diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,37913 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.6187745924676786,
+  "eval_steps": 500,
+  "global_step": 5400,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00029979389169945664,
+      "grad_norm": 4.617544174194336,
+      "learning_rate": 2.9999999999999997e-05,
+      "loss": 1.26,
+      "step": 1
+    },
+    {
+      "epoch": 0.0005995877833989133,
+      "grad_norm": 5.574990272521973,
+      "learning_rate": 5.9999999999999995e-05,
+      "loss": 1.2193,
+      "step": 2
+    },
+    {
+      "epoch": 0.0008993816750983699,
+      "grad_norm": 2.685803174972534,
+      "learning_rate": 8.999999999999999e-05,
+      "loss": 1.2098,
+      "step": 3
+    },
+    {
+      "epoch": 0.0011991755667978266,
+      "grad_norm": 1.2137216329574585,
+      "learning_rate": 0.00011999999999999999,
+      "loss": 1.0505,
+      "step": 4
+    },
+    {
+      "epoch": 0.001498969458497283,
+      "grad_norm": 1.1230342388153076,
+      "learning_rate": 0.00015,
+      "loss": 0.8918,
+      "step": 5
+    },
+    {
+      "epoch": 0.0017987633501967398,
+      "grad_norm": 0.9145472049713135,
+      "learning_rate": 0.00017999999999999998,
+      "loss": 0.8306,
+      "step": 6
+    },
+    {
+      "epoch": 0.0020985572418961962,
+      "grad_norm": 1.5094902515411377,
+      "learning_rate": 0.00020999999999999998,
+      "loss": 0.788,
+      "step": 7
+    },
+    {
+      "epoch": 0.002398351133595653,
+      "grad_norm": 0.5805755257606506,
+      "learning_rate": 0.00023999999999999998,
+      "loss": 0.8026,
+      "step": 8
+    },
+    {
+      "epoch": 0.0026981450252951096,
+      "grad_norm": 0.338878333568573,
+      "learning_rate": 0.00027,
+      "loss": 0.723,
+      "step": 9
+    },
+    {
+      "epoch": 0.002997938916994566,
+      "grad_norm": 0.34122711420059204,
+      "learning_rate": 0.0003,
+      "loss": 0.705,
+      "step": 10
+    },
+    {
+      "epoch": 0.003297732808694023,
+      "grad_norm": 0.3165666162967682,
+      "learning_rate": 0.0002999549549549549,
+      "loss": 0.7282,
+      "step": 11
+    },
+    {
+      "epoch": 0.0035975267003934795,
+      "grad_norm": 0.3415158689022064,
+      "learning_rate": 0.0002999099099099099,
+      "loss": 0.6833,
+      "step": 12
+    },
+    {
+      "epoch": 0.003897320592092936,
+      "grad_norm": 0.3431508243083954,
+      "learning_rate": 0.00029986486486486484,
+      "loss": 0.713,
+      "step": 13
+    },
+    {
+      "epoch": 0.0041971144837923925,
+      "grad_norm": 0.3494497239589691,
+      "learning_rate": 0.0002998198198198198,
+      "loss": 0.6822,
+      "step": 14
+    },
+    {
+      "epoch": 0.004496908375491849,
+      "grad_norm": 0.29291579127311707,
+      "learning_rate": 0.00029977477477477477,
+      "loss": 0.6348,
+      "step": 15
+    },
+    {
+      "epoch": 0.004796702267191306,
+      "grad_norm": 0.3061606287956238,
+      "learning_rate": 0.0002997297297297297,
+      "loss": 0.6584,
+      "step": 16
+    },
+    {
+      "epoch": 0.005096496158890762,
+      "grad_norm": 0.32851532101631165,
+      "learning_rate": 0.00029968468468468464,
+      "loss": 0.6792,
+      "step": 17
+    },
+    {
+      "epoch": 0.005396290050590219,
+      "grad_norm": 0.3087822198867798,
+      "learning_rate": 0.00029963963963963963,
+      "loss": 0.686,
+      "step": 18
+    },
+    {
+      "epoch": 0.005696083942289676,
+      "grad_norm": 0.2965148687362671,
+      "learning_rate": 0.00029959459459459457,
+      "loss": 0.6549,
+      "step": 19
+    },
+    {
+      "epoch": 0.005995877833989132,
+      "grad_norm": 0.32235461473464966,
+      "learning_rate": 0.0002995495495495495,
+      "loss": 0.6653,
+      "step": 20
+    },
+    {
+      "epoch": 0.006295671725688589,
+      "grad_norm": 0.29656529426574707,
+      "learning_rate": 0.0002995045045045045,
+      "loss": 0.6423,
+      "step": 21
+    },
+    {
+      "epoch": 0.006595465617388046,
+      "grad_norm": 0.3357860743999481,
+      "learning_rate": 0.00029945945945945943,
+      "loss": 0.6957,
+      "step": 22
+    },
+    {
+      "epoch": 0.006895259509087502,
+      "grad_norm": 0.28682708740234375,
+      "learning_rate": 0.00029941441441441437,
+      "loss": 0.6475,
+      "step": 23
+    },
+    {
+      "epoch": 0.007195053400786959,
+      "grad_norm": 0.29643768072128296,
+      "learning_rate": 0.0002993693693693693,
+      "loss": 0.6651,
+      "step": 24
+    },
+    {
+      "epoch": 0.007494847292486416,
+      "grad_norm": 0.2794937789440155,
+      "learning_rate": 0.0002993243243243243,
+      "loss": 0.6423,
+      "step": 25
+    },
+    {
+      "epoch": 0.007794641184185872,
+      "grad_norm": 0.3023228943347931,
+      "learning_rate": 0.00029927927927927923,
+      "loss": 0.6728,
+      "step": 26
+    },
+    {
+      "epoch": 0.008094435075885328,
+      "grad_norm": 0.27937522530555725,
+      "learning_rate": 0.00029923423423423417,
+      "loss": 0.6538,
+      "step": 27
+    },
+    {
+      "epoch": 0.008394228967584785,
+      "grad_norm": 0.2845768332481384,
+      "learning_rate": 0.00029918918918918916,
+      "loss": 0.6588,
+      "step": 28
+    },
+    {
+      "epoch": 0.008694022859284242,
+      "grad_norm": 0.2637081742286682,
+      "learning_rate": 0.0002991441441441441,
+      "loss": 0.6253,
+      "step": 29
+    },
+    {
+      "epoch": 0.008993816750983699,
+      "grad_norm": 0.25597694516181946,
+      "learning_rate": 0.00029909909909909903,
+      "loss": 0.5973,
+      "step": 30
+    },
+    {
+      "epoch": 0.009293610642683156,
+      "grad_norm": 0.28932246565818787,
+      "learning_rate": 0.000299054054054054,
+      "loss": 0.6607,
+      "step": 31
+    },
+    {
+      "epoch": 0.009593404534382613,
+      "grad_norm": 0.28410351276397705,
+      "learning_rate": 0.00029900900900900896,
+      "loss": 0.6496,
+      "step": 32
+    },
+    {
+      "epoch": 0.009893198426082068,
+      "grad_norm": 0.2708141803741455,
+      "learning_rate": 0.00029896396396396395,
+      "loss": 0.6161,
+      "step": 33
+    },
+    {
+      "epoch": 0.010192992317781525,
+      "grad_norm": 0.27171120047569275,
+      "learning_rate": 0.0002989189189189189,
+      "loss": 0.6179,
+      "step": 34
+    },
+    {
+      "epoch": 0.010492786209480982,
+      "grad_norm": 0.2806681990623474,
+      "learning_rate": 0.0002988738738738738,
+      "loss": 0.6111,
+      "step": 35
+    },
+    {
+      "epoch": 0.010792580101180439,
+      "grad_norm": 0.36722084879875183,
+      "learning_rate": 0.0002988288288288288,
+      "loss": 0.5778,
+      "step": 36
+    },
+    {
+      "epoch": 0.011092373992879895,
+      "grad_norm": 0.26182547211647034,
+      "learning_rate": 0.00029878378378378375,
+      "loss": 0.5941,
+      "step": 37
+    },
+    {
+      "epoch": 0.011392167884579352,
+      "grad_norm": 0.26753902435302734,
+      "learning_rate": 0.0002987387387387387,
+      "loss": 0.6019,
+      "step": 38
+    },
+    {
+      "epoch": 0.011691961776278808,
+      "grad_norm": 0.28038090467453003,
+      "learning_rate": 0.0002986936936936937,
+      "loss": 0.6379,
+      "step": 39
+    },
+    {
+      "epoch": 0.011991755667978264,
+      "grad_norm": 0.29290881752967834,
+      "learning_rate": 0.0002986486486486486,
+      "loss": 0.6743,
+      "step": 40
+    },
+    {
+      "epoch": 0.012291549559677721,
+      "grad_norm": 0.28465205430984497,
+      "learning_rate": 0.00029860360360360356,
+      "loss": 0.6214,
+      "step": 41
+    },
+    {
+      "epoch": 0.012591343451377178,
+      "grad_norm": 0.26730677485466003,
+      "learning_rate": 0.00029855855855855855,
+      "loss": 0.6085,
+      "step": 42
+    },
+    {
+      "epoch": 0.012891137343076635,
+      "grad_norm": 0.2801668643951416,
+      "learning_rate": 0.0002985135135135135,
+      "loss": 0.6541,
+      "step": 43
+    },
+    {
+      "epoch": 0.013190931234776092,
+      "grad_norm": 0.2741893529891968,
+      "learning_rate": 0.0002984684684684684,
+      "loss": 0.6102,
+      "step": 44
+    },
+    {
+      "epoch": 0.013490725126475547,
+      "grad_norm": 0.26284873485565186,
+      "learning_rate": 0.0002984234234234234,
+      "loss": 0.5762,
+      "step": 45
+    },
+    {
+      "epoch": 0.013790519018175004,
+      "grad_norm": 0.26464149355888367,
+      "learning_rate": 0.00029837837837837835,
+      "loss": 0.6238,
+      "step": 46
+    },
+    {
+      "epoch": 0.014090312909874461,
+      "grad_norm": 0.2674684226512909,
+      "learning_rate": 0.00029833333333333334,
+      "loss": 0.6342,
+      "step": 47
+    },
+    {
+      "epoch": 0.014390106801573918,
+      "grad_norm": 0.2591281533241272,
+      "learning_rate": 0.0002982882882882883,
+      "loss": 0.5993,
+      "step": 48
+    },
+    {
+      "epoch": 0.014689900693273375,
+      "grad_norm": 0.29611843824386597,
+      "learning_rate": 0.0002982432432432432,
+      "loss": 0.6376,
+      "step": 49
+    },
+    {
+      "epoch": 0.014989694584972832,
+      "grad_norm": 0.26478683948516846,
+      "learning_rate": 0.0002981981981981982,
+      "loss": 0.5935,
+      "step": 50
+    },
+    {
+      "epoch": 0.015289488476672287,
+      "grad_norm": 0.25142884254455566,
+      "learning_rate": 0.00029815315315315314,
+      "loss": 0.5522,
+      "step": 51
+    },
+    {
+      "epoch": 0.015589282368371744,
+      "grad_norm": 0.26863306760787964,
+      "learning_rate": 0.0002981081081081081,
+      "loss": 0.5808,
+      "step": 52
+    },
+    {
+      "epoch": 0.015889076260071203,
+      "grad_norm": 0.26126888394355774,
+      "learning_rate": 0.00029806306306306307,
+      "loss": 0.6101,
+      "step": 53
+    },
+    {
+      "epoch": 0.016188870151770656,
+      "grad_norm": 0.24878369271755219,
+      "learning_rate": 0.000298018018018018,
+      "loss": 0.5787,
+      "step": 54
+    },
+    {
+      "epoch": 0.016488664043470113,
+      "grad_norm": 0.2513170838356018,
+      "learning_rate": 0.00029797297297297294,
+      "loss": 0.5743,
+      "step": 55
+    },
+    {
+      "epoch": 0.01678845793516957,
+      "grad_norm": 0.26510271430015564,
+      "learning_rate": 0.00029792792792792793,
+      "loss": 0.6023,
+      "step": 56
+    },
+    {
+      "epoch": 0.017088251826869027,
+      "grad_norm": 0.2809905409812927,
+      "learning_rate": 0.00029788288288288287,
+      "loss": 0.6336,
+      "step": 57
+    },
+    {
+      "epoch": 0.017388045718568484,
+      "grad_norm": 0.250021755695343,
+      "learning_rate": 0.0002978378378378378,
+      "loss": 0.5385,
+      "step": 58
+    },
+    {
+      "epoch": 0.01768783961026794,
+      "grad_norm": 0.248112291097641,
+      "learning_rate": 0.0002977927927927928,
+      "loss": 0.5633,
+      "step": 59
+    },
+    {
+      "epoch": 0.017987633501967398,
+      "grad_norm": 0.2756728529930115,
+      "learning_rate": 0.00029774774774774773,
+      "loss": 0.5988,
+      "step": 60
+    },
+    {
+      "epoch": 0.018287427393666855,
+      "grad_norm": 0.26302388310432434,
+      "learning_rate": 0.00029770270270270267,
+      "loss": 0.5738,
+      "step": 61
+    },
+    {
+      "epoch": 0.01858722128536631,
+      "grad_norm": 0.2736607789993286,
+      "learning_rate": 0.00029765765765765766,
+      "loss": 0.5908,
+      "step": 62
+    },
+    {
+      "epoch": 0.01888701517706577,
+      "grad_norm": 0.26531675457954407,
+      "learning_rate": 0.0002976126126126126,
+      "loss": 0.6112,
+      "step": 63
+    },
+    {
+      "epoch": 0.019186809068765225,
+      "grad_norm": 0.252244234085083,
+      "learning_rate": 0.00029756756756756753,
+      "loss": 0.5921,
+      "step": 64
+    },
+    {
+      "epoch": 0.019486602960464682,
+      "grad_norm": 0.26548662781715393,
+      "learning_rate": 0.0002975225225225225,
+      "loss": 0.604,
+      "step": 65
+    },
+    {
+      "epoch": 0.019786396852164136,
+      "grad_norm": 0.2575795352458954,
+      "learning_rate": 0.00029747747747747746,
+      "loss": 0.564,
+      "step": 66
+    },
+    {
+      "epoch": 0.020086190743863593,
+      "grad_norm": 0.25535911321640015,
+      "learning_rate": 0.0002974324324324324,
+      "loss": 0.5962,
+      "step": 67
+    },
+    {
+      "epoch": 0.02038598463556305,
+      "grad_norm": 0.26686182618141174,
+      "learning_rate": 0.0002973873873873874,
+      "loss": 0.6268,
+      "step": 68
+    },
+    {
+      "epoch": 0.020685778527262506,
+      "grad_norm": 0.25553348660469055,
+      "learning_rate": 0.00029734234234234233,
+      "loss": 0.5583,
+      "step": 69
+    },
+    {
+      "epoch": 0.020985572418961963,
+      "grad_norm": 0.2537127435207367,
+      "learning_rate": 0.00029729729729729726,
+      "loss": 0.5734,
+      "step": 70
+    },
+    {
+      "epoch": 0.02128536631066142,
+      "grad_norm": 0.27586349844932556,
+      "learning_rate": 0.0002972522522522522,
+      "loss": 0.5667,
+      "step": 71
+    },
+    {
+      "epoch": 0.021585160202360877,
+      "grad_norm": 0.2614293694496155,
+      "learning_rate": 0.0002972072072072072,
+      "loss": 0.5831,
+      "step": 72
+    },
+    {
+      "epoch": 0.021884954094060334,
+      "grad_norm": 0.263944536447525,
+      "learning_rate": 0.00029716216216216213,
+      "loss": 0.5862,
+      "step": 73
+    },
+    {
+      "epoch": 0.02218474798575979,
+      "grad_norm": 0.24802720546722412,
+      "learning_rate": 0.00029711711711711707,
+      "loss": 0.5788,
+      "step": 74
+    },
+    {
+      "epoch": 0.022484541877459248,
+      "grad_norm": 0.2746400833129883,
+      "learning_rate": 0.00029707207207207206,
+      "loss": 0.591,
+      "step": 75
+    },
+    {
+      "epoch": 0.022784335769158705,
+      "grad_norm": 0.2553657591342926,
+      "learning_rate": 0.000297027027027027,
+      "loss": 0.5735,
+      "step": 76
+    },
+    {
+      "epoch": 0.02308412966085816,
+      "grad_norm": 0.2704835534095764,
+      "learning_rate": 0.00029698198198198193,
+      "loss": 0.5809,
+      "step": 77
+    },
+    {
+      "epoch": 0.023383923552557615,
+      "grad_norm": 0.2556537985801697,
+      "learning_rate": 0.0002969369369369369,
+      "loss": 0.5496,
+      "step": 78
+    },
+    {
+      "epoch": 0.023683717444257072,
+      "grad_norm": 0.2523523271083832,
+      "learning_rate": 0.00029689189189189186,
+      "loss": 0.5758,
+      "step": 79
+    },
+    {
+      "epoch": 0.02398351133595653,
+      "grad_norm": 0.25214141607284546,
+      "learning_rate": 0.0002968468468468468,
+      "loss": 0.5844,
+      "step": 80
+    },
+    {
+      "epoch": 0.024283305227655986,
+      "grad_norm": 0.2590884268283844,
+      "learning_rate": 0.0002968018018018018,
+      "loss": 0.5862,
+      "step": 81
+    },
+    {
+      "epoch": 0.024583099119355443,
+      "grad_norm": 0.24399450421333313,
+      "learning_rate": 0.0002967567567567567,
+      "loss": 0.5685,
+      "step": 82
+    },
+    {
+      "epoch": 0.0248828930110549,
+      "grad_norm": 0.24457746744155884,
+      "learning_rate": 0.00029671171171171166,
+      "loss": 0.5684,
+      "step": 83
+    },
+    {
+      "epoch": 0.025182686902754357,
+      "grad_norm": 0.23214662075042725,
+      "learning_rate": 0.00029666666666666665,
+      "loss": 0.5682,
+      "step": 84
+    },
+    {
+      "epoch": 0.025482480794453814,
+      "grad_norm": 0.24058258533477783,
+      "learning_rate": 0.0002966216216216216,
+      "loss": 0.5688,
+      "step": 85
+    },
+    {
+      "epoch": 0.02578227468615327,
+      "grad_norm": 0.2338346391916275,
+      "learning_rate": 0.0002965765765765765,
+      "loss": 0.5485,
+      "step": 86
+    },
+    {
+      "epoch": 0.026082068577852727,
+      "grad_norm": 0.2553529143333435,
+      "learning_rate": 0.0002965315315315315,
+      "loss": 0.5985,
+      "step": 87
+    },
+    {
+      "epoch": 0.026381862469552184,
+      "grad_norm": 0.2518393099308014,
+      "learning_rate": 0.00029648648648648645,
+      "loss": 0.6079,
+      "step": 88
+    },
+    {
+      "epoch": 0.02668165636125164,
+      "grad_norm": 0.23418371379375458,
+      "learning_rate": 0.0002964414414414414,
+      "loss": 0.5942,
+      "step": 89
+    },
+    {
+      "epoch": 0.026981450252951095,
+      "grad_norm": 0.2454022616147995,
+      "learning_rate": 0.0002963963963963964,
+      "loss": 0.5519,
+      "step": 90
+    },
+    {
+      "epoch": 0.02728124414465055,
+      "grad_norm": 0.25474220514297485,
+      "learning_rate": 0.0002963513513513513,
+      "loss": 0.5771,
+      "step": 91
+    },
+    {
+      "epoch": 0.02758103803635001,
+      "grad_norm": 0.2332638055086136,
+      "learning_rate": 0.00029630630630630625,
+      "loss": 0.5477,
+      "step": 92
+    },
+    {
+      "epoch": 0.027880831928049465,
+      "grad_norm": 0.23931057751178741,
+      "learning_rate": 0.00029626126126126124,
+      "loss": 0.5391,
+      "step": 93
+    },
+    {
+      "epoch": 0.028180625819748922,
+      "grad_norm": 0.23103167116641998,
+      "learning_rate": 0.0002962162162162162,
+      "loss": 0.5622,
+      "step": 94
+    },
+    {
+      "epoch": 0.02848041971144838,
+      "grad_norm": 0.2356046438217163,
+      "learning_rate": 0.0002961711711711711,
+      "loss": 0.5729,
+      "step": 95
+    },
+    {
+      "epoch": 0.028780213603147836,
+      "grad_norm": 0.26015767455101013,
+      "learning_rate": 0.0002961261261261261,
+      "loss": 0.5878,
+      "step": 96
+    },
+    {
+      "epoch": 0.029080007494847293,
+      "grad_norm": 0.2299613654613495,
+      "learning_rate": 0.00029608108108108104,
+      "loss": 0.5385,
+      "step": 97
+    },
+    {
+      "epoch": 0.02937980138654675,
+      "grad_norm": 0.24516573548316956,
+      "learning_rate": 0.000296036036036036,
+      "loss": 0.5688,
+      "step": 98
+    },
+    {
+      "epoch": 0.029679595278246207,
+      "grad_norm": 0.25019732117652893,
+      "learning_rate": 0.00029599099099099097,
+      "loss": 0.5562,
+      "step": 99
+    },
+    {
+      "epoch": 0.029979389169945664,
+      "grad_norm": 0.23934195935726166,
+      "learning_rate": 0.0002959459459459459,
+      "loss": 0.5572,
+      "step": 100
+    },
+    {
+      "epoch": 0.030279183061645117,
+      "grad_norm": 0.23651225864887238,
+      "learning_rate": 0.00029590090090090085,
+      "loss": 0.5517,
+      "step": 101
+    },
+    {
+      "epoch": 0.030578976953344574,
+      "grad_norm": 0.2545163631439209,
+      "learning_rate": 0.00029585585585585584,
+      "loss": 0.5553,
+      "step": 102
+    },
+    {
+      "epoch": 0.03087877084504403,
+      "grad_norm": 0.2259800285100937,
+      "learning_rate": 0.0002958108108108108,
+      "loss": 0.5216,
+      "step": 103
+    },
+    {
+      "epoch": 0.031178564736743488,
+      "grad_norm": 0.25137075781822205,
+      "learning_rate": 0.00029576576576576576,
+      "loss": 0.559,
+      "step": 104
+    },
+    {
+      "epoch": 0.03147835862844295,
+      "grad_norm": 0.2526615262031555,
+      "learning_rate": 0.0002957207207207207,
+      "loss": 0.6025,
+      "step": 105
+    },
+    {
+      "epoch": 0.031778152520142405,
+      "grad_norm": 0.2406662553548813,
+      "learning_rate": 0.00029567567567567564,
+      "loss": 0.562,
+      "step": 106
+    },
+    {
+      "epoch": 0.032077946411841855,
+      "grad_norm": 0.23598824441432953,
+      "learning_rate": 0.00029563063063063063,
+      "loss": 0.5749,
+      "step": 107
+    },
+    {
+      "epoch": 0.03237774030354131,
+      "grad_norm": 0.24079738557338715,
+      "learning_rate": 0.00029558558558558557,
+      "loss": 0.5697,
+      "step": 108
+    },
+    {
+      "epoch": 0.03267753419524077,
+      "grad_norm": 0.25064727663993835,
+      "learning_rate": 0.0002955405405405405,
+      "loss": 0.5821,
+      "step": 109
+    },
+    {
+      "epoch": 0.032977328086940226,
+      "grad_norm": 0.22557033598423004,
+      "learning_rate": 0.0002954954954954955,
+      "loss": 0.5367,
+      "step": 110
+    },
+    {
+      "epoch": 0.03327712197863968,
+      "grad_norm": 0.22542916238307953,
+      "learning_rate": 0.00029545045045045043,
+      "loss": 0.55,
+      "step": 111
+    },
+    {
+      "epoch": 0.03357691587033914,
+      "grad_norm": 0.24270494282245636,
+      "learning_rate": 0.00029540540540540537,
+      "loss": 0.5773,
+      "step": 112
+    },
+    {
+      "epoch": 0.0338767097620386,
+      "grad_norm": 0.23397719860076904,
+      "learning_rate": 0.00029536036036036036,
+      "loss": 0.5516,
+      "step": 113
+    },
+    {
+      "epoch": 0.034176503653738054,
+      "grad_norm": 0.2322990894317627,
+      "learning_rate": 0.0002953153153153153,
+      "loss": 0.5307,
+      "step": 114
+    },
+    {
+      "epoch": 0.03447629754543751,
+      "grad_norm": 0.2167540043592453,
+      "learning_rate": 0.0002952702702702703,
+      "loss": 0.5197,
+      "step": 115
+    },
+    {
+      "epoch": 0.03477609143713697,
+      "grad_norm": 0.2705434262752533,
+      "learning_rate": 0.0002952252252252252,
+      "loss": 0.564,
+      "step": 116
+    },
+    {
+      "epoch": 0.035075885328836424,
+      "grad_norm": 0.23195774853229523,
+      "learning_rate": 0.00029518018018018016,
+      "loss": 0.5397,
+      "step": 117
+    },
+    {
+      "epoch": 0.03537567922053588,
+      "grad_norm": 0.22944559156894684,
+      "learning_rate": 0.00029513513513513515,
+      "loss": 0.5197,
+      "step": 118
+    },
+    {
+      "epoch": 0.03567547311223534,
+      "grad_norm": 0.25140368938446045,
+      "learning_rate": 0.0002950900900900901,
+      "loss": 0.5661,
+      "step": 119
+    },
+    {
+      "epoch": 0.035975267003934795,
+      "grad_norm": 0.23137278854846954,
+      "learning_rate": 0.000295045045045045,
+      "loss": 0.5284,
+      "step": 120
+    },
+    {
+      "epoch": 0.03627506089563425,
+      "grad_norm": 0.2358487993478775,
+      "learning_rate": 0.00029499999999999996,
+      "loss": 0.551,
+      "step": 121
+    },
+    {
+      "epoch": 0.03657485478733371,
+      "grad_norm": 0.23212139308452606,
+      "learning_rate": 0.00029495495495495495,
+      "loss": 0.5624,
+      "step": 122
+    },
+    {
+      "epoch": 0.036874648679033166,
+      "grad_norm": 0.2504674196243286,
+      "learning_rate": 0.0002949099099099099,
+      "loss": 0.5649,
+      "step": 123
+    },
+    {
+      "epoch": 0.03717444257073262,
+      "grad_norm": 0.24574224650859833,
+      "learning_rate": 0.0002948648648648648,
+      "loss": 0.5415,
+      "step": 124
+    },
+    {
+      "epoch": 0.03747423646243208,
+      "grad_norm": 0.23007921874523163,
+      "learning_rate": 0.0002948198198198198,
+      "loss": 0.5624,
+      "step": 125
+    },
+    {
+      "epoch": 0.03777403035413154,
+      "grad_norm": 0.2503686547279358,
+      "learning_rate": 0.00029477477477477475,
+      "loss": 0.5484,
+      "step": 126
+    },
+    {
+      "epoch": 0.038073824245830994,
+      "grad_norm": 0.2467029094696045,
+      "learning_rate": 0.0002947297297297297,
+      "loss": 0.5442,
+      "step": 127
+    },
+    {
+      "epoch": 0.03837361813753045,
+      "grad_norm": 0.23382776975631714,
+      "learning_rate": 0.0002946846846846847,
+      "loss": 0.5485,
+      "step": 128
+    },
+    {
+      "epoch": 0.03867341202922991,
+      "grad_norm": 0.25080442428588867,
+      "learning_rate": 0.0002946396396396396,
+      "loss": 0.5726,
+      "step": 129
+    },
+    {
+      "epoch": 0.038973205920929364,
+      "grad_norm": 0.2410043627023697,
+      "learning_rate": 0.00029459459459459455,
+      "loss": 0.5475,
+      "step": 130
+    },
+    {
+      "epoch": 0.039272999812628814,
+      "grad_norm": 0.24639956653118134,
+      "learning_rate": 0.00029454954954954955,
+      "loss": 0.5461,
+      "step": 131
+    },
+    {
+      "epoch": 0.03957279370432827,
+      "grad_norm": 0.23122674226760864,
+      "learning_rate": 0.0002945045045045045,
+      "loss": 0.5605,
+      "step": 132
+    },
+    {
+      "epoch": 0.03987258759602773,
+      "grad_norm": 0.22207331657409668,
+      "learning_rate": 0.0002944594594594594,
+      "loss": 0.5045,
+      "step": 133
+    },
+    {
+      "epoch": 0.040172381487727185,
+      "grad_norm": 0.2344479262828827,
+      "learning_rate": 0.0002944144144144144,
+      "loss": 0.5724,
+      "step": 134
+    },
+    {
+      "epoch": 0.04047217537942664,
+      "grad_norm": 0.2410653978586197,
+      "learning_rate": 0.00029436936936936935,
+      "loss": 0.5266,
+      "step": 135
+    },
+    {
+      "epoch": 0.0407719692711261,
+      "grad_norm": 0.23028722405433655,
+      "learning_rate": 0.0002943243243243243,
+      "loss": 0.545,
+      "step": 136
+    },
+    {
+      "epoch": 0.041071763162825556,
+      "grad_norm": 0.23669356107711792,
+      "learning_rate": 0.0002942792792792793,
+      "loss": 0.5429,
+      "step": 137
+    },
+    {
+      "epoch": 0.04137155705452501,
+      "grad_norm": 0.24232889711856842,
+      "learning_rate": 0.0002942342342342342,
+      "loss": 0.5815,
+      "step": 138
+    },
+    {
+      "epoch": 0.04167135094622447,
+      "grad_norm": 0.22076158225536346,
+      "learning_rate": 0.00029418918918918915,
+      "loss": 0.5028,
+      "step": 139
+    },
+    {
+      "epoch": 0.04197114483792393,
+      "grad_norm": 0.23322072625160217,
+      "learning_rate": 0.00029414414414414414,
+      "loss": 0.5293,
+      "step": 140
+    },
+    {
+      "epoch": 0.042270938729623384,
+      "grad_norm": 0.24910598993301392,
+      "learning_rate": 0.0002940990990990991,
+      "loss": 0.5403,
+      "step": 141
+    },
+    {
+      "epoch": 0.04257073262132284,
+      "grad_norm": 0.24588559567928314,
+      "learning_rate": 0.000294054054054054,
+      "loss": 0.538,
+      "step": 142
+    },
+    {
+      "epoch": 0.0428705265130223,
+      "grad_norm": 0.20689445734024048,
+      "learning_rate": 0.00029400900900900895,
+      "loss": 0.4921,
+      "step": 143
+    },
+    {
+      "epoch": 0.043170320404721754,
+      "grad_norm": 0.2518969178199768,
+      "learning_rate": 0.00029396396396396394,
+      "loss": 0.5625,
+      "step": 144
+    },
+    {
+      "epoch": 0.04347011429642121,
+      "grad_norm": 0.25027644634246826,
+      "learning_rate": 0.0002939189189189189,
+      "loss": 0.5823,
+      "step": 145
+    },
+    {
+      "epoch": 0.04376990818812067,
+      "grad_norm": 0.24390611052513123,
+      "learning_rate": 0.0002938738738738738,
+      "loss": 0.5205,
+      "step": 146
+    },
+    {
+      "epoch": 0.044069702079820125,
+      "grad_norm": 0.24781368672847748,
+      "learning_rate": 0.0002938288288288288,
+      "loss": 0.5559,
+      "step": 147
+    },
+    {
+      "epoch": 0.04436949597151958,
+      "grad_norm": 0.23696455359458923,
+      "learning_rate": 0.00029378378378378374,
+      "loss": 0.5443,
+      "step": 148
+    },
+    {
+      "epoch": 0.04466928986321904,
+      "grad_norm": 0.23888202011585236,
+      "learning_rate": 0.0002937387387387387,
+      "loss": 0.5397,
+      "step": 149
+    },
+    {
+      "epoch": 0.044969083754918496,
+      "grad_norm": 0.22896665334701538,
+      "learning_rate": 0.00029369369369369367,
+      "loss": 0.5462,
+      "step": 150
+    },
+    {
+      "epoch": 0.04526887764661795,
+      "grad_norm": 0.23967792093753815,
+      "learning_rate": 0.0002936486486486486,
+      "loss": 0.5653,
+      "step": 151
+    },
+    {
+      "epoch": 0.04556867153831741,
+      "grad_norm": 0.2244715392589569,
+      "learning_rate": 0.00029360360360360354,
+      "loss": 0.5251,
+      "step": 152
+    },
+    {
+      "epoch": 0.045868465430016866,
+      "grad_norm": 0.2287164330482483,
+      "learning_rate": 0.00029355855855855853,
+      "loss": 0.5313,
+      "step": 153
+    },
+    {
+      "epoch": 0.04616825932171632,
+      "grad_norm": 0.2419881671667099,
+      "learning_rate": 0.00029351351351351347,
+      "loss": 0.5397,
+      "step": 154
+    },
+    {
+      "epoch": 0.04646805321341577,
+      "grad_norm": 0.2414129674434662,
+      "learning_rate": 0.0002934684684684684,
+      "loss": 0.574,
+      "step": 155
+    },
+    {
+      "epoch": 0.04676784710511523,
+      "grad_norm": 0.237702414393425,
+      "learning_rate": 0.0002934234234234234,
+      "loss": 0.5377,
+      "step": 156
+    },
+    {
+      "epoch": 0.04706764099681469,
+      "grad_norm": 0.24712331593036652,
+      "learning_rate": 0.00029337837837837833,
+      "loss": 0.5722,
+      "step": 157
+    },
+    {
+      "epoch": 0.047367434888514144,
+      "grad_norm": 0.23250596225261688,
+      "learning_rate": 0.00029333333333333327,
+      "loss": 0.5663,
+      "step": 158
+    },
+    {
+      "epoch": 0.0476672287802136,
+      "grad_norm": 0.22616314888000488,
+      "learning_rate": 0.00029328828828828826,
+      "loss": 0.4984,
+      "step": 159
+    },
+    {
+      "epoch": 0.04796702267191306,
+      "grad_norm": 0.23835696280002594,
+      "learning_rate": 0.0002932432432432432,
+      "loss": 0.5512,
+      "step": 160
+    },
+    {
+      "epoch": 0.048266816563612515,
+      "grad_norm": 0.23458854854106903,
+      "learning_rate": 0.0002931981981981982,
+      "loss": 0.5472,
+      "step": 161
+    },
+    {
+      "epoch": 0.04856661045531197,
+      "grad_norm": 0.2530427873134613,
+      "learning_rate": 0.0002931531531531531,
+      "loss": 0.5807,
+      "step": 162
+    },
+    {
+      "epoch": 0.04886640434701143,
+      "grad_norm": 0.23911581933498383,
+      "learning_rate": 0.00029310810810810806,
+      "loss": 0.5236,
+      "step": 163
+    },
+    {
+      "epoch": 0.049166198238710886,
+      "grad_norm": 0.24235881865024567,
+      "learning_rate": 0.00029306306306306305,
+      "loss": 0.5509,
+      "step": 164
+    },
+    {
+      "epoch": 0.04946599213041034,
+      "grad_norm": 0.23077982664108276,
+      "learning_rate": 0.000293018018018018,
+      "loss": 0.5265,
+      "step": 165
+    },
+    {
+      "epoch": 0.0497657860221098,
+      "grad_norm": 0.23159649968147278,
+      "learning_rate": 0.00029297297297297293,
+      "loss": 0.5521,
+      "step": 166
+    },
+    {
+      "epoch": 0.050065579913809256,
+      "grad_norm": 0.23901638388633728,
+      "learning_rate": 0.0002929279279279279,
+      "loss": 0.517,
+      "step": 167
+    },
+    {
+      "epoch": 0.05036537380550871,
+      "grad_norm": 0.2441125512123108,
+      "learning_rate": 0.00029288288288288286,
+      "loss": 0.5376,
+      "step": 168
+    },
+    {
+      "epoch": 0.05066516769720817,
+      "grad_norm": 0.2408357411623001,
+      "learning_rate": 0.0002928378378378378,
+      "loss": 0.5119,
+      "step": 169
+    },
+    {
+      "epoch": 0.05096496158890763,
+      "grad_norm": 0.24620820581912994,
+      "learning_rate": 0.0002927927927927928,
+      "loss": 0.5454,
+      "step": 170
+    },
+    {
+      "epoch": 0.051264755480607084,
+      "grad_norm": 0.23092059791088104,
+      "learning_rate": 0.0002927477477477477,
+      "loss": 0.5464,
+      "step": 171
+    },
+    {
+      "epoch": 0.05156454937230654,
+      "grad_norm": 0.23076176643371582,
+      "learning_rate": 0.0002927027027027027,
+      "loss": 0.5153,
+      "step": 172
+    },
+    {
+      "epoch": 0.051864343264006,
+      "grad_norm": 0.2518848478794098,
+      "learning_rate": 0.00029265765765765765,
+      "loss": 0.5416,
+      "step": 173
+    },
+    {
+      "epoch": 0.052164137155705455,
+      "grad_norm": 0.24004964530467987,
+      "learning_rate": 0.0002926126126126126,
+      "loss": 0.5625,
+      "step": 174
+    },
+    {
+      "epoch": 0.05246393104740491,
+      "grad_norm": 0.26169759035110474,
+      "learning_rate": 0.0002925675675675676,
+      "loss": 0.5322,
+      "step": 175
+    },
+    {
+      "epoch": 0.05276372493910437,
+      "grad_norm": 0.23733891546726227,
+      "learning_rate": 0.0002925225225225225,
+      "loss": 0.5526,
+      "step": 176
+    },
+    {
+      "epoch": 0.053063518830803826,
+      "grad_norm": 0.22720226645469666,
+      "learning_rate": 0.00029247747747747745,
+      "loss": 0.5108,
+      "step": 177
+    },
+    {
+      "epoch": 0.05336331272250328,
+      "grad_norm": 0.2425220012664795,
+      "learning_rate": 0.00029243243243243244,
+      "loss": 0.5337,
+      "step": 178
+    },
+    {
+      "epoch": 0.05366310661420273,
+      "grad_norm": 0.24859829246997833,
+      "learning_rate": 0.0002923873873873874,
+      "loss": 0.5239,
+      "step": 179
+    },
+    {
+      "epoch": 0.05396290050590219,
+      "grad_norm": 0.23471564054489136,
+      "learning_rate": 0.0002923423423423423,
+      "loss": 0.5372,
+      "step": 180
+    },
+    {
+      "epoch": 0.054262694397601646,
+      "grad_norm": 0.23243340849876404,
+      "learning_rate": 0.0002922972972972973,
+      "loss": 0.5283,
+      "step": 181
+    },
+    {
+      "epoch": 0.0545624882893011,
+      "grad_norm": 0.240774467587471,
+      "learning_rate": 0.00029225225225225224,
+      "loss": 0.523,
+      "step": 182
+    },
+    {
+      "epoch": 0.05486228218100056,
+      "grad_norm": 0.2518199384212494,
+      "learning_rate": 0.0002922072072072072,
+      "loss": 0.56,
+      "step": 183
+    },
+    {
+      "epoch": 0.05516207607270002,
+      "grad_norm": 0.22320473194122314,
+      "learning_rate": 0.00029216216216216217,
+      "loss": 0.5442,
+      "step": 184
+    },
+    {
+      "epoch": 0.055461869964399474,
+      "grad_norm": 0.24681055545806885,
+      "learning_rate": 0.0002921171171171171,
+      "loss": 0.5596,
+      "step": 185
+    },
+    {
+      "epoch": 0.05576166385609893,
+      "grad_norm": 0.23595312237739563,
+      "learning_rate": 0.00029207207207207204,
+      "loss": 0.5343,
+      "step": 186
+    },
+    {
+      "epoch": 0.05606145774779839,
+      "grad_norm": 0.2307872772216797,
+      "learning_rate": 0.00029202702702702703,
+      "loss": 0.5206,
+      "step": 187
+    },
+    {
+      "epoch": 0.056361251639497845,
+      "grad_norm": 0.24639058113098145,
+      "learning_rate": 0.00029198198198198197,
+      "loss": 0.5449,
+      "step": 188
+    },
+    {
+      "epoch": 0.0566610455311973,
+      "grad_norm": 0.24432207643985748,
+      "learning_rate": 0.0002919369369369369,
+      "loss": 0.5455,
+      "step": 189
+    },
+    {
+      "epoch": 0.05696083942289676,
+      "grad_norm": 0.23387478291988373,
+      "learning_rate": 0.0002918918918918919,
+      "loss": 0.5402,
+      "step": 190
+    },
+    {
+      "epoch": 0.057260633314596215,
+      "grad_norm": 0.2183016538619995,
+      "learning_rate": 0.00029184684684684684,
+      "loss": 0.4989,
+      "step": 191
+    },
+    {
+      "epoch": 0.05756042720629567,
+      "grad_norm": 0.23916368186473846,
+      "learning_rate": 0.00029180180180180177,
+      "loss": 0.5397,
+      "step": 192
+    },
+    {
+      "epoch": 0.05786022109799513,
+      "grad_norm": 0.24624724686145782,
+      "learning_rate": 0.0002917567567567567,
+      "loss": 0.5248,
+      "step": 193
+    },
+    {
+      "epoch": 0.058160014989694586,
+      "grad_norm": 0.2571977376937866,
+      "learning_rate": 0.0002917117117117117,
+      "loss": 0.587,
+      "step": 194
+    },
+    {
+      "epoch": 0.05845980888139404,
+      "grad_norm": 0.22828777134418488,
+      "learning_rate": 0.00029166666666666664,
+      "loss": 0.5424,
+      "step": 195
+    },
+    {
+      "epoch": 0.0587596027730935,
+      "grad_norm": 0.23892144858837128,
+      "learning_rate": 0.0002916216216216216,
+      "loss": 0.5228,
+      "step": 196
+    },
+    {
+      "epoch": 0.05905939666479296,
+      "grad_norm": 0.22904744744300842,
+      "learning_rate": 0.00029157657657657656,
+      "loss": 0.5052,
+      "step": 197
+    },
+    {
+      "epoch": 0.059359190556492414,
+      "grad_norm": 0.2399347573518753,
+      "learning_rate": 0.0002915315315315315,
+      "loss": 0.541,
+      "step": 198
+    },
+    {
+      "epoch": 0.05965898444819187,
+      "grad_norm": 0.2253408432006836,
+      "learning_rate": 0.00029148648648648644,
+      "loss": 0.5204,
+      "step": 199
+    },
+    {
+      "epoch": 0.05995877833989133,
+      "grad_norm": 0.23211434483528137,
+      "learning_rate": 0.00029144144144144143,
+      "loss": 0.5248,
+      "step": 200
+    },
+    {
+      "epoch": 0.060258572231590785,
+      "grad_norm": 0.24016128480434418,
+      "learning_rate": 0.00029139639639639637,
+      "loss": 0.5386,
+      "step": 201
+    },
+    {
+      "epoch": 0.060558366123290235,
+      "grad_norm": 0.2374415397644043,
+      "learning_rate": 0.0002913513513513513,
+      "loss": 0.4904,
+      "step": 202
+    },
+    {
+      "epoch": 0.06085816001498969,
+      "grad_norm": 0.257735013961792,
+      "learning_rate": 0.0002913063063063063,
+      "loss": 0.5253,
+      "step": 203
+    },
+    {
+      "epoch": 0.06115795390668915,
+      "grad_norm": 0.23416665196418762,
+      "learning_rate": 0.00029126126126126123,
+      "loss": 0.5164,
+      "step": 204
+    },
+    {
+      "epoch": 0.061457747798388605,
+      "grad_norm": 0.24382436275482178,
+      "learning_rate": 0.00029121621621621617,
+      "loss": 0.5507,
+      "step": 205
+    },
+    {
+      "epoch": 0.06175754169008806,
+      "grad_norm": 0.2488170564174652,
+      "learning_rate": 0.00029117117117117116,
+      "loss": 0.543,
+      "step": 206
+    },
+    {
+      "epoch": 0.06205733558178752,
+      "grad_norm": 0.24262337386608124,
+      "learning_rate": 0.0002911261261261261,
+      "loss": 0.5533,
+      "step": 207
+    },
+    {
+      "epoch": 0.062357129473486976,
+      "grad_norm": 0.2239450216293335,
+      "learning_rate": 0.00029108108108108103,
+      "loss": 0.4955,
+      "step": 208
+    },
+    {
+      "epoch": 0.06265692336518644,
+      "grad_norm": 0.22644869983196259,
+      "learning_rate": 0.000291036036036036,
+      "loss": 0.5362,
+      "step": 209
+    },
+    {
+      "epoch": 0.0629567172568859,
+      "grad_norm": 0.23194913566112518,
+      "learning_rate": 0.00029099099099099096,
+      "loss": 0.5226,
+      "step": 210
+    },
+    {
+      "epoch": 0.06325651114858535,
+      "grad_norm": 0.22114375233650208,
+      "learning_rate": 0.0002909459459459459,
+      "loss": 0.472,
+      "step": 211
+    },
+    {
+      "epoch": 0.06355630504028481,
+      "grad_norm": 0.24242356419563293,
+      "learning_rate": 0.0002909009009009009,
+      "loss": 0.5041,
+      "step": 212
+    },
+    {
+      "epoch": 0.06385609893198427,
+      "grad_norm": 0.24195095896720886,
+      "learning_rate": 0.0002908558558558558,
+      "loss": 0.511,
+      "step": 213
+    },
+    {
+      "epoch": 0.06415589282368371,
+      "grad_norm": 0.23720628023147583,
+      "learning_rate": 0.00029081081081081076,
+      "loss": 0.5257,
+      "step": 214
+    },
+    {
+      "epoch": 0.06445568671538317,
+      "grad_norm": 0.22656656801700592,
+      "learning_rate": 0.00029076576576576575,
+      "loss": 0.5336,
+      "step": 215
+    },
+    {
+      "epoch": 0.06475548060708262,
+      "grad_norm": 0.23326708376407623,
+      "learning_rate": 0.0002907207207207207,
+      "loss": 0.5184,
+      "step": 216
+    },
+    {
+      "epoch": 0.06505527449878208,
+      "grad_norm": 0.2327839583158493,
+      "learning_rate": 0.0002906756756756756,
+      "loss": 0.5434,
+      "step": 217
+    },
+    {
+      "epoch": 0.06535506839048154,
+      "grad_norm": 0.24181734025478363,
+      "learning_rate": 0.0002906306306306306,
+      "loss": 0.534,
+      "step": 218
+    },
+    {
+      "epoch": 0.065654862282181,
+      "grad_norm": 0.23325061798095703,
+      "learning_rate": 0.00029058558558558555,
+      "loss": 0.5438,
+      "step": 219
+    },
+    {
+      "epoch": 0.06595465617388045,
+      "grad_norm": 0.2559988498687744,
+      "learning_rate": 0.0002905405405405405,
+      "loss": 0.5229,
+      "step": 220
+    },
+    {
+      "epoch": 0.06625445006557991,
+      "grad_norm": 0.2432825118303299,
+      "learning_rate": 0.0002904954954954955,
+      "loss": 0.5223,
+      "step": 221
+    },
+    {
+      "epoch": 0.06655424395727937,
+      "grad_norm": 0.23795264959335327,
+      "learning_rate": 0.0002904504504504504,
+      "loss": 0.5199,
+      "step": 222
+    },
+    {
+      "epoch": 0.06685403784897882,
+      "grad_norm": 0.22999484837055206,
+      "learning_rate": 0.00029040540540540535,
+      "loss": 0.506,
+      "step": 223
+    },
+    {
+      "epoch": 0.06715383174067828,
+      "grad_norm": 0.22168515622615814,
+      "learning_rate": 0.00029036036036036034,
+      "loss": 0.4873,
+      "step": 224
+    },
+    {
+      "epoch": 0.06745362563237774,
+      "grad_norm": 0.23448102176189423,
+      "learning_rate": 0.0002903153153153153,
+      "loss": 0.5533,
+      "step": 225
+    },
+    {
+      "epoch": 0.0677534195240772,
+      "grad_norm": 0.23822058737277985,
+      "learning_rate": 0.0002902702702702702,
+      "loss": 0.5221,
+      "step": 226
+    },
+    {
+      "epoch": 0.06805321341577665,
+      "grad_norm": 0.22952431440353394,
+      "learning_rate": 0.0002902252252252252,
+      "loss": 0.5402,
+      "step": 227
+    },
+    {
+      "epoch": 0.06835300730747611,
+      "grad_norm": 0.2197679728269577,
+      "learning_rate": 0.00029018018018018015,
+      "loss": 0.5115,
+      "step": 228
+    },
+    {
+      "epoch": 0.06865280119917556,
+      "grad_norm": 0.23495204746723175,
+      "learning_rate": 0.00029013513513513514,
+      "loss": 0.4963,
+      "step": 229
+    },
+    {
+      "epoch": 0.06895259509087502,
+      "grad_norm": 0.2233276218175888,
+      "learning_rate": 0.0002900900900900901,
+      "loss": 0.5257,
+      "step": 230
+    },
+    {
+      "epoch": 0.06925238898257448,
+      "grad_norm": 0.23294100165367126,
+      "learning_rate": 0.000290045045045045,
+      "loss": 0.5231,
+      "step": 231
+    },
+    {
+      "epoch": 0.06955218287427394,
+      "grad_norm": 0.24349254369735718,
+      "learning_rate": 0.00029,
+      "loss": 0.5198,
+      "step": 232
+    },
+    {
+      "epoch": 0.06985197676597339,
+      "grad_norm": 0.22396647930145264,
+      "learning_rate": 0.00028995495495495494,
+      "loss": 0.5032,
+      "step": 233
+    },
+    {
+      "epoch": 0.07015177065767285,
+      "grad_norm": 0.23737500607967377,
+      "learning_rate": 0.0002899099099099099,
+      "loss": 0.5071,
+      "step": 234
+    },
+    {
+      "epoch": 0.0704515645493723,
+      "grad_norm": 0.22798973321914673,
+      "learning_rate": 0.00028986486486486487,
+      "loss": 0.4844,
+      "step": 235
+    },
+    {
+      "epoch": 0.07075135844107176,
+      "grad_norm": 0.24095383286476135,
+      "learning_rate": 0.0002898198198198198,
+      "loss": 0.5351,
+      "step": 236
+    },
+    {
+      "epoch": 0.07105115233277122,
+      "grad_norm": 0.23701000213623047,
+      "learning_rate": 0.00028977477477477474,
+      "loss": 0.5051,
+      "step": 237
+    },
+    {
+      "epoch": 0.07135094622447068,
+      "grad_norm": 0.23588530719280243,
+      "learning_rate": 0.00028972972972972973,
+      "loss": 0.4992,
+      "step": 238
+    },
+    {
+      "epoch": 0.07165074011617013,
+      "grad_norm": 0.27736473083496094,
+      "learning_rate": 0.00028968468468468467,
+      "loss": 0.5399,
+      "step": 239
+    },
+    {
+      "epoch": 0.07195053400786959,
+      "grad_norm": 0.2486957311630249,
+      "learning_rate": 0.0002896396396396396,
+      "loss": 0.543,
+      "step": 240
+    },
+    {
+      "epoch": 0.07225032789956905,
+      "grad_norm": 0.23474164307117462,
+      "learning_rate": 0.0002895945945945946,
+      "loss": 0.5275,
+      "step": 241
+    },
+    {
+      "epoch": 0.0725501217912685,
+      "grad_norm": 0.26560667157173157,
+      "learning_rate": 0.00028954954954954953,
+      "loss": 0.5576,
+      "step": 242
+    },
+    {
+      "epoch": 0.07284991568296796,
+      "grad_norm": 0.2260173261165619,
+      "learning_rate": 0.00028950450450450447,
+      "loss": 0.5046,
+      "step": 243
+    },
+    {
+      "epoch": 0.07314970957466742,
+      "grad_norm": 0.24497725069522858,
+      "learning_rate": 0.00028945945945945946,
+      "loss": 0.5093,
+      "step": 244
+    },
+    {
+      "epoch": 0.07344950346636687,
+      "grad_norm": 0.24871525168418884,
+      "learning_rate": 0.0002894144144144144,
+      "loss": 0.5353,
+      "step": 245
+    },
+    {
+      "epoch": 0.07374929735806633,
+      "grad_norm": 0.2592950761318207,
+      "learning_rate": 0.00028936936936936933,
+      "loss": 0.553,
+      "step": 246
+    },
+    {
+      "epoch": 0.07404909124976579,
+      "grad_norm": 0.23617064952850342,
+      "learning_rate": 0.0002893243243243243,
+      "loss": 0.5276,
+      "step": 247
+    },
+    {
+      "epoch": 0.07434888514146525,
+      "grad_norm": 0.23108692467212677,
+      "learning_rate": 0.00028927927927927926,
+      "loss": 0.5027,
+      "step": 248
+    },
+    {
+      "epoch": 0.0746486790331647,
+      "grad_norm": 0.2418566793203354,
+      "learning_rate": 0.0002892342342342342,
+      "loss": 0.5142,
+      "step": 249
+    },
+    {
+      "epoch": 0.07494847292486416,
+      "grad_norm": 0.2367243766784668,
+      "learning_rate": 0.0002891891891891892,
+      "loss": 0.5062,
+      "step": 250
+    },
+    {
+      "epoch": 0.07524826681656362,
+      "grad_norm": 0.2405940294265747,
+      "learning_rate": 0.0002891441441441441,
+      "loss": 0.4964,
+      "step": 251
+    },
+    {
+      "epoch": 0.07554806070826307,
+      "grad_norm": 0.22956568002700806,
+      "learning_rate": 0.00028909909909909906,
+      "loss": 0.5081,
+      "step": 252
+    },
+    {
+      "epoch": 0.07584785459996253,
+      "grad_norm": 0.24594880640506744,
+      "learning_rate": 0.00028905405405405405,
+      "loss": 0.5522,
+      "step": 253
+    },
+    {
+      "epoch": 0.07614764849166199,
+      "grad_norm": 0.25554656982421875,
+      "learning_rate": 0.000289009009009009,
+      "loss": 0.5579,
+      "step": 254
+    },
+    {
+      "epoch": 0.07644744238336144,
+      "grad_norm": 0.24296589195728302,
+      "learning_rate": 0.0002889639639639639,
+      "loss": 0.515,
+      "step": 255
+    },
+    {
+      "epoch": 0.0767472362750609,
+      "grad_norm": 0.2128724306821823,
+      "learning_rate": 0.0002889189189189189,
+      "loss": 0.5046,
+      "step": 256
+    },
+    {
+      "epoch": 0.07704703016676036,
+      "grad_norm": 0.2304687201976776,
+      "learning_rate": 0.00028887387387387385,
+      "loss": 0.5275,
+      "step": 257
+    },
+    {
+      "epoch": 0.07734682405845981,
+      "grad_norm": 0.23740506172180176,
+      "learning_rate": 0.0002888288288288288,
+      "loss": 0.5308,
+      "step": 258
+    },
+    {
+      "epoch": 0.07764661795015927,
+      "grad_norm": 0.2270033061504364,
+      "learning_rate": 0.0002887837837837838,
+      "loss": 0.5232,
+      "step": 259
+    },
+    {
+      "epoch": 0.07794641184185873,
+      "grad_norm": 0.24655793607234955,
+      "learning_rate": 0.0002887387387387387,
+      "loss": 0.5087,
+      "step": 260
+    },
+    {
+      "epoch": 0.07824620573355819,
+      "grad_norm": 0.25199684500694275,
+      "learning_rate": 0.00028869369369369366,
+      "loss": 0.5616,
+      "step": 261
+    },
+    {
+      "epoch": 0.07854599962525763,
+      "grad_norm": 0.2466064840555191,
+      "learning_rate": 0.00028864864864864865,
+      "loss": 0.5343,
+      "step": 262
+    },
+    {
+      "epoch": 0.07884579351695709,
+      "grad_norm": 0.2401546686887741,
+      "learning_rate": 0.0002886036036036036,
+      "loss": 0.5326,
+      "step": 263
+    },
+    {
+      "epoch": 0.07914558740865654,
+      "grad_norm": 0.2478310465812683,
+      "learning_rate": 0.0002885585585585585,
+      "loss": 0.5377,
+      "step": 264
+    },
+    {
+      "epoch": 0.079445381300356,
+      "grad_norm": 0.25078094005584717,
+      "learning_rate": 0.00028851351351351346,
+      "loss": 0.5065,
+      "step": 265
+    },
+    {
+      "epoch": 0.07974517519205546,
+      "grad_norm": 0.24395832419395447,
+      "learning_rate": 0.00028846846846846845,
+      "loss": 0.5066,
+      "step": 266
+    },
+    {
+      "epoch": 0.08004496908375491,
+      "grad_norm": 0.23809655010700226,
+      "learning_rate": 0.0002884234234234234,
+      "loss": 0.5032,
+      "step": 267
+    },
+    {
+      "epoch": 0.08034476297545437,
+      "grad_norm": 0.25203442573547363,
+      "learning_rate": 0.0002883783783783783,
+      "loss": 0.5539,
+      "step": 268
+    },
+    {
+      "epoch": 0.08064455686715383,
+      "grad_norm": 0.22731390595436096,
+      "learning_rate": 0.0002883333333333333,
+      "loss": 0.5056,
+      "step": 269
+    },
+    {
+      "epoch": 0.08094435075885328,
+      "grad_norm": 0.26038557291030884,
+      "learning_rate": 0.00028828828828828825,
+      "loss": 0.5554,
+      "step": 270
+    },
+    {
+      "epoch": 0.08124414465055274,
+      "grad_norm": 0.22735048830509186,
+      "learning_rate": 0.0002882432432432432,
+      "loss": 0.513,
+      "step": 271
+    },
+    {
+      "epoch": 0.0815439385422522,
+      "grad_norm": 0.2459019273519516,
+      "learning_rate": 0.0002881981981981982,
+      "loss": 0.5074,
+      "step": 272
+    },
+    {
+      "epoch": 0.08184373243395165,
+      "grad_norm": 0.23949427902698517,
+      "learning_rate": 0.0002881531531531531,
+      "loss": 0.5289,
+      "step": 273
+    },
+    {
+      "epoch": 0.08214352632565111,
+      "grad_norm": 0.22845999896526337,
+      "learning_rate": 0.00028810810810810805,
+      "loss": 0.5046,
+      "step": 274
+    },
+    {
+      "epoch": 0.08244332021735057,
+      "grad_norm": 0.23616977035999298,
+      "learning_rate": 0.00028806306306306304,
+      "loss": 0.4766,
+      "step": 275
+    },
+    {
+      "epoch": 0.08274311410905003,
+      "grad_norm": 0.24239955842494965,
+      "learning_rate": 0.000288018018018018,
+      "loss": 0.5177,
+      "step": 276
+    },
+    {
+      "epoch": 0.08304290800074948,
+      "grad_norm": 0.22668063640594482,
+      "learning_rate": 0.0002879729729729729,
+      "loss": 0.5343,
+      "step": 277
+    },
+    {
+      "epoch": 0.08334270189244894,
+      "grad_norm": 0.23789212107658386,
+      "learning_rate": 0.0002879279279279279,
+      "loss": 0.5325,
+      "step": 278
+    },
+    {
+      "epoch": 0.0836424957841484,
+      "grad_norm": 0.23370423913002014,
+      "learning_rate": 0.00028788288288288284,
+      "loss": 0.5054,
+      "step": 279
+    },
+    {
+      "epoch": 0.08394228967584785,
+      "grad_norm": 0.2391810417175293,
+      "learning_rate": 0.0002878378378378378,
+      "loss": 0.5128,
+      "step": 280
+    },
+    {
+      "epoch": 0.08424208356754731,
+      "grad_norm": 0.2511520981788635,
+      "learning_rate": 0.00028779279279279277,
+      "loss": 0.5205,
+      "step": 281
+    },
+    {
+      "epoch": 0.08454187745924677,
+      "grad_norm": 0.25624001026153564,
+      "learning_rate": 0.0002877477477477477,
+      "loss": 0.5341,
+      "step": 282
+    },
+    {
+      "epoch": 0.08484167135094622,
+      "grad_norm": 0.22454805672168732,
+      "learning_rate": 0.00028770270270270264,
+      "loss": 0.4776,
+      "step": 283
+    },
+    {
+      "epoch": 0.08514146524264568,
+      "grad_norm": 0.24822303652763367,
+      "learning_rate": 0.00028765765765765764,
+      "loss": 0.5545,
+      "step": 284
+    },
+    {
+      "epoch": 0.08544125913434514,
+      "grad_norm": 0.24899134039878845,
+      "learning_rate": 0.00028761261261261257,
+      "loss": 0.5081,
+      "step": 285
+    },
+    {
+      "epoch": 0.0857410530260446,
+      "grad_norm": 0.22409197688102722,
+      "learning_rate": 0.00028756756756756756,
+      "loss": 0.4964,
+      "step": 286
+    },
+    {
+      "epoch": 0.08604084691774405,
+      "grad_norm": 0.24713397026062012,
+      "learning_rate": 0.0002875225225225225,
+      "loss": 0.5604,
+      "step": 287
+    },
+    {
+      "epoch": 0.08634064080944351,
+      "grad_norm": 0.2386777251958847,
+      "learning_rate": 0.00028747747747747744,
+      "loss": 0.5158,
+      "step": 288
+    },
+    {
+      "epoch": 0.08664043470114297,
+      "grad_norm": 0.2695513963699341,
+      "learning_rate": 0.00028743243243243243,
+      "loss": 0.5289,
+      "step": 289
+    },
+    {
+      "epoch": 0.08694022859284242,
+      "grad_norm": 0.23554596304893494,
+      "learning_rate": 0.00028738738738738736,
+      "loss": 0.5002,
+      "step": 290
+    },
+    {
+      "epoch": 0.08724002248454188,
+      "grad_norm": 0.2348342090845108,
+      "learning_rate": 0.0002873423423423423,
+      "loss": 0.5026,
+      "step": 291
+    },
+    {
+      "epoch": 0.08753981637624134,
+      "grad_norm": 0.24448086321353912,
+      "learning_rate": 0.0002872972972972973,
+      "loss": 0.508,
+      "step": 292
+    },
+    {
+      "epoch": 0.0878396102679408,
+      "grad_norm": 0.22410009801387787,
+      "learning_rate": 0.00028725225225225223,
+      "loss": 0.4932,
+      "step": 293
+    },
+    {
+      "epoch": 0.08813940415964025,
+      "grad_norm": 0.25097909569740295,
+      "learning_rate": 0.00028720720720720717,
+      "loss": 0.5196,
+      "step": 294
+    },
+    {
+      "epoch": 0.0884391980513397,
+      "grad_norm": 0.23564878106117249,
+      "learning_rate": 0.00028716216216216216,
+      "loss": 0.4871,
+      "step": 295
+    },
+    {
+      "epoch": 0.08873899194303916,
+      "grad_norm": 0.22870436310768127,
+      "learning_rate": 0.0002871171171171171,
+      "loss": 0.4717,
+      "step": 296
+    },
+    {
+      "epoch": 0.08903878583473862,
+      "grad_norm": 0.24509447813034058,
+      "learning_rate": 0.00028707207207207203,
+      "loss": 0.5417,
+      "step": 297
+    },
+    {
+      "epoch": 0.08933857972643808,
+      "grad_norm": 0.22554557025432587,
+      "learning_rate": 0.000287027027027027,
+      "loss": 0.5081,
+      "step": 298
+    },
+    {
+      "epoch": 0.08963837361813753,
+      "grad_norm": 0.2863612473011017,
+      "learning_rate": 0.00028698198198198196,
+      "loss": 0.5453,
+      "step": 299
+    },
+    {
+      "epoch": 0.08993816750983699,
+      "grad_norm": 0.23862136900424957,
+      "learning_rate": 0.00028693693693693695,
+      "loss": 0.4938,
+      "step": 300
+    },
+    {
+      "epoch": 0.09023796140153645,
+      "grad_norm": 0.2628205120563507,
+      "learning_rate": 0.0002868918918918919,
+      "loss": 0.5392,
+      "step": 301
+    },
+    {
+      "epoch": 0.0905377552932359,
+      "grad_norm": 0.23156946897506714,
+      "learning_rate": 0.0002868468468468468,
+      "loss": 0.4983,
+      "step": 302
+    },
+    {
+      "epoch": 0.09083754918493536,
+      "grad_norm": 0.27579790353775024,
+      "learning_rate": 0.0002868018018018018,
+      "loss": 0.5343,
+      "step": 303
+    },
+    {
+      "epoch": 0.09113734307663482,
+      "grad_norm": 0.22691994905471802,
+      "learning_rate": 0.00028675675675675675,
+      "loss": 0.4828,
+      "step": 304
+    },
+    {
+      "epoch": 0.09143713696833428,
+      "grad_norm": 0.23544538021087646,
+      "learning_rate": 0.0002867117117117117,
+      "loss": 0.5143,
+      "step": 305
+    },
+    {
+      "epoch": 0.09173693086003373,
+      "grad_norm": 0.22597934305667877,
+      "learning_rate": 0.0002866666666666667,
+      "loss": 0.5059,
+      "step": 306
+    },
+    {
+      "epoch": 0.09203672475173319,
+      "grad_norm": 0.23851321637630463,
+      "learning_rate": 0.0002866216216216216,
+      "loss": 0.5127,
+      "step": 307
+    },
+    {
+      "epoch": 0.09233651864343265,
+      "grad_norm": 0.2260097861289978,
+      "learning_rate": 0.00028657657657657655,
+      "loss": 0.4901,
+      "step": 308
+    },
+    {
+      "epoch": 0.09263631253513209,
+      "grad_norm": 0.2568291127681732,
+      "learning_rate": 0.00028653153153153154,
+      "loss": 0.556,
+      "step": 309
+    },
+    {
+      "epoch": 0.09293610642683155,
+      "grad_norm": 0.23510834574699402,
+      "learning_rate": 0.0002864864864864865,
+      "loss": 0.5254,
+      "step": 310
+    },
+    {
+      "epoch": 0.093235900318531,
+      "grad_norm": 0.24556247889995575,
+      "learning_rate": 0.0002864414414414414,
+      "loss": 0.5292,
+      "step": 311
+    },
+    {
+      "epoch": 0.09353569421023046,
+      "grad_norm": 0.2357538938522339,
+      "learning_rate": 0.00028639639639639635,
+      "loss": 0.5321,
+      "step": 312
+    },
+    {
+      "epoch": 0.09383548810192992,
+      "grad_norm": 0.25484514236450195,
+      "learning_rate": 0.00028635135135135134,
+      "loss": 0.5449,
+      "step": 313
+    },
+    {
+      "epoch": 0.09413528199362937,
+      "grad_norm": 0.243759885430336,
+      "learning_rate": 0.0002863063063063063,
+      "loss": 0.4766,
+      "step": 314
+    },
+    {
+      "epoch": 0.09443507588532883,
+      "grad_norm": 0.232033833861351,
+      "learning_rate": 0.0002862612612612612,
+      "loss": 0.481,
+      "step": 315
+    },
+    {
+      "epoch": 0.09473486977702829,
+      "grad_norm": 0.24136248230934143,
+      "learning_rate": 0.0002862162162162162,
+      "loss": 0.541,
+      "step": 316
+    },
+    {
+      "epoch": 0.09503466366872775,
+      "grad_norm": 0.23708300292491913,
+      "learning_rate": 0.00028617117117117114,
+      "loss": 0.5298,
+      "step": 317
+    },
+    {
+      "epoch": 0.0953344575604272,
+      "grad_norm": 0.22408358752727509,
+      "learning_rate": 0.0002861261261261261,
+      "loss": 0.4818,
+      "step": 318
+    },
+    {
+      "epoch": 0.09563425145212666,
+      "grad_norm": 0.22969000041484833,
+      "learning_rate": 0.00028608108108108107,
+      "loss": 0.4668,
+      "step": 319
+    },
+    {
+      "epoch": 0.09593404534382612,
+      "grad_norm": 0.233534574508667,
+      "learning_rate": 0.000286036036036036,
+      "loss": 0.506,
+      "step": 320
+    },
+    {
+      "epoch": 0.09623383923552557,
+      "grad_norm": 0.2442328929901123,
+      "learning_rate": 0.00028599099099099095,
+      "loss": 0.5208,
+      "step": 321
+    },
+    {
+      "epoch": 0.09653363312722503,
+      "grad_norm": 0.22215022146701813,
+      "learning_rate": 0.00028594594594594594,
+      "loss": 0.4963,
+      "step": 322
+    },
+    {
+      "epoch": 0.09683342701892449,
+      "grad_norm": 0.23993152379989624,
+      "learning_rate": 0.0002859009009009009,
+      "loss": 0.4945,
+      "step": 323
+    },
+    {
+      "epoch": 0.09713322091062394,
+      "grad_norm": 0.22987253963947296,
+      "learning_rate": 0.0002858558558558558,
+      "loss": 0.4895,
+      "step": 324
+    },
+    {
+      "epoch": 0.0974330148023234,
+      "grad_norm": 0.25025674700737,
+      "learning_rate": 0.0002858108108108108,
+      "loss": 0.5053,
+      "step": 325
+    },
+    {
+      "epoch": 0.09773280869402286,
+      "grad_norm": 0.21882835030555725,
+      "learning_rate": 0.00028576576576576574,
+      "loss": 0.4844,
+      "step": 326
+    },
+    {
+      "epoch": 0.09803260258572231,
+      "grad_norm": 0.23361052572727203,
+      "learning_rate": 0.0002857207207207207,
+      "loss": 0.5159,
+      "step": 327
+    },
+    {
+      "epoch": 0.09833239647742177,
+      "grad_norm": 0.23713140189647675,
+      "learning_rate": 0.00028567567567567567,
+      "loss": 0.5419,
+      "step": 328
+    },
+    {
+      "epoch": 0.09863219036912123,
+      "grad_norm": 0.2439422458410263,
+      "learning_rate": 0.0002856306306306306,
+      "loss": 0.5068,
+      "step": 329
+    },
+    {
+      "epoch": 0.09893198426082069,
+      "grad_norm": 0.23442302644252777,
+      "learning_rate": 0.00028558558558558554,
+      "loss": 0.5041,
+      "step": 330
+    },
+    {
+      "epoch": 0.09923177815252014,
+      "grad_norm": 0.2295604944229126,
+      "learning_rate": 0.00028554054054054053,
+      "loss": 0.4838,
+      "step": 331
+    },
+    {
+      "epoch": 0.0995315720442196,
+      "grad_norm": 0.2328769713640213,
+      "learning_rate": 0.00028549549549549547,
+      "loss": 0.5118,
+      "step": 332
+    },
+    {
+      "epoch": 0.09983136593591906,
+      "grad_norm": 0.21605005860328674,
+      "learning_rate": 0.0002854504504504504,
+      "loss": 0.4786,
+      "step": 333
+    },
+    {
+      "epoch": 0.10013115982761851,
+      "grad_norm": 0.22740024328231812,
+      "learning_rate": 0.0002854054054054054,
+      "loss": 0.5116,
+      "step": 334
+    },
+    {
+      "epoch": 0.10043095371931797,
+      "grad_norm": 0.24437595903873444,
+      "learning_rate": 0.00028536036036036033,
+      "loss": 0.5467,
+      "step": 335
+    },
+    {
+      "epoch": 0.10073074761101743,
+      "grad_norm": 0.2281450480222702,
+      "learning_rate": 0.00028531531531531527,
+      "loss": 0.5269,
+      "step": 336
+    },
+    {
+      "epoch": 0.10103054150271688,
+      "grad_norm": 0.2343592643737793,
+      "learning_rate": 0.0002852702702702702,
+      "loss": 0.5176,
+      "step": 337
+    },
+    {
+      "epoch": 0.10133033539441634,
+      "grad_norm": 0.24275358021259308,
+      "learning_rate": 0.0002852252252252252,
+      "loss": 0.5103,
+      "step": 338
+    },
+    {
+      "epoch": 0.1016301292861158,
+      "grad_norm": 0.2498926818370819,
+      "learning_rate": 0.00028518018018018013,
+      "loss": 0.5282,
+      "step": 339
+    },
+    {
+      "epoch": 0.10192992317781525,
+      "grad_norm": 0.23029617965221405,
+      "learning_rate": 0.00028513513513513507,
+      "loss": 0.512,
+      "step": 340
+    },
+    {
+      "epoch": 0.10222971706951471,
+      "grad_norm": 0.2608179450035095,
+      "learning_rate": 0.00028509009009009006,
+      "loss": 0.5419,
+      "step": 341
+    },
+    {
+      "epoch": 0.10252951096121417,
+      "grad_norm": 0.24309523403644562,
+      "learning_rate": 0.000285045045045045,
+      "loss": 0.5178,
+      "step": 342
+    },
+    {
+      "epoch": 0.10282930485291362,
+      "grad_norm": 0.22357851266860962,
+      "learning_rate": 0.000285,
+      "loss": 0.487,
+      "step": 343
+    },
+    {
+      "epoch": 0.10312909874461308,
+      "grad_norm": 0.2575152814388275,
+      "learning_rate": 0.0002849549549549549,
+      "loss": 0.5163,
+      "step": 344
+    },
+    {
+      "epoch": 0.10342889263631254,
+      "grad_norm": 0.25661131739616394,
+      "learning_rate": 0.00028490990990990986,
+      "loss": 0.5229,
+      "step": 345
+    },
+    {
+      "epoch": 0.103728686528012,
+      "grad_norm": 0.25328096747398376,
+      "learning_rate": 0.00028486486486486485,
+      "loss": 0.5456,
+      "step": 346
+    },
+    {
+      "epoch": 0.10402848041971145,
+      "grad_norm": 0.2357887625694275,
+      "learning_rate": 0.0002848198198198198,
+      "loss": 0.4824,
+      "step": 347
+    },
+    {
+      "epoch": 0.10432827431141091,
+      "grad_norm": 0.2275196760892868,
+      "learning_rate": 0.0002847747747747747,
+      "loss": 0.4899,
+      "step": 348
+    },
+    {
+      "epoch": 0.10462806820311037,
+      "grad_norm": 0.2616022229194641,
+      "learning_rate": 0.0002847297297297297,
+      "loss": 0.543,
+      "step": 349
+    },
+    {
+      "epoch": 0.10492786209480982,
+      "grad_norm": 0.23371827602386475,
+      "learning_rate": 0.00028468468468468465,
+      "loss": 0.4876,
+      "step": 350
+    },
+    {
+      "epoch": 0.10522765598650928,
+      "grad_norm": 0.21652457118034363,
+      "learning_rate": 0.0002846396396396396,
+      "loss": 0.5051,
+      "step": 351
+    },
+    {
+      "epoch": 0.10552744987820874,
+      "grad_norm": 0.2464732825756073,
+      "learning_rate": 0.0002845945945945946,
+      "loss": 0.4993,
+      "step": 352
+    },
+    {
+      "epoch": 0.1058272437699082,
+      "grad_norm": 0.24194246530532837,
+      "learning_rate": 0.0002845495495495495,
+      "loss": 0.492,
+      "step": 353
+    },
+    {
+      "epoch": 0.10612703766160765,
+      "grad_norm": 0.21986526250839233,
+      "learning_rate": 0.00028450450450450446,
+      "loss": 0.4668,
+      "step": 354
+    },
+    {
+      "epoch": 0.10642683155330711,
+      "grad_norm": 0.2386123389005661,
+      "learning_rate": 0.00028445945945945945,
+      "loss": 0.4823,
+      "step": 355
+    },
+    {
+      "epoch": 0.10672662544500656,
+      "grad_norm": 0.2368634194135666,
+      "learning_rate": 0.0002844144144144144,
+      "loss": 0.5119,
+      "step": 356
+    },
+    {
+      "epoch": 0.10702641933670601,
+      "grad_norm": 0.24802769720554352,
+      "learning_rate": 0.0002843693693693694,
+      "loss": 0.5298,
+      "step": 357
+    },
+    {
+      "epoch": 0.10732621322840546,
+      "grad_norm": 0.25534310936927795,
+      "learning_rate": 0.0002843243243243243,
+      "loss": 0.5107,
+      "step": 358
+    },
+    {
+      "epoch": 0.10762600712010492,
+      "grad_norm": 0.24845390021800995,
+      "learning_rate": 0.00028427927927927925,
+      "loss": 0.539,
+      "step": 359
+    },
+    {
+      "epoch": 0.10792580101180438,
+      "grad_norm": 0.22670955955982208,
+      "learning_rate": 0.00028423423423423424,
+      "loss": 0.5013,
+      "step": 360
+    },
+    {
+      "epoch": 0.10822559490350384,
+      "grad_norm": 0.21695859730243683,
+      "learning_rate": 0.0002841891891891892,
+      "loss": 0.468,
+      "step": 361
+    },
+    {
+      "epoch": 0.10852538879520329,
+      "grad_norm": 0.24691784381866455,
+      "learning_rate": 0.0002841441441441441,
+      "loss": 0.5011,
+      "step": 362
+    },
+    {
+      "epoch": 0.10882518268690275,
+      "grad_norm": 0.23244348168373108,
+      "learning_rate": 0.0002840990990990991,
+      "loss": 0.481,
+      "step": 363
+    },
+    {
+      "epoch": 0.1091249765786022,
+      "grad_norm": 0.23756597936153412,
+      "learning_rate": 0.00028405405405405404,
+      "loss": 0.4869,
+      "step": 364
+    },
+    {
+      "epoch": 0.10942477047030166,
+      "grad_norm": 0.26336169242858887,
+      "learning_rate": 0.000284009009009009,
+      "loss": 0.5216,
+      "step": 365
+    },
+    {
+      "epoch": 0.10972456436200112,
+      "grad_norm": 0.23734593391418457,
+      "learning_rate": 0.00028396396396396397,
+      "loss": 0.502,
+      "step": 366
+    },
+    {
+      "epoch": 0.11002435825370058,
+      "grad_norm": 0.2188960462808609,
+      "learning_rate": 0.0002839189189189189,
+      "loss": 0.4784,
+      "step": 367
+    },
+    {
+      "epoch": 0.11032415214540003,
+      "grad_norm": 0.26715338230133057,
+      "learning_rate": 0.00028387387387387384,
+      "loss": 0.4997,
+      "step": 368
+    },
+    {
+      "epoch": 0.11062394603709949,
+      "grad_norm": 0.23621448874473572,
+      "learning_rate": 0.00028382882882882883,
+      "loss": 0.4819,
+      "step": 369
+    },
+    {
+      "epoch": 0.11092373992879895,
+      "grad_norm": 0.23166733980178833,
+      "learning_rate": 0.00028378378378378377,
+      "loss": 0.4949,
+      "step": 370
+    },
+    {
+      "epoch": 0.1112235338204984,
+      "grad_norm": 0.22275973856449127,
+      "learning_rate": 0.0002837387387387387,
+      "loss": 0.4893,
+      "step": 371
+    },
+    {
+      "epoch": 0.11152332771219786,
+      "grad_norm": 0.22097566723823547,
+      "learning_rate": 0.0002836936936936937,
+      "loss": 0.5011,
+      "step": 372
+    },
+    {
+      "epoch": 0.11182312160389732,
+      "grad_norm": 0.25464650988578796,
+      "learning_rate": 0.00028364864864864863,
+      "loss": 0.5369,
+      "step": 373
+    },
+    {
+      "epoch": 0.11212291549559678,
+      "grad_norm": 0.2584247589111328,
+      "learning_rate": 0.00028360360360360357,
+      "loss": 0.5271,
+      "step": 374
+    },
+    {
+      "epoch": 0.11242270938729623,
+      "grad_norm": 0.2229800671339035,
+      "learning_rate": 0.00028355855855855856,
+      "loss": 0.4919,
+      "step": 375
+    },
+    {
+      "epoch": 0.11272250327899569,
+      "grad_norm": 0.25196340680122375,
+      "learning_rate": 0.0002835135135135135,
+      "loss": 0.4935,
+      "step": 376
+    },
+    {
+      "epoch": 0.11302229717069515,
+      "grad_norm": 0.23945283889770508,
+      "learning_rate": 0.00028346846846846843,
+      "loss": 0.5158,
+      "step": 377
+    },
+    {
+      "epoch": 0.1133220910623946,
+      "grad_norm": 0.22441525757312775,
+      "learning_rate": 0.0002834234234234234,
+      "loss": 0.5106,
+      "step": 378
+    },
+    {
+      "epoch": 0.11362188495409406,
+      "grad_norm": 0.24109874665737152,
+      "learning_rate": 0.00028337837837837836,
+      "loss": 0.4646,
+      "step": 379
+    },
+    {
+      "epoch": 0.11392167884579352,
+      "grad_norm": 0.23156128823757172,
+      "learning_rate": 0.0002833333333333333,
+      "loss": 0.4752,
+      "step": 380
+    },
+    {
+      "epoch": 0.11422147273749297,
+      "grad_norm": 0.24328507483005524,
+      "learning_rate": 0.0002832882882882883,
+      "loss": 0.5225,
+      "step": 381
+    },
+    {
+      "epoch": 0.11452126662919243,
+      "grad_norm": 0.2364337146282196,
+      "learning_rate": 0.00028324324324324323,
+      "loss": 0.4965,
+      "step": 382
+    },
+    {
+      "epoch": 0.11482106052089189,
+      "grad_norm": 0.24130620062351227,
+      "learning_rate": 0.00028319819819819816,
+      "loss": 0.5058,
+      "step": 383
+    },
+    {
+      "epoch": 0.11512085441259134,
+      "grad_norm": 0.2482600063085556,
+      "learning_rate": 0.00028315315315315315,
+      "loss": 0.5142,
+      "step": 384
+    },
+    {
+      "epoch": 0.1154206483042908,
+      "grad_norm": 0.2426941990852356,
+      "learning_rate": 0.0002831081081081081,
+      "loss": 0.5185,
+      "step": 385
+    },
+    {
+      "epoch": 0.11572044219599026,
+      "grad_norm": 0.24431252479553223,
+      "learning_rate": 0.00028306306306306303,
+      "loss": 0.4946,
+      "step": 386
+    },
+    {
+      "epoch": 0.11602023608768972,
+      "grad_norm": 0.22974024713039398,
+      "learning_rate": 0.00028301801801801797,
+      "loss": 0.4845,
+      "step": 387
+    },
+    {
+      "epoch": 0.11632002997938917,
+      "grad_norm": 0.26395878195762634,
+      "learning_rate": 0.00028297297297297296,
+      "loss": 0.512,
+      "step": 388
+    },
+    {
+      "epoch": 0.11661982387108863,
+      "grad_norm": 0.2300662398338318,
+      "learning_rate": 0.0002829279279279279,
+      "loss": 0.5161,
+      "step": 389
+    },
+    {
+      "epoch": 0.11691961776278809,
+      "grad_norm": 0.23376531898975372,
+      "learning_rate": 0.00028288288288288283,
+      "loss": 0.5165,
+      "step": 390
+    },
+    {
+      "epoch": 0.11721941165448754,
+      "grad_norm": 0.23775231838226318,
+      "learning_rate": 0.0002828378378378378,
+      "loss": 0.5028,
+      "step": 391
+    },
+    {
+      "epoch": 0.117519205546187,
+      "grad_norm": 0.22368961572647095,
+      "learning_rate": 0.00028279279279279276,
+      "loss": 0.4724,
+      "step": 392
+    },
+    {
+      "epoch": 0.11781899943788646,
+      "grad_norm": 0.25789904594421387,
+      "learning_rate": 0.0002827477477477477,
+      "loss": 0.5545,
+      "step": 393
+    },
+    {
+      "epoch": 0.11811879332958591,
+      "grad_norm": 0.22045502066612244,
+      "learning_rate": 0.0002827027027027027,
+      "loss": 0.4739,
+      "step": 394
+    },
+    {
+      "epoch": 0.11841858722128537,
+      "grad_norm": 0.24297311902046204,
+      "learning_rate": 0.0002826576576576576,
+      "loss": 0.5284,
+      "step": 395
+    },
+    {
+      "epoch": 0.11871838111298483,
+      "grad_norm": 0.22944821417331696,
+      "learning_rate": 0.00028261261261261256,
+      "loss": 0.5071,
+      "step": 396
+    },
+    {
+      "epoch": 0.11901817500468428,
+      "grad_norm": 0.22093501687049866,
+      "learning_rate": 0.00028256756756756755,
+      "loss": 0.5016,
+      "step": 397
+    },
+    {
+      "epoch": 0.11931796889638374,
+      "grad_norm": 0.24359877407550812,
+      "learning_rate": 0.0002825225225225225,
+      "loss": 0.494,
+      "step": 398
+    },
+    {
+      "epoch": 0.1196177627880832,
+      "grad_norm": 0.23019848763942719,
+      "learning_rate": 0.0002824774774774774,
+      "loss": 0.4774,
+      "step": 399
+    },
+    {
+      "epoch": 0.11991755667978266,
+      "grad_norm": 0.22894737124443054,
+      "learning_rate": 0.0002824324324324324,
+      "loss": 0.4804,
+      "step": 400
+    },
+    {
+      "epoch": 0.12021735057148211,
+      "grad_norm": 0.23049385845661163,
+      "learning_rate": 0.00028238738738738735,
+      "loss": 0.4816,
+      "step": 401
+    },
+    {
+      "epoch": 0.12051714446318157,
+      "grad_norm": 0.21988703310489655,
+      "learning_rate": 0.0002823423423423423,
+      "loss": 0.479,
+      "step": 402
+    },
+    {
+      "epoch": 0.12081693835488103,
+      "grad_norm": 0.2389378845691681,
+      "learning_rate": 0.0002822972972972973,
+      "loss": 0.4993,
+      "step": 403
+    },
+    {
+      "epoch": 0.12111673224658047,
+      "grad_norm": 0.2271224707365036,
+      "learning_rate": 0.0002822522522522522,
+      "loss": 0.479,
+      "step": 404
+    },
+    {
+      "epoch": 0.12141652613827993,
+      "grad_norm": 0.253499835729599,
+      "learning_rate": 0.00028220720720720715,
+      "loss": 0.5289,
+      "step": 405
+    },
+    {
+      "epoch": 0.12171632002997938,
+      "grad_norm": 0.2213011533021927,
+      "learning_rate": 0.00028216216216216214,
+      "loss": 0.4759,
+      "step": 406
+    },
+    {
+      "epoch": 0.12201611392167884,
+      "grad_norm": 0.2327415943145752,
+      "learning_rate": 0.0002821171171171171,
+      "loss": 0.4822,
+      "step": 407
+    },
+    {
+      "epoch": 0.1223159078133783,
+      "grad_norm": 0.2592346668243408,
+      "learning_rate": 0.000282072072072072,
+      "loss": 0.5331,
+      "step": 408
+    },
+    {
+      "epoch": 0.12261570170507775,
+      "grad_norm": 0.23039428889751434,
+      "learning_rate": 0.000282027027027027,
+      "loss": 0.4882,
+      "step": 409
+    },
+    {
+      "epoch": 0.12291549559677721,
+      "grad_norm": 0.2345624566078186,
+      "learning_rate": 0.00028198198198198194,
+      "loss": 0.4859,
+      "step": 410
+    },
+    {
+      "epoch": 0.12321528948847667,
+      "grad_norm": 0.22759179770946503,
+      "learning_rate": 0.0002819369369369369,
+      "loss": 0.484,
+      "step": 411
+    },
+    {
+      "epoch": 0.12351508338017612,
+      "grad_norm": 0.24670295417308807,
+      "learning_rate": 0.00028189189189189187,
+      "loss": 0.5047,
+      "step": 412
+    },
+    {
+      "epoch": 0.12381487727187558,
+      "grad_norm": 0.2306670844554901,
+      "learning_rate": 0.0002818468468468468,
+      "loss": 0.5041,
+      "step": 413
+    },
+    {
+      "epoch": 0.12411467116357504,
+      "grad_norm": 0.2440173178911209,
+      "learning_rate": 0.0002818018018018018,
+      "loss": 0.4942,
+      "step": 414
+    },
+    {
+      "epoch": 0.1244144650552745,
+      "grad_norm": 0.2239232212305069,
+      "learning_rate": 0.00028175675675675674,
+      "loss": 0.4677,
+      "step": 415
+    },
+    {
+      "epoch": 0.12471425894697395,
+      "grad_norm": 0.22838331758975983,
+      "learning_rate": 0.0002817117117117117,
+      "loss": 0.5061,
+      "step": 416
+    },
+    {
+      "epoch": 0.12501405283867342,
+      "grad_norm": 0.23874101042747498,
+      "learning_rate": 0.00028166666666666666,
+      "loss": 0.5261,
+      "step": 417
+    },
+    {
+      "epoch": 0.12531384673037288,
+      "grad_norm": 0.2522803544998169,
+      "learning_rate": 0.0002816216216216216,
+      "loss": 0.5525,
+      "step": 418
+    },
+    {
+      "epoch": 0.12561364062207234,
+      "grad_norm": 0.2339329719543457,
+      "learning_rate": 0.00028157657657657654,
+      "loss": 0.5157,
+      "step": 419
+    },
+    {
+      "epoch": 0.1259134345137718,
+      "grad_norm": 0.23074860870838165,
+      "learning_rate": 0.00028153153153153153,
+      "loss": 0.5062,
+      "step": 420
+    },
+    {
+      "epoch": 0.12621322840547125,
+      "grad_norm": 0.22877013683319092,
+      "learning_rate": 0.00028148648648648647,
+      "loss": 0.4827,
+      "step": 421
+    },
+    {
+      "epoch": 0.1265130222971707,
+      "grad_norm": 0.2384471893310547,
+      "learning_rate": 0.0002814414414414414,
+      "loss": 0.5035,
+      "step": 422
+    },
+    {
+      "epoch": 0.12681281618887016,
+      "grad_norm": 0.24479855597019196,
+      "learning_rate": 0.0002813963963963964,
+      "loss": 0.4645,
+      "step": 423
+    },
+    {
+      "epoch": 0.12711261008056962,
+      "grad_norm": 0.23424111306667328,
+      "learning_rate": 0.00028135135135135133,
+      "loss": 0.4927,
+      "step": 424
+    },
+    {
+      "epoch": 0.12741240397226908,
+      "grad_norm": 0.24012430012226105,
+      "learning_rate": 0.0002813063063063063,
+      "loss": 0.4904,
+      "step": 425
+    },
+    {
+      "epoch": 0.12771219786396854,
+      "grad_norm": 0.24606235325336456,
+      "learning_rate": 0.00028126126126126126,
+      "loss": 0.5143,
+      "step": 426
+    },
+    {
+      "epoch": 0.128011991755668,
+      "grad_norm": 0.24893403053283691,
+      "learning_rate": 0.0002812162162162162,
+      "loss": 0.5232,
+      "step": 427
+    },
+    {
+      "epoch": 0.12831178564736742,
+      "grad_norm": 0.2424110770225525,
+      "learning_rate": 0.0002811711711711712,
+      "loss": 0.5232,
+      "step": 428
+    },
+    {
+      "epoch": 0.12861157953906688,
+      "grad_norm": 0.2312486171722412,
+      "learning_rate": 0.0002811261261261261,
+      "loss": 0.4645,
+      "step": 429
+    },
+    {
+      "epoch": 0.12891137343076634,
+      "grad_norm": 0.22639836370944977,
+      "learning_rate": 0.00028108108108108106,
+      "loss": 0.4642,
+      "step": 430
+    },
+    {
+      "epoch": 0.1292111673224658,
+      "grad_norm": 0.23626941442489624,
+      "learning_rate": 0.00028103603603603605,
+      "loss": 0.5038,
+      "step": 431
+    },
+    {
+      "epoch": 0.12951096121416525,
+      "grad_norm": 0.2625383138656616,
+      "learning_rate": 0.000280990990990991,
+      "loss": 0.4867,
+      "step": 432
+    },
+    {
+      "epoch": 0.1298107551058647,
+      "grad_norm": 0.24292655289173126,
+      "learning_rate": 0.0002809459459459459,
+      "loss": 0.5081,
+      "step": 433
+    },
+    {
+      "epoch": 0.13011054899756416,
+      "grad_norm": 0.23609769344329834,
+      "learning_rate": 0.00028090090090090086,
+      "loss": 0.4832,
+      "step": 434
+    },
+    {
+      "epoch": 0.13041034288926362,
+      "grad_norm": 0.22934478521347046,
+      "learning_rate": 0.00028085585585585585,
+      "loss": 0.4872,
+      "step": 435
+    },
+    {
+      "epoch": 0.13071013678096308,
+      "grad_norm": 0.22949008643627167,
+      "learning_rate": 0.0002808108108108108,
+      "loss": 0.5129,
+      "step": 436
+    },
+    {
+      "epoch": 0.13100993067266253,
+      "grad_norm": 0.2302381694316864,
+      "learning_rate": 0.0002807657657657657,
+      "loss": 0.4793,
+      "step": 437
+    },
+    {
+      "epoch": 0.131309724564362,
+      "grad_norm": 0.23368242383003235,
+      "learning_rate": 0.0002807207207207207,
+      "loss": 0.4989,
+      "step": 438
+    },
+    {
+      "epoch": 0.13160951845606145,
+      "grad_norm": 0.21572020649909973,
+      "learning_rate": 0.00028067567567567565,
+      "loss": 0.4546,
+      "step": 439
+    },
+    {
+      "epoch": 0.1319093123477609,
+      "grad_norm": 0.2268449366092682,
+      "learning_rate": 0.0002806306306306306,
+      "loss": 0.4954,
+      "step": 440
+    },
+    {
+      "epoch": 0.13220910623946036,
+      "grad_norm": 0.23617544770240784,
+      "learning_rate": 0.0002805855855855856,
+      "loss": 0.4865,
+      "step": 441
+    },
+    {
+      "epoch": 0.13250890013115982,
+      "grad_norm": 0.24015142023563385,
+      "learning_rate": 0.0002805405405405405,
+      "loss": 0.5211,
+      "step": 442
+    },
+    {
+      "epoch": 0.13280869402285927,
+      "grad_norm": 0.21798421442508698,
+      "learning_rate": 0.00028049549549549545,
+      "loss": 0.4482,
+      "step": 443
+    },
+    {
+      "epoch": 0.13310848791455873,
+      "grad_norm": 0.23476584255695343,
+      "learning_rate": 0.00028045045045045045,
+      "loss": 0.4871,
+      "step": 444
+    },
+    {
+      "epoch": 0.1334082818062582,
+      "grad_norm": 0.2404216080904007,
+      "learning_rate": 0.0002804054054054054,
+      "loss": 0.488,
+      "step": 445
+    },
+    {
+      "epoch": 0.13370807569795765,
+      "grad_norm": 0.25073060393333435,
+      "learning_rate": 0.0002803603603603603,
+      "loss": 0.5194,
+      "step": 446
+    },
+    {
+      "epoch": 0.1340078695896571,
+      "grad_norm": 0.24332286417484283,
+      "learning_rate": 0.0002803153153153153,
+      "loss": 0.5347,
+      "step": 447
+    },
+    {
+      "epoch": 0.13430766348135656,
+      "grad_norm": 0.2420525699853897,
+      "learning_rate": 0.00028027027027027025,
+      "loss": 0.5109,
+      "step": 448
+    },
+    {
+      "epoch": 0.13460745737305602,
+      "grad_norm": 0.22389326989650726,
+      "learning_rate": 0.0002802252252252252,
+      "loss": 0.5075,
+      "step": 449
+    },
+    {
+      "epoch": 0.13490725126475547,
+      "grad_norm": 0.23522739112377167,
+      "learning_rate": 0.0002801801801801802,
+      "loss": 0.4794,
+      "step": 450
+    },
+    {
+      "epoch": 0.13520704515645493,
+      "grad_norm": 0.221591517329216,
+      "learning_rate": 0.0002801351351351351,
+      "loss": 0.4821,
+      "step": 451
+    },
+    {
+      "epoch": 0.1355068390481544,
+      "grad_norm": 0.24136802554130554,
+      "learning_rate": 0.00028009009009009005,
+      "loss": 0.4949,
+      "step": 452
+    },
+    {
+      "epoch": 0.13580663293985384,
+      "grad_norm": 0.23207104206085205,
+      "learning_rate": 0.00028004504504504504,
+      "loss": 0.4803,
+      "step": 453
+    },
+    {
+      "epoch": 0.1361064268315533,
+      "grad_norm": 0.25318193435668945,
+      "learning_rate": 0.00028,
+      "loss": 0.5345,
+      "step": 454
+    },
+    {
+      "epoch": 0.13640622072325276,
+      "grad_norm": 0.25274619460105896,
+      "learning_rate": 0.0002799549549549549,
+      "loss": 0.5159,
+      "step": 455
+    },
+    {
+      "epoch": 0.13670601461495221,
+      "grad_norm": 0.22540399432182312,
+      "learning_rate": 0.0002799099099099099,
+      "loss": 0.4772,
+      "step": 456
+    },
+    {
+      "epoch": 0.13700580850665167,
+      "grad_norm": 0.2346925288438797,
+      "learning_rate": 0.00027986486486486484,
+      "loss": 0.4916,
+      "step": 457
+    },
+    {
+      "epoch": 0.13730560239835113,
+      "grad_norm": 0.2226891815662384,
+      "learning_rate": 0.0002798198198198198,
+      "loss": 0.4651,
+      "step": 458
+    },
+    {
+      "epoch": 0.13760539629005059,
+      "grad_norm": 0.25540515780448914,
+      "learning_rate": 0.0002797747747747747,
+      "loss": 0.5316,
+      "step": 459
+    },
+    {
+      "epoch": 0.13790519018175004,
+      "grad_norm": 0.22934426367282867,
+      "learning_rate": 0.0002797297297297297,
+      "loss": 0.477,
+      "step": 460
+    },
+    {
+      "epoch": 0.1382049840734495,
+      "grad_norm": 0.22268570959568024,
+      "learning_rate": 0.00027968468468468464,
+      "loss": 0.4611,
+      "step": 461
+    },
+    {
+      "epoch": 0.13850477796514896,
+      "grad_norm": 0.23548570275306702,
+      "learning_rate": 0.0002796396396396396,
+      "loss": 0.5012,
+      "step": 462
+    },
+    {
+      "epoch": 0.1388045718568484,
+      "grad_norm": 0.22782792150974274,
+      "learning_rate": 0.00027959459459459457,
+      "loss": 0.4878,
+      "step": 463
+    },
+    {
+      "epoch": 0.13910436574854787,
+      "grad_norm": 0.24569828808307648,
+      "learning_rate": 0.0002795495495495495,
+      "loss": 0.4882,
+      "step": 464
+    },
+    {
+      "epoch": 0.13940415964024733,
+      "grad_norm": 0.23523476719856262,
+      "learning_rate": 0.00027950450450450444,
+      "loss": 0.498,
+      "step": 465
+    },
+    {
+      "epoch": 0.13970395353194678,
+      "grad_norm": 0.24249842762947083,
+      "learning_rate": 0.00027945945945945943,
+      "loss": 0.5156,
+      "step": 466
+    },
+    {
+      "epoch": 0.14000374742364624,
+      "grad_norm": 0.22582505643367767,
+      "learning_rate": 0.00027941441441441437,
+      "loss": 0.4428,
+      "step": 467
+    },
+    {
+      "epoch": 0.1403035413153457,
+      "grad_norm": 0.2527635395526886,
+      "learning_rate": 0.0002793693693693693,
+      "loss": 0.507,
+      "step": 468
+    },
+    {
+      "epoch": 0.14060333520704515,
+      "grad_norm": 0.2490163892507553,
+      "learning_rate": 0.0002793243243243243,
+      "loss": 0.5119,
+      "step": 469
+    },
+    {
+      "epoch": 0.1409031290987446,
+      "grad_norm": 0.26502713561058044,
+      "learning_rate": 0.00027927927927927923,
+      "loss": 0.4994,
+      "step": 470
+    },
+    {
+      "epoch": 0.14120292299044407,
+      "grad_norm": 0.25225281715393066,
+      "learning_rate": 0.0002792342342342342,
+      "loss": 0.5029,
+      "step": 471
+    },
+    {
+      "epoch": 0.14150271688214353,
+      "grad_norm": 0.235763818025589,
+      "learning_rate": 0.00027918918918918916,
+      "loss": 0.5154,
+      "step": 472
+    },
+    {
+      "epoch": 0.14180251077384298,
+      "grad_norm": 0.24403534829616547,
+      "learning_rate": 0.0002791441441441441,
+      "loss": 0.5241,
+      "step": 473
+    },
+    {
+      "epoch": 0.14210230466554244,
+      "grad_norm": 0.24656488001346588,
+      "learning_rate": 0.0002790990990990991,
+      "loss": 0.4767,
+      "step": 474
+    },
+    {
+      "epoch": 0.1424020985572419,
+      "grad_norm": 0.2506386935710907,
+      "learning_rate": 0.000279054054054054,
+      "loss": 0.5066,
+      "step": 475
+    },
+    {
+      "epoch": 0.14270189244894135,
+      "grad_norm": 0.24634157121181488,
+      "learning_rate": 0.00027900900900900896,
+      "loss": 0.4815,
+      "step": 476
+    },
+    {
+      "epoch": 0.1430016863406408,
+      "grad_norm": 0.23619256913661957,
+      "learning_rate": 0.00027896396396396395,
+      "loss": 0.4929,
+      "step": 477
+    },
+    {
+      "epoch": 0.14330148023234027,
+      "grad_norm": 0.23421134054660797,
+      "learning_rate": 0.0002789189189189189,
+      "loss": 0.4841,
+      "step": 478
+    },
+    {
+      "epoch": 0.14360127412403972,
+      "grad_norm": 0.2287687510251999,
+      "learning_rate": 0.00027887387387387383,
+      "loss": 0.4976,
+      "step": 479
+    },
+    {
+      "epoch": 0.14390106801573918,
+      "grad_norm": 0.2362293004989624,
+      "learning_rate": 0.0002788288288288288,
+      "loss": 0.5054,
+      "step": 480
+    },
+    {
+      "epoch": 0.14420086190743864,
+      "grad_norm": 0.23907198011875153,
+      "learning_rate": 0.00027878378378378376,
+      "loss": 0.4879,
+      "step": 481
+    },
+    {
+      "epoch": 0.1445006557991381,
+      "grad_norm": 0.21802479028701782,
+      "learning_rate": 0.00027873873873873875,
+      "loss": 0.4807,
+      "step": 482
+    },
+    {
+      "epoch": 0.14480044969083755,
+      "grad_norm": 0.2445833832025528,
+      "learning_rate": 0.0002786936936936937,
+      "loss": 0.519,
+      "step": 483
+    },
+    {
+      "epoch": 0.145100243582537,
+      "grad_norm": 0.24606822431087494,
+      "learning_rate": 0.0002786486486486486,
+      "loss": 0.4956,
+      "step": 484
+    },
+    {
+      "epoch": 0.14540003747423647,
+      "grad_norm": 0.24663852155208588,
+      "learning_rate": 0.0002786036036036036,
+      "loss": 0.5029,
+      "step": 485
+    },
+    {
+      "epoch": 0.14569983136593592,
+      "grad_norm": 0.22668293118476868,
+      "learning_rate": 0.00027855855855855855,
+      "loss": 0.4987,
+      "step": 486
+    },
+    {
+      "epoch": 0.14599962525763538,
+      "grad_norm": 0.2292596995830536,
+      "learning_rate": 0.0002785135135135135,
+      "loss": 0.4808,
+      "step": 487
+    },
+    {
+      "epoch": 0.14629941914933484,
+      "grad_norm": 0.2296249270439148,
+      "learning_rate": 0.0002784684684684685,
+      "loss": 0.4733,
+      "step": 488
+    },
+    {
+      "epoch": 0.1465992130410343,
+      "grad_norm": 0.2463514357805252,
+      "learning_rate": 0.0002784234234234234,
+      "loss": 0.4856,
+      "step": 489
+    },
+    {
+      "epoch": 0.14689900693273375,
+      "grad_norm": 0.2578481435775757,
+      "learning_rate": 0.00027837837837837835,
+      "loss": 0.52,
+      "step": 490
+    },
+    {
+      "epoch": 0.1471988008244332,
+      "grad_norm": 0.22988936305046082,
+      "learning_rate": 0.00027833333333333334,
+      "loss": 0.469,
+      "step": 491
+    },
+    {
+      "epoch": 0.14749859471613266,
+      "grad_norm": 0.24489805102348328,
+      "learning_rate": 0.0002782882882882883,
+      "loss": 0.5108,
+      "step": 492
+    },
+    {
+      "epoch": 0.14779838860783212,
+      "grad_norm": 0.24594642221927643,
+      "learning_rate": 0.0002782432432432432,
+      "loss": 0.4977,
+      "step": 493
+    },
+    {
+      "epoch": 0.14809818249953158,
+      "grad_norm": 0.23341059684753418,
+      "learning_rate": 0.0002781981981981982,
+      "loss": 0.4784,
+      "step": 494
+    },
+    {
+      "epoch": 0.14839797639123103,
+      "grad_norm": 0.24211278557777405,
+      "learning_rate": 0.00027815315315315314,
+      "loss": 0.4966,
+      "step": 495
+    },
+    {
+      "epoch": 0.1486977702829305,
+      "grad_norm": 0.24049176275730133,
+      "learning_rate": 0.0002781081081081081,
+      "loss": 0.4807,
+      "step": 496
+    },
+    {
+      "epoch": 0.14899756417462995,
+      "grad_norm": 0.2326640784740448,
+      "learning_rate": 0.00027806306306306307,
+      "loss": 0.4571,
+      "step": 497
+    },
+    {
+      "epoch": 0.1492973580663294,
+      "grad_norm": 0.23826268315315247,
+      "learning_rate": 0.000278018018018018,
+      "loss": 0.4656,
+      "step": 498
+    },
+    {
+      "epoch": 0.14959715195802886,
+      "grad_norm": 0.2514077425003052,
+      "learning_rate": 0.00027797297297297294,
+      "loss": 0.5057,
+      "step": 499
+    },
+    {
+      "epoch": 0.14989694584972832,
+      "grad_norm": 0.22455474734306335,
+      "learning_rate": 0.00027792792792792793,
+      "loss": 0.4754,
+      "step": 500
+    },
+    {
+      "epoch": 0.14989694584972832,
+      "eval_loss": 0.49213707447052,
+      "eval_runtime": 564.4362,
+      "eval_samples_per_second": 3.825,
+      "eval_steps_per_second": 0.478,
+      "step": 500
+    },
+    {
+      "epoch": 0.15019673974142778,
+      "grad_norm": 0.24720892310142517,
+      "learning_rate": 0.00027788288288288287,
+      "loss": 0.5022,
+      "step": 501
+    },
+    {
+      "epoch": 0.15049653363312723,
+      "grad_norm": 0.24081699550151825,
+      "learning_rate": 0.0002778378378378378,
+      "loss": 0.4928,
+      "step": 502
+    },
+    {
+      "epoch": 0.1507963275248267,
+      "grad_norm": 0.24245139956474304,
+      "learning_rate": 0.0002777927927927928,
+      "loss": 0.5005,
+      "step": 503
+    },
+    {
+      "epoch": 0.15109612141652615,
+      "grad_norm": 0.23554831743240356,
+      "learning_rate": 0.00027774774774774774,
+      "loss": 0.4914,
+      "step": 504
+    },
+    {
+      "epoch": 0.1513959153082256,
+      "grad_norm": 0.2291078418493271,
+      "learning_rate": 0.00027770270270270267,
+      "loss": 0.4807,
+      "step": 505
+    },
+    {
+      "epoch": 0.15169570919992506,
+      "grad_norm": 0.23393088579177856,
+      "learning_rate": 0.0002776576576576576,
+      "loss": 0.496,
+      "step": 506
+    },
+    {
+      "epoch": 0.15199550309162452,
+      "grad_norm": 0.24490569531917572,
+      "learning_rate": 0.0002776126126126126,
+      "loss": 0.5,
+      "step": 507
+    },
+    {
+      "epoch": 0.15229529698332397,
+      "grad_norm": 0.21680289506912231,
+      "learning_rate": 0.00027756756756756754,
+      "loss": 0.4549,
+      "step": 508
+    },
+    {
+      "epoch": 0.15259509087502343,
+      "grad_norm": 0.22479134798049927,
+      "learning_rate": 0.0002775225225225225,
+      "loss": 0.471,
+      "step": 509
+    },
+    {
+      "epoch": 0.1528948847667229,
+      "grad_norm": 0.2381131947040558,
+      "learning_rate": 0.00027747747747747746,
+      "loss": 0.4949,
+      "step": 510
+    },
+    {
+      "epoch": 0.15319467865842235,
+      "grad_norm": 0.22718404233455658,
+      "learning_rate": 0.0002774324324324324,
+      "loss": 0.4883,
+      "step": 511
+    },
+    {
+      "epoch": 0.1534944725501218,
+      "grad_norm": 0.2359694391489029,
+      "learning_rate": 0.00027738738738738734,
+      "loss": 0.5052,
+      "step": 512
+    },
+    {
+      "epoch": 0.15379426644182126,
+      "grad_norm": 0.229795902967453,
+      "learning_rate": 0.00027734234234234233,
+      "loss": 0.4814,
+      "step": 513
+    },
+    {
+      "epoch": 0.15409406033352072,
+      "grad_norm": 0.2197142243385315,
+      "learning_rate": 0.00027729729729729727,
+      "loss": 0.4617,
+      "step": 514
+    },
+    {
+      "epoch": 0.15439385422522017,
+      "grad_norm": 0.23570996522903442,
+      "learning_rate": 0.0002772522522522522,
+      "loss": 0.4731,
+      "step": 515
+    },
+    {
+      "epoch": 0.15469364811691963,
+      "grad_norm": 0.23566411435604095,
+      "learning_rate": 0.0002772072072072072,
+      "loss": 0.4921,
+      "step": 516
+    },
+    {
+      "epoch": 0.1549934420086191,
+      "grad_norm": 0.21966999769210815,
+      "learning_rate": 0.00027716216216216213,
+      "loss": 0.4683,
+      "step": 517
+    },
+    {
+      "epoch": 0.15529323590031854,
+      "grad_norm": 0.2531338036060333,
+      "learning_rate": 0.00027711711711711707,
+      "loss": 0.5254,
+      "step": 518
+    },
+    {
+      "epoch": 0.155593029792018,
+      "grad_norm": 0.2375670224428177,
+      "learning_rate": 0.00027707207207207206,
+      "loss": 0.4988,
+      "step": 519
+    },
+    {
+      "epoch": 0.15589282368371746,
+      "grad_norm": 0.2455272376537323,
+      "learning_rate": 0.000277027027027027,
+      "loss": 0.501,
+      "step": 520
+    },
+    {
+      "epoch": 0.15619261757541691,
+      "grad_norm": 0.21289831399917603,
+      "learning_rate": 0.00027698198198198193,
+      "loss": 0.4575,
+      "step": 521
+    },
+    {
+      "epoch": 0.15649241146711637,
+      "grad_norm": 0.2653936743736267,
+      "learning_rate": 0.0002769369369369369,
+      "loss": 0.5251,
+      "step": 522
+    },
+    {
+      "epoch": 0.1567922053588158,
+      "grad_norm": 0.23822923004627228,
+      "learning_rate": 0.00027689189189189186,
+      "loss": 0.5095,
+      "step": 523
+    },
+    {
+      "epoch": 0.15709199925051526,
+      "grad_norm": 0.25067201256752014,
+      "learning_rate": 0.0002768468468468468,
+      "loss": 0.4841,
+      "step": 524
+    },
+    {
+      "epoch": 0.15739179314221471,
+      "grad_norm": 0.2340254783630371,
+      "learning_rate": 0.0002768018018018018,
+      "loss": 0.4959,
+      "step": 525
+    },
+    {
+      "epoch": 0.15769158703391417,
+      "grad_norm": 0.2431899458169937,
+      "learning_rate": 0.0002767567567567567,
+      "loss": 0.5035,
+      "step": 526
+    },
+    {
+      "epoch": 0.15799138092561363,
+      "grad_norm": 0.22817112505435944,
+      "learning_rate": 0.00027671171171171166,
+      "loss": 0.49,
+      "step": 527
+    },
+    {
+      "epoch": 0.15829117481731309,
+      "grad_norm": 0.21927404403686523,
+      "learning_rate": 0.00027666666666666665,
+      "loss": 0.4785,
+      "step": 528
+    },
+    {
+      "epoch": 0.15859096870901254,
+      "grad_norm": 0.2402762919664383,
+      "learning_rate": 0.0002766216216216216,
+      "loss": 0.4799,
+      "step": 529
+    },
+    {
+      "epoch": 0.158890762600712,
+      "grad_norm": 0.2559228241443634,
+      "learning_rate": 0.0002765765765765765,
+      "loss": 0.5065,
+      "step": 530
+    },
+    {
+      "epoch": 0.15919055649241146,
+      "grad_norm": 0.22883668541908264,
+      "learning_rate": 0.0002765315315315315,
+      "loss": 0.4797,
+      "step": 531
+    },
+    {
+      "epoch": 0.1594903503841109,
+      "grad_norm": 0.24328212440013885,
+      "learning_rate": 0.00027648648648648645,
+      "loss": 0.4796,
+      "step": 532
+    },
+    {
+      "epoch": 0.15979014427581037,
+      "grad_norm": 0.2543148398399353,
+      "learning_rate": 0.0002764414414414414,
+      "loss": 0.4785,
+      "step": 533
+    },
+    {
+      "epoch": 0.16008993816750983,
+      "grad_norm": 0.24784719944000244,
+      "learning_rate": 0.0002763963963963964,
+      "loss": 0.4981,
+      "step": 534
+    },
+    {
+      "epoch": 0.16038973205920928,
+      "grad_norm": 0.24210456013679504,
+      "learning_rate": 0.0002763513513513513,
+      "loss": 0.4939,
+      "step": 535
+    },
+    {
+      "epoch": 0.16068952595090874,
+      "grad_norm": 0.22924496233463287,
+      "learning_rate": 0.00027630630630630625,
+      "loss": 0.4524,
+      "step": 536
+    },
+    {
+      "epoch": 0.1609893198426082,
+      "grad_norm": 0.270022451877594,
+      "learning_rate": 0.00027626126126126124,
+      "loss": 0.5184,
+      "step": 537
+    },
+    {
+      "epoch": 0.16128911373430765,
+      "grad_norm": 0.2689591646194458,
+      "learning_rate": 0.0002762162162162162,
+      "loss": 0.4966,
+      "step": 538
+    },
+    {
+      "epoch": 0.1615889076260071,
+      "grad_norm": 0.23465842008590698,
+      "learning_rate": 0.00027617117117117117,
+      "loss": 0.4865,
+      "step": 539
+    },
+    {
+      "epoch": 0.16188870151770657,
+      "grad_norm": 0.23281508684158325,
+      "learning_rate": 0.0002761261261261261,
+      "loss": 0.4859,
+      "step": 540
+    },
+    {
+      "epoch": 0.16218849540940602,
+      "grad_norm": 0.25370529294013977,
+      "learning_rate": 0.00027608108108108105,
+      "loss": 0.4609,
+      "step": 541
+    },
+    {
+      "epoch": 0.16248828930110548,
+      "grad_norm": 0.2646511495113373,
+      "learning_rate": 0.00027603603603603604,
+      "loss": 0.4967,
+      "step": 542
+    },
+    {
+      "epoch": 0.16278808319280494,
+      "grad_norm": 0.22836188971996307,
+      "learning_rate": 0.000275990990990991,
+      "loss": 0.4982,
+      "step": 543
+    },
+    {
+      "epoch": 0.1630878770845044,
+      "grad_norm": 0.22948142886161804,
+      "learning_rate": 0.0002759459459459459,
+      "loss": 0.471,
+      "step": 544
+    },
+    {
+      "epoch": 0.16338767097620385,
+      "grad_norm": 0.2623734474182129,
+      "learning_rate": 0.0002759009009009009,
+      "loss": 0.4938,
+      "step": 545
+    },
+    {
+      "epoch": 0.1636874648679033,
+      "grad_norm": 0.2337695211172104,
+      "learning_rate": 0.00027585585585585584,
+      "loss": 0.4584,
+      "step": 546
+    },
+    {
+      "epoch": 0.16398725875960277,
+      "grad_norm": 0.2507021129131317,
+      "learning_rate": 0.0002758108108108108,
+      "loss": 0.5002,
+      "step": 547
+    },
+    {
+      "epoch": 0.16428705265130222,
+      "grad_norm": 0.23930178582668304,
+      "learning_rate": 0.00027576576576576577,
+      "loss": 0.4724,
+      "step": 548
+    },
+    {
+      "epoch": 0.16458684654300168,
+      "grad_norm": 0.24984320998191833,
+      "learning_rate": 0.0002757207207207207,
+      "loss": 0.4768,
+      "step": 549
+    },
+    {
+      "epoch": 0.16488664043470114,
+      "grad_norm": 0.2434365600347519,
+      "learning_rate": 0.00027567567567567564,
+      "loss": 0.4667,
+      "step": 550
+    },
+    {
+      "epoch": 0.1651864343264006,
+      "grad_norm": 0.22952896356582642,
+      "learning_rate": 0.00027563063063063063,
+      "loss": 0.4624,
+      "step": 551
+    },
+    {
+      "epoch": 0.16548622821810005,
+      "grad_norm": 0.2372165471315384,
+      "learning_rate": 0.00027558558558558557,
+      "loss": 0.477,
+      "step": 552
+    },
+    {
+      "epoch": 0.1657860221097995,
+      "grad_norm": 0.24741259217262268,
+      "learning_rate": 0.0002755405405405405,
+      "loss": 0.5077,
+      "step": 553
+    },
+    {
+      "epoch": 0.16608581600149896,
+      "grad_norm": 0.2387109249830246,
+      "learning_rate": 0.0002754954954954955,
+      "loss": 0.4817,
+      "step": 554
+    },
+    {
+      "epoch": 0.16638560989319842,
+      "grad_norm": 0.24962367117404938,
+      "learning_rate": 0.00027545045045045043,
+      "loss": 0.4775,
+      "step": 555
+    },
+    {
+      "epoch": 0.16668540378489788,
+      "grad_norm": 0.2375505119562149,
+      "learning_rate": 0.00027540540540540537,
+      "loss": 0.4843,
+      "step": 556
+    },
+    {
+      "epoch": 0.16698519767659734,
+      "grad_norm": 0.24189910292625427,
+      "learning_rate": 0.00027536036036036036,
+      "loss": 0.4952,
+      "step": 557
+    },
+    {
+      "epoch": 0.1672849915682968,
+      "grad_norm": 0.2314407229423523,
+      "learning_rate": 0.0002753153153153153,
+      "loss": 0.4676,
+      "step": 558
+    },
+    {
+      "epoch": 0.16758478545999625,
+      "grad_norm": 0.24112465977668762,
+      "learning_rate": 0.00027527027027027023,
+      "loss": 0.479,
+      "step": 559
+    },
+    {
+      "epoch": 0.1678845793516957,
+      "grad_norm": 0.22687260806560516,
+      "learning_rate": 0.0002752252252252252,
+      "loss": 0.4651,
+      "step": 560
+    },
+    {
+      "epoch": 0.16818437324339516,
+      "grad_norm": 0.23146574199199677,
+      "learning_rate": 0.00027518018018018016,
+      "loss": 0.4724,
+      "step": 561
+    },
+    {
+      "epoch": 0.16848416713509462,
+      "grad_norm": 0.23164650797843933,
+      "learning_rate": 0.0002751351351351351,
+      "loss": 0.4595,
+      "step": 562
+    },
+    {
+      "epoch": 0.16878396102679408,
+      "grad_norm": 0.2290349006652832,
+      "learning_rate": 0.0002750900900900901,
+      "loss": 0.469,
+      "step": 563
+    },
+    {
+      "epoch": 0.16908375491849353,
+      "grad_norm": 0.22324185073375702,
+      "learning_rate": 0.000275045045045045,
+      "loss": 0.4616,
+      "step": 564
+    },
+    {
+      "epoch": 0.169383548810193,
+      "grad_norm": 0.2320687472820282,
+      "learning_rate": 0.00027499999999999996,
+      "loss": 0.4706,
+      "step": 565
+    },
+    {
+      "epoch": 0.16968334270189245,
+      "grad_norm": 0.2461112141609192,
+      "learning_rate": 0.00027495495495495495,
+      "loss": 0.4886,
+      "step": 566
+    },
+    {
+      "epoch": 0.1699831365935919,
+      "grad_norm": 0.22541654109954834,
+      "learning_rate": 0.0002749099099099099,
+      "loss": 0.4676,
+      "step": 567
+    },
+    {
+      "epoch": 0.17028293048529136,
+      "grad_norm": 0.24664641916751862,
+      "learning_rate": 0.0002748648648648648,
+      "loss": 0.509,
+      "step": 568
+    },
+    {
+      "epoch": 0.17058272437699082,
+      "grad_norm": 0.23051698505878448,
+      "learning_rate": 0.0002748198198198198,
+      "loss": 0.4742,
+      "step": 569
+    },
+    {
+      "epoch": 0.17088251826869028,
+      "grad_norm": 0.21268148720264435,
+      "learning_rate": 0.00027477477477477475,
+      "loss": 0.4536,
+      "step": 570
+    },
+    {
+      "epoch": 0.17118231216038973,
+      "grad_norm": 0.25143638253211975,
+      "learning_rate": 0.0002747297297297297,
+      "loss": 0.5233,
+      "step": 571
+    },
+    {
+      "epoch": 0.1714821060520892,
+      "grad_norm": 0.21673695743083954,
+      "learning_rate": 0.0002746846846846847,
+      "loss": 0.445,
+      "step": 572
+    },
+    {
+      "epoch": 0.17178189994378865,
+      "grad_norm": 0.24307781457901,
+      "learning_rate": 0.0002746396396396396,
+      "loss": 0.493,
+      "step": 573
+    },
+    {
+      "epoch": 0.1720816938354881,
+      "grad_norm": 0.24256987869739532,
+      "learning_rate": 0.00027459459459459456,
+      "loss": 0.5249,
+      "step": 574
+    },
+    {
+      "epoch": 0.17238148772718756,
+      "grad_norm": 0.23426513373851776,
+      "learning_rate": 0.00027454954954954955,
+      "loss": 0.4956,
+      "step": 575
+    },
+    {
+      "epoch": 0.17268128161888702,
+      "grad_norm": 0.23137056827545166,
+      "learning_rate": 0.0002745045045045045,
+      "loss": 0.4909,
+      "step": 576
+    },
+    {
+      "epoch": 0.17298107551058647,
+      "grad_norm": 0.22946982085704803,
+      "learning_rate": 0.0002744594594594594,
+      "loss": 0.4733,
+      "step": 577
+    },
+    {
+      "epoch": 0.17328086940228593,
+      "grad_norm": 0.23843489587306976,
+      "learning_rate": 0.00027441441441441436,
+      "loss": 0.4944,
+      "step": 578
+    },
+    {
+      "epoch": 0.1735806632939854,
+      "grad_norm": 0.21571891009807587,
+      "learning_rate": 0.00027436936936936935,
+      "loss": 0.4488,
+      "step": 579
+    },
+    {
+      "epoch": 0.17388045718568484,
+      "grad_norm": 0.25007542967796326,
+      "learning_rate": 0.0002743243243243243,
+      "loss": 0.488,
+      "step": 580
+    },
+    {
+      "epoch": 0.1741802510773843,
+      "grad_norm": 0.24017852544784546,
+      "learning_rate": 0.0002742792792792792,
+      "loss": 0.4969,
+      "step": 581
+    },
+    {
+      "epoch": 0.17448004496908376,
+      "grad_norm": 0.23361638188362122,
+      "learning_rate": 0.0002742342342342342,
+      "loss": 0.5348,
+      "step": 582
+    },
+    {
+      "epoch": 0.17477983886078322,
+      "grad_norm": 0.22652795910835266,
+      "learning_rate": 0.00027418918918918915,
+      "loss": 0.4776,
+      "step": 583
+    },
+    {
+      "epoch": 0.17507963275248267,
+      "grad_norm": 0.23973309993743896,
+      "learning_rate": 0.0002741441441441441,
+      "loss": 0.4803,
+      "step": 584
+    },
+    {
+      "epoch": 0.17537942664418213,
+      "grad_norm": 0.23487752676010132,
+      "learning_rate": 0.0002740990990990991,
+      "loss": 0.482,
+      "step": 585
+    },
+    {
+      "epoch": 0.1756792205358816,
+      "grad_norm": 0.2464274764060974,
+      "learning_rate": 0.000274054054054054,
+      "loss": 0.5167,
+      "step": 586
+    },
+    {
+      "epoch": 0.17597901442758104,
+      "grad_norm": 0.2280922681093216,
+      "learning_rate": 0.00027400900900900895,
+      "loss": 0.4941,
+      "step": 587
+    },
+    {
+      "epoch": 0.1762788083192805,
+      "grad_norm": 0.24017813801765442,
+      "learning_rate": 0.00027396396396396394,
+      "loss": 0.4721,
+      "step": 588
+    },
+    {
+      "epoch": 0.17657860221097996,
+      "grad_norm": 0.24262118339538574,
+      "learning_rate": 0.0002739189189189189,
+      "loss": 0.5053,
+      "step": 589
+    },
+    {
+      "epoch": 0.1768783961026794,
+      "grad_norm": 0.24060070514678955,
+      "learning_rate": 0.0002738738738738738,
+      "loss": 0.4932,
+      "step": 590
+    },
+    {
+      "epoch": 0.17717818999437887,
+      "grad_norm": 0.2486894428730011,
+      "learning_rate": 0.0002738288288288288,
+      "loss": 0.4963,
+      "step": 591
+    },
+    {
+      "epoch": 0.17747798388607833,
+      "grad_norm": 0.22934255003929138,
+      "learning_rate": 0.00027378378378378374,
+      "loss": 0.4911,
+      "step": 592
+    },
+    {
+      "epoch": 0.17777777777777778,
+      "grad_norm": 0.23473136126995087,
+      "learning_rate": 0.0002737387387387387,
+      "loss": 0.4967,
+      "step": 593
+    },
+    {
+      "epoch": 0.17807757166947724,
+      "grad_norm": 0.24307146668434143,
+      "learning_rate": 0.00027369369369369367,
+      "loss": 0.5051,
+      "step": 594
+    },
+    {
+      "epoch": 0.1783773655611767,
+      "grad_norm": 0.25658494234085083,
+      "learning_rate": 0.0002736486486486486,
+      "loss": 0.4959,
+      "step": 595
+    },
+    {
+      "epoch": 0.17867715945287616,
+      "grad_norm": 0.23326924443244934,
+      "learning_rate": 0.0002736036036036036,
+      "loss": 0.5024,
+      "step": 596
+    },
+    {
+      "epoch": 0.1789769533445756,
+      "grad_norm": 0.2539668083190918,
+      "learning_rate": 0.00027355855855855854,
+      "loss": 0.5059,
+      "step": 597
+    },
+    {
+      "epoch": 0.17927674723627507,
+      "grad_norm": 0.24097499251365662,
+      "learning_rate": 0.00027351351351351347,
+      "loss": 0.4888,
+      "step": 598
+    },
+    {
+      "epoch": 0.17957654112797453,
+      "grad_norm": 0.24816173315048218,
+      "learning_rate": 0.00027346846846846846,
+      "loss": 0.4943,
+      "step": 599
+    },
+    {
+      "epoch": 0.17987633501967398,
+      "grad_norm": 0.25391021370887756,
+      "learning_rate": 0.0002734234234234234,
+      "loss": 0.4909,
+      "step": 600
+    },
+    {
+      "epoch": 0.18017612891137344,
+      "grad_norm": 0.25449302792549133,
+      "learning_rate": 0.00027337837837837834,
+      "loss": 0.499,
+      "step": 601
+    },
+    {
+      "epoch": 0.1804759228030729,
+      "grad_norm": 0.2581718862056732,
+      "learning_rate": 0.00027333333333333333,
+      "loss": 0.5093,
+      "step": 602
+    },
+    {
+      "epoch": 0.18077571669477235,
+      "grad_norm": 0.2509480118751526,
+      "learning_rate": 0.00027328828828828826,
+      "loss": 0.4974,
+      "step": 603
+    },
+    {
+      "epoch": 0.1810755105864718,
+      "grad_norm": 0.23391370475292206,
+      "learning_rate": 0.0002732432432432432,
+      "loss": 0.5176,
+      "step": 604
+    },
+    {
+      "epoch": 0.18137530447817127,
+      "grad_norm": 0.2365102618932724,
+      "learning_rate": 0.0002731981981981982,
+      "loss": 0.5021,
+      "step": 605
+    },
+    {
+      "epoch": 0.18167509836987072,
+      "grad_norm": 0.23687691986560822,
+      "learning_rate": 0.00027315315315315313,
+      "loss": 0.4728,
+      "step": 606
+    },
+    {
+      "epoch": 0.18197489226157018,
+      "grad_norm": 0.24404986202716827,
+      "learning_rate": 0.00027310810810810807,
+      "loss": 0.4842,
+      "step": 607
+    },
+    {
+      "epoch": 0.18227468615326964,
+      "grad_norm": 0.2473643720149994,
+      "learning_rate": 0.00027306306306306306,
+      "loss": 0.4823,
+      "step": 608
+    },
+    {
+      "epoch": 0.1825744800449691,
+      "grad_norm": 0.22777344286441803,
+      "learning_rate": 0.000273018018018018,
+      "loss": 0.4995,
+      "step": 609
+    },
+    {
+      "epoch": 0.18287427393666855,
+      "grad_norm": 0.22545696794986725,
+      "learning_rate": 0.000272972972972973,
+      "loss": 0.4634,
+      "step": 610
+    },
+    {
+      "epoch": 0.183174067828368,
+      "grad_norm": 0.2380336970090866,
+      "learning_rate": 0.0002729279279279279,
+      "loss": 0.4831,
+      "step": 611
+    },
+    {
+      "epoch": 0.18347386172006747,
+      "grad_norm": 0.22387194633483887,
+      "learning_rate": 0.00027288288288288286,
+      "loss": 0.4567,
+      "step": 612
+    },
+    {
+      "epoch": 0.18377365561176692,
+      "grad_norm": 0.2482718676328659,
+      "learning_rate": 0.00027283783783783785,
+      "loss": 0.4974,
+      "step": 613
+    },
+    {
+      "epoch": 0.18407344950346638,
+      "grad_norm": 0.2622338533401489,
+      "learning_rate": 0.0002727927927927928,
+      "loss": 0.5004,
+      "step": 614
+    },
+    {
+      "epoch": 0.18437324339516584,
+      "grad_norm": 0.22443141043186188,
+      "learning_rate": 0.0002727477477477477,
+      "loss": 0.4717,
+      "step": 615
+    },
+    {
+      "epoch": 0.1846730372868653,
+      "grad_norm": 0.23443573713302612,
+      "learning_rate": 0.0002727027027027027,
+      "loss": 0.5023,
+      "step": 616
+    },
+    {
+      "epoch": 0.18497283117856475,
+      "grad_norm": 0.2201087921857834,
+      "learning_rate": 0.00027265765765765765,
+      "loss": 0.4812,
+      "step": 617
+    },
+    {
+      "epoch": 0.18527262507026418,
+      "grad_norm": 0.24011440575122833,
+      "learning_rate": 0.0002726126126126126,
+      "loss": 0.4865,
+      "step": 618
+    },
+    {
+      "epoch": 0.18557241896196364,
+      "grad_norm": 0.2697189450263977,
+      "learning_rate": 0.0002725675675675676,
+      "loss": 0.5256,
+      "step": 619
+    },
+    {
+      "epoch": 0.1858722128536631,
+      "grad_norm": 0.22377456724643707,
+      "learning_rate": 0.0002725225225225225,
+      "loss": 0.4701,
+      "step": 620
+    },
+    {
+      "epoch": 0.18617200674536255,
+      "grad_norm": 0.2551979720592499,
+      "learning_rate": 0.00027247747747747745,
+      "loss": 0.4878,
+      "step": 621
+    },
+    {
+      "epoch": 0.186471800637062,
+      "grad_norm": 0.2368023544549942,
+      "learning_rate": 0.00027243243243243244,
+      "loss": 0.4768,
+      "step": 622
+    },
+    {
+      "epoch": 0.18677159452876146,
+      "grad_norm": 0.23817569017410278,
+      "learning_rate": 0.0002723873873873874,
+      "loss": 0.4914,
+      "step": 623
+    },
+    {
+      "epoch": 0.18707138842046092,
+      "grad_norm": 0.23484331369400024,
+      "learning_rate": 0.0002723423423423423,
+      "loss": 0.4896,
+      "step": 624
+    },
+    {
+      "epoch": 0.18737118231216038,
+      "grad_norm": 0.2473037838935852,
+      "learning_rate": 0.0002722972972972973,
+      "loss": 0.4841,
+      "step": 625
+    },
+    {
+      "epoch": 0.18767097620385984,
+      "grad_norm": 0.2387157678604126,
+      "learning_rate": 0.00027225225225225224,
+      "loss": 0.4913,
+      "step": 626
+    },
+    {
+      "epoch": 0.1879707700955593,
+      "grad_norm": 0.2485678642988205,
+      "learning_rate": 0.0002722072072072072,
+      "loss": 0.5155,
+      "step": 627
+    },
+    {
+      "epoch": 0.18827056398725875,
+      "grad_norm": 0.22908784449100494,
+      "learning_rate": 0.0002721621621621621,
+      "loss": 0.4874,
+      "step": 628
+    },
+    {
+      "epoch": 0.1885703578789582,
+      "grad_norm": 0.22057555615901947,
+      "learning_rate": 0.0002721171171171171,
+      "loss": 0.4846,
+      "step": 629
+    },
+    {
+      "epoch": 0.18887015177065766,
+      "grad_norm": 0.23972582817077637,
+      "learning_rate": 0.00027207207207207204,
+      "loss": 0.4785,
+      "step": 630
+    },
+    {
+      "epoch": 0.18916994566235712,
+      "grad_norm": 0.2453726977109909,
+      "learning_rate": 0.000272027027027027,
+      "loss": 0.4855,
+      "step": 631
+    },
+    {
+      "epoch": 0.18946973955405658,
+      "grad_norm": 0.23710183799266815,
+      "learning_rate": 0.00027198198198198197,
+      "loss": 0.4979,
+      "step": 632
+    },
+    {
+      "epoch": 0.18976953344575603,
+      "grad_norm": 0.24524274468421936,
+      "learning_rate": 0.0002719369369369369,
+      "loss": 0.4806,
+      "step": 633
+    },
+    {
+      "epoch": 0.1900693273374555,
+      "grad_norm": 0.22835515439510345,
+      "learning_rate": 0.00027189189189189185,
+      "loss": 0.4746,
+      "step": 634
+    },
+    {
+      "epoch": 0.19036912122915495,
+      "grad_norm": 0.23380154371261597,
+      "learning_rate": 0.00027184684684684684,
+      "loss": 0.4931,
+      "step": 635
+    },
+    {
+      "epoch": 0.1906689151208544,
+      "grad_norm": 0.22659477591514587,
+      "learning_rate": 0.0002718018018018018,
+      "loss": 0.4443,
+      "step": 636
+    },
+    {
+      "epoch": 0.19096870901255386,
+      "grad_norm": 0.22725367546081543,
+      "learning_rate": 0.0002717567567567567,
+      "loss": 0.4888,
+      "step": 637
+    },
+    {
+      "epoch": 0.19126850290425332,
+      "grad_norm": 0.233082115650177,
+      "learning_rate": 0.0002717117117117117,
+      "loss": 0.4752,
+      "step": 638
+    },
+    {
+      "epoch": 0.19156829679595277,
+      "grad_norm": 0.22560617327690125,
+      "learning_rate": 0.00027166666666666664,
+      "loss": 0.4585,
+      "step": 639
+    },
+    {
+      "epoch": 0.19186809068765223,
+      "grad_norm": 0.22963936626911163,
+      "learning_rate": 0.0002716216216216216,
+      "loss": 0.4774,
+      "step": 640
+    },
+    {
+      "epoch": 0.1921678845793517,
+      "grad_norm": 0.2543715238571167,
+      "learning_rate": 0.00027157657657657657,
+      "loss": 0.4896,
+      "step": 641
+    },
+    {
+      "epoch": 0.19246767847105115,
+      "grad_norm": 0.24594075977802277,
+      "learning_rate": 0.0002715315315315315,
+      "loss": 0.4753,
+      "step": 642
+    },
+    {
+      "epoch": 0.1927674723627506,
+      "grad_norm": 0.2333337366580963,
+      "learning_rate": 0.00027148648648648644,
+      "loss": 0.4759,
+      "step": 643
+    },
+    {
+      "epoch": 0.19306726625445006,
+      "grad_norm": 0.23100800812244415,
+      "learning_rate": 0.00027144144144144143,
+      "loss": 0.4536,
+      "step": 644
+    },
+    {
+      "epoch": 0.19336706014614952,
+      "grad_norm": 0.26426073908805847,
+      "learning_rate": 0.00027139639639639637,
+      "loss": 0.4882,
+      "step": 645
+    },
+    {
+      "epoch": 0.19366685403784897,
+      "grad_norm": 0.22670197486877441,
+      "learning_rate": 0.0002713513513513513,
+      "loss": 0.473,
+      "step": 646
+    },
+    {
+      "epoch": 0.19396664792954843,
+      "grad_norm": 0.23645341396331787,
+      "learning_rate": 0.0002713063063063063,
+      "loss": 0.4964,
+      "step": 647
+    },
+    {
+      "epoch": 0.1942664418212479,
+      "grad_norm": 0.2535463869571686,
+      "learning_rate": 0.00027126126126126123,
+      "loss": 0.5001,
+      "step": 648
+    },
+    {
+      "epoch": 0.19456623571294734,
+      "grad_norm": 0.25858014822006226,
+      "learning_rate": 0.00027121621621621617,
+      "loss": 0.5007,
+      "step": 649
+    },
+    {
+      "epoch": 0.1948660296046468,
+      "grad_norm": 0.2457359880208969,
+      "learning_rate": 0.0002711711711711711,
+      "loss": 0.4849,
+      "step": 650
+    },
+    {
+      "epoch": 0.19516582349634626,
+      "grad_norm": 0.25070154666900635,
+      "learning_rate": 0.0002711261261261261,
+      "loss": 0.4888,
+      "step": 651
+    },
+    {
+      "epoch": 0.19546561738804571,
+      "grad_norm": 0.22972378134727478,
+      "learning_rate": 0.00027108108108108103,
+      "loss": 0.4719,
+      "step": 652
+    },
+    {
+      "epoch": 0.19576541127974517,
+      "grad_norm": 0.23261764645576477,
+      "learning_rate": 0.000271036036036036,
+      "loss": 0.4633,
+      "step": 653
+    },
+    {
+      "epoch": 0.19606520517144463,
+      "grad_norm": 0.25028276443481445,
+      "learning_rate": 0.00027099099099099096,
+      "loss": 0.5209,
+      "step": 654
+    },
+    {
+      "epoch": 0.19636499906314409,
+      "grad_norm": 0.23486949503421783,
+      "learning_rate": 0.0002709459459459459,
+      "loss": 0.4757,
+      "step": 655
+    },
+    {
+      "epoch": 0.19666479295484354,
+      "grad_norm": 0.2393907606601715,
+      "learning_rate": 0.0002709009009009009,
+      "loss": 0.464,
+      "step": 656
+    },
+    {
+      "epoch": 0.196964586846543,
+      "grad_norm": 0.24317681789398193,
+      "learning_rate": 0.0002708558558558558,
+      "loss": 0.4843,
+      "step": 657
+    },
+    {
+      "epoch": 0.19726438073824246,
+      "grad_norm": 0.23742252588272095,
+      "learning_rate": 0.00027081081081081076,
+      "loss": 0.5025,
+      "step": 658
+    },
+    {
+      "epoch": 0.1975641746299419,
+      "grad_norm": 0.23426878452301025,
+      "learning_rate": 0.00027076576576576575,
+      "loss": 0.4776,
+      "step": 659
+    },
+    {
+      "epoch": 0.19786396852164137,
+      "grad_norm": 0.250949501991272,
+      "learning_rate": 0.0002707207207207207,
+      "loss": 0.4968,
+      "step": 660
+    },
+    {
+      "epoch": 0.19816376241334083,
+      "grad_norm": 0.23000746965408325,
+      "learning_rate": 0.0002706756756756756,
+      "loss": 0.4895,
+      "step": 661
+    },
+    {
+      "epoch": 0.19846355630504028,
+      "grad_norm": 0.26243215799331665,
+      "learning_rate": 0.0002706306306306306,
+      "loss": 0.5076,
+      "step": 662
+    },
+    {
+      "epoch": 0.19876335019673974,
+      "grad_norm": 0.22115159034729004,
+      "learning_rate": 0.00027058558558558555,
+      "loss": 0.4823,
+      "step": 663
+    },
+    {
+      "epoch": 0.1990631440884392,
+      "grad_norm": 0.22618655860424042,
+      "learning_rate": 0.0002705405405405405,
+      "loss": 0.4994,
+      "step": 664
+    },
+    {
+      "epoch": 0.19936293798013865,
+      "grad_norm": 0.22989815473556519,
+      "learning_rate": 0.0002704954954954955,
+      "loss": 0.4701,
+      "step": 665
+    },
+    {
+      "epoch": 0.1996627318718381,
+      "grad_norm": 0.24214977025985718,
+      "learning_rate": 0.0002704504504504504,
+      "loss": 0.4806,
+      "step": 666
+    },
+    {
+      "epoch": 0.19996252576353757,
+      "grad_norm": 0.21489010751247406,
+      "learning_rate": 0.0002704054054054054,
+      "loss": 0.4511,
+      "step": 667
+    },
+    {
+      "epoch": 0.20026231965523703,
+      "grad_norm": 0.2397059053182602,
+      "learning_rate": 0.00027036036036036035,
+      "loss": 0.4427,
+      "step": 668
+    },
+    {
+      "epoch": 0.20056211354693648,
+      "grad_norm": 0.2419203370809555,
+      "learning_rate": 0.0002703153153153153,
+      "loss": 0.4993,
+      "step": 669
+    },
+    {
+      "epoch": 0.20086190743863594,
+      "grad_norm": 0.24709810316562653,
+      "learning_rate": 0.0002702702702702703,
+      "loss": 0.5195,
+      "step": 670
+    },
+    {
+      "epoch": 0.2011617013303354,
+      "grad_norm": 0.25068792700767517,
+      "learning_rate": 0.0002702252252252252,
+      "loss": 0.508,
+      "step": 671
+    },
+    {
+      "epoch": 0.20146149522203485,
+      "grad_norm": 0.210756316781044,
+      "learning_rate": 0.00027018018018018015,
+      "loss": 0.4413,
+      "step": 672
+    },
+    {
+      "epoch": 0.2017612891137343,
+      "grad_norm": 0.2557854652404785,
+      "learning_rate": 0.00027013513513513514,
+      "loss": 0.4971,
+      "step": 673
+    },
+    {
+      "epoch": 0.20206108300543377,
+      "grad_norm": 0.23103776574134827,
+      "learning_rate": 0.0002700900900900901,
+      "loss": 0.4683,
+      "step": 674
+    },
+    {
+      "epoch": 0.20236087689713322,
+      "grad_norm": 0.21560733020305634,
+      "learning_rate": 0.000270045045045045,
+      "loss": 0.4443,
+      "step": 675
+    },
+    {
