diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,8434 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9573653306101608,
+  "eval_steps": 500,
+  "global_step": 30000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.000797804442175134,
+      "grad_norm": 1.1620235443115234,
+      "learning_rate": 3.125e-06,
+      "loss": 1.0717,
+      "step": 25
+    },
+    {
+      "epoch": 0.001595608884350268,
+      "grad_norm": 0.1761351078748703,
+      "learning_rate": 6.25e-06,
+      "loss": 1.0035,
+      "step": 50
+    },
+    {
+      "epoch": 0.0023934133265254023,
+      "grad_norm": 0.7293746471405029,
+      "learning_rate": 9.375000000000001e-06,
+      "loss": 1.0267,
+      "step": 75
+    },
+    {
+      "epoch": 0.003191217768700536,
+      "grad_norm": 0.5050331950187683,
+      "learning_rate": 1.25e-05,
+      "loss": 0.8625,
+      "step": 100
+    },
+    {
+      "epoch": 0.00398902221087567,
+      "grad_norm": 0.5066962242126465,
+      "learning_rate": 1.5625e-05,
+      "loss": 0.7529,
+      "step": 125
+    },
+    {
+      "epoch": 0.0047868266530508045,
+      "grad_norm": 0.7367194294929504,
+      "learning_rate": 1.8750000000000002e-05,
+      "loss": 0.765,
+      "step": 150
+    },
+    {
+      "epoch": 0.005584631095225938,
+      "grad_norm": 0.5524747967720032,
+      "learning_rate": 2.1875e-05,
+      "loss": 0.7062,
+      "step": 175
+    },
+    {
+      "epoch": 0.006382435537401072,
+      "grad_norm": 0.6297128200531006,
+      "learning_rate": 2.5e-05,
+      "loss": 0.7283,
+      "step": 200
+    },
+    {
+      "epoch": 0.007180239979576206,
+      "grad_norm": 0.3561699092388153,
+      "learning_rate": 2.4979926772867422e-05,
+      "loss": 0.6711,
+      "step": 225
+    },
+    {
+      "epoch": 0.00797804442175134,
+      "grad_norm": 0.6796361207962036,
+      "learning_rate": 2.4959853545734843e-05,
+      "loss": 0.6566,
+      "step": 250
+    },
+    {
+      "epoch": 0.008775848863926475,
+      "grad_norm": 0.9527813792228699,
+      "learning_rate": 2.493978031860226e-05,
+      "loss": 0.6333,
+      "step": 275
+    },
+    {
+      "epoch": 0.009573653306101609,
+      "grad_norm": 0.5738817453384399,
+      "learning_rate": 2.4919707091469682e-05,
+      "loss": 0.6126,
+      "step": 300
+    },
+    {
+      "epoch": 0.010371457748276743,
+      "grad_norm": 0.8721635937690735,
+      "learning_rate": 2.4899633864337103e-05,
+      "loss": 0.5869,
+      "step": 325
+    },
+    {
+      "epoch": 0.011169262190451876,
+      "grad_norm": 0.8487357497215271,
+      "learning_rate": 2.4879560637204525e-05,
+      "loss": 0.6693,
+      "step": 350
+    },
+    {
+      "epoch": 0.01196706663262701,
+      "grad_norm": 0.8497329950332642,
+      "learning_rate": 2.4859487410071942e-05,
+      "loss": 0.6331,
+      "step": 375
+    },
+    {
+      "epoch": 0.012764871074802144,
+      "grad_norm": 0.7911163568496704,
+      "learning_rate": 2.4839414182939363e-05,
+      "loss": 0.6252,
+      "step": 400
+    },
+    {
+      "epoch": 0.013562675516977278,
+      "grad_norm": 1.0417559146881104,
+      "learning_rate": 2.4819340955806785e-05,
+      "loss": 0.654,
+      "step": 425
+    },
+    {
+      "epoch": 0.014360479959152413,
+      "grad_norm": 0.7947675585746765,
+      "learning_rate": 2.4799267728674206e-05,
+      "loss": 0.6265,
+      "step": 450
+    },
+    {
+      "epoch": 0.015158284401327547,
+      "grad_norm": 0.654082179069519,
+      "learning_rate": 2.4779194501541623e-05,
+      "loss": 0.5948,
+      "step": 475
+    },
+    {
+      "epoch": 0.01595608884350268,
+      "grad_norm": 0.5939388275146484,
+      "learning_rate": 2.4759121274409045e-05,
+      "loss": 0.5704,
+      "step": 500
+    },
+    {
+      "epoch": 0.016753893285677814,
+      "grad_norm": 0.5853065848350525,
+      "learning_rate": 2.4739048047276466e-05,
+      "loss": 0.6123,
+      "step": 525
+    },
+    {
+      "epoch": 0.01755169772785295,
+      "grad_norm": 0.924404501914978,
+      "learning_rate": 2.4718974820143887e-05,
+      "loss": 0.578,
+      "step": 550
+    },
+    {
+      "epoch": 0.018349502170028082,
+      "grad_norm": 0.6280034184455872,
+      "learning_rate": 2.4698901593011305e-05,
+      "loss": 0.612,
+      "step": 575
+    },
+    {
+      "epoch": 0.019147306612203218,
+      "grad_norm": 1.4167883396148682,
+      "learning_rate": 2.4678828365878726e-05,
+      "loss": 0.6074,
+      "step": 600
+    },
+    {
+      "epoch": 0.01994511105437835,
+      "grad_norm": 1.254233717918396,
+      "learning_rate": 2.4658755138746147e-05,
+      "loss": 0.6437,
+      "step": 625
+    },
+    {
+      "epoch": 0.020742915496553486,
+      "grad_norm": 0.6111056208610535,
+      "learning_rate": 2.4638681911613568e-05,
+      "loss": 0.6093,
+      "step": 650
+    },
+    {
+      "epoch": 0.02154071993872862,
+      "grad_norm": 0.7576286792755127,
+      "learning_rate": 2.4618608684480986e-05,
+      "loss": 0.5317,
+      "step": 675
+    },
+    {
+      "epoch": 0.02233852438090375,
+      "grad_norm": 1.1031715869903564,
+      "learning_rate": 2.459853545734841e-05,
+      "loss": 0.6359,
+      "step": 700
+    },
+    {
+      "epoch": 0.023136328823078887,
+      "grad_norm": 0.9123847484588623,
+      "learning_rate": 2.4578462230215828e-05,
+      "loss": 0.5288,
+      "step": 725
+    },
+    {
+      "epoch": 0.02393413326525402,
+      "grad_norm": 1.0893287658691406,
+      "learning_rate": 2.455838900308325e-05,
+      "loss": 0.5866,
+      "step": 750
+    },
+    {
+      "epoch": 0.024731937707429156,
+      "grad_norm": 0.6093726754188538,
+      "learning_rate": 2.4538315775950667e-05,
+      "loss": 0.524,
+      "step": 775
+    },
+    {
+      "epoch": 0.02552974214960429,
+      "grad_norm": 0.8649279475212097,
+      "learning_rate": 2.451824254881809e-05,
+      "loss": 0.6248,
+      "step": 800
+    },
+    {
+      "epoch": 0.026327546591779424,
+      "grad_norm": 0.7364605069160461,
+      "learning_rate": 2.449816932168551e-05,
+      "loss": 0.5878,
+      "step": 825
+    },
+    {
+      "epoch": 0.027125351033954557,
+      "grad_norm": 1.0632132291793823,
+      "learning_rate": 2.447809609455293e-05,
+      "loss": 0.6738,
+      "step": 850
+    },
+    {
+      "epoch": 0.027923155476129693,
+      "grad_norm": 0.5434523224830627,
+      "learning_rate": 2.4458022867420348e-05,
+      "loss": 0.6436,
+      "step": 875
+    },
+    {
+      "epoch": 0.028720959918304825,
+      "grad_norm": 1.3886839151382446,
+      "learning_rate": 2.4437949640287773e-05,
+      "loss": 0.6421,
+      "step": 900
+    },
+    {
+      "epoch": 0.029518764360479958,
+      "grad_norm": 0.66085284948349,
+      "learning_rate": 2.441787641315519e-05,
+      "loss": 0.5792,
+      "step": 925
+    },
+    {
+      "epoch": 0.030316568802655094,
+      "grad_norm": 0.39016640186309814,
+      "learning_rate": 2.439780318602261e-05,
+      "loss": 0.5376,
+      "step": 950
+    },
+    {
+      "epoch": 0.031114373244830226,
+      "grad_norm": 1.0862053632736206,
+      "learning_rate": 2.4377729958890033e-05,
+      "loss": 0.4904,
+      "step": 975
+    },
+    {
+      "epoch": 0.03191217768700536,
+      "grad_norm": 0.7027643918991089,
+      "learning_rate": 2.4357656731757454e-05,
+      "loss": 0.6461,
+      "step": 1000
+    },
+    {
+      "epoch": 0.0327099821291805,
+      "grad_norm": 1.035848617553711,
+      "learning_rate": 2.433758350462487e-05,
+      "loss": 0.6061,
+      "step": 1025
+    },
+    {
+      "epoch": 0.03350778657135563,
+      "grad_norm": 0.9431428909301758,
+      "learning_rate": 2.4317510277492293e-05,
+      "loss": 0.585,
+      "step": 1050
+    },
+    {
+      "epoch": 0.03430559101353076,
+      "grad_norm": 1.2081215381622314,
+      "learning_rate": 2.4297437050359714e-05,
+      "loss": 0.5293,
+      "step": 1075
+    },
+    {
+      "epoch": 0.0351033954557059,
+      "grad_norm": 0.6889535784721375,
+      "learning_rate": 2.4277363823227135e-05,
+      "loss": 0.5132,
+      "step": 1100
+    },
+    {
+      "epoch": 0.03590119989788103,
+      "grad_norm": 1.3305994272232056,
+      "learning_rate": 2.4257290596094553e-05,
+      "loss": 0.6028,
+      "step": 1125
+    },
+    {
+      "epoch": 0.036699004340056164,
+      "grad_norm": 0.4918406307697296,
+      "learning_rate": 2.4237217368961977e-05,
+      "loss": 0.6059,
+      "step": 1150
+    },
+    {
+      "epoch": 0.0374968087822313,
+      "grad_norm": 0.9576552510261536,
+      "learning_rate": 2.4217144141829395e-05,
+      "loss": 0.581,
+      "step": 1175
+    },
+    {
+      "epoch": 0.038294613224406436,
+      "grad_norm": 0.533026397228241,
+      "learning_rate": 2.4197070914696816e-05,
+      "loss": 0.5114,
+      "step": 1200
+    },
+    {
+      "epoch": 0.039092417666581565,
+      "grad_norm": 1.2196937799453735,
+      "learning_rate": 2.4176997687564234e-05,
+      "loss": 0.6211,
+      "step": 1225
+    },
+    {
+      "epoch": 0.0398902221087567,
+      "grad_norm": 1.5606813430786133,
+      "learning_rate": 2.4156924460431658e-05,
+      "loss": 0.5383,
+      "step": 1250
+    },
+    {
+      "epoch": 0.04068802655093184,
+      "grad_norm": 0.4488339424133301,
+      "learning_rate": 2.4136851233299076e-05,
+      "loss": 0.563,
+      "step": 1275
+    },
+    {
+      "epoch": 0.04148583099310697,
+      "grad_norm": 0.9547097086906433,
+      "learning_rate": 2.4116778006166497e-05,
+      "loss": 0.5983,
+      "step": 1300
+    },
+    {
+      "epoch": 0.0422836354352821,
+      "grad_norm": 0.40564727783203125,
+      "learning_rate": 2.4096704779033915e-05,
+      "loss": 0.5555,
+      "step": 1325
+    },
+    {
+      "epoch": 0.04308143987745724,
+      "grad_norm": 0.889912486076355,
+      "learning_rate": 2.407663155190134e-05,
+      "loss": 0.5673,
+      "step": 1350
+    },
+    {
+      "epoch": 0.043879244319632374,
+      "grad_norm": 1.0422492027282715,
+      "learning_rate": 2.4056558324768757e-05,
+      "loss": 0.5523,
+      "step": 1375
+    },
+    {
+      "epoch": 0.0446770487618075,
+      "grad_norm": 1.0268012285232544,
+      "learning_rate": 2.4036485097636178e-05,
+      "loss": 0.674,
+      "step": 1400
+    },
+    {
+      "epoch": 0.04547485320398264,
+      "grad_norm": 0.4538346230983734,
+      "learning_rate": 2.40164118705036e-05,
+      "loss": 0.5532,
+      "step": 1425
+    },
+    {
+      "epoch": 0.046272657646157775,
+      "grad_norm": 0.4324471652507782,
+      "learning_rate": 2.399633864337102e-05,
+      "loss": 0.5489,
+      "step": 1450
+    },
+    {
+      "epoch": 0.04707046208833291,
+      "grad_norm": 0.9547377824783325,
+      "learning_rate": 2.3976265416238438e-05,
+      "loss": 0.5658,
+      "step": 1475
+    },
+    {
+      "epoch": 0.04786826653050804,
+      "grad_norm": 0.3644287586212158,
+      "learning_rate": 2.395619218910586e-05,
+      "loss": 0.4947,
+      "step": 1500
+    },
+    {
+      "epoch": 0.048666070972683176,
+      "grad_norm": 0.9548143744468689,
+      "learning_rate": 2.393611896197328e-05,
+      "loss": 0.5477,
+      "step": 1525
+    },
+    {
+      "epoch": 0.04946387541485831,
+      "grad_norm": 0.7494068145751953,
+      "learning_rate": 2.3916045734840702e-05,
+      "loss": 0.4489,
+      "step": 1550
+    },
+    {
+      "epoch": 0.05026167985703344,
+      "grad_norm": 1.4067238569259644,
+      "learning_rate": 2.389597250770812e-05,
+      "loss": 0.5193,
+      "step": 1575
+    },
+    {
+      "epoch": 0.05105948429920858,
+      "grad_norm": 0.6029401421546936,
+      "learning_rate": 2.387589928057554e-05,
+      "loss": 0.4929,
+      "step": 1600
+    },
+    {
+      "epoch": 0.05185728874138371,
+      "grad_norm": 0.5118242502212524,
+      "learning_rate": 2.3855826053442962e-05,
+      "loss": 0.5729,
+      "step": 1625
+    },
+    {
+      "epoch": 0.05265509318355885,
+      "grad_norm": 0.3905925452709198,
+      "learning_rate": 2.3835752826310383e-05,
+      "loss": 0.5315,
+      "step": 1650
+    },
+    {
+      "epoch": 0.05345289762573398,
+      "grad_norm": 1.9033564329147339,
+      "learning_rate": 2.38156795991778e-05,
+      "loss": 0.573,
+      "step": 1675
+    },
+    {
+      "epoch": 0.054250702067909114,
+      "grad_norm": 0.6796665787696838,
+      "learning_rate": 2.3795606372045222e-05,
+      "loss": 0.5212,
+      "step": 1700
+    },
+    {
+      "epoch": 0.05504850651008425,
+      "grad_norm": 0.7749186158180237,
+      "learning_rate": 2.3775533144912643e-05,
+      "loss": 0.5417,
+      "step": 1725
+    },
+    {
+      "epoch": 0.055846310952259386,
+      "grad_norm": 0.8586596250534058,
+      "learning_rate": 2.3755459917780064e-05,
+      "loss": 0.6107,
+      "step": 1750
+    },
+    {
+      "epoch": 0.056644115394434515,
+      "grad_norm": 0.5593289136886597,
+      "learning_rate": 2.3735386690647482e-05,
+      "loss": 0.555,
+      "step": 1775
+    },
+    {
+      "epoch": 0.05744191983660965,
+      "grad_norm": 0.39711669087409973,
+      "learning_rate": 2.3715313463514903e-05,
+      "loss": 0.5285,
+      "step": 1800
+    },
+    {
+      "epoch": 0.05823972427878479,
+      "grad_norm": 0.6996438503265381,
+      "learning_rate": 2.3695240236382324e-05,
+      "loss": 0.561,
+      "step": 1825
+    },
+    {
+      "epoch": 0.059037528720959916,
+      "grad_norm": 0.6649994850158691,
+      "learning_rate": 2.3675167009249745e-05,
+      "loss": 0.6026,
+      "step": 1850
+    },
+    {
+      "epoch": 0.05983533316313505,
+      "grad_norm": 0.7238967418670654,
+      "learning_rate": 2.3655093782117166e-05,
+      "loss": 0.5061,
+      "step": 1875
+    },
+    {
+      "epoch": 0.06063313760531019,
+      "grad_norm": 1.1705780029296875,
+      "learning_rate": 2.3635020554984584e-05,
+      "loss": 0.6363,
+      "step": 1900
+    },
+    {
+      "epoch": 0.061430942047485323,
+      "grad_norm": 1.1129088401794434,
+      "learning_rate": 2.3614947327852005e-05,
+      "loss": 0.4697,
+      "step": 1925
+    },
+    {
+      "epoch": 0.06222874648966045,
+      "grad_norm": 0.6562258005142212,
+      "learning_rate": 2.3594874100719426e-05,
+      "loss": 0.5425,
+      "step": 1950
+    },
+    {
+      "epoch": 0.06302655093183558,
+      "grad_norm": 1.2954466342926025,
+      "learning_rate": 2.3574800873586847e-05,
+      "loss": 0.508,
+      "step": 1975
+    },
+    {
+      "epoch": 0.06382435537401072,
+      "grad_norm": 0.428022176027298,
+      "learning_rate": 2.3554727646454265e-05,
+      "loss": 0.5646,
+      "step": 2000
+    },
+    {
+      "epoch": 0.06462215981618585,
+      "grad_norm": 1.1137559413909912,
+      "learning_rate": 2.3534654419321686e-05,
+      "loss": 0.5514,
+      "step": 2025
+    },
+    {
+      "epoch": 0.065419964258361,
+      "grad_norm": 0.9217244982719421,
+      "learning_rate": 2.3514581192189107e-05,
+      "loss": 0.4994,
+      "step": 2050
+    },
+    {
+      "epoch": 0.06621776870053613,
+      "grad_norm": 1.9023886919021606,
+      "learning_rate": 2.349450796505653e-05,
+      "loss": 0.4763,
+      "step": 2075
+    },
+    {
+      "epoch": 0.06701557314271125,
+      "grad_norm": 0.49815502762794495,
+      "learning_rate": 2.3474434737923946e-05,
+      "loss": 0.5196,
+      "step": 2100
+    },
+    {
+      "epoch": 0.0678133775848864,
+      "grad_norm": 0.4530934691429138,
+      "learning_rate": 2.3454361510791367e-05,
+      "loss": 0.6507,
+      "step": 2125
+    },
+    {
+      "epoch": 0.06861118202706153,
+      "grad_norm": 1.0055551528930664,
+      "learning_rate": 2.343428828365879e-05,
+      "loss": 0.5758,
+      "step": 2150
+    },
+    {
+      "epoch": 0.06940898646923666,
+      "grad_norm": 0.8732419013977051,
+      "learning_rate": 2.341421505652621e-05,
+      "loss": 0.5036,
+      "step": 2175
+    },
+    {
+      "epoch": 0.0702067909114118,
+      "grad_norm": 0.7871633768081665,
+      "learning_rate": 2.3394141829393627e-05,
+      "loss": 0.5806,
+      "step": 2200
+    },
+    {
+      "epoch": 0.07100459535358693,
+      "grad_norm": 0.6435555815696716,
+      "learning_rate": 2.337406860226105e-05,
+      "loss": 0.6648,
+      "step": 2225
+    },
+    {
+      "epoch": 0.07180239979576206,
+      "grad_norm": 1.1382043361663818,
+      "learning_rate": 2.335399537512847e-05,
+      "loss": 0.5407,
+      "step": 2250
+    },
+    {
+      "epoch": 0.0726002042379372,
+      "grad_norm": 0.8095251321792603,
+      "learning_rate": 2.333392214799589e-05,
+      "loss": 0.5153,
+      "step": 2275
+    },
+    {
+      "epoch": 0.07339800868011233,
+      "grad_norm": 1.1035778522491455,
+      "learning_rate": 2.331384892086331e-05,
+      "loss": 0.5654,
+      "step": 2300
+    },
+    {
+      "epoch": 0.07419581312228747,
+      "grad_norm": 1.2554080486297607,
+      "learning_rate": 2.329377569373073e-05,
+      "loss": 0.5228,
+      "step": 2325
+    },
+    {
+      "epoch": 0.0749936175644626,
+      "grad_norm": 1.3043935298919678,
+      "learning_rate": 2.327370246659815e-05,
+      "loss": 0.5288,
+      "step": 2350
+    },
+    {
+      "epoch": 0.07579142200663773,
+      "grad_norm": 0.9910039901733398,
+      "learning_rate": 2.3253629239465572e-05,
+      "loss": 0.5839,
+      "step": 2375
+    },
+    {
+      "epoch": 0.07658922644881287,
+      "grad_norm": 0.8184835910797119,
+      "learning_rate": 2.323355601233299e-05,
+      "loss": 0.5325,
+      "step": 2400
+    },
+    {
+      "epoch": 0.077387030890988,
+      "grad_norm": 0.8235958814620972,
+      "learning_rate": 2.3213482785200414e-05,
+      "loss": 0.5109,
+      "step": 2425
+    },
+    {
+      "epoch": 0.07818483533316313,
+      "grad_norm": 0.581178605556488,
+      "learning_rate": 2.3193409558067832e-05,
+      "loss": 0.5618,
+      "step": 2450
+    },
+    {
+      "epoch": 0.07898263977533827,
+      "grad_norm": 1.0203675031661987,
+      "learning_rate": 2.3173336330935253e-05,
+      "loss": 0.485,
+      "step": 2475
+    },
+    {
+      "epoch": 0.0797804442175134,
+      "grad_norm": 1.0625289678573608,
+      "learning_rate": 2.315326310380267e-05,
+      "loss": 0.5517,
+      "step": 2500
+    },
+    {
+      "epoch": 0.08057824865968853,
+      "grad_norm": 0.6422656178474426,
+      "learning_rate": 2.3133189876670095e-05,
+      "loss": 0.5651,
+      "step": 2525
+    },
+    {
+      "epoch": 0.08137605310186367,
+      "grad_norm": 0.6603285670280457,
+      "learning_rate": 2.3113116649537513e-05,
+      "loss": 0.52,
+      "step": 2550
+    },
+    {
+      "epoch": 0.0821738575440388,
+      "grad_norm": 0.593621015548706,
+      "learning_rate": 2.3093043422404934e-05,
+      "loss": 0.4592,
+      "step": 2575
+    },
+    {
+      "epoch": 0.08297166198621395,
+      "grad_norm": 0.5782598853111267,
+      "learning_rate": 2.3072970195272355e-05,
+      "loss": 0.5465,
+      "step": 2600
+    },
+    {
+      "epoch": 0.08376946642838907,
+      "grad_norm": 0.6208533644676208,
+      "learning_rate": 2.3052896968139777e-05,
+      "loss": 0.5752,
+      "step": 2625
+    },
+    {
+      "epoch": 0.0845672708705642,
+      "grad_norm": 0.5738264918327332,
+      "learning_rate": 2.3032823741007194e-05,
+      "loss": 0.5624,
+      "step": 2650
+    },
+    {
+      "epoch": 0.08536507531273935,
+      "grad_norm": 0.9327691197395325,
+      "learning_rate": 2.3012750513874615e-05,
+      "loss": 0.5183,
+      "step": 2675
+    },
+    {
+      "epoch": 0.08616287975491448,
+      "grad_norm": 0.3608091473579407,
+      "learning_rate": 2.2992677286742037e-05,
+      "loss": 0.4504,
+      "step": 2700
+    },
+    {
+      "epoch": 0.0869606841970896,
+      "grad_norm": 1.4098608493804932,
+      "learning_rate": 2.2972604059609458e-05,
+      "loss": 0.5134,
+      "step": 2725
+    },
+    {
+      "epoch": 0.08775848863926475,
+      "grad_norm": 0.7239978909492493,
+      "learning_rate": 2.2952530832476875e-05,
+      "loss": 0.544,
+      "step": 2750
+    },
+    {
+      "epoch": 0.08855629308143988,
+      "grad_norm": 0.9918779134750366,
+      "learning_rate": 2.2932457605344297e-05,
+      "loss": 0.5694,
+      "step": 2775
+    },
+    {
+      "epoch": 0.089354097523615,
+      "grad_norm": 0.7250977158546448,
+      "learning_rate": 2.2912384378211718e-05,
+      "loss": 0.5262,
+      "step": 2800
+    },
+    {
+      "epoch": 0.09015190196579015,
+      "grad_norm": 0.8188421726226807,
+      "learning_rate": 2.289231115107914e-05,
+      "loss": 0.6284,
+      "step": 2825
+    },
+    {
+      "epoch": 0.09094970640796528,
+      "grad_norm": 0.46473366022109985,
+      "learning_rate": 2.2872237923946557e-05,
+      "loss": 0.4849,
+      "step": 2850
+    },
+    {
+      "epoch": 0.0917475108501404,
+      "grad_norm": 0.5546078085899353,
+      "learning_rate": 2.285216469681398e-05,
+      "loss": 0.5263,
+      "step": 2875
+    },
+    {
+      "epoch": 0.09254531529231555,
+      "grad_norm": 0.9259676337242126,
+      "learning_rate": 2.28320914696814e-05,
+      "loss": 0.5464,
+      "step": 2900
+    },
+    {
+      "epoch": 0.09334311973449068,
+      "grad_norm": 1.2702957391738892,
+      "learning_rate": 2.281201824254882e-05,
+      "loss": 0.5304,
+      "step": 2925
+    },
+    {
+      "epoch": 0.09414092417666582,
+      "grad_norm": 1.6516447067260742,
+      "learning_rate": 2.2791945015416238e-05,
+      "loss": 0.5297,
+      "step": 2950
+    },
+    {
+      "epoch": 0.09493872861884095,
+      "grad_norm": 0.70711350440979,
+      "learning_rate": 2.2771871788283662e-05,
+      "loss": 0.5308,
+      "step": 2975
+    },
+    {
+      "epoch": 0.09573653306101608,
+      "grad_norm": 0.931492805480957,
+      "learning_rate": 2.275179856115108e-05,
+      "loss": 0.5609,
+      "step": 3000
+    },
+    {
+      "epoch": 0.09653433750319122,
+      "grad_norm": 0.3455384373664856,
+      "learning_rate": 2.27317253340185e-05,
+      "loss": 0.5528,
+      "step": 3025
+    },
+    {
+      "epoch": 0.09733214194536635,
+      "grad_norm": 0.7669233083724976,
+      "learning_rate": 2.2711652106885922e-05,
+      "loss": 0.4318,
+      "step": 3050
+    },
+    {
+      "epoch": 0.09812994638754148,
+      "grad_norm": 0.5945572853088379,
+      "learning_rate": 2.2691578879753343e-05,
+      "loss": 0.5695,
+      "step": 3075
+    },
+    {
+      "epoch": 0.09892775082971662,
+      "grad_norm": 0.8934717774391174,
+      "learning_rate": 2.267150565262076e-05,
+      "loss": 0.5284,
+      "step": 3100
+    },
+    {
+      "epoch": 0.09972555527189175,
+      "grad_norm": 0.6024275422096252,
+      "learning_rate": 2.2651432425488182e-05,
+      "loss": 0.5028,
+      "step": 3125
+    },
+    {
+      "epoch": 0.10052335971406688,
+      "grad_norm": 0.6754337549209595,
+      "learning_rate": 2.2631359198355603e-05,
+      "loss": 0.5949,
+      "step": 3150
+    },
+    {
+      "epoch": 0.10132116415624202,
+      "grad_norm": 1.1694283485412598,
+      "learning_rate": 2.2611285971223024e-05,
+      "loss": 0.5488,
+      "step": 3175
+    },
+    {
+      "epoch": 0.10211896859841715,
+      "grad_norm": 0.8880589604377747,
+      "learning_rate": 2.2591212744090442e-05,
+      "loss": 0.5348,
+      "step": 3200
+    },
+    {
+      "epoch": 0.1029167730405923,
+      "grad_norm": 0.425998330116272,
+      "learning_rate": 2.2571139516957863e-05,
+      "loss": 0.6288,
+      "step": 3225
+    },
+    {
+      "epoch": 0.10371457748276743,
+      "grad_norm": 0.9844962358474731,
+      "learning_rate": 2.2551066289825284e-05,
+      "loss": 0.6145,
+      "step": 3250
+    },
+    {
+      "epoch": 0.10451238192494255,
+      "grad_norm": 0.7182367444038391,
+      "learning_rate": 2.2530993062692706e-05,
+      "loss": 0.6248,
+      "step": 3275
+    },
+    {
+      "epoch": 0.1053101863671177,
+      "grad_norm": 0.493770956993103,
+      "learning_rate": 2.2510919835560123e-05,
+      "loss": 0.5463,
+      "step": 3300
+    },
+    {
+      "epoch": 0.10610799080929283,
+      "grad_norm": 1.0892950296401978,
+      "learning_rate": 2.2490846608427545e-05,
+      "loss": 0.543,
+      "step": 3325
+    },
+    {
+      "epoch": 0.10690579525146796,
+      "grad_norm": 0.9539535045623779,
+      "learning_rate": 2.2470773381294966e-05,
+      "loss": 0.5055,
+      "step": 3350
+    },
+    {
+      "epoch": 0.1077035996936431,
+      "grad_norm": 0.9406744837760925,
+      "learning_rate": 2.2450700154162387e-05,
+      "loss": 0.4861,
+      "step": 3375
+    },
+    {
+      "epoch": 0.10850140413581823,
+      "grad_norm": 1.0666565895080566,
+      "learning_rate": 2.2430626927029805e-05,
+      "loss": 0.4699,
+      "step": 3400
+    },
+    {
+      "epoch": 0.10929920857799336,
+      "grad_norm": 0.6132543683052063,
+      "learning_rate": 2.2410553699897226e-05,
+      "loss": 0.5105,
+      "step": 3425
+    },
+    {
+      "epoch": 0.1100970130201685,
+      "grad_norm": 1.0401798486709595,
+      "learning_rate": 2.2390480472764647e-05,
+      "loss": 0.5095,
+      "step": 3450
+    },
+    {
+      "epoch": 0.11089481746234363,
+      "grad_norm": 0.6453993320465088,
+      "learning_rate": 2.2370407245632068e-05,
+      "loss": 0.5078,
+      "step": 3475
+    },
+    {
+      "epoch": 0.11169262190451877,
+      "grad_norm": 1.104200839996338,
+      "learning_rate": 2.2350334018499486e-05,
+      "loss": 0.5562,
+      "step": 3500
+    },
+    {
+      "epoch": 0.1124904263466939,
+      "grad_norm": 0.8307452201843262,
+      "learning_rate": 2.2330260791366907e-05,
+      "loss": 0.5327,
+      "step": 3525
+    },
+    {
+      "epoch": 0.11328823078886903,
+      "grad_norm": 0.5826612710952759,
+      "learning_rate": 2.2310187564234328e-05,
+      "loss": 0.4893,
+      "step": 3550
+    },
+    {
+      "epoch": 0.11408603523104417,
+      "grad_norm": 0.5139064192771912,
+      "learning_rate": 2.2290114337101746e-05,
+      "loss": 0.5041,
+      "step": 3575
+    },
+    {
+      "epoch": 0.1148838396732193,
+      "grad_norm": 0.40968069434165955,
+      "learning_rate": 2.227004110996917e-05,
+      "loss": 0.519,
+      "step": 3600
+    },
+    {
+      "epoch": 0.11568164411539443,
+      "grad_norm": 0.7572267055511475,
+      "learning_rate": 2.2249967882836588e-05,
+      "loss": 0.4424,
+      "step": 3625
+    },
+    {
+      "epoch": 0.11647944855756957,
+      "grad_norm": 1.1099541187286377,
+      "learning_rate": 2.222989465570401e-05,
+      "loss": 0.5423,
+      "step": 3650
+    },
+    {
+      "epoch": 0.1172772529997447,
+      "grad_norm": 0.8752992749214172,
+      "learning_rate": 2.2209821428571427e-05,
+      "loss": 0.5944,
+      "step": 3675
+    },
+    {
+      "epoch": 0.11807505744191983,
+      "grad_norm": 1.1976771354675293,
+      "learning_rate": 2.218974820143885e-05,
+      "loss": 0.5575,
+      "step": 3700
+    },
+    {
+      "epoch": 0.11887286188409497,
+      "grad_norm": 1.2560482025146484,
+      "learning_rate": 2.216967497430627e-05,
+      "loss": 0.502,
+      "step": 3725
+    },
+    {
+      "epoch": 0.1196706663262701,
+      "grad_norm": 0.7299981713294983,
