
+      "grad_norm": 0.44549729019844936,
+      "learning_rate": 1.0638864991272698e-05,
+      "loss": 0.1647,
+      "step": 2366
+    },
+    {
+      "epoch": 2.658430436613786,
+      "grad_norm": 0.4723488402722565,
+      "learning_rate": 1.063103664342245e-05,
+      "loss": 0.1757,
+      "step": 2367
+    },
+    {
+      "epoch": 2.6595535588937245,
+      "grad_norm": 0.46272844102717686,
+      "learning_rate": 1.0623207907288409e-05,
+      "loss": 0.1719,
+      "step": 2368
+    },
+    {
+      "epoch": 2.660676681173663,
+      "grad_norm": 0.45549848321049613,
+      "learning_rate": 1.061537878768769e-05,
+      "loss": 0.1697,
+      "step": 2369
+    },
+    {
+      "epoch": 2.661799803453601,
+      "grad_norm": 0.45675590525963283,
+      "learning_rate": 1.0607549289437626e-05,
+      "loss": 0.174,
+      "step": 2370
+    },
+    {
+      "epoch": 2.662922925733539,
+      "grad_norm": 0.47841033221623763,
+      "learning_rate": 1.0599719417355801e-05,
+      "loss": 0.1778,
+      "step": 2371
+    },
+    {
+      "epoch": 2.6640460480134776,
+      "grad_norm": 0.4536400418140574,
+      "learning_rate": 1.0591889176260017e-05,
+      "loss": 0.1737,
+      "step": 2372
+    },
+    {
+      "epoch": 2.6651691702934155,
+      "grad_norm": 0.4630804441617724,
+      "learning_rate": 1.0584058570968312e-05,
+      "loss": 0.1792,
+      "step": 2373
+    },
+    {
+      "epoch": 2.666292292573354,
+      "grad_norm": 0.4382451498975878,
+      "learning_rate": 1.0576227606298937e-05,
+      "loss": 0.1567,
+      "step": 2374
+    },
+    {
+      "epoch": 2.6674154148532923,
+      "grad_norm": 0.4584632279254452,
+      "learning_rate": 1.0568396287070377e-05,
+      "loss": 0.1751,
+      "step": 2375
+    },
+    {
+      "epoch": 2.66853853713323,
+      "grad_norm": 0.46530159309615415,
+      "learning_rate": 1.0560564618101328e-05,
+      "loss": 0.1796,
+      "step": 2376
+    },
+    {
+      "epoch": 2.6696616594131686,
+      "grad_norm": 0.4635165893290825,
+      "learning_rate": 1.0552732604210701e-05,
+      "loss": 0.1737,
+      "step": 2377
+    },
+    {
+      "epoch": 2.670784781693107,
+      "grad_norm": 0.4395153429761878,
+      "learning_rate": 1.0544900250217615e-05,
+      "loss": 0.1599,
+      "step": 2378
+    },
+    {
+      "epoch": 2.671907903973045,
+      "grad_norm": 0.45345637640530895,
+      "learning_rate": 1.0537067560941416e-05,
+      "loss": 0.1746,
+      "step": 2379
+    },
+    {
+      "epoch": 2.6730310262529833,
+      "grad_norm": 0.45176101895188636,
+      "learning_rate": 1.0529234541201631e-05,
+      "loss": 0.1724,
+      "step": 2380
+    },
+    {
+      "epoch": 2.6741541485329217,
+      "grad_norm": 0.46146241332136667,
+      "learning_rate": 1.0521401195818014e-05,
+      "loss": 0.1748,
+      "step": 2381
+    },
+    {
+      "epoch": 2.6752772708128596,
+      "grad_norm": 0.4644112669184593,
+      "learning_rate": 1.0513567529610498e-05,
+      "loss": 0.1792,
+      "step": 2382
+    },
+    {
+      "epoch": 2.676400393092798,
+      "grad_norm": 0.474578955286889,
+      "learning_rate": 1.050573354739923e-05,
+      "loss": 0.1785,
+      "step": 2383
+    },
+    {
+      "epoch": 2.6775235153727364,
+      "grad_norm": 0.4361002502010669,
+      "learning_rate": 1.049789925400455e-05,
+      "loss": 0.1625,
+      "step": 2384
+    },
+    {
+      "epoch": 2.6786466376526743,
+      "grad_norm": 0.45846134672913397,
+      "learning_rate": 1.0490064654246976e-05,
+      "loss": 0.18,
+      "step": 2385
+    },
+    {
+      "epoch": 2.6797697599326127,
+      "grad_norm": 0.46526875476056234,
+      "learning_rate": 1.0482229752947228e-05,
+      "loss": 0.1712,
+      "step": 2386
+    },
+    {
+      "epoch": 2.680892882212551,
+      "grad_norm": 0.4661257112898557,
+      "learning_rate": 1.0474394554926206e-05,
+      "loss": 0.1815,
+      "step": 2387
+    },
+    {
+      "epoch": 2.682016004492489,
+      "grad_norm": 0.4648307273335071,
+      "learning_rate": 1.0466559065004995e-05,
+      "loss": 0.17,
+      "step": 2388
+    },
+    {
+      "epoch": 2.6831391267724274,
+      "grad_norm": 0.4673176281630868,
+      "learning_rate": 1.0458723288004858e-05,
+      "loss": 0.1689,
+      "step": 2389
+    },
+    {
+      "epoch": 2.6842622490523658,
+      "grad_norm": 0.4580586103480894,
+      "learning_rate": 1.0450887228747229e-05,
+      "loss": 0.1608,
+      "step": 2390
+    },
+    {
+      "epoch": 2.6853853713323037,
+      "grad_norm": 0.46591951479828025,
+      "learning_rate": 1.0443050892053733e-05,
+      "loss": 0.1752,
+      "step": 2391
+    },
+    {
+      "epoch": 2.686508493612242,
+      "grad_norm": 0.4515547388684425,
+      "learning_rate": 1.0435214282746142e-05,
+      "loss": 0.1598,
+      "step": 2392
+    },
+    {
+      "epoch": 2.6876316158921805,
+      "grad_norm": 0.4951158425551178,
+      "learning_rate": 1.0427377405646414e-05,
+      "loss": 0.1933,
+      "step": 2393
+    },
+    {
+      "epoch": 2.6887547381721184,
+      "grad_norm": 0.47544695446998514,
+      "learning_rate": 1.0419540265576666e-05,
+      "loss": 0.1797,
+      "step": 2394
+    },
+    {
+      "epoch": 2.689877860452057,
+      "grad_norm": 0.45123909591416184,
+      "learning_rate": 1.041170286735918e-05,
+      "loss": 0.1724,
+      "step": 2395
+    },
+    {
+      "epoch": 2.691000982731995,
+      "grad_norm": 0.462230580491664,
+      "learning_rate": 1.0403865215816382e-05,
+      "loss": 0.1846,
+      "step": 2396
+    },
+    {
+      "epoch": 2.692124105011933,
+      "grad_norm": 0.452353253692668,
+      "learning_rate": 1.0396027315770876e-05,
+      "loss": 0.1757,
+      "step": 2397
+    },
+    {
+      "epoch": 2.6932472272918715,
+      "grad_norm": 0.45256519362751757,
+      "learning_rate": 1.0388189172045407e-05,
+      "loss": 0.1815,
+      "step": 2398
+    },
+    {
+      "epoch": 2.69437034957181,
+      "grad_norm": 0.47210582761980324,
+      "learning_rate": 1.0380350789462865e-05,
+      "loss": 0.178,
+      "step": 2399
+    },
+    {
+      "epoch": 2.695493471851748,
+      "grad_norm": 0.4544816107564523,
+      "learning_rate": 1.0372512172846296e-05,
+      "loss": 0.1659,
+      "step": 2400
+    },
+    {
+      "epoch": 2.696616594131686,
+      "grad_norm": 0.47543248227366125,
+      "learning_rate": 1.0364673327018891e-05,
+      "loss": 0.1823,
+      "step": 2401
+    },
+    {
+      "epoch": 2.6977397164116246,
+      "grad_norm": 0.4645680632421113,
+      "learning_rate": 1.0356834256803974e-05,
+      "loss": 0.1817,
+      "step": 2402
+    },
+    {
+      "epoch": 2.6988628386915625,
+      "grad_norm": 0.47229910569982986,
+      "learning_rate": 1.0348994967025012e-05,
+      "loss": 0.1705,
+      "step": 2403
+    },
+    {
+      "epoch": 2.699985960971501,
+      "grad_norm": 0.4709976896475921,
+      "learning_rate": 1.0341155462505606e-05,
+      "loss": 0.1816,
+      "step": 2404
+    },
+    {
+      "epoch": 2.701109083251439,
+      "grad_norm": 0.43745184102209106,
+      "learning_rate": 1.033331574806949e-05,
+      "loss": 0.1542,
+      "step": 2405
+    },
+    {
+      "epoch": 2.702232205531377,
+      "grad_norm": 0.47528735474280753,
+      "learning_rate": 1.0325475828540524e-05,
+      "loss": 0.1746,
+      "step": 2406
+    },
+    {
+      "epoch": 2.7033553278113156,
+      "grad_norm": 0.4864779824250153,
+      "learning_rate": 1.03176357087427e-05,
+      "loss": 0.1861,
+      "step": 2407
+    },
+    {
+      "epoch": 2.7044784500912535,
+      "grad_norm": 0.44098989955704543,
+      "learning_rate": 1.030979539350013e-05,
+      "loss": 0.1665,
+      "step": 2408
+    },
+    {
+      "epoch": 2.705601572371192,
+      "grad_norm": 0.45665136007590384,
+      "learning_rate": 1.0301954887637045e-05,
+      "loss": 0.1701,
+      "step": 2409
+    },
+    {
+      "epoch": 2.7067246946511303,
+      "grad_norm": 0.457919080022194,
+      "learning_rate": 1.0294114195977796e-05,
+      "loss": 0.1647,
+      "step": 2410
+    },
+    {
+      "epoch": 2.7078478169310682,
+      "grad_norm": 0.47134438824824476,
+      "learning_rate": 1.0286273323346843e-05,
+      "loss": 0.1811,
+      "step": 2411
+    },
+    {
+      "epoch": 2.7089709392110066,
+      "grad_norm": 0.48672871636486603,
+      "learning_rate": 1.0278432274568765e-05,
+      "loss": 0.1743,
+      "step": 2412
+    },
+    {
+      "epoch": 2.7100940614909446,
+      "grad_norm": 0.4619105109658203,
+      "learning_rate": 1.0270591054468244e-05,
+      "loss": 0.1684,
+      "step": 2413
+    },
+    {
+      "epoch": 2.711217183770883,
+      "grad_norm": 0.4693014599142233,
+      "learning_rate": 1.0262749667870071e-05,
+      "loss": 0.1735,
+      "step": 2414
+    },
+    {
+      "epoch": 2.7123403060508213,
+      "grad_norm": 0.4536596675091747,
+      "learning_rate": 1.0254908119599134e-05,
+      "loss": 0.1663,
+      "step": 2415
+    },
+    {
+      "epoch": 2.7134634283307593,
+      "grad_norm": 0.4762674495287221,
+      "learning_rate": 1.0247066414480424e-05,
+      "loss": 0.1733,
+      "step": 2416
+    },
+    {
+      "epoch": 2.7145865506106976,
+      "grad_norm": 0.4689655559574979,
+      "learning_rate": 1.0239224557339035e-05,
+      "loss": 0.1784,
+      "step": 2417
+    },
+    {
+      "epoch": 2.715709672890636,
+      "grad_norm": 0.4673344907187584,
+      "learning_rate": 1.0231382553000143e-05,
+      "loss": 0.1837,
+      "step": 2418
+    },
+    {
+      "epoch": 2.716832795170574,
+      "grad_norm": 0.46231195313403367,
+      "learning_rate": 1.0223540406289017e-05,
+      "loss": 0.1691,
+      "step": 2419
+    },
+    {
+      "epoch": 2.7179559174505123,
+      "grad_norm": 0.4722466169309128,
+      "learning_rate": 1.0215698122031021e-05,
+      "loss": 0.1765,
+      "step": 2420
+    },
+    {
+      "epoch": 2.7190790397304507,
+      "grad_norm": 0.4464937713270927,
+      "learning_rate": 1.0207855705051595e-05,
+      "loss": 0.1692,
+      "step": 2421
+    },
+    {
+      "epoch": 2.7202021620103887,
+      "grad_norm": 0.43287348633124384,
+      "learning_rate": 1.020001316017627e-05,
+      "loss": 0.161,
+      "step": 2422
+    },
+    {
+      "epoch": 2.721325284290327,
+      "grad_norm": 0.46989012910634004,
+      "learning_rate": 1.0192170492230643e-05,
+      "loss": 0.1809,
+      "step": 2423
+    },
+    {
+      "epoch": 2.7224484065702654,
+      "grad_norm": 0.4705286884689803,
+      "learning_rate": 1.0184327706040397e-05,
+      "loss": 0.174,
+      "step": 2424
+    },
+    {
+      "epoch": 2.7235715288502034,
+      "grad_norm": 0.43611447014588406,
+      "learning_rate": 1.0176484806431288e-05,
+      "loss": 0.1604,
+      "step": 2425
+    },
+    {
+      "epoch": 2.7246946511301418,
+      "grad_norm": 0.4794627114691746,
+      "learning_rate": 1.0168641798229133e-05,
+      "loss": 0.1802,
+      "step": 2426
+    },
+    {
+      "epoch": 2.72581777341008,
+      "grad_norm": 0.4381810224958243,
+      "learning_rate": 1.0160798686259825e-05,
+      "loss": 0.1609,
+      "step": 2427
+    },
+    {
+      "epoch": 2.726940895690018,
+      "grad_norm": 0.48565778276163146,
+      "learning_rate": 1.0152955475349316e-05,
+      "loss": 0.1878,
+      "step": 2428
+    },
+    {
+      "epoch": 2.7280640179699565,
+      "grad_norm": 0.46551401986940266,
+      "learning_rate": 1.014511217032362e-05,
+      "loss": 0.1744,
+      "step": 2429
+    },
+    {
+      "epoch": 2.729187140249895,
+      "grad_norm": 0.45724136025751433,
+      "learning_rate": 1.0137268776008809e-05,
+      "loss": 0.17,
+      "step": 2430
+    },
+    {
+      "epoch": 2.7303102625298328,
+      "grad_norm": 0.45946275216502946,
+      "learning_rate": 1.0129425297231005e-05,
+      "loss": 0.1766,
+      "step": 2431
+    },
+    {
+      "epoch": 2.731433384809771,
+      "grad_norm": 0.46397417169666205,
+      "learning_rate": 1.0121581738816397e-05,
+      "loss": 0.174,
+      "step": 2432
+    },
+    {
+      "epoch": 2.7325565070897095,
+      "grad_norm": 0.46245651199187193,
+      "learning_rate": 1.0113738105591203e-05,
+      "loss": 0.1792,
+      "step": 2433
+    },
+    {
+      "epoch": 2.7336796293696475,
+      "grad_norm": 0.4524764262786222,
+      "learning_rate": 1.0105894402381703e-05,
+      "loss": 0.1694,
+      "step": 2434
+    },
+    {
+      "epoch": 2.734802751649586,
+      "grad_norm": 0.4529695543319877,
+      "learning_rate": 1.0098050634014216e-05,
+      "loss": 0.1753,
+      "step": 2435
+    },
+    {
+      "epoch": 2.7359258739295242,
+      "grad_norm": 0.45224833550454835,
+      "learning_rate": 1.0090206805315087e-05,
+      "loss": 0.1688,
+      "step": 2436
+    },
+    {
+      "epoch": 2.737048996209462,
+      "grad_norm": 0.4559944373716082,
+      "learning_rate": 1.0082362921110721e-05,
+      "loss": 0.1715,
+      "step": 2437
+    },
+    {
+      "epoch": 2.7381721184894006,
+      "grad_norm": 0.45958405120251117,
+      "learning_rate": 1.0074518986227546e-05,
+      "loss": 0.17,
+      "step": 2438
+    },
+    {
+      "epoch": 2.739295240769339,
+      "grad_norm": 0.4490783585500566,
+      "learning_rate": 1.0066675005492017e-05,
+      "loss": 0.1715,
+      "step": 2439
+    },
+    {
+      "epoch": 2.740418363049277,
+      "grad_norm": 0.448796044847132,
+      "learning_rate": 1.0058830983730622e-05,
+      "loss": 0.178,
+      "step": 2440
+    },
+    {
+      "epoch": 2.7415414853292153,
+      "grad_norm": 0.45884028647372394,
+      "learning_rate": 1.0050986925769877e-05,
+      "loss": 0.1731,
+      "step": 2441
+    },
+    {
+      "epoch": 2.7426646076091536,
+      "grad_norm": 0.4426418856703806,
+      "learning_rate": 1.0043142836436316e-05,
+      "loss": 0.16,
+      "step": 2442
+    },
+    {
+      "epoch": 2.7437877298890916,
+      "grad_norm": 0.4678363877463838,
+      "learning_rate": 1.0035298720556493e-05,
+      "loss": 0.1859,
+      "step": 2443
+    },
+    {
+      "epoch": 2.74491085216903,
+      "grad_norm": 0.4676021495394005,
+      "learning_rate": 1.002745458295698e-05,
+      "loss": 0.1706,
+      "step": 2444
+    },
+    {
+      "epoch": 2.7460339744489684,
+      "grad_norm": 0.4623563788627261,
+      "learning_rate": 1.0019610428464354e-05,
+      "loss": 0.1755,
+      "step": 2445
+    },
+    {
+      "epoch": 2.7471570967289063,
+      "grad_norm": 0.4567998634977174,
+      "learning_rate": 1.001176626190522e-05,
+      "loss": 0.1749,
+      "step": 2446
+    },
+    {
+      "epoch": 2.7482802190088447,
+      "grad_norm": 0.47580917906104647,
+      "learning_rate": 1.0003922088106178e-05,
+      "loss": 0.1795,
+      "step": 2447
+    },
+    {
+      "epoch": 2.749403341288783,
+      "grad_norm": 0.4595058756812876,
+      "learning_rate": 9.996077911893829e-06,
+      "loss": 0.1758,
+      "step": 2448
+    },
+    {
+      "epoch": 2.750526463568721,
+      "grad_norm": 0.45863685177045854,
+      "learning_rate": 9.988233738094782e-06,
+      "loss": 0.177,
+      "step": 2449
+    },
+    {
+      "epoch": 2.7516495858486594,
+      "grad_norm": 0.4547228686150152,
+      "learning_rate": 9.980389571535647e-06,
+      "loss": 0.1738,
+      "step": 2450
+    },
+    {
+      "epoch": 2.7527727081285978,
+      "grad_norm": 0.4425914309842023,
+      "learning_rate": 9.972545417043024e-06,
+      "loss": 0.1591,
+      "step": 2451
+    },
+    {
+      "epoch": 2.7538958304085357,
+      "grad_norm": 0.4739331962858264,
+      "learning_rate": 9.964701279443509e-06,
+      "loss": 0.1862,
+      "step": 2452
+    },
+    {
+      "epoch": 2.755018952688474,
+      "grad_norm": 0.44719582544834424,
+      "learning_rate": 9.956857163563689e-06,
+      "loss": 0.1621,
+      "step": 2453
+    },
+    {
+      "epoch": 2.7561420749684125,
+      "grad_norm": 0.4493034917391583,
+      "learning_rate": 9.949013074230127e-06,
+      "loss": 0.1678,
+      "step": 2454
+    },
+    {
+      "epoch": 2.7572651972483504,
+      "grad_norm": 0.4341676937820874,
+      "learning_rate": 9.94116901626938e-06,
+      "loss": 0.1635,
+      "step": 2455
+    },
+    {
+      "epoch": 2.758388319528289,
+      "grad_norm": 0.44123424149038765,
+      "learning_rate": 9.933324994507984e-06,
+      "loss": 0.1617,
+      "step": 2456
+    },
+    {
+      "epoch": 2.7595114418082267,
+      "grad_norm": 0.4520594389268704,
+      "learning_rate": 9.925481013772456e-06,
+      "loss": 0.166,
+      "step": 2457
+    },
+    {
+      "epoch": 2.760634564088165,
+      "grad_norm": 0.47723497756567185,
+      "learning_rate": 9.91763707888928e-06,
+      "loss": 0.1865,
+      "step": 2458
+    },
+    {
+      "epoch": 2.7617576863681035,
+      "grad_norm": 0.4555339710437883,
+      "learning_rate": 9.909793194684914e-06,
+      "loss": 0.1806,
+      "step": 2459
+    },
+    {
+      "epoch": 2.7628808086480414,
+      "grad_norm": 0.4636447483901148,
+      "learning_rate": 9.901949365985787e-06,
+      "loss": 0.1672,
+      "step": 2460
+    },
+    {
+      "epoch": 2.76400393092798,
+      "grad_norm": 0.468526515916128,
+      "learning_rate": 9.894105597618297e-06,
+      "loss": 0.1739,
+      "step": 2461
+    },
+    {
+      "epoch": 2.7651270532079177,
+      "grad_norm": 0.4436410518618047,
+      "learning_rate": 9.886261894408798e-06,
+      "loss": 0.1604,
+      "step": 2462
+    },
+    {
+      "epoch": 2.766250175487856,
+      "grad_norm": 0.4584305461946224,
+      "learning_rate": 9.878418261183606e-06,
+      "loss": 0.1645,
+      "step": 2463
+    },
+    {
+      "epoch": 2.7673732977677945,
+      "grad_norm": 0.48434226395422153,
+      "learning_rate": 9.870574702768997e-06,
+      "loss": 0.1757,
+      "step": 2464
+    },
+    {
+      "epoch": 2.7684964200477324,
+      "grad_norm": 0.48156835406806114,
+      "learning_rate": 9.862731223991196e-06,
+      "loss": 0.1845,
+      "step": 2465
+    },
+    {
+      "epoch": 2.769619542327671,
+      "grad_norm": 0.44955147765140463,
+      "learning_rate": 9.854887829676382e-06,
+      "loss": 0.1654,
+      "step": 2466
+    },
+    {
+      "epoch": 2.770742664607609,
+      "grad_norm": 0.46769768486111885,
+      "learning_rate": 9.847044524650689e-06,
+      "loss": 0.1839,
+      "step": 2467
+    },
+    {
+      "epoch": 2.771865786887547,
+      "grad_norm": 0.4802433910241491,
+      "learning_rate": 9.839201313740179e-06,
+      "loss": 0.1848,
+      "step": 2468
+    },
+    {
+      "epoch": 2.7729889091674855,
+      "grad_norm": 0.4730216994614374,
+      "learning_rate": 9.83135820177087e-06,
+      "loss": 0.1784,
+      "step": 2469
+    },
+    {
+      "epoch": 2.774112031447424,
+      "grad_norm": 0.49525261555778016,
+      "learning_rate": 9.823515193568715e-06,
+      "loss": 0.1855,
+      "step": 2470
+    },
+    {
+      "epoch": 2.775235153727362,
+      "grad_norm": 0.44336826641873517,
+      "learning_rate": 9.815672293959605e-06,
+      "loss": 0.1696,
+      "step": 2471
+    },
+    {
+      "epoch": 2.7763582760073002,
+      "grad_norm": 0.4157328847494749,
+      "learning_rate": 9.807829507769362e-06,
+      "loss": 0.1611,
+      "step": 2472
+    },
+    {
+      "epoch": 2.7774813982872386,
+      "grad_norm": 0.42886202625170655,
+      "learning_rate": 9.799986839823736e-06,
+      "loss": 0.1618,
+      "step": 2473
+    },
+    {
+      "epoch": 2.7786045205671766,
+      "grad_norm": 0.4677846081299561,
+      "learning_rate": 9.792144294948408e-06,
+      "loss": 0.1748,
+      "step": 2474
+    },
+    {
+      "epoch": 2.779727642847115,
+      "grad_norm": 0.4583869711242806,
+      "learning_rate": 9.784301877968982e-06,
+      "loss": 0.1778,
+      "step": 2475
+    },
+    {
+      "epoch": 2.7808507651270533,
+      "grad_norm": 0.4495391147428708,
+      "learning_rate": 9.776459593710985e-06,
+      "loss": 0.1697,
+      "step": 2476
+    },
+    {
+      "epoch": 2.7819738874069913,
+      "grad_norm": 0.46075629132037005,
+      "learning_rate": 9.768617446999862e-06,
+      "loss": 0.1666,
+      "step": 2477
+    },
+    {
+      "epoch": 2.7830970096869296,
+      "grad_norm": 0.4673513087773124,
+      "learning_rate": 9.760775442660966e-06,
+      "loss": 0.1815,
+      "step": 2478
+    },
+    {
+      "epoch": 2.784220131966868,
+      "grad_norm": 0.4723634910358823,
+      "learning_rate": 9.752933585519578e-06,
+      "loss": 0.1796,
+      "step": 2479
+    },
+    {
+      "epoch": 2.785343254246806,
+      "grad_norm": 0.467537758740658,
+      "learning_rate": 9.74509188040087e-06,
+      "loss": 0.1734,
+      "step": 2480
+    },
+    {
+      "epoch": 2.7864663765267443,
+      "grad_norm": 0.4600429623850505,
+      "learning_rate": 9.737250332129932e-06,
+      "loss": 0.1653,
+      "step": 2481
+    },
+    {
+      "epoch": 2.7875894988066827,
+      "grad_norm": 0.43060840156496566,
+      "learning_rate": 9.72940894553176e-06,
+      "loss": 0.1608,
+      "step": 2482
+    },
+    {
+      "epoch": 2.7887126210866207,
+      "grad_norm": 0.4950921987658919,
+      "learning_rate": 9.721567725431239e-06,
+      "loss": 0.1915,
+      "step": 2483
+    },
+    {
+      "epoch": 2.789835743366559,
+      "grad_norm": 0.44612774116221526,
+      "learning_rate": 9.71372667665316e-06,
+      "loss": 0.1669,
+      "step": 2484
+    },
+    {
+      "epoch": 2.7909588656464974,
+      "grad_norm": 0.4585740919900049,
+      "learning_rate": 9.705885804022207e-06,
+      "loss": 0.1733,
+      "step": 2485
+    },
+    {
+      "epoch": 2.7920819879264354,
+      "grad_norm": 0.47309965981508717,
+      "learning_rate": 9.698045112362956e-06,
+      "loss": 0.1752,
+      "step": 2486
+    },
+    {
+      "epoch": 2.7932051102063737,
+      "grad_norm": 0.4541221114294698,
+      "learning_rate": 9.690204606499875e-06,
+      "loss": 0.1665,
+      "step": 2487
+    },
+    {
+      "epoch": 2.794328232486312,
+      "grad_norm": 0.4816033103876313,
+      "learning_rate": 9.682364291257304e-06,
+      "loss": 0.1854,
+      "step": 2488
+    },
+    {
+      "epoch": 2.79545135476625,
+      "grad_norm": 0.46392038210393416,
+      "learning_rate": 9.674524171459478e-06,
+      "loss": 0.1736,
+      "step": 2489
+    },
+    {
+      "epoch": 2.7965744770461884,
+      "grad_norm": 0.49408949325331897,
+      "learning_rate": 9.666684251930514e-06,
+      "loss": 0.187,
+      "step": 2490
+    },
+    {
+      "epoch": 2.797697599326127,
+      "grad_norm": 0.47463312580832867,
+      "learning_rate": 9.658844537494396e-06,
+      "loss": 0.1789,
+      "step": 2491
+    },
+    {
+      "epoch": 2.7988207216060648,
+      "grad_norm": 0.48036454400289674,
+      "learning_rate": 9.651005032974994e-06,
+      "loss": 0.178,
+      "step": 2492
+    },
+    {
+      "epoch": 2.799943843886003,
+      "grad_norm": 0.4386654951501184,
+      "learning_rate": 9.64316574319603e-06,
+      "loss": 0.1611,
+      "step": 2493
+    },
+    {
+      "epoch": 2.8010669661659415,
+      "grad_norm": 0.46695275500508854,
+      "learning_rate": 9.63532667298111e-06,
+      "loss": 0.1771,
+      "step": 2494
+    },
+    {
+      "epoch": 2.8021900884458795,
+      "grad_norm": 0.45981170005783556,
+      "learning_rate": 9.627487827153704e-06,
+      "loss": 0.1818,
+      "step": 2495
+    },
+    {
+      "epoch": 2.803313210725818,
+      "grad_norm": 0.44791412818213827,
+      "learning_rate": 9.619649210537136e-06,
+      "loss": 0.1686,
+      "step": 2496
+    },
+    {
+      "epoch": 2.8044363330057562,
+      "grad_norm": 0.4694605987355865,
+      "learning_rate": 9.6118108279546e-06,
+      "loss": 0.1804,
+      "step": 2497
+    },
+    {
+      "epoch": 2.805559455285694,
+      "grad_norm": 0.46461819232187124,
+      "learning_rate": 9.603972684229127e-06,
+      "loss": 0.1718,
+      "step": 2498
+    },
+    {
+      "epoch": 2.8066825775656326,
+      "grad_norm": 0.4640887745511831,
+      "learning_rate": 9.59613478418362e-06,
+      "loss": 0.1729,
+      "step": 2499
+    },
+    {
+      "epoch": 2.807805699845571,
+      "grad_norm": 0.46473184965851605,
+      "learning_rate": 9.588297132640824e-06,
+      "loss": 0.1799,
+      "step": 2500
+    },
+    {
+      "epoch": 2.808928822125509,
+      "grad_norm": 0.44859390204733884,
+      "learning_rate": 9.580459734423334e-06,
+      "loss": 0.1692,
+      "step": 2501
+    },
+    {
+      "epoch": 2.8100519444054473,
+      "grad_norm": 0.48966480162143305,
+      "learning_rate": 9.572622594353589e-06,
+      "loss": 0.1863,
+      "step": 2502
+    },
+    {
+      "epoch": 2.8111750666853856,
+      "grad_norm": 0.4707076422301409,
+      "learning_rate": 9.564785717253862e-06,
+      "loss": 0.1737,
+      "step": 2503
+    },
+    {
+      "epoch": 2.8122981889653236,
+      "grad_norm": 0.4798544695330848,
+      "learning_rate": 9.556949107946272e-06,
+      "loss": 0.1762,
+      "step": 2504
+    },
+    {
+      "epoch": 2.813421311245262,
+      "grad_norm": 0.4460329229033776,
+      "learning_rate": 9.549112771252771e-06,
+      "loss": 0.1645,
+      "step": 2505
+    },
+    {
+      "epoch": 2.8145444335252,
+      "grad_norm": 0.46898014918292236,
+      "learning_rate": 9.541276711995149e-06,
+      "loss": 0.1773,
+      "step": 2506
+    },
+    {
+      "epoch": 2.8156675558051383,
+      "grad_norm": 0.45833210070182195,
+      "learning_rate": 9.53344093499501e-06,
+      "loss": 0.1764,
+      "step": 2507
+    },
+    {
+      "epoch": 2.8167906780850767,
+      "grad_norm": 0.4479886546084605,
+      "learning_rate": 9.525605445073797e-06,
+      "loss": 0.1648,
+      "step": 2508
+    },
+    {
+      "epoch": 2.8179138003650146,
+      "grad_norm": 0.4701076235299984,
+      "learning_rate": 9.517770247052775e-06,
+      "loss": 0.1798,
+      "step": 2509
+    },
+    {
+      "epoch": 2.819036922644953,
+      "grad_norm": 0.44716178784872795,
+      "learning_rate": 9.509935345753026e-06,
+      "loss": 0.1668,
+      "step": 2510
+    },
+    {
+      "epoch": 2.8201600449248914,
+      "grad_norm": 0.4575598504302553,
+      "learning_rate": 9.502100745995456e-06,
+      "loss": 0.1781,
+      "step": 2511
+    },
+    {
+      "epoch": 2.8212831672048293,
+      "grad_norm": 0.46438759653279443,
+      "learning_rate": 9.494266452600771e-06,
+      "loss": 0.1759,
+      "step": 2512
+    },
+    {
+      "epoch": 2.8224062894847677,
+      "grad_norm": 0.4463213888843625,
+      "learning_rate": 9.486432470389505e-06,
+      "loss": 0.1739,
+      "step": 2513
+    },
+    {
+      "epoch": 2.8235294117647056,
+      "grad_norm": 0.44363386556020745,
+      "learning_rate": 9.47859880418199e-06,
+      "loss": 0.1677,
+      "step": 2514
+    },
+    {
+      "epoch": 2.824652534044644,
+      "grad_norm": 0.47574851732895207,
+      "learning_rate": 9.470765458798369e-06,
+      "loss": 0.1851,
+      "step": 2515
+    },
+    {
+      "epoch": 2.8257756563245824,
+      "grad_norm": 0.47371348230350735,
+      "learning_rate": 9.46293243905859e-06,
+      "loss": 0.1728,
+      "step": 2516
+    },
+    {
+      "epoch": 2.8268987786045203,
+      "grad_norm": 0.4529641523386059,
+      "learning_rate": 9.455099749782387e-06,
+      "loss": 0.171,
+      "step": 2517
+    },
+    {
+      "epoch": 2.8280219008844587,
+      "grad_norm": 0.46948443592441114,
+      "learning_rate": 9.447267395789304e-06,
+      "loss": 0.1727,
+      "step": 2518
+    },
+    {
+      "epoch": 2.829145023164397,
+      "grad_norm": 0.46792399231713533,
+      "learning_rate": 9.439435381898674e-06,
+      "loss": 0.1732,
+      "step": 2519
+    },
+    {
+      "epoch": 2.830268145444335,
+      "grad_norm": 0.46471285429284986,
+      "learning_rate": 9.431603712929623e-06,
+      "loss": 0.1792,
+      "step": 2520
+    },
+    {
+      "epoch": 2.8313912677242734,
+      "grad_norm": 0.4710096469718015,
+      "learning_rate": 9.423772393701064e-06,
+      "loss": 0.1761,
+      "step": 2521
+    },
+    {
+      "epoch": 2.832514390004212,
+      "grad_norm": 0.45657758770291,
+      "learning_rate": 9.415941429031693e-06,
+      "loss": 0.1747,
+      "step": 2522
+    },
+    {
+      "epoch": 2.8336375122841497,
+      "grad_norm": 0.4380363624369588,
+      "learning_rate": 9.408110823739985e-06,
+      "loss": 0.167,
+      "step": 2523
+    },
+    {
+      "epoch": 2.834760634564088,
+      "grad_norm": 0.4654221971780646,
+      "learning_rate": 9.400280582644204e-06,
+      "loss": 0.1704,
+      "step": 2524
+    },
+    {
+      "epoch": 2.8358837568440265,
+      "grad_norm": 0.4630864998764016,
+      "learning_rate": 9.392450710562377e-06,
+      "loss": 0.1747,
+      "step": 2525
+    },
+    {
+      "epoch": 2.8370068791239644,
+      "grad_norm": 0.47819793126694177,
+      "learning_rate": 9.384621212312316e-06,
+      "loss": 0.1806,
+      "step": 2526
+    },
+    {
+      "epoch": 2.838130001403903,
+      "grad_norm": 0.44989059188440433,
+      "learning_rate": 9.376792092711593e-06,
+      "loss": 0.1719,
+      "step": 2527
+    },
+    {
+      "epoch": 2.839253123683841,
+      "grad_norm": 0.4493159352390759,
+      "learning_rate": 9.368963356577554e-06,
+      "loss": 0.1627,
+      "step": 2528
+    },
+    {
+      "epoch": 2.840376245963779,
+      "grad_norm": 0.462068887667578,
+      "learning_rate": 9.361135008727304e-06,
+      "loss": 0.1744,
+      "step": 2529
+    },
+    {
+      "epoch": 2.8414993682437175,
+      "grad_norm": 0.46328463138927967,
+      "learning_rate": 9.353307053977717e-06,
+      "loss": 0.1698,
+      "step": 2530
+    },
+    {
+      "epoch": 2.842622490523656,
+      "grad_norm": 0.4535786753854389,
+      "learning_rate": 9.345479497145417e-06,
+      "loss": 0.1598,
+      "step": 2531
+    },
+    {
+      "epoch": 2.843745612803594,
+      "grad_norm": 0.4659782442541839,
+      "learning_rate": 9.337652343046782e-06,
+      "loss": 0.1829,
+      "step": 2532
+    },
+    {
+      "epoch": 2.844868735083532,
+      "grad_norm": 0.43264542173054715,
+      "learning_rate": 9.32982559649795e-06,
+      "loss": 0.1633,
+      "step": 2533
+    },
+    {
+      "epoch": 2.8459918573634706,
+      "grad_norm": 0.4467290239010384,
+      "learning_rate": 9.321999262314803e-06,
+      "loss": 0.1668,
+      "step": 2534
+    },
+    {
+      "epoch": 2.8471149796434085,
+      "grad_norm": 0.4546808256348493,
+      "learning_rate": 9.314173345312972e-06,
+      "loss": 0.1729,
+      "step": 2535
+    },
+    {
+      "epoch": 2.848238101923347,
+      "grad_norm": 0.46681169539201894,
+      "learning_rate": 9.30634785030783e-06,
+      "loss": 0.1768,
+      "step": 2536
+    },
+    {
+      "epoch": 2.8493612242032853,
+      "grad_norm": 0.48967009130593026,
+      "learning_rate": 9.298522782114488e-06,
+      "loss": 0.1876,
+      "step": 2537
+    },
+    {
+      "epoch": 2.8504843464832232,
+      "grad_norm": 0.4476752114365759,
+      "learning_rate": 9.290698145547796e-06,
+      "loss": 0.1624,
+      "step": 2538
+    },
+    {
+      "epoch": 2.8516074687631616,
+      "grad_norm": 0.478382573772851,
+      "learning_rate": 9.282873945422341e-06,
+      "loss": 0.1843,
+      "step": 2539
+    },
+    {
+      "epoch": 2.8527305910431,
+      "grad_norm": 0.45741308655522045,
+      "learning_rate": 9.27505018655244e-06,
+      "loss": 0.1733,
+      "step": 2540
+    },
+    {
+      "epoch": 2.853853713323038,
+      "grad_norm": 0.4683894832083663,
+      "learning_rate": 9.267226873752137e-06,
+      "loss": 0.1729,
+      "step": 2541
+    },
+    {
+      "epoch": 2.8549768356029763,
+      "grad_norm": 0.4621198618194314,
+      "learning_rate": 9.259404011835203e-06,
+      "loss": 0.1683,
+      "step": 2542
+    },
+    {
+      "epoch": 2.8560999578829147,
+      "grad_norm": 0.47870289959867185,
+      "learning_rate": 9.251581605615128e-06,
+      "loss": 0.1751,
+      "step": 2543
+    },
+    {
+      "epoch": 2.8572230801628526,
+      "grad_norm": 0.48939618598564266,
+      "learning_rate": 9.243759659905126e-06,
+      "loss": 0.1851,
+      "step": 2544
+    },
+    {
+      "epoch": 2.858346202442791,
+      "grad_norm": 0.48615941707120763,
+      "learning_rate": 9.235938179518131e-06,
+      "loss": 0.1825,
+      "step": 2545
+    },
+    {
+      "epoch": 2.8594693247227294,
+      "grad_norm": 0.46996292093877523,
+      "learning_rate": 9.228117169266782e-06,
+      "loss": 0.1706,
+      "step": 2546
+    },
+    {
+      "epoch": 2.8605924470026673,
+      "grad_norm": 0.4577345045610456,
+      "learning_rate": 9.22029663396343e-06,
+      "loss": 0.1731,
+      "step": 2547
+    },
+    {
+      "epoch": 2.8617155692826057,
+      "grad_norm": 0.4593857700548628,
+      "learning_rate": 9.21247657842014e-06,
+      "loss": 0.1711,
+      "step": 2548
+    },
+    {
+      "epoch": 2.862838691562544,
+      "grad_norm": 0.48560219199017984,
+      "learning_rate": 9.204657007448678e-06,
+      "loss": 0.1932,
+      "step": 2549
+    },
+    {
+      "epoch": 2.863961813842482,
+      "grad_norm": 0.4747666905904359,
+      "learning_rate": 9.196837925860516e-06,
+      "loss": 0.1844,
+      "step": 2550
+    },
+    {
+      "epoch": 2.8650849361224204,
+      "grad_norm": 0.4818639171590592,
+      "learning_rate": 9.189019338466812e-06,
+      "loss": 0.1791,
+      "step": 2551
+    },
+    {
+      "epoch": 2.866208058402359,
+      "grad_norm": 0.4643395199305748,
+      "learning_rate": 9.181201250078435e-06,
+      "loss": 0.1704,
+      "step": 2552
+    },
+    {
+      "epoch": 2.8673311806822968,
+      "grad_norm": 0.4611993802683832,
+      "learning_rate": 9.173383665505937e-06,
+      "loss": 0.1705,
+      "step": 2553
+    },
+    {
+      "epoch": 2.868454302962235,
+      "grad_norm": 0.47676505658180535,
+      "learning_rate": 9.165566589559568e-06,
+      "loss": 0.1831,
+      "step": 2554
+    },
+    {
+      "epoch": 2.8695774252421735,
+      "grad_norm": 0.4674787023765812,
+      "learning_rate": 9.15775002704926e-06,
+      "loss": 0.1632,
+      "step": 2555
+    },
+    {
+      "epoch": 2.8707005475221115,
+      "grad_norm": 0.4905453400165015,
+      "learning_rate": 9.14993398278463e-06,
+      "loss": 0.1819,
+      "step": 2556
+    },
+    {
+      "epoch": 2.87182366980205,
+      "grad_norm": 0.47733632103578205,
+      "learning_rate": 9.142118461574971e-06,
+      "loss": 0.1792,
+      "step": 2557
+    },
+    {
+      "epoch": 2.8729467920819878,
+      "grad_norm": 0.4558817946417188,
+      "learning_rate": 9.134303468229264e-06,
+      "loss": 0.1792,
+      "step": 2558
+    },
+    {
+      "epoch": 2.874069914361926,
+      "grad_norm": 0.47700983129939084,
+      "learning_rate": 9.12648900755616e-06,
+      "loss": 0.182,
+      "step": 2559
+    },
+    {
+      "epoch": 2.8751930366418645,
+      "grad_norm": 0.4553018015995702,
+      "learning_rate": 9.118675084363986e-06,
+      "loss": 0.1667,
+      "step": 2560
+    },
+    {
+      "epoch": 2.8763161589218025,
+      "grad_norm": 0.4593598102751128,
+      "learning_rate": 9.110861703460727e-06,
+      "loss": 0.1728,
+      "step": 2561
+    },
+    {
+      "epoch": 2.877439281201741,
+      "grad_norm": 0.4648771153530072,
+      "learning_rate": 9.103048869654047e-06,
+      "loss": 0.1728,
+      "step": 2562
+    },
+    {
+      "epoch": 2.878562403481679,
+      "grad_norm": 0.436076180749989,
+      "learning_rate": 9.095236587751267e-06,
+      "loss": 0.1615,
+      "step": 2563
+    },
+    {
+      "epoch": 2.879685525761617,
+      "grad_norm": 0.4734747981831071,
+      "learning_rate": 9.08742486255937e-06,
+      "loss": 0.1802,
+      "step": 2564
+    },
+    {
+      "epoch": 2.8808086480415556,
+      "grad_norm": 0.4877652162687277,
+      "learning_rate": 9.079613698885002e-06,
+      "loss": 0.1869,
+      "step": 2565
+    },
+    {
+      "epoch": 2.8819317703214935,
+      "grad_norm": 0.45765721191013503,
+      "learning_rate": 9.071803101534451e-06,
+      "loss": 0.1755,
+      "step": 2566
+    },
+    {
+      "epoch": 2.883054892601432,
+      "grad_norm": 0.45021196688829096,
+      "learning_rate": 9.063993075313666e-06,
+      "loss": 0.1727,
+      "step": 2567
+    },
+    {
+      "epoch": 2.8841780148813703,
+      "grad_norm": 0.4522674069841106,
+      "learning_rate": 9.056183625028243e-06,
+      "loss": 0.166,
+      "step": 2568
+    },
+    {
+      "epoch": 2.885301137161308,
+      "grad_norm": 0.46231761385594666,
+      "learning_rate": 9.04837475548342e-06,
+      "loss": 0.1808,
+      "step": 2569
+    },
+    {
+      "epoch": 2.8864242594412466,
+      "grad_norm": 0.47474127509623437,
+      "learning_rate": 9.040566471484085e-06,
+      "loss": 0.183,
+      "step": 2570
+    },
+    {
+      "epoch": 2.887547381721185,
+      "grad_norm": 0.45900525335603964,
+      "learning_rate": 9.032758777834754e-06,
+      "loss": 0.1737,
+      "step": 2571
+    },
+    {
+      "epoch": 2.888670504001123,
+      "grad_norm": 0.45812880138007356,
+      "learning_rate": 9.024951679339594e-06,
+      "loss": 0.1677,
+      "step": 2572
+    },
+    {
+      "epoch": 2.8897936262810613,
+      "grad_norm": 0.47586663132708334,
+      "learning_rate": 9.017145180802393e-06,
+      "loss": 0.1834,
+      "step": 2573
+    },
+    {
+      "epoch": 2.8909167485609997,
+      "grad_norm": 0.4116685169052402,
+      "learning_rate": 9.00933928702658e-06,
+      "loss": 0.1496,
+      "step": 2574
+    },
+    {
+      "epoch": 2.8920398708409376,
+      "grad_norm": 0.46861344574621616,
+      "learning_rate": 9.001534002815209e-06,
+      "loss": 0.1789,
+      "step": 2575
+    },
+    {
+      "epoch": 2.893162993120876,
+      "grad_norm": 0.46679304982614994,
+      "learning_rate": 8.993729332970948e-06,
+      "loss": 0.1742,
+      "step": 2576
+    },
+    {
+      "epoch": 2.8942861154008144,
+      "grad_norm": 0.4580429287886535,
+      "learning_rate": 8.985925282296105e-06,
+      "loss": 0.1838,
+      "step": 2577
+    },
+    {
+      "epoch": 2.8954092376807523,
+      "grad_norm": 0.4569743853099299,
+      "learning_rate": 8.978121855592593e-06,
+      "loss": 0.1761,
+      "step": 2578
+    },
+    {
+      "epoch": 2.8965323599606907,
+      "grad_norm": 0.4743176355550856,
+      "learning_rate": 8.970319057661954e-06,
+      "loss": 0.1831,
+      "step": 2579
+    },
+    {
+      "epoch": 2.897655482240629,
+      "grad_norm": 0.4553016611576038,
+      "learning_rate": 8.962516893305324e-06,
+      "loss": 0.1692,
+      "step": 2580
+    },
+    {
+      "epoch": 2.898778604520567,
+      "grad_norm": 0.45473502109690267,
+      "learning_rate": 8.954715367323468e-06,
+      "loss": 0.1678,
+      "step": 2581
+    },
+    {
+      "epoch": 2.8999017268005054,
+      "grad_norm": 0.4807942668241952,
+      "learning_rate": 8.946914484516748e-06,
+      "loss": 0.1807,
+      "step": 2582
+    },
+    {
+      "epoch": 2.901024849080444,
+      "grad_norm": 0.429941902019654,
+      "learning_rate": 8.939114249685135e-06,
+      "loss": 0.1516,
+      "step": 2583
+    },
+    {
+      "epoch": 2.9021479713603817,
+      "grad_norm": 0.4653157438770528,
+      "learning_rate": 8.931314667628201e-06,
+      "loss": 0.1767,
+      "step": 2584
+    },
+    {
+      "epoch": 2.90327109364032,
+      "grad_norm": 0.44312625971069464,
+      "learning_rate": 8.923515743145113e-06,
+      "loss": 0.1697,
+      "step": 2585
+    },
+    {
+      "epoch": 2.9043942159202585,
+      "grad_norm": 0.43811033758732193,
+      "learning_rate": 8.915717481034632e-06,
+      "loss": 0.1604,
+      "step": 2586
+    },
+    {
+      "epoch": 2.9055173382001964,
+      "grad_norm": 0.47216339814248476,
+      "learning_rate": 8.907919886095115e-06,
+      "loss": 0.1731,
+      "step": 2587
+    },
+    {
+      "epoch": 2.906640460480135,
+      "grad_norm": 0.4621783530580539,
+      "learning_rate": 8.900122963124513e-06,
+      "loss": 0.1759,
+      "step": 2588
+    },
+    {
+      "epoch": 2.907763582760073,
+      "grad_norm": 0.45497514323875493,
+      "learning_rate": 8.892326716920356e-06,
+      "loss": 0.1651,
+      "step": 2589
+    },
+    {
+      "epoch": 2.908886705040011,
+      "grad_norm": 0.457851254832388,
+      "learning_rate": 8.884531152279757e-06,
+      "loss": 0.1729,
+      "step": 2590
+    },
+    {
+      "epoch": 2.9100098273199495,
+      "grad_norm": 0.45822650406750187,
+      "learning_rate": 8.876736273999415e-06,
+      "loss": 0.1693,
+      "step": 2591
+    },
+    {
+      "epoch": 2.911132949599888,
+      "grad_norm": 0.46856109234379667,
+      "learning_rate": 8.868942086875605e-06,
+      "loss": 0.1848,
+      "step": 2592
+    },
+    {
+      "epoch": 2.912256071879826,
+      "grad_norm": 0.4576565366829723,
+      "learning_rate": 8.861148595704176e-06,
+      "loss": 0.1731,
+      "step": 2593
+    },
+    {
+      "epoch": 2.913379194159764,
+      "grad_norm": 0.4523806979663706,
+      "learning_rate": 8.853355805280553e-06,
+      "loss": 0.1744,
+      "step": 2594
+    },
+    {
+      "epoch": 2.9145023164397026,
+      "grad_norm": 0.45200684428131077,
+      "learning_rate": 8.845563720399715e-06,
+      "loss": 0.173,
+      "step": 2595
+    },
+    {
+      "epoch": 2.9156254387196405,
+      "grad_norm": 0.47579624922930075,
+      "learning_rate": 8.837772345856226e-06,
+      "loss": 0.1917,
+      "step": 2596
+    },
+    {
+      "epoch": 2.916748560999579,
+      "grad_norm": 0.45572922351885753,
+      "learning_rate": 8.829981686444201e-06,
+      "loss": 0.174,
+      "step": 2597
+    },
+    {
+      "epoch": 2.9178716832795173,
+      "grad_norm": 0.4638260376523015,
+      "learning_rate": 8.822191746957321e-06,
+      "loss": 0.1868,
+      "step": 2598
+    },
+    {
+      "epoch": 2.9189948055594552,
+      "grad_norm": 0.45148558403348366,
+      "learning_rate": 8.814402532188824e-06,
+      "loss": 0.1717,
+      "step": 2599
+    },
+    {
+      "epoch": 2.9201179278393936,
+      "grad_norm": 0.4573379490292585,
+      "learning_rate": 8.806614046931491e-06,
+      "loss": 0.1713,
+      "step": 2600
+    },
+    {
+      "epoch": 2.921241050119332,
+      "grad_norm": 0.4767926454940618,
+      "learning_rate": 8.79882629597767e-06,
+      "loss": 0.1763,
+      "step": 2601
+    },
+    {
+      "epoch": 2.92236417239927,
+      "grad_norm": 0.48398291984858927,
+      "learning_rate": 8.791039284119244e-06,
+      "loss": 0.1898,
+      "step": 2602
+    },
+    {
+      "epoch": 2.9234872946792083,
+      "grad_norm": 0.4645986763409629,
+      "learning_rate": 8.783253016147652e-06,
+      "loss": 0.1704,
+      "step": 2603
+    },
+    {
+      "epoch": 2.9246104169591467,
+      "grad_norm": 0.5308394707341152,
+      "learning_rate": 8.775467496853873e-06,
+      "loss": 0.191,
+      "step": 2604
+    },
+    {
+      "epoch": 2.9257335392390846,
+      "grad_norm": 0.4670451326042578,
+      "learning_rate": 8.767682731028415e-06,
+      "loss": 0.1711,
+      "step": 2605
+    },
+    {
+      "epoch": 2.926856661519023,
+      "grad_norm": 0.438698398612858,
+      "learning_rate": 8.759898723461333e-06,
+      "loss": 0.1644,
+      "step": 2606
+    },
+    {
+      "epoch": 2.927979783798961,
+      "grad_norm": 0.44189027854801244,
+      "learning_rate": 8.752115478942213e-06,
+      "loss": 0.1649,
+      "step": 2607
+    },
+    {
+      "epoch": 2.9291029060788993,
+      "grad_norm": 0.4581859363210317,
+      "learning_rate": 8.744333002260172e-06,
+      "loss": 0.1708,
+      "step": 2608
+    },
+    {
+      "epoch": 2.9302260283588377,
+      "grad_norm": 0.45676884931636114,
+      "learning_rate": 8.736551298203854e-06,
+      "loss": 0.1742,
+      "step": 2609
+    },
+    {
+      "epoch": 2.9313491506387757,
+      "grad_norm": 0.4497025780421319,
+      "learning_rate": 8.728770371561424e-06,
+      "loss": 0.1651,
+      "step": 2610
+    },
+    {
+      "epoch": 2.932472272918714,
+      "grad_norm": 0.4781813885457078,
+      "learning_rate": 8.720990227120575e-06,
+      "loss": 0.1815,
+      "step": 2611
+    },
+    {
+      "epoch": 2.933595395198652,
+      "grad_norm": 0.46022287061474587,
+      "learning_rate": 8.71321086966851e-06,
+      "loss": 0.1807,
+      "step": 2612
+    },
+    {
+      "epoch": 2.9347185174785904,
+      "grad_norm": 0.48446443471996553,
+      "learning_rate": 8.705432303991967e-06,
+      "loss": 0.1915,
+      "step": 2613
+    },
+    {
+      "epoch": 2.9358416397585287,
+      "grad_norm": 0.46344897666541895,
+      "learning_rate": 8.697654534877166e-06,
+      "loss": 0.1857,
+      "step": 2614
+    },
+    {
+      "epoch": 2.9369647620384667,
+      "grad_norm": 0.4425948152166924,
+      "learning_rate": 8.689877567109861e-06,
+      "loss": 0.1724,
+      "step": 2615
+    },
+    {
+      "epoch": 2.938087884318405,
+      "grad_norm": 0.44998509556224764,
+      "learning_rate": 8.682101405475308e-06,
+      "loss": 0.1718,
+      "step": 2616
+    },
+    {
+      "epoch": 2.9392110065983434,
+      "grad_norm": 0.4675922727964663,
+      "learning_rate": 8.674326054758261e-06,
+      "loss": 0.1771,
+      "step": 2617
+    },
+    {
+      "epoch": 2.9403341288782814,
+      "grad_norm": 0.47264154820508586,
+      "learning_rate": 8.666551519742988e-06,
+      "loss": 0.1816,
+      "step": 2618
+    },
+    {
+      "epoch": 2.9414572511582198,
+      "grad_norm": 0.46063885171682406,
+      "learning_rate": 8.658777805213233e-06,
+      "loss": 0.1744,
+      "step": 2619
+    },
+    {
+      "epoch": 2.942580373438158,
+      "grad_norm": 0.4568310611912791,
+      "learning_rate": 8.651004915952252e-06,
+      "loss": 0.1779,
+      "step": 2620
+    },
+    {
+      "epoch": 2.943703495718096,
+      "grad_norm": 0.48101099147671783,
+      "learning_rate": 8.643232856742794e-06,
+      "loss": 0.1912,
+      "step": 2621
+    },
+    {
+      "epoch": 2.9448266179980345,
+      "grad_norm": 0.46646815088946425,
+      "learning_rate": 8.635461632367087e-06,
+      "loss": 0.1859,
+      "step": 2622
+    },
+    {
+      "epoch": 2.945949740277973,
+      "grad_norm": 0.4601186273096568,
+      "learning_rate": 8.627691247606862e-06,
+      "loss": 0.1649,
+      "step": 2623
+    },
+    {
+      "epoch": 2.947072862557911,
+      "grad_norm": 0.6344634698370505,
+      "learning_rate": 8.619921707243308e-06,
+      "loss": 0.1859,
+      "step": 2624
+    },
+    {
+      "epoch": 2.948195984837849,
+      "grad_norm": 0.4461101591994598,
+      "learning_rate": 8.612153016057114e-06,
+      "loss": 0.1586,
+      "step": 2625
+    },
+    {
+      "epoch": 2.9493191071177876,
+      "grad_norm": 0.43938035234144796,
+      "learning_rate": 8.604385178828441e-06,
+      "loss": 0.165,
+      "step": 2626