+      "epoch": 0.20266067078883268,
+      "grad_norm": 0.2477121204137802,
+      "learning_rate": 0.00027,
+      "loss": 0.5131,
+      "step": 676
+    },
+    {
+      "epoch": 0.20296046468053214,
+      "grad_norm": 0.24966078996658325,
+      "learning_rate": 0.00026995495495495494,
+      "loss": 0.4856,
+      "step": 677
+    },
+    {
+      "epoch": 0.2032602585722316,
+      "grad_norm": 0.23841539025306702,
+      "learning_rate": 0.0002699099099099099,
+      "loss": 0.4658,
+      "step": 678
+    },
+    {
+      "epoch": 0.20356005246393105,
+      "grad_norm": 0.2685762047767639,
+      "learning_rate": 0.00026986486486486487,
+      "loss": 0.5096,
+      "step": 679
+    },
+    {
+      "epoch": 0.2038598463556305,
+      "grad_norm": 0.25834083557128906,
+      "learning_rate": 0.0002698198198198198,
+      "loss": 0.5036,
+      "step": 680
+    },
+    {
+      "epoch": 0.20415964024732997,
+      "grad_norm": 0.2324528843164444,
+      "learning_rate": 0.00026977477477477474,
+      "loss": 0.4618,
+      "step": 681
+    },
+    {
+      "epoch": 0.20445943413902942,
+      "grad_norm": 0.22903920710086823,
+      "learning_rate": 0.00026972972972972973,
+      "loss": 0.4662,
+      "step": 682
+    },
+    {
+      "epoch": 0.20475922803072888,
+      "grad_norm": 0.24908147752285004,
+      "learning_rate": 0.00026968468468468467,
+      "loss": 0.4684,
+      "step": 683
+    },
+    {
+      "epoch": 0.20505902192242834,
+      "grad_norm": 0.2278299629688263,
+      "learning_rate": 0.0002696396396396396,
+      "loss": 0.4447,
+      "step": 684
+    },
+    {
+      "epoch": 0.2053588158141278,
+      "grad_norm": 0.2315731793642044,
+      "learning_rate": 0.0002695945945945946,
+      "loss": 0.4875,
+      "step": 685
+    },
+    {
+      "epoch": 0.20565860970582725,
+      "grad_norm": 0.23152673244476318,
+      "learning_rate": 0.00026954954954954953,
+      "loss": 0.4796,
+      "step": 686
+    },
+    {
+      "epoch": 0.2059584035975267,
+      "grad_norm": 0.23902982473373413,
+      "learning_rate": 0.00026950450450450447,
+      "loss": 0.5169,
+      "step": 687
+    },
+    {
+      "epoch": 0.20625819748922616,
+      "grad_norm": 0.23636193573474884,
+      "learning_rate": 0.00026945945945945946,
+      "loss": 0.493,
+      "step": 688
+    },
+    {
+      "epoch": 0.20655799138092562,
+      "grad_norm": 0.21632736921310425,
+      "learning_rate": 0.0002694144144144144,
+      "loss": 0.4594,
+      "step": 689
+    },
+    {
+      "epoch": 0.20685778527262508,
+      "grad_norm": 0.2258147895336151,
+      "learning_rate": 0.00026936936936936934,
+      "loss": 0.4625,
+      "step": 690
+    },
+    {
+      "epoch": 0.20715757916432453,
+      "grad_norm": 0.21552099287509918,
+      "learning_rate": 0.0002693243243243243,
+      "loss": 0.4481,
+      "step": 691
+    },
+    {
+      "epoch": 0.207457373056024,
+      "grad_norm": 0.23030760884284973,
+      "learning_rate": 0.00026927927927927926,
+      "loss": 0.4644,
+      "step": 692
+    },
+    {
+      "epoch": 0.20775716694772345,
+      "grad_norm": 0.23163190484046936,
+      "learning_rate": 0.0002692342342342342,
+      "loss": 0.4483,
+      "step": 693
+    },
+    {
+      "epoch": 0.2080569608394229,
+      "grad_norm": 0.2412249743938446,
+      "learning_rate": 0.0002691891891891892,
+      "loss": 0.4886,
+      "step": 694
+    },
+    {
+      "epoch": 0.20835675473112236,
+      "grad_norm": 0.23279330134391785,
+      "learning_rate": 0.00026914414414414413,
+      "loss": 0.4733,
+      "step": 695
+    },
+    {
+      "epoch": 0.20865654862282182,
+      "grad_norm": 0.2269987165927887,
+      "learning_rate": 0.00026909909909909906,
+      "loss": 0.4866,
+      "step": 696
+    },
+    {
+      "epoch": 0.20895634251452128,
+      "grad_norm": 0.23355835676193237,
+      "learning_rate": 0.00026905405405405406,
+      "loss": 0.4918,
+      "step": 697
+    },
+    {
+      "epoch": 0.20925613640622073,
+      "grad_norm": 0.26988187432289124,
+      "learning_rate": 0.000269009009009009,
+      "loss": 0.4953,
+      "step": 698
+    },
+    {
+      "epoch": 0.2095559302979202,
+      "grad_norm": 0.22978806495666504,
+      "learning_rate": 0.00026896396396396393,
+      "loss": 0.4622,
+      "step": 699
+    },
+    {
+      "epoch": 0.20985572418961965,
+      "grad_norm": 0.2823212146759033,
+      "learning_rate": 0.00026891891891891887,
+      "loss": 0.5105,
+      "step": 700
+    },
+    {
+      "epoch": 0.2101555180813191,
+      "grad_norm": 0.23818424344062805,
+      "learning_rate": 0.00026887387387387386,
+      "loss": 0.479,
+      "step": 701
+    },
+    {
+      "epoch": 0.21045531197301856,
+      "grad_norm": 0.23730318248271942,
+      "learning_rate": 0.0002688288288288288,
+      "loss": 0.4991,
+      "step": 702
+    },
+    {
+      "epoch": 0.21075510586471802,
+      "grad_norm": 0.23275551199913025,
+      "learning_rate": 0.00026878378378378373,
+      "loss": 0.466,
+      "step": 703
+    },
+    {
+      "epoch": 0.21105489975641747,
+      "grad_norm": 0.2296077013015747,
+      "learning_rate": 0.0002687387387387387,
+      "loss": 0.4964,
+      "step": 704
+    },
+    {
+      "epoch": 0.21135469364811693,
+      "grad_norm": 0.24341174960136414,
+      "learning_rate": 0.00026869369369369366,
+      "loss": 0.505,
+      "step": 705
+    },
+    {
+      "epoch": 0.2116544875398164,
+      "grad_norm": 0.22542104125022888,
+      "learning_rate": 0.0002686486486486486,
+      "loss": 0.4533,
+      "step": 706
+    },
+    {
+      "epoch": 0.21195428143151585,
+      "grad_norm": 0.22414691746234894,
+      "learning_rate": 0.0002686036036036036,
+      "loss": 0.4504,
+      "step": 707
+    },
+    {
+      "epoch": 0.2122540753232153,
+      "grad_norm": 0.2119213044643402,
+      "learning_rate": 0.0002685585585585585,
+      "loss": 0.4457,
+      "step": 708
+    },
+    {
+      "epoch": 0.21255386921491476,
+      "grad_norm": 0.21771720051765442,
+      "learning_rate": 0.00026851351351351346,
+      "loss": 0.4411,
+      "step": 709
+    },
+    {
+      "epoch": 0.21285366310661422,
+      "grad_norm": 0.2484733760356903,
+      "learning_rate": 0.00026846846846846845,
+      "loss": 0.5016,
+      "step": 710
+    },
+    {
+      "epoch": 0.21315345699831367,
+      "grad_norm": 0.23043473064899445,
+      "learning_rate": 0.0002684234234234234,
+      "loss": 0.4717,
+      "step": 711
+    },
+    {
+      "epoch": 0.21345325089001313,
+      "grad_norm": 0.2769162952899933,
+      "learning_rate": 0.0002683783783783783,
+      "loss": 0.5135,
+      "step": 712
+    },
+    {
+      "epoch": 0.21375304478171256,
+      "grad_norm": 0.22561033070087433,
+      "learning_rate": 0.0002683333333333333,
+      "loss": 0.4928,
+      "step": 713
+    },
+    {
+      "epoch": 0.21405283867341202,
+      "grad_norm": 0.2365075945854187,
+      "learning_rate": 0.00026828828828828825,
+      "loss": 0.4843,
+      "step": 714
+    },
+    {
+      "epoch": 0.21435263256511147,
+      "grad_norm": 0.23628349602222443,
+      "learning_rate": 0.0002682432432432432,
+      "loss": 0.4782,
+      "step": 715
+    },
+    {
+      "epoch": 0.21465242645681093,
+      "grad_norm": 0.22449685633182526,
+      "learning_rate": 0.0002681981981981982,
+      "loss": 0.4533,
+      "step": 716
+    },
+    {
+      "epoch": 0.2149522203485104,
+      "grad_norm": 0.23336151242256165,
+      "learning_rate": 0.0002681531531531531,
+      "loss": 0.4818,
+      "step": 717
+    },
+    {
+      "epoch": 0.21525201424020984,
+      "grad_norm": 0.2387206107378006,
+      "learning_rate": 0.00026810810810810805,
+      "loss": 0.4537,
+      "step": 718
+    },
+    {
+      "epoch": 0.2155518081319093,
+      "grad_norm": 0.23359011113643646,
+      "learning_rate": 0.00026806306306306304,
+      "loss": 0.4895,
+      "step": 719
+    },
+    {
+      "epoch": 0.21585160202360876,
+      "grad_norm": 0.240494042634964,
+      "learning_rate": 0.000268018018018018,
+      "loss": 0.4878,
+      "step": 720
+    },
+    {
+      "epoch": 0.21615139591530821,
+      "grad_norm": 0.23335424065589905,
+      "learning_rate": 0.0002679729729729729,
+      "loss": 0.4747,
+      "step": 721
+    },
+    {
+      "epoch": 0.21645118980700767,
+      "grad_norm": 0.2620643079280853,
+      "learning_rate": 0.0002679279279279279,
+      "loss": 0.4968,
+      "step": 722
+    },
+    {
+      "epoch": 0.21675098369870713,
+      "grad_norm": 0.2350034862756729,
+      "learning_rate": 0.00026788288288288284,
+      "loss": 0.4801,
+      "step": 723
+    },
+    {
+      "epoch": 0.21705077759040659,
+      "grad_norm": 0.2358752340078354,
+      "learning_rate": 0.00026783783783783784,
+      "loss": 0.5265,
+      "step": 724
+    },
+    {
+      "epoch": 0.21735057148210604,
+      "grad_norm": 0.2392471730709076,
+      "learning_rate": 0.00026779279279279277,
+      "loss": 0.4667,
+      "step": 725
+    },
+    {
+      "epoch": 0.2176503653738055,
+      "grad_norm": 0.23733973503112793,
+      "learning_rate": 0.0002677477477477477,
+      "loss": 0.4712,
+      "step": 726
+    },
+    {
+      "epoch": 0.21795015926550496,
+      "grad_norm": 0.2224283516407013,
+      "learning_rate": 0.0002677027027027027,
+      "loss": 0.4519,
+      "step": 727
+    },
+    {
+      "epoch": 0.2182499531572044,
+      "grad_norm": 0.22749215364456177,
+      "learning_rate": 0.00026765765765765764,
+      "loss": 0.4656,
+      "step": 728
+    },
+    {
+      "epoch": 0.21854974704890387,
+      "grad_norm": 0.29321786761283875,
+      "learning_rate": 0.0002676126126126126,
+      "loss": 0.5085,
+      "step": 729
+    },
+    {
+      "epoch": 0.21884954094060333,
+      "grad_norm": 0.23674741387367249,
+      "learning_rate": 0.00026756756756756756,
+      "loss": 0.4958,
+      "step": 730
+    },
+    {
+      "epoch": 0.21914933483230278,
+      "grad_norm": 0.21558566391468048,
+      "learning_rate": 0.0002675225225225225,
+      "loss": 0.4679,
+      "step": 731
+    },
+    {
+      "epoch": 0.21944912872400224,
+      "grad_norm": 0.2383924126625061,
+      "learning_rate": 0.00026747747747747744,
+      "loss": 0.4827,
+      "step": 732
+    },
+    {
+      "epoch": 0.2197489226157017,
+      "grad_norm": 0.23788924515247345,
+      "learning_rate": 0.00026743243243243243,
+      "loss": 0.486,
+      "step": 733
+    },
+    {
+      "epoch": 0.22004871650740115,
+      "grad_norm": 0.23550404608249664,
+      "learning_rate": 0.00026738738738738737,
+      "loss": 0.4603,
+      "step": 734
+    },
+    {
+      "epoch": 0.2203485103991006,
+      "grad_norm": 0.2342066466808319,
+      "learning_rate": 0.00026734234234234236,
+      "loss": 0.4591,
+      "step": 735
+    },
+    {
+      "epoch": 0.22064830429080007,
+      "grad_norm": 0.25759053230285645,
+      "learning_rate": 0.0002672972972972973,
+      "loss": 0.483,
+      "step": 736
+    },
+    {
+      "epoch": 0.22094809818249952,
+      "grad_norm": 0.22325725853443146,
+      "learning_rate": 0.00026725225225225223,
+      "loss": 0.4609,
+      "step": 737
+    },
+    {
+      "epoch": 0.22124789207419898,
+      "grad_norm": 0.22235055267810822,
+      "learning_rate": 0.0002672072072072072,
+      "loss": 0.4512,
+      "step": 738
+    },
+    {
+      "epoch": 0.22154768596589844,
+      "grad_norm": 0.23441246151924133,
+      "learning_rate": 0.00026716216216216216,
+      "loss": 0.4517,
+      "step": 739
+    },
+    {
+      "epoch": 0.2218474798575979,
+      "grad_norm": 0.2520740330219269,
+      "learning_rate": 0.0002671171171171171,
+      "loss": 0.4712,
+      "step": 740
+    },
+    {
+      "epoch": 0.22214727374929735,
+      "grad_norm": 0.22782452404499054,
+      "learning_rate": 0.0002670720720720721,
+      "loss": 0.4723,
+      "step": 741
+    },
+    {
+      "epoch": 0.2224470676409968,
+      "grad_norm": 0.2406499981880188,
+      "learning_rate": 0.000267027027027027,
+      "loss": 0.4909,
+      "step": 742
+    },
+    {
+      "epoch": 0.22274686153269627,
+      "grad_norm": 0.21733756363391876,
+      "learning_rate": 0.00026698198198198196,
+      "loss": 0.4402,
+      "step": 743
+    },
+    {
+      "epoch": 0.22304665542439572,
+      "grad_norm": 0.2329728901386261,
+      "learning_rate": 0.00026693693693693695,
+      "loss": 0.4659,
+      "step": 744
+    },
+    {
+      "epoch": 0.22334644931609518,
+      "grad_norm": 0.23359104990959167,
+      "learning_rate": 0.0002668918918918919,
+      "loss": 0.4848,
+      "step": 745
+    },
+    {
+      "epoch": 0.22364624320779464,
+      "grad_norm": 0.23723845183849335,
+      "learning_rate": 0.0002668468468468468,
+      "loss": 0.4674,
+      "step": 746
+    },
+    {
+      "epoch": 0.2239460370994941,
+      "grad_norm": 0.2128835916519165,
+      "learning_rate": 0.00026680180180180176,
+      "loss": 0.4617,
+      "step": 747
+    },
+    {
+      "epoch": 0.22424583099119355,
+      "grad_norm": 0.2343822568655014,
+      "learning_rate": 0.00026675675675675675,
+      "loss": 0.4624,
+      "step": 748
+    },
+    {
+      "epoch": 0.224545624882893,
+      "grad_norm": 0.24932916462421417,
+      "learning_rate": 0.0002667117117117117,
+      "loss": 0.476,
+      "step": 749
+    },
+    {
+      "epoch": 0.22484541877459246,
+      "grad_norm": 0.24181415140628815,
+      "learning_rate": 0.0002666666666666666,
+      "loss": 0.4681,
+      "step": 750
+    },
+    {
+      "epoch": 0.22514521266629192,
+      "grad_norm": 0.23665620386600494,
+      "learning_rate": 0.0002666216216216216,
+      "loss": 0.4858,
+      "step": 751
+    },
+    {
+      "epoch": 0.22544500655799138,
+      "grad_norm": 0.24904295802116394,
+      "learning_rate": 0.00026657657657657655,
+      "loss": 0.4957,
+      "step": 752
+    },
+    {
+      "epoch": 0.22574480044969084,
+      "grad_norm": 0.2285979986190796,
+      "learning_rate": 0.0002665315315315315,
+      "loss": 0.4515,
+      "step": 753
+    },
+    {
+      "epoch": 0.2260445943413903,
+      "grad_norm": 0.2505464553833008,
+      "learning_rate": 0.0002664864864864865,
+      "loss": 0.4884,
+      "step": 754
+    },
+    {
+      "epoch": 0.22634438823308975,
+      "grad_norm": 0.22328858077526093,
+      "learning_rate": 0.0002664414414414414,
+      "loss": 0.4463,
+      "step": 755
+    },
+    {
+      "epoch": 0.2266441821247892,
+      "grad_norm": 0.2543044984340668,
+      "learning_rate": 0.00026639639639639635,
+      "loss": 0.493,
+      "step": 756
+    },
+    {
+      "epoch": 0.22694397601648866,
+      "grad_norm": 0.2348204255104065,
+      "learning_rate": 0.00026635135135135135,
+      "loss": 0.4511,
+      "step": 757
+    },
+    {
+      "epoch": 0.22724376990818812,
+      "grad_norm": 0.25663718581199646,
+      "learning_rate": 0.0002663063063063063,
+      "loss": 0.5006,
+      "step": 758
+    },
+    {
+      "epoch": 0.22754356379988758,
+      "grad_norm": 0.24245639145374298,
+      "learning_rate": 0.0002662612612612612,
+      "loss": 0.4687,
+      "step": 759
+    },
+    {
+      "epoch": 0.22784335769158703,
+      "grad_norm": 0.2461511641740799,
+      "learning_rate": 0.0002662162162162162,
+      "loss": 0.5136,
+      "step": 760
+    },
+    {
+      "epoch": 0.2281431515832865,
+      "grad_norm": 0.22325679659843445,
+      "learning_rate": 0.00026617117117117115,
+      "loss": 0.472,
+      "step": 761
+    },
+    {
+      "epoch": 0.22844294547498595,
+      "grad_norm": 0.2652730345726013,
+      "learning_rate": 0.0002661261261261261,
+      "loss": 0.4732,
+      "step": 762
+    },
+    {
+      "epoch": 0.2287427393666854,
+      "grad_norm": 0.24870134890079498,
+      "learning_rate": 0.0002660810810810811,
+      "loss": 0.467,
+      "step": 763
+    },
+    {
+      "epoch": 0.22904253325838486,
+      "grad_norm": 0.23315280675888062,
+      "learning_rate": 0.000266036036036036,
+      "loss": 0.4994,
+      "step": 764
+    },
+    {
+      "epoch": 0.22934232715008432,
+      "grad_norm": 0.23781219124794006,
+      "learning_rate": 0.00026599099099099095,
+      "loss": 0.4923,
+      "step": 765
+    },
+    {
+      "epoch": 0.22964212104178378,
+      "grad_norm": 0.2272898256778717,
+      "learning_rate": 0.00026594594594594594,
+      "loss": 0.4683,
+      "step": 766
+    },
+    {
+      "epoch": 0.22994191493348323,
+      "grad_norm": 0.2321631759405136,
+      "learning_rate": 0.0002659009009009009,
+      "loss": 0.5018,
+      "step": 767
+    },
+    {
+      "epoch": 0.2302417088251827,
+      "grad_norm": 0.22698219120502472,
+      "learning_rate": 0.0002658558558558558,
+      "loss": 0.479,
+      "step": 768
+    },
+    {
+      "epoch": 0.23054150271688215,
+      "grad_norm": 0.23421627283096313,
+      "learning_rate": 0.0002658108108108108,
+      "loss": 0.4613,
+      "step": 769
+    },
+    {
+      "epoch": 0.2308412966085816,
+      "grad_norm": 0.21950644254684448,
+      "learning_rate": 0.00026576576576576574,
+      "loss": 0.4497,
+      "step": 770
+    },
+    {
+      "epoch": 0.23114109050028106,
+      "grad_norm": 0.2207535058259964,
+      "learning_rate": 0.0002657207207207207,
+      "loss": 0.4777,
+      "step": 771
+    },
+    {
+      "epoch": 0.23144088439198052,
+      "grad_norm": 0.22216112911701202,
+      "learning_rate": 0.0002656756756756756,
+      "loss": 0.4682,
+      "step": 772
+    },
+    {
+      "epoch": 0.23174067828367997,
+      "grad_norm": 0.2619054317474365,
+      "learning_rate": 0.0002656306306306306,
+      "loss": 0.4791,
+      "step": 773
+    },
+    {
+      "epoch": 0.23204047217537943,
+      "grad_norm": 0.2443225234746933,
+      "learning_rate": 0.00026558558558558554,
+      "loss": 0.4777,
+      "step": 774
+    },
+    {
+      "epoch": 0.2323402660670789,
+      "grad_norm": 0.21427664160728455,
+      "learning_rate": 0.0002655405405405405,
+      "loss": 0.4407,
+      "step": 775
+    },
+    {
+      "epoch": 0.23264005995877834,
+      "grad_norm": 0.21477638185024261,
+      "learning_rate": 0.00026549549549549547,
+      "loss": 0.4333,
+      "step": 776
+    },
+    {
+      "epoch": 0.2329398538504778,
+      "grad_norm": 0.23390546441078186,
+      "learning_rate": 0.0002654504504504504,
+      "loss": 0.4839,
+      "step": 777
+    },
+    {
+      "epoch": 0.23323964774217726,
+      "grad_norm": 0.2529938220977783,
+      "learning_rate": 0.00026540540540540534,
+      "loss": 0.4701,
+      "step": 778
+    },
+    {
+      "epoch": 0.23353944163387672,
+      "grad_norm": 0.24290771782398224,
+      "learning_rate": 0.00026536036036036033,
+      "loss": 0.492,
+      "step": 779
+    },
+    {
+      "epoch": 0.23383923552557617,
+      "grad_norm": 0.2573592960834503,
+      "learning_rate": 0.00026531531531531527,
+      "loss": 0.5105,
+      "step": 780
+    },
+    {
+      "epoch": 0.23413902941727563,
+      "grad_norm": 0.25404611229896545,
+      "learning_rate": 0.00026527027027027026,
+      "loss": 0.5054,
+      "step": 781
+    },
+    {
+      "epoch": 0.2344388233089751,
+      "grad_norm": 0.2394997775554657,
+      "learning_rate": 0.0002652252252252252,
+      "loss": 0.4786,
+      "step": 782
+    },
+    {
+      "epoch": 0.23473861720067454,
+      "grad_norm": 0.2353266179561615,
+      "learning_rate": 0.00026518018018018013,
+      "loss": 0.4613,
+      "step": 783
+    },
+    {
+      "epoch": 0.235038411092374,
+      "grad_norm": 0.22989432513713837,
+      "learning_rate": 0.0002651351351351351,
+      "loss": 0.4505,
+      "step": 784
+    },
+    {
+      "epoch": 0.23533820498407346,
+      "grad_norm": 0.21917951107025146,
+      "learning_rate": 0.00026509009009009006,
+      "loss": 0.438,
+      "step": 785
+    },
+    {
+      "epoch": 0.2356379988757729,
+      "grad_norm": 0.23011858761310577,
+      "learning_rate": 0.000265045045045045,
+      "loss": 0.4477,
+      "step": 786
+    },
+    {
+      "epoch": 0.23593779276747237,
+      "grad_norm": 0.22732798755168915,
+      "learning_rate": 0.000265,
+      "loss": 0.454,
+      "step": 787
+    },
+    {
+      "epoch": 0.23623758665917183,
+      "grad_norm": 0.22975054383277893,
+      "learning_rate": 0.0002649549549549549,
+      "loss": 0.4419,
+      "step": 788
+    },
+    {
+      "epoch": 0.23653738055087128,
+      "grad_norm": 0.25520968437194824,
+      "learning_rate": 0.00026490990990990986,
+      "loss": 0.4987,
+      "step": 789
+    },
+    {
+      "epoch": 0.23683717444257074,
+      "grad_norm": 0.2375541776418686,
+      "learning_rate": 0.00026486486486486485,
+      "loss": 0.473,
+      "step": 790
+    },
+    {
+      "epoch": 0.2371369683342702,
+      "grad_norm": 0.2304588407278061,
+      "learning_rate": 0.0002648198198198198,
+      "loss": 0.4734,
+      "step": 791
+    },
+    {
+      "epoch": 0.23743676222596966,
+      "grad_norm": 0.22878654301166534,
+      "learning_rate": 0.0002647747747747748,
+      "loss": 0.4684,
+      "step": 792
+    },
+    {
+      "epoch": 0.2377365561176691,
+      "grad_norm": 0.25825339555740356,
+      "learning_rate": 0.0002647297297297297,
+      "loss": 0.5006,
+      "step": 793
+    },
+    {
+      "epoch": 0.23803635000936857,
+      "grad_norm": 0.2332850843667984,
+      "learning_rate": 0.00026468468468468466,
+      "loss": 0.4603,
+      "step": 794
+    },
+    {
+      "epoch": 0.23833614390106803,
+      "grad_norm": 0.23115694522857666,
+      "learning_rate": 0.00026463963963963965,
+      "loss": 0.4879,
+      "step": 795
+    },
+    {
+      "epoch": 0.23863593779276748,
+      "grad_norm": 0.2409309297800064,
+      "learning_rate": 0.0002645945945945946,
+      "loss": 0.4662,
+      "step": 796
+    },
+    {
+      "epoch": 0.23893573168446694,
+      "grad_norm": 0.23094283044338226,
+      "learning_rate": 0.0002645495495495495,
+      "loss": 0.4477,
+      "step": 797
+    },
+    {
+      "epoch": 0.2392355255761664,
+      "grad_norm": 0.2324245125055313,
+      "learning_rate": 0.0002645045045045045,
+      "loss": 0.4539,
+      "step": 798
+    },
+    {
+      "epoch": 0.23953531946786585,
+      "grad_norm": 0.23273488879203796,
+      "learning_rate": 0.00026445945945945945,
+      "loss": 0.4926,
+      "step": 799
+    },
+    {
+      "epoch": 0.2398351133595653,
+      "grad_norm": 0.2254081666469574,
+      "learning_rate": 0.0002644144144144144,
+      "loss": 0.4487,
+      "step": 800
+    },
+    {
+      "epoch": 0.24013490725126477,
+      "grad_norm": 0.22009852528572083,
+      "learning_rate": 0.0002643693693693694,
+      "loss": 0.4373,
+      "step": 801
+    },
+    {
+      "epoch": 0.24043470114296422,
+      "grad_norm": 0.24840936064720154,
+      "learning_rate": 0.0002643243243243243,
+      "loss": 0.4803,
+      "step": 802
+    },
+    {
+      "epoch": 0.24073449503466368,
+      "grad_norm": 0.2305980920791626,
+      "learning_rate": 0.00026427927927927925,
+      "loss": 0.4727,
+      "step": 803
+    },
+    {
+      "epoch": 0.24103428892636314,
+      "grad_norm": 0.23277850449085236,
+      "learning_rate": 0.00026423423423423424,
+      "loss": 0.4775,
+      "step": 804
+    },
+    {
+      "epoch": 0.2413340828180626,
+      "grad_norm": 0.24016259610652924,
+      "learning_rate": 0.0002641891891891892,
+      "loss": 0.5002,
+      "step": 805
+    },
+    {
+      "epoch": 0.24163387670976205,
+      "grad_norm": 0.239017054438591,
+      "learning_rate": 0.0002641441441441441,
+      "loss": 0.4586,
+      "step": 806
+    },
+    {
+      "epoch": 0.2419336706014615,
+      "grad_norm": 0.23575717210769653,
+      "learning_rate": 0.0002640990990990991,
+      "loss": 0.481,
+      "step": 807
+    },
+    {
+      "epoch": 0.24223346449316094,
+      "grad_norm": 0.23028531670570374,
+      "learning_rate": 0.00026405405405405404,
+      "loss": 0.4546,
+      "step": 808
+    },
+    {
+      "epoch": 0.2425332583848604,
+      "grad_norm": 0.23798401653766632,
+      "learning_rate": 0.000264009009009009,
+      "loss": 0.4806,
+      "step": 809
+    },
+    {
+      "epoch": 0.24283305227655985,
+      "grad_norm": 0.23191827535629272,
+      "learning_rate": 0.00026396396396396397,
+      "loss": 0.4509,
+      "step": 810
+    },
+    {
+      "epoch": 0.2431328461682593,
+      "grad_norm": 0.2182149440050125,
+      "learning_rate": 0.0002639189189189189,
+      "loss": 0.4448,
+      "step": 811
+    },
+    {
+      "epoch": 0.24343264005995877,
+      "grad_norm": 0.2463945746421814,
+      "learning_rate": 0.00026387387387387384,
+      "loss": 0.5088,
+      "step": 812
+    },
+    {
+      "epoch": 0.24373243395165822,
+      "grad_norm": 0.2388424575328827,
+      "learning_rate": 0.00026382882882882883,
+      "loss": 0.4931,
+      "step": 813
+    },
+    {
+      "epoch": 0.24403222784335768,
+      "grad_norm": 0.2515762746334076,
+      "learning_rate": 0.00026378378378378377,
+      "loss": 0.4742,
+      "step": 814
+    },
+    {
+      "epoch": 0.24433202173505714,
+      "grad_norm": 0.23625001311302185,
+      "learning_rate": 0.0002637387387387387,
+      "loss": 0.4937,
+      "step": 815
+    },
+    {
+      "epoch": 0.2446318156267566,
+      "grad_norm": 0.2393738478422165,
+      "learning_rate": 0.0002636936936936937,
+      "loss": 0.4588,
+      "step": 816
+    },
+    {
+      "epoch": 0.24493160951845605,
+      "grad_norm": 0.23316219449043274,
+      "learning_rate": 0.00026364864864864864,
+      "loss": 0.4577,
+      "step": 817
+    },
+    {
+      "epoch": 0.2452314034101555,
+      "grad_norm": 0.2306746244430542,
+      "learning_rate": 0.00026360360360360357,
+      "loss": 0.4454,
+      "step": 818
+    },
+    {
+      "epoch": 0.24553119730185496,
+      "grad_norm": 0.26293689012527466,
+      "learning_rate": 0.0002635585585585585,
+      "loss": 0.4723,
+      "step": 819
+    },
+    {
+      "epoch": 0.24583099119355442,
+      "grad_norm": 0.23483715951442719,
+      "learning_rate": 0.0002635135135135135,
+      "loss": 0.4704,
+      "step": 820
+    },
+    {
+      "epoch": 0.24613078508525388,
+      "grad_norm": 0.2556680738925934,
+      "learning_rate": 0.00026346846846846844,
+      "loss": 0.5077,
+      "step": 821
+    },
+    {
+      "epoch": 0.24643057897695334,
+      "grad_norm": 0.25275811553001404,
+      "learning_rate": 0.0002634234234234234,
+      "loss": 0.483,
+      "step": 822
+    },
+    {
+      "epoch": 0.2467303728686528,
+      "grad_norm": 0.22292107343673706,
+      "learning_rate": 0.00026337837837837836,
+      "loss": 0.4684,
+      "step": 823
+    },
+    {
+      "epoch": 0.24703016676035225,
+      "grad_norm": 0.23125959932804108,
+      "learning_rate": 0.0002633333333333333,
+      "loss": 0.4618,
+      "step": 824
+    },
+    {
+      "epoch": 0.2473299606520517,
+      "grad_norm": 0.2515474259853363,
+      "learning_rate": 0.00026328828828828824,
+      "loss": 0.5111,
+      "step": 825
+    },
+    {
+      "epoch": 0.24762975454375116,
+      "grad_norm": 0.23193036019802094,
+      "learning_rate": 0.00026324324324324323,
+      "loss": 0.451,
+      "step": 826
+    },
+    {
+      "epoch": 0.24792954843545062,
+      "grad_norm": 0.22238105535507202,
+      "learning_rate": 0.00026319819819819817,
+      "loss": 0.4472,
+      "step": 827
+    },
+    {
+      "epoch": 0.24822934232715008,
+      "grad_norm": 0.23125764727592468,
+      "learning_rate": 0.0002631531531531531,
+      "loss": 0.4711,
+      "step": 828
+    },
+    {
+      "epoch": 0.24852913621884953,
+      "grad_norm": 0.23620037734508514,
+      "learning_rate": 0.0002631081081081081,
+      "loss": 0.4723,
+      "step": 829
+    },
+    {
+      "epoch": 0.248828930110549,
+      "grad_norm": 0.22470439970493317,
+      "learning_rate": 0.00026306306306306303,
+      "loss": 0.4814,
+      "step": 830
+    },
+    {
+      "epoch": 0.24912872400224845,
+      "grad_norm": 0.23267348110675812,
+      "learning_rate": 0.00026301801801801797,
+      "loss": 0.4669,
+      "step": 831
+    },
+    {
+      "epoch": 0.2494285178939479,
+      "grad_norm": 0.23558740317821503,
+      "learning_rate": 0.00026297297297297296,
+      "loss": 0.4407,
+      "step": 832
+    },
+    {
+      "epoch": 0.24972831178564736,
+      "grad_norm": 0.2202112227678299,
+      "learning_rate": 0.0002629279279279279,
+      "loss": 0.4546,
+      "step": 833
+    },
+    {
+      "epoch": 0.25002810567734685,
+      "grad_norm": 0.2349451333284378,
+      "learning_rate": 0.00026288288288288283,
+      "loss": 0.4741,
+      "step": 834
+    },
+    {
+      "epoch": 0.2503278995690463,
+      "grad_norm": 0.2210862636566162,
+      "learning_rate": 0.0002628378378378378,
+      "loss": 0.4624,
+      "step": 835
+    },
+    {
+      "epoch": 0.25062769346074576,
+      "grad_norm": 0.25249290466308594,
+      "learning_rate": 0.00026279279279279276,
+      "loss": 0.5213,
+      "step": 836
+    },
+    {
+      "epoch": 0.2509274873524452,
+      "grad_norm": 0.2458237111568451,
+      "learning_rate": 0.0002627477477477477,
+      "loss": 0.4937,
+      "step": 837
+    },
+    {
+      "epoch": 0.2512272812441447,
+      "grad_norm": 0.22827856242656708,
+      "learning_rate": 0.0002627027027027027,
+      "loss": 0.4286,
+      "step": 838
+    },
+    {
+      "epoch": 0.2515270751358441,
+      "grad_norm": 0.22871458530426025,
+      "learning_rate": 0.0002626576576576576,
+      "loss": 0.4181,
+      "step": 839
+    },
+    {
+      "epoch": 0.2518268690275436,
+      "grad_norm": 0.24196332693099976,
+      "learning_rate": 0.00026261261261261256,
+      "loss": 0.4704,
+      "step": 840
+    },
+    {
+      "epoch": 0.252126662919243,
+      "grad_norm": 0.24222321808338165,
+      "learning_rate": 0.00026256756756756755,
+      "loss": 0.474,
+      "step": 841
+    },
+    {
+      "epoch": 0.2524264568109425,
+      "grad_norm": 0.2258533090353012,
+      "learning_rate": 0.0002625225225225225,
+      "loss": 0.4644,
+      "step": 842
+    },
+    {
+      "epoch": 0.25272625070264193,
+      "grad_norm": 0.2234419882297516,
+      "learning_rate": 0.0002624774774774774,
+      "loss": 0.4506,
+      "step": 843
+    },
+    {
+      "epoch": 0.2530260445943414,
+      "grad_norm": 0.24231363832950592,
+      "learning_rate": 0.0002624324324324324,
+      "loss": 0.4975,
+      "step": 844
+    },
+    {
+      "epoch": 0.25332583848604084,
+      "grad_norm": 0.2430192083120346,
+      "learning_rate": 0.00026238738738738735,
+      "loss": 0.462,
+      "step": 845
+    },
+    {
+      "epoch": 0.25362563237774033,
+      "grad_norm": 0.23717942833900452,
+      "learning_rate": 0.0002623423423423423,
+      "loss": 0.48,
+      "step": 846
+    },
+    {
+      "epoch": 0.25392542626943976,
+      "grad_norm": 0.23983192443847656,
+      "learning_rate": 0.0002622972972972973,
+      "loss": 0.491,
+      "step": 847
+    },
+    {
+      "epoch": 0.25422522016113924,
+      "grad_norm": 0.24544605612754822,
+      "learning_rate": 0.0002622522522522522,
+      "loss": 0.4525,
+      "step": 848
+    },
+    {
+      "epoch": 0.25452501405283867,
+      "grad_norm": 0.25106650590896606,
+      "learning_rate": 0.0002622072072072072,
+      "loss": 0.4716,
+      "step": 849
+    },
+    {
+      "epoch": 0.25482480794453816,
+      "grad_norm": 0.2644721269607544,
+      "learning_rate": 0.00026216216216216215,
+      "loss": 0.4888,
+      "step": 850
+    },
+    {
+      "epoch": 0.2551246018362376,
+      "grad_norm": 0.2338344007730484,
+      "learning_rate": 0.0002621171171171171,
+      "loss": 0.4779,
+      "step": 851
+    },
+    {
+      "epoch": 0.25542439572793707,
+      "grad_norm": 0.2368081659078598,
+      "learning_rate": 0.00026207207207207207,
+      "loss": 0.458,
+      "step": 852
+    },
+    {
+      "epoch": 0.2557241896196365,
+      "grad_norm": 0.2628321051597595,
+      "learning_rate": 0.000262027027027027,
+      "loss": 0.4786,
+      "step": 853
+    },
+    {
+      "epoch": 0.256023983511336,
+      "grad_norm": 0.23109877109527588,
+      "learning_rate": 0.00026198198198198195,
+      "loss": 0.4396,
+      "step": 854
+    },
+    {
+      "epoch": 0.2563237774030354,
+      "grad_norm": 0.23273521661758423,
+      "learning_rate": 0.00026193693693693694,
+      "loss": 0.4579,
+      "step": 855
+    },
+    {
+      "epoch": 0.25662357129473484,
+      "grad_norm": 0.229389026761055,
+      "learning_rate": 0.0002618918918918919,
+      "loss": 0.4515,
+      "step": 856
+    },
+    {
+      "epoch": 0.2569233651864343,
+      "grad_norm": 0.24866041541099548,
+      "learning_rate": 0.0002618468468468468,
+      "loss": 0.4402,
+      "step": 857
+    },
+    {
+      "epoch": 0.25722315907813376,
+      "grad_norm": 0.24374257028102875,
+      "learning_rate": 0.0002618018018018018,
+      "loss": 0.5025,
+      "step": 858
+    },
+    {
+      "epoch": 0.25752295296983324,
+      "grad_norm": 0.2753133177757263,
+      "learning_rate": 0.00026175675675675674,
+      "loss": 0.5058,
+      "step": 859
+    },
+    {
+      "epoch": 0.25782274686153267,
+      "grad_norm": 0.236386239528656,
+      "learning_rate": 0.0002617117117117117,
+      "loss": 0.4741,
+      "step": 860
+    },
+    {
+      "epoch": 0.25812254075323215,
+      "grad_norm": 0.21907605230808258,
+      "learning_rate": 0.00026166666666666667,
+      "loss": 0.463,
+      "step": 861
+    },
+    {
+      "epoch": 0.2584223346449316,
+      "grad_norm": 0.25744542479515076,
+      "learning_rate": 0.0002616216216216216,
+      "loss": 0.4585,
+      "step": 862
+    },
+    {
+      "epoch": 0.25872212853663107,
+      "grad_norm": 0.25060373544692993,
+      "learning_rate": 0.0002615765765765766,
+      "loss": 0.476,
+      "step": 863
+    },
+    {
+      "epoch": 0.2590219224283305,
+      "grad_norm": 0.21545180678367615,
+      "learning_rate": 0.00026153153153153153,
+      "loss": 0.4379,
+      "step": 864
+    },
+    {
+      "epoch": 0.25932171632003,
+      "grad_norm": 0.2536545991897583,
+      "learning_rate": 0.00026148648648648647,
+      "loss": 0.4965,
+      "step": 865
+    },
+    {
+      "epoch": 0.2596215102117294,
+      "grad_norm": 0.22960424423217773,
+      "learning_rate": 0.00026144144144144146,
+      "loss": 0.4441,
+      "step": 866
+    },
+    {
+      "epoch": 0.2599213041034289,
+      "grad_norm": 0.22601282596588135,
+      "learning_rate": 0.0002613963963963964,
+      "loss": 0.4462,
+      "step": 867
+    },
+    {
+      "epoch": 0.2602210979951283,
+      "grad_norm": 0.23997683823108673,
+      "learning_rate": 0.00026135135135135133,
+      "loss": 0.4496,
+      "step": 868
+    },
+    {
+      "epoch": 0.2605208918868278,
+      "grad_norm": 0.24241064488887787,
+      "learning_rate": 0.00026130630630630627,
+      "loss": 0.4538,
+      "step": 869
+    },
+    {
+      "epoch": 0.26082068577852724,
+      "grad_norm": 0.2302720993757248,
+      "learning_rate": 0.00026126126126126126,
+      "loss": 0.4547,
+      "step": 870
+    },
+    {
+      "epoch": 0.2611204796702267,
+      "grad_norm": 0.2532520294189453,
+      "learning_rate": 0.0002612162162162162,
+      "loss": 0.4919,
+      "step": 871
+    },
+    {
+      "epoch": 0.26142027356192615,
+      "grad_norm": 0.2543450891971588,
+      "learning_rate": 0.00026117117117117113,
+      "loss": 0.4638,
+      "step": 872
+    },
+    {
+      "epoch": 0.26172006745362564,
+      "grad_norm": 0.24088874459266663,
+      "learning_rate": 0.0002611261261261261,
+      "loss": 0.4603,
+      "step": 873
+    },
+    {
+      "epoch": 0.26201986134532507,
+      "grad_norm": 0.22305645048618317,
+      "learning_rate": 0.00026108108108108106,
+      "loss": 0.4394,
+      "step": 874
+    },
+    {
+      "epoch": 0.26231965523702455,
+      "grad_norm": 0.261001318693161,
+      "learning_rate": 0.000261036036036036,
+      "loss": 0.5244,
+      "step": 875
+    },
+    {
+      "epoch": 0.262619449128724,
+      "grad_norm": 0.2550908029079437,
+      "learning_rate": 0.000260990990990991,
+      "loss": 0.4716,
+      "step": 876
+    },
+    {
+      "epoch": 0.26291924302042347,
+      "grad_norm": 0.2264460027217865,
+      "learning_rate": 0.0002609459459459459,
+      "loss": 0.4527,
+      "step": 877
+    },
+    {
+      "epoch": 0.2632190369121229,
+      "grad_norm": 0.2598486542701721,
+      "learning_rate": 0.00026090090090090086,
+      "loss": 0.51,
+      "step": 878
+    },
+    {
+      "epoch": 0.2635188308038224,
+      "grad_norm": 0.2528247833251953,
+      "learning_rate": 0.00026085585585585585,
+      "loss": 0.4619,
+      "step": 879
+    },
+    {
+      "epoch": 0.2638186246955218,
+      "grad_norm": 0.22703434526920319,
+      "learning_rate": 0.0002608108108108108,
+      "loss": 0.4519,
+      "step": 880
+    },
+    {
+      "epoch": 0.2641184185872213,
+      "grad_norm": 0.24291987717151642,
+      "learning_rate": 0.0002607657657657657,
+      "loss": 0.4588,
+      "step": 881
+    },
+    {
+      "epoch": 0.2644182124789207,
+      "grad_norm": 0.265899121761322,
+      "learning_rate": 0.0002607207207207207,
+      "loss": 0.4741,
+      "step": 882
+    },
+    {
+      "epoch": 0.2647180063706202,
+      "grad_norm": 0.24852798879146576,
+      "learning_rate": 0.00026067567567567565,
+      "loss": 0.4772,
+      "step": 883
+    },
+    {
+      "epoch": 0.26501780026231964,
+      "grad_norm": 0.24373799562454224,
+      "learning_rate": 0.0002606306306306306,
+      "loss": 0.475,
+      "step": 884
+    },
+    {
+      "epoch": 0.2653175941540191,
+      "grad_norm": 0.24994871020317078,
+      "learning_rate": 0.0002605855855855856,
+      "loss": 0.5094,
+      "step": 885
+    },
+    {
+      "epoch": 0.26561738804571855,
+      "grad_norm": 0.23686103522777557,
+      "learning_rate": 0.0002605405405405405,
+      "loss": 0.4583,
+      "step": 886
+    },
+    {
+      "epoch": 0.26591718193741803,
+      "grad_norm": 0.2280004322528839,
+      "learning_rate": 0.00026049549549549546,
+      "loss": 0.4474,
+      "step": 887
+    },
+    {
+      "epoch": 0.26621697582911746,
+      "grad_norm": 0.25110939145088196,
+      "learning_rate": 0.00026045045045045045,
+      "loss": 0.4722,
+      "step": 888
+    },
+    {
+      "epoch": 0.26651676972081695,
+      "grad_norm": 0.25370022654533386,
+      "learning_rate": 0.0002604054054054054,
+      "loss": 0.4651,
+      "step": 889
+    },
+    {
+      "epoch": 0.2668165636125164,
+      "grad_norm": 0.24179215729236603,
+      "learning_rate": 0.0002603603603603603,
+      "loss": 0.4727,
+      "step": 890
+    },
+    {
+      "epoch": 0.26711635750421586,
+      "grad_norm": 0.2525777518749237,
+      "learning_rate": 0.00026031531531531526,
+      "loss": 0.4681,
+      "step": 891
+    },
+    {
+      "epoch": 0.2674161513959153,
+      "grad_norm": 0.21325957775115967,
+      "learning_rate": 0.00026027027027027025,
+      "loss": 0.4374,
+      "step": 892
+    },
+    {
+      "epoch": 0.2677159452876148,
+      "grad_norm": 0.2358642816543579,
+      "learning_rate": 0.0002602252252252252,
+      "loss": 0.4492,
+      "step": 893
+    },
+    {
+      "epoch": 0.2680157391793142,
+      "grad_norm": 0.2625977694988251,
+      "learning_rate": 0.0002601801801801801,
+      "loss": 0.522,
+      "step": 894
+    },
+    {
+      "epoch": 0.2683155330710137,
+      "grad_norm": 0.22606413066387177,
+      "learning_rate": 0.0002601351351351351,
+      "loss": 0.4539,
+      "step": 895
+    },
+    {
+      "epoch": 0.2686153269627131,
+      "grad_norm": 0.24337491393089294,
+      "learning_rate": 0.00026009009009009005,
+      "loss": 0.4988,
+      "step": 896
+    },
+    {
+      "epoch": 0.2689151208544126,
+      "grad_norm": 0.23522725701332092,
+      "learning_rate": 0.000260045045045045,
+      "loss": 0.4665,
+      "step": 897
+    },
+    {
+      "epoch": 0.26921491474611203,
+      "grad_norm": 0.25222131609916687,
+      "learning_rate": 0.00026,
+      "loss": 0.4715,
+      "step": 898
+    },
+    {
+      "epoch": 0.2695147086378115,
+      "grad_norm": 0.22760646045207977,
+      "learning_rate": 0.0002599549549549549,
+      "loss": 0.4633,
+      "step": 899
+    },
+    {
+      "epoch": 0.26981450252951095,
+      "grad_norm": 0.2398597002029419,
+      "learning_rate": 0.00025990990990990985,
+      "loss": 0.4682,
+      "step": 900
+    },
+    {
+      "epoch": 0.27011429642121043,
+      "grad_norm": 0.24494816362857819,
+      "learning_rate": 0.00025986486486486484,
+      "loss": 0.4833,
+      "step": 901
+    },
+    {
+      "epoch": 0.27041409031290986,
+      "grad_norm": 0.23173199594020844,
+      "learning_rate": 0.0002598198198198198,
+      "loss": 0.4583,
+      "step": 902
+    },
+    {
+      "epoch": 0.27071388420460935,
+      "grad_norm": 0.242969810962677,
+      "learning_rate": 0.0002597747747747747,
+      "loss": 0.4748,
+      "step": 903
+    },
+    {
+      "epoch": 0.2710136780963088,
+      "grad_norm": 0.2286025583744049,
+      "learning_rate": 0.0002597297297297297,
+      "loss": 0.4802,
+      "step": 904
+    },
+    {
+      "epoch": 0.27131347198800826,
+      "grad_norm": 0.241167351603508,
+      "learning_rate": 0.00025968468468468464,
+      "loss": 0.4947,
+      "step": 905
+    },
+    {
+      "epoch": 0.2716132658797077,
+      "grad_norm": 0.2599638104438782,
+      "learning_rate": 0.00025963963963963963,
+      "loss": 0.5028,
+      "step": 906
+    },
+    {
+      "epoch": 0.2719130597714072,
+      "grad_norm": 0.22766104340553284,
+      "learning_rate": 0.00025959459459459457,
+      "loss": 0.4586,
+      "step": 907
+    },
+    {
+      "epoch": 0.2722128536631066,
+      "grad_norm": 0.24524454772472382,
+      "learning_rate": 0.0002595495495495495,
+      "loss": 0.4918,
+      "step": 908
+    },
+    {
+      "epoch": 0.2725126475548061,
+      "grad_norm": 0.24995583295822144,
+      "learning_rate": 0.0002595045045045045,
+      "loss": 0.4851,
+      "step": 909
+    },
+    {
+      "epoch": 0.2728124414465055,
+      "grad_norm": 0.24542704224586487,
+      "learning_rate": 0.00025945945945945944,
+      "loss": 0.4891,
+      "step": 910
+    },
+    {
+      "epoch": 0.273112235338205,
+      "grad_norm": 0.2262720763683319,
+      "learning_rate": 0.00025941441441441437,
+      "loss": 0.457,
+      "step": 911
+    },
+    {
+      "epoch": 0.27341202922990443,
+      "grad_norm": 0.2677282989025116,
+      "learning_rate": 0.00025936936936936936,
+      "loss": 0.4781,
+      "step": 912
+    },
+    {
+      "epoch": 0.2737118231216039,
+      "grad_norm": 0.22617483139038086,
+      "learning_rate": 0.0002593243243243243,
+      "loss": 0.4666,
+      "step": 913
+    },
+    {
+      "epoch": 0.27401161701330334,
+      "grad_norm": 0.24579745531082153,
+      "learning_rate": 0.00025927927927927924,
+      "loss": 0.4991,
+      "step": 914
+    },
+    {
+      "epoch": 0.27431141090500283,
+      "grad_norm": 0.22964158654212952,
+      "learning_rate": 0.00025923423423423423,
+      "loss": 0.4757,
+      "step": 915
+    },
+    {
+      "epoch": 0.27461120479670226,
+      "grad_norm": 0.24435275793075562,
+      "learning_rate": 0.00025918918918918916,
+      "loss": 0.4636,
+      "step": 916
+    },
+    {
+      "epoch": 0.27491099868840174,
+      "grad_norm": 0.23039105534553528,
+      "learning_rate": 0.0002591441441441441,
+      "loss": 0.4591,
+      "step": 917
+    },
+    {
+      "epoch": 0.27521079258010117,
+      "grad_norm": 0.24856770038604736,
+      "learning_rate": 0.0002590990990990991,
+      "loss": 0.4866,
+      "step": 918
+    },
+    {
+      "epoch": 0.27551058647180066,
+      "grad_norm": 0.22115269303321838,
+      "learning_rate": 0.00025905405405405403,
+      "loss": 0.4322,
+      "step": 919
+    },
+    {
+      "epoch": 0.2758103803635001,
+      "grad_norm": 0.2645402252674103,
+      "learning_rate": 0.000259009009009009,
+      "loss": 0.4891,
+      "step": 920
+    },
+    {
+      "epoch": 0.27611017425519957,
+      "grad_norm": 0.24427354335784912,
+      "learning_rate": 0.00025896396396396396,
+      "loss": 0.4636,
+      "step": 921
+    },
+    {
+      "epoch": 0.276409968146899,
+      "grad_norm": 0.23059400916099548,
+      "learning_rate": 0.0002589189189189189,
+      "loss": 0.4502,
+      "step": 922
+    },
+    {
+      "epoch": 0.2767097620385985,
+      "grad_norm": 0.21996812522411346,
+      "learning_rate": 0.0002588738738738739,
+      "loss": 0.4461,
+      "step": 923
+    },
+    {
+      "epoch": 0.2770095559302979,
+      "grad_norm": 0.24204552173614502,
+      "learning_rate": 0.0002588288288288288,
+      "loss": 0.4678,
+      "step": 924
+    },
+    {
+      "epoch": 0.2773093498219974,
+      "grad_norm": 0.26428595185279846,
+      "learning_rate": 0.00025878378378378376,
+      "loss": 0.476,
+      "step": 925
+    },
+    {
+      "epoch": 0.2776091437136968,
+      "grad_norm": 0.2542773187160492,
+      "learning_rate": 0.00025873873873873875,
+      "loss": 0.464,
+      "step": 926
+    },
+    {
+      "epoch": 0.2779089376053963,
+      "grad_norm": 0.2621975839138031,
+      "learning_rate": 0.0002586936936936937,
+      "loss": 0.4735,
+      "step": 927
+    },
+    {
+      "epoch": 0.27820873149709574,
+      "grad_norm": 0.24359507858753204,
+      "learning_rate": 0.0002586486486486486,
+      "loss": 0.4768,
+      "step": 928
+    },
+    {
+      "epoch": 0.2785085253887952,
+      "grad_norm": 0.24825096130371094,
+      "learning_rate": 0.0002586036036036036,
+      "loss": 0.471,
+      "step": 929
+    },
+    {
+      "epoch": 0.27880831928049465,
+      "grad_norm": 0.2950778007507324,
+      "learning_rate": 0.00025855855855855855,
+      "loss": 0.4625,
+      "step": 930
+    },
+    {
+      "epoch": 0.27910811317219414,
+      "grad_norm": 0.23210273683071136,
+      "learning_rate": 0.0002585135135135135,
+      "loss": 0.4325,
+      "step": 931
+    },
+    {
+      "epoch": 0.27940790706389357,
+      "grad_norm": 0.26627883315086365,
+      "learning_rate": 0.0002584684684684685,
+      "loss": 0.4903,
+      "step": 932
+    },
+    {
+      "epoch": 0.27970770095559305,
+      "grad_norm": 0.2619936466217041,
+      "learning_rate": 0.0002584234234234234,
+      "loss": 0.4777,
+      "step": 933
+    },
+    {
+      "epoch": 0.2800074948472925,
+      "grad_norm": 0.23771123588085175,
+      "learning_rate": 0.00025837837837837835,
+      "loss": 0.4319,
+      "step": 934
+    },
+    {
+      "epoch": 0.28030728873899197,
+      "grad_norm": 0.2495034784078598,
+      "learning_rate": 0.00025833333333333334,
+      "loss": 0.4503,
+      "step": 935
+    },
+    {
+      "epoch": 0.2806070826306914,
+      "grad_norm": 0.26627448201179504,
+      "learning_rate": 0.0002582882882882883,
+      "loss": 0.4501,
+      "step": 936
+    },
+    {
+      "epoch": 0.2809068765223909,
+      "grad_norm": 0.22482304275035858,
+      "learning_rate": 0.0002582432432432432,
+      "loss": 0.4615,
+      "step": 937
+    },
+    {
+      "epoch": 0.2812066704140903,
+      "grad_norm": 0.23891392350196838,
+      "learning_rate": 0.0002581981981981982,
+      "loss": 0.4732,
+      "step": 938
+    },
+    {
+      "epoch": 0.2815064643057898,
+      "grad_norm": 0.2233395278453827,
+      "learning_rate": 0.00025815315315315314,
+      "loss": 0.4521,
+      "step": 939
+    },
+    {
+      "epoch": 0.2818062581974892,
+      "grad_norm": 0.22510269284248352,
+      "learning_rate": 0.0002581081081081081,
+      "loss": 0.4307,
+      "step": 940
+    },
+    {
+      "epoch": 0.2821060520891887,
+      "grad_norm": 0.24909009039402008,
+      "learning_rate": 0.000258063063063063,
+      "loss": 0.4574,
+      "step": 941
+    },
+    {
+      "epoch": 0.28240584598088814,
+      "grad_norm": 0.2461954653263092,
+      "learning_rate": 0.000258018018018018,
+      "loss": 0.4782,
+      "step": 942
+    },
+    {
+      "epoch": 0.2827056398725876,
+      "grad_norm": 0.23391996324062347,
+      "learning_rate": 0.00025797297297297294,
+      "loss": 0.4692,
+      "step": 943
+    },
+    {
+      "epoch": 0.28300543376428705,
+      "grad_norm": 0.2419288158416748,
+      "learning_rate": 0.0002579279279279279,
+      "loss": 0.4742,
+      "step": 944
+    },
+    {
+      "epoch": 0.28330522765598654,
+      "grad_norm": 0.24654226005077362,
+      "learning_rate": 0.00025788288288288287,
+      "loss": 0.4793,
+      "step": 945
+    },
+    {
+      "epoch": 0.28360502154768596,
+      "grad_norm": 0.23362454771995544,
+      "learning_rate": 0.0002578378378378378,
+      "loss": 0.4398,
+      "step": 946
+    },
+    {
+      "epoch": 0.28390481543938545,
+      "grad_norm": 0.23269514739513397,
+      "learning_rate": 0.00025779279279279275,
+      "loss": 0.4642,
+      "step": 947
+    },
+    {
+      "epoch": 0.2842046093310849,
+      "grad_norm": 0.22531503438949585,
+      "learning_rate": 0.00025774774774774774,
+      "loss": 0.436,
+      "step": 948
+    },
+    {
+      "epoch": 0.28450440322278436,
+      "grad_norm": 0.24250438809394836,
+      "learning_rate": 0.0002577027027027027,
+      "loss": 0.4708,
+      "step": 949
+    },
+    {
+      "epoch": 0.2848041971144838,
+      "grad_norm": 0.2370329648256302,
+      "learning_rate": 0.0002576576576576576,
+      "loss": 0.4737,
+      "step": 950
+    },
+    {
+      "epoch": 0.2851039910061832,
+      "grad_norm": 0.2553395628929138,
+      "learning_rate": 0.0002576126126126126,
+      "loss": 0.4927,
+      "step": 951
+    },
+    {
+      "epoch": 0.2854037848978827,
+      "grad_norm": 0.24398140609264374,
+      "learning_rate": 0.00025756756756756754,
+      "loss": 0.4471,
+      "step": 952
+    },
+    {
+      "epoch": 0.28570357878958214,
+      "grad_norm": 0.2420070916414261,
+      "learning_rate": 0.0002575225225225225,
+      "loss": 0.4573,
+      "step": 953
+    },
+    {
+      "epoch": 0.2860033726812816,
+      "grad_norm": 0.22280406951904297,
+      "learning_rate": 0.00025747747747747747,
+      "loss": 0.4637,
+      "step": 954
+    },
+    {
+      "epoch": 0.28630316657298105,
+      "grad_norm": 0.268107146024704,
+      "learning_rate": 0.0002574324324324324,
+      "loss": 0.4751,
+      "step": 955
+    },
+    {
+      "epoch": 0.28660296046468053,
+      "grad_norm": 0.224797785282135,
+      "learning_rate": 0.00025738738738738734,
+      "loss": 0.4567,
+      "step": 956
+    },
+    {
+      "epoch": 0.28690275435637996,
+      "grad_norm": 0.2350010722875595,
+      "learning_rate": 0.00025734234234234233,
+      "loss": 0.4669,
+      "step": 957
+    },
+    {
+      "epoch": 0.28720254824807945,
+      "grad_norm": 0.23346953094005585,
+      "learning_rate": 0.00025729729729729727,
+      "loss": 0.4711,
+      "step": 958
+    },
+    {
+      "epoch": 0.2875023421397789,
+      "grad_norm": 0.26031309366226196,
+      "learning_rate": 0.0002572522522522522,
+      "loss": 0.4921,
+      "step": 959
+    },
+    {
+      "epoch": 0.28780213603147836,
+      "grad_norm": 0.21255329251289368,
+      "learning_rate": 0.0002572072072072072,
+      "loss": 0.444,
+      "step": 960
+    },
+    {
+      "epoch": 0.2881019299231778,
+      "grad_norm": 0.24799884855747223,
+      "learning_rate": 0.00025716216216216213,
+      "loss": 0.469,
+      "step": 961
+    },
+    {
+      "epoch": 0.2884017238148773,
+      "grad_norm": 0.2208838164806366,
+      "learning_rate": 0.00025711711711711707,
+      "loss": 0.442,
+      "step": 962
+    },
+    {
+      "epoch": 0.2887015177065767,
+      "grad_norm": 0.2880913317203522,
+      "learning_rate": 0.00025707207207207206,
+      "loss": 0.4424,
+      "step": 963
+    },
+    {
+      "epoch": 0.2890013115982762,
+      "grad_norm": 0.26574239134788513,
+      "learning_rate": 0.000257027027027027,
+      "loss": 0.452,
+      "step": 964
+    },
+    {
+      "epoch": 0.2893011054899756,
+      "grad_norm": 0.23267340660095215,
+      "learning_rate": 0.00025698198198198193,
+      "loss": 0.455,
+      "step": 965
+    },
+    {
+      "epoch": 0.2896008993816751,
+      "grad_norm": 0.26304900646209717,
+      "learning_rate": 0.0002569369369369369,
+      "loss": 0.5211,
+      "step": 966
+    },
+    {
+      "epoch": 0.28990069327337453,
+      "grad_norm": 0.2575905919075012,
+      "learning_rate": 0.00025689189189189186,
+      "loss": 0.483,
+      "step": 967
+    },
+    {
+      "epoch": 0.290200487165074,
+      "grad_norm": 0.22459660470485687,
+      "learning_rate": 0.0002568468468468468,
+      "loss": 0.4636,
+      "step": 968
+    },
+    {
+      "epoch": 0.29050028105677345,
+      "grad_norm": 0.220341295003891,
+      "learning_rate": 0.0002568018018018018,
+      "loss": 0.4615,
+      "step": 969
+    },
+    {
+      "epoch": 0.29080007494847293,
+      "grad_norm": 0.239531472325325,
+      "learning_rate": 0.0002567567567567567,
+      "loss": 0.4446,
+      "step": 970
+    },
+    {
+      "epoch": 0.29109986884017236,
+      "grad_norm": 0.23338812589645386,
+      "learning_rate": 0.00025671171171171166,
+      "loss": 0.4704,
+      "step": 971
+    },
+    {
+      "epoch": 0.29139966273187184,
+      "grad_norm": 0.24035978317260742,
+      "learning_rate": 0.00025666666666666665,
+      "loss": 0.4717,
+      "step": 972
+    },
+    {
+      "epoch": 0.2916994566235713,
+      "grad_norm": 0.23094506561756134,
+      "learning_rate": 0.0002566216216216216,
+      "loss": 0.4493,
+      "step": 973
+    },
+    {
+      "epoch": 0.29199925051527076,
+      "grad_norm": 0.25101473927497864,
+      "learning_rate": 0.0002565765765765765,
+      "loss": 0.4979,
+      "step": 974
+    },
+    {
+      "epoch": 0.2922990444069702,
+      "grad_norm": 0.24172839522361755,
+      "learning_rate": 0.0002565315315315315,
+      "loss": 0.4666,
+      "step": 975
+    },
+    {
+      "epoch": 0.2925988382986697,
+      "grad_norm": 0.2213078737258911,
+      "learning_rate": 0.00025648648648648645,
+      "loss": 0.4335,
+      "step": 976
+    },
+    {
+      "epoch": 0.2928986321903691,
+      "grad_norm": 0.2230203002691269,
+      "learning_rate": 0.00025644144144144145,
+      "loss": 0.4183,
+      "step": 977
+    },
+    {
+      "epoch": 0.2931984260820686,
+      "grad_norm": 0.24735258519649506,
+      "learning_rate": 0.0002563963963963964,
+      "loss": 0.4612,
+      "step": 978
+    },
+    {
+      "epoch": 0.293498219973768,
+      "grad_norm": 0.24861575663089752,
+      "learning_rate": 0.0002563513513513513,
+      "loss": 0.4713,
+      "step": 979
+    },
+    {
+      "epoch": 0.2937980138654675,
+      "grad_norm": 0.2333897352218628,
+      "learning_rate": 0.0002563063063063063,
+      "loss": 0.4502,
+      "step": 980
+    },
+    {
+      "epoch": 0.29409780775716693,
+      "grad_norm": 0.23923064768314362,
+      "learning_rate": 0.00025626126126126125,
+      "loss": 0.4574,
+      "step": 981
+    },
+    {
+      "epoch": 0.2943976016488664,
+      "grad_norm": 0.24568355083465576,
+      "learning_rate": 0.0002562162162162162,
+      "loss": 0.4358,
+      "step": 982
+    },
+    {
+      "epoch": 0.29469739554056584,
+      "grad_norm": 0.24993112683296204,
+      "learning_rate": 0.0002561711711711712,
+      "loss": 0.4663,
+      "step": 983
+    },
+    {
+      "epoch": 0.2949971894322653,
+      "grad_norm": 0.2531440854072571,
+      "learning_rate": 0.0002561261261261261,
+      "loss": 0.4964,
+      "step": 984
+    },
+    {
+      "epoch": 0.29529698332396476,
+      "grad_norm": 0.26995500922203064,
+      "learning_rate": 0.00025608108108108105,
+      "loss": 0.4971,
+      "step": 985
+    },
+    {
+      "epoch": 0.29559677721566424,
+      "grad_norm": 0.23319192230701447,
+      "learning_rate": 0.00025603603603603604,
+      "loss": 0.4605,
+      "step": 986
+    },
+    {
+      "epoch": 0.29589657110736367,
+      "grad_norm": 0.2496713548898697,
+      "learning_rate": 0.000255990990990991,
+      "loss": 0.4468,
+      "step": 987
+    },
+    {
+      "epoch": 0.29619636499906316,
+      "grad_norm": 0.23224860429763794,
+      "learning_rate": 0.0002559459459459459,
+      "loss": 0.4519,
+      "step": 988
+    },
+    {
+      "epoch": 0.2964961588907626,
+      "grad_norm": 0.24383842945098877,
+      "learning_rate": 0.0002559009009009009,
+      "loss": 0.467,
+      "step": 989
+    },
+    {
+      "epoch": 0.29679595278246207,
+      "grad_norm": 0.2240372598171234,
+      "learning_rate": 0.00025585585585585584,
+      "loss": 0.4541,
+      "step": 990
+    },
+    {
+      "epoch": 0.2970957466741615,
+      "grad_norm": 0.23554278910160065,
+      "learning_rate": 0.0002558108108108108,
+      "loss": 0.4627,
+      "step": 991
+    },
+    {
+      "epoch": 0.297395540565861,
+      "grad_norm": 0.24655288457870483,
+      "learning_rate": 0.00025576576576576577,
+      "loss": 0.4697,
+      "step": 992
+    },
+    {
+      "epoch": 0.2976953344575604,
+      "grad_norm": 0.2397495061159134,
+      "learning_rate": 0.0002557207207207207,
+      "loss": 0.4821,
+      "step": 993
+    },
+    {
+      "epoch": 0.2979951283492599,
+      "grad_norm": 0.24440963566303253,
+      "learning_rate": 0.00025567567567567564,
+      "loss": 0.4293,
+      "step": 994
+    },
+    {
+      "epoch": 0.2982949222409593,
+      "grad_norm": 0.230440154671669,
+      "learning_rate": 0.00025563063063063063,
+      "loss": 0.4592,
+      "step": 995
+    },
+    {
+      "epoch": 0.2985947161326588,
+      "grad_norm": 0.22211557626724243,
+      "learning_rate": 0.00025558558558558557,
+      "loss": 0.4324,
+      "step": 996
+    },
+    {
+      "epoch": 0.29889451002435824,
+      "grad_norm": 0.22826789319515228,
+      "learning_rate": 0.0002555405405405405,
+      "loss": 0.447,
+      "step": 997
+    },
+    {
+      "epoch": 0.2991943039160577,
+      "grad_norm": 0.24060975015163422,
+      "learning_rate": 0.0002554954954954955,
+      "loss": 0.4954,
+      "step": 998
+    },
+    {
+      "epoch": 0.29949409780775715,
+      "grad_norm": 0.2227400243282318,
+      "learning_rate": 0.00025545045045045043,
+      "loss": 0.4503,
+      "step": 999
+    },
+    {
+      "epoch": 0.29979389169945664,
+      "grad_norm": 0.23061898350715637,
+      "learning_rate": 0.00025540540540540537,
+      "loss": 0.4356,
+      "step": 1000
+    },
+    {
+      "epoch": 0.29979389169945664,
+      "eval_loss": 0.46561944484710693,
+      "eval_runtime": 566.7946,
+      "eval_samples_per_second": 3.809,
+      "eval_steps_per_second": 0.476,
+      "step": 1000
+    },
+    {
+      "epoch": 0.30009368559115607,
+      "grad_norm": 0.25651443004608154,
+      "learning_rate": 0.00025536036036036036,
+      "loss": 0.4813,
+      "step": 1001
+    },
+    {
+      "epoch": 0.30039347948285555,
+      "grad_norm": 0.23068277537822723,
+      "learning_rate": 0.0002553153153153153,
+      "loss": 0.4563,
+      "step": 1002
+    },
+    {
+      "epoch": 0.300693273374555,
+      "grad_norm": 0.2346974015235901,
+      "learning_rate": 0.00025527027027027024,
+      "loss": 0.4471,
+      "step": 1003
+    },
+    {
+      "epoch": 0.30099306726625447,
+      "grad_norm": 0.22291089594364166,
+      "learning_rate": 0.0002552252252252252,
+      "loss": 0.4399,
+      "step": 1004
+    },
+    {
+      "epoch": 0.3012928611579539,
+      "grad_norm": 0.24533626437187195,
+      "learning_rate": 0.00025518018018018016,
+      "loss": 0.4949,
+      "step": 1005
+    },
+    {
+      "epoch": 0.3015926550496534,
+      "grad_norm": 0.2337205857038498,
+      "learning_rate": 0.0002551351351351351,
+      "loss": 0.4815,
+      "step": 1006
+    },
+    {
+      "epoch": 0.3018924489413528,
+      "grad_norm": 0.23926278948783875,
+      "learning_rate": 0.0002550900900900901,
+      "loss": 0.4742,
+      "step": 1007
+    },
+    {
+      "epoch": 0.3021922428330523,
+      "grad_norm": 0.23630262911319733,
+      "learning_rate": 0.00025504504504504503,
+      "loss": 0.4643,
+      "step": 1008
+    },
+    {
+      "epoch": 0.3024920367247517,
+      "grad_norm": 0.2534981071949005,
+      "learning_rate": 0.00025499999999999996,
+      "loss": 0.4806,
+      "step": 1009
+    },
+    {
+      "epoch": 0.3027918306164512,
+      "grad_norm": 0.22765369713306427,
+      "learning_rate": 0.00025495495495495496,
+      "loss": 0.4562,
+      "step": 1010
+    },
+    {
+      "epoch": 0.30309162450815064,
+      "grad_norm": 0.23825423419475555,
+      "learning_rate": 0.0002549099099099099,
+      "loss": 0.4642,
+      "step": 1011
+    },
+    {
+      "epoch": 0.3033914183998501,
+      "grad_norm": 0.2275952249765396,
+      "learning_rate": 0.00025486486486486483,
+      "loss": 0.4716,
+      "step": 1012
+    },
+    {
+      "epoch": 0.30369121229154955,
+      "grad_norm": 0.23309756815433502,
+      "learning_rate": 0.00025481981981981977,
+      "loss": 0.467,
+      "step": 1013
+    },
+    {
+      "epoch": 0.30399100618324904,
+      "grad_norm": 0.2582738995552063,
+      "learning_rate": 0.00025477477477477476,
+      "loss": 0.4756,
+      "step": 1014
+    },
+    {
+      "epoch": 0.30429080007494846,
+      "grad_norm": 0.21543192863464355,
+      "learning_rate": 0.0002547297297297297,
+      "loss": 0.4184,
+      "step": 1015
+    },
+    {
+      "epoch": 0.30459059396664795,
+      "grad_norm": 0.22537867724895477,
+      "learning_rate": 0.00025468468468468463,
+      "loss": 0.451,
+      "step": 1016
+    },
+    {
+      "epoch": 0.3048903878583474,
+      "grad_norm": 0.22783374786376953,
+      "learning_rate": 0.0002546396396396396,
+      "loss": 0.4595,
+      "step": 1017
+    },
+    {
+      "epoch": 0.30519018175004686,
+      "grad_norm": 0.2382606863975525,
+      "learning_rate": 0.00025459459459459456,
+      "loss": 0.4465,
+      "step": 1018
+    },
+    {
+      "epoch": 0.3054899756417463,
+      "grad_norm": 0.23583681881427765,
+      "learning_rate": 0.0002545495495495495,
+      "loss": 0.453,
+      "step": 1019
+    },
+    {
+      "epoch": 0.3057897695334458,
+      "grad_norm": 0.24536754190921783,
+      "learning_rate": 0.0002545045045045045,
+      "loss": 0.4829,
+      "step": 1020
+    },
+    {
+      "epoch": 0.3060895634251452,
+      "grad_norm": 0.21961447596549988,
+      "learning_rate": 0.0002544594594594594,
+      "loss": 0.4555,
+      "step": 1021
+    },
+    {
+      "epoch": 0.3063893573168447,
+      "grad_norm": 0.22678659856319427,
+      "learning_rate": 0.00025441441441441436,
+      "loss": 0.4423,
+      "step": 1022
+    },
+    {
+      "epoch": 0.3066891512085441,
+      "grad_norm": 0.23871202766895294,
+      "learning_rate": 0.00025436936936936935,
+      "loss": 0.4657,
+      "step": 1023
+    },
+    {
+      "epoch": 0.3069889451002436,
+      "grad_norm": 0.23637500405311584,
+      "learning_rate": 0.0002543243243243243,
+      "loss": 0.4665,
+      "step": 1024
+    },
+    {
+      "epoch": 0.30728873899194303,
+      "grad_norm": 0.22538509964942932,
+      "learning_rate": 0.0002542792792792792,
+      "loss": 0.4349,
+      "step": 1025
+    },
+    {
+      "epoch": 0.3075885328836425,
+      "grad_norm": 0.2449694573879242,
+      "learning_rate": 0.0002542342342342342,
+      "loss": 0.465,
+      "step": 1026
+    },
+    {
+      "epoch": 0.30788832677534195,
+      "grad_norm": 0.22211913764476776,
+      "learning_rate": 0.00025418918918918915,
+      "loss": 0.4276,
+      "step": 1027
+    },
+    {
+      "epoch": 0.30818812066704143,
+      "grad_norm": 0.25039684772491455,
+      "learning_rate": 0.0002541441441441441,
+      "loss": 0.4799,
+      "step": 1028
+    },
+    {
+      "epoch": 0.30848791455874086,
+      "grad_norm": 0.22903893887996674,
+      "learning_rate": 0.0002540990990990991,
+      "loss": 0.4679,
+      "step": 1029
+    },
+    {
+      "epoch": 0.30878770845044035,
+      "grad_norm": 0.2479073703289032,
+      "learning_rate": 0.000254054054054054,
+      "loss": 0.4722,
+      "step": 1030
+    },
+    {
+      "epoch": 0.3090875023421398,
+      "grad_norm": 0.22166845202445984,
+      "learning_rate": 0.00025400900900900895,
+      "loss": 0.4354,
+      "step": 1031
+    },
+    {
+      "epoch": 0.30938729623383926,
+      "grad_norm": 0.22735866904258728,
+      "learning_rate": 0.00025396396396396394,
+      "loss": 0.4344,
+      "step": 1032
+    },
+    {
+      "epoch": 0.3096870901255387,
+      "grad_norm": 0.2471131831407547,
+      "learning_rate": 0.0002539189189189189,
+      "loss": 0.4751,
+      "step": 1033
+    },
+    {
+      "epoch": 0.3099868840172382,
+      "grad_norm": 0.22858619689941406,
+      "learning_rate": 0.00025387387387387387,
+      "loss": 0.4589,
+      "step": 1034
+    },
+    {
+      "epoch": 0.3102866779089376,
+      "grad_norm": 0.2418094426393509,
+      "learning_rate": 0.0002538288288288288,
+      "loss": 0.4873,
+      "step": 1035
+    },
+    {
+      "epoch": 0.3105864718006371,
+      "grad_norm": 0.2313918024301529,
+      "learning_rate": 0.00025378378378378374,
+      "loss": 0.4819,
+      "step": 1036
+    },
+    {
+      "epoch": 0.3108862656923365,
+      "grad_norm": 0.22201773524284363,
+      "learning_rate": 0.00025373873873873874,
+      "loss": 0.4553,
+      "step": 1037
+    },
+    {
+      "epoch": 0.311186059584036,
+      "grad_norm": 0.2284671664237976,
+      "learning_rate": 0.00025369369369369367,
+      "loss": 0.4485,
+      "step": 1038
+    },
+    {
+      "epoch": 0.31148585347573543,
+      "grad_norm": 0.2529887855052948,
+      "learning_rate": 0.0002536486486486486,
+      "loss": 0.4674,
+      "step": 1039
+    },
+    {
+      "epoch": 0.3117856473674349,
+      "grad_norm": 0.23350189626216888,
+      "learning_rate": 0.0002536036036036036,
+      "loss": 0.4835,
+      "step": 1040
+    },
+    {
+      "epoch": 0.31208544125913434,
+      "grad_norm": 0.23258428275585175,
+      "learning_rate": 0.00025355855855855854,
+      "loss": 0.4458,
+      "step": 1041
+    },
+    {
+      "epoch": 0.31238523515083383,
+      "grad_norm": 0.23113112151622772,
+      "learning_rate": 0.0002535135135135135,
+      "loss": 0.4581,
+      "step": 1042
+    },
+    {
+      "epoch": 0.31268502904253326,
+      "grad_norm": 0.22711153328418732,
+      "learning_rate": 0.00025346846846846846,
+      "loss": 0.4485,
+      "step": 1043
+    },
+    {
+      "epoch": 0.31298482293423274,
+      "grad_norm": 0.23305024206638336,
+      "learning_rate": 0.0002534234234234234,
+      "loss": 0.4316,
+      "step": 1044
+    },
+    {
+      "epoch": 0.31328461682593217,
+      "grad_norm": 0.24723917245864868,
+      "learning_rate": 0.0002533783783783784,
+      "loss": 0.4512,
+      "step": 1045
+    },
+    {
+      "epoch": 0.3135844107176316,
+      "grad_norm": 0.21640846133232117,
+      "learning_rate": 0.00025333333333333333,
+      "loss": 0.4485,
+      "step": 1046
+    },
+    {
+      "epoch": 0.3138842046093311,
+      "grad_norm": 0.25021156668663025,
+      "learning_rate": 0.00025328828828828827,
+      "loss": 0.4708,
+      "step": 1047
+    },
+    {
+      "epoch": 0.3141839985010305,
+      "grad_norm": 0.24005773663520813,
+      "learning_rate": 0.00025324324324324326,
+      "loss": 0.4698,
+      "step": 1048
+    },
+    {
+      "epoch": 0.31448379239273,
+      "grad_norm": 0.24885396659374237,
+      "learning_rate": 0.0002531981981981982,
+      "loss": 0.4899,
+      "step": 1049
+    },
+    {
+      "epoch": 0.31478358628442943,
+      "grad_norm": 0.2413524091243744,
+      "learning_rate": 0.00025315315315315313,
+      "loss": 0.4776,
+      "step": 1050
+    },
+    {
+      "epoch": 0.3150833801761289,
+      "grad_norm": 0.25239062309265137,
+      "learning_rate": 0.0002531081081081081,
+      "loss": 0.4788,
+      "step": 1051
+    },
+    {
+      "epoch": 0.31538317406782834,
+      "grad_norm": 0.23389939963817596,
+      "learning_rate": 0.00025306306306306306,
+      "loss": 0.4299,
+      "step": 1052
+    },
+    {
+      "epoch": 0.3156829679595278,
+      "grad_norm": 0.2468218207359314,
+      "learning_rate": 0.000253018018018018,
+      "loss": 0.4759,
+      "step": 1053
+    },
+    {
+      "epoch": 0.31598276185122726,
+      "grad_norm": 0.2298142910003662,
+      "learning_rate": 0.000252972972972973,
+      "loss": 0.4658,
+      "step": 1054
+    },
+    {
+      "epoch": 0.31628255574292674,
+      "grad_norm": 0.22888216376304626,
+      "learning_rate": 0.0002529279279279279,
+      "loss": 0.4419,
+      "step": 1055
+    },
+    {
+      "epoch": 0.31658234963462617,
+      "grad_norm": 0.23855063319206238,
+      "learning_rate": 0.00025288288288288286,
+      "loss": 0.4642,
+      "step": 1056
+    },
+    {
+      "epoch": 0.31688214352632565,
+      "grad_norm": 0.24454447627067566,
+      "learning_rate": 0.00025283783783783785,
+      "loss": 0.4495,
+      "step": 1057
+    },
+    {
+      "epoch": 0.3171819374180251,
+      "grad_norm": 0.22794046998023987,
+      "learning_rate": 0.0002527927927927928,
+      "loss": 0.4474,
+      "step": 1058
+    },
+    {
+      "epoch": 0.31748173130972457,
+      "grad_norm": 0.248634934425354,
+      "learning_rate": 0.0002527477477477477,
+      "loss": 0.4694,
+      "step": 1059
+    },
+    {
+      "epoch": 0.317781525201424,
+      "grad_norm": 0.24363334476947784,
+      "learning_rate": 0.00025270270270270266,
+      "loss": 0.4549,
+      "step": 1060
+    },
+    {
+      "epoch": 0.3180813190931235,
+      "grad_norm": 0.23220765590667725,
+      "learning_rate": 0.00025265765765765765,
+      "loss": 0.4478,
+      "step": 1061
+    },
+    {
+      "epoch": 0.3183811129848229,
+      "grad_norm": 0.22161665558815002,
+      "learning_rate": 0.0002526126126126126,
+      "loss": 0.4549,
+      "step": 1062
+    },
+    {
+      "epoch": 0.3186809068765224,
+      "grad_norm": 0.24613521993160248,
+      "learning_rate": 0.0002525675675675675,
+      "loss": 0.4505,
+      "step": 1063
+    },
+    {
+      "epoch": 0.3189807007682218,
+      "grad_norm": 0.26228928565979004,
+      "learning_rate": 0.0002525225225225225,
+      "loss": 0.4878,
+      "step": 1064
+    },
+    {
+      "epoch": 0.3192804946599213,
+      "grad_norm": 0.2279721200466156,
+      "learning_rate": 0.00025247747747747745,
+      "loss": 0.4481,
+      "step": 1065
+    },
+    {
+      "epoch": 0.31958028855162074,
+      "grad_norm": 0.24583470821380615,
+      "learning_rate": 0.0002524324324324324,
+      "loss": 0.4643,
+      "step": 1066
+    },
+    {
+      "epoch": 0.3198800824433202,
+      "grad_norm": 0.24150992929935455,
+      "learning_rate": 0.0002523873873873874,
+      "loss": 0.4705,
+      "step": 1067
+    },
+    {
+      "epoch": 0.32017987633501965,
+      "grad_norm": 0.2419997900724411,
+      "learning_rate": 0.0002523423423423423,
+      "loss": 0.4648,
+      "step": 1068
+    },
+    {
+      "epoch": 0.32047967022671914,
+      "grad_norm": 0.26776254177093506,
+      "learning_rate": 0.00025229729729729725,
+      "loss": 0.5013,
+      "step": 1069
+    },
+    {
+      "epoch": 0.32077946411841857,
+      "grad_norm": 0.24678350985050201,
+      "learning_rate": 0.00025225225225225225,
+      "loss": 0.449,
+      "step": 1070
+    },
+    {
+      "epoch": 0.32107925801011805,
+      "grad_norm": 0.24101199209690094,
+      "learning_rate": 0.0002522072072072072,
+      "loss": 0.469,
+      "step": 1071
+    },
+    {
+      "epoch": 0.3213790519018175,
+      "grad_norm": 0.24230711162090302,
+      "learning_rate": 0.0002521621621621621,
+      "loss": 0.481,
+      "step": 1072
+    },
+    {
+      "epoch": 0.32167884579351697,
+      "grad_norm": 0.22988542914390564,
+      "learning_rate": 0.0002521171171171171,
+      "loss": 0.4442,
+      "step": 1073
+    },
+    {
+      "epoch": 0.3219786396852164,
+      "grad_norm": 0.23284588754177094,
+      "learning_rate": 0.00025207207207207205,
+      "loss": 0.4392,
+      "step": 1074
+    },
+    {
+      "epoch": 0.3222784335769159,
+      "grad_norm": 0.24554894864559174,
+      "learning_rate": 0.000252027027027027,
+      "loss": 0.4544,
+      "step": 1075
+    },
+    {
+      "epoch": 0.3225782274686153,
+      "grad_norm": 0.22776508331298828,
+      "learning_rate": 0.000251981981981982,
+      "loss": 0.4503,
+      "step": 1076
+    },
+    {
+      "epoch": 0.3228780213603148,
+      "grad_norm": 0.2508374750614166,
+      "learning_rate": 0.0002519369369369369,
+      "loss": 0.4779,
+      "step": 1077
+    },
+    {
+      "epoch": 0.3231778152520142,
+      "grad_norm": 0.22543244063854218,
+      "learning_rate": 0.00025189189189189185,
+      "loss": 0.4672,
+      "step": 1078
+    },
+    {
+      "epoch": 0.3234776091437137,
+      "grad_norm": 0.2409958839416504,
+      "learning_rate": 0.00025184684684684684,
+      "loss": 0.4631,
+      "step": 1079
+    },
+    {
+      "epoch": 0.32377740303541314,
+      "grad_norm": 0.2308938056230545,
+      "learning_rate": 0.0002518018018018018,
+      "loss": 0.4244,
+      "step": 1080
+    },
+    {
+      "epoch": 0.3240771969271126,
+      "grad_norm": 0.2354745715856552,
+      "learning_rate": 0.0002517567567567567,
+      "loss": 0.4499,
+      "step": 1081
+    },
+    {
+      "epoch": 0.32437699081881205,
+      "grad_norm": 0.24564653635025024,
+      "learning_rate": 0.0002517117117117117,
+      "loss": 0.4655,
+      "step": 1082
+    },
+    {
+      "epoch": 0.32467678471051153,
+      "grad_norm": 0.2388393133878708,
+      "learning_rate": 0.00025166666666666664,
+      "loss": 0.4747,
+      "step": 1083
+    },
+    {
+      "epoch": 0.32497657860221096,
+      "grad_norm": 0.23941588401794434,
+      "learning_rate": 0.0002516216216216216,
+      "loss": 0.4646,
+      "step": 1084
+    },
+    {
+      "epoch": 0.32527637249391045,
+      "grad_norm": 0.24191126227378845,
+      "learning_rate": 0.0002515765765765765,
+      "loss": 0.4686,
+      "step": 1085
+    },
+    {
+      "epoch": 0.3255761663856099,
+      "grad_norm": 0.2466372847557068,
+      "learning_rate": 0.0002515315315315315,
+      "loss": 0.4989,
+      "step": 1086
+    },
+    {
+      "epoch": 0.32587596027730936,
+      "grad_norm": 0.2441006302833557,
+      "learning_rate": 0.00025148648648648644,
+      "loss": 0.4562,
+      "step": 1087
+    },
+    {
+      "epoch": 0.3261757541690088,
+      "grad_norm": 0.26483842730522156,
+      "learning_rate": 0.0002514414414414414,
+      "loss": 0.4872,
+      "step": 1088
+    },
+    {
+      "epoch": 0.3264755480607083,
+      "grad_norm": 0.2481161653995514,
+      "learning_rate": 0.00025139639639639637,
+      "loss": 0.4486,
+      "step": 1089
+    },
+    {
+      "epoch": 0.3267753419524077,
+      "grad_norm": 0.23705454170703888,
+      "learning_rate": 0.0002513513513513513,
+      "loss": 0.4407,
+      "step": 1090
+    },
+    {
+      "epoch": 0.3270751358441072,
+      "grad_norm": 0.25678539276123047,
+      "learning_rate": 0.0002513063063063063,
+      "loss": 0.4899,
+      "step": 1091
+    },
+    {
+      "epoch": 0.3273749297358066,
+      "grad_norm": 0.22578591108322144,
+      "learning_rate": 0.00025126126126126123,
+      "loss": 0.4326,
+      "step": 1092
+    },
+    {
+      "epoch": 0.3276747236275061,
+      "grad_norm": 0.23661458492279053,
+      "learning_rate": 0.00025121621621621617,
+      "loss": 0.4649,
+      "step": 1093
+    },
+    {
+      "epoch": 0.32797451751920553,
+      "grad_norm": 0.2496035248041153,
+      "learning_rate": 0.00025117117117117116,
+      "loss": 0.4978,
+      "step": 1094
+    },
+    {
+      "epoch": 0.328274311410905,
+      "grad_norm": 0.225214421749115,
+      "learning_rate": 0.0002511261261261261,
+      "loss": 0.4608,
+      "step": 1095
+    },
+    {
+      "epoch": 0.32857410530260445,
+      "grad_norm": 0.24089965224266052,
+      "learning_rate": 0.00025108108108108103,
+      "loss": 0.4854,
+      "step": 1096
+    },
+    {
+      "epoch": 0.32887389919430393,
+      "grad_norm": 0.23737536370754242,
+      "learning_rate": 0.000251036036036036,
+      "loss": 0.4692,
+      "step": 1097
+    },
+    {
+      "epoch": 0.32917369308600336,
+      "grad_norm": 0.23569715023040771,
+      "learning_rate": 0.00025099099099099096,
+      "loss": 0.44,
+      "step": 1098
+    },
+    {
+      "epoch": 0.32947348697770285,
+      "grad_norm": 0.22477473318576813,
+      "learning_rate": 0.0002509459459459459,
+      "loss": 0.4168,
+      "step": 1099
+    },
+    {
+      "epoch": 0.3297732808694023,
+      "grad_norm": 0.25336310267448425,
+      "learning_rate": 0.0002509009009009009,
+      "loss": 0.4493,
+      "step": 1100
+    },
+    {
+      "epoch": 0.33007307476110176,
+      "grad_norm": 0.2452186793088913,
+      "learning_rate": 0.00025085585585585583,
+      "loss": 0.484,
+      "step": 1101
+    },
+    {
+      "epoch": 0.3303728686528012,
+      "grad_norm": 0.23870813846588135,
+      "learning_rate": 0.0002508108108108108,
+      "loss": 0.4228,
+      "step": 1102
+    },
+    {
+      "epoch": 0.3306726625445007,
+      "grad_norm": 0.2262556552886963,
+      "learning_rate": 0.00025076576576576575,
+      "loss": 0.4525,
+      "step": 1103
+    },
+    {
+      "epoch": 0.3309724564362001,
+      "grad_norm": 0.2720157504081726,
+      "learning_rate": 0.0002507207207207207,
+      "loss": 0.4961,
+      "step": 1104
+    },
+    {
+      "epoch": 0.3312722503278996,
+      "grad_norm": 0.23624925315380096,
+      "learning_rate": 0.0002506756756756757,
+      "loss": 0.4524,
+      "step": 1105
+    },
+    {
+      "epoch": 0.331572044219599,
+      "grad_norm": 0.2634661793708801,
+      "learning_rate": 0.0002506306306306306,
+      "loss": 0.4603,
+      "step": 1106
+    },
+    {
+      "epoch": 0.3318718381112985,
+      "grad_norm": 0.24629558622837067,
+      "learning_rate": 0.00025058558558558556,
+      "loss": 0.453,
+      "step": 1107
+    },
+    {
+      "epoch": 0.33217163200299793,
+      "grad_norm": 0.2503926157951355,
+      "learning_rate": 0.00025054054054054055,
+      "loss": 0.4756,
+      "step": 1108
+    },
+    {
+      "epoch": 0.3324714258946974,
+      "grad_norm": 0.22624404728412628,
+      "learning_rate": 0.0002504954954954955,
+      "loss": 0.4391,
+      "step": 1109
+    },
+    {
+      "epoch": 0.33277121978639684,
+      "grad_norm": 0.22146141529083252,
+      "learning_rate": 0.0002504504504504504,
+      "loss": 0.4496,
+      "step": 1110
+    },
+    {
+      "epoch": 0.33307101367809633,
+      "grad_norm": 0.23911216855049133,
+      "learning_rate": 0.0002504054054054054,
+      "loss": 0.4584,
+      "step": 1111
+    },
+    {
+      "epoch": 0.33337080756979576,
+      "grad_norm": 0.23537759482860565,
+      "learning_rate": 0.00025036036036036035,
+      "loss": 0.4735,
+      "step": 1112
+    },
+    {
+      "epoch": 0.33367060146149524,
+      "grad_norm": 0.23113307356834412,
+      "learning_rate": 0.0002503153153153153,
+      "loss": 0.4393,
+      "step": 1113
+    },
+    {
+      "epoch": 0.33397039535319467,
+      "grad_norm": 0.24185238778591156,
+      "learning_rate": 0.0002502702702702703,
+      "loss": 0.4661,
+      "step": 1114
+    },
+    {
+      "epoch": 0.33427018924489416,
+      "grad_norm": 0.25618115067481995,
+      "learning_rate": 0.0002502252252252252,
+      "loss": 0.4777,
+      "step": 1115
+    },
+    {
+      "epoch": 0.3345699831365936,
+      "grad_norm": 0.24486567080020905,
+      "learning_rate": 0.00025018018018018015,
+      "loss": 0.4589,
+      "step": 1116
+    },
+    {
+      "epoch": 0.33486977702829307,
+      "grad_norm": 0.2608473300933838,
+      "learning_rate": 0.00025013513513513514,
+      "loss": 0.4716,
+      "step": 1117
+    },
+    {
+      "epoch": 0.3351695709199925,
+      "grad_norm": 0.2427588254213333,
+      "learning_rate": 0.0002500900900900901,
+      "loss": 0.4856,
+      "step": 1118
+    },
+    {
+      "epoch": 0.335469364811692,
+      "grad_norm": 0.2493797391653061,
+      "learning_rate": 0.000250045045045045,
+      "loss": 0.4925,
+      "step": 1119
+    },
+    {
+      "epoch": 0.3357691587033914,
+      "grad_norm": 0.2610112428665161,
+      "learning_rate": 0.00025,
+      "loss": 0.4685,
+      "step": 1120
+    },
+    {
+      "epoch": 0.3360689525950909,
+      "grad_norm": 0.2435634434223175,
+      "learning_rate": 0.00024995495495495494,
+      "loss": 0.4678,
+      "step": 1121
+    },
+    {
+      "epoch": 0.3363687464867903,
+      "grad_norm": 0.24447932839393616,
+      "learning_rate": 0.0002499099099099099,
+      "loss": 0.4656,
+      "step": 1122
+    },
+    {
+      "epoch": 0.3366685403784898,
+      "grad_norm": 0.24005521833896637,
+      "learning_rate": 0.00024986486486486487,
+      "loss": 0.4712,
+      "step": 1123
+    },
+    {
+      "epoch": 0.33696833427018924,
+      "grad_norm": 0.2303832322359085,
+      "learning_rate": 0.0002498198198198198,
+      "loss": 0.4527,
+      "step": 1124
+    },
+    {
+      "epoch": 0.3372681281618887,
+      "grad_norm": 0.221012681722641,
+      "learning_rate": 0.00024977477477477474,
+      "loss": 0.4376,
+      "step": 1125
+    },
+    {
+      "epoch": 0.33756792205358815,
+      "grad_norm": 0.23421809077262878,
+      "learning_rate": 0.00024972972972972973,
+      "loss": 0.4449,
+      "step": 1126
+    },
+    {
+      "epoch": 0.33786771594528764,
+      "grad_norm": 0.23941418528556824,
+      "learning_rate": 0.00024968468468468467,
+      "loss": 0.4679,
+      "step": 1127
+    },
+    {
+      "epoch": 0.33816750983698707,
+      "grad_norm": 0.2479025423526764,
+      "learning_rate": 0.0002496396396396396,
+      "loss": 0.4872,
+      "step": 1128
+    },
+    {
+      "epoch": 0.33846730372868655,
+      "grad_norm": 0.24516451358795166,
+      "learning_rate": 0.0002495945945945946,
+      "loss": 0.4488,
+      "step": 1129
+    },
+    {
+      "epoch": 0.338767097620386,
+      "grad_norm": 0.2436760663986206,
+      "learning_rate": 0.00024954954954954954,
+      "loss": 0.4477,
+      "step": 1130
+    },
+    {
+      "epoch": 0.33906689151208547,
+      "grad_norm": 0.23894813656806946,
+      "learning_rate": 0.00024950450450450447,
+      "loss": 0.4295,
+      "step": 1131
+    },
+    {
+      "epoch": 0.3393666854037849,
+      "grad_norm": 0.24569731950759888,
+      "learning_rate": 0.00024945945945945946,
+      "loss": 0.46,
+      "step": 1132
+    },
+    {
+      "epoch": 0.3396664792954844,
+      "grad_norm": 0.24807578325271606,
+      "learning_rate": 0.0002494144144144144,
+      "loss": 0.4512,
+      "step": 1133
+    },
+    {
+      "epoch": 0.3399662731871838,
+      "grad_norm": 0.23641791939735413,
+      "learning_rate": 0.00024936936936936934,
+      "loss": 0.454,
+      "step": 1134
+    },
+    {
+      "epoch": 0.3402660670788833,
+      "grad_norm": 0.26076388359069824,
+      "learning_rate": 0.0002493243243243243,
+      "loss": 0.4771,
+      "step": 1135
+    },
+    {
+      "epoch": 0.3405658609705827,
+      "grad_norm": 0.24686305224895477,
+      "learning_rate": 0.00024927927927927926,
+      "loss": 0.4783,
+      "step": 1136
+    },
+    {
+      "epoch": 0.3408656548622822,
+      "grad_norm": 0.2262791097164154,
+      "learning_rate": 0.0002492342342342342,
+      "loss": 0.4272,
+      "step": 1137
+    },
+    {
+      "epoch": 0.34116544875398164,
+      "grad_norm": 0.23418664932250977,
+      "learning_rate": 0.00024918918918918914,
+      "loss": 0.4666,
+      "step": 1138
+    },
+    {
+      "epoch": 0.3414652426456811,
+      "grad_norm": 0.23737958073616028,
+      "learning_rate": 0.00024914414414414413,
+      "loss": 0.4433,
+      "step": 1139
+    },
+    {
+      "epoch": 0.34176503653738055,
+      "grad_norm": 0.2579478919506073,
+      "learning_rate": 0.00024909909909909907,
+      "loss": 0.4702,
+      "step": 1140
+    },
+    {
+      "epoch": 0.34206483042908,
+      "grad_norm": 0.2627730667591095,
+      "learning_rate": 0.000249054054054054,
+      "loss": 0.4607,
+      "step": 1141
+    },
+    {
+      "epoch": 0.34236462432077946,
+      "grad_norm": 0.2283281534910202,
+      "learning_rate": 0.000249009009009009,
+      "loss": 0.4462,
+      "step": 1142
+    },
+    {
+      "epoch": 0.3426644182124789,
+      "grad_norm": 0.24083632230758667,
+      "learning_rate": 0.00024896396396396393,
+      "loss": 0.4663,
+      "step": 1143
+    },
+    {
+      "epoch": 0.3429642121041784,
+      "grad_norm": 0.24331289529800415,
+      "learning_rate": 0.00024891891891891887,
+      "loss": 0.4414,
+      "step": 1144
+    },
+    {
+      "epoch": 0.3432640059958778,
+      "grad_norm": 0.24059659242630005,
+      "learning_rate": 0.00024887387387387386,
+      "loss": 0.4728,
+      "step": 1145
+    },
+    {
+      "epoch": 0.3435637998875773,
+      "grad_norm": 0.23175103962421417,
+      "learning_rate": 0.0002488288288288288,
+      "loss": 0.437,
+      "step": 1146
+    },
+    {
+      "epoch": 0.3438635937792767,
+      "grad_norm": 0.23247724771499634,
+      "learning_rate": 0.00024878378378378373,
+      "loss": 0.4454,
+      "step": 1147
+    },
+    {
+      "epoch": 0.3441633876709762,
+      "grad_norm": 0.22808894515037537,
+      "learning_rate": 0.0002487387387387387,
+      "loss": 0.4379,
+      "step": 1148
+    },
+    {
+      "epoch": 0.34446318156267564,
+      "grad_norm": 0.2515697777271271,
+      "learning_rate": 0.00024869369369369366,
+      "loss": 0.4938,
+      "step": 1149
+    },
+    {
+      "epoch": 0.3447629754543751,
+      "grad_norm": 0.22830873727798462,
+      "learning_rate": 0.0002486486486486486,
+      "loss": 0.4335,
+      "step": 1150
+    },
+    {
+      "epoch": 0.34506276934607455,
+      "grad_norm": 0.2352674901485443,
+      "learning_rate": 0.0002486036036036036,
+      "loss": 0.4661,
+      "step": 1151
+    },
+    {
+      "epoch": 0.34536256323777403,
+      "grad_norm": 0.233395054936409,
+      "learning_rate": 0.0002485585585585585,
+      "loss": 0.4429,
+      "step": 1152
+    },
+    {
+      "epoch": 0.34566235712947346,
+      "grad_norm": 0.2488911747932434,
+      "learning_rate": 0.00024851351351351346,
+      "loss": 0.4832,
+      "step": 1153
+    },
+    {
+      "epoch": 0.34596215102117295,
+      "grad_norm": 0.2402927577495575,
+      "learning_rate": 0.00024846846846846845,
+      "loss": 0.4633,
+      "step": 1154
+    },
+    {
+      "epoch": 0.3462619449128724,
+      "grad_norm": 0.23628897964954376,
+      "learning_rate": 0.0002484234234234234,
+      "loss": 0.4683,
+      "step": 1155
+    },
+    {
+      "epoch": 0.34656173880457186,
+      "grad_norm": 0.23966571688652039,
+      "learning_rate": 0.0002483783783783783,
+      "loss": 0.4716,
+      "step": 1156
+    },
+    {
+      "epoch": 0.3468615326962713,
+      "grad_norm": 0.23786477744579315,
+      "learning_rate": 0.0002483333333333333,
+      "loss": 0.4829,
+      "step": 1157
+    },
+    {
+      "epoch": 0.3471613265879708,
+      "grad_norm": 0.24706590175628662,
+      "learning_rate": 0.00024828828828828825,
+      "loss": 0.4741,
+      "step": 1158
+    },
+    {
+      "epoch": 0.3474611204796702,
+      "grad_norm": 0.25192269682884216,
+      "learning_rate": 0.00024824324324324324,
+      "loss": 0.4725,
+      "step": 1159
+    },
+    {
+      "epoch": 0.3477609143713697,
+      "grad_norm": 0.2672080099582672,
+      "learning_rate": 0.0002481981981981982,
+      "loss": 0.4677,
+      "step": 1160
+    },
+    {
+      "epoch": 0.3480607082630691,
+      "grad_norm": 0.23710590600967407,
+      "learning_rate": 0.0002481531531531531,
+      "loss": 0.4629,
+      "step": 1161
+    },
+    {
+      "epoch": 0.3483605021547686,
+      "grad_norm": 0.2306007295846939,
+      "learning_rate": 0.0002481081081081081,
+      "loss": 0.4425,
+      "step": 1162
+    },
+    {
+      "epoch": 0.34866029604646803,
+      "grad_norm": 0.2416974902153015,
+      "learning_rate": 0.00024806306306306305,
+      "loss": 0.4452,
+      "step": 1163
+    },
+    {
+      "epoch": 0.3489600899381675,
+      "grad_norm": 0.2495068907737732,
+      "learning_rate": 0.000248018018018018,
+      "loss": 0.4875,
+      "step": 1164
+    },
+    {
+      "epoch": 0.34925988382986695,
+      "grad_norm": 0.26994964480400085,
+      "learning_rate": 0.00024797297297297297,
+      "loss": 0.4957,
+      "step": 1165
+    },
+    {
+      "epoch": 0.34955967772156643,
+      "grad_norm": 0.2546437978744507,
+      "learning_rate": 0.0002479279279279279,
+      "loss": 0.498,
+      "step": 1166
+    },
+    {
+      "epoch": 0.34985947161326586,
+      "grad_norm": 0.2271340936422348,
+      "learning_rate": 0.00024788288288288285,
+      "loss": 0.4615,
+      "step": 1167
+    },
+    {
+      "epoch": 0.35015926550496534,
+      "grad_norm": 0.24870999157428741,
+      "learning_rate": 0.00024783783783783784,
+      "loss": 0.4688,
+      "step": 1168
+    },
+    {
+      "epoch": 0.3504590593966648,
+      "grad_norm": 0.23978911340236664,
+      "learning_rate": 0.0002477927927927928,
+      "loss": 0.4882,
+      "step": 1169
+    },
+    {
+      "epoch": 0.35075885328836426,
+      "grad_norm": 0.2773337662220001,
+      "learning_rate": 0.0002477477477477477,
+      "loss": 0.4695,
+      "step": 1170
+    },
+    {
+      "epoch": 0.3510586471800637,
+      "grad_norm": 0.24570350348949432,
+      "learning_rate": 0.0002477027027027027,
+      "loss": 0.4679,
+      "step": 1171
+    },
+    {
+      "epoch": 0.3513584410717632,
+      "grad_norm": 0.25563982129096985,
+      "learning_rate": 0.00024765765765765764,
+      "loss": 0.4731,
+      "step": 1172
+    },
+    {
+      "epoch": 0.3516582349634626,
+      "grad_norm": 0.23189115524291992,
+      "learning_rate": 0.00024761261261261263,
+      "loss": 0.4476,
+      "step": 1173
+    },
+    {
+      "epoch": 0.3519580288551621,
+      "grad_norm": 0.24074453115463257,
+      "learning_rate": 0.00024756756756756757,
+      "loss": 0.4419,
+      "step": 1174
+    },
+    {
+      "epoch": 0.3522578227468615,
+      "grad_norm": 0.2376662790775299,
+      "learning_rate": 0.0002475225225225225,
+      "loss": 0.4571,
+      "step": 1175
+    },
+    {
+      "epoch": 0.352557616638561,
+      "grad_norm": 0.2344047725200653,
+      "learning_rate": 0.0002474774774774775,
+      "loss": 0.4389,
+      "step": 1176
+    },
+    {
+      "epoch": 0.35285741053026043,
+      "grad_norm": 0.23310165107250214,
+      "learning_rate": 0.00024743243243243243,
+      "loss": 0.4583,
+      "step": 1177
+    },
+    {
+      "epoch": 0.3531572044219599,
+      "grad_norm": 0.21277011930942535,
+      "learning_rate": 0.00024738738738738737,
+      "loss": 0.4306,
+      "step": 1178
+    },
+    {
+      "epoch": 0.35345699831365934,
+      "grad_norm": 0.23581352829933167,
+      "learning_rate": 0.00024734234234234236,
+      "loss": 0.475,
+      "step": 1179
+    },
+    {
+      "epoch": 0.3537567922053588,
+      "grad_norm": 0.23194879293441772,
+      "learning_rate": 0.0002472972972972973,
+      "loss": 0.4518,
+      "step": 1180
+    },
+    {
+      "epoch": 0.35405658609705826,
+      "grad_norm": 0.22603453695774078,
+      "learning_rate": 0.00024725225225225223,
+      "loss": 0.4464,
+      "step": 1181
+    },
+    {
+      "epoch": 0.35435637998875774,
+      "grad_norm": 0.23987054824829102,
+      "learning_rate": 0.00024720720720720717,
+      "loss": 0.4586,
+      "step": 1182
+    },
+    {
+      "epoch": 0.35465617388045717,
+      "grad_norm": 0.22986359894275665,
+      "learning_rate": 0.00024716216216216216,
+      "loss": 0.4664,
+      "step": 1183
+    },
+    {
+      "epoch": 0.35495596777215666,
+      "grad_norm": 0.22636739909648895,
+      "learning_rate": 0.0002471171171171171,
+      "loss": 0.4236,
+      "step": 1184
+    },
+    {
+      "epoch": 0.3552557616638561,
+      "grad_norm": 0.2346397340297699,
+      "learning_rate": 0.00024707207207207203,
+      "loss": 0.4703,
+      "step": 1185
+    },
+    {
+      "epoch": 0.35555555555555557,
+      "grad_norm": 0.2564719617366791,
+      "learning_rate": 0.000247027027027027,
+      "loss": 0.4775,
+      "step": 1186
+    },
+    {
+      "epoch": 0.355855349447255,
+      "grad_norm": 0.22305525839328766,
+      "learning_rate": 0.00024698198198198196,
+      "loss": 0.4694,
+      "step": 1187
+    },
+    {
+      "epoch": 0.3561551433389545,
+      "grad_norm": 0.2369467169046402,
+      "learning_rate": 0.0002469369369369369,
+      "loss": 0.4498,
+      "step": 1188
+    },
+    {
+      "epoch": 0.3564549372306539,
+      "grad_norm": 0.25123798847198486,
+      "learning_rate": 0.0002468918918918919,
+      "loss": 0.4619,
+      "step": 1189
+    },
+    {
+      "epoch": 0.3567547311223534,
+      "grad_norm": 0.21925069391727448,
+      "learning_rate": 0.0002468468468468468,
+      "loss": 0.4498,
+      "step": 1190
+    },
+    {
+      "epoch": 0.3570545250140528,
+      "grad_norm": 0.2385261207818985,
+      "learning_rate": 0.00024680180180180176,
+      "loss": 0.4537,
+      "step": 1191
+    },
+    {
+      "epoch": 0.3573543189057523,
+      "grad_norm": 0.23894301056861877,
+      "learning_rate": 0.00024675675675675675,
+      "loss": 0.4665,
+      "step": 1192
+    },
+    {
+      "epoch": 0.35765411279745174,
+      "grad_norm": 0.23315206170082092,
+      "learning_rate": 0.0002467117117117117,
+      "loss": 0.4421,
+      "step": 1193
+    },
+    {
+      "epoch": 0.3579539066891512,
+      "grad_norm": 0.23406696319580078,
+      "learning_rate": 0.0002466666666666666,
+      "loss": 0.4377,
+      "step": 1194
+    },
+    {
+      "epoch": 0.35825370058085065,
+      "grad_norm": 0.25852885842323303,
+      "learning_rate": 0.0002466216216216216,
+      "loss": 0.4838,
+      "step": 1195
+    },
+    {
+      "epoch": 0.35855349447255014,
+      "grad_norm": 0.24008771777153015,
+      "learning_rate": 0.00024657657657657655,
+      "loss": 0.4733,
+      "step": 1196
+    },
+    {
+      "epoch": 0.35885328836424957,
+      "grad_norm": 0.228665292263031,
+      "learning_rate": 0.0002465315315315315,
+      "loss": 0.4753,
+      "step": 1197
+    },
+    {
+      "epoch": 0.35915308225594905,
+      "grad_norm": 0.2344791293144226,
+      "learning_rate": 0.0002464864864864865,
+      "loss": 0.4484,
+      "step": 1198
+    },
+    {
+      "epoch": 0.3594528761476485,
+      "grad_norm": 0.22843588888645172,
+      "learning_rate": 0.0002464414414414414,
+      "loss": 0.4349,
+      "step": 1199
+    },
+    {
+      "epoch": 0.35975267003934797,
+      "grad_norm": 0.23127557337284088,
+      "learning_rate": 0.00024639639639639636,
+      "loss": 0.4466,
+      "step": 1200
+    },
+    {
+      "epoch": 0.3600524639310474,
+      "grad_norm": 0.2092585414648056,
+      "learning_rate": 0.00024635135135135135,
+      "loss": 0.4101,
+      "step": 1201
+    },
+    {
+      "epoch": 0.3603522578227469,
+      "grad_norm": 0.24416851997375488,
+      "learning_rate": 0.0002463063063063063,
+      "loss": 0.4507,
+      "step": 1202
+    },
+    {
+      "epoch": 0.3606520517144463,
+      "grad_norm": 0.2409181445837021,
+      "learning_rate": 0.0002462612612612612,
+      "loss": 0.4641,
+      "step": 1203
+    },
+    {
+      "epoch": 0.3609518456061458,
+      "grad_norm": 0.2390405684709549,
+      "learning_rate": 0.0002462162162162162,
+      "loss": 0.4398,
+      "step": 1204
+    },
+    {
+      "epoch": 0.3612516394978452,
+      "grad_norm": 0.25821688771247864,
+      "learning_rate": 0.00024617117117117115,
+      "loss": 0.4441,
+      "step": 1205
+    },
+    {
+      "epoch": 0.3615514333895447,
+      "grad_norm": 0.24180777370929718,
+      "learning_rate": 0.0002461261261261261,
+      "loss": 0.4852,
+      "step": 1206
+    },
+    {
+      "epoch": 0.36185122728124414,
+      "grad_norm": 0.2260608822107315,
+      "learning_rate": 0.000246081081081081,
+      "loss": 0.4214,
+      "step": 1207
+    },
+    {
+      "epoch": 0.3621510211729436,
+      "grad_norm": 0.2266250103712082,
+      "learning_rate": 0.000246036036036036,
+      "loss": 0.4104,
+      "step": 1208
+    },
+    {
+      "epoch": 0.36245081506464305,
+      "grad_norm": 0.247540682554245,
+      "learning_rate": 0.00024599099099099095,
+      "loss": 0.4563,
+      "step": 1209
+    },
+    {
+      "epoch": 0.36275060895634254,
+      "grad_norm": 0.22714072465896606,
+      "learning_rate": 0.0002459459459459459,
+      "loss": 0.4528,
+      "step": 1210
+    },
+    {
+      "epoch": 0.36305040284804196,
+      "grad_norm": 0.22302433848381042,
+      "learning_rate": 0.0002459009009009009,
+      "loss": 0.3982,
+      "step": 1211
+    },
+    {
+      "epoch": 0.36335019673974145,
+      "grad_norm": 0.2646171748638153,
+      "learning_rate": 0.0002458558558558558,
+      "loss": 0.4837,
+      "step": 1212
+    },
+    {
+      "epoch": 0.3636499906314409,
+      "grad_norm": 0.24546460807323456,
+      "learning_rate": 0.00024581081081081075,
+      "loss": 0.4716,
+      "step": 1213
+    },
+    {
+      "epoch": 0.36394978452314036,
+      "grad_norm": 0.2416929006576538,
+      "learning_rate": 0.00024576576576576574,
+      "loss": 0.4634,
+      "step": 1214
+    },
+    {
+      "epoch": 0.3642495784148398,
+      "grad_norm": 0.2360236495733261,
+      "learning_rate": 0.0002457207207207207,
+      "loss": 0.4409,
+      "step": 1215
+    },
+    {
+      "epoch": 0.3645493723065393,
+      "grad_norm": 0.24383249878883362,
+      "learning_rate": 0.00024567567567567567,
+      "loss": 0.4553,
+      "step": 1216
+    },
+    {
+      "epoch": 0.3648491661982387,
+      "grad_norm": 0.2516370117664337,
+      "learning_rate": 0.0002456306306306306,
+      "loss": 0.4553,
+      "step": 1217
+    },
+    {
+      "epoch": 0.3651489600899382,
+      "grad_norm": 0.2524015009403229,
+      "learning_rate": 0.00024558558558558554,
+      "loss": 0.4766,
+      "step": 1218
+    },
+    {
+      "epoch": 0.3654487539816376,
+      "grad_norm": 0.23386207222938538,
+      "learning_rate": 0.00024554054054054053,
+      "loss": 0.439,
+      "step": 1219
+    },
+    {
+      "epoch": 0.3657485478733371,
+      "grad_norm": 0.23544320464134216,
+      "learning_rate": 0.00024549549549549547,
+      "loss": 0.4414,
+      "step": 1220
+    },
+    {
+      "epoch": 0.36604834176503653,
+      "grad_norm": 0.24809177219867706,
+      "learning_rate": 0.0002454504504504504,
+      "loss": 0.4577,
+      "step": 1221
+    },
+    {
+      "epoch": 0.366348135656736,
+      "grad_norm": 0.24466872215270996,
+      "learning_rate": 0.0002454054054054054,
+      "loss": 0.4609,
+      "step": 1222
+    },
+    {
+      "epoch": 0.36664792954843545,
+      "grad_norm": 0.24159879982471466,
+      "learning_rate": 0.00024536036036036034,
+      "loss": 0.448,
+      "step": 1223
+    },
+    {
+      "epoch": 0.36694772344013493,
+      "grad_norm": 0.2456122189760208,
+      "learning_rate": 0.00024531531531531527,
+      "loss": 0.4563,
+      "step": 1224
+    },
+    {
+      "epoch": 0.36724751733183436,
+      "grad_norm": 0.23266494274139404,
+      "learning_rate": 0.00024527027027027026,
+      "loss": 0.457,
+      "step": 1225
+    },
+    {
+      "epoch": 0.36754731122353385,
+      "grad_norm": 0.24822424352169037,
+      "learning_rate": 0.0002452252252252252,
+      "loss": 0.4577,
+      "step": 1226
+    },
+    {
+      "epoch": 0.3678471051152333,
+      "grad_norm": 0.2528662383556366,
+      "learning_rate": 0.00024518018018018014,
+      "loss": 0.4717,
+      "step": 1227
+    },
+    {
+      "epoch": 0.36814689900693276,
+      "grad_norm": 0.22255463898181915,
+      "learning_rate": 0.00024513513513513513,
+      "loss": 0.4287,
+      "step": 1228
+    },
+    {
+      "epoch": 0.3684466928986322,
+      "grad_norm": 0.23769044876098633,
+      "learning_rate": 0.00024509009009009006,
+      "loss": 0.4348,
+      "step": 1229
+    },
+    {
+      "epoch": 0.3687464867903317,
+      "grad_norm": 0.23163922131061554,
+      "learning_rate": 0.00024504504504504506,
+      "loss": 0.4631,
+      "step": 1230
+    },
+    {
+      "epoch": 0.3690462806820311,
+      "grad_norm": 0.22742880880832672,
+      "learning_rate": 0.000245,
+      "loss": 0.4344,
+      "step": 1231
+    },
+    {
+      "epoch": 0.3693460745737306,
+      "grad_norm": 0.2314460575580597,
+      "learning_rate": 0.00024495495495495493,
+      "loss": 0.4507,
+      "step": 1232
+    },
+    {
+      "epoch": 0.36964586846543,
+      "grad_norm": 0.21651825308799744,
+      "learning_rate": 0.0002449099099099099,
+      "loss": 0.422,
+      "step": 1233
+    },
+    {
+      "epoch": 0.3699456623571295,
+      "grad_norm": 0.24304571747779846,
+      "learning_rate": 0.00024486486486486486,
+      "loss": 0.4322,
+      "step": 1234
+    },
+    {
+      "epoch": 0.37024545624882893,
+      "grad_norm": 0.24105508625507355,
+      "learning_rate": 0.0002448198198198198,
+      "loss": 0.4367,
+      "step": 1235
+    },
+    {
+      "epoch": 0.37054525014052836,
+      "grad_norm": 0.2495047152042389,
+      "learning_rate": 0.0002447747747747748,
+      "loss": 0.456,
+      "step": 1236
+    },
+    {
+      "epoch": 0.37084504403222784,
+      "grad_norm": 0.2545395791530609,
+      "learning_rate": 0.0002447297297297297,
+      "loss": 0.4608,
+      "step": 1237
+    },
+    {
+      "epoch": 0.3711448379239273,
+      "grad_norm": 0.2552499771118164,
+      "learning_rate": 0.00024468468468468466,
+      "loss": 0.4898,
+      "step": 1238
+    },
+    {
+      "epoch": 0.37144463181562676,
+      "grad_norm": 0.2469811737537384,
+      "learning_rate": 0.00024463963963963965,
+      "loss": 0.4427,
+      "step": 1239
+    },
+    {
+      "epoch": 0.3717444257073262,
+      "grad_norm": 0.23857302963733673,
+      "learning_rate": 0.0002445945945945946,
+      "loss": 0.4509,
+      "step": 1240
+    },
+    {
+      "epoch": 0.37204421959902567,
+      "grad_norm": 0.2521422803401947,
+      "learning_rate": 0.0002445495495495495,
+      "loss": 0.4425,
+      "step": 1241
+    },
+    {
+      "epoch": 0.3723440134907251,
+      "grad_norm": 0.24907280504703522,
+      "learning_rate": 0.0002445045045045045,
+      "loss": 0.4542,
+      "step": 1242
+    },
+    {
+      "epoch": 0.3726438073824246,
+      "grad_norm": 0.23783591389656067,
+      "learning_rate": 0.00024445945945945945,
+      "loss": 0.4831,
+      "step": 1243
+    },
+    {
+      "epoch": 0.372943601274124,
+      "grad_norm": 0.2376372069120407,
+      "learning_rate": 0.0002444144144144144,
+      "loss": 0.4514,
+      "step": 1244
+    },
+    {
+      "epoch": 0.3732433951658235,
+      "grad_norm": 0.2387792468070984,
+      "learning_rate": 0.0002443693693693694,
+      "loss": 0.4593,
+      "step": 1245
+    },
+    {
+      "epoch": 0.37354318905752293,
+      "grad_norm": 0.22432541847229004,
+      "learning_rate": 0.0002443243243243243,
+      "loss": 0.4462,
+      "step": 1246
+    },
+    {
+      "epoch": 0.3738429829492224,
+      "grad_norm": 0.24190527200698853,
+      "learning_rate": 0.00024427927927927925,
+      "loss": 0.4645,
+      "step": 1247
+    },
+    {
+      "epoch": 0.37414277684092184,
+      "grad_norm": 0.23738646507263184,
+      "learning_rate": 0.00024423423423423424,
+      "loss": 0.4594,
+      "step": 1248
+    },
+    {
+      "epoch": 0.3744425707326213,
+      "grad_norm": 0.24582220613956451,
+      "learning_rate": 0.0002441891891891892,
+      "loss": 0.4632,
+      "step": 1249
+    },
+    {
+      "epoch": 0.37474236462432076,
+      "grad_norm": 0.22717328369617462,
+      "learning_rate": 0.0002441441441441441,
+      "loss": 0.4372,
+      "step": 1250
+    },
+    {
+      "epoch": 0.37504215851602024,
+      "grad_norm": 0.24414947628974915,
+      "learning_rate": 0.0002440990990990991,
+      "loss": 0.4458,
+      "step": 1251
+    },
+    {
+      "epoch": 0.37534195240771967,
+      "grad_norm": 0.23710165917873383,
+      "learning_rate": 0.00024405405405405404,
+      "loss": 0.4468,
+      "step": 1252
+    },
+    {
+      "epoch": 0.37564174629941915,
+      "grad_norm": 0.25462841987609863,
+      "learning_rate": 0.00024400900900900898,
+      "loss": 0.4616,
+      "step": 1253
+    },
+    {
+      "epoch": 0.3759415401911186,
+      "grad_norm": 0.22636806964874268,
+      "learning_rate": 0.00024396396396396392,
+      "loss": 0.4259,
+      "step": 1254
+    },
+    {
+      "epoch": 0.37624133408281807,
+      "grad_norm": 0.24978788197040558,
+      "learning_rate": 0.0002439189189189189,
+      "loss": 0.464,
+      "step": 1255
+    },
+    {
+      "epoch": 0.3765411279745175,
+      "grad_norm": 0.2312556803226471,
+      "learning_rate": 0.00024387387387387384,
+      "loss": 0.4531,
+      "step": 1256
+    },
+    {
+      "epoch": 0.376840921866217,
+      "grad_norm": 0.22814960777759552,
+      "learning_rate": 0.00024382882882882878,
+      "loss": 0.4608,
+      "step": 1257
+    },
+    {
+      "epoch": 0.3771407157579164,
+      "grad_norm": 0.25135213136672974,
+      "learning_rate": 0.00024378378378378377,
+      "loss": 0.4551,
+      "step": 1258
+    },
+    {
+      "epoch": 0.3774405096496159,
+      "grad_norm": 0.2209721952676773,
+      "learning_rate": 0.0002437387387387387,
+      "loss": 0.4679,
+      "step": 1259
+    },
+    {
+      "epoch": 0.3777403035413153,
+      "grad_norm": 0.2256690412759781,
+      "learning_rate": 0.00024369369369369365,
+      "loss": 0.4388,
+      "step": 1260
+    },
+    {
+      "epoch": 0.3780400974330148,
+      "grad_norm": 0.23604658246040344,
+      "learning_rate": 0.00024364864864864864,
+      "loss": 0.4393,
+      "step": 1261
+    },
+    {
+      "epoch": 0.37833989132471424,
+      "grad_norm": 0.22875599563121796,
+      "learning_rate": 0.00024360360360360357,
+      "loss": 0.4176,
+      "step": 1262
+    },
+    {
+      "epoch": 0.3786396852164137,
+      "grad_norm": 0.2428806722164154,
+      "learning_rate": 0.00024355855855855854,
+      "loss": 0.461,
+      "step": 1263
+    },
+    {
+      "epoch": 0.37893947910811315,
+      "grad_norm": 0.2470446228981018,
+      "learning_rate": 0.0002435135135135135,
+      "loss": 0.462,
+      "step": 1264
+    },
+    {
+      "epoch": 0.37923927299981264,
+      "grad_norm": 0.22954460978507996,
+      "learning_rate": 0.00024346846846846844,
+      "loss": 0.4529,
+      "step": 1265
+    },
+    {
+      "epoch": 0.37953906689151207,
+      "grad_norm": 0.23748372495174408,
+      "learning_rate": 0.0002434234234234234,
+      "loss": 0.4598,
+      "step": 1266
+    },
+    {
+      "epoch": 0.37983886078321155,
+      "grad_norm": 0.26300883293151855,
+      "learning_rate": 0.00024337837837837837,
+      "loss": 0.482,
+      "step": 1267
+    },
+    {
+      "epoch": 0.380138654674911,
+      "grad_norm": 0.2546245753765106,
+      "learning_rate": 0.0002433333333333333,
+      "loss": 0.4607,
+      "step": 1268
+    },
+    {
+      "epoch": 0.38043844856661047,
+      "grad_norm": 0.24974289536476135,
+      "learning_rate": 0.00024328828828828827,
+      "loss": 0.4713,
+      "step": 1269
+    },
+    {
+      "epoch": 0.3807382424583099,
+      "grad_norm": 0.2457670271396637,
+      "learning_rate": 0.00024324324324324323,
+      "loss": 0.4709,
+      "step": 1270
+    },
+    {
+      "epoch": 0.3810380363500094,
+      "grad_norm": 0.2360873520374298,
+      "learning_rate": 0.00024319819819819817,
+      "loss": 0.4653,
+      "step": 1271
+    },
+    {
+      "epoch": 0.3813378302417088,
+      "grad_norm": 0.24448256194591522,
+      "learning_rate": 0.00024315315315315313,
+      "loss": 0.4651,
+      "step": 1272
+    },
+    {
+      "epoch": 0.3816376241334083,
+      "grad_norm": 0.22578337788581848,
+      "learning_rate": 0.0002431081081081081,
+      "loss": 0.4163,
+      "step": 1273
+    },
+    {
+      "epoch": 0.3819374180251077,
+      "grad_norm": 0.24973514676094055,
+      "learning_rate": 0.00024306306306306306,
+      "loss": 0.4638,
+      "step": 1274
+    },
+    {
+      "epoch": 0.3822372119168072,
+      "grad_norm": 0.21938931941986084,
+      "learning_rate": 0.000243018018018018,
+      "loss": 0.4325,
+      "step": 1275
+    },
+    {
+      "epoch": 0.38253700580850664,
+      "grad_norm": 0.23947425186634064,
+      "learning_rate": 0.00024297297297297296,
+      "loss": 0.4408,
+      "step": 1276
+    },
+    {
+      "epoch": 0.3828367997002061,
+      "grad_norm": 0.23008406162261963,
+      "learning_rate": 0.00024292792792792792,
+      "loss": 0.4377,
+      "step": 1277
+    },
+    {
+      "epoch": 0.38313659359190555,
+      "grad_norm": 0.24068063497543335,
+      "learning_rate": 0.00024288288288288286,
+      "loss": 0.4606,
+      "step": 1278
+    },
+    {
+      "epoch": 0.38343638748360503,
+      "grad_norm": 0.2432139664888382,
+      "learning_rate": 0.0002428378378378378,
+      "loss": 0.4494,
+      "step": 1279
+    },
+    {
+      "epoch": 0.38373618137530446,
+      "grad_norm": 0.22731392085552216,
+      "learning_rate": 0.0002427927927927928,
+      "loss": 0.426,
+      "step": 1280
+    },
+    {
+      "epoch": 0.38403597526700395,
+      "grad_norm": 0.2352358102798462,
+      "learning_rate": 0.00024274774774774772,
+      "loss": 0.4537,
+      "step": 1281
+    },
+    {
+      "epoch": 0.3843357691587034,
+      "grad_norm": 0.23868781328201294,
+      "learning_rate": 0.00024270270270270266,
+      "loss": 0.4641,
+      "step": 1282
+    },
+    {
+      "epoch": 0.38463556305040286,
+      "grad_norm": 0.23498302698135376,
+      "learning_rate": 0.00024265765765765765,
+      "loss": 0.4563,
+      "step": 1283
+    },
+    {
+      "epoch": 0.3849353569421023,
+      "grad_norm": 0.24769316613674164,
+      "learning_rate": 0.0002426126126126126,
+      "loss": 0.4567,
+      "step": 1284
+    },
+    {
+      "epoch": 0.3852351508338018,
+      "grad_norm": 0.21658611297607422,
+      "learning_rate": 0.00024256756756756753,
+      "loss": 0.4437,
+      "step": 1285
+    },
+    {
+      "epoch": 0.3855349447255012,
+      "grad_norm": 0.2677985727787018,
+      "learning_rate": 0.00024252252252252252,
+      "loss": 0.45,
+      "step": 1286
+    },
+    {
+      "epoch": 0.3858347386172007,
+      "grad_norm": 0.23147153854370117,
+      "learning_rate": 0.00024247747747747745,
+      "loss": 0.4341,
+      "step": 1287
+    },
+    {
+      "epoch": 0.3861345325089001,
+      "grad_norm": 0.2465144395828247,
+      "learning_rate": 0.0002424324324324324,
+      "loss": 0.4629,
+      "step": 1288
+    },
+    {
+      "epoch": 0.3864343264005996,
+      "grad_norm": 0.23633845150470734,
+      "learning_rate": 0.00024238738738738738,
+      "loss": 0.4484,
+      "step": 1289
+    },
+    {
+      "epoch": 0.38673412029229903,
+      "grad_norm": 0.22743773460388184,
+      "learning_rate": 0.00024234234234234232,
+      "loss": 0.4451,
+      "step": 1290
+    },
+    {
+      "epoch": 0.3870339141839985,
+      "grad_norm": 0.233259379863739,
+      "learning_rate": 0.00024229729729729726,
+      "loss": 0.4546,
+      "step": 1291
+    },
+    {
+      "epoch": 0.38733370807569795,
+      "grad_norm": 0.24213840067386627,
+      "learning_rate": 0.00024225225225225225,
+      "loss": 0.4378,
+      "step": 1292
+    },
+    {
+      "epoch": 0.38763350196739743,
+      "grad_norm": 0.23812246322631836,
+      "learning_rate": 0.00024220720720720718,
+      "loss": 0.4613,
+      "step": 1293
+    },
+    {
+      "epoch": 0.38793329585909686,
+      "grad_norm": 0.25436896085739136,
+      "learning_rate": 0.00024216216216216212,
+      "loss": 0.4462,
+      "step": 1294
+    },
+    {
+      "epoch": 0.38823308975079635,
+      "grad_norm": 0.24321508407592773,
+      "learning_rate": 0.0002421171171171171,
+      "loss": 0.4674,
+      "step": 1295
+    },
+    {
+      "epoch": 0.3885328836424958,
+      "grad_norm": 0.2434927523136139,
+      "learning_rate": 0.00024207207207207205,
+      "loss": 0.4317,
+      "step": 1296
+    },
+    {
+      "epoch": 0.38883267753419526,
+      "grad_norm": 0.2564734220504761,
+      "learning_rate": 0.000242027027027027,
+      "loss": 0.4556,
+      "step": 1297
+    },
+    {
+      "epoch": 0.3891324714258947,
+      "grad_norm": 0.26102596521377563,
+      "learning_rate": 0.00024198198198198198,
+      "loss": 0.4866,
+      "step": 1298
+    },
+    {
+      "epoch": 0.3894322653175942,
+      "grad_norm": 0.26838192343711853,
+      "learning_rate": 0.0002419369369369369,
+      "loss": 0.4824,
+      "step": 1299
+    },
+    {
+      "epoch": 0.3897320592092936,
+      "grad_norm": 0.24552933871746063,
+      "learning_rate": 0.00024189189189189188,
+      "loss": 0.4581,
+      "step": 1300
+    },
+    {
+      "epoch": 0.3900318531009931,
+      "grad_norm": 0.2453169822692871,
+      "learning_rate": 0.0002418468468468468,
+      "loss": 0.4605,
+      "step": 1301
+    },
+    {
+      "epoch": 0.3903316469926925,
+      "grad_norm": 0.24922487139701843,
+      "learning_rate": 0.00024180180180180178,
+      "loss": 0.4555,
+      "step": 1302
+    },
+    {
+      "epoch": 0.390631440884392,
+      "grad_norm": 0.2683355212211609,
+      "learning_rate": 0.00024175675675675674,
+      "loss": 0.4674,
+      "step": 1303
+    },
+    {
+      "epoch": 0.39093123477609143,
+      "grad_norm": 0.23587100207805634,
+      "learning_rate": 0.00024171171171171168,
+      "loss": 0.4564,
+      "step": 1304
+    },
+    {
+      "epoch": 0.3912310286677909,
+      "grad_norm": 0.2331109195947647,
+      "learning_rate": 0.00024166666666666664,
+      "loss": 0.4499,
+      "step": 1305
+    },
+    {
+      "epoch": 0.39153082255949034,
+      "grad_norm": 0.24744129180908203,
+      "learning_rate": 0.0002416216216216216,
+      "loss": 0.453,
+      "step": 1306
+    },
+    {
+      "epoch": 0.39183061645118983,
+      "grad_norm": 0.228163942694664,
+      "learning_rate": 0.00024157657657657654,
+      "loss": 0.4219,
+      "step": 1307
+    },
+    {
+      "epoch": 0.39213041034288926,
+      "grad_norm": 0.2409953773021698,
+      "learning_rate": 0.00024153153153153153,
+      "loss": 0.4322,
+      "step": 1308
+    },
+    {
+      "epoch": 0.39243020423458874,
+      "grad_norm": 0.22519168257713318,
+      "learning_rate": 0.00024148648648648647,
+      "loss": 0.4177,
+      "step": 1309
+    },
+    {
+      "epoch": 0.39272999812628817,
+      "grad_norm": 0.2466001957654953,
+      "learning_rate": 0.0002414414414414414,
+      "loss": 0.4747,
+      "step": 1310
+    },
+    {
+      "epoch": 0.39302979201798766,
+      "grad_norm": 0.247292622923851,
+      "learning_rate": 0.0002413963963963964,
+      "loss": 0.4663,
+      "step": 1311
+    },
+    {
+      "epoch": 0.3933295859096871,
+      "grad_norm": 0.23905499279499054,
+      "learning_rate": 0.00024135135135135133,
+      "loss": 0.4631,
+      "step": 1312
+    },
+    {
+      "epoch": 0.39362937980138657,
+      "grad_norm": 0.27826932072639465,
+      "learning_rate": 0.00024130630630630627,
+      "loss": 0.4863,
+      "step": 1313
+    },
+    {
+      "epoch": 0.393929173693086,
+      "grad_norm": 0.3269807994365692,
+      "learning_rate": 0.00024126126126126126,
+      "loss": 0.4751,
+      "step": 1314
+    },
+    {
+      "epoch": 0.3942289675847855,
+      "grad_norm": 0.2593337595462799,
+      "learning_rate": 0.0002412162162162162,
+      "loss": 0.4754,
+      "step": 1315
+    },
+    {
+      "epoch": 0.3945287614764849,
+      "grad_norm": 0.24511882662773132,
+      "learning_rate": 0.00024117117117117114,
+      "loss": 0.4704,
+      "step": 1316
+    },
+    {
+      "epoch": 0.3948285553681844,
+      "grad_norm": 0.2439701408147812,
+      "learning_rate": 0.00024112612612612613,
+      "loss": 0.4367,
+      "step": 1317
+    },
+    {
+      "epoch": 0.3951283492598838,
+      "grad_norm": 0.24651208519935608,
+      "learning_rate": 0.00024108108108108106,
+      "loss": 0.4635,
+      "step": 1318
+    },
+    {
+      "epoch": 0.3954281431515833,
+      "grad_norm": 0.23849599063396454,
+      "learning_rate": 0.000241036036036036,
+      "loss": 0.4377,
+      "step": 1319
+    },
+    {
+      "epoch": 0.39572793704328274,
+      "grad_norm": 0.23733258247375488,
+      "learning_rate": 0.000240990990990991,
+      "loss": 0.4401,
+      "step": 1320
+    },
+    {
+      "epoch": 0.3960277309349822,
+      "grad_norm": 0.2643173336982727,
+      "learning_rate": 0.00024094594594594593,
+      "loss": 0.4572,
+      "step": 1321
+    },
+    {
+      "epoch": 0.39632752482668165,
+      "grad_norm": 0.23576226830482483,
+      "learning_rate": 0.00024090090090090086,
+      "loss": 0.4693,
+      "step": 1322
+    },
+    {
+      "epoch": 0.39662731871838114,
+      "grad_norm": 0.2418884038925171,
+      "learning_rate": 0.00024085585585585586,
+      "loss": 0.4453,
+      "step": 1323
+    },
+    {
+      "epoch": 0.39692711261008057,
+      "grad_norm": 0.23336432874202728,
+      "learning_rate": 0.0002408108108108108,
+      "loss": 0.4663,
+      "step": 1324
+    },
+    {
+      "epoch": 0.39722690650178005,
+      "grad_norm": 0.2603462040424347,
+      "learning_rate": 0.00024076576576576573,
+      "loss": 0.4643,
+      "step": 1325
+    },
+    {
+      "epoch": 0.3975267003934795,
+      "grad_norm": 0.24288874864578247,
+      "learning_rate": 0.0002407207207207207,
+      "loss": 0.4442,
+      "step": 1326
+    },
+    {
+      "epoch": 0.39782649428517897,
+      "grad_norm": 0.24161702394485474,
+      "learning_rate": 0.00024067567567567566,
+      "loss": 0.4773,
+      "step": 1327
+    },
+    {
+      "epoch": 0.3981262881768784,
+      "grad_norm": 0.23539233207702637,
+      "learning_rate": 0.0002406306306306306,
+      "loss": 0.4416,
+      "step": 1328
+    },
+    {
+      "epoch": 0.3984260820685779,
+      "grad_norm": 0.26161664724349976,
+      "learning_rate": 0.00024058558558558556,
+      "loss": 0.4672,
+      "step": 1329
+    },
+    {
+      "epoch": 0.3987258759602773,
+      "grad_norm": 0.24159055948257446,
+      "learning_rate": 0.00024054054054054052,
+      "loss": 0.4563,
+      "step": 1330
+    },
+    {
+      "epoch": 0.39902566985197674,
+      "grad_norm": 0.24944022297859192,
+      "learning_rate": 0.00024049549549549548,
+      "loss": 0.4554,
+      "step": 1331
+    },
+    {
+      "epoch": 0.3993254637436762,
+      "grad_norm": 0.24242587387561798,
+      "learning_rate": 0.00024045045045045042,
+      "loss": 0.4343,
+      "step": 1332
+    },
+    {
+      "epoch": 0.39962525763537565,
+      "grad_norm": 0.2406960427761078,
+      "learning_rate": 0.00024040540540540539,
+      "loss": 0.4469,
+      "step": 1333
+    },
+    {
+      "epoch": 0.39992505152707514,
+      "grad_norm": 0.24025918543338776,
+      "learning_rate": 0.00024036036036036035,
+      "loss": 0.4056,
+      "step": 1334
+    },
+    {
+      "epoch": 0.40022484541877457,
+      "grad_norm": 0.267782062292099,
+      "learning_rate": 0.00024031531531531529,
+      "loss": 0.4319,
+      "step": 1335
+    },
+    {
+      "epoch": 0.40052463931047405,
+      "grad_norm": 0.2527433931827545,
+      "learning_rate": 0.00024027027027027025,
+      "loss": 0.4679,
+      "step": 1336
+    },
+    {
+      "epoch": 0.4008244332021735,
+      "grad_norm": 0.2520921528339386,
+      "learning_rate": 0.00024022522522522521,
+      "loss": 0.4802,
+      "step": 1337
+    },
+    {
+      "epoch": 0.40112422709387296,
+      "grad_norm": 0.24161456525325775,
+      "learning_rate": 0.00024018018018018015,
+      "loss": 0.4415,
+      "step": 1338
+    },
+    {
+      "epoch": 0.4014240209855724,
+      "grad_norm": 0.24513588845729828,
+      "learning_rate": 0.00024013513513513511,
+      "loss": 0.445,
+      "step": 1339
+    },
+    {
+      "epoch": 0.4017238148772719,
+      "grad_norm": 0.24400116503238678,
+      "learning_rate": 0.00024009009009009008,
+      "loss": 0.4283,
+      "step": 1340
+    },
+    {
+      "epoch": 0.4020236087689713,
+      "grad_norm": 0.24796655774116516,
+      "learning_rate": 0.00024004504504504502,
+      "loss": 0.4434,
+      "step": 1341
+    },
+    {
+      "epoch": 0.4023234026606708,
+      "grad_norm": 0.2471655309200287,
+      "learning_rate": 0.00023999999999999998,
+      "loss": 0.4378,
+      "step": 1342
+    },
+    {
+      "epoch": 0.4026231965523702,
+      "grad_norm": 0.2507822811603546,
+      "learning_rate": 0.00023995495495495494,
+      "loss": 0.4623,
+      "step": 1343
+    },
+    {
+      "epoch": 0.4029229904440697,
+      "grad_norm": 0.24304543435573578,
+      "learning_rate": 0.00023990990990990988,
+      "loss": 0.4355,
+      "step": 1344
+    },
+    {
+      "epoch": 0.40322278433576914,
+      "grad_norm": 0.240738183259964,
+      "learning_rate": 0.00023986486486486487,
+      "loss": 0.4454,
+      "step": 1345
+    },
+    {
+      "epoch": 0.4035225782274686,
+      "grad_norm": 0.2353314906358719,
+      "learning_rate": 0.0002398198198198198,
+      "loss": 0.4493,
+      "step": 1346
+    },
+    {
+      "epoch": 0.40382237211916805,
+      "grad_norm": 0.24467633664608002,
+      "learning_rate": 0.00023977477477477474,
+      "loss": 0.4771,
+      "step": 1347
+    },
+    {
+      "epoch": 0.40412216601086753,
+      "grad_norm": 0.22876618802547455,
+      "learning_rate": 0.00023972972972972974,
+      "loss": 0.4311,
+      "step": 1348
+    },
+    {
+      "epoch": 0.40442195990256696,
+      "grad_norm": 0.24602358043193817,
+      "learning_rate": 0.00023968468468468467,
+      "loss": 0.4676,
+      "step": 1349
+    },
+    {
+      "epoch": 0.40472175379426645,
+      "grad_norm": 0.2410927563905716,
+      "learning_rate": 0.0002396396396396396,
+      "loss": 0.4564,
+      "step": 1350
+    },
+    {
+      "epoch": 0.4050215476859659,
+      "grad_norm": 0.22765080630779266,