+      "learning_rate": 2.214960174717369e-05,
+      "loss": 0.5669,
+      "step": 3750
+    },
+    {
+      "epoch": 0.12046847076844523,
+      "grad_norm": 0.9207307696342468,
+      "learning_rate": 2.212952852004111e-05,
+      "loss": 0.5062,
+      "step": 3775
+    },
+    {
+      "epoch": 0.12126627521062038,
+      "grad_norm": 1.2774001359939575,
+      "learning_rate": 2.2109455292908532e-05,
+      "loss": 0.5401,
+      "step": 3800
+    },
+    {
+      "epoch": 0.1220640796527955,
+      "grad_norm": 0.797244668006897,
+      "learning_rate": 2.208938206577595e-05,
+      "loss": 0.5483,
+      "step": 3825
+    },
+    {
+      "epoch": 0.12286188409497065,
+      "grad_norm": 0.9657193422317505,
+      "learning_rate": 2.206930883864337e-05,
+      "loss": 0.5249,
+      "step": 3850
+    },
+    {
+      "epoch": 0.12365968853714578,
+      "grad_norm": 0.7845392227172852,
+      "learning_rate": 2.2049235611510792e-05,
+      "loss": 0.5519,
+      "step": 3875
+    },
+    {
+      "epoch": 0.1244574929793209,
+      "grad_norm": 0.8090838193893433,
+      "learning_rate": 2.2029162384378214e-05,
+      "loss": 0.565,
+      "step": 3900
+    },
+    {
+      "epoch": 0.12525529742149605,
+      "grad_norm": 0.8069561719894409,
+      "learning_rate": 2.200908915724563e-05,
+      "loss": 0.6185,
+      "step": 3925
+    },
+    {
+      "epoch": 0.12605310186367116,
+      "grad_norm": 0.8270429372787476,
+      "learning_rate": 2.1989015930113052e-05,
+      "loss": 0.5291,
+      "step": 3950
+    },
+    {
+      "epoch": 0.1268509063058463,
+      "grad_norm": 0.7712761163711548,
+      "learning_rate": 2.1968942702980474e-05,
+      "loss": 0.581,
+      "step": 3975
+    },
+    {
+      "epoch": 0.12764871074802145,
+      "grad_norm": 0.6054518222808838,
+      "learning_rate": 2.1948869475847895e-05,
+      "loss": 0.4576,
+      "step": 4000
+    },
+    {
+      "epoch": 0.1284465151901966,
+      "grad_norm": 1.1705981492996216,
+      "learning_rate": 2.1928796248715312e-05,
+      "loss": 0.481,
+      "step": 4025
+    },
+    {
+      "epoch": 0.1292443196323717,
+      "grad_norm": 1.2569853067398071,
+      "learning_rate": 2.1908723021582737e-05,
+      "loss": 0.5749,
+      "step": 4050
+    },
+    {
+      "epoch": 0.13004212407454685,
+      "grad_norm": 1.009342908859253,
+      "learning_rate": 2.1888649794450155e-05,
+      "loss": 0.5604,
+      "step": 4075
+    },
+    {
+      "epoch": 0.130839928516722,
+      "grad_norm": 0.8715687394142151,
+      "learning_rate": 2.1868576567317576e-05,
+      "loss": 0.4648,
+      "step": 4100
+    },
+    {
+      "epoch": 0.1316377329588971,
+      "grad_norm": 0.7666533589363098,
+      "learning_rate": 2.1848503340184994e-05,
+      "loss": 0.5616,
+      "step": 4125
+    },
+    {
+      "epoch": 0.13243553740107225,
+      "grad_norm": 0.9141665101051331,
+      "learning_rate": 2.1828430113052418e-05,
+      "loss": 0.517,
+      "step": 4150
+    },
+    {
+      "epoch": 0.1332333418432474,
+      "grad_norm": 0.7477203011512756,
+      "learning_rate": 2.1808356885919836e-05,
+      "loss": 0.4709,
+      "step": 4175
+    },
+    {
+      "epoch": 0.1340311462854225,
+      "grad_norm": 1.1471550464630127,
+      "learning_rate": 2.1788283658787257e-05,
+      "loss": 0.4974,
+      "step": 4200
+    },
+    {
+      "epoch": 0.13482895072759765,
+      "grad_norm": 0.5586664080619812,
+      "learning_rate": 2.1768210431654678e-05,
+      "loss": 0.5779,
+      "step": 4225
+    },
+    {
+      "epoch": 0.1356267551697728,
+      "grad_norm": 0.8173395395278931,
+      "learning_rate": 2.17481372045221e-05,
+      "loss": 0.4929,
+      "step": 4250
+    },
+    {
+      "epoch": 0.1364245596119479,
+      "grad_norm": 0.8017388582229614,
+      "learning_rate": 2.1728063977389517e-05,
+      "loss": 0.423,
+      "step": 4275
+    },
+    {
+      "epoch": 0.13722236405412305,
+      "grad_norm": 0.6071333885192871,
+      "learning_rate": 2.1707990750256938e-05,
+      "loss": 0.5407,
+      "step": 4300
+    },
+    {
+      "epoch": 0.1380201684962982,
+      "grad_norm": 0.6837856769561768,
+      "learning_rate": 2.168791752312436e-05,
+      "loss": 0.4963,
+      "step": 4325
+    },
+    {
+      "epoch": 0.1388179729384733,
+      "grad_norm": 0.5120430588722229,
+      "learning_rate": 2.166784429599178e-05,
+      "loss": 0.5045,
+      "step": 4350
+    },
+    {
+      "epoch": 0.13961577738064845,
+      "grad_norm": 0.8727617263793945,
+      "learning_rate": 2.1647771068859198e-05,
+      "loss": 0.4111,
+      "step": 4375
+    },
+    {
+      "epoch": 0.1404135818228236,
+      "grad_norm": 1.0532296895980835,
+      "learning_rate": 2.162769784172662e-05,
+      "loss": 0.4489,
+      "step": 4400
+    },
+    {
+      "epoch": 0.1412113862649987,
+      "grad_norm": 0.57016921043396,
+      "learning_rate": 2.160762461459404e-05,
+      "loss": 0.477,
+      "step": 4425
+    },
+    {
+      "epoch": 0.14200919070717385,
+      "grad_norm": 0.8259162902832031,
+      "learning_rate": 2.158755138746146e-05,
+      "loss": 0.5164,
+      "step": 4450
+    },
+    {
+      "epoch": 0.142806995149349,
+      "grad_norm": 0.3785061538219452,
+      "learning_rate": 2.156747816032888e-05,
+      "loss": 0.4683,
+      "step": 4475
+    },
+    {
+      "epoch": 0.1436047995915241,
+      "grad_norm": 1.1364110708236694,
+      "learning_rate": 2.1547404933196304e-05,
+      "loss": 0.4723,
+      "step": 4500
+    },
+    {
+      "epoch": 0.14440260403369926,
+      "grad_norm": 0.8808683156967163,
+      "learning_rate": 2.152733170606372e-05,
+      "loss": 0.5151,
+      "step": 4525
+    },
+    {
+      "epoch": 0.1452004084758744,
+      "grad_norm": 0.8673577308654785,
+      "learning_rate": 2.1507258478931143e-05,
+      "loss": 0.5108,
+      "step": 4550
+    },
+    {
+      "epoch": 0.14599821291804954,
+      "grad_norm": 1.2383835315704346,
+      "learning_rate": 2.148718525179856e-05,
+      "loss": 0.5185,
+      "step": 4575
+    },
+    {
+      "epoch": 0.14679601736022466,
+      "grad_norm": 1.3869210481643677,
+      "learning_rate": 2.1467112024665985e-05,
+      "loss": 0.5453,
+      "step": 4600
+    },
+    {
+      "epoch": 0.1475938218023998,
+      "grad_norm": 0.4665587842464447,
+      "learning_rate": 2.1447038797533403e-05,
+      "loss": 0.4585,
+      "step": 4625
+    },
+    {
+      "epoch": 0.14839162624457494,
+      "grad_norm": 0.29424333572387695,
+      "learning_rate": 2.1426965570400824e-05,
+      "loss": 0.5438,
+      "step": 4650
+    },
+    {
+      "epoch": 0.14918943068675006,
+      "grad_norm": 1.7765225172042847,
+      "learning_rate": 2.140689234326824e-05,
+      "loss": 0.5499,
+      "step": 4675
+    },
+    {
+      "epoch": 0.1499872351289252,
+      "grad_norm": 0.9864098429679871,
+      "learning_rate": 2.1386819116135666e-05,
+      "loss": 0.4991,
+      "step": 4700
+    },
+    {
+      "epoch": 0.15078503957110034,
+      "grad_norm": 0.8926157355308533,
+      "learning_rate": 2.1366745889003084e-05,
+      "loss": 0.5008,
+      "step": 4725
+    },
+    {
+      "epoch": 0.15158284401327546,
+      "grad_norm": 0.7439494729042053,
+      "learning_rate": 2.1346672661870505e-05,
+      "loss": 0.5324,
+      "step": 4750
+    },
+    {
+      "epoch": 0.1523806484554506,
+      "grad_norm": 0.6221075057983398,
+      "learning_rate": 2.1326599434737926e-05,
+      "loss": 0.5064,
+      "step": 4775
+    },
+    {
+      "epoch": 0.15317845289762574,
+      "grad_norm": 0.8994887471199036,
+      "learning_rate": 2.1306526207605347e-05,
+      "loss": 0.5377,
+      "step": 4800
+    },
+    {
+      "epoch": 0.15397625733980086,
+      "grad_norm": 1.5914185047149658,
+      "learning_rate": 2.1286452980472765e-05,
+      "loss": 0.4884,
+      "step": 4825
+    },
+    {
+      "epoch": 0.154774061781976,
+      "grad_norm": 0.37632209062576294,
+      "learning_rate": 2.1266379753340186e-05,
+      "loss": 0.5193,
+      "step": 4850
+    },
+    {
+      "epoch": 0.15557186622415115,
+      "grad_norm": 1.095083475112915,
+      "learning_rate": 2.1246306526207607e-05,
+      "loss": 0.4342,
+      "step": 4875
+    },
+    {
+      "epoch": 0.15636967066632626,
+      "grad_norm": 1.2659378051757812,
+      "learning_rate": 2.122623329907503e-05,
+      "loss": 0.5291,
+      "step": 4900
+    },
+    {
+      "epoch": 0.1571674751085014,
+      "grad_norm": 0.8220809698104858,
+      "learning_rate": 2.1206160071942446e-05,
+      "loss": 0.5319,
+      "step": 4925
+    },
+    {
+      "epoch": 0.15796527955067655,
+      "grad_norm": 0.8385465145111084,
+      "learning_rate": 2.1186086844809867e-05,
+      "loss": 0.4569,
+      "step": 4950
+    },
+    {
+      "epoch": 0.15876308399285166,
+      "grad_norm": 0.5405530333518982,
+      "learning_rate": 2.116601361767729e-05,
+      "loss": 0.4429,
+      "step": 4975
+    },
+    {
+      "epoch": 0.1595608884350268,
+      "grad_norm": 1.7731573581695557,
+      "learning_rate": 2.114594039054471e-05,
+      "loss": 0.4868,
+      "step": 5000
+    },
+    {
+      "epoch": 0.16035869287720195,
+      "grad_norm": 1.1661796569824219,
+      "learning_rate": 2.1125867163412127e-05,
+      "loss": 0.438,
+      "step": 5025
+    },
+    {
+      "epoch": 0.16115649731937706,
+      "grad_norm": 0.8461349606513977,
+      "learning_rate": 2.110579393627955e-05,
+      "loss": 0.566,
+      "step": 5050
+    },
+    {
+      "epoch": 0.1619543017615522,
+      "grad_norm": 0.503616452217102,
+      "learning_rate": 2.108572070914697e-05,
+      "loss": 0.4791,
+      "step": 5075
+    },
+    {
+      "epoch": 0.16275210620372735,
+      "grad_norm": 1.0445747375488281,
+      "learning_rate": 2.1065647482014387e-05,
+      "loss": 0.5193,
+      "step": 5100
+    },
+    {
+      "epoch": 0.16354991064590246,
+      "grad_norm": 0.9016474485397339,
+      "learning_rate": 2.104557425488181e-05,
+      "loss": 0.5409,
+      "step": 5125
+    },
+    {
+      "epoch": 0.1643477150880776,
+      "grad_norm": 1.2945493459701538,
+      "learning_rate": 2.102550102774923e-05,
+      "loss": 0.528,
+      "step": 5150
+    },
+    {
+      "epoch": 0.16514551953025275,
+      "grad_norm": 1.1654711961746216,
+      "learning_rate": 2.100542780061665e-05,
+      "loss": 0.5258,
+      "step": 5175
+    },
+    {
+      "epoch": 0.1659433239724279,
+      "grad_norm": 0.4375602602958679,
+      "learning_rate": 2.098535457348407e-05,
+      "loss": 0.5405,
+      "step": 5200
+    },
+    {
+      "epoch": 0.166741128414603,
+      "grad_norm": 0.808623194694519,
+      "learning_rate": 2.0965281346351493e-05,
+      "loss": 0.4869,
+      "step": 5225
+    },
+    {
+      "epoch": 0.16753893285677815,
+      "grad_norm": 0.881857693195343,
+      "learning_rate": 2.094520811921891e-05,
+      "loss": 0.492,
+      "step": 5250
+    },
+    {
+      "epoch": 0.1683367372989533,
+      "grad_norm": 0.7097049355506897,
+      "learning_rate": 2.0925134892086332e-05,
+      "loss": 0.4401,
+      "step": 5275
+    },
+    {
+      "epoch": 0.1691345417411284,
+      "grad_norm": 0.6862661242485046,
+      "learning_rate": 2.090506166495375e-05,
+      "loss": 0.5644,
+      "step": 5300
+    },
+    {
+      "epoch": 0.16993234618330355,
+      "grad_norm": 0.5271662473678589,
+      "learning_rate": 2.0884988437821174e-05,
+      "loss": 0.536,
+      "step": 5325
+    },
+    {
+      "epoch": 0.1707301506254787,
+      "grad_norm": 1.1128803491592407,
+      "learning_rate": 2.0864915210688592e-05,
+      "loss": 0.5166,
+      "step": 5350
+    },
+    {
+      "epoch": 0.1715279550676538,
+      "grad_norm": 0.656862199306488,
+      "learning_rate": 2.0844841983556013e-05,
+      "loss": 0.4488,
+      "step": 5375
+    },
+    {
+      "epoch": 0.17232575950982895,
+      "grad_norm": 0.5935942530632019,
+      "learning_rate": 2.0824768756423434e-05,
+      "loss": 0.5376,
+      "step": 5400
+    },
+    {
+      "epoch": 0.1731235639520041,
+      "grad_norm": 0.6691839098930359,
+      "learning_rate": 2.0804695529290855e-05,
+      "loss": 0.512,
+      "step": 5425
+    },
+    {
+      "epoch": 0.1739213683941792,
+      "grad_norm": 0.6211944818496704,
+      "learning_rate": 2.0784622302158273e-05,
+      "loss": 0.535,
+      "step": 5450
+    },
+    {
+      "epoch": 0.17471917283635435,
+      "grad_norm": 0.8693767786026001,
+      "learning_rate": 2.0764549075025694e-05,
+      "loss": 0.5033,
+      "step": 5475
+    },
+    {
+      "epoch": 0.1755169772785295,
+      "grad_norm": 0.5080140829086304,
+      "learning_rate": 2.0744475847893115e-05,
+      "loss": 0.5029,
+      "step": 5500
+    },
+    {
+      "epoch": 0.1763147817207046,
+      "grad_norm": 0.9701049327850342,
+      "learning_rate": 2.0724402620760536e-05,
+      "loss": 0.4856,
+      "step": 5525
+    },
+    {
+      "epoch": 0.17711258616287975,
+      "grad_norm": 0.9497528672218323,
+      "learning_rate": 2.0704329393627954e-05,
+      "loss": 0.5074,
+      "step": 5550
+    },
+    {
+      "epoch": 0.1779103906050549,
+      "grad_norm": 1.4564626216888428,
+      "learning_rate": 2.0684256166495375e-05,
+      "loss": 0.4149,
+      "step": 5575
+    },
+    {
+      "epoch": 0.17870819504723,
+      "grad_norm": 1.11367928981781,
+      "learning_rate": 2.0664182939362796e-05,
+      "loss": 0.4164,
+      "step": 5600
+    },
+    {
+      "epoch": 0.17950599948940515,
+      "grad_norm": 0.6194526553153992,
+      "learning_rate": 2.0644109712230218e-05,
+      "loss": 0.5459,
+      "step": 5625
+    },
+    {
+      "epoch": 0.1803038039315803,
+      "grad_norm": 1.52586030960083,
+      "learning_rate": 2.0624036485097635e-05,
+      "loss": 0.5643,
+      "step": 5650
+    },
+    {
+      "epoch": 0.1811016083737554,
+      "grad_norm": 0.6483252644538879,
+      "learning_rate": 2.060396325796506e-05,
+      "loss": 0.4655,
+      "step": 5675
+    },
+    {
+      "epoch": 0.18189941281593056,
+      "grad_norm": 0.6978493928909302,
+      "learning_rate": 2.0583890030832478e-05,
+      "loss": 0.5387,
+      "step": 5700
+    },
+    {
+      "epoch": 0.1826972172581057,
+      "grad_norm": 0.3132692277431488,
+      "learning_rate": 2.05638168036999e-05,
+      "loss": 0.4972,
+      "step": 5725
+    },
+    {
+      "epoch": 0.1834950217002808,
+      "grad_norm": 0.7360209226608276,
+      "learning_rate": 2.0543743576567316e-05,
+      "loss": 0.6053,
+      "step": 5750
+    },
+    {
+      "epoch": 0.18429282614245596,
+      "grad_norm": 0.4095138907432556,
+      "learning_rate": 2.052367034943474e-05,
+      "loss": 0.4819,
+      "step": 5775
+    },
+    {
+      "epoch": 0.1850906305846311,
+      "grad_norm": 0.6342286467552185,
+      "learning_rate": 2.050359712230216e-05,
+      "loss": 0.4445,
+      "step": 5800
+    },
+    {
+      "epoch": 0.18588843502680624,
+      "grad_norm": 1.3439861536026,
+      "learning_rate": 2.048352389516958e-05,
+      "loss": 0.407,
+      "step": 5825
+    },
+    {
+      "epoch": 0.18668623946898136,
+      "grad_norm": 1.3485324382781982,
+      "learning_rate": 2.0463450668036998e-05,
+      "loss": 0.4515,
+      "step": 5850
+    },
+    {
+      "epoch": 0.1874840439111565,
+      "grad_norm": 1.4493141174316406,
+      "learning_rate": 2.0443377440904422e-05,
+      "loss": 0.4998,
+      "step": 5875
+    },
+    {
+      "epoch": 0.18828184835333164,
+      "grad_norm": 0.6169376969337463,
+      "learning_rate": 2.042330421377184e-05,
+      "loss": 0.4751,
+      "step": 5900
+    },
+    {
+      "epoch": 0.18907965279550676,
+      "grad_norm": 0.8687784671783447,
+      "learning_rate": 2.040323098663926e-05,
+      "loss": 0.5946,
+      "step": 5925
+    },
+    {
+      "epoch": 0.1898774572376819,
+      "grad_norm": 1.027100920677185,
+      "learning_rate": 2.0383157759506682e-05,
+      "loss": 0.4756,
+      "step": 5950
+    },
+    {
+      "epoch": 0.19067526167985704,
+      "grad_norm": 2.688061475753784,
+      "learning_rate": 2.0363084532374103e-05,
+      "loss": 0.505,
+      "step": 5975
+    },
+    {
+      "epoch": 0.19147306612203216,
+      "grad_norm": 0.5821250081062317,
+      "learning_rate": 2.034301130524152e-05,
+      "loss": 0.4667,
+      "step": 6000
+    },
+    {
+      "epoch": 0.1922708705642073,
+      "grad_norm": 1.2581595182418823,
+      "learning_rate": 2.0322938078108942e-05,
+      "loss": 0.5072,
+      "step": 6025
+    },
+    {
+      "epoch": 0.19306867500638245,
+      "grad_norm": 0.5629422068595886,
+      "learning_rate": 2.0302864850976363e-05,
+      "loss": 0.4727,
+      "step": 6050
+    },
+    {
+      "epoch": 0.19386647944855756,
+      "grad_norm": 1.0601611137390137,
+      "learning_rate": 2.0282791623843784e-05,
+      "loss": 0.516,
+      "step": 6075
+    },
+    {
+      "epoch": 0.1946642838907327,
+      "grad_norm": 0.5984941720962524,
+      "learning_rate": 2.0262718396711202e-05,
+      "loss": 0.4991,
+      "step": 6100
+    },
+    {
+      "epoch": 0.19546208833290785,
+      "grad_norm": 1.2396142482757568,
+      "learning_rate": 2.0242645169578627e-05,
+      "loss": 0.4942,
+      "step": 6125
+    },
+    {
+      "epoch": 0.19625989277508296,
+      "grad_norm": 0.722944974899292,
+      "learning_rate": 2.0222571942446044e-05,
+      "loss": 0.5489,
+      "step": 6150
+    },
+    {
+      "epoch": 0.1970576972172581,
+      "grad_norm": 0.28916776180267334,
+      "learning_rate": 2.0202498715313466e-05,
+      "loss": 0.4739,
+      "step": 6175
+    },
+    {
+      "epoch": 0.19785550165943325,
+      "grad_norm": 0.5348458886146545,
+      "learning_rate": 2.0182425488180883e-05,
+      "loss": 0.4926,
+      "step": 6200
+    },
+    {
+      "epoch": 0.19865330610160836,
+      "grad_norm": 1.1795445680618286,
+      "learning_rate": 2.0162352261048308e-05,
+      "loss": 0.5097,
+      "step": 6225
+    },
+    {
+      "epoch": 0.1994511105437835,
+      "grad_norm": 0.9219532608985901,
+      "learning_rate": 2.0142279033915726e-05,
+      "loss": 0.441,
+      "step": 6250
+    },
+    {
+      "epoch": 0.20024891498595865,
+      "grad_norm": 0.9598010778427124,
+      "learning_rate": 2.0122205806783147e-05,
+      "loss": 0.4856,
+      "step": 6275
+    },
+    {
+      "epoch": 0.20104671942813376,
+      "grad_norm": 1.0245938301086426,
+      "learning_rate": 2.0102132579650564e-05,
+      "loss": 0.449,
+      "step": 6300
+    },
+    {
+      "epoch": 0.2018445238703089,
+      "grad_norm": 1.4804080724716187,
+      "learning_rate": 2.008205935251799e-05,
+      "loss": 0.5155,
+      "step": 6325
+    },
+    {
+      "epoch": 0.20264232831248405,
+      "grad_norm": 0.5288428664207458,
+      "learning_rate": 2.0061986125385407e-05,
+      "loss": 0.4514,
+      "step": 6350
+    },
+    {
+      "epoch": 0.20344013275465916,
+      "grad_norm": 0.9050682783126831,
+      "learning_rate": 2.0041912898252828e-05,
+      "loss": 0.4946,
+      "step": 6375
+    },
+    {
+      "epoch": 0.2042379371968343,
+      "grad_norm": 1.1348626613616943,
+      "learning_rate": 2.002183967112025e-05,
+      "loss": 0.492,
+      "step": 6400
+    },
+    {
+      "epoch": 0.20503574163900945,
+      "grad_norm": 0.9863408803939819,
+      "learning_rate": 2.000176644398767e-05,
+      "loss": 0.4849,
+      "step": 6425
+    },
+    {
+      "epoch": 0.2058335460811846,
+      "grad_norm": 0.8158233165740967,
+      "learning_rate": 1.9981693216855088e-05,
+      "loss": 0.4371,
+      "step": 6450
+    },
+    {
+      "epoch": 0.2066313505233597,
+      "grad_norm": 0.5699653625488281,
+      "learning_rate": 1.996161998972251e-05,
+      "loss": 0.4869,
+      "step": 6475
+    },
+    {
+      "epoch": 0.20742915496553485,
+      "grad_norm": 0.6135886311531067,
+      "learning_rate": 1.994154676258993e-05,
+      "loss": 0.4866,
+      "step": 6500
+    },
+    {
+      "epoch": 0.20822695940771,
+      "grad_norm": 0.8482078909873962,
+      "learning_rate": 1.9921473535457348e-05,
+      "loss": 0.5002,
+      "step": 6525
+    },
+    {
+      "epoch": 0.2090247638498851,
+      "grad_norm": 0.9109760522842407,
+      "learning_rate": 1.990140030832477e-05,
+      "loss": 0.4985,
+      "step": 6550
+    },
+    {
+      "epoch": 0.20982256829206025,
+      "grad_norm": 0.7344380617141724,
+      "learning_rate": 1.988132708119219e-05,
+      "loss": 0.494,
+      "step": 6575
+    },
+    {
+      "epoch": 0.2106203727342354,
+      "grad_norm": 0.7968775629997253,
+      "learning_rate": 1.986125385405961e-05,
+      "loss": 0.5183,
+      "step": 6600
+    },
+    {
+      "epoch": 0.2114181771764105,
+      "grad_norm": 1.089021921157837,
+      "learning_rate": 1.984118062692703e-05,
+      "loss": 0.5522,
+      "step": 6625
+    },
+    {
+      "epoch": 0.21221598161858565,
+      "grad_norm": 0.886903703212738,
+      "learning_rate": 1.982110739979445e-05,
+      "loss": 0.516,
+      "step": 6650
+    },
+    {
+      "epoch": 0.2130137860607608,
+      "grad_norm": 0.8863084316253662,
+      "learning_rate": 1.980103417266187e-05,
+      "loss": 0.4592,
+      "step": 6675
+    },
+    {
+      "epoch": 0.2138115905029359,
+      "grad_norm": 0.4921455383300781,
+      "learning_rate": 1.9780960945529292e-05,
+      "loss": 0.483,
+      "step": 6700
+    },
+    {
+      "epoch": 0.21460939494511105,
+      "grad_norm": 0.9074678421020508,
+      "learning_rate": 1.976088771839671e-05,
+      "loss": 0.5242,
+      "step": 6725
+    },
+    {
+      "epoch": 0.2154071993872862,
+      "grad_norm": 0.6239631772041321,
+      "learning_rate": 1.974081449126413e-05,
+      "loss": 0.4701,
+      "step": 6750
+    },
+    {
+      "epoch": 0.2162050038294613,
+      "grad_norm": 0.6419113278388977,
+      "learning_rate": 1.9720741264131552e-05,
+      "loss": 0.5541,
+      "step": 6775
+    },
+    {
+      "epoch": 0.21700280827163645,
+      "grad_norm": 1.3911529779434204,
+      "learning_rate": 1.9700668036998974e-05,
+      "loss": 0.4585,
+      "step": 6800
+    },
+    {
+      "epoch": 0.2178006127138116,
+      "grad_norm": 1.0443439483642578,
+      "learning_rate": 1.968059480986639e-05,
+      "loss": 0.4677,
+      "step": 6825
+    },
+    {
+      "epoch": 0.2185984171559867,
+      "grad_norm": 1.1934226751327515,
+      "learning_rate": 1.9660521582733816e-05,
+      "loss": 0.4793,
+      "step": 6850
+    },
+    {
+      "epoch": 0.21939622159816186,
+      "grad_norm": 0.3158310055732727,
+      "learning_rate": 1.9640448355601234e-05,
+      "loss": 0.5371,
+      "step": 6875
+    },
+    {
+      "epoch": 0.220194026040337,
+      "grad_norm": 0.8096803426742554,
+      "learning_rate": 1.9620375128468655e-05,
+      "loss": 0.6083,
+      "step": 6900
+    },
+    {
+      "epoch": 0.2209918304825121,
+      "grad_norm": 0.4177459478378296,
+      "learning_rate": 1.9600301901336072e-05,
+      "loss": 0.5113,
+      "step": 6925
+    },
+    {
+      "epoch": 0.22178963492468726,
+      "grad_norm": 0.9590117931365967,
+      "learning_rate": 1.9580228674203497e-05,
+      "loss": 0.4658,
+      "step": 6950
+    },
+    {
+      "epoch": 0.2225874393668624,
+      "grad_norm": 1.7520601749420166,
+      "learning_rate": 1.9560155447070915e-05,
+      "loss": 0.5562,
+      "step": 6975
+    },
+    {
+      "epoch": 0.22338524380903754,
+      "grad_norm": 0.8236017823219299,
+      "learning_rate": 1.9540082219938336e-05,
+      "loss": 0.4285,
+      "step": 7000
+    },
+    {
+      "epoch": 0.22418304825121266,
+      "grad_norm": 0.48859933018684387,
+      "learning_rate": 1.9520008992805754e-05,
+      "loss": 0.4806,
+      "step": 7025
+    },
+    {
+      "epoch": 0.2249808526933878,
+      "grad_norm": 1.1268525123596191,
+      "learning_rate": 1.9499935765673178e-05,
+      "loss": 0.5602,
+      "step": 7050
+    },
+    {
+      "epoch": 0.22577865713556294,
+      "grad_norm": 1.7955353260040283,
+      "learning_rate": 1.9479862538540596e-05,
+      "loss": 0.4894,
+      "step": 7075
+    },
+    {
+      "epoch": 0.22657646157773806,
+      "grad_norm": 0.9489879012107849,
+      "learning_rate": 1.9459789311408017e-05,
+      "loss": 0.5958,
+      "step": 7100
+    },
+    {
+      "epoch": 0.2273742660199132,
+      "grad_norm": 0.6150937080383301,
+      "learning_rate": 1.9439716084275438e-05,
+      "loss": 0.4284,
+      "step": 7125
+    },
+    {
+      "epoch": 0.22817207046208834,
+      "grad_norm": 0.865825891494751,
+      "learning_rate": 1.941964285714286e-05,
+      "loss": 0.5286,
+      "step": 7150
+    },
+    {
+      "epoch": 0.22896987490426346,
+      "grad_norm": 0.8219313621520996,
+      "learning_rate": 1.9399569630010277e-05,
+      "loss": 0.4516,
+      "step": 7175
+    },
+    {
+      "epoch": 0.2297676793464386,
+      "grad_norm": 0.9921424984931946,
+      "learning_rate": 1.9379496402877698e-05,
+      "loss": 0.4889,
+      "step": 7200
+    },
+    {
+      "epoch": 0.23056548378861375,
+      "grad_norm": 0.645586371421814,
+      "learning_rate": 1.935942317574512e-05,
+      "loss": 0.5168,
+      "step": 7225
+    },
+    {
+      "epoch": 0.23136328823078886,
+      "grad_norm": 0.7225667238235474,
+      "learning_rate": 1.933934994861254e-05,
+      "loss": 0.5267,
+      "step": 7250
+    },
+    {
+      "epoch": 0.232161092672964,
+      "grad_norm": 0.6475264430046082,
+      "learning_rate": 1.9319276721479958e-05,
+      "loss": 0.4812,
+      "step": 7275
+    },
+    {
+      "epoch": 0.23295889711513915,
+      "grad_norm": 0.6617599725723267,
+      "learning_rate": 1.9299203494347383e-05,
+      "loss": 0.4795,
+      "step": 7300
+    },
+    {
+      "epoch": 0.23375670155731426,
+      "grad_norm": 0.8409857153892517,
+      "learning_rate": 1.92791302672148e-05,
+      "loss": 0.501,
+      "step": 7325
+    },
+    {
+      "epoch": 0.2345545059994894,
+      "grad_norm": 0.6069079637527466,
+      "learning_rate": 1.925905704008222e-05,
+      "loss": 0.4619,
+      "step": 7350
+    },
+    {
+      "epoch": 0.23535231044166455,
+      "grad_norm": 0.4156222343444824,
+      "learning_rate": 1.923898381294964e-05,
+      "loss": 0.5283,
+      "step": 7375
+    },
+    {
+      "epoch": 0.23615011488383966,
+      "grad_norm": 0.8497269749641418,
+      "learning_rate": 1.9218910585817064e-05,
+      "loss": 0.5375,
+      "step": 7400
+    },
+    {
+      "epoch": 0.2369479193260148,
+      "grad_norm": 0.9058747887611389,
+      "learning_rate": 1.919883735868448e-05,
+      "loss": 0.522,
+      "step": 7425
+    },
+    {
+      "epoch": 0.23774572376818995,
+      "grad_norm": 0.41128915548324585,
+      "learning_rate": 1.9178764131551903e-05,
+      "loss": 0.517,
+      "step": 7450
+    },
+    {
+      "epoch": 0.23854352821036506,
+      "grad_norm": 0.824787974357605,
+      "learning_rate": 1.915869090441932e-05,
+      "loss": 0.4799,
+      "step": 7475
+    },
+    {
+      "epoch": 0.2393413326525402,
+      "grad_norm": 1.1245702505111694,
+      "learning_rate": 1.9138617677286745e-05,
+      "loss": 0.4543,
+      "step": 7500