+    },
+    {
+      "epoch": 2.9504422293977255,
+      "grad_norm": 0.44943044279325145,
+      "learning_rate": 8.596618200336925e-06,
+      "loss": 0.1764,
+      "step": 2627
+    },
+    {
+      "epoch": 2.951565351677664,
+      "grad_norm": 0.4746096642980948,
+      "learning_rate": 8.588852085361679e-06,
+      "loss": 0.1748,
+      "step": 2628
+    },
+    {
+      "epoch": 2.9526884739576023,
+      "grad_norm": 0.46112492600966043,
+      "learning_rate": 8.58108683868127e-06,
+      "loss": 0.1707,
+      "step": 2629
+    },
+    {
+      "epoch": 2.95381159623754,
+      "grad_norm": 0.45177789162071297,
+      "learning_rate": 8.573322465073746e-06,
+      "loss": 0.1726,
+      "step": 2630
+    },
+    {
+      "epoch": 2.9549347185174786,
+      "grad_norm": 0.450023349792469,
+      "learning_rate": 8.565558969316607e-06,
+      "loss": 0.1728,
+      "step": 2631
+    },
+    {
+      "epoch": 2.956057840797417,
+      "grad_norm": 0.4770351172578664,
+      "learning_rate": 8.557796356186818e-06,
+      "loss": 0.1877,
+      "step": 2632
+    },
+    {
+      "epoch": 2.957180963077355,
+      "grad_norm": 0.4513830973776379,
+      "learning_rate": 8.550034630460806e-06,
+      "loss": 0.1685,
+      "step": 2633
+    },
+    {
+      "epoch": 2.9583040853572933,
+      "grad_norm": 0.4857445983384114,
+      "learning_rate": 8.542273796914439e-06,
+      "loss": 0.1718,
+      "step": 2634
+    },
+    {
+      "epoch": 2.9594272076372317,
+      "grad_norm": 0.47068053305099244,
+      "learning_rate": 8.534513860323047e-06,
+      "loss": 0.1702,
+      "step": 2635
+    },
+    {
+      "epoch": 2.9605503299171696,
+      "grad_norm": 0.47923791043815955,
+      "learning_rate": 8.526754825461402e-06,
+      "loss": 0.1811,
+      "step": 2636
+    },
+    {
+      "epoch": 2.961673452197108,
+      "grad_norm": 0.4548065001435798,
+      "learning_rate": 8.518996697103726e-06,
+      "loss": 0.1756,
+      "step": 2637
+    },
+    {
+      "epoch": 2.9627965744770464,
+      "grad_norm": 0.4460233301614257,
+      "learning_rate": 8.511239480023686e-06,
+      "loss": 0.1625,
+      "step": 2638
+    },
+    {
+      "epoch": 2.9639196967569843,
+      "grad_norm": 0.4548163435033796,
+      "learning_rate": 8.50348317899437e-06,
+      "loss": 0.1658,
+      "step": 2639
+    },
+    {
+      "epoch": 2.9650428190369227,
+      "grad_norm": 0.4655678091479219,
+      "learning_rate": 8.495727798788323e-06,
+      "loss": 0.1683,
+      "step": 2640
+    },
+    {
+      "epoch": 2.966165941316861,
+      "grad_norm": 0.4680379062633164,
+      "learning_rate": 8.487973344177517e-06,
+      "loss": 0.1869,
+      "step": 2641
+    },
+    {
+      "epoch": 2.967289063596799,
+      "grad_norm": 0.44275214343881863,
+      "learning_rate": 8.48021981993335e-06,
+      "loss": 0.1735,
+      "step": 2642
+    },
+    {
+      "epoch": 2.9684121858767374,
+      "grad_norm": 0.4393359117417517,
+      "learning_rate": 8.472467230826656e-06,
+      "loss": 0.1596,
+      "step": 2643
+    },
+    {
+      "epoch": 2.9695353081566758,
+      "grad_norm": 0.4738828812631789,
+      "learning_rate": 8.464715581627682e-06,
+      "loss": 0.176,
+      "step": 2644
+    },
+    {
+      "epoch": 2.9706584304366137,
+      "grad_norm": 0.4573620252964248,
+      "learning_rate": 8.456964877106104e-06,
+      "loss": 0.1727,
+      "step": 2645
+    },
+    {
+      "epoch": 2.971781552716552,
+      "grad_norm": 0.46158734380298694,
+      "learning_rate": 8.44921512203102e-06,
+      "loss": 0.1759,
+      "step": 2646
+    },
+    {
+      "epoch": 2.9729046749964905,
+      "grad_norm": 0.4627319651256285,
+      "learning_rate": 8.441466321170935e-06,
+      "loss": 0.1755,
+      "step": 2647
+    },
+    {
+      "epoch": 2.9740277972764284,
+      "grad_norm": 0.45445850820324185,
+      "learning_rate": 8.433718479293777e-06,
+      "loss": 0.1807,
+      "step": 2648
+    },
+    {
+      "epoch": 2.975150919556367,
+      "grad_norm": 0.4428450885299262,
+      "learning_rate": 8.425971601166872e-06,
+      "loss": 0.1684,
+      "step": 2649
+    },
+    {
+      "epoch": 2.976274041836305,
+      "grad_norm": 0.4541319747937619,
+      "learning_rate": 8.418225691556962e-06,
+      "loss": 0.1805,
+      "step": 2650
+    },
+    {
+      "epoch": 2.977397164116243,
+      "grad_norm": 0.4573856493833637,
+      "learning_rate": 8.41048075523019e-06,
+      "loss": 0.1742,
+      "step": 2651
+    },
+    {
+      "epoch": 2.9785202863961815,
+      "grad_norm": 0.4574620336290775,
+      "learning_rate": 8.402736796952104e-06,
+      "loss": 0.1684,
+      "step": 2652
+    },
+    {
+      "epoch": 2.97964340867612,
+      "grad_norm": 0.4514989212744961,
+      "learning_rate": 8.39499382148764e-06,
+      "loss": 0.1741,
+      "step": 2653
+    },
+    {
+      "epoch": 2.980766530956058,
+      "grad_norm": 0.4559456519447681,
+      "learning_rate": 8.387251833601142e-06,
+      "loss": 0.1761,
+      "step": 2654
+    },
+    {
+      "epoch": 2.981889653235996,
+      "grad_norm": 0.45131114670300065,
+      "learning_rate": 8.379510838056338e-06,
+      "loss": 0.1718,
+      "step": 2655
+    },
+    {
+      "epoch": 2.983012775515934,
+      "grad_norm": 0.44966905961207143,
+      "learning_rate": 8.371770839616348e-06,
+      "loss": 0.1717,
+      "step": 2656
+    },
+    {
+      "epoch": 2.9841358977958725,
+      "grad_norm": 0.46120801970633646,
+      "learning_rate": 8.364031843043683e-06,
+      "loss": 0.1678,
+      "step": 2657
+    },
+    {
+      "epoch": 2.985259020075811,
+      "grad_norm": 0.4421962165877643,
+      "learning_rate": 8.356293853100223e-06,
+      "loss": 0.1606,
+      "step": 2658
+    },
+    {
+      "epoch": 2.986382142355749,
+      "grad_norm": 0.46999315292829197,
+      "learning_rate": 8.348556874547242e-06,
+      "loss": 0.181,
+      "step": 2659
+    },
+    {
+      "epoch": 2.987505264635687,
+      "grad_norm": 0.4506186892281384,
+      "learning_rate": 8.340820912145391e-06,
+      "loss": 0.1712,
+      "step": 2660
+    },
+    {
+      "epoch": 2.9886283869156256,
+      "grad_norm": 0.4626508396110205,
+      "learning_rate": 8.333085970654691e-06,
+      "loss": 0.1716,
+      "step": 2661
+    },
+    {
+      "epoch": 2.9897515091955635,
+      "grad_norm": 0.45107479921301236,
+      "learning_rate": 8.325352054834542e-06,
+      "loss": 0.1762,
+      "step": 2662
+    },
+    {
+      "epoch": 2.990874631475502,
+      "grad_norm": 0.4595705443642629,
+      "learning_rate": 8.317619169443696e-06,
+      "loss": 0.1746,
+      "step": 2663
+    },
+    {
+      "epoch": 2.99199775375544,
+      "grad_norm": 0.4694949935187253,
+      "learning_rate": 8.309887319240291e-06,
+      "loss": 0.1747,
+      "step": 2664
+    },
+    {
+      "epoch": 2.9931208760353782,
+      "grad_norm": 0.4822701328030329,
+      "learning_rate": 8.302156508981816e-06,
+      "loss": 0.177,
+      "step": 2665
+    },
+    {
+      "epoch": 2.9942439983153166,
+      "grad_norm": 0.4623869423886324,
+      "learning_rate": 8.294426743425125e-06,
+      "loss": 0.1653,
+      "step": 2666
+    },
+    {
+      "epoch": 2.9953671205952546,
+      "grad_norm": 0.48059683012598114,
+      "learning_rate": 8.286698027326432e-06,
+      "loss": 0.1763,
+      "step": 2667
+    },
+    {
+      "epoch": 2.996490242875193,
+      "grad_norm": 0.4600386216074317,
+      "learning_rate": 8.278970365441292e-06,
+      "loss": 0.1671,
+      "step": 2668
+    },
+    {
+      "epoch": 2.9976133651551313,
+      "grad_norm": 0.48001415671174696,
+      "learning_rate": 8.271243762524627e-06,
+      "loss": 0.1731,
+      "step": 2669
+    },
+    {
+      "epoch": 2.9987364874350693,
+      "grad_norm": 0.4566979760209381,
+      "learning_rate": 8.263518223330698e-06,
+      "loss": 0.1827,
+      "step": 2670
+    },
+    {
+      "epoch": 2.9998596097150076,
+      "grad_norm": 0.7596103631516063,
+      "learning_rate": 8.255793752613115e-06,
+      "loss": 0.2575,
+      "step": 2671
+    },
+    {
+      "epoch": 3.000982731994946,
+      "grad_norm": 0.6259154790515346,
+      "learning_rate": 8.248070355124832e-06,
+      "loss": 0.1158,
+      "step": 2672
+    },
+    {
+      "epoch": 3.002105854274884,
+      "grad_norm": 0.49503499354650105,
+      "learning_rate": 8.240348035618138e-06,
+      "loss": 0.0951,
+      "step": 2673
+    },
+    {
+      "epoch": 3.0032289765548223,
+      "grad_norm": 0.4242302786603668,
+      "learning_rate": 8.232626798844661e-06,
+      "loss": 0.0806,
+      "step": 2674
+    },
+    {
+      "epoch": 3.0043520988347607,
+      "grad_norm": 0.4269038054152872,
+      "learning_rate": 8.224906649555365e-06,
+      "loss": 0.0878,
+      "step": 2675
+    },
+    {
+      "epoch": 3.0054752211146987,
+      "grad_norm": 0.4530862959158299,
+      "learning_rate": 8.21718759250054e-06,
+      "loss": 0.0855,
+      "step": 2676
+    },
+    {
+      "epoch": 3.006598343394637,
+      "grad_norm": 0.4886548006885315,
+      "learning_rate": 8.209469632429811e-06,
+      "loss": 0.086,
+      "step": 2677
+    },
+    {
+      "epoch": 3.0077214656745754,
+      "grad_norm": 0.554545304546421,
+      "learning_rate": 8.201752774092118e-06,
+      "loss": 0.0994,
+      "step": 2678
+    },
+    {
+      "epoch": 3.0088445879545134,
+      "grad_norm": 0.5854424397857068,
+      "learning_rate": 8.194037022235732e-06,
+      "loss": 0.0839,
+      "step": 2679
+    },
+    {
+      "epoch": 3.0099677102344518,
+      "grad_norm": 0.5920483383090562,
+      "learning_rate": 8.18632238160824e-06,
+      "loss": 0.0873,
+      "step": 2680
+    },
+    {
+      "epoch": 3.01109083251439,
+      "grad_norm": 0.5324726059263992,
+      "learning_rate": 8.178608856956547e-06,
+      "loss": 0.0843,
+      "step": 2681
+    },
+    {
+      "epoch": 3.012213954794328,
+      "grad_norm": 0.4991909599622551,
+      "learning_rate": 8.17089645302687e-06,
+      "loss": 0.087,
+      "step": 2682
+    },
+    {
+      "epoch": 3.0133370770742665,
+      "grad_norm": 0.4832948640805594,
+      "learning_rate": 8.163185174564731e-06,
+      "loss": 0.0857,
+      "step": 2683
+    },
+    {
+      "epoch": 3.014460199354205,
+      "grad_norm": 0.4362830353220936,
+      "learning_rate": 8.155475026314966e-06,
+      "loss": 0.0814,
+      "step": 2684
+    },
+    {
+      "epoch": 3.0155833216341428,
+      "grad_norm": 0.424359667514001,
+      "learning_rate": 8.147766013021716e-06,
+      "loss": 0.0829,
+      "step": 2685
+    },
+    {
+      "epoch": 3.016706443914081,
+      "grad_norm": 0.41618233800829574,
+      "learning_rate": 8.140058139428425e-06,
+      "loss": 0.0832,
+      "step": 2686
+    },
+    {
+      "epoch": 3.0178295661940195,
+      "grad_norm": 0.40606151522127315,
+      "learning_rate": 8.132351410277824e-06,
+      "loss": 0.0858,
+      "step": 2687
+    },
+    {
+      "epoch": 3.0189526884739575,
+      "grad_norm": 0.42085251727822387,
+      "learning_rate": 8.124645830311954e-06,
+      "loss": 0.0868,
+      "step": 2688
+    },
+    {
+      "epoch": 3.020075810753896,
+      "grad_norm": 0.43662806253277353,
+      "learning_rate": 8.116941404272142e-06,
+      "loss": 0.0884,
+      "step": 2689
+    },
+    {
+      "epoch": 3.0211989330338342,
+      "grad_norm": 0.4174983822990382,
+      "learning_rate": 8.109238136899004e-06,
+      "loss": 0.0849,
+      "step": 2690
+    },
+    {
+      "epoch": 3.022322055313772,
+      "grad_norm": 0.41649346119332137,
+      "learning_rate": 8.101536032932452e-06,
+      "loss": 0.0791,
+      "step": 2691
+    },
+    {
+      "epoch": 3.0234451775937106,
+      "grad_norm": 0.4363115097166906,
+      "learning_rate": 8.093835097111668e-06,
+      "loss": 0.0842,
+      "step": 2692
+    },
+    {
+      "epoch": 3.024568299873649,
+      "grad_norm": 0.4507250434052767,
+      "learning_rate": 8.086135334175126e-06,
+      "loss": 0.0809,
+      "step": 2693
+    },
+    {
+      "epoch": 3.025691422153587,
+      "grad_norm": 0.46412219030100466,
+      "learning_rate": 8.078436748860572e-06,
+      "loss": 0.0808,
+      "step": 2694
+    },
+    {
+      "epoch": 3.0268145444335253,
+      "grad_norm": 0.46817713622059637,
+      "learning_rate": 8.070739345905032e-06,
+      "loss": 0.082,
+      "step": 2695
+    },
+    {
+      "epoch": 3.0279376667134636,
+      "grad_norm": 0.4359811445564712,
+      "learning_rate": 8.063043130044806e-06,
+      "loss": 0.0816,
+      "step": 2696
+    },
+    {
+      "epoch": 3.0290607889934016,
+      "grad_norm": 0.44282045477705045,
+      "learning_rate": 8.055348106015455e-06,
+      "loss": 0.0738,
+      "step": 2697
+    },
+    {
+      "epoch": 3.03018391127334,
+      "grad_norm": 0.4687677335027077,
+      "learning_rate": 8.047654278551814e-06,
+      "loss": 0.083,
+      "step": 2698
+    },
+    {
+      "epoch": 3.031307033553278,
+      "grad_norm": 0.45388563511970487,
+      "learning_rate": 8.03996165238798e-06,
+      "loss": 0.0811,
+      "step": 2699
+    },
+    {
+      "epoch": 3.0324301558332163,
+      "grad_norm": 0.4511628549306528,
+      "learning_rate": 8.032270232257312e-06,
+      "loss": 0.0811,
+      "step": 2700
+    },
+    {
+      "epoch": 3.0335532781131547,
+      "grad_norm": 0.4397177997054855,
+      "learning_rate": 8.024580022892427e-06,
+      "loss": 0.0836,
+      "step": 2701
+    },
+    {
+      "epoch": 3.0346764003930926,
+      "grad_norm": 0.43549667655174495,
+      "learning_rate": 8.01689102902519e-06,
+      "loss": 0.0772,
+      "step": 2702
+    },
+    {
+      "epoch": 3.035799522673031,
+      "grad_norm": 0.4197204578796909,
+      "learning_rate": 8.009203255386723e-06,
+      "loss": 0.0783,
+      "step": 2703
+    },
+    {
+      "epoch": 3.0369226449529694,
+      "grad_norm": 0.42838286805922127,
+      "learning_rate": 8.001516706707401e-06,
+      "loss": 0.0821,
+      "step": 2704
+    },
+    {
+      "epoch": 3.0380457672329073,
+      "grad_norm": 0.404870980751977,
+      "learning_rate": 7.993831387716844e-06,
+      "loss": 0.077,
+      "step": 2705
+    },
+    {
+      "epoch": 3.0391688895128457,
+      "grad_norm": 0.4167100368716409,
+      "learning_rate": 7.986147303143913e-06,
+      "loss": 0.0761,
+      "step": 2706
+    },
+    {
+      "epoch": 3.040292011792784,
+      "grad_norm": 0.40996511641383154,
+      "learning_rate": 7.978464457716704e-06,
+      "loss": 0.0802,
+      "step": 2707
+    },
+    {
+      "epoch": 3.041415134072722,
+      "grad_norm": 0.44035804475350526,
+      "learning_rate": 7.97078285616256e-06,
+      "loss": 0.0787,
+      "step": 2708
+    },
+    {
+      "epoch": 3.0425382563526604,
+      "grad_norm": 0.43789822457853483,
+      "learning_rate": 7.963102503208058e-06,
+      "loss": 0.0827,
+      "step": 2709
+    },
+    {
+      "epoch": 3.043661378632599,
+      "grad_norm": 0.4463729892875018,
+      "learning_rate": 7.955423403578998e-06,
+      "loss": 0.081,
+      "step": 2710
+    },
+    {
+      "epoch": 3.0447845009125367,
+      "grad_norm": 0.4593657665400022,
+      "learning_rate": 7.947745562000421e-06,
+      "loss": 0.0839,
+      "step": 2711
+    },
+    {
+      "epoch": 3.045907623192475,
+      "grad_norm": 0.4426263669678776,
+      "learning_rate": 7.940068983196581e-06,
+      "loss": 0.0793,
+      "step": 2712
+    },
+    {
+      "epoch": 3.0470307454724135,
+      "grad_norm": 0.4546022445193033,
+      "learning_rate": 7.932393671890965e-06,
+      "loss": 0.0823,
+      "step": 2713
+    },
+    {
+      "epoch": 3.0481538677523514,
+      "grad_norm": 0.46071246386796366,
+      "learning_rate": 7.924719632806274e-06,
+      "loss": 0.0843,
+      "step": 2714
+    },
+    {
+      "epoch": 3.04927699003229,
+      "grad_norm": 0.43222793934163595,
+      "learning_rate": 7.917046870664431e-06,
+      "loss": 0.0818,
+      "step": 2715
+    },
+    {
+      "epoch": 3.050400112312228,
+      "grad_norm": 0.4468470825780685,
+      "learning_rate": 7.909375390186572e-06,
+      "loss": 0.0812,
+      "step": 2716
+    },
+    {
+      "epoch": 3.051523234592166,
+      "grad_norm": 0.4288047135096686,
+      "learning_rate": 7.90170519609304e-06,
+      "loss": 0.0816,
+      "step": 2717
+    },
+    {
+      "epoch": 3.0526463568721045,
+      "grad_norm": 0.445145856226943,
+      "learning_rate": 7.894036293103393e-06,
+      "loss": 0.0844,
+      "step": 2718
+    },
+    {
+      "epoch": 3.053769479152043,
+      "grad_norm": 0.4565081459753225,
+      "learning_rate": 7.88636868593639e-06,
+      "loss": 0.0864,
+      "step": 2719
+    },
+    {
+      "epoch": 3.054892601431981,
+      "grad_norm": 0.4515132111689418,
+      "learning_rate": 7.878702379309992e-06,
+      "loss": 0.0874,
+      "step": 2720
+    },
+    {
+      "epoch": 3.056015723711919,
+      "grad_norm": 0.44344069780688466,
+      "learning_rate": 7.871037377941367e-06,
+      "loss": 0.0808,
+      "step": 2721
+    },
+    {
+      "epoch": 3.057138845991857,
+      "grad_norm": 0.448178193139511,
+      "learning_rate": 7.863373686546868e-06,
+      "loss": 0.0855,
+      "step": 2722
+    },
+    {
+      "epoch": 3.0582619682717955,
+      "grad_norm": 0.3940384087377325,
+      "learning_rate": 7.855711309842054e-06,
+      "loss": 0.0763,
+      "step": 2723
+    },
+    {
+      "epoch": 3.059385090551734,
+      "grad_norm": 0.45681289631612565,
+      "learning_rate": 7.848050252541666e-06,
+      "loss": 0.0845,
+      "step": 2724
+    },
+    {
+      "epoch": 3.060508212831672,
+      "grad_norm": 0.4362986797057327,
+      "learning_rate": 7.840390519359644e-06,
+      "loss": 0.0772,
+      "step": 2725
+    },
+    {
+      "epoch": 3.0616313351116102,
+      "grad_norm": 0.43138388625168317,
+      "learning_rate": 7.832732115009096e-06,
+      "loss": 0.0817,
+      "step": 2726
+    },
+    {
+      "epoch": 3.0627544573915486,
+      "grad_norm": 0.4242754319993129,
+      "learning_rate": 7.825075044202329e-06,
+      "loss": 0.079,
+      "step": 2727
+    },
+    {
+      "epoch": 3.0638775796714866,
+      "grad_norm": 0.4371286146302931,
+      "learning_rate": 7.817419311650819e-06,
+      "loss": 0.0835,
+      "step": 2728
+    },
+    {
+      "epoch": 3.065000701951425,
+      "grad_norm": 0.42440135053295946,
+      "learning_rate": 7.809764922065226e-06,
+      "loss": 0.0774,
+      "step": 2729
+    },
+    {
+      "epoch": 3.0661238242313633,
+      "grad_norm": 0.43591296320616574,
+      "learning_rate": 7.802111880155382e-06,
+      "loss": 0.0812,
+      "step": 2730
+    },
+    {
+      "epoch": 3.0672469465113013,
+      "grad_norm": 0.4264395948747793,
+      "learning_rate": 7.794460190630283e-06,
+      "loss": 0.0785,
+      "step": 2731
+    },
+    {
+      "epoch": 3.0683700687912396,
+      "grad_norm": 0.47067358685640387,
+      "learning_rate": 7.786809858198096e-06,
+      "loss": 0.0835,
+      "step": 2732
+    },
+    {
+      "epoch": 3.069493191071178,
+      "grad_norm": 0.45020780231345264,
+      "learning_rate": 7.779160887566161e-06,
+      "loss": 0.0842,
+      "step": 2733
+    },
+    {
+      "epoch": 3.070616313351116,
+      "grad_norm": 0.4598985349986955,
+      "learning_rate": 7.77151328344097e-06,
+      "loss": 0.0818,
+      "step": 2734
+    },
+    {
+      "epoch": 3.0717394356310543,
+      "grad_norm": 0.44764895115040065,
+      "learning_rate": 7.763867050528184e-06,
+      "loss": 0.0845,
+      "step": 2735
+    },
+    {
+      "epoch": 3.0728625579109927,
+      "grad_norm": 0.45919564184903605,
+      "learning_rate": 7.756222193532606e-06,
+      "loss": 0.078,
+      "step": 2736
+    },
+    {
+      "epoch": 3.0739856801909307,
+      "grad_norm": 0.43399382955613974,
+      "learning_rate": 7.748578717158204e-06,
+      "loss": 0.0796,
+      "step": 2737
+    },
+    {
+      "epoch": 3.075108802470869,
+      "grad_norm": 0.4415661509626846,
+      "learning_rate": 7.74093662610809e-06,
+      "loss": 0.081,
+      "step": 2738
+    },
+    {
+      "epoch": 3.0762319247508074,
+      "grad_norm": 0.4623493322073683,
+      "learning_rate": 7.733295925084534e-06,
+      "loss": 0.0787,
+      "step": 2739
+    },
+    {
+      "epoch": 3.0773550470307454,
+      "grad_norm": 0.42863153572886326,
+      "learning_rate": 7.725656618788938e-06,
+      "loss": 0.0822,
+      "step": 2740
+    },
+    {
+      "epoch": 3.0784781693106837,
+      "grad_norm": 0.4471791832208811,
+      "learning_rate": 7.718018711921852e-06,
+      "loss": 0.0832,
+      "step": 2741
+    },
+    {
+      "epoch": 3.079601291590622,
+      "grad_norm": 0.4478220162373831,
+      "learning_rate": 7.710382209182964e-06,
+      "loss": 0.0874,
+      "step": 2742
+    },
+    {
+      "epoch": 3.08072441387056,
+      "grad_norm": 0.4214162588600572,
+      "learning_rate": 7.702747115271098e-06,
+      "loss": 0.0826,
+      "step": 2743
+    },
+    {
+      "epoch": 3.0818475361504984,
+      "grad_norm": 0.4469938818454545,
+      "learning_rate": 7.695113434884214e-06,
+      "loss": 0.0804,
+      "step": 2744
+    },
+    {
+      "epoch": 3.082970658430437,
+      "grad_norm": 0.46167710102896964,
+      "learning_rate": 7.687481172719402e-06,
+      "loss": 0.0819,
+      "step": 2745
+    },
+    {
+      "epoch": 3.0840937807103748,
+      "grad_norm": 0.4605465891709998,
+      "learning_rate": 7.679850333472867e-06,
+      "loss": 0.0855,
+      "step": 2746
+    },
+    {
+      "epoch": 3.085216902990313,
+      "grad_norm": 0.43521315545514894,
+      "learning_rate": 7.672220921839955e-06,
+      "loss": 0.0736,
+      "step": 2747
+    },
+    {
+      "epoch": 3.086340025270251,
+      "grad_norm": 0.4604041931459674,
+      "learning_rate": 7.664592942515125e-06,
+      "loss": 0.0819,
+      "step": 2748
+    },
+    {
+      "epoch": 3.0874631475501895,
+      "grad_norm": 0.4352802112893712,
+      "learning_rate": 7.656966400191956e-06,
+      "loss": 0.0767,
+      "step": 2749
+    },
+    {
+      "epoch": 3.088586269830128,
+      "grad_norm": 0.43183285295275914,
+      "learning_rate": 7.649341299563151e-06,
+      "loss": 0.0815,
+      "step": 2750
+    },
+    {
+      "epoch": 3.089709392110066,
+      "grad_norm": 0.4398374191229837,
+      "learning_rate": 7.641717645320508e-06,
+      "loss": 0.0826,
+      "step": 2751
+    },
+    {
+      "epoch": 3.090832514390004,
+      "grad_norm": 0.44375949908128276,
+      "learning_rate": 7.634095442154949e-06,
+      "loss": 0.0796,
+      "step": 2752
+    },
+    {
+      "epoch": 3.0919556366699426,
+      "grad_norm": 0.4341552574208931,
+      "learning_rate": 7.626474694756501e-06,
+      "loss": 0.0798,
+      "step": 2753
+    },
+    {
+      "epoch": 3.0930787589498805,
+      "grad_norm": 0.45510308680572736,
+      "learning_rate": 7.6188554078142915e-06,
+      "loss": 0.0818,
+      "step": 2754
+    },
+    {
+      "epoch": 3.094201881229819,
+      "grad_norm": 0.4375905218338109,
+      "learning_rate": 7.611237586016558e-06,
+      "loss": 0.0811,
+      "step": 2755
+    },
+    {
+      "epoch": 3.0953250035097573,
+      "grad_norm": 0.4441137124518761,
+      "learning_rate": 7.60362123405062e-06,
+      "loss": 0.079,
+      "step": 2756
+    },
+    {
+      "epoch": 3.096448125789695,
+      "grad_norm": 0.4502413713509533,
+      "learning_rate": 7.596006356602908e-06,
+      "loss": 0.0797,
+      "step": 2757
+    },
+    {
+      "epoch": 3.0975712480696336,
+      "grad_norm": 0.42048708566816295,
+      "learning_rate": 7.58839295835894e-06,
+      "loss": 0.0769,
+      "step": 2758
+    },
+    {
+      "epoch": 3.098694370349572,
+      "grad_norm": 0.4573125387642322,
+      "learning_rate": 7.580781044003324e-06,
+      "loss": 0.0799,
+      "step": 2759
+    },
+    {
+      "epoch": 3.09981749262951,
+      "grad_norm": 0.506317107529294,
+      "learning_rate": 7.573170618219754e-06,
+      "loss": 0.0858,
+      "step": 2760
+    },
+    {
+      "epoch": 3.1009406149094483,
+      "grad_norm": 0.43658684401047776,
+      "learning_rate": 7.565561685691008e-06,
+      "loss": 0.0797,
+      "step": 2761
+    },
+    {
+      "epoch": 3.1020637371893867,
+      "grad_norm": 0.42157837097318585,
+      "learning_rate": 7.557954251098946e-06,
+      "loss": 0.0785,
+      "step": 2762
+    },
+    {
+      "epoch": 3.1031868594693246,
+      "grad_norm": 0.43891360296769083,
+      "learning_rate": 7.550348319124506e-06,
+      "loss": 0.0801,
+      "step": 2763
+    },
+    {
+      "epoch": 3.104309981749263,
+      "grad_norm": 0.4616891997168113,
+      "learning_rate": 7.5427438944477086e-06,
+      "loss": 0.0794,
+      "step": 2764
+    },
+    {
+      "epoch": 3.1054331040292014,
+      "grad_norm": 0.43768658044133474,
+      "learning_rate": 7.535140981747627e-06,
+      "loss": 0.0789,
+      "step": 2765
+    },
+    {
+      "epoch": 3.1065562263091393,
+      "grad_norm": 0.43116002218960375,
+      "learning_rate": 7.527539585702426e-06,
+      "loss": 0.0766,
+      "step": 2766
+    },
+    {
+      "epoch": 3.1076793485890777,
+      "grad_norm": 0.46226193562781304,
+      "learning_rate": 7.519939710989326e-06,
+      "loss": 0.0861,
+      "step": 2767
+    },
+    {
+      "epoch": 3.108802470869016,
+      "grad_norm": 0.4584385515521631,
+      "learning_rate": 7.512341362284612e-06,
+      "loss": 0.0824,
+      "step": 2768
+    },
+    {
+      "epoch": 3.109925593148954,
+      "grad_norm": 0.42541505578878597,
+      "learning_rate": 7.504744544263639e-06,
+      "loss": 0.0793,
+      "step": 2769
+    },
+    {
+      "epoch": 3.1110487154288924,
+      "grad_norm": 0.44149954225470445,
+      "learning_rate": 7.497149261600803e-06,
+      "loss": 0.0761,
+      "step": 2770
+    },
+    {
+      "epoch": 3.1121718377088303,
+      "grad_norm": 0.4367339709220696,
+      "learning_rate": 7.489555518969568e-06,
+      "loss": 0.0768,
+      "step": 2771
+    },
+    {
+      "epoch": 3.1132949599887687,
+      "grad_norm": 0.44918411334128344,
+      "learning_rate": 7.481963321042449e-06,
+      "loss": 0.0792,
+      "step": 2772
+    },
+    {
+      "epoch": 3.114418082268707,
+      "grad_norm": 0.44367325496090004,
+      "learning_rate": 7.474372672491008e-06,
+      "loss": 0.084,
+      "step": 2773
+    },
+    {
+      "epoch": 3.115541204548645,
+      "grad_norm": 0.4398077189907013,
+      "learning_rate": 7.4667835779858585e-06,
+      "loss": 0.0823,
+      "step": 2774
+    },
+    {
+      "epoch": 3.1166643268285834,
+      "grad_norm": 0.4690551629213953,
+      "learning_rate": 7.459196042196647e-06,
+      "loss": 0.08,
+      "step": 2775
+    },
+    {
+      "epoch": 3.117787449108522,
+      "grad_norm": 0.4473577302083683,
+      "learning_rate": 7.45161006979207e-06,
+      "loss": 0.0799,
+      "step": 2776
+    },
+    {
+      "epoch": 3.1189105713884597,
+      "grad_norm": 0.4388023068250219,
+      "learning_rate": 7.444025665439862e-06,
+      "loss": 0.0795,
+      "step": 2777
+    },
+    {
+      "epoch": 3.120033693668398,
+      "grad_norm": 0.4491722860474616,
+      "learning_rate": 7.43644283380679e-06,
+      "loss": 0.0777,
+      "step": 2778
+    },
+    {
+      "epoch": 3.1211568159483365,
+      "grad_norm": 0.45449915733821294,
+      "learning_rate": 7.428861579558653e-06,
+      "loss": 0.0792,
+      "step": 2779
+    },
+    {
+      "epoch": 3.1222799382282744,
+      "grad_norm": 0.44290763583095166,
+      "learning_rate": 7.42128190736028e-06,
+      "loss": 0.0767,
+      "step": 2780
+    },
+    {
+      "epoch": 3.123403060508213,
+      "grad_norm": 0.46051935717803083,
+      "learning_rate": 7.413703821875526e-06,
+      "loss": 0.0818,
+      "step": 2781
+    },
+    {
+      "epoch": 3.124526182788151,
+      "grad_norm": 0.4400112926339292,
+      "learning_rate": 7.406127327767272e-06,
+      "loss": 0.0804,
+      "step": 2782
+    },
+    {
+      "epoch": 3.125649305068089,
+      "grad_norm": 0.4589606315894283,
+      "learning_rate": 7.398552429697416e-06,
+      "loss": 0.0829,
+      "step": 2783
+    },
+    {
+      "epoch": 3.1267724273480275,
+      "grad_norm": 0.43320506570345907,
+      "learning_rate": 7.390979132326881e-06,
+      "loss": 0.0803,
+      "step": 2784
+    },
+    {
+      "epoch": 3.127895549627966,
+      "grad_norm": 0.43767705767402443,
+      "learning_rate": 7.383407440315595e-06,
+      "loss": 0.0797,
+      "step": 2785
+    },
+    {
+      "epoch": 3.129018671907904,
+      "grad_norm": 0.4551489448501023,
+      "learning_rate": 7.375837358322504e-06,
+      "loss": 0.0803,
+      "step": 2786
+    },
+    {
+      "epoch": 3.130141794187842,
+      "grad_norm": 0.4483555435289511,
+      "learning_rate": 7.368268891005565e-06,
+      "loss": 0.0851,
+      "step": 2787
+    },
+    {
+      "epoch": 3.1312649164677806,
+      "grad_norm": 0.44696608573573154,
+      "learning_rate": 7.360702043021738e-06,
+      "loss": 0.0824,
+      "step": 2788
+    },
+    {
+      "epoch": 3.1323880387477185,
+      "grad_norm": 0.4175887163998629,
+      "learning_rate": 7.353136819026991e-06,
+      "loss": 0.081,
+      "step": 2789
+    },
+    {
+      "epoch": 3.133511161027657,
+      "grad_norm": 0.42858708958351066,
+      "learning_rate": 7.345573223676284e-06,
+      "loss": 0.0793,
+      "step": 2790
+    },
+    {
+      "epoch": 3.1346342833075953,
+      "grad_norm": 0.4315573518809542,
+      "learning_rate": 7.338011261623583e-06,
+      "loss": 0.0749,
+      "step": 2791
+    },
+    {
+      "epoch": 3.1357574055875332,
+      "grad_norm": 0.46061705300883715,
+      "learning_rate": 7.330450937521847e-06,
+      "loss": 0.0802,
+      "step": 2792
+    },
+    {
+      "epoch": 3.1368805278674716,
+      "grad_norm": 0.45499828672626436,
+      "learning_rate": 7.322892256023025e-06,
+      "loss": 0.082,
+      "step": 2793
+    },
+    {
+      "epoch": 3.13800365014741,
+      "grad_norm": 0.45703139441971474,
+      "learning_rate": 7.315335221778064e-06,
+      "loss": 0.0837,
+      "step": 2794
+    },
+    {
+      "epoch": 3.139126772427348,
+      "grad_norm": 0.44422721534446546,
+      "learning_rate": 7.307779839436878e-06,
+      "loss": 0.0789,
+      "step": 2795
+    },
+    {
+      "epoch": 3.1402498947072863,
+      "grad_norm": 0.432747277056686,
+      "learning_rate": 7.300226113648384e-06,
+      "loss": 0.0814,
+      "step": 2796
+    },
+    {
+      "epoch": 3.1413730169872247,
+      "grad_norm": 0.4354763866952526,
+      "learning_rate": 7.292674049060473e-06,
+      "loss": 0.0809,
+      "step": 2797
+    },
+    {
+      "epoch": 3.1424961392671626,
+      "grad_norm": 0.42129390641539777,
+      "learning_rate": 7.285123650320017e-06,
+      "loss": 0.0759,
+      "step": 2798
+    },
+    {
+      "epoch": 3.143619261547101,
+      "grad_norm": 0.4299331690685501,
+      "learning_rate": 7.277574922072847e-06,
+      "loss": 0.0811,
+      "step": 2799
+    },
+    {
+      "epoch": 3.144742383827039,
+      "grad_norm": 0.4224270029033599,
+      "learning_rate": 7.27002786896379e-06,
+      "loss": 0.076,
+      "step": 2800
+    },
+    {
+      "epoch": 3.1458655061069773,
+      "grad_norm": 0.42177062287547257,
+      "learning_rate": 7.262482495636627e-06,
+      "loss": 0.08,
+      "step": 2801
+    },
+    {
+      "epoch": 3.1469886283869157,
+      "grad_norm": 0.44438476148551986,
+      "learning_rate": 7.254938806734108e-06,
+      "loss": 0.0826,
+      "step": 2802
+    },
+    {
+      "epoch": 3.1481117506668537,
+      "grad_norm": 0.44860619869928176,
+      "learning_rate": 7.247396806897953e-06,
+      "loss": 0.0795,
+      "step": 2803
+    },
+    {
+      "epoch": 3.149234872946792,
+      "grad_norm": 0.45127852568778315,
+      "learning_rate": 7.239856500768829e-06,
+      "loss": 0.084,
+      "step": 2804
+    },
+    {
+      "epoch": 3.1503579952267304,
+      "grad_norm": 0.4516223814508983,
+      "learning_rate": 7.232317892986376e-06,
+      "loss": 0.0781,
+      "step": 2805
+    },
+    {
+      "epoch": 3.1514811175066684,
+      "grad_norm": 0.45355410301989446,
+      "learning_rate": 7.2247809881891805e-06,
+      "loss": 0.0839,
+      "step": 2806
+    },
+    {
+      "epoch": 3.1526042397866068,
+      "grad_norm": 0.4772117453912696,
+      "learning_rate": 7.217245791014782e-06,
+      "loss": 0.0804,
+      "step": 2807
+    },
+    {
+      "epoch": 3.153727362066545,
+      "grad_norm": 0.4208304719348036,
+      "learning_rate": 7.2097123060996764e-06,
+      "loss": 0.0701,
+      "step": 2808
+    },
+    {
+      "epoch": 3.154850484346483,
+      "grad_norm": 0.44934404159731467,
+      "learning_rate": 7.20218053807929e-06,
+      "loss": 0.0809,
+      "step": 2809
+    },
+    {
+      "epoch": 3.1559736066264215,
+      "grad_norm": 0.4313912104136727,
+      "learning_rate": 7.194650491588007e-06,
+      "loss": 0.0771,
+      "step": 2810
+    },
+    {
+      "epoch": 3.15709672890636,
+      "grad_norm": 0.4758022567163495,
+      "learning_rate": 7.1871221712591474e-06,
+      "loss": 0.087,
+      "step": 2811
+    },
+    {
+      "epoch": 3.1582198511862978,
+      "grad_norm": 0.4145445019750945,
+      "learning_rate": 7.179595581724971e-06,
+      "loss": 0.0699,
+      "step": 2812
+    },
+    {
+      "epoch": 3.159342973466236,
+      "grad_norm": 0.44264907459716,
+      "learning_rate": 7.1720707276166736e-06,
+      "loss": 0.0758,
+      "step": 2813
+    },
+    {
+      "epoch": 3.1604660957461745,
+      "grad_norm": 0.4482812080110177,
+      "learning_rate": 7.164547613564374e-06,
+      "loss": 0.0803,
+      "step": 2814
+    },
+    {
+      "epoch": 3.1615892180261125,
+      "grad_norm": 0.43483597301420784,
+      "learning_rate": 7.157026244197132e-06,
+      "loss": 0.0806,
+      "step": 2815
+    },
+    {
+      "epoch": 3.162712340306051,
+      "grad_norm": 0.4958849116746515,
+      "learning_rate": 7.149506624142924e-06,
+      "loss": 0.0844,
+      "step": 2816
+    },
+    {
+      "epoch": 3.1638354625859892,
+      "grad_norm": 0.4628156948836129,
+      "learning_rate": 7.1419887580286615e-06,
+      "loss": 0.0792,
+      "step": 2817
+    },
+    {
+      "epoch": 3.164958584865927,
+      "grad_norm": 0.44851043959036796,
+      "learning_rate": 7.13447265048017e-06,
+      "loss": 0.0815,
+      "step": 2818
+    },
+    {
+      "epoch": 3.1660817071458656,
+      "grad_norm": 0.4600530891320242,
+      "learning_rate": 7.126958306122186e-06,
+      "loss": 0.0847,
+      "step": 2819
+    },
+    {
+      "epoch": 3.1672048294258035,
+      "grad_norm": 0.43714011172496053,
+      "learning_rate": 7.119445729578374e-06,
+      "loss": 0.0779,
+      "step": 2820
+    },
+    {
+      "epoch": 3.168327951705742,
+      "grad_norm": 0.4453922999565407,
+      "learning_rate": 7.111934925471302e-06,
+      "loss": 0.0812,
+      "step": 2821
+    },
+    {
+      "epoch": 3.1694510739856803,
+      "grad_norm": 0.45599177872389823,
+      "learning_rate": 7.1044258984224524e-06,
+      "loss": 0.081,
+      "step": 2822
+    },
+    {
+      "epoch": 3.170574196265618,
+      "grad_norm": 0.44089126347694846,
+      "learning_rate": 7.096918653052214e-06,
+      "loss": 0.0771,
+      "step": 2823
+    },
+    {
+      "epoch": 3.1716973185455566,
+      "grad_norm": 0.43082375817441887,
+      "learning_rate": 7.089413193979874e-06,
+      "loss": 0.0778,
+      "step": 2824
+    },
+    {
+      "epoch": 3.172820440825495,
+      "grad_norm": 0.4425588712821237,
+      "learning_rate": 7.081909525823625e-06,
+      "loss": 0.0821,
+      "step": 2825
+    },
+    {
+      "epoch": 3.173943563105433,
+      "grad_norm": 0.4382912863146195,
+      "learning_rate": 7.074407653200559e-06,
+      "loss": 0.0807,
+      "step": 2826
+    },
+    {
+      "epoch": 3.1750666853853713,
+      "grad_norm": 0.45424971866156333,
+      "learning_rate": 7.066907580726656e-06,
+      "loss": 0.082,
+      "step": 2827
+    },
+    {
+      "epoch": 3.1761898076653097,
+      "grad_norm": 0.41419138678049927,
+      "learning_rate": 7.059409313016798e-06,
+      "loss": 0.072,
+      "step": 2828
+    },
+    {
+      "epoch": 3.1773129299452476,
+      "grad_norm": 0.44000690960252986,
+      "learning_rate": 7.051912854684748e-06,
+      "loss": 0.0808,
+      "step": 2829
+    },
+    {
+      "epoch": 3.178436052225186,
+      "grad_norm": 0.44970359244435276,
+      "learning_rate": 7.044418210343161e-06,
+      "loss": 0.086,
+      "step": 2830
+    },
+    {
+      "epoch": 3.1795591745051244,
+      "grad_norm": 0.4443948815798674,
+      "learning_rate": 7.036925384603572e-06,
+      "loss": 0.079,
+      "step": 2831
+    },
+    {
+      "epoch": 3.1806822967850623,
+      "grad_norm": 0.4554250454666911,
+      "learning_rate": 7.029434382076408e-06,
+      "loss": 0.0836,
+      "step": 2832
+    },
+    {
+      "epoch": 3.1818054190650007,
+      "grad_norm": 0.4354644467232646,
+      "learning_rate": 7.021945207370951e-06,
+      "loss": 0.0763,
+      "step": 2833
+    },
+    {
+      "epoch": 3.182928541344939,
+      "grad_norm": 0.44780451430095414,
+      "learning_rate": 7.014457865095382e-06,
+      "loss": 0.0798,
+      "step": 2834
+    },
+    {
+      "epoch": 3.184051663624877,
+      "grad_norm": 0.44114833638296336,
+      "learning_rate": 7.006972359856743e-06,
+      "loss": 0.0837,
+      "step": 2835
+    },
+    {
+      "epoch": 3.1851747859048154,
+      "grad_norm": 0.4571651151911437,
+      "learning_rate": 6.999488696260947e-06,
+      "loss": 0.0843,
+      "step": 2836
+    },
+    {
+      "epoch": 3.186297908184754,
+      "grad_norm": 0.45247999536502137,
+      "learning_rate": 6.99200687891278e-06,
+      "loss": 0.0864,
+      "step": 2837
+    },
+    {
+      "epoch": 3.1874210304646917,
+      "grad_norm": 0.4439747981411603,
+      "learning_rate": 6.984526912415878e-06,
+      "loss": 0.0815,
+      "step": 2838
+    },
+    {
+      "epoch": 3.18854415274463,
+      "grad_norm": 0.43498072881051186,
+      "learning_rate": 6.97704880137275e-06,
+      "loss": 0.0822,
+      "step": 2839
+    },
+    {
+      "epoch": 3.1896672750245685,
+      "grad_norm": 0.44156927868282614,
+      "learning_rate": 6.96957255038476e-06,
+      "loss": 0.0736,
+      "step": 2840
+    },
+    {
+      "epoch": 3.1907903973045064,
+      "grad_norm": 0.4575306361920231,
+      "learning_rate": 6.962098164052129e-06,
+      "loss": 0.0792,
+      "step": 2841
+    },
+    {
+      "epoch": 3.191913519584445,
+      "grad_norm": 0.4503884973223931,
+      "learning_rate": 6.954625646973931e-06,
+      "loss": 0.0811,
+      "step": 2842
+    },
+    {
+      "epoch": 3.193036641864383,
+      "grad_norm": 0.4419481555121357,
+      "learning_rate": 6.947155003748083e-06,
+      "loss": 0.0812,
+      "step": 2843
+    },
+    {
+      "epoch": 3.194159764144321,
+      "grad_norm": 0.4517228406602635,
+      "learning_rate": 6.939686238971356e-06,
+      "loss": 0.0784,
+      "step": 2844
+    },
+    {
+      "epoch": 3.1952828864242595,
+      "grad_norm": 0.44544948534956574,
+      "learning_rate": 6.932219357239362e-06,
+      "loss": 0.0816,
+      "step": 2845
+    },
+    {
+      "epoch": 3.196406008704198,
+      "grad_norm": 0.445448430389794,
+      "learning_rate": 6.924754363146559e-06,
+      "loss": 0.0897,
+      "step": 2846
+    },
+    {
+      "epoch": 3.197529130984136,
+      "grad_norm": 0.4576732225943825,
+      "learning_rate": 6.917291261286239e-06,
+      "loss": 0.0908,
+      "step": 2847
+    },
+    {
+      "epoch": 3.198652253264074,
+      "grad_norm": 0.4479613547612505,
+      "learning_rate": 6.909830056250527e-06,
+      "loss": 0.0784,
+      "step": 2848
+    },
+    {
+      "epoch": 3.1997753755440126,
+      "grad_norm": 0.4335107505865122,
+      "learning_rate": 6.902370752630387e-06,
+      "loss": 0.0815,
+      "step": 2849
+    },
+    {
+      "epoch": 3.2008984978239505,
+      "grad_norm": 0.42987674122594727,
+      "learning_rate": 6.894913355015611e-06,
+      "loss": 0.0803,
+      "step": 2850
+    },
+    {
+      "epoch": 3.202021620103889,
+      "grad_norm": 0.4432200727897701,
+      "learning_rate": 6.887457867994819e-06,
+      "loss": 0.078,
+      "step": 2851
+    },
+    {
+      "epoch": 3.203144742383827,
+      "grad_norm": 0.4229390999012076,
+      "learning_rate": 6.880004296155456e-06,
+      "loss": 0.077,
+      "step": 2852
+    },
+    {
+      "epoch": 3.2042678646637652,
+      "grad_norm": 0.4424366303313029,
+      "learning_rate": 6.872552644083779e-06,
+      "loss": 0.0838,
+      "step": 2853
+    },
+    {
+      "epoch": 3.2053909869437036,
+      "grad_norm": 0.4483537337256006,
+      "learning_rate": 6.865102916364876e-06,
+      "loss": 0.0813,
+      "step": 2854
+    },
+    {
+      "epoch": 3.2065141092236416,
+      "grad_norm": 0.4459517380181533,
+      "learning_rate": 6.857655117582647e-06,
+      "loss": 0.0856,
+      "step": 2855
+    },
+    {
+      "epoch": 3.20763723150358,
+      "grad_norm": 0.45366324323089674,
+      "learning_rate": 6.850209252319804e-06,
+      "loss": 0.0814,
+      "step": 2856
+    },
+    {
+      "epoch": 3.2087603537835183,
+      "grad_norm": 0.46685596344158564,
+      "learning_rate": 6.842765325157874e-06,
+      "loss": 0.0829,
+      "step": 2857
+    },
+    {
+      "epoch": 3.2098834760634563,
+      "grad_norm": 0.45685853191410364,
+      "learning_rate": 6.83532334067718e-06,
+      "loss": 0.0819,
+      "step": 2858
+    },
+    {
+      "epoch": 3.2110065983433946,
+      "grad_norm": 0.450175843547502,
+      "learning_rate": 6.8278833034568595e-06,
+      "loss": 0.082,
+      "step": 2859
+    },
+    {
+      "epoch": 3.212129720623333,
+      "grad_norm": 0.45393030619312513,
+      "learning_rate": 6.820445218074849e-06,
+      "loss": 0.0838,
+      "step": 2860
+    },
+    {
+      "epoch": 3.213252842903271,
+      "grad_norm": 0.4349893777321262,
+      "learning_rate": 6.813009089107887e-06,
+      "loss": 0.0756,
+      "step": 2861
+    },
+    {
+      "epoch": 3.2143759651832093,
+      "grad_norm": 0.42814105018993454,
+      "learning_rate": 6.805574921131506e-06,
+      "loss": 0.0789,
+      "step": 2862
+    },
+    {
+      "epoch": 3.2154990874631477,
+      "grad_norm": 0.4538034830964177,
+      "learning_rate": 6.798142718720027e-06,
+      "loss": 0.0844,
+      "step": 2863
+    },
+    {
+      "epoch": 3.2166222097430857,
+      "grad_norm": 0.45315166207455276,
+      "learning_rate": 6.790712486446567e-06,
+      "loss": 0.0855,
+      "step": 2864
+    },
+    {
+      "epoch": 3.217745332023024,
+      "grad_norm": 0.4527713015472594,
+      "learning_rate": 6.783284228883029e-06,
+      "loss": 0.0794,
+      "step": 2865
+    },
+    {
+      "epoch": 3.2188684543029624,
+      "grad_norm": 0.4292889184279728,
+      "learning_rate": 6.775857950600107e-06,
+      "loss": 0.0723,
+      "step": 2866
+    },
+    {
+      "epoch": 3.2199915765829004,
+      "grad_norm": 0.4384777852700717,
+      "learning_rate": 6.768433656167267e-06,
+      "loss": 0.0801,
+      "step": 2867
+    },
+    {
+      "epoch": 3.2211146988628387,
+      "grad_norm": 0.4749472746328939,
+      "learning_rate": 6.76101135015276e-06,
+      "loss": 0.0853,
+      "step": 2868
+    },
+    {
+      "epoch": 3.222237821142777,
+      "grad_norm": 0.447676493402192,
+      "learning_rate": 6.7535910371236105e-06,
+      "loss": 0.085,
+      "step": 2869
+    },
+    {
+      "epoch": 3.223360943422715,
+      "grad_norm": 0.4448626847467645,
+      "learning_rate": 6.746172721645625e-06,
+      "loss": 0.0845,
+      "step": 2870
+    },
+    {
+      "epoch": 3.2244840657026534,
+      "grad_norm": 0.4369101325966336,
+      "learning_rate": 6.73875640828337e-06,
+      "loss": 0.0789,
+      "step": 2871
+    },
+    {
+      "epoch": 3.2256071879825914,
+      "grad_norm": 0.4650100486596405,
+      "learning_rate": 6.731342101600183e-06,
+      "loss": 0.0846,
+      "step": 2872
+    },
+    {
+      "epoch": 3.2267303102625298,
+      "grad_norm": 0.45213980309912594,
+      "learning_rate": 6.7239298061581716e-06,
+      "loss": 0.0821,
+      "step": 2873
+    },
+    {
+      "epoch": 3.227853432542468,
+      "grad_norm": 0.4660318930820804,
+      "learning_rate": 6.716519526518201e-06,
+      "loss": 0.0803,
+      "step": 2874
+    },
+    {
+      "epoch": 3.228976554822406,
+      "grad_norm": 0.43454024938308855,
+      "learning_rate": 6.7091112672399e-06,
+      "loss": 0.0792,
+      "step": 2875
+    },
+    {
+      "epoch": 3.2300996771023445,
+      "grad_norm": 0.44078038350930815,
+      "learning_rate": 6.701705032881654e-06,
+      "loss": 0.0798,
+      "step": 2876
+    },
+    {
+      "epoch": 3.231222799382283,
+      "grad_norm": 0.452985339480439,
+      "learning_rate": 6.694300828000594e-06,
+      "loss": 0.0811,
+      "step": 2877
+    },
+    {
+      "epoch": 3.232345921662221,
+      "grad_norm": 0.4432797452038033,
+      "learning_rate": 6.686898657152612e-06,
+      "loss": 0.0768,
+      "step": 2878
+    },
+    {
+      "epoch": 3.233469043942159,
+      "grad_norm": 0.43857516156481385,
+      "learning_rate": 6.679498524892345e-06,
+      "loss": 0.0776,
+      "step": 2879
+    },
+    {
+      "epoch": 3.2345921662220976,
+      "grad_norm": 0.46446080341677276,
+      "learning_rate": 6.672100435773176e-06,
+      "loss": 0.0847,
+      "step": 2880
+    },
+    {
+      "epoch": 3.2357152885020355,
+      "grad_norm": 0.45625351695476635,
+      "learning_rate": 6.664704394347235e-06,
+      "loss": 0.0803,
+      "step": 2881
+    },
+    {
+      "epoch": 3.236838410781974,
+      "grad_norm": 0.45358260601300854,
+      "learning_rate": 6.657310405165379e-06,
+      "loss": 0.0794,
+      "step": 2882
+    },
+    {
+      "epoch": 3.2379615330619123,
+      "grad_norm": 0.430765669099815,
+      "learning_rate": 6.649918472777216e-06,
+      "loss": 0.078,
+      "step": 2883
+    },
+    {
+      "epoch": 3.23908465534185,
+      "grad_norm": 0.4417554716567722,
+      "learning_rate": 6.642528601731082e-06,
+      "loss": 0.0799,
+      "step": 2884
+    },
+    {
+      "epoch": 3.2402077776217886,
+      "grad_norm": 0.4426439861111362,
+      "learning_rate": 6.6351407965740465e-06,
+      "loss": 0.0736,
+      "step": 2885
+    },
+    {
+      "epoch": 3.241330899901727,
+      "grad_norm": 0.4461344386327623,
+      "learning_rate": 6.627755061851911e-06,
+      "loss": 0.0826,
+      "step": 2886
+    },
+    {
+      "epoch": 3.242454022181665,
+      "grad_norm": 0.4428801843373427,
+      "learning_rate": 6.620371402109195e-06,
+      "loss": 0.083,
+      "step": 2887
+    },
+    {
+      "epoch": 3.2435771444616033,
