diff --git "a/adapter_info/trainer_state.json" "b/adapter_info/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/adapter_info/trainer_state.json"
@@ -0,0 +1,10948 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 1558,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0006418485237483953,
+      "grad_norm": 0.32000207901000977,
+      "learning_rate": 2.1276595744680853e-06,
+      "loss": 9.1875,
+      "step": 1
+    },
+    {
+      "epoch": 0.0012836970474967907,
+      "grad_norm": 0.31870341300964355,
+      "learning_rate": 4.255319148936171e-06,
+      "loss": 9.25,
+      "step": 2
+    },
+    {
+      "epoch": 0.0019255455712451862,
+      "grad_norm": 0.30978211760520935,
+      "learning_rate": 6.3829787234042555e-06,
+      "loss": 9.125,
+      "step": 3
+    },
+    {
+      "epoch": 0.0025673940949935813,
+      "grad_norm": 0.3134618401527405,
+      "learning_rate": 8.510638297872341e-06,
+      "loss": 9.125,
+      "step": 4
+    },
+    {
+      "epoch": 0.003209242618741977,
+      "grad_norm": 0.3147919178009033,
+      "learning_rate": 1.0638297872340426e-05,
+      "loss": 9.125,
+      "step": 5
+    },
+    {
+      "epoch": 0.0038510911424903724,
+      "grad_norm": 0.3050297796726227,
+      "learning_rate": 1.2765957446808511e-05,
+      "loss": 9.125,
+      "step": 6
+    },
+    {
+      "epoch": 0.004492939666238768,
+      "grad_norm": 0.29189687967300415,
+      "learning_rate": 1.4893617021276596e-05,
+      "loss": 9.125,
+      "step": 7
+    },
+    {
+      "epoch": 0.005134788189987163,
+      "grad_norm": 0.2791145145893097,
+      "learning_rate": 1.7021276595744682e-05,
+      "loss": 9.0,
+      "step": 8
+    },
+    {
+      "epoch": 0.005776636713735558,
+      "grad_norm": 0.24908146262168884,
+      "learning_rate": 1.9148936170212766e-05,
+      "loss": 8.875,
+      "step": 9
+    },
+    {
+      "epoch": 0.006418485237483954,
+      "grad_norm": 0.2112804502248764,
+      "learning_rate": 2.1276595744680852e-05,
+      "loss": 8.8125,
+      "step": 10
+    },
+    {
+      "epoch": 0.007060333761232349,
+      "grad_norm": 0.15402039885520935,
+      "learning_rate": 2.340425531914894e-05,
+      "loss": 8.625,
+      "step": 11
+    },
+    {
+      "epoch": 0.007702182284980745,
+      "grad_norm": 0.10005302727222443,
+      "learning_rate": 2.5531914893617022e-05,
+      "loss": 8.5,
+      "step": 12
+    },
+    {
+      "epoch": 0.00834403080872914,
+      "grad_norm": 0.061644140630960464,
+      "learning_rate": 2.765957446808511e-05,
+      "loss": 8.4375,
+      "step": 13
+    },
+    {
+      "epoch": 0.008985879332477536,
+      "grad_norm": 0.04229239746928215,
+      "learning_rate": 2.9787234042553192e-05,
+      "loss": 8.375,
+      "step": 14
+    },
+    {
+      "epoch": 0.009627727856225931,
+      "grad_norm": 0.04680441692471504,
+      "learning_rate": 3.191489361702128e-05,
+      "loss": 8.375,
+      "step": 15
+    },
+    {
+      "epoch": 0.010269576379974325,
+      "grad_norm": 0.0876237079501152,
+      "learning_rate": 3.4042553191489365e-05,
+      "loss": 8.3125,
+      "step": 16
+    },
+    {
+      "epoch": 0.010911424903722721,
+      "grad_norm": 0.12335914373397827,
+      "learning_rate": 3.617021276595745e-05,
+      "loss": 8.375,
+      "step": 17
+    },
+    {
+      "epoch": 0.011553273427471117,
+      "grad_norm": 0.0966993197798729,
+      "learning_rate": 3.829787234042553e-05,
+      "loss": 8.3125,
+      "step": 18
+    },
+    {
+      "epoch": 0.012195121951219513,
+      "grad_norm": 0.047657210379838943,
+      "learning_rate": 4.0425531914893614e-05,
+      "loss": 8.25,
+      "step": 19
+    },
+    {
+      "epoch": 0.012836970474967908,
+      "grad_norm": 0.024272233247756958,
+      "learning_rate": 4.2553191489361704e-05,
+      "loss": 8.25,
+      "step": 20
+    },
+    {
+      "epoch": 0.013478818998716302,
+      "grad_norm": 0.02427881769835949,
+      "learning_rate": 4.468085106382979e-05,
+      "loss": 8.25,
+      "step": 21
+    },
+    {
+      "epoch": 0.014120667522464698,
+      "grad_norm": 0.026496585458517075,
+      "learning_rate": 4.680851063829788e-05,
+      "loss": 8.25,
+      "step": 22
+    },
+    {
+      "epoch": 0.014762516046213094,
+      "grad_norm": 0.03327532112598419,
+      "learning_rate": 4.893617021276596e-05,
+      "loss": 8.25,
+      "step": 23
+    },
+    {
+      "epoch": 0.01540436456996149,
+      "grad_norm": 0.03587386757135391,
+      "learning_rate": 5.1063829787234044e-05,
+      "loss": 8.25,
+      "step": 24
+    },
+    {
+      "epoch": 0.016046213093709884,
+      "grad_norm": 0.03461160138249397,
+      "learning_rate": 5.319148936170213e-05,
+      "loss": 8.25,
+      "step": 25
+    },
+    {
+      "epoch": 0.01668806161745828,
+      "grad_norm": 0.029515139758586884,
+      "learning_rate": 5.531914893617022e-05,
+      "loss": 8.25,
+      "step": 26
+    },
+    {
+      "epoch": 0.017329910141206675,
+      "grad_norm": 0.027110258117318153,
+      "learning_rate": 5.744680851063831e-05,
+      "loss": 8.25,
+      "step": 27
+    },
+    {
+      "epoch": 0.01797175866495507,
+      "grad_norm": 0.02379121631383896,
+      "learning_rate": 5.9574468085106384e-05,
+      "loss": 8.1875,
+      "step": 28
+    },
+    {
+      "epoch": 0.018613607188703467,
+      "grad_norm": 0.02544436603784561,
+      "learning_rate": 6.170212765957447e-05,
+      "loss": 8.1875,
+      "step": 29
+    },
+    {
+      "epoch": 0.019255455712451863,
+      "grad_norm": 0.030174219980835915,
+      "learning_rate": 6.382978723404256e-05,
+      "loss": 8.1875,
+      "step": 30
+    },
+    {
+      "epoch": 0.01989730423620026,
+      "grad_norm": 0.03604472056031227,
+      "learning_rate": 6.595744680851063e-05,
+      "loss": 8.1875,
+      "step": 31
+    },
+    {
+      "epoch": 0.02053915275994865,
+      "grad_norm": 0.037900377064943314,
+      "learning_rate": 6.808510638297873e-05,
+      "loss": 8.125,
+      "step": 32
+    },
+    {
+      "epoch": 0.021181001283697046,
+      "grad_norm": 0.03888288140296936,
+      "learning_rate": 7.021276595744681e-05,
+      "loss": 8.125,
+      "step": 33
+    },
+    {
+      "epoch": 0.021822849807445442,
+      "grad_norm": 0.04872816056013107,
+      "learning_rate": 7.23404255319149e-05,
+      "loss": 8.0625,
+      "step": 34
+    },
+    {
+      "epoch": 0.022464698331193838,
+      "grad_norm": 0.07964374125003815,
+      "learning_rate": 7.446808510638298e-05,
+      "loss": 7.9375,
+      "step": 35
+    },
+    {
+      "epoch": 0.023106546854942234,
+      "grad_norm": 0.08704540133476257,
+      "learning_rate": 7.659574468085106e-05,
+      "loss": 7.875,
+      "step": 36
+    },
+    {
+      "epoch": 0.02374839537869063,
+      "grad_norm": 0.14245420694351196,
+      "learning_rate": 7.872340425531916e-05,
+      "loss": 7.625,
+      "step": 37
+    },
+    {
+      "epoch": 0.024390243902439025,
+      "grad_norm": 0.20091165602207184,
+      "learning_rate": 8.085106382978723e-05,
+      "loss": 7.25,
+      "step": 38
+    },
+    {
+      "epoch": 0.02503209242618742,
+      "grad_norm": 0.3081730902194977,
+      "learning_rate": 8.297872340425533e-05,
+      "loss": 6.4688,
+      "step": 39
+    },
+    {
+      "epoch": 0.025673940949935817,
+      "grad_norm": 0.3360954523086548,
+      "learning_rate": 8.510638297872341e-05,
+      "loss": 5.25,
+      "step": 40
+    },
+    {
+      "epoch": 0.02631578947368421,
+      "grad_norm": 0.28741300106048584,
+      "learning_rate": 8.723404255319149e-05,
+      "loss": 4.0625,
+      "step": 41
+    },
+    {
+      "epoch": 0.026957637997432605,
+      "grad_norm": 0.2542773187160492,
+      "learning_rate": 8.936170212765958e-05,
+      "loss": 3.1094,
+      "step": 42
+    },
+    {
+      "epoch": 0.027599486521181,
+      "grad_norm": 0.16312512755393982,
+      "learning_rate": 9.148936170212766e-05,
+      "loss": 2.5,
+      "step": 43
+    },
+    {
+      "epoch": 0.028241335044929396,
+      "grad_norm": 0.10257931798696518,
+      "learning_rate": 9.361702127659576e-05,
+      "loss": 2.1562,
+      "step": 44
+    },
+    {
+      "epoch": 0.028883183568677792,
+      "grad_norm": 0.07631522417068481,
+      "learning_rate": 9.574468085106384e-05,
+      "loss": 2.0,
+      "step": 45
+    },
+    {
+      "epoch": 0.029525032092426188,
+      "grad_norm": 0.06334506720304489,
+      "learning_rate": 9.787234042553192e-05,
+      "loss": 1.8281,
+      "step": 46
+    },
+    {
+      "epoch": 0.030166880616174584,
+      "grad_norm": 0.05079680308699608,
+      "learning_rate": 0.0001,
+      "loss": 1.7031,
+      "step": 47
+    },
+    {
+      "epoch": 0.03080872913992298,
+      "grad_norm": 0.050113581120967865,
+      "learning_rate": 9.999989192862701e-05,
+      "loss": 1.6719,
+      "step": 48
+    },
+    {
+      "epoch": 0.031450577663671375,
+      "grad_norm": 0.041200995445251465,
+      "learning_rate": 9.999956771497528e-05,
+      "loss": 1.6094,
+      "step": 49
+    },
+    {
+      "epoch": 0.03209242618741977,
+      "grad_norm": 0.03537944331765175,
+      "learning_rate": 9.999902736044627e-05,
+      "loss": 1.5312,
+      "step": 50
+    },
+    {
+      "epoch": 0.03273427471116817,
+      "grad_norm": 0.03154836595058441,
+      "learning_rate": 9.999827086737589e-05,
+      "loss": 1.5,
+      "step": 51
+    },
+    {
+      "epoch": 0.03337612323491656,
+      "grad_norm": 0.027803990989923477,
+      "learning_rate": 9.999729823903436e-05,
+      "loss": 1.4766,
+      "step": 52
+    },
+    {
+      "epoch": 0.03401797175866496,
+      "grad_norm": 0.023708201944828033,
+      "learning_rate": 9.99961094796262e-05,
+      "loss": 1.375,
+      "step": 53
+    },
+    {
+      "epoch": 0.03465982028241335,
+      "grad_norm": 0.021473145112395287,
+      "learning_rate": 9.999470459429021e-05,
+      "loss": 1.3281,
+      "step": 54
+    },
+    {
+      "epoch": 0.03530166880616174,
+      "grad_norm": 0.019184721633791924,
+      "learning_rate": 9.999308358909955e-05,
+      "loss": 1.2891,
+      "step": 55
+    },
+    {
+      "epoch": 0.03594351732991014,
+      "grad_norm": 0.01847364939749241,
+      "learning_rate": 9.999124647106159e-05,
+      "loss": 1.2656,
+      "step": 56
+    },
+    {
+      "epoch": 0.036585365853658534,
+      "grad_norm": 0.016471078619360924,
+      "learning_rate": 9.99891932481179e-05,
+      "loss": 1.2344,
+      "step": 57
+    },
+    {
+      "epoch": 0.037227214377406934,
+      "grad_norm": 0.016381070017814636,
+      "learning_rate": 9.99869239291443e-05,
+      "loss": 1.2812,
+      "step": 58
+    },
+    {
+      "epoch": 0.037869062901155326,
+      "grad_norm": 0.01620948687195778,
+      "learning_rate": 9.998443852395067e-05,
+      "loss": 1.2188,
+      "step": 59
+    },
+    {
+      "epoch": 0.038510911424903725,
+      "grad_norm": 0.01424532663077116,
+      "learning_rate": 9.998173704328112e-05,
+      "loss": 1.1797,
+      "step": 60
+    },
+    {
+      "epoch": 0.03915275994865212,
+      "grad_norm": 0.014445069245994091,
+      "learning_rate": 9.997881949881371e-05,
+      "loss": 1.1953,
+      "step": 61
+    },
+    {
+      "epoch": 0.03979460847240052,
+      "grad_norm": 0.013568165712058544,
+      "learning_rate": 9.99756859031606e-05,
+      "loss": 1.1562,
+      "step": 62
+    },
+    {
+      "epoch": 0.04043645699614891,
+      "grad_norm": 0.01342830155044794,
+      "learning_rate": 9.997233626986781e-05,
+      "loss": 1.1719,
+      "step": 63
+    },
+    {
+      "epoch": 0.0410783055198973,
+      "grad_norm": 0.013186159543693066,
+      "learning_rate": 9.99687706134154e-05,
+      "loss": 1.0938,
+      "step": 64
+    },
+    {
+      "epoch": 0.0417201540436457,
+      "grad_norm": 0.013615522533655167,
+      "learning_rate": 9.996498894921713e-05,
+      "loss": 1.1172,
+      "step": 65
+    },
+    {
+      "epoch": 0.04236200256739409,
+      "grad_norm": 0.011763419024646282,
+      "learning_rate": 9.996099129362059e-05,
+      "loss": 1.0781,
+      "step": 66
+    },
+    {
+      "epoch": 0.04300385109114249,
+      "grad_norm": 0.009965788573026657,
+      "learning_rate": 9.995677766390707e-05,
+      "loss": 1.0938,
+      "step": 67
+    },
+    {
+      "epoch": 0.043645699614890884,
+      "grad_norm": 0.011237729340791702,
+      "learning_rate": 9.99523480782915e-05,
+      "loss": 1.0938,
+      "step": 68
+    },
+    {
+      "epoch": 0.044287548138639284,
+      "grad_norm": 0.010530981235206127,
+      "learning_rate": 9.994770255592233e-05,
+      "loss": 1.0938,
+      "step": 69
+    },
+    {
+      "epoch": 0.044929396662387676,
+      "grad_norm": 0.010274014435708523,
+      "learning_rate": 9.994284111688145e-05,
+      "loss": 1.0234,
+      "step": 70
+    },
+    {
+      "epoch": 0.045571245186136075,
+      "grad_norm": 0.009948200546205044,
+      "learning_rate": 9.993776378218418e-05,
+      "loss": 1.0469,
+      "step": 71
+    },
+    {
+      "epoch": 0.04621309370988447,
+      "grad_norm": 0.010145426727831364,
+      "learning_rate": 9.99324705737791e-05,
+      "loss": 1.0625,
+      "step": 72
+    },
+    {
+      "epoch": 0.04685494223363286,
+      "grad_norm": 0.01015029288828373,
+      "learning_rate": 9.992696151454799e-05,
+      "loss": 1.0703,
+      "step": 73
+    },
+    {
+      "epoch": 0.04749679075738126,
+      "grad_norm": 0.011197876185178757,
+      "learning_rate": 9.992123662830568e-05,
+      "loss": 1.0469,
+      "step": 74
+    },
+    {
+      "epoch": 0.04813863928112965,
+      "grad_norm": 0.010476046241819859,
+      "learning_rate": 9.991529593980006e-05,
+      "loss": 1.0234,
+      "step": 75
+    },
+    {
+      "epoch": 0.04878048780487805,
+      "grad_norm": 0.009835699573159218,
+      "learning_rate": 9.990913947471184e-05,
+      "loss": 0.9766,
+      "step": 76
+    },
+    {
+      "epoch": 0.04942233632862644,
+      "grad_norm": 0.009792305529117584,
+      "learning_rate": 9.990276725965455e-05,
+      "loss": 0.9805,
+      "step": 77
+    },
+    {
+      "epoch": 0.05006418485237484,
+      "grad_norm": 0.00878097303211689,
+      "learning_rate": 9.989617932217432e-05,
+      "loss": 0.918,
+      "step": 78
+    },
+    {
+      "epoch": 0.050706033376123234,
+      "grad_norm": 0.009141056798398495,
+      "learning_rate": 9.988937569074988e-05,
+      "loss": 0.9297,
+      "step": 79
+    },
+    {
+      "epoch": 0.051347881899871634,
+      "grad_norm": 0.008934931829571724,
+      "learning_rate": 9.98823563947923e-05,
+      "loss": 0.9727,
+      "step": 80
+    },
+    {
+      "epoch": 0.051989730423620026,
+      "grad_norm": 0.008960246108472347,
+      "learning_rate": 9.987512146464504e-05,
+      "loss": 0.9922,
+      "step": 81
+    },
+    {
+      "epoch": 0.05263157894736842,
+      "grad_norm": 0.00925202202051878,
+      "learning_rate": 9.986767093158359e-05,
+      "loss": 0.9453,
+      "step": 82
+    },
+    {
+      "epoch": 0.05327342747111682,
+      "grad_norm": 0.008465278893709183,
+      "learning_rate": 9.986000482781557e-05,
+      "loss": 0.9766,
+      "step": 83
+    },
+    {
+      "epoch": 0.05391527599486521,
+      "grad_norm": 0.008937939070165157,
+      "learning_rate": 9.985212318648039e-05,
+      "loss": 0.8906,
+      "step": 84
+    },
+    {
+      "epoch": 0.05455712451861361,
+      "grad_norm": 0.00869137141853571,
+      "learning_rate": 9.984402604164928e-05,
+      "loss": 0.8828,
+      "step": 85
+    },
+    {
+      "epoch": 0.055198973042362,
+      "grad_norm": 0.00921856239438057,
+      "learning_rate": 9.983571342832501e-05,
+      "loss": 0.875,
+      "step": 86
+    },
+    {
+      "epoch": 0.0558408215661104,
+      "grad_norm": 0.008353888988494873,
+      "learning_rate": 9.982718538244181e-05,
+      "loss": 0.9219,
+      "step": 87
+    },
+    {
+      "epoch": 0.05648267008985879,
+      "grad_norm": 0.008289925754070282,
+      "learning_rate": 9.981844194086517e-05,
+      "loss": 0.875,
+      "step": 88
+    },
+    {
+      "epoch": 0.05712451861360719,
+      "grad_norm": 0.009131287224590778,
+      "learning_rate": 9.980948314139173e-05,
+      "loss": 0.9375,
+      "step": 89
+    },
+    {
+      "epoch": 0.057766367137355584,
+      "grad_norm": 0.008783861063420773,
+      "learning_rate": 9.980030902274907e-05,
+      "loss": 0.9141,
+      "step": 90
+    },
+    {
+      "epoch": 0.05840821566110398,
+      "grad_norm": 0.009424981661140919,
+      "learning_rate": 9.979091962459557e-05,
+      "loss": 0.8672,
+      "step": 91
+    },
+    {
+      "epoch": 0.059050064184852376,
+      "grad_norm": 0.008605397306382656,
+      "learning_rate": 9.978131498752026e-05,
+      "loss": 0.8672,
+      "step": 92
+    },
+    {
+      "epoch": 0.05969191270860077,
+      "grad_norm": 0.007921144366264343,
+      "learning_rate": 9.977149515304257e-05,
+      "loss": 0.9102,
+      "step": 93
+    },
+    {
+      "epoch": 0.06033376123234917,
+      "grad_norm": 0.008213751949369907,
+      "learning_rate": 9.976146016361223e-05,
+      "loss": 0.8555,
+      "step": 94
+    },
+    {
+      "epoch": 0.06097560975609756,
+      "grad_norm": 0.0076583558693528175,
+      "learning_rate": 9.975121006260905e-05,
+      "loss": 0.8359,
+      "step": 95
+    },
+    {
+      "epoch": 0.06161745827984596,
+      "grad_norm": 0.00856377650052309,
+      "learning_rate": 9.97407448943427e-05,
+      "loss": 0.8203,
+      "step": 96
+    },
+    {
+      "epoch": 0.06225930680359435,
+      "grad_norm": 0.008036739192903042,
+      "learning_rate": 9.973006470405264e-05,
+      "loss": 0.8516,
+      "step": 97
+    },
+    {
+      "epoch": 0.06290115532734275,
+      "grad_norm": 0.008434290066361427,
+      "learning_rate": 9.971916953790772e-05,
+      "loss": 0.8086,
+      "step": 98
+    },
+    {
+      "epoch": 0.06354300385109114,
+      "grad_norm": 0.008587395772337914,
+      "learning_rate": 9.970805944300621e-05,
+      "loss": 0.8203,
+      "step": 99
+    },
+    {
+      "epoch": 0.06418485237483953,
+      "grad_norm": 0.0088276332244277,
+      "learning_rate": 9.96967344673754e-05,
+      "loss": 0.8086,
+      "step": 100
+    },
+    {
+      "epoch": 0.06482670089858793,
+      "grad_norm": 0.009358148090541363,
+      "learning_rate": 9.968519465997155e-05,
+      "loss": 0.8047,
+      "step": 101
+    },
+    {
+      "epoch": 0.06546854942233633,
+      "grad_norm": 0.008407055400311947,
+      "learning_rate": 9.967344007067954e-05,
+      "loss": 0.7617,
+      "step": 102
+    },
+    {
+      "epoch": 0.06611039794608473,
+      "grad_norm": 0.00826259609311819,
+      "learning_rate": 9.96614707503128e-05,
+      "loss": 0.8359,
+      "step": 103
+    },
+    {
+      "epoch": 0.06675224646983312,
+      "grad_norm": 0.009418654255568981,
+      "learning_rate": 9.964928675061291e-05,
+      "loss": 0.7812,
+      "step": 104
+    },
+    {
+      "epoch": 0.06739409499358151,
+      "grad_norm": 0.012697753496468067,
+      "learning_rate": 9.963688812424958e-05,
+      "loss": 0.7969,
+      "step": 105
+    },
+    {
+      "epoch": 0.06803594351732992,
+      "grad_norm": 0.008924663066864014,
+      "learning_rate": 9.962427492482026e-05,
+      "loss": 0.8359,
+      "step": 106
+    },
+    {
+      "epoch": 0.06867779204107831,
+      "grad_norm": 0.009173823520541191,
+      "learning_rate": 9.961144720684997e-05,
+      "loss": 0.7656,
+      "step": 107
+    },
+    {
+      "epoch": 0.0693196405648267,
+      "grad_norm": 0.009439458139240742,
+      "learning_rate": 9.959840502579108e-05,
+      "loss": 0.7578,
+      "step": 108
+    },
+    {
+      "epoch": 0.0699614890885751,
+      "grad_norm": 0.01025363802909851,
+      "learning_rate": 9.958514843802305e-05,
+      "loss": 0.8203,
+      "step": 109
+    },
+    {
+      "epoch": 0.07060333761232349,
+      "grad_norm": 0.01139143854379654,
+      "learning_rate": 9.957167750085217e-05,
+      "loss": 0.7891,
+      "step": 110
+    },
+    {
+      "epoch": 0.07124518613607189,
+      "grad_norm": 0.012185357511043549,
+      "learning_rate": 9.955799227251137e-05,
+      "loss": 0.75,
+      "step": 111
+    },
+    {
+      "epoch": 0.07188703465982028,
+      "grad_norm": 0.011034708470106125,
+      "learning_rate": 9.954409281215989e-05,
+      "loss": 0.8047,
+      "step": 112
+    },
+    {
+      "epoch": 0.07252888318356868,
+      "grad_norm": 0.009615590795874596,
+      "learning_rate": 9.952997917988308e-05,
+      "loss": 0.75,
+      "step": 113
+    },
+    {
+      "epoch": 0.07317073170731707,
+      "grad_norm": 0.010590028017759323,
+      "learning_rate": 9.951565143669213e-05,
+      "loss": 0.7422,
+      "step": 114
+    },
+    {
+      "epoch": 0.07381258023106547,
+      "grad_norm": 0.010860998183488846,
+      "learning_rate": 9.950110964452382e-05,
+      "loss": 0.7188,
+      "step": 115
+    },
+    {
+      "epoch": 0.07445442875481387,
+      "grad_norm": 0.010860625654459,
+      "learning_rate": 9.948635386624016e-05,
+      "loss": 0.7305,
+      "step": 116
+    },
+    {
+      "epoch": 0.07509627727856226,
+      "grad_norm": 0.011553982272744179,
+      "learning_rate": 9.947138416562826e-05,
+      "loss": 0.7617,
+      "step": 117
+    },
+    {
+      "epoch": 0.07573812580231065,
+      "grad_norm": 0.012888479046523571,
+      "learning_rate": 9.945620060739999e-05,
+      "loss": 0.6953,
+      "step": 118
+    },
+    {
+      "epoch": 0.07637997432605904,
+      "grad_norm": 0.01304571982473135,
+      "learning_rate": 9.944080325719163e-05,
+      "loss": 0.7461,
+      "step": 119
+    },
+    {
+      "epoch": 0.07702182284980745,
+      "grad_norm": 0.01109766960144043,
+      "learning_rate": 9.94251921815637e-05,
+      "loss": 0.7344,
+      "step": 120
+    },
+    {
+      "epoch": 0.07766367137355584,
+      "grad_norm": 0.010833417065441608,
+      "learning_rate": 9.940936744800064e-05,
+      "loss": 0.7148,
+      "step": 121
+    },
+    {
+      "epoch": 0.07830551989730423,
+      "grad_norm": 0.012143843807280064,
+      "learning_rate": 9.939332912491043e-05,
+      "loss": 0.7578,
+      "step": 122
+    },
+    {
+      "epoch": 0.07894736842105263,
+      "grad_norm": 0.01161271333694458,
+      "learning_rate": 9.937707728162448e-05,
+      "loss": 0.7617,
+      "step": 123
+    },
+    {
+      "epoch": 0.07958921694480103,
+      "grad_norm": 0.011860181577503681,
+      "learning_rate": 9.93606119883971e-05,
+      "loss": 0.7422,
+      "step": 124
+    },
+    {
+      "epoch": 0.08023106546854943,
+      "grad_norm": 0.009615721181035042,
+      "learning_rate": 9.934393331640536e-05,
+      "loss": 0.668,
+      "step": 125
+    },
+    {
+      "epoch": 0.08087291399229782,
+      "grad_norm": 0.01393971312791109,
+      "learning_rate": 9.932704133774877e-05,
+      "loss": 0.6953,
+      "step": 126
+    },
+    {
+      "epoch": 0.08151476251604621,
+      "grad_norm": 0.012269717641174793,
+      "learning_rate": 9.93099361254489e-05,
+      "loss": 0.7266,
+      "step": 127
+    },
+    {
+      "epoch": 0.0821566110397946,
+      "grad_norm": 0.014152932912111282,
+      "learning_rate": 9.929261775344909e-05,
+      "loss": 0.6562,
+      "step": 128
+    },
+    {
+      "epoch": 0.08279845956354301,
+      "grad_norm": 0.011507104150950909,
+      "learning_rate": 9.927508629661413e-05,
+      "loss": 0.7266,
+      "step": 129
+    },
+    {
+      "epoch": 0.0834403080872914,
+      "grad_norm": 0.014678554609417915,
+      "learning_rate": 9.925734183073001e-05,
+      "loss": 0.7109,
+      "step": 130
+    },
+    {
+      "epoch": 0.0840821566110398,
+      "grad_norm": 0.01359044574201107,
+      "learning_rate": 9.923938443250345e-05,
+      "loss": 0.6875,
+      "step": 131
+    },
+    {
+      "epoch": 0.08472400513478819,
+      "grad_norm": 0.012736028991639614,
+      "learning_rate": 9.922121417956168e-05,
+      "loss": 0.6641,
+      "step": 132
+    },
+    {
+      "epoch": 0.08536585365853659,
+      "grad_norm": 0.01216636598110199,
+      "learning_rate": 9.920283115045206e-05,
+      "loss": 0.7031,
+      "step": 133
+    },
+    {
+      "epoch": 0.08600770218228498,
+      "grad_norm": 0.011741497553884983,
+      "learning_rate": 9.918423542464177e-05,
+      "loss": 0.6719,
+      "step": 134
+    },
+    {
+      "epoch": 0.08664955070603338,
+      "grad_norm": 0.010003621689975262,
+      "learning_rate": 9.916542708251745e-05,
+      "loss": 0.6719,
+      "step": 135
+    },
+    {
+      "epoch": 0.08729139922978177,
+      "grad_norm": 0.012072731740772724,
+      "learning_rate": 9.91464062053848e-05,
+      "loss": 0.7188,
+      "step": 136
+    },
+    {
+      "epoch": 0.08793324775353016,
+      "grad_norm": 0.010896646417677402,
+      "learning_rate": 9.912717287546835e-05,
+      "loss": 0.7031,
+      "step": 137
+    },
+    {
+      "epoch": 0.08857509627727857,
+      "grad_norm": 0.011950834654271603,
+      "learning_rate": 9.910772717591097e-05,
+      "loss": 0.6836,
+      "step": 138
+    },
+    {
+      "epoch": 0.08921694480102696,
+      "grad_norm": 0.01246932614594698,
+      "learning_rate": 9.90880691907736e-05,
+      "loss": 0.6797,
+      "step": 139
+    },
+    {
+      "epoch": 0.08985879332477535,
+      "grad_norm": 0.013044403865933418,
+      "learning_rate": 9.906819900503486e-05,
+      "loss": 0.7109,
+      "step": 140
+    },
+    {
+      "epoch": 0.09050064184852374,
+      "grad_norm": 0.012423249892890453,
+      "learning_rate": 9.904811670459067e-05,
+      "loss": 0.6562,
+      "step": 141
+    },
+    {
+      "epoch": 0.09114249037227215,
