diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,13503 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.996102883865939,
+  "eval_steps": 500,
+  "global_step": 1923,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.001558846453624318,
+      "grad_norm": 52.22184914143673,
+      "learning_rate": 2.590673575129534e-07,
+      "loss": 11.3056,
+      "step": 1
+    },
+    {
+      "epoch": 0.003117692907248636,
+      "grad_norm": 55.3801908988395,
+      "learning_rate": 5.181347150259068e-07,
+      "loss": 11.1148,
+      "step": 2
+    },
+    {
+      "epoch": 0.004676539360872954,
+      "grad_norm": 53.76354651632375,
+      "learning_rate": 7.772020725388602e-07,
+      "loss": 11.215,
+      "step": 3
+    },
+    {
+      "epoch": 0.006235385814497272,
+      "grad_norm": 52.06232793344078,
+      "learning_rate": 1.0362694300518136e-06,
+      "loss": 11.2145,
+      "step": 4
+    },
+    {
+      "epoch": 0.00779423226812159,
+      "grad_norm": 55.06385321064369,
+      "learning_rate": 1.2953367875647669e-06,
+      "loss": 11.1779,
+      "step": 5
+    },
+    {
+      "epoch": 0.009353078721745909,
+      "grad_norm": 53.1236773756813,
+      "learning_rate": 1.5544041450777204e-06,
+      "loss": 11.1812,
+      "step": 6
+    },
+    {
+      "epoch": 0.010911925175370226,
+      "grad_norm": 58.28685190281901,
+      "learning_rate": 1.8134715025906736e-06,
+      "loss": 10.864,
+      "step": 7
+    },
+    {
+      "epoch": 0.012470771628994544,
+      "grad_norm": 57.49860454762129,
+      "learning_rate": 2.0725388601036273e-06,
+      "loss": 10.8997,
+      "step": 8
+    },
+    {
+      "epoch": 0.014029618082618862,
+      "grad_norm": 61.33406437121332,
+      "learning_rate": 2.3316062176165805e-06,
+      "loss": 10.7659,
+      "step": 9
+    },
+    {
+      "epoch": 0.01558846453624318,
+      "grad_norm": 83.51848150551032,
+      "learning_rate": 2.5906735751295338e-06,
+      "loss": 9.7158,
+      "step": 10
+    },
+    {
+      "epoch": 0.0171473109898675,
+      "grad_norm": 87.40268379033651,
+      "learning_rate": 2.8497409326424875e-06,
+      "loss": 9.3576,
+      "step": 11
+    },
+    {
+      "epoch": 0.018706157443491817,
+      "grad_norm": 94.86364555239567,
+      "learning_rate": 3.1088082901554407e-06,
+      "loss": 9.0476,
+      "step": 12
+    },
+    {
+      "epoch": 0.020265003897116135,
+      "grad_norm": 66.58727850770262,
+      "learning_rate": 3.367875647668394e-06,
+      "loss": 3.7998,
+      "step": 13
+    },
+    {
+      "epoch": 0.021823850350740453,
+      "grad_norm": 60.23863297900514,
+      "learning_rate": 3.626943005181347e-06,
+      "loss": 3.5631,
+      "step": 14
+    },
+    {
+      "epoch": 0.02338269680436477,
+      "grad_norm": 42.715651899746156,
+      "learning_rate": 3.886010362694301e-06,
+      "loss": 2.8496,
+      "step": 15
+    },
+    {
+      "epoch": 0.02494154325798909,
+      "grad_norm": 35.86036496654402,
+      "learning_rate": 4.1450777202072546e-06,
+      "loss": 2.5679,
+      "step": 16
+    },
+    {
+      "epoch": 0.026500389711613406,
+      "grad_norm": 8.204333077091338,
+      "learning_rate": 4.404145077720207e-06,
+      "loss": 1.4646,
+      "step": 17
+    },
+    {
+      "epoch": 0.028059236165237724,
+      "grad_norm": 5.413485907710649,
+      "learning_rate": 4.663212435233161e-06,
+      "loss": 1.3395,
+      "step": 18
+    },
+    {
+      "epoch": 0.029618082618862042,
+      "grad_norm": 4.181557631031653,
+      "learning_rate": 4.922279792746115e-06,
+      "loss": 1.282,
+      "step": 19
+    },
+    {
+      "epoch": 0.03117692907248636,
+      "grad_norm": 3.3106118750979614,
+      "learning_rate": 5.1813471502590676e-06,
+      "loss": 1.192,
+      "step": 20
+    },
+    {
+      "epoch": 0.03273577552611068,
+      "grad_norm": 2.5377335055220023,
+      "learning_rate": 5.440414507772021e-06,
+      "loss": 1.1684,
+      "step": 21
+    },
+    {
+      "epoch": 0.034294621979735,
+      "grad_norm": 2.1309505059833436,
+      "learning_rate": 5.699481865284975e-06,
+      "loss": 1.0969,
+      "step": 22
+    },
+    {
+      "epoch": 0.03585346843335931,
+      "grad_norm": 3.8622301532964998,
+      "learning_rate": 5.958549222797928e-06,
+      "loss": 1.0732,
+      "step": 23
+    },
+    {
+      "epoch": 0.037412314886983634,
+      "grad_norm": 47.58759142573856,
+      "learning_rate": 6.217616580310881e-06,
+      "loss": 1.0184,
+      "step": 24
+    },
+    {
+      "epoch": 0.03897116134060795,
+      "grad_norm": 16.958376825756968,
+      "learning_rate": 6.476683937823834e-06,
+      "loss": 0.9784,
+      "step": 25
+    },
+    {
+      "epoch": 0.04053000779423227,
+      "grad_norm": 1.7724891024655978,
+      "learning_rate": 6.735751295336788e-06,
+      "loss": 0.9303,
+      "step": 26
+    },
+    {
+      "epoch": 0.042088854247856584,
+      "grad_norm": 1.2517023402408312,
+      "learning_rate": 6.994818652849741e-06,
+      "loss": 0.8952,
+      "step": 27
+    },
+    {
+      "epoch": 0.043647700701480906,
+      "grad_norm": 1.0169924063509375,
+      "learning_rate": 7.253886010362694e-06,
+      "loss": 0.8697,
+      "step": 28
+    },
+    {
+      "epoch": 0.04520654715510522,
+      "grad_norm": 0.8637449789891917,
+      "learning_rate": 7.512953367875648e-06,
+      "loss": 0.8588,
+      "step": 29
+    },
+    {
+      "epoch": 0.04676539360872954,
+      "grad_norm": 0.8722633450269738,
+      "learning_rate": 7.772020725388602e-06,
+      "loss": 0.8327,
+      "step": 30
+    },
+    {
+      "epoch": 0.048324240062353856,
+      "grad_norm": 0.7979319477336262,
+      "learning_rate": 8.031088082901555e-06,
+      "loss": 0.815,
+      "step": 31
+    },
+    {
+      "epoch": 0.04988308651597818,
+      "grad_norm": 0.7636414965508945,
+      "learning_rate": 8.290155440414509e-06,
+      "loss": 0.7699,
+      "step": 32
+    },
+    {
+      "epoch": 0.05144193296960249,
+      "grad_norm": 0.6723426478607053,
+      "learning_rate": 8.549222797927462e-06,
+      "loss": 0.7777,
+      "step": 33
+    },
+    {
+      "epoch": 0.05300077942322681,
+      "grad_norm": 0.671680033034605,
+      "learning_rate": 8.808290155440415e-06,
+      "loss": 0.7429,
+      "step": 34
+    },
+    {
+      "epoch": 0.05455962587685113,
+      "grad_norm": 0.6367482766420466,
+      "learning_rate": 9.06735751295337e-06,
+      "loss": 0.7442,
+      "step": 35
+    },
+    {
+      "epoch": 0.05611847233047545,
+      "grad_norm": 0.5985756418410207,
+      "learning_rate": 9.326424870466322e-06,
+      "loss": 0.7354,
+      "step": 36
+    },
+    {
+      "epoch": 0.05767731878409977,
+      "grad_norm": 0.5965999524800757,
+      "learning_rate": 9.585492227979275e-06,
+      "loss": 0.7226,
+      "step": 37
+    },
+    {
+      "epoch": 0.059236165237724084,
+      "grad_norm": 0.5390447035986822,
+      "learning_rate": 9.84455958549223e-06,
+      "loss": 0.6732,
+      "step": 38
+    },
+    {
+      "epoch": 0.060795011691348405,
+      "grad_norm": 0.4740676330023613,
+      "learning_rate": 1.0103626943005182e-05,
+      "loss": 0.6923,
+      "step": 39
+    },
+    {
+      "epoch": 0.06235385814497272,
+      "grad_norm": 0.4870908612880822,
+      "learning_rate": 1.0362694300518135e-05,
+      "loss": 0.6898,
+      "step": 40
+    },
+    {
+      "epoch": 0.06391270459859703,
+      "grad_norm": 0.49048808676105743,
+      "learning_rate": 1.062176165803109e-05,
+      "loss": 0.6817,
+      "step": 41
+    },
+    {
+      "epoch": 0.06547155105222136,
+      "grad_norm": 0.4775360211701637,
+      "learning_rate": 1.0880829015544042e-05,
+      "loss": 0.6806,
+      "step": 42
+    },
+    {
+      "epoch": 0.06703039750584568,
+      "grad_norm": 0.540847742943483,
+      "learning_rate": 1.1139896373056995e-05,
+      "loss": 0.6548,
+      "step": 43
+    },
+    {
+      "epoch": 0.06858924395947,
+      "grad_norm": 0.4168099062285217,
+      "learning_rate": 1.139896373056995e-05,
+      "loss": 0.6584,
+      "step": 44
+    },
+    {
+      "epoch": 0.0701480904130943,
+      "grad_norm": 0.4163477313182357,
+      "learning_rate": 1.1658031088082903e-05,
+      "loss": 0.6511,
+      "step": 45
+    },
+    {
+      "epoch": 0.07170693686671863,
+      "grad_norm": 0.4425510534250163,
+      "learning_rate": 1.1917098445595855e-05,
+      "loss": 0.6432,
+      "step": 46
+    },
+    {
+      "epoch": 0.07326578332034295,
+      "grad_norm": 0.3769848201392934,
+      "learning_rate": 1.2176165803108808e-05,
+      "loss": 0.6004,
+      "step": 47
+    },
+    {
+      "epoch": 0.07482462977396727,
+      "grad_norm": 0.36307003967305024,
+      "learning_rate": 1.2435233160621763e-05,
+      "loss": 0.6339,
+      "step": 48
+    },
+    {
+      "epoch": 0.07638347622759158,
+      "grad_norm": 0.36365771900218874,
+      "learning_rate": 1.2694300518134716e-05,
+      "loss": 0.6573,
+      "step": 49
+    },
+    {
+      "epoch": 0.0779423226812159,
+      "grad_norm": 0.3554573096821652,
+      "learning_rate": 1.2953367875647668e-05,
+      "loss": 0.6606,
+      "step": 50
+    },
+    {
+      "epoch": 0.07950116913484022,
+      "grad_norm": 0.4045859236452954,
+      "learning_rate": 1.3212435233160623e-05,
+      "loss": 0.6475,
+      "step": 51
+    },
+    {
+      "epoch": 0.08106001558846454,
+      "grad_norm": 0.32890043969215965,
+      "learning_rate": 1.3471502590673576e-05,
+      "loss": 0.6148,
+      "step": 52
+    },
+    {
+      "epoch": 0.08261886204208885,
+      "grad_norm": 0.3083895203558219,
+      "learning_rate": 1.3730569948186529e-05,
+      "loss": 0.6347,
+      "step": 53
+    },
+    {
+      "epoch": 0.08417770849571317,
+      "grad_norm": 0.33340100324523925,
+      "learning_rate": 1.3989637305699481e-05,
+      "loss": 0.6165,
+      "step": 54
+    },
+    {
+      "epoch": 0.08573655494933749,
+      "grad_norm": 0.29517015584889045,
+      "learning_rate": 1.4248704663212436e-05,
+      "loss": 0.5802,
+      "step": 55
+    },
+    {
+      "epoch": 0.08729540140296181,
+      "grad_norm": 0.28430427065168046,
+      "learning_rate": 1.4507772020725389e-05,
+      "loss": 0.5895,
+      "step": 56
+    },
+    {
+      "epoch": 0.08885424785658613,
+      "grad_norm": 0.3338831939267004,
+      "learning_rate": 1.4766839378238342e-05,
+      "loss": 0.5715,
+      "step": 57
+    },
+    {
+      "epoch": 0.09041309431021044,
+      "grad_norm": 0.3093980059650245,
+      "learning_rate": 1.5025906735751296e-05,
+      "loss": 0.5847,
+      "step": 58
+    },
+    {
+      "epoch": 0.09197194076383476,
+      "grad_norm": 0.2948901838276863,
+      "learning_rate": 1.528497409326425e-05,
+      "loss": 0.5995,
+      "step": 59
+    },
+    {
+      "epoch": 0.09353078721745908,
+      "grad_norm": 0.31762615759493373,
+      "learning_rate": 1.5544041450777204e-05,
+      "loss": 0.6043,
+      "step": 60
+    },
+    {
+      "epoch": 0.0950896336710834,
+      "grad_norm": 0.29491512166284634,
+      "learning_rate": 1.5803108808290158e-05,
+      "loss": 0.5992,
+      "step": 61
+    },
+    {
+      "epoch": 0.09664848012470771,
+      "grad_norm": 0.29897052685770187,
+      "learning_rate": 1.606217616580311e-05,
+      "loss": 0.5931,
+      "step": 62
+    },
+    {
+      "epoch": 0.09820732657833203,
+      "grad_norm": 0.30055023786228835,
+      "learning_rate": 1.6321243523316064e-05,
+      "loss": 0.5849,
+      "step": 63
+    },
+    {
+      "epoch": 0.09976617303195635,
+      "grad_norm": 0.26046107559568865,
+      "learning_rate": 1.6580310880829018e-05,
+      "loss": 0.5512,
+      "step": 64
+    },
+    {
+      "epoch": 0.10132501948558068,
+      "grad_norm": 0.30065721102538445,
+      "learning_rate": 1.683937823834197e-05,
+      "loss": 0.5814,
+      "step": 65
+    },
+    {
+      "epoch": 0.10288386593920498,
+      "grad_norm": 0.29570037471107435,
+      "learning_rate": 1.7098445595854924e-05,
+      "loss": 0.5889,
+      "step": 66
+    },
+    {
+      "epoch": 0.1044427123928293,
+      "grad_norm": 0.2675735266669228,
+      "learning_rate": 1.735751295336788e-05,
+      "loss": 0.5442,
+      "step": 67
+    },
+    {
+      "epoch": 0.10600155884645363,
+      "grad_norm": 0.3100468880818413,
+      "learning_rate": 1.761658031088083e-05,
+      "loss": 0.5776,
+      "step": 68
+    },
+    {
+      "epoch": 0.10756040530007795,
+      "grad_norm": 0.27731104939451146,
+      "learning_rate": 1.7875647668393784e-05,
+      "loss": 0.5644,
+      "step": 69
+    },
+    {
+      "epoch": 0.10911925175370225,
+      "grad_norm": 0.3214790758034934,
+      "learning_rate": 1.813471502590674e-05,
+      "loss": 0.5361,
+      "step": 70
+    },
+    {
+      "epoch": 0.11067809820732658,
+      "grad_norm": 0.29991490625376055,
+      "learning_rate": 1.839378238341969e-05,
+      "loss": 0.5567,
+      "step": 71
+    },
+    {
+      "epoch": 0.1122369446609509,
+      "grad_norm": 0.2826573502857747,
+      "learning_rate": 1.8652849740932644e-05,
+      "loss": 0.5901,
+      "step": 72
+    },
+    {
+      "epoch": 0.11379579111457522,
+      "grad_norm": 0.30249730284978976,
+      "learning_rate": 1.89119170984456e-05,
+      "loss": 0.5699,
+      "step": 73
+    },
+    {
+      "epoch": 0.11535463756819954,
+      "grad_norm": 0.32336196262393,
+      "learning_rate": 1.917098445595855e-05,
+      "loss": 0.5842,
+      "step": 74
+    },
+    {
+      "epoch": 0.11691348402182385,
+      "grad_norm": 0.2940197733300188,
+      "learning_rate": 1.9430051813471504e-05,
+      "loss": 0.56,
+      "step": 75
+    },
+    {
+      "epoch": 0.11847233047544817,
+      "grad_norm": 0.27234922818526347,
+      "learning_rate": 1.968911917098446e-05,
+      "loss": 0.5596,
+      "step": 76
+    },
+    {
+      "epoch": 0.12003117692907249,
+      "grad_norm": 0.31604251280714674,
+      "learning_rate": 1.994818652849741e-05,
+      "loss": 0.5638,
+      "step": 77
+    },
+    {
+      "epoch": 0.12159002338269681,
+      "grad_norm": 0.2640909764215779,
+      "learning_rate": 2.0207253886010365e-05,
+      "loss": 0.5357,
+      "step": 78
+    },
+    {
+      "epoch": 0.12314886983632112,
+      "grad_norm": 0.3325125655849214,
+      "learning_rate": 2.046632124352332e-05,
+      "loss": 0.5634,
+      "step": 79
+    },
+    {
+      "epoch": 0.12470771628994544,
+      "grad_norm": 0.2728623091830791,
+      "learning_rate": 2.072538860103627e-05,
+      "loss": 0.567,
+      "step": 80
+    },
+    {
+      "epoch": 0.12626656274356976,
+      "grad_norm": 0.3120421349580941,
+      "learning_rate": 2.0984455958549225e-05,
+      "loss": 0.5428,
+      "step": 81
+    },
+    {
+      "epoch": 0.12782540919719407,
+      "grad_norm": 0.2771731681198735,
+      "learning_rate": 2.124352331606218e-05,
+      "loss": 0.5526,
+      "step": 82
+    },
+    {
+      "epoch": 0.1293842556508184,
+      "grad_norm": 0.3027775941751013,
+      "learning_rate": 2.150259067357513e-05,
+      "loss": 0.5432,
+      "step": 83
+    },
+    {
+      "epoch": 0.1309431021044427,
+      "grad_norm": 0.3063900417637035,
+      "learning_rate": 2.1761658031088085e-05,
+      "loss": 0.5633,
+      "step": 84
+    },
+    {
+      "epoch": 0.13250194855806702,
+      "grad_norm": 0.29978974426938987,
+      "learning_rate": 2.202072538860104e-05,
+      "loss": 0.5521,
+      "step": 85
+    },
+    {
+      "epoch": 0.13406079501169135,
+      "grad_norm": 0.3066667325804301,
+      "learning_rate": 2.227979274611399e-05,
+      "loss": 0.5576,
+      "step": 86
+    },
+    {
+      "epoch": 0.13561964146531566,
+      "grad_norm": 0.3257258293999494,
+      "learning_rate": 2.2538860103626945e-05,
+      "loss": 0.5409,
+      "step": 87
+    },
+    {
+      "epoch": 0.13717848791894,
+      "grad_norm": 0.32964630331063166,
+      "learning_rate": 2.27979274611399e-05,
+      "loss": 0.5524,
+      "step": 88
+    },
+    {
+      "epoch": 0.1387373343725643,
+      "grad_norm": 0.3390745114943544,
+      "learning_rate": 2.305699481865285e-05,
+      "loss": 0.5539,
+      "step": 89
+    },
+    {
+      "epoch": 0.1402961808261886,
+      "grad_norm": 0.296430130882943,
+      "learning_rate": 2.3316062176165805e-05,
+      "loss": 0.5519,
+      "step": 90
+    },
+    {
+      "epoch": 0.14185502727981295,
+      "grad_norm": 0.3159884102421548,
+      "learning_rate": 2.3575129533678756e-05,
+      "loss": 0.5265,
+      "step": 91
+    },
+    {
+      "epoch": 0.14341387373343725,
+      "grad_norm": 0.3136672635617097,
+      "learning_rate": 2.383419689119171e-05,
+      "loss": 0.5331,
+      "step": 92
+    },
+    {
+      "epoch": 0.1449727201870616,
+      "grad_norm": 0.3203194252949439,
+      "learning_rate": 2.4093264248704665e-05,
+      "loss": 0.5472,
+      "step": 93
+    },
+    {
+      "epoch": 0.1465315666406859,
+      "grad_norm": 0.34456319750216174,
+      "learning_rate": 2.4352331606217617e-05,
+      "loss": 0.5575,
+      "step": 94
+    },
+    {
+      "epoch": 0.1480904130943102,
+      "grad_norm": 0.3228736119075391,
+      "learning_rate": 2.461139896373057e-05,
+      "loss": 0.5293,
+      "step": 95
+    },
+    {
+      "epoch": 0.14964925954793454,
+      "grad_norm": 0.34523710171851074,
+      "learning_rate": 2.4870466321243526e-05,
+      "loss": 0.5287,
+      "step": 96
+    },
+    {
+      "epoch": 0.15120810600155885,
+      "grad_norm": 0.3667004301126633,
+      "learning_rate": 2.5129533678756477e-05,
+      "loss": 0.5385,
+      "step": 97
+    },
+    {
+      "epoch": 0.15276695245518315,
+      "grad_norm": 0.36137195700706454,
+      "learning_rate": 2.538860103626943e-05,
+      "loss": 0.5182,
+      "step": 98
+    },
+    {
+      "epoch": 0.1543257989088075,
+      "grad_norm": 0.27897710920322044,
+      "learning_rate": 2.5647668393782386e-05,
+      "loss": 0.5301,
+      "step": 99
+    },
+    {
+      "epoch": 0.1558846453624318,
+      "grad_norm": 0.28832759123856166,
+      "learning_rate": 2.5906735751295337e-05,
+      "loss": 0.5118,
+      "step": 100
+    },
+    {
+      "epoch": 0.15744349181605613,
+      "grad_norm": 0.34377322242370634,
+      "learning_rate": 2.616580310880829e-05,
+      "loss": 0.532,
+      "step": 101
+    },
+    {
+      "epoch": 0.15900233826968044,
+      "grad_norm": 0.3036711410586048,
+      "learning_rate": 2.6424870466321246e-05,
+      "loss": 0.5165,
+      "step": 102
+    },
+    {
+      "epoch": 0.16056118472330475,
+      "grad_norm": 0.36103093391647273,
+      "learning_rate": 2.6683937823834197e-05,
+      "loss": 0.5287,
+      "step": 103
+    },
+    {
+      "epoch": 0.16212003117692908,
+      "grad_norm": 0.29337030826817656,
+      "learning_rate": 2.694300518134715e-05,
+      "loss": 0.528,
+      "step": 104
+    },
+    {
+      "epoch": 0.1636788776305534,
+      "grad_norm": 0.33805107509606247,
+      "learning_rate": 2.7202072538860106e-05,
+      "loss": 0.5354,
+      "step": 105
+    },
+    {
+      "epoch": 0.1652377240841777,
+      "grad_norm": 0.31771328671144855,
+      "learning_rate": 2.7461139896373057e-05,
+      "loss": 0.5282,
+      "step": 106
+    },
+    {
+      "epoch": 0.16679657053780203,
+      "grad_norm": 0.28313518361082185,
+      "learning_rate": 2.7720207253886012e-05,
+      "loss": 0.5086,
+      "step": 107
+    },
+    {
+      "epoch": 0.16835541699142634,
+      "grad_norm": 0.34173304219782913,
+      "learning_rate": 2.7979274611398963e-05,
+      "loss": 0.5266,
+      "step": 108
+    },
+    {
+      "epoch": 0.16991426344505067,
+      "grad_norm": 0.2895019297309633,
+      "learning_rate": 2.8238341968911917e-05,
+      "loss": 0.5305,
+      "step": 109
+    },
+    {
+      "epoch": 0.17147310989867498,
+      "grad_norm": 0.3271405004182382,
+      "learning_rate": 2.8497409326424872e-05,
+      "loss": 0.5412,
+      "step": 110
+    },
+    {
+      "epoch": 0.1730319563522993,
+      "grad_norm": 0.3254108424111249,
+      "learning_rate": 2.8756476683937827e-05,
+      "loss": 0.5267,
+      "step": 111
+    },
+    {
+      "epoch": 0.17459080280592362,
+      "grad_norm": 0.43636064754159903,
+      "learning_rate": 2.9015544041450778e-05,
+      "loss": 0.546,
+      "step": 112
+    },
+    {
+      "epoch": 0.17614964925954793,
+      "grad_norm": 0.35895062463773536,
+      "learning_rate": 2.9274611398963732e-05,
+      "loss": 0.5379,
+      "step": 113
+    },
+    {
+      "epoch": 0.17770849571317227,
+      "grad_norm": 0.48008348852035654,
+      "learning_rate": 2.9533678756476683e-05,
+      "loss": 0.5251,
+      "step": 114
+    },
+    {
+      "epoch": 0.17926734216679657,
+      "grad_norm": 0.40042199319895483,
+      "learning_rate": 2.9792746113989638e-05,
+      "loss": 0.5043,
+      "step": 115
+    },
+    {
+      "epoch": 0.18082618862042088,
+      "grad_norm": 0.44332954681106973,
+      "learning_rate": 3.0051813471502592e-05,
+      "loss": 0.5135,
+      "step": 116
+    },
+    {
+      "epoch": 0.18238503507404522,
+      "grad_norm": 0.43275195875835987,
+      "learning_rate": 3.0310880829015547e-05,
+      "loss": 0.5196,
+      "step": 117
+    },
+    {
+      "epoch": 0.18394388152766952,
+      "grad_norm": 0.3939079245195568,
+      "learning_rate": 3.05699481865285e-05,
+      "loss": 0.5261,
+      "step": 118
+    },
+    {
+      "epoch": 0.18550272798129383,
+      "grad_norm": 0.45003761498432127,
+      "learning_rate": 3.082901554404145e-05,
+      "loss": 0.5285,
+      "step": 119
+    },
+    {
+      "epoch": 0.18706157443491817,
+      "grad_norm": 0.38744988267232733,
+      "learning_rate": 3.108808290155441e-05,
+      "loss": 0.5183,
+      "step": 120
+    },
+    {
+      "epoch": 0.18862042088854247,
+      "grad_norm": 0.42417740462219883,
+      "learning_rate": 3.134715025906736e-05,
+      "loss": 0.5366,
+      "step": 121
+    },
+    {
+      "epoch": 0.1901792673421668,
+      "grad_norm": 0.38878746316666607,
+      "learning_rate": 3.1606217616580316e-05,
+      "loss": 0.5405,
+      "step": 122
+    },
+    {
+      "epoch": 0.19173811379579112,
+      "grad_norm": 0.39470480511600553,
+      "learning_rate": 3.186528497409327e-05,
+      "loss": 0.5177,
+      "step": 123
+    },
+    {
+      "epoch": 0.19329696024941542,
+      "grad_norm": 0.3425139995303132,
+      "learning_rate": 3.212435233160622e-05,
+      "loss": 0.5278,
+      "step": 124
+    },
+    {
+      "epoch": 0.19485580670303976,
+      "grad_norm": 0.3855003673988402,
+      "learning_rate": 3.238341968911917e-05,
+      "loss": 0.5303,
+      "step": 125
+    },
+    {
+      "epoch": 0.19641465315666407,
+      "grad_norm": 0.3708329304695599,
+      "learning_rate": 3.264248704663213e-05,
+      "loss": 0.5308,
+      "step": 126
+    },
+    {
+      "epoch": 0.1979734996102884,
+      "grad_norm": 0.42197888435438696,
+      "learning_rate": 3.290155440414508e-05,
+      "loss": 0.5199,
+      "step": 127
+    },
+    {
+      "epoch": 0.1995323460639127,
+      "grad_norm": 0.39698217513178863,
+      "learning_rate": 3.3160621761658036e-05,
+      "loss": 0.506,
+      "step": 128
+    },
+    {
+      "epoch": 0.20109119251753702,
+      "grad_norm": 0.4043130382762473,
+      "learning_rate": 3.341968911917099e-05,
+      "loss": 0.5065,
+      "step": 129
+    },
+    {
+      "epoch": 0.20265003897116135,
+      "grad_norm": 0.34367190967592626,
+      "learning_rate": 3.367875647668394e-05,
+      "loss": 0.509,
+      "step": 130
+    },
+    {
+      "epoch": 0.20420888542478566,
+      "grad_norm": 0.49458742478228435,
+      "learning_rate": 3.393782383419689e-05,
+      "loss": 0.5057,
+      "step": 131
+    },
+    {
+      "epoch": 0.20576773187840997,
+      "grad_norm": 0.3946670126655184,
+      "learning_rate": 3.419689119170985e-05,
+      "loss": 0.5074,
+      "step": 132
+    },
+    {
+      "epoch": 0.2073265783320343,
+      "grad_norm": 0.4590062123156051,
+      "learning_rate": 3.44559585492228e-05,
+      "loss": 0.5196,
+      "step": 133
+    },
+    {
+      "epoch": 0.2088854247856586,
+      "grad_norm": 0.41794015093559495,
+      "learning_rate": 3.471502590673576e-05,
+      "loss": 0.5293,
+      "step": 134
+    },
+    {
+      "epoch": 0.21044427123928294,
+      "grad_norm": 0.3976262150889235,
+      "learning_rate": 3.497409326424871e-05,
+      "loss": 0.4958,
+      "step": 135
+    },
+    {
+      "epoch": 0.21200311769290725,
+      "grad_norm": 0.3801765387110536,
+      "learning_rate": 3.523316062176166e-05,
+      "loss": 0.5283,
+      "step": 136
+    },
+    {
+      "epoch": 0.21356196414653156,
+      "grad_norm": 0.4393098932539108,
+      "learning_rate": 3.549222797927461e-05,
+      "loss": 0.5106,
+      "step": 137
+    },
+    {
+      "epoch": 0.2151208106001559,
+      "grad_norm": 0.37730905366487805,
+      "learning_rate": 3.575129533678757e-05,
+      "loss": 0.5004,
+      "step": 138
+    },
+    {
+      "epoch": 0.2166796570537802,
+      "grad_norm": 0.47264900412277805,
+      "learning_rate": 3.601036269430052e-05,
+      "loss": 0.5107,
+      "step": 139
+    },
+    {
+      "epoch": 0.2182385035074045,
+      "grad_norm": 0.4572211602618532,
+      "learning_rate": 3.626943005181348e-05,
+      "loss": 0.5258,
+      "step": 140
+    },
+    {
+      "epoch": 0.21979734996102884,
+      "grad_norm": 0.566903195196336,
+      "learning_rate": 3.652849740932643e-05,
+      "loss": 0.4926,
+      "step": 141
+    },
+    {
+      "epoch": 0.22135619641465315,
+      "grad_norm": 0.40493119123467375,
+      "learning_rate": 3.678756476683938e-05,
+      "loss": 0.5167,
+      "step": 142
+    },
+    {
+      "epoch": 0.22291504286827749,
+      "grad_norm": 0.5800300099928885,
+      "learning_rate": 3.704663212435233e-05,
+      "loss": 0.4888,
+      "step": 143
+    },
+    {
+      "epoch": 0.2244738893219018,
+      "grad_norm": 0.49267006079774145,
+      "learning_rate": 3.730569948186529e-05,
+      "loss": 0.4931,
+      "step": 144
+    },
+    {
+      "epoch": 0.2260327357755261,
+      "grad_norm": 0.5509681396457901,
+      "learning_rate": 3.756476683937824e-05,
+      "loss": 0.515,
+      "step": 145
+    },
+    {
+      "epoch": 0.22759158222915044,
+      "grad_norm": 0.7109027605533598,
+      "learning_rate": 3.78238341968912e-05,
+      "loss": 0.4932,
+      "step": 146
+    },
+    {
+      "epoch": 0.22915042868277474,
+      "grad_norm": 0.35899713776420916,
+      "learning_rate": 3.808290155440415e-05,
+      "loss": 0.4893,
+      "step": 147
+    },
+    {
+      "epoch": 0.23070927513639908,
+      "grad_norm": 0.5392148285010113,
+      "learning_rate": 3.83419689119171e-05,
+      "loss": 0.4865,
+      "step": 148
+    },
+    {
+      "epoch": 0.23226812159002339,
+      "grad_norm": 0.4200166849207286,
+      "learning_rate": 3.860103626943005e-05,
+      "loss": 0.516,
+      "step": 149
+    },
+    {
+      "epoch": 0.2338269680436477,
+      "grad_norm": 0.4668106441613405,
+      "learning_rate": 3.886010362694301e-05,
+      "loss": 0.5104,
+      "step": 150
+    },
+    {
+      "epoch": 0.23538581449727203,
+      "grad_norm": 0.388638095128956,
+      "learning_rate": 3.911917098445596e-05,
+      "loss": 0.5251,
+      "step": 151
+    },
+    {
+      "epoch": 0.23694466095089634,
+      "grad_norm": 0.3927756772978556,
+      "learning_rate": 3.937823834196892e-05,
+      "loss": 0.4898,
+      "step": 152
+    },
+    {
+      "epoch": 0.23850350740452064,
+      "grad_norm": 0.4474650309815615,
+      "learning_rate": 3.963730569948187e-05,
+      "loss": 0.5002,
+      "step": 153
+    },
+    {
+      "epoch": 0.24006235385814498,
+      "grad_norm": 0.40465455399123634,
+      "learning_rate": 3.989637305699482e-05,
+      "loss": 0.4805,
+      "step": 154
+    },
+    {
+      "epoch": 0.24162120031176929,
+      "grad_norm": 0.46329230044256475,
+      "learning_rate": 4.015544041450777e-05,
+      "loss": 0.5148,
+      "step": 155
+    },
+    {
+      "epoch": 0.24318004676539362,
+      "grad_norm": 0.4221898620021292,
+      "learning_rate": 4.041450777202073e-05,
+      "loss": 0.5087,
+      "step": 156
+    },
+    {
+      "epoch": 0.24473889321901793,
+      "grad_norm": 0.41405351180679445,
+      "learning_rate": 4.067357512953368e-05,
+      "loss": 0.5206,
+      "step": 157
+    },
+    {
+      "epoch": 0.24629773967264224,
+      "grad_norm": 0.3718428876018395,
+      "learning_rate": 4.093264248704664e-05,
+      "loss": 0.5018,
+      "step": 158
+    },
+    {
+      "epoch": 0.24785658612626657,
+      "grad_norm": 0.4108732557264086,
+      "learning_rate": 4.119170984455959e-05,
+      "loss": 0.4721,
+      "step": 159
+    },
+    {
+      "epoch": 0.24941543257989088,
+      "grad_norm": 0.5196664214895078,
+      "learning_rate": 4.145077720207254e-05,
+      "loss": 0.501,
+      "step": 160
+    },
+    {
+      "epoch": 0.2509742790335152,
+      "grad_norm": 0.34375475907972514,
+      "learning_rate": 4.170984455958549e-05,
+      "loss": 0.4944,
+      "step": 161
+    },
+    {
+      "epoch": 0.2525331254871395,
+      "grad_norm": 0.5064449794040312,
+      "learning_rate": 4.196891191709845e-05,
+      "loss": 0.5029,
+      "step": 162
+    },
+    {
+      "epoch": 0.2540919719407638,
+      "grad_norm": 0.4647723748878803,
+      "learning_rate": 4.22279792746114e-05,
+      "loss": 0.5024,
+      "step": 163
+    },
+    {
+      "epoch": 0.25565081839438814,
+      "grad_norm": 0.4006925999521958,
+      "learning_rate": 4.248704663212436e-05,
+      "loss": 0.4655,
+      "step": 164
+    },
+    {
+      "epoch": 0.2572096648480125,
+      "grad_norm": 0.45646986120048183,
+      "learning_rate": 4.274611398963731e-05,
+      "loss": 0.5059,
+      "step": 165
+    },
+    {
+      "epoch": 0.2587685113016368,
+      "grad_norm": 0.4337454689280252,
+      "learning_rate": 4.300518134715026e-05,
+      "loss": 0.4847,
+      "step": 166
+    },
+    {
+      "epoch": 0.2603273577552611,
+      "grad_norm": 0.4818011779118697,
+      "learning_rate": 4.326424870466321e-05,
+      "loss": 0.5039,
+      "step": 167
+    },
+    {
+      "epoch": 0.2618862042088854,
+      "grad_norm": 0.4004895532370599,
+      "learning_rate": 4.352331606217617e-05,
+      "loss": 0.4997,
+      "step": 168
+    },
+    {
+      "epoch": 0.2634450506625097,
+      "grad_norm": 0.4776133221437833,
+      "learning_rate": 4.378238341968912e-05,
+      "loss": 0.4747,
+      "step": 169
+    },
+    {
+      "epoch": 0.26500389711613404,
+      "grad_norm": 0.3542482573234714,
+      "learning_rate": 4.404145077720208e-05,
+      "loss": 0.4995,
+      "step": 170
+    },
+    {
+      "epoch": 0.2665627435697584,
+      "grad_norm": 0.47546134775525584,
+      "learning_rate": 4.430051813471503e-05,
+      "loss": 0.5021,
+      "step": 171
+    },
+    {
+      "epoch": 0.2681215900233827,
+      "grad_norm": 0.37668384089912865,
+      "learning_rate": 4.455958549222798e-05,
+      "loss": 0.4847,
+      "step": 172
+    },
+    {
+      "epoch": 0.269680436477007,
+      "grad_norm": 0.5250094003553915,
+      "learning_rate": 4.481865284974093e-05,
+      "loss": 0.4965,
+      "step": 173
+    },
+    {
+      "epoch": 0.2712392829306313,
+      "grad_norm": 0.5118060286220544,
+      "learning_rate": 4.507772020725389e-05,
+      "loss": 0.5002,
+      "step": 174
+    },
+    {
+      "epoch": 0.2727981293842556,
+      "grad_norm": 0.5439924404821223,
+      "learning_rate": 4.533678756476684e-05,
+      "loss": 0.4935,
+      "step": 175
+    },
+    {
+      "epoch": 0.27435697583788,
+      "grad_norm": 0.46651416281963,
+      "learning_rate": 4.55958549222798e-05,
+      "loss": 0.4936,
+      "step": 176
+    },
+    {
+      "epoch": 0.2759158222915043,
+      "grad_norm": 0.6025736893384493,
+      "learning_rate": 4.585492227979275e-05,
+      "loss": 0.4936,
+      "step": 177
+    },
+    {
+      "epoch": 0.2774746687451286,
+      "grad_norm": 0.49477813313240576,
+      "learning_rate": 4.61139896373057e-05,
+      "loss": 0.5014,
+      "step": 178
+    },
+    {
+      "epoch": 0.2790335151987529,
+      "grad_norm": 0.38453152894256626,
+      "learning_rate": 4.637305699481865e-05,
+      "loss": 0.5026,
+      "step": 179
+    },
+    {
+      "epoch": 0.2805923616523772,
+      "grad_norm": 0.6693985713554009,
+      "learning_rate": 4.663212435233161e-05,
+      "loss": 0.4966,
+      "step": 180
+    },
+    {
+      "epoch": 0.2821512081060016,
+      "grad_norm": 0.49890072866209745,
+      "learning_rate": 4.689119170984456e-05,
+      "loss": 0.4951,
+      "step": 181
+    },
+    {
+      "epoch": 0.2837100545596259,
+      "grad_norm": 0.4986049703889912,
+      "learning_rate": 4.715025906735751e-05,
+      "loss": 0.4848,
+      "step": 182
+    },
+    {
+      "epoch": 0.2852689010132502,
+      "grad_norm": 0.7123183153704133,
+      "learning_rate": 4.740932642487047e-05,
+      "loss": 0.5087,
+      "step": 183
+    },
+    {
+      "epoch": 0.2868277474668745,
+      "grad_norm": 0.5101437582246441,
+      "learning_rate": 4.766839378238342e-05,
+      "loss": 0.496,
+      "step": 184
+    },
+    {
+      "epoch": 0.2883865939204988,
+      "grad_norm": 0.5462740650609419,
+      "learning_rate": 4.792746113989637e-05,
+      "loss": 0.4891,
+      "step": 185
+    },
+    {
+      "epoch": 0.2899454403741232,
+      "grad_norm": 0.6739608210631324,
+      "learning_rate": 4.818652849740933e-05,
+      "loss": 0.4809,
+      "step": 186
+    },
+    {
+      "epoch": 0.2915042868277475,
+      "grad_norm": 0.36454838249845395,
+      "learning_rate": 4.844559585492228e-05,
+      "loss": 0.4826,
+      "step": 187
+    },
+    {
+      "epoch": 0.2930631332813718,
+      "grad_norm": 0.8180118832431814,
+      "learning_rate": 4.870466321243523e-05,
+      "loss": 0.5007,
+      "step": 188
+    },
+    {
+      "epoch": 0.2946219797349961,
+      "grad_norm": 0.8582599255383014,
+      "learning_rate": 4.896373056994819e-05,
+      "loss": 0.5101,
+      "step": 189
+    },
+    {
+      "epoch": 0.2961808261886204,
+      "grad_norm": 0.5339198441824459,
+      "learning_rate": 4.922279792746114e-05,
+      "loss": 0.4755,
+      "step": 190
+    },
+    {
+      "epoch": 0.2977396726422447,
+      "grad_norm": 0.5038691245377587,
+      "learning_rate": 4.948186528497409e-05,
+      "loss": 0.4987,
+      "step": 191
+    },
+    {
+      "epoch": 0.2992985190958691,
+      "grad_norm": 0.7512130026001776,
+      "learning_rate": 4.974093264248705e-05,
+      "loss": 0.5168,
+      "step": 192
+    },
+    {
+      "epoch": 0.3008573655494934,
+      "grad_norm": 0.5965449485798306,
+      "learning_rate": 5e-05,
+      "loss": 0.4825,
+      "step": 193
+    },
+    {
+      "epoch": 0.3024162120031177,
+      "grad_norm": 0.5874775355232069,
+      "learning_rate": 4.9971098265895956e-05,
+      "loss": 0.484,
+      "step": 194
+    },
+    {
+      "epoch": 0.303975058456742,
+      "grad_norm": 0.6794164991276913,
+      "learning_rate": 4.994219653179191e-05,
+      "loss": 0.4956,
+      "step": 195
+    },
+    {
+      "epoch": 0.3055339049103663,
+      "grad_norm": 0.5324211441774743,
+      "learning_rate": 4.991329479768786e-05,
+      "loss": 0.469,
+      "step": 196
+    },
+    {
+      "epoch": 0.30709275136399067,
+      "grad_norm": 0.7381982819903451,
+      "learning_rate": 4.9884393063583816e-05,
+      "loss": 0.4859,
+      "step": 197
+    },
+    {
+      "epoch": 0.308651597817615,
+      "grad_norm": 0.7316455279398204,
+      "learning_rate": 4.985549132947977e-05,
+      "loss": 0.4962,
+      "step": 198
+    },
+    {
+      "epoch": 0.3102104442712393,
+      "grad_norm": 0.5513173476365167,
+      "learning_rate": 4.982658959537572e-05,
+      "loss": 0.4973,
+      "step": 199
+    },
+    {
+      "epoch": 0.3117692907248636,
+      "grad_norm": 0.6961187825800337,
+      "learning_rate": 4.979768786127168e-05,
+      "loss": 0.504,
+      "step": 200
+    },
+    {
+      "epoch": 0.3133281371784879,
+      "grad_norm": 0.6675059607265288,
+      "learning_rate": 4.976878612716763e-05,
+      "loss": 0.4818,
+      "step": 201
+    },
+    {
+      "epoch": 0.31488698363211226,
+      "grad_norm": 0.5775315130950791,
+      "learning_rate": 4.9739884393063583e-05,
+      "loss": 0.5108,
+      "step": 202
+    },
+    {
+      "epoch": 0.31644583008573657,
+      "grad_norm": 0.5493567678427346,
+      "learning_rate": 4.971098265895954e-05,
+      "loss": 0.4933,
+      "step": 203
+    },
+    {
+      "epoch": 0.3180046765393609,
+      "grad_norm": 0.5788493631561638,
+      "learning_rate": 4.96820809248555e-05,
+      "loss": 0.4921,
+      "step": 204
+    },
+    {
+      "epoch": 0.3195635229929852,
+      "grad_norm": 0.5111462707368106,
+      "learning_rate": 4.965317919075145e-05,
+      "loss": 0.4617,
+      "step": 205
+    },
+    {
+      "epoch": 0.3211223694466095,
+      "grad_norm": 0.5647593330505578,
+      "learning_rate": 4.96242774566474e-05,
+      "loss": 0.4953,
+      "step": 206
+    },
+    {
+      "epoch": 0.32268121590023385,
+      "grad_norm": 0.4207725153414849,
+      "learning_rate": 4.959537572254335e-05,
+      "loss": 0.5115,
+      "step": 207
+    },
+    {
+      "epoch": 0.32424006235385816,
+      "grad_norm": 0.4624044434848398,
+      "learning_rate": 4.956647398843931e-05,
+      "loss": 0.4722,
+      "step": 208
+    },
+    {
+      "epoch": 0.32579890880748247,
+      "grad_norm": 0.4376386673543866,
+      "learning_rate": 4.9537572254335264e-05,
+      "loss": 0.4832,
+      "step": 209
+    },
+    {
+      "epoch": 0.3273577552611068,
+      "grad_norm": 0.531865499974187,
+      "learning_rate": 4.950867052023122e-05,
+      "loss": 0.5055,
+      "step": 210
+    },
+    {
+      "epoch": 0.3289166017147311,
+      "grad_norm": 0.5050602424671378,
+      "learning_rate": 4.9479768786127164e-05,
+      "loss": 0.4697,
+      "step": 211
+    },
+    {
+      "epoch": 0.3304754481683554,
+      "grad_norm": 0.38168626861209426,
+      "learning_rate": 4.9450867052023125e-05,
+      "loss": 0.4907,
+      "step": 212
+    },
+    {
+      "epoch": 0.33203429462197975,
+      "grad_norm": 0.5281790174146758,
+      "learning_rate": 4.942196531791908e-05,
+      "loss": 0.4876,
+      "step": 213
+    },
+    {
+      "epoch": 0.33359314107560406,
+      "grad_norm": 0.5297198653114203,
+      "learning_rate": 4.939306358381503e-05,
+      "loss": 0.4913,
+      "step": 214
+    },
+    {
+      "epoch": 0.33515198752922837,
+      "grad_norm": 0.45152753639301624,
+      "learning_rate": 4.9364161849710985e-05,
+      "loss": 0.4731,
+      "step": 215
+    },
+    {
+      "epoch": 0.3367108339828527,
+      "grad_norm": 0.3683086256111994,
+      "learning_rate": 4.933526011560694e-05,
+      "loss": 0.4779,
+      "step": 216
+    },
+    {
+      "epoch": 0.338269680436477,
+      "grad_norm": 0.4739643840797579,
+      "learning_rate": 4.930635838150289e-05,
+      "loss": 0.4992,
+      "step": 217
+    },
+    {
+      "epoch": 0.33982852689010135,
+      "grad_norm": 0.4068837265657986,
+      "learning_rate": 4.9277456647398845e-05,
+      "loss": 0.4996,
+      "step": 218
+    },
+    {
+      "epoch": 0.34138737334372565,
+      "grad_norm": 0.4885011148922096,
+      "learning_rate": 4.92485549132948e-05,
+      "loss": 0.4865,
+      "step": 219
+    },
+    {
+      "epoch": 0.34294621979734996,
+      "grad_norm": 0.493814836908405,
+      "learning_rate": 4.921965317919075e-05,
+      "loss": 0.4731,
+      "step": 220
+    },
+    {
+      "epoch": 0.34450506625097427,
+      "grad_norm": 0.46892553233657863,
+      "learning_rate": 4.9190751445086706e-05,
+      "loss": 0.4699,
+      "step": 221
+    },
+    {
+      "epoch": 0.3460639127045986,
+      "grad_norm": 0.464609243189492,
+      "learning_rate": 4.916184971098266e-05,
+      "loss": 0.478,
+      "step": 222
+    },
+    {
+      "epoch": 0.34762275915822294,
+      "grad_norm": 0.37152543938557336,
+      "learning_rate": 4.913294797687861e-05,
+      "loss": 0.4694,
+      "step": 223
+    },
+    {
+      "epoch": 0.34918160561184725,
+      "grad_norm": 0.5025252675976578,
+      "learning_rate": 4.910404624277457e-05,
+      "loss": 0.4715,
+      "step": 224
+    },
+    {
+      "epoch": 0.35074045206547155,
+      "grad_norm": 0.3402466722282753,
+      "learning_rate": 4.9075144508670526e-05,
+      "loss": 0.463,
+      "step": 225
+    },
+    {
+      "epoch": 0.35229929851909586,
+      "grad_norm": 0.3965061942369021,
+      "learning_rate": 4.904624277456647e-05,
+      "loss": 0.4792,
+      "step": 226
+    },
+    {
+      "epoch": 0.35385814497272017,
+      "grad_norm": 0.4206144149046459,
+      "learning_rate": 4.9017341040462426e-05,
+      "loss": 0.4901,
+      "step": 227
+    },
+    {
+      "epoch": 0.35541699142634453,
+      "grad_norm": 0.4286051420798536,
+      "learning_rate": 4.8988439306358387e-05,
+      "loss": 0.4742,
+      "step": 228
+    },
+    {
+      "epoch": 0.35697583787996884,
+      "grad_norm": 0.47141158644523007,
+      "learning_rate": 4.895953757225434e-05,
+      "loss": 0.4968,
+      "step": 229
+    },
+    {
+      "epoch": 0.35853468433359315,
+      "grad_norm": 0.3415005284976459,
+      "learning_rate": 4.8930635838150293e-05,
+      "loss": 0.4568,
+      "step": 230
+    },
+    {
+      "epoch": 0.36009353078721745,
+      "grad_norm": 0.48287718381612604,
+      "learning_rate": 4.890173410404624e-05,
+      "loss": 0.4993,
+      "step": 231
+    },
+    {
+      "epoch": 0.36165237724084176,
+      "grad_norm": 0.5803189730929726,
+      "learning_rate": 4.88728323699422e-05,
+      "loss": 0.4653,
+      "step": 232
+    },
+    {
+      "epoch": 0.36321122369446607,
+      "grad_norm": 0.3907951855741137,
+      "learning_rate": 4.8843930635838154e-05,
+      "loss": 0.4794,
+      "step": 233
+    },
+    {
+      "epoch": 0.36477007014809043,
+      "grad_norm": 0.5028194362808669,
+      "learning_rate": 4.881502890173411e-05,
+      "loss": 0.4942,
+      "step": 234
+    },
+    {
+      "epoch": 0.36632891660171474,
+      "grad_norm": 0.4945111081672912,
+      "learning_rate": 4.878612716763006e-05,
+      "loss": 0.4771,
+      "step": 235
+    },
+    {
+      "epoch": 0.36788776305533905,
+      "grad_norm": 0.43827490144476094,
+      "learning_rate": 4.8757225433526014e-05,
+      "loss": 0.459,
+      "step": 236
+    },
+    {
+      "epoch": 0.36944660950896335,
+      "grad_norm": 0.5256559700101235,
+      "learning_rate": 4.872832369942197e-05,
+      "loss": 0.4696,
+      "step": 237
+    },
+    {
+      "epoch": 0.37100545596258766,
+      "grad_norm": 0.5899114320019979,
+      "learning_rate": 4.869942196531792e-05,
+      "loss": 0.4769,
+      "step": 238
+    },
+    {
+      "epoch": 0.372564302416212,
+      "grad_norm": 0.4326187078710821,
+      "learning_rate": 4.8670520231213874e-05,
+      "loss": 0.4761,
+      "step": 239
+    },
+    {
+      "epoch": 0.37412314886983633,
+      "grad_norm": 0.4922313590375816,
+      "learning_rate": 4.864161849710983e-05,
+      "loss": 0.4871,
+      "step": 240
+    },
+    {