+      "grad_norm": 0.45311188824038895,
+      "learning_rate": 6.612989821889144e-06,
+      "loss": 0.0877,
+      "step": 2888
+    },
+    {
+      "epoch": 3.2447002667415417,
+      "grad_norm": 0.43522073849830734,
+      "learning_rate": 6.605610325733728e-06,
+      "loss": 0.0761,
+      "step": 2889
+    },
+    {
+      "epoch": 3.2458233890214796,
+      "grad_norm": 0.43937974707906036,
+      "learning_rate": 6.5982329181836325e-06,
+      "loss": 0.078,
+      "step": 2890
+    },
+    {
+      "epoch": 3.246946511301418,
+      "grad_norm": 0.44838131831874334,
+      "learning_rate": 6.590857603778259e-06,
+      "loss": 0.0765,
+      "step": 2891
+    },
+    {
+      "epoch": 3.2480696335813564,
+      "grad_norm": 0.4487567179771303,
+      "learning_rate": 6.583484387055716e-06,
+      "loss": 0.0829,
+      "step": 2892
+    },
+    {
+      "epoch": 3.2491927558612943,
+      "grad_norm": 0.43973073043498445,
+      "learning_rate": 6.5761132725528265e-06,
+      "loss": 0.0775,
+      "step": 2893
+    },
+    {
+      "epoch": 3.2503158781412327,
+      "grad_norm": 0.44761002978377296,
+      "learning_rate": 6.568744264805118e-06,
+      "loss": 0.0888,
+      "step": 2894
+    },
+    {
+      "epoch": 3.251439000421171,
+      "grad_norm": 0.4314843009567722,
+      "learning_rate": 6.561377368346824e-06,
+      "loss": 0.0752,
+      "step": 2895
+    },
+    {
+      "epoch": 3.252562122701109,
+      "grad_norm": 0.4484597216270316,
+      "learning_rate": 6.554012587710879e-06,
+      "loss": 0.0786,
+      "step": 2896
+    },
+    {
+      "epoch": 3.2536852449810474,
+      "grad_norm": 0.4533967440227061,
+      "learning_rate": 6.546649927428905e-06,
+      "loss": 0.084,
+      "step": 2897
+    },
+    {
+      "epoch": 3.2548083672609858,
+      "grad_norm": 0.43537375245345705,
+      "learning_rate": 6.539289392031234e-06,
+      "loss": 0.0737,
+      "step": 2898
+    },
+    {
+      "epoch": 3.2559314895409237,
+      "grad_norm": 0.44392098653435624,
+      "learning_rate": 6.531930986046884e-06,
+      "loss": 0.0798,
+      "step": 2899
+    },
+    {
+      "epoch": 3.257054611820862,
+      "grad_norm": 0.43078694542955465,
+      "learning_rate": 6.524574714003562e-06,
+      "loss": 0.0787,
+      "step": 2900
+    },
+    {
+      "epoch": 3.2581777341008005,
+      "grad_norm": 0.4525304342259154,
+      "learning_rate": 6.517220580427669e-06,
+      "loss": 0.0852,
+      "step": 2901
+    },
+    {
+      "epoch": 3.2593008563807384,
+      "grad_norm": 0.44680111766540903,
+      "learning_rate": 6.509868589844274e-06,
+      "loss": 0.0732,
+      "step": 2902
+    },
+    {
+      "epoch": 3.260423978660677,
+      "grad_norm": 0.44095965893761685,
+      "learning_rate": 6.502518746777143e-06,
+      "loss": 0.085,
+      "step": 2903
+    },
+    {
+      "epoch": 3.2615471009406147,
+      "grad_norm": 0.45173240369816303,
+      "learning_rate": 6.495171055748714e-06,
+      "loss": 0.0794,
+      "step": 2904
+    },
+    {
+      "epoch": 3.262670223220553,
+      "grad_norm": 0.4439061819150142,
+      "learning_rate": 6.487825521280109e-06,
+      "loss": 0.0826,
+      "step": 2905
+    },
+    {
+      "epoch": 3.2637933455004915,
+      "grad_norm": 0.4334545536696232,
+      "learning_rate": 6.480482147891106e-06,
+      "loss": 0.0801,
+      "step": 2906
+    },
+    {
+      "epoch": 3.2649164677804294,
+      "grad_norm": 0.4329477851970838,
+      "learning_rate": 6.473140940100169e-06,
+      "loss": 0.0822,
+      "step": 2907
+    },
+    {
+      "epoch": 3.266039590060368,
+      "grad_norm": 0.44476467976568695,
+      "learning_rate": 6.4658019024244214e-06,
+      "loss": 0.0816,
+      "step": 2908
+    },
+    {
+      "epoch": 3.267162712340306,
+      "grad_norm": 0.4373699550356578,
+      "learning_rate": 6.458465039379655e-06,
+      "loss": 0.0749,
+      "step": 2909
+    },
+    {
+      "epoch": 3.268285834620244,
+      "grad_norm": 0.4289188369981385,
+      "learning_rate": 6.451130355480326e-06,
+      "loss": 0.0787,
+      "step": 2910
+    },
+    {
+      "epoch": 3.2694089569001825,
+      "grad_norm": 0.46145266949341135,
+      "learning_rate": 6.44379785523954e-06,
+      "loss": 0.0805,
+      "step": 2911
+    },
+    {
+      "epoch": 3.270532079180121,
+      "grad_norm": 0.4319621203979465,
+      "learning_rate": 6.4364675431690684e-06,
+      "loss": 0.0735,
+      "step": 2912
+    },
+    {
+      "epoch": 3.271655201460059,
+      "grad_norm": 0.4507565266528019,
+      "learning_rate": 6.429139423779332e-06,
+      "loss": 0.0809,
+      "step": 2913
+    },
+    {
+      "epoch": 3.272778323739997,
+      "grad_norm": 0.44904183335465914,
+      "learning_rate": 6.421813501579403e-06,
+      "loss": 0.0818,
+      "step": 2914
+    },
+    {
+      "epoch": 3.2739014460199356,
+      "grad_norm": 0.44237241058884846,
+      "learning_rate": 6.414489781077009e-06,
+      "loss": 0.076,
+      "step": 2915
+    },
+    {
+      "epoch": 3.2750245682998735,
+      "grad_norm": 0.45079073834235045,
+      "learning_rate": 6.407168266778503e-06,
+      "loss": 0.0807,
+      "step": 2916
+    },
+    {
+      "epoch": 3.276147690579812,
+      "grad_norm": 0.4325942699296943,
+      "learning_rate": 6.399848963188902e-06,
+      "loss": 0.0752,
+      "step": 2917
+    },
+    {
+      "epoch": 3.27727081285975,
+      "grad_norm": 0.4436273348713645,
+      "learning_rate": 6.392531874811849e-06,
+      "loss": 0.0759,
+      "step": 2918
+    },
+    {
+      "epoch": 3.2783939351396882,
+      "grad_norm": 0.4399049256998244,
+      "learning_rate": 6.385217006149633e-06,
+      "loss": 0.0768,
+      "step": 2919
+    },
+    {
+      "epoch": 3.2795170574196266,
+      "grad_norm": 0.45422810539143077,
+      "learning_rate": 6.3779043617031775e-06,
+      "loss": 0.081,
+      "step": 2920
+    },
+    {
+      "epoch": 3.2806401796995646,
+      "grad_norm": 0.4643323939153561,
+      "learning_rate": 6.370593945972022e-06,
+      "loss": 0.0824,
+      "step": 2921
+    },
+    {
+      "epoch": 3.281763301979503,
+      "grad_norm": 0.4384243413962703,
+      "learning_rate": 6.363285763454352e-06,
+      "loss": 0.0777,
+      "step": 2922
+    },
+    {
+      "epoch": 3.2828864242594413,
+      "grad_norm": 0.44988902779488643,
+      "learning_rate": 6.355979818646972e-06,
+      "loss": 0.0797,
+      "step": 2923
+    },
+    {
+      "epoch": 3.2840095465393793,
+      "grad_norm": 0.4533162798690284,
+      "learning_rate": 6.34867611604531e-06,
+      "loss": 0.0842,
+      "step": 2924
+    },
+    {
+      "epoch": 3.2851326688193176,
+      "grad_norm": 0.4470213268981242,
+      "learning_rate": 6.341374660143419e-06,
+      "loss": 0.0846,
+      "step": 2925
+    },
+    {
+      "epoch": 3.286255791099256,
+      "grad_norm": 0.44475456824817117,
+      "learning_rate": 6.334075455433957e-06,
+      "loss": 0.0848,
+      "step": 2926
+    },
+    {
+      "epoch": 3.287378913379194,
+      "grad_norm": 0.4345189413360516,
+      "learning_rate": 6.326778506408209e-06,
+      "loss": 0.0763,
+      "step": 2927
+    },
+    {
+      "epoch": 3.2885020356591323,
+      "grad_norm": 0.4443083844205599,
+      "learning_rate": 6.319483817556067e-06,
+      "loss": 0.0836,
+      "step": 2928
+    },
+    {
+      "epoch": 3.2896251579390707,
+      "grad_norm": 0.44868349523571427,
+      "learning_rate": 6.312191393366036e-06,
+      "loss": 0.0838,
+      "step": 2929
+    },
+    {
+      "epoch": 3.2907482802190087,
+      "grad_norm": 0.43446703024226885,
+      "learning_rate": 6.304901238325224e-06,
+      "loss": 0.0802,
+      "step": 2930
+    },
+    {
+      "epoch": 3.291871402498947,
+      "grad_norm": 0.4482045860075501,
+      "learning_rate": 6.297613356919341e-06,
+      "loss": 0.0808,
+      "step": 2931
+    },
+    {
+      "epoch": 3.2929945247788854,
+      "grad_norm": 0.4427022564744176,
+      "learning_rate": 6.290327753632705e-06,
+      "loss": 0.08,
+      "step": 2932
+    },
+    {
+      "epoch": 3.2941176470588234,
+      "grad_norm": 0.45857038552731477,
+      "learning_rate": 6.283044432948222e-06,
+      "loss": 0.0791,
+      "step": 2933
+    },
+    {
+      "epoch": 3.2952407693387618,
+      "grad_norm": 0.4495217764887851,
+      "learning_rate": 6.275763399347403e-06,
+      "loss": 0.0809,
+      "step": 2934
+    },
+    {
+      "epoch": 3.2963638916187,
+      "grad_norm": 0.4377261844495198,
+      "learning_rate": 6.268484657310351e-06,
+      "loss": 0.0771,
+      "step": 2935
+    },
+    {
+      "epoch": 3.297487013898638,
+      "grad_norm": 0.4693294460359561,
+      "learning_rate": 6.26120821131575e-06,
+      "loss": 0.0896,
+      "step": 2936
+    },
+    {
+      "epoch": 3.2986101361785765,
+      "grad_norm": 0.4548392670428416,
+      "learning_rate": 6.25393406584088e-06,
+      "loss": 0.0804,
+      "step": 2937
+    },
+    {
+      "epoch": 3.299733258458515,
+      "grad_norm": 0.4338443235373701,
+      "learning_rate": 6.246662225361603e-06,
+      "loss": 0.0758,
+      "step": 2938
+    },
+    {
+      "epoch": 3.3008563807384528,
+      "grad_norm": 0.4094192153958026,
+      "learning_rate": 6.239392694352362e-06,
+      "loss": 0.0681,
+      "step": 2939
+    },
+    {
+      "epoch": 3.301979503018391,
+      "grad_norm": 0.46354022190108407,
+      "learning_rate": 6.232125477286184e-06,
+      "loss": 0.0901,
+      "step": 2940
+    },
+    {
+      "epoch": 3.3031026252983295,
+      "grad_norm": 0.43420174617897006,
+      "learning_rate": 6.224860578634659e-06,
+      "loss": 0.0784,
+      "step": 2941
+    },
+    {
+      "epoch": 3.3042257475782675,
+      "grad_norm": 0.4497917258205283,
+      "learning_rate": 6.217598002867965e-06,
+      "loss": 0.0847,
+      "step": 2942
+    },
+    {
+      "epoch": 3.305348869858206,
+      "grad_norm": 0.44585339258016715,
+      "learning_rate": 6.210337754454842e-06,
+      "loss": 0.0809,
+      "step": 2943
+    },
+    {
+      "epoch": 3.3064719921381442,
+      "grad_norm": 0.4134829693581107,
+      "learning_rate": 6.203079837862607e-06,
+      "loss": 0.0766,
+      "step": 2944
+    },
+    {
+      "epoch": 3.307595114418082,
+      "grad_norm": 0.4509752340647484,
+      "learning_rate": 6.195824257557126e-06,
+      "loss": 0.0861,
+      "step": 2945
+    },
+    {
+      "epoch": 3.3087182366980206,
+      "grad_norm": 0.4664138175037376,
+      "learning_rate": 6.188571018002843e-06,
+      "loss": 0.0853,
+      "step": 2946
+    },
+    {
+      "epoch": 3.309841358977959,
+      "grad_norm": 0.4426772347810475,
+      "learning_rate": 6.181320123662755e-06,
+      "loss": 0.0755,
+      "step": 2947
+    },
+    {
+      "epoch": 3.310964481257897,
+      "grad_norm": 0.4456457712399848,
+      "learning_rate": 6.174071578998419e-06,
+      "loss": 0.0787,
+      "step": 2948
+    },
+    {
+      "epoch": 3.3120876035378353,
+      "grad_norm": 0.4375785593556192,
+      "learning_rate": 6.166825388469946e-06,
+      "loss": 0.0796,
+      "step": 2949
+    },
+    {
+      "epoch": 3.3132107258177736,
+      "grad_norm": 0.45268483396403963,
+      "learning_rate": 6.159581556535989e-06,
+      "loss": 0.084,
+      "step": 2950
+    },
+    {
+      "epoch": 3.3143338480977116,
+      "grad_norm": 0.42407864679075813,
+      "learning_rate": 6.152340087653762e-06,
+      "loss": 0.0794,
+      "step": 2951
+    },
+    {
+      "epoch": 3.31545697037765,
+      "grad_norm": 0.4509299352401977,
+      "learning_rate": 6.145100986279021e-06,
+      "loss": 0.0861,
+      "step": 2952
+    },
+    {
+      "epoch": 3.316580092657588,
+      "grad_norm": 0.43447301128730714,
+      "learning_rate": 6.137864256866065e-06,
+      "loss": 0.084,
+      "step": 2953
+    },
+    {
+      "epoch": 3.3177032149375263,
+      "grad_norm": 0.4325020912385428,
+      "learning_rate": 6.130629903867734e-06,
+      "loss": 0.0799,
+      "step": 2954
+    },
+    {
+      "epoch": 3.3188263372174647,
+      "grad_norm": 0.43598451645884106,
+      "learning_rate": 6.123397931735402e-06,
+      "loss": 0.0776,
+      "step": 2955
+    },
+    {
+      "epoch": 3.3199494594974026,
+      "grad_norm": 0.4469736893147317,
+      "learning_rate": 6.116168344918982e-06,
+      "loss": 0.0831,
+      "step": 2956
+    },
+    {
+      "epoch": 3.321072581777341,
+      "grad_norm": 0.4626495409312179,
+      "learning_rate": 6.10894114786692e-06,
+      "loss": 0.085,
+      "step": 2957
+    },
+    {
+      "epoch": 3.3221957040572794,
+      "grad_norm": 0.4474334248967847,
+      "learning_rate": 6.101716345026189e-06,
+      "loss": 0.0818,
+      "step": 2958
+    },
+    {
+      "epoch": 3.3233188263372173,
+      "grad_norm": 0.4512283659979245,
+      "learning_rate": 6.094493940842293e-06,
+      "loss": 0.0859,
+      "step": 2959
+    },
+    {
+      "epoch": 3.3244419486171557,
+      "grad_norm": 0.4505970094609106,
+      "learning_rate": 6.08727393975925e-06,
+      "loss": 0.0783,
+      "step": 2960
+    },
+    {
+      "epoch": 3.325565070897094,
+      "grad_norm": 0.4250274459193349,
+      "learning_rate": 6.080056346219608e-06,
+      "loss": 0.0816,
+      "step": 2961
+    },
+    {
+      "epoch": 3.326688193177032,
+      "grad_norm": 0.47814033258276745,
+      "learning_rate": 6.072841164664433e-06,
+      "loss": 0.0821,
+      "step": 2962
+    },
+    {
+      "epoch": 3.3278113154569704,
+      "grad_norm": 0.45580838414860525,
+      "learning_rate": 6.065628399533307e-06,
+      "loss": 0.0825,
+      "step": 2963
+    },
+    {
+      "epoch": 3.328934437736909,
+      "grad_norm": 0.4281087882366829,
+      "learning_rate": 6.058418055264328e-06,
+      "loss": 0.0767,
+      "step": 2964
+    },
+    {
+      "epoch": 3.3300575600168467,
+      "grad_norm": 0.459821896563383,
+      "learning_rate": 6.051210136294089e-06,
+      "loss": 0.0909,
+      "step": 2965
+    },
+    {
+      "epoch": 3.331180682296785,
+      "grad_norm": 0.43302194673000116,
+      "learning_rate": 6.044004647057709e-06,
+      "loss": 0.0772,
+      "step": 2966
+    },
+    {
+      "epoch": 3.3323038045767235,
+      "grad_norm": 0.44404264162176665,
+      "learning_rate": 6.036801591988802e-06,
+      "loss": 0.0793,
+      "step": 2967
+    },
+    {
+      "epoch": 3.3334269268566614,
+      "grad_norm": 0.4387838908116998,
+      "learning_rate": 6.0296009755194875e-06,
+      "loss": 0.0801,
+      "step": 2968
+    },
+    {
+      "epoch": 3.3345500491366,
+      "grad_norm": 0.4532205199732577,
+      "learning_rate": 6.022402802080392e-06,
+      "loss": 0.0835,
+      "step": 2969
+    },
+    {
+      "epoch": 3.3356731714165377,
+      "grad_norm": 0.4445884667652232,
+      "learning_rate": 6.015207076100618e-06,
+      "loss": 0.0841,
+      "step": 2970
+    },
+    {
+      "epoch": 3.336796293696476,
+      "grad_norm": 0.4396390636961917,
+      "learning_rate": 6.00801380200778e-06,
+      "loss": 0.0764,
+      "step": 2971
+    },
+    {
+      "epoch": 3.3379194159764145,
+      "grad_norm": 0.4140514704401163,
+      "learning_rate": 6.000822984227981e-06,
+      "loss": 0.0726,
+      "step": 2972
+    },
+    {
+      "epoch": 3.3390425382563524,
+      "grad_norm": 0.4482407427874633,
+      "learning_rate": 5.993634627185807e-06,
+      "loss": 0.0846,
+      "step": 2973
+    },
+    {
+      "epoch": 3.340165660536291,
+      "grad_norm": 0.4736302287268389,
+      "learning_rate": 5.986448735304339e-06,
+      "loss": 0.08,
+      "step": 2974
+    },
+    {
+      "epoch": 3.341288782816229,
+      "grad_norm": 0.45919790518048875,
+      "learning_rate": 5.979265313005128e-06,
+      "loss": 0.0821,
+      "step": 2975
+    },
+    {
+      "epoch": 3.342411905096167,
+      "grad_norm": 0.4370562219096713,
+      "learning_rate": 5.972084364708217e-06,
+      "loss": 0.082,
+      "step": 2976
+    },
+    {
+      "epoch": 3.3435350273761055,
+      "grad_norm": 0.4497513684415123,
+      "learning_rate": 5.9649058948321225e-06,
+      "loss": 0.0794,
+      "step": 2977
+    },
+    {
+      "epoch": 3.344658149656044,
+      "grad_norm": 0.45727145539332165,
+      "learning_rate": 5.957729907793837e-06,
+      "loss": 0.0749,
+      "step": 2978
+    },
+    {
+      "epoch": 3.345781271935982,
+      "grad_norm": 0.4601952594639126,
+      "learning_rate": 5.950556408008818e-06,
+      "loss": 0.0856,
+      "step": 2979
+    },
+    {
+      "epoch": 3.3469043942159202,
+      "grad_norm": 0.43940530937369787,
+      "learning_rate": 5.943385399891004e-06,
+      "loss": 0.077,
+      "step": 2980
+    },
+    {
+      "epoch": 3.3480275164958586,
+      "grad_norm": 0.463228542311302,
+      "learning_rate": 5.9362168878527944e-06,
+      "loss": 0.0894,
+      "step": 2981
+    },
+    {
+      "epoch": 3.3491506387757966,
+      "grad_norm": 0.44397605607111024,
+      "learning_rate": 5.929050876305056e-06,
+      "loss": 0.075,
+      "step": 2982
+    },
+    {
+      "epoch": 3.350273761055735,
+      "grad_norm": 0.4474755147706091,
+      "learning_rate": 5.921887369657113e-06,
+      "loss": 0.0823,
+      "step": 2983
+    },
+    {
+      "epoch": 3.3513968833356733,
+      "grad_norm": 0.43075761139504265,
+      "learning_rate": 5.914726372316747e-06,
+      "loss": 0.0797,
+      "step": 2984
+    },
+    {
+      "epoch": 3.3525200056156113,
+      "grad_norm": 0.4551763638672436,
+      "learning_rate": 5.9075678886902e-06,
+      "loss": 0.0875,
+      "step": 2985
+    },
+    {
+      "epoch": 3.3536431278955496,
+      "grad_norm": 0.4372584770340282,
+      "learning_rate": 5.900411923182166e-06,
+      "loss": 0.0797,
+      "step": 2986
+    },
+    {
+      "epoch": 3.354766250175488,
+      "grad_norm": 0.42796756138216163,
+      "learning_rate": 5.893258480195789e-06,
+      "loss": 0.0859,
+      "step": 2987
+    },
+    {
+      "epoch": 3.355889372455426,
+      "grad_norm": 0.4621345317165231,
+      "learning_rate": 5.886107564132667e-06,
+      "loss": 0.0861,
+      "step": 2988
+    },
+    {
+      "epoch": 3.3570124947353643,
+      "grad_norm": 0.4416529130015374,
+      "learning_rate": 5.878959179392828e-06,
+      "loss": 0.0821,
+      "step": 2989
+    },
+    {
+      "epoch": 3.3581356170153027,
+      "grad_norm": 0.44707394897677344,
+      "learning_rate": 5.871813330374756e-06,
+      "loss": 0.0764,
+      "step": 2990
+    },
+    {
+      "epoch": 3.3592587392952407,
+      "grad_norm": 0.4674292247994994,
+      "learning_rate": 5.86467002147537e-06,
+      "loss": 0.1027,
+      "step": 2991
+    },
+    {
+      "epoch": 3.360381861575179,
+      "grad_norm": 0.4473398857345525,
+      "learning_rate": 5.857529257090027e-06,
+      "loss": 0.0808,
+      "step": 2992
+    },
+    {
+      "epoch": 3.3615049838551174,
+      "grad_norm": 0.4474406567869844,
+      "learning_rate": 5.85039104161252e-06,
+      "loss": 0.0791,
+      "step": 2993
+    },
+    {
+      "epoch": 3.3626281061350554,
+      "grad_norm": 0.4284252832694052,
+      "learning_rate": 5.843255379435069e-06,
+      "loss": 0.0805,
+      "step": 2994
+    },
+    {
+      "epoch": 3.3637512284149937,
+      "grad_norm": 0.4470373964616598,
+      "learning_rate": 5.8361222749483246e-06,
+      "loss": 0.0811,
+      "step": 2995
+    },
+    {
+      "epoch": 3.364874350694932,
+      "grad_norm": 0.45169365813301077,
+      "learning_rate": 5.8289917325413655e-06,
+      "loss": 0.0809,
+      "step": 2996
+    },
+    {
+      "epoch": 3.36599747297487,
+      "grad_norm": 0.45414154287290753,
+      "learning_rate": 5.821863756601694e-06,
+      "loss": 0.0827,
+      "step": 2997
+    },
+    {
+      "epoch": 3.3671205952548084,
+      "grad_norm": 0.4542864047987009,
+      "learning_rate": 5.814738351515234e-06,
+      "loss": 0.0901,
+      "step": 2998
+    },
+    {
+      "epoch": 3.368243717534747,
+      "grad_norm": 0.4403191156268337,
+      "learning_rate": 5.807615521666321e-06,
+      "loss": 0.0737,
+      "step": 2999
+    },
+    {
+      "epoch": 3.3693668398146848,
+      "grad_norm": 0.47140768833660335,
+      "learning_rate": 5.800495271437712e-06,
+      "loss": 0.0857,
+      "step": 3000
+    },
+    {
+      "epoch": 3.370489962094623,
+      "grad_norm": 0.44449931199515863,
+      "learning_rate": 5.793377605210575e-06,
+      "loss": 0.0763,
+      "step": 3001
+    },
+    {
+      "epoch": 3.371613084374561,
+      "grad_norm": 0.43375063637654476,
+      "learning_rate": 5.786262527364489e-06,
+      "loss": 0.0768,
+      "step": 3002
+    },
+    {
+      "epoch": 3.3727362066544995,
+      "grad_norm": 0.4634645012272231,
+      "learning_rate": 5.779150042277445e-06,
+      "loss": 0.0832,
+      "step": 3003
+    },
+    {
+      "epoch": 3.373859328934438,
+      "grad_norm": 0.46021930500381747,
+      "learning_rate": 5.7720401543258245e-06,
+      "loss": 0.0864,
+      "step": 3004
+    },
+    {
+      "epoch": 3.374982451214376,
+      "grad_norm": 0.45355421975815485,
+      "learning_rate": 5.764932867884423e-06,
+      "loss": 0.0819,
+      "step": 3005
+    },
+    {
+      "epoch": 3.376105573494314,
+      "grad_norm": 0.4399739477662287,
+      "learning_rate": 5.757828187326433e-06,
+      "loss": 0.0783,
+      "step": 3006
+    },
+    {
+      "epoch": 3.3772286957742526,
+      "grad_norm": 0.43934576191385377,
+      "learning_rate": 5.750726117023442e-06,
+      "loss": 0.0769,
+      "step": 3007
+    },
+    {
+      "epoch": 3.3783518180541905,
+      "grad_norm": 0.4567949271653509,
+      "learning_rate": 5.743626661345433e-06,
+      "loss": 0.0824,
+      "step": 3008
+    },
+    {
+      "epoch": 3.379474940334129,
+      "grad_norm": 0.4424452473412179,
+      "learning_rate": 5.736529824660778e-06,
+      "loss": 0.0812,
+      "step": 3009
+    },
+    {
+      "epoch": 3.3805980626140673,
+      "grad_norm": 0.44503663152596046,
+      "learning_rate": 5.729435611336239e-06,
+      "loss": 0.0794,
+      "step": 3010
+    },
+    {
+      "epoch": 3.381721184894005,
+      "grad_norm": 0.44895576795697817,
+      "learning_rate": 5.722344025736965e-06,
+      "loss": 0.0821,
+      "step": 3011
+    },
+    {
+      "epoch": 3.3828443071739436,
+      "grad_norm": 0.4744276589880692,
+      "learning_rate": 5.715255072226489e-06,
+      "loss": 0.0914,
+      "step": 3012
+    },
+    {
+      "epoch": 3.383967429453882,
+      "grad_norm": 0.45473296934213203,
+      "learning_rate": 5.708168755166714e-06,
+      "loss": 0.0762,
+      "step": 3013
+    },
+    {
+      "epoch": 3.38509055173382,
+      "grad_norm": 0.4669408291943065,
+      "learning_rate": 5.701085078917934e-06,
+      "loss": 0.0812,
+      "step": 3014
+    },
+    {
+      "epoch": 3.3862136740137583,
+      "grad_norm": 0.4985094979807337,
+      "learning_rate": 5.694004047838812e-06,
+      "loss": 0.081,
+      "step": 3015
+    },
+    {
+      "epoch": 3.3873367962936967,
+      "grad_norm": 0.4664501659830471,
+      "learning_rate": 5.686925666286385e-06,
+      "loss": 0.0782,
+      "step": 3016
+    },
+    {
+      "epoch": 3.3884599185736346,
+      "grad_norm": 0.4814941024166765,
+      "learning_rate": 5.679849938616062e-06,
+      "loss": 0.0907,
+      "step": 3017
+    },
+    {
+      "epoch": 3.389583040853573,
+      "grad_norm": 0.4481377810964817,
+      "learning_rate": 5.672776869181609e-06,
+      "loss": 0.079,
+      "step": 3018
+    },
+    {
+      "epoch": 3.390706163133511,
+      "grad_norm": 0.4104819286802798,
+      "learning_rate": 5.665706462335167e-06,
+      "loss": 0.0704,
+      "step": 3019
+    },
+    {
+      "epoch": 3.3918292854134493,
+      "grad_norm": 0.4308997552132428,
+      "learning_rate": 5.658638722427237e-06,
+      "loss": 0.08,
+      "step": 3020
+    },
+    {
+      "epoch": 3.3929524076933877,
+      "grad_norm": 0.44189383865557846,
+      "learning_rate": 5.651573653806675e-06,
+      "loss": 0.0785,
+      "step": 3021
+    },
+    {
+      "epoch": 3.3940755299733256,
+      "grad_norm": 0.45421284153104713,
+      "learning_rate": 5.6445112608207e-06,
+      "loss": 0.078,
+      "step": 3022
+    },
+    {
+      "epoch": 3.395198652253264,
+      "grad_norm": 0.4813721347164326,
+      "learning_rate": 5.6374515478148714e-06,
+      "loss": 0.0932,
+      "step": 3023
+    },
+    {
+      "epoch": 3.3963217745332024,
+      "grad_norm": 0.44235907345540504,
+      "learning_rate": 5.630394519133114e-06,
+      "loss": 0.0824,
+      "step": 3024
+    },
+    {
+      "epoch": 3.3974448968131403,
+      "grad_norm": 0.4338332189073094,
+      "learning_rate": 5.6233401791176946e-06,
+      "loss": 0.0786,
+      "step": 3025
+    },
+    {
+      "epoch": 3.3985680190930787,
+      "grad_norm": 0.4394631553406978,
+      "learning_rate": 5.616288532109225e-06,
+      "loss": 0.0794,
+      "step": 3026
+    },
+    {
+      "epoch": 3.399691141373017,
+      "grad_norm": 0.44813371027964505,
+      "learning_rate": 5.609239582446666e-06,
+      "loss": 0.0836,
+      "step": 3027
+    },
+    {
+      "epoch": 3.400814263652955,
+      "grad_norm": 0.45706449566282054,
+      "learning_rate": 5.602193334467307e-06,
+      "loss": 0.0845,
+      "step": 3028
+    },
+    {
+      "epoch": 3.4019373859328934,
+      "grad_norm": 0.4739118872159749,
+      "learning_rate": 5.595149792506785e-06,
+      "loss": 0.0884,
+      "step": 3029
+    },
+    {
+      "epoch": 3.403060508212832,
+      "grad_norm": 0.4456277523613716,
+      "learning_rate": 5.588108960899069e-06,
+      "loss": 0.0781,
+      "step": 3030
+    },
+    {
+      "epoch": 3.4041836304927697,
+      "grad_norm": 0.45013136112068963,
+      "learning_rate": 5.58107084397646e-06,
+      "loss": 0.0792,
+      "step": 3031
+    },
+    {
+      "epoch": 3.405306752772708,
+      "grad_norm": 0.46808037494992766,
+      "learning_rate": 5.574035446069593e-06,
+      "loss": 0.0827,
+      "step": 3032
+    },
+    {
+      "epoch": 3.4064298750526465,
+      "grad_norm": 0.44644484340636836,
+      "learning_rate": 5.567002771507416e-06,
+      "loss": 0.0778,
+      "step": 3033
+    },
+    {
+      "epoch": 3.4075529973325844,
+      "grad_norm": 0.42953138306404914,
+      "learning_rate": 5.559972824617217e-06,
+      "loss": 0.0798,
+      "step": 3034
+    },
+    {
+      "epoch": 3.408676119612523,
+      "grad_norm": 0.4336870120406191,
+      "learning_rate": 5.552945609724601e-06,
+      "loss": 0.0834,
+      "step": 3035
+    },
+    {
+      "epoch": 3.409799241892461,
+      "grad_norm": 0.4533973433553709,
+      "learning_rate": 5.545921131153487e-06,
+      "loss": 0.0876,
+      "step": 3036
+    },
+    {
+      "epoch": 3.410922364172399,
+      "grad_norm": 0.5136103261731987,
+      "learning_rate": 5.538899393226122e-06,
+      "loss": 0.0853,
+      "step": 3037
+    },
+    {
+      "epoch": 3.4120454864523375,
+      "grad_norm": 0.4320180339485597,
+      "learning_rate": 5.5318804002630465e-06,
+      "loss": 0.0818,
+      "step": 3038
+    },
+    {
+      "epoch": 3.413168608732276,
+      "grad_norm": 0.4378199677787273,
+      "learning_rate": 5.524864156583132e-06,
+      "loss": 0.0809,
+      "step": 3039
+    },
+    {
+      "epoch": 3.414291731012214,
+      "grad_norm": 0.4428589037894112,
+      "learning_rate": 5.517850666503547e-06,
+      "loss": 0.0773,
+      "step": 3040
+    },
+    {
+      "epoch": 3.415414853292152,
+      "grad_norm": 0.43356282741829844,
+      "learning_rate": 5.510839934339771e-06,
+      "loss": 0.0788,
+      "step": 3041
+    },
+    {
+      "epoch": 3.4165379755720906,
+      "grad_norm": 0.4421358222104903,
+      "learning_rate": 5.503831964405588e-06,
+      "loss": 0.0813,
+      "step": 3042
+    },
+    {
+      "epoch": 3.4176610978520285,
+      "grad_norm": 0.45263566082349915,
+      "learning_rate": 5.4968267610130736e-06,
+      "loss": 0.0877,
+      "step": 3043
+    },
+    {
+      "epoch": 3.418784220131967,
+      "grad_norm": 0.42884381173143404,
+      "learning_rate": 5.489824328472606e-06,
+      "loss": 0.0783,
+      "step": 3044
+    },
+    {
+      "epoch": 3.4199073424119053,
+      "grad_norm": 0.45905652234442995,
+      "learning_rate": 5.482824671092862e-06,
+      "loss": 0.0829,
+      "step": 3045
+    },
+    {
+      "epoch": 3.4210304646918432,
+      "grad_norm": 0.43427893805487594,
+      "learning_rate": 5.475827793180808e-06,
+      "loss": 0.0762,
+      "step": 3046
+    },
+    {
+      "epoch": 3.4221535869717816,
+      "grad_norm": 0.4383974688917182,
+      "learning_rate": 5.468833699041702e-06,
+      "loss": 0.0787,
+      "step": 3047
+    },
+    {
+      "epoch": 3.42327670925172,
+      "grad_norm": 0.42636151758099317,
+      "learning_rate": 5.461842392979081e-06,
+      "loss": 0.0801,
+      "step": 3048
+    },
+    {
+      "epoch": 3.424399831531658,
+      "grad_norm": 0.43538975548705666,
+      "learning_rate": 5.454853879294776e-06,
+      "loss": 0.0764,
+      "step": 3049
+    },
+    {
+      "epoch": 3.4255229538115963,
+      "grad_norm": 0.43470614632530913,
+      "learning_rate": 5.447868162288895e-06,
+      "loss": 0.0821,
+      "step": 3050
+    },
+    {
+      "epoch": 3.4266460760915347,
+      "grad_norm": 0.4258384808643859,
+      "learning_rate": 5.440885246259828e-06,
+      "loss": 0.0738,
+      "step": 3051
+    },
+    {
+      "epoch": 3.4277691983714726,
+      "grad_norm": 0.47091027809265673,
+      "learning_rate": 5.433905135504241e-06,
+      "loss": 0.0875,
+      "step": 3052
+    },
+    {
+      "epoch": 3.428892320651411,
+      "grad_norm": 0.42250636073037695,
+      "learning_rate": 5.42692783431707e-06,
+      "loss": 0.0776,
+      "step": 3053
+    },
+    {
+      "epoch": 3.430015442931349,
+      "grad_norm": 0.4418674818368583,
+      "learning_rate": 5.419953346991529e-06,
+      "loss": 0.0753,
+      "step": 3054
+    },
+    {
+      "epoch": 3.4311385652112874,
+      "grad_norm": 0.44643900506814016,
+      "learning_rate": 5.412981677819094e-06,
+      "loss": 0.0839,
+      "step": 3055
+    },
+    {
+      "epoch": 3.4322616874912257,
+      "grad_norm": 0.4509416128414816,
+      "learning_rate": 5.406012831089514e-06,
+      "loss": 0.0786,
+      "step": 3056
+    },
+    {
+      "epoch": 3.4333848097711637,
+      "grad_norm": 0.4526654447129122,
+      "learning_rate": 5.399046811090789e-06,
+      "loss": 0.0825,
+      "step": 3057
+    },
+    {
+      "epoch": 3.434507932051102,
+      "grad_norm": 0.43008105275608804,
+      "learning_rate": 5.392083622109192e-06,
+      "loss": 0.0754,
+      "step": 3058
+    },
+    {
+      "epoch": 3.4356310543310404,
+      "grad_norm": 0.44425939125974356,
+      "learning_rate": 5.385123268429251e-06,
+      "loss": 0.0796,
+      "step": 3059
+    },
+    {
+      "epoch": 3.4367541766109784,
+      "grad_norm": 0.4459296964868783,
+      "learning_rate": 5.3781657543337484e-06,
+      "loss": 0.0806,
+      "step": 3060
+    },
+    {
+      "epoch": 3.4378772988909168,
+      "grad_norm": 0.44459091799716943,
+      "learning_rate": 5.37121108410372e-06,
+      "loss": 0.0811,
+      "step": 3061
+    },
+    {
+      "epoch": 3.439000421170855,
+      "grad_norm": 0.4570923904403929,
+      "learning_rate": 5.364259262018448e-06,
+      "loss": 0.082,
+      "step": 3062
+    },
+    {
+      "epoch": 3.440123543450793,
+      "grad_norm": 0.44006880702846607,
+      "learning_rate": 5.357310292355463e-06,
+      "loss": 0.0791,
+      "step": 3063
+    },
+    {
+      "epoch": 3.4412466657307315,
+      "grad_norm": 0.4606550393242539,
+      "learning_rate": 5.3503641793905485e-06,
+      "loss": 0.0825,
+      "step": 3064
+    },
+    {
+      "epoch": 3.44236978801067,
+      "grad_norm": 0.4254705701825656,
+      "learning_rate": 5.343420927397718e-06,
+      "loss": 0.0859,
+      "step": 3065
+    },
+    {
+      "epoch": 3.443492910290608,
+      "grad_norm": 0.425582782498819,
+      "learning_rate": 5.33648054064924e-06,
+      "loss": 0.0803,
+      "step": 3066
+    },
+    {
+      "epoch": 3.444616032570546,
+      "grad_norm": 0.4400458710522822,
+      "learning_rate": 5.329543023415602e-06,
+      "loss": 0.0805,
+      "step": 3067
+    },
+    {
+      "epoch": 3.445739154850484,
+      "grad_norm": 0.4176064311503563,
+      "learning_rate": 5.322608379965537e-06,
+      "loss": 0.075,
+      "step": 3068
+    },
+    {
+      "epoch": 3.4468622771304225,
+      "grad_norm": 0.45054223519188685,
+      "learning_rate": 5.315676614566008e-06,
+      "loss": 0.0799,
+      "step": 3069
+    },
+    {
+      "epoch": 3.447985399410361,
+      "grad_norm": 0.4423125763067608,
+      "learning_rate": 5.308747731482207e-06,
+      "loss": 0.0787,
+      "step": 3070
+    },
+    {
+      "epoch": 3.449108521690299,
+      "grad_norm": 0.4216181686130632,
+      "learning_rate": 5.301821734977555e-06,
+      "loss": 0.0755,
+      "step": 3071
+    },
+    {
+      "epoch": 3.450231643970237,
+      "grad_norm": 0.4484194253723383,
+      "learning_rate": 5.2948986293136876e-06,
+      "loss": 0.0805,
+      "step": 3072
+    },
+    {
+      "epoch": 3.4513547662501756,
+      "grad_norm": 0.46771531925041326,
+      "learning_rate": 5.28797841875047e-06,
+      "loss": 0.0813,
+      "step": 3073
+    },
+    {
+      "epoch": 3.4524778885301135,
+      "grad_norm": 0.4340387002286317,
+      "learning_rate": 5.281061107545985e-06,
+      "loss": 0.0784,
+      "step": 3074
+    },
+    {
+      "epoch": 3.453601010810052,
+      "grad_norm": 0.4312273423989433,
+      "learning_rate": 5.274146699956531e-06,
+      "loss": 0.0784,
+      "step": 3075
+    },
+    {
+      "epoch": 3.4547241330899903,
+      "grad_norm": 0.45797427049587475,
+      "learning_rate": 5.26723520023662e-06,
+      "loss": 0.0841,
+      "step": 3076
+    },
+    {
+      "epoch": 3.455847255369928,
+      "grad_norm": 0.4478121802861111,
+      "learning_rate": 5.260326612638971e-06,
+      "loss": 0.0779,
+      "step": 3077
+    },
+    {
+      "epoch": 3.4569703776498666,
+      "grad_norm": 0.4352907201834758,
+      "learning_rate": 5.253420941414513e-06,
+      "loss": 0.0757,
+      "step": 3078
+    },
+    {
+      "epoch": 3.458093499929805,
+      "grad_norm": 0.438323892417082,
+      "learning_rate": 5.246518190812384e-06,
+      "loss": 0.077,
+      "step": 3079
+    },
+    {
+      "epoch": 3.459216622209743,
+      "grad_norm": 0.45721881289918465,
+      "learning_rate": 5.239618365079921e-06,
+      "loss": 0.086,
+      "step": 3080
+    },
+    {
+      "epoch": 3.4603397444896813,
+      "grad_norm": 0.45134533377379477,
+      "learning_rate": 5.232721468462669e-06,
+      "loss": 0.0837,
+      "step": 3081
+    },
+    {
+      "epoch": 3.4614628667696197,
+      "grad_norm": 0.4427173473895729,
+      "learning_rate": 5.225827505204355e-06,
+      "loss": 0.0743,
+      "step": 3082
+    },
+    {
+      "epoch": 3.4625859890495576,
+      "grad_norm": 0.4591916370110219,
+      "learning_rate": 5.218936479546913e-06,
+      "loss": 0.0832,
+      "step": 3083
+    },
+    {
+      "epoch": 3.463709111329496,
+      "grad_norm": 0.45693111857381774,
+      "learning_rate": 5.212048395730469e-06,
+      "loss": 0.0768,
+      "step": 3084
+    },
+    {
+      "epoch": 3.4648322336094344,
+      "grad_norm": 0.4522432872093845,
+      "learning_rate": 5.205163257993341e-06,
+      "loss": 0.0822,
+      "step": 3085
+    },
+    {
+      "epoch": 3.4659553558893723,
+      "grad_norm": 0.45600705538504716,
+      "learning_rate": 5.19828107057202e-06,
+      "loss": 0.0798,
+      "step": 3086
+    },
+    {
+      "epoch": 3.4670784781693107,
+      "grad_norm": 0.45123885398109936,
+      "learning_rate": 5.191401837701197e-06,
+      "loss": 0.0823,
+      "step": 3087
+    },
+    {
+      "epoch": 3.468201600449249,
+      "grad_norm": 0.4480694693802906,
+      "learning_rate": 5.18452556361374e-06,
+      "loss": 0.0768,
+      "step": 3088
+    },
+    {
+      "epoch": 3.469324722729187,
+      "grad_norm": 0.443774248075364,
+      "learning_rate": 5.177652252540697e-06,
+      "loss": 0.078,
+      "step": 3089
+    },
+    {
+      "epoch": 3.4704478450091254,
+      "grad_norm": 0.4651048676199529,
+      "learning_rate": 5.170781908711289e-06,
+      "loss": 0.0814,
+      "step": 3090
+    },
+    {
+      "epoch": 3.471570967289064,
+      "grad_norm": 0.4858930041394996,
+      "learning_rate": 5.163914536352919e-06,
+      "loss": 0.0831,
+      "step": 3091
+    },
+    {
+      "epoch": 3.4726940895690017,
+      "grad_norm": 0.45944862905297834,
+      "learning_rate": 5.157050139691151e-06,
+      "loss": 0.0847,
+      "step": 3092
+    },
+    {
+      "epoch": 3.47381721184894,
+      "grad_norm": 0.43905597445283906,
+      "learning_rate": 5.150188722949725e-06,
+      "loss": 0.0791,
+      "step": 3093
+    },
+    {
+      "epoch": 3.4749403341288785,
+      "grad_norm": 0.43904792852959645,
+      "learning_rate": 5.143330290350548e-06,
+      "loss": 0.0767,
+      "step": 3094
+    },
+    {
+      "epoch": 3.4760634564088164,
+      "grad_norm": 0.4487339919699267,
+      "learning_rate": 5.136474846113688e-06,
+      "loss": 0.0811,
+      "step": 3095
+    },
+    {
+      "epoch": 3.477186578688755,
+      "grad_norm": 0.4280955715492865,
+      "learning_rate": 5.129622394457377e-06,
+      "loss": 0.074,
+      "step": 3096
+    },
+    {
+      "epoch": 3.478309700968693,
+      "grad_norm": 0.44114191944550274,
+      "learning_rate": 5.122772939598003e-06,
+      "loss": 0.0811,
+      "step": 3097
+    },
+    {
+      "epoch": 3.479432823248631,
+      "grad_norm": 0.4381981300103117,
+      "learning_rate": 5.11592648575011e-06,
+      "loss": 0.0769,
+      "step": 3098
+    },
+    {
+      "epoch": 3.4805559455285695,
+      "grad_norm": 0.41185510445743156,
+      "learning_rate": 5.109083037126397e-06,
+      "loss": 0.0731,
+      "step": 3099
+    },
+    {
+      "epoch": 3.481679067808508,
+      "grad_norm": 0.4438897661024262,
+      "learning_rate": 5.1022425979377174e-06,
+      "loss": 0.0809,
+      "step": 3100
+    },
+    {
+      "epoch": 3.482802190088446,
+      "grad_norm": 0.43375334791452175,
+      "learning_rate": 5.095405172393062e-06,
+      "loss": 0.0828,
+      "step": 3101
+    },
+    {
+      "epoch": 3.483925312368384,
+      "grad_norm": 0.4326713003014848,
+      "learning_rate": 5.088570764699574e-06,
+      "loss": 0.0741,
+      "step": 3102
+    },
+    {
+      "epoch": 3.485048434648322,
+      "grad_norm": 0.4553368231780253,
+      "learning_rate": 5.081739379062545e-06,
+      "loss": 0.0797,
+      "step": 3103
+    },
+    {
+      "epoch": 3.4861715569282605,
+      "grad_norm": 0.45948155021628384,
+      "learning_rate": 5.074911019685398e-06,
+      "loss": 0.079,
+      "step": 3104
+    },
+    {
+      "epoch": 3.487294679208199,
+      "grad_norm": 0.4435769006877875,
+      "learning_rate": 5.068085690769702e-06,
+      "loss": 0.0828,
+      "step": 3105
+    },
+    {
+      "epoch": 3.488417801488137,
+      "grad_norm": 0.4676805551105927,
+      "learning_rate": 5.06126339651515e-06,
+      "loss": 0.0876,
+      "step": 3106
+    },
+    {
+      "epoch": 3.4895409237680752,
+      "grad_norm": 0.44708470564753106,
+      "learning_rate": 5.054444141119579e-06,
+      "loss": 0.0789,
+      "step": 3107
+    },
+    {
+      "epoch": 3.4906640460480136,
+      "grad_norm": 0.44694690179062085,
+      "learning_rate": 5.047627928778951e-06,
+      "loss": 0.0797,
+      "step": 3108
+    },
+    {
+      "epoch": 3.4917871683279516,
+      "grad_norm": 0.45987872193849977,
+      "learning_rate": 5.040814763687358e-06,
+      "loss": 0.0804,
+      "step": 3109
+    },
+    {
+      "epoch": 3.49291029060789,
+      "grad_norm": 0.46368555836155595,
+      "learning_rate": 5.034004650037016e-06,
+      "loss": 0.0838,
+      "step": 3110
+    },
+    {
+      "epoch": 3.4940334128878283,
+      "grad_norm": 0.4670280413811003,
+      "learning_rate": 5.02719759201826e-06,
+      "loss": 0.0854,
+      "step": 3111
+    },
+    {
+      "epoch": 3.4951565351677663,
+      "grad_norm": 0.4655508082978832,
+      "learning_rate": 5.020393593819547e-06,
+      "loss": 0.085,
+      "step": 3112
+    },
+    {
+      "epoch": 3.4962796574477046,
+      "grad_norm": 0.4506202886596756,
+      "learning_rate": 5.013592659627454e-06,
+      "loss": 0.0796,
+      "step": 3113
+    },
+    {
+      "epoch": 3.497402779727643,
+      "grad_norm": 0.44760914223329473,
+      "learning_rate": 5.006794793626671e-06,
+      "loss": 0.0773,
+      "step": 3114
+    },
+    {
+      "epoch": 3.498525902007581,
+      "grad_norm": 0.44369165321868126,
+      "learning_rate": 5.000000000000003e-06,
+      "loss": 0.0787,
+      "step": 3115
+    },
+    {
+      "epoch": 3.4996490242875193,
+      "grad_norm": 0.42826531752580976,
+      "learning_rate": 4.9932082829283524e-06,
+      "loss": 0.0812,
+      "step": 3116
+    },
+    {
+      "epoch": 3.5007721465674573,
+      "grad_norm": 0.4370975011654157,
+      "learning_rate": 4.986419646590744e-06,
+      "loss": 0.0811,
+      "step": 3117
+    },
+    {
+      "epoch": 3.5018952688473957,
+      "grad_norm": 0.4431924357682965,
+      "learning_rate": 4.979634095164298e-06,
+      "loss": 0.0811,
+      "step": 3118
+    },
+    {
+      "epoch": 3.503018391127334,
+      "grad_norm": 0.43224211139362045,
+      "learning_rate": 4.972851632824241e-06,
+      "loss": 0.0797,
+      "step": 3119
+    },
+    {
+      "epoch": 3.504141513407272,
+      "grad_norm": 0.44424653895974464,
+      "learning_rate": 4.966072263743899e-06,
+      "loss": 0.0828,
+      "step": 3120
+    },
+    {
+      "epoch": 3.5052646356872104,
+      "grad_norm": 0.4572424329181564,
+      "learning_rate": 4.959295992094685e-06,
+      "loss": 0.0801,
+      "step": 3121
+    },
+    {
+      "epoch": 3.5063877579671487,
+      "grad_norm": 0.460448950085812,
+      "learning_rate": 4.952522822046117e-06,
+      "loss": 0.0764,
+      "step": 3122
+    },
+    {
+      "epoch": 3.5075108802470867,
+      "grad_norm": 0.4249602220115872,
+      "learning_rate": 4.945752757765802e-06,
+      "loss": 0.0771,
+      "step": 3123
+    },
+    {
+      "epoch": 3.508634002527025,
+      "grad_norm": 0.4413605494574981,
+      "learning_rate": 4.93898580341944e-06,
+      "loss": 0.0822,
+      "step": 3124
+    },
+    {
+      "epoch": 3.5097571248069634,
+      "grad_norm": 0.4291289570379504,
+      "learning_rate": 4.932221963170801e-06,
+      "loss": 0.0797,
+      "step": 3125
+    },
+    {
+      "epoch": 3.5108802470869014,
+      "grad_norm": 0.4721851004921407,
+      "learning_rate": 4.925461241181757e-06,
+      "loss": 0.0843,
+      "step": 3126
+    },
+    {
+      "epoch": 3.5120033693668398,
+      "grad_norm": 0.4392592949377926,
+      "learning_rate": 4.918703641612255e-06,
+      "loss": 0.0786,
+      "step": 3127
+    },
+    {
+      "epoch": 3.513126491646778,
+      "grad_norm": 0.43903462957651435,
+      "learning_rate": 4.9119491686203195e-06,
+      "loss": 0.0804,
+      "step": 3128
+    },
+    {
+      "epoch": 3.514249613926716,
+      "grad_norm": 0.4430529984909637,
+      "learning_rate": 4.9051978263620545e-06,
+      "loss": 0.0766,
+      "step": 3129
+    },
+    {
+      "epoch": 3.5153727362066545,
+      "grad_norm": 0.4394580449820179,
+      "learning_rate": 4.89844961899163e-06,
+      "loss": 0.0762,
+      "step": 3130
+    },
+    {
+      "epoch": 3.516495858486593,
+      "grad_norm": 0.4686215670552109,
+      "learning_rate": 4.891704550661294e-06,
+      "loss": 0.0847,
+      "step": 3131
+    },
+    {
+      "epoch": 3.517618980766531,
+      "grad_norm": 0.41231062902627313,
+      "learning_rate": 4.884962625521363e-06,
+      "loss": 0.0793,
+      "step": 3132
+    },
+    {
+      "epoch": 3.518742103046469,
+      "grad_norm": 0.4569282712376777,
+      "learning_rate": 4.878223847720217e-06,
+      "loss": 0.0876,
+      "step": 3133
+    },
+    {
+      "epoch": 3.5198652253264076,
+      "grad_norm": 0.4208575740257057,
+      "learning_rate": 4.8714882214043e-06,
+      "loss": 0.078,
+      "step": 3134
+    },
+    {
+      "epoch": 3.5209883476063455,
+      "grad_norm": 0.43267410608868484,
+      "learning_rate": 4.8647557507181164e-06,
+      "loss": 0.0814,
+      "step": 3135
+    },
+    {
+      "epoch": 3.522111469886284,
+      "grad_norm": 0.42634413807778826,
+      "learning_rate": 4.8580264398042355e-06,
+      "loss": 0.0745,
+      "step": 3136
+    },
+    {
+      "epoch": 3.5232345921662223,
+      "grad_norm": 0.45182552642987417,
+      "learning_rate": 4.851300292803266e-06,
+      "loss": 0.081,
+      "step": 3137
+    },
+    {
+      "epoch": 3.52435771444616,
+      "grad_norm": 0.43109856616030523,
+      "learning_rate": 4.844577313853886e-06,
+      "loss": 0.0814,
+      "step": 3138
+    },
+    {
+      "epoch": 3.5254808367260986,
+      "grad_norm": 0.4470250084652514,
+      "learning_rate": 4.837857507092817e-06,
+      "loss": 0.0822,
+      "step": 3139
+    },
+    {
+      "epoch": 3.526603959006037,
+      "grad_norm": 0.44862752921716353,
+      "learning_rate": 4.831140876654831e-06,
+      "loss": 0.0802,
+      "step": 3140
+    },
+    {
+      "epoch": 3.527727081285975,
+      "grad_norm": 0.4769651304639455,
+      "learning_rate": 4.824427426672743e-06,
+      "loss": 0.083,
+      "step": 3141
+    },
+    {
+      "epoch": 3.5288502035659133,
+      "grad_norm": 0.4709080102440193,
+      "learning_rate": 4.8177171612774155e-06,
+      "loss": 0.0833,
+      "step": 3142
+    },
+    {
+      "epoch": 3.5299733258458517,
+      "grad_norm": 0.4393104624450094,
+      "learning_rate": 4.811010084597747e-06,
+      "loss": 0.0729,
+      "step": 3143
+    },
+    {
+      "epoch": 3.5310964481257896,
+      "grad_norm": 0.4600533673879964,
+      "learning_rate": 4.80430620076068e-06,
+      "loss": 0.0807,
+      "step": 3144
+    },
+    {
+      "epoch": 3.532219570405728,
+      "grad_norm": 0.47092645697860963,
+      "learning_rate": 4.797605513891179e-06,
+      "loss": 0.0786,
+      "step": 3145
+    },
+    {
+      "epoch": 3.5333426926856664,
+      "grad_norm": 0.44424095531161306,
+      "learning_rate": 4.790908028112256e-06,
+      "loss": 0.0817,
+      "step": 3146
+    },
+    {
+      "epoch": 3.5344658149656043,
+      "grad_norm": 0.4536227579660774,
+      "learning_rate": 4.7842137475449444e-06,
+      "loss": 0.0782,
+      "step": 3147
+    },
+    {
+      "epoch": 3.5355889372455427,
+      "grad_norm": 0.45564989250792665,
+      "learning_rate": 4.777522676308314e-06,
+      "loss": 0.0823,
+      "step": 3148
+    },
+    {
+      "epoch": 3.536712059525481,