+      "grad_norm": 0.010445546358823776,
+      "learning_rate": 9.902782237625393e-05,
+      "loss": 0.6211,
+      "step": 142
+    },
+    {
+      "epoch": 0.09178433889602054,
+      "grad_norm": 0.012769351713359356,
+      "learning_rate": 9.900731610775405e-05,
+      "loss": 0.6602,
+      "step": 143
+    },
+    {
+      "epoch": 0.09242618741976893,
+      "grad_norm": 0.01104013156145811,
+      "learning_rate": 9.898659798773667e-05,
+      "loss": 0.6875,
+      "step": 144
+    },
+    {
+      "epoch": 0.09306803594351733,
+      "grad_norm": 0.009402669034898281,
+      "learning_rate": 9.89656681057632e-05,
+      "loss": 0.668,
+      "step": 145
+    },
+    {
+      "epoch": 0.09370988446726572,
+      "grad_norm": 0.008899720385670662,
+      "learning_rate": 9.894452655231051e-05,
+      "loss": 0.6406,
+      "step": 146
+    },
+    {
+      "epoch": 0.09435173299101413,
+      "grad_norm": 0.009158738888800144,
+      "learning_rate": 9.892317341877045e-05,
+      "loss": 0.6484,
+      "step": 147
+    },
+    {
+      "epoch": 0.09499358151476252,
+      "grad_norm": 0.010559624060988426,
+      "learning_rate": 9.890160879744951e-05,
+      "loss": 0.6484,
+      "step": 148
+    },
+    {
+      "epoch": 0.09563543003851091,
+      "grad_norm": 0.010568162426352501,
+      "learning_rate": 9.887983278156842e-05,
+      "loss": 0.6406,
+      "step": 149
+    },
+    {
+      "epoch": 0.0962772785622593,
+      "grad_norm": 0.00897563062608242,
+      "learning_rate": 9.885784546526177e-05,
+      "loss": 0.6328,
+      "step": 150
+    },
+    {
+      "epoch": 0.09691912708600771,
+      "grad_norm": 0.009579532779753208,
+      "learning_rate": 9.88356469435775e-05,
+      "loss": 0.6367,
+      "step": 151
+    },
+    {
+      "epoch": 0.0975609756097561,
+      "grad_norm": 0.010554454289376736,
+      "learning_rate": 9.881323731247663e-05,
+      "loss": 0.6289,
+      "step": 152
+    },
+    {
+      "epoch": 0.0982028241335045,
+      "grad_norm": 0.009940484538674355,
+      "learning_rate": 9.879061666883272e-05,
+      "loss": 0.6016,
+      "step": 153
+    },
+    {
+      "epoch": 0.09884467265725289,
+      "grad_norm": 0.010707387700676918,
+      "learning_rate": 9.876778511043153e-05,
+      "loss": 0.6641,
+      "step": 154
+    },
+    {
+      "epoch": 0.09948652118100128,
+      "grad_norm": 0.009775303304195404,
+      "learning_rate": 9.874474273597059e-05,
+      "loss": 0.6406,
+      "step": 155
+    },
+    {
+      "epoch": 0.10012836970474968,
+      "grad_norm": 0.011378953233361244,
+      "learning_rate": 9.872148964505872e-05,
+      "loss": 0.6719,
+      "step": 156
+    },
+    {
+      "epoch": 0.10077021822849808,
+      "grad_norm": 0.009989460930228233,
+      "learning_rate": 9.869802593821568e-05,
+      "loss": 0.6641,
+      "step": 157
+    },
+    {
+      "epoch": 0.10141206675224647,
+      "grad_norm": 0.011383592151105404,
+      "learning_rate": 9.867435171687168e-05,
+      "loss": 0.625,
+      "step": 158
+    },
+    {
+      "epoch": 0.10205391527599486,
+      "grad_norm": 0.011871813796460629,
+      "learning_rate": 9.865046708336691e-05,
+      "loss": 0.6602,
+      "step": 159
+    },
+    {
+      "epoch": 0.10269576379974327,
+      "grad_norm": 0.009828433394432068,
+      "learning_rate": 9.862637214095121e-05,
+      "loss": 0.625,
+      "step": 160
+    },
+    {
+      "epoch": 0.10333761232349166,
+      "grad_norm": 0.010281259194016457,
+      "learning_rate": 9.860206699378349e-05,
+      "loss": 0.6328,
+      "step": 161
+    },
+    {
+      "epoch": 0.10397946084724005,
+      "grad_norm": 0.013299995101988316,
+      "learning_rate": 9.85775517469314e-05,
+      "loss": 0.625,
+      "step": 162
+    },
+    {
+      "epoch": 0.10462130937098844,
+      "grad_norm": 0.009967153891921043,
+      "learning_rate": 9.855282650637079e-05,
+      "loss": 0.5977,
+      "step": 163
+    },
+    {
+      "epoch": 0.10526315789473684,
+      "grad_norm": 0.011748317629098892,
+      "learning_rate": 9.852789137898526e-05,
+      "loss": 0.6133,
+      "step": 164
+    },
+    {
+      "epoch": 0.10590500641848524,
+      "grad_norm": 0.00852824654430151,
+      "learning_rate": 9.85027464725658e-05,
+      "loss": 0.6094,
+      "step": 165
+    },
+    {
+      "epoch": 0.10654685494223363,
+      "grad_norm": 0.011019607074558735,
+      "learning_rate": 9.847739189581013e-05,
+      "loss": 0.6445,
+      "step": 166
+    },
+    {
+      "epoch": 0.10718870346598203,
+      "grad_norm": 0.011637663468718529,
+      "learning_rate": 9.845182775832244e-05,
+      "loss": 0.625,
+      "step": 167
+    },
+    {
+      "epoch": 0.10783055198973042,
+      "grad_norm": 0.011865214444696903,
+      "learning_rate": 9.842605417061282e-05,
+      "loss": 0.6484,
+      "step": 168
+    },
+    {
+      "epoch": 0.10847240051347883,
+      "grad_norm": 0.011490019969642162,
+      "learning_rate": 9.840007124409669e-05,
+      "loss": 0.625,
+      "step": 169
+    },
+    {
+      "epoch": 0.10911424903722722,
+      "grad_norm": 0.010043886490166187,
+      "learning_rate": 9.837387909109452e-05,
+      "loss": 0.6094,
+      "step": 170
+    },
+    {
+      "epoch": 0.10975609756097561,
+      "grad_norm": 0.01057278923690319,
+      "learning_rate": 9.834747782483113e-05,
+      "loss": 0.6328,
+      "step": 171
+    },
+    {
+      "epoch": 0.110397946084724,
+      "grad_norm": 0.013821721076965332,
+      "learning_rate": 9.832086755943543e-05,
+      "loss": 0.625,
+      "step": 172
+    },
+    {
+      "epoch": 0.1110397946084724,
+      "grad_norm": 0.011414537206292152,
+      "learning_rate": 9.82940484099397e-05,
+      "loss": 0.5938,
+      "step": 173
+    },
+    {
+      "epoch": 0.1116816431322208,
+      "grad_norm": 0.011716284789144993,
+      "learning_rate": 9.826702049227925e-05,
+      "loss": 0.6484,
+      "step": 174
+    },
+    {
+      "epoch": 0.1123234916559692,
+      "grad_norm": 0.009901394136250019,
+      "learning_rate": 9.823978392329183e-05,
+      "loss": 0.6016,
+      "step": 175
+    },
+    {
+      "epoch": 0.11296534017971759,
+      "grad_norm": 0.011477384716272354,
+      "learning_rate": 9.821233882071717e-05,
+      "loss": 0.625,
+      "step": 176
+    },
+    {
+      "epoch": 0.11360718870346598,
+      "grad_norm": 0.011600608006119728,
+      "learning_rate": 9.81846853031965e-05,
+      "loss": 0.6094,
+      "step": 177
+    },
+    {
+      "epoch": 0.11424903722721438,
+      "grad_norm": 0.011693180538713932,
+      "learning_rate": 9.815682349027193e-05,
+      "loss": 0.6328,
+      "step": 178
+    },
+    {
+      "epoch": 0.11489088575096278,
+      "grad_norm": 0.013446972705423832,
+      "learning_rate": 9.812875350238604e-05,
+      "loss": 0.582,
+      "step": 179
+    },
+    {
+      "epoch": 0.11553273427471117,
+      "grad_norm": 0.012557251378893852,
+      "learning_rate": 9.810047546088133e-05,
+      "loss": 0.6016,
+      "step": 180
+    },
+    {
+      "epoch": 0.11617458279845956,
+      "grad_norm": 0.010684821754693985,
+      "learning_rate": 9.807198948799968e-05,
+      "loss": 0.6133,
+      "step": 181
+    },
+    {
+      "epoch": 0.11681643132220795,
+      "grad_norm": 0.011835362762212753,
+      "learning_rate": 9.804329570688177e-05,
+      "loss": 0.6133,
+      "step": 182
+    },
+    {
+      "epoch": 0.11745827984595636,
+      "grad_norm": 0.016150448471307755,
+      "learning_rate": 9.801439424156672e-05,
+      "loss": 0.6602,
+      "step": 183
+    },
+    {
+      "epoch": 0.11810012836970475,
+      "grad_norm": 0.00978788174688816,
+      "learning_rate": 9.798528521699132e-05,
+      "loss": 0.582,
+      "step": 184
+    },
+    {
+      "epoch": 0.11874197689345314,
+      "grad_norm": 0.012773280963301659,
+      "learning_rate": 9.795596875898967e-05,
+      "loss": 0.6094,
+      "step": 185
+    },
+    {
+      "epoch": 0.11938382541720154,
+      "grad_norm": 0.013691268861293793,
+      "learning_rate": 9.792644499429259e-05,
+      "loss": 0.6016,
+      "step": 186
+    },
+    {
+      "epoch": 0.12002567394094994,
+      "grad_norm": 0.012030140496790409,
+      "learning_rate": 9.789671405052701e-05,
+      "loss": 0.5703,
+      "step": 187
+    },
+    {
+      "epoch": 0.12066752246469833,
+      "grad_norm": 0.01144426316022873,
+      "learning_rate": 9.786677605621547e-05,
+      "loss": 0.5469,
+      "step": 188
+    },
+    {
+      "epoch": 0.12130937098844673,
+      "grad_norm": 0.010707143694162369,
+      "learning_rate": 9.783663114077562e-05,
+      "loss": 0.6172,
+      "step": 189
+    },
+    {
+      "epoch": 0.12195121951219512,
+      "grad_norm": 0.011395948939025402,
+      "learning_rate": 9.78062794345195e-05,
+      "loss": 0.543,
+      "step": 190
+    },
+    {
+      "epoch": 0.12259306803594351,
+      "grad_norm": 0.010497979819774628,
+      "learning_rate": 9.777572106865318e-05,
+      "loss": 0.5859,
+      "step": 191
+    },
+    {
+      "epoch": 0.12323491655969192,
+      "grad_norm": 0.011176045052707195,
+      "learning_rate": 9.774495617527603e-05,
+      "loss": 0.5781,
+      "step": 192
+    },
+    {
+      "epoch": 0.12387676508344031,
+      "grad_norm": 0.009919974952936172,
+      "learning_rate": 9.771398488738022e-05,
+      "loss": 0.5938,
+      "step": 193
+    },
+    {
+      "epoch": 0.1245186136071887,
+      "grad_norm": 0.009419148787856102,
+      "learning_rate": 9.768280733885014e-05,
+      "loss": 0.5508,
+      "step": 194
+    },
+    {
+      "epoch": 0.1251604621309371,
+      "grad_norm": 0.011592664755880833,
+      "learning_rate": 9.765142366446178e-05,
+      "loss": 0.5664,
+      "step": 195
+    },
+    {
+      "epoch": 0.1258023106546855,
+      "grad_norm": 0.01011605840176344,
+      "learning_rate": 9.761983399988223e-05,
+      "loss": 0.582,
+      "step": 196
+    },
+    {
+      "epoch": 0.12644415917843388,
+      "grad_norm": 0.011448321864008904,
+      "learning_rate": 9.758803848166904e-05,
+      "loss": 0.5625,
+      "step": 197
+    },
+    {
+      "epoch": 0.12708600770218229,
+      "grad_norm": 0.011436822824180126,
+      "learning_rate": 9.75560372472696e-05,
+      "loss": 0.5781,
+      "step": 198
+    },
+    {
+      "epoch": 0.1277278562259307,
+      "grad_norm": 0.01132872048765421,
+      "learning_rate": 9.752383043502063e-05,
+      "loss": 0.5938,
+      "step": 199
+    },
+    {
+      "epoch": 0.12836970474967907,
+      "grad_norm": 0.010940883308649063,
+      "learning_rate": 9.749141818414749e-05,
+      "loss": 0.543,
+      "step": 200
+    },
+    {
+      "epoch": 0.12901155327342748,
+      "grad_norm": 0.011073155328631401,
+      "learning_rate": 9.745880063476362e-05,
+      "loss": 0.5469,
+      "step": 201
+    },
+    {
+      "epoch": 0.12965340179717585,
+      "grad_norm": 0.01133064553141594,
+      "learning_rate": 9.742597792786999e-05,
+      "loss": 0.5859,
+      "step": 202
+    },
+    {
+      "epoch": 0.13029525032092426,
+      "grad_norm": 0.011374837718904018,
+      "learning_rate": 9.739295020535437e-05,
+      "loss": 0.5508,
+      "step": 203
+    },
+    {
+      "epoch": 0.13093709884467267,
+      "grad_norm": 0.013687084428966045,
+      "learning_rate": 9.735971760999083e-05,
+      "loss": 0.5859,
+      "step": 204
+    },
+    {
+      "epoch": 0.13157894736842105,
+      "grad_norm": 0.010709963738918304,
+      "learning_rate": 9.732628028543906e-05,
+      "loss": 0.5312,
+      "step": 205
+    },
+    {
+      "epoch": 0.13222079589216945,
+      "grad_norm": 0.012992851436138153,
+      "learning_rate": 9.729263837624376e-05,
+      "loss": 0.5938,
+      "step": 206
+    },
+    {
+      "epoch": 0.13286264441591783,
+      "grad_norm": 0.011751007288694382,
+      "learning_rate": 9.725879202783401e-05,
+      "loss": 0.5859,
+      "step": 207
+    },
+    {
+      "epoch": 0.13350449293966624,
+      "grad_norm": 0.01132382545620203,
+      "learning_rate": 9.722474138652267e-05,
+      "loss": 0.582,
+      "step": 208
+    },
+    {
+      "epoch": 0.13414634146341464,
+      "grad_norm": 0.011845381930470467,
+      "learning_rate": 9.719048659950573e-05,
+      "loss": 0.5625,
+      "step": 209
+    },
+    {
+      "epoch": 0.13478818998716302,
+      "grad_norm": 0.011076336726546288,
+      "learning_rate": 9.715602781486166e-05,
+      "loss": 0.5703,
+      "step": 210
+    },
+    {
+      "epoch": 0.13543003851091143,
+      "grad_norm": 0.01224234327673912,
+      "learning_rate": 9.712136518155079e-05,
+      "loss": 0.5625,
+      "step": 211
+    },
+    {
+      "epoch": 0.13607188703465983,
+      "grad_norm": 0.012131668627262115,
+      "learning_rate": 9.708649884941466e-05,
+      "loss": 0.5547,
+      "step": 212
+    },
+    {
+      "epoch": 0.1367137355584082,
+      "grad_norm": 0.014434052631258965,
+      "learning_rate": 9.705142896917534e-05,
+      "loss": 0.5625,
+      "step": 213
+    },
+    {
+      "epoch": 0.13735558408215662,
+      "grad_norm": 0.013702210038900375,
+      "learning_rate": 9.701615569243485e-05,
+      "loss": 0.5781,
+      "step": 214
+    },
+    {
+      "epoch": 0.137997432605905,
+      "grad_norm": 0.01172378659248352,
+      "learning_rate": 9.698067917167446e-05,
+      "loss": 0.5703,
+      "step": 215
+    },
+    {
+      "epoch": 0.1386392811296534,
+      "grad_norm": 0.01423399057239294,
+      "learning_rate": 9.6944999560254e-05,
+      "loss": 0.5352,
+      "step": 216
+    },
+    {
+      "epoch": 0.1392811296534018,
+      "grad_norm": 0.013110863976180553,
+      "learning_rate": 9.690911701241128e-05,
+      "loss": 0.5625,
+      "step": 217
+    },
+    {
+      "epoch": 0.1399229781771502,
+      "grad_norm": 0.013581229373812675,
+      "learning_rate": 9.687303168326133e-05,
+      "loss": 0.5312,
+      "step": 218
+    },
+    {
+      "epoch": 0.1405648267008986,
+      "grad_norm": 0.010977387428283691,
+      "learning_rate": 9.683674372879579e-05,
+      "loss": 0.5391,
+      "step": 219
+    },
+    {
+      "epoch": 0.14120667522464697,
+      "grad_norm": 0.010781140998005867,
+      "learning_rate": 9.680025330588223e-05,
+      "loss": 0.5664,
+      "step": 220
+    },
+    {
+      "epoch": 0.14184852374839538,
+      "grad_norm": 0.010498926043510437,
+      "learning_rate": 9.676356057226346e-05,
+      "loss": 0.5234,
+      "step": 221
+    },
+    {
+      "epoch": 0.14249037227214378,
+      "grad_norm": 0.017316415905952454,
+      "learning_rate": 9.672666568655684e-05,
+      "loss": 0.5859,
+      "step": 222
+    },
+    {
+      "epoch": 0.14313222079589216,
+      "grad_norm": 0.010606053285300732,
+      "learning_rate": 9.66895688082536e-05,
+      "loss": 0.5859,
+      "step": 223
+    },
+    {
+      "epoch": 0.14377406931964057,
+      "grad_norm": 0.010790948756039143,
+      "learning_rate": 9.665227009771815e-05,
+      "loss": 0.5469,
+      "step": 224
+    },
+    {
+      "epoch": 0.14441591784338895,
+      "grad_norm": 0.016792552545666695,
+      "learning_rate": 9.661476971618744e-05,
+      "loss": 0.5625,
+      "step": 225
+    },
+    {
+      "epoch": 0.14505776636713735,
+      "grad_norm": 0.011027633212506771,
+      "learning_rate": 9.657706782577017e-05,
+      "loss": 0.5703,
+      "step": 226
+    },
+    {
+      "epoch": 0.14569961489088576,
+      "grad_norm": 0.012505598366260529,
+      "learning_rate": 9.653916458944612e-05,
+      "loss": 0.5664,
+      "step": 227
+    },
+    {
+      "epoch": 0.14634146341463414,
+      "grad_norm": 0.010994997806847095,
+      "learning_rate": 9.650106017106548e-05,
+      "loss": 0.5312,
+      "step": 228
+    },
+    {
+      "epoch": 0.14698331193838254,
+      "grad_norm": 0.019628558307886124,
+      "learning_rate": 9.646275473534817e-05,
+      "loss": 0.5312,
+      "step": 229
+    },
+    {
+      "epoch": 0.14762516046213095,
+      "grad_norm": 0.012861509807407856,
+      "learning_rate": 9.642424844788298e-05,
+      "loss": 0.5508,
+      "step": 230
+    },
+    {
+      "epoch": 0.14826700898587933,
+      "grad_norm": 0.014055141247808933,
+      "learning_rate": 9.638554147512702e-05,
+      "loss": 0.5625,
+      "step": 231
+    },
+    {
+      "epoch": 0.14890885750962773,
+      "grad_norm": 0.015493751503527164,
+      "learning_rate": 9.634663398440493e-05,
+      "loss": 0.543,
+      "step": 232
+    },
+    {
+      "epoch": 0.1495507060333761,
+      "grad_norm": 0.012860543094575405,
+      "learning_rate": 9.630752614390813e-05,
+      "loss": 0.4922,
+      "step": 233
+    },
+    {
+      "epoch": 0.15019255455712452,
+      "grad_norm": 0.01053924672305584,
+      "learning_rate": 9.626821812269415e-05,
+      "loss": 0.543,
+      "step": 234
+    },
+    {
+      "epoch": 0.15083440308087293,
+      "grad_norm": 0.011286425404250622,
+      "learning_rate": 9.622871009068588e-05,
+      "loss": 0.5039,
+      "step": 235
+    },
+    {
+      "epoch": 0.1514762516046213,
+      "grad_norm": 0.012574701569974422,
+      "learning_rate": 9.618900221867077e-05,
+      "loss": 0.5391,
+      "step": 236
+    },
+    {
+      "epoch": 0.1521181001283697,
+      "grad_norm": 0.012315727770328522,
+      "learning_rate": 9.614909467830022e-05,
+      "loss": 0.5273,
+      "step": 237
+    },
+    {
+      "epoch": 0.1527599486521181,
+      "grad_norm": 0.010991916060447693,
+      "learning_rate": 9.610898764208873e-05,
+      "loss": 0.5469,
+      "step": 238
+    },
+    {
+      "epoch": 0.1534017971758665,
+      "grad_norm": 0.011746605858206749,
+      "learning_rate": 9.60686812834132e-05,
+      "loss": 0.5547,
+      "step": 239
+    },
+    {
+      "epoch": 0.1540436456996149,
+      "grad_norm": 0.011498944833874702,
+      "learning_rate": 9.602817577651217e-05,
+      "loss": 0.5586,
+      "step": 240
+    },
+    {
+      "epoch": 0.15468549422336328,
+      "grad_norm": 0.011620343662798405,
+      "learning_rate": 9.598747129648505e-05,
+      "loss": 0.5312,
+      "step": 241
+    },
+    {
+      "epoch": 0.15532734274711169,
+      "grad_norm": 0.010271632112562656,
+      "learning_rate": 9.594656801929145e-05,
+      "loss": 0.5586,
+      "step": 242
+    },
+    {
+      "epoch": 0.15596919127086006,
+      "grad_norm": 0.013585187494754791,
+      "learning_rate": 9.590546612175024e-05,
+      "loss": 0.5625,
+      "step": 243
+    },
+    {
+      "epoch": 0.15661103979460847,
+      "grad_norm": 0.012161128222942352,
+      "learning_rate": 9.586416578153903e-05,
+      "loss": 0.5508,
+      "step": 244
+    },
+    {
+      "epoch": 0.15725288831835688,
+      "grad_norm": 0.011398044414818287,
+      "learning_rate": 9.582266717719314e-05,
+      "loss": 0.5703,
+      "step": 245
+    },
+    {
+      "epoch": 0.15789473684210525,
+      "grad_norm": 0.012302602641284466,
+      "learning_rate": 9.578097048810503e-05,
+      "loss": 0.5469,
+      "step": 246
+    },
+    {
+      "epoch": 0.15853658536585366,
+      "grad_norm": 0.012995855882763863,
+      "learning_rate": 9.573907589452347e-05,
+      "loss": 0.5195,
+      "step": 247
+    },
+    {
+      "epoch": 0.15917843388960207,
+      "grad_norm": 0.011635944247245789,
+      "learning_rate": 9.569698357755267e-05,
+      "loss": 0.5273,
+      "step": 248
+    },
+    {
+      "epoch": 0.15982028241335045,
+      "grad_norm": 0.011981082148849964,
+      "learning_rate": 9.565469371915164e-05,
+      "loss": 0.5234,
+      "step": 249
+    },
+    {
+      "epoch": 0.16046213093709885,
+      "grad_norm": 0.013345803134143353,
+      "learning_rate": 9.561220650213326e-05,
+      "loss": 0.5547,
+      "step": 250
+    },
+    {
+      "epoch": 0.16110397946084723,
+      "grad_norm": 0.010194964706897736,
+      "learning_rate": 9.556952211016366e-05,
+      "loss": 0.4805,
+      "step": 251
+    },
+    {
+      "epoch": 0.16174582798459564,
+      "grad_norm": 0.010166475549340248,
+      "learning_rate": 9.552664072776123e-05,
+      "loss": 0.4844,
+      "step": 252
+    },
+    {
+      "epoch": 0.16238767650834404,
+      "grad_norm": 0.011069612577557564,
+      "learning_rate": 9.548356254029598e-05,
+      "loss": 0.5,
+      "step": 253
+    },
+    {
+      "epoch": 0.16302952503209242,
+      "grad_norm": 0.014476699754595757,
+      "learning_rate": 9.544028773398867e-05,
+      "loss": 0.5352,
+      "step": 254
+    },
+    {
+      "epoch": 0.16367137355584083,
+      "grad_norm": 0.013018032535910606,
+      "learning_rate": 9.539681649591002e-05,
+      "loss": 0.5234,
+      "step": 255
+    },
+    {
+      "epoch": 0.1643132220795892,
+      "grad_norm": 0.009963922202587128,
+      "learning_rate": 9.535314901397985e-05,
+      "loss": 0.5273,
+      "step": 256
+    },
+    {
+      "epoch": 0.1649550706033376,
+      "grad_norm": 0.011673172004520893,
+      "learning_rate": 9.530928547696639e-05,
+      "loss": 0.5586,
+      "step": 257
+    },
+    {
+      "epoch": 0.16559691912708602,
+      "grad_norm": 0.011493822559714317,
+      "learning_rate": 9.52652260744853e-05,
+      "loss": 0.5391,
+      "step": 258
+    },
+    {
+      "epoch": 0.1662387676508344,
+      "grad_norm": 0.015019495040178299,
+      "learning_rate": 9.522097099699903e-05,
+      "loss": 0.5234,
+      "step": 259
+    },
+    {
+      "epoch": 0.1668806161745828,
+      "grad_norm": 0.012797566130757332,
+      "learning_rate": 9.517652043581583e-05,
+      "loss": 0.5,
+      "step": 260
+    },
+    {
+      "epoch": 0.16752246469833118,
+      "grad_norm": 0.012790082022547722,
+      "learning_rate": 9.513187458308904e-05,
+      "loss": 0.5,
+      "step": 261
+    },
+    {
+      "epoch": 0.1681643132220796,
+      "grad_norm": 0.013730279169976711,
+      "learning_rate": 9.508703363181621e-05,
+      "loss": 0.5469,
+      "step": 262
+    },
+    {
+      "epoch": 0.168806161745828,
+      "grad_norm": 0.011931111104786396,
+      "learning_rate": 9.504199777583824e-05,
+      "loss": 0.5625,
+      "step": 263
+    },
+    {
+      "epoch": 0.16944801026957637,
+      "grad_norm": 0.012304815463721752,
+      "learning_rate": 9.499676720983864e-05,
+      "loss": 0.5234,
+      "step": 264
+    },
+    {
+      "epoch": 0.17008985879332478,
+      "grad_norm": 0.012368976138532162,
+      "learning_rate": 9.495134212934256e-05,
+      "loss": 0.5273,
+      "step": 265
+    },
+    {
+      "epoch": 0.17073170731707318,
+      "grad_norm": 0.01270821038633585,
+      "learning_rate": 9.490572273071603e-05,
+      "loss": 0.5234,
+      "step": 266
+    },
+    {
+      "epoch": 0.17137355584082156,
+      "grad_norm": 0.01076431479305029,
+      "learning_rate": 9.48599092111651e-05,
+      "loss": 0.5156,
+      "step": 267
+    },
+    {
+      "epoch": 0.17201540436456997,
+      "grad_norm": 0.013268671929836273,
+      "learning_rate": 9.481390176873498e-05,
+      "loss": 0.4961,
+      "step": 268
+    },
+    {
+      "epoch": 0.17265725288831835,
+      "grad_norm": 0.012573538348078728,
+      "learning_rate": 9.476770060230915e-05,
+      "loss": 0.4883,
+      "step": 269
+    },
+    {
+      "epoch": 0.17329910141206675,
+      "grad_norm": 0.011922149918973446,
+      "learning_rate": 9.472130591160855e-05,
+      "loss": 0.5625,
+      "step": 270
+    },
+    {
+      "epoch": 0.17394094993581516,
+      "grad_norm": 0.012727407738566399,
+      "learning_rate": 9.467471789719072e-05,
+      "loss": 0.5586,
+      "step": 271
+    },
+    {
+      "epoch": 0.17458279845956354,
+      "grad_norm": 0.01120324619114399,
+      "learning_rate": 9.462793676044885e-05,
+      "loss": 0.4902,
+      "step": 272
+    },
+    {
+      "epoch": 0.17522464698331194,
+      "grad_norm": 0.013182485476136208,
+      "learning_rate": 9.458096270361104e-05,
+      "loss": 0.5391,
+      "step": 273
+    },
+    {
+      "epoch": 0.17586649550706032,
+      "grad_norm": 0.010877820663154125,
+      "learning_rate": 9.453379592973928e-05,
+      "loss": 0.5234,
+      "step": 274
+    },
+    {
+      "epoch": 0.17650834403080873,
+      "grad_norm": 0.010631302371621132,
+      "learning_rate": 9.448643664272876e-05,
+      "loss": 0.5117,
+      "step": 275
+    },
+    {
+      "epoch": 0.17715019255455713,
+      "grad_norm": 0.011218084953725338,
+      "learning_rate": 9.443888504730673e-05,
+      "loss": 0.4707,
+      "step": 276
+    },
+    {
+      "epoch": 0.1777920410783055,
+      "grad_norm": 0.013051497749984264,
+      "learning_rate": 9.439114134903191e-05,
+      "loss": 0.4668,
+      "step": 277
+    },
+    {
+      "epoch": 0.17843388960205392,
+      "grad_norm": 0.014543252997100353,
+      "learning_rate": 9.434320575429334e-05,
+      "loss": 0.5469,
+      "step": 278
+    },
+    {
+      "epoch": 0.1790757381258023,
+      "grad_norm": 0.010615730658173561,
+      "learning_rate": 9.429507847030963e-05,
+      "loss": 0.5352,
+      "step": 279
+    },
+    {
+      "epoch": 0.1797175866495507,
+      "grad_norm": 0.014637456275522709,
+      "learning_rate": 9.424675970512808e-05,
+      "loss": 0.5039,
+      "step": 280
+    },
+    {
+      "epoch": 0.1803594351732991,
+      "grad_norm": 0.01204176526516676,
+      "learning_rate": 9.419824966762367e-05,
+      "loss": 0.5039,
+      "step": 281
+    },
+    {
+      "epoch": 0.1810012836970475,
+      "grad_norm": 0.013622627593576908,
+      "learning_rate": 9.414954856749828e-05,
+      "loss": 0.4883,
+      "step": 282
+    },
+    {
+      "epoch": 0.1816431322207959,
+      "grad_norm": 0.012876857072114944,
+      "learning_rate": 9.41006566152797e-05,
+      "loss": 0.5039,
+      "step": 283
+    },
+    {
+      "epoch": 0.1822849807445443,