+      "epoch": 0.37568199532346064,
+      "grad_norm": 0.4403169878610757,
+      "learning_rate": 4.861271676300578e-05,
+      "loss": 0.4918,
+      "step": 241
+    },
+    {
+      "epoch": 0.37724084177708495,
+      "grad_norm": 0.39128925993823416,
+      "learning_rate": 4.8583815028901735e-05,
+      "loss": 0.4695,
+      "step": 242
+    },
+    {
+      "epoch": 0.37879968823070925,
+      "grad_norm": 0.4947169815070576,
+      "learning_rate": 4.855491329479769e-05,
+      "loss": 0.4785,
+      "step": 243
+    },
+    {
+      "epoch": 0.3803585346843336,
+      "grad_norm": 0.36561636850124596,
+      "learning_rate": 4.852601156069365e-05,
+      "loss": 0.4687,
+      "step": 244
+    },
+    {
+      "epoch": 0.3819173811379579,
+      "grad_norm": 0.5740065574334354,
+      "learning_rate": 4.8497109826589595e-05,
+      "loss": 0.4777,
+      "step": 245
+    },
+    {
+      "epoch": 0.38347622759158223,
+      "grad_norm": 0.37164208100522045,
+      "learning_rate": 4.846820809248555e-05,
+      "loss": 0.4769,
+      "step": 246
+    },
+    {
+      "epoch": 0.38503507404520654,
+      "grad_norm": 0.45893063215148555,
+      "learning_rate": 4.84393063583815e-05,
+      "loss": 0.4754,
+      "step": 247
+    },
+    {
+      "epoch": 0.38659392049883085,
+      "grad_norm": 0.4592126499581869,
+      "learning_rate": 4.841040462427746e-05,
+      "loss": 0.4667,
+      "step": 248
+    },
+    {
+      "epoch": 0.3881527669524552,
+      "grad_norm": 0.42130274935477224,
+      "learning_rate": 4.8381502890173416e-05,
+      "loss": 0.4727,
+      "step": 249
+    },
+    {
+      "epoch": 0.3897116134060795,
+      "grad_norm": 0.5562234140465554,
+      "learning_rate": 4.835260115606937e-05,
+      "loss": 0.5142,
+      "step": 250
+    },
+    {
+      "epoch": 0.3912704598597038,
+      "grad_norm": 0.35335610698932396,
+      "learning_rate": 4.8323699421965316e-05,
+      "loss": 0.4726,
+      "step": 251
+    },
+    {
+      "epoch": 0.39282930631332813,
+      "grad_norm": 0.5355853337204556,
+      "learning_rate": 4.8294797687861276e-05,
+      "loss": 0.468,
+      "step": 252
+    },
+    {
+      "epoch": 0.39438815276695244,
+      "grad_norm": 0.3940457770663424,
+      "learning_rate": 4.826589595375723e-05,
+      "loss": 0.4795,
+      "step": 253
+    },
+    {
+      "epoch": 0.3959469992205768,
+      "grad_norm": 0.48308987476155735,
+      "learning_rate": 4.823699421965318e-05,
+      "loss": 0.5007,
+      "step": 254
+    },
+    {
+      "epoch": 0.3975058456742011,
+      "grad_norm": 0.529695325269085,
+      "learning_rate": 4.8208092485549136e-05,
+      "loss": 0.4928,
+      "step": 255
+    },
+    {
+      "epoch": 0.3990646921278254,
+      "grad_norm": 0.46950723041884723,
+      "learning_rate": 4.817919075144509e-05,
+      "loss": 0.4637,
+      "step": 256
+    },
+    {
+      "epoch": 0.4006235385814497,
+      "grad_norm": 0.5659006167244565,
+      "learning_rate": 4.815028901734104e-05,
+      "loss": 0.4719,
+      "step": 257
+    },
+    {
+      "epoch": 0.40218238503507403,
+      "grad_norm": 0.39896571715761564,
+      "learning_rate": 4.8121387283237e-05,
+      "loss": 0.4847,
+      "step": 258
+    },
+    {
+      "epoch": 0.40374123148869834,
+      "grad_norm": 0.5485795723559732,
+      "learning_rate": 4.809248554913295e-05,
+      "loss": 0.4782,
+      "step": 259
+    },
+    {
+      "epoch": 0.4053000779423227,
+      "grad_norm": 0.3776722335922268,
+      "learning_rate": 4.8063583815028904e-05,
+      "loss": 0.4728,
+      "step": 260
+    },
+    {
+      "epoch": 0.406858924395947,
+      "grad_norm": 0.42769770440042826,
+      "learning_rate": 4.803468208092486e-05,
+      "loss": 0.4718,
+      "step": 261
+    },
+    {
+      "epoch": 0.4084177708495713,
+      "grad_norm": 0.38140213686869234,
+      "learning_rate": 4.800578034682081e-05,
+      "loss": 0.4798,
+      "step": 262
+    },
+    {
+      "epoch": 0.4099766173031956,
+      "grad_norm": 0.3693912240886004,
+      "learning_rate": 4.7976878612716764e-05,
+      "loss": 0.4677,
+      "step": 263
+    },
+    {
+      "epoch": 0.41153546375681993,
+      "grad_norm": 0.540507172297762,
+      "learning_rate": 4.7947976878612724e-05,
+      "loss": 0.4783,
+      "step": 264
+    },
+    {
+      "epoch": 0.4130943102104443,
+      "grad_norm": 0.38027657459985387,
+      "learning_rate": 4.791907514450867e-05,
+      "loss": 0.4977,
+      "step": 265
+    },
+    {
+      "epoch": 0.4146531566640686,
+      "grad_norm": 0.5498501393924234,
+      "learning_rate": 4.7890173410404624e-05,
+      "loss": 0.5207,
+      "step": 266
+    },
+    {
+      "epoch": 0.4162120031176929,
+      "grad_norm": 0.44454634948662736,
+      "learning_rate": 4.786127167630058e-05,
+      "loss": 0.449,
+      "step": 267
+    },
+    {
+      "epoch": 0.4177708495713172,
+      "grad_norm": 0.5112086728008995,
+      "learning_rate": 4.783236994219654e-05,
+      "loss": 0.4803,
+      "step": 268
+    },
+    {
+      "epoch": 0.4193296960249415,
+      "grad_norm": 0.6274330925404005,
+      "learning_rate": 4.780346820809249e-05,
+      "loss": 0.477,
+      "step": 269
+    },
+    {
+      "epoch": 0.4208885424785659,
+      "grad_norm": 0.45606522557073254,
+      "learning_rate": 4.777456647398844e-05,
+      "loss": 0.4794,
+      "step": 270
+    },
+    {
+      "epoch": 0.4224473889321902,
+      "grad_norm": 0.5630490443389089,
+      "learning_rate": 4.774566473988439e-05,
+      "loss": 0.48,
+      "step": 271
+    },
+    {
+      "epoch": 0.4240062353858145,
+      "grad_norm": 0.5538678824248618,
+      "learning_rate": 4.771676300578035e-05,
+      "loss": 0.4972,
+      "step": 272
+    },
+    {
+      "epoch": 0.4255650818394388,
+      "grad_norm": 0.5075041955013312,
+      "learning_rate": 4.7687861271676305e-05,
+      "loss": 0.4935,
+      "step": 273
+    },
+    {
+      "epoch": 0.4271239282930631,
+      "grad_norm": 0.5416003321516639,
+      "learning_rate": 4.765895953757226e-05,
+      "loss": 0.4732,
+      "step": 274
+    },
+    {
+      "epoch": 0.4286827747466875,
+      "grad_norm": 0.4993962101559131,
+      "learning_rate": 4.763005780346821e-05,
+      "loss": 0.477,
+      "step": 275
+    },
+    {
+      "epoch": 0.4302416212003118,
+      "grad_norm": 0.5254171958882733,
+      "learning_rate": 4.7601156069364166e-05,
+      "loss": 0.4899,
+      "step": 276
+    },
+    {
+      "epoch": 0.4318004676539361,
+      "grad_norm": 0.5484868681986969,
+      "learning_rate": 4.757225433526012e-05,
+      "loss": 0.4608,
+      "step": 277
+    },
+    {
+      "epoch": 0.4333593141075604,
+      "grad_norm": 0.6405428710095884,
+      "learning_rate": 4.754335260115607e-05,
+      "loss": 0.4816,
+      "step": 278
+    },
+    {
+      "epoch": 0.4349181605611847,
+      "grad_norm": 0.48377922320779415,
+      "learning_rate": 4.7514450867052026e-05,
+      "loss": 0.4838,
+      "step": 279
+    },
+    {
+      "epoch": 0.436477007014809,
+      "grad_norm": 0.7085436078292358,
+      "learning_rate": 4.748554913294798e-05,
+      "loss": 0.4718,
+      "step": 280
+    },
+    {
+      "epoch": 0.4380358534684334,
+      "grad_norm": 0.4043009368444505,
+      "learning_rate": 4.745664739884393e-05,
+      "loss": 0.4442,
+      "step": 281
+    },
+    {
+      "epoch": 0.4395946999220577,
+      "grad_norm": 0.5866137208959042,
+      "learning_rate": 4.7427745664739886e-05,
+      "loss": 0.461,
+      "step": 282
+    },
+    {
+      "epoch": 0.441153546375682,
+      "grad_norm": 0.5810371524784574,
+      "learning_rate": 4.739884393063584e-05,
+      "loss": 0.4839,
+      "step": 283
+    },
+    {
+      "epoch": 0.4427123928293063,
+      "grad_norm": 0.438670614683413,
+      "learning_rate": 4.73699421965318e-05,
+      "loss": 0.4615,
+      "step": 284
+    },
+    {
+      "epoch": 0.4442712392829306,
+      "grad_norm": 0.6406824783627434,
+      "learning_rate": 4.7341040462427747e-05,
+      "loss": 0.4611,
+      "step": 285
+    },
+    {
+      "epoch": 0.44583008573655497,
+      "grad_norm": 0.529477326835477,
+      "learning_rate": 4.73121387283237e-05,
+      "loss": 0.4788,
+      "step": 286
+    },
+    {
+      "epoch": 0.4473889321901793,
+      "grad_norm": 0.4838077152970437,
+      "learning_rate": 4.7283236994219653e-05,
+      "loss": 0.4727,
+      "step": 287
+    },
+    {
+      "epoch": 0.4489477786438036,
+      "grad_norm": 0.47119490593405255,
+      "learning_rate": 4.7254335260115614e-05,
+      "loss": 0.473,
+      "step": 288
+    },
+    {
+      "epoch": 0.4505066250974279,
+      "grad_norm": 0.4366369786625669,
+      "learning_rate": 4.722543352601157e-05,
+      "loss": 0.4526,
+      "step": 289
+    },
+    {
+      "epoch": 0.4520654715510522,
+      "grad_norm": 0.4281616340000569,
+      "learning_rate": 4.7196531791907514e-05,
+      "loss": 0.482,
+      "step": 290
+    },
+    {
+      "epoch": 0.45362431800467656,
+      "grad_norm": 0.5820798503771535,
+      "learning_rate": 4.716763005780347e-05,
+      "loss": 0.4716,
+      "step": 291
+    },
+    {
+      "epoch": 0.45518316445830087,
+      "grad_norm": 0.45736556696474184,
+      "learning_rate": 4.713872832369942e-05,
+      "loss": 0.4569,
+      "step": 292
+    },
+    {
+      "epoch": 0.4567420109119252,
+      "grad_norm": 0.5568676859423672,
+      "learning_rate": 4.710982658959538e-05,
+      "loss": 0.4738,
+      "step": 293
+    },
+    {
+      "epoch": 0.4583008573655495,
+      "grad_norm": 0.6028680330451817,
+      "learning_rate": 4.7080924855491334e-05,
+      "loss": 0.476,
+      "step": 294
+    },
+    {
+      "epoch": 0.4598597038191738,
+      "grad_norm": 0.47393834722502826,
+      "learning_rate": 4.705202312138728e-05,
+      "loss": 0.4896,
+      "step": 295
+    },
+    {
+      "epoch": 0.46141855027279816,
+      "grad_norm": 0.5313704646163404,
+      "learning_rate": 4.7023121387283234e-05,
+      "loss": 0.4805,
+      "step": 296
+    },
+    {
+      "epoch": 0.46297739672642246,
+      "grad_norm": 0.5051367616146644,
+      "learning_rate": 4.6994219653179195e-05,
+      "loss": 0.4569,
+      "step": 297
+    },
+    {
+      "epoch": 0.46453624318004677,
+      "grad_norm": 0.3831029585767748,
+      "learning_rate": 4.696531791907515e-05,
+      "loss": 0.4744,
+      "step": 298
+    },
+    {
+      "epoch": 0.4660950896336711,
+      "grad_norm": 0.521461479028099,
+      "learning_rate": 4.69364161849711e-05,
+      "loss": 0.4483,
+      "step": 299
+    },
+    {
+      "epoch": 0.4676539360872954,
+      "grad_norm": 0.38896958455232095,
+      "learning_rate": 4.690751445086705e-05,
+      "loss": 0.4498,
+      "step": 300
+    },
+    {
+      "epoch": 0.4692127825409197,
+      "grad_norm": 0.39085029504274005,
+      "learning_rate": 4.687861271676301e-05,
+      "loss": 0.4671,
+      "step": 301
+    },
+    {
+      "epoch": 0.47077162899454406,
+      "grad_norm": 0.4788937605695569,
+      "learning_rate": 4.684971098265896e-05,
+      "loss": 0.476,
+      "step": 302
+    },
+    {
+      "epoch": 0.47233047544816836,
+      "grad_norm": 0.34601184680325847,
+      "learning_rate": 4.6820809248554915e-05,
+      "loss": 0.4638,
+      "step": 303
+    },
+    {
+      "epoch": 0.47388932190179267,
+      "grad_norm": 0.43700546489204795,
+      "learning_rate": 4.679190751445087e-05,
+      "loss": 0.4814,
+      "step": 304
+    },
+    {
+      "epoch": 0.475448168355417,
+      "grad_norm": 0.3838353985965739,
+      "learning_rate": 4.676300578034682e-05,
+      "loss": 0.451,
+      "step": 305
+    },
+    {
+      "epoch": 0.4770070148090413,
+      "grad_norm": 0.4177639814145164,
+      "learning_rate": 4.6734104046242776e-05,
+      "loss": 0.4758,
+      "step": 306
+    },
+    {
+      "epoch": 0.47856586126266565,
+      "grad_norm": 0.3922398233693093,
+      "learning_rate": 4.670520231213873e-05,
+      "loss": 0.4657,
+      "step": 307
+    },
+    {
+      "epoch": 0.48012470771628996,
+      "grad_norm": 0.3757724195621053,
+      "learning_rate": 4.667630057803468e-05,
+      "loss": 0.4793,
+      "step": 308
+    },
+    {
+      "epoch": 0.48168355416991426,
+      "grad_norm": 0.3414831861075698,
+      "learning_rate": 4.664739884393064e-05,
+      "loss": 0.4636,
+      "step": 309
+    },
+    {
+      "epoch": 0.48324240062353857,
+      "grad_norm": 0.41270011789282307,
+      "learning_rate": 4.661849710982659e-05,
+      "loss": 0.4508,
+      "step": 310
+    },
+    {
+      "epoch": 0.4848012470771629,
+      "grad_norm": 0.40583792569372235,
+      "learning_rate": 4.658959537572254e-05,
+      "loss": 0.4849,
+      "step": 311
+    },
+    {
+      "epoch": 0.48636009353078724,
+      "grad_norm": 0.44731637705198873,
+      "learning_rate": 4.6560693641618496e-05,
+      "loss": 0.4956,
+      "step": 312
+    },
+    {
+      "epoch": 0.48791893998441155,
+      "grad_norm": 0.39572009537399316,
+      "learning_rate": 4.653179190751446e-05,
+      "loss": 0.4752,
+      "step": 313
+    },
+    {
+      "epoch": 0.48947778643803586,
+      "grad_norm": 0.39239985837197133,
+      "learning_rate": 4.650289017341041e-05,
+      "loss": 0.4448,
+      "step": 314
+    },
+    {
+      "epoch": 0.49103663289166016,
+      "grad_norm": 0.47254520649722037,
+      "learning_rate": 4.647398843930636e-05,
+      "loss": 0.4424,
+      "step": 315
+    },
+    {
+      "epoch": 0.49259547934528447,
+      "grad_norm": 0.3655766693795544,
+      "learning_rate": 4.644508670520231e-05,
+      "loss": 0.4585,
+      "step": 316
+    },
+    {
+      "epoch": 0.49415432579890883,
+      "grad_norm": 0.5024262207840765,
+      "learning_rate": 4.641618497109827e-05,
+      "loss": 0.4711,
+      "step": 317
+    },
+    {
+      "epoch": 0.49571317225253314,
+      "grad_norm": 0.41554970235113214,
+      "learning_rate": 4.6387283236994224e-05,
+      "loss": 0.4606,
+      "step": 318
+    },
+    {
+      "epoch": 0.49727201870615745,
+      "grad_norm": 0.4597741099913984,
+      "learning_rate": 4.635838150289018e-05,
+      "loss": 0.4527,
+      "step": 319
+    },
+    {
+      "epoch": 0.49883086515978176,
+      "grad_norm": 0.4425235361753541,
+      "learning_rate": 4.6329479768786124e-05,
+      "loss": 0.4769,
+      "step": 320
+    },
+    {
+      "epoch": 0.5003897116134061,
+      "grad_norm": 0.38187283401074307,
+      "learning_rate": 4.6300578034682084e-05,
+      "loss": 0.4594,
+      "step": 321
+    },
+    {
+      "epoch": 0.5019485580670304,
+      "grad_norm": 0.3797910266033653,
+      "learning_rate": 4.627167630057804e-05,
+      "loss": 0.462,
+      "step": 322
+    },
+    {
+      "epoch": 0.5035074045206547,
+      "grad_norm": 0.39511922436954366,
+      "learning_rate": 4.624277456647399e-05,
+      "loss": 0.4703,
+      "step": 323
+    },
+    {
+      "epoch": 0.505066250974279,
+      "grad_norm": 0.4061473825369823,
+      "learning_rate": 4.6213872832369945e-05,
+      "loss": 0.4497,
+      "step": 324
+    },
+    {
+      "epoch": 0.5066250974279034,
+      "grad_norm": 0.36187280391895393,
+      "learning_rate": 4.61849710982659e-05,
+      "loss": 0.4536,
+      "step": 325
+    },
+    {
+      "epoch": 0.5081839438815277,
+      "grad_norm": 0.4400520460188613,
+      "learning_rate": 4.615606936416185e-05,
+      "loss": 0.4608,
+      "step": 326
+    },
+    {
+      "epoch": 0.509742790335152,
+      "grad_norm": 0.3663575693429021,
+      "learning_rate": 4.6127167630057805e-05,
+      "loss": 0.4626,
+      "step": 327
+    },
+    {
+      "epoch": 0.5113016367887763,
+      "grad_norm": 0.4766184153960843,
+      "learning_rate": 4.609826589595376e-05,
+      "loss": 0.4773,
+      "step": 328
+    },
+    {
+      "epoch": 0.5128604832424006,
+      "grad_norm": 0.45206345178046475,
+      "learning_rate": 4.606936416184972e-05,
+      "loss": 0.4469,
+      "step": 329
+    },
+    {
+      "epoch": 0.514419329696025,
+      "grad_norm": 0.39026528576069053,
+      "learning_rate": 4.6040462427745665e-05,
+      "loss": 0.4604,
+      "step": 330
+    },
+    {
+      "epoch": 0.5159781761496492,
+      "grad_norm": 0.5361204288029048,
+      "learning_rate": 4.601156069364162e-05,
+      "loss": 0.4613,
+      "step": 331
+    },
+    {
+      "epoch": 0.5175370226032736,
+      "grad_norm": 0.43012235247767305,
+      "learning_rate": 4.598265895953757e-05,
+      "loss": 0.4579,
+      "step": 332
+    },
+    {
+      "epoch": 0.5190958690568979,
+      "grad_norm": 0.527243694253446,
+      "learning_rate": 4.595375722543353e-05,
+      "loss": 0.4602,
+      "step": 333
+    },
+    {
+      "epoch": 0.5206547155105222,
+      "grad_norm": 0.4179764821236328,
+      "learning_rate": 4.5924855491329486e-05,
+      "loss": 0.4585,
+      "step": 334
+    },
+    {
+      "epoch": 0.5222135619641465,
+      "grad_norm": 0.4510997406406127,
+      "learning_rate": 4.589595375722543e-05,
+      "loss": 0.4636,
+      "step": 335
+    },
+    {
+      "epoch": 0.5237724084177708,
+      "grad_norm": 0.3187587883976425,
+      "learning_rate": 4.5867052023121386e-05,
+      "loss": 0.4612,
+      "step": 336
+    },
+    {
+      "epoch": 0.5253312548713952,
+      "grad_norm": 0.46286631819853313,
+      "learning_rate": 4.5838150289017346e-05,
+      "loss": 0.4455,
+      "step": 337
+    },
+    {
+      "epoch": 0.5268901013250195,
+      "grad_norm": 0.43393995615442665,
+      "learning_rate": 4.58092485549133e-05,
+      "loss": 0.481,
+      "step": 338
+    },
+    {
+      "epoch": 0.5284489477786438,
+      "grad_norm": 0.5297318671699833,
+      "learning_rate": 4.578034682080925e-05,
+      "loss": 0.464,
+      "step": 339
+    },
+    {
+      "epoch": 0.5300077942322681,
+      "grad_norm": 0.42416355410040585,
+      "learning_rate": 4.57514450867052e-05,
+      "loss": 0.4644,
+      "step": 340
+    },
+    {
+      "epoch": 0.5315666406858924,
+      "grad_norm": 0.3712827662390051,
+      "learning_rate": 4.572254335260116e-05,
+      "loss": 0.4675,
+      "step": 341
+    },
+    {
+      "epoch": 0.5331254871395168,
+      "grad_norm": 0.39762495205102294,
+      "learning_rate": 4.569364161849711e-05,
+      "loss": 0.469,
+      "step": 342
+    },
+    {
+      "epoch": 0.534684333593141,
+      "grad_norm": 0.3254212965910093,
+      "learning_rate": 4.566473988439307e-05,
+      "loss": 0.4534,
+      "step": 343
+    },
+    {
+      "epoch": 0.5362431800467654,
+      "grad_norm": 0.4848016190354553,
+      "learning_rate": 4.563583815028902e-05,
+      "loss": 0.4616,
+      "step": 344
+    },
+    {
+      "epoch": 0.5378020265003897,
+      "grad_norm": 0.31661190588364,
+      "learning_rate": 4.5606936416184974e-05,
+      "loss": 0.4568,
+      "step": 345
+    },
+    {
+      "epoch": 0.539360872954014,
+      "grad_norm": 0.4676445060911043,
+      "learning_rate": 4.557803468208093e-05,
+      "loss": 0.472,
+      "step": 346
+    },
+    {
+      "epoch": 0.5409197194076384,
+      "grad_norm": 0.3160669772415641,
+      "learning_rate": 4.554913294797688e-05,
+      "loss": 0.444,
+      "step": 347
+    },
+    {
+      "epoch": 0.5424785658612626,
+      "grad_norm": 0.4323013049552334,
+      "learning_rate": 4.5520231213872834e-05,
+      "loss": 0.4392,
+      "step": 348
+    },
+    {
+      "epoch": 0.544037412314887,
+      "grad_norm": 0.4523674768546962,
+      "learning_rate": 4.549132947976879e-05,
+      "loss": 0.4682,
+      "step": 349
+    },
+    {
+      "epoch": 0.5455962587685113,
+      "grad_norm": 0.3575280341982122,
+      "learning_rate": 4.546242774566474e-05,
+      "loss": 0.4483,
+      "step": 350
+    },
+    {
+      "epoch": 0.5471551052221356,
+      "grad_norm": 0.4418729569254676,
+      "learning_rate": 4.5433526011560694e-05,
+      "loss": 0.4489,
+      "step": 351
+    },
+    {
+      "epoch": 0.54871395167576,
+      "grad_norm": 0.3280270922434657,
+      "learning_rate": 4.540462427745665e-05,
+      "loss": 0.4616,
+      "step": 352
+    },
+    {
+      "epoch": 0.5502727981293842,
+      "grad_norm": 0.4873805692806027,
+      "learning_rate": 4.537572254335261e-05,
+      "loss": 0.4802,
+      "step": 353
+    },
+    {
+      "epoch": 0.5518316445830086,
+      "grad_norm": 0.36573476600102434,
+      "learning_rate": 4.5346820809248555e-05,
+      "loss": 0.475,
+      "step": 354
+    },
+    {
+      "epoch": 0.5533904910366328,
+      "grad_norm": 0.41050124588561904,
+      "learning_rate": 4.531791907514451e-05,
+      "loss": 0.4535,
+      "step": 355
+    },
+    {
+      "epoch": 0.5549493374902572,
+      "grad_norm": 0.3620545249321528,
+      "learning_rate": 4.528901734104046e-05,
+      "loss": 0.4428,
+      "step": 356
+    },
+    {
+      "epoch": 0.5565081839438816,
+      "grad_norm": 0.467231893264037,
+      "learning_rate": 4.526011560693642e-05,
+      "loss": 0.4566,
+      "step": 357
+    },
+    {
+      "epoch": 0.5580670303975058,
+      "grad_norm": 0.4030396125824727,
+      "learning_rate": 4.5231213872832375e-05,
+      "loss": 0.4717,
+      "step": 358
+    },
+    {
+      "epoch": 0.5596258768511302,
+      "grad_norm": 0.34726035881210865,
+      "learning_rate": 4.520231213872833e-05,
+      "loss": 0.4497,
+      "step": 359
+    },
+    {
+      "epoch": 0.5611847233047544,
+      "grad_norm": 0.39289264377072153,
+      "learning_rate": 4.5173410404624275e-05,
+      "loss": 0.4318,
+      "step": 360
+    },
+    {
+      "epoch": 0.5627435697583788,
+      "grad_norm": 0.3890869555190299,
+      "learning_rate": 4.5144508670520236e-05,
+      "loss": 0.4569,
+      "step": 361
+    },
+    {
+      "epoch": 0.5643024162120032,
+      "grad_norm": 0.39781174379277523,
+      "learning_rate": 4.511560693641619e-05,
+      "loss": 0.4638,
+      "step": 362
+    },
+    {
+      "epoch": 0.5658612626656274,
+      "grad_norm": 0.36372440090872343,
+      "learning_rate": 4.508670520231214e-05,
+      "loss": 0.441,
+      "step": 363
+    },
+    {
+      "epoch": 0.5674201091192518,
+      "grad_norm": 0.4284734591617738,
+      "learning_rate": 4.5057803468208096e-05,
+      "loss": 0.4745,
+      "step": 364
+    },
+    {
+      "epoch": 0.568978955572876,
+      "grad_norm": 0.3838382539954402,
+      "learning_rate": 4.502890173410405e-05,
+      "loss": 0.4538,
+      "step": 365
+    },
+    {
+      "epoch": 0.5705378020265004,
+      "grad_norm": 0.4837580518409696,
+      "learning_rate": 4.5e-05,
+      "loss": 0.4291,
+      "step": 366
+    },
+    {
+      "epoch": 0.5720966484801248,
+      "grad_norm": 0.3807661371797664,
+      "learning_rate": 4.4971098265895956e-05,
+      "loss": 0.4681,
+      "step": 367
+    },
+    {
+      "epoch": 0.573655494933749,
+      "grad_norm": 0.4409972991204447,
+      "learning_rate": 4.494219653179191e-05,
+      "loss": 0.4425,
+      "step": 368
+    },
+    {
+      "epoch": 0.5752143413873734,
+      "grad_norm": 0.371972546132041,
+      "learning_rate": 4.491329479768786e-05,
+      "loss": 0.4543,
+      "step": 369
+    },
+    {
+      "epoch": 0.5767731878409976,
+      "grad_norm": 0.37683721417276733,
+      "learning_rate": 4.488439306358382e-05,
+      "loss": 0.4575,
+      "step": 370
+    },
+    {
+      "epoch": 0.578332034294622,
+      "grad_norm": 0.45855850398477666,
+      "learning_rate": 4.485549132947977e-05,
+      "loss": 0.4847,
+      "step": 371
+    },
+    {
+      "epoch": 0.5798908807482464,
+      "grad_norm": 0.4223420944336656,
+      "learning_rate": 4.4826589595375724e-05,
+      "loss": 0.4649,
+      "step": 372
+    },
+    {
+      "epoch": 0.5814497272018706,
+      "grad_norm": 0.4597887227283376,
+      "learning_rate": 4.4797687861271684e-05,
+      "loss": 0.4828,
+      "step": 373
+    },
+    {
+      "epoch": 0.583008573655495,
+      "grad_norm": 0.4998303192690309,
+      "learning_rate": 4.476878612716763e-05,
+      "loss": 0.452,
+      "step": 374
+    },
+    {
+      "epoch": 0.5845674201091192,
+      "grad_norm": 0.3756104589899762,
+      "learning_rate": 4.4739884393063584e-05,
+      "loss": 0.4411,
+      "step": 375
+    },
+    {
+      "epoch": 0.5861262665627436,
+      "grad_norm": 0.4224811548039192,
+      "learning_rate": 4.471098265895954e-05,
+      "loss": 0.4564,
+      "step": 376
+    },
+    {
+      "epoch": 0.5876851130163678,
+      "grad_norm": 0.4569199095589889,
+      "learning_rate": 4.46820809248555e-05,
+      "loss": 0.4664,
+      "step": 377
+    },
+    {
+      "epoch": 0.5892439594699922,
+      "grad_norm": 0.3608347158374211,
+      "learning_rate": 4.465317919075145e-05,
+      "loss": 0.4734,
+      "step": 378
+    },
+    {
+      "epoch": 0.5908028059236166,
+      "grad_norm": 0.4552865050320062,
+      "learning_rate": 4.46242774566474e-05,
+      "loss": 0.4723,
+      "step": 379
+    },
+    {
+      "epoch": 0.5923616523772408,
+      "grad_norm": 0.3539125767256611,
+      "learning_rate": 4.459537572254335e-05,
+      "loss": 0.4762,
+      "step": 380
+    },
+    {
+      "epoch": 0.5939204988308652,
+      "grad_norm": 0.3880352014434992,
+      "learning_rate": 4.456647398843931e-05,
+      "loss": 0.4388,
+      "step": 381
+    },
+    {
+      "epoch": 0.5954793452844894,
+      "grad_norm": 0.37410095368321145,
+      "learning_rate": 4.4537572254335265e-05,
+      "loss": 0.4533,
+      "step": 382
+    },
+    {
+      "epoch": 0.5970381917381138,
+      "grad_norm": 0.3133270680014226,
+      "learning_rate": 4.450867052023122e-05,
+      "loss": 0.4525,
+      "step": 383
+    },
+    {
+      "epoch": 0.5985970381917382,
+      "grad_norm": 0.41920980039661443,
+      "learning_rate": 4.447976878612717e-05,
+      "loss": 0.4333,
+      "step": 384
+    },
+    {
+      "epoch": 0.6001558846453624,
+      "grad_norm": 0.3774275243895145,
+      "learning_rate": 4.4450867052023125e-05,
+      "loss": 0.4459,
+      "step": 385
+    },
+    {
+      "epoch": 0.6017147310989868,
+      "grad_norm": 0.369753586583569,
+      "learning_rate": 4.442196531791908e-05,
+      "loss": 0.4628,
+      "step": 386
+    },
+    {
+      "epoch": 0.603273577552611,
+      "grad_norm": 0.43933813534803284,
+      "learning_rate": 4.439306358381503e-05,
+      "loss": 0.4866,
+      "step": 387
+    },
+    {
+      "epoch": 0.6048324240062354,
+      "grad_norm": 0.36877162976354194,
+      "learning_rate": 4.4364161849710985e-05,
+      "loss": 0.4489,
+      "step": 388
+    },
+    {
+      "epoch": 0.6063912704598597,
+      "grad_norm": 0.36797218493960576,
+      "learning_rate": 4.433526011560694e-05,
+      "loss": 0.4681,
+      "step": 389
+    },
+    {
+      "epoch": 0.607950116913484,
+      "grad_norm": 0.3598201801544274,
+      "learning_rate": 4.430635838150289e-05,
+      "loss": 0.4451,
+      "step": 390
+    },
+    {
+      "epoch": 0.6095089633671084,
+      "grad_norm": 0.37415526691302536,
+      "learning_rate": 4.4277456647398846e-05,
+      "loss": 0.4649,
+      "step": 391
+    },
+    {
+      "epoch": 0.6110678098207326,
+      "grad_norm": 0.3217861980835781,
+      "learning_rate": 4.42485549132948e-05,
+      "loss": 0.4579,
+      "step": 392
+    },
+    {
+      "epoch": 0.612626656274357,
+      "grad_norm": 0.4455765725783138,
+      "learning_rate": 4.421965317919075e-05,
+      "loss": 0.4603,
+      "step": 393
+    },
+    {
+      "epoch": 0.6141855027279813,
+      "grad_norm": 0.37219414271910367,
+      "learning_rate": 4.4190751445086706e-05,
+      "loss": 0.4548,
+      "step": 394
+    },
+    {
+      "epoch": 0.6157443491816056,
+      "grad_norm": 0.45080978894718876,
+      "learning_rate": 4.416184971098266e-05,
+      "loss": 0.4504,
+      "step": 395
+    },
+    {
+      "epoch": 0.61730319563523,
+      "grad_norm": 0.4358441423072049,
+      "learning_rate": 4.413294797687861e-05,
+      "loss": 0.4446,
+      "step": 396
+    },
+    {
+      "epoch": 0.6188620420888542,
+      "grad_norm": 0.37738329645269036,
+      "learning_rate": 4.4104046242774566e-05,
+      "loss": 0.4436,
+      "step": 397
+    },
+    {
+      "epoch": 0.6204208885424786,
+      "grad_norm": 0.4204646254215696,
+      "learning_rate": 4.407514450867053e-05,
+      "loss": 0.4512,
+      "step": 398
+    },
+    {
+      "epoch": 0.6219797349961029,
+      "grad_norm": 0.39901225883937147,
+      "learning_rate": 4.404624277456647e-05,
+      "loss": 0.463,
+      "step": 399
+    },
+    {
+      "epoch": 0.6235385814497272,
+      "grad_norm": 0.320265409159274,
+      "learning_rate": 4.401734104046243e-05,
+      "loss": 0.4329,
+      "step": 400
+    },
+    {
+      "epoch": 0.6250974279033515,
+      "grad_norm": 0.49013995892880263,
+      "learning_rate": 4.398843930635838e-05,
+      "loss": 0.4542,
+      "step": 401
+    },
+    {
+      "epoch": 0.6266562743569758,
+      "grad_norm": 0.3510400983584625,
+      "learning_rate": 4.395953757225434e-05,
+      "loss": 0.4589,
+      "step": 402
+    },
+    {
+      "epoch": 0.6282151208106002,
+      "grad_norm": 0.4402933914634823,
+      "learning_rate": 4.3930635838150294e-05,
+      "loss": 0.4593,
+      "step": 403
+    },
+    {
+      "epoch": 0.6297739672642245,
+      "grad_norm": 0.47377011043879347,
+      "learning_rate": 4.390173410404624e-05,
+      "loss": 0.4675,
+      "step": 404
+    },
+    {
+      "epoch": 0.6313328137178488,
+      "grad_norm": 0.37916907989218646,
+      "learning_rate": 4.3872832369942194e-05,
+      "loss": 0.4412,
+      "step": 405
+    },
+    {
+      "epoch": 0.6328916601714731,
+      "grad_norm": 0.35621367384404673,
+      "learning_rate": 4.3843930635838154e-05,
+      "loss": 0.4618,
+      "step": 406
+    },
+    {
+      "epoch": 0.6344505066250974,
+      "grad_norm": 0.3834554558658467,
+      "learning_rate": 4.381502890173411e-05,
+      "loss": 0.4591,
+      "step": 407
+    },
+    {
+      "epoch": 0.6360093530787218,
+      "grad_norm": 0.3273576509892712,
+      "learning_rate": 4.378612716763006e-05,
+      "loss": 0.4307,
+      "step": 408
+    },
+    {
+      "epoch": 0.6375681995323461,
+      "grad_norm": 0.3260428886276838,
+      "learning_rate": 4.3757225433526015e-05,
+      "loss": 0.4315,
+      "step": 409
+    },
+    {
+      "epoch": 0.6391270459859704,
+      "grad_norm": 0.3828677152566117,
+      "learning_rate": 4.372832369942197e-05,
+      "loss": 0.431,
+      "step": 410
+    },
+    {
+      "epoch": 0.6406858924395947,
+      "grad_norm": 0.3428782387615178,
+      "learning_rate": 4.369942196531792e-05,
+      "loss": 0.4535,
+      "step": 411
+    },
+    {
+      "epoch": 0.642244738893219,
+      "grad_norm": 0.3568689766352524,
+      "learning_rate": 4.3670520231213875e-05,
+      "loss": 0.4471,
+      "step": 412
+    },
+    {
+      "epoch": 0.6438035853468433,
+      "grad_norm": 0.3200198497931682,
+      "learning_rate": 4.364161849710983e-05,
+      "loss": 0.4675,
+      "step": 413
+    },
+    {
+      "epoch": 0.6453624318004677,
+      "grad_norm": 0.3439904016830579,
+      "learning_rate": 4.361271676300578e-05,
+      "loss": 0.4434,
+      "step": 414
+    },
+    {
+      "epoch": 0.646921278254092,
+      "grad_norm": 0.3910285780339108,
+      "learning_rate": 4.3583815028901735e-05,
+      "loss": 0.4291,
+      "step": 415
+    },
+    {
+      "epoch": 0.6484801247077163,
+      "grad_norm": 0.35633702898935654,
+      "learning_rate": 4.355491329479769e-05,
+      "loss": 0.4465,
+      "step": 416
+    },
+    {
+      "epoch": 0.6500389711613406,
+      "grad_norm": 0.4280187254199856,
+      "learning_rate": 4.352601156069364e-05,
+      "loss": 0.4467,
+      "step": 417
+    },
+    {
+      "epoch": 0.6515978176149649,
+      "grad_norm": 0.3860941239745589,
+      "learning_rate": 4.34971098265896e-05,
+      "loss": 0.4443,
+      "step": 418
+    },
+    {
+      "epoch": 0.6531566640685893,
+      "grad_norm": 0.44940034037874554,
+      "learning_rate": 4.346820809248555e-05,
+      "loss": 0.48,
+      "step": 419
+    },
+    {
+      "epoch": 0.6547155105222136,
+      "grad_norm": 0.4625797028642205,
+      "learning_rate": 4.34393063583815e-05,
+      "loss": 0.4568,
+      "step": 420
+    },
+    {
+      "epoch": 0.6562743569758379,
+      "grad_norm": 0.36493835250776474,
+      "learning_rate": 4.3410404624277456e-05,
+      "loss": 0.4622,
+      "step": 421
+    },
+    {
+      "epoch": 0.6578332034294622,
+      "grad_norm": 0.4451390677090527,
+      "learning_rate": 4.3381502890173416e-05,
+      "loss": 0.4399,
+      "step": 422
+    },
+    {
+      "epoch": 0.6593920498830865,
+      "grad_norm": 0.40509517400153716,
+      "learning_rate": 4.335260115606937e-05,
+      "loss": 0.4548,
+      "step": 423
+    },
+    {
+      "epoch": 0.6609508963367108,
+      "grad_norm": 0.4745439842667765,
+      "learning_rate": 4.3323699421965316e-05,
+      "loss": 0.4409,
+      "step": 424
+    },
+    {
+      "epoch": 0.6625097427903351,
+      "grad_norm": 0.40589601301353223,
+      "learning_rate": 4.329479768786127e-05,
+      "loss": 0.4386,
+      "step": 425
+    },
+    {
+      "epoch": 0.6640685892439595,
+      "grad_norm": 0.410367648733596,
+      "learning_rate": 4.326589595375723e-05,
+      "loss": 0.4359,
+      "step": 426
+    },
+    {
+      "epoch": 0.6656274356975838,
+      "grad_norm": 0.39329540338329594,
+      "learning_rate": 4.3236994219653183e-05,
+      "loss": 0.4503,
+      "step": 427
+    },
+    {
+      "epoch": 0.6671862821512081,
+      "grad_norm": 0.4110086370635056,
+      "learning_rate": 4.320809248554914e-05,
+      "loss": 0.4447,
+      "step": 428
+    },
+    {
+      "epoch": 0.6687451286048324,
+      "grad_norm": 0.47772877851625145,
+      "learning_rate": 4.3179190751445084e-05,
+      "loss": 0.4687,
+      "step": 429
+    },
+    {
+      "epoch": 0.6703039750584567,
+      "grad_norm": 0.3866102449562328,
+      "learning_rate": 4.3150289017341044e-05,
+      "loss": 0.4393,
+      "step": 430
+    },
+    {
+      "epoch": 0.6718628215120811,
+      "grad_norm": 0.45739240597886627,
+      "learning_rate": 4.3121387283237e-05,
+      "loss": 0.4668,
+      "step": 431
+    },
+    {
+      "epoch": 0.6734216679657054,
+      "grad_norm": 0.4314476402559539,
+      "learning_rate": 4.309248554913295e-05,
+      "loss": 0.4649,
+      "step": 432
+    },
+    {
+      "epoch": 0.6749805144193297,
+      "grad_norm": 0.45634509121859473,
+      "learning_rate": 4.3063583815028904e-05,
+      "loss": 0.4441,
+      "step": 433
+    },
+    {
+      "epoch": 0.676539360872954,
+      "grad_norm": 0.4811305736485706,
+      "learning_rate": 4.303468208092486e-05,
+      "loss": 0.4529,
+      "step": 434
+    },
+    {
+      "epoch": 0.6780982073265783,
+      "grad_norm": 0.39207564033387,
+      "learning_rate": 4.300578034682081e-05,
+      "loss": 0.4588,
+      "step": 435
+    },
+    {
+      "epoch": 0.6796570537802027,
+      "grad_norm": 0.4658600522478725,
+      "learning_rate": 4.2976878612716764e-05,
+      "loss": 0.4379,
+      "step": 436
+    },
+    {
+      "epoch": 0.6812159002338269,
+      "grad_norm": 0.41119750867971183,
+      "learning_rate": 4.294797687861272e-05,
+      "loss": 0.4418,
+      "step": 437
+    },
+    {
+      "epoch": 0.6827747466874513,
+      "grad_norm": 0.4230535920977332,
+      "learning_rate": 4.291907514450868e-05,
+      "loss": 0.4536,
+      "step": 438
+    },
+    {
+      "epoch": 0.6843335931410756,
+      "grad_norm": 0.5092465838859024,
+      "learning_rate": 4.2890173410404625e-05,
+      "loss": 0.4584,
+      "step": 439
+    },
+    {
+      "epoch": 0.6858924395946999,
+      "grad_norm": 0.3601271310299026,
+      "learning_rate": 4.286127167630058e-05,
+      "loss": 0.4368,
+      "step": 440
+    },
+    {
+      "epoch": 0.6874512860483243,
+      "grad_norm": 0.5033625655714722,
+      "learning_rate": 4.283236994219653e-05,
+      "loss": 0.4484,
+      "step": 441
+    },
+    {
+      "epoch": 0.6890101325019485,
+      "grad_norm": 0.3994249544123872,
+      "learning_rate": 4.280346820809249e-05,
+      "loss": 0.4418,
+      "step": 442
+    },
+    {
+      "epoch": 0.6905689789555729,
+      "grad_norm": 0.45539241489062715,
+      "learning_rate": 4.2774566473988445e-05,
+      "loss": 0.4473,
+      "step": 443
+    },
+    {
+      "epoch": 0.6921278254091972,
+      "grad_norm": 0.4417751638358608,
+      "learning_rate": 4.274566473988439e-05,
+      "loss": 0.4496,
+      "step": 444
+    },
+    {
+      "epoch": 0.6936866718628215,
+      "grad_norm": 0.40882073328312935,
+      "learning_rate": 4.2716763005780345e-05,
+      "loss": 0.4622,
+      "step": 445
+    },
+    {
+      "epoch": 0.6952455183164459,
+      "grad_norm": 0.39444852000587594,
+      "learning_rate": 4.2687861271676306e-05,
+      "loss": 0.4769,
+      "step": 446
+    },
+    {
+      "epoch": 0.6968043647700701,
+      "grad_norm": 0.37669038630165647,
+      "learning_rate": 4.265895953757226e-05,
+      "loss": 0.4463,
+      "step": 447
+    },
+    {
+      "epoch": 0.6983632112236945,
+      "grad_norm": 0.4180560176644135,
+      "learning_rate": 4.263005780346821e-05,
+      "loss": 0.4502,
+      "step": 448
+    },
+    {
+      "epoch": 0.6999220576773187,
+      "grad_norm": 0.3606230739957135,
+      "learning_rate": 4.260115606936416e-05,
+      "loss": 0.4686,
+      "step": 449
+    },
+    {
+      "epoch": 0.7014809041309431,
+      "grad_norm": 0.4763884079373346,
+      "learning_rate": 4.257225433526012e-05,
+      "loss": 0.4465,
+      "step": 450
+    },
+    {
+      "epoch": 0.7030397505845675,
+      "grad_norm": 0.36030780349125946,
+      "learning_rate": 4.254335260115607e-05,
+      "loss": 0.4475,
+      "step": 451
+    },
+    {
+      "epoch": 0.7045985970381917,
+      "grad_norm": 0.4451176525342345,
+      "learning_rate": 4.2514450867052026e-05,
+      "loss": 0.4268,
+      "step": 452
+    },
+    {
+      "epoch": 0.7061574434918161,
+      "grad_norm": 0.33875598529236145,
+      "learning_rate": 4.248554913294798e-05,
+      "loss": 0.4499,
+      "step": 453
+    },
+    {
+      "epoch": 0.7077162899454403,
+      "grad_norm": 0.5071769434422747,
+      "learning_rate": 4.245664739884393e-05,
+      "loss": 0.4537,
+      "step": 454
+    },
+    {
+      "epoch": 0.7092751363990647,
+      "grad_norm": 0.3488438534196029,
+      "learning_rate": 4.242774566473989e-05,
+      "loss": 0.4623,
+      "step": 455
+    },
+    {
+      "epoch": 0.7108339828526891,
+      "grad_norm": 0.4454725104594618,
+      "learning_rate": 4.239884393063584e-05,
+      "loss": 0.4516,
+      "step": 456
+    },
+    {
+      "epoch": 0.7123928293063133,
+      "grad_norm": 0.3583891995078908,
+      "learning_rate": 4.2369942196531794e-05,
+      "loss": 0.4581,
+      "step": 457
+    },
+    {
+      "epoch": 0.7139516757599377,
+      "grad_norm": 0.36757937011684644,
+      "learning_rate": 4.234104046242775e-05,
+      "loss": 0.4505,
+      "step": 458
+    },
+    {
+      "epoch": 0.7155105222135619,
+      "grad_norm": 0.39437787323637835,
+      "learning_rate": 4.23121387283237e-05,
+      "loss": 0.4555,
+      "step": 459
+    },
+    {
+      "epoch": 0.7170693686671863,
+      "grad_norm": 0.3526260112578523,
+      "learning_rate": 4.2283236994219654e-05,
+      "loss": 0.4462,
+      "step": 460
+    },
+    {
+      "epoch": 0.7186282151208107,
+      "grad_norm": 0.3610517367016722,
+      "learning_rate": 4.225433526011561e-05,
+      "loss": 0.4439,
+      "step": 461
+    },
+    {
+      "epoch": 0.7201870615744349,
+      "grad_norm": 0.33046116660611863,
+      "learning_rate": 4.222543352601157e-05,
+      "loss": 0.4553,
+      "step": 462
+    },
+    {
+      "epoch": 0.7217459080280593,
+      "grad_norm": 0.3479382421778036,
+      "learning_rate": 4.2196531791907514e-05,
+      "loss": 0.4615,
+      "step": 463
+    },
+    {
+      "epoch": 0.7233047544816835,
+      "grad_norm": 0.3563546046165941,
+      "learning_rate": 4.216763005780347e-05,
+      "loss": 0.4502,
+      "step": 464
+    },
+    {
+      "epoch": 0.7248636009353079,
+      "grad_norm": 0.33104574079840804,
+      "learning_rate": 4.213872832369942e-05,
+      "loss": 0.4522,
+      "step": 465
+    },
+    {
+      "epoch": 0.7264224473889321,
+      "grad_norm": 0.3615689785424514,
+      "learning_rate": 4.210982658959538e-05,
+      "loss": 0.4456,
+      "step": 466
+    },
+    {
+      "epoch": 0.7279812938425565,
+      "grad_norm": 0.33245430385705754,
+      "learning_rate": 4.2080924855491335e-05,
+      "loss": 0.451,
+      "step": 467
+    },
+    {
+      "epoch": 0.7295401402961809,
+      "grad_norm": 0.31570139553177473,
+      "learning_rate": 4.205202312138729e-05,
+      "loss": 0.4387,
+      "step": 468
+    },
+    {
+      "epoch": 0.7310989867498051,
+      "grad_norm": 0.38612290372770497,
+      "learning_rate": 4.2023121387283235e-05,
+      "loss": 0.4391,
+      "step": 469
+    },
+    {
+      "epoch": 0.7326578332034295,
+      "grad_norm": 0.4573667515885595,
+      "learning_rate": 4.1994219653179195e-05,
+      "loss": 0.4689,
+      "step": 470
+    },
+    {
+      "epoch": 0.7342166796570537,
+      "grad_norm": 0.3807562323188079,
+      "learning_rate": 4.196531791907515e-05,
+      "loss": 0.4632,
+      "step": 471
+    },
+    {
+      "epoch": 0.7357755261106781,
+      "grad_norm": 0.3525392715340891,
+      "learning_rate": 4.19364161849711e-05,
+      "loss": 0.4381,
+      "step": 472
+    },
+    {
+      "epoch": 0.7373343725643025,
+      "grad_norm": 0.3782760064170116,
+      "learning_rate": 4.1907514450867055e-05,
+      "loss": 0.4477,
+      "step": 473
+    },
+    {
+      "epoch": 0.7388932190179267,
+      "grad_norm": 0.41115398176210943,
+      "learning_rate": 4.187861271676301e-05,
+      "loss": 0.4602,
+      "step": 474
+    },
+    {
+      "epoch": 0.7404520654715511,
+      "grad_norm": 0.32624650584367754,
+      "learning_rate": 4.184971098265896e-05,
+      "loss": 0.4569,
+      "step": 475
+    },
+    {
+      "epoch": 0.7420109119251753,
+      "grad_norm": 0.378893068450349,
+      "learning_rate": 4.1820809248554916e-05,
+      "loss": 0.4489,
+      "step": 476
+    },
+    {
+      "epoch": 0.7435697583787997,
+      "grad_norm": 0.3829792535253302,
+      "learning_rate": 4.179190751445087e-05,
+      "loss": 0.4466,
+      "step": 477
+    },
+    {
+      "epoch": 0.745128604832424,
+      "grad_norm": 0.33863233675349563,
+      "learning_rate": 4.176300578034682e-05,
+      "loss": 0.4383,
+      "step": 478
+    },
+    {
+      "epoch": 0.7466874512860483,
+      "grad_norm": 0.40164735104812754,
+      "learning_rate": 4.1734104046242776e-05,
+      "loss": 0.4203,
+      "step": 479
+    },
+    {
+      "epoch": 0.7482462977396727,
+      "grad_norm": 0.35580789823430714,
+      "learning_rate": 4.170520231213873e-05,
+      "loss": 0.4744,
+      "step": 480
+    },
+    {
+      "epoch": 0.7498051441932969,
+      "grad_norm": 0.31246598578051804,
+      "learning_rate": 4.167630057803468e-05,
+      "loss": 0.4219,
+      "step": 481
+    },
+    {
+      "epoch": 0.7513639906469213,
+      "grad_norm": 0.329237795782184,