+      "grad_norm": 0.42912926619367286,
+      "learning_rate": 4.770834818519454e-06,
+      "loss": 0.0815,
+      "step": 3149
+    },
+    {
+      "epoch": 3.537835181805419,
+      "grad_norm": 0.44561480322449887,
+      "learning_rate": 4.764150178293471e-06,
+      "loss": 0.0783,
+      "step": 3150
+    },
+    {
+      "epoch": 3.5389583040853574,
+      "grad_norm": 0.4553565188197437,
+      "learning_rate": 4.757468759743501e-06,
+      "loss": 0.0837,
+      "step": 3151
+    },
+    {
+      "epoch": 3.5400814263652958,
+      "grad_norm": 0.46776113922754914,
+      "learning_rate": 4.750790566980694e-06,
+      "loss": 0.0809,
+      "step": 3152
+    },
+    {
+      "epoch": 3.5412045486452337,
+      "grad_norm": 0.43159302458196586,
+      "learning_rate": 4.744115604114218e-06,
+      "loss": 0.0775,
+      "step": 3153
+    },
+    {
+      "epoch": 3.542327670925172,
+      "grad_norm": 0.451930193919335,
+      "learning_rate": 4.737443875251251e-06,
+      "loss": 0.076,
+      "step": 3154
+    },
+    {
+      "epoch": 3.5434507932051105,
+      "grad_norm": 0.46172685056937696,
+      "learning_rate": 4.730775384496976e-06,
+      "loss": 0.0796,
+      "step": 3155
+    },
+    {
+      "epoch": 3.5445739154850484,
+      "grad_norm": 0.4262475873889323,
+      "learning_rate": 4.724110135954593e-06,
+      "loss": 0.076,
+      "step": 3156
+    },
+    {
+      "epoch": 3.545697037764987,
+      "grad_norm": 0.4533319005981512,
+      "learning_rate": 4.717448133725302e-06,
+      "loss": 0.0864,
+      "step": 3157
+    },
+    {
+      "epoch": 3.5468201600449247,
+      "grad_norm": 0.44873241402246483,
+      "learning_rate": 4.710789381908308e-06,
+      "loss": 0.0827,
+      "step": 3158
+    },
+    {
+      "epoch": 3.547943282324863,
+      "grad_norm": 0.4438311385849649,
+      "learning_rate": 4.704133884600811e-06,
+      "loss": 0.0807,
+      "step": 3159
+    },
+    {
+      "epoch": 3.5490664046048015,
+      "grad_norm": 0.42639669279917847,
+      "learning_rate": 4.697481645898012e-06,
+      "loss": 0.0789,
+      "step": 3160
+    },
+    {
+      "epoch": 3.5501895268847394,
+      "grad_norm": 0.410746116785441,
+      "learning_rate": 4.690832669893108e-06,
+      "loss": 0.0805,
+      "step": 3161
+    },
+    {
+      "epoch": 3.551312649164678,
+      "grad_norm": 0.45365920946155613,
+      "learning_rate": 4.684186960677287e-06,
+      "loss": 0.0866,
+      "step": 3162
+    },
+    {
+      "epoch": 3.552435771444616,
+      "grad_norm": 0.46568428236414383,
+      "learning_rate": 4.6775445223397306e-06,
+      "loss": 0.084,
+      "step": 3163
+    },
+    {
+      "epoch": 3.553558893724554,
+      "grad_norm": 0.45425121145084363,
+      "learning_rate": 4.670905358967598e-06,
+      "loss": 0.0856,
+      "step": 3164
+    },
+    {
+      "epoch": 3.5546820160044925,
+      "grad_norm": 0.44337140062483155,
+      "learning_rate": 4.66426947464604e-06,
+      "loss": 0.0823,
+      "step": 3165
+    },
+    {
+      "epoch": 3.5558051382844305,
+      "grad_norm": 0.45052653263910536,
+      "learning_rate": 4.6576368734581935e-06,
+      "loss": 0.0829,
+      "step": 3166
+    },
+    {
+      "epoch": 3.556928260564369,
+      "grad_norm": 0.4421176966358397,
+      "learning_rate": 4.651007559485168e-06,
+      "loss": 0.081,
+      "step": 3167
+    },
+    {
+      "epoch": 3.558051382844307,
+      "grad_norm": 0.4417611457609658,
+      "learning_rate": 4.644381536806058e-06,
+      "loss": 0.0798,
+      "step": 3168
+    },
+    {
+      "epoch": 3.559174505124245,
+      "grad_norm": 0.4362310044034491,
+      "learning_rate": 4.637758809497919e-06,
+      "loss": 0.0789,
+      "step": 3169
+    },
+    {
+      "epoch": 3.5602976274041835,
+      "grad_norm": 0.48187030579374374,
+      "learning_rate": 4.631139381635795e-06,
+      "loss": 0.0826,
+      "step": 3170
+    },
+    {
+      "epoch": 3.561420749684122,
+      "grad_norm": 0.4338280376779694,
+      "learning_rate": 4.62452325729269e-06,
+      "loss": 0.0782,
+      "step": 3171
+    },
+    {
+      "epoch": 3.56254387196406,
+      "grad_norm": 0.44241538181060147,
+      "learning_rate": 4.61791044053958e-06,
+      "loss": 0.0813,
+      "step": 3172
+    },
+    {
+      "epoch": 3.5636669942439982,
+      "grad_norm": 0.4572788174432963,
+      "learning_rate": 4.611300935445407e-06,
+      "loss": 0.0822,
+      "step": 3173
+    },
+    {
+      "epoch": 3.5647901165239366,
+      "grad_norm": 0.4441694387134764,
+      "learning_rate": 4.604694746077064e-06,
+      "loss": 0.0838,
+      "step": 3174
+    },
+    {
+      "epoch": 3.5659132388038746,
+      "grad_norm": 0.45759098365330836,
+      "learning_rate": 4.598091876499417e-06,
+      "loss": 0.0799,
+      "step": 3175
+    },
+    {
+      "epoch": 3.567036361083813,
+      "grad_norm": 0.4343009539002791,
+      "learning_rate": 4.591492330775283e-06,
+      "loss": 0.0756,
+      "step": 3176
+    },
+    {
+      "epoch": 3.5681594833637513,
+      "grad_norm": 0.4473461995617762,
+      "learning_rate": 4.5848961129654365e-06,
+      "loss": 0.0812,
+      "step": 3177
+    },
+    {
+      "epoch": 3.5692826056436893,
+      "grad_norm": 0.4546004165413151,
+      "learning_rate": 4.578303227128603e-06,
+      "loss": 0.0833,
+      "step": 3178
+    },
+    {
+      "epoch": 3.5704057279236276,
+      "grad_norm": 0.46424501547792263,
+      "learning_rate": 4.571713677321455e-06,
+      "loss": 0.0812,
+      "step": 3179
+    },
+    {
+      "epoch": 3.571528850203566,
+      "grad_norm": 0.4361206138265064,
+      "learning_rate": 4.565127467598619e-06,
+      "loss": 0.0774,
+      "step": 3180
+    },
+    {
+      "epoch": 3.572651972483504,
+      "grad_norm": 0.4286236286675256,
+      "learning_rate": 4.5585446020126634e-06,
+      "loss": 0.0733,
+      "step": 3181
+    },
+    {
+      "epoch": 3.5737750947634424,
+      "grad_norm": 0.4617841876606261,
+      "learning_rate": 4.551965084614089e-06,
+      "loss": 0.083,
+      "step": 3182
+    },
+    {
+      "epoch": 3.5748982170433807,
+      "grad_norm": 0.46431636230368917,
+      "learning_rate": 4.545388919451353e-06,
+      "loss": 0.0772,
+      "step": 3183
+    },
+    {
+      "epoch": 3.5760213393233187,
+      "grad_norm": 0.42398130900439096,
+      "learning_rate": 4.538816110570841e-06,
+      "loss": 0.0787,
+      "step": 3184
+    },
+    {
+      "epoch": 3.577144461603257,
+      "grad_norm": 0.4340576396237869,
+      "learning_rate": 4.532246662016872e-06,
+      "loss": 0.0792,
+      "step": 3185
+    },
+    {
+      "epoch": 3.5782675838831954,
+      "grad_norm": 0.4706268135249468,
+      "learning_rate": 4.5256805778317015e-06,
+      "loss": 0.0879,
+      "step": 3186
+    },
+    {
+      "epoch": 3.5793907061631334,
+      "grad_norm": 0.43976864195858995,
+      "learning_rate": 4.519117862055514e-06,
+      "loss": 0.0811,
+      "step": 3187
+    },
+    {
+      "epoch": 3.5805138284430718,
+      "grad_norm": 0.45376420671959067,
+      "learning_rate": 4.512558518726425e-06,
+      "loss": 0.0806,
+      "step": 3188
+    },
+    {
+      "epoch": 3.58163695072301,
+      "grad_norm": 0.468528511091486,
+      "learning_rate": 4.506002551880462e-06,
+      "loss": 0.0832,
+      "step": 3189
+    },
+    {
+      "epoch": 3.582760073002948,
+      "grad_norm": 0.43294967832559633,
+      "learning_rate": 4.4994499655515865e-06,
+      "loss": 0.074,
+      "step": 3190
+    },
+    {
+      "epoch": 3.5838831952828865,
+      "grad_norm": 0.42827553987381783,
+      "learning_rate": 4.492900763771679e-06,
+      "loss": 0.0746,
+      "step": 3191
+    },
+    {
+      "epoch": 3.585006317562825,
+      "grad_norm": 0.4389937609121276,
+      "learning_rate": 4.486354950570534e-06,
+      "loss": 0.0768,
+      "step": 3192
+    },
+    {
+      "epoch": 3.586129439842763,
+      "grad_norm": 0.44454906671504224,
+      "learning_rate": 4.4798125299758666e-06,
+      "loss": 0.0741,
+      "step": 3193
+    },
+    {
+      "epoch": 3.587252562122701,
+      "grad_norm": 0.4423414531424126,
+      "learning_rate": 4.473273506013294e-06,
+      "loss": 0.0785,
+      "step": 3194
+    },
+    {
+      "epoch": 3.5883756844026395,
+      "grad_norm": 0.45927341409394545,
+      "learning_rate": 4.46673788270635e-06,
+      "loss": 0.088,
+      "step": 3195
+    },
+    {
+      "epoch": 3.5894988066825775,
+      "grad_norm": 0.4420335156563938,
+      "learning_rate": 4.460205664076479e-06,
+      "loss": 0.0767,
+      "step": 3196
+    },
+    {
+      "epoch": 3.590621928962516,
+      "grad_norm": 0.46538751797421757,
+      "learning_rate": 4.453676854143029e-06,
+      "loss": 0.0848,
+      "step": 3197
+    },
+    {
+      "epoch": 3.5917450512424542,
+      "grad_norm": 0.45668682965397933,
+      "learning_rate": 4.447151456923241e-06,
+      "loss": 0.0815,
+      "step": 3198
+    },
+    {
+      "epoch": 3.592868173522392,
+      "grad_norm": 0.45937139886413936,
+      "learning_rate": 4.440629476432268e-06,
+      "loss": 0.079,
+      "step": 3199
+    },
+    {
+      "epoch": 3.5939912958023306,
+      "grad_norm": 0.4309123966029538,
+      "learning_rate": 4.4341109166831565e-06,
+      "loss": 0.0752,
+      "step": 3200
+    },
+    {
+      "epoch": 3.595114418082269,
+      "grad_norm": 0.45123899659954364,
+      "learning_rate": 4.427595781686848e-06,
+      "loss": 0.0819,
+      "step": 3201
+    },
+    {
+      "epoch": 3.596237540362207,
+      "grad_norm": 0.4301342537854829,
+      "learning_rate": 4.42108407545218e-06,
+      "loss": 0.0792,
+      "step": 3202
+    },
+    {
+      "epoch": 3.5973606626421453,
+      "grad_norm": 0.4234732951722343,
+      "learning_rate": 4.41457580198587e-06,
+      "loss": 0.0759,
+      "step": 3203
+    },
+    {
+      "epoch": 3.5984837849220837,
+      "grad_norm": 0.4267600293298272,
+      "learning_rate": 4.408070965292534e-06,
+      "loss": 0.08,
+      "step": 3204
+    },
+    {
+      "epoch": 3.5996069072020216,
+      "grad_norm": 0.43857035097482017,
+      "learning_rate": 4.4015695693746685e-06,
+      "loss": 0.0782,
+      "step": 3205
+    },
+    {
+      "epoch": 3.60073002948196,
+      "grad_norm": 0.4662120376628465,
+      "learning_rate": 4.395071618232656e-06,
+      "loss": 0.0829,
+      "step": 3206
+    },
+    {
+      "epoch": 3.6018531517618984,
+      "grad_norm": 0.4353076779897321,
+      "learning_rate": 4.3885771158647595e-06,
+      "loss": 0.0766,
+      "step": 3207
+    },
+    {
+      "epoch": 3.6029762740418363,
+      "grad_norm": 0.420745107025468,
+      "learning_rate": 4.38208606626711e-06,
+      "loss": 0.0788,
+      "step": 3208
+    },
+    {
+      "epoch": 3.6040993963217747,
+      "grad_norm": 0.43420825398499724,
+      "learning_rate": 4.375598473433727e-06,
+      "loss": 0.082,
+      "step": 3209
+    },
+    {
+      "epoch": 3.6052225186017126,
+      "grad_norm": 0.43624668378544307,
+      "learning_rate": 4.369114341356497e-06,
+      "loss": 0.0782,
+      "step": 3210
+    },
+    {
+      "epoch": 3.606345640881651,
+      "grad_norm": 0.4286224207837259,
+      "learning_rate": 4.362633674025178e-06,
+      "loss": 0.0752,
+      "step": 3211
+    },
+    {
+      "epoch": 3.6074687631615894,
+      "grad_norm": 0.45618447258162387,
+      "learning_rate": 4.3561564754274e-06,
+      "loss": 0.0821,
+      "step": 3212
+    },
+    {
+      "epoch": 3.6085918854415273,
+      "grad_norm": 0.4245976031185963,
+      "learning_rate": 4.349682749548647e-06,
+      "loss": 0.0782,
+      "step": 3213
+    },
+    {
+      "epoch": 3.6097150077214657,
+      "grad_norm": 0.4576618374401797,
+      "learning_rate": 4.3432125003722754e-06,
+      "loss": 0.0843,
+      "step": 3214
+    },
+    {
+      "epoch": 3.6108381300014036,
+      "grad_norm": 0.48128056616977466,
+      "learning_rate": 4.3367457318795034e-06,
+      "loss": 0.0855,
+      "step": 3215
+    },
+    {
+      "epoch": 3.611961252281342,
+      "grad_norm": 0.4573589667687167,
+      "learning_rate": 4.330282448049405e-06,
+      "loss": 0.0869,
+      "step": 3216
+    },
+    {
+      "epoch": 3.6130843745612804,
+      "grad_norm": 0.41682956812824573,
+      "learning_rate": 4.323822652858911e-06,
+      "loss": 0.079,
+      "step": 3217
+    },
+    {
+      "epoch": 3.6142074968412183,
+      "grad_norm": 0.44417693491582083,
+      "learning_rate": 4.3173663502827985e-06,
+      "loss": 0.0795,
+      "step": 3218
+    },
+    {
+      "epoch": 3.6153306191211567,
+      "grad_norm": 0.45615707711819087,
+      "learning_rate": 4.310913544293706e-06,
+      "loss": 0.0842,
+      "step": 3219
+    },
+    {
+      "epoch": 3.616453741401095,
+      "grad_norm": 0.46322398924470976,
+      "learning_rate": 4.304464238862115e-06,
+      "loss": 0.0813,
+      "step": 3220
+    },
+    {
+      "epoch": 3.617576863681033,
+      "grad_norm": 0.43509351671563573,
+      "learning_rate": 4.298018437956354e-06,
+      "loss": 0.0788,
+      "step": 3221
+    },
+    {
+      "epoch": 3.6186999859609714,
+      "grad_norm": 0.43286932500602177,
+      "learning_rate": 4.2915761455425965e-06,
+      "loss": 0.0739,
+      "step": 3222
+    },
+    {
+      "epoch": 3.61982310824091,
+      "grad_norm": 0.4596133947427022,
+      "learning_rate": 4.285137365584854e-06,
+      "loss": 0.0875,
+      "step": 3223
+    },
+    {
+      "epoch": 3.6209462305208477,
+      "grad_norm": 0.4481272875661536,
+      "learning_rate": 4.27870210204498e-06,
+      "loss": 0.0832,
+      "step": 3224
+    },
+    {
+      "epoch": 3.622069352800786,
+      "grad_norm": 0.43075573098301184,
+      "learning_rate": 4.272270358882667e-06,
+      "loss": 0.0785,
+      "step": 3225
+    },
+    {
+      "epoch": 3.6231924750807245,
+      "grad_norm": 0.44064062012364946,
+      "learning_rate": 4.265842140055428e-06,
+      "loss": 0.0743,
+      "step": 3226
+    },
+    {
+      "epoch": 3.6243155973606624,
+      "grad_norm": 0.44599199599462813,
+      "learning_rate": 4.2594174495186225e-06,
+      "loss": 0.0768,
+      "step": 3227
+    },
+    {
+      "epoch": 3.625438719640601,
+      "grad_norm": 0.4756212129886129,
+      "learning_rate": 4.252996291225433e-06,
+      "loss": 0.0832,
+      "step": 3228
+    },
+    {
+      "epoch": 3.626561841920539,
+      "grad_norm": 0.44272449482968135,
+      "learning_rate": 4.2465786691268675e-06,
+      "loss": 0.0798,
+      "step": 3229
+    },
+    {
+      "epoch": 3.627684964200477,
+      "grad_norm": 0.44797258468509793,
+      "learning_rate": 4.240164587171761e-06,
+      "loss": 0.0745,
+      "step": 3230
+    },
+    {
+      "epoch": 3.6288080864804155,
+      "grad_norm": 0.44058249923913984,
+      "learning_rate": 4.233754049306772e-06,
+      "loss": 0.076,
+      "step": 3231
+    },
+    {
+      "epoch": 3.629931208760354,
+      "grad_norm": 0.43397062915249096,
+      "learning_rate": 4.227347059476367e-06,
+      "loss": 0.078,
+      "step": 3232
+    },
+    {
+      "epoch": 3.631054331040292,
+      "grad_norm": 0.43034776436202615,
+      "learning_rate": 4.220943621622841e-06,
+      "loss": 0.0777,
+      "step": 3233
+    },
+    {
+      "epoch": 3.6321774533202302,
+      "grad_norm": 0.46052076524797797,
+      "learning_rate": 4.214543739686302e-06,
+      "loss": 0.0769,
+      "step": 3234
+    },
+    {
+      "epoch": 3.6333005756001686,
+      "grad_norm": 0.4761059654293087,
+      "learning_rate": 4.208147417604665e-06,
+      "loss": 0.0866,
+      "step": 3235
+    },
+    {
+      "epoch": 3.6344236978801066,
+      "grad_norm": 0.4582874720771719,
+      "learning_rate": 4.2017546593136615e-06,
+      "loss": 0.0816,
+      "step": 3236
+    },
+    {
+      "epoch": 3.635546820160045,
+      "grad_norm": 0.4596850731102858,
+      "learning_rate": 4.195365468746821e-06,
+      "loss": 0.0836,
+      "step": 3237
+    },
+    {
+      "epoch": 3.6366699424399833,
+      "grad_norm": 0.4451422703440366,
+      "learning_rate": 4.188979849835483e-06,
+      "loss": 0.084,
+      "step": 3238
+    },
+    {
+      "epoch": 3.6377930647199213,
+      "grad_norm": 0.45215038572014205,
+      "learning_rate": 4.182597806508791e-06,
+      "loss": 0.0803,
+      "step": 3239
+    },
+    {
+      "epoch": 3.6389161869998596,
+      "grad_norm": 0.4545896148997309,
+      "learning_rate": 4.176219342693687e-06,
+      "loss": 0.0817,
+      "step": 3240
+    },
+    {
+      "epoch": 3.640039309279798,
+      "grad_norm": 0.4415219136001395,
+      "learning_rate": 4.1698444623149125e-06,
+      "loss": 0.0777,
+      "step": 3241
+    },
+    {
+      "epoch": 3.641162431559736,
+      "grad_norm": 0.4290457562106187,
+      "learning_rate": 4.163473169294995e-06,
+      "loss": 0.0756,
+      "step": 3242
+    },
+    {
+      "epoch": 3.6422855538396743,
+      "grad_norm": 0.4362701926456306,
+      "learning_rate": 4.1571054675542645e-06,
+      "loss": 0.0749,
+      "step": 3243
+    },
+    {
+      "epoch": 3.6434086761196127,
+      "grad_norm": 0.4439307907074262,
+      "learning_rate": 4.150741361010837e-06,
+      "loss": 0.0795,
+      "step": 3244
+    },
+    {
+      "epoch": 3.6445317983995507,
+      "grad_norm": 0.4603100158815367,
+      "learning_rate": 4.1443808535806195e-06,
+      "loss": 0.0877,
+      "step": 3245
+    },
+    {
+      "epoch": 3.645654920679489,
+      "grad_norm": 0.42378544801862755,
+      "learning_rate": 4.138023949177303e-06,
+      "loss": 0.073,
+      "step": 3246
+    },
+    {
+      "epoch": 3.6467780429594274,
+      "grad_norm": 0.4225346697676541,
+      "learning_rate": 4.131670651712357e-06,
+      "loss": 0.0748,
+      "step": 3247
+    },
+    {
+      "epoch": 3.6479011652393654,
+      "grad_norm": 0.4518554843426758,
+      "learning_rate": 4.125320965095037e-06,
+      "loss": 0.0785,
+      "step": 3248
+    },
+    {
+      "epoch": 3.6490242875193037,
+      "grad_norm": 0.470687474826635,
+      "learning_rate": 4.1189748932323735e-06,
+      "loss": 0.0827,
+      "step": 3249
+    },
+    {
+      "epoch": 3.650147409799242,
+      "grad_norm": 0.4506569877630462,
+      "learning_rate": 4.112632440029176e-06,
+      "loss": 0.078,
+      "step": 3250
+    },
+    {
+      "epoch": 3.65127053207918,
+      "grad_norm": 0.44578298786080667,
+      "learning_rate": 4.10629360938803e-06,
+      "loss": 0.0823,
+      "step": 3251
+    },
+    {
+      "epoch": 3.6523936543591184,
+      "grad_norm": 0.44292577788828275,
+      "learning_rate": 4.099958405209281e-06,
+      "loss": 0.0796,
+      "step": 3252
+    },
+    {
+      "epoch": 3.653516776639057,
+      "grad_norm": 0.4464121497898923,
+      "learning_rate": 4.093626831391051e-06,
+      "loss": 0.0729,
+      "step": 3253
+    },
+    {
+      "epoch": 3.6546398989189948,
+      "grad_norm": 0.44330661042864056,
+      "learning_rate": 4.087298891829229e-06,
+      "loss": 0.0753,
+      "step": 3254
+    },
+    {
+      "epoch": 3.655763021198933,
+      "grad_norm": 0.45305772566991315,
+      "learning_rate": 4.080974590417464e-06,
+      "loss": 0.0766,
+      "step": 3255
+    },
+    {
+      "epoch": 3.6568861434788715,
+      "grad_norm": 0.45177947052718287,
+      "learning_rate": 4.074653931047175e-06,
+      "loss": 0.0755,
+      "step": 3256
+    },
+    {
+      "epoch": 3.6580092657588095,
+      "grad_norm": 0.45801600980210305,
+      "learning_rate": 4.068336917607521e-06,
+      "loss": 0.0765,
+      "step": 3257
+    },
+    {
+      "epoch": 3.659132388038748,
+      "grad_norm": 0.4554130421647156,
+      "learning_rate": 4.0620235539854394e-06,
+      "loss": 0.0773,
+      "step": 3258
+    },
+    {
+      "epoch": 3.660255510318686,
+      "grad_norm": 0.4463096526208407,
+      "learning_rate": 4.0557138440656084e-06,
+      "loss": 0.0831,
+      "step": 3259
+    },
+    {
+      "epoch": 3.661378632598624,
+      "grad_norm": 0.45167705223108046,
+      "learning_rate": 4.049407791730464e-06,
+      "loss": 0.0774,
+      "step": 3260
+    },
+    {
+      "epoch": 3.6625017548785626,
+      "grad_norm": 0.45315081343108904,
+      "learning_rate": 4.043105400860191e-06,
+      "loss": 0.0804,
+      "step": 3261
+    },
+    {
+      "epoch": 3.6636248771585005,
+      "grad_norm": 0.43944579518546123,
+      "learning_rate": 4.036806675332715e-06,
+      "loss": 0.0771,
+      "step": 3262
+    },
+    {
+      "epoch": 3.664747999438439,
+      "grad_norm": 0.4617022800989824,
+      "learning_rate": 4.030511619023714e-06,
+      "loss": 0.0838,
+      "step": 3263
+    },
+    {
+      "epoch": 3.665871121718377,
+      "grad_norm": 0.8198470447722894,
+      "learning_rate": 4.0242202358066026e-06,
+      "loss": 0.089,
+      "step": 3264
+    },
+    {
+      "epoch": 3.666994243998315,
+      "grad_norm": 0.4638960398681334,
+      "learning_rate": 4.017932529552543e-06,
+      "loss": 0.0725,
+      "step": 3265
+    },
+    {
+      "epoch": 3.6681173662782536,
+      "grad_norm": 0.4669132218710685,
+      "learning_rate": 4.011648504130427e-06,
+      "loss": 0.0883,
+      "step": 3266
+    },
+    {
+      "epoch": 3.6692404885581915,
+      "grad_norm": 0.4316332260721309,
+      "learning_rate": 4.0053681634068854e-06,
+      "loss": 0.0817,
+      "step": 3267
+    },
+    {
+      "epoch": 3.67036361083813,
+      "grad_norm": 0.4386821038556611,
+      "learning_rate": 3.9990915112462805e-06,
+      "loss": 0.0784,
+      "step": 3268
+    },
+    {
+      "epoch": 3.6714867331180683,
+      "grad_norm": 0.43782337955114087,
+      "learning_rate": 3.992818551510705e-06,
+      "loss": 0.0799,
+      "step": 3269
+    },
+    {
+      "epoch": 3.672609855398006,
+      "grad_norm": 0.46001973641130106,
+      "learning_rate": 3.986549288059985e-06,
+      "loss": 0.0814,
+      "step": 3270
+    },
+    {
+      "epoch": 3.6737329776779446,
+      "grad_norm": 0.45245647722083326,
+      "learning_rate": 3.98028372475166e-06,
+      "loss": 0.0813,
+      "step": 3271
+    },
+    {
+      "epoch": 3.674856099957883,
+      "grad_norm": 0.43678629728053836,
+      "learning_rate": 3.974021865441003e-06,
+      "loss": 0.0799,
+      "step": 3272
+    },
+    {
+      "epoch": 3.675979222237821,
+      "grad_norm": 0.4176324355614767,
+      "learning_rate": 3.967763713981005e-06,
+      "loss": 0.0796,
+      "step": 3273
+    },
+    {
+      "epoch": 3.6771023445177593,
+      "grad_norm": 0.43709690178412436,
+      "learning_rate": 3.961509274222376e-06,
+      "loss": 0.0835,
+      "step": 3274
+    },
+    {
+      "epoch": 3.6782254667976977,
+      "grad_norm": 0.47658335386852513,
+      "learning_rate": 3.955258550013544e-06,
+      "loss": 0.0859,
+      "step": 3275
+    },
+    {
+      "epoch": 3.6793485890776356,
+      "grad_norm": 0.441801085296526,
+      "learning_rate": 3.949011545200643e-06,
+      "loss": 0.0792,
+      "step": 3276
+    },
+    {
+      "epoch": 3.680471711357574,
+      "grad_norm": 0.42725668162684416,
+      "learning_rate": 3.9427682636275256e-06,
+      "loss": 0.0757,
+      "step": 3277
+    },
+    {
+      "epoch": 3.6815948336375124,
+      "grad_norm": 0.4409037890558187,
+      "learning_rate": 3.936528709135752e-06,
+      "loss": 0.0764,
+      "step": 3278
+    },
+    {
+      "epoch": 3.6827179559174503,
+      "grad_norm": 0.4215867066752958,
+      "learning_rate": 3.93029288556459e-06,
+      "loss": 0.0821,
+      "step": 3279
+    },
+    {
+      "epoch": 3.6838410781973887,
+      "grad_norm": 0.4602625173462566,
+      "learning_rate": 3.924060796751012e-06,
+      "loss": 0.08,
+      "step": 3280
+    },
+    {
+      "epoch": 3.684964200477327,
+      "grad_norm": 0.4689840196628956,
+      "learning_rate": 3.9178324465296854e-06,
+      "loss": 0.079,
+      "step": 3281
+    },
+    {
+      "epoch": 3.686087322757265,
+      "grad_norm": 0.45249515865265794,
+      "learning_rate": 3.911607838732986e-06,
+      "loss": 0.0861,
+      "step": 3282
+    },
+    {
+      "epoch": 3.6872104450372034,
+      "grad_norm": 0.4505909397106748,
+      "learning_rate": 3.9053869771909835e-06,
+      "loss": 0.0816,
+      "step": 3283
+    },
+    {
+      "epoch": 3.688333567317142,
+      "grad_norm": 0.4357114391116514,
+      "learning_rate": 3.899169865731441e-06,
+      "loss": 0.0793,
+      "step": 3284
+    },
+    {
+      "epoch": 3.6894566895970797,
+      "grad_norm": 0.45612069756713314,
+      "learning_rate": 3.892956508179822e-06,
+      "loss": 0.0817,
+      "step": 3285
+    },
+    {
+      "epoch": 3.690579811877018,
+      "grad_norm": 0.4559467390418769,
+      "learning_rate": 3.886746908359264e-06,
+      "loss": 0.0832,
+      "step": 3286
+    },
+    {
+      "epoch": 3.6917029341569565,
+      "grad_norm": 0.43075052200607683,
+      "learning_rate": 3.880541070090607e-06,
+      "loss": 0.0753,
+      "step": 3287
+    },
+    {
+      "epoch": 3.6928260564368944,
+      "grad_norm": 0.4635796785903508,
+      "learning_rate": 3.874338997192371e-06,
+      "loss": 0.084,
+      "step": 3288
+    },
+    {
+      "epoch": 3.693949178716833,
+      "grad_norm": 0.4313194092470317,
+      "learning_rate": 3.8681406934807585e-06,
+      "loss": 0.0743,
+      "step": 3289
+    },
+    {
+      "epoch": 3.695072300996771,
+      "grad_norm": 0.42259202051656924,
+      "learning_rate": 3.8619461627696605e-06,
+      "loss": 0.0755,
+      "step": 3290
+    },
+    {
+      "epoch": 3.696195423276709,
+      "grad_norm": 0.4529194564392549,
+      "learning_rate": 3.855755408870631e-06,
+      "loss": 0.0792,
+      "step": 3291
+    },
+    {
+      "epoch": 3.6973185455566475,
+      "grad_norm": 0.46556671089771,
+      "learning_rate": 3.8495684355929115e-06,
+      "loss": 0.0853,
+      "step": 3292
+    },
+    {
+      "epoch": 3.698441667836586,
+      "grad_norm": 0.4448844328456529,
+      "learning_rate": 3.8433852467434175e-06,
+      "loss": 0.0786,
+      "step": 3293
+    },
+    {
+      "epoch": 3.699564790116524,
+      "grad_norm": 0.4432713263856399,
+      "learning_rate": 3.837205846126731e-06,
+      "loss": 0.0782,
+      "step": 3294
+    },
+    {
+      "epoch": 3.700687912396462,
+      "grad_norm": 0.4211879529566886,
+      "learning_rate": 3.83103023754511e-06,
+      "loss": 0.0798,
+      "step": 3295
+    },
+    {
+      "epoch": 3.7018110346764006,
+      "grad_norm": 0.4743647041115318,
+      "learning_rate": 3.824858424798467e-06,
+      "loss": 0.0883,
+      "step": 3296
+    },
+    {
+      "epoch": 3.7029341569563385,
+      "grad_norm": 0.4248995644146303,
+      "learning_rate": 3.8186904116843895e-06,
+      "loss": 0.0728,
+      "step": 3297
+    },
+    {
+      "epoch": 3.704057279236277,
+      "grad_norm": 0.4378423244099125,
+      "learning_rate": 3.8125262019981224e-06,
+      "loss": 0.0767,
+      "step": 3298
+    },
+    {
+      "epoch": 3.7051804015162153,
+      "grad_norm": 0.4518707599907417,
+      "learning_rate": 3.8063657995325743e-06,
+      "loss": 0.0856,
+      "step": 3299
+    },
+    {
+      "epoch": 3.7063035237961532,
+      "grad_norm": 0.45215273429838926,
+      "learning_rate": 3.8002092080783116e-06,
+      "loss": 0.0765,
+      "step": 3300
+    },
+    {
+      "epoch": 3.7074266460760916,
+      "grad_norm": 0.4447025106277428,
+      "learning_rate": 3.794056431423545e-06,
+      "loss": 0.0786,
+      "step": 3301
+    },
+    {
+      "epoch": 3.70854976835603,
+      "grad_norm": 0.4459028238128283,
+      "learning_rate": 3.787907473354149e-06,
+      "loss": 0.082,
+      "step": 3302
+    },
+    {
+      "epoch": 3.709672890635968,
+      "grad_norm": 0.4404022767102967,
+      "learning_rate": 3.781762337653646e-06,
+      "loss": 0.0785,
+      "step": 3303
+    },
+    {
+      "epoch": 3.7107960129159063,
+      "grad_norm": 0.4450961184830428,
+      "learning_rate": 3.7756210281032092e-06,
+      "loss": 0.0828,
+      "step": 3304
+    },
+    {
+      "epoch": 3.7119191351958447,
+      "grad_norm": 0.423474128079963,
+      "learning_rate": 3.769483548481646e-06,
+      "loss": 0.0727,
+      "step": 3305
+    },
+    {
+      "epoch": 3.7130422574757826,
+      "grad_norm": 0.4556402049278725,
+      "learning_rate": 3.7633499025654186e-06,
+      "loss": 0.0801,
+      "step": 3306
+    },
+    {
+      "epoch": 3.714165379755721,
+      "grad_norm": 0.45401931123707706,
+      "learning_rate": 3.7572200941286284e-06,
+      "loss": 0.0828,
+      "step": 3307
+    },
+    {
+      "epoch": 3.715288502035659,
+      "grad_norm": 0.4549636700971652,
+      "learning_rate": 3.7510941269430124e-06,
+      "loss": 0.0837,
+      "step": 3308
+    },
+    {
+      "epoch": 3.7164116243155974,
+      "grad_norm": 0.45729117494453114,
+      "learning_rate": 3.744972004777947e-06,
+      "loss": 0.0787,
+      "step": 3309
+    },
+    {
+      "epoch": 3.7175347465955357,
+      "grad_norm": 0.45161195568530854,
+      "learning_rate": 3.7388537314004394e-06,
+      "loss": 0.0768,
+      "step": 3310
+    },
+    {
+      "epoch": 3.7186578688754737,
+      "grad_norm": 0.4400234197426347,
+      "learning_rate": 3.732739310575132e-06,
+      "loss": 0.0745,
+      "step": 3311
+    },
+    {
+      "epoch": 3.719780991155412,
+      "grad_norm": 0.4343707184077051,
+      "learning_rate": 3.7266287460642956e-06,
+      "loss": 0.0817,
+      "step": 3312
+    },
+    {
+      "epoch": 3.7209041134353504,
+      "grad_norm": 0.4394923766740628,
+      "learning_rate": 3.7205220416278263e-06,
+      "loss": 0.0747,
+      "step": 3313
+    },
+    {
+      "epoch": 3.7220272357152884,
+      "grad_norm": 0.43402411147419445,
+      "learning_rate": 3.7144192010232515e-06,
+      "loss": 0.0816,
+      "step": 3314
+    },
+    {
+      "epoch": 3.7231503579952268,
+      "grad_norm": 0.4560774435649313,
+      "learning_rate": 3.7083202280057084e-06,
+      "loss": 0.0788,
+      "step": 3315
+    },
+    {
+      "epoch": 3.7242734802751647,
+      "grad_norm": 0.4414913360932274,
+      "learning_rate": 3.702225126327965e-06,
+      "loss": 0.0742,
+      "step": 3316
+    },
+    {
+      "epoch": 3.725396602555103,
+      "grad_norm": 0.43653829484825973,
+      "learning_rate": 3.6961338997404062e-06,
+      "loss": 0.0766,
+      "step": 3317
+    },
+    {
+      "epoch": 3.7265197248350415,
+      "grad_norm": 0.4597215590802913,
+      "learning_rate": 3.6900465519910288e-06,
+      "loss": 0.086,
+      "step": 3318
+    },
+    {
+      "epoch": 3.7276428471149794,
+      "grad_norm": 0.45688319712825454,
+      "learning_rate": 3.6839630868254505e-06,
+      "loss": 0.086,
+      "step": 3319
+    },
+    {
+      "epoch": 3.728765969394918,
+      "grad_norm": 0.4613315368292363,
+      "learning_rate": 3.6778835079868857e-06,
+      "loss": 0.0794,
+      "step": 3320
+    },
+    {
+      "epoch": 3.729889091674856,
+      "grad_norm": 0.45526897285581736,
+      "learning_rate": 3.67180781921617e-06,
+      "loss": 0.0816,
+      "step": 3321
+    },
+    {
+      "epoch": 3.731012213954794,
+      "grad_norm": 0.4467162169946555,
+      "learning_rate": 3.6657360242517413e-06,
+      "loss": 0.0795,
+      "step": 3322
+    },
+    {
+      "epoch": 3.7321353362347325,
+      "grad_norm": 0.46971814383232896,
+      "learning_rate": 3.6596681268296443e-06,
+      "loss": 0.0985,
+      "step": 3323
+    },
+    {
+      "epoch": 3.733258458514671,
+      "grad_norm": 0.4354505523936359,
+      "learning_rate": 3.6536041306835226e-06,
+      "loss": 0.0784,
+      "step": 3324
+    },
+    {
+      "epoch": 3.734381580794609,
+      "grad_norm": 0.42196736134846446,
+      "learning_rate": 3.647544039544615e-06,
+      "loss": 0.077,
+      "step": 3325
+    },
+    {
+      "epoch": 3.735504703074547,
+      "grad_norm": 0.4436527706104773,
+      "learning_rate": 3.6414878571417667e-06,
+      "loss": 0.0851,
+      "step": 3326
+    },
+    {
+      "epoch": 3.7366278253544856,
+      "grad_norm": 0.429806201884575,
+      "learning_rate": 3.6354355872014113e-06,
+      "loss": 0.0726,
+      "step": 3327
+    },
+    {
+      "epoch": 3.7377509476344235,
+      "grad_norm": 0.4336025736673827,
+      "learning_rate": 3.629387233447578e-06,
+      "loss": 0.0821,
+      "step": 3328
+    },
+    {
+      "epoch": 3.738874069914362,
+      "grad_norm": 0.4602608383711589,
+      "learning_rate": 3.623342799601889e-06,
+      "loss": 0.0804,
+      "step": 3329
+    },
+    {
+      "epoch": 3.7399971921943003,
+      "grad_norm": 0.4114914340076706,
+      "learning_rate": 3.617302289383543e-06,
+      "loss": 0.074,
+      "step": 3330
+    },
+    {
+      "epoch": 3.741120314474238,
+      "grad_norm": 0.4429172740958033,
+      "learning_rate": 3.6112657065093382e-06,
+      "loss": 0.0788,
+      "step": 3331
+    },
+    {
+      "epoch": 3.7422434367541766,
+      "grad_norm": 0.4504323974328796,
+      "learning_rate": 3.6052330546936476e-06,
+      "loss": 0.0791,
+      "step": 3332
+    },
+    {
+      "epoch": 3.743366559034115,
+      "grad_norm": 0.4112295106510093,
+      "learning_rate": 3.599204337648431e-06,
+      "loss": 0.0743,
+      "step": 3333
+    },
+    {
+      "epoch": 3.744489681314053,
+      "grad_norm": 0.44159456652322426,
+      "learning_rate": 3.593179559083225e-06,
+      "loss": 0.0735,
+      "step": 3334
+    },
+    {
+      "epoch": 3.7456128035939913,
+      "grad_norm": 0.4148152260642522,
+      "learning_rate": 3.5871587227051385e-06,
+      "loss": 0.0749,
+      "step": 3335
+    },
+    {
+      "epoch": 3.7467359258739297,
+      "grad_norm": 0.47005121017013696,
+      "learning_rate": 3.5811418322188585e-06,
+      "loss": 0.0823,
+      "step": 3336
+    },
+    {
+      "epoch": 3.7478590481538676,
+      "grad_norm": 0.4606593735634737,
+      "learning_rate": 3.575128891326647e-06,
+      "loss": 0.0849,
+      "step": 3337
+    },
+    {
+      "epoch": 3.748982170433806,
+      "grad_norm": 0.5019905326101969,
+      "learning_rate": 3.56911990372833e-06,
+      "loss": 0.0792,
+      "step": 3338
+    },
+    {
+      "epoch": 3.7501052927137444,
+      "grad_norm": 0.4497014234992369,
+      "learning_rate": 3.563114873121308e-06,
+      "loss": 0.0768,
+      "step": 3339
+    },
+    {
+      "epoch": 3.7512284149936823,
+      "grad_norm": 0.4280427061355485,
+      "learning_rate": 3.557113803200537e-06,
+      "loss": 0.0757,
+      "step": 3340
+    },
+    {
+      "epoch": 3.7523515372736207,
+      "grad_norm": 0.4591535352490078,
+      "learning_rate": 3.5511166976585432e-06,
+      "loss": 0.0836,
+      "step": 3341
+    },
+    {
+      "epoch": 3.753474659553559,
+      "grad_norm": 0.4659249443970378,
+      "learning_rate": 3.5451235601854118e-06,
+      "loss": 0.0847,
+      "step": 3342
+    },
+    {
+      "epoch": 3.754597781833497,
+      "grad_norm": 0.4465299372843637,
+      "learning_rate": 3.5391343944687906e-06,
+      "loss": 0.075,
+      "step": 3343
+    },
+    {
+      "epoch": 3.7557209041134354,
+      "grad_norm": 0.46314659971337047,
+      "learning_rate": 3.533149204193871e-06,
+      "loss": 0.0813,
+      "step": 3344
+    },
+    {
+      "epoch": 3.756844026393374,
+      "grad_norm": 0.454679241847546,
+      "learning_rate": 3.527167993043411e-06,
+      "loss": 0.0779,
+      "step": 3345
+    },
+    {
+      "epoch": 3.7579671486733117,
+      "grad_norm": 0.432912317003138,
+      "learning_rate": 3.5211907646977152e-06,
+      "loss": 0.0772,
+      "step": 3346
+    },
+    {
+      "epoch": 3.75909027095325,
+      "grad_norm": 0.43149778700891545,
+      "learning_rate": 3.5152175228346375e-06,
+      "loss": 0.0754,
+      "step": 3347
+    },
+    {
+      "epoch": 3.7602133932331885,
+      "grad_norm": 0.4296614839842937,
+      "learning_rate": 3.5092482711295805e-06,
+      "loss": 0.0777,
+      "step": 3348
+    },
+    {
+      "epoch": 3.7613365155131264,
+      "grad_norm": 0.4210158808787179,
+      "learning_rate": 3.5032830132554928e-06,
+      "loss": 0.0722,
+      "step": 3349
+    },
+    {
+      "epoch": 3.762459637793065,
+      "grad_norm": 0.4390732400163971,
+      "learning_rate": 3.497321752882856e-06,
+      "loss": 0.0716,
+      "step": 3350
+    },
+    {
+      "epoch": 3.763582760073003,
+      "grad_norm": 0.44295131957694994,
+      "learning_rate": 3.4913644936797054e-06,
+      "loss": 0.0751,
+      "step": 3351
+    },
+    {
+      "epoch": 3.764705882352941,
+      "grad_norm": 0.4534307190117399,
+      "learning_rate": 3.4854112393116047e-06,
+      "loss": 0.0782,
+      "step": 3352
+    },
+    {
+      "epoch": 3.7658290046328795,
+      "grad_norm": 0.4385591150069758,
+      "learning_rate": 3.4794619934416586e-06,
+      "loss": 0.0738,
+      "step": 3353
+    },
+    {
+      "epoch": 3.766952126912818,
+      "grad_norm": 0.43649347579395936,
+      "learning_rate": 3.473516759730503e-06,
+      "loss": 0.0768,
+      "step": 3354
+    },
+    {
+      "epoch": 3.768075249192756,
+      "grad_norm": 0.437118874941968,
+      "learning_rate": 3.4675755418363054e-06,
+      "loss": 0.0769,
+      "step": 3355
+    },
+    {
+      "epoch": 3.769198371472694,
+      "grad_norm": 0.4299095911687841,
+      "learning_rate": 3.461638343414764e-06,
+      "loss": 0.0786,
+      "step": 3356
+    },
+    {
+      "epoch": 3.7703214937526326,
+      "grad_norm": 0.4697008287502685,
+      "learning_rate": 3.455705168119101e-06,
+      "loss": 0.0885,
+      "step": 3357
+    },
+    {
+      "epoch": 3.7714446160325705,
+      "grad_norm": 0.46789742839025383,
+      "learning_rate": 3.4497760196000686e-06,
+      "loss": 0.0846,
+      "step": 3358
+    },
+    {
+      "epoch": 3.772567738312509,
+      "grad_norm": 0.42160537923493707,
+      "learning_rate": 3.4438509015059284e-06,
+      "loss": 0.0742,
+      "step": 3359
+    },
+    {
+      "epoch": 3.773690860592447,
+      "grad_norm": 0.42954624698813265,
+      "learning_rate": 3.437929817482477e-06,
+      "loss": 0.078,
+      "step": 3360
+    },
+    {
+      "epoch": 3.7748139828723852,
+      "grad_norm": 0.436902871143199,
+      "learning_rate": 3.432012771173021e-06,
+      "loss": 0.0786,
+      "step": 3361
+    },
+    {
+      "epoch": 3.7759371051523236,
+      "grad_norm": 0.4617151544078479,
+      "learning_rate": 3.4260997662183836e-06,
+      "loss": 0.0817,
+      "step": 3362
+    },
+    {
+      "epoch": 3.7770602274322616,
+      "grad_norm": 0.4527611616144744,
+      "learning_rate": 3.4201908062569066e-06,
+      "loss": 0.0751,
+      "step": 3363
+    },
+    {
+      "epoch": 3.7781833497122,
+      "grad_norm": 0.4238491787758921,
+      "learning_rate": 3.4142858949244305e-06,
+      "loss": 0.0733,
+      "step": 3364
+    },
+    {
+      "epoch": 3.779306471992138,
+      "grad_norm": 0.4466064502716907,
+      "learning_rate": 3.408385035854317e-06,
+      "loss": 0.0743,
+      "step": 3365
+    },
+    {
+      "epoch": 3.7804295942720763,
+      "grad_norm": 0.4426815137139256,
+      "learning_rate": 3.4024882326774266e-06,
+      "loss": 0.0714,
+      "step": 3366
+    },
+    {
+      "epoch": 3.7815527165520146,
+      "grad_norm": 0.4699960233638405,
+      "learning_rate": 3.39659548902213e-06,
+      "loss": 0.0803,
+      "step": 3367
+    },
+    {
+      "epoch": 3.7826758388319526,
+      "grad_norm": 0.4482404399293802,
+      "learning_rate": 3.390706808514299e-06,
+      "loss": 0.0819,
+      "step": 3368
+    },
+    {
+      "epoch": 3.783798961111891,
+      "grad_norm": 0.45212437238597697,
+      "learning_rate": 3.3848221947772976e-06,
+      "loss": 0.0761,
+      "step": 3369
+    },
+    {
+      "epoch": 3.7849220833918293,
+      "grad_norm": 0.4533539920505649,
+      "learning_rate": 3.378941651431996e-06,
+      "loss": 0.0814,
+      "step": 3370
+    },
+    {
+      "epoch": 3.7860452056717673,
+      "grad_norm": 0.448203569611718,
+      "learning_rate": 3.3730651820967588e-06,
+      "loss": 0.076,
+      "step": 3371
+    },
+    {
+      "epoch": 3.7871683279517057,
+      "grad_norm": 0.42790870357703314,
+      "learning_rate": 3.36719279038744e-06,
+      "loss": 0.0789,
+      "step": 3372
+    },
+    {
+      "epoch": 3.788291450231644,
+      "grad_norm": 0.42797461979429224,
+      "learning_rate": 3.361324479917393e-06,
+      "loss": 0.0734,
+      "step": 3373
+    },
+    {
+      "epoch": 3.789414572511582,
+      "grad_norm": 0.44621043727284554,
+      "learning_rate": 3.3554602542974436e-06,
+      "loss": 0.086,
+      "step": 3374
+    },
+    {
+      "epoch": 3.7905376947915204,
+      "grad_norm": 0.42240375254920737,
+      "learning_rate": 3.3496001171359204e-06,
+      "loss": 0.0768,
+      "step": 3375
+    },
+    {
+      "epoch": 3.7916608170714587,
+      "grad_norm": 0.4386956304607235,
+      "learning_rate": 3.3437440720386294e-06,
+      "loss": 0.076,
+      "step": 3376
+    },
+    {
+      "epoch": 3.7927839393513967,
+      "grad_norm": 0.45986967497222725,
+      "learning_rate": 3.3378921226088632e-06,
+      "loss": 0.0827,
+      "step": 3377
+    },
+    {
+      "epoch": 3.793907061631335,
+      "grad_norm": 0.44325952459332413,
+      "learning_rate": 3.3320442724473843e-06,
+      "loss": 0.0842,
+      "step": 3378
+    },
+    {
+      "epoch": 3.7950301839112734,
+      "grad_norm": 0.43529759847083,
+      "learning_rate": 3.326200525152441e-06,
+      "loss": 0.0812,
+      "step": 3379
+    },
+    {
+      "epoch": 3.7961533061912114,
+      "grad_norm": 0.4637261053857433,
+      "learning_rate": 3.3203608843197575e-06,
+      "loss": 0.0801,
+      "step": 3380
+    },
+    {
+      "epoch": 3.7972764284711498,
+      "grad_norm": 0.44756524271819015,
+      "learning_rate": 3.314525353542527e-06,
+      "loss": 0.0811,
+      "step": 3381
+    },
+    {
+      "epoch": 3.798399550751088,
+      "grad_norm": 0.4205094044217173,
+      "learning_rate": 3.308693936411421e-06,
+      "loss": 0.0826,
+      "step": 3382
+    },
+    {
+      "epoch": 3.799522673031026,
+      "grad_norm": 0.43291470049655334,
+      "learning_rate": 3.302866636514567e-06,
+      "loss": 0.0833,
+      "step": 3383
+    },
+    {
+      "epoch": 3.8006457953109645,
+      "grad_norm": 0.42670722217149304,
+      "learning_rate": 3.29704345743757e-06,
+      "loss": 0.073,
+      "step": 3384
+    },
+    {
+      "epoch": 3.801768917590903,
+      "grad_norm": 0.4343627394675364,
+      "learning_rate": 3.2912244027634953e-06,
+      "loss": 0.0799,
+      "step": 3385
+    },
+    {
+      "epoch": 3.802892039870841,
+      "grad_norm": 0.4488560245180038,
+      "learning_rate": 3.285409476072874e-06,
+      "loss": 0.0779,
+      "step": 3386
+    },
+    {
+      "epoch": 3.804015162150779,
+      "grad_norm": 0.4503612616723187,
+      "learning_rate": 3.2795986809436953e-06,
+      "loss": 0.0808,
+      "step": 3387
+    },
+    {
+      "epoch": 3.8051382844307176,
+      "grad_norm": 0.4407667707573012,
+      "learning_rate": 3.2737920209513994e-06,
+      "loss": 0.0746,
+      "step": 3388
+    },
+    {
+      "epoch": 3.8062614067106555,
+      "grad_norm": 0.4491010129063716,
+      "learning_rate": 3.267989499668892e-06,
+      "loss": 0.0722,
+      "step": 3389
+    },
+    {
+      "epoch": 3.807384528990594,
+      "grad_norm": 0.4329898689532073,
+      "learning_rate": 3.262191120666528e-06,
+      "loss": 0.0774,
+      "step": 3390
+    },
+    {
+      "epoch": 3.8085076512705323,
+      "grad_norm": 0.4383306886957197,
+      "learning_rate": 3.2563968875121133e-06,
+      "loss": 0.0766,
+      "step": 3391
+    },
+    {
+      "epoch": 3.80963077355047,
+      "grad_norm": 0.4401170472954754,
+      "learning_rate": 3.250606803770904e-06,
+      "loss": 0.0791,
+      "step": 3392
+    },
+    {
+      "epoch": 3.8107538958304086,
+      "grad_norm": 0.4532511703655005,
+      "learning_rate": 3.2448208730056053e-06,
+      "loss": 0.0799,
+      "step": 3393
+    },
+    {
+      "epoch": 3.811877018110347,
+      "grad_norm": 0.42610017647616477,
+      "learning_rate": 3.2390390987763578e-06,
+      "loss": 0.0785,
+      "step": 3394
+    },
+    {
+      "epoch": 3.813000140390285,
+      "grad_norm": 0.4230591818086143,
+      "learning_rate": 3.233261484640753e-06,
+      "loss": 0.0667,
+      "step": 3395
+    },
+    {
+      "epoch": 3.8141232626702233,
+      "grad_norm": 0.4625893901658147,
+      "learning_rate": 3.2274880341538216e-06,
+      "loss": 0.0769,
+      "step": 3396
+    },
+    {
+      "epoch": 3.8152463849501617,
+      "grad_norm": 0.4400378121499265,
+      "learning_rate": 3.2217187508680314e-06,
+      "loss": 0.0779,
+      "step": 3397
+    },
+    {
+      "epoch": 3.8163695072300996,
+      "grad_norm": 0.4490227215560117,
+      "learning_rate": 3.2159536383332846e-06,
+      "loss": 0.0779,
+      "step": 3398
+    },
+    {
+      "epoch": 3.817492629510038,
+      "grad_norm": 0.41925773594382665,
+      "learning_rate": 3.21019270009692e-06,
+      "loss": 0.0741,
+      "step": 3399
+    },
+    {
+      "epoch": 3.8186157517899764,
+      "grad_norm": 0.4415558039457624,
+      "learning_rate": 3.204435939703705e-06,
+      "loss": 0.0759,
+      "step": 3400
+    },
+    {
+      "epoch": 3.8197388740699143,
+      "grad_norm": 0.4451411473808772,
+      "learning_rate": 3.198683360695839e-06,
+      "loss": 0.0753,
+      "step": 3401
+    },
+    {
+      "epoch": 3.8208619963498527,
+      "grad_norm": 0.42552871943963727,
+      "learning_rate": 3.192934966612948e-06,
+      "loss": 0.075,
+      "step": 3402
+    },
+    {
+      "epoch": 3.821985118629791,
+      "grad_norm": 0.4700312235719863,
+      "learning_rate": 3.187190760992078e-06,
+      "loss": 0.0828,
+      "step": 3403
+    },
+    {
+      "epoch": 3.823108240909729,
+      "grad_norm": 0.45339677778386067,
+      "learning_rate": 3.1814507473677047e-06,
+      "loss": 0.0789,
+      "step": 3404
+    },
+    {
+      "epoch": 3.8242313631896674,
+      "grad_norm": 0.4331334874484187,
+      "learning_rate": 3.1757149292717216e-06,
+      "loss": 0.0754,
+      "step": 3405
+    },
+    {
+      "epoch": 3.8253544854696058,
+      "grad_norm": 0.44163707435590865,
+      "learning_rate": 3.1699833102334397e-06,
+      "loss": 0.079,
+      "step": 3406
+    },
+    {
+      "epoch": 3.8264776077495437,
+      "grad_norm": 0.42901426073349475,
+      "learning_rate": 3.164255893779591e-06,
+      "loss": 0.0748,
+      "step": 3407
+    },
+    {
+      "epoch": 3.827600730029482,
+      "grad_norm": 0.4246549205961034,
+      "learning_rate": 3.1585326834343123e-06,
+      "loss": 0.0758,
+      "step": 3408
+    },
+    {
+      "epoch": 3.82872385230942,
+      "grad_norm": 0.43824663282905246,
+      "learning_rate": 3.152813682719159e-06,
+      "loss": 0.0817,