+    },
+    {
+      "epoch": 0.24013913709471535,
+      "grad_norm": 1.861336350440979,
+      "learning_rate": 1.9118544450154163e-05,
+      "loss": 0.4628,
+      "step": 7525
+    },
+    {
+      "epoch": 0.24093694153689046,
+      "grad_norm": 1.8422930240631104,
+      "learning_rate": 1.9098471223021584e-05,
+      "loss": 0.4926,
+      "step": 7550
+    },
+    {
+      "epoch": 0.2417347459790656,
+      "grad_norm": 0.821029007434845,
+      "learning_rate": 1.9078397995889005e-05,
+      "loss": 0.5141,
+      "step": 7575
+    },
+    {
+      "epoch": 0.24253255042124075,
+      "grad_norm": 0.7867022752761841,
+      "learning_rate": 1.9058324768756426e-05,
+      "loss": 0.4917,
+      "step": 7600
+    },
+    {
+      "epoch": 0.2433303548634159,
+      "grad_norm": 0.5639078617095947,
+      "learning_rate": 1.9038251541623844e-05,
+      "loss": 0.4945,
+      "step": 7625
+    },
+    {
+      "epoch": 0.244128159305591,
+      "grad_norm": 1.0543819665908813,
+      "learning_rate": 1.9018178314491265e-05,
+      "loss": 0.5713,
+      "step": 7650
+    },
+    {
+      "epoch": 0.24492596374776615,
+      "grad_norm": 0.8011125326156616,
+      "learning_rate": 1.8998105087358686e-05,
+      "loss": 0.4423,
+      "step": 7675
+    },
+    {
+      "epoch": 0.2457237681899413,
+      "grad_norm": 0.4197518527507782,
+      "learning_rate": 1.8978031860226107e-05,
+      "loss": 0.4802,
+      "step": 7700
+    },
+    {
+      "epoch": 0.2465215726321164,
+      "grad_norm": 0.43537619709968567,
+      "learning_rate": 1.8957958633093525e-05,
+      "loss": 0.491,
+      "step": 7725
+    },
+    {
+      "epoch": 0.24731937707429155,
+      "grad_norm": 0.9926440119743347,
+      "learning_rate": 1.8937885405960946e-05,
+      "loss": 0.5206,
+      "step": 7750
+    },
+    {
+      "epoch": 0.2481171815164667,
+      "grad_norm": 0.6548684239387512,
+      "learning_rate": 1.8917812178828367e-05,
+      "loss": 0.5278,
+      "step": 7775
+    },
+    {
+      "epoch": 0.2489149859586418,
+      "grad_norm": 0.6337385773658752,
+      "learning_rate": 1.889773895169579e-05,
+      "loss": 0.5716,
+      "step": 7800
+    },
+    {
+      "epoch": 0.24971279040081695,
+      "grad_norm": 0.8676078915596008,
+      "learning_rate": 1.8877665724563206e-05,
+      "loss": 0.4849,
+      "step": 7825
+    },
+    {
+      "epoch": 0.2505105948429921,
+      "grad_norm": 0.860718846321106,
+      "learning_rate": 1.885759249743063e-05,
+      "loss": 0.5069,
+      "step": 7850
+    },
+    {
+      "epoch": 0.25130839928516724,
+      "grad_norm": 0.6479659080505371,
+      "learning_rate": 1.883751927029805e-05,
+      "loss": 0.4364,
+      "step": 7875
+    },
+    {
+      "epoch": 0.2521062037273423,
+      "grad_norm": 0.5919177532196045,
+      "learning_rate": 1.881744604316547e-05,
+      "loss": 0.5035,
+      "step": 7900
+    },
+    {
+      "epoch": 0.25290400816951747,
+      "grad_norm": 0.6908952593803406,
+      "learning_rate": 1.8797372816032887e-05,
+      "loss": 0.4602,
+      "step": 7925
+    },
+    {
+      "epoch": 0.2537018126116926,
+      "grad_norm": 0.9885756373405457,
+      "learning_rate": 1.8777299588900312e-05,
+      "loss": 0.52,
+      "step": 7950
+    },
+    {
+      "epoch": 0.25449961705386775,
+      "grad_norm": 1.0296975374221802,
+      "learning_rate": 1.875722636176773e-05,
+      "loss": 0.542,
+      "step": 7975
+    },
+    {
+      "epoch": 0.2552974214960429,
+      "grad_norm": 0.44542399048805237,
+      "learning_rate": 1.873715313463515e-05,
+      "loss": 0.4894,
+      "step": 8000
+    },
+    {
+      "epoch": 0.25609522593821804,
+      "grad_norm": 1.1537628173828125,
+      "learning_rate": 1.8717079907502572e-05,
+      "loss": 0.5045,
+      "step": 8025
+    },
+    {
+      "epoch": 0.2568930303803932,
+      "grad_norm": 0.4203762114048004,
+      "learning_rate": 1.869700668036999e-05,
+      "loss": 0.4983,
+      "step": 8050
+    },
+    {
+      "epoch": 0.25769083482256827,
+      "grad_norm": 0.9488811492919922,
+      "learning_rate": 1.867693345323741e-05,
+      "loss": 0.4102,
+      "step": 8075
+    },
+    {
+      "epoch": 0.2584886392647434,
+      "grad_norm": 1.0219916105270386,
+      "learning_rate": 1.8656860226104832e-05,
+      "loss": 0.5188,
+      "step": 8100
+    },
+    {
+      "epoch": 0.25928644370691856,
+      "grad_norm": 2.179837465286255,
+      "learning_rate": 1.8636786998972253e-05,
+      "loss": 0.4998,
+      "step": 8125
+    },
+    {
+      "epoch": 0.2600842481490937,
+      "grad_norm": 0.5699672102928162,
+      "learning_rate": 1.861671377183967e-05,
+      "loss": 0.5445,
+      "step": 8150
+    },
+    {
+      "epoch": 0.26088205259126884,
+      "grad_norm": 1.0542255640029907,
+      "learning_rate": 1.8596640544707092e-05,
+      "loss": 0.4803,
+      "step": 8175
+    },
+    {
+      "epoch": 0.261679857033444,
+      "grad_norm": 0.6809638142585754,
+      "learning_rate": 1.8576567317574513e-05,
+      "loss": 0.4494,
+      "step": 8200
+    },
+    {
+      "epoch": 0.2624776614756191,
+      "grad_norm": 0.8445626497268677,
+      "learning_rate": 1.8556494090441934e-05,
+      "loss": 0.5055,
+      "step": 8225
+    },
+    {
+      "epoch": 0.2632754659177942,
+      "grad_norm": 0.6919910311698914,
+      "learning_rate": 1.8536420863309352e-05,
+      "loss": 0.448,
+      "step": 8250
+    },
+    {
+      "epoch": 0.26407327035996936,
+      "grad_norm": 1.0985883474349976,
+      "learning_rate": 1.8516347636176773e-05,
+      "loss": 0.4849,
+      "step": 8275
+    },
+    {
+      "epoch": 0.2648710748021445,
+      "grad_norm": 0.7515255212783813,
+      "learning_rate": 1.8496274409044194e-05,
+      "loss": 0.4677,
+      "step": 8300
+    },
+    {
+      "epoch": 0.26566887924431964,
+      "grad_norm": 0.825547993183136,
+      "learning_rate": 1.8476201181911615e-05,
+      "loss": 0.5624,
+      "step": 8325
+    },
+    {
+      "epoch": 0.2664666836864948,
+      "grad_norm": 1.1458021402359009,
+      "learning_rate": 1.8456127954779033e-05,
+      "loss": 0.4882,
+      "step": 8350
+    },
+    {
+      "epoch": 0.2672644881286699,
+      "grad_norm": 0.3550206422805786,
+      "learning_rate": 1.8436054727646454e-05,
+      "loss": 0.4423,
+      "step": 8375
+    },
+    {
+      "epoch": 0.268062292570845,
+      "grad_norm": 0.5570346713066101,
+      "learning_rate": 1.8415981500513875e-05,
+      "loss": 0.565,
+      "step": 8400
+    },
+    {
+      "epoch": 0.26886009701302016,
+      "grad_norm": 0.5592429637908936,
+      "learning_rate": 1.8395908273381296e-05,
+      "loss": 0.4471,
+      "step": 8425
+    },
+    {
+      "epoch": 0.2696579014551953,
+      "grad_norm": 0.3842066526412964,
+      "learning_rate": 1.8375835046248714e-05,
+      "loss": 0.5283,
+      "step": 8450
+    },
+    {
+      "epoch": 0.27045570589737045,
+      "grad_norm": 0.478157103061676,
+      "learning_rate": 1.835576181911614e-05,
+      "loss": 0.4916,
+      "step": 8475
+    },
+    {
+      "epoch": 0.2712535103395456,
+      "grad_norm": 1.1027177572250366,
+      "learning_rate": 1.8335688591983556e-05,
+      "loss": 0.5794,
+      "step": 8500
+    },
+    {
+      "epoch": 0.27205131478172073,
+      "grad_norm": 1.0234731435775757,
+      "learning_rate": 1.8315615364850977e-05,
+      "loss": 0.4383,
+      "step": 8525
+    },
+    {
+      "epoch": 0.2728491192238958,
+      "grad_norm": 0.9634427428245544,
+      "learning_rate": 1.8295542137718395e-05,
+      "loss": 0.4818,
+      "step": 8550
+    },
+    {
+      "epoch": 0.27364692366607096,
+      "grad_norm": 0.6669744253158569,
+      "learning_rate": 1.827546891058582e-05,
+      "loss": 0.5153,
+      "step": 8575
+    },
+    {
+      "epoch": 0.2744447281082461,
+      "grad_norm": 0.6819443106651306,
+      "learning_rate": 1.8255395683453237e-05,
+      "loss": 0.4161,
+      "step": 8600
+    },
+    {
+      "epoch": 0.27524253255042125,
+      "grad_norm": 0.6380025148391724,
+      "learning_rate": 1.823532245632066e-05,
+      "loss": 0.4378,
+      "step": 8625
+    },
+    {
+      "epoch": 0.2760403369925964,
+      "grad_norm": 1.0884222984313965,
+      "learning_rate": 1.8215249229188076e-05,
+      "loss": 0.4557,
+      "step": 8650
+    },
+    {
+      "epoch": 0.27683814143477153,
+      "grad_norm": 0.3991362750530243,
+      "learning_rate": 1.81951760020555e-05,
+      "loss": 0.4468,
+      "step": 8675
+    },
+    {
+      "epoch": 0.2776359458769466,
+      "grad_norm": 0.6306272149085999,
+      "learning_rate": 1.817510277492292e-05,
+      "loss": 0.4402,
+      "step": 8700
+    },
+    {
+      "epoch": 0.27843375031912176,
+      "grad_norm": 0.9752461910247803,
+      "learning_rate": 1.815502954779034e-05,
+      "loss": 0.502,
+      "step": 8725
+    },
+    {
+      "epoch": 0.2792315547612969,
+      "grad_norm": 1.1096750497817993,
+      "learning_rate": 1.813495632065776e-05,
+      "loss": 0.4984,
+      "step": 8750
+    },
+    {
+      "epoch": 0.28002935920347205,
+      "grad_norm": 0.7078149914741516,
+      "learning_rate": 1.8114883093525182e-05,
+      "loss": 0.4901,
+      "step": 8775
+    },
+    {
+      "epoch": 0.2808271636456472,
+      "grad_norm": 0.8923904299736023,
+      "learning_rate": 1.80948098663926e-05,
+      "loss": 0.4476,
+      "step": 8800
+    },
+    {
+      "epoch": 0.28162496808782234,
+      "grad_norm": 1.1634156703948975,
+      "learning_rate": 1.807473663926002e-05,
+      "loss": 0.5004,
+      "step": 8825
+    },
+    {
+      "epoch": 0.2824227725299974,
+      "grad_norm": 0.6043126583099365,
+      "learning_rate": 1.8054663412127442e-05,
+      "loss": 0.4892,
+      "step": 8850
+    },
+    {
+      "epoch": 0.28322057697217257,
+      "grad_norm": 0.7623210549354553,
+      "learning_rate": 1.8034590184994863e-05,
+      "loss": 0.4507,
+      "step": 8875
+    },
+    {
+      "epoch": 0.2840183814143477,
+      "grad_norm": 0.7996333837509155,
+      "learning_rate": 1.801451695786228e-05,
+      "loss": 0.4039,
+      "step": 8900
+    },
+    {
+      "epoch": 0.28481618585652285,
+      "grad_norm": 1.5519790649414062,
+      "learning_rate": 1.7994443730729702e-05,
+      "loss": 0.4859,
+      "step": 8925
+    },
+    {
+      "epoch": 0.285613990298698,
+      "grad_norm": 0.7139633297920227,
+      "learning_rate": 1.7974370503597123e-05,
+      "loss": 0.466,
+      "step": 8950
+    },
+    {
+      "epoch": 0.28641179474087314,
+      "grad_norm": 1.1657685041427612,
+      "learning_rate": 1.7954297276464544e-05,
+      "loss": 0.4362,
+      "step": 8975
+    },
+    {
+      "epoch": 0.2872095991830482,
+      "grad_norm": 1.1409554481506348,
+      "learning_rate": 1.7934224049331962e-05,
+      "loss": 0.41,
+      "step": 9000
+    },
+    {
+      "epoch": 0.28800740362522337,
+      "grad_norm": 0.5146012902259827,
+      "learning_rate": 1.7914150822199387e-05,
+      "loss": 0.4733,
+      "step": 9025
+    },
+    {
+      "epoch": 0.2888052080673985,
+      "grad_norm": 1.1914423704147339,
+      "learning_rate": 1.7894077595066804e-05,
+      "loss": 0.505,
+      "step": 9050
+    },
+    {
+      "epoch": 0.28960301250957365,
+      "grad_norm": 0.5126356482505798,
+      "learning_rate": 1.7874004367934225e-05,
+      "loss": 0.4781,
+      "step": 9075
+    },
+    {
+      "epoch": 0.2904008169517488,
+      "grad_norm": 0.3558446168899536,
+      "learning_rate": 1.7853931140801643e-05,
+      "loss": 0.4813,
+      "step": 9100
+    },
+    {
+      "epoch": 0.29119862139392394,
+      "grad_norm": 0.9246950745582581,
+      "learning_rate": 1.7833857913669068e-05,
+      "loss": 0.4645,
+      "step": 9125
+    },
+    {
+      "epoch": 0.2919964258360991,
+      "grad_norm": 0.4734017550945282,
+      "learning_rate": 1.7813784686536485e-05,
+      "loss": 0.4598,
+      "step": 9150
+    },
+    {
+      "epoch": 0.29279423027827417,
+      "grad_norm": 0.7660199403762817,
+      "learning_rate": 1.7793711459403907e-05,
+      "loss": 0.5033,
+      "step": 9175
+    },
+    {
+      "epoch": 0.2935920347204493,
+      "grad_norm": 0.7290098071098328,
+      "learning_rate": 1.7773638232271328e-05,
+      "loss": 0.4783,
+      "step": 9200
+    },
+    {
+      "epoch": 0.29438983916262446,
+      "grad_norm": 0.7562490105628967,
+      "learning_rate": 1.775356500513875e-05,
+      "loss": 0.4961,
+      "step": 9225
+    },
+    {
+      "epoch": 0.2951876436047996,
+      "grad_norm": 0.31955620646476746,
+      "learning_rate": 1.7733491778006167e-05,
+      "loss": 0.5139,
+      "step": 9250
+    },
+    {
+      "epoch": 0.29598544804697474,
+      "grad_norm": 1.351357102394104,
+      "learning_rate": 1.7713418550873588e-05,
+      "loss": 0.4708,
+      "step": 9275
+    },
+    {
+      "epoch": 0.2967832524891499,
+      "grad_norm": 0.41533350944519043,
+      "learning_rate": 1.769334532374101e-05,
+      "loss": 0.4064,
+      "step": 9300
+    },
+    {
+      "epoch": 0.29758105693132497,
+      "grad_norm": 1.161253809928894,
+      "learning_rate": 1.767327209660843e-05,
+      "loss": 0.4731,
+      "step": 9325
+    },
+    {
+      "epoch": 0.2983788613735001,
+      "grad_norm": 0.7671948075294495,
+      "learning_rate": 1.7653198869475848e-05,
+      "loss": 0.491,
+      "step": 9350
+    },
+    {
+      "epoch": 0.29917666581567526,
+      "grad_norm": 0.7606444954872131,
+      "learning_rate": 1.763312564234327e-05,
+      "loss": 0.5046,
+      "step": 9375
+    },
+    {
+      "epoch": 0.2999744702578504,
+      "grad_norm": 1.3677698373794556,
+      "learning_rate": 1.761305241521069e-05,
+      "loss": 0.3982,
+      "step": 9400
+    },
+    {
+      "epoch": 0.30077227470002554,
+      "grad_norm": 0.811142086982727,
+      "learning_rate": 1.759297918807811e-05,
+      "loss": 0.4427,
+      "step": 9425
+    },
+    {
+      "epoch": 0.3015700791422007,
+      "grad_norm": 0.864321768283844,
+      "learning_rate": 1.757290596094553e-05,
+      "loss": 0.4531,
+      "step": 9450
+    },
+    {
+      "epoch": 0.3023678835843758,
+      "grad_norm": 1.6318188905715942,
+      "learning_rate": 1.7552832733812953e-05,
+      "loss": 0.499,
+      "step": 9475
+    },
+    {
+      "epoch": 0.3031656880265509,
+      "grad_norm": 1.0822746753692627,
+      "learning_rate": 1.753275950668037e-05,
+      "loss": 0.546,
+      "step": 9500
+    },
+    {
+      "epoch": 0.30396349246872606,
+      "grad_norm": 0.5256426334381104,
+      "learning_rate": 1.7512686279547792e-05,
+      "loss": 0.4819,
+      "step": 9525
+    },
+    {
+      "epoch": 0.3047612969109012,
+      "grad_norm": 1.2467024326324463,
+      "learning_rate": 1.749261305241521e-05,
+      "loss": 0.4946,
+      "step": 9550
+    },
+    {
+      "epoch": 0.30555910135307635,
+      "grad_norm": 0.6196355819702148,
+      "learning_rate": 1.747253982528263e-05,
+      "loss": 0.4924,
+      "step": 9575
+    },
+    {
+      "epoch": 0.3063569057952515,
+      "grad_norm": 0.48964831233024597,
+      "learning_rate": 1.7452466598150052e-05,
+      "loss": 0.4554,
+      "step": 9600
+    },
+    {
+      "epoch": 0.3071547102374266,
+      "grad_norm": 1.2648134231567383,
+      "learning_rate": 1.7432393371017473e-05,
+      "loss": 0.5184,
+      "step": 9625
+    },
+    {
+      "epoch": 0.3079525146796017,
+      "grad_norm": 0.8756480813026428,
+      "learning_rate": 1.7412320143884895e-05,
+      "loss": 0.4393,
+      "step": 9650
+    },
+    {
+      "epoch": 0.30875031912177686,
+      "grad_norm": 0.6415950059890747,
+      "learning_rate": 1.7392246916752312e-05,
+      "loss": 0.519,
+      "step": 9675
+    },
+    {
+      "epoch": 0.309548123563952,
+      "grad_norm": 0.8496715426445007,
+      "learning_rate": 1.7372173689619733e-05,
+      "loss": 0.5031,
+      "step": 9700
+    },
+    {
+      "epoch": 0.31034592800612715,
+      "grad_norm": 0.42665472626686096,
+      "learning_rate": 1.7352100462487155e-05,
+      "loss": 0.5048,
+      "step": 9725
+    },
+    {
+      "epoch": 0.3111437324483023,
+      "grad_norm": 0.6046943664550781,
+      "learning_rate": 1.7332027235354576e-05,
+      "loss": 0.499,
+      "step": 9750
+    },
+    {
+      "epoch": 0.31194153689047743,
+      "grad_norm": 0.6538355350494385,
+      "learning_rate": 1.7311954008221993e-05,
+      "loss": 0.4287,
+      "step": 9775
+    },
+    {
+      "epoch": 0.3127393413326525,
+      "grad_norm": 0.4587744176387787,
+      "learning_rate": 1.7291880781089415e-05,
+      "loss": 0.4803,
+      "step": 9800
+    },
+    {
+      "epoch": 0.31353714577482766,
+      "grad_norm": 0.9619668126106262,
+      "learning_rate": 1.7271807553956836e-05,
+      "loss": 0.4662,
+      "step": 9825
+    },
+    {
+      "epoch": 0.3143349502170028,
+      "grad_norm": 0.7719420194625854,
+      "learning_rate": 1.7251734326824257e-05,
+      "loss": 0.5092,
+      "step": 9850
+    },
+    {
+      "epoch": 0.31513275465917795,
+      "grad_norm": 0.7614120841026306,
+      "learning_rate": 1.7231661099691675e-05,
+      "loss": 0.4709,
+      "step": 9875
+    },
+    {
+      "epoch": 0.3159305591013531,
+      "grad_norm": 1.1873565912246704,
+      "learning_rate": 1.7211587872559096e-05,
+      "loss": 0.4459,
+      "step": 9900
+    },
+    {
+      "epoch": 0.31672836354352824,
+      "grad_norm": 0.6782514452934265,
+      "learning_rate": 1.7191514645426517e-05,
+      "loss": 0.4737,
+      "step": 9925
+    },
+    {
+      "epoch": 0.3175261679857033,
+      "grad_norm": 0.46671539545059204,
+      "learning_rate": 1.7171441418293938e-05,
+      "loss": 0.5144,
+      "step": 9950
+    },
+    {
+      "epoch": 0.31832397242787847,
+      "grad_norm": 0.7853887677192688,
+      "learning_rate": 1.7151368191161356e-05,
+      "loss": 0.5039,
+      "step": 9975
+    },
+    {
+      "epoch": 0.3191217768700536,
+      "grad_norm": 0.9797842502593994,
+      "learning_rate": 1.7131294964028777e-05,
+      "loss": 0.5097,
+      "step": 10000
+    },
+    {
+      "epoch": 0.31991958131222875,
+      "grad_norm": 1.0205621719360352,
+      "learning_rate": 1.7111221736896198e-05,
+      "loss": 0.4735,
+      "step": 10025
+    },
+    {
+      "epoch": 0.3207173857544039,
+      "grad_norm": 0.7101638913154602,
+      "learning_rate": 1.709114850976362e-05,
+      "loss": 0.5876,
+      "step": 10050
+    },
+    {
+      "epoch": 0.32151519019657904,
+      "grad_norm": 1.0455431938171387,
+      "learning_rate": 1.7071075282631037e-05,
+      "loss": 0.5183,
+      "step": 10075
+    },
+    {
+      "epoch": 0.3223129946387541,
+      "grad_norm": 1.2804579734802246,
+      "learning_rate": 1.7051002055498458e-05,
+      "loss": 0.472,
+      "step": 10100
+    },
+    {
+      "epoch": 0.32311079908092927,
+      "grad_norm": 0.707480788230896,
+      "learning_rate": 1.703092882836588e-05,
+      "loss": 0.4995,
+      "step": 10125
+    },
+    {
+      "epoch": 0.3239086035231044,
+      "grad_norm": 0.5702756643295288,
+      "learning_rate": 1.70108556012333e-05,
+      "loss": 0.5285,
+      "step": 10150
+    },
+    {
+      "epoch": 0.32470640796527955,
+      "grad_norm": 1.105216383934021,
+      "learning_rate": 1.6990782374100718e-05,
+      "loss": 0.4843,
+      "step": 10175
+    },
+    {
+      "epoch": 0.3255042124074547,
+      "grad_norm": 0.6410677433013916,
+      "learning_rate": 1.6970709146968142e-05,
+      "loss": 0.4855,
+      "step": 10200
+    },
+    {
+      "epoch": 0.32630201684962984,
+      "grad_norm": 1.1038342714309692,
+      "learning_rate": 1.695063591983556e-05,
+      "loss": 0.4942,
+      "step": 10225
+    },
+    {
+      "epoch": 0.3270998212918049,
+      "grad_norm": 0.8862643837928772,
+      "learning_rate": 1.693056269270298e-05,
+      "loss": 0.5562,
+      "step": 10250
+    },
+    {
+      "epoch": 0.32789762573398007,
+      "grad_norm": 0.8062041997909546,
+      "learning_rate": 1.69104894655704e-05,
+      "loss": 0.5021,
+      "step": 10275
+    },
+    {
+      "epoch": 0.3286954301761552,
+      "grad_norm": 0.8131182789802551,
+      "learning_rate": 1.6890416238437824e-05,
+      "loss": 0.47,
+      "step": 10300
+    },
+    {
+      "epoch": 0.32949323461833036,
+      "grad_norm": 0.7970181703567505,
+      "learning_rate": 1.687034301130524e-05,
+      "loss": 0.4677,
+      "step": 10325
+    },
+    {
+      "epoch": 0.3302910390605055,
+      "grad_norm": 0.6498174071311951,
+      "learning_rate": 1.6850269784172663e-05,
+      "loss": 0.4772,
+      "step": 10350
+    },
+    {
+      "epoch": 0.33108884350268064,
+      "grad_norm": 0.6998928189277649,
+      "learning_rate": 1.6830196557040084e-05,
+      "loss": 0.4691,
+      "step": 10375
+    },
+    {
+      "epoch": 0.3318866479448558,
+      "grad_norm": 0.653899073600769,
+      "learning_rate": 1.6810123329907505e-05,
+      "loss": 0.4978,
+      "step": 10400
+    },
+    {
+      "epoch": 0.33268445238703087,
+      "grad_norm": 0.8953115344047546,
+      "learning_rate": 1.6790050102774923e-05,
+      "loss": 0.4758,
+      "step": 10425
+    },
+    {
+      "epoch": 0.333482256829206,
+      "grad_norm": 1.1513452529907227,
+      "learning_rate": 1.6769976875642344e-05,
+      "loss": 0.4322,
+      "step": 10450
+    },
+    {
+      "epoch": 0.33428006127138116,
+      "grad_norm": 0.9625234007835388,
+      "learning_rate": 1.6749903648509765e-05,
+      "loss": 0.5192,
+      "step": 10475
+    },
+    {
+      "epoch": 0.3350778657135563,
+      "grad_norm": 0.9107385277748108,
+      "learning_rate": 1.6729830421377186e-05,
+      "loss": 0.5478,
+      "step": 10500
+    },
+    {
+      "epoch": 0.33587567015573144,
+      "grad_norm": 1.4335834980010986,
+      "learning_rate": 1.6709757194244604e-05,
+      "loss": 0.482,
+      "step": 10525
+    },
+    {
+      "epoch": 0.3366734745979066,
+      "grad_norm": 0.3528570830821991,
+      "learning_rate": 1.6689683967112025e-05,
+      "loss": 0.4542,
+      "step": 10550
+    },
+    {
+      "epoch": 0.3374712790400817,
+      "grad_norm": 0.7097840309143066,
+      "learning_rate": 1.6669610739979446e-05,
+      "loss": 0.4464,
+      "step": 10575
+    },
+    {
+      "epoch": 0.3382690834822568,
+      "grad_norm": 0.835563063621521,
+      "learning_rate": 1.6649537512846867e-05,
+      "loss": 0.4649,
+      "step": 10600
+    },
+    {
+      "epoch": 0.33906688792443196,
+      "grad_norm": 0.5886844992637634,
+      "learning_rate": 1.6629464285714285e-05,
+      "loss": 0.5257,
+      "step": 10625
+    },
+    {
+      "epoch": 0.3398646923666071,
+      "grad_norm": 1.1818501949310303,
+      "learning_rate": 1.660939105858171e-05,
+      "loss": 0.4614,
+      "step": 10650
+    },
+    {
+      "epoch": 0.34066249680878224,
+      "grad_norm": 1.0059629678726196,
+      "learning_rate": 1.6589317831449127e-05,
+      "loss": 0.4473,
+      "step": 10675
+    },
+    {
+      "epoch": 0.3414603012509574,
+      "grad_norm": 0.6061350107192993,
+      "learning_rate": 1.6569244604316548e-05,
+      "loss": 0.4962,
+      "step": 10700
+    },
+    {
+      "epoch": 0.3422581056931325,
+      "grad_norm": 0.7166891694068909,
+      "learning_rate": 1.6549171377183966e-05,
+      "loss": 0.5111,
+      "step": 10725
+    },
+    {
+      "epoch": 0.3430559101353076,
+      "grad_norm": 1.4252043962478638,
+      "learning_rate": 1.652909815005139e-05,
+      "loss": 0.46,
+      "step": 10750
+    },
+    {
+      "epoch": 0.34385371457748276,
+      "grad_norm": 0.9708260893821716,
+      "learning_rate": 1.6509024922918808e-05,
+      "loss": 0.4923,
+      "step": 10775
+    },
+    {
+      "epoch": 0.3446515190196579,
+      "grad_norm": 1.8117674589157104,
+      "learning_rate": 1.648895169578623e-05,
+      "loss": 0.4932,
+      "step": 10800
+    },
+    {
+      "epoch": 0.34544932346183305,
+      "grad_norm": 1.0717155933380127,
+      "learning_rate": 1.6468878468653647e-05,
+      "loss": 0.4686,
+      "step": 10825
+    },
+    {
+      "epoch": 0.3462471279040082,
+      "grad_norm": 1.0861729383468628,
+      "learning_rate": 1.644880524152107e-05,
+      "loss": 0.513,
+      "step": 10850
+    },
+    {
+      "epoch": 0.3470449323461833,
+      "grad_norm": 0.8470234274864197,
+      "learning_rate": 1.642873201438849e-05,
+      "loss": 0.3456,
+      "step": 10875
+    },
+    {
+      "epoch": 0.3478427367883584,
+      "grad_norm": 1.0085972547531128,
+      "learning_rate": 1.640865878725591e-05,
+      "loss": 0.4574,
+      "step": 10900
+    },
+    {
+      "epoch": 0.34864054123053356,
+      "grad_norm": 0.8033475279808044,
+      "learning_rate": 1.638858556012333e-05,
+      "loss": 0.6031,
+      "step": 10925
+    },
+    {
+      "epoch": 0.3494383456727087,
+      "grad_norm": 0.9935612082481384,
+      "learning_rate": 1.6368512332990753e-05,
+      "loss": 0.4731,
+      "step": 10950
+    },
+    {
+      "epoch": 0.35023615011488385,
+      "grad_norm": 0.6845170259475708,
+      "learning_rate": 1.634843910585817e-05,
+      "loss": 0.4265,
+      "step": 10975
+    },
+    {
+      "epoch": 0.351033954557059,
+      "grad_norm": 0.9165354371070862,
+      "learning_rate": 1.632836587872559e-05,
+      "loss": 0.4642,
+      "step": 11000
+    },
+    {
+      "epoch": 0.35183175899923413,
+      "grad_norm": 0.9118735194206238,
+      "learning_rate": 1.6308292651593013e-05,
+      "loss": 0.5264,
+      "step": 11025
+    },
+    {
+      "epoch": 0.3526295634414092,
+      "grad_norm": 0.4230351448059082,
+      "learning_rate": 1.6288219424460434e-05,
+      "loss": 0.5,
+      "step": 11050
+    },
+    {
+      "epoch": 0.35342736788358436,
+      "grad_norm": 0.555743396282196,
+      "learning_rate": 1.626814619732785e-05,
+      "loss": 0.4524,
+      "step": 11075
+    },
+    {
+      "epoch": 0.3542251723257595,
+      "grad_norm": 0.5540426969528198,
+      "learning_rate": 1.6248072970195273e-05,
+      "loss": 0.4752,
+      "step": 11100
+    },
+    {
+      "epoch": 0.35502297676793465,
+      "grad_norm": 0.6602437496185303,
+      "learning_rate": 1.6227999743062694e-05,
+      "loss": 0.4597,
+      "step": 11125
+    },
+    {
+      "epoch": 0.3558207812101098,
+      "grad_norm": 1.2678382396697998,
+      "learning_rate": 1.6207926515930115e-05,
+      "loss": 0.4918,
+      "step": 11150
+    },
+    {
+      "epoch": 0.35661858565228494,
+      "grad_norm": 0.7993351817131042,
+      "learning_rate": 1.6187853288797533e-05,
+      "loss": 0.4044,
+      "step": 11175
+    },
+    {
+      "epoch": 0.35741639009446,
+      "grad_norm": 0.4799681007862091,
+      "learning_rate": 1.6167780061664954e-05,
+      "loss": 0.4845,
+      "step": 11200
+    },
+    {
+      "epoch": 0.35821419453663517,
+      "grad_norm": 0.4807746708393097,
+      "learning_rate": 1.6147706834532375e-05,
+      "loss": 0.4885,
+      "step": 11225
+    },
+    {
+      "epoch": 0.3590119989788103,
+      "grad_norm": 0.6480098962783813,
+      "learning_rate": 1.6127633607399796e-05,
+      "loss": 0.4902,
+      "step": 11250
+    },
+    {
+      "epoch": 0.35980980342098545,