+      "learning_rate": 0.00023959459459459455,
+      "loss": 0.447,
+      "step": 1351
+    },
+    {
+      "epoch": 0.40532134157766536,
+      "grad_norm": 0.24429920315742493,
+      "learning_rate": 0.00023954954954954954,
+      "loss": 0.4734,
+      "step": 1352
+    },
+    {
+      "epoch": 0.4056211354693648,
+      "grad_norm": 0.2417088747024536,
+      "learning_rate": 0.00023950450450450447,
+      "loss": 0.4465,
+      "step": 1353
+    },
+    {
+      "epoch": 0.4059209293610643,
+      "grad_norm": 0.25090181827545166,
+      "learning_rate": 0.00023945945945945944,
+      "loss": 0.4754,
+      "step": 1354
+    },
+    {
+      "epoch": 0.4062207232527637,
+      "grad_norm": 0.255610853433609,
+      "learning_rate": 0.0002394144144144144,
+      "loss": 0.4891,
+      "step": 1355
+    },
+    {
+      "epoch": 0.4065205171444632,
+      "grad_norm": 0.22734206914901733,
+      "learning_rate": 0.00023936936936936934,
+      "loss": 0.4279,
+      "step": 1356
+    },
+    {
+      "epoch": 0.4068203110361626,
+      "grad_norm": 0.24400131404399872,
+      "learning_rate": 0.0002393243243243243,
+      "loss": 0.4493,
+      "step": 1357
+    },
+    {
+      "epoch": 0.4071201049278621,
+      "grad_norm": 0.2442186176776886,
+      "learning_rate": 0.00023927927927927927,
+      "loss": 0.4347,
+      "step": 1358
+    },
+    {
+      "epoch": 0.40741989881956153,
+      "grad_norm": 0.2278696894645691,
+      "learning_rate": 0.0002392342342342342,
+      "loss": 0.4346,
+      "step": 1359
+    },
+    {
+      "epoch": 0.407719692711261,
+      "grad_norm": 0.23576432466506958,
+      "learning_rate": 0.00023918918918918917,
+      "loss": 0.4279,
+      "step": 1360
+    },
+    {
+      "epoch": 0.40801948660296045,
+      "grad_norm": 0.25753775238990784,
+      "learning_rate": 0.00023914414414414413,
+      "loss": 0.465,
+      "step": 1361
+    },
+    {
+      "epoch": 0.40831928049465993,
+      "grad_norm": 0.232134148478508,
+      "learning_rate": 0.00023909909909909907,
+      "loss": 0.4467,
+      "step": 1362
+    },
+    {
+      "epoch": 0.40861907438635936,
+      "grad_norm": 0.25058963894844055,
+      "learning_rate": 0.00023905405405405403,
+      "loss": 0.4496,
+      "step": 1363
+    },
+    {
+      "epoch": 0.40891886827805884,
+      "grad_norm": 0.24595490097999573,
+      "learning_rate": 0.000239009009009009,
+      "loss": 0.4625,
+      "step": 1364
+    },
+    {
+      "epoch": 0.4092186621697583,
+      "grad_norm": 0.2425229847431183,
+      "learning_rate": 0.00023896396396396393,
+      "loss": 0.4757,
+      "step": 1365
+    },
+    {
+      "epoch": 0.40951845606145776,
+      "grad_norm": 0.26115790009498596,
+      "learning_rate": 0.0002389189189189189,
+      "loss": 0.4619,
+      "step": 1366
+    },
+    {
+      "epoch": 0.4098182499531572,
+      "grad_norm": 0.22355914115905762,
+      "learning_rate": 0.00023887387387387386,
+      "loss": 0.4234,
+      "step": 1367
+    },
+    {
+      "epoch": 0.4101180438448567,
+      "grad_norm": 0.22979210317134857,
+      "learning_rate": 0.00023882882882882882,
+      "loss": 0.4472,
+      "step": 1368
+    },
+    {
+      "epoch": 0.4104178377365561,
+      "grad_norm": 0.2713291645050049,
+      "learning_rate": 0.00023878378378378376,
+      "loss": 0.4819,
+      "step": 1369
+    },
+    {
+      "epoch": 0.4107176316282556,
+      "grad_norm": 0.22952406108379364,
+      "learning_rate": 0.00023873873873873872,
+      "loss": 0.4428,
+      "step": 1370
+    },
+    {
+      "epoch": 0.411017425519955,
+      "grad_norm": 0.24338556826114655,
+      "learning_rate": 0.0002386936936936937,
+      "loss": 0.4703,
+      "step": 1371
+    },
+    {
+      "epoch": 0.4113172194116545,
+      "grad_norm": 0.24610304832458496,
+      "learning_rate": 0.00023864864864864862,
+      "loss": 0.4583,
+      "step": 1372
+    },
+    {
+      "epoch": 0.41161701330335393,
+      "grad_norm": 0.23116329312324524,
+      "learning_rate": 0.0002386036036036036,
+      "loss": 0.4234,
+      "step": 1373
+    },
+    {
+      "epoch": 0.4119168071950534,
+      "grad_norm": 0.24718856811523438,
+      "learning_rate": 0.00023855855855855855,
+      "loss": 0.4507,
+      "step": 1374
+    },
+    {
+      "epoch": 0.41221660108675284,
+      "grad_norm": 0.25558093190193176,
+      "learning_rate": 0.0002385135135135135,
+      "loss": 0.4609,
+      "step": 1375
+    },
+    {
+      "epoch": 0.4125163949784523,
+      "grad_norm": 0.2391168475151062,
+      "learning_rate": 0.00023846846846846843,
+      "loss": 0.4334,
+      "step": 1376
+    },
+    {
+      "epoch": 0.41281618887015176,
+      "grad_norm": 0.242578387260437,
+      "learning_rate": 0.00023842342342342342,
+      "loss": 0.472,
+      "step": 1377
+    },
+    {
+      "epoch": 0.41311598276185124,
+      "grad_norm": 0.24465034902095795,
+      "learning_rate": 0.00023837837837837835,
+      "loss": 0.4584,
+      "step": 1378
+    },
+    {
+      "epoch": 0.41341577665355067,
+      "grad_norm": 0.2500922679901123,
+      "learning_rate": 0.0002383333333333333,
+      "loss": 0.468,
+      "step": 1379
+    },
+    {
+      "epoch": 0.41371557054525016,
+      "grad_norm": 0.23939989507198334,
+      "learning_rate": 0.00023828828828828828,
+      "loss": 0.443,
+      "step": 1380
+    },
+    {
+      "epoch": 0.4140153644369496,
+      "grad_norm": 0.272876113653183,
+      "learning_rate": 0.00023824324324324322,
+      "loss": 0.4665,
+      "step": 1381
+    },
+    {
+      "epoch": 0.41431515832864907,
+      "grad_norm": 0.2664034068584442,
+      "learning_rate": 0.00023819819819819815,
+      "loss": 0.4855,
+      "step": 1382
+    },
+    {
+      "epoch": 0.4146149522203485,
+      "grad_norm": 0.23809301853179932,
+      "learning_rate": 0.00023815315315315315,
+      "loss": 0.4483,
+      "step": 1383
+    },
+    {
+      "epoch": 0.414914746112048,
+      "grad_norm": 0.23019112646579742,
+      "learning_rate": 0.00023810810810810808,
+      "loss": 0.4324,
+      "step": 1384
+    },
+    {
+      "epoch": 0.4152145400037474,
+      "grad_norm": 0.25093144178390503,
+      "learning_rate": 0.00023806306306306302,
+      "loss": 0.4732,
+      "step": 1385
+    },
+    {
+      "epoch": 0.4155143338954469,
+      "grad_norm": 0.23062798380851746,
+      "learning_rate": 0.000238018018018018,
+      "loss": 0.4489,
+      "step": 1386
+    },
+    {
+      "epoch": 0.4158141277871463,
+      "grad_norm": 0.23424816131591797,
+      "learning_rate": 0.00023797297297297295,
+      "loss": 0.4586,
+      "step": 1387
+    },
+    {
+      "epoch": 0.4161139216788458,
+      "grad_norm": 0.23515811562538147,
+      "learning_rate": 0.0002379279279279279,
+      "loss": 0.4478,
+      "step": 1388
+    },
+    {
+      "epoch": 0.41641371557054524,
+      "grad_norm": 0.23988017439842224,
+      "learning_rate": 0.00023788288288288287,
+      "loss": 0.4382,
+      "step": 1389
+    },
+    {
+      "epoch": 0.4167135094622447,
+      "grad_norm": 0.23148085176944733,
+      "learning_rate": 0.0002378378378378378,
+      "loss": 0.4476,
+      "step": 1390
+    },
+    {
+      "epoch": 0.41701330335394415,
+      "grad_norm": 0.22274036705493927,
+      "learning_rate": 0.00023779279279279278,
+      "loss": 0.4356,
+      "step": 1391
+    },
+    {
+      "epoch": 0.41731309724564364,
+      "grad_norm": 0.22446554899215698,
+      "learning_rate": 0.00023774774774774774,
+      "loss": 0.438,
+      "step": 1392
+    },
+    {
+      "epoch": 0.41761289113734307,
+      "grad_norm": 0.2218816876411438,
+      "learning_rate": 0.00023770270270270268,
+      "loss": 0.4137,
+      "step": 1393
+    },
+    {
+      "epoch": 0.41791268502904255,
+      "grad_norm": 0.23347264528274536,
+      "learning_rate": 0.00023765765765765764,
+      "loss": 0.4361,
+      "step": 1394
+    },
+    {
+      "epoch": 0.418212478920742,
+      "grad_norm": 0.26775455474853516,
+      "learning_rate": 0.0002376126126126126,
+      "loss": 0.5153,
+      "step": 1395
+    },
+    {
+      "epoch": 0.41851227281244147,
+      "grad_norm": 0.2503002882003784,
+      "learning_rate": 0.00023756756756756754,
+      "loss": 0.4705,
+      "step": 1396
+    },
+    {
+      "epoch": 0.4188120667041409,
+      "grad_norm": 0.2259148210287094,
+      "learning_rate": 0.0002375225225225225,
+      "loss": 0.4251,
+      "step": 1397
+    },
+    {
+      "epoch": 0.4191118605958404,
+      "grad_norm": 0.24358846247196198,
+      "learning_rate": 0.00023747747747747744,
+      "loss": 0.4483,
+      "step": 1398
+    },
+    {
+      "epoch": 0.4194116544875398,
+      "grad_norm": 0.24675339460372925,
+      "learning_rate": 0.0002374324324324324,
+      "loss": 0.4379,
+      "step": 1399
+    },
+    {
+      "epoch": 0.4197114483792393,
+      "grad_norm": 0.24181531369686127,
+      "learning_rate": 0.00023738738738738737,
+      "loss": 0.4498,
+      "step": 1400
+    },
+    {
+      "epoch": 0.4200112422709387,
+      "grad_norm": 0.24917852878570557,
+      "learning_rate": 0.0002373423423423423,
+      "loss": 0.4436,
+      "step": 1401
+    },
+    {
+      "epoch": 0.4203110361626382,
+      "grad_norm": 0.24392594397068024,
+      "learning_rate": 0.0002372972972972973,
+      "loss": 0.4791,
+      "step": 1402
+    },
+    {
+      "epoch": 0.42061083005433764,
+      "grad_norm": 0.2411157488822937,
+      "learning_rate": 0.00023725225225225223,
+      "loss": 0.472,
+      "step": 1403
+    },
+    {
+      "epoch": 0.4209106239460371,
+      "grad_norm": 0.2558772563934326,
+      "learning_rate": 0.00023720720720720717,
+      "loss": 0.4914,
+      "step": 1404
+    },
+    {
+      "epoch": 0.42121041783773655,
+      "grad_norm": 0.2580840587615967,
+      "learning_rate": 0.00023716216216216216,
+      "loss": 0.4502,
+      "step": 1405
+    },
+    {
+      "epoch": 0.42151021172943604,
+      "grad_norm": 0.23793622851371765,
+      "learning_rate": 0.0002371171171171171,
+      "loss": 0.4599,
+      "step": 1406
+    },
+    {
+      "epoch": 0.42181000562113546,
+      "grad_norm": 0.23233507573604584,
+      "learning_rate": 0.00023707207207207203,
+      "loss": 0.4268,
+      "step": 1407
+    },
+    {
+      "epoch": 0.42210979951283495,
+      "grad_norm": 0.24053210020065308,
+      "learning_rate": 0.00023702702702702703,
+      "loss": 0.4233,
+      "step": 1408
+    },
+    {
+      "epoch": 0.4224095934045344,
+      "grad_norm": 0.2314370572566986,
+      "learning_rate": 0.00023698198198198196,
+      "loss": 0.427,
+      "step": 1409
+    },
+    {
+      "epoch": 0.42270938729623386,
+      "grad_norm": 0.2242741882801056,
+      "learning_rate": 0.0002369369369369369,
+      "loss": 0.4097,
+      "step": 1410
+    },
+    {
+      "epoch": 0.4230091811879333,
+      "grad_norm": 0.24178822338581085,
+      "learning_rate": 0.0002368918918918919,
+      "loss": 0.4259,
+      "step": 1411
+    },
+    {
+      "epoch": 0.4233089750796328,
+      "grad_norm": 0.23346510529518127,
+      "learning_rate": 0.00023684684684684683,
+      "loss": 0.4442,
+      "step": 1412
+    },
+    {
+      "epoch": 0.4236087689713322,
+      "grad_norm": 0.23908735811710358,
+      "learning_rate": 0.00023680180180180176,
+      "loss": 0.432,
+      "step": 1413
+    },
+    {
+      "epoch": 0.4239085628630317,
+      "grad_norm": 0.24085824191570282,
+      "learning_rate": 0.00023675675675675675,
+      "loss": 0.4648,
+      "step": 1414
+    },
+    {
+      "epoch": 0.4242083567547311,
+      "grad_norm": 0.2325652688741684,
+      "learning_rate": 0.0002367117117117117,
+      "loss": 0.4499,
+      "step": 1415
+    },
+    {
+      "epoch": 0.4245081506464306,
+      "grad_norm": 0.2413867712020874,
+      "learning_rate": 0.00023666666666666663,
+      "loss": 0.4407,
+      "step": 1416
+    },
+    {
+      "epoch": 0.42480794453813003,
+      "grad_norm": 0.2494458556175232,
+      "learning_rate": 0.00023662162162162162,
+      "loss": 0.4828,
+      "step": 1417
+    },
+    {
+      "epoch": 0.4251077384298295,
+      "grad_norm": 0.24031957983970642,
+      "learning_rate": 0.00023657657657657656,
+      "loss": 0.4453,
+      "step": 1418
+    },
+    {
+      "epoch": 0.42540753232152895,
+      "grad_norm": 0.2490081638097763,
+      "learning_rate": 0.0002365315315315315,
+      "loss": 0.4595,
+      "step": 1419
+    },
+    {
+      "epoch": 0.42570732621322843,
+      "grad_norm": 0.2513922452926636,
+      "learning_rate": 0.00023648648648648648,
+      "loss": 0.4507,
+      "step": 1420
+    },
+    {
+      "epoch": 0.42600712010492786,
+      "grad_norm": 0.23692888021469116,
+      "learning_rate": 0.00023644144144144142,
+      "loss": 0.4339,
+      "step": 1421
+    },
+    {
+      "epoch": 0.42630691399662735,
+      "grad_norm": 0.22867028415203094,
+      "learning_rate": 0.00023639639639639636,
+      "loss": 0.437,
+      "step": 1422
+    },
+    {
+      "epoch": 0.4266067078883268,
+      "grad_norm": 0.23194801807403564,
+      "learning_rate": 0.00023635135135135132,
+      "loss": 0.4558,
+      "step": 1423
+    },
+    {
+      "epoch": 0.42690650178002626,
+      "grad_norm": 0.23193570971488953,
+      "learning_rate": 0.00023630630630630628,
+      "loss": 0.4446,
+      "step": 1424
+    },
+    {
+      "epoch": 0.4272062956717257,
+      "grad_norm": 0.2337258905172348,
+      "learning_rate": 0.00023626126126126125,
+      "loss": 0.4337,
+      "step": 1425
+    },
+    {
+      "epoch": 0.4275060895634251,
+      "grad_norm": 0.23658838868141174,
+      "learning_rate": 0.00023621621621621619,
+      "loss": 0.4288,
+      "step": 1426
+    },
+    {
+      "epoch": 0.4278058834551246,
+      "grad_norm": 0.26249489188194275,
+      "learning_rate": 0.00023617117117117115,
+      "loss": 0.4673,
+      "step": 1427
+    },
+    {
+      "epoch": 0.42810567734682403,
+      "grad_norm": 0.2465837299823761,
+      "learning_rate": 0.0002361261261261261,
+      "loss": 0.4357,
+      "step": 1428
+    },
+    {
+      "epoch": 0.4284054712385235,
+      "grad_norm": 0.26788344979286194,
+      "learning_rate": 0.00023608108108108105,
+      "loss": 0.4465,
+      "step": 1429
+    },
+    {
+      "epoch": 0.42870526513022295,
+      "grad_norm": 0.2597041726112366,
+      "learning_rate": 0.00023603603603603601,
+      "loss": 0.4433,
+      "step": 1430
+    },
+    {
+      "epoch": 0.42900505902192243,
+      "grad_norm": 0.235942080616951,
+      "learning_rate": 0.00023599099099099098,
+      "loss": 0.4322,
+      "step": 1431
+    },
+    {
+      "epoch": 0.42930485291362186,
+      "grad_norm": 0.25687772035598755,
+      "learning_rate": 0.00023594594594594591,
+      "loss": 0.478,
+      "step": 1432
+    },
+    {
+      "epoch": 0.42960464680532134,
+      "grad_norm": 0.2209557294845581,
+      "learning_rate": 0.00023590090090090088,
+      "loss": 0.4254,
+      "step": 1433
+    },
+    {
+      "epoch": 0.4299044406970208,
+      "grad_norm": 0.2533595860004425,
+      "learning_rate": 0.00023585585585585584,
+      "loss": 0.4545,
+      "step": 1434
+    },
+    {
+      "epoch": 0.43020423458872026,
+      "grad_norm": 0.2461264431476593,
+      "learning_rate": 0.00023581081081081078,
+      "loss": 0.4642,
+      "step": 1435
+    },
+    {
+      "epoch": 0.4305040284804197,
+      "grad_norm": 0.23638790845870972,
+      "learning_rate": 0.00023576576576576577,
+      "loss": 0.4403,
+      "step": 1436
+    },
+    {
+      "epoch": 0.43080382237211917,
+      "grad_norm": 0.23931315541267395,
+      "learning_rate": 0.0002357207207207207,
+      "loss": 0.4356,
+      "step": 1437
+    },
+    {
+      "epoch": 0.4311036162638186,
+      "grad_norm": 0.26020053029060364,
+      "learning_rate": 0.00023567567567567564,
+      "loss": 0.4562,
+      "step": 1438
+    },
+    {
+      "epoch": 0.4314034101555181,
+      "grad_norm": 0.25653496384620667,
+      "learning_rate": 0.00023563063063063063,
+      "loss": 0.473,
+      "step": 1439
+    },
+    {
+      "epoch": 0.4317032040472175,
+      "grad_norm": 0.23115600645542145,
+      "learning_rate": 0.00023558558558558557,
+      "loss": 0.4259,
+      "step": 1440
+    },
+    {
+      "epoch": 0.432002997938917,
+      "grad_norm": 0.240982785820961,
+      "learning_rate": 0.0002355405405405405,
+      "loss": 0.4507,
+      "step": 1441
+    },
+    {
+      "epoch": 0.43230279183061643,
+      "grad_norm": 0.2819494307041168,
+      "learning_rate": 0.0002354954954954955,
+      "loss": 0.4539,
+      "step": 1442
+    },
+    {
+      "epoch": 0.4326025857223159,
+      "grad_norm": 0.24356286227703094,
+      "learning_rate": 0.00023545045045045044,
+      "loss": 0.4377,
+      "step": 1443
+    },
+    {
+      "epoch": 0.43290237961401534,
+      "grad_norm": 0.23919035494327545,
+      "learning_rate": 0.00023540540540540537,
+      "loss": 0.4141,
+      "step": 1444
+    },
+    {
+      "epoch": 0.4332021735057148,
+      "grad_norm": 0.27333101630210876,
+      "learning_rate": 0.00023536036036036036,
+      "loss": 0.4655,
+      "step": 1445
+    },
+    {
+      "epoch": 0.43350196739741426,
+      "grad_norm": 0.24128217995166779,
+      "learning_rate": 0.0002353153153153153,
+      "loss": 0.4483,
+      "step": 1446
+    },
+    {
+      "epoch": 0.43380176128911374,
+      "grad_norm": 0.2448810636997223,
+      "learning_rate": 0.00023527027027027024,
+      "loss": 0.4376,
+      "step": 1447
+    },
+    {
+      "epoch": 0.43410155518081317,
+      "grad_norm": 0.24084526300430298,
+      "learning_rate": 0.0002352252252252252,
+      "loss": 0.4228,
+      "step": 1448
+    },
+    {
+      "epoch": 0.43440134907251265,
+      "grad_norm": 0.2719487249851227,
+      "learning_rate": 0.00023518018018018016,
+      "loss": 0.4564,
+      "step": 1449
+    },
+    {
+      "epoch": 0.4347011429642121,
+      "grad_norm": 0.23745928704738617,
+      "learning_rate": 0.0002351351351351351,
+      "loss": 0.4283,
+      "step": 1450
+    },
+    {
+      "epoch": 0.43500093685591157,
+      "grad_norm": 0.22788289189338684,
+      "learning_rate": 0.00023509009009009007,
+      "loss": 0.4215,
+      "step": 1451
+    },
+    {
+      "epoch": 0.435300730747611,
+      "grad_norm": 0.2702344059944153,
+      "learning_rate": 0.00023504504504504503,
+      "loss": 0.4425,
+      "step": 1452
+    },
+    {
+      "epoch": 0.4356005246393105,
+      "grad_norm": 0.2503633499145508,
+      "learning_rate": 0.00023499999999999997,
+      "loss": 0.449,
+      "step": 1453
+    },
+    {
+      "epoch": 0.4359003185310099,
+      "grad_norm": 0.2610470950603485,
+      "learning_rate": 0.00023495495495495493,
+      "loss": 0.4708,
+      "step": 1454
+    },
+    {
+      "epoch": 0.4362001124227094,
+      "grad_norm": 0.2375205010175705,
+      "learning_rate": 0.0002349099099099099,
+      "loss": 0.4426,
+      "step": 1455
+    },
+    {
+      "epoch": 0.4364999063144088,
+      "grad_norm": 0.2531331777572632,
+      "learning_rate": 0.00023486486486486483,
+      "loss": 0.4516,
+      "step": 1456
+    },
+    {
+      "epoch": 0.4367997002061083,
+      "grad_norm": 0.23722784221172333,
+      "learning_rate": 0.0002348198198198198,
+      "loss": 0.4345,
+      "step": 1457
+    },
+    {
+      "epoch": 0.43709949409780774,
+      "grad_norm": 0.2411564141511917,
+      "learning_rate": 0.00023477477477477476,
+      "loss": 0.4227,
+      "step": 1458
+    },
+    {
+      "epoch": 0.4373992879895072,
+      "grad_norm": 0.22307904064655304,
+      "learning_rate": 0.00023472972972972972,
+      "loss": 0.4254,
+      "step": 1459
+    },
+    {
+      "epoch": 0.43769908188120665,
+      "grad_norm": 0.273569256067276,
+      "learning_rate": 0.00023468468468468466,
+      "loss": 0.4384,
+      "step": 1460
+    },
+    {
+      "epoch": 0.43799887577290614,
+      "grad_norm": 0.24101589620113373,
+      "learning_rate": 0.00023463963963963962,
+      "loss": 0.4694,
+      "step": 1461
+    },
+    {
+      "epoch": 0.43829866966460557,
+      "grad_norm": 0.2409631907939911,
+      "learning_rate": 0.0002345945945945946,
+      "loss": 0.4515,
+      "step": 1462
+    },
+    {
+      "epoch": 0.43859846355630505,
+      "grad_norm": 0.24057404696941376,
+      "learning_rate": 0.00023454954954954952,
+      "loss": 0.4516,
+      "step": 1463
+    },
+    {
+      "epoch": 0.4388982574480045,
+      "grad_norm": 0.24539843201637268,
+      "learning_rate": 0.0002345045045045045,
+      "loss": 0.4078,
+      "step": 1464
+    },
+    {
+      "epoch": 0.43919805133970397,
+      "grad_norm": 0.23763391375541687,
+      "learning_rate": 0.00023445945945945945,
+      "loss": 0.4538,
+      "step": 1465
+    },
+    {
+      "epoch": 0.4394978452314034,
+      "grad_norm": 0.25087833404541016,
+      "learning_rate": 0.0002344144144144144,
+      "loss": 0.4474,
+      "step": 1466
+    },
+    {
+      "epoch": 0.4397976391231029,
+      "grad_norm": 0.24220441281795502,
+      "learning_rate": 0.00023436936936936935,
+      "loss": 0.4527,
+      "step": 1467
+    },
+    {
+      "epoch": 0.4400974330148023,
+      "grad_norm": 0.23056988418102264,
+      "learning_rate": 0.00023432432432432432,
+      "loss": 0.4375,
+      "step": 1468
+    },
+    {
+      "epoch": 0.4403972269065018,
+      "grad_norm": 0.23940956592559814,
+      "learning_rate": 0.00023427927927927925,
+      "loss": 0.4561,
+      "step": 1469
+    },
+    {
+      "epoch": 0.4406970207982012,
+      "grad_norm": 0.23279373347759247,
+      "learning_rate": 0.0002342342342342342,
+      "loss": 0.4342,
+      "step": 1470
+    },
+    {
+      "epoch": 0.4409968146899007,
+      "grad_norm": 0.23729127645492554,
+      "learning_rate": 0.00023418918918918918,
+      "loss": 0.4196,
+      "step": 1471
+    },
+    {
+      "epoch": 0.44129660858160014,
+      "grad_norm": 0.2296978086233139,
+      "learning_rate": 0.00023414414414414412,
+      "loss": 0.4308,
+      "step": 1472
+    },
+    {
+      "epoch": 0.4415964024732996,
+      "grad_norm": 0.24595269560813904,
+      "learning_rate": 0.00023409909909909905,
+      "loss": 0.4607,
+      "step": 1473
+    },
+    {
+      "epoch": 0.44189619636499905,
+      "grad_norm": 0.2266230583190918,
+      "learning_rate": 0.00023405405405405404,
+      "loss": 0.4233,
+      "step": 1474
+    },
+    {
+      "epoch": 0.44219599025669853,
+      "grad_norm": 0.23812752962112427,
+      "learning_rate": 0.00023400900900900898,
+      "loss": 0.4341,
+      "step": 1475
+    },
+    {
+      "epoch": 0.44249578414839796,
+      "grad_norm": 0.2498784214258194,
+      "learning_rate": 0.00023396396396396392,
+      "loss": 0.4406,
+      "step": 1476
+    },
+    {
+      "epoch": 0.44279557804009745,
+      "grad_norm": 0.26205873489379883,
+      "learning_rate": 0.0002339189189189189,
+      "loss": 0.463,
+      "step": 1477
+    },
+    {
+      "epoch": 0.4430953719317969,
+      "grad_norm": 0.22054405510425568,
+      "learning_rate": 0.00023387387387387385,
+      "loss": 0.4142,
+      "step": 1478
+    },
+    {
+      "epoch": 0.44339516582349636,
+      "grad_norm": 0.2611534893512726,
+      "learning_rate": 0.00023382882882882878,
+      "loss": 0.4985,
+      "step": 1479
+    },
+    {
+      "epoch": 0.4436949597151958,
+      "grad_norm": 0.2515600323677063,
+      "learning_rate": 0.00023378378378378377,
+      "loss": 0.4567,
+      "step": 1480
+    },
+    {
+      "epoch": 0.4439947536068953,
+      "grad_norm": 0.23842091858386993,
+      "learning_rate": 0.0002337387387387387,
+      "loss": 0.4301,
+      "step": 1481
+    },
+    {
+      "epoch": 0.4442945474985947,
+      "grad_norm": 0.21718434989452362,
+      "learning_rate": 0.00023369369369369367,
+      "loss": 0.4444,
+      "step": 1482
+    },
+    {
+      "epoch": 0.4445943413902942,
+      "grad_norm": 0.2340681105852127,
+      "learning_rate": 0.00023364864864864864,
+      "loss": 0.4555,
+      "step": 1483
+    },
+    {
+      "epoch": 0.4448941352819936,
+      "grad_norm": 0.24483737349510193,
+      "learning_rate": 0.00023360360360360357,
+      "loss": 0.4419,
+      "step": 1484
+    },
+    {
+      "epoch": 0.4451939291736931,
+      "grad_norm": 0.24512210488319397,
+      "learning_rate": 0.00023355855855855854,
+      "loss": 0.427,
+      "step": 1485
+    },
+    {
+      "epoch": 0.44549372306539253,
+      "grad_norm": 0.21723505854606628,
+      "learning_rate": 0.0002335135135135135,
+      "loss": 0.4075,
+      "step": 1486
+    },
+    {
+      "epoch": 0.445793516957092,
+      "grad_norm": 0.23696058988571167,
+      "learning_rate": 0.00023346846846846844,
+      "loss": 0.4332,
+      "step": 1487
+    },
+    {
+      "epoch": 0.44609331084879145,
+      "grad_norm": 0.23997806012630463,
+      "learning_rate": 0.0002334234234234234,
+      "loss": 0.4438,
+      "step": 1488
+    },
+    {
+      "epoch": 0.44639310474049093,
+      "grad_norm": 0.2866729497909546,
+      "learning_rate": 0.00023337837837837837,
+      "loss": 0.4851,
+      "step": 1489
+    },
+    {
+      "epoch": 0.44669289863219036,
+      "grad_norm": 0.22988809645175934,
+      "learning_rate": 0.0002333333333333333,
+      "loss": 0.4185,
+      "step": 1490
+    },
+    {
+      "epoch": 0.44699269252388985,
+      "grad_norm": 0.24402973055839539,
+      "learning_rate": 0.00023328828828828827,
+      "loss": 0.4277,
+      "step": 1491
+    },
+    {
+      "epoch": 0.4472924864155893,
+      "grad_norm": 0.25479069352149963,
+      "learning_rate": 0.00023324324324324323,
+      "loss": 0.4593,
+      "step": 1492
+    },
+    {
+      "epoch": 0.44759228030728876,
+      "grad_norm": 0.27010777592658997,
+      "learning_rate": 0.0002331981981981982,
+      "loss": 0.4391,
+      "step": 1493
+    },
+    {
+      "epoch": 0.4478920741989882,
+      "grad_norm": 0.2443162202835083,
+      "learning_rate": 0.00023315315315315313,
+      "loss": 0.4485,
+      "step": 1494
+    },
+    {
+      "epoch": 0.4481918680906877,
+      "grad_norm": 0.23816895484924316,
+      "learning_rate": 0.00023310810810810807,
+      "loss": 0.4606,
+      "step": 1495
+    },
+    {
+      "epoch": 0.4484916619823871,
+      "grad_norm": 0.25230711698532104,
+      "learning_rate": 0.00023306306306306306,
+      "loss": 0.4479,
+      "step": 1496
+    },
+    {
+      "epoch": 0.4487914558740866,
+      "grad_norm": 0.23312939703464508,
+      "learning_rate": 0.000233018018018018,
+      "loss": 0.4157,
+      "step": 1497
+    },
+    {
+      "epoch": 0.449091249765786,
+      "grad_norm": 0.2355630099773407,
+      "learning_rate": 0.00023297297297297293,
+      "loss": 0.4428,
+      "step": 1498
+    },
+    {
+      "epoch": 0.4493910436574855,
+      "grad_norm": 0.21646787226200104,
+      "learning_rate": 0.00023292792792792792,
+      "loss": 0.4166,
+      "step": 1499
+    },
+    {
+      "epoch": 0.44969083754918493,
+      "grad_norm": 0.2547577917575836,
+      "learning_rate": 0.00023288288288288286,
+      "loss": 0.4517,
+      "step": 1500
+    },
+    {
+      "epoch": 0.44969083754918493,
+      "eval_loss": 0.4507916569709778,
+      "eval_runtime": 565.1356,
+      "eval_samples_per_second": 3.82,
+      "eval_steps_per_second": 0.478,
+      "step": 1500
+    },
+    {
+      "epoch": 0.4499906314408844,
+      "grad_norm": 0.22076524794101715,
+      "learning_rate": 0.0002328378378378378,
+      "loss": 0.4395,
+      "step": 1501
+    },
+    {
+      "epoch": 0.45029042533258384,
+      "grad_norm": 0.23838326334953308,
+      "learning_rate": 0.0002327927927927928,
+      "loss": 0.456,
+      "step": 1502
+    },
+    {
+      "epoch": 0.45059021922428333,
+      "grad_norm": 0.23766019940376282,
+      "learning_rate": 0.00023274774774774773,
+      "loss": 0.4506,
+      "step": 1503
+    },
+    {
+      "epoch": 0.45089001311598276,
+      "grad_norm": 0.2391175776720047,
+      "learning_rate": 0.00023270270270270266,
+      "loss": 0.4484,
+      "step": 1504
+    },
+    {
+      "epoch": 0.45118980700768224,
+      "grad_norm": 0.24999289214611053,
+      "learning_rate": 0.00023265765765765765,
+      "loss": 0.4524,
+      "step": 1505
+    },
+    {
+      "epoch": 0.45148960089938167,
+      "grad_norm": 0.24920374155044556,
+      "learning_rate": 0.0002326126126126126,
+      "loss": 0.4626,
+      "step": 1506
+    },
+    {
+      "epoch": 0.45178939479108116,
+      "grad_norm": 0.24506935477256775,
+      "learning_rate": 0.00023256756756756753,
+      "loss": 0.438,
+      "step": 1507
+    },
+    {
+      "epoch": 0.4520891886827806,
+      "grad_norm": 0.2410869002342224,
+      "learning_rate": 0.00023252252252252252,
+      "loss": 0.4579,
+      "step": 1508
+    },
+    {
+      "epoch": 0.45238898257448007,
+      "grad_norm": 0.2394392192363739,
+      "learning_rate": 0.00023247747747747745,
+      "loss": 0.4701,
+      "step": 1509
+    },
+    {
+      "epoch": 0.4526887764661795,
+      "grad_norm": 0.24809177219867706,
+      "learning_rate": 0.0002324324324324324,
+      "loss": 0.4521,
+      "step": 1510
+    },
+    {
+      "epoch": 0.452988570357879,
+      "grad_norm": 0.24093541502952576,
+      "learning_rate": 0.00023238738738738738,
+      "loss": 0.4364,
+      "step": 1511
+    },
+    {
+      "epoch": 0.4532883642495784,
+      "grad_norm": 0.24750453233718872,
+      "learning_rate": 0.00023234234234234232,
+      "loss": 0.454,
+      "step": 1512
+    },
+    {
+      "epoch": 0.4535881581412779,
+      "grad_norm": 0.24669384956359863,
+      "learning_rate": 0.00023229729729729726,
+      "loss": 0.4687,
+      "step": 1513
+    },
+    {
+      "epoch": 0.4538879520329773,
+      "grad_norm": 0.258184015750885,
+      "learning_rate": 0.00023225225225225225,
+      "loss": 0.4741,
+      "step": 1514
+    },
+    {
+      "epoch": 0.4541877459246768,
+      "grad_norm": 0.23264341056346893,
+      "learning_rate": 0.00023220720720720718,
+      "loss": 0.4255,
+      "step": 1515
+    },
+    {
+      "epoch": 0.45448753981637624,
+      "grad_norm": 0.24050508439540863,
+      "learning_rate": 0.00023216216216216215,
+      "loss": 0.4465,
+      "step": 1516
+    },
+    {
+      "epoch": 0.4547873337080757,
+      "grad_norm": 0.23079554736614227,
+      "learning_rate": 0.0002321171171171171,
+      "loss": 0.4232,
+      "step": 1517
+    },
+    {
+      "epoch": 0.45508712759977515,
+      "grad_norm": 0.22280898690223694,
+      "learning_rate": 0.00023207207207207205,
+      "loss": 0.4233,
+      "step": 1518
+    },
+    {
+      "epoch": 0.45538692149147464,
+      "grad_norm": 0.24419550597667694,
+      "learning_rate": 0.000232027027027027,
+      "loss": 0.453,
+      "step": 1519
+    },
+    {
+      "epoch": 0.45568671538317407,
+      "grad_norm": 0.2578713595867157,
+      "learning_rate": 0.00023198198198198195,
+      "loss": 0.4701,
+      "step": 1520
+    },
+    {
+      "epoch": 0.4559865092748735,
+      "grad_norm": 0.24617789685726166,
+      "learning_rate": 0.0002319369369369369,
+      "loss": 0.4744,
+      "step": 1521
+    },
+    {
+      "epoch": 0.456286303166573,
+      "grad_norm": 0.2564181387424469,
+      "learning_rate": 0.00023189189189189188,
+      "loss": 0.4672,
+      "step": 1522
+    },
+    {
+      "epoch": 0.4565860970582724,
+      "grad_norm": 0.23741687834262848,
+      "learning_rate": 0.0002318468468468468,
+      "loss": 0.4337,
+      "step": 1523
+    },
+    {
+      "epoch": 0.4568858909499719,
+      "grad_norm": 0.2671225070953369,
+      "learning_rate": 0.00023180180180180178,
+      "loss": 0.4723,
+      "step": 1524
+    },
+    {
+      "epoch": 0.4571856848416713,
+      "grad_norm": 0.2585636377334595,
+      "learning_rate": 0.00023175675675675674,
+      "loss": 0.4585,
+      "step": 1525
+    },
+    {
+      "epoch": 0.4574854787333708,
+      "grad_norm": 0.25808751583099365,
+      "learning_rate": 0.00023171171171171168,
+      "loss": 0.4667,
+      "step": 1526
+    },
+    {
+      "epoch": 0.45778527262507024,
+      "grad_norm": 0.25702106952667236,
+      "learning_rate": 0.00023166666666666667,
+      "loss": 0.4218,
+      "step": 1527
+    },
+    {
+      "epoch": 0.4580850665167697,
+      "grad_norm": 0.2685486972332001,
+      "learning_rate": 0.0002316216216216216,
+      "loss": 0.4765,
+      "step": 1528
+    },
+    {
+      "epoch": 0.45838486040846915,
+      "grad_norm": 0.25075605511665344,
+      "learning_rate": 0.00023157657657657654,
+      "loss": 0.4795,
+      "step": 1529
+    },
+    {
+      "epoch": 0.45868465430016864,
+      "grad_norm": 0.25849252939224243,
+      "learning_rate": 0.00023153153153153153,
+      "loss": 0.4362,
+      "step": 1530
+    },
+    {
+      "epoch": 0.45898444819186807,
+      "grad_norm": 0.25761592388153076,
+      "learning_rate": 0.00023148648648648647,
+      "loss": 0.4791,
+      "step": 1531
+    },
+    {
+      "epoch": 0.45928424208356755,
+      "grad_norm": 0.228532075881958,
+      "learning_rate": 0.0002314414414414414,
+      "loss": 0.4258,
+      "step": 1532
+    },
+    {
+      "epoch": 0.459584035975267,
+      "grad_norm": 0.24463020265102386,
+      "learning_rate": 0.0002313963963963964,
+      "loss": 0.4199,
+      "step": 1533
+    },
+    {
+      "epoch": 0.45988382986696646,
+      "grad_norm": 0.26668593287467957,
+      "learning_rate": 0.00023135135135135133,
+      "loss": 0.4686,
+      "step": 1534
+    },
+    {
+      "epoch": 0.4601836237586659,
+      "grad_norm": 0.24953673779964447,
+      "learning_rate": 0.00023130630630630627,
+      "loss": 0.4433,
+      "step": 1535
+    },
+    {
+      "epoch": 0.4604834176503654,
+      "grad_norm": 0.2565534710884094,
+      "learning_rate": 0.00023126126126126126,
+      "loss": 0.4638,
+      "step": 1536
+    },
+    {
+      "epoch": 0.4607832115420648,
+      "grad_norm": 0.241172194480896,
+      "learning_rate": 0.0002312162162162162,
+      "loss": 0.4065,
+      "step": 1537
+    },
+    {
+      "epoch": 0.4610830054337643,
+      "grad_norm": 0.2695203125476837,
+      "learning_rate": 0.00023117117117117114,
+      "loss": 0.4782,
+      "step": 1538
+    },
+    {
+      "epoch": 0.4613827993254637,
+      "grad_norm": 0.25559231638908386,
+      "learning_rate": 0.00023112612612612613,
+      "loss": 0.425,
+      "step": 1539
+    },
+    {
+      "epoch": 0.4616825932171632,
+      "grad_norm": 0.2544387876987457,
+      "learning_rate": 0.00023108108108108106,
+      "loss": 0.451,
+      "step": 1540
+    },
+    {
+      "epoch": 0.46198238710886264,
+      "grad_norm": 0.27124300599098206,
+      "learning_rate": 0.000231036036036036,
+      "loss": 0.4492,
+      "step": 1541
+    },
+    {
+      "epoch": 0.4622821810005621,
+      "grad_norm": 0.2581422030925751,
+      "learning_rate": 0.000230990990990991,
+      "loss": 0.4317,
+      "step": 1542
+    },
+    {
+      "epoch": 0.46258197489226155,
+      "grad_norm": 0.2657614052295685,
+      "learning_rate": 0.00023094594594594593,
+      "loss": 0.4308,
+      "step": 1543
+    },
+    {
+      "epoch": 0.46288176878396103,
+      "grad_norm": 0.25568437576293945,
+      "learning_rate": 0.00023090090090090087,
+      "loss": 0.4808,
+      "step": 1544
+    },
+    {
+      "epoch": 0.46318156267566046,
+      "grad_norm": 0.26639649271965027,
+      "learning_rate": 0.00023085585585585583,
+      "loss": 0.4808,
+      "step": 1545
+    },
+    {
+      "epoch": 0.46348135656735995,
+      "grad_norm": 0.23767255246639252,
+      "learning_rate": 0.0002308108108108108,
+      "loss": 0.4382,
+      "step": 1546
+    },
+    {
+      "epoch": 0.4637811504590594,
+      "grad_norm": 0.22875267267227173,
+      "learning_rate": 0.00023076576576576573,
+      "loss": 0.4132,
+      "step": 1547
+    },
+    {
+      "epoch": 0.46408094435075886,
+      "grad_norm": 0.27917149662971497,
+      "learning_rate": 0.0002307207207207207,
+      "loss": 0.4899,
+      "step": 1548
+    },
+    {
+      "epoch": 0.4643807382424583,
+      "grad_norm": 0.24044126272201538,
+      "learning_rate": 0.00023067567567567566,
+      "loss": 0.4481,
+      "step": 1549
+    },
+    {
+      "epoch": 0.4646805321341578,
+      "grad_norm": 0.2577783465385437,
+      "learning_rate": 0.00023063063063063062,
+      "loss": 0.4403,
+      "step": 1550
+    },
+    {
+      "epoch": 0.4649803260258572,
+      "grad_norm": 0.2589547038078308,
+      "learning_rate": 0.00023058558558558556,
+      "loss": 0.438,
+      "step": 1551
+    },
+    {
+      "epoch": 0.4652801199175567,
+      "grad_norm": 0.22958579659461975,
+      "learning_rate": 0.00023054054054054052,
+      "loss": 0.4374,
+      "step": 1552
+    },
+    {
+      "epoch": 0.4655799138092561,
+      "grad_norm": 0.28952687978744507,
+      "learning_rate": 0.00023049549549549549,
+      "loss": 0.4463,
+      "step": 1553
+    },
+    {
+      "epoch": 0.4658797077009556,
+      "grad_norm": 0.2680447995662689,
+      "learning_rate": 0.00023045045045045042,
+      "loss": 0.4758,
+      "step": 1554
+    },
+    {
+      "epoch": 0.46617950159265503,
+      "grad_norm": 0.23771092295646667,
+      "learning_rate": 0.00023040540540540539,
+      "loss": 0.4322,
+      "step": 1555
+    },
+    {
+      "epoch": 0.4664792954843545,
+      "grad_norm": 0.2691210210323334,
+      "learning_rate": 0.00023036036036036035,
+      "loss": 0.4739,
+      "step": 1556
+    },
+    {
+      "epoch": 0.46677908937605395,
+      "grad_norm": 0.24547810852527618,
+      "learning_rate": 0.0002303153153153153,
+      "loss": 0.4507,
+      "step": 1557
+    },
+    {
+      "epoch": 0.46707888326775343,
+      "grad_norm": 0.2669890224933624,
+      "learning_rate": 0.00023027027027027025,
+      "loss": 0.4809,
+      "step": 1558
+    },
+    {
+      "epoch": 0.46737867715945286,
+      "grad_norm": 0.25527825951576233,
+      "learning_rate": 0.00023022522522522521,
+      "loss": 0.4558,
+      "step": 1559
+    },
+    {
+      "epoch": 0.46767847105115234,
+      "grad_norm": 0.23491926491260529,
+      "learning_rate": 0.00023018018018018015,
+      "loss": 0.4594,
+      "step": 1560
+    },
+    {
+      "epoch": 0.4679782649428518,
+      "grad_norm": 0.27634891867637634,
+      "learning_rate": 0.00023013513513513514,
+      "loss": 0.4503,
+      "step": 1561
+    },
+    {
+      "epoch": 0.46827805883455126,
+      "grad_norm": 0.2656886577606201,
+      "learning_rate": 0.00023009009009009008,
+      "loss": 0.4674,
+      "step": 1562
+    },
+    {
+      "epoch": 0.4685778527262507,
+      "grad_norm": 0.23933476209640503,
+      "learning_rate": 0.00023004504504504502,
+      "loss": 0.4124,
+      "step": 1563
+    },
+    {
+      "epoch": 0.4688776466179502,
+      "grad_norm": 0.2596864700317383,
+      "learning_rate": 0.00023,
+      "loss": 0.4288,
+      "step": 1564
+    },
+    {
+      "epoch": 0.4691774405096496,
+      "grad_norm": 0.25186148285865784,
+      "learning_rate": 0.00022995495495495494,
+      "loss": 0.4571,
+      "step": 1565
+    },
+    {
+      "epoch": 0.4694772344013491,
+      "grad_norm": 0.28007790446281433,
+      "learning_rate": 0.00022990990990990988,
+      "loss": 0.4785,
+      "step": 1566
+    },
+    {
+      "epoch": 0.4697770282930485,
+      "grad_norm": 0.26225724816322327,
+      "learning_rate": 0.00022986486486486482,
+      "loss": 0.4378,
+      "step": 1567
+    },
+    {
+      "epoch": 0.470076822184748,
+      "grad_norm": 0.24554051458835602,
+      "learning_rate": 0.0002298198198198198,
+      "loss": 0.4655,
+      "step": 1568
+    },
+    {
+      "epoch": 0.47037661607644743,
+      "grad_norm": 0.24976593255996704,
+      "learning_rate": 0.00022977477477477475,
+      "loss": 0.4356,
+      "step": 1569
+    },
+    {
+      "epoch": 0.4706764099681469,
+      "grad_norm": 0.23914846777915955,
+      "learning_rate": 0.00022972972972972968,
+      "loss": 0.4239,
+      "step": 1570
+    },
+    {
+      "epoch": 0.47097620385984634,
+      "grad_norm": 0.24698884785175323,
+      "learning_rate": 0.00022968468468468467,
+      "loss": 0.4607,
+      "step": 1571
+    },
+    {
+      "epoch": 0.4712759977515458,
+      "grad_norm": 0.24240310490131378,
+      "learning_rate": 0.0002296396396396396,
+      "loss": 0.4218,
+      "step": 1572
+    },
+    {
+      "epoch": 0.47157579164324526,
+      "grad_norm": 0.24838680028915405,
+      "learning_rate": 0.00022959459459459457,
+      "loss": 0.4387,
+      "step": 1573
+    },
+    {
+      "epoch": 0.47187558553494474,
+      "grad_norm": 0.25536447763442993,
+      "learning_rate": 0.00022954954954954954,
+      "loss": 0.4529,
+      "step": 1574
+    },
+    {
+      "epoch": 0.47217537942664417,
+      "grad_norm": 0.24535490572452545,
+      "learning_rate": 0.00022950450450450447,
+      "loss": 0.4505,
+      "step": 1575
+    },
+    {
+      "epoch": 0.47247517331834366,
+      "grad_norm": 0.258878618478775,
+      "learning_rate": 0.00022945945945945944,
+      "loss": 0.4794,
+      "step": 1576
+    },
+    {
+      "epoch": 0.4727749672100431,
+      "grad_norm": 0.23862193524837494,
+      "learning_rate": 0.0002294144144144144,
+      "loss": 0.4555,
+      "step": 1577
+    },
+    {
+      "epoch": 0.47307476110174257,
+      "grad_norm": 0.2369290292263031,
+      "learning_rate": 0.00022936936936936934,
+      "loss": 0.4111,
+      "step": 1578
+    },
+    {
+      "epoch": 0.473374554993442,
+      "grad_norm": 0.2591108977794647,
+      "learning_rate": 0.0002293243243243243,
+      "loss": 0.4738,
+      "step": 1579
+    },
+    {
+      "epoch": 0.4736743488851415,
+      "grad_norm": 0.2639445662498474,
+      "learning_rate": 0.00022927927927927927,
+      "loss": 0.489,
+      "step": 1580
+    },
+    {
+      "epoch": 0.4739741427768409,
+      "grad_norm": 0.2452382892370224,
+      "learning_rate": 0.0002292342342342342,
+      "loss": 0.4499,
+      "step": 1581
+    },
+    {
+      "epoch": 0.4742739366685404,
+      "grad_norm": 0.24414241313934326,
+      "learning_rate": 0.00022918918918918917,
+      "loss": 0.4246,
+      "step": 1582
+    },
+    {
+      "epoch": 0.4745737305602398,
+      "grad_norm": 0.24609197676181793,
+      "learning_rate": 0.00022914414414414413,
+      "loss": 0.4615,
+      "step": 1583
+    },
+    {
+      "epoch": 0.4748735244519393,
+      "grad_norm": 0.2610466480255127,
+      "learning_rate": 0.0002290990990990991,
+      "loss": 0.4597,
+      "step": 1584
+    },
+    {
+      "epoch": 0.47517331834363874,
+      "grad_norm": 0.24946355819702148,
+      "learning_rate": 0.00022905405405405403,
+      "loss": 0.4382,
+      "step": 1585
+    },
+    {
+      "epoch": 0.4754731122353382,
+      "grad_norm": 0.24156548082828522,
+      "learning_rate": 0.000229009009009009,
+      "loss": 0.4421,
+      "step": 1586
+    },
+    {
+      "epoch": 0.47577290612703765,
+      "grad_norm": 0.2650264799594879,
+      "learning_rate": 0.00022896396396396396,
+      "loss": 0.4735,
+      "step": 1587
+    },
+    {
+      "epoch": 0.47607270001873714,
+      "grad_norm": 0.2677678167819977,
+      "learning_rate": 0.0002289189189189189,
+      "loss": 0.4581,
+      "step": 1588
+    },
+    {
+      "epoch": 0.47637249391043657,
+      "grad_norm": 0.2421169877052307,
+      "learning_rate": 0.00022887387387387386,
+      "loss": 0.4342,
+      "step": 1589
+    },
+    {
+      "epoch": 0.47667228780213605,
+      "grad_norm": 0.2284284085035324,
+      "learning_rate": 0.00022882882882882882,
+      "loss": 0.434,
+      "step": 1590
+    },
+    {
+      "epoch": 0.4769720816938355,
+      "grad_norm": 0.235052689909935,
+      "learning_rate": 0.00022878378378378376,
+      "loss": 0.4439,
+      "step": 1591
+    },
+    {
+      "epoch": 0.47727187558553497,
+      "grad_norm": 0.24947918951511383,
+      "learning_rate": 0.0002287387387387387,
+      "loss": 0.4704,
+      "step": 1592
+    },
+    {
+      "epoch": 0.4775716694772344,
+      "grad_norm": 0.24523784220218658,
+      "learning_rate": 0.0002286936936936937,
+      "loss": 0.4688,
+      "step": 1593
+    },
+    {
+      "epoch": 0.4778714633689339,
+      "grad_norm": 0.2427687793970108,
+      "learning_rate": 0.00022864864864864862,
+      "loss": 0.4145,
+      "step": 1594
+    },
+    {
+      "epoch": 0.4781712572606333,
+      "grad_norm": 0.2589262127876282,
+      "learning_rate": 0.00022860360360360356,
+      "loss": 0.4602,
+      "step": 1595
+    },
+    {
+      "epoch": 0.4784710511523328,
+      "grad_norm": 0.22775280475616455,
+      "learning_rate": 0.00022855855855855855,
+      "loss": 0.4118,
+      "step": 1596
+    },
+    {
+      "epoch": 0.4787708450440322,
+      "grad_norm": 0.26483890414237976,
+      "learning_rate": 0.0002285135135135135,
+      "loss": 0.4342,
+      "step": 1597
+    },
+    {
+      "epoch": 0.4790706389357317,
+      "grad_norm": 0.2529483735561371,
+      "learning_rate": 0.00022846846846846843,
+      "loss": 0.4474,
+      "step": 1598
+    },
+    {
+      "epoch": 0.47937043282743114,
+      "grad_norm": 0.24874089658260345,
+      "learning_rate": 0.00022842342342342342,
+      "loss": 0.4383,
+      "step": 1599
+    },
+    {
+      "epoch": 0.4796702267191306,
+      "grad_norm": 0.2583334445953369,
+      "learning_rate": 0.00022837837837837835,
+      "loss": 0.4406,
+      "step": 1600
+    },
+    {
+      "epoch": 0.47997002061083005,
+      "grad_norm": 0.2603740990161896,
+      "learning_rate": 0.0002283333333333333,
+      "loss": 0.4418,
+      "step": 1601
+    },
+    {
+      "epoch": 0.48026981450252954,
+      "grad_norm": 0.24915438890457153,
+      "learning_rate": 0.00022828828828828828,
+      "loss": 0.458,
+      "step": 1602
+    },
+    {
+      "epoch": 0.48056960839422896,
+      "grad_norm": 0.2531617283821106,
+      "learning_rate": 0.00022824324324324322,
+      "loss": 0.4639,
+      "step": 1603
+    },
+    {
+      "epoch": 0.48086940228592845,
+      "grad_norm": 0.23148907721042633,
+      "learning_rate": 0.00022819819819819816,
+      "loss": 0.4345,
+      "step": 1604
+    },
+    {
+      "epoch": 0.4811691961776279,
+      "grad_norm": 0.26466086506843567,
+      "learning_rate": 0.00022815315315315315,
+      "loss": 0.4625,
+      "step": 1605
+    },
+    {
+      "epoch": 0.48146899006932736,
+      "grad_norm": 0.2558290660381317,
+      "learning_rate": 0.00022810810810810808,
+      "loss": 0.4438,
+      "step": 1606
+    },
+    {
+      "epoch": 0.4817687839610268,
+      "grad_norm": 0.2294214814901352,
+      "learning_rate": 0.00022806306306306305,
+      "loss": 0.4343,
+      "step": 1607
+    },
+    {
+      "epoch": 0.4820685778527263,
+      "grad_norm": 0.25497883558273315,
+      "learning_rate": 0.000228018018018018,
+      "loss": 0.4498,
+      "step": 1608
+    },
+    {
+      "epoch": 0.4823683717444257,
+      "grad_norm": 0.25641798973083496,
+      "learning_rate": 0.00022797297297297295,
+      "loss": 0.4077,
+      "step": 1609
+    },
+    {
+      "epoch": 0.4826681656361252,
+      "grad_norm": 0.2556770145893097,
+      "learning_rate": 0.0002279279279279279,
+      "loss": 0.4125,
+      "step": 1610
+    },
+    {
+      "epoch": 0.4829679595278246,
+      "grad_norm": 0.2528950273990631,
+      "learning_rate": 0.00022788288288288288,
+      "loss": 0.4759,
+      "step": 1611
+    },
+    {
+      "epoch": 0.4832677534195241,
+      "grad_norm": 0.22939835488796234,
+      "learning_rate": 0.0002278378378378378,
+      "loss": 0.4422,
+      "step": 1612
+    },
+    {
+      "epoch": 0.48356754731122353,
+      "grad_norm": 0.24314181506633759,
+      "learning_rate": 0.00022779279279279278,
+      "loss": 0.4358,
+      "step": 1613
+    },
+    {
+      "epoch": 0.483867341202923,
+      "grad_norm": 0.2460578978061676,
+      "learning_rate": 0.00022774774774774774,
+      "loss": 0.4571,
+      "step": 1614
+    },
+    {
+      "epoch": 0.48416713509462245,
+      "grad_norm": 0.2627396583557129,
+      "learning_rate": 0.00022770270270270268,
+      "loss": 0.4616,
+      "step": 1615
+    },
+    {
+      "epoch": 0.4844669289863219,
+      "grad_norm": 0.23525434732437134,
+      "learning_rate": 0.00022765765765765764,
+      "loss": 0.4243,
+      "step": 1616
+    },
+    {
+      "epoch": 0.48476672287802136,
+      "grad_norm": 0.2397356927394867,
+      "learning_rate": 0.00022761261261261258,
+      "loss": 0.4382,
+      "step": 1617
+    },
+    {
+      "epoch": 0.4850665167697208,
+      "grad_norm": 0.2398831993341446,
+      "learning_rate": 0.00022756756756756757,
+      "loss": 0.4412,
+      "step": 1618
+    },
+    {
+      "epoch": 0.4853663106614203,
+      "grad_norm": 0.259376585483551,
+      "learning_rate": 0.0002275225225225225,
+      "loss": 0.4259,
+      "step": 1619
+    },
+    {
+      "epoch": 0.4856661045531197,
+      "grad_norm": 0.23204876482486725,
+      "learning_rate": 0.00022747747747747744,
+      "loss": 0.4068,
+      "step": 1620
+    },
+    {
+      "epoch": 0.4859658984448192,
+      "grad_norm": 0.2531450688838959,
+      "learning_rate": 0.00022743243243243243,
+      "loss": 0.4383,
+      "step": 1621
+    },
+    {
+      "epoch": 0.4862656923365186,
+      "grad_norm": 0.24866148829460144,
+      "learning_rate": 0.00022738738738738737,
+      "loss": 0.4856,
+      "step": 1622
+    },
+    {
+      "epoch": 0.4865654862282181,
+      "grad_norm": 0.23114702105522156,
+      "learning_rate": 0.0002273423423423423,
+      "loss": 0.4167,
+      "step": 1623
+    },
+    {
+      "epoch": 0.48686528011991753,
+      "grad_norm": 0.24857169389724731,
+      "learning_rate": 0.0002272972972972973,
+      "loss": 0.4678,
+      "step": 1624
+    },
+    {
+      "epoch": 0.487165074011617,
+      "grad_norm": 0.24516363441944122,
+      "learning_rate": 0.00022725225225225223,
+      "loss": 0.4236,
+      "step": 1625
+    },
+    {
+      "epoch": 0.48746486790331645,
+      "grad_norm": 0.2485557198524475,
+      "learning_rate": 0.00022720720720720717,
+      "loss": 0.4562,
+      "step": 1626
+    },
+    {
+      "epoch": 0.48776466179501593,
+      "grad_norm": 0.2518232762813568,
+      "learning_rate": 0.00022716216216216216,
+      "loss": 0.4703,
+      "step": 1627
+    },
+    {
+      "epoch": 0.48806445568671536,
+      "grad_norm": 0.23469193279743195,
+      "learning_rate": 0.0002271171171171171,
+      "loss": 0.4301,
+      "step": 1628
+    },
+    {
+      "epoch": 0.48836424957841484,
+      "grad_norm": 0.25275421142578125,
+      "learning_rate": 0.00022707207207207204,
+      "loss": 0.4576,
+      "step": 1629
+    },
+    {
+      "epoch": 0.4886640434701143,
+      "grad_norm": 0.25434768199920654,
+      "learning_rate": 0.00022702702702702703,
+      "loss": 0.4654,
+      "step": 1630
+    },
+    {
+      "epoch": 0.48896383736181376,
+      "grad_norm": 0.24483707547187805,
+      "learning_rate": 0.00022698198198198196,
+      "loss": 0.45,
+      "step": 1631
+    },
+    {
+      "epoch": 0.4892636312535132,
+      "grad_norm": 0.24300119280815125,
+      "learning_rate": 0.0002269369369369369,
+      "loss": 0.4795,
+      "step": 1632
+    },
+    {
+      "epoch": 0.48956342514521267,
+      "grad_norm": 0.23230652511119843,
+      "learning_rate": 0.0002268918918918919,
+      "loss": 0.4318,
+      "step": 1633
+    },
+    {
+      "epoch": 0.4898632190369121,
+      "grad_norm": 0.2618871331214905,
+      "learning_rate": 0.00022684684684684683,
+      "loss": 0.4846,
+      "step": 1634
+    },
+    {
+      "epoch": 0.4901630129286116,
+      "grad_norm": 0.2477121651172638,
+      "learning_rate": 0.00022680180180180176,
+      "loss": 0.4392,
+      "step": 1635
+    },
+    {
+      "epoch": 0.490462806820311,
+      "grad_norm": 0.2603763937950134,
+      "learning_rate": 0.00022675675675675676,
+      "loss": 0.4875,
+      "step": 1636
+    },
+    {
+      "epoch": 0.4907626007120105,
+      "grad_norm": 0.2470758855342865,
+      "learning_rate": 0.0002267117117117117,
+      "loss": 0.4575,
+      "step": 1637
+    },
+    {
+      "epoch": 0.49106239460370993,
+      "grad_norm": 0.2612755000591278,
+      "learning_rate": 0.00022666666666666663,
+      "loss": 0.4652,
+      "step": 1638
+    },
+    {
+      "epoch": 0.4913621884954094,
+      "grad_norm": 0.2424866259098053,
+      "learning_rate": 0.0002266216216216216,
+      "loss": 0.4346,
+      "step": 1639
+    },
+    {
+      "epoch": 0.49166198238710884,
+      "grad_norm": 0.2529725730419159,
+      "learning_rate": 0.00022657657657657656,
+      "loss": 0.4427,
+      "step": 1640
+    },
+    {
+      "epoch": 0.4919617762788083,
+      "grad_norm": 0.2564398944377899,
+      "learning_rate": 0.00022653153153153152,
+      "loss": 0.4641,
+      "step": 1641
+    },
+    {
+      "epoch": 0.49226157017050776,
+      "grad_norm": 0.2599097490310669,
+      "learning_rate": 0.00022648648648648646,
+      "loss": 0.4564,
+      "step": 1642
+    },
+    {
+      "epoch": 0.49256136406220724,
+      "grad_norm": 0.2355022430419922,
+      "learning_rate": 0.00022644144144144142,
+      "loss": 0.4313,
+      "step": 1643
+    },
+    {
+      "epoch": 0.49286115795390667,
+      "grad_norm": 0.24770522117614746,
+      "learning_rate": 0.00022639639639639638,
+      "loss": 0.4468,
+      "step": 1644
+    },
+    {
+      "epoch": 0.49316095184560615,
+      "grad_norm": 0.2558223009109497,
+      "learning_rate": 0.00022635135135135132,
+      "loss": 0.4506,
+      "step": 1645
+    },
+    {
+      "epoch": 0.4934607457373056,
+      "grad_norm": 0.24741050601005554,
+      "learning_rate": 0.00022630630630630629,
+      "loss": 0.4375,
+      "step": 1646
+    },
+    {
+      "epoch": 0.49376053962900507,
+      "grad_norm": 0.24249225854873657,
+      "learning_rate": 0.00022626126126126125,
+      "loss": 0.4283,
+      "step": 1647
+    },
+    {
+      "epoch": 0.4940603335207045,
+      "grad_norm": 0.2792401611804962,
+      "learning_rate": 0.00022621621621621619,
+      "loss": 0.4757,
+      "step": 1648
+    },
+    {
+      "epoch": 0.494360127412404,
+      "grad_norm": 0.23249030113220215,
+      "learning_rate": 0.00022617117117117115,
+      "loss": 0.4417,
+      "step": 1649
+    },
+    {
+      "epoch": 0.4946599213041034,
+      "grad_norm": 0.2646411955356598,
+      "learning_rate": 0.00022612612612612611,
+      "loss": 0.4338,
+      "step": 1650
+    },
+    {
+      "epoch": 0.4949597151958029,
+      "grad_norm": 0.2633950710296631,
+      "learning_rate": 0.00022608108108108105,
+      "loss": 0.476,
+      "step": 1651
+    },
+    {
+      "epoch": 0.4952595090875023,
+      "grad_norm": 0.24183906614780426,
+      "learning_rate": 0.00022603603603603601,
+      "loss": 0.4363,
+      "step": 1652
+    },
+    {
+      "epoch": 0.4955593029792018,
+      "grad_norm": 0.26413261890411377,
+      "learning_rate": 0.00022599099099099098,
+      "loss": 0.4387,
+      "step": 1653
+    },
+    {
+      "epoch": 0.49585909687090124,
+      "grad_norm": 0.2538875937461853,
+      "learning_rate": 0.00022594594594594592,
+      "loss": 0.4284,
+      "step": 1654
+    },
+    {
+      "epoch": 0.4961588907626007,
+      "grad_norm": 0.2566417157649994,
+      "learning_rate": 0.0002259009009009009,
+      "loss": 0.4352,
+      "step": 1655
+    },
+    {
+      "epoch": 0.49645868465430015,
+      "grad_norm": 0.2526787519454956,
+      "learning_rate": 0.00022585585585585584,
+      "loss": 0.4567,
+      "step": 1656
+    },
+    {
+      "epoch": 0.49675847854599964,
+      "grad_norm": 0.2506735324859619,
+      "learning_rate": 0.00022581081081081078,
+      "loss": 0.4346,
+      "step": 1657
+    },
+    {
+      "epoch": 0.49705827243769907,
+      "grad_norm": 0.25367823243141174,
+      "learning_rate": 0.00022576576576576577,
+      "loss": 0.4392,
+      "step": 1658
+    },
+    {
+      "epoch": 0.49735806632939855,
+      "grad_norm": 0.24888494610786438,
+      "learning_rate": 0.0002257207207207207,
+      "loss": 0.4532,
+      "step": 1659
+    },
+    {
+      "epoch": 0.497657860221098,
+      "grad_norm": 0.22846034169197083,
+      "learning_rate": 0.00022567567567567564,
+      "loss": 0.4163,
+      "step": 1660
+    },
+    {
+      "epoch": 0.49795765411279747,
+      "grad_norm": 0.23993121087551117,
+      "learning_rate": 0.00022563063063063064,
+      "loss": 0.4385,
+      "step": 1661
+    },
+    {
+      "epoch": 0.4982574480044969,
+      "grad_norm": 0.2544318437576294,
+      "learning_rate": 0.00022558558558558557,
+      "loss": 0.4058,
+      "step": 1662
+    },
+    {
+      "epoch": 0.4985572418961964,
+      "grad_norm": 0.2419573813676834,
+      "learning_rate": 0.0002255405405405405,
+      "loss": 0.4428,
+      "step": 1663
+    },
+    {
+      "epoch": 0.4988570357878958,
+      "grad_norm": 0.23838767409324646,
+      "learning_rate": 0.00022549549549549547,
+      "loss": 0.4293,
+      "step": 1664
+    },
+    {
+      "epoch": 0.4991568296795953,
+      "grad_norm": 0.23000121116638184,
+      "learning_rate": 0.00022545045045045044,
+      "loss": 0.436,
+      "step": 1665
+    },
+    {
+      "epoch": 0.4994566235712947,
+      "grad_norm": 0.2665446102619171,
+      "learning_rate": 0.00022540540540540537,
+      "loss": 0.4841,
+      "step": 1666
+    },
+    {
+      "epoch": 0.4997564174629942,
+      "grad_norm": 0.25025802850723267,
+      "learning_rate": 0.00022536036036036034,
+      "loss": 0.4357,
+      "step": 1667
+    },
+    {
+      "epoch": 0.5000562113546937,
+      "grad_norm": 0.2441510409116745,
+      "learning_rate": 0.0002253153153153153,
+      "loss": 0.4345,
+      "step": 1668
+    },
+    {
+      "epoch": 0.5003560052463931,
+      "grad_norm": 0.24143864214420319,
+      "learning_rate": 0.00022527027027027024,
+      "loss": 0.4416,
+      "step": 1669
+    },
+    {
+      "epoch": 0.5006557991380925,
+      "grad_norm": 0.246190145611763,
+      "learning_rate": 0.0002252252252252252,
+      "loss": 0.4124,
+      "step": 1670
+    },
+    {
+      "epoch": 0.500955593029792,
+      "grad_norm": 0.2695963382720947,
+      "learning_rate": 0.00022518018018018017,
+      "loss": 0.4683,
+      "step": 1671
+    },
+    {
+      "epoch": 0.5012553869214915,
+      "grad_norm": 0.23124708235263824,
+      "learning_rate": 0.0002251351351351351,
+      "loss": 0.4347,
+      "step": 1672
+    },
+    {
+      "epoch": 0.501555180813191,
+      "grad_norm": 0.25153648853302,
+      "learning_rate": 0.00022509009009009007,
+      "loss": 0.4408,
+      "step": 1673
+    },
+    {
+      "epoch": 0.5018549747048904,
+      "grad_norm": 0.2653743028640747,
+      "learning_rate": 0.00022504504504504503,
+      "loss": 0.4482,
+      "step": 1674
+    },
+    {
+      "epoch": 0.5021547685965898,
+      "grad_norm": 0.24365836381912231,
+      "learning_rate": 0.000225,
+      "loss": 0.4225,
+      "step": 1675
+    },
+    {
+      "epoch": 0.5024545624882893,
+      "grad_norm": 0.2595864236354828,
+      "learning_rate": 0.00022495495495495493,
+      "loss": 0.4624,
+      "step": 1676
+    },
+    {
+      "epoch": 0.5027543563799888,
+      "grad_norm": 0.24555698037147522,
+      "learning_rate": 0.0002249099099099099,
+      "loss": 0.4495,
+      "step": 1677
+    },
+    {
+      "epoch": 0.5030541502716882,
+      "grad_norm": 0.25951844453811646,
+      "learning_rate": 0.00022486486486486486,
+      "loss": 0.4503,
+      "step": 1678
+    },
+    {
+      "epoch": 0.5033539441633876,
+      "grad_norm": 0.28872594237327576,
+      "learning_rate": 0.0002248198198198198,
+      "loss": 0.4916,
+      "step": 1679
+    },
+    {
+      "epoch": 0.5036537380550872,
+      "grad_norm": 0.24827940762043,
+      "learning_rate": 0.00022477477477477476,
+      "loss": 0.4024,
+      "step": 1680
+    },
+    {
+      "epoch": 0.5039535319467866,
+      "grad_norm": 0.2625780999660492,
+      "learning_rate": 0.00022472972972972972,
+      "loss": 0.4624,
+      "step": 1681
+    },
+    {
+      "epoch": 0.504253325838486,
+      "grad_norm": 0.26651719212532043,
+      "learning_rate": 0.00022468468468468466,
+      "loss": 0.4526,
+      "step": 1682
+    },
+    {
+      "epoch": 0.5045531197301855,
+      "grad_norm": 0.2538798153400421,
+      "learning_rate": 0.00022463963963963962,
+      "loss": 0.4328,
+      "step": 1683
+    },
+    {
+      "epoch": 0.504852913621885,
+      "grad_norm": 0.25730088353157043,
+      "learning_rate": 0.0002245945945945946,
+      "loss": 0.4642,
+      "step": 1684
+    },
+    {
+      "epoch": 0.5051527075135844,
+      "grad_norm": 0.24593298137187958,
+      "learning_rate": 0.00022454954954954952,
+      "loss": 0.4422,
+      "step": 1685
+    },
+    {
+      "epoch": 0.5054525014052839,
+      "grad_norm": 0.25883376598358154,
+      "learning_rate": 0.0002245045045045045,
+      "loss": 0.4439,
+      "step": 1686
+    },
+    {
+      "epoch": 0.5057522952969833,
+      "grad_norm": 0.2680940330028534,
+      "learning_rate": 0.00022445945945945945,
+      "loss": 0.4459,
+      "step": 1687
+    },
+    {
+      "epoch": 0.5060520891886828,
+      "grad_norm": 0.23688143491744995,
+      "learning_rate": 0.0002244144144144144,
+      "loss": 0.4404,
+      "step": 1688
+    },
+    {
+      "epoch": 0.5063518830803823,
+      "grad_norm": 0.24362222850322723,
+      "learning_rate": 0.00022436936936936933,
+      "loss": 0.473,
+      "step": 1689
+    },
+    {
+      "epoch": 0.5066516769720817,
+      "grad_norm": 0.24516165256500244,
+      "learning_rate": 0.00022432432432432432,
+      "loss": 0.4288,
+      "step": 1690
+    },
+    {
+      "epoch": 0.5069514708637811,
+      "grad_norm": 0.23739798367023468,
+      "learning_rate": 0.00022427927927927925,
+      "loss": 0.4231,
+      "step": 1691
+    },
+    {
+      "epoch": 0.5072512647554807,
+      "grad_norm": 0.25751793384552,
+      "learning_rate": 0.0002242342342342342,
+      "loss": 0.4406,
+      "step": 1692
+    },
+    {
+      "epoch": 0.5075510586471801,
+      "grad_norm": 0.23062650859355927,
+      "learning_rate": 0.00022418918918918918,
+      "loss": 0.4273,
+      "step": 1693
+    },
+    {
+      "epoch": 0.5078508525388795,
+      "grad_norm": 0.2453629970550537,
+      "learning_rate": 0.00022414414414414412,
+      "loss": 0.4526,
+      "step": 1694
+    },
+    {
+      "epoch": 0.508150646430579,
+      "grad_norm": 0.26255685091018677,
+      "learning_rate": 0.00022409909909909905,
+      "loss": 0.461,
+      "step": 1695
+    },
+    {
+      "epoch": 0.5084504403222785,
+      "grad_norm": 0.2384280562400818,
+      "learning_rate": 0.00022405405405405405,
+      "loss": 0.4333,
+      "step": 1696
+    },
+    {
+      "epoch": 0.5087502342139779,
+      "grad_norm": 0.24055825173854828,
+      "learning_rate": 0.00022400900900900898,
+      "loss": 0.4472,
+      "step": 1697
+    },
+    {
+      "epoch": 0.5090500281056773,
+      "grad_norm": 0.24356570839881897,
+      "learning_rate": 0.00022396396396396395,
+      "loss": 0.4159,
+      "step": 1698
+    },
+    {
+      "epoch": 0.5093498219973768,
+      "grad_norm": 0.3042013645172119,
+      "learning_rate": 0.0002239189189189189,
+      "loss": 0.474,
+      "step": 1699
+    },
+    {
+      "epoch": 0.5096496158890763,
+      "grad_norm": 0.2500532567501068,
+      "learning_rate": 0.00022387387387387385,
+      "loss": 0.4223,
+      "step": 1700
+    },
+    {
+      "epoch": 0.5099494097807757,
+      "grad_norm": 0.25324761867523193,
+      "learning_rate": 0.0002238288288288288,
+      "loss": 0.452,
+      "step": 1701
+    },
+    {
+      "epoch": 0.5102492036724752,
+      "grad_norm": 0.26007258892059326,
+      "learning_rate": 0.00022378378378378377,
+      "loss": 0.4516,
+      "step": 1702
+    },
+    {
+      "epoch": 0.5105489975641746,
+      "grad_norm": 0.2657194435596466,
+      "learning_rate": 0.0002237387387387387,
+      "loss": 0.4663,
+      "step": 1703
+    },
+    {
+      "epoch": 0.5108487914558741,
+      "grad_norm": 0.28216373920440674,
+      "learning_rate": 0.00022369369369369368,
+      "loss": 0.4341,
+      "step": 1704
+    },
+    {
+      "epoch": 0.5111485853475736,
+      "grad_norm": 0.2571386992931366,
+      "learning_rate": 0.00022364864864864864,
+      "loss": 0.4532,
+      "step": 1705
+    },
+    {
+      "epoch": 0.511448379239273,
+      "grad_norm": 0.27189430594444275,
+      "learning_rate": 0.00022360360360360358,
+      "loss": 0.4593,
+      "step": 1706
+    },
+    {
+      "epoch": 0.5117481731309724,
+      "grad_norm": 0.2536429166793823,
+      "learning_rate": 0.00022355855855855854,
+      "loss": 0.4603,
+      "step": 1707
+    },
+    {
+      "epoch": 0.512047967022672,
+      "grad_norm": 0.2552615702152252,
+      "learning_rate": 0.0002235135135135135,
+      "loss": 0.4265,
+      "step": 1708
+    },
+    {
+      "epoch": 0.5123477609143714,
+      "grad_norm": 0.24926409125328064,
+      "learning_rate": 0.00022346846846846844,
+      "loss": 0.444,
+      "step": 1709
+    },
+    {
+      "epoch": 0.5126475548060708,
+      "grad_norm": 0.27449071407318115,
+      "learning_rate": 0.0002234234234234234,
+      "loss": 0.4605,
+      "step": 1710
+    },
+    {
+      "epoch": 0.5129473486977703,
+      "grad_norm": 0.255071222782135,
+      "learning_rate": 0.00022337837837837837,
+      "loss": 0.4409,
+      "step": 1711
+    },
+    {
+      "epoch": 0.5132471425894697,
+      "grad_norm": 0.2560432553291321,
+      "learning_rate": 0.00022333333333333333,
+      "loss": 0.467,
+      "step": 1712
+    },
+    {
+      "epoch": 0.5135469364811692,
+      "grad_norm": 0.2458151876926422,
+      "learning_rate": 0.00022328828828828827,
+      "loss": 0.4626,
+      "step": 1713
+    },
+    {
+      "epoch": 0.5138467303728687,
+      "grad_norm": 0.2361784130334854,
+      "learning_rate": 0.0002232432432432432,
+      "loss": 0.4305,
+      "step": 1714
+    },
+    {
+      "epoch": 0.5141465242645681,
+      "grad_norm": 0.25834766030311584,
+      "learning_rate": 0.0002231981981981982,
+      "loss": 0.4569,
+      "step": 1715
+    },
+    {
+      "epoch": 0.5144463181562675,
+      "grad_norm": 0.253662109375,
+      "learning_rate": 0.00022315315315315313,
+      "loss": 0.4289,
+      "step": 1716
+    },
+    {
+      "epoch": 0.514746112047967,
+      "grad_norm": 0.2442713975906372,
+      "learning_rate": 0.00022310810810810807,
+      "loss": 0.4356,
+      "step": 1717
+    },
+    {
+      "epoch": 0.5150459059396665,
+      "grad_norm": 0.25813472270965576,
+      "learning_rate": 0.00022306306306306306,
+      "loss": 0.4721,
+      "step": 1718
+    },
+    {
+      "epoch": 0.5153456998313659,
+      "grad_norm": 0.269927054643631,
+      "learning_rate": 0.000223018018018018,
+      "loss": 0.4603,
+      "step": 1719
+    },
+    {
+      "epoch": 0.5156454937230653,
+      "grad_norm": 0.25108182430267334,
+      "learning_rate": 0.00022297297297297293,
+      "loss": 0.4611,
+      "step": 1720
+    },
+    {
+      "epoch": 0.5159452876147649,
+      "grad_norm": 0.23553571105003357,
+      "learning_rate": 0.00022292792792792793,
+      "loss": 0.431,
+      "step": 1721
+    },
+    {
+      "epoch": 0.5162450815064643,
+      "grad_norm": 0.2411264032125473,
+      "learning_rate": 0.00022288288288288286,
+      "loss": 0.4405,
+      "step": 1722
+    },
+    {
+      "epoch": 0.5165448753981637,
+      "grad_norm": 0.24999505281448364,
+      "learning_rate": 0.0002228378378378378,
+      "loss": 0.4487,
+      "step": 1723
+    },
+    {
+      "epoch": 0.5168446692898632,
+      "grad_norm": 0.23619996011257172,
+      "learning_rate": 0.0002227927927927928,
+      "loss": 0.4399,
+      "step": 1724
+    },
+    {
+      "epoch": 0.5171444631815627,
+      "grad_norm": 0.23623579740524292,
+      "learning_rate": 0.00022274774774774773,
+      "loss": 0.4358,
+      "step": 1725
+    },
+    {
+      "epoch": 0.5174442570732621,
+      "grad_norm": 0.2538294494152069,
+      "learning_rate": 0.00022270270270270266,
+      "loss": 0.4715,
+      "step": 1726
+    },
+    {
+      "epoch": 0.5177440509649616,
+      "grad_norm": 0.23572376370429993,
+      "learning_rate": 0.00022265765765765765,
+      "loss": 0.4529,
+      "step": 1727
+    },
+    {
+      "epoch": 0.518043844856661,
+      "grad_norm": 0.25421515107154846,
+      "learning_rate": 0.0002226126126126126,
+      "loss": 0.4427,
+      "step": 1728
+    },
+    {
+      "epoch": 0.5183436387483605,
+      "grad_norm": 0.23936650156974792,
+      "learning_rate": 0.00022256756756756753,
+      "loss": 0.4416,
+      "step": 1729
+    },
+    {
+      "epoch": 0.51864343264006,
+      "grad_norm": 0.26791784167289734,
+      "learning_rate": 0.00022252252252252252,
+      "loss": 0.4697,
+      "step": 1730
+    },
+    {
+      "epoch": 0.5189432265317594,
+      "grad_norm": 0.2541804015636444,
+      "learning_rate": 0.00022247747747747746,
+      "loss": 0.4801,
+      "step": 1731
+    },
+    {
+      "epoch": 0.5192430204234588,
+      "grad_norm": 0.24337895214557648,
+      "learning_rate": 0.0002224324324324324,
+      "loss": 0.4316,
+      "step": 1732
+    },
+    {
+      "epoch": 0.5195428143151584,
+      "grad_norm": 0.2545047998428345,
+      "learning_rate": 0.00022238738738738738,
+      "loss": 0.4465,
+      "step": 1733
+    },
+    {
+      "epoch": 0.5198426082068578,
+      "grad_norm": 0.24640010297298431,
+      "learning_rate": 0.00022234234234234232,
+      "loss": 0.4574,
+      "step": 1734
+    },
+    {
+      "epoch": 0.5201424020985572,
+      "grad_norm": 0.2528793215751648,
+      "learning_rate": 0.00022229729729729728,
+      "loss": 0.4423,
+      "step": 1735
+    },
+    {
+      "epoch": 0.5204421959902567,
+      "grad_norm": 0.24697841703891754,
+      "learning_rate": 0.00022225225225225222,
+      "loss": 0.4259,
+      "step": 1736
+    },
+    {
+      "epoch": 0.5207419898819562,
+      "grad_norm": 0.24195986986160278,
+      "learning_rate": 0.00022220720720720718,
+      "loss": 0.439,
+      "step": 1737
+    },
+    {
+      "epoch": 0.5210417837736556,
+      "grad_norm": 0.2523336410522461,
+      "learning_rate": 0.00022216216216216215,
+      "loss": 0.4441,
+      "step": 1738
+    },
+    {
+      "epoch": 0.521341577665355,
+      "grad_norm": 0.22917968034744263,
+      "learning_rate": 0.00022211711711711709,
+      "loss": 0.4152,
+      "step": 1739
+    },
+    {
+      "epoch": 0.5216413715570545,
+      "grad_norm": 0.2633557915687561,
+      "learning_rate": 0.00022207207207207205,
+      "loss": 0.4593,
+      "step": 1740
+    },
+    {
+      "epoch": 0.521941165448754,
+      "grad_norm": 0.23756282031536102,
+      "learning_rate": 0.000222027027027027,
+      "loss": 0.4244,
+      "step": 1741
+    },
+    {
+      "epoch": 0.5222409593404534,
+      "grad_norm": 0.24200837314128876,
+      "learning_rate": 0.00022198198198198195,
+      "loss": 0.4056,
+      "step": 1742
+    },
+    {
+      "epoch": 0.5225407532321529,
+      "grad_norm": 0.2658364474773407,
+      "learning_rate": 0.00022193693693693691,
+      "loss": 0.4432,
+      "step": 1743
+    },
+    {
+      "epoch": 0.5228405471238523,
+      "grad_norm": 0.2403399795293808,
+      "learning_rate": 0.00022189189189189188,
+      "loss": 0.4395,
+      "step": 1744
+    },
+    {
+      "epoch": 0.5231403410155518,
+      "grad_norm": 0.23752082884311676,
+      "learning_rate": 0.00022184684684684681,
+      "loss": 0.4284,
+      "step": 1745
+    },
+    {
+      "epoch": 0.5234401349072513,
+      "grad_norm": 0.2684466540813446,
+      "learning_rate": 0.0002218018018018018,
+      "loss": 0.4266,
+      "step": 1746
+    },
+    {
+      "epoch": 0.5237399287989507,
+      "grad_norm": 0.24474304914474487,
+      "learning_rate": 0.00022175675675675674,
+      "loss": 0.4472,
+      "step": 1747
+    },
+    {
+      "epoch": 0.5240397226906501,
+      "grad_norm": 0.24958543479442596,
+      "learning_rate": 0.00022171171171171168,
+      "loss": 0.4253,
+      "step": 1748
+    },
+    {
+      "epoch": 0.5243395165823497,
+      "grad_norm": 0.25974011421203613,
+      "learning_rate": 0.00022166666666666667,
+      "loss": 0.4477,
+      "step": 1749
+    },
+    {
+      "epoch": 0.5246393104740491,
+      "grad_norm": 0.23762384057044983,
+      "learning_rate": 0.0002216216216216216,
+      "loss": 0.425,
+      "step": 1750
+    },
+    {
+      "epoch": 0.5249391043657485,
+      "grad_norm": 0.2543089687824249,
+      "learning_rate": 0.00022157657657657654,
+      "loss": 0.3947,
+      "step": 1751
+    },
+    {
+      "epoch": 0.525238898257448,
+      "grad_norm": 0.26260843873023987,
+      "learning_rate": 0.00022153153153153153,
+      "loss": 0.4642,
+      "step": 1752
+    },
+    {
+      "epoch": 0.5255386921491475,
+      "grad_norm": 0.2474740445613861,
+      "learning_rate": 0.00022148648648648647,
+      "loss": 0.4468,
+      "step": 1753
+    },
+    {
+      "epoch": 0.5258384860408469,
+      "grad_norm": 0.24307985603809357,
+      "learning_rate": 0.0002214414414414414,
+      "loss": 0.4537,
+      "step": 1754
+    },
+    {
+      "epoch": 0.5261382799325464,
+      "grad_norm": 0.27346494793891907,
+      "learning_rate": 0.0002213963963963964,
+      "loss": 0.4789,
+      "step": 1755
+    },
+    {
+      "epoch": 0.5264380738242458,
+      "grad_norm": 0.2407042384147644,
+      "learning_rate": 0.00022135135135135134,
+      "loss": 0.4334,
+      "step": 1756
+    },
+    {
+      "epoch": 0.5267378677159453,
+      "grad_norm": 0.2311742603778839,
+      "learning_rate": 0.00022130630630630627,
+      "loss": 0.4117,
+      "step": 1757
+    },
+    {
+      "epoch": 0.5270376616076448,
+      "grad_norm": 0.23060280084609985,
+      "learning_rate": 0.00022126126126126126,
+      "loss": 0.4413,
+      "step": 1758
+    },
+    {
+      "epoch": 0.5273374554993442,
+      "grad_norm": 0.286851167678833,
+      "learning_rate": 0.0002212162162162162,
+      "loss": 0.461,
+      "step": 1759
+    },
+    {
+      "epoch": 0.5276372493910436,
+      "grad_norm": 0.23764149844646454,
+      "learning_rate": 0.00022117117117117114,
+      "loss": 0.4428,
+      "step": 1760
+    },
+    {
+      "epoch": 0.5279370432827432,
+      "grad_norm": 0.24021115899085999,
+      "learning_rate": 0.0002211261261261261,
+      "loss": 0.4416,
+      "step": 1761
+    },
+    {
+      "epoch": 0.5282368371744426,
+      "grad_norm": 0.25310957431793213,
+      "learning_rate": 0.00022108108108108106,
+      "loss": 0.4474,
+      "step": 1762
+    },
+    {
+      "epoch": 0.528536631066142,
+      "grad_norm": 0.2636144161224365,
+      "learning_rate": 0.000221036036036036,
+      "loss": 0.4629,
+      "step": 1763
+    },
+    {
+      "epoch": 0.5288364249578414,
+      "grad_norm": 0.254807710647583,
+      "learning_rate": 0.00022099099099099097,
+      "loss": 0.4558,
+      "step": 1764
+    },
+    {
+      "epoch": 0.529136218849541,
+      "grad_norm": 0.2389029860496521,
+      "learning_rate": 0.00022094594594594593,
+      "loss": 0.4519,
+      "step": 1765
+    },
+    {
+      "epoch": 0.5294360127412404,
+      "grad_norm": 0.24269163608551025,
+      "learning_rate": 0.00022090090090090087,
+      "loss": 0.4371,
+      "step": 1766
+    },
+    {
+      "epoch": 0.5297358066329398,
+      "grad_norm": 0.2602348029613495,
+      "learning_rate": 0.00022085585585585583,
+      "loss": 0.4474,
+      "step": 1767
+    },
+    {
+      "epoch": 0.5300356005246393,
+      "grad_norm": 0.2557549774646759,
+      "learning_rate": 0.0002208108108108108,
+      "loss": 0.4709,
+      "step": 1768
+    },
+    {
+      "epoch": 0.5303353944163388,
+      "grad_norm": 0.27289271354675293,
+      "learning_rate": 0.00022076576576576576,
+      "loss": 0.4248,
+      "step": 1769
+    },
+    {
+      "epoch": 0.5306351883080382,
+      "grad_norm": 0.27305862307548523,
+      "learning_rate": 0.0002207207207207207,
+      "loss": 0.4566,
+      "step": 1770
+    },
+    {
+      "epoch": 0.5309349821997377,
+      "grad_norm": 0.2508034408092499,
+      "learning_rate": 0.00022067567567567566,
+      "loss": 0.4334,
+      "step": 1771
+    },
+    {
+      "epoch": 0.5312347760914371,
+      "grad_norm": 0.24729134142398834,
+      "learning_rate": 0.00022063063063063062,
+      "loss": 0.4251,
+      "step": 1772
+    },
+    {
+      "epoch": 0.5315345699831366,
+      "grad_norm": 0.23145246505737305,
+      "learning_rate": 0.00022058558558558556,
+      "loss": 0.3992,
+      "step": 1773
+    },
+    {
+      "epoch": 0.5318343638748361,
+      "grad_norm": 0.24190931022167206,
+      "learning_rate": 0.00022054054054054052,
+      "loss": 0.4183,
+      "step": 1774
+    },
+    {
+      "epoch": 0.5321341577665355,
+      "grad_norm": 0.2671104073524475,
+      "learning_rate": 0.0002204954954954955,
+      "loss": 0.4548,
+      "step": 1775
+    },
+    {
+      "epoch": 0.5324339516582349,
+      "grad_norm": 0.2361837476491928,
+      "learning_rate": 0.00022045045045045042,
+      "loss": 0.414,
+      "step": 1776
+    },
+    {
+      "epoch": 0.5327337455499345,
+      "grad_norm": 0.25529375672340393,
+      "learning_rate": 0.0002204054054054054,
+      "loss": 0.4345,
+      "step": 1777
+    },
+    {
+      "epoch": 0.5330335394416339,
+      "grad_norm": 0.26258841156959534,
+      "learning_rate": 0.00022036036036036035,
+      "loss": 0.4522,
+      "step": 1778
+    },
+    {
+      "epoch": 0.5333333333333333,
+      "grad_norm": 0.24532164633274078,
+      "learning_rate": 0.0002203153153153153,
+      "loss": 0.4303,
+      "step": 1779
+    },
+    {
+      "epoch": 0.5336331272250328,
+      "grad_norm": 0.24562229216098785,
+      "learning_rate": 0.00022027027027027028,
+      "loss": 0.427,
+      "step": 1780
+    },
+    {
+      "epoch": 0.5339329211167323,
+      "grad_norm": 0.25493496656417847,
+      "learning_rate": 0.00022022522522522522,
+      "loss": 0.45,
+      "step": 1781
+    },
+    {
+      "epoch": 0.5342327150084317,
+      "grad_norm": 0.25944793224334717,
+      "learning_rate": 0.00022018018018018015,
+      "loss": 0.449,
+      "step": 1782
+    },
+    {
+      "epoch": 0.5345325089001312,
+      "grad_norm": 0.24805709719657898,
+      "learning_rate": 0.00022013513513513514,
+      "loss": 0.4431,
+      "step": 1783
+    },
+    {
+      "epoch": 0.5348323027918306,
+      "grad_norm": 0.27083903551101685,
+      "learning_rate": 0.00022009009009009008,
+      "loss": 0.465,
+      "step": 1784
+    },
+    {
+      "epoch": 0.5351320966835301,
+      "grad_norm": 0.2876584827899933,
+      "learning_rate": 0.00022004504504504502,
+      "loss": 0.4859,
+      "step": 1785
+    },
+    {
+      "epoch": 0.5354318905752296,
+      "grad_norm": 0.2499544471502304,
+      "learning_rate": 0.00021999999999999995,
+      "loss": 0.4551,
+      "step": 1786
+    },
+    {
+      "epoch": 0.535731684466929,
+      "grad_norm": 0.2812490165233612,
+      "learning_rate": 0.00021995495495495494,
+      "loss": 0.4554,
+      "step": 1787
+    },
+    {
+      "epoch": 0.5360314783586284,
+      "grad_norm": 0.2530219256877899,
+      "learning_rate": 0.00021990990990990988,
+      "loss": 0.4335,
+      "step": 1788
+    },
+    {
+      "epoch": 0.536331272250328,
+      "grad_norm": 0.24573327600955963,
+      "learning_rate": 0.00021986486486486482,
+      "loss": 0.4574,
+      "step": 1789
+    },
+    {
+      "epoch": 0.5366310661420274,
+      "grad_norm": 0.3053630292415619,
+      "learning_rate": 0.0002198198198198198,
+      "loss": 0.4745,
+      "step": 1790
+    },
+    {
+      "epoch": 0.5369308600337268,
+      "grad_norm": 0.24176299571990967,
+      "learning_rate": 0.00021977477477477475,
+      "loss": 0.4244,
+      "step": 1791
+    },
+    {
+      "epoch": 0.5372306539254262,
+      "grad_norm": 0.27554550766944885,
+      "learning_rate": 0.0002197297297297297,
+      "loss": 0.469,
+      "step": 1792
+    },
+    {
+      "epoch": 0.5375304478171258,
+      "grad_norm": 0.28003454208374023,
+      "learning_rate": 0.00021968468468468467,
+      "loss": 0.4477,
+      "step": 1793
+    },
+    {
+      "epoch": 0.5378302417088252,
+      "grad_norm": 0.27388301491737366,
+      "learning_rate": 0.0002196396396396396,
+      "loss": 0.4602,
+      "step": 1794
+    },
+    {
+      "epoch": 0.5381300356005246,
+      "grad_norm": 0.2521568238735199,
+      "learning_rate": 0.00021959459459459457,
+      "loss": 0.4507,
+      "step": 1795
+    },
+    {
+      "epoch": 0.5384298294922241,
+      "grad_norm": 0.24554435908794403,
+      "learning_rate": 0.00021954954954954954,
+      "loss": 0.4387,
+      "step": 1796
+    },
+    {
+      "epoch": 0.5387296233839236,
+      "grad_norm": 0.24148909747600555,
+      "learning_rate": 0.00021950450450450447,
+      "loss": 0.4316,
+      "step": 1797
+    },
+    {
+      "epoch": 0.539029417275623,
+      "grad_norm": 0.24902650713920593,
+      "learning_rate": 0.00021945945945945944,
+      "loss": 0.448,
+      "step": 1798
+    },
+    {
+      "epoch": 0.5393292111673225,
+      "grad_norm": 0.25397536158561707,
+      "learning_rate": 0.0002194144144144144,
+      "loss": 0.479,
+      "step": 1799
+    },
+    {
+      "epoch": 0.5396290050590219,
+      "grad_norm": 0.2324562668800354,
+      "learning_rate": 0.00021936936936936934,
+      "loss": 0.4077,
+      "step": 1800
+    },
+    {
+      "epoch": 0.5399287989507214,
+      "grad_norm": 0.26509541273117065,
+      "learning_rate": 0.0002193243243243243,
+      "loss": 0.4628,
+      "step": 1801
+    },
+    {
+      "epoch": 0.5402285928424209,
+      "grad_norm": 0.24714629352092743,
+      "learning_rate": 0.00021927927927927927,
+      "loss": 0.4393,
+      "step": 1802
+    },
+    {
+      "epoch": 0.5405283867341203,
+      "grad_norm": 0.2634679079055786,
+      "learning_rate": 0.00021923423423423423,
+      "loss": 0.4508,
+      "step": 1803
+    },
+    {
+      "epoch": 0.5408281806258197,
+      "grad_norm": 0.2673392593860626,
+      "learning_rate": 0.00021918918918918917,
+      "loss": 0.4587,
+      "step": 1804
+    },
+    {
+      "epoch": 0.5411279745175192,
+      "grad_norm": 0.2438841462135315,
+      "learning_rate": 0.00021914414414414413,
+      "loss": 0.4271,
+      "step": 1805
+    },
+    {
+      "epoch": 0.5414277684092187,
+      "grad_norm": 0.2463088482618332,
+      "learning_rate": 0.0002190990990990991,
+      "loss": 0.4216,
+      "step": 1806
+    },
+    {
+      "epoch": 0.5417275623009181,
+      "grad_norm": 0.22443120181560516,
+      "learning_rate": 0.00021905405405405403,
+      "loss": 0.3935,
+      "step": 1807
+    },
+    {
+      "epoch": 0.5420273561926175,
+      "grad_norm": 0.2647371292114258,
+      "learning_rate": 0.00021900900900900897,
+      "loss": 0.4474,
+      "step": 1808
+    },
+    {
+      "epoch": 0.542327150084317,
+      "grad_norm": 0.2505868077278137,
+      "learning_rate": 0.00021896396396396396,
+      "loss": 0.4623,
+      "step": 1809
+    },
+    {
+      "epoch": 0.5426269439760165,
+      "grad_norm": 0.2395186722278595,
+      "learning_rate": 0.0002189189189189189,
+      "loss": 0.4469,
+      "step": 1810
+    },
+    {
+      "epoch": 0.542926737867716,
+      "grad_norm": 0.2345222383737564,
+      "learning_rate": 0.00021887387387387383,
+      "loss": 0.4127,
+      "step": 1811
+    },
+    {
+      "epoch": 0.5432265317594154,
+      "grad_norm": 0.25592970848083496,
+      "learning_rate": 0.00021882882882882882,
+      "loss": 0.4642,
+      "step": 1812
+    },
+    {
+      "epoch": 0.5435263256511148,
+      "grad_norm": 0.2604975700378418,
+      "learning_rate": 0.00021878378378378376,
+      "loss": 0.461,
+      "step": 1813
+    },
+    {
+      "epoch": 0.5438261195428143,
+      "grad_norm": 0.2413945347070694,
+      "learning_rate": 0.0002187387387387387,
+      "loss": 0.4281,
+      "step": 1814
+    },
+    {
+      "epoch": 0.5441259134345138,
+      "grad_norm": 0.2543574571609497,
+      "learning_rate": 0.0002186936936936937,
+      "loss": 0.4657,
+      "step": 1815
+    },
+    {
+      "epoch": 0.5444257073262132,
+      "grad_norm": 0.2783868610858917,
+      "learning_rate": 0.00021864864864864863,
+      "loss": 0.4917,
+      "step": 1816
+    },
+    {
+      "epoch": 0.5447255012179126,
+      "grad_norm": 0.23379693925380707,
+      "learning_rate": 0.00021860360360360356,
+      "loss": 0.4324,
+      "step": 1817
+    },
+    {
+      "epoch": 0.5450252951096122,
+      "grad_norm": 0.22307537496089935,
+      "learning_rate": 0.00021855855855855855,
+      "loss": 0.4175,
+      "step": 1818
+    },
+    {
+      "epoch": 0.5453250890013116,
+      "grad_norm": 0.23716312646865845,
+      "learning_rate": 0.0002185135135135135,
+      "loss": 0.4195,
+      "step": 1819
+    },
+    {
+      "epoch": 0.545624882893011,
+      "grad_norm": 0.2469712346792221,
+      "learning_rate": 0.00021846846846846843,
+      "loss": 0.4386,
+      "step": 1820
+    },
+    {
+      "epoch": 0.5459246767847105,
+      "grad_norm": 0.24840980768203735,
+      "learning_rate": 0.00021842342342342342,
+      "loss": 0.4193,
+      "step": 1821
+    },
+    {
+      "epoch": 0.54622447067641,
+      "grad_norm": 0.25362369418144226,
+      "learning_rate": 0.00021837837837837835,
+      "loss": 0.445,
+      "step": 1822
+    },
+    {
+      "epoch": 0.5465242645681094,
+      "grad_norm": 0.23812243342399597,
+      "learning_rate": 0.0002183333333333333,
+      "loss": 0.4511,
+      "step": 1823
+    },
+    {
+      "epoch": 0.5468240584598089,
+      "grad_norm": 0.25310468673706055,
+      "learning_rate": 0.00021828828828828828,
+      "loss": 0.4463,
+      "step": 1824
+    },
+    {
+      "epoch": 0.5471238523515083,
+      "grad_norm": 0.25943437218666077,
+      "learning_rate": 0.00021824324324324322,
+      "loss": 0.42,
+      "step": 1825
+    },
+    {
+      "epoch": 0.5474236462432078,
+      "grad_norm": 0.270178884267807,
+      "learning_rate": 0.00021819819819819818,
+      "loss": 0.4666,
+      "step": 1826
+    },
+    {
+      "epoch": 0.5477234401349073,
+      "grad_norm": 0.23301197588443756,
+      "learning_rate": 0.00021815315315315315,
+      "loss": 0.407,
+      "step": 1827
+    },
+    {
+      "epoch": 0.5480232340266067,
+      "grad_norm": 0.2634783983230591,
+      "learning_rate": 0.00021810810810810808,
+      "loss": 0.4596,
+      "step": 1828
+    },
+    {
+      "epoch": 0.5483230279183061,
+      "grad_norm": 0.24297159910202026,
+      "learning_rate": 0.00021806306306306305,
+      "loss": 0.4107,
+      "step": 1829
+    },
+    {
+      "epoch": 0.5486228218100057,
+      "grad_norm": 0.2781638503074646,
+      "learning_rate": 0.000218018018018018,
+      "loss": 0.4686,
+      "step": 1830
+    },
+    {
+      "epoch": 0.5489226157017051,
+      "grad_norm": 0.25801247358322144,
+      "learning_rate": 0.00021797297297297295,
+      "loss": 0.4644,
+      "step": 1831
+    },
+    {
+      "epoch": 0.5492224095934045,
+      "grad_norm": 0.27080485224723816,
+      "learning_rate": 0.0002179279279279279,
+      "loss": 0.428,
+      "step": 1832
+    },
+    {
+      "epoch": 0.5495222034851039,
+      "grad_norm": 0.2617061138153076,
+      "learning_rate": 0.00021788288288288285,
+      "loss": 0.4437,
+      "step": 1833
+    },
+    {
+      "epoch": 0.5498219973768035,
+      "grad_norm": 0.26249265670776367,
+      "learning_rate": 0.0002178378378378378,
+      "loss": 0.4413,
+      "step": 1834
+    },
+    {
+      "epoch": 0.5501217912685029,
+      "grad_norm": 0.2705504894256592,
+      "learning_rate": 0.00021779279279279278,
+      "loss": 0.4716,
+      "step": 1835
+    },
+    {
+      "epoch": 0.5504215851602023,
+      "grad_norm": 0.25836360454559326,
+      "learning_rate": 0.0002177477477477477,
+      "loss": 0.454,
+      "step": 1836
+    },
+    {
+      "epoch": 0.5507213790519018,
+      "grad_norm": 0.23872222006320953,
+      "learning_rate": 0.0002177027027027027,
+      "loss": 0.4173,
+      "step": 1837
+    },
+    {
+      "epoch": 0.5510211729436013,
+      "grad_norm": 0.2438534051179886,
+      "learning_rate": 0.00021765765765765764,
+      "loss": 0.4114,
+      "step": 1838
+    },
+    {
+      "epoch": 0.5513209668353007,
+      "grad_norm": 0.2600691318511963,
+      "learning_rate": 0.00021761261261261258,
+      "loss": 0.4364,
+      "step": 1839
+    },
+    {
+      "epoch": 0.5516207607270002,
+      "grad_norm": 0.2483833283185959,
+      "learning_rate": 0.00021756756756756757,
+      "loss": 0.4544,
+      "step": 1840
+    },
+    {
+      "epoch": 0.5519205546186996,
+      "grad_norm": 0.2506955564022064,
+      "learning_rate": 0.0002175225225225225,
+      "loss": 0.4561,
+      "step": 1841
+    },
+    {
+      "epoch": 0.5522203485103991,
+      "grad_norm": 0.2800386846065521,
+      "learning_rate": 0.00021747747747747744,
+      "loss": 0.438,
+      "step": 1842
+    },
+    {
+      "epoch": 0.5525201424020986,
+      "grad_norm": 0.2458580881357193,
+      "learning_rate": 0.00021743243243243243,
+      "loss": 0.4431,
+      "step": 1843
+    },
+    {
+      "epoch": 0.552819936293798,
+      "grad_norm": 0.24995484948158264,
+      "learning_rate": 0.00021738738738738737,
+      "loss": 0.4349,
+      "step": 1844
+    },
+    {
+      "epoch": 0.5531197301854974,
+      "grad_norm": 0.2502653896808624,
+      "learning_rate": 0.0002173423423423423,
+      "loss": 0.4366,
+      "step": 1845
+    },
+    {
+      "epoch": 0.553419524077197,
+      "grad_norm": 0.2538207471370697,
+      "learning_rate": 0.0002172972972972973,
+      "loss": 0.4381,
+      "step": 1846
+    },
+    {
+      "epoch": 0.5537193179688964,
+      "grad_norm": 0.2417684644460678,
+      "learning_rate": 0.00021725225225225223,
+      "loss": 0.4277,
+      "step": 1847
+    },
+    {
+      "epoch": 0.5540191118605958,
+      "grad_norm": 0.2531186044216156,
+      "learning_rate": 0.00021720720720720717,
+      "loss": 0.4361,
+      "step": 1848
+    },
+    {
+      "epoch": 0.5543189057522953,
+      "grad_norm": 0.2551022171974182,
+      "learning_rate": 0.00021716216216216216,
+      "loss": 0.4483,
+      "step": 1849
+    },
+    {
+      "epoch": 0.5546186996439948,
+      "grad_norm": 0.23375505208969116,
+      "learning_rate": 0.0002171171171171171,
+      "loss": 0.4157,
+      "step": 1850
+    },
+    {
+      "epoch": 0.5549184935356942,
+      "grad_norm": 0.23428985476493835,
+      "learning_rate": 0.00021707207207207204,
+      "loss": 0.4092,
+      "step": 1851
+    },
+    {
+      "epoch": 0.5552182874273937,
+      "grad_norm": 0.24895919859409332,
+      "learning_rate": 0.00021702702702702703,
+      "loss": 0.4377,
+      "step": 1852
+    },
+    {
+      "epoch": 0.5555180813190931,
+      "grad_norm": 0.23306754231452942,
+      "learning_rate": 0.00021698198198198196,
+      "loss": 0.4316,
+      "step": 1853
+    },
+    {
+      "epoch": 0.5558178752107926,
+      "grad_norm": 0.2527741491794586,
+      "learning_rate": 0.0002169369369369369,
+      "loss": 0.4245,
+      "step": 1854
+    },
+    {
+      "epoch": 0.556117669102492,
+      "grad_norm": 0.2319868952035904,
+      "learning_rate": 0.0002168918918918919,
+      "loss": 0.4074,
+      "step": 1855
+    },
+    {
+      "epoch": 0.5564174629941915,
+      "grad_norm": 0.25132691860198975,
+      "learning_rate": 0.00021684684684684683,
+      "loss": 0.4378,
+      "step": 1856
+    },
+    {
+      "epoch": 0.5567172568858909,
+      "grad_norm": 0.24814215302467346,
+      "learning_rate": 0.00021680180180180177,
+      "loss": 0.4639,
+      "step": 1857
+    },
+    {
+      "epoch": 0.5570170507775905,
+      "grad_norm": 0.23483987152576447,
+      "learning_rate": 0.00021675675675675673,
+      "loss": 0.4419,
+      "step": 1858
+    },
+    {
+      "epoch": 0.5573168446692899,
+      "grad_norm": 0.24670268595218658,
+      "learning_rate": 0.0002167117117117117,
+      "loss": 0.436,
+      "step": 1859
+    },
+    {
+      "epoch": 0.5576166385609893,
+      "grad_norm": 0.23737584054470062,
+      "learning_rate": 0.00021666666666666666,
+      "loss": 0.4246,
+      "step": 1860
+    },
+    {
+      "epoch": 0.5579164324526887,
+      "grad_norm": 0.2634913921356201,
+      "learning_rate": 0.0002166216216216216,
+      "loss": 0.4252,
+      "step": 1861
+    },
+    {
+      "epoch": 0.5582162263443883,
+      "grad_norm": 0.25565293431282043,
+      "learning_rate": 0.00021657657657657656,
+      "loss": 0.4287,
+      "step": 1862
+    },
+    {
+      "epoch": 0.5585160202360877,
+      "grad_norm": 0.2408502846956253,
+      "learning_rate": 0.00021653153153153152,
+      "loss": 0.4259,
+      "step": 1863
+    },
+    {
+      "epoch": 0.5588158141277871,
+      "grad_norm": 0.2516701817512512,
+      "learning_rate": 0.00021648648648648646,
+      "loss": 0.4361,
+      "step": 1864
+    },
+    {
+      "epoch": 0.5591156080194866,
+      "grad_norm": 0.2645619213581085,
+      "learning_rate": 0.00021644144144144142,
+      "loss": 0.4336,
+      "step": 1865
+    },
+    {
+      "epoch": 0.5594154019111861,
+      "grad_norm": 0.27260729670524597,
+      "learning_rate": 0.00021639639639639639,
+      "loss": 0.4494,
+      "step": 1866
+    },
+    {
+      "epoch": 0.5597151958028855,
+      "grad_norm": 0.2506016492843628,
+      "learning_rate": 0.00021635135135135132,
+      "loss": 0.4027,
+      "step": 1867
+    },
+    {
+      "epoch": 0.560014989694585,
+      "grad_norm": 0.2552817463874817,
+      "learning_rate": 0.00021630630630630629,
+      "loss": 0.426,
+      "step": 1868
+    },
+    {
+      "epoch": 0.5603147835862844,
+      "grad_norm": 0.2529433071613312,
+      "learning_rate": 0.00021626126126126125,
+      "loss": 0.4327,
+      "step": 1869
+    },
+    {
+      "epoch": 0.5606145774779839,
+      "grad_norm": 0.2483612447977066,
+      "learning_rate": 0.0002162162162162162,
+      "loss": 0.4492,
+      "step": 1870
+    },
+    {
+      "epoch": 0.5609143713696834,
+      "grad_norm": 0.25298991799354553,
+      "learning_rate": 0.00021617117117117118,
+      "loss": 0.4442,
+      "step": 1871
+    },
+    {
+      "epoch": 0.5612141652613828,
+      "grad_norm": 0.243782639503479,
+      "learning_rate": 0.00021612612612612611,
+      "loss": 0.4462,
+      "step": 1872
+    },
+    {
+      "epoch": 0.5615139591530822,
+      "grad_norm": 0.24677562713623047,
+      "learning_rate": 0.00021608108108108105,
+      "loss": 0.433,
+      "step": 1873
+    },
+    {
+      "epoch": 0.5618137530447818,
+      "grad_norm": 0.2761828601360321,
+      "learning_rate": 0.00021603603603603604,
+      "loss": 0.4436,
+      "step": 1874
+    },
+    {
+      "epoch": 0.5621135469364812,
+      "grad_norm": 0.24887466430664062,
+      "learning_rate": 0.00021599099099099098,
+      "loss": 0.4211,
+      "step": 1875
+    },
+    {
+      "epoch": 0.5624133408281806,
+      "grad_norm": 0.2687954902648926,
+      "learning_rate": 0.00021594594594594592,
+      "loss": 0.4545,
+      "step": 1876
+    },
+    {
+      "epoch": 0.56271313471988,
+      "grad_norm": 0.2376958280801773,
+      "learning_rate": 0.0002159009009009009,
+      "loss": 0.4062,
+      "step": 1877
+    },
+    {
+      "epoch": 0.5630129286115796,
+      "grad_norm": 0.25497403740882874,
+      "learning_rate": 0.00021585585585585584,
+      "loss": 0.4512,
+      "step": 1878
+    },
+    {
+      "epoch": 0.563312722503279,
+      "grad_norm": 0.27644020318984985,
+      "learning_rate": 0.00021581081081081078,
+      "loss": 0.4317,
+      "step": 1879
+    },
+    {
+      "epoch": 0.5636125163949784,
+      "grad_norm": 0.274746298789978,
+      "learning_rate": 0.00021576576576576577,
+      "loss": 0.4523,
+      "step": 1880
+    },
+    {
+      "epoch": 0.5639123102866779,
+      "grad_norm": 0.24522797763347626,
+      "learning_rate": 0.0002157207207207207,
+      "loss": 0.4193,
+      "step": 1881
+    },
+    {
+      "epoch": 0.5642121041783774,
+      "grad_norm": 0.26614081859588623,
+      "learning_rate": 0.00021567567567567565,
+      "loss": 0.4474,
+      "step": 1882
+    },
+    {
+      "epoch": 0.5645118980700768,
+      "grad_norm": 0.25119736790657043,
+      "learning_rate": 0.0002156306306306306,
+      "loss": 0.4349,
+      "step": 1883
+    },
+    {
+      "epoch": 0.5648116919617763,
+      "grad_norm": 0.2535664737224579,
+      "learning_rate": 0.00021558558558558557,
+      "loss": 0.4395,
+      "step": 1884
+    },
+    {
+      "epoch": 0.5651114858534757,
+      "grad_norm": 0.27441659569740295,
+      "learning_rate": 0.0002155405405405405,
+      "loss": 0.4591,
+      "step": 1885
+    },
+    {
+      "epoch": 0.5654112797451752,
+      "grad_norm": 0.2535187900066376,
+      "learning_rate": 0.00021549549549549547,
+      "loss": 0.4256,
+      "step": 1886
+    },
+    {
+      "epoch": 0.5657110736368747,
+      "grad_norm": 0.2877647876739502,
+      "learning_rate": 0.00021545045045045044,
+      "loss": 0.4877,
+      "step": 1887
+    },
+    {
+      "epoch": 0.5660108675285741,
+      "grad_norm": 0.23665641248226166,
+      "learning_rate": 0.00021540540540540537,
+      "loss": 0.4506,
+      "step": 1888
+    },
+    {
+      "epoch": 0.5663106614202735,
+      "grad_norm": 0.2620941996574402,
+      "learning_rate": 0.00021536036036036034,
+      "loss": 0.4415,
+      "step": 1889
+    },
+    {
+      "epoch": 0.5666104553119731,
+      "grad_norm": 0.2704925239086151,
+      "learning_rate": 0.0002153153153153153,
+      "loss": 0.4855,
+      "step": 1890
+    },
+    {
+      "epoch": 0.5669102492036725,
+      "grad_norm": 0.2636096775531769,
+      "learning_rate": 0.00021527027027027024,
+      "loss": 0.4196,
+      "step": 1891
+    },
+    {
+      "epoch": 0.5672100430953719,
+      "grad_norm": 0.2897530496120453,
+      "learning_rate": 0.0002152252252252252,
+      "loss": 0.4556,
+      "step": 1892
+    },
+    {
+      "epoch": 0.5675098369870714,
+      "grad_norm": 0.2407667189836502,
+      "learning_rate": 0.00021518018018018017,
+      "loss": 0.4406,
+      "step": 1893
+    },
+    {
+      "epoch": 0.5678096308787709,
+      "grad_norm": 0.277865469455719,
+      "learning_rate": 0.00021513513513513513,
+      "loss": 0.4033,
+      "step": 1894
+    },
+    {
+      "epoch": 0.5681094247704703,
+      "grad_norm": 0.2540576159954071,
+      "learning_rate": 0.00021509009009009007,
+      "loss": 0.4355,
+      "step": 1895
+    },
+    {
+      "epoch": 0.5684092186621698,
+      "grad_norm": 0.26155397295951843,
+      "learning_rate": 0.00021504504504504503,
+      "loss": 0.4235,
+      "step": 1896
+    },
+    {
+      "epoch": 0.5687090125538692,
+      "grad_norm": 0.25544247031211853,
+      "learning_rate": 0.000215,
+      "loss": 0.4279,
+      "step": 1897
+    },
+    {
+      "epoch": 0.5690088064455687,
+      "grad_norm": 0.2406405806541443,
+      "learning_rate": 0.00021495495495495493,
+      "loss": 0.4208,
+      "step": 1898
+    },
+    {
+      "epoch": 0.5693086003372682,
+      "grad_norm": 0.24712449312210083,
+      "learning_rate": 0.0002149099099099099,
+      "loss": 0.4344,
+      "step": 1899
+    },
+    {
+      "epoch": 0.5696083942289676,
+      "grad_norm": 0.26752716302871704,
+      "learning_rate": 0.00021486486486486486,
+      "loss": 0.4934,
+      "step": 1900
+    },
+    {
+      "epoch": 0.569908188120667,
+      "grad_norm": 0.24652232229709625,
+      "learning_rate": 0.0002148198198198198,
+      "loss": 0.4546,
+      "step": 1901
+    },
+    {
+      "epoch": 0.5702079820123664,
+      "grad_norm": 0.251396507024765,
+      "learning_rate": 0.00021477477477477476,
+      "loss": 0.4738,
+      "step": 1902
+    },
+    {
+      "epoch": 0.570507775904066,
+      "grad_norm": 0.24406296014785767,
+      "learning_rate": 0.00021472972972972972,
+      "loss": 0.4536,
+      "step": 1903
+    },
+    {
+      "epoch": 0.5708075697957654,
+      "grad_norm": 0.2634785771369934,
+      "learning_rate": 0.00021468468468468466,
+      "loss": 0.4653,
+      "step": 1904
+    },
+    {
+      "epoch": 0.5711073636874648,
+      "grad_norm": 0.24010764062404633,
+      "learning_rate": 0.0002146396396396396,
+      "loss": 0.4188,
+      "step": 1905
+    },
+    {
+      "epoch": 0.5714071575791643,
+      "grad_norm": 0.24869616329669952,
+      "learning_rate": 0.0002145945945945946,
+      "loss": 0.4474,
+      "step": 1906
+    },
+    {
+      "epoch": 0.5717069514708638,
+      "grad_norm": 0.24654364585876465,
+      "learning_rate": 0.00021454954954954953,
+      "loss": 0.4409,
+      "step": 1907
+    },
+    {
+      "epoch": 0.5720067453625632,
+      "grad_norm": 0.29856958985328674,
+      "learning_rate": 0.00021450450450450446,
+      "loss": 0.4912,
+      "step": 1908
+    },
+    {
+      "epoch": 0.5723065392542627,
+      "grad_norm": 0.2449256181716919,
+      "learning_rate": 0.00021445945945945945,
+      "loss": 0.4501,
+      "step": 1909
+    },
+    {
+      "epoch": 0.5726063331459621,
+      "grad_norm": 0.29776662588119507,
+      "learning_rate": 0.0002144144144144144,
+      "loss": 0.4403,
+      "step": 1910
+    },
+    {
+      "epoch": 0.5729061270376616,
+      "grad_norm": 0.26075392961502075,
+      "learning_rate": 0.00021436936936936933,
+      "loss": 0.4325,
+      "step": 1911
+    },
+    {
+      "epoch": 0.5732059209293611,
+      "grad_norm": 0.23287932574748993,
+      "learning_rate": 0.00021432432432432432,
+      "loss": 0.4265,
+      "step": 1912
+    },
+    {
+      "epoch": 0.5735057148210605,
+      "grad_norm": 0.2624457776546478,
+      "learning_rate": 0.00021427927927927925,
+      "loss": 0.4433,
+      "step": 1913
+    },
+    {
+      "epoch": 0.5738055087127599,
+      "grad_norm": 0.23440878093242645,
+      "learning_rate": 0.0002142342342342342,
+      "loss": 0.4136,
+      "step": 1914
+    },
+    {
+      "epoch": 0.5741053026044595,
+      "grad_norm": 0.2694714069366455,
+      "learning_rate": 0.00021418918918918918,
+      "loss": 0.4508,
+      "step": 1915
+    },
+    {
+      "epoch": 0.5744050964961589,
+      "grad_norm": 0.261952668428421,
+      "learning_rate": 0.00021414414414414412,
+      "loss": 0.4697,
+      "step": 1916
+    },
+    {
+      "epoch": 0.5747048903878583,
+      "grad_norm": 0.2526634931564331,
+      "learning_rate": 0.00021409909909909908,
+      "loss": 0.4518,
+      "step": 1917
+    },
+    {
+      "epoch": 0.5750046842795578,
+      "grad_norm": 0.2554527223110199,
+      "learning_rate": 0.00021405405405405405,
+      "loss": 0.4802,
+      "step": 1918
+    },
+    {
+      "epoch": 0.5753044781712573,
+      "grad_norm": 0.2729927599430084,
+      "learning_rate": 0.00021400900900900898,
+      "loss": 0.4622,
+      "step": 1919
+    },
+    {
+      "epoch": 0.5756042720629567,
+      "grad_norm": 0.26347601413726807,
+      "learning_rate": 0.00021396396396396395,
+      "loss": 0.4709,
+      "step": 1920
+    },
+    {
+      "epoch": 0.5759040659546562,
+      "grad_norm": 0.27795466780662537,
+      "learning_rate": 0.0002139189189189189,
+      "loss": 0.4555,
+      "step": 1921
+    },
+    {
+      "epoch": 0.5762038598463556,
+      "grad_norm": 0.25687262415885925,
+      "learning_rate": 0.00021387387387387385,
+      "loss": 0.4459,
+      "step": 1922
+    },
+    {
+      "epoch": 0.5765036537380551,
+      "grad_norm": 0.24855007231235504,
+      "learning_rate": 0.0002138288288288288,
+      "loss": 0.3944,
+      "step": 1923
+    },
+    {
+      "epoch": 0.5768034476297546,
+      "grad_norm": 0.26772525906562805,
+      "learning_rate": 0.00021378378378378378,
+      "loss": 0.4557,
+      "step": 1924
+    },
+    {
+      "epoch": 0.577103241521454,
+      "grad_norm": 0.2783243954181671,
+      "learning_rate": 0.0002137387387387387,
+      "loss": 0.4507,
+      "step": 1925
+    },
+    {
+      "epoch": 0.5774030354131534,
+      "grad_norm": 0.26769769191741943,
+      "learning_rate": 0.00021369369369369368,
+      "loss": 0.4371,
+      "step": 1926
+    },
+    {
+      "epoch": 0.577702829304853,
+      "grad_norm": 0.2783718705177307,
+      "learning_rate": 0.00021364864864864864,
+      "loss": 0.4269,
+      "step": 1927
+    },
+    {
+      "epoch": 0.5780026231965524,
+      "grad_norm": 0.2654764652252197,
+      "learning_rate": 0.0002136036036036036,
+      "loss": 0.4407,
+      "step": 1928
+    },
+    {
+      "epoch": 0.5783024170882518,
+      "grad_norm": 0.2694533169269562,
+      "learning_rate": 0.00021355855855855854,
+      "loss": 0.4822,
+      "step": 1929
+    },
+    {
+      "epoch": 0.5786022109799512,
+      "grad_norm": 0.2403596192598343,
+      "learning_rate": 0.00021351351351351348,
+      "loss": 0.3985,
+      "step": 1930
+    },
+    {
+      "epoch": 0.5789020048716508,
+      "grad_norm": 0.26435697078704834,
+      "learning_rate": 0.00021346846846846847,
+      "loss": 0.4547,
+      "step": 1931
+    },
+    {
+      "epoch": 0.5792017987633502,
+      "grad_norm": 0.2416696548461914,
+      "learning_rate": 0.0002134234234234234,
+      "loss": 0.406,
+      "step": 1932
+    },
+    {
+      "epoch": 0.5795015926550496,
+      "grad_norm": 0.2664091885089874,
+      "learning_rate": 0.00021337837837837834,
+      "loss": 0.412,
+      "step": 1933
+    },
+    {
+      "epoch": 0.5798013865467491,
+      "grad_norm": 0.24384742975234985,
+      "learning_rate": 0.00021333333333333333,
+      "loss": 0.4621,
+      "step": 1934
+    },
+    {
+      "epoch": 0.5801011804384486,
+      "grad_norm": 0.27569761872291565,
+      "learning_rate": 0.00021328828828828827,
+      "loss": 0.4643,
+      "step": 1935
+    },
+    {
+      "epoch": 0.580400974330148,
+      "grad_norm": 0.2567230761051178,
+      "learning_rate": 0.0002132432432432432,
+      "loss": 0.4231,
+      "step": 1936
+    },
+    {
+      "epoch": 0.5807007682218475,
+      "grad_norm": 0.25014641880989075,
+      "learning_rate": 0.0002131981981981982,
+      "loss": 0.4461,
+      "step": 1937
+    },
+    {
+      "epoch": 0.5810005621135469,
+      "grad_norm": 0.2679264545440674,
+      "learning_rate": 0.00021315315315315313,
+      "loss": 0.4321,
+      "step": 1938
+    },
+    {
+      "epoch": 0.5813003560052464,
+      "grad_norm": 0.2585947513580322,
+      "learning_rate": 0.00021310810810810807,
+      "loss": 0.4486,
+      "step": 1939
+    },
+    {
+      "epoch": 0.5816001498969459,
+      "grad_norm": 0.2640276253223419,
+      "learning_rate": 0.00021306306306306306,
+      "loss": 0.4439,
+      "step": 1940
+    },
+    {
+      "epoch": 0.5818999437886453,
+      "grad_norm": 0.2556924521923065,
+      "learning_rate": 0.000213018018018018,
+      "loss": 0.4557,
+      "step": 1941
+    },
+    {
+      "epoch": 0.5821997376803447,
+      "grad_norm": 0.2560097873210907,
+      "learning_rate": 0.00021297297297297294,
+      "loss": 0.4384,
+      "step": 1942
+    },
+    {
+      "epoch": 0.5824995315720443,
+      "grad_norm": 0.25773030519485474,
+      "learning_rate": 0.00021292792792792793,
+      "loss": 0.4289,
+      "step": 1943
+    },
+    {
+      "epoch": 0.5827993254637437,
+      "grad_norm": 0.26476195454597473,
+      "learning_rate": 0.00021288288288288286,
+      "loss": 0.4513,
+      "step": 1944
+    },
+    {
+      "epoch": 0.5830991193554431,
+      "grad_norm": 0.26929306983947754,
+      "learning_rate": 0.0002128378378378378,
+      "loss": 0.4464,
+      "step": 1945
+    },
+    {
+      "epoch": 0.5833989132471425,
+      "grad_norm": 0.25713875889778137,
+      "learning_rate": 0.0002127927927927928,
+      "loss": 0.4422,
+      "step": 1946
+    },
+    {
+      "epoch": 0.5836987071388421,
+      "grad_norm": 0.24906396865844727,
+      "learning_rate": 0.00021274774774774773,
+      "loss": 0.4454,
+      "step": 1947
+    },
+    {
+      "epoch": 0.5839985010305415,
+      "grad_norm": 0.24760214984416962,
+      "learning_rate": 0.00021270270270270266,
+      "loss": 0.4343,
+      "step": 1948
+    },
+    {
+      "epoch": 0.584298294922241,
+      "grad_norm": 0.25164082646369934,
+      "learning_rate": 0.00021265765765765766,
+      "loss": 0.4455,
+      "step": 1949
+    },
+    {
+      "epoch": 0.5845980888139404,
+      "grad_norm": 0.24564093351364136,
+      "learning_rate": 0.0002126126126126126,
+      "loss": 0.4389,
+      "step": 1950
+    },
+    {
+      "epoch": 0.5848978827056399,
+      "grad_norm": 0.2754795551300049,
+      "learning_rate": 0.00021256756756756756,
+      "loss": 0.466,
+      "step": 1951
+    },
+    {
+      "epoch": 0.5851976765973393,
+      "grad_norm": 0.2439223974943161,
+      "learning_rate": 0.00021252252252252252,
+      "loss": 0.4425,
+      "step": 1952
+    },
+    {
+      "epoch": 0.5854974704890388,
+      "grad_norm": 0.24320489168167114,
+      "learning_rate": 0.00021247747747747746,
+      "loss": 0.4085,
+      "step": 1953
+    },
+    {
+      "epoch": 0.5857972643807382,
+      "grad_norm": 0.2758175730705261,
+      "learning_rate": 0.00021243243243243242,
+      "loss": 0.4606,
+      "step": 1954
+    },
+    {
+      "epoch": 0.5860970582724377,
+      "grad_norm": 0.2623588442802429,
+      "learning_rate": 0.00021238738738738736,
+      "loss": 0.4167,
+      "step": 1955
+    },
+    {
+      "epoch": 0.5863968521641372,
+      "grad_norm": 0.2561275064945221,
+      "learning_rate": 0.00021234234234234232,
+      "loss": 0.4399,
+      "step": 1956
+    },
+    {
+      "epoch": 0.5866966460558366,
+      "grad_norm": 0.2818892300128937,
+      "learning_rate": 0.00021229729729729728,
+      "loss": 0.4628,
+      "step": 1957
+    },
+    {
+      "epoch": 0.586996439947536,
+      "grad_norm": 0.26609155535697937,
+      "learning_rate": 0.00021225225225225222,
+      "loss": 0.4285,
+      "step": 1958
+    },
+    {
+      "epoch": 0.5872962338392356,
+      "grad_norm": 0.2503769099712372,
+      "learning_rate": 0.00021220720720720719,
+      "loss": 0.4632,
+      "step": 1959
+    },
+    {
+      "epoch": 0.587596027730935,
+      "grad_norm": 0.2810426950454712,
+      "learning_rate": 0.00021216216216216215,
+      "loss": 0.4591,
+      "step": 1960
+    },
+    {
+      "epoch": 0.5878958216226344,
+      "grad_norm": 0.2517390251159668,
+      "learning_rate": 0.00021211711711711709,
+      "loss": 0.4323,
+      "step": 1961
+    },
+    {
+      "epoch": 0.5881956155143339,
+      "grad_norm": 0.26425543427467346,
+      "learning_rate": 0.00021207207207207205,
+      "loss": 0.4466,
+      "step": 1962
+    },
+    {
+      "epoch": 0.5884954094060334,
+      "grad_norm": 0.25192978978157043,
+      "learning_rate": 0.00021202702702702701,
+      "loss": 0.4524,
+      "step": 1963
+    },
+    {
+      "epoch": 0.5887952032977328,
+      "grad_norm": 0.26266035437583923,
+      "learning_rate": 0.00021198198198198195,
+      "loss": 0.4417,
+      "step": 1964
+    },
+    {
+      "epoch": 0.5890949971894323,
+      "grad_norm": 0.31183677911758423,
+      "learning_rate": 0.00021193693693693694,
+      "loss": 0.4473,
+      "step": 1965
+    },
+    {
+      "epoch": 0.5893947910811317,
+      "grad_norm": 0.24742548167705536,
+      "learning_rate": 0.00021189189189189188,
+      "loss": 0.4209,
+      "step": 1966
+    },
+    {
+      "epoch": 0.5896945849728312,
+      "grad_norm": 0.27282243967056274,
+      "learning_rate": 0.00021184684684684682,
+      "loss": 0.4316,
+      "step": 1967
+    },
+    {
+      "epoch": 0.5899943788645307,
+      "grad_norm": 0.2748444974422455,
+      "learning_rate": 0.0002118018018018018,
+      "loss": 0.4276,
+      "step": 1968
+    },
+    {
+      "epoch": 0.5902941727562301,
+      "grad_norm": 0.2746492624282837,
+      "learning_rate": 0.00021175675675675674,
+      "loss": 0.4597,
+      "step": 1969
+    },
+    {
+      "epoch": 0.5905939666479295,
+      "grad_norm": 0.25905507802963257,
+      "learning_rate": 0.00021171171171171168,
+      "loss": 0.4668,
+      "step": 1970
+    },
+    {
+      "epoch": 0.590893760539629,
+      "grad_norm": 0.2456134557723999,
+      "learning_rate": 0.00021166666666666667,
+      "loss": 0.4057,
+      "step": 1971
+    },
+    {
+      "epoch": 0.5911935544313285,
+      "grad_norm": 0.25221529603004456,
+      "learning_rate": 0.0002116216216216216,
+      "loss": 0.4427,
+      "step": 1972
+    },
+    {
+      "epoch": 0.5914933483230279,
+      "grad_norm": 0.24493472278118134,
+      "learning_rate": 0.00021157657657657654,
+      "loss": 0.4206,
+      "step": 1973
+    },
+    {
+      "epoch": 0.5917931422147273,
+      "grad_norm": 0.2561238408088684,
+      "learning_rate": 0.00021153153153153154,
+      "loss": 0.4537,
+      "step": 1974
+    },
+    {
+      "epoch": 0.5920929361064269,
+      "grad_norm": 0.2350313663482666,
+      "learning_rate": 0.00021148648648648647,
+      "loss": 0.4163,
+      "step": 1975
+    },
+    {
+      "epoch": 0.5923927299981263,
+      "grad_norm": 0.2535955309867859,
+      "learning_rate": 0.0002114414414414414,
+      "loss": 0.4435,
+      "step": 1976
+    },
+    {
+      "epoch": 0.5926925238898257,
+      "grad_norm": 0.2498706728219986,
+      "learning_rate": 0.00021139639639639637,
+      "loss": 0.4507,
+      "step": 1977
+    },
+    {
+      "epoch": 0.5929923177815252,
+      "grad_norm": 0.2510056495666504,
+      "learning_rate": 0.00021135135135135134,
+      "loss": 0.4268,
+      "step": 1978
+    },
+    {
+      "epoch": 0.5932921116732247,
+      "grad_norm": 0.252605676651001,
+      "learning_rate": 0.00021130630630630627,
+      "loss": 0.4457,
+      "step": 1979
+    },
+    {
+      "epoch": 0.5935919055649241,
+      "grad_norm": 0.2618269622325897,
+      "learning_rate": 0.00021126126126126124,
+      "loss": 0.4447,
+      "step": 1980
+    },
+    {
+      "epoch": 0.5938916994566236,
+      "grad_norm": 0.25301989912986755,
+      "learning_rate": 0.0002112162162162162,
+      "loss": 0.4468,
+      "step": 1981
+    },
+    {
+      "epoch": 0.594191493348323,
+      "grad_norm": 0.2434222400188446,
+      "learning_rate": 0.00021117117117117114,
+      "loss": 0.4348,
+      "step": 1982
+    },
+    {
+      "epoch": 0.5944912872400225,
+      "grad_norm": 0.24123191833496094,
+      "learning_rate": 0.0002111261261261261,
+      "loss": 0.4485,
+      "step": 1983
+    },
+    {
+      "epoch": 0.594791081131722,
+      "grad_norm": 0.2643167972564697,
+      "learning_rate": 0.00021108108108108107,
+      "loss": 0.4464,
+      "step": 1984
+    },
+    {
+      "epoch": 0.5950908750234214,
+      "grad_norm": 0.2503741979598999,
+      "learning_rate": 0.00021103603603603603,
+      "loss": 0.4247,
+      "step": 1985
+    },
+    {
+      "epoch": 0.5953906689151208,
+      "grad_norm": 0.23599058389663696,
+      "learning_rate": 0.00021099099099099097,
+      "loss": 0.4089,
+      "step": 1986
+    },
+    {
+      "epoch": 0.5956904628068204,
+      "grad_norm": 0.25657573342323303,
+      "learning_rate": 0.00021094594594594593,
+      "loss": 0.445,
+      "step": 1987
+    },
+    {
+      "epoch": 0.5959902566985198,
+      "grad_norm": 0.26530593633651733,
+      "learning_rate": 0.0002109009009009009,
+      "loss": 0.4386,
+      "step": 1988
+    },
+    {
+      "epoch": 0.5962900505902192,
+      "grad_norm": 0.2542692720890045,
+      "learning_rate": 0.00021085585585585583,
+      "loss": 0.4431,
+      "step": 1989
+    },
+    {
+      "epoch": 0.5965898444819187,
+      "grad_norm": 0.2595093548297882,
+      "learning_rate": 0.0002108108108108108,
+      "loss": 0.4243,
+      "step": 1990
+    },
+    {
+      "epoch": 0.5968896383736182,
+      "grad_norm": 0.2481933832168579,
+      "learning_rate": 0.00021076576576576576,
+      "loss": 0.4215,
+      "step": 1991
+    },
+    {
+      "epoch": 0.5971894322653176,
+      "grad_norm": 0.2607382535934448,
+      "learning_rate": 0.0002107207207207207,
+      "loss": 0.4505,
+      "step": 1992
+    },
+    {
+      "epoch": 0.597489226157017,
+      "grad_norm": 0.26596030592918396,
+      "learning_rate": 0.00021067567567567566,
+      "loss": 0.4703,
+      "step": 1993
+    },
+    {
+      "epoch": 0.5977890200487165,
+      "grad_norm": 0.24884359538555145,
+      "learning_rate": 0.00021063063063063062,
+      "loss": 0.4237,
+      "step": 1994
+    },
+    {
+      "epoch": 0.5980888139404159,
+      "grad_norm": 0.23510870337486267,
+      "learning_rate": 0.00021058558558558556,
+      "loss": 0.4235,
+      "step": 1995
+    },
+    {
+      "epoch": 0.5983886078321154,
+      "grad_norm": 0.25982344150543213,
+      "learning_rate": 0.00021054054054054052,
+      "loss": 0.4262,
+      "step": 1996
+    },
+    {
+      "epoch": 0.5986884017238149,
+      "grad_norm": 0.24918124079704285,
+      "learning_rate": 0.0002104954954954955,
+      "loss": 0.4225,
+      "step": 1997
+    },
+    {
+      "epoch": 0.5989881956155143,
+      "grad_norm": 0.26578548550605774,
+      "learning_rate": 0.00021045045045045042,
+      "loss": 0.4685,
+      "step": 1998
+    },
+    {
+      "epoch": 0.5992879895072137,
+      "grad_norm": 0.2682873010635376,
+      "learning_rate": 0.00021040540540540542,
+      "loss": 0.4291,
+      "step": 1999
+    },
+    {
+      "epoch": 0.5995877833989133,
+      "grad_norm": 0.247017040848732,
+      "learning_rate": 0.00021036036036036035,
+      "loss": 0.4457,
+      "step": 2000
+    },
+    {
+      "epoch": 0.5995877833989133,
+      "eval_loss": 0.4406769275665283,
+      "eval_runtime": 564.9558,
+      "eval_samples_per_second": 3.822,
+      "eval_steps_per_second": 0.478,
+      "step": 2000
+    },
+    {
+      "epoch": 0.5998875772906127,
+      "grad_norm": 0.25887760519981384,
+      "learning_rate": 0.0002103153153153153,
+      "loss": 0.4261,
+      "step": 2001
+    },
+    {
+      "epoch": 0.6001873711823121,
+      "grad_norm": 0.26813459396362305,
+      "learning_rate": 0.00021027027027027023,
+      "loss": 0.4091,
+      "step": 2002
+    },
+    {
+      "epoch": 0.6004871650740116,
+      "grad_norm": 0.2717922329902649,
+      "learning_rate": 0.00021022522522522522,
+      "loss": 0.4257,
+      "step": 2003
+    },
+    {
+      "epoch": 0.6007869589657111,
+      "grad_norm": 0.2423432320356369,
+      "learning_rate": 0.00021018018018018015,
+      "loss": 0.4187,
+      "step": 2004
+    },
+    {
+      "epoch": 0.6010867528574105,
+      "grad_norm": 0.2616721987724304,
+      "learning_rate": 0.0002101351351351351,
+      "loss": 0.4414,
+      "step": 2005
+    },
+    {
+      "epoch": 0.60138654674911,
+      "grad_norm": 0.2668519914150238,
+      "learning_rate": 0.00021009009009009008,
+      "loss": 0.4566,
+      "step": 2006
+    },
+    {
+      "epoch": 0.6016863406408094,
+      "grad_norm": 0.24378737807273865,
+      "learning_rate": 0.00021004504504504502,
+      "loss": 0.4379,
+      "step": 2007
+    },
+    {
+      "epoch": 0.6019861345325089,
+      "grad_norm": 0.24571847915649414,
+      "learning_rate": 0.00020999999999999998,
+      "loss": 0.4491,
+      "step": 2008
+    },
+    {
+      "epoch": 0.6022859284242084,
+      "grad_norm": 0.23369182646274567,
+      "learning_rate": 0.00020995495495495495,
+      "loss": 0.4023,
+      "step": 2009
+    },
+    {
+      "epoch": 0.6025857223159078,
+      "grad_norm": 0.27274230122566223,
+      "learning_rate": 0.00020990990990990988,
+      "loss": 0.4514,
+      "step": 2010
+    },
+    {
+      "epoch": 0.6028855162076072,
+      "grad_norm": 0.29650700092315674,
+      "learning_rate": 0.00020986486486486485,
+      "loss": 0.4553,
+      "step": 2011
+    },
+    {
+      "epoch": 0.6031853100993068,
+      "grad_norm": 0.2543351650238037,
+      "learning_rate": 0.0002098198198198198,
+      "loss": 0.4522,
+      "step": 2012
+    },
+    {
+      "epoch": 0.6034851039910062,
+      "grad_norm": 0.25208958983421326,
+      "learning_rate": 0.00020977477477477475,
+      "loss": 0.4342,
+      "step": 2013
+    },
+    {
+      "epoch": 0.6037848978827056,
+      "grad_norm": 0.24142181873321533,
+      "learning_rate": 0.0002097297297297297,
+      "loss": 0.4106,
+      "step": 2014
+    },
+    {
+      "epoch": 0.604084691774405,
+      "grad_norm": 0.2537544071674347,
+      "learning_rate": 0.00020968468468468467,
+      "loss": 0.4373,
+      "step": 2015
+    },
+    {
+      "epoch": 0.6043844856661046,
+      "grad_norm": 0.27682605385780334,
+      "learning_rate": 0.0002096396396396396,
+      "loss": 0.4491,
+      "step": 2016
+    },
+    {
+      "epoch": 0.604684279557804,
+      "grad_norm": 0.24604292213916779,
+      "learning_rate": 0.00020959459459459458,
+      "loss": 0.4059,
+      "step": 2017
+    },
+    {
+      "epoch": 0.6049840734495034,
+      "grad_norm": 0.24552318453788757,
+      "learning_rate": 0.00020954954954954954,
+      "loss": 0.4361,
+      "step": 2018
+    },
+    {
+      "epoch": 0.6052838673412029,
+      "grad_norm": 0.24892392754554749,
+      "learning_rate": 0.00020950450450450448,
+      "loss": 0.4579,
+      "step": 2019
+    },
+    {
+      "epoch": 0.6055836612329024,
+      "grad_norm": 0.26011762022972107,
+      "learning_rate": 0.00020945945945945944,
+      "loss": 0.4444,
+      "step": 2020
+    },
+    {
+      "epoch": 0.6058834551246018,
+      "grad_norm": 0.25486037135124207,
+      "learning_rate": 0.0002094144144144144,
+      "loss": 0.4256,
+      "step": 2021
+    },
+    {
+      "epoch": 0.6061832490163013,
+      "grad_norm": 0.24915416538715363,
+      "learning_rate": 0.00020936936936936937,
+      "loss": 0.4295,
+      "step": 2022
+    },
+    {
+      "epoch": 0.6064830429080007,
+      "grad_norm": 0.24448558688163757,
+      "learning_rate": 0.0002093243243243243,
+      "loss": 0.422,
+      "step": 2023
+    },
+    {
+      "epoch": 0.6067828367997002,
+      "grad_norm": 0.24826547503471375,