+      "grad_norm": 0.011097314767539501,
+      "learning_rate": 9.405157402232072e-05,
+      "loss": 0.5234,
+      "step": 284
+    },
+    {
+      "epoch": 0.18292682926829268,
+      "grad_norm": 0.012606339529156685,
+      "learning_rate": 9.400230100079829e-05,
+      "loss": 0.5078,
+      "step": 285
+    },
+    {
+      "epoch": 0.18356867779204109,
+      "grad_norm": 0.013987897895276546,
+      "learning_rate": 9.395283776371253e-05,
+      "loss": 0.5156,
+      "step": 286
+    },
+    {
+      "epoch": 0.18421052631578946,
+      "grad_norm": 0.010777951218187809,
+      "learning_rate": 9.390318452488583e-05,
+      "loss": 0.5195,
+      "step": 287
+    },
+    {
+      "epoch": 0.18485237483953787,
+      "grad_norm": 0.012862320989370346,
+      "learning_rate": 9.385334149896195e-05,
+      "loss": 0.543,
+      "step": 288
+    },
+    {
+      "epoch": 0.18549422336328628,
+      "grad_norm": 0.013322705402970314,
+      "learning_rate": 9.380330890140505e-05,
+      "loss": 0.5156,
+      "step": 289
+    },
+    {
+      "epoch": 0.18613607188703465,
+      "grad_norm": 0.013143341057002544,
+      "learning_rate": 9.37530869484988e-05,
+      "loss": 0.5117,
+      "step": 290
+    },
+    {
+      "epoch": 0.18677792041078306,
+      "grad_norm": 0.012022369541227818,
+      "learning_rate": 9.37026758573454e-05,
+      "loss": 0.4707,
+      "step": 291
+    },
+    {
+      "epoch": 0.18741976893453144,
+      "grad_norm": 0.012463578954339027,
+      "learning_rate": 9.365207584586471e-05,
+      "loss": 0.5,
+      "step": 292
+    },
+    {
+      "epoch": 0.18806161745827984,
+      "grad_norm": 0.01562550477683544,
+      "learning_rate": 9.360128713279321e-05,
+      "loss": 0.5391,
+      "step": 293
+    },
+    {
+      "epoch": 0.18870346598202825,
+      "grad_norm": 0.01224803552031517,
+      "learning_rate": 9.355030993768314e-05,
+      "loss": 0.4922,
+      "step": 294
+    },
+    {
+      "epoch": 0.18934531450577663,
+      "grad_norm": 0.012806670740246773,
+      "learning_rate": 9.349914448090156e-05,
+      "loss": 0.5078,
+      "step": 295
+    },
+    {
+      "epoch": 0.18998716302952504,
+      "grad_norm": 0.014918792992830276,
+      "learning_rate": 9.344779098362926e-05,
+      "loss": 0.4902,
+      "step": 296
+    },
+    {
+      "epoch": 0.19062901155327341,
+      "grad_norm": 0.015880340710282326,
+      "learning_rate": 9.339624966785999e-05,
+      "loss": 0.4961,
+      "step": 297
+    },
+    {
+      "epoch": 0.19127086007702182,
+      "grad_norm": 0.016379036009311676,
+      "learning_rate": 9.334452075639938e-05,
+      "loss": 0.5117,
+      "step": 298
+    },
+    {
+      "epoch": 0.19191270860077023,
+      "grad_norm": 0.013551746495068073,
+      "learning_rate": 9.329260447286399e-05,
+      "loss": 0.5391,
+      "step": 299
+    },
+    {
+      "epoch": 0.1925545571245186,
+      "grad_norm": 0.012869910337030888,
+      "learning_rate": 9.32405010416804e-05,
+      "loss": 0.4961,
+      "step": 300
+    },
+    {
+      "epoch": 0.193196405648267,
+      "grad_norm": 0.01776876486837864,
+      "learning_rate": 9.318821068808418e-05,
+      "loss": 0.4805,
+      "step": 301
+    },
+    {
+      "epoch": 0.19383825417201542,
+      "grad_norm": 0.01472089160233736,
+      "learning_rate": 9.313573363811895e-05,
+      "loss": 0.5078,
+      "step": 302
+    },
+    {
+      "epoch": 0.1944801026957638,
+      "grad_norm": 0.012855904176831245,
+      "learning_rate": 9.308307011863537e-05,
+      "loss": 0.5,
+      "step": 303
+    },
+    {
+      "epoch": 0.1951219512195122,
+      "grad_norm": 0.015342450700700283,
+      "learning_rate": 9.303022035729019e-05,
+      "loss": 0.5156,
+      "step": 304
+    },
+    {
+      "epoch": 0.19576379974326058,
+      "grad_norm": 0.014027988538146019,
+      "learning_rate": 9.297718458254528e-05,
+      "loss": 0.4922,
+      "step": 305
+    },
+    {
+      "epoch": 0.196405648267009,
+      "grad_norm": 0.014229445718228817,
+      "learning_rate": 9.29239630236666e-05,
+      "loss": 0.4922,
+      "step": 306
+    },
+    {
+      "epoch": 0.1970474967907574,
+      "grad_norm": 0.011061770841479301,
+      "learning_rate": 9.287055591072319e-05,
+      "loss": 0.457,
+      "step": 307
+    },
+    {
+      "epoch": 0.19768934531450577,
+      "grad_norm": 0.012393632903695107,
+      "learning_rate": 9.28169634745863e-05,
+      "loss": 0.5078,
+      "step": 308
+    },
+    {
+      "epoch": 0.19833119383825418,
+      "grad_norm": 0.014692421071231365,
+      "learning_rate": 9.276318594692822e-05,
+      "loss": 0.4707,
+      "step": 309
+    },
+    {
+      "epoch": 0.19897304236200256,
+      "grad_norm": 0.013025031425058842,
+      "learning_rate": 9.270922356022142e-05,
+      "loss": 0.4883,
+      "step": 310
+    },
+    {
+      "epoch": 0.19961489088575096,
+      "grad_norm": 0.011297275312244892,
+      "learning_rate": 9.265507654773747e-05,
+      "loss": 0.5,
+      "step": 311
+    },
+    {
+      "epoch": 0.20025673940949937,
+      "grad_norm": 0.012788046151399612,
+      "learning_rate": 9.260074514354603e-05,
+      "loss": 0.5,
+      "step": 312
+    },
+    {
+      "epoch": 0.20089858793324775,
+      "grad_norm": 0.012072072364389896,
+      "learning_rate": 9.254622958251389e-05,
+      "loss": 0.4902,
+      "step": 313
+    },
+    {
+      "epoch": 0.20154043645699615,
+      "grad_norm": 0.011217731982469559,
+      "learning_rate": 9.249153010030391e-05,
+      "loss": 0.5,
+      "step": 314
+    },
+    {
+      "epoch": 0.20218228498074453,
+      "grad_norm": 0.015587392263114452,
+      "learning_rate": 9.243664693337404e-05,
+      "loss": 0.4766,
+      "step": 315
+    },
+    {
+      "epoch": 0.20282413350449294,
+      "grad_norm": 0.014844152145087719,
+      "learning_rate": 9.23815803189762e-05,
+      "loss": 0.4668,
+      "step": 316
+    },
+    {
+      "epoch": 0.20346598202824134,
+      "grad_norm": 0.012265946716070175,
+      "learning_rate": 9.232633049515541e-05,
+      "loss": 0.5078,
+      "step": 317
+    },
+    {
+      "epoch": 0.20410783055198972,
+      "grad_norm": 0.016097310930490494,
+      "learning_rate": 9.227089770074864e-05,
+      "loss": 0.4746,
+      "step": 318
+    },
+    {
+      "epoch": 0.20474967907573813,
+      "grad_norm": 0.013965766876935959,
+      "learning_rate": 9.22152821753838e-05,
+      "loss": 0.4766,
+      "step": 319
+    },
+    {
+      "epoch": 0.20539152759948653,
+      "grad_norm": 0.012059083208441734,
+      "learning_rate": 9.215948415947875e-05,
+      "loss": 0.5078,
+      "step": 320
+    },
+    {
+      "epoch": 0.2060333761232349,
+      "grad_norm": 0.013087877072393894,
+      "learning_rate": 9.210350389424021e-05,
+      "loss": 0.5273,
+      "step": 321
+    },
+    {
+      "epoch": 0.20667522464698332,
+      "grad_norm": 0.016379237174987793,
+      "learning_rate": 9.204734162166276e-05,
+      "loss": 0.5039,
+      "step": 322
+    },
+    {
+      "epoch": 0.2073170731707317,
+      "grad_norm": 0.014362157322466373,
+      "learning_rate": 9.199099758452774e-05,
+      "loss": 0.4883,
+      "step": 323
+    },
+    {
+      "epoch": 0.2079589216944801,
+      "grad_norm": 0.017146103084087372,
+      "learning_rate": 9.193447202640226e-05,
+      "loss": 0.5,
+      "step": 324
+    },
+    {
+      "epoch": 0.2086007702182285,
+      "grad_norm": 0.011025834828615189,
+      "learning_rate": 9.187776519163811e-05,
+      "loss": 0.4648,
+      "step": 325
+    },
+    {
+      "epoch": 0.2092426187419769,
+      "grad_norm": 0.013537570834159851,
+      "learning_rate": 9.182087732537068e-05,
+      "loss": 0.4922,
+      "step": 326
+    },
+    {
+      "epoch": 0.2098844672657253,
+      "grad_norm": 0.012621462345123291,
+      "learning_rate": 9.176380867351801e-05,
+      "loss": 0.5078,
+      "step": 327
+    },
+    {
+      "epoch": 0.21052631578947367,
+      "grad_norm": 0.013818283565342426,
+      "learning_rate": 9.170655948277956e-05,
+      "loss": 0.4922,
+      "step": 328
+    },
+    {
+      "epoch": 0.21116816431322208,
+      "grad_norm": 0.011845164000988007,
+      "learning_rate": 9.164913000063529e-05,
+      "loss": 0.5195,
+      "step": 329
+    },
+    {
+      "epoch": 0.21181001283697048,
+      "grad_norm": 0.014658152125775814,
+      "learning_rate": 9.159152047534454e-05,
+      "loss": 0.4883,
+      "step": 330
+    },
+    {
+      "epoch": 0.21245186136071886,
+      "grad_norm": 0.013154874555766582,
+      "learning_rate": 9.153373115594489e-05,
+      "loss": 0.5312,
+      "step": 331
+    },
+    {
+      "epoch": 0.21309370988446727,
+      "grad_norm": 0.013952475972473621,
+      "learning_rate": 9.147576229225121e-05,
+      "loss": 0.4727,
+      "step": 332
+    },
+    {
+      "epoch": 0.21373555840821565,
+      "grad_norm": 0.012911539524793625,
+      "learning_rate": 9.14176141348545e-05,
+      "loss": 0.5156,
+      "step": 333
+    },
+    {
+      "epoch": 0.21437740693196405,
+      "grad_norm": 0.012119744904339314,
+      "learning_rate": 9.135928693512077e-05,
+      "loss": 0.4922,
+      "step": 334
+    },
+    {
+      "epoch": 0.21501925545571246,
+      "grad_norm": 0.013211081735789776,
+      "learning_rate": 9.130078094519008e-05,
+      "loss": 0.5078,
+      "step": 335
+    },
+    {
+      "epoch": 0.21566110397946084,
+      "grad_norm": 0.010815898887813091,
+      "learning_rate": 9.12420964179753e-05,
+      "loss": 0.4688,
+      "step": 336
+    },
+    {
+      "epoch": 0.21630295250320924,
+      "grad_norm": 0.011093908920884132,
+      "learning_rate": 9.118323360716117e-05,
+      "loss": 0.4883,
+      "step": 337
+    },
+    {
+      "epoch": 0.21694480102695765,
+      "grad_norm": 0.011569825932383537,
+      "learning_rate": 9.112419276720303e-05,
+      "loss": 0.4727,
+      "step": 338
+    },
+    {
+      "epoch": 0.21758664955070603,
+      "grad_norm": 0.01549893245100975,
+      "learning_rate": 9.106497415332591e-05,
+      "loss": 0.4688,
+      "step": 339
+    },
+    {
+      "epoch": 0.21822849807445444,
+      "grad_norm": 0.012803218327462673,
+      "learning_rate": 9.100557802152328e-05,
+      "loss": 0.5078,
+      "step": 340
+    },
+    {
+      "epoch": 0.21887034659820281,
+      "grad_norm": 0.015469119884073734,
+      "learning_rate": 9.094600462855598e-05,
+      "loss": 0.5156,
+      "step": 341
+    },
+    {
+      "epoch": 0.21951219512195122,
+      "grad_norm": 0.013518411666154861,
+      "learning_rate": 9.088625423195115e-05,
+      "loss": 0.4766,
+      "step": 342
+    },
+    {
+      "epoch": 0.22015404364569963,
+      "grad_norm": 0.01218617707490921,
+      "learning_rate": 9.08263270900011e-05,
+      "loss": 0.4824,
+      "step": 343
+    },
+    {
+      "epoch": 0.220795892169448,
+      "grad_norm": 0.011522636748850346,
+      "learning_rate": 9.076622346176216e-05,
+      "loss": 0.4395,
+      "step": 344
+    },
+    {
+      "epoch": 0.2214377406931964,
+      "grad_norm": 0.01178770512342453,
+      "learning_rate": 9.070594360705358e-05,
+      "loss": 0.4531,
+      "step": 345
+    },
+    {
+      "epoch": 0.2220795892169448,
+      "grad_norm": 0.012270083650946617,
+      "learning_rate": 9.064548778645646e-05,
+      "loss": 0.4648,
+      "step": 346
+    },
+    {
+      "epoch": 0.2227214377406932,
+      "grad_norm": 0.012514191679656506,
+      "learning_rate": 9.058485626131252e-05,
+      "loss": 0.4766,
+      "step": 347
+    },
+    {
+      "epoch": 0.2233632862644416,
+      "grad_norm": 0.01187925785779953,
+      "learning_rate": 9.052404929372305e-05,
+      "loss": 0.4609,
+      "step": 348
+    },
+    {
+      "epoch": 0.22400513478818998,
+      "grad_norm": 0.013160323724150658,
+      "learning_rate": 9.046306714654775e-05,
+      "loss": 0.4609,
+      "step": 349
+    },
+    {
+      "epoch": 0.2246469833119384,
+      "grad_norm": 0.01244362536817789,
+      "learning_rate": 9.04019100834036e-05,
+      "loss": 0.4609,
+      "step": 350
+    },
+    {
+      "epoch": 0.22528883183568676,
+      "grad_norm": 0.01528861839324236,
+      "learning_rate": 9.03405783686637e-05,
+      "loss": 0.4844,
+      "step": 351
+    },
+    {
+      "epoch": 0.22593068035943517,
+      "grad_norm": 0.010898303240537643,
+      "learning_rate": 9.027907226745616e-05,
+      "loss": 0.4785,
+      "step": 352
+    },
+    {
+      "epoch": 0.22657252888318358,
+      "grad_norm": 0.013835811987519264,
+      "learning_rate": 9.021739204566295e-05,
+      "loss": 0.4805,
+      "step": 353
+    },
+    {
+      "epoch": 0.22721437740693196,
+      "grad_norm": 0.012497610412538052,
+      "learning_rate": 9.015553796991869e-05,
+      "loss": 0.4922,
+      "step": 354
+    },
+    {
+      "epoch": 0.22785622593068036,
+      "grad_norm": 0.011201409623026848,
+      "learning_rate": 9.009351030760958e-05,
+      "loss": 0.4648,
+      "step": 355
+    },
+    {
+      "epoch": 0.22849807445442877,
+      "grad_norm": 0.015274932608008385,
+      "learning_rate": 9.003130932687223e-05,
+      "loss": 0.4883,
+      "step": 356
+    },
+    {
+      "epoch": 0.22913992297817715,
+      "grad_norm": 0.013798701576888561,
+      "learning_rate": 8.996893529659243e-05,
+      "loss": 0.4766,
+      "step": 357
+    },
+    {
+      "epoch": 0.22978177150192555,
+      "grad_norm": 0.013277494348585606,
+      "learning_rate": 8.990638848640407e-05,
+      "loss": 0.457,
+      "step": 358
+    },
+    {
+      "epoch": 0.23042362002567393,
+      "grad_norm": 0.013865608721971512,
+      "learning_rate": 8.984366916668795e-05,
+      "loss": 0.4824,
+      "step": 359
+    },
+    {
+      "epoch": 0.23106546854942234,
+      "grad_norm": 0.01533426996320486,
+      "learning_rate": 8.978077760857058e-05,
+      "loss": 0.4609,
+      "step": 360
+    },
+    {
+      "epoch": 0.23170731707317074,
+      "grad_norm": 0.013055570423603058,
+      "learning_rate": 8.971771408392302e-05,
+      "loss": 0.4629,
+      "step": 361
+    },
+    {
+      "epoch": 0.23234916559691912,
+      "grad_norm": 0.011497069150209427,
+      "learning_rate": 8.965447886535978e-05,
+      "loss": 0.4727,
+      "step": 362
+    },
+    {
+      "epoch": 0.23299101412066753,
+      "grad_norm": 0.01564948260784149,
+      "learning_rate": 8.95910722262375e-05,
+      "loss": 0.4863,
+      "step": 363
+    },
+    {
+      "epoch": 0.2336328626444159,
+      "grad_norm": 0.010947753675282001,
+      "learning_rate": 8.95274944406539e-05,
+      "loss": 0.4512,
+      "step": 364
+    },
+    {
+      "epoch": 0.2342747111681643,
+      "grad_norm": 0.011532683856785297,
+      "learning_rate": 8.946374578344653e-05,
+      "loss": 0.4551,
+      "step": 365
+    },
+    {
+      "epoch": 0.23491655969191272,
+      "grad_norm": 0.012891596183180809,
+      "learning_rate": 8.939982653019157e-05,
+      "loss": 0.4883,
+      "step": 366
+    },
+    {
+      "epoch": 0.2355584082156611,
+      "grad_norm": 0.014495361596345901,
+      "learning_rate": 8.933573695720267e-05,
+      "loss": 0.4727,
+      "step": 367
+    },
+    {
+      "epoch": 0.2362002567394095,
+      "grad_norm": 0.01764221116900444,
+      "learning_rate": 8.927147734152978e-05,
+      "loss": 0.4766,
+      "step": 368
+    },
+    {
+      "epoch": 0.23684210526315788,
+      "grad_norm": 0.014389213174581528,
+      "learning_rate": 8.920704796095788e-05,
+      "loss": 0.4922,
+      "step": 369
+    },
+    {
+      "epoch": 0.2374839537869063,
+      "grad_norm": 0.01599840074777603,
+      "learning_rate": 8.914244909400585e-05,
+      "loss": 0.4551,
+      "step": 370
+    },
+    {
+      "epoch": 0.2381258023106547,
+      "grad_norm": 0.011929795145988464,
+      "learning_rate": 8.90776810199252e-05,
+      "loss": 0.4805,
+      "step": 371
+    },
+    {
+      "epoch": 0.23876765083440307,
+      "grad_norm": 0.011343288235366344,
+      "learning_rate": 8.901274401869893e-05,
+      "loss": 0.4805,
+      "step": 372
+    },
+    {
+      "epoch": 0.23940949935815148,
+      "grad_norm": 0.014292542822659016,
+      "learning_rate": 8.894763837104026e-05,
+      "loss": 0.4883,
+      "step": 373
+    },
+    {
+      "epoch": 0.24005134788189988,
+      "grad_norm": 0.013415781781077385,
+      "learning_rate": 8.888236435839147e-05,
+      "loss": 0.4766,
+      "step": 374
+    },
+    {
+      "epoch": 0.24069319640564826,
+      "grad_norm": 0.013569642789661884,
+      "learning_rate": 8.881692226292269e-05,
+      "loss": 0.5,
+      "step": 375
+    },
+    {
+      "epoch": 0.24133504492939667,
+      "grad_norm": 0.011331136338412762,
+      "learning_rate": 8.875131236753051e-05,
+      "loss": 0.4609,
+      "step": 376
+    },
+    {
+      "epoch": 0.24197689345314505,
+      "grad_norm": 0.013599894009530544,
+      "learning_rate": 8.868553495583707e-05,
+      "loss": 0.4766,
+      "step": 377
+    },
+    {
+      "epoch": 0.24261874197689345,
+      "grad_norm": 0.011140760965645313,
+      "learning_rate": 8.861959031218855e-05,
+      "loss": 0.4746,
+      "step": 378
+    },
+    {
+      "epoch": 0.24326059050064186,
+      "grad_norm": 0.01980205811560154,
+      "learning_rate": 8.855347872165406e-05,
+      "loss": 0.4727,
+      "step": 379
+    },
+    {
+      "epoch": 0.24390243902439024,
+      "grad_norm": 0.013902513310313225,
+      "learning_rate": 8.848720047002446e-05,
+      "loss": 0.4727,
+      "step": 380
+    },
+    {
+      "epoch": 0.24454428754813864,
+      "grad_norm": 0.01137533038854599,
+      "learning_rate": 8.842075584381096e-05,
+      "loss": 0.4648,
+      "step": 381
+    },
+    {
+      "epoch": 0.24518613607188702,
+      "grad_norm": 0.019017167389392853,
+      "learning_rate": 8.835414513024409e-05,
+      "loss": 0.4922,
+      "step": 382
+    },
+    {
+      "epoch": 0.24582798459563543,
+      "grad_norm": 0.014492962509393692,
+      "learning_rate": 8.828736861727225e-05,
+      "loss": 0.4844,
+      "step": 383
+    },
+    {
+      "epoch": 0.24646983311938384,
+      "grad_norm": 0.018102675676345825,
+      "learning_rate": 8.822042659356065e-05,
+      "loss": 0.4746,
+      "step": 384
+    },
+    {
+      "epoch": 0.2471116816431322,
+      "grad_norm": 0.012978147715330124,
+      "learning_rate": 8.815331934848996e-05,
+      "loss": 0.459,
+      "step": 385
+    },
+    {
+      "epoch": 0.24775353016688062,
+      "grad_norm": 0.011396810412406921,
+      "learning_rate": 8.808604717215503e-05,
+      "loss": 0.4492,
+      "step": 386
+    },
+    {
+      "epoch": 0.248395378690629,
+      "grad_norm": 0.013627592474222183,
+      "learning_rate": 8.801861035536374e-05,
+      "loss": 0.4805,
+      "step": 387
+    },
+    {
+      "epoch": 0.2490372272143774,
+      "grad_norm": 0.013477048836648464,
+      "learning_rate": 8.795100918963566e-05,
+      "loss": 0.4785,
+      "step": 388
+    },
+    {
+      "epoch": 0.2496790757381258,
+      "grad_norm": 0.012124343775212765,
+      "learning_rate": 8.78832439672008e-05,
+      "loss": 0.4727,
+      "step": 389
+    },
+    {
+      "epoch": 0.2503209242618742,
+      "grad_norm": 0.011770759709179401,
+      "learning_rate": 8.781531498099844e-05,
+      "loss": 0.4629,
+      "step": 390
+    },
+    {
+      "epoch": 0.25096277278562257,
+      "grad_norm": 0.011974097229540348,
+      "learning_rate": 8.774722252467566e-05,
+      "loss": 0.4766,
+      "step": 391
+    },
+    {
+      "epoch": 0.251604621309371,
+      "grad_norm": 0.012057865038514137,
+      "learning_rate": 8.767896689258635e-05,
+      "loss": 0.4531,
+      "step": 392
+    },
+    {
+      "epoch": 0.2522464698331194,
+      "grad_norm": 0.012654995545744896,
+      "learning_rate": 8.761054837978964e-05,
+      "loss": 0.4961,
+      "step": 393
+    },
+    {
+      "epoch": 0.25288831835686776,
+      "grad_norm": 0.01383557915687561,
+      "learning_rate": 8.754196728204886e-05,
+      "loss": 0.4727,
+      "step": 394
+    },
+    {
+      "epoch": 0.2535301668806162,
+      "grad_norm": 0.011721700429916382,
+      "learning_rate": 8.747322389583013e-05,
+      "loss": 0.4766,
+      "step": 395
+    },
+    {
+      "epoch": 0.25417201540436457,
+      "grad_norm": 0.012775659561157227,
+      "learning_rate": 8.740431851830118e-05,
+      "loss": 0.4707,
+      "step": 396
+    },
+    {
+      "epoch": 0.25481386392811295,
+      "grad_norm": 0.013818454928696156,
+      "learning_rate": 8.733525144732991e-05,
+      "loss": 0.4453,
+      "step": 397
+    },
+    {
+      "epoch": 0.2554557124518614,
+      "grad_norm": 0.014101814478635788,
+      "learning_rate": 8.726602298148325e-05,
+      "loss": 0.457,
+      "step": 398
+    },
+    {
+      "epoch": 0.25609756097560976,
+      "grad_norm": 0.013880562037229538,
+      "learning_rate": 8.719663342002585e-05,
+      "loss": 0.4727,
+      "step": 399
+    },
+    {
+      "epoch": 0.25673940949935814,
+      "grad_norm": 0.015757685527205467,
+      "learning_rate": 8.71270830629187e-05,
+      "loss": 0.4941,
+      "step": 400
+    },
+    {
+      "epoch": 0.2573812580231066,
+      "grad_norm": 0.01359548605978489,
+      "learning_rate": 8.705737221081789e-05,
+      "loss": 0.4883,
+      "step": 401
+    },
+    {
+      "epoch": 0.25802310654685495,
+      "grad_norm": 0.0140383904799819,
+      "learning_rate": 8.698750116507333e-05,
+      "loss": 0.4219,
+      "step": 402
+    },
+    {
+      "epoch": 0.25866495507060333,
+      "grad_norm": 0.016375742852687836,
+      "learning_rate": 8.691747022772742e-05,
+      "loss": 0.4883,
+      "step": 403
+    },
+    {
+      "epoch": 0.2593068035943517,
+      "grad_norm": 0.0133128697052598,
+      "learning_rate": 8.684727970151374e-05,
+      "loss": 0.4805,
+      "step": 404
+    },
+    {
+      "epoch": 0.25994865211810014,
+      "grad_norm": 0.013061772100627422,
+      "learning_rate": 8.677692988985575e-05,
+      "loss": 0.4766,
+      "step": 405
+    },
+    {
+      "epoch": 0.2605905006418485,
+      "grad_norm": 0.013719551265239716,
+      "learning_rate": 8.670642109686546e-05,
+      "loss": 0.457,
+      "step": 406
+    },
+    {
+      "epoch": 0.2612323491655969,
+      "grad_norm": 0.013394895009696484,
+      "learning_rate": 8.663575362734219e-05,
+      "loss": 0.4844,
+      "step": 407
+    },
+    {
+      "epoch": 0.26187419768934533,
+      "grad_norm": 0.014179762452840805,
+      "learning_rate": 8.656492778677112e-05,
+      "loss": 0.4512,
+      "step": 408
+    },
+    {
+      "epoch": 0.2625160462130937,
+      "grad_norm": 0.011698760092258453,
+      "learning_rate": 8.649394388132211e-05,
+      "loss": 0.4609,
+      "step": 409
+    },
+    {
+      "epoch": 0.2631578947368421,
+      "grad_norm": 0.012449960224330425,
+      "learning_rate": 8.642280221784828e-05,
+      "loss": 0.4922,
+      "step": 410
+    },
+    {
+      "epoch": 0.2637997432605905,
+      "grad_norm": 0.013107222504913807,
+      "learning_rate": 8.635150310388471e-05,
+      "loss": 0.4531,
+      "step": 411
+    },
+    {
+      "epoch": 0.2644415917843389,
+      "grad_norm": 0.012861665338277817,
+      "learning_rate": 8.628004684764715e-05,
+      "loss": 0.4258,
+      "step": 412
+    },
+    {
+      "epoch": 0.2650834403080873,
+      "grad_norm": 0.013430675491690636,
+      "learning_rate": 8.620843375803058e-05,
+      "loss": 0.4805,
+      "step": 413
+    },
+    {
+      "epoch": 0.26572528883183566,
+      "grad_norm": 0.01297419611364603,
+      "learning_rate": 8.613666414460806e-05,
+      "loss": 0.457,
+      "step": 414
+    },
+    {
+      "epoch": 0.2663671373555841,
+      "grad_norm": 0.01470587681978941,
+      "learning_rate": 8.606473831762916e-05,
+      "loss": 0.4531,
+      "step": 415
+    },
+    {
+      "epoch": 0.26700898587933247,
+      "grad_norm": 0.012723923660814762,
+      "learning_rate": 8.599265658801883e-05,
+      "loss": 0.4531,
+      "step": 416
+    },
+    {
+      "epoch": 0.26765083440308085,
+      "grad_norm": 0.010943865403532982,
+      "learning_rate": 8.592041926737591e-05,
+      "loss": 0.4688,
+      "step": 417
+    },
+    {
+      "epoch": 0.2682926829268293,
+      "grad_norm": 0.015221865847706795,
+      "learning_rate": 8.584802666797188e-05,
+      "loss": 0.4648,
+      "step": 418
+    },
+    {
+      "epoch": 0.26893453145057766,
+      "grad_norm": 0.016463516280055046,
+      "learning_rate": 8.577547910274941e-05,
+      "loss": 0.4824,
+      "step": 419
+    },
+    {
+      "epoch": 0.26957637997432604,
+      "grad_norm": 0.01197727955877781,
+      "learning_rate": 8.570277688532112e-05,
+      "loss": 0.4648,
+      "step": 420
+    },
+    {
+      "epoch": 0.2702182284980745,
+      "grad_norm": 0.016626371070742607,
+      "learning_rate": 8.562992032996815e-05,
+      "loss": 0.4824,
+      "step": 421
+    },
+    {
+      "epoch": 0.27086007702182285,
+      "grad_norm": 0.012758683413267136,
+      "learning_rate": 8.555690975163883e-05,
+      "loss": 0.4688,
+      "step": 422
+    },
+    {
+      "epoch": 0.27150192554557123,
+      "grad_norm": 0.01449665054678917,
+      "learning_rate": 8.548374546594727e-05,
+      "loss": 0.4648,
+      "step": 423
+    },
+    {
+      "epoch": 0.27214377406931967,
+      "grad_norm": 0.012013441883027554,
+      "learning_rate": 8.541042778917207e-05,
+      "loss": 0.4805,
+      "step": 424
+    },
+    {
+      "epoch": 0.27278562259306804,
+      "grad_norm": 0.01310022547841072,
+      "learning_rate": 8.533695703825493e-05,
+      "loss": 0.4551,
+      "step": 425
+    },
+    {
+      "epoch": 0.2734274711168164,