+      "grad_norm": 0.921784520149231,
+      "learning_rate": 1.6107560380267214e-05,
+      "loss": 0.4527,
+      "step": 11275
+    },
+    {
+      "epoch": 0.3606076078631606,
+      "grad_norm": 0.7924018502235413,
+      "learning_rate": 1.6087487153134635e-05,
+      "loss": 0.4114,
+      "step": 11300
+    },
+    {
+      "epoch": 0.36140541230533574,
+      "grad_norm": 0.680823564529419,
+      "learning_rate": 1.6067413926002056e-05,
+      "loss": 0.4844,
+      "step": 11325
+    },
+    {
+      "epoch": 0.3622032167475108,
+      "grad_norm": 0.6174274682998657,
+      "learning_rate": 1.6047340698869477e-05,
+      "loss": 0.4905,
+      "step": 11350
+    },
+    {
+      "epoch": 0.36300102118968597,
+      "grad_norm": 0.9357193112373352,
+      "learning_rate": 1.60272674717369e-05,
+      "loss": 0.4846,
+      "step": 11375
+    },
+    {
+      "epoch": 0.3637988256318611,
+      "grad_norm": 0.8845779895782471,
+      "learning_rate": 1.6007194244604316e-05,
+      "loss": 0.48,
+      "step": 11400
+    },
+    {
+      "epoch": 0.36459663007403625,
+      "grad_norm": 1.3138706684112549,
+      "learning_rate": 1.5987121017471737e-05,
+      "loss": 0.4544,
+      "step": 11425
+    },
+    {
+      "epoch": 0.3653944345162114,
+      "grad_norm": 0.6036775708198547,
+      "learning_rate": 1.596704779033916e-05,
+      "loss": 0.4878,
+      "step": 11450
+    },
+    {
+      "epoch": 0.36619223895838654,
+      "grad_norm": 0.646816611289978,
+      "learning_rate": 1.594697456320658e-05,
+      "loss": 0.5299,
+      "step": 11475
+    },
+    {
+      "epoch": 0.3669900434005616,
+      "grad_norm": 0.8093705177307129,
+      "learning_rate": 1.5926901336073997e-05,
+      "loss": 0.4264,
+      "step": 11500
+    },
+    {
+      "epoch": 0.36778784784273677,
+      "grad_norm": 0.7628122568130493,
+      "learning_rate": 1.590682810894142e-05,
+      "loss": 0.467,
+      "step": 11525
+    },
+    {
+      "epoch": 0.3685856522849119,
+      "grad_norm": 0.45453178882598877,
+      "learning_rate": 1.588675488180884e-05,
+      "loss": 0.514,
+      "step": 11550
+    },
+    {
+      "epoch": 0.36938345672708706,
+      "grad_norm": 0.6202525496482849,
+      "learning_rate": 1.586668165467626e-05,
+      "loss": 0.4646,
+      "step": 11575
+    },
+    {
+      "epoch": 0.3701812611692622,
+      "grad_norm": 0.7522080540657043,
+      "learning_rate": 1.584660842754368e-05,
+      "loss": 0.4502,
+      "step": 11600
+    },
+    {
+      "epoch": 0.37097906561143734,
+      "grad_norm": 0.9508293867111206,
+      "learning_rate": 1.58265352004111e-05,
+      "loss": 0.4064,
+      "step": 11625
+    },
+    {
+      "epoch": 0.3717768700536125,
+      "grad_norm": 1.5886996984481812,
+      "learning_rate": 1.580646197327852e-05,
+      "loss": 0.4133,
+      "step": 11650
+    },
+    {
+      "epoch": 0.37257467449578757,
+      "grad_norm": 0.9361369013786316,
+      "learning_rate": 1.5786388746145942e-05,
+      "loss": 0.4868,
+      "step": 11675
+    },
+    {
+      "epoch": 0.3733724789379627,
+      "grad_norm": 1.056349515914917,
+      "learning_rate": 1.576631551901336e-05,
+      "loss": 0.46,
+      "step": 11700
+    },
+    {
+      "epoch": 0.37417028338013786,
+      "grad_norm": 0.6318427324295044,
+      "learning_rate": 1.574624229188078e-05,
+      "loss": 0.48,
+      "step": 11725
+    },
+    {
+      "epoch": 0.374968087822313,
+      "grad_norm": 5.217873573303223,
+      "learning_rate": 1.5726169064748202e-05,
+      "loss": 0.5168,
+      "step": 11750
+    },
+    {
+      "epoch": 0.37576589226448814,
+      "grad_norm": 0.5438673496246338,
+      "learning_rate": 1.5706095837615623e-05,
+      "loss": 0.4524,
+      "step": 11775
+    },
+    {
+      "epoch": 0.3765636967066633,
+      "grad_norm": 0.9590070843696594,
+      "learning_rate": 1.568602261048304e-05,
+      "loss": 0.5341,
+      "step": 11800
+    },
+    {
+      "epoch": 0.3773615011488384,
+      "grad_norm": 1.6194452047348022,
+      "learning_rate": 1.5665949383350465e-05,
+      "loss": 0.5055,
+      "step": 11825
+    },
+    {
+      "epoch": 0.3781593055910135,
+      "grad_norm": 0.5337164998054504,
+      "learning_rate": 1.5645876156217883e-05,
+      "loss": 0.4146,
+      "step": 11850
+    },
+    {
+      "epoch": 0.37895711003318866,
+      "grad_norm": 0.4194962680339813,
+      "learning_rate": 1.5625802929085304e-05,
+      "loss": 0.4396,
+      "step": 11875
+    },
+    {
+      "epoch": 0.3797549144753638,
+      "grad_norm": 0.8845074772834778,
+      "learning_rate": 1.5605729701952722e-05,
+      "loss": 0.4649,
+      "step": 11900
+    },
+    {
+      "epoch": 0.38055271891753895,
+      "grad_norm": 0.4936656653881073,
+      "learning_rate": 1.5585656474820146e-05,
+      "loss": 0.5033,
+      "step": 11925
+    },
+    {
+      "epoch": 0.3813505233597141,
+      "grad_norm": 1.3536922931671143,
+      "learning_rate": 1.5565583247687564e-05,
+      "loss": 0.5048,
+      "step": 11950
+    },
+    {
+      "epoch": 0.3821483278018892,
+      "grad_norm": 0.27829474210739136,
+      "learning_rate": 1.5545510020554985e-05,
+      "loss": 0.4022,
+      "step": 11975
+    },
+    {
+      "epoch": 0.3829461322440643,
+      "grad_norm": 1.678222894668579,
+      "learning_rate": 1.5525436793422403e-05,
+      "loss": 0.4964,
+      "step": 12000
+    },
+    {
+      "epoch": 0.38374393668623946,
+      "grad_norm": 0.45845431089401245,
+      "learning_rate": 1.5505363566289828e-05,
+      "loss": 0.4632,
+      "step": 12025
+    },
+    {
+      "epoch": 0.3845417411284146,
+      "grad_norm": 0.9954239726066589,
+      "learning_rate": 1.5485290339157245e-05,
+      "loss": 0.5619,
+      "step": 12050
+    },
+    {
+      "epoch": 0.38533954557058975,
+      "grad_norm": 0.4554518461227417,
+      "learning_rate": 1.5465217112024666e-05,
+      "loss": 0.4775,
+      "step": 12075
+    },
+    {
+      "epoch": 0.3861373500127649,
+      "grad_norm": 1.5614060163497925,
+      "learning_rate": 1.5445143884892088e-05,
+      "loss": 0.4309,
+      "step": 12100
+    },
+    {
+      "epoch": 0.38693515445494,
+      "grad_norm": 0.525596559047699,
+      "learning_rate": 1.542507065775951e-05,
+      "loss": 0.4973,
+      "step": 12125
+    },
+    {
+      "epoch": 0.3877329588971151,
+      "grad_norm": 1.0024532079696655,
+      "learning_rate": 1.5404997430626926e-05,
+      "loss": 0.4464,
+      "step": 12150
+    },
+    {
+      "epoch": 0.38853076333929026,
+      "grad_norm": 1.3986321687698364,
+      "learning_rate": 1.5384924203494348e-05,
+      "loss": 0.4882,
+      "step": 12175
+    },
+    {
+      "epoch": 0.3893285677814654,
+      "grad_norm": 1.0822752714157104,
+      "learning_rate": 1.536485097636177e-05,
+      "loss": 0.4517,
+      "step": 12200
+    },
+    {
+      "epoch": 0.39012637222364055,
+      "grad_norm": 1.104766607284546,
+      "learning_rate": 1.534477774922919e-05,
+      "loss": 0.4684,
+      "step": 12225
+    },
+    {
+      "epoch": 0.3909241766658157,
+      "grad_norm": 0.4974013566970825,
+      "learning_rate": 1.5324704522096608e-05,
+      "loss": 0.4644,
+      "step": 12250
+    },
+    {
+      "epoch": 0.39172198110799084,
+      "grad_norm": 1.233022689819336,
+      "learning_rate": 1.5304631294964032e-05,
+      "loss": 0.4296,
+      "step": 12275
+    },
+    {
+      "epoch": 0.3925197855501659,
+      "grad_norm": 1.320533275604248,
+      "learning_rate": 1.528455806783145e-05,
+      "loss": 0.4937,
+      "step": 12300
+    },
+    {
+      "epoch": 0.39331758999234107,
+      "grad_norm": 0.3013911545276642,
+      "learning_rate": 1.526448484069887e-05,
+      "loss": 0.4021,
+      "step": 12325
+    },
+    {
+      "epoch": 0.3941153944345162,
+      "grad_norm": 0.8026270270347595,
+      "learning_rate": 1.5244411613566289e-05,
+      "loss": 0.4949,
+      "step": 12350
+    },
+    {
+      "epoch": 0.39491319887669135,
+      "grad_norm": 0.7568235993385315,
+      "learning_rate": 1.5224338386433712e-05,
+      "loss": 0.4236,
+      "step": 12375
+    },
+    {
+      "epoch": 0.3957110033188665,
+      "grad_norm": 0.8384756445884705,
+      "learning_rate": 1.5204265159301131e-05,
+      "loss": 0.4697,
+      "step": 12400
+    },
+    {
+      "epoch": 0.39650880776104164,
+      "grad_norm": 0.9342674016952515,
+      "learning_rate": 1.518419193216855e-05,
+      "loss": 0.4424,
+      "step": 12425
+    },
+    {
+      "epoch": 0.3973066122032167,
+      "grad_norm": 0.8164283037185669,
+      "learning_rate": 1.516411870503597e-05,
+      "loss": 0.4843,
+      "step": 12450
+    },
+    {
+      "epoch": 0.39810441664539187,
+      "grad_norm": 0.8556693196296692,
+      "learning_rate": 1.5144045477903393e-05,
+      "loss": 0.4874,
+      "step": 12475
+    },
+    {
+      "epoch": 0.398902221087567,
+      "grad_norm": 1.1020561456680298,
+      "learning_rate": 1.5123972250770812e-05,
+      "loss": 0.4821,
+      "step": 12500
+    },
+    {
+      "epoch": 0.39970002552974215,
+      "grad_norm": 0.43005600571632385,
+      "learning_rate": 1.5103899023638232e-05,
+      "loss": 0.4774,
+      "step": 12525
+    },
+    {
+      "epoch": 0.4004978299719173,
+      "grad_norm": 1.0019757747650146,
+      "learning_rate": 1.5083825796505654e-05,
+      "loss": 0.47,
+      "step": 12550
+    },
+    {
+      "epoch": 0.40129563441409244,
+      "grad_norm": 0.9805948734283447,
+      "learning_rate": 1.5063752569373074e-05,
+      "loss": 0.4655,
+      "step": 12575
+    },
+    {
+      "epoch": 0.4020934388562675,
+      "grad_norm": 1.0946297645568848,
+      "learning_rate": 1.5043679342240493e-05,
+      "loss": 0.4656,
+      "step": 12600
+    },
+    {
+      "epoch": 0.40289124329844267,
+      "grad_norm": 0.7530800700187683,
+      "learning_rate": 1.5023606115107913e-05,
+      "loss": 0.4989,
+      "step": 12625
+    },
+    {
+      "epoch": 0.4036890477406178,
+      "grad_norm": 0.7736749053001404,
+      "learning_rate": 1.5003532887975336e-05,
+      "loss": 0.474,
+      "step": 12650
+    },
+    {
+      "epoch": 0.40448685218279296,
+      "grad_norm": 0.6549740433692932,
+      "learning_rate": 1.4983459660842755e-05,
+      "loss": 0.4824,
+      "step": 12675
+    },
+    {
+      "epoch": 0.4052846566249681,
+      "grad_norm": 0.556654691696167,
+      "learning_rate": 1.4963386433710174e-05,
+      "loss": 0.4596,
+      "step": 12700
+    },
+    {
+      "epoch": 0.40608246106714324,
+      "grad_norm": 0.8068644404411316,
+      "learning_rate": 1.4943313206577597e-05,
+      "loss": 0.4255,
+      "step": 12725
+    },
+    {
+      "epoch": 0.40688026550931833,
+      "grad_norm": 0.7896491289138794,
+      "learning_rate": 1.4923239979445017e-05,
+      "loss": 0.4676,
+      "step": 12750
+    },
+    {
+      "epoch": 0.40767806995149347,
+      "grad_norm": 0.8864936232566833,
+      "learning_rate": 1.4903166752312436e-05,
+      "loss": 0.4526,
+      "step": 12775
+    },
+    {
+      "epoch": 0.4084758743936686,
+      "grad_norm": 0.7941861748695374,
+      "learning_rate": 1.4883093525179856e-05,
+      "loss": 0.4269,
+      "step": 12800
+    },
+    {
+      "epoch": 0.40927367883584376,
+      "grad_norm": 0.8171241879463196,
+      "learning_rate": 1.4863020298047278e-05,
+      "loss": 0.4955,
+      "step": 12825
+    },
+    {
+      "epoch": 0.4100714832780189,
+      "grad_norm": 0.7162429094314575,
+      "learning_rate": 1.4842947070914698e-05,
+      "loss": 0.4507,
+      "step": 12850
+    },
+    {
+      "epoch": 0.41086928772019404,
+      "grad_norm": 0.6082909107208252,
+      "learning_rate": 1.4822873843782117e-05,
+      "loss": 0.4514,
+      "step": 12875
+    },
+    {
+      "epoch": 0.4116670921623692,
+      "grad_norm": 0.9657214879989624,
+      "learning_rate": 1.4802800616649537e-05,
+      "loss": 0.4681,
+      "step": 12900
+    },
+    {
+      "epoch": 0.4124648966045443,
+      "grad_norm": 0.9833455681800842,
+      "learning_rate": 1.478272738951696e-05,
+      "loss": 0.5292,
+      "step": 12925
+    },
+    {
+      "epoch": 0.4132627010467194,
+      "grad_norm": 0.7374243140220642,
+      "learning_rate": 1.4762654162384379e-05,
+      "loss": 0.4767,
+      "step": 12950
+    },
+    {
+      "epoch": 0.41406050548889456,
+      "grad_norm": 0.7715759873390198,
+      "learning_rate": 1.4742580935251798e-05,
+      "loss": 0.4711,
+      "step": 12975
+    },
+    {
+      "epoch": 0.4148583099310697,
+      "grad_norm": 0.7317653298377991,
+      "learning_rate": 1.4722507708119221e-05,
+      "loss": 0.4605,
+      "step": 13000
+    },
+    {
+      "epoch": 0.41565611437324484,
+      "grad_norm": 0.5078895092010498,
+      "learning_rate": 1.470243448098664e-05,
+      "loss": 0.4192,
+      "step": 13025
+    },
+    {
+      "epoch": 0.41645391881542,
+      "grad_norm": 0.4797663390636444,
+      "learning_rate": 1.468236125385406e-05,
+      "loss": 0.4826,
+      "step": 13050
+    },
+    {
+      "epoch": 0.4172517232575951,
+      "grad_norm": 0.7433570027351379,
+      "learning_rate": 1.466228802672148e-05,
+      "loss": 0.4962,
+      "step": 13075
+    },
+    {
+      "epoch": 0.4180495276997702,
+      "grad_norm": 0.7199302315711975,
+      "learning_rate": 1.4642214799588902e-05,
+      "loss": 0.4215,
+      "step": 13100
+    },
+    {
+      "epoch": 0.41884733214194536,
+      "grad_norm": 1.2280867099761963,
+      "learning_rate": 1.4622141572456322e-05,
+      "loss": 0.5665,
+      "step": 13125
+    },
+    {
+      "epoch": 0.4196451365841205,
+      "grad_norm": 1.1554296016693115,
+      "learning_rate": 1.4602068345323741e-05,
+      "loss": 0.5012,
+      "step": 13150
+    },
+    {
+      "epoch": 0.42044294102629565,
+      "grad_norm": 0.9082030057907104,
+      "learning_rate": 1.458199511819116e-05,
+      "loss": 0.4835,
+      "step": 13175
+    },
+    {
+      "epoch": 0.4212407454684708,
+      "grad_norm": 1.0009952783584595,
+      "learning_rate": 1.4561921891058584e-05,
+      "loss": 0.5488,
+      "step": 13200
+    },
+    {
+      "epoch": 0.4220385499106459,
+      "grad_norm": 0.5102426409721375,
+      "learning_rate": 1.4541848663926003e-05,
+      "loss": 0.5196,
+      "step": 13225
+    },
+    {
+      "epoch": 0.422836354352821,
+      "grad_norm": 1.0560364723205566,
+      "learning_rate": 1.4521775436793422e-05,
+      "loss": 0.4803,
+      "step": 13250
+    },
+    {
+      "epoch": 0.42363415879499616,
+      "grad_norm": 1.5170994997024536,
+      "learning_rate": 1.4501702209660845e-05,
+      "loss": 0.4307,
+      "step": 13275
+    },
+    {
+      "epoch": 0.4244319632371713,
+      "grad_norm": 0.8432527184486389,
+      "learning_rate": 1.4481628982528265e-05,
+      "loss": 0.5236,
+      "step": 13300
+    },
+    {
+      "epoch": 0.42522976767934645,
+      "grad_norm": 0.5388304591178894,
+      "learning_rate": 1.4461555755395684e-05,
+      "loss": 0.5089,
+      "step": 13325
+    },
+    {
+      "epoch": 0.4260275721215216,
+      "grad_norm": 1.233600378036499,
+      "learning_rate": 1.4441482528263104e-05,
+      "loss": 0.433,
+      "step": 13350
+    },
+    {
+      "epoch": 0.4268253765636967,
+      "grad_norm": 1.7418749332427979,
+      "learning_rate": 1.4421409301130526e-05,
+      "loss": 0.3748,
+      "step": 13375
+    },
+    {
+      "epoch": 0.4276231810058718,
+      "grad_norm": 0.7089687585830688,
+      "learning_rate": 1.4401336073997946e-05,
+      "loss": 0.405,
+      "step": 13400
+    },
+    {
+      "epoch": 0.42842098544804696,
+      "grad_norm": 0.848208487033844,
+      "learning_rate": 1.4381262846865365e-05,
+      "loss": 0.4831,
+      "step": 13425
+    },
+    {
+      "epoch": 0.4292187898902221,
+      "grad_norm": 0.5146478414535522,
+      "learning_rate": 1.4361189619732786e-05,
+      "loss": 0.4587,
+      "step": 13450
+    },
+    {
+      "epoch": 0.43001659433239725,
+      "grad_norm": 0.4021094739437103,
+      "learning_rate": 1.4341116392600208e-05,
+      "loss": 0.4515,
+      "step": 13475
+    },
+    {
+      "epoch": 0.4308143987745724,
+      "grad_norm": 0.797254204750061,
+      "learning_rate": 1.4321043165467627e-05,
+      "loss": 0.4606,
+      "step": 13500
+    },
+    {
+      "epoch": 0.43161220321674754,
+      "grad_norm": 0.7970134019851685,
+      "learning_rate": 1.4300969938335046e-05,
+      "loss": 0.4588,
+      "step": 13525
+    },
+    {
+      "epoch": 0.4324100076589226,
+      "grad_norm": 0.7470701336860657,
+      "learning_rate": 1.4280896711202468e-05,
+      "loss": 0.4378,
+      "step": 13550
+    },
+    {
+      "epoch": 0.43320781210109777,
+      "grad_norm": 0.998225748538971,
+      "learning_rate": 1.4260823484069889e-05,
+      "loss": 0.4534,
+      "step": 13575
+    },
+    {
+      "epoch": 0.4340056165432729,
+      "grad_norm": 1.5322880744934082,
+      "learning_rate": 1.4240750256937308e-05,
+      "loss": 0.4574,
+      "step": 13600
+    },
+    {
+      "epoch": 0.43480342098544805,
+      "grad_norm": 2.1111321449279785,
+      "learning_rate": 1.4220677029804728e-05,
+      "loss": 0.548,
+      "step": 13625
+    },
+    {
+      "epoch": 0.4356012254276232,
+      "grad_norm": 1.3010362386703491,
+      "learning_rate": 1.4200603802672149e-05,
+      "loss": 0.3906,
+      "step": 13650
+    },
+    {
+      "epoch": 0.43639902986979834,
+      "grad_norm": 0.4892048239707947,
+      "learning_rate": 1.418053057553957e-05,
+      "loss": 0.4437,
+      "step": 13675
+    },
+    {
+      "epoch": 0.4371968343119734,
+      "grad_norm": 0.603466808795929,
+      "learning_rate": 1.416045734840699e-05,
+      "loss": 0.4823,
+      "step": 13700
+    },
+    {
+      "epoch": 0.43799463875414857,
+      "grad_norm": 0.5488519072532654,
+      "learning_rate": 1.414038412127441e-05,
+      "loss": 0.4642,
+      "step": 13725
+    },
+    {
+      "epoch": 0.4387924431963237,
+      "grad_norm": 0.9897418022155762,
+      "learning_rate": 1.412031089414183e-05,
+      "loss": 0.5128,
+      "step": 13750
+    },
+    {
+      "epoch": 0.43959024763849885,
+      "grad_norm": 0.32410895824432373,
+      "learning_rate": 1.410023766700925e-05,
+      "loss": 0.3808,
+      "step": 13775
+    },
+    {
+      "epoch": 0.440388052080674,
+      "grad_norm": 1.3885278701782227,
+      "learning_rate": 1.408016443987667e-05,
+      "loss": 0.4951,
+      "step": 13800
+    },
+    {
+      "epoch": 0.44118585652284914,
+      "grad_norm": 0.7485048770904541,
+      "learning_rate": 1.4060091212744092e-05,
+      "loss": 0.5199,
+      "step": 13825
+    },
+    {
+      "epoch": 0.4419836609650242,
+      "grad_norm": 0.6236168146133423,
+      "learning_rate": 1.4040017985611511e-05,
+      "loss": 0.4795,
+      "step": 13850
+    },
+    {
+      "epoch": 0.44278146540719937,
+      "grad_norm": 0.7369155287742615,
+      "learning_rate": 1.401994475847893e-05,
+      "loss": 0.4874,
+      "step": 13875
+    },
+    {
+      "epoch": 0.4435792698493745,
+      "grad_norm": 1.1749411821365356,
+      "learning_rate": 1.3999871531346353e-05,
+      "loss": 0.4792,
+      "step": 13900
+    },
+    {
+      "epoch": 0.44437707429154966,
+      "grad_norm": 0.394364595413208,
+      "learning_rate": 1.3979798304213773e-05,
+      "loss": 0.4384,
+      "step": 13925
+    },
+    {
+      "epoch": 0.4451748787337248,
+      "grad_norm": 0.717343807220459,
+      "learning_rate": 1.3959725077081192e-05,
+      "loss": 0.5035,
+      "step": 13950
+    },
+    {
+      "epoch": 0.44597268317589994,
+      "grad_norm": 0.9770955443382263,
+      "learning_rate": 1.3939651849948612e-05,
+      "loss": 0.4876,
+      "step": 13975
+    },
+    {
+      "epoch": 0.4467704876180751,
+      "grad_norm": 0.1990404576063156,
+      "learning_rate": 1.3919578622816034e-05,
+      "loss": 0.5259,
+      "step": 14000
+    },
+    {
+      "epoch": 0.4475682920602502,
+      "grad_norm": 1.2714977264404297,
+      "learning_rate": 1.3899505395683454e-05,
+      "loss": 0.5204,
+      "step": 14025
+    },
+    {
+      "epoch": 0.4483660965024253,
+      "grad_norm": 0.9181020855903625,
+      "learning_rate": 1.3879432168550873e-05,
+      "loss": 0.5153,
+      "step": 14050
+    },
+    {
+      "epoch": 0.44916390094460046,
+      "grad_norm": 0.7125169038772583,
+      "learning_rate": 1.3859358941418293e-05,
+      "loss": 0.4085,
+      "step": 14075
+    },
+    {
+      "epoch": 0.4499617053867756,
+      "grad_norm": 0.9389515519142151,
+      "learning_rate": 1.3839285714285715e-05,
+      "loss": 0.4255,
+      "step": 14100
+    },
+    {
+      "epoch": 0.45075950982895074,
+      "grad_norm": 0.9015195369720459,
+      "learning_rate": 1.3819212487153135e-05,
+      "loss": 0.4657,
+      "step": 14125
+    },
+    {
+      "epoch": 0.4515573142711259,
+      "grad_norm": 0.5635702610015869,
+      "learning_rate": 1.3799139260020554e-05,
+      "loss": 0.4839,
+      "step": 14150
+    },
+    {
+      "epoch": 0.452355118713301,
+      "grad_norm": 0.6795171499252319,
+      "learning_rate": 1.3779066032887977e-05,
+      "loss": 0.4617,
+      "step": 14175
+    },
+    {
+      "epoch": 0.4531529231554761,
+      "grad_norm": 1.020156741142273,
+      "learning_rate": 1.3758992805755397e-05,
+      "loss": 0.519,
+      "step": 14200
+    },
+    {
+      "epoch": 0.45395072759765126,
+      "grad_norm": 1.1862505674362183,
+      "learning_rate": 1.3738919578622816e-05,
+      "loss": 0.5016,
+      "step": 14225
+    },
+    {
+      "epoch": 0.4547485320398264,
+      "grad_norm": 0.9645688533782959,
+      "learning_rate": 1.3718846351490236e-05,
+      "loss": 0.516,
+      "step": 14250
+    },
+    {
+      "epoch": 0.45554633648200155,
+      "grad_norm": 0.8609448671340942,
+      "learning_rate": 1.3698773124357658e-05,
+      "loss": 0.4662,
+      "step": 14275
+    },
+    {
+      "epoch": 0.4563441409241767,
+      "grad_norm": 0.7543959617614746,
+      "learning_rate": 1.3678699897225078e-05,
+      "loss": 0.4866,
+      "step": 14300
+    },
+    {
+      "epoch": 0.4571419453663518,
+      "grad_norm": 1.1085060834884644,
+      "learning_rate": 1.3658626670092497e-05,
+      "loss": 0.4216,
+      "step": 14325
+    },
+    {
+      "epoch": 0.4579397498085269,
+      "grad_norm": 1.7177371978759766,
+      "learning_rate": 1.3638553442959917e-05,
+      "loss": 0.5042,
+      "step": 14350
+    },
+    {
+      "epoch": 0.45873755425070206,
+      "grad_norm": 0.8194246888160706,
+      "learning_rate": 1.361848021582734e-05,
+      "loss": 0.4852,
+      "step": 14375
+    },
+    {
+      "epoch": 0.4595353586928772,
+      "grad_norm": 0.5089772343635559,
+      "learning_rate": 1.3598406988694759e-05,
+      "loss": 0.5001,
+      "step": 14400
+    },
+    {
+      "epoch": 0.46033316313505235,
+      "grad_norm": 0.7217407822608948,
+      "learning_rate": 1.3578333761562178e-05,
+      "loss": 0.3952,
+      "step": 14425
+    },
+    {
+      "epoch": 0.4611309675772275,
+      "grad_norm": 0.9532568454742432,
+      "learning_rate": 1.3558260534429601e-05,
+      "loss": 0.3844,
+      "step": 14450
+    },
+    {
+      "epoch": 0.4619287720194026,
+      "grad_norm": 0.8932861685752869,
+      "learning_rate": 1.353818730729702e-05,
+      "loss": 0.4268,
+      "step": 14475
+    },
+    {
+      "epoch": 0.4627265764615777,
+      "grad_norm": 0.7293800115585327,
+      "learning_rate": 1.351811408016444e-05,
+      "loss": 0.4509,
+      "step": 14500
+    },
+    {
+      "epoch": 0.46352438090375286,
+      "grad_norm": 0.616687536239624,
+      "learning_rate": 1.349804085303186e-05,
+      "loss": 0.4004,
+      "step": 14525
+    },
+    {
+      "epoch": 0.464322185345928,
+      "grad_norm": 0.6240351796150208,
+      "learning_rate": 1.3477967625899282e-05,
+      "loss": 0.4765,
+      "step": 14550
+    },
+    {
+      "epoch": 0.46511998978810315,
+      "grad_norm": 0.7241207361221313,
+      "learning_rate": 1.3457894398766702e-05,
+      "loss": 0.5139,
+      "step": 14575
+    },
+    {
+      "epoch": 0.4659177942302783,
+      "grad_norm": 0.6566533446311951,
+      "learning_rate": 1.3437821171634121e-05,
+      "loss": 0.4583,
+      "step": 14600
+    },
+    {
+      "epoch": 0.46671559867245344,
+      "grad_norm": 0.5932489633560181,
+      "learning_rate": 1.3417747944501544e-05,
+      "loss": 0.4816,
+      "step": 14625
+    },
+    {
+      "epoch": 0.4675134031146285,
+      "grad_norm": 0.9913176894187927,
+      "learning_rate": 1.3397674717368963e-05,
+      "loss": 0.4806,
+      "step": 14650
+    },
+    {
+      "epoch": 0.46831120755680367,
+      "grad_norm": 0.5182284116744995,
+      "learning_rate": 1.3377601490236383e-05,
+      "loss": 0.4503,
+      "step": 14675
+    },
+    {
+      "epoch": 0.4691090119989788,
+      "grad_norm": 0.3942260444164276,
+      "learning_rate": 1.3357528263103802e-05,
+      "loss": 0.4434,
+      "step": 14700
+    },
+    {
+      "epoch": 0.46990681644115395,
+      "grad_norm": 0.7585189342498779,
+      "learning_rate": 1.3337455035971225e-05,
+      "loss": 0.5389,
+      "step": 14725
+    },
+    {
+      "epoch": 0.4707046208833291,
+      "grad_norm": 0.6170709133148193,
+      "learning_rate": 1.3317381808838645e-05,
+      "loss": 0.4664,
+      "step": 14750
+    },
+    {
+      "epoch": 0.47150242532550424,
+      "grad_norm": 0.7014161348342896,
+      "learning_rate": 1.3297308581706064e-05,
+      "loss": 0.4745,
+      "step": 14775
+    },
+    {
+      "epoch": 0.4723002297676793,
+      "grad_norm": 1.9517532587051392,
+      "learning_rate": 1.3277235354573483e-05,
+      "loss": 0.4908,
+      "step": 14800
+    },
+    {
+      "epoch": 0.47309803420985447,
+      "grad_norm": 0.47267842292785645,
+      "learning_rate": 1.3257162127440906e-05,
+      "loss": 0.3736,
+      "step": 14825
+    },
+    {
+      "epoch": 0.4738958386520296,
+      "grad_norm": 0.5750212073326111,
+      "learning_rate": 1.3237088900308326e-05,
+      "loss": 0.5174,
+      "step": 14850
+    },
+    {
+      "epoch": 0.47469364309420475,
+      "grad_norm": 0.7819316983222961,
+      "learning_rate": 1.3217015673175745e-05,
+      "loss": 0.4964,
+      "step": 14875
+    },
+    {
+      "epoch": 0.4754914475363799,
+      "grad_norm": 1.4765163660049438,
+      "learning_rate": 1.3196942446043168e-05,
+      "loss": 0.4414,
+      "step": 14900
+    },
+    {
+      "epoch": 0.47628925197855504,
+      "grad_norm": 0.7535708546638489,
+      "learning_rate": 1.3176869218910587e-05,
+      "loss": 0.4577,
+      "step": 14925
+    },
+    {
+      "epoch": 0.4770870564207301,
+      "grad_norm": 1.512585997581482,
+      "learning_rate": 1.3156795991778007e-05,
+      "loss": 0.4547,
+      "step": 14950
+    },
+    {
+      "epoch": 0.47788486086290527,
+      "grad_norm": 0.7475982308387756,
+      "learning_rate": 1.3136722764645426e-05,
+      "loss": 0.4786,
+      "step": 14975
+    },
+    {
+      "epoch": 0.4786826653050804,
+      "grad_norm": 0.29426124691963196,
+      "learning_rate": 1.311664953751285e-05,
+      "loss": 0.4754,