+      "learning_rate": 0.00020927927927927927,
+      "loss": 0.3927,
+      "step": 2024
+    },
+    {
+      "epoch": 0.6070826306913997,
+      "grad_norm": 0.24415460228919983,
+      "learning_rate": 0.00020923423423423423,
+      "loss": 0.4128,
+      "step": 2025
+    },
+    {
+      "epoch": 0.6073824245830991,
+      "grad_norm": 0.25376150012016296,
+      "learning_rate": 0.00020918918918918917,
+      "loss": 0.4363,
+      "step": 2026
+    },
+    {
+      "epoch": 0.6076822184747985,
+      "grad_norm": 0.2526264786720276,
+      "learning_rate": 0.0002091441441441441,
+      "loss": 0.4354,
+      "step": 2027
+    },
+    {
+      "epoch": 0.6079820123664981,
+      "grad_norm": 0.25586017966270447,
+      "learning_rate": 0.0002090990990990991,
+      "loss": 0.401,
+      "step": 2028
+    },
+    {
+      "epoch": 0.6082818062581975,
+      "grad_norm": 0.2642875909805298,
+      "learning_rate": 0.00020905405405405403,
+      "loss": 0.4476,
+      "step": 2029
+    },
+    {
+      "epoch": 0.6085816001498969,
+      "grad_norm": 0.2795594334602356,
+      "learning_rate": 0.00020900900900900897,
+      "loss": 0.4845,
+      "step": 2030
+    },
+    {
+      "epoch": 0.6088813940415964,
+      "grad_norm": 0.23246948421001434,
+      "learning_rate": 0.00020896396396396396,
+      "loss": 0.3906,
+      "step": 2031
+    },
+    {
+      "epoch": 0.6091811879332959,
+      "grad_norm": 0.25595536828041077,
+      "learning_rate": 0.0002089189189189189,
+      "loss": 0.4331,
+      "step": 2032
+    },
+    {
+      "epoch": 0.6094809818249953,
+      "grad_norm": 0.24884217977523804,
+      "learning_rate": 0.00020887387387387383,
+      "loss": 0.4361,
+      "step": 2033
+    },
+    {
+      "epoch": 0.6097807757166948,
+      "grad_norm": 0.25283509492874146,
+      "learning_rate": 0.00020882882882882883,
+      "loss": 0.422,
+      "step": 2034
+    },
+    {
+      "epoch": 0.6100805696083942,
+      "grad_norm": 0.25681042671203613,
+      "learning_rate": 0.00020878378378378376,
+      "loss": 0.4301,
+      "step": 2035
+    },
+    {
+      "epoch": 0.6103803635000937,
+      "grad_norm": 0.2695556879043579,
+      "learning_rate": 0.0002087387387387387,
+      "loss": 0.4624,
+      "step": 2036
+    },
+    {
+      "epoch": 0.6106801573917932,
+      "grad_norm": 0.28260117769241333,
+      "learning_rate": 0.0002086936936936937,
+      "loss": 0.4608,
+      "step": 2037
+    },
+    {
+      "epoch": 0.6109799512834926,
+      "grad_norm": 0.2578640878200531,
+      "learning_rate": 0.00020864864864864863,
+      "loss": 0.4638,
+      "step": 2038
+    },
+    {
+      "epoch": 0.611279745175192,
+      "grad_norm": 0.2544034421443939,
+      "learning_rate": 0.00020860360360360356,
+      "loss": 0.4222,
+      "step": 2039
+    },
+    {
+      "epoch": 0.6115795390668916,
+      "grad_norm": 0.2726188600063324,
+      "learning_rate": 0.00020855855855855855,
+      "loss": 0.4802,
+      "step": 2040
+    },
+    {
+      "epoch": 0.611879332958591,
+      "grad_norm": 0.26486557722091675,
+      "learning_rate": 0.0002085135135135135,
+      "loss": 0.4475,
+      "step": 2041
+    },
+    {
+      "epoch": 0.6121791268502904,
+      "grad_norm": 0.2517718970775604,
+      "learning_rate": 0.00020846846846846843,
+      "loss": 0.4445,
+      "step": 2042
+    },
+    {
+      "epoch": 0.6124789207419898,
+      "grad_norm": 0.2507137060165405,
+      "learning_rate": 0.00020842342342342342,
+      "loss": 0.4243,
+      "step": 2043
+    },
+    {
+      "epoch": 0.6127787146336894,
+      "grad_norm": 0.2707221806049347,
+      "learning_rate": 0.00020837837837837836,
+      "loss": 0.4706,
+      "step": 2044
+    },
+    {
+      "epoch": 0.6130785085253888,
+      "grad_norm": 0.24258100986480713,
+      "learning_rate": 0.00020833333333333332,
+      "loss": 0.423,
+      "step": 2045
+    },
+    {
+      "epoch": 0.6133783024170882,
+      "grad_norm": 0.25611022114753723,
+      "learning_rate": 0.00020828828828828828,
+      "loss": 0.4553,
+      "step": 2046
+    },
+    {
+      "epoch": 0.6136780963087877,
+      "grad_norm": 0.26037782430648804,
+      "learning_rate": 0.00020824324324324322,
+      "loss": 0.4173,
+      "step": 2047
+    },
+    {
+      "epoch": 0.6139778902004872,
+      "grad_norm": 0.26126980781555176,
+      "learning_rate": 0.00020819819819819818,
+      "loss": 0.4556,
+      "step": 2048
+    },
+    {
+      "epoch": 0.6142776840921866,
+      "grad_norm": 0.283407598733902,
+      "learning_rate": 0.00020815315315315315,
+      "loss": 0.4421,
+      "step": 2049
+    },
+    {
+      "epoch": 0.6145774779838861,
+      "grad_norm": 0.2533104717731476,
+      "learning_rate": 0.00020810810810810808,
+      "loss": 0.4253,
+      "step": 2050
+    },
+    {
+      "epoch": 0.6148772718755855,
+      "grad_norm": 0.24108561873435974,
+      "learning_rate": 0.00020806306306306305,
+      "loss": 0.438,
+      "step": 2051
+    },
+    {
+      "epoch": 0.615177065767285,
+      "grad_norm": 0.2516781687736511,
+      "learning_rate": 0.00020801801801801799,
+      "loss": 0.4245,
+      "step": 2052
+    },
+    {
+      "epoch": 0.6154768596589845,
+      "grad_norm": 0.25689882040023804,
+      "learning_rate": 0.00020797297297297295,
+      "loss": 0.4412,
+      "step": 2053
+    },
+    {
+      "epoch": 0.6157766535506839,
+      "grad_norm": 0.2661590576171875,
+      "learning_rate": 0.0002079279279279279,
+      "loss": 0.47,
+      "step": 2054
+    },
+    {
+      "epoch": 0.6160764474423833,
+      "grad_norm": 0.2580914795398712,
+      "learning_rate": 0.00020788288288288285,
+      "loss": 0.4537,
+      "step": 2055
+    },
+    {
+      "epoch": 0.6163762413340829,
+      "grad_norm": 0.23960340023040771,
+      "learning_rate": 0.00020783783783783784,
+      "loss": 0.4535,
+      "step": 2056
+    },
+    {
+      "epoch": 0.6166760352257823,
+      "grad_norm": 0.2658163905143738,
+      "learning_rate": 0.00020779279279279278,
+      "loss": 0.4483,
+      "step": 2057
+    },
+    {
+      "epoch": 0.6169758291174817,
+      "grad_norm": 0.2633172869682312,
+      "learning_rate": 0.00020774774774774771,
+      "loss": 0.4408,
+      "step": 2058
+    },
+    {
+      "epoch": 0.6172756230091812,
+      "grad_norm": 0.2618810832500458,
+      "learning_rate": 0.0002077027027027027,
+      "loss": 0.4497,
+      "step": 2059
+    },
+    {
+      "epoch": 0.6175754169008807,
+      "grad_norm": 0.2505929470062256,
+      "learning_rate": 0.00020765765765765764,
+      "loss": 0.4411,
+      "step": 2060
+    },
+    {
+      "epoch": 0.6178752107925801,
+      "grad_norm": 0.2556394338607788,
+      "learning_rate": 0.00020761261261261258,
+      "loss": 0.4427,
+      "step": 2061
+    },
+    {
+      "epoch": 0.6181750046842795,
+      "grad_norm": 0.2584443688392639,
+      "learning_rate": 0.00020756756756756757,
+      "loss": 0.4707,
+      "step": 2062
+    },
+    {
+      "epoch": 0.618474798575979,
+      "grad_norm": 0.2633087635040283,
+      "learning_rate": 0.0002075225225225225,
+      "loss": 0.4576,
+      "step": 2063
+    },
+    {
+      "epoch": 0.6187745924676785,
+      "grad_norm": 0.2487655133008957,
+      "learning_rate": 0.00020747747747747744,
+      "loss": 0.4374,
+      "step": 2064
+    },
+    {
+      "epoch": 0.619074386359378,
+      "grad_norm": 0.24193575978279114,
+      "learning_rate": 0.00020743243243243243,
+      "loss": 0.4267,
+      "step": 2065
+    },
+    {
+      "epoch": 0.6193741802510774,
+      "grad_norm": 0.2395676076412201,
+      "learning_rate": 0.00020738738738738737,
+      "loss": 0.4236,
+      "step": 2066
+    },
+    {
+      "epoch": 0.6196739741427768,
+      "grad_norm": 0.24150055646896362,
+      "learning_rate": 0.0002073423423423423,
+      "loss": 0.4138,
+      "step": 2067
+    },
+    {
+      "epoch": 0.6199737680344763,
+      "grad_norm": 0.2652229368686676,
+      "learning_rate": 0.0002072972972972973,
+      "loss": 0.4248,
+      "step": 2068
+    },
+    {
+      "epoch": 0.6202735619261758,
+      "grad_norm": 0.24750439822673798,
+      "learning_rate": 0.00020725225225225224,
+      "loss": 0.4273,
+      "step": 2069
+    },
+    {
+      "epoch": 0.6205733558178752,
+      "grad_norm": 0.25777891278266907,
+      "learning_rate": 0.00020720720720720717,
+      "loss": 0.4471,
+      "step": 2070
+    },
+    {
+      "epoch": 0.6208731497095746,
+      "grad_norm": 0.25587761402130127,
+      "learning_rate": 0.00020716216216216216,
+      "loss": 0.4246,
+      "step": 2071
+    },
+    {
+      "epoch": 0.6211729436012742,
+      "grad_norm": 0.2518714368343353,
+      "learning_rate": 0.0002071171171171171,
+      "loss": 0.4645,
+      "step": 2072
+    },
+    {
+      "epoch": 0.6214727374929736,
+      "grad_norm": 0.24348096549510956,
+      "learning_rate": 0.00020707207207207204,
+      "loss": 0.4454,
+      "step": 2073
+    },
+    {
+      "epoch": 0.621772531384673,
+      "grad_norm": 0.2423403263092041,
+      "learning_rate": 0.000207027027027027,
+      "loss": 0.405,
+      "step": 2074
+    },
+    {
+      "epoch": 0.6220723252763725,
+      "grad_norm": 0.26894888281822205,
+      "learning_rate": 0.00020698198198198196,
+      "loss": 0.4699,
+      "step": 2075
+    },
+    {
+      "epoch": 0.622372119168072,
+      "grad_norm": 0.2578108608722687,
+      "learning_rate": 0.0002069369369369369,
+      "loss": 0.4365,
+      "step": 2076
+    },
+    {
+      "epoch": 0.6226719130597714,
+      "grad_norm": 0.27144232392311096,
+      "learning_rate": 0.00020689189189189187,
+      "loss": 0.4291,
+      "step": 2077
+    },
+    {
+      "epoch": 0.6229717069514709,
+      "grad_norm": 0.25405609607696533,
+      "learning_rate": 0.00020684684684684683,
+      "loss": 0.4389,
+      "step": 2078
+    },
+    {
+      "epoch": 0.6232715008431703,
+      "grad_norm": 0.24775098264217377,
+      "learning_rate": 0.0002068018018018018,
+      "loss": 0.416,
+      "step": 2079
+    },
+    {
+      "epoch": 0.6235712947348698,
+      "grad_norm": 0.25767782330513,
+      "learning_rate": 0.00020675675675675673,
+      "loss": 0.455,
+      "step": 2080
+    },
+    {
+      "epoch": 0.6238710886265693,
+      "grad_norm": 0.2590571641921997,
+      "learning_rate": 0.0002067117117117117,
+      "loss": 0.4436,
+      "step": 2081
+    },
+    {
+      "epoch": 0.6241708825182687,
+      "grad_norm": 0.24577729403972626,
+      "learning_rate": 0.00020666666666666666,
+      "loss": 0.4389,
+      "step": 2082
+    },
+    {
+      "epoch": 0.6244706764099681,
+      "grad_norm": 0.25743165612220764,
+      "learning_rate": 0.0002066216216216216,
+      "loss": 0.4467,
+      "step": 2083
+    },
+    {
+      "epoch": 0.6247704703016677,
+      "grad_norm": 0.24127769470214844,
+      "learning_rate": 0.00020657657657657656,
+      "loss": 0.4056,
+      "step": 2084
+    },
+    {
+      "epoch": 0.6250702641933671,
+      "grad_norm": 0.25930923223495483,
+      "learning_rate": 0.00020653153153153152,
+      "loss": 0.4352,
+      "step": 2085
+    },
+    {
+      "epoch": 0.6253700580850665,
+      "grad_norm": 0.23111720383167267,
+      "learning_rate": 0.00020648648648648646,
+      "loss": 0.4035,
+      "step": 2086
+    },
+    {
+      "epoch": 0.625669851976766,
+      "grad_norm": 0.25705838203430176,
+      "learning_rate": 0.00020644144144144142,
+      "loss": 0.4402,
+      "step": 2087
+    },
+    {
+      "epoch": 0.6259696458684655,
+      "grad_norm": 0.260232150554657,
+      "learning_rate": 0.0002063963963963964,
+      "loss": 0.444,
+      "step": 2088
+    },
+    {
+      "epoch": 0.6262694397601649,
+      "grad_norm": 0.24525994062423706,
+      "learning_rate": 0.00020635135135135132,
+      "loss": 0.4461,
+      "step": 2089
+    },
+    {
+      "epoch": 0.6265692336518643,
+      "grad_norm": 0.2550069987773895,
+      "learning_rate": 0.00020630630630630631,
+      "loss": 0.4391,
+      "step": 2090
+    },
+    {
+      "epoch": 0.6268690275435638,
+      "grad_norm": 0.24561873078346252,
+      "learning_rate": 0.00020626126126126125,
+      "loss": 0.424,
+      "step": 2091
+    },
+    {
+      "epoch": 0.6271688214352632,
+      "grad_norm": 0.2500348687171936,
+      "learning_rate": 0.0002062162162162162,
+      "loss": 0.4415,
+      "step": 2092
+    },
+    {
+      "epoch": 0.6274686153269627,
+      "grad_norm": 0.23684681951999664,
+      "learning_rate": 0.00020617117117117118,
+      "loss": 0.4282,
+      "step": 2093
+    },
+    {
+      "epoch": 0.6277684092186622,
+      "grad_norm": 0.2459501475095749,
+      "learning_rate": 0.00020612612612612612,
+      "loss": 0.4229,
+      "step": 2094
+    },
+    {
+      "epoch": 0.6280682031103616,
+      "grad_norm": 0.23220065236091614,
+      "learning_rate": 0.00020608108108108105,
+      "loss": 0.4083,
+      "step": 2095
+    },
+    {
+      "epoch": 0.628367997002061,
+      "grad_norm": 0.26996707916259766,
+      "learning_rate": 0.00020603603603603604,
+      "loss": 0.4581,
+      "step": 2096
+    },
+    {
+      "epoch": 0.6286677908937606,
+      "grad_norm": 0.2551998496055603,
+      "learning_rate": 0.00020599099099099098,
+      "loss": 0.4444,
+      "step": 2097
+    },
+    {
+      "epoch": 0.62896758478546,
+      "grad_norm": 0.2604631781578064,
+      "learning_rate": 0.00020594594594594592,
+      "loss": 0.4367,
+      "step": 2098
+    },
+    {
+      "epoch": 0.6292673786771594,
+      "grad_norm": 0.25958266854286194,
+      "learning_rate": 0.00020590090090090085,
+      "loss": 0.4618,
+      "step": 2099
+    },
+    {
+      "epoch": 0.6295671725688589,
+      "grad_norm": 0.2714870274066925,
+      "learning_rate": 0.00020585585585585584,
+      "loss": 0.4382,
+      "step": 2100
+    },
+    {
+      "epoch": 0.6298669664605584,
+      "grad_norm": 0.2739773392677307,
+      "learning_rate": 0.00020581081081081078,
+      "loss": 0.4752,
+      "step": 2101
+    },
+    {
+      "epoch": 0.6301667603522578,
+      "grad_norm": 0.24829277396202087,
+      "learning_rate": 0.00020576576576576575,
+      "loss": 0.4285,
+      "step": 2102
+    },
+    {
+      "epoch": 0.6304665542439573,
+      "grad_norm": 0.2551855742931366,
+      "learning_rate": 0.0002057207207207207,
+      "loss": 0.4259,
+      "step": 2103
+    },
+    {
+      "epoch": 0.6307663481356567,
+      "grad_norm": 0.243735671043396,
+      "learning_rate": 0.00020567567567567565,
+      "loss": 0.4201,
+      "step": 2104
+    },
+    {
+      "epoch": 0.6310661420273562,
+      "grad_norm": 0.2511364817619324,
+      "learning_rate": 0.0002056306306306306,
+      "loss": 0.4351,
+      "step": 2105
+    },
+    {
+      "epoch": 0.6313659359190557,
+      "grad_norm": 0.2456447184085846,
+      "learning_rate": 0.00020558558558558557,
+      "loss": 0.448,
+      "step": 2106
+    },
+    {
+      "epoch": 0.6316657298107551,
+      "grad_norm": 0.26450565457344055,
+      "learning_rate": 0.0002055405405405405,
+      "loss": 0.4576,
+      "step": 2107
+    },
+    {
+      "epoch": 0.6319655237024545,
+      "grad_norm": 0.25267186760902405,
+      "learning_rate": 0.00020549549549549547,
+      "loss": 0.4511,
+      "step": 2108
+    },
+    {
+      "epoch": 0.632265317594154,
+      "grad_norm": 0.2436206340789795,
+      "learning_rate": 0.00020545045045045044,
+      "loss": 0.4147,
+      "step": 2109
+    },
+    {
+      "epoch": 0.6325651114858535,
+      "grad_norm": 0.27077367901802063,
+      "learning_rate": 0.00020540540540540537,
+      "loss": 0.4687,
+      "step": 2110
+    },
+    {
+      "epoch": 0.6328649053775529,
+      "grad_norm": 0.25476735830307007,
+      "learning_rate": 0.00020536036036036034,
+      "loss": 0.4357,
+      "step": 2111
+    },
+    {
+      "epoch": 0.6331646992692523,
+      "grad_norm": 0.23889677226543427,
+      "learning_rate": 0.0002053153153153153,
+      "loss": 0.4109,
+      "step": 2112
+    },
+    {
+      "epoch": 0.6334644931609519,
+      "grad_norm": 0.2620011270046234,
+      "learning_rate": 0.00020527027027027027,
+      "loss": 0.4418,
+      "step": 2113
+    },
+    {
+      "epoch": 0.6337642870526513,
+      "grad_norm": 0.24259789288043976,
+      "learning_rate": 0.0002052252252252252,
+      "loss": 0.4158,
+      "step": 2114
+    },
+    {
+      "epoch": 0.6340640809443507,
+      "grad_norm": 0.26212331652641296,
+      "learning_rate": 0.00020518018018018017,
+      "loss": 0.4407,
+      "step": 2115
+    },
+    {
+      "epoch": 0.6343638748360502,
+      "grad_norm": 0.2421627789735794,
+      "learning_rate": 0.00020513513513513513,
+      "loss": 0.4162,
+      "step": 2116
+    },
+    {
+      "epoch": 0.6346636687277497,
+      "grad_norm": 0.25949686765670776,
+      "learning_rate": 0.00020509009009009007,
+      "loss": 0.4436,
+      "step": 2117
+    },
+    {
+      "epoch": 0.6349634626194491,
+      "grad_norm": 0.26797404885292053,
+      "learning_rate": 0.00020504504504504503,
+      "loss": 0.4684,
+      "step": 2118
+    },
+    {
+      "epoch": 0.6352632565111486,
+      "grad_norm": 0.2433563470840454,
+      "learning_rate": 0.000205,
+      "loss": 0.4312,
+      "step": 2119
+    },
+    {
+      "epoch": 0.635563050402848,
+      "grad_norm": 0.25377362966537476,
+      "learning_rate": 0.00020495495495495493,
+      "loss": 0.4583,
+      "step": 2120
+    },
+    {
+      "epoch": 0.6358628442945475,
+      "grad_norm": 0.2583523392677307,
+      "learning_rate": 0.0002049099099099099,
+      "loss": 0.4587,
+      "step": 2121
+    },
+    {
+      "epoch": 0.636162638186247,
+      "grad_norm": 0.2512962520122528,
+      "learning_rate": 0.00020486486486486486,
+      "loss": 0.4148,
+      "step": 2122
+    },
+    {
+      "epoch": 0.6364624320779464,
+      "grad_norm": 0.24238212406635284,
+      "learning_rate": 0.0002048198198198198,
+      "loss": 0.4057,
+      "step": 2123
+    },
+    {
+      "epoch": 0.6367622259696458,
+      "grad_norm": 0.25228351354599,
+      "learning_rate": 0.00020477477477477473,
+      "loss": 0.4291,
+      "step": 2124
+    },
+    {
+      "epoch": 0.6370620198613454,
+      "grad_norm": 0.26437970995903015,
+      "learning_rate": 0.00020472972972972972,
+      "loss": 0.4425,
+      "step": 2125
+    },
+    {
+      "epoch": 0.6373618137530448,
+      "grad_norm": 0.2502360939979553,
+      "learning_rate": 0.00020468468468468466,
+      "loss": 0.4102,
+      "step": 2126
+    },
+    {
+      "epoch": 0.6376616076447442,
+      "grad_norm": 0.24895761907100677,
+      "learning_rate": 0.0002046396396396396,
+      "loss": 0.4423,
+      "step": 2127
+    },
+    {
+      "epoch": 0.6379614015364437,
+      "grad_norm": 0.2674795985221863,
+      "learning_rate": 0.0002045945945945946,
+      "loss": 0.4572,
+      "step": 2128
+    },
+    {
+      "epoch": 0.6382611954281432,
+      "grad_norm": 0.2468012422323227,
+      "learning_rate": 0.00020454954954954953,
+      "loss": 0.4525,
+      "step": 2129
+    },
+    {
+      "epoch": 0.6385609893198426,
+      "grad_norm": 0.2560267448425293,
+      "learning_rate": 0.00020450450450450446,
+      "loss": 0.4553,
+      "step": 2130
+    },
+    {
+      "epoch": 0.638860783211542,
+      "grad_norm": 0.25216519832611084,
+      "learning_rate": 0.00020445945945945945,
+      "loss": 0.4319,
+      "step": 2131
+    },
+    {
+      "epoch": 0.6391605771032415,
+      "grad_norm": 0.22927191853523254,
+      "learning_rate": 0.0002044144144144144,
+      "loss": 0.4088,
+      "step": 2132
+    },
+    {
+      "epoch": 0.639460370994941,
+      "grad_norm": 0.2495734840631485,
+      "learning_rate": 0.00020436936936936933,
+      "loss": 0.4301,
+      "step": 2133
+    },
+    {
+      "epoch": 0.6397601648866404,
+      "grad_norm": 0.2604648470878601,
+      "learning_rate": 0.00020432432432432432,
+      "loss": 0.4115,
+      "step": 2134
+    },
+    {
+      "epoch": 0.6400599587783399,
+      "grad_norm": 0.27185025811195374,
+      "learning_rate": 0.00020427927927927925,
+      "loss": 0.4452,
+      "step": 2135
+    },
+    {
+      "epoch": 0.6403597526700393,
+      "grad_norm": 0.26344043016433716,
+      "learning_rate": 0.00020423423423423422,
+      "loss": 0.4178,
+      "step": 2136
+    },
+    {
+      "epoch": 0.6406595465617388,
+      "grad_norm": 0.26041179895401,
+      "learning_rate": 0.00020418918918918918,
+      "loss": 0.4384,
+      "step": 2137
+    },
+    {
+      "epoch": 0.6409593404534383,
+      "grad_norm": 0.25143757462501526,
+      "learning_rate": 0.00020414414414414412,
+      "loss": 0.4486,
+      "step": 2138
+    },
+    {
+      "epoch": 0.6412591343451377,
+      "grad_norm": 0.2869662642478943,
+      "learning_rate": 0.00020409909909909908,
+      "loss": 0.487,
+      "step": 2139
+    },
+    {
+      "epoch": 0.6415589282368371,
+      "grad_norm": 0.2788044214248657,
+      "learning_rate": 0.00020405405405405405,
+      "loss": 0.4255,
+      "step": 2140
+    },
+    {
+      "epoch": 0.6418587221285367,
+      "grad_norm": 0.27385109663009644,
+      "learning_rate": 0.00020400900900900898,
+      "loss": 0.4535,
+      "step": 2141
+    },
+    {
+      "epoch": 0.6421585160202361,
+      "grad_norm": 0.26231497526168823,
+      "learning_rate": 0.00020396396396396395,
+      "loss": 0.4435,
+      "step": 2142
+    },
+    {
+      "epoch": 0.6424583099119355,
+      "grad_norm": 0.26582983136177063,
+      "learning_rate": 0.0002039189189189189,
+      "loss": 0.4394,
+      "step": 2143
+    },
+    {
+      "epoch": 0.642758103803635,
+      "grad_norm": 0.2665957510471344,
+      "learning_rate": 0.00020387387387387385,
+      "loss": 0.4401,
+      "step": 2144
+    },
+    {
+      "epoch": 0.6430578976953345,
+      "grad_norm": 0.2581312358379364,
+      "learning_rate": 0.0002038288288288288,
+      "loss": 0.4302,
+      "step": 2145
+    },
+    {
+      "epoch": 0.6433576915870339,
+      "grad_norm": 0.25063759088516235,
+      "learning_rate": 0.00020378378378378375,
+      "loss": 0.4558,
+      "step": 2146
+    },
+    {
+      "epoch": 0.6436574854787334,
+      "grad_norm": 0.2655949890613556,
+      "learning_rate": 0.00020373873873873874,
+      "loss": 0.4559,
+      "step": 2147
+    },
+    {
+      "epoch": 0.6439572793704328,
+      "grad_norm": 0.25797712802886963,
+      "learning_rate": 0.00020369369369369368,
+      "loss": 0.4127,
+      "step": 2148
+    },
+    {
+      "epoch": 0.6442570732621323,
+      "grad_norm": 0.2774755358695984,
+      "learning_rate": 0.00020364864864864861,
+      "loss": 0.4807,
+      "step": 2149
+    },
+    {
+      "epoch": 0.6445568671538318,
+      "grad_norm": 0.27283889055252075,
+      "learning_rate": 0.0002036036036036036,
+      "loss": 0.4483,
+      "step": 2150
+    },
+    {
+      "epoch": 0.6448566610455312,
+      "grad_norm": 0.2647114396095276,
+      "learning_rate": 0.00020355855855855854,
+      "loss": 0.4629,
+      "step": 2151
+    },
+    {
+      "epoch": 0.6451564549372306,
+      "grad_norm": 0.2683508098125458,
+      "learning_rate": 0.00020351351351351348,
+      "loss": 0.4755,
+      "step": 2152
+    },
+    {
+      "epoch": 0.6454562488289302,
+      "grad_norm": 0.26375094056129456,
+      "learning_rate": 0.00020346846846846847,
+      "loss": 0.4354,
+      "step": 2153
+    },
+    {
+      "epoch": 0.6457560427206296,
+      "grad_norm": 0.2688734233379364,
+      "learning_rate": 0.0002034234234234234,
+      "loss": 0.4515,
+      "step": 2154
+    },
+    {
+      "epoch": 0.646055836612329,
+      "grad_norm": 0.2545178532600403,
+      "learning_rate": 0.00020337837837837834,
+      "loss": 0.4281,
+      "step": 2155
+    },
+    {
+      "epoch": 0.6463556305040284,
+      "grad_norm": 0.24753254652023315,
+      "learning_rate": 0.00020333333333333333,
+      "loss": 0.4332,
+      "step": 2156
+    },
+    {
+      "epoch": 0.646655424395728,
+      "grad_norm": 0.2619148790836334,
+      "learning_rate": 0.00020328828828828827,
+      "loss": 0.4531,
+      "step": 2157
+    },
+    {
+      "epoch": 0.6469552182874274,
+      "grad_norm": 0.2698518633842468,
+      "learning_rate": 0.0002032432432432432,
+      "loss": 0.4807,
+      "step": 2158
+    },
+    {
+      "epoch": 0.6472550121791268,
+      "grad_norm": 0.2625516355037689,
+      "learning_rate": 0.0002031981981981982,
+      "loss": 0.4368,
+      "step": 2159
+    },
+    {
+      "epoch": 0.6475548060708263,
+      "grad_norm": 0.25750574469566345,
+      "learning_rate": 0.00020315315315315313,
+      "loss": 0.4489,
+      "step": 2160
+    },
+    {
+      "epoch": 0.6478545999625258,
+      "grad_norm": 0.2963887155056,
+      "learning_rate": 0.00020310810810810807,
+      "loss": 0.4567,
+      "step": 2161
+    },
+    {
+      "epoch": 0.6481543938542252,
+      "grad_norm": 0.28631022572517395,
+      "learning_rate": 0.00020306306306306306,
+      "loss": 0.457,
+      "step": 2162
+    },
+    {
+      "epoch": 0.6484541877459247,
+      "grad_norm": 0.2652076184749603,
+      "learning_rate": 0.000203018018018018,
+      "loss": 0.4008,
+      "step": 2163
+    },
+    {
+      "epoch": 0.6487539816376241,
+      "grad_norm": 0.25599247217178345,
+      "learning_rate": 0.00020297297297297294,
+      "loss": 0.4236,
+      "step": 2164
+    },
+    {
+      "epoch": 0.6490537755293236,
+      "grad_norm": 0.2655317783355713,
+      "learning_rate": 0.00020292792792792793,
+      "loss": 0.4401,
+      "step": 2165
+    },
+    {
+      "epoch": 0.6493535694210231,
+      "grad_norm": 0.27608954906463623,
+      "learning_rate": 0.00020288288288288286,
+      "loss": 0.4469,
+      "step": 2166
+    },
+    {
+      "epoch": 0.6496533633127225,
+      "grad_norm": 0.2523987293243408,
+      "learning_rate": 0.0002028378378378378,
+      "loss": 0.4134,
+      "step": 2167
+    },
+    {
+      "epoch": 0.6499531572044219,
+      "grad_norm": 0.26536789536476135,
+      "learning_rate": 0.0002027927927927928,
+      "loss": 0.4416,
+      "step": 2168
+    },
+    {
+      "epoch": 0.6502529510961215,
+      "grad_norm": 0.24977469444274902,
+      "learning_rate": 0.00020274774774774773,
+      "loss": 0.4318,
+      "step": 2169
+    },
+    {
+      "epoch": 0.6505527449878209,
+      "grad_norm": 0.27510321140289307,
+      "learning_rate": 0.0002027027027027027,
+      "loss": 0.4535,
+      "step": 2170
+    },
+    {
+      "epoch": 0.6508525388795203,
+      "grad_norm": 0.24680355191230774,
+      "learning_rate": 0.00020265765765765763,
+      "loss": 0.4247,
+      "step": 2171
+    },
+    {
+      "epoch": 0.6511523327712198,
+      "grad_norm": 0.24580539762973785,
+      "learning_rate": 0.0002026126126126126,
+      "loss": 0.4356,
+      "step": 2172
+    },
+    {
+      "epoch": 0.6514521266629193,
+      "grad_norm": 0.2560003101825714,
+      "learning_rate": 0.00020256756756756756,
+      "loss": 0.4019,
+      "step": 2173
+    },
+    {
+      "epoch": 0.6517519205546187,
+      "grad_norm": 0.2403692603111267,
+      "learning_rate": 0.0002025225225225225,
+      "loss": 0.4088,
+      "step": 2174
+    },
+    {
+      "epoch": 0.6520517144463182,
+      "grad_norm": 0.25952261686325073,
+      "learning_rate": 0.00020247747747747746,
+      "loss": 0.4585,
+      "step": 2175
+    },
+    {
+      "epoch": 0.6523515083380176,
+      "grad_norm": 0.2858710289001465,
+      "learning_rate": 0.00020243243243243242,
+      "loss": 0.4616,
+      "step": 2176
+    },
+    {
+      "epoch": 0.6526513022297171,
+      "grad_norm": 0.2761286795139313,
+      "learning_rate": 0.00020238738738738736,
+      "loss": 0.4694,
+      "step": 2177
+    },
+    {
+      "epoch": 0.6529510961214166,
+      "grad_norm": 0.2620023190975189,
+      "learning_rate": 0.00020234234234234232,
+      "loss": 0.4441,
+      "step": 2178
+    },
+    {
+      "epoch": 0.653250890013116,
+      "grad_norm": 0.2743069529533386,
+      "learning_rate": 0.00020229729729729729,
+      "loss": 0.4433,
+      "step": 2179
+    },
+    {
+      "epoch": 0.6535506839048154,
+      "grad_norm": 0.25842079520225525,
+      "learning_rate": 0.00020225225225225222,
+      "loss": 0.475,
+      "step": 2180
+    },
+    {
+      "epoch": 0.653850477796515,
+      "grad_norm": 0.267971932888031,
+      "learning_rate": 0.0002022072072072072,
+      "loss": 0.4384,
+      "step": 2181
+    },
+    {
+      "epoch": 0.6541502716882144,
+      "grad_norm": 0.26229405403137207,
+      "learning_rate": 0.00020216216216216215,
+      "loss": 0.4525,
+      "step": 2182
+    },
+    {
+      "epoch": 0.6544500655799138,
+      "grad_norm": 0.26667454838752747,
+      "learning_rate": 0.0002021171171171171,
+      "loss": 0.4384,
+      "step": 2183
+    },
+    {
+      "epoch": 0.6547498594716132,
+      "grad_norm": 0.2867937982082367,
+      "learning_rate": 0.00020207207207207208,
+      "loss": 0.4575,
+      "step": 2184
+    },
+    {
+      "epoch": 0.6550496533633127,
+      "grad_norm": 0.2772350013256073,
+      "learning_rate": 0.00020202702702702701,
+      "loss": 0.4515,
+      "step": 2185
+    },
+    {
+      "epoch": 0.6553494472550122,
+      "grad_norm": 0.2599480450153351,
+      "learning_rate": 0.00020198198198198195,
+      "loss": 0.4614,
+      "step": 2186
+    },
+    {
+      "epoch": 0.6556492411467116,
+      "grad_norm": 0.25755733251571655,
+      "learning_rate": 0.00020193693693693694,
+      "loss": 0.4543,
+      "step": 2187
+    },
+    {
+      "epoch": 0.6559490350384111,
+      "grad_norm": 0.2630176246166229,
+      "learning_rate": 0.00020189189189189188,
+      "loss": 0.4317,
+      "step": 2188
+    },
+    {
+      "epoch": 0.6562488289301105,
+      "grad_norm": 0.25234025716781616,
+      "learning_rate": 0.00020184684684684682,
+      "loss": 0.4187,
+      "step": 2189
+    },
+    {
+      "epoch": 0.65654862282181,
+      "grad_norm": 0.2525855302810669,
+      "learning_rate": 0.0002018018018018018,
+      "loss": 0.4364,
+      "step": 2190
+    },
+    {
+      "epoch": 0.6568484167135095,
+      "grad_norm": 0.2502008378505707,
+      "learning_rate": 0.00020175675675675674,
+      "loss": 0.4162,
+      "step": 2191
+    },
+    {
+      "epoch": 0.6571482106052089,
+      "grad_norm": 0.24894630908966064,
+      "learning_rate": 0.00020171171171171168,
+      "loss": 0.4287,
+      "step": 2192
+    },
+    {
+      "epoch": 0.6574480044969083,
+      "grad_norm": 0.24618011713027954,
+      "learning_rate": 0.00020166666666666667,
+      "loss": 0.4225,
+      "step": 2193
+    },
+    {
+      "epoch": 0.6577477983886079,
+      "grad_norm": 0.3233836591243744,
+      "learning_rate": 0.0002016216216216216,
+      "loss": 0.5051,
+      "step": 2194
+    },
+    {
+      "epoch": 0.6580475922803073,
+      "grad_norm": 0.2600880265235901,
+      "learning_rate": 0.00020157657657657655,
+      "loss": 0.4437,
+      "step": 2195
+    },
+    {
+      "epoch": 0.6583473861720067,
+      "grad_norm": 0.2576145529747009,
+      "learning_rate": 0.0002015315315315315,
+      "loss": 0.4035,
+      "step": 2196
+    },
+    {
+      "epoch": 0.6586471800637062,
+      "grad_norm": 0.2555258572101593,
+      "learning_rate": 0.00020148648648648647,
+      "loss": 0.409,
+      "step": 2197
+    },
+    {
+      "epoch": 0.6589469739554057,
+      "grad_norm": 0.2691539525985718,
+      "learning_rate": 0.0002014414414414414,
+      "loss": 0.4281,
+      "step": 2198
+    },
+    {
+      "epoch": 0.6592467678471051,
+      "grad_norm": 0.2896745502948761,
+      "learning_rate": 0.00020139639639639637,
+      "loss": 0.4529,
+      "step": 2199
+    },
+    {
+      "epoch": 0.6595465617388045,
+      "grad_norm": 0.2579841911792755,
+      "learning_rate": 0.00020135135135135134,
+      "loss": 0.4479,
+      "step": 2200
+    },
+    {
+      "epoch": 0.659846355630504,
+      "grad_norm": 0.24643990397453308,
+      "learning_rate": 0.00020130630630630627,
+      "loss": 0.4205,
+      "step": 2201
+    },
+    {
+      "epoch": 0.6601461495222035,
+      "grad_norm": 0.2523176968097687,
+      "learning_rate": 0.00020126126126126124,
+      "loss": 0.428,
+      "step": 2202
+    },
+    {
+      "epoch": 0.660445943413903,
+      "grad_norm": 0.27141350507736206,
+      "learning_rate": 0.0002012162162162162,
+      "loss": 0.4345,
+      "step": 2203
+    },
+    {
+      "epoch": 0.6607457373056024,
+      "grad_norm": 0.2594592571258545,
+      "learning_rate": 0.00020117117117117117,
+      "loss": 0.428,
+      "step": 2204
+    },
+    {
+      "epoch": 0.6610455311973018,
+      "grad_norm": 0.2673409581184387,
+      "learning_rate": 0.0002011261261261261,
+      "loss": 0.4229,
+      "step": 2205
+    },
+    {
+      "epoch": 0.6613453250890013,
+      "grad_norm": 0.26660290360450745,
+      "learning_rate": 0.00020108108108108107,
+      "loss": 0.4635,
+      "step": 2206
+    },
+    {
+      "epoch": 0.6616451189807008,
+      "grad_norm": 0.2599898874759674,
+      "learning_rate": 0.00020103603603603603,
+      "loss": 0.4398,
+      "step": 2207
+    },
+    {
+      "epoch": 0.6619449128724002,
+      "grad_norm": 0.2786708474159241,
+      "learning_rate": 0.00020099099099099097,
+      "loss": 0.4221,
+      "step": 2208
+    },
+    {
+      "epoch": 0.6622447067640996,
+      "grad_norm": 0.2445315271615982,
+      "learning_rate": 0.00020094594594594593,
+      "loss": 0.4143,
+      "step": 2209
+    },
+    {
+      "epoch": 0.6625445006557992,
+      "grad_norm": 0.2615225315093994,
+      "learning_rate": 0.0002009009009009009,
+      "loss": 0.4388,
+      "step": 2210
+    },
+    {
+      "epoch": 0.6628442945474986,
+      "grad_norm": 0.25724494457244873,
+      "learning_rate": 0.00020085585585585583,
+      "loss": 0.4359,
+      "step": 2211
+    },
+    {
+      "epoch": 0.663144088439198,
+      "grad_norm": 0.2562429904937744,
+      "learning_rate": 0.0002008108108108108,
+      "loss": 0.4213,
+      "step": 2212
+    },
+    {
+      "epoch": 0.6634438823308975,
+      "grad_norm": 0.25428321957588196,
+      "learning_rate": 0.00020076576576576576,
+      "loss": 0.4398,
+      "step": 2213
+    },
+    {
+      "epoch": 0.663743676222597,
+      "grad_norm": 0.2503153383731842,
+      "learning_rate": 0.0002007207207207207,
+      "loss": 0.4266,
+      "step": 2214
+    },
+    {
+      "epoch": 0.6640434701142964,
+      "grad_norm": 0.26181626319885254,
+      "learning_rate": 0.00020067567567567566,
+      "loss": 0.4398,
+      "step": 2215
+    },
+    {
+      "epoch": 0.6643432640059959,
+      "grad_norm": 0.26600703597068787,
+      "learning_rate": 0.00020063063063063062,
+      "loss": 0.4418,
+      "step": 2216
+    },
+    {
+      "epoch": 0.6646430578976953,
+      "grad_norm": 0.27476897835731506,
+      "learning_rate": 0.00020058558558558556,
+      "loss": 0.4405,
+      "step": 2217
+    },
+    {
+      "epoch": 0.6649428517893948,
+      "grad_norm": 0.23994912207126617,
+      "learning_rate": 0.0002005405405405405,
+      "loss": 0.4171,
+      "step": 2218
+    },
+    {
+      "epoch": 0.6652426456810943,
+      "grad_norm": 0.2647401988506317,
+      "learning_rate": 0.0002004954954954955,
+      "loss": 0.4514,
+      "step": 2219
+    },
+    {
+      "epoch": 0.6655424395727937,
+      "grad_norm": 0.27040109038352966,
+      "learning_rate": 0.00020045045045045043,
+      "loss": 0.4665,
+      "step": 2220
+    },
+    {
+      "epoch": 0.6658422334644931,
+      "grad_norm": 0.252128005027771,
+      "learning_rate": 0.00020040540540540536,
+      "loss": 0.4316,
+      "step": 2221
+    },
+    {
+      "epoch": 0.6661420273561927,
+      "grad_norm": 0.2605067193508148,
+      "learning_rate": 0.00020036036036036035,
+      "loss": 0.4242,
+      "step": 2222
+    },
+    {
+      "epoch": 0.6664418212478921,
+      "grad_norm": 0.26456427574157715,
+      "learning_rate": 0.0002003153153153153,
+      "loss": 0.4379,
+      "step": 2223
+    },
+    {
+      "epoch": 0.6667416151395915,
+      "grad_norm": 0.24853096902370453,
+      "learning_rate": 0.00020027027027027023,
+      "loss": 0.4339,
+      "step": 2224
+    },
+    {
+      "epoch": 0.6670414090312909,
+      "grad_norm": 0.2573966979980469,
+      "learning_rate": 0.00020022522522522522,
+      "loss": 0.4427,
+      "step": 2225
+    },
+    {
+      "epoch": 0.6673412029229905,
+      "grad_norm": 0.2402806282043457,
+      "learning_rate": 0.00020018018018018015,
+      "loss": 0.4195,
+      "step": 2226
+    },
+    {
+      "epoch": 0.6676409968146899,
+      "grad_norm": 0.2738892138004303,
+      "learning_rate": 0.00020013513513513512,
+      "loss": 0.4374,
+      "step": 2227
+    },
+    {
+      "epoch": 0.6679407907063893,
+      "grad_norm": 0.2548046410083771,
+      "learning_rate": 0.00020009009009009008,
+      "loss": 0.4312,
+      "step": 2228
+    },
+    {
+      "epoch": 0.6682405845980888,
+      "grad_norm": 0.24718345701694489,
+      "learning_rate": 0.00020004504504504502,
+      "loss": 0.4533,
+      "step": 2229
+    },
+    {
+      "epoch": 0.6685403784897883,
+      "grad_norm": 0.2544947862625122,
+      "learning_rate": 0.00019999999999999998,
+      "loss": 0.4255,
+      "step": 2230
+    },
+    {
+      "epoch": 0.6688401723814877,
+      "grad_norm": 0.2689761817455292,
+      "learning_rate": 0.00019995495495495495,
+      "loss": 0.451,
+      "step": 2231
+    },
+    {
+      "epoch": 0.6691399662731872,
+      "grad_norm": 0.2824248969554901,
+      "learning_rate": 0.00019990990990990988,
+      "loss": 0.4374,
+      "step": 2232
+    },
+    {
+      "epoch": 0.6694397601648866,
+      "grad_norm": 0.25299206376075745,
+      "learning_rate": 0.00019986486486486485,
+      "loss": 0.4301,
+      "step": 2233
+    },
+    {
+      "epoch": 0.6697395540565861,
+      "grad_norm": 0.26550740003585815,
+      "learning_rate": 0.0001998198198198198,
+      "loss": 0.4332,
+      "step": 2234
+    },
+    {
+      "epoch": 0.6700393479482856,
+      "grad_norm": 0.2446885108947754,
+      "learning_rate": 0.00019977477477477475,
+      "loss": 0.4207,
+      "step": 2235
+    },
+    {
+      "epoch": 0.670339141839985,
+      "grad_norm": 0.23779337108135223,
+      "learning_rate": 0.0001997297297297297,
+      "loss": 0.432,
+      "step": 2236
+    },
+    {
+      "epoch": 0.6706389357316844,
+      "grad_norm": 0.2619169354438782,
+      "learning_rate": 0.00019968468468468468,
+      "loss": 0.4332,
+      "step": 2237
+    },
+    {
+      "epoch": 0.670938729623384,
+      "grad_norm": 0.258789598941803,
+      "learning_rate": 0.00019963963963963964,
+      "loss": 0.4328,
+      "step": 2238
+    },
+    {
+      "epoch": 0.6712385235150834,
+      "grad_norm": 0.27282267808914185,
+      "learning_rate": 0.00019959459459459458,
+      "loss": 0.4423,
+      "step": 2239
+    },
+    {
+      "epoch": 0.6715383174067828,
+      "grad_norm": 0.2593368887901306,
+      "learning_rate": 0.00019954954954954954,
+      "loss": 0.4473,
+      "step": 2240
+    },
+    {
+      "epoch": 0.6718381112984823,
+      "grad_norm": 0.25508931279182434,
+      "learning_rate": 0.0001995045045045045,
+      "loss": 0.4371,
+      "step": 2241
+    },
+    {
+      "epoch": 0.6721379051901818,
+      "grad_norm": 0.25891220569610596,
+      "learning_rate": 0.00019945945945945944,
+      "loss": 0.4143,
+      "step": 2242
+    },
+    {
+      "epoch": 0.6724376990818812,
+      "grad_norm": 0.2661759853363037,
+      "learning_rate": 0.00019941441441441438,
+      "loss": 0.4585,
+      "step": 2243
+    },
+    {
+      "epoch": 0.6727374929735807,
+      "grad_norm": 0.25630587339401245,
+      "learning_rate": 0.00019936936936936937,
+      "loss": 0.4442,
+      "step": 2244
+    },
+    {
+      "epoch": 0.6730372868652801,
+      "grad_norm": 0.27050697803497314,
+      "learning_rate": 0.0001993243243243243,
+      "loss": 0.4437,
+      "step": 2245
+    },
+    {
+      "epoch": 0.6733370807569796,
+      "grad_norm": 0.2564866244792938,
+      "learning_rate": 0.00019927927927927924,
+      "loss": 0.4352,
+      "step": 2246
+    },
+    {
+      "epoch": 0.673636874648679,
+      "grad_norm": 0.2438693344593048,
+      "learning_rate": 0.00019923423423423423,
+      "loss": 0.4138,
+      "step": 2247
+    },
+    {
+      "epoch": 0.6739366685403785,
+      "grad_norm": 0.26038050651550293,
+      "learning_rate": 0.00019918918918918917,
+      "loss": 0.4309,
+      "step": 2248
+    },
+    {
+      "epoch": 0.6742364624320779,
+      "grad_norm": 0.2429644614458084,
+      "learning_rate": 0.0001991441441441441,
+      "loss": 0.4036,
+      "step": 2249
+    },
+    {
+      "epoch": 0.6745362563237775,
+      "grad_norm": 0.24670763313770294,
+      "learning_rate": 0.0001990990990990991,
+      "loss": 0.4054,
+      "step": 2250
+    },
+    {
+      "epoch": 0.6748360502154769,
+      "grad_norm": 0.2641808092594147,
+      "learning_rate": 0.00019905405405405403,
+      "loss": 0.4431,
+      "step": 2251
+    },
+    {
+      "epoch": 0.6751358441071763,
+      "grad_norm": 0.2601335644721985,
+      "learning_rate": 0.00019900900900900897,
+      "loss": 0.403,
+      "step": 2252
+    },
+    {
+      "epoch": 0.6754356379988757,
+      "grad_norm": 0.2698741555213928,
+      "learning_rate": 0.00019896396396396396,
+      "loss": 0.4328,
+      "step": 2253
+    },
+    {
+      "epoch": 0.6757354318905753,
+      "grad_norm": 0.26246562600135803,
+      "learning_rate": 0.0001989189189189189,
+      "loss": 0.4264,
+      "step": 2254
+    },
+    {
+      "epoch": 0.6760352257822747,
+      "grad_norm": 0.26453712582588196,
+      "learning_rate": 0.00019887387387387384,
+      "loss": 0.4552,
+      "step": 2255
+    },
+    {
+      "epoch": 0.6763350196739741,
+      "grad_norm": 0.24887052178382874,
+      "learning_rate": 0.00019882882882882883,
+      "loss": 0.3918,
+      "step": 2256
+    },
+    {
+      "epoch": 0.6766348135656736,
+      "grad_norm": 0.26789602637290955,
+      "learning_rate": 0.00019878378378378376,
+      "loss": 0.4416,
+      "step": 2257
+    },
+    {
+      "epoch": 0.6769346074573731,
+      "grad_norm": 0.24345183372497559,
+      "learning_rate": 0.0001987387387387387,
+      "loss": 0.4355,
+      "step": 2258
+    },
+    {
+      "epoch": 0.6772344013490725,
+      "grad_norm": 0.2475103735923767,
+      "learning_rate": 0.0001986936936936937,
+      "loss": 0.4277,
+      "step": 2259
+    },
+    {
+      "epoch": 0.677534195240772,
+      "grad_norm": 0.2740587294101715,
+      "learning_rate": 0.00019864864864864863,
+      "loss": 0.4536,
+      "step": 2260
+    },
+    {
+      "epoch": 0.6778339891324714,
+      "grad_norm": 0.23657365143299103,
+      "learning_rate": 0.0001986036036036036,
+      "loss": 0.3887,
+      "step": 2261
+    },
+    {
+      "epoch": 0.6781337830241709,
+      "grad_norm": 0.267630398273468,
+      "learning_rate": 0.00019855855855855856,
+      "loss": 0.443,
+      "step": 2262
+    },
+    {
+      "epoch": 0.6784335769158704,
+      "grad_norm": 0.2708898186683655,
+      "learning_rate": 0.0001985135135135135,
+      "loss": 0.4398,
+      "step": 2263
+    },
+    {
+      "epoch": 0.6787333708075698,
+      "grad_norm": 0.26607415080070496,
+      "learning_rate": 0.00019846846846846846,
+      "loss": 0.4249,
+      "step": 2264
+    },
+    {
+      "epoch": 0.6790331646992692,
+      "grad_norm": 0.2398756742477417,
+      "learning_rate": 0.00019842342342342342,
+      "loss": 0.419,
+      "step": 2265
+    },
+    {
+      "epoch": 0.6793329585909688,
+      "grad_norm": 0.2509295344352722,
+      "learning_rate": 0.00019837837837837836,
+      "loss": 0.3858,
+      "step": 2266
+    },
+    {
+      "epoch": 0.6796327524826682,
+      "grad_norm": 0.30269870162010193,
+      "learning_rate": 0.00019833333333333332,
+      "loss": 0.4564,
+      "step": 2267
+    },
+    {
+      "epoch": 0.6799325463743676,
+      "grad_norm": 0.2576700448989868,
+      "learning_rate": 0.00019828828828828826,
+      "loss": 0.4296,
+      "step": 2268
+    },
+    {
+      "epoch": 0.680232340266067,
+      "grad_norm": 0.29139164090156555,
+      "learning_rate": 0.00019824324324324322,
+      "loss": 0.4583,
+      "step": 2269
+    },
+    {
+      "epoch": 0.6805321341577666,
+      "grad_norm": 0.2578124701976776,
+      "learning_rate": 0.00019819819819819818,
+      "loss": 0.4419,
+      "step": 2270
+    },
+    {
+      "epoch": 0.680831928049466,
+      "grad_norm": 0.2546633780002594,
+      "learning_rate": 0.00019815315315315312,
+      "loss": 0.4311,
+      "step": 2271
+    },
+    {
+      "epoch": 0.6811317219411654,
+      "grad_norm": 0.293409526348114,
+      "learning_rate": 0.00019810810810810809,
+      "loss": 0.4756,
+      "step": 2272
+    },
+    {
+      "epoch": 0.6814315158328649,
+      "grad_norm": 0.249635249376297,
+      "learning_rate": 0.00019806306306306305,
+      "loss": 0.4434,
+      "step": 2273
+    },
+    {
+      "epoch": 0.6817313097245644,
+      "grad_norm": 0.2729664146900177,
+      "learning_rate": 0.00019801801801801799,
+      "loss": 0.4721,
+      "step": 2274
+    },
+    {
+      "epoch": 0.6820311036162638,
+      "grad_norm": 0.24961845576763153,
+      "learning_rate": 0.00019797297297297298,
+      "loss": 0.4124,
+      "step": 2275
+    },
+    {
+      "epoch": 0.6823308975079633,
+      "grad_norm": 0.26508617401123047,
+      "learning_rate": 0.00019792792792792791,
+      "loss": 0.4311,
+      "step": 2276
+    },
+    {
+      "epoch": 0.6826306913996627,
+      "grad_norm": 0.24888217449188232,
+      "learning_rate": 0.00019788288288288285,
+      "loss": 0.4334,
+      "step": 2277
+    },
+    {
+      "epoch": 0.6829304852913622,
+      "grad_norm": 0.2550651431083679,
+      "learning_rate": 0.00019783783783783784,
+      "loss": 0.4289,
+      "step": 2278
+    },
+    {
+      "epoch": 0.6832302791830617,
+      "grad_norm": 0.25816190242767334,
+      "learning_rate": 0.00019779279279279278,
+      "loss": 0.4425,
+      "step": 2279
+    },
+    {
+      "epoch": 0.6835300730747611,
+      "grad_norm": 0.25145018100738525,
+      "learning_rate": 0.00019774774774774772,
+      "loss": 0.4123,
+      "step": 2280
+    },
+    {
+      "epoch": 0.6838298669664605,
+      "grad_norm": 0.24678850173950195,
+      "learning_rate": 0.0001977027027027027,
+      "loss": 0.4309,
+      "step": 2281
+    },
+    {
+      "epoch": 0.68412966085816,
+      "grad_norm": 0.2629925012588501,
+      "learning_rate": 0.00019765765765765764,
+      "loss": 0.4184,
+      "step": 2282
+    },
+    {
+      "epoch": 0.6844294547498595,
+      "grad_norm": 0.2568414807319641,
+      "learning_rate": 0.00019761261261261258,
+      "loss": 0.4164,
+      "step": 2283
+    },
+    {
+      "epoch": 0.6847292486415589,
+      "grad_norm": 0.25906744599342346,
+      "learning_rate": 0.00019756756756756757,
+      "loss": 0.4547,
+      "step": 2284
+    },
+    {
+      "epoch": 0.6850290425332584,
+      "grad_norm": 0.2697434723377228,
+      "learning_rate": 0.0001975225225225225,
+      "loss": 0.4444,
+      "step": 2285
+    },
+    {
+      "epoch": 0.6853288364249578,
+      "grad_norm": 0.2573794424533844,
+      "learning_rate": 0.00019747747747747744,
+      "loss": 0.4309,
+      "step": 2286
+    },
+    {
+      "epoch": 0.6856286303166573,
+      "grad_norm": 0.2532881796360016,
+      "learning_rate": 0.00019743243243243244,
+      "loss": 0.4172,
+      "step": 2287
+    },
+    {
+      "epoch": 0.6859284242083568,
+      "grad_norm": 0.253292977809906,
+      "learning_rate": 0.00019738738738738737,
+      "loss": 0.4305,
+      "step": 2288
+    },
+    {
+      "epoch": 0.6862282181000562,
+      "grad_norm": 0.246769517660141,
+      "learning_rate": 0.0001973423423423423,
+      "loss": 0.4106,
+      "step": 2289
+    },
+    {
+      "epoch": 0.6865280119917556,
+      "grad_norm": 0.2593647539615631,
+      "learning_rate": 0.0001972972972972973,
+      "loss": 0.4537,
+      "step": 2290
+    },
+    {
+      "epoch": 0.6868278058834552,
+      "grad_norm": 0.25611796975135803,
+      "learning_rate": 0.00019725225225225224,
+      "loss": 0.43,
+      "step": 2291
+    },
+    {
+      "epoch": 0.6871275997751546,
+      "grad_norm": 0.25119319558143616,
+      "learning_rate": 0.00019720720720720717,
+      "loss": 0.4142,
+      "step": 2292
+    },
+    {
+      "epoch": 0.687427393666854,
+      "grad_norm": 0.250675231218338,
+      "learning_rate": 0.00019716216216216214,
+      "loss": 0.4134,
+      "step": 2293
+    },
+    {
+      "epoch": 0.6877271875585534,
+      "grad_norm": 0.2680164873600006,
+      "learning_rate": 0.0001971171171171171,
+      "loss": 0.4201,
+      "step": 2294
+    },
+    {
+      "epoch": 0.688026981450253,
+      "grad_norm": 0.26599201560020447,
+      "learning_rate": 0.00019707207207207206,
+      "loss": 0.45,
+      "step": 2295
+    },
+    {
+      "epoch": 0.6883267753419524,
+      "grad_norm": 0.24248278141021729,
+      "learning_rate": 0.000197027027027027,
+      "loss": 0.4129,
+      "step": 2296
+    },
+    {
+      "epoch": 0.6886265692336518,
+      "grad_norm": 0.25668129324913025,
+      "learning_rate": 0.00019698198198198197,
+      "loss": 0.4354,
+      "step": 2297
+    },
+    {
+      "epoch": 0.6889263631253513,
+      "grad_norm": 0.26304370164871216,
+      "learning_rate": 0.00019693693693693693,
+      "loss": 0.4423,
+      "step": 2298
+    },
+    {
+      "epoch": 0.6892261570170508,
+      "grad_norm": 0.2509578466415405,
+      "learning_rate": 0.00019689189189189187,
+      "loss": 0.4263,
+      "step": 2299
+    },
+    {
+      "epoch": 0.6895259509087502,
+      "grad_norm": 0.2629247009754181,
+      "learning_rate": 0.00019684684684684683,
+      "loss": 0.4323,
+      "step": 2300
+    },
+    {
+      "epoch": 0.6898257448004497,
+      "grad_norm": 0.24706493318080902,
+      "learning_rate": 0.0001968018018018018,
+      "loss": 0.3913,
+      "step": 2301
+    },
+    {
+      "epoch": 0.6901255386921491,
+      "grad_norm": 0.29551559686660767,
+      "learning_rate": 0.00019675675675675673,
+      "loss": 0.419,
+      "step": 2302
+    },
+    {
+      "epoch": 0.6904253325838486,
+      "grad_norm": 0.2612929046154022,
+      "learning_rate": 0.0001967117117117117,
+      "loss": 0.4223,
+      "step": 2303
+    },
+    {
+      "epoch": 0.6907251264755481,
+      "grad_norm": 0.28399109840393066,
+      "learning_rate": 0.00019666666666666666,
+      "loss": 0.4715,
+      "step": 2304
+    },
+    {
+      "epoch": 0.6910249203672475,
+      "grad_norm": 0.24555319547653198,
+      "learning_rate": 0.0001966216216216216,
+      "loss": 0.4559,
+      "step": 2305
+    },
+    {
+      "epoch": 0.6913247142589469,
+      "grad_norm": 0.2576359808444977,
+      "learning_rate": 0.00019657657657657656,
+      "loss": 0.4314,
+      "step": 2306
+    },
+    {
+      "epoch": 0.6916245081506465,
+      "grad_norm": 0.25595325231552124,
+      "learning_rate": 0.00019653153153153152,
+      "loss": 0.4286,
+      "step": 2307
+    },
+    {
+      "epoch": 0.6919243020423459,
+      "grad_norm": 0.23903168737888336,
+      "learning_rate": 0.00019648648648648646,
+      "loss": 0.4259,
+      "step": 2308
+    },
+    {
+      "epoch": 0.6922240959340453,
+      "grad_norm": 0.2797984480857849,
+      "learning_rate": 0.00019644144144144145,
+      "loss": 0.4621,
+      "step": 2309
+    },
+    {
+      "epoch": 0.6925238898257448,
+      "grad_norm": 0.25375935435295105,
+      "learning_rate": 0.0001963963963963964,
+      "loss": 0.4356,
+      "step": 2310
+    },
+    {
+      "epoch": 0.6928236837174443,
+      "grad_norm": 0.2765314280986786,
+      "learning_rate": 0.00019635135135135132,
+      "loss": 0.4474,
+      "step": 2311
+    },
+    {
+      "epoch": 0.6931234776091437,
+      "grad_norm": 0.23902222514152527,
+      "learning_rate": 0.00019630630630630632,
+      "loss": 0.4176,
+      "step": 2312
+    },
+    {
+      "epoch": 0.6934232715008432,
+      "grad_norm": 0.278622031211853,
+      "learning_rate": 0.00019626126126126125,
+      "loss": 0.4177,
+      "step": 2313
+    },
+    {
+      "epoch": 0.6937230653925426,
+      "grad_norm": 0.25161993503570557,
+      "learning_rate": 0.0001962162162162162,
+      "loss": 0.4235,
+      "step": 2314
+    },
+    {
+      "epoch": 0.6940228592842421,
+      "grad_norm": 0.28174108266830444,
+      "learning_rate": 0.00019617117117117113,
+      "loss": 0.4469,
+      "step": 2315
+    },
+    {
+      "epoch": 0.6943226531759416,
+      "grad_norm": 0.24297156929969788,
+      "learning_rate": 0.00019612612612612612,
+      "loss": 0.4348,
+      "step": 2316
+    },
+    {
+      "epoch": 0.694622447067641,
+      "grad_norm": 0.2582569122314453,
+      "learning_rate": 0.00019608108108108105,
+      "loss": 0.4373,
+      "step": 2317
+    },
+    {
+      "epoch": 0.6949222409593404,
+      "grad_norm": 0.2808705270290375,
+      "learning_rate": 0.00019603603603603602,
+      "loss": 0.4574,
+      "step": 2318
+    },
+    {
+      "epoch": 0.69522203485104,
+      "grad_norm": 0.27071914076805115,
+      "learning_rate": 0.00019599099099099098,
+      "loss": 0.4449,
+      "step": 2319
+    },
+    {
+      "epoch": 0.6955218287427394,
+      "grad_norm": 0.27735450863838196,
+      "learning_rate": 0.00019594594594594592,
+      "loss": 0.4154,
+      "step": 2320
+    },
+    {
+      "epoch": 0.6958216226344388,
+      "grad_norm": 0.25535905361175537,
+      "learning_rate": 0.00019590090090090088,
+      "loss": 0.4432,
+      "step": 2321
+    },
+    {
+      "epoch": 0.6961214165261382,
+      "grad_norm": 0.24208863079547882,
+      "learning_rate": 0.00019585585585585585,
+      "loss": 0.4248,
+      "step": 2322
+    },
+    {
+      "epoch": 0.6964212104178378,
+      "grad_norm": 0.26040393114089966,
+      "learning_rate": 0.00019581081081081078,
+      "loss": 0.4215,
+      "step": 2323
+    },
+    {
+      "epoch": 0.6967210043095372,
+      "grad_norm": 0.24389687180519104,
+      "learning_rate": 0.00019576576576576575,
+      "loss": 0.422,
+      "step": 2324
+    },
+    {
+      "epoch": 0.6970207982012366,
+      "grad_norm": 0.2545843720436096,
+      "learning_rate": 0.0001957207207207207,
+      "loss": 0.4384,
+      "step": 2325
+    },
+    {
+      "epoch": 0.6973205920929361,
+      "grad_norm": 0.2566373348236084,
+      "learning_rate": 0.00019567567567567565,
+      "loss": 0.4383,
+      "step": 2326
+    },
+    {
+      "epoch": 0.6976203859846356,
+      "grad_norm": 0.2538570463657379,
+      "learning_rate": 0.0001956306306306306,
+      "loss": 0.4285,
+      "step": 2327
+    },
+    {
+      "epoch": 0.697920179876335,
+      "grad_norm": 0.25821006298065186,
+      "learning_rate": 0.00019558558558558557,
+      "loss": 0.4304,
+      "step": 2328
+    },
+    {
+      "epoch": 0.6982199737680345,
+      "grad_norm": 0.26139143109321594,
+      "learning_rate": 0.0001955405405405405,
+      "loss": 0.4416,
+      "step": 2329
+    },
+    {
+      "epoch": 0.6985197676597339,
+      "grad_norm": 0.2557656168937683,
+      "learning_rate": 0.00019549549549549548,
+      "loss": 0.4166,
+      "step": 2330
+    },
+    {
+      "epoch": 0.6988195615514334,
+      "grad_norm": 0.2611480951309204,
+      "learning_rate": 0.00019545045045045044,
+      "loss": 0.4463,
+      "step": 2331
+    },
+    {
+      "epoch": 0.6991193554431329,
+      "grad_norm": 0.24384301900863647,
+      "learning_rate": 0.0001954054054054054,
+      "loss": 0.4027,
+      "step": 2332
+    },
+    {
+      "epoch": 0.6994191493348323,
+      "grad_norm": 0.2693532407283783,
+      "learning_rate": 0.00019536036036036034,
+      "loss": 0.4449,
+      "step": 2333
+    },
+    {
+      "epoch": 0.6997189432265317,
+      "grad_norm": 0.2669787108898163,
+      "learning_rate": 0.0001953153153153153,
+      "loss": 0.4266,
+      "step": 2334
+    },
+    {
+      "epoch": 0.7000187371182313,
+      "grad_norm": 0.23384937644004822,
+      "learning_rate": 0.00019527027027027027,
+      "loss": 0.4103,
+      "step": 2335
+    },
+    {
+      "epoch": 0.7003185310099307,
+      "grad_norm": 0.2738743722438812,
+      "learning_rate": 0.0001952252252252252,
+      "loss": 0.4664,
+      "step": 2336
+    },
+    {
+      "epoch": 0.7006183249016301,
+      "grad_norm": 0.2557884752750397,
+      "learning_rate": 0.00019518018018018017,
+      "loss": 0.4121,
+      "step": 2337
+    },
+    {
+      "epoch": 0.7009181187933295,
+      "grad_norm": 0.24830694496631622,
+      "learning_rate": 0.00019513513513513513,
+      "loss": 0.4,
+      "step": 2338
+    },
+    {
+      "epoch": 0.7012179126850291,
+      "grad_norm": 0.2636083960533142,
+      "learning_rate": 0.00019509009009009007,
+      "loss": 0.4418,
+      "step": 2339
+    },
+    {
+      "epoch": 0.7015177065767285,
+      "grad_norm": 0.252029687166214,
+      "learning_rate": 0.000195045045045045,
+      "loss": 0.4205,
+      "step": 2340
+    },
+    {
+      "epoch": 0.701817500468428,
+      "grad_norm": 0.25448256731033325,
+      "learning_rate": 0.000195,
+      "loss": 0.4517,
+      "step": 2341
+    },
+    {
+      "epoch": 0.7021172943601274,
+      "grad_norm": 0.26609039306640625,
+      "learning_rate": 0.00019495495495495493,
+      "loss": 0.4213,
+      "step": 2342
+    },
+    {
+      "epoch": 0.7024170882518269,
+      "grad_norm": 0.2746337652206421,
+      "learning_rate": 0.00019490990990990987,
+      "loss": 0.4183,
+      "step": 2343
+    },
+    {
+      "epoch": 0.7027168821435263,
+      "grad_norm": 0.2514724135398865,
+      "learning_rate": 0.00019486486486486486,
+      "loss": 0.4329,
+      "step": 2344
+    },
+    {
+      "epoch": 0.7030166760352258,
+      "grad_norm": 0.27683892846107483,
+      "learning_rate": 0.0001948198198198198,
+      "loss": 0.4207,
+      "step": 2345
+    },
+    {
+      "epoch": 0.7033164699269252,
+      "grad_norm": 0.2525181174278259,
+      "learning_rate": 0.00019477477477477473,
+      "loss": 0.4193,
+      "step": 2346
+    },
+    {
+      "epoch": 0.7036162638186247,
+      "grad_norm": 0.2759072482585907,
+      "learning_rate": 0.00019472972972972973,
+      "loss": 0.4508,
+      "step": 2347
+    },
+    {
+      "epoch": 0.7039160577103242,
+      "grad_norm": 0.2594849169254303,
+      "learning_rate": 0.00019468468468468466,
+      "loss": 0.4289,
+      "step": 2348
+    },
+    {
+      "epoch": 0.7042158516020236,
+      "grad_norm": 0.26971113681793213,
+      "learning_rate": 0.0001946396396396396,
+      "loss": 0.4514,
+      "step": 2349
+    },
+    {
+      "epoch": 0.704515645493723,
+      "grad_norm": 0.25291457772254944,
+      "learning_rate": 0.0001945945945945946,
+      "loss": 0.4664,
+      "step": 2350
+    },
+    {
+      "epoch": 0.7048154393854226,
+      "grad_norm": 0.2617851495742798,
+      "learning_rate": 0.00019454954954954953,
+      "loss": 0.4301,
+      "step": 2351
+    },
+    {
+      "epoch": 0.705115233277122,
+      "grad_norm": 0.24216975271701813,
+      "learning_rate": 0.00019450450450450446,
+      "loss": 0.4144,
+      "step": 2352
+    },
+    {
+      "epoch": 0.7054150271688214,
+      "grad_norm": 0.2737904489040375,
+      "learning_rate": 0.00019445945945945945,
+      "loss": 0.417,
+      "step": 2353
+    },
+    {
+      "epoch": 0.7057148210605209,
+      "grad_norm": 0.27587682008743286,
+      "learning_rate": 0.0001944144144144144,
+      "loss": 0.4421,
+      "step": 2354
+    },
+    {
+      "epoch": 0.7060146149522204,
+      "grad_norm": 0.24917447566986084,
+      "learning_rate": 0.00019436936936936936,
+      "loss": 0.406,
+      "step": 2355
+    },
+    {
+      "epoch": 0.7063144088439198,
+      "grad_norm": 0.27958497405052185,
+      "learning_rate": 0.00019432432432432432,
+      "loss": 0.4591,
+      "step": 2356
+    },
+    {
+      "epoch": 0.7066142027356193,
+      "grad_norm": 0.27273818850517273,
+      "learning_rate": 0.00019427927927927926,
+      "loss": 0.4219,
+      "step": 2357
+    },
+    {
+      "epoch": 0.7069139966273187,
+      "grad_norm": 0.24517607688903809,
+      "learning_rate": 0.00019423423423423422,
+      "loss": 0.4063,
+      "step": 2358
+    },
+    {
+      "epoch": 0.7072137905190182,
+      "grad_norm": 0.28854820132255554,
+      "learning_rate": 0.00019418918918918918,
+      "loss": 0.4383,
+      "step": 2359
+    },
+    {
+      "epoch": 0.7075135844107177,
+      "grad_norm": 0.27321329712867737,
+      "learning_rate": 0.00019414414414414412,
+      "loss": 0.4479,
+      "step": 2360
+    },
+    {
+      "epoch": 0.7078133783024171,
+      "grad_norm": 0.2749727666378021,
+      "learning_rate": 0.00019409909909909908,
+      "loss": 0.4272,
+      "step": 2361
+    },
+    {
+      "epoch": 0.7081131721941165,
+      "grad_norm": 0.27384045720100403,
+      "learning_rate": 0.00019405405405405405,
+      "loss": 0.444,
+      "step": 2362
+    },
+    {
+      "epoch": 0.708412966085816,
+      "grad_norm": 0.2604135274887085,
+      "learning_rate": 0.00019400900900900898,
+      "loss": 0.4248,
+      "step": 2363
+    },
+    {
+      "epoch": 0.7087127599775155,
+      "grad_norm": 0.2598932385444641,
+      "learning_rate": 0.00019396396396396395,
+      "loss": 0.4401,
+      "step": 2364
+    },
+    {
+      "epoch": 0.7090125538692149,
+      "grad_norm": 0.253755122423172,
+      "learning_rate": 0.00019391891891891889,
+      "loss": 0.4316,
+      "step": 2365
+    },
+    {
+      "epoch": 0.7093123477609143,
+      "grad_norm": 0.2677047848701477,
+      "learning_rate": 0.00019387387387387388,
+      "loss": 0.4264,
+      "step": 2366
+    },
+    {
+      "epoch": 0.7096121416526139,
+      "grad_norm": 0.24191899597644806,
+      "learning_rate": 0.0001938288288288288,
+      "loss": 0.4172,
+      "step": 2367
+    },
+    {
+      "epoch": 0.7099119355443133,
+      "grad_norm": 0.2684822082519531,
+      "learning_rate": 0.00019378378378378375,
+      "loss": 0.4466,
+      "step": 2368
+    },
+    {
+      "epoch": 0.7102117294360127,
+      "grad_norm": 0.2859225273132324,
+      "learning_rate": 0.00019373873873873874,
+      "loss": 0.4348,
+      "step": 2369
+    },
+    {
+      "epoch": 0.7105115233277122,
+      "grad_norm": 0.2676656246185303,
+      "learning_rate": 0.00019369369369369368,
+      "loss": 0.4322,
+      "step": 2370
+    },
+    {
+      "epoch": 0.7108113172194117,
+      "grad_norm": 0.27145159244537354,
+      "learning_rate": 0.00019364864864864861,
+      "loss": 0.4485,
+      "step": 2371
+    },
+    {
+      "epoch": 0.7111111111111111,
+      "grad_norm": 0.2646559178829193,
+      "learning_rate": 0.0001936036036036036,
+      "loss": 0.4255,
+      "step": 2372
+    },
+    {
+      "epoch": 0.7114109050028106,
+      "grad_norm": 0.26495832204818726,
+      "learning_rate": 0.00019355855855855854,
+      "loss": 0.4681,
+      "step": 2373
+    },
+    {
+      "epoch": 0.71171069889451,
+      "grad_norm": 0.26360276341438293,
+      "learning_rate": 0.00019351351351351348,
+      "loss": 0.4219,
+      "step": 2374
+    },
+    {
+      "epoch": 0.7120104927862094,
+      "grad_norm": 0.25372758507728577,
+      "learning_rate": 0.00019346846846846847,
+      "loss": 0.4542,
+      "step": 2375
+    },
+    {
+      "epoch": 0.712310286677909,
+      "grad_norm": 0.25859972834587097,
+      "learning_rate": 0.0001934234234234234,
+      "loss": 0.4121,
+      "step": 2376
+    },
+    {
+      "epoch": 0.7126100805696084,
+      "grad_norm": 0.2402067929506302,
+      "learning_rate": 0.00019337837837837834,
+      "loss": 0.4132,
+      "step": 2377
+    },
+    {
+      "epoch": 0.7129098744613078,
+      "grad_norm": 0.25183209776878357,
+      "learning_rate": 0.00019333333333333333,
+      "loss": 0.4106,
+      "step": 2378
+    },
+    {
+      "epoch": 0.7132096683530073,
+      "grad_norm": 0.24842600524425507,
+      "learning_rate": 0.00019328828828828827,
+      "loss": 0.3895,
+      "step": 2379
+    },
+    {
+      "epoch": 0.7135094622447068,
+      "grad_norm": 0.2635684609413147,
+      "learning_rate": 0.0001932432432432432,
+      "loss": 0.4112,
+      "step": 2380
+    },
+    {
+      "epoch": 0.7138092561364062,
+      "grad_norm": 0.2578394412994385,
+      "learning_rate": 0.0001931981981981982,
+      "loss": 0.4429,
+      "step": 2381
+    },
+    {
+      "epoch": 0.7141090500281057,
+      "grad_norm": 0.2912173867225647,
+      "learning_rate": 0.00019315315315315314,
+      "loss": 0.4256,
+      "step": 2382
+    },
+    {
+      "epoch": 0.7144088439198051,
+      "grad_norm": 0.2592369019985199,
+      "learning_rate": 0.00019310810810810807,
+      "loss": 0.4357,
+      "step": 2383
+    },
+    {
+      "epoch": 0.7147086378115046,
+      "grad_norm": 0.25853514671325684,
+      "learning_rate": 0.00019306306306306306,
+      "loss": 0.3816,
+      "step": 2384
+    },
+    {
+      "epoch": 0.715008431703204,
+      "grad_norm": 0.26601412892341614,
+      "learning_rate": 0.000193018018018018,
+      "loss": 0.4344,
+      "step": 2385
+    },
+    {
+      "epoch": 0.7153082255949035,
+      "grad_norm": 0.2789902091026306,
+      "learning_rate": 0.00019297297297297294,
+      "loss": 0.435,
+      "step": 2386
+    },
+    {
+      "epoch": 0.7156080194866029,
+      "grad_norm": 0.285659521818161,
+      "learning_rate": 0.0001929279279279279,
+      "loss": 0.4194,
+      "step": 2387
+    },
+    {
+      "epoch": 0.7159078133783024,
+      "grad_norm": 0.24971207976341248,
+      "learning_rate": 0.00019288288288288286,
+      "loss": 0.4322,
+      "step": 2388
+    },
+    {
+      "epoch": 0.7162076072700019,
+      "grad_norm": 0.30142828822135925,
+      "learning_rate": 0.00019283783783783783,
+      "loss": 0.415,
+      "step": 2389
+    },
+    {
+      "epoch": 0.7165074011617013,
+      "grad_norm": 0.2843458950519562,
+      "learning_rate": 0.00019279279279279277,
+      "loss": 0.4605,
+      "step": 2390
+    },
+    {
+      "epoch": 0.7168071950534007,
+      "grad_norm": 0.24795542657375336,
+      "learning_rate": 0.00019274774774774773,
+      "loss": 0.4147,
+      "step": 2391
+    },
+    {
+      "epoch": 0.7171069889451003,
+      "grad_norm": 0.2908228635787964,
+      "learning_rate": 0.0001927027027027027,
+      "loss": 0.4573,
+      "step": 2392
+    },
+    {
+      "epoch": 0.7174067828367997,
+      "grad_norm": 0.2770592272281647,
+      "learning_rate": 0.00019265765765765763,
+      "loss": 0.4248,
+      "step": 2393
+    },
+    {
+      "epoch": 0.7177065767284991,
+      "grad_norm": 0.2638701796531677,
+      "learning_rate": 0.0001926126126126126,
+      "loss": 0.4409,
+      "step": 2394
+    },
+    {
+      "epoch": 0.7180063706201986,
+      "grad_norm": 0.2952423393726349,
+      "learning_rate": 0.00019256756756756756,
+      "loss": 0.4651,
+      "step": 2395
+    },
+    {
+      "epoch": 0.7183061645118981,
+      "grad_norm": 0.2776016294956207,
+      "learning_rate": 0.0001925225225225225,
+      "loss": 0.4453,
+      "step": 2396
+    },
+    {
+      "epoch": 0.7186059584035975,
+      "grad_norm": 0.2560594081878662,
+      "learning_rate": 0.00019247747747747746,
+      "loss": 0.4441,
+      "step": 2397
+    },
+    {
+      "epoch": 0.718905752295297,
+      "grad_norm": 0.2397725135087967,
+      "learning_rate": 0.00019243243243243242,
+      "loss": 0.3933,
+      "step": 2398
+    },
+    {
+      "epoch": 0.7192055461869964,
+      "grad_norm": 0.2630067467689514,
+      "learning_rate": 0.00019238738738738736,
+      "loss": 0.4334,
+      "step": 2399
+    },
+    {
+      "epoch": 0.7195053400786959,
+      "grad_norm": 0.2648056149482727,
+      "learning_rate": 0.00019234234234234235,
+      "loss": 0.4456,
+      "step": 2400
+    },
+    {
+      "epoch": 0.7198051339703954,
+      "grad_norm": 0.27124911546707153,
+      "learning_rate": 0.0001922972972972973,
+      "loss": 0.4324,
+      "step": 2401
+    },
+    {
+      "epoch": 0.7201049278620948,
+      "grad_norm": 0.2809448540210724,
+      "learning_rate": 0.00019225225225225222,
+      "loss": 0.4419,
+      "step": 2402
+    },
+    {
+      "epoch": 0.7204047217537942,
+      "grad_norm": 0.2577565610408783,
+      "learning_rate": 0.00019220720720720721,
+      "loss": 0.441,
+      "step": 2403
+    },
+    {
+      "epoch": 0.7207045156454938,
+      "grad_norm": 0.261111855506897,
+      "learning_rate": 0.00019216216216216215,
+      "loss": 0.4234,
+      "step": 2404
+    },
+    {
+      "epoch": 0.7210043095371932,
+      "grad_norm": 0.2598218321800232,
+      "learning_rate": 0.0001921171171171171,
+      "loss": 0.4183,
+      "step": 2405
+    },
+    {
+      "epoch": 0.7213041034288926,
+      "grad_norm": 0.2712027132511139,
+      "learning_rate": 0.00019207207207207208,
+      "loss": 0.4354,
+      "step": 2406
+    },
+    {
+      "epoch": 0.721603897320592,
+      "grad_norm": 0.27912774682044983,
+      "learning_rate": 0.00019202702702702702,
+      "loss": 0.4594,
+      "step": 2407
+    },
+    {
+      "epoch": 0.7219036912122916,
+      "grad_norm": 0.25328657031059265,
+      "learning_rate": 0.00019198198198198195,
+      "loss": 0.408,
+      "step": 2408
+    },
+    {
+      "epoch": 0.722203485103991,
+      "grad_norm": 0.2425694465637207,
+      "learning_rate": 0.00019193693693693694,
+      "loss": 0.4197,
+      "step": 2409
+    },
+    {
+      "epoch": 0.7225032789956904,
+      "grad_norm": 0.2552039623260498,
+      "learning_rate": 0.00019189189189189188,
+      "loss": 0.4383,
+      "step": 2410
+    },
+    {
+      "epoch": 0.7228030728873899,
+      "grad_norm": 0.26128917932510376,
+      "learning_rate": 0.00019184684684684682,
+      "loss": 0.4204,
+      "step": 2411
+    },
+    {
+      "epoch": 0.7231028667790894,
+      "grad_norm": 0.2713935077190399,
+      "learning_rate": 0.00019180180180180178,
+      "loss": 0.4325,
+      "step": 2412
+    },
+    {
+      "epoch": 0.7234026606707888,
+      "grad_norm": 0.257618248462677,
+      "learning_rate": 0.00019175675675675674,
+      "loss": 0.4443,
+      "step": 2413
+    },
+    {
+      "epoch": 0.7237024545624883,
+      "grad_norm": 0.2597353458404541,
+      "learning_rate": 0.00019171171171171168,
+      "loss": 0.4415,
+      "step": 2414
+    },
+    {
+      "epoch": 0.7240022484541877,
+      "grad_norm": 0.26509061455726624,
+      "learning_rate": 0.00019166666666666665,
+      "loss": 0.4711,
+      "step": 2415
+    },
+    {
+      "epoch": 0.7243020423458872,
+      "grad_norm": 0.26354658603668213,
+      "learning_rate": 0.0001916216216216216,
+      "loss": 0.4341,
+      "step": 2416
+    },
+    {
+      "epoch": 0.7246018362375867,
+      "grad_norm": 0.2364490032196045,
+      "learning_rate": 0.00019157657657657655,
+      "loss": 0.4016,
+      "step": 2417
+    },
+    {
+      "epoch": 0.7249016301292861,
+      "grad_norm": 0.24982942640781403,
+      "learning_rate": 0.0001915315315315315,
+      "loss": 0.427,
+      "step": 2418
+    },
+    {
+      "epoch": 0.7252014240209855,
+      "grad_norm": 0.27166748046875,
+      "learning_rate": 0.00019148648648648647,
+      "loss": 0.4807,
+      "step": 2419
+    },
+    {
+      "epoch": 0.7255012179126851,
+      "grad_norm": 0.24789117276668549,
+      "learning_rate": 0.0001914414414414414,
+      "loss": 0.4265,
+      "step": 2420
+    },
+    {
+      "epoch": 0.7258010118043845,
+      "grad_norm": 0.2433491349220276,
+      "learning_rate": 0.00019139639639639637,
+      "loss": 0.4405,
+      "step": 2421
+    },
+    {
+      "epoch": 0.7261008056960839,
+      "grad_norm": 0.24121679365634918,
+      "learning_rate": 0.00019135135135135134,
+      "loss": 0.4215,
+      "step": 2422
+    },
+    {
+      "epoch": 0.7264005995877834,
+      "grad_norm": 0.25895169377326965,
+      "learning_rate": 0.0001913063063063063,
+      "loss": 0.4168,
+      "step": 2423
+    },
+    {
+      "epoch": 0.7267003934794829,
+      "grad_norm": 0.24981217086315155,
+      "learning_rate": 0.00019126126126126124,
+      "loss": 0.4268,
+      "step": 2424
+    },
+    {
+      "epoch": 0.7270001873711823,
+      "grad_norm": 0.25490307807922363,
+      "learning_rate": 0.0001912162162162162,
+      "loss": 0.4464,
+      "step": 2425
+    },
+    {
+      "epoch": 0.7272999812628818,
+      "grad_norm": 0.2552802562713623,
+      "learning_rate": 0.00019117117117117117,
+      "loss": 0.4461,
+      "step": 2426
+    },
+    {
+      "epoch": 0.7275997751545812,
+      "grad_norm": 0.27454614639282227,
+      "learning_rate": 0.0001911261261261261,
+      "loss": 0.4683,
+      "step": 2427
+    },
+    {
+      "epoch": 0.7278995690462807,
+      "grad_norm": 0.2501683831214905,
+      "learning_rate": 0.00019108108108108107,
+      "loss": 0.4244,
+      "step": 2428
+    },
+    {
+      "epoch": 0.7281993629379802,
+      "grad_norm": 0.24820026755332947,
+      "learning_rate": 0.00019103603603603603,
+      "loss": 0.4416,
+      "step": 2429
+    },
+    {
+      "epoch": 0.7284991568296796,
+      "grad_norm": 0.25755947828292847,
+      "learning_rate": 0.00019099099099099097,
+      "loss": 0.4377,
+      "step": 2430
+    },
+    {
+      "epoch": 0.728798950721379,
+      "grad_norm": 0.268839031457901,
+      "learning_rate": 0.00019094594594594593,
+      "loss": 0.46,
+      "step": 2431
+    },
+    {
+      "epoch": 0.7290987446130786,
+      "grad_norm": 0.2707115113735199,
+      "learning_rate": 0.0001909009009009009,
+      "loss": 0.4375,
+      "step": 2432
+    },
+    {
+      "epoch": 0.729398538504778,
+      "grad_norm": 0.25406280159950256,
+      "learning_rate": 0.00019085585585585583,
+      "loss": 0.441,
+      "step": 2433
+    },
+    {
+      "epoch": 0.7296983323964774,
+      "grad_norm": 0.2569238841533661,
+      "learning_rate": 0.00019081081081081082,
+      "loss": 0.4446,
+      "step": 2434
+    },
+    {
+      "epoch": 0.7299981262881768,
+      "grad_norm": 0.2784389555454254,
+      "learning_rate": 0.00019076576576576576,
+      "loss": 0.4396,
+      "step": 2435
+    },
+    {
+      "epoch": 0.7302979201798764,
+      "grad_norm": 0.25094011425971985,
+      "learning_rate": 0.0001907207207207207,
+      "loss": 0.4189,
+      "step": 2436
+    },
+    {
+      "epoch": 0.7305977140715758,
+      "grad_norm": 0.2696321904659271,
+      "learning_rate": 0.00019067567567567563,
+      "loss": 0.4575,
+      "step": 2437
+    },
+    {
+      "epoch": 0.7308975079632752,
+      "grad_norm": 0.2526487112045288,
+      "learning_rate": 0.00019063063063063062,
+      "loss": 0.4081,
+      "step": 2438
+    },
+    {
+      "epoch": 0.7311973018549747,
+      "grad_norm": 0.2577098608016968,
+      "learning_rate": 0.00019058558558558556,
+      "loss": 0.4478,
+      "step": 2439
+    },
+    {
+      "epoch": 0.7314970957466742,
+      "grad_norm": 0.26381629705429077,
+      "learning_rate": 0.0001905405405405405,
+      "loss": 0.4573,
+      "step": 2440
+    },
+    {
+      "epoch": 0.7317968896383736,
+      "grad_norm": 0.26065248250961304,
+      "learning_rate": 0.0001904954954954955,
+      "loss": 0.4413,
+      "step": 2441
+    },
+    {
+      "epoch": 0.7320966835300731,
+      "grad_norm": 0.24994462728500366,
+      "learning_rate": 0.00019045045045045043,
+      "loss": 0.4016,
+      "step": 2442
+    },
+    {
+      "epoch": 0.7323964774217725,
+      "grad_norm": 0.2742599844932556,
+      "learning_rate": 0.00019040540540540536,
+      "loss": 0.4578,
+      "step": 2443
+    },
+    {
+      "epoch": 0.732696271313472,
+      "grad_norm": 0.23844504356384277,
+      "learning_rate": 0.00019036036036036035,
+      "loss": 0.4078,
+      "step": 2444
+    },
+    {
+      "epoch": 0.7329960652051715,
+      "grad_norm": 0.2562139630317688,
+      "learning_rate": 0.0001903153153153153,
+      "loss": 0.4363,
+      "step": 2445
+    },
+    {
+      "epoch": 0.7332958590968709,
+      "grad_norm": 0.2525213956832886,
+      "learning_rate": 0.00019027027027027025,
+      "loss": 0.4585,
+      "step": 2446
+    },
+    {
+      "epoch": 0.7335956529885703,
+      "grad_norm": 0.2562119662761688,
+      "learning_rate": 0.00019022522522522522,
+      "loss": 0.4158,
+      "step": 2447
+    },
+    {
+      "epoch": 0.7338954468802699,
+      "grad_norm": 0.2771570682525635,
+      "learning_rate": 0.00019018018018018015,
+      "loss": 0.4601,
+      "step": 2448
+    },
+    {
+      "epoch": 0.7341952407719693,
+      "grad_norm": 0.2594900131225586,
+      "learning_rate": 0.00019013513513513512,
+      "loss": 0.4252,
+      "step": 2449
+    },
+    {
+      "epoch": 0.7344950346636687,
+      "grad_norm": 0.27634164690971375,
+      "learning_rate": 0.00019009009009009008,
+      "loss": 0.4614,
+      "step": 2450
+    },
+    {
+      "epoch": 0.7347948285553682,
+      "grad_norm": 0.27118000388145447,
+      "learning_rate": 0.00019004504504504502,
+      "loss": 0.425,
+      "step": 2451
+    },
+    {
+      "epoch": 0.7350946224470677,
+      "grad_norm": 0.26404282450675964,
+      "learning_rate": 0.00018999999999999998,
+      "loss": 0.435,
+      "step": 2452
+    },
+    {
+      "epoch": 0.7353944163387671,
+      "grad_norm": 0.24286137521266937,
+      "learning_rate": 0.00018995495495495495,
+      "loss": 0.4099,
+      "step": 2453
+    },
+    {
+      "epoch": 0.7356942102304665,
+      "grad_norm": 0.2554706335067749,
+      "learning_rate": 0.00018990990990990988,
+      "loss": 0.4031,
+      "step": 2454
+    },
+    {
+      "epoch": 0.735994004122166,
+      "grad_norm": 0.2666279375553131,
+      "learning_rate": 0.00018986486486486485,
+      "loss": 0.4397,
+      "step": 2455
+    },
+    {
+      "epoch": 0.7362937980138655,
+      "grad_norm": 0.24479645490646362,
+      "learning_rate": 0.0001898198198198198,
+      "loss": 0.4059,
+      "step": 2456
+    },
+    {
+      "epoch": 0.736593591905565,
+      "grad_norm": 0.27331724762916565,
+      "learning_rate": 0.00018977477477477478,
+      "loss": 0.4337,
+      "step": 2457
+    },
+    {
+      "epoch": 0.7368933857972644,
+      "grad_norm": 0.2546418309211731,
+      "learning_rate": 0.0001897297297297297,
+      "loss": 0.4206,
+      "step": 2458
+    },
+    {
+      "epoch": 0.7371931796889638,
+      "grad_norm": 0.2593313753604889,
+      "learning_rate": 0.00018968468468468468,
+      "loss": 0.39,
+      "step": 2459
+    },
+    {
+      "epoch": 0.7374929735806633,
+      "grad_norm": 0.2757156789302826,
+      "learning_rate": 0.00018963963963963964,
+      "loss": 0.441,
+      "step": 2460
+    },
+    {
+      "epoch": 0.7377927674723628,
+      "grad_norm": 0.2826617956161499,
+      "learning_rate": 0.00018959459459459458,
+      "loss": 0.4497,
+      "step": 2461
+    },
+    {
+      "epoch": 0.7380925613640622,
+      "grad_norm": 0.26498305797576904,
+      "learning_rate": 0.00018954954954954951,
+      "loss": 0.4429,
+      "step": 2462
+    },
+    {
+      "epoch": 0.7383923552557616,
+      "grad_norm": 0.22784557938575745,
+      "learning_rate": 0.0001895045045045045,
+      "loss": 0.4069,
+      "step": 2463
+    },
+    {
+      "epoch": 0.7386921491474612,
+      "grad_norm": 0.277037113904953,
+      "learning_rate": 0.00018945945945945944,
+      "loss": 0.4257,
+      "step": 2464
+    },
+    {
+      "epoch": 0.7389919430391606,
+      "grad_norm": 0.25758159160614014,
+      "learning_rate": 0.00018941441441441438,
+      "loss": 0.4255,
+      "step": 2465
+    },
+    {
+      "epoch": 0.73929173693086,
+      "grad_norm": 0.24654820561408997,
+      "learning_rate": 0.00018936936936936937,
+      "loss": 0.402,
+      "step": 2466
+    },
+    {
+      "epoch": 0.7395915308225595,
+      "grad_norm": 0.259376585483551,
+      "learning_rate": 0.0001893243243243243,
+      "loss": 0.416,
+      "step": 2467
+    },
+    {
+      "epoch": 0.739891324714259,
+      "grad_norm": 0.28223109245300293,
+      "learning_rate": 0.00018927927927927924,
+      "loss": 0.4675,
+      "step": 2468
+    },
+    {
+      "epoch": 0.7401911186059584,
+      "grad_norm": 0.2680475413799286,
+      "learning_rate": 0.00018923423423423423,
+      "loss": 0.4147,
+      "step": 2469
+    },
+    {
+      "epoch": 0.7404909124976579,
+      "grad_norm": 0.2528432309627533,
+      "learning_rate": 0.00018918918918918917,
+      "loss": 0.4374,
+      "step": 2470
+    },
+    {
+      "epoch": 0.7407907063893573,
+      "grad_norm": 0.26637372374534607,
+      "learning_rate": 0.0001891441441441441,
+      "loss": 0.4189,
+      "step": 2471
+    },
+    {
+      "epoch": 0.7410905002810567,
+      "grad_norm": 0.2570081055164337,
+      "learning_rate": 0.0001890990990990991,
+      "loss": 0.4388,
+      "step": 2472
+    },
+    {
+      "epoch": 0.7413902941727563,
+      "grad_norm": 0.27075570821762085,
+      "learning_rate": 0.00018905405405405403,
+      "loss": 0.4599,
+      "step": 2473
+    },
+    {
+      "epoch": 0.7416900880644557,
+      "grad_norm": 0.2676197290420532,
+      "learning_rate": 0.00018900900900900897,
+      "loss": 0.4105,
+      "step": 2474
+    },
+    {
+      "epoch": 0.7419898819561551,
+      "grad_norm": 0.24458040297031403,
+      "learning_rate": 0.00018896396396396396,
+      "loss": 0.4152,
+      "step": 2475
+    },
+    {
+      "epoch": 0.7422896758478545,
+      "grad_norm": 0.2793339788913727,
+      "learning_rate": 0.0001889189189189189,
+      "loss": 0.4468,
+      "step": 2476
+    },
+    {
+      "epoch": 0.7425894697395541,
+      "grad_norm": 0.25252237915992737,
+      "learning_rate": 0.00018887387387387384,
+      "loss": 0.4215,
+      "step": 2477
+    },
+    {
+      "epoch": 0.7428892636312535,
+      "grad_norm": 0.27801933884620667,
+      "learning_rate": 0.00018882882882882883,
+      "loss": 0.4362,
+      "step": 2478
+    },
+    {
+      "epoch": 0.743189057522953,
+      "grad_norm": 0.26056137681007385,
+      "learning_rate": 0.00018878378378378376,
+      "loss": 0.4342,
+      "step": 2479
+    },
+    {
+      "epoch": 0.7434888514146524,
+      "grad_norm": 0.26250821352005005,
+      "learning_rate": 0.00018873873873873873,
+      "loss": 0.4196,
+      "step": 2480
+    },
+    {
+      "epoch": 0.7437886453063519,
+      "grad_norm": 0.2682492733001709,
+      "learning_rate": 0.0001886936936936937,
+      "loss": 0.4368,
+      "step": 2481
+    },
+    {
+      "epoch": 0.7440884391980513,
+      "grad_norm": 0.2572811245918274,
+      "learning_rate": 0.00018864864864864863,
+      "loss": 0.4363,
+      "step": 2482
+    },
+    {
+      "epoch": 0.7443882330897508,
+      "grad_norm": 0.25746074318885803,
+      "learning_rate": 0.0001886036036036036,
+      "loss": 0.4045,
+      "step": 2483
+    },
+    {
+      "epoch": 0.7446880269814502,
+      "grad_norm": 0.25470736622810364,
+      "learning_rate": 0.00018855855855855853,
+      "loss": 0.4062,
+      "step": 2484
+    },
+    {
+      "epoch": 0.7449878208731497,
+      "grad_norm": 0.2766227722167969,
+      "learning_rate": 0.0001885135135135135,
+      "loss": 0.4605,
+      "step": 2485
+    },
+    {
+      "epoch": 0.7452876147648492,
+      "grad_norm": 0.28737902641296387,
+      "learning_rate": 0.00018846846846846846,
+      "loss": 0.4481,
+      "step": 2486
+    },
+    {
+      "epoch": 0.7455874086565486,
+      "grad_norm": 0.2646963894367218,
+      "learning_rate": 0.0001884234234234234,
+      "loss": 0.4212,
+      "step": 2487
+    },
+    {
+      "epoch": 0.745887202548248,
+      "grad_norm": 0.2569124698638916,
+      "learning_rate": 0.00018837837837837836,
+      "loss": 0.4268,
+      "step": 2488
+    },
+    {
+      "epoch": 0.7461869964399476,
+      "grad_norm": 0.25343701243400574,
+      "learning_rate": 0.00018833333333333332,
+      "loss": 0.4285,
+      "step": 2489
+    },
+    {
+      "epoch": 0.746486790331647,
+      "grad_norm": 0.27101901173591614,
+      "learning_rate": 0.00018828828828828826,
+      "loss": 0.4426,
+      "step": 2490
+    },
+    {
+      "epoch": 0.7467865842233464,
+      "grad_norm": 0.2594289779663086,
+      "learning_rate": 0.00018824324324324325,
+      "loss": 0.4333,
+      "step": 2491
+    },
+    {
+      "epoch": 0.7470863781150459,
+      "grad_norm": 0.2643277049064636,
+      "learning_rate": 0.00018819819819819819,
+      "loss": 0.4407,
+      "step": 2492
+    },
+    {
+      "epoch": 0.7473861720067454,
+      "grad_norm": 0.267240047454834,
+      "learning_rate": 0.00018815315315315312,
+      "loss": 0.4448,
+      "step": 2493
+    },
+    {
+      "epoch": 0.7476859658984448,
+      "grad_norm": 0.24963083863258362,
+      "learning_rate": 0.00018810810810810811,
+      "loss": 0.4313,
+      "step": 2494
+    },
+    {
+      "epoch": 0.7479857597901443,
+      "grad_norm": 0.2673603892326355,
+      "learning_rate": 0.00018806306306306305,
+      "loss": 0.4327,
+      "step": 2495
+    },
+    {
+      "epoch": 0.7482855536818437,
+      "grad_norm": 0.25436538457870483,
+      "learning_rate": 0.000188018018018018,
+      "loss": 0.4251,
+      "step": 2496
+    },
+    {
+      "epoch": 0.7485853475735432,
+      "grad_norm": 0.25511813163757324,
+      "learning_rate": 0.00018797297297297298,
+      "loss": 0.441,
+      "step": 2497
+    },
+    {
+      "epoch": 0.7488851414652427,
+      "grad_norm": 0.26634255051612854,
+      "learning_rate": 0.00018792792792792791,
+      "loss": 0.4142,
+      "step": 2498
+    },
+    {
+      "epoch": 0.7491849353569421,
+      "grad_norm": 0.2738245129585266,
+      "learning_rate": 0.00018788288288288285,
+      "loss": 0.446,
+      "step": 2499
+    },
+    {
+      "epoch": 0.7494847292486415,
+      "grad_norm": 0.2478281557559967,
+      "learning_rate": 0.00018783783783783784,
+      "loss": 0.4355,
+      "step": 2500
+    },
+    {
+      "epoch": 0.7494847292486415,
+      "eval_loss": 0.43330296874046326,
+      "eval_runtime": 567.7062,
+      "eval_samples_per_second": 3.803,
+      "eval_steps_per_second": 0.476,
+      "step": 2500
+    },
+    {
+      "epoch": 0.749784523140341,
+      "grad_norm": 0.24915559589862823,
+      "learning_rate": 0.00018779279279279278,
+      "loss": 0.431,
+      "step": 2501
+    },
+    {
+      "epoch": 0.7500843170320405,
+      "grad_norm": 0.25979626178741455,
+      "learning_rate": 0.00018774774774774772,
+      "loss": 0.4336,
+      "step": 2502
+    },
+    {
+      "epoch": 0.7503841109237399,
+      "grad_norm": 0.2514503002166748,
+      "learning_rate": 0.0001877027027027027,
+      "loss": 0.4104,
+      "step": 2503
+    },
+    {
+      "epoch": 0.7506839048154393,
+      "grad_norm": 0.2693893015384674,
+      "learning_rate": 0.00018765765765765764,
+      "loss": 0.4706,
+      "step": 2504
+    },
+    {
+      "epoch": 0.7509836987071389,
+      "grad_norm": 0.2881157696247101,
+      "learning_rate": 0.00018761261261261258,
+      "loss": 0.4346,
+      "step": 2505
+    },
+    {
+      "epoch": 0.7512834925988383,
+      "grad_norm": 0.232576385140419,
+      "learning_rate": 0.00018756756756756757,
+      "loss": 0.42,
+      "step": 2506
+    },
+    {
+      "epoch": 0.7515832864905377,
+      "grad_norm": 0.30108994245529175,
+      "learning_rate": 0.0001875225225225225,
+      "loss": 0.425,
+      "step": 2507
+    },
+    {
+      "epoch": 0.7518830803822372,
+      "grad_norm": 0.28761547803878784,
+      "learning_rate": 0.00018747747747747745,
+      "loss": 0.47,
+      "step": 2508
+    },
+    {
+      "epoch": 0.7521828742739367,
+      "grad_norm": 0.2556571662425995,
+      "learning_rate": 0.0001874324324324324,
+      "loss": 0.4236,
+      "step": 2509
+    },
+    {
+      "epoch": 0.7524826681656361,
+      "grad_norm": 0.27593177556991577,
+      "learning_rate": 0.00018738738738738737,
+      "loss": 0.4209,
+      "step": 2510
+    },
+    {
+      "epoch": 0.7527824620573356,
+      "grad_norm": 0.26506245136260986,
+      "learning_rate": 0.0001873423423423423,
+      "loss": 0.4182,
+      "step": 2511
+    },
+    {
+      "epoch": 0.753082255949035,
+      "grad_norm": 0.31767213344573975,
+      "learning_rate": 0.00018729729729729727,
+      "loss": 0.4574,
+      "step": 2512
+    },
+    {
+      "epoch": 0.7533820498407345,
+      "grad_norm": 0.25470229983329773,
+      "learning_rate": 0.00018725225225225224,
+      "loss": 0.4064,
+      "step": 2513
+    },
+    {
+      "epoch": 0.753681843732434,
+      "grad_norm": 0.25668561458587646,
+      "learning_rate": 0.0001872072072072072,
+      "loss": 0.4247,
+      "step": 2514
+    },
+    {
+      "epoch": 0.7539816376241334,
+      "grad_norm": 0.29275453090667725,
+      "learning_rate": 0.00018716216216216214,
+      "loss": 0.4232,
+      "step": 2515
+    },
+    {
+      "epoch": 0.7542814315158328,
+      "grad_norm": 0.2815520167350769,
+      "learning_rate": 0.0001871171171171171,
+      "loss": 0.4402,
+      "step": 2516
+    },
+    {
+      "epoch": 0.7545812254075324,
+      "grad_norm": 0.24637946486473083,
+      "learning_rate": 0.00018707207207207207,
+      "loss": 0.4209,
+      "step": 2517
+    },
+    {
+      "epoch": 0.7548810192992318,
+      "grad_norm": 0.27894946932792664,
+      "learning_rate": 0.000187027027027027,
+      "loss": 0.453,
+      "step": 2518
+    },
+    {
+      "epoch": 0.7551808131909312,
+      "grad_norm": 0.3131442070007324,
+      "learning_rate": 0.00018698198198198197,
+      "loss": 0.4728,
+      "step": 2519
+    },
+    {
+      "epoch": 0.7554806070826307,
+      "grad_norm": 0.2698810398578644,
+      "learning_rate": 0.00018693693693693693,
+      "loss": 0.4387,
+      "step": 2520
+    },
+    {
+      "epoch": 0.7557804009743302,
+      "grad_norm": 0.24818141758441925,
+      "learning_rate": 0.00018689189189189187,
+      "loss": 0.4065,
+      "step": 2521
+    },
+    {
+      "epoch": 0.7560801948660296,
+      "grad_norm": 0.2903098464012146,
+      "learning_rate": 0.00018684684684684683,
+      "loss": 0.426,
+      "step": 2522
+    },
+    {
+      "epoch": 0.756379988757729,
+      "grad_norm": 0.2602495551109314,
+      "learning_rate": 0.0001868018018018018,
+      "loss": 0.4144,
+      "step": 2523
+    },
+    {
+      "epoch": 0.7566797826494285,
+      "grad_norm": 0.2648625373840332,
+      "learning_rate": 0.00018675675675675673,
+      "loss": 0.4476,
+      "step": 2524
+    },
+    {
+      "epoch": 0.756979576541128,
+      "grad_norm": 0.26782098412513733,
+      "learning_rate": 0.0001867117117117117,
+      "loss": 0.4238,
+      "step": 2525
+    },
+    {
+      "epoch": 0.7572793704328274,
+      "grad_norm": 0.27317121624946594,
+      "learning_rate": 0.00018666666666666666,
+      "loss": 0.4251,
+      "step": 2526
+    },
+    {
+      "epoch": 0.7575791643245269,
+      "grad_norm": 0.2720593214035034,
+      "learning_rate": 0.0001866216216216216,
+      "loss": 0.4667,
+      "step": 2527
+    },
+    {
+      "epoch": 0.7578789582162263,
+      "grad_norm": 0.2541276812553406,
+      "learning_rate": 0.0001865765765765766,
+      "loss": 0.431,
+      "step": 2528
+    },
+    {
+      "epoch": 0.7581787521079258,
+      "grad_norm": 0.27258971333503723,
+      "learning_rate": 0.00018653153153153152,
+      "loss": 0.4263,
+      "step": 2529
+    },
+    {
+      "epoch": 0.7584785459996253,
+      "grad_norm": 0.28021714091300964,
+      "learning_rate": 0.00018648648648648646,
+      "loss": 0.4276,
+      "step": 2530
+    },
+    {
+      "epoch": 0.7587783398913247,
+      "grad_norm": 0.2505019009113312,
+      "learning_rate": 0.00018644144144144145,
+      "loss": 0.4105,
+      "step": 2531
+    },
+    {
+      "epoch": 0.7590781337830241,
+      "grad_norm": 0.28030917048454285,
+      "learning_rate": 0.0001863963963963964,
+      "loss": 0.4428,
+      "step": 2532
+    },
+    {
+      "epoch": 0.7593779276747237,
+      "grad_norm": 0.27447059750556946,
+      "learning_rate": 0.00018635135135135133,
+      "loss": 0.4551,
+      "step": 2533
+    },
+    {
+      "epoch": 0.7596777215664231,
+      "grad_norm": 0.26824313402175903,
+      "learning_rate": 0.00018630630630630626,
+      "loss": 0.4445,
+      "step": 2534
+    },
+    {
+      "epoch": 0.7599775154581225,
+      "grad_norm": 0.2674945294857025,
+      "learning_rate": 0.00018626126126126125,
+      "loss": 0.4513,
+      "step": 2535
+    },
+    {
+      "epoch": 0.760277309349822,
+      "grad_norm": 0.2604798972606659,
+      "learning_rate": 0.0001862162162162162,
+      "loss": 0.4381,
+      "step": 2536
+    },
+    {
+      "epoch": 0.7605771032415215,
+      "grad_norm": 0.27609342336654663,
+      "learning_rate": 0.00018617117117117115,
+      "loss": 0.4441,
+      "step": 2537
+    },
+    {
+      "epoch": 0.7608768971332209,
+      "grad_norm": 0.2614879012107849,
+      "learning_rate": 0.00018612612612612612,
+      "loss": 0.435,
+      "step": 2538
+    },
+    {
+      "epoch": 0.7611766910249204,
+      "grad_norm": 0.25386688113212585,
+      "learning_rate": 0.00018608108108108105,
+      "loss": 0.4434,
+      "step": 2539
+    },
+    {
+      "epoch": 0.7614764849166198,
+      "grad_norm": 0.24181143939495087,
+      "learning_rate": 0.00018603603603603602,
+      "loss": 0.4175,
+      "step": 2540
+    },
+    {
+      "epoch": 0.7617762788083193,
+      "grad_norm": 0.2645350694656372,
+      "learning_rate": 0.00018599099099099098,
+      "loss": 0.4136,
+      "step": 2541
+    },
+    {
+      "epoch": 0.7620760727000188,
+      "grad_norm": 0.2677913010120392,
+      "learning_rate": 0.00018594594594594592,
+      "loss": 0.4436,
+      "step": 2542
+    },
+    {
+      "epoch": 0.7623758665917182,
+      "grad_norm": 0.2717260420322418,
+      "learning_rate": 0.00018590090090090088,
+      "loss": 0.4565,
+      "step": 2543
+    },
+    {
+      "epoch": 0.7626756604834176,
+      "grad_norm": 0.25026705861091614,
+      "learning_rate": 0.00018585585585585585,
+      "loss": 0.4119,
+      "step": 2544
+    },
+    {
+      "epoch": 0.7629754543751172,
+      "grad_norm": 0.24770689010620117,
+      "learning_rate": 0.00018581081081081078,
+      "loss": 0.4097,
+      "step": 2545
+    },
+    {
+      "epoch": 0.7632752482668166,
+      "grad_norm": 0.27625271677970886,
+      "learning_rate": 0.00018576576576576575,
+      "loss": 0.4269,
+      "step": 2546
+    },
+    {
+      "epoch": 0.763575042158516,
+      "grad_norm": 0.27056175470352173,
+      "learning_rate": 0.0001857207207207207,
+      "loss": 0.4499,
+      "step": 2547
+    },
+    {
+      "epoch": 0.7638748360502154,
+      "grad_norm": 0.2812648415565491,
+      "learning_rate": 0.00018567567567567567,
+      "loss": 0.4736,
+      "step": 2548
+    },
+    {
+      "epoch": 0.764174629941915,
+      "grad_norm": 0.26717478036880493,
+      "learning_rate": 0.0001856306306306306,
+      "loss": 0.4072,
+      "step": 2549
+    },
+    {
+      "epoch": 0.7644744238336144,
+      "grad_norm": 0.2870055139064789,
+      "learning_rate": 0.00018558558558558558,
+      "loss": 0.4229,
+      "step": 2550
+    },
+    {
+      "epoch": 0.7647742177253138,
+      "grad_norm": 0.2580265700817108,
+      "learning_rate": 0.00018554054054054054,
+      "loss": 0.4068,
+      "step": 2551
+    },
+    {
+      "epoch": 0.7650740116170133,
+      "grad_norm": 0.28002214431762695,
+      "learning_rate": 0.00018549549549549548,
+      "loss": 0.432,
+      "step": 2552
+    },
+    {
+      "epoch": 0.7653738055087128,
+      "grad_norm": 0.27384141087532043,
+      "learning_rate": 0.00018545045045045044,
+      "loss": 0.4363,
+      "step": 2553
+    },
+    {
+      "epoch": 0.7656735994004122,
+      "grad_norm": 0.2627524137496948,
+      "learning_rate": 0.0001854054054054054,
+      "loss": 0.4197,
+      "step": 2554
+    },
+    {
+      "epoch": 0.7659733932921117,
+      "grad_norm": 0.2666347324848175,
+      "learning_rate": 0.00018536036036036034,
+      "loss": 0.3912,
+      "step": 2555
+    },
+    {
+      "epoch": 0.7662731871838111,
+      "grad_norm": 0.2756651043891907,
+      "learning_rate": 0.00018531531531531528,
+      "loss": 0.4364,
+      "step": 2556
+    },
+    {
+      "epoch": 0.7665729810755106,
+      "grad_norm": 0.2617150545120239,
+      "learning_rate": 0.00018527027027027027,
+      "loss": 0.4468,
+      "step": 2557
+    },
+    {
+      "epoch": 0.7668727749672101,
+      "grad_norm": 0.27227911353111267,
+      "learning_rate": 0.0001852252252252252,
+      "loss": 0.4246,
+      "step": 2558
+    },
+    {
+      "epoch": 0.7671725688589095,
+      "grad_norm": 0.2841823697090149,
+      "learning_rate": 0.00018518018018018014,
+      "loss": 0.4363,
+      "step": 2559
+    },
+    {
+      "epoch": 0.7674723627506089,
+      "grad_norm": 0.253366082906723,
+      "learning_rate": 0.00018513513513513513,
+      "loss": 0.419,
+      "step": 2560
+    },
+    {
+      "epoch": 0.7677721566423085,
+      "grad_norm": 0.2522357106208801,
+      "learning_rate": 0.00018509009009009007,
+      "loss": 0.4124,
+      "step": 2561
+    },
+    {
+      "epoch": 0.7680719505340079,
+      "grad_norm": 0.2550141215324402,
+      "learning_rate": 0.000185045045045045,
+      "loss": 0.4256,
+      "step": 2562
+    },
+    {
+      "epoch": 0.7683717444257073,
+      "grad_norm": 0.27578258514404297,
+      "learning_rate": 0.000185,
+      "loss": 0.4447,
+      "step": 2563
+    },
+    {
+      "epoch": 0.7686715383174068,
+      "grad_norm": 0.2517780661582947,
+      "learning_rate": 0.00018495495495495493,
+      "loss": 0.4167,
+      "step": 2564
+    },
+    {
+      "epoch": 0.7689713322091062,
+      "grad_norm": 0.2627197802066803,
+      "learning_rate": 0.00018490990990990987,
+      "loss": 0.421,
+      "step": 2565
+    },
+    {
+      "epoch": 0.7692711261008057,
+      "grad_norm": 0.2572929263114929,
+      "learning_rate": 0.00018486486486486486,
+      "loss": 0.4295,
+      "step": 2566
+    },
+    {
+      "epoch": 0.7695709199925052,
+      "grad_norm": 0.2549370229244232,
+      "learning_rate": 0.0001848198198198198,
+      "loss": 0.4353,
+      "step": 2567
+    },
+    {
+      "epoch": 0.7698707138842046,
+      "grad_norm": 0.25990357995033264,
+      "learning_rate": 0.00018477477477477474,
+      "loss": 0.4598,
+      "step": 2568
+    },
+    {
+      "epoch": 0.770170507775904,
+      "grad_norm": 0.26102861762046814,
+      "learning_rate": 0.00018472972972972973,
+      "loss": 0.4342,
+      "step": 2569
+    },
+    {
+      "epoch": 0.7704703016676036,
+      "grad_norm": 0.26292112469673157,
+      "learning_rate": 0.00018468468468468466,
+      "loss": 0.4217,
+      "step": 2570
+    },
+    {
+      "epoch": 0.770770095559303,
+      "grad_norm": 0.24879471957683563,
+      "learning_rate": 0.00018463963963963963,
+      "loss": 0.4034,
+      "step": 2571
+    },
+    {
+      "epoch": 0.7710698894510024,
+      "grad_norm": 0.249162495136261,
+      "learning_rate": 0.0001845945945945946,
+      "loss": 0.4213,
+      "step": 2572
+    },
+    {
+      "epoch": 0.7713696833427018,
+      "grad_norm": 0.25036314129829407,
+      "learning_rate": 0.00018454954954954953,
+      "loss": 0.449,
+      "step": 2573
+    },
+    {
+      "epoch": 0.7716694772344014,
+      "grad_norm": 0.2511482238769531,
+      "learning_rate": 0.0001845045045045045,
+      "loss": 0.4285,
+      "step": 2574
+    },
+    {
+      "epoch": 0.7719692711261008,
+      "grad_norm": 0.25358885526657104,
+      "learning_rate": 0.00018445945945945946,
+      "loss": 0.4392,
+      "step": 2575
+    },
+    {
+      "epoch": 0.7722690650178002,
+      "grad_norm": 0.25731760263442993,
+      "learning_rate": 0.0001844144144144144,
+      "loss": 0.4326,
+      "step": 2576
+    },
+    {
+      "epoch": 0.7725688589094997,
+      "grad_norm": 0.24017149209976196,
+      "learning_rate": 0.00018436936936936936,
+      "loss": 0.4285,
+      "step": 2577
+    },
+    {
+      "epoch": 0.7728686528011992,
+      "grad_norm": 0.24697363376617432,
+      "learning_rate": 0.00018432432432432432,
+      "loss": 0.4326,
+      "step": 2578
+    },
+    {
+      "epoch": 0.7731684466928986,
+      "grad_norm": 0.2622368335723877,
+      "learning_rate": 0.00018427927927927926,
+      "loss": 0.4051,
+      "step": 2579
+    },
+    {
+      "epoch": 0.7734682405845981,
+      "grad_norm": 0.26079848408699036,
+      "learning_rate": 0.00018423423423423422,
+      "loss": 0.4296,
+      "step": 2580
+    },
+    {
+      "epoch": 0.7737680344762975,
+      "grad_norm": 0.26790016889572144,
+      "learning_rate": 0.00018418918918918916,
+      "loss": 0.4487,
+      "step": 2581
+    },
+    {
+      "epoch": 0.774067828367997,
+      "grad_norm": 0.26801207661628723,
+      "learning_rate": 0.00018414414414414412,
+      "loss": 0.441,
+      "step": 2582
+    },
+    {
+      "epoch": 0.7743676222596965,
+      "grad_norm": 0.2615436911582947,
+      "learning_rate": 0.00018409909909909909,
+      "loss": 0.4356,
+      "step": 2583
+    },
+    {
+      "epoch": 0.7746674161513959,
+      "grad_norm": 0.26157858967781067,
+      "learning_rate": 0.00018405405405405402,
+      "loss": 0.4624,
+      "step": 2584
+    },
+    {
+      "epoch": 0.7749672100430953,
+      "grad_norm": 0.2570144832134247,
+      "learning_rate": 0.000184009009009009,
+      "loss": 0.4159,
+      "step": 2585
+    },
+    {
+      "epoch": 0.7752670039347949,
+      "grad_norm": 0.25635403394699097,
+      "learning_rate": 0.00018396396396396395,
+      "loss": 0.4479,
+      "step": 2586
+    },
+    {
+      "epoch": 0.7755667978264943,
+      "grad_norm": 0.24715913832187653,
+      "learning_rate": 0.00018391891891891889,
+      "loss": 0.4258,
+      "step": 2587
+    },
+    {
+      "epoch": 0.7758665917181937,
+      "grad_norm": 0.2577861547470093,
+      "learning_rate": 0.00018387387387387388,
+      "loss": 0.4107,
+      "step": 2588
+    },
+    {
+      "epoch": 0.7761663856098932,
+      "grad_norm": 0.24768322706222534,
+      "learning_rate": 0.00018382882882882881,
+      "loss": 0.4208,
+      "step": 2589
+    },
+    {
+      "epoch": 0.7764661795015927,
+      "grad_norm": 0.24486133456230164,
+      "learning_rate": 0.00018378378378378375,
+      "loss": 0.4128,
+      "step": 2590
+    },
+    {
+      "epoch": 0.7767659733932921,
+      "grad_norm": 0.2598220109939575,
+      "learning_rate": 0.00018373873873873874,
+      "loss": 0.4398,
+      "step": 2591
+    },
+    {
+      "epoch": 0.7770657672849915,
+      "grad_norm": 0.2616111636161804,
+      "learning_rate": 0.00018369369369369368,
+      "loss": 0.4472,
+      "step": 2592
+    },
+    {
+      "epoch": 0.777365561176691,
+      "grad_norm": 0.2481420487165451,
+      "learning_rate": 0.00018364864864864862,
+      "loss": 0.4073,
+      "step": 2593
+    },
+    {
+      "epoch": 0.7776653550683905,
+      "grad_norm": 0.26911380887031555,
+      "learning_rate": 0.0001836036036036036,
+      "loss": 0.4584,
+      "step": 2594
+    },
+    {
+      "epoch": 0.77796514896009,
+      "grad_norm": 0.2654714584350586,
+      "learning_rate": 0.00018355855855855854,
+      "loss": 0.4541,
+      "step": 2595
+    },
+    {
+      "epoch": 0.7782649428517894,
+      "grad_norm": 0.2782737612724304,
+      "learning_rate": 0.00018351351351351348,
+      "loss": 0.4496,
+      "step": 2596
+    },
+    {
+      "epoch": 0.7785647367434888,
+      "grad_norm": 0.24328523874282837,
+      "learning_rate": 0.00018346846846846847,
+      "loss": 0.4198,
+      "step": 2597
+    },
+    {
+      "epoch": 0.7788645306351883,
+      "grad_norm": 0.2627139985561371,
+      "learning_rate": 0.0001834234234234234,
+      "loss": 0.4383,
+      "step": 2598
+    },
+    {
+      "epoch": 0.7791643245268878,
+      "grad_norm": 0.25555041432380676,
+      "learning_rate": 0.00018337837837837834,
+      "loss": 0.4097,
+      "step": 2599
+    },
+    {
+      "epoch": 0.7794641184185872,
+      "grad_norm": 0.24429571628570557,
+      "learning_rate": 0.00018333333333333334,
+      "loss": 0.3932,
+      "step": 2600
+    },
+    {
+      "epoch": 0.7797639123102866,
+      "grad_norm": 0.25940433144569397,
+      "learning_rate": 0.00018328828828828827,
+      "loss": 0.4142,
+      "step": 2601
+    },
+    {
+      "epoch": 0.7800637062019862,
+      "grad_norm": 0.25724172592163086,
+      "learning_rate": 0.0001832432432432432,
+      "loss": 0.4316,
+      "step": 2602
+    },
+    {
+      "epoch": 0.7803635000936856,
+      "grad_norm": 0.2488878071308136,
+      "learning_rate": 0.0001831981981981982,
+      "loss": 0.4259,
+      "step": 2603
+    },
+    {
+      "epoch": 0.780663293985385,
+      "grad_norm": 0.25847765803337097,
+      "learning_rate": 0.00018315315315315314,
+      "loss": 0.4015,
+      "step": 2604
+    },
+    {
+      "epoch": 0.7809630878770845,
+      "grad_norm": 0.2530449628829956,
+      "learning_rate": 0.0001831081081081081,
+      "loss": 0.4015,
+      "step": 2605
+    },
+    {
+      "epoch": 0.781262881768784,
+      "grad_norm": 0.2667160630226135,
+      "learning_rate": 0.00018306306306306304,
+      "loss": 0.4333,
+      "step": 2606
+    },
+    {
+      "epoch": 0.7815626756604834,
+      "grad_norm": 0.27646440267562866,
+      "learning_rate": 0.000183018018018018,
+      "loss": 0.4585,
+      "step": 2607
+    },
+    {
+      "epoch": 0.7818624695521829,
+      "grad_norm": 0.2871904969215393,
+      "learning_rate": 0.00018297297297297297,
+      "loss": 0.4443,
+      "step": 2608
+    },
+    {
+      "epoch": 0.7821622634438823,
+      "grad_norm": 0.2631723880767822,
+      "learning_rate": 0.0001829279279279279,
+      "loss": 0.4312,
+      "step": 2609
+    },
+    {
+      "epoch": 0.7824620573355818,
+      "grad_norm": 0.27293872833251953,
+      "learning_rate": 0.00018288288288288287,
+      "loss": 0.4121,
+      "step": 2610
+    },
+    {
+      "epoch": 0.7827618512272813,
+      "grad_norm": 0.2640814781188965,
+      "learning_rate": 0.00018283783783783783,
+      "loss": 0.4253,
+      "step": 2611
+    },
+    {
+      "epoch": 0.7830616451189807,
+      "grad_norm": 0.25551357865333557,
+      "learning_rate": 0.00018279279279279277,
+      "loss": 0.4207,
+      "step": 2612
+    },
+    {
+      "epoch": 0.7833614390106801,
+      "grad_norm": 0.27317047119140625,
+      "learning_rate": 0.00018274774774774773,
+      "loss": 0.4144,
+      "step": 2613
+    },
+    {
+      "epoch": 0.7836612329023797,
+      "grad_norm": 0.2567535638809204,
+      "learning_rate": 0.0001827027027027027,
+      "loss": 0.4238,
+      "step": 2614
+    },
+    {
+      "epoch": 0.7839610267940791,
+      "grad_norm": 0.2705351710319519,
+      "learning_rate": 0.00018265765765765763,
+      "loss": 0.4431,
+      "step": 2615
+    },
+    {
+      "epoch": 0.7842608206857785,
+      "grad_norm": 0.2697280943393707,
+      "learning_rate": 0.0001826126126126126,
+      "loss": 0.4153,
+      "step": 2616
+    },
+    {
+      "epoch": 0.7845606145774779,
+      "grad_norm": 0.2661687731742859,
+      "learning_rate": 0.00018256756756756756,
+      "loss": 0.4213,
+      "step": 2617
+    },
+    {
+      "epoch": 0.7848604084691775,
+      "grad_norm": 0.2607208788394928,
+      "learning_rate": 0.0001825225225225225,
+      "loss": 0.4248,
+      "step": 2618
+    },
+    {
+      "epoch": 0.7851602023608769,
+      "grad_norm": 0.250765860080719,
+      "learning_rate": 0.00018247747747747749,
+      "loss": 0.4208,
+      "step": 2619
+    },
+    {
+      "epoch": 0.7854599962525763,
+      "grad_norm": 0.29620206356048584,
+      "learning_rate": 0.00018243243243243242,
+      "loss": 0.4231,
+      "step": 2620
+    },
+    {
+      "epoch": 0.7857597901442758,
+      "grad_norm": 0.26590675115585327,
+      "learning_rate": 0.00018238738738738736,
+      "loss": 0.4526,
+      "step": 2621
+    },
+    {
+      "epoch": 0.7860595840359753,
+      "grad_norm": 0.2643924057483673,
+      "learning_rate": 0.00018234234234234235,
+      "loss": 0.453,
+      "step": 2622
+    },
+    {
+      "epoch": 0.7863593779276747,
+      "grad_norm": 0.25070053339004517,
+      "learning_rate": 0.0001822972972972973,
+      "loss": 0.4142,
+      "step": 2623
+    },
+    {
+      "epoch": 0.7866591718193742,
+      "grad_norm": 0.2581370770931244,
+      "learning_rate": 0.00018225225225225222,
+      "loss": 0.4266,
+      "step": 2624
+    },
+    {
+      "epoch": 0.7869589657110736,
+      "grad_norm": 0.2647317945957184,
+      "learning_rate": 0.00018220720720720722,
+      "loss": 0.4648,
+      "step": 2625
+    },
+    {
+      "epoch": 0.7872587596027731,
+      "grad_norm": 0.25011420249938965,
+      "learning_rate": 0.00018216216216216215,
+      "loss": 0.441,
+      "step": 2626
+    },
+    {
+      "epoch": 0.7875585534944726,
+      "grad_norm": 0.25477978587150574,
+      "learning_rate": 0.0001821171171171171,
+      "loss": 0.4305,
+      "step": 2627
+    },
+    {
+      "epoch": 0.787858347386172,
+      "grad_norm": 0.24877221882343292,
+      "learning_rate": 0.00018207207207207208,
+      "loss": 0.4232,
+      "step": 2628
+    },
+    {
+      "epoch": 0.7881581412778714,
+      "grad_norm": 0.26181760430336,
+      "learning_rate": 0.00018202702702702702,
+      "loss": 0.4417,
+      "step": 2629
+    },
+    {
+      "epoch": 0.788457935169571,
+      "grad_norm": 0.2556387782096863,
+      "learning_rate": 0.00018198198198198195,
+      "loss": 0.4213,
+      "step": 2630
+    },
+    {
+      "epoch": 0.7887577290612704,
+      "grad_norm": 0.2542903423309326,
+      "learning_rate": 0.00018193693693693692,
+      "loss": 0.4179,
+      "step": 2631
+    },
+    {
+      "epoch": 0.7890575229529698,
+      "grad_norm": 0.2722475528717041,
+      "learning_rate": 0.00018189189189189188,
+      "loss": 0.4424,
+      "step": 2632
+    },
+    {
+      "epoch": 0.7893573168446693,
+      "grad_norm": 0.26245227456092834,
+      "learning_rate": 0.00018184684684684682,
+      "loss": 0.4263,
+      "step": 2633
+    },
+    {
+      "epoch": 0.7896571107363688,
+      "grad_norm": 0.27598729729652405,
+      "learning_rate": 0.00018180180180180178,
+      "loss": 0.438,
+      "step": 2634
+    },
+    {
+      "epoch": 0.7899569046280682,
+      "grad_norm": 0.25007325410842896,
+      "learning_rate": 0.00018175675675675675,
+      "loss": 0.4519,
+      "step": 2635
+    },
+    {
+      "epoch": 0.7902566985197677,
+      "grad_norm": 0.25102177262306213,
+      "learning_rate": 0.00018171171171171168,
+      "loss": 0.3859,
+      "step": 2636
+    },
+    {
+      "epoch": 0.7905564924114671,
+      "grad_norm": 0.26916658878326416,
+      "learning_rate": 0.00018166666666666665,
+      "loss": 0.4643,
+      "step": 2637
+    },
+    {
+      "epoch": 0.7908562863031666,
+      "grad_norm": 0.2508098781108856,
+      "learning_rate": 0.0001816216216216216,
+      "loss": 0.4211,
+      "step": 2638
+    },
+    {
+      "epoch": 0.791156080194866,
+      "grad_norm": 0.24109424650669098,
+      "learning_rate": 0.00018157657657657655,
+      "loss": 0.418,
+      "step": 2639
+    },
+    {
+      "epoch": 0.7914558740865655,
+      "grad_norm": 0.24954918026924133,
+      "learning_rate": 0.0001815315315315315,
+      "loss": 0.4219,
+      "step": 2640
+    },
+    {
+      "epoch": 0.7917556679782649,
+      "grad_norm": 0.24477514624595642,
+      "learning_rate": 0.00018148648648648647,
+      "loss": 0.4165,
+      "step": 2641
+    },
+    {
+      "epoch": 0.7920554618699645,
+      "grad_norm": 0.24159999191761017,
+      "learning_rate": 0.00018144144144144144,
+      "loss": 0.4327,
+      "step": 2642
+    },
+    {
+      "epoch": 0.7923552557616639,
+      "grad_norm": 0.2405584752559662,
+      "learning_rate": 0.00018139639639639638,
+      "loss": 0.4041,
+      "step": 2643
+    },
+    {
+      "epoch": 0.7926550496533633,
+      "grad_norm": 0.2744334638118744,
+      "learning_rate": 0.00018135135135135134,
+      "loss": 0.4128,
+      "step": 2644
+    },
+    {
+      "epoch": 0.7929548435450627,
+      "grad_norm": 0.24337726831436157,
+      "learning_rate": 0.0001813063063063063,
+      "loss": 0.4254,
+      "step": 2645
+    },
+    {
+      "epoch": 0.7932546374367623,
+      "grad_norm": 0.24543915688991547,
+      "learning_rate": 0.00018126126126126124,
+      "loss": 0.4296,
+      "step": 2646
+    },
+    {
+      "epoch": 0.7935544313284617,
+      "grad_norm": 0.2545223832130432,
+      "learning_rate": 0.0001812162162162162,
+      "loss": 0.4446,
+      "step": 2647
+    },
+    {
+      "epoch": 0.7938542252201611,
+      "grad_norm": 0.2627246379852295,
+      "learning_rate": 0.00018117117117117117,
+      "loss": 0.4304,
+      "step": 2648
+    },
+    {
+      "epoch": 0.7941540191118606,
+      "grad_norm": 0.269690603017807,
+      "learning_rate": 0.0001811261261261261,
+      "loss": 0.4417,
+      "step": 2649
+    },
+    {
+      "epoch": 0.7944538130035601,
+      "grad_norm": 0.26495659351348877,
+      "learning_rate": 0.00018108108108108107,
+      "loss": 0.4274,
+      "step": 2650
+    },
+    {
+      "epoch": 0.7947536068952595,
+      "grad_norm": 0.24386748671531677,
+      "learning_rate": 0.00018103603603603603,
+      "loss": 0.402,
+      "step": 2651
+    },
+    {
+      "epoch": 0.795053400786959,
+      "grad_norm": 0.2787957489490509,
+      "learning_rate": 0.00018099099099099097,
+      "loss": 0.4285,
+      "step": 2652
+    },
+    {
+      "epoch": 0.7953531946786584,
+      "grad_norm": 0.28965434432029724,
+      "learning_rate": 0.0001809459459459459,
+      "loss": 0.4296,
+      "step": 2653
+    },
+    {
+      "epoch": 0.7956529885703579,
+      "grad_norm": 0.27486535906791687,
+      "learning_rate": 0.0001809009009009009,
+      "loss": 0.4383,
+      "step": 2654
+    },
+    {
+      "epoch": 0.7959527824620574,
+      "grad_norm": 0.29988306760787964,
+      "learning_rate": 0.00018085585585585583,
+      "loss": 0.4326,
+      "step": 2655
+    },
+    {
+      "epoch": 0.7962525763537568,
+      "grad_norm": 0.2822316884994507,
+      "learning_rate": 0.00018081081081081077,
+      "loss": 0.4497,
+      "step": 2656
+    },
+    {
+      "epoch": 0.7965523702454562,
+      "grad_norm": 0.27767157554626465,
+      "learning_rate": 0.00018076576576576576,
+      "loss": 0.4461,
+      "step": 2657
+    },
+    {
+      "epoch": 0.7968521641371558,
+      "grad_norm": 0.27223673462867737,
+      "learning_rate": 0.0001807207207207207,
+      "loss": 0.4338,
+      "step": 2658
+    },
+    {
+      "epoch": 0.7971519580288552,
+      "grad_norm": 0.26280853152275085,
+      "learning_rate": 0.00018067567567567563,
+      "loss": 0.4556,
+      "step": 2659
+    },
+    {
+      "epoch": 0.7974517519205546,
+      "grad_norm": 0.29061955213546753,
+      "learning_rate": 0.00018063063063063063,
+      "loss": 0.4494,
+      "step": 2660
+    },
+    {
+      "epoch": 0.797751545812254,
+      "grad_norm": 0.2875960171222687,
+      "learning_rate": 0.00018058558558558556,
+      "loss": 0.4286,
+      "step": 2661
+    },
+    {
+      "epoch": 0.7980513397039535,
+      "grad_norm": 0.26884058117866516,
+      "learning_rate": 0.0001805405405405405,
+      "loss": 0.4196,
+      "step": 2662
+    },
+    {
+      "epoch": 0.798351133595653,
+      "grad_norm": 0.25288286805152893,
+      "learning_rate": 0.0001804954954954955,
+      "loss": 0.4313,
+      "step": 2663
+    },
+    {
+      "epoch": 0.7986509274873524,
+      "grad_norm": 0.2743682861328125,
+      "learning_rate": 0.00018045045045045043,
+      "loss": 0.4016,
+      "step": 2664
+    },
+    {
+      "epoch": 0.7989507213790519,
+      "grad_norm": 0.2604522109031677,
+      "learning_rate": 0.0001804054054054054,
+      "loss": 0.4199,
+      "step": 2665
+    },
+    {
+      "epoch": 0.7992505152707513,
+      "grad_norm": 0.2633124589920044,
+      "learning_rate": 0.00018036036036036035,
+      "loss": 0.4512,
+      "step": 2666
+    },
+    {
+      "epoch": 0.7995503091624508,
+      "grad_norm": 0.2687436640262604,
+      "learning_rate": 0.0001803153153153153,
+      "loss": 0.4512,
+      "step": 2667
+    },
+    {
+      "epoch": 0.7998501030541503,
+      "grad_norm": 0.26707741618156433,
+      "learning_rate": 0.00018027027027027026,
+      "loss": 0.4096,
+      "step": 2668
+    },
+    {
+      "epoch": 0.8001498969458497,
+      "grad_norm": 0.25287461280822754,
+      "learning_rate": 0.00018022522522522522,
+      "loss": 0.4235,
+      "step": 2669
+    },
+    {
+      "epoch": 0.8004496908375491,
+      "grad_norm": 0.2665085792541504,
+      "learning_rate": 0.00018018018018018016,
+      "loss": 0.4421,
+      "step": 2670
+    },
+    {
+      "epoch": 0.8007494847292487,
+      "grad_norm": 0.2533876895904541,
+      "learning_rate": 0.00018013513513513512,
+      "loss": 0.4031,
+      "step": 2671
+    },
+    {
+      "epoch": 0.8010492786209481,
+      "grad_norm": 0.27074354887008667,
+      "learning_rate": 0.00018009009009009008,
+      "loss": 0.4811,
+      "step": 2672
+    },
+    {
+      "epoch": 0.8013490725126475,
+      "grad_norm": 0.24895958602428436,
+      "learning_rate": 0.00018004504504504502,
+      "loss": 0.4346,
+      "step": 2673
+    },
+    {
+      "epoch": 0.801648866404347,
+      "grad_norm": 0.2514555752277374,
+      "learning_rate": 0.00017999999999999998,
+      "loss": 0.4188,
+      "step": 2674
+    },
+    {
+      "epoch": 0.8019486602960465,
+      "grad_norm": 0.25880131125450134,
+      "learning_rate": 0.00017995495495495495,
+      "loss": 0.4125,
+      "step": 2675
+    },
+    {
+      "epoch": 0.8022484541877459,
+      "grad_norm": 0.2755208909511566,
+      "learning_rate": 0.0001799099099099099,
+      "loss": 0.4448,
+      "step": 2676
+    },
+    {
+      "epoch": 0.8025482480794454,
+      "grad_norm": 0.251765638589859,
+      "learning_rate": 0.00017986486486486485,
+      "loss": 0.4216,
+      "step": 2677
+    },
+    {
+      "epoch": 0.8028480419711448,
+      "grad_norm": 0.27061545848846436,
+      "learning_rate": 0.00017981981981981979,
+      "loss": 0.4392,
+      "step": 2678
+    },
+    {
+      "epoch": 0.8031478358628443,
+      "grad_norm": 0.2883546054363251,
+      "learning_rate": 0.00017977477477477478,
+      "loss": 0.4486,
+      "step": 2679
+    },
+    {
+      "epoch": 0.8034476297545438,
+      "grad_norm": 0.2492683082818985,
+      "learning_rate": 0.0001797297297297297,
+      "loss": 0.441,
+      "step": 2680
+    },
+    {
+      "epoch": 0.8037474236462432,
+      "grad_norm": 0.24431340396404266,
+      "learning_rate": 0.00017968468468468465,
+      "loss": 0.4181,
+      "step": 2681
+    },
+    {
+      "epoch": 0.8040472175379426,
+      "grad_norm": 0.24715867638587952,
+      "learning_rate": 0.00017963963963963964,
+      "loss": 0.4137,
+      "step": 2682
+    },
+    {
+      "epoch": 0.8043470114296422,
+      "grad_norm": 0.27124032378196716,
+      "learning_rate": 0.00017959459459459458,
+      "loss": 0.4308,
+      "step": 2683
+    },
+    {
+      "epoch": 0.8046468053213416,
+      "grad_norm": 0.25120750069618225,
+      "learning_rate": 0.00017954954954954951,
+      "loss": 0.4183,
+      "step": 2684
+    },
+    {
+      "epoch": 0.804946599213041,
+      "grad_norm": 0.2443256676197052,
+      "learning_rate": 0.0001795045045045045,
+      "loss": 0.3974,
+      "step": 2685
+    },
+    {
+      "epoch": 0.8052463931047404,
+      "grad_norm": 0.2588818669319153,
+      "learning_rate": 0.00017945945945945944,
+      "loss": 0.4281,
+      "step": 2686
+    },
+    {
+      "epoch": 0.80554618699644,
+      "grad_norm": 0.26433560252189636,
+      "learning_rate": 0.00017941441441441438,
+      "loss": 0.4313,
+      "step": 2687
+    },
+    {
+      "epoch": 0.8058459808881394,
+      "grad_norm": 0.2529260814189911,
+      "learning_rate": 0.00017936936936936937,
+      "loss": 0.4398,
+      "step": 2688
+    },
+    {
+      "epoch": 0.8061457747798388,
+      "grad_norm": 0.25349390506744385,
+      "learning_rate": 0.0001793243243243243,
+      "loss": 0.4388,
+      "step": 2689
+    },
+    {
+      "epoch": 0.8064455686715383,
+      "grad_norm": 0.2616881728172302,
+      "learning_rate": 0.00017927927927927924,
+      "loss": 0.4135,
+      "step": 2690
+    },
+    {
+      "epoch": 0.8067453625632378,
+      "grad_norm": 0.24046184122562408,
+      "learning_rate": 0.00017923423423423423,
+      "loss": 0.4276,
+      "step": 2691
+    },
+    {
+      "epoch": 0.8070451564549372,
+      "grad_norm": 0.2653987407684326,
+      "learning_rate": 0.00017918918918918917,
+      "loss": 0.4515,
+      "step": 2692
+    },
+    {
+      "epoch": 0.8073449503466367,
+      "grad_norm": 0.27617138624191284,
+      "learning_rate": 0.0001791441441441441,
+      "loss": 0.4767,
+      "step": 2693
+    },
+    {
+      "epoch": 0.8076447442383361,
+      "grad_norm": 0.24791288375854492,
+      "learning_rate": 0.0001790990990990991,
+      "loss": 0.4203,
+      "step": 2694
+    },
+    {
+      "epoch": 0.8079445381300356,
+      "grad_norm": 0.2514311671257019,
+      "learning_rate": 0.00017905405405405404,
+      "loss": 0.4157,
+      "step": 2695
+    },
+    {
+      "epoch": 0.8082443320217351,
+      "grad_norm": 0.25074705481529236,
+      "learning_rate": 0.00017900900900900897,
+      "loss": 0.4327,
+      "step": 2696
+    },
+    {
+      "epoch": 0.8085441259134345,
+      "grad_norm": 0.2415165901184082,
+      "learning_rate": 0.00017896396396396396,
+      "loss": 0.4204,
+      "step": 2697
+    },
+    {
+      "epoch": 0.8088439198051339,
+      "grad_norm": 0.23986560106277466,
+      "learning_rate": 0.0001789189189189189,
+      "loss": 0.4143,
+      "step": 2698
+    },
+    {
+      "epoch": 0.8091437136968335,
+      "grad_norm": 0.24843257665634155,
+      "learning_rate": 0.00017887387387387386,