+      "step": 3409
+    },
+    {
+      "epoch": 3.8298469745893584,
+      "grad_norm": 0.4487556162344414,
+      "learning_rate": 3.147098895153098e-06,
+      "loss": 0.0808,
+      "step": 3410
+    },
+    {
+      "epoch": 3.830970096869297,
+      "grad_norm": 0.41728849401494356,
+      "learning_rate": 3.141388324252499e-06,
+      "loss": 0.0734,
+      "step": 3411
+    },
+    {
+      "epoch": 3.8320932191492347,
+      "grad_norm": 0.42655605860029716,
+      "learning_rate": 3.135681973531144e-06,
+      "loss": 0.081,
+      "step": 3412
+    },
+    {
+      "epoch": 3.833216341429173,
+      "grad_norm": 0.4241664691861822,
+      "learning_rate": 3.129979846500205e-06,
+      "loss": 0.0779,
+      "step": 3413
+    },
+    {
+      "epoch": 3.8343394637091115,
+      "grad_norm": 0.4153186762640536,
+      "learning_rate": 3.1242819466682673e-06,
+      "loss": 0.0771,
+      "step": 3414
+    },
+    {
+      "epoch": 3.8354625859890494,
+      "grad_norm": 0.4267033688189136,
+      "learning_rate": 3.1185882775413123e-06,
+      "loss": 0.073,
+      "step": 3415
+    },
+    {
+      "epoch": 3.836585708268988,
+      "grad_norm": 0.4457824337627224,
+      "learning_rate": 3.1128988426227193e-06,
+      "loss": 0.0805,
+      "step": 3416
+    },
+    {
+      "epoch": 3.8377088305489258,
+      "grad_norm": 0.4281819990435502,
+      "learning_rate": 3.107213645413254e-06,
+      "loss": 0.0783,
+      "step": 3417
+    },
+    {
+      "epoch": 3.838831952828864,
+      "grad_norm": 0.42960944794714556,
+      "learning_rate": 3.101532689411085e-06,
+      "loss": 0.0758,
+      "step": 3418
+    },
+    {
+      "epoch": 3.8399550751088025,
+      "grad_norm": 0.4270593518801406,
+      "learning_rate": 3.095855978111767e-06,
+      "loss": 0.0763,
+      "step": 3419
+    },
+    {
+      "epoch": 3.8410781973887405,
+      "grad_norm": 0.445387965169147,
+      "learning_rate": 3.0901835150082417e-06,
+      "loss": 0.0755,
+      "step": 3420
+    },
+    {
+      "epoch": 3.842201319668679,
+      "grad_norm": 0.44342244606556647,
+      "learning_rate": 3.0845153035908415e-06,
+      "loss": 0.0711,
+      "step": 3421
+    },
+    {
+      "epoch": 3.843324441948617,
+      "grad_norm": 0.4222486589242842,
+      "learning_rate": 3.0788513473472747e-06,
+      "loss": 0.0778,
+      "step": 3422
+    },
+    {
+      "epoch": 3.844447564228555,
+      "grad_norm": 0.43608936121729586,
+      "learning_rate": 3.0731916497626356e-06,
+      "loss": 0.0798,
+      "step": 3423
+    },
+    {
+      "epoch": 3.8455706865084935,
+      "grad_norm": 0.4601934819466419,
+      "learning_rate": 3.067536214319402e-06,
+      "loss": 0.0834,
+      "step": 3424
+    },
+    {
+      "epoch": 3.846693808788432,
+      "grad_norm": 0.41891016230414746,
+      "learning_rate": 3.0618850444974237e-06,
+      "loss": 0.0703,
+      "step": 3425
+    },
+    {
+      "epoch": 3.84781693106837,
+      "grad_norm": 0.4572426188664163,
+      "learning_rate": 3.056238143773932e-06,
+      "loss": 0.0792,
+      "step": 3426
+    },
+    {
+      "epoch": 3.8489400533483082,
+      "grad_norm": 0.4518709283094167,
+      "learning_rate": 3.050595515623519e-06,
+      "loss": 0.0793,
+      "step": 3427
+    },
+    {
+      "epoch": 3.8500631756282466,
+      "grad_norm": 0.8261186535001159,
+      "learning_rate": 3.0449571635181595e-06,
+      "loss": 0.0906,
+      "step": 3428
+    },
+    {
+      "epoch": 3.8511862979081846,
+      "grad_norm": 0.43746333274368837,
+      "learning_rate": 3.0393230909271953e-06,
+      "loss": 0.0806,
+      "step": 3429
+    },
+    {
+      "epoch": 3.852309420188123,
+      "grad_norm": 0.4328585063719159,
+      "learning_rate": 3.0336933013173307e-06,
+      "loss": 0.0777,
+      "step": 3430
+    },
+    {
+      "epoch": 3.8534325424680613,
+      "grad_norm": 0.43314046732664285,
+      "learning_rate": 3.028067798152643e-06,
+      "loss": 0.0809,
+      "step": 3431
+    },
+    {
+      "epoch": 3.8545556647479993,
+      "grad_norm": 0.4373077254389021,
+      "learning_rate": 3.022446584894557e-06,
+      "loss": 0.0788,
+      "step": 3432
+    },
+    {
+      "epoch": 3.8556787870279376,
+      "grad_norm": 0.43228751269770593,
+      "learning_rate": 3.0168296650018736e-06,
+      "loss": 0.0777,
+      "step": 3433
+    },
+    {
+      "epoch": 3.856801909307876,
+      "grad_norm": 0.43151574407618987,
+      "learning_rate": 3.011217041930743e-06,
+      "loss": 0.0733,
+      "step": 3434
+    },
+    {
+      "epoch": 3.857925031587814,
+      "grad_norm": 0.4424834168473701,
+      "learning_rate": 3.005608719134674e-06,
+      "loss": 0.0756,
+      "step": 3435
+    },
+    {
+      "epoch": 3.8590481538677524,
+      "grad_norm": 0.4685629925803647,
+      "learning_rate": 3.0000047000645305e-06,
+      "loss": 0.0827,
+      "step": 3436
+    },
+    {
+      "epoch": 3.8601712761476907,
+      "grad_norm": 0.44357998574627705,
+      "learning_rate": 2.9944049881685265e-06,
+      "loss": 0.0787,
+      "step": 3437
+    },
+    {
+      "epoch": 3.8612943984276287,
+      "grad_norm": 0.4253461038308456,
+      "learning_rate": 2.9888095868922297e-06,
+      "loss": 0.0726,
+      "step": 3438
+    },
+    {
+      "epoch": 3.862417520707567,
+      "grad_norm": 0.4260652279185222,
+      "learning_rate": 2.9832184996785453e-06,
+      "loss": 0.077,
+      "step": 3439
+    },
+    {
+      "epoch": 3.8635406429875054,
+      "grad_norm": 0.471295717516851,
+      "learning_rate": 2.977631729967735e-06,
+      "loss": 0.0836,
+      "step": 3440
+    },
+    {
+      "epoch": 3.8646637652674434,
+      "grad_norm": 0.43840929560138525,
+      "learning_rate": 2.9720492811973977e-06,
+      "loss": 0.076,
+      "step": 3441
+    },
+    {
+      "epoch": 3.8657868875473818,
+      "grad_norm": 0.453205839637241,
+      "learning_rate": 2.966471156802477e-06,
+      "loss": 0.075,
+      "step": 3442
+    },
+    {
+      "epoch": 3.86691000982732,
+      "grad_norm": 0.4288959454696245,
+      "learning_rate": 2.960897360215255e-06,
+      "loss": 0.0803,
+      "step": 3443
+    },
+    {
+      "epoch": 3.868033132107258,
+      "grad_norm": 0.43608110599439925,
+      "learning_rate": 2.9553278948653484e-06,
+      "loss": 0.0804,
+      "step": 3444
+    },
+    {
+      "epoch": 3.8691562543871965,
+      "grad_norm": 0.4411598218041702,
+      "learning_rate": 2.949762764179711e-06,
+      "loss": 0.0791,
+      "step": 3445
+    },
+    {
+      "epoch": 3.870279376667135,
+      "grad_norm": 0.45413949768533496,
+      "learning_rate": 2.9442019715826318e-06,
+      "loss": 0.0777,
+      "step": 3446
+    },
+    {
+      "epoch": 3.871402498947073,
+      "grad_norm": 0.4448584359461084,
+      "learning_rate": 2.938645520495723e-06,
+      "loss": 0.0769,
+      "step": 3447
+    },
+    {
+      "epoch": 3.872525621227011,
+      "grad_norm": 0.4551946661458706,
+      "learning_rate": 2.933093414337932e-06,
+      "loss": 0.0783,
+      "step": 3448
+    },
+    {
+      "epoch": 3.8736487435069495,
+      "grad_norm": 0.4458905959273584,
+      "learning_rate": 2.927545656525531e-06,
+      "loss": 0.0765,
+      "step": 3449
+    },
+    {
+      "epoch": 3.8747718657868875,
+      "grad_norm": 0.4281315602250711,
+      "learning_rate": 2.922002250472119e-06,
+      "loss": 0.0749,
+      "step": 3450
+    },
+    {
+      "epoch": 3.875894988066826,
+      "grad_norm": 0.43195709338842564,
+      "learning_rate": 2.9164631995886095e-06,
+      "loss": 0.0784,
+      "step": 3451
+    },
+    {
+      "epoch": 3.8770181103467642,
+      "grad_norm": 0.415873215959344,
+      "learning_rate": 2.9109285072832437e-06,
+      "loss": 0.0783,
+      "step": 3452
+    },
+    {
+      "epoch": 3.878141232626702,
+      "grad_norm": 0.450371175353607,
+      "learning_rate": 2.9053981769615792e-06,
+      "loss": 0.0779,
+      "step": 3453
+    },
+    {
+      "epoch": 3.8792643549066406,
+      "grad_norm": 0.47287548262861223,
+      "learning_rate": 2.899872212026489e-06,
+      "loss": 0.089,
+      "step": 3454
+    },
+    {
+      "epoch": 3.880387477186579,
+      "grad_norm": 0.44671822443338133,
+      "learning_rate": 2.894350615878163e-06,
+      "loss": 0.0771,
+      "step": 3455
+    },
+    {
+      "epoch": 3.881510599466517,
+      "grad_norm": 0.4287761263930842,
+      "learning_rate": 2.8888333919140954e-06,
+      "loss": 0.0745,
+      "step": 3456
+    },
+    {
+      "epoch": 3.8826337217464553,
+      "grad_norm": 0.42522941098200495,
+      "learning_rate": 2.883320543529096e-06,
+      "loss": 0.0733,
+      "step": 3457
+    },
+    {
+      "epoch": 3.8837568440263937,
+      "grad_norm": 0.46005795014191836,
+      "learning_rate": 2.8778120741152805e-06,
+      "loss": 0.0813,
+      "step": 3458
+    },
+    {
+      "epoch": 3.8848799663063316,
+      "grad_norm": 0.43782850129874495,
+      "learning_rate": 2.872307987062073e-06,
+      "loss": 0.0809,
+      "step": 3459
+    },
+    {
+      "epoch": 3.88600308858627,
+      "grad_norm": 0.42888285045593805,
+      "learning_rate": 2.8668082857562006e-06,
+      "loss": 0.0759,
+      "step": 3460
+    },
+    {
+      "epoch": 3.887126210866208,
+      "grad_norm": 0.4536689216993471,
+      "learning_rate": 2.8613129735816838e-06,
+      "loss": 0.0819,
+      "step": 3461
+    },
+    {
+      "epoch": 3.8882493331461463,
+      "grad_norm": 0.43211539049911346,
+      "learning_rate": 2.855822053919852e-06,
+      "loss": 0.0855,
+      "step": 3462
+    },
+    {
+      "epoch": 3.8893724554260847,
+      "grad_norm": 0.42986836248333826,
+      "learning_rate": 2.8503355301493298e-06,
+      "loss": 0.0747,
+      "step": 3463
+    },
+    {
+      "epoch": 3.8904955777060226,
+      "grad_norm": 0.441832461225257,
+      "learning_rate": 2.8448534056460332e-06,
+      "loss": 0.0767,
+      "step": 3464
+    },
+    {
+      "epoch": 3.891618699985961,
+      "grad_norm": 0.4620375105987588,
+      "learning_rate": 2.839375683783179e-06,
+      "loss": 0.0832,
+      "step": 3465
+    },
+    {
+      "epoch": 3.892741822265899,
+      "grad_norm": 0.43964940953553355,
+      "learning_rate": 2.833902367931262e-06,
+      "loss": 0.0746,
+      "step": 3466
+    },
+    {
+      "epoch": 3.8938649445458373,
+      "grad_norm": 0.464163071227373,
+      "learning_rate": 2.8284334614580767e-06,
+      "loss": 0.082,
+      "step": 3467
+    },
+    {
+      "epoch": 3.8949880668257757,
+      "grad_norm": 0.4290775892676479,
+      "learning_rate": 2.822968967728703e-06,
+      "loss": 0.0742,
+      "step": 3468
+    },
+    {
+      "epoch": 3.8961111891057136,
+      "grad_norm": 0.4208068309684712,
+      "learning_rate": 2.8175088901055026e-06,
+      "loss": 0.0751,
+      "step": 3469
+    },
+    {
+      "epoch": 3.897234311385652,
+      "grad_norm": 0.4313279478607281,
+      "learning_rate": 2.8120532319481255e-06,
+      "loss": 0.0784,
+      "step": 3470
+    },
+    {
+      "epoch": 3.8983574336655904,
+      "grad_norm": 0.4576364929353952,
+      "learning_rate": 2.8066019966134907e-06,
+      "loss": 0.0752,
+      "step": 3471
+    },
+    {
+      "epoch": 3.8994805559455283,
+      "grad_norm": 0.45124729742636177,
+      "learning_rate": 2.801155187455807e-06,
+      "loss": 0.0812,
+      "step": 3472
+    },
+    {
+      "epoch": 3.9006036782254667,
+      "grad_norm": 0.4274779842231408,
+      "learning_rate": 2.7957128078265574e-06,
+      "loss": 0.076,
+      "step": 3473
+    },
+    {
+      "epoch": 3.901726800505405,
+      "grad_norm": 0.4378140528997512,
+      "learning_rate": 2.790274861074497e-06,
+      "loss": 0.0736,
+      "step": 3474
+    },
+    {
+      "epoch": 3.902849922785343,
+      "grad_norm": 0.4343770792886928,
+      "learning_rate": 2.7848413505456564e-06,
+      "loss": 0.0743,
+      "step": 3475
+    },
+    {
+      "epoch": 3.9039730450652814,
+      "grad_norm": 0.4269837178553136,
+      "learning_rate": 2.7794122795833276e-06,
+      "loss": 0.0785,
+      "step": 3476
+    },
+    {
+      "epoch": 3.90509616734522,
+      "grad_norm": 0.44957322685953266,
+      "learning_rate": 2.7739876515280838e-06,
+      "loss": 0.0744,
+      "step": 3477
+    },
+    {
+      "epoch": 3.9062192896251577,
+      "grad_norm": 0.4336825511389323,
+      "learning_rate": 2.7685674697177568e-06,
+      "loss": 0.0721,
+      "step": 3478
+    },
+    {
+      "epoch": 3.907342411905096,
+      "grad_norm": 0.44723299851333226,
+      "learning_rate": 2.7631517374874427e-06,
+      "loss": 0.0744,
+      "step": 3479
+    },
+    {
+      "epoch": 3.9084655341850345,
+      "grad_norm": 0.44066946939094515,
+      "learning_rate": 2.7577404581695035e-06,
+      "loss": 0.0768,
+      "step": 3480
+    },
+    {
+      "epoch": 3.9095886564649724,
+      "grad_norm": 0.4500736384140416,
+      "learning_rate": 2.752333635093558e-06,
+      "loss": 0.0754,
+      "step": 3481
+    },
+    {
+      "epoch": 3.910711778744911,
+      "grad_norm": 0.4594645625800169,
+      "learning_rate": 2.746931271586484e-06,
+      "loss": 0.0803,
+      "step": 3482
+    },
+    {
+      "epoch": 3.911834901024849,
+      "grad_norm": 0.42877398862098437,
+      "learning_rate": 2.7415333709724168e-06,
+      "loss": 0.0736,
+      "step": 3483
+    },
+    {
+      "epoch": 3.912958023304787,
+      "grad_norm": 0.42734145184760597,
+      "learning_rate": 2.7361399365727404e-06,
+      "loss": 0.0791,
+      "step": 3484
+    },
+    {
+      "epoch": 3.9140811455847255,
+      "grad_norm": 0.4340445049242047,
+      "learning_rate": 2.7307509717060954e-06,
+      "loss": 0.0819,
+      "step": 3485
+    },
+    {
+      "epoch": 3.915204267864664,
+      "grad_norm": 0.4341246214018538,
+      "learning_rate": 2.725366479688373e-06,
+      "loss": 0.0756,
+      "step": 3486
+    },
+    {
+      "epoch": 3.916327390144602,
+      "grad_norm": 0.4236324146454293,
+      "learning_rate": 2.719986463832708e-06,
+      "loss": 0.0757,
+      "step": 3487
+    },
+    {
+      "epoch": 3.9174505124245402,
+      "grad_norm": 0.46626173343907984,
+      "learning_rate": 2.714610927449486e-06,
+      "loss": 0.0853,
+      "step": 3488
+    },
+    {
+      "epoch": 3.9185736347044786,
+      "grad_norm": 0.4377237420668139,
+      "learning_rate": 2.7092398738463345e-06,
+      "loss": 0.0767,
+      "step": 3489
+    },
+    {
+      "epoch": 3.9196967569844166,
+      "grad_norm": 0.44860358330503586,
+      "learning_rate": 2.7038733063281177e-06,
+      "loss": 0.073,
+      "step": 3490
+    },
+    {
+      "epoch": 3.920819879264355,
+      "grad_norm": 0.43604700182859185,
+      "learning_rate": 2.698511228196945e-06,
+      "loss": 0.0763,
+      "step": 3491
+    },
+    {
+      "epoch": 3.9219430015442933,
+      "grad_norm": 0.4401389309270918,
+      "learning_rate": 2.6931536427521632e-06,
+      "loss": 0.0846,
+      "step": 3492
+    },
+    {
+      "epoch": 3.9230661238242313,
+      "grad_norm": 0.4331666233322056,
+      "learning_rate": 2.687800553290353e-06,
+      "loss": 0.0713,
+      "step": 3493
+    },
+    {
+      "epoch": 3.9241892461041696,
+      "grad_norm": 0.44560495275667983,
+      "learning_rate": 2.6824519631053324e-06,
+      "loss": 0.0756,
+      "step": 3494
+    },
+    {
+      "epoch": 3.925312368384108,
+      "grad_norm": 0.4209276031297463,
+      "learning_rate": 2.6771078754881417e-06,
+      "loss": 0.0712,
+      "step": 3495
+    },
+    {
+      "epoch": 3.926435490664046,
+      "grad_norm": 0.4531187078058862,
+      "learning_rate": 2.6717682937270605e-06,
+      "loss": 0.0865,
+      "step": 3496
+    },
+    {
+      "epoch": 3.9275586129439843,
+      "grad_norm": 0.42791348822861297,
+      "learning_rate": 2.6664332211075915e-06,
+      "loss": 0.0734,
+      "step": 3497
+    },
+    {
+      "epoch": 3.9286817352239227,
+      "grad_norm": 0.42910458284895564,
+      "learning_rate": 2.6611026609124647e-06,
+      "loss": 0.0767,
+      "step": 3498
+    },
+    {
+      "epoch": 3.9298048575038607,
+      "grad_norm": 0.4360003667897708,
+      "learning_rate": 2.6557766164216334e-06,
+      "loss": 0.0821,
+      "step": 3499
+    },
+    {
+      "epoch": 3.930927979783799,
+      "grad_norm": 0.4473465009354422,
+      "learning_rate": 2.6504550909122674e-06,
+      "loss": 0.0824,
+      "step": 3500
+    },
+    {
+      "epoch": 3.9320511020637374,
+      "grad_norm": 0.4559995591973532,
+      "learning_rate": 2.6451380876587617e-06,
+      "loss": 0.0773,
+      "step": 3501
+    },
+    {
+      "epoch": 3.9331742243436754,
+      "grad_norm": 0.44013072714573725,
+      "learning_rate": 2.639825609932727e-06,
+      "loss": 0.0773,
+      "step": 3502
+    },
+    {
+      "epoch": 3.9342973466236137,
+      "grad_norm": 0.4385601175313549,
+      "learning_rate": 2.63451766100299e-06,
+      "loss": 0.0766,
+      "step": 3503
+    },
+    {
+      "epoch": 3.935420468903552,
+      "grad_norm": 0.46222160717340305,
+      "learning_rate": 2.6292142441355915e-06,
+      "loss": 0.0834,
+      "step": 3504
+    },
+    {
+      "epoch": 3.93654359118349,
+      "grad_norm": 0.44847101380558047,
+      "learning_rate": 2.6239153625937786e-06,
+      "loss": 0.0777,
+      "step": 3505
+    },
+    {
+      "epoch": 3.9376667134634284,
+      "grad_norm": 0.4335638035588458,
+      "learning_rate": 2.6186210196380135e-06,
+      "loss": 0.0745,
+      "step": 3506
+    },
+    {
+      "epoch": 3.938789835743367,
+      "grad_norm": 0.45155174333318204,
+      "learning_rate": 2.613331218525963e-06,
+      "loss": 0.0813,
+      "step": 3507
+    },
+    {
+      "epoch": 3.9399129580233048,
+      "grad_norm": 0.44021250781644283,
+      "learning_rate": 2.6080459625124997e-06,
+      "loss": 0.0824,
+      "step": 3508
+    },
+    {
+      "epoch": 3.941036080303243,
+      "grad_norm": 0.42626782736040425,
+      "learning_rate": 2.602765254849704e-06,
+      "loss": 0.0759,
+      "step": 3509
+    },
+    {
+      "epoch": 3.942159202583181,
+      "grad_norm": 0.4502768942497821,
+      "learning_rate": 2.597489098786847e-06,
+      "loss": 0.0808,
+      "step": 3510
+    },
+    {
+      "epoch": 3.9432823248631195,
+      "grad_norm": 0.46342459388260226,
+      "learning_rate": 2.5922174975704083e-06,
+      "loss": 0.0765,
+      "step": 3511
+    },
+    {
+      "epoch": 3.944405447143058,
+      "grad_norm": 0.43079292525854407,
+      "learning_rate": 2.5869504544440625e-06,
+      "loss": 0.0804,
+      "step": 3512
+    },
+    {
+      "epoch": 3.945528569422996,
+      "grad_norm": 0.43315675855851205,
+      "learning_rate": 2.58168797264868e-06,
+      "loss": 0.0726,
+      "step": 3513
+    },
+    {
+      "epoch": 3.946651691702934,
+      "grad_norm": 0.44735874470841014,
+      "learning_rate": 2.576430055422324e-06,
+      "loss": 0.0743,
+      "step": 3514
+    },
+    {
+      "epoch": 3.947774813982872,
+      "grad_norm": 0.4564757971460192,
+      "learning_rate": 2.5711767060002457e-06,
+      "loss": 0.0756,
+      "step": 3515
+    },
+    {
+      "epoch": 3.9488979362628105,
+      "grad_norm": 0.44223205696360607,
+      "learning_rate": 2.5659279276148896e-06,
+      "loss": 0.0718,
+      "step": 3516
+    },
+    {
+      "epoch": 3.950021058542749,
+      "grad_norm": 0.452807039724236,
+      "learning_rate": 2.5606837234958893e-06,
+      "loss": 0.0807,
+      "step": 3517
+    },
+    {
+      "epoch": 3.951144180822687,
+      "grad_norm": 0.44686811295375434,
+      "learning_rate": 2.5554440968700587e-06,
+      "loss": 0.0763,
+      "step": 3518
+    },
+    {
+      "epoch": 3.952267303102625,
+      "grad_norm": 0.43505801163376684,
+      "learning_rate": 2.550209050961403e-06,
+      "loss": 0.0706,
+      "step": 3519
+    },
+    {
+      "epoch": 3.9533904253825636,
+      "grad_norm": 0.4247883961735843,
+      "learning_rate": 2.544978588991096e-06,
+      "loss": 0.0744,
+      "step": 3520
+    },
+    {
+      "epoch": 3.9545135476625015,
+      "grad_norm": 0.45479121928721894,
+      "learning_rate": 2.5397527141775025e-06,
+      "loss": 0.0843,
+      "step": 3521
+    },
+    {
+      "epoch": 3.95563666994244,
+      "grad_norm": 0.4374273400030285,
+      "learning_rate": 2.534531429736159e-06,
+      "loss": 0.074,
+      "step": 3522
+    },
+    {
+      "epoch": 3.9567597922223783,
+      "grad_norm": 0.4149773313963219,
+      "learning_rate": 2.5293147388797813e-06,
+      "loss": 0.0708,
+      "step": 3523
+    },
+    {
+      "epoch": 3.957882914502316,
+      "grad_norm": 0.4295010740541775,
+      "learning_rate": 2.524102644818256e-06,
+      "loss": 0.0758,
+      "step": 3524
+    },
+    {
+      "epoch": 3.9590060367822546,
+      "grad_norm": 0.4486488019169518,
+      "learning_rate": 2.5188951507586422e-06,
+      "loss": 0.0808,
+      "step": 3525
+    },
+    {
+      "epoch": 3.960129159062193,
+      "grad_norm": 0.4462521068537958,
+      "learning_rate": 2.5136922599051684e-06,
+      "loss": 0.0733,
+      "step": 3526
+    },
+    {
+      "epoch": 3.961252281342131,
+      "grad_norm": 0.44838324716484473,
+      "learning_rate": 2.508493975459232e-06,
+      "loss": 0.0813,
+      "step": 3527
+    },
+    {
+      "epoch": 3.9623754036220693,
+      "grad_norm": 0.4313406939987042,
+      "learning_rate": 2.50330030061939e-06,
+      "loss": 0.0764,
+      "step": 3528
+    },
+    {
+      "epoch": 3.9634985259020077,
+      "grad_norm": 0.429785435524192,
+      "learning_rate": 2.498111238581371e-06,
+      "loss": 0.0758,
+      "step": 3529
+    },
+    {
+      "epoch": 3.9646216481819456,
+      "grad_norm": 0.4417273682496011,
+      "learning_rate": 2.492926792538061e-06,
+      "loss": 0.0812,
+      "step": 3530
+    },
+    {
+      "epoch": 3.965744770461884,
+      "grad_norm": 0.42584217521194967,
+      "learning_rate": 2.487746965679507e-06,
+      "loss": 0.0844,
+      "step": 3531
+    },
+    {
+      "epoch": 3.9668678927418224,
+      "grad_norm": 0.4512431805814872,
+      "learning_rate": 2.4825717611929144e-06,
+      "loss": 0.0844,
+      "step": 3532
+    },
+    {
+      "epoch": 3.9679910150217603,
+      "grad_norm": 0.44924315703937545,
+      "learning_rate": 2.4774011822626455e-06,
+      "loss": 0.0827,
+      "step": 3533
+    },
+    {
+      "epoch": 3.9691141373016987,
+      "grad_norm": 0.45652249852342425,
+      "learning_rate": 2.472235232070208e-06,
+      "loss": 0.085,
+      "step": 3534
+    },
+    {
+      "epoch": 3.970237259581637,
+      "grad_norm": 0.4313367900118166,
+      "learning_rate": 2.4670739137942723e-06,
+      "loss": 0.0784,
+      "step": 3535
+    },
+    {
+      "epoch": 3.971360381861575,
+      "grad_norm": 0.43828415959844774,
+      "learning_rate": 2.4619172306106533e-06,
+      "loss": 0.0759,
+      "step": 3536
+    },
+    {
+      "epoch": 3.9724835041415134,
+      "grad_norm": 0.4205420662994824,
+      "learning_rate": 2.456765185692315e-06,
+      "loss": 0.0765,
+      "step": 3537
+    },
+    {
+      "epoch": 3.973606626421452,
+      "grad_norm": 0.4438761928014099,
+      "learning_rate": 2.451617782209371e-06,
+      "loss": 0.077,
+      "step": 3538
+    },
+    {
+      "epoch": 3.9747297487013897,
+      "grad_norm": 0.4164454101954422,
+      "learning_rate": 2.446475023329068e-06,
+      "loss": 0.0764,
+      "step": 3539
+    },
+    {
+      "epoch": 3.975852870981328,
+      "grad_norm": 0.4441939122586264,
+      "learning_rate": 2.441336912215807e-06,
+      "loss": 0.0797,
+      "step": 3540
+    },
+    {
+      "epoch": 3.9769759932612665,
+      "grad_norm": 0.42960608918880705,
+      "learning_rate": 2.4362034520311216e-06,
+      "loss": 0.0729,
+      "step": 3541
+    },
+    {
+      "epoch": 3.9780991155412044,
+      "grad_norm": 0.44957178328850944,
+      "learning_rate": 2.4310746459336896e-06,
+      "loss": 0.0802,
+      "step": 3542
+    },
+    {
+      "epoch": 3.979222237821143,
+      "grad_norm": 0.4329168884237763,
+      "learning_rate": 2.4259504970793226e-06,
+      "loss": 0.0727,
+      "step": 3543
+    },
+    {
+      "epoch": 3.980345360101081,
+      "grad_norm": 0.45775710829454436,
+      "learning_rate": 2.4208310086209607e-06,
+      "loss": 0.0765,
+      "step": 3544
+    },
+    {
+      "epoch": 3.981468482381019,
+      "grad_norm": 0.4377294306740636,
+      "learning_rate": 2.415716183708684e-06,
+      "loss": 0.0752,
+      "step": 3545
+    },
+    {
+      "epoch": 3.9825916046609575,
+      "grad_norm": 0.443470429696325,
+      "learning_rate": 2.4106060254897002e-06,
+      "loss": 0.0803,
+      "step": 3546
+    },
+    {
+      "epoch": 3.983714726940896,
+      "grad_norm": 0.45578066038549137,
+      "learning_rate": 2.405500537108347e-06,
+      "loss": 0.0805,
+      "step": 3547
+    },
+    {
+      "epoch": 3.984837849220834,
+      "grad_norm": 0.45327149698238034,
+      "learning_rate": 2.4003997217060893e-06,
+      "loss": 0.0744,
+      "step": 3548
+    },
+    {
+      "epoch": 3.985960971500772,
+      "grad_norm": 0.4679219665710609,
+      "learning_rate": 2.395303582421511e-06,
+      "loss": 0.0817,
+      "step": 3549
+    },
+    {
+      "epoch": 3.9870840937807106,
+      "grad_norm": 0.4461073700564401,
+      "learning_rate": 2.390212122390323e-06,
+      "loss": 0.0795,
+      "step": 3550
+    },
+    {
+      "epoch": 3.9882072160606485,
+      "grad_norm": 0.44258640272166866,
+      "learning_rate": 2.385125344745359e-06,
+      "loss": 0.0764,
+      "step": 3551
+    },
+    {
+      "epoch": 3.989330338340587,
+      "grad_norm": 0.4287951915573227,
+      "learning_rate": 2.3800432526165683e-06,
+      "loss": 0.0819,
+      "step": 3552
+    },
+    {
+      "epoch": 3.9904534606205253,
+      "grad_norm": 0.4196233297951179,
+      "learning_rate": 2.37496584913102e-06,
+      "loss": 0.0768,
+      "step": 3553
+    },
+    {
+      "epoch": 3.9915765829004632,
+      "grad_norm": 0.43264036028578096,
+      "learning_rate": 2.369893137412893e-06,
+      "loss": 0.0827,
+      "step": 3554
+    },
+    {
+      "epoch": 3.9926997051804016,
+      "grad_norm": 0.44435177476439064,
+      "learning_rate": 2.3648251205834827e-06,
+      "loss": 0.0747,
+      "step": 3555
+    },
+    {
+      "epoch": 3.99382282746034,
+      "grad_norm": 0.4207067931671747,
+      "learning_rate": 2.3597618017611977e-06,
+      "loss": 0.0734,
+      "step": 3556
+    },
+    {
+      "epoch": 3.994945949740278,
+      "grad_norm": 0.42953451358834877,
+      "learning_rate": 2.3547031840615532e-06,
+      "loss": 0.0737,
+      "step": 3557
+    },
+    {
+      "epoch": 3.9960690720202163,
+      "grad_norm": 0.45552658779976596,
+      "learning_rate": 2.3496492705971753e-06,
+      "loss": 0.0766,
+      "step": 3558
+    },
+    {
+      "epoch": 3.9971921943001543,
+      "grad_norm": 0.4389923110529137,
+      "learning_rate": 2.3446000644777856e-06,
+      "loss": 0.0765,
+      "step": 3559
+    },
+    {
+      "epoch": 3.9983153165800926,
+      "grad_norm": 0.44857662285405525,
+      "learning_rate": 2.339555568810221e-06,
+      "loss": 0.0774,
+      "step": 3560
+    },
+    {
+      "epoch": 3.999438438860031,
+      "grad_norm": 0.4393713835259846,
+      "learning_rate": 2.334515786698415e-06,
+      "loss": 0.0767,
+      "step": 3561
+    },
+    {
+      "epoch": 4.000561561139969,
+      "grad_norm": 0.7414061671456206,
+      "learning_rate": 2.329480721243401e-06,
+      "loss": 0.0959,
+      "step": 3562
+    },
+    {
+      "epoch": 4.001684683419907,
+      "grad_norm": 0.35974986114637436,
+      "learning_rate": 2.3244503755433077e-06,
+      "loss": 0.0433,
+      "step": 3563
+    },
+    {
+      "epoch": 4.002807805699845,
+      "grad_norm": 0.35204376720792546,
+      "learning_rate": 2.3194247526933644e-06,
+      "loss": 0.0427,
+      "step": 3564
+    },
+    {
+      "epoch": 4.003930927979784,
+      "grad_norm": 0.30547691019965295,
+      "learning_rate": 2.3144038557858915e-06,
+      "loss": 0.037,
+      "step": 3565
+    },
+    {
+      "epoch": 4.005054050259722,
+      "grad_norm": 0.3503089261671818,
+      "learning_rate": 2.3093876879103027e-06,
+      "loss": 0.045,
+      "step": 3566
+    },
+    {
+      "epoch": 4.00617717253966,
+      "grad_norm": 0.28985834180869824,
+      "learning_rate": 2.3043762521531e-06,
+      "loss": 0.0356,
+      "step": 3567
+    },
+    {
+      "epoch": 4.007300294819599,
+      "grad_norm": 0.3078892585372151,
+      "learning_rate": 2.2993695515978767e-06,
+      "loss": 0.041,
+      "step": 3568
+    },
+    {
+      "epoch": 4.008423417099537,
+      "grad_norm": 0.2793695957969276,
+      "learning_rate": 2.2943675893253094e-06,
+      "loss": 0.035,
+      "step": 3569
+    },
+    {
+      "epoch": 4.009546539379475,
+      "grad_norm": 0.3126139214392695,
+      "learning_rate": 2.2893703684131608e-06,
+      "loss": 0.0409,
+      "step": 3570
+    },
+    {
+      "epoch": 4.0106696616594135,
+      "grad_norm": 0.32235788779278907,
+      "learning_rate": 2.284377891936277e-06,
+      "loss": 0.039,
+      "step": 3571
+    },
+    {
+      "epoch": 4.0117927839393515,
+      "grad_norm": 0.3386774994355639,
+      "learning_rate": 2.2793901629665847e-06,
+      "loss": 0.0394,
+      "step": 3572
+    },
+    {
+      "epoch": 4.012915906219289,
+      "grad_norm": 0.33966921939041494,
+      "learning_rate": 2.2744071845730843e-06,
+      "loss": 0.0378,
+      "step": 3573
+    },
+    {
+      "epoch": 4.014039028499228,
+      "grad_norm": 0.3650876310831837,
+      "learning_rate": 2.26942895982186e-06,
+      "loss": 0.0352,
+      "step": 3574
+    },
+    {
+      "epoch": 4.015162150779166,
+      "grad_norm": 0.39322271754253874,
+      "learning_rate": 2.2644554917760674e-06,
+      "loss": 0.0388,
+      "step": 3575
+    },
+    {
+      "epoch": 4.016285273059104,
+      "grad_norm": 0.3603318333009823,
+      "learning_rate": 2.2594867834959367e-06,
+      "loss": 0.0349,
+      "step": 3576
+    },
+    {
+      "epoch": 4.017408395339043,
+      "grad_norm": 0.39766487225833574,
+      "learning_rate": 2.2545228380387706e-06,
+      "loss": 0.0392,
+      "step": 3577
+    },
+    {
+      "epoch": 4.018531517618981,
+      "grad_norm": 0.3879795346602286,
+      "learning_rate": 2.2495636584589353e-06,
+      "loss": 0.0344,
+      "step": 3578
+    },
+    {
+      "epoch": 4.019654639898919,
+      "grad_norm": 0.38693724167691756,
+      "learning_rate": 2.2446092478078706e-06,
+      "loss": 0.0391,
+      "step": 3579
+    },
+    {
+      "epoch": 4.020777762178858,
+      "grad_norm": 0.3794613149433384,
+      "learning_rate": 2.2396596091340805e-06,
+      "loss": 0.0354,
+      "step": 3580
+    },
+    {
+      "epoch": 4.021900884458796,
+      "grad_norm": 0.3992223472216564,
+      "learning_rate": 2.2347147454831306e-06,
+      "loss": 0.0418,
+      "step": 3581
+    },
+    {
+      "epoch": 4.0230240067387335,
+      "grad_norm": 0.3661442590112595,
+      "learning_rate": 2.2297746598976545e-06,
+      "loss": 0.0352,
+      "step": 3582
+    },
+    {
+      "epoch": 4.024147129018672,
+      "grad_norm": 0.37467677700540847,
+      "learning_rate": 2.2248393554173344e-06,
+      "loss": 0.039,
+      "step": 3583
+    },
+    {
+      "epoch": 4.02527025129861,
+      "grad_norm": 0.3601512962433673,
+      "learning_rate": 2.219908835078921e-06,
+      "loss": 0.0358,
+      "step": 3584
+    },
+    {
+      "epoch": 4.026393373578548,
+      "grad_norm": 0.37668176536008646,
+      "learning_rate": 2.2149831019162173e-06,
+      "loss": 0.0418,
+      "step": 3585
+    },
+    {
+      "epoch": 4.027516495858487,
+      "grad_norm": 0.3258768222612785,
+      "learning_rate": 2.2100621589600813e-06,
+      "loss": 0.0334,
+      "step": 3586
+    },
+    {
+      "epoch": 4.028639618138425,
+      "grad_norm": 0.3370222447649643,
+      "learning_rate": 2.205146009238426e-06,
+      "loss": 0.0373,
+      "step": 3587
+    },
+    {
+      "epoch": 4.029762740418363,
+      "grad_norm": 0.32311730754858564,
+      "learning_rate": 2.2002346557762068e-06,
+      "loss": 0.0344,
+      "step": 3588
+    },
+    {
+      "epoch": 4.030885862698301,
+      "grad_norm": 0.3123818705115929,
+      "learning_rate": 2.1953281015954364e-06,
+      "loss": 0.0326,
+      "step": 3589
+    },
+    {
+      "epoch": 4.03200898497824,
+      "grad_norm": 0.3241220437041912,
+      "learning_rate": 2.190426349715171e-06,
+      "loss": 0.0341,
+      "step": 3590
+    },
+    {
+      "epoch": 4.033132107258178,
+      "grad_norm": 0.33250107457893724,
+      "learning_rate": 2.185529403151514e-06,
+      "loss": 0.0373,
+      "step": 3591
+    },
+    {
+      "epoch": 4.0342552295381156,
+      "grad_norm": 0.3170426069163533,
+      "learning_rate": 2.1806372649176124e-06,
+      "loss": 0.0327,
+      "step": 3592
+    },
+    {
+      "epoch": 4.035378351818054,
+      "grad_norm": 0.32095635332227473,
+      "learning_rate": 2.175749938023647e-06,
+      "loss": 0.036,
+      "step": 3593
+    },
+    {
+      "epoch": 4.036501474097992,
+      "grad_norm": 0.3296174400793843,
+      "learning_rate": 2.170867425476847e-06,
+      "loss": 0.0345,
+      "step": 3594
+    },
+    {
+      "epoch": 4.03762459637793,
+      "grad_norm": 0.32420977388665184,
+      "learning_rate": 2.165989730281475e-06,
+      "loss": 0.0358,
+      "step": 3595
+    },
+    {
+      "epoch": 4.038747718657869,
+      "grad_norm": 0.32486539485385396,
+      "learning_rate": 2.1611168554388353e-06,
+      "loss": 0.0375,
+      "step": 3596
+    },
+    {
+      "epoch": 4.039870840937807,
+      "grad_norm": 0.33856386057317117,
+      "learning_rate": 2.156248803947254e-06,
+      "loss": 0.0357,
+      "step": 3597
+    },
+    {
+      "epoch": 4.040993963217745,
+      "grad_norm": 0.32586631725378973,
+      "learning_rate": 2.1513855788021e-06,
+      "loss": 0.0354,
+      "step": 3598
+    },
+    {
+      "epoch": 4.042117085497684,
+      "grad_norm": 0.3308607327766132,
+      "learning_rate": 2.14652718299577e-06,
+      "loss": 0.0347,
+      "step": 3599
+    },
+    {
+      "epoch": 4.043240207777622,
+      "grad_norm": 0.33048976251932244,
+      "learning_rate": 2.141673619517687e-06,
+      "loss": 0.0361,
+      "step": 3600
+    },
+    {
+      "epoch": 4.04436333005756,
+      "grad_norm": 0.32515150803710624,
+      "learning_rate": 2.1368248913543065e-06,
+      "loss": 0.035,
+      "step": 3601
+    },
+    {
+      "epoch": 4.0454864523374985,
+      "grad_norm": 0.3027989162513895,
+      "learning_rate": 2.1319810014890972e-06,
+      "loss": 0.0331,
+      "step": 3602
+    },
+    {
+      "epoch": 4.046609574617436,
+      "grad_norm": 0.32565678045910307,
+      "learning_rate": 2.127141952902563e-06,
+      "loss": 0.0371,
+      "step": 3603
+    },
+    {
+      "epoch": 4.047732696897374,
+      "grad_norm": 0.30745483238302934,
+      "learning_rate": 2.12230774857222e-06,
+      "loss": 0.0322,
+      "step": 3604
+    },
+    {
+      "epoch": 4.048855819177313,
+      "grad_norm": 0.3418209827856129,
+      "learning_rate": 2.1174783914726106e-06,
+      "loss": 0.0357,
+      "step": 3605
+    },
+    {
+      "epoch": 4.049978941457251,
+      "grad_norm": 0.34150940597932833,
+      "learning_rate": 2.1126538845752918e-06,
+      "loss": 0.0361,
+      "step": 3606
+    },
+    {
+      "epoch": 4.051102063737189,
+      "grad_norm": 0.3319716006832906,
+      "learning_rate": 2.107834230848833e-06,
+      "loss": 0.0355,
+      "step": 3607
+    },
+    {
+      "epoch": 4.052225186017128,
+      "grad_norm": 0.3256155221300326,
+      "learning_rate": 2.1030194332588203e-06,
+      "loss": 0.0341,
+      "step": 3608
+    },
+    {
+      "epoch": 4.053348308297066,
+      "grad_norm": 0.3511915328594416,
+      "learning_rate": 2.098209494767853e-06,
+      "loss": 0.0357,
+      "step": 3609
+    },
+    {
+      "epoch": 4.054471430577004,
+      "grad_norm": 0.3389534143051318,
+      "learning_rate": 2.0934044183355384e-06,
+      "loss": 0.0331,
+      "step": 3610
+    },
+    {
+      "epoch": 4.055594552856943,
+      "grad_norm": 0.3311633004682248,
+      "learning_rate": 2.088604206918494e-06,
+      "loss": 0.0297,
+      "step": 3611
+    },
+    {
+      "epoch": 4.0567176751368805,
+      "grad_norm": 0.37079247885517713,
+      "learning_rate": 2.0838088634703412e-06,
+      "loss": 0.0495,
+      "step": 3612
+    },
+    {
+      "epoch": 4.0578407974168185,
+      "grad_norm": 0.3436254547868691,
+      "learning_rate": 2.0790183909417096e-06,
+      "loss": 0.0354,
+      "step": 3613
+    },
+    {
+      "epoch": 4.058963919696757,
+      "grad_norm": 0.3537084660888286,
+      "learning_rate": 2.0742327922802285e-06,
+      "loss": 0.0364,
+      "step": 3614
+    },
+    {
+      "epoch": 4.060087041976695,
+      "grad_norm": 0.33720406412003606,
+      "learning_rate": 2.069452070430529e-06,
+      "loss": 0.0357,
+      "step": 3615
+    },
+    {
+      "epoch": 4.061210164256633,
+      "grad_norm": 0.361772020239582,
+      "learning_rate": 2.0646762283342448e-06,
+      "loss": 0.0406,
+      "step": 3616
+    },
+    {
+      "epoch": 4.062333286536572,
+      "grad_norm": 0.3305905026607228,
+      "learning_rate": 2.059905268929999e-06,
+      "loss": 0.0291,
+      "step": 3617
+    },
+    {
+      "epoch": 4.06345640881651,
+      "grad_norm": 0.35558319079029205,
+      "learning_rate": 2.055139195153417e-06,
+      "loss": 0.0314,
+      "step": 3618
+    },
+    {
+      "epoch": 4.064579531096448,
+      "grad_norm": 0.347660893552983,
+      "learning_rate": 2.0503780099371196e-06,
+      "loss": 0.0357,
+      "step": 3619
+    },
+    {
+      "epoch": 4.065702653376387,
+      "grad_norm": 0.35259174178160463,
+      "learning_rate": 2.045621716210713e-06,
+      "loss": 0.0314,
+      "step": 3620
+    },
+    {
+      "epoch": 4.066825775656325,
+      "grad_norm": 0.32281631904843283,
+      "learning_rate": 2.0408703169008015e-06,
+      "loss": 0.033,
+      "step": 3621
+    },
+    {
+      "epoch": 4.067948897936263,
+      "grad_norm": 0.34725153445115975,
+      "learning_rate": 2.036123814930967e-06,
+      "loss": 0.0375,
+      "step": 3622
+    },
+    {
+      "epoch": 4.069072020216201,
+      "grad_norm": 0.35654253964781996,
+      "learning_rate": 2.0313822132217887e-06,
+      "loss": 0.0328,
+      "step": 3623
+    },
+    {
+      "epoch": 4.070195142496139,
+      "grad_norm": 0.33133915489525906,
+      "learning_rate": 2.0266455146908248e-06,
+      "loss": 0.0345,
+      "step": 3624
+    },
+    {
+      "epoch": 4.071318264776077,
+      "grad_norm": 0.36110178776774343,
+      "learning_rate": 2.0219137222526188e-06,
+      "loss": 0.0396,
+      "step": 3625
+    },
+    {
+      "epoch": 4.072441387056016,
+      "grad_norm": 0.32077710176540647,
+      "learning_rate": 2.0171868388186953e-06,
+      "loss": 0.0358,
+      "step": 3626
+    },
+    {
+      "epoch": 4.073564509335954,
+      "grad_norm": 0.3444116573250157,
+      "learning_rate": 2.0124648672975567e-06,
+      "loss": 0.0355,
+      "step": 3627
+    },
+    {
+      "epoch": 4.074687631615892,
+      "grad_norm": 0.3515163459356766,
+      "learning_rate": 2.007747810594682e-06,
+      "loss": 0.0367,
+      "step": 3628
+    },
+    {
+      "epoch": 4.075810753895831,
+      "grad_norm": 0.3450125011830339,
+      "learning_rate": 2.003035671612532e-06,
+      "loss": 0.036,
+      "step": 3629
+    },
+    {
+      "epoch": 4.076933876175769,
+      "grad_norm": 0.3547038161654613,
+      "learning_rate": 1.9983284532505343e-06,
+      "loss": 0.0372,
+      "step": 3630
+    },
+    {
+      "epoch": 4.078056998455707,
+      "grad_norm": 0.3417279392843428,
+      "learning_rate": 1.9936261584050974e-06,
+      "loss": 0.0374,
+      "step": 3631
+    },
+    {
+      "epoch": 4.0791801207356455,
+      "grad_norm": 0.34281518207305883,
+      "learning_rate": 1.9889287899695887e-06,
+      "loss": 0.0371,
+      "step": 3632
+    },
+    {
+      "epoch": 4.0803032430155834,
+      "grad_norm": 0.3272532226131082,
+      "learning_rate": 1.9842363508343532e-06,
+      "loss": 0.0344,
+      "step": 3633
+    },
+    {
+      "epoch": 4.081426365295521,
+      "grad_norm": 0.3362181119467591,
+      "learning_rate": 1.9795488438867005e-06,
+      "loss": 0.0325,
+      "step": 3634
+    },
+    {
+      "epoch": 4.08254948757546,
+      "grad_norm": 0.34910761973320087,
+      "learning_rate": 1.974866272010908e-06,
+      "loss": 0.0335,
+      "step": 3635
+    },
+    {
+      "epoch": 4.083672609855398,
+      "grad_norm": 0.3730488558950869,
+      "learning_rate": 1.9701886380882073e-06,
+      "loss": 0.0402,
+      "step": 3636
+    },
+    {
+      "epoch": 4.084795732135336,
+      "grad_norm": 0.3413923732565504,
+      "learning_rate": 1.965515944996803e-06,
+      "loss": 0.0361,
+      "step": 3637
+    },
+    {
+      "epoch": 4.085918854415274,
+      "grad_norm": 0.33096031668167075,
+      "learning_rate": 1.960848195611853e-06,
+      "loss": 0.0329,
+      "step": 3638
+    },
+    {
+      "epoch": 4.087041976695213,
+      "grad_norm": 0.34388985312179227,
+      "learning_rate": 1.9561853928054753e-06,
+      "loss": 0.0377,
+      "step": 3639
+    },
+    {
+      "epoch": 4.088165098975151,
+      "grad_norm": 0.32146386739388727,
+      "learning_rate": 1.9515275394467446e-06,
+      "loss": 0.0349,
+      "step": 3640
+    },
+    {
+      "epoch": 4.089288221255089,
+      "grad_norm": 0.33807086229307787,
+      "learning_rate": 1.946874638401688e-06,
+      "loss": 0.0336,
+      "step": 3641
+    },
+    {
+      "epoch": 4.0904113435350276,
+      "grad_norm": 0.34620877675756956,
+      "learning_rate": 1.9422266925332857e-06,
+      "loss": 0.0383,
+      "step": 3642
+    },
+    {
+      "epoch": 4.0915344658149655,
+      "grad_norm": 0.36100090203902546,
+      "learning_rate": 1.9375837047014712e-06,
+      "loss": 0.0393,
+      "step": 3643
+    },
+    {
+      "epoch": 4.092657588094903,
+      "grad_norm": 0.34073891008207474,
+      "learning_rate": 1.9329456777631273e-06,
+      "loss": 0.0342,
+      "step": 3644
+    },
+    {
+      "epoch": 4.093780710374842,
+      "grad_norm": 0.343909859895942,
+      "learning_rate": 1.928312614572083e-06,
+      "loss": 0.0346,
+      "step": 3645
+    },
+    {
+      "epoch": 4.09490383265478,
+      "grad_norm": 0.3405997491217741,
+      "learning_rate": 1.92368451797911e-06,
+      "loss": 0.0333,
+      "step": 3646
+    },
+    {
+      "epoch": 4.096026954934718,
+      "grad_norm": 0.3532535917803961,
+      "learning_rate": 1.919061390831929e-06,
+      "loss": 0.0371,
+      "step": 3647
+    },
+    {
+      "epoch": 4.097150077214657,
+      "grad_norm": 0.3402279776506247,
+      "learning_rate": 1.914443235975201e-06,
+      "loss": 0.0345,
+      "step": 3648
+    },
+    {
+      "epoch": 4.098273199494595,
+      "grad_norm": 0.33955904857288616,
+      "learning_rate": 1.9098300562505266e-06,
+      "loss": 0.0354,
+      "step": 3649
+    },
+    {
+      "epoch": 4.099396321774533,
+      "grad_norm": 0.3548495987201884,
+      "learning_rate": 1.9052218544964473e-06,
+      "loss": 0.0378,
+      "step": 3650
+    },
+    {
+      "epoch": 4.100519444054472,
+      "grad_norm": 0.3565039554766224,
+      "learning_rate": 1.9006186335484422e-06,
+      "loss": 0.0389,
+      "step": 3651
+    },
+    {
+      "epoch": 4.10164256633441,
+      "grad_norm": 0.3176237453159487,
+      "learning_rate": 1.89602039623892e-06,
+      "loss": 0.0329,
+      "step": 3652
+    },
+    {
+      "epoch": 4.1027656886143475,
+      "grad_norm": 0.34171584057174337,
+      "learning_rate": 1.8914271453972277e-06,
+      "loss": 0.0374,
+      "step": 3653
+    },
+    {
+      "epoch": 4.103888810894286,
+      "grad_norm": 0.3573392279296198,
+      "learning_rate": 1.8868388838496433e-06,
+      "loss": 0.0341,
+      "step": 3654
+    },
+    {
+      "epoch": 4.105011933174224,
+      "grad_norm": 0.3319986984776665,
+      "learning_rate": 1.882255614419376e-06,
+      "loss": 0.0343,
+      "step": 3655
+    },
+    {
+      "epoch": 4.106135055454162,
+      "grad_norm": 0.33666673936394453,
+      "learning_rate": 1.8776773399265601e-06,
+      "loss": 0.0354,
+      "step": 3656
+    },
+    {
+      "epoch": 4.107258177734101,
+      "grad_norm": 0.36043615873431406,
+      "learning_rate": 1.8731040631882591e-06,
+      "loss": 0.0354,
+      "step": 3657
+    },
+    {
+      "epoch": 4.108381300014039,
+      "grad_norm": 0.3719854481200747,
+      "learning_rate": 1.8685357870184605e-06,
+      "loss": 0.0384,
+      "step": 3658
+    },
+    {
+      "epoch": 4.109504422293977,
+      "grad_norm": 0.3351525535966926,
+      "learning_rate": 1.8639725142280752e-06,
+      "loss": 0.0336,
+      "step": 3659
+    },
+    {
+      "epoch": 4.110627544573916,
+      "grad_norm": 0.33593296146741825,
+      "learning_rate": 1.8594142476249365e-06,
+      "loss": 0.0318,
+      "step": 3660
+    },
+    {
+      "epoch": 4.111750666853854,
+      "grad_norm": 0.34070125655000394,
+      "learning_rate": 1.8548609900137926e-06,
+      "loss": 0.0381,
+      "step": 3661
+    },
+    {
+      "epoch": 4.112873789133792,
+      "grad_norm": 0.3438466556073073,
+      "learning_rate": 1.8503127441963153e-06,
+      "loss": 0.0317,
+      "step": 3662
+    },
+    {
+      "epoch": 4.1139969114137305,
+      "grad_norm": 0.3422040619381074,
+      "learning_rate": 1.8457695129710885e-06,
+      "loss": 0.0341,
+      "step": 3663
+    },
+    {
+      "epoch": 4.115120033693668,
+      "grad_norm": 0.3336224223024909,
+      "learning_rate": 1.8412312991336146e-06,
+      "loss": 0.0335,
+      "step": 3664
+    },
+    {
+      "epoch": 4.116243155973606,
+      "grad_norm": 0.3243595593137732,
+      "learning_rate": 1.8366981054763077e-06,
+      "loss": 0.034,
+      "step": 3665
+    },
+    {
+      "epoch": 4.117366278253545,
+      "grad_norm": 0.3408130322069507,
+      "learning_rate": 1.8321699347884869e-06,
+      "loss": 0.0359,
+      "step": 3666
+    },
+    {
+      "epoch": 4.118489400533483,
+      "grad_norm": 0.33585943316941114,
+      "learning_rate": 1.8276467898563887e-06,
+      "loss": 0.0369,
+      "step": 3667
+    },
+    {
+      "epoch": 4.119612522813421,
+      "grad_norm": 0.3440889810814874,
+      "learning_rate": 1.8231286734631526e-06,
+      "loss": 0.0323,
+      "step": 3668
+    },
+    {
+      "epoch": 4.12073564509336,
+      "grad_norm": 0.34877435967999315,
+      "learning_rate": 1.818615588388829e-06,
+      "loss": 0.0391,
+      "step": 3669
+    },
+    {
+      "epoch": 4.121858767373298,
+      "grad_norm": 0.33501771510850326,