+      "grad_norm": 0.01266622543334961,
+      "learning_rate": 8.526333353079922e-05,
+      "loss": 0.416,
+      "step": 426
+    },
+    {
+      "epoch": 0.2740693196405648,
+      "grad_norm": 0.013401447795331478,
+      "learning_rate": 8.518955758506871e-05,
+      "loss": 0.4531,
+      "step": 427
+    },
+    {
+      "epoch": 0.27471116816431324,
+      "grad_norm": 0.013581424951553345,
+      "learning_rate": 8.511562951998607e-05,
+      "loss": 0.4688,
+      "step": 428
+    },
+    {
+      "epoch": 0.2753530166880616,
+      "grad_norm": 0.013305437751114368,
+      "learning_rate": 8.504154965513163e-05,
+      "loss": 0.5,
+      "step": 429
+    },
+    {
+      "epoch": 0.27599486521181,
+      "grad_norm": 0.012767552398145199,
+      "learning_rate": 8.496731831074189e-05,
+      "loss": 0.4805,
+      "step": 430
+    },
+    {
+      "epoch": 0.2766367137355584,
+      "grad_norm": 0.014782493934035301,
+      "learning_rate": 8.48929358077082e-05,
+      "loss": 0.4414,
+      "step": 431
+    },
+    {
+      "epoch": 0.2772785622593068,
+      "grad_norm": 0.013910866342484951,
+      "learning_rate": 8.481840246757531e-05,
+      "loss": 0.4414,
+      "step": 432
+    },
+    {
+      "epoch": 0.2779204107830552,
+      "grad_norm": 0.012829374521970749,
+      "learning_rate": 8.474371861254002e-05,
+      "loss": 0.4746,
+      "step": 433
+    },
+    {
+      "epoch": 0.2785622593068036,
+      "grad_norm": 0.011781183071434498,
+      "learning_rate": 8.466888456544983e-05,
+      "loss": 0.457,
+      "step": 434
+    },
+    {
+      "epoch": 0.279204107830552,
+      "grad_norm": 0.011600065045058727,
+      "learning_rate": 8.459390064980146e-05,
+      "loss": 0.4531,
+      "step": 435
+    },
+    {
+      "epoch": 0.2798459563543004,
+      "grad_norm": 0.014579742215573788,
+      "learning_rate": 8.45187671897395e-05,
+      "loss": 0.4922,
+      "step": 436
+    },
+    {
+      "epoch": 0.2804878048780488,
+      "grad_norm": 0.01168655976653099,
+      "learning_rate": 8.444348451005499e-05,
+      "loss": 0.4316,
+      "step": 437
+    },
+    {
+      "epoch": 0.2811296534017972,
+      "grad_norm": 0.013155391439795494,
+      "learning_rate": 8.436805293618404e-05,
+      "loss": 0.4688,
+      "step": 438
+    },
+    {
+      "epoch": 0.28177150192554556,
+      "grad_norm": 0.012940961867570877,
+      "learning_rate": 8.429247279420637e-05,
+      "loss": 0.4492,
+      "step": 439
+    },
+    {
+      "epoch": 0.28241335044929394,
+      "grad_norm": 0.01248475257307291,
+      "learning_rate": 8.421674441084404e-05,
+      "loss": 0.457,
+      "step": 440
+    },
+    {
+      "epoch": 0.2830551989730424,
+      "grad_norm": 0.014048158191144466,
+      "learning_rate": 8.414086811345978e-05,
+      "loss": 0.4863,
+      "step": 441
+    },
+    {
+      "epoch": 0.28369704749679076,
+      "grad_norm": 0.012956113554537296,
+      "learning_rate": 8.406484423005587e-05,
+      "loss": 0.4688,
+      "step": 442
+    },
+    {
+      "epoch": 0.28433889602053913,
+      "grad_norm": 0.01437971368432045,
+      "learning_rate": 8.398867308927252e-05,
+      "loss": 0.4609,
+      "step": 443
+    },
+    {
+      "epoch": 0.28498074454428757,
+      "grad_norm": 0.01513160765171051,
+      "learning_rate": 8.391235502038651e-05,
+      "loss": 0.4727,
+      "step": 444
+    },
+    {
+      "epoch": 0.28562259306803595,
+      "grad_norm": 0.012992394156754017,
+      "learning_rate": 8.383589035330977e-05,
+      "loss": 0.4766,
+      "step": 445
+    },
+    {
+      "epoch": 0.2862644415917843,
+      "grad_norm": 0.012563800439238548,
+      "learning_rate": 8.375927941858797e-05,
+      "loss": 0.4492,
+      "step": 446
+    },
+    {
+      "epoch": 0.28690629011553276,
+      "grad_norm": 0.014958107843995094,
+      "learning_rate": 8.368252254739908e-05,
+      "loss": 0.4688,
+      "step": 447
+    },
+    {
+      "epoch": 0.28754813863928114,
+      "grad_norm": 0.01127879973500967,
+      "learning_rate": 8.360562007155192e-05,
+      "loss": 0.418,
+      "step": 448
+    },
+    {
+      "epoch": 0.2881899871630295,
+      "grad_norm": 0.012075050733983517,
+      "learning_rate": 8.352857232348472e-05,
+      "loss": 0.4297,
+      "step": 449
+    },
+    {
+      "epoch": 0.2888318356867779,
+      "grad_norm": 0.01276320219039917,
+      "learning_rate": 8.345137963626372e-05,
+      "loss": 0.457,
+      "step": 450
+    },
+    {
+      "epoch": 0.2894736842105263,
+      "grad_norm": 0.012094990350306034,
+      "learning_rate": 8.337404234358172e-05,
+      "loss": 0.4609,
+      "step": 451
+    },
+    {
+      "epoch": 0.2901155327342747,
+      "grad_norm": 0.012510258704423904,
+      "learning_rate": 8.329656077975658e-05,
+      "loss": 0.4531,
+      "step": 452
+    },
+    {
+      "epoch": 0.2907573812580231,
+      "grad_norm": 0.013942587189376354,
+      "learning_rate": 8.32189352797299e-05,
+      "loss": 0.4688,
+      "step": 453
+    },
+    {
+      "epoch": 0.2913992297817715,
+      "grad_norm": 0.011335965245962143,
+      "learning_rate": 8.314116617906544e-05,
+      "loss": 0.4355,
+      "step": 454
+    },
+    {
+      "epoch": 0.2920410783055199,
+      "grad_norm": 0.013128997758030891,
+      "learning_rate": 8.306325381394774e-05,
+      "loss": 0.4473,
+      "step": 455
+    },
+    {
+      "epoch": 0.2926829268292683,
+      "grad_norm": 0.013161028735339642,
+      "learning_rate": 8.298519852118065e-05,
+      "loss": 0.4609,
+      "step": 456
+    },
+    {
+      "epoch": 0.2933247753530167,
+      "grad_norm": 0.012427788227796555,
+      "learning_rate": 8.290700063818589e-05,
+      "loss": 0.4434,
+      "step": 457
+    },
+    {
+      "epoch": 0.2939666238767651,
+      "grad_norm": 0.014244086109101772,
+      "learning_rate": 8.282866050300152e-05,
+      "loss": 0.4766,
+      "step": 458
+    },
+    {
+      "epoch": 0.29460847240051347,
+      "grad_norm": 0.012252756394445896,
+      "learning_rate": 8.275017845428063e-05,
+      "loss": 0.4766,
+      "step": 459
+    },
+    {
+      "epoch": 0.2952503209242619,
+      "grad_norm": 0.012408980168402195,
+      "learning_rate": 8.26715548312897e-05,
+      "loss": 0.4746,
+      "step": 460
+    },
+    {
+      "epoch": 0.2958921694480103,
+      "grad_norm": 0.014483646489679813,
+      "learning_rate": 8.259278997390725e-05,
+      "loss": 0.4316,
+      "step": 461
+    },
+    {
+      "epoch": 0.29653401797175866,
+      "grad_norm": 0.011408063583076,
+      "learning_rate": 8.251388422262234e-05,
+      "loss": 0.4453,
+      "step": 462
+    },
+    {
+      "epoch": 0.29717586649550704,
+      "grad_norm": 0.016647979617118835,
+      "learning_rate": 8.243483791853308e-05,
+      "loss": 0.4688,
+      "step": 463
+    },
+    {
+      "epoch": 0.29781771501925547,
+      "grad_norm": 0.012057177722454071,
+      "learning_rate": 8.235565140334518e-05,
+      "loss": 0.4512,
+      "step": 464
+    },
+    {
+      "epoch": 0.29845956354300385,
+      "grad_norm": 0.01220938190817833,
+      "learning_rate": 8.227632501937045e-05,
+      "loss": 0.4473,
+      "step": 465
+    },
+    {
+      "epoch": 0.2991014120667522,
+      "grad_norm": 0.012876022607088089,
+      "learning_rate": 8.219685910952532e-05,
+      "loss": 0.5078,
+      "step": 466
+    },
+    {
+      "epoch": 0.29974326059050066,
+      "grad_norm": 0.011478842236101627,
+      "learning_rate": 8.211725401732944e-05,
+      "loss": 0.4531,
+      "step": 467
+    },
+    {
+      "epoch": 0.30038510911424904,
+      "grad_norm": 0.012432406656444073,
+      "learning_rate": 8.203751008690403e-05,
+      "loss": 0.4297,
+      "step": 468
+    },
+    {
+      "epoch": 0.3010269576379974,
+      "grad_norm": 0.011501291766762733,
+      "learning_rate": 8.195762766297055e-05,
+      "loss": 0.4297,
+      "step": 469
+    },
+    {
+      "epoch": 0.30166880616174585,
+      "grad_norm": 0.013943162746727467,
+      "learning_rate": 8.187760709084911e-05,
+      "loss": 0.4707,
+      "step": 470
+    },
+    {
+      "epoch": 0.30231065468549423,
+      "grad_norm": 0.011519819498062134,
+      "learning_rate": 8.179744871645707e-05,
+      "loss": 0.4453,
+      "step": 471
+    },
+    {
+      "epoch": 0.3029525032092426,
+      "grad_norm": 0.011311845853924751,
+      "learning_rate": 8.171715288630742e-05,
+      "loss": 0.4414,
+      "step": 472
+    },
+    {
+      "epoch": 0.30359435173299104,
+      "grad_norm": 0.01158730685710907,
+      "learning_rate": 8.16367199475074e-05,
+      "loss": 0.4609,
+      "step": 473
+    },
+    {
+      "epoch": 0.3042362002567394,
+      "grad_norm": 0.013310915790498257,
+      "learning_rate": 8.155615024775693e-05,
+      "loss": 0.4473,
+      "step": 474
+    },
+    {
+      "epoch": 0.3048780487804878,
+      "grad_norm": 0.014084615744650364,
+      "learning_rate": 8.147544413534714e-05,
+      "loss": 0.4375,
+      "step": 475
+    },
+    {
+      "epoch": 0.3055198973042362,
+      "grad_norm": 0.011800910346210003,
+      "learning_rate": 8.139460195915883e-05,
+      "loss": 0.4453,
+      "step": 476
+    },
+    {
+      "epoch": 0.3061617458279846,
+      "grad_norm": 0.012746773660182953,
+      "learning_rate": 8.1313624068661e-05,
+      "loss": 0.4473,
+      "step": 477
+    },
+    {
+      "epoch": 0.306803594351733,
+      "grad_norm": 0.0162535198032856,
+      "learning_rate": 8.123251081390937e-05,
+      "loss": 0.4805,
+      "step": 478
+    },
+    {
+      "epoch": 0.30744544287548137,
+      "grad_norm": 0.011776765808463097,
+      "learning_rate": 8.115126254554471e-05,
+      "loss": 0.4336,
+      "step": 479
+    },
+    {
+      "epoch": 0.3080872913992298,
+      "grad_norm": 0.015100241638720036,
+      "learning_rate": 8.10698796147915e-05,
+      "loss": 0.4453,
+      "step": 480
+    },
+    {
+      "epoch": 0.3087291399229782,
+      "grad_norm": 0.014546082355082035,
+      "learning_rate": 8.09883623734564e-05,
+      "loss": 0.418,
+      "step": 481
+    },
+    {
+      "epoch": 0.30937098844672656,
+      "grad_norm": 0.01146666705608368,
+      "learning_rate": 8.090671117392655e-05,
+      "loss": 0.4336,
+      "step": 482
+    },
+    {
+      "epoch": 0.310012836970475,
+      "grad_norm": 0.012188490480184555,
+      "learning_rate": 8.082492636916828e-05,
+      "loss": 0.4551,
+      "step": 483
+    },
+    {
+      "epoch": 0.31065468549422337,
+      "grad_norm": 0.012891625985503197,
+      "learning_rate": 8.074300831272542e-05,
+      "loss": 0.4316,
+      "step": 484
+    },
+    {
+      "epoch": 0.31129653401797175,
+      "grad_norm": 0.012159287929534912,
+      "learning_rate": 8.066095735871786e-05,
+      "loss": 0.4414,
+      "step": 485
+    },
+    {
+      "epoch": 0.3119383825417201,
+      "grad_norm": 0.014158441685140133,
+      "learning_rate": 8.057877386183995e-05,
+      "loss": 0.457,
+      "step": 486
+    },
+    {
+      "epoch": 0.31258023106546856,
+      "grad_norm": 0.014469098299741745,
+      "learning_rate": 8.049645817735903e-05,
+      "loss": 0.4492,
+      "step": 487
+    },
+    {
+      "epoch": 0.31322207958921694,
+      "grad_norm": 0.012358460575342178,
+      "learning_rate": 8.041401066111387e-05,
+      "loss": 0.4316,
+      "step": 488
+    },
+    {
+      "epoch": 0.3138639281129653,
+      "grad_norm": 0.014756308868527412,
+      "learning_rate": 8.033143166951311e-05,
+      "loss": 0.4297,
+      "step": 489
+    },
+    {
+      "epoch": 0.31450577663671375,
+      "grad_norm": 0.012180181220173836,
+      "learning_rate": 8.024872155953376e-05,
+      "loss": 0.4375,
+      "step": 490
+    },
+    {
+      "epoch": 0.31514762516046213,
+      "grad_norm": 0.014177980832755566,
+      "learning_rate": 8.016588068871961e-05,
+      "loss": 0.4492,
+      "step": 491
+    },
+    {
+      "epoch": 0.3157894736842105,
+      "grad_norm": 0.014273082837462425,
+      "learning_rate": 8.008290941517976e-05,
+      "loss": 0.4336,
+      "step": 492
+    },
+    {
+      "epoch": 0.31643132220795894,
+      "grad_norm": 0.012356355786323547,
+      "learning_rate": 7.999980809758694e-05,
+      "loss": 0.4375,
+      "step": 493
+    },
+    {
+      "epoch": 0.3170731707317073,
+      "grad_norm": 0.013944054022431374,
+      "learning_rate": 7.991657709517613e-05,
+      "loss": 0.4375,
+      "step": 494
+    },
+    {
+      "epoch": 0.3177150192554557,
+      "grad_norm": 0.015265170484781265,
+      "learning_rate": 7.983321676774285e-05,
+      "loss": 0.4609,
+      "step": 495
+    },
+    {
+      "epoch": 0.31835686777920413,
+      "grad_norm": 0.011430501006543636,
+      "learning_rate": 7.974972747564174e-05,
+      "loss": 0.4355,
+      "step": 496
+    },
+    {
+      "epoch": 0.3189987163029525,
+      "grad_norm": 0.015532695688307285,
+      "learning_rate": 7.966610957978483e-05,
+      "loss": 0.4922,
+      "step": 497
+    },
+    {
+      "epoch": 0.3196405648267009,
+      "grad_norm": 0.014049970544874668,
+      "learning_rate": 7.958236344164021e-05,
+      "loss": 0.4473,
+      "step": 498
+    },
+    {
+      "epoch": 0.32028241335044927,
+      "grad_norm": 0.011538274586200714,
+      "learning_rate": 7.949848942323029e-05,
+      "loss": 0.4375,
+      "step": 499
+    },
+    {
+      "epoch": 0.3209242618741977,
+      "grad_norm": 0.01410175021737814,
+      "learning_rate": 7.941448788713024e-05,
+      "loss": 0.4512,
+      "step": 500
+    },
+    {
+      "epoch": 0.3215661103979461,
+      "grad_norm": 0.017135094851255417,
+      "learning_rate": 7.933035919646654e-05,
+      "loss": 0.4453,
+      "step": 501
+    },
+    {
+      "epoch": 0.32220795892169446,
+      "grad_norm": 0.013886646367609501,
+      "learning_rate": 7.92461037149153e-05,
+      "loss": 0.457,
+      "step": 502
+    },
+    {
+      "epoch": 0.3228498074454429,
+      "grad_norm": 0.02123275399208069,
+      "learning_rate": 7.916172180670076e-05,
+      "loss": 0.4805,
+      "step": 503
+    },
+    {
+      "epoch": 0.32349165596919127,
+      "grad_norm": 0.012984886765480042,
+      "learning_rate": 7.907721383659368e-05,
+      "loss": 0.4453,
+      "step": 504
+    },
+    {
+      "epoch": 0.32413350449293965,
+      "grad_norm": 0.014154748991131783,
+      "learning_rate": 7.899258016990969e-05,
+      "loss": 0.4453,
+      "step": 505
+    },
+    {
+      "epoch": 0.3247753530166881,
+      "grad_norm": 0.01399321760982275,
+      "learning_rate": 7.890782117250793e-05,
+      "loss": 0.4492,
+      "step": 506
+    },
+    {
+      "epoch": 0.32541720154043646,
+      "grad_norm": 0.012078939937055111,
+      "learning_rate": 7.882293721078921e-05,
+      "loss": 0.4238,
+      "step": 507
+    },
+    {
+      "epoch": 0.32605905006418484,
+      "grad_norm": 0.013892695307731628,
+      "learning_rate": 7.873792865169458e-05,
+      "loss": 0.4766,
+      "step": 508
+    },
+    {
+      "epoch": 0.3267008985879333,
+      "grad_norm": 0.012964858673512936,
+      "learning_rate": 7.865279586270371e-05,
+      "loss": 0.4629,
+      "step": 509
+    },
+    {
+      "epoch": 0.32734274711168165,
+      "grad_norm": 0.012767773121595383,
+      "learning_rate": 7.856753921183331e-05,
+      "loss": 0.4727,
+      "step": 510
+    },
+    {
+      "epoch": 0.32798459563543003,
+      "grad_norm": 0.015065948478877544,
+      "learning_rate": 7.84821590676355e-05,
+      "loss": 0.4336,
+      "step": 511
+    },
+    {
+      "epoch": 0.3286264441591784,
+      "grad_norm": 0.01637822948396206,
+      "learning_rate": 7.839665579919626e-05,
+      "loss": 0.4277,
+      "step": 512
+    },
+    {
+      "epoch": 0.32926829268292684,
+      "grad_norm": 0.015638669952750206,
+      "learning_rate": 7.83110297761338e-05,
+      "loss": 0.4199,
+      "step": 513
+    },
+    {
+      "epoch": 0.3299101412066752,
+      "grad_norm": 0.014630009420216084,
+      "learning_rate": 7.822528136859702e-05,
+      "loss": 0.418,
+      "step": 514
+    },
+    {
+      "epoch": 0.3305519897304236,
+      "grad_norm": 0.017467154189944267,
+      "learning_rate": 7.813941094726384e-05,
+      "loss": 0.4707,
+      "step": 515
+    },
+    {
+      "epoch": 0.33119383825417203,
+      "grad_norm": 0.012475800700485706,
+      "learning_rate": 7.805341888333961e-05,
+      "loss": 0.4297,
+      "step": 516
+    },
+    {
+      "epoch": 0.3318356867779204,
+      "grad_norm": 0.011913339607417583,
+      "learning_rate": 7.796730554855558e-05,
+      "loss": 0.4414,
+      "step": 517
+    },
+    {
+      "epoch": 0.3324775353016688,
+      "grad_norm": 0.016746627166867256,
+      "learning_rate": 7.788107131516717e-05,
+      "loss": 0.4219,
+      "step": 518
+    },
+    {
+      "epoch": 0.3331193838254172,
+      "grad_norm": 0.015730712562799454,
+      "learning_rate": 7.779471655595248e-05,
+      "loss": 0.4414,
+      "step": 519
+    },
+    {
+      "epoch": 0.3337612323491656,
+      "grad_norm": 0.013719049282371998,
+      "learning_rate": 7.770824164421062e-05,
+      "loss": 0.4609,
+      "step": 520
+    },
+    {
+      "epoch": 0.334403080872914,
+      "grad_norm": 0.01619836688041687,
+      "learning_rate": 7.762164695376005e-05,
+      "loss": 0.459,
+      "step": 521
+    },
+    {
+      "epoch": 0.33504492939666236,
+      "grad_norm": 0.015834450721740723,
+      "learning_rate": 7.753493285893708e-05,
+      "loss": 0.4219,
+      "step": 522
+    },
+    {
+      "epoch": 0.3356867779204108,
+      "grad_norm": 0.012220130302011967,
+      "learning_rate": 7.744809973459415e-05,
+      "loss": 0.4238,
+      "step": 523
+    },
+    {
+      "epoch": 0.3363286264441592,
+      "grad_norm": 0.014210267923772335,
+      "learning_rate": 7.736114795609828e-05,
+      "loss": 0.4297,
+      "step": 524
+    },
+    {
+      "epoch": 0.33697047496790755,
+      "grad_norm": 0.012427953071892262,
+      "learning_rate": 7.727407789932935e-05,
+      "loss": 0.4531,
+      "step": 525
+    },
+    {
+      "epoch": 0.337612323491656,
+      "grad_norm": 0.014018227346241474,
+      "learning_rate": 7.718688994067863e-05,
+      "loss": 0.4258,
+      "step": 526
+    },
+    {
+      "epoch": 0.33825417201540436,
+      "grad_norm": 0.011482173576951027,
+      "learning_rate": 7.709958445704698e-05,
+      "loss": 0.4219,
+      "step": 527
+    },
+    {
+      "epoch": 0.33889602053915274,
+      "grad_norm": 0.01443139836192131,
+      "learning_rate": 7.701216182584336e-05,
+      "loss": 0.4375,
+      "step": 528
+    },
+    {
+      "epoch": 0.3395378690629012,
+      "grad_norm": 0.012723613530397415,
+      "learning_rate": 7.692462242498313e-05,
+      "loss": 0.4512,
+      "step": 529
+    },
+    {
+      "epoch": 0.34017971758664955,
+      "grad_norm": 0.013509729877114296,
+      "learning_rate": 7.68369666328864e-05,
+      "loss": 0.4492,
+      "step": 530
+    },
+    {
+      "epoch": 0.34082156611039793,
+      "grad_norm": 0.012013684958219528,
+      "learning_rate": 7.674919482847645e-05,
+      "loss": 0.4336,
+      "step": 531
+    },
+    {
+      "epoch": 0.34146341463414637,
+      "grad_norm": 0.012167776003479958,
+      "learning_rate": 7.666130739117805e-05,
+      "loss": 0.418,
+      "step": 532
+    },
+    {
+      "epoch": 0.34210526315789475,
+      "grad_norm": 0.013328084722161293,
+      "learning_rate": 7.657330470091584e-05,
+      "loss": 0.4258,
+      "step": 533
+    },
+    {
+      "epoch": 0.3427471116816431,
+      "grad_norm": 0.011257752776145935,
+      "learning_rate": 7.648518713811267e-05,
+      "loss": 0.4062,
+      "step": 534
+    },
+    {
+      "epoch": 0.3433889602053915,
+      "grad_norm": 0.013994047418236732,
+      "learning_rate": 7.639695508368803e-05,
+      "loss": 0.4258,
+      "step": 535
+    },
+    {
+      "epoch": 0.34403080872913994,
+      "grad_norm": 0.014302888885140419,
+      "learning_rate": 7.630860891905625e-05,
+      "loss": 0.4492,
+      "step": 536
+    },
+    {
+      "epoch": 0.3446726572528883,
+      "grad_norm": 0.012811757624149323,
+      "learning_rate": 7.6220149026125e-05,
+      "loss": 0.4414,
+      "step": 537
+    },
+    {
+      "epoch": 0.3453145057766367,
+      "grad_norm": 0.011713909916579723,
+      "learning_rate": 7.613157578729353e-05,
+      "loss": 0.4219,
+      "step": 538
+    },
+    {
+      "epoch": 0.3459563543003851,
+      "grad_norm": 0.01381740439683199,
+      "learning_rate": 7.604288958545113e-05,
+      "loss": 0.4199,
+      "step": 539
+    },
+    {
+      "epoch": 0.3465982028241335,
+      "grad_norm": 0.011871659196913242,
+      "learning_rate": 7.59540908039754e-05,
+      "loss": 0.375,
+      "step": 540
+    },
+    {
+      "epoch": 0.3472400513478819,
+      "grad_norm": 0.012422285042703152,
+      "learning_rate": 7.586517982673054e-05,
+      "loss": 0.4453,
+      "step": 541
+    },
+    {
+      "epoch": 0.3478818998716303,
+      "grad_norm": 0.014537949115037918,
+      "learning_rate": 7.577615703806586e-05,
+      "loss": 0.4316,
+      "step": 542
+    },
+    {
+      "epoch": 0.3485237483953787,
+      "grad_norm": 0.013570103794336319,
+      "learning_rate": 7.568702282281392e-05,
+      "loss": 0.4336,
+      "step": 543
+    },
+    {
+      "epoch": 0.3491655969191271,
+      "grad_norm": 0.011618518270552158,
+      "learning_rate": 7.559777756628901e-05,
+      "loss": 0.4258,
+      "step": 544
+    },
+    {
+      "epoch": 0.3498074454428755,
+      "grad_norm": 0.014149384573101997,
+      "learning_rate": 7.550842165428543e-05,
+      "loss": 0.4297,
+      "step": 545
+    },
+    {
+      "epoch": 0.3504492939666239,
+      "grad_norm": 0.012584755197167397,
+      "learning_rate": 7.541895547307584e-05,
+      "loss": 0.4414,
+      "step": 546
+    },
+    {
+      "epoch": 0.35109114249037227,
+      "grad_norm": 0.01141996867954731,
+      "learning_rate": 7.532937940940953e-05,
+      "loss": 0.4102,
+      "step": 547
+    },
+    {
+      "epoch": 0.35173299101412064,
+      "grad_norm": 0.012579156085848808,
+      "learning_rate": 7.523969385051084e-05,
+      "loss": 0.4199,
+      "step": 548
+    },
+    {
+      "epoch": 0.3523748395378691,
+      "grad_norm": 0.015211129561066628,
+      "learning_rate": 7.514989918407744e-05,
+      "loss": 0.3906,
+      "step": 549
+    },
+    {
+      "epoch": 0.35301668806161746,
+      "grad_norm": 0.013005176559090614,
+      "learning_rate": 7.505999579827863e-05,
+      "loss": 0.4297,
+      "step": 550
+    },
+    {
+      "epoch": 0.35365853658536583,
+      "grad_norm": 0.015413876622915268,
+      "learning_rate": 7.496998408175373e-05,
+      "loss": 0.4531,
+      "step": 551
+    },
+    {
+      "epoch": 0.35430038510911427,
+      "grad_norm": 0.014693383127450943,
+      "learning_rate": 7.48798644236103e-05,
+      "loss": 0.4707,
+      "step": 552
+    },
+    {
+      "epoch": 0.35494223363286265,
+      "grad_norm": 0.012898659333586693,
+      "learning_rate": 7.478963721342256e-05,
+      "loss": 0.416,
+      "step": 553
+    },
+    {
+      "epoch": 0.355584082156611,
+      "grad_norm": 0.011905779130756855,
+      "learning_rate": 7.469930284122966e-05,
+      "loss": 0.4141,
+      "step": 554
+    },
+    {
+      "epoch": 0.35622593068035946,
+      "grad_norm": 0.01483968086540699,
+      "learning_rate": 7.460886169753397e-05,
+      "loss": 0.4141,
+      "step": 555
+    },
+    {
+      "epoch": 0.35686777920410784,
+      "grad_norm": 0.014148566871881485,
+      "learning_rate": 7.451831417329943e-05,
+      "loss": 0.4531,
+      "step": 556
+    },
+    {
+      "epoch": 0.3575096277278562,
+      "grad_norm": 0.013587851077318192,
+      "learning_rate": 7.442766065994985e-05,
+      "loss": 0.4316,
+      "step": 557
+    },
+    {
+      "epoch": 0.3581514762516046,
+      "grad_norm": 0.012541595846414566,
+      "learning_rate": 7.433690154936724e-05,
+      "loss": 0.4141,
+      "step": 558
+    },
+    {
+      "epoch": 0.35879332477535303,
+      "grad_norm": 0.012243524193763733,
+      "learning_rate": 7.424603723389006e-05,
+      "loss": 0.4277,
+      "step": 559
+    },
+    {
+      "epoch": 0.3594351732991014,
+      "grad_norm": 0.014602742157876492,
+      "learning_rate": 7.415506810631155e-05,
+      "loss": 0.4102,
+      "step": 560
+    },
+    {
+      "epoch": 0.3600770218228498,
+      "grad_norm": 0.015661180019378662,
+      "learning_rate": 7.406399455987804e-05,
+      "loss": 0.4336,
+      "step": 561
+    },
+    {
+      "epoch": 0.3607188703465982,
+      "grad_norm": 0.014495295472443104,
+      "learning_rate": 7.39728169882873e-05,
+      "loss": 0.4492,
+      "step": 562
+    },
+    {
+      "epoch": 0.3613607188703466,
+      "grad_norm": 0.013921634294092655,
+      "learning_rate": 7.388153578568671e-05,
+      "loss": 0.4531,
+      "step": 563
+    },
+    {
+      "epoch": 0.362002567394095,
+      "grad_norm": 0.014414452016353607,
+      "learning_rate": 7.379015134667167e-05,
+      "loss": 0.4297,
+      "step": 564
+    },
+    {
+      "epoch": 0.3626444159178434,
+      "grad_norm": 0.01415167935192585,
+      "learning_rate": 7.369866406628385e-05,
+      "loss": 0.4297,
+      "step": 565
+    },
+    {
+      "epoch": 0.3632862644415918,
+      "grad_norm": 0.017361560836434364,
+      "learning_rate": 7.36070743400095e-05,
+      "loss": 0.4062,
+      "step": 566
+    },
+    {
+      "epoch": 0.36392811296534017,
+      "grad_norm": 0.01368169579654932,
+      "learning_rate": 7.351538256377771e-05,
+      "loss": 0.4297,
+      "step": 567
+    },
+    {
+      "epoch": 0.3645699614890886,
+      "grad_norm": 0.013606247492134571,
+      "learning_rate": 7.342358913395874e-05,