+      "loss": 0.4406,
+      "step": 2699
+    },
+    {
+      "epoch": 0.8094435075885329,
+      "grad_norm": 0.2549048960208893,
+      "learning_rate": 0.00017882882882882883,
+      "loss": 0.4293,
+      "step": 2700
+    },
+    {
+      "epoch": 0.8097433014802323,
+      "grad_norm": 0.2509189248085022,
+      "learning_rate": 0.00017878378378378376,
+      "loss": 0.4236,
+      "step": 2701
+    },
+    {
+      "epoch": 0.8100430953719318,
+      "grad_norm": 0.2555721402168274,
+      "learning_rate": 0.00017873873873873873,
+      "loss": 0.4387,
+      "step": 2702
+    },
+    {
+      "epoch": 0.8103428892636313,
+      "grad_norm": 0.2514118552207947,
+      "learning_rate": 0.00017869369369369367,
+      "loss": 0.4182,
+      "step": 2703
+    },
+    {
+      "epoch": 0.8106426831553307,
+      "grad_norm": 0.2508332133293152,
+      "learning_rate": 0.00017864864864864863,
+      "loss": 0.4283,
+      "step": 2704
+    },
+    {
+      "epoch": 0.8109424770470302,
+      "grad_norm": 0.2642100155353546,
+      "learning_rate": 0.0001786036036036036,
+      "loss": 0.4069,
+      "step": 2705
+    },
+    {
+      "epoch": 0.8112422709387296,
+      "grad_norm": 0.24579380452632904,
+      "learning_rate": 0.00017855855855855853,
+      "loss": 0.4229,
+      "step": 2706
+    },
+    {
+      "epoch": 0.8115420648304291,
+      "grad_norm": 0.25910264253616333,
+      "learning_rate": 0.0001785135135135135,
+      "loss": 0.422,
+      "step": 2707
+    },
+    {
+      "epoch": 0.8118418587221286,
+      "grad_norm": 0.256245493888855,
+      "learning_rate": 0.00017846846846846846,
+      "loss": 0.4223,
+      "step": 2708
+    },
+    {
+      "epoch": 0.812141652613828,
+      "grad_norm": 0.2674013376235962,
+      "learning_rate": 0.0001784234234234234,
+      "loss": 0.4256,
+      "step": 2709
+    },
+    {
+      "epoch": 0.8124414465055274,
+      "grad_norm": 0.25418075919151306,
+      "learning_rate": 0.00017837837837837839,
+      "loss": 0.4173,
+      "step": 2710
+    },
+    {
+      "epoch": 0.812741240397227,
+      "grad_norm": 0.25557002425193787,
+      "learning_rate": 0.00017833333333333332,
+      "loss": 0.4434,
+      "step": 2711
+    },
+    {
+      "epoch": 0.8130410342889264,
+      "grad_norm": 0.25232791900634766,
+      "learning_rate": 0.00017828828828828826,
+      "loss": 0.4366,
+      "step": 2712
+    },
+    {
+      "epoch": 0.8133408281806258,
+      "grad_norm": 0.24985311925411224,
+      "learning_rate": 0.00017824324324324325,
+      "loss": 0.4352,
+      "step": 2713
+    },
+    {
+      "epoch": 0.8136406220723252,
+      "grad_norm": 0.27299973368644714,
+      "learning_rate": 0.0001781981981981982,
+      "loss": 0.4017,
+      "step": 2714
+    },
+    {
+      "epoch": 0.8139404159640248,
+      "grad_norm": 0.24529805779457092,
+      "learning_rate": 0.00017815315315315312,
+      "loss": 0.4089,
+      "step": 2715
+    },
+    {
+      "epoch": 0.8142402098557242,
+      "grad_norm": 0.28326576948165894,
+      "learning_rate": 0.00017810810810810811,
+      "loss": 0.427,
+      "step": 2716
+    },
+    {
+      "epoch": 0.8145400037474236,
+      "grad_norm": 0.25834745168685913,
+      "learning_rate": 0.00017806306306306305,
+      "loss": 0.4257,
+      "step": 2717
+    },
+    {
+      "epoch": 0.8148397976391231,
+      "grad_norm": 0.31774160265922546,
+      "learning_rate": 0.000178018018018018,
+      "loss": 0.4321,
+      "step": 2718
+    },
+    {
+      "epoch": 0.8151395915308226,
+      "grad_norm": 0.26752525568008423,
+      "learning_rate": 0.00017797297297297298,
+      "loss": 0.4263,
+      "step": 2719
+    },
+    {
+      "epoch": 0.815439385422522,
+      "grad_norm": 0.26342013478279114,
+      "learning_rate": 0.00017792792792792792,
+      "loss": 0.4214,
+      "step": 2720
+    },
+    {
+      "epoch": 0.8157391793142215,
+      "grad_norm": 0.2852500081062317,
+      "learning_rate": 0.00017788288288288285,
+      "loss": 0.4123,
+      "step": 2721
+    },
+    {
+      "epoch": 0.8160389732059209,
+      "grad_norm": 0.26281121373176575,
+      "learning_rate": 0.00017783783783783784,
+      "loss": 0.4199,
+      "step": 2722
+    },
+    {
+      "epoch": 0.8163387670976204,
+      "grad_norm": 0.2701656222343445,
+      "learning_rate": 0.00017779279279279278,
+      "loss": 0.4002,
+      "step": 2723
+    },
+    {
+      "epoch": 0.8166385609893199,
+      "grad_norm": 0.28216591477394104,
+      "learning_rate": 0.00017774774774774772,
+      "loss": 0.4383,
+      "step": 2724
+    },
+    {
+      "epoch": 0.8169383548810193,
+      "grad_norm": 0.2807302176952362,
+      "learning_rate": 0.00017770270270270268,
+      "loss": 0.4481,
+      "step": 2725
+    },
+    {
+      "epoch": 0.8172381487727187,
+      "grad_norm": 0.27764102816581726,
+      "learning_rate": 0.00017765765765765764,
+      "loss": 0.4168,
+      "step": 2726
+    },
+    {
+      "epoch": 0.8175379426644183,
+      "grad_norm": 0.25099071860313416,
+      "learning_rate": 0.00017761261261261258,
+      "loss": 0.4052,
+      "step": 2727
+    },
+    {
+      "epoch": 0.8178377365561177,
+      "grad_norm": 0.24437499046325684,
+      "learning_rate": 0.00017756756756756755,
+      "loss": 0.4181,
+      "step": 2728
+    },
+    {
+      "epoch": 0.8181375304478171,
+      "grad_norm": 0.2668071389198303,
+      "learning_rate": 0.0001775225225225225,
+      "loss": 0.415,
+      "step": 2729
+    },
+    {
+      "epoch": 0.8184373243395165,
+      "grad_norm": 0.24040313065052032,
+      "learning_rate": 0.00017747747747747745,
+      "loss": 0.4251,
+      "step": 2730
+    },
+    {
+      "epoch": 0.8187371182312161,
+      "grad_norm": 0.2401929497718811,
+      "learning_rate": 0.0001774324324324324,
+      "loss": 0.4153,
+      "step": 2731
+    },
+    {
+      "epoch": 0.8190369121229155,
+      "grad_norm": 0.2651936709880829,
+      "learning_rate": 0.00017738738738738737,
+      "loss": 0.4327,
+      "step": 2732
+    },
+    {
+      "epoch": 0.819336706014615,
+      "grad_norm": 0.24846699833869934,
+      "learning_rate": 0.00017734234234234234,
+      "loss": 0.4213,
+      "step": 2733
+    },
+    {
+      "epoch": 0.8196364999063144,
+      "grad_norm": 0.24691832065582275,
+      "learning_rate": 0.00017729729729729727,
+      "loss": 0.4218,
+      "step": 2734
+    },
+    {
+      "epoch": 0.8199362937980139,
+      "grad_norm": 0.25850343704223633,
+      "learning_rate": 0.00017725225225225224,
+      "loss": 0.4363,
+      "step": 2735
+    },
+    {
+      "epoch": 0.8202360876897133,
+      "grad_norm": 0.27458709478378296,
+      "learning_rate": 0.0001772072072072072,
+      "loss": 0.4353,
+      "step": 2736
+    },
+    {
+      "epoch": 0.8205358815814128,
+      "grad_norm": 0.26524078845977783,
+      "learning_rate": 0.00017716216216216214,
+      "loss": 0.4367,
+      "step": 2737
+    },
+    {
+      "epoch": 0.8208356754731122,
+      "grad_norm": 0.24760177731513977,
+      "learning_rate": 0.0001771171171171171,
+      "loss": 0.4272,
+      "step": 2738
+    },
+    {
+      "epoch": 0.8211354693648117,
+      "grad_norm": 0.2548927068710327,
+      "learning_rate": 0.00017707207207207207,
+      "loss": 0.4247,
+      "step": 2739
+    },
+    {
+      "epoch": 0.8214352632565112,
+      "grad_norm": 0.2804696559906006,
+      "learning_rate": 0.000177027027027027,
+      "loss": 0.4296,
+      "step": 2740
+    },
+    {
+      "epoch": 0.8217350571482106,
+      "grad_norm": 0.27337226271629333,
+      "learning_rate": 0.00017698198198198197,
+      "loss": 0.4573,
+      "step": 2741
+    },
+    {
+      "epoch": 0.82203485103991,
+      "grad_norm": 0.26369425654411316,
+      "learning_rate": 0.00017693693693693693,
+      "loss": 0.4569,
+      "step": 2742
+    },
+    {
+      "epoch": 0.8223346449316096,
+      "grad_norm": 0.2607530653476715,
+      "learning_rate": 0.00017689189189189187,
+      "loss": 0.43,
+      "step": 2743
+    },
+    {
+      "epoch": 0.822634438823309,
+      "grad_norm": 0.2663254737854004,
+      "learning_rate": 0.00017684684684684686,
+      "loss": 0.4144,
+      "step": 2744
+    },
+    {
+      "epoch": 0.8229342327150084,
+      "grad_norm": 0.25760191679000854,
+      "learning_rate": 0.0001768018018018018,
+      "loss": 0.4257,
+      "step": 2745
+    },
+    {
+      "epoch": 0.8232340266067079,
+      "grad_norm": 0.27796119451522827,
+      "learning_rate": 0.00017675675675675673,
+      "loss": 0.4204,
+      "step": 2746
+    },
+    {
+      "epoch": 0.8235338204984074,
+      "grad_norm": 0.27072879672050476,
+      "learning_rate": 0.00017671171171171172,
+      "loss": 0.4296,
+      "step": 2747
+    },
+    {
+      "epoch": 0.8238336143901068,
+      "grad_norm": 0.2645798623561859,
+      "learning_rate": 0.00017666666666666666,
+      "loss": 0.4181,
+      "step": 2748
+    },
+    {
+      "epoch": 0.8241334082818063,
+      "grad_norm": 0.2641856372356415,
+      "learning_rate": 0.0001766216216216216,
+      "loss": 0.4341,
+      "step": 2749
+    },
+    {
+      "epoch": 0.8244332021735057,
+      "grad_norm": 0.26364126801490784,
+      "learning_rate": 0.00017657657657657653,
+      "loss": 0.3884,
+      "step": 2750
+    },
+    {
+      "epoch": 0.8247329960652052,
+      "grad_norm": 0.23896346986293793,
+      "learning_rate": 0.00017653153153153152,
+      "loss": 0.4193,
+      "step": 2751
+    },
+    {
+      "epoch": 0.8250327899569047,
+      "grad_norm": 0.2746429443359375,
+      "learning_rate": 0.00017648648648648646,
+      "loss": 0.4248,
+      "step": 2752
+    },
+    {
+      "epoch": 0.8253325838486041,
+      "grad_norm": 0.251149445772171,
+      "learning_rate": 0.0001764414414414414,
+      "loss": 0.4223,
+      "step": 2753
+    },
+    {
+      "epoch": 0.8256323777403035,
+      "grad_norm": 0.26169952750205994,
+      "learning_rate": 0.0001763963963963964,
+      "loss": 0.4257,
+      "step": 2754
+    },
+    {
+      "epoch": 0.8259321716320029,
+      "grad_norm": 0.2569408416748047,
+      "learning_rate": 0.00017635135135135133,
+      "loss": 0.4032,
+      "step": 2755
+    },
+    {
+      "epoch": 0.8262319655237025,
+      "grad_norm": 0.24922534823417664,
+      "learning_rate": 0.0001763063063063063,
+      "loss": 0.414,
+      "step": 2756
+    },
+    {
+      "epoch": 0.8265317594154019,
+      "grad_norm": 0.26194071769714355,
+      "learning_rate": 0.00017626126126126125,
+      "loss": 0.3991,
+      "step": 2757
+    },
+    {
+      "epoch": 0.8268315533071013,
+      "grad_norm": 0.2403927892446518,
+      "learning_rate": 0.0001762162162162162,
+      "loss": 0.3935,
+      "step": 2758
+    },
+    {
+      "epoch": 0.8271313471988008,
+      "grad_norm": 0.25983864068984985,
+      "learning_rate": 0.00017617117117117115,
+      "loss": 0.4249,
+      "step": 2759
+    },
+    {
+      "epoch": 0.8274311410905003,
+      "grad_norm": 0.2580634355545044,
+      "learning_rate": 0.00017612612612612612,
+      "loss": 0.4209,
+      "step": 2760
+    },
+    {
+      "epoch": 0.8277309349821997,
+      "grad_norm": 0.2543208599090576,
+      "learning_rate": 0.00017608108108108106,
+      "loss": 0.4126,
+      "step": 2761
+    },
+    {
+      "epoch": 0.8280307288738992,
+      "grad_norm": 0.27977827191352844,
+      "learning_rate": 0.00017603603603603602,
+      "loss": 0.4161,
+      "step": 2762
+    },
+    {
+      "epoch": 0.8283305227655986,
+      "grad_norm": 0.27077874541282654,
+      "learning_rate": 0.00017599099099099098,
+      "loss": 0.4341,
+      "step": 2763
+    },
+    {
+      "epoch": 0.8286303166572981,
+      "grad_norm": 0.2763903737068176,
+      "learning_rate": 0.00017594594594594592,
+      "loss": 0.4444,
+      "step": 2764
+    },
+    {
+      "epoch": 0.8289301105489976,
+      "grad_norm": 0.26092103123664856,
+      "learning_rate": 0.00017590090090090088,
+      "loss": 0.4542,
+      "step": 2765
+    },
+    {
+      "epoch": 0.829229904440697,
+      "grad_norm": 0.25859686732292175,
+      "learning_rate": 0.00017585585585585585,
+      "loss": 0.443,
+      "step": 2766
+    },
+    {
+      "epoch": 0.8295296983323964,
+      "grad_norm": 0.26898616552352905,
+      "learning_rate": 0.0001758108108108108,
+      "loss": 0.421,
+      "step": 2767
+    },
+    {
+      "epoch": 0.829829492224096,
+      "grad_norm": 0.2678453028202057,
+      "learning_rate": 0.00017576576576576575,
+      "loss": 0.4431,
+      "step": 2768
+    },
+    {
+      "epoch": 0.8301292861157954,
+      "grad_norm": 0.27072227001190186,
+      "learning_rate": 0.0001757207207207207,
+      "loss": 0.4353,
+      "step": 2769
+    },
+    {
+      "epoch": 0.8304290800074948,
+      "grad_norm": 0.2593313455581665,
+      "learning_rate": 0.00017567567567567568,
+      "loss": 0.4102,
+      "step": 2770
+    },
+    {
+      "epoch": 0.8307288738991943,
+      "grad_norm": 0.2503952085971832,
+      "learning_rate": 0.0001756306306306306,
+      "loss": 0.4102,
+      "step": 2771
+    },
+    {
+      "epoch": 0.8310286677908938,
+      "grad_norm": 0.2745724618434906,
+      "learning_rate": 0.00017558558558558558,
+      "loss": 0.4364,
+      "step": 2772
+    },
+    {
+      "epoch": 0.8313284616825932,
+      "grad_norm": 0.27861306071281433,
+      "learning_rate": 0.00017554054054054054,
+      "loss": 0.4501,
+      "step": 2773
+    },
+    {
+      "epoch": 0.8316282555742927,
+      "grad_norm": 0.2669075131416321,
+      "learning_rate": 0.00017549549549549548,
+      "loss": 0.4388,
+      "step": 2774
+    },
+    {
+      "epoch": 0.8319280494659921,
+      "grad_norm": 0.25937843322753906,
+      "learning_rate": 0.00017545045045045041,
+      "loss": 0.4528,
+      "step": 2775
+    },
+    {
+      "epoch": 0.8322278433576916,
+      "grad_norm": 0.26331016421318054,
+      "learning_rate": 0.0001754054054054054,
+      "loss": 0.4609,
+      "step": 2776
+    },
+    {
+      "epoch": 0.832527637249391,
+      "grad_norm": 0.27154049277305603,
+      "learning_rate": 0.00017536036036036034,
+      "loss": 0.4908,
+      "step": 2777
+    },
+    {
+      "epoch": 0.8328274311410905,
+      "grad_norm": 0.27023017406463623,
+      "learning_rate": 0.00017531531531531528,
+      "loss": 0.4294,
+      "step": 2778
+    },
+    {
+      "epoch": 0.8331272250327899,
+      "grad_norm": 0.26888102293014526,
+      "learning_rate": 0.00017527027027027027,
+      "loss": 0.4371,
+      "step": 2779
+    },
+    {
+      "epoch": 0.8334270189244894,
+      "grad_norm": 0.24405723810195923,
+      "learning_rate": 0.0001752252252252252,
+      "loss": 0.4035,
+      "step": 2780
+    },
+    {
+      "epoch": 0.8337268128161889,
+      "grad_norm": 0.26839005947113037,
+      "learning_rate": 0.00017518018018018014,
+      "loss": 0.4109,
+      "step": 2781
+    },
+    {
+      "epoch": 0.8340266067078883,
+      "grad_norm": 0.26782795786857605,
+      "learning_rate": 0.00017513513513513513,
+      "loss": 0.4072,
+      "step": 2782
+    },
+    {
+      "epoch": 0.8343264005995877,
+      "grad_norm": 0.2640438973903656,
+      "learning_rate": 0.00017509009009009007,
+      "loss": 0.4183,
+      "step": 2783
+    },
+    {
+      "epoch": 0.8346261944912873,
+      "grad_norm": 0.26554253697395325,
+      "learning_rate": 0.000175045045045045,
+      "loss": 0.3979,
+      "step": 2784
+    },
+    {
+      "epoch": 0.8349259883829867,
+      "grad_norm": 0.25945335626602173,
+      "learning_rate": 0.000175,
+      "loss": 0.4196,
+      "step": 2785
+    },
+    {
+      "epoch": 0.8352257822746861,
+      "grad_norm": 0.29560789465904236,
+      "learning_rate": 0.00017495495495495493,
+      "loss": 0.4584,
+      "step": 2786
+    },
+    {
+      "epoch": 0.8355255761663856,
+      "grad_norm": 0.2612930238246918,
+      "learning_rate": 0.00017490990990990987,
+      "loss": 0.4172,
+      "step": 2787
+    },
+    {
+      "epoch": 0.8358253700580851,
+      "grad_norm": 0.23529918491840363,
+      "learning_rate": 0.00017486486486486486,
+      "loss": 0.3909,
+      "step": 2788
+    },
+    {
+      "epoch": 0.8361251639497845,
+      "grad_norm": 0.2858099341392517,
+      "learning_rate": 0.0001748198198198198,
+      "loss": 0.444,
+      "step": 2789
+    },
+    {
+      "epoch": 0.836424957841484,
+      "grad_norm": 0.26430973410606384,
+      "learning_rate": 0.00017477477477477476,
+      "loss": 0.4329,
+      "step": 2790
+    },
+    {
+      "epoch": 0.8367247517331834,
+      "grad_norm": 0.256142795085907,
+      "learning_rate": 0.00017472972972972973,
+      "loss": 0.4299,
+      "step": 2791
+    },
+    {
+      "epoch": 0.8370245456248829,
+      "grad_norm": 0.26407790184020996,
+      "learning_rate": 0.00017468468468468466,
+      "loss": 0.4296,
+      "step": 2792
+    },
+    {
+      "epoch": 0.8373243395165824,
+      "grad_norm": 0.2579902708530426,
+      "learning_rate": 0.00017463963963963963,
+      "loss": 0.4243,
+      "step": 2793
+    },
+    {
+      "epoch": 0.8376241334082818,
+      "grad_norm": 0.25428980588912964,
+      "learning_rate": 0.0001745945945945946,
+      "loss": 0.4091,
+      "step": 2794
+    },
+    {
+      "epoch": 0.8379239272999812,
+      "grad_norm": 0.2652067542076111,
+      "learning_rate": 0.00017454954954954953,
+      "loss": 0.4341,
+      "step": 2795
+    },
+    {
+      "epoch": 0.8382237211916808,
+      "grad_norm": 0.2717462182044983,
+      "learning_rate": 0.0001745045045045045,
+      "loss": 0.4426,
+      "step": 2796
+    },
+    {
+      "epoch": 0.8385235150833802,
+      "grad_norm": 0.24994242191314697,
+      "learning_rate": 0.00017445945945945946,
+      "loss": 0.3897,
+      "step": 2797
+    },
+    {
+      "epoch": 0.8388233089750796,
+      "grad_norm": 0.26126357913017273,
+      "learning_rate": 0.0001744144144144144,
+      "loss": 0.4238,
+      "step": 2798
+    },
+    {
+      "epoch": 0.839123102866779,
+      "grad_norm": 0.25549226999282837,
+      "learning_rate": 0.00017436936936936936,
+      "loss": 0.4342,
+      "step": 2799
+    },
+    {
+      "epoch": 0.8394228967584786,
+      "grad_norm": 0.26869869232177734,
+      "learning_rate": 0.0001743243243243243,
+      "loss": 0.436,
+      "step": 2800
+    },
+    {
+      "epoch": 0.839722690650178,
+      "grad_norm": 0.2841474115848541,
+      "learning_rate": 0.00017427927927927928,
+      "loss": 0.4325,
+      "step": 2801
+    },
+    {
+      "epoch": 0.8400224845418774,
+      "grad_norm": 0.25196996331214905,
+      "learning_rate": 0.00017423423423423422,
+      "loss": 0.3917,
+      "step": 2802
+    },
+    {
+      "epoch": 0.8403222784335769,
+      "grad_norm": 0.26642677187919617,
+      "learning_rate": 0.00017418918918918916,
+      "loss": 0.4429,
+      "step": 2803
+    },
+    {
+      "epoch": 0.8406220723252764,
+      "grad_norm": 0.27364763617515564,
+      "learning_rate": 0.00017414414414414415,
+      "loss": 0.4326,
+      "step": 2804
+    },
+    {
+      "epoch": 0.8409218662169758,
+      "grad_norm": 0.28387343883514404,
+      "learning_rate": 0.00017409909909909909,
+      "loss": 0.465,
+      "step": 2805
+    },
+    {
+      "epoch": 0.8412216601086753,
+      "grad_norm": 0.2638298273086548,
+      "learning_rate": 0.00017405405405405402,
+      "loss": 0.4394,
+      "step": 2806
+    },
+    {
+      "epoch": 0.8415214540003747,
+      "grad_norm": 0.23615007102489471,
+      "learning_rate": 0.00017400900900900901,
+      "loss": 0.4126,
+      "step": 2807
+    },
+    {
+      "epoch": 0.8418212478920742,
+      "grad_norm": 0.2662603259086609,
+      "learning_rate": 0.00017396396396396395,
+      "loss": 0.4601,
+      "step": 2808
+    },
+    {
+      "epoch": 0.8421210417837737,
+      "grad_norm": 0.2668648660182953,
+      "learning_rate": 0.0001739189189189189,
+      "loss": 0.4152,
+      "step": 2809
+    },
+    {
+      "epoch": 0.8424208356754731,
+      "grad_norm": 0.23364293575286865,
+      "learning_rate": 0.00017387387387387388,
+      "loss": 0.3843,
+      "step": 2810
+    },
+    {
+      "epoch": 0.8427206295671725,
+      "grad_norm": 0.2743973731994629,
+      "learning_rate": 0.00017382882882882881,
+      "loss": 0.4041,
+      "step": 2811
+    },
+    {
+      "epoch": 0.8430204234588721,
+      "grad_norm": 0.25221478939056396,
+      "learning_rate": 0.00017378378378378375,
+      "loss": 0.4082,
+      "step": 2812
+    },
+    {
+      "epoch": 0.8433202173505715,
+      "grad_norm": 0.27902311086654663,
+      "learning_rate": 0.00017373873873873874,
+      "loss": 0.4437,
+      "step": 2813
+    },
+    {
+      "epoch": 0.8436200112422709,
+      "grad_norm": 0.26741036772727966,
+      "learning_rate": 0.00017369369369369368,
+      "loss": 0.4305,
+      "step": 2814
+    },
+    {
+      "epoch": 0.8439198051339704,
+      "grad_norm": 0.2413070648908615,
+      "learning_rate": 0.00017364864864864862,
+      "loss": 0.3938,
+      "step": 2815
+    },
+    {
+      "epoch": 0.8442195990256699,
+      "grad_norm": 0.2679976224899292,
+      "learning_rate": 0.0001736036036036036,
+      "loss": 0.4505,
+      "step": 2816
+    },
+    {
+      "epoch": 0.8445193929173693,
+      "grad_norm": 0.25923097133636475,
+      "learning_rate": 0.00017355855855855854,
+      "loss": 0.4025,
+      "step": 2817
+    },
+    {
+      "epoch": 0.8448191868090688,
+      "grad_norm": 0.2735726833343506,
+      "learning_rate": 0.00017351351351351348,
+      "loss": 0.4189,
+      "step": 2818
+    },
+    {
+      "epoch": 0.8451189807007682,
+      "grad_norm": 0.24195829033851624,
+      "learning_rate": 0.00017346846846846847,
+      "loss": 0.4193,
+      "step": 2819
+    },
+    {
+      "epoch": 0.8454187745924677,
+      "grad_norm": 0.26585015654563904,
+      "learning_rate": 0.0001734234234234234,
+      "loss": 0.4529,
+      "step": 2820
+    },
+    {
+      "epoch": 0.8457185684841672,
+      "grad_norm": 0.27688056230545044,
+      "learning_rate": 0.00017337837837837835,
+      "loss": 0.4234,
+      "step": 2821
+    },
+    {
+      "epoch": 0.8460183623758666,
+      "grad_norm": 0.2597983181476593,
+      "learning_rate": 0.0001733333333333333,
+      "loss": 0.4262,
+      "step": 2822
+    },
+    {
+      "epoch": 0.846318156267566,
+      "grad_norm": 0.26929470896720886,
+      "learning_rate": 0.00017328828828828827,
+      "loss": 0.4063,
+      "step": 2823
+    },
+    {
+      "epoch": 0.8466179501592656,
+      "grad_norm": 0.2761951982975006,
+      "learning_rate": 0.00017324324324324324,
+      "loss": 0.4005,
+      "step": 2824
+    },
+    {
+      "epoch": 0.846917744050965,
+      "grad_norm": 0.24759140610694885,
+      "learning_rate": 0.00017319819819819817,
+      "loss": 0.4171,
+      "step": 2825
+    },
+    {
+      "epoch": 0.8472175379426644,
+      "grad_norm": 0.2795017659664154,
+      "learning_rate": 0.00017315315315315314,
+      "loss": 0.3981,
+      "step": 2826
+    },
+    {
+      "epoch": 0.8475173318343638,
+      "grad_norm": 0.2652096450328827,
+      "learning_rate": 0.0001731081081081081,
+      "loss": 0.4013,
+      "step": 2827
+    },
+    {
+      "epoch": 0.8478171257260634,
+      "grad_norm": 0.2893044650554657,
+      "learning_rate": 0.00017306306306306304,
+      "loss": 0.4317,
+      "step": 2828
+    },
+    {
+      "epoch": 0.8481169196177628,
+      "grad_norm": 0.28209424018859863,
+      "learning_rate": 0.000173018018018018,
+      "loss": 0.4563,
+      "step": 2829
+    },
+    {
+      "epoch": 0.8484167135094622,
+      "grad_norm": 0.2647794485092163,
+      "learning_rate": 0.00017297297297297297,
+      "loss": 0.4261,
+      "step": 2830
+    },
+    {
+      "epoch": 0.8487165074011617,
+      "grad_norm": 0.25158509612083435,
+      "learning_rate": 0.0001729279279279279,
+      "loss": 0.4109,
+      "step": 2831
+    },
+    {
+      "epoch": 0.8490163012928612,
+      "grad_norm": 0.26878440380096436,
+      "learning_rate": 0.00017288288288288287,
+      "loss": 0.4322,
+      "step": 2832
+    },
+    {
+      "epoch": 0.8493160951845606,
+      "grad_norm": 0.25768351554870605,
+      "learning_rate": 0.00017283783783783783,
+      "loss": 0.4266,
+      "step": 2833
+    },
+    {
+      "epoch": 0.8496158890762601,
+      "grad_norm": 0.26001647114753723,
+      "learning_rate": 0.00017279279279279277,
+      "loss": 0.4299,
+      "step": 2834
+    },
+    {
+      "epoch": 0.8499156829679595,
+      "grad_norm": 0.250027060508728,
+      "learning_rate": 0.00017274774774774773,
+      "loss": 0.4119,
+      "step": 2835
+    },
+    {
+      "epoch": 0.850215476859659,
+      "grad_norm": 0.23465260863304138,
+      "learning_rate": 0.0001727027027027027,
+      "loss": 0.3822,
+      "step": 2836
+    },
+    {
+      "epoch": 0.8505152707513585,
+      "grad_norm": 0.23945166170597076,
+      "learning_rate": 0.00017265765765765763,
+      "loss": 0.4063,
+      "step": 2837
+    },
+    {
+      "epoch": 0.8508150646430579,
+      "grad_norm": 0.2597178518772125,
+      "learning_rate": 0.00017261261261261262,
+      "loss": 0.4011,
+      "step": 2838
+    },
+    {
+      "epoch": 0.8511148585347573,
+      "grad_norm": 0.2437753677368164,
+      "learning_rate": 0.00017256756756756756,
+      "loss": 0.4199,
+      "step": 2839
+    },
+    {
+      "epoch": 0.8514146524264569,
+      "grad_norm": 0.26032543182373047,
+      "learning_rate": 0.0001725225225225225,
+      "loss": 0.4666,
+      "step": 2840
+    },
+    {
+      "epoch": 0.8517144463181563,
+      "grad_norm": 0.2443908005952835,
+      "learning_rate": 0.0001724774774774775,
+      "loss": 0.4109,
+      "step": 2841
+    },
+    {
+      "epoch": 0.8520142402098557,
+      "grad_norm": 0.25420573353767395,
+      "learning_rate": 0.00017243243243243242,
+      "loss": 0.4357,
+      "step": 2842
+    },
+    {
+      "epoch": 0.8523140341015552,
+      "grad_norm": 0.2527938783168793,
+      "learning_rate": 0.00017238738738738736,
+      "loss": 0.4299,
+      "step": 2843
+    },
+    {
+      "epoch": 0.8526138279932547,
+      "grad_norm": 0.24754786491394043,
+      "learning_rate": 0.00017234234234234235,
+      "loss": 0.4161,
+      "step": 2844
+    },
+    {
+      "epoch": 0.8529136218849541,
+      "grad_norm": 0.2448507398366928,
+      "learning_rate": 0.0001722972972972973,
+      "loss": 0.4238,
+      "step": 2845
+    },
+    {
+      "epoch": 0.8532134157766535,
+      "grad_norm": 0.2616099715232849,
+      "learning_rate": 0.00017225225225225223,
+      "loss": 0.4328,
+      "step": 2846
+    },
+    {
+      "epoch": 0.853513209668353,
+      "grad_norm": 0.263560950756073,
+      "learning_rate": 0.0001722072072072072,
+      "loss": 0.4204,
+      "step": 2847
+    },
+    {
+      "epoch": 0.8538130035600525,
+      "grad_norm": 0.25410616397857666,
+      "learning_rate": 0.00017216216216216215,
+      "loss": 0.4333,
+      "step": 2848
+    },
+    {
+      "epoch": 0.854112797451752,
+      "grad_norm": 0.2678585946559906,
+      "learning_rate": 0.0001721171171171171,
+      "loss": 0.4498,
+      "step": 2849
+    },
+    {
+      "epoch": 0.8544125913434514,
+      "grad_norm": 0.2663200795650482,
+      "learning_rate": 0.00017207207207207205,
+      "loss": 0.455,
+      "step": 2850
+    },
+    {
+      "epoch": 0.8547123852351508,
+      "grad_norm": 0.25283655524253845,
+      "learning_rate": 0.00017202702702702702,
+      "loss": 0.4167,
+      "step": 2851
+    },
+    {
+      "epoch": 0.8550121791268502,
+      "grad_norm": 0.25023430585861206,
+      "learning_rate": 0.00017198198198198195,
+      "loss": 0.3981,
+      "step": 2852
+    },
+    {
+      "epoch": 0.8553119730185498,
+      "grad_norm": 0.23520569503307343,
+      "learning_rate": 0.00017193693693693692,
+      "loss": 0.3895,
+      "step": 2853
+    },
+    {
+      "epoch": 0.8556117669102492,
+      "grad_norm": 0.26464787125587463,
+      "learning_rate": 0.00017189189189189188,
+      "loss": 0.4639,
+      "step": 2854
+    },
+    {
+      "epoch": 0.8559115608019486,
+      "grad_norm": 0.24785351753234863,
+      "learning_rate": 0.00017184684684684682,
+      "loss": 0.4202,
+      "step": 2855
+    },
+    {
+      "epoch": 0.8562113546936481,
+      "grad_norm": 0.2626986503601074,
+      "learning_rate": 0.00017180180180180178,
+      "loss": 0.4222,
+      "step": 2856
+    },
+    {
+      "epoch": 0.8565111485853476,
+      "grad_norm": 0.2554967403411865,
+      "learning_rate": 0.00017175675675675675,
+      "loss": 0.4182,
+      "step": 2857
+    },
+    {
+      "epoch": 0.856810942477047,
+      "grad_norm": 0.2597710192203522,
+      "learning_rate": 0.0001717117117117117,
+      "loss": 0.4296,
+      "step": 2858
+    },
+    {
+      "epoch": 0.8571107363687465,
+      "grad_norm": 0.26038360595703125,
+      "learning_rate": 0.00017166666666666665,
+      "loss": 0.4275,
+      "step": 2859
+    },
+    {
+      "epoch": 0.8574105302604459,
+      "grad_norm": 0.24397097527980804,
+      "learning_rate": 0.0001716216216216216,
+      "loss": 0.3991,
+      "step": 2860
+    },
+    {
+      "epoch": 0.8577103241521454,
+      "grad_norm": 0.2713565528392792,
+      "learning_rate": 0.00017157657657657657,
+      "loss": 0.4211,
+      "step": 2861
+    },
+    {
+      "epoch": 0.8580101180438449,
+      "grad_norm": 0.24685804545879364,
+      "learning_rate": 0.0001715315315315315,
+      "loss": 0.4008,
+      "step": 2862
+    },
+    {
+      "epoch": 0.8583099119355443,
+      "grad_norm": 0.2701510488986969,
+      "learning_rate": 0.00017148648648648648,
+      "loss": 0.422,
+      "step": 2863
+    },
+    {
+      "epoch": 0.8586097058272437,
+      "grad_norm": 0.2619568705558777,
+      "learning_rate": 0.00017144144144144144,
+      "loss": 0.4271,
+      "step": 2864
+    },
+    {
+      "epoch": 0.8589094997189433,
+      "grad_norm": 0.2503330707550049,
+      "learning_rate": 0.00017139639639639638,
+      "loss": 0.4517,
+      "step": 2865
+    },
+    {
+      "epoch": 0.8592092936106427,
+      "grad_norm": 0.2657839059829712,
+      "learning_rate": 0.00017135135135135134,
+      "loss": 0.431,
+      "step": 2866
+    },
+    {
+      "epoch": 0.8595090875023421,
+      "grad_norm": 0.24868468940258026,
+      "learning_rate": 0.0001713063063063063,
+      "loss": 0.415,
+      "step": 2867
+    },
+    {
+      "epoch": 0.8598088813940415,
+      "grad_norm": 0.2581307888031006,
+      "learning_rate": 0.00017126126126126124,
+      "loss": 0.4465,
+      "step": 2868
+    },
+    {
+      "epoch": 0.8601086752857411,
+      "grad_norm": 0.24836871027946472,
+      "learning_rate": 0.0001712162162162162,
+      "loss": 0.418,
+      "step": 2869
+    },
+    {
+      "epoch": 0.8604084691774405,
+      "grad_norm": 0.2541236877441406,
+      "learning_rate": 0.00017117117117117117,
+      "loss": 0.3917,
+      "step": 2870
+    },
+    {
+      "epoch": 0.86070826306914,
+      "grad_norm": 0.24935103952884674,
+      "learning_rate": 0.0001711261261261261,
+      "loss": 0.4002,
+      "step": 2871
+    },
+    {
+      "epoch": 0.8610080569608394,
+      "grad_norm": 0.2818658649921417,
+      "learning_rate": 0.00017108108108108104,
+      "loss": 0.441,
+      "step": 2872
+    },
+    {
+      "epoch": 0.8613078508525389,
+      "grad_norm": 0.2605019509792328,
+      "learning_rate": 0.00017103603603603603,
+      "loss": 0.4324,
+      "step": 2873
+    },
+    {
+      "epoch": 0.8616076447442383,
+      "grad_norm": 0.26078200340270996,
+      "learning_rate": 0.00017099099099099097,
+      "loss": 0.4206,
+      "step": 2874
+    },
+    {
+      "epoch": 0.8619074386359378,
+      "grad_norm": 0.26391538977622986,
+      "learning_rate": 0.0001709459459459459,
+      "loss": 0.3951,
+      "step": 2875
+    },
+    {
+      "epoch": 0.8622072325276372,
+      "grad_norm": 0.28473007678985596,
+      "learning_rate": 0.0001709009009009009,
+      "loss": 0.4655,
+      "step": 2876
+    },
+    {
+      "epoch": 0.8625070264193367,
+      "grad_norm": 0.24676088988780975,
+      "learning_rate": 0.00017085585585585583,
+      "loss": 0.4196,
+      "step": 2877
+    },
+    {
+      "epoch": 0.8628068203110362,
+      "grad_norm": 0.26333364844322205,
+      "learning_rate": 0.00017081081081081077,
+      "loss": 0.4204,
+      "step": 2878
+    },
+    {
+      "epoch": 0.8631066142027356,
+      "grad_norm": 0.24308471381664276,
+      "learning_rate": 0.00017076576576576576,
+      "loss": 0.4047,
+      "step": 2879
+    },
+    {
+      "epoch": 0.863406408094435,
+      "grad_norm": 0.24952733516693115,
+      "learning_rate": 0.0001707207207207207,
+      "loss": 0.417,
+      "step": 2880
+    },
+    {
+      "epoch": 0.8637062019861346,
+      "grad_norm": 0.26905426383018494,
+      "learning_rate": 0.00017067567567567566,
+      "loss": 0.432,
+      "step": 2881
+    },
+    {
+      "epoch": 0.864005995877834,
+      "grad_norm": 0.2668468654155731,
+      "learning_rate": 0.00017063063063063063,
+      "loss": 0.4278,
+      "step": 2882
+    },
+    {
+      "epoch": 0.8643057897695334,
+      "grad_norm": 0.2533245384693146,
+      "learning_rate": 0.00017058558558558556,
+      "loss": 0.4228,
+      "step": 2883
+    },
+    {
+      "epoch": 0.8646055836612329,
+      "grad_norm": 0.25289660692214966,
+      "learning_rate": 0.00017054054054054053,
+      "loss": 0.4136,
+      "step": 2884
+    },
+    {
+      "epoch": 0.8649053775529324,
+      "grad_norm": 0.25045090913772583,
+      "learning_rate": 0.0001704954954954955,
+      "loss": 0.425,
+      "step": 2885
+    },
+    {
+      "epoch": 0.8652051714446318,
+      "grad_norm": 0.2620987296104431,
+      "learning_rate": 0.00017045045045045043,
+      "loss": 0.4115,
+      "step": 2886
+    },
+    {
+      "epoch": 0.8655049653363313,
+      "grad_norm": 0.2613299489021301,
+      "learning_rate": 0.0001704054054054054,
+      "loss": 0.4311,
+      "step": 2887
+    },
+    {
+      "epoch": 0.8658047592280307,
+      "grad_norm": 0.25921955704689026,
+      "learning_rate": 0.00017036036036036036,
+      "loss": 0.4254,
+      "step": 2888
+    },
+    {
+      "epoch": 0.8661045531197302,
+      "grad_norm": 0.24337226152420044,
+      "learning_rate": 0.0001703153153153153,
+      "loss": 0.4282,
+      "step": 2889
+    },
+    {
+      "epoch": 0.8664043470114297,
+      "grad_norm": 0.25295788049697876,
+      "learning_rate": 0.00017027027027027026,
+      "loss": 0.3921,
+      "step": 2890
+    },
+    {
+      "epoch": 0.8667041409031291,
+      "grad_norm": 0.26155343651771545,
+      "learning_rate": 0.00017022522522522522,
+      "loss": 0.4379,
+      "step": 2891
+    },
+    {
+      "epoch": 0.8670039347948285,
+      "grad_norm": 0.25492122769355774,
+      "learning_rate": 0.00017018018018018016,
+      "loss": 0.4225,
+      "step": 2892
+    },
+    {
+      "epoch": 0.867303728686528,
+      "grad_norm": 0.27565327286720276,
+      "learning_rate": 0.00017013513513513512,
+      "loss": 0.4254,
+      "step": 2893
+    },
+    {
+      "epoch": 0.8676035225782275,
+      "grad_norm": 0.24523693323135376,
+      "learning_rate": 0.00017009009009009006,
+      "loss": 0.4103,
+      "step": 2894
+    },
+    {
+      "epoch": 0.8679033164699269,
+      "grad_norm": 0.29129156470298767,
+      "learning_rate": 0.00017004504504504505,
+      "loss": 0.4366,
+      "step": 2895
+    },
+    {
+      "epoch": 0.8682031103616263,
+      "grad_norm": 0.2691640555858612,
+      "learning_rate": 0.00016999999999999999,
+      "loss": 0.4152,
+      "step": 2896
+    },
+    {
+      "epoch": 0.8685029042533259,
+      "grad_norm": 0.2639869153499603,
+      "learning_rate": 0.00016995495495495492,
+      "loss": 0.3987,
+      "step": 2897
+    },
+    {
+      "epoch": 0.8688026981450253,
+      "grad_norm": 0.254861056804657,
+      "learning_rate": 0.0001699099099099099,
+      "loss": 0.387,
+      "step": 2898
+    },
+    {
+      "epoch": 0.8691024920367247,
+      "grad_norm": 0.2625315189361572,
+      "learning_rate": 0.00016986486486486485,
+      "loss": 0.4306,
+      "step": 2899
+    },
+    {
+      "epoch": 0.8694022859284242,
+      "grad_norm": 0.26464176177978516,
+      "learning_rate": 0.00016981981981981979,
+      "loss": 0.4035,
+      "step": 2900
+    },
+    {
+      "epoch": 0.8697020798201237,
+      "grad_norm": 0.2657167613506317,
+      "learning_rate": 0.00016977477477477478,
+      "loss": 0.4321,
+      "step": 2901
+    },
+    {
+      "epoch": 0.8700018737118231,
+      "grad_norm": 0.268087238073349,
+      "learning_rate": 0.00016972972972972971,
+      "loss": 0.4187,
+      "step": 2902
+    },
+    {
+      "epoch": 0.8703016676035226,
+      "grad_norm": 0.2585284411907196,
+      "learning_rate": 0.00016968468468468465,
+      "loss": 0.4213,
+      "step": 2903
+    },
+    {
+      "epoch": 0.870601461495222,
+      "grad_norm": 0.26495012640953064,
+      "learning_rate": 0.00016963963963963964,
+      "loss": 0.42,
+      "step": 2904
+    },
+    {
+      "epoch": 0.8709012553869215,
+      "grad_norm": 0.27982714772224426,
+      "learning_rate": 0.00016959459459459458,
+      "loss": 0.402,
+      "step": 2905
+    },
+    {
+      "epoch": 0.871201049278621,
+      "grad_norm": 0.2580246925354004,
+      "learning_rate": 0.00016954954954954952,
+      "loss": 0.4028,
+      "step": 2906
+    },
+    {
+      "epoch": 0.8715008431703204,
+      "grad_norm": 0.254881888628006,
+      "learning_rate": 0.0001695045045045045,
+      "loss": 0.4186,
+      "step": 2907
+    },
+    {
+      "epoch": 0.8718006370620198,
+      "grad_norm": 0.26736170053482056,
+      "learning_rate": 0.00016945945945945944,
+      "loss": 0.4322,
+      "step": 2908
+    },
+    {
+      "epoch": 0.8721004309537194,
+      "grad_norm": 0.27539244294166565,
+      "learning_rate": 0.00016941441441441438,
+      "loss": 0.4362,
+      "step": 2909
+    },
+    {
+      "epoch": 0.8724002248454188,
+      "grad_norm": 0.2490614801645279,
+      "learning_rate": 0.00016936936936936937,
+      "loss": 0.4193,
+      "step": 2910
+    },
+    {
+      "epoch": 0.8727000187371182,
+      "grad_norm": 0.25294286012649536,
+      "learning_rate": 0.0001693243243243243,
+      "loss": 0.4202,
+      "step": 2911
+    },
+    {
+      "epoch": 0.8729998126288177,
+      "grad_norm": 0.2698332965373993,
+      "learning_rate": 0.00016927927927927924,
+      "loss": 0.4393,
+      "step": 2912
+    },
+    {
+      "epoch": 0.8732996065205172,
+      "grad_norm": 0.2698095738887787,
+      "learning_rate": 0.00016923423423423424,
+      "loss": 0.4265,
+      "step": 2913
+    },
+    {
+      "epoch": 0.8735994004122166,
+      "grad_norm": 0.25060606002807617,
+      "learning_rate": 0.00016918918918918917,
+      "loss": 0.4201,
+      "step": 2914
+    },
+    {
+      "epoch": 0.873899194303916,
+      "grad_norm": 0.2536526918411255,
+      "learning_rate": 0.00016914414414414414,
+      "loss": 0.4038,
+      "step": 2915
+    },
+    {
+      "epoch": 0.8741989881956155,
+      "grad_norm": 0.25597628951072693,
+      "learning_rate": 0.0001690990990990991,
+      "loss": 0.422,
+      "step": 2916
+    },
+    {
+      "epoch": 0.874498782087315,
+      "grad_norm": 0.25855180621147156,
+      "learning_rate": 0.00016905405405405404,
+      "loss": 0.4017,
+      "step": 2917
+    },
+    {
+      "epoch": 0.8747985759790144,
+      "grad_norm": 0.23959973454475403,
+      "learning_rate": 0.000169009009009009,
+      "loss": 0.3882,
+      "step": 2918
+    },
+    {
+      "epoch": 0.8750983698707139,
+      "grad_norm": 0.24938194453716278,
+      "learning_rate": 0.00016896396396396394,
+      "loss": 0.435,
+      "step": 2919
+    },
+    {
+      "epoch": 0.8753981637624133,
+      "grad_norm": 0.268046110868454,
+      "learning_rate": 0.0001689189189189189,
+      "loss": 0.4323,
+      "step": 2920
+    },
+    {
+      "epoch": 0.8756979576541128,
+      "grad_norm": 0.28519049286842346,
+      "learning_rate": 0.00016887387387387387,
+      "loss": 0.446,
+      "step": 2921
+    },
+    {
+      "epoch": 0.8759977515458123,
+      "grad_norm": 0.2485249787569046,
+      "learning_rate": 0.0001688288288288288,
+      "loss": 0.4328,
+      "step": 2922
+    },
+    {
+      "epoch": 0.8762975454375117,
+      "grad_norm": 0.28231462836265564,
+      "learning_rate": 0.00016878378378378377,
+      "loss": 0.4201,
+      "step": 2923
+    },
+    {
+      "epoch": 0.8765973393292111,
+      "grad_norm": 0.2693428695201874,
+      "learning_rate": 0.00016873873873873873,
+      "loss": 0.451,
+      "step": 2924
+    },
+    {
+      "epoch": 0.8768971332209107,
+      "grad_norm": 0.2595715820789337,
+      "learning_rate": 0.00016869369369369367,
+      "loss": 0.4319,
+      "step": 2925
+    },
+    {
+      "epoch": 0.8771969271126101,
+      "grad_norm": 0.2597898840904236,
+      "learning_rate": 0.00016864864864864863,
+      "loss": 0.4056,
+      "step": 2926
+    },
+    {
+      "epoch": 0.8774967210043095,
+      "grad_norm": 0.2534911632537842,
+      "learning_rate": 0.0001686036036036036,
+      "loss": 0.4325,
+      "step": 2927
+    },
+    {
+      "epoch": 0.877796514896009,
+      "grad_norm": 0.2742319107055664,
+      "learning_rate": 0.00016855855855855853,
+      "loss": 0.456,
+      "step": 2928
+    },
+    {
+      "epoch": 0.8780963087877085,
+      "grad_norm": 0.25808632373809814,
+      "learning_rate": 0.00016851351351351352,
+      "loss": 0.4419,
+      "step": 2929
+    },
+    {
+      "epoch": 0.8783961026794079,
+      "grad_norm": 0.2860279083251953,
+      "learning_rate": 0.00016846846846846846,
+      "loss": 0.4534,
+      "step": 2930
+    },
+    {
+      "epoch": 0.8786958965711074,
+      "grad_norm": 0.26722362637519836,
+      "learning_rate": 0.0001684234234234234,
+      "loss": 0.445,
+      "step": 2931
+    },
+    {
+      "epoch": 0.8789956904628068,
+      "grad_norm": 0.2761351764202118,
+      "learning_rate": 0.00016837837837837839,
+      "loss": 0.4175,
+      "step": 2932
+    },
+    {
+      "epoch": 0.8792954843545063,
+      "grad_norm": 0.2346809059381485,
+      "learning_rate": 0.00016833333333333332,
+      "loss": 0.4107,
+      "step": 2933
+    },
+    {
+      "epoch": 0.8795952782462058,
+      "grad_norm": 0.24212133884429932,
+      "learning_rate": 0.00016828828828828826,
+      "loss": 0.393,
+      "step": 2934
+    },
+    {
+      "epoch": 0.8798950721379052,
+      "grad_norm": 0.252184122800827,
+      "learning_rate": 0.00016824324324324325,
+      "loss": 0.4069,
+      "step": 2935
+    },
+    {
+      "epoch": 0.8801948660296046,
+      "grad_norm": 0.26833096146583557,
+      "learning_rate": 0.0001681981981981982,
+      "loss": 0.4455,
+      "step": 2936
+    },
+    {
+      "epoch": 0.8804946599213042,
+      "grad_norm": 0.2804100811481476,
+      "learning_rate": 0.00016815315315315312,
+      "loss": 0.4348,
+      "step": 2937
+    },
+    {
+      "epoch": 0.8807944538130036,
+      "grad_norm": 0.2611362338066101,
+      "learning_rate": 0.00016810810810810812,
+      "loss": 0.4407,
+      "step": 2938
+    },
+    {
+      "epoch": 0.881094247704703,
+      "grad_norm": 0.2497948855161667,
+      "learning_rate": 0.00016806306306306305,
+      "loss": 0.437,
+      "step": 2939
+    },
+    {
+      "epoch": 0.8813940415964024,
+      "grad_norm": 0.2518768906593323,
+      "learning_rate": 0.000168018018018018,
+      "loss": 0.4279,
+      "step": 2940
+    },
+    {
+      "epoch": 0.881693835488102,
+      "grad_norm": 0.2658750116825104,
+      "learning_rate": 0.00016797297297297298,
+      "loss": 0.4444,
+      "step": 2941
+    },
+    {
+      "epoch": 0.8819936293798014,
+      "grad_norm": 0.25899848341941833,
+      "learning_rate": 0.00016792792792792792,
+      "loss": 0.4145,
+      "step": 2942
+    },
+    {
+      "epoch": 0.8822934232715008,
+      "grad_norm": 0.2500343322753906,
+      "learning_rate": 0.00016788288288288285,
+      "loss": 0.4142,
+      "step": 2943
+    },
+    {
+      "epoch": 0.8825932171632003,
+      "grad_norm": 0.2614184319972992,
+      "learning_rate": 0.00016783783783783782,
+      "loss": 0.4125,
+      "step": 2944
+    },
+    {
+      "epoch": 0.8828930110548997,
+      "grad_norm": 0.2568380832672119,
+      "learning_rate": 0.00016779279279279278,
+      "loss": 0.4184,
+      "step": 2945
+    },
+    {
+      "epoch": 0.8831928049465992,
+      "grad_norm": 0.2491275817155838,
+      "learning_rate": 0.00016774774774774772,
+      "loss": 0.4386,
+      "step": 2946
+    },
+    {
+      "epoch": 0.8834925988382987,
+      "grad_norm": 0.2676237225532532,
+      "learning_rate": 0.00016770270270270268,
+      "loss": 0.4395,
+      "step": 2947
+    },
+    {
+      "epoch": 0.8837923927299981,
+      "grad_norm": 0.26965218782424927,
+      "learning_rate": 0.00016765765765765765,
+      "loss": 0.4235,
+      "step": 2948
+    },
+    {
+      "epoch": 0.8840921866216975,
+      "grad_norm": 0.26237690448760986,
+      "learning_rate": 0.00016761261261261258,
+      "loss": 0.4246,
+      "step": 2949
+    },
+    {
+      "epoch": 0.8843919805133971,
+      "grad_norm": 0.2397497445344925,
+      "learning_rate": 0.00016756756756756755,
+      "loss": 0.4092,
+      "step": 2950
+    },
+    {
+      "epoch": 0.8846917744050965,
+      "grad_norm": 0.25894200801849365,
+      "learning_rate": 0.0001675225225225225,
+      "loss": 0.4355,
+      "step": 2951
+    },
+    {
+      "epoch": 0.8849915682967959,
+      "grad_norm": 0.24452635645866394,
+      "learning_rate": 0.00016747747747747747,
+      "loss": 0.4062,
+      "step": 2952
+    },
+    {
+      "epoch": 0.8852913621884954,
+      "grad_norm": 0.25013741850852966,
+      "learning_rate": 0.0001674324324324324,
+      "loss": 0.4195,
+      "step": 2953
+    },
+    {
+      "epoch": 0.8855911560801949,
+      "grad_norm": 0.2652290463447571,
+      "learning_rate": 0.00016738738738738737,
+      "loss": 0.4442,
+      "step": 2954
+    },
+    {
+      "epoch": 0.8858909499718943,
+      "grad_norm": 0.25907742977142334,
+      "learning_rate": 0.00016734234234234234,
+      "loss": 0.4353,
+      "step": 2955
+    },
+    {
+      "epoch": 0.8861907438635938,
+      "grad_norm": 0.26591479778289795,
+      "learning_rate": 0.00016729729729729728,
+      "loss": 0.4427,
+      "step": 2956
+    },
+    {
+      "epoch": 0.8864905377552932,
+      "grad_norm": 0.25649645924568176,
+      "learning_rate": 0.00016725225225225224,
+      "loss": 0.4194,
+      "step": 2957
+    },
+    {
+      "epoch": 0.8867903316469927,
+      "grad_norm": 0.23913215100765228,
+      "learning_rate": 0.0001672072072072072,
+      "loss": 0.4114,
+      "step": 2958
+    },
+    {
+      "epoch": 0.8870901255386922,
+      "grad_norm": 0.2712022364139557,
+      "learning_rate": 0.00016716216216216214,
+      "loss": 0.4358,
+      "step": 2959
+    },
+    {
+      "epoch": 0.8873899194303916,
+      "grad_norm": 0.2589244544506073,
+      "learning_rate": 0.0001671171171171171,
+      "loss": 0.4325,
+      "step": 2960
+    },
+    {
+      "epoch": 0.887689713322091,
+      "grad_norm": 0.2385113388299942,
+      "learning_rate": 0.00016707207207207207,
+      "loss": 0.4057,
+      "step": 2961
+    },
+    {
+      "epoch": 0.8879895072137906,
+      "grad_norm": 0.2598143219947815,
+      "learning_rate": 0.000167027027027027,
+      "loss": 0.436,
+      "step": 2962
+    },
+    {
+      "epoch": 0.88828930110549,
+      "grad_norm": 0.24032726883888245,
+      "learning_rate": 0.000166981981981982,
+      "loss": 0.3991,
+      "step": 2963
+    },
+    {
+      "epoch": 0.8885890949971894,
+      "grad_norm": 0.25453847646713257,
+      "learning_rate": 0.00016693693693693693,
+      "loss": 0.3913,
+      "step": 2964
+    },
+    {
+      "epoch": 0.8888888888888888,
+      "grad_norm": 0.24269893765449524,
+      "learning_rate": 0.00016689189189189187,
+      "loss": 0.4134,
+      "step": 2965
+    },
+    {
+      "epoch": 0.8891886827805884,
+      "grad_norm": 0.25419285893440247,
+      "learning_rate": 0.0001668468468468468,
+      "loss": 0.4218,
+      "step": 2966
+    },
+    {
+      "epoch": 0.8894884766722878,
+      "grad_norm": 0.2587622106075287,
+      "learning_rate": 0.0001668018018018018,
+      "loss": 0.4269,
+      "step": 2967
+    },
+    {
+      "epoch": 0.8897882705639872,
+      "grad_norm": 0.25632867217063904,
+      "learning_rate": 0.00016675675675675673,
+      "loss": 0.4379,
+      "step": 2968
+    },
+    {
+      "epoch": 0.8900880644556867,
+      "grad_norm": 0.2640867531299591,
+      "learning_rate": 0.00016671171171171167,
+      "loss": 0.4345,
+      "step": 2969
+    },
+    {
+      "epoch": 0.8903878583473862,
+      "grad_norm": 0.24803832173347473,
+      "learning_rate": 0.00016666666666666666,
+      "loss": 0.4126,
+      "step": 2970
+    },
+    {
+      "epoch": 0.8906876522390856,
+      "grad_norm": 0.24913029372692108,
+      "learning_rate": 0.0001666216216216216,
+      "loss": 0.3953,
+      "step": 2971
+    },
+    {
+      "epoch": 0.8909874461307851,
+      "grad_norm": 0.2572920620441437,
+      "learning_rate": 0.00016657657657657653,
+      "loss": 0.4184,
+      "step": 2972
+    },
+    {
+      "epoch": 0.8912872400224845,
+      "grad_norm": 0.26253217458724976,
+      "learning_rate": 0.00016653153153153153,
+      "loss": 0.4509,
+      "step": 2973
+    },
+    {
+      "epoch": 0.891587033914184,
+      "grad_norm": 0.24650625884532928,
+      "learning_rate": 0.00016648648648648646,
+      "loss": 0.4371,
+      "step": 2974
+    },
+    {
+      "epoch": 0.8918868278058835,
+      "grad_norm": 0.26514360308647156,
+      "learning_rate": 0.00016644144144144143,
+      "loss": 0.4547,
+      "step": 2975
+    },
+    {
+      "epoch": 0.8921866216975829,
+      "grad_norm": 0.2522089183330536,
+      "learning_rate": 0.0001663963963963964,
+      "loss": 0.4141,
+      "step": 2976
+    },
+    {
+      "epoch": 0.8924864155892823,
+      "grad_norm": 0.2426896095275879,
+      "learning_rate": 0.00016635135135135133,
+      "loss": 0.4124,
+      "step": 2977
+    },
+    {
+      "epoch": 0.8927862094809819,
+      "grad_norm": 0.24892792105674744,
+      "learning_rate": 0.0001663063063063063,
+      "loss": 0.3971,
+      "step": 2978
+    },
+    {
+      "epoch": 0.8930860033726813,
+      "grad_norm": 0.25551697611808777,
+      "learning_rate": 0.00016626126126126125,
+      "loss": 0.4191,
+      "step": 2979
+    },
+    {
+      "epoch": 0.8933857972643807,
+      "grad_norm": 0.2599818706512451,
+      "learning_rate": 0.0001662162162162162,
+      "loss": 0.4616,
+      "step": 2980
+    },
+    {
+      "epoch": 0.8936855911560802,
+      "grad_norm": 0.24369800090789795,
+      "learning_rate": 0.00016617117117117116,
+      "loss": 0.4092,
+      "step": 2981
+    },
+    {
+      "epoch": 0.8939853850477797,
+      "grad_norm": 0.24979013204574585,
+      "learning_rate": 0.00016612612612612612,
+      "loss": 0.4184,
+      "step": 2982
+    },
+    {
+      "epoch": 0.8942851789394791,
+      "grad_norm": 0.24617372453212738,
+      "learning_rate": 0.00016608108108108106,
+      "loss": 0.4022,
+      "step": 2983
+    },
+    {
+      "epoch": 0.8945849728311785,
+      "grad_norm": 0.25377964973449707,
+      "learning_rate": 0.00016603603603603602,
+      "loss": 0.4246,
+      "step": 2984
+    },
+    {
+      "epoch": 0.894884766722878,
+      "grad_norm": 0.25180289149284363,
+      "learning_rate": 0.00016599099099099098,
+      "loss": 0.4262,
+      "step": 2985
+    },
+    {
+      "epoch": 0.8951845606145775,
+      "grad_norm": 0.26268377900123596,
+      "learning_rate": 0.00016594594594594595,
+      "loss": 0.4083,
+      "step": 2986
+    },
+    {
+      "epoch": 0.895484354506277,
+      "grad_norm": 0.2539224624633789,
+      "learning_rate": 0.00016590090090090088,
+      "loss": 0.4328,
+      "step": 2987
+    },
+    {
+      "epoch": 0.8957841483979764,
+      "grad_norm": 0.2591584324836731,
+      "learning_rate": 0.00016585585585585585,
+      "loss": 0.4312,
+      "step": 2988
+    },
+    {
+      "epoch": 0.8960839422896758,
+      "grad_norm": 0.246408611536026,
+      "learning_rate": 0.0001658108108108108,
+      "loss": 0.3991,
+      "step": 2989
+    },
+    {
+      "epoch": 0.8963837361813753,
+      "grad_norm": 0.25761210918426514,
+      "learning_rate": 0.00016576576576576575,
+      "loss": 0.4225,
+      "step": 2990
+    },
+    {
+      "epoch": 0.8966835300730748,
+      "grad_norm": 0.2380562722682953,
+      "learning_rate": 0.00016572072072072069,
+      "loss": 0.4174,
+      "step": 2991
+    },
+    {
+      "epoch": 0.8969833239647742,
+      "grad_norm": 0.2515964210033417,
+      "learning_rate": 0.00016567567567567568,
+      "loss": 0.4347,
+      "step": 2992
+    },
+    {
+      "epoch": 0.8972831178564736,
+      "grad_norm": 0.24080772697925568,
+      "learning_rate": 0.0001656306306306306,
+      "loss": 0.4331,
+      "step": 2993
+    },
+    {
+      "epoch": 0.8975829117481732,
+      "grad_norm": 0.25413286685943604,
+      "learning_rate": 0.00016558558558558555,
+      "loss": 0.4101,
+      "step": 2994
+    },
+    {
+      "epoch": 0.8978827056398726,
+      "grad_norm": 0.2696084678173065,
+      "learning_rate": 0.00016554054054054054,
+      "loss": 0.4446,
+      "step": 2995
+    },
+    {
+      "epoch": 0.898182499531572,
+      "grad_norm": 0.24439360201358795,
+      "learning_rate": 0.00016549549549549548,
+      "loss": 0.4136,
+      "step": 2996
+    },
+    {
+      "epoch": 0.8984822934232715,
+      "grad_norm": 0.25744253396987915,
+      "learning_rate": 0.00016545045045045041,
+      "loss": 0.4321,
+      "step": 2997
+    },
+    {
+      "epoch": 0.898782087314971,
+      "grad_norm": 0.26146844029426575,
+      "learning_rate": 0.0001654054054054054,
+      "loss": 0.4214,
+      "step": 2998
+    },
+    {
+      "epoch": 0.8990818812066704,
+      "grad_norm": 0.24922147393226624,
+      "learning_rate": 0.00016536036036036034,
+      "loss": 0.3922,
+      "step": 2999
+    },
+    {
+      "epoch": 0.8993816750983699,
+      "grad_norm": 0.24045661091804504,
+      "learning_rate": 0.00016531531531531528,
+      "loss": 0.423,
+      "step": 3000
+    },
+    {
+      "epoch": 0.8993816750983699,
+      "eval_loss": 0.4257502257823944,
+      "eval_runtime": 565.3599,
+      "eval_samples_per_second": 3.819,
+      "eval_steps_per_second": 0.478,
+      "step": 3000
+    },
+    {
+      "epoch": 0.8996814689900693,
+      "grad_norm": 0.2597513198852539,
+      "learning_rate": 0.00016527027027027027,
+      "loss": 0.4506,
+      "step": 3001
+    },
+    {
+      "epoch": 0.8999812628817688,
+      "grad_norm": 0.24458399415016174,
+      "learning_rate": 0.0001652252252252252,
+      "loss": 0.419,
+      "step": 3002
+    },
+    {
+      "epoch": 0.9002810567734683,
+      "grad_norm": 0.24708233773708344,
+      "learning_rate": 0.00016518018018018014,
+      "loss": 0.4016,
+      "step": 3003
+    },
+    {
+      "epoch": 0.9005808506651677,
+      "grad_norm": 0.25798776745796204,
+      "learning_rate": 0.00016513513513513513,
+      "loss": 0.4394,
+      "step": 3004
+    },
+    {
+      "epoch": 0.9008806445568671,
+      "grad_norm": 0.27492186427116394,
+      "learning_rate": 0.00016509009009009007,
+      "loss": 0.4262,
+      "step": 3005
+    },
+    {
+      "epoch": 0.9011804384485667,
+      "grad_norm": 0.2706008553504944,
+      "learning_rate": 0.000165045045045045,
+      "loss": 0.4416,
+      "step": 3006
+    },
+    {
+      "epoch": 0.9014802323402661,
+      "grad_norm": 0.269711971282959,
+      "learning_rate": 0.000165,
+      "loss": 0.4568,
+      "step": 3007
+    },
+    {
+      "epoch": 0.9017800262319655,
+      "grad_norm": 0.2955131530761719,
+      "learning_rate": 0.00016495495495495494,
+      "loss": 0.4043,
+      "step": 3008
+    },
+    {
+      "epoch": 0.9020798201236649,
+      "grad_norm": 0.2553322911262512,
+      "learning_rate": 0.0001649099099099099,
+      "loss": 0.4135,
+      "step": 3009
+    },
+    {
+      "epoch": 0.9023796140153645,
+      "grad_norm": 0.2554074227809906,
+      "learning_rate": 0.00016486486486486486,
+      "loss": 0.4277,
+      "step": 3010
+    },
+    {
+      "epoch": 0.9026794079070639,
+      "grad_norm": 0.2844032645225525,
+      "learning_rate": 0.0001648198198198198,
+      "loss": 0.4327,
+      "step": 3011
+    },
+    {
+      "epoch": 0.9029792017987633,
+      "grad_norm": 0.28705543279647827,
+      "learning_rate": 0.00016477477477477476,
+      "loss": 0.426,
+      "step": 3012
+    },
+    {
+      "epoch": 0.9032789956904628,
+      "grad_norm": 0.2566993832588196,
+      "learning_rate": 0.00016472972972972973,
+      "loss": 0.4094,
+      "step": 3013
+    },
+    {
+      "epoch": 0.9035787895821623,
+      "grad_norm": 0.24438917636871338,
+      "learning_rate": 0.00016468468468468466,
+      "loss": 0.4009,
+      "step": 3014
+    },
+    {
+      "epoch": 0.9038785834738617,
+      "grad_norm": 0.2831672430038452,
+      "learning_rate": 0.00016463963963963963,
+      "loss": 0.4533,
+      "step": 3015
+    },
+    {
+      "epoch": 0.9041783773655612,
+      "grad_norm": 0.24414128065109253,
+      "learning_rate": 0.00016459459459459457,
+      "loss": 0.4161,
+      "step": 3016
+    },
+    {
+      "epoch": 0.9044781712572606,
+      "grad_norm": 0.27103421092033386,
+      "learning_rate": 0.00016454954954954953,
+      "loss": 0.396,
+      "step": 3017
+    },
+    {
+      "epoch": 0.9047779651489601,
+      "grad_norm": 0.2692296504974365,
+      "learning_rate": 0.0001645045045045045,
+      "loss": 0.4136,
+      "step": 3018
+    },
+    {
+      "epoch": 0.9050777590406596,
+      "grad_norm": 0.267905056476593,
+      "learning_rate": 0.00016445945945945943,
+      "loss": 0.4126,
+      "step": 3019
+    },
+    {
+      "epoch": 0.905377552932359,
+      "grad_norm": 0.286996990442276,
+      "learning_rate": 0.00016441441441441442,
+      "loss": 0.4518,
+      "step": 3020
+    },
+    {
+      "epoch": 0.9056773468240584,
+      "grad_norm": 0.2529492676258087,
+      "learning_rate": 0.00016436936936936936,
+      "loss": 0.422,
+      "step": 3021
+    },
+    {
+      "epoch": 0.905977140715758,
+      "grad_norm": 0.23985780775547028,
+      "learning_rate": 0.0001643243243243243,
+      "loss": 0.3948,
+      "step": 3022
+    },
+    {
+      "epoch": 0.9062769346074574,
+      "grad_norm": 0.2592794895172119,
+      "learning_rate": 0.00016427927927927929,
+      "loss": 0.426,
+      "step": 3023
+    },
+    {
+      "epoch": 0.9065767284991568,
+      "grad_norm": 0.2730545401573181,
+      "learning_rate": 0.00016423423423423422,
+      "loss": 0.4524,
+      "step": 3024
+    },
+    {
+      "epoch": 0.9068765223908563,
+      "grad_norm": 0.2558513581752777,
+      "learning_rate": 0.00016418918918918916,
+      "loss": 0.4151,
+      "step": 3025
+    },
+    {
+      "epoch": 0.9071763162825558,
+      "grad_norm": 0.26334816217422485,
+      "learning_rate": 0.00016414414414414415,
+      "loss": 0.4069,
+      "step": 3026
+    },
+    {
+      "epoch": 0.9074761101742552,
+      "grad_norm": 0.25048863887786865,
+      "learning_rate": 0.0001640990990990991,
+      "loss": 0.4251,
+      "step": 3027
+    },
+    {
+      "epoch": 0.9077759040659547,
+      "grad_norm": 0.24627207219600677,
+      "learning_rate": 0.00016405405405405402,
+      "loss": 0.4181,
+      "step": 3028
+    },
+    {
+      "epoch": 0.9080756979576541,
+      "grad_norm": 0.2565348744392395,
+      "learning_rate": 0.00016400900900900901,
+      "loss": 0.4088,
+      "step": 3029
+    },
+    {
+      "epoch": 0.9083754918493536,
+      "grad_norm": 0.23633405566215515,
+      "learning_rate": 0.00016396396396396395,
+      "loss": 0.3953,
+      "step": 3030
+    },
+    {
+      "epoch": 0.908675285741053,
+      "grad_norm": 0.25836828351020813,
+      "learning_rate": 0.0001639189189189189,
+      "loss": 0.4177,
+      "step": 3031
+    },
+    {
+      "epoch": 0.9089750796327525,
+      "grad_norm": 0.2472517192363739,
+      "learning_rate": 0.00016387387387387388,
+      "loss": 0.401,
+      "step": 3032
+    },
+    {
+      "epoch": 0.9092748735244519,
+      "grad_norm": 0.2629941403865814,
+      "learning_rate": 0.00016382882882882882,
+      "loss": 0.4367,
+      "step": 3033
+    },
+    {
+      "epoch": 0.9095746674161515,
+      "grad_norm": 0.2588827908039093,
+      "learning_rate": 0.00016378378378378375,
+      "loss": 0.4569,
+      "step": 3034
+    },
+    {
+      "epoch": 0.9098744613078509,
+      "grad_norm": 0.2503022253513336,
+      "learning_rate": 0.00016373873873873874,
+      "loss": 0.4193,
+      "step": 3035
+    },
+    {
+      "epoch": 0.9101742551995503,
+      "grad_norm": 0.2534733712673187,
+      "learning_rate": 0.00016369369369369368,
+      "loss": 0.4277,
+      "step": 3036
+    },
+    {
+      "epoch": 0.9104740490912497,
+      "grad_norm": 0.2703090310096741,
+      "learning_rate": 0.00016364864864864862,
+      "loss": 0.4363,
+      "step": 3037
+    },
+    {
+      "epoch": 0.9107738429829493,
+      "grad_norm": 0.25499293208122253,
+      "learning_rate": 0.0001636036036036036,
+      "loss": 0.4213,
+      "step": 3038
+    },
+    {
+      "epoch": 0.9110736368746487,
+      "grad_norm": 0.25429412722587585,
+      "learning_rate": 0.00016355855855855854,
+      "loss": 0.4242,
+      "step": 3039
+    },
+    {
+      "epoch": 0.9113734307663481,
+      "grad_norm": 0.2607458829879761,
+      "learning_rate": 0.00016351351351351348,
+      "loss": 0.4148,
+      "step": 3040
+    },
+    {
+      "epoch": 0.9116732246580476,
+      "grad_norm": 0.2532136142253876,
+      "learning_rate": 0.00016346846846846845,
+      "loss": 0.4274,
+      "step": 3041
+    },
+    {
+      "epoch": 0.911973018549747,
+      "grad_norm": 0.25032955408096313,
+      "learning_rate": 0.0001634234234234234,
+      "loss": 0.4325,
+      "step": 3042
+    },
+    {
+      "epoch": 0.9122728124414465,
+      "grad_norm": 0.23373675346374512,
+      "learning_rate": 0.00016337837837837837,
+      "loss": 0.4207,
+      "step": 3043
+    },
+    {
+      "epoch": 0.912572606333146,
+      "grad_norm": 0.2524081766605377,
+      "learning_rate": 0.0001633333333333333,
+      "loss": 0.4308,
+      "step": 3044
+    },
+    {
+      "epoch": 0.9128724002248454,
+      "grad_norm": 0.2704564034938812,
+      "learning_rate": 0.00016328828828828827,
+      "loss": 0.4296,
+      "step": 3045
+    },
+    {
+      "epoch": 0.9131721941165448,
+      "grad_norm": 0.241916686296463,
+      "learning_rate": 0.00016324324324324324,
+      "loss": 0.4113,
+      "step": 3046
+    },
+    {
+      "epoch": 0.9134719880082444,
+      "grad_norm": 0.29074499011039734,
+      "learning_rate": 0.00016319819819819817,
+      "loss": 0.4772,
+      "step": 3047
+    },
+    {
+      "epoch": 0.9137717818999438,
+      "grad_norm": 0.2655307948589325,
+      "learning_rate": 0.00016315315315315314,
+      "loss": 0.4378,
+      "step": 3048
+    },
+    {
+      "epoch": 0.9140715757916432,
+      "grad_norm": 0.23870833218097687,
+      "learning_rate": 0.0001631081081081081,
+      "loss": 0.3933,
+      "step": 3049
+    },
+    {
+      "epoch": 0.9143713696833426,
+      "grad_norm": 0.28048643469810486,
+      "learning_rate": 0.00016306306306306304,
+      "loss": 0.4366,
+      "step": 3050
+    },
+    {
+      "epoch": 0.9146711635750422,
+      "grad_norm": 0.2783832550048828,
+      "learning_rate": 0.000163018018018018,
+      "loss": 0.4435,
+      "step": 3051
+    },
+    {
+      "epoch": 0.9149709574667416,
+      "grad_norm": 0.25969600677490234,
+      "learning_rate": 0.00016297297297297297,
+      "loss": 0.4226,
+      "step": 3052
+    },
+    {
+      "epoch": 0.915270751358441,
+      "grad_norm": 0.2698688209056854,
+      "learning_rate": 0.0001629279279279279,
+      "loss": 0.4153,
+      "step": 3053
+    },
+    {
+      "epoch": 0.9155705452501405,
+      "grad_norm": 0.26543521881103516,
+      "learning_rate": 0.0001628828828828829,
+      "loss": 0.4439,
+      "step": 3054
+    },
+    {
+      "epoch": 0.91587033914184,
+      "grad_norm": 0.2702194154262543,
+      "learning_rate": 0.00016283783783783783,
+      "loss": 0.4416,
+      "step": 3055
+    },
+    {
+      "epoch": 0.9161701330335394,
+      "grad_norm": 0.2544569671154022,
+      "learning_rate": 0.00016279279279279277,
+      "loss": 0.4245,
+      "step": 3056
+    },
+    {
+      "epoch": 0.9164699269252389,
+      "grad_norm": 0.2640305161476135,
+      "learning_rate": 0.00016274774774774776,
+      "loss": 0.4342,
+      "step": 3057
+    },
+    {
+      "epoch": 0.9167697208169383,
+      "grad_norm": 0.26513299345970154,
+      "learning_rate": 0.0001627027027027027,
+      "loss": 0.448,
+      "step": 3058
+    },
+    {
+      "epoch": 0.9170695147086378,
+      "grad_norm": 0.25756070017814636,
+      "learning_rate": 0.00016265765765765763,
+      "loss": 0.4196,
+      "step": 3059
+    },
+    {
+      "epoch": 0.9173693086003373,
+      "grad_norm": 0.26089876890182495,
+      "learning_rate": 0.00016261261261261262,
+      "loss": 0.359,
+      "step": 3060
+    },
+    {
+      "epoch": 0.9176691024920367,
+      "grad_norm": 0.28472572565078735,
+      "learning_rate": 0.00016256756756756756,
+      "loss": 0.454,
+      "step": 3061
+    },
+    {
+      "epoch": 0.9179688963837361,
+      "grad_norm": 0.26528477668762207,
+      "learning_rate": 0.0001625225225225225,
+      "loss": 0.4422,
+      "step": 3062
+    },
+    {
+      "epoch": 0.9182686902754357,
+      "grad_norm": 0.26731163263320923,
+      "learning_rate": 0.00016247747747747743,
+      "loss": 0.4211,
+      "step": 3063
+    },
+    {
+      "epoch": 0.9185684841671351,
+      "grad_norm": 0.26028740406036377,
+      "learning_rate": 0.00016243243243243242,
+      "loss": 0.4338,
+      "step": 3064
+    },
+    {
+      "epoch": 0.9188682780588345,
+      "grad_norm": 0.25657540559768677,
+      "learning_rate": 0.00016238738738738736,
+      "loss": 0.4262,
+      "step": 3065
+    },
+    {
+      "epoch": 0.919168071950534,
+      "grad_norm": 0.2732202410697937,
+      "learning_rate": 0.00016234234234234233,
+      "loss": 0.441,
+      "step": 3066
+    },
+    {
+      "epoch": 0.9194678658422335,
+      "grad_norm": 0.25451597571372986,
+      "learning_rate": 0.0001622972972972973,
+      "loss": 0.4412,
+      "step": 3067
+    },
+    {
+      "epoch": 0.9197676597339329,
+      "grad_norm": 0.2619662582874298,
+      "learning_rate": 0.00016225225225225223,
+      "loss": 0.4642,
+      "step": 3068
+    },
+    {
+      "epoch": 0.9200674536256324,
+      "grad_norm": 0.25545254349708557,
+      "learning_rate": 0.0001622072072072072,
+      "loss": 0.4091,
+      "step": 3069
+    },
+    {
+      "epoch": 0.9203672475173318,
+      "grad_norm": 0.25901320576667786,
+      "learning_rate": 0.00016216216216216215,
+      "loss": 0.3968,
+      "step": 3070
+    },
+    {
+      "epoch": 0.9206670414090313,
+      "grad_norm": 0.26680001616477966,
+      "learning_rate": 0.0001621171171171171,
+      "loss": 0.4391,
+      "step": 3071
+    },
+    {
+      "epoch": 0.9209668353007308,
+      "grad_norm": 0.2471819370985031,
+      "learning_rate": 0.00016207207207207205,
+      "loss": 0.4167,
+      "step": 3072
+    },
+    {
+      "epoch": 0.9212666291924302,
+      "grad_norm": 0.25497692823410034,
+      "learning_rate": 0.00016202702702702702,
+      "loss": 0.4061,
+      "step": 3073
+    },
+    {
+      "epoch": 0.9215664230841296,
+      "grad_norm": 0.25427013635635376,
+      "learning_rate": 0.00016198198198198196,
+      "loss": 0.4223,
+      "step": 3074
+    },
+    {
+      "epoch": 0.9218662169758292,
+      "grad_norm": 0.24736852943897247,
+      "learning_rate": 0.00016193693693693692,
+      "loss": 0.415,
+      "step": 3075
+    },
+    {
+      "epoch": 0.9221660108675286,
+      "grad_norm": 0.26432523131370544,
+      "learning_rate": 0.00016189189189189188,
+      "loss": 0.4289,
+      "step": 3076
+    },
+    {
+      "epoch": 0.922465804759228,
+      "grad_norm": 0.2562811076641083,
+      "learning_rate": 0.00016184684684684685,
+      "loss": 0.4291,
+      "step": 3077
+    },
+    {
+      "epoch": 0.9227655986509274,
+      "grad_norm": 0.25394096970558167,
+      "learning_rate": 0.00016180180180180178,
+      "loss": 0.4386,
+      "step": 3078
+    },
+    {
+      "epoch": 0.923065392542627,
+      "grad_norm": 0.23978635668754578,
+      "learning_rate": 0.00016175675675675675,
+      "loss": 0.4207,
+      "step": 3079
+    },
+    {
+      "epoch": 0.9233651864343264,
+      "grad_norm": 0.26115700602531433,
+      "learning_rate": 0.0001617117117117117,
+      "loss": 0.4368,
+      "step": 3080
+    },
+    {
+      "epoch": 0.9236649803260258,
+      "grad_norm": 0.2565377652645111,
+      "learning_rate": 0.00016166666666666665,
+      "loss": 0.4203,
+      "step": 3081
+    },
+    {
+      "epoch": 0.9239647742177253,
+      "grad_norm": 0.24304890632629395,
+      "learning_rate": 0.0001616216216216216,
+      "loss": 0.4194,
+      "step": 3082
+    },
+    {
+      "epoch": 0.9242645681094248,
+      "grad_norm": 0.25062617659568787,
+      "learning_rate": 0.00016157657657657658,
+      "loss": 0.4465,
+      "step": 3083
+    },
+    {
+      "epoch": 0.9245643620011242,
+      "grad_norm": 0.24126526713371277,
+      "learning_rate": 0.0001615315315315315,
+      "loss": 0.4245,
+      "step": 3084
+    },
+    {
+      "epoch": 0.9248641558928237,
+      "grad_norm": 0.2625589072704315,
+      "learning_rate": 0.00016148648648648648,
+      "loss": 0.4211,
+      "step": 3085
+    },
+    {
+      "epoch": 0.9251639497845231,
+      "grad_norm": 0.25375208258628845,
+      "learning_rate": 0.00016144144144144144,
+      "loss": 0.3876,
+      "step": 3086
+    },
+    {
+      "epoch": 0.9254637436762226,
+      "grad_norm": 0.25839680433273315,
+      "learning_rate": 0.00016139639639639638,
+      "loss": 0.4405,
+      "step": 3087
+    },
+    {
+      "epoch": 0.9257635375679221,
+      "grad_norm": 0.248674213886261,
+      "learning_rate": 0.00016135135135135131,
+      "loss": 0.4173,
+      "step": 3088
+    },
+    {
+      "epoch": 0.9260633314596215,
+      "grad_norm": 0.2578020989894867,
+      "learning_rate": 0.0001613063063063063,
+      "loss": 0.4059,
+      "step": 3089
+    },
+    {
+      "epoch": 0.9263631253513209,
+      "grad_norm": 0.24993066489696503,
+      "learning_rate": 0.00016126126126126124,
+      "loss": 0.4385,
+      "step": 3090
+    },
+    {
+      "epoch": 0.9266629192430205,
+      "grad_norm": 0.24710343778133392,
+      "learning_rate": 0.00016121621621621618,
+      "loss": 0.4163,
+      "step": 3091
+    },
+    {
+      "epoch": 0.9269627131347199,
+      "grad_norm": 0.26110345125198364,
+      "learning_rate": 0.00016117117117117117,
+      "loss": 0.464,
+      "step": 3092
+    },
+    {
+      "epoch": 0.9272625070264193,
+      "grad_norm": 0.22636494040489197,
+      "learning_rate": 0.0001611261261261261,
+      "loss": 0.3878,
+      "step": 3093
+    },
+    {
+      "epoch": 0.9275623009181188,
+      "grad_norm": 0.2674306333065033,
+      "learning_rate": 0.00016108108108108104,
+      "loss": 0.4009,
+      "step": 3094
+    },
+    {
+      "epoch": 0.9278620948098183,
+      "grad_norm": 0.2749866843223572,
+      "learning_rate": 0.00016103603603603603,
+      "loss": 0.4189,
+      "step": 3095
+    },
+    {
+      "epoch": 0.9281618887015177,
+      "grad_norm": 0.24979329109191895,
+      "learning_rate": 0.00016099099099099097,
+      "loss": 0.4296,
+      "step": 3096
+    },
+    {
+      "epoch": 0.9284616825932172,
+      "grad_norm": 0.2526085376739502,
+      "learning_rate": 0.0001609459459459459,
+      "loss": 0.4158,
+      "step": 3097
+    },
+    {
+      "epoch": 0.9287614764849166,
+      "grad_norm": 0.274911493062973,
+      "learning_rate": 0.0001609009009009009,
+      "loss": 0.4162,
+      "step": 3098
+    },
+    {
+      "epoch": 0.9290612703766161,
+      "grad_norm": 0.25422587990760803,
+      "learning_rate": 0.00016085585585585584,
+      "loss": 0.4305,
+      "step": 3099
+    },
+    {
+      "epoch": 0.9293610642683156,
+      "grad_norm": 0.26116907596588135,
+      "learning_rate": 0.0001608108108108108,
+      "loss": 0.4612,
+      "step": 3100
+    },
+    {
+      "epoch": 0.929660858160015,
+      "grad_norm": 0.24997106194496155,
+      "learning_rate": 0.00016076576576576576,
+      "loss": 0.4039,
+      "step": 3101
+    },
+    {
+      "epoch": 0.9299606520517144,
+      "grad_norm": 0.24362730979919434,
+      "learning_rate": 0.0001607207207207207,
+      "loss": 0.3995,
+      "step": 3102
+    },
+    {
+      "epoch": 0.930260445943414,
+      "grad_norm": 0.2644561529159546,
+      "learning_rate": 0.00016067567567567566,
+      "loss": 0.4282,
+      "step": 3103
+    },
+    {
+      "epoch": 0.9305602398351134,
+      "grad_norm": 0.2574602961540222,
+      "learning_rate": 0.00016063063063063063,
+      "loss": 0.4243,
+      "step": 3104
+    },
+    {
+      "epoch": 0.9308600337268128,
+      "grad_norm": 0.27288708090782166,
+      "learning_rate": 0.00016058558558558556,
+      "loss": 0.4057,
+      "step": 3105
+    },
+    {
+      "epoch": 0.9311598276185122,
+      "grad_norm": 0.2651262879371643,
+      "learning_rate": 0.00016054054054054053,
+      "loss": 0.407,
+      "step": 3106
+    },
+    {
+      "epoch": 0.9314596215102118,
+      "grad_norm": 0.26550403237342834,
+      "learning_rate": 0.0001604954954954955,
+      "loss": 0.4294,
+      "step": 3107
+    },
+    {
+      "epoch": 0.9317594154019112,
+      "grad_norm": 0.25612547993659973,
+      "learning_rate": 0.00016045045045045043,
+      "loss": 0.4224,
+      "step": 3108
+    },
+    {
+      "epoch": 0.9320592092936106,
+      "grad_norm": 0.25533682107925415,
+      "learning_rate": 0.0001604054054054054,
+      "loss": 0.4134,
+      "step": 3109
+    },
+    {
+      "epoch": 0.9323590031853101,
+      "grad_norm": 0.24888603389263153,
+      "learning_rate": 0.00016036036036036036,
+      "loss": 0.3807,
+      "step": 3110
+    },
+    {
+      "epoch": 0.9326587970770096,
+      "grad_norm": 0.25240257382392883,
+      "learning_rate": 0.00016031531531531532,
+      "loss": 0.406,
+      "step": 3111
+    },
+    {
+      "epoch": 0.932958590968709,
+      "grad_norm": 0.25701963901519775,
+      "learning_rate": 0.00016027027027027026,
+      "loss": 0.4319,
+      "step": 3112
+    },
+    {
+      "epoch": 0.9332583848604085,
+      "grad_norm": 0.2428182065486908,
+      "learning_rate": 0.0001602252252252252,
+      "loss": 0.4,
+      "step": 3113
+    },
+    {
+      "epoch": 0.9335581787521079,
+      "grad_norm": 0.2379184514284134,
+      "learning_rate": 0.00016018018018018018,
+      "loss": 0.4209,
+      "step": 3114
+    },
+    {
+      "epoch": 0.9338579726438074,
+      "grad_norm": 0.28138455748558044,
+      "learning_rate": 0.00016013513513513512,
+      "loss": 0.446,
+      "step": 3115
+    },
+    {
+      "epoch": 0.9341577665355069,
+      "grad_norm": 0.2478276938199997,
+      "learning_rate": 0.00016009009009009006,
+      "loss": 0.384,
+      "step": 3116
+    },
+    {
+      "epoch": 0.9344575604272063,
+      "grad_norm": 0.2511236369609833,
+      "learning_rate": 0.00016004504504504505,
+      "loss": 0.4289,
+      "step": 3117
+    },
+    {
+      "epoch": 0.9347573543189057,
+      "grad_norm": 0.24487215280532837,
+      "learning_rate": 0.00015999999999999999,
+      "loss": 0.4271,
+      "step": 3118
+    },
+    {
+      "epoch": 0.9350571482106053,
+      "grad_norm": 0.27695828676223755,
+      "learning_rate": 0.00015995495495495492,
+      "loss": 0.4384,
+      "step": 3119
+    },
+    {
+      "epoch": 0.9353569421023047,
+      "grad_norm": 0.24011778831481934,
+      "learning_rate": 0.00015990990990990991,
+      "loss": 0.3903,
+      "step": 3120
+    },
+    {
+      "epoch": 0.9356567359940041,
+      "grad_norm": 0.2450532615184784,
+      "learning_rate": 0.00015986486486486485,
+      "loss": 0.4258,
+      "step": 3121
+    },
+    {
+      "epoch": 0.9359565298857035,
+      "grad_norm": 0.26375943422317505,
+      "learning_rate": 0.0001598198198198198,
+      "loss": 0.4395,
+      "step": 3122
+    },
+    {
+      "epoch": 0.9362563237774031,
+      "grad_norm": 0.2335769087076187,
+      "learning_rate": 0.00015977477477477478,
+      "loss": 0.3922,
+      "step": 3123
+    },
+    {
+      "epoch": 0.9365561176691025,
+      "grad_norm": 0.2576378285884857,
+      "learning_rate": 0.00015972972972972972,
+      "loss": 0.412,
+      "step": 3124
+    },
+    {
+      "epoch": 0.936855911560802,
+      "grad_norm": 0.24350136518478394,
+      "learning_rate": 0.00015968468468468465,
+      "loss": 0.4207,
+      "step": 3125
+    },
+    {
+      "epoch": 0.9371557054525014,
+      "grad_norm": 0.24622796475887299,
+      "learning_rate": 0.00015963963963963964,
+      "loss": 0.4137,
+      "step": 3126
+    },
+    {
+      "epoch": 0.9374554993442009,
+      "grad_norm": 0.24007757008075714,
+      "learning_rate": 0.00015959459459459458,
+      "loss": 0.4297,
+      "step": 3127
+    },
+    {
+      "epoch": 0.9377552932359003,
+      "grad_norm": 0.2597469091415405,
+      "learning_rate": 0.00015954954954954952,
+      "loss": 0.4262,
+      "step": 3128
+    },
+    {
+      "epoch": 0.9380550871275998,
+      "grad_norm": 0.25026780366897583,
+      "learning_rate": 0.0001595045045045045,
+      "loss": 0.4206,
+      "step": 3129
+    },
+    {
+      "epoch": 0.9383548810192992,
+      "grad_norm": 0.2386496216058731,
+      "learning_rate": 0.00015945945945945944,
+      "loss": 0.4266,
+      "step": 3130
+    },
+    {
+      "epoch": 0.9386546749109987,
+      "grad_norm": 0.25480231642723083,
+      "learning_rate": 0.00015941441441441438,
+      "loss": 0.4373,
+      "step": 3131
+    },
+    {
+      "epoch": 0.9389544688026982,
+      "grad_norm": 0.23915326595306396,
+      "learning_rate": 0.00015936936936936937,
+      "loss": 0.411,
+      "step": 3132
+    },
+    {
+      "epoch": 0.9392542626943976,
+      "grad_norm": 0.2645804286003113,
+      "learning_rate": 0.0001593243243243243,
+      "loss": 0.4334,
+      "step": 3133
+    },
+    {
+      "epoch": 0.939554056586097,
+      "grad_norm": 0.26099300384521484,
+      "learning_rate": 0.00015927927927927927,
+      "loss": 0.4264,
+      "step": 3134
+    },
+    {
+      "epoch": 0.9398538504777965,
+      "grad_norm": 0.2576426863670349,
+      "learning_rate": 0.0001592342342342342,
+      "loss": 0.4276,
+      "step": 3135
+    },
+    {
+      "epoch": 0.940153644369496,
+      "grad_norm": 0.2772998511791229,
+      "learning_rate": 0.00015918918918918917,
+      "loss": 0.4317,
+      "step": 3136
+    },
+    {
+      "epoch": 0.9404534382611954,
+      "grad_norm": 0.2620845437049866,
+      "learning_rate": 0.00015914414414414414,
+      "loss": 0.4139,
+      "step": 3137
+    },
+    {
+      "epoch": 0.9407532321528949,
+      "grad_norm": 0.24328507483005524,
+      "learning_rate": 0.00015909909909909907,
+      "loss": 0.3952,
+      "step": 3138
+    },
+    {
+      "epoch": 0.9410530260445943,
+      "grad_norm": 0.25061240792274475,
+      "learning_rate": 0.00015905405405405404,
+      "loss": 0.3669,
+      "step": 3139
+    },
+    {
+      "epoch": 0.9413528199362938,
+      "grad_norm": 0.24149778485298157,
+      "learning_rate": 0.000159009009009009,
+      "loss": 0.416,
+      "step": 3140
+    },
+    {
+      "epoch": 0.9416526138279933,
+      "grad_norm": 0.2483665943145752,
+      "learning_rate": 0.00015896396396396394,
+      "loss": 0.3935,
+      "step": 3141
+    },
+    {
+      "epoch": 0.9419524077196927,
+      "grad_norm": 0.270732045173645,
+      "learning_rate": 0.0001589189189189189,
+      "loss": 0.3991,
+      "step": 3142
+    },
+    {
+      "epoch": 0.9422522016113921,
+      "grad_norm": 0.2482568472623825,
+      "learning_rate": 0.00015887387387387387,
+      "loss": 0.417,
+      "step": 3143
+    },
+    {
+      "epoch": 0.9425519955030917,
+      "grad_norm": 0.241281196475029,
+      "learning_rate": 0.0001588288288288288,
+      "loss": 0.4144,
+      "step": 3144
+    },
+    {
+      "epoch": 0.9428517893947911,
+      "grad_norm": 0.2527291774749756,
+      "learning_rate": 0.00015878378378378377,
+      "loss": 0.4318,
+      "step": 3145
+    },
+    {
+      "epoch": 0.9431515832864905,
+      "grad_norm": 0.25932496786117554,
+      "learning_rate": 0.00015873873873873873,
+      "loss": 0.4305,
+      "step": 3146
+    },
+    {
+      "epoch": 0.9434513771781899,
+      "grad_norm": 0.2633473873138428,
+      "learning_rate": 0.00015869369369369367,
+      "loss": 0.4395,
+      "step": 3147
+    },
+    {
+      "epoch": 0.9437511710698895,
+      "grad_norm": 0.23622636497020721,
+      "learning_rate": 0.00015864864864864866,
+      "loss": 0.3921,
+      "step": 3148
+    },
+    {
+      "epoch": 0.9440509649615889,
+      "grad_norm": 0.2766309976577759,
+      "learning_rate": 0.0001586036036036036,
+      "loss": 0.477,
+      "step": 3149
+    },
+    {
+      "epoch": 0.9443507588532883,
+      "grad_norm": 0.24325770139694214,
+      "learning_rate": 0.00015855855855855853,
+      "loss": 0.3976,
+      "step": 3150
+    },
+    {
+      "epoch": 0.9446505527449878,
+      "grad_norm": 0.25284042954444885,
+      "learning_rate": 0.00015851351351351352,
+      "loss": 0.4077,
+      "step": 3151
+    },
+    {
+      "epoch": 0.9449503466366873,
+      "grad_norm": 0.2570466995239258,
+      "learning_rate": 0.00015846846846846846,
+      "loss": 0.4361,
+      "step": 3152
+    },
+    {
+      "epoch": 0.9452501405283867,
+      "grad_norm": 0.2617407441139221,
+      "learning_rate": 0.0001584234234234234,
+      "loss": 0.4358,
+      "step": 3153
+    },
+    {
+      "epoch": 0.9455499344200862,
+      "grad_norm": 0.25398409366607666,
+      "learning_rate": 0.0001583783783783784,
+      "loss": 0.4148,
+      "step": 3154
+    },
+    {
+      "epoch": 0.9458497283117856,
+      "grad_norm": 0.23910127580165863,
+      "learning_rate": 0.00015833333333333332,
+      "loss": 0.4149,
+      "step": 3155
+    },
+    {
+      "epoch": 0.9461495222034851,
+      "grad_norm": 0.24659039080142975,
+      "learning_rate": 0.00015828828828828826,
+      "loss": 0.4026,
+      "step": 3156
+    },
+    {
+      "epoch": 0.9464493160951846,
+      "grad_norm": 0.23952017724514008,
+      "learning_rate": 0.00015824324324324325,
+      "loss": 0.3825,
+      "step": 3157
+    },
+    {
+      "epoch": 0.946749109986884,
+      "grad_norm": 0.2537391185760498,
+      "learning_rate": 0.0001581981981981982,
+      "loss": 0.4131,
+      "step": 3158
+    },
+    {
+      "epoch": 0.9470489038785834,
+      "grad_norm": 0.25354787707328796,
+      "learning_rate": 0.00015815315315315313,
+      "loss": 0.4318,
+      "step": 3159
+    },
+    {
+      "epoch": 0.947348697770283,
+      "grad_norm": 0.23185314238071442,
+      "learning_rate": 0.0001581081081081081,
+      "loss": 0.4062,
+      "step": 3160
+    },
+    {
+      "epoch": 0.9476484916619824,
+      "grad_norm": 0.23187117278575897,
+      "learning_rate": 0.00015806306306306305,
+      "loss": 0.4265,
+      "step": 3161
+    },
+    {
+      "epoch": 0.9479482855536818,
+      "grad_norm": 0.2618250548839569,
+      "learning_rate": 0.000158018018018018,
+      "loss": 0.3997,
+      "step": 3162
+    },
+    {
+      "epoch": 0.9482480794453813,
+      "grad_norm": 0.2492150217294693,
+      "learning_rate": 0.00015797297297297295,
+      "loss": 0.4418,
+      "step": 3163
+    },
+    {
+      "epoch": 0.9485478733370808,
+      "grad_norm": 0.23397988080978394,
+      "learning_rate": 0.00015792792792792792,
+      "loss": 0.4056,
+      "step": 3164
+    },
+    {
+      "epoch": 0.9488476672287802,
+      "grad_norm": 0.2404291331768036,
+      "learning_rate": 0.00015788288288288285,
+      "loss": 0.4234,
+      "step": 3165
+    },
+    {
+      "epoch": 0.9491474611204797,
+      "grad_norm": 0.25739723443984985,
+      "learning_rate": 0.00015783783783783782,
+      "loss": 0.4376,
+      "step": 3166
+    },
+    {
+      "epoch": 0.9494472550121791,
+      "grad_norm": 0.2516269385814667,
+      "learning_rate": 0.00015779279279279278,
+      "loss": 0.4371,
+      "step": 3167
+    },
+    {
+      "epoch": 0.9497470489038786,
+      "grad_norm": 0.2476969063282013,
+      "learning_rate": 0.00015774774774774775,
+      "loss": 0.4042,
+      "step": 3168
+    },
+    {
+      "epoch": 0.950046842795578,
+      "grad_norm": 0.2627060115337372,
+      "learning_rate": 0.00015770270270270268,
+      "loss": 0.4651,
+      "step": 3169
+    },
+    {
+      "epoch": 0.9503466366872775,
+      "grad_norm": 0.2544974982738495,
+      "learning_rate": 0.00015765765765765765,
+      "loss": 0.439,
+      "step": 3170
+    },
+    {
+      "epoch": 0.9506464305789769,
+      "grad_norm": 0.2589476406574249,
+      "learning_rate": 0.0001576126126126126,
+      "loss": 0.4093,
+      "step": 3171
+    },
+    {
+      "epoch": 0.9509462244706764,
+      "grad_norm": 0.24550233781337738,
+      "learning_rate": 0.00015756756756756755,
+      "loss": 0.4176,
+      "step": 3172
+    },
+    {
+      "epoch": 0.9512460183623759,
+      "grad_norm": 0.2545708119869232,
+      "learning_rate": 0.0001575225225225225,
+      "loss": 0.4308,
+      "step": 3173
+    },
+    {
+      "epoch": 0.9515458122540753,
+      "grad_norm": 0.24934959411621094,
+      "learning_rate": 0.00015747747747747747,
+      "loss": 0.3848,
+      "step": 3174
+    },
+    {
+      "epoch": 0.9518456061457747,
+      "grad_norm": 0.2537474036216736,
+      "learning_rate": 0.0001574324324324324,
+      "loss": 0.4149,
+      "step": 3175
+    },
+    {
+      "epoch": 0.9521454000374743,
+      "grad_norm": 0.2342180609703064,
+      "learning_rate": 0.00015738738738738738,
+      "loss": 0.3815,
+      "step": 3176
+    },
+    {
+      "epoch": 0.9524451939291737,
+      "grad_norm": 0.2533189356327057,
+      "learning_rate": 0.00015734234234234234,
+      "loss": 0.3993,
+      "step": 3177
+    },
+    {
+      "epoch": 0.9527449878208731,
+      "grad_norm": 0.2543899714946747,
+      "learning_rate": 0.00015729729729729728,
+      "loss": 0.4062,
+      "step": 3178
+    },
+    {
+      "epoch": 0.9530447817125726,
+      "grad_norm": 0.2690375745296478,
+      "learning_rate": 0.00015725225225225224,
+      "loss": 0.4109,
+      "step": 3179
+    },
+    {
+      "epoch": 0.9533445756042721,
+      "grad_norm": 0.2458389550447464,
+      "learning_rate": 0.0001572072072072072,
+      "loss": 0.4056,
+      "step": 3180
+    },
+    {
+      "epoch": 0.9536443694959715,
+      "grad_norm": 0.2488129585981369,
+      "learning_rate": 0.00015716216216216214,
+      "loss": 0.4312,
+      "step": 3181
+    },
+    {
+      "epoch": 0.953944163387671,
+      "grad_norm": 0.23376654088497162,
+      "learning_rate": 0.00015711711711711713,
+      "loss": 0.4254,
+      "step": 3182
+    },
+    {
+      "epoch": 0.9542439572793704,
+      "grad_norm": 0.2395961731672287,
+      "learning_rate": 0.00015707207207207207,
+      "loss": 0.4169,
+      "step": 3183
+    },
+    {
+      "epoch": 0.9545437511710699,
+      "grad_norm": 0.2538315951824188,
+      "learning_rate": 0.000157027027027027,
+      "loss": 0.4418,
+      "step": 3184
+    },
+    {
+      "epoch": 0.9548435450627694,
+      "grad_norm": 0.25304052233695984,
+      "learning_rate": 0.00015698198198198194,
+      "loss": 0.4082,
+      "step": 3185
+    },
+    {
+      "epoch": 0.9551433389544688,
+      "grad_norm": 0.25075477361679077,
+      "learning_rate": 0.00015693693693693693,
+      "loss": 0.4288,
+      "step": 3186
+    },
+    {
+      "epoch": 0.9554431328461682,
+      "grad_norm": 0.2536257803440094,
+      "learning_rate": 0.00015689189189189187,
+      "loss": 0.4326,
+      "step": 3187
+    },
+    {
+      "epoch": 0.9557429267378678,
+      "grad_norm": 0.24698510766029358,
+      "learning_rate": 0.0001568468468468468,
+      "loss": 0.4245,
+      "step": 3188
+    },
+    {
+      "epoch": 0.9560427206295672,
+      "grad_norm": 0.2647172212600708,
+      "learning_rate": 0.0001568018018018018,
+      "loss": 0.4018,
+      "step": 3189
+    },
+    {
+      "epoch": 0.9563425145212666,
+      "grad_norm": 0.2701962888240814,
+      "learning_rate": 0.00015675675675675673,
+      "loss": 0.4289,
+      "step": 3190
+    },
+    {
+      "epoch": 0.956642308412966,
+      "grad_norm": 0.24717958271503448,
+      "learning_rate": 0.0001567117117117117,
+      "loss": 0.4202,
+      "step": 3191
+    },
+    {
+      "epoch": 0.9569421023046656,
+      "grad_norm": 0.24269205331802368,
+      "learning_rate": 0.00015666666666666666,
+      "loss": 0.3864,
+      "step": 3192
+    },
+    {
+      "epoch": 0.957241896196365,
+      "grad_norm": 0.2608867287635803,
+      "learning_rate": 0.0001566216216216216,
+      "loss": 0.4403,
+      "step": 3193
+    },
+    {
+      "epoch": 0.9575416900880644,
+      "grad_norm": 0.2656077444553375,
+      "learning_rate": 0.00015657657657657656,
+      "loss": 0.4203,
+      "step": 3194
+    },
+    {
+      "epoch": 0.9578414839797639,
+      "grad_norm": 0.2502601444721222,
+      "learning_rate": 0.00015653153153153153,
+      "loss": 0.4367,
+      "step": 3195
+    },
+    {
+      "epoch": 0.9581412778714634,
+      "grad_norm": 0.26323074102401733,
+      "learning_rate": 0.00015648648648648646,
+      "loss": 0.4299,
+      "step": 3196
+    },
+    {
+      "epoch": 0.9584410717631628,
+      "grad_norm": 0.251655250787735,
+      "learning_rate": 0.00015644144144144143,
+      "loss": 0.4003,
+      "step": 3197
+    },
+    {
+      "epoch": 0.9587408656548623,
+      "grad_norm": 0.25573211908340454,
+      "learning_rate": 0.0001563963963963964,
+      "loss": 0.4005,
+      "step": 3198
+    },
+    {
+      "epoch": 0.9590406595465617,
+      "grad_norm": 0.2425386756658554,
+      "learning_rate": 0.00015635135135135133,
+      "loss": 0.3839,
+      "step": 3199
+    },
+    {
+      "epoch": 0.9593404534382612,
+      "grad_norm": 0.25007614493370056,
+      "learning_rate": 0.0001563063063063063,
+      "loss": 0.4108,
+      "step": 3200
+    },
+    {
+      "epoch": 0.9596402473299607,
+      "grad_norm": 0.26201489567756653,
+      "learning_rate": 0.00015626126126126126,
+      "loss": 0.4448,
+      "step": 3201
+    },
+    {
+      "epoch": 0.9599400412216601,
+      "grad_norm": 0.25759240984916687,
+      "learning_rate": 0.0001562162162162162,
+      "loss": 0.4318,
+      "step": 3202
+    },
+    {
+      "epoch": 0.9602398351133595,
+      "grad_norm": 0.24872949719429016,
+      "learning_rate": 0.00015617117117117116,
+      "loss": 0.4018,
+      "step": 3203
+    },
+    {
+      "epoch": 0.9605396290050591,
+      "grad_norm": 0.2599150538444519,
+      "learning_rate": 0.00015612612612612612,
+      "loss": 0.4071,
+      "step": 3204
+    },
+    {
+      "epoch": 0.9608394228967585,
+      "grad_norm": 0.2560199499130249,
+      "learning_rate": 0.00015608108108108108,
+      "loss": 0.4251,
+      "step": 3205
+    },
+    {
+      "epoch": 0.9611392167884579,
+      "grad_norm": 0.24916759133338928,
+      "learning_rate": 0.00015603603603603602,
+      "loss": 0.3954,
+      "step": 3206
+    },
+    {
+      "epoch": 0.9614390106801574,
+      "grad_norm": 0.24545198678970337,
+      "learning_rate": 0.00015599099099099098,
+      "loss": 0.3978,
+      "step": 3207
+    },
+    {
+      "epoch": 0.9617388045718569,
+      "grad_norm": 0.2455730438232422,
+      "learning_rate": 0.00015594594594594595,
+      "loss": 0.3949,
+      "step": 3208
+    },
+    {
+      "epoch": 0.9620385984635563,
+      "grad_norm": 0.23964105546474457,
+      "learning_rate": 0.00015590090090090089,
+      "loss": 0.413,
+      "step": 3209
+    },
+    {
+      "epoch": 0.9623383923552558,
+      "grad_norm": 0.2658955156803131,
+      "learning_rate": 0.00015585585585585582,
+      "loss": 0.4455,
+      "step": 3210
+    },
+    {
+      "epoch": 0.9626381862469552,
+      "grad_norm": 0.2649344205856323,
+      "learning_rate": 0.0001558108108108108,
+      "loss": 0.409,
+      "step": 3211
+    },
+    {
+      "epoch": 0.9629379801386547,
+      "grad_norm": 0.24537350237369537,
+      "learning_rate": 0.00015576576576576575,
+      "loss": 0.404,
+      "step": 3212
+    },
+    {
+      "epoch": 0.9632377740303542,
+      "grad_norm": 0.2801573574542999,
+      "learning_rate": 0.00015572072072072069,
+      "loss": 0.4254,
+      "step": 3213
+    },
+    {
+      "epoch": 0.9635375679220536,
+      "grad_norm": 0.2631642818450928,
+      "learning_rate": 0.00015567567567567568,
+      "loss": 0.4201,
+      "step": 3214
+    },
+    {
+      "epoch": 0.963837361813753,
+      "grad_norm": 0.25088247656822205,
+      "learning_rate": 0.00015563063063063061,
+      "loss": 0.3975,
+      "step": 3215
+    },
+    {
+      "epoch": 0.9641371557054526,
+      "grad_norm": 0.271305650472641,
+      "learning_rate": 0.00015558558558558555,
+      "loss": 0.4317,
+      "step": 3216
+    },
+    {
+      "epoch": 0.964436949597152,
+      "grad_norm": 0.2699548304080963,
+      "learning_rate": 0.00015554054054054054,
+      "loss": 0.4175,
+      "step": 3217
+    },
+    {
+      "epoch": 0.9647367434888514,
+      "grad_norm": 0.25501522421836853,
+      "learning_rate": 0.00015549549549549548,
+      "loss": 0.4228,
+      "step": 3218
+    },
+    {
+      "epoch": 0.9650365373805508,
+      "grad_norm": 0.2544115483760834,
+      "learning_rate": 0.00015545045045045042,
+      "loss": 0.3931,
+      "step": 3219
+    },
+    {
+      "epoch": 0.9653363312722504,
+      "grad_norm": 0.26797887682914734,
+      "learning_rate": 0.0001554054054054054,
+      "loss": 0.3828,
+      "step": 3220
+    },
+    {
+      "epoch": 0.9656361251639498,
+      "grad_norm": 0.2449193000793457,
+      "learning_rate": 0.00015536036036036034,
+      "loss": 0.4103,
+      "step": 3221
+    },
+    {
+      "epoch": 0.9659359190556492,
+      "grad_norm": 0.24695518612861633,
+      "learning_rate": 0.00015531531531531528,
+      "loss": 0.4088,
+      "step": 3222
+    },
+    {
+      "epoch": 0.9662357129473487,
+      "grad_norm": 0.24960634112358093,
+      "learning_rate": 0.00015527027027027027,
+      "loss": 0.4193,
+      "step": 3223
+    },
+    {
+      "epoch": 0.9665355068390482,
+      "grad_norm": 0.255014568567276,
+      "learning_rate": 0.0001552252252252252,
+      "loss": 0.4041,
+      "step": 3224
+    },
+    {
+      "epoch": 0.9668353007307476,
+      "grad_norm": 0.2603205442428589,
+      "learning_rate": 0.00015518018018018017,
+      "loss": 0.4438,
+      "step": 3225
+    },
+    {
+      "epoch": 0.9671350946224471,
+      "grad_norm": 0.2639630436897278,
+      "learning_rate": 0.00015513513513513514,
+      "loss": 0.4313,
+      "step": 3226
+    },
+    {
+      "epoch": 0.9674348885141465,
+      "grad_norm": 0.25526097416877747,
+      "learning_rate": 0.00015509009009009007,
+      "loss": 0.4165,
+      "step": 3227
+    },
+    {
+      "epoch": 0.967734682405846,
+      "grad_norm": 0.26529547572135925,
+      "learning_rate": 0.00015504504504504504,
+      "loss": 0.4046,
+      "step": 3228
+    },
+    {
+      "epoch": 0.9680344762975455,
+      "grad_norm": 0.25639456510543823,
+      "learning_rate": 0.000155,
+      "loss": 0.4092,
+      "step": 3229
+    },
+    {
+      "epoch": 0.9683342701892449,
+      "grad_norm": 0.24461041390895844,
+      "learning_rate": 0.00015495495495495494,
+      "loss": 0.4155,
+      "step": 3230
+    },
+    {
+      "epoch": 0.9686340640809443,
+      "grad_norm": 0.25683319568634033,
+      "learning_rate": 0.0001549099099099099,
+      "loss": 0.4152,
+      "step": 3231
+    },
+    {
+      "epoch": 0.9689338579726438,
+      "grad_norm": 0.2518060505390167,
+      "learning_rate": 0.00015486486486486484,
+      "loss": 0.4299,
+      "step": 3232
+    },
+    {
+      "epoch": 0.9692336518643433,
+      "grad_norm": 0.24693354964256287,
+      "learning_rate": 0.0001548198198198198,
+      "loss": 0.4178,
+      "step": 3233
+    },
+    {
+      "epoch": 0.9695334457560427,
+      "grad_norm": 0.2540181577205658,
+      "learning_rate": 0.00015477477477477477,
+      "loss": 0.4158,
+      "step": 3234
+    },
+    {
+      "epoch": 0.9698332396477422,
+      "grad_norm": 0.27640795707702637,
+      "learning_rate": 0.0001547297297297297,
+      "loss": 0.4387,
+      "step": 3235
+    },
+    {
+      "epoch": 0.9701330335394416,
+      "grad_norm": 0.23969754576683044,
+      "learning_rate": 0.00015468468468468467,
+      "loss": 0.4322,
+      "step": 3236
+    },
+    {
+      "epoch": 0.9704328274311411,
+      "grad_norm": 0.27284395694732666,
+      "learning_rate": 0.00015463963963963963,
+      "loss": 0.4359,
+      "step": 3237
+    },
+    {
+      "epoch": 0.9707326213228405,
+      "grad_norm": 0.2576393187046051,
+      "learning_rate": 0.00015459459459459457,
+      "loss": 0.4282,
+      "step": 3238
+    },
+    {
+      "epoch": 0.97103241521454,
+      "grad_norm": 0.2445439100265503,
+      "learning_rate": 0.00015454954954954956,
+      "loss": 0.3871,
+      "step": 3239
+    },
+    {
+      "epoch": 0.9713322091062394,
+      "grad_norm": 0.25176161527633667,
+      "learning_rate": 0.0001545045045045045,
+      "loss": 0.4512,
+      "step": 3240
+    },
+    {
+      "epoch": 0.971632002997939,
+      "grad_norm": 0.25426721572875977,
+      "learning_rate": 0.00015445945945945943,
+      "loss": 0.4265,
+      "step": 3241
+    },
+    {
+      "epoch": 0.9719317968896384,
+      "grad_norm": 0.25714048743247986,
+      "learning_rate": 0.00015441441441441442,
+      "loss": 0.4169,
+      "step": 3242
+    },
+    {
+      "epoch": 0.9722315907813378,
+      "grad_norm": 0.2531634569168091,
+      "learning_rate": 0.00015436936936936936,
+      "loss": 0.4404,
+      "step": 3243
+    },
+    {
+      "epoch": 0.9725313846730372,
+      "grad_norm": 0.2426653355360031,
+      "learning_rate": 0.0001543243243243243,
+      "loss": 0.3981,
+      "step": 3244
+    },
+    {
+      "epoch": 0.9728311785647368,
+      "grad_norm": 0.24552041292190552,
+      "learning_rate": 0.00015427927927927929,
+      "loss": 0.4124,
+      "step": 3245
+    },
+    {
+      "epoch": 0.9731309724564362,
+      "grad_norm": 0.2520458698272705,
+      "learning_rate": 0.00015423423423423422,
+      "loss": 0.4132,
+      "step": 3246
+    },
+    {
+      "epoch": 0.9734307663481356,
+      "grad_norm": 0.2456391304731369,
+      "learning_rate": 0.00015418918918918916,
+      "loss": 0.3962,
+      "step": 3247
+    },
+    {
+      "epoch": 0.9737305602398351,
+      "grad_norm": 0.2388329654932022,
+      "learning_rate": 0.00015414414414414415,
+      "loss": 0.4302,
+      "step": 3248
+    },
+    {
+      "epoch": 0.9740303541315346,
+      "grad_norm": 0.24677686393260956,
+      "learning_rate": 0.0001540990990990991,
+      "loss": 0.3775,
+      "step": 3249
+    },
+    {
+      "epoch": 0.974330148023234,
+      "grad_norm": 0.260043203830719,
+      "learning_rate": 0.00015405405405405402,
+      "loss": 0.4242,
+      "step": 3250
+    },
+    {
+      "epoch": 0.9746299419149335,
+      "grad_norm": 0.25858423113822937,
+      "learning_rate": 0.00015400900900900902,
+      "loss": 0.4033,
+      "step": 3251
+    },
+    {
+      "epoch": 0.9749297358066329,
+      "grad_norm": 0.245976984500885,
+      "learning_rate": 0.00015396396396396395,
+      "loss": 0.4111,
+      "step": 3252
+    },
+    {
+      "epoch": 0.9752295296983324,
+      "grad_norm": 0.2492712438106537,
+      "learning_rate": 0.0001539189189189189,
+      "loss": 0.4013,
+      "step": 3253
+    },
+    {
+      "epoch": 0.9755293235900319,
+      "grad_norm": 0.2476668655872345,
+      "learning_rate": 0.00015387387387387388,
+      "loss": 0.3959,
+      "step": 3254
+    },
+    {
+      "epoch": 0.9758291174817313,
+      "grad_norm": 0.2594514787197113,
+      "learning_rate": 0.00015382882882882882,
+      "loss": 0.4646,
+      "step": 3255
+    },
+    {
+      "epoch": 0.9761289113734307,
+      "grad_norm": 0.27506035566329956,
+      "learning_rate": 0.00015378378378378375,
+      "loss": 0.4511,
+      "step": 3256
+    },
+    {
+      "epoch": 0.9764287052651303,
+      "grad_norm": 0.27209872007369995,
+      "learning_rate": 0.00015373873873873872,
+      "loss": 0.4084,
+      "step": 3257
+    },
+    {
+      "epoch": 0.9767284991568297,
+      "grad_norm": 0.24144190549850464,
+      "learning_rate": 0.00015369369369369368,
+      "loss": 0.3934,
+      "step": 3258
+    },
+    {
+      "epoch": 0.9770282930485291,
+      "grad_norm": 0.25071820616722107,
+      "learning_rate": 0.00015364864864864862,
+      "loss": 0.4203,
+      "step": 3259
+    },
+    {
+      "epoch": 0.9773280869402285,
+      "grad_norm": 0.24017009139060974,
+      "learning_rate": 0.00015360360360360358,
+      "loss": 0.4086,
+      "step": 3260
+    },
+    {
+      "epoch": 0.9776278808319281,
+      "grad_norm": 0.2667907774448395,
+      "learning_rate": 0.00015355855855855855,
+      "loss": 0.4008,
+      "step": 3261
+    },
+    {
+      "epoch": 0.9779276747236275,
+      "grad_norm": 0.26080140471458435,
+      "learning_rate": 0.0001535135135135135,
+      "loss": 0.4328,
+      "step": 3262
+    },
+    {
+      "epoch": 0.978227468615327,
+      "grad_norm": 0.25462666153907776,
+      "learning_rate": 0.00015346846846846845,
+      "loss": 0.446,
+      "step": 3263
+    },
+    {
+      "epoch": 0.9785272625070264,
+      "grad_norm": 0.23574258387088776,
+      "learning_rate": 0.0001534234234234234,
+      "loss": 0.3997,
+      "step": 3264
+    },
+    {
+      "epoch": 0.9788270563987259,
+      "grad_norm": 0.2527666389942169,
+      "learning_rate": 0.00015337837837837837,
+      "loss": 0.4495,
+      "step": 3265
+    },
+    {
+      "epoch": 0.9791268502904253,
+      "grad_norm": 0.24528340995311737,
+      "learning_rate": 0.0001533333333333333,
+      "loss": 0.4119,
+      "step": 3266
+    },
+    {
+      "epoch": 0.9794266441821248,
+      "grad_norm": 0.2558531165122986,
+      "learning_rate": 0.00015328828828828827,
+      "loss": 0.4028,
+      "step": 3267
+    },
+    {
+      "epoch": 0.9797264380738242,
+      "grad_norm": 0.2466735988855362,
+      "learning_rate": 0.00015324324324324324,
+      "loss": 0.4302,
+      "step": 3268
+    },
+    {
+      "epoch": 0.9800262319655237,
+      "grad_norm": 0.2911660075187683,
+      "learning_rate": 0.00015319819819819818,
+      "loss": 0.4551,
+      "step": 3269
+    },
+    {
+      "epoch": 0.9803260258572232,
+      "grad_norm": 0.2526625990867615,
+      "learning_rate": 0.00015315315315315314,
+      "loss": 0.41,
+      "step": 3270
+    },
+    {
+      "epoch": 0.9806258197489226,
+      "grad_norm": 0.24766401946544647,
+      "learning_rate": 0.0001531081081081081,
+      "loss": 0.3924,
+      "step": 3271
+    },
+    {
+      "epoch": 0.980925613640622,
+      "grad_norm": 0.2853473424911499,
+      "learning_rate": 0.00015306306306306304,
+      "loss": 0.4287,
+      "step": 3272
+    },
+    {
+      "epoch": 0.9812254075323216,
+      "grad_norm": 0.25968533754348755,
+      "learning_rate": 0.00015301801801801803,
+      "loss": 0.4395,
+      "step": 3273
+    },
+    {
+      "epoch": 0.981525201424021,
+      "grad_norm": 0.2610696852207184,
+      "learning_rate": 0.00015297297297297297,
+      "loss": 0.4075,
+      "step": 3274
+    },
+    {
+      "epoch": 0.9818249953157204,
+      "grad_norm": 0.23721027374267578,
+      "learning_rate": 0.0001529279279279279,
+      "loss": 0.4071,
+      "step": 3275
+    },
+    {
+      "epoch": 0.9821247892074199,
+      "grad_norm": 0.26066192984580994,
+      "learning_rate": 0.0001528828828828829,
+      "loss": 0.4205,
+      "step": 3276
+    },
+    {
+      "epoch": 0.9824245830991194,
+      "grad_norm": 0.285672128200531,
+      "learning_rate": 0.00015283783783783783,
+      "loss": 0.4046,
+      "step": 3277
+    },
+    {
+      "epoch": 0.9827243769908188,
+      "grad_norm": 0.22871406376361847,
+      "learning_rate": 0.00015279279279279277,
+      "loss": 0.3944,
+      "step": 3278
+    },
+    {
+      "epoch": 0.9830241708825183,
+      "grad_norm": 0.26142746210098267,
+      "learning_rate": 0.00015274774774774776,
+      "loss": 0.4084,
+      "step": 3279
+    },
+    {
+      "epoch": 0.9833239647742177,
+      "grad_norm": 0.2570386528968811,
+      "learning_rate": 0.0001527027027027027,
+      "loss": 0.4225,
+      "step": 3280
+    },
+    {
+      "epoch": 0.9836237586659172,
+      "grad_norm": 0.25191813707351685,
+      "learning_rate": 0.00015265765765765763,
+      "loss": 0.4146,
+      "step": 3281
+    },
+    {
+      "epoch": 0.9839235525576167,
+      "grad_norm": 0.2769797444343567,
+      "learning_rate": 0.00015261261261261257,
+      "loss": 0.4379,
+      "step": 3282
+    },
+    {
+      "epoch": 0.9842233464493161,
+      "grad_norm": 0.2423688769340515,
+      "learning_rate": 0.00015256756756756756,
+      "loss": 0.4221,
+      "step": 3283
+    },
+    {
+      "epoch": 0.9845231403410155,
+      "grad_norm": 0.2752641439437866,
+      "learning_rate": 0.0001525225225225225,
+      "loss": 0.4269,
+      "step": 3284
+    },
+    {
+      "epoch": 0.984822934232715,
+      "grad_norm": 0.27877339720726013,
+      "learning_rate": 0.00015247747747747746,
+      "loss": 0.4391,
+      "step": 3285
+    },
+    {
+      "epoch": 0.9851227281244145,
+      "grad_norm": 0.24187946319580078,
+      "learning_rate": 0.00015243243243243243,
+      "loss": 0.4196,
+      "step": 3286
+    },
+    {
+      "epoch": 0.9854225220161139,
+      "grad_norm": 0.24265043437480927,
+      "learning_rate": 0.00015238738738738736,
+      "loss": 0.4081,
+      "step": 3287
+    },
+    {
+      "epoch": 0.9857223159078133,
+      "grad_norm": 0.24394306540489197,
+      "learning_rate": 0.00015234234234234233,
+      "loss": 0.3875,
+      "step": 3288
+    },
+    {
+      "epoch": 0.9860221097995129,
+      "grad_norm": 0.23722290992736816,
+      "learning_rate": 0.0001522972972972973,
+      "loss": 0.3906,
+      "step": 3289
+    },
+    {
+      "epoch": 0.9863219036912123,
+      "grad_norm": 0.2603495121002197,
+      "learning_rate": 0.00015225225225225223,
+      "loss": 0.4435,
+      "step": 3290
+    },
+    {
+      "epoch": 0.9866216975829117,
+      "grad_norm": 0.23921896517276764,
+      "learning_rate": 0.0001522072072072072,
+      "loss": 0.4234,
+      "step": 3291
+    },
+    {
+      "epoch": 0.9869214914746112,
+      "grad_norm": 0.24632927775382996,
+      "learning_rate": 0.00015216216216216215,
+      "loss": 0.4411,
+      "step": 3292
+    },
+    {
+      "epoch": 0.9872212853663107,
+      "grad_norm": 0.2542307674884796,
+      "learning_rate": 0.0001521171171171171,
+      "loss": 0.4192,
+      "step": 3293
+    },
+    {
+      "epoch": 0.9875210792580101,
+      "grad_norm": 0.26318350434303284,
+      "learning_rate": 0.00015207207207207206,
+      "loss": 0.4286,
+      "step": 3294
+    },
+    {
+      "epoch": 0.9878208731497096,
+      "grad_norm": 0.2523062527179718,
+      "learning_rate": 0.00015202702702702702,
+      "loss": 0.42,
+      "step": 3295
+    },
+    {
+      "epoch": 0.988120667041409,
+      "grad_norm": 0.2566467225551605,
+      "learning_rate": 0.00015198198198198198,
+      "loss": 0.3847,
+      "step": 3296
+    },
+    {
+      "epoch": 0.9884204609331085,
+      "grad_norm": 0.2654307782649994,
+      "learning_rate": 0.00015193693693693692,
+      "loss": 0.4448,
+      "step": 3297
+    },
+    {
+      "epoch": 0.988720254824808,
+      "grad_norm": 0.25399646162986755,
+      "learning_rate": 0.00015189189189189188,
+      "loss": 0.4374,
+      "step": 3298
+    },
+    {
+      "epoch": 0.9890200487165074,
+      "grad_norm": 0.24990344047546387,
+      "learning_rate": 0.00015184684684684685,
+      "loss": 0.4365,
+      "step": 3299
+    },
+    {
+      "epoch": 0.9893198426082068,
+      "grad_norm": 0.24241049587726593,
+      "learning_rate": 0.00015180180180180178,
+      "loss": 0.399,
+      "step": 3300
+    },
+    {
+      "epoch": 0.9896196364999064,
+      "grad_norm": 0.24599871039390564,
+      "learning_rate": 0.00015175675675675675,
+      "loss": 0.4081,
+      "step": 3301
+    },
+    {
+      "epoch": 0.9899194303916058,
+      "grad_norm": 0.26060429215431213,
+      "learning_rate": 0.0001517117117117117,
+      "loss": 0.4247,
+      "step": 3302
+    },
+    {
+      "epoch": 0.9902192242833052,
+      "grad_norm": 0.2448931187391281,
+      "learning_rate": 0.00015166666666666665,
+      "loss": 0.4289,
+      "step": 3303
+    },
+    {
+      "epoch": 0.9905190181750047,
+      "grad_norm": 0.2747792899608612,
+      "learning_rate": 0.00015162162162162159,
+      "loss": 0.4398,
+      "step": 3304
+    },
+    {
+      "epoch": 0.9908188120667042,
+      "grad_norm": 0.29383328557014465,
+      "learning_rate": 0.00015157657657657658,
+      "loss": 0.4737,
+      "step": 3305
+    },
+    {
+      "epoch": 0.9911186059584036,
+      "grad_norm": 0.24790361523628235,
+      "learning_rate": 0.0001515315315315315,
+      "loss": 0.3991,
+      "step": 3306
+    },
+    {
+      "epoch": 0.991418399850103,
+      "grad_norm": 0.24991101026535034,
+      "learning_rate": 0.00015148648648648645,
+      "loss": 0.4146,
+      "step": 3307
+    },
+    {
+      "epoch": 0.9917181937418025,
+      "grad_norm": 0.2633141279220581,
+      "learning_rate": 0.00015144144144144144,
+      "loss": 0.4349,
+      "step": 3308
+    },
+    {
+      "epoch": 0.992017987633502,
+      "grad_norm": 0.25757932662963867,
+      "learning_rate": 0.00015139639639639638,
+      "loss": 0.417,
+      "step": 3309
+    },
+    {
+      "epoch": 0.9923177815252014,
+      "grad_norm": 0.2572961747646332,
+      "learning_rate": 0.00015135135135135131,
+      "loss": 0.413,
+      "step": 3310
+    },
+    {
+      "epoch": 0.9926175754169009,
+      "grad_norm": 0.2692258358001709,
+      "learning_rate": 0.0001513063063063063,
+      "loss": 0.4364,
+      "step": 3311
+    },
+    {
+      "epoch": 0.9929173693086003,
+      "grad_norm": 0.2538521885871887,
+      "learning_rate": 0.00015126126126126124,
+      "loss": 0.4182,
+      "step": 3312
+    },
+    {
+      "epoch": 0.9932171632002998,
+      "grad_norm": 0.2588602304458618,
+      "learning_rate": 0.00015121621621621618,
+      "loss": 0.4508,
+      "step": 3313
+    },
+    {
+      "epoch": 0.9935169570919993,
+      "grad_norm": 0.26588940620422363,
+      "learning_rate": 0.00015117117117117117,
+      "loss": 0.4345,
+      "step": 3314
+    },
+    {
+      "epoch": 0.9938167509836987,
+      "grad_norm": 0.24780629575252533,
+      "learning_rate": 0.0001511261261261261,
+      "loss": 0.4148,
+      "step": 3315
+    },
+    {
+      "epoch": 0.9941165448753981,
+      "grad_norm": 0.2527424693107605,
+      "learning_rate": 0.00015108108108108104,
+      "loss": 0.4369,
+      "step": 3316
+    },
+    {
+      "epoch": 0.9944163387670977,
+      "grad_norm": 0.24510394036769867,
+      "learning_rate": 0.00015103603603603603,
+      "loss": 0.3947,
+      "step": 3317
+    },
+    {
+      "epoch": 0.9947161326587971,
+      "grad_norm": 0.25519701838493347,
+      "learning_rate": 0.00015099099099099097,
+      "loss": 0.4395,
+      "step": 3318
+    },
+    {
+      "epoch": 0.9950159265504965,
+      "grad_norm": 0.2711610794067383,
+      "learning_rate": 0.00015094594594594594,
+      "loss": 0.4171,
+      "step": 3319
+    },
+    {
+      "epoch": 0.995315720442196,
+      "grad_norm": 0.25308021903038025,
+      "learning_rate": 0.0001509009009009009,
+      "loss": 0.4334,
+      "step": 3320
+    },
+    {
+      "epoch": 0.9956155143338955,
+      "grad_norm": 0.24517041444778442,
+      "learning_rate": 0.00015085585585585584,
+      "loss": 0.4122,
+      "step": 3321
+    },
+    {
+      "epoch": 0.9959153082255949,
+      "grad_norm": 0.26738440990448,
+      "learning_rate": 0.0001508108108108108,
+      "loss": 0.4221,
+      "step": 3322
+    },
+    {
+      "epoch": 0.9962151021172944,
+      "grad_norm": 0.2602112293243408,
+      "learning_rate": 0.00015076576576576576,
+      "loss": 0.4566,
+      "step": 3323
+    },
+    {
+      "epoch": 0.9965148960089938,
+      "grad_norm": 0.2651210129261017,
+      "learning_rate": 0.0001507207207207207,
+      "loss": 0.4169,
+      "step": 3324
+    },
+    {
+      "epoch": 0.9968146899006932,
+      "grad_norm": 0.23928852379322052,
+      "learning_rate": 0.00015067567567567566,
+      "loss": 0.4146,
+      "step": 3325
+    },
+    {
+      "epoch": 0.9971144837923928,
+      "grad_norm": 0.24811214208602905,
+      "learning_rate": 0.00015063063063063063,
+      "loss": 0.4192,
+      "step": 3326
+    },
+    {
+      "epoch": 0.9974142776840922,
+      "grad_norm": 0.24762240052223206,
+      "learning_rate": 0.00015058558558558556,
+      "loss": 0.4094,
+      "step": 3327
+    },
+    {
+      "epoch": 0.9977140715757916,
+      "grad_norm": 0.27879321575164795,
+      "learning_rate": 0.00015054054054054053,
+      "loss": 0.4221,
+      "step": 3328
+    },
+    {
+      "epoch": 0.998013865467491,
+      "grad_norm": 0.23711088299751282,
+      "learning_rate": 0.00015049549549549547,
+      "loss": 0.4065,
+      "step": 3329
+    },
+    {
+      "epoch": 0.9983136593591906,
+      "grad_norm": 0.2728164792060852,
+      "learning_rate": 0.00015045045045045046,
+      "loss": 0.4159,
+      "step": 3330
+    },
+    {
+      "epoch": 0.99861345325089,
+      "grad_norm": 0.24097463488578796,
+      "learning_rate": 0.0001504054054054054,
+      "loss": 0.4325,
+      "step": 3331
+    },
+    {
+      "epoch": 0.9989132471425894,
+      "grad_norm": 0.2429143637418747,
+      "learning_rate": 0.00015036036036036033,
+      "loss": 0.4285,
+      "step": 3332
+    },
+    {
+      "epoch": 0.9992130410342889,
+      "grad_norm": 0.24508728086948395,
+      "learning_rate": 0.00015031531531531532,
+      "loss": 0.4049,
+      "step": 3333
+    },
+    {
+      "epoch": 0.9995128349259884,
+      "grad_norm": 0.2516220510005951,
+      "learning_rate": 0.00015027027027027026,
+      "loss": 0.442,
+      "step": 3334
+    },
+    {
+      "epoch": 0.9998126288176878,
+      "grad_norm": 0.24752123653888702,
+      "learning_rate": 0.0001502252252252252,
+      "loss": 0.3956,
+      "step": 3335
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.3306880593299866,
+      "learning_rate": 0.00015018018018018019,
+      "loss": 0.4501,
+      "step": 3336
+    },
+    {
+      "epoch": 1.0002997938916995,
+      "grad_norm": 0.24475271999835968,
+      "learning_rate": 0.00015013513513513512,
+      "loss": 0.3665,
+      "step": 3337
+    },
+    {
+      "epoch": 1.0005995877833989,
+      "grad_norm": 0.263034850358963,
+      "learning_rate": 0.00015009009009009006,
+      "loss": 0.3607,
+      "step": 3338
+    },
+    {
+      "epoch": 1.0008993816750984,
+      "grad_norm": 0.22813622653484344,
+      "learning_rate": 0.00015004504504504505,
+      "loss": 0.3377,
+      "step": 3339
+    },
+    {
+      "epoch": 1.0011991755667977,
+      "grad_norm": 0.2284983992576599,
+      "learning_rate": 0.00015,
+      "loss": 0.3635,
+      "step": 3340
+    },
+    {
+      "epoch": 1.0014989694584973,
+      "grad_norm": 0.2420538365840912,
+      "learning_rate": 0.00014995495495495495,
+      "loss": 0.3567,
+      "step": 3341
+    },
+    {
+      "epoch": 1.0017987633501968,
+      "grad_norm": 0.24065744876861572,
+      "learning_rate": 0.0001499099099099099,
+      "loss": 0.3633,
+      "step": 3342
+    },
+    {
+      "epoch": 1.0020985572418961,
+      "grad_norm": 0.22252792119979858,
+      "learning_rate": 0.00014986486486486485,
+      "loss": 0.3598,
+      "step": 3343
+    },
+    {
+      "epoch": 1.0023983511335957,
+      "grad_norm": 0.2273607850074768,
+      "learning_rate": 0.00014981981981981982,
+      "loss": 0.3306,
+      "step": 3344
+    },
+    {
+      "epoch": 1.0026981450252952,
+      "grad_norm": 0.25264090299606323,
+      "learning_rate": 0.00014977477477477475,
+      "loss": 0.3911,
+      "step": 3345
+    },
+    {
+      "epoch": 1.0029979389169945,
+      "grad_norm": 0.2376517355442047,
+      "learning_rate": 0.00014972972972972972,
+      "loss": 0.355,
+      "step": 3346
+    },
+    {
+      "epoch": 1.003297732808694,
+      "grad_norm": 0.23404935002326965,
+      "learning_rate": 0.00014968468468468465,
+      "loss": 0.3946,
+      "step": 3347
+    },
+    {
+      "epoch": 1.0035975267003934,
+      "grad_norm": 0.23313717544078827,
+      "learning_rate": 0.00014963963963963962,
+      "loss": 0.3639,
+      "step": 3348
+    },
+    {
+      "epoch": 1.003897320592093,
+      "grad_norm": 0.25040316581726074,
+      "learning_rate": 0.00014959459459459458,
+      "loss": 0.3761,
+      "step": 3349
+    },
+    {
+      "epoch": 1.0041971144837925,
+      "grad_norm": 0.22730188071727753,
+      "learning_rate": 0.00014954954954954952,
+      "loss": 0.3625,
+      "step": 3350
+    },
+    {
+      "epoch": 1.0044969083754918,
+      "grad_norm": 0.22673781216144562,
+      "learning_rate": 0.00014950450450450448,
+      "loss": 0.3451,
+      "step": 3351
+    },
+    {
+      "epoch": 1.0047967022671913,
+      "grad_norm": 0.2615843713283539,
+      "learning_rate": 0.00014945945945945944,
+      "loss": 0.3629,
+      "step": 3352
+    },
+    {
+      "epoch": 1.0050964961588909,
+      "grad_norm": 0.2444346845149994,
+      "learning_rate": 0.0001494144144144144,
+      "loss": 0.3594,
+      "step": 3353
+    },
+    {
+      "epoch": 1.0053962900505902,
+      "grad_norm": 0.24913236498832703,
+      "learning_rate": 0.00014936936936936935,
+      "loss": 0.3615,
+      "step": 3354
+    },
+    {
+      "epoch": 1.0056960839422897,
+      "grad_norm": 0.2454083263874054,
+      "learning_rate": 0.0001493243243243243,
+      "loss": 0.3855,
+      "step": 3355
+    },
+    {
+      "epoch": 1.005995877833989,
+      "grad_norm": 0.2320137619972229,
+      "learning_rate": 0.00014927927927927927,
+      "loss": 0.3697,
+      "step": 3356
+    },
+    {
+      "epoch": 1.0062956717256886,
+      "grad_norm": 0.2395658791065216,
+      "learning_rate": 0.0001492342342342342,
+      "loss": 0.3744,
+      "step": 3357
+    },
+    {
+      "epoch": 1.006595465617388,
+      "grad_norm": 0.24167458713054657,
+      "learning_rate": 0.00014918918918918917,
+      "loss": 0.3782,
+      "step": 3358
+    },
+    {
+      "epoch": 1.0068952595090874,
+      "grad_norm": 0.2351546287536621,
+      "learning_rate": 0.00014914414414414414,
+      "loss": 0.3539,
+      "step": 3359
+    },
+    {
+      "epoch": 1.007195053400787,
+      "grad_norm": 0.249009907245636,
+      "learning_rate": 0.0001490990990990991,
+      "loss": 0.3881,
+      "step": 3360
+    },
+    {
+      "epoch": 1.0074948472924865,
+      "grad_norm": 0.2699730396270752,
+      "learning_rate": 0.00014905405405405404,
+      "loss": 0.3797,
+      "step": 3361
+    },
+    {
+      "epoch": 1.0077946411841858,
+      "grad_norm": 0.23312336206436157,
+      "learning_rate": 0.000149009009009009,
+      "loss": 0.3502,
+      "step": 3362
+    },
+    {
+      "epoch": 1.0080944350758854,
+      "grad_norm": 0.27450332045555115,
+      "learning_rate": 0.00014896396396396397,
+      "loss": 0.3715,
+      "step": 3363
+    },
+    {
+      "epoch": 1.0083942289675847,
+      "grad_norm": 0.2660040259361267,
+      "learning_rate": 0.0001489189189189189,
+      "loss": 0.4085,
+      "step": 3364
+    },
+    {
+      "epoch": 1.0086940228592842,
+      "grad_norm": 0.2341001033782959,
+      "learning_rate": 0.00014887387387387387,
+      "loss": 0.3729,
+      "step": 3365
+    },
+    {
+      "epoch": 1.0089938167509838,
+      "grad_norm": 0.27408385276794434,
+      "learning_rate": 0.00014882882882882883,
+      "loss": 0.4096,
+      "step": 3366
+    },
+    {
+      "epoch": 1.009293610642683,
+      "grad_norm": 0.2749234437942505,
+      "learning_rate": 0.00014878378378378377,
+      "loss": 0.3569,
+      "step": 3367
+    },
+    {
+      "epoch": 1.0095934045343826,
+      "grad_norm": 0.24223123490810394,
+      "learning_rate": 0.00014873873873873873,
+      "loss": 0.3605,
+      "step": 3368
+    },
+    {
+      "epoch": 1.0098931984260822,
+      "grad_norm": 0.2597927749156952,
+      "learning_rate": 0.0001486936936936937,
+      "loss": 0.3578,
+      "step": 3369
+    },
+    {
+      "epoch": 1.0101929923177815,
+      "grad_norm": 0.2509942054748535,
+      "learning_rate": 0.00014864864864864863,
+      "loss": 0.3782,
+      "step": 3370
+    },
+    {
+      "epoch": 1.010492786209481,
+      "grad_norm": 0.24225856363773346,
+      "learning_rate": 0.0001486036036036036,
+      "loss": 0.3745,
+      "step": 3371
+    },
+    {
+      "epoch": 1.0107925801011803,
+      "grad_norm": 0.23656389117240906,
+      "learning_rate": 0.00014855855855855853,
+      "loss": 0.3808,
+      "step": 3372
+    },
+    {
+      "epoch": 1.0110923739928799,
+      "grad_norm": 0.24159061908721924,
+      "learning_rate": 0.0001485135135135135,
+      "loss": 0.3841,
+      "step": 3373
+    },
+    {
+      "epoch": 1.0113921678845794,
+      "grad_norm": 0.23979109525680542,
+      "learning_rate": 0.00014846846846846846,
+      "loss": 0.3828,
+      "step": 3374