+      "learning_rate": 1.8141075374103634e-06,
+      "loss": 0.0334,
+      "step": 3670
+    },
+    {
+      "epoch": 4.122981889653236,
+      "grad_norm": 0.3477079472833074,
+      "learning_rate": 1.8096045233016123e-06,
+      "loss": 0.0372,
+      "step": 3671
+    },
+    {
+      "epoch": 4.124105011933175,
+      "grad_norm": 0.352773874107671,
+      "learning_rate": 1.8051065488333285e-06,
+      "loss": 0.039,
+      "step": 3672
+    },
+    {
+      "epoch": 4.1252281342131125,
+      "grad_norm": 0.3249660660258163,
+      "learning_rate": 1.8006136167731658e-06,
+      "loss": 0.0325,
+      "step": 3673
+    },
+    {
+      "epoch": 4.1263512564930505,
+      "grad_norm": 0.352952838092871,
+      "learning_rate": 1.7961257298856783e-06,
+      "loss": 0.0406,
+      "step": 3674
+    },
+    {
+      "epoch": 4.127474378772989,
+      "grad_norm": 0.36490521962442574,
+      "learning_rate": 1.7916428909323057e-06,
+      "loss": 0.0374,
+      "step": 3675
+    },
+    {
+      "epoch": 4.128597501052927,
+      "grad_norm": 0.33698323334336555,
+      "learning_rate": 1.787165102671391e-06,
+      "loss": 0.0336,
+      "step": 3676
+    },
+    {
+      "epoch": 4.129720623332865,
+      "grad_norm": 0.32813513335813027,
+      "learning_rate": 1.7826923678581664e-06,
+      "loss": 0.034,
+      "step": 3677
+    },
+    {
+      "epoch": 4.130843745612804,
+      "grad_norm": 0.3358051137115379,
+      "learning_rate": 1.7782246892447564e-06,
+      "loss": 0.0323,
+      "step": 3678
+    },
+    {
+      "epoch": 4.131966867892742,
+      "grad_norm": 0.31172849246423406,
+      "learning_rate": 1.7737620695801737e-06,
+      "loss": 0.0306,
+      "step": 3679
+    },
+    {
+      "epoch": 4.13308999017268,
+      "grad_norm": 0.3726210328067485,
+      "learning_rate": 1.7693045116103125e-06,
+      "loss": 0.0382,
+      "step": 3680
+    },
+    {
+      "epoch": 4.134213112452619,
+      "grad_norm": 0.33518709615258124,
+      "learning_rate": 1.7648520180779605e-06,
+      "loss": 0.0319,
+      "step": 3681
+    },
+    {
+      "epoch": 4.135336234732557,
+      "grad_norm": 0.37299298008128623,
+      "learning_rate": 1.7604045917227852e-06,
+      "loss": 0.0364,
+      "step": 3682
+    },
+    {
+      "epoch": 4.136459357012495,
+      "grad_norm": 0.3471679342038857,
+      "learning_rate": 1.7559622352813366e-06,
+      "loss": 0.0346,
+      "step": 3683
+    },
+    {
+      "epoch": 4.137582479292433,
+      "grad_norm": 0.36575324426037753,
+      "learning_rate": 1.7515249514870504e-06,
+      "loss": 0.036,
+      "step": 3684
+    },
+    {
+      "epoch": 4.138705601572371,
+      "grad_norm": 0.33169330081699266,
+      "learning_rate": 1.7470927430702277e-06,
+      "loss": 0.0359,
+      "step": 3685
+    },
+    {
+      "epoch": 4.139828723852309,
+      "grad_norm": 0.34483958264370257,
+      "learning_rate": 1.7426656127580598e-06,
+      "loss": 0.0341,
+      "step": 3686
+    },
+    {
+      "epoch": 4.140951846132248,
+      "grad_norm": 0.3562439460847369,
+      "learning_rate": 1.7382435632746086e-06,
+      "loss": 0.0379,
+      "step": 3687
+    },
+    {
+      "epoch": 4.142074968412186,
+      "grad_norm": 0.31582095689600725,
+      "learning_rate": 1.7338265973408097e-06,
+      "loss": 0.0314,
+      "step": 3688
+    },
+    {
+      "epoch": 4.143198090692124,
+      "grad_norm": 0.3433619286521538,
+      "learning_rate": 1.7294147176744725e-06,
+      "loss": 0.0356,
+      "step": 3689
+    },
+    {
+      "epoch": 4.144321212972062,
+      "grad_norm": 0.3489147400059873,
+      "learning_rate": 1.7250079269902708e-06,
+      "loss": 0.0367,
+      "step": 3690
+    },
+    {
+      "epoch": 4.145444335252001,
+      "grad_norm": 0.3352194034975239,
+      "learning_rate": 1.7206062279997538e-06,
+      "loss": 0.0326,
+      "step": 3691
+    },
+    {
+      "epoch": 4.146567457531939,
+      "grad_norm": 0.355514581203735,
+      "learning_rate": 1.7162096234113358e-06,
+      "loss": 0.0372,
+      "step": 3692
+    },
+    {
+      "epoch": 4.147690579811877,
+      "grad_norm": 0.3359309769139185,
+      "learning_rate": 1.7118181159302948e-06,
+      "loss": 0.0353,
+      "step": 3693
+    },
+    {
+      "epoch": 4.148813702091815,
+      "grad_norm": 0.3380299997227094,
+      "learning_rate": 1.7074317082587755e-06,
+      "loss": 0.0375,
+      "step": 3694
+    },
+    {
+      "epoch": 4.149936824371753,
+      "grad_norm": 0.3179117693436562,
+      "learning_rate": 1.703050403095783e-06,
+      "loss": 0.0319,
+      "step": 3695
+    },
+    {
+      "epoch": 4.151059946651691,
+      "grad_norm": 0.34267472556917217,
+      "learning_rate": 1.6986742031371794e-06,
+      "loss": 0.0375,
+      "step": 3696
+    },
+    {
+      "epoch": 4.15218306893163,
+      "grad_norm": 0.3550491112940439,
+      "learning_rate": 1.6943031110756902e-06,
+      "loss": 0.0384,
+      "step": 3697
+    },
+    {
+      "epoch": 4.153306191211568,
+      "grad_norm": 0.3441009447815663,
+      "learning_rate": 1.689937129600897e-06,
+      "loss": 0.036,
+      "step": 3698
+    },
+    {
+      "epoch": 4.154429313491506,
+      "grad_norm": 0.35632790653268437,
+      "learning_rate": 1.6855762613992367e-06,
+      "loss": 0.0383,
+      "step": 3699
+    },
+    {
+      "epoch": 4.155552435771445,
+      "grad_norm": 0.3380140910478162,
+      "learning_rate": 1.6812205091539979e-06,
+      "loss": 0.0366,
+      "step": 3700
+    },
+    {
+      "epoch": 4.156675558051383,
+      "grad_norm": 0.36642723566904717,
+      "learning_rate": 1.676869875545324e-06,
+      "loss": 0.0425,
+      "step": 3701
+    },
+    {
+      "epoch": 4.157798680331321,
+      "grad_norm": 0.3486015385918872,
+      "learning_rate": 1.6725243632502074e-06,
+      "loss": 0.0365,
+      "step": 3702
+    },
+    {
+      "epoch": 4.1589218026112595,
+      "grad_norm": 0.34087715875690777,
+      "learning_rate": 1.668183974942491e-06,
+      "loss": 0.0354,
+      "step": 3703
+    },
+    {
+      "epoch": 4.1600449248911975,
+      "grad_norm": 0.3301015274275597,
+      "learning_rate": 1.6638487132928638e-06,
+      "loss": 0.0337,
+      "step": 3704
+    },
+    {
+      "epoch": 4.161168047171135,
+      "grad_norm": 0.32631689981447143,
+      "learning_rate": 1.6595185809688564e-06,
+      "loss": 0.0359,
+      "step": 3705
+    },
+    {
+      "epoch": 4.162291169451074,
+      "grad_norm": 0.33677899223356966,
+      "learning_rate": 1.6551935806348485e-06,
+      "loss": 0.0352,
+      "step": 3706
+    },
+    {
+      "epoch": 4.163414291731012,
+      "grad_norm": 0.344706832968584,
+      "learning_rate": 1.6508737149520615e-06,
+      "loss": 0.0344,
+      "step": 3707
+    },
+    {
+      "epoch": 4.16453741401095,
+      "grad_norm": 0.3319801271011935,
+      "learning_rate": 1.6465589865785581e-06,
+      "loss": 0.0311,
+      "step": 3708
+    },
+    {
+      "epoch": 4.165660536290889,
+      "grad_norm": 0.3469710548961725,
+      "learning_rate": 1.6422493981692333e-06,
+      "loss": 0.0343,
+      "step": 3709
+    },
+    {
+      "epoch": 4.166783658570827,
+      "grad_norm": 0.3459491276034882,
+      "learning_rate": 1.6379449523758262e-06,
+      "loss": 0.0341,
+      "step": 3710
+    },
+    {
+      "epoch": 4.167906780850765,
+      "grad_norm": 0.3664777228112351,
+      "learning_rate": 1.6336456518469112e-06,
+      "loss": 0.0398,
+      "step": 3711
+    },
+    {
+      "epoch": 4.169029903130704,
+      "grad_norm": 0.3404376801657948,
+      "learning_rate": 1.6293514992278935e-06,
+      "loss": 0.0326,
+      "step": 3712
+    },
+    {
+      "epoch": 4.170153025410642,
+      "grad_norm": 0.3568606518647232,
+      "learning_rate": 1.6250624971610152e-06,
+      "loss": 0.0331,
+      "step": 3713
+    },
+    {
+      "epoch": 4.1712761476905795,
+      "grad_norm": 0.34901751847554274,
+      "learning_rate": 1.6207786482853428e-06,
+      "loss": 0.0376,
+      "step": 3714
+    },
+    {
+      "epoch": 4.172399269970518,
+      "grad_norm": 0.3399558991517041,
+      "learning_rate": 1.6164999552367767e-06,
+      "loss": 0.0314,
+      "step": 3715
+    },
+    {
+      "epoch": 4.173522392250456,
+      "grad_norm": 0.33061099011882433,
+      "learning_rate": 1.6122264206480443e-06,
+      "loss": 0.0329,
+      "step": 3716
+    },
+    {
+      "epoch": 4.174645514530394,
+      "grad_norm": 0.3710949614187633,
+      "learning_rate": 1.6079580471486988e-06,
+      "loss": 0.033,
+      "step": 3717
+    },
+    {
+      "epoch": 4.175768636810333,
+      "grad_norm": 0.3769411467365772,
+      "learning_rate": 1.6036948373651195e-06,
+      "loss": 0.037,
+      "step": 3718
+    },
+    {
+      "epoch": 4.176891759090271,
+      "grad_norm": 0.3349294639560482,
+      "learning_rate": 1.5994367939205012e-06,
+      "loss": 0.0308,
+      "step": 3719
+    },
+    {
+      "epoch": 4.178014881370209,
+      "grad_norm": 0.322230527234023,
+      "learning_rate": 1.5951839194348684e-06,
+      "loss": 0.033,
+      "step": 3720
+    },
+    {
+      "epoch": 4.179138003650148,
+      "grad_norm": 0.3423969186367366,
+      "learning_rate": 1.5909362165250609e-06,
+      "loss": 0.0333,
+      "step": 3721
+    },
+    {
+      "epoch": 4.180261125930086,
+      "grad_norm": 0.34016184747977096,
+      "learning_rate": 1.5866936878047368e-06,
+      "loss": 0.0329,
+      "step": 3722
+    },
+    {
+      "epoch": 4.181384248210024,
+      "grad_norm": 0.3606283637266968,
+      "learning_rate": 1.5824563358843725e-06,
+      "loss": 0.0385,
+      "step": 3723
+    },
+    {
+      "epoch": 4.1825073704899625,
+      "grad_norm": 0.36699375787006455,
+      "learning_rate": 1.5782241633712536e-06,
+      "loss": 0.0323,
+      "step": 3724
+    },
+    {
+      "epoch": 4.1836304927699,
+      "grad_norm": 0.34062675166929457,
+      "learning_rate": 1.5739971728694848e-06,
+      "loss": 0.0323,
+      "step": 3725
+    },
+    {
+      "epoch": 4.184753615049838,
+      "grad_norm": 0.3585907476688318,
+      "learning_rate": 1.5697753669799788e-06,
+      "loss": 0.0352,
+      "step": 3726
+    },
+    {
+      "epoch": 4.185876737329777,
+      "grad_norm": 0.3627326102547834,
+      "learning_rate": 1.5655587483004608e-06,
+      "loss": 0.0352,
+      "step": 3727
+    },
+    {
+      "epoch": 4.186999859609715,
+      "grad_norm": 0.3202203931611738,
+      "learning_rate": 1.5613473194254636e-06,
+      "loss": 0.0323,
+      "step": 3728
+    },
+    {
+      "epoch": 4.188122981889653,
+      "grad_norm": 0.35906124847404014,
+      "learning_rate": 1.5571410829463218e-06,
+      "loss": 0.033,
+      "step": 3729
+    },
+    {
+      "epoch": 4.189246104169592,
+      "grad_norm": 0.3407268596851762,
+      "learning_rate": 1.5529400414511809e-06,
+      "loss": 0.0363,
+      "step": 3730
+    },
+    {
+      "epoch": 4.19036922644953,
+      "grad_norm": 0.3301264157569526,
+      "learning_rate": 1.5487441975249885e-06,
+      "loss": 0.0308,
+      "step": 3731
+    },
+    {
+      "epoch": 4.191492348729468,
+      "grad_norm": 0.35754773559343533,
+      "learning_rate": 1.5445535537494926e-06,
+      "loss": 0.0341,
+      "step": 3732
+    },
+    {
+      "epoch": 4.192615471009407,
+      "grad_norm": 0.3546319476031264,
+      "learning_rate": 1.5403681127032466e-06,
+      "loss": 0.0354,
+      "step": 3733
+    },
+    {
+      "epoch": 4.1937385932893445,
+      "grad_norm": 0.3334688214498469,
+      "learning_rate": 1.5361878769615913e-06,
+      "loss": 0.0363,
+      "step": 3734
+    },
+    {
+      "epoch": 4.1948617155692824,
+      "grad_norm": 0.359559039318054,
+      "learning_rate": 1.5320128490966768e-06,
+      "loss": 0.0402,
+      "step": 3735
+    },
+    {
+      "epoch": 4.19598483784922,
+      "grad_norm": 0.3559528249651165,
+      "learning_rate": 1.5278430316774406e-06,
+      "loss": 0.0336,
+      "step": 3736
+    },
+    {
+      "epoch": 4.197107960129159,
+      "grad_norm": 0.3389181856407088,
+      "learning_rate": 1.5236784272696204e-06,
+      "loss": 0.0363,
+      "step": 3737
+    },
+    {
+      "epoch": 4.198231082409097,
+      "grad_norm": 0.33514425271731724,
+      "learning_rate": 1.5195190384357405e-06,
+      "loss": 0.0315,
+      "step": 3738
+    },
+    {
+      "epoch": 4.199354204689035,
+      "grad_norm": 0.34519876807313443,
+      "learning_rate": 1.5153648677351196e-06,
+      "loss": 0.0355,
+      "step": 3739
+    },
+    {
+      "epoch": 4.200477326968974,
+      "grad_norm": 0.32555795899970486,
+      "learning_rate": 1.5112159177238683e-06,
+      "loss": 0.0322,
+      "step": 3740
+    },
+    {
+      "epoch": 4.201600449248912,
+      "grad_norm": 0.3499784570691095,
+      "learning_rate": 1.5070721909548747e-06,
+      "loss": 0.036,
+      "step": 3741
+    },
+    {
+      "epoch": 4.20272357152885,
+      "grad_norm": 0.3454531769390773,
+      "learning_rate": 1.5029336899778224e-06,
+      "loss": 0.0351,
+      "step": 3742
+    },
+    {
+      "epoch": 4.203846693808789,
+      "grad_norm": 0.33874193071072906,
+      "learning_rate": 1.4988004173391769e-06,
+      "loss": 0.0342,
+      "step": 3743
+    },
+    {
+      "epoch": 4.2049698160887266,
+      "grad_norm": 0.3427966066704258,
+      "learning_rate": 1.4946723755821858e-06,
+      "loss": 0.0361,
+      "step": 3744
+    },
+    {
+      "epoch": 4.2060929383686645,
+      "grad_norm": 0.33371125691338493,
+      "learning_rate": 1.4905495672468784e-06,
+      "loss": 0.0335,
+      "step": 3745
+    },
+    {
+      "epoch": 4.207216060648603,
+      "grad_norm": 0.348963362736691,
+      "learning_rate": 1.4864319948700656e-06,
+      "loss": 0.0355,
+      "step": 3746
+    },
+    {
+      "epoch": 4.208339182928541,
+      "grad_norm": 0.3440752736809589,
+      "learning_rate": 1.4823196609853362e-06,
+      "loss": 0.0306,
+      "step": 3747
+    },
+    {
+      "epoch": 4.209462305208479,
+      "grad_norm": 0.3531490219546731,
+      "learning_rate": 1.4782125681230497e-06,
+      "loss": 0.0315,
+      "step": 3748
+    },
+    {
+      "epoch": 4.210585427488418,
+      "grad_norm": 0.3507377236203947,
+      "learning_rate": 1.4741107188103477e-06,
+      "loss": 0.0337,
+      "step": 3749
+    },
+    {
+      "epoch": 4.211708549768356,
+      "grad_norm": 0.33761169130704055,
+      "learning_rate": 1.4700141155711433e-06,
+      "loss": 0.0344,
+      "step": 3750
+    },
+    {
+      "epoch": 4.212831672048294,
+      "grad_norm": 0.5661341702283312,
+      "learning_rate": 1.465922760926123e-06,
+      "loss": 0.0341,
+      "step": 3751
+    },
+    {
+      "epoch": 4.213954794328233,
+      "grad_norm": 0.3575202836894078,
+      "learning_rate": 1.4618366573927423e-06,
+      "loss": 0.0367,
+      "step": 3752
+    },
+    {
+      "epoch": 4.215077916608171,
+      "grad_norm": 0.3513549935655655,
+      "learning_rate": 1.4577558074852228e-06,
+      "loss": 0.0374,
+      "step": 3753
+    },
+    {
+      "epoch": 4.216201038888109,
+      "grad_norm": 0.3315282047974937,
+      "learning_rate": 1.453680213714559e-06,
+      "loss": 0.0319,
+      "step": 3754
+    },
+    {
+      "epoch": 4.217324161168047,
+      "grad_norm": 0.3718187160987202,
+      "learning_rate": 1.449609878588506e-06,
+      "loss": 0.0375,
+      "step": 3755
+    },
+    {
+      "epoch": 4.218447283447985,
+      "grad_norm": 0.3375869813485876,
+      "learning_rate": 1.4455448046115884e-06,
+      "loss": 0.0336,
+      "step": 3756
+    },
+    {
+      "epoch": 4.219570405727923,
+      "grad_norm": 0.34005174509994746,
+      "learning_rate": 1.4414849942850927e-06,
+      "loss": 0.0341,
+      "step": 3757
+    },
+    {
+      "epoch": 4.220693528007862,
+      "grad_norm": 0.33034188119181235,
+      "learning_rate": 1.4374304501070592e-06,
+      "loss": 0.0329,
+      "step": 3758
+    },
+    {
+      "epoch": 4.2218166502878,
+      "grad_norm": 0.3356259815303574,
+      "learning_rate": 1.433381174572297e-06,
+      "loss": 0.0342,
+      "step": 3759
+    },
+    {
+      "epoch": 4.222939772567738,
+      "grad_norm": 0.3410011093554527,
+      "learning_rate": 1.4293371701723701e-06,
+      "loss": 0.0335,
+      "step": 3760
+    },
+    {
+      "epoch": 4.224062894847677,
+      "grad_norm": 0.34434008829412244,
+      "learning_rate": 1.425298439395597e-06,
+      "loss": 0.0342,
+      "step": 3761
+    },
+    {
+      "epoch": 4.225186017127615,
+      "grad_norm": 0.32274850049291476,
+      "learning_rate": 1.4212649847270576e-06,
+      "loss": 0.0344,
+      "step": 3762
+    },
+    {
+      "epoch": 4.226309139407553,
+      "grad_norm": 0.36333470622076525,
+      "learning_rate": 1.4172368086485755e-06,
+      "loss": 0.0363,
+      "step": 3763
+    },
+    {
+      "epoch": 4.2274322616874915,
+      "grad_norm": 0.3479702931501896,
+      "learning_rate": 1.4132139136387334e-06,
+      "loss": 0.0359,
+      "step": 3764
+    },
+    {
+      "epoch": 4.2285553839674295,
+      "grad_norm": 0.33884738891643645,
+      "learning_rate": 1.4091963021728639e-06,
+      "loss": 0.0331,
+      "step": 3765
+    },
+    {
+      "epoch": 4.229678506247367,
+      "grad_norm": 0.35343625411719487,
+      "learning_rate": 1.4051839767230479e-06,
+      "loss": 0.0352,
+      "step": 3766
+    },
+    {
+      "epoch": 4.230801628527306,
+      "grad_norm": 0.3370621206104113,
+      "learning_rate": 1.4011769397581143e-06,
+      "loss": 0.035,
+      "step": 3767
+    },
+    {
+      "epoch": 4.231924750807244,
+      "grad_norm": 0.3693570022771069,
+      "learning_rate": 1.397175193743633e-06,
+      "loss": 0.0379,
+      "step": 3768
+    },
+    {
+      "epoch": 4.233047873087182,
+      "grad_norm": 0.34977327665429203,
+      "learning_rate": 1.3931787411419252e-06,
+      "loss": 0.0335,
+      "step": 3769
+    },
+    {
+      "epoch": 4.234170995367121,
+      "grad_norm": 0.35935908738327177,
+      "learning_rate": 1.3891875844120517e-06,
+      "loss": 0.0354,
+      "step": 3770
+    },
+    {
+      "epoch": 4.235294117647059,
+      "grad_norm": 0.33995180258642155,
+      "learning_rate": 1.385201726009815e-06,
+      "loss": 0.0319,
+      "step": 3771
+    },
+    {
+      "epoch": 4.236417239926997,
+      "grad_norm": 0.33986459891753673,
+      "learning_rate": 1.3812211683877608e-06,
+      "loss": 0.0322,
+      "step": 3772
+    },
+    {
+      "epoch": 4.237540362206936,
+      "grad_norm": 0.3418058114878745,
+      "learning_rate": 1.3772459139951643e-06,
+      "loss": 0.0351,
+      "step": 3773
+    },
+    {
+      "epoch": 4.238663484486874,
+      "grad_norm": 0.3556017630987595,
+      "learning_rate": 1.3732759652780458e-06,
+      "loss": 0.0384,
+      "step": 3774
+    },
+    {
+      "epoch": 4.2397866067668115,
+      "grad_norm": 0.3293972750031755,
+      "learning_rate": 1.369311324679159e-06,
+      "loss": 0.0337,
+      "step": 3775
+    },
+    {
+      "epoch": 4.24090972904675,
+      "grad_norm": 0.3212509101961188,
+      "learning_rate": 1.3653519946379912e-06,
+      "loss": 0.0345,
+      "step": 3776
+    },
+    {
+      "epoch": 4.242032851326688,
+      "grad_norm": 0.34182476674452744,
+      "learning_rate": 1.3613979775907627e-06,
+      "loss": 0.0366,
+      "step": 3777
+    },
+    {
+      "epoch": 4.243155973606626,
+      "grad_norm": 0.3369835165724027,
+      "learning_rate": 1.3574492759704194e-06,
+      "loss": 0.0336,
+      "step": 3778
+    },
+    {
+      "epoch": 4.244279095886565,
+      "grad_norm": 0.3377925126914695,
+      "learning_rate": 1.3535058922066447e-06,
+      "loss": 0.035,
+      "step": 3779
+    },
+    {
+      "epoch": 4.245402218166503,
+      "grad_norm": 0.33513094535575855,
+      "learning_rate": 1.349567828725844e-06,
+      "loss": 0.033,
+      "step": 3780
+    },
+    {
+      "epoch": 4.246525340446441,
+      "grad_norm": 0.3396136926844531,
+      "learning_rate": 1.3456350879511526e-06,
+      "loss": 0.0343,
+      "step": 3781
+    },
+    {
+      "epoch": 4.24764846272638,
+      "grad_norm": 0.34445639039011566,
+      "learning_rate": 1.3417076723024281e-06,
+      "loss": 0.0337,
+      "step": 3782
+    },
+    {
+      "epoch": 4.248771585006318,
+      "grad_norm": 0.3451407077304778,
+      "learning_rate": 1.3377855841962528e-06,
+      "loss": 0.0321,
+      "step": 3783
+    },
+    {
+      "epoch": 4.249894707286256,
+      "grad_norm": 0.35973782888492206,
+      "learning_rate": 1.333868826045932e-06,
+      "loss": 0.0384,
+      "step": 3784
+    },
+    {
+      "epoch": 4.2510178295661945,
+      "grad_norm": 0.351803523339287,
+      "learning_rate": 1.3299574002614901e-06,
+      "loss": 0.0355,
+      "step": 3785
+    },
+    {
+      "epoch": 4.252140951846132,
+      "grad_norm": 0.34276230294757215,
+      "learning_rate": 1.3260513092496674e-06,
+      "loss": 0.0334,
+      "step": 3786
+    },
+    {
+      "epoch": 4.25326407412607,
+      "grad_norm": 0.3690545859860941,
+      "learning_rate": 1.322150555413927e-06,
+      "loss": 0.0347,
+      "step": 3787
+    },
+    {
+      "epoch": 4.254387196406009,
+      "grad_norm": 0.3630986947686924,
+      "learning_rate": 1.3182551411544454e-06,
+      "loss": 0.0359,
+      "step": 3788
+    },
+    {
+      "epoch": 4.255510318685947,
+      "grad_norm": 0.3479212681472171,
+      "learning_rate": 1.314365068868113e-06,
+      "loss": 0.0313,
+      "step": 3789
+    },
+    {
+      "epoch": 4.256633440965885,
+      "grad_norm": 0.34790711153559306,
+      "learning_rate": 1.3104803409485357e-06,
+      "loss": 0.0364,
+      "step": 3790
+    },
+    {
+      "epoch": 4.257756563245824,
+      "grad_norm": 0.3306603188465356,
+      "learning_rate": 1.3066009597860295e-06,
+      "loss": 0.0356,
+      "step": 3791
+    },
+    {
+      "epoch": 4.258879685525762,
+      "grad_norm": 0.34564641948043884,
+      "learning_rate": 1.302726927767618e-06,
+      "loss": 0.0357,
+      "step": 3792
+    },
+    {
+      "epoch": 4.2600028078057,
+      "grad_norm": 0.3585307062453091,
+      "learning_rate": 1.2988582472770372e-06,
+      "loss": 0.0376,
+      "step": 3793
+    },
+    {
+      "epoch": 4.261125930085638,
+      "grad_norm": 0.3297599087847977,
+      "learning_rate": 1.2949949206947276e-06,
+      "loss": 0.035,
+      "step": 3794
+    },
+    {
+      "epoch": 4.2622490523655765,
+      "grad_norm": 0.3581042327920125,
+      "learning_rate": 1.2911369503978389e-06,
+      "loss": 0.0427,
+      "step": 3795
+    },
+    {
+      "epoch": 4.263372174645514,
+      "grad_norm": 0.34205869714374937,
+      "learning_rate": 1.287284338760222e-06,
+      "loss": 0.0333,
+      "step": 3796
+    },
+    {
+      "epoch": 4.264495296925452,
+      "grad_norm": 0.37851885751818304,
+      "learning_rate": 1.2834370881524294e-06,
+      "loss": 0.0344,
+      "step": 3797
+    },
+    {
+      "epoch": 4.265618419205391,
+      "grad_norm": 0.3658512028956833,
+      "learning_rate": 1.2795952009417178e-06,
+      "loss": 0.0357,
+      "step": 3798
+    },
+    {
+      "epoch": 4.266741541485329,
+      "grad_norm": 0.35155118073032193,
+      "learning_rate": 1.275758679492043e-06,
+      "loss": 0.0323,
+      "step": 3799
+    },
+    {
+      "epoch": 4.267864663765267,
+      "grad_norm": 0.3634666090752805,
+      "learning_rate": 1.2719275261640584e-06,
+      "loss": 0.0368,
+      "step": 3800
+    },
+    {
+      "epoch": 4.268987786045206,
+      "grad_norm": 0.3434528301477464,
+      "learning_rate": 1.2681017433151166e-06,
+      "loss": 0.0332,
+      "step": 3801
+    },
+    {
+      "epoch": 4.270110908325144,
+      "grad_norm": 0.3551781416800868,
+      "learning_rate": 1.264281333299261e-06,
+      "loss": 0.0358,
+      "step": 3802
+    },
+    {
+      "epoch": 4.271234030605082,
+      "grad_norm": 0.35818119334156057,
+      "learning_rate": 1.2604662984672333e-06,
+      "loss": 0.0358,
+      "step": 3803
+    },
+    {
+      "epoch": 4.272357152885021,
+      "grad_norm": 0.37688093949994683,
+      "learning_rate": 1.256656641166466e-06,
+      "loss": 0.0389,
+      "step": 3804
+    },
+    {
+      "epoch": 4.2734802751649585,
+      "grad_norm": 0.33982096154376207,
+      "learning_rate": 1.252852363741084e-06,
+      "loss": 0.0328,
+      "step": 3805
+    },
+    {
+      "epoch": 4.2746033974448965,
+      "grad_norm": 0.33879902680647445,
+      "learning_rate": 1.2490534685319022e-06,
+      "loss": 0.034,
+      "step": 3806
+    },
+    {
+      "epoch": 4.275726519724835,
+      "grad_norm": 0.341933822830432,
+      "learning_rate": 1.2452599578764191e-06,
+      "loss": 0.0359,
+      "step": 3807
+    },
+    {
+      "epoch": 4.276849642004773,
+      "grad_norm": 0.35541574499799117,
+      "learning_rate": 1.241471834108825e-06,
+      "loss": 0.0375,
+      "step": 3808
+    },
+    {
+      "epoch": 4.277972764284711,
+      "grad_norm": 0.33786723192624557,
+      "learning_rate": 1.2376890995599955e-06,
+      "loss": 0.0334,
+      "step": 3809
+    },
+    {
+      "epoch": 4.27909588656465,
+      "grad_norm": 0.3586483875205426,
+      "learning_rate": 1.2339117565574877e-06,
+      "loss": 0.0344,
+      "step": 3810
+    },
+    {
+      "epoch": 4.280219008844588,
+      "grad_norm": 0.33925970086487844,
+      "learning_rate": 1.2301398074255444e-06,
+      "loss": 0.0317,
+      "step": 3811
+    },
+    {
+      "epoch": 4.281342131124526,
+      "grad_norm": 0.34451634085780747,
+      "learning_rate": 1.2263732544850826e-06,
+      "loss": 0.0324,
+      "step": 3812
+    },
+    {
+      "epoch": 4.282465253404465,
+      "grad_norm": 0.3534467917065109,
+      "learning_rate": 1.2226121000537082e-06,
+      "loss": 0.0353,
+      "step": 3813
+    },
+    {
+      "epoch": 4.283588375684403,
+      "grad_norm": 0.34295732835300846,
+      "learning_rate": 1.2188563464456993e-06,
+      "loss": 0.0344,
+      "step": 3814
+    },
+    {
+      "epoch": 4.284711497964341,
+      "grad_norm": 0.3467884753627846,
+      "learning_rate": 1.2151059959720136e-06,
+      "loss": 0.0353,
+      "step": 3815
+    },
+    {
+      "epoch": 4.285834620244279,
+      "grad_norm": 0.36125586893387135,
+      "learning_rate": 1.2113610509402806e-06,
+      "loss": 0.0352,
+      "step": 3816
+    },
+    {
+      "epoch": 4.286957742524217,
+      "grad_norm": 0.3550225603303854,
+      "learning_rate": 1.2076215136548076e-06,
+      "loss": 0.0332,
+      "step": 3817
+    },
+    {
+      "epoch": 4.288080864804155,
+      "grad_norm": 0.31427786748412406,
+      "learning_rate": 1.2038873864165734e-06,
+      "loss": 0.0297,
+      "step": 3818
+    },
+    {
+      "epoch": 4.289203987084094,
+      "grad_norm": 0.3748675811463692,
+      "learning_rate": 1.200158671523226e-06,
+      "loss": 0.0397,
+      "step": 3819
+    },
+    {
+      "epoch": 4.290327109364032,
+      "grad_norm": 0.33911019219908817,
+      "learning_rate": 1.196435371269089e-06,
+      "loss": 0.0335,
+      "step": 3820
+    },
+    {
+      "epoch": 4.29145023164397,
+      "grad_norm": 0.3578468474443971,
+      "learning_rate": 1.1927174879451442e-06,
+      "loss": 0.0385,
+      "step": 3821
+    },
+    {
+      "epoch": 4.292573353923909,
+      "grad_norm": 0.325444432765532,
+      "learning_rate": 1.1890050238390493e-06,
+      "loss": 0.033,
+      "step": 3822
+    },
+    {
+      "epoch": 4.293696476203847,
+      "grad_norm": 0.38444358306981036,
+      "learning_rate": 1.185297981235124e-06,
+      "loss": 0.0414,
+      "step": 3823
+    },
+    {
+      "epoch": 4.294819598483785,
+      "grad_norm": 0.31738949785516596,
+      "learning_rate": 1.1815963624143522e-06,
+      "loss": 0.0309,
+      "step": 3824
+    },
+    {
+      "epoch": 4.2959427207637235,
+      "grad_norm": 0.3407919871654333,
+      "learning_rate": 1.1779001696543802e-06,
+      "loss": 0.0374,
+      "step": 3825
+    },
+    {
+      "epoch": 4.2970658430436615,
+      "grad_norm": 0.33587739133134814,
+      "learning_rate": 1.1742094052295172e-06,
+      "loss": 0.031,
+      "step": 3826
+    },
+    {
+      "epoch": 4.298188965323599,
+      "grad_norm": 0.36769806503943114,
+      "learning_rate": 1.1705240714107301e-06,
+      "loss": 0.0355,
+      "step": 3827
+    },
+    {
+      "epoch": 4.299312087603538,
+      "grad_norm": 0.35570550048123933,
+      "learning_rate": 1.1668441704656463e-06,
+      "loss": 0.033,
+      "step": 3828
+    },
+    {
+      "epoch": 4.300435209883476,
+      "grad_norm": 0.3535248321200532,
+      "learning_rate": 1.1631697046585511e-06,
+      "loss": 0.0345,
+      "step": 3829
+    },
+    {
+      "epoch": 4.301558332163414,
+      "grad_norm": 0.34749247467018113,
+      "learning_rate": 1.1595006762503791e-06,
+      "loss": 0.0346,
+      "step": 3830
+    },
+    {
+      "epoch": 4.302681454443353,
+      "grad_norm": 0.338129590666788,
+      "learning_rate": 1.1558370874987268e-06,
+      "loss": 0.0326,
+      "step": 3831
+    },
+    {
+      "epoch": 4.303804576723291,
+      "grad_norm": 0.3494915631570644,
+      "learning_rate": 1.1521789406578399e-06,
+      "loss": 0.0351,
+      "step": 3832
+    },
+    {
+      "epoch": 4.304927699003229,
+      "grad_norm": 0.3378744608121184,
+      "learning_rate": 1.1485262379786166e-06,
+      "loss": 0.0341,
+      "step": 3833
+    },
+    {
+      "epoch": 4.306050821283167,
+      "grad_norm": 0.3419544672603244,
+      "learning_rate": 1.1448789817086048e-06,
+      "loss": 0.0344,
+      "step": 3834
+    },
+    {
+      "epoch": 4.307173943563106,
+      "grad_norm": 0.3291924046121765,
+      "learning_rate": 1.1412371740920036e-06,
+      "loss": 0.0316,
+      "step": 3835
+    },
+    {
+      "epoch": 4.3082970658430435,
+      "grad_norm": 0.35966027874793216,
+      "learning_rate": 1.137600817369654e-06,
+      "loss": 0.0362,
+      "step": 3836
+    },
+    {
+      "epoch": 4.309420188122981,
+      "grad_norm": 0.3557579953353349,
+      "learning_rate": 1.1339699137790483e-06,
+      "loss": 0.0348,
+      "step": 3837
+    },
+    {
+      "epoch": 4.31054331040292,
+      "grad_norm": 0.3437370526418855,
+      "learning_rate": 1.1303444655543206e-06,
+      "loss": 0.0351,
+      "step": 3838
+    },
+    {
+      "epoch": 4.311666432682858,
+      "grad_norm": 0.3322992260093866,
+      "learning_rate": 1.1267244749262496e-06,
+      "loss": 0.0361,
+      "step": 3839
+    },
+    {
+      "epoch": 4.312789554962796,
+      "grad_norm": 0.3506274769092776,
+      "learning_rate": 1.123109944122256e-06,
+      "loss": 0.0346,
+      "step": 3840
+    },
+    {
+      "epoch": 4.313912677242735,
+      "grad_norm": 0.3470389405075643,
+      "learning_rate": 1.1195008753663982e-06,
+      "loss": 0.0328,
+      "step": 3841
+    },
+    {
+      "epoch": 4.315035799522673,
+      "grad_norm": 0.3428900746970874,
+      "learning_rate": 1.115897270879378e-06,
+      "loss": 0.0358,
+      "step": 3842
+    },
+    {
+      "epoch": 4.316158921802611,
+      "grad_norm": 0.3381168228309907,
+      "learning_rate": 1.1122991328785315e-06,
+      "loss": 0.0318,
+      "step": 3843
+    },
+    {
+      "epoch": 4.31728204408255,
+      "grad_norm": 0.35370094895864085,
+      "learning_rate": 1.1087064635778333e-06,
+      "loss": 0.0364,
+      "step": 3844
+    },
+    {
+      "epoch": 4.318405166362488,
+      "grad_norm": 0.34626426769086593,
+      "learning_rate": 1.1051192651878938e-06,
+      "loss": 0.0348,
+      "step": 3845
+    },
+    {
+      "epoch": 4.3195282886424256,
+      "grad_norm": 0.34679291214864055,
+      "learning_rate": 1.1015375399159533e-06,
+      "loss": 0.0347,
+      "step": 3846
+    },
+    {
+      "epoch": 4.320651410922364,
+      "grad_norm": 0.3267291332902231,
+      "learning_rate": 1.0979612899658875e-06,
+      "loss": 0.0312,
+      "step": 3847
+    },
+    {
+      "epoch": 4.321774533202302,
+      "grad_norm": 0.33086846753238264,
+      "learning_rate": 1.0943905175382018e-06,
+      "loss": 0.0368,
+      "step": 3848
+    },
+    {
+      "epoch": 4.32289765548224,
+      "grad_norm": 0.34911582216245135,
+      "learning_rate": 1.0908252248300332e-06,
+      "loss": 0.0357,
+      "step": 3849
+    },
+    {
+      "epoch": 4.324020777762179,
+      "grad_norm": 0.330549307989673,
+      "learning_rate": 1.0872654140351458e-06,
+      "loss": 0.0328,
+      "step": 3850
+    },
+    {
+      "epoch": 4.325143900042117,
+      "grad_norm": 0.3377978034641037,
+      "learning_rate": 1.0837110873439282e-06,
+      "loss": 0.0323,
+      "step": 3851
+    },
+    {
+      "epoch": 4.326267022322055,
+      "grad_norm": 0.3641220302640594,
+      "learning_rate": 1.080162246943398e-06,
+      "loss": 0.0386,
+      "step": 3852
+    },
+    {
+      "epoch": 4.327390144601994,
+      "grad_norm": 0.3597585412481393,
+      "learning_rate": 1.0766188950171952e-06,
+      "loss": 0.034,
+      "step": 3853
+    },
+    {
+      "epoch": 4.328513266881932,
+      "grad_norm": 0.34322666886228304,
+      "learning_rate": 1.0730810337455856e-06,
+      "loss": 0.0321,
+      "step": 3854
+    },
+    {
+      "epoch": 4.32963638916187,
+      "grad_norm": 0.3210457881854714,
+      "learning_rate": 1.069548665305451e-06,
+      "loss": 0.0295,
+      "step": 3855
+    },
+    {
+      "epoch": 4.3307595114418085,
+      "grad_norm": 0.36948984897642523,
+      "learning_rate": 1.0660217918702965e-06,
+      "loss": 0.0361,
+      "step": 3856
+    },
+    {
+      "epoch": 4.331882633721746,
+      "grad_norm": 0.4271058260932982,
+      "learning_rate": 1.0625004156102492e-06,
+      "loss": 0.0354,
+      "step": 3857
+    },
+    {
+      "epoch": 4.333005756001684,
+      "grad_norm": 0.3549117455827815,
+      "learning_rate": 1.0589845386920473e-06,
+      "loss": 0.0387,
+      "step": 3858
+    },
+    {
+      "epoch": 4.334128878281623,
+      "grad_norm": 0.35534169762226714,
+      "learning_rate": 1.0554741632790532e-06,
+      "loss": 0.0361,
+      "step": 3859
+    },
+    {
+      "epoch": 4.335252000561561,
+      "grad_norm": 0.3399213392541908,
+      "learning_rate": 1.051969291531234e-06,
+      "loss": 0.0346,
+      "step": 3860
+    },
+    {
+      "epoch": 4.336375122841499,
+      "grad_norm": 0.3706100498973919,
+      "learning_rate": 1.0484699256051788e-06,
+      "loss": 0.0353,
+      "step": 3861
+    },
+    {
+      "epoch": 4.337498245121438,
+      "grad_norm": 0.3578491650225699,
+      "learning_rate": 1.0449760676540844e-06,
+      "loss": 0.0342,
+      "step": 3862
+    },
+    {
+      "epoch": 4.338621367401376,
+      "grad_norm": 0.32948715354128266,
+      "learning_rate": 1.0414877198277629e-06,
+      "loss": 0.0329,
+      "step": 3863
+    },
+    {
+      "epoch": 4.339744489681314,
+      "grad_norm": 0.3651556904940354,
+      "learning_rate": 1.038004884272632e-06,
+      "loss": 0.0348,
+      "step": 3864
+    },
+    {
+      "epoch": 4.340867611961253,
+      "grad_norm": 0.3745318904885555,
+      "learning_rate": 1.0345275631317165e-06,
+      "loss": 0.0371,
+      "step": 3865
+    },
+    {
+      "epoch": 4.3419907342411905,
+      "grad_norm": 0.3408489479974244,
+      "learning_rate": 1.0310557585446523e-06,
+      "loss": 0.0343,
+      "step": 3866
+    },
+    {
+      "epoch": 4.3431138565211285,
+      "grad_norm": 0.3616137231183904,
+      "learning_rate": 1.0275894726476787e-06,
+      "loss": 0.0382,
+      "step": 3867
+    },
+    {
+      "epoch": 4.344236978801067,
+      "grad_norm": 0.35337736664475866,
+      "learning_rate": 1.0241287075736384e-06,
+      "loss": 0.0327,
+      "step": 3868
+    },
+    {
+      "epoch": 4.345360101081005,
+      "grad_norm": 0.34065649357736527,
+      "learning_rate": 1.0206734654519802e-06,
+      "loss": 0.0362,
+      "step": 3869
+    },
+    {
+      "epoch": 4.346483223360943,
+      "grad_norm": 0.3677984704496751,
+      "learning_rate": 1.0172237484087522e-06,
+      "loss": 0.0383,
+      "step": 3870
+    },
+    {
+      "epoch": 4.347606345640882,
+      "grad_norm": 0.3504661828261147,
+      "learning_rate": 1.0137795585666023e-06,
+      "loss": 0.0355,
+      "step": 3871
+    },
+    {
+      "epoch": 4.34872946792082,
+      "grad_norm": 0.3547684524121742,
+      "learning_rate": 1.0103408980447793e-06,
+      "loss": 0.0358,
+      "step": 3872
+    },
+    {
+      "epoch": 4.349852590200758,
+      "grad_norm": 0.33899684366263133,
+      "learning_rate": 1.0069077689591279e-06,
+      "loss": 0.0309,
+      "step": 3873
+    },
+    {
+      "epoch": 4.350975712480697,
+      "grad_norm": 0.3498349592633247,
+      "learning_rate": 1.0034801734220922e-06,
+      "loss": 0.035,
+      "step": 3874
+    },
+    {
+      "epoch": 4.352098834760635,
+      "grad_norm": 0.34975151345498384,
+      "learning_rate": 1.0000581135427067e-06,
+      "loss": 0.0365,
+      "step": 3875
+    },
+    {
+      "epoch": 4.353221957040573,
+      "grad_norm": 0.3369442861209052,
+      "learning_rate": 9.966415914266049e-07,
+      "loss": 0.0349,
+      "step": 3876
+    },
+    {
+      "epoch": 4.354345079320511,
+      "grad_norm": 0.36960971150127886,
+      "learning_rate": 9.93230609176008e-07,
+      "loss": 0.0359,
+      "step": 3877
+    },
+    {
+      "epoch": 4.355468201600449,
+      "grad_norm": 0.3499527159771044,
+      "learning_rate": 9.898251688897332e-07,
+      "loss": 0.0357,
+      "step": 3878
+    },
+    {
+      "epoch": 4.356591323880387,
+      "grad_norm": 0.3676928856512341,
+      "learning_rate": 9.86425272663185e-07,
+      "loss": 0.035,
+      "step": 3879
+    },
+    {
+      "epoch": 4.357714446160326,
+      "grad_norm": 0.35146864143042605,
+      "learning_rate": 9.830309225883562e-07,
+      "loss": 0.0348,
+      "step": 3880
+    },
+    {
+      "epoch": 4.358837568440264,
+      "grad_norm": 0.3506185390903599,
+      "learning_rate": 9.796421207538265e-07,
+      "loss": 0.0343,
+      "step": 3881
+    },
+    {
+      "epoch": 4.359960690720202,
+      "grad_norm": 0.35319022540725176,
+      "learning_rate": 9.762588692447661e-07,
+      "loss": 0.0341,
+      "step": 3882
+    },
+    {
+      "epoch": 4.361083813000141,
+      "grad_norm": 0.335262563237058,
+      "learning_rate": 9.728811701429242e-07,
+      "loss": 0.0345,
+      "step": 3883
+    },
+    {
+      "epoch": 4.362206935280079,
+      "grad_norm": 0.35087168008448233,
+      "learning_rate": 9.695090255266394e-07,
+      "loss": 0.0356,
+      "step": 3884
+    },
+    {
+      "epoch": 4.363330057560017,
+      "grad_norm": 0.32755003839475955,
+      "learning_rate": 9.66142437470825e-07,
+      "loss": 0.0306,
+      "step": 3885
+    },
+    {
+      "epoch": 4.3644531798399555,
+      "grad_norm": 0.36618851591535595,
+      "learning_rate": 9.627814080469822e-07,
+      "loss": 0.0382,
+      "step": 3886
+    },
+    {
+      "epoch": 4.3655763021198934,
+      "grad_norm": 0.3411679485731454,
+      "learning_rate": 9.594259393231897e-07,
+      "loss": 0.031,
+      "step": 3887
+    },
+    {
+      "epoch": 4.366699424399831,
+      "grad_norm": 0.3488732598923187,
+      "learning_rate": 9.56076033364105e-07,
+      "loss": 0.0348,
+      "step": 3888
+    },
+    {
+      "epoch": 4.36782254667977,
+      "grad_norm": 0.351431363191658,
+      "learning_rate": 9.527316922309593e-07,
+      "loss": 0.0358,
+      "step": 3889
+    },
+    {
+      "epoch": 4.368945668959708,
+      "grad_norm": 0.36029895486382335,
+      "learning_rate": 9.493929179815631e-07,
+      "loss": 0.0417,
+      "step": 3890
+    },
+    {
+      "epoch": 4.370068791239646,
+      "grad_norm": 0.33022530950278417,
+      "learning_rate": 9.46059712670303e-07,
+      "loss": 0.0335,
+      "step": 3891
+    },
+    {
+      "epoch": 4.371191913519585,
+      "grad_norm": 0.31602307458227624,
+      "learning_rate": 9.427320783481353e-07,
+      "loss": 0.0296,
+      "step": 3892
+    },
+    {
+      "epoch": 4.372315035799523,
+      "grad_norm": 0.357480507018708,
+      "learning_rate": 9.394100170625931e-07,
+      "loss": 0.0383,
+      "step": 3893
+    },
+    {
+      "epoch": 4.373438158079461,
+      "grad_norm": 0.35949006197745925,
+      "learning_rate": 9.360935308577723e-07,
+      "loss": 0.0391,
+      "step": 3894
+    },
+    {
+      "epoch": 4.374561280359399,
+      "grad_norm": 0.33189753767969865,
+      "learning_rate": 9.327826217743452e-07,
+      "loss": 0.0329,
+      "step": 3895
+    },
+    {
+      "epoch": 4.3756844026393376,
+      "grad_norm": 0.3508226977683308,
+      "learning_rate": 9.294772918495521e-07,
+      "loss": 0.0414,
+      "step": 3896
+    },
+    {
+      "epoch": 4.3768075249192755,
+      "grad_norm": 0.3643064354558861,
+      "learning_rate": 9.26177543117197e-07,
+      "loss": 0.0336,
+      "step": 3897
+    },
+    {
+      "epoch": 4.377930647199213,
+      "grad_norm": 0.3413054142419292,
+      "learning_rate": 9.228833776076551e-07,
+      "loss": 0.0329,
+      "step": 3898
+    },
+    {
+      "epoch": 4.379053769479152,
+      "grad_norm": 0.3346550439678259,
+      "learning_rate": 9.195947973478592e-07,
+      "loss": 0.033,
+      "step": 3899
+    },
+    {
+      "epoch": 4.38017689175909,
+      "grad_norm": 0.3401382520992714,
+      "learning_rate": 9.163118043613084e-07,
+      "loss": 0.0339,
+      "step": 3900
+    },
+    {
+      "epoch": 4.381300014039028,
+      "grad_norm": 0.37105306794446297,
+      "learning_rate": 9.130344006680658e-07,
+      "loss": 0.035,
+      "step": 3901
+    },
+    {
+      "epoch": 4.382423136318967,
+      "grad_norm": 0.35559365927896963,
+      "learning_rate": 9.09762588284755e-07,
+      "loss": 0.0308,
+      "step": 3902
+    },
+    {
+      "epoch": 4.383546258598905,
+      "grad_norm": 0.34778900056099654,
+      "learning_rate": 9.064963692245588e-07,
+      "loss": 0.0367,
+      "step": 3903
+    },
+    {
+      "epoch": 4.384669380878843,
+      "grad_norm": 0.3431968229053822,
+      "learning_rate": 9.03235745497213e-07,
+      "loss": 0.0327,
+      "step": 3904
+    },
+    {
+      "epoch": 4.385792503158782,
+      "grad_norm": 0.32014372957978515,
+      "learning_rate": 8.999807191090193e-07,
+      "loss": 0.0309,
+      "step": 3905
+    },
+    {
+      "epoch": 4.38691562543872,
+      "grad_norm": 0.3299346644592865,
+      "learning_rate": 8.967312920628312e-07,
+      "loss": 0.0351,
+      "step": 3906
+    },
+    {
+      "epoch": 4.3880387477186575,
+      "grad_norm": 0.34173628201013206,
+      "learning_rate": 8.934874663580551e-07,
+      "loss": 0.0337,
+      "step": 3907
+    },
+    {
+      "epoch": 4.389161869998596,
+      "grad_norm": 0.34178452730956427,
+      "learning_rate": 8.902492439906552e-07,
+      "loss": 0.031,
+      "step": 3908
+    },
+    {
+      "epoch": 4.390284992278534,
+      "grad_norm": 0.35561085437557965,
+      "learning_rate": 8.870166269531421e-07,
+      "loss": 0.0366,
+      "step": 3909
+    },
+    {
+      "epoch": 4.391408114558472,
+      "grad_norm": 0.3607348657261128,
+      "learning_rate": 8.837896172345827e-07,
+      "loss": 0.0338,
+      "step": 3910
+    },
+    {
+      "epoch": 4.392531236838411,
+      "grad_norm": 0.3535133481860961,
+      "learning_rate": 8.805682168205909e-07,
+      "loss": 0.0336,
+      "step": 3911
+    },
+    {
+      "epoch": 4.393654359118349,
+      "grad_norm": 0.33450525221475136,
+      "learning_rate": 8.773524276933299e-07,
+      "loss": 0.0321,
+      "step": 3912
+    },
+    {
+      "epoch": 4.394777481398287,
+      "grad_norm": 0.33740549152276617,
+      "learning_rate": 8.741422518315113e-07,
+      "loss": 0.0322,
+      "step": 3913
+    },
+    {
+      "epoch": 4.395900603678226,
+      "grad_norm": 0.34668595275206976,
+      "learning_rate": 8.709376912103895e-07,
+      "loss": 0.0343,
+      "step": 3914
+    },
+    {
+      "epoch": 4.397023725958164,
+      "grad_norm": 0.3720699327652378,
+      "learning_rate": 8.677387478017673e-07,
+      "loss": 0.0383,
+      "step": 3915
+    },
+    {
+      "epoch": 4.398146848238102,
+      "grad_norm": 0.3310372019137898,
+      "learning_rate": 8.645454235739903e-07,
+      "loss": 0.0309,
+      "step": 3916
+    },
+    {
+      "epoch": 4.3992699705180405,
+      "grad_norm": 0.3446218941497761,
+      "learning_rate": 8.613577204919455e-07,
+      "loss": 0.0336,
+      "step": 3917
+    },
+    {
+      "epoch": 4.400393092797978,
+      "grad_norm": 0.3321270678180654,
+      "learning_rate": 8.581756405170627e-07,
+      "loss": 0.0326,
+      "step": 3918
+    },
+    {
+      "epoch": 4.401516215077916,
+      "grad_norm": 0.3429813545514485,
+      "learning_rate": 8.54999185607307e-07,
+      "loss": 0.0337,
+      "step": 3919
+    },
+    {
+      "epoch": 4.402639337357855,
+      "grad_norm": 0.329083241283035,
+      "learning_rate": 8.518283577171894e-07,
+      "loss": 0.032,
+      "step": 3920
+    },
+    {
+      "epoch": 4.403762459637793,
+      "grad_norm": 0.35028547291655737,
+      "learning_rate": 8.486631587977545e-07,
+      "loss": 0.0331,
+      "step": 3921
+    },
+    {
+      "epoch": 4.404885581917731,
+      "grad_norm": 0.36812246717749864,
+      "learning_rate": 8.455035907965837e-07,
+      "loss": 0.0322,
+      "step": 3922
+    },
+    {
+      "epoch": 4.40600870419767,
+      "grad_norm": 0.3354095575220094,
+      "learning_rate": 8.423496556577959e-07,
+      "loss": 0.0315,
+      "step": 3923
+    },
+    {
+      "epoch": 4.407131826477608,
+      "grad_norm": 0.36423209726760586,
+      "learning_rate": 8.392013553220391e-07,
+      "loss": 0.0363,
+      "step": 3924
+    },
+    {
+      "epoch": 4.408254948757546,
+      "grad_norm": 0.33578769827695937,
+      "learning_rate": 8.360586917264979e-07,
+      "loss": 0.0327,
+      "step": 3925
+    },
+    {
+      "epoch": 4.409378071037485,
+      "grad_norm": 0.35083436656255,
+      "learning_rate": 8.329216668048878e-07,
+      "loss": 0.0337,
+      "step": 3926
+    },
+    {
+      "epoch": 4.4105011933174225,
+      "grad_norm": 0.352145989291215,
+      "learning_rate": 8.297902824874582e-07,
+      "loss": 0.0362,
+      "step": 3927
+    },
+    {
+      "epoch": 4.4116243155973605,
+      "grad_norm": 0.35269237221340394,
+      "learning_rate": 8.266645407009788e-07,
+      "loss": 0.0326,
+      "step": 3928
+    },
+    {
+      "epoch": 4.412747437877299,
+      "grad_norm": 0.34801427609667296,
+      "learning_rate": 8.235444433687556e-07,
+      "loss": 0.0366,
+      "step": 3929
+    },
+    {
+      "epoch": 4.413870560157237,
+      "grad_norm": 0.3399050508413452,
+      "learning_rate": 8.204299924106196e-07,
+      "loss": 0.0347,
+      "step": 3930
+    },
+    {
+      "epoch": 4.414993682437175,