+      "loss": 0.4258,
+      "step": 568
+    },
+    {
+      "epoch": 0.365211810012837,
+      "grad_norm": 0.014146490022540092,
+      "learning_rate": 7.333169444736225e-05,
+      "loss": 0.4531,
+      "step": 569
+    },
+    {
+      "epoch": 0.36585365853658536,
+      "grad_norm": 0.013685311190783978,
+      "learning_rate": 7.323969890123565e-05,
+      "loss": 0.4297,
+      "step": 570
+    },
+    {
+      "epoch": 0.36649550706033374,
+      "grad_norm": 0.012898226268589497,
+      "learning_rate": 7.314760289326236e-05,
+      "loss": 0.418,
+      "step": 571
+    },
+    {
+      "epoch": 0.36713735558408217,
+      "grad_norm": 0.014033549465239048,
+      "learning_rate": 7.305540682156e-05,
+      "loss": 0.4414,
+      "step": 572
+    },
+    {
+      "epoch": 0.36777920410783055,
+      "grad_norm": 0.012177499942481518,
+      "learning_rate": 7.296311108467888e-05,
+      "loss": 0.4102,
+      "step": 573
+    },
+    {
+      "epoch": 0.3684210526315789,
+      "grad_norm": 0.019322926178574562,
+      "learning_rate": 7.287071608160003e-05,
+      "loss": 0.4453,
+      "step": 574
+    },
+    {
+      "epoch": 0.36906290115532736,
+      "grad_norm": 0.014528055675327778,
+      "learning_rate": 7.277822221173367e-05,
+      "loss": 0.4434,
+      "step": 575
+    },
+    {
+      "epoch": 0.36970474967907574,
+      "grad_norm": 0.014417881146073341,
+      "learning_rate": 7.268562987491739e-05,
+      "loss": 0.4609,
+      "step": 576
+    },
+    {
+      "epoch": 0.3703465982028241,
+      "grad_norm": 0.013631274923682213,
+      "learning_rate": 7.259293947141441e-05,
+      "loss": 0.4375,
+      "step": 577
+    },
+    {
+      "epoch": 0.37098844672657255,
+      "grad_norm": 0.016857121139764786,
+      "learning_rate": 7.250015140191188e-05,
+      "loss": 0.4531,
+      "step": 578
+    },
+    {
+      "epoch": 0.37163029525032093,
+      "grad_norm": 0.014084297232329845,
+      "learning_rate": 7.24072660675192e-05,
+      "loss": 0.4219,
+      "step": 579
+    },
+    {
+      "epoch": 0.3722721437740693,
+      "grad_norm": 0.014484019950032234,
+      "learning_rate": 7.231428386976618e-05,
+      "loss": 0.4551,
+      "step": 580
+    },
+    {
+      "epoch": 0.37291399229781774,
+      "grad_norm": 0.016134057193994522,
+      "learning_rate": 7.222120521060134e-05,
+      "loss": 0.4531,
+      "step": 581
+    },
+    {
+      "epoch": 0.3735558408215661,
+      "grad_norm": 0.015156952664256096,
+      "learning_rate": 7.212803049239028e-05,
+      "loss": 0.4277,
+      "step": 582
+    },
+    {
+      "epoch": 0.3741976893453145,
+      "grad_norm": 0.014183555729687214,
+      "learning_rate": 7.203476011791373e-05,
+      "loss": 0.4102,
+      "step": 583
+    },
+    {
+      "epoch": 0.3748395378690629,
+      "grad_norm": 0.016075201332569122,
+      "learning_rate": 7.194139449036603e-05,
+      "loss": 0.4375,
+      "step": 584
+    },
+    {
+      "epoch": 0.3754813863928113,
+      "grad_norm": 0.01476068701595068,
+      "learning_rate": 7.184793401335322e-05,
+      "loss": 0.418,
+      "step": 585
+    },
+    {
+      "epoch": 0.3761232349165597,
+      "grad_norm": 0.013279945589601994,
+      "learning_rate": 7.175437909089139e-05,
+      "loss": 0.4395,
+      "step": 586
+    },
+    {
+      "epoch": 0.37676508344030807,
+      "grad_norm": 0.01367487758398056,
+      "learning_rate": 7.166073012740491e-05,
+      "loss": 0.4531,
+      "step": 587
+    },
+    {
+      "epoch": 0.3774069319640565,
+      "grad_norm": 0.012248165905475616,
+      "learning_rate": 7.156698752772463e-05,
+      "loss": 0.4102,
+      "step": 588
+    },
+    {
+      "epoch": 0.3780487804878049,
+      "grad_norm": 0.01692807860672474,
+      "learning_rate": 7.147315169708622e-05,
+      "loss": 0.4336,
+      "step": 589
+    },
+    {
+      "epoch": 0.37869062901155326,
+      "grad_norm": 0.011725704185664654,
+      "learning_rate": 7.137922304112838e-05,
+      "loss": 0.4062,
+      "step": 590
+    },
+    {
+      "epoch": 0.3793324775353017,
+      "grad_norm": 0.015160840004682541,
+      "learning_rate": 7.128520196589105e-05,
+      "loss": 0.4062,
+      "step": 591
+    },
+    {
+      "epoch": 0.37997432605905007,
+      "grad_norm": 0.01488816924393177,
+      "learning_rate": 7.119108887781371e-05,
+      "loss": 0.4375,
+      "step": 592
+    },
+    {
+      "epoch": 0.38061617458279845,
+      "grad_norm": 0.014125654473900795,
+      "learning_rate": 7.109688418373355e-05,
+      "loss": 0.4453,
+      "step": 593
+    },
+    {
+      "epoch": 0.38125802310654683,
+      "grad_norm": 0.015549080446362495,
+      "learning_rate": 7.100258829088384e-05,
+      "loss": 0.4492,
+      "step": 594
+    },
+    {
+      "epoch": 0.38189987163029526,
+      "grad_norm": 0.012530624866485596,
+      "learning_rate": 7.090820160689201e-05,
+      "loss": 0.418,
+      "step": 595
+    },
+    {
+      "epoch": 0.38254172015404364,
+      "grad_norm": 0.011856675148010254,
+      "learning_rate": 7.081372453977803e-05,
+      "loss": 0.4238,
+      "step": 596
+    },
+    {
+      "epoch": 0.383183568677792,
+      "grad_norm": 0.013924370519816875,
+      "learning_rate": 7.071915749795253e-05,
+      "loss": 0.4258,
+      "step": 597
+    },
+    {
+      "epoch": 0.38382541720154045,
+      "grad_norm": 0.012145726941525936,
+      "learning_rate": 7.062450089021511e-05,
+      "loss": 0.4258,
+      "step": 598
+    },
+    {
+      "epoch": 0.38446726572528883,
+      "grad_norm": 0.01391034759581089,
+      "learning_rate": 7.052975512575258e-05,
+      "loss": 0.4219,
+      "step": 599
+    },
+    {
+      "epoch": 0.3851091142490372,
+      "grad_norm": 0.012797384522855282,
+      "learning_rate": 7.04349206141371e-05,
+      "loss": 0.4355,
+      "step": 600
+    },
+    {
+      "epoch": 0.38575096277278564,
+      "grad_norm": 0.0131190475076437,
+      "learning_rate": 7.033999776532454e-05,
+      "loss": 0.4336,
+      "step": 601
+    },
+    {
+      "epoch": 0.386392811296534,
+      "grad_norm": 0.012565034441649914,
+      "learning_rate": 7.024498698965258e-05,
+      "loss": 0.4395,
+      "step": 602
+    },
+    {
+      "epoch": 0.3870346598202824,
+      "grad_norm": 0.014118792489171028,
+      "learning_rate": 7.014988869783901e-05,
+      "loss": 0.4473,
+      "step": 603
+    },
+    {
+      "epoch": 0.38767650834403083,
+      "grad_norm": 0.01384700182825327,
+      "learning_rate": 7.005470330098e-05,
+      "loss": 0.3984,
+      "step": 604
+    },
+    {
+      "epoch": 0.3883183568677792,
+      "grad_norm": 0.0131995165720582,
+      "learning_rate": 6.995943121054816e-05,
+      "loss": 0.4395,
+      "step": 605
+    },
+    {
+      "epoch": 0.3889602053915276,
+      "grad_norm": 0.013334594666957855,
+      "learning_rate": 6.986407283839091e-05,
+      "loss": 0.4336,
+      "step": 606
+    },
+    {
+      "epoch": 0.38960205391527597,
+      "grad_norm": 0.012548287399113178,
+      "learning_rate": 6.976862859672869e-05,
+      "loss": 0.4316,
+      "step": 607
+    },
+    {
+      "epoch": 0.3902439024390244,
+      "grad_norm": 0.01311532687395811,
+      "learning_rate": 6.96730988981531e-05,
+      "loss": 0.4316,
+      "step": 608
+    },
+    {
+      "epoch": 0.3908857509627728,
+      "grad_norm": 0.012342913076281548,
+      "learning_rate": 6.957748415562517e-05,
+      "loss": 0.3867,
+      "step": 609
+    },
+    {
+      "epoch": 0.39152759948652116,
+      "grad_norm": 0.012902142480015755,
+      "learning_rate": 6.948178478247355e-05,
+      "loss": 0.3906,
+      "step": 610
+    },
+    {
+      "epoch": 0.3921694480102696,
+      "grad_norm": 0.011220437474548817,
+      "learning_rate": 6.938600119239273e-05,
+      "loss": 0.416,
+      "step": 611
+    },
+    {
+      "epoch": 0.392811296534018,
+      "grad_norm": 0.012904628179967403,
+      "learning_rate": 6.929013379944132e-05,
+      "loss": 0.4277,
+      "step": 612
+    },
+    {
+      "epoch": 0.39345314505776635,
+      "grad_norm": 0.019070589914917946,
+      "learning_rate": 6.919418301804012e-05,
+      "loss": 0.4238,
+      "step": 613
+    },
+    {
+      "epoch": 0.3940949935815148,
+      "grad_norm": 0.013551855459809303,
+      "learning_rate": 6.909814926297044e-05,
+      "loss": 0.4199,
+      "step": 614
+    },
+    {
+      "epoch": 0.39473684210526316,
+      "grad_norm": 0.011821141466498375,
+      "learning_rate": 6.900203294937229e-05,
+      "loss": 0.3906,
+      "step": 615
+    },
+    {
+      "epoch": 0.39537869062901154,
+      "grad_norm": 0.01720435544848442,
+      "learning_rate": 6.89058344927425e-05,
+      "loss": 0.4375,
+      "step": 616
+    },
+    {
+      "epoch": 0.39602053915276,
+      "grad_norm": 0.013075001537799835,
+      "learning_rate": 6.88095543089331e-05,
+      "loss": 0.4473,
+      "step": 617
+    },
+    {
+      "epoch": 0.39666238767650835,
+      "grad_norm": 0.015591239556670189,
+      "learning_rate": 6.871319281414932e-05,
+      "loss": 0.4219,
+      "step": 618
+    },
+    {
+      "epoch": 0.39730423620025673,
+      "grad_norm": 0.014195159077644348,
+      "learning_rate": 6.861675042494794e-05,
+      "loss": 0.4414,
+      "step": 619
+    },
+    {
+      "epoch": 0.3979460847240051,
+      "grad_norm": 0.013499117456376553,
+      "learning_rate": 6.85202275582354e-05,
+      "loss": 0.3867,
+      "step": 620
+    },
+    {
+      "epoch": 0.39858793324775355,
+      "grad_norm": 0.014024997130036354,
+      "learning_rate": 6.842362463126604e-05,
+      "loss": 0.4199,
+      "step": 621
+    },
+    {
+      "epoch": 0.3992297817715019,
+      "grad_norm": 0.011665104888379574,
+      "learning_rate": 6.832694206164035e-05,
+      "loss": 0.3984,
+      "step": 622
+    },
+    {
+      "epoch": 0.3998716302952503,
+      "grad_norm": 0.01300954632461071,
+      "learning_rate": 6.8230180267303e-05,
+      "loss": 0.4062,
+      "step": 623
+    },
+    {
+      "epoch": 0.40051347881899874,
+      "grad_norm": 0.018235405907034874,
+      "learning_rate": 6.813333966654122e-05,
+      "loss": 0.457,
+      "step": 624
+    },
+    {
+      "epoch": 0.4011553273427471,
+      "grad_norm": 0.013178367167711258,
+      "learning_rate": 6.803642067798284e-05,
+      "loss": 0.4434,
+      "step": 625
+    },
+    {
+      "epoch": 0.4017971758664955,
+      "grad_norm": 0.014165622182190418,
+      "learning_rate": 6.793942372059462e-05,
+      "loss": 0.4102,
+      "step": 626
+    },
+    {
+      "epoch": 0.4024390243902439,
+      "grad_norm": 0.013849026523530483,
+      "learning_rate": 6.784234921368033e-05,
+      "loss": 0.4258,
+      "step": 627
+    },
+    {
+      "epoch": 0.4030808729139923,
+      "grad_norm": 0.014649467542767525,
+      "learning_rate": 6.774519757687897e-05,
+      "loss": 0.418,
+      "step": 628
+    },
+    {
+      "epoch": 0.4037227214377407,
+      "grad_norm": 0.01384900975972414,
+      "learning_rate": 6.764796923016298e-05,
+      "loss": 0.4199,
+      "step": 629
+    },
+    {
+      "epoch": 0.40436456996148906,
+      "grad_norm": 0.011851613409817219,
+      "learning_rate": 6.755066459383637e-05,
+      "loss": 0.4258,
+      "step": 630
+    },
+    {
+      "epoch": 0.4050064184852375,
+      "grad_norm": 0.01206874568015337,
+      "learning_rate": 6.7453284088533e-05,
+      "loss": 0.4121,
+      "step": 631
+    },
+    {
+      "epoch": 0.4056482670089859,
+      "grad_norm": 0.012382757849991322,
+      "learning_rate": 6.735582813521467e-05,
+      "loss": 0.4141,
+      "step": 632
+    },
+    {
+      "epoch": 0.40629011553273425,
+      "grad_norm": 0.01458633504807949,
+      "learning_rate": 6.72582971551693e-05,
+      "loss": 0.418,
+      "step": 633
+    },
+    {
+      "epoch": 0.4069319640564827,
+      "grad_norm": 0.012113570235669613,
+      "learning_rate": 6.716069157000917e-05,
+      "loss": 0.4219,
+      "step": 634
+    },
+    {
+      "epoch": 0.40757381258023107,
+      "grad_norm": 0.011811967939138412,
+      "learning_rate": 6.706301180166909e-05,
+      "loss": 0.4395,
+      "step": 635
+    },
+    {
+      "epoch": 0.40821566110397944,
+      "grad_norm": 0.012068016454577446,
+      "learning_rate": 6.69652582724045e-05,
+      "loss": 0.4062,
+      "step": 636
+    },
+    {
+      "epoch": 0.4088575096277279,
+      "grad_norm": 0.012213320471346378,
+      "learning_rate": 6.686743140478972e-05,
+      "loss": 0.4297,
+      "step": 637
+    },
+    {
+      "epoch": 0.40949935815147626,
+      "grad_norm": 0.012913153506815434,
+      "learning_rate": 6.676953162171613e-05,
+      "loss": 0.3926,
+      "step": 638
+    },
+    {
+      "epoch": 0.41014120667522463,
+      "grad_norm": 0.013523731380701065,
+      "learning_rate": 6.667155934639026e-05,
+      "loss": 0.4141,
+      "step": 639
+    },
+    {
+      "epoch": 0.41078305519897307,
+      "grad_norm": 0.01334290113300085,
+      "learning_rate": 6.65735150023321e-05,
+      "loss": 0.4258,
+      "step": 640
+    },
+    {
+      "epoch": 0.41142490372272145,
+      "grad_norm": 0.011239216662943363,
+      "learning_rate": 6.647539901337307e-05,
+      "loss": 0.4004,
+      "step": 641
+    },
+    {
+      "epoch": 0.4120667522464698,
+      "grad_norm": 0.013552368618547916,
+      "learning_rate": 6.637721180365437e-05,
+      "loss": 0.3984,
+      "step": 642
+    },
+    {
+      "epoch": 0.4127086007702182,
+      "grad_norm": 0.012600836344063282,
+      "learning_rate": 6.627895379762506e-05,
+      "loss": 0.4375,
+      "step": 643
+    },
+    {
+      "epoch": 0.41335044929396664,
+      "grad_norm": 0.012270119972527027,
+      "learning_rate": 6.618062542004024e-05,
+      "loss": 0.4453,
+      "step": 644
+    },
+    {
+      "epoch": 0.413992297817715,
+      "grad_norm": 0.013107932172715664,
+      "learning_rate": 6.608222709595925e-05,
+      "loss": 0.418,
+      "step": 645
+    },
+    {
+      "epoch": 0.4146341463414634,
+      "grad_norm": 0.011738761328160763,
+      "learning_rate": 6.598375925074373e-05,
+      "loss": 0.4297,
+      "step": 646
+    },
+    {
+      "epoch": 0.41527599486521183,
+      "grad_norm": 0.014674089848995209,
+      "learning_rate": 6.588522231005591e-05,
+      "loss": 0.4492,
+      "step": 647
+    },
+    {
+      "epoch": 0.4159178433889602,
+      "grad_norm": 0.014927718788385391,
+      "learning_rate": 6.578661669985669e-05,
+      "loss": 0.4395,
+      "step": 648
+    },
+    {
+      "epoch": 0.4165596919127086,
+      "grad_norm": 0.011675119400024414,
+      "learning_rate": 6.568794284640383e-05,
+      "loss": 0.4102,
+      "step": 649
+    },
+    {
+      "epoch": 0.417201540436457,
+      "grad_norm": 0.012475242838263512,
+      "learning_rate": 6.558920117625005e-05,
+      "loss": 0.4082,
+      "step": 650
+    },
+    {
+      "epoch": 0.4178433889602054,
+      "grad_norm": 0.014828507788479328,
+      "learning_rate": 6.549039211624129e-05,
+      "loss": 0.4062,
+      "step": 651
+    },
+    {
+      "epoch": 0.4184852374839538,
+      "grad_norm": 0.01544868666678667,
+      "learning_rate": 6.539151609351476e-05,
+      "loss": 0.4336,
+      "step": 652
+    },
+    {
+      "epoch": 0.4191270860077022,
+      "grad_norm": 0.011812376789748669,
+      "learning_rate": 6.529257353549717e-05,
+      "loss": 0.3789,
+      "step": 653
+    },
+    {
+      "epoch": 0.4197689345314506,
+      "grad_norm": 0.013554428704082966,
+      "learning_rate": 6.519356486990287e-05,
+      "loss": 0.4297,
+      "step": 654
+    },
+    {
+      "epoch": 0.42041078305519897,
+      "grad_norm": 0.013231674209237099,
+      "learning_rate": 6.509449052473193e-05,
+      "loss": 0.4102,
+      "step": 655
+    },
+    {
+      "epoch": 0.42105263157894735,
+      "grad_norm": 0.013440310023725033,
+      "learning_rate": 6.499535092826835e-05,
+      "loss": 0.3984,
+      "step": 656
+    },
+    {
+      "epoch": 0.4216944801026958,
+      "grad_norm": 0.01264109555631876,
+      "learning_rate": 6.489614650907825e-05,
+      "loss": 0.4277,
+      "step": 657
+    },
+    {
+      "epoch": 0.42233632862644416,
+      "grad_norm": 0.01602267101407051,
+      "learning_rate": 6.479687769600795e-05,
+      "loss": 0.4258,
+      "step": 658
+    },
+    {
+      "epoch": 0.42297817715019254,
+      "grad_norm": 0.013843649998307228,
+      "learning_rate": 6.469754491818212e-05,
+      "loss": 0.3945,
+      "step": 659
+    },
+    {
+      "epoch": 0.42362002567394097,
+      "grad_norm": 0.011772658675909042,
+      "learning_rate": 6.459814860500194e-05,
+      "loss": 0.4004,
+      "step": 660
+    },
+    {
+      "epoch": 0.42426187419768935,
+      "grad_norm": 0.011879803612828255,
+      "learning_rate": 6.449868918614325e-05,
+      "loss": 0.4023,
+      "step": 661
+    },
+    {
+      "epoch": 0.4249037227214377,
+      "grad_norm": 0.016992155462503433,
+      "learning_rate": 6.439916709155468e-05,
+      "loss": 0.416,
+      "step": 662
+    },
+    {
+      "epoch": 0.42554557124518616,
+      "grad_norm": 0.013572323136031628,
+      "learning_rate": 6.429958275145583e-05,
+      "loss": 0.4082,
+      "step": 663
+    },
+    {
+      "epoch": 0.42618741976893454,
+      "grad_norm": 0.012573149986565113,
+      "learning_rate": 6.419993659633535e-05,
+      "loss": 0.4336,
+      "step": 664
+    },
+    {
+      "epoch": 0.4268292682926829,
+      "grad_norm": 0.011085672304034233,
+      "learning_rate": 6.41002290569491e-05,
+      "loss": 0.4121,
+      "step": 665
+    },
+    {
+      "epoch": 0.4274711168164313,
+      "grad_norm": 0.01807100512087345,
+      "learning_rate": 6.400046056431829e-05,
+      "loss": 0.4277,
+      "step": 666
+    },
+    {
+      "epoch": 0.42811296534017973,
+      "grad_norm": 0.01302388682961464,
+      "learning_rate": 6.390063154972767e-05,
+      "loss": 0.4219,
+      "step": 667
+    },
+    {
+      "epoch": 0.4287548138639281,
+      "grad_norm": 0.01238784659653902,
+      "learning_rate": 6.38007424447236e-05,
+      "loss": 0.4023,
+      "step": 668
+    },
+    {
+      "epoch": 0.4293966623876765,
+      "grad_norm": 0.014977477490901947,
+      "learning_rate": 6.370079368111214e-05,
+      "loss": 0.4062,
+      "step": 669
+    },
+    {
+      "epoch": 0.4300385109114249,
+      "grad_norm": 0.014952096156775951,
+      "learning_rate": 6.360078569095734e-05,
+      "loss": 0.4219,
+      "step": 670
+    },
+    {
+      "epoch": 0.4306803594351733,
+      "grad_norm": 0.012696454301476479,
+      "learning_rate": 6.350071890657918e-05,
+      "loss": 0.3848,
+      "step": 671
+    },
+    {
+      "epoch": 0.4313222079589217,
+      "grad_norm": 0.013992464169859886,
+      "learning_rate": 6.340059376055193e-05,
+      "loss": 0.4336,
+      "step": 672
+    },
+    {
+      "epoch": 0.4319640564826701,
+      "grad_norm": 0.015323866158723831,
+      "learning_rate": 6.330041068570198e-05,
+      "loss": 0.4258,
+      "step": 673
+    },
+    {
+      "epoch": 0.4326059050064185,
+      "grad_norm": 0.013826565816998482,
+      "learning_rate": 6.32001701151063e-05,
+      "loss": 0.418,
+      "step": 674
+    },
+    {
+      "epoch": 0.43324775353016687,
+      "grad_norm": 0.015464494936168194,
+      "learning_rate": 6.309987248209029e-05,
+      "loss": 0.4023,
+      "step": 675
+    },
+    {
+      "epoch": 0.4338896020539153,
+      "grad_norm": 0.015331785194575787,
+      "learning_rate": 6.299951822022609e-05,
+      "loss": 0.4102,
+      "step": 676
+    },
+    {
+      "epoch": 0.4345314505776637,
+      "grad_norm": 0.012050331570208073,
+      "learning_rate": 6.289910776333062e-05,
+      "loss": 0.418,
+      "step": 677
+    },
+    {
+      "epoch": 0.43517329910141206,
+      "grad_norm": 0.016262125223875046,
+      "learning_rate": 6.279864154546366e-05,
+      "loss": 0.4102,
+      "step": 678
+    },
+    {
+      "epoch": 0.43581514762516044,
+      "grad_norm": 0.014397810213267803,
+      "learning_rate": 6.269812000092619e-05,
+      "loss": 0.4219,
+      "step": 679
+    },
+    {
+      "epoch": 0.43645699614890887,
+      "grad_norm": 0.012025640346109867,
+      "learning_rate": 6.259754356425818e-05,
+      "loss": 0.4199,
+      "step": 680
+    },
+    {
+      "epoch": 0.43709884467265725,
+      "grad_norm": 0.012223280966281891,
+      "learning_rate": 6.249691267023701e-05,
+      "loss": 0.4141,
+      "step": 681
+    },
+    {
+      "epoch": 0.43774069319640563,
+      "grad_norm": 0.013894766569137573,
+      "learning_rate": 6.239622775387543e-05,
+      "loss": 0.4141,
+      "step": 682
+    },
+    {
+      "epoch": 0.43838254172015406,
+      "grad_norm": 0.013338884338736534,
+      "learning_rate": 6.229548925041973e-05,
+      "loss": 0.4141,
+      "step": 683
+    },
+    {
+      "epoch": 0.43902439024390244,
+      "grad_norm": 0.012654834426939487,
+      "learning_rate": 6.219469759534784e-05,
+      "loss": 0.4102,
+      "step": 684
+    },
+    {
+      "epoch": 0.4396662387676508,
+      "grad_norm": 0.012363187968730927,
+      "learning_rate": 6.209385322436746e-05,
+      "loss": 0.3809,
+      "step": 685
+    },
+    {
+      "epoch": 0.44030808729139925,
+      "grad_norm": 0.012646438553929329,
+      "learning_rate": 6.199295657341419e-05,
+      "loss": 0.418,
+      "step": 686
+    },
+    {
+      "epoch": 0.44094993581514763,
+      "grad_norm": 0.013392324559390545,
+      "learning_rate": 6.189200807864959e-05,
+      "loss": 0.4023,
+      "step": 687
+    },
+    {
+      "epoch": 0.441591784338896,
+      "grad_norm": 0.013896627351641655,
+      "learning_rate": 6.179100817645938e-05,
+      "loss": 0.4082,
+      "step": 688
+    },
+    {
+      "epoch": 0.44223363286264444,
+      "grad_norm": 0.014535841532051563,
+      "learning_rate": 6.168995730345145e-05,
+      "loss": 0.418,
+      "step": 689
+    },
+    {
+      "epoch": 0.4428754813863928,
+      "grad_norm": 0.015824053436517715,
+      "learning_rate": 6.15888558964541e-05,
+      "loss": 0.3926,
+      "step": 690
+    },
+    {
+      "epoch": 0.4435173299101412,
+      "grad_norm": 0.012450765818357468,
+      "learning_rate": 6.148770439251405e-05,
+      "loss": 0.416,
+      "step": 691
+    },
+    {
+      "epoch": 0.4441591784338896,
+      "grad_norm": 0.012254000641405582,
+      "learning_rate": 6.138650322889453e-05,
+      "loss": 0.4219,
+      "step": 692
+    },
+    {
+      "epoch": 0.444801026957638,
+      "grad_norm": 0.018290603533387184,
+      "learning_rate": 6.128525284307354e-05,
+      "loss": 0.4219,
+      "step": 693
+    },
+    {
+      "epoch": 0.4454428754813864,
+      "grad_norm": 0.012934917584061623,
+      "learning_rate": 6.118395367274177e-05,
+      "loss": 0.4297,
+      "step": 694
+    },
+    {
+      "epoch": 0.44608472400513477,
+      "grad_norm": 0.01413138210773468,
+      "learning_rate": 6.108260615580086e-05,
+      "loss": 0.4336,
+      "step": 695
+    },
+    {
+      "epoch": 0.4467265725288832,
+      "grad_norm": 0.0147935152053833,
+      "learning_rate": 6.09812107303614e-05,
+      "loss": 0.4102,
+      "step": 696
+    },
+    {
+      "epoch": 0.4473684210526316,
+      "grad_norm": 0.013712236657738686,
+      "learning_rate": 6.087976783474114e-05,
+      "loss": 0.418,
+      "step": 697
+    },
+    {
+      "epoch": 0.44801026957637996,
+      "grad_norm": 0.012871808372437954,
+      "learning_rate": 6.0778277907462945e-05,
+      "loss": 0.4023,
+      "step": 698
+    },
+    {
+      "epoch": 0.4486521181001284,
+      "grad_norm": 0.013505810871720314,
+      "learning_rate": 6.06767413872531e-05,
+      "loss": 0.4121,
+      "step": 699
+    },
+    {
+      "epoch": 0.4492939666238768,
+      "grad_norm": 0.014261540025472641,
+      "learning_rate": 6.0575158713039234e-05,
+      "loss": 0.4102,
+      "step": 700
+    },
+    {
+      "epoch": 0.44993581514762515,
+      "grad_norm": 0.014325248077511787,
+      "learning_rate": 6.047353032394849e-05,
+      "loss": 0.4238,
+      "step": 701
+    },
+    {
+      "epoch": 0.45057766367137353,
+      "grad_norm": 0.014693713746964931,
+      "learning_rate": 6.037185665930567e-05,
+      "loss": 0.4297,
+      "step": 702
+    },
+    {
+      "epoch": 0.45121951219512196,
+      "grad_norm": 0.01406776811927557,
+      "learning_rate": 6.027013815863128e-05,
+      "loss": 0.3848,
+      "step": 703
+    },
+    {
+      "epoch": 0.45186136071887034,
+      "grad_norm": 0.011017663404345512,
+      "learning_rate": 6.016837526163962e-05,
+      "loss": 0.4023,
+      "step": 704
+    },
+    {
+      "epoch": 0.4525032092426187,
+      "grad_norm": 0.01230404619127512,
+      "learning_rate": 6.006656840823696e-05,
+      "loss": 0.418,
+      "step": 705
+    },
+    {
+      "epoch": 0.45314505776636715,
+      "grad_norm": 0.012102317065000534,
+      "learning_rate": 5.996471803851951e-05,
+      "loss": 0.3789,
+      "step": 706
+    },
+    {
+      "epoch": 0.45378690629011553,
+      "grad_norm": 0.01572810672223568,
+      "learning_rate": 5.986282459277168e-05,
+      "loss": 0.4258,
+      "step": 707
+    },
+    {
+      "epoch": 0.4544287548138639,
+      "grad_norm": 0.013043870218098164,
+      "learning_rate": 5.976088851146405e-05,
+      "loss": 0.3926,
+      "step": 708
+    },
+    {
+      "epoch": 0.45507060333761234,
+      "grad_norm": 0.013452388346195221,
+      "learning_rate": 5.9658910235251495e-05,
+      "loss": 0.4297,
+      "step": 709
+    },
+    {
+      "epoch": 0.4557124518613607,
+      "grad_norm": 0.011983458884060383,
+      "learning_rate": 5.9556890204971326e-05,
+      "loss": 0.4219,