+    },
+    {
+      "epoch": 1.0116919617762787,
+      "grad_norm": 0.25807490944862366,
+      "learning_rate": 0.0001484234234234234,
+      "loss": 0.3494,
+      "step": 3375
+    },
+    {
+      "epoch": 1.0119917556679783,
+      "grad_norm": 0.23768498003482819,
+      "learning_rate": 0.00014837837837837836,
+      "loss": 0.3718,
+      "step": 3376
+    },
+    {
+      "epoch": 1.0122915495596778,
+      "grad_norm": 0.23397590219974518,
+      "learning_rate": 0.00014833333333333332,
+      "loss": 0.3737,
+      "step": 3377
+    },
+    {
+      "epoch": 1.0125913434513771,
+      "grad_norm": 0.26081061363220215,
+      "learning_rate": 0.00014828828828828826,
+      "loss": 0.3721,
+      "step": 3378
+    },
+    {
+      "epoch": 1.0128911373430767,
+      "grad_norm": 0.24995768070220947,
+      "learning_rate": 0.00014824324324324323,
+      "loss": 0.3839,
+      "step": 3379
+    },
+    {
+      "epoch": 1.013190931234776,
+      "grad_norm": 0.22104698419570923,
+      "learning_rate": 0.0001481981981981982,
+      "loss": 0.3325,
+      "step": 3380
+    },
+    {
+      "epoch": 1.0134907251264755,
+      "grad_norm": 0.2202545702457428,
+      "learning_rate": 0.00014815315315315313,
+      "loss": 0.3552,
+      "step": 3381
+    },
+    {
+      "epoch": 1.013790519018175,
+      "grad_norm": 0.2345467209815979,
+      "learning_rate": 0.0001481081081081081,
+      "loss": 0.3616,
+      "step": 3382
+    },
+    {
+      "epoch": 1.0140903129098744,
+      "grad_norm": 0.24063590168952942,
+      "learning_rate": 0.00014806306306306305,
+      "loss": 0.3837,
+      "step": 3383
+    },
+    {
+      "epoch": 1.014390106801574,
+      "grad_norm": 0.25035032629966736,
+      "learning_rate": 0.000148018018018018,
+      "loss": 0.3869,
+      "step": 3384
+    },
+    {
+      "epoch": 1.0146899006932735,
+      "grad_norm": 0.235214963555336,
+      "learning_rate": 0.00014797297297297295,
+      "loss": 0.3799,
+      "step": 3385
+    },
+    {
+      "epoch": 1.0149896945849728,
+      "grad_norm": 0.2631278336048126,
+      "learning_rate": 0.00014792792792792792,
+      "loss": 0.3736,
+      "step": 3386
+    },
+    {
+      "epoch": 1.0152894884766723,
+      "grad_norm": 0.23787197470664978,
+      "learning_rate": 0.00014788288288288288,
+      "loss": 0.3606,
+      "step": 3387
+    },
+    {
+      "epoch": 1.0155892823683716,
+      "grad_norm": 0.21479521691799164,
+      "learning_rate": 0.00014783783783783782,
+      "loss": 0.3588,
+      "step": 3388
+    },
+    {
+      "epoch": 1.0158890762600712,
+      "grad_norm": 0.25930336117744446,
+      "learning_rate": 0.00014779279279279278,
+      "loss": 0.3515,
+      "step": 3389
+    },
+    {
+      "epoch": 1.0161888701517707,
+      "grad_norm": 0.25570812821388245,
+      "learning_rate": 0.00014774774774774775,
+      "loss": 0.3626,
+      "step": 3390
+    },
+    {
+      "epoch": 1.01648866404347,
+      "grad_norm": 0.2151179164648056,
+      "learning_rate": 0.00014770270270270268,
+      "loss": 0.3366,
+      "step": 3391
+    },
+    {
+      "epoch": 1.0167884579351696,
+      "grad_norm": 0.24506697058677673,
+      "learning_rate": 0.00014765765765765765,
+      "loss": 0.3446,
+      "step": 3392
+    },
+    {
+      "epoch": 1.0170882518268691,
+      "grad_norm": 0.22405314445495605,
+      "learning_rate": 0.0001476126126126126,
+      "loss": 0.3493,
+      "step": 3393
+    },
+    {
+      "epoch": 1.0173880457185684,
+      "grad_norm": 0.24243111908435822,
+      "learning_rate": 0.00014756756756756758,
+      "loss": 0.3628,
+      "step": 3394
+    },
+    {
+      "epoch": 1.017687839610268,
+      "grad_norm": 0.2257484495639801,
+      "learning_rate": 0.0001475225225225225,
+      "loss": 0.3364,
+      "step": 3395
+    },
+    {
+      "epoch": 1.0179876335019673,
+      "grad_norm": 0.24955156445503235,
+      "learning_rate": 0.00014747747747747748,
+      "loss": 0.3629,
+      "step": 3396
+    },
+    {
+      "epoch": 1.0182874273936668,
+      "grad_norm": 0.2347087413072586,
+      "learning_rate": 0.0001474324324324324,
+      "loss": 0.3602,
+      "step": 3397
+    },
+    {
+      "epoch": 1.0185872212853664,
+      "grad_norm": 0.23353172838687897,
+      "learning_rate": 0.00014738738738738738,
+      "loss": 0.3715,
+      "step": 3398
+    },
+    {
+      "epoch": 1.0188870151770657,
+      "grad_norm": 0.2449275553226471,
+      "learning_rate": 0.00014734234234234234,
+      "loss": 0.38,
+      "step": 3399
+    },
+    {
+      "epoch": 1.0191868090687652,
+      "grad_norm": 0.2428295761346817,
+      "learning_rate": 0.00014729729729729728,
+      "loss": 0.3774,
+      "step": 3400
+    },
+    {
+      "epoch": 1.0194866029604648,
+      "grad_norm": 0.22880619764328003,
+      "learning_rate": 0.00014725225225225224,
+      "loss": 0.3718,
+      "step": 3401
+    },
+    {
+      "epoch": 1.019786396852164,
+      "grad_norm": 0.24155092239379883,
+      "learning_rate": 0.0001472072072072072,
+      "loss": 0.3724,
+      "step": 3402
+    },
+    {
+      "epoch": 1.0200861907438636,
+      "grad_norm": 0.2445833534002304,
+      "learning_rate": 0.00014716216216216214,
+      "loss": 0.3568,
+      "step": 3403
+    },
+    {
+      "epoch": 1.020385984635563,
+      "grad_norm": 0.2260163277387619,
+      "learning_rate": 0.0001471171171171171,
+      "loss": 0.3424,
+      "step": 3404
+    },
+    {
+      "epoch": 1.0206857785272625,
+      "grad_norm": 0.2485581338405609,
+      "learning_rate": 0.00014707207207207207,
+      "loss": 0.3806,
+      "step": 3405
+    },
+    {
+      "epoch": 1.020985572418962,
+      "grad_norm": 0.2234206348657608,
+      "learning_rate": 0.000147027027027027,
+      "loss": 0.3482,
+      "step": 3406
+    },
+    {
+      "epoch": 1.0212853663106614,
+      "grad_norm": 0.2420557290315628,
+      "learning_rate": 0.00014698198198198197,
+      "loss": 0.3747,
+      "step": 3407
+    },
+    {
+      "epoch": 1.021585160202361,
+      "grad_norm": 0.2375592589378357,
+      "learning_rate": 0.0001469369369369369,
+      "loss": 0.3455,
+      "step": 3408
+    },
+    {
+      "epoch": 1.0218849540940604,
+      "grad_norm": 0.21995361149311066,
+      "learning_rate": 0.00014689189189189187,
+      "loss": 0.3837,
+      "step": 3409
+    },
+    {
+      "epoch": 1.0221847479857598,
+      "grad_norm": 0.24082349240779877,
+      "learning_rate": 0.00014684684684684683,
+      "loss": 0.3594,
+      "step": 3410
+    },
+    {
+      "epoch": 1.0224845418774593,
+      "grad_norm": 0.24804848432540894,
+      "learning_rate": 0.00014680180180180177,
+      "loss": 0.3747,
+      "step": 3411
+    },
+    {
+      "epoch": 1.0227843357691586,
+      "grad_norm": 0.2306135892868042,
+      "learning_rate": 0.00014675675675675674,
+      "loss": 0.3665,
+      "step": 3412
+    },
+    {
+      "epoch": 1.0230841296608582,
+      "grad_norm": 0.23374010622501373,
+      "learning_rate": 0.0001467117117117117,
+      "loss": 0.3755,
+      "step": 3413
+    },
+    {
+      "epoch": 1.0233839235525577,
+      "grad_norm": 0.24325768649578094,
+      "learning_rate": 0.00014666666666666664,
+      "loss": 0.3761,
+      "step": 3414
+    },
+    {
+      "epoch": 1.023683717444257,
+      "grad_norm": 0.21893738210201263,
+      "learning_rate": 0.0001466216216216216,
+      "loss": 0.3527,
+      "step": 3415
+    },
+    {
+      "epoch": 1.0239835113359566,
+      "grad_norm": 0.22672438621520996,
+      "learning_rate": 0.00014657657657657656,
+      "loss": 0.3718,
+      "step": 3416
+    },
+    {
+      "epoch": 1.024283305227656,
+      "grad_norm": 0.24228642880916595,
+      "learning_rate": 0.00014653153153153153,
+      "loss": 0.3856,
+      "step": 3417
+    },
+    {
+      "epoch": 1.0245830991193554,
+      "grad_norm": 0.24045732617378235,
+      "learning_rate": 0.00014648648648648646,
+      "loss": 0.4047,
+      "step": 3418
+    },
+    {
+      "epoch": 1.024882893011055,
+      "grad_norm": 0.23358234763145447,
+      "learning_rate": 0.00014644144144144143,
+      "loss": 0.3719,
+      "step": 3419
+    },
+    {
+      "epoch": 1.0251826869027543,
+      "grad_norm": 0.22614139318466187,
+      "learning_rate": 0.0001463963963963964,
+      "loss": 0.3677,
+      "step": 3420
+    },
+    {
+      "epoch": 1.0254824807944538,
+      "grad_norm": 0.23576593399047852,
+      "learning_rate": 0.00014635135135135136,
+      "loss": 0.366,
+      "step": 3421
+    },
+    {
+      "epoch": 1.0257822746861534,
+      "grad_norm": 0.23267975449562073,
+      "learning_rate": 0.0001463063063063063,
+      "loss": 0.3725,
+      "step": 3422
+    },
+    {
+      "epoch": 1.0260820685778527,
+      "grad_norm": 0.23192216455936432,
+      "learning_rate": 0.00014626126126126126,
+      "loss": 0.3689,
+      "step": 3423
+    },
+    {
+      "epoch": 1.0263818624695522,
+      "grad_norm": 0.22963547706604004,
+      "learning_rate": 0.00014621621621621622,
+      "loss": 0.3804,
+      "step": 3424
+    },
+    {
+      "epoch": 1.0266816563612517,
+      "grad_norm": 0.22560270130634308,
+      "learning_rate": 0.00014617117117117116,
+      "loss": 0.3264,
+      "step": 3425
+    },
+    {
+      "epoch": 1.026981450252951,
+      "grad_norm": 0.24126774072647095,
+      "learning_rate": 0.00014612612612612612,
+      "loss": 0.3646,
+      "step": 3426
+    },
+    {
+      "epoch": 1.0272812441446506,
+      "grad_norm": 0.2244587242603302,
+      "learning_rate": 0.00014608108108108108,
+      "loss": 0.3728,
+      "step": 3427
+    },
+    {
+      "epoch": 1.02758103803635,
+      "grad_norm": 0.23782998323440552,
+      "learning_rate": 0.00014603603603603602,
+      "loss": 0.3628,
+      "step": 3428
+    },
+    {
+      "epoch": 1.0278808319280495,
+      "grad_norm": 0.22830158472061157,
+      "learning_rate": 0.00014599099099099099,
+      "loss": 0.3852,
+      "step": 3429
+    },
+    {
+      "epoch": 1.028180625819749,
+      "grad_norm": 0.23033784329891205,
+      "learning_rate": 0.00014594594594594595,
+      "loss": 0.367,
+      "step": 3430
+    },
+    {
+      "epoch": 1.0284804197114483,
+      "grad_norm": 0.22531138360500336,
+      "learning_rate": 0.00014590090090090089,
+      "loss": 0.3553,
+      "step": 3431
+    },
+    {
+      "epoch": 1.0287802136031479,
+      "grad_norm": 0.21490846574306488,
+      "learning_rate": 0.00014585585585585585,
+      "loss": 0.3541,
+      "step": 3432
+    },
+    {
+      "epoch": 1.0290800074948474,
+      "grad_norm": 0.23411285877227783,
+      "learning_rate": 0.0001458108108108108,
+      "loss": 0.3669,
+      "step": 3433
+    },
+    {
+      "epoch": 1.0293798013865467,
+      "grad_norm": 0.2277381718158722,
+      "learning_rate": 0.00014576576576576575,
+      "loss": 0.3566,
+      "step": 3434
+    },
+    {
+      "epoch": 1.0296795952782463,
+      "grad_norm": 0.24335390329360962,
+      "learning_rate": 0.00014572072072072071,
+      "loss": 0.3533,
+      "step": 3435
+    },
+    {
+      "epoch": 1.0299793891699456,
+      "grad_norm": 0.2254333347082138,
+      "learning_rate": 0.00014567567567567565,
+      "loss": 0.3484,
+      "step": 3436
+    },
+    {
+      "epoch": 1.0302791830616451,
+      "grad_norm": 0.22899997234344482,
+      "learning_rate": 0.00014563063063063062,
+      "loss": 0.3839,
+      "step": 3437
+    },
+    {
+      "epoch": 1.0305789769533447,
+      "grad_norm": 0.23717020452022552,
+      "learning_rate": 0.00014558558558558558,
+      "loss": 0.3697,
+      "step": 3438
+    },
+    {
+      "epoch": 1.030878770845044,
+      "grad_norm": 0.23469127714633942,
+      "learning_rate": 0.00014554054054054052,
+      "loss": 0.3912,
+      "step": 3439
+    },
+    {
+      "epoch": 1.0311785647367435,
+      "grad_norm": 0.21906235814094543,
+      "learning_rate": 0.00014549549549549548,
+      "loss": 0.3555,
+      "step": 3440
+    },
+    {
+      "epoch": 1.0314783586284428,
+      "grad_norm": 0.2294732928276062,
+      "learning_rate": 0.00014545045045045044,
+      "loss": 0.3521,
+      "step": 3441
+    },
+    {
+      "epoch": 1.0317781525201424,
+      "grad_norm": 0.22533249855041504,
+      "learning_rate": 0.00014540540540540538,
+      "loss": 0.3574,
+      "step": 3442
+    },
+    {
+      "epoch": 1.032077946411842,
+      "grad_norm": 0.21970278024673462,
+      "learning_rate": 0.00014536036036036034,
+      "loss": 0.3414,
+      "step": 3443
+    },
+    {
+      "epoch": 1.0323777403035412,
+      "grad_norm": 0.2248661071062088,
+      "learning_rate": 0.0001453153153153153,
+      "loss": 0.3603,
+      "step": 3444
+    },
+    {
+      "epoch": 1.0326775341952408,
+      "grad_norm": 0.237972229719162,
+      "learning_rate": 0.00014527027027027024,
+      "loss": 0.3716,
+      "step": 3445
+    },
+    {
+      "epoch": 1.0329773280869403,
+      "grad_norm": 0.22870191931724548,
+      "learning_rate": 0.0001452252252252252,
+      "loss": 0.3642,
+      "step": 3446
+    },
+    {
+      "epoch": 1.0332771219786396,
+      "grad_norm": 0.20847219228744507,
+      "learning_rate": 0.00014518018018018017,
+      "loss": 0.3238,
+      "step": 3447
+    },
+    {
+      "epoch": 1.0335769158703392,
+      "grad_norm": 0.22527335584163666,
+      "learning_rate": 0.0001451351351351351,
+      "loss": 0.3771,
+      "step": 3448
+    },
+    {
+      "epoch": 1.0338767097620385,
+      "grad_norm": 0.21900662779808044,
+      "learning_rate": 0.00014509009009009007,
+      "loss": 0.3759,
+      "step": 3449
+    },
+    {
+      "epoch": 1.034176503653738,
+      "grad_norm": 0.23750483989715576,
+      "learning_rate": 0.00014504504504504504,
+      "loss": 0.357,
+      "step": 3450
+    },
+    {
+      "epoch": 1.0344762975454376,
+      "grad_norm": 0.22595860064029694,
+      "learning_rate": 0.000145,
+      "loss": 0.3491,
+      "step": 3451
+    },
+    {
+      "epoch": 1.034776091437137,
+      "grad_norm": 0.22662843763828278,
+      "learning_rate": 0.00014495495495495494,
+      "loss": 0.3629,
+      "step": 3452
+    },
+    {
+      "epoch": 1.0350758853288364,
+      "grad_norm": 0.22160004079341888,
+      "learning_rate": 0.0001449099099099099,
+      "loss": 0.3508,
+      "step": 3453
+    },
+    {
+      "epoch": 1.035375679220536,
+      "grad_norm": 0.22288091480731964,
+      "learning_rate": 0.00014486486486486487,
+      "loss": 0.3681,
+      "step": 3454
+    },
+    {
+      "epoch": 1.0356754731122353,
+      "grad_norm": 0.2303716242313385,
+      "learning_rate": 0.0001448198198198198,
+      "loss": 0.3635,
+      "step": 3455
+    },
+    {
+      "epoch": 1.0359752670039348,
+      "grad_norm": 0.22402909398078918,
+      "learning_rate": 0.00014477477477477477,
+      "loss": 0.3712,
+      "step": 3456
+    },
+    {
+      "epoch": 1.0362750608956341,
+      "grad_norm": 0.2331233024597168,
+      "learning_rate": 0.00014472972972972973,
+      "loss": 0.369,
+      "step": 3457
+    },
+    {
+      "epoch": 1.0365748547873337,
+      "grad_norm": 0.2336449921131134,
+      "learning_rate": 0.00014468468468468467,
+      "loss": 0.3672,
+      "step": 3458
+    },
+    {
+      "epoch": 1.0368746486790332,
+      "grad_norm": 0.22813144326210022,
+      "learning_rate": 0.00014463963963963963,
+      "loss": 0.3692,
+      "step": 3459
+    },
+    {
+      "epoch": 1.0371744425707325,
+      "grad_norm": 0.23772816359996796,
+      "learning_rate": 0.0001445945945945946,
+      "loss": 0.3726,
+      "step": 3460
+    },
+    {
+      "epoch": 1.037474236462432,
+      "grad_norm": 0.23593769967556,
+      "learning_rate": 0.00014454954954954953,
+      "loss": 0.3857,
+      "step": 3461
+    },
+    {
+      "epoch": 1.0377740303541316,
+      "grad_norm": 0.2305610328912735,
+      "learning_rate": 0.0001445045045045045,
+      "loss": 0.3595,
+      "step": 3462
+    },
+    {
+      "epoch": 1.038073824245831,
+      "grad_norm": 0.22642318904399872,
+      "learning_rate": 0.00014445945945945946,
+      "loss": 0.3483,
+      "step": 3463
+    },
+    {
+      "epoch": 1.0383736181375305,
+      "grad_norm": 0.23105306923389435,
+      "learning_rate": 0.0001444144144144144,
+      "loss": 0.3714,
+      "step": 3464
+    },
+    {
+      "epoch": 1.0386734120292298,
+      "grad_norm": 0.2172197848558426,
+      "learning_rate": 0.00014436936936936936,
+      "loss": 0.3439,
+      "step": 3465
+    },
+    {
+      "epoch": 1.0389732059209293,
+      "grad_norm": 0.2232641875743866,
+      "learning_rate": 0.00014432432432432432,
+      "loss": 0.3552,
+      "step": 3466
+    },
+    {
+      "epoch": 1.0392729998126289,
+      "grad_norm": 0.21522095799446106,
+      "learning_rate": 0.00014427927927927926,
+      "loss": 0.3323,
+      "step": 3467
+    },
+    {
+      "epoch": 1.0395727937043282,
+      "grad_norm": 0.24464233219623566,
+      "learning_rate": 0.00014423423423423422,
+      "loss": 0.3632,
+      "step": 3468
+    },
+    {
+      "epoch": 1.0398725875960277,
+      "grad_norm": 0.21524198353290558,
+      "learning_rate": 0.00014418918918918916,
+      "loss": 0.3408,
+      "step": 3469
+    },
+    {
+      "epoch": 1.0401723814877273,
+      "grad_norm": 0.21136979758739471,
+      "learning_rate": 0.00014414414414414412,
+      "loss": 0.345,
+      "step": 3470
+    },
+    {
+      "epoch": 1.0404721753794266,
+      "grad_norm": 0.23737655580043793,
+      "learning_rate": 0.0001440990990990991,
+      "loss": 0.3643,
+      "step": 3471
+    },
+    {
+      "epoch": 1.0407719692711261,
+      "grad_norm": 0.22656260430812836,
+      "learning_rate": 0.00014405405405405403,
+      "loss": 0.356,
+      "step": 3472
+    },
+    {
+      "epoch": 1.0410717631628255,
+      "grad_norm": 0.24382594227790833,
+      "learning_rate": 0.000144009009009009,
+      "loss": 0.367,
+      "step": 3473
+    },
+    {
+      "epoch": 1.041371557054525,
+      "grad_norm": 0.2477714568376541,
+      "learning_rate": 0.00014396396396396395,
+      "loss": 0.3744,
+      "step": 3474
+    },
+    {
+      "epoch": 1.0416713509462245,
+      "grad_norm": 0.22636958956718445,
+      "learning_rate": 0.0001439189189189189,
+      "loss": 0.3599,
+      "step": 3475
+    },
+    {
+      "epoch": 1.0419711448379239,
+      "grad_norm": 0.24886015057563782,
+      "learning_rate": 0.00014387387387387385,
+      "loss": 0.3772,
+      "step": 3476
+    },
+    {
+      "epoch": 1.0422709387296234,
+      "grad_norm": 0.21825815737247467,
+      "learning_rate": 0.00014382882882882882,
+      "loss": 0.3517,
+      "step": 3477
+    },
+    {
+      "epoch": 1.042570732621323,
+      "grad_norm": 0.23218809068202972,
+      "learning_rate": 0.00014378378378378378,
+      "loss": 0.364,
+      "step": 3478
+    },
+    {
+      "epoch": 1.0428705265130223,
+      "grad_norm": 0.23707517981529236,
+      "learning_rate": 0.00014373873873873872,
+      "loss": 0.3711,
+      "step": 3479
+    },
+    {
+      "epoch": 1.0431703204047218,
+      "grad_norm": 0.2252931147813797,
+      "learning_rate": 0.00014369369369369368,
+      "loss": 0.3774,
+      "step": 3480
+    },
+    {
+      "epoch": 1.0434701142964211,
+      "grad_norm": 0.2177061289548874,
+      "learning_rate": 0.00014364864864864865,
+      "loss": 0.3448,
+      "step": 3481
+    },
+    {
+      "epoch": 1.0437699081881207,
+      "grad_norm": 0.22449122369289398,
+      "learning_rate": 0.00014360360360360358,
+      "loss": 0.3588,
+      "step": 3482
+    },
+    {
+      "epoch": 1.0440697020798202,
+      "grad_norm": 0.2354591339826584,
+      "learning_rate": 0.00014355855855855855,
+      "loss": 0.3743,
+      "step": 3483
+    },
+    {
+      "epoch": 1.0443694959715195,
+      "grad_norm": 0.2261870801448822,
+      "learning_rate": 0.0001435135135135135,
+      "loss": 0.3554,
+      "step": 3484
+    },
+    {
+      "epoch": 1.044669289863219,
+      "grad_norm": 0.23348073661327362,
+      "learning_rate": 0.00014346846846846847,
+      "loss": 0.387,
+      "step": 3485
+    },
+    {
+      "epoch": 1.0449690837549186,
+      "grad_norm": 0.23522013425827026,
+      "learning_rate": 0.0001434234234234234,
+      "loss": 0.3567,
+      "step": 3486
+    },
+    {
+      "epoch": 1.045268877646618,
+      "grad_norm": 0.2213398516178131,
+      "learning_rate": 0.00014337837837837837,
+      "loss": 0.3543,
+      "step": 3487
+    },
+    {
+      "epoch": 1.0455686715383175,
+      "grad_norm": 0.24790069460868835,
+      "learning_rate": 0.00014333333333333334,
+      "loss": 0.3769,
+      "step": 3488
+    },
+    {
+      "epoch": 1.0458684654300168,
+      "grad_norm": 0.23697051405906677,
+      "learning_rate": 0.00014328828828828828,
+      "loss": 0.3662,
+      "step": 3489
+    },
+    {
+      "epoch": 1.0461682593217163,
+      "grad_norm": 0.2302704155445099,
+      "learning_rate": 0.00014324324324324324,
+      "loss": 0.3761,
+      "step": 3490
+    },
+    {
+      "epoch": 1.0464680532134158,
+      "grad_norm": 0.22645635902881622,
+      "learning_rate": 0.00014319819819819818,
+      "loss": 0.3678,
+      "step": 3491
+    },
+    {
+      "epoch": 1.0467678471051152,
+      "grad_norm": 0.23438039422035217,
+      "learning_rate": 0.00014315315315315314,
+      "loss": 0.3633,
+      "step": 3492
+    },
+    {
+      "epoch": 1.0470676409968147,
+      "grad_norm": 0.23394618928432465,
+      "learning_rate": 0.0001431081081081081,
+      "loss": 0.3636,
+      "step": 3493
+    },
+    {
+      "epoch": 1.0473674348885142,
+      "grad_norm": 0.24473746120929718,
+      "learning_rate": 0.00014306306306306304,
+      "loss": 0.3769,
+      "step": 3494
+    },
+    {
+      "epoch": 1.0476672287802136,
+      "grad_norm": 0.22074007987976074,
+      "learning_rate": 0.000143018018018018,
+      "loss": 0.3611,
+      "step": 3495
+    },
+    {
+      "epoch": 1.047967022671913,
+      "grad_norm": 0.2372681349515915,
+      "learning_rate": 0.00014297297297297297,
+      "loss": 0.3639,
+      "step": 3496
+    },
+    {
+      "epoch": 1.0482668165636124,
+      "grad_norm": 0.2267588973045349,
+      "learning_rate": 0.0001429279279279279,
+      "loss": 0.3734,
+      "step": 3497
+    },
+    {
+      "epoch": 1.048566610455312,
+      "grad_norm": 0.22533351182937622,
+      "learning_rate": 0.00014288288288288287,
+      "loss": 0.3763,
+      "step": 3498
+    },
+    {
+      "epoch": 1.0488664043470115,
+      "grad_norm": 0.22511237859725952,
+      "learning_rate": 0.00014283783783783783,
+      "loss": 0.3501,
+      "step": 3499
+    },
+    {
+      "epoch": 1.0491661982387108,
+      "grad_norm": 0.2493421882390976,
+      "learning_rate": 0.00014279279279279277,
+      "loss": 0.3953,
+      "step": 3500
+    },
+    {
+      "epoch": 1.0491661982387108,
+      "eval_loss": 0.42186981439590454,
+      "eval_runtime": 565.6283,
+      "eval_samples_per_second": 3.817,
+      "eval_steps_per_second": 0.477,
+      "step": 3500
+    },
+    {
+      "epoch": 1.0494659921304104,
+      "grad_norm": 0.22452357411384583,
+      "learning_rate": 0.00014274774774774773,
+      "loss": 0.3531,
+      "step": 3501
+    },
+    {
+      "epoch": 1.04976578602211,
+      "grad_norm": 0.2269439995288849,
+      "learning_rate": 0.0001427027027027027,
+      "loss": 0.3756,
+      "step": 3502
+    },
+    {
+      "epoch": 1.0500655799138092,
+      "grad_norm": 0.2126823365688324,
+      "learning_rate": 0.00014265765765765763,
+      "loss": 0.3563,
+      "step": 3503
+    },
+    {
+      "epoch": 1.0503653738055088,
+      "grad_norm": 0.2116537094116211,
+      "learning_rate": 0.0001426126126126126,
+      "loss": 0.3619,
+      "step": 3504
+    },
+    {
+      "epoch": 1.050665167697208,
+      "grad_norm": 0.2395385056734085,
+      "learning_rate": 0.00014256756756756753,
+      "loss": 0.3746,
+      "step": 3505
+    },
+    {
+      "epoch": 1.0509649615889076,
+      "grad_norm": 0.22017928957939148,
+      "learning_rate": 0.0001425225225225225,
+      "loss": 0.3582,
+      "step": 3506
+    },
+    {
+      "epoch": 1.0512647554806072,
+      "grad_norm": 0.22230744361877441,
+      "learning_rate": 0.00014247747747747746,
+      "loss": 0.3367,
+      "step": 3507
+    },
+    {
+      "epoch": 1.0515645493723065,
+      "grad_norm": 0.22529073059558868,
+      "learning_rate": 0.00014243243243243243,
+      "loss": 0.3423,
+      "step": 3508
+    },
+    {
+      "epoch": 1.051864343264006,
+      "grad_norm": 0.24858680367469788,
+      "learning_rate": 0.00014238738738738736,
+      "loss": 0.4004,
+      "step": 3509
+    },
+    {
+      "epoch": 1.0521641371557056,
+      "grad_norm": 0.22725269198417664,
+      "learning_rate": 0.00014234234234234233,
+      "loss": 0.3673,
+      "step": 3510
+    },
+    {
+      "epoch": 1.0524639310474049,
+      "grad_norm": 0.2208700180053711,
+      "learning_rate": 0.0001422972972972973,
+      "loss": 0.3492,
+      "step": 3511
+    },
+    {
+      "epoch": 1.0527637249391044,
+      "grad_norm": 0.253497451543808,
+      "learning_rate": 0.00014225225225225223,
+      "loss": 0.3884,
+      "step": 3512
+    },
+    {
+      "epoch": 1.0530635188308037,
+      "grad_norm": 0.2408270537853241,
+      "learning_rate": 0.0001422072072072072,
+      "loss": 0.3813,
+      "step": 3513
+    },
+    {
+      "epoch": 1.0533633127225033,
+      "grad_norm": 0.21874786913394928,
+      "learning_rate": 0.00014216216216216216,
+      "loss": 0.3291,
+      "step": 3514
+    },
+    {
+      "epoch": 1.0536631066142028,
+      "grad_norm": 0.2416338324546814,
+      "learning_rate": 0.00014211711711711712,
+      "loss": 0.3801,
+      "step": 3515
+    },
+    {
+      "epoch": 1.0539629005059021,
+      "grad_norm": 0.2278902381658554,
+      "learning_rate": 0.00014207207207207206,
+      "loss": 0.3547,
+      "step": 3516
+    },
+    {
+      "epoch": 1.0542626943976017,
+      "grad_norm": 0.2442399263381958,
+      "learning_rate": 0.00014202702702702702,
+      "loss": 0.3818,
+      "step": 3517
+    },
+    {
+      "epoch": 1.0545624882893012,
+      "grad_norm": 0.2202155590057373,
+      "learning_rate": 0.00014198198198198198,
+      "loss": 0.3486,
+      "step": 3518
+    },
+    {
+      "epoch": 1.0548622821810005,
+      "grad_norm": 0.22215566039085388,
+      "learning_rate": 0.00014193693693693692,
+      "loss": 0.3729,
+      "step": 3519
+    },
+    {
+      "epoch": 1.0551620760727,
+      "grad_norm": 0.23282161355018616,
+      "learning_rate": 0.00014189189189189188,
+      "loss": 0.3695,
+      "step": 3520
+    },
+    {
+      "epoch": 1.0554618699643994,
+      "grad_norm": 0.21966330707073212,
+      "learning_rate": 0.00014184684684684685,
+      "loss": 0.3463,
+      "step": 3521
+    },
+    {
+      "epoch": 1.055761663856099,
+      "grad_norm": 0.2348901331424713,
+      "learning_rate": 0.00014180180180180179,
+      "loss": 0.3836,
+      "step": 3522
+    },
+    {
+      "epoch": 1.0560614577477985,
+      "grad_norm": 0.23036788403987885,
+      "learning_rate": 0.00014175675675675675,
+      "loss": 0.3875,
+      "step": 3523
+    },
+    {
+      "epoch": 1.0563612516394978,
+      "grad_norm": 0.2271084189414978,
+      "learning_rate": 0.0001417117117117117,
+      "loss": 0.3631,
+      "step": 3524
+    },
+    {
+      "epoch": 1.0566610455311973,
+      "grad_norm": 0.23065337538719177,
+      "learning_rate": 0.00014166666666666665,
+      "loss": 0.384,
+      "step": 3525
+    },
+    {
+      "epoch": 1.0569608394228966,
+      "grad_norm": 0.23186425864696503,
+      "learning_rate": 0.00014162162162162161,
+      "loss": 0.3733,
+      "step": 3526
+    },
+    {
+      "epoch": 1.0572606333145962,
+      "grad_norm": 0.23528079688549042,
+      "learning_rate": 0.00014157657657657658,
+      "loss": 0.377,
+      "step": 3527
+    },
+    {
+      "epoch": 1.0575604272062957,
+      "grad_norm": 0.24329833686351776,
+      "learning_rate": 0.00014153153153153151,
+      "loss": 0.3605,
+      "step": 3528
+    },
+    {
+      "epoch": 1.057860221097995,
+      "grad_norm": 0.24599115550518036,
+      "learning_rate": 0.00014148648648648648,
+      "loss": 0.3859,
+      "step": 3529
+    },
+    {
+      "epoch": 1.0581600149896946,
+      "grad_norm": 0.22416117787361145,
+      "learning_rate": 0.00014144144144144141,
+      "loss": 0.3788,
+      "step": 3530
+    },
+    {
+      "epoch": 1.0584598088813941,
+      "grad_norm": 0.23425617814064026,
+      "learning_rate": 0.00014139639639639638,
+      "loss": 0.3745,
+      "step": 3531
+    },
+    {
+      "epoch": 1.0587596027730934,
+      "grad_norm": 0.2239811271429062,
+      "learning_rate": 0.00014135135135135134,
+      "loss": 0.3617,
+      "step": 3532
+    },
+    {
+      "epoch": 1.059059396664793,
+      "grad_norm": 0.2348106950521469,
+      "learning_rate": 0.00014130630630630628,
+      "loss": 0.3637,
+      "step": 3533
+    },
+    {
+      "epoch": 1.0593591905564925,
+      "grad_norm": 0.2254401445388794,
+      "learning_rate": 0.00014126126126126124,
+      "loss": 0.3395,
+      "step": 3534
+    },
+    {
+      "epoch": 1.0596589844481918,
+      "grad_norm": 0.22252045571804047,
+      "learning_rate": 0.0001412162162162162,
+      "loss": 0.3716,
+      "step": 3535
+    },
+    {
+      "epoch": 1.0599587783398914,
+      "grad_norm": 0.2396722286939621,
+      "learning_rate": 0.00014117117117117114,
+      "loss": 0.3808,
+      "step": 3536
+    },
+    {
+      "epoch": 1.0602585722315907,
+      "grad_norm": 0.24962125718593597,
+      "learning_rate": 0.0001411261261261261,
+      "loss": 0.3934,
+      "step": 3537
+    },
+    {
+      "epoch": 1.0605583661232902,
+      "grad_norm": 0.2402314692735672,
+      "learning_rate": 0.00014108108108108107,
+      "loss": 0.3801,
+      "step": 3538
+    },
+    {
+      "epoch": 1.0608581600149898,
+      "grad_norm": 0.22211486101150513,
+      "learning_rate": 0.000141036036036036,
+      "loss": 0.3721,
+      "step": 3539
+    },
+    {
+      "epoch": 1.061157953906689,
+      "grad_norm": 0.22576668858528137,
+      "learning_rate": 0.00014099099099099097,
+      "loss": 0.3477,
+      "step": 3540
+    },
+    {
+      "epoch": 1.0614577477983886,
+      "grad_norm": 0.2367440015077591,
+      "learning_rate": 0.00014094594594594594,
+      "loss": 0.3793,
+      "step": 3541
+    },
+    {
+      "epoch": 1.061757541690088,
+      "grad_norm": 0.21483252942562103,
+      "learning_rate": 0.0001409009009009009,
+      "loss": 0.3429,
+      "step": 3542
+    },
+    {
+      "epoch": 1.0620573355817875,
+      "grad_norm": 0.2304966300725937,
+      "learning_rate": 0.00014085585585585584,
+      "loss": 0.3603,
+      "step": 3543
+    },
+    {
+      "epoch": 1.062357129473487,
+      "grad_norm": 0.24434131383895874,
+      "learning_rate": 0.0001408108108108108,
+      "loss": 0.3763,
+      "step": 3544
+    },
+    {
+      "epoch": 1.0626569233651864,
+      "grad_norm": 0.2312142550945282,
+      "learning_rate": 0.00014076576576576576,
+      "loss": 0.3856,
+      "step": 3545
+    },
+    {
+      "epoch": 1.062956717256886,
+      "grad_norm": 0.22286604344844818,
+      "learning_rate": 0.0001407207207207207,
+      "loss": 0.3367,
+      "step": 3546
+    },
+    {
+      "epoch": 1.0632565111485854,
+      "grad_norm": 0.2351447492837906,
+      "learning_rate": 0.00014067567567567567,
+      "loss": 0.3725,
+      "step": 3547
+    },
+    {
+      "epoch": 1.0635563050402848,
+      "grad_norm": 0.22424407303333282,
+      "learning_rate": 0.00014063063063063063,
+      "loss": 0.3638,
+      "step": 3548
+    },
+    {
+      "epoch": 1.0638560989319843,
+      "grad_norm": 0.22963954508304596,
+      "learning_rate": 0.0001405855855855856,
+      "loss": 0.3613,
+      "step": 3549
+    },
+    {
+      "epoch": 1.0641558928236836,
+      "grad_norm": 0.23109790682792664,
+      "learning_rate": 0.00014054054054054053,
+      "loss": 0.369,
+      "step": 3550
+    },
+    {
+      "epoch": 1.0644556867153832,
+      "grad_norm": 0.2307424396276474,
+      "learning_rate": 0.0001404954954954955,
+      "loss": 0.3688,
+      "step": 3551
+    },
+    {
+      "epoch": 1.0647554806070827,
+      "grad_norm": 0.21975605189800262,
+      "learning_rate": 0.00014045045045045043,
+      "loss": 0.3301,
+      "step": 3552
+    },
+    {
+      "epoch": 1.065055274498782,
+      "grad_norm": 0.2344069480895996,
+      "learning_rate": 0.0001404054054054054,
+      "loss": 0.4069,
+      "step": 3553
+    },
+    {
+      "epoch": 1.0653550683904816,
+      "grad_norm": 0.22718745470046997,
+      "learning_rate": 0.00014036036036036036,
+      "loss": 0.3613,
+      "step": 3554
+    },
+    {
+      "epoch": 1.065654862282181,
+      "grad_norm": 0.22939084470272064,
+      "learning_rate": 0.0001403153153153153,
+      "loss": 0.3678,
+      "step": 3555
+    },
+    {
+      "epoch": 1.0659546561738804,
+      "grad_norm": 0.2331932783126831,
+      "learning_rate": 0.00014027027027027026,
+      "loss": 0.3744,
+      "step": 3556
+    },
+    {
+      "epoch": 1.06625445006558,
+      "grad_norm": 0.24038521945476532,
+      "learning_rate": 0.00014022522522522522,
+      "loss": 0.3608,
+      "step": 3557
+    },
+    {
+      "epoch": 1.0665542439572793,
+      "grad_norm": 0.24839916825294495,
+      "learning_rate": 0.00014018018018018016,
+      "loss": 0.369,
+      "step": 3558
+    },
+    {
+      "epoch": 1.0668540378489788,
+      "grad_norm": 0.22673757374286652,
+      "learning_rate": 0.00014013513513513512,
+      "loss": 0.3573,
+      "step": 3559
+    },
+    {
+      "epoch": 1.0671538317406783,
+      "grad_norm": 0.23155052959918976,
+      "learning_rate": 0.0001400900900900901,
+      "loss": 0.3507,
+      "step": 3560
+    },
+    {
+      "epoch": 1.0674536256323777,
+      "grad_norm": 0.23616757988929749,
+      "learning_rate": 0.00014004504504504502,
+      "loss": 0.3675,
+      "step": 3561
+    },
+    {
+      "epoch": 1.0677534195240772,
+      "grad_norm": 0.22857490181922913,
+      "learning_rate": 0.00014,
+      "loss": 0.3508,
+      "step": 3562
+    },
+    {
+      "epoch": 1.0680532134157767,
+      "grad_norm": 0.23677614331245422,
+      "learning_rate": 0.00013995495495495495,
+      "loss": 0.3545,
+      "step": 3563
+    },
+    {
+      "epoch": 1.068353007307476,
+      "grad_norm": 0.23888587951660156,
+      "learning_rate": 0.0001399099099099099,
+      "loss": 0.3679,
+      "step": 3564
+    },
+    {
+      "epoch": 1.0686528011991756,
+      "grad_norm": 0.24625924229621887,
+      "learning_rate": 0.00013986486486486485,
+      "loss": 0.349,
+      "step": 3565
+    },
+    {
+      "epoch": 1.068952595090875,
+      "grad_norm": 0.24238745868206024,
+      "learning_rate": 0.0001398198198198198,
+      "loss": 0.3867,
+      "step": 3566
+    },
+    {
+      "epoch": 1.0692523889825745,
+      "grad_norm": 0.2739328444004059,
+      "learning_rate": 0.00013977477477477475,
+      "loss": 0.395,
+      "step": 3567
+    },
+    {
+      "epoch": 1.069552182874274,
+      "grad_norm": 0.24995070695877075,
+      "learning_rate": 0.00013972972972972972,
+      "loss": 0.3247,
+      "step": 3568
+    },
+    {
+      "epoch": 1.0698519767659733,
+      "grad_norm": 0.2289031744003296,
+      "learning_rate": 0.00013968468468468465,
+      "loss": 0.3629,
+      "step": 3569
+    },
+    {
+      "epoch": 1.0701517706576729,
+      "grad_norm": 0.23290695250034332,
+      "learning_rate": 0.00013963963963963962,
+      "loss": 0.3619,
+      "step": 3570
+    },
+    {
+      "epoch": 1.0704515645493724,
+      "grad_norm": 0.250203013420105,
+      "learning_rate": 0.00013959459459459458,
+      "loss": 0.3592,
+      "step": 3571
+    },
+    {
+      "epoch": 1.0707513584410717,
+      "grad_norm": 0.24303412437438965,
+      "learning_rate": 0.00013954954954954955,
+      "loss": 0.3439,
+      "step": 3572
+    },
+    {
+      "epoch": 1.0710511523327713,
+      "grad_norm": 0.23078800737857819,
+      "learning_rate": 0.00013950450450450448,
+      "loss": 0.3933,
+      "step": 3573
+    },
+    {
+      "epoch": 1.0713509462244706,
+      "grad_norm": 0.23219646513462067,
+      "learning_rate": 0.00013945945945945945,
+      "loss": 0.3565,
+      "step": 3574
+    },
+    {
+      "epoch": 1.0716507401161701,
+      "grad_norm": 0.25298643112182617,
+      "learning_rate": 0.0001394144144144144,
+      "loss": 0.3848,
+      "step": 3575
+    },
+    {
+      "epoch": 1.0719505340078697,
+      "grad_norm": 0.23267370462417603,
+      "learning_rate": 0.00013936936936936937,
+      "loss": 0.3515,
+      "step": 3576
+    },
+    {
+      "epoch": 1.072250327899569,
+      "grad_norm": 0.23744705319404602,
+      "learning_rate": 0.0001393243243243243,
+      "loss": 0.3963,
+      "step": 3577
+    },
+    {
+      "epoch": 1.0725501217912685,
+      "grad_norm": 0.23229120671749115,
+      "learning_rate": 0.00013927927927927927,
+      "loss": 0.3395,
+      "step": 3578
+    },
+    {
+      "epoch": 1.072849915682968,
+      "grad_norm": 0.25329065322875977,
+      "learning_rate": 0.00013923423423423424,
+      "loss": 0.352,
+      "step": 3579
+    },
+    {
+      "epoch": 1.0731497095746674,
+      "grad_norm": 0.22956810891628265,
+      "learning_rate": 0.00013918918918918917,
+      "loss": 0.3792,
+      "step": 3580
+    },
+    {
+      "epoch": 1.073449503466367,
+      "grad_norm": 0.23131117224693298,
+      "learning_rate": 0.00013914414414414414,
+      "loss": 0.3503,
+      "step": 3581
+    },
+    {
+      "epoch": 1.0737492973580662,
+      "grad_norm": 0.22814302146434784,
+      "learning_rate": 0.0001390990990990991,
+      "loss": 0.3675,
+      "step": 3582
+    },
+    {
+      "epoch": 1.0740490912497658,
+      "grad_norm": 0.23262296617031097,
+      "learning_rate": 0.00013905405405405404,
+      "loss": 0.3607,
+      "step": 3583
+    },
+    {
+      "epoch": 1.0743488851414653,
+      "grad_norm": 0.23462675511837006,
+      "learning_rate": 0.000139009009009009,
+      "loss": 0.3831,
+      "step": 3584
+    },
+    {
+      "epoch": 1.0746486790331646,
+      "grad_norm": 0.24443736672401428,
+      "learning_rate": 0.00013896396396396397,
+      "loss": 0.3735,
+      "step": 3585
+    },
+    {
+      "epoch": 1.0749484729248642,
+      "grad_norm": 0.22071777284145355,
+      "learning_rate": 0.0001389189189189189,
+      "loss": 0.3607,
+      "step": 3586
+    },
+    {
+      "epoch": 1.0752482668165637,
+      "grad_norm": 0.25556886196136475,
+      "learning_rate": 0.00013887387387387387,
+      "loss": 0.383,
+      "step": 3587
+    },
+    {
+      "epoch": 1.075548060708263,
+      "grad_norm": 0.23284411430358887,
+      "learning_rate": 0.0001388288288288288,
+      "loss": 0.376,
+      "step": 3588
+    },
+    {
+      "epoch": 1.0758478545999626,
+      "grad_norm": 0.22534599900245667,
+      "learning_rate": 0.00013878378378378377,
+      "loss": 0.3647,
+      "step": 3589
+    },
+    {
+      "epoch": 1.076147648491662,
+      "grad_norm": 0.22576001286506653,
+      "learning_rate": 0.00013873873873873873,
+      "loss": 0.356,
+      "step": 3590
+    },
+    {
+      "epoch": 1.0764474423833614,
+      "grad_norm": 0.231540247797966,
+      "learning_rate": 0.00013869369369369367,
+      "loss": 0.3917,
+      "step": 3591
+    },
+    {
+      "epoch": 1.076747236275061,
+      "grad_norm": 0.22077928483486176,
+      "learning_rate": 0.00013864864864864863,
+      "loss": 0.3639,
+      "step": 3592
+    },
+    {
+      "epoch": 1.0770470301667603,
+      "grad_norm": 0.22737953066825867,
+      "learning_rate": 0.0001386036036036036,
+      "loss": 0.3544,
+      "step": 3593
+    },
+    {
+      "epoch": 1.0773468240584598,
+      "grad_norm": 0.250492125749588,
+      "learning_rate": 0.00013855855855855853,
+      "loss": 0.3821,
+      "step": 3594
+    },
+    {
+      "epoch": 1.0776466179501594,
+      "grad_norm": 0.23965615034103394,
+      "learning_rate": 0.0001385135135135135,
+      "loss": 0.3623,
+      "step": 3595
+    },
+    {
+      "epoch": 1.0779464118418587,
+      "grad_norm": 0.24302807450294495,
+      "learning_rate": 0.00013846846846846846,
+      "loss": 0.38,
+      "step": 3596
+    },
+    {
+      "epoch": 1.0782462057335582,
+      "grad_norm": 0.2204785794019699,
+      "learning_rate": 0.0001384234234234234,
+      "loss": 0.3435,
+      "step": 3597
+    },
+    {
+      "epoch": 1.0785459996252575,
+      "grad_norm": 0.2180601954460144,
+      "learning_rate": 0.00013837837837837836,
+      "loss": 0.3477,
+      "step": 3598
+    },
+    {
+      "epoch": 1.078845793516957,
+      "grad_norm": 0.22917163372039795,
+      "learning_rate": 0.00013833333333333333,
+      "loss": 0.3551,
+      "step": 3599
+    },
+    {
+      "epoch": 1.0791455874086566,
+      "grad_norm": 0.24759288132190704,
+      "learning_rate": 0.00013828828828828826,
+      "loss": 0.3666,
+      "step": 3600
+    },
+    {
+      "epoch": 1.079445381300356,
+      "grad_norm": 0.22639593482017517,
+      "learning_rate": 0.00013824324324324323,
+      "loss": 0.3419,
+      "step": 3601
+    },
+    {
+      "epoch": 1.0797451751920555,
+      "grad_norm": 0.24272778630256653,
+      "learning_rate": 0.0001381981981981982,
+      "loss": 0.3755,
+      "step": 3602
+    },
+    {
+      "epoch": 1.080044969083755,
+      "grad_norm": 0.23822729289531708,
+      "learning_rate": 0.00013815315315315313,
+      "loss": 0.3642,
+      "step": 3603
+    },
+    {
+      "epoch": 1.0803447629754543,
+      "grad_norm": 0.25243374705314636,
+      "learning_rate": 0.0001381081081081081,
+      "loss": 0.3754,
+      "step": 3604
+    },
+    {
+      "epoch": 1.0806445568671539,
+      "grad_norm": 0.24384267628192902,
+      "learning_rate": 0.00013806306306306305,
+      "loss": 0.3727,
+      "step": 3605
+    },
+    {
+      "epoch": 1.0809443507588532,
+      "grad_norm": 0.2210264801979065,
+      "learning_rate": 0.00013801801801801802,
+      "loss": 0.3527,
+      "step": 3606
+    },
+    {
+      "epoch": 1.0812441446505527,
+      "grad_norm": 0.23943836987018585,
+      "learning_rate": 0.00013797297297297296,
+      "loss": 0.3624,
+      "step": 3607
+    },
+    {
+      "epoch": 1.0815439385422523,
+      "grad_norm": 0.2331855595111847,
+      "learning_rate": 0.00013792792792792792,
+      "loss": 0.3677,
+      "step": 3608
+    },
+    {
+      "epoch": 1.0818437324339516,
+      "grad_norm": 0.23624837398529053,
+      "learning_rate": 0.00013788288288288288,
+      "loss": 0.3539,
+      "step": 3609
+    },
+    {
+      "epoch": 1.0821435263256511,
+      "grad_norm": 0.24074693024158478,
+      "learning_rate": 0.00013783783783783782,
+      "loss": 0.3615,
+      "step": 3610
+    },
+    {
+      "epoch": 1.0824433202173505,
+      "grad_norm": 0.22781312465667725,
+      "learning_rate": 0.00013779279279279278,
+      "loss": 0.3633,
+      "step": 3611
+    },
+    {
+      "epoch": 1.08274311410905,
+      "grad_norm": 0.2411901205778122,
+      "learning_rate": 0.00013774774774774775,
+      "loss": 0.352,
+      "step": 3612
+    },
+    {
+      "epoch": 1.0830429080007495,
+      "grad_norm": 0.24481438100337982,
+      "learning_rate": 0.00013770270270270268,
+      "loss": 0.353,
+      "step": 3613
+    },
+    {
+      "epoch": 1.0833427018924489,
+      "grad_norm": 0.2313372939825058,
+      "learning_rate": 0.00013765765765765765,
+      "loss": 0.3857,
+      "step": 3614
+    },
+    {
+      "epoch": 1.0836424957841484,
+      "grad_norm": 0.23338955640792847,
+      "learning_rate": 0.0001376126126126126,
+      "loss": 0.3852,
+      "step": 3615
+    },
+    {
+      "epoch": 1.083942289675848,
+      "grad_norm": 0.24845603108406067,
+      "learning_rate": 0.00013756756756756755,
+      "loss": 0.3418,
+      "step": 3616
+    },
+    {
+      "epoch": 1.0842420835675473,
+      "grad_norm": 0.22615694999694824,
+      "learning_rate": 0.0001375225225225225,
+      "loss": 0.3652,
+      "step": 3617
+    },
+    {
+      "epoch": 1.0845418774592468,
+      "grad_norm": 0.21894417703151703,
+      "learning_rate": 0.00013747747747747748,
+      "loss": 0.3343,
+      "step": 3618
+    },
+    {
+      "epoch": 1.0848416713509463,
+      "grad_norm": 0.23376759886741638,
+      "learning_rate": 0.0001374324324324324,
+      "loss": 0.3806,
+      "step": 3619
+    },
+    {
+      "epoch": 1.0851414652426457,
+      "grad_norm": 0.24642081558704376,
+      "learning_rate": 0.00013738738738738738,
+      "loss": 0.3869,
+      "step": 3620
+    },
+    {
+      "epoch": 1.0854412591343452,
+      "grad_norm": 0.2276991456747055,
+      "learning_rate": 0.00013734234234234234,
+      "loss": 0.3676,
+      "step": 3621
+    },
+    {
+      "epoch": 1.0857410530260445,
+      "grad_norm": 0.22866936028003693,
+      "learning_rate": 0.00013729729729729728,
+      "loss": 0.3566,
+      "step": 3622
+    },
+    {
+      "epoch": 1.086040846917744,
+      "grad_norm": 0.23887404799461365,
+      "learning_rate": 0.00013725225225225224,
+      "loss": 0.3738,
+      "step": 3623
+    },
+    {
+      "epoch": 1.0863406408094436,
+      "grad_norm": 0.23940208554267883,
+      "learning_rate": 0.00013720720720720718,
+      "loss": 0.3732,
+      "step": 3624
+    },
+    {
+      "epoch": 1.086640434701143,
+      "grad_norm": 0.2483586072921753,
+      "learning_rate": 0.00013716216216216214,
+      "loss": 0.3778,
+      "step": 3625
+    },
+    {
+      "epoch": 1.0869402285928425,
+      "grad_norm": 0.23491543531417847,
+      "learning_rate": 0.0001371171171171171,
+      "loss": 0.3532,
+      "step": 3626
+    },
+    {
+      "epoch": 1.0872400224845418,
+      "grad_norm": 0.21283003687858582,
+      "learning_rate": 0.00013707207207207204,
+      "loss": 0.3328,
+      "step": 3627
+    },
+    {
+      "epoch": 1.0875398163762413,
+      "grad_norm": 0.22129040956497192,
+      "learning_rate": 0.000137027027027027,
+      "loss": 0.356,
+      "step": 3628
+    },
+    {
+      "epoch": 1.0878396102679408,
+      "grad_norm": 0.2425287812948227,
+      "learning_rate": 0.00013698198198198197,
+      "loss": 0.3915,
+      "step": 3629
+    },
+    {
+      "epoch": 1.0881394041596402,
+      "grad_norm": 0.22848957777023315,
+      "learning_rate": 0.0001369369369369369,
+      "loss": 0.382,
+      "step": 3630
+    },
+    {
+      "epoch": 1.0884391980513397,
+      "grad_norm": 0.21875622868537903,
+      "learning_rate": 0.00013689189189189187,
+      "loss": 0.3399,
+      "step": 3631
+    },
+    {
+      "epoch": 1.0887389919430392,
+      "grad_norm": 0.2495780885219574,
+      "learning_rate": 0.00013684684684684684,
+      "loss": 0.3583,
+      "step": 3632
+    },
+    {
+      "epoch": 1.0890387858347386,
+      "grad_norm": 0.22478289902210236,
+      "learning_rate": 0.0001368018018018018,
+      "loss": 0.3594,
+      "step": 3633
+    },
+    {
+      "epoch": 1.089338579726438,
+      "grad_norm": 0.24992315471172333,
+      "learning_rate": 0.00013675675675675674,
+      "loss": 0.3892,
+      "step": 3634
+    },
+    {
+      "epoch": 1.0896383736181376,
+      "grad_norm": 0.21412153542041779,
+      "learning_rate": 0.0001367117117117117,
+      "loss": 0.3504,
+      "step": 3635
+    },
+    {
+      "epoch": 1.089938167509837,
+      "grad_norm": 0.246387779712677,
+      "learning_rate": 0.00013666666666666666,
+      "loss": 0.3786,
+      "step": 3636
+    },
+    {
+      "epoch": 1.0902379614015365,
+      "grad_norm": 0.21780602633953094,
+      "learning_rate": 0.0001366216216216216,
+      "loss": 0.3603,
+      "step": 3637
+    },
+    {
+      "epoch": 1.0905377552932358,
+      "grad_norm": 0.23555536568164825,
+      "learning_rate": 0.00013657657657657656,
+      "loss": 0.3879,
+      "step": 3638
+    },
+    {
+      "epoch": 1.0908375491849354,
+      "grad_norm": 0.22709296643733978,
+      "learning_rate": 0.00013653153153153153,
+      "loss": 0.3619,
+      "step": 3639
+    },
+    {
+      "epoch": 1.091137343076635,
+      "grad_norm": 0.2261640876531601,
+      "learning_rate": 0.0001364864864864865,
+      "loss": 0.362,
+      "step": 3640
+    },
+    {
+      "epoch": 1.0914371369683342,
+      "grad_norm": 0.21862421929836273,
+      "learning_rate": 0.00013644144144144143,
+      "loss": 0.3627,
+      "step": 3641
+    },
+    {
+      "epoch": 1.0917369308600338,
+      "grad_norm": 0.2231677919626236,
+      "learning_rate": 0.0001363963963963964,
+      "loss": 0.3522,
+      "step": 3642
+    },
+    {
+      "epoch": 1.092036724751733,
+      "grad_norm": 0.23300901055335999,
+      "learning_rate": 0.00013635135135135136,
+      "loss": 0.3718,
+      "step": 3643
+    },
+    {
+      "epoch": 1.0923365186434326,
+      "grad_norm": 0.2250256985425949,
+      "learning_rate": 0.0001363063063063063,
+      "loss": 0.3666,
+      "step": 3644
+    },
+    {
+      "epoch": 1.0926363125351322,
+      "grad_norm": 0.22397859394550323,
+      "learning_rate": 0.00013626126126126126,
+      "loss": 0.3627,
+      "step": 3645
+    },
+    {
+      "epoch": 1.0929361064268315,
+      "grad_norm": 0.22734856605529785,
+      "learning_rate": 0.00013621621621621622,
+      "loss": 0.3621,
+      "step": 3646
+    },
+    {
+      "epoch": 1.093235900318531,
+      "grad_norm": 0.2459496110677719,
+      "learning_rate": 0.00013617117117117116,
+      "loss": 0.3275,
+      "step": 3647
+    },
+    {
+      "epoch": 1.0935356942102306,
+      "grad_norm": 0.23530100286006927,
+      "learning_rate": 0.00013612612612612612,
+      "loss": 0.3507,
+      "step": 3648
+    },
+    {
+      "epoch": 1.0938354881019299,
+      "grad_norm": 0.2462019920349121,
+      "learning_rate": 0.00013608108108108106,
+      "loss": 0.3558,
+      "step": 3649
+    },
+    {
+      "epoch": 1.0941352819936294,
+      "grad_norm": 0.2784384489059448,
+      "learning_rate": 0.00013603603603603602,
+      "loss": 0.3739,
+      "step": 3650
+    },
+    {
+      "epoch": 1.0944350758853287,
+      "grad_norm": 0.2294548749923706,
+      "learning_rate": 0.00013599099099099099,
+      "loss": 0.3749,
+      "step": 3651
+    },
+    {
+      "epoch": 1.0947348697770283,
+      "grad_norm": 0.26810503005981445,
+      "learning_rate": 0.00013594594594594592,
+      "loss": 0.3896,
+      "step": 3652
+    },
+    {
+      "epoch": 1.0950346636687278,
+      "grad_norm": 0.2433824986219406,
+      "learning_rate": 0.0001359009009009009,
+      "loss": 0.3547,
+      "step": 3653
+    },
+    {
+      "epoch": 1.0953344575604271,
+      "grad_norm": 0.22635456919670105,
+      "learning_rate": 0.00013585585585585585,
+      "loss": 0.3529,
+      "step": 3654
+    },
+    {
+      "epoch": 1.0956342514521267,
+      "grad_norm": 0.2657754421234131,
+      "learning_rate": 0.0001358108108108108,
+      "loss": 0.3692,
+      "step": 3655
+    },
+    {
+      "epoch": 1.0959340453438262,
+      "grad_norm": 0.23322702944278717,
+      "learning_rate": 0.00013576576576576575,
+      "loss": 0.3598,
+      "step": 3656
+    },
+    {
+      "epoch": 1.0962338392355255,
+      "grad_norm": 0.23769034445285797,
+      "learning_rate": 0.00013572072072072072,
+      "loss": 0.3685,
+      "step": 3657
+    },
+    {
+      "epoch": 1.096533633127225,
+      "grad_norm": 0.2438647300004959,
+      "learning_rate": 0.00013567567567567565,
+      "loss": 0.3744,
+      "step": 3658
+    },
+    {
+      "epoch": 1.0968334270189244,
+      "grad_norm": 0.2461502104997635,
+      "learning_rate": 0.00013563063063063062,
+      "loss": 0.3686,
+      "step": 3659
+    },
+    {
+      "epoch": 1.097133220910624,
+      "grad_norm": 0.24611039459705353,
+      "learning_rate": 0.00013558558558558555,
+      "loss": 0.3745,
+      "step": 3660
+    },
+    {
+      "epoch": 1.0974330148023235,
+      "grad_norm": 0.25352588295936584,
+      "learning_rate": 0.00013554054054054052,
+      "loss": 0.3853,
+      "step": 3661
+    },
+    {
+      "epoch": 1.0977328086940228,
+      "grad_norm": 0.22174997627735138,
+      "learning_rate": 0.00013549549549549548,
+      "loss": 0.3601,
+      "step": 3662
+    },
+    {
+      "epoch": 1.0980326025857223,
+      "grad_norm": 0.22847199440002441,
+      "learning_rate": 0.00013545045045045044,
+      "loss": 0.3519,
+      "step": 3663
+    },
+    {
+      "epoch": 1.0983323964774219,
+      "grad_norm": 0.2461545169353485,
+      "learning_rate": 0.00013540540540540538,
+      "loss": 0.3754,
+      "step": 3664
+    },
+    {
+      "epoch": 1.0986321903691212,
+      "grad_norm": 0.21870985627174377,
+      "learning_rate": 0.00013536036036036034,
+      "loss": 0.3632,
+      "step": 3665
+    },
+    {
+      "epoch": 1.0989319842608207,
+      "grad_norm": 0.23304589092731476,
+      "learning_rate": 0.0001353153153153153,
+      "loss": 0.3328,
+      "step": 3666
+    },
+    {
+      "epoch": 1.09923177815252,
+      "grad_norm": 0.2161308377981186,
+      "learning_rate": 0.00013527027027027025,
+      "loss": 0.3376,
+      "step": 3667
+    },
+    {
+      "epoch": 1.0995315720442196,
+      "grad_norm": 0.22993774712085724,
+      "learning_rate": 0.0001352252252252252,
+      "loss": 0.3711,
+      "step": 3668
+    },
+    {
+      "epoch": 1.0998313659359191,
+      "grad_norm": 0.22846129536628723,
+      "learning_rate": 0.00013518018018018017,
+      "loss": 0.3326,
+      "step": 3669
+    },
+    {
+      "epoch": 1.1001311598276184,
+      "grad_norm": 0.22285179793834686,
+      "learning_rate": 0.00013513513513513514,
+      "loss": 0.3624,
+      "step": 3670
+    },
+    {
+      "epoch": 1.100430953719318,
+      "grad_norm": 0.21985436975955963,
+      "learning_rate": 0.00013509009009009007,
+      "loss": 0.3645,
+      "step": 3671
+    },
+    {
+      "epoch": 1.1007307476110175,
+      "grad_norm": 0.24420645833015442,
+      "learning_rate": 0.00013504504504504504,
+      "loss": 0.3842,
+      "step": 3672
+    },
+    {
+      "epoch": 1.1010305415027168,
+      "grad_norm": 0.22566132247447968,
+      "learning_rate": 0.000135,
+      "loss": 0.3743,
+      "step": 3673
+    },
+    {
+      "epoch": 1.1013303353944164,
+      "grad_norm": 0.23107251524925232,
+      "learning_rate": 0.00013495495495495494,
+      "loss": 0.3735,
+      "step": 3674
+    },
+    {
+      "epoch": 1.1016301292861157,
+      "grad_norm": 0.22111618518829346,
+      "learning_rate": 0.0001349099099099099,
+      "loss": 0.3556,
+      "step": 3675
+    },
+    {
+      "epoch": 1.1019299231778152,
+      "grad_norm": 0.22829733788967133,
+      "learning_rate": 0.00013486486486486487,
+      "loss": 0.3611,
+      "step": 3676
+    },
+    {
+      "epoch": 1.1022297170695148,
+      "grad_norm": 0.22518396377563477,
+      "learning_rate": 0.0001348198198198198,
+      "loss": 0.3681,
+      "step": 3677
+    },
+    {
+      "epoch": 1.102529510961214,
+      "grad_norm": 0.21075007319450378,
+      "learning_rate": 0.00013477477477477477,
+      "loss": 0.3343,
+      "step": 3678
+    },
+    {
+      "epoch": 1.1028293048529136,
+      "grad_norm": 0.2217082977294922,
+      "learning_rate": 0.00013472972972972973,
+      "loss": 0.3608,
+      "step": 3679
+    },
+    {
+      "epoch": 1.1031290987446132,
+      "grad_norm": 0.22440844774246216,
+      "learning_rate": 0.00013468468468468467,
+      "loss": 0.3548,
+      "step": 3680
+    },
+    {
+      "epoch": 1.1034288926363125,
+      "grad_norm": 0.22126342356204987,
+      "learning_rate": 0.00013463963963963963,
+      "loss": 0.3495,
+      "step": 3681
+    },
+    {
+      "epoch": 1.103728686528012,
+      "grad_norm": 0.23776847124099731,
+      "learning_rate": 0.0001345945945945946,
+      "loss": 0.393,
+      "step": 3682
+    },
+    {
+      "epoch": 1.1040284804197114,
+      "grad_norm": 0.21629567444324493,
+      "learning_rate": 0.00013454954954954953,
+      "loss": 0.3678,
+      "step": 3683
+    },
+    {
+      "epoch": 1.104328274311411,
+      "grad_norm": 0.2131320685148239,
+      "learning_rate": 0.0001345045045045045,
+      "loss": 0.3419,
+      "step": 3684
+    },
+    {
+      "epoch": 1.1046280682031104,
+      "grad_norm": 0.22412894666194916,
+      "learning_rate": 0.00013445945945945943,
+      "loss": 0.3656,
+      "step": 3685
+    },
+    {
+      "epoch": 1.1049278620948098,
+      "grad_norm": 0.2328217476606369,
+      "learning_rate": 0.0001344144144144144,
+      "loss": 0.3534,
+      "step": 3686
+    },
+    {
+      "epoch": 1.1052276559865093,
+      "grad_norm": 0.24423038959503174,
+      "learning_rate": 0.00013436936936936936,
+      "loss": 0.3858,
+      "step": 3687
+    },
+    {
+      "epoch": 1.1055274498782088,
+      "grad_norm": 0.23228362202644348,
+      "learning_rate": 0.0001343243243243243,
+      "loss": 0.3808,
+      "step": 3688
+    },
+    {
+      "epoch": 1.1058272437699082,
+      "grad_norm": 0.22738321125507355,
+      "learning_rate": 0.00013427927927927926,
+      "loss": 0.3892,
+      "step": 3689
+    },
+    {
+      "epoch": 1.1061270376616077,
+      "grad_norm": 0.2281411737203598,
+      "learning_rate": 0.00013423423423423422,
+      "loss": 0.3624,
+      "step": 3690
+    },
+    {
+      "epoch": 1.106426831553307,
+      "grad_norm": 0.2366478443145752,
+      "learning_rate": 0.00013418918918918916,
+      "loss": 0.3901,
+      "step": 3691
+    },
+    {
+      "epoch": 1.1067266254450066,
+      "grad_norm": 0.21975024044513702,
+      "learning_rate": 0.00013414414414414413,
+      "loss": 0.3815,
+      "step": 3692
+    },
+    {
+      "epoch": 1.107026419336706,
+      "grad_norm": 0.24421407282352448,
+      "learning_rate": 0.0001340990990990991,
+      "loss": 0.3763,
+      "step": 3693
+    },
+    {
+      "epoch": 1.1073262132284054,
+      "grad_norm": 0.23885339498519897,
+      "learning_rate": 0.00013405405405405403,
+      "loss": 0.3576,
+      "step": 3694
+    },
+    {
+      "epoch": 1.107626007120105,
+      "grad_norm": 0.2415476143360138,
+      "learning_rate": 0.000134009009009009,
+      "loss": 0.3766,
+      "step": 3695
+    },
+    {
+      "epoch": 1.1079258010118043,
+      "grad_norm": 0.23010794818401337,
+      "learning_rate": 0.00013396396396396395,
+      "loss": 0.3846,
+      "step": 3696
+    },
+    {
+      "epoch": 1.1082255949035038,
+      "grad_norm": 0.24465374648571014,
+      "learning_rate": 0.00013391891891891892,
+      "loss": 0.3803,
+      "step": 3697
+    },
+    {
+      "epoch": 1.1085253887952033,
+      "grad_norm": 0.23296433687210083,
+      "learning_rate": 0.00013387387387387385,
+      "loss": 0.3702,
+      "step": 3698
+    },
+    {
+      "epoch": 1.1088251826869027,
+      "grad_norm": 0.23366378247737885,
+      "learning_rate": 0.00013382882882882882,
+      "loss": 0.3618,
+      "step": 3699
+    },
+    {
+      "epoch": 1.1091249765786022,
+      "grad_norm": 0.21989014744758606,
+      "learning_rate": 0.00013378378378378378,
+      "loss": 0.3547,
+      "step": 3700
+    },
+    {
+      "epoch": 1.1094247704703017,
+      "grad_norm": 0.23640766739845276,
+      "learning_rate": 0.00013373873873873872,
+      "loss": 0.3863,
+      "step": 3701
+    },
+    {
+      "epoch": 1.109724564362001,
+      "grad_norm": 0.22326450049877167,
+      "learning_rate": 0.00013369369369369368,
+      "loss": 0.3579,
+      "step": 3702
+    },
+    {
+      "epoch": 1.1100243582537006,
+      "grad_norm": 0.23940657079219818,
+      "learning_rate": 0.00013364864864864865,
+      "loss": 0.3679,
+      "step": 3703
+    },
+    {
+      "epoch": 1.1103241521454001,
+      "grad_norm": 0.24894234538078308,
+      "learning_rate": 0.0001336036036036036,
+      "loss": 0.3873,
+      "step": 3704
+    },
+    {
+      "epoch": 1.1106239460370995,
+      "grad_norm": 0.22873175144195557,
+      "learning_rate": 0.00013355855855855855,
+      "loss": 0.3797,
+      "step": 3705
+    },
+    {
+      "epoch": 1.110923739928799,
+      "grad_norm": 0.23857954144477844,
+      "learning_rate": 0.0001335135135135135,
+      "loss": 0.368,
+      "step": 3706
+    },
+    {
+      "epoch": 1.1112235338204983,
+      "grad_norm": 0.22301368415355682,
+      "learning_rate": 0.00013346846846846848,
+      "loss": 0.3535,
+      "step": 3707
+    },
+    {
+      "epoch": 1.1115233277121979,
+      "grad_norm": 0.23414717614650726,
+      "learning_rate": 0.0001334234234234234,
+      "loss": 0.3637,
+      "step": 3708
+    },
+    {
+      "epoch": 1.1118231216038974,
+      "grad_norm": 0.2625446021556854,
+      "learning_rate": 0.00013337837837837838,
+      "loss": 0.3959,
+      "step": 3709
+    },
+    {
+      "epoch": 1.1121229154955967,
+      "grad_norm": 0.21231703460216522,
+      "learning_rate": 0.0001333333333333333,
+      "loss": 0.3427,
+      "step": 3710
+    },
+    {
+      "epoch": 1.1124227093872963,
+      "grad_norm": 0.23999685049057007,
+      "learning_rate": 0.00013328828828828828,
+      "loss": 0.402,
+      "step": 3711
+    },
+    {
+      "epoch": 1.1127225032789956,
+      "grad_norm": 0.23020519316196442,
+      "learning_rate": 0.00013324324324324324,
+      "loss": 0.361,
+      "step": 3712
+    },
+    {
+      "epoch": 1.1130222971706951,
+      "grad_norm": 0.2160668969154358,
+      "learning_rate": 0.00013319819819819818,
+      "loss": 0.3559,
+      "step": 3713
+    },
+    {
+      "epoch": 1.1133220910623947,
+      "grad_norm": 0.23094019293785095,
+      "learning_rate": 0.00013315315315315314,
+      "loss": 0.3799,
+      "step": 3714
+    },
+    {
+      "epoch": 1.113621884954094,
+      "grad_norm": 0.23295539617538452,
+      "learning_rate": 0.0001331081081081081,
+      "loss": 0.3782,
+      "step": 3715
+    },
+    {
+      "epoch": 1.1139216788457935,
+      "grad_norm": 0.24856853485107422,
+      "learning_rate": 0.00013306306306306304,
+      "loss": 0.3878,
+      "step": 3716
+    },
+    {
+      "epoch": 1.114221472737493,
+      "grad_norm": 0.22519513964653015,
+      "learning_rate": 0.000133018018018018,
+      "loss": 0.3637,
+      "step": 3717
+    },
+    {
+      "epoch": 1.1145212666291924,
+      "grad_norm": 0.235112726688385,
+      "learning_rate": 0.00013297297297297297,
+      "loss": 0.3623,
+      "step": 3718
+    },
+    {
+      "epoch": 1.114821060520892,
+      "grad_norm": 0.2207203060388565,
+      "learning_rate": 0.0001329279279279279,
+      "loss": 0.3711,
+      "step": 3719
+    },
+    {
+      "epoch": 1.1151208544125915,
+      "grad_norm": 0.2256298065185547,
+      "learning_rate": 0.00013288288288288287,
+      "loss": 0.3407,
+      "step": 3720
+    },
+    {
+      "epoch": 1.1154206483042908,
+      "grad_norm": 0.21627315878868103,
+      "learning_rate": 0.0001328378378378378,
+      "loss": 0.3514,
+      "step": 3721
+    },
+    {
+      "epoch": 1.1157204421959903,
+      "grad_norm": 0.22296972572803497,
+      "learning_rate": 0.00013279279279279277,
+      "loss": 0.3598,
+      "step": 3722
+    },
+    {
+      "epoch": 1.1160202360876896,
+      "grad_norm": 0.23051004111766815,
+      "learning_rate": 0.00013274774774774773,
+      "loss": 0.3703,
+      "step": 3723
+    },
+    {
+      "epoch": 1.1163200299793892,
+      "grad_norm": 0.22421251237392426,
+      "learning_rate": 0.00013270270270270267,
+      "loss": 0.3671,
+      "step": 3724
+    },
+    {
+      "epoch": 1.1166198238710887,
+      "grad_norm": 0.22339527308940887,
+      "learning_rate": 0.00013265765765765764,
+      "loss": 0.3396,
+      "step": 3725
+    },
+    {
+      "epoch": 1.116919617762788,
+      "grad_norm": 0.22906826436519623,
+      "learning_rate": 0.0001326126126126126,
+      "loss": 0.3492,
+      "step": 3726
+    },
+    {
+      "epoch": 1.1172194116544876,
+      "grad_norm": 0.22747504711151123,
+      "learning_rate": 0.00013256756756756756,
+      "loss": 0.3592,
+      "step": 3727
+    },
+    {
+      "epoch": 1.1175192055461869,
+      "grad_norm": 0.2335633784532547,
+      "learning_rate": 0.0001325225225225225,
+      "loss": 0.3636,
+      "step": 3728
+    },
+    {
+      "epoch": 1.1178189994378864,
+      "grad_norm": 0.22873105108737946,
+      "learning_rate": 0.00013247747747747746,
+      "loss": 0.3716,
+      "step": 3729
+    },
+    {
+      "epoch": 1.118118793329586,
+      "grad_norm": 0.23329488933086395,
+      "learning_rate": 0.00013243243243243243,
+      "loss": 0.3573,
+      "step": 3730
+    },
+    {
+      "epoch": 1.1184185872212853,
+      "grad_norm": 0.21844343841075897,
+      "learning_rate": 0.0001323873873873874,
+      "loss": 0.3355,
+      "step": 3731
+    },
+    {
+      "epoch": 1.1187183811129848,
+      "grad_norm": 0.23338133096694946,
+      "learning_rate": 0.00013234234234234233,
+      "loss": 0.3792,
+      "step": 3732
+    },
+    {
+      "epoch": 1.1190181750046844,
+      "grad_norm": 0.22400586307048798,
+      "learning_rate": 0.0001322972972972973,
+      "loss": 0.3753,
+      "step": 3733
+    },
+    {
+      "epoch": 1.1193179688963837,
+      "grad_norm": 0.2163958102464676,
+      "learning_rate": 0.00013225225225225226,
+      "loss": 0.353,
+      "step": 3734
+    },
+    {
+      "epoch": 1.1196177627880832,
+      "grad_norm": 0.24042215943336487,
+      "learning_rate": 0.0001322072072072072,
+      "loss": 0.3607,
+      "step": 3735
+    },
+    {
+      "epoch": 1.1199175566797828,
+      "grad_norm": 0.22561825811862946,
+      "learning_rate": 0.00013216216216216216,
+      "loss": 0.3701,
+      "step": 3736
+    },
+    {
+      "epoch": 1.120217350571482,
+      "grad_norm": 0.22830486297607422,
+      "learning_rate": 0.00013211711711711712,
+      "loss": 0.3649,
+      "step": 3737
+    },
+    {
+      "epoch": 1.1205171444631816,
+      "grad_norm": 0.21653513610363007,
+      "learning_rate": 0.00013207207207207206,
+      "loss": 0.3464,
+      "step": 3738
+    },
+    {
+      "epoch": 1.120816938354881,
+      "grad_norm": 0.23094573616981506,
+      "learning_rate": 0.00013202702702702702,
+      "loss": 0.3829,
+      "step": 3739
+    },
+    {
+      "epoch": 1.1211167322465805,
+      "grad_norm": 0.22236552834510803,
+      "learning_rate": 0.00013198198198198198,
+      "loss": 0.341,
+      "step": 3740
+    },
+    {
+      "epoch": 1.12141652613828,
+      "grad_norm": 0.22838713228702545,
+      "learning_rate": 0.00013193693693693692,
+      "loss": 0.3724,
+      "step": 3741
+    },
+    {
+      "epoch": 1.1217163200299793,
+      "grad_norm": 0.2387278527021408,
+      "learning_rate": 0.00013189189189189189,
+      "loss": 0.3535,
+      "step": 3742
+    },
+    {
+      "epoch": 1.1220161139216789,
+      "grad_norm": 0.2243206948041916,
+      "learning_rate": 0.00013184684684684685,
+      "loss": 0.3484,
+      "step": 3743
+    },
+    {
+      "epoch": 1.1223159078133782,
+      "grad_norm": 0.21640804409980774,
+      "learning_rate": 0.00013180180180180179,
+      "loss": 0.3558,
+      "step": 3744
+    },
+    {
+      "epoch": 1.1226157017050777,
+      "grad_norm": 0.23123346269130707,
+      "learning_rate": 0.00013175675675675675,
+      "loss": 0.3549,
+      "step": 3745
+    },
+    {
+      "epoch": 1.1229154955967773,
+      "grad_norm": 0.22995956242084503,
+      "learning_rate": 0.0001317117117117117,
+      "loss": 0.3502,
+      "step": 3746
+    },
+    {
+      "epoch": 1.1232152894884766,
+      "grad_norm": 0.2435770183801651,
+      "learning_rate": 0.00013166666666666665,
+      "loss": 0.4184,
+      "step": 3747
+    },
+    {
+      "epoch": 1.1235150833801761,
+      "grad_norm": 0.21576935052871704,
+      "learning_rate": 0.00013162162162162161,
+      "loss": 0.3598,
+      "step": 3748
+    },
+    {
+      "epoch": 1.1238148772718757,
+      "grad_norm": 0.22539666295051575,
+      "learning_rate": 0.00013157657657657655,
+      "loss": 0.3527,
+      "step": 3749
+    },
+    {
+      "epoch": 1.124114671163575,
+      "grad_norm": 0.24922697246074677,
+      "learning_rate": 0.00013153153153153152,
+      "loss": 0.3499,
+      "step": 3750
+    },
+    {
+      "epoch": 1.1244144650552745,
+      "grad_norm": 0.2371659129858017,
+      "learning_rate": 0.00013148648648648648,
+      "loss": 0.3837,
+      "step": 3751
+    },
+    {
+      "epoch": 1.1247142589469739,
+      "grad_norm": 0.22361376881599426,
+      "learning_rate": 0.00013144144144144142,
+      "loss": 0.3675,
+      "step": 3752
+    },
+    {
+      "epoch": 1.1250140528386734,
+      "grad_norm": 0.24542655050754547,
+      "learning_rate": 0.00013139639639639638,
+      "loss": 0.3922,
+      "step": 3753
+    },
+    {
+      "epoch": 1.125313846730373,
+      "grad_norm": 0.22801847755908966,
+      "learning_rate": 0.00013135135135135134,
+      "loss": 0.357,
+      "step": 3754
+    },
+    {
+      "epoch": 1.1256136406220723,
+      "grad_norm": 0.23783008754253387,
+      "learning_rate": 0.00013130630630630628,
+      "loss": 0.3834,
+      "step": 3755
+    },
+    {
+      "epoch": 1.1259134345137718,
+      "grad_norm": 0.2301386147737503,
+      "learning_rate": 0.00013126126126126124,
+      "loss": 0.3671,
+      "step": 3756
+    },
+    {
+      "epoch": 1.1262132284054713,
+      "grad_norm": 0.2202630639076233,
+      "learning_rate": 0.0001312162162162162,
+      "loss": 0.3564,
+      "step": 3757
+    },
+    {
+      "epoch": 1.1265130222971707,
+      "grad_norm": 0.22803302109241486,
+      "learning_rate": 0.00013117117117117114,
+      "loss": 0.3716,
+      "step": 3758
+    },
+    {
+      "epoch": 1.1268128161888702,
+      "grad_norm": 0.23013994097709656,
+      "learning_rate": 0.0001311261261261261,
+      "loss": 0.3668,
+      "step": 3759
+    },
+    {
+      "epoch": 1.1271126100805695,
+      "grad_norm": 0.2381478101015091,
+      "learning_rate": 0.00013108108108108107,
+      "loss": 0.364,
+      "step": 3760
+    },
+    {
+      "epoch": 1.127412403972269,
+      "grad_norm": 0.22907504439353943,
+      "learning_rate": 0.00013103603603603604,
+      "loss": 0.3582,
+      "step": 3761
+    },
+    {
+      "epoch": 1.1277121978639686,
+      "grad_norm": 0.23410449922084808,
+      "learning_rate": 0.00013099099099099097,
+      "loss": 0.3604,
+      "step": 3762
+    },
+    {
+      "epoch": 1.128011991755668,
+      "grad_norm": 0.2244759202003479,
+      "learning_rate": 0.00013094594594594594,
+      "loss": 0.3534,
+      "step": 3763
+    },
+    {
+      "epoch": 1.1283117856473674,
+      "grad_norm": 0.22788846492767334,
+      "learning_rate": 0.0001309009009009009,
+      "loss": 0.375,
+      "step": 3764
+    },
+    {
+      "epoch": 1.1286115795390668,
+      "grad_norm": 0.23673410713672638,
+      "learning_rate": 0.00013085585585585584,
+      "loss": 0.3754,
+      "step": 3765
+    },
+    {
+      "epoch": 1.1289113734307663,
+      "grad_norm": 0.21583424508571625,
+      "learning_rate": 0.0001308108108108108,
+      "loss": 0.3523,
+      "step": 3766
+    },
+    {
+      "epoch": 1.1292111673224658,
+      "grad_norm": 0.22422835230827332,
+      "learning_rate": 0.00013076576576576577,
+      "loss": 0.3659,
+      "step": 3767
+    },
+    {
+      "epoch": 1.1295109612141652,
+      "grad_norm": 0.23020607233047485,
+      "learning_rate": 0.00013072072072072073,
+      "loss": 0.3736,
+      "step": 3768
+    },
+    {
+      "epoch": 1.1298107551058647,
+      "grad_norm": 0.22912679612636566,
+      "learning_rate": 0.00013067567567567567,
+      "loss": 0.3822,
+      "step": 3769
+    },
+    {
+      "epoch": 1.1301105489975642,
+      "grad_norm": 0.221907839179039,
+      "learning_rate": 0.00013063063063063063,
+      "loss": 0.3718,
+      "step": 3770
+    },
+    {
+      "epoch": 1.1304103428892636,
+      "grad_norm": 0.23240089416503906,
+      "learning_rate": 0.00013058558558558557,
+      "loss": 0.3703,
+      "step": 3771
+    },
+    {
+      "epoch": 1.130710136780963,
+      "grad_norm": 0.23240825533866882,
+      "learning_rate": 0.00013054054054054053,
+      "loss": 0.3412,
+      "step": 3772
+    },
+    {
+      "epoch": 1.1310099306726626,
+      "grad_norm": 0.22136814892292023,
+      "learning_rate": 0.0001304954954954955,
+      "loss": 0.3667,
+      "step": 3773
+    },
+    {
+      "epoch": 1.131309724564362,
+      "grad_norm": 0.23060092329978943,
+      "learning_rate": 0.00013045045045045043,
+      "loss": 0.3689,
+      "step": 3774
+    },
+    {
+      "epoch": 1.1316095184560615,
+      "grad_norm": 0.23062026500701904,
+      "learning_rate": 0.0001304054054054054,
+      "loss": 0.3546,
+      "step": 3775
+    },
+    {
+      "epoch": 1.1319093123477608,
+      "grad_norm": 0.22831834852695465,
+      "learning_rate": 0.00013036036036036036,
+      "loss": 0.3699,
+      "step": 3776
+    },
+    {
+      "epoch": 1.1322091062394604,
+      "grad_norm": 0.22447456419467926,
+      "learning_rate": 0.0001303153153153153,
+      "loss": 0.368,
+      "step": 3777
+    },
+    {
+      "epoch": 1.13250890013116,
+      "grad_norm": 0.21648916602134705,
+      "learning_rate": 0.00013027027027027026,
+      "loss": 0.3498,
+      "step": 3778
+    },
+    {
+      "epoch": 1.1328086940228592,
+      "grad_norm": 0.2411709874868393,
+      "learning_rate": 0.00013022522522522522,
+      "loss": 0.3729,
+      "step": 3779
+    },
+    {
+      "epoch": 1.1331084879145588,
+      "grad_norm": 0.2325679361820221,
+      "learning_rate": 0.00013018018018018016,
+      "loss": 0.3459,
+      "step": 3780
+    },
+    {
+      "epoch": 1.133408281806258,
+      "grad_norm": 0.22246721386909485,
+      "learning_rate": 0.00013013513513513512,
+      "loss": 0.3518,
+      "step": 3781
+    },
+    {
+      "epoch": 1.1337080756979576,
+      "grad_norm": 0.23394489288330078,
+      "learning_rate": 0.00013009009009009006,
+      "loss": 0.3825,
+      "step": 3782
+    },
+    {
+      "epoch": 1.1340078695896572,
+      "grad_norm": 0.21840216219425201,
+      "learning_rate": 0.00013004504504504502,
+      "loss": 0.3584,
+      "step": 3783
+    },
+    {
+      "epoch": 1.1343076634813565,
+      "grad_norm": 0.2106565684080124,
+      "learning_rate": 0.00013,
+      "loss": 0.3458,
+      "step": 3784
+    },
+    {
+      "epoch": 1.134607457373056,
+      "grad_norm": 0.24276579916477203,
+      "learning_rate": 0.00012995495495495493,
+      "loss": 0.3997,
+      "step": 3785
+    },
+    {
+      "epoch": 1.1349072512647556,
+      "grad_norm": 0.226302832365036,
+      "learning_rate": 0.0001299099099099099,
+      "loss": 0.3594,
+      "step": 3786
+    },
+    {
+      "epoch": 1.1352070451564549,
+      "grad_norm": 0.23036521673202515,
+      "learning_rate": 0.00012986486486486485,
+      "loss": 0.3705,
+      "step": 3787
+    },
+    {
+      "epoch": 1.1355068390481544,
+      "grad_norm": 0.21445749700069427,
+      "learning_rate": 0.00012981981981981982,
+      "loss": 0.3596,
+      "step": 3788
+    },
+    {
+      "epoch": 1.135806632939854,
+      "grad_norm": 0.23045356571674347,
+      "learning_rate": 0.00012977477477477475,
+      "loss": 0.3622,
+      "step": 3789
+    },
+    {
+      "epoch": 1.1361064268315533,
+      "grad_norm": 0.21341310441493988,
+      "learning_rate": 0.00012972972972972972,
+      "loss": 0.3351,
+      "step": 3790
+    },
+    {
+      "epoch": 1.1364062207232528,
+      "grad_norm": 0.2475743591785431,
+      "learning_rate": 0.00012968468468468468,
+      "loss": 0.384,
+      "step": 3791
+    },
+    {
+      "epoch": 1.1367060146149521,
+      "grad_norm": 0.23405541479587555,
+      "learning_rate": 0.00012963963963963962,
+      "loss": 0.381,
+      "step": 3792
+    },
+    {
+      "epoch": 1.1370058085066517,
+      "grad_norm": 0.22105379402637482,
+      "learning_rate": 0.00012959459459459458,
+      "loss": 0.3682,
+      "step": 3793
+    },
+    {
+      "epoch": 1.1373056023983512,
+      "grad_norm": 0.2180909365415573,
+      "learning_rate": 0.00012954954954954955,
+      "loss": 0.3549,
+      "step": 3794
+    },
+    {
+      "epoch": 1.1376053962900505,
+      "grad_norm": 0.23662352561950684,
+      "learning_rate": 0.0001295045045045045,
+      "loss": 0.404,
+      "step": 3795
+    },
+    {
+      "epoch": 1.13790519018175,
+      "grad_norm": 0.23034508526325226,
+      "learning_rate": 0.00012945945945945945,
+      "loss": 0.3422,
+      "step": 3796
+    },
+    {
+      "epoch": 1.1382049840734494,
+      "grad_norm": 0.2209431529045105,
+      "learning_rate": 0.0001294144144144144,
+      "loss": 0.3522,
+      "step": 3797
+    },
+    {
+      "epoch": 1.138504777965149,
+      "grad_norm": 0.2166588455438614,
+      "learning_rate": 0.00012936936936936937,
+      "loss": 0.3507,
+      "step": 3798
+    },
+    {
+      "epoch": 1.1388045718568485,
+      "grad_norm": 0.22025123238563538,
+      "learning_rate": 0.0001293243243243243,
+      "loss": 0.3624,
+      "step": 3799
+    },
+    {
+      "epoch": 1.1391043657485478,
+      "grad_norm": 0.22763989865779877,
+      "learning_rate": 0.00012927927927927928,
+      "loss": 0.3695,
+      "step": 3800
+    },
+    {
+      "epoch": 1.1394041596402473,
+      "grad_norm": 0.2226601541042328,
+      "learning_rate": 0.00012923423423423424,
+      "loss": 0.3573,
+      "step": 3801
+    },
+    {
+      "epoch": 1.1397039535319469,
+      "grad_norm": 0.2142391949892044,
+      "learning_rate": 0.00012918918918918918,
+      "loss": 0.3257,
+      "step": 3802
+    },
+    {
+      "epoch": 1.1400037474236462,
+      "grad_norm": 0.23330844938755035,
+      "learning_rate": 0.00012914414414414414,
+      "loss": 0.3619,
+      "step": 3803
+    },
+    {
+      "epoch": 1.1403035413153457,
+      "grad_norm": 0.22902002930641174,
+      "learning_rate": 0.0001290990990990991,
+      "loss": 0.3563,
+      "step": 3804
+    },
+    {
+      "epoch": 1.1406033352070453,
+      "grad_norm": 0.2570081353187561,
+      "learning_rate": 0.00012905405405405404,
+      "loss": 0.3784,
+      "step": 3805
+    },
+    {
+      "epoch": 1.1409031290987446,
+      "grad_norm": 0.2169354110956192,
+      "learning_rate": 0.000129009009009009,
+      "loss": 0.3751,
+      "step": 3806
+    },
+    {
+      "epoch": 1.1412029229904441,
+      "grad_norm": 0.2247963547706604,
+      "learning_rate": 0.00012896396396396394,
+      "loss": 0.3797,
+      "step": 3807
+    },
+    {
+      "epoch": 1.1415027168821434,
+      "grad_norm": 0.23023879528045654,
+      "learning_rate": 0.0001289189189189189,
+      "loss": 0.3571,
+      "step": 3808
+    },
+    {
+      "epoch": 1.141802510773843,
+      "grad_norm": 0.21960099041461945,
+      "learning_rate": 0.00012887387387387387,
+      "loss": 0.3531,
+      "step": 3809
+    },
+    {
+      "epoch": 1.1421023046655425,
+      "grad_norm": 0.22244469821453094,
+      "learning_rate": 0.0001288288288288288,
+      "loss": 0.3544,
+      "step": 3810
+    },
+    {
+      "epoch": 1.1424020985572418,
+      "grad_norm": 0.23384922742843628,
+      "learning_rate": 0.00012878378378378377,
+      "loss": 0.382,
+      "step": 3811
+    },
+    {
+      "epoch": 1.1427018924489414,
+      "grad_norm": 0.2118406891822815,
+      "learning_rate": 0.00012873873873873873,
+      "loss": 0.3372,
+      "step": 3812
+    },
+    {
+      "epoch": 1.1430016863406407,
+      "grad_norm": 0.22800342738628387,
+      "learning_rate": 0.00012869369369369367,
+      "loss": 0.3664,
+      "step": 3813
+    },
+    {
+      "epoch": 1.1433014802323402,
+      "grad_norm": 0.24370819330215454,
+      "learning_rate": 0.00012864864864864863,
+      "loss": 0.364,
+      "step": 3814
+    },
+    {
+      "epoch": 1.1436012741240398,
+      "grad_norm": 0.22153517603874207,
+      "learning_rate": 0.0001286036036036036,
+      "loss": 0.3664,
+      "step": 3815
+    },
+    {
+      "epoch": 1.143901068015739,
+      "grad_norm": 0.22754263877868652,
+      "learning_rate": 0.00012855855855855853,
+      "loss": 0.367,
+      "step": 3816
+    },
+    {
+      "epoch": 1.1442008619074386,
+      "grad_norm": 0.22332321107387543,
+      "learning_rate": 0.0001285135135135135,
+      "loss": 0.3625,
+      "step": 3817
+    },
+    {
+      "epoch": 1.1445006557991382,
+      "grad_norm": 0.22275400161743164,
+      "learning_rate": 0.00012846846846846846,
+      "loss": 0.3575,
+      "step": 3818
+    },
+    {
+      "epoch": 1.1448004496908375,
+      "grad_norm": 0.21963873505592346,
+      "learning_rate": 0.0001284234234234234,
+      "loss": 0.341,
+      "step": 3819
+    },
+    {
+      "epoch": 1.145100243582537,
+      "grad_norm": 0.2317275106906891,
+      "learning_rate": 0.00012837837837837836,
+      "loss": 0.3631,
+      "step": 3820
+    },
+    {
+      "epoch": 1.1454000374742366,
+      "grad_norm": 0.21779996156692505,
+      "learning_rate": 0.00012833333333333333,
+      "loss": 0.3643,
+      "step": 3821
+    },
+    {
+      "epoch": 1.145699831365936,
+      "grad_norm": 0.22643230855464935,
+      "learning_rate": 0.00012828828828828826,
+      "loss": 0.362,
+      "step": 3822
+    },
+    {
+      "epoch": 1.1459996252576354,
+      "grad_norm": 0.23639261722564697,
+      "learning_rate": 0.00012824324324324323,
+      "loss": 0.36,
+      "step": 3823
+    },
+    {
+      "epoch": 1.1462994191493348,
+      "grad_norm": 0.241719588637352,
+      "learning_rate": 0.0001281981981981982,
+      "loss": 0.4015,
+      "step": 3824
+    },
+    {
+      "epoch": 1.1465992130410343,
+      "grad_norm": 0.22257046401500702,
+      "learning_rate": 0.00012815315315315315,
+      "loss": 0.3648,
+      "step": 3825
+    },
+    {
+      "epoch": 1.1468990069327338,
+      "grad_norm": 0.2236374020576477,
+      "learning_rate": 0.0001281081081081081,
+      "loss": 0.3684,
+      "step": 3826
+    },
+    {
+      "epoch": 1.1471988008244332,
+      "grad_norm": 0.2196348011493683,
+      "learning_rate": 0.00012806306306306306,
+      "loss": 0.3627,
+      "step": 3827
+    },
+    {
+      "epoch": 1.1474985947161327,
+      "grad_norm": 0.24044060707092285,
+      "learning_rate": 0.00012801801801801802,
+      "loss": 0.3706,
+      "step": 3828
+    },
+    {
+      "epoch": 1.147798388607832,
+      "grad_norm": 0.2156895101070404,
+      "learning_rate": 0.00012797297297297296,
+      "loss": 0.3697,
+      "step": 3829
+    },
+    {
+      "epoch": 1.1480981824995315,
+      "grad_norm": 0.2214498221874237,
+      "learning_rate": 0.00012792792792792792,
+      "loss": 0.3719,
+      "step": 3830
+    },
+    {
+      "epoch": 1.148397976391231,
+      "grad_norm": 0.2220916897058487,
+      "learning_rate": 0.00012788288288288288,
+      "loss": 0.367,
+      "step": 3831
+    },
+    {
+      "epoch": 1.1486977702829304,
+      "grad_norm": 0.23982498049736023,
+      "learning_rate": 0.00012783783783783782,
+      "loss": 0.3856,
+      "step": 3832
+    },
+    {
+      "epoch": 1.14899756417463,
+      "grad_norm": 0.22089318931102753,
+      "learning_rate": 0.00012779279279279278,
+      "loss": 0.3337,
+      "step": 3833
+    },
+    {
+      "epoch": 1.1492973580663295,
+      "grad_norm": 0.20539483428001404,
+      "learning_rate": 0.00012774774774774775,
+      "loss": 0.3382,
+      "step": 3834
+    },
+    {
+      "epoch": 1.1495971519580288,
+      "grad_norm": 0.2173803299665451,
+      "learning_rate": 0.00012770270270270269,
+      "loss": 0.3595,
+      "step": 3835
+    },
+    {
+      "epoch": 1.1498969458497283,
+      "grad_norm": 0.2402333915233612,
+      "learning_rate": 0.00012765765765765765,
+      "loss": 0.3714,
+      "step": 3836
+    },
+    {
+      "epoch": 1.1501967397414279,
+      "grad_norm": 0.2177773267030716,
+      "learning_rate": 0.0001276126126126126,
+      "loss": 0.348,
+      "step": 3837
+    },
+    {
+      "epoch": 1.1504965336331272,
+      "grad_norm": 0.22336289286613464,
+      "learning_rate": 0.00012756756756756755,
+      "loss": 0.3722,
+      "step": 3838
+    },
+    {
+      "epoch": 1.1507963275248267,
+      "grad_norm": 0.20858174562454224,
+      "learning_rate": 0.00012752252252252251,
+      "loss": 0.3337,
+      "step": 3839
+    },
+    {
+      "epoch": 1.151096121416526,
+      "grad_norm": 0.22657938301563263,
+      "learning_rate": 0.00012747747747747748,
+      "loss": 0.3641,
+      "step": 3840
+    },
+    {
+      "epoch": 1.1513959153082256,
+      "grad_norm": 0.23537993431091309,
+      "learning_rate": 0.00012743243243243241,
+      "loss": 0.3679,
+      "step": 3841
+    },
+    {
+      "epoch": 1.1516957091999251,
+      "grad_norm": 0.23289118707180023,
+      "learning_rate": 0.00012738738738738738,
+      "loss": 0.3783,
+      "step": 3842
+    },
+    {
+      "epoch": 1.1519955030916245,
+      "grad_norm": 0.2241375595331192,
+      "learning_rate": 0.00012734234234234231,
+      "loss": 0.351,
+      "step": 3843
+    },
+    {
+      "epoch": 1.152295296983324,
+      "grad_norm": 0.22272761166095734,
+      "learning_rate": 0.00012729729729729728,
+      "loss": 0.3466,
+      "step": 3844
+    },
+    {
+      "epoch": 1.1525950908750233,
+      "grad_norm": 0.22620238363742828,
+      "learning_rate": 0.00012725225225225224,
+      "loss": 0.3567,
+      "step": 3845
+    },
+    {
+      "epoch": 1.1528948847667229,
+      "grad_norm": 0.23571644723415375,
+      "learning_rate": 0.00012720720720720718,
+      "loss": 0.3679,
+      "step": 3846
+    },
+    {
+      "epoch": 1.1531946786584224,
+      "grad_norm": 0.2184414118528366,
+      "learning_rate": 0.00012716216216216214,
+      "loss": 0.3329,
+      "step": 3847
+    },
+    {
+      "epoch": 1.1534944725501217,
+      "grad_norm": 0.23073722422122955,
+      "learning_rate": 0.0001271171171171171,
+      "loss": 0.3613,
+      "step": 3848
+    },
+    {
+      "epoch": 1.1537942664418213,
+      "grad_norm": 0.21744020283222198,
+      "learning_rate": 0.00012707207207207204,
+      "loss": 0.378,
+      "step": 3849
+    },
+    {
+      "epoch": 1.1540940603335208,
+      "grad_norm": 0.23165687918663025,
+      "learning_rate": 0.000127027027027027,
+      "loss": 0.3623,
+      "step": 3850
+    },
+    {
+      "epoch": 1.1543938542252201,
+      "grad_norm": 0.2230186015367508,
+      "learning_rate": 0.00012698198198198197,
+      "loss": 0.3849,
+      "step": 3851
+    },
+    {
+      "epoch": 1.1546936481169197,
+      "grad_norm": 0.2341298907995224,
+      "learning_rate": 0.00012693693693693694,
+      "loss": 0.3744,
+      "step": 3852
+    },
+    {
+      "epoch": 1.1549934420086192,
+      "grad_norm": 0.2185104489326477,
+      "learning_rate": 0.00012689189189189187,
+      "loss": 0.3675,
+      "step": 3853
+    },
+    {
+      "epoch": 1.1552932359003185,
+      "grad_norm": 0.21571005880832672,
+      "learning_rate": 0.00012684684684684684,
+      "loss": 0.3428,
+      "step": 3854
+    },
+    {
+      "epoch": 1.155593029792018,
+      "grad_norm": 0.221488818526268,
+      "learning_rate": 0.0001268018018018018,
+      "loss": 0.3449,
+      "step": 3855
+    },
+    {
+      "epoch": 1.1558928236837174,
+      "grad_norm": 0.2154919058084488,
+      "learning_rate": 0.00012675675675675674,
+      "loss": 0.3392,
+      "step": 3856
+    },
+    {
+      "epoch": 1.156192617575417,
+      "grad_norm": 0.22724199295043945,
+      "learning_rate": 0.0001267117117117117,
+      "loss": 0.3618,
+      "step": 3857
+    },
+    {
+      "epoch": 1.1564924114671165,
+      "grad_norm": 0.22333818674087524,
+      "learning_rate": 0.00012666666666666666,
+      "loss": 0.3547,
+      "step": 3858
+    },
+    {
+      "epoch": 1.1567922053588158,
+      "grad_norm": 0.22069819271564484,
+      "learning_rate": 0.00012662162162162163,
+      "loss": 0.3636,
+      "step": 3859
+    },
+    {
+      "epoch": 1.1570919992505153,
+      "grad_norm": 0.21591633558273315,
+      "learning_rate": 0.00012657657657657657,
+      "loss": 0.3574,
+      "step": 3860
+    },
+    {
+      "epoch": 1.1573917931422146,
+      "grad_norm": 0.24098694324493408,
+      "learning_rate": 0.00012653153153153153,
+      "loss": 0.3836,
+      "step": 3861
+    },
+    {
+      "epoch": 1.1576915870339142,
+      "grad_norm": 0.2127639651298523,
+      "learning_rate": 0.0001264864864864865,
+      "loss": 0.3508,
+      "step": 3862
+    },
+    {
+      "epoch": 1.1579913809256137,
+      "grad_norm": 0.2245662361383438,
+      "learning_rate": 0.00012644144144144143,
+      "loss": 0.3617,
+      "step": 3863
+    },
+    {
+      "epoch": 1.158291174817313,
+      "grad_norm": 0.2253742665052414,
+      "learning_rate": 0.0001263963963963964,
+      "loss": 0.3689,
+      "step": 3864
+    },
+    {
+      "epoch": 1.1585909687090126,
+      "grad_norm": 0.22618187963962555,
+      "learning_rate": 0.00012635135135135133,
+      "loss": 0.3588,
+      "step": 3865
+    },
+    {
+      "epoch": 1.1588907626007119,
+      "grad_norm": 0.21666477620601654,
+      "learning_rate": 0.0001263063063063063,
+      "loss": 0.3447,
+      "step": 3866
+    },
+    {
+      "epoch": 1.1591905564924114,
+      "grad_norm": 0.20630092918872833,
+      "learning_rate": 0.00012626126126126126,
+      "loss": 0.3466,
+      "step": 3867
+    },
+    {
+      "epoch": 1.159490350384111,
+      "grad_norm": 0.22365500032901764,
+      "learning_rate": 0.0001262162162162162,
+      "loss": 0.363,
+      "step": 3868
+    },
+    {
+      "epoch": 1.1597901442758103,
+      "grad_norm": 0.22614437341690063,
+      "learning_rate": 0.00012617117117117116,
+      "loss": 0.3689,
+      "step": 3869
+    },
+    {
+      "epoch": 1.1600899381675098,
+      "grad_norm": 0.21964935958385468,
+      "learning_rate": 0.00012612612612612612,
+      "loss": 0.3512,
+      "step": 3870
+    },
+    {
+      "epoch": 1.1603897320592094,
+      "grad_norm": 0.22623321413993835,
+      "learning_rate": 0.00012608108108108106,
+      "loss": 0.3808,
+      "step": 3871
+    },
+    {
+      "epoch": 1.1606895259509087,
+      "grad_norm": 0.24244821071624756,
+      "learning_rate": 0.00012603603603603602,
+      "loss": 0.391,
+      "step": 3872
+    },
+    {
+      "epoch": 1.1609893198426082,
+      "grad_norm": 0.21943898499011993,
+      "learning_rate": 0.000125990990990991,
+      "loss": 0.3544,
+      "step": 3873
+    },
+    {
+      "epoch": 1.1612891137343078,
+      "grad_norm": 0.2190704196691513,
+      "learning_rate": 0.00012594594594594592,
+      "loss": 0.3668,
+      "step": 3874
+    },
+    {
+      "epoch": 1.161588907626007,
+      "grad_norm": 0.23364375531673431,
+      "learning_rate": 0.0001259009009009009,
+      "loss": 0.3464,
+      "step": 3875
+    },
+    {
+      "epoch": 1.1618887015177066,
+      "grad_norm": 0.22535942494869232,
+      "learning_rate": 0.00012585585585585585,
+      "loss": 0.3547,
+      "step": 3876
+    },
+    {
+      "epoch": 1.162188495409406,
+      "grad_norm": 0.2408933788537979,
+      "learning_rate": 0.0001258108108108108,
+      "loss": 0.381,
+      "step": 3877
+    },
+    {
+      "epoch": 1.1624882893011055,
+      "grad_norm": 0.22482524812221527,
+      "learning_rate": 0.00012576576576576575,
+      "loss": 0.3638,
+      "step": 3878
+    },
+    {
+      "epoch": 1.162788083192805,
+      "grad_norm": 0.22486832737922668,
+      "learning_rate": 0.0001257207207207207,
+      "loss": 0.3801,
+      "step": 3879
+    },
+    {
+      "epoch": 1.1630878770845043,
+      "grad_norm": 0.22509555518627167,
+      "learning_rate": 0.00012567567567567565,
+      "loss": 0.3776,
+      "step": 3880
+    },
+    {
+      "epoch": 1.1633876709762039,
+      "grad_norm": 0.21738441288471222,
+      "learning_rate": 0.00012563063063063062,
+      "loss": 0.3394,
+      "step": 3881
+    },
+    {
+      "epoch": 1.1636874648679032,
+      "grad_norm": 0.24172024428844452,
+      "learning_rate": 0.00012558558558558558,
+      "loss": 0.3598,
+      "step": 3882
+    },
+    {
+      "epoch": 1.1639872587596027,
+      "grad_norm": 0.21539001166820526,
+      "learning_rate": 0.00012554054054054052,
+      "loss": 0.3463,
+      "step": 3883
+    },
+    {
+      "epoch": 1.1642870526513023,
+      "grad_norm": 0.22196736931800842,
+      "learning_rate": 0.00012549549549549548,
+      "loss": 0.3507,
+      "step": 3884
+    },
+    {
+      "epoch": 1.1645868465430016,
+      "grad_norm": 0.23154208064079285,
+      "learning_rate": 0.00012545045045045045,
+      "loss": 0.3763,
+      "step": 3885
+    },
+    {
+      "epoch": 1.1648866404347011,
+      "grad_norm": 0.23062720894813538,
+      "learning_rate": 0.0001254054054054054,
+      "loss": 0.3465,
+      "step": 3886
+    },
+    {
+      "epoch": 1.1651864343264007,
+      "grad_norm": 0.21816441416740417,
+      "learning_rate": 0.00012536036036036035,
+      "loss": 0.3488,
+      "step": 3887
+    },
+    {
+      "epoch": 1.1654862282181,
+      "grad_norm": 0.2304592877626419,
+      "learning_rate": 0.0001253153153153153,
+      "loss": 0.386,
+      "step": 3888
+    },
+    {
+      "epoch": 1.1657860221097995,
+      "grad_norm": 0.2256111055612564,
+      "learning_rate": 0.00012527027027027027,
+      "loss": 0.3537,
+      "step": 3889
+    },
+    {
+      "epoch": 1.166085816001499,
+      "grad_norm": 0.22202616930007935,
+      "learning_rate": 0.0001252252252252252,
+      "loss": 0.364,
+      "step": 3890
+    },
+    {
+      "epoch": 1.1663856098931984,
+      "grad_norm": 0.28027182817459106,
+      "learning_rate": 0.00012518018018018017,
+      "loss": 0.3567,
+      "step": 3891
+    },
+    {
+      "epoch": 1.166685403784898,
+      "grad_norm": 0.2304534763097763,
+      "learning_rate": 0.00012513513513513514,
+      "loss": 0.3668,
+      "step": 3892
+    },
+    {
+      "epoch": 1.1669851976765973,
+      "grad_norm": 0.22690708935260773,
+      "learning_rate": 0.00012509009009009007,
+      "loss": 0.3801,
+      "step": 3893
+    },
+    {
+      "epoch": 1.1672849915682968,
+      "grad_norm": 0.2346915751695633,
+      "learning_rate": 0.00012504504504504504,
+      "loss": 0.3595,
+      "step": 3894
+    },
+    {
+      "epoch": 1.1675847854599963,
+      "grad_norm": 0.22019316256046295,
+      "learning_rate": 0.000125,
+      "loss": 0.36,
+      "step": 3895
+    },
+    {
+      "epoch": 1.1678845793516957,
+      "grad_norm": 0.2175881713628769,
+      "learning_rate": 0.00012495495495495494,
+      "loss": 0.3526,
+      "step": 3896
+    },
+    {
+      "epoch": 1.1681843732433952,
+      "grad_norm": 0.22282525897026062,
+      "learning_rate": 0.0001249099099099099,
+      "loss": 0.3552,
+      "step": 3897
+    },
+    {
+      "epoch": 1.1684841671350945,
+      "grad_norm": 0.21789103746414185,
+      "learning_rate": 0.00012486486486486487,
+      "loss": 0.3591,
+      "step": 3898
+    },
+    {
+      "epoch": 1.168783961026794,
+      "grad_norm": 0.21931247413158417,
+      "learning_rate": 0.0001248198198198198,
+      "loss": 0.3583,
+      "step": 3899
+    },
+    {
+      "epoch": 1.1690837549184936,
+      "grad_norm": 0.22268635034561157,
+      "learning_rate": 0.00012477477477477477,
+      "loss": 0.3497,
+      "step": 3900
+    },
+    {
+      "epoch": 1.169383548810193,
+      "grad_norm": 0.23430244624614716,
+      "learning_rate": 0.00012472972972972973,
+      "loss": 0.3827,
+      "step": 3901
+    },
+    {
+      "epoch": 1.1696833427018924,
+      "grad_norm": 0.21991029381752014,
+      "learning_rate": 0.00012468468468468467,
+      "loss": 0.356,
+      "step": 3902
+    },
+    {
+      "epoch": 1.169983136593592,
+      "grad_norm": 0.20806784927845,
+      "learning_rate": 0.00012463963963963963,
+      "loss": 0.3456,
+      "step": 3903
+    },
+    {
+      "epoch": 1.1702829304852913,
+      "grad_norm": 0.22096467018127441,
+      "learning_rate": 0.00012459459459459457,
+      "loss": 0.3751,
+      "step": 3904
+    },
+    {
+      "epoch": 1.1705827243769908,
+      "grad_norm": 0.22532299160957336,
+      "learning_rate": 0.00012454954954954953,
+      "loss": 0.3598,
+      "step": 3905
+    },
+    {
+      "epoch": 1.1708825182686904,
+      "grad_norm": 0.23222416639328003,
+      "learning_rate": 0.0001245045045045045,
+      "loss": 0.368,
+      "step": 3906
+    },
+    {
+      "epoch": 1.1711823121603897,
+      "grad_norm": 0.21905198693275452,
+      "learning_rate": 0.00012445945945945943,
+      "loss": 0.3648,
+      "step": 3907
+    },
+    {
+      "epoch": 1.1714821060520892,
+      "grad_norm": 0.2258433699607849,
+      "learning_rate": 0.0001244144144144144,
+      "loss": 0.3768,
+      "step": 3908
+    },
+    {
+      "epoch": 1.1717818999437886,
+      "grad_norm": 0.21878860890865326,
+      "learning_rate": 0.00012436936936936936,
+      "loss": 0.3729,
+      "step": 3909
+    },
+    {
+      "epoch": 1.172081693835488,
+      "grad_norm": 0.22248351573944092,
+      "learning_rate": 0.0001243243243243243,
+      "loss": 0.3472,
+      "step": 3910
+    },
+    {
+      "epoch": 1.1723814877271876,
+      "grad_norm": 0.23260651528835297,
+      "learning_rate": 0.00012427927927927926,
+      "loss": 0.3664,
+      "step": 3911
+    },
+    {
+      "epoch": 1.172681281618887,
+      "grad_norm": 0.22885175049304962,
+      "learning_rate": 0.00012423423423423423,
+      "loss": 0.4064,
+      "step": 3912
+    },
+    {
+      "epoch": 1.1729810755105865,
+      "grad_norm": 0.22039201855659485,
+      "learning_rate": 0.00012418918918918916,
+      "loss": 0.3432,
+      "step": 3913
+    },
+    {
+      "epoch": 1.1732808694022858,
+      "grad_norm": 0.2224431037902832,
+      "learning_rate": 0.00012414414414414413,
+      "loss": 0.3673,
+      "step": 3914
+    },
+    {
+      "epoch": 1.1735806632939854,
+      "grad_norm": 0.21871282160282135,
+      "learning_rate": 0.0001240990990990991,
+      "loss": 0.3665,
+      "step": 3915
+    },
+    {
+      "epoch": 1.173880457185685,
+      "grad_norm": 0.22808001935482025,
+      "learning_rate": 0.00012405405405405405,
+      "loss": 0.3595,
+      "step": 3916
+    },
+    {
+      "epoch": 1.1741802510773842,
+      "grad_norm": 0.21829096972942352,
+      "learning_rate": 0.000124009009009009,
+      "loss": 0.3708,
+      "step": 3917
+    },
+    {
+      "epoch": 1.1744800449690838,
+      "grad_norm": 0.23878386616706848,
+      "learning_rate": 0.00012396396396396395,
+      "loss": 0.3469,
+      "step": 3918
+    },
+    {
+      "epoch": 1.1747798388607833,
+      "grad_norm": 0.22917965054512024,
+      "learning_rate": 0.00012391891891891892,
+      "loss": 0.3734,
+      "step": 3919
+    },
+    {
+      "epoch": 1.1750796327524826,
+      "grad_norm": 0.2284645140171051,
+      "learning_rate": 0.00012387387387387386,
+      "loss": 0.3721,
+      "step": 3920
+    },
+    {
+      "epoch": 1.1753794266441822,
+      "grad_norm": 0.21777629852294922,
+      "learning_rate": 0.00012382882882882882,
+      "loss": 0.3631,
+      "step": 3921
+    },
+    {
+      "epoch": 1.1756792205358817,
+      "grad_norm": 0.22827209532260895,
+      "learning_rate": 0.00012378378378378378,
+      "loss": 0.3962,
+      "step": 3922
+    },
+    {
+      "epoch": 1.175979014427581,
+      "grad_norm": 0.22360359132289886,
+      "learning_rate": 0.00012373873873873875,
+      "loss": 0.3624,
+      "step": 3923
+    },
+    {
+      "epoch": 1.1762788083192806,
+      "grad_norm": 0.20297862589359283,
+      "learning_rate": 0.00012369369369369368,
+      "loss": 0.3348,
+      "step": 3924
+    },
+    {
+      "epoch": 1.1765786022109799,
+      "grad_norm": 0.22274337708950043,
+      "learning_rate": 0.00012364864864864865,
+      "loss": 0.3659,
+      "step": 3925
+    },
+    {
+      "epoch": 1.1768783961026794,
+      "grad_norm": 0.21908465027809143,
+      "learning_rate": 0.00012360360360360358,
+      "loss": 0.367,
+      "step": 3926
+    },
+    {
+      "epoch": 1.177178189994379,
+      "grad_norm": 0.2173020988702774,
+      "learning_rate": 0.00012355855855855855,
+      "loss": 0.3718,
+      "step": 3927
+    },
+    {
+      "epoch": 1.1774779838860783,
+      "grad_norm": 0.21801786124706268,
+      "learning_rate": 0.0001235135135135135,
+      "loss": 0.3358,
+      "step": 3928
+    },
+    {
+      "epoch": 1.1777777777777778,
+      "grad_norm": 0.22594833374023438,
+      "learning_rate": 0.00012346846846846845,
+      "loss": 0.3646,
+      "step": 3929
+    },
+    {
+      "epoch": 1.1780775716694771,
+      "grad_norm": 0.2159031629562378,
+      "learning_rate": 0.0001234234234234234,
+      "loss": 0.3459,
+      "step": 3930
+    },
+    {
+      "epoch": 1.1783773655611767,
+      "grad_norm": 0.23376664519309998,
+      "learning_rate": 0.00012337837837837838,
+      "loss": 0.3899,
+      "step": 3931
+    },
+    {
+      "epoch": 1.1786771594528762,
+      "grad_norm": 0.2555432915687561,
+      "learning_rate": 0.0001233333333333333,
+      "loss": 0.3729,
+      "step": 3932
+    },
+    {
+      "epoch": 1.1789769533445755,
+      "grad_norm": 0.2331700474023819,
+      "learning_rate": 0.00012328828828828828,
+      "loss": 0.3783,
+      "step": 3933
+    },
+    {
+      "epoch": 1.179276747236275,
+      "grad_norm": 0.2171568125486374,
+      "learning_rate": 0.00012324324324324324,
+      "loss": 0.3588,
+      "step": 3934
+    },
+    {
+      "epoch": 1.1795765411279746,
+      "grad_norm": 0.2255590707063675,
+      "learning_rate": 0.00012319819819819818,
+      "loss": 0.3486,
+      "step": 3935
+    },
+    {
+      "epoch": 1.179876335019674,
+      "grad_norm": 0.23395510017871857,
+      "learning_rate": 0.00012315315315315314,
+      "loss": 0.3399,
+      "step": 3936
+    },
+    {
+      "epoch": 1.1801761289113735,
+      "grad_norm": 0.22677238285541534,
+      "learning_rate": 0.0001231081081081081,
+      "loss": 0.3917,
+      "step": 3937
+    },
+    {
+      "epoch": 1.180475922803073,
+      "grad_norm": 0.22630028426647186,
+      "learning_rate": 0.00012306306306306304,
+      "loss": 0.3717,
+      "step": 3938
+    },
+    {
+      "epoch": 1.1807757166947723,
+      "grad_norm": 0.22929327189922333,
+      "learning_rate": 0.000123018018018018,
+      "loss": 0.3797,
+      "step": 3939
+    },
+    {
+      "epoch": 1.1810755105864719,
+      "grad_norm": 0.24179096519947052,
+      "learning_rate": 0.00012297297297297294,
+      "loss": 0.4066,
+      "step": 3940
+    },
+    {
+      "epoch": 1.1813753044781712,
+      "grad_norm": 0.2346000075340271,
+      "learning_rate": 0.0001229279279279279,
+      "loss": 0.3587,
+      "step": 3941
+    },
+    {
+      "epoch": 1.1816750983698707,
+      "grad_norm": 0.21742184460163116,
+      "learning_rate": 0.00012288288288288287,
+      "loss": 0.3498,
+      "step": 3942
+    },
+    {
+      "epoch": 1.1819748922615703,
+      "grad_norm": 0.22498059272766113,
+      "learning_rate": 0.00012283783783783783,
+      "loss": 0.3469,
+      "step": 3943
+    },
+    {
+      "epoch": 1.1822746861532696,
+      "grad_norm": 0.22407019138336182,
+      "learning_rate": 0.00012279279279279277,
+      "loss": 0.3871,
+      "step": 3944
+    },
+    {
+      "epoch": 1.1825744800449691,
+      "grad_norm": 0.2243795543909073,
+      "learning_rate": 0.00012274774774774774,
+      "loss": 0.3562,
+      "step": 3945
+    },
+    {
+      "epoch": 1.1828742739366684,
+      "grad_norm": 0.22106344997882843,
+      "learning_rate": 0.0001227027027027027,
+      "loss": 0.3585,
+      "step": 3946
+    },
+    {
+      "epoch": 1.183174067828368,
+      "grad_norm": 0.21711674332618713,
+      "learning_rate": 0.00012265765765765764,
+      "loss": 0.3635,
+      "step": 3947
+    },
+    {
+      "epoch": 1.1834738617200675,
+      "grad_norm": 0.22720924019813538,
+      "learning_rate": 0.0001226126126126126,
+      "loss": 0.3617,
+      "step": 3948
+    },
+    {
+      "epoch": 1.1837736556117668,
+      "grad_norm": 0.21701271831989288,
+      "learning_rate": 0.00012256756756756756,
+      "loss": 0.3605,
+      "step": 3949
+    },
+    {
+      "epoch": 1.1840734495034664,
+      "grad_norm": 0.23393398523330688,
+      "learning_rate": 0.00012252252252252253,
+      "loss": 0.3743,
+      "step": 3950
+    },
+    {
+      "epoch": 1.184373243395166,
+      "grad_norm": 0.21987706422805786,
+      "learning_rate": 0.00012247747747747746,
+      "loss": 0.3536,
+      "step": 3951
+    },
+    {
+      "epoch": 1.1846730372868652,
+      "grad_norm": 0.22225283086299896,
+      "learning_rate": 0.00012243243243243243,
+      "loss": 0.3538,
+      "step": 3952
+    },
+    {
+      "epoch": 1.1849728311785648,
+      "grad_norm": 0.24060380458831787,
+      "learning_rate": 0.0001223873873873874,
+      "loss": 0.3733,
+      "step": 3953
+    },
+    {
+      "epoch": 1.185272625070264,
+      "grad_norm": 0.2302178144454956,
+      "learning_rate": 0.00012234234234234233,
+      "loss": 0.3707,
+      "step": 3954
+    },
+    {
+      "epoch": 1.1855724189619636,
+      "grad_norm": 0.22814485430717468,
+      "learning_rate": 0.0001222972972972973,
+      "loss": 0.3678,
+      "step": 3955
+    },
+    {
+      "epoch": 1.1858722128536632,
+      "grad_norm": 0.24073362350463867,
+      "learning_rate": 0.00012225225225225226,
+      "loss": 0.3746,
+      "step": 3956
+    },
+    {
+      "epoch": 1.1861720067453625,
+      "grad_norm": 0.21149156987667084,
+      "learning_rate": 0.0001222072072072072,
+      "loss": 0.3266,
+      "step": 3957
+    },
+    {
+      "epoch": 1.186471800637062,
+      "grad_norm": 0.21905194222927094,
+      "learning_rate": 0.00012216216216216216,
+      "loss": 0.3426,
+      "step": 3958
+    },
+    {
+      "epoch": 1.1867715945287616,
+      "grad_norm": 0.22855554521083832,
+      "learning_rate": 0.00012211711711711712,
+      "loss": 0.3624,
+      "step": 3959
+    },
+    {
+      "epoch": 1.187071388420461,
+      "grad_norm": 0.23453767597675323,
+      "learning_rate": 0.00012207207207207206,
+      "loss": 0.3742,
+      "step": 3960
+    },
+    {
+      "epoch": 1.1873711823121604,
+      "grad_norm": 0.23686885833740234,
+      "learning_rate": 0.00012202702702702702,
+      "loss": 0.3956,
+      "step": 3961
+    },
+    {
+      "epoch": 1.1876709762038598,
+      "grad_norm": 0.2404448688030243,
+      "learning_rate": 0.00012198198198198196,
+      "loss": 0.3717,
+      "step": 3962
+    },
+    {
+      "epoch": 1.1879707700955593,
+      "grad_norm": 0.2192368060350418,
+      "learning_rate": 0.00012193693693693692,
+      "loss": 0.3692,
+      "step": 3963
+    },
+    {
+      "epoch": 1.1882705639872588,
+      "grad_norm": 0.2268245369195938,
+      "learning_rate": 0.00012189189189189189,
+      "loss": 0.3722,
+      "step": 3964
+    },
+    {
+      "epoch": 1.1885703578789582,
+      "grad_norm": 0.23420900106430054,
+      "learning_rate": 0.00012184684684684682,
+      "loss": 0.3696,
+      "step": 3965
+    },
+    {
+      "epoch": 1.1888701517706577,
+      "grad_norm": 0.23050832748413086,
+      "learning_rate": 0.00012180180180180179,
+      "loss": 0.3836,
+      "step": 3966
+    },
+    {
+      "epoch": 1.189169945662357,
+      "grad_norm": 0.25578629970550537,
+      "learning_rate": 0.00012175675675675675,
+      "loss": 0.3358,
+      "step": 3967
+    },
+    {
+      "epoch": 1.1894697395540565,
+      "grad_norm": 0.21223682165145874,
+      "learning_rate": 0.0001217117117117117,
+      "loss": 0.354,
+      "step": 3968
+    },
+    {
+      "epoch": 1.189769533445756,
+      "grad_norm": 0.2102338820695877,
+      "learning_rate": 0.00012166666666666665,
+      "loss": 0.3485,
+      "step": 3969
+    },
+    {
+      "epoch": 1.1900693273374554,
+      "grad_norm": 0.22144131362438202,
+      "learning_rate": 0.00012162162162162162,
+      "loss": 0.3503,
+      "step": 3970
+    },
+    {
+      "epoch": 1.190369121229155,
+      "grad_norm": 0.22547723352909088,
+      "learning_rate": 0.00012157657657657657,
+      "loss": 0.3569,
+      "step": 3971
+    },
+    {
+      "epoch": 1.1906689151208545,
+      "grad_norm": 0.20872855186462402,
+      "learning_rate": 0.00012153153153153153,
+      "loss": 0.343,
+      "step": 3972
+    },
+    {
+      "epoch": 1.1909687090125538,
+      "grad_norm": 0.2280842363834381,
+      "learning_rate": 0.00012148648648648648,
+      "loss": 0.3911,
+      "step": 3973
+    },
+    {
+      "epoch": 1.1912685029042533,
+      "grad_norm": 0.23819836974143982,
+      "learning_rate": 0.00012144144144144143,
+      "loss": 0.3952,
+      "step": 3974
+    },
+    {
+      "epoch": 1.1915682967959529,
+      "grad_norm": 0.22785331308841705,
+      "learning_rate": 0.0001213963963963964,
+      "loss": 0.3559,
+      "step": 3975
+    },
+    {
+      "epoch": 1.1918680906876522,
+      "grad_norm": 0.21852603554725647,
+      "learning_rate": 0.00012135135135135133,
+      "loss": 0.3413,
+      "step": 3976
+    },
+    {
+      "epoch": 1.1921678845793517,
+      "grad_norm": 0.2668752670288086,
+      "learning_rate": 0.0001213063063063063,
+      "loss": 0.4048,
+      "step": 3977
+    },
+    {
+      "epoch": 1.192467678471051,
+      "grad_norm": 0.223458930850029,
+      "learning_rate": 0.00012126126126126126,
+      "loss": 0.3763,
+      "step": 3978
+    },
+    {
+      "epoch": 1.1927674723627506,
+      "grad_norm": 0.2441006898880005,
+      "learning_rate": 0.0001212162162162162,
+      "loss": 0.3752,
+      "step": 3979
+    },
+    {
+      "epoch": 1.1930672662544501,
+      "grad_norm": 0.241143599152565,
+      "learning_rate": 0.00012117117117117116,
+      "loss": 0.39,
+      "step": 3980
+    },
+    {
+      "epoch": 1.1933670601461495,
+      "grad_norm": 0.22745396196842194,
+      "learning_rate": 0.00012112612612612612,
+      "loss": 0.3755,
+      "step": 3981
+    },
+    {
+      "epoch": 1.193666854037849,
+      "grad_norm": 0.21644215285778046,
+      "learning_rate": 0.00012108108108108106,
+      "loss": 0.3448,
+      "step": 3982
+    },
+    {
+      "epoch": 1.1939666479295483,
+      "grad_norm": 0.23860111832618713,
+      "learning_rate": 0.00012103603603603602,
+      "loss": 0.3878,
+      "step": 3983
+    },
+    {
+      "epoch": 1.1942664418212479,
+      "grad_norm": 0.21523654460906982,
+      "learning_rate": 0.00012099099099099099,
+      "loss": 0.3428,
+      "step": 3984
+    },
+    {
+      "epoch": 1.1945662357129474,
+      "grad_norm": 0.24360989034175873,
+      "learning_rate": 0.00012094594594594594,
+      "loss": 0.3778,
+      "step": 3985
+    },
+    {
+      "epoch": 1.1948660296046467,
+      "grad_norm": 0.23723161220550537,
+      "learning_rate": 0.00012090090090090089,
+      "loss": 0.3658,
+      "step": 3986
+    },
+    {
+      "epoch": 1.1951658234963463,
+      "grad_norm": 0.22299665212631226,
+      "learning_rate": 0.00012085585585585584,
+      "loss": 0.3742,
+      "step": 3987
+    },
+    {
+      "epoch": 1.1954656173880458,
+      "grad_norm": 0.2324703186750412,
+      "learning_rate": 0.0001208108108108108,
+      "loss": 0.3709,
+      "step": 3988
+    },
+    {
+      "epoch": 1.1957654112797451,
+      "grad_norm": 0.22136402130126953,
+      "learning_rate": 0.00012076576576576577,
+      "loss": 0.3662,
+      "step": 3989
+    },
+    {
+      "epoch": 1.1960652051714447,
+      "grad_norm": 0.23628540337085724,
+      "learning_rate": 0.0001207207207207207,
+      "loss": 0.3409,
+      "step": 3990
+    },
+    {
+      "epoch": 1.1963649990631442,
+      "grad_norm": 0.21257908642292023,
+      "learning_rate": 0.00012067567567567567,
+      "loss": 0.3382,
+      "step": 3991
+    },
+    {
+      "epoch": 1.1966647929548435,
+      "grad_norm": 0.22111082077026367,
+      "learning_rate": 0.00012063063063063063,
+      "loss": 0.3529,
+      "step": 3992
+    },
+    {
+      "epoch": 1.196964586846543,
+      "grad_norm": 0.23919793963432312,
+      "learning_rate": 0.00012058558558558557,
+      "loss": 0.3831,
+      "step": 3993
+    },
+    {
+      "epoch": 1.1972643807382424,
+      "grad_norm": 0.20639801025390625,
+      "learning_rate": 0.00012054054054054053,
+      "loss": 0.333,
+      "step": 3994
+    },
+    {
+      "epoch": 1.197564174629942,
+      "grad_norm": 0.2506180703639984,
+      "learning_rate": 0.0001204954954954955,
+      "loss": 0.3755,
+      "step": 3995
+    },
+    {
+      "epoch": 1.1978639685216415,
+      "grad_norm": 0.2168271392583847,
+      "learning_rate": 0.00012045045045045043,
+      "loss": 0.3483,
+      "step": 3996
+    },
+    {
+      "epoch": 1.1981637624133408,
+      "grad_norm": 0.21680712699890137,
+      "learning_rate": 0.0001204054054054054,
+      "loss": 0.3411,
+      "step": 3997
+    },
+    {
+      "epoch": 1.1984635563050403,
+      "grad_norm": 0.2226821780204773,
+      "learning_rate": 0.00012036036036036035,
+      "loss": 0.3511,
+      "step": 3998
+    },
+    {
+      "epoch": 1.1987633501967396,
+      "grad_norm": 0.23245584964752197,
+      "learning_rate": 0.0001203153153153153,
+      "loss": 0.3554,
+      "step": 3999
+    },
+    {
+      "epoch": 1.1990631440884392,
+      "grad_norm": 0.2246054708957672,
+      "learning_rate": 0.00012027027027027026,
+      "loss": 0.3684,
+      "step": 4000
+    },
+    {
+      "epoch": 1.1990631440884392,
+      "eval_loss": 0.41834643483161926,
+      "eval_runtime": 565.2826,
+      "eval_samples_per_second": 3.819,
+      "eval_steps_per_second": 0.478,
+      "step": 4000
+    },
+    {
+      "epoch": 1.1993629379801387,
+      "grad_norm": 0.22603508830070496,
+      "learning_rate": 0.00012022522522522521,
+      "loss": 0.3897,
+      "step": 4001
+    },
+    {
+      "epoch": 1.199662731871838,
+      "grad_norm": 0.2275640070438385,
+      "learning_rate": 0.00012018018018018017,
+      "loss": 0.3567,
+      "step": 4002
+    },
+    {
+      "epoch": 1.1999625257635376,
+      "grad_norm": 0.23352980613708496,
+      "learning_rate": 0.00012013513513513512,
+      "loss": 0.3807,
+      "step": 4003
+    },
+    {
+      "epoch": 1.200262319655237,
+      "grad_norm": 0.2211960405111313,
+      "learning_rate": 0.00012009009009009008,
+      "loss": 0.3546,
+      "step": 4004
+    },
+    {
+      "epoch": 1.2005621135469364,
+      "grad_norm": 0.2441369891166687,
+      "learning_rate": 0.00012004504504504504,
+      "loss": 0.3776,
+      "step": 4005
+    },
+    {
+      "epoch": 1.200861907438636,
+      "grad_norm": 0.22006067633628845,
+      "learning_rate": 0.00011999999999999999,
+      "loss": 0.3705,
+      "step": 4006
+    },
+    {
+      "epoch": 1.2011617013303355,
+      "grad_norm": 0.23932883143424988,
+      "learning_rate": 0.00011995495495495494,
+      "loss": 0.3665,
+      "step": 4007
+    },
+    {
+      "epoch": 1.2014614952220348,
+      "grad_norm": 0.2247323989868164,
+      "learning_rate": 0.0001199099099099099,
+      "loss": 0.3859,
+      "step": 4008
+    },
+    {
+      "epoch": 1.2017612891137344,
+      "grad_norm": 0.22732797265052795,
+      "learning_rate": 0.00011986486486486487,
+      "loss": 0.3782,
+      "step": 4009
+    },
+    {
+      "epoch": 1.2020610830054337,
+      "grad_norm": 0.22873279452323914,
+      "learning_rate": 0.0001198198198198198,
+      "loss": 0.3625,
+      "step": 4010
+    },
+    {
+      "epoch": 1.2023608768971332,
+      "grad_norm": 0.22076311707496643,
+      "learning_rate": 0.00011977477477477477,
+      "loss": 0.3578,
+      "step": 4011
+    },
+    {
+      "epoch": 1.2026606707888328,
+      "grad_norm": 0.2291504144668579,
+      "learning_rate": 0.00011972972972972972,
+      "loss": 0.3457,
+      "step": 4012
+    },
+    {
+      "epoch": 1.202960464680532,
+      "grad_norm": 0.22902348637580872,
+      "learning_rate": 0.00011968468468468467,
+      "loss": 0.3823,
+      "step": 4013
+    },
+    {
+      "epoch": 1.2032602585722316,
+      "grad_norm": 0.22112970054149628,
+      "learning_rate": 0.00011963963963963963,
+      "loss": 0.3687,
+      "step": 4014
+    },
+    {
+      "epoch": 1.203560052463931,
+      "grad_norm": 0.2165578454732895,
+      "learning_rate": 0.00011959459459459458,
+      "loss": 0.3469,
+      "step": 4015
+    },
+    {
+      "epoch": 1.2038598463556305,
+      "grad_norm": 0.2269822061061859,
+      "learning_rate": 0.00011954954954954953,
+      "loss": 0.3643,
+      "step": 4016
+    },
+    {
+      "epoch": 1.20415964024733,
+      "grad_norm": 0.23120154440402985,
+      "learning_rate": 0.0001195045045045045,
+      "loss": 0.3833,
+      "step": 4017
+    },
+    {
+      "epoch": 1.2044594341390293,
+      "grad_norm": 0.22641071677207947,
+      "learning_rate": 0.00011945945945945945,
+      "loss": 0.3718,
+      "step": 4018
+    },
+    {
+      "epoch": 1.2047592280307289,
+      "grad_norm": 0.23450811207294464,
+      "learning_rate": 0.00011941441441441441,
+      "loss": 0.3817,
+      "step": 4019
+    },
+    {
+      "epoch": 1.2050590219224284,
+      "grad_norm": 0.2254931777715683,
+      "learning_rate": 0.00011936936936936936,
+      "loss": 0.3693,
+      "step": 4020
+    },
+    {
+      "epoch": 1.2053588158141277,
+      "grad_norm": 0.21437515318393707,
+      "learning_rate": 0.00011932432432432431,
+      "loss": 0.3499,
+      "step": 4021
+    },
+    {
+      "epoch": 1.2056586097058273,
+      "grad_norm": 0.21812491118907928,
+      "learning_rate": 0.00011927927927927928,
+      "loss": 0.3615,
+      "step": 4022
+    },
+    {
+      "epoch": 1.2059584035975268,
+      "grad_norm": 0.23963527381420135,
+      "learning_rate": 0.00011923423423423421,
+      "loss": 0.3799,
+      "step": 4023
+    },
+    {
+      "epoch": 1.2062581974892261,
+      "grad_norm": 0.2252485454082489,
+      "learning_rate": 0.00011918918918918918,
+      "loss": 0.3422,
+      "step": 4024
+    },
+    {
+      "epoch": 1.2065579913809257,
+      "grad_norm": 0.23507623374462128,
+      "learning_rate": 0.00011914414414414414,
+      "loss": 0.3572,
+      "step": 4025
+    },
+    {
+      "epoch": 1.206857785272625,
+      "grad_norm": 0.2158452719449997,
+      "learning_rate": 0.00011909909909909908,
+      "loss": 0.3771,
+      "step": 4026
+    },
+    {
+      "epoch": 1.2071575791643245,
+      "grad_norm": 0.24523043632507324,
+      "learning_rate": 0.00011905405405405404,
+      "loss": 0.3615,
+      "step": 4027
+    },
+    {
+      "epoch": 1.207457373056024,
+      "grad_norm": 0.2354147732257843,
+      "learning_rate": 0.000119009009009009,
+      "loss": 0.3614,
+      "step": 4028
+    },
+    {
+      "epoch": 1.2077571669477234,
+      "grad_norm": 0.22273007035255432,
+      "learning_rate": 0.00011896396396396396,
+      "loss": 0.3398,
+      "step": 4029
+    },
+    {
+      "epoch": 1.208056960839423,
+      "grad_norm": 0.22855806350708008,
+      "learning_rate": 0.0001189189189189189,
+      "loss": 0.3706,
+      "step": 4030
+    },
+    {
+      "epoch": 1.2083567547311223,
+      "grad_norm": 0.23568007349967957,
+      "learning_rate": 0.00011887387387387387,
+      "loss": 0.3952,
+      "step": 4031
+    },
+    {
+      "epoch": 1.2086565486228218,
+      "grad_norm": 0.215668722987175,
+      "learning_rate": 0.00011882882882882882,
+      "loss": 0.3492,
+      "step": 4032
+    },
+    {
+      "epoch": 1.2089563425145213,
+      "grad_norm": 0.2247469574213028,
+      "learning_rate": 0.00011878378378378377,
+      "loss": 0.3547,
+      "step": 4033
+    },
+    {
+      "epoch": 1.2092561364062206,
+      "grad_norm": 0.2222067266702652,
+      "learning_rate": 0.00011873873873873872,
+      "loss": 0.35,
+      "step": 4034
+    },
+    {
+      "epoch": 1.2095559302979202,
+      "grad_norm": 0.22392643988132477,
+      "learning_rate": 0.00011869369369369368,
+      "loss": 0.3445,
+      "step": 4035
+    },
+    {
+      "epoch": 1.2098557241896197,
+      "grad_norm": 0.22978229820728302,
+      "learning_rate": 0.00011864864864864865,
+      "loss": 0.3806,
+      "step": 4036
+    },
+    {
+      "epoch": 1.210155518081319,
+      "grad_norm": 0.23751720786094666,
+      "learning_rate": 0.00011860360360360358,
+      "loss": 0.3567,
+      "step": 4037
+    },
+    {
+      "epoch": 1.2104553119730186,
+      "grad_norm": 0.2222888469696045,
+      "learning_rate": 0.00011855855855855855,
+      "loss": 0.3583,
+      "step": 4038
+    },
+    {
+      "epoch": 1.2107551058647181,
+      "grad_norm": 0.24681641161441803,
+      "learning_rate": 0.00011851351351351351,
+      "loss": 0.3746,
+      "step": 4039
+    },
+    {
+      "epoch": 1.2110548997564174,
+      "grad_norm": 0.2182309329509735,
+      "learning_rate": 0.00011846846846846845,
+      "loss": 0.3348,
+      "step": 4040
+    },
+    {
+      "epoch": 1.211354693648117,
+      "grad_norm": 0.21371771395206451,
+      "learning_rate": 0.00011842342342342341,
+      "loss": 0.3338,
+      "step": 4041
+    },
+    {
+      "epoch": 1.2116544875398163,
+      "grad_norm": 0.23629151284694672,
+      "learning_rate": 0.00011837837837837838,
+      "loss": 0.3408,
+      "step": 4042
+    },
+    {
+      "epoch": 1.2119542814315158,
+      "grad_norm": 0.23609225451946259,
+      "learning_rate": 0.00011833333333333331,
+      "loss": 0.38,
+      "step": 4043
+    },
+    {
+      "epoch": 1.2122540753232154,
+      "grad_norm": 0.21824340522289276,
+      "learning_rate": 0.00011828828828828828,
+      "loss": 0.3355,
+      "step": 4044
+    },
+    {
+      "epoch": 1.2125538692149147,
+      "grad_norm": 0.2105160504579544,
+      "learning_rate": 0.00011824324324324324,
+      "loss": 0.353,
+      "step": 4045
+    },
+    {
+      "epoch": 1.2128536631066142,
+      "grad_norm": 0.2521781027317047,
+      "learning_rate": 0.00011819819819819818,
+      "loss": 0.3899,
+      "step": 4046
+    },
+    {
+      "epoch": 1.2131534569983136,
+      "grad_norm": 0.21568675339221954,
+      "learning_rate": 0.00011815315315315314,
+      "loss": 0.3575,
+      "step": 4047
+    },
+    {
+      "epoch": 1.213453250890013,
+      "grad_norm": 0.21583522856235504,
+      "learning_rate": 0.00011810810810810809,