+      "step": 15000
+    },
+    {
+      "epoch": 0.47948046974725556,
+      "grad_norm": 0.7986690402030945,
+      "learning_rate": 1.3096576310380269e-05,
+      "loss": 0.4829,
+      "step": 15025
+    },
+    {
+      "epoch": 0.4802782741894307,
+      "grad_norm": 0.7937321662902832,
+      "learning_rate": 1.3076503083247688e-05,
+      "loss": 0.479,
+      "step": 15050
+    },
+    {
+      "epoch": 0.48107607863160584,
+      "grad_norm": 0.8871631026268005,
+      "learning_rate": 1.3056429856115107e-05,
+      "loss": 0.4428,
+      "step": 15075
+    },
+    {
+      "epoch": 0.48187388307378093,
+      "grad_norm": 1.7166088819503784,
+      "learning_rate": 1.303635662898253e-05,
+      "loss": 0.4494,
+      "step": 15100
+    },
+    {
+      "epoch": 0.48267168751595607,
+      "grad_norm": 1.0895076990127563,
+      "learning_rate": 1.301628340184995e-05,
+      "loss": 0.4473,
+      "step": 15125
+    },
+    {
+      "epoch": 0.4834694919581312,
+      "grad_norm": 0.8884141445159912,
+      "learning_rate": 1.299621017471737e-05,
+      "loss": 0.4681,
+      "step": 15150
+    },
+    {
+      "epoch": 0.48426729640030636,
+      "grad_norm": 1.1340069770812988,
+      "learning_rate": 1.297613694758479e-05,
+      "loss": 0.5118,
+      "step": 15175
+    },
+    {
+      "epoch": 0.4850651008424815,
+      "grad_norm": 0.3954940140247345,
+      "learning_rate": 1.2956063720452211e-05,
+      "loss": 0.4181,
+      "step": 15200
+    },
+    {
+      "epoch": 0.48586290528465664,
+      "grad_norm": 0.9146122932434082,
+      "learning_rate": 1.2935990493319631e-05,
+      "loss": 0.4984,
+      "step": 15225
+    },
+    {
+      "epoch": 0.4866607097268318,
+      "grad_norm": 0.6080213189125061,
+      "learning_rate": 1.291591726618705e-05,
+      "loss": 0.4609,
+      "step": 15250
+    },
+    {
+      "epoch": 0.4874585141690069,
+      "grad_norm": 1.3363434076309204,
+      "learning_rate": 1.2895844039054471e-05,
+      "loss": 0.5055,
+      "step": 15275
+    },
+    {
+      "epoch": 0.488256318611182,
+      "grad_norm": 0.7919199466705322,
+      "learning_rate": 1.2875770811921891e-05,
+      "loss": 0.4164,
+      "step": 15300
+    },
+    {
+      "epoch": 0.48905412305335716,
+      "grad_norm": 0.7286228537559509,
+      "learning_rate": 1.2855697584789312e-05,
+      "loss": 0.488,
+      "step": 15325
+    },
+    {
+      "epoch": 0.4898519274955323,
+      "grad_norm": 0.8509761691093445,
+      "learning_rate": 1.2835624357656733e-05,
+      "loss": 0.4565,
+      "step": 15350
+    },
+    {
+      "epoch": 0.49064973193770745,
+      "grad_norm": 0.9142221808433533,
+      "learning_rate": 1.2815551130524153e-05,
+      "loss": 0.5023,
+      "step": 15375
+    },
+    {
+      "epoch": 0.4914475363798826,
+      "grad_norm": 0.9267359972000122,
+      "learning_rate": 1.2795477903391572e-05,
+      "loss": 0.4025,
+      "step": 15400
+    },
+    {
+      "epoch": 0.4922453408220577,
+      "grad_norm": 0.964512288570404,
+      "learning_rate": 1.2775404676258993e-05,
+      "loss": 0.5004,
+      "step": 15425
+    },
+    {
+      "epoch": 0.4930431452642328,
+      "grad_norm": 0.857150137424469,
+      "learning_rate": 1.2755331449126414e-05,
+      "loss": 0.4275,
+      "step": 15450
+    },
+    {
+      "epoch": 0.49384094970640796,
+      "grad_norm": 0.49004054069519043,
+      "learning_rate": 1.2735258221993834e-05,
+      "loss": 0.4849,
+      "step": 15475
+    },
+    {
+      "epoch": 0.4946387541485831,
+      "grad_norm": 0.9604198932647705,
+      "learning_rate": 1.2715184994861253e-05,
+      "loss": 0.4593,
+      "step": 15500
+    },
+    {
+      "epoch": 0.49543655859075825,
+      "grad_norm": 0.5274483561515808,
+      "learning_rate": 1.2695111767728674e-05,
+      "loss": 0.4779,
+      "step": 15525
+    },
+    {
+      "epoch": 0.4962343630329334,
+      "grad_norm": 2.0652174949645996,
+      "learning_rate": 1.2675038540596095e-05,
+      "loss": 0.4492,
+      "step": 15550
+    },
+    {
+      "epoch": 0.4970321674751085,
+      "grad_norm": 0.8072428703308105,
+      "learning_rate": 1.2654965313463515e-05,
+      "loss": 0.492,
+      "step": 15575
+    },
+    {
+      "epoch": 0.4978299719172836,
+      "grad_norm": 0.5541747212409973,
+      "learning_rate": 1.2634892086330934e-05,
+      "loss": 0.449,
+      "step": 15600
+    },
+    {
+      "epoch": 0.49862777635945876,
+      "grad_norm": 0.773730456829071,
+      "learning_rate": 1.2614818859198357e-05,
+      "loss": 0.4562,
+      "step": 15625
+    },
+    {
+      "epoch": 0.4994255808016339,
+      "grad_norm": 0.9741753339767456,
+      "learning_rate": 1.2594745632065777e-05,
+      "loss": 0.4142,
+      "step": 15650
+    },
+    {
+      "epoch": 0.500223385243809,
+      "grad_norm": 1.2298176288604736,
+      "learning_rate": 1.2574672404933196e-05,
+      "loss": 0.5186,
+      "step": 15675
+    },
+    {
+      "epoch": 0.5010211896859842,
+      "grad_norm": 0.8459392189979553,
+      "learning_rate": 1.2554599177800615e-05,
+      "loss": 0.5335,
+      "step": 15700
+    },
+    {
+      "epoch": 0.5018189941281593,
+      "grad_norm": 0.9389364123344421,
+      "learning_rate": 1.2534525950668038e-05,
+      "loss": 0.4762,
+      "step": 15725
+    },
+    {
+      "epoch": 0.5026167985703345,
+      "grad_norm": 1.0771899223327637,
+      "learning_rate": 1.2514452723535458e-05,
+      "loss": 0.5067,
+      "step": 15750
+    },
+    {
+      "epoch": 0.5034146030125096,
+      "grad_norm": 1.1614314317703247,
+      "learning_rate": 1.2494379496402877e-05,
+      "loss": 0.4776,
+      "step": 15775
+    },
+    {
+      "epoch": 0.5042124074546847,
+      "grad_norm": 1.7263872623443604,
+      "learning_rate": 1.2474306269270298e-05,
+      "loss": 0.4122,
+      "step": 15800
+    },
+    {
+      "epoch": 0.5050102118968598,
+      "grad_norm": 1.081225037574768,
+      "learning_rate": 1.245423304213772e-05,
+      "loss": 0.4653,
+      "step": 15825
+    },
+    {
+      "epoch": 0.5058080163390349,
+      "grad_norm": 0.6398253440856934,
+      "learning_rate": 1.2434159815005139e-05,
+      "loss": 0.4563,
+      "step": 15850
+    },
+    {
+      "epoch": 0.5066058207812101,
+      "grad_norm": 1.2320051193237305,
+      "learning_rate": 1.241408658787256e-05,
+      "loss": 0.4612,
+      "step": 15875
+    },
+    {
+      "epoch": 0.5074036252233852,
+      "grad_norm": 0.6390859484672546,
+      "learning_rate": 1.239401336073998e-05,
+      "loss": 0.4836,
+      "step": 15900
+    },
+    {
+      "epoch": 0.5082014296655604,
+      "grad_norm": 0.7116482853889465,
+      "learning_rate": 1.23739401336074e-05,
+      "loss": 0.4157,
+      "step": 15925
+    },
+    {
+      "epoch": 0.5089992341077355,
+      "grad_norm": 0.8386733531951904,
+      "learning_rate": 1.235386690647482e-05,
+      "loss": 0.4989,
+      "step": 15950
+    },
+    {
+      "epoch": 0.5097970385499107,
+      "grad_norm": 0.46999993920326233,
+      "learning_rate": 1.2333793679342241e-05,
+      "loss": 0.5006,
+      "step": 15975
+    },
+    {
+      "epoch": 0.5105948429920858,
+      "grad_norm": 0.6487120985984802,
+      "learning_rate": 1.231372045220966e-05,
+      "loss": 0.4817,
+      "step": 16000
+    },
+    {
+      "epoch": 0.5113926474342609,
+      "grad_norm": 0.7748807668685913,
+      "learning_rate": 1.2293647225077082e-05,
+      "loss": 0.435,
+      "step": 16025
+    },
+    {
+      "epoch": 0.5121904518764361,
+      "grad_norm": 1.0600996017456055,
+      "learning_rate": 1.2273573997944503e-05,
+      "loss": 0.4186,
+      "step": 16050
+    },
+    {
+      "epoch": 0.5129882563186112,
+      "grad_norm": 0.6807442307472229,
+      "learning_rate": 1.2253500770811922e-05,
+      "loss": 0.4519,
+      "step": 16075
+    },
+    {
+      "epoch": 0.5137860607607864,
+      "grad_norm": 0.8556498885154724,
+      "learning_rate": 1.2233427543679343e-05,
+      "loss": 0.4238,
+      "step": 16100
+    },
+    {
+      "epoch": 0.5145838652029614,
+      "grad_norm": 1.2915552854537964,
+      "learning_rate": 1.2213354316546763e-05,
+      "loss": 0.4428,
+      "step": 16125
+    },
+    {
+      "epoch": 0.5153816696451365,
+      "grad_norm": 1.086190104484558,
+      "learning_rate": 1.2193281089414184e-05,
+      "loss": 0.5217,
+      "step": 16150
+    },
+    {
+      "epoch": 0.5161794740873117,
+      "grad_norm": 0.7254406809806824,
+      "learning_rate": 1.2173207862281603e-05,
+      "loss": 0.4927,
+      "step": 16175
+    },
+    {
+      "epoch": 0.5169772785294868,
+      "grad_norm": 1.0535979270935059,
+      "learning_rate": 1.2153134635149025e-05,
+      "loss": 0.5605,
+      "step": 16200
+    },
+    {
+      "epoch": 0.517775082971662,
+      "grad_norm": 0.9690276980400085,
+      "learning_rate": 1.2133061408016444e-05,
+      "loss": 0.4528,
+      "step": 16225
+    },
+    {
+      "epoch": 0.5185728874138371,
+      "grad_norm": 0.6356890201568604,
+      "learning_rate": 1.2112988180883865e-05,
+      "loss": 0.5514,
+      "step": 16250
+    },
+    {
+      "epoch": 0.5193706918560123,
+      "grad_norm": 0.5850731134414673,
+      "learning_rate": 1.2092914953751286e-05,
+      "loss": 0.4528,
+      "step": 16275
+    },
+    {
+      "epoch": 0.5201684962981874,
+      "grad_norm": 1.5722885131835938,
+      "learning_rate": 1.2072841726618706e-05,
+      "loss": 0.419,
+      "step": 16300
+    },
+    {
+      "epoch": 0.5209663007403625,
+      "grad_norm": 0.645149290561676,
+      "learning_rate": 1.2052768499486127e-05,
+      "loss": 0.4778,
+      "step": 16325
+    },
+    {
+      "epoch": 0.5217641051825377,
+      "grad_norm": 0.8022610545158386,
+      "learning_rate": 1.2032695272353546e-05,
+      "loss": 0.4892,
+      "step": 16350
+    },
+    {
+      "epoch": 0.5225619096247128,
+      "grad_norm": 1.2941786050796509,
+      "learning_rate": 1.2012622045220967e-05,
+      "loss": 0.4753,
+      "step": 16375
+    },
+    {
+      "epoch": 0.523359714066888,
+      "grad_norm": 0.8070077896118164,
+      "learning_rate": 1.1992548818088387e-05,
+      "loss": 0.4174,
+      "step": 16400
+    },
+    {
+      "epoch": 0.5241575185090631,
+      "grad_norm": 0.7832496166229248,
+      "learning_rate": 1.1972475590955808e-05,
+      "loss": 0.461,
+      "step": 16425
+    },
+    {
+      "epoch": 0.5249553229512381,
+      "grad_norm": 0.6133424639701843,
+      "learning_rate": 1.1952402363823227e-05,
+      "loss": 0.4594,
+      "step": 16450
+    },
+    {
+      "epoch": 0.5257531273934133,
+      "grad_norm": 0.5747678279876709,
+      "learning_rate": 1.1932329136690649e-05,
+      "loss": 0.4576,
+      "step": 16475
+    },
+    {
+      "epoch": 0.5265509318355884,
+      "grad_norm": 0.6148069500923157,
+      "learning_rate": 1.191225590955807e-05,
+      "loss": 0.4757,
+      "step": 16500
+    },
+    {
+      "epoch": 0.5273487362777636,
+      "grad_norm": 0.8133326172828674,
+      "learning_rate": 1.1892182682425489e-05,
+      "loss": 0.4392,
+      "step": 16525
+    },
+    {
+      "epoch": 0.5281465407199387,
+      "grad_norm": 1.6861692667007446,
+      "learning_rate": 1.187210945529291e-05,
+      "loss": 0.4725,
+      "step": 16550
+    },
+    {
+      "epoch": 0.5289443451621139,
+      "grad_norm": 0.44647324085235596,
+      "learning_rate": 1.185203622816033e-05,
+      "loss": 0.4729,
+      "step": 16575
+    },
+    {
+      "epoch": 0.529742149604289,
+      "grad_norm": 0.7046321034431458,
+      "learning_rate": 1.183196300102775e-05,
+      "loss": 0.4553,
+      "step": 16600
+    },
+    {
+      "epoch": 0.5305399540464641,
+      "grad_norm": 0.5717901587486267,
+      "learning_rate": 1.181188977389517e-05,
+      "loss": 0.4812,
+      "step": 16625
+    },
+    {
+      "epoch": 0.5313377584886393,
+      "grad_norm": 0.5891947150230408,
+      "learning_rate": 1.1791816546762591e-05,
+      "loss": 0.3988,
+      "step": 16650
+    },
+    {
+      "epoch": 0.5321355629308144,
+      "grad_norm": 0.5775773525238037,
+      "learning_rate": 1.177174331963001e-05,
+      "loss": 0.4744,
+      "step": 16675
+    },
+    {
+      "epoch": 0.5329333673729896,
+      "grad_norm": 0.7174286246299744,
+      "learning_rate": 1.1751670092497432e-05,
+      "loss": 0.3991,
+      "step": 16700
+    },
+    {
+      "epoch": 0.5337311718151647,
+      "grad_norm": 0.754780650138855,
+      "learning_rate": 1.1731596865364851e-05,
+      "loss": 0.4513,
+      "step": 16725
+    },
+    {
+      "epoch": 0.5345289762573397,
+      "grad_norm": 0.8494898080825806,
+      "learning_rate": 1.1711523638232273e-05,
+      "loss": 0.4149,
+      "step": 16750
+    },
+    {
+      "epoch": 0.5353267806995149,
+      "grad_norm": 0.32787203788757324,
+      "learning_rate": 1.1691450411099692e-05,
+      "loss": 0.4747,
+      "step": 16775
+    },
+    {
+      "epoch": 0.53612458514169,
+      "grad_norm": 1.0056346654891968,
+      "learning_rate": 1.1671377183967113e-05,
+      "loss": 0.4527,
+      "step": 16800
+    },
+    {
+      "epoch": 0.5369223895838652,
+      "grad_norm": 1.3246283531188965,
+      "learning_rate": 1.1651303956834533e-05,
+      "loss": 0.4984,
+      "step": 16825
+    },
+    {
+      "epoch": 0.5377201940260403,
+      "grad_norm": 1.2716175317764282,
+      "learning_rate": 1.1631230729701954e-05,
+      "loss": 0.3923,
+      "step": 16850
+    },
+    {
+      "epoch": 0.5385179984682155,
+      "grad_norm": 0.9225130677223206,
+      "learning_rate": 1.1611157502569373e-05,
+      "loss": 0.4539,
+      "step": 16875
+    },
+    {
+      "epoch": 0.5393158029103906,
+      "grad_norm": 0.7265071272850037,
+      "learning_rate": 1.1591084275436794e-05,
+      "loss": 0.4778,
+      "step": 16900
+    },
+    {
+      "epoch": 0.5401136073525657,
+      "grad_norm": 0.8638312816619873,
+      "learning_rate": 1.1571011048304214e-05,
+      "loss": 0.4632,
+      "step": 16925
+    },
+    {
+      "epoch": 0.5409114117947409,
+      "grad_norm": 0.6868686079978943,
+      "learning_rate": 1.1550937821171635e-05,
+      "loss": 0.5054,
+      "step": 16950
+    },
+    {
+      "epoch": 0.541709216236916,
+      "grad_norm": 1.0026332139968872,
+      "learning_rate": 1.1530864594039054e-05,
+      "loss": 0.4213,
+      "step": 16975
+    },
+    {
+      "epoch": 0.5425070206790912,
+      "grad_norm": 1.02359938621521,
+      "learning_rate": 1.1510791366906475e-05,
+      "loss": 0.4539,
+      "step": 17000
+    },
+    {
+      "epoch": 0.5433048251212663,
+      "grad_norm": 0.5522026419639587,
+      "learning_rate": 1.1490718139773895e-05,
+      "loss": 0.4533,
+      "step": 17025
+    },
+    {
+      "epoch": 0.5441026295634415,
+      "grad_norm": 0.847406268119812,
+      "learning_rate": 1.1470644912641316e-05,
+      "loss": 0.4679,
+      "step": 17050
+    },
+    {
+      "epoch": 0.5449004340056165,
+      "grad_norm": 0.3250424861907959,
+      "learning_rate": 1.1450571685508735e-05,
+      "loss": 0.4526,
+      "step": 17075
+    },
+    {
+      "epoch": 0.5456982384477916,
+      "grad_norm": 1.1642229557037354,
+      "learning_rate": 1.1430498458376157e-05,
+      "loss": 0.4731,
+      "step": 17100
+    },
+    {
+      "epoch": 0.5464960428899668,
+      "grad_norm": 1.393276572227478,
+      "learning_rate": 1.1410425231243576e-05,
+      "loss": 0.4393,
+      "step": 17125
+    },
+    {
+      "epoch": 0.5472938473321419,
+      "grad_norm": 0.9260122179985046,
+      "learning_rate": 1.1390352004110997e-05,
+      "loss": 0.415,
+      "step": 17150
+    },
+    {
+      "epoch": 0.5480916517743171,
+      "grad_norm": 0.6754615902900696,
+      "learning_rate": 1.1370278776978417e-05,
+      "loss": 0.4119,
+      "step": 17175
+    },
+    {
+      "epoch": 0.5488894562164922,
+      "grad_norm": 0.4611823856830597,
+      "learning_rate": 1.1350205549845838e-05,
+      "loss": 0.4634,
+      "step": 17200
+    },
+    {
+      "epoch": 0.5496872606586674,
+      "grad_norm": 0.439230740070343,
+      "learning_rate": 1.1330132322713259e-05,
+      "loss": 0.3912,
+      "step": 17225
+    },
+    {
+      "epoch": 0.5504850651008425,
+      "grad_norm": 0.7567656636238098,
+      "learning_rate": 1.1310059095580678e-05,
+      "loss": 0.4156,
+      "step": 17250
+    },
+    {
+      "epoch": 0.5512828695430176,
+      "grad_norm": 0.6174827814102173,
+      "learning_rate": 1.12899858684481e-05,
+      "loss": 0.4712,
+      "step": 17275
+    },
+    {
+      "epoch": 0.5520806739851928,
+      "grad_norm": 1.0842621326446533,
+      "learning_rate": 1.1269912641315519e-05,
+      "loss": 0.4809,
+      "step": 17300
+    },
+    {
+      "epoch": 0.5528784784273679,
+      "grad_norm": 1.0989375114440918,
+      "learning_rate": 1.124983941418294e-05,
+      "loss": 0.4579,
+      "step": 17325
+    },
+    {
+      "epoch": 0.5536762828695431,
+      "grad_norm": 1.2708433866500854,
+      "learning_rate": 1.122976618705036e-05,
+      "loss": 0.4065,
+      "step": 17350
+    },
+    {
+      "epoch": 0.5544740873117181,
+      "grad_norm": 0.6591207981109619,
+      "learning_rate": 1.120969295991778e-05,
+      "loss": 0.5051,
+      "step": 17375
+    },
+    {
+      "epoch": 0.5552718917538932,
+      "grad_norm": 0.7222548127174377,
+      "learning_rate": 1.11896197327852e-05,
+      "loss": 0.5097,
+      "step": 17400
+    },
+    {
+      "epoch": 0.5560696961960684,
+      "grad_norm": 1.1408518552780151,
+      "learning_rate": 1.1169546505652621e-05,
+      "loss": 0.4664,
+      "step": 17425
+    },
+    {
+      "epoch": 0.5568675006382435,
+      "grad_norm": 0.2580488622188568,
+      "learning_rate": 1.1149473278520042e-05,
+      "loss": 0.4534,
+      "step": 17450
+    },
+    {
+      "epoch": 0.5576653050804187,
+      "grad_norm": 0.9623468518257141,
+      "learning_rate": 1.1129400051387462e-05,
+      "loss": 0.4357,
+      "step": 17475
+    },
+    {
+      "epoch": 0.5584631095225938,
+      "grad_norm": 1.5668182373046875,
+      "learning_rate": 1.1109326824254883e-05,
+      "loss": 0.5258,
+      "step": 17500
+    },
+    {
+      "epoch": 0.559260913964769,
+      "grad_norm": 0.4907796382904053,
+      "learning_rate": 1.1089253597122302e-05,
+      "loss": 0.4705,
+      "step": 17525
+    },
+    {
+      "epoch": 0.5600587184069441,
+      "grad_norm": 0.6818260550498962,
+      "learning_rate": 1.1069180369989723e-05,
+      "loss": 0.5015,
+      "step": 17550
+    },
+    {
+      "epoch": 0.5608565228491192,
+      "grad_norm": 1.1441423892974854,
+      "learning_rate": 1.1049107142857143e-05,
+      "loss": 0.4585,
+      "step": 17575
+    },
+    {
+      "epoch": 0.5616543272912944,
+      "grad_norm": 0.5130887627601624,
+      "learning_rate": 1.1029033915724564e-05,
+      "loss": 0.4527,
+      "step": 17600
+    },
+    {
+      "epoch": 0.5624521317334695,
+      "grad_norm": 1.3574963808059692,
+      "learning_rate": 1.1008960688591983e-05,
+      "loss": 0.4047,
+      "step": 17625
+    },
+    {
+      "epoch": 0.5632499361756447,
+      "grad_norm": 0.6711642146110535,
+      "learning_rate": 1.0988887461459405e-05,
+      "loss": 0.4798,
+      "step": 17650
+    },
+    {
+      "epoch": 0.5640477406178198,
+      "grad_norm": 0.6693195104598999,
+      "learning_rate": 1.0968814234326824e-05,
+      "loss": 0.4238,
+      "step": 17675
+    },
+    {
+      "epoch": 0.5648455450599948,
+      "grad_norm": 1.0625276565551758,
+      "learning_rate": 1.0948741007194245e-05,
+      "loss": 0.3812,
+      "step": 17700
+    },
+    {
+      "epoch": 0.56564334950217,
+      "grad_norm": 0.9710107445716858,
+      "learning_rate": 1.0928667780061666e-05,
+      "loss": 0.4445,
+      "step": 17725
+    },
+    {
+      "epoch": 0.5664411539443451,
+      "grad_norm": 0.9373927116394043,
+      "learning_rate": 1.0908594552929086e-05,
+      "loss": 0.4968,
+      "step": 17750
+    },
+    {
+      "epoch": 0.5672389583865203,
+      "grad_norm": 0.7126112580299377,
+      "learning_rate": 1.0888521325796507e-05,
+      "loss": 0.4338,
+      "step": 17775
+    },
+    {
+      "epoch": 0.5680367628286954,
+      "grad_norm": 0.8321107029914856,
+      "learning_rate": 1.0868448098663926e-05,
+      "loss": 0.4766,
+      "step": 17800
+    },
+    {
+      "epoch": 0.5688345672708706,
+      "grad_norm": 1.0770740509033203,
+      "learning_rate": 1.0848374871531347e-05,
+      "loss": 0.444,
+      "step": 17825
+    },
+    {
+      "epoch": 0.5696323717130457,
+      "grad_norm": 0.8870105147361755,
+      "learning_rate": 1.0828301644398767e-05,
+      "loss": 0.4883,
+      "step": 17850
+    },
+    {
+      "epoch": 0.5704301761552208,
+      "grad_norm": 1.0617618560791016,
+      "learning_rate": 1.0808228417266188e-05,
+      "loss": 0.4949,
+      "step": 17875
+    },
+    {
+      "epoch": 0.571227980597396,
+      "grad_norm": 0.30951830744743347,
+      "learning_rate": 1.0788155190133607e-05,
+      "loss": 0.5067,
+      "step": 17900
+    },
+    {
+      "epoch": 0.5720257850395711,
+      "grad_norm": 0.9690567255020142,
+      "learning_rate": 1.0768081963001028e-05,
+      "loss": 0.4336,
+      "step": 17925
+    },
+    {
+      "epoch": 0.5728235894817463,
+      "grad_norm": 1.1942867040634155,
+      "learning_rate": 1.074800873586845e-05,
+      "loss": 0.4603,
+      "step": 17950
+    },
+    {
+      "epoch": 0.5736213939239214,
+      "grad_norm": 0.9600849151611328,
+      "learning_rate": 1.0727935508735869e-05,
+      "loss": 0.4351,
+      "step": 17975
+    },
+    {
+      "epoch": 0.5744191983660965,
+      "grad_norm": 1.325777530670166,
+      "learning_rate": 1.070786228160329e-05,
+      "loss": 0.4366,
+      "step": 18000
+    },
+    {
+      "epoch": 0.5752170028082716,
+      "grad_norm": 0.8732290267944336,
+      "learning_rate": 1.068778905447071e-05,
+      "loss": 0.4119,
+      "step": 18025
+    },
+    {
+      "epoch": 0.5760148072504467,
+      "grad_norm": 0.822232186794281,
+      "learning_rate": 1.066771582733813e-05,
+      "loss": 0.453,
+      "step": 18050
+    },
+    {
+      "epoch": 0.5768126116926219,
+      "grad_norm": 0.5754446387290955,
+      "learning_rate": 1.064764260020555e-05,
+      "loss": 0.4386,
+      "step": 18075
+    },
+    {
+      "epoch": 0.577610416134797,
+      "grad_norm": 0.9932488203048706,
+      "learning_rate": 1.0627569373072971e-05,
+      "loss": 0.4224,
+      "step": 18100
+    },
+    {
+      "epoch": 0.5784082205769722,
+      "grad_norm": 1.197025179862976,
+      "learning_rate": 1.060749614594039e-05,
+      "loss": 0.4029,
+      "step": 18125
+    },
+    {
+      "epoch": 0.5792060250191473,
+      "grad_norm": 0.6170285940170288,
+      "learning_rate": 1.0587422918807812e-05,
+      "loss": 0.4688,
+      "step": 18150
+    },
+    {
+      "epoch": 0.5800038294613225,
+      "grad_norm": 0.6788281798362732,
+      "learning_rate": 1.0567349691675233e-05,
+      "loss": 0.4127,
+      "step": 18175
+    },
+    {
+      "epoch": 0.5808016339034976,
+      "grad_norm": 0.9156305193901062,
+      "learning_rate": 1.0547276464542652e-05,
+      "loss": 0.4977,
+      "step": 18200
+    },
+    {
+      "epoch": 0.5815994383456727,
+      "grad_norm": 1.309981107711792,
+      "learning_rate": 1.0527203237410074e-05,
+      "loss": 0.3932,
+      "step": 18225
+    },
+    {
+      "epoch": 0.5823972427878479,
+      "grad_norm": 1.0941970348358154,
+      "learning_rate": 1.0507130010277493e-05,
+      "loss": 0.4313,
+      "step": 18250
+    },
+    {
+      "epoch": 0.583195047230023,
+      "grad_norm": 0.6311997771263123,
+      "learning_rate": 1.0487056783144914e-05,
+      "loss": 0.4303,
+      "step": 18275
+    },
+    {
+      "epoch": 0.5839928516721982,
+      "grad_norm": 0.6561720967292786,
+      "learning_rate": 1.0466983556012334e-05,
+      "loss": 0.5383,
+      "step": 18300
+    },
+    {
+      "epoch": 0.5847906561143732,
+      "grad_norm": 0.528823733329773,
+      "learning_rate": 1.0446910328879755e-05,
+      "loss": 0.467,
+      "step": 18325
+    },
+    {
+      "epoch": 0.5855884605565483,
+      "grad_norm": 0.3194543123245239,
+      "learning_rate": 1.0426837101747174e-05,
+      "loss": 0.484,
+      "step": 18350
+    },
+    {
+      "epoch": 0.5863862649987235,
+      "grad_norm": 1.203895092010498,
+      "learning_rate": 1.0406763874614595e-05,
+      "loss": 0.5813,
+      "step": 18375
+    },
+    {
+      "epoch": 0.5871840694408986,
+      "grad_norm": 0.5720348954200745,
+      "learning_rate": 1.0386690647482015e-05,
+      "loss": 0.4371,
+      "step": 18400
+    },
+    {
+      "epoch": 0.5879818738830738,
+      "grad_norm": 0.49713271856307983,
+      "learning_rate": 1.0366617420349436e-05,
+      "loss": 0.51,
+      "step": 18425
+    },
+    {
+      "epoch": 0.5887796783252489,
+      "grad_norm": 0.9100812673568726,
+      "learning_rate": 1.0346544193216855e-05,
+      "loss": 0.456,
+      "step": 18450
+    },
+    {
+      "epoch": 0.589577482767424,
+      "grad_norm": 0.6779456734657288,
+      "learning_rate": 1.0326470966084276e-05,
+      "loss": 0.4508,
+      "step": 18475
+    },
+    {
+      "epoch": 0.5903752872095992,
+      "grad_norm": 0.5142289996147156,
+      "learning_rate": 1.0306397738951696e-05,
+      "loss": 0.5069,
+      "step": 18500
+    },
+    {
+      "epoch": 0.5911730916517743,
+      "grad_norm": 0.8972587585449219,
+      "learning_rate": 1.0286324511819117e-05,
+      "loss": 0.5365,
+      "step": 18525
+    },
+    {
+      "epoch": 0.5919708960939495,
+      "grad_norm": 0.6554467678070068,
+      "learning_rate": 1.0266251284686536e-05,
+      "loss": 0.4959,
+      "step": 18550
+    },
+    {
+      "epoch": 0.5927687005361246,
+      "grad_norm": 0.8800398111343384,
+      "learning_rate": 1.0246178057553958e-05,
+      "loss": 0.4706,
+      "step": 18575
+    },
+    {
+      "epoch": 0.5935665049782998,
+      "grad_norm": 0.9559227824211121,
+      "learning_rate": 1.0226104830421377e-05,
+      "loss": 0.4496,
+      "step": 18600
+    },
+    {
+      "epoch": 0.5943643094204748,
+      "grad_norm": 1.1470880508422852,
+      "learning_rate": 1.0206031603288798e-05,
+      "loss": 0.4553,
+      "step": 18625
+    },
+    {
+      "epoch": 0.5951621138626499,
+      "grad_norm": 0.7805154919624329,
+      "learning_rate": 1.0185958376156218e-05,
+      "loss": 0.5258,
+      "step": 18650
+    },
+    {
+      "epoch": 0.5959599183048251,
+      "grad_norm": 1.723807454109192,
+      "learning_rate": 1.0165885149023639e-05,
+      "loss": 0.5571,
+      "step": 18675
+    },
+    {
+      "epoch": 0.5967577227470002,
+      "grad_norm": 0.3630259931087494,
+      "learning_rate": 1.0145811921891058e-05,
+      "loss": 0.427,
+      "step": 18700
+    },
+    {
+      "epoch": 0.5975555271891754,
+      "grad_norm": 1.5315433740615845,
+      "learning_rate": 1.012573869475848e-05,
+      "loss": 0.4529,
+      "step": 18725
+    },
+    {
+      "epoch": 0.5983533316313505,
+      "grad_norm": 0.9309017658233643,
+      "learning_rate": 1.0105665467625899e-05,