+      "learning_rate": 4.1647398843930636e-05,
+      "loss": 0.4462,
+      "step": 482
+    },
+    {
+      "epoch": 0.7529228371005456,
+      "grad_norm": 0.32134761939251955,
+      "learning_rate": 4.161849710982659e-05,
+      "loss": 0.4533,
+      "step": 483
+    },
+    {
+      "epoch": 0.7544816835541699,
+      "grad_norm": 0.34349515426855415,
+      "learning_rate": 4.1589595375722543e-05,
+      "loss": 0.4381,
+      "step": 484
+    },
+    {
+      "epoch": 0.7560405300077943,
+      "grad_norm": 0.3706250923736234,
+      "learning_rate": 4.15606936416185e-05,
+      "loss": 0.4406,
+      "step": 485
+    },
+    {
+      "epoch": 0.7575993764614185,
+      "grad_norm": 0.30867896526600563,
+      "learning_rate": 4.153179190751445e-05,
+      "loss": 0.4328,
+      "step": 486
+    },
+    {
+      "epoch": 0.7591582229150429,
+      "grad_norm": 0.3756083443390548,
+      "learning_rate": 4.150289017341041e-05,
+      "loss": 0.4593,
+      "step": 487
+    },
+    {
+      "epoch": 0.7607170693686672,
+      "grad_norm": 0.3359590028525238,
+      "learning_rate": 4.147398843930636e-05,
+      "loss": 0.4452,
+      "step": 488
+    },
+    {
+      "epoch": 0.7622759158222915,
+      "grad_norm": 0.3805120865629876,
+      "learning_rate": 4.144508670520231e-05,
+      "loss": 0.448,
+      "step": 489
+    },
+    {
+      "epoch": 0.7638347622759158,
+      "grad_norm": 0.36856409628466563,
+      "learning_rate": 4.1416184971098264e-05,
+      "loss": 0.4361,
+      "step": 490
+    },
+    {
+      "epoch": 0.7653936087295401,
+      "grad_norm": 0.3130789506998712,
+      "learning_rate": 4.1387283236994224e-05,
+      "loss": 0.4422,
+      "step": 491
+    },
+    {
+      "epoch": 0.7669524551831645,
+      "grad_norm": 0.3420964053109533,
+      "learning_rate": 4.135838150289018e-05,
+      "loss": 0.44,
+      "step": 492
+    },
+    {
+      "epoch": 0.7685113016367888,
+      "grad_norm": 0.3761751528633877,
+      "learning_rate": 4.132947976878613e-05,
+      "loss": 0.4512,
+      "step": 493
+    },
+    {
+      "epoch": 0.7700701480904131,
+      "grad_norm": 0.3217947372011256,
+      "learning_rate": 4.130057803468208e-05,
+      "loss": 0.4542,
+      "step": 494
+    },
+    {
+      "epoch": 0.7716289945440374,
+      "grad_norm": 0.36411196358092746,
+      "learning_rate": 4.127167630057804e-05,
+      "loss": 0.4453,
+      "step": 495
+    },
+    {
+      "epoch": 0.7731878409976617,
+      "grad_norm": 0.36062847920871005,
+      "learning_rate": 4.124277456647399e-05,
+      "loss": 0.432,
+      "step": 496
+    },
+    {
+      "epoch": 0.774746687451286,
+      "grad_norm": 0.31936162560311365,
+      "learning_rate": 4.1213872832369945e-05,
+      "loss": 0.4477,
+      "step": 497
+    },
+    {
+      "epoch": 0.7763055339049104,
+      "grad_norm": 0.3761636426405107,
+      "learning_rate": 4.11849710982659e-05,
+      "loss": 0.4353,
+      "step": 498
+    },
+    {
+      "epoch": 0.7778643803585347,
+      "grad_norm": 0.32932231835478515,
+      "learning_rate": 4.115606936416185e-05,
+      "loss": 0.4467,
+      "step": 499
+    },
+    {
+      "epoch": 0.779423226812159,
+      "grad_norm": 0.38070437317284483,
+      "learning_rate": 4.1127167630057805e-05,
+      "loss": 0.4395,
+      "step": 500
+    },
+    {
+      "epoch": 0.7809820732657833,
+      "grad_norm": 0.3246837506466452,
+      "learning_rate": 4.109826589595376e-05,
+      "loss": 0.4447,
+      "step": 501
+    },
+    {
+      "epoch": 0.7825409197194076,
+      "grad_norm": 0.34953071175085054,
+      "learning_rate": 4.106936416184971e-05,
+      "loss": 0.4333,
+      "step": 502
+    },
+    {
+      "epoch": 0.784099766173032,
+      "grad_norm": 0.37305833228096014,
+      "learning_rate": 4.1040462427745666e-05,
+      "loss": 0.4314,
+      "step": 503
+    },
+    {
+      "epoch": 0.7856586126266563,
+      "grad_norm": 0.3472419499146332,
+      "learning_rate": 4.101156069364162e-05,
+      "loss": 0.4363,
+      "step": 504
+    },
+    {
+      "epoch": 0.7872174590802806,
+      "grad_norm": 0.38866158249187055,
+      "learning_rate": 4.098265895953757e-05,
+      "loss": 0.4429,
+      "step": 505
+    },
+    {
+      "epoch": 0.7887763055339049,
+      "grad_norm": 0.330824609948283,
+      "learning_rate": 4.0953757225433526e-05,
+      "loss": 0.4442,
+      "step": 506
+    },
+    {
+      "epoch": 0.7903351519875292,
+      "grad_norm": 0.39663219339611827,
+      "learning_rate": 4.0924855491329486e-05,
+      "loss": 0.4434,
+      "step": 507
+    },
+    {
+      "epoch": 0.7918939984411536,
+      "grad_norm": 0.3619324831325355,
+      "learning_rate": 4.089595375722543e-05,
+      "loss": 0.4416,
+      "step": 508
+    },
+    {
+      "epoch": 0.7934528448947779,
+      "grad_norm": 0.35246995122654845,
+      "learning_rate": 4.0867052023121386e-05,
+      "loss": 0.4265,
+      "step": 509
+    },
+    {
+      "epoch": 0.7950116913484022,
+      "grad_norm": 0.3612965001810029,
+      "learning_rate": 4.083815028901734e-05,
+      "loss": 0.433,
+      "step": 510
+    },
+    {
+      "epoch": 0.7965705378020265,
+      "grad_norm": 0.38138895157014485,
+      "learning_rate": 4.08092485549133e-05,
+      "loss": 0.4341,
+      "step": 511
+    },
+    {
+      "epoch": 0.7981293842556508,
+      "grad_norm": 0.3449144364761348,
+      "learning_rate": 4.0780346820809253e-05,
+      "loss": 0.457,
+      "step": 512
+    },
+    {
+      "epoch": 0.7996882307092751,
+      "grad_norm": 0.4015780437671286,
+      "learning_rate": 4.07514450867052e-05,
+      "loss": 0.43,
+      "step": 513
+    },
+    {
+      "epoch": 0.8012470771628994,
+      "grad_norm": 0.31072059103382504,
+      "learning_rate": 4.0722543352601154e-05,
+      "loss": 0.4424,
+      "step": 514
+    },
+    {
+      "epoch": 0.8028059236165238,
+      "grad_norm": 0.3758869241826082,
+      "learning_rate": 4.0693641618497114e-05,
+      "loss": 0.4568,
+      "step": 515
+    },
+    {
+      "epoch": 0.8043647700701481,
+      "grad_norm": 0.3383461562057852,
+      "learning_rate": 4.066473988439307e-05,
+      "loss": 0.4513,
+      "step": 516
+    },
+    {
+      "epoch": 0.8059236165237724,
+      "grad_norm": 0.35697829825213634,
+      "learning_rate": 4.063583815028902e-05,
+      "loss": 0.4512,
+      "step": 517
+    },
+    {
+      "epoch": 0.8074824629773967,
+      "grad_norm": 0.3930071983033836,
+      "learning_rate": 4.0606936416184974e-05,
+      "loss": 0.4386,
+      "step": 518
+    },
+    {
+      "epoch": 0.809041309431021,
+      "grad_norm": 0.34936814110167913,
+      "learning_rate": 4.057803468208093e-05,
+      "loss": 0.4249,
+      "step": 519
+    },
+    {
+      "epoch": 0.8106001558846454,
+      "grad_norm": 0.3477156813236279,
+      "learning_rate": 4.054913294797688e-05,
+      "loss": 0.4421,
+      "step": 520
+    },
+    {
+      "epoch": 0.8121590023382697,
+      "grad_norm": 0.3771551756284452,
+      "learning_rate": 4.0520231213872834e-05,
+      "loss": 0.4422,
+      "step": 521
+    },
+    {
+      "epoch": 0.813717848791894,
+      "grad_norm": 0.3915428314100825,
+      "learning_rate": 4.049132947976879e-05,
+      "loss": 0.4489,
+      "step": 522
+    },
+    {
+      "epoch": 0.8152766952455183,
+      "grad_norm": 0.3438944943929536,
+      "learning_rate": 4.046242774566474e-05,
+      "loss": 0.4403,
+      "step": 523
+    },
+    {
+      "epoch": 0.8168355416991426,
+      "grad_norm": 0.41104576558641404,
+      "learning_rate": 4.0433526011560695e-05,
+      "loss": 0.4606,
+      "step": 524
+    },
+    {
+      "epoch": 0.818394388152767,
+      "grad_norm": 0.37712289992365605,
+      "learning_rate": 4.040462427745665e-05,
+      "loss": 0.4325,
+      "step": 525
+    },
+    {
+      "epoch": 0.8199532346063912,
+      "grad_norm": 0.3008825974501281,
+      "learning_rate": 4.03757225433526e-05,
+      "loss": 0.4348,
+      "step": 526
+    },
+    {
+      "epoch": 0.8215120810600156,
+      "grad_norm": 0.3701724452827897,
+      "learning_rate": 4.034682080924856e-05,
+      "loss": 0.444,
+      "step": 527
+    },
+    {
+      "epoch": 0.8230709275136399,
+      "grad_norm": 0.4810485926905224,
+      "learning_rate": 4.031791907514451e-05,
+      "loss": 0.4608,
+      "step": 528
+    },
+    {
+      "epoch": 0.8246297739672642,
+      "grad_norm": 0.3549215989456713,
+      "learning_rate": 4.028901734104046e-05,
+      "loss": 0.4539,
+      "step": 529
+    },
+    {
+      "epoch": 0.8261886204208886,
+      "grad_norm": 0.37028937018443525,
+      "learning_rate": 4.0260115606936415e-05,
+      "loss": 0.4485,
+      "step": 530
+    },
+    {
+      "epoch": 0.8277474668745128,
+      "grad_norm": 0.4323084487068229,
+      "learning_rate": 4.0231213872832376e-05,
+      "loss": 0.4622,
+      "step": 531
+    },
+    {
+      "epoch": 0.8293063133281372,
+      "grad_norm": 0.37099996438121063,
+      "learning_rate": 4.020231213872833e-05,
+      "loss": 0.4389,
+      "step": 532
+    },
+    {
+      "epoch": 0.8308651597817615,
+      "grad_norm": 0.4485476616523096,
+      "learning_rate": 4.0173410404624276e-05,
+      "loss": 0.458,
+      "step": 533
+    },
+    {
+      "epoch": 0.8324240062353858,
+      "grad_norm": 0.47843701934513816,
+      "learning_rate": 4.014450867052023e-05,
+      "loss": 0.4313,
+      "step": 534
+    },
+    {
+      "epoch": 0.8339828526890102,
+      "grad_norm": 0.39097267691711673,
+      "learning_rate": 4.011560693641619e-05,
+      "loss": 0.457,
+      "step": 535
+    },
+    {
+      "epoch": 0.8355416991426344,
+      "grad_norm": 0.44705834158808144,
+      "learning_rate": 4.008670520231214e-05,
+      "loss": 0.4411,
+      "step": 536
+    },
+    {
+      "epoch": 0.8371005455962588,
+      "grad_norm": 0.39009820610398244,
+      "learning_rate": 4.0057803468208096e-05,
+      "loss": 0.4666,
+      "step": 537
+    },
+    {
+      "epoch": 0.838659392049883,
+      "grad_norm": 0.40676152233104423,
+      "learning_rate": 4.002890173410404e-05,
+      "loss": 0.4464,
+      "step": 538
+    },
+    {
+      "epoch": 0.8402182385035074,
+      "grad_norm": 0.38980475885919486,
+      "learning_rate": 4e-05,
+      "loss": 0.4287,
+      "step": 539
+    },
+    {
+      "epoch": 0.8417770849571318,
+      "grad_norm": 0.4155553631446244,
+      "learning_rate": 3.997109826589596e-05,
+      "loss": 0.4592,
+      "step": 540
+    },
+    {
+      "epoch": 0.843335931410756,
+      "grad_norm": 0.4683782615674356,
+      "learning_rate": 3.994219653179191e-05,
+      "loss": 0.4434,
+      "step": 541
+    },
+    {
+      "epoch": 0.8448947778643804,
+      "grad_norm": 0.4270570741139377,
+      "learning_rate": 3.9913294797687864e-05,
+      "loss": 0.4471,
+      "step": 542
+    },
+    {
+      "epoch": 0.8464536243180046,
+      "grad_norm": 0.3328227432824558,
+      "learning_rate": 3.988439306358382e-05,
+      "loss": 0.4342,
+      "step": 543
+    },
+    {
+      "epoch": 0.848012470771629,
+      "grad_norm": 0.383218659320037,
+      "learning_rate": 3.985549132947977e-05,
+      "loss": 0.4418,
+      "step": 544
+    },
+    {
+      "epoch": 0.8495713172252534,
+      "grad_norm": 0.41491251710244376,
+      "learning_rate": 3.9826589595375724e-05,
+      "loss": 0.4562,
+      "step": 545
+    },
+    {
+      "epoch": 0.8511301636788776,
+      "grad_norm": 0.4581749465782902,
+      "learning_rate": 3.979768786127168e-05,
+      "loss": 0.4543,
+      "step": 546
+    },
+    {
+      "epoch": 0.852689010132502,
+      "grad_norm": 0.39063683965855545,
+      "learning_rate": 3.976878612716764e-05,
+      "loss": 0.4403,
+      "step": 547
+    },
+    {
+      "epoch": 0.8542478565861262,
+      "grad_norm": 0.4827779522726547,
+      "learning_rate": 3.9739884393063584e-05,
+      "loss": 0.4371,
+      "step": 548
+    },
+    {
+      "epoch": 0.8558067030397506,
+      "grad_norm": 0.4619643532977416,
+      "learning_rate": 3.971098265895954e-05,
+      "loss": 0.4481,
+      "step": 549
+    },
+    {
+      "epoch": 0.857365549493375,
+      "grad_norm": 0.38833379166687454,
+      "learning_rate": 3.968208092485549e-05,
+      "loss": 0.4298,
+      "step": 550
+    },
+    {
+      "epoch": 0.8589243959469992,
+      "grad_norm": 0.47500496297331074,
+      "learning_rate": 3.965317919075145e-05,
+      "loss": 0.4242,
+      "step": 551
+    },
+    {
+      "epoch": 0.8604832424006236,
+      "grad_norm": 0.43102309998132354,
+      "learning_rate": 3.9624277456647405e-05,
+      "loss": 0.4417,
+      "step": 552
+    },
+    {
+      "epoch": 0.8620420888542478,
+      "grad_norm": 0.5161140253561636,
+      "learning_rate": 3.959537572254335e-05,
+      "loss": 0.4497,
+      "step": 553
+    },
+    {
+      "epoch": 0.8636009353078722,
+      "grad_norm": 0.5456731985809448,
+      "learning_rate": 3.9566473988439305e-05,
+      "loss": 0.4179,
+      "step": 554
+    },
+    {
+      "epoch": 0.8651597817614964,
+      "grad_norm": 0.4207485633634238,
+      "learning_rate": 3.9537572254335265e-05,
+      "loss": 0.4427,
+      "step": 555
+    },
+    {
+      "epoch": 0.8667186282151208,
+      "grad_norm": 0.5759488662135421,
+      "learning_rate": 3.950867052023122e-05,
+      "loss": 0.4477,
+      "step": 556
+    },
+    {
+      "epoch": 0.8682774746687452,
+      "grad_norm": 0.4021704450074726,
+      "learning_rate": 3.947976878612717e-05,
+      "loss": 0.4463,
+      "step": 557
+    },
+    {
+      "epoch": 0.8698363211223694,
+      "grad_norm": 0.42177083348273137,
+      "learning_rate": 3.945086705202312e-05,
+      "loss": 0.4598,
+      "step": 558
+    },
+    {
+      "epoch": 0.8713951675759938,
+      "grad_norm": 0.4701952158517483,
+      "learning_rate": 3.942196531791908e-05,
+      "loss": 0.4394,
+      "step": 559
+    },
+    {
+      "epoch": 0.872954014029618,
+      "grad_norm": 0.44080844830392035,
+      "learning_rate": 3.939306358381503e-05,
+      "loss": 0.4493,
+      "step": 560
+    },
+    {
+      "epoch": 0.8745128604832424,
+      "grad_norm": 0.3977306328304877,
+      "learning_rate": 3.9364161849710986e-05,
+      "loss": 0.4524,
+      "step": 561
+    },
+    {
+      "epoch": 0.8760717069368668,
+      "grad_norm": 0.43018670105627926,
+      "learning_rate": 3.933526011560694e-05,
+      "loss": 0.4267,
+      "step": 562
+    },
+    {
+      "epoch": 0.877630553390491,
+      "grad_norm": 0.4268384770791736,
+      "learning_rate": 3.930635838150289e-05,
+      "loss": 0.4607,
+      "step": 563
+    },
+    {
+      "epoch": 0.8791893998441154,
+      "grad_norm": 0.38783009673051294,
+      "learning_rate": 3.9277456647398846e-05,
+      "loss": 0.4476,
+      "step": 564
+    },
+    {
+      "epoch": 0.8807482462977396,
+      "grad_norm": 0.35687832758423854,
+      "learning_rate": 3.92485549132948e-05,
+      "loss": 0.435,
+      "step": 565
+    },
+    {
+      "epoch": 0.882307092751364,
+      "grad_norm": 0.43764488183272643,
+      "learning_rate": 3.921965317919075e-05,
+      "loss": 0.4466,
+      "step": 566
+    },
+    {
+      "epoch": 0.8838659392049883,
+      "grad_norm": 0.40005960295573123,
+      "learning_rate": 3.9190751445086707e-05,
+      "loss": 0.4288,
+      "step": 567
+    },
+    {
+      "epoch": 0.8854247856586126,
+      "grad_norm": 0.46289540585836514,
+      "learning_rate": 3.916184971098266e-05,
+      "loss": 0.4545,
+      "step": 568
+    },
+    {
+      "epoch": 0.886983632112237,
+      "grad_norm": 0.3824987032253999,
+      "learning_rate": 3.9132947976878613e-05,
+      "loss": 0.4414,
+      "step": 569
+    },
+    {
+      "epoch": 0.8885424785658612,
+      "grad_norm": 0.47993501337835254,
+      "learning_rate": 3.910404624277457e-05,
+      "loss": 0.4411,
+      "step": 570
+    },
+    {
+      "epoch": 0.8901013250194856,
+      "grad_norm": 0.3194097378426562,
+      "learning_rate": 3.907514450867053e-05,
+      "loss": 0.4532,
+      "step": 571
+    },
+    {
+      "epoch": 0.8916601714731099,
+      "grad_norm": 0.5232202230976029,
+      "learning_rate": 3.904624277456648e-05,
+      "loss": 0.423,
+      "step": 572
+    },
+    {
+      "epoch": 0.8932190179267342,
+      "grad_norm": 0.3126690343426845,
+      "learning_rate": 3.901734104046243e-05,
+      "loss": 0.4333,
+      "step": 573
+    },
+    {
+      "epoch": 0.8947778643803586,
+      "grad_norm": 0.45718182082215625,
+      "learning_rate": 3.898843930635838e-05,
+      "loss": 0.4579,
+      "step": 574
+    },
+    {
+      "epoch": 0.8963367108339828,
+      "grad_norm": 0.3578668564536035,
+      "learning_rate": 3.895953757225434e-05,
+      "loss": 0.4365,
+      "step": 575
+    },
+    {
+      "epoch": 0.8978955572876072,
+      "grad_norm": 0.39528003293735503,
+      "learning_rate": 3.8930635838150294e-05,
+      "loss": 0.4373,
+      "step": 576
+    },
+    {
+      "epoch": 0.8994544037412315,
+      "grad_norm": 0.4144241684007145,
+      "learning_rate": 3.890173410404625e-05,
+      "loss": 0.4261,
+      "step": 577
+    },
+    {
+      "epoch": 0.9010132501948558,
+      "grad_norm": 0.37062244540768974,
+      "learning_rate": 3.8872832369942194e-05,
+      "loss": 0.4495,
+      "step": 578
+    },
+    {
+      "epoch": 0.9025720966484801,
+      "grad_norm": 0.38183987870941294,
+      "learning_rate": 3.884393063583815e-05,
+      "loss": 0.4338,
+      "step": 579
+    },
+    {
+      "epoch": 0.9041309431021044,
+      "grad_norm": 0.39082808942712405,
+      "learning_rate": 3.881502890173411e-05,
+      "loss": 0.4444,
+      "step": 580
+    },
+    {
+      "epoch": 0.9056897895557288,
+      "grad_norm": 0.4477498201218956,
+      "learning_rate": 3.878612716763006e-05,
+      "loss": 0.4519,
+      "step": 581
+    },
+    {
+      "epoch": 0.9072486360093531,
+      "grad_norm": 0.44239045295490875,
+      "learning_rate": 3.8757225433526015e-05,
+      "loss": 0.4481,
+      "step": 582
+    },
+    {
+      "epoch": 0.9088074824629774,
+      "grad_norm": 0.3680081228407661,
+      "learning_rate": 3.872832369942196e-05,
+      "loss": 0.44,
+      "step": 583
+    },
+    {
+      "epoch": 0.9103663289166017,
+      "grad_norm": 0.49017226032397,
+      "learning_rate": 3.869942196531792e-05,
+      "loss": 0.4523,
+      "step": 584
+    },
+    {
+      "epoch": 0.911925175370226,
+      "grad_norm": 0.36889936915401544,
+      "learning_rate": 3.8670520231213875e-05,
+      "loss": 0.4252,
+      "step": 585
+    },
+    {
+      "epoch": 0.9134840218238504,
+      "grad_norm": 0.42106930607963194,
+      "learning_rate": 3.864161849710983e-05,
+      "loss": 0.4366,
+      "step": 586
+    },
+    {
+      "epoch": 0.9150428682774747,
+      "grad_norm": 0.33426016889258103,
+      "learning_rate": 3.861271676300578e-05,
+      "loss": 0.4364,
+      "step": 587
+    },
+    {
+      "epoch": 0.916601714731099,
+      "grad_norm": 0.38283532007129306,
+      "learning_rate": 3.8583815028901736e-05,
+      "loss": 0.4299,
+      "step": 588
+    },
+    {
+      "epoch": 0.9181605611847233,
+      "grad_norm": 0.3843431490728921,
+      "learning_rate": 3.855491329479769e-05,
+      "loss": 0.4423,
+      "step": 589
+    },
+    {
+      "epoch": 0.9197194076383476,
+      "grad_norm": 0.39391225219289805,
+      "learning_rate": 3.852601156069364e-05,
+      "loss": 0.4553,
+      "step": 590
+    },
+    {
+      "epoch": 0.921278254091972,
+      "grad_norm": 0.3921483022893861,
+      "learning_rate": 3.8497109826589596e-05,
+      "loss": 0.4374,
+      "step": 591
+    },
+    {
+      "epoch": 0.9228371005455963,
+      "grad_norm": 0.4586348772859173,
+      "learning_rate": 3.846820809248555e-05,
+      "loss": 0.4391,
+      "step": 592
+    },
+    {
+      "epoch": 0.9243959469992206,
+      "grad_norm": 0.3870941468579518,
+      "learning_rate": 3.84393063583815e-05,
+      "loss": 0.4588,
+      "step": 593
+    },
+    {
+      "epoch": 0.9259547934528449,
+      "grad_norm": 0.4121790906090321,
+      "learning_rate": 3.8410404624277456e-05,
+      "loss": 0.4192,
+      "step": 594
+    },
+    {
+      "epoch": 0.9275136399064692,
+      "grad_norm": 0.3287491996044717,
+      "learning_rate": 3.838150289017341e-05,
+      "loss": 0.4376,
+      "step": 595
+    },
+    {
+      "epoch": 0.9290724863600935,
+      "grad_norm": 0.4668763195586917,
+      "learning_rate": 3.835260115606937e-05,
+      "loss": 0.4513,
+      "step": 596
+    },
+    {
+      "epoch": 0.9306313328137179,
+      "grad_norm": 0.39213339962767724,
+      "learning_rate": 3.832369942196532e-05,
+      "loss": 0.4349,
+      "step": 597
+    },
+    {
+      "epoch": 0.9321901792673422,
+      "grad_norm": 0.42578108338194876,
+      "learning_rate": 3.829479768786127e-05,
+      "loss": 0.4365,
+      "step": 598
+    },
+    {
+      "epoch": 0.9337490257209665,
+      "grad_norm": 0.3949037922002061,
+      "learning_rate": 3.8265895953757224e-05,
+      "loss": 0.4366,
+      "step": 599
+    },
+    {
+      "epoch": 0.9353078721745908,
+      "grad_norm": 0.3826228697949301,
+      "learning_rate": 3.8236994219653184e-05,
+      "loss": 0.4426,
+      "step": 600
+    },
+    {
+      "epoch": 0.9368667186282151,
+      "grad_norm": 0.3650038749614227,
+      "learning_rate": 3.820809248554914e-05,
+      "loss": 0.4339,
+      "step": 601
+    },
+    {
+      "epoch": 0.9384255650818394,
+      "grad_norm": 0.40293156157563836,
+      "learning_rate": 3.817919075144509e-05,
+      "loss": 0.4744,
+      "step": 602
+    },
+    {
+      "epoch": 0.9399844115354637,
+      "grad_norm": 0.40210846514825743,
+      "learning_rate": 3.815028901734104e-05,
+      "loss": 0.4404,
+      "step": 603
+    },
+    {
+      "epoch": 0.9415432579890881,
+      "grad_norm": 0.39097895288839746,
+      "learning_rate": 3.8121387283237e-05,
+      "loss": 0.4401,
+      "step": 604
+    },
+    {
+      "epoch": 0.9431021044427124,
+      "grad_norm": 0.3227109176868314,
+      "learning_rate": 3.809248554913295e-05,
+      "loss": 0.4312,
+      "step": 605
+    },
+    {
+      "epoch": 0.9446609508963367,
+      "grad_norm": 0.4608466312979325,
+      "learning_rate": 3.8063583815028905e-05,
+      "loss": 0.4591,
+      "step": 606
+    },
+    {
+      "epoch": 0.946219797349961,
+      "grad_norm": 0.34756293281666123,
+      "learning_rate": 3.803468208092486e-05,
+      "loss": 0.4566,
+      "step": 607
+    },
+    {
+      "epoch": 0.9477786438035853,
+      "grad_norm": 0.37558636760339875,
+      "learning_rate": 3.800578034682081e-05,
+      "loss": 0.4205,
+      "step": 608
+    },
+    {
+      "epoch": 0.9493374902572097,
+      "grad_norm": 0.4205365075958774,
+      "learning_rate": 3.7976878612716765e-05,
+      "loss": 0.4388,
+      "step": 609
+    },
+    {
+      "epoch": 0.950896336710834,
+      "grad_norm": 0.45096900527646666,
+      "learning_rate": 3.794797687861272e-05,
+      "loss": 0.4316,
+      "step": 610
+    },
+    {
+      "epoch": 0.9524551831644583,
+      "grad_norm": 0.3812977384877062,
+      "learning_rate": 3.791907514450867e-05,
+      "loss": 0.4363,
+      "step": 611
+    },
+    {
+      "epoch": 0.9540140296180826,
+      "grad_norm": 0.48866253778453445,
+      "learning_rate": 3.7890173410404625e-05,
+      "loss": 0.428,
+      "step": 612
+    },
+    {
+      "epoch": 0.9555728760717069,
+      "grad_norm": 0.40499141402747296,
+      "learning_rate": 3.786127167630058e-05,
+      "loss": 0.4381,
+      "step": 613
+    },
+    {
+      "epoch": 0.9571317225253313,
+      "grad_norm": 0.46430129913008983,
+      "learning_rate": 3.783236994219653e-05,
+      "loss": 0.4429,
+      "step": 614
+    },
+    {
+      "epoch": 0.9586905689789555,
+      "grad_norm": 0.34966737213764454,
+      "learning_rate": 3.7803468208092486e-05,
+      "loss": 0.4381,
+      "step": 615
+    },
+    {
+      "epoch": 0.9602494154325799,
+      "grad_norm": 0.4138090253112475,
+      "learning_rate": 3.7774566473988446e-05,
+      "loss": 0.4258,
+      "step": 616
+    },
+    {
+      "epoch": 0.9618082618862042,
+      "grad_norm": 0.3297979382576517,
+      "learning_rate": 3.774566473988439e-05,
+      "loss": 0.4413,
+      "step": 617
+    },
+    {
+      "epoch": 0.9633671083398285,
+      "grad_norm": 0.3863735923761598,
+      "learning_rate": 3.7716763005780346e-05,
+      "loss": 0.431,
+      "step": 618
+    },
+    {
+      "epoch": 0.9649259547934529,
+      "grad_norm": 0.33528959794283103,
+      "learning_rate": 3.76878612716763e-05,
+      "loss": 0.4377,
+      "step": 619
+    },
+    {
+      "epoch": 0.9664848012470771,
+      "grad_norm": 0.3292524974297176,
+      "learning_rate": 3.765895953757226e-05,
+      "loss": 0.4326,
+      "step": 620
+    },
+    {
+      "epoch": 0.9680436477007015,
+      "grad_norm": 0.3156599159357699,
+      "learning_rate": 3.763005780346821e-05,
+      "loss": 0.4236,
+      "step": 621
+    },
+    {
+      "epoch": 0.9696024941543258,
+      "grad_norm": 0.349891638695045,
+      "learning_rate": 3.760115606936416e-05,
+      "loss": 0.4226,
+      "step": 622
+    },
+    {
+      "epoch": 0.9711613406079501,
+      "grad_norm": 0.3027036773941601,
+      "learning_rate": 3.757225433526011e-05,
+      "loss": 0.4357,
+      "step": 623
+    },
+    {
+      "epoch": 0.9727201870615745,
+      "grad_norm": 0.3418983655828593,
+      "learning_rate": 3.754335260115607e-05,
+      "loss": 0.4072,
+      "step": 624
+    },
+    {
+      "epoch": 0.9742790335151987,
+      "grad_norm": 0.32424716914194174,
+      "learning_rate": 3.751445086705203e-05,
+      "loss": 0.4229,
+      "step": 625
+    },
+    {
+      "epoch": 0.9758378799688231,
+      "grad_norm": 0.35397027714852347,
+      "learning_rate": 3.748554913294798e-05,
+      "loss": 0.443,
+      "step": 626
+    },
+    {
+      "epoch": 0.9773967264224473,
+      "grad_norm": 0.35983171780105105,
+      "learning_rate": 3.7456647398843934e-05,
+      "loss": 0.4218,
+      "step": 627
+    },
+    {
+      "epoch": 0.9789555728760717,
+      "grad_norm": 0.36410259602213924,
+      "learning_rate": 3.742774566473989e-05,
+      "loss": 0.4293,
+      "step": 628
+    },
+    {
+      "epoch": 0.9805144193296961,
+      "grad_norm": 0.3865124494448919,
+      "learning_rate": 3.739884393063584e-05,
+      "loss": 0.4229,
+      "step": 629
+    },
+    {
+      "epoch": 0.9820732657833203,
+      "grad_norm": 0.3646826145021347,
+      "learning_rate": 3.7369942196531794e-05,
+      "loss": 0.4308,
+      "step": 630
+    },
+    {
+      "epoch": 0.9836321122369447,
+      "grad_norm": 0.35897147135714,
+      "learning_rate": 3.734104046242775e-05,
+      "loss": 0.4407,
+      "step": 631
+    },
+    {
+      "epoch": 0.9851909586905689,
+      "grad_norm": 0.3391861968595338,
+      "learning_rate": 3.73121387283237e-05,
+      "loss": 0.4354,
+      "step": 632
+    },
+    {
+      "epoch": 0.9867498051441933,
+      "grad_norm": 0.43647757995986053,
+      "learning_rate": 3.7283236994219654e-05,
+      "loss": 0.4274,
+      "step": 633
+    },
+    {
+      "epoch": 0.9883086515978177,
+      "grad_norm": 0.3618260493378971,
+      "learning_rate": 3.725433526011561e-05,
+      "loss": 0.4203,
+      "step": 634
+    },
+    {
+      "epoch": 0.9898674980514419,
+      "grad_norm": 0.40188293786147516,
+      "learning_rate": 3.722543352601156e-05,
+      "loss": 0.4306,
+      "step": 635
+    },
+    {
+      "epoch": 0.9914263445050663,
+      "grad_norm": 0.4129221508285703,
+      "learning_rate": 3.719653179190752e-05,
+      "loss": 0.4248,
+      "step": 636
+    },
+    {
+      "epoch": 0.9929851909586905,
+      "grad_norm": 0.39766604692773766,
+      "learning_rate": 3.716763005780347e-05,
+      "loss": 0.4574,
+      "step": 637
+    },
+    {
+      "epoch": 0.9945440374123149,
+      "grad_norm": 0.5559578417632596,
+      "learning_rate": 3.713872832369942e-05,
+      "loss": 0.469,
+      "step": 638
+    },
+    {
+      "epoch": 0.9961028838659393,
+      "grad_norm": 0.3580889923606767,
+      "learning_rate": 3.7109826589595375e-05,
+      "loss": 0.4091,
+      "step": 639
+    },
+    {
+      "epoch": 0.9976617303195635,
+      "grad_norm": 0.5076402218866569,
+      "learning_rate": 3.7080924855491335e-05,
+      "loss": 0.4305,
+      "step": 640
+    },
+    {
+      "epoch": 0.9992205767731879,
+      "grad_norm": 0.3789119376045301,
+      "learning_rate": 3.705202312138729e-05,
+      "loss": 0.4306,
+      "step": 641
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.3789119376045301,
+      "learning_rate": 3.7023121387283235e-05,
+      "loss": 0.4336,
+      "step": 642
+    },
+    {
+      "epoch": 1.0015588464536243,
+      "grad_norm": 0.628371201501632,
+      "learning_rate": 3.699421965317919e-05,
+      "loss": 0.3748,
+      "step": 643
+    },
+    {
+      "epoch": 1.0031176929072487,
+      "grad_norm": 0.49964431541022886,
+      "learning_rate": 3.696531791907515e-05,
+      "loss": 0.3842,
+      "step": 644
+    },
+    {
+      "epoch": 1.004676539360873,
+      "grad_norm": 0.4115388078983886,
+      "learning_rate": 3.69364161849711e-05,
+      "loss": 0.3921,
+      "step": 645
+    },
+    {
+      "epoch": 1.0062353858144972,
+      "grad_norm": 0.36847896294616717,
+      "learning_rate": 3.6907514450867056e-05,
+      "loss": 0.3917,
+      "step": 646
+    },
+    {
+      "epoch": 1.0077942322681215,
+      "grad_norm": 0.41235306439266933,
+      "learning_rate": 3.6878612716763e-05,
+      "loss": 0.3661,
+      "step": 647
+    },
+    {
+      "epoch": 1.009353078721746,
+      "grad_norm": 0.36043638686025053,
+      "learning_rate": 3.684971098265896e-05,
+      "loss": 0.3793,
+      "step": 648
+    },
+    {
+      "epoch": 1.0109119251753702,
+      "grad_norm": 0.5072805021746171,
+      "learning_rate": 3.6820809248554916e-05,
+      "loss": 0.3731,
+      "step": 649
+    },
+    {
+      "epoch": 1.0124707716289945,
+      "grad_norm": 0.3774967084990004,
+      "learning_rate": 3.679190751445087e-05,
+      "loss": 0.3837,
+      "step": 650
+    },
+    {
+      "epoch": 1.014029618082619,
+      "grad_norm": 0.47148649785731656,
+      "learning_rate": 3.676300578034682e-05,
+      "loss": 0.3655,
+      "step": 651
+    },
+    {
+      "epoch": 1.0155884645362432,
+      "grad_norm": 0.37322483271463347,
+      "learning_rate": 3.6734104046242777e-05,
+      "loss": 0.3955,
+      "step": 652
+    },
+    {
+      "epoch": 1.0171473109898674,
+      "grad_norm": 0.3705138885107383,
+      "learning_rate": 3.670520231213873e-05,
+      "loss": 0.3753,
+      "step": 653
+    },
+    {
+      "epoch": 1.018706157443492,
+      "grad_norm": 0.5310690376603894,
+      "learning_rate": 3.6676300578034683e-05,
+      "loss": 0.365,
+      "step": 654
+    },
+    {
+      "epoch": 1.0202650038971162,
+      "grad_norm": 0.3622397954678426,
+      "learning_rate": 3.664739884393064e-05,
+      "loss": 0.3691,
+      "step": 655
+    },
+    {
+      "epoch": 1.0218238503507404,
+      "grad_norm": 0.48489876341543203,
+      "learning_rate": 3.66184971098266e-05,
+      "loss": 0.3757,
+      "step": 656
+    },
+    {
+      "epoch": 1.0233826968043647,
+      "grad_norm": 0.3846528970782493,
+      "learning_rate": 3.6589595375722544e-05,
+      "loss": 0.3766,
+      "step": 657
+    },
+    {
+      "epoch": 1.0249415432579891,
+      "grad_norm": 0.4581460525204425,
+      "learning_rate": 3.65606936416185e-05,
+      "loss": 0.3513,
+      "step": 658
+    },
+    {
+      "epoch": 1.0265003897116134,
+      "grad_norm": 0.3465258787798717,
+      "learning_rate": 3.653179190751445e-05,
+      "loss": 0.3624,
+      "step": 659
+    },
+    {
+      "epoch": 1.0280592361652376,
+      "grad_norm": 0.336442810655498,
+      "learning_rate": 3.650289017341041e-05,
+      "loss": 0.3572,
+      "step": 660
+    },
+    {
+      "epoch": 1.0296180826188621,
+      "grad_norm": 0.3003091419390878,
+      "learning_rate": 3.6473988439306364e-05,
+      "loss": 0.367,
+      "step": 661
+    },
+    {
+      "epoch": 1.0311769290724864,
+      "grad_norm": 0.32146010358650395,
+      "learning_rate": 3.644508670520231e-05,
+      "loss": 0.3758,
+      "step": 662
+    },
+    {
+      "epoch": 1.0327357755261106,
+      "grad_norm": 0.3233773848232951,
+      "learning_rate": 3.6416184971098265e-05,
+      "loss": 0.3602,
+      "step": 663
+    },
+    {
+      "epoch": 1.034294621979735,
+      "grad_norm": 0.3511338330800115,
+      "learning_rate": 3.6387283236994225e-05,
+      "loss": 0.3496,
+      "step": 664
+    },
+    {
+      "epoch": 1.0358534684333593,
+      "grad_norm": 0.2996717533713208,
+      "learning_rate": 3.635838150289018e-05,
+      "loss": 0.3462,
+      "step": 665
+    },
+    {
+      "epoch": 1.0374123148869836,
+      "grad_norm": 0.32342558777316216,
+      "learning_rate": 3.632947976878613e-05,
+      "loss": 0.3747,
+      "step": 666
+    },
+    {
+      "epoch": 1.0389711613406079,
+      "grad_norm": 0.3367570922014961,
+      "learning_rate": 3.630057803468208e-05,
+      "loss": 0.3614,
+      "step": 667
+    },
+    {
+      "epoch": 1.0405300077942323,
+      "grad_norm": 0.3441547440118924,
+      "learning_rate": 3.627167630057804e-05,
+      "loss": 0.3917,
+      "step": 668
+    },
+    {
+      "epoch": 1.0420888542478566,
+      "grad_norm": 0.3361181158280621,
+      "learning_rate": 3.624277456647399e-05,
+      "loss": 0.3832,
+      "step": 669
+    },
+    {
+      "epoch": 1.0436477007014808,
+      "grad_norm": 0.3291404615432134,
+      "learning_rate": 3.6213872832369945e-05,
+      "loss": 0.3622,
+      "step": 670
+    },
+    {
+      "epoch": 1.0452065471551053,
+      "grad_norm": 0.30771644853510227,
+      "learning_rate": 3.61849710982659e-05,
+      "loss": 0.3796,
+      "step": 671
+    },
+    {
+      "epoch": 1.0467653936087296,
+      "grad_norm": 0.33176245015646244,
+      "learning_rate": 3.615606936416185e-05,
+      "loss": 0.3788,
+      "step": 672
+    },
+    {
+      "epoch": 1.0483242400623538,
+      "grad_norm": 0.9925708666951382,
+      "learning_rate": 3.6127167630057806e-05,
+      "loss": 0.4087,
+      "step": 673
+    },
+    {
+      "epoch": 1.0498830865159783,
+      "grad_norm": 0.3681749130556859,
+      "learning_rate": 3.609826589595376e-05,
+      "loss": 0.3647,
+      "step": 674
+    },
+    {
+      "epoch": 1.0514419329696025,
+      "grad_norm": 0.3804125438000747,
+      "learning_rate": 3.606936416184971e-05,
+      "loss": 0.387,
+      "step": 675
+    },
+    {
+      "epoch": 1.0530007794232268,
+      "grad_norm": 0.3792677722995949,
+      "learning_rate": 3.6040462427745666e-05,
+      "loss": 0.3806,
+      "step": 676
+    },
+    {
+      "epoch": 1.054559625876851,
+      "grad_norm": 0.3546135860064871,
+      "learning_rate": 3.601156069364162e-05,
+      "loss": 0.3858,
+      "step": 677
+    },
+    {
+      "epoch": 1.0561184723304755,
+      "grad_norm": 0.3603149324311687,
+      "learning_rate": 3.598265895953757e-05,
+      "loss": 0.3744,
+      "step": 678
+    },
+    {
+      "epoch": 1.0576773187840998,
+      "grad_norm": 0.34480487604338456,
+      "learning_rate": 3.5953757225433526e-05,
+      "loss": 0.3642,
+      "step": 679
+    },
+    {
+      "epoch": 1.059236165237724,
+      "grad_norm": 0.3405204986814321,
+      "learning_rate": 3.592485549132948e-05,
+      "loss": 0.364,
+      "step": 680
+    },
+    {
+      "epoch": 1.0607950116913485,
+      "grad_norm": 0.3605005587786407,
+      "learning_rate": 3.589595375722544e-05,
+      "loss": 0.3619,
+      "step": 681
+    },
+    {
+      "epoch": 1.0623538581449727,
+      "grad_norm": 0.4006257143609181,
+      "learning_rate": 3.586705202312139e-05,
+      "loss": 0.3818,
+      "step": 682
+    },
+    {
+      "epoch": 1.063912704598597,
+      "grad_norm": 0.29455634443366796,
+      "learning_rate": 3.583815028901734e-05,
+      "loss": 0.3673,
+      "step": 683
+    },
+    {
+      "epoch": 1.0654715510522212,
+      "grad_norm": 0.36050845424709793,
+      "learning_rate": 3.5809248554913294e-05,
+      "loss": 0.367,
+      "step": 684
+    },
+    {
+      "epoch": 1.0670303975058457,
+      "grad_norm": 0.3404885010186498,
+      "learning_rate": 3.5780346820809254e-05,
+      "loss": 0.3931,
+      "step": 685
+    },
+    {
+      "epoch": 1.06858924395947,
+      "grad_norm": 0.3209534029364564,
+      "learning_rate": 3.575144508670521e-05,
+      "loss": 0.371,
+      "step": 686
+    },
+    {
+      "epoch": 1.0701480904130942,
+      "grad_norm": 0.33657202823646476,
+      "learning_rate": 3.5722543352601154e-05,
+      "loss": 0.3789,
+      "step": 687
+    },
+    {
+      "epoch": 1.0717069368667187,
+      "grad_norm": 0.3056381970556524,
+      "learning_rate": 3.569364161849711e-05,
+      "loss": 0.3781,
+      "step": 688
+    },
+    {
+      "epoch": 1.073265783320343,
+      "grad_norm": 0.2869272425111672,
+      "learning_rate": 3.566473988439307e-05,
+      "loss": 0.3795,
+      "step": 689
+    },
+    {
+      "epoch": 1.0748246297739672,
+      "grad_norm": 0.3423224557400581,
+      "learning_rate": 3.563583815028902e-05,
+      "loss": 0.3569,
+      "step": 690
+    },
+    {
+      "epoch": 1.0763834762275917,
+      "grad_norm": 0.3620299786614207,
+      "learning_rate": 3.5606936416184975e-05,
+      "loss": 0.3813,
+      "step": 691
+    },
+    {
+      "epoch": 1.077942322681216,
+      "grad_norm": 0.3164910374888722,
+      "learning_rate": 3.557803468208092e-05,
+      "loss": 0.3548,
+      "step": 692
+    },
+    {
+      "epoch": 1.0795011691348402,
+      "grad_norm": 0.32077378022129677,
+      "learning_rate": 3.554913294797688e-05,
+      "loss": 0.3714,
+      "step": 693
+    },
+    {
+      "epoch": 1.0810600155884647,
+      "grad_norm": 0.3449220187992264,
+      "learning_rate": 3.5520231213872835e-05,
+      "loss": 0.363,
+      "step": 694
+    },
+    {
+      "epoch": 1.082618862042089,
+      "grad_norm": 0.3342998444359934,
+      "learning_rate": 3.549132947976879e-05,
+      "loss": 0.3482,
+      "step": 695
+    },
+    {
+      "epoch": 1.0841777084957132,
+      "grad_norm": 0.35266941801482377,
+      "learning_rate": 3.546242774566474e-05,
+      "loss": 0.3784,
+      "step": 696
+    },
+    {
+      "epoch": 1.0857365549493374,
+      "grad_norm": 0.4488757216839682,
+      "learning_rate": 3.5433526011560695e-05,
+      "loss": 0.3788,
+      "step": 697
+    },
+    {
+      "epoch": 1.0872954014029619,
+      "grad_norm": 0.3288290304532232,
+      "learning_rate": 3.540462427745665e-05,
+      "loss": 0.3812,
+      "step": 698
+    },
+    {
+      "epoch": 1.0888542478565861,
+      "grad_norm": 0.4417200363539964,
+      "learning_rate": 3.53757225433526e-05,
+      "loss": 0.377,
+      "step": 699
+    },
+    {
+      "epoch": 1.0904130943102104,
+      "grad_norm": 0.31742949476273763,
+      "learning_rate": 3.5346820809248556e-05,
+      "loss": 0.3639,
+      "step": 700
+    },
+    {
+      "epoch": 1.0919719407638349,
+      "grad_norm": 0.3308705930780861,
+      "learning_rate": 3.531791907514451e-05,
+      "loss": 0.3634,
+      "step": 701
+    },
+    {
+      "epoch": 1.093530787217459,
+      "grad_norm": 0.3382413793786327,
+      "learning_rate": 3.528901734104046e-05,
+      "loss": 0.3598,
+      "step": 702
+    },
+    {
+      "epoch": 1.0950896336710834,
+      "grad_norm": 0.3621055356977116,
+      "learning_rate": 3.5260115606936416e-05,
+      "loss": 0.3657,
+      "step": 703
+    },
+    {
+      "epoch": 1.0966484801247076,
+      "grad_norm": 0.30903567078078115,
+      "learning_rate": 3.523121387283237e-05,
+      "loss": 0.3746,
+      "step": 704
+    },
+    {
+      "epoch": 1.098207326578332,
+      "grad_norm": 0.43168788302923716,
+      "learning_rate": 3.520231213872833e-05,
+      "loss": 0.3786,
+      "step": 705
+    },
+    {
+      "epoch": 1.0997661730319563,
+      "grad_norm": 0.3735078603768068,
+      "learning_rate": 3.5173410404624276e-05,
+      "loss": 0.3554,
+      "step": 706
+    },
+    {
+      "epoch": 1.1013250194855806,
+      "grad_norm": 0.3705144355540284,
+      "learning_rate": 3.514450867052023e-05,
+      "loss": 0.3491,
+      "step": 707
+    },
+    {
+      "epoch": 1.102883865939205,
+      "grad_norm": 0.36212715939080664,
+      "learning_rate": 3.511560693641618e-05,
+      "loss": 0.3665,
+      "step": 708
+    },
+    {
+      "epoch": 1.1044427123928293,
+      "grad_norm": 0.3844205403442548,
+      "learning_rate": 3.508670520231214e-05,
+      "loss": 0.369,
+      "step": 709
+    },
+    {
+      "epoch": 1.1060015588464536,
+      "grad_norm": 0.3434945710583639,
+      "learning_rate": 3.50578034682081e-05,
+      "loss": 0.3724,
+      "step": 710
+    },
+    {
+      "epoch": 1.107560405300078,
+      "grad_norm": 0.4376944680100341,
+      "learning_rate": 3.502890173410405e-05,
+      "loss": 0.3829,
+      "step": 711
+    },
+    {
+      "epoch": 1.1091192517537023,
+      "grad_norm": 0.3162790873067193,
+      "learning_rate": 3.5e-05,
+      "loss": 0.3794,
+      "step": 712
+    },
+    {
+      "epoch": 1.1106780982073265,
+      "grad_norm": 0.30611808166388454,
+      "learning_rate": 3.497109826589596e-05,
+      "loss": 0.3652,
+      "step": 713
+    },
+    {
+      "epoch": 1.1122369446609508,
+      "grad_norm": 0.3100967615655972,
+      "learning_rate": 3.494219653179191e-05,
+      "loss": 0.3715,
+      "step": 714
+    },
+    {
+      "epoch": 1.1137957911145753,
+      "grad_norm": 0.29998053632892707,
+      "learning_rate": 3.4913294797687864e-05,
+      "loss": 0.385,
+      "step": 715
+    },
+    {
+      "epoch": 1.1153546375681995,
+      "grad_norm": 0.3531103029098729,
+      "learning_rate": 3.488439306358382e-05,
+      "loss": 0.3852,
+      "step": 716
+    },
+    {
+      "epoch": 1.1169134840218238,
+      "grad_norm": 0.28836989657125967,
+      "learning_rate": 3.485549132947977e-05,
+      "loss": 0.369,
+      "step": 717
+    },
+    {
+      "epoch": 1.1184723304754483,
+      "grad_norm": 0.3726105295277464,
+      "learning_rate": 3.4826589595375724e-05,
+      "loss": 0.3901,
+      "step": 718
+    },
+    {
+      "epoch": 1.1200311769290725,
+      "grad_norm": 0.30082206554969804,
+      "learning_rate": 3.479768786127168e-05,
+      "loss": 0.3672,
+      "step": 719
+    },
+    {
+      "epoch": 1.1215900233826968,
+      "grad_norm": 0.3229975948532406,
+      "learning_rate": 3.476878612716763e-05,
+      "loss": 0.3741,
+      "step": 720
+    },
+    {
+      "epoch": 1.1231488698363212,
+      "grad_norm": 0.30690351827295265,
+      "learning_rate": 3.4739884393063585e-05,
+      "loss": 0.3612,
+      "step": 721
+    },
+    {
+      "epoch": 1.1247077162899455,
+      "grad_norm": 0.3558269454615068,
+      "learning_rate": 3.471098265895954e-05,
+      "loss": 0.3887,
+      "step": 722
+    },
+    {
+      "epoch": 1.1262665627435697,
+      "grad_norm": 0.34137813880164614,
+      "learning_rate": 3.468208092485549e-05,
+      "loss": 0.3835,
+      "step": 723
+    },
+    {
+      "epoch": 1.127825409197194,
+      "grad_norm": 0.3060779272980073,