+      "step": 710
+    },
+    {
+      "epoch": 0.4563543003851091,
+      "grad_norm": 0.012922801077365875,
+      "learning_rate": 5.945482886164132e-05,
+      "loss": 0.4336,
+      "step": 711
+    },
+    {
+      "epoch": 0.45699614890885754,
+      "grad_norm": 0.01430261880159378,
+      "learning_rate": 5.935272664645786e-05,
+      "loss": 0.4141,
+      "step": 712
+    },
+    {
+      "epoch": 0.4576379974326059,
+      "grad_norm": 0.013975469395518303,
+      "learning_rate": 5.9250584000794017e-05,
+      "loss": 0.4238,
+      "step": 713
+    },
+    {
+      "epoch": 0.4582798459563543,
+      "grad_norm": 0.013664472848176956,
+      "learning_rate": 5.914840136619761e-05,
+      "loss": 0.4453,
+      "step": 714
+    },
+    {
+      "epoch": 0.45892169448010267,
+      "grad_norm": 0.01542427483946085,
+      "learning_rate": 5.904617918438936e-05,
+      "loss": 0.3945,
+      "step": 715
+    },
+    {
+      "epoch": 0.4595635430038511,
+      "grad_norm": 0.012004110030829906,
+      "learning_rate": 5.894391789726093e-05,
+      "loss": 0.3945,
+      "step": 716
+    },
+    {
+      "epoch": 0.4602053915275995,
+      "grad_norm": 0.014152607880532742,
+      "learning_rate": 5.8841617946873015e-05,
+      "loss": 0.4023,
+      "step": 717
+    },
+    {
+      "epoch": 0.46084724005134786,
+      "grad_norm": 0.014494920149445534,
+      "learning_rate": 5.873927977545346e-05,
+      "loss": 0.4219,
+      "step": 718
+    },
+    {
+      "epoch": 0.4614890885750963,
+      "grad_norm": 0.01685180701315403,
+      "learning_rate": 5.863690382539535e-05,
+      "loss": 0.3945,
+      "step": 719
+    },
+    {
+      "epoch": 0.4621309370988447,
+      "grad_norm": 0.013664394617080688,
+      "learning_rate": 5.853449053925505e-05,
+      "loss": 0.4219,
+      "step": 720
+    },
+    {
+      "epoch": 0.46277278562259305,
+      "grad_norm": 0.014847121201455593,
+      "learning_rate": 5.843204035975033e-05,
+      "loss": 0.4258,
+      "step": 721
+    },
+    {
+      "epoch": 0.4634146341463415,
+      "grad_norm": 0.015490569174289703,
+      "learning_rate": 5.832955372975848e-05,
+      "loss": 0.418,
+      "step": 722
+    },
+    {
+      "epoch": 0.46405648267008986,
+      "grad_norm": 0.015505144372582436,
+      "learning_rate": 5.822703109231431e-05,
+      "loss": 0.4258,
+      "step": 723
+    },
+    {
+      "epoch": 0.46469833119383824,
+      "grad_norm": 0.012789933010935783,
+      "learning_rate": 5.812447289060832e-05,
+      "loss": 0.4023,
+      "step": 724
+    },
+    {
+      "epoch": 0.4653401797175867,
+      "grad_norm": 0.01355286780744791,
+      "learning_rate": 5.80218795679847e-05,
+      "loss": 0.3984,
+      "step": 725
+    },
+    {
+      "epoch": 0.46598202824133506,
+      "grad_norm": 0.013972820714116096,
+      "learning_rate": 5.791925156793956e-05,
+      "loss": 0.4258,
+      "step": 726
+    },
+    {
+      "epoch": 0.46662387676508343,
+      "grad_norm": 0.012723516672849655,
+      "learning_rate": 5.781658933411882e-05,
+      "loss": 0.3633,
+      "step": 727
+    },
+    {
+      "epoch": 0.4672657252888318,
+      "grad_norm": 0.013931083492934704,
+      "learning_rate": 5.7713893310316426e-05,
+      "loss": 0.4258,
+      "step": 728
+    },
+    {
+      "epoch": 0.46790757381258025,
+      "grad_norm": 0.014217558316886425,
+      "learning_rate": 5.761116394047238e-05,
+      "loss": 0.4434,
+      "step": 729
+    },
+    {
+      "epoch": 0.4685494223363286,
+      "grad_norm": 0.012869459576904774,
+      "learning_rate": 5.750840166867085e-05,
+      "loss": 0.4062,
+      "step": 730
+    },
+    {
+      "epoch": 0.469191270860077,
+      "grad_norm": 0.014253105036914349,
+      "learning_rate": 5.740560693913825e-05,
+      "loss": 0.4375,
+      "step": 731
+    },
+    {
+      "epoch": 0.46983311938382544,
+      "grad_norm": 0.012103088200092316,
+      "learning_rate": 5.7302780196241245e-05,
+      "loss": 0.4297,
+      "step": 732
+    },
+    {
+      "epoch": 0.4704749679075738,
+      "grad_norm": 0.012361271306872368,
+      "learning_rate": 5.7199921884484975e-05,
+      "loss": 0.4258,
+      "step": 733
+    },
+    {
+      "epoch": 0.4711168164313222,
+      "grad_norm": 0.015202773734927177,
+      "learning_rate": 5.7097032448510945e-05,
+      "loss": 0.4102,
+      "step": 734
+    },
+    {
+      "epoch": 0.47175866495507063,
+      "grad_norm": 0.012599785812199116,
+      "learning_rate": 5.699411233309528e-05,
+      "loss": 0.3984,
+      "step": 735
+    },
+    {
+      "epoch": 0.472400513478819,
+      "grad_norm": 0.011520575731992722,
+      "learning_rate": 5.689116198314673e-05,
+      "loss": 0.377,
+      "step": 736
+    },
+    {
+      "epoch": 0.4730423620025674,
+      "grad_norm": 0.01304789911955595,
+      "learning_rate": 5.678818184370469e-05,
+      "loss": 0.4102,
+      "step": 737
+    },
+    {
+      "epoch": 0.47368421052631576,
+      "grad_norm": 0.011445467360317707,
+      "learning_rate": 5.668517235993739e-05,
+      "loss": 0.3945,
+      "step": 738
+    },
+    {
+      "epoch": 0.4743260590500642,
+      "grad_norm": 0.012655424885451794,
+      "learning_rate": 5.658213397713985e-05,
+      "loss": 0.4023,
+      "step": 739
+    },
+    {
+      "epoch": 0.4749679075738126,
+      "grad_norm": 0.014334771782159805,
+      "learning_rate": 5.647906714073208e-05,
+      "loss": 0.4082,
+      "step": 740
+    },
+    {
+      "epoch": 0.47560975609756095,
+      "grad_norm": 0.011826477013528347,
+      "learning_rate": 5.637597229625705e-05,
+      "loss": 0.4375,
+      "step": 741
+    },
+    {
+      "epoch": 0.4762516046213094,
+      "grad_norm": 0.013866581954061985,
+      "learning_rate": 5.627284988937882e-05,
+      "loss": 0.4102,
+      "step": 742
+    },
+    {
+      "epoch": 0.47689345314505777,
+      "grad_norm": 0.011929787695407867,
+      "learning_rate": 5.616970036588058e-05,
+      "loss": 0.3945,
+      "step": 743
+    },
+    {
+      "epoch": 0.47753530166880614,
+      "grad_norm": 0.014417732134461403,
+      "learning_rate": 5.606652417166276e-05,
+      "loss": 0.418,
+      "step": 744
+    },
+    {
+      "epoch": 0.4781771501925546,
+      "grad_norm": 0.012242990545928478,
+      "learning_rate": 5.59633217527411e-05,
+      "loss": 0.4277,
+      "step": 745
+    },
+    {
+      "epoch": 0.47881899871630296,
+      "grad_norm": 0.012755023315548897,
+      "learning_rate": 5.586009355524465e-05,
+      "loss": 0.3945,
+      "step": 746
+    },
+    {
+      "epoch": 0.47946084724005134,
+      "grad_norm": 0.012235814705491066,
+      "learning_rate": 5.575684002541397e-05,
+      "loss": 0.3945,
+      "step": 747
+    },
+    {
+      "epoch": 0.48010269576379977,
+      "grad_norm": 0.013560086488723755,
+      "learning_rate": 5.5653561609599046e-05,
+      "loss": 0.4023,
+      "step": 748
+    },
+    {
+      "epoch": 0.48074454428754815,
+      "grad_norm": 0.013584466651082039,
+      "learning_rate": 5.555025875425751e-05,
+      "loss": 0.4102,
+      "step": 749
+    },
+    {
+      "epoch": 0.4813863928112965,
+      "grad_norm": 0.012287539429962635,
+      "learning_rate": 5.5446931905952624e-05,
+      "loss": 0.4023,
+      "step": 750
+    },
+    {
+      "epoch": 0.4820282413350449,
+      "grad_norm": 0.015574885532259941,
+      "learning_rate": 5.534358151135135e-05,
+      "loss": 0.4023,
+      "step": 751
+    },
+    {
+      "epoch": 0.48267008985879334,
+      "grad_norm": 0.0164936576038599,
+      "learning_rate": 5.524020801722246e-05,
+      "loss": 0.4102,
+      "step": 752
+    },
+    {
+      "epoch": 0.4833119383825417,
+      "grad_norm": 0.012463854625821114,
+      "learning_rate": 5.513681187043456e-05,
+      "loss": 0.3965,
+      "step": 753
+    },
+    {
+      "epoch": 0.4839537869062901,
+      "grad_norm": 0.013957749120891094,
+      "learning_rate": 5.503339351795419e-05,
+      "loss": 0.4062,
+      "step": 754
+    },
+    {
+      "epoch": 0.48459563543003853,
+      "grad_norm": 0.013575349003076553,
+      "learning_rate": 5.4929953406843906e-05,
+      "loss": 0.4375,
+      "step": 755
+    },
+    {
+      "epoch": 0.4852374839537869,
+      "grad_norm": 0.012251514010131359,
+      "learning_rate": 5.4826491984260284e-05,
+      "loss": 0.4043,
+      "step": 756
+    },
+    {
+      "epoch": 0.4858793324775353,
+      "grad_norm": 0.012463661842048168,
+      "learning_rate": 5.472300969745204e-05,
+      "loss": 0.418,
+      "step": 757
+    },
+    {
+      "epoch": 0.4865211810012837,
+      "grad_norm": 0.014550065621733665,
+      "learning_rate": 5.46195069937581e-05,
+      "loss": 0.4297,
+      "step": 758
+    },
+    {
+      "epoch": 0.4871630295250321,
+      "grad_norm": 0.012784469872713089,
+      "learning_rate": 5.451598432060563e-05,
+      "loss": 0.4473,
+      "step": 759
+    },
+    {
+      "epoch": 0.4878048780487805,
+      "grad_norm": 0.014856231398880482,
+      "learning_rate": 5.4412442125508113e-05,
+      "loss": 0.4023,
+      "step": 760
+    },
+    {
+      "epoch": 0.4884467265725289,
+      "grad_norm": 0.014101446606218815,
+      "learning_rate": 5.430888085606346e-05,
+      "loss": 0.4336,
+      "step": 761
+    },
+    {
+      "epoch": 0.4890885750962773,
+      "grad_norm": 0.013721944764256477,
+      "learning_rate": 5.4205300959951974e-05,
+      "loss": 0.3945,
+      "step": 762
+    },
+    {
+      "epoch": 0.48973042362002567,
+      "grad_norm": 0.013042719103395939,
+      "learning_rate": 5.410170288493458e-05,
+      "loss": 0.3867,
+      "step": 763
+    },
+    {
+      "epoch": 0.49037227214377405,
+      "grad_norm": 0.013128210790455341,
+      "learning_rate": 5.399808707885069e-05,
+      "loss": 0.3789,
+      "step": 764
+    },
+    {
+      "epoch": 0.4910141206675225,
+      "grad_norm": 0.016061117872595787,
+      "learning_rate": 5.389445398961639e-05,
+      "loss": 0.3965,
+      "step": 765
+    },
+    {
+      "epoch": 0.49165596919127086,
+      "grad_norm": 0.013035389594733715,
+      "learning_rate": 5.3790804065222524e-05,
+      "loss": 0.4023,
+      "step": 766
+    },
+    {
+      "epoch": 0.49229781771501924,
+      "grad_norm": 0.012277752161026001,
+      "learning_rate": 5.368713775373264e-05,
+      "loss": 0.375,
+      "step": 767
+    },
+    {
+      "epoch": 0.49293966623876767,
+      "grad_norm": 0.013250727206468582,
+      "learning_rate": 5.3583455503281176e-05,
+      "loss": 0.4219,
+      "step": 768
+    },
+    {
+      "epoch": 0.49358151476251605,
+      "grad_norm": 0.01238930132240057,
+      "learning_rate": 5.347975776207148e-05,
+      "loss": 0.416,
+      "step": 769
+    },
+    {
+      "epoch": 0.4942233632862644,
+      "grad_norm": 0.011913436464965343,
+      "learning_rate": 5.337604497837383e-05,
+      "loss": 0.4023,
+      "step": 770
+    },
+    {
+      "epoch": 0.49486521181001286,
+      "grad_norm": 0.01252955012023449,
+      "learning_rate": 5.3272317600523505e-05,
+      "loss": 0.3828,
+      "step": 771
+    },
+    {
+      "epoch": 0.49550706033376124,
+      "grad_norm": 0.013348378241062164,
+      "learning_rate": 5.3168576076918966e-05,
+      "loss": 0.4102,
+      "step": 772
+    },
+    {
+      "epoch": 0.4961489088575096,
+      "grad_norm": 0.014592018909752369,
+      "learning_rate": 5.3064820856019745e-05,
+      "loss": 0.3828,
+      "step": 773
+    },
+    {
+      "epoch": 0.496790757381258,
+      "grad_norm": 0.014033477753400803,
+      "learning_rate": 5.2961052386344615e-05,
+      "loss": 0.4121,
+      "step": 774
+    },
+    {
+      "epoch": 0.49743260590500643,
+      "grad_norm": 0.013774393126368523,
+      "learning_rate": 5.28572711164696e-05,
+      "loss": 0.4102,
+      "step": 775
+    },
+    {
+      "epoch": 0.4980744544287548,
+      "grad_norm": 0.013182034716010094,
+      "learning_rate": 5.2753477495026084e-05,
+      "loss": 0.4297,
+      "step": 776
+    },
+    {
+      "epoch": 0.4987163029525032,
+      "grad_norm": 0.01512659341096878,
+      "learning_rate": 5.264967197069884e-05,
+      "loss": 0.4004,
+      "step": 777
+    },
+    {
+      "epoch": 0.4993581514762516,
+      "grad_norm": 0.013438140042126179,
+      "learning_rate": 5.25458549922241e-05,
+      "loss": 0.3945,
+      "step": 778
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.012077905237674713,
+      "learning_rate": 5.244202700838756e-05,
+      "loss": 0.3867,
+      "step": 779
+    },
+    {
+      "epoch": 0.5006418485237484,
+      "grad_norm": 0.015471718274056911,
+      "learning_rate": 5.233818846802255e-05,
+      "loss": 0.3887,
+      "step": 780
+    },
+    {
+      "epoch": 0.5012836970474968,
+      "grad_norm": 0.015281775034964085,
+      "learning_rate": 5.223433982000804e-05,
+      "loss": 0.3926,
+      "step": 781
+    },
+    {
+      "epoch": 0.5019255455712451,
+      "grad_norm": 0.014022503979504108,
+      "learning_rate": 5.213048151326664e-05,
+      "loss": 0.4043,
+      "step": 782
+    },
+    {
+      "epoch": 0.5025673940949936,
+      "grad_norm": 0.012185904197394848,
+      "learning_rate": 5.2026613996762754e-05,
+      "loss": 0.375,
+      "step": 783
+    },
+    {
+      "epoch": 0.503209242618742,
+      "grad_norm": 0.01691463030874729,
+      "learning_rate": 5.192273771950057e-05,
+      "loss": 0.4141,
+      "step": 784
+    },
+    {
+      "epoch": 0.5038510911424904,
+      "grad_norm": 0.013279242441058159,
+      "learning_rate": 5.1818853130522184e-05,
+      "loss": 0.4023,
+      "step": 785
+    },
+    {
+      "epoch": 0.5044929396662388,
+      "grad_norm": 0.014866221696138382,
+      "learning_rate": 5.17149606789056e-05,
+      "loss": 0.4023,
+      "step": 786
+    },
+    {
+      "epoch": 0.5051347881899871,
+      "grad_norm": 0.014251725748181343,
+      "learning_rate": 5.161106081376281e-05,
+      "loss": 0.4199,
+      "step": 787
+    },
+    {
+      "epoch": 0.5057766367137355,
+      "grad_norm": 0.01272731926292181,
+      "learning_rate": 5.1507153984237857e-05,
+      "loss": 0.4023,
+      "step": 788
+    },
+    {
+      "epoch": 0.506418485237484,
+      "grad_norm": 0.014095312915742397,
+      "learning_rate": 5.140324063950488e-05,
+      "loss": 0.4082,
+      "step": 789
+    },
+    {
+      "epoch": 0.5070603337612324,
+      "grad_norm": 0.015140281058847904,
+      "learning_rate": 5.1299321228766194e-05,
+      "loss": 0.3809,
+      "step": 790
+    },
+    {
+      "epoch": 0.5077021822849808,
+      "grad_norm": 0.013064335100352764,
+      "learning_rate": 5.119539620125037e-05,
+      "loss": 0.4297,
+      "step": 791
+    },
+    {
+      "epoch": 0.5083440308087291,
+      "grad_norm": 0.013767260126769543,
+      "learning_rate": 5.1091466006210185e-05,
+      "loss": 0.3965,
+      "step": 792
+    },
+    {
+      "epoch": 0.5089858793324775,
+      "grad_norm": 0.014455810189247131,
+      "learning_rate": 5.0987531092920806e-05,
+      "loss": 0.3945,
+      "step": 793
+    },
+    {
+      "epoch": 0.5096277278562259,
+      "grad_norm": 0.012534044682979584,
+      "learning_rate": 5.0883591910677774e-05,
+      "loss": 0.3945,
+      "step": 794
+    },
+    {
+      "epoch": 0.5102695763799743,
+      "grad_norm": 0.015598188154399395,
+      "learning_rate": 5.0779648908795116e-05,
+      "loss": 0.3945,
+      "step": 795
+    },
+    {
+      "epoch": 0.5109114249037228,
+      "grad_norm": 0.013126207515597343,
+      "learning_rate": 5.067570253660333e-05,
+      "loss": 0.4102,
+      "step": 796
+    },
+    {
+      "epoch": 0.5115532734274711,
+      "grad_norm": 0.013315839692950249,
+      "learning_rate": 5.0571753243447515e-05,
+      "loss": 0.4043,
+      "step": 797
+    },
+    {
+      "epoch": 0.5121951219512195,
+      "grad_norm": 0.016860026866197586,
+      "learning_rate": 5.046780147868537e-05,
+      "loss": 0.3848,
+      "step": 798
+    },
+    {
+      "epoch": 0.5128369704749679,
+      "grad_norm": 0.01219270471483469,
+      "learning_rate": 5.0363847691685305e-05,
+      "loss": 0.3945,
+      "step": 799
+    },
+    {
+      "epoch": 0.5134788189987163,
+      "grad_norm": 0.011769594624638557,
+      "learning_rate": 5.0259892331824474e-05,
+      "loss": 0.3867,
+      "step": 800
+    },
+    {
+      "epoch": 0.5141206675224647,
+      "grad_norm": 0.012503019534051418,
+      "learning_rate": 5.015593584848679e-05,
+      "loss": 0.416,
+      "step": 801
+    },
+    {
+      "epoch": 0.5147625160462131,
+      "grad_norm": 0.01306526642292738,
+      "learning_rate": 5.0051978691061054e-05,
+      "loss": 0.4141,
+      "step": 802
+    },
+    {
+      "epoch": 0.5154043645699615,
+      "grad_norm": 0.012884598225355148,
+      "learning_rate": 4.9948021308938965e-05,
+      "loss": 0.4043,
+      "step": 803
+    },
+    {
+      "epoch": 0.5160462130937099,
+      "grad_norm": 0.010572467930614948,
+      "learning_rate": 4.984406415151323e-05,
+      "loss": 0.3828,
+      "step": 804
+    },
+    {
+      "epoch": 0.5166880616174583,
+      "grad_norm": 0.014578374102711678,
+      "learning_rate": 4.974010766817555e-05,
+      "loss": 0.4141,
+      "step": 805
+    },
+    {
+      "epoch": 0.5173299101412067,
+      "grad_norm": 0.012406845577061176,
+      "learning_rate": 4.9636152308314687e-05,
+      "loss": 0.3926,
+      "step": 806
+    },
+    {
+      "epoch": 0.517971758664955,
+      "grad_norm": 0.012090642005205154,
+      "learning_rate": 4.9532198521314635e-05,
+      "loss": 0.4102,
+      "step": 807
+    },
+    {
+      "epoch": 0.5186136071887034,
+      "grad_norm": 0.013852028176188469,
+      "learning_rate": 4.9428246756552496e-05,
+      "loss": 0.3945,
+      "step": 808
+    },
+    {
+      "epoch": 0.5192554557124519,
+      "grad_norm": 0.01639687456190586,
+      "learning_rate": 4.9324297463396685e-05,
+      "loss": 0.4062,
+      "step": 809
+    },
+    {
+      "epoch": 0.5198973042362003,
+      "grad_norm": 0.01494478527456522,
+      "learning_rate": 4.922035109120491e-05,
+      "loss": 0.4277,
+      "step": 810
+    },
+    {
+      "epoch": 0.5205391527599487,
+      "grad_norm": 0.012924427166581154,
+      "learning_rate": 4.911640808932223e-05,
+      "loss": 0.4062,
+      "step": 811
+    },
+    {
+      "epoch": 0.521181001283697,
+      "grad_norm": 0.012701825238764286,
+      "learning_rate": 4.901246890707921e-05,
+      "loss": 0.3984,
+      "step": 812
+    },
+    {
+      "epoch": 0.5218228498074454,
+      "grad_norm": 0.013332185335457325,
+      "learning_rate": 4.8908533993789826e-05,
+      "loss": 0.4023,
+      "step": 813
+    },
+    {
+      "epoch": 0.5224646983311938,
+      "grad_norm": 0.015123789198696613,
+      "learning_rate": 4.880460379874965e-05,
+      "loss": 0.4102,
+      "step": 814
+    },
+    {
+      "epoch": 0.5231065468549422,
+      "grad_norm": 0.011819974519312382,
+      "learning_rate": 4.870067877123382e-05,
+      "loss": 0.3789,
+      "step": 815
+    },
+    {
+      "epoch": 0.5237483953786907,
+      "grad_norm": 0.016871098428964615,
+      "learning_rate": 4.859675936049514e-05,
+      "loss": 0.3945,
+      "step": 816
+    },
+    {
+      "epoch": 0.524390243902439,
+      "grad_norm": 0.013095321133732796,
+      "learning_rate": 4.849284601576215e-05,
+      "loss": 0.3809,
+      "step": 817
+    },
+    {
+      "epoch": 0.5250320924261874,
+      "grad_norm": 0.014770804904401302,
+      "learning_rate": 4.8388939186237195e-05,
+      "loss": 0.3711,
+      "step": 818
+    },
+    {
+      "epoch": 0.5256739409499358,
+      "grad_norm": 0.015598641708493233,
+      "learning_rate": 4.828503932109441e-05,
+      "loss": 0.4121,
+      "step": 819
+    },
+    {
+      "epoch": 0.5263157894736842,
+      "grad_norm": 0.012911478988826275,
+      "learning_rate": 4.818114686947783e-05,
+      "loss": 0.3887,
+      "step": 820
+    },
+    {
+      "epoch": 0.5269576379974326,
+      "grad_norm": 0.013151862658560276,
+      "learning_rate": 4.8077262280499444e-05,
+      "loss": 0.4062,
+      "step": 821
+    },
+    {
+      "epoch": 0.527599486521181,
+      "grad_norm": 0.012958858162164688,
+      "learning_rate": 4.797338600323727e-05,
+      "loss": 0.3984,
+      "step": 822
+    },
+    {
+      "epoch": 0.5282413350449294,
+      "grad_norm": 0.013672449626028538,
+      "learning_rate": 4.7869518486733364e-05,
+      "loss": 0.418,
+      "step": 823
+    },
+    {
+      "epoch": 0.5288831835686778,
+      "grad_norm": 0.01231690589338541,
+      "learning_rate": 4.776566017999197e-05,
+      "loss": 0.373,
+      "step": 824
+    },
+    {
+      "epoch": 0.5295250320924262,
+      "grad_norm": 0.012452663853764534,
+      "learning_rate": 4.766181153197746e-05,
+      "loss": 0.3711,
+      "step": 825
+    },
+    {
+      "epoch": 0.5301668806161746,
+      "grad_norm": 0.011798235587775707,
+      "learning_rate": 4.755797299161246e-05,
+      "loss": 0.3828,
+      "step": 826
+    },
+    {
+      "epoch": 0.5308087291399229,
+      "grad_norm": 0.013432770036160946,
+      "learning_rate": 4.7454145007775925e-05,
+      "loss": 0.375,
+      "step": 827
+    },
+    {
+      "epoch": 0.5314505776636713,
+      "grad_norm": 0.013223826885223389,
+      "learning_rate": 4.7350328029301156e-05,
+      "loss": 0.3867,
+      "step": 828
+    },
+    {
+      "epoch": 0.5320924261874198,
+      "grad_norm": 0.012126738205552101,
+      "learning_rate": 4.7246522504973914e-05,
+      "loss": 0.3926,
+      "step": 829
+    },
+    {
+      "epoch": 0.5327342747111682,
+      "grad_norm": 0.011325270868837833,
+      "learning_rate": 4.714272888353041e-05,
+      "loss": 0.373,
+      "step": 830
+    },
+    {
+      "epoch": 0.5333761232349166,
+      "grad_norm": 0.01423851028084755,
+      "learning_rate": 4.70389476136554e-05,
+      "loss": 0.418,
+      "step": 831
+    },
+    {
+      "epoch": 0.5340179717586649,
+      "grad_norm": 0.01402893103659153,
+      "learning_rate": 4.6935179143980266e-05,
+      "loss": 0.4258,
+      "step": 832
+    },
+    {
+      "epoch": 0.5346598202824133,
+      "grad_norm": 0.012596989050507545,
+      "learning_rate": 4.683142392308105e-05,
+      "loss": 0.375,
+      "step": 833
+    },
+    {
+      "epoch": 0.5353016688061617,
+      "grad_norm": 0.012625569477677345,
+      "learning_rate": 4.672768239947649e-05,
+      "loss": 0.4043,
+      "step": 834
+    },
+    {
+      "epoch": 0.5359435173299102,
+      "grad_norm": 0.012342373840510845,
+      "learning_rate": 4.6623955021626184e-05,
+      "loss": 0.4023,
+      "step": 835
+    },
+    {
+      "epoch": 0.5365853658536586,
+      "grad_norm": 0.014212844893336296,
+      "learning_rate": 4.652024223792853e-05,
+      "loss": 0.4258,
+      "step": 836
+    },
+    {
+      "epoch": 0.537227214377407,
+      "grad_norm": 0.01258305087685585,
+      "learning_rate": 4.6416544496718835e-05,
+      "loss": 0.3945,
+      "step": 837
+    },
+    {
+      "epoch": 0.5378690629011553,
+      "grad_norm": 0.013877546414732933,
+      "learning_rate": 4.631286224626739e-05,
+      "loss": 0.3906,
+      "step": 838
+    },
+    {
+      "epoch": 0.5385109114249037,
+      "grad_norm": 0.013766865245997906,
+      "learning_rate": 4.620919593477749e-05,
+      "loss": 0.4141,
+      "step": 839
+    },
+    {
+      "epoch": 0.5391527599486521,
+      "grad_norm": 0.011848713271319866,
+      "learning_rate": 4.610554601038361e-05,
+      "loss": 0.3594,
+      "step": 840
+    },
+    {
+      "epoch": 0.5397946084724005,
+      "grad_norm": 0.013195286504924297,
+      "learning_rate": 4.600191292114932e-05,
+      "loss": 0.4219,
+      "step": 841
+    },
+    {
+      "epoch": 0.540436456996149,
+      "grad_norm": 0.012025582604110241,
+      "learning_rate": 4.5898297115065434e-05,
+      "loss": 0.3828,
+      "step": 842
+    },
+    {
+      "epoch": 0.5410783055198973,
+      "grad_norm": 0.014689466916024685,
+      "learning_rate": 4.579469904004803e-05,
+      "loss": 0.418,
+      "step": 843
+    },
+    {
+      "epoch": 0.5417201540436457,
+      "grad_norm": 0.01123505737632513,
+      "learning_rate": 4.5691119143936565e-05,
+      "loss": 0.3867,
+      "step": 844
+    },
+    {
+      "epoch": 0.5423620025673941,
+      "grad_norm": 0.012202415615320206,
+      "learning_rate": 4.558755787449189e-05,
+      "loss": 0.3848,
+      "step": 845
+    },
+    {
+      "epoch": 0.5430038510911425,
+      "grad_norm": 0.011510019190609455,
+      "learning_rate": 4.548401567939439e-05,
+      "loss": 0.3691,
+      "step": 846
+    },
+    {
+      "epoch": 0.5436456996148908,
+      "grad_norm": 0.014318268746137619,
+      "learning_rate": 4.538049300624192e-05,
+      "loss": 0.4297,
+      "step": 847
+    },
+    {
+      "epoch": 0.5442875481386393,
+      "grad_norm": 0.01239953376352787,
+      "learning_rate": 4.5276990302547964e-05,
+      "loss": 0.4102,
+      "step": 848
+    },
+    {
+      "epoch": 0.5449293966623877,
+      "grad_norm": 0.012069017626345158,
+      "learning_rate": 4.5173508015739735e-05,
+      "loss": 0.3789,
+      "step": 849
+    },
+    {
+      "epoch": 0.5455712451861361,
+      "grad_norm": 0.011770493350923061,
+      "learning_rate": 4.507004659315611e-05,
+      "loss": 0.3711,
+      "step": 850
+    },
+    {
+      "epoch": 0.5462130937098845,
+      "grad_norm": 0.012663186527788639,
+      "learning_rate": 4.496660648204581e-05,
+      "loss": 0.3828,
+      "step": 851
+    },
+    {
+      "epoch": 0.5468549422336328,
+      "grad_norm": 0.012398538179695606,
+      "learning_rate": 4.4863188129565454e-05,
+      "loss": 0.3906,
+      "step": 852
+    },
+    {