+      "loss": 0.4576,
+      "step": 18750
+    },
+    {
+      "epoch": 0.5991511360735257,
+      "grad_norm": 1.712268352508545,
+      "learning_rate": 1.008559224049332e-05,
+      "loss": 0.4551,
+      "step": 18775
+    },
+    {
+      "epoch": 0.5999489405157008,
+      "grad_norm": 1.235117793083191,
+      "learning_rate": 1.006551901336074e-05,
+      "loss": 0.4807,
+      "step": 18800
+    },
+    {
+      "epoch": 0.6007467449578759,
+      "grad_norm": 1.1038379669189453,
+      "learning_rate": 1.004544578622816e-05,
+      "loss": 0.4631,
+      "step": 18825
+    },
+    {
+      "epoch": 0.6015445494000511,
+      "grad_norm": 0.6753078699111938,
+      "learning_rate": 1.002537255909558e-05,
+      "loss": 0.5024,
+      "step": 18850
+    },
+    {
+      "epoch": 0.6023423538422262,
+      "grad_norm": 0.7761247754096985,
+      "learning_rate": 1.0005299331963001e-05,
+      "loss": 0.4557,
+      "step": 18875
+    },
+    {
+      "epoch": 0.6031401582844014,
+      "grad_norm": 1.1328458786010742,
+      "learning_rate": 9.985226104830422e-06,
+      "loss": 0.5145,
+      "step": 18900
+    },
+    {
+      "epoch": 0.6039379627265765,
+      "grad_norm": 0.63515704870224,
+      "learning_rate": 9.965152877697842e-06,
+      "loss": 0.4617,
+      "step": 18925
+    },
+    {
+      "epoch": 0.6047357671687515,
+      "grad_norm": 0.4074954688549042,
+      "learning_rate": 9.945079650565263e-06,
+      "loss": 0.4264,
+      "step": 18950
+    },
+    {
+      "epoch": 0.6055335716109267,
+      "grad_norm": 0.9943312406539917,
+      "learning_rate": 9.925006423432682e-06,
+      "loss": 0.5184,
+      "step": 18975
+    },
+    {
+      "epoch": 0.6063313760531018,
+      "grad_norm": 0.9145040512084961,
+      "learning_rate": 9.904933196300103e-06,
+      "loss": 0.3868,
+      "step": 19000
+    },
+    {
+      "epoch": 0.607129180495277,
+      "grad_norm": 1.028710126876831,
+      "learning_rate": 9.884859969167523e-06,
+      "loss": 0.4998,
+      "step": 19025
+    },
+    {
+      "epoch": 0.6079269849374521,
+      "grad_norm": 0.7836230397224426,
+      "learning_rate": 9.864786742034944e-06,
+      "loss": 0.467,
+      "step": 19050
+    },
+    {
+      "epoch": 0.6087247893796273,
+      "grad_norm": 1.2801554203033447,
+      "learning_rate": 9.844713514902363e-06,
+      "loss": 0.4495,
+      "step": 19075
+    },
+    {
+      "epoch": 0.6095225938218024,
+      "grad_norm": 0.748471200466156,
+      "learning_rate": 9.824640287769784e-06,
+      "loss": 0.4979,
+      "step": 19100
+    },
+    {
+      "epoch": 0.6103203982639775,
+      "grad_norm": 0.5333442091941833,
+      "learning_rate": 9.804567060637206e-06,
+      "loss": 0.5054,
+      "step": 19125
+    },
+    {
+      "epoch": 0.6111182027061527,
+      "grad_norm": 0.6367368698120117,
+      "learning_rate": 9.784493833504625e-06,
+      "loss": 0.4753,
+      "step": 19150
+    },
+    {
+      "epoch": 0.6119160071483278,
+      "grad_norm": 0.6109923720359802,
+      "learning_rate": 9.764420606372046e-06,
+      "loss": 0.4178,
+      "step": 19175
+    },
+    {
+      "epoch": 0.612713811590503,
+      "grad_norm": 0.8069015145301819,
+      "learning_rate": 9.744347379239466e-06,
+      "loss": 0.5275,
+      "step": 19200
+    },
+    {
+      "epoch": 0.6135116160326781,
+      "grad_norm": 0.6866354942321777,
+      "learning_rate": 9.724274152106887e-06,
+      "loss": 0.3898,
+      "step": 19225
+    },
+    {
+      "epoch": 0.6143094204748532,
+      "grad_norm": 0.7845311164855957,
+      "learning_rate": 9.704200924974306e-06,
+      "loss": 0.5202,
+      "step": 19250
+    },
+    {
+      "epoch": 0.6151072249170283,
+      "grad_norm": 0.6064120531082153,
+      "learning_rate": 9.684127697841727e-06,
+      "loss": 0.4977,
+      "step": 19275
+    },
+    {
+      "epoch": 0.6159050293592034,
+      "grad_norm": 0.8461109399795532,
+      "learning_rate": 9.664054470709147e-06,
+      "loss": 0.4322,
+      "step": 19300
+    },
+    {
+      "epoch": 0.6167028338013786,
+      "grad_norm": 0.3266942501068115,
+      "learning_rate": 9.643981243576568e-06,
+      "loss": 0.473,
+      "step": 19325
+    },
+    {
+      "epoch": 0.6175006382435537,
+      "grad_norm": 0.7122315764427185,
+      "learning_rate": 9.623908016443989e-06,
+      "loss": 0.518,
+      "step": 19350
+    },
+    {
+      "epoch": 0.6182984426857289,
+      "grad_norm": 0.25452858209609985,
+      "learning_rate": 9.603834789311408e-06,
+      "loss": 0.4029,
+      "step": 19375
+    },
+    {
+      "epoch": 0.619096247127904,
+      "grad_norm": 0.5196482539176941,
+      "learning_rate": 9.58376156217883e-06,
+      "loss": 0.3705,
+      "step": 19400
+    },
+    {
+      "epoch": 0.6198940515700792,
+      "grad_norm": 1.0940996408462524,
+      "learning_rate": 9.563688335046249e-06,
+      "loss": 0.4584,
+      "step": 19425
+    },
+    {
+      "epoch": 0.6206918560122543,
+      "grad_norm": 0.5919874310493469,
+      "learning_rate": 9.54361510791367e-06,
+      "loss": 0.5453,
+      "step": 19450
+    },
+    {
+      "epoch": 0.6214896604544294,
+      "grad_norm": 1.0671135187149048,
+      "learning_rate": 9.52354188078109e-06,
+      "loss": 0.4539,
+      "step": 19475
+    },
+    {
+      "epoch": 0.6222874648966046,
+      "grad_norm": 1.9059410095214844,
+      "learning_rate": 9.50346865364851e-06,
+      "loss": 0.4905,
+      "step": 19500
+    },
+    {
+      "epoch": 0.6230852693387797,
+      "grad_norm": 0.8389720320701599,
+      "learning_rate": 9.48339542651593e-06,
+      "loss": 0.4132,
+      "step": 19525
+    },
+    {
+      "epoch": 0.6238830737809549,
+      "grad_norm": 0.7651321291923523,
+      "learning_rate": 9.463322199383351e-06,
+      "loss": 0.4647,
+      "step": 19550
+    },
+    {
+      "epoch": 0.6246808782231299,
+      "grad_norm": 0.4866482615470886,
+      "learning_rate": 9.443248972250772e-06,
+      "loss": 0.4135,
+      "step": 19575
+    },
+    {
+      "epoch": 0.625478682665305,
+      "grad_norm": 0.42190396785736084,
+      "learning_rate": 9.423175745118192e-06,
+      "loss": 0.4283,
+      "step": 19600
+    },
+    {
+      "epoch": 0.6262764871074802,
+      "grad_norm": 2.0557570457458496,
+      "learning_rate": 9.403102517985613e-06,
+      "loss": 0.4617,
+      "step": 19625
+    },
+    {
+      "epoch": 0.6270742915496553,
+      "grad_norm": 0.7080968022346497,
+      "learning_rate": 9.383029290853032e-06,
+      "loss": 0.4704,
+      "step": 19650
+    },
+    {
+      "epoch": 0.6278720959918305,
+      "grad_norm": 0.35508260130882263,
+      "learning_rate": 9.362956063720454e-06,
+      "loss": 0.514,
+      "step": 19675
+    },
+    {
+      "epoch": 0.6286699004340056,
+      "grad_norm": 0.3968958556652069,
+      "learning_rate": 9.342882836587873e-06,
+      "loss": 0.4835,
+      "step": 19700
+    },
+    {
+      "epoch": 0.6294677048761808,
+      "grad_norm": 0.7077760696411133,
+      "learning_rate": 9.322809609455294e-06,
+      "loss": 0.332,
+      "step": 19725
+    },
+    {
+      "epoch": 0.6302655093183559,
+      "grad_norm": 0.5143846869468689,
+      "learning_rate": 9.302736382322714e-06,
+      "loss": 0.5054,
+      "step": 19750
+    },
+    {
+      "epoch": 0.631063313760531,
+      "grad_norm": 0.4242660105228424,
+      "learning_rate": 9.282663155190135e-06,
+      "loss": 0.4281,
+      "step": 19775
+    },
+    {
+      "epoch": 0.6318611182027062,
+      "grad_norm": 0.679987907409668,
+      "learning_rate": 9.262589928057554e-06,
+      "loss": 0.4453,
+      "step": 19800
+    },
+    {
+      "epoch": 0.6326589226448813,
+      "grad_norm": 1.0942387580871582,
+      "learning_rate": 9.242516700924975e-06,
+      "loss": 0.4916,
+      "step": 19825
+    },
+    {
+      "epoch": 0.6334567270870565,
+      "grad_norm": 0.6843435168266296,
+      "learning_rate": 9.222443473792396e-06,
+      "loss": 0.4365,
+      "step": 19850
+    },
+    {
+      "epoch": 0.6342545315292315,
+      "grad_norm": 0.5189023613929749,
+      "learning_rate": 9.202370246659816e-06,
+      "loss": 0.4858,
+      "step": 19875
+    },
+    {
+      "epoch": 0.6350523359714066,
+      "grad_norm": 0.9441674947738647,
+      "learning_rate": 9.182297019527237e-06,
+      "loss": 0.3874,
+      "step": 19900
+    },
+    {
+      "epoch": 0.6358501404135818,
+      "grad_norm": 0.35448402166366577,
+      "learning_rate": 9.162223792394656e-06,
+      "loss": 0.3836,
+      "step": 19925
+    },
+    {
+      "epoch": 0.6366479448557569,
+      "grad_norm": 1.9092568159103394,
+      "learning_rate": 9.142150565262078e-06,
+      "loss": 0.4742,
+      "step": 19950
+    },
+    {
+      "epoch": 0.6374457492979321,
+      "grad_norm": 0.3969165086746216,
+      "learning_rate": 9.122077338129497e-06,
+      "loss": 0.3687,
+      "step": 19975
+    },
+    {
+      "epoch": 0.6382435537401072,
+      "grad_norm": 1.0867652893066406,
+      "learning_rate": 9.102004110996918e-06,
+      "loss": 0.4653,
+      "step": 20000
+    },
+    {
+      "epoch": 0.6390413581822824,
+      "grad_norm": 1.0081027746200562,
+      "learning_rate": 9.081930883864338e-06,
+      "loss": 0.4174,
+      "step": 20025
+    },
+    {
+      "epoch": 0.6398391626244575,
+      "grad_norm": 0.42013636231422424,
+      "learning_rate": 9.061857656731759e-06,
+      "loss": 0.4373,
+      "step": 20050
+    },
+    {
+      "epoch": 0.6406369670666326,
+      "grad_norm": 0.9158220291137695,
+      "learning_rate": 9.041784429599178e-06,
+      "loss": 0.4604,
+      "step": 20075
+    },
+    {
+      "epoch": 0.6414347715088078,
+      "grad_norm": 1.103516697883606,
+      "learning_rate": 9.021711202466598e-06,
+      "loss": 0.4662,
+      "step": 20100
+    },
+    {
+      "epoch": 0.6422325759509829,
+      "grad_norm": 1.128838300704956,
+      "learning_rate": 9.001637975334019e-06,
+      "loss": 0.4665,
+      "step": 20125
+    },
+    {
+      "epoch": 0.6430303803931581,
+      "grad_norm": 0.66585773229599,
+      "learning_rate": 8.981564748201438e-06,
+      "loss": 0.4627,
+      "step": 20150
+    },
+    {
+      "epoch": 0.6438281848353332,
+      "grad_norm": 1.4553017616271973,
+      "learning_rate": 8.96149152106886e-06,
+      "loss": 0.4319,
+      "step": 20175
+    },
+    {
+      "epoch": 0.6446259892775082,
+      "grad_norm": 0.5869995355606079,
+      "learning_rate": 8.941418293936279e-06,
+      "loss": 0.4512,
+      "step": 20200
+    },
+    {
+      "epoch": 0.6454237937196834,
+      "grad_norm": 0.7448079586029053,
+      "learning_rate": 8.9213450668037e-06,
+      "loss": 0.4283,
+      "step": 20225
+    },
+    {
+      "epoch": 0.6462215981618585,
+      "grad_norm": 0.9250773191452026,
+      "learning_rate": 8.90127183967112e-06,
+      "loss": 0.4558,
+      "step": 20250
+    },
+    {
+      "epoch": 0.6470194026040337,
+      "grad_norm": 1.2568118572235107,
+      "learning_rate": 8.88119861253854e-06,
+      "loss": 0.4673,
+      "step": 20275
+    },
+    {
+      "epoch": 0.6478172070462088,
+      "grad_norm": 0.7910193800926208,
+      "learning_rate": 8.861125385405962e-06,
+      "loss": 0.4134,
+      "step": 20300
+    },
+    {
+      "epoch": 0.648615011488384,
+      "grad_norm": 0.8021867871284485,
+      "learning_rate": 8.841052158273381e-06,
+      "loss": 0.4737,
+      "step": 20325
+    },
+    {
+      "epoch": 0.6494128159305591,
+      "grad_norm": 1.035273790359497,
+      "learning_rate": 8.820978931140802e-06,
+      "loss": 0.451,
+      "step": 20350
+    },
+    {
+      "epoch": 0.6502106203727342,
+      "grad_norm": 0.932902455329895,
+      "learning_rate": 8.800905704008222e-06,
+      "loss": 0.4868,
+      "step": 20375
+    },
+    {
+      "epoch": 0.6510084248149094,
+      "grad_norm": 0.8025002479553223,
+      "learning_rate": 8.780832476875643e-06,
+      "loss": 0.4211,
+      "step": 20400
+    },
+    {
+      "epoch": 0.6518062292570845,
+      "grad_norm": 0.7193844318389893,
+      "learning_rate": 8.760759249743062e-06,
+      "loss": 0.4245,
+      "step": 20425
+    },
+    {
+      "epoch": 0.6526040336992597,
+      "grad_norm": 0.7800945043563843,
+      "learning_rate": 8.740686022610483e-06,
+      "loss": 0.4597,
+      "step": 20450
+    },
+    {
+      "epoch": 0.6534018381414348,
+      "grad_norm": 1.033271074295044,
+      "learning_rate": 8.720612795477903e-06,
+      "loss": 0.4414,
+      "step": 20475
+    },
+    {
+      "epoch": 0.6541996425836099,
+      "grad_norm": 2.8177645206451416,
+      "learning_rate": 8.700539568345324e-06,
+      "loss": 0.4914,
+      "step": 20500
+    },
+    {
+      "epoch": 0.654997447025785,
+      "grad_norm": 0.8891062140464783,
+      "learning_rate": 8.680466341212745e-06,
+      "loss": 0.4424,
+      "step": 20525
+    },
+    {
+      "epoch": 0.6557952514679601,
+      "grad_norm": 1.363903522491455,
+      "learning_rate": 8.660393114080164e-06,
+      "loss": 0.4735,
+      "step": 20550
+    },
+    {
+      "epoch": 0.6565930559101353,
+      "grad_norm": 0.9103833436965942,
+      "learning_rate": 8.640319886947586e-06,
+      "loss": 0.4518,
+      "step": 20575
+    },
+    {
+      "epoch": 0.6573908603523104,
+      "grad_norm": 0.7822914719581604,
+      "learning_rate": 8.620246659815005e-06,
+      "loss": 0.4292,
+      "step": 20600
+    },
+    {
+      "epoch": 0.6581886647944856,
+      "grad_norm": 2.1470134258270264,
+      "learning_rate": 8.600173432682426e-06,
+      "loss": 0.4664,
+      "step": 20625
+    },
+    {
+      "epoch": 0.6589864692366607,
+      "grad_norm": 0.6423518657684326,
+      "learning_rate": 8.580100205549846e-06,
+      "loss": 0.4308,
+      "step": 20650
+    },
+    {
+      "epoch": 0.6597842736788359,
+      "grad_norm": 1.1730583906173706,
+      "learning_rate": 8.560026978417267e-06,
+      "loss": 0.4317,
+      "step": 20675
+    },
+    {
+      "epoch": 0.660582078121011,
+      "grad_norm": 1.6694974899291992,
+      "learning_rate": 8.539953751284686e-06,
+      "loss": 0.4499,
+      "step": 20700
+    },
+    {
+      "epoch": 0.6613798825631861,
+      "grad_norm": 1.1325433254241943,
+      "learning_rate": 8.519880524152107e-06,
+      "loss": 0.505,
+      "step": 20725
+    },
+    {
+      "epoch": 0.6621776870053613,
+      "grad_norm": 1.1263487339019775,
+      "learning_rate": 8.499807297019528e-06,
+      "loss": 0.4222,
+      "step": 20750
+    },
+    {
+      "epoch": 0.6629754914475364,
+      "grad_norm": 0.5488644242286682,
+      "learning_rate": 8.479734069886948e-06,
+      "loss": 0.4639,
+      "step": 20775
+    },
+    {
+      "epoch": 0.6637732958897116,
+      "grad_norm": 0.7831270694732666,
+      "learning_rate": 8.459660842754369e-06,
+      "loss": 0.4212,
+      "step": 20800
+    },
+    {
+      "epoch": 0.6645711003318866,
+      "grad_norm": 0.8199292421340942,
+      "learning_rate": 8.439587615621788e-06,
+      "loss": 0.4419,
+      "step": 20825
+    },
+    {
+      "epoch": 0.6653689047740617,
+      "grad_norm": 0.6708464026451111,
+      "learning_rate": 8.41951438848921e-06,
+      "loss": 0.4738,
+      "step": 20850
+    },
+    {
+      "epoch": 0.6661667092162369,
+      "grad_norm": 0.7316299676895142,
+      "learning_rate": 8.399441161356629e-06,
+      "loss": 0.4844,
+      "step": 20875
+    },
+    {
+      "epoch": 0.666964513658412,
+      "grad_norm": 1.4401187896728516,
+      "learning_rate": 8.37936793422405e-06,
+      "loss": 0.4703,
+      "step": 20900
+    },
+    {
+      "epoch": 0.6677623181005872,
+      "grad_norm": 0.5707149505615234,
+      "learning_rate": 8.35929470709147e-06,
+      "loss": 0.4759,
+      "step": 20925
+    },
+    {
+      "epoch": 0.6685601225427623,
+      "grad_norm": 0.7490055561065674,
+      "learning_rate": 8.33922147995889e-06,
+      "loss": 0.3137,
+      "step": 20950
+    },
+    {
+      "epoch": 0.6693579269849375,
+      "grad_norm": 0.6585310101509094,
+      "learning_rate": 8.31914825282631e-06,
+      "loss": 0.3908,
+      "step": 20975
+    },
+    {
+      "epoch": 0.6701557314271126,
+      "grad_norm": 0.8592582941055298,
+      "learning_rate": 8.299075025693731e-06,
+      "loss": 0.5232,
+      "step": 21000
+    },
+    {
+      "epoch": 0.6709535358692877,
+      "grad_norm": 0.35057422518730164,
+      "learning_rate": 8.279001798561152e-06,
+      "loss": 0.3998,
+      "step": 21025
+    },
+    {
+      "epoch": 0.6717513403114629,
+      "grad_norm": 0.4567711651325226,
+      "learning_rate": 8.258928571428572e-06,
+      "loss": 0.4052,
+      "step": 21050
+    },
+    {
+      "epoch": 0.672549144753638,
+      "grad_norm": 0.9955174922943115,
+      "learning_rate": 8.238855344295993e-06,
+      "loss": 0.4841,
+      "step": 21075
+    },
+    {
+      "epoch": 0.6733469491958132,
+      "grad_norm": 0.7141025066375732,
+      "learning_rate": 8.218782117163412e-06,
+      "loss": 0.3902,
+      "step": 21100
+    },
+    {
+      "epoch": 0.6741447536379882,
+      "grad_norm": 0.9155953526496887,
+      "learning_rate": 8.198708890030833e-06,
+      "loss": 0.4197,
+      "step": 21125
+    },
+    {
+      "epoch": 0.6749425580801633,
+      "grad_norm": 1.427937388420105,
+      "learning_rate": 8.178635662898253e-06,
+      "loss": 0.4639,
+      "step": 21150
+    },
+    {
+      "epoch": 0.6757403625223385,
+      "grad_norm": 0.2833626866340637,
+      "learning_rate": 8.158562435765674e-06,
+      "loss": 0.4684,
+      "step": 21175
+    },
+    {
+      "epoch": 0.6765381669645136,
+      "grad_norm": 0.4364416301250458,
+      "learning_rate": 8.138489208633094e-06,
+      "loss": 0.3881,
+      "step": 21200
+    },
+    {
+      "epoch": 0.6773359714066888,
+      "grad_norm": 0.6703497171401978,
+      "learning_rate": 8.118415981500515e-06,
+      "loss": 0.4907,
+      "step": 21225
+    },
+    {
+      "epoch": 0.6781337758488639,
+      "grad_norm": 1.2733856439590454,
+      "learning_rate": 8.098342754367936e-06,
+      "loss": 0.4545,
+      "step": 21250
+    },
+    {
+      "epoch": 0.6789315802910391,
+      "grad_norm": 1.015809416770935,
+      "learning_rate": 8.078269527235355e-06,
+      "loss": 0.49,
+      "step": 21275
+    },
+    {
+      "epoch": 0.6797293847332142,
+      "grad_norm": 0.4571249783039093,
+      "learning_rate": 8.058196300102776e-06,
+      "loss": 0.5203,
+      "step": 21300
+    },
+    {
+      "epoch": 0.6805271891753893,
+      "grad_norm": 0.6945416927337646,
+      "learning_rate": 8.038123072970196e-06,
+      "loss": 0.45,
+      "step": 21325
+    },
+    {
+      "epoch": 0.6813249936175645,
+      "grad_norm": 0.37875500321388245,
+      "learning_rate": 8.018049845837617e-06,
+      "loss": 0.4111,
+      "step": 21350
+    },
+    {
+      "epoch": 0.6821227980597396,
+      "grad_norm": 0.6942138671875,
+      "learning_rate": 7.997976618705036e-06,
+      "loss": 0.4605,
+      "step": 21375
+    },
+    {
+      "epoch": 0.6829206025019148,
+      "grad_norm": 1.9715650081634521,
+      "learning_rate": 7.977903391572457e-06,
+      "loss": 0.4446,
+      "step": 21400
+    },
+    {
+      "epoch": 0.6837184069440899,
+      "grad_norm": 0.6993141174316406,
+      "learning_rate": 7.957830164439877e-06,
+      "loss": 0.3837,
+      "step": 21425
+    },
+    {
+      "epoch": 0.684516211386265,
+      "grad_norm": 1.2793349027633667,
+      "learning_rate": 7.937756937307298e-06,
+      "loss": 0.4387,
+      "step": 21450
+    },
+    {
+      "epoch": 0.6853140158284401,
+      "grad_norm": 1.1474782228469849,
+      "learning_rate": 7.91768371017472e-06,
+      "loss": 0.4459,
+      "step": 21475
+    },
+    {
+      "epoch": 0.6861118202706152,
+      "grad_norm": 1.2606806755065918,
+      "learning_rate": 7.897610483042139e-06,
+      "loss": 0.4069,
+      "step": 21500
+    },
+    {
+      "epoch": 0.6869096247127904,
+      "grad_norm": 0.9630894064903259,
+      "learning_rate": 7.87753725590956e-06,
+      "loss": 0.4406,
+      "step": 21525
+    },
+    {
+      "epoch": 0.6877074291549655,
+      "grad_norm": 1.2839369773864746,
+      "learning_rate": 7.85746402877698e-06,
+      "loss": 0.3923,
+      "step": 21550
+    },
+    {
+      "epoch": 0.6885052335971407,
+      "grad_norm": 0.9209822416305542,
+      "learning_rate": 7.8373908016444e-06,
+      "loss": 0.498,
+      "step": 21575
+    },
+    {
+      "epoch": 0.6893030380393158,
+      "grad_norm": 1.2149327993392944,
+      "learning_rate": 7.81731757451182e-06,
+      "loss": 0.4475,
+      "step": 21600
+    },
+    {
+      "epoch": 0.690100842481491,
+      "grad_norm": 0.6048392057418823,
+      "learning_rate": 7.79724434737924e-06,
+      "loss": 0.5274,
+      "step": 21625
+    },
+    {
+      "epoch": 0.6908986469236661,
+      "grad_norm": 1.0528596639633179,
+      "learning_rate": 7.77717112024666e-06,
+      "loss": 0.5082,
+      "step": 21650
+    },
+    {
+      "epoch": 0.6916964513658412,
+      "grad_norm": 0.7477854490280151,
+      "learning_rate": 7.75709789311408e-06,
+      "loss": 0.4999,
+      "step": 21675
+    },
+    {
+      "epoch": 0.6924942558080164,
+      "grad_norm": 0.8400072455406189,
+      "learning_rate": 7.737024665981501e-06,
+      "loss": 0.4639,
+      "step": 21700
+    },
+    {
+      "epoch": 0.6932920602501915,
+      "grad_norm": 0.7565206289291382,
+      "learning_rate": 7.71695143884892e-06,
+      "loss": 0.4452,
+      "step": 21725
+    },
+    {
+      "epoch": 0.6940898646923666,
+      "grad_norm": 0.8357092142105103,
+      "learning_rate": 7.696878211716341e-06,
+      "loss": 0.483,
+      "step": 21750
+    },
+    {
+      "epoch": 0.6948876691345417,
+      "grad_norm": 0.7302252650260925,
+      "learning_rate": 7.676804984583761e-06,
+      "loss": 0.5125,
+      "step": 21775
+    },
+    {
+      "epoch": 0.6956854735767168,
+      "grad_norm": 0.8201086521148682,
+      "learning_rate": 7.656731757451182e-06,
+      "loss": 0.415,
+      "step": 21800
+    },
+    {
+      "epoch": 0.696483278018892,
+      "grad_norm": 0.5483965277671814,
+      "learning_rate": 7.636658530318601e-06,
+      "loss": 0.4885,
+      "step": 21825
+    },
+    {
+      "epoch": 0.6972810824610671,
+      "grad_norm": 0.669844925403595,
+      "learning_rate": 7.6165853031860235e-06,
+      "loss": 0.4792,
+      "step": 21850
+    },
+    {
+      "epoch": 0.6980788869032423,
+      "grad_norm": 1.3633980751037598,
+      "learning_rate": 7.596512076053443e-06,
+      "loss": 0.4563,
+      "step": 21875
+    },
+    {
+      "epoch": 0.6988766913454174,
+      "grad_norm": 0.5376991033554077,
+      "learning_rate": 7.576438848920864e-06,
+      "loss": 0.4399,
+      "step": 21900
+    },
+    {
+      "epoch": 0.6996744957875926,
+      "grad_norm": 0.7209357619285583,
+      "learning_rate": 7.5563656217882835e-06,
+      "loss": 0.4284,
+      "step": 21925
+    },
+    {
+      "epoch": 0.7004723002297677,
+      "grad_norm": 1.0455883741378784,
+      "learning_rate": 7.536292394655705e-06,
+      "loss": 0.4254,
+      "step": 21950
+    },
+    {
+      "epoch": 0.7012701046719428,
+      "grad_norm": 1.3988233804702759,
+      "learning_rate": 7.516219167523125e-06,
+      "loss": 0.4694,
+      "step": 21975
+    },
+    {
+      "epoch": 0.702067909114118,
+      "grad_norm": 0.7280498743057251,
+      "learning_rate": 7.496145940390545e-06,
+      "loss": 0.4521,
+      "step": 22000
+    },
+    {
+      "epoch": 0.7028657135562931,
+      "grad_norm": 0.457893431186676,
+      "learning_rate": 7.4760727132579655e-06,
+      "loss": 0.3892,
+      "step": 22025
+    },
+    {
+      "epoch": 0.7036635179984683,
+      "grad_norm": 0.6572154760360718,
+      "learning_rate": 7.455999486125386e-06,
+      "loss": 0.419,
+      "step": 22050
+    },
+    {
+      "epoch": 0.7044613224406433,
+      "grad_norm": 0.5123388171195984,
+      "learning_rate": 7.435926258992806e-06,
+      "loss": 0.4823,
+      "step": 22075
+    },
+    {
+      "epoch": 0.7052591268828184,
+      "grad_norm": 1.2624458074569702,
+      "learning_rate": 7.415853031860226e-06,
+      "loss": 0.4367,
+      "step": 22100
+    },
+    {
+      "epoch": 0.7060569313249936,
+      "grad_norm": 0.9980996251106262,
+      "learning_rate": 7.395779804727647e-06,
+      "loss": 0.4898,
+      "step": 22125
+    },
+    {
+      "epoch": 0.7068547357671687,
+      "grad_norm": 1.3691335916519165,
+      "learning_rate": 7.375706577595067e-06,
+      "loss": 0.4773,
+      "step": 22150
+    },
+    {
+      "epoch": 0.7076525402093439,
+      "grad_norm": 0.6638593673706055,
+      "learning_rate": 7.355633350462487e-06,
+      "loss": 0.4036,
+      "step": 22175
+    },
+    {
+      "epoch": 0.708450344651519,
+      "grad_norm": 0.773307740688324,
+      "learning_rate": 7.335560123329908e-06,
+      "loss": 0.4258,
+      "step": 22200
+    },
+    {
+      "epoch": 0.7092481490936942,
+      "grad_norm": 0.8988430500030518,
+      "learning_rate": 7.315486896197328e-06,
+      "loss": 0.3869,
+      "step": 22225
+    },
+    {
+      "epoch": 0.7100459535358693,
+      "grad_norm": 0.46655961871147156,
+      "learning_rate": 7.295413669064749e-06,
+      "loss": 0.4413,
+      "step": 22250
+    },
+    {
+      "epoch": 0.7108437579780444,
+      "grad_norm": 0.7027944326400757,
+      "learning_rate": 7.275340441932168e-06,
+      "loss": 0.4794,
+      "step": 22275
+    },
+    {
+      "epoch": 0.7116415624202196,
+      "grad_norm": 0.5659714937210083,
+      "learning_rate": 7.2552672147995895e-06,
+      "loss": 0.5137,
+      "step": 22300
+    },
+    {
+      "epoch": 0.7124393668623947,
+      "grad_norm": 0.7782139182090759,
+      "learning_rate": 7.235193987667009e-06,
+      "loss": 0.4064,
+      "step": 22325
+    },
+    {
+      "epoch": 0.7132371713045699,
+      "grad_norm": 0.8942545652389526,
+      "learning_rate": 7.21512076053443e-06,
+      "loss": 0.5079,
+      "step": 22350
+    },
+    {
+      "epoch": 0.7140349757467449,
+      "grad_norm": 0.7615993022918701,
+      "learning_rate": 7.1950475334018495e-06,
+      "loss": 0.4202,
+      "step": 22375
+    },
+    {
+      "epoch": 0.71483278018892,
+      "grad_norm": 0.9601730704307556,
+      "learning_rate": 7.174974306269271e-06,
+      "loss": 0.4549,
+      "step": 22400
+    },
+    {
+      "epoch": 0.7156305846310952,
+      "grad_norm": 1.1959699392318726,
+      "learning_rate": 7.154901079136692e-06,
+      "loss": 0.4714,
+      "step": 22425
+    },
+    {
+      "epoch": 0.7164283890732703,
+      "grad_norm": 0.8447968363761902,
+      "learning_rate": 7.134827852004111e-06,
+      "loss": 0.4699,
+      "step": 22450
+    },
+    {
+      "epoch": 0.7172261935154455,
+      "grad_norm": 1.6587258577346802,
+      "learning_rate": 7.114754624871532e-06,
+      "loss": 0.3632,
+      "step": 22475
+    },
+    {
+      "epoch": 0.7180239979576206,
+      "grad_norm": 0.9805980324745178,
+      "learning_rate": 7.094681397738952e-06,
+      "loss": 0.3702,
+      "step": 22500
+    },
+    {
+      "epoch": 0.7188218023997958,
+      "grad_norm": 0.6011649370193481,