+      "grad_norm": 0.3487557779248968,
+      "learning_rate": 8.173211897429245e-07,
+      "loss": 0.0342,
+      "step": 3931
+    },
+    {
+      "epoch": 4.416116804717114,
+      "grad_norm": 0.3667385663949515,
+      "learning_rate": 8.142180372785547e-07,
+      "loss": 0.0359,
+      "step": 3932
+    },
+    {
+      "epoch": 4.417239926997052,
+      "grad_norm": 0.3333295175583706,
+      "learning_rate": 8.111205369269104e-07,
+      "loss": 0.0308,
+      "step": 3933
+    },
+    {
+      "epoch": 4.41836304927699,
+      "grad_norm": 0.33057100673502793,
+      "learning_rate": 8.080286905939172e-07,
+      "loss": 0.0322,
+      "step": 3934
+    },
+    {
+      "epoch": 4.419486171556928,
+      "grad_norm": 0.3420534039639189,
+      "learning_rate": 8.049425001820255e-07,
+      "loss": 0.0374,
+      "step": 3935
+    },
+    {
+      "epoch": 4.420609293836867,
+      "grad_norm": 0.33462359777080614,
+      "learning_rate": 8.018619675901995e-07,
+      "loss": 0.034,
+      "step": 3936
+    },
+    {
+      "epoch": 4.421732416116805,
+      "grad_norm": 0.3634343403035784,
+      "learning_rate": 7.987870947139276e-07,
+      "loss": 0.032,
+      "step": 3937
+    },
+    {
+      "epoch": 4.4228555383967425,
+      "grad_norm": 0.3732288458675281,
+      "learning_rate": 7.957178834452095e-07,
+      "loss": 0.036,
+      "step": 3938
+    },
+    {
+      "epoch": 4.423978660676681,
+      "grad_norm": 0.3481104484204922,
+      "learning_rate": 7.926543356725658e-07,
+      "loss": 0.0315,
+      "step": 3939
+    },
+    {
+      "epoch": 4.425101782956619,
+      "grad_norm": 0.3400656410063691,
+      "learning_rate": 7.895964532810318e-07,
+      "loss": 0.0331,
+      "step": 3940
+    },
+    {
+      "epoch": 4.426224905236557,
+      "grad_norm": 0.32595754035332963,
+      "learning_rate": 7.86544238152157e-07,
+      "loss": 0.0327,
+      "step": 3941
+    },
+    {
+      "epoch": 4.427348027516496,
+      "grad_norm": 0.35250908793584956,
+      "learning_rate": 7.834976921640025e-07,
+      "loss": 0.0357,
+      "step": 3942
+    },
+    {
+      "epoch": 4.428471149796434,
+      "grad_norm": 0.3520988496330288,
+      "learning_rate": 7.804568171911398e-07,
+      "loss": 0.0379,
+      "step": 3943
+    },
+    {
+      "epoch": 4.429594272076372,
+      "grad_norm": 0.36252580809370655,
+      "learning_rate": 7.774216151046543e-07,
+      "loss": 0.034,
+      "step": 3944
+    },
+    {
+      "epoch": 4.430717394356311,
+      "grad_norm": 0.34372024882244934,
+      "learning_rate": 7.743920877721378e-07,
+      "loss": 0.0337,
+      "step": 3945
+    },
+    {
+      "epoch": 4.431840516636249,
+      "grad_norm": 0.3573305329011033,
+      "learning_rate": 7.713682370576947e-07,
+      "loss": 0.034,
+      "step": 3946
+    },
+    {
+      "epoch": 4.432963638916187,
+      "grad_norm": 0.35271907669417196,
+      "learning_rate": 7.683500648219322e-07,
+      "loss": 0.0366,
+      "step": 3947
+    },
+    {
+      "epoch": 4.434086761196125,
+      "grad_norm": 0.3387988216872993,
+      "learning_rate": 7.653375729219636e-07,
+      "loss": 0.0327,
+      "step": 3948
+    },
+    {
+      "epoch": 4.435209883476063,
+      "grad_norm": 0.34034343693680486,
+      "learning_rate": 7.623307632114085e-07,
+      "loss": 0.0345,
+      "step": 3949
+    },
+    {
+      "epoch": 4.436333005756001,
+      "grad_norm": 0.3391577418176705,
+      "learning_rate": 7.593296375403914e-07,
+      "loss": 0.034,
+      "step": 3950
+    },
+    {
+      "epoch": 4.43745612803594,
+      "grad_norm": 0.34486562074073207,
+      "learning_rate": 7.563341977555372e-07,
+      "loss": 0.0345,
+      "step": 3951
+    },
+    {
+      "epoch": 4.438579250315878,
+      "grad_norm": 0.3412224433899792,
+      "learning_rate": 7.533444456999728e-07,
+      "loss": 0.0337,
+      "step": 3952
+    },
+    {
+      "epoch": 4.439702372595816,
+      "grad_norm": 0.3572984260577703,
+      "learning_rate": 7.503603832133277e-07,
+      "loss": 0.0362,
+      "step": 3953
+    },
+    {
+      "epoch": 4.440825494875755,
+      "grad_norm": 0.3378929389056555,
+      "learning_rate": 7.473820121317243e-07,
+      "loss": 0.0335,
+      "step": 3954
+    },
+    {
+      "epoch": 4.441948617155693,
+      "grad_norm": 0.3586567601428513,
+      "learning_rate": 7.4440933428779e-07,
+      "loss": 0.0357,
+      "step": 3955
+    },
+    {
+      "epoch": 4.443071739435631,
+      "grad_norm": 0.35329443808804784,
+      "learning_rate": 7.41442351510645e-07,
+      "loss": 0.0317,
+      "step": 3956
+    },
+    {
+      "epoch": 4.4441948617155695,
+      "grad_norm": 0.3554120702375195,
+      "learning_rate": 7.384810656259078e-07,
+      "loss": 0.0351,
+      "step": 3957
+    },
+    {
+      "epoch": 4.4453179839955075,
+      "grad_norm": 0.32801148615792985,
+      "learning_rate": 7.355254784556887e-07,
+      "loss": 0.0323,
+      "step": 3958
+    },
+    {
+      "epoch": 4.446441106275445,
+      "grad_norm": 0.35260937314651863,
+      "learning_rate": 7.325755918185928e-07,
+      "loss": 0.0361,
+      "step": 3959
+    },
+    {
+      "epoch": 4.447564228555384,
+      "grad_norm": 0.35121127734872754,
+      "learning_rate": 7.296314075297196e-07,
+      "loss": 0.0338,
+      "step": 3960
+    },
+    {
+      "epoch": 4.448687350835322,
+      "grad_norm": 0.3538367483299584,
+      "learning_rate": 7.266929274006595e-07,
+      "loss": 0.0363,
+      "step": 3961
+    },
+    {
+      "epoch": 4.44981047311526,
+      "grad_norm": 0.3518231420814252,
+      "learning_rate": 7.237601532394866e-07,
+      "loss": 0.0393,
+      "step": 3962
+    },
+    {
+      "epoch": 4.450933595395199,
+      "grad_norm": 0.35867753461888296,
+      "learning_rate": 7.208330868507718e-07,
+      "loss": 0.0355,
+      "step": 3963
+    },
+    {
+      "epoch": 4.452056717675137,
+      "grad_norm": 0.3399901393948421,
+      "learning_rate": 7.17911730035572e-07,
+      "loss": 0.0323,
+      "step": 3964
+    },
+    {
+      "epoch": 4.453179839955075,
+      "grad_norm": 0.36423866949072375,
+      "learning_rate": 7.149960845914294e-07,
+      "loss": 0.0374,
+      "step": 3965
+    },
+    {
+      "epoch": 4.454302962235014,
+      "grad_norm": 0.3566191322953024,
+      "learning_rate": 7.120861523123735e-07,
+      "loss": 0.0362,
+      "step": 3966
+    },
+    {
+      "epoch": 4.455426084514952,
+      "grad_norm": 0.35784219880826335,
+      "learning_rate": 7.091819349889162e-07,
+      "loss": 0.0362,
+      "step": 3967
+    },
+    {
+      "epoch": 4.4565492067948895,
+      "grad_norm": 0.3382634880819678,
+      "learning_rate": 7.062834344080549e-07,
+      "loss": 0.0315,
+      "step": 3968
+    },
+    {
+      "epoch": 4.457672329074828,
+      "grad_norm": 0.3569943519463488,
+      "learning_rate": 7.033906523532697e-07,
+      "loss": 0.0349,
+      "step": 3969
+    },
+    {
+      "epoch": 4.458795451354766,
+      "grad_norm": 0.3650433682446708,
+      "learning_rate": 7.005035906045199e-07,
+      "loss": 0.0346,
+      "step": 3970
+    },
+    {
+      "epoch": 4.459918573634704,
+      "grad_norm": 0.33813321468776913,
+      "learning_rate": 6.976222509382491e-07,
+      "loss": 0.0317,
+      "step": 3971
+    },
+    {
+      "epoch": 4.461041695914643,
+      "grad_norm": 0.3619159838292503,
+      "learning_rate": 6.947466351273735e-07,
+      "loss": 0.0359,
+      "step": 3972
+    },
+    {
+      "epoch": 4.462164818194581,
+      "grad_norm": 0.3763835667376468,
+      "learning_rate": 6.918767449412933e-07,
+      "loss": 0.0361,
+      "step": 3973
+    },
+    {
+      "epoch": 4.463287940474519,
+      "grad_norm": 0.3436876392723535,
+      "learning_rate": 6.890125821458826e-07,
+      "loss": 0.0336,
+      "step": 3974
+    },
+    {
+      "epoch": 4.464411062754458,
+      "grad_norm": 0.32509347109822845,
+      "learning_rate": 6.86154148503494e-07,
+      "loss": 0.0308,
+      "step": 3975
+    },
+    {
+      "epoch": 4.465534185034396,
+      "grad_norm": 0.3525907199308526,
+      "learning_rate": 6.833014457729525e-07,
+      "loss": 0.0356,
+      "step": 3976
+    },
+    {
+      "epoch": 4.466657307314334,
+      "grad_norm": 0.35208517679189183,
+      "learning_rate": 6.804544757095566e-07,
+      "loss": 0.0328,
+      "step": 3977
+    },
+    {
+      "epoch": 4.4677804295942725,
+      "grad_norm": 0.3599858916473201,
+      "learning_rate": 6.776132400650781e-07,
+      "loss": 0.0352,
+      "step": 3978
+    },
+    {
+      "epoch": 4.46890355187421,
+      "grad_norm": 0.3641931626113949,
+      "learning_rate": 6.747777405877609e-07,
+      "loss": 0.0393,
+      "step": 3979
+    },
+    {
+      "epoch": 4.470026674154148,
+      "grad_norm": 0.34746423533313237,
+      "learning_rate": 6.719479790223204e-07,
+      "loss": 0.0364,
+      "step": 3980
+    },
+    {
+      "epoch": 4.471149796434087,
+      "grad_norm": 0.3564297638222409,
+      "learning_rate": 6.691239571099395e-07,
+      "loss": 0.0365,
+      "step": 3981
+    },
+    {
+      "epoch": 4.472272918714025,
+      "grad_norm": 0.34573485421602407,
+      "learning_rate": 6.663056765882692e-07,
+      "loss": 0.0329,
+      "step": 3982
+    },
+    {
+      "epoch": 4.473396040993963,
+      "grad_norm": 0.3597984417319236,
+      "learning_rate": 6.634931391914278e-07,
+      "loss": 0.037,
+      "step": 3983
+    },
+    {
+      "epoch": 4.474519163273902,
+      "grad_norm": 0.3378924186798682,
+      "learning_rate": 6.606863466500013e-07,
+      "loss": 0.0343,
+      "step": 3984
+    },
+    {
+      "epoch": 4.47564228555384,
+      "grad_norm": 0.34863970957372714,
+      "learning_rate": 6.578853006910402e-07,
+      "loss": 0.0343,
+      "step": 3985
+    },
+    {
+      "epoch": 4.476765407833778,
+      "grad_norm": 0.33950262649507096,
+      "learning_rate": 6.550900030380614e-07,
+      "loss": 0.0378,
+      "step": 3986
+    },
+    {
+      "epoch": 4.477888530113717,
+      "grad_norm": 0.35574459663793684,
+      "learning_rate": 6.523004554110379e-07,
+      "loss": 0.0351,
+      "step": 3987
+    },
+    {
+      "epoch": 4.4790116523936545,
+      "grad_norm": 0.32764803743478366,
+      "learning_rate": 6.495166595264102e-07,
+      "loss": 0.0318,
+      "step": 3988
+    },
+    {
+      "epoch": 4.4801347746735924,
+      "grad_norm": 0.3648664364754083,
+      "learning_rate": 6.467386170970802e-07,
+      "loss": 0.0359,
+      "step": 3989
+    },
+    {
+      "epoch": 4.481257896953531,
+      "grad_norm": 0.33735699056875545,
+      "learning_rate": 6.439663298324061e-07,
+      "loss": 0.0307,
+      "step": 3990
+    },
+    {
+      "epoch": 4.482381019233469,
+      "grad_norm": 0.36038648251848654,
+      "learning_rate": 6.411997994382102e-07,
+      "loss": 0.0366,
+      "step": 3991
+    },
+    {
+      "epoch": 4.483504141513407,
+      "grad_norm": 0.3582401185410582,
+      "learning_rate": 6.384390276167651e-07,
+      "loss": 0.0373,
+      "step": 3992
+    },
+    {
+      "epoch": 4.484627263793345,
+      "grad_norm": 0.3514649132145566,
+      "learning_rate": 6.356840160668054e-07,
+      "loss": 0.0351,
+      "step": 3993
+    },
+    {
+      "epoch": 4.485750386073284,
+      "grad_norm": 0.3316545137948871,
+      "learning_rate": 6.329347664835206e-07,
+      "loss": 0.0335,
+      "step": 3994
+    },
+    {
+      "epoch": 4.486873508353222,
+      "grad_norm": 0.38390972496268505,
+      "learning_rate": 6.30191280558553e-07,
+      "loss": 0.033,
+      "step": 3995
+    },
+    {
+      "epoch": 4.48799663063316,
+      "grad_norm": 0.34745468335455126,
+      "learning_rate": 6.274535599800014e-07,
+      "loss": 0.0339,
+      "step": 3996
+    },
+    {
+      "epoch": 4.489119752913099,
+      "grad_norm": 0.3469881466439082,
+      "learning_rate": 6.247216064324158e-07,
+      "loss": 0.0348,
+      "step": 3997
+    },
+    {
+      "epoch": 4.4902428751930366,
+      "grad_norm": 0.33480434506463175,
+      "learning_rate": 6.219954215967949e-07,
+      "loss": 0.0324,
+      "step": 3998
+    },
+    {
+      "epoch": 4.4913659974729745,
+      "grad_norm": 0.33430612710636604,
+      "learning_rate": 6.192750071505904e-07,
+      "loss": 0.0327,
+      "step": 3999
+    },
+    {
+      "epoch": 4.492489119752913,
+      "grad_norm": 0.37230150478039675,
+      "learning_rate": 6.165603647677054e-07,
+      "loss": 0.0343,
+      "step": 4000
+    },
+    {
+      "epoch": 4.493612242032851,
+      "grad_norm": 0.3715774781533714,
+      "learning_rate": 6.138514961184872e-07,
+      "loss": 0.0388,
+      "step": 4001
+    },
+    {
+      "epoch": 4.494735364312789,
+      "grad_norm": 0.34588801151813375,
+      "learning_rate": 6.111484028697334e-07,
+      "loss": 0.0357,
+      "step": 4002
+    },
+    {
+      "epoch": 4.495858486592728,
+      "grad_norm": 0.3473978219479411,
+      "learning_rate": 6.084510866846882e-07,
+      "loss": 0.0337,
+      "step": 4003
+    },
+    {
+      "epoch": 4.496981608872666,
+      "grad_norm": 0.3558220793979365,
+      "learning_rate": 6.057595492230372e-07,
+      "loss": 0.035,
+      "step": 4004
+    },
+    {
+      "epoch": 4.498104731152604,
+      "grad_norm": 0.3617212667424299,
+      "learning_rate": 6.030737921409169e-07,
+      "loss": 0.0366,
+      "step": 4005
+    },
+    {
+      "epoch": 4.499227853432543,
+      "grad_norm": 0.372375829726114,
+      "learning_rate": 6.003938170908985e-07,
+      "loss": 0.0372,
+      "step": 4006
+    },
+    {
+      "epoch": 4.500350975712481,
+      "grad_norm": 0.33813041058043286,
+      "learning_rate": 5.97719625722003e-07,
+      "loss": 0.0339,
+      "step": 4007
+    },
+    {
+      "epoch": 4.501474097992419,
+      "grad_norm": 0.3336235673204082,
+      "learning_rate": 5.950512196796898e-07,
+      "loss": 0.0337,
+      "step": 4008
+    },
+    {
+      "epoch": 4.502597220272357,
+      "grad_norm": 0.3582259509306251,
+      "learning_rate": 5.923886006058566e-07,
+      "loss": 0.035,
+      "step": 4009
+    },
+    {
+      "epoch": 4.503720342552295,
+      "grad_norm": 0.34292021535584205,
+      "learning_rate": 5.897317701388461e-07,
+      "loss": 0.0325,
+      "step": 4010
+    },
+    {
+      "epoch": 4.504843464832233,
+      "grad_norm": 0.3345731631163473,
+      "learning_rate": 5.870807299134307e-07,
+      "loss": 0.0337,
+      "step": 4011
+    },
+    {
+      "epoch": 4.505966587112172,
+      "grad_norm": 0.32035494944471327,
+      "learning_rate": 5.844354815608267e-07,
+      "loss": 0.0304,
+      "step": 4012
+    },
+    {
+      "epoch": 4.50708970939211,
+      "grad_norm": 0.3447654015600138,
+      "learning_rate": 5.817960267086853e-07,
+      "loss": 0.0329,
+      "step": 4013
+    },
+    {
+      "epoch": 4.508212831672048,
+      "grad_norm": 0.34544101773876196,
+      "learning_rate": 5.791623669810908e-07,
+      "loss": 0.0339,
+      "step": 4014
+    },
+    {
+      "epoch": 4.509335953951987,
+      "grad_norm": 0.3728532295897952,
+      "learning_rate": 5.765345039985648e-07,
+      "loss": 0.0343,
+      "step": 4015
+    },
+    {
+      "epoch": 4.510459076231925,
+      "grad_norm": 0.35317009146829786,
+      "learning_rate": 5.739124393780571e-07,
+      "loss": 0.0335,
+      "step": 4016
+    },
+    {
+      "epoch": 4.511582198511863,
+      "grad_norm": 0.3545765005968472,
+      "learning_rate": 5.71296174732956e-07,
+      "loss": 0.0345,
+      "step": 4017
+    },
+    {
+      "epoch": 4.5127053207918015,
+      "grad_norm": 0.342843964056811,
+      "learning_rate": 5.68685711673076e-07,
+      "loss": 0.0323,
+      "step": 4018
+    },
+    {
+      "epoch": 4.5138284430717395,
+      "grad_norm": 0.3639774486483642,
+      "learning_rate": 5.660810518046644e-07,
+      "loss": 0.0355,
+      "step": 4019
+    },
+    {
+      "epoch": 4.514951565351677,
+      "grad_norm": 0.3332083193281611,
+      "learning_rate": 5.634821967303994e-07,
+      "loss": 0.0337,
+      "step": 4020
+    },
+    {
+      "epoch": 4.516074687631616,
+      "grad_norm": 0.34776766829793077,
+      "learning_rate": 5.608891480493816e-07,
+      "loss": 0.0351,
+      "step": 4021
+    },
+    {
+      "epoch": 4.517197809911554,
+      "grad_norm": 0.3483906727553061,
+      "learning_rate": 5.583019073571427e-07,
+      "loss": 0.034,
+      "step": 4022
+    },
+    {
+      "epoch": 4.518320932191492,
+      "grad_norm": 0.35109229875700276,
+      "learning_rate": 5.557204762456425e-07,
+      "loss": 0.0326,
+      "step": 4023
+    },
+    {
+      "epoch": 4.519444054471431,
+      "grad_norm": 0.32870715662554173,
+      "learning_rate": 5.531448563032626e-07,
+      "loss": 0.035,
+      "step": 4024
+    },
+    {
+      "epoch": 4.520567176751369,
+      "grad_norm": 0.35831485527315554,
+      "learning_rate": 5.505750491148121e-07,
+      "loss": 0.0371,
+      "step": 4025
+    },
+    {
+      "epoch": 4.521690299031307,
+      "grad_norm": 0.3717078413941628,
+      "learning_rate": 5.480110562615182e-07,
+      "loss": 0.0401,
+      "step": 4026
+    },
+    {
+      "epoch": 4.522813421311246,
+      "grad_norm": 0.34683971898871724,
+      "learning_rate": 5.454528793210356e-07,
+      "loss": 0.0308,
+      "step": 4027
+    },
+    {
+      "epoch": 4.523936543591184,
+      "grad_norm": 0.33865977054207247,
+      "learning_rate": 5.429005198674398e-07,
+      "loss": 0.0334,
+      "step": 4028
+    },
+    {
+      "epoch": 4.5250596658711215,
+      "grad_norm": 0.35261816775158816,
+      "learning_rate": 5.403539794712243e-07,
+      "loss": 0.036,
+      "step": 4029
+    },
+    {
+      "epoch": 4.5261827881510595,
+      "grad_norm": 0.34517842856104286,
+      "learning_rate": 5.378132596993047e-07,
+      "loss": 0.0375,
+      "step": 4030
+    },
+    {
+      "epoch": 4.527305910430998,
+      "grad_norm": 0.35750348534406,
+      "learning_rate": 5.352783621150126e-07,
+      "loss": 0.0355,
+      "step": 4031
+    },
+    {
+      "epoch": 4.528429032710936,
+      "grad_norm": 0.3454480091845214,
+      "learning_rate": 5.327492882780993e-07,
+      "loss": 0.0356,
+      "step": 4032
+    },
+    {
+      "epoch": 4.529552154990874,
+      "grad_norm": 0.32573268708480096,
+      "learning_rate": 5.3022603974473e-07,
+      "loss": 0.0308,
+      "step": 4033
+    },
+    {
+      "epoch": 4.530675277270813,
+      "grad_norm": 0.34576124173105804,
+      "learning_rate": 5.277086180674906e-07,
+      "loss": 0.0336,
+      "step": 4034
+    },
+    {
+      "epoch": 4.531798399550751,
+      "grad_norm": 0.34623159718083096,
+      "learning_rate": 5.251970247953752e-07,
+      "loss": 0.036,
+      "step": 4035
+    },
+    {
+      "epoch": 4.532921521830689,
+      "grad_norm": 0.35168976903132443,
+      "learning_rate": 5.226912614737956e-07,
+      "loss": 0.0326,
+      "step": 4036
+    },
+    {
+      "epoch": 4.534044644110628,
+      "grad_norm": 0.5382472283002754,
+      "learning_rate": 5.20191329644577e-07,
+      "loss": 0.0353,
+      "step": 4037
+    },
+    {
+      "epoch": 4.535167766390566,
+      "grad_norm": 0.37754077895496924,
+      "learning_rate": 5.176972308459527e-07,
+      "loss": 0.0365,
+      "step": 4038
+    },
+    {
+      "epoch": 4.536290888670504,
+      "grad_norm": 0.37416484246623505,
+      "learning_rate": 5.152089666125704e-07,
+      "loss": 0.0394,
+      "step": 4039
+    },
+    {
+      "epoch": 4.537414010950442,
+      "grad_norm": 0.3590048250258893,
+      "learning_rate": 5.127265384754865e-07,
+      "loss": 0.0366,
+      "step": 4040
+    },
+    {
+      "epoch": 4.53853713323038,
+      "grad_norm": 0.3342358936092815,
+      "learning_rate": 5.102499479621658e-07,
+      "loss": 0.0335,
+      "step": 4041
+    },
+    {
+      "epoch": 4.539660255510318,
+      "grad_norm": 0.3493925300990296,
+      "learning_rate": 5.07779196596484e-07,
+      "loss": 0.0339,
+      "step": 4042
+    },
+    {
+      "epoch": 4.540783377790257,
+      "grad_norm": 0.33636877105880414,
+      "learning_rate": 5.053142858987192e-07,
+      "loss": 0.0317,
+      "step": 4043
+    },
+    {
+      "epoch": 4.541906500070195,
+      "grad_norm": 0.34263176501366066,
+      "learning_rate": 5.028552173855572e-07,
+      "loss": 0.0336,
+      "step": 4044
+    },
+    {
+      "epoch": 4.543029622350133,
+      "grad_norm": 0.37257701817237154,
+      "learning_rate": 5.004019925700921e-07,
+      "loss": 0.0392,
+      "step": 4045
+    },
+    {
+      "epoch": 4.544152744630072,
+      "grad_norm": 0.3369340532849318,
+      "learning_rate": 4.979546129618184e-07,
+      "loss": 0.0325,
+      "step": 4046
+    },
+    {
+      "epoch": 4.54527586691001,
+      "grad_norm": 0.33726138907008607,
+      "learning_rate": 4.955130800666374e-07,
+      "loss": 0.0329,
+      "step": 4047
+    },
+    {
+      "epoch": 4.546398989189948,
+      "grad_norm": 0.38713309676317126,
+      "learning_rate": 4.930773953868506e-07,
+      "loss": 0.0378,
+      "step": 4048
+    },
+    {
+      "epoch": 4.5475221114698865,
+      "grad_norm": 0.36438821244482394,
+      "learning_rate": 4.906475604211624e-07,
+      "loss": 0.0341,
+      "step": 4049
+    },
+    {
+      "epoch": 4.548645233749824,
+      "grad_norm": 0.33904817801576353,
+      "learning_rate": 4.882235766646748e-07,
+      "loss": 0.0356,
+      "step": 4050
+    },
+    {
+      "epoch": 4.549768356029762,
+      "grad_norm": 0.35475025219773854,
+      "learning_rate": 4.858054456088923e-07,
+      "loss": 0.0338,
+      "step": 4051
+    },
+    {
+      "epoch": 4.550891478309701,
+      "grad_norm": 0.3458201973740821,
+      "learning_rate": 4.833931687417182e-07,
+      "loss": 0.0345,
+      "step": 4052
+    },
+    {
+      "epoch": 4.552014600589639,
+      "grad_norm": 0.36111440231472997,
+      "learning_rate": 4.809867475474539e-07,
+      "loss": 0.0355,
+      "step": 4053
+    },
+    {
+      "epoch": 4.553137722869577,
+      "grad_norm": 0.3535234629211531,
+      "learning_rate": 4.785861835067962e-07,
+      "loss": 0.0343,
+      "step": 4054
+    },
+    {
+      "epoch": 4.554260845149516,
+      "grad_norm": 0.3528696097512272,
+      "learning_rate": 4.761914780968369e-07,
+      "loss": 0.0333,
+      "step": 4055
+    },
+    {
+      "epoch": 4.555383967429454,
+      "grad_norm": 0.34838927834260514,
+      "learning_rate": 4.738026327910661e-07,
+      "loss": 0.0358,
+      "step": 4056
+    },
+    {
+      "epoch": 4.556507089709392,
+      "grad_norm": 0.3431964796657279,
+      "learning_rate": 4.7141964905936697e-07,
+      "loss": 0.0338,
+      "step": 4057
+    },
+    {
+      "epoch": 4.557630211989331,
+      "grad_norm": 0.349604850461702,
+      "learning_rate": 4.6904252836801446e-07,
+      "loss": 0.0351,
+      "step": 4058
+    },
+    {
+      "epoch": 4.5587533342692685,
+      "grad_norm": 0.3284644735862563,
+      "learning_rate": 4.6667127217967844e-07,
+      "loss": 0.0312,
+      "step": 4059
+    },
+    {
+      "epoch": 4.5598764565492065,
+      "grad_norm": 0.3381032631698376,
+      "learning_rate": 4.6430588195341853e-07,
+      "loss": 0.0347,
+      "step": 4060
+    },
+    {
+      "epoch": 4.560999578829145,
+      "grad_norm": 0.3259962293243492,
+      "learning_rate": 4.6194635914468377e-07,
+      "loss": 0.0321,
+      "step": 4061
+    },
+    {
+      "epoch": 4.562122701109083,
+      "grad_norm": 0.34016741280577595,
+      "learning_rate": 4.595927052053162e-07,
+      "loss": 0.0317,
+      "step": 4062
+    },
+    {
+      "epoch": 4.563245823389021,
+      "grad_norm": 0.3783286672968007,
+      "learning_rate": 4.5724492158354397e-07,
+      "loss": 0.0346,
+      "step": 4063
+    },
+    {
+      "epoch": 4.56436894566896,
+      "grad_norm": 0.3627087896517975,
+      "learning_rate": 4.5490300972398705e-07,
+      "loss": 0.0337,
+      "step": 4064
+    },
+    {
+      "epoch": 4.565492067948898,
+      "grad_norm": 0.362771085697257,
+      "learning_rate": 4.52566971067645e-07,
+      "loss": 0.0353,
+      "step": 4065
+    },
+    {
+      "epoch": 4.566615190228836,
+      "grad_norm": 0.3102609073574198,
+      "learning_rate": 4.502368070519114e-07,
+      "loss": 0.0274,
+      "step": 4066
+    },
+    {
+      "epoch": 4.567738312508775,
+      "grad_norm": 0.33733629821182154,
+      "learning_rate": 4.4791251911056043e-07,
+      "loss": 0.0344,
+      "step": 4067
+    },
+    {
+      "epoch": 4.568861434788713,
+      "grad_norm": 0.34048282314449596,
+      "learning_rate": 4.4559410867375365e-07,
+      "loss": 0.0378,
+      "step": 4068
+    },
+    {
+      "epoch": 4.569984557068651,
+      "grad_norm": 0.34656193172912014,
+      "learning_rate": 4.432815771680321e-07,
+      "loss": 0.0346,
+      "step": 4069
+    },
+    {
+      "epoch": 4.571107679348589,
+      "grad_norm": 0.32758773549200243,
+      "learning_rate": 4.409749260163232e-07,
+      "loss": 0.0345,
+      "step": 4070
+    },
+    {
+      "epoch": 4.572230801628527,
+      "grad_norm": 0.3654433415452595,
+      "learning_rate": 4.386741566379338e-07,
+      "loss": 0.0387,
+      "step": 4071
+    },
+    {
+      "epoch": 4.573353923908465,
+      "grad_norm": 0.3367092688064399,
+      "learning_rate": 4.3637927044855476e-07,
+      "loss": 0.0324,
+      "step": 4072
+    },
+    {
+      "epoch": 4.574477046188404,
+      "grad_norm": 0.3640805465047991,
+      "learning_rate": 4.340902688602544e-07,
+      "loss": 0.0339,
+      "step": 4073
+    },
+    {
+      "epoch": 4.575600168468342,
+      "grad_norm": 0.34452856807576676,
+      "learning_rate": 4.3180715328147826e-07,
+      "loss": 0.034,
+      "step": 4074
+    },
+    {
+      "epoch": 4.57672329074828,
+      "grad_norm": 0.34328210378409857,
+      "learning_rate": 4.295299251170537e-07,
+      "loss": 0.0327,
+      "step": 4075
+    },
+    {
+      "epoch": 4.577846413028219,
+      "grad_norm": 0.35304122276887184,
+      "learning_rate": 4.272585857681844e-07,
+      "loss": 0.0344,
+      "step": 4076
+    },
+    {
+      "epoch": 4.578969535308157,
+      "grad_norm": 0.3253336789166046,
+      "learning_rate": 4.249931366324511e-07,
+      "loss": 0.0331,
+      "step": 4077
+    },
+    {
+      "epoch": 4.580092657588095,
+      "grad_norm": 0.3388987083083182,
+      "learning_rate": 4.2273357910380896e-07,
+      "loss": 0.0343,
+      "step": 4078
+    },
+    {
+      "epoch": 4.5812157798680335,
+      "grad_norm": 0.3543772419693709,
+      "learning_rate": 4.2047991457258905e-07,
+      "loss": 0.0374,
+      "step": 4079
+    },
+    {
+      "epoch": 4.5823389021479715,
+      "grad_norm": 0.3765630360863373,
+      "learning_rate": 4.182321444254944e-07,
+      "loss": 0.0354,
+      "step": 4080
+    },
+    {
+      "epoch": 4.583462024427909,
+      "grad_norm": 0.3540818181461558,
+      "learning_rate": 4.1599027004560535e-07,
+      "loss": 0.0428,
+      "step": 4081
+    },
+    {
+      "epoch": 4.584585146707848,
+      "grad_norm": 0.34719828160408894,
+      "learning_rate": 4.1375429281236946e-07,
+      "loss": 0.0366,
+      "step": 4082
+    },
+    {
+      "epoch": 4.585708268987786,
+      "grad_norm": 0.3264230696168353,
+      "learning_rate": 4.115242141016085e-07,
+      "loss": 0.0316,
+      "step": 4083
+    },
+    {
+      "epoch": 4.586831391267724,
+      "grad_norm": 0.34719908923103476,
+      "learning_rate": 4.0930003528551587e-07,
+      "loss": 0.0359,
+      "step": 4084
+    },
+    {
+      "epoch": 4.587954513547663,
+      "grad_norm": 0.35625052906788557,
+      "learning_rate": 4.0708175773265246e-07,
+      "loss": 0.0348,
+      "step": 4085
+    },
+    {
+      "epoch": 4.589077635827601,
+      "grad_norm": 0.3650173507417219,
+      "learning_rate": 4.0486938280794754e-07,
+      "loss": 0.038,
+      "step": 4086
+    },
+    {
+      "epoch": 4.590200758107539,
+      "grad_norm": 0.33970468745861326,
+      "learning_rate": 4.0266291187270435e-07,
+      "loss": 0.033,
+      "step": 4087
+    },
+    {
+      "epoch": 4.591323880387478,
+      "grad_norm": 0.3841339639866425,
+      "learning_rate": 4.004623462845825e-07,
+      "loss": 0.0344,
+      "step": 4088
+    },
+    {
+      "epoch": 4.592447002667416,
+      "grad_norm": 0.3247187954531285,
+      "learning_rate": 3.9826768739761765e-07,
+      "loss": 0.0345,
+      "step": 4089
+    },
+    {
+      "epoch": 4.5935701249473535,
+      "grad_norm": 0.3431276872602053,
+      "learning_rate": 3.960789365622075e-07,
+      "loss": 0.0311,
+      "step": 4090
+    },
+    {
+      "epoch": 4.594693247227292,
+      "grad_norm": 0.34267005186851196,
+      "learning_rate": 3.938960951251136e-07,
+      "loss": 0.0376,
+      "step": 4091
+    },
+    {
+      "epoch": 4.59581636950723,
+      "grad_norm": 0.34136270274596336,
+      "learning_rate": 3.917191644294627e-07,
+      "loss": 0.0316,
+      "step": 4092
+    },
+    {
+      "epoch": 4.596939491787168,
+      "grad_norm": 0.34219281794077305,
+      "learning_rate": 3.895481458147454e-07,
+      "loss": 0.0339,
+      "step": 4093
+    },
+    {
+      "epoch": 4.598062614067107,
+      "grad_norm": 0.3695903719561843,
+      "learning_rate": 3.8738304061681107e-07,
+      "loss": 0.0391,
+      "step": 4094
+    },
+    {
+      "epoch": 4.599185736347045,
+      "grad_norm": 0.3426126714468979,
+      "learning_rate": 3.852238501678751e-07,
+      "loss": 0.0339,
+      "step": 4095
+    },
+    {
+      "epoch": 4.600308858626983,
+      "grad_norm": 0.3600465378066275,
+      "learning_rate": 3.830705757965081e-07,
+      "loss": 0.0345,
+      "step": 4096
+    },
+    {
+      "epoch": 4.601431980906922,
+      "grad_norm": 0.4555888651642329,
+      "learning_rate": 3.809232188276468e-07,
+      "loss": 0.0375,
+      "step": 4097
+    },
+    {
+      "epoch": 4.60255510318686,
+      "grad_norm": 0.3441834185775792,
+      "learning_rate": 3.7878178058258217e-07,
+      "loss": 0.0376,
+      "step": 4098
+    },
+    {
+      "epoch": 4.603678225466798,
+      "grad_norm": 0.34364984838035073,
+      "learning_rate": 3.766462623789646e-07,
+      "loss": 0.0334,
+      "step": 4099
+    },
+    {
+      "epoch": 4.6048013477467356,
+      "grad_norm": 0.36747528881747177,
+      "learning_rate": 3.745166655308019e-07,
+      "loss": 0.0404,
+      "step": 4100
+    },
+    {
+      "epoch": 4.605924470026674,
+      "grad_norm": 0.3341683012479436,
+      "learning_rate": 3.723929913484581e-07,
+      "loss": 0.0301,
+      "step": 4101
+    },
+    {
+      "epoch": 4.607047592306612,
+      "grad_norm": 0.3505021440089559,
+      "learning_rate": 3.702752411386534e-07,
+      "loss": 0.0367,
+      "step": 4102
+    },
+    {
+      "epoch": 4.60817071458655,
+      "grad_norm": 0.3474780616469266,
+      "learning_rate": 3.681634162044645e-07,
+      "loss": 0.0358,
+      "step": 4103
+    },
+    {
+      "epoch": 4.609293836866489,
+      "grad_norm": 0.33887657295088386,
+      "learning_rate": 3.6605751784531853e-07,
+      "loss": 0.0362,
+      "step": 4104
+    },
+    {
+      "epoch": 4.610416959146427,
+      "grad_norm": 0.34961137504173107,
+      "learning_rate": 3.6395754735699896e-07,
+      "loss": 0.0379,
+      "step": 4105
+    },
+    {
+      "epoch": 4.611540081426365,
+      "grad_norm": 0.3302529939325971,
+      "learning_rate": 3.6186350603164e-07,
+      "loss": 0.0334,
+      "step": 4106
+    },
+    {
+      "epoch": 4.612663203706304,
+      "grad_norm": 0.33644579720443657,
+      "learning_rate": 3.5977539515772874e-07,
+      "loss": 0.0345,
+      "step": 4107
+    },
+    {
+      "epoch": 4.613786325986242,
+      "grad_norm": 0.3278842607379112,
+      "learning_rate": 3.57693216020103e-07,
+      "loss": 0.0332,
+      "step": 4108
+    },
+    {
+      "epoch": 4.61490944826618,
+      "grad_norm": 0.36040005767567596,
+      "learning_rate": 3.556169698999501e-07,
+      "loss": 0.035,
+      "step": 4109
+    },
+    {
+      "epoch": 4.6160325705461185,
+      "grad_norm": 0.35158408144549563,
+      "learning_rate": 3.535466580748059e-07,
+      "loss": 0.0345,
+      "step": 4110
+    },
+    {
+      "epoch": 4.617155692826056,
+      "grad_norm": 0.36439910096515155,
+      "learning_rate": 3.5148228181855923e-07,
+      "loss": 0.0373,
+      "step": 4111
+    },
+    {
+      "epoch": 4.618278815105994,
+      "grad_norm": 0.33513522124762823,
+      "learning_rate": 3.4942384240144176e-07,
+      "loss": 0.0347,
+      "step": 4112
+    },
+    {
+      "epoch": 4.619401937385933,
+      "grad_norm": 0.34260874749603937,
+      "learning_rate": 3.473713410900326e-07,
+      "loss": 0.0334,
+      "step": 4113
+    },
+    {
+      "epoch": 4.620525059665871,
+      "grad_norm": 0.3352507044783814,
+      "learning_rate": 3.453247791472603e-07,
+      "loss": 0.0357,
+      "step": 4114
+    },
+    {
+      "epoch": 4.621648181945809,
+      "grad_norm": 0.34750467435944776,
+      "learning_rate": 3.4328415783239646e-07,
+      "loss": 0.033,
+      "step": 4115
+    },
+    {
+      "epoch": 4.622771304225748,
+      "grad_norm": 0.3428620974144906,
+      "learning_rate": 3.4124947840105673e-07,
+      "loss": 0.0314,
+      "step": 4116
+    },
+    {
+      "epoch": 4.623894426505686,
+      "grad_norm": 0.36245758828541697,
+      "learning_rate": 3.3922074210520407e-07,
+      "loss": 0.0349,
+      "step": 4117
+    },
+    {
+      "epoch": 4.625017548785624,
+      "grad_norm": 0.35741628717665924,
+      "learning_rate": 3.3719795019313993e-07,
+      "loss": 0.0319,
+      "step": 4118
+    },
+    {
+      "epoch": 4.626140671065563,
+      "grad_norm": 0.3571691112754241,
+      "learning_rate": 3.351811039095121e-07,
+      "loss": 0.0341,
+      "step": 4119
+    },
+    {
+      "epoch": 4.6272637933455005,
+      "grad_norm": 0.3361597736689054,
+      "learning_rate": 3.3317020449530666e-07,
+      "loss": 0.0337,
+      "step": 4120
+    },
+    {
+      "epoch": 4.6283869156254385,
+      "grad_norm": 0.352268450177105,
+      "learning_rate": 3.3116525318785286e-07,
+      "loss": 0.0315,
+      "step": 4121
+    },
+    {
+      "epoch": 4.629510037905377,
+      "grad_norm": 0.3496979817415821,
+      "learning_rate": 3.291662512208216e-07,
+      "loss": 0.0377,
+      "step": 4122
+    },
+    {
+      "epoch": 4.630633160185315,
+      "grad_norm": 0.3415382592263034,
+      "learning_rate": 3.271731998242167e-07,
+      "loss": 0.0349,
+      "step": 4123
+    },
+    {
+      "epoch": 4.631756282465253,
+      "grad_norm": 0.3329970180899988,
+      "learning_rate": 3.2518610022438724e-07,
+      "loss": 0.0308,
+      "step": 4124
+    },
+    {
+      "epoch": 4.632879404745192,
+      "grad_norm": 0.3554638729226304,
+      "learning_rate": 3.2320495364401625e-07,
+      "loss": 0.0396,
+      "step": 4125
+    },
+    {
+      "epoch": 4.63400252702513,
+      "grad_norm": 0.3467526836273791,
+      "learning_rate": 3.2122976130212644e-07,
+      "loss": 0.0334,
+      "step": 4126
+    },
+    {
+      "epoch": 4.635125649305068,
+      "grad_norm": 0.34256948945547966,
+      "learning_rate": 3.192605244140745e-07,
+      "loss": 0.0329,
+      "step": 4127
+    },
+    {
+      "epoch": 4.636248771585007,
+      "grad_norm": 0.35204121377326897,
+      "learning_rate": 3.172972441915523e-07,
+      "loss": 0.0332,
+      "step": 4128
+    },
+    {
+      "epoch": 4.637371893864945,
+      "grad_norm": 0.37489357665496786,
+      "learning_rate": 3.153399218425901e-07,
+      "loss": 0.0365,
+      "step": 4129
+    },
+    {
+      "epoch": 4.638495016144883,
+      "grad_norm": 0.32498835574013707,
+      "learning_rate": 3.13388558571549e-07,
+      "loss": 0.0307,
+      "step": 4130
+    },
+    {
+      "epoch": 4.6396181384248205,
+      "grad_norm": 0.33414718587616266,
+      "learning_rate": 3.114431555791253e-07,
+      "loss": 0.037,
+      "step": 4131
+    },
+    {
+      "epoch": 4.640741260704759,
+      "grad_norm": 0.3384428063498482,
+      "learning_rate": 3.0950371406234357e-07,
+      "loss": 0.0323,
+      "step": 4132
+    },
+    {
+      "epoch": 4.641864382984697,
+      "grad_norm": 0.3493656495995995,
+      "learning_rate": 3.075702352145671e-07,
+      "loss": 0.0337,
+      "step": 4133
+    },
+    {
+      "epoch": 4.642987505264635,
+      "grad_norm": 0.34456307900275657,
+      "learning_rate": 3.0564272022548414e-07,
+      "loss": 0.0321,
+      "step": 4134
+    },
+    {
+      "epoch": 4.644110627544574,
+      "grad_norm": 0.35749872404688643,
+      "learning_rate": 3.0372117028111825e-07,
+      "loss": 0.0353,
+      "step": 4135
+    },
+    {
+      "epoch": 4.645233749824512,
+      "grad_norm": 0.3702263663076714,
+      "learning_rate": 3.0180558656381806e-07,
+      "loss": 0.0365,
+      "step": 4136
+    },
+    {
+      "epoch": 4.64635687210445,
+      "grad_norm": 0.32586875886680355,
+      "learning_rate": 2.9989597025226523e-07,
+      "loss": 0.032,
+      "step": 4137
+    },
+    {
+      "epoch": 4.647479994384389,
+      "grad_norm": 0.33538912472499605,
+      "learning_rate": 2.979923225214665e-07,
+      "loss": 0.0313,
+      "step": 4138
+    },
+    {
+      "epoch": 4.648603116664327,
+      "grad_norm": 0.37507751939830075,
+      "learning_rate": 2.9609464454275707e-07,
+      "loss": 0.0375,
+      "step": 4139
+    },
+    {
+      "epoch": 4.649726238944265,
+      "grad_norm": 0.3481121639949164,
+      "learning_rate": 2.942029374837996e-07,
+      "loss": 0.0314,
+      "step": 4140
+    },
+    {
+      "epoch": 4.6508493612242034,
+      "grad_norm": 0.3423593643914144,
+      "learning_rate": 2.9231720250858296e-07,
+      "loss": 0.0343,
+      "step": 4141
+    },
+    {
+      "epoch": 4.651972483504141,
+      "grad_norm": 0.35719566136893616,
+      "learning_rate": 2.904374407774191e-07,
+      "loss": 0.0366,
+      "step": 4142
+    },
+    {
+      "epoch": 4.653095605784079,
+      "grad_norm": 0.3658780078221181,
+      "learning_rate": 2.8856365344694604e-07,
+      "loss": 0.0348,
+      "step": 4143
+    },
+    {
+      "epoch": 4.654218728064018,
+      "grad_norm": 0.34290940357613336,
+      "learning_rate": 2.866958416701271e-07,
+      "loss": 0.0335,
+      "step": 4144
+    },
+    {
+      "epoch": 4.655341850343956,
+      "grad_norm": 0.33735481225106245,
+      "learning_rate": 2.8483400659624737e-07,
+      "loss": 0.0354,
+      "step": 4145
+    },
+    {
+      "epoch": 4.656464972623894,
+      "grad_norm": 0.35331730640988807,
+      "learning_rate": 2.8297814937091495e-07,
+      "loss": 0.0363,
+      "step": 4146
+    },
+    {
+      "epoch": 4.657588094903833,
+      "grad_norm": 0.3582352972813522,
+      "learning_rate": 2.8112827113605637e-07,
+      "loss": 0.0321,
+      "step": 4147
+    },
+    {
+      "epoch": 4.658711217183771,
+      "grad_norm": 0.369439070620047,
+      "learning_rate": 2.792843730299244e-07,
+      "loss": 0.0409,
+      "step": 4148
+    },
+    {
+      "epoch": 4.659834339463709,
+      "grad_norm": 0.36467256732278813,
+      "learning_rate": 2.774464561870893e-07,
+      "loss": 0.0331,
+      "step": 4149
+    },
+    {
+      "epoch": 4.660957461743648,
+      "grad_norm": 0.3578494257105483,
+      "learning_rate": 2.7561452173844206e-07,
+      "loss": 0.0348,
+      "step": 4150
+    },
+    {
+      "epoch": 4.6620805840235855,
+      "grad_norm": 0.33043652998644596,
+      "learning_rate": 2.7378857081119204e-07,
+      "loss": 0.0291,
+      "step": 4151
+    },
+    {
+      "epoch": 4.663203706303523,
+      "grad_norm": 0.32682915997898,
+      "learning_rate": 2.7196860452886496e-07,
+      "loss": 0.0295,
+      "step": 4152
+    },
+    {
+      "epoch": 4.664326828583462,
+      "grad_norm": 0.35842979933675306,
+      "learning_rate": 2.7015462401130843e-07,
+      "loss": 0.0347,
+      "step": 4153
+    },
+    {
+      "epoch": 4.6654499508634,
+      "grad_norm": 0.3608036317898208,
+      "learning_rate": 2.68346630374684e-07,
+      "loss": 0.0351,
+      "step": 4154
+    },
+    {
+      "epoch": 4.666573073143338,
+      "grad_norm": 0.3445352154282416,
+      "learning_rate": 2.665446247314696e-07,
+      "loss": 0.0368,
+      "step": 4155
+    },
+    {
+      "epoch": 4.667696195423277,
+      "grad_norm": 0.36637403299593924,
+      "learning_rate": 2.6474860819046157e-07,
+      "loss": 0.0384,
+      "step": 4156
+    },
+    {
+      "epoch": 4.668819317703215,
+      "grad_norm": 0.3402971768771307,
+      "learning_rate": 2.629585818567637e-07,
+      "loss": 0.0336,
+      "step": 4157
+    },
+    {
+      "epoch": 4.669942439983153,
+      "grad_norm": 0.32819300144957086,
+      "learning_rate": 2.6117454683180274e-07,
+      "loss": 0.0312,
+      "step": 4158
+    },
+    {
+      "epoch": 4.671065562263092,
+      "grad_norm": 0.3449374419407223,
+      "learning_rate": 2.5939650421331395e-07,
+      "loss": 0.0334,
+      "step": 4159
+    },
+    {
+      "epoch": 4.67218868454303,
+      "grad_norm": 0.3596838423275439,
+      "learning_rate": 2.576244550953466e-07,
+      "loss": 0.0375,
+      "step": 4160
+    },
+    {
+      "epoch": 4.6733118068229675,
+      "grad_norm": 0.3685410673567921,
+      "learning_rate": 2.5585840056826295e-07,
+      "loss": 0.0381,
+      "step": 4161
+    },
+    {
+      "epoch": 4.674434929102906,
+      "grad_norm": 0.352525835209393,
+      "learning_rate": 2.540983417187348e-07,
+      "loss": 0.033,
+      "step": 4162
+    },
+    {
+      "epoch": 4.675558051382844,
+      "grad_norm": 0.3406022335981378,
+      "learning_rate": 2.5234427962974486e-07,
+      "loss": 0.03,
+      "step": 4163
+    },
+    {
+      "epoch": 4.676681173662782,
+      "grad_norm": 0.3732219938497303,
+      "learning_rate": 2.5059621538058743e-07,
+      "loss": 0.0327,
+      "step": 4164
+    },
+    {
+      "epoch": 4.677804295942721,
+      "grad_norm": 0.3454388239084282,
+      "learning_rate": 2.488541500468666e-07,
+      "loss": 0.0321,
+      "step": 4165
+    },
+    {
+      "epoch": 4.678927418222659,
+      "grad_norm": 0.3482806759890393,
+      "learning_rate": 2.47118084700495e-07,
+      "loss": 0.0334,
+      "step": 4166
+    },
+    {
+      "epoch": 4.680050540502597,
+      "grad_norm": 0.3423367434752234,
+      "learning_rate": 2.453880204096892e-07,
+      "loss": 0.0325,
+      "step": 4167
+    },
+    {
+      "epoch": 4.681173662782536,
+      "grad_norm": 0.35270956062249814,
+      "learning_rate": 2.4366395823898104e-07,
+      "loss": 0.0339,
+      "step": 4168
+    },
+    {
+      "epoch": 4.682296785062474,
+      "grad_norm": 0.3296367877167982,
+      "learning_rate": 2.419458992492019e-07,
+      "loss": 0.0329,
+      "step": 4169
+    },
+    {
+      "epoch": 4.683419907342412,
+      "grad_norm": 0.33676841186669126,
+      "learning_rate": 2.402338444974928e-07,
+      "loss": 0.0304,
+      "step": 4170
+    },
+    {
+      "epoch": 4.6845430296223505,
+      "grad_norm": 0.3432291676295886,
+      "learning_rate": 2.3852779503730217e-07,
+      "loss": 0.0357,
+      "step": 4171
+    },
+    {
+      "epoch": 4.685666151902288,
+      "grad_norm": 0.3415736112710834,
+      "learning_rate": 2.3682775191837814e-07,
+      "loss": 0.0331,
+      "step": 4172
+    },
+    {
+      "epoch": 4.686789274182226,
+      "grad_norm": 0.3718654445263105,
+      "learning_rate": 2.3513371618677838e-07,
+      "loss": 0.036,
+      "step": 4173
+    },
+    {
+      "epoch": 4.687912396462165,
+      "grad_norm": 0.3659974123825445,
+      "learning_rate": 2.3344568888485907e-07,
+      "loss": 0.0361,
+      "step": 4174
+    },
+    {
+      "epoch": 4.689035518742103,
+      "grad_norm": 0.34842971301416137,
+      "learning_rate": 2.3176367105128494e-07,
+      "loss": 0.0353,
+      "step": 4175
+    },
+    {
+      "epoch": 4.690158641022041,
+      "grad_norm": 0.35992964433081653,
+      "learning_rate": 2.300876637210181e-07,
+      "loss": 0.0352,
+      "step": 4176
+    },
+    {
+      "epoch": 4.69128176330198,
+      "grad_norm": 0.35358447232276063,
+      "learning_rate": 2.2841766792532472e-07,
+      "loss": 0.0367,
+      "step": 4177
+    },
+    {
+      "epoch": 4.692404885581918,
+      "grad_norm": 0.3458896703817227,
+      "learning_rate": 2.2675368469177171e-07,
+      "loss": 0.0319,
+      "step": 4178
+    },
+    {
+      "epoch": 4.693528007861856,
+      "grad_norm": 0.33681153882758513,
+      "learning_rate": 2.2509571504422678e-07,
+      "loss": 0.035,
+      "step": 4179
+    },
+    {
+      "epoch": 4.694651130141795,
+      "grad_norm": 0.34571417197318716,
+      "learning_rate": 2.2344376000285606e-07,
+      "loss": 0.035,
+      "step": 4180
+    },
+    {
+      "epoch": 4.6957742524217325,
+      "grad_norm": 0.33932945848040214,
+      "learning_rate": 2.2179782058412646e-07,
+      "loss": 0.0327,
+      "step": 4181
+    },
+    {
+      "epoch": 4.6968973747016705,
+      "grad_norm": 0.34489622098163436,
+      "learning_rate": 2.201578978008012e-07,
+      "loss": 0.0337,
+      "step": 4182
+    },
+    {
+      "epoch": 4.698020496981609,
+      "grad_norm": 0.3480940006392897,
+      "learning_rate": 2.1852399266194312e-07,
+      "loss": 0.0318,
+      "step": 4183
+    },
+    {
+      "epoch": 4.699143619261547,
+      "grad_norm": 0.33290094072681525,
+      "learning_rate": 2.1689610617291357e-07,
+      "loss": 0.0332,
+      "step": 4184
+    },
+    {
+      "epoch": 4.700266741541485,
+      "grad_norm": 0.3599868420281495,
+      "learning_rate": 2.1527423933536906e-07,
+      "loss": 0.0385,
+      "step": 4185
+    },
+    {
+      "epoch": 4.701389863821424,
+      "grad_norm": 0.3337751304343682,
+      "learning_rate": 2.1365839314726021e-07,
+      "loss": 0.03,
+      "step": 4186
+    },
+    {
+      "epoch": 4.702512986101362,
+      "grad_norm": 0.33760321339006033,
+      "learning_rate": 2.1204856860283506e-07,
+      "loss": 0.0353,
+      "step": 4187
+    },
+    {
+      "epoch": 4.7036361083813,
+      "grad_norm": 0.3357458834520037,
+      "learning_rate": 2.1044476669263793e-07,
+      "loss": 0.0316,
+      "step": 4188
+    },
+    {
+      "epoch": 4.704759230661239,
+      "grad_norm": 0.3494253152613926,
+      "learning_rate": 2.0884698840350492e-07,
+      "loss": 0.0352,
+      "step": 4189
+    },
+    {
+      "epoch": 4.705882352941177,
+      "grad_norm": 0.3313642700196441,
+      "learning_rate": 2.0725523471856744e-07,
+      "loss": 0.0344,
+      "step": 4190
+    },
+    {
+      "epoch": 4.707005475221115,
+      "grad_norm": 0.3515014279713326,
+      "learning_rate": 2.056695066172476e-07,
+      "loss": 0.0375,
+      "step": 4191
+    },
+    {
+      "epoch": 4.708128597501053,
+      "grad_norm": 0.3393284033675587,