+      "epoch": 0.5474967907573812,
+      "grad_norm": 0.013944141566753387,
+      "learning_rate": 4.4759791982777546e-05,
+      "loss": 0.375,
+      "step": 853
+    },
+    {
+      "epoch": 0.5481386392811296,
+      "grad_norm": 0.012811684980988503,
+      "learning_rate": 4.465641848864866e-05,
+      "loss": 0.4199,
+      "step": 854
+    },
+    {
+      "epoch": 0.5487804878048781,
+      "grad_norm": 0.013038134202361107,
+      "learning_rate": 4.4553068094047394e-05,
+      "loss": 0.3965,
+      "step": 855
+    },
+    {
+      "epoch": 0.5494223363286265,
+      "grad_norm": 0.010878165252506733,
+      "learning_rate": 4.4449741245742494e-05,
+      "loss": 0.3828,
+      "step": 856
+    },
+    {
+      "epoch": 0.5500641848523748,
+      "grad_norm": 0.013144944794476032,
+      "learning_rate": 4.4346438390400965e-05,
+      "loss": 0.375,
+      "step": 857
+    },
+    {
+      "epoch": 0.5507060333761232,
+      "grad_norm": 0.013227356597781181,
+      "learning_rate": 4.4243159974586044e-05,
+      "loss": 0.3867,
+      "step": 858
+    },
+    {
+      "epoch": 0.5513478818998716,
+      "grad_norm": 0.013552071526646614,
+      "learning_rate": 4.413990644475536e-05,
+      "loss": 0.4102,
+      "step": 859
+    },
+    {
+      "epoch": 0.55198973042362,
+      "grad_norm": 0.016889218240976334,
+      "learning_rate": 4.4036678247258924e-05,
+      "loss": 0.4062,
+      "step": 860
+    },
+    {
+      "epoch": 0.5526315789473685,
+      "grad_norm": 0.01147947646677494,
+      "learning_rate": 4.393347582833724e-05,
+      "loss": 0.3867,
+      "step": 861
+    },
+    {
+      "epoch": 0.5532734274711169,
+      "grad_norm": 0.011723281815648079,
+      "learning_rate": 4.383029963411943e-05,
+      "loss": 0.3867,
+      "step": 862
+    },
+    {
+      "epoch": 0.5539152759948652,
+      "grad_norm": 0.012720635160803795,
+      "learning_rate": 4.372715011062118e-05,
+      "loss": 0.4102,
+      "step": 863
+    },
+    {
+      "epoch": 0.5545571245186136,
+      "grad_norm": 0.016574840992689133,
+      "learning_rate": 4.3624027703742953e-05,
+      "loss": 0.3945,
+      "step": 864
+    },
+    {
+      "epoch": 0.555198973042362,
+      "grad_norm": 0.01387743279337883,
+      "learning_rate": 4.352093285926793e-05,
+      "loss": 0.4004,
+      "step": 865
+    },
+    {
+      "epoch": 0.5558408215661104,
+      "grad_norm": 0.014131239615380764,
+      "learning_rate": 4.3417866022860156e-05,
+      "loss": 0.4297,
+      "step": 866
+    },
+    {
+      "epoch": 0.5564826700898587,
+      "grad_norm": 0.013099314644932747,
+      "learning_rate": 4.331482764006262e-05,
+      "loss": 0.4141,
+      "step": 867
+    },
+    {
+      "epoch": 0.5571245186136072,
+      "grad_norm": 0.012790219858288765,
+      "learning_rate": 4.3211818156295307e-05,
+      "loss": 0.3945,
+      "step": 868
+    },
+    {
+      "epoch": 0.5577663671373556,
+      "grad_norm": 0.017325740307569504,
+      "learning_rate": 4.310883801685328e-05,
+      "loss": 0.3906,
+      "step": 869
+    },
+    {
+      "epoch": 0.558408215661104,
+      "grad_norm": 0.012533944100141525,
+      "learning_rate": 4.300588766690473e-05,
+      "loss": 0.3711,
+      "step": 870
+    },
+    {
+      "epoch": 0.5590500641848524,
+      "grad_norm": 0.013145022094249725,
+      "learning_rate": 4.290296755148907e-05,
+      "loss": 0.3984,
+      "step": 871
+    },
+    {
+      "epoch": 0.5596919127086007,
+      "grad_norm": 0.014711151830852032,
+      "learning_rate": 4.280007811551505e-05,
+      "loss": 0.3945,
+      "step": 872
+    },
+    {
+      "epoch": 0.5603337612323491,
+      "grad_norm": 0.01361773069947958,
+      "learning_rate": 4.2697219803758746e-05,
+      "loss": 0.4043,
+      "step": 873
+    },
+    {
+      "epoch": 0.5609756097560976,
+      "grad_norm": 0.014319811947643757,
+      "learning_rate": 4.259439306086176e-05,
+      "loss": 0.3809,
+      "step": 874
+    },
+    {
+      "epoch": 0.561617458279846,
+      "grad_norm": 0.01249017659574747,
+      "learning_rate": 4.2491598331329154e-05,
+      "loss": 0.3906,
+      "step": 875
+    },
+    {
+      "epoch": 0.5622593068035944,
+      "grad_norm": 0.013194131664931774,
+      "learning_rate": 4.2388836059527634e-05,
+      "loss": 0.3906,
+      "step": 876
+    },
+    {
+      "epoch": 0.5629011553273428,
+      "grad_norm": 0.016752328723669052,
+      "learning_rate": 4.2286106689683605e-05,
+      "loss": 0.4082,
+      "step": 877
+    },
+    {
+      "epoch": 0.5635430038510911,
+      "grad_norm": 0.014421268366277218,
+      "learning_rate": 4.218341066588121e-05,
+      "loss": 0.3984,
+      "step": 878
+    },
+    {
+      "epoch": 0.5641848523748395,
+      "grad_norm": 0.012712839990854263,
+      "learning_rate": 4.2080748432060444e-05,
+      "loss": 0.3984,
+      "step": 879
+    },
+    {
+      "epoch": 0.5648267008985879,
+      "grad_norm": 0.016034195199608803,
+      "learning_rate": 4.19781204320153e-05,
+      "loss": 0.3945,
+      "step": 880
+    },
+    {
+      "epoch": 0.5654685494223364,
+      "grad_norm": 0.014965656213462353,
+      "learning_rate": 4.18755271093917e-05,
+      "loss": 0.3828,
+      "step": 881
+    },
+    {
+      "epoch": 0.5661103979460848,
+      "grad_norm": 0.013681135140359402,
+      "learning_rate": 4.1772968907685704e-05,
+      "loss": 0.3828,
+      "step": 882
+    },
+    {
+      "epoch": 0.5667522464698331,
+      "grad_norm": 0.012078775092959404,
+      "learning_rate": 4.1670446270241545e-05,
+      "loss": 0.4062,
+      "step": 883
+    },
+    {
+      "epoch": 0.5673940949935815,
+      "grad_norm": 0.013224365189671516,
+      "learning_rate": 4.156795964024967e-05,
+      "loss": 0.4004,
+      "step": 884
+    },
+    {
+      "epoch": 0.5680359435173299,
+      "grad_norm": 0.012720691040158272,
+      "learning_rate": 4.1465509460744963e-05,
+      "loss": 0.3555,
+      "step": 885
+    },
+    {
+      "epoch": 0.5686777920410783,
+      "grad_norm": 0.015442232601344585,
+      "learning_rate": 4.1363096174604654e-05,
+      "loss": 0.4023,
+      "step": 886
+    },
+    {
+      "epoch": 0.5693196405648266,
+      "grad_norm": 0.013440394774079323,
+      "learning_rate": 4.126072022454655e-05,
+      "loss": 0.3789,
+      "step": 887
+    },
+    {
+      "epoch": 0.5699614890885751,
+      "grad_norm": 0.014425133354961872,
+      "learning_rate": 4.115838205312701e-05,
+      "loss": 0.4102,
+      "step": 888
+    },
+    {
+      "epoch": 0.5706033376123235,
+      "grad_norm": 0.012587243691086769,
+      "learning_rate": 4.105608210273909e-05,
+      "loss": 0.4004,
+      "step": 889
+    },
+    {
+      "epoch": 0.5712451861360719,
+      "grad_norm": 0.01208252739161253,
+      "learning_rate": 4.0953820815610636e-05,
+      "loss": 0.375,
+      "step": 890
+    },
+    {
+      "epoch": 0.5718870346598203,
+      "grad_norm": 0.012063434347510338,
+      "learning_rate": 4.085159863380239e-05,
+      "loss": 0.3789,
+      "step": 891
+    },
+    {
+      "epoch": 0.5725288831835686,
+      "grad_norm": 0.012063032016158104,
+      "learning_rate": 4.0749415999205995e-05,
+      "loss": 0.4141,
+      "step": 892
+    },
+    {
+      "epoch": 0.573170731707317,
+      "grad_norm": 0.012117046862840652,
+      "learning_rate": 4.064727335354215e-05,
+      "loss": 0.4121,
+      "step": 893
+    },
+    {
+      "epoch": 0.5738125802310655,
+      "grad_norm": 0.014085904695093632,
+      "learning_rate": 4.054517113835869e-05,
+      "loss": 0.3594,
+      "step": 894
+    },
+    {
+      "epoch": 0.5744544287548139,
+      "grad_norm": 0.014206532388925552,
+      "learning_rate": 4.0443109795028665e-05,
+      "loss": 0.375,
+      "step": 895
+    },
+    {
+      "epoch": 0.5750962772785623,
+      "grad_norm": 0.014762453734874725,
+      "learning_rate": 4.03410897647485e-05,
+      "loss": 0.3633,
+      "step": 896
+    },
+    {
+      "epoch": 0.5757381258023107,
+      "grad_norm": 0.014051433652639389,
+      "learning_rate": 4.023911148853596e-05,
+      "loss": 0.4141,
+      "step": 897
+    },
+    {
+      "epoch": 0.576379974326059,
+      "grad_norm": 0.013479063287377357,
+      "learning_rate": 4.013717540722833e-05,
+      "loss": 0.418,
+      "step": 898
+    },
+    {
+      "epoch": 0.5770218228498074,
+      "grad_norm": 0.016634047031402588,
+      "learning_rate": 4.0035281961480496e-05,
+      "loss": 0.4219,
+      "step": 899
+    },
+    {
+      "epoch": 0.5776636713735558,
+      "grad_norm": 0.015428063459694386,
+      "learning_rate": 3.993343159176307e-05,
+      "loss": 0.4004,
+      "step": 900
+    },
+    {
+      "epoch": 0.5783055198973043,
+      "grad_norm": 0.013261674903333187,
+      "learning_rate": 3.983162473836038e-05,
+      "loss": 0.3945,
+      "step": 901
+    },
+    {
+      "epoch": 0.5789473684210527,
+      "grad_norm": 0.015347424894571304,
+      "learning_rate": 3.972986184136873e-05,
+      "loss": 0.3867,
+      "step": 902
+    },
+    {
+      "epoch": 0.579589216944801,
+      "grad_norm": 0.014665666036307812,
+      "learning_rate": 3.962814334069434e-05,
+      "loss": 0.3828,
+      "step": 903
+    },
+    {
+      "epoch": 0.5802310654685494,
+      "grad_norm": 0.01409863866865635,
+      "learning_rate": 3.952646967605152e-05,
+      "loss": 0.3906,
+      "step": 904
+    },
+    {
+      "epoch": 0.5808729139922978,
+      "grad_norm": 0.012653130106627941,
+      "learning_rate": 3.9424841286960784e-05,
+      "loss": 0.4102,
+      "step": 905
+    },
+    {
+      "epoch": 0.5815147625160462,
+      "grad_norm": 0.015162872150540352,
+      "learning_rate": 3.9323258612746916e-05,
+      "loss": 0.3887,
+      "step": 906
+    },
+    {
+      "epoch": 0.5821566110397947,
+      "grad_norm": 0.012417521327733994,
+      "learning_rate": 3.922172209253705e-05,
+      "loss": 0.4121,
+      "step": 907
+    },
+    {
+      "epoch": 0.582798459563543,
+      "grad_norm": 0.012824651785194874,
+      "learning_rate": 3.912023216525887e-05,
+      "loss": 0.3906,
+      "step": 908
+    },
+    {
+      "epoch": 0.5834403080872914,
+      "grad_norm": 0.013151159510016441,
+      "learning_rate": 3.90187892696386e-05,
+      "loss": 0.4062,
+      "step": 909
+    },
+    {
+      "epoch": 0.5840821566110398,
+      "grad_norm": 0.012618138454854488,
+      "learning_rate": 3.8917393844199156e-05,
+      "loss": 0.4102,
+      "step": 910
+    },
+    {
+      "epoch": 0.5847240051347882,
+      "grad_norm": 0.013840214349329472,
+      "learning_rate": 3.881604632725825e-05,
+      "loss": 0.3809,
+      "step": 911
+    },
+    {
+      "epoch": 0.5853658536585366,
+      "grad_norm": 0.016128696501255035,
+      "learning_rate": 3.8714747156926466e-05,
+      "loss": 0.3984,
+      "step": 912
+    },
+    {
+      "epoch": 0.5860077021822849,
+      "grad_norm": 0.013604899868369102,
+      "learning_rate": 3.8613496771105464e-05,
+      "loss": 0.3867,
+      "step": 913
+    },
+    {
+      "epoch": 0.5866495507060334,
+      "grad_norm": 0.0137571319937706,
+      "learning_rate": 3.8512295607485965e-05,
+      "loss": 0.3906,
+      "step": 914
+    },
+    {
+      "epoch": 0.5872913992297818,
+      "grad_norm": 0.022923102602362633,
+      "learning_rate": 3.8411144103545904e-05,
+      "loss": 0.4082,
+      "step": 915
+    },
+    {
+      "epoch": 0.5879332477535302,
+      "grad_norm": 0.013792094774544239,
+      "learning_rate": 3.8310042696548565e-05,
+      "loss": 0.418,
+      "step": 916
+    },
+    {
+      "epoch": 0.5885750962772786,
+      "grad_norm": 0.013917778618633747,
+      "learning_rate": 3.8208991823540646e-05,
+      "loss": 0.3691,
+      "step": 917
+    },
+    {
+      "epoch": 0.5892169448010269,
+      "grad_norm": 0.012257051654160023,
+      "learning_rate": 3.8107991921350416e-05,
+      "loss": 0.3945,
+      "step": 918
+    },
+    {
+      "epoch": 0.5898587933247753,
+      "grad_norm": 0.015123442746698856,
+      "learning_rate": 3.8007043426585824e-05,
+      "loss": 0.3945,
+      "step": 919
+    },
+    {
+      "epoch": 0.5905006418485238,
+      "grad_norm": 0.011402972973883152,
+      "learning_rate": 3.7906146775632554e-05,
+      "loss": 0.4004,
+      "step": 920
+    },
+    {
+      "epoch": 0.5911424903722722,
+      "grad_norm": 0.014657670632004738,
+      "learning_rate": 3.780530240465217e-05,
+      "loss": 0.3945,
+      "step": 921
+    },
+    {
+      "epoch": 0.5917843388960206,
+      "grad_norm": 0.011596706695854664,
+      "learning_rate": 3.770451074958029e-05,
+      "loss": 0.3711,
+      "step": 922
+    },
+    {
+      "epoch": 0.5924261874197689,
+      "grad_norm": 0.012447874061763287,
+      "learning_rate": 3.760377224612457e-05,
+      "loss": 0.3945,
+      "step": 923
+    },
+    {
+      "epoch": 0.5930680359435173,
+      "grad_norm": 0.012190941721200943,
+      "learning_rate": 3.7503087329763e-05,
+      "loss": 0.3984,
+      "step": 924
+    },
+    {
+      "epoch": 0.5937098844672657,
+      "grad_norm": 0.012762784026563168,
+      "learning_rate": 3.740245643574184e-05,
+      "loss": 0.3789,
+      "step": 925
+    },
+    {
+      "epoch": 0.5943517329910141,
+      "grad_norm": 0.013606292195618153,
+      "learning_rate": 3.730187999907383e-05,
+      "loss": 0.3652,
+      "step": 926
+    },
+    {
+      "epoch": 0.5949935815147626,
+      "grad_norm": 0.013056452386081219,
+      "learning_rate": 3.720135845453634e-05,
+      "loss": 0.3828,
+      "step": 927
+    },
+    {
+      "epoch": 0.5956354300385109,
+      "grad_norm": 0.012558141723275185,
+      "learning_rate": 3.7100892236669415e-05,
+      "loss": 0.3867,
+      "step": 928
+    },
+    {
+      "epoch": 0.5962772785622593,
+      "grad_norm": 0.013830955140292645,
+      "learning_rate": 3.700048177977391e-05,
+      "loss": 0.4102,
+      "step": 929
+    },
+    {
+      "epoch": 0.5969191270860077,
+      "grad_norm": 0.013005700893700123,
+      "learning_rate": 3.690012751790972e-05,
+      "loss": 0.3789,
+      "step": 930
+    },
+    {
+      "epoch": 0.5975609756097561,
+      "grad_norm": 0.014384974725544453,
+      "learning_rate": 3.679982988489371e-05,
+      "loss": 0.3945,
+      "step": 931
+    },
+    {
+      "epoch": 0.5982028241335045,
+      "grad_norm": 0.011707819066941738,
+      "learning_rate": 3.6699589314298026e-05,
+      "loss": 0.3926,
+      "step": 932
+    },
+    {
+      "epoch": 0.5988446726572529,
+      "grad_norm": 0.015173131600022316,
+      "learning_rate": 3.65994062394481e-05,
+      "loss": 0.3945,
+      "step": 933
+    },
+    {
+      "epoch": 0.5994865211810013,
+      "grad_norm": 0.01235125120729208,
+      "learning_rate": 3.649928109342082e-05,
+      "loss": 0.3867,
+      "step": 934
+    },
+    {
+      "epoch": 0.6001283697047497,
+      "grad_norm": 0.016087165102362633,
+      "learning_rate": 3.639921430904268e-05,
+      "loss": 0.3984,
+      "step": 935
+    },
+    {
+      "epoch": 0.6007702182284981,
+      "grad_norm": 0.013020205311477184,
+      "learning_rate": 3.629920631888787e-05,
+      "loss": 0.4062,
+      "step": 936
+    },
+    {
+      "epoch": 0.6014120667522465,
+      "grad_norm": 0.01325344666838646,
+      "learning_rate": 3.619925755527642e-05,
+      "loss": 0.3926,
+      "step": 937
+    },
+    {
+      "epoch": 0.6020539152759948,
+      "grad_norm": 0.014157435856759548,
+      "learning_rate": 3.609936845027234e-05,
+      "loss": 0.3984,
+      "step": 938
+    },
+    {
+      "epoch": 0.6026957637997432,
+      "grad_norm": 0.01504555344581604,
+      "learning_rate": 3.5999539435681717e-05,
+      "loss": 0.4062,
+      "step": 939
+    },
+    {
+      "epoch": 0.6033376123234917,
+      "grad_norm": 0.012598301284015179,
+      "learning_rate": 3.5899770943050924e-05,
+      "loss": 0.3945,
+      "step": 940
+    },
+    {
+      "epoch": 0.6039794608472401,
+      "grad_norm": 0.012591421604156494,
+      "learning_rate": 3.5800063403664666e-05,
+      "loss": 0.3867,
+      "step": 941
+    },
+    {
+      "epoch": 0.6046213093709885,
+      "grad_norm": 0.013082444667816162,
+      "learning_rate": 3.5700417248544174e-05,
+      "loss": 0.3828,
+      "step": 942
+    },
+    {
+      "epoch": 0.6052631578947368,
+      "grad_norm": 0.013451972976326942,
+      "learning_rate": 3.560083290844534e-05,
+      "loss": 0.3945,
+      "step": 943
+    },
+    {
+      "epoch": 0.6059050064184852,
+      "grad_norm": 0.014088460244238377,
+      "learning_rate": 3.5501310813856766e-05,
+      "loss": 0.4023,
+      "step": 944
+    },
+    {
+      "epoch": 0.6065468549422336,
+      "grad_norm": 0.013572737574577332,
+      "learning_rate": 3.5401851394998084e-05,
+      "loss": 0.3809,
+      "step": 945
+    },
+    {
+      "epoch": 0.6071887034659821,
+      "grad_norm": 0.013333848677575588,
+      "learning_rate": 3.530245508181789e-05,
+      "loss": 0.3867,
+      "step": 946
+    },
+    {
+      "epoch": 0.6078305519897305,
+      "grad_norm": 0.015051721595227718,
+      "learning_rate": 3.5203122303992056e-05,
+      "loss": 0.3906,
+      "step": 947
+    },
+    {
+      "epoch": 0.6084724005134788,
+      "grad_norm": 0.01500693429261446,
+      "learning_rate": 3.5103853490921756e-05,
+      "loss": 0.3828,
+      "step": 948
+    },
+    {
+      "epoch": 0.6091142490372272,
+      "grad_norm": 0.012870650738477707,
+      "learning_rate": 3.5004649071731664e-05,
+      "loss": 0.3867,
+      "step": 949
+    },
+    {
+      "epoch": 0.6097560975609756,
+      "grad_norm": 0.012895474210381508,
+      "learning_rate": 3.4905509475268104e-05,
+      "loss": 0.3594,
+      "step": 950
+    },
+    {
+      "epoch": 0.610397946084724,
+      "grad_norm": 0.016247158870100975,
+      "learning_rate": 3.4806435130097134e-05,
+      "loss": 0.3926,
+      "step": 951
+    },
+    {
+      "epoch": 0.6110397946084724,
+      "grad_norm": 0.013967982493340969,
+      "learning_rate": 3.470742646450282e-05,
+      "loss": 0.3867,
+      "step": 952
+    },
+    {
+      "epoch": 0.6116816431322208,
+      "grad_norm": 0.01287417858839035,
+      "learning_rate": 3.4608483906485254e-05,
+      "loss": 0.3945,
+      "step": 953
+    },
+    {
+      "epoch": 0.6123234916559692,
+      "grad_norm": 0.013941469602286816,
+      "learning_rate": 3.450960788375872e-05,
+      "loss": 0.3691,
+      "step": 954
+    },
+    {
+      "epoch": 0.6129653401797176,
+      "grad_norm": 0.017454208806157112,
+      "learning_rate": 3.4410798823749964e-05,
+      "loss": 0.3867,
+      "step": 955
+    },
+    {
+      "epoch": 0.613607188703466,
+      "grad_norm": 0.013686290942132473,
+      "learning_rate": 3.4312057153596186e-05,
+      "loss": 0.3789,
+      "step": 956
+    },
+    {
+      "epoch": 0.6142490372272144,
+      "grad_norm": 0.013105151243507862,
+      "learning_rate": 3.42133833001433e-05,
+      "loss": 0.3672,
+      "step": 957
+    },
+    {
+      "epoch": 0.6148908857509627,
+      "grad_norm": 0.012791011482477188,
+      "learning_rate": 3.411477768994409e-05,
+      "loss": 0.373,
+      "step": 958
+    },
+    {
+      "epoch": 0.6155327342747111,
+      "grad_norm": 0.01345115527510643,
+      "learning_rate": 3.4016240749256266e-05,
+      "loss": 0.3555,
+      "step": 959
+    },
+    {
+      "epoch": 0.6161745827984596,
+      "grad_norm": 0.012273910455405712,
+      "learning_rate": 3.391777290404077e-05,
+      "loss": 0.4004,
+      "step": 960
+    },
+    {
+      "epoch": 0.616816431322208,
+      "grad_norm": 0.013736803084611893,
+      "learning_rate": 3.381937457995977e-05,
+      "loss": 0.4062,
+      "step": 961
+    },
+    {
+      "epoch": 0.6174582798459564,
+      "grad_norm": 0.014002666808664799,
+      "learning_rate": 3.372104620237495e-05,
+      "loss": 0.3789,
+      "step": 962
+    },
+    {
+      "epoch": 0.6181001283697047,
+      "grad_norm": 0.01376549992710352,
+      "learning_rate": 3.362278819634563e-05,
+      "loss": 0.4004,
+      "step": 963
+    },
+    {
+      "epoch": 0.6187419768934531,
+      "grad_norm": 0.014606526121497154,
+      "learning_rate": 3.352460098662694e-05,
+      "loss": 0.3926,
+      "step": 964
+    },
+    {
+      "epoch": 0.6193838254172015,
+      "grad_norm": 0.013210095465183258,
+      "learning_rate": 3.342648499766791e-05,
+      "loss": 0.3711,
+      "step": 965
+    },
+    {
+      "epoch": 0.62002567394095,
+      "grad_norm": 0.012232472188770771,
+      "learning_rate": 3.3328440653609735e-05,
+      "loss": 0.3809,
+      "step": 966
+    },
+    {
+      "epoch": 0.6206675224646984,
+      "grad_norm": 0.015114069916307926,
+      "learning_rate": 3.323046837828388e-05,
+      "loss": 0.4453,
+      "step": 967
+    },
+    {
+      "epoch": 0.6213093709884467,
+      "grad_norm": 0.014265330508351326,
+      "learning_rate": 3.313256859521028e-05,
+      "loss": 0.4141,
+      "step": 968
+    },
+    {
+      "epoch": 0.6219512195121951,
+      "grad_norm": 0.012651746161282063,
+      "learning_rate": 3.303474172759552e-05,
+      "loss": 0.3789,
+      "step": 969
+    },
+    {
+      "epoch": 0.6225930680359435,
+      "grad_norm": 0.01358629297465086,
+      "learning_rate": 3.293698819833093e-05,
+      "loss": 0.3984,
+      "step": 970
+    },
+    {
+      "epoch": 0.6232349165596919,
+      "grad_norm": 0.012160924263298512,
+      "learning_rate": 3.2839308429990846e-05,
+      "loss": 0.3555,
+      "step": 971
+    },
+    {
+      "epoch": 0.6238767650834403,
+      "grad_norm": 0.012310308404266834,
+      "learning_rate": 3.274170284483071e-05,
+      "loss": 0.3906,
+      "step": 972
+    },
+    {
+      "epoch": 0.6245186136071887,
+      "grad_norm": 0.015214286744594574,
+      "learning_rate": 3.264417186478535e-05,
+      "loss": 0.3945,
+      "step": 973
+    },
+    {
+      "epoch": 0.6251604621309371,
+      "grad_norm": 0.013628803193569183,
+      "learning_rate": 3.254671591146699e-05,
+      "loss": 0.3945,
+      "step": 974
+    },
+    {
+      "epoch": 0.6258023106546855,
+      "grad_norm": 0.012002642266452312,
+      "learning_rate": 3.244933540616363e-05,
+      "loss": 0.4121,
+      "step": 975
+    },
+    {
+      "epoch": 0.6264441591784339,
+      "grad_norm": 0.012065154500305653,
+      "learning_rate": 3.235203076983704e-05,
+      "loss": 0.3555,
+      "step": 976
+    },
+    {
+      "epoch": 0.6270860077021823,
+      "grad_norm": 0.012343095615506172,
+      "learning_rate": 3.2254802423121045e-05,
+      "loss": 0.3828,
+      "step": 977
+    },
+    {
+      "epoch": 0.6277278562259306,
+      "grad_norm": 0.012293039821088314,
+      "learning_rate": 3.2157650786319694e-05,
+      "loss": 0.3984,
+      "step": 978
+    },
+    {
+      "epoch": 0.6283697047496791,
+      "grad_norm": 0.013871513307094574,
+      "learning_rate": 3.206057627940539e-05,
+      "loss": 0.3711,
+      "step": 979
+    },
+    {
+      "epoch": 0.6290115532734275,
+      "grad_norm": 0.014145727269351482,
+      "learning_rate": 3.196357932201717e-05,
+      "loss": 0.3867,
+      "step": 980
+    },
+    {
+      "epoch": 0.6296534017971759,
+      "grad_norm": 0.015412476845085621,
+      "learning_rate": 3.18666603334588e-05,
+      "loss": 0.3691,
+      "step": 981
+    },
+    {
+      "epoch": 0.6302952503209243,
+      "grad_norm": 0.011903384700417519,
+      "learning_rate": 3.176981973269701e-05,
+      "loss": 0.3672,
+      "step": 982
+    },
+    {
+      "epoch": 0.6309370988446726,
+      "grad_norm": 0.012548285536468029,
+      "learning_rate": 3.167305793835967e-05,
+      "loss": 0.3945,
+      "step": 983
+    },
+    {
+      "epoch": 0.631578947368421,
+      "grad_norm": 0.015092693269252777,
+      "learning_rate": 3.157637536873397e-05,
+      "loss": 0.4062,
+      "step": 984
+    },
+    {
+      "epoch": 0.6322207958921694,
+      "grad_norm": 0.012332094833254814,
+      "learning_rate": 3.147977244176461e-05,
+      "loss": 0.375,
+      "step": 985
+    },
+    {
+      "epoch": 0.6328626444159179,
+      "grad_norm": 0.012180821038782597,
+      "learning_rate": 3.138324957505207e-05,
+      "loss": 0.4004,
+      "step": 986
+    },
+    {
+      "epoch": 0.6335044929396663,
+      "grad_norm": 0.014315093867480755,
+      "learning_rate": 3.128680718585068e-05,
+      "loss": 0.4355,
+      "step": 987
+    },
+    {
+      "epoch": 0.6341463414634146,
+      "grad_norm": 0.012419315055012703,
+      "learning_rate": 3.1190445691066916e-05,
+      "loss": 0.375,
+      "step": 988
+    },
+    {
+      "epoch": 0.634788189987163,
+      "grad_norm": 0.012981019914150238,
+      "learning_rate": 3.1094165507257514e-05,
+      "loss": 0.3711,
+      "step": 989
+    },
+    {
+      "epoch": 0.6354300385109114,
+      "grad_norm": 0.014806430786848068,
+      "learning_rate": 3.099796705062773e-05,
+      "loss": 0.3789,
+      "step": 990
+    },
+    {
+      "epoch": 0.6360718870346598,
+      "grad_norm": 0.01672835648059845,
+      "learning_rate": 3.090185073702956e-05,
+      "loss": 0.4062,
+      "step": 991
+    },
+    {
+      "epoch": 0.6367137355584083,
+      "grad_norm": 0.013994491659104824,
+      "learning_rate": 3.080581698195989e-05,
+      "loss": 0.3867,
+      "step": 992
+    },
+    {
+      "epoch": 0.6373555840821566,
+      "grad_norm": 0.012993131764233112,
+      "learning_rate": 3.070986620055869e-05,
+      "loss": 0.375,
+      "step": 993
+    },
+    {
+      "epoch": 0.637997432605905,
+      "grad_norm": 0.016973624005913734,
+      "learning_rate": 3.0613998807607267e-05,
+      "loss": 0.3945,
+      "step": 994
+    },
+    {
+      "epoch": 0.6386392811296534,