+      "learning_rate": 3.4653179190751445e-05,
+      "loss": 0.3825,
+      "step": 724
+    },
+    {
+      "epoch": 1.1293842556508185,
+      "grad_norm": 0.34107259706747284,
+      "learning_rate": 3.4624277456647405e-05,
+      "loss": 0.3716,
+      "step": 725
+    },
+    {
+      "epoch": 1.1309431021044427,
+      "grad_norm": 0.28915982893470976,
+      "learning_rate": 3.459537572254335e-05,
+      "loss": 0.3658,
+      "step": 726
+    },
+    {
+      "epoch": 1.132501948558067,
+      "grad_norm": 0.2927164690689312,
+      "learning_rate": 3.4566473988439305e-05,
+      "loss": 0.3604,
+      "step": 727
+    },
+    {
+      "epoch": 1.1340607950116914,
+      "grad_norm": 0.3191085857306229,
+      "learning_rate": 3.453757225433526e-05,
+      "loss": 0.3837,
+      "step": 728
+    },
+    {
+      "epoch": 1.1356196414653157,
+      "grad_norm": 0.3079933409669794,
+      "learning_rate": 3.450867052023122e-05,
+      "loss": 0.374,
+      "step": 729
+    },
+    {
+      "epoch": 1.13717848791894,
+      "grad_norm": 0.2904492631644862,
+      "learning_rate": 3.447976878612717e-05,
+      "loss": 0.3778,
+      "step": 730
+    },
+    {
+      "epoch": 1.1387373343725642,
+      "grad_norm": 0.29735218541832303,
+      "learning_rate": 3.445086705202312e-05,
+      "loss": 0.3573,
+      "step": 731
+    },
+    {
+      "epoch": 1.1402961808261887,
+      "grad_norm": 0.3152575212565156,
+      "learning_rate": 3.442196531791907e-05,
+      "loss": 0.373,
+      "step": 732
+    },
+    {
+      "epoch": 1.141855027279813,
+      "grad_norm": 0.32039618688552635,
+      "learning_rate": 3.439306358381503e-05,
+      "loss": 0.3779,
+      "step": 733
+    },
+    {
+      "epoch": 1.1434138737334372,
+      "grad_norm": 0.3168459279894005,
+      "learning_rate": 3.4364161849710986e-05,
+      "loss": 0.3612,
+      "step": 734
+    },
+    {
+      "epoch": 1.1449727201870616,
+      "grad_norm": 0.3382435631010463,
+      "learning_rate": 3.433526011560694e-05,
+      "loss": 0.375,
+      "step": 735
+    },
+    {
+      "epoch": 1.146531566640686,
+      "grad_norm": 0.3451916527844319,
+      "learning_rate": 3.430635838150289e-05,
+      "loss": 0.3656,
+      "step": 736
+    },
+    {
+      "epoch": 1.1480904130943101,
+      "grad_norm": 0.29795458770948763,
+      "learning_rate": 3.427745664739885e-05,
+      "loss": 0.3517,
+      "step": 737
+    },
+    {
+      "epoch": 1.1496492595479346,
+      "grad_norm": 0.3204048119901522,
+      "learning_rate": 3.42485549132948e-05,
+      "loss": 0.3753,
+      "step": 738
+    },
+    {
+      "epoch": 1.1512081060015589,
+      "grad_norm": 0.3509132683834904,
+      "learning_rate": 3.4219653179190754e-05,
+      "loss": 0.3718,
+      "step": 739
+    },
+    {
+      "epoch": 1.1527669524551831,
+      "grad_norm": 0.3252836365240409,
+      "learning_rate": 3.419075144508671e-05,
+      "loss": 0.3884,
+      "step": 740
+    },
+    {
+      "epoch": 1.1543257989088076,
+      "grad_norm": 0.2938786860494216,
+      "learning_rate": 3.416184971098266e-05,
+      "loss": 0.3508,
+      "step": 741
+    },
+    {
+      "epoch": 1.1558846453624319,
+      "grad_norm": 0.3286847180631177,
+      "learning_rate": 3.4132947976878614e-05,
+      "loss": 0.3887,
+      "step": 742
+    },
+    {
+      "epoch": 1.157443491816056,
+      "grad_norm": 0.293702845880183,
+      "learning_rate": 3.410404624277457e-05,
+      "loss": 0.3625,
+      "step": 743
+    },
+    {
+      "epoch": 1.1590023382696804,
+      "grad_norm": 0.3637581698851474,
+      "learning_rate": 3.407514450867052e-05,
+      "loss": 0.3807,
+      "step": 744
+    },
+    {
+      "epoch": 1.1605611847233048,
+      "grad_norm": 0.33030454774450496,
+      "learning_rate": 3.404624277456648e-05,
+      "loss": 0.3625,
+      "step": 745
+    },
+    {
+      "epoch": 1.162120031176929,
+      "grad_norm": 0.3413072692858561,
+      "learning_rate": 3.401734104046243e-05,
+      "loss": 0.3799,
+      "step": 746
+    },
+    {
+      "epoch": 1.1636788776305533,
+      "grad_norm": 0.3282042861228878,
+      "learning_rate": 3.398843930635838e-05,
+      "loss": 0.3744,
+      "step": 747
+    },
+    {
+      "epoch": 1.1652377240841778,
+      "grad_norm": 0.4051993980268251,
+      "learning_rate": 3.3959537572254335e-05,
+      "loss": 0.3658,
+      "step": 748
+    },
+    {
+      "epoch": 1.166796570537802,
+      "grad_norm": 0.3513826597230518,
+      "learning_rate": 3.3930635838150295e-05,
+      "loss": 0.3641,
+      "step": 749
+    },
+    {
+      "epoch": 1.1683554169914263,
+      "grad_norm": 0.35299247249790444,
+      "learning_rate": 3.390173410404625e-05,
+      "loss": 0.3654,
+      "step": 750
+    },
+    {
+      "epoch": 1.1699142634450506,
+      "grad_norm": 0.3014386293388838,
+      "learning_rate": 3.3872832369942195e-05,
+      "loss": 0.3765,
+      "step": 751
+    },
+    {
+      "epoch": 1.171473109898675,
+      "grad_norm": 0.3603523634257503,
+      "learning_rate": 3.384393063583815e-05,
+      "loss": 0.3845,
+      "step": 752
+    },
+    {
+      "epoch": 1.1730319563522993,
+      "grad_norm": 0.29665311038360115,
+      "learning_rate": 3.381502890173411e-05,
+      "loss": 0.3797,
+      "step": 753
+    },
+    {
+      "epoch": 1.1745908028059235,
+      "grad_norm": 0.2935340087785823,
+      "learning_rate": 3.378612716763006e-05,
+      "loss": 0.3602,
+      "step": 754
+    },
+    {
+      "epoch": 1.176149649259548,
+      "grad_norm": 0.33634845355876714,
+      "learning_rate": 3.3757225433526015e-05,
+      "loss": 0.3796,
+      "step": 755
+    },
+    {
+      "epoch": 1.1777084957131723,
+      "grad_norm": 0.33117541570348524,
+      "learning_rate": 3.372832369942196e-05,
+      "loss": 0.4078,
+      "step": 756
+    },
+    {
+      "epoch": 1.1792673421667965,
+      "grad_norm": 0.2734330693617624,
+      "learning_rate": 3.369942196531792e-05,
+      "loss": 0.3731,
+      "step": 757
+    },
+    {
+      "epoch": 1.1808261886204208,
+      "grad_norm": 0.35008178961287134,
+      "learning_rate": 3.3670520231213876e-05,
+      "loss": 0.3764,
+      "step": 758
+    },
+    {
+      "epoch": 1.1823850350740452,
+      "grad_norm": 0.3094977957875163,
+      "learning_rate": 3.364161849710983e-05,
+      "loss": 0.3565,
+      "step": 759
+    },
+    {
+      "epoch": 1.1839438815276695,
+      "grad_norm": 0.29463076564571844,
+      "learning_rate": 3.361271676300578e-05,
+      "loss": 0.3705,
+      "step": 760
+    },
+    {
+      "epoch": 1.1855027279812937,
+      "grad_norm": 0.3164209650580971,
+      "learning_rate": 3.3583815028901736e-05,
+      "loss": 0.3764,
+      "step": 761
+    },
+    {
+      "epoch": 1.1870615744349182,
+      "grad_norm": 0.28906263199367943,
+      "learning_rate": 3.355491329479769e-05,
+      "loss": 0.3889,
+      "step": 762
+    },
+    {
+      "epoch": 1.1886204208885425,
+      "grad_norm": 0.3138486170125581,
+      "learning_rate": 3.352601156069364e-05,
+      "loss": 0.3781,
+      "step": 763
+    },
+    {
+      "epoch": 1.1901792673421667,
+      "grad_norm": 0.31679716557668985,
+      "learning_rate": 3.3497109826589596e-05,
+      "loss": 0.3725,
+      "step": 764
+    },
+    {
+      "epoch": 1.1917381137957912,
+      "grad_norm": 0.2819822969115435,
+      "learning_rate": 3.346820809248556e-05,
+      "loss": 0.3621,
+      "step": 765
+    },
+    {
+      "epoch": 1.1932969602494155,
+      "grad_norm": 0.2900191750020783,
+      "learning_rate": 3.34393063583815e-05,
+      "loss": 0.3864,
+      "step": 766
+    },
+    {
+      "epoch": 1.1948558067030397,
+      "grad_norm": 0.3475437886006249,
+      "learning_rate": 3.341040462427746e-05,
+      "loss": 0.3501,
+      "step": 767
+    },
+    {
+      "epoch": 1.1964146531566642,
+      "grad_norm": 0.27289955081057704,
+      "learning_rate": 3.338150289017341e-05,
+      "loss": 0.3738,
+      "step": 768
+    },
+    {
+      "epoch": 1.1979734996102884,
+      "grad_norm": 0.37715118746357845,
+      "learning_rate": 3.335260115606937e-05,
+      "loss": 0.3566,
+      "step": 769
+    },
+    {
+      "epoch": 1.1995323460639127,
+      "grad_norm": 0.30510592253569463,
+      "learning_rate": 3.3323699421965324e-05,
+      "loss": 0.3579,
+      "step": 770
+    },
+    {
+      "epoch": 1.201091192517537,
+      "grad_norm": 0.3705292948456225,
+      "learning_rate": 3.329479768786127e-05,
+      "loss": 0.3655,
+      "step": 771
+    },
+    {
+      "epoch": 1.2026500389711614,
+      "grad_norm": 0.32470886193713566,
+      "learning_rate": 3.3265895953757224e-05,
+      "loss": 0.3512,
+      "step": 772
+    },
+    {
+      "epoch": 1.2042088854247857,
+      "grad_norm": 0.3267984083578711,
+      "learning_rate": 3.323699421965318e-05,
+      "loss": 0.3559,
+      "step": 773
+    },
+    {
+      "epoch": 1.20576773187841,
+      "grad_norm": 0.316676369330289,
+      "learning_rate": 3.320809248554914e-05,
+      "loss": 0.3504,
+      "step": 774
+    },
+    {
+      "epoch": 1.2073265783320344,
+      "grad_norm": 0.3083068143586948,
+      "learning_rate": 3.317919075144509e-05,
+      "loss": 0.3497,
+      "step": 775
+    },
+    {
+      "epoch": 1.2088854247856586,
+      "grad_norm": 0.3112543283121555,
+      "learning_rate": 3.315028901734104e-05,
+      "loss": 0.3626,
+      "step": 776
+    },
+    {
+      "epoch": 1.2104442712392829,
+      "grad_norm": 0.2930169607949705,
+      "learning_rate": 3.312138728323699e-05,
+      "loss": 0.3767,
+      "step": 777
+    },
+    {
+      "epoch": 1.2120031176929071,
+      "grad_norm": 0.3740917898852477,
+      "learning_rate": 3.309248554913295e-05,
+      "loss": 0.3623,
+      "step": 778
+    },
+    {
+      "epoch": 1.2135619641465316,
+      "grad_norm": 0.2909666222558865,
+      "learning_rate": 3.3063583815028905e-05,
+      "loss": 0.3571,
+      "step": 779
+    },
+    {
+      "epoch": 1.2151208106001559,
+      "grad_norm": 0.3349084531099002,
+      "learning_rate": 3.303468208092486e-05,
+      "loss": 0.3809,
+      "step": 780
+    },
+    {
+      "epoch": 1.2166796570537801,
+      "grad_norm": 0.32395231786814505,
+      "learning_rate": 3.3005780346820805e-05,
+      "loss": 0.3701,
+      "step": 781
+    },
+    {
+      "epoch": 1.2182385035074046,
+      "grad_norm": 0.35641499162371804,
+      "learning_rate": 3.2976878612716765e-05,
+      "loss": 0.3767,
+      "step": 782
+    },
+    {
+      "epoch": 1.2197973499610288,
+      "grad_norm": 0.371466143629439,
+      "learning_rate": 3.294797687861272e-05,
+      "loss": 0.3716,
+      "step": 783
+    },
+    {
+      "epoch": 1.221356196414653,
+      "grad_norm": 0.3246192454526884,
+      "learning_rate": 3.291907514450867e-05,
+      "loss": 0.3742,
+      "step": 784
+    },
+    {
+      "epoch": 1.2229150428682776,
+      "grad_norm": 0.35868229918432093,
+      "learning_rate": 3.2890173410404626e-05,
+      "loss": 0.38,
+      "step": 785
+    },
+    {
+      "epoch": 1.2244738893219018,
+      "grad_norm": 0.353282741567031,
+      "learning_rate": 3.286127167630058e-05,
+      "loss": 0.3621,
+      "step": 786
+    },
+    {
+      "epoch": 1.226032735775526,
+      "grad_norm": 0.34468589401361616,
+      "learning_rate": 3.283236994219653e-05,
+      "loss": 0.3739,
+      "step": 787
+    },
+    {
+      "epoch": 1.2275915822291505,
+      "grad_norm": 0.2790582877348234,
+      "learning_rate": 3.2803468208092486e-05,
+      "loss": 0.3759,
+      "step": 788
+    },
+    {
+      "epoch": 1.2291504286827748,
+      "grad_norm": 0.455365887136131,
+      "learning_rate": 3.277456647398844e-05,
+      "loss": 0.371,
+      "step": 789
+    },
+    {
+      "epoch": 1.230709275136399,
+      "grad_norm": 0.32167141704971974,
+      "learning_rate": 3.27456647398844e-05,
+      "loss": 0.3728,
+      "step": 790
+    },
+    {
+      "epoch": 1.2322681215900233,
+      "grad_norm": 0.3938144396135692,
+      "learning_rate": 3.2716763005780346e-05,
+      "loss": 0.3613,
+      "step": 791
+    },
+    {
+      "epoch": 1.2338269680436478,
+      "grad_norm": 0.35340743429724525,
+      "learning_rate": 3.26878612716763e-05,
+      "loss": 0.362,
+      "step": 792
+    },
+    {
+      "epoch": 1.235385814497272,
+      "grad_norm": 0.3887479835449903,
+      "learning_rate": 3.265895953757225e-05,
+      "loss": 0.3534,
+      "step": 793
+    },
+    {
+      "epoch": 1.2369446609508963,
+      "grad_norm": 0.3810148179704415,
+      "learning_rate": 3.2630057803468213e-05,
+      "loss": 0.3708,
+      "step": 794
+    },
+    {
+      "epoch": 1.2385035074045208,
+      "grad_norm": 0.45308168447987623,
+      "learning_rate": 3.260115606936417e-05,
+      "loss": 0.363,
+      "step": 795
+    },
+    {
+      "epoch": 1.240062353858145,
+      "grad_norm": 0.3530892057764186,
+      "learning_rate": 3.2572254335260114e-05,
+      "loss": 0.3748,
+      "step": 796
+    },
+    {
+      "epoch": 1.2416212003117693,
+      "grad_norm": 0.4323443583689573,
+      "learning_rate": 3.254335260115607e-05,
+      "loss": 0.3727,
+      "step": 797
+    },
+    {
+      "epoch": 1.2431800467653935,
+      "grad_norm": 0.3213425990151126,
+      "learning_rate": 3.251445086705203e-05,
+      "loss": 0.391,
+      "step": 798
+    },
+    {
+      "epoch": 1.244738893219018,
+      "grad_norm": 0.33159050533895373,
+      "learning_rate": 3.248554913294798e-05,
+      "loss": 0.378,
+      "step": 799
+    },
+    {
+      "epoch": 1.2462977396726422,
+      "grad_norm": 0.35918958392112355,
+      "learning_rate": 3.2456647398843934e-05,
+      "loss": 0.372,
+      "step": 800
+    },
+    {
+      "epoch": 1.2478565861262665,
+      "grad_norm": 0.3065526765937574,
+      "learning_rate": 3.242774566473988e-05,
+      "loss": 0.3678,
+      "step": 801
+    },
+    {
+      "epoch": 1.249415432579891,
+      "grad_norm": 0.3162819465443684,
+      "learning_rate": 3.239884393063584e-05,
+      "loss": 0.3554,
+      "step": 802
+    },
+    {
+      "epoch": 1.2509742790335152,
+      "grad_norm": 0.3623636030150004,
+      "learning_rate": 3.2369942196531794e-05,
+      "loss": 0.3742,
+      "step": 803
+    },
+    {
+      "epoch": 1.2525331254871395,
+      "grad_norm": 0.2955704493514888,
+      "learning_rate": 3.234104046242775e-05,
+      "loss": 0.3704,
+      "step": 804
+    },
+    {
+      "epoch": 1.2540919719407637,
+      "grad_norm": 0.36701054340991607,
+      "learning_rate": 3.23121387283237e-05,
+      "loss": 0.3728,
+      "step": 805
+    },
+    {
+      "epoch": 1.2556508183943882,
+      "grad_norm": 0.3328636144558375,
+      "learning_rate": 3.2283236994219655e-05,
+      "loss": 0.3816,
+      "step": 806
+    },
+    {
+      "epoch": 1.2572096648480124,
+      "grad_norm": 0.3578865290981359,
+      "learning_rate": 3.225433526011561e-05,
+      "loss": 0.3667,
+      "step": 807
+    },
+    {
+      "epoch": 1.258768511301637,
+      "grad_norm": 0.3408336723453489,
+      "learning_rate": 3.222543352601156e-05,
+      "loss": 0.3817,
+      "step": 808
+    },
+    {
+      "epoch": 1.2603273577552612,
+      "grad_norm": 0.3216030123089057,
+      "learning_rate": 3.2196531791907515e-05,
+      "loss": 0.3691,
+      "step": 809
+    },
+    {
+      "epoch": 1.2618862042088854,
+      "grad_norm": 0.37724572418313224,
+      "learning_rate": 3.216763005780347e-05,
+      "loss": 0.3824,
+      "step": 810
+    },
+    {
+      "epoch": 1.2634450506625097,
+      "grad_norm": 0.3056707439719446,
+      "learning_rate": 3.213872832369942e-05,
+      "loss": 0.3718,
+      "step": 811
+    },
+    {
+      "epoch": 1.265003897116134,
+      "grad_norm": 0.3614010471408866,
+      "learning_rate": 3.2109826589595375e-05,
+      "loss": 0.378,
+      "step": 812
+    },
+    {
+      "epoch": 1.2665627435697584,
+      "grad_norm": 0.30433373829385135,
+      "learning_rate": 3.208092485549133e-05,
+      "loss": 0.4059,
+      "step": 813
+    },
+    {
+      "epoch": 1.2681215900233826,
+      "grad_norm": 0.3789679238319672,
+      "learning_rate": 3.205202312138729e-05,
+      "loss": 0.3739,
+      "step": 814
+    },
+    {
+      "epoch": 1.2696804364770071,
+      "grad_norm": 0.31090210614669855,
+      "learning_rate": 3.202312138728324e-05,
+      "loss": 0.3796,
+      "step": 815
+    },
+    {
+      "epoch": 1.2712392829306314,
+      "grad_norm": 0.3395673581732934,
+      "learning_rate": 3.199421965317919e-05,
+      "loss": 0.3842,
+      "step": 816
+    },
+    {
+      "epoch": 1.2727981293842556,
+      "grad_norm": 0.3814013963708837,
+      "learning_rate": 3.196531791907514e-05,
+      "loss": 0.3818,
+      "step": 817
+    },
+    {
+      "epoch": 1.2743569758378799,
+      "grad_norm": 0.2881312269831677,
+      "learning_rate": 3.19364161849711e-05,
+      "loss": 0.3755,
+      "step": 818
+    },
+    {
+      "epoch": 1.2759158222915044,
+      "grad_norm": 0.3386831635441485,
+      "learning_rate": 3.1907514450867056e-05,
+      "loss": 0.3666,
+      "step": 819
+    },
+    {
+      "epoch": 1.2774746687451286,
+      "grad_norm": 0.30697028027634937,
+      "learning_rate": 3.187861271676301e-05,
+      "loss": 0.3843,
+      "step": 820
+    },
+    {
+      "epoch": 1.2790335151987529,
+      "grad_norm": 0.29980913900639733,
+      "learning_rate": 3.1849710982658956e-05,
+      "loss": 0.3666,
+      "step": 821
+    },
+    {
+      "epoch": 1.2805923616523773,
+      "grad_norm": 0.3327607375144203,
+      "learning_rate": 3.182080924855492e-05,
+      "loss": 0.3858,
+      "step": 822
+    },
+    {
+      "epoch": 1.2821512081060016,
+      "grad_norm": 0.29432860827743634,
+      "learning_rate": 3.179190751445087e-05,
+      "loss": 0.3614,
+      "step": 823
+    },
+    {
+      "epoch": 1.2837100545596258,
+      "grad_norm": 0.3621120310764712,
+      "learning_rate": 3.1763005780346824e-05,
+      "loss": 0.3737,
+      "step": 824
+    },
+    {
+      "epoch": 1.28526890101325,
+      "grad_norm": 0.3445821791263814,
+      "learning_rate": 3.173410404624278e-05,
+      "loss": 0.3672,
+      "step": 825
+    },
+    {
+      "epoch": 1.2868277474668746,
+      "grad_norm": 0.2954327384154805,
+      "learning_rate": 3.170520231213873e-05,
+      "loss": 0.3607,
+      "step": 826
+    },
+    {
+      "epoch": 1.2883865939204988,
+      "grad_norm": 0.328451959411728,
+      "learning_rate": 3.1676300578034684e-05,
+      "loss": 0.3702,
+      "step": 827
+    },
+    {
+      "epoch": 1.2899454403741233,
+      "grad_norm": 0.2944126016476037,
+      "learning_rate": 3.164739884393064e-05,
+      "loss": 0.3889,
+      "step": 828
+    },
+    {
+      "epoch": 1.2915042868277475,
+      "grad_norm": 0.3969777724956226,
+      "learning_rate": 3.161849710982659e-05,
+      "loss": 0.3547,
+      "step": 829
+    },
+    {
+      "epoch": 1.2930631332813718,
+      "grad_norm": 0.2951235605441101,
+      "learning_rate": 3.1589595375722544e-05,
+      "loss": 0.378,
+      "step": 830
+    },
+    {
+      "epoch": 1.294621979734996,
+      "grad_norm": 0.28308239600029,
+      "learning_rate": 3.15606936416185e-05,
+      "loss": 0.3828,
+      "step": 831
+    },
+    {
+      "epoch": 1.2961808261886203,
+      "grad_norm": 0.3901401668691733,
+      "learning_rate": 3.153179190751445e-05,
+      "loss": 0.3842,
+      "step": 832
+    },
+    {
+      "epoch": 1.2977396726422448,
+      "grad_norm": 0.3137907326756719,
+      "learning_rate": 3.1502890173410405e-05,
+      "loss": 0.3517,
+      "step": 833
+    },
+    {
+      "epoch": 1.299298519095869,
+      "grad_norm": 0.30348942495737735,
+      "learning_rate": 3.1473988439306365e-05,
+      "loss": 0.3584,
+      "step": 834
+    },
+    {
+      "epoch": 1.3008573655494935,
+      "grad_norm": 0.3134452203812044,
+      "learning_rate": 3.144508670520231e-05,
+      "loss": 0.3668,
+      "step": 835
+    },
+    {
+      "epoch": 1.3024162120031177,
+      "grad_norm": 0.3330164303286052,
+      "learning_rate": 3.1416184971098265e-05,
+      "loss": 0.3791,
+      "step": 836
+    },
+    {
+      "epoch": 1.303975058456742,
+      "grad_norm": 0.3242958410056007,
+      "learning_rate": 3.138728323699422e-05,
+      "loss": 0.3644,
+      "step": 837
+    },
+    {
+      "epoch": 1.3055339049103662,
+      "grad_norm": 0.3372448607214704,
+      "learning_rate": 3.135838150289018e-05,
+      "loss": 0.3755,
+      "step": 838
+    },
+    {
+      "epoch": 1.3070927513639907,
+      "grad_norm": 0.34805951436795246,
+      "learning_rate": 3.132947976878613e-05,
+      "loss": 0.4029,
+      "step": 839
+    },
+    {
+      "epoch": 1.308651597817615,
+      "grad_norm": 0.3375892298162743,
+      "learning_rate": 3.130057803468208e-05,
+      "loss": 0.3688,
+      "step": 840
+    },
+    {
+      "epoch": 1.3102104442712392,
+      "grad_norm": 0.3427372693638577,
+      "learning_rate": 3.127167630057803e-05,
+      "loss": 0.3583,
+      "step": 841
+    },
+    {
+      "epoch": 1.3117692907248637,
+      "grad_norm": 0.39528017078512534,
+      "learning_rate": 3.124277456647399e-05,
+      "loss": 0.3815,
+      "step": 842
+    },
+    {
+      "epoch": 1.313328137178488,
+      "grad_norm": 0.300797549857878,
+      "learning_rate": 3.1213872832369946e-05,
+      "loss": 0.3789,
+      "step": 843
+    },
+    {
+      "epoch": 1.3148869836321122,
+      "grad_norm": 0.37253735153697626,
+      "learning_rate": 3.11849710982659e-05,
+      "loss": 0.38,
+      "step": 844
+    },
+    {
+      "epoch": 1.3164458300857365,
+      "grad_norm": 0.3636396680715121,
+      "learning_rate": 3.115606936416185e-05,
+      "loss": 0.3584,
+      "step": 845
+    },
+    {
+      "epoch": 1.318004676539361,
+      "grad_norm": 0.3225603467896192,
+      "learning_rate": 3.1127167630057806e-05,
+      "loss": 0.3801,
+      "step": 846
+    },
+    {
+      "epoch": 1.3195635229929852,
+      "grad_norm": 0.3923444106986942,
+      "learning_rate": 3.109826589595376e-05,
+      "loss": 0.3838,
+      "step": 847
+    },
+    {
+      "epoch": 1.3211223694466094,
+      "grad_norm": 0.29792571705489457,
+      "learning_rate": 3.106936416184971e-05,
+      "loss": 0.3877,
+      "step": 848
+    },
+    {
+      "epoch": 1.322681215900234,
+      "grad_norm": 0.41101804688404436,
+      "learning_rate": 3.1040462427745667e-05,
+      "loss": 0.3711,
+      "step": 849
+    },
+    {
+      "epoch": 1.3242400623538582,
+      "grad_norm": 0.31527820005688345,
+      "learning_rate": 3.101156069364162e-05,
+      "loss": 0.3672,
+      "step": 850
+    },
+    {
+      "epoch": 1.3257989088074824,
+      "grad_norm": 0.3247891491953755,
+      "learning_rate": 3.0982658959537573e-05,
+      "loss": 0.3875,
+      "step": 851
+    },
+    {
+      "epoch": 1.3273577552611067,
+      "grad_norm": 0.31678580564634845,
+      "learning_rate": 3.095375722543353e-05,
+      "loss": 0.3767,
+      "step": 852
+    },
+    {
+      "epoch": 1.3289166017147311,
+      "grad_norm": 0.42289684401021704,
+      "learning_rate": 3.092485549132948e-05,
+      "loss": 0.3896,
+      "step": 853
+    },
+    {
+      "epoch": 1.3304754481683554,
+      "grad_norm": 0.34344843446611606,
+      "learning_rate": 3.089595375722544e-05,
+      "loss": 0.3837,
+      "step": 854
+    },
+    {
+      "epoch": 1.3320342946219799,
+      "grad_norm": 0.46291047151845677,
+      "learning_rate": 3.086705202312139e-05,
+      "loss": 0.3757,
+      "step": 855
+    },
+    {
+      "epoch": 1.3335931410756041,
+      "grad_norm": 0.30760449980458326,
+      "learning_rate": 3.083815028901734e-05,
+      "loss": 0.3784,
+      "step": 856
+    },
+    {
+      "epoch": 1.3351519875292284,
+      "grad_norm": 0.36618124968476307,
+      "learning_rate": 3.0809248554913294e-05,
+      "loss": 0.3646,
+      "step": 857
+    },
+    {
+      "epoch": 1.3367108339828526,
+      "grad_norm": 0.3517633452852137,
+      "learning_rate": 3.0780346820809254e-05,
+      "loss": 0.3678,
+      "step": 858
+    },
+    {
+      "epoch": 1.3382696804364769,
+      "grad_norm": 0.36650573052741386,
+      "learning_rate": 3.075144508670521e-05,
+      "loss": 0.3666,
+      "step": 859
+    },
+    {
+      "epoch": 1.3398285268901013,
+      "grad_norm": 0.32309525776364156,
+      "learning_rate": 3.0722543352601154e-05,
+      "loss": 0.3596,
+      "step": 860
+    },
+    {
+      "epoch": 1.3413873733437256,
+      "grad_norm": 0.2918364251771976,
+      "learning_rate": 3.069364161849711e-05,
+      "loss": 0.363,
+      "step": 861
+    },
+    {
+      "epoch": 1.34294621979735,
+      "grad_norm": 0.34546628975016486,
+      "learning_rate": 3.066473988439307e-05,
+      "loss": 0.3553,
+      "step": 862
+    },
+    {
+      "epoch": 1.3445050662509743,
+      "grad_norm": 0.2923967277989431,
+      "learning_rate": 3.063583815028902e-05,
+      "loss": 0.3489,
+      "step": 863
+    },
+    {
+      "epoch": 1.3460639127045986,
+      "grad_norm": 0.29914296381783523,
+      "learning_rate": 3.0606936416184975e-05,
+      "loss": 0.3787,
+      "step": 864
+    },
+    {
+      "epoch": 1.3476227591582228,
+      "grad_norm": 0.31943444292620105,
+      "learning_rate": 3.057803468208092e-05,
+      "loss": 0.3846,
+      "step": 865
+    },
+    {
+      "epoch": 1.3491816056118473,
+      "grad_norm": 0.29039633067671383,
+      "learning_rate": 3.0549132947976875e-05,
+      "loss": 0.3639,
+      "step": 866
+    },
+    {
+      "epoch": 1.3507404520654716,
+      "grad_norm": 0.3421893491096786,
+      "learning_rate": 3.0520231213872835e-05,
+      "loss": 0.3734,
+      "step": 867
+    },
+    {
+      "epoch": 1.3522992985190958,
+      "grad_norm": 0.3170017953585186,
+      "learning_rate": 3.049132947976879e-05,
+      "loss": 0.369,
+      "step": 868
+    },
+    {
+      "epoch": 1.3538581449727203,
+      "grad_norm": 0.31067447367056183,
+      "learning_rate": 3.046242774566474e-05,
+      "loss": 0.3879,
+      "step": 869
+    },
+    {
+      "epoch": 1.3554169914263445,
+      "grad_norm": 0.3212256602786557,
+      "learning_rate": 3.0433526011560692e-05,
+      "loss": 0.3805,
+      "step": 870
+    },
+    {
+      "epoch": 1.3569758378799688,
+      "grad_norm": 0.32052080705580777,
+      "learning_rate": 3.040462427745665e-05,
+      "loss": 0.3879,
+      "step": 871
+    },
+    {
+      "epoch": 1.358534684333593,
+      "grad_norm": 0.27323028010679573,
+      "learning_rate": 3.0375722543352603e-05,
+      "loss": 0.3753,
+      "step": 872
+    },
+    {
+      "epoch": 1.3600935307872175,
+      "grad_norm": 0.31472694270289026,
+      "learning_rate": 3.0346820809248556e-05,
+      "loss": 0.3676,
+      "step": 873
+    },
+    {
+      "epoch": 1.3616523772408418,
+      "grad_norm": 0.2817828326023501,
+      "learning_rate": 3.0317919075144506e-05,
+      "loss": 0.3602,
+      "step": 874
+    },
+    {
+      "epoch": 1.363211223694466,
+      "grad_norm": 0.31483550680330125,
+      "learning_rate": 3.0289017341040466e-05,
+      "loss": 0.3801,
+      "step": 875
+    },
+    {
+      "epoch": 1.3647700701480905,
+      "grad_norm": 0.31361863279521,
+      "learning_rate": 3.0260115606936416e-05,
+      "loss": 0.3574,
+      "step": 876
+    },
+    {
+      "epoch": 1.3663289166017147,
+      "grad_norm": 0.3795186696828694,
+      "learning_rate": 3.023121387283237e-05,
+      "loss": 0.3608,
+      "step": 877
+    },
+    {
+      "epoch": 1.367887763055339,
+      "grad_norm": 0.3024681697597954,
+      "learning_rate": 3.0202312138728323e-05,
+      "loss": 0.3597,
+      "step": 878
+    },
+    {
+      "epoch": 1.3694466095089632,
+      "grad_norm": 0.2839484436314578,
+      "learning_rate": 3.017341040462428e-05,
+      "loss": 0.354,
+      "step": 879
+    },
+    {
+      "epoch": 1.3710054559625877,
+      "grad_norm": 0.2989185656953038,
+      "learning_rate": 3.0144508670520234e-05,
+      "loss": 0.3997,
+      "step": 880
+    },
+    {
+      "epoch": 1.372564302416212,
+      "grad_norm": 0.3332125838479124,
+      "learning_rate": 3.0115606936416184e-05,
+      "loss": 0.3866,
+      "step": 881
+    },
+    {
+      "epoch": 1.3741231488698364,
+      "grad_norm": 0.32196595663758937,
+      "learning_rate": 3.0086705202312137e-05,
+      "loss": 0.3703,
+      "step": 882
+    },
+    {
+      "epoch": 1.3756819953234607,
+      "grad_norm": 0.3218885635026921,
+      "learning_rate": 3.0057803468208097e-05,
+      "loss": 0.3644,
+      "step": 883
+    },
+    {
+      "epoch": 1.377240841777085,
+      "grad_norm": 0.3653050901483605,
+      "learning_rate": 3.0028901734104047e-05,
+      "loss": 0.3589,
+      "step": 884
+    },
+    {
+      "epoch": 1.3787996882307092,
+      "grad_norm": 0.3686597692797899,
+      "learning_rate": 3e-05,
+      "loss": 0.3691,
+      "step": 885
+    },
+    {
+      "epoch": 1.3803585346843337,
+      "grad_norm": 0.31536889278803354,
+      "learning_rate": 2.9971098265895954e-05,
+      "loss": 0.3719,
+      "step": 886
+    },
+    {
+      "epoch": 1.381917381137958,
+      "grad_norm": 0.33299163893566386,
+      "learning_rate": 2.994219653179191e-05,
+      "loss": 0.376,
+      "step": 887
+    },
+    {
+      "epoch": 1.3834762275915822,
+      "grad_norm": 0.3637154017248889,
+      "learning_rate": 2.9913294797687864e-05,
+      "loss": 0.3763,
+      "step": 888
+    },
+    {
+      "epoch": 1.3850350740452066,
+      "grad_norm": 0.31459018536275674,
+      "learning_rate": 2.9884393063583815e-05,
+      "loss": 0.3719,
+      "step": 889
+    },
+    {
+      "epoch": 1.386593920498831,
+      "grad_norm": 0.38863024038763483,
+      "learning_rate": 2.9855491329479768e-05,
+      "loss": 0.3897,
+      "step": 890
+    },
+    {
+      "epoch": 1.3881527669524552,
+      "grad_norm": 0.2713739072144362,
+      "learning_rate": 2.9826589595375725e-05,
+      "loss": 0.3774,
+      "step": 891
+    },
+    {
+      "epoch": 1.3897116134060794,
+      "grad_norm": 0.3161296579386695,
+      "learning_rate": 2.9797687861271678e-05,
+      "loss": 0.367,
+      "step": 892
+    },
+    {
+      "epoch": 1.3912704598597039,
+      "grad_norm": 0.32321623162651764,
+      "learning_rate": 2.9768786127167632e-05,
+      "loss": 0.3908,
+      "step": 893
+    },
+    {
+      "epoch": 1.3928293063133281,
+      "grad_norm": 0.28969209370393095,
+      "learning_rate": 2.9739884393063582e-05,
+      "loss": 0.3715,
+      "step": 894
+    },
+    {
+      "epoch": 1.3943881527669524,
+      "grad_norm": 0.35841207954661103,
+      "learning_rate": 2.9710982658959542e-05,
+      "loss": 0.3714,
+      "step": 895
+    },
+    {
+      "epoch": 1.3959469992205769,
+      "grad_norm": 0.3324679248139433,
+      "learning_rate": 2.9682080924855492e-05,
+      "loss": 0.3717,
+      "step": 896
+    },
+    {
+      "epoch": 1.397505845674201,
+      "grad_norm": 0.43923218037881906,
+      "learning_rate": 2.9653179190751446e-05,
+      "loss": 0.3619,
+      "step": 897
+    },
+    {
+      "epoch": 1.3990646921278254,
+      "grad_norm": 0.31489100441245577,
+      "learning_rate": 2.96242774566474e-05,
+      "loss": 0.3523,
+      "step": 898
+    },
+    {
+      "epoch": 1.4006235385814496,
+      "grad_norm": 0.3292376441511829,
+      "learning_rate": 2.9595375722543356e-05,
+      "loss": 0.3661,
+      "step": 899
+    },
+    {
+      "epoch": 1.402182385035074,
+      "grad_norm": 0.2988144575593582,
+      "learning_rate": 2.956647398843931e-05,
+      "loss": 0.3749,
+      "step": 900
+    },
+    {
+      "epoch": 1.4037412314886983,
+      "grad_norm": 0.41020723022893746,
+      "learning_rate": 2.953757225433526e-05,
+      "loss": 0.366,
+      "step": 901
+    },
+    {
+      "epoch": 1.4053000779423228,
+      "grad_norm": 0.310023862219356,
+      "learning_rate": 2.9508670520231213e-05,
+      "loss": 0.3884,
+      "step": 902
+    },
+    {
+      "epoch": 1.406858924395947,
+      "grad_norm": 0.3454601431005727,
+      "learning_rate": 2.947976878612717e-05,
+      "loss": 0.3634,
+      "step": 903
+    },
+    {
+      "epoch": 1.4084177708495713,
+      "grad_norm": 0.34356963535549423,
+      "learning_rate": 2.9450867052023123e-05,
+      "loss": 0.3588,
+      "step": 904
+    },
+    {
+      "epoch": 1.4099766173031956,
+      "grad_norm": 0.3283997995226211,
+      "learning_rate": 2.9421965317919076e-05,
+      "loss": 0.3637,
+      "step": 905
+    },
+    {
+      "epoch": 1.4115354637568198,
+      "grad_norm": 0.3554228840104719,
+      "learning_rate": 2.9393063583815027e-05,
+      "loss": 0.343,
+      "step": 906
+    },
+    {
+      "epoch": 1.4130943102104443,
+      "grad_norm": 0.36071645685077747,
+      "learning_rate": 2.9364161849710987e-05,
+      "loss": 0.3596,
+      "step": 907
+    },
+    {
+      "epoch": 1.4146531566640685,
+      "grad_norm": 0.3112359765865261,
+      "learning_rate": 2.9335260115606937e-05,
+      "loss": 0.3526,
+      "step": 908
+    },
+    {
+      "epoch": 1.416212003117693,
+      "grad_norm": 0.34359020030558024,
+      "learning_rate": 2.930635838150289e-05,
+      "loss": 0.3844,
+      "step": 909
+    },
+    {
+      "epoch": 1.4177708495713173,
+      "grad_norm": 0.3008878572757736,
+      "learning_rate": 2.9277456647398844e-05,
+      "loss": 0.3772,
+      "step": 910
+    },
+    {
+      "epoch": 1.4193296960249415,
+      "grad_norm": 0.349466144708206,
+      "learning_rate": 2.92485549132948e-05,
+      "loss": 0.3473,
+      "step": 911
+    },
+    {
+      "epoch": 1.4208885424785658,
+      "grad_norm": 0.3382202790340046,
+      "learning_rate": 2.9219653179190754e-05,
+      "loss": 0.3757,
+      "step": 912
+    },
+    {
+      "epoch": 1.4224473889321902,
+      "grad_norm": 0.32188902141101255,
+      "learning_rate": 2.9190751445086707e-05,
+      "loss": 0.3712,
+      "step": 913
+    },
+    {
+      "epoch": 1.4240062353858145,
+      "grad_norm": 0.29754261721596836,
+      "learning_rate": 2.9161849710982657e-05,
+      "loss": 0.3674,
+      "step": 914
+    },
+    {
+      "epoch": 1.4255650818394388,
+      "grad_norm": 0.41389196935864453,
+      "learning_rate": 2.9132947976878618e-05,
+      "loss": 0.3835,
+      "step": 915
+    },
+    {
+      "epoch": 1.4271239282930632,
+      "grad_norm": 0.2943029270951012,
+      "learning_rate": 2.9104046242774568e-05,
+      "loss": 0.359,
+      "step": 916
+    },
+    {
+      "epoch": 1.4286827747466875,
+      "grad_norm": 0.28723479498577376,
+      "learning_rate": 2.907514450867052e-05,
+      "loss": 0.3695,
+      "step": 917
+    },
+    {
+      "epoch": 1.4302416212003117,
+      "grad_norm": 0.35137358020533505,
+      "learning_rate": 2.9046242774566475e-05,
+      "loss": 0.3626,
+      "step": 918
+    },
+    {
+      "epoch": 1.431800467653936,
+      "grad_norm": 0.3198068865556259,
+      "learning_rate": 2.901734104046243e-05,
+      "loss": 0.3855,
+      "step": 919
+    },
+    {
+      "epoch": 1.4333593141075605,
+      "grad_norm": 0.30362892326696816,
+      "learning_rate": 2.8988439306358385e-05,
+      "loss": 0.3882,
+      "step": 920
+    },
+    {
+      "epoch": 1.4349181605611847,
+      "grad_norm": 0.3191735883892925,
+      "learning_rate": 2.8959537572254335e-05,
+      "loss": 0.381,
+      "step": 921
+    },
+    {
+      "epoch": 1.436477007014809,
+      "grad_norm": 0.3034703934282239,
+      "learning_rate": 2.893063583815029e-05,
+      "loss": 0.3756,
+      "step": 922
+    },
+    {
+      "epoch": 1.4380358534684334,
+      "grad_norm": 0.32215086435520635,
+      "learning_rate": 2.8901734104046245e-05,
+      "loss": 0.3505,
+      "step": 923
+    },
+    {
+      "epoch": 1.4395946999220577,
+      "grad_norm": 0.3038326438989388,
+      "learning_rate": 2.88728323699422e-05,
+      "loss": 0.3578,
+      "step": 924
+    },
+    {
+      "epoch": 1.441153546375682,
+      "grad_norm": 0.32683148429237635,
+      "learning_rate": 2.8843930635838152e-05,
+      "loss": 0.358,
+      "step": 925
+    },
+    {
+      "epoch": 1.4427123928293062,
+      "grad_norm": 0.3682668995271643,
+      "learning_rate": 2.8815028901734102e-05,
+      "loss": 0.3693,
+      "step": 926
+    },
+    {
+      "epoch": 1.4442712392829307,
+      "grad_norm": 0.2743528467933646,
+      "learning_rate": 2.8786127167630062e-05,
+      "loss": 0.366,
+      "step": 927
+    },
+    {
+      "epoch": 1.445830085736555,
+      "grad_norm": 0.3966807244623201,
+      "learning_rate": 2.8757225433526013e-05,
+      "loss": 0.386,
+      "step": 928
+    },
+    {
+      "epoch": 1.4473889321901794,
+      "grad_norm": 0.32237328497617107,
+      "learning_rate": 2.8728323699421966e-05,
+      "loss": 0.3659,
+      "step": 929
+    },
+    {
+      "epoch": 1.4489477786438036,
+      "grad_norm": 0.31054470894310937,
+      "learning_rate": 2.869942196531792e-05,
+      "loss": 0.3512,
+      "step": 930
+    },
+    {
+      "epoch": 1.450506625097428,
+      "grad_norm": 0.43587348795942593,
+      "learning_rate": 2.8670520231213876e-05,
+      "loss": 0.3673,
+      "step": 931
+    },
+    {
+      "epoch": 1.4520654715510521,
+      "grad_norm": 0.2889720776490319,
+      "learning_rate": 2.864161849710983e-05,
+      "loss": 0.3834,
+      "step": 932
+    },
+    {
+      "epoch": 1.4536243180046766,
+      "grad_norm": 0.41243764294739627,
+      "learning_rate": 2.861271676300578e-05,
+      "loss": 0.3718,
+      "step": 933
+    },
+    {
+      "epoch": 1.4551831644583009,
+      "grad_norm": 0.3609717686516069,
+      "learning_rate": 2.8583815028901733e-05,
+      "loss": 0.3666,
+      "step": 934
+    },
+    {
+      "epoch": 1.4567420109119251,
+      "grad_norm": 0.3210455895295045,
+      "learning_rate": 2.855491329479769e-05,
+      "loss": 0.3512,
+      "step": 935
+    },
+    {
+      "epoch": 1.4583008573655496,
+      "grad_norm": 0.36374442891877706,
+      "learning_rate": 2.8526011560693643e-05,
+      "loss": 0.3789,
+      "step": 936
+    },
+    {
+      "epoch": 1.4598597038191738,
+      "grad_norm": 0.302661164884665,
+      "learning_rate": 2.8497109826589597e-05,
+      "loss": 0.3849,
+      "step": 937
+    },
+    {
+      "epoch": 1.461418550272798,
+      "grad_norm": 0.2922986064549059,
+      "learning_rate": 2.846820809248555e-05,
+      "loss": 0.3799,
+      "step": 938
+    },
+    {
+      "epoch": 1.4629773967264224,
+      "grad_norm": 0.3106937568151618,
+      "learning_rate": 2.8439306358381507e-05,
+      "loss": 0.3934,
+      "step": 939
+    },
+    {
+      "epoch": 1.4645362431800468,
+      "grad_norm": 0.3894062278825112,
+      "learning_rate": 2.841040462427746e-05,
+      "loss": 0.366,
+      "step": 940
+    },
+    {
+      "epoch": 1.466095089633671,
+      "grad_norm": 0.3094641338569837,
+      "learning_rate": 2.838150289017341e-05,
+      "loss": 0.3667,
+      "step": 941
+    },
+    {
+      "epoch": 1.4676539360872953,
+      "grad_norm": 0.2993216786356107,
+      "learning_rate": 2.8352601156069364e-05,
+      "loss": 0.3778,
+      "step": 942
+    },
+    {
+      "epoch": 1.4692127825409198,
+      "grad_norm": 0.290935367620464,
+      "learning_rate": 2.832369942196532e-05,
+      "loss": 0.3705,
+      "step": 943
+    },
+    {
+      "epoch": 1.470771628994544,
+      "grad_norm": 0.3157318710559096,
+      "learning_rate": 2.8294797687861274e-05,
+      "loss": 0.3595,
+      "step": 944
+    },
+    {
+      "epoch": 1.4723304754481683,
+      "grad_norm": 0.2890101359777515,
+      "learning_rate": 2.8265895953757228e-05,
+      "loss": 0.3773,
+      "step": 945
+    },
+    {
+      "epoch": 1.4738893219017926,
+      "grad_norm": 0.3502615396853657,
+      "learning_rate": 2.8236994219653178e-05,
+      "loss": 0.3605,
+      "step": 946
+    },
+    {
+      "epoch": 1.475448168355417,
+      "grad_norm": 0.34510628572007923,
+      "learning_rate": 2.8208092485549138e-05,
+      "loss": 0.3567,
+      "step": 947
+    },
+    {
+      "epoch": 1.4770070148090413,
+      "grad_norm": 0.30987576563699276,
+      "learning_rate": 2.8179190751445088e-05,
+      "loss": 0.3876,
+      "step": 948
+    },
+    {
+      "epoch": 1.4785658612626658,
+      "grad_norm": 0.38023748103458505,
+      "learning_rate": 2.815028901734104e-05,
+      "loss": 0.3692,
+      "step": 949
+    },
+    {
+      "epoch": 1.48012470771629,
+      "grad_norm": 0.30194848904733007,
+      "learning_rate": 2.8121387283236995e-05,
+      "loss": 0.3639,
+      "step": 950
+    },
+    {
+      "epoch": 1.4816835541699143,
+      "grad_norm": 0.36247398033072276,
+      "learning_rate": 2.8092485549132952e-05,
+      "loss": 0.3647,
+      "step": 951
+    },
+    {
+      "epoch": 1.4832424006235385,
+      "grad_norm": 0.3225418884761915,
+      "learning_rate": 2.8063583815028905e-05,
+      "loss": 0.3728,
+      "step": 952
+    },
+    {
+      "epoch": 1.4848012470771628,
+      "grad_norm": 0.33143640861599943,
+      "learning_rate": 2.8034682080924855e-05,
+      "loss": 0.3786,
+      "step": 953
+    },
+    {
+      "epoch": 1.4863600935307872,
+      "grad_norm": 0.37859874084366085,
+      "learning_rate": 2.800578034682081e-05,
+      "loss": 0.3676,
+      "step": 954
+    },
+    {
+      "epoch": 1.4879189399844115,
+      "grad_norm": 0.376572202242789,
+      "learning_rate": 2.7976878612716766e-05,
+      "loss": 0.3741,
+      "step": 955
+    },
+    {
+      "epoch": 1.489477786438036,
+      "grad_norm": 0.33305139704253905,
+      "learning_rate": 2.794797687861272e-05,
+      "loss": 0.3676,
+      "step": 956
+    },
+    {
+      "epoch": 1.4910366328916602,
+      "grad_norm": 0.5151154703791708,
+      "learning_rate": 2.7919075144508673e-05,
+      "loss": 0.3712,
+      "step": 957
+    },
+    {
+      "epoch": 1.4925954793452845,
+      "grad_norm": 0.289014937897802,
+      "learning_rate": 2.7890173410404623e-05,
+      "loss": 0.3522,
+      "step": 958
+    },
+    {
+      "epoch": 1.4941543257989087,
+      "grad_norm": 0.4050759220300978,
+      "learning_rate": 2.7861271676300583e-05,
+      "loss": 0.3695,
+      "step": 959
+    },
+    {
+      "epoch": 1.4957131722525332,
+      "grad_norm": 0.4162682771540926,
+      "learning_rate": 2.7832369942196533e-05,
+      "loss": 0.3864,
+      "step": 960
+    },
+    {
+      "epoch": 1.4972720187061574,
+      "grad_norm": 0.3267115752346114,
+      "learning_rate": 2.7803468208092486e-05,
+      "loss": 0.398,
+      "step": 961
+    },
+    {
+      "epoch": 1.4988308651597817,
+      "grad_norm": 0.46629831946057043,
+      "learning_rate": 2.777456647398844e-05,
+      "loss": 0.3667,
+      "step": 962
+    },
+    {
+      "epoch": 1.5003897116134062,
+      "grad_norm": 0.3126837897461622,
+      "learning_rate": 2.7745664739884393e-05,
+      "loss": 0.3903,
+      "step": 963
+    },
+    {
+      "epoch": 1.5019485580670304,
+      "grad_norm": 0.3025185534818139,
+      "learning_rate": 2.771676300578035e-05,
+      "loss": 0.4,
+      "step": 964
+    },
+    {
+      "epoch": 1.5035074045206547,
+      "grad_norm": 0.4008151783936587,
+      "learning_rate": 2.7687861271676304e-05,
+      "loss": 0.3839,
+      "step": 965
+    },