+      "learning_rate": 7.074608170606373e-06,
+      "loss": 0.4965,
+      "step": 22525
+    },
+    {
+      "epoch": 0.7196196068419709,
+      "grad_norm": 0.5311054587364197,
+      "learning_rate": 7.054534943473792e-06,
+      "loss": 0.4595,
+      "step": 22550
+    },
+    {
+      "epoch": 0.720417411284146,
+      "grad_norm": 1.2185113430023193,
+      "learning_rate": 7.0344617163412134e-06,
+      "loss": 0.4478,
+      "step": 22575
+    },
+    {
+      "epoch": 0.7212152157263212,
+      "grad_norm": 0.746266782283783,
+      "learning_rate": 7.014388489208633e-06,
+      "loss": 0.4443,
+      "step": 22600
+    },
+    {
+      "epoch": 0.7220130201684963,
+      "grad_norm": 0.499446302652359,
+      "learning_rate": 6.994315262076054e-06,
+      "loss": 0.401,
+      "step": 22625
+    },
+    {
+      "epoch": 0.7228108246106715,
+      "grad_norm": 0.8455522060394287,
+      "learning_rate": 6.974242034943475e-06,
+      "loss": 0.4382,
+      "step": 22650
+    },
+    {
+      "epoch": 0.7236086290528466,
+      "grad_norm": 1.090509295463562,
+      "learning_rate": 6.954168807810895e-06,
+      "loss": 0.4279,
+      "step": 22675
+    },
+    {
+      "epoch": 0.7244064334950217,
+      "grad_norm": 0.9620064496994019,
+      "learning_rate": 6.934095580678316e-06,
+      "loss": 0.3785,
+      "step": 22700
+    },
+    {
+      "epoch": 0.7252042379371968,
+      "grad_norm": 0.455472856760025,
+      "learning_rate": 6.914022353545735e-06,
+      "loss": 0.4455,
+      "step": 22725
+    },
+    {
+      "epoch": 0.7260020423793719,
+      "grad_norm": 1.5191274881362915,
+      "learning_rate": 6.893949126413156e-06,
+      "loss": 0.4911,
+      "step": 22750
+    },
+    {
+      "epoch": 0.7267998468215471,
+      "grad_norm": 0.9269769191741943,
+      "learning_rate": 6.873875899280576e-06,
+      "loss": 0.4514,
+      "step": 22775
+    },
+    {
+      "epoch": 0.7275976512637222,
+      "grad_norm": 0.774607241153717,
+      "learning_rate": 6.853802672147996e-06,
+      "loss": 0.4568,
+      "step": 22800
+    },
+    {
+      "epoch": 0.7283954557058974,
+      "grad_norm": 1.2156281471252441,
+      "learning_rate": 6.833729445015416e-06,
+      "loss": 0.4349,
+      "step": 22825
+    },
+    {
+      "epoch": 0.7291932601480725,
+      "grad_norm": 0.797739565372467,
+      "learning_rate": 6.813656217882837e-06,
+      "loss": 0.4247,
+      "step": 22850
+    },
+    {
+      "epoch": 0.7299910645902477,
+      "grad_norm": 0.942914605140686,
+      "learning_rate": 6.793582990750258e-06,
+      "loss": 0.4915,
+      "step": 22875
+    },
+    {
+      "epoch": 0.7307888690324228,
+      "grad_norm": 1.1048330068588257,
+      "learning_rate": 6.773509763617677e-06,
+      "loss": 0.49,
+      "step": 22900
+    },
+    {
+      "epoch": 0.7315866734745979,
+      "grad_norm": 0.7522678375244141,
+      "learning_rate": 6.753436536485098e-06,
+      "loss": 0.5611,
+      "step": 22925
+    },
+    {
+      "epoch": 0.7323844779167731,
+      "grad_norm": 0.7625138163566589,
+      "learning_rate": 6.733363309352518e-06,
+      "loss": 0.4254,
+      "step": 22950
+    },
+    {
+      "epoch": 0.7331822823589482,
+      "grad_norm": 0.7532039284706116,
+      "learning_rate": 6.713290082219939e-06,
+      "loss": 0.4184,
+      "step": 22975
+    },
+    {
+      "epoch": 0.7339800868011233,
+      "grad_norm": 1.1807453632354736,
+      "learning_rate": 6.693216855087358e-06,
+      "loss": 0.5173,
+      "step": 23000
+    },
+    {
+      "epoch": 0.7347778912432984,
+      "grad_norm": 0.6093919277191162,
+      "learning_rate": 6.6731436279547794e-06,
+      "loss": 0.415,
+      "step": 23025
+    },
+    {
+      "epoch": 0.7355756956854735,
+      "grad_norm": 1.1248325109481812,
+      "learning_rate": 6.653070400822199e-06,
+      "loss": 0.4539,
+      "step": 23050
+    },
+    {
+      "epoch": 0.7363735001276487,
+      "grad_norm": 0.8926781415939331,
+      "learning_rate": 6.63299717368962e-06,
+      "loss": 0.4423,
+      "step": 23075
+    },
+    {
+      "epoch": 0.7371713045698238,
+      "grad_norm": 0.7217565178871155,
+      "learning_rate": 6.6129239465570394e-06,
+      "loss": 0.4549,
+      "step": 23100
+    },
+    {
+      "epoch": 0.737969109011999,
+      "grad_norm": 1.1932436227798462,
+      "learning_rate": 6.5928507194244606e-06,
+      "loss": 0.4441,
+      "step": 23125
+    },
+    {
+      "epoch": 0.7387669134541741,
+      "grad_norm": 0.7477916479110718,
+      "learning_rate": 6.572777492291882e-06,
+      "loss": 0.4147,
+      "step": 23150
+    },
+    {
+      "epoch": 0.7395647178963493,
+      "grad_norm": 0.721010684967041,
+      "learning_rate": 6.552704265159301e-06,
+      "loss": 0.4351,
+      "step": 23175
+    },
+    {
+      "epoch": 0.7403625223385244,
+      "grad_norm": 0.8074814677238464,
+      "learning_rate": 6.532631038026722e-06,
+      "loss": 0.5054,
+      "step": 23200
+    },
+    {
+      "epoch": 0.7411603267806995,
+      "grad_norm": 1.1390972137451172,
+      "learning_rate": 6.512557810894142e-06,
+      "loss": 0.5007,
+      "step": 23225
+    },
+    {
+      "epoch": 0.7419581312228747,
+      "grad_norm": 0.6260811686515808,
+      "learning_rate": 6.492484583761563e-06,
+      "loss": 0.4868,
+      "step": 23250
+    },
+    {
+      "epoch": 0.7427559356650498,
+      "grad_norm": 0.871371865272522,
+      "learning_rate": 6.472411356628982e-06,
+      "loss": 0.4647,
+      "step": 23275
+    },
+    {
+      "epoch": 0.743553740107225,
+      "grad_norm": 0.4815622866153717,
+      "learning_rate": 6.452338129496403e-06,
+      "loss": 0.4833,
+      "step": 23300
+    },
+    {
+      "epoch": 0.7443515445494,
+      "grad_norm": 0.49014878273010254,
+      "learning_rate": 6.432264902363823e-06,
+      "loss": 0.4405,
+      "step": 23325
+    },
+    {
+      "epoch": 0.7451493489915751,
+      "grad_norm": 0.7321723103523254,
+      "learning_rate": 6.412191675231244e-06,
+      "loss": 0.4431,
+      "step": 23350
+    },
+    {
+      "epoch": 0.7459471534337503,
+      "grad_norm": 0.6357221007347107,
+      "learning_rate": 6.392118448098665e-06,
+      "loss": 0.4405,
+      "step": 23375
+    },
+    {
+      "epoch": 0.7467449578759254,
+      "grad_norm": 0.8464056253433228,
+      "learning_rate": 6.3720452209660846e-06,
+      "loss": 0.3762,
+      "step": 23400
+    },
+    {
+      "epoch": 0.7475427623181006,
+      "grad_norm": 0.36736053228378296,
+      "learning_rate": 6.351971993833506e-06,
+      "loss": 0.5167,
+      "step": 23425
+    },
+    {
+      "epoch": 0.7483405667602757,
+      "grad_norm": 0.4179846942424774,
+      "learning_rate": 6.331898766700925e-06,
+      "loss": 0.4696,
+      "step": 23450
+    },
+    {
+      "epoch": 0.7491383712024509,
+      "grad_norm": 0.5624365210533142,
+      "learning_rate": 6.311825539568346e-06,
+      "loss": 0.4453,
+      "step": 23475
+    },
+    {
+      "epoch": 0.749936175644626,
+      "grad_norm": 0.6948854923248291,
+      "learning_rate": 6.291752312435766e-06,
+      "loss": 0.4108,
+      "step": 23500
+    },
+    {
+      "epoch": 0.7507339800868011,
+      "grad_norm": 0.8782447576522827,
+      "learning_rate": 6.271679085303187e-06,
+      "loss": 0.4025,
+      "step": 23525
+    },
+    {
+      "epoch": 0.7515317845289763,
+      "grad_norm": 0.5253655910491943,
+      "learning_rate": 6.251605858170606e-06,
+      "loss": 0.4388,
+      "step": 23550
+    },
+    {
+      "epoch": 0.7523295889711514,
+      "grad_norm": 0.7023342847824097,
+      "learning_rate": 6.231532631038027e-06,
+      "loss": 0.4273,
+      "step": 23575
+    },
+    {
+      "epoch": 0.7531273934133266,
+      "grad_norm": 0.8842118978500366,
+      "learning_rate": 6.211459403905448e-06,
+      "loss": 0.5036,
+      "step": 23600
+    },
+    {
+      "epoch": 0.7539251978555016,
+      "grad_norm": 1.4032319784164429,
+      "learning_rate": 6.191386176772868e-06,
+      "loss": 0.4888,
+      "step": 23625
+    },
+    {
+      "epoch": 0.7547230022976767,
+      "grad_norm": 0.44206473231315613,
+      "learning_rate": 6.171312949640288e-06,
+      "loss": 0.4463,
+      "step": 23650
+    },
+    {
+      "epoch": 0.7555208067398519,
+      "grad_norm": 1.5239768028259277,
+      "learning_rate": 6.1512397225077085e-06,
+      "loss": 0.4516,
+      "step": 23675
+    },
+    {
+      "epoch": 0.756318611182027,
+      "grad_norm": 0.9154671430587769,
+      "learning_rate": 6.131166495375129e-06,
+      "loss": 0.5288,
+      "step": 23700
+    },
+    {
+      "epoch": 0.7571164156242022,
+      "grad_norm": 0.9400933384895325,
+      "learning_rate": 6.111093268242549e-06,
+      "loss": 0.4362,
+      "step": 23725
+    },
+    {
+      "epoch": 0.7579142200663773,
+      "grad_norm": 1.0026756525039673,
+      "learning_rate": 6.091020041109969e-06,
+      "loss": 0.4442,
+      "step": 23750
+    },
+    {
+      "epoch": 0.7587120245085525,
+      "grad_norm": 0.9670106172561646,
+      "learning_rate": 6.07094681397739e-06,
+      "loss": 0.4778,
+      "step": 23775
+    },
+    {
+      "epoch": 0.7595098289507276,
+      "grad_norm": 0.8731892704963684,
+      "learning_rate": 6.05087358684481e-06,
+      "loss": 0.5162,
+      "step": 23800
+    },
+    {
+      "epoch": 0.7603076333929027,
+      "grad_norm": 0.8117145895957947,
+      "learning_rate": 6.03080035971223e-06,
+      "loss": 0.4648,
+      "step": 23825
+    },
+    {
+      "epoch": 0.7611054378350779,
+      "grad_norm": 0.8562591671943665,
+      "learning_rate": 6.0107271325796505e-06,
+      "loss": 0.4668,
+      "step": 23850
+    },
+    {
+      "epoch": 0.761903242277253,
+      "grad_norm": 1.0122448205947876,
+      "learning_rate": 5.990653905447071e-06,
+      "loss": 0.4951,
+      "step": 23875
+    },
+    {
+      "epoch": 0.7627010467194282,
+      "grad_norm": 0.6894921064376831,
+      "learning_rate": 5.970580678314491e-06,
+      "loss": 0.4328,
+      "step": 23900
+    },
+    {
+      "epoch": 0.7634988511616033,
+      "grad_norm": 1.251848816871643,
+      "learning_rate": 5.950507451181911e-06,
+      "loss": 0.5219,
+      "step": 23925
+    },
+    {
+      "epoch": 0.7642966556037784,
+      "grad_norm": 0.6008142232894897,
+      "learning_rate": 5.9304342240493325e-06,
+      "loss": 0.4343,
+      "step": 23950
+    },
+    {
+      "epoch": 0.7650944600459535,
+      "grad_norm": 1.4150334596633911,
+      "learning_rate": 5.910360996916753e-06,
+      "loss": 0.4935,
+      "step": 23975
+    },
+    {
+      "epoch": 0.7658922644881286,
+      "grad_norm": 1.140459656715393,
+      "learning_rate": 5.890287769784173e-06,
+      "loss": 0.4381,
+      "step": 24000
+    },
+    {
+      "epoch": 0.7666900689303038,
+      "grad_norm": 1.0855172872543335,
+      "learning_rate": 5.870214542651593e-06,
+      "loss": 0.4398,
+      "step": 24025
+    },
+    {
+      "epoch": 0.7674878733724789,
+      "grad_norm": 1.5270390510559082,
+      "learning_rate": 5.850141315519014e-06,
+      "loss": 0.4635,
+      "step": 24050
+    },
+    {
+      "epoch": 0.7682856778146541,
+      "grad_norm": 0.49588742852211,
+      "learning_rate": 5.830068088386434e-06,
+      "loss": 0.4111,
+      "step": 24075
+    },
+    {
+      "epoch": 0.7690834822568292,
+      "grad_norm": 0.7938015460968018,
+      "learning_rate": 5.809994861253854e-06,
+      "loss": 0.4,
+      "step": 24100
+    },
+    {
+      "epoch": 0.7698812866990044,
+      "grad_norm": 1.051199197769165,
+      "learning_rate": 5.7899216341212745e-06,
+      "loss": 0.4471,
+      "step": 24125
+    },
+    {
+      "epoch": 0.7706790911411795,
+      "grad_norm": 0.41370636224746704,
+      "learning_rate": 5.769848406988695e-06,
+      "loss": 0.3582,
+      "step": 24150
+    },
+    {
+      "epoch": 0.7714768955833546,
+      "grad_norm": 0.6135370135307312,
+      "learning_rate": 5.749775179856116e-06,
+      "loss": 0.4763,
+      "step": 24175
+    },
+    {
+      "epoch": 0.7722747000255298,
+      "grad_norm": 0.6476448178291321,
+      "learning_rate": 5.729701952723536e-06,
+      "loss": 0.4751,
+      "step": 24200
+    },
+    {
+      "epoch": 0.7730725044677049,
+      "grad_norm": 0.5266830325126648,
+      "learning_rate": 5.7096287255909565e-06,
+      "loss": 0.4136,
+      "step": 24225
+    },
+    {
+      "epoch": 0.77387030890988,
+      "grad_norm": 0.7233723998069763,
+      "learning_rate": 5.689555498458377e-06,
+      "loss": 0.4664,
+      "step": 24250
+    },
+    {
+      "epoch": 0.7746681133520551,
+      "grad_norm": 1.1774080991744995,
+      "learning_rate": 5.669482271325797e-06,
+      "loss": 0.4115,
+      "step": 24275
+    },
+    {
+      "epoch": 0.7754659177942302,
+      "grad_norm": 0.9674144983291626,
+      "learning_rate": 5.649409044193217e-06,
+      "loss": 0.4505,
+      "step": 24300
+    },
+    {
+      "epoch": 0.7762637222364054,
+      "grad_norm": 0.6825979351997375,
+      "learning_rate": 5.629335817060638e-06,
+      "loss": 0.5304,
+      "step": 24325
+    },
+    {
+      "epoch": 0.7770615266785805,
+      "grad_norm": 0.5845705270767212,
+      "learning_rate": 5.609262589928058e-06,
+      "loss": 0.4467,
+      "step": 24350
+    },
+    {
+      "epoch": 0.7778593311207557,
+      "grad_norm": 1.3499436378479004,
+      "learning_rate": 5.589189362795478e-06,
+      "loss": 0.376,
+      "step": 24375
+    },
+    {
+      "epoch": 0.7786571355629308,
+      "grad_norm": 0.782960832118988,
+      "learning_rate": 5.5691161356628985e-06,
+      "loss": 0.4922,
+      "step": 24400
+    },
+    {
+      "epoch": 0.779454940005106,
+      "grad_norm": 0.5972276926040649,
+      "learning_rate": 5.549042908530319e-06,
+      "loss": 0.3807,
+      "step": 24425
+    },
+    {
+      "epoch": 0.7802527444472811,
+      "grad_norm": 0.9225984811782837,
+      "learning_rate": 5.528969681397739e-06,
+      "loss": 0.3951,
+      "step": 24450
+    },
+    {
+      "epoch": 0.7810505488894562,
+      "grad_norm": 0.6132213473320007,
+      "learning_rate": 5.508896454265159e-06,
+      "loss": 0.4606,
+      "step": 24475
+    },
+    {
+      "epoch": 0.7818483533316314,
+      "grad_norm": 0.8292983174324036,
+      "learning_rate": 5.48882322713258e-06,
+      "loss": 0.4301,
+      "step": 24500
+    },
+    {
+      "epoch": 0.7826461577738065,
+      "grad_norm": 0.8979394435882568,
+      "learning_rate": 5.46875e-06,
+      "loss": 0.4381,
+      "step": 24525
+    },
+    {
+      "epoch": 0.7834439622159817,
+      "grad_norm": 0.7707974910736084,
+      "learning_rate": 5.44867677286742e-06,
+      "loss": 0.4548,
+      "step": 24550
+    },
+    {
+      "epoch": 0.7842417666581567,
+      "grad_norm": 0.7816491723060608,
+      "learning_rate": 5.4286035457348405e-06,
+      "loss": 0.4446,
+      "step": 24575
+    },
+    {
+      "epoch": 0.7850395711003318,
+      "grad_norm": 1.1294130086898804,
+      "learning_rate": 5.408530318602261e-06,
+      "loss": 0.5052,
+      "step": 24600
+    },
+    {
+      "epoch": 0.785837375542507,
+      "grad_norm": 0.662937343120575,
+      "learning_rate": 5.388457091469681e-06,
+      "loss": 0.3954,
+      "step": 24625
+    },
+    {
+      "epoch": 0.7866351799846821,
+      "grad_norm": 0.5076242685317993,
+      "learning_rate": 5.368383864337102e-06,
+      "loss": 0.488,
+      "step": 24650
+    },
+    {
+      "epoch": 0.7874329844268573,
+      "grad_norm": 1.2003448009490967,
+      "learning_rate": 5.3483106372045225e-06,
+      "loss": 0.4162,
+      "step": 24675
+    },
+    {
+      "epoch": 0.7882307888690324,
+      "grad_norm": 0.9463032484054565,
+      "learning_rate": 5.328237410071943e-06,
+      "loss": 0.4827,
+      "step": 24700
+    },
+    {
+      "epoch": 0.7890285933112076,
+      "grad_norm": 0.827299177646637,
+      "learning_rate": 5.308164182939363e-06,
+      "loss": 0.4346,
+      "step": 24725
+    },
+    {
+      "epoch": 0.7898263977533827,
+      "grad_norm": 0.7343999147415161,
+      "learning_rate": 5.288090955806783e-06,
+      "loss": 0.4749,
+      "step": 24750
+    },
+    {
+      "epoch": 0.7906242021955578,
+      "grad_norm": 1.009519100189209,
+      "learning_rate": 5.268017728674204e-06,
+      "loss": 0.4613,
+      "step": 24775
+    },
+    {
+      "epoch": 0.791422006637733,
+      "grad_norm": 0.7076058387756348,
+      "learning_rate": 5.247944501541624e-06,
+      "loss": 0.4468,
+      "step": 24800
+    },
+    {
+      "epoch": 0.7922198110799081,
+      "grad_norm": 0.6096115708351135,
+      "learning_rate": 5.227871274409044e-06,
+      "loss": 0.4518,
+      "step": 24825
+    },
+    {
+      "epoch": 0.7930176155220833,
+      "grad_norm": 0.813082218170166,
+      "learning_rate": 5.2077980472764645e-06,
+      "loss": 0.4212,
+      "step": 24850
+    },
+    {
+      "epoch": 0.7938154199642583,
+      "grad_norm": 0.5449020862579346,
+      "learning_rate": 5.187724820143885e-06,
+      "loss": 0.4565,
+      "step": 24875
+    },
+    {
+      "epoch": 0.7946132244064334,
+      "grad_norm": 1.2484004497528076,
+      "learning_rate": 5.167651593011306e-06,
+      "loss": 0.506,
+      "step": 24900
+    },
+    {
+      "epoch": 0.7954110288486086,
+      "grad_norm": 0.6107988357543945,
+      "learning_rate": 5.147578365878726e-06,
+      "loss": 0.4591,
+      "step": 24925
+    },
+    {
+      "epoch": 0.7962088332907837,
+      "grad_norm": 1.039955735206604,
+      "learning_rate": 5.1275051387461465e-06,
+      "loss": 0.3877,
+      "step": 24950
+    },
+    {
+      "epoch": 0.7970066377329589,
+      "grad_norm": 0.9229916930198669,
+      "learning_rate": 5.107431911613567e-06,
+      "loss": 0.5025,
+      "step": 24975
+    },
+    {
+      "epoch": 0.797804442175134,
+      "grad_norm": 1.3384779691696167,
+      "learning_rate": 5.087358684480987e-06,
+      "loss": 0.4399,
+      "step": 25000
+    },
+    {
+      "epoch": 0.7986022466173092,
+      "grad_norm": 0.9055022597312927,
+      "learning_rate": 5.067285457348407e-06,
+      "loss": 0.3905,
+      "step": 25025
+    },
+    {
+      "epoch": 0.7994000510594843,
+      "grad_norm": 1.076596736907959,
+      "learning_rate": 5.047212230215828e-06,
+      "loss": 0.4646,
+      "step": 25050
+    },
+    {
+      "epoch": 0.8001978555016594,
+      "grad_norm": 0.8877248764038086,
+      "learning_rate": 5.027139003083248e-06,
+      "loss": 0.4294,
+      "step": 25075
+    },
+    {
+      "epoch": 0.8009956599438346,
+      "grad_norm": 0.8110154867172241,
+      "learning_rate": 5.007065775950668e-06,
+      "loss": 0.4358,
+      "step": 25100
+    },
+    {
+      "epoch": 0.8017934643860097,
+      "grad_norm": 0.4719976782798767,
+      "learning_rate": 4.986992548818089e-06,
+      "loss": 0.4037,
+      "step": 25125
+    },
+    {
+      "epoch": 0.8025912688281849,
+      "grad_norm": 0.9919301867485046,
+      "learning_rate": 4.96691932168551e-06,
+      "loss": 0.4317,
+      "step": 25150
+    },
+    {
+      "epoch": 0.80338907327036,
+      "grad_norm": 0.7624319195747375,
+      "learning_rate": 4.94684609455293e-06,
+      "loss": 0.4069,
+      "step": 25175
+    },
+    {
+      "epoch": 0.804186877712535,
+      "grad_norm": 1.2844278812408447,
+      "learning_rate": 4.926772867420349e-06,
+      "loss": 0.4065,
+      "step": 25200
+    },
+    {
+      "epoch": 0.8049846821547102,
+      "grad_norm": 1.9668337106704712,
+      "learning_rate": 4.90669964028777e-06,
+      "loss": 0.5158,
+      "step": 25225
+    },
+    {
+      "epoch": 0.8057824865968853,
+      "grad_norm": 0.7960777282714844,
+      "learning_rate": 4.88662641315519e-06,
+      "loss": 0.4561,
+      "step": 25250
+    },
+    {
+      "epoch": 0.8065802910390605,
+      "grad_norm": 1.1442573070526123,
+      "learning_rate": 4.86655318602261e-06,
+      "loss": 0.4389,
+      "step": 25275
+    },
+    {
+      "epoch": 0.8073780954812356,
+      "grad_norm": 0.6414112448692322,
+      "learning_rate": 4.8464799588900305e-06,
+      "loss": 0.3954,
+      "step": 25300
+    },
+    {
+      "epoch": 0.8081758999234108,
+      "grad_norm": 1.3013173341751099,
+      "learning_rate": 4.826406731757451e-06,
+      "loss": 0.4712,
+      "step": 25325
+    },
+    {
+      "epoch": 0.8089737043655859,
+      "grad_norm": 1.218284249305725,
+      "learning_rate": 4.806333504624871e-06,
+      "loss": 0.4355,
+      "step": 25350
+    },
+    {
+      "epoch": 0.809771508807761,
+      "grad_norm": 0.918598473072052,
+      "learning_rate": 4.786260277492292e-06,
+      "loss": 0.4291,
+      "step": 25375
+    },
+    {
+      "epoch": 0.8105693132499362,
+      "grad_norm": 0.6754209399223328,
+      "learning_rate": 4.7661870503597125e-06,
+      "loss": 0.4094,
+      "step": 25400
+    },
+    {
+      "epoch": 0.8113671176921113,
+      "grad_norm": 0.45365118980407715,
+      "learning_rate": 4.746113823227133e-06,
+      "loss": 0.4213,
+      "step": 25425
+    },
+    {
+      "epoch": 0.8121649221342865,
+      "grad_norm": 0.9628251791000366,
+      "learning_rate": 4.726040596094553e-06,
+      "loss": 0.467,
+      "step": 25450
+    },
+    {
+      "epoch": 0.8129627265764616,
+      "grad_norm": 0.6204245686531067,
+      "learning_rate": 4.705967368961973e-06,
+      "loss": 0.4692,
+      "step": 25475
+    },
+    {
+      "epoch": 0.8137605310186367,
+      "grad_norm": 0.8180040717124939,
+      "learning_rate": 4.685894141829394e-06,
+      "loss": 0.4403,
+      "step": 25500
+    },
+    {
+      "epoch": 0.8145583354608118,
+      "grad_norm": 0.6478278040885925,
+      "learning_rate": 4.665820914696814e-06,
+      "loss": 0.4135,
+      "step": 25525
+    },
+    {
+      "epoch": 0.8153561399029869,
+      "grad_norm": 0.9499540328979492,
+      "learning_rate": 4.645747687564234e-06,
+      "loss": 0.4316,
+      "step": 25550
+    },
+    {
+      "epoch": 0.8161539443451621,
+      "grad_norm": 0.9487488865852356,
+      "learning_rate": 4.6256744604316545e-06,
+      "loss": 0.4531,
+      "step": 25575
+    },
+    {
+      "epoch": 0.8169517487873372,
+      "grad_norm": 1.0303360223770142,
+      "learning_rate": 4.605601233299076e-06,
+      "loss": 0.4601,
+      "step": 25600
+    },
+    {
+      "epoch": 0.8177495532295124,
+      "grad_norm": 0.5638821125030518,
+      "learning_rate": 4.585528006166496e-06,
+      "loss": 0.4367,
+      "step": 25625
+    },
+    {
+      "epoch": 0.8185473576716875,
+      "grad_norm": 0.6594163179397583,
+      "learning_rate": 4.565454779033916e-06,
+      "loss": 0.4313,
+      "step": 25650
+    },
+    {
+      "epoch": 0.8193451621138627,
+      "grad_norm": 0.9298909306526184,
+      "learning_rate": 4.5453815519013365e-06,
+      "loss": 0.553,
+      "step": 25675
+    },
+    {
+      "epoch": 0.8201429665560378,
+      "grad_norm": 0.6085266470909119,
+      "learning_rate": 4.525308324768757e-06,
+      "loss": 0.4837,
+      "step": 25700
+    },
+    {
+      "epoch": 0.8209407709982129,
+      "grad_norm": 0.8190747499465942,
+      "learning_rate": 4.505235097636177e-06,
+      "loss": 0.4788,
+      "step": 25725
+    },
+    {
+      "epoch": 0.8217385754403881,
+      "grad_norm": 1.0059787034988403,
+      "learning_rate": 4.485161870503597e-06,
+      "loss": 0.4349,
+      "step": 25750
+    },
+    {
+      "epoch": 0.8225363798825632,
+      "grad_norm": 0.9931736588478088,
+      "learning_rate": 4.465088643371018e-06,
+      "loss": 0.4796,
+      "step": 25775
+    },
+    {
+      "epoch": 0.8233341843247384,
+      "grad_norm": 1.0960803031921387,
+      "learning_rate": 4.445015416238438e-06,
+      "loss": 0.4288,
+      "step": 25800
+    },
+    {
+      "epoch": 0.8241319887669134,
+      "grad_norm": 0.7804691791534424,
+      "learning_rate": 4.424942189105859e-06,
+      "loss": 0.4688,
+      "step": 25825
+    },
+    {
+      "epoch": 0.8249297932090885,
+      "grad_norm": 0.49692556262016296,
+      "learning_rate": 4.404868961973279e-06,
+      "loss": 0.453,
+      "step": 25850
+    },
+    {
+      "epoch": 0.8257275976512637,
+      "grad_norm": 0.8641229867935181,
+      "learning_rate": 4.3847957348407e-06,
+      "loss": 0.4549,
+      "step": 25875
+    },
+    {
+      "epoch": 0.8265254020934388,
+      "grad_norm": 0.7968979477882385,
+      "learning_rate": 4.36472250770812e-06,
+      "loss": 0.3873,
+      "step": 25900
+    },
+    {
+      "epoch": 0.827323206535614,
+      "grad_norm": 0.6417710781097412,
+      "learning_rate": 4.34464928057554e-06,
+      "loss": 0.3927,
+      "step": 25925
+    },
+    {
+      "epoch": 0.8281210109777891,
+      "grad_norm": 0.7604640126228333,
+      "learning_rate": 4.3245760534429605e-06,
+      "loss": 0.466,
+      "step": 25950
+    },
+    {
+      "epoch": 0.8289188154199643,
+      "grad_norm": 1.0356528759002686,
+      "learning_rate": 4.304502826310381e-06,
+      "loss": 0.4877,
+      "step": 25975
+    },
+    {
+      "epoch": 0.8297166198621394,
+      "grad_norm": 0.8598850965499878,
+      "learning_rate": 4.284429599177801e-06,
+      "loss": 0.3717,
+      "step": 26000
+    },
+    {
+      "epoch": 0.8305144243043145,
+      "grad_norm": 1.647703766822815,
+      "learning_rate": 4.264356372045221e-06,
+      "loss": 0.4876,
+      "step": 26025
+    },
+    {
+      "epoch": 0.8313122287464897,
+      "grad_norm": 0.6927606463432312,
+      "learning_rate": 4.244283144912642e-06,
+      "loss": 0.4737,
+      "step": 26050
+    },
+    {
+      "epoch": 0.8321100331886648,
+      "grad_norm": 1.9317525625228882,
+      "learning_rate": 4.224209917780062e-06,
+      "loss": 0.4072,
+      "step": 26075
+    },
+    {
+      "epoch": 0.83290783763084,
+      "grad_norm": 1.4476104974746704,
+      "learning_rate": 4.204136690647482e-06,
+      "loss": 0.4776,
+      "step": 26100
+    },
+    {
+      "epoch": 0.833705642073015,
+      "grad_norm": 1.3182293176651,
+      "learning_rate": 4.1840634635149024e-06,
+      "loss": 0.4551,
+      "step": 26125
+    },
+    {
+      "epoch": 0.8345034465151902,
+      "grad_norm": 0.4591107666492462,
+      "learning_rate": 4.163990236382323e-06,
+      "loss": 0.5006,
+      "step": 26150
+    },
+    {
+      "epoch": 0.8353012509573653,
+      "grad_norm": 2.0317952632904053,
+      "learning_rate": 4.143917009249743e-06,
+      "loss": 0.3685,
+      "step": 26175
+    },
+    {
+      "epoch": 0.8360990553995404,
+      "grad_norm": 1.3375020027160645,
+      "learning_rate": 4.123843782117163e-06,
+      "loss": 0.391,
+      "step": 26200
+    },
+    {
+      "epoch": 0.8368968598417156,
+      "grad_norm": 0.8244248628616333,
+      "learning_rate": 4.103770554984584e-06,
+      "loss": 0.4395,
+      "step": 26225
+    },
+    {
+      "epoch": 0.8376946642838907,
+      "grad_norm": 0.8836075663566589,
+      "learning_rate": 4.083697327852004e-06,
+      "loss": 0.4241,
+      "step": 26250
+    },
+    {
+      "epoch": 0.8384924687260659,
+      "grad_norm": 0.8446002006530762,
+      "learning_rate": 4.063624100719424e-06,
+      "loss": 0.4999,
+      "step": 26275
+    },
+    {