+      "loss": 0.365,
+      "step": 4048
+    },
+    {
+      "epoch": 1.2137530447817126,
+      "grad_norm": 0.23436518013477325,
+      "learning_rate": 0.00011806306306306306,
+      "loss": 0.3783,
+      "step": 4049
+    },
+    {
+      "epoch": 1.214052838673412,
+      "grad_norm": 0.21352584660053253,
+      "learning_rate": 0.00011801801801801801,
+      "loss": 0.3446,
+      "step": 4050
+    },
+    {
+      "epoch": 1.2143526325651115,
+      "grad_norm": 0.2259955108165741,
+      "learning_rate": 0.00011797297297297296,
+      "loss": 0.3866,
+      "step": 4051
+    },
+    {
+      "epoch": 1.2146524264568108,
+      "grad_norm": 0.215003103017807,
+      "learning_rate": 0.00011792792792792792,
+      "loss": 0.3372,
+      "step": 4052
+    },
+    {
+      "epoch": 1.2149522203485104,
+      "grad_norm": 0.21452820301055908,
+      "learning_rate": 0.00011788288288288288,
+      "loss": 0.333,
+      "step": 4053
+    },
+    {
+      "epoch": 1.21525201424021,
+      "grad_norm": 0.21691389381885529,
+      "learning_rate": 0.00011783783783783782,
+      "loss": 0.3588,
+      "step": 4054
+    },
+    {
+      "epoch": 1.2155518081319092,
+      "grad_norm": 0.22161614894866943,
+      "learning_rate": 0.00011779279279279279,
+      "loss": 0.366,
+      "step": 4055
+    },
+    {
+      "epoch": 1.2158516020236088,
+      "grad_norm": 0.22550690174102783,
+      "learning_rate": 0.00011774774774774775,
+      "loss": 0.3535,
+      "step": 4056
+    },
+    {
+      "epoch": 1.2161513959153083,
+      "grad_norm": 0.24499079585075378,
+      "learning_rate": 0.00011770270270270269,
+      "loss": 0.387,
+      "step": 4057
+    },
+    {
+      "epoch": 1.2164511898070076,
+      "grad_norm": 0.21006570756435394,
+      "learning_rate": 0.00011765765765765765,
+      "loss": 0.3313,
+      "step": 4058
+    },
+    {
+      "epoch": 1.2167509836987072,
+      "grad_norm": 0.24210864305496216,
+      "learning_rate": 0.0001176126126126126,
+      "loss": 0.3861,
+      "step": 4059
+    },
+    {
+      "epoch": 1.2170507775904067,
+      "grad_norm": 0.22041113674640656,
+      "learning_rate": 0.00011756756756756755,
+      "loss": 0.357,
+      "step": 4060
+    },
+    {
+      "epoch": 1.217350571482106,
+      "grad_norm": 0.22228474915027618,
+      "learning_rate": 0.00011752252252252251,
+      "loss": 0.3738,
+      "step": 4061
+    },
+    {
+      "epoch": 1.2176503653738056,
+      "grad_norm": 0.22721122205257416,
+      "learning_rate": 0.00011747747747747746,
+      "loss": 0.3829,
+      "step": 4062
+    },
+    {
+      "epoch": 1.2179501592655049,
+      "grad_norm": 0.23477396368980408,
+      "learning_rate": 0.00011743243243243242,
+      "loss": 0.405,
+      "step": 4063
+    },
+    {
+      "epoch": 1.2182499531572044,
+      "grad_norm": 0.22051817178726196,
+      "learning_rate": 0.00011738738738738738,
+      "loss": 0.3666,
+      "step": 4064
+    },
+    {
+      "epoch": 1.218549747048904,
+      "grad_norm": 0.2251734584569931,
+      "learning_rate": 0.00011734234234234233,
+      "loss": 0.3843,
+      "step": 4065
+    },
+    {
+      "epoch": 1.2188495409406033,
+      "grad_norm": 0.223738893866539,
+      "learning_rate": 0.0001172972972972973,
+      "loss": 0.3734,
+      "step": 4066
+    },
+    {
+      "epoch": 1.2191493348323028,
+      "grad_norm": 0.2110142558813095,
+      "learning_rate": 0.00011725225225225224,
+      "loss": 0.3412,
+      "step": 4067
+    },
+    {
+      "epoch": 1.2194491287240021,
+      "grad_norm": 0.2286280393600464,
+      "learning_rate": 0.0001172072072072072,
+      "loss": 0.3668,
+      "step": 4068
+    },
+    {
+      "epoch": 1.2197489226157017,
+      "grad_norm": 0.22667136788368225,
+      "learning_rate": 0.00011716216216216216,
+      "loss": 0.3575,
+      "step": 4069
+    },
+    {
+      "epoch": 1.2200487165074012,
+      "grad_norm": 0.2305767983198166,
+      "learning_rate": 0.0001171171171171171,
+      "loss": 0.3581,
+      "step": 4070
+    },
+    {
+      "epoch": 1.2203485103991005,
+      "grad_norm": 0.2244352102279663,
+      "learning_rate": 0.00011707207207207206,
+      "loss": 0.3569,
+      "step": 4071
+    },
+    {
+      "epoch": 1.2206483042908,
+      "grad_norm": 0.2392815798521042,
+      "learning_rate": 0.00011702702702702702,
+      "loss": 0.3671,
+      "step": 4072
+    },
+    {
+      "epoch": 1.2209480981824996,
+      "grad_norm": 0.2159206122159958,
+      "learning_rate": 0.00011698198198198196,
+      "loss": 0.3403,
+      "step": 4073
+    },
+    {
+      "epoch": 1.221247892074199,
+      "grad_norm": 0.21920053660869598,
+      "learning_rate": 0.00011693693693693692,
+      "loss": 0.3549,
+      "step": 4074
+    },
+    {
+      "epoch": 1.2215476859658985,
+      "grad_norm": 0.2131916582584381,
+      "learning_rate": 0.00011689189189189189,
+      "loss": 0.337,
+      "step": 4075
+    },
+    {
+      "epoch": 1.221847479857598,
+      "grad_norm": 0.22338977456092834,
+      "learning_rate": 0.00011684684684684684,
+      "loss": 0.3591,
+      "step": 4076
+    },
+    {
+      "epoch": 1.2221472737492973,
+      "grad_norm": 0.22058121860027313,
+      "learning_rate": 0.00011680180180180179,
+      "loss": 0.3529,
+      "step": 4077
+    },
+    {
+      "epoch": 1.2224470676409969,
+      "grad_norm": 0.2266795039176941,
+      "learning_rate": 0.00011675675675675675,
+      "loss": 0.372,
+      "step": 4078
+    },
+    {
+      "epoch": 1.2227468615326962,
+      "grad_norm": 0.23163147270679474,
+      "learning_rate": 0.0001167117117117117,
+      "loss": 0.3737,
+      "step": 4079
+    },
+    {
+      "epoch": 1.2230466554243957,
+      "grad_norm": 0.22143979370594025,
+      "learning_rate": 0.00011666666666666665,
+      "loss": 0.3743,
+      "step": 4080
+    },
+    {
+      "epoch": 1.2233464493160953,
+      "grad_norm": 0.21844099462032318,
+      "learning_rate": 0.00011662162162162162,
+      "loss": 0.3538,
+      "step": 4081
+    },
+    {
+      "epoch": 1.2236462432077946,
+      "grad_norm": 0.22864669561386108,
+      "learning_rate": 0.00011657657657657657,
+      "loss": 0.3325,
+      "step": 4082
+    },
+    {
+      "epoch": 1.2239460370994941,
+      "grad_norm": 0.22677049040794373,
+      "learning_rate": 0.00011653153153153153,
+      "loss": 0.3708,
+      "step": 4083
+    },
+    {
+      "epoch": 1.2242458309911934,
+      "grad_norm": 0.2209145426750183,
+      "learning_rate": 0.00011648648648648647,
+      "loss": 0.3529,
+      "step": 4084
+    },
+    {
+      "epoch": 1.224545624882893,
+      "grad_norm": 0.22299690544605255,
+      "learning_rate": 0.00011644144144144143,
+      "loss": 0.3471,
+      "step": 4085
+    },
+    {
+      "epoch": 1.2248454187745925,
+      "grad_norm": 0.2258932739496231,
+      "learning_rate": 0.0001163963963963964,
+      "loss": 0.3729,
+      "step": 4086
+    },
+    {
+      "epoch": 1.2251452126662918,
+      "grad_norm": 0.21733281016349792,
+      "learning_rate": 0.00011635135135135133,
+      "loss": 0.3464,
+      "step": 4087
+    },
+    {
+      "epoch": 1.2254450065579914,
+      "grad_norm": 0.2246559113264084,
+      "learning_rate": 0.0001163063063063063,
+      "loss": 0.3618,
+      "step": 4088
+    },
+    {
+      "epoch": 1.225744800449691,
+      "grad_norm": 0.2373102605342865,
+      "learning_rate": 0.00011626126126126126,
+      "loss": 0.3695,
+      "step": 4089
+    },
+    {
+      "epoch": 1.2260445943413902,
+      "grad_norm": 0.21920005977153778,
+      "learning_rate": 0.0001162162162162162,
+      "loss": 0.3697,
+      "step": 4090
+    },
+    {
+      "epoch": 1.2263443882330898,
+      "grad_norm": 0.22167542576789856,
+      "learning_rate": 0.00011617117117117116,
+      "loss": 0.3613,
+      "step": 4091
+    },
+    {
+      "epoch": 1.2266441821247893,
+      "grad_norm": 0.21665841341018677,
+      "learning_rate": 0.00011612612612612612,
+      "loss": 0.36,
+      "step": 4092
+    },
+    {
+      "epoch": 1.2269439760164886,
+      "grad_norm": 0.22142474353313446,
+      "learning_rate": 0.00011608108108108107,
+      "loss": 0.3577,
+      "step": 4093
+    },
+    {
+      "epoch": 1.2272437699081882,
+      "grad_norm": 0.21807880699634552,
+      "learning_rate": 0.00011603603603603602,
+      "loss": 0.3598,
+      "step": 4094
+    },
+    {
+      "epoch": 1.2275435637998875,
+      "grad_norm": 0.21711617708206177,
+      "learning_rate": 0.00011599099099099097,
+      "loss": 0.3609,
+      "step": 4095
+    },
+    {
+      "epoch": 1.227843357691587,
+      "grad_norm": 0.2247924506664276,
+      "learning_rate": 0.00011594594594594594,
+      "loss": 0.3485,
+      "step": 4096
+    },
+    {
+      "epoch": 1.2281431515832866,
+      "grad_norm": 0.22750286757946014,
+      "learning_rate": 0.00011590090090090089,
+      "loss": 0.3669,
+      "step": 4097
+    },
+    {
+      "epoch": 1.228442945474986,
+      "grad_norm": 0.21605989336967468,
+      "learning_rate": 0.00011585585585585584,
+      "loss": 0.3658,
+      "step": 4098
+    },
+    {
+      "epoch": 1.2287427393666854,
+      "grad_norm": 0.217095747590065,
+      "learning_rate": 0.0001158108108108108,
+      "loss": 0.3469,
+      "step": 4099
+    },
+    {
+      "epoch": 1.2290425332583848,
+      "grad_norm": 0.22495174407958984,
+      "learning_rate": 0.00011576576576576577,
+      "loss": 0.3741,
+      "step": 4100
+    },
+    {
+      "epoch": 1.2293423271500843,
+      "grad_norm": 0.21024355292320251,
+      "learning_rate": 0.0001157207207207207,
+      "loss": 0.3494,
+      "step": 4101
+    },
+    {
+      "epoch": 1.2296421210417838,
+      "grad_norm": 0.22852125763893127,
+      "learning_rate": 0.00011567567567567567,
+      "loss": 0.3852,
+      "step": 4102
+    },
+    {
+      "epoch": 1.2299419149334831,
+      "grad_norm": 0.22062285244464874,
+      "learning_rate": 0.00011563063063063063,
+      "loss": 0.3512,
+      "step": 4103
+    },
+    {
+      "epoch": 1.2302417088251827,
+      "grad_norm": 0.21184411644935608,
+      "learning_rate": 0.00011558558558558557,
+      "loss": 0.359,
+      "step": 4104
+    },
+    {
+      "epoch": 1.2305415027168822,
+      "grad_norm": 0.21569395065307617,
+      "learning_rate": 0.00011554054054054053,
+      "loss": 0.3636,
+      "step": 4105
+    },
+    {
+      "epoch": 1.2308412966085815,
+      "grad_norm": 0.22727784514427185,
+      "learning_rate": 0.0001154954954954955,
+      "loss": 0.3616,
+      "step": 4106
+    },
+    {
+      "epoch": 1.231141090500281,
+      "grad_norm": 0.2448568046092987,
+      "learning_rate": 0.00011545045045045043,
+      "loss": 0.3546,
+      "step": 4107
+    },
+    {
+      "epoch": 1.2314408843919806,
+      "grad_norm": 0.2317483127117157,
+      "learning_rate": 0.0001154054054054054,
+      "loss": 0.3619,
+      "step": 4108
+    },
+    {
+      "epoch": 1.23174067828368,
+      "grad_norm": 0.22012370824813843,
+      "learning_rate": 0.00011536036036036035,
+      "loss": 0.3571,
+      "step": 4109
+    },
+    {
+      "epoch": 1.2320404721753795,
+      "grad_norm": 0.2250751107931137,
+      "learning_rate": 0.00011531531531531531,
+      "loss": 0.3697,
+      "step": 4110
+    },
+    {
+      "epoch": 1.2323402660670788,
+      "grad_norm": 0.2103440761566162,
+      "learning_rate": 0.00011527027027027026,
+      "loss": 0.3539,
+      "step": 4111
+    },
+    {
+      "epoch": 1.2326400599587783,
+      "grad_norm": 0.22503060102462769,
+      "learning_rate": 0.00011522522522522521,
+      "loss": 0.3778,
+      "step": 4112
+    },
+    {
+      "epoch": 1.2329398538504779,
+      "grad_norm": 0.22393330931663513,
+      "learning_rate": 0.00011518018018018018,
+      "loss": 0.3791,
+      "step": 4113
+    },
+    {
+      "epoch": 1.2332396477421772,
+      "grad_norm": 0.21146221458911896,
+      "learning_rate": 0.00011513513513513513,
+      "loss": 0.3353,
+      "step": 4114
+    },
+    {
+      "epoch": 1.2335394416338767,
+      "grad_norm": 0.23022443056106567,
+      "learning_rate": 0.00011509009009009008,
+      "loss": 0.369,
+      "step": 4115
+    },
+    {
+      "epoch": 1.233839235525576,
+      "grad_norm": 0.24283228814601898,
+      "learning_rate": 0.00011504504504504504,
+      "loss": 0.3553,
+      "step": 4116
+    },
+    {
+      "epoch": 1.2341390294172756,
+      "grad_norm": 0.2372848093509674,
+      "learning_rate": 0.000115,
+      "loss": 0.3657,
+      "step": 4117
+    },
+    {
+      "epoch": 1.2344388233089751,
+      "grad_norm": 0.22817353904247284,
+      "learning_rate": 0.00011495495495495494,
+      "loss": 0.3542,
+      "step": 4118
+    },
+    {
+      "epoch": 1.2347386172006745,
+      "grad_norm": 0.21485529839992523,
+      "learning_rate": 0.0001149099099099099,
+      "loss": 0.3315,
+      "step": 4119
+    },
+    {
+      "epoch": 1.235038411092374,
+      "grad_norm": 0.2125268578529358,
+      "learning_rate": 0.00011486486486486484,
+      "loss": 0.367,
+      "step": 4120
+    },
+    {
+      "epoch": 1.2353382049840735,
+      "grad_norm": 0.2177802473306656,
+      "learning_rate": 0.0001148198198198198,
+      "loss": 0.3667,
+      "step": 4121
+    },
+    {
+      "epoch": 1.2356379988757729,
+      "grad_norm": 0.22439855337142944,
+      "learning_rate": 0.00011477477477477477,
+      "loss": 0.3595,
+      "step": 4122
+    },
+    {
+      "epoch": 1.2359377927674724,
+      "grad_norm": 0.22252248227596283,
+      "learning_rate": 0.00011472972972972972,
+      "loss": 0.3689,
+      "step": 4123
+    },
+    {
+      "epoch": 1.236237586659172,
+      "grad_norm": 0.21893781423568726,
+      "learning_rate": 0.00011468468468468467,
+      "loss": 0.3363,
+      "step": 4124
+    },
+    {
+      "epoch": 1.2365373805508713,
+      "grad_norm": 0.20217002928256989,
+      "learning_rate": 0.00011463963963963963,
+      "loss": 0.3522,
+      "step": 4125
+    },
+    {
+      "epoch": 1.2368371744425708,
+      "grad_norm": 0.2127171903848648,
+      "learning_rate": 0.00011459459459459458,
+      "loss": 0.345,
+      "step": 4126
+    },
+    {
+      "epoch": 1.2371369683342701,
+      "grad_norm": 0.2248011827468872,
+      "learning_rate": 0.00011454954954954955,
+      "loss": 0.3326,
+      "step": 4127
+    },
+    {
+      "epoch": 1.2374367622259697,
+      "grad_norm": 0.22507144510746002,
+      "learning_rate": 0.0001145045045045045,
+      "loss": 0.3728,
+      "step": 4128
+    },
+    {
+      "epoch": 1.2377365561176692,
+      "grad_norm": 0.22254809737205505,
+      "learning_rate": 0.00011445945945945945,
+      "loss": 0.3707,
+      "step": 4129
+    },
+    {
+      "epoch": 1.2380363500093685,
+      "grad_norm": 0.2163870483636856,
+      "learning_rate": 0.00011441441441441441,
+      "loss": 0.3686,
+      "step": 4130
+    },
+    {
+      "epoch": 1.238336143901068,
+      "grad_norm": 0.23733799159526825,
+      "learning_rate": 0.00011436936936936935,
+      "loss": 0.3502,
+      "step": 4131
+    },
+    {
+      "epoch": 1.2386359377927674,
+      "grad_norm": 0.22752176225185394,
+      "learning_rate": 0.00011432432432432431,
+      "loss": 0.3817,
+      "step": 4132
+    },
+    {
+      "epoch": 1.238935731684467,
+      "grad_norm": 0.2311381846666336,
+      "learning_rate": 0.00011427927927927928,
+      "loss": 0.3544,
+      "step": 4133
+    },
+    {
+      "epoch": 1.2392355255761665,
+      "grad_norm": 0.21874240040779114,
+      "learning_rate": 0.00011423423423423421,
+      "loss": 0.3665,
+      "step": 4134
+    },
+    {
+      "epoch": 1.2395353194678658,
+      "grad_norm": 0.21653451025485992,
+      "learning_rate": 0.00011418918918918918,
+      "loss": 0.354,
+      "step": 4135
+    },
+    {
+      "epoch": 1.2398351133595653,
+      "grad_norm": 0.21141435205936432,
+      "learning_rate": 0.00011414414414414414,
+      "loss": 0.34,
+      "step": 4136
+    },
+    {
+      "epoch": 1.2401349072512649,
+      "grad_norm": 0.21471178531646729,
+      "learning_rate": 0.00011409909909909908,
+      "loss": 0.343,
+      "step": 4137
+    },
+    {
+      "epoch": 1.2404347011429642,
+      "grad_norm": 0.22427885234355927,
+      "learning_rate": 0.00011405405405405404,
+      "loss": 0.3719,
+      "step": 4138
+    },
+    {
+      "epoch": 1.2407344950346637,
+      "grad_norm": 0.21209987998008728,
+      "learning_rate": 0.000114009009009009,
+      "loss": 0.3724,
+      "step": 4139
+    },
+    {
+      "epoch": 1.2410342889263632,
+      "grad_norm": 0.2354046106338501,
+      "learning_rate": 0.00011396396396396396,
+      "loss": 0.3735,
+      "step": 4140
+    },
+    {
+      "epoch": 1.2413340828180626,
+      "grad_norm": 0.21986010670661926,
+      "learning_rate": 0.0001139189189189189,
+      "loss": 0.3605,
+      "step": 4141
+    },
+    {
+      "epoch": 1.241633876709762,
+      "grad_norm": 0.2356344759464264,
+      "learning_rate": 0.00011387387387387387,
+      "loss": 0.3589,
+      "step": 4142
+    },
+    {
+      "epoch": 1.2419336706014614,
+      "grad_norm": 0.2403137981891632,
+      "learning_rate": 0.00011382882882882882,
+      "loss": 0.342,
+      "step": 4143
+    },
+    {
+      "epoch": 1.242233464493161,
+      "grad_norm": 0.22544018924236298,
+      "learning_rate": 0.00011378378378378378,
+      "loss": 0.3639,
+      "step": 4144
+    },
+    {
+      "epoch": 1.2425332583848605,
+      "grad_norm": 0.21682587265968323,
+      "learning_rate": 0.00011373873873873872,
+      "loss": 0.3466,
+      "step": 4145
+    },
+    {
+      "epoch": 1.2428330522765598,
+      "grad_norm": 0.21896804869174957,
+      "learning_rate": 0.00011369369369369368,
+      "loss": 0.36,
+      "step": 4146
+    },
+    {
+      "epoch": 1.2431328461682594,
+      "grad_norm": 0.2302297204732895,
+      "learning_rate": 0.00011364864864864865,
+      "loss": 0.3799,
+      "step": 4147
+    },
+    {
+      "epoch": 1.2434326400599587,
+      "grad_norm": 0.23035024106502533,
+      "learning_rate": 0.00011360360360360359,
+      "loss": 0.3488,
+      "step": 4148
+    },
+    {
+      "epoch": 1.2437324339516582,
+      "grad_norm": 0.21760596334934235,
+      "learning_rate": 0.00011355855855855855,
+      "loss": 0.3612,
+      "step": 4149
+    },
+    {
+      "epoch": 1.2440322278433578,
+      "grad_norm": 0.22977299988269806,
+      "learning_rate": 0.00011351351351351351,
+      "loss": 0.3621,
+      "step": 4150
+    },
+    {
+      "epoch": 1.244332021735057,
+      "grad_norm": 0.22122150659561157,
+      "learning_rate": 0.00011346846846846845,
+      "loss": 0.3574,
+      "step": 4151
+    },
+    {
+      "epoch": 1.2446318156267566,
+      "grad_norm": 0.21653127670288086,
+      "learning_rate": 0.00011342342342342341,
+      "loss": 0.3341,
+      "step": 4152
+    },
+    {
+      "epoch": 1.244931609518456,
+      "grad_norm": 0.20750166475772858,
+      "learning_rate": 0.00011337837837837838,
+      "loss": 0.3352,
+      "step": 4153
+    },
+    {
+      "epoch": 1.2452314034101555,
+      "grad_norm": 0.2003934681415558,
+      "learning_rate": 0.00011333333333333331,
+      "loss": 0.3224,
+      "step": 4154
+    },
+    {
+      "epoch": 1.245531197301855,
+      "grad_norm": 0.21856017410755157,
+      "learning_rate": 0.00011328828828828828,
+      "loss": 0.351,
+      "step": 4155
+    },
+    {
+      "epoch": 1.2458309911935543,
+      "grad_norm": 0.21926572918891907,
+      "learning_rate": 0.00011324324324324323,
+      "loss": 0.3587,
+      "step": 4156
+    },
+    {
+      "epoch": 1.2461307850852539,
+      "grad_norm": 0.2154683619737625,
+      "learning_rate": 0.00011319819819819819,
+      "loss": 0.3571,
+      "step": 4157
+    },
+    {
+      "epoch": 1.2464305789769534,
+      "grad_norm": 0.20265419781208038,
+      "learning_rate": 0.00011315315315315314,
+      "loss": 0.3338,
+      "step": 4158
+    },
+    {
+      "epoch": 1.2467303728686527,
+      "grad_norm": 0.22431926429271698,
+      "learning_rate": 0.00011310810810810809,
+      "loss": 0.3604,
+      "step": 4159
+    },
+    {
+      "epoch": 1.2470301667603523,
+      "grad_norm": 0.2245953381061554,
+      "learning_rate": 0.00011306306306306306,
+      "loss": 0.3575,
+      "step": 4160
+    },
+    {
+      "epoch": 1.2473299606520518,
+      "grad_norm": 0.2239590287208557,
+      "learning_rate": 0.00011301801801801801,
+      "loss": 0.3487,
+      "step": 4161
+    },
+    {
+      "epoch": 1.2476297545437511,
+      "grad_norm": 0.22053176164627075,
+      "learning_rate": 0.00011297297297297296,
+      "loss": 0.3493,
+      "step": 4162
+    },
+    {
+      "epoch": 1.2479295484354507,
+      "grad_norm": 0.22241485118865967,
+      "learning_rate": 0.00011292792792792792,
+      "loss": 0.3605,
+      "step": 4163
+    },
+    {
+      "epoch": 1.24822934232715,
+      "grad_norm": 0.22865022718906403,
+      "learning_rate": 0.00011288288288288289,
+      "loss": 0.3591,
+      "step": 4164
+    },
+    {
+      "epoch": 1.2485291362188495,
+      "grad_norm": 0.2038031369447708,
+      "learning_rate": 0.00011283783783783782,
+      "loss": 0.3484,
+      "step": 4165
+    },
+    {
+      "epoch": 1.248828930110549,
+      "grad_norm": 0.2149299681186676,
+      "learning_rate": 0.00011279279279279279,
+      "loss": 0.3583,
+      "step": 4166
+    },
+    {
+      "epoch": 1.2491287240022484,
+      "grad_norm": 0.22604189813137054,
+      "learning_rate": 0.00011274774774774774,
+      "loss": 0.375,
+      "step": 4167
+    },
+    {
+      "epoch": 1.249428517893948,
+      "grad_norm": 0.23565004765987396,
+      "learning_rate": 0.00011270270270270269,
+      "loss": 0.3691,
+      "step": 4168
+    },
+    {
+      "epoch": 1.2497283117856473,
+      "grad_norm": 0.21931523084640503,
+      "learning_rate": 0.00011265765765765765,
+      "loss": 0.3667,
+      "step": 4169
+    },
+    {
+      "epoch": 1.2500281056773468,
+      "grad_norm": 0.22204811871051788,
+      "learning_rate": 0.0001126126126126126,
+      "loss": 0.357,
+      "step": 4170
+    },
+    {
+      "epoch": 1.2503278995690463,
+      "grad_norm": 0.21145589649677277,
+      "learning_rate": 0.00011256756756756755,
+      "loss": 0.3445,
+      "step": 4171
+    },
+    {
+      "epoch": 1.2506276934607459,
+      "grad_norm": 0.22172702848911285,
+      "learning_rate": 0.00011252252252252251,
+      "loss": 0.3684,
+      "step": 4172
+    },
+    {
+      "epoch": 1.2509274873524452,
+      "grad_norm": 0.22264516353607178,
+      "learning_rate": 0.00011247747747747747,
+      "loss": 0.3675,
+      "step": 4173
+    },
+    {
+      "epoch": 1.2512272812441447,
+      "grad_norm": 0.2182130068540573,
+      "learning_rate": 0.00011243243243243243,
+      "loss": 0.3688,
+      "step": 4174
+    },
+    {
+      "epoch": 1.251527075135844,
+      "grad_norm": 0.20346961915493011,
+      "learning_rate": 0.00011238738738738738,
+      "loss": 0.3342,
+      "step": 4175
+    },
+    {
+      "epoch": 1.2518268690275436,
+      "grad_norm": 0.2261209338903427,
+      "learning_rate": 0.00011234234234234233,
+      "loss": 0.3894,
+      "step": 4176
+    },
+    {
+      "epoch": 1.2521266629192431,
+      "grad_norm": 0.22320590913295746,
+      "learning_rate": 0.0001122972972972973,
+      "loss": 0.35,
+      "step": 4177
+    },
+    {
+      "epoch": 1.2524264568109424,
+      "grad_norm": 0.22703339159488678,
+      "learning_rate": 0.00011225225225225224,
+      "loss": 0.3718,
+      "step": 4178
+    },
+    {
+      "epoch": 1.252726250702642,
+      "grad_norm": 0.21656812727451324,
+      "learning_rate": 0.0001122072072072072,
+      "loss": 0.3573,
+      "step": 4179
+    },
+    {
+      "epoch": 1.2530260445943413,
+      "grad_norm": 0.21183227002620697,
+      "learning_rate": 0.00011216216216216216,
+      "loss": 0.3774,
+      "step": 4180
+    },
+    {
+      "epoch": 1.2533258384860408,
+      "grad_norm": 0.21337759494781494,
+      "learning_rate": 0.0001121171171171171,
+      "loss": 0.3614,
+      "step": 4181
+    },
+    {
+      "epoch": 1.2536256323777404,
+      "grad_norm": 0.2244948446750641,
+      "learning_rate": 0.00011207207207207206,
+      "loss": 0.3562,
+      "step": 4182
+    },
+    {
+      "epoch": 1.2539254262694397,
+      "grad_norm": 0.20882092416286469,
+      "learning_rate": 0.00011202702702702702,
+      "loss": 0.3456,
+      "step": 4183
+    },
+    {
+      "epoch": 1.2542252201611392,
+      "grad_norm": 0.2338513433933258,
+      "learning_rate": 0.00011198198198198197,
+      "loss": 0.3714,
+      "step": 4184
+    },
+    {
+      "epoch": 1.2545250140528386,
+      "grad_norm": 0.20936037600040436,
+      "learning_rate": 0.00011193693693693692,
+      "loss": 0.3447,
+      "step": 4185
+    },
+    {
+      "epoch": 1.254824807944538,
+      "grad_norm": 0.217252716422081,
+      "learning_rate": 0.00011189189189189189,
+      "loss": 0.3592,
+      "step": 4186
+    },
+    {
+      "epoch": 1.2551246018362376,
+      "grad_norm": 0.22383899986743927,
+      "learning_rate": 0.00011184684684684684,
+      "loss": 0.3537,
+      "step": 4187
+    },
+    {
+      "epoch": 1.2554243957279372,
+      "grad_norm": 0.22584226727485657,
+      "learning_rate": 0.00011180180180180179,
+      "loss": 0.3615,
+      "step": 4188
+    },
+    {
+      "epoch": 1.2557241896196365,
+      "grad_norm": 0.2147902399301529,
+      "learning_rate": 0.00011175675675675675,
+      "loss": 0.3726,
+      "step": 4189
+    },
+    {
+      "epoch": 1.256023983511336,
+      "grad_norm": 0.23081621527671814,
+      "learning_rate": 0.0001117117117117117,
+      "loss": 0.3909,
+      "step": 4190
+    },
+    {
+      "epoch": 1.2563237774030354,
+      "grad_norm": 0.21398404240608215,
+      "learning_rate": 0.00011166666666666667,
+      "loss": 0.3763,
+      "step": 4191
+    },
+    {
+      "epoch": 1.256623571294735,
+      "grad_norm": 0.20779982209205627,
+      "learning_rate": 0.0001116216216216216,
+      "loss": 0.3367,
+      "step": 4192
+    },
+    {
+      "epoch": 1.2569233651864344,
+      "grad_norm": 0.2094753086566925,
+      "learning_rate": 0.00011157657657657657,
+      "loss": 0.3401,
+      "step": 4193
+    },
+    {
+      "epoch": 1.2572231590781338,
+      "grad_norm": 0.21873033046722412,
+      "learning_rate": 0.00011153153153153153,
+      "loss": 0.3735,
+      "step": 4194
+    },
+    {
+      "epoch": 1.2575229529698333,
+      "grad_norm": 0.21792501211166382,
+      "learning_rate": 0.00011148648648648647,
+      "loss": 0.3689,
+      "step": 4195
+    },
+    {
+      "epoch": 1.2578227468615326,
+      "grad_norm": 0.21120500564575195,
+      "learning_rate": 0.00011144144144144143,
+      "loss": 0.3443,
+      "step": 4196
+    },
+    {
+      "epoch": 1.2581225407532322,
+      "grad_norm": 0.22521494328975677,
+      "learning_rate": 0.0001113963963963964,
+      "loss": 0.3584,
+      "step": 4197
+    },
+    {
+      "epoch": 1.2584223346449317,
+      "grad_norm": 0.2140861600637436,
+      "learning_rate": 0.00011135135135135133,
+      "loss": 0.3599,
+      "step": 4198
+    },
+    {
+      "epoch": 1.258722128536631,
+      "grad_norm": 0.21226727962493896,
+      "learning_rate": 0.0001113063063063063,
+      "loss": 0.3673,
+      "step": 4199
+    },
+    {
+      "epoch": 1.2590219224283306,
+      "grad_norm": 0.22516387701034546,
+      "learning_rate": 0.00011126126126126126,
+      "loss": 0.3678,
+      "step": 4200
+    },
+    {
+      "epoch": 1.2593217163200299,
+      "grad_norm": 0.22374635934829712,
+      "learning_rate": 0.0001112162162162162,
+      "loss": 0.3548,
+      "step": 4201
+    },
+    {
+      "epoch": 1.2596215102117294,
+      "grad_norm": 0.2221246212720871,
+      "learning_rate": 0.00011117117117117116,
+      "loss": 0.3745,
+      "step": 4202
+    },
+    {
+      "epoch": 1.259921304103429,
+      "grad_norm": 0.22684535384178162,
+      "learning_rate": 0.00011112612612612611,
+      "loss": 0.3557,
+      "step": 4203
+    },
+    {
+      "epoch": 1.2602210979951283,
+      "grad_norm": 0.22062942385673523,
+      "learning_rate": 0.00011108108108108107,
+      "loss": 0.3814,
+      "step": 4204
+    },
+    {
+      "epoch": 1.2605208918868278,
+      "grad_norm": 0.21318332850933075,
+      "learning_rate": 0.00011103603603603602,
+      "loss": 0.3628,
+      "step": 4205
+    },
+    {
+      "epoch": 1.2608206857785271,
+      "grad_norm": 0.20761235058307648,
+      "learning_rate": 0.00011099099099099097,
+      "loss": 0.3486,
+      "step": 4206
+    },
+    {
+      "epoch": 1.2611204796702267,
+      "grad_norm": 0.2155384123325348,
+      "learning_rate": 0.00011094594594594594,
+      "loss": 0.364,
+      "step": 4207
+    },
+    {
+      "epoch": 1.2614202735619262,
+      "grad_norm": 0.2154616266489029,
+      "learning_rate": 0.0001109009009009009,
+      "loss": 0.3583,
+      "step": 4208
+    },
+    {
+      "epoch": 1.2617200674536257,
+      "grad_norm": 0.22318878769874573,
+      "learning_rate": 0.00011085585585585584,
+      "loss": 0.3677,
+      "step": 4209
+    },
+    {
+      "epoch": 1.262019861345325,
+      "grad_norm": 0.21294383704662323,
+      "learning_rate": 0.0001108108108108108,
+      "loss": 0.3745,
+      "step": 4210
+    },
+    {
+      "epoch": 1.2623196552370246,
+      "grad_norm": 0.2227228581905365,
+      "learning_rate": 0.00011076576576576577,
+      "loss": 0.3825,
+      "step": 4211
+    },
+    {
+      "epoch": 1.262619449128724,
+      "grad_norm": 0.21078257262706757,
+      "learning_rate": 0.0001107207207207207,
+      "loss": 0.3403,
+      "step": 4212
+    },
+    {
+      "epoch": 1.2629192430204235,
+      "grad_norm": 0.23595395684242249,
+      "learning_rate": 0.00011067567567567567,
+      "loss": 0.3718,
+      "step": 4213
+    },
+    {
+      "epoch": 1.263219036912123,
+      "grad_norm": 0.22779153287410736,
+      "learning_rate": 0.00011063063063063063,
+      "loss": 0.3546,
+      "step": 4214
+    },
+    {
+      "epoch": 1.2635188308038223,
+      "grad_norm": 0.21909108757972717,
+      "learning_rate": 0.00011058558558558557,
+      "loss": 0.3701,
+      "step": 4215
+    },
+    {
+      "epoch": 1.2638186246955219,
+      "grad_norm": 0.22383159399032593,
+      "learning_rate": 0.00011054054054054053,
+      "loss": 0.3735,
+      "step": 4216
+    },
+    {
+      "epoch": 1.2641184185872212,
+      "grad_norm": 0.22993333637714386,
+      "learning_rate": 0.00011049549549549548,
+      "loss": 0.3859,
+      "step": 4217
+    },
+    {
+      "epoch": 1.2644182124789207,
+      "grad_norm": 0.22387270629405975,
+      "learning_rate": 0.00011045045045045043,
+      "loss": 0.3517,
+      "step": 4218
+    },
+    {
+      "epoch": 1.2647180063706203,
+      "grad_norm": 0.22424739599227905,
+      "learning_rate": 0.0001104054054054054,
+      "loss": 0.3758,
+      "step": 4219
+    },
+    {
+      "epoch": 1.2650178002623196,
+      "grad_norm": 0.21295391023159027,
+      "learning_rate": 0.00011036036036036035,
+      "loss": 0.3594,
+      "step": 4220
+    },
+    {
+      "epoch": 1.2653175941540191,
+      "grad_norm": 0.2224559783935547,
+      "learning_rate": 0.00011031531531531531,
+      "loss": 0.368,
+      "step": 4221
+    },
+    {
+      "epoch": 1.2656173880457184,
+      "grad_norm": 0.22176241874694824,
+      "learning_rate": 0.00011027027027027026,
+      "loss": 0.3644,
+      "step": 4222
+    },
+    {
+      "epoch": 1.265917181937418,
+      "grad_norm": 0.2132066935300827,
+      "learning_rate": 0.00011022522522522521,
+      "loss": 0.3545,
+      "step": 4223
+    },
+    {
+      "epoch": 1.2662169758291175,
+      "grad_norm": 0.2125639170408249,
+      "learning_rate": 0.00011018018018018018,
+      "loss": 0.3694,
+      "step": 4224
+    },
+    {
+      "epoch": 1.266516769720817,
+      "grad_norm": 0.21990853548049927,
+      "learning_rate": 0.00011013513513513514,
+      "loss": 0.3772,
+      "step": 4225
+    },
+    {
+      "epoch": 1.2668165636125164,
+      "grad_norm": 0.20371896028518677,
+      "learning_rate": 0.00011009009009009008,
+      "loss": 0.3463,
+      "step": 4226
+    },
+    {
+      "epoch": 1.267116357504216,
+      "grad_norm": 0.21950723230838776,
+      "learning_rate": 0.00011004504504504504,
+      "loss": 0.3609,
+      "step": 4227
+    },
+    {
+      "epoch": 1.2674161513959152,
+      "grad_norm": 0.20541849732398987,
+      "learning_rate": 0.00010999999999999998,
+      "loss": 0.3411,
+      "step": 4228
+    },
+    {
+      "epoch": 1.2677159452876148,
+      "grad_norm": 0.2173074632883072,
+      "learning_rate": 0.00010995495495495494,
+      "loss": 0.3694,
+      "step": 4229
+    },
+    {
+      "epoch": 1.2680157391793143,
+      "grad_norm": 0.21789297461509705,
+      "learning_rate": 0.0001099099099099099,
+      "loss": 0.3732,
+      "step": 4230
+    },
+    {
+      "epoch": 1.2683155330710136,
+      "grad_norm": 0.21867066621780396,
+      "learning_rate": 0.00010986486486486485,
+      "loss": 0.371,
+      "step": 4231
+    },
+    {
+      "epoch": 1.2686153269627132,
+      "grad_norm": 0.2214251011610031,
+      "learning_rate": 0.0001098198198198198,
+      "loss": 0.3645,
+      "step": 4232
+    },
+    {
+      "epoch": 1.2689151208544125,
+      "grad_norm": 0.21950241923332214,
+      "learning_rate": 0.00010977477477477477,
+      "loss": 0.3584,
+      "step": 4233
+    },
+    {
+      "epoch": 1.269214914746112,
+      "grad_norm": 0.21484404802322388,
+      "learning_rate": 0.00010972972972972972,
+      "loss": 0.3451,
+      "step": 4234
+    },
+    {
+      "epoch": 1.2695147086378116,
+      "grad_norm": 0.21472983062267303,
+      "learning_rate": 0.00010968468468468467,
+      "loss": 0.3334,
+      "step": 4235
+    },
+    {
+      "epoch": 1.269814502529511,
+      "grad_norm": 0.224507138133049,
+      "learning_rate": 0.00010963963963963963,
+      "loss": 0.3746,
+      "step": 4236
+    },
+    {
+      "epoch": 1.2701142964212104,
+      "grad_norm": 0.22031918168067932,
+      "learning_rate": 0.00010959459459459458,
+      "loss": 0.3401,
+      "step": 4237
+    },
+    {
+      "epoch": 1.2704140903129097,
+      "grad_norm": 0.22102873027324677,
+      "learning_rate": 0.00010954954954954955,
+      "loss": 0.3778,
+      "step": 4238
+    },
+    {
+      "epoch": 1.2707138842046093,
+      "grad_norm": 0.2392221838235855,
+      "learning_rate": 0.00010950450450450448,
+      "loss": 0.3587,
+      "step": 4239
+    },
+    {
+      "epoch": 1.2710136780963088,
+      "grad_norm": 0.2183891385793686,
+      "learning_rate": 0.00010945945945945945,
+      "loss": 0.3407,
+      "step": 4240
+    },
+    {
+      "epoch": 1.2713134719880084,
+      "grad_norm": 0.22965964674949646,
+      "learning_rate": 0.00010941441441441441,
+      "loss": 0.3641,
+      "step": 4241
+    },
+    {
+      "epoch": 1.2716132658797077,
+      "grad_norm": 0.22997382283210754,
+      "learning_rate": 0.00010936936936936935,
+      "loss": 0.3698,
+      "step": 4242
+    },
+    {
+      "epoch": 1.2719130597714072,
+      "grad_norm": 0.2099880427122116,
+      "learning_rate": 0.00010932432432432431,
+      "loss": 0.3441,
+      "step": 4243
+    },
+    {
+      "epoch": 1.2722128536631065,
+      "grad_norm": 0.2355547547340393,
+      "learning_rate": 0.00010927927927927928,
+      "loss": 0.3905,
+      "step": 4244
+    },
+    {
+      "epoch": 1.272512647554806,
+      "grad_norm": 0.2386329025030136,
+      "learning_rate": 0.00010923423423423421,
+      "loss": 0.3716,
+      "step": 4245
+    },
+    {
+      "epoch": 1.2728124414465056,
+      "grad_norm": 0.21616117656230927,
+      "learning_rate": 0.00010918918918918918,
+      "loss": 0.3542,
+      "step": 4246
+    },
+    {
+      "epoch": 1.273112235338205,
+      "grad_norm": 0.22752898931503296,
+      "learning_rate": 0.00010914414414414414,
+      "loss": 0.3713,
+      "step": 4247
+    },
+    {
+      "epoch": 1.2734120292299045,
+      "grad_norm": 0.21322378516197205,
+      "learning_rate": 0.00010909909909909909,
+      "loss": 0.3541,
+      "step": 4248
+    },
+    {
+      "epoch": 1.2737118231216038,
+      "grad_norm": 0.22594040632247925,
+      "learning_rate": 0.00010905405405405404,
+      "loss": 0.3702,
+      "step": 4249
+    },
+    {
+      "epoch": 1.2740116170133033,
+      "grad_norm": 0.22437477111816406,
+      "learning_rate": 0.000109009009009009,
+      "loss": 0.3849,
+      "step": 4250
+    },
+    {
+      "epoch": 1.2743114109050029,
+      "grad_norm": 0.20456473529338837,
+      "learning_rate": 0.00010896396396396396,
+      "loss": 0.3297,
+      "step": 4251
+    },
+    {
+      "epoch": 1.2746112047967022,
+      "grad_norm": 0.22588905692100525,
+      "learning_rate": 0.0001089189189189189,
+      "loss": 0.3807,
+      "step": 4252
+    },
+    {
+      "epoch": 1.2749109986884017,
+      "grad_norm": 0.2254696935415268,
+      "learning_rate": 0.00010887387387387386,
+      "loss": 0.3501,
+      "step": 4253
+    },
+    {
+      "epoch": 1.275210792580101,
+      "grad_norm": 0.2138838917016983,
+      "learning_rate": 0.00010882882882882882,
+      "loss": 0.3368,
+      "step": 4254
+    },
+    {
+      "epoch": 1.2755105864718006,
+      "grad_norm": 0.221001535654068,
+      "learning_rate": 0.00010878378378378378,
+      "loss": 0.3531,
+      "step": 4255
+    },
+    {
+      "epoch": 1.2758103803635001,
+      "grad_norm": 0.226884663105011,
+      "learning_rate": 0.00010873873873873872,
+      "loss": 0.366,
+      "step": 4256
+    },
+    {
+      "epoch": 1.2761101742551997,
+      "grad_norm": 0.22693143784999847,
+      "learning_rate": 0.00010869369369369369,
+      "loss": 0.3906,
+      "step": 4257
+    },
+    {
+      "epoch": 1.276409968146899,
+      "grad_norm": 0.22758977115154266,
+      "learning_rate": 0.00010864864864864865,
+      "loss": 0.3821,
+      "step": 4258
+    },
+    {
+      "epoch": 1.2767097620385985,
+      "grad_norm": 0.22596494853496552,
+      "learning_rate": 0.00010860360360360359,
+      "loss": 0.3825,
+      "step": 4259
+    },
+    {
+      "epoch": 1.2770095559302979,
+      "grad_norm": 0.22047007083892822,
+      "learning_rate": 0.00010855855855855855,
+      "loss": 0.3412,
+      "step": 4260
+    },
+    {
+      "epoch": 1.2773093498219974,
+      "grad_norm": 0.22519131004810333,
+      "learning_rate": 0.00010851351351351351,
+      "loss": 0.3592,
+      "step": 4261
+    },
+    {
+      "epoch": 1.277609143713697,
+      "grad_norm": 0.22262904047966003,
+      "learning_rate": 0.00010846846846846845,
+      "loss": 0.3766,
+      "step": 4262
+    },
+    {
+      "epoch": 1.2779089376053963,
+      "grad_norm": 0.23117446899414062,
+      "learning_rate": 0.00010842342342342341,
+      "loss": 0.3897,
+      "step": 4263
+    },
+    {
+      "epoch": 1.2782087314970958,
+      "grad_norm": 0.21441631019115448,
+      "learning_rate": 0.00010837837837837836,
+      "loss": 0.3616,
+      "step": 4264
+    },
+    {
+      "epoch": 1.2785085253887951,
+      "grad_norm": 0.21647268533706665,
+      "learning_rate": 0.00010833333333333333,
+      "loss": 0.3571,
+      "step": 4265
+    },
+    {
+      "epoch": 1.2788083192804947,
+      "grad_norm": 0.23482079803943634,
+      "learning_rate": 0.00010828828828828828,
+      "loss": 0.354,
+      "step": 4266
+    },
+    {
+      "epoch": 1.2791081131721942,
+      "grad_norm": 0.20919762551784515,
+      "learning_rate": 0.00010824324324324323,
+      "loss": 0.3666,
+      "step": 4267
+    },
+    {
+      "epoch": 1.2794079070638935,
+      "grad_norm": 0.21219047904014587,
+      "learning_rate": 0.00010819819819819819,
+      "loss": 0.3602,
+      "step": 4268
+    },
+    {
+      "epoch": 1.279707700955593,
+      "grad_norm": 0.279299259185791,
+      "learning_rate": 0.00010815315315315314,
+      "loss": 0.3867,
+      "step": 4269
+    },
+    {
+      "epoch": 1.2800074948472924,
+      "grad_norm": 0.2186315357685089,
+      "learning_rate": 0.0001081081081081081,
+      "loss": 0.3494,
+      "step": 4270
+    },
+    {
+      "epoch": 1.280307288738992,
+      "grad_norm": 0.24019218981266022,
+      "learning_rate": 0.00010806306306306306,
+      "loss": 0.3813,
+      "step": 4271
+    },
+    {
+      "epoch": 1.2806070826306915,
+      "grad_norm": 0.2710072696208954,
+      "learning_rate": 0.00010801801801801802,
+      "loss": 0.4117,
+      "step": 4272
+    },
+    {
+      "epoch": 1.280906876522391,
+      "grad_norm": 0.21920080482959747,
+      "learning_rate": 0.00010797297297297296,
+      "loss": 0.3701,
+      "step": 4273
+    },
+    {
+      "epoch": 1.2812066704140903,
+      "grad_norm": 0.2656455934047699,
+      "learning_rate": 0.00010792792792792792,
+      "loss": 0.3556,
+      "step": 4274
+    },
+    {
+      "epoch": 1.2815064643057898,
+      "grad_norm": 0.24394501745700836,
+      "learning_rate": 0.00010788288288288289,
+      "loss": 0.3978,
+      "step": 4275
+    },
+    {
+      "epoch": 1.2818062581974892,
+      "grad_norm": 0.2048931121826172,
+      "learning_rate": 0.00010783783783783782,
+      "loss": 0.3491,
+      "step": 4276
+    },
+    {
+      "epoch": 1.2821060520891887,
+      "grad_norm": 0.25899815559387207,
+      "learning_rate": 0.00010779279279279279,
+      "loss": 0.3599,
+      "step": 4277
+    },
+    {
+      "epoch": 1.2824058459808882,
+      "grad_norm": 0.2182856947183609,
+      "learning_rate": 0.00010774774774774774,
+      "loss": 0.3547,
+      "step": 4278
+    },
+    {
+      "epoch": 1.2827056398725876,
+      "grad_norm": 0.22078947722911835,
+      "learning_rate": 0.00010770270270270269,
+      "loss": 0.3822,
+      "step": 4279
+    },
+    {
+      "epoch": 1.283005433764287,
+      "grad_norm": 0.21624134480953217,
+      "learning_rate": 0.00010765765765765765,
+      "loss": 0.3594,
+      "step": 4280
+    },
+    {
+      "epoch": 1.2833052276559864,
+      "grad_norm": 0.22964727878570557,
+      "learning_rate": 0.0001076126126126126,
+      "loss": 0.3908,
+      "step": 4281
+    },
+    {
+      "epoch": 1.283605021547686,
+      "grad_norm": 0.21670106053352356,
+      "learning_rate": 0.00010756756756756757,
+      "loss": 0.3864,
+      "step": 4282
+    },
+    {
+      "epoch": 1.2839048154393855,
+      "grad_norm": 0.21967041492462158,
+      "learning_rate": 0.00010752252252252252,
+      "loss": 0.3639,
+      "step": 4283
+    },
+    {
+      "epoch": 1.2842046093310848,
+      "grad_norm": 0.20801186561584473,
+      "learning_rate": 0.00010747747747747747,
+      "loss": 0.3328,
+      "step": 4284
+    },
+    {
+      "epoch": 1.2845044032227844,
+      "grad_norm": 0.21895767748355865,
+      "learning_rate": 0.00010743243243243243,
+      "loss": 0.3734,
+      "step": 4285
+    },
+    {
+      "epoch": 1.2848041971144837,
+      "grad_norm": 0.2117338627576828,
+      "learning_rate": 0.00010738738738738738,
+      "loss": 0.3443,
+      "step": 4286
+    },
+    {
+      "epoch": 1.2851039910061832,
+      "grad_norm": 0.22575978934764862,
+      "learning_rate": 0.00010734234234234233,
+      "loss": 0.3619,
+      "step": 4287
+    },
+    {
+      "epoch": 1.2854037848978828,
+      "grad_norm": 0.21902771294116974,
+      "learning_rate": 0.0001072972972972973,
+      "loss": 0.3545,
+      "step": 4288
+    },
+    {
+      "epoch": 1.285703578789582,
+      "grad_norm": 0.22649185359477997,
+      "learning_rate": 0.00010725225225225223,
+      "loss": 0.3569,
+      "step": 4289
+    },
+    {
+      "epoch": 1.2860033726812816,
+      "grad_norm": 0.2184680551290512,
+      "learning_rate": 0.0001072072072072072,
+      "loss": 0.3868,
+      "step": 4290
+    },
+    {
+      "epoch": 1.286303166572981,
+      "grad_norm": 0.2234402894973755,
+      "learning_rate": 0.00010716216216216216,
+      "loss": 0.3599,
+      "step": 4291
+    },
+    {
+      "epoch": 1.2866029604646805,
+      "grad_norm": 0.2185603380203247,
+      "learning_rate": 0.0001071171171171171,
+      "loss": 0.3288,
+      "step": 4292
+    },
+    {
+      "epoch": 1.28690275435638,
+      "grad_norm": 0.216938316822052,
+      "learning_rate": 0.00010707207207207206,
+      "loss": 0.3657,
+      "step": 4293
+    },
+    {
+      "epoch": 1.2872025482480796,
+      "grad_norm": 0.22174493968486786,
+      "learning_rate": 0.00010702702702702702,
+      "loss": 0.3486,
+      "step": 4294
+    },
+    {
+      "epoch": 1.2875023421397789,
+      "grad_norm": 0.23581112921237946,
+      "learning_rate": 0.00010698198198198197,
+      "loss": 0.3934,
+      "step": 4295
+    },
+    {
+      "epoch": 1.2878021360314784,
+      "grad_norm": 0.2387666404247284,
+      "learning_rate": 0.00010693693693693692,
+      "loss": 0.3561,
+      "step": 4296
+    },
+    {
+      "epoch": 1.2881019299231777,
+      "grad_norm": 0.21351338922977448,
+      "learning_rate": 0.00010689189189189189,
+      "loss": 0.3673,
+      "step": 4297
+    },
+    {
+      "epoch": 1.2884017238148773,
+      "grad_norm": 0.22528427839279175,
+      "learning_rate": 0.00010684684684684684,
+      "loss": 0.3493,
+      "step": 4298
+    },
+    {
+      "epoch": 1.2887015177065768,
+      "grad_norm": 0.22709199786186218,
+      "learning_rate": 0.0001068018018018018,
+      "loss": 0.3645,
+      "step": 4299
+    },
+    {
+      "epoch": 1.2890013115982761,
+      "grad_norm": 0.2260282188653946,
+      "learning_rate": 0.00010675675675675674,
+      "loss": 0.3904,
+      "step": 4300
+    },
+    {
+      "epoch": 1.2893011054899757,
+      "grad_norm": 0.23056663572788239,
+      "learning_rate": 0.0001067117117117117,
+      "loss": 0.3642,
+      "step": 4301
+    },
+    {
+      "epoch": 1.289600899381675,
+      "grad_norm": 0.21464140713214874,
+      "learning_rate": 0.00010666666666666667,
+      "loss": 0.364,
+      "step": 4302
+    },
+    {
+      "epoch": 1.2899006932733745,
+      "grad_norm": 0.20631635189056396,
+      "learning_rate": 0.0001066216216216216,
+      "loss": 0.3576,
+      "step": 4303
+    },
+    {
+      "epoch": 1.290200487165074,
+      "grad_norm": 0.22520391643047333,
+      "learning_rate": 0.00010657657657657657,
+      "loss": 0.3706,
+      "step": 4304
+    },
+    {
+      "epoch": 1.2905002810567734,
+      "grad_norm": 0.2366526871919632,
+      "learning_rate": 0.00010653153153153153,
+      "loss": 0.3667,
+      "step": 4305
+    },
+    {
+      "epoch": 1.290800074948473,
+      "grad_norm": 0.21495036780834198,
+      "learning_rate": 0.00010648648648648647,
+      "loss": 0.3808,
+      "step": 4306
+    },
+    {
+      "epoch": 1.2910998688401722,
+      "grad_norm": 0.2272660732269287,
+      "learning_rate": 0.00010644144144144143,
+      "loss": 0.3683,
+      "step": 4307
+    },
+    {
+      "epoch": 1.2913996627318718,
+      "grad_norm": 0.22838729619979858,
+      "learning_rate": 0.0001063963963963964,
+      "loss": 0.3537,
+      "step": 4308
+    },
+    {
+      "epoch": 1.2916994566235713,
+      "grad_norm": 0.20135217905044556,
+      "learning_rate": 0.00010635135135135133,
+      "loss": 0.3258,
+      "step": 4309
+    },
+    {
+      "epoch": 1.2919992505152709,
+      "grad_norm": 0.22161005437374115,
+      "learning_rate": 0.0001063063063063063,
+      "loss": 0.3766,
+      "step": 4310
+    },
+    {
+      "epoch": 1.2922990444069702,
+      "grad_norm": 0.2209654301404953,
+      "learning_rate": 0.00010626126126126126,
+      "loss": 0.3558,
+      "step": 4311
+    },
+    {
+      "epoch": 1.2925988382986697,
+      "grad_norm": 0.21990704536437988,
+      "learning_rate": 0.00010621621621621621,
+      "loss": 0.3503,
+      "step": 4312
+    },
+    {
+      "epoch": 1.292898632190369,
+      "grad_norm": 0.25278615951538086,
+      "learning_rate": 0.00010617117117117116,
+      "loss": 0.3738,
+      "step": 4313
+    },
+    {
+      "epoch": 1.2931984260820686,
+      "grad_norm": 0.22431500256061554,
+      "learning_rate": 0.00010612612612612611,
+      "loss": 0.3625,
+      "step": 4314
+    },
+    {
+      "epoch": 1.2934982199737681,
+      "grad_norm": 0.20802560448646545,
+      "learning_rate": 0.00010608108108108107,
+      "loss": 0.3367,
+      "step": 4315
+    },
+    {
+      "epoch": 1.2937980138654674,
+      "grad_norm": 0.248360276222229,
+      "learning_rate": 0.00010603603603603603,
+      "loss": 0.3693,
+      "step": 4316
+    },
+    {
+      "epoch": 1.294097807757167,
+      "grad_norm": 0.21649615466594696,
+      "learning_rate": 0.00010599099099099098,
+      "loss": 0.3542,
+      "step": 4317
+    },
+    {
+      "epoch": 1.2943976016488663,
+      "grad_norm": 0.20109300315380096,
+      "learning_rate": 0.00010594594594594594,
+      "loss": 0.3314,
+      "step": 4318
+    },
+    {
+      "epoch": 1.2946973955405658,
+      "grad_norm": 0.2151678502559662,
+      "learning_rate": 0.0001059009009009009,
+      "loss": 0.344,
+      "step": 4319
+    },
+    {
+      "epoch": 1.2949971894322654,
+      "grad_norm": 0.2119542509317398,
+      "learning_rate": 0.00010585585585585584,
+      "loss": 0.341,
+      "step": 4320
+    },
+    {
+      "epoch": 1.2952969833239647,
+      "grad_norm": 0.2245444804430008,
+      "learning_rate": 0.0001058108108108108,
+      "loss": 0.364,
+      "step": 4321
+    },
+    {
+      "epoch": 1.2955967772156642,
+      "grad_norm": 0.20092026889324188,
+      "learning_rate": 0.00010576576576576577,
+      "loss": 0.328,
+      "step": 4322
+    },
+    {
+      "epoch": 1.2958965711073636,
+      "grad_norm": 0.22229532897472382,
+      "learning_rate": 0.0001057207207207207,
+      "loss": 0.3758,
+      "step": 4323
+    },
+    {
+      "epoch": 1.296196364999063,
+      "grad_norm": 0.21315178275108337,
+      "learning_rate": 0.00010567567567567567,
+      "loss": 0.3523,
+      "step": 4324
+    },
+    {
+      "epoch": 1.2964961588907626,
+      "grad_norm": 0.22185546159744263,
+      "learning_rate": 0.00010563063063063062,
+      "loss": 0.3764,
+      "step": 4325
+    },
+    {
+      "epoch": 1.2967959527824622,
+      "grad_norm": 0.20400555431842804,
+      "learning_rate": 0.00010558558558558557,
+      "loss": 0.3061,
+      "step": 4326
+    },
+    {
+      "epoch": 1.2970957466741615,
+      "grad_norm": 0.22479741275310516,
+      "learning_rate": 0.00010554054054054053,
+      "loss": 0.3716,
+      "step": 4327
+    },
+    {
+      "epoch": 1.297395540565861,
+      "grad_norm": 0.2396492063999176,
+      "learning_rate": 0.00010549549549549548,
+      "loss": 0.3491,
+      "step": 4328
+    },
+    {
+      "epoch": 1.2976953344575604,
+      "grad_norm": 0.22863778471946716,
+      "learning_rate": 0.00010545045045045045,
+      "loss": 0.3699,
+      "step": 4329
+    },
+    {
+      "epoch": 1.29799512834926,
+      "grad_norm": 0.21430253982543945,
+      "learning_rate": 0.0001054054054054054,
+      "loss": 0.3682,
+      "step": 4330
+    },
+    {
+      "epoch": 1.2982949222409594,
+      "grad_norm": 0.23227596282958984,
+      "learning_rate": 0.00010536036036036035,
+      "loss": 0.3621,
+      "step": 4331
+    },
+    {
+      "epoch": 1.2985947161326588,
+      "grad_norm": 0.22787782549858093,
+      "learning_rate": 0.00010531531531531531,
+      "loss": 0.3599,
+      "step": 4332
+    },
+    {
+      "epoch": 1.2988945100243583,
+      "grad_norm": 0.21019242703914642,
+      "learning_rate": 0.00010527027027027026,
+      "loss": 0.3426,
+      "step": 4333
+    },
+    {
+      "epoch": 1.2991943039160576,
+      "grad_norm": 0.22081467509269714,
+      "learning_rate": 0.00010522522522522521,
+      "loss": 0.3589,
+      "step": 4334
+    },
+    {
+      "epoch": 1.2994940978077572,
+      "grad_norm": 0.21910566091537476,
+      "learning_rate": 0.00010518018018018018,
+      "loss": 0.3415,
+      "step": 4335
+    },
+    {
+      "epoch": 1.2997938916994567,
+      "grad_norm": 0.22049827873706818,
+      "learning_rate": 0.00010513513513513511,
+      "loss": 0.3602,
+      "step": 4336
+    },
+    {
+      "epoch": 1.300093685591156,
+      "grad_norm": 0.21187494695186615,
+      "learning_rate": 0.00010509009009009008,
+      "loss": 0.3446,
+      "step": 4337
+    },
+    {
+      "epoch": 1.3003934794828556,
+      "grad_norm": 0.22202420234680176,
+      "learning_rate": 0.00010504504504504504,
+      "loss": 0.3918,
+      "step": 4338
+    },
+    {
+      "epoch": 1.3006932733745549,
+      "grad_norm": 0.21001745760440826,
+      "learning_rate": 0.00010499999999999999,
+      "loss": 0.3631,
+      "step": 4339
+    },
+    {
+      "epoch": 1.3009930672662544,
+      "grad_norm": 0.21252772212028503,
+      "learning_rate": 0.00010495495495495494,
+      "loss": 0.3491,
+      "step": 4340
+    },
+    {
+      "epoch": 1.301292861157954,
+      "grad_norm": 0.22069324553012848,
+      "learning_rate": 0.0001049099099099099,
+      "loss": 0.3696,
+      "step": 4341
+    },
+    {
+      "epoch": 1.3015926550496535,
+      "grad_norm": 0.221624955534935,
+      "learning_rate": 0.00010486486486486486,
+      "loss": 0.3736,
+      "step": 4342
+    },
+    {
+      "epoch": 1.3018924489413528,
+      "grad_norm": 0.22644497454166412,
+      "learning_rate": 0.0001048198198198198,
+      "loss": 0.3855,
+      "step": 4343
+    },
+    {
+      "epoch": 1.3021922428330523,
+      "grad_norm": 0.21738597750663757,
+      "learning_rate": 0.00010477477477477477,
+      "loss": 0.3628,
+      "step": 4344
+    },
+    {
+      "epoch": 1.3024920367247517,
+      "grad_norm": 0.23884043097496033,
+      "learning_rate": 0.00010472972972972972,
+      "loss": 0.3596,
+      "step": 4345
+    },
+    {
+      "epoch": 1.3027918306164512,
+      "grad_norm": 0.21880125999450684,
+      "learning_rate": 0.00010468468468468468,
+      "loss": 0.3749,
+      "step": 4346
+    },
+    {
+      "epoch": 1.3030916245081507,
+      "grad_norm": 0.23063544929027557,
+      "learning_rate": 0.00010463963963963963,
+      "loss": 0.3672,
+      "step": 4347
+    },
+    {
+      "epoch": 1.30339141839985,
+      "grad_norm": 0.21442480385303497,
+      "learning_rate": 0.00010459459459459458,
+      "loss": 0.3772,
+      "step": 4348
+    },
+    {
+      "epoch": 1.3036912122915496,
+      "grad_norm": 0.21598872542381287,
+      "learning_rate": 0.00010454954954954955,
+      "loss": 0.3814,
+      "step": 4349
+    },
+    {
+      "epoch": 1.303991006183249,
+      "grad_norm": 0.21835830807685852,
+      "learning_rate": 0.00010450450450450448,
+      "loss": 0.3425,
+      "step": 4350
+    },
+    {
+      "epoch": 1.3042908000749485,
+      "grad_norm": 0.22444991767406464,
+      "learning_rate": 0.00010445945945945945,
+      "loss": 0.3784,
+      "step": 4351
+    },
+    {
+      "epoch": 1.304590593966648,
+      "grad_norm": 0.2318716198205948,
+      "learning_rate": 0.00010441441441441441,
+      "loss": 0.3706,
+      "step": 4352
+    },
+    {
+      "epoch": 1.3048903878583473,
+      "grad_norm": 0.21502730250358582,
+      "learning_rate": 0.00010436936936936935,
+      "loss": 0.3497,
+      "step": 4353
+    },
+    {
+      "epoch": 1.3051901817500469,
+      "grad_norm": 0.213004007935524,
+      "learning_rate": 0.00010432432432432431,
+      "loss": 0.3552,
+      "step": 4354
+    },
+    {
+      "epoch": 1.3054899756417462,
+      "grad_norm": 0.22040680050849915,
+      "learning_rate": 0.00010427927927927928,
+      "loss": 0.3719,
+      "step": 4355
+    },
+    {
+      "epoch": 1.3057897695334457,
+      "grad_norm": 0.22093693912029266,
+      "learning_rate": 0.00010423423423423421,
+      "loss": 0.349,
+      "step": 4356
+    },
+    {
+      "epoch": 1.3060895634251453,
+      "grad_norm": 0.21267752349376678,
+      "learning_rate": 0.00010418918918918918,
+      "loss": 0.3441,
+      "step": 4357
+    },
+    {
+      "epoch": 1.3063893573168448,
+      "grad_norm": 0.21792468428611755,
+      "learning_rate": 0.00010414414414414414,
+      "loss": 0.3587,
+      "step": 4358
+    },
+    {
+      "epoch": 1.3066891512085441,
+      "grad_norm": 0.2234465628862381,
+      "learning_rate": 0.00010409909909909909,
+      "loss": 0.3594,
+      "step": 4359
+    },
+    {
+      "epoch": 1.3069889451002437,
+      "grad_norm": 0.22853820025920868,
+      "learning_rate": 0.00010405405405405404,
+      "loss": 0.396,
+      "step": 4360
+    },
+    {
+      "epoch": 1.307288738991943,
+      "grad_norm": 0.21059495210647583,
+      "learning_rate": 0.00010400900900900899,
+      "loss": 0.3778,
+      "step": 4361
+    },
+    {
+      "epoch": 1.3075885328836425,
+      "grad_norm": 0.23244240880012512,
+      "learning_rate": 0.00010396396396396396,
+      "loss": 0.3898,
+      "step": 4362
+    },
+    {
+      "epoch": 1.307888326775342,
+      "grad_norm": 0.21645265817642212,
+      "learning_rate": 0.00010391891891891892,
+      "loss": 0.3388,
+      "step": 4363
+    },
+    {
+      "epoch": 1.3081881206670414,
+      "grad_norm": 0.22847887873649597,
+      "learning_rate": 0.00010387387387387386,
+      "loss": 0.37,
+      "step": 4364
+    },
+    {
+      "epoch": 1.308487914558741,
+      "grad_norm": 0.20887666940689087,
+      "learning_rate": 0.00010382882882882882,
+      "loss": 0.3496,
+      "step": 4365
+    },
+    {
+      "epoch": 1.3087877084504402,
+      "grad_norm": 0.21099483966827393,
+      "learning_rate": 0.00010378378378378378,
+      "loss": 0.362,
+      "step": 4366
+    },
+    {
+      "epoch": 1.3090875023421398,
+      "grad_norm": 0.2116171419620514,
+      "learning_rate": 0.00010373873873873872,
+      "loss": 0.3456,
+      "step": 4367
+    },
+    {
+      "epoch": 1.3093872962338393,
+      "grad_norm": 0.19885142147541046,
+      "learning_rate": 0.00010369369369369369,
+      "loss": 0.3329,
+      "step": 4368
+    },
+    {
+      "epoch": 1.3096870901255386,
+      "grad_norm": 0.21213248372077942,
+      "learning_rate": 0.00010364864864864865,
+      "loss": 0.3403,
+      "step": 4369
+    },
+    {
+      "epoch": 1.3099868840172382,
+      "grad_norm": 0.2295825481414795,
+      "learning_rate": 0.00010360360360360359,
+      "loss": 0.3823,
+      "step": 4370
+    },
+    {
+      "epoch": 1.3102866779089375,
+      "grad_norm": 0.2063182145357132,
+      "learning_rate": 0.00010355855855855855,
+      "loss": 0.3421,
+      "step": 4371
+    },
+    {
+      "epoch": 1.310586471800637,
+      "grad_norm": 0.20554274320602417,
+      "learning_rate": 0.0001035135135135135,
+      "loss": 0.3638,
+      "step": 4372
+    },
+    {
+      "epoch": 1.3108862656923366,
+      "grad_norm": 0.21071697771549225,
+      "learning_rate": 0.00010346846846846845,
+      "loss": 0.3582,
+      "step": 4373
+    },
+    {
+      "epoch": 1.3111860595840361,
+      "grad_norm": 0.21579106152057648,
+      "learning_rate": 0.00010342342342342341,
+      "loss": 0.3512,
+      "step": 4374
+    },
+    {
+      "epoch": 1.3114858534757354,
+      "grad_norm": 0.21299222111701965,
+      "learning_rate": 0.00010337837837837836,
+      "loss": 0.3424,
+      "step": 4375
+    },
+    {
+      "epoch": 1.311785647367435,
+      "grad_norm": 0.21030524373054504,
+      "learning_rate": 0.00010333333333333333,
+      "loss": 0.3581,
+      "step": 4376
+    },
+    {
+      "epoch": 1.3120854412591343,
+      "grad_norm": 0.21630938351154327,
+      "learning_rate": 0.00010328828828828828,
+      "loss": 0.3747,
+      "step": 4377
+    },
+    {
+      "epoch": 1.3123852351508338,
+      "grad_norm": 0.22339695692062378,
+      "learning_rate": 0.00010324324324324323,
+      "loss": 0.3833,
+      "step": 4378
+    },
+    {
+      "epoch": 1.3126850290425334,
+      "grad_norm": 0.21773970127105713,
+      "learning_rate": 0.0001031981981981982,
+      "loss": 0.3729,
+      "step": 4379
+    },
+    {
+      "epoch": 1.3129848229342327,
+      "grad_norm": 0.2207316905260086,
+      "learning_rate": 0.00010315315315315316,
+      "loss": 0.3553,
+      "step": 4380
+    },
+    {
+      "epoch": 1.3132846168259322,
+      "grad_norm": 0.21742987632751465,
+      "learning_rate": 0.0001031081081081081,
+      "loss": 0.3674,
+      "step": 4381
+    },
+    {
+      "epoch": 1.3135844107176315,
+      "grad_norm": 0.21643692255020142,
+      "learning_rate": 0.00010306306306306306,
+      "loss": 0.3557,
+      "step": 4382
+    },
+    {
+      "epoch": 1.313884204609331,
+      "grad_norm": 0.2156946212053299,
+      "learning_rate": 0.00010301801801801802,
+      "loss": 0.3688,
+      "step": 4383
+    },
+    {
+      "epoch": 1.3141839985010306,
+      "grad_norm": 0.21070463955402374,
+      "learning_rate": 0.00010297297297297296,
+      "loss": 0.3524,
+      "step": 4384
+    },
+    {
+      "epoch": 1.31448379239273,
+      "grad_norm": 0.21057476103305817,
+      "learning_rate": 0.00010292792792792792,
+      "loss": 0.3716,
+      "step": 4385
+    },
+    {
+      "epoch": 1.3147835862844295,
+      "grad_norm": 0.2108742743730545,
+      "learning_rate": 0.00010288288288288287,
+      "loss": 0.3596,
+      "step": 4386
+    },
+    {
+      "epoch": 1.3150833801761288,
+      "grad_norm": 0.22613100707530975,
+      "learning_rate": 0.00010283783783783782,
+      "loss": 0.3755,
+      "step": 4387
+    },
+    {
+      "epoch": 1.3153831740678283,
+      "grad_norm": 0.21166808903217316,
+      "learning_rate": 0.00010279279279279279,
+      "loss": 0.3528,
+      "step": 4388
+    },
+    {
+      "epoch": 1.3156829679595279,
+      "grad_norm": 0.22526302933692932,
+      "learning_rate": 0.00010274774774774774,
+      "loss": 0.3857,
+      "step": 4389
+    },
+    {
+      "epoch": 1.3159827618512272,
+      "grad_norm": 0.22209607064723969,
+      "learning_rate": 0.00010270270270270269,
+      "loss": 0.3613,
+      "step": 4390
+    },
+    {
+      "epoch": 1.3162825557429267,
+      "grad_norm": 0.22119589149951935,
+      "learning_rate": 0.00010265765765765765,
+      "loss": 0.3781,
+      "step": 4391
+    },
+    {
+      "epoch": 1.316582349634626,
+      "grad_norm": 0.2123972475528717,
+      "learning_rate": 0.0001026126126126126,
+      "loss": 0.3425,
+      "step": 4392
+    },
+    {
+      "epoch": 1.3168821435263256,
+      "grad_norm": 0.22913959622383118,
+      "learning_rate": 0.00010256756756756757,
+      "loss": 0.3951,
+      "step": 4393
+    },
+    {
+      "epoch": 1.3171819374180251,
+      "grad_norm": 0.2116348147392273,
+      "learning_rate": 0.00010252252252252252,
+      "loss": 0.3619,
+      "step": 4394
+    },
+    {
+      "epoch": 1.3174817313097247,
+      "grad_norm": 0.2163851112127304,
+      "learning_rate": 0.00010247747747747747,
+      "loss": 0.3378,
+      "step": 4395
+    },
+    {
+      "epoch": 1.317781525201424,
+      "grad_norm": 0.22200891375541687,
+      "learning_rate": 0.00010243243243243243,
+      "loss": 0.3685,
+      "step": 4396
+    },
+    {
+      "epoch": 1.3180813190931235,
+      "grad_norm": 0.2236941158771515,
+      "learning_rate": 0.00010238738738738737,
+      "loss": 0.3597,
+      "step": 4397
+    },
+    {
+      "epoch": 1.3183811129848229,
+      "grad_norm": 0.20215103030204773,
+      "learning_rate": 0.00010234234234234233,
+      "loss": 0.3534,
+      "step": 4398
+    },
+    {
+      "epoch": 1.3186809068765224,
+      "grad_norm": 0.2127383053302765,
+      "learning_rate": 0.0001022972972972973,
+      "loss": 0.3462,
+      "step": 4399
+    },
+    {
+      "epoch": 1.318980700768222,
+      "grad_norm": 0.21308203041553497,
+      "learning_rate": 0.00010225225225225223,
+      "loss": 0.3411,
+      "step": 4400
+    },
+    {
+      "epoch": 1.3192804946599213,
+      "grad_norm": 0.21060781180858612,
+      "learning_rate": 0.0001022072072072072,
+      "loss": 0.3585,
+      "step": 4401
+    },
+    {
+      "epoch": 1.3195802885516208,
+      "grad_norm": 0.2206219881772995,
+      "learning_rate": 0.00010216216216216216,
+      "loss": 0.3643,
+      "step": 4402
+    },
+    {
+      "epoch": 1.3198800824433201,
+      "grad_norm": 0.2104436755180359,
+      "learning_rate": 0.00010211711711711711,
+      "loss": 0.3727,
+      "step": 4403
+    },
+    {
+      "epoch": 1.3201798763350197,
+      "grad_norm": 0.21713684499263763,
+      "learning_rate": 0.00010207207207207206,
+      "loss": 0.3579,
+      "step": 4404
+    },
+    {
+      "epoch": 1.3204796702267192,
+      "grad_norm": 0.19761869311332703,
+      "learning_rate": 0.00010202702702702702,
+      "loss": 0.3367,
+      "step": 4405
+    },
+    {
+      "epoch": 1.3207794641184185,
+      "grad_norm": 0.21974486112594604,
+      "learning_rate": 0.00010198198198198197,
+      "loss": 0.3823,
+      "step": 4406
+    },
+    {
+      "epoch": 1.321079258010118,
+      "grad_norm": 0.21158014237880707,
+      "learning_rate": 0.00010193693693693692,
+      "loss": 0.3702,
+      "step": 4407
+    },
+    {
+      "epoch": 1.3213790519018174,
+      "grad_norm": 0.21842747926712036,
+      "learning_rate": 0.00010189189189189187,
+      "loss": 0.3489,
+      "step": 4408
+    },
+    {
+      "epoch": 1.321678845793517,
+      "grad_norm": 0.22330352663993835,
+      "learning_rate": 0.00010184684684684684,
+      "loss": 0.3434,
+      "step": 4409
+    },
+    {
+      "epoch": 1.3219786396852165,
+      "grad_norm": 0.20083275437355042,
+      "learning_rate": 0.0001018018018018018,
+      "loss": 0.3297,
+      "step": 4410
+    },
+    {
+      "epoch": 1.322278433576916,
+      "grad_norm": 0.2106667160987854,
+      "learning_rate": 0.00010175675675675674,
+      "loss": 0.3613,
+      "step": 4411
+    },
+    {
+      "epoch": 1.3225782274686153,
+      "grad_norm": 0.21465076506137848,
+      "learning_rate": 0.0001017117117117117,
+      "loss": 0.357,
+      "step": 4412
+    },
+    {
+      "epoch": 1.3228780213603148,
+      "grad_norm": 0.21720650792121887,
+      "learning_rate": 0.00010166666666666667,
+      "loss": 0.3498,
+      "step": 4413
+    },
+    {
+      "epoch": 1.3231778152520142,
+      "grad_norm": 0.20751236379146576,
+      "learning_rate": 0.0001016216216216216,
+      "loss": 0.3534,
+      "step": 4414
+    },
+    {
+      "epoch": 1.3234776091437137,
+      "grad_norm": 0.21749740839004517,
+      "learning_rate": 0.00010157657657657657,
+      "loss": 0.3816,
+      "step": 4415
+    },
+    {
+      "epoch": 1.3237774030354132,
+      "grad_norm": 0.21746374666690826,
+      "learning_rate": 0.00010153153153153153,
+      "loss": 0.3583,
+      "step": 4416
+    },
+    {
+      "epoch": 1.3240771969271126,
+      "grad_norm": 0.21947112679481506,
+      "learning_rate": 0.00010148648648648647,
+      "loss": 0.3692,
+      "step": 4417
+    },
+    {
+      "epoch": 1.324376990818812,
+      "grad_norm": 0.2228492647409439,
+      "learning_rate": 0.00010144144144144143,
+      "loss": 0.3693,
+      "step": 4418
+    },
+    {
+      "epoch": 1.3246767847105114,
+      "grad_norm": 0.21328027546405792,
+      "learning_rate": 0.0001013963963963964,
+      "loss": 0.3658,
+      "step": 4419
+    },
+    {
+      "epoch": 1.324976578602211,
+      "grad_norm": 0.22100898623466492,
+      "learning_rate": 0.00010135135135135135,
+      "loss": 0.3682,
+      "step": 4420
+    },
+    {
+      "epoch": 1.3252763724939105,
+      "grad_norm": 0.23616699874401093,
+      "learning_rate": 0.0001013063063063063,
+      "loss": 0.3875,
+      "step": 4421
+    },
+    {
+      "epoch": 1.3255761663856098,
+      "grad_norm": 0.2213396579027176,
+      "learning_rate": 0.00010126126126126125,
+      "loss": 0.3697,
+      "step": 4422
+    },
+    {
+      "epoch": 1.3258759602773094,
+      "grad_norm": 0.21712253987789154,
+      "learning_rate": 0.00010121621621621621,
+      "loss": 0.3588,
+      "step": 4423
+    },
+    {
+      "epoch": 1.3261757541690087,
+      "grad_norm": 0.21829205751419067,
+      "learning_rate": 0.00010117117117117116,
+      "loss": 0.3518,
+      "step": 4424
+    },
+    {
+      "epoch": 1.3264755480607082,
+      "grad_norm": 0.21578383445739746,
+      "learning_rate": 0.00010112612612612611,
+      "loss": 0.3846,
+      "step": 4425
+    },
+    {
+      "epoch": 1.3267753419524078,
+      "grad_norm": 0.21372157335281372,
+      "learning_rate": 0.00010108108108108108,
+      "loss": 0.3482,
+      "step": 4426
+    },
+    {
+      "epoch": 1.3270751358441073,
+      "grad_norm": 0.22385689616203308,
+      "learning_rate": 0.00010103603603603604,
+      "loss": 0.3743,
+      "step": 4427
+    },
+    {
+      "epoch": 1.3273749297358066,
+      "grad_norm": 0.22532759606838226,
+      "learning_rate": 0.00010099099099099098,
+      "loss": 0.3685,
+      "step": 4428
+    },
+    {
+      "epoch": 1.3276747236275062,
+      "grad_norm": 0.22591564059257507,
+      "learning_rate": 0.00010094594594594594,
+      "loss": 0.3879,
+      "step": 4429
+    },
+    {
+      "epoch": 1.3279745175192055,
+      "grad_norm": 0.2188117355108261,
+      "learning_rate": 0.0001009009009009009,
+      "loss": 0.361,
+      "step": 4430
+    },
+    {
+      "epoch": 1.328274311410905,
+      "grad_norm": 0.21723529696464539,
+      "learning_rate": 0.00010085585585585584,
+      "loss": 0.3612,
+      "step": 4431
+    },
+    {
+      "epoch": 1.3285741053026046,
+      "grad_norm": 0.21275779604911804,
+      "learning_rate": 0.0001008108108108108,
+      "loss": 0.3656,
+      "step": 4432
+    },
+    {
+      "epoch": 1.3288738991943039,
+      "grad_norm": 0.2086101621389389,
+      "learning_rate": 0.00010076576576576575,
+      "loss": 0.3599,
+      "step": 4433
+    },
+    {
+      "epoch": 1.3291736930860034,
+      "grad_norm": 0.21776525676250458,
+      "learning_rate": 0.0001007207207207207,
+      "loss": 0.3638,
+      "step": 4434
+    },
+    {
+      "epoch": 1.3294734869777027,
+      "grad_norm": 0.21323978900909424,
+      "learning_rate": 0.00010067567567567567,
+      "loss": 0.3695,
+      "step": 4435
+    },
+    {
+      "epoch": 1.3297732808694023,
+      "grad_norm": 0.2201107293367386,
+      "learning_rate": 0.00010063063063063062,
+      "loss": 0.3821,
+      "step": 4436
+    },
+    {
+      "epoch": 1.3300730747611018,
+      "grad_norm": 0.22386541962623596,
+      "learning_rate": 0.00010058558558558558,
+      "loss": 0.3608,
+      "step": 4437
+    },
+    {
+      "epoch": 1.3303728686528011,
+      "grad_norm": 0.2155165672302246,
+      "learning_rate": 0.00010054054054054053,
+      "loss": 0.3562,
+      "step": 4438
+    },
+    {
+      "epoch": 1.3306726625445007,
+      "grad_norm": 0.21298325061798096,
+      "learning_rate": 0.00010049549549549548,
+      "loss": 0.3521,
+      "step": 4439
+    },
+    {
+      "epoch": 1.3309724564362,
+      "grad_norm": 0.2154662311077118,
+      "learning_rate": 0.00010045045045045045,
+      "loss": 0.3765,
+      "step": 4440
+    },
+    {
+      "epoch": 1.3312722503278995,
+      "grad_norm": 0.20846322178840637,
+      "learning_rate": 0.0001004054054054054,
+      "loss": 0.3715,
+      "step": 4441
+    },
+    {
+      "epoch": 1.331572044219599,
+      "grad_norm": 0.21915653347969055,
+      "learning_rate": 0.00010036036036036035,
+      "loss": 0.3246,
+      "step": 4442
+    },
+    {
+      "epoch": 1.3318718381112986,
+      "grad_norm": 0.2159089297056198,
+      "learning_rate": 0.00010031531531531531,
+      "loss": 0.3636,
+      "step": 4443
+    },
+    {
+      "epoch": 1.332171632002998,
+      "grad_norm": 0.22735199332237244,
+      "learning_rate": 0.00010027027027027025,
+      "loss": 0.3687,
+      "step": 4444
+    },
+    {
+      "epoch": 1.3324714258946975,
+      "grad_norm": 0.218096524477005,
+      "learning_rate": 0.00010022522522522521,
+      "loss": 0.3914,
+      "step": 4445
+    },
+    {
+      "epoch": 1.3327712197863968,
+      "grad_norm": 0.21156595647335052,
+      "learning_rate": 0.00010018018018018018,
+      "loss": 0.3915,
+      "step": 4446
+    },
+    {
+      "epoch": 1.3330710136780963,
+      "grad_norm": 0.2183390110731125,
+      "learning_rate": 0.00010013513513513511,
+      "loss": 0.3611,
+      "step": 4447
+    },
+    {
+      "epoch": 1.3333708075697959,
+      "grad_norm": 0.21990258991718292,
+      "learning_rate": 0.00010009009009009008,
+      "loss": 0.3614,
+      "step": 4448
+    },
+    {
+      "epoch": 1.3336706014614952,
+      "grad_norm": 0.21382641792297363,
+      "learning_rate": 0.00010004504504504504,
+      "loss": 0.3541,
+      "step": 4449
+    },
+    {
+      "epoch": 1.3339703953531947,
+      "grad_norm": 0.23284634947776794,
+      "learning_rate": 9.999999999999999e-05,
+      "loss": 0.3694,
+      "step": 4450
+    },
+    {
+      "epoch": 1.334270189244894,
+      "grad_norm": 0.22020602226257324,
+      "learning_rate": 9.995495495495494e-05,
+      "loss": 0.3559,
+      "step": 4451
+    },
+    {
+      "epoch": 1.3345699831365936,
+      "grad_norm": 0.2088702917098999,
+      "learning_rate": 9.99099099099099e-05,
+      "loss": 0.3566,
+      "step": 4452
+    },
+    {
+      "epoch": 1.3348697770282931,
+      "grad_norm": 0.26016759872436523,
+      "learning_rate": 9.986486486486486e-05,
+      "loss": 0.3741,
+      "step": 4453
+    },
+    {
+      "epoch": 1.3351695709199924,
+      "grad_norm": 0.22935625910758972,
+      "learning_rate": 9.981981981981982e-05,
+      "loss": 0.3789,
+      "step": 4454
+    },
+    {
+      "epoch": 1.335469364811692,
+      "grad_norm": 0.2152242511510849,
+      "learning_rate": 9.977477477477477e-05,
+      "loss": 0.3712,
+      "step": 4455
+    },
+    {
+      "epoch": 1.3357691587033913,
+      "grad_norm": 0.2259458303451538,
+      "learning_rate": 9.972972972972972e-05,
+      "loss": 0.3712,
+      "step": 4456
+    },
+    {
+      "epoch": 1.3360689525950908,
+      "grad_norm": 0.21118693053722382,
+      "learning_rate": 9.968468468468468e-05,
+      "loss": 0.3546,
+      "step": 4457
+    },
+    {
+      "epoch": 1.3363687464867904,
+      "grad_norm": 0.21529187262058258,
+      "learning_rate": 9.963963963963962e-05,
+      "loss": 0.3721,
+      "step": 4458
+    },
+    {
+      "epoch": 1.33666854037849,
+      "grad_norm": 0.21422810852527618,
+      "learning_rate": 9.959459459459458e-05,
+      "loss": 0.3645,
+      "step": 4459
+    },
+    {
+      "epoch": 1.3369683342701892,
+      "grad_norm": 0.2009521722793579,
+      "learning_rate": 9.954954954954955e-05,
+      "loss": 0.3561,
+      "step": 4460
+    },
+    {
+      "epoch": 1.3372681281618888,
+      "grad_norm": 0.22312965989112854,
+      "learning_rate": 9.950450450450449e-05,
+      "loss": 0.3546,
+      "step": 4461
+    },
+    {
+      "epoch": 1.337567922053588,
+      "grad_norm": 0.21248088777065277,
+      "learning_rate": 9.945945945945945e-05,
+      "loss": 0.3603,
+      "step": 4462
+    },
+    {
+      "epoch": 1.3378677159452876,
+      "grad_norm": 0.21879053115844727,
+      "learning_rate": 9.941441441441441e-05,
+      "loss": 0.3847,
+      "step": 4463
+    },
+    {
+      "epoch": 1.3381675098369872,
+      "grad_norm": 0.2076069414615631,
+      "learning_rate": 9.936936936936935e-05,
+      "loss": 0.3678,
+      "step": 4464
+    },
+    {
+      "epoch": 1.3384673037286865,
+      "grad_norm": 0.19983349740505219,
+      "learning_rate": 9.932432432432431e-05,
+      "loss": 0.3483,
+      "step": 4465
+    },
+    {
+      "epoch": 1.338767097620386,
+      "grad_norm": 0.20354819297790527,
+      "learning_rate": 9.927927927927928e-05,
+      "loss": 0.3098,
+      "step": 4466
+    },
+    {
+      "epoch": 1.3390668915120854,
+      "grad_norm": 0.2120402306318283,
+      "learning_rate": 9.923423423423423e-05,
+      "loss": 0.3742,
+      "step": 4467
+    },
+    {
+      "epoch": 1.339366685403785,
+      "grad_norm": 0.20712968707084656,
+      "learning_rate": 9.918918918918918e-05,
+      "loss": 0.3682,
+      "step": 4468
+    },
+    {
+      "epoch": 1.3396664792954844,
+      "grad_norm": 0.205884650349617,
+      "learning_rate": 9.914414414414413e-05,
+      "loss": 0.3232,
+      "step": 4469
+    },
+    {
+      "epoch": 1.3399662731871838,
+      "grad_norm": 0.20404751598834991,
+      "learning_rate": 9.909909909909909e-05,
+      "loss": 0.3413,
+      "step": 4470
+    },
+    {
+      "epoch": 1.3402660670788833,
+      "grad_norm": 0.20457707345485687,
+      "learning_rate": 9.905405405405404e-05,
+      "loss": 0.3455,
+      "step": 4471
+    },
+    {
+      "epoch": 1.3405658609705826,
+      "grad_norm": 0.20822665095329285,
+      "learning_rate": 9.900900900900899e-05,
+      "loss": 0.3559,
+      "step": 4472
+    },
+    {
+      "epoch": 1.3408656548622822,
+      "grad_norm": 0.21473756432533264,
+      "learning_rate": 9.896396396396396e-05,
+      "loss": 0.3723,
+      "step": 4473
+    },
+    {
+      "epoch": 1.3411654487539817,
+      "grad_norm": 0.22079592943191528,
+      "learning_rate": 9.891891891891892e-05,
+      "loss": 0.3715,
+      "step": 4474
+    },
+    {
+      "epoch": 1.3414652426456812,
+      "grad_norm": 0.22699683904647827,
+      "learning_rate": 9.887387387387386e-05,
+      "loss": 0.4057,
+      "step": 4475
+    },
+    {
+      "epoch": 1.3417650365373806,
+      "grad_norm": 0.20466183125972748,
+      "learning_rate": 9.882882882882882e-05,
+      "loss": 0.3454,
+      "step": 4476
+    },
+    {
+      "epoch": 1.3420648304290799,
+      "grad_norm": 0.2107648253440857,
+      "learning_rate": 9.878378378378379e-05,
+      "loss": 0.3393,
+      "step": 4477
+    },
+    {
+      "epoch": 1.3423646243207794,
+      "grad_norm": 0.20453116297721863,
+      "learning_rate": 9.873873873873872e-05,
+      "loss": 0.3393,
+      "step": 4478
+    },
+    {
+      "epoch": 1.342664418212479,
+      "grad_norm": 0.21234673261642456,
+      "learning_rate": 9.869369369369369e-05,
+      "loss": 0.3625,
+      "step": 4479
+    },
+    {
+      "epoch": 1.3429642121041785,
+      "grad_norm": 0.2178509384393692,
+      "learning_rate": 9.864864864864865e-05,
+      "loss": 0.351,
+      "step": 4480
+    },
+    {
+      "epoch": 1.3432640059958778,
+      "grad_norm": 0.2194160670042038,
+      "learning_rate": 9.860360360360359e-05,
+      "loss": 0.3498,
+      "step": 4481
+    },
+    {
+      "epoch": 1.3435637998875773,
+      "grad_norm": 0.19866818189620972,
+      "learning_rate": 9.855855855855855e-05,
+      "loss": 0.3126,
+      "step": 4482
+    },
+    {
+      "epoch": 1.3438635937792767,
+      "grad_norm": 0.20963077247142792,
+      "learning_rate": 9.85135135135135e-05,
+      "loss": 0.3505,
+      "step": 4483
+    },
+    {
+      "epoch": 1.3441633876709762,
+      "grad_norm": 0.23428975045681,
+      "learning_rate": 9.846846846846846e-05,
+      "loss": 0.3655,
+      "step": 4484
+    },
+    {
+      "epoch": 1.3444631815626757,
+      "grad_norm": 0.21178551018238068,
+      "learning_rate": 9.842342342342342e-05,
+      "loss": 0.3542,
+      "step": 4485
+    },
+    {
+      "epoch": 1.344762975454375,
+      "grad_norm": 0.20787645876407623,
+      "learning_rate": 9.837837837837837e-05,
+      "loss": 0.3543,
+      "step": 4486
+    },
+    {
+      "epoch": 1.3450627693460746,
+      "grad_norm": 0.2064124494791031,
+      "learning_rate": 9.833333333333333e-05,
+      "loss": 0.3534,
+      "step": 4487
+    },
+    {
+      "epoch": 1.345362563237774,
+      "grad_norm": 0.2198820412158966,
+      "learning_rate": 9.828828828828828e-05,
+      "loss": 0.3539,
+      "step": 4488
+    },
+    {
+      "epoch": 1.3456623571294735,
+      "grad_norm": 0.21668361127376556,
+      "learning_rate": 9.824324324324323e-05,
+      "loss": 0.3707,
+      "step": 4489
+    },
+    {
+      "epoch": 1.345962151021173,
+      "grad_norm": 0.21410995721817017,
+      "learning_rate": 9.81981981981982e-05,
+      "loss": 0.3572,
+      "step": 4490
+    },
+    {
+      "epoch": 1.3462619449128723,
+      "grad_norm": 0.21897783875465393,
+      "learning_rate": 9.815315315315316e-05,
+      "loss": 0.3775,
+      "step": 4491
+    },
+    {
+      "epoch": 1.3465617388045719,
+      "grad_norm": 0.20904091000556946,
+      "learning_rate": 9.81081081081081e-05,
+      "loss": 0.3521,
+      "step": 4492
+    },
+    {
+      "epoch": 1.3468615326962712,
+      "grad_norm": 0.23392531275749207,
+      "learning_rate": 9.806306306306306e-05,
+      "loss": 0.3709,
+      "step": 4493
+    },
+    {
+      "epoch": 1.3471613265879707,
+      "grad_norm": 0.21514087915420532,
+      "learning_rate": 9.801801801801801e-05,
+      "loss": 0.3661,
+      "step": 4494
+    },
+    {
+      "epoch": 1.3474611204796703,
+      "grad_norm": 0.2174890786409378,
+      "learning_rate": 9.797297297297296e-05,
+      "loss": 0.3611,
+      "step": 4495
+    },
+    {
+      "epoch": 1.3477609143713698,
+      "grad_norm": 0.21337510645389557,
+      "learning_rate": 9.792792792792792e-05,
+      "loss": 0.3856,
+      "step": 4496
+    },
+    {
+      "epoch": 1.3480607082630691,
+      "grad_norm": 0.22505298256874084,
+      "learning_rate": 9.788288288288287e-05,
+      "loss": 0.3736,
+      "step": 4497
+    },
+    {
+      "epoch": 1.3483605021547687,
+      "grad_norm": 0.21831883490085602,
+      "learning_rate": 9.783783783783782e-05,
+      "loss": 0.3548,
+      "step": 4498
+    },
+    {
+      "epoch": 1.348660296046468,
+      "grad_norm": 0.2187795341014862,
+      "learning_rate": 9.779279279279279e-05,
+      "loss": 0.3482,
+      "step": 4499
+    },
+    {
+      "epoch": 1.3489600899381675,
+      "grad_norm": 0.21773041784763336,
+      "learning_rate": 9.774774774774774e-05,
+      "loss": 0.3732,
+      "step": 4500
+    },
+    {
+      "epoch": 1.3489600899381675,
+      "eval_loss": 0.4129393398761749,
+      "eval_runtime": 565.7837,
+      "eval_samples_per_second": 3.816,
+      "eval_steps_per_second": 0.477,
+      "step": 4500
+    },
+    {
+      "epoch": 1.349259883829867,
+      "grad_norm": 0.20802539587020874,
+      "learning_rate": 9.77027027027027e-05,
+      "loss": 0.3597,
+      "step": 4501
+    },
+    {
+      "epoch": 1.3495596777215664,
+      "grad_norm": 0.2075861245393753,
+      "learning_rate": 9.765765765765765e-05,
+      "loss": 0.3471,
+      "step": 4502
+    },
+    {
+      "epoch": 1.349859471613266,
+      "grad_norm": 0.2349012792110443,
+      "learning_rate": 9.76126126126126e-05,
+      "loss": 0.3839,
+      "step": 4503
+    },
+    {
+      "epoch": 1.3501592655049652,
+      "grad_norm": 0.22266164422035217,
+      "learning_rate": 9.756756756756757e-05,
+      "loss": 0.3602,
+      "step": 4504
+    },
+    {
+      "epoch": 1.3504590593966648,
+      "grad_norm": 0.23045696318149567,
+      "learning_rate": 9.75225225225225e-05,
+      "loss": 0.379,
+      "step": 4505
+    },
+    {
+      "epoch": 1.3507588532883643,
+      "grad_norm": 0.2137947827577591,
+      "learning_rate": 9.747747747747747e-05,
+      "loss": 0.3318,
+      "step": 4506
+    },
+    {
+      "epoch": 1.3510586471800636,
+      "grad_norm": 0.21073168516159058,
+      "learning_rate": 9.743243243243243e-05,
+      "loss": 0.3563,
+      "step": 4507
+    },
+    {
+      "epoch": 1.3513584410717632,
+      "grad_norm": 0.21682259440422058,
+      "learning_rate": 9.738738738738737e-05,
+      "loss": 0.356,
+      "step": 4508
+    },
+    {
+      "epoch": 1.3516582349634625,
+      "grad_norm": 0.2167338877916336,
+      "learning_rate": 9.734234234234233e-05,
+      "loss": 0.3348,
+      "step": 4509
+    },
+    {
+      "epoch": 1.351958028855162,
+      "grad_norm": 0.21236591041088104,
+      "learning_rate": 9.72972972972973e-05,
+      "loss": 0.3618,
+      "step": 4510
+    },
+    {
+      "epoch": 1.3522578227468616,
+      "grad_norm": 0.20849741995334625,
+      "learning_rate": 9.725225225225223e-05,
+      "loss": 0.3405,
+      "step": 4511
+    },
+    {
+      "epoch": 1.352557616638561,
+      "grad_norm": 0.20645050704479218,
+      "learning_rate": 9.72072072072072e-05,
+      "loss": 0.3501,
+      "step": 4512
+    },
+    {
+      "epoch": 1.3528574105302604,
+      "grad_norm": 0.2172173261642456,
+      "learning_rate": 9.716216216216216e-05,
+      "loss": 0.3414,
+      "step": 4513
+    },
+    {
+      "epoch": 1.35315720442196,
+      "grad_norm": 0.21994781494140625,
+      "learning_rate": 9.711711711711711e-05,
+      "loss": 0.3745,
+      "step": 4514
+    },
+    {
+      "epoch": 1.3534569983136593,
+      "grad_norm": 0.22667646408081055,
+      "learning_rate": 9.707207207207206e-05,
+      "loss": 0.363,
+      "step": 4515
+    },
+    {
+      "epoch": 1.3537567922053588,
+      "grad_norm": 0.21792596578598022,
+      "learning_rate": 9.702702702702702e-05,
+      "loss": 0.364,
+      "step": 4516
+    },
+    {
+      "epoch": 1.3540565860970584,
+      "grad_norm": 0.216531440615654,
+      "learning_rate": 9.698198198198197e-05,
+      "loss": 0.3672,
+      "step": 4517
+    },
+    {
+      "epoch": 1.3543563799887577,
+      "grad_norm": 0.20364072918891907,
+      "learning_rate": 9.693693693693694e-05,
+      "loss": 0.3262,
+      "step": 4518
+    },
+    {
+      "epoch": 1.3546561738804572,
+      "grad_norm": 0.22151176631450653,
+      "learning_rate": 9.689189189189187e-05,
+      "loss": 0.3748,
+      "step": 4519
+    },
+    {
+      "epoch": 1.3549559677721565,
+      "grad_norm": 0.21700510382652283,
+      "learning_rate": 9.684684684684684e-05,
+      "loss": 0.3585,
+      "step": 4520
+    },
+    {
+      "epoch": 1.355255761663856,
+      "grad_norm": 0.21197721362113953,
+      "learning_rate": 9.68018018018018e-05,
+      "loss": 0.3448,
+      "step": 4521
+    },
+    {
+      "epoch": 1.3555555555555556,
+      "grad_norm": 0.21541689336299896,
+      "learning_rate": 9.675675675675674e-05,
+      "loss": 0.3622,
+      "step": 4522
+    },
+    {
+      "epoch": 1.355855349447255,
+      "grad_norm": 0.2156587541103363,
+      "learning_rate": 9.67117117117117e-05,
+      "loss": 0.363,
+      "step": 4523
+    },
+    {
+      "epoch": 1.3561551433389545,
+      "grad_norm": 0.21375496685504913,
+      "learning_rate": 9.666666666666667e-05,
+      "loss": 0.3576,
+      "step": 4524
+    },
+    {
+      "epoch": 1.3564549372306538,
+      "grad_norm": 0.208513543009758,
+      "learning_rate": 9.66216216216216e-05,
+      "loss": 0.3604,
+      "step": 4525
+    },
+    {
+      "epoch": 1.3567547311223533,
+      "grad_norm": 0.2164967805147171,
+      "learning_rate": 9.657657657657657e-05,
+      "loss": 0.3814,
+      "step": 4526
+    },
+    {
+      "epoch": 1.3570545250140529,
+      "grad_norm": 0.2145373374223709,
+      "learning_rate": 9.653153153153153e-05,
+      "loss": 0.3636,
+      "step": 4527
+    },
+    {
+      "epoch": 1.3573543189057524,
+      "grad_norm": 0.20882003009319305,
+      "learning_rate": 9.648648648648647e-05,
+      "loss": 0.347,
+      "step": 4528
+    },
+    {
+      "epoch": 1.3576541127974517,
+      "grad_norm": 0.2010725885629654,
+      "learning_rate": 9.644144144144143e-05,
+      "loss": 0.3423,
+      "step": 4529
+    },
+    {
+      "epoch": 1.3579539066891513,
+      "grad_norm": 0.2104565054178238,
+      "learning_rate": 9.639639639639638e-05,
+      "loss": 0.3505,
+      "step": 4530
+    },
+    {
+      "epoch": 1.3582537005808506,
+      "grad_norm": 0.2170221358537674,
+      "learning_rate": 9.635135135135135e-05,
+      "loss": 0.3708,
+      "step": 4531
+    },
+    {
+      "epoch": 1.3585534944725501,
+      "grad_norm": 0.21563062071800232,
+      "learning_rate": 9.63063063063063e-05,
+      "loss": 0.3764,
+      "step": 4532
+    },
+    {
+      "epoch": 1.3588532883642497,
+      "grad_norm": 0.21485255658626556,
+      "learning_rate": 9.626126126126125e-05,
+      "loss": 0.3438,
+      "step": 4533
+    },
+    {
+      "epoch": 1.359153082255949,
+      "grad_norm": 0.20779499411582947,
+      "learning_rate": 9.621621621621621e-05,
+      "loss": 0.3597,
+      "step": 4534
+    },
+    {
+      "epoch": 1.3594528761476485,
+      "grad_norm": 0.22415833175182343,
+      "learning_rate": 9.617117117117117e-05,
+      "loss": 0.3897,
+      "step": 4535
+    },
+    {
+      "epoch": 1.3597526700393479,
+      "grad_norm": 0.21616344153881073,
+      "learning_rate": 9.612612612612611e-05,
+      "loss": 0.3397,
+      "step": 4536
+    },
+    {
+      "epoch": 1.3600524639310474,
+      "grad_norm": 0.20136888325214386,
+      "learning_rate": 9.608108108108108e-05,
+      "loss": 0.3543,
+      "step": 4537
+    },
+    {
+      "epoch": 1.360352257822747,
+      "grad_norm": 0.21425513923168182,
+      "learning_rate": 9.603603603603604e-05,
+      "loss": 0.365,
+      "step": 4538
+    },
+    {
+      "epoch": 1.3606520517144463,
+      "grad_norm": 0.2139771580696106,
+      "learning_rate": 9.599099099099098e-05,
+      "loss": 0.3724,
+      "step": 4539
+    },
+    {
+      "epoch": 1.3609518456061458,
+      "grad_norm": 0.20793326199054718,
+      "learning_rate": 9.594594594594594e-05,
+      "loss": 0.3242,
+      "step": 4540
+    },
+    {
+      "epoch": 1.3612516394978451,
+      "grad_norm": 0.21706359088420868,
+      "learning_rate": 9.590090090090089e-05,
+      "loss": 0.3471,
+      "step": 4541
+    },
+    {
+      "epoch": 1.3615514333895447,
+      "grad_norm": 0.21347346901893616,
+      "learning_rate": 9.585585585585584e-05,
+      "loss": 0.343,
+      "step": 4542
+    },
+    {
+      "epoch": 1.3618512272812442,
+      "grad_norm": 0.22152802348136902,
+      "learning_rate": 9.58108108108108e-05,
+      "loss": 0.3857,
+      "step": 4543
+    },
+    {
+      "epoch": 1.3621510211729437,
+      "grad_norm": 0.21097399294376373,
+      "learning_rate": 9.576576576576575e-05,
+      "loss": 0.3675,
+      "step": 4544
+    },
+    {
+      "epoch": 1.362450815064643,
+      "grad_norm": 0.22488509118556976,
+      "learning_rate": 9.57207207207207e-05,
+      "loss": 0.3796,
+      "step": 4545
+    },
+    {
+      "epoch": 1.3627506089563426,
+      "grad_norm": 0.20570115745067596,
+      "learning_rate": 9.567567567567567e-05,
+      "loss": 0.3662,
+      "step": 4546
+    },
+    {
+      "epoch": 1.363050402848042,
+      "grad_norm": 0.20886462926864624,
+      "learning_rate": 9.563063063063062e-05,
+      "loss": 0.3735,
+      "step": 4547
+    },
+    {
+      "epoch": 1.3633501967397414,
+      "grad_norm": 0.22209970653057098,
+      "learning_rate": 9.558558558558558e-05,
+      "loss": 0.3648,
+      "step": 4548
+    },
+    {
+      "epoch": 1.363649990631441,
+      "grad_norm": 0.2033938765525818,
+      "learning_rate": 9.554054054054053e-05,
+      "loss": 0.3484,
+      "step": 4549
+    },
+    {
+      "epoch": 1.3639497845231403,
+      "grad_norm": 0.21983402967453003,
+      "learning_rate": 9.549549549549548e-05,
+      "loss": 0.3743,
+      "step": 4550
+    },
+    {
+      "epoch": 1.3642495784148398,
+      "grad_norm": 0.22346578538417816,
+      "learning_rate": 9.545045045045045e-05,
+      "loss": 0.3806,
+      "step": 4551
+    },
+    {
+      "epoch": 1.3645493723065392,
+      "grad_norm": 0.26459768414497375,
+      "learning_rate": 9.540540540540541e-05,
+      "loss": 0.3382,
+      "step": 4552
+    },
+    {
+      "epoch": 1.3648491661982387,
+      "grad_norm": 0.20662111043930054,
+      "learning_rate": 9.536036036036035e-05,
+      "loss": 0.3322,
+      "step": 4553
+    },
+    {
+      "epoch": 1.3651489600899382,
+      "grad_norm": 0.21967481076717377,
+      "learning_rate": 9.531531531531531e-05,
+      "loss": 0.3699,
+      "step": 4554
+    },
+    {
+      "epoch": 1.3654487539816376,
+      "grad_norm": 0.21974623203277588,
+      "learning_rate": 9.527027027027025e-05,
+      "loss": 0.3481,
+      "step": 4555
+    },
+    {
+      "epoch": 1.365748547873337,
+      "grad_norm": 0.19991014897823334,
+      "learning_rate": 9.522522522522521e-05,
+      "loss": 0.3348,
+      "step": 4556
+    },
+    {
+      "epoch": 1.3660483417650364,
+      "grad_norm": 0.22287118434906006,
+      "learning_rate": 9.518018018018018e-05,
+      "loss": 0.3791,
+      "step": 4557
+    },
+    {
+      "epoch": 1.366348135656736,
+      "grad_norm": 0.22346952557563782,
+      "learning_rate": 9.513513513513513e-05,
+      "loss": 0.3663,
+      "step": 4558
+    },
+    {
+      "epoch": 1.3666479295484355,
+      "grad_norm": 0.2181125432252884,
+      "learning_rate": 9.509009009009008e-05,
+      "loss": 0.3678,
+      "step": 4559
+    },
+    {
+      "epoch": 1.366947723440135,
+      "grad_norm": 0.2311943769454956,
+      "learning_rate": 9.504504504504504e-05,
+      "loss": 0.3705,
+      "step": 4560
+    },
+    {
+      "epoch": 1.3672475173318344,
+      "grad_norm": 0.20911650359630585,
+      "learning_rate": 9.499999999999999e-05,
+      "loss": 0.3471,
+      "step": 4561
+    },
+    {
+      "epoch": 1.367547311223534,
+      "grad_norm": 0.22261382639408112,
+      "learning_rate": 9.495495495495494e-05,
+      "loss": 0.3691,
+      "step": 4562
+    },
+    {
+      "epoch": 1.3678471051152332,
+      "grad_norm": 0.2156069576740265,
+      "learning_rate": 9.49099099099099e-05,
+      "loss": 0.3537,
+      "step": 4563
+    },
+    {
+      "epoch": 1.3681468990069328,
+      "grad_norm": 0.21138927340507507,
+      "learning_rate": 9.486486486486486e-05,
+      "loss": 0.3507,
+      "step": 4564
+    },
+    {
+      "epoch": 1.3684466928986323,
+      "grad_norm": 0.2109546810388565,
+      "learning_rate": 9.481981981981982e-05,
+      "loss": 0.3669,
+      "step": 4565
+    },
+    {
+      "epoch": 1.3687464867903316,
+      "grad_norm": 0.2068871259689331,
+      "learning_rate": 9.477477477477476e-05,
+      "loss": 0.3529,
+      "step": 4566
+    },
+    {
+      "epoch": 1.3690462806820312,
+      "grad_norm": 0.21869945526123047,
+      "learning_rate": 9.472972972972972e-05,
+      "loss": 0.3913,
+      "step": 4567
+    },
+    {
+      "epoch": 1.3693460745737305,
+      "grad_norm": 0.2108869105577469,
+      "learning_rate": 9.468468468468468e-05,
+      "loss": 0.3622,
+      "step": 4568
+    },
+    {
+      "epoch": 1.36964586846543,
+      "grad_norm": 0.21931153535842896,
+      "learning_rate": 9.463963963963962e-05,
+      "loss": 0.3598,
+      "step": 4569
+    },
+    {
+      "epoch": 1.3699456623571296,
+      "grad_norm": 0.2235441952943802,
+      "learning_rate": 9.459459459459459e-05,
+      "loss": 0.3696,
+      "step": 4570
+    },
+    {
+      "epoch": 1.3702454562488289,
+      "grad_norm": 0.21451959013938904,
+      "learning_rate": 9.454954954954955e-05,
+      "loss": 0.3565,
+      "step": 4571
+    },
+    {
+      "epoch": 1.3705452501405284,
+      "grad_norm": 0.22164882719516754,
+      "learning_rate": 9.450450450450449e-05,
+      "loss": 0.3669,
+      "step": 4572
+    },
+    {
+      "epoch": 1.3708450440322277,
+      "grad_norm": 0.20921410620212555,
+      "learning_rate": 9.445945945945945e-05,
+      "loss": 0.3552,
+      "step": 4573
+    },
+    {
+      "epoch": 1.3711448379239273,
+      "grad_norm": 0.22980685532093048,
+      "learning_rate": 9.441441441441441e-05,
+      "loss": 0.3638,
+      "step": 4574
+    },
+    {
+      "epoch": 1.3714446318156268,
+      "grad_norm": 0.20418162643909454,
+      "learning_rate": 9.436936936936936e-05,
+      "loss": 0.3414,
+      "step": 4575
+    },
+    {
+      "epoch": 1.3717444257073261,
+      "grad_norm": 0.20581036806106567,
+      "learning_rate": 9.432432432432431e-05,
+      "loss": 0.3271,
+      "step": 4576
+    },
+    {
+      "epoch": 1.3720442195990257,
+      "grad_norm": 0.21641530096530914,
+      "learning_rate": 9.427927927927926e-05,
+      "loss": 0.35,
+      "step": 4577
+    },
+    {
+      "epoch": 1.372344013490725,
+      "grad_norm": 0.21078841388225555,
+      "learning_rate": 9.423423423423423e-05,
+      "loss": 0.3677,
+      "step": 4578
+    },
+    {
+      "epoch": 1.3726438073824245,
+      "grad_norm": 0.22394929826259613,
+      "learning_rate": 9.418918918918918e-05,
+      "loss": 0.3811,
+      "step": 4579
+    },
+    {
+      "epoch": 1.372943601274124,
+      "grad_norm": 0.21448741853237152,
+      "learning_rate": 9.414414414414413e-05,
+      "loss": 0.3627,
+      "step": 4580
+    },
+    {
+      "epoch": 1.3732433951658236,
+      "grad_norm": 0.23157335817813873,
+      "learning_rate": 9.409909909909909e-05,
+      "loss": 0.3881,
+      "step": 4581
+    },
+    {
+      "epoch": 1.373543189057523,
+      "grad_norm": 0.21925771236419678,
+      "learning_rate": 9.405405405405406e-05,
+      "loss": 0.3546,
+      "step": 4582
+    },
+    {
+      "epoch": 1.3738429829492225,
+      "grad_norm": 0.21934622526168823,
+      "learning_rate": 9.4009009009009e-05,
+      "loss": 0.3487,
+      "step": 4583
+    },
+    {
+      "epoch": 1.3741427768409218,
+      "grad_norm": 0.22106409072875977,
+      "learning_rate": 9.396396396396396e-05,
+      "loss": 0.3689,
+      "step": 4584
+    },
+    {
+      "epoch": 1.3744425707326213,
+      "grad_norm": 0.2363518327474594,
+      "learning_rate": 9.391891891891892e-05,
+      "loss": 0.36,
+      "step": 4585
+    },
+    {
+      "epoch": 1.3747423646243209,
+      "grad_norm": 0.22158819437026978,
+      "learning_rate": 9.387387387387386e-05,
+      "loss": 0.37,
+      "step": 4586
+    },
+    {
+      "epoch": 1.3750421585160202,
+      "grad_norm": 0.2105867564678192,
+      "learning_rate": 9.382882882882882e-05,
+      "loss": 0.3588,
+      "step": 4587
+    },
+    {
+      "epoch": 1.3753419524077197,
+      "grad_norm": 0.22234956920146942,
+      "learning_rate": 9.378378378378379e-05,
+      "loss": 0.3523,
+      "step": 4588
+    },
+    {
+      "epoch": 1.375641746299419,
+      "grad_norm": 0.20997658371925354,
+      "learning_rate": 9.373873873873872e-05,
+      "loss": 0.3504,
+      "step": 4589
+    },
+    {
+      "epoch": 1.3759415401911186,
+      "grad_norm": 0.22327353060245514,
+      "learning_rate": 9.369369369369369e-05,
+      "loss": 0.365,
+      "step": 4590
+    },
+    {
+      "epoch": 1.3762413340828181,
+      "grad_norm": 0.218084916472435,
+      "learning_rate": 9.364864864864864e-05,
+      "loss": 0.3526,
+      "step": 4591
+    },
+    {
+      "epoch": 1.3765411279745174,
+      "grad_norm": 0.2157234251499176,
+      "learning_rate": 9.36036036036036e-05,
+      "loss": 0.3539,
+      "step": 4592
+    },
+    {
+      "epoch": 1.376840921866217,
+      "grad_norm": 0.22182735800743103,
+      "learning_rate": 9.355855855855855e-05,
+      "loss": 0.3468,
+      "step": 4593
+    },
+    {
+      "epoch": 1.3771407157579163,
+      "grad_norm": 0.21038967370986938,
+      "learning_rate": 9.35135135135135e-05,
+      "loss": 0.3549,
+      "step": 4594
+    },
+    {
+      "epoch": 1.3774405096496158,
+      "grad_norm": 0.2077571451663971,
+      "learning_rate": 9.346846846846847e-05,
+      "loss": 0.3517,
+      "step": 4595
+    },
+    {
+      "epoch": 1.3777403035413154,
+      "grad_norm": 0.20815788209438324,
+      "learning_rate": 9.342342342342342e-05,
+      "loss": 0.3568,
+      "step": 4596
+    },
+    {
+      "epoch": 1.378040097433015,
+      "grad_norm": 0.21737034618854523,
+      "learning_rate": 9.337837837837837e-05,
+      "loss": 0.3792,
+      "step": 4597
+    },
+    {
+      "epoch": 1.3783398913247142,
+      "grad_norm": 0.2084619402885437,
+      "learning_rate": 9.333333333333333e-05,
+      "loss": 0.3389,
+      "step": 4598
+    },
+    {
+      "epoch": 1.3786396852164138,
+      "grad_norm": 0.222722589969635,
+      "learning_rate": 9.32882882882883e-05,
+      "loss": 0.3812,
+      "step": 4599
+    },
+    {
+      "epoch": 1.378939479108113,
+      "grad_norm": 0.21024678647518158,
+      "learning_rate": 9.324324324324323e-05,
+      "loss": 0.3675,
+      "step": 4600
+    },
+    {
+      "epoch": 1.3792392729998126,
+      "grad_norm": 0.21897338330745697,
+      "learning_rate": 9.31981981981982e-05,
+      "loss": 0.3664,
+      "step": 4601
+    },
+    {
+      "epoch": 1.3795390668915122,
+      "grad_norm": 0.21706995368003845,
+      "learning_rate": 9.315315315315313e-05,
+      "loss": 0.346,
+      "step": 4602
+    },
+    {
+      "epoch": 1.3798388607832115,
+      "grad_norm": 0.21143165230751038,
+      "learning_rate": 9.31081081081081e-05,
+      "loss": 0.3465,
+      "step": 4603
+    },
+    {
+      "epoch": 1.380138654674911,
+      "grad_norm": 0.2121080905199051,
+      "learning_rate": 9.306306306306306e-05,
+      "loss": 0.349,
+      "step": 4604
+    },
+    {
+      "epoch": 1.3804384485666104,
+      "grad_norm": 0.21112461388111115,
+      "learning_rate": 9.301801801801801e-05,
+      "loss": 0.3558,
+      "step": 4605
+    },
+    {
+      "epoch": 1.38073824245831,
+      "grad_norm": 0.21366949379444122,
+      "learning_rate": 9.297297297297296e-05,
+      "loss": 0.3601,
+      "step": 4606
+    },
+    {
+      "epoch": 1.3810380363500094,
+      "grad_norm": 0.21894964575767517,
+      "learning_rate": 9.292792792792792e-05,
+      "loss": 0.3619,
+      "step": 4607
+    },
+    {
+      "epoch": 1.3813378302417088,
+      "grad_norm": 0.21191251277923584,
+      "learning_rate": 9.288288288288287e-05,
+      "loss": 0.3592,
+      "step": 4608
+    },
+    {
+      "epoch": 1.3816376241334083,
+      "grad_norm": 0.22147145867347717,
+      "learning_rate": 9.283783783783784e-05,
+      "loss": 0.3592,
+      "step": 4609
+    },
+    {
+      "epoch": 1.3819374180251076,
+      "grad_norm": 0.2090921849012375,
+      "learning_rate": 9.279279279279279e-05,
+      "loss": 0.3391,
+      "step": 4610
+    },
+    {
+      "epoch": 1.3822372119168072,
+      "grad_norm": 0.20255540311336517,
+      "learning_rate": 9.274774774774774e-05,
+      "loss": 0.3383,
+      "step": 4611
+    },
+    {
+      "epoch": 1.3825370058085067,
+      "grad_norm": 0.21785931289196014,
+      "learning_rate": 9.27027027027027e-05,
+      "loss": 0.3639,
+      "step": 4612
+    },
+    {
+      "epoch": 1.3828367997002062,
+      "grad_norm": 0.21439585089683533,
+      "learning_rate": 9.265765765765764e-05,
+      "loss": 0.3805,
+      "step": 4613
+    },
+    {
+      "epoch": 1.3831365935919055,
+      "grad_norm": 0.2247477024793625,
+      "learning_rate": 9.26126126126126e-05,
+      "loss": 0.3709,
+      "step": 4614
+    },
+    {
+      "epoch": 1.383436387483605,
+      "grad_norm": 0.22263139486312866,
+      "learning_rate": 9.256756756756757e-05,
+      "loss": 0.3801,
+      "step": 4615
+    },
+    {
+      "epoch": 1.3837361813753044,
+      "grad_norm": 0.22057713568210602,
+      "learning_rate": 9.25225225225225e-05,
+      "loss": 0.3709,
+      "step": 4616
+    },
+    {
+      "epoch": 1.384035975267004,
+      "grad_norm": 0.21248671412467957,
+      "learning_rate": 9.247747747747747e-05,
+      "loss": 0.3655,
+      "step": 4617
+    },
+    {
+      "epoch": 1.3843357691587035,
+      "grad_norm": 0.22102704644203186,
+      "learning_rate": 9.243243243243243e-05,
+      "loss": 0.3594,
+      "step": 4618
+    },
+    {
+      "epoch": 1.3846355630504028,
+      "grad_norm": 0.21167469024658203,
+      "learning_rate": 9.238738738738737e-05,
+      "loss": 0.3515,
+      "step": 4619
+    },
+    {
+      "epoch": 1.3849353569421023,
+      "grad_norm": 0.225198894739151,
+      "learning_rate": 9.234234234234233e-05,
+      "loss": 0.3936,
+      "step": 4620
+    },
+    {
+      "epoch": 1.3852351508338017,
+      "grad_norm": 0.22111676633358002,
+      "learning_rate": 9.22972972972973e-05,
+      "loss": 0.3449,
+      "step": 4621
+    },
+    {
+      "epoch": 1.3855349447255012,
+      "grad_norm": 0.2069976031780243,
+      "learning_rate": 9.225225225225225e-05,
+      "loss": 0.3571,
+      "step": 4622
+    },
+    {
+      "epoch": 1.3858347386172007,
+      "grad_norm": 0.21398650109767914,
+      "learning_rate": 9.22072072072072e-05,
+      "loss": 0.3696,
+      "step": 4623
+    },
+    {
+      "epoch": 1.3861345325089,
+      "grad_norm": 0.1977337896823883,
+      "learning_rate": 9.216216216216216e-05,
+      "loss": 0.3493,
+      "step": 4624
+    },
+    {
+      "epoch": 1.3864343264005996,
+      "grad_norm": 0.21368259191513062,
+      "learning_rate": 9.211711711711711e-05,
+      "loss": 0.3657,
+      "step": 4625
+    },
+    {
+      "epoch": 1.386734120292299,
+      "grad_norm": 0.2091464400291443,
+      "learning_rate": 9.207207207207206e-05,
+      "loss": 0.3497,
+      "step": 4626
+    },
+    {
+      "epoch": 1.3870339141839985,
+      "grad_norm": 0.1985793113708496,
+      "learning_rate": 9.202702702702701e-05,
+      "loss": 0.3193,
+      "step": 4627
+    },
+    {
+      "epoch": 1.387333708075698,
+      "grad_norm": 0.2133084088563919,
+      "learning_rate": 9.198198198198197e-05,
+      "loss": 0.3502,
+      "step": 4628
+    },
+    {
+      "epoch": 1.3876335019673975,
+      "grad_norm": 0.2148822396993637,
+      "learning_rate": 9.193693693693694e-05,
+      "loss": 0.3713,
+      "step": 4629
+    },
+    {
+      "epoch": 1.3879332958590969,
+      "grad_norm": 0.20007802546024323,
+      "learning_rate": 9.189189189189188e-05,
+      "loss": 0.3348,
+      "step": 4630
+    },
+    {
+      "epoch": 1.3882330897507964,
+      "grad_norm": 0.21505482494831085,
+      "learning_rate": 9.184684684684684e-05,
+      "loss": 0.3381,
+      "step": 4631
+    },
+    {
+      "epoch": 1.3885328836424957,
+      "grad_norm": 0.21337425708770752,
+      "learning_rate": 9.18018018018018e-05,
+      "loss": 0.3535,
+      "step": 4632
+    },
+    {
+      "epoch": 1.3888326775341953,
+      "grad_norm": 0.2174317091703415,
+      "learning_rate": 9.175675675675674e-05,
+      "loss": 0.3692,
+      "step": 4633
+    },
+    {
+      "epoch": 1.3891324714258948,
+      "grad_norm": 0.2055487334728241,
+      "learning_rate": 9.17117117117117e-05,
+      "loss": 0.3505,
+      "step": 4634
+    },
+    {
+      "epoch": 1.3894322653175941,
+      "grad_norm": 0.21364519000053406,
+      "learning_rate": 9.166666666666667e-05,
+      "loss": 0.3436,
+      "step": 4635
+    },
+    {
+      "epoch": 1.3897320592092937,
+      "grad_norm": 0.2120402306318283,
+      "learning_rate": 9.16216216216216e-05,
+      "loss": 0.361,
+      "step": 4636
+    },
+    {
+      "epoch": 1.390031853100993,
+      "grad_norm": 0.210201233625412,
+      "learning_rate": 9.157657657657657e-05,
+      "loss": 0.3496,
+      "step": 4637
+    },
+    {
+      "epoch": 1.3903316469926925,
+      "grad_norm": 0.22710666060447693,
+      "learning_rate": 9.153153153153152e-05,
+      "loss": 0.3787,
+      "step": 4638
+    },
+    {
+      "epoch": 1.390631440884392,
+      "grad_norm": 0.21441099047660828,
+      "learning_rate": 9.148648648648648e-05,
+      "loss": 0.3516,
+      "step": 4639
+    },
+    {
+      "epoch": 1.3909312347760914,
+      "grad_norm": 0.21323607861995697,
+      "learning_rate": 9.144144144144143e-05,
+      "loss": 0.3751,
+      "step": 4640
+    },
+    {
+      "epoch": 1.391231028667791,
+      "grad_norm": 0.21749518811702728,
+      "learning_rate": 9.139639639639638e-05,
+      "loss": 0.3544,
+      "step": 4641
+    },
+    {
+      "epoch": 1.3915308225594902,
+      "grad_norm": 0.2149432897567749,
+      "learning_rate": 9.135135135135135e-05,
+      "loss": 0.3548,
+      "step": 4642
+    },
+    {
+      "epoch": 1.3918306164511898,
+      "grad_norm": 0.2081831991672516,
+      "learning_rate": 9.13063063063063e-05,
+      "loss": 0.3496,
+      "step": 4643
+    },
+    {
+      "epoch": 1.3921304103428893,
+      "grad_norm": 0.19373933970928192,
+      "learning_rate": 9.126126126126125e-05,
+      "loss": 0.3397,
+      "step": 4644
+    },
+    {
+      "epoch": 1.3924302042345889,
+      "grad_norm": 0.2193603813648224,
+      "learning_rate": 9.121621621621621e-05,
+      "loss": 0.3663,
+      "step": 4645
+    },
+    {
+      "epoch": 1.3927299981262882,
+      "grad_norm": 0.2071419656276703,
+      "learning_rate": 9.117117117117118e-05,
+      "loss": 0.3459,
+      "step": 4646
+    },
+    {
+      "epoch": 1.3930297920179877,
+      "grad_norm": 0.21072034537792206,
+      "learning_rate": 9.112612612612611e-05,
+      "loss": 0.3491,
+      "step": 4647
+    },
+    {
+      "epoch": 1.393329585909687,
+      "grad_norm": 0.21739129722118378,
+      "learning_rate": 9.108108108108108e-05,
+      "loss": 0.358,
+      "step": 4648
+    },
+    {
+      "epoch": 1.3936293798013866,
+      "grad_norm": 0.21196670830249786,
+      "learning_rate": 9.103603603603604e-05,
+      "loss": 0.3595,
+      "step": 4649
+    },
+    {
+      "epoch": 1.393929173693086,
+      "grad_norm": 0.20686453580856323,
+      "learning_rate": 9.099099099099098e-05,
+      "loss": 0.3514,
+      "step": 4650
+    },
+    {
+      "epoch": 1.3942289675847854,
+      "grad_norm": 0.21052870154380798,
+      "learning_rate": 9.094594594594594e-05,
+      "loss": 0.3654,
+      "step": 4651
+    },
+    {
+      "epoch": 1.394528761476485,
+      "grad_norm": 0.21836452186107635,
+      "learning_rate": 9.090090090090089e-05,
+      "loss": 0.3569,
+      "step": 4652
+    },
+    {
+      "epoch": 1.3948285553681843,
+      "grad_norm": 0.21096113324165344,
+      "learning_rate": 9.085585585585584e-05,
+      "loss": 0.3436,
+      "step": 4653
+    },
+    {
+      "epoch": 1.3951283492598838,
+      "grad_norm": 0.20840652287006378,
+      "learning_rate": 9.08108108108108e-05,
+      "loss": 0.3413,
+      "step": 4654
+    },
+    {
+      "epoch": 1.3954281431515834,
+      "grad_norm": 0.2149275839328766,
+      "learning_rate": 9.076576576576576e-05,
+      "loss": 0.3708,
+      "step": 4655
+    },
+    {
+      "epoch": 1.3957279370432827,
+      "grad_norm": 0.2034139484167099,
+      "learning_rate": 9.072072072072072e-05,
+      "loss": 0.3678,
+      "step": 4656
+    },
+    {
+      "epoch": 1.3960277309349822,
+      "grad_norm": 0.21449275314807892,
+      "learning_rate": 9.067567567567567e-05,
+      "loss": 0.3314,
+      "step": 4657
+    },
+    {
+      "epoch": 1.3963275248266815,
+      "grad_norm": 0.22322918474674225,
+      "learning_rate": 9.063063063063062e-05,
+      "loss": 0.3792,
+      "step": 4658
+    },
+    {
+      "epoch": 1.396627318718381,
+      "grad_norm": 0.1963275969028473,
+      "learning_rate": 9.058558558558558e-05,
+      "loss": 0.3371,
+      "step": 4659
+    },
+    {
+      "epoch": 1.3969271126100806,
+      "grad_norm": 0.2096429467201233,
+      "learning_rate": 9.054054054054053e-05,
+      "loss": 0.3615,
+      "step": 4660
+    },
+    {
+      "epoch": 1.3972269065017802,
+      "grad_norm": 0.2007439285516739,
+      "learning_rate": 9.049549549549548e-05,
+      "loss": 0.3225,
+      "step": 4661
+    },
+    {
+      "epoch": 1.3975267003934795,
+      "grad_norm": 0.22279343008995056,
+      "learning_rate": 9.045045045045045e-05,
+      "loss": 0.3572,
+      "step": 4662
+    },
+    {
+      "epoch": 1.397826494285179,
+      "grad_norm": 0.217242032289505,
+      "learning_rate": 9.040540540540539e-05,
+      "loss": 0.3578,
+      "step": 4663
+    },
+    {
+      "epoch": 1.3981262881768783,
+      "grad_norm": 0.2023712396621704,
+      "learning_rate": 9.036036036036035e-05,
+      "loss": 0.3249,
+      "step": 4664
+    },
+    {
+      "epoch": 1.3984260820685779,
+      "grad_norm": 0.20755234360694885,
+      "learning_rate": 9.031531531531531e-05,
+      "loss": 0.3523,
+      "step": 4665
+    },
+    {
+      "epoch": 1.3987258759602774,
+      "grad_norm": 0.20431917905807495,
+      "learning_rate": 9.027027027027025e-05,
+      "loss": 0.3305,
+      "step": 4666
+    },
+    {
+      "epoch": 1.3990256698519767,
+      "grad_norm": 0.21976415812969208,
+      "learning_rate": 9.022522522522521e-05,
+      "loss": 0.3601,
+      "step": 4667
+    },
+    {
+      "epoch": 1.3993254637436763,
+      "grad_norm": 0.21623222529888153,
+      "learning_rate": 9.018018018018018e-05,
+      "loss": 0.3684,
+      "step": 4668
+    },
+    {
+      "epoch": 1.3996252576353756,
+      "grad_norm": 0.20926150679588318,
+      "learning_rate": 9.013513513513513e-05,
+      "loss": 0.3714,
+      "step": 4669
+    },
+    {
+      "epoch": 1.3999250515270751,
+      "grad_norm": 0.22803552448749542,
+      "learning_rate": 9.009009009009008e-05,
+      "loss": 0.3691,
+      "step": 4670
+    },
+    {
+      "epoch": 1.4002248454187747,
+      "grad_norm": 0.21862058341503143,
+      "learning_rate": 9.004504504504504e-05,
+      "loss": 0.3746,
+      "step": 4671
+    },
+    {
+      "epoch": 1.400524639310474,
+      "grad_norm": 0.22191566228866577,
+      "learning_rate": 8.999999999999999e-05,
+      "loss": 0.3745,
+      "step": 4672
+    },
+    {
+      "epoch": 1.4008244332021735,
+      "grad_norm": 0.21257881820201874,
+      "learning_rate": 8.995495495495496e-05,
+      "loss": 0.3724,
+      "step": 4673
+    },
+    {
+      "epoch": 1.4011242270938729,
+      "grad_norm": 0.20726177096366882,
+      "learning_rate": 8.990990990990989e-05,
+      "loss": 0.363,
+      "step": 4674
+    },
+    {
+      "epoch": 1.4014240209855724,
+      "grad_norm": 0.21418994665145874,
+      "learning_rate": 8.986486486486486e-05,
+      "loss": 0.3771,
+      "step": 4675
+    },
+    {
+      "epoch": 1.401723814877272,
+      "grad_norm": 0.2140100598335266,
+      "learning_rate": 8.981981981981982e-05,
+      "loss": 0.3518,
+      "step": 4676
+    },
+    {
+      "epoch": 1.4020236087689713,
+      "grad_norm": 0.20784786343574524,
+      "learning_rate": 8.977477477477476e-05,
+      "loss": 0.3565,
+      "step": 4677
+    },
+    {
+      "epoch": 1.4023234026606708,
+      "grad_norm": 0.2127532958984375,
+      "learning_rate": 8.972972972972972e-05,
+      "loss": 0.3675,
+      "step": 4678
+    },
+    {
+      "epoch": 1.40262319655237,
+      "grad_norm": 0.1980638951063156,
+      "learning_rate": 8.968468468468469e-05,
+      "loss": 0.3402,
+      "step": 4679
+    },
+    {
+      "epoch": 1.4029229904440697,
+      "grad_norm": 0.21542289853096008,
+      "learning_rate": 8.963963963963962e-05,
+      "loss": 0.3871,
+      "step": 4680
+    },
+    {
+      "epoch": 1.4032227843357692,
+      "grad_norm": 0.2163427472114563,
+      "learning_rate": 8.959459459459459e-05,
+      "loss": 0.3631,
+      "step": 4681
+    },
+    {
+      "epoch": 1.4035225782274687,
+      "grad_norm": 0.2084680050611496,
+      "learning_rate": 8.954954954954955e-05,
+      "loss": 0.3426,
+      "step": 4682
+    },
+    {
+      "epoch": 1.403822372119168,
+      "grad_norm": 0.2164963334798813,
+      "learning_rate": 8.950450450450449e-05,
+      "loss": 0.372,
+      "step": 4683
+    },
+    {
+      "epoch": 1.4041221660108676,
+      "grad_norm": 0.20649103820323944,
+      "learning_rate": 8.945945945945945e-05,
+      "loss": 0.3465,
+      "step": 4684
+    },
+    {
+      "epoch": 1.404421959902567,
+      "grad_norm": 0.21171538531780243,
+      "learning_rate": 8.941441441441441e-05,
+      "loss": 0.3644,
+      "step": 4685
+    },
+    {
+      "epoch": 1.4047217537942664,
+      "grad_norm": 0.2132863998413086,
+      "learning_rate": 8.936936936936936e-05,
+      "loss": 0.3751,
+      "step": 4686
+    },
+    {
+      "epoch": 1.405021547685966,
+      "grad_norm": 0.2170388251543045,
+      "learning_rate": 8.932432432432431e-05,
+      "loss": 0.3664,
+      "step": 4687
+    },
+    {
+      "epoch": 1.4053213415776653,
+      "grad_norm": 0.21167437732219696,
+      "learning_rate": 8.927927927927926e-05,
+      "loss": 0.36,
+      "step": 4688
+    },
+    {
+      "epoch": 1.4056211354693648,
+      "grad_norm": 0.21021796762943268,
+      "learning_rate": 8.923423423423423e-05,
+      "loss": 0.3773,
+      "step": 4689
+    },
+    {
+      "epoch": 1.4059209293610642,
+      "grad_norm": 0.21205608546733856,
+      "learning_rate": 8.918918918918919e-05,
+      "loss": 0.3418,
+      "step": 4690
+    },
+    {
+      "epoch": 1.4062207232527637,
+      "grad_norm": 0.20514236390590668,
+      "learning_rate": 8.914414414414413e-05,
+      "loss": 0.3519,
+      "step": 4691
+    },
+    {
+      "epoch": 1.4065205171444632,
+      "grad_norm": 0.21129223704338074,
+      "learning_rate": 8.90990990990991e-05,
+      "loss": 0.3532,
+      "step": 4692
+    },
+    {
+      "epoch": 1.4068203110361626,
+      "grad_norm": 0.2106056809425354,
+      "learning_rate": 8.905405405405406e-05,
+      "loss": 0.3514,
+      "step": 4693
+    },
+    {
+      "epoch": 1.407120104927862,
+      "grad_norm": 0.21528342366218567,
+      "learning_rate": 8.9009009009009e-05,
+      "loss": 0.3759,
+      "step": 4694
+    },
+    {
+      "epoch": 1.4074198988195614,
+      "grad_norm": 0.21354471147060394,
+      "learning_rate": 8.896396396396396e-05,
+      "loss": 0.372,
+      "step": 4695
+    },
+    {
+      "epoch": 1.407719692711261,
+      "grad_norm": 0.2052365392446518,
+      "learning_rate": 8.891891891891892e-05,
+      "loss": 0.3421,
+      "step": 4696
+    },
+    {
+      "epoch": 1.4080194866029605,
+      "grad_norm": 0.2149754762649536,
+      "learning_rate": 8.887387387387386e-05,
+      "loss": 0.3505,
+      "step": 4697
+    },
+    {
+      "epoch": 1.40831928049466,
+      "grad_norm": 0.20838052034378052,
+      "learning_rate": 8.882882882882882e-05,
+      "loss": 0.3604,
+      "step": 4698
+    },
+    {
+      "epoch": 1.4086190743863594,
+      "grad_norm": 0.21180404722690582,
+      "learning_rate": 8.878378378378377e-05,
+      "loss": 0.342,
+      "step": 4699
+    },
+    {
+      "epoch": 1.408918868278059,
+      "grad_norm": 0.2073618769645691,
+      "learning_rate": 8.873873873873872e-05,
+      "loss": 0.3694,
+      "step": 4700
+    },
+    {
+      "epoch": 1.4092186621697582,
+      "grad_norm": 0.20314088463783264,
+      "learning_rate": 8.869369369369369e-05,
+      "loss": 0.366,
+      "step": 4701
+    },
+    {
+      "epoch": 1.4095184560614578,
+      "grad_norm": 0.22556687891483307,
+      "learning_rate": 8.864864864864864e-05,
+      "loss": 0.3434,
+      "step": 4702
+    },
+    {
+      "epoch": 1.4098182499531573,
+      "grad_norm": 0.21841377019882202,
+      "learning_rate": 8.86036036036036e-05,
+      "loss": 0.3622,
+      "step": 4703
+    },
+    {
+      "epoch": 1.4101180438448566,
+      "grad_norm": 0.21003997325897217,
+      "learning_rate": 8.855855855855855e-05,
+      "loss": 0.3658,
+      "step": 4704
+    },
+    {
+      "epoch": 1.4104178377365562,
+      "grad_norm": 0.20329275727272034,
+      "learning_rate": 8.85135135135135e-05,
+      "loss": 0.3793,
+      "step": 4705
+    },
+    {
+      "epoch": 1.4107176316282555,
+      "grad_norm": 0.22247040271759033,
+      "learning_rate": 8.846846846846847e-05,
+      "loss": 0.3687,
+      "step": 4706
+    },
+    {
+      "epoch": 1.411017425519955,
+      "grad_norm": 0.21038077771663666,
+      "learning_rate": 8.842342342342343e-05,
+      "loss": 0.3435,
+      "step": 4707
+    },
+    {
+      "epoch": 1.4113172194116546,
+      "grad_norm": 0.21580614149570465,
+      "learning_rate": 8.837837837837837e-05,
+      "loss": 0.353,
+      "step": 4708
+    },
+    {
+      "epoch": 1.4116170133033539,
+      "grad_norm": 0.20951011776924133,
+      "learning_rate": 8.833333333333333e-05,
+      "loss": 0.3314,
+      "step": 4709
+    },
+    {
+      "epoch": 1.4119168071950534,
+      "grad_norm": 0.21354609727859497,
+      "learning_rate": 8.828828828828827e-05,
+      "loss": 0.3393,
+      "step": 4710
+    },
+    {
+      "epoch": 1.4122166010867527,
+      "grad_norm": 0.20141154527664185,
+      "learning_rate": 8.824324324324323e-05,
+      "loss": 0.3392,
+      "step": 4711
+    },
+    {
+      "epoch": 1.4125163949784523,
+      "grad_norm": 0.21682001650333405,
+      "learning_rate": 8.81981981981982e-05,
+      "loss": 0.3473,
+      "step": 4712
+    },
+    {
+      "epoch": 1.4128161888701518,
+      "grad_norm": 0.24055300652980804,
+      "learning_rate": 8.815315315315314e-05,
+      "loss": 0.3727,
+      "step": 4713
+    },
+    {
+      "epoch": 1.4131159827618514,
+      "grad_norm": 0.21340611577033997,
+      "learning_rate": 8.81081081081081e-05,
+      "loss": 0.3668,
+      "step": 4714
+    },
+    {
+      "epoch": 1.4134157766535507,
+      "grad_norm": 0.22251774370670319,
+      "learning_rate": 8.806306306306306e-05,
+      "loss": 0.3509,
+      "step": 4715
+    },
+    {
+      "epoch": 1.4137155705452502,
+      "grad_norm": 0.21125243604183197,
+      "learning_rate": 8.801801801801801e-05,
+      "loss": 0.3619,
+      "step": 4716
+    },
+    {
+      "epoch": 1.4140153644369495,
+      "grad_norm": 0.22132828831672668,
+      "learning_rate": 8.797297297297296e-05,
+      "loss": 0.3874,
+      "step": 4717
+    },
+    {
+      "epoch": 1.414315158328649,
+      "grad_norm": 0.21483568847179413,
+      "learning_rate": 8.792792792792792e-05,
+      "loss": 0.3462,
+      "step": 4718
+    },
+    {
+      "epoch": 1.4146149522203486,
+      "grad_norm": 0.2184942066669464,
+      "learning_rate": 8.788288288288287e-05,
+      "loss": 0.3579,
+      "step": 4719
+    },
+    {
+      "epoch": 1.414914746112048,
+      "grad_norm": 0.2186221331357956,
+      "learning_rate": 8.783783783783784e-05,
+      "loss": 0.3633,
+      "step": 4720
+    },
+    {
+      "epoch": 1.4152145400037475,
+      "grad_norm": 0.20850953459739685,
+      "learning_rate": 8.779279279279279e-05,
+      "loss": 0.3622,
+      "step": 4721
+    },
+    {
+      "epoch": 1.4155143338954468,
+      "grad_norm": 0.21589501202106476,
+      "learning_rate": 8.774774774774774e-05,
+      "loss": 0.3308,
+      "step": 4722
+    },
+    {
+      "epoch": 1.4158141277871463,
+      "grad_norm": 0.20245422422885895,
+      "learning_rate": 8.77027027027027e-05,
+      "loss": 0.3527,
+      "step": 4723
+    },
+    {
+      "epoch": 1.4161139216788459,
+      "grad_norm": 0.21961507201194763,