+      "learning_rate": 2.0408980507526267e-07,
+      "loss": 0.0324,
+      "step": 4192
+    },
+    {
+      "epoch": 4.709251719780991,
+      "grad_norm": 0.3433377101697511,
+      "learning_rate": 2.0251613106461955e-07,
+      "loss": 0.0351,
+      "step": 4193
+    },
+    {
+      "epoch": 4.710374842060929,
+      "grad_norm": 0.3609914609514239,
+      "learning_rate": 2.0094848555361702e-07,
+      "loss": 0.0343,
+      "step": 4194
+    },
+    {
+      "epoch": 4.711497964340868,
+      "grad_norm": 0.33674994452184703,
+      "learning_rate": 1.993868695068457e-07,
+      "loss": 0.0332,
+      "step": 4195
+    },
+    {
+      "epoch": 4.712621086620806,
+      "grad_norm": 0.3405640938227505,
+      "learning_rate": 1.978312838851837e-07,
+      "loss": 0.0338,
+      "step": 4196
+    },
+    {
+      "epoch": 4.713744208900744,
+      "grad_norm": 0.3518234996906055,
+      "learning_rate": 1.9628172964580082e-07,
+      "loss": 0.0323,
+      "step": 4197
+    },
+    {
+      "epoch": 4.714867331180682,
+      "grad_norm": 0.37220183549793007,
+      "learning_rate": 1.9473820774215557e-07,
+      "loss": 0.0342,
+      "step": 4198
+    },
+    {
+      "epoch": 4.715990453460621,
+      "grad_norm": 0.32965656846842434,
+      "learning_rate": 1.932007191239915e-07,
+      "loss": 0.0295,
+      "step": 4199
+    },
+    {
+      "epoch": 4.717113575740559,
+      "grad_norm": 0.3521785347247238,
+      "learning_rate": 1.9166926473734636e-07,
+      "loss": 0.033,
+      "step": 4200
+    },
+    {
+      "epoch": 4.718236698020497,
+      "grad_norm": 0.35522561854409634,
+      "learning_rate": 1.9014384552453635e-07,
+      "loss": 0.0338,
+      "step": 4201
+    },
+    {
+      "epoch": 4.719359820300435,
+      "grad_norm": 0.3310871063889209,
+      "learning_rate": 1.8862446242417175e-07,
+      "loss": 0.0327,
+      "step": 4202
+    },
+    {
+      "epoch": 4.720482942580373,
+      "grad_norm": 0.3239117817488521,
+      "learning_rate": 1.8711111637114364e-07,
+      "loss": 0.0334,
+      "step": 4203
+    },
+    {
+      "epoch": 4.721606064860311,
+      "grad_norm": 0.33473520840189047,
+      "learning_rate": 1.856038082966327e-07,
+      "loss": 0.0315,
+      "step": 4204
+    },
+    {
+      "epoch": 4.72272918714025,
+      "grad_norm": 0.3546234500423055,
+      "learning_rate": 1.841025391281015e-07,
+      "loss": 0.0347,
+      "step": 4205
+    },
+    {
+      "epoch": 4.723852309420188,
+      "grad_norm": 0.35300752775260824,
+      "learning_rate": 1.8260730978929664e-07,
+      "loss": 0.0373,
+      "step": 4206
+    },
+    {
+      "epoch": 4.724975431700126,
+      "grad_norm": 0.31962674294469384,
+      "learning_rate": 1.8111812120024884e-07,
+      "loss": 0.0304,
+      "step": 4207
+    },
+    {
+      "epoch": 4.726098553980065,
+      "grad_norm": 0.3606715730669359,
+      "learning_rate": 1.7963497427727294e-07,
+      "loss": 0.0337,
+      "step": 4208
+    },
+    {
+      "epoch": 4.727221676260003,
+      "grad_norm": 0.32723932407883416,
+      "learning_rate": 1.781578699329667e-07,
+      "loss": 0.032,
+      "step": 4209
+    },
+    {
+      "epoch": 4.728344798539941,
+      "grad_norm": 0.32660649785881374,
+      "learning_rate": 1.766868090762075e-07,
+      "loss": 0.0304,
+      "step": 4210
+    },
+    {
+      "epoch": 4.7294679208198795,
+      "grad_norm": 0.3834637625549436,
+      "learning_rate": 1.7522179261215467e-07,
+      "loss": 0.043,
+      "step": 4211
+    },
+    {
+      "epoch": 4.7305910430998175,
+      "grad_norm": 0.31299817068786057,
+      "learning_rate": 1.7376282144224933e-07,
+      "loss": 0.0302,
+      "step": 4212
+    },
+    {
+      "epoch": 4.731714165379755,
+      "grad_norm": 0.34864103394231777,
+      "learning_rate": 1.7230989646421337e-07,
+      "loss": 0.0329,
+      "step": 4213
+    },
+    {
+      "epoch": 4.732837287659694,
+      "grad_norm": 0.35855522330101297,
+      "learning_rate": 1.7086301857204725e-07,
+      "loss": 0.0327,
+      "step": 4214
+    },
+    {
+      "epoch": 4.733960409939632,
+      "grad_norm": 0.35553712042165425,
+      "learning_rate": 1.694221886560299e-07,
+      "loss": 0.037,
+      "step": 4215
+    },
+    {
+      "epoch": 4.73508353221957,
+      "grad_norm": 0.3348937621754762,
+      "learning_rate": 1.6798740760272104e-07,
+      "loss": 0.0329,
+      "step": 4216
+    },
+    {
+      "epoch": 4.736206654499509,
+      "grad_norm": 0.3809215375029254,
+      "learning_rate": 1.665586762949567e-07,
+      "loss": 0.0366,
+      "step": 4217
+    },
+    {
+      "epoch": 4.737329776779447,
+      "grad_norm": 0.34204433938299095,
+      "learning_rate": 1.6513599561185034e-07,
+      "loss": 0.0328,
+      "step": 4218
+    },
+    {
+      "epoch": 4.738452899059385,
+      "grad_norm": 0.3550432910580413,
+      "learning_rate": 1.6371936642879504e-07,
+      "loss": 0.0321,
+      "step": 4219
+    },
+    {
+      "epoch": 4.739576021339324,
+      "grad_norm": 0.3318127050500399,
+      "learning_rate": 1.623087896174558e-07,
+      "loss": 0.0365,
+      "step": 4220
+    },
+    {
+      "epoch": 4.740699143619262,
+      "grad_norm": 0.36204671764832996,
+      "learning_rate": 1.6090426604577714e-07,
+      "loss": 0.0346,
+      "step": 4221
+    },
+    {
+      "epoch": 4.7418222658991995,
+      "grad_norm": 0.3387248418153391,
+      "learning_rate": 1.5950579657797894e-07,
+      "loss": 0.0306,
+      "step": 4222
+    },
+    {
+      "epoch": 4.742945388179138,
+      "grad_norm": 0.3427148305083139,
+      "learning_rate": 1.5811338207455284e-07,
+      "loss": 0.0345,
+      "step": 4223
+    },
+    {
+      "epoch": 4.744068510459076,
+      "grad_norm": 0.35411961450361146,
+      "learning_rate": 1.5672702339226909e-07,
+      "loss": 0.0371,
+      "step": 4224
+    },
+    {
+      "epoch": 4.745191632739014,
+      "grad_norm": 0.3721507850238725,
+      "learning_rate": 1.553467213841664e-07,
+      "loss": 0.0368,
+      "step": 4225
+    },
+    {
+      "epoch": 4.746314755018953,
+      "grad_norm": 0.3466898365608997,
+      "learning_rate": 1.5397247689956318e-07,
+      "loss": 0.0308,
+      "step": 4226
+    },
+    {
+      "epoch": 4.747437877298891,
+      "grad_norm": 0.3574266108394466,
+      "learning_rate": 1.5260429078404416e-07,
+      "loss": 0.0335,
+      "step": 4227
+    },
+    {
+      "epoch": 4.748560999578829,
+      "grad_norm": 0.3400286952615133,
+      "learning_rate": 1.5124216387947143e-07,
+      "loss": 0.0307,
+      "step": 4228
+    },
+    {
+      "epoch": 4.749684121858768,
+      "grad_norm": 0.36309544829463203,
+      "learning_rate": 1.4988609702397683e-07,
+      "loss": 0.04,
+      "step": 4229
+    },
+    {
+      "epoch": 4.750807244138706,
+      "grad_norm": 0.3417693922064799,
+      "learning_rate": 1.4853609105196175e-07,
+      "loss": 0.0367,
+      "step": 4230
+    },
+    {
+      "epoch": 4.751930366418644,
+      "grad_norm": 0.3602988292077408,
+      "learning_rate": 1.4719214679409954e-07,
+      "loss": 0.0358,
+      "step": 4231
+    },
+    {
+      "epoch": 4.753053488698582,
+      "grad_norm": 0.3735538962893692,
+      "learning_rate": 1.4585426507733536e-07,
+      "loss": 0.0311,
+      "step": 4232
+    },
+    {
+      "epoch": 4.75417661097852,
+      "grad_norm": 0.39448460405940855,
+      "learning_rate": 1.445224467248818e-07,
+      "loss": 0.0385,
+      "step": 4233
+    },
+    {
+      "epoch": 4.755299733258458,
+      "grad_norm": 0.36862373252888875,
+      "learning_rate": 1.4319669255622115e-07,
+      "loss": 0.0378,
+      "step": 4234
+    },
+    {
+      "epoch": 4.756422855538396,
+      "grad_norm": 0.3606558317899239,
+      "learning_rate": 1.418770033871053e-07,
+      "loss": 0.0351,
+      "step": 4235
+    },
+    {
+      "epoch": 4.757545977818335,
+      "grad_norm": 0.357546164191823,
+      "learning_rate": 1.405633800295525e-07,
+      "loss": 0.0377,
+      "step": 4236
+    },
+    {
+      "epoch": 4.758669100098273,
+      "grad_norm": 0.33523607494626556,
+      "learning_rate": 1.392558232918506e-07,
+      "loss": 0.0313,
+      "step": 4237
+    },
+    {
+      "epoch": 4.759792222378211,
+      "grad_norm": 0.3554788435214293,
+      "learning_rate": 1.3795433397855274e-07,
+      "loss": 0.0363,
+      "step": 4238
+    },
+    {
+      "epoch": 4.76091534465815,
+      "grad_norm": 0.334541232594485,
+      "learning_rate": 1.366589128904805e-07,
+      "loss": 0.0331,
+      "step": 4239
+    },
+    {
+      "epoch": 4.762038466938088,
+      "grad_norm": 0.36173611814235396,
+      "learning_rate": 1.3536956082472074e-07,
+      "loss": 0.0369,
+      "step": 4240
+    },
+    {
+      "epoch": 4.763161589218026,
+      "grad_norm": 0.3250258742570545,
+      "learning_rate": 1.3408627857462443e-07,
+      "loss": 0.0318,
+      "step": 4241
+    },
+    {
+      "epoch": 4.7642847114979645,
+      "grad_norm": 0.3565828415127001,
+      "learning_rate": 1.32809066929811e-07,
+      "loss": 0.0387,
+      "step": 4242
+    },
+    {
+      "epoch": 4.7654078337779024,
+      "grad_norm": 0.3383956134948221,
+      "learning_rate": 1.3153792667616183e-07,
+      "loss": 0.0332,
+      "step": 4243
+    },
+    {
+      "epoch": 4.76653095605784,
+      "grad_norm": 0.3476033007253469,
+      "learning_rate": 1.302728585958246e-07,
+      "loss": 0.0365,
+      "step": 4244
+    },
+    {
+      "epoch": 4.767654078337779,
+      "grad_norm": 0.3516963013877261,
+      "learning_rate": 1.290138634672089e-07,
+      "loss": 0.0386,
+      "step": 4245
+    },
+    {
+      "epoch": 4.768777200617717,
+      "grad_norm": 0.34481467608140326,
+      "learning_rate": 1.2776094206498834e-07,
+      "loss": 0.0339,
+      "step": 4246
+    },
+    {
+      "epoch": 4.769900322897655,
+      "grad_norm": 0.3925266775395916,
+      "learning_rate": 1.2651409516009848e-07,
+      "loss": 0.0345,
+      "step": 4247
+    },
+    {
+      "epoch": 4.771023445177594,
+      "grad_norm": 0.356302010857859,
+      "learning_rate": 1.2527332351973899e-07,
+      "loss": 0.0342,
+      "step": 4248
+    },
+    {
+      "epoch": 4.772146567457532,
+      "grad_norm": 0.3385658467723331,
+      "learning_rate": 1.2403862790737021e-07,
+      "loss": 0.0297,
+      "step": 4249
+    },
+    {
+      "epoch": 4.77326968973747,
+      "grad_norm": 0.34274297140098986,
+      "learning_rate": 1.2281000908271336e-07,
+      "loss": 0.0335,
+      "step": 4250
+    },
+    {
+      "epoch": 4.774392812017409,
+      "grad_norm": 0.32904157582350113,
+      "learning_rate": 1.2158746780175257e-07,
+      "loss": 0.0324,
+      "step": 4251
+    },
+    {
+      "epoch": 4.7755159342973466,
+      "grad_norm": 0.3583962692152138,
+      "learning_rate": 1.2037100481672836e-07,
+      "loss": 0.0349,
+      "step": 4252
+    },
+    {
+      "epoch": 4.7766390565772845,
+      "grad_norm": 0.3411729792940599,
+      "learning_rate": 1.1916062087614644e-07,
+      "loss": 0.0344,
+      "step": 4253
+    },
+    {
+      "epoch": 4.777762178857223,
+      "grad_norm": 0.34168029879821443,
+      "learning_rate": 1.1795631672476771e-07,
+      "loss": 0.0364,
+      "step": 4254
+    },
+    {
+      "epoch": 4.778885301137161,
+      "grad_norm": 0.35060334074323224,
+      "learning_rate": 1.1675809310361497e-07,
+      "loss": 0.0466,
+      "step": 4255
+    },
+    {
+      "epoch": 4.780008423417099,
+      "grad_norm": 0.34336606647480034,
+      "learning_rate": 1.1556595074996624e-07,
+      "loss": 0.0331,
+      "step": 4256
+    },
+    {
+      "epoch": 4.781131545697038,
+      "grad_norm": 0.32382578010358776,
+      "learning_rate": 1.1437989039736253e-07,
+      "loss": 0.0325,
+      "step": 4257
+    },
+    {
+      "epoch": 4.782254667976976,
+      "grad_norm": 0.34851470783798044,
+      "learning_rate": 1.1319991277559783e-07,
+      "loss": 0.036,
+      "step": 4258
+    },
+    {
+      "epoch": 4.783377790256914,
+      "grad_norm": 0.35953669765583024,
+      "learning_rate": 1.1202601861072693e-07,
+      "loss": 0.0364,
+      "step": 4259
+    },
+    {
+      "epoch": 4.784500912536853,
+      "grad_norm": 0.3420834182439802,
+      "learning_rate": 1.108582086250587e-07,
+      "loss": 0.0304,
+      "step": 4260
+    },
+    {
+      "epoch": 4.785624034816791,
+      "grad_norm": 0.3474035202936612,
+      "learning_rate": 1.0969648353715945e-07,
+      "loss": 0.0353,
+      "step": 4261
+    },
+    {
+      "epoch": 4.786747157096729,
+      "grad_norm": 0.33774829169251014,
+      "learning_rate": 1.0854084406185184e-07,
+      "loss": 0.0309,
+      "step": 4262
+    },
+    {
+      "epoch": 4.787870279376667,
+      "grad_norm": 0.3323004162031955,
+      "learning_rate": 1.0739129091021372e-07,
+      "loss": 0.0301,
+      "step": 4263
+    },
+    {
+      "epoch": 4.788993401656605,
+      "grad_norm": 0.35491038528232177,
+      "learning_rate": 1.0624782478957818e-07,
+      "loss": 0.0362,
+      "step": 4264
+    },
+    {
+      "epoch": 4.790116523936543,
+      "grad_norm": 0.3148060898080571,
+      "learning_rate": 1.051104464035313e-07,
+      "loss": 0.03,
+      "step": 4265
+    },
+    {
+      "epoch": 4.791239646216482,
+      "grad_norm": 0.3454633600900965,
+      "learning_rate": 1.0397915645191437e-07,
+      "loss": 0.0325,
+      "step": 4266
+    },
+    {
+      "epoch": 4.79236276849642,
+      "grad_norm": 0.3387657241758695,
+      "learning_rate": 1.028539556308239e-07,
+      "loss": 0.0366,
+      "step": 4267
+    },
+    {
+      "epoch": 4.793485890776358,
+      "grad_norm": 0.343688514356464,
+      "learning_rate": 1.017348446326083e-07,
+      "loss": 0.0355,
+      "step": 4268
+    },
+    {
+      "epoch": 4.794609013056297,
+      "grad_norm": 0.33804452702394205,
+      "learning_rate": 1.0062182414586786e-07,
+      "loss": 0.0317,
+      "step": 4269
+    },
+    {
+      "epoch": 4.795732135336235,
+      "grad_norm": 0.34775190604371375,
+      "learning_rate": 9.951489485545696e-08,
+      "loss": 0.0357,
+      "step": 4270
+    },
+    {
+      "epoch": 4.796855257616173,
+      "grad_norm": 0.34403813226107627,
+      "learning_rate": 9.841405744248078e-08,
+      "loss": 0.036,
+      "step": 4271
+    },
+    {
+      "epoch": 4.7979783798961115,
+      "grad_norm": 0.337690442201065,
+      "learning_rate": 9.731931258429638e-08,
+      "loss": 0.0336,
+      "step": 4272
+    },
+    {
+      "epoch": 4.7991015021760495,
+      "grad_norm": 0.36856828496724703,
+      "learning_rate": 9.623066095451494e-08,
+      "loss": 0.04,
+      "step": 4273
+    },
+    {
+      "epoch": 4.800224624455987,
+      "grad_norm": 0.3454697388417179,
+      "learning_rate": 9.514810322299283e-08,
+      "loss": 0.0327,
+      "step": 4274
+    },
+    {
+      "epoch": 4.801347746735926,
+      "grad_norm": 0.3462964343590165,
+      "learning_rate": 9.407164005584057e-08,
+      "loss": 0.0333,
+      "step": 4275
+    },
+    {
+      "epoch": 4.802470869015864,
+      "grad_norm": 0.3431968345054756,
+      "learning_rate": 9.300127211541832e-08,
+      "loss": 0.0314,
+      "step": 4276
+    },
+    {
+      "epoch": 4.803593991295802,
+      "grad_norm": 0.3651225919722223,
+      "learning_rate": 9.193700006033368e-08,
+      "loss": 0.0412,
+      "step": 4277
+    },
+    {
+      "epoch": 4.804717113575741,
+      "grad_norm": 0.33522720006840734,
+      "learning_rate": 9.087882454544839e-08,
+      "loss": 0.0298,
+      "step": 4278
+    },
+    {
+      "epoch": 4.805840235855679,
+      "grad_norm": 0.3320171412387023,
+      "learning_rate": 8.982674622186605e-08,
+      "loss": 0.0309,
+      "step": 4279
+    },
+    {
+      "epoch": 4.806963358135617,
+      "grad_norm": 0.3506412235588877,
+      "learning_rate": 8.878076573694328e-08,
+      "loss": 0.0344,
+      "step": 4280
+    },
+    {
+      "epoch": 4.808086480415556,
+      "grad_norm": 0.36747870462605653,
+      "learning_rate": 8.774088373428413e-08,
+      "loss": 0.0401,
+      "step": 4281
+    },
+    {
+      "epoch": 4.809209602695494,
+      "grad_norm": 0.3537987285313262,
+      "learning_rate": 8.67071008537379e-08,
+      "loss": 0.0342,
+      "step": 4282
+    },
+    {
+      "epoch": 4.8103327249754315,
+      "grad_norm": 0.33629875351700084,
+      "learning_rate": 8.567941773140465e-08,
+      "loss": 0.0332,
+      "step": 4283
+    },
+    {
+      "epoch": 4.81145584725537,
+      "grad_norm": 0.3534627818646823,
+      "learning_rate": 8.465783499962633e-08,
+      "loss": 0.0354,
+      "step": 4284
+    },
+    {
+      "epoch": 4.812578969535308,
+      "grad_norm": 0.3630589407739355,
+      "learning_rate": 8.364235328699566e-08,
+      "loss": 0.0352,
+      "step": 4285
+    },
+    {
+      "epoch": 4.813702091815246,
+      "grad_norm": 0.33985452597445087,
+      "learning_rate": 8.263297321835062e-08,
+      "loss": 0.032,
+      "step": 4286
+    },
+    {
+      "epoch": 4.814825214095185,
+      "grad_norm": 0.355931019590897,
+      "learning_rate": 8.162969541477217e-08,
+      "loss": 0.0323,
+      "step": 4287
+    },
+    {
+      "epoch": 4.815948336375123,
+      "grad_norm": 0.36125431699041477,
+      "learning_rate": 8.063252049358983e-08,
+      "loss": 0.04,
+      "step": 4288
+    },
+    {
+      "epoch": 4.817071458655061,
+      "grad_norm": 0.3354243066230309,
+      "learning_rate": 7.96414490683739e-08,
+      "loss": 0.0334,
+      "step": 4289
+    },
+    {
+      "epoch": 4.818194580935,
+      "grad_norm": 0.35478986196267864,
+      "learning_rate": 7.865648174894325e-08,
+      "loss": 0.0338,
+      "step": 4290
+    },
+    {
+      "epoch": 4.819317703214938,
+      "grad_norm": 0.34998123448801854,
+      "learning_rate": 7.767761914135974e-08,
+      "loss": 0.0339,
+      "step": 4291
+    },
+    {
+      "epoch": 4.820440825494876,
+      "grad_norm": 0.35949343783541043,
+      "learning_rate": 7.670486184792713e-08,
+      "loss": 0.0333,
+      "step": 4292
+    },
+    {
+      "epoch": 4.8215639477748145,
+      "grad_norm": 0.3651218793634851,
+      "learning_rate": 7.573821046719332e-08,
+      "loss": 0.0394,
+      "step": 4293
+    },
+    {
+      "epoch": 4.822687070054752,
+      "grad_norm": 0.34634749486357275,
+      "learning_rate": 7.477766559395139e-08,
+      "loss": 0.0327,
+      "step": 4294
+    },
+    {
+      "epoch": 4.82381019233469,
+      "grad_norm": 0.3413706560361767,
+      "learning_rate": 7.382322781923301e-08,
+      "loss": 0.0367,
+      "step": 4295
+    },
+    {
+      "epoch": 4.824933314614629,
+      "grad_norm": 0.3475100318642269,
+      "learning_rate": 7.287489773031508e-08,
+      "loss": 0.0344,
+      "step": 4296
+    },
+    {
+      "epoch": 4.826056436894567,
+      "grad_norm": 0.35140910720099416,
+      "learning_rate": 7.193267591071529e-08,
+      "loss": 0.0313,
+      "step": 4297
+    },
+    {
+      "epoch": 4.827179559174505,
+      "grad_norm": 0.30755190051409365,
+      "learning_rate": 7.09965629401943e-08,
+      "loss": 0.0299,
+      "step": 4298
+    },
+    {
+      "epoch": 4.828302681454443,
+      "grad_norm": 0.3208778262419729,
+      "learning_rate": 7.006655939475248e-08,
+      "loss": 0.0304,
+      "step": 4299
+    },
+    {
+      "epoch": 4.829425803734382,
+      "grad_norm": 0.33024767193628707,
+      "learning_rate": 6.914266584662988e-08,
+      "loss": 0.0317,
+      "step": 4300
+    },
+    {
+      "epoch": 4.83054892601432,
+      "grad_norm": 0.3600374450371469,
+      "learning_rate": 6.82248828643084e-08,
+      "loss": 0.0358,
+      "step": 4301
+    },
+    {
+      "epoch": 4.831672048294258,
+      "grad_norm": 0.343893665355897,
+      "learning_rate": 6.731321101251187e-08,
+      "loss": 0.0348,
+      "step": 4302
+    },
+    {
+      "epoch": 4.8327951705741965,
+      "grad_norm": 0.3401993628458368,
+      "learning_rate": 6.640765085220047e-08,
+      "loss": 0.0334,
+      "step": 4303
+    },
+    {
+      "epoch": 4.833918292854134,
+      "grad_norm": 0.37117274391604826,
+      "learning_rate": 6.550820294057625e-08,
+      "loss": 0.038,
+      "step": 4304
+    },
+    {
+      "epoch": 4.835041415134072,
+      "grad_norm": 0.3270123504885918,
+      "learning_rate": 6.461486783107762e-08,
+      "loss": 0.0318,
+      "step": 4305
+    },
+    {
+      "epoch": 4.836164537414011,
+      "grad_norm": 0.35563694733212037,
+      "learning_rate": 6.3727646073386e-08,
+      "loss": 0.0354,
+      "step": 4306
+    },
+    {
+      "epoch": 4.837287659693949,
+      "grad_norm": 0.3312605443855957,
+      "learning_rate": 6.284653821341691e-08,
+      "loss": 0.0341,
+      "step": 4307
+    },
+    {
+      "epoch": 4.838410781973887,
+      "grad_norm": 0.34887322226150547,
+      "learning_rate": 6.197154479332667e-08,
+      "loss": 0.0356,
+      "step": 4308
+    },
+    {
+      "epoch": 4.839533904253826,
+      "grad_norm": 0.34169094625118857,
+      "learning_rate": 6.110266635150796e-08,
+      "loss": 0.0332,
+      "step": 4309
+    },
+    {
+      "epoch": 4.840657026533764,
+      "grad_norm": 0.3460642146322915,
+      "learning_rate": 6.02399034225909e-08,
+      "loss": 0.0356,
+      "step": 4310
+    },
+    {
+      "epoch": 4.841780148813702,
+      "grad_norm": 0.33286836691758975,
+      "learning_rate": 5.9383256537444144e-08,
+      "loss": 0.0297,
+      "step": 4311
+    },
+    {
+      "epoch": 4.842903271093641,
+      "grad_norm": 0.3402965072293783,
+      "learning_rate": 5.853272622317052e-08,
+      "loss": 0.0349,
+      "step": 4312
+    },
+    {
+      "epoch": 4.8440263933735785,
+      "grad_norm": 0.35934649357906984,
+      "learning_rate": 5.7688313003112506e-08,
+      "loss": 0.0374,
+      "step": 4313
+    },
+    {
+      "epoch": 4.8451495156535165,
+      "grad_norm": 0.3536391767479085,
+      "learning_rate": 5.685001739684448e-08,
+      "loss": 0.037,
+      "step": 4314
+    },
+    {
+      "epoch": 4.846272637933455,
+      "grad_norm": 0.41892157440986844,
+      "learning_rate": 5.6017839920180506e-08,
+      "loss": 0.0332,
+      "step": 4315
+    },
+    {
+      "epoch": 4.847395760213393,
+      "grad_norm": 0.362207049616821,
+      "learning_rate": 5.519178108516765e-08,
+      "loss": 0.0358,
+      "step": 4316
+    },
+    {
+      "epoch": 4.848518882493331,
+      "grad_norm": 0.339492102312975,
+      "learning_rate": 5.437184140009044e-08,
+      "loss": 0.0309,
+      "step": 4317
+    },
+    {
+      "epoch": 4.84964200477327,
+      "grad_norm": 0.3487122546387408,
+      "learning_rate": 5.355802136946531e-08,
+      "loss": 0.0345,
+      "step": 4318
+    },
+    {
+      "epoch": 4.850765127053208,
+      "grad_norm": 0.3793793262602845,
+      "learning_rate": 5.2750321494046133e-08,
+      "loss": 0.0381,
+      "step": 4319
+    },
+    {
+      "epoch": 4.851888249333146,
+      "grad_norm": 0.36623706813318585,
+      "learning_rate": 5.1948742270817584e-08,
+      "loss": 0.0346,
+      "step": 4320
+    },
+    {
+      "epoch": 4.853011371613085,
+      "grad_norm": 0.3730438215325291,
+      "learning_rate": 5.1153284193001803e-08,
+      "loss": 0.0391,
+      "step": 4321
+    },
+    {
+      "epoch": 4.854134493893023,
+      "grad_norm": 0.31906053615524266,
+      "learning_rate": 5.036394775005282e-08,
+      "loss": 0.0305,
+      "step": 4322
+    },
+    {
+      "epoch": 4.855257616172961,
+      "grad_norm": 0.3256027247837537,
+      "learning_rate": 4.958073342765768e-08,
+      "loss": 0.0334,
+      "step": 4323
+    },
+    {
+      "epoch": 4.856380738452899,
+      "grad_norm": 0.3507274265615372,
+      "learning_rate": 4.880364170773533e-08,
+      "loss": 0.0349,
+      "step": 4324
+    },
+    {
+      "epoch": 4.857503860732837,
+      "grad_norm": 0.3429380538021881,
+      "learning_rate": 4.803267306844106e-08,
+      "loss": 0.0323,
+      "step": 4325
+    },
+    {
+      "epoch": 4.858626983012775,
+      "grad_norm": 0.3537057379449312,
+      "learning_rate": 4.726782798415985e-08,
+      "loss": 0.0374,
+      "step": 4326
+    },
+    {
+      "epoch": 4.859750105292714,
+      "grad_norm": 0.31967178022973664,
+      "learning_rate": 4.650910692550858e-08,
+      "loss": 0.0312,
+      "step": 4327
+    },
+    {
+      "epoch": 4.860873227572652,
+      "grad_norm": 0.3465360661382547,
+      "learning_rate": 4.5756510359337145e-08,
+      "loss": 0.0383,
+      "step": 4328
+    },
+    {
+      "epoch": 4.86199634985259,
+      "grad_norm": 0.3480958307636944,
+      "learning_rate": 4.501003874872623e-08,
+      "loss": 0.0361,
+      "step": 4329
+    },
+    {
+      "epoch": 4.863119472132528,
+      "grad_norm": 0.351546101719163,
+      "learning_rate": 4.426969255298841e-08,
+      "loss": 0.0365,
+      "step": 4330
+    },
+    {
+      "epoch": 4.864242594412467,
+      "grad_norm": 0.33709962673459676,
+      "learning_rate": 4.3535472227667075e-08,
+      "loss": 0.031,
+      "step": 4331
+    },
+    {
+      "epoch": 4.865365716692405,
+      "grad_norm": 0.3592936899620035,
+      "learning_rate": 4.280737822453529e-08,
+      "loss": 0.0402,
+      "step": 4332
+    },
+    {
+      "epoch": 4.866488838972343,
+      "grad_norm": 0.3400368919719095,
+      "learning_rate": 4.208541099159691e-08,
+      "loss": 0.0324,
+      "step": 4333
+    },
+    {
+      "epoch": 4.8676119612522815,
+      "grad_norm": 0.3289295108551006,
+      "learning_rate": 4.136957097308769e-08,
+      "loss": 0.0316,
+      "step": 4334
+    },
+    {
+      "epoch": 4.868735083532219,
+      "grad_norm": 0.3564469735283429,
+      "learning_rate": 4.065985860947086e-08,
+      "loss": 0.0383,
+      "step": 4335
+    },
+    {
+      "epoch": 4.869858205812157,
+      "grad_norm": 0.3456958920658182,
+      "learning_rate": 3.9956274337441533e-08,
+      "loss": 0.0373,
+      "step": 4336
+    },
+    {
+      "epoch": 4.870981328092096,
+      "grad_norm": 0.3546463192740376,
+      "learning_rate": 3.92588185899212e-08,
+      "loss": 0.0351,
+      "step": 4337
+    },
+    {
+      "epoch": 4.872104450372034,
+      "grad_norm": 0.3634352198807142,
+      "learning_rate": 3.856749179606323e-08,
+      "loss": 0.0351,
+      "step": 4338
+    },
+    {
+      "epoch": 4.873227572651972,
+      "grad_norm": 0.34143099459117354,
+      "learning_rate": 3.7882294381247355e-08,
+      "loss": 0.0325,
+      "step": 4339
+    },
+    {
+      "epoch": 4.874350694931911,
+      "grad_norm": 0.3445137921102463,
+      "learning_rate": 3.72032267670841e-08,
+      "loss": 0.0342,
+      "step": 4340
+    },
+    {
+      "epoch": 4.875473817211849,
+      "grad_norm": 0.34957356832593245,
+      "learning_rate": 3.6530289371411453e-08,
+      "loss": 0.0335,
+      "step": 4341
+    },
+    {
+      "epoch": 4.876596939491787,
+      "grad_norm": 0.3559100988101287,
+      "learning_rate": 3.586348260829486e-08,
+      "loss": 0.0341,
+      "step": 4342
+    },
+    {
+      "epoch": 4.877720061771726,
+      "grad_norm": 0.3582325578381939,
+      "learning_rate": 3.520280688802724e-08,
+      "loss": 0.0389,
+      "step": 4343
+    },
+    {
+      "epoch": 4.8788431840516635,
+      "grad_norm": 0.3434812907253102,
+      "learning_rate": 3.4548262617131176e-08,
+      "loss": 0.0313,
+      "step": 4344
+    },
+    {
+      "epoch": 4.8799663063316014,
+      "grad_norm": 0.32607848224810887,
+      "learning_rate": 3.38998501983534e-08,
+      "loss": 0.0306,
+      "step": 4345
+    },
+    {
+      "epoch": 4.88108942861154,
+      "grad_norm": 0.3701550768586464,
+      "learning_rate": 3.3257570030670316e-08,
+      "loss": 0.0361,
+      "step": 4346
+    },
+    {
+      "epoch": 4.882212550891478,
+      "grad_norm": 0.36030791267577544,
+      "learning_rate": 3.2621422509282464e-08,
+      "loss": 0.0366,
+      "step": 4347
+    },
+    {
+      "epoch": 4.883335673171416,
+      "grad_norm": 0.3451337361364224,
+      "learning_rate": 3.199140802562006e-08,
+      "loss": 0.0314,
+      "step": 4348
+    },
+    {
+      "epoch": 4.884458795451355,
+      "grad_norm": 0.3452767709477446,
+      "learning_rate": 3.1367526967336356e-08,
+      "loss": 0.0354,
+      "step": 4349
+    },
+    {
+      "epoch": 4.885581917731293,
+      "grad_norm": 0.3547638758596286,
+      "learning_rate": 3.0749779718314273e-08,
+      "loss": 0.0353,
+      "step": 4350
+    },
+    {
+      "epoch": 4.886705040011231,
+      "grad_norm": 0.3506982695342505,
+      "learning_rate": 3.013816665865976e-08,
+      "loss": 0.0344,
+      "step": 4351
+    },
+    {
+      "epoch": 4.88782816229117,
+      "grad_norm": 0.3737683890517672,
+      "learning_rate": 2.9532688164704005e-08,
+      "loss": 0.0378,
+      "step": 4352
+    },
+    {
+      "epoch": 4.888951284571108,
+      "grad_norm": 0.35014209109341093,
+      "learning_rate": 2.8933344609004545e-08,
+      "loss": 0.0367,
+      "step": 4353
+    },
+    {
+      "epoch": 4.8900744068510456,
+      "grad_norm": 0.3264875513879221,
+      "learning_rate": 2.834013636034527e-08,
+      "loss": 0.0328,
+      "step": 4354
+    },
+    {
+      "epoch": 4.891197529130984,
+      "grad_norm": 0.32367274530213846,
+      "learning_rate": 2.7753063783734212e-08,
+      "loss": 0.0308,
+      "step": 4355
+    },
+    {
+      "epoch": 4.892320651410922,
+      "grad_norm": 0.32601616170265874,
+      "learning_rate": 2.7172127240401302e-08,
+      "loss": 0.0306,
+      "step": 4356
+    },
+    {
+      "epoch": 4.89344377369086,
+      "grad_norm": 0.3514262599500851,
+      "learning_rate": 2.6597327087805048e-08,
+      "loss": 0.034,
+      "step": 4357
+    },
+    {
+      "epoch": 4.894566895970799,
+      "grad_norm": 0.34413690056428353,
+      "learning_rate": 2.6028663679625865e-08,
+      "loss": 0.0358,
+      "step": 4358
+    },
+    {
+      "epoch": 4.895690018250737,
+      "grad_norm": 0.33388792579998283,
+      "learning_rate": 2.5466137365768307e-08,
+      "loss": 0.0331,
+      "step": 4359
+    },
+    {
+      "epoch": 4.896813140530675,
+      "grad_norm": 0.3558627099214127,
+      "learning_rate": 2.4909748492362162e-08,
+      "loss": 0.0335,
+      "step": 4360
+    },
+    {
+      "epoch": 4.897936262810614,
+      "grad_norm": 0.3568173099350905,
+      "learning_rate": 2.4359497401758026e-08,
+      "loss": 0.0335,
+      "step": 4361
+    },
+    {
+      "epoch": 4.899059385090552,
+      "grad_norm": 0.3576774734027632,
+      "learning_rate": 2.3815384432531728e-08,
+      "loss": 0.0352,
+      "step": 4362
+    },
+    {
+      "epoch": 4.90018250737049,
+      "grad_norm": 0.35519782847619524,
+      "learning_rate": 2.327740991948213e-08,
+      "loss": 0.0346,
+      "step": 4363
+    },
+    {
+      "epoch": 4.9013056296504285,
+      "grad_norm": 0.3450872223409437,
+      "learning_rate": 2.2745574193632215e-08,
+      "loss": 0.035,
+      "step": 4364
+    },
+    {
+      "epoch": 4.902428751930366,
+      "grad_norm": 0.32911920269693273,
+      "learning_rate": 2.2219877582224657e-08,
+      "loss": 0.0307,
+      "step": 4365
+    },
+    {
+      "epoch": 4.903551874210304,
+      "grad_norm": 0.33655879907579667,
+      "learning_rate": 2.170032040872627e-08,
+      "loss": 0.0317,
+      "step": 4366
+    },
+    {
+      "epoch": 4.904674996490243,
+      "grad_norm": 0.34894129051496664,
+      "learning_rate": 2.1186902992827995e-08,
+      "loss": 0.0356,
+      "step": 4367
+    },
+    {
+      "epoch": 4.905798118770181,
+      "grad_norm": 0.3720141793260485,
+      "learning_rate": 2.067962565043935e-08,
+      "loss": 0.0343,
+      "step": 4368
+    },
+    {
+      "epoch": 4.906921241050119,
+      "grad_norm": 0.34874029206243984,
+      "learning_rate": 2.0178488693695096e-08,
+      "loss": 0.0361,
+      "step": 4369
+    },
+    {
+      "epoch": 4.908044363330058,
+      "grad_norm": 0.3474003843287452,
+      "learning_rate": 1.968349243094969e-08,
+      "loss": 0.0334,
+      "step": 4370
+    },
+    {
+      "epoch": 4.909167485609996,
+      "grad_norm": 0.3423160120197686,
+      "learning_rate": 1.9194637166780606e-08,
+      "loss": 0.0337,
+      "step": 4371
+    },
+    {
+      "epoch": 4.910290607889934,
+      "grad_norm": 0.33591485159670115,
+      "learning_rate": 1.8711923201983895e-08,
+      "loss": 0.0343,
+      "step": 4372
+    },
+    {
+      "epoch": 4.911413730169873,
+      "grad_norm": 0.33191680893555436,
+      "learning_rate": 1.8235350833579745e-08,
+      "loss": 0.0321,
+      "step": 4373
+    },
+    {
+      "epoch": 4.9125368524498105,
+      "grad_norm": 0.3316935624017572,
+      "learning_rate": 1.7764920354809146e-08,
+      "loss": 0.0337,
+      "step": 4374
+    },
+    {
+      "epoch": 4.9136599747297485,
+      "grad_norm": 0.35967697736265597,
+      "learning_rate": 1.730063205513277e-08,
+      "loss": 0.0365,
+      "step": 4375
+    },
+    {
+      "epoch": 4.914783097009687,
+      "grad_norm": 0.359094588149611,
+      "learning_rate": 1.6842486220232102e-08,
+      "loss": 0.0321,
+      "step": 4376
+    },
+    {
+      "epoch": 4.915906219289625,
+      "grad_norm": 0.3604389646619046,
+      "learning_rate": 1.6390483132009415e-08,
+      "loss": 0.0369,
+      "step": 4377
+    },
+    {
+      "epoch": 4.917029341569563,
+      "grad_norm": 0.3457113375936627,
+      "learning_rate": 1.5944623068586683e-08,
+      "loss": 0.0339,
+      "step": 4378
+    },
+    {
+      "epoch": 4.918152463849502,
+      "grad_norm": 0.31938400735132355,
+      "learning_rate": 1.5504906304306677e-08,
+      "loss": 0.0317,
+      "step": 4379
+    },
+    {
+      "epoch": 4.91927558612944,
+      "grad_norm": 0.340077310318995,
+      "learning_rate": 1.5071333109732966e-08,
+      "loss": 0.0361,
+      "step": 4380
+    },
+    {
+      "epoch": 4.920398708409378,
+      "grad_norm": 0.35535503166517346,
+      "learning_rate": 1.4643903751647703e-08,
+      "loss": 0.0361,
+      "step": 4381
+    },
+    {
+      "epoch": 4.921521830689317,
+      "grad_norm": 0.3537974935188405,
+      "learning_rate": 1.422261849305162e-08,
+      "loss": 0.0342,
+      "step": 4382
+    },
+    {
+      "epoch": 4.922644952969255,
+      "grad_norm": 0.32843491322571805,
+      "learning_rate": 1.3807477593166252e-08,
+      "loss": 0.0319,
+      "step": 4383
+    },
+    {
+      "epoch": 4.923768075249193,
+      "grad_norm": 0.3642690465675649,
+      "learning_rate": 1.339848130743393e-08,
+      "loss": 0.0363,
+      "step": 4384
+    },
+    {
+      "epoch": 4.924891197529131,
+      "grad_norm": 0.3431037322105767,
+      "learning_rate": 1.299562988751335e-08,
+      "loss": 0.0342,
+      "step": 4385
+    },
+    {
+      "epoch": 4.926014319809069,
+      "grad_norm": 0.37231643890134525,
+      "learning_rate": 1.2598923581284006e-08,
+      "loss": 0.0341,
+      "step": 4386
+    },
+    {
+      "epoch": 4.927137442089007,
+      "grad_norm": 0.3500700062527457,
+      "learning_rate": 1.2208362632842863e-08,
+      "loss": 0.0363,
+      "step": 4387
+    },
+    {
+      "epoch": 4.928260564368946,
+      "grad_norm": 0.33425871580503613,
+      "learning_rate": 1.1823947282506576e-08,
+      "loss": 0.0312,
+      "step": 4388
+    },
+    {
+      "epoch": 4.929383686648884,
+      "grad_norm": 0.3633980633957219,
+      "learning_rate": 1.144567776681149e-08,
+      "loss": 0.0376,
+      "step": 4389
+    },
+    {
+      "epoch": 4.930506808928822,
+      "grad_norm": 0.3349073890426962,
+      "learning_rate": 1.1073554318509206e-08,
+      "loss": 0.0326,
+      "step": 4390
+    },
+    {
+      "epoch": 4.931629931208761,
+      "grad_norm": 0.34098481332204433,
+      "learning_rate": 1.0707577166572114e-08,
+      "loss": 0.0341,
+      "step": 4391
+    },
+    {
+      "epoch": 4.932753053488699,
+      "grad_norm": 0.3285118951852116,
+      "learning_rate": 1.0347746536191195e-08,
+      "loss": 0.0329,
+      "step": 4392
+    },
+    {
+      "epoch": 4.933876175768637,
+      "grad_norm": 0.3451360635835923,
+      "learning_rate": 9.994062648771563e-09,
+      "loss": 0.0345,
+      "step": 4393
+    },
+    {
+      "epoch": 4.9349992980485755,
+      "grad_norm": 0.3451649817059978,
+      "learning_rate": 9.646525721940247e-09,
+      "loss": 0.0339,
+      "step": 4394
+    },
+    {
+      "epoch": 4.9361224203285134,
+      "grad_norm": 0.3329314247830526,
+      "learning_rate": 9.305135969541746e-09,
+      "loss": 0.0331,
+      "step": 4395
+    },
+    {
+      "epoch": 4.937245542608451,
+      "grad_norm": 0.3703275176605861,
+      "learning_rate": 8.969893601634694e-09,
+      "loss": 0.0321,
+      "step": 4396
+    },
+    {
+      "epoch": 4.93836866488839,
+      "grad_norm": 0.38004031161852586,
+      "learning_rate": 8.64079882449853e-09,
+      "loss": 0.0343,
+      "step": 4397
+    },
+    {
+      "epoch": 4.939491787168328,
+      "grad_norm": 0.36158113782294066,
+      "learning_rate": 8.317851840629055e-09,
+      "loss": 0.0324,
+      "step": 4398
+    },
+    {
+      "epoch": 4.940614909448266,
+      "grad_norm": 0.35802038640100103,
+      "learning_rate": 8.001052848739532e-09,
+      "loss": 0.0355,
+      "step": 4399
+    },
+    {
+      "epoch": 4.941738031728204,
+      "grad_norm": 0.3534681414083675,
+      "learning_rate": 7.690402043758482e-09,
+      "loss": 0.0342,
+      "step": 4400
+    },
+    {
+      "epoch": 4.942861154008143,
+      "grad_norm": 0.3409750506973528,
+      "learning_rate": 7.385899616833003e-09,
+      "loss": 0.0322,
+      "step": 4401
+    },
+    {
+      "epoch": 4.943984276288081,
+      "grad_norm": 0.35696111455006185,
+      "learning_rate": 7.087545755327663e-09,
+      "loss": 0.0343,
+      "step": 4402
+    },
+    {
+      "epoch": 4.945107398568019,
+      "grad_norm": 0.3628732180685351,
+      "learning_rate": 6.795340642823389e-09,
+      "loss": 0.033,
+      "step": 4403
+    },
+    {
+      "epoch": 4.946230520847958,
+      "grad_norm": 0.3561531152392209,
+      "learning_rate": 6.50928445911525e-09,
+      "loss": 0.0347,
+      "step": 4404
+    },
+    {
+      "epoch": 4.9473536431278955,
+      "grad_norm": 0.3604397825246946,
+      "learning_rate": 6.229377380218005e-09,
+      "loss": 0.0349,
+      "step": 4405
+    },
+    {
+      "epoch": 4.948476765407833,
+      "grad_norm": 0.3515361148095995,
+      "learning_rate": 5.95561957836055e-09,
+      "loss": 0.0349,
+      "step": 4406
+    },
+    {
+      "epoch": 4.949599887687772,
+      "grad_norm": 0.34835719474984594,
+      "learning_rate": 5.688011221991474e-09,
+      "loss": 0.034,
+      "step": 4407
+    },
+    {
+      "epoch": 4.95072300996771,
+      "grad_norm": 0.36776092832286844,
+      "learning_rate": 5.426552475770175e-09,
+      "loss": 0.0405,
+      "step": 4408
+    },
+    {
+      "epoch": 4.951846132247648,
+      "grad_norm": 0.33964356120947575,
+      "learning_rate": 5.1712435005768504e-09,
+      "loss": 0.0314,
+      "step": 4409
+    },
+    {
+      "epoch": 4.952969254527587,
+      "grad_norm": 0.3576271748655273,
+      "learning_rate": 4.922084453505838e-09,
+      "loss": 0.0348,
+      "step": 4410
+    },
+    {
+      "epoch": 4.954092376807525,
+      "grad_norm": 0.3429879169889465,
+      "learning_rate": 4.679075487866725e-09,
+      "loss": 0.0337,
+      "step": 4411
+    },
+    {
+      "epoch": 4.955215499087463,
+      "grad_norm": 0.3745978618879177,
+      "learning_rate": 4.4422167531865675e-09,
+      "loss": 0.0368,
+      "step": 4412
+    },
+    {
+      "epoch": 4.956338621367402,
+      "grad_norm": 0.32986700838548433,
+      "learning_rate": 4.211508395206565e-09,
+      "loss": 0.0327,
+      "step": 4413
+    },
+    {
+      "epoch": 4.95746174364734,
+      "grad_norm": 0.33311799847039564,
+      "learning_rate": 3.986950555883162e-09,
+      "loss": 0.0348,
+      "step": 4414
+    },
+    {
+      "epoch": 4.9585848659272775,
+      "grad_norm": 0.3851607304135312,
+      "learning_rate": 3.768543373391387e-09,
+      "loss": 0.0424,
+      "step": 4415
+    },
+    {
+      "epoch": 4.959707988207216,
+      "grad_norm": 0.31347433372100486,
+      "learning_rate": 3.5562869821181843e-09,
+      "loss": 0.0315,
+      "step": 4416
+    },
+    {
+      "epoch": 4.960831110487154,
+      "grad_norm": 0.34107319972689115,
+      "learning_rate": 3.3501815126668613e-09,
+      "loss": 0.0339,
+      "step": 4417
+    },
+    {
+      "epoch": 4.961954232767092,
+      "grad_norm": 0.33193990363553355,
+      "learning_rate": 3.150227091857083e-09,
+      "loss": 0.0329,
+      "step": 4418
+    },
+    {
+      "epoch": 4.963077355047031,
+      "grad_norm": 0.3715468555033886,
+      "learning_rate": 2.9564238427237657e-09,
+      "loss": 0.038,
+      "step": 4419
+    },
+    {
+      "epoch": 4.964200477326969,
+      "grad_norm": 0.3372262605486213,
+      "learning_rate": 2.7687718845148538e-09,
+      "loss": 0.0314,
+      "step": 4420
+    },
+    {
+      "epoch": 4.965323599606907,
+      "grad_norm": 0.350019697957,
+      "learning_rate": 2.587271332694652e-09,
+      "loss": 0.0347,
+      "step": 4421
+    },
+    {
+      "epoch": 4.966446721886846,
+      "grad_norm": 0.3446161586717599,
+      "learning_rate": 2.411922298943825e-09,
+      "loss": 0.0344,
+      "step": 4422
+    },
+    {
+      "epoch": 4.967569844166784,
+      "grad_norm": 0.319098731802588,
+      "learning_rate": 2.242724891156067e-09,
+      "loss": 0.0323,
+      "step": 4423
+    },
+    {
+      "epoch": 4.968692966446722,
+      "grad_norm": 0.3801202120943219,
+      "learning_rate": 2.079679213439212e-09,
+      "loss": 0.0372,
+      "step": 4424
+    },
+    {
+      "epoch": 4.9698160887266605,
+      "grad_norm": 0.33750070670739035,
+      "learning_rate": 1.9227853661174524e-09,
+      "loss": 0.0352,
+      "step": 4425
+    },
+    {
+      "epoch": 4.970939211006598,
+      "grad_norm": 0.3519190781169123,
+      "learning_rate": 1.7720434457302315e-09,
+      "loss": 0.0373,
+      "step": 4426
+    },
+    {
+      "epoch": 4.972062333286536,
+      "grad_norm": 0.3437633504102528,
+      "learning_rate": 1.6274535450311324e-09,
+      "loss": 0.0325,
+      "step": 4427
+    },
+    {
+      "epoch": 4.973185455566475,
+      "grad_norm": 0.3677752690761468,
+      "learning_rate": 1.4890157529856563e-09,
+      "loss": 0.0402,
+      "step": 4428
+    },
+    {
+      "epoch": 4.974308577846413,
+      "grad_norm": 0.3683625506073574,
+      "learning_rate": 1.3567301547778856e-09,
+      "loss": 0.037,
+      "step": 4429
+    },
+    {
+      "epoch": 4.975431700126351,
+      "grad_norm": 0.3560600240146814,
+      "learning_rate": 1.230596831804931e-09,
+      "loss": 0.0358,
+      "step": 4430
+    },
+    {
+      "epoch": 4.976554822406289,
+      "grad_norm": 0.3625253632661801,
+      "learning_rate": 1.1106158616758235e-09,
+      "loss": 0.0366,
+      "step": 4431
+    },
+    {
+      "epoch": 4.977677944686228,
+      "grad_norm": 0.3444846000256938,
+      "learning_rate": 9.96787318218173e-10,
+      "loss": 0.0324,
+      "step": 4432
+    },
+    {
+      "epoch": 4.978801066966166,
+      "grad_norm": 0.35833357272901073,
+      "learning_rate": 8.891112714726203e-10,
+      "loss": 0.035,
+      "step": 4433
+    },
+    {
+      "epoch": 4.979924189246104,
+      "grad_norm": 0.3425329647959898,
+      "learning_rate": 7.875877876906135e-10,
+      "loss": 0.0331,
+      "step": 4434
+    },
+    {
+      "epoch": 4.9810473115260425,
+      "grad_norm": 0.35679200669540573,
+      "learning_rate": 6.922169293421821e-10,
+      "loss": 0.0364,
+      "step": 4435
+    },
+    {
+      "epoch": 4.9821704338059805,
+      "grad_norm": 0.3259794017034792,
+      "learning_rate": 6.029987551103844e-10,
+      "loss": 0.0319,
+      "step": 4436
+    },
+    {
+      "epoch": 4.983293556085918,
+      "grad_norm": 0.3480641727719168,
+      "learning_rate": 5.199333198924183e-10,
+      "loss": 0.0336,
+      "step": 4437
+    },
+    {
+      "epoch": 4.984416678365857,
+      "grad_norm": 0.348847347957358,
+      "learning_rate": 4.4302067479851105e-10,
+      "loss": 0.0321,
+      "step": 4438
+    },
+    {
+      "epoch": 4.985539800645795,
+      "grad_norm": 0.3783464710272088,
+      "learning_rate": 3.7226086715413945e-10,
+      "loss": 0.0381,
+      "step": 4439
+    },
+    {
+      "epoch": 4.986662922925733,
+      "grad_norm": 0.35344733158741803,
+      "learning_rate": 3.076539404989198e-10,
+      "loss": 0.0309,
+      "step": 4440
+    },
+    {
+      "epoch": 4.987786045205672,
+      "grad_norm": 0.33442902082308484,
+      "learning_rate": 2.4919993458549783e-10,
+      "loss": 0.0343,
+      "step": 4441
+    },
+    {
+      "epoch": 4.98890916748561,
+      "grad_norm": 0.35360484233382744,
+      "learning_rate": 1.968988853806586e-10,
+      "loss": 0.036,
+      "step": 4442
+    },
+    {
+      "epoch": 4.990032289765548,
+      "grad_norm": 0.3421628593665695,
+      "learning_rate": 1.5075082506865734e-10,
+      "loss": 0.0356,
+      "step": 4443
+    },
+    {
+      "epoch": 4.991155412045487,
+      "grad_norm": 0.33672336609006215,
+      "learning_rate": 1.1075578204233772e-10,
+      "loss": 0.0337,
+      "step": 4444
+    },
+    {
+      "epoch": 4.992278534325425,
+      "grad_norm": 0.3641808930391885,
+      "learning_rate": 7.691378091090329e-11,
+      "loss": 0.0328,
+      "step": 4445
+    },
+    {
+      "epoch": 4.9934016566053625,
+      "grad_norm": 0.32873267052415456,
+      "learning_rate": 4.9224842499917546e-11,
+      "loss": 0.0315,
+      "step": 4446
+    },
+    {
+      "epoch": 4.994524778885301,
+      "grad_norm": 0.36446284662411277,
+      "learning_rate": 2.768898384464258e-11,
+      "loss": 0.035,
+      "step": 4447
+    },
+    {
+      "epoch": 4.995647901165239,
+      "grad_norm": 0.33359437371199696,
+      "learning_rate": 1.2306218196700414e-11,
+      "loss": 0.031,
+      "step": 4448
+    },
+    {
+      "epoch": 4.996771023445177,
+      "grad_norm": 0.3510829003384089,
+      "learning_rate": 3.0765550229627793e-12,
+      "loss": 0.0348,
+      "step": 4449
+    },
+    {
+      "epoch": 4.997894145725116,
+      "grad_norm": 0.3451047076893161,
+      "learning_rate": 0.0,
+      "loss": 0.0326,
+      "step": 4450
+    },
+    {
+      "epoch": 4.997894145725116,
+      "step": 4450,
+      "total_flos": 2.0283773111783916e+18,
+      "train_loss": 0.1949592823937033,
+      "train_runtime": 25560.7624,
+      "train_samples_per_second": 22.291,
+      "train_steps_per_second": 0.174
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 4450,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 5,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.0283773111783916e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}