+    {
+      "epoch": 1.505066250974279,
+      "grad_norm": 0.26063162497897896,
+      "learning_rate": 2.7658959537572254e-05,
+      "loss": 0.361,
+      "step": 966
+    },
+    {
+      "epoch": 1.5066250974279034,
+      "grad_norm": 0.2888143269712675,
+      "learning_rate": 2.7630057803468207e-05,
+      "loss": 0.365,
+      "step": 967
+    },
+    {
+      "epoch": 1.5081839438815277,
+      "grad_norm": 0.3227561551129935,
+      "learning_rate": 2.7601156069364164e-05,
+      "loss": 0.3867,
+      "step": 968
+    },
+    {
+      "epoch": 1.5097427903351521,
+      "grad_norm": 0.29415161701639214,
+      "learning_rate": 2.7572254335260117e-05,
+      "loss": 0.3766,
+      "step": 969
+    },
+    {
+      "epoch": 1.5113016367887764,
+      "grad_norm": 0.30622334937549417,
+      "learning_rate": 2.754335260115607e-05,
+      "loss": 0.362,
+      "step": 970
+    },
+    {
+      "epoch": 1.5128604832424006,
+      "grad_norm": 0.29765107975268507,
+      "learning_rate": 2.751445086705202e-05,
+      "loss": 0.3691,
+      "step": 971
+    },
+    {
+      "epoch": 1.5144193296960249,
+      "grad_norm": 0.28887375199419957,
+      "learning_rate": 2.748554913294798e-05,
+      "loss": 0.376,
+      "step": 972
+    },
+    {
+      "epoch": 1.5159781761496491,
+      "grad_norm": 0.30491806233389507,
+      "learning_rate": 2.745664739884393e-05,
+      "loss": 0.3767,
+      "step": 973
+    },
+    {
+      "epoch": 1.5175370226032736,
+      "grad_norm": 0.2645879031934002,
+      "learning_rate": 2.7427745664739885e-05,
+      "loss": 0.3812,
+      "step": 974
+    },
+    {
+      "epoch": 1.5190958690568979,
+      "grad_norm": 0.28565659154478995,
+      "learning_rate": 2.7398843930635838e-05,
+      "loss": 0.3785,
+      "step": 975
+    },
+    {
+      "epoch": 1.5206547155105223,
+      "grad_norm": 0.28811620089387,
+      "learning_rate": 2.7369942196531795e-05,
+      "loss": 0.3537,
+      "step": 976
+    },
+    {
+      "epoch": 1.5222135619641466,
+      "grad_norm": 0.2577090443673922,
+      "learning_rate": 2.734104046242775e-05,
+      "loss": 0.355,
+      "step": 977
+    },
+    {
+      "epoch": 1.5237724084177708,
+      "grad_norm": 0.2624546459669209,
+      "learning_rate": 2.73121387283237e-05,
+      "loss": 0.3665,
+      "step": 978
+    },
+    {
+      "epoch": 1.525331254871395,
+      "grad_norm": 0.29221496113302586,
+      "learning_rate": 2.7283236994219652e-05,
+      "loss": 0.3683,
+      "step": 979
+    },
+    {
+      "epoch": 1.5268901013250193,
+      "grad_norm": 0.2680361810204351,
+      "learning_rate": 2.725433526011561e-05,
+      "loss": 0.3699,
+      "step": 980
+    },
+    {
+      "epoch": 1.5284489477786438,
+      "grad_norm": 0.3207752203807436,
+      "learning_rate": 2.7225433526011562e-05,
+      "loss": 0.3741,
+      "step": 981
+    },
+    {
+      "epoch": 1.530007794232268,
+      "grad_norm": 0.282031401318972,
+      "learning_rate": 2.7196531791907516e-05,
+      "loss": 0.3853,
+      "step": 982
+    },
+    {
+      "epoch": 1.5315666406858925,
+      "grad_norm": 0.27982574114644004,
+      "learning_rate": 2.7167630057803466e-05,
+      "loss": 0.3832,
+      "step": 983
+    },
+    {
+      "epoch": 1.5331254871395168,
+      "grad_norm": 0.2804241252514004,
+      "learning_rate": 2.7138728323699426e-05,
+      "loss": 0.3593,
+      "step": 984
+    },
+    {
+      "epoch": 1.534684333593141,
+      "grad_norm": 0.2628827070596599,
+      "learning_rate": 2.7109826589595376e-05,
+      "loss": 0.3669,
+      "step": 985
+    },
+    {
+      "epoch": 1.5362431800467653,
+      "grad_norm": 0.29615449790999526,
+      "learning_rate": 2.708092485549133e-05,
+      "loss": 0.3559,
+      "step": 986
+    },
+    {
+      "epoch": 1.5378020265003896,
+      "grad_norm": 0.28549992629405135,
+      "learning_rate": 2.7052023121387283e-05,
+      "loss": 0.3859,
+      "step": 987
+    },
+    {
+      "epoch": 1.539360872954014,
+      "grad_norm": 0.34007692401926715,
+      "learning_rate": 2.702312138728324e-05,
+      "loss": 0.3715,
+      "step": 988
+    },
+    {
+      "epoch": 1.5409197194076385,
+      "grad_norm": 0.40294192708679744,
+      "learning_rate": 2.6994219653179193e-05,
+      "loss": 0.3821,
+      "step": 989
+    },
+    {
+      "epoch": 1.5424785658612628,
+      "grad_norm": 0.265218381447006,
+      "learning_rate": 2.6965317919075143e-05,
+      "loss": 0.369,
+      "step": 990
+    },
+    {
+      "epoch": 1.544037412314887,
+      "grad_norm": 0.37011390060826654,
+      "learning_rate": 2.6936416184971097e-05,
+      "loss": 0.3643,
+      "step": 991
+    },
+    {
+      "epoch": 1.5455962587685113,
+      "grad_norm": 0.31157711298908103,
+      "learning_rate": 2.6907514450867057e-05,
+      "loss": 0.3688,
+      "step": 992
+    },
+    {
+      "epoch": 1.5471551052221355,
+      "grad_norm": 0.30367205453380675,
+      "learning_rate": 2.6878612716763007e-05,
+      "loss": 0.362,
+      "step": 993
+    },
+    {
+      "epoch": 1.54871395167576,
+      "grad_norm": 0.40306120286710007,
+      "learning_rate": 2.684971098265896e-05,
+      "loss": 0.3679,
+      "step": 994
+    },
+    {
+      "epoch": 1.5502727981293842,
+      "grad_norm": 0.3044105968804601,
+      "learning_rate": 2.6820809248554914e-05,
+      "loss": 0.3719,
+      "step": 995
+    },
+    {
+      "epoch": 1.5518316445830087,
+      "grad_norm": 0.3384424306040368,
+      "learning_rate": 2.679190751445087e-05,
+      "loss": 0.3669,
+      "step": 996
+    },
+    {
+      "epoch": 1.553390491036633,
+      "grad_norm": 0.2879848276257589,
+      "learning_rate": 2.6763005780346824e-05,
+      "loss": 0.3698,
+      "step": 997
+    },
+    {
+      "epoch": 1.5549493374902572,
+      "grad_norm": 0.3215450830190318,
+      "learning_rate": 2.6734104046242774e-05,
+      "loss": 0.3687,
+      "step": 998
+    },
+    {
+      "epoch": 1.5565081839438815,
+      "grad_norm": 0.2820349630596858,
+      "learning_rate": 2.6705202312138728e-05,
+      "loss": 0.3665,
+      "step": 999
+    },
+    {
+      "epoch": 1.5580670303975057,
+      "grad_norm": 0.3606469331809056,
+      "learning_rate": 2.6676300578034684e-05,
+      "loss": 0.351,
+      "step": 1000
+    },
+    {
+      "epoch": 1.5596258768511302,
+      "grad_norm": 0.3028843154170325,
+      "learning_rate": 2.6647398843930638e-05,
+      "loss": 0.3736,
+      "step": 1001
+    },
+    {
+      "epoch": 1.5611847233047544,
+      "grad_norm": 0.3174824654872977,
+      "learning_rate": 2.661849710982659e-05,
+      "loss": 0.3629,
+      "step": 1002
+    },
+    {
+      "epoch": 1.562743569758379,
+      "grad_norm": 0.34721718744253455,
+      "learning_rate": 2.658959537572254e-05,
+      "loss": 0.379,
+      "step": 1003
+    },
+    {
+      "epoch": 1.5643024162120032,
+      "grad_norm": 0.2978398739891535,
+      "learning_rate": 2.65606936416185e-05,
+      "loss": 0.3683,
+      "step": 1004
+    },
+    {
+      "epoch": 1.5658612626656274,
+      "grad_norm": 0.39334684334782377,
+      "learning_rate": 2.653179190751445e-05,
+      "loss": 0.3569,
+      "step": 1005
+    },
+    {
+      "epoch": 1.5674201091192517,
+      "grad_norm": 0.2717962848929705,
+      "learning_rate": 2.6502890173410405e-05,
+      "loss": 0.3561,
+      "step": 1006
+    },
+    {
+      "epoch": 1.568978955572876,
+      "grad_norm": 0.3736251973751076,
+      "learning_rate": 2.647398843930636e-05,
+      "loss": 0.3669,
+      "step": 1007
+    },
+    {
+      "epoch": 1.5705378020265004,
+      "grad_norm": 0.29899355625093155,
+      "learning_rate": 2.6445086705202315e-05,
+      "loss": 0.3604,
+      "step": 1008
+    },
+    {
+      "epoch": 1.5720966484801249,
+      "grad_norm": 0.38876556071538754,
+      "learning_rate": 2.641618497109827e-05,
+      "loss": 0.3605,
+      "step": 1009
+    },
+    {
+      "epoch": 1.5736554949337491,
+      "grad_norm": 0.325884443622183,
+      "learning_rate": 2.638728323699422e-05,
+      "loss": 0.3858,
+      "step": 1010
+    },
+    {
+      "epoch": 1.5752143413873734,
+      "grad_norm": 0.3267871100560373,
+      "learning_rate": 2.6358381502890172e-05,
+      "loss": 0.3504,
+      "step": 1011
+    },
+    {
+      "epoch": 1.5767731878409976,
+      "grad_norm": 0.33526147496951697,
+      "learning_rate": 2.632947976878613e-05,
+      "loss": 0.3559,
+      "step": 1012
+    },
+    {
+      "epoch": 1.5783320342946219,
+      "grad_norm": 0.2944668930671466,
+      "learning_rate": 2.6300578034682083e-05,
+      "loss": 0.3755,
+      "step": 1013
+    },
+    {
+      "epoch": 1.5798908807482464,
+      "grad_norm": 0.3138451746470972,
+      "learning_rate": 2.6271676300578036e-05,
+      "loss": 0.3643,
+      "step": 1014
+    },
+    {
+      "epoch": 1.5814497272018706,
+      "grad_norm": 0.26776940795700943,
+      "learning_rate": 2.6242774566473986e-05,
+      "loss": 0.3466,
+      "step": 1015
+    },
+    {
+      "epoch": 1.583008573655495,
+      "grad_norm": 0.26784179313482087,
+      "learning_rate": 2.6213872832369946e-05,
+      "loss": 0.3486,
+      "step": 1016
+    },
+    {
+      "epoch": 1.5845674201091193,
+      "grad_norm": 0.38607190770403105,
+      "learning_rate": 2.6184971098265896e-05,
+      "loss": 0.353,
+      "step": 1017
+    },
+    {
+      "epoch": 1.5861262665627436,
+      "grad_norm": 0.3065474352566968,
+      "learning_rate": 2.615606936416185e-05,
+      "loss": 0.3589,
+      "step": 1018
+    },
+    {
+      "epoch": 1.5876851130163678,
+      "grad_norm": 0.2832505582830171,
+      "learning_rate": 2.6127167630057803e-05,
+      "loss": 0.3678,
+      "step": 1019
+    },
+    {
+      "epoch": 1.589243959469992,
+      "grad_norm": 0.36412176076022207,
+      "learning_rate": 2.609826589595376e-05,
+      "loss": 0.3798,
+      "step": 1020
+    },
+    {
+      "epoch": 1.5908028059236166,
+      "grad_norm": 0.28000224882500524,
+      "learning_rate": 2.6069364161849714e-05,
+      "loss": 0.3497,
+      "step": 1021
+    },
+    {
+      "epoch": 1.5923616523772408,
+      "grad_norm": 1.8214222618798932,
+      "learning_rate": 2.6040462427745667e-05,
+      "loss": 0.4108,
+      "step": 1022
+    },
+    {
+      "epoch": 1.5939204988308653,
+      "grad_norm": 0.44865333521920003,
+      "learning_rate": 2.6011560693641617e-05,
+      "loss": 0.3772,
+      "step": 1023
+    },
+    {
+      "epoch": 1.5954793452844895,
+      "grad_norm": 0.25598641910972036,
+      "learning_rate": 2.5982658959537577e-05,
+      "loss": 0.3657,
+      "step": 1024
+    },
+    {
+      "epoch": 1.5970381917381138,
+      "grad_norm": 0.33942270049254686,
+      "learning_rate": 2.5953757225433527e-05,
+      "loss": 0.3995,
+      "step": 1025
+    },
+    {
+      "epoch": 1.598597038191738,
+      "grad_norm": 0.3657115461703349,
+      "learning_rate": 2.592485549132948e-05,
+      "loss": 0.3941,
+      "step": 1026
+    },
+    {
+      "epoch": 1.6001558846453623,
+      "grad_norm": 0.28531237255153097,
+      "learning_rate": 2.5895953757225434e-05,
+      "loss": 0.3806,
+      "step": 1027
+    },
+    {
+      "epoch": 1.6017147310989868,
+      "grad_norm": 0.35592372312382153,
+      "learning_rate": 2.586705202312139e-05,
+      "loss": 0.3781,
+      "step": 1028
+    },
+    {
+      "epoch": 1.603273577552611,
+      "grad_norm": 0.29099766422855,
+      "learning_rate": 2.5838150289017344e-05,
+      "loss": 0.3638,
+      "step": 1029
+    },
+    {
+      "epoch": 1.6048324240062355,
+      "grad_norm": 0.3174734446542449,
+      "learning_rate": 2.5809248554913295e-05,
+      "loss": 0.3435,
+      "step": 1030
+    },
+    {
+      "epoch": 1.6063912704598597,
+      "grad_norm": 0.4203285741115712,
+      "learning_rate": 2.5780346820809248e-05,
+      "loss": 0.3788,
+      "step": 1031
+    },
+    {
+      "epoch": 1.607950116913484,
+      "grad_norm": 0.32251687798177475,
+      "learning_rate": 2.5751445086705205e-05,
+      "loss": 0.3632,
+      "step": 1032
+    },
+    {
+      "epoch": 1.6095089633671082,
+      "grad_norm": 0.3531877282131029,
+      "learning_rate": 2.5722543352601158e-05,
+      "loss": 0.366,
+      "step": 1033
+    },
+    {
+      "epoch": 1.6110678098207325,
+      "grad_norm": 0.34051504848522857,
+      "learning_rate": 2.5693641618497112e-05,
+      "loss": 0.3591,
+      "step": 1034
+    },
+    {
+      "epoch": 1.612626656274357,
+      "grad_norm": 0.3072876546153309,
+      "learning_rate": 2.5664739884393062e-05,
+      "loss": 0.3807,
+      "step": 1035
+    },
+    {
+      "epoch": 1.6141855027279814,
+      "grad_norm": 0.3229272762875624,
+      "learning_rate": 2.5635838150289022e-05,
+      "loss": 0.3707,
+      "step": 1036
+    },
+    {
+      "epoch": 1.6157443491816057,
+      "grad_norm": 0.3309237615111176,
+      "learning_rate": 2.5606936416184972e-05,
+      "loss": 0.3612,
+      "step": 1037
+    },
+    {
+      "epoch": 1.61730319563523,
+      "grad_norm": 0.3111215230340758,
+      "learning_rate": 2.5578034682080925e-05,
+      "loss": 0.3803,
+      "step": 1038
+    },
+    {
+      "epoch": 1.6188620420888542,
+      "grad_norm": 0.32805176015176535,
+      "learning_rate": 2.554913294797688e-05,
+      "loss": 0.3695,
+      "step": 1039
+    },
+    {
+      "epoch": 1.6204208885424785,
+      "grad_norm": 0.2835660539290897,
+      "learning_rate": 2.5520231213872836e-05,
+      "loss": 0.3807,
+      "step": 1040
+    },
+    {
+      "epoch": 1.621979734996103,
+      "grad_norm": 0.3032998483914643,
+      "learning_rate": 2.549132947976879e-05,
+      "loss": 0.3695,
+      "step": 1041
+    },
+    {
+      "epoch": 1.6235385814497272,
+      "grad_norm": 0.27617969523468905,
+      "learning_rate": 2.546242774566474e-05,
+      "loss": 0.3723,
+      "step": 1042
+    },
+    {
+      "epoch": 1.6250974279033517,
+      "grad_norm": 0.32157009299636713,
+      "learning_rate": 2.5433526011560693e-05,
+      "loss": 0.3572,
+      "step": 1043
+    },
+    {
+      "epoch": 1.626656274356976,
+      "grad_norm": 0.2842521187908904,
+      "learning_rate": 2.540462427745665e-05,
+      "loss": 0.3656,
+      "step": 1044
+    },
+    {
+      "epoch": 1.6282151208106002,
+      "grad_norm": 0.3288780675289122,
+      "learning_rate": 2.5375722543352603e-05,
+      "loss": 0.385,
+      "step": 1045
+    },
+    {
+      "epoch": 1.6297739672642244,
+      "grad_norm": 0.32678871102435747,
+      "learning_rate": 2.5346820809248556e-05,
+      "loss": 0.3844,
+      "step": 1046
+    },
+    {
+      "epoch": 1.6313328137178487,
+      "grad_norm": 0.39119637546888175,
+      "learning_rate": 2.531791907514451e-05,
+      "loss": 0.3542,
+      "step": 1047
+    },
+    {
+      "epoch": 1.6328916601714731,
+      "grad_norm": 0.288052529248597,
+      "learning_rate": 2.5289017341040467e-05,
+      "loss": 0.3572,
+      "step": 1048
+    },
+    {
+      "epoch": 1.6344505066250974,
+      "grad_norm": 0.3487713869931321,
+      "learning_rate": 2.526011560693642e-05,
+      "loss": 0.3542,
+      "step": 1049
+    },
+    {
+      "epoch": 1.6360093530787219,
+      "grad_norm": 0.26513201445656637,
+      "learning_rate": 2.523121387283237e-05,
+      "loss": 0.3731,
+      "step": 1050
+    },
+    {
+      "epoch": 1.6375681995323461,
+      "grad_norm": 0.32844205151640604,
+      "learning_rate": 2.5202312138728324e-05,
+      "loss": 0.3691,
+      "step": 1051
+    },
+    {
+      "epoch": 1.6391270459859704,
+      "grad_norm": 0.30490163955005906,
+      "learning_rate": 2.517341040462428e-05,
+      "loss": 0.3813,
+      "step": 1052
+    },
+    {
+      "epoch": 1.6406858924395946,
+      "grad_norm": 0.4555429114132785,
+      "learning_rate": 2.5144508670520234e-05,
+      "loss": 0.3818,
+      "step": 1053
+    },
+    {
+      "epoch": 1.6422447388932189,
+      "grad_norm": 0.29026515692355687,
+      "learning_rate": 2.5115606936416187e-05,
+      "loss": 0.3626,
+      "step": 1054
+    },
+    {
+      "epoch": 1.6438035853468433,
+      "grad_norm": 0.39179865186142615,
+      "learning_rate": 2.5086705202312137e-05,
+      "loss": 0.3502,
+      "step": 1055
+    },
+    {
+      "epoch": 1.6453624318004678,
+      "grad_norm": 0.32433692268192194,
+      "learning_rate": 2.5057803468208098e-05,
+      "loss": 0.3781,
+      "step": 1056
+    },
+    {
+      "epoch": 1.646921278254092,
+      "grad_norm": 0.33527613308127024,
+      "learning_rate": 2.5028901734104048e-05,
+      "loss": 0.3619,
+      "step": 1057
+    },
+    {
+      "epoch": 1.6484801247077163,
+      "grad_norm": 0.3159961287138508,
+      "learning_rate": 2.5e-05,
+      "loss": 0.3632,
+      "step": 1058
+    },
+    {
+      "epoch": 1.6500389711613406,
+      "grad_norm": 0.31940038082116945,
+      "learning_rate": 2.4971098265895955e-05,
+      "loss": 0.3633,
+      "step": 1059
+    },
+    {
+      "epoch": 1.6515978176149648,
+      "grad_norm": 0.3303935987168502,
+      "learning_rate": 2.4942196531791908e-05,
+      "loss": 0.3745,
+      "step": 1060
+    },
+    {
+      "epoch": 1.6531566640685893,
+      "grad_norm": 0.34507486262396847,
+      "learning_rate": 2.491329479768786e-05,
+      "loss": 0.3507,
+      "step": 1061
+    },
+    {
+      "epoch": 1.6547155105222136,
+      "grad_norm": 0.2898972340965827,
+      "learning_rate": 2.4884393063583815e-05,
+      "loss": 0.3583,
+      "step": 1062
+    },
+    {
+      "epoch": 1.656274356975838,
+      "grad_norm": 0.29250387172577985,
+      "learning_rate": 2.485549132947977e-05,
+      "loss": 0.3754,
+      "step": 1063
+    },
+    {
+      "epoch": 1.6578332034294623,
+      "grad_norm": 0.3144587373068325,
+      "learning_rate": 2.4826589595375725e-05,
+      "loss": 0.3675,
+      "step": 1064
+    },
+    {
+      "epoch": 1.6593920498830865,
+      "grad_norm": 0.28878104517546993,
+      "learning_rate": 2.4797687861271675e-05,
+      "loss": 0.3535,
+      "step": 1065
+    },
+    {
+      "epoch": 1.6609508963367108,
+      "grad_norm": 0.2915446431012894,
+      "learning_rate": 2.4768786127167632e-05,
+      "loss": 0.3461,
+      "step": 1066
+    },
+    {
+      "epoch": 1.662509742790335,
+      "grad_norm": 0.3180916504846977,
+      "learning_rate": 2.4739884393063582e-05,
+      "loss": 0.3482,
+      "step": 1067
+    },
+    {
+      "epoch": 1.6640685892439595,
+      "grad_norm": 0.35825594962772983,
+      "learning_rate": 2.471098265895954e-05,
+      "loss": 0.3822,
+      "step": 1068
+    },
+    {
+      "epoch": 1.6656274356975838,
+      "grad_norm": 0.2716382477039464,
+      "learning_rate": 2.4682080924855492e-05,
+      "loss": 0.3712,
+      "step": 1069
+    },
+    {
+      "epoch": 1.6671862821512082,
+      "grad_norm": 0.3832340579480356,
+      "learning_rate": 2.4653179190751446e-05,
+      "loss": 0.3744,
+      "step": 1070
+    },
+    {
+      "epoch": 1.6687451286048325,
+      "grad_norm": 0.30678572957271294,
+      "learning_rate": 2.46242774566474e-05,
+      "loss": 0.3579,
+      "step": 1071
+    },
+    {
+      "epoch": 1.6703039750584567,
+      "grad_norm": 0.2843857794151713,
+      "learning_rate": 2.4595375722543353e-05,
+      "loss": 0.3826,
+      "step": 1072
+    },
+    {
+      "epoch": 1.671862821512081,
+      "grad_norm": 0.34064253973495867,
+      "learning_rate": 2.4566473988439306e-05,
+      "loss": 0.3761,
+      "step": 1073
+    },
+    {
+      "epoch": 1.6734216679657052,
+      "grad_norm": 0.29119739353949486,
+      "learning_rate": 2.4537572254335263e-05,
+      "loss": 0.3755,
+      "step": 1074
+    },
+    {
+      "epoch": 1.6749805144193297,
+      "grad_norm": 0.2816956584902144,
+      "learning_rate": 2.4508670520231213e-05,
+      "loss": 0.3741,
+      "step": 1075
+    },
+    {
+      "epoch": 1.676539360872954,
+      "grad_norm": 0.3111411062453413,
+      "learning_rate": 2.447976878612717e-05,
+      "loss": 0.3825,
+      "step": 1076
+    },
+    {
+      "epoch": 1.6780982073265784,
+      "grad_norm": 0.2726926618664558,
+      "learning_rate": 2.445086705202312e-05,
+      "loss": 0.3879,
+      "step": 1077
+    },
+    {
+      "epoch": 1.6796570537802027,
+      "grad_norm": 0.2834300086794794,
+      "learning_rate": 2.4421965317919077e-05,
+      "loss": 0.3705,
+      "step": 1078
+    },
+    {
+      "epoch": 1.681215900233827,
+      "grad_norm": 0.31189081369621346,
+      "learning_rate": 2.439306358381503e-05,
+      "loss": 0.3663,
+      "step": 1079
+    },
+    {
+      "epoch": 1.6827747466874512,
+      "grad_norm": 0.2829807568606232,
+      "learning_rate": 2.4364161849710984e-05,
+      "loss": 0.3626,
+      "step": 1080
+    },
+    {
+      "epoch": 1.6843335931410754,
+      "grad_norm": 0.32489614588339644,
+      "learning_rate": 2.4335260115606937e-05,
+      "loss": 0.3733,
+      "step": 1081
+    },
+    {
+      "epoch": 1.6858924395947,
+      "grad_norm": 0.3379997934308957,
+      "learning_rate": 2.430635838150289e-05,
+      "loss": 0.3843,
+      "step": 1082
+    },
+    {
+      "epoch": 1.6874512860483244,
+      "grad_norm": 0.2730933075226193,
+      "learning_rate": 2.4277456647398844e-05,
+      "loss": 0.3825,
+      "step": 1083
+    },
+    {
+      "epoch": 1.6890101325019486,
+      "grad_norm": 0.3404758440705323,
+      "learning_rate": 2.4248554913294798e-05,
+      "loss": 0.3692,
+      "step": 1084
+    },
+    {
+      "epoch": 1.690568978955573,
+      "grad_norm": 0.2678166339094811,
+      "learning_rate": 2.421965317919075e-05,
+      "loss": 0.358,
+      "step": 1085
+    },
+    {
+      "epoch": 1.6921278254091972,
+      "grad_norm": 0.3179761438396162,
+      "learning_rate": 2.4190751445086708e-05,
+      "loss": 0.3615,
+      "step": 1086
+    },
+    {
+      "epoch": 1.6936866718628214,
+      "grad_norm": 0.30230071697163113,
+      "learning_rate": 2.4161849710982658e-05,
+      "loss": 0.3778,
+      "step": 1087
+    },
+    {
+      "epoch": 1.6952455183164459,
+      "grad_norm": 0.291898879189332,
+      "learning_rate": 2.4132947976878615e-05,
+      "loss": 0.373,
+      "step": 1088
+    },
+    {
+      "epoch": 1.6968043647700701,
+      "grad_norm": 0.2503981788139211,
+      "learning_rate": 2.4104046242774568e-05,
+      "loss": 0.3568,
+      "step": 1089
+    },
+    {
+      "epoch": 1.6983632112236946,
+      "grad_norm": 0.28111682299046925,
+      "learning_rate": 2.407514450867052e-05,
+      "loss": 0.352,
+      "step": 1090
+    },
+    {
+      "epoch": 1.6999220576773189,
+      "grad_norm": 0.2448550241175499,
+      "learning_rate": 2.4046242774566475e-05,
+      "loss": 0.3657,
+      "step": 1091
+    },
+    {
+      "epoch": 1.701480904130943,
+      "grad_norm": 0.3065398908103042,
+      "learning_rate": 2.401734104046243e-05,
+      "loss": 0.3632,
+      "step": 1092
+    },
+    {
+      "epoch": 1.7030397505845674,
+      "grad_norm": 0.2822693214761903,
+      "learning_rate": 2.3988439306358382e-05,
+      "loss": 0.401,
+      "step": 1093
+    },
+    {
+      "epoch": 1.7045985970381916,
+      "grad_norm": 0.29360065434429594,
+      "learning_rate": 2.3959537572254335e-05,
+      "loss": 0.3682,
+      "step": 1094
+    },
+    {
+      "epoch": 1.706157443491816,
+      "grad_norm": 0.24618104109851893,
+      "learning_rate": 2.393063583815029e-05,
+      "loss": 0.3653,
+      "step": 1095
+    },
+    {
+      "epoch": 1.7077162899454403,
+      "grad_norm": 0.27658727281555634,
+      "learning_rate": 2.3901734104046246e-05,
+      "loss": 0.3703,
+      "step": 1096
+    },
+    {
+      "epoch": 1.7092751363990648,
+      "grad_norm": 0.2777982792931725,
+      "learning_rate": 2.3872832369942196e-05,
+      "loss": 0.382,
+      "step": 1097
+    },
+    {
+      "epoch": 1.710833982852689,
+      "grad_norm": 0.30238858038905553,
+      "learning_rate": 2.3843930635838153e-05,
+      "loss": 0.3646,
+      "step": 1098
+    },
+    {
+      "epoch": 1.7123928293063133,
+      "grad_norm": 0.27806470539498046,
+      "learning_rate": 2.3815028901734106e-05,
+      "loss": 0.3745,
+      "step": 1099
+    },
+    {
+      "epoch": 1.7139516757599376,
+      "grad_norm": 0.29767876896518725,
+      "learning_rate": 2.378612716763006e-05,
+      "loss": 0.3758,
+      "step": 1100
+    },
+    {
+      "epoch": 1.7155105222135618,
+      "grad_norm": 0.2829737040629793,
+      "learning_rate": 2.3757225433526013e-05,
+      "loss": 0.3666,
+      "step": 1101
+    },
+    {
+      "epoch": 1.7170693686671863,
+      "grad_norm": 0.3578161092071975,
+      "learning_rate": 2.3728323699421966e-05,
+      "loss": 0.3886,
+      "step": 1102
+    },
+    {
+      "epoch": 1.7186282151208108,
+      "grad_norm": 0.29906079097318256,
+      "learning_rate": 2.369942196531792e-05,
+      "loss": 0.3671,
+      "step": 1103
+    },
+    {
+      "epoch": 1.720187061574435,
+      "grad_norm": 0.28648360330595535,
+      "learning_rate": 2.3670520231213873e-05,
+      "loss": 0.3552,
+      "step": 1104
+    },
+    {
+      "epoch": 1.7217459080280593,
+      "grad_norm": 0.2690083195767578,
+      "learning_rate": 2.3641618497109827e-05,
+      "loss": 0.3642,
+      "step": 1105
+    },
+    {
+      "epoch": 1.7233047544816835,
+      "grad_norm": 0.3412710940527978,
+      "learning_rate": 2.3612716763005784e-05,
+      "loss": 0.3851,
+      "step": 1106
+    },
+    {
+      "epoch": 1.7248636009353078,
+      "grad_norm": 0.2945093059468343,
+      "learning_rate": 2.3583815028901734e-05,
+      "loss": 0.3623,
+      "step": 1107
+    },
+    {
+      "epoch": 1.726422447388932,
+      "grad_norm": 0.2745278267359284,
+      "learning_rate": 2.355491329479769e-05,
+      "loss": 0.3573,
+      "step": 1108
+    },
+    {
+      "epoch": 1.7279812938425565,
+      "grad_norm": 0.2718225903508495,
+      "learning_rate": 2.352601156069364e-05,
+      "loss": 0.3641,
+      "step": 1109
+    },
+    {
+      "epoch": 1.729540140296181,
+      "grad_norm": 0.27774676623694455,
+      "learning_rate": 2.3497109826589597e-05,
+      "loss": 0.3621,
+      "step": 1110
+    },
+    {
+      "epoch": 1.7310989867498052,
+      "grad_norm": 0.26445890337068356,
+      "learning_rate": 2.346820809248555e-05,
+      "loss": 0.3616,
+      "step": 1111
+    },
+    {
+      "epoch": 1.7326578332034295,
+      "grad_norm": 0.34740235201216224,
+      "learning_rate": 2.3439306358381504e-05,
+      "loss": 0.3549,
+      "step": 1112
+    },
+    {
+      "epoch": 1.7342166796570537,
+      "grad_norm": 0.27846442379370767,
+      "learning_rate": 2.3410404624277458e-05,
+      "loss": 0.3582,
+      "step": 1113
+    },
+    {
+      "epoch": 1.735775526110678,
+      "grad_norm": 0.3157341078739885,
+      "learning_rate": 2.338150289017341e-05,
+      "loss": 0.3777,
+      "step": 1114
+    },
+    {
+      "epoch": 1.7373343725643025,
+      "grad_norm": 0.27171759230918247,
+      "learning_rate": 2.3352601156069365e-05,
+      "loss": 0.3752,
+      "step": 1115
+    },
+    {
+      "epoch": 1.7388932190179267,
+      "grad_norm": 0.3016475132282448,
+      "learning_rate": 2.332369942196532e-05,
+      "loss": 0.3815,
+      "step": 1116
+    },
+    {
+      "epoch": 1.7404520654715512,
+      "grad_norm": 0.28434294314632325,
+      "learning_rate": 2.329479768786127e-05,
+      "loss": 0.3771,
+      "step": 1117
+    },
+    {
+      "epoch": 1.7420109119251754,
+      "grad_norm": 0.3013627601659192,
+      "learning_rate": 2.326589595375723e-05,
+      "loss": 0.3668,
+      "step": 1118
+    },
+    {
+      "epoch": 1.7435697583787997,
+      "grad_norm": 0.2907093179335693,
+      "learning_rate": 2.323699421965318e-05,
+      "loss": 0.3939,
+      "step": 1119
+    },
+    {
+      "epoch": 1.745128604832424,
+      "grad_norm": 0.2727965148056477,
+      "learning_rate": 2.3208092485549135e-05,
+      "loss": 0.3892,
+      "step": 1120
+    },
+    {
+      "epoch": 1.7466874512860482,
+      "grad_norm": 0.2993429686182799,
+      "learning_rate": 2.317919075144509e-05,
+      "loss": 0.3519,
+      "step": 1121
+    },
+    {
+      "epoch": 1.7482462977396727,
+      "grad_norm": 0.2693671384061593,
+      "learning_rate": 2.3150289017341042e-05,
+      "loss": 0.3549,
+      "step": 1122
+    },
+    {
+      "epoch": 1.749805144193297,
+      "grad_norm": 0.3209589746292965,
+      "learning_rate": 2.3121387283236996e-05,
+      "loss": 0.375,
+      "step": 1123
+    },
+    {
+      "epoch": 1.7513639906469214,
+      "grad_norm": 0.24312926167519938,
+      "learning_rate": 2.309248554913295e-05,
+      "loss": 0.3499,
+      "step": 1124
+    },
+    {
+      "epoch": 1.7529228371005456,
+      "grad_norm": 0.3104273862594975,
+      "learning_rate": 2.3063583815028902e-05,
+      "loss": 0.389,
+      "step": 1125
+    },
+    {
+      "epoch": 1.75448168355417,
+      "grad_norm": 0.2752541588528496,
+      "learning_rate": 2.303468208092486e-05,
+      "loss": 0.3646,
+      "step": 1126
+    },
+    {
+      "epoch": 1.7560405300077941,
+      "grad_norm": 0.27374708814696136,
+      "learning_rate": 2.300578034682081e-05,
+      "loss": 0.3576,
+      "step": 1127
+    },
+    {
+      "epoch": 1.7575993764614184,
+      "grad_norm": 0.31492934426283914,
+      "learning_rate": 2.2976878612716766e-05,
+      "loss": 0.3872,
+      "step": 1128
+    },
+    {
+      "epoch": 1.7591582229150429,
+      "grad_norm": 0.28256761554526544,
+      "learning_rate": 2.2947976878612716e-05,
+      "loss": 0.3659,
+      "step": 1129
+    },
+    {
+      "epoch": 1.7607170693686673,
+      "grad_norm": 0.2543705987468934,
+      "learning_rate": 2.2919075144508673e-05,
+      "loss": 0.3607,
+      "step": 1130
+    },
+    {
+      "epoch": 1.7622759158222916,
+      "grad_norm": 0.298933402272287,
+      "learning_rate": 2.2890173410404627e-05,
+      "loss": 0.3683,
+      "step": 1131
+    },
+    {
+      "epoch": 1.7638347622759158,
+      "grad_norm": 0.311816355725973,
+      "learning_rate": 2.286127167630058e-05,
+      "loss": 0.3597,
+      "step": 1132
+    },
+    {
+      "epoch": 1.76539360872954,
+      "grad_norm": 0.3067218252259407,
+      "learning_rate": 2.2832369942196533e-05,
+      "loss": 0.3786,
+      "step": 1133
+    },
+    {
+      "epoch": 1.7669524551831643,
+      "grad_norm": 0.2861975440862343,
+      "learning_rate": 2.2803468208092487e-05,
+      "loss": 0.3731,
+      "step": 1134
+    },
+    {
+      "epoch": 1.7685113016367888,
+      "grad_norm": 0.316953066205446,
+      "learning_rate": 2.277456647398844e-05,
+      "loss": 0.3518,
+      "step": 1135
+    },
+    {
+      "epoch": 1.770070148090413,
+      "grad_norm": 0.2708691409104658,
+      "learning_rate": 2.2745664739884394e-05,
+      "loss": 0.373,
+      "step": 1136
+    },
+    {
+      "epoch": 1.7716289945440375,
+      "grad_norm": 0.26797311923183315,
+      "learning_rate": 2.2716763005780347e-05,
+      "loss": 0.3665,
+      "step": 1137
+    },
+    {
+      "epoch": 1.7731878409976618,
+      "grad_norm": 0.3099497339980096,
+      "learning_rate": 2.2687861271676304e-05,
+      "loss": 0.3951,
+      "step": 1138
+    },
+    {
+      "epoch": 1.774746687451286,
+      "grad_norm": 0.34054924476007226,
+      "learning_rate": 2.2658959537572254e-05,
+      "loss": 0.3594,
+      "step": 1139
+    },
+    {
+      "epoch": 1.7763055339049103,
+      "grad_norm": 0.25223898072450013,
+      "learning_rate": 2.263005780346821e-05,
+      "loss": 0.361,
+      "step": 1140
+    },
+    {
+      "epoch": 1.7778643803585346,
+      "grad_norm": 0.29284462313911747,
+      "learning_rate": 2.2601156069364164e-05,
+      "loss": 0.3575,
+      "step": 1141
+    },
+    {
+      "epoch": 1.779423226812159,
+      "grad_norm": 0.2960130777592313,
+      "learning_rate": 2.2572254335260118e-05,
+      "loss": 0.3715,
+      "step": 1142
+    },
+    {
+      "epoch": 1.7809820732657833,
+      "grad_norm": 0.26708445021070565,
+      "learning_rate": 2.254335260115607e-05,
+      "loss": 0.3685,
+      "step": 1143
+    },
+    {
+      "epoch": 1.7825409197194078,
+      "grad_norm": 0.23850951752764862,
+      "learning_rate": 2.2514450867052025e-05,
+      "loss": 0.3528,
+      "step": 1144
+    },
+    {
+      "epoch": 1.784099766173032,
+      "grad_norm": 0.28725957396773283,
+      "learning_rate": 2.2485549132947978e-05,
+      "loss": 0.3575,
+      "step": 1145
+    },
+    {
+      "epoch": 1.7856586126266563,
+      "grad_norm": 0.27977866068073226,
+      "learning_rate": 2.245664739884393e-05,
+      "loss": 0.3605,
+      "step": 1146
+    },
+    {
+      "epoch": 1.7872174590802805,
+      "grad_norm": 0.27572937920147894,
+      "learning_rate": 2.2427745664739885e-05,
+      "loss": 0.3797,
+      "step": 1147
+    },
+    {
+      "epoch": 1.7887763055339048,
+      "grad_norm": 0.3044698433558107,
+      "learning_rate": 2.2398843930635842e-05,
+      "loss": 0.3561,
+      "step": 1148
+    },
+    {
+      "epoch": 1.7903351519875292,
+      "grad_norm": 0.3034254736946843,
+      "learning_rate": 2.2369942196531792e-05,
+      "loss": 0.3729,
+      "step": 1149
+    },
+    {
+      "epoch": 1.7918939984411537,
+      "grad_norm": 0.29184309753197535,
+      "learning_rate": 2.234104046242775e-05,
+      "loss": 0.3602,
+      "step": 1150
+    },
+    {
+      "epoch": 1.793452844894778,
+      "grad_norm": 0.2752185880115312,
+      "learning_rate": 2.23121387283237e-05,
+      "loss": 0.3705,
+      "step": 1151
+    },
+    {
+      "epoch": 1.7950116913484022,
+      "grad_norm": 0.2809903038728107,
+      "learning_rate": 2.2283236994219656e-05,
+      "loss": 0.3545,
+      "step": 1152
+    },
+    {
+      "epoch": 1.7965705378020265,
+      "grad_norm": 0.3185758597066887,
+      "learning_rate": 2.225433526011561e-05,
+      "loss": 0.3775,
+      "step": 1153
+    },
+    {
+      "epoch": 1.7981293842556507,
+      "grad_norm": 0.32522837307739677,
+      "learning_rate": 2.2225433526011563e-05,
+      "loss": 0.3762,
+      "step": 1154
+    },
+    {
+      "epoch": 1.799688230709275,
+      "grad_norm": 0.29829669655375685,
+      "learning_rate": 2.2196531791907516e-05,
+      "loss": 0.3431,
+      "step": 1155
+    },
+    {
+      "epoch": 1.8012470771628994,
+      "grad_norm": 0.2923880518129942,
+      "learning_rate": 2.216763005780347e-05,
+      "loss": 0.3625,
+      "step": 1156
+    },
+    {
+      "epoch": 1.802805923616524,
+      "grad_norm": 0.31193429426470687,
+      "learning_rate": 2.2138728323699423e-05,
+      "loss": 0.3687,
+      "step": 1157
+    },
+    {
+      "epoch": 1.8043647700701482,
+      "grad_norm": 0.3028214816745831,
+      "learning_rate": 2.2109826589595376e-05,
+      "loss": 0.3835,
+      "step": 1158
+    },
+    {
+      "epoch": 1.8059236165237724,
+      "grad_norm": 0.35099054840606936,
+      "learning_rate": 2.208092485549133e-05,
+      "loss": 0.3785,
+      "step": 1159
+    },
+    {
+      "epoch": 1.8074824629773967,
+      "grad_norm": 0.3153451160744472,
+      "learning_rate": 2.2052023121387283e-05,
+      "loss": 0.367,
+      "step": 1160
+    },
+    {
+      "epoch": 1.809041309431021,
+      "grad_norm": 0.3253785936574865,
+      "learning_rate": 2.2023121387283237e-05,
+      "loss": 0.3877,
+      "step": 1161
+    },
+    {
+      "epoch": 1.8106001558846454,
+      "grad_norm": 0.30760323974470033,
+      "learning_rate": 2.199421965317919e-05,
+      "loss": 0.3643,
+      "step": 1162
+    },
+    {
+      "epoch": 1.8121590023382697,
+      "grad_norm": 0.3598054851684999,
+      "learning_rate": 2.1965317919075147e-05,
+      "loss": 0.3586,
+      "step": 1163
+    },
+    {
+      "epoch": 1.8137178487918941,
+      "grad_norm": 0.26304147816154894,
+      "learning_rate": 2.1936416184971097e-05,
+      "loss": 0.3621,
+      "step": 1164
+    },
+    {
+      "epoch": 1.8152766952455184,
+      "grad_norm": 0.3322641073371267,
+      "learning_rate": 2.1907514450867054e-05,
+      "loss": 0.3787,
+      "step": 1165
+    },
+    {
+      "epoch": 1.8168355416991426,
+      "grad_norm": 0.2972445280530977,
+      "learning_rate": 2.1878612716763007e-05,
+      "loss": 0.3623,
+      "step": 1166
+    },
+    {
+      "epoch": 1.8183943881527669,
+      "grad_norm": 0.29860325366761714,
+      "learning_rate": 2.184971098265896e-05,
+      "loss": 0.3693,
+      "step": 1167
+    },
+    {
+      "epoch": 1.8199532346063911,
+      "grad_norm": 0.31932300688288,
+      "learning_rate": 2.1820809248554914e-05,
+      "loss": 0.3687,
+      "step": 1168
+    },
+    {
+      "epoch": 1.8215120810600156,
+      "grad_norm": 0.25058398882971866,
+      "learning_rate": 2.1791907514450868e-05,
+      "loss": 0.3688,
+      "step": 1169
+    },
+    {
+      "epoch": 1.8230709275136399,
+      "grad_norm": 0.30451618236243205,
+      "learning_rate": 2.176300578034682e-05,
+      "loss": 0.3604,
+      "step": 1170
+    },
+    {
+      "epoch": 1.8246297739672643,
+      "grad_norm": 0.30692079739121414,
+      "learning_rate": 2.1734104046242775e-05,
+      "loss": 0.3515,
+      "step": 1171
+    },
+    {
+      "epoch": 1.8261886204208886,
+      "grad_norm": 0.2807778315885177,
+      "learning_rate": 2.1705202312138728e-05,
+      "loss": 0.3509,
+      "step": 1172
+    },
+    {
+      "epoch": 1.8277474668745128,
+      "grad_norm": 0.2905244115992112,
+      "learning_rate": 2.1676300578034685e-05,
+      "loss": 0.3661,
+      "step": 1173
+    },
+    {
+      "epoch": 1.829306313328137,
+      "grad_norm": 0.3105261582464057,
+      "learning_rate": 2.1647398843930635e-05,
+      "loss": 0.372,
+      "step": 1174
+    },
+    {
+      "epoch": 1.8308651597817613,
+      "grad_norm": 0.3095524114344275,
+      "learning_rate": 2.1618497109826592e-05,
+      "loss": 0.3683,
+      "step": 1175
+    },
+    {
+      "epoch": 1.8324240062353858,
+      "grad_norm": 0.3224707852323462,
+      "learning_rate": 2.1589595375722542e-05,
+      "loss": 0.3941,
+      "step": 1176
+    },
+    {
+      "epoch": 1.8339828526890103,
+      "grad_norm": 0.37265316733469883,
+      "learning_rate": 2.15606936416185e-05,
+      "loss": 0.3729,
+      "step": 1177
+    },
+    {
+      "epoch": 1.8355416991426345,
+      "grad_norm": 0.3484805036172097,
+      "learning_rate": 2.1531791907514452e-05,
+      "loss": 0.3695,
+      "step": 1178
+    },
+    {
+      "epoch": 1.8371005455962588,
+      "grad_norm": 0.3062185230582236,
+      "learning_rate": 2.1502890173410405e-05,
+      "loss": 0.3722,
+      "step": 1179
+    },
+    {
+      "epoch": 1.838659392049883,
+      "grad_norm": 0.3514017844930539,
+      "learning_rate": 2.147398843930636e-05,
+      "loss": 0.3716,
+      "step": 1180
+    },
+    {
+      "epoch": 1.8402182385035073,
+      "grad_norm": 0.29370679610653583,
+      "learning_rate": 2.1445086705202312e-05,
+      "loss": 0.3511,
+      "step": 1181
+    },
+    {
+      "epoch": 1.8417770849571318,
+      "grad_norm": 0.3650311060921678,
+      "learning_rate": 2.1416184971098266e-05,
+      "loss": 0.3746,
+      "step": 1182
+    },
+    {
+      "epoch": 1.843335931410756,
+      "grad_norm": 0.28361922423446095,
+      "learning_rate": 2.1387283236994223e-05,
+      "loss": 0.354,
+      "step": 1183
+    },
+    {
+      "epoch": 1.8448947778643805,
+      "grad_norm": 0.2915388311697381,
+      "learning_rate": 2.1358381502890173e-05,
+      "loss": 0.36,
+      "step": 1184
+    },
+    {
+      "epoch": 1.8464536243180047,
+      "grad_norm": 0.33038140947084105,
+      "learning_rate": 2.132947976878613e-05,
+      "loss": 0.3554,
+      "step": 1185
+    },
+    {
+      "epoch": 1.848012470771629,
+      "grad_norm": 0.419071588239709,
+      "learning_rate": 2.130057803468208e-05,
+      "loss": 0.3834,
+      "step": 1186
+    },
+    {
+      "epoch": 1.8495713172252533,
+      "grad_norm": 0.3514728435923063,
+      "learning_rate": 2.1271676300578036e-05,
+      "loss": 0.3627,
+      "step": 1187
+    },
+    {
+      "epoch": 1.8511301636788775,
+      "grad_norm": 0.35707340267403825,
+      "learning_rate": 2.124277456647399e-05,
+      "loss": 0.3874,
+      "step": 1188
+    },
+    {
+      "epoch": 1.852689010132502,
+      "grad_norm": 0.37099787424788894,
+      "learning_rate": 2.1213872832369943e-05,
+      "loss": 0.3726,
+      "step": 1189
+    },
+    {
+      "epoch": 1.8542478565861262,
+      "grad_norm": 0.330071756650617,
+      "learning_rate": 2.1184971098265897e-05,
+      "loss": 0.3715,
+      "step": 1190
+    },
+    {
+      "epoch": 1.8558067030397507,
+      "grad_norm": 0.37948140477016307,
+      "learning_rate": 2.115606936416185e-05,
+      "loss": 0.3561,
+      "step": 1191
+    },
+    {
+      "epoch": 1.857365549493375,
+      "grad_norm": 0.27396995420812964,
+      "learning_rate": 2.1127167630057804e-05,
+      "loss": 0.38,
+      "step": 1192
+    },
+    {
+      "epoch": 1.8589243959469992,
+      "grad_norm": 0.35094585580789595,
+      "learning_rate": 2.1098265895953757e-05,
+      "loss": 0.3646,
+      "step": 1193
+    },
+    {
+      "epoch": 1.8604832424006235,
+      "grad_norm": 0.3047789152628225,
+      "learning_rate": 2.106936416184971e-05,
+      "loss": 0.3778,
+      "step": 1194
+    },
+    {
+      "epoch": 1.8620420888542477,
+      "grad_norm": 0.33234879657650934,
+      "learning_rate": 2.1040462427745667e-05,
+      "loss": 0.3713,
+      "step": 1195
+    },
+    {
+      "epoch": 1.8636009353078722,
+      "grad_norm": 0.32766164310333334,
+      "learning_rate": 2.1011560693641617e-05,
+      "loss": 0.3736,
+      "step": 1196
+    },
+    {
+      "epoch": 1.8651597817614964,
+      "grad_norm": 0.2759861630025658,
+      "learning_rate": 2.0982658959537574e-05,
+      "loss": 0.3796,
+      "step": 1197
+    },
+    {
+      "epoch": 1.866718628215121,
+      "grad_norm": 0.32521839552334825,
+      "learning_rate": 2.0953757225433528e-05,
+      "loss": 0.362,
+      "step": 1198
+    },
+    {
+      "epoch": 1.8682774746687452,
+      "grad_norm": 0.2931283862783075,
+      "learning_rate": 2.092485549132948e-05,
+      "loss": 0.3502,
+      "step": 1199
+    },
+    {
+      "epoch": 1.8698363211223694,
+      "grad_norm": 0.2991343395943882,
+      "learning_rate": 2.0895953757225435e-05,
+      "loss": 0.3714,
+      "step": 1200
+    },
+    {
+      "epoch": 1.8713951675759937,
+      "grad_norm": 0.2932605788220294,
+      "learning_rate": 2.0867052023121388e-05,
+      "loss": 0.3819,
+      "step": 1201
+    },
+    {
+      "epoch": 1.872954014029618,
+      "grad_norm": 0.30785885245721334,
+      "learning_rate": 2.083815028901734e-05,
+      "loss": 0.3471,
+      "step": 1202
+    },
+    {
+      "epoch": 1.8745128604832424,
+      "grad_norm": 0.32014124433096525,
+      "learning_rate": 2.0809248554913295e-05,
+      "loss": 0.3561,
+      "step": 1203
+    },
+    {
+      "epoch": 1.8760717069368669,
+      "grad_norm": 0.3235541208435597,
+      "learning_rate": 2.078034682080925e-05,
+      "loss": 0.3866,
+      "step": 1204
+    },
+    {
+      "epoch": 1.8776305533904911,
+      "grad_norm": 0.30949299738115854,
+      "learning_rate": 2.0751445086705205e-05,
+      "loss": 0.3639,
+      "step": 1205
+    },
+    {