+      "learning_rate": 8.765765765765764e-05,
+      "loss": 0.3637,
+      "step": 4724
+    },
+    {
+      "epoch": 1.4164137155705452,
+      "grad_norm": 0.21779102087020874,
+      "learning_rate": 8.76126126126126e-05,
+      "loss": 0.3713,
+      "step": 4725
+    },
+    {
+      "epoch": 1.4167135094622447,
+      "grad_norm": 0.23682500422000885,
+      "learning_rate": 8.756756756756757e-05,
+      "loss": 0.3533,
+      "step": 4726
+    },
+    {
+      "epoch": 1.417013303353944,
+      "grad_norm": 0.20653976500034332,
+      "learning_rate": 8.75225225225225e-05,
+      "loss": 0.3365,
+      "step": 4727
+    },
+    {
+      "epoch": 1.4173130972456436,
+      "grad_norm": 0.2318013608455658,
+      "learning_rate": 8.747747747747747e-05,
+      "loss": 0.3705,
+      "step": 4728
+    },
+    {
+      "epoch": 1.4176128911373431,
+      "grad_norm": 0.2127133309841156,
+      "learning_rate": 8.743243243243243e-05,
+      "loss": 0.3496,
+      "step": 4729
+    },
+    {
+      "epoch": 1.4179126850290427,
+      "grad_norm": 0.2204805463552475,
+      "learning_rate": 8.738738738738738e-05,
+      "loss": 0.3477,
+      "step": 4730
+    },
+    {
+      "epoch": 1.418212478920742,
+      "grad_norm": 0.21495847404003143,
+      "learning_rate": 8.734234234234233e-05,
+      "loss": 0.3638,
+      "step": 4731
+    },
+    {
+      "epoch": 1.4185122728124415,
+      "grad_norm": 0.21796728670597076,
+      "learning_rate": 8.72972972972973e-05,
+      "loss": 0.3608,
+      "step": 4732
+    },
+    {
+      "epoch": 1.4188120667041408,
+      "grad_norm": 0.2098185420036316,
+      "learning_rate": 8.725225225225225e-05,
+      "loss": 0.3585,
+      "step": 4733
+    },
+    {
+      "epoch": 1.4191118605958404,
+      "grad_norm": 0.2242172807455063,
+      "learning_rate": 8.72072072072072e-05,
+      "loss": 0.3719,
+      "step": 4734
+    },
+    {
+      "epoch": 1.41941165448754,
+      "grad_norm": 0.2186833620071411,
+      "learning_rate": 8.716216216216215e-05,
+      "loss": 0.3833,
+      "step": 4735
+    },
+    {
+      "epoch": 1.4197114483792392,
+      "grad_norm": 0.1976839154958725,
+      "learning_rate": 8.711711711711711e-05,
+      "loss": 0.3142,
+      "step": 4736
+    },
+    {
+      "epoch": 1.4200112422709388,
+      "grad_norm": 0.19263608753681183,
+      "learning_rate": 8.707207207207207e-05,
+      "loss": 0.3205,
+      "step": 4737
+    },
+    {
+      "epoch": 1.420311036162638,
+      "grad_norm": 0.21386316418647766,
+      "learning_rate": 8.702702702702701e-05,
+      "loss": 0.344,
+      "step": 4738
+    },
+    {
+      "epoch": 1.4206108300543376,
+      "grad_norm": 0.20507782697677612,
+      "learning_rate": 8.698198198198198e-05,
+      "loss": 0.3434,
+      "step": 4739
+    },
+    {
+      "epoch": 1.4209106239460372,
+      "grad_norm": 0.20067331194877625,
+      "learning_rate": 8.693693693693694e-05,
+      "loss": 0.3435,
+      "step": 4740
+    },
+    {
+      "epoch": 1.4212104178377365,
+      "grad_norm": 0.20949102938175201,
+      "learning_rate": 8.689189189189188e-05,
+      "loss": 0.358,
+      "step": 4741
+    },
+    {
+      "epoch": 1.421510211729436,
+      "grad_norm": 0.21897202730178833,
+      "learning_rate": 8.684684684684684e-05,
+      "loss": 0.357,
+      "step": 4742
+    },
+    {
+      "epoch": 1.4218100056211354,
+      "grad_norm": 0.2232789844274521,
+      "learning_rate": 8.68018018018018e-05,
+      "loss": 0.3495,
+      "step": 4743
+    },
+    {
+      "epoch": 1.422109799512835,
+      "grad_norm": 0.20372246205806732,
+      "learning_rate": 8.675675675675674e-05,
+      "loss": 0.3285,
+      "step": 4744
+    },
+    {
+      "epoch": 1.4224095934045344,
+      "grad_norm": 0.21029895544052124,
+      "learning_rate": 8.67117117117117e-05,
+      "loss": 0.3475,
+      "step": 4745
+    },
+    {
+      "epoch": 1.422709387296234,
+      "grad_norm": 0.2134687602519989,
+      "learning_rate": 8.666666666666665e-05,
+      "loss": 0.3529,
+      "step": 4746
+    },
+    {
+      "epoch": 1.4230091811879333,
+      "grad_norm": 0.21589897572994232,
+      "learning_rate": 8.662162162162162e-05,
+      "loss": 0.3446,
+      "step": 4747
+    },
+    {
+      "epoch": 1.4233089750796328,
+      "grad_norm": 0.22031275928020477,
+      "learning_rate": 8.657657657657657e-05,
+      "loss": 0.3759,
+      "step": 4748
+    },
+    {
+      "epoch": 1.4236087689713322,
+      "grad_norm": 0.20217642188072205,
+      "learning_rate": 8.653153153153152e-05,
+      "loss": 0.3324,
+      "step": 4749
+    },
+    {
+      "epoch": 1.4239085628630317,
+      "grad_norm": 0.22420530021190643,
+      "learning_rate": 8.648648648648648e-05,
+      "loss": 0.3884,
+      "step": 4750
+    },
+    {
+      "epoch": 1.4242083567547312,
+      "grad_norm": 0.2160356044769287,
+      "learning_rate": 8.644144144144143e-05,
+      "loss": 0.3546,
+      "step": 4751
+    },
+    {
+      "epoch": 1.4245081506464305,
+      "grad_norm": 0.2071686089038849,
+      "learning_rate": 8.639639639639638e-05,
+      "loss": 0.3741,
+      "step": 4752
+    },
+    {
+      "epoch": 1.42480794453813,
+      "grad_norm": 0.22385691106319427,
+      "learning_rate": 8.635135135135135e-05,
+      "loss": 0.3441,
+      "step": 4753
+    },
+    {
+      "epoch": 1.4251077384298294,
+      "grad_norm": 0.20757465064525604,
+      "learning_rate": 8.630630630630631e-05,
+      "loss": 0.3466,
+      "step": 4754
+    },
+    {
+      "epoch": 1.425407532321529,
+      "grad_norm": 0.20419491827487946,
+      "learning_rate": 8.626126126126125e-05,
+      "loss": 0.3246,
+      "step": 4755
+    },
+    {
+      "epoch": 1.4257073262132285,
+      "grad_norm": 0.2099572867155075,
+      "learning_rate": 8.621621621621621e-05,
+      "loss": 0.3642,
+      "step": 4756
+    },
+    {
+      "epoch": 1.4260071201049278,
+      "grad_norm": 0.21940596401691437,
+      "learning_rate": 8.617117117117118e-05,
+      "loss": 0.3569,
+      "step": 4757
+    },
+    {
+      "epoch": 1.4263069139966273,
+      "grad_norm": 0.21012279391288757,
+      "learning_rate": 8.612612612612611e-05,
+      "loss": 0.3658,
+      "step": 4758
+    },
+    {
+      "epoch": 1.4266067078883267,
+      "grad_norm": 0.20855596661567688,
+      "learning_rate": 8.608108108108108e-05,
+      "loss": 0.3697,
+      "step": 4759
+    },
+    {
+      "epoch": 1.4269065017800262,
+      "grad_norm": 0.22132547199726105,
+      "learning_rate": 8.603603603603603e-05,
+      "loss": 0.348,
+      "step": 4760
+    },
+    {
+      "epoch": 1.4272062956717257,
+      "grad_norm": 0.2173125445842743,
+      "learning_rate": 8.599099099099098e-05,
+      "loss": 0.3516,
+      "step": 4761
+    },
+    {
+      "epoch": 1.427506089563425,
+      "grad_norm": 0.2090543806552887,
+      "learning_rate": 8.594594594594594e-05,
+      "loss": 0.3521,
+      "step": 4762
+    },
+    {
+      "epoch": 1.4278058834551246,
+      "grad_norm": 0.21249796450138092,
+      "learning_rate": 8.590090090090089e-05,
+      "loss": 0.3688,
+      "step": 4763
+    },
+    {
+      "epoch": 1.428105677346824,
+      "grad_norm": 0.22098830342292786,
+      "learning_rate": 8.585585585585586e-05,
+      "loss": 0.3585,
+      "step": 4764
+    },
+    {
+      "epoch": 1.4284054712385235,
+      "grad_norm": 0.21827562153339386,
+      "learning_rate": 8.58108108108108e-05,
+      "loss": 0.3416,
+      "step": 4765
+    },
+    {
+      "epoch": 1.428705265130223,
+      "grad_norm": 0.22327540814876556,
+      "learning_rate": 8.576576576576576e-05,
+      "loss": 0.3714,
+      "step": 4766
+    },
+    {
+      "epoch": 1.4290050590219225,
+      "grad_norm": 0.20923393964767456,
+      "learning_rate": 8.572072072072072e-05,
+      "loss": 0.3596,
+      "step": 4767
+    },
+    {
+      "epoch": 1.4293048529136219,
+      "grad_norm": 0.22230814397335052,
+      "learning_rate": 8.567567567567567e-05,
+      "loss": 0.362,
+      "step": 4768
+    },
+    {
+      "epoch": 1.4296046468053214,
+      "grad_norm": 0.2018887847661972,
+      "learning_rate": 8.563063063063062e-05,
+      "loss": 0.3394,
+      "step": 4769
+    },
+    {
+      "epoch": 1.4299044406970207,
+      "grad_norm": 0.22293516993522644,
+      "learning_rate": 8.558558558558558e-05,
+      "loss": 0.35,
+      "step": 4770
+    },
+    {
+      "epoch": 1.4302042345887203,
+      "grad_norm": 0.20164437592029572,
+      "learning_rate": 8.554054054054052e-05,
+      "loss": 0.3486,
+      "step": 4771
+    },
+    {
+      "epoch": 1.4305040284804198,
+      "grad_norm": 0.22350816428661346,
+      "learning_rate": 8.549549549549548e-05,
+      "loss": 0.3784,
+      "step": 4772
+    },
+    {
+      "epoch": 1.4308038223721191,
+      "grad_norm": 0.2070097029209137,
+      "learning_rate": 8.545045045045045e-05,
+      "loss": 0.3488,
+      "step": 4773
+    },
+    {
+      "epoch": 1.4311036162638187,
+      "grad_norm": 0.20798823237419128,
+      "learning_rate": 8.540540540540539e-05,
+      "loss": 0.3468,
+      "step": 4774
+    },
+    {
+      "epoch": 1.431403410155518,
+      "grad_norm": 0.21909308433532715,
+      "learning_rate": 8.536036036036035e-05,
+      "loss": 0.3654,
+      "step": 4775
+    },
+    {
+      "epoch": 1.4317032040472175,
+      "grad_norm": 0.1986744999885559,
+      "learning_rate": 8.531531531531531e-05,
+      "loss": 0.3392,
+      "step": 4776
+    },
+    {
+      "epoch": 1.432002997938917,
+      "grad_norm": 0.2093539983034134,
+      "learning_rate": 8.527027027027026e-05,
+      "loss": 0.3651,
+      "step": 4777
+    },
+    {
+      "epoch": 1.4323027918306164,
+      "grad_norm": 0.20547617971897125,
+      "learning_rate": 8.522522522522521e-05,
+      "loss": 0.3425,
+      "step": 4778
+    },
+    {
+      "epoch": 1.432602585722316,
+      "grad_norm": 0.21190671622753143,
+      "learning_rate": 8.518018018018018e-05,
+      "loss": 0.3336,
+      "step": 4779
+    },
+    {
+      "epoch": 1.4329023796140152,
+      "grad_norm": 0.20590968430042267,
+      "learning_rate": 8.513513513513513e-05,
+      "loss": 0.3588,
+      "step": 4780
+    },
+    {
+      "epoch": 1.4332021735057148,
+      "grad_norm": 0.20922647416591644,
+      "learning_rate": 8.509009009009008e-05,
+      "loss": 0.3222,
+      "step": 4781
+    },
+    {
+      "epoch": 1.4335019673974143,
+      "grad_norm": 0.209987074136734,
+      "learning_rate": 8.504504504504503e-05,
+      "loss": 0.3416,
+      "step": 4782
+    },
+    {
+      "epoch": 1.4338017612891139,
+      "grad_norm": 0.2170223444700241,
+      "learning_rate": 8.499999999999999e-05,
+      "loss": 0.3554,
+      "step": 4783
+    },
+    {
+      "epoch": 1.4341015551808132,
+      "grad_norm": 0.2070566564798355,
+      "learning_rate": 8.495495495495496e-05,
+      "loss": 0.3472,
+      "step": 4784
+    },
+    {
+      "epoch": 1.4344013490725127,
+      "grad_norm": 0.222892165184021,
+      "learning_rate": 8.490990990990989e-05,
+      "loss": 0.3625,
+      "step": 4785
+    },
+    {
+      "epoch": 1.434701142964212,
+      "grad_norm": 0.21506567299365997,
+      "learning_rate": 8.486486486486486e-05,
+      "loss": 0.3613,
+      "step": 4786
+    },
+    {
+      "epoch": 1.4350009368559116,
+      "grad_norm": 0.19889682531356812,
+      "learning_rate": 8.481981981981982e-05,
+      "loss": 0.3184,
+      "step": 4787
+    },
+    {
+      "epoch": 1.435300730747611,
+      "grad_norm": 0.2095138281583786,
+      "learning_rate": 8.477477477477476e-05,
+      "loss": 0.3507,
+      "step": 4788
+    },
+    {
+      "epoch": 1.4356005246393104,
+      "grad_norm": 0.21511593461036682,
+      "learning_rate": 8.472972972972972e-05,
+      "loss": 0.3577,
+      "step": 4789
+    },
+    {
+      "epoch": 1.43590031853101,
+      "grad_norm": 0.21792061626911163,
+      "learning_rate": 8.468468468468469e-05,
+      "loss": 0.3768,
+      "step": 4790
+    },
+    {
+      "epoch": 1.4362001124227093,
+      "grad_norm": 0.22104412317276,
+      "learning_rate": 8.463963963963962e-05,
+      "loss": 0.4004,
+      "step": 4791
+    },
+    {
+      "epoch": 1.4364999063144088,
+      "grad_norm": 0.21523074805736542,
+      "learning_rate": 8.459459459459459e-05,
+      "loss": 0.3608,
+      "step": 4792
+    },
+    {
+      "epoch": 1.4367997002061084,
+      "grad_norm": 0.22235362231731415,
+      "learning_rate": 8.454954954954955e-05,
+      "loss": 0.3556,
+      "step": 4793
+    },
+    {
+      "epoch": 1.4370994940978077,
+      "grad_norm": 0.20765119791030884,
+      "learning_rate": 8.45045045045045e-05,
+      "loss": 0.3514,
+      "step": 4794
+    },
+    {
+      "epoch": 1.4373992879895072,
+      "grad_norm": 0.2087615430355072,
+      "learning_rate": 8.445945945945945e-05,
+      "loss": 0.3556,
+      "step": 4795
+    },
+    {
+      "epoch": 1.4376990818812065,
+      "grad_norm": 0.20913684368133545,
+      "learning_rate": 8.44144144144144e-05,
+      "loss": 0.3271,
+      "step": 4796
+    },
+    {
+      "epoch": 1.437998875772906,
+      "grad_norm": 0.20327967405319214,
+      "learning_rate": 8.436936936936936e-05,
+      "loss": 0.3427,
+      "step": 4797
+    },
+    {
+      "epoch": 1.4382986696646056,
+      "grad_norm": 0.21335206925868988,
+      "learning_rate": 8.432432432432432e-05,
+      "loss": 0.3577,
+      "step": 4798
+    },
+    {
+      "epoch": 1.4385984635563052,
+      "grad_norm": 0.21867401897907257,
+      "learning_rate": 8.427927927927927e-05,
+      "loss": 0.3687,
+      "step": 4799
+    },
+    {
+      "epoch": 1.4388982574480045,
+      "grad_norm": 0.21505360305309296,
+      "learning_rate": 8.423423423423423e-05,
+      "loss": 0.3362,
+      "step": 4800
+    },
+    {
+      "epoch": 1.439198051339704,
+      "grad_norm": 0.20245307683944702,
+      "learning_rate": 8.418918918918919e-05,
+      "loss": 0.3432,
+      "step": 4801
+    },
+    {
+      "epoch": 1.4394978452314033,
+      "grad_norm": 0.20703735947608948,
+      "learning_rate": 8.414414414414413e-05,
+      "loss": 0.34,
+      "step": 4802
+    },
+    {
+      "epoch": 1.4397976391231029,
+      "grad_norm": 0.20130011439323425,
+      "learning_rate": 8.40990990990991e-05,
+      "loss": 0.3335,
+      "step": 4803
+    },
+    {
+      "epoch": 1.4400974330148024,
+      "grad_norm": 0.20115235447883606,
+      "learning_rate": 8.405405405405406e-05,
+      "loss": 0.3479,
+      "step": 4804
+    },
+    {
+      "epoch": 1.4403972269065017,
+      "grad_norm": 0.20624947547912598,
+      "learning_rate": 8.4009009009009e-05,
+      "loss": 0.3533,
+      "step": 4805
+    },
+    {
+      "epoch": 1.4406970207982013,
+      "grad_norm": 0.19826063513755798,
+      "learning_rate": 8.396396396396396e-05,
+      "loss": 0.3412,
+      "step": 4806
+    },
+    {
+      "epoch": 1.4409968146899006,
+      "grad_norm": 0.21055233478546143,
+      "learning_rate": 8.391891891891891e-05,
+      "loss": 0.3611,
+      "step": 4807
+    },
+    {
+      "epoch": 1.4412966085816001,
+      "grad_norm": 0.2094549536705017,
+      "learning_rate": 8.387387387387386e-05,
+      "loss": 0.3353,
+      "step": 4808
+    },
+    {
+      "epoch": 1.4415964024732997,
+      "grad_norm": 0.20850618183612823,
+      "learning_rate": 8.382882882882882e-05,
+      "loss": 0.3345,
+      "step": 4809
+    },
+    {
+      "epoch": 1.441896196364999,
+      "grad_norm": 0.2107923924922943,
+      "learning_rate": 8.378378378378377e-05,
+      "loss": 0.3531,
+      "step": 4810
+    },
+    {
+      "epoch": 1.4421959902566985,
+      "grad_norm": 0.21351805329322815,
+      "learning_rate": 8.373873873873874e-05,
+      "loss": 0.3561,
+      "step": 4811
+    },
+    {
+      "epoch": 1.4424957841483979,
+      "grad_norm": 0.21343296766281128,
+      "learning_rate": 8.369369369369369e-05,
+      "loss": 0.3648,
+      "step": 4812
+    },
+    {
+      "epoch": 1.4427955780400974,
+      "grad_norm": 0.20149439573287964,
+      "learning_rate": 8.364864864864864e-05,
+      "loss": 0.327,
+      "step": 4813
+    },
+    {
+      "epoch": 1.443095371931797,
+      "grad_norm": 0.20820508897304535,
+      "learning_rate": 8.36036036036036e-05,
+      "loss": 0.3756,
+      "step": 4814
+    },
+    {
+      "epoch": 1.4433951658234965,
+      "grad_norm": 0.21000804007053375,
+      "learning_rate": 8.355855855855855e-05,
+      "loss": 0.3539,
+      "step": 4815
+    },
+    {
+      "epoch": 1.4436949597151958,
+      "grad_norm": 0.21678487956523895,
+      "learning_rate": 8.35135135135135e-05,
+      "loss": 0.3662,
+      "step": 4816
+    },
+    {
+      "epoch": 1.4439947536068953,
+      "grad_norm": 0.2124273180961609,
+      "learning_rate": 8.346846846846847e-05,
+      "loss": 0.3702,
+      "step": 4817
+    },
+    {
+      "epoch": 1.4442945474985946,
+      "grad_norm": 0.2035249024629593,
+      "learning_rate": 8.34234234234234e-05,
+      "loss": 0.3401,
+      "step": 4818
+    },
+    {
+      "epoch": 1.4445943413902942,
+      "grad_norm": 0.2147194743156433,
+      "learning_rate": 8.337837837837837e-05,
+      "loss": 0.3565,
+      "step": 4819
+    },
+    {
+      "epoch": 1.4448941352819937,
+      "grad_norm": 0.20793719589710236,
+      "learning_rate": 8.333333333333333e-05,
+      "loss": 0.3507,
+      "step": 4820
+    },
+    {
+      "epoch": 1.445193929173693,
+      "grad_norm": 0.20299087464809418,
+      "learning_rate": 8.328828828828827e-05,
+      "loss": 0.3528,
+      "step": 4821
+    },
+    {
+      "epoch": 1.4454937230653926,
+      "grad_norm": 0.20743577182292938,
+      "learning_rate": 8.324324324324323e-05,
+      "loss": 0.3585,
+      "step": 4822
+    },
+    {
+      "epoch": 1.445793516957092,
+      "grad_norm": 0.21002289652824402,
+      "learning_rate": 8.31981981981982e-05,
+      "loss": 0.3496,
+      "step": 4823
+    },
+    {
+      "epoch": 1.4460933108487914,
+      "grad_norm": 0.2117779701948166,
+      "learning_rate": 8.315315315315315e-05,
+      "loss": 0.3683,
+      "step": 4824
+    },
+    {
+      "epoch": 1.446393104740491,
+      "grad_norm": 0.21782147884368896,
+      "learning_rate": 8.31081081081081e-05,
+      "loss": 0.3545,
+      "step": 4825
+    },
+    {
+      "epoch": 1.4466928986321903,
+      "grad_norm": 0.20695964992046356,
+      "learning_rate": 8.306306306306306e-05,
+      "loss": 0.3553,
+      "step": 4826
+    },
+    {
+      "epoch": 1.4469926925238898,
+      "grad_norm": 0.19961470365524292,
+      "learning_rate": 8.301801801801801e-05,
+      "loss": 0.3447,
+      "step": 4827
+    },
+    {
+      "epoch": 1.4472924864155892,
+      "grad_norm": 0.2148684710264206,
+      "learning_rate": 8.297297297297297e-05,
+      "loss": 0.3538,
+      "step": 4828
+    },
+    {
+      "epoch": 1.4475922803072887,
+      "grad_norm": 0.21752296388149261,
+      "learning_rate": 8.292792792792792e-05,
+      "loss": 0.3677,
+      "step": 4829
+    },
+    {
+      "epoch": 1.4478920741989882,
+      "grad_norm": 0.20284780859947205,
+      "learning_rate": 8.288288288288287e-05,
+      "loss": 0.3581,
+      "step": 4830
+    },
+    {
+      "epoch": 1.4481918680906878,
+      "grad_norm": 0.21310214698314667,
+      "learning_rate": 8.283783783783784e-05,
+      "loss": 0.3538,
+      "step": 4831
+    },
+    {
+      "epoch": 1.448491661982387,
+      "grad_norm": 0.20693115890026093,
+      "learning_rate": 8.279279279279278e-05,
+      "loss": 0.3669,
+      "step": 4832
+    },
+    {
+      "epoch": 1.4487914558740866,
+      "grad_norm": 0.2130792886018753,
+      "learning_rate": 8.274774774774774e-05,
+      "loss": 0.3514,
+      "step": 4833
+    },
+    {
+      "epoch": 1.449091249765786,
+      "grad_norm": 0.21330896019935608,
+      "learning_rate": 8.27027027027027e-05,
+      "loss": 0.3523,
+      "step": 4834
+    },
+    {
+      "epoch": 1.4493910436574855,
+      "grad_norm": 0.20298290252685547,
+      "learning_rate": 8.265765765765764e-05,
+      "loss": 0.3388,
+      "step": 4835
+    },
+    {
+      "epoch": 1.449690837549185,
+      "grad_norm": 0.20353178679943085,
+      "learning_rate": 8.26126126126126e-05,
+      "loss": 0.3461,
+      "step": 4836
+    },
+    {
+      "epoch": 1.4499906314408844,
+      "grad_norm": 0.2044645994901657,
+      "learning_rate": 8.256756756756757e-05,
+      "loss": 0.3302,
+      "step": 4837
+    },
+    {
+      "epoch": 1.450290425332584,
+      "grad_norm": 0.2172108143568039,
+      "learning_rate": 8.25225225225225e-05,
+      "loss": 0.3573,
+      "step": 4838
+    },
+    {
+      "epoch": 1.4505902192242832,
+      "grad_norm": 0.20277266204357147,
+      "learning_rate": 8.247747747747747e-05,
+      "loss": 0.345,
+      "step": 4839
+    },
+    {
+      "epoch": 1.4508900131159828,
+      "grad_norm": 0.22137466073036194,
+      "learning_rate": 8.243243243243243e-05,
+      "loss": 0.3754,
+      "step": 4840
+    },
+    {
+      "epoch": 1.4511898070076823,
+      "grad_norm": 0.20912005007266998,
+      "learning_rate": 8.238738738738738e-05,
+      "loss": 0.353,
+      "step": 4841
+    },
+    {
+      "epoch": 1.4514896008993816,
+      "grad_norm": 0.1990968883037567,
+      "learning_rate": 8.234234234234233e-05,
+      "loss": 0.3336,
+      "step": 4842
+    },
+    {
+      "epoch": 1.4517893947910812,
+      "grad_norm": 0.21815185248851776,
+      "learning_rate": 8.229729729729728e-05,
+      "loss": 0.3884,
+      "step": 4843
+    },
+    {
+      "epoch": 1.4520891886827805,
+      "grad_norm": 0.21321304142475128,
+      "learning_rate": 8.225225225225225e-05,
+      "loss": 0.3624,
+      "step": 4844
+    },
+    {
+      "epoch": 1.45238898257448,
+      "grad_norm": 0.20276996493339539,
+      "learning_rate": 8.220720720720721e-05,
+      "loss": 0.342,
+      "step": 4845
+    },
+    {
+      "epoch": 1.4526887764661796,
+      "grad_norm": 0.19751514494419098,
+      "learning_rate": 8.216216216216215e-05,
+      "loss": 0.3381,
+      "step": 4846
+    },
+    {
+      "epoch": 1.452988570357879,
+      "grad_norm": 0.2045748233795166,
+      "learning_rate": 8.211711711711711e-05,
+      "loss": 0.3579,
+      "step": 4847
+    },
+    {
+      "epoch": 1.4532883642495784,
+      "grad_norm": 0.2179010659456253,
+      "learning_rate": 8.207207207207208e-05,
+      "loss": 0.3746,
+      "step": 4848
+    },
+    {
+      "epoch": 1.453588158141278,
+      "grad_norm": 0.20252852141857147,
+      "learning_rate": 8.202702702702701e-05,
+      "loss": 0.3512,
+      "step": 4849
+    },
+    {
+      "epoch": 1.4538879520329773,
+      "grad_norm": 0.2126975953578949,
+      "learning_rate": 8.198198198198198e-05,
+      "loss": 0.3473,
+      "step": 4850
+    },
+    {
+      "epoch": 1.4541877459246768,
+      "grad_norm": 0.2115095853805542,
+      "learning_rate": 8.193693693693694e-05,
+      "loss": 0.376,
+      "step": 4851
+    },
+    {
+      "epoch": 1.4544875398163764,
+      "grad_norm": 0.21630194783210754,
+      "learning_rate": 8.189189189189188e-05,
+      "loss": 0.3726,
+      "step": 4852
+    },
+    {
+      "epoch": 1.4547873337080757,
+      "grad_norm": 0.21030595898628235,
+      "learning_rate": 8.184684684684684e-05,
+      "loss": 0.363,
+      "step": 4853
+    },
+    {
+      "epoch": 1.4550871275997752,
+      "grad_norm": 0.20921146869659424,
+      "learning_rate": 8.18018018018018e-05,
+      "loss": 0.3435,
+      "step": 4854
+    },
+    {
+      "epoch": 1.4553869214914745,
+      "grad_norm": 0.22109493613243103,
+      "learning_rate": 8.175675675675674e-05,
+      "loss": 0.3523,
+      "step": 4855
+    },
+    {
+      "epoch": 1.455686715383174,
+      "grad_norm": 0.203207865357399,
+      "learning_rate": 8.17117117117117e-05,
+      "loss": 0.3276,
+      "step": 4856
+    },
+    {
+      "epoch": 1.4559865092748736,
+      "grad_norm": 0.21486714482307434,
+      "learning_rate": 8.166666666666665e-05,
+      "loss": 0.3696,
+      "step": 4857
+    },
+    {
+      "epoch": 1.456286303166573,
+      "grad_norm": 0.20264849066734314,
+      "learning_rate": 8.162162162162162e-05,
+      "loss": 0.3339,
+      "step": 4858
+    },
+    {
+      "epoch": 1.4565860970582725,
+      "grad_norm": 0.20054593682289124,
+      "learning_rate": 8.157657657657657e-05,
+      "loss": 0.3353,
+      "step": 4859
+    },
+    {
+      "epoch": 1.4568858909499718,
+      "grad_norm": 0.20756486058235168,
+      "learning_rate": 8.153153153153152e-05,
+      "loss": 0.3571,
+      "step": 4860
+    },
+    {
+      "epoch": 1.4571856848416713,
+      "grad_norm": 0.21069984138011932,
+      "learning_rate": 8.148648648648648e-05,
+      "loss": 0.3495,
+      "step": 4861
+    },
+    {
+      "epoch": 1.4574854787333709,
+      "grad_norm": 0.21038687229156494,
+      "learning_rate": 8.144144144144145e-05,
+      "loss": 0.3639,
+      "step": 4862
+    },
+    {
+      "epoch": 1.4577852726250702,
+      "grad_norm": 0.21815302968025208,
+      "learning_rate": 8.139639639639638e-05,
+      "loss": 0.3768,
+      "step": 4863
+    },
+    {
+      "epoch": 1.4580850665167697,
+      "grad_norm": 0.20300470292568207,
+      "learning_rate": 8.135135135135135e-05,
+      "loss": 0.349,
+      "step": 4864
+    },
+    {
+      "epoch": 1.458384860408469,
+      "grad_norm": 0.21326471865177155,
+      "learning_rate": 8.130630630630631e-05,
+      "loss": 0.364,
+      "step": 4865
+    },
+    {
+      "epoch": 1.4586846543001686,
+      "grad_norm": 0.2107071727514267,
+      "learning_rate": 8.126126126126125e-05,
+      "loss": 0.3701,
+      "step": 4866
+    },
+    {
+      "epoch": 1.4589844481918681,
+      "grad_norm": 0.21025468409061432,
+      "learning_rate": 8.121621621621621e-05,
+      "loss": 0.3736,
+      "step": 4867
+    },
+    {
+      "epoch": 1.4592842420835677,
+      "grad_norm": 0.2088591605424881,
+      "learning_rate": 8.117117117117116e-05,
+      "loss": 0.3498,
+      "step": 4868
+    },
+    {
+      "epoch": 1.459584035975267,
+      "grad_norm": 0.22014079988002777,
+      "learning_rate": 8.112612612612611e-05,
+      "loss": 0.3635,
+      "step": 4869
+    },
+    {
+      "epoch": 1.4598838298669665,
+      "grad_norm": 0.20154501497745514,
+      "learning_rate": 8.108108108108108e-05,
+      "loss": 0.3429,
+      "step": 4870
+    },
+    {
+      "epoch": 1.4601836237586658,
+      "grad_norm": 0.2102486938238144,
+      "learning_rate": 8.103603603603603e-05,
+      "loss": 0.3574,
+      "step": 4871
+    },
+    {
+      "epoch": 1.4604834176503654,
+      "grad_norm": 0.2018607258796692,
+      "learning_rate": 8.099099099099098e-05,
+      "loss": 0.3337,
+      "step": 4872
+    },
+    {
+      "epoch": 1.460783211542065,
+      "grad_norm": 0.20702728629112244,
+      "learning_rate": 8.094594594594594e-05,
+      "loss": 0.3473,
+      "step": 4873
+    },
+    {
+      "epoch": 1.4610830054337642,
+      "grad_norm": 0.20739267766475677,
+      "learning_rate": 8.090090090090089e-05,
+      "loss": 0.3506,
+      "step": 4874
+    },
+    {
+      "epoch": 1.4613827993254638,
+      "grad_norm": 0.20889143645763397,
+      "learning_rate": 8.085585585585586e-05,
+      "loss": 0.3711,
+      "step": 4875
+    },
+    {
+      "epoch": 1.461682593217163,
+      "grad_norm": 0.20718500018119812,
+      "learning_rate": 8.08108108108108e-05,
+      "loss": 0.3618,
+      "step": 4876
+    },
+    {
+      "epoch": 1.4619823871088626,
+      "grad_norm": 0.22436094284057617,
+      "learning_rate": 8.076576576576576e-05,
+      "loss": 0.3752,
+      "step": 4877
+    },
+    {
+      "epoch": 1.4622821810005622,
+      "grad_norm": 0.21661464869976044,
+      "learning_rate": 8.072072072072072e-05,
+      "loss": 0.3945,
+      "step": 4878
+    },
+    {
+      "epoch": 1.4625819748922615,
+      "grad_norm": 0.21557655930519104,
+      "learning_rate": 8.067567567567566e-05,
+      "loss": 0.33,
+      "step": 4879
+    },
+    {
+      "epoch": 1.462881768783961,
+      "grad_norm": 0.20167309045791626,
+      "learning_rate": 8.063063063063062e-05,
+      "loss": 0.3333,
+      "step": 4880
+    },
+    {
+      "epoch": 1.4631815626756604,
+      "grad_norm": 0.20474180579185486,
+      "learning_rate": 8.058558558558558e-05,
+      "loss": 0.3634,
+      "step": 4881
+    },
+    {
+      "epoch": 1.46348135656736,
+      "grad_norm": 0.2092469483613968,
+      "learning_rate": 8.054054054054052e-05,
+      "loss": 0.3638,
+      "step": 4882
+    },
+    {
+      "epoch": 1.4637811504590594,
+      "grad_norm": 0.21168501675128937,
+      "learning_rate": 8.049549549549549e-05,
+      "loss": 0.3653,
+      "step": 4883
+    },
+    {
+      "epoch": 1.464080944350759,
+      "grad_norm": 0.2195328325033188,
+      "learning_rate": 8.045045045045045e-05,
+      "loss": 0.3888,
+      "step": 4884
+    },
+    {
+      "epoch": 1.4643807382424583,
+      "grad_norm": 0.2088628113269806,
+      "learning_rate": 8.04054054054054e-05,
+      "loss": 0.3611,
+      "step": 4885
+    },
+    {
+      "epoch": 1.4646805321341578,
+      "grad_norm": 0.20401327311992645,
+      "learning_rate": 8.036036036036035e-05,
+      "loss": 0.3487,
+      "step": 4886
+    },
+    {
+      "epoch": 1.4649803260258571,
+      "grad_norm": 0.2168956845998764,
+      "learning_rate": 8.031531531531531e-05,
+      "loss": 0.3709,
+      "step": 4887
+    },
+    {
+      "epoch": 1.4652801199175567,
+      "grad_norm": 0.199428528547287,
+      "learning_rate": 8.027027027027026e-05,
+      "loss": 0.3379,
+      "step": 4888
+    },
+    {
+      "epoch": 1.4655799138092562,
+      "grad_norm": 0.20674806833267212,
+      "learning_rate": 8.022522522522521e-05,
+      "loss": 0.3648,
+      "step": 4889
+    },
+    {
+      "epoch": 1.4658797077009555,
+      "grad_norm": 0.1945922076702118,
+      "learning_rate": 8.018018018018018e-05,
+      "loss": 0.3477,
+      "step": 4890
+    },
+    {
+      "epoch": 1.466179501592655,
+      "grad_norm": 0.2134675234556198,
+      "learning_rate": 8.013513513513513e-05,
+      "loss": 0.3796,
+      "step": 4891
+    },
+    {
+      "epoch": 1.4664792954843544,
+      "grad_norm": 0.2022518515586853,
+      "learning_rate": 8.009009009009009e-05,
+      "loss": 0.3458,
+      "step": 4892
+    },
+    {
+      "epoch": 1.466779089376054,
+      "grad_norm": 0.21644069254398346,
+      "learning_rate": 8.004504504504503e-05,
+      "loss": 0.3625,
+      "step": 4893
+    },
+    {
+      "epoch": 1.4670788832677535,
+      "grad_norm": 0.21158376336097717,
+      "learning_rate": 7.999999999999999e-05,
+      "loss": 0.3616,
+      "step": 4894
+    },
+    {
+      "epoch": 1.4673786771594528,
+      "grad_norm": 0.21450527012348175,
+      "learning_rate": 7.995495495495496e-05,
+      "loss": 0.3662,
+      "step": 4895
+    },
+    {
+      "epoch": 1.4676784710511523,
+      "grad_norm": 0.21180300414562225,
+      "learning_rate": 7.99099099099099e-05,
+      "loss": 0.3544,
+      "step": 4896
+    },
+    {
+      "epoch": 1.4679782649428517,
+      "grad_norm": 0.20244859158992767,
+      "learning_rate": 7.986486486486486e-05,
+      "loss": 0.3475,
+      "step": 4897
+    },
+    {
+      "epoch": 1.4682780588345512,
+      "grad_norm": 0.21008580923080444,
+      "learning_rate": 7.981981981981982e-05,
+      "loss": 0.3545,
+      "step": 4898
+    },
+    {
+      "epoch": 1.4685778527262507,
+      "grad_norm": 0.21382389962673187,
+      "learning_rate": 7.977477477477476e-05,
+      "loss": 0.3796,
+      "step": 4899
+    },
+    {
+      "epoch": 1.4688776466179503,
+      "grad_norm": 0.21943014860153198,
+      "learning_rate": 7.972972972972972e-05,
+      "loss": 0.3769,
+      "step": 4900
+    },
+    {
+      "epoch": 1.4691774405096496,
+      "grad_norm": 0.20606811344623566,
+      "learning_rate": 7.968468468468469e-05,
+      "loss": 0.3511,
+      "step": 4901
+    },
+    {
+      "epoch": 1.4694772344013491,
+      "grad_norm": 0.2154122292995453,
+      "learning_rate": 7.963963963963964e-05,
+      "loss": 0.367,
+      "step": 4902
+    },
+    {
+      "epoch": 1.4697770282930485,
+      "grad_norm": 0.21306243538856506,
+      "learning_rate": 7.959459459459459e-05,
+      "loss": 0.3667,
+      "step": 4903
+    },
+    {
+      "epoch": 1.470076822184748,
+      "grad_norm": 0.20610538125038147,
+      "learning_rate": 7.954954954954954e-05,
+      "loss": 0.3608,
+      "step": 4904
+    },
+    {
+      "epoch": 1.4703766160764475,
+      "grad_norm": 0.20182529091835022,
+      "learning_rate": 7.95045045045045e-05,
+      "loss": 0.3404,
+      "step": 4905
+    },
+    {
+      "epoch": 1.4706764099681469,
+      "grad_norm": 0.22032389044761658,
+      "learning_rate": 7.945945945945945e-05,
+      "loss": 0.371,
+      "step": 4906
+    },
+    {
+      "epoch": 1.4709762038598464,
+      "grad_norm": 0.22178107500076294,
+      "learning_rate": 7.94144144144144e-05,
+      "loss": 0.348,
+      "step": 4907
+    },
+    {
+      "epoch": 1.4712759977515457,
+      "grad_norm": 0.2226249873638153,
+      "learning_rate": 7.936936936936937e-05,
+      "loss": 0.3342,
+      "step": 4908
+    },
+    {
+      "epoch": 1.4715757916432453,
+      "grad_norm": 0.21837739646434784,
+      "learning_rate": 7.932432432432433e-05,
+      "loss": 0.3316,
+      "step": 4909
+    },
+    {
+      "epoch": 1.4718755855349448,
+      "grad_norm": 0.2183731645345688,
+      "learning_rate": 7.927927927927927e-05,
+      "loss": 0.3615,
+      "step": 4910
+    },
+    {
+      "epoch": 1.4721753794266441,
+      "grad_norm": 0.22412636876106262,
+      "learning_rate": 7.923423423423423e-05,
+      "loss": 0.3466,
+      "step": 4911
+    },
+    {
+      "epoch": 1.4724751733183437,
+      "grad_norm": 0.21817882359027863,
+      "learning_rate": 7.91891891891892e-05,
+      "loss": 0.3699,
+      "step": 4912
+    },
+    {
+      "epoch": 1.472774967210043,
+      "grad_norm": 0.2297516018152237,
+      "learning_rate": 7.914414414414413e-05,
+      "loss": 0.3617,
+      "step": 4913
+    },
+    {
+      "epoch": 1.4730747611017425,
+      "grad_norm": 0.21283744275569916,
+      "learning_rate": 7.90990990990991e-05,
+      "loss": 0.3578,
+      "step": 4914
+    },
+    {
+      "epoch": 1.473374554993442,
+      "grad_norm": 0.21200676262378693,
+      "learning_rate": 7.905405405405404e-05,
+      "loss": 0.3389,
+      "step": 4915
+    },
+    {
+      "epoch": 1.4736743488851416,
+      "grad_norm": 0.22450967133045197,
+      "learning_rate": 7.9009009009009e-05,
+      "loss": 0.352,
+      "step": 4916
+    },
+    {
+      "epoch": 1.473974142776841,
+      "grad_norm": 0.23047909140586853,
+      "learning_rate": 7.896396396396396e-05,
+      "loss": 0.371,
+      "step": 4917
+    },
+    {
+      "epoch": 1.4742739366685405,
+      "grad_norm": 0.2088591456413269,
+      "learning_rate": 7.891891891891891e-05,
+      "loss": 0.3543,
+      "step": 4918
+    },
+    {
+      "epoch": 1.4745737305602398,
+      "grad_norm": 0.22470034658908844,
+      "learning_rate": 7.887387387387387e-05,
+      "loss": 0.3477,
+      "step": 4919
+    },
+    {
+      "epoch": 1.4748735244519393,
+      "grad_norm": 0.22082637250423431,
+      "learning_rate": 7.882882882882882e-05,
+      "loss": 0.361,
+      "step": 4920
+    },
+    {
+      "epoch": 1.4751733183436389,
+      "grad_norm": 0.22368364036083221,
+      "learning_rate": 7.878378378378377e-05,
+      "loss": 0.3703,
+      "step": 4921
+    },
+    {
+      "epoch": 1.4754731122353382,
+      "grad_norm": 0.22301022708415985,
+      "learning_rate": 7.873873873873874e-05,
+      "loss": 0.3558,
+      "step": 4922
+    },
+    {
+      "epoch": 1.4757729061270377,
+      "grad_norm": 0.21291525661945343,
+      "learning_rate": 7.869369369369369e-05,
+      "loss": 0.3634,
+      "step": 4923
+    },
+    {
+      "epoch": 1.476072700018737,
+      "grad_norm": 0.2003738433122635,
+      "learning_rate": 7.864864864864864e-05,
+      "loss": 0.3432,
+      "step": 4924
+    },
+    {
+      "epoch": 1.4763724939104366,
+      "grad_norm": 0.21728284657001495,
+      "learning_rate": 7.86036036036036e-05,
+      "loss": 0.3527,
+      "step": 4925
+    },
+    {
+      "epoch": 1.476672287802136,
+      "grad_norm": 0.21595244109630585,
+      "learning_rate": 7.855855855855857e-05,
+      "loss": 0.3567,
+      "step": 4926
+    },
+    {
+      "epoch": 1.4769720816938354,
+      "grad_norm": 0.21208882331848145,
+      "learning_rate": 7.85135135135135e-05,
+      "loss": 0.3611,
+      "step": 4927
+    },
+    {
+      "epoch": 1.477271875585535,
+      "grad_norm": 0.21161866188049316,
+      "learning_rate": 7.846846846846847e-05,
+      "loss": 0.3587,
+      "step": 4928
+    },
+    {
+      "epoch": 1.4775716694772343,
+      "grad_norm": 0.2157314121723175,
+      "learning_rate": 7.84234234234234e-05,
+      "loss": 0.3595,
+      "step": 4929
+    },
+    {
+      "epoch": 1.4778714633689338,
+      "grad_norm": 0.22329649329185486,
+      "learning_rate": 7.837837837837837e-05,
+      "loss": 0.3563,
+      "step": 4930
+    },
+    {
+      "epoch": 1.4781712572606334,
+      "grad_norm": 0.20477770268917084,
+      "learning_rate": 7.833333333333333e-05,
+      "loss": 0.3265,
+      "step": 4931
+    },
+    {
+      "epoch": 1.478471051152333,
+      "grad_norm": 0.20577353239059448,
+      "learning_rate": 7.828828828828828e-05,
+      "loss": 0.3553,
+      "step": 4932
+    },
+    {
+      "epoch": 1.4787708450440322,
+      "grad_norm": 0.22000664472579956,
+      "learning_rate": 7.824324324324323e-05,
+      "loss": 0.3523,
+      "step": 4933
+    },
+    {
+      "epoch": 1.4790706389357318,
+      "grad_norm": 0.21802487969398499,
+      "learning_rate": 7.81981981981982e-05,
+      "loss": 0.3964,
+      "step": 4934
+    },
+    {
+      "epoch": 1.479370432827431,
+      "grad_norm": 0.21147708594799042,
+      "learning_rate": 7.815315315315315e-05,
+      "loss": 0.3463,
+      "step": 4935
+    },
+    {
+      "epoch": 1.4796702267191306,
+      "grad_norm": 0.21361719071865082,
+      "learning_rate": 7.81081081081081e-05,
+      "loss": 0.3512,
+      "step": 4936
+    },
+    {
+      "epoch": 1.4799700206108302,
+      "grad_norm": 0.20637863874435425,
+      "learning_rate": 7.806306306306306e-05,
+      "loss": 0.3357,
+      "step": 4937
+    },
+    {
+      "epoch": 1.4802698145025295,
+      "grad_norm": 0.20978131890296936,
+      "learning_rate": 7.801801801801801e-05,
+      "loss": 0.3778,
+      "step": 4938
+    },
+    {
+      "epoch": 1.480569608394229,
+      "grad_norm": 0.22668473422527313,
+      "learning_rate": 7.797297297297297e-05,
+      "loss": 0.3844,
+      "step": 4939
+    },
+    {
+      "epoch": 1.4808694022859283,
+      "grad_norm": 0.21885189414024353,
+      "learning_rate": 7.792792792792791e-05,
+      "loss": 0.3337,
+      "step": 4940
+    },
+    {
+      "epoch": 1.4811691961776279,
+      "grad_norm": 0.19497647881507874,
+      "learning_rate": 7.788288288288287e-05,
+      "loss": 0.3193,
+      "step": 4941
+    },
+    {
+      "epoch": 1.4814689900693274,
+      "grad_norm": 0.2125978320837021,
+      "learning_rate": 7.783783783783784e-05,
+      "loss": 0.3722,
+      "step": 4942
+    },
+    {
+      "epoch": 1.4817687839610267,
+      "grad_norm": 0.20573614537715912,
+      "learning_rate": 7.779279279279278e-05,
+      "loss": 0.3535,
+      "step": 4943
+    },
+    {
+      "epoch": 1.4820685778527263,
+      "grad_norm": 0.20022504031658173,
+      "learning_rate": 7.774774774774774e-05,
+      "loss": 0.358,
+      "step": 4944
+    },
+    {
+      "epoch": 1.4823683717444256,
+      "grad_norm": 0.19802720844745636,
+      "learning_rate": 7.77027027027027e-05,
+      "loss": 0.3268,
+      "step": 4945
+    },
+    {
+      "epoch": 1.4826681656361251,
+      "grad_norm": 0.2035313844680786,
+      "learning_rate": 7.765765765765764e-05,
+      "loss": 0.3432,
+      "step": 4946
+    },
+    {
+      "epoch": 1.4829679595278247,
+      "grad_norm": 0.19907894730567932,
+      "learning_rate": 7.76126126126126e-05,
+      "loss": 0.3341,
+      "step": 4947
+    },
+    {
+      "epoch": 1.4832677534195242,
+      "grad_norm": 0.22520247101783752,
+      "learning_rate": 7.756756756756757e-05,
+      "loss": 0.3673,
+      "step": 4948
+    },
+    {
+      "epoch": 1.4835675473112235,
+      "grad_norm": 0.21920591592788696,
+      "learning_rate": 7.752252252252252e-05,
+      "loss": 0.3789,
+      "step": 4949
+    },
+    {
+      "epoch": 1.483867341202923,
+      "grad_norm": 0.2152647227048874,
+      "learning_rate": 7.747747747747747e-05,
+      "loss": 0.3697,
+      "step": 4950
+    },
+    {
+      "epoch": 1.4841671350946224,
+      "grad_norm": 0.20558972656726837,
+      "learning_rate": 7.743243243243242e-05,
+      "loss": 0.3421,
+      "step": 4951
+    },
+    {
+      "epoch": 1.484466928986322,
+      "grad_norm": 0.21162059903144836,
+      "learning_rate": 7.738738738738738e-05,
+      "loss": 0.3531,
+      "step": 4952
+    },
+    {
+      "epoch": 1.4847667228780215,
+      "grad_norm": 0.20541666448116302,
+      "learning_rate": 7.734234234234233e-05,
+      "loss": 0.3557,
+      "step": 4953
+    },
+    {
+      "epoch": 1.4850665167697208,
+      "grad_norm": 0.21199767291545868,
+      "learning_rate": 7.729729729729728e-05,
+      "loss": 0.3593,
+      "step": 4954
+    },
+    {
+      "epoch": 1.4853663106614203,
+      "grad_norm": 0.21618947386741638,
+      "learning_rate": 7.725225225225225e-05,
+      "loss": 0.3663,
+      "step": 4955
+    },
+    {
+      "epoch": 1.4856661045531196,
+      "grad_norm": 0.20753389596939087,
+      "learning_rate": 7.720720720720721e-05,
+      "loss": 0.3508,
+      "step": 4956
+    },
+    {
+      "epoch": 1.4859658984448192,
+      "grad_norm": 0.22099077701568604,
+      "learning_rate": 7.716216216216215e-05,
+      "loss": 0.3516,
+      "step": 4957
+    },
+    {
+      "epoch": 1.4862656923365187,
+      "grad_norm": 0.2257046103477478,
+      "learning_rate": 7.711711711711711e-05,
+      "loss": 0.3551,
+      "step": 4958
+    },
+    {
+      "epoch": 1.486565486228218,
+      "grad_norm": 0.2128523290157318,
+      "learning_rate": 7.707207207207208e-05,
+      "loss": 0.36,
+      "step": 4959
+    },
+    {
+      "epoch": 1.4868652801199176,
+      "grad_norm": 0.2099178284406662,
+      "learning_rate": 7.702702702702701e-05,
+      "loss": 0.3645,
+      "step": 4960
+    },
+    {
+      "epoch": 1.487165074011617,
+      "grad_norm": 0.20770512521266937,
+      "learning_rate": 7.698198198198198e-05,
+      "loss": 0.36,
+      "step": 4961
+    },
+    {
+      "epoch": 1.4874648679033164,
+      "grad_norm": 0.20869390666484833,
+      "learning_rate": 7.693693693693694e-05,
+      "loss": 0.3542,
+      "step": 4962
+    },
+    {
+      "epoch": 1.487764661795016,
+      "grad_norm": 0.1958959698677063,
+      "learning_rate": 7.689189189189188e-05,
+      "loss": 0.3298,
+      "step": 4963
+    },
+    {
+      "epoch": 1.4880644556867153,
+      "grad_norm": 0.22579431533813477,
+      "learning_rate": 7.684684684684684e-05,
+      "loss": 0.3886,
+      "step": 4964
+    },
+    {
+      "epoch": 1.4883642495784148,
+      "grad_norm": 0.21450291574001312,
+      "learning_rate": 7.680180180180179e-05,
+      "loss": 0.3742,
+      "step": 4965
+    },
+    {
+      "epoch": 1.4886640434701142,
+      "grad_norm": 0.21159665286540985,
+      "learning_rate": 7.675675675675675e-05,
+      "loss": 0.362,
+      "step": 4966
+    },
+    {
+      "epoch": 1.4889638373618137,
+      "grad_norm": 0.2096509039402008,
+      "learning_rate": 7.67117117117117e-05,
+      "loss": 0.373,
+      "step": 4967
+    },
+    {
+      "epoch": 1.4892636312535132,
+      "grad_norm": 0.21596965193748474,
+      "learning_rate": 7.666666666666666e-05,
+      "loss": 0.3584,
+      "step": 4968
+    },
+    {
+      "epoch": 1.4895634251452128,
+      "grad_norm": 0.20583121478557587,
+      "learning_rate": 7.662162162162162e-05,
+      "loss": 0.344,
+      "step": 4969
+    },
+    {
+      "epoch": 1.489863219036912,
+      "grad_norm": 0.2186458855867386,
+      "learning_rate": 7.657657657657657e-05,
+      "loss": 0.3943,
+      "step": 4970
+    },
+    {
+      "epoch": 1.4901630129286116,
+      "grad_norm": 0.21535223722457886,
+      "learning_rate": 7.653153153153152e-05,
+      "loss": 0.3337,
+      "step": 4971
+    },
+    {
+      "epoch": 1.490462806820311,
+      "grad_norm": 0.21319331228733063,
+      "learning_rate": 7.648648648648648e-05,
+      "loss": 0.3697,
+      "step": 4972
+    },
+    {
+      "epoch": 1.4907626007120105,
+      "grad_norm": 0.21556057035923004,
+      "learning_rate": 7.644144144144145e-05,
+      "loss": 0.3758,
+      "step": 4973
+    },
+    {
+      "epoch": 1.49106239460371,
+      "grad_norm": 0.2040267437696457,
+      "learning_rate": 7.639639639639638e-05,
+      "loss": 0.3605,
+      "step": 4974
+    },
+    {
+      "epoch": 1.4913621884954094,
+      "grad_norm": 0.22170892357826233,
+      "learning_rate": 7.635135135135135e-05,
+      "loss": 0.3907,
+      "step": 4975
+    },
+    {
+      "epoch": 1.491661982387109,
+      "grad_norm": 0.21407882869243622,
+      "learning_rate": 7.630630630630629e-05,
+      "loss": 0.3591,
+      "step": 4976
+    },
+    {
+      "epoch": 1.4919617762788082,
+      "grad_norm": 0.20388971269130707,
+      "learning_rate": 7.626126126126125e-05,
+      "loss": 0.3491,
+      "step": 4977
+    },
+    {
+      "epoch": 1.4922615701705078,
+      "grad_norm": 0.21065190434455872,
+      "learning_rate": 7.621621621621621e-05,
+      "loss": 0.3413,
+      "step": 4978
+    },
+    {
+      "epoch": 1.4925613640622073,
+      "grad_norm": 0.21436721086502075,
+      "learning_rate": 7.617117117117116e-05,
+      "loss": 0.3576,
+      "step": 4979
+    },
+    {
+      "epoch": 1.4928611579539066,
+      "grad_norm": 0.205039381980896,
+      "learning_rate": 7.612612612612611e-05,
+      "loss": 0.3722,
+      "step": 4980
+    },
+    {
+      "epoch": 1.4931609518456062,
+      "grad_norm": 0.2189406454563141,
+      "learning_rate": 7.608108108108108e-05,
+      "loss": 0.3728,
+      "step": 4981
+    },
+    {
+      "epoch": 1.4934607457373055,
+      "grad_norm": 0.21848881244659424,
+      "learning_rate": 7.603603603603603e-05,
+      "loss": 0.3592,
+      "step": 4982
+    },
+    {
+      "epoch": 1.493760539629005,
+      "grad_norm": 0.21232780814170837,
+      "learning_rate": 7.599099099099099e-05,
+      "loss": 0.3704,
+      "step": 4983
+    },
+    {
+      "epoch": 1.4940603335207046,
+      "grad_norm": 0.2145267128944397,
+      "learning_rate": 7.594594594594594e-05,
+      "loss": 0.368,
+      "step": 4984
+    },
+    {
+      "epoch": 1.494360127412404,
+      "grad_norm": 0.22516867518424988,
+      "learning_rate": 7.590090090090089e-05,
+      "loss": 0.3801,
+      "step": 4985
+    },
+    {
+      "epoch": 1.4946599213041034,
+      "grad_norm": 0.21354909241199493,
+      "learning_rate": 7.585585585585586e-05,
+      "loss": 0.3709,
+      "step": 4986
+    },
+    {
+      "epoch": 1.494959715195803,
+      "grad_norm": 0.21226109564304352,
+      "learning_rate": 7.581081081081079e-05,
+      "loss": 0.3591,
+      "step": 4987
+    },
+    {
+      "epoch": 1.4952595090875023,
+      "grad_norm": 0.21242061257362366,
+      "learning_rate": 7.576576576576576e-05,
+      "loss": 0.3492,
+      "step": 4988
+    },
+    {
+      "epoch": 1.4955593029792018,
+      "grad_norm": 0.20486439764499664,
+      "learning_rate": 7.572072072072072e-05,
+      "loss": 0.3624,
+      "step": 4989
+    },
+    {
+      "epoch": 1.4958590968709014,
+      "grad_norm": 0.20076008141040802,
+      "learning_rate": 7.567567567567566e-05,
+      "loss": 0.3296,
+      "step": 4990
+    },
+    {
+      "epoch": 1.4961588907626007,
+      "grad_norm": 0.21576011180877686,
+      "learning_rate": 7.563063063063062e-05,
+      "loss": 0.3839,
+      "step": 4991
+    },
+    {
+      "epoch": 1.4964586846543002,
+      "grad_norm": 0.20013126730918884,
+      "learning_rate": 7.558558558558559e-05,
+      "loss": 0.3606,
+      "step": 4992
+    },
+    {
+      "epoch": 1.4967584785459995,
+      "grad_norm": 0.2180420160293579,
+      "learning_rate": 7.554054054054052e-05,
+      "loss": 0.3652,
+      "step": 4993
+    },
+    {
+      "epoch": 1.497058272437699,
+      "grad_norm": 0.21717427670955658,
+      "learning_rate": 7.549549549549549e-05,
+      "loss": 0.3481,
+      "step": 4994
+    },
+    {
+      "epoch": 1.4973580663293986,
+      "grad_norm": 0.2011396586894989,
+      "learning_rate": 7.545045045045045e-05,
+      "loss": 0.3143,
+      "step": 4995
+    },
+    {
+      "epoch": 1.497657860221098,
+      "grad_norm": 0.2071286290884018,
+      "learning_rate": 7.54054054054054e-05,
+      "loss": 0.3822,
+      "step": 4996
+    },
+    {
+      "epoch": 1.4979576541127975,
+      "grad_norm": 0.20190057158470154,
+      "learning_rate": 7.536036036036035e-05,
+      "loss": 0.3512,
+      "step": 4997
+    },
+    {
+      "epoch": 1.4982574480044968,
+      "grad_norm": 0.20391745865345,
+      "learning_rate": 7.531531531531531e-05,
+      "loss": 0.3613,
+      "step": 4998
+    },
+    {
+      "epoch": 1.4985572418961963,
+      "grad_norm": 0.20113232731819153,
+      "learning_rate": 7.527027027027026e-05,
+      "loss": 0.3377,
+      "step": 4999
+    },
+    {
+      "epoch": 1.4988570357878959,
+      "grad_norm": 0.20599274337291718,
+      "learning_rate": 7.522522522522523e-05,
+      "loss": 0.34,
+      "step": 5000
+    },
+    {
+      "epoch": 1.4988570357878959,
+      "eval_loss": 0.409088671207428,
+      "eval_runtime": 565.6657,
+      "eval_samples_per_second": 3.817,
+      "eval_steps_per_second": 0.477,
+      "step": 5000
+    },
+    {
+      "epoch": 1.4991568296795954,
+      "grad_norm": 0.21036118268966675,
+      "learning_rate": 7.518018018018017e-05,
+      "loss": 0.349,
+      "step": 5001
+    },
+    {
+      "epoch": 1.4994566235712947,
+      "grad_norm": 0.20376059412956238,
+      "learning_rate": 7.513513513513513e-05,
+      "loss": 0.3423,
+      "step": 5002
+    },
+    {
+      "epoch": 1.4997564174629943,
+      "grad_norm": 0.20630986988544464,
+      "learning_rate": 7.509009009009009e-05,
+      "loss": 0.3514,
+      "step": 5003
+    },
+    {
+      "epoch": 1.5000562113546936,
+      "grad_norm": 0.20768223702907562,
+      "learning_rate": 7.504504504504503e-05,
+      "loss": 0.3557,
+      "step": 5004
+    },
+    {
+      "epoch": 1.5003560052463931,
+      "grad_norm": 0.22584180533885956,
+      "learning_rate": 7.5e-05,
+      "loss": 0.3641,
+      "step": 5005
+    },
+    {
+      "epoch": 1.5006557991380927,
+      "grad_norm": 0.20455649495124817,
+      "learning_rate": 7.495495495495494e-05,
+      "loss": 0.3396,
+      "step": 5006
+    },
+    {
+      "epoch": 1.500955593029792,
+      "grad_norm": 0.21096675097942352,
+      "learning_rate": 7.490990990990991e-05,
+      "loss": 0.3527,
+      "step": 5007
+    },
+    {
+      "epoch": 1.5012553869214915,
+      "grad_norm": 0.20545406639575958,
+      "learning_rate": 7.486486486486486e-05,
+      "loss": 0.3494,
+      "step": 5008
+    },
+    {
+      "epoch": 1.5015551808131908,
+      "grad_norm": 0.20965568721294403,
+      "learning_rate": 7.481981981981981e-05,
+      "loss": 0.3613,
+      "step": 5009
+    },
+    {
+      "epoch": 1.5018549747048904,
+      "grad_norm": 0.21544276177883148,
+      "learning_rate": 7.477477477477476e-05,
+      "loss": 0.3565,
+      "step": 5010
+    },
+    {
+      "epoch": 1.50215476859659,
+      "grad_norm": 0.20230863988399506,
+      "learning_rate": 7.472972972972972e-05,
+      "loss": 0.3376,
+      "step": 5011
+    },
+    {
+      "epoch": 1.5024545624882895,
+      "grad_norm": 0.20163989067077637,
+      "learning_rate": 7.468468468468467e-05,
+      "loss": 0.3276,
+      "step": 5012
+    },
+    {
+      "epoch": 1.5027543563799888,
+      "grad_norm": 0.21091212332248688,
+      "learning_rate": 7.463963963963964e-05,
+      "loss": 0.3613,
+      "step": 5013
+    },
+    {
+      "epoch": 1.503054150271688,
+      "grad_norm": 0.21473614871501923,
+      "learning_rate": 7.459459459459459e-05,
+      "loss": 0.345,
+      "step": 5014
+    },
+    {
+      "epoch": 1.5033539441633876,
+      "grad_norm": 0.19872497022151947,
+      "learning_rate": 7.454954954954955e-05,
+      "loss": 0.349,
+      "step": 5015
+    },
+    {
+      "epoch": 1.5036537380550872,
+      "grad_norm": 0.21866753697395325,
+      "learning_rate": 7.45045045045045e-05,
+      "loss": 0.3704,
+      "step": 5016
+    },
+    {
+      "epoch": 1.5039535319467867,
+      "grad_norm": 0.19531461596488953,
+      "learning_rate": 7.445945945945945e-05,
+      "loss": 0.3293,
+      "step": 5017
+    },
+    {
+      "epoch": 1.504253325838486,
+      "grad_norm": 0.20132407546043396,
+      "learning_rate": 7.441441441441442e-05,
+      "loss": 0.3535,
+      "step": 5018
+    },
+    {
+      "epoch": 1.5045531197301854,
+      "grad_norm": 0.20986546576023102,
+      "learning_rate": 7.436936936936937e-05,
+      "loss": 0.3635,
+      "step": 5019
+    },
+    {
+      "epoch": 1.504852913621885,
+      "grad_norm": 0.21108418703079224,
+      "learning_rate": 7.432432432432432e-05,
+      "loss": 0.3592,
+      "step": 5020
+    },
+    {
+      "epoch": 1.5051527075135844,
+      "grad_norm": 0.2011735737323761,
+      "learning_rate": 7.427927927927927e-05,
+      "loss": 0.3311,
+      "step": 5021
+    },
+    {
+      "epoch": 1.505452501405284,
+      "grad_norm": 0.20430687069892883,
+      "learning_rate": 7.423423423423423e-05,
+      "loss": 0.3586,
+      "step": 5022
+    },
+    {
+      "epoch": 1.5057522952969833,
+      "grad_norm": 0.19233015179634094,
+      "learning_rate": 7.418918918918918e-05,
+      "loss": 0.3203,
+      "step": 5023
+    },
+    {
+      "epoch": 1.5060520891886828,
+      "grad_norm": 0.21224820613861084,
+      "learning_rate": 7.414414414414413e-05,
+      "loss": 0.352,
+      "step": 5024
+    },
+    {
+      "epoch": 1.5063518830803821,
+      "grad_norm": 0.2162107676267624,
+      "learning_rate": 7.40990990990991e-05,
+      "loss": 0.3716,
+      "step": 5025
+    },
+    {
+      "epoch": 1.5066516769720817,
+      "grad_norm": 0.21990056335926056,
+      "learning_rate": 7.405405405405405e-05,
+      "loss": 0.371,
+      "step": 5026
+    },
+    {
+      "epoch": 1.5069514708637812,
+      "grad_norm": 0.2139321267604828,
+      "learning_rate": 7.4009009009009e-05,
+      "loss": 0.3406,
+      "step": 5027
+    },
+    {
+      "epoch": 1.5072512647554808,
+      "grad_norm": 0.21176601946353912,
+      "learning_rate": 7.396396396396396e-05,
+      "loss": 0.3863,
+      "step": 5028
+    },
+    {
+      "epoch": 1.50755105864718,
+      "grad_norm": 0.21457381546497345,
+      "learning_rate": 7.391891891891891e-05,
+      "loss": 0.3651,
+      "step": 5029
+    },
+    {
+      "epoch": 1.5078508525388794,
+      "grad_norm": 0.22571976482868195,
+      "learning_rate": 7.387387387387387e-05,
+      "loss": 0.3626,
+      "step": 5030
+    },
+    {
+      "epoch": 1.508150646430579,
+      "grad_norm": 0.20284788310527802,
+      "learning_rate": 7.382882882882882e-05,
+      "loss": 0.3339,
+      "step": 5031
+    },
+    {
+      "epoch": 1.5084504403222785,
+      "grad_norm": 0.2027716338634491,
+      "learning_rate": 7.378378378378379e-05,
+      "loss": 0.3439,
+      "step": 5032
+    },
+    {
+      "epoch": 1.508750234213978,
+      "grad_norm": 0.21533340215682983,
+      "learning_rate": 7.373873873873874e-05,
+      "loss": 0.3524,
+      "step": 5033
+    },
+    {
+      "epoch": 1.5090500281056773,
+      "grad_norm": 0.2093924582004547,
+      "learning_rate": 7.369369369369369e-05,
+      "loss": 0.37,
+      "step": 5034
+    },
+    {
+      "epoch": 1.5093498219973767,
+      "grad_norm": 0.20701274275779724,
+      "learning_rate": 7.364864864864864e-05,
+      "loss": 0.3559,
+      "step": 5035
+    },
+    {
+      "epoch": 1.5096496158890762,
+      "grad_norm": 0.20343904197216034,
+      "learning_rate": 7.36036036036036e-05,
+      "loss": 0.3713,
+      "step": 5036
+    },
+    {
+      "epoch": 1.5099494097807757,
+      "grad_norm": 0.20688888430595398,
+      "learning_rate": 7.355855855855855e-05,
+      "loss": 0.3541,
+      "step": 5037
+    },
+    {
+      "epoch": 1.5102492036724753,
+      "grad_norm": 0.21339473128318787,
+      "learning_rate": 7.35135135135135e-05,
+      "loss": 0.3674,
+      "step": 5038
+    },
+    {
+      "epoch": 1.5105489975641746,
+      "grad_norm": 0.2124233990907669,
+      "learning_rate": 7.346846846846845e-05,
+      "loss": 0.3723,
+      "step": 5039
+    },
+    {
+      "epoch": 1.5108487914558741,
+      "grad_norm": 0.20129559934139252,
+      "learning_rate": 7.342342342342342e-05,
+      "loss": 0.3266,
+      "step": 5040
+    },
+    {
+      "epoch": 1.5111485853475735,
+      "grad_norm": 0.23245026171207428,
+      "learning_rate": 7.337837837837837e-05,
+      "loss": 0.3649,
+      "step": 5041
+    },
+    {
+      "epoch": 1.511448379239273,
+      "grad_norm": 0.2057090848684311,
+      "learning_rate": 7.333333333333332e-05,
+      "loss": 0.3694,
+      "step": 5042
+    },
+    {
+      "epoch": 1.5117481731309725,
+      "grad_norm": 0.20243839919567108,
+      "learning_rate": 7.328828828828828e-05,
+      "loss": 0.3258,
+      "step": 5043
+    },
+    {
+      "epoch": 1.512047967022672,
+      "grad_norm": 0.20519401133060455,
+      "learning_rate": 7.324324324324323e-05,
+      "loss": 0.3383,
+      "step": 5044
+    },
+    {
+      "epoch": 1.5123477609143714,
+      "grad_norm": 0.2056913673877716,
+      "learning_rate": 7.31981981981982e-05,
+      "loss": 0.3543,
+      "step": 5045
+    },
+    {
+      "epoch": 1.5126475548060707,
+      "grad_norm": 0.20329561829566956,
+      "learning_rate": 7.315315315315315e-05,
+      "loss": 0.339,
+      "step": 5046
+    },
+    {
+      "epoch": 1.5129473486977703,
+      "grad_norm": 0.20037339627742767,
+      "learning_rate": 7.310810810810811e-05,
+      "loss": 0.3371,
+      "step": 5047
+    },
+    {
+      "epoch": 1.5132471425894698,
+      "grad_norm": 0.21403111517429352,
+      "learning_rate": 7.306306306306306e-05,
+      "loss": 0.3854,
+      "step": 5048
+    },
+    {
+      "epoch": 1.5135469364811693,
+      "grad_norm": 0.20956586301326752,
+      "learning_rate": 7.301801801801801e-05,
+      "loss": 0.3696,
+      "step": 5049
+    },
+    {
+      "epoch": 1.5138467303728687,
+      "grad_norm": 0.208208367228508,
+      "learning_rate": 7.297297297297297e-05,
+      "loss": 0.3495,
+      "step": 5050
+    },
+    {
+      "epoch": 1.514146524264568,
+      "grad_norm": 0.2096438705921173,
+      "learning_rate": 7.292792792792792e-05,
+      "loss": 0.3275,
+      "step": 5051
+    },
+    {
+      "epoch": 1.5144463181562675,
+      "grad_norm": 0.22821904718875885,
+      "learning_rate": 7.288288288288288e-05,
+      "loss": 0.3513,
+      "step": 5052
+    },
+    {
+      "epoch": 1.514746112047967,
+      "grad_norm": 0.20898018777370453,
+      "learning_rate": 7.283783783783783e-05,
+      "loss": 0.366,
+      "step": 5053
+    },
+    {
+      "epoch": 1.5150459059396666,
+      "grad_norm": 0.21257898211479187,
+      "learning_rate": 7.279279279279279e-05,
+      "loss": 0.3604,
+      "step": 5054
+    },
+    {
+      "epoch": 1.515345699831366,
+      "grad_norm": 0.2063610702753067,
+      "learning_rate": 7.274774774774774e-05,
+      "loss": 0.3513,
+      "step": 5055
+    },
+    {
+      "epoch": 1.5156454937230652,
+      "grad_norm": 0.21631956100463867,
+      "learning_rate": 7.270270270270269e-05,
+      "loss": 0.3586,
+      "step": 5056
+    },
+    {
+      "epoch": 1.5159452876147648,
+      "grad_norm": 0.21602363884449005,
+      "learning_rate": 7.265765765765765e-05,
+      "loss": 0.3659,
+      "step": 5057
+    },
+    {
+      "epoch": 1.5162450815064643,
+      "grad_norm": 0.20453397929668427,
+      "learning_rate": 7.26126126126126e-05,
+      "loss": 0.346,
+      "step": 5058
+    },
+    {
+      "epoch": 1.5165448753981638,
+      "grad_norm": 0.20310211181640625,
+      "learning_rate": 7.256756756756755e-05,
+      "loss": 0.3352,
+      "step": 5059
+    },
+    {
+      "epoch": 1.5168446692898632,
+      "grad_norm": 0.2058761715888977,
+      "learning_rate": 7.252252252252252e-05,
+      "loss": 0.3479,
+      "step": 5060
+    },
+    {
+      "epoch": 1.5171444631815627,
+      "grad_norm": 0.20629523694515228,
+      "learning_rate": 7.247747747747747e-05,
+      "loss": 0.3784,
+      "step": 5061
+    },
+    {
+      "epoch": 1.517444257073262,
+      "grad_norm": 0.2118360996246338,
+      "learning_rate": 7.243243243243243e-05,
+      "loss": 0.3586,
+      "step": 5062
+    },
+    {
+      "epoch": 1.5177440509649616,
+      "grad_norm": 0.22766007483005524,
+      "learning_rate": 7.238738738738738e-05,
+      "loss": 0.3553,
+      "step": 5063
+    },
+    {
+      "epoch": 1.518043844856661,
+      "grad_norm": 0.21020479500293732,
+      "learning_rate": 7.234234234234233e-05,
+      "loss": 0.3518,
+      "step": 5064
+    },
+    {
+      "epoch": 1.5183436387483606,
+      "grad_norm": 0.19974002242088318,
+      "learning_rate": 7.22972972972973e-05,
+      "loss": 0.3441,
+      "step": 5065
+    },
+    {
+      "epoch": 1.51864343264006,
+      "grad_norm": 0.20955109596252441,
+      "learning_rate": 7.225225225225225e-05,
+      "loss": 0.3786,
+      "step": 5066
+    },
+    {
+      "epoch": 1.5189432265317593,
+      "grad_norm": 0.19488339126110077,
+      "learning_rate": 7.22072072072072e-05,
+      "loss": 0.3461,
+      "step": 5067
+    },
+    {
+      "epoch": 1.5192430204234588,
+      "grad_norm": 0.20264093577861786,
+      "learning_rate": 7.216216216216216e-05,
+      "loss": 0.3526,
+      "step": 5068
+    },
+    {
+      "epoch": 1.5195428143151584,
+      "grad_norm": 0.20072831213474274,
+      "learning_rate": 7.211711711711711e-05,
+      "loss": 0.3539,
+      "step": 5069
+    },
+    {
+      "epoch": 1.519842608206858,
+      "grad_norm": 0.19965948164463043,
+      "learning_rate": 7.207207207207206e-05,
+      "loss": 0.3529,
+      "step": 5070
+    },
+    {
+      "epoch": 1.5201424020985572,
+      "grad_norm": 0.21437616646289825,
+      "learning_rate": 7.202702702702701e-05,
+      "loss": 0.3363,
+      "step": 5071
+    },
+    {
+      "epoch": 1.5204421959902565,
+      "grad_norm": 0.20634238421916962,
+      "learning_rate": 7.198198198198198e-05,
+      "loss": 0.3653,
+      "step": 5072
+    },
+    {
+      "epoch": 1.520741989881956,
+      "grad_norm": 0.20045985281467438,
+      "learning_rate": 7.193693693693693e-05,
+      "loss": 0.3508,
+      "step": 5073
+    },
+    {
+      "epoch": 1.5210417837736556,
+      "grad_norm": 0.2013823390007019,
+      "learning_rate": 7.189189189189189e-05,
+      "loss": 0.3434,
+      "step": 5074
+    },
+    {
+      "epoch": 1.5213415776653552,
+      "grad_norm": 0.20428241789340973,
+      "learning_rate": 7.184684684684684e-05,
+      "loss": 0.3478,
+      "step": 5075
+    },
+    {
+      "epoch": 1.5216413715570545,
+      "grad_norm": 0.20228740572929382,
+      "learning_rate": 7.180180180180179e-05,
+      "loss": 0.3502,
+      "step": 5076
+    },
+    {
+      "epoch": 1.521941165448754,
+      "grad_norm": 0.21271152794361115,
+      "learning_rate": 7.175675675675676e-05,
+      "loss": 0.3606,
+      "step": 5077
+    },
+    {
+      "epoch": 1.5222409593404533,
+      "grad_norm": 0.20367246866226196,
+      "learning_rate": 7.17117117117117e-05,
+      "loss": 0.3658,
+      "step": 5078
+    },
+    {
+      "epoch": 1.5225407532321529,
+      "grad_norm": 0.20234622061252594,
+      "learning_rate": 7.166666666666667e-05,
+      "loss": 0.3669,
+      "step": 5079
+    },
+    {
+      "epoch": 1.5228405471238524,
+      "grad_norm": 0.20251305401325226,
+      "learning_rate": 7.162162162162162e-05,
+      "loss": 0.353,
+      "step": 5080
+    },
+    {
+      "epoch": 1.523140341015552,
+      "grad_norm": 0.2070438116788864,
+      "learning_rate": 7.157657657657657e-05,
+      "loss": 0.3654,
+      "step": 5081
+    },
+    {
+      "epoch": 1.5234401349072513,
+      "grad_norm": 0.21502596139907837,
+      "learning_rate": 7.153153153153152e-05,
+      "loss": 0.3666,
+      "step": 5082
+    },
+    {
+      "epoch": 1.5237399287989506,
+      "grad_norm": 0.2074059396982193,
+      "learning_rate": 7.148648648648648e-05,
+      "loss": 0.3868,
+      "step": 5083
+    },
+    {
+      "epoch": 1.5240397226906501,
+      "grad_norm": 0.20912374556064606,
+      "learning_rate": 7.144144144144143e-05,
+      "loss": 0.3563,
+      "step": 5084
+    },
+    {
+      "epoch": 1.5243395165823497,
+      "grad_norm": 0.19438451528549194,
+      "learning_rate": 7.139639639639638e-05,
+      "loss": 0.3388,
+      "step": 5085
+    },
+    {
+      "epoch": 1.5246393104740492,
+      "grad_norm": 0.20785526931285858,
+      "learning_rate": 7.135135135135135e-05,
+      "loss": 0.3535,
+      "step": 5086
+    },
+    {
+      "epoch": 1.5249391043657485,
+      "grad_norm": 0.21372053027153015,
+      "learning_rate": 7.13063063063063e-05,
+      "loss": 0.385,
+      "step": 5087
+    },
+    {
+      "epoch": 1.5252388982574479,
+      "grad_norm": 0.2027290165424347,
+      "learning_rate": 7.126126126126125e-05,
+      "loss": 0.3561,
+      "step": 5088
+    },
+    {
+      "epoch": 1.5255386921491474,
+      "grad_norm": 0.21128539741039276,
+      "learning_rate": 7.121621621621621e-05,
+      "loss": 0.3608,
+      "step": 5089
+    },
+    {
+      "epoch": 1.525838486040847,
+      "grad_norm": 0.2072833627462387,
+      "learning_rate": 7.117117117117116e-05,
+      "loss": 0.3572,
+      "step": 5090
+    },
+    {
+      "epoch": 1.5261382799325465,
+      "grad_norm": 0.21812467277050018,
+      "learning_rate": 7.112612612612611e-05,
+      "loss": 0.3916,
+      "step": 5091
+    },
+    {
+      "epoch": 1.5264380738242458,
+      "grad_norm": 0.1990809589624405,
+      "learning_rate": 7.108108108108108e-05,
+      "loss": 0.3457,
+      "step": 5092
+    },
+    {
+      "epoch": 1.5267378677159453,
+      "grad_norm": 0.21215678751468658,
+      "learning_rate": 7.103603603603603e-05,
+      "loss": 0.361,
+      "step": 5093
+    },
+    {
+      "epoch": 1.5270376616076446,
+      "grad_norm": 0.20100034773349762,
+      "learning_rate": 7.099099099099099e-05,
+      "loss": 0.346,
+      "step": 5094
+    },
+    {
+      "epoch": 1.5273374554993442,
+      "grad_norm": 0.25413811206817627,
+      "learning_rate": 7.094594594594594e-05,
+      "loss": 0.339,
+      "step": 5095
+    },
+    {
+      "epoch": 1.5276372493910437,
+      "grad_norm": 0.19817399978637695,
+      "learning_rate": 7.090090090090089e-05,
+      "loss": 0.3539,
+      "step": 5096
+    },
+    {
+      "epoch": 1.5279370432827433,
+      "grad_norm": 0.20904068648815155,
+      "learning_rate": 7.085585585585586e-05,
+      "loss": 0.3499,
+      "step": 5097
+    },
+    {
+      "epoch": 1.5282368371744426,
+      "grad_norm": 0.21889416873455048,
+      "learning_rate": 7.081081081081081e-05,
+      "loss": 0.3659,
+      "step": 5098
+    },
+    {
+      "epoch": 1.528536631066142,
+      "grad_norm": 0.2134886085987091,
+      "learning_rate": 7.076576576576576e-05,
+      "loss": 0.3583,
+      "step": 5099
+    },
+    {
+      "epoch": 1.5288364249578414,
+      "grad_norm": 0.20167680084705353,
+      "learning_rate": 7.072072072072071e-05,
+      "loss": 0.3578,
+      "step": 5100
+    },
+    {
+      "epoch": 1.529136218849541,
+      "grad_norm": 0.21113313734531403,
+      "learning_rate": 7.067567567567567e-05,
+      "loss": 0.329,
+      "step": 5101
+    },
+    {
+      "epoch": 1.5294360127412405,
+      "grad_norm": 0.1989641785621643,
+      "learning_rate": 7.063063063063062e-05,
+      "loss": 0.3332,
+      "step": 5102
+    },
+    {
+      "epoch": 1.5297358066329398,
+      "grad_norm": 0.20343445241451263,
+      "learning_rate": 7.058558558558557e-05,
+      "loss": 0.3505,
+      "step": 5103
+    },
+    {
+      "epoch": 1.5300356005246392,
+      "grad_norm": 0.21245552599430084,
+      "learning_rate": 7.054054054054054e-05,
+      "loss": 0.3468,
+      "step": 5104
+    },
+    {
+      "epoch": 1.5303353944163387,
+      "grad_norm": 0.209833025932312,
+      "learning_rate": 7.049549549549549e-05,
+      "loss": 0.3684,
+      "step": 5105
+    },
+    {
+      "epoch": 1.5306351883080382,
+      "grad_norm": 0.19674982130527496,
+      "learning_rate": 7.045045045045045e-05,
+      "loss": 0.3338,
+      "step": 5106
+    },
+    {
+      "epoch": 1.5309349821997378,
+      "grad_norm": 0.2160000056028366,
+      "learning_rate": 7.04054054054054e-05,
+      "loss": 0.3603,
+      "step": 5107
+    },
+    {
+      "epoch": 1.531234776091437,
+      "grad_norm": 0.19963768124580383,
+      "learning_rate": 7.036036036036035e-05,
+      "loss": 0.3641,
+      "step": 5108
+    },
+    {
+      "epoch": 1.5315345699831366,
+      "grad_norm": 0.1939668208360672,
+      "learning_rate": 7.031531531531531e-05,
+      "loss": 0.3339,
+      "step": 5109
+    },
+    {
+      "epoch": 1.531834363874836,
+      "grad_norm": 0.2016298770904541,
+      "learning_rate": 7.027027027027026e-05,
+      "loss": 0.3602,
+      "step": 5110
+    },
+    {
+      "epoch": 1.5321341577665355,
+      "grad_norm": 0.21612954139709473,
+      "learning_rate": 7.022522522522522e-05,
+      "loss": 0.3846,
+      "step": 5111
+    },
+    {
+      "epoch": 1.532433951658235,
+      "grad_norm": 0.1958528757095337,
+      "learning_rate": 7.018018018018018e-05,
+      "loss": 0.3414,
+      "step": 5112
+    },
+    {
+      "epoch": 1.5327337455499346,
+      "grad_norm": 0.19213689863681793,
+      "learning_rate": 7.013513513513513e-05,
+      "loss": 0.3327,
+      "step": 5113
+    },
+    {
+      "epoch": 1.533033539441634,
+      "grad_norm": 0.2063470482826233,
+      "learning_rate": 7.009009009009008e-05,
+      "loss": 0.3538,
+      "step": 5114
+    },
+    {
+      "epoch": 1.5333333333333332,
+      "grad_norm": 0.21728667616844177,
+      "learning_rate": 7.004504504504504e-05,
+      "loss": 0.3551,
+      "step": 5115
+    },
+    {
+      "epoch": 1.5336331272250328,
+      "grad_norm": 0.20425064861774445,
+      "learning_rate": 7e-05,
+      "loss": 0.3249,
+      "step": 5116
+    },
+    {
+      "epoch": 1.5339329211167323,
+      "grad_norm": 0.20801694691181183,
+      "learning_rate": 6.995495495495494e-05,
+      "loss": 0.3611,
+      "step": 5117
+    },
+    {
+      "epoch": 1.5342327150084318,
+      "grad_norm": 0.19504021108150482,
+      "learning_rate": 6.99099099099099e-05,
+      "loss": 0.348,
+      "step": 5118
+    },
+    {
+      "epoch": 1.5345325089001312,
+      "grad_norm": 0.21642310917377472,
+      "learning_rate": 6.986486486486486e-05,
+      "loss": 0.3574,
+      "step": 5119
+    },
+    {
+      "epoch": 1.5348323027918305,
+      "grad_norm": 0.2077680230140686,
+      "learning_rate": 6.981981981981981e-05,
+      "loss": 0.3348,
+      "step": 5120
+    },
+    {
+      "epoch": 1.53513209668353,
+      "grad_norm": 0.2039538472890854,
+      "learning_rate": 6.977477477477477e-05,
+      "loss": 0.3443,
+      "step": 5121
+    },
+    {
+      "epoch": 1.5354318905752296,
+      "grad_norm": 0.20255398750305176,
+      "learning_rate": 6.972972972972972e-05,
+      "loss": 0.3563,
+      "step": 5122
+    },
+    {
+      "epoch": 1.535731684466929,
+      "grad_norm": 0.21409064531326294,
+      "learning_rate": 6.968468468468469e-05,
+      "loss": 0.3481,
+      "step": 5123
+    },
+    {
+      "epoch": 1.5360314783586284,
+      "grad_norm": 0.20551519095897675,
+      "learning_rate": 6.963963963963964e-05,
+      "loss": 0.3597,
+      "step": 5124
+    },
+    {
+      "epoch": 1.536331272250328,
+      "grad_norm": 0.2170633226633072,
+      "learning_rate": 6.959459459459459e-05,
+      "loss": 0.3652,
+      "step": 5125
+    },
+    {
+      "epoch": 1.5366310661420273,
+      "grad_norm": 0.20070041716098785,
+      "learning_rate": 6.954954954954955e-05,
+      "loss": 0.3412,
+      "step": 5126
+    },
+    {
+      "epoch": 1.5369308600337268,
+      "grad_norm": 0.22287438809871674,
+      "learning_rate": 6.95045045045045e-05,
+      "loss": 0.3788,
+      "step": 5127
+    },
+    {
+      "epoch": 1.5372306539254263,
+      "grad_norm": 0.2027093470096588,
+      "learning_rate": 6.945945945945945e-05,
+      "loss": 0.3439,
+      "step": 5128
+    },
+    {
+      "epoch": 1.537530447817126,
+      "grad_norm": 0.19838665425777435,
+      "learning_rate": 6.94144144144144e-05,
+      "loss": 0.3213,
+      "step": 5129
+    },
+    {
+      "epoch": 1.5378302417088252,
+      "grad_norm": 0.20357826352119446,
+      "learning_rate": 6.936936936936937e-05,
+      "loss": 0.35,
+      "step": 5130
+    },
+    {
+      "epoch": 1.5381300356005245,
+      "grad_norm": 0.2061091959476471,
+      "learning_rate": 6.932432432432432e-05,
+      "loss": 0.3341,
+      "step": 5131
+    },
+    {
+      "epoch": 1.538429829492224,
+      "grad_norm": 0.205326110124588,
+      "learning_rate": 6.927927927927927e-05,
+      "loss": 0.3473,
+      "step": 5132
+    },
+    {
+      "epoch": 1.5387296233839236,
+      "grad_norm": 0.21463227272033691,
+      "learning_rate": 6.923423423423423e-05,
+      "loss": 0.3708,
+      "step": 5133
+    },
+    {
+      "epoch": 1.5390294172756231,
+      "grad_norm": 0.1967410147190094,
+      "learning_rate": 6.918918918918918e-05,
+      "loss": 0.3439,
+      "step": 5134
+    },
+    {
+      "epoch": 1.5393292111673225,
+      "grad_norm": 0.20796793699264526,
+      "learning_rate": 6.914414414414413e-05,
+      "loss": 0.3573,
+      "step": 5135
+    },
+    {
+      "epoch": 1.5396290050590218,
+      "grad_norm": 0.2030736207962036,
+      "learning_rate": 6.90990990990991e-05,
+      "loss": 0.3634,
+      "step": 5136
+    },
+    {
+      "epoch": 1.5399287989507213,
+      "grad_norm": 0.20618915557861328,
+      "learning_rate": 6.905405405405405e-05,
+      "loss": 0.3503,
+      "step": 5137
+    },
+    {
+      "epoch": 1.5402285928424209,
+      "grad_norm": 0.21252769231796265,
+      "learning_rate": 6.900900900900901e-05,
+      "loss": 0.3655,
+      "step": 5138
+    },
+    {
+      "epoch": 1.5405283867341204,
+      "grad_norm": 0.2006307691335678,
+      "learning_rate": 6.896396396396396e-05,
+      "loss": 0.3445,
+      "step": 5139
+    },
+    {
+      "epoch": 1.5408281806258197,
+      "grad_norm": 0.19492262601852417,
+      "learning_rate": 6.891891891891891e-05,
+      "loss": 0.3275,
+      "step": 5140
+    },
+    {
+      "epoch": 1.541127974517519,
+      "grad_norm": 0.20005106925964355,
+      "learning_rate": 6.887387387387387e-05,
+      "loss": 0.3486,
+      "step": 5141
+    },
+    {
+      "epoch": 1.5414277684092186,
+      "grad_norm": 0.21888670325279236,
+      "learning_rate": 6.882882882882882e-05,
+      "loss": 0.3486,
+      "step": 5142
+    },
+    {
+      "epoch": 1.5417275623009181,
+      "grad_norm": 0.19712640345096588,
+      "learning_rate": 6.878378378378377e-05,
+      "loss": 0.3372,
+      "step": 5143
+    },
+    {
+      "epoch": 1.5420273561926177,
+      "grad_norm": 0.19769833981990814,
+      "learning_rate": 6.873873873873874e-05,
+      "loss": 0.3589,
+      "step": 5144
+    },
+    {
+      "epoch": 1.542327150084317,
+      "grad_norm": 0.2131979912519455,
+      "learning_rate": 6.869369369369369e-05,
+      "loss": 0.3463,
+      "step": 5145
+    },
+    {
+      "epoch": 1.5426269439760165,
+      "grad_norm": 0.20438240468502045,
+      "learning_rate": 6.864864864864864e-05,
+      "loss": 0.3632,
+      "step": 5146
+    },
+    {
+      "epoch": 1.5429267378677158,
+      "grad_norm": 0.2102312594652176,
+      "learning_rate": 6.860360360360359e-05,
+      "loss": 0.3709,
+      "step": 5147
+    },
+    {
+      "epoch": 1.5432265317594154,
+      "grad_norm": 0.20745466649532318,
+      "learning_rate": 6.855855855855855e-05,
+      "loss": 0.373,
+      "step": 5148
+    },
+    {
+      "epoch": 1.543526325651115,
+      "grad_norm": 0.20155175030231476,
+      "learning_rate": 6.85135135135135e-05,
+      "loss": 0.3678,
+      "step": 5149
+    },
+    {
+      "epoch": 1.5438261195428145,
+      "grad_norm": 0.21392785012722015,
+      "learning_rate": 6.846846846846845e-05,
+      "loss": 0.3831,
+      "step": 5150
+    },
+    {
+      "epoch": 1.5441259134345138,
+      "grad_norm": 0.20007534325122833,
+      "learning_rate": 6.842342342342342e-05,
+      "loss": 0.352,
+      "step": 5151
+    },
+    {
+      "epoch": 1.544425707326213,
+      "grad_norm": 0.20434188842773438,
+      "learning_rate": 6.837837837837837e-05,
+      "loss": 0.3534,
+      "step": 5152
+    },
+    {
+      "epoch": 1.5447255012179126,
+      "grad_norm": 0.2060280591249466,
+      "learning_rate": 6.833333333333333e-05,
+      "loss": 0.3425,
+      "step": 5153
+    },
+    {
+      "epoch": 1.5450252951096122,
+      "grad_norm": 0.20043106377124786,
+      "learning_rate": 6.828828828828828e-05,
+      "loss": 0.3471,
+      "step": 5154
+    },
+    {
+      "epoch": 1.5453250890013117,
+      "grad_norm": 0.20862244069576263,
+      "learning_rate": 6.824324324324325e-05,
+      "loss": 0.3743,
+      "step": 5155
+    },
+    {
+      "epoch": 1.545624882893011,
+      "grad_norm": 0.20181861519813538,
+      "learning_rate": 6.81981981981982e-05,
+      "loss": 0.3345,
+      "step": 5156
+    },
+    {
+      "epoch": 1.5459246767847103,
+      "grad_norm": 0.19661659002304077,
+      "learning_rate": 6.815315315315315e-05,
+      "loss": 0.3485,
+      "step": 5157
+    },
+    {
+      "epoch": 1.54622447067641,
+      "grad_norm": 0.19832420349121094,
+      "learning_rate": 6.810810810810811e-05,
+      "loss": 0.3398,
+      "step": 5158
+    },
+    {
+      "epoch": 1.5465242645681094,
+      "grad_norm": 0.19693104922771454,
+      "learning_rate": 6.806306306306306e-05,
+      "loss": 0.3493,
+      "step": 5159
+    },
+    {
+      "epoch": 1.546824058459809,
+      "grad_norm": 0.20385287702083588,
+      "learning_rate": 6.801801801801801e-05,
+      "loss": 0.3658,
+      "step": 5160
+    },
+    {
+      "epoch": 1.5471238523515083,
+      "grad_norm": 0.19576039910316467,
+      "learning_rate": 6.797297297297296e-05,
+      "loss": 0.3485,
+      "step": 5161
+    },
+    {
+      "epoch": 1.5474236462432078,
+      "grad_norm": 0.19672639667987823,
+      "learning_rate": 6.792792792792793e-05,
+      "loss": 0.3395,
+      "step": 5162
+    },
+    {
+      "epoch": 1.5477234401349071,
+      "grad_norm": 0.19690831005573273,
+      "learning_rate": 6.788288288288288e-05,
+      "loss": 0.3501,
+      "step": 5163
+    },
+    {
+      "epoch": 1.5480232340266067,
+      "grad_norm": 0.20069430768489838,
+      "learning_rate": 6.783783783783783e-05,
+      "loss": 0.3425,
+      "step": 5164
+    },
+    {
+      "epoch": 1.5483230279183062,
+      "grad_norm": 0.20951126515865326,
+      "learning_rate": 6.779279279279278e-05,
+      "loss": 0.3594,
+      "step": 5165
+    },
+    {
+      "epoch": 1.5486228218100058,
+      "grad_norm": 0.2090735137462616,
+      "learning_rate": 6.774774774774774e-05,
+      "loss": 0.3967,
+      "step": 5166
+    },
+    {
+      "epoch": 1.548922615701705,
+      "grad_norm": 0.2065633237361908,
+      "learning_rate": 6.770270270270269e-05,
+      "loss": 0.3609,
+      "step": 5167
+    },
+    {
+      "epoch": 1.5492224095934044,
+      "grad_norm": 0.20373541116714478,
+      "learning_rate": 6.765765765765765e-05,
+      "loss": 0.3446,
+      "step": 5168
+    },
+    {
+      "epoch": 1.549522203485104,
+      "grad_norm": 0.20226061344146729,
+      "learning_rate": 6.76126126126126e-05,
+      "loss": 0.351,
+      "step": 5169
+    },
+    {
+      "epoch": 1.5498219973768035,
+      "grad_norm": 0.19589018821716309,
+      "learning_rate": 6.756756756756757e-05,
+      "loss": 0.3165,
+      "step": 5170
+    },
+    {
+      "epoch": 1.550121791268503,
+      "grad_norm": 0.19932496547698975,
+      "learning_rate": 6.752252252252252e-05,
+      "loss": 0.3503,
+      "step": 5171
+    },
+    {
+      "epoch": 1.5504215851602023,
+      "grad_norm": 0.20237968862056732,
+      "learning_rate": 6.747747747747747e-05,
+      "loss": 0.3637,
+      "step": 5172
+    },
+    {
+      "epoch": 1.5507213790519017,
+      "grad_norm": 0.20368225872516632,
+      "learning_rate": 6.743243243243243e-05,
+      "loss": 0.3595,
+      "step": 5173
+    },
+    {
+      "epoch": 1.5510211729436012,
+      "grad_norm": 0.20362980663776398,
+      "learning_rate": 6.738738738738738e-05,
+      "loss": 0.3522,
+      "step": 5174
+    },
+    {
+      "epoch": 1.5513209668353007,
+      "grad_norm": 0.20698122680187225,
+      "learning_rate": 6.734234234234233e-05,
+      "loss": 0.3547,
+      "step": 5175
+    },
+    {
+      "epoch": 1.5516207607270003,
+      "grad_norm": 0.20629571378231049,
+      "learning_rate": 6.72972972972973e-05,
+      "loss": 0.3708,
+      "step": 5176
+    },
+    {
+      "epoch": 1.5519205546186996,
+      "grad_norm": 0.20679518580436707,
+      "learning_rate": 6.725225225225225e-05,
+      "loss": 0.3418,
+      "step": 5177
+    },
+    {
+      "epoch": 1.5522203485103991,
+      "grad_norm": 0.2120271623134613,
+      "learning_rate": 6.72072072072072e-05,
+      "loss": 0.3554,
+      "step": 5178
+    },
+    {
+      "epoch": 1.5525201424020985,
+      "grad_norm": 0.2077147364616394,
+      "learning_rate": 6.716216216216215e-05,
+      "loss": 0.3508,
+      "step": 5179
+    },
+    {
+      "epoch": 1.552819936293798,
+      "grad_norm": 0.199779212474823,
+      "learning_rate": 6.711711711711711e-05,
+      "loss": 0.3535,
+      "step": 5180
+    },
+    {
+      "epoch": 1.5531197301854975,
+      "grad_norm": 0.20204727351665497,
+      "learning_rate": 6.707207207207206e-05,
+      "loss": 0.3523,
+      "step": 5181
+    },
+    {
+      "epoch": 1.553419524077197,
+      "grad_norm": 0.21412204205989838,
+      "learning_rate": 6.702702702702701e-05,
+      "loss": 0.3569,
+      "step": 5182
+    },
+    {
+      "epoch": 1.5537193179688964,
+      "grad_norm": 0.21525458991527557,
+      "learning_rate": 6.698198198198198e-05,
+      "loss": 0.3581,
+      "step": 5183
+    },
+    {
+      "epoch": 1.5540191118605957,
+      "grad_norm": 0.20961333811283112,
+      "learning_rate": 6.693693693693693e-05,
+      "loss": 0.3573,
+      "step": 5184
+    },
+    {
+      "epoch": 1.5543189057522953,
+      "grad_norm": 0.20356057584285736,
+      "learning_rate": 6.689189189189189e-05,
+      "loss": 0.3278,
+      "step": 5185
+    },
+    {
+      "epoch": 1.5546186996439948,
+      "grad_norm": 0.20768241584300995,
+      "learning_rate": 6.684684684684684e-05,
+      "loss": 0.3584,
+      "step": 5186
+    },
+    {
+      "epoch": 1.5549184935356943,
+      "grad_norm": 0.2090872973203659,
+      "learning_rate": 6.68018018018018e-05,
+      "loss": 0.3532,
+      "step": 5187
+    },
+    {
+      "epoch": 1.5552182874273937,
+      "grad_norm": 0.21495124697685242,
+      "learning_rate": 6.675675675675676e-05,
+      "loss": 0.3308,
+      "step": 5188
+    },
+    {
+      "epoch": 1.555518081319093,
+      "grad_norm": 0.19636289775371552,
+      "learning_rate": 6.67117117117117e-05,
+      "loss": 0.3379,
+      "step": 5189
+    },
+    {
+      "epoch": 1.5558178752107925,
+      "grad_norm": 0.20264169573783875,
+      "learning_rate": 6.666666666666666e-05,
+      "loss": 0.3437,
+      "step": 5190
+    },
+    {
+      "epoch": 1.556117669102492,
+      "grad_norm": 0.2122523933649063,
+      "learning_rate": 6.662162162162162e-05,
+      "loss": 0.364,
+      "step": 5191
+    },
+    {
+      "epoch": 1.5564174629941916,
+      "grad_norm": 0.208254873752594,
+      "learning_rate": 6.657657657657657e-05,
+      "loss": 0.3399,
+      "step": 5192
+    },
+    {
+      "epoch": 1.556717256885891,
+      "grad_norm": 0.21715570986270905,
+      "learning_rate": 6.653153153153152e-05,
+      "loss": 0.3684,
+      "step": 5193
+    },
+    {
+      "epoch": 1.5570170507775905,
+      "grad_norm": 0.21001993119716644,
+      "learning_rate": 6.648648648648648e-05,
+      "loss": 0.3473,
+      "step": 5194
+    },
+    {
+      "epoch": 1.5573168446692898,
+      "grad_norm": 0.20231187343597412,
+      "learning_rate": 6.644144144144144e-05,
+      "loss": 0.3334,
+      "step": 5195
+    },
+    {
+      "epoch": 1.5576166385609893,
+      "grad_norm": 0.20209276676177979,
+      "learning_rate": 6.639639639639639e-05,
+      "loss": 0.3453,
+      "step": 5196
+    },
+    {
+      "epoch": 1.5579164324526888,
+      "grad_norm": 0.1989961862564087,
+      "learning_rate": 6.635135135135134e-05,
+      "loss": 0.3562,
+      "step": 5197
+    },
+    {
+      "epoch": 1.5582162263443884,
+      "grad_norm": 0.19940902292728424,
+      "learning_rate": 6.63063063063063e-05,
+      "loss": 0.3448,
+      "step": 5198
+    },
+    {
+      "epoch": 1.5585160202360877,
+      "grad_norm": 0.19980254769325256,
+      "learning_rate": 6.626126126126125e-05,
+      "loss": 0.3424,
+      "step": 5199
+    },
+    {
+      "epoch": 1.558815814127787,
+      "grad_norm": 0.2070973515510559,
+      "learning_rate": 6.621621621621621e-05,
+      "loss": 0.3654,
+      "step": 5200
+    },
+    {
+      "epoch": 1.5591156080194866,
+      "grad_norm": 0.20411080121994019,
+      "learning_rate": 6.617117117117116e-05,
+      "loss": 0.3549,
+      "step": 5201
+    },
+    {
+      "epoch": 1.559415401911186,
+      "grad_norm": 0.20583924651145935,
+      "learning_rate": 6.612612612612613e-05,
+      "loss": 0.3567,
+      "step": 5202
+    },
+    {
+      "epoch": 1.5597151958028856,
+      "grad_norm": 0.2240842878818512,
+      "learning_rate": 6.608108108108108e-05,
+      "loss": 0.3793,
+      "step": 5203
+    },
+    {
+      "epoch": 1.560014989694585,
+      "grad_norm": 0.1986006647348404,
+      "learning_rate": 6.603603603603603e-05,
+      "loss": 0.3539,
+      "step": 5204
+    },
+    {
+      "epoch": 1.5603147835862843,
+      "grad_norm": 0.20004892349243164,
+      "learning_rate": 6.599099099099099e-05,
+      "loss": 0.3512,
+      "step": 5205
+    },
+    {
+      "epoch": 1.5606145774779838,
+      "grad_norm": 0.19041694700717926,
+      "learning_rate": 6.594594594594594e-05,
+      "loss": 0.3341,
+      "step": 5206
+    },
+    {
+      "epoch": 1.5609143713696834,
+      "grad_norm": 0.208636075258255,
+      "learning_rate": 6.590090090090089e-05,
+      "loss": 0.3746,
+      "step": 5207
+    },
+    {
+      "epoch": 1.561214165261383,
+      "grad_norm": 0.20262156426906586,
+      "learning_rate": 6.585585585585584e-05,
+      "loss": 0.3455,
+      "step": 5208
+    },
+    {
+      "epoch": 1.5615139591530822,
+      "grad_norm": 0.21155297756195068,
+      "learning_rate": 6.581081081081081e-05,
+      "loss": 0.3844,
+      "step": 5209
+    },
+    {
+      "epoch": 1.5618137530447818,
+      "grad_norm": 0.20110183954238892,
+      "learning_rate": 6.576576576576576e-05,
+      "loss": 0.3615,
+      "step": 5210
+    },
+    {
+      "epoch": 1.562113546936481,
+      "grad_norm": 0.20476794242858887,
+      "learning_rate": 6.572072072072071e-05,
+      "loss": 0.344,
+      "step": 5211
+    },
+    {
+      "epoch": 1.5624133408281806,
+      "grad_norm": 0.21922746300697327,
+      "learning_rate": 6.567567567567567e-05,
+      "loss": 0.3739,
+      "step": 5212
+    },
+    {
+      "epoch": 1.5627131347198802,
+      "grad_norm": 0.20494097471237183,
+      "learning_rate": 6.563063063063062e-05,
+      "loss": 0.3659,
+      "step": 5213
+    },
+    {
+      "epoch": 1.5630129286115797,
+      "grad_norm": 0.21062615513801575,
+      "learning_rate": 6.558558558558557e-05,
+      "loss": 0.3713,
+      "step": 5214
+    },
+    {
+      "epoch": 1.563312722503279,
+      "grad_norm": 0.22318191826343536,
+      "learning_rate": 6.554054054054054e-05,
+      "loss": 0.3676,
+      "step": 5215
+    },
+    {
+      "epoch": 1.5636125163949783,
+      "grad_norm": 0.19631384313106537,
+      "learning_rate": 6.549549549549549e-05,
+      "loss": 0.3633,
+      "step": 5216
+    },
+    {
+      "epoch": 1.5639123102866779,
+      "grad_norm": 0.2068559229373932,
+      "learning_rate": 6.545045045045045e-05,
+      "loss": 0.3758,
+      "step": 5217
+    },
+    {
+      "epoch": 1.5642121041783774,
+      "grad_norm": 0.2027987688779831,
+      "learning_rate": 6.54054054054054e-05,
+      "loss": 0.3514,
+      "step": 5218
+    },
+    {
+      "epoch": 1.564511898070077,
+      "grad_norm": 0.20169684290885925,
+      "learning_rate": 6.536036036036036e-05,
+      "loss": 0.3362,
+      "step": 5219
+    },
+    {
+      "epoch": 1.5648116919617763,
+      "grad_norm": 0.2035885900259018,
+      "learning_rate": 6.531531531531531e-05,
+      "loss": 0.351,
+      "step": 5220
+    },
+    {
+      "epoch": 1.5651114858534756,
+      "grad_norm": 0.20314770936965942,
+      "learning_rate": 6.527027027027027e-05,
+      "loss": 0.3644,
+      "step": 5221
+    },
+    {
+      "epoch": 1.5654112797451751,
+      "grad_norm": 0.19844964146614075,
+      "learning_rate": 6.522522522522522e-05,
+      "loss": 0.339,
+      "step": 5222
+    },
+    {
+      "epoch": 1.5657110736368747,
+      "grad_norm": 0.2092464417219162,
+      "learning_rate": 6.518018018018018e-05,
+      "loss": 0.3653,
+      "step": 5223
+    },
+    {
+      "epoch": 1.5660108675285742,
+      "grad_norm": 0.22174417972564697,
+      "learning_rate": 6.513513513513513e-05,
+      "loss": 0.3876,
+      "step": 5224
+    },
+    {
+      "epoch": 1.5663106614202735,
+      "grad_norm": 0.2033134251832962,
+      "learning_rate": 6.509009009009008e-05,
+      "loss": 0.3635,
+      "step": 5225
+    },
+    {
+      "epoch": 1.566610455311973,
+      "grad_norm": 0.20326606929302216,
+      "learning_rate": 6.504504504504503e-05,
+      "loss": 0.3377,
+      "step": 5226
+    },
+    {
+      "epoch": 1.5669102492036724,
+      "grad_norm": 0.2041843980550766,
+      "learning_rate": 6.5e-05,
+      "loss": 0.3396,
+      "step": 5227
+    },
+    {
+      "epoch": 1.567210043095372,
+      "grad_norm": 0.2067858725786209,
+      "learning_rate": 6.495495495495494e-05,
+      "loss": 0.3544,
+      "step": 5228
+    },
+    {
+      "epoch": 1.5675098369870715,
+      "grad_norm": 0.208845317363739,
+      "learning_rate": 6.490990990990991e-05,
+      "loss": 0.3445,
+      "step": 5229
+    },
+    {
+      "epoch": 1.567809630878771,
+      "grad_norm": 0.2178359478712082,
+      "learning_rate": 6.486486486486486e-05,
+      "loss": 0.3334,
+      "step": 5230
+    },
+    {
+      "epoch": 1.5681094247704703,
+      "grad_norm": 0.20035913586616516,
+      "learning_rate": 6.481981981981981e-05,
+      "loss": 0.3561,
+      "step": 5231
+    },
+    {
+      "epoch": 1.5684092186621696,
+      "grad_norm": 0.20526796579360962,
+      "learning_rate": 6.477477477477477e-05,
+      "loss": 0.3622,
+      "step": 5232
+    },
+    {
+      "epoch": 1.5687090125538692,
+      "grad_norm": 0.2058054804801941,
+      "learning_rate": 6.472972972972972e-05,
+      "loss": 0.3576,
+      "step": 5233
+    },
+    {
+      "epoch": 1.5690088064455687,
+      "grad_norm": 0.20559555292129517,
+      "learning_rate": 6.468468468468469e-05,
+      "loss": 0.3695,
+      "step": 5234
+    },
+    {
+      "epoch": 1.5693086003372683,
+      "grad_norm": 0.21687480807304382,
+      "learning_rate": 6.463963963963964e-05,
+      "loss": 0.3643,
+      "step": 5235
+    },
+    {
+      "epoch": 1.5696083942289676,
+      "grad_norm": 0.20145055651664734,
+      "learning_rate": 6.459459459459459e-05,
+      "loss": 0.342,
+      "step": 5236
+    },
+    {
+      "epoch": 1.569908188120667,
+      "grad_norm": 0.2083156406879425,
+      "learning_rate": 6.454954954954955e-05,
+      "loss": 0.3694,
+      "step": 5237
+    },
+    {
+      "epoch": 1.5702079820123664,
+      "grad_norm": 0.2008083611726761,
+      "learning_rate": 6.45045045045045e-05,
+      "loss": 0.358,
+      "step": 5238
+    },
+    {
+      "epoch": 1.570507775904066,
+      "grad_norm": 0.20801851153373718,
+      "learning_rate": 6.445945945945945e-05,
+      "loss": 0.3444,
+      "step": 5239
+    },
+    {
+      "epoch": 1.5708075697957655,
+      "grad_norm": 0.20249147713184357,
+      "learning_rate": 6.44144144144144e-05,
+      "loss": 0.3594,
+      "step": 5240
+    },
+    {
+      "epoch": 1.5711073636874648,
+      "grad_norm": 0.20679299533367157,
+      "learning_rate": 6.436936936936937e-05,
+      "loss": 0.3604,
+      "step": 5241
+    },
+    {
+      "epoch": 1.5714071575791642,
+      "grad_norm": 0.20185597240924835,
+      "learning_rate": 6.432432432432432e-05,
+      "loss": 0.3361,
+      "step": 5242
+    },
+    {
+      "epoch": 1.5717069514708637,
+      "grad_norm": 0.20025381445884705,
+      "learning_rate": 6.427927927927927e-05,
+      "loss": 0.3515,
+      "step": 5243
+    },
+    {
+      "epoch": 1.5720067453625632,
+      "grad_norm": 0.19662436842918396,
+      "learning_rate": 6.423423423423423e-05,
+      "loss": 0.3188,
+      "step": 5244
+    },
+    {
+      "epoch": 1.5723065392542628,
+      "grad_norm": 0.20173777639865875,
+      "learning_rate": 6.418918918918918e-05,
+      "loss": 0.3509,
+      "step": 5245
+    },
+    {
+      "epoch": 1.572606333145962,
+      "grad_norm": 0.21695034205913544,
+      "learning_rate": 6.414414414414413e-05,
+      "loss": 0.3589,
+      "step": 5246
+    },
+    {
+      "epoch": 1.5729061270376616,
+      "grad_norm": 0.21321168541908264,
+      "learning_rate": 6.40990990990991e-05,
+      "loss": 0.371,
+      "step": 5247
+    },
+    {
+      "epoch": 1.573205920929361,
+      "grad_norm": 0.21553568542003632,
+      "learning_rate": 6.405405405405405e-05,
+      "loss": 0.3673,
+      "step": 5248
+    },
+    {
+      "epoch": 1.5735057148210605,
+      "grad_norm": 0.21144214272499084,
+      "learning_rate": 6.400900900900901e-05,
+      "loss": 0.3649,
+      "step": 5249
+    },
+    {
+      "epoch": 1.57380550871276,
+      "grad_norm": 0.20765072107315063,
+      "learning_rate": 6.396396396396396e-05,
+      "loss": 0.3642,
+      "step": 5250
+    },
+    {
+      "epoch": 1.5741053026044596,
+      "grad_norm": 0.2129376381635666,
+      "learning_rate": 6.391891891891891e-05,
+      "loss": 0.3612,
+      "step": 5251
+    },
+    {
+      "epoch": 1.574405096496159,
+      "grad_norm": 0.20372062921524048,
+      "learning_rate": 6.387387387387387e-05,
+      "loss": 0.3579,
+      "step": 5252
+    },
+    {
+      "epoch": 1.5747048903878582,
+      "grad_norm": 0.1967153400182724,
+      "learning_rate": 6.382882882882882e-05,
+      "loss": 0.3413,
+      "step": 5253
+    },
+    {
+      "epoch": 1.5750046842795578,
+      "grad_norm": 0.19723547995090485,
+      "learning_rate": 6.378378378378377e-05,
+      "loss": 0.3423,
+      "step": 5254
+    },
+    {
+      "epoch": 1.5753044781712573,
+      "grad_norm": 0.20694804191589355,
+      "learning_rate": 6.373873873873874e-05,
+      "loss": 0.3431,
+      "step": 5255
+    },
+    {
+      "epoch": 1.5756042720629568,
+      "grad_norm": 0.209686741232872,
+      "learning_rate": 6.369369369369369e-05,
+      "loss": 0.3634,
+      "step": 5256
+    },
+    {
+      "epoch": 1.5759040659546562,
+      "grad_norm": 0.20350812375545502,
+      "learning_rate": 6.364864864864864e-05,
+      "loss": 0.3499,
+      "step": 5257
+    },
+    {
+      "epoch": 1.5762038598463555,
+      "grad_norm": 0.19632326066493988,
+      "learning_rate": 6.360360360360359e-05,
+      "loss": 0.3564,
+      "step": 5258
+    },
+    {
+      "epoch": 1.576503653738055,
+      "grad_norm": 0.19848290085792542,
+      "learning_rate": 6.355855855855855e-05,
+      "loss": 0.3611,
+      "step": 5259
+    },
+    {
+      "epoch": 1.5768034476297546,
+      "grad_norm": 0.22999663650989532,
+      "learning_rate": 6.35135135135135e-05,
+      "loss": 0.3523,
+      "step": 5260
+    },
+    {
+      "epoch": 1.577103241521454,
+      "grad_norm": 0.20401522517204285,
+      "learning_rate": 6.346846846846847e-05,
+      "loss": 0.3361,
+      "step": 5261
+    },
+    {
+      "epoch": 1.5774030354131534,
+      "grad_norm": 0.19645123183727264,
+      "learning_rate": 6.342342342342342e-05,
+      "loss": 0.3451,
+      "step": 5262
+    },
+    {
+      "epoch": 1.577702829304853,
+      "grad_norm": 0.2110058218240738,
+      "learning_rate": 6.337837837837837e-05,
+      "loss": 0.3687,
+      "step": 5263
+    },
+    {
+      "epoch": 1.5780026231965523,
+      "grad_norm": 0.20960693061351776,
+      "learning_rate": 6.333333333333333e-05,
+      "loss": 0.3732,
+      "step": 5264
+    },
+    {
+      "epoch": 1.5783024170882518,
+      "grad_norm": 0.20382042229175568,
+      "learning_rate": 6.328828828828828e-05,
+      "loss": 0.3454,
+      "step": 5265
+    },
+    {
+      "epoch": 1.5786022109799513,
+      "grad_norm": 0.20214159786701202,
+      "learning_rate": 6.324324324324325e-05,
+      "loss": 0.3532,
+      "step": 5266
+    },
+    {
+      "epoch": 1.5789020048716509,
+      "grad_norm": 0.20783282816410065,
+      "learning_rate": 6.31981981981982e-05,
+      "loss": 0.3861,
+      "step": 5267
+    },
+    {
+      "epoch": 1.5792017987633502,
+      "grad_norm": 0.19975163042545319,
+      "learning_rate": 6.315315315315315e-05,
+      "loss": 0.3567,
+      "step": 5268
+    },
+    {
+      "epoch": 1.5795015926550495,
+      "grad_norm": 0.20556297898292542,
+      "learning_rate": 6.31081081081081e-05,
+      "loss": 0.3658,
+      "step": 5269
+    },
+    {
+      "epoch": 1.579801386546749,
+      "grad_norm": 0.20620745420455933,
+      "learning_rate": 6.306306306306306e-05,
+      "loss": 0.3593,
+      "step": 5270
+    },
+    {
+      "epoch": 1.5801011804384486,
+      "grad_norm": 0.2031751573085785,
+      "learning_rate": 6.301801801801801e-05,
+      "loss": 0.3426,
+      "step": 5271
+    },
+    {
+      "epoch": 1.5804009743301481,
+      "grad_norm": 0.2001752108335495,
+      "learning_rate": 6.297297297297296e-05,
+      "loss": 0.3696,
+      "step": 5272
+    },
+    {
+      "epoch": 1.5807007682218475,
+      "grad_norm": 0.20031659305095673,
+      "learning_rate": 6.292792792792793e-05,
+      "loss": 0.3616,
+      "step": 5273
+    },
+    {
+      "epoch": 1.5810005621135468,
+      "grad_norm": 0.20846503973007202,
+      "learning_rate": 6.288288288288288e-05,
+      "loss": 0.3689,
+      "step": 5274
+    },
+    {
+      "epoch": 1.5813003560052463,
+      "grad_norm": 0.21977601945400238,
+      "learning_rate": 6.283783783783783e-05,
+      "loss": 0.374,
+      "step": 5275
+    },
+    {
+      "epoch": 1.5816001498969459,
+      "grad_norm": 0.20855718851089478,
+      "learning_rate": 6.279279279279279e-05,
+      "loss": 0.3654,
+      "step": 5276
+    },
+    {
+      "epoch": 1.5818999437886454,
+      "grad_norm": 0.22264760732650757,
+      "learning_rate": 6.274774774774774e-05,
+      "loss": 0.3838,
+      "step": 5277
+    },
+    {
+      "epoch": 1.5821997376803447,
+      "grad_norm": 0.2073168158531189,
+      "learning_rate": 6.27027027027027e-05,
+      "loss": 0.3424,
+      "step": 5278
+    },
+    {
+      "epoch": 1.5824995315720443,
+      "grad_norm": 0.20966382324695587,
+      "learning_rate": 6.265765765765765e-05,
+      "loss": 0.3666,
+      "step": 5279
+    },
+    {
+      "epoch": 1.5827993254637436,
+      "grad_norm": 0.19850985705852509,
+      "learning_rate": 6.26126126126126e-05,
+      "loss": 0.3304,
+      "step": 5280
+    },
+    {
+      "epoch": 1.5830991193554431,
+      "grad_norm": 0.2018197774887085,
+      "learning_rate": 6.256756756756757e-05,
+      "loss": 0.3496,
+      "step": 5281
+    },
+    {
+      "epoch": 1.5833989132471427,
+      "grad_norm": 0.21336401998996735,
+      "learning_rate": 6.252252252252252e-05,
+      "loss": 0.3907,
+      "step": 5282
+    },
+    {
+      "epoch": 1.5836987071388422,
+      "grad_norm": 0.2028983235359192,
+      "learning_rate": 6.247747747747747e-05,
+      "loss": 0.3528,
+      "step": 5283
+    },
+    {
+      "epoch": 1.5839985010305415,
+      "grad_norm": 0.20132046937942505,
+      "learning_rate": 6.243243243243243e-05,
+      "loss": 0.3358,
+      "step": 5284
+    },
+    {
+      "epoch": 1.5842982949222408,
+      "grad_norm": 0.2028990089893341,
+      "learning_rate": 6.238738738738738e-05,
+      "loss": 0.3534,
+      "step": 5285
+    },
+    {
+      "epoch": 1.5845980888139404,
+      "grad_norm": 0.20739594101905823,
+      "learning_rate": 6.234234234234233e-05,
+      "loss": 0.3595,
+      "step": 5286
+    },
+    {
+      "epoch": 1.58489788270564,
+      "grad_norm": 0.2152402400970459,
+      "learning_rate": 6.229729729729728e-05,
+      "loss": 0.3585,
+      "step": 5287
+    },
+    {
+      "epoch": 1.5851976765973395,
+      "grad_norm": 0.2000325322151184,
+      "learning_rate": 6.225225225225225e-05,
+      "loss": 0.3307,
+      "step": 5288
+    },
+    {
+      "epoch": 1.5854974704890388,
+      "grad_norm": 0.1998959332704544,
+      "learning_rate": 6.22072072072072e-05,
+      "loss": 0.3517,
+      "step": 5289
+    },
+    {
+      "epoch": 1.585797264380738,
+      "grad_norm": 0.1964779496192932,
+      "learning_rate": 6.216216216216215e-05,
+      "loss": 0.331,
+      "step": 5290
+    },
+    {
+      "epoch": 1.5860970582724376,
+      "grad_norm": 0.21091748774051666,
+      "learning_rate": 6.211711711711711e-05,
+      "loss": 0.3483,
+      "step": 5291
+    },
+    {
+      "epoch": 1.5863968521641372,
+      "grad_norm": 0.20698702335357666,
+      "learning_rate": 6.207207207207206e-05,
+      "loss": 0.3513,
+      "step": 5292
+    },
+    {
+      "epoch": 1.5866966460558367,
+      "grad_norm": 0.21151360869407654,
+      "learning_rate": 6.202702702702703e-05,
+      "loss": 0.3753,
+      "step": 5293
+    },
+    {
+      "epoch": 1.586996439947536,
+      "grad_norm": 0.20693033933639526,
+      "learning_rate": 6.198198198198198e-05,
+      "loss": 0.3428,
+      "step": 5294
+    },
+    {
+      "epoch": 1.5872962338392356,
+      "grad_norm": 0.20240649580955505,
+      "learning_rate": 6.193693693693693e-05,
+      "loss": 0.3344,
+      "step": 5295
+    },
+    {
+      "epoch": 1.5875960277309349,
+      "grad_norm": 0.2075674682855606,
+      "learning_rate": 6.189189189189189e-05,
+      "loss": 0.358,
+      "step": 5296
+    },
+    {
+      "epoch": 1.5878958216226344,
+      "grad_norm": 0.2162725031375885,
+      "learning_rate": 6.184684684684684e-05,
+      "loss": 0.3737,
+      "step": 5297
+    },
+    {
+      "epoch": 1.588195615514334,
+      "grad_norm": 0.20333421230316162,
+      "learning_rate": 6.180180180180179e-05,
+      "loss": 0.3486,
+      "step": 5298
+    },
+    {
+      "epoch": 1.5884954094060335,
+      "grad_norm": 0.2127985805273056,
+      "learning_rate": 6.175675675675676e-05,
+      "loss": 0.3723,
+      "step": 5299
+    },
+    {
+      "epoch": 1.5887952032977328,
+      "grad_norm": 0.20037934184074402,
+      "learning_rate": 6.17117117117117e-05,
+      "loss": 0.3639,
+      "step": 5300
+    },
+    {
+      "epoch": 1.5890949971894321,
+      "grad_norm": 0.21006622910499573,
+      "learning_rate": 6.166666666666666e-05,
+      "loss": 0.3539,
+      "step": 5301
+    },
+    {
+      "epoch": 1.5893947910811317,
+      "grad_norm": 0.21559639275074005,
+      "learning_rate": 6.162162162162162e-05,
+      "loss": 0.3706,
+      "step": 5302
+    },
+    {
+      "epoch": 1.5896945849728312,
+      "grad_norm": 0.19809530675411224,
+      "learning_rate": 6.157657657657657e-05,
+      "loss": 0.3262,
+      "step": 5303
+    },
+    {
+      "epoch": 1.5899943788645308,
+      "grad_norm": 0.2111746221780777,
+      "learning_rate": 6.153153153153152e-05,
+      "loss": 0.367,
+      "step": 5304
+    },
+    {
+      "epoch": 1.59029417275623,
+      "grad_norm": 0.2136707603931427,
+      "learning_rate": 6.148648648648647e-05,
+      "loss": 0.3775,
+      "step": 5305
+    },
+    {
+      "epoch": 1.5905939666479294,
+      "grad_norm": 0.21586540341377258,
+      "learning_rate": 6.144144144144144e-05,
+      "loss": 0.3507,
+      "step": 5306
+    },
+    {
+      "epoch": 1.590893760539629,
+      "grad_norm": 0.21576684713363647,
+      "learning_rate": 6.139639639639639e-05,
+      "loss": 0.3788,
+      "step": 5307
+    },
+    {
+      "epoch": 1.5911935544313285,
+      "grad_norm": 0.2173517942428589,
+      "learning_rate": 6.135135135135135e-05,
+      "loss": 0.3512,
+      "step": 5308
+    },
+    {
+      "epoch": 1.591493348323028,
+      "grad_norm": 0.19617997109889984,
+      "learning_rate": 6.13063063063063e-05,
+      "loss": 0.3464,
+      "step": 5309
+    },
+    {
+      "epoch": 1.5917931422147273,
+      "grad_norm": 0.1915069967508316,
+      "learning_rate": 6.126126126126126e-05,
+      "loss": 0.3131,
+      "step": 5310
+    },
+    {
+      "epoch": 1.5920929361064269,
+      "grad_norm": 0.21343593299388885,
+      "learning_rate": 6.121621621621621e-05,
+      "loss": 0.3542,
+      "step": 5311
+    },
+    {
+      "epoch": 1.5923927299981262,
+      "grad_norm": 0.19464311003684998,
+      "learning_rate": 6.117117117117116e-05,
+      "loss": 0.3474,
+      "step": 5312
+    },
+    {
+      "epoch": 1.5926925238898257,
+      "grad_norm": 0.2090366929769516,
+      "learning_rate": 6.112612612612613e-05,
+      "loss": 0.3418,
+      "step": 5313
+    },
+    {
+      "epoch": 1.5929923177815253,
+      "grad_norm": 0.2039409577846527,
+      "learning_rate": 6.108108108108108e-05,
+      "loss": 0.3397,
+      "step": 5314
+    },
+    {
+      "epoch": 1.5932921116732248,
+      "grad_norm": 0.1947229951620102,
+      "learning_rate": 6.103603603603603e-05,
+      "loss": 0.328,
+      "step": 5315
+    },
+    {
+      "epoch": 1.5935919055649241,
+      "grad_norm": 0.20912787318229675,
+      "learning_rate": 6.099099099099098e-05,
+      "loss": 0.344,
+      "step": 5316
+    },
+    {
+      "epoch": 1.5938916994566235,
+      "grad_norm": 0.21064409613609314,
+      "learning_rate": 6.094594594594594e-05,
+      "loss": 0.3577,
+      "step": 5317
+    },
+    {
+      "epoch": 1.594191493348323,
+      "grad_norm": 0.2055872529745102,
+      "learning_rate": 6.0900900900900893e-05,
+      "loss": 0.365,
+      "step": 5318
+    },
+    {
+      "epoch": 1.5944912872400225,
+      "grad_norm": 0.20798040926456451,
+      "learning_rate": 6.085585585585585e-05,
+      "loss": 0.3503,
+      "step": 5319
+    },
+    {
+      "epoch": 1.594791081131722,
+      "grad_norm": 0.20974093675613403,
+      "learning_rate": 6.081081081081081e-05,
+      "loss": 0.3587,
+      "step": 5320
+    },
+    {
+      "epoch": 1.5950908750234214,
+      "grad_norm": 0.19876185059547424,
+      "learning_rate": 6.0765765765765765e-05,
+      "loss": 0.3497,
+      "step": 5321
+    },
+    {
+      "epoch": 1.5953906689151207,
+      "grad_norm": 0.1975713074207306,
+      "learning_rate": 6.0720720720720715e-05,
+      "loss": 0.3295,
+      "step": 5322
+    },
+    {
+      "epoch": 1.5956904628068203,
+      "grad_norm": 0.20481306314468384,
+      "learning_rate": 6.0675675675675665e-05,
+      "loss": 0.3272,
+      "step": 5323
+    },
+    {
+      "epoch": 1.5959902566985198,
+      "grad_norm": 0.20750102400779724,
+      "learning_rate": 6.063063063063063e-05,
+      "loss": 0.356,
+      "step": 5324
+    },
+    {
+      "epoch": 1.5962900505902193,
+      "grad_norm": 0.21194814145565033,
+      "learning_rate": 6.058558558558558e-05,
+      "loss": 0.3695,
+      "step": 5325
+    },
+    {
+      "epoch": 1.5965898444819187,
+      "grad_norm": 0.1984264999628067,
+      "learning_rate": 6.054054054054053e-05,
+      "loss": 0.3514,
+      "step": 5326
+    },
+    {
+      "epoch": 1.5968896383736182,
+      "grad_norm": 0.2006942182779312,
+      "learning_rate": 6.0495495495495494e-05,
+      "loss": 0.3538,
+      "step": 5327
+    },
+    {
+      "epoch": 1.5971894322653175,
+      "grad_norm": 0.20723217725753784,
+      "learning_rate": 6.0450450450450444e-05,
+      "loss": 0.3668,
+      "step": 5328
+    },
+    {
+      "epoch": 1.597489226157017,
+      "grad_norm": 0.21457070112228394,
+      "learning_rate": 6.04054054054054e-05,
+      "loss": 0.3976,
+      "step": 5329
+    },
+    {
+      "epoch": 1.5977890200487166,
+      "grad_norm": 0.19332286715507507,
+      "learning_rate": 6.036036036036035e-05,
+      "loss": 0.3395,
+      "step": 5330
+    },
+    {
+      "epoch": 1.598088813940416,
+      "grad_norm": 0.20550230145454407,
+      "learning_rate": 6.0315315315315315e-05,
+      "loss": 0.3395,
+      "step": 5331
+    },
+    {
+      "epoch": 1.5983886078321154,
+      "grad_norm": 0.19099922478199005,
+      "learning_rate": 6.0270270270270266e-05,
+      "loss": 0.3172,
+      "step": 5332
+    },
+    {
+      "epoch": 1.5986884017238148,
+      "grad_norm": 0.20576585829257965,
+      "learning_rate": 6.0225225225225216e-05,
+      "loss": 0.3544,
+      "step": 5333
+    },
+    {
+      "epoch": 1.5989881956155143,
+      "grad_norm": 0.20336727797985077,
+      "learning_rate": 6.018018018018017e-05,
+      "loss": 0.3568,
+      "step": 5334
+    },
+    {
+      "epoch": 1.5992879895072138,
+      "grad_norm": 0.21562445163726807,
+      "learning_rate": 6.013513513513513e-05,
+      "loss": 0.363,
+      "step": 5335
+    },
+    {
+      "epoch": 1.5995877833989134,
+      "grad_norm": 0.20894883573055267,
+      "learning_rate": 6.009009009009009e-05,
+      "loss": 0.357,
+      "step": 5336
+    },
+    {
+      "epoch": 1.5998875772906127,
+      "grad_norm": 0.21003037691116333,
+      "learning_rate": 6.004504504504504e-05,
+      "loss": 0.349,
+      "step": 5337
+    },
+    {
+      "epoch": 1.600187371182312,
+      "grad_norm": 0.1921328455209732,
+      "learning_rate": 5.9999999999999995e-05,
+      "loss": 0.3329,
+      "step": 5338
+    },
+    {
+      "epoch": 1.6004871650740116,
+      "grad_norm": 0.21384379267692566,
+      "learning_rate": 5.995495495495495e-05,
+      "loss": 0.361,
+      "step": 5339
+    },
+    {
+      "epoch": 1.600786958965711,
+      "grad_norm": 0.2054745852947235,
+      "learning_rate": 5.99099099099099e-05,
+      "loss": 0.3383,
+      "step": 5340
+    },
+    {
+      "epoch": 1.6010867528574106,
+      "grad_norm": 0.20083633065223694,
+      "learning_rate": 5.986486486486486e-05,
+      "loss": 0.3442,
+      "step": 5341
+    },
+    {
+      "epoch": 1.60138654674911,
+      "grad_norm": 0.19561520218849182,
+      "learning_rate": 5.9819819819819816e-05,
+      "loss": 0.3416,
+      "step": 5342
+    },
+    {
+      "epoch": 1.6016863406408093,
+      "grad_norm": 0.22439317405223846,
+      "learning_rate": 5.977477477477477e-05,
+      "loss": 0.3699,
+      "step": 5343
+    },
+    {
+      "epoch": 1.6019861345325088,
+      "grad_norm": 0.20617817342281342,
+      "learning_rate": 5.9729729729729724e-05,
+      "loss": 0.3572,
+      "step": 5344
+    },
+    {
+      "epoch": 1.6022859284242084,
+      "grad_norm": 0.21624356508255005,
+      "learning_rate": 5.968468468468468e-05,
+      "loss": 0.3911,
+      "step": 5345
+    },
+    {
+      "epoch": 1.602585722315908,
+      "grad_norm": 0.21435832977294922,
+      "learning_rate": 5.963963963963964e-05,
+      "loss": 0.3497,
+      "step": 5346
+    },
+    {
+      "epoch": 1.6028855162076072,
+      "grad_norm": 0.20786425471305847,
+      "learning_rate": 5.959459459459459e-05,
+      "loss": 0.335,
+      "step": 5347
+    },
+    {
+      "epoch": 1.6031853100993068,
+      "grad_norm": 0.20502492785453796,
+      "learning_rate": 5.954954954954954e-05,
+      "loss": 0.3595,
+      "step": 5348
+    },
+    {
+      "epoch": 1.603485103991006,
+      "grad_norm": 0.20398326218128204,
+      "learning_rate": 5.95045045045045e-05,
+      "loss": 0.3562,
+      "step": 5349
+    },
+    {
+      "epoch": 1.6037848978827056,
+      "grad_norm": 0.19771265983581543,
+      "learning_rate": 5.945945945945945e-05,
+      "loss": 0.338,
+      "step": 5350
+    },
+    {
+      "epoch": 1.6040846917744052,
+      "grad_norm": 0.20617644488811493,
+      "learning_rate": 5.941441441441441e-05,
+      "loss": 0.3617,
+      "step": 5351
+    },
+    {
+      "epoch": 1.6043844856661047,
+      "grad_norm": 0.20850981771945953,
+      "learning_rate": 5.936936936936936e-05,
+      "loss": 0.3757,
+      "step": 5352
+    },
+    {
+      "epoch": 1.604684279557804,
+      "grad_norm": 0.20652638375759125,
+      "learning_rate": 5.9324324324324324e-05,
+      "loss": 0.3576,
+      "step": 5353
+    },
+    {
+      "epoch": 1.6049840734495033,
+      "grad_norm": 0.20817598700523376,
+      "learning_rate": 5.9279279279279274e-05,
+      "loss": 0.3645,
+      "step": 5354
+    },
+    {
+      "epoch": 1.6052838673412029,
+      "grad_norm": 0.19501417875289917,
+      "learning_rate": 5.9234234234234225e-05,
+      "loss": 0.346,
+      "step": 5355
+    },
+    {
+      "epoch": 1.6055836612329024,
+      "grad_norm": 0.20347361266613007,
+      "learning_rate": 5.918918918918919e-05,
+      "loss": 0.3521,
+      "step": 5356
+    },
+    {
+      "epoch": 1.605883455124602,
+      "grad_norm": 0.19727610051631927,
+      "learning_rate": 5.914414414414414e-05,
+      "loss": 0.3488,
+      "step": 5357
+    },
+    {
+      "epoch": 1.6061832490163013,
+      "grad_norm": 0.1962079256772995,
+      "learning_rate": 5.909909909909909e-05,
+      "loss": 0.3359,
+      "step": 5358
+    },
+    {
+      "epoch": 1.6064830429080006,
+      "grad_norm": 0.19446706771850586,
+      "learning_rate": 5.9054054054054046e-05,
+      "loss": 0.3439,
+      "step": 5359
+    },
+    {
+      "epoch": 1.6067828367997001,
+      "grad_norm": 0.2100493609905243,
+      "learning_rate": 5.9009009009009003e-05,
+      "loss": 0.3559,
+      "step": 5360
+    },
+    {
+      "epoch": 1.6070826306913997,
+      "grad_norm": 0.20190827548503876,
+      "learning_rate": 5.896396396396396e-05,
+      "loss": 0.3572,
+      "step": 5361
+    },
+    {
+      "epoch": 1.6073824245830992,
+      "grad_norm": 0.19307351112365723,
+      "learning_rate": 5.891891891891891e-05,
+      "loss": 0.3234,
+      "step": 5362
+    },
+    {
+      "epoch": 1.6076822184747985,
+      "grad_norm": 0.2018936276435852,
+      "learning_rate": 5.8873873873873875e-05,
+      "loss": 0.3509,
+      "step": 5363
+    },
+    {
+      "epoch": 1.607982012366498,
+      "grad_norm": 0.19602860510349274,
+      "learning_rate": 5.8828828828828825e-05,
+      "loss": 0.3462,
+      "step": 5364
+    },
+    {
+      "epoch": 1.6082818062581974,
+      "grad_norm": 0.19937603175640106,
+      "learning_rate": 5.8783783783783775e-05,
+      "loss": 0.3514,
+      "step": 5365
+    },
+    {
+      "epoch": 1.608581600149897,
+      "grad_norm": 0.20160555839538574,
+      "learning_rate": 5.873873873873873e-05,
+      "loss": 0.3628,
+      "step": 5366
+    },
+    {
+      "epoch": 1.6088813940415965,
+      "grad_norm": 0.20256981253623962,
+      "learning_rate": 5.869369369369369e-05,
+      "loss": 0.3315,
+      "step": 5367
+    },
+    {
+      "epoch": 1.609181187933296,
+      "grad_norm": 0.21089547872543335,
+      "learning_rate": 5.864864864864865e-05,
+      "loss": 0.3704,
+      "step": 5368
+    },
+    {
+      "epoch": 1.6094809818249953,
+      "grad_norm": 0.20721639692783356,
+      "learning_rate": 5.86036036036036e-05,
+      "loss": 0.3485,
+      "step": 5369
+    },
+    {
+      "epoch": 1.6097807757166946,
+      "grad_norm": 0.19964303076267242,
+      "learning_rate": 5.855855855855855e-05,
+      "loss": 0.3449,
+      "step": 5370
+    },
+    {
+      "epoch": 1.6100805696083942,
+      "grad_norm": 0.21870988607406616,
+      "learning_rate": 5.851351351351351e-05,
+      "loss": 0.3566,
+      "step": 5371
+    },
+    {
+      "epoch": 1.6103803635000937,
+      "grad_norm": 0.21295110881328583,
+      "learning_rate": 5.846846846846846e-05,
+      "loss": 0.3539,
+      "step": 5372
+    },
+    {
+      "epoch": 1.6106801573917933,
+      "grad_norm": 0.19819046556949615,
+      "learning_rate": 5.842342342342342e-05,
+      "loss": 0.3508,
+      "step": 5373
+    },
+    {
+      "epoch": 1.6109799512834926,
+      "grad_norm": 0.21221822500228882,
+      "learning_rate": 5.8378378378378376e-05,
+      "loss": 0.3716,
+      "step": 5374
+    },
+    {
+      "epoch": 1.611279745175192,
+      "grad_norm": 0.20993143320083618,
+      "learning_rate": 5.8333333333333326e-05,
+      "loss": 0.3714,
+      "step": 5375
+    },
+    {
+      "epoch": 1.6115795390668914,
+      "grad_norm": 0.20888566970825195,
+      "learning_rate": 5.828828828828828e-05,
+      "loss": 0.3561,
+      "step": 5376
+    },
+    {
+      "epoch": 1.611879332958591,
+      "grad_norm": 0.1925889551639557,
+      "learning_rate": 5.824324324324323e-05,
+      "loss": 0.3458,
+      "step": 5377
+    },
+    {
+      "epoch": 1.6121791268502905,
+      "grad_norm": 0.19851090013980865,
+      "learning_rate": 5.81981981981982e-05,
+      "loss": 0.3354,
+      "step": 5378
+    },
+    {
+      "epoch": 1.6124789207419898,
+      "grad_norm": 0.20619840919971466,
+      "learning_rate": 5.815315315315315e-05,
+      "loss": 0.3571,
+      "step": 5379
+    },
+    {
+      "epoch": 1.6127787146336894,
+      "grad_norm": 0.20241157710552216,
+      "learning_rate": 5.81081081081081e-05,
+      "loss": 0.3478,
+      "step": 5380
+    },
+    {
+      "epoch": 1.6130785085253887,
+      "grad_norm": 0.2011449635028839,
+      "learning_rate": 5.806306306306306e-05,
+      "loss": 0.354,
+      "step": 5381
+    },
+    {
+      "epoch": 1.6133783024170882,
+      "grad_norm": 0.19988194108009338,
+      "learning_rate": 5.801801801801801e-05,
+      "loss": 0.3399,
+      "step": 5382
+    },
+    {
+      "epoch": 1.6136780963087878,
+      "grad_norm": 0.21191838383674622,
+      "learning_rate": 5.797297297297297e-05,
+      "loss": 0.3748,
+      "step": 5383
+    },
+    {
+      "epoch": 1.6139778902004873,
+      "grad_norm": 0.20201589167118073,
+      "learning_rate": 5.792792792792792e-05,
+      "loss": 0.3636,
+      "step": 5384
+    },
+    {
+      "epoch": 1.6142776840921866,
+      "grad_norm": 0.22016596794128418,
+      "learning_rate": 5.788288288288288e-05,
+      "loss": 0.3713,
+      "step": 5385
+    },
+    {
+      "epoch": 1.614577477983886,
+      "grad_norm": 0.20050211250782013,
+      "learning_rate": 5.7837837837837834e-05,
+      "loss": 0.337,
+      "step": 5386
+    },
+    {
+      "epoch": 1.6148772718755855,
+      "grad_norm": 0.20136626064777374,
+      "learning_rate": 5.7792792792792784e-05,
+      "loss": 0.3672,
+      "step": 5387
+    },
+    {
+      "epoch": 1.615177065767285,
+      "grad_norm": 0.20241838693618774,
+      "learning_rate": 5.774774774774775e-05,
+      "loss": 0.3585,
+      "step": 5388
+    },
+    {
+      "epoch": 1.6154768596589846,
+      "grad_norm": 0.20410926640033722,
+      "learning_rate": 5.77027027027027e-05,
+      "loss": 0.352,
+      "step": 5389
+    },
+    {
+      "epoch": 1.615776653550684,
+      "grad_norm": 0.20114073157310486,
+      "learning_rate": 5.7657657657657655e-05,
+      "loss": 0.3386,
+      "step": 5390
+    },
+    {
+      "epoch": 1.6160764474423832,
+      "grad_norm": 0.20846368372440338,
+      "learning_rate": 5.7612612612612606e-05,
+      "loss": 0.3623,
+      "step": 5391
+    },
+    {
+      "epoch": 1.6163762413340828,
+      "grad_norm": 0.20690296590328217,
+      "learning_rate": 5.756756756756756e-05,
+      "loss": 0.3526,
+      "step": 5392
+    },
+    {
+      "epoch": 1.6166760352257823,
+      "grad_norm": 0.21066324412822723,
+      "learning_rate": 5.752252252252252e-05,
+      "loss": 0.3576,
+      "step": 5393
+    },
+    {
+      "epoch": 1.6169758291174818,
+      "grad_norm": 0.1903742104768753,
+      "learning_rate": 5.747747747747747e-05,
+      "loss": 0.3385,
+      "step": 5394
+    },
+    {
+      "epoch": 1.6172756230091812,
+      "grad_norm": 0.19906085729599,
+      "learning_rate": 5.743243243243242e-05,
+      "loss": 0.333,
+      "step": 5395
+    },
+    {
+      "epoch": 1.6175754169008807,
+      "grad_norm": 0.20286022126674652,
+      "learning_rate": 5.7387387387387384e-05,
+      "loss": 0.3405,
+      "step": 5396
+    },
+    {
+      "epoch": 1.61787521079258,
+      "grad_norm": 0.2012995481491089,
+      "learning_rate": 5.7342342342342335e-05,
+      "loss": 0.3622,
+      "step": 5397
+    },
+    {
+      "epoch": 1.6181750046842795,
+      "grad_norm": 0.20434437692165375,
+      "learning_rate": 5.729729729729729e-05,
+      "loss": 0.3456,
+      "step": 5398
+    },
+    {
+      "epoch": 1.618474798575979,
+      "grad_norm": 0.21659024059772491,
+      "learning_rate": 5.725225225225225e-05,
+      "loss": 0.3764,
+      "step": 5399
+    },
+    {
+      "epoch": 1.6187745924676786,
+      "grad_norm": 0.20373186469078064,
+      "learning_rate": 5.7207207207207206e-05,
+      "loss": 0.3489,
+      "step": 5400
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 6670,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 200,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.4368809647228741e+20,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}