+      "epoch": 0.839290273168241,
+      "grad_norm": 0.6555794477462769,
+      "learning_rate": 4.043550873586845e-06,
+      "loss": 0.5398,
+      "step": 26300
+    },
+    {
+      "epoch": 0.8400880776104162,
+      "grad_norm": 1.4175652265548706,
+      "learning_rate": 4.023477646454266e-06,
+      "loss": 0.4261,
+      "step": 26325
+    },
+    {
+      "epoch": 0.8408858820525913,
+      "grad_norm": 0.6591663956642151,
+      "learning_rate": 4.003404419321686e-06,
+      "loss": 0.4852,
+      "step": 26350
+    },
+    {
+      "epoch": 0.8416836864947664,
+      "grad_norm": 0.9750247597694397,
+      "learning_rate": 3.983331192189106e-06,
+      "loss": 0.423,
+      "step": 26375
+    },
+    {
+      "epoch": 0.8424814909369416,
+      "grad_norm": 0.7487388253211975,
+      "learning_rate": 3.9632579650565264e-06,
+      "loss": 0.407,
+      "step": 26400
+    },
+    {
+      "epoch": 0.8432792953791167,
+      "grad_norm": 0.8284870982170105,
+      "learning_rate": 3.943184737923947e-06,
+      "loss": 0.4602,
+      "step": 26425
+    },
+    {
+      "epoch": 0.8440770998212918,
+      "grad_norm": 0.8264731168746948,
+      "learning_rate": 3.923111510791367e-06,
+      "loss": 0.4068,
+      "step": 26450
+    },
+    {
+      "epoch": 0.8448749042634669,
+      "grad_norm": 0.648423969745636,
+      "learning_rate": 3.903038283658787e-06,
+      "loss": 0.4144,
+      "step": 26475
+    },
+    {
+      "epoch": 0.845672708705642,
+      "grad_norm": 1.2434682846069336,
+      "learning_rate": 3.882965056526208e-06,
+      "loss": 0.4697,
+      "step": 26500
+    },
+    {
+      "epoch": 0.8464705131478172,
+      "grad_norm": 1.1299934387207031,
+      "learning_rate": 3.862891829393628e-06,
+      "loss": 0.479,
+      "step": 26525
+    },
+    {
+      "epoch": 0.8472683175899923,
+      "grad_norm": 0.8019355535507202,
+      "learning_rate": 3.842818602261049e-06,
+      "loss": 0.4719,
+      "step": 26550
+    },
+    {
+      "epoch": 0.8480661220321675,
+      "grad_norm": 1.5749025344848633,
+      "learning_rate": 3.822745375128469e-06,
+      "loss": 0.4612,
+      "step": 26575
+    },
+    {
+      "epoch": 0.8488639264743426,
+      "grad_norm": 1.2036354541778564,
+      "learning_rate": 3.802672147995889e-06,
+      "loss": 0.4539,
+      "step": 26600
+    },
+    {
+      "epoch": 0.8496617309165178,
+      "grad_norm": 1.055600881576538,
+      "learning_rate": 3.7825989208633094e-06,
+      "loss": 0.4538,
+      "step": 26625
+    },
+    {
+      "epoch": 0.8504595353586929,
+      "grad_norm": 1.1564183235168457,
+      "learning_rate": 3.7625256937307297e-06,
+      "loss": 0.5394,
+      "step": 26650
+    },
+    {
+      "epoch": 0.851257339800868,
+      "grad_norm": 1.1138954162597656,
+      "learning_rate": 3.74245246659815e-06,
+      "loss": 0.4532,
+      "step": 26675
+    },
+    {
+      "epoch": 0.8520551442430432,
+      "grad_norm": 0.5815456509590149,
+      "learning_rate": 3.7223792394655703e-06,
+      "loss": 0.4968,
+      "step": 26700
+    },
+    {
+      "epoch": 0.8528529486852183,
+      "grad_norm": 0.4686177372932434,
+      "learning_rate": 3.7023060123329906e-06,
+      "loss": 0.4501,
+      "step": 26725
+    },
+    {
+      "epoch": 0.8536507531273934,
+      "grad_norm": 0.5037088990211487,
+      "learning_rate": 3.682232785200411e-06,
+      "loss": 0.4265,
+      "step": 26750
+    },
+    {
+      "epoch": 0.8544485575695685,
+      "grad_norm": 0.9735825061798096,
+      "learning_rate": 3.662159558067832e-06,
+      "loss": 0.3994,
+      "step": 26775
+    },
+    {
+      "epoch": 0.8552463620117436,
+      "grad_norm": 1.6671348810195923,
+      "learning_rate": 3.6420863309352523e-06,
+      "loss": 0.4917,
+      "step": 26800
+    },
+    {
+      "epoch": 0.8560441664539188,
+      "grad_norm": 0.9147958159446716,
+      "learning_rate": 3.6220131038026726e-06,
+      "loss": 0.4934,
+      "step": 26825
+    },
+    {
+      "epoch": 0.8568419708960939,
+      "grad_norm": 0.7258955240249634,
+      "learning_rate": 3.601939876670093e-06,
+      "loss": 0.4512,
+      "step": 26850
+    },
+    {
+      "epoch": 0.8576397753382691,
+      "grad_norm": 1.2786023616790771,
+      "learning_rate": 3.581866649537513e-06,
+      "loss": 0.449,
+      "step": 26875
+    },
+    {
+      "epoch": 0.8584375797804442,
+      "grad_norm": 1.0247211456298828,
+      "learning_rate": 3.5617934224049334e-06,
+      "loss": 0.4312,
+      "step": 26900
+    },
+    {
+      "epoch": 0.8592353842226194,
+      "grad_norm": 0.9334677457809448,
+      "learning_rate": 3.5417201952723537e-06,
+      "loss": 0.4484,
+      "step": 26925
+    },
+    {
+      "epoch": 0.8600331886647945,
+      "grad_norm": 1.0831176042556763,
+      "learning_rate": 3.521646968139774e-06,
+      "loss": 0.439,
+      "step": 26950
+    },
+    {
+      "epoch": 0.8608309931069696,
+      "grad_norm": 0.9988906979560852,
+      "learning_rate": 3.5015737410071943e-06,
+      "loss": 0.4584,
+      "step": 26975
+    },
+    {
+      "epoch": 0.8616287975491448,
+      "grad_norm": 0.7575721740722656,
+      "learning_rate": 3.4815005138746146e-06,
+      "loss": 0.4597,
+      "step": 27000
+    },
+    {
+      "epoch": 0.8624266019913199,
+      "grad_norm": 0.5323251485824585,
+      "learning_rate": 3.4614272867420353e-06,
+      "loss": 0.4362,
+      "step": 27025
+    },
+    {
+      "epoch": 0.8632244064334951,
+      "grad_norm": 0.7563241720199585,
+      "learning_rate": 3.4413540596094556e-06,
+      "loss": 0.3917,
+      "step": 27050
+    },
+    {
+      "epoch": 0.8640222108756701,
+      "grad_norm": 0.6328034996986389,
+      "learning_rate": 3.421280832476876e-06,
+      "loss": 0.506,
+      "step": 27075
+    },
+    {
+      "epoch": 0.8648200153178452,
+      "grad_norm": 0.6230515241622925,
+      "learning_rate": 3.401207605344296e-06,
+      "loss": 0.4231,
+      "step": 27100
+    },
+    {
+      "epoch": 0.8656178197600204,
+      "grad_norm": 1.0412312746047974,
+      "learning_rate": 3.3811343782117164e-06,
+      "loss": 0.3983,
+      "step": 27125
+    },
+    {
+      "epoch": 0.8664156242021955,
+      "grad_norm": 1.1394895315170288,
+      "learning_rate": 3.3610611510791367e-06,
+      "loss": 0.447,
+      "step": 27150
+    },
+    {
+      "epoch": 0.8672134286443707,
+      "grad_norm": 0.7843630909919739,
+      "learning_rate": 3.340987923946557e-06,
+      "loss": 0.4271,
+      "step": 27175
+    },
+    {
+      "epoch": 0.8680112330865458,
+      "grad_norm": 0.6892642378807068,
+      "learning_rate": 3.3209146968139773e-06,
+      "loss": 0.4125,
+      "step": 27200
+    },
+    {
+      "epoch": 0.868809037528721,
+      "grad_norm": 1.255333662033081,
+      "learning_rate": 3.3008414696813976e-06,
+      "loss": 0.492,
+      "step": 27225
+    },
+    {
+      "epoch": 0.8696068419708961,
+      "grad_norm": 1.151365041732788,
+      "learning_rate": 3.2807682425488187e-06,
+      "loss": 0.4146,
+      "step": 27250
+    },
+    {
+      "epoch": 0.8704046464130712,
+      "grad_norm": 1.0154856443405151,
+      "learning_rate": 3.260695015416239e-06,
+      "loss": 0.4375,
+      "step": 27275
+    },
+    {
+      "epoch": 0.8712024508552464,
+      "grad_norm": 1.9048150777816772,
+      "learning_rate": 3.2406217882836593e-06,
+      "loss": 0.4004,
+      "step": 27300
+    },
+    {
+      "epoch": 0.8720002552974215,
+      "grad_norm": 1.4371875524520874,
+      "learning_rate": 3.2205485611510795e-06,
+      "loss": 0.5112,
+      "step": 27325
+    },
+    {
+      "epoch": 0.8727980597395967,
+      "grad_norm": 0.8767828345298767,
+      "learning_rate": 3.2004753340185e-06,
+      "loss": 0.4611,
+      "step": 27350
+    },
+    {
+      "epoch": 0.8735958641817718,
+      "grad_norm": 0.9504615664482117,
+      "learning_rate": 3.18040210688592e-06,
+      "loss": 0.4734,
+      "step": 27375
+    },
+    {
+      "epoch": 0.8743936686239469,
+      "grad_norm": 0.7097592353820801,
+      "learning_rate": 3.1603288797533404e-06,
+      "loss": 0.4001,
+      "step": 27400
+    },
+    {
+      "epoch": 0.875191473066122,
+      "grad_norm": 0.9382915496826172,
+      "learning_rate": 3.1402556526207607e-06,
+      "loss": 0.4468,
+      "step": 27425
+    },
+    {
+      "epoch": 0.8759892775082971,
+      "grad_norm": 0.840699315071106,
+      "learning_rate": 3.120182425488181e-06,
+      "loss": 0.4581,
+      "step": 27450
+    },
+    {
+      "epoch": 0.8767870819504723,
+      "grad_norm": 0.5212263464927673,
+      "learning_rate": 3.1001091983556013e-06,
+      "loss": 0.4597,
+      "step": 27475
+    },
+    {
+      "epoch": 0.8775848863926474,
+      "grad_norm": 1.300909161567688,
+      "learning_rate": 3.0800359712230215e-06,
+      "loss": 0.4125,
+      "step": 27500
+    },
+    {
+      "epoch": 0.8783826908348226,
+      "grad_norm": 1.2471950054168701,
+      "learning_rate": 3.059962744090442e-06,
+      "loss": 0.4136,
+      "step": 27525
+    },
+    {
+      "epoch": 0.8791804952769977,
+      "grad_norm": 1.0161088705062866,
+      "learning_rate": 3.0398895169578625e-06,
+      "loss": 0.4226,
+      "step": 27550
+    },
+    {
+      "epoch": 0.8799782997191729,
+      "grad_norm": 0.7740815877914429,
+      "learning_rate": 3.019816289825283e-06,
+      "loss": 0.4074,
+      "step": 27575
+    },
+    {
+      "epoch": 0.880776104161348,
+      "grad_norm": 0.5264682769775391,
+      "learning_rate": 2.999743062692703e-06,
+      "loss": 0.4391,
+      "step": 27600
+    },
+    {
+      "epoch": 0.8815739086035231,
+      "grad_norm": 0.3435772657394409,
+      "learning_rate": 2.9796698355601234e-06,
+      "loss": 0.5446,
+      "step": 27625
+    },
+    {
+      "epoch": 0.8823717130456983,
+      "grad_norm": 2.251746654510498,
+      "learning_rate": 2.9595966084275437e-06,
+      "loss": 0.4649,
+      "step": 27650
+    },
+    {
+      "epoch": 0.8831695174878734,
+      "grad_norm": 0.6416149139404297,
+      "learning_rate": 2.9395233812949644e-06,
+      "loss": 0.4889,
+      "step": 27675
+    },
+    {
+      "epoch": 0.8839673219300485,
+      "grad_norm": 1.3342888355255127,
+      "learning_rate": 2.9194501541623847e-06,
+      "loss": 0.4606,
+      "step": 27700
+    },
+    {
+      "epoch": 0.8847651263722236,
+      "grad_norm": 0.7645418047904968,
+      "learning_rate": 2.899376927029805e-06,
+      "loss": 0.4643,
+      "step": 27725
+    },
+    {
+      "epoch": 0.8855629308143987,
+      "grad_norm": 1.366817593574524,
+      "learning_rate": 2.8793036998972252e-06,
+      "loss": 0.4273,
+      "step": 27750
+    },
+    {
+      "epoch": 0.8863607352565739,
+      "grad_norm": 0.8227196931838989,
+      "learning_rate": 2.8592304727646455e-06,
+      "loss": 0.4334,
+      "step": 27775
+    },
+    {
+      "epoch": 0.887158539698749,
+      "grad_norm": 0.8876248002052307,
+      "learning_rate": 2.839157245632066e-06,
+      "loss": 0.4379,
+      "step": 27800
+    },
+    {
+      "epoch": 0.8879563441409242,
+      "grad_norm": 0.5903315544128418,
+      "learning_rate": 2.819084018499486e-06,
+      "loss": 0.4365,
+      "step": 27825
+    },
+    {
+      "epoch": 0.8887541485830993,
+      "grad_norm": 0.8090491890907288,
+      "learning_rate": 2.7990107913669064e-06,
+      "loss": 0.4082,
+      "step": 27850
+    },
+    {
+      "epoch": 0.8895519530252745,
+      "grad_norm": 1.499947190284729,
+      "learning_rate": 2.7789375642343267e-06,
+      "loss": 0.4417,
+      "step": 27875
+    },
+    {
+      "epoch": 0.8903497574674496,
+      "grad_norm": 1.3385909795761108,
+      "learning_rate": 2.7588643371017474e-06,
+      "loss": 0.5155,
+      "step": 27900
+    },
+    {
+      "epoch": 0.8911475619096247,
+      "grad_norm": 0.8796870112419128,
+      "learning_rate": 2.7387911099691677e-06,
+      "loss": 0.4104,
+      "step": 27925
+    },
+    {
+      "epoch": 0.8919453663517999,
+      "grad_norm": 0.5240228772163391,
+      "learning_rate": 2.718717882836588e-06,
+      "loss": 0.4295,
+      "step": 27950
+    },
+    {
+      "epoch": 0.892743170793975,
+      "grad_norm": 1.00859534740448,
+      "learning_rate": 2.6986446557040082e-06,
+      "loss": 0.4137,
+      "step": 27975
+    },
+    {
+      "epoch": 0.8935409752361502,
+      "grad_norm": 0.6330901980400085,
+      "learning_rate": 2.6785714285714285e-06,
+      "loss": 0.4976,
+      "step": 28000
+    },
+    {
+      "epoch": 0.8943387796783252,
+      "grad_norm": 1.969018578529358,
+      "learning_rate": 2.6584982014388492e-06,
+      "loss": 0.3857,
+      "step": 28025
+    },
+    {
+      "epoch": 0.8951365841205003,
+      "grad_norm": 0.9891221523284912,
+      "learning_rate": 2.6384249743062695e-06,
+      "loss": 0.365,
+      "step": 28050
+    },
+    {
+      "epoch": 0.8959343885626755,
+      "grad_norm": 0.9894170165061951,
+      "learning_rate": 2.61835174717369e-06,
+      "loss": 0.378,
+      "step": 28075
+    },
+    {
+      "epoch": 0.8967321930048506,
+      "grad_norm": 0.8851298093795776,
+      "learning_rate": 2.59827852004111e-06,
+      "loss": 0.4099,
+      "step": 28100
+    },
+    {
+      "epoch": 0.8975299974470258,
+      "grad_norm": 1.1948890686035156,
+      "learning_rate": 2.5782052929085304e-06,
+      "loss": 0.4556,
+      "step": 28125
+    },
+    {
+      "epoch": 0.8983278018892009,
+      "grad_norm": 0.6201117038726807,
+      "learning_rate": 2.558132065775951e-06,
+      "loss": 0.5078,
+      "step": 28150
+    },
+    {
+      "epoch": 0.8991256063313761,
+      "grad_norm": 0.5613217353820801,
+      "learning_rate": 2.5380588386433714e-06,
+      "loss": 0.4271,
+      "step": 28175
+    },
+    {
+      "epoch": 0.8999234107735512,
+      "grad_norm": 0.7274844646453857,
+      "learning_rate": 2.5179856115107916e-06,
+      "loss": 0.4116,
+      "step": 28200
+    },
+    {
+      "epoch": 0.9007212152157263,
+      "grad_norm": 0.6684631705284119,
+      "learning_rate": 2.4979123843782115e-06,
+      "loss": 0.4697,
+      "step": 28225
+    },
+    {
+      "epoch": 0.9015190196579015,
+      "grad_norm": 0.8778204321861267,
+      "learning_rate": 2.4778391572456322e-06,
+      "loss": 0.3628,
+      "step": 28250
+    },
+    {
+      "epoch": 0.9023168241000766,
+      "grad_norm": 1.1124216318130493,
+      "learning_rate": 2.4577659301130525e-06,
+      "loss": 0.4588,
+      "step": 28275
+    },
+    {
+      "epoch": 0.9031146285422518,
+      "grad_norm": 0.9175296425819397,
+      "learning_rate": 2.437692702980473e-06,
+      "loss": 0.4123,
+      "step": 28300
+    },
+    {
+      "epoch": 0.9039124329844268,
+      "grad_norm": 0.8441956639289856,
+      "learning_rate": 2.417619475847893e-06,
+      "loss": 0.3731,
+      "step": 28325
+    },
+    {
+      "epoch": 0.904710237426602,
+      "grad_norm": 0.42998984456062317,
+      "learning_rate": 2.3975462487153134e-06,
+      "loss": 0.464,
+      "step": 28350
+    },
+    {
+      "epoch": 0.9055080418687771,
+      "grad_norm": 1.1030632257461548,
+      "learning_rate": 2.377473021582734e-06,
+      "loss": 0.4391,
+      "step": 28375
+    },
+    {
+      "epoch": 0.9063058463109522,
+      "grad_norm": 1.7289245128631592,
+      "learning_rate": 2.3573997944501544e-06,
+      "loss": 0.4695,
+      "step": 28400
+    },
+    {
+      "epoch": 0.9071036507531274,
+      "grad_norm": 0.6781057715415955,
+      "learning_rate": 2.3373265673175746e-06,
+      "loss": 0.4401,
+      "step": 28425
+    },
+    {
+      "epoch": 0.9079014551953025,
+      "grad_norm": 1.0768778324127197,
+      "learning_rate": 2.317253340184995e-06,
+      "loss": 0.4557,
+      "step": 28450
+    },
+    {
+      "epoch": 0.9086992596374777,
+      "grad_norm": 0.7264516353607178,
+      "learning_rate": 2.2971801130524152e-06,
+      "loss": 0.482,
+      "step": 28475
+    },
+    {
+      "epoch": 0.9094970640796528,
+      "grad_norm": 1.0383821725845337,
+      "learning_rate": 2.277106885919836e-06,
+      "loss": 0.5131,
+      "step": 28500
+    },
+    {
+      "epoch": 0.910294868521828,
+      "grad_norm": 0.6867876052856445,
+      "learning_rate": 2.257033658787256e-06,
+      "loss": 0.4676,
+      "step": 28525
+    },
+    {
+      "epoch": 0.9110926729640031,
+      "grad_norm": 0.840083122253418,
+      "learning_rate": 2.2369604316546765e-06,
+      "loss": 0.4954,
+      "step": 28550
+    },
+    {
+      "epoch": 0.9118904774061782,
+      "grad_norm": 1.2478915452957153,
+      "learning_rate": 2.2168872045220968e-06,
+      "loss": 0.4991,
+      "step": 28575
+    },
+    {
+      "epoch": 0.9126882818483534,
+      "grad_norm": 0.5823152661323547,
+      "learning_rate": 2.196813977389517e-06,
+      "loss": 0.4539,
+      "step": 28600
+    },
+    {
+      "epoch": 0.9134860862905285,
+      "grad_norm": 1.4774327278137207,
+      "learning_rate": 2.1767407502569373e-06,
+      "loss": 0.5059,
+      "step": 28625
+    },
+    {
+      "epoch": 0.9142838907327036,
+      "grad_norm": 0.9385523796081543,
+      "learning_rate": 2.1566675231243576e-06,
+      "loss": 0.4734,
+      "step": 28650
+    },
+    {
+      "epoch": 0.9150816951748787,
+      "grad_norm": 0.7799694538116455,
+      "learning_rate": 2.136594295991778e-06,
+      "loss": 0.459,
+      "step": 28675
+    },
+    {
+      "epoch": 0.9158794996170538,
+      "grad_norm": 0.8785132765769958,
+      "learning_rate": 2.116521068859198e-06,
+      "loss": 0.4647,
+      "step": 28700
+    },
+    {
+      "epoch": 0.916677304059229,
+      "grad_norm": 0.6985654830932617,
+      "learning_rate": 2.096447841726619e-06,
+      "loss": 0.395,
+      "step": 28725
+    },
+    {
+      "epoch": 0.9174751085014041,
+      "grad_norm": 1.901870608329773,
+      "learning_rate": 2.076374614594039e-06,
+      "loss": 0.4138,
+      "step": 28750
+    },
+    {
+      "epoch": 0.9182729129435793,
+      "grad_norm": 0.8224325776100159,
+      "learning_rate": 2.0563013874614595e-06,
+      "loss": 0.4232,
+      "step": 28775
+    },
+    {
+      "epoch": 0.9190707173857544,
+      "grad_norm": 1.8118170499801636,
+      "learning_rate": 2.0362281603288798e-06,
+      "loss": 0.4041,
+      "step": 28800
+    },
+    {
+      "epoch": 0.9198685218279296,
+      "grad_norm": 1.5147721767425537,
+      "learning_rate": 2.0161549331963e-06,
+      "loss": 0.4933,
+      "step": 28825
+    },
+    {
+      "epoch": 0.9206663262701047,
+      "grad_norm": 0.9193384051322937,
+      "learning_rate": 1.9960817060637208e-06,
+      "loss": 0.4328,
+      "step": 28850
+    },
+    {
+      "epoch": 0.9214641307122798,
+      "grad_norm": 0.6495897769927979,
+      "learning_rate": 1.976008478931141e-06,
+      "loss": 0.4864,
+      "step": 28875
+    },
+    {
+      "epoch": 0.922261935154455,
+      "grad_norm": 1.427866816520691,
+      "learning_rate": 1.9559352517985613e-06,
+      "loss": 0.5036,
+      "step": 28900
+    },
+    {
+      "epoch": 0.9230597395966301,
+      "grad_norm": 0.9476510286331177,
+      "learning_rate": 1.9358620246659816e-06,
+      "loss": 0.4221,
+      "step": 28925
+    },
+    {
+      "epoch": 0.9238575440388052,
+      "grad_norm": 1.2440739870071411,
+      "learning_rate": 1.915788797533402e-06,
+      "loss": 0.3542,
+      "step": 28950
+    },
+    {
+      "epoch": 0.9246553484809803,
+      "grad_norm": 1.3304252624511719,
+      "learning_rate": 1.8957155704008224e-06,
+      "loss": 0.4482,
+      "step": 28975
+    },
+    {
+      "epoch": 0.9254531529231554,
+      "grad_norm": 0.4619370400905609,
+      "learning_rate": 1.8756423432682427e-06,
+      "loss": 0.4353,
+      "step": 29000
+    },
+    {
+      "epoch": 0.9262509573653306,
+      "grad_norm": 0.46142280101776123,
+      "learning_rate": 1.855569116135663e-06,
+      "loss": 0.4172,
+      "step": 29025
+    },
+    {
+      "epoch": 0.9270487618075057,
+      "grad_norm": 0.9833303093910217,
+      "learning_rate": 1.8354958890030833e-06,
+      "loss": 0.5051,
+      "step": 29050
+    },
+    {
+      "epoch": 0.9278465662496809,
+      "grad_norm": 0.7043890357017517,
+      "learning_rate": 1.8154226618705038e-06,
+      "loss": 0.5368,
+      "step": 29075
+    },
+    {
+      "epoch": 0.928644370691856,
+      "grad_norm": 0.7420092821121216,
+      "learning_rate": 1.795349434737924e-06,
+      "loss": 0.4528,
+      "step": 29100
+    },
+    {
+      "epoch": 0.9294421751340312,
+      "grad_norm": 1.0412980318069458,
+      "learning_rate": 1.7752762076053443e-06,
+      "loss": 0.6043,
+      "step": 29125
+    },
+    {
+      "epoch": 0.9302399795762063,
+      "grad_norm": 1.3793296813964844,
+      "learning_rate": 1.7552029804727646e-06,
+      "loss": 0.3907,
+      "step": 29150
+    },
+    {
+      "epoch": 0.9310377840183814,
+      "grad_norm": 1.4980815649032593,
+      "learning_rate": 1.735129753340185e-06,
+      "loss": 0.4657,
+      "step": 29175
+    },
+    {
+      "epoch": 0.9318355884605566,
+      "grad_norm": 0.6619548797607422,
+      "learning_rate": 1.7150565262076056e-06,
+      "loss": 0.4292,
+      "step": 29200
+    },
+    {
+      "epoch": 0.9326333929027317,
+      "grad_norm": 0.5773054361343384,
+      "learning_rate": 1.694983299075026e-06,
+      "loss": 0.3851,
+      "step": 29225
+    },
+    {
+      "epoch": 0.9334311973449069,
+      "grad_norm": 0.8674346804618835,
+      "learning_rate": 1.6749100719424462e-06,
+      "loss": 0.4751,
+      "step": 29250
+    },
+    {
+      "epoch": 0.9342290017870819,
+      "grad_norm": 0.7046436667442322,
+      "learning_rate": 1.6548368448098665e-06,
+      "loss": 0.4367,
+      "step": 29275
+    },
+    {
+      "epoch": 0.935026806229257,
+      "grad_norm": 0.8323100209236145,
+      "learning_rate": 1.6347636176772868e-06,
+      "loss": 0.5169,
+      "step": 29300
+    },
+    {
+      "epoch": 0.9358246106714322,
+      "grad_norm": 0.6985541582107544,
+      "learning_rate": 1.6146903905447072e-06,
+      "loss": 0.4482,
+      "step": 29325
+    },
+    {
+      "epoch": 0.9366224151136073,
+      "grad_norm": 1.2575738430023193,
+      "learning_rate": 1.5946171634121275e-06,
+      "loss": 0.4144,
+      "step": 29350
+    },
+    {
+      "epoch": 0.9374202195557825,
+      "grad_norm": 0.5694016814231873,
+      "learning_rate": 1.5745439362795478e-06,
+      "loss": 0.4731,
+      "step": 29375
+    },
+    {
+      "epoch": 0.9382180239979576,
+      "grad_norm": 0.4448241889476776,
+      "learning_rate": 1.5544707091469683e-06,
+      "loss": 0.4668,
+      "step": 29400
+    },
+    {
+      "epoch": 0.9390158284401328,
+      "grad_norm": 0.6371973752975464,
+      "learning_rate": 1.5343974820143886e-06,
+      "loss": 0.4542,
+      "step": 29425
+    },
+    {
+      "epoch": 0.9398136328823079,
+      "grad_norm": 0.6415957808494568,
+      "learning_rate": 1.5143242548818089e-06,
+      "loss": 0.403,
+      "step": 29450
+    },
+    {
+      "epoch": 0.940611437324483,
+      "grad_norm": 0.5851134657859802,
+      "learning_rate": 1.4942510277492294e-06,
+      "loss": 0.4501,
+      "step": 29475
+    },
+    {
+      "epoch": 0.9414092417666582,
+      "grad_norm": 1.1694976091384888,
+      "learning_rate": 1.4741778006166495e-06,
+      "loss": 0.474,
+      "step": 29500
+    },
+    {
+      "epoch": 0.9422070462088333,
+      "grad_norm": 0.941901683807373,
+      "learning_rate": 1.45410457348407e-06,
+      "loss": 0.469,
+      "step": 29525
+    },
+    {
+      "epoch": 0.9430048506510085,
+      "grad_norm": 0.8407112956047058,
+      "learning_rate": 1.4340313463514902e-06,
+      "loss": 0.4813,
+      "step": 29550
+    },
+    {
+      "epoch": 0.9438026550931835,
+      "grad_norm": 1.4796777963638306,
+      "learning_rate": 1.4139581192189107e-06,
+      "loss": 0.4733,
+      "step": 29575
+    },
+    {
+      "epoch": 0.9446004595353586,
+      "grad_norm": 0.6006361842155457,
+      "learning_rate": 1.393884892086331e-06,
+      "loss": 0.4195,
+      "step": 29600
+    },
+    {
+      "epoch": 0.9453982639775338,
+      "grad_norm": 0.835727334022522,
+      "learning_rate": 1.3738116649537513e-06,
+      "loss": 0.4281,
+      "step": 29625
+    },
+    {
+      "epoch": 0.9461960684197089,
+      "grad_norm": 0.8317933678627014,
+      "learning_rate": 1.3537384378211718e-06,
+      "loss": 0.4221,
+      "step": 29650
+    },
+    {
+      "epoch": 0.9469938728618841,
+      "grad_norm": 2.135981559753418,
+      "learning_rate": 1.333665210688592e-06,
+      "loss": 0.3963,
+      "step": 29675
+    },
+    {
+      "epoch": 0.9477916773040592,
+      "grad_norm": 1.3075852394104004,
+      "learning_rate": 1.3135919835560124e-06,
+      "loss": 0.4323,
+      "step": 29700
+    },
+    {
+      "epoch": 0.9485894817462344,
+      "grad_norm": 0.9642304182052612,
+      "learning_rate": 1.2935187564234327e-06,
+      "loss": 0.4221,
+      "step": 29725
+    },
+    {
+      "epoch": 0.9493872861884095,
+      "grad_norm": 1.0338194370269775,
+      "learning_rate": 1.273445529290853e-06,
+      "loss": 0.393,
+      "step": 29750
+    },
+    {
+      "epoch": 0.9501850906305847,
+      "grad_norm": 0.676669180393219,
+      "learning_rate": 1.2533723021582734e-06,
+      "loss": 0.4709,
+      "step": 29775
+    },
+    {
+      "epoch": 0.9509828950727598,
+      "grad_norm": 1.3416296243667603,
+      "learning_rate": 1.2332990750256937e-06,
+      "loss": 0.5081,
+      "step": 29800
+    },
+    {
+      "epoch": 0.9517806995149349,
+      "grad_norm": 0.7452605962753296,
+      "learning_rate": 1.2132258478931142e-06,
+      "loss": 0.4765,
+      "step": 29825
+    },
+    {
+      "epoch": 0.9525785039571101,
+      "grad_norm": 1.2132765054702759,
+      "learning_rate": 1.1931526207605345e-06,
+      "loss": 0.4737,
+      "step": 29850
+    },
+    {
+      "epoch": 0.9533763083992852,
+      "grad_norm": 1.0834319591522217,
+      "learning_rate": 1.1730793936279548e-06,
+      "loss": 0.3876,
+      "step": 29875
+    },
+    {
+      "epoch": 0.9541741128414603,
+      "grad_norm": 0.4042583405971527,
+      "learning_rate": 1.153006166495375e-06,
+      "loss": 0.4086,
+      "step": 29900
+    },
+    {
+      "epoch": 0.9549719172836354,
+      "grad_norm": 0.618110716342926,
+      "learning_rate": 1.1329329393627954e-06,
+      "loss": 0.5247,
+      "step": 29925
+    },
+    {
+      "epoch": 0.9557697217258105,
+      "grad_norm": 0.5366015434265137,
+      "learning_rate": 1.1128597122302159e-06,
+      "loss": 0.3987,
+      "step": 29950
+    },
+    {
+      "epoch": 0.9565675261679857,
+      "grad_norm": 0.7964014410972595,
+      "learning_rate": 1.0927864850976362e-06,
+      "loss": 0.4654,
+      "step": 29975
+    },
+    {
+      "epoch": 0.9573653306101608,
+      "grad_norm": 1.018946886062622,
+      "learning_rate": 1.0727132579650566e-06,
+      "loss": 0.4827,
+      "step": 30000
+    }
+  ],
+  "logging_steps": 25,
+  "max_steps": 31336,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 1000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 4.93204147273728e+18,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}