+      "epoch": 1.8791893998441154,
+      "grad_norm": 0.2817005798429569,
+      "learning_rate": 2.0722543352601155e-05,
+      "loss": 0.3597,
+      "step": 1206
+    },
+    {
+      "epoch": 1.8807482462977396,
+      "grad_norm": 0.30497938827386945,
+      "learning_rate": 2.0693641618497112e-05,
+      "loss": 0.383,
+      "step": 1207
+    },
+    {
+      "epoch": 1.8823070927513639,
+      "grad_norm": 0.25436146038693197,
+      "learning_rate": 2.0664739884393066e-05,
+      "loss": 0.3692,
+      "step": 1208
+    },
+    {
+      "epoch": 1.8838659392049883,
+      "grad_norm": 0.3056404459036848,
+      "learning_rate": 2.063583815028902e-05,
+      "loss": 0.3704,
+      "step": 1209
+    },
+    {
+      "epoch": 1.8854247856586126,
+      "grad_norm": 0.26875488415904514,
+      "learning_rate": 2.0606936416184972e-05,
+      "loss": 0.3594,
+      "step": 1210
+    },
+    {
+      "epoch": 1.886983632112237,
+      "grad_norm": 0.3511619874207967,
+      "learning_rate": 2.0578034682080926e-05,
+      "loss": 0.3798,
+      "step": 1211
+    },
+    {
+      "epoch": 1.8885424785658613,
+      "grad_norm": 0.26158426756531344,
+      "learning_rate": 2.054913294797688e-05,
+      "loss": 0.3856,
+      "step": 1212
+    },
+    {
+      "epoch": 1.8901013250194856,
+      "grad_norm": 0.3122928786540747,
+      "learning_rate": 2.0520231213872833e-05,
+      "loss": 0.3677,
+      "step": 1213
+    },
+    {
+      "epoch": 1.8916601714731098,
+      "grad_norm": 0.2684224099385814,
+      "learning_rate": 2.0491329479768786e-05,
+      "loss": 0.3534,
+      "step": 1214
+    },
+    {
+      "epoch": 1.893219017926734,
+      "grad_norm": 0.2571315731383385,
+      "learning_rate": 2.0462427745664743e-05,
+      "loss": 0.3722,
+      "step": 1215
+    },
+    {
+      "epoch": 1.8947778643803586,
+      "grad_norm": 0.25588546444512905,
+      "learning_rate": 2.0433526011560693e-05,
+      "loss": 0.3681,
+      "step": 1216
+    },
+    {
+      "epoch": 1.8963367108339828,
+      "grad_norm": 0.3187680200863735,
+      "learning_rate": 2.040462427745665e-05,
+      "loss": 0.3634,
+      "step": 1217
+    },
+    {
+      "epoch": 1.8978955572876073,
+      "grad_norm": 0.2611861721574427,
+      "learning_rate": 2.03757225433526e-05,
+      "loss": 0.3879,
+      "step": 1218
+    },
+    {
+      "epoch": 1.8994544037412315,
+      "grad_norm": 0.30654388499949137,
+      "learning_rate": 2.0346820809248557e-05,
+      "loss": 0.363,
+      "step": 1219
+    },
+    {
+      "epoch": 1.9010132501948558,
+      "grad_norm": 0.32726906906242276,
+      "learning_rate": 2.031791907514451e-05,
+      "loss": 0.369,
+      "step": 1220
+    },
+    {
+      "epoch": 1.90257209664848,
+      "grad_norm": 0.28441311387650836,
+      "learning_rate": 2.0289017341040464e-05,
+      "loss": 0.3559,
+      "step": 1221
+    },
+    {
+      "epoch": 1.9041309431021043,
+      "grad_norm": 0.2657998910454343,
+      "learning_rate": 2.0260115606936417e-05,
+      "loss": 0.3794,
+      "step": 1222
+    },
+    {
+      "epoch": 1.9056897895557288,
+      "grad_norm": 0.31053785670596307,
+      "learning_rate": 2.023121387283237e-05,
+      "loss": 0.3658,
+      "step": 1223
+    },
+    {
+      "epoch": 1.9072486360093532,
+      "grad_norm": 0.30798708262091334,
+      "learning_rate": 2.0202312138728324e-05,
+      "loss": 0.3603,
+      "step": 1224
+    },
+    {
+      "epoch": 1.9088074824629775,
+      "grad_norm": 0.2523008465363555,
+      "learning_rate": 2.017341040462428e-05,
+      "loss": 0.3724,
+      "step": 1225
+    },
+    {
+      "epoch": 1.9103663289166017,
+      "grad_norm": 0.2763374659384913,
+      "learning_rate": 2.014450867052023e-05,
+      "loss": 0.3548,
+      "step": 1226
+    },
+    {
+      "epoch": 1.911925175370226,
+      "grad_norm": 0.24964103816057576,
+      "learning_rate": 2.0115606936416188e-05,
+      "loss": 0.3694,
+      "step": 1227
+    },
+    {
+      "epoch": 1.9134840218238502,
+      "grad_norm": 0.28584599942242367,
+      "learning_rate": 2.0086705202312138e-05,
+      "loss": 0.3644,
+      "step": 1228
+    },
+    {
+      "epoch": 1.9150428682774747,
+      "grad_norm": 0.253224234190821,
+      "learning_rate": 2.0057803468208095e-05,
+      "loss": 0.3651,
+      "step": 1229
+    },
+    {
+      "epoch": 1.916601714731099,
+      "grad_norm": 0.2853460997991044,
+      "learning_rate": 2.0028901734104048e-05,
+      "loss": 0.3865,
+      "step": 1230
+    },
+    {
+      "epoch": 1.9181605611847234,
+      "grad_norm": 0.2914953043711197,
+      "learning_rate": 2e-05,
+      "loss": 0.3705,
+      "step": 1231
+    },
+    {
+      "epoch": 1.9197194076383477,
+      "grad_norm": 0.2703628033375394,
+      "learning_rate": 1.9971098265895955e-05,
+      "loss": 0.35,
+      "step": 1232
+    },
+    {
+      "epoch": 1.921278254091972,
+      "grad_norm": 0.3403993339455389,
+      "learning_rate": 1.994219653179191e-05,
+      "loss": 0.367,
+      "step": 1233
+    },
+    {
+      "epoch": 1.9228371005455962,
+      "grad_norm": 0.2403088129031985,
+      "learning_rate": 1.9913294797687862e-05,
+      "loss": 0.3554,
+      "step": 1234
+    },
+    {
+      "epoch": 1.9243959469992205,
+      "grad_norm": 0.2811341273215568,
+      "learning_rate": 1.988439306358382e-05,
+      "loss": 0.3617,
+      "step": 1235
+    },
+    {
+      "epoch": 1.925954793452845,
+      "grad_norm": 0.2878197944500182,
+      "learning_rate": 1.985549132947977e-05,
+      "loss": 0.3574,
+      "step": 1236
+    },
+    {
+      "epoch": 1.9275136399064692,
+      "grad_norm": 0.2774310293587101,
+      "learning_rate": 1.9826589595375726e-05,
+      "loss": 0.3748,
+      "step": 1237
+    },
+    {
+      "epoch": 1.9290724863600937,
+      "grad_norm": 0.3115605490343856,
+      "learning_rate": 1.9797687861271676e-05,
+      "loss": 0.353,
+      "step": 1238
+    },
+    {
+      "epoch": 1.930631332813718,
+      "grad_norm": 0.25590953954693557,
+      "learning_rate": 1.9768786127167633e-05,
+      "loss": 0.3662,
+      "step": 1239
+    },
+    {
+      "epoch": 1.9321901792673422,
+      "grad_norm": 0.2555007229781606,
+      "learning_rate": 1.9739884393063586e-05,
+      "loss": 0.3494,
+      "step": 1240
+    },
+    {
+      "epoch": 1.9337490257209664,
+      "grad_norm": 0.2691775864628584,
+      "learning_rate": 1.971098265895954e-05,
+      "loss": 0.3695,
+      "step": 1241
+    },
+    {
+      "epoch": 1.9353078721745907,
+      "grad_norm": 0.27904072318741047,
+      "learning_rate": 1.9682080924855493e-05,
+      "loss": 0.3553,
+      "step": 1242
+    },
+    {
+      "epoch": 1.9368667186282151,
+      "grad_norm": 0.2940422073737695,
+      "learning_rate": 1.9653179190751446e-05,
+      "loss": 0.3732,
+      "step": 1243
+    },
+    {
+      "epoch": 1.9384255650818394,
+      "grad_norm": 0.28761825900324317,
+      "learning_rate": 1.96242774566474e-05,
+      "loss": 0.3739,
+      "step": 1244
+    },
+    {
+      "epoch": 1.9399844115354639,
+      "grad_norm": 0.30744221511987735,
+      "learning_rate": 1.9595375722543353e-05,
+      "loss": 0.3598,
+      "step": 1245
+    },
+    {
+      "epoch": 1.9415432579890881,
+      "grad_norm": 0.2669547151465641,
+      "learning_rate": 1.9566473988439307e-05,
+      "loss": 0.3625,
+      "step": 1246
+    },
+    {
+      "epoch": 1.9431021044427124,
+      "grad_norm": 0.3163867104214392,
+      "learning_rate": 1.9537572254335264e-05,
+      "loss": 0.3527,
+      "step": 1247
+    },
+    {
+      "epoch": 1.9446609508963366,
+      "grad_norm": 0.2724508920694651,
+      "learning_rate": 1.9508670520231214e-05,
+      "loss": 0.349,
+      "step": 1248
+    },
+    {
+      "epoch": 1.9462197973499609,
+      "grad_norm": 0.31987516157242796,
+      "learning_rate": 1.947976878612717e-05,
+      "loss": 0.3578,
+      "step": 1249
+    },
+    {
+      "epoch": 1.9477786438035853,
+      "grad_norm": 0.27194855512008154,
+      "learning_rate": 1.9450867052023124e-05,
+      "loss": 0.3717,
+      "step": 1250
+    },
+    {
+      "epoch": 1.9493374902572098,
+      "grad_norm": 0.42931537141861814,
+      "learning_rate": 1.9421965317919074e-05,
+      "loss": 0.3762,
+      "step": 1251
+    },
+    {
+      "epoch": 1.950896336710834,
+      "grad_norm": 0.263639511648232,
+      "learning_rate": 1.939306358381503e-05,
+      "loss": 0.37,
+      "step": 1252
+    },
+    {
+      "epoch": 1.9524551831644583,
+      "grad_norm": 0.3046013215628072,
+      "learning_rate": 1.936416184971098e-05,
+      "loss": 0.3657,
+      "step": 1253
+    },
+    {
+      "epoch": 1.9540140296180826,
+      "grad_norm": 0.27822048356909573,
+      "learning_rate": 1.9335260115606938e-05,
+      "loss": 0.3702,
+      "step": 1254
+    },
+    {
+      "epoch": 1.9555728760717068,
+      "grad_norm": 0.2503963888170451,
+      "learning_rate": 1.930635838150289e-05,
+      "loss": 0.3533,
+      "step": 1255
+    },
+    {
+      "epoch": 1.9571317225253313,
+      "grad_norm": 0.3608440008147506,
+      "learning_rate": 1.9277456647398845e-05,
+      "loss": 0.3617,
+      "step": 1256
+    },
+    {
+      "epoch": 1.9586905689789555,
+      "grad_norm": 0.2618137124636811,
+      "learning_rate": 1.9248554913294798e-05,
+      "loss": 0.3536,
+      "step": 1257
+    },
+    {
+      "epoch": 1.96024941543258,
+      "grad_norm": 0.31638615142182935,
+      "learning_rate": 1.921965317919075e-05,
+      "loss": 0.3646,
+      "step": 1258
+    },
+    {
+      "epoch": 1.9618082618862043,
+      "grad_norm": 0.2902981949466536,
+      "learning_rate": 1.9190751445086705e-05,
+      "loss": 0.3576,
+      "step": 1259
+    },
+    {
+      "epoch": 1.9633671083398285,
+      "grad_norm": 0.24681016366367486,
+      "learning_rate": 1.916184971098266e-05,
+      "loss": 0.37,
+      "step": 1260
+    },
+    {
+      "epoch": 1.9649259547934528,
+      "grad_norm": 0.30886035269615025,
+      "learning_rate": 1.9132947976878612e-05,
+      "loss": 0.3735,
+      "step": 1261
+    },
+    {
+      "epoch": 1.966484801247077,
+      "grad_norm": 0.3261707681560328,
+      "learning_rate": 1.910404624277457e-05,
+      "loss": 0.3802,
+      "step": 1262
+    },
+    {
+      "epoch": 1.9680436477007015,
+      "grad_norm": 0.27343056543520194,
+      "learning_rate": 1.907514450867052e-05,
+      "loss": 0.3689,
+      "step": 1263
+    },
+    {
+      "epoch": 1.9696024941543258,
+      "grad_norm": 0.2919468083028249,
+      "learning_rate": 1.9046242774566476e-05,
+      "loss": 0.3599,
+      "step": 1264
+    },
+    {
+      "epoch": 1.9711613406079502,
+      "grad_norm": 0.2994649798188123,
+      "learning_rate": 1.901734104046243e-05,
+      "loss": 0.3713,
+      "step": 1265
+    },
+    {
+      "epoch": 1.9727201870615745,
+      "grad_norm": 0.3564440516448399,
+      "learning_rate": 1.8988439306358382e-05,
+      "loss": 0.3451,
+      "step": 1266
+    },
+    {
+      "epoch": 1.9742790335151987,
+      "grad_norm": 0.28761498604656505,
+      "learning_rate": 1.8959537572254336e-05,
+      "loss": 0.3691,
+      "step": 1267
+    },
+    {
+      "epoch": 1.975837879968823,
+      "grad_norm": 0.29458672500997857,
+      "learning_rate": 1.893063583815029e-05,
+      "loss": 0.3694,
+      "step": 1268
+    },
+    {
+      "epoch": 1.9773967264224472,
+      "grad_norm": 0.28968659606971875,
+      "learning_rate": 1.8901734104046243e-05,
+      "loss": 0.3757,
+      "step": 1269
+    },
+    {
+      "epoch": 1.9789555728760717,
+      "grad_norm": 0.25074130760016594,
+      "learning_rate": 1.8872832369942196e-05,
+      "loss": 0.3671,
+      "step": 1270
+    },
+    {
+      "epoch": 1.9805144193296962,
+      "grad_norm": 0.2727775661900619,
+      "learning_rate": 1.884393063583815e-05,
+      "loss": 0.3763,
+      "step": 1271
+    },
+    {
+      "epoch": 1.9820732657833204,
+      "grad_norm": 0.2540694971749628,
+      "learning_rate": 1.8815028901734106e-05,
+      "loss": 0.361,
+      "step": 1272
+    },
+    {
+      "epoch": 1.9836321122369447,
+      "grad_norm": 0.26620323560623,
+      "learning_rate": 1.8786127167630057e-05,
+      "loss": 0.3804,
+      "step": 1273
+    },
+    {
+      "epoch": 1.985190958690569,
+      "grad_norm": 0.2770753239201809,
+      "learning_rate": 1.8757225433526013e-05,
+      "loss": 0.3511,
+      "step": 1274
+    },
+    {
+      "epoch": 1.9867498051441932,
+      "grad_norm": 0.28316981515390427,
+      "learning_rate": 1.8728323699421967e-05,
+      "loss": 0.3527,
+      "step": 1275
+    },
+    {
+      "epoch": 1.9883086515978177,
+      "grad_norm": 0.24806656551525558,
+      "learning_rate": 1.869942196531792e-05,
+      "loss": 0.3888,
+      "step": 1276
+    },
+    {
+      "epoch": 1.989867498051442,
+      "grad_norm": 0.2751522114384385,
+      "learning_rate": 1.8670520231213874e-05,
+      "loss": 0.3591,
+      "step": 1277
+    },
+    {
+      "epoch": 1.9914263445050664,
+      "grad_norm": 0.29543750469737046,
+      "learning_rate": 1.8641618497109827e-05,
+      "loss": 0.3604,
+      "step": 1278
+    },
+    {
+      "epoch": 1.9929851909586906,
+      "grad_norm": 0.27104672954069825,
+      "learning_rate": 1.861271676300578e-05,
+      "loss": 0.3686,
+      "step": 1279
+    },
+    {
+      "epoch": 1.994544037412315,
+      "grad_norm": 0.26770020711721565,
+      "learning_rate": 1.8583815028901734e-05,
+      "loss": 0.3919,
+      "step": 1280
+    },
+    {
+      "epoch": 1.9961028838659391,
+      "grad_norm": 0.3002772154980313,
+      "learning_rate": 1.8554913294797688e-05,
+      "loss": 0.378,
+      "step": 1281
+    },
+    {
+      "epoch": 1.9976617303195634,
+      "grad_norm": 0.2934928063887189,
+      "learning_rate": 1.8526011560693644e-05,
+      "loss": 0.3851,
+      "step": 1282
+    },
+    {
+      "epoch": 1.9992205767731879,
+      "grad_norm": 0.26071286159069573,
+      "learning_rate": 1.8497109826589594e-05,
+      "loss": 0.3661,
+      "step": 1283
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.2407353049559013,
+      "learning_rate": 1.846820809248555e-05,
+      "loss": 0.3315,
+      "step": 1284
+    },
+    {
+      "epoch": 2.0015588464536243,
+      "grad_norm": 0.3019354121209205,
+      "learning_rate": 1.84393063583815e-05,
+      "loss": 0.3132,
+      "step": 1285
+    },
+    {
+      "epoch": 2.0031176929072485,
+      "grad_norm": 0.3221654485813362,
+      "learning_rate": 1.8410404624277458e-05,
+      "loss": 0.3049,
+      "step": 1286
+    },
+    {
+      "epoch": 2.0046765393608728,
+      "grad_norm": 0.3107212962226372,
+      "learning_rate": 1.838150289017341e-05,
+      "loss": 0.2962,
+      "step": 1287
+    },
+    {
+      "epoch": 2.0062353858144975,
+      "grad_norm": 0.2652451153347466,
+      "learning_rate": 1.8352601156069365e-05,
+      "loss": 0.2848,
+      "step": 1288
+    },
+    {
+      "epoch": 2.0077942322681217,
+      "grad_norm": 0.29115331241478903,
+      "learning_rate": 1.832369942196532e-05,
+      "loss": 0.2974,
+      "step": 1289
+    },
+    {
+      "epoch": 2.009353078721746,
+      "grad_norm": 0.29937191649859674,
+      "learning_rate": 1.8294797687861272e-05,
+      "loss": 0.28,
+      "step": 1290
+    },
+    {
+      "epoch": 2.01091192517537,
+      "grad_norm": 0.3181952793594262,
+      "learning_rate": 1.8265895953757225e-05,
+      "loss": 0.2795,
+      "step": 1291
+    },
+    {
+      "epoch": 2.0124707716289945,
+      "grad_norm": 0.3030873579967208,
+      "learning_rate": 1.8236994219653182e-05,
+      "loss": 0.3004,
+      "step": 1292
+    },
+    {
+      "epoch": 2.0140296180826187,
+      "grad_norm": 0.30093086334228963,
+      "learning_rate": 1.8208092485549132e-05,
+      "loss": 0.2987,
+      "step": 1293
+    },
+    {
+      "epoch": 2.015588464536243,
+      "grad_norm": 0.31992474845831625,
+      "learning_rate": 1.817919075144509e-05,
+      "loss": 0.2921,
+      "step": 1294
+    },
+    {
+      "epoch": 2.0171473109898677,
+      "grad_norm": 0.31971945976797334,
+      "learning_rate": 1.815028901734104e-05,
+      "loss": 0.2938,
+      "step": 1295
+    },
+    {
+      "epoch": 2.018706157443492,
+      "grad_norm": 0.2784528999275469,
+      "learning_rate": 1.8121387283236996e-05,
+      "loss": 0.2969,
+      "step": 1296
+    },
+    {
+      "epoch": 2.020265003897116,
+      "grad_norm": 0.28547739386324084,
+      "learning_rate": 1.809248554913295e-05,
+      "loss": 0.2819,
+      "step": 1297
+    },
+    {
+      "epoch": 2.0218238503507404,
+      "grad_norm": 0.35064000763654246,
+      "learning_rate": 1.8063583815028903e-05,
+      "loss": 0.2886,
+      "step": 1298
+    },
+    {
+      "epoch": 2.0233826968043647,
+      "grad_norm": 0.298377942201609,
+      "learning_rate": 1.8034682080924856e-05,
+      "loss": 0.2888,
+      "step": 1299
+    },
+    {
+      "epoch": 2.024941543257989,
+      "grad_norm": 0.2741254018613014,
+      "learning_rate": 1.800578034682081e-05,
+      "loss": 0.2846,
+      "step": 1300
+    },
+    {
+      "epoch": 2.0265003897116136,
+      "grad_norm": 0.3278384877642372,
+      "learning_rate": 1.7976878612716763e-05,
+      "loss": 0.2996,
+      "step": 1301
+    },
+    {
+      "epoch": 2.028059236165238,
+      "grad_norm": 0.31062098942293614,
+      "learning_rate": 1.794797687861272e-05,
+      "loss": 0.3004,
+      "step": 1302
+    },
+    {
+      "epoch": 2.029618082618862,
+      "grad_norm": 0.2840675944649226,
+      "learning_rate": 1.791907514450867e-05,
+      "loss": 0.2915,
+      "step": 1303
+    },
+    {
+      "epoch": 2.0311769290724864,
+      "grad_norm": 0.3066133656454931,
+      "learning_rate": 1.7890173410404627e-05,
+      "loss": 0.2887,
+      "step": 1304
+    },
+    {
+      "epoch": 2.0327357755261106,
+      "grad_norm": 0.2729351456670734,
+      "learning_rate": 1.7861271676300577e-05,
+      "loss": 0.2984,
+      "step": 1305
+    },
+    {
+      "epoch": 2.034294621979735,
+      "grad_norm": 0.2636535310871996,
+      "learning_rate": 1.7832369942196534e-05,
+      "loss": 0.291,
+      "step": 1306
+    },
+    {
+      "epoch": 2.035853468433359,
+      "grad_norm": 0.2964005719112254,
+      "learning_rate": 1.7803468208092487e-05,
+      "loss": 0.2878,
+      "step": 1307
+    },
+    {
+      "epoch": 2.037412314886984,
+      "grad_norm": 0.2362770470391929,
+      "learning_rate": 1.777456647398844e-05,
+      "loss": 0.2889,
+      "step": 1308
+    },
+    {
+      "epoch": 2.038971161340608,
+      "grad_norm": 0.2877360765498249,
+      "learning_rate": 1.7745664739884394e-05,
+      "loss": 0.306,
+      "step": 1309
+    },
+    {
+      "epoch": 2.0405300077942323,
+      "grad_norm": 0.26635846301125965,
+      "learning_rate": 1.7716763005780348e-05,
+      "loss": 0.2835,
+      "step": 1310
+    },
+    {
+      "epoch": 2.0420888542478566,
+      "grad_norm": 0.2364083261820846,
+      "learning_rate": 1.76878612716763e-05,
+      "loss": 0.2813,
+      "step": 1311
+    },
+    {
+      "epoch": 2.043647700701481,
+      "grad_norm": 0.25071829512152555,
+      "learning_rate": 1.7658959537572255e-05,
+      "loss": 0.2722,
+      "step": 1312
+    },
+    {
+      "epoch": 2.045206547155105,
+      "grad_norm": 0.2640068234739541,
+      "learning_rate": 1.7630057803468208e-05,
+      "loss": 0.3078,
+      "step": 1313
+    },
+    {
+      "epoch": 2.0467653936087293,
+      "grad_norm": 0.2733838421400439,
+      "learning_rate": 1.7601156069364165e-05,
+      "loss": 0.2936,
+      "step": 1314
+    },
+    {
+      "epoch": 2.048324240062354,
+      "grad_norm": 0.2550755795463892,
+      "learning_rate": 1.7572254335260115e-05,
+      "loss": 0.2819,
+      "step": 1315
+    },
+    {
+      "epoch": 2.0498830865159783,
+      "grad_norm": 0.25545376074388176,
+      "learning_rate": 1.754335260115607e-05,
+      "loss": 0.2823,
+      "step": 1316
+    },
+    {
+      "epoch": 2.0514419329696025,
+      "grad_norm": 0.22661427370917822,
+      "learning_rate": 1.7514450867052025e-05,
+      "loss": 0.2824,
+      "step": 1317
+    },
+    {
+      "epoch": 2.053000779423227,
+      "grad_norm": 0.29774507660120497,
+      "learning_rate": 1.748554913294798e-05,
+      "loss": 0.2838,
+      "step": 1318
+    },
+    {
+      "epoch": 2.054559625876851,
+      "grad_norm": 0.2598671788363091,
+      "learning_rate": 1.7456647398843932e-05,
+      "loss": 0.2903,
+      "step": 1319
+    },
+    {
+      "epoch": 2.0561184723304753,
+      "grad_norm": 0.2519961479463723,
+      "learning_rate": 1.7427745664739885e-05,
+      "loss": 0.2571,
+      "step": 1320
+    },
+    {
+      "epoch": 2.0576773187841,
+      "grad_norm": 0.25326241935804067,
+      "learning_rate": 1.739884393063584e-05,
+      "loss": 0.2977,
+      "step": 1321
+    },
+    {
+      "epoch": 2.0592361652377242,
+      "grad_norm": 0.24057089058554867,
+      "learning_rate": 1.7369942196531792e-05,
+      "loss": 0.2866,
+      "step": 1322
+    },
+    {
+      "epoch": 2.0607950116913485,
+      "grad_norm": 0.22295403456879342,
+      "learning_rate": 1.7341040462427746e-05,
+      "loss": 0.28,
+      "step": 1323
+    },
+    {
+      "epoch": 2.0623538581449727,
+      "grad_norm": 0.2868118439567236,
+      "learning_rate": 1.7312138728323703e-05,
+      "loss": 0.3044,
+      "step": 1324
+    },
+    {
+      "epoch": 2.063912704598597,
+      "grad_norm": 0.2708711985497848,
+      "learning_rate": 1.7283236994219653e-05,
+      "loss": 0.2823,
+      "step": 1325
+    },
+    {
+      "epoch": 2.0654715510522212,
+      "grad_norm": 0.42537603462611046,
+      "learning_rate": 1.725433526011561e-05,
+      "loss": 0.2984,
+      "step": 1326
+    },
+    {
+      "epoch": 2.0670303975058455,
+      "grad_norm": 0.25936897272013576,
+      "learning_rate": 1.722543352601156e-05,
+      "loss": 0.2996,
+      "step": 1327
+    },
+    {
+      "epoch": 2.06858924395947,
+      "grad_norm": 0.29205530294187826,
+      "learning_rate": 1.7196531791907516e-05,
+      "loss": 0.2815,
+      "step": 1328
+    },
+    {
+      "epoch": 2.0701480904130944,
+      "grad_norm": 0.2498440090336756,
+      "learning_rate": 1.716763005780347e-05,
+      "loss": 0.2908,
+      "step": 1329
+    },
+    {
+      "epoch": 2.0717069368667187,
+      "grad_norm": 0.2803332912261779,
+      "learning_rate": 1.7138728323699423e-05,
+      "loss": 0.2919,
+      "step": 1330
+    },
+    {
+      "epoch": 2.073265783320343,
+      "grad_norm": 0.2714375808140227,
+      "learning_rate": 1.7109826589595377e-05,
+      "loss": 0.285,
+      "step": 1331
+    },
+    {
+      "epoch": 2.074824629773967,
+      "grad_norm": 0.23047443561287922,
+      "learning_rate": 1.708092485549133e-05,
+      "loss": 0.2982,
+      "step": 1332
+    },
+    {
+      "epoch": 2.0763834762275915,
+      "grad_norm": 0.27974401775468655,
+      "learning_rate": 1.7052023121387284e-05,
+      "loss": 0.2816,
+      "step": 1333
+    },
+    {
+      "epoch": 2.0779423226812157,
+      "grad_norm": 0.27728883446303665,
+      "learning_rate": 1.702312138728324e-05,
+      "loss": 0.2874,
+      "step": 1334
+    },
+    {
+      "epoch": 2.0795011691348404,
+      "grad_norm": 0.22428181416464069,
+      "learning_rate": 1.699421965317919e-05,
+      "loss": 0.2959,
+      "step": 1335
+    },
+    {
+      "epoch": 2.0810600155884647,
+      "grad_norm": 0.4070173071535244,
+      "learning_rate": 1.6965317919075147e-05,
+      "loss": 0.2853,
+      "step": 1336
+    },
+    {
+      "epoch": 2.082618862042089,
+      "grad_norm": 0.33199185273932175,
+      "learning_rate": 1.6936416184971097e-05,
+      "loss": 0.2825,
+      "step": 1337
+    },
+    {
+      "epoch": 2.084177708495713,
+      "grad_norm": 0.23889026155636925,
+      "learning_rate": 1.6907514450867054e-05,
+      "loss": 0.2878,
+      "step": 1338
+    },
+    {
+      "epoch": 2.0857365549493374,
+      "grad_norm": 0.2588106933535315,
+      "learning_rate": 1.6878612716763008e-05,
+      "loss": 0.291,
+      "step": 1339
+    },
+    {
+      "epoch": 2.0872954014029617,
+      "grad_norm": 0.2521130177780321,
+      "learning_rate": 1.684971098265896e-05,
+      "loss": 0.2927,
+      "step": 1340
+    },
+    {
+      "epoch": 2.0888542478565864,
+      "grad_norm": 0.28556085649359164,
+      "learning_rate": 1.6820809248554915e-05,
+      "loss": 0.3141,
+      "step": 1341
+    },
+    {
+      "epoch": 2.0904130943102106,
+      "grad_norm": 0.23613222180841584,
+      "learning_rate": 1.6791907514450868e-05,
+      "loss": 0.2861,
+      "step": 1342
+    },
+    {
+      "epoch": 2.091971940763835,
+      "grad_norm": 0.24684587451326986,
+      "learning_rate": 1.676300578034682e-05,
+      "loss": 0.2853,
+      "step": 1343
+    },
+    {
+      "epoch": 2.093530787217459,
+      "grad_norm": 0.23325395667135312,
+      "learning_rate": 1.673410404624278e-05,
+      "loss": 0.2908,
+      "step": 1344
+    },
+    {
+      "epoch": 2.0950896336710834,
+      "grad_norm": 0.2594871779004086,
+      "learning_rate": 1.670520231213873e-05,
+      "loss": 0.3087,
+      "step": 1345
+    },
+    {
+      "epoch": 2.0966484801247076,
+      "grad_norm": 0.23321943315628127,
+      "learning_rate": 1.6676300578034685e-05,
+      "loss": 0.2958,
+      "step": 1346
+    },
+    {
+      "epoch": 2.098207326578332,
+      "grad_norm": 0.23482451299773394,
+      "learning_rate": 1.6647398843930635e-05,
+      "loss": 0.3037,
+      "step": 1347
+    },
+    {
+      "epoch": 2.0997661730319566,
+      "grad_norm": 0.2237656119837914,
+      "learning_rate": 1.661849710982659e-05,
+      "loss": 0.2788,
+      "step": 1348
+    },
+    {
+      "epoch": 2.101325019485581,
+      "grad_norm": 0.25698435105220185,
+      "learning_rate": 1.6589595375722546e-05,
+      "loss": 0.2747,
+      "step": 1349
+    },
+    {
+      "epoch": 2.102883865939205,
+      "grad_norm": 0.23963767268717012,
+      "learning_rate": 1.6560693641618496e-05,
+      "loss": 0.292,
+      "step": 1350
+    },
+    {
+      "epoch": 2.1044427123928293,
+      "grad_norm": 0.2588056696749546,
+      "learning_rate": 1.6531791907514452e-05,
+      "loss": 0.3035,
+      "step": 1351
+    },
+    {
+      "epoch": 2.1060015588464536,
+      "grad_norm": 0.25164368263083836,
+      "learning_rate": 1.6502890173410403e-05,
+      "loss": 0.297,
+      "step": 1352
+    },
+    {
+      "epoch": 2.107560405300078,
+      "grad_norm": 0.2570592630746674,
+      "learning_rate": 1.647398843930636e-05,
+      "loss": 0.2982,
+      "step": 1353
+    },
+    {
+      "epoch": 2.109119251753702,
+      "grad_norm": 0.2848315307917638,
+      "learning_rate": 1.6445086705202313e-05,
+      "loss": 0.3007,
+      "step": 1354
+    },
+    {
+      "epoch": 2.1106780982073268,
+      "grad_norm": 0.2634643134825361,
+      "learning_rate": 1.6416184971098266e-05,
+      "loss": 0.2818,
+      "step": 1355
+    },
+    {
+      "epoch": 2.112236944660951,
+      "grad_norm": 0.25838881698032484,
+      "learning_rate": 1.638728323699422e-05,
+      "loss": 0.2675,
+      "step": 1356
+    },
+    {
+      "epoch": 2.1137957911145753,
+      "grad_norm": 0.2438857581178678,
+      "learning_rate": 1.6358381502890173e-05,
+      "loss": 0.2931,
+      "step": 1357
+    },
+    {
+      "epoch": 2.1153546375681995,
+      "grad_norm": 0.23647456464356711,
+      "learning_rate": 1.6329479768786127e-05,
+      "loss": 0.2758,
+      "step": 1358
+    },
+    {
+      "epoch": 2.1169134840218238,
+      "grad_norm": 0.245819384339159,
+      "learning_rate": 1.6300578034682083e-05,
+      "loss": 0.2935,
+      "step": 1359
+    },
+    {
+      "epoch": 2.118472330475448,
+      "grad_norm": 0.21611647939557374,
+      "learning_rate": 1.6271676300578033e-05,
+      "loss": 0.2761,
+      "step": 1360
+    },
+    {
+      "epoch": 2.1200311769290723,
+      "grad_norm": 0.24308227386319844,
+      "learning_rate": 1.624277456647399e-05,
+      "loss": 0.2911,
+      "step": 1361
+    },
+    {
+      "epoch": 2.121590023382697,
+      "grad_norm": 0.28166564601668537,
+      "learning_rate": 1.621387283236994e-05,
+      "loss": 0.291,
+      "step": 1362
+    },
+    {
+      "epoch": 2.1231488698363212,
+      "grad_norm": 0.2333910024555345,
+      "learning_rate": 1.6184971098265897e-05,
+      "loss": 0.282,
+      "step": 1363
+    },
+    {
+      "epoch": 2.1247077162899455,
+      "grad_norm": 0.24801232636156478,
+      "learning_rate": 1.615606936416185e-05,
+      "loss": 0.286,
+      "step": 1364
+    },
+    {
+      "epoch": 2.1262665627435697,
+      "grad_norm": 0.2529976131043813,
+      "learning_rate": 1.6127167630057804e-05,
+      "loss": 0.2907,
+      "step": 1365
+    },
+    {
+      "epoch": 2.127825409197194,
+      "grad_norm": 0.24496746678348746,
+      "learning_rate": 1.6098265895953758e-05,
+      "loss": 0.3096,
+      "step": 1366
+    },
+    {
+      "epoch": 2.1293842556508182,
+      "grad_norm": 0.22826596743843666,
+      "learning_rate": 1.606936416184971e-05,
+      "loss": 0.2837,
+      "step": 1367
+    },
+    {
+      "epoch": 2.1309431021044425,
+      "grad_norm": 0.22953989118587162,
+      "learning_rate": 1.6040462427745664e-05,
+      "loss": 0.3016,
+      "step": 1368
+    },
+    {
+      "epoch": 2.132501948558067,
+      "grad_norm": 0.24938076633148093,
+      "learning_rate": 1.601156069364162e-05,
+      "loss": 0.2938,
+      "step": 1369
+    },
+    {
+      "epoch": 2.1340607950116914,
+      "grad_norm": 0.21892102115489234,
+      "learning_rate": 1.598265895953757e-05,
+      "loss": 0.2815,
+      "step": 1370
+    },
+    {
+      "epoch": 2.1356196414653157,
+      "grad_norm": 0.2590092923549659,
+      "learning_rate": 1.5953757225433528e-05,
+      "loss": 0.2954,
+      "step": 1371
+    },
+    {
+      "epoch": 2.13717848791894,
+      "grad_norm": 0.24890666085552407,
+      "learning_rate": 1.5924855491329478e-05,
+      "loss": 0.2906,
+      "step": 1372
+    },
+    {
+      "epoch": 2.138737334372564,
+      "grad_norm": 0.2266359560410207,
+      "learning_rate": 1.5895953757225435e-05,
+      "loss": 0.2817,
+      "step": 1373
+    },
+    {
+      "epoch": 2.1402961808261884,
+      "grad_norm": 0.24930550949426433,
+      "learning_rate": 1.586705202312139e-05,
+      "loss": 0.2902,
+      "step": 1374
+    },
+    {
+      "epoch": 2.141855027279813,
+      "grad_norm": 0.22270650770572073,
+      "learning_rate": 1.5838150289017342e-05,
+      "loss": 0.2963,
+      "step": 1375
+    },
+    {
+      "epoch": 2.1434138737334374,
+      "grad_norm": 0.21082837821856862,
+      "learning_rate": 1.5809248554913295e-05,
+      "loss": 0.2823,
+      "step": 1376
+    },
+    {
+      "epoch": 2.1449727201870616,
+      "grad_norm": 0.2506283104261599,
+      "learning_rate": 1.578034682080925e-05,
+      "loss": 0.3003,
+      "step": 1377
+    },
+    {
+      "epoch": 2.146531566640686,
+      "grad_norm": 0.2389054933616818,
+      "learning_rate": 1.5751445086705202e-05,
+      "loss": 0.2885,
+      "step": 1378
+    },
+    {
+      "epoch": 2.14809041309431,
+      "grad_norm": 0.2334569505339403,
+      "learning_rate": 1.5722543352601156e-05,
+      "loss": 0.3052,
+      "step": 1379
+    },
+    {
+      "epoch": 2.1496492595479344,
+      "grad_norm": 0.23175590298956983,
+      "learning_rate": 1.569364161849711e-05,
+      "loss": 0.2739,
+      "step": 1380
+    },
+    {
+      "epoch": 2.1512081060015587,
+      "grad_norm": 0.23033041297872123,
+      "learning_rate": 1.5664739884393066e-05,
+      "loss": 0.286,
+      "step": 1381
+    },
+    {
+      "epoch": 2.1527669524551833,
+      "grad_norm": 0.2465779921606415,
+      "learning_rate": 1.5635838150289016e-05,
+      "loss": 0.3009,
+      "step": 1382
+    },
+    {
+      "epoch": 2.1543257989088076,
+      "grad_norm": 0.21179227741678436,
+      "learning_rate": 1.5606936416184973e-05,
+      "loss": 0.2847,
+      "step": 1383
+    },
+    {
+      "epoch": 2.155884645362432,
+      "grad_norm": 0.25428536535374435,
+      "learning_rate": 1.5578034682080926e-05,
+      "loss": 0.2954,
+      "step": 1384
+    },
+    {
+      "epoch": 2.157443491816056,
+      "grad_norm": 0.22930157749520183,
+      "learning_rate": 1.554913294797688e-05,
+      "loss": 0.2818,
+      "step": 1385
+    },
+    {
+      "epoch": 2.1590023382696804,
+      "grad_norm": 0.23105202022376808,
+      "learning_rate": 1.5520231213872833e-05,
+      "loss": 0.3053,
+      "step": 1386
+    },
+    {
+      "epoch": 2.1605611847233046,
+      "grad_norm": 0.23246069929983043,
+      "learning_rate": 1.5491329479768787e-05,
+      "loss": 0.2803,
+      "step": 1387
+    },
+    {
+      "epoch": 2.1621200311769293,
+      "grad_norm": 0.22704083693916596,
+      "learning_rate": 1.546242774566474e-05,
+      "loss": 0.2891,
+      "step": 1388
+    },
+    {
+      "epoch": 2.1636788776305536,
+      "grad_norm": 0.2540369139980788,
+      "learning_rate": 1.5433526011560694e-05,
+      "loss": 0.2847,
+      "step": 1389
+    },
+    {
+      "epoch": 2.165237724084178,
+      "grad_norm": 0.21737272977125724,
+      "learning_rate": 1.5404624277456647e-05,
+      "loss": 0.2681,
+      "step": 1390
+    },
+    {
+      "epoch": 2.166796570537802,
+      "grad_norm": 0.2515481263822816,
+      "learning_rate": 1.5375722543352604e-05,
+      "loss": 0.2763,
+      "step": 1391
+    },
+    {
+      "epoch": 2.1683554169914263,
+      "grad_norm": 0.26477321713964047,
+      "learning_rate": 1.5346820809248554e-05,
+      "loss": 0.2808,
+      "step": 1392
+    },
+    {
+      "epoch": 2.1699142634450506,
+      "grad_norm": 0.22854862492021247,
+      "learning_rate": 1.531791907514451e-05,
+      "loss": 0.2943,
+      "step": 1393
+    },
+    {
+      "epoch": 2.171473109898675,
+      "grad_norm": 0.27158405015797865,
+      "learning_rate": 1.528901734104046e-05,
+      "loss": 0.2925,
+      "step": 1394
+    },
+    {
+      "epoch": 2.1730319563522995,
+      "grad_norm": 0.2347337281042923,
+      "learning_rate": 1.5260115606936418e-05,
+      "loss": 0.2943,
+      "step": 1395
+    },
+    {
+      "epoch": 2.1745908028059238,
+      "grad_norm": 0.23975916222494428,
+      "learning_rate": 1.523121387283237e-05,
+      "loss": 0.2804,
+      "step": 1396
+    },
+    {
+      "epoch": 2.176149649259548,
+      "grad_norm": 0.22779116644041048,
+      "learning_rate": 1.5202312138728325e-05,
+      "loss": 0.2847,
+      "step": 1397
+    },
+    {
+      "epoch": 2.1777084957131723,
+      "grad_norm": 0.2576587147864288,
+      "learning_rate": 1.5173410404624278e-05,
+      "loss": 0.2996,
+      "step": 1398
+    },
+    {
+      "epoch": 2.1792673421667965,
+      "grad_norm": 0.21115000700822473,
+      "learning_rate": 1.5144508670520233e-05,
+      "loss": 0.2787,
+      "step": 1399
+    },
+    {
+      "epoch": 2.1808261886204208,
+      "grad_norm": 0.24108973804933587,
+      "learning_rate": 1.5115606936416185e-05,
+      "loss": 0.287,
+      "step": 1400
+    },
+    {
+      "epoch": 2.182385035074045,
+      "grad_norm": 0.24288289350228465,
+      "learning_rate": 1.508670520231214e-05,
+      "loss": 0.2852,
+      "step": 1401
+    },
+    {
+      "epoch": 2.1839438815276697,
+      "grad_norm": 0.23706474437913794,
+      "learning_rate": 1.5057803468208092e-05,
+      "loss": 0.2968,
+      "step": 1402
+    },
+    {
+      "epoch": 2.185502727981294,
+      "grad_norm": 0.2183208225549947,
+      "learning_rate": 1.5028901734104049e-05,
+      "loss": 0.2865,
+      "step": 1403
+    },
+    {
+      "epoch": 2.187061574434918,
+      "grad_norm": 0.24253894248682953,
+      "learning_rate": 1.5e-05,
+      "loss": 0.2847,
+      "step": 1404
+    },
+    {
+      "epoch": 2.1886204208885425,
+      "grad_norm": 0.22598056970641814,
+      "learning_rate": 1.4971098265895956e-05,
+      "loss": 0.2982,
+      "step": 1405
+    },
+    {
+      "epoch": 2.1901792673421667,
+      "grad_norm": 0.2525617645303132,
+      "learning_rate": 1.4942196531791907e-05,
+      "loss": 0.2861,
+      "step": 1406
+    },
+    {
+      "epoch": 2.191738113795791,
+      "grad_norm": 0.2192754370403558,
+      "learning_rate": 1.4913294797687862e-05,
+      "loss": 0.2872,
+      "step": 1407
+    },
+    {
+      "epoch": 2.1932969602494152,
+      "grad_norm": 0.22900250201312447,
+      "learning_rate": 1.4884393063583816e-05,
+      "loss": 0.2845,
+      "step": 1408
+    },
+    {
+      "epoch": 2.19485580670304,
+      "grad_norm": 0.30054519236450217,
+      "learning_rate": 1.4855491329479771e-05,
+      "loss": 0.2943,
+      "step": 1409
+    },
+    {
+      "epoch": 2.196414653156664,
+      "grad_norm": 0.24441363373477734,
+      "learning_rate": 1.4826589595375723e-05,
+      "loss": 0.2975,
+      "step": 1410
+    },
+    {
+      "epoch": 2.1979734996102884,
+      "grad_norm": 0.2267613912682233,
+      "learning_rate": 1.4797687861271678e-05,
+      "loss": 0.3009,
+      "step": 1411
+    },
+    {
+      "epoch": 2.1995323460639127,
+      "grad_norm": 0.21100387037469506,
+      "learning_rate": 1.476878612716763e-05,
+      "loss": 0.2971,
+      "step": 1412
+    },
+    {
+      "epoch": 2.201091192517537,
+      "grad_norm": 0.21361711019848872,
+      "learning_rate": 1.4739884393063585e-05,
+      "loss": 0.2986,
+      "step": 1413
+    },
+    {
+      "epoch": 2.202650038971161,
+      "grad_norm": 0.23486378364056842,
+      "learning_rate": 1.4710982658959538e-05,
+      "loss": 0.2983,
+      "step": 1414
+    },
+    {
+      "epoch": 2.2042088854247854,
+      "grad_norm": 0.2194691571773862,
+      "learning_rate": 1.4682080924855493e-05,
+      "loss": 0.2805,
+      "step": 1415
+    },
+    {
+      "epoch": 2.20576773187841,
+      "grad_norm": 0.22839671067672832,
+      "learning_rate": 1.4653179190751445e-05,
+      "loss": 0.2821,
+      "step": 1416
+    },
+    {
+      "epoch": 2.2073265783320344,
+      "grad_norm": 0.22721582376446814,
+      "learning_rate": 1.46242774566474e-05,
+      "loss": 0.2883,
+      "step": 1417
+    },
+    {
+      "epoch": 2.2088854247856586,
+      "grad_norm": 0.21273654291505045,
+      "learning_rate": 1.4595375722543354e-05,
+      "loss": 0.2871,
+      "step": 1418
+    },
+    {
+      "epoch": 2.210444271239283,
+      "grad_norm": 0.22181506285423208,
+      "learning_rate": 1.4566473988439309e-05,
+      "loss": 0.2937,
+      "step": 1419
+    },
+    {
+      "epoch": 2.212003117692907,
+      "grad_norm": 0.21682833270568624,
+      "learning_rate": 1.453757225433526e-05,
+      "loss": 0.295,
+      "step": 1420
+    },
+    {
+      "epoch": 2.2135619641465314,
+      "grad_norm": 0.2544117406843311,
+      "learning_rate": 1.4508670520231216e-05,
+      "loss": 0.2959,
+      "step": 1421
+    },
+    {
+      "epoch": 2.215120810600156,
+      "grad_norm": 0.2496480472797846,
+      "learning_rate": 1.4479768786127167e-05,
+      "loss": 0.2955,
+      "step": 1422
+    },
+    {
+      "epoch": 2.2166796570537803,
+      "grad_norm": 0.25230712728942756,
+      "learning_rate": 1.4450867052023123e-05,
+      "loss": 0.2834,
+      "step": 1423
+    },
+    {
+      "epoch": 2.2182385035074046,
+      "grad_norm": 0.22497012216327553,
+      "learning_rate": 1.4421965317919076e-05,
+      "loss": 0.292,
+      "step": 1424
+    },
+    {
+      "epoch": 2.219797349961029,
+      "grad_norm": 0.2374016981159356,
+      "learning_rate": 1.4393063583815031e-05,
+      "loss": 0.2825,
+      "step": 1425
+    },
+    {
+      "epoch": 2.221356196414653,
+      "grad_norm": 0.25465184050475065,
+      "learning_rate": 1.4364161849710983e-05,
+      "loss": 0.2903,
+      "step": 1426
+    },
+    {
+      "epoch": 2.2229150428682773,
+      "grad_norm": 0.23870397107279917,
+      "learning_rate": 1.4335260115606938e-05,
+      "loss": 0.2936,
+      "step": 1427
+    },
+    {
+      "epoch": 2.2244738893219016,
+      "grad_norm": 0.2318933878119134,
+      "learning_rate": 1.430635838150289e-05,
+      "loss": 0.2849,
+      "step": 1428
+    },
+    {
+      "epoch": 2.2260327357755263,
+      "grad_norm": 0.23210851294757337,
+      "learning_rate": 1.4277456647398845e-05,
+      "loss": 0.2919,
+      "step": 1429
+    },
+    {
+      "epoch": 2.2275915822291505,
+      "grad_norm": 0.20510591201151088,
+      "learning_rate": 1.4248554913294798e-05,
+      "loss": 0.291,
+      "step": 1430
+    },
+    {
+      "epoch": 2.229150428682775,
+      "grad_norm": 0.2254853152020144,
+      "learning_rate": 1.4219653179190754e-05,
+      "loss": 0.2855,
+      "step": 1431
+    },
+    {
+      "epoch": 2.230709275136399,
+      "grad_norm": 0.22024099508735234,
+      "learning_rate": 1.4190751445086705e-05,
+      "loss": 0.2927,
+      "step": 1432
+    },
+    {
+      "epoch": 2.2322681215900233,
+      "grad_norm": 0.26043420023310093,
+      "learning_rate": 1.416184971098266e-05,
+      "loss": 0.3052,
+      "step": 1433
+    },
+    {
+      "epoch": 2.2338269680436476,
+      "grad_norm": 0.2133056223198868,
+      "learning_rate": 1.4132947976878614e-05,
+      "loss": 0.2927,
+      "step": 1434
+    },
+    {
+      "epoch": 2.2353858144972722,
+      "grad_norm": 0.2259340299782071,
+      "learning_rate": 1.4104046242774569e-05,
+      "loss": 0.2857,
+      "step": 1435
+    },
+    {
+      "epoch": 2.2369446609508965,
+      "grad_norm": 0.22200399124864398,
+      "learning_rate": 1.407514450867052e-05,
+      "loss": 0.2833,
+      "step": 1436
+    },
+    {
+      "epoch": 2.2385035074045208,
+      "grad_norm": 0.22962299042832857,
+      "learning_rate": 1.4046242774566476e-05,
+      "loss": 0.2799,
+      "step": 1437
+    },
+    {
+      "epoch": 2.240062353858145,
+      "grad_norm": 0.21438991547193448,
+      "learning_rate": 1.4017341040462428e-05,
+      "loss": 0.2859,
+      "step": 1438
+    },
+    {
+      "epoch": 2.2416212003117693,
+      "grad_norm": 0.22069076358373807,
+      "learning_rate": 1.3988439306358383e-05,
+      "loss": 0.2902,
+      "step": 1439
+    },
+    {
+      "epoch": 2.2431800467653935,
+      "grad_norm": 0.21552495322148416,
+      "learning_rate": 1.3959537572254336e-05,
+      "loss": 0.2717,
+      "step": 1440
+    },
+    {
+      "epoch": 2.2447388932190178,
+      "grad_norm": 0.22292189496523734,
+      "learning_rate": 1.3930635838150291e-05,
+      "loss": 0.2837,
+      "step": 1441
+    },
+    {
+      "epoch": 2.2462977396726425,
+      "grad_norm": 0.2235003322526891,
+      "learning_rate": 1.3901734104046243e-05,
+      "loss": 0.2785,
+      "step": 1442
+    },
+    {
+      "epoch": 2.2478565861262667,
+      "grad_norm": 0.21304911072284627,
+      "learning_rate": 1.3872832369942197e-05,
+      "loss": 0.2879,
+      "step": 1443
+    },
+    {
+      "epoch": 2.249415432579891,
+      "grad_norm": 0.25029914205006026,
+      "learning_rate": 1.3843930635838152e-05,
+      "loss": 0.2988,
+      "step": 1444
+    },
+    {
+      "epoch": 2.250974279033515,
+      "grad_norm": 0.2165276488818481,
+      "learning_rate": 1.3815028901734104e-05,
+      "loss": 0.272,
+      "step": 1445
+    },
+    {
+      "epoch": 2.2525331254871395,
+      "grad_norm": 0.24444489582018986,
+      "learning_rate": 1.3786127167630059e-05,
+      "loss": 0.3014,
+      "step": 1446
+    },
+    {
+      "epoch": 2.2540919719407637,
+      "grad_norm": 0.24771789944118036,
+      "learning_rate": 1.375722543352601e-05,
+      "loss": 0.2984,
+      "step": 1447
+    },
+    {
+      "epoch": 2.255650818394388,
+      "grad_norm": 0.23706904927685987,
+      "learning_rate": 1.3728323699421966e-05,
+      "loss": 0.3033,
+      "step": 1448
+    },
+    {
+      "epoch": 2.2572096648480127,
+      "grad_norm": 0.23835195294951741,
+      "learning_rate": 1.3699421965317919e-05,
+      "loss": 0.2774,
+      "step": 1449
+    },
+    {
+      "epoch": 2.258768511301637,
+      "grad_norm": 0.2492838137926293,
+      "learning_rate": 1.3670520231213874e-05,
+      "loss": 0.2843,
+      "step": 1450
+    },
+    {
+      "epoch": 2.260327357755261,
+      "grad_norm": 0.22427981907836916,
+      "learning_rate": 1.3641618497109826e-05,
+      "loss": 0.2862,
+      "step": 1451
+    },
+    {
+      "epoch": 2.2618862042088854,
+      "grad_norm": 0.2425076666411249,
+      "learning_rate": 1.3612716763005781e-05,
+      "loss": 0.2908,
+      "step": 1452
+    },
+    {
+      "epoch": 2.2634450506625097,
+      "grad_norm": 0.2686851435498773,
+      "learning_rate": 1.3583815028901733e-05,
+      "loss": 0.2853,
+      "step": 1453
+    },
+    {
+      "epoch": 2.265003897116134,
+      "grad_norm": 0.23258321364166276,
+      "learning_rate": 1.3554913294797688e-05,
+      "loss": 0.292,
+      "step": 1454
+    },
+    {
+      "epoch": 2.266562743569758,
+      "grad_norm": 0.2178252049659836,
+      "learning_rate": 1.3526011560693641e-05,
+      "loss": 0.2798,
+      "step": 1455
+    },
+    {
+      "epoch": 2.268121590023383,
+      "grad_norm": 0.23464644401450185,
+      "learning_rate": 1.3497109826589597e-05,
+      "loss": 0.279,
+      "step": 1456
+    },
+    {
+      "epoch": 2.269680436477007,
+      "grad_norm": 0.25551325960953597,
+      "learning_rate": 1.3468208092485548e-05,
+      "loss": 0.2997,
+      "step": 1457
+    },
+    {
+      "epoch": 2.2712392829306314,
+      "grad_norm": 0.2352887691568654,
+      "learning_rate": 1.3439306358381503e-05,
+      "loss": 0.2843,
+      "step": 1458
+    },
+    {
+      "epoch": 2.2727981293842556,
+      "grad_norm": 0.24160307618137192,
+      "learning_rate": 1.3410404624277457e-05,
+      "loss": 0.2979,
+      "step": 1459
+    },
+    {
+      "epoch": 2.27435697583788,
+      "grad_norm": 0.247635002623451,
+      "learning_rate": 1.3381502890173412e-05,
+      "loss": 0.2819,
+      "step": 1460
+    },
+    {
+      "epoch": 2.275915822291504,
+      "grad_norm": 0.2335426458405105,
+      "learning_rate": 1.3352601156069364e-05,
+      "loss": 0.2813,
+      "step": 1461
+    },
+    {
+      "epoch": 2.2774746687451284,
+      "grad_norm": 0.22663182041299806,
+      "learning_rate": 1.3323699421965319e-05,
+      "loss": 0.2981,
+      "step": 1462
+    },
+    {
+      "epoch": 2.279033515198753,
+      "grad_norm": 0.21171630873754166,
+      "learning_rate": 1.329479768786127e-05,
+      "loss": 0.2826,
+      "step": 1463
+    },
+    {
+      "epoch": 2.2805923616523773,
+      "grad_norm": 0.22757568106572626,
+      "learning_rate": 1.3265895953757226e-05,
+      "loss": 0.2923,
+      "step": 1464
+    },
+    {
+      "epoch": 2.2821512081060016,
+      "grad_norm": 0.22020055614811698,
+      "learning_rate": 1.323699421965318e-05,
+      "loss": 0.2807,
+      "step": 1465
+    },
+    {
+      "epoch": 2.283710054559626,
+      "grad_norm": 0.23965818423420435,
+      "learning_rate": 1.3208092485549134e-05,
+      "loss": 0.295,
+      "step": 1466
+    },
+    {
+      "epoch": 2.28526890101325,
+      "grad_norm": 0.21933495891348634,
+      "learning_rate": 1.3179190751445086e-05,
+      "loss": 0.2808,
+      "step": 1467
+    },
+    {
+      "epoch": 2.2868277474668743,
+      "grad_norm": 0.22394396080019968,
+      "learning_rate": 1.3150289017341041e-05,
+      "loss": 0.2828,
+      "step": 1468
+    },
+    {
+      "epoch": 2.2883865939204986,
+      "grad_norm": 0.2341244557636366,
+      "learning_rate": 1.3121387283236993e-05,
+      "loss": 0.2872,
+      "step": 1469
+    },
+    {
+      "epoch": 2.2899454403741233,
+      "grad_norm": 0.24597918258046544,
+      "learning_rate": 1.3092485549132948e-05,
+      "loss": 0.2854,
+      "step": 1470
+    },
+    {
+      "epoch": 2.2915042868277475,
+      "grad_norm": 0.23301804978413262,
+      "learning_rate": 1.3063583815028902e-05,
+      "loss": 0.3053,
+      "step": 1471
+    },
+    {
+      "epoch": 2.293063133281372,
+      "grad_norm": 0.2360765813557192,
+      "learning_rate": 1.3034682080924857e-05,
+      "loss": 0.2831,
+      "step": 1472
+    },
+    {
+      "epoch": 2.294621979734996,
+      "grad_norm": 0.2658423968018713,
+      "learning_rate": 1.3005780346820809e-05,
+      "loss": 0.2869,
+      "step": 1473
+    },
+    {
+      "epoch": 2.2961808261886203,
+      "grad_norm": 0.25016881594307117,
+      "learning_rate": 1.2976878612716764e-05,
+      "loss": 0.2763,
+      "step": 1474
+    },
+    {
+      "epoch": 2.2977396726422445,
+      "grad_norm": 0.2382698683460339,
+      "learning_rate": 1.2947976878612717e-05,
+      "loss": 0.3028,
+      "step": 1475
+    },
+    {
+      "epoch": 2.2992985190958692,
+      "grad_norm": 0.24221989883612682,
+      "learning_rate": 1.2919075144508672e-05,
+      "loss": 0.3088,
+      "step": 1476
+    },
+    {
+      "epoch": 2.3008573655494935,
+      "grad_norm": 0.25017938488171704,
+      "learning_rate": 1.2890173410404624e-05,
+      "loss": 0.2869,
+      "step": 1477
+    },
+    {
+      "epoch": 2.3024162120031177,
+      "grad_norm": 0.23057606820343765,
+      "learning_rate": 1.2861271676300579e-05,
+      "loss": 0.2812,
+      "step": 1478
+    },
+    {
+      "epoch": 2.303975058456742,
+      "grad_norm": 0.22851951981455979,
+      "learning_rate": 1.2832369942196531e-05,
+      "loss": 0.2903,
+      "step": 1479
+    },
+    {
+      "epoch": 2.3055339049103662,
+      "grad_norm": 0.25023954290485484,
+      "learning_rate": 1.2803468208092486e-05,
+      "loss": 0.2995,
+      "step": 1480
+    },
+    {
+      "epoch": 2.3070927513639905,
+      "grad_norm": 0.22267748488279954,
+      "learning_rate": 1.277456647398844e-05,
+      "loss": 0.2946,
+      "step": 1481
+    },
+    {
+      "epoch": 2.308651597817615,
+      "grad_norm": 0.2433472732667005,
+      "learning_rate": 1.2745664739884395e-05,
+      "loss": 0.2805,
+      "step": 1482
+    },
+    {
+      "epoch": 2.3102104442712394,
+      "grad_norm": 0.22084205335592666,
+      "learning_rate": 1.2716763005780346e-05,
+      "loss": 0.2718,
+      "step": 1483
+    },
+    {
+      "epoch": 2.3117692907248637,
+      "grad_norm": 0.24304148596183653,
+      "learning_rate": 1.2687861271676302e-05,
+      "loss": 0.2988,
+      "step": 1484
+    },
+    {
+      "epoch": 2.313328137178488,
+      "grad_norm": 0.2424544214432883,
+      "learning_rate": 1.2658959537572255e-05,
+      "loss": 0.2899,
+      "step": 1485
+    },
+    {
+      "epoch": 2.314886983632112,
+      "grad_norm": 0.2328910388765159,
+      "learning_rate": 1.263005780346821e-05,
+      "loss": 0.3127,
+      "step": 1486
+    },
+    {
+      "epoch": 2.3164458300857365,
+      "grad_norm": 0.23286533311552443,
+      "learning_rate": 1.2601156069364162e-05,
+      "loss": 0.2995,
+      "step": 1487
+    },
+    {
+      "epoch": 2.3180046765393607,
+      "grad_norm": 0.2549382777228154,
+      "learning_rate": 1.2572254335260117e-05,
+      "loss": 0.2928,
+      "step": 1488
+    },
+    {
+      "epoch": 2.3195635229929854,
+      "grad_norm": 0.2559163608878674,
+      "learning_rate": 1.2543352601156069e-05,
+      "loss": 0.3113,
+      "step": 1489
+    },
+    {
+      "epoch": 2.3211223694466097,
+      "grad_norm": 0.2265685017446726,
+      "learning_rate": 1.2514450867052024e-05,
+      "loss": 0.295,
+      "step": 1490
+    },
+    {
+      "epoch": 2.322681215900234,
+      "grad_norm": 0.22273564838191307,
+      "learning_rate": 1.2485549132947977e-05,
+      "loss": 0.2717,
+      "step": 1491
+    },
+    {
+      "epoch": 2.324240062353858,
+      "grad_norm": 0.21674653038960545,
+      "learning_rate": 1.245664739884393e-05,
+      "loss": 0.2845,
+      "step": 1492
+    },
+    {
+      "epoch": 2.3257989088074824,
+      "grad_norm": 0.21909928945574206,
+      "learning_rate": 1.2427745664739884e-05,
+      "loss": 0.284,
+      "step": 1493
+    },
+    {
+      "epoch": 2.3273577552611067,
+      "grad_norm": 0.2750007906217208,
+      "learning_rate": 1.2398843930635838e-05,
+      "loss": 0.2862,
+      "step": 1494
+    },
+    {
+      "epoch": 2.328916601714731,
+      "grad_norm": 0.21703802020708915,
+      "learning_rate": 1.2369942196531791e-05,
+      "loss": 0.2829,
+      "step": 1495
+    },
+    {
+      "epoch": 2.3304754481683556,
+      "grad_norm": 0.22950831232595262,
+      "learning_rate": 1.2341040462427746e-05,
+      "loss": 0.2883,
+      "step": 1496
+    },
+    {
+      "epoch": 2.33203429462198,
+      "grad_norm": 0.24426647745397967,
+      "learning_rate": 1.23121387283237e-05,
+      "loss": 0.2955,
+      "step": 1497
+    },
+    {
+      "epoch": 2.333593141075604,
+      "grad_norm": 0.23949518800459763,
+      "learning_rate": 1.2283236994219653e-05,
+      "loss": 0.2771,
+      "step": 1498
+    },
+    {
+      "epoch": 2.3351519875292284,
+      "grad_norm": 0.21894296552516587,
+      "learning_rate": 1.2254335260115607e-05,
+      "loss": 0.2899,
+      "step": 1499
+    },
+    {
+      "epoch": 2.3367108339828526,
+      "grad_norm": 0.22130870442542108,
+      "learning_rate": 1.222543352601156e-05,
+      "loss": 0.2836,
+      "step": 1500
+    },
+    {
+      "epoch": 2.338269680436477,
+      "grad_norm": 0.2659075059435682,
+      "learning_rate": 1.2196531791907515e-05,
+      "loss": 0.2838,
+      "step": 1501
+    },
+    {
+      "epoch": 2.339828526890101,
+      "grad_norm": 0.23607600423290415,
+      "learning_rate": 1.2167630057803469e-05,
+      "loss": 0.2972,
+      "step": 1502
+    },
+    {
+      "epoch": 2.341387373343726,
+      "grad_norm": 0.2352342145609656,
+      "learning_rate": 1.2138728323699422e-05,
+      "loss": 0.3003,
+      "step": 1503
+    },
+    {
+      "epoch": 2.34294621979735,
+      "grad_norm": 0.23180782759300908,
+      "learning_rate": 1.2109826589595376e-05,
+      "loss": 0.2925,
+      "step": 1504
+    },
+    {
+      "epoch": 2.3445050662509743,
+      "grad_norm": 0.2421492426629086,
+      "learning_rate": 1.2080924855491329e-05,
+      "loss": 0.3153,
+      "step": 1505
+    },
+    {
+      "epoch": 2.3460639127045986,
+      "grad_norm": 0.220047149003109,
+      "learning_rate": 1.2052023121387284e-05,
+      "loss": 0.2915,
+      "step": 1506
+    },
+    {
+      "epoch": 2.347622759158223,
+      "grad_norm": 0.2081371282315123,
+      "learning_rate": 1.2023121387283238e-05,
+      "loss": 0.2786,
+      "step": 1507
+    },
+    {
+      "epoch": 2.349181605611847,
+      "grad_norm": 0.2209132364079271,
+      "learning_rate": 1.1994219653179191e-05,
+      "loss": 0.2919,
+      "step": 1508
+    },
+    {
+      "epoch": 2.3507404520654713,
+      "grad_norm": 0.24203159776015792,
+      "learning_rate": 1.1965317919075144e-05,
+      "loss": 0.2944,
+      "step": 1509
+    },
+    {
+      "epoch": 2.352299298519096,
+      "grad_norm": 0.21125490785458723,
+      "learning_rate": 1.1936416184971098e-05,
+      "loss": 0.2819,
+      "step": 1510
+    },
+    {
+      "epoch": 2.3538581449727203,
+      "grad_norm": 0.21084827155820668,
+      "learning_rate": 1.1907514450867053e-05,
+      "loss": 0.2735,
+      "step": 1511
+    },
+    {
+      "epoch": 2.3554169914263445,
+      "grad_norm": 0.2341059203258637,
+      "learning_rate": 1.1878612716763006e-05,
+      "loss": 0.2991,
+      "step": 1512
+    },
+    {
+      "epoch": 2.356975837879969,
+      "grad_norm": 0.21296853713090008,
+      "learning_rate": 1.184971098265896e-05,
+      "loss": 0.2833,
+      "step": 1513
+    },
+    {
+      "epoch": 2.358534684333593,
+      "grad_norm": 0.2567119485230043,
+      "learning_rate": 1.1820809248554913e-05,
+      "loss": 0.2947,
+      "step": 1514
+    },
+    {
+      "epoch": 2.3600935307872173,
+      "grad_norm": 0.2333031479163138,
+      "learning_rate": 1.1791907514450867e-05,
+      "loss": 0.2864,
+      "step": 1515
+    },
+    {
+      "epoch": 2.3616523772408415,
+      "grad_norm": 0.24056595254646906,
+      "learning_rate": 1.176300578034682e-05,
+      "loss": 0.2802,
+      "step": 1516
+    },
+    {
+      "epoch": 2.3632112236944662,
+      "grad_norm": 0.23247035275872605,
+      "learning_rate": 1.1734104046242775e-05,
+      "loss": 0.2821,
+      "step": 1517
+    },
+    {
+      "epoch": 2.3647700701480905,
+      "grad_norm": 0.23551357842178408,
+      "learning_rate": 1.1705202312138729e-05,
+      "loss": 0.2799,
+      "step": 1518
+    },
+    {
+      "epoch": 2.3663289166017147,
+      "grad_norm": 0.22199252143905965,
+      "learning_rate": 1.1676300578034682e-05,
+      "loss": 0.2921,
+      "step": 1519
+    },
+    {
+      "epoch": 2.367887763055339,
+      "grad_norm": 0.2557073653528457,
+      "learning_rate": 1.1647398843930636e-05,
+      "loss": 0.2819,
+      "step": 1520
+    },
+    {
+      "epoch": 2.3694466095089632,
+      "grad_norm": 0.23177632971310883,
+      "learning_rate": 1.161849710982659e-05,
+      "loss": 0.3,
+      "step": 1521
+    },
+    {
+      "epoch": 2.3710054559625875,
+      "grad_norm": 0.2141949825685525,
+      "learning_rate": 1.1589595375722544e-05,
+      "loss": 0.2804,
+      "step": 1522
+    },
+    {
+      "epoch": 2.372564302416212,
+      "grad_norm": 0.22456594982131964,
+      "learning_rate": 1.1560693641618498e-05,
+      "loss": 0.2919,
+      "step": 1523
+    },
+    {
+      "epoch": 2.3741231488698364,
+      "grad_norm": 0.2328422263169891,
+      "learning_rate": 1.1531791907514451e-05,
+      "loss": 0.298,
+      "step": 1524
+    },
+    {
+      "epoch": 2.3756819953234607,
+      "grad_norm": 0.23696530509869898,
+      "learning_rate": 1.1502890173410405e-05,
+      "loss": 0.2826,
+      "step": 1525
+    },
+    {
+      "epoch": 2.377240841777085,
+      "grad_norm": 0.23471588381502256,
+      "learning_rate": 1.1473988439306358e-05,
+      "loss": 0.2954,
+      "step": 1526
+    },
+    {
+      "epoch": 2.378799688230709,
+      "grad_norm": 0.24324817997977416,
+      "learning_rate": 1.1445086705202313e-05,
+      "loss": 0.3002,
+      "step": 1527
+    },
+    {
+      "epoch": 2.3803585346843334,
+      "grad_norm": 0.23818184986570196,
+      "learning_rate": 1.1416184971098267e-05,
+      "loss": 0.2901,
+      "step": 1528
+    },
+    {
+      "epoch": 2.381917381137958,
+      "grad_norm": 0.2177316843513184,
+      "learning_rate": 1.138728323699422e-05,
+      "loss": 0.2837,
+      "step": 1529
+    },
+    {
+      "epoch": 2.3834762275915824,
+      "grad_norm": 0.2230616433485158,
+      "learning_rate": 1.1358381502890174e-05,
+      "loss": 0.2926,
+      "step": 1530
+    },
+    {
+      "epoch": 2.3850350740452066,
+      "grad_norm": 0.2092816344234883,
+      "learning_rate": 1.1329479768786127e-05,
+      "loss": 0.2771,
+      "step": 1531
+    },
+    {
+      "epoch": 2.386593920498831,
+      "grad_norm": 0.24426299651005548,
+      "learning_rate": 1.1300578034682082e-05,
+      "loss": 0.2893,
+      "step": 1532
+    },
+    {
+      "epoch": 2.388152766952455,
+      "grad_norm": 0.2257044731930893,
+      "learning_rate": 1.1271676300578036e-05,
+      "loss": 0.3099,
+      "step": 1533
+    },
+    {
+      "epoch": 2.3897116134060794,
+      "grad_norm": 0.22709943701947127,
+      "learning_rate": 1.1242774566473989e-05,
+      "loss": 0.2815,
+      "step": 1534
+    },
+    {
+      "epoch": 2.3912704598597037,
+      "grad_norm": 0.23081380246295977,
+      "learning_rate": 1.1213872832369943e-05,
+      "loss": 0.2892,
+      "step": 1535
+    },
+    {
+      "epoch": 2.3928293063133284,
+      "grad_norm": 0.2404518884149201,
+      "learning_rate": 1.1184971098265896e-05,
+      "loss": 0.2784,
+      "step": 1536
+    },
+    {
+      "epoch": 2.3943881527669526,
+      "grad_norm": 0.21874834698362006,
+      "learning_rate": 1.115606936416185e-05,
+      "loss": 0.2838,
+      "step": 1537
+    },
+    {
+      "epoch": 2.395946999220577,
+      "grad_norm": 0.2170285592301821,
+      "learning_rate": 1.1127167630057805e-05,
+      "loss": 0.2836,
+      "step": 1538
+    },
+    {
+      "epoch": 2.397505845674201,
+      "grad_norm": 0.22151366255017815,
+      "learning_rate": 1.1098265895953758e-05,
+      "loss": 0.2945,
+      "step": 1539
+    },
+    {
+      "epoch": 2.3990646921278254,
+      "grad_norm": 0.2272145173574314,
+      "learning_rate": 1.1069364161849711e-05,
+      "loss": 0.2836,
+      "step": 1540
+    },
+    {
+      "epoch": 2.4006235385814496,
+      "grad_norm": 0.22611588795882234,
+      "learning_rate": 1.1040462427745665e-05,
+      "loss": 0.3038,
+      "step": 1541
+    },
+    {
+      "epoch": 2.402182385035074,
+      "grad_norm": 0.21786269575794123,
+      "learning_rate": 1.1011560693641618e-05,
+      "loss": 0.2955,
+      "step": 1542
+    },
+    {
+      "epoch": 2.4037412314886986,
+      "grad_norm": 0.2081735505823338,
+      "learning_rate": 1.0982658959537573e-05,
+      "loss": 0.2845,
+      "step": 1543
+    },
+    {
+      "epoch": 2.405300077942323,
+      "grad_norm": 0.21791564332112892,
+      "learning_rate": 1.0953757225433527e-05,
+      "loss": 0.2888,
+      "step": 1544
+    },
+    {
+      "epoch": 2.406858924395947,
+      "grad_norm": 0.21343814806147593,
+      "learning_rate": 1.092485549132948e-05,
+      "loss": 0.3027,
+      "step": 1545
+    },
+    {
+      "epoch": 2.4084177708495713,
+      "grad_norm": 0.24305021560110593,
+      "learning_rate": 1.0895953757225434e-05,
+      "loss": 0.3096,
+      "step": 1546
+    },
+    {
+      "epoch": 2.4099766173031956,
+      "grad_norm": 0.2428823537320873,
+      "learning_rate": 1.0867052023121387e-05,
+      "loss": 0.2787,
+      "step": 1547
+    },
+    {
+      "epoch": 2.41153546375682,
+      "grad_norm": 0.22533176874540484,
+      "learning_rate": 1.0838150289017342e-05,
+      "loss": 0.3003,
+      "step": 1548
+    },
+    {
+      "epoch": 2.413094310210444,
+      "grad_norm": 0.23052788059445348,
+      "learning_rate": 1.0809248554913296e-05,
+      "loss": 0.2937,
+      "step": 1549
+    },
+    {
+      "epoch": 2.4146531566640688,
+      "grad_norm": 0.21873125390687612,
+      "learning_rate": 1.078034682080925e-05,
+      "loss": 0.2852,
+      "step": 1550
+    },
+    {
+      "epoch": 2.416212003117693,
+      "grad_norm": 0.23262040078023866,
+      "learning_rate": 1.0751445086705203e-05,
+      "loss": 0.2867,
+      "step": 1551
+    },
+    {
+      "epoch": 2.4177708495713173,
+      "grad_norm": 0.2061110656033279,
+      "learning_rate": 1.0722543352601156e-05,
+      "loss": 0.2834,
+      "step": 1552
+    },
+    {
+      "epoch": 2.4193296960249415,
+      "grad_norm": 0.21742353821980845,
+      "learning_rate": 1.0693641618497111e-05,
+      "loss": 0.286,
+      "step": 1553
+    },
+    {
+      "epoch": 2.4208885424785658,
+      "grad_norm": 0.21761958606143322,
+      "learning_rate": 1.0664739884393065e-05,
+      "loss": 0.2934,
+      "step": 1554
+    },
+    {
+      "epoch": 2.42244738893219,
+      "grad_norm": 0.22413107731125181,
+      "learning_rate": 1.0635838150289018e-05,
+      "loss": 0.3017,
+      "step": 1555
+    },
+    {
+      "epoch": 2.4240062353858143,
+      "grad_norm": 0.21834093708352184,
+      "learning_rate": 1.0606936416184972e-05,
+      "loss": 0.304,
+      "step": 1556
+    },
+    {
+      "epoch": 2.425565081839439,
+      "grad_norm": 0.22975540033681002,
+      "learning_rate": 1.0578034682080925e-05,
+      "loss": 0.2926,
+      "step": 1557
+    },
+    {
+      "epoch": 2.4271239282930632,
+      "grad_norm": 0.2274868161114814,
+      "learning_rate": 1.0549132947976879e-05,
+      "loss": 0.2999,
+      "step": 1558
+    },
+    {
+      "epoch": 2.4286827747466875,
+      "grad_norm": 0.22082603952846616,
+      "learning_rate": 1.0520231213872834e-05,
+      "loss": 0.2782,
+      "step": 1559
+    },
+    {
+      "epoch": 2.4302416212003117,
+      "grad_norm": 0.23461317288098227,
+      "learning_rate": 1.0491329479768787e-05,
+      "loss": 0.2842,
+      "step": 1560
+    },
+    {
+      "epoch": 2.431800467653936,
+      "grad_norm": 0.20440782512445527,
+      "learning_rate": 1.046242774566474e-05,
+      "loss": 0.2853,
+      "step": 1561
+    },
+    {
+      "epoch": 2.4333593141075602,
+      "grad_norm": 0.23611049627402994,
+      "learning_rate": 1.0433526011560694e-05,
+      "loss": 0.2841,
+      "step": 1562
+    },
+    {
+      "epoch": 2.4349181605611845,
+      "grad_norm": 0.23437404672528528,
+      "learning_rate": 1.0404624277456647e-05,
+      "loss": 0.2963,
+      "step": 1563
+    },
+    {
+      "epoch": 2.436477007014809,
+      "grad_norm": 0.2204494699300127,
+      "learning_rate": 1.0375722543352603e-05,
+      "loss": 0.2806,
+      "step": 1564
+    },
+    {
+      "epoch": 2.4380358534684334,
+      "grad_norm": 0.20490594460203101,
+      "learning_rate": 1.0346820809248556e-05,
+      "loss": 0.2761,
+      "step": 1565
+    },
+    {
+      "epoch": 2.4395946999220577,
+      "grad_norm": 0.6956010247527492,
+      "learning_rate": 1.031791907514451e-05,
+      "loss": 0.2901,
+      "step": 1566
+    },
+    {
+      "epoch": 2.441153546375682,
+      "grad_norm": 0.26224821751440014,
+      "learning_rate": 1.0289017341040463e-05,
+      "loss": 0.3034,
+      "step": 1567
+    },
+    {
+      "epoch": 2.442712392829306,
+      "grad_norm": 0.25126259259462486,
+      "learning_rate": 1.0260115606936416e-05,
+      "loss": 0.2899,
+      "step": 1568
+    },
+    {
+      "epoch": 2.4442712392829304,
+      "grad_norm": 0.22870053657745457,
+      "learning_rate": 1.0231213872832372e-05,
+      "loss": 0.2822,
+      "step": 1569
+    },
+    {
+      "epoch": 2.445830085736555,
+      "grad_norm": 0.23587258742639972,
+      "learning_rate": 1.0202312138728325e-05,
+      "loss": 0.2593,
+      "step": 1570
+    },
+    {
+      "epoch": 2.4473889321901794,
+      "grad_norm": 0.22897162723371334,
+      "learning_rate": 1.0173410404624278e-05,
+      "loss": 0.2912,
+      "step": 1571
+    },
+    {
+      "epoch": 2.4489477786438036,
+      "grad_norm": 0.24149402834986947,
+      "learning_rate": 1.0144508670520232e-05,
+      "loss": 0.2878,
+      "step": 1572
+    },
+    {
+      "epoch": 2.450506625097428,
+      "grad_norm": 0.23350074250406747,
+      "learning_rate": 1.0115606936416185e-05,
+      "loss": 0.2869,
+      "step": 1573
+    },
+    {
+      "epoch": 2.452065471551052,
+      "grad_norm": 0.22195416376987834,
+      "learning_rate": 1.008670520231214e-05,
+      "loss": 0.2956,
+      "step": 1574
+    },
+    {
+      "epoch": 2.4536243180046764,
+      "grad_norm": 1.3487509546561793,
+      "learning_rate": 1.0057803468208094e-05,
+      "loss": 0.3012,
+      "step": 1575
+    },
+    {
+      "epoch": 2.455183164458301,
+      "grad_norm": 0.2873786952181812,
+      "learning_rate": 1.0028901734104047e-05,
+      "loss": 0.284,
+      "step": 1576
+    },
+    {
+      "epoch": 2.4567420109119253,
+      "grad_norm": 0.281096191572859,
+      "learning_rate": 1e-05,
+      "loss": 0.2839,
+      "step": 1577
+    },
+    {
+      "epoch": 2.4583008573655496,
+      "grad_norm": 0.26582966306416833,
+      "learning_rate": 9.971098265895954e-06,
+      "loss": 0.2868,
+      "step": 1578
+    },
+    {
+      "epoch": 2.459859703819174,
+      "grad_norm": 0.23786440274496637,
+      "learning_rate": 9.94219653179191e-06,
+      "loss": 0.2874,
+      "step": 1579
+    },
+    {
+      "epoch": 2.461418550272798,
+      "grad_norm": 0.26380977501613667,
+      "learning_rate": 9.913294797687863e-06,
+      "loss": 0.298,
+      "step": 1580
+    },
+    {
+      "epoch": 2.4629773967264224,
+      "grad_norm": 0.28090632623520245,
+      "learning_rate": 9.884393063583816e-06,
+      "loss": 0.2824,
+      "step": 1581
+    },
+    {
+      "epoch": 2.4645362431800466,
+      "grad_norm": 0.24654785691378972,
+      "learning_rate": 9.85549132947977e-06,
+      "loss": 0.2842,
+      "step": 1582
+    },
+    {
+      "epoch": 2.4660950896336713,
+      "grad_norm": 0.22185993368845872,
+      "learning_rate": 9.826589595375723e-06,
+      "loss": 0.2784,
+      "step": 1583
+    },
+    {
+      "epoch": 2.4676539360872956,
+      "grad_norm": 0.25759773055873747,
+      "learning_rate": 9.797687861271677e-06,
+      "loss": 0.3065,
+      "step": 1584
+    },
+    {
+      "epoch": 2.46921278254092,
+      "grad_norm": 0.2448676371644166,
+      "learning_rate": 9.768786127167632e-06,
+      "loss": 0.2951,
+      "step": 1585
+    },
+    {
+      "epoch": 2.470771628994544,
+      "grad_norm": 0.22038630080496294,
+      "learning_rate": 9.739884393063585e-06,
+      "loss": 0.2766,
+      "step": 1586
+    },
+    {
+      "epoch": 2.4723304754481683,
+      "grad_norm": 0.21067846046967328,
+      "learning_rate": 9.710982658959537e-06,
+      "loss": 0.2772,
+      "step": 1587
+    },
+    {
+      "epoch": 2.4738893219017926,
+      "grad_norm": 0.2196571148859021,
+      "learning_rate": 9.68208092485549e-06,
+      "loss": 0.2944,
+      "step": 1588
+    },
+    {
+      "epoch": 2.475448168355417,
+      "grad_norm": 0.20661464746790095,
+      "learning_rate": 9.653179190751446e-06,
+      "loss": 0.3023,
+      "step": 1589
+    },
+    {
+      "epoch": 2.4770070148090415,
+      "grad_norm": 0.20946455345772869,
+      "learning_rate": 9.624277456647399e-06,
+      "loss": 0.2768,
+      "step": 1590
+    },
+    {
+      "epoch": 2.4785658612626658,
+      "grad_norm": 0.21800665843244352,
+      "learning_rate": 9.595375722543352e-06,
+      "loss": 0.3014,
+      "step": 1591
+    },
+    {
+      "epoch": 2.48012470771629,
+      "grad_norm": 0.21469256658865496,
+      "learning_rate": 9.566473988439306e-06,
+      "loss": 0.2865,
+      "step": 1592
+    },
+    {
+      "epoch": 2.4816835541699143,
+      "grad_norm": 0.20483902813627355,
+      "learning_rate": 9.53757225433526e-06,
+      "loss": 0.2734,
+      "step": 1593
+    },
+    {
+      "epoch": 2.4832424006235385,
+      "grad_norm": 0.22797586254976032,
+      "learning_rate": 9.508670520231214e-06,
+      "loss": 0.2839,
+      "step": 1594
+    },
+    {
+      "epoch": 2.4848012470771628,
+      "grad_norm": 0.20269734125823377,
+      "learning_rate": 9.479768786127168e-06,
+      "loss": 0.2853,
+      "step": 1595
+    },
+    {
+      "epoch": 2.486360093530787,
+      "grad_norm": 0.21209747614140953,
+      "learning_rate": 9.450867052023121e-06,
+      "loss": 0.2846,
+      "step": 1596
+    },
+    {
+      "epoch": 2.4879189399844117,
+      "grad_norm": 0.20886949964376036,
+      "learning_rate": 9.421965317919075e-06,
+      "loss": 0.2768,
+      "step": 1597
+    },
+    {
+      "epoch": 2.489477786438036,
+      "grad_norm": 0.21987185193696723,
+      "learning_rate": 9.393063583815028e-06,
+      "loss": 0.289,
+      "step": 1598
+    },
+    {
+      "epoch": 2.49103663289166,
+      "grad_norm": 0.22992122520081743,
+      "learning_rate": 9.364161849710983e-06,
+      "loss": 0.317,
+      "step": 1599
+    },
+    {
+      "epoch": 2.4925954793452845,
+      "grad_norm": 0.2216415692962526,
+      "learning_rate": 9.335260115606937e-06,
+      "loss": 0.2872,
+      "step": 1600
+    },
+    {
+      "epoch": 2.4941543257989087,
+      "grad_norm": 0.22333309804066181,
+      "learning_rate": 9.30635838150289e-06,
+      "loss": 0.2762,
+      "step": 1601
+    },
+    {
+      "epoch": 2.495713172252533,
+      "grad_norm": 0.2088510278237382,
+      "learning_rate": 9.277456647398844e-06,
+      "loss": 0.2849,
+      "step": 1602
+    },
+    {
+      "epoch": 2.4972720187061572,
+      "grad_norm": 0.21834916127206458,
+      "learning_rate": 9.248554913294797e-06,
+      "loss": 0.2945,
+      "step": 1603
+    },
+    {
+      "epoch": 2.498830865159782,
+      "grad_norm": 0.23784512272220631,
+      "learning_rate": 9.21965317919075e-06,
+      "loss": 0.29,
+      "step": 1604
+    },
+    {
+      "epoch": 2.500389711613406,
+      "grad_norm": 0.22622134161606003,
+      "learning_rate": 9.190751445086706e-06,
+      "loss": 0.2855,
+      "step": 1605
+    },
+    {
+      "epoch": 2.5019485580670304,
+      "grad_norm": 0.20987648910752202,
+      "learning_rate": 9.16184971098266e-06,
+      "loss": 0.2886,
+      "step": 1606
+    },
+    {
+      "epoch": 2.5035074045206547,
+      "grad_norm": 0.2271896769680405,
+      "learning_rate": 9.132947976878613e-06,
+      "loss": 0.2849,
+      "step": 1607
+    },
+    {
+      "epoch": 2.505066250974279,
+      "grad_norm": 0.20703729358511513,
+      "learning_rate": 9.104046242774566e-06,
+      "loss": 0.285,
+      "step": 1608
+    },
+    {
+      "epoch": 2.5066250974279036,
+      "grad_norm": 0.2210043646064696,
+      "learning_rate": 9.07514450867052e-06,
+      "loss": 0.2793,
+      "step": 1609
+    },
+    {
+      "epoch": 2.5081839438815274,
+      "grad_norm": 0.20832438994217642,
+      "learning_rate": 9.046242774566475e-06,
+      "loss": 0.2706,
+      "step": 1610
+    },
+    {
+      "epoch": 2.509742790335152,
+      "grad_norm": 0.22663700329890946,
+      "learning_rate": 9.017341040462428e-06,
+      "loss": 0.2926,
+      "step": 1611
+    },
+    {
+      "epoch": 2.5113016367887764,
+      "grad_norm": 0.2321310919459722,
+      "learning_rate": 8.988439306358382e-06,
+      "loss": 0.2929,
+      "step": 1612
+    },
+    {
+      "epoch": 2.5128604832424006,
+      "grad_norm": 0.20367730242743393,
+      "learning_rate": 8.959537572254335e-06,
+      "loss": 0.273,
+      "step": 1613
+    },
+    {
+      "epoch": 2.514419329696025,
+      "grad_norm": 0.2573291858309685,
+      "learning_rate": 8.930635838150288e-06,
+      "loss": 0.2937,
+      "step": 1614
+    },
+    {
+      "epoch": 2.515978176149649,
+      "grad_norm": 0.21765613776271725,
+      "learning_rate": 8.901734104046244e-06,
+      "loss": 0.3041,
+      "step": 1615
+    },
+    {
+      "epoch": 2.517537022603274,
+      "grad_norm": 0.22719439248378498,
+      "learning_rate": 8.872832369942197e-06,
+      "loss": 0.3025,
+      "step": 1616
+    },
+    {
+      "epoch": 2.5190958690568976,
+      "grad_norm": 0.2233584630382556,
+      "learning_rate": 8.84393063583815e-06,
+      "loss": 0.2901,
+      "step": 1617
+    },
+    {
+      "epoch": 2.5206547155105223,
+      "grad_norm": 0.2756546250082013,
+      "learning_rate": 8.815028901734104e-06,
+      "loss": 0.299,
+      "step": 1618
+    },
+    {
+      "epoch": 2.5222135619641466,
+      "grad_norm": 0.2471624019129573,
+      "learning_rate": 8.786127167630057e-06,
+      "loss": 0.2856,
+      "step": 1619
+    },
+    {
+      "epoch": 2.523772408417771,
+      "grad_norm": 0.22928484558281476,
+      "learning_rate": 8.757225433526013e-06,
+      "loss": 0.2885,
+      "step": 1620
+    },
+    {
+      "epoch": 2.525331254871395,
+      "grad_norm": 0.22619496618870039,
+      "learning_rate": 8.728323699421966e-06,
+      "loss": 0.2858,
+      "step": 1621
+    },
+    {
+      "epoch": 2.5268901013250193,
+      "grad_norm": 0.23523904557731568,
+      "learning_rate": 8.69942196531792e-06,
+      "loss": 0.2879,
+      "step": 1622
+    },
+    {
+      "epoch": 2.528448947778644,
+      "grad_norm": 0.23102152485027722,
+      "learning_rate": 8.670520231213873e-06,
+      "loss": 0.2978,
+      "step": 1623
+    },
+    {
+      "epoch": 2.530007794232268,
+      "grad_norm": 0.23596999054388898,
+      "learning_rate": 8.641618497109826e-06,
+      "loss": 0.2955,
+      "step": 1624
+    },
+    {
+      "epoch": 2.5315666406858925,
+      "grad_norm": 0.23775289137632036,
+      "learning_rate": 8.61271676300578e-06,
+      "loss": 0.303,
+      "step": 1625
+    },
+    {
+      "epoch": 2.533125487139517,
+      "grad_norm": 0.22544643461170122,
+      "learning_rate": 8.583815028901735e-06,
+      "loss": 0.2907,
+      "step": 1626
+    },
+    {
+      "epoch": 2.534684333593141,
+      "grad_norm": 0.2598040215067987,
+      "learning_rate": 8.554913294797688e-06,
+      "loss": 0.2946,
+      "step": 1627
+    },
+    {
+      "epoch": 2.5362431800467653,
+      "grad_norm": 0.21413673157343954,
+      "learning_rate": 8.526011560693642e-06,
+      "loss": 0.2836,
+      "step": 1628
+    },
+    {
+      "epoch": 2.5378020265003896,
+      "grad_norm": 0.4266061422700911,
+      "learning_rate": 8.497109826589595e-06,
+      "loss": 0.2864,
+      "step": 1629
+    },
+    {
+      "epoch": 2.5393608729540142,
+      "grad_norm": 0.21799108371646567,
+      "learning_rate": 8.468208092485549e-06,
+      "loss": 0.2813,
+      "step": 1630
+    },
+    {
+      "epoch": 2.5409197194076385,
+      "grad_norm": 0.2322583381309059,
+      "learning_rate": 8.439306358381504e-06,
+      "loss": 0.3071,
+      "step": 1631
+    },
+    {
+      "epoch": 2.5424785658612628,
+      "grad_norm": 0.2202382316089639,
+      "learning_rate": 8.410404624277457e-06,
+      "loss": 0.2926,
+      "step": 1632
+    },
+    {
+      "epoch": 2.544037412314887,
+      "grad_norm": 0.22009637471013366,
+      "learning_rate": 8.38150289017341e-06,
+      "loss": 0.2963,
+      "step": 1633
+    },
+    {
+      "epoch": 2.5455962587685113,
+      "grad_norm": 0.20540911119232894,
+      "learning_rate": 8.352601156069364e-06,
+      "loss": 0.275,
+      "step": 1634
+    },
+    {
+      "epoch": 2.5471551052221355,
+      "grad_norm": 0.21777131689422363,
+      "learning_rate": 8.323699421965318e-06,
+      "loss": 0.2996,
+      "step": 1635
+    },
+    {
+      "epoch": 2.5487139516757598,
+      "grad_norm": 0.20908427654761658,
+      "learning_rate": 8.294797687861273e-06,
+      "loss": 0.2742,
+      "step": 1636
+    },
+    {
+      "epoch": 2.5502727981293845,
+      "grad_norm": 0.21275798202193402,
+      "learning_rate": 8.265895953757226e-06,
+      "loss": 0.2921,
+      "step": 1637
+    },
+    {
+      "epoch": 2.5518316445830087,
+      "grad_norm": 0.22079763984850975,
+      "learning_rate": 8.23699421965318e-06,
+      "loss": 0.3023,
+      "step": 1638
+    },
+    {
+      "epoch": 2.553390491036633,
+      "grad_norm": 0.20167698447825708,
+      "learning_rate": 8.208092485549133e-06,
+      "loss": 0.3018,
+      "step": 1639
+    },
+    {
+      "epoch": 2.554949337490257,
+      "grad_norm": 0.204370882797588,
+      "learning_rate": 8.179190751445087e-06,
+      "loss": 0.2934,
+      "step": 1640
+    },
+    {
+      "epoch": 2.5565081839438815,
+      "grad_norm": 0.20933271636814302,
+      "learning_rate": 8.150289017341042e-06,
+      "loss": 0.2884,
+      "step": 1641
+    },
+    {
+      "epoch": 2.5580670303975057,
+      "grad_norm": 0.2364244855094284,
+      "learning_rate": 8.121387283236995e-06,
+      "loss": 0.291,
+      "step": 1642
+    },
+    {
+      "epoch": 2.55962587685113,
+      "grad_norm": 0.21280205055543394,
+      "learning_rate": 8.092485549132949e-06,
+      "loss": 0.2953,
+      "step": 1643
+    },
+    {
+      "epoch": 2.5611847233047547,
+      "grad_norm": 0.2225834531719899,
+      "learning_rate": 8.063583815028902e-06,
+      "loss": 0.2931,
+      "step": 1644
+    },
+    {
+      "epoch": 2.562743569758379,
+      "grad_norm": 0.20003247002848737,
+      "learning_rate": 8.034682080924856e-06,
+      "loss": 0.2769,
+      "step": 1645
+    },
+    {
+      "epoch": 2.564302416212003,
+      "grad_norm": 0.21372713472537674,
+      "learning_rate": 8.00578034682081e-06,
+      "loss": 0.2832,
+      "step": 1646
+    },
+    {
+      "epoch": 2.5658612626656274,
+      "grad_norm": 0.21988186302060084,
+      "learning_rate": 7.976878612716764e-06,
+      "loss": 0.2816,
+      "step": 1647
+    },
+    {
+      "epoch": 2.5674201091192517,
+      "grad_norm": 0.2114883685070932,
+      "learning_rate": 7.947976878612718e-06,
+      "loss": 0.2716,
+      "step": 1648
+    },
+    {
+      "epoch": 2.568978955572876,
+      "grad_norm": 0.22479614996029462,
+      "learning_rate": 7.919075144508671e-06,
+      "loss": 0.2895,
+      "step": 1649
+    },
+    {
+      "epoch": 2.5705378020265,
+      "grad_norm": 0.20995160083973538,
+      "learning_rate": 7.890173410404624e-06,
+      "loss": 0.2938,
+      "step": 1650
+    },
+    {
+      "epoch": 2.572096648480125,
+      "grad_norm": 0.19074058893303877,
+      "learning_rate": 7.861271676300578e-06,
+      "loss": 0.2879,
+      "step": 1651
+    },
+    {
+      "epoch": 2.573655494933749,
+      "grad_norm": 0.21433796239796818,
+      "learning_rate": 7.832369942196533e-06,
+      "loss": 0.2801,
+      "step": 1652
+    },
+    {
+      "epoch": 2.5752143413873734,
+      "grad_norm": 0.23148461464412992,
+      "learning_rate": 7.803468208092486e-06,
+      "loss": 0.3049,
+      "step": 1653
+    },
+    {
+      "epoch": 2.5767731878409976,
+      "grad_norm": 0.22404297663771336,
+      "learning_rate": 7.77456647398844e-06,
+      "loss": 0.2844,
+      "step": 1654
+    },
+    {
+      "epoch": 2.578332034294622,
+      "grad_norm": 0.22573478963343901,
+      "learning_rate": 7.745664739884393e-06,
+      "loss": 0.2907,
+      "step": 1655
+    },
+    {
+      "epoch": 2.5798908807482466,
+      "grad_norm": 0.19465572962012764,
+      "learning_rate": 7.716763005780347e-06,
+      "loss": 0.2786,
+      "step": 1656
+    },
+    {
+      "epoch": 2.5814497272018704,
+      "grad_norm": 0.21095454192122828,
+      "learning_rate": 7.687861271676302e-06,
+      "loss": 0.2845,
+      "step": 1657
+    },
+    {
+      "epoch": 2.583008573655495,
+      "grad_norm": 0.20373213583538066,
+      "learning_rate": 7.658959537572255e-06,
+      "loss": 0.2809,
+      "step": 1658
+    },
+    {
+      "epoch": 2.5845674201091193,
+      "grad_norm": 0.20909977447044456,
+      "learning_rate": 7.630057803468209e-06,
+      "loss": 0.2918,
+      "step": 1659
+    },
+    {
+      "epoch": 2.5861262665627436,
+      "grad_norm": 0.20642027240366984,
+      "learning_rate": 7.601156069364162e-06,
+      "loss": 0.2951,
+      "step": 1660
+    },
+    {
+      "epoch": 2.587685113016368,
+      "grad_norm": 0.20983785591383902,
+      "learning_rate": 7.5722543352601166e-06,
+      "loss": 0.2958,
+      "step": 1661
+    },
+    {
+      "epoch": 2.589243959469992,
+      "grad_norm": 0.23042841089901345,
+      "learning_rate": 7.54335260115607e-06,
+      "loss": 0.3101,
+      "step": 1662
+    },
+    {
+      "epoch": 2.590802805923617,
+      "grad_norm": 0.24194824337916648,
+      "learning_rate": 7.514450867052024e-06,
+      "loss": 0.3112,
+      "step": 1663
+    },
+    {
+      "epoch": 2.5923616523772406,
+      "grad_norm": 0.21401069536687056,
+      "learning_rate": 7.485549132947978e-06,
+      "loss": 0.2846,
+      "step": 1664
+    },
+    {
+      "epoch": 2.5939204988308653,
+      "grad_norm": 0.20182835176068856,
+      "learning_rate": 7.456647398843931e-06,
+      "loss": 0.2884,
+      "step": 1665
+    },
+    {
+      "epoch": 2.5954793452844895,
+      "grad_norm": 0.2438186993870951,
+      "learning_rate": 7.4277456647398855e-06,
+      "loss": 0.2914,
+      "step": 1666
+    },
+    {
+      "epoch": 2.597038191738114,
+      "grad_norm": 0.24739122146108788,
+      "learning_rate": 7.398843930635839e-06,
+      "loss": 0.2981,
+      "step": 1667
+    },
+    {
+      "epoch": 2.598597038191738,
+      "grad_norm": 0.25582873700332,
+      "learning_rate": 7.369942196531792e-06,
+      "loss": 0.2857,
+      "step": 1668
+    },
+    {
+      "epoch": 2.6001558846453623,
+      "grad_norm": 0.23788093624422235,
+      "learning_rate": 7.341040462427747e-06,
+      "loss": 0.2992,
+      "step": 1669
+    },
+    {
+      "epoch": 2.601714731098987,
+      "grad_norm": 0.2200261335858021,
+      "learning_rate": 7.3121387283237e-06,
+      "loss": 0.2869,
+      "step": 1670
+    },
+    {
+      "epoch": 2.603273577552611,
+      "grad_norm": 0.22579760963949985,
+      "learning_rate": 7.2832369942196544e-06,
+      "loss": 0.2982,
+      "step": 1671
+    },
+    {
+      "epoch": 2.6048324240062355,
+      "grad_norm": 0.23297289522530723,
+      "learning_rate": 7.254335260115608e-06,
+      "loss": 0.2825,
+      "step": 1672
+    },
+    {
+      "epoch": 2.6063912704598597,
+      "grad_norm": 0.20686966606749702,
+      "learning_rate": 7.225433526011561e-06,
+      "loss": 0.2852,
+      "step": 1673
+    },
+    {
+      "epoch": 2.607950116913484,
+      "grad_norm": 0.2266441955874942,
+      "learning_rate": 7.196531791907516e-06,
+      "loss": 0.3128,
+      "step": 1674
+    },
+    {
+      "epoch": 2.6095089633671082,
+      "grad_norm": 0.23213362522888187,
+      "learning_rate": 7.167630057803469e-06,
+      "loss": 0.2865,
+      "step": 1675
+    },
+    {
+      "epoch": 2.6110678098207325,
+      "grad_norm": 0.22468659409132077,
+      "learning_rate": 7.1387283236994225e-06,
+      "loss": 0.2654,
+      "step": 1676
+    },
+    {
+      "epoch": 2.612626656274357,
+      "grad_norm": 0.22695879901341415,
+      "learning_rate": 7.109826589595377e-06,
+      "loss": 0.2846,
+      "step": 1677
+    },
+    {
+      "epoch": 2.6141855027279814,
+      "grad_norm": 0.2055809379257179,
+      "learning_rate": 7.08092485549133e-06,
+      "loss": 0.2799,
+      "step": 1678
+    },
+    {
+      "epoch": 2.6157443491816057,
+      "grad_norm": 0.21550005636993283,
+      "learning_rate": 7.0520231213872845e-06,
+      "loss": 0.276,
+      "step": 1679
+    },
+    {
+      "epoch": 2.61730319563523,
+      "grad_norm": 0.2057285652964607,
+      "learning_rate": 7.023121387283238e-06,
+      "loss": 0.2871,
+      "step": 1680
+    },
+    {
+      "epoch": 2.618862042088854,
+      "grad_norm": 0.2329408962798249,
+      "learning_rate": 6.9942196531791914e-06,
+      "loss": 0.2969,
+      "step": 1681
+    },
+    {
+      "epoch": 2.6204208885424785,
+      "grad_norm": 0.21413261205015083,
+      "learning_rate": 6.965317919075146e-06,
+      "loss": 0.2887,
+      "step": 1682
+    },
+    {
+      "epoch": 2.6219797349961027,
+      "grad_norm": 0.20032603391697887,
+      "learning_rate": 6.936416184971098e-06,
+      "loss": 0.268,
+      "step": 1683
+    },
+    {
+      "epoch": 2.6235385814497274,
+      "grad_norm": 0.22915281553263556,
+      "learning_rate": 6.907514450867052e-06,
+      "loss": 0.2961,
+      "step": 1684
+    },
+    {
+      "epoch": 2.6250974279033517,
+      "grad_norm": 0.22923520682538653,
+      "learning_rate": 6.878612716763005e-06,
+      "loss": 0.2789,
+      "step": 1685
+    },
+    {
+      "epoch": 2.626656274356976,
+      "grad_norm": 0.20610407028884006,
+      "learning_rate": 6.8497109826589595e-06,
+      "loss": 0.2793,
+      "step": 1686
+    },
+    {
+      "epoch": 2.6282151208106,
+      "grad_norm": 0.21052444570238973,
+      "learning_rate": 6.820809248554913e-06,
+      "loss": 0.2864,
+      "step": 1687
+    },
+    {
+      "epoch": 2.6297739672642244,
+      "grad_norm": 0.2116958939607887,
+      "learning_rate": 6.791907514450866e-06,
+      "loss": 0.2893,
+      "step": 1688
+    },
+    {
+      "epoch": 2.6313328137178487,
+      "grad_norm": 0.2277230086812389,
+      "learning_rate": 6.763005780346821e-06,
+      "loss": 0.2912,
+      "step": 1689
+    },
+    {
+      "epoch": 2.632891660171473,
+      "grad_norm": 0.22320147334511584,
+      "learning_rate": 6.734104046242774e-06,
+      "loss": 0.3009,
+      "step": 1690
+    },
+    {
+      "epoch": 2.6344505066250976,
+      "grad_norm": 0.22390700490954193,
+      "learning_rate": 6.7052023121387284e-06,
+      "loss": 0.2818,
+      "step": 1691
+    },
+    {
+      "epoch": 2.636009353078722,
+      "grad_norm": 0.22080083950392762,
+      "learning_rate": 6.676300578034682e-06,
+      "loss": 0.3002,
+      "step": 1692
+    },
+    {
+      "epoch": 2.637568199532346,
+      "grad_norm": 0.2045566499206754,
+      "learning_rate": 6.647398843930635e-06,
+      "loss": 0.2845,
+      "step": 1693
+    },
+    {
+      "epoch": 2.6391270459859704,
+      "grad_norm": 0.2058593680536001,
+      "learning_rate": 6.61849710982659e-06,
+      "loss": 0.2963,
+      "step": 1694
+    },
+    {
+      "epoch": 2.6406858924395946,
+      "grad_norm": 0.21761267846258828,
+      "learning_rate": 6.589595375722543e-06,
+      "loss": 0.2958,
+      "step": 1695
+    },
+    {
+      "epoch": 2.642244738893219,
+      "grad_norm": 0.21267130911051824,
+      "learning_rate": 6.5606936416184965e-06,
+      "loss": 0.282,
+      "step": 1696
+    },
+    {
+      "epoch": 2.643803585346843,
+      "grad_norm": 0.2725520542350063,
+      "learning_rate": 6.531791907514451e-06,
+      "loss": 0.3082,
+      "step": 1697
+    },
+    {
+      "epoch": 2.645362431800468,
+      "grad_norm": 0.22632776119108647,
+      "learning_rate": 6.502890173410404e-06,
+      "loss": 0.2888,
+      "step": 1698
+    },
+    {
+      "epoch": 2.646921278254092,
+      "grad_norm": 0.21287462722249717,
+      "learning_rate": 6.4739884393063585e-06,
+      "loss": 0.2873,
+      "step": 1699
+    },
+    {
+      "epoch": 2.6484801247077163,
+      "grad_norm": 0.20006610540308692,
+      "learning_rate": 6.445086705202312e-06,
+      "loss": 0.2891,
+      "step": 1700
+    },
+    {
+      "epoch": 2.6500389711613406,
+      "grad_norm": 0.23392897268756666,
+      "learning_rate": 6.4161849710982654e-06,
+      "loss": 0.299,
+      "step": 1701
+    },
+    {
+      "epoch": 2.651597817614965,
+      "grad_norm": 0.2094557655435245,
+      "learning_rate": 6.38728323699422e-06,
+      "loss": 0.2893,
+      "step": 1702
+    },
+    {
+      "epoch": 2.6531566640685895,
+      "grad_norm": 0.2162667697074946,
+      "learning_rate": 6.358381502890173e-06,
+      "loss": 0.3018,
+      "step": 1703
+    },
+    {
+      "epoch": 2.6547155105222133,
+      "grad_norm": 0.20292202156574402,
+      "learning_rate": 6.3294797687861275e-06,
+      "loss": 0.289,
+      "step": 1704
+    },
+    {
+      "epoch": 2.656274356975838,
+      "grad_norm": 0.19930348804964854,
+      "learning_rate": 6.300578034682081e-06,
+      "loss": 0.2911,
+      "step": 1705
+    },
+    {
+      "epoch": 2.6578332034294623,
+      "grad_norm": 0.21358543639853841,
+      "learning_rate": 6.271676300578034e-06,
+      "loss": 0.2961,
+      "step": 1706
+    },
+    {
+      "epoch": 2.6593920498830865,
+      "grad_norm": 0.20342382560761232,
+      "learning_rate": 6.242774566473989e-06,
+      "loss": 0.2964,
+      "step": 1707
+    },
+    {
+      "epoch": 2.660950896336711,
+      "grad_norm": 0.21527984552285181,
+      "learning_rate": 6.213872832369942e-06,
+      "loss": 0.2757,
+      "step": 1708
+    },
+    {
+      "epoch": 2.662509742790335,
+      "grad_norm": 0.21467701298854663,
+      "learning_rate": 6.1849710982658956e-06,
+      "loss": 0.2901,
+      "step": 1709
+    },
+    {
+      "epoch": 2.6640685892439597,
+      "grad_norm": 0.19947891783905247,
+      "learning_rate": 6.15606936416185e-06,
+      "loss": 0.2821,
+      "step": 1710
+    },
+    {
+      "epoch": 2.6656274356975835,
+      "grad_norm": 0.2082997257189345,
+      "learning_rate": 6.127167630057803e-06,
+      "loss": 0.2845,
+      "step": 1711
+    },
+    {
+      "epoch": 2.6671862821512082,
+      "grad_norm": 0.21230363198815436,
+      "learning_rate": 6.098265895953758e-06,
+      "loss": 0.2933,
+      "step": 1712
+    },
+    {
+      "epoch": 2.6687451286048325,
+      "grad_norm": 0.19933909904198954,
+      "learning_rate": 6.069364161849711e-06,
+      "loss": 0.2837,
+      "step": 1713
+    },
+    {
+      "epoch": 2.6703039750584567,
+      "grad_norm": 0.22579536637132155,
+      "learning_rate": 6.0404624277456645e-06,
+      "loss": 0.2944,
+      "step": 1714
+    },
+    {
+      "epoch": 2.671862821512081,
+      "grad_norm": 0.2083571147340846,
+      "learning_rate": 6.011560693641619e-06,
+      "loss": 0.2858,
+      "step": 1715
+    },
+    {
+      "epoch": 2.6734216679657052,
+      "grad_norm": 0.21447279328154586,
+      "learning_rate": 5.982658959537572e-06,
+      "loss": 0.2796,
+      "step": 1716
+    },
+    {
+      "epoch": 2.67498051441933,
+      "grad_norm": 0.2577104816893999,
+      "learning_rate": 5.9537572254335265e-06,
+      "loss": 0.2777,
+      "step": 1717
+    },
+    {
+      "epoch": 2.6765393608729537,
+      "grad_norm": 0.2134999870561008,
+      "learning_rate": 5.92485549132948e-06,
+      "loss": 0.288,
+      "step": 1718
+    },
+    {
+      "epoch": 2.6780982073265784,
+      "grad_norm": 0.2057246513397782,
+      "learning_rate": 5.895953757225433e-06,
+      "loss": 0.2758,
+      "step": 1719
+    },
+    {
+      "epoch": 2.6796570537802027,
+      "grad_norm": 0.2291012037170924,
+      "learning_rate": 5.867052023121388e-06,
+      "loss": 0.2913,
+      "step": 1720
+    },
+    {
+      "epoch": 2.681215900233827,
+      "grad_norm": 0.2100404936453596,
+      "learning_rate": 5.838150289017341e-06,
+      "loss": 0.2817,
+      "step": 1721
+    },
+    {
+      "epoch": 2.682774746687451,
+      "grad_norm": 0.20420008955113567,
+      "learning_rate": 5.809248554913295e-06,
+      "loss": 0.2963,
+      "step": 1722
+    },
+    {
+      "epoch": 2.6843335931410754,
+      "grad_norm": 0.21002723265147977,
+      "learning_rate": 5.780346820809249e-06,
+      "loss": 0.3049,
+      "step": 1723
+    },
+    {
+      "epoch": 2.6858924395947,
+      "grad_norm": 0.20340290578451115,
+      "learning_rate": 5.751445086705202e-06,
+      "loss": 0.2847,
+      "step": 1724
+    },
+    {
+      "epoch": 2.6874512860483244,
+      "grad_norm": 0.21104439985364004,
+      "learning_rate": 5.722543352601157e-06,
+      "loss": 0.2821,
+      "step": 1725
+    },
+    {
+      "epoch": 2.6890101325019486,
+      "grad_norm": 0.21696962161072067,
+      "learning_rate": 5.69364161849711e-06,
+      "loss": 0.2721,
+      "step": 1726
+    },
+    {
+      "epoch": 2.690568978955573,
+      "grad_norm": 0.23015064920584716,
+      "learning_rate": 5.6647398843930635e-06,
+      "loss": 0.2922,
+      "step": 1727
+    },
+    {
+      "epoch": 2.692127825409197,
+      "grad_norm": 0.20813991146352637,
+      "learning_rate": 5.635838150289018e-06,
+      "loss": 0.2971,
+      "step": 1728
+    },
+    {
+      "epoch": 2.6936866718628214,
+      "grad_norm": 0.20649336747594163,
+      "learning_rate": 5.606936416184971e-06,
+      "loss": 0.3036,
+      "step": 1729
+    },
+    {
+      "epoch": 2.6952455183164457,
+      "grad_norm": 0.20331720787397545,
+      "learning_rate": 5.578034682080925e-06,
+      "loss": 0.2886,
+      "step": 1730
+    },
+    {
+      "epoch": 2.6968043647700704,
+      "grad_norm": 0.21078699708857532,
+      "learning_rate": 5.549132947976879e-06,
+      "loss": 0.2849,
+      "step": 1731
+    },
+    {
+      "epoch": 2.6983632112236946,
+      "grad_norm": 0.22036177053042216,
+      "learning_rate": 5.5202312138728324e-06,
+      "loss": 0.2861,
+      "step": 1732
+    },
+    {
+      "epoch": 2.699922057677319,
+      "grad_norm": 0.20172166515293294,
+      "learning_rate": 5.491329479768787e-06,
+      "loss": 0.2866,
+      "step": 1733
+    },
+    {
+      "epoch": 2.701480904130943,
+      "grad_norm": 0.19522835656935336,
+      "learning_rate": 5.46242774566474e-06,
+      "loss": 0.2681,
+      "step": 1734
+    },
+    {
+      "epoch": 2.7030397505845674,
+      "grad_norm": 0.2212443755898533,
+      "learning_rate": 5.433526011560694e-06,
+      "loss": 0.2958,
+      "step": 1735
+    },
+    {
+      "epoch": 2.7045985970381916,
+      "grad_norm": 0.19883348976389542,
+      "learning_rate": 5.404624277456648e-06,
+      "loss": 0.283,
+      "step": 1736
+    },
+    {
+      "epoch": 2.706157443491816,
+      "grad_norm": 0.22774567151755043,
+      "learning_rate": 5.375722543352601e-06,
+      "loss": 0.2817,
+      "step": 1737
+    },
+    {
+      "epoch": 2.7077162899454406,
+      "grad_norm": 0.19653848981003447,
+      "learning_rate": 5.346820809248556e-06,
+      "loss": 0.2807,
+      "step": 1738
+    },
+    {
+      "epoch": 2.709275136399065,
+      "grad_norm": 0.2155411949245046,
+      "learning_rate": 5.317919075144509e-06,
+      "loss": 0.2874,
+      "step": 1739
+    },
+    {
+      "epoch": 2.710833982852689,
+      "grad_norm": 0.21309269576853238,
+      "learning_rate": 5.2890173410404626e-06,
+      "loss": 0.2922,
+      "step": 1740
+    },
+    {
+      "epoch": 2.7123928293063133,
+      "grad_norm": 0.20084602651111383,
+      "learning_rate": 5.260115606936417e-06,
+      "loss": 0.2861,
+      "step": 1741
+    },
+    {
+      "epoch": 2.7139516757599376,
+      "grad_norm": 0.20949324216365292,
+      "learning_rate": 5.23121387283237e-06,
+      "loss": 0.2906,
+      "step": 1742
+    },
+    {
+      "epoch": 2.715510522213562,
+      "grad_norm": 0.19787967361130746,
+      "learning_rate": 5.202312138728324e-06,
+      "loss": 0.2851,
+      "step": 1743
+    },
+    {
+      "epoch": 2.717069368667186,
+      "grad_norm": 0.19777902230004707,
+      "learning_rate": 5.173410404624278e-06,
+      "loss": 0.2724,
+      "step": 1744
+    },
+    {
+      "epoch": 2.7186282151208108,
+      "grad_norm": 0.21000085676745223,
+      "learning_rate": 5.1445086705202315e-06,
+      "loss": 0.2863,
+      "step": 1745
+    },
+    {
+      "epoch": 2.720187061574435,
+      "grad_norm": 0.20127920458103757,
+      "learning_rate": 5.115606936416186e-06,
+      "loss": 0.2857,
+      "step": 1746
+    },
+    {
+      "epoch": 2.7217459080280593,
+      "grad_norm": 0.2090573509510013,
+      "learning_rate": 5.086705202312139e-06,
+      "loss": 0.2866,
+      "step": 1747
+    },
+    {
+      "epoch": 2.7233047544816835,
+      "grad_norm": 0.21739367096274437,
+      "learning_rate": 5.057803468208093e-06,
+      "loss": 0.2814,
+      "step": 1748
+    },
+    {
+      "epoch": 2.7248636009353078,
+      "grad_norm": 0.20833581410322963,
+      "learning_rate": 5.028901734104047e-06,
+      "loss": 0.2888,
+      "step": 1749
+    },
+    {
+      "epoch": 2.726422447388932,
+      "grad_norm": 0.19909078368387395,
+      "learning_rate": 5e-06,
+      "loss": 0.2911,
+      "step": 1750
+    },
+    {
+      "epoch": 2.7279812938425563,
+      "grad_norm": 0.20109886859179885,
+      "learning_rate": 4.971098265895955e-06,
+      "loss": 0.2895,
+      "step": 1751
+    },
+    {
+      "epoch": 2.729540140296181,
+      "grad_norm": 0.20247469803387927,
+      "learning_rate": 4.942196531791908e-06,
+      "loss": 0.276,
+      "step": 1752
+    },
+    {
+      "epoch": 2.7310989867498052,
+      "grad_norm": 0.20421110219784483,
+      "learning_rate": 4.913294797687862e-06,
+      "loss": 0.2792,
+      "step": 1753
+    },
+    {
+      "epoch": 2.7326578332034295,
+      "grad_norm": 0.19858031749032556,
+      "learning_rate": 4.884393063583816e-06,
+      "loss": 0.285,
+      "step": 1754
+    },
+    {
+      "epoch": 2.7342166796570537,
+      "grad_norm": 0.22215186509011026,
+      "learning_rate": 4.8554913294797685e-06,
+      "loss": 0.3038,
+      "step": 1755
+    },
+    {
+      "epoch": 2.735775526110678,
+      "grad_norm": 0.20256389152709856,
+      "learning_rate": 4.826589595375723e-06,
+      "loss": 0.2726,
+      "step": 1756
+    },
+    {
+      "epoch": 2.7373343725643027,
+      "grad_norm": 0.21034046831696598,
+      "learning_rate": 4.797687861271676e-06,
+      "loss": 0.296,
+      "step": 1757
+    },
+    {
+      "epoch": 2.7388932190179265,
+      "grad_norm": 0.20958728755008627,
+      "learning_rate": 4.76878612716763e-06,
+      "loss": 0.2868,
+      "step": 1758
+    },
+    {
+      "epoch": 2.740452065471551,
+      "grad_norm": 0.20383033185004612,
+      "learning_rate": 4.739884393063584e-06,
+      "loss": 0.2854,
+      "step": 1759
+    },
+    {
+      "epoch": 2.7420109119251754,
+      "grad_norm": 0.19907487917140468,
+      "learning_rate": 4.710982658959537e-06,
+      "loss": 0.285,
+      "step": 1760
+    },
+    {
+      "epoch": 2.7435697583787997,
+      "grad_norm": 0.20044771305150624,
+      "learning_rate": 4.682080924855492e-06,
+      "loss": 0.2816,
+      "step": 1761
+    },
+    {
+      "epoch": 2.745128604832424,
+      "grad_norm": 0.2164138120105804,
+      "learning_rate": 4.653179190751445e-06,
+      "loss": 0.2939,
+      "step": 1762
+    },
+    {
+      "epoch": 2.746687451286048,
+      "grad_norm": 0.24028484525744273,
+      "learning_rate": 4.624277456647399e-06,
+      "loss": 0.3133,
+      "step": 1763
+    },
+    {
+      "epoch": 2.748246297739673,
+      "grad_norm": 0.19984994036225665,
+      "learning_rate": 4.595375722543353e-06,
+      "loss": 0.2868,
+      "step": 1764
+    },
+    {
+      "epoch": 2.7498051441932967,
+      "grad_norm": 0.21636457592977681,
+      "learning_rate": 4.566473988439306e-06,
+      "loss": 0.2871,
+      "step": 1765
+    },
+    {
+      "epoch": 2.7513639906469214,
+      "grad_norm": 0.19990931527574785,
+      "learning_rate": 4.53757225433526e-06,
+      "loss": 0.2708,
+      "step": 1766
+    },
+    {
+      "epoch": 2.7529228371005456,
+      "grad_norm": 0.217527953437702,
+      "learning_rate": 4.508670520231214e-06,
+      "loss": 0.2972,
+      "step": 1767
+    },
+    {
+      "epoch": 2.75448168355417,
+      "grad_norm": 0.19223678719292475,
+      "learning_rate": 4.4797687861271675e-06,
+      "loss": 0.2898,
+      "step": 1768
+    },
+    {
+      "epoch": 2.756040530007794,
+      "grad_norm": 0.21317778941850382,
+      "learning_rate": 4.450867052023122e-06,
+      "loss": 0.2857,
+      "step": 1769
+    },
+    {
+      "epoch": 2.7575993764614184,
+      "grad_norm": 0.21322464171901906,
+      "learning_rate": 4.421965317919075e-06,
+      "loss": 0.2889,
+      "step": 1770
+    },
+    {
+      "epoch": 2.759158222915043,
+      "grad_norm": 0.21331351232714513,
+      "learning_rate": 4.393063583815029e-06,
+      "loss": 0.2787,
+      "step": 1771
+    },
+    {
+      "epoch": 2.7607170693686673,
+      "grad_norm": 0.20670483032146136,
+      "learning_rate": 4.364161849710983e-06,
+      "loss": 0.2961,
+      "step": 1772
+    },
+    {
+      "epoch": 2.7622759158222916,
+      "grad_norm": 0.22435785586885243,
+      "learning_rate": 4.3352601156069365e-06,
+      "loss": 0.2943,
+      "step": 1773
+    },
+    {
+      "epoch": 2.763834762275916,
+      "grad_norm": 0.19815210166141742,
+      "learning_rate": 4.30635838150289e-06,
+      "loss": 0.2935,
+      "step": 1774
+    },
+    {
+      "epoch": 2.76539360872954,
+      "grad_norm": 0.20921721259317946,
+      "learning_rate": 4.277456647398844e-06,
+      "loss": 0.2711,
+      "step": 1775
+    },
+    {
+      "epoch": 2.7669524551831643,
+      "grad_norm": 0.2239327598983314,
+      "learning_rate": 4.248554913294798e-06,
+      "loss": 0.3043,
+      "step": 1776
+    },
+    {
+      "epoch": 2.7685113016367886,
+      "grad_norm": 0.2004589789574725,
+      "learning_rate": 4.219653179190752e-06,
+      "loss": 0.2964,
+      "step": 1777
+    },
+    {
+      "epoch": 2.7700701480904133,
+      "grad_norm": 0.20625941199900238,
+      "learning_rate": 4.190751445086705e-06,
+      "loss": 0.2907,
+      "step": 1778
+    },
+    {
+      "epoch": 2.7716289945440375,
+      "grad_norm": 0.20519545434025108,
+      "learning_rate": 4.161849710982659e-06,
+      "loss": 0.3068,
+      "step": 1779
+    },
+    {
+      "epoch": 2.773187840997662,
+      "grad_norm": 0.21554367954746417,
+      "learning_rate": 4.132947976878613e-06,
+      "loss": 0.2952,
+      "step": 1780
+    },
+    {
+      "epoch": 2.774746687451286,
+      "grad_norm": 0.21272563121891813,
+      "learning_rate": 4.1040462427745666e-06,
+      "loss": 0.2842,
+      "step": 1781
+    },
+    {
+      "epoch": 2.7763055339049103,
+      "grad_norm": 0.21611401595167964,
+      "learning_rate": 4.075144508670521e-06,
+      "loss": 0.3046,
+      "step": 1782
+    },
+    {
+      "epoch": 2.7778643803585346,
+      "grad_norm": 0.1997924444181988,
+      "learning_rate": 4.046242774566474e-06,
+      "loss": 0.2931,
+      "step": 1783
+    },
+    {
+      "epoch": 2.779423226812159,
+      "grad_norm": 0.21506298028146192,
+      "learning_rate": 4.017341040462428e-06,
+      "loss": 0.2832,
+      "step": 1784
+    },
+    {
+      "epoch": 2.7809820732657835,
+      "grad_norm": 0.20021184702963296,
+      "learning_rate": 3.988439306358382e-06,
+      "loss": 0.29,
+      "step": 1785
+    },
+    {
+      "epoch": 2.7825409197194078,
+      "grad_norm": 0.20970397205026278,
+      "learning_rate": 3.9595375722543355e-06,
+      "loss": 0.3047,
+      "step": 1786
+    },
+    {
+      "epoch": 2.784099766173032,
+      "grad_norm": 0.20062422216247,
+      "learning_rate": 3.930635838150289e-06,
+      "loss": 0.2844,
+      "step": 1787
+    },
+    {
+      "epoch": 2.7856586126266563,
+      "grad_norm": 0.20179644813019854,
+      "learning_rate": 3.901734104046243e-06,
+      "loss": 0.2842,
+      "step": 1788
+    },
+    {
+      "epoch": 2.7872174590802805,
+      "grad_norm": 0.20018746664410972,
+      "learning_rate": 3.872832369942197e-06,
+      "loss": 0.2892,
+      "step": 1789
+    },
+    {
+      "epoch": 2.7887763055339048,
+      "grad_norm": 0.1946501903871324,
+      "learning_rate": 3.843930635838151e-06,
+      "loss": 0.29,
+      "step": 1790
+    },
+    {
+      "epoch": 2.790335151987529,
+      "grad_norm": 0.19085065858233305,
+      "learning_rate": 3.815028901734104e-06,
+      "loss": 0.275,
+      "step": 1791
+    },
+    {
+      "epoch": 2.7918939984411537,
+      "grad_norm": 0.20722259921075883,
+      "learning_rate": 3.7861271676300583e-06,
+      "loss": 0.2804,
+      "step": 1792
+    },
+    {
+      "epoch": 2.793452844894778,
+      "grad_norm": 0.2103035874157165,
+      "learning_rate": 3.757225433526012e-06,
+      "loss": 0.2896,
+      "step": 1793
+    },
+    {
+      "epoch": 2.795011691348402,
+      "grad_norm": 0.2102336939252641,
+      "learning_rate": 3.7283236994219656e-06,
+      "loss": 0.2779,
+      "step": 1794
+    },
+    {
+      "epoch": 2.7965705378020265,
+      "grad_norm": 0.20336255858279942,
+      "learning_rate": 3.6994219653179195e-06,
+      "loss": 0.2933,
+      "step": 1795
+    },
+    {
+      "epoch": 2.7981293842556507,
+      "grad_norm": 0.21206172168707907,
+      "learning_rate": 3.6705202312138733e-06,
+      "loss": 0.2786,
+      "step": 1796
+    },
+    {
+      "epoch": 2.799688230709275,
+      "grad_norm": 0.19753691873632548,
+      "learning_rate": 3.6416184971098272e-06,
+      "loss": 0.2925,
+      "step": 1797
+    },
+    {
+      "epoch": 2.8012470771628992,
+      "grad_norm": 0.21025169769964883,
+      "learning_rate": 3.6127167630057807e-06,
+      "loss": 0.2968,
+      "step": 1798
+    },
+    {
+      "epoch": 2.802805923616524,
+      "grad_norm": 0.22618633232114826,
+      "learning_rate": 3.5838150289017345e-06,
+      "loss": 0.2959,
+      "step": 1799
+    },
+    {
+      "epoch": 2.804364770070148,
+      "grad_norm": 0.21822500972127065,
+      "learning_rate": 3.5549132947976884e-06,
+      "loss": 0.3092,
+      "step": 1800
+    },
+    {
+      "epoch": 2.8059236165237724,
+      "grad_norm": 0.220482613302332,
+      "learning_rate": 3.5260115606936423e-06,
+      "loss": 0.2955,
+      "step": 1801
+    },
+    {
+      "epoch": 2.8074824629773967,
+      "grad_norm": 0.20780910132598235,
+      "learning_rate": 3.4971098265895957e-06,
+      "loss": 0.2761,
+      "step": 1802
+    },
+    {
+      "epoch": 2.809041309431021,
+      "grad_norm": 0.2072002447217666,
+      "learning_rate": 3.468208092485549e-06,
+      "loss": 0.2915,
+      "step": 1803
+    },
+    {
+      "epoch": 2.8106001558846456,
+      "grad_norm": 0.20725271527567482,
+      "learning_rate": 3.4393063583815026e-06,
+      "loss": 0.2876,
+      "step": 1804
+    },
+    {
+      "epoch": 2.8121590023382694,
+      "grad_norm": 0.2101105745995178,
+      "learning_rate": 3.4104046242774565e-06,
+      "loss": 0.2856,
+      "step": 1805
+    },
+    {
+      "epoch": 2.813717848791894,
+      "grad_norm": 0.20498941201400542,
+      "learning_rate": 3.3815028901734103e-06,
+      "loss": 0.2796,
+      "step": 1806
+    },
+    {
+      "epoch": 2.8152766952455184,
+      "grad_norm": 0.20462911144204035,
+      "learning_rate": 3.3526011560693642e-06,
+      "loss": 0.3016,
+      "step": 1807
+    },
+    {
+      "epoch": 2.8168355416991426,
+      "grad_norm": 0.2019060859936445,
+      "learning_rate": 3.3236994219653177e-06,
+      "loss": 0.2805,
+      "step": 1808
+    },
+    {
+      "epoch": 2.818394388152767,
+      "grad_norm": 0.2124738066306522,
+      "learning_rate": 3.2947976878612715e-06,
+      "loss": 0.2893,
+      "step": 1809
+    },
+    {
+      "epoch": 2.819953234606391,
+      "grad_norm": 0.2088568665930041,
+      "learning_rate": 3.2658959537572254e-06,
+      "loss": 0.2844,
+      "step": 1810
+    },
+    {
+      "epoch": 2.821512081060016,
+      "grad_norm": 0.19645688715836024,
+      "learning_rate": 3.2369942196531793e-06,
+      "loss": 0.2749,
+      "step": 1811
+    },
+    {
+      "epoch": 2.8230709275136396,
+      "grad_norm": 0.19409466738213488,
+      "learning_rate": 3.2080924855491327e-06,
+      "loss": 0.2812,
+      "step": 1812
+    },
+    {
+      "epoch": 2.8246297739672643,
+      "grad_norm": 0.2120230001481611,
+      "learning_rate": 3.1791907514450866e-06,
+      "loss": 0.2954,
+      "step": 1813
+    },
+    {
+      "epoch": 2.8261886204208886,
+      "grad_norm": 0.1984563439139553,
+      "learning_rate": 3.1502890173410405e-06,
+      "loss": 0.2984,
+      "step": 1814
+    },
+    {
+      "epoch": 2.827747466874513,
+      "grad_norm": 0.20703019708211207,
+      "learning_rate": 3.1213872832369943e-06,
+      "loss": 0.2818,
+      "step": 1815
+    },
+    {
+      "epoch": 2.829306313328137,
+      "grad_norm": 0.2073483200049451,
+      "learning_rate": 3.0924855491329478e-06,
+      "loss": 0.2945,
+      "step": 1816
+    },
+    {
+      "epoch": 2.8308651597817613,
+      "grad_norm": 0.20322071664588748,
+      "learning_rate": 3.0635838150289016e-06,
+      "loss": 0.28,
+      "step": 1817
+    },
+    {
+      "epoch": 2.832424006235386,
+      "grad_norm": 0.19200180838930353,
+      "learning_rate": 3.0346820809248555e-06,
+      "loss": 0.2899,
+      "step": 1818
+    },
+    {
+      "epoch": 2.8339828526890103,
+      "grad_norm": 0.19563958688827382,
+      "learning_rate": 3.0057803468208094e-06,
+      "loss": 0.282,
+      "step": 1819
+    },
+    {
+      "epoch": 2.8355416991426345,
+      "grad_norm": 0.1999256570053609,
+      "learning_rate": 2.9768786127167633e-06,
+      "loss": 0.2915,
+      "step": 1820
+    },
+    {
+      "epoch": 2.837100545596259,
+      "grad_norm": 0.19601345181511165,
+      "learning_rate": 2.9479768786127167e-06,
+      "loss": 0.2846,
+      "step": 1821
+    },
+    {
+      "epoch": 2.838659392049883,
+      "grad_norm": 0.2099789860276315,
+      "learning_rate": 2.9190751445086706e-06,
+      "loss": 0.2807,
+      "step": 1822
+    },
+    {
+      "epoch": 2.8402182385035073,
+      "grad_norm": 0.20747261090512478,
+      "learning_rate": 2.8901734104046244e-06,
+      "loss": 0.279,
+      "step": 1823
+    },
+    {
+      "epoch": 2.8417770849571315,
+      "grad_norm": 0.22257411158836346,
+      "learning_rate": 2.8612716763005783e-06,
+      "loss": 0.2997,
+      "step": 1824
+    },
+    {
+      "epoch": 2.8433359314107562,
+      "grad_norm": 0.20554349028187938,
+      "learning_rate": 2.8323699421965318e-06,
+      "loss": 0.2831,
+      "step": 1825
+    },
+    {
+      "epoch": 2.8448947778643805,
+      "grad_norm": 0.20735600722529712,
+      "learning_rate": 2.8034682080924856e-06,
+      "loss": 0.294,
+      "step": 1826
+    },
+    {
+      "epoch": 2.8464536243180047,
+      "grad_norm": 0.21980476595040552,
+      "learning_rate": 2.7745664739884395e-06,
+      "loss": 0.3026,
+      "step": 1827
+    },
+    {
+      "epoch": 2.848012470771629,
+      "grad_norm": 0.1925143943177367,
+      "learning_rate": 2.7456647398843934e-06,
+      "loss": 0.2876,
+      "step": 1828
+    },
+    {
+      "epoch": 2.8495713172252533,
+      "grad_norm": 0.1957010171931748,
+      "learning_rate": 2.716763005780347e-06,
+      "loss": 0.2859,
+      "step": 1829
+    },
+    {
+      "epoch": 2.8511301636788775,
+      "grad_norm": 0.20358594514165249,
+      "learning_rate": 2.6878612716763007e-06,
+      "loss": 0.282,
+      "step": 1830
+    },
+    {
+      "epoch": 2.8526890101325018,
+      "grad_norm": 0.20541392956189472,
+      "learning_rate": 2.6589595375722546e-06,
+      "loss": 0.2795,
+      "step": 1831
+    },
+    {
+      "epoch": 2.8542478565861265,
+      "grad_norm": 0.1956976623325346,
+      "learning_rate": 2.6300578034682084e-06,
+      "loss": 0.2753,
+      "step": 1832
+    },
+    {
+      "epoch": 2.8558067030397507,
+      "grad_norm": 0.1944502240686237,
+      "learning_rate": 2.601156069364162e-06,
+      "loss": 0.2779,
+      "step": 1833
+    },
+    {
+      "epoch": 2.857365549493375,
+      "grad_norm": 0.1842303161830082,
+      "learning_rate": 2.5722543352601157e-06,
+      "loss": 0.2747,
+      "step": 1834
+    },
+    {
+      "epoch": 2.858924395946999,
+      "grad_norm": 0.19270326923177308,
+      "learning_rate": 2.5433526011560696e-06,
+      "loss": 0.2895,
+      "step": 1835
+    },
+    {
+      "epoch": 2.8604832424006235,
+      "grad_norm": 0.20578389463380026,
+      "learning_rate": 2.5144508670520235e-06,
+      "loss": 0.2784,
+      "step": 1836
+    },
+    {
+      "epoch": 2.8620420888542477,
+      "grad_norm": 0.18959748823440806,
+      "learning_rate": 2.4855491329479774e-06,
+      "loss": 0.2941,
+      "step": 1837
+    },
+    {
+      "epoch": 2.863600935307872,
+      "grad_norm": 0.189041300234095,
+      "learning_rate": 2.456647398843931e-06,
+      "loss": 0.2863,
+      "step": 1838
+    },
+    {
+      "epoch": 2.8651597817614967,
+      "grad_norm": 0.21031320630114367,
+      "learning_rate": 2.4277456647398842e-06,
+      "loss": 0.2996,
+      "step": 1839
+    },
+    {
+      "epoch": 2.866718628215121,
+      "grad_norm": 0.20690718603806962,
+      "learning_rate": 2.398843930635838e-06,
+      "loss": 0.2812,
+      "step": 1840
+    },
+    {
+      "epoch": 2.868277474668745,
+      "grad_norm": 0.2073404491146693,
+      "learning_rate": 2.369942196531792e-06,
+      "loss": 0.2947,
+      "step": 1841
+    },
+    {
+      "epoch": 2.8698363211223694,
+      "grad_norm": 0.19029334104582032,
+      "learning_rate": 2.341040462427746e-06,
+      "loss": 0.2747,
+      "step": 1842
+    },
+    {
+      "epoch": 2.8713951675759937,
+      "grad_norm": 0.19017238272079218,
+      "learning_rate": 2.3121387283236993e-06,
+      "loss": 0.2847,
+      "step": 1843
+    },
+    {
+      "epoch": 2.872954014029618,
+      "grad_norm": 0.1979579277272625,
+      "learning_rate": 2.283236994219653e-06,
+      "loss": 0.2966,
+      "step": 1844
+    },
+    {
+      "epoch": 2.874512860483242,
+      "grad_norm": 0.20597366624414118,
+      "learning_rate": 2.254335260115607e-06,
+      "loss": 0.2967,
+      "step": 1845
+    },
+    {
+      "epoch": 2.876071706936867,
+      "grad_norm": 0.1914660504884028,
+      "learning_rate": 2.225433526011561e-06,
+      "loss": 0.2847,
+      "step": 1846
+    },
+    {
+      "epoch": 2.877630553390491,
+      "grad_norm": 0.21004523817562415,
+      "learning_rate": 2.1965317919075144e-06,
+      "loss": 0.2844,
+      "step": 1847
+    },
+    {
+      "epoch": 2.8791893998441154,
+      "grad_norm": 0.2050510627988167,
+      "learning_rate": 2.1676300578034682e-06,
+      "loss": 0.3026,
+      "step": 1848
+    },
+    {
+      "epoch": 2.8807482462977396,
+      "grad_norm": 0.19875825849528217,
+      "learning_rate": 2.138728323699422e-06,
+      "loss": 0.29,
+      "step": 1849
+    },
+    {
+      "epoch": 2.882307092751364,
+      "grad_norm": 0.19201938735986301,
+      "learning_rate": 2.109826589595376e-06,
+      "loss": 0.2814,
+      "step": 1850
+    },
+    {
+      "epoch": 2.8838659392049886,
+      "grad_norm": 0.19276619456964975,
+      "learning_rate": 2.0809248554913294e-06,
+      "loss": 0.2794,
+      "step": 1851
+    },
+    {
+      "epoch": 2.8854247856586124,
+      "grad_norm": 0.20837326668805686,
+      "learning_rate": 2.0520231213872833e-06,
+      "loss": 0.274,
+      "step": 1852
+    },
+    {
+      "epoch": 2.886983632112237,
+      "grad_norm": 0.20834168208438708,
+      "learning_rate": 2.023121387283237e-06,
+      "loss": 0.2817,
+      "step": 1853
+    },
+    {
+      "epoch": 2.8885424785658613,
+      "grad_norm": 0.20432457017704334,
+      "learning_rate": 1.994219653179191e-06,
+      "loss": 0.2949,
+      "step": 1854
+    },
+    {
+      "epoch": 2.8901013250194856,
+      "grad_norm": 0.20163276098807092,
+      "learning_rate": 1.9653179190751445e-06,
+      "loss": 0.2706,
+      "step": 1855
+    },
+    {
+      "epoch": 2.89166017147311,
+      "grad_norm": 0.20977626122832665,
+      "learning_rate": 1.9364161849710983e-06,
+      "loss": 0.3086,
+      "step": 1856
+    },
+    {
+      "epoch": 2.893219017926734,
+      "grad_norm": 0.20060594756785594,
+      "learning_rate": 1.907514450867052e-06,
+      "loss": 0.2911,
+      "step": 1857
+    },
+    {
+      "epoch": 2.894777864380359,
+      "grad_norm": 0.20293612568917765,
+      "learning_rate": 1.878612716763006e-06,
+      "loss": 0.2911,
+      "step": 1858
+    },
+    {
+      "epoch": 2.8963367108339826,
+      "grad_norm": 0.2096162037257313,
+      "learning_rate": 1.8497109826589597e-06,
+      "loss": 0.2918,
+      "step": 1859
+    },
+    {
+      "epoch": 2.8978955572876073,
+      "grad_norm": 0.21042720816672483,
+      "learning_rate": 1.8208092485549136e-06,
+      "loss": 0.3119,
+      "step": 1860
+    },
+    {
+      "epoch": 2.8994544037412315,
+      "grad_norm": 0.18793473718435338,
+      "learning_rate": 1.7919075144508673e-06,
+      "loss": 0.2752,
+      "step": 1861
+    },
+    {
+      "epoch": 2.901013250194856,
+      "grad_norm": 0.2079973480121488,
+      "learning_rate": 1.7630057803468211e-06,
+      "loss": 0.3029,
+      "step": 1862
+    },
+    {
+      "epoch": 2.90257209664848,
+      "grad_norm": 0.1961639560813693,
+      "learning_rate": 1.7341040462427746e-06,
+      "loss": 0.2916,
+      "step": 1863
+    },
+    {
+      "epoch": 2.9041309431021043,
+      "grad_norm": 0.1907492865888087,
+      "learning_rate": 1.7052023121387282e-06,
+      "loss": 0.2789,
+      "step": 1864
+    },
+    {
+      "epoch": 2.905689789555729,
+      "grad_norm": 0.20235805559023728,
+      "learning_rate": 1.6763005780346821e-06,
+      "loss": 0.2849,
+      "step": 1865
+    },
+    {
+      "epoch": 2.9072486360093532,
+      "grad_norm": 0.20288572024690002,
+      "learning_rate": 1.6473988439306358e-06,
+      "loss": 0.2654,
+      "step": 1866
+    },
+    {
+      "epoch": 2.9088074824629775,
+      "grad_norm": 0.20289343651469097,
+      "learning_rate": 1.6184971098265896e-06,
+      "loss": 0.2951,
+      "step": 1867
+    },
+    {
+      "epoch": 2.9103663289166017,
+      "grad_norm": 0.18922720141621407,
+      "learning_rate": 1.5895953757225433e-06,
+      "loss": 0.2762,
+      "step": 1868
+    },
+    {
+      "epoch": 2.911925175370226,
+      "grad_norm": 0.19241261968232612,
+      "learning_rate": 1.5606936416184972e-06,
+      "loss": 0.2881,
+      "step": 1869
+    },
+    {
+      "epoch": 2.9134840218238502,
+      "grad_norm": 0.19674852841020815,
+      "learning_rate": 1.5317919075144508e-06,
+      "loss": 0.2832,
+      "step": 1870
+    },
+    {
+      "epoch": 2.9150428682774745,
+      "grad_norm": 0.18740622725667494,
+      "learning_rate": 1.5028901734104047e-06,
+      "loss": 0.283,
+      "step": 1871
+    },
+    {
+      "epoch": 2.916601714731099,
+      "grad_norm": 0.20669295826778195,
+      "learning_rate": 1.4739884393063584e-06,
+      "loss": 0.2808,
+      "step": 1872
+    },
+    {
+      "epoch": 2.9181605611847234,
+      "grad_norm": 0.18108268743966033,
+      "learning_rate": 1.4450867052023122e-06,
+      "loss": 0.2708,
+      "step": 1873
+    },
+    {
+      "epoch": 2.9197194076383477,
+      "grad_norm": 0.19216377105239985,
+      "learning_rate": 1.4161849710982659e-06,
+      "loss": 0.2798,
+      "step": 1874
+    },
+    {
+      "epoch": 2.921278254091972,
+      "grad_norm": 0.2036871353826851,
+      "learning_rate": 1.3872832369942197e-06,
+      "loss": 0.2871,
+      "step": 1875
+    },
+    {
+      "epoch": 2.922837100545596,
+      "grad_norm": 0.19352818374916894,
+      "learning_rate": 1.3583815028901734e-06,
+      "loss": 0.2982,
+      "step": 1876
+    },
+    {
+      "epoch": 2.9243959469992205,
+      "grad_norm": 0.19567432642222407,
+      "learning_rate": 1.3294797687861273e-06,
+      "loss": 0.2832,
+      "step": 1877
+    },
+    {
+      "epoch": 2.9259547934528447,
+      "grad_norm": 0.19512011366905269,
+      "learning_rate": 1.300578034682081e-06,
+      "loss": 0.2839,
+      "step": 1878
+    },
+    {
+      "epoch": 2.9275136399064694,
+      "grad_norm": 0.202990327713183,
+      "learning_rate": 1.2716763005780348e-06,
+      "loss": 0.2823,
+      "step": 1879
+    },
+    {
+      "epoch": 2.9290724863600937,
+      "grad_norm": 0.20583389454978893,
+      "learning_rate": 1.2427745664739887e-06,
+      "loss": 0.2988,
+      "step": 1880
+    },
+    {
+      "epoch": 2.930631332813718,
+      "grad_norm": 0.2014751871494088,
+      "learning_rate": 1.2138728323699421e-06,
+      "loss": 0.293,
+      "step": 1881
+    },
+    {
+      "epoch": 2.932190179267342,
+      "grad_norm": 0.19892906821370257,
+      "learning_rate": 1.184971098265896e-06,
+      "loss": 0.2949,
+      "step": 1882
+    },
+    {
+      "epoch": 2.9337490257209664,
+      "grad_norm": 0.18958700311963878,
+      "learning_rate": 1.1560693641618497e-06,
+      "loss": 0.2826,
+      "step": 1883
+    },
+    {
+      "epoch": 2.9353078721745907,
+      "grad_norm": 0.19062254575039358,
+      "learning_rate": 1.1271676300578035e-06,
+      "loss": 0.2734,
+      "step": 1884
+    },
+    {
+      "epoch": 2.936866718628215,
+      "grad_norm": 0.19081647607459015,
+      "learning_rate": 1.0982658959537572e-06,
+      "loss": 0.2719,
+      "step": 1885
+    },
+    {
+      "epoch": 2.9384255650818396,
+      "grad_norm": 0.20749991571237006,
+      "learning_rate": 1.069364161849711e-06,
+      "loss": 0.2902,
+      "step": 1886
+    },
+    {
+      "epoch": 2.939984411535464,
+      "grad_norm": 0.201338883635779,
+      "learning_rate": 1.0404624277456647e-06,
+      "loss": 0.2866,
+      "step": 1887
+    },
+    {
+      "epoch": 2.941543257989088,
+      "grad_norm": 0.185374324630895,
+      "learning_rate": 1.0115606936416186e-06,
+      "loss": 0.291,
+      "step": 1888
+    },
+    {
+      "epoch": 2.9431021044427124,
+      "grad_norm": 0.19169182164611995,
+      "learning_rate": 9.826589595375722e-07,
+      "loss": 0.2919,
+      "step": 1889
+    },
+    {
+      "epoch": 2.9446609508963366,
+      "grad_norm": 0.19500681675014075,
+      "learning_rate": 9.53757225433526e-07,
+      "loss": 0.289,
+      "step": 1890
+    },
+    {
+      "epoch": 2.946219797349961,
+      "grad_norm": 0.20336352790343418,
+      "learning_rate": 9.248554913294799e-07,
+      "loss": 0.2888,
+      "step": 1891
+    },
+    {
+      "epoch": 2.947778643803585,
+      "grad_norm": 0.18996694302336917,
+      "learning_rate": 8.959537572254336e-07,
+      "loss": 0.2837,
+      "step": 1892
+    },
+    {
+      "epoch": 2.94933749025721,
+      "grad_norm": 0.2156301542305652,
+      "learning_rate": 8.670520231213873e-07,
+      "loss": 0.2799,
+      "step": 1893
+    },
+    {
+      "epoch": 2.950896336710834,
+      "grad_norm": 0.19727279525402547,
+      "learning_rate": 8.381502890173411e-07,
+      "loss": 0.2906,
+      "step": 1894
+    },
+    {
+      "epoch": 2.9524551831644583,
+      "grad_norm": 0.19284045282449663,
+      "learning_rate": 8.092485549132948e-07,
+      "loss": 0.2795,
+      "step": 1895
+    },
+    {
+      "epoch": 2.9540140296180826,
+      "grad_norm": 0.20089648135420463,
+      "learning_rate": 7.803468208092486e-07,
+      "loss": 0.2857,
+      "step": 1896
+    },
+    {
+      "epoch": 2.955572876071707,
+      "grad_norm": 0.18636348684721002,
+      "learning_rate": 7.514450867052023e-07,
+      "loss": 0.28,
+      "step": 1897
+    },
+    {
+      "epoch": 2.9571317225253315,
+      "grad_norm": 0.1842759397599328,
+      "learning_rate": 7.225433526011561e-07,
+      "loss": 0.2811,
+      "step": 1898
+    },
+    {
+      "epoch": 2.9586905689789553,
+      "grad_norm": 0.198586721483869,
+      "learning_rate": 6.936416184971099e-07,
+      "loss": 0.2795,
+      "step": 1899
+    },
+    {
+      "epoch": 2.96024941543258,
+      "grad_norm": 0.20175920964436003,
+      "learning_rate": 6.647398843930636e-07,
+      "loss": 0.2963,
+      "step": 1900
+    },
+    {
+      "epoch": 2.9618082618862043,
+      "grad_norm": 0.1894255377343828,
+      "learning_rate": 6.358381502890174e-07,
+      "loss": 0.2786,
+      "step": 1901
+    },
+    {
+      "epoch": 2.9633671083398285,
+      "grad_norm": 0.18846655912423485,
+      "learning_rate": 6.069364161849711e-07,
+      "loss": 0.2755,
+      "step": 1902
+    },
+    {
+      "epoch": 2.964925954793453,
+      "grad_norm": 0.1928139829341696,
+      "learning_rate": 5.780346820809248e-07,
+      "loss": 0.2999,
+      "step": 1903
+    },
+    {
+      "epoch": 2.966484801247077,
+      "grad_norm": 0.19323847526640792,
+      "learning_rate": 5.491329479768786e-07,
+      "loss": 0.2937,
+      "step": 1904
+    },
+    {
+      "epoch": 2.9680436477007017,
+      "grad_norm": 0.20120482130965053,
+      "learning_rate": 5.202312138728324e-07,
+      "loss": 0.2947,
+      "step": 1905
+    },
+    {
+      "epoch": 2.9696024941543255,
+      "grad_norm": 0.21061414202296763,
+      "learning_rate": 4.913294797687861e-07,
+      "loss": 0.2905,
+      "step": 1906
+    },
+    {
+      "epoch": 2.9711613406079502,
+      "grad_norm": 0.21063326211832292,
+      "learning_rate": 4.6242774566473993e-07,
+      "loss": 0.2965,
+      "step": 1907
+    },
+    {
+      "epoch": 2.9727201870615745,
+      "grad_norm": 0.20410862279580308,
+      "learning_rate": 4.3352601156069365e-07,
+      "loss": 0.2891,
+      "step": 1908
+    },
+    {
+      "epoch": 2.9742790335151987,
+      "grad_norm": 0.19836165458820448,
+      "learning_rate": 4.046242774566474e-07,
+      "loss": 0.2962,
+      "step": 1909
+    },
+    {
+      "epoch": 2.975837879968823,
+      "grad_norm": 0.19040393729360364,
+      "learning_rate": 3.7572254335260117e-07,
+      "loss": 0.2986,
+      "step": 1910
+    },
+    {
+      "epoch": 2.9773967264224472,
+      "grad_norm": 0.18498159624839047,
+      "learning_rate": 3.4682080924855494e-07,
+      "loss": 0.2855,
+      "step": 1911
+    },
+    {
+      "epoch": 2.978955572876072,
+      "grad_norm": 0.19089075468074318,
+      "learning_rate": 3.179190751445087e-07,
+      "loss": 0.2913,
+      "step": 1912
+    },
+    {
+      "epoch": 2.980514419329696,
+      "grad_norm": 0.19143218775583679,
+      "learning_rate": 2.890173410404624e-07,
+      "loss": 0.2739,
+      "step": 1913
+    },
+    {
+      "epoch": 2.9820732657833204,
+      "grad_norm": 0.19760714438972718,
+      "learning_rate": 2.601156069364162e-07,
+      "loss": 0.2955,
+      "step": 1914
+    },
+    {
+      "epoch": 2.9836321122369447,
+      "grad_norm": 0.18341652706228365,
+      "learning_rate": 2.3121387283236997e-07,
+      "loss": 0.2713,
+      "step": 1915
+    },
+    {
+      "epoch": 2.985190958690569,
+      "grad_norm": 0.19428086051572635,
+      "learning_rate": 2.023121387283237e-07,
+      "loss": 0.2867,
+      "step": 1916
+    },
+    {
+      "epoch": 2.986749805144193,
+      "grad_norm": 0.19485137014886658,
+      "learning_rate": 1.7341040462427747e-07,
+      "loss": 0.2761,
+      "step": 1917
+    },
+    {
+      "epoch": 2.9883086515978174,
+      "grad_norm": 0.19777851419652856,
+      "learning_rate": 1.445086705202312e-07,
+      "loss": 0.2823,
+      "step": 1918
+    },
+    {
+      "epoch": 2.989867498051442,
+      "grad_norm": 0.19065668774947628,
+      "learning_rate": 1.1560693641618498e-07,
+      "loss": 0.2954,
+      "step": 1919
+    },
+    {
+      "epoch": 2.9914263445050664,
+      "grad_norm": 0.1974490266526018,
+      "learning_rate": 8.670520231213873e-08,
+      "loss": 0.2813,
+      "step": 1920
+    },
+    {
+      "epoch": 2.9929851909586906,
+      "grad_norm": 0.19818812256280618,
+      "learning_rate": 5.780346820809249e-08,
+      "loss": 0.2773,
+      "step": 1921
+    },
+    {
+      "epoch": 2.994544037412315,
+      "grad_norm": 0.1943618741709864,
+      "learning_rate": 2.8901734104046246e-08,
+      "loss": 0.2787,
+      "step": 1922
+    },
+    {
+      "epoch": 2.996102883865939,
+      "grad_norm": 0.1954427128274177,
+      "learning_rate": 0.0,
+      "loss": 0.278,
+      "step": 1923
+    },
+    {
+      "epoch": 2.996102883865939,
+      "step": 1923,
+      "total_flos": 1.6424334483479593e+18,
+      "train_loss": 0.4532742842497208,
+      "train_runtime": 111641.3311,
+      "train_samples_per_second": 0.276,
+      "train_steps_per_second": 0.017
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1923,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.6424334483479593e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}