+      "grad_norm": 0.013423938304185867,
+      "learning_rate": 3.051821521752647e-05,
+      "loss": 0.3945,
+      "step": 995
+    },
+    {
+      "epoch": 0.6392811296534018,
+      "grad_norm": 0.014576129615306854,
+      "learning_rate": 3.042251584437484e-05,
+      "loss": 0.373,
+      "step": 996
+    },
+    {
+      "epoch": 0.6399229781771502,
+      "grad_norm": 0.01368379034101963,
+      "learning_rate": 3.0326901101846905e-05,
+      "loss": 0.3984,
+      "step": 997
+    },
+    {
+      "epoch": 0.6405648267008985,
+      "grad_norm": 0.01375710591673851,
+      "learning_rate": 3.023137140327132e-05,
+      "loss": 0.375,
+      "step": 998
+    },
+    {
+      "epoch": 0.641206675224647,
+      "grad_norm": 0.013690740801393986,
+      "learning_rate": 3.0135927161609097e-05,
+      "loss": 0.3984,
+      "step": 999
+    },
+    {
+      "epoch": 0.6418485237483954,
+      "grad_norm": 0.01352259423583746,
+      "learning_rate": 3.0040568789451862e-05,
+      "loss": 0.3867,
+      "step": 1000
+    },
+    {
+      "epoch": 0.6424903722721438,
+      "grad_norm": 0.013356149196624756,
+      "learning_rate": 2.9945296699020027e-05,
+      "loss": 0.3438,
+      "step": 1001
+    },
+    {
+      "epoch": 0.6431322207958922,
+      "grad_norm": 0.013253252021968365,
+      "learning_rate": 2.9850111302160976e-05,
+      "loss": 0.4043,
+      "step": 1002
+    },
+    {
+      "epoch": 0.6437740693196405,
+      "grad_norm": 0.012618259526789188,
+      "learning_rate": 2.9755013010347434e-05,
+      "loss": 0.4023,
+      "step": 1003
+    },
+    {
+      "epoch": 0.6444159178433889,
+      "grad_norm": 0.013053328730165958,
+      "learning_rate": 2.9660002234675465e-05,
+      "loss": 0.3535,
+      "step": 1004
+    },
+    {
+      "epoch": 0.6450577663671374,
+      "grad_norm": 0.01481606811285019,
+      "learning_rate": 2.9565079385862903e-05,
+      "loss": 0.3594,
+      "step": 1005
+    },
+    {
+      "epoch": 0.6456996148908858,
+      "grad_norm": 0.012548556551337242,
+      "learning_rate": 2.9470244874247443e-05,
+      "loss": 0.3867,
+      "step": 1006
+    },
+    {
+      "epoch": 0.6463414634146342,
+      "grad_norm": 0.01402729656547308,
+      "learning_rate": 2.9375499109784886e-05,
+      "loss": 0.4023,
+      "step": 1007
+    },
+    {
+      "epoch": 0.6469833119383825,
+      "grad_norm": 0.014539182186126709,
+      "learning_rate": 2.928084250204749e-05,
+      "loss": 0.3984,
+      "step": 1008
+    },
+    {
+      "epoch": 0.6476251604621309,
+      "grad_norm": 0.015400463715195656,
+      "learning_rate": 2.918627546022199e-05,
+      "loss": 0.3828,
+      "step": 1009
+    },
+    {
+      "epoch": 0.6482670089858793,
+      "grad_norm": 0.016256695613265038,
+      "learning_rate": 2.9091798393107994e-05,
+      "loss": 0.377,
+      "step": 1010
+    },
+    {
+      "epoch": 0.6489088575096277,
+      "grad_norm": 0.012321251444518566,
+      "learning_rate": 2.8997411709116168e-05,
+      "loss": 0.3867,
+      "step": 1011
+    },
+    {
+      "epoch": 0.6495507060333762,
+      "grad_norm": 0.012926090508699417,
+      "learning_rate": 2.890311581626647e-05,
+      "loss": 0.3672,
+      "step": 1012
+    },
+    {
+      "epoch": 0.6501925545571245,
+      "grad_norm": 0.01560135930776596,
+      "learning_rate": 2.8808911122186293e-05,
+      "loss": 0.3906,
+      "step": 1013
+    },
+    {
+      "epoch": 0.6508344030808729,
+      "grad_norm": 0.01321703102439642,
+      "learning_rate": 2.871479803410896e-05,
+      "loss": 0.4219,
+      "step": 1014
+    },
+    {
+      "epoch": 0.6514762516046213,
+      "grad_norm": 0.013517164625227451,
+      "learning_rate": 2.8620776958871627e-05,
+      "loss": 0.375,
+      "step": 1015
+    },
+    {
+      "epoch": 0.6521181001283697,
+      "grad_norm": 0.013148191384971142,
+      "learning_rate": 2.852684830291378e-05,
+      "loss": 0.3633,
+      "step": 1016
+    },
+    {
+      "epoch": 0.6527599486521181,
+      "grad_norm": 0.01478677149862051,
+      "learning_rate": 2.84330124722754e-05,
+      "loss": 0.3848,
+      "step": 1017
+    },
+    {
+      "epoch": 0.6534017971758665,
+      "grad_norm": 0.013707499951124191,
+      "learning_rate": 2.8339269872595097e-05,
+      "loss": 0.3945,
+      "step": 1018
+    },
+    {
+      "epoch": 0.6540436456996149,
+      "grad_norm": 0.014768145978450775,
+      "learning_rate": 2.8245620909108618e-05,
+      "loss": 0.3906,
+      "step": 1019
+    },
+    {
+      "epoch": 0.6546854942233633,
+      "grad_norm": 0.01512923650443554,
+      "learning_rate": 2.8152065986646788e-05,
+      "loss": 0.3789,
+      "step": 1020
+    },
+    {
+      "epoch": 0.6553273427471117,
+      "grad_norm": 0.015050279907882214,
+      "learning_rate": 2.8058605509633974e-05,
+      "loss": 0.3984,
+      "step": 1021
+    },
+    {
+      "epoch": 0.6559691912708601,
+      "grad_norm": 0.012544546276330948,
+      "learning_rate": 2.7965239882086292e-05,
+      "loss": 0.3633,
+      "step": 1022
+    },
+    {
+      "epoch": 0.6566110397946084,
+      "grad_norm": 0.013999714516103268,
+      "learning_rate": 2.7871969507609747e-05,
+      "loss": 0.3887,
+      "step": 1023
+    },
+    {
+      "epoch": 0.6572528883183568,
+      "grad_norm": 0.012318492867052555,
+      "learning_rate": 2.7778794789398667e-05,
+      "loss": 0.3789,
+      "step": 1024
+    },
+    {
+      "epoch": 0.6578947368421053,
+      "grad_norm": 0.013553598895668983,
+      "learning_rate": 2.7685716130233842e-05,
+      "loss": 0.3828,
+      "step": 1025
+    },
+    {
+      "epoch": 0.6585365853658537,
+      "grad_norm": 0.01511886902153492,
+      "learning_rate": 2.759273393248081e-05,
+      "loss": 0.373,
+      "step": 1026
+    },
+    {
+      "epoch": 0.6591784338896021,
+      "grad_norm": 0.012835157103836536,
+      "learning_rate": 2.7499848598088136e-05,
+      "loss": 0.3633,
+      "step": 1027
+    },
+    {
+      "epoch": 0.6598202824133504,
+      "grad_norm": 0.012799869291484356,
+      "learning_rate": 2.7407060528585616e-05,
+      "loss": 0.3848,
+      "step": 1028
+    },
+    {
+      "epoch": 0.6604621309370988,
+      "grad_norm": 0.012743955478072166,
+      "learning_rate": 2.731437012508262e-05,
+      "loss": 0.3691,
+      "step": 1029
+    },
+    {
+      "epoch": 0.6611039794608472,
+      "grad_norm": 0.014323204755783081,
+      "learning_rate": 2.7221777788266324e-05,
+      "loss": 0.3828,
+      "step": 1030
+    },
+    {
+      "epoch": 0.6617458279845956,
+      "grad_norm": 0.015074282884597778,
+      "learning_rate": 2.712928391839996e-05,
+      "loss": 0.4043,
+      "step": 1031
+    },
+    {
+      "epoch": 0.6623876765083441,
+      "grad_norm": 0.01439194567501545,
+      "learning_rate": 2.7036888915321136e-05,
+      "loss": 0.4062,
+      "step": 1032
+    },
+    {
+      "epoch": 0.6630295250320924,
+      "grad_norm": 0.013209599070250988,
+      "learning_rate": 2.6944593178440003e-05,
+      "loss": 0.3828,
+      "step": 1033
+    },
+    {
+      "epoch": 0.6636713735558408,
+      "grad_norm": 0.013193206861615181,
+      "learning_rate": 2.685239710673766e-05,
+      "loss": 0.3555,
+      "step": 1034
+    },
+    {
+      "epoch": 0.6643132220795892,
+      "grad_norm": 0.012524771504104137,
+      "learning_rate": 2.676030109876434e-05,
+      "loss": 0.4062,
+      "step": 1035
+    },
+    {
+      "epoch": 0.6649550706033376,
+      "grad_norm": 0.012812083587050438,
+      "learning_rate": 2.666830555263774e-05,
+      "loss": 0.3945,
+      "step": 1036
+    },
+    {
+      "epoch": 0.665596919127086,
+      "grad_norm": 0.014166711829602718,
+      "learning_rate": 2.6576410866041274e-05,
+      "loss": 0.3789,
+      "step": 1037
+    },
+    {
+      "epoch": 0.6662387676508345,
+      "grad_norm": 0.013674836605787277,
+      "learning_rate": 2.6484617436222293e-05,
+      "loss": 0.3672,
+      "step": 1038
+    },
+    {
+      "epoch": 0.6668806161745828,
+      "grad_norm": 0.012029345147311687,
+      "learning_rate": 2.639292565999051e-05,
+      "loss": 0.3711,
+      "step": 1039
+    },
+    {
+      "epoch": 0.6675224646983312,
+      "grad_norm": 0.015155037865042686,
+      "learning_rate": 2.6301335933716176e-05,
+      "loss": 0.4102,
+      "step": 1040
+    },
+    {
+      "epoch": 0.6681643132220796,
+      "grad_norm": 0.011492251418530941,
+      "learning_rate": 2.620984865332834e-05,
+      "loss": 0.3594,
+      "step": 1041
+    },
+    {
+      "epoch": 0.668806161745828,
+      "grad_norm": 0.012895763851702213,
+      "learning_rate": 2.611846421431331e-05,
+      "loss": 0.3926,
+      "step": 1042
+    },
+    {
+      "epoch": 0.6694480102695763,
+      "grad_norm": 0.013804377056658268,
+      "learning_rate": 2.6027183011712708e-05,
+      "loss": 0.3867,
+      "step": 1043
+    },
+    {
+      "epoch": 0.6700898587933247,
+      "grad_norm": 0.015160754323005676,
+      "learning_rate": 2.593600544012196e-05,
+      "loss": 0.3906,
+      "step": 1044
+    },
+    {
+      "epoch": 0.6707317073170732,
+      "grad_norm": 0.013987107202410698,
+      "learning_rate": 2.5844931893688473e-05,
+      "loss": 0.3711,
+      "step": 1045
+    },
+    {
+      "epoch": 0.6713735558408216,
+      "grad_norm": 0.01572549343109131,
+      "learning_rate": 2.575396276610994e-05,
+      "loss": 0.3906,
+      "step": 1046
+    },
+    {
+      "epoch": 0.67201540436457,
+      "grad_norm": 0.013872017152607441,
+      "learning_rate": 2.5663098450632762e-05,
+      "loss": 0.3594,
+      "step": 1047
+    },
+    {
+      "epoch": 0.6726572528883183,
+      "grad_norm": 0.013739910908043385,
+      "learning_rate": 2.557233934005015e-05,
+      "loss": 0.3906,
+      "step": 1048
+    },
+    {
+      "epoch": 0.6732991014120667,
+      "grad_norm": 0.013132906518876553,
+      "learning_rate": 2.548168582670058e-05,
+      "loss": 0.3789,
+      "step": 1049
+    },
+    {
+      "epoch": 0.6739409499358151,
+      "grad_norm": 0.013742578215897083,
+      "learning_rate": 2.5391138302466062e-05,
+      "loss": 0.3984,
+      "step": 1050
+    },
+    {
+      "epoch": 0.6745827984595636,
+      "grad_norm": 0.013859684579074383,
+      "learning_rate": 2.5300697158770365e-05,
+      "loss": 0.3711,
+      "step": 1051
+    },
+    {
+      "epoch": 0.675224646983312,
+      "grad_norm": 0.012405592948198318,
+      "learning_rate": 2.5210362786577452e-05,
+      "loss": 0.4043,
+      "step": 1052
+    },
+    {
+      "epoch": 0.6758664955070603,
+      "grad_norm": 0.012248203158378601,
+      "learning_rate": 2.5120135576389715e-05,
+      "loss": 0.3867,
+      "step": 1053
+    },
+    {
+      "epoch": 0.6765083440308087,
+      "grad_norm": 0.01508716307580471,
+      "learning_rate": 2.503001591824628e-05,
+      "loss": 0.3789,
+      "step": 1054
+    },
+    {
+      "epoch": 0.6771501925545571,
+      "grad_norm": 0.013256301172077656,
+      "learning_rate": 2.4940004201721384e-05,
+      "loss": 0.4141,
+      "step": 1055
+    },
+    {
+      "epoch": 0.6777920410783055,
+      "grad_norm": 0.01401185430586338,
+      "learning_rate": 2.4850100815922577e-05,
+      "loss": 0.4004,
+      "step": 1056
+    },
+    {
+      "epoch": 0.6784338896020539,
+      "grad_norm": 0.012873097322881222,
+      "learning_rate": 2.476030614948917e-05,
+      "loss": 0.3809,
+      "step": 1057
+    },
+    {
+      "epoch": 0.6790757381258024,
+      "grad_norm": 0.01284098532050848,
+      "learning_rate": 2.4670620590590482e-05,
+      "loss": 0.3867,
+      "step": 1058
+    },
+    {
+      "epoch": 0.6797175866495507,
+      "grad_norm": 0.011478393338620663,
+      "learning_rate": 2.4581044526924175e-05,
+      "loss": 0.4102,
+      "step": 1059
+    },
+    {
+      "epoch": 0.6803594351732991,
+      "grad_norm": 0.0136007871478796,
+      "learning_rate": 2.4491578345714587e-05,
+      "loss": 0.3965,
+      "step": 1060
+    },
+    {
+      "epoch": 0.6810012836970475,
+      "grad_norm": 0.012089556083083153,
+      "learning_rate": 2.4402222433711008e-05,
+      "loss": 0.3809,
+      "step": 1061
+    },
+    {
+      "epoch": 0.6816431322207959,
+      "grad_norm": 0.013884073123335838,
+      "learning_rate": 2.4312977177186095e-05,
+      "loss": 0.3906,
+      "step": 1062
+    },
+    {
+      "epoch": 0.6822849807445442,
+      "grad_norm": 0.013468881137669086,
+      "learning_rate": 2.422384296193415e-05,
+      "loss": 0.4102,
+      "step": 1063
+    },
+    {
+      "epoch": 0.6829268292682927,
+      "grad_norm": 0.012956452555954456,
+      "learning_rate": 2.4134820173269456e-05,
+      "loss": 0.3652,
+      "step": 1064
+    },
+    {
+      "epoch": 0.6835686777920411,
+      "grad_norm": 0.012719823978841305,
+      "learning_rate": 2.4045909196024624e-05,
+      "loss": 0.3984,
+      "step": 1065
+    },
+    {
+      "epoch": 0.6842105263157895,
+      "grad_norm": 0.013100896961987019,
+      "learning_rate": 2.3957110414548874e-05,
+      "loss": 0.373,
+      "step": 1066
+    },
+    {
+      "epoch": 0.6848523748395379,
+      "grad_norm": 0.013353249989449978,
+      "learning_rate": 2.3868424212706476e-05,
+      "loss": 0.377,
+      "step": 1067
+    },
+    {
+      "epoch": 0.6854942233632862,
+      "grad_norm": 0.012005530297756195,
+      "learning_rate": 2.3779850973875036e-05,
+      "loss": 0.3789,
+      "step": 1068
+    },
+    {
+      "epoch": 0.6861360718870346,
+      "grad_norm": 0.012415427714586258,
+      "learning_rate": 2.369139108094375e-05,
+      "loss": 0.3438,
+      "step": 1069
+    },
+    {
+      "epoch": 0.686777920410783,
+      "grad_norm": 0.012416977435350418,
+      "learning_rate": 2.3603044916311963e-05,
+      "loss": 0.373,
+      "step": 1070
+    },
+    {
+      "epoch": 0.6874197689345315,
+      "grad_norm": 0.012292545288801193,
+      "learning_rate": 2.3514812861887327e-05,
+      "loss": 0.375,
+      "step": 1071
+    },
+    {
+      "epoch": 0.6880616174582799,
+      "grad_norm": 0.012595480307936668,
+      "learning_rate": 2.3426695299084173e-05,
+      "loss": 0.3828,
+      "step": 1072
+    },
+    {
+      "epoch": 0.6887034659820283,
+      "grad_norm": 0.012537018395960331,
+      "learning_rate": 2.3338692608821982e-05,
+      "loss": 0.3711,
+      "step": 1073
+    },
+    {
+      "epoch": 0.6893453145057766,
+      "grad_norm": 0.011396169662475586,
+      "learning_rate": 2.325080517152356e-05,
+      "loss": 0.3887,
+      "step": 1074
+    },
+    {
+      "epoch": 0.689987163029525,
+      "grad_norm": 0.012768317945301533,
+      "learning_rate": 2.3163033367113602e-05,
+      "loss": 0.3828,
+      "step": 1075
+    },
+    {
+      "epoch": 0.6906290115532734,
+      "grad_norm": 0.012074895203113556,
+      "learning_rate": 2.307537757501688e-05,
+      "loss": 0.3867,
+      "step": 1076
+    },
+    {
+      "epoch": 0.6912708600770219,
+      "grad_norm": 0.011272402480244637,
+      "learning_rate": 2.298783817415664e-05,
+      "loss": 0.3691,
+      "step": 1077
+    },
+    {
+      "epoch": 0.6919127086007703,
+      "grad_norm": 0.012021319009363651,
+      "learning_rate": 2.2900415542953035e-05,
+      "loss": 0.3574,
+      "step": 1078
+    },
+    {
+      "epoch": 0.6925545571245186,
+      "grad_norm": 0.013903750106692314,
+      "learning_rate": 2.281311005932139e-05,
+      "loss": 0.3887,
+      "step": 1079
+    },
+    {
+      "epoch": 0.693196405648267,
+      "grad_norm": 0.013219778425991535,
+      "learning_rate": 2.2725922100670644e-05,
+      "loss": 0.3867,
+      "step": 1080
+    },
+    {
+      "epoch": 0.6938382541720154,
+      "grad_norm": 0.013222387991845608,
+      "learning_rate": 2.263885204390174e-05,
+      "loss": 0.3633,
+      "step": 1081
+    },
+    {
+      "epoch": 0.6944801026957638,
+      "grad_norm": 0.015906382352113724,
+      "learning_rate": 2.255190026540585e-05,
+      "loss": 0.3867,
+      "step": 1082
+    },
+    {
+      "epoch": 0.6951219512195121,
+      "grad_norm": 0.013349148444831371,
+      "learning_rate": 2.246506714106294e-05,
+      "loss": 0.4141,
+      "step": 1083
+    },
+    {
+      "epoch": 0.6957637997432606,
+      "grad_norm": 0.016905279830098152,
+      "learning_rate": 2.2378353046239963e-05,
+      "loss": 0.3945,
+      "step": 1084
+    },
+    {
+      "epoch": 0.696405648267009,
+      "grad_norm": 0.011738389730453491,
+      "learning_rate": 2.22917583557894e-05,
+      "loss": 0.3887,
+      "step": 1085
+    },
+    {
+      "epoch": 0.6970474967907574,
+      "grad_norm": 0.013519138097763062,
+      "learning_rate": 2.220528344404752e-05,
+      "loss": 0.373,
+      "step": 1086
+    },
+    {
+      "epoch": 0.6976893453145058,
+      "grad_norm": 0.013963024131953716,
+      "learning_rate": 2.211892868483283e-05,
+      "loss": 0.3789,
+      "step": 1087
+    },
+    {
+      "epoch": 0.6983311938382541,
+      "grad_norm": 0.01420444156974554,
+      "learning_rate": 2.2032694451444424e-05,
+      "loss": 0.3633,
+      "step": 1088
+    },
+    {
+      "epoch": 0.6989730423620025,
+      "grad_norm": 0.012128225527703762,
+      "learning_rate": 2.1946581116660403e-05,
+      "loss": 0.3555,
+      "step": 1089
+    },
+    {
+      "epoch": 0.699614890885751,
+      "grad_norm": 0.012816647067666054,
+      "learning_rate": 2.186058905273618e-05,
+      "loss": 0.3789,
+      "step": 1090
+    },
+    {
+      "epoch": 0.7002567394094994,
+      "grad_norm": 0.013052361086010933,
+      "learning_rate": 2.1774718631402987e-05,
+      "loss": 0.3906,
+      "step": 1091
+    },
+    {
+      "epoch": 0.7008985879332478,
+      "grad_norm": 0.012910526245832443,
+      "learning_rate": 2.16889702238662e-05,
+      "loss": 0.3984,
+      "step": 1092
+    },
+    {
+      "epoch": 0.7015404364569962,
+      "grad_norm": 0.014237569645047188,
+      "learning_rate": 2.1603344200803743e-05,
+      "loss": 0.4062,
+      "step": 1093
+    },
+    {
+      "epoch": 0.7021822849807445,
+      "grad_norm": 0.012514214962720871,
+      "learning_rate": 2.151784093236452e-05,
+      "loss": 0.3711,
+      "step": 1094
+    },
+    {
+      "epoch": 0.7028241335044929,
+      "grad_norm": 0.011893239803612232,
+      "learning_rate": 2.1432460788166704e-05,
+      "loss": 0.3555,
+      "step": 1095
+    },
+    {
+      "epoch": 0.7034659820282413,
+      "grad_norm": 0.013944214209914207,
+      "learning_rate": 2.134720413729631e-05,
+      "loss": 0.3828,
+      "step": 1096
+    },
+    {
+      "epoch": 0.7041078305519898,
+      "grad_norm": 0.012904972769320011,
+      "learning_rate": 2.1262071348305425e-05,
+      "loss": 0.3828,
+      "step": 1097
+    },
+    {
+      "epoch": 0.7047496790757382,
+      "grad_norm": 0.014260875061154366,
+      "learning_rate": 2.1177062789210796e-05,
+      "loss": 0.3828,
+      "step": 1098
+    },
+    {
+      "epoch": 0.7053915275994865,
+      "grad_norm": 0.012384374625980854,
+      "learning_rate": 2.109217882749208e-05,
+      "loss": 0.3965,
+      "step": 1099
+    },
+    {
+      "epoch": 0.7060333761232349,
+      "grad_norm": 0.013365059159696102,
+      "learning_rate": 2.1007419830090306e-05,
+      "loss": 0.3594,
+      "step": 1100
+    },
+    {
+      "epoch": 0.7066752246469833,
+      "grad_norm": 0.015220331028103828,
+      "learning_rate": 2.0922786163406337e-05,
+      "loss": 0.3945,
+      "step": 1101
+    },
+    {
+      "epoch": 0.7073170731707317,
+      "grad_norm": 0.012258613482117653,
+      "learning_rate": 2.0838278193299237e-05,
+      "loss": 0.3535,
+      "step": 1102
+    },
+    {
+      "epoch": 0.70795892169448,
+      "grad_norm": 0.012427720241248608,
+      "learning_rate": 2.0753896285084694e-05,
+      "loss": 0.3848,
+      "step": 1103
+    },
+    {
+      "epoch": 0.7086007702182285,
+      "grad_norm": 0.011899521574378014,
+      "learning_rate": 2.0669640803533474e-05,
+      "loss": 0.3965,
+      "step": 1104
+    },
+    {
+      "epoch": 0.7092426187419769,
+      "grad_norm": 0.012814830057322979,
+      "learning_rate": 2.058551211286977e-05,
+      "loss": 0.3613,
+      "step": 1105
+    },
+    {
+      "epoch": 0.7098844672657253,
+      "grad_norm": 0.015035499818623066,
+      "learning_rate": 2.0501510576769722e-05,
+      "loss": 0.3535,
+      "step": 1106
+    },
+    {
+      "epoch": 0.7105263157894737,
+      "grad_norm": 0.016760557889938354,
+      "learning_rate": 2.04176365583598e-05,
+      "loss": 0.3984,
+      "step": 1107
+    },
+    {
+      "epoch": 0.711168164313222,
+      "grad_norm": 0.012370377779006958,
+      "learning_rate": 2.0333890420215164e-05,
+      "loss": 0.3691,
+      "step": 1108
+    },
+    {
+      "epoch": 0.7118100128369704,
+      "grad_norm": 0.013664168305695057,
+      "learning_rate": 2.025027252435829e-05,
+      "loss": 0.3984,
+      "step": 1109
+    },
+    {
+      "epoch": 0.7124518613607189,
+      "grad_norm": 0.01310388557612896,
+      "learning_rate": 2.0166783232257154e-05,
+      "loss": 0.3711,
+      "step": 1110
+    },
+    {
+      "epoch": 0.7130937098844673,
+      "grad_norm": 0.013425314798951149,
+      "learning_rate": 2.008342290482388e-05,
+      "loss": 0.3984,
+      "step": 1111
+    },
+    {
+      "epoch": 0.7137355584082157,
+      "grad_norm": 0.013783198781311512,
+      "learning_rate": 2.0000191902413073e-05,
+      "loss": 0.3945,
+      "step": 1112
+    },
+    {
+      "epoch": 0.714377406931964,
+      "grad_norm": 0.013114030472934246,
+      "learning_rate": 1.991709058482026e-05,
+      "loss": 0.3711,
+      "step": 1113
+    },
+    {
+      "epoch": 0.7150192554557124,
+      "grad_norm": 0.012237347662448883,
+      "learning_rate": 1.9834119311280392e-05,
+      "loss": 0.4043,
+      "step": 1114
+    },
+    {
+      "epoch": 0.7156611039794608,
+      "grad_norm": 0.01424240879714489,
+      "learning_rate": 1.9751278440466248e-05,
+      "loss": 0.3906,
+      "step": 1115
+    },
+    {
+      "epoch": 0.7163029525032092,
+      "grad_norm": 0.013711946085095406,
+      "learning_rate": 1.966856833048689e-05,
+      "loss": 0.3672,
+      "step": 1116
+    },
+    {
+      "epoch": 0.7169448010269577,
+      "grad_norm": 0.01330238301306963,
+      "learning_rate": 1.9585989338886147e-05,
+      "loss": 0.4102,
+      "step": 1117
+    },
+    {
+      "epoch": 0.7175866495507061,
+      "grad_norm": 0.013692202046513557,
+      "learning_rate": 1.950354182264098e-05,
+      "loss": 0.3965,
+      "step": 1118
+    },
+    {
+      "epoch": 0.7182284980744544,
+      "grad_norm": 0.013284241780638695,
+      "learning_rate": 1.942122613816006e-05,
+      "loss": 0.3613,
+      "step": 1119
+    },
+    {
+      "epoch": 0.7188703465982028,
+      "grad_norm": 0.012372693978250027,
+      "learning_rate": 1.9339042641282146e-05,
+      "loss": 0.3652,
+      "step": 1120
+    },
+    {
+      "epoch": 0.7195121951219512,
+      "grad_norm": 0.013219374231994152,
+      "learning_rate": 1.925699168727458e-05,
+      "loss": 0.3477,
+      "step": 1121
+    },
+    {
+      "epoch": 0.7201540436456996,
+      "grad_norm": 0.014225227758288383,
+      "learning_rate": 1.9175073630831736e-05,
+      "loss": 0.4023,
+      "step": 1122
+    },
+    {
+      "epoch": 0.7207958921694481,
+      "grad_norm": 0.01241610199213028,
+      "learning_rate": 1.9093288826073462e-05,
+      "loss": 0.3516,
+      "step": 1123
+    },
+    {
+      "epoch": 0.7214377406931964,
+      "grad_norm": 0.013540954329073429,
+      "learning_rate": 1.9011637626543617e-05,
+      "loss": 0.3906,
+      "step": 1124
+    },
+    {
+      "epoch": 0.7220795892169448,
+      "grad_norm": 0.015003510750830173,
+      "learning_rate": 1.8930120385208495e-05,
+      "loss": 0.3867,
+      "step": 1125
+    },
+    {
+      "epoch": 0.7227214377406932,
+      "grad_norm": 0.013304078951478004,
+      "learning_rate": 1.88487374544553e-05,
+      "loss": 0.3867,
+      "step": 1126
+    },
+    {
+      "epoch": 0.7233632862644416,
+      "grad_norm": 0.012056106701493263,
+      "learning_rate": 1.8767489186090654e-05,
+      "loss": 0.3887,
+      "step": 1127
+    },
+    {
+      "epoch": 0.72400513478819,
+      "grad_norm": 0.01176505722105503,
+      "learning_rate": 1.8686375931338997e-05,
+      "loss": 0.3945,
+      "step": 1128
+    },
+    {
+      "epoch": 0.7246469833119383,
+      "grad_norm": 0.013599003665149212,
+      "learning_rate": 1.8605398040841172e-05,
+      "loss": 0.3789,
+      "step": 1129
+    },
+    {
+      "epoch": 0.7252888318356868,
+      "grad_norm": 0.014635767787694931,
+      "learning_rate": 1.8524555864652865e-05,
+      "loss": 0.416,
+      "step": 1130
+    },
+    {
+      "epoch": 0.7259306803594352,
+      "grad_norm": 0.013412141241133213,
+      "learning_rate": 1.844384975224307e-05,
+      "loss": 0.4062,
+      "step": 1131
+    },
+    {
+      "epoch": 0.7265725288831836,
+      "grad_norm": 0.013095933943986893,
+      "learning_rate": 1.8363280052492617e-05,
+      "loss": 0.3711,
+      "step": 1132
+    },
+    {
+      "epoch": 0.727214377406932,
+      "grad_norm": 0.01511120330542326,
+      "learning_rate": 1.8282847113692593e-05,
+      "loss": 0.4023,
+      "step": 1133
+    },
+    {
+      "epoch": 0.7278562259306803,
+      "grad_norm": 0.012698481790721416,
+      "learning_rate": 1.820255128354294e-05,
+      "loss": 0.3828,
+      "step": 1134
+    },
+    {
+      "epoch": 0.7284980744544287,
+      "grad_norm": 0.013212742283940315,
+      "learning_rate": 1.8122392909150904e-05,
+      "loss": 0.375,
+      "step": 1135
+    },
+    {
+      "epoch": 0.7291399229781772,
+      "grad_norm": 0.012589886784553528,
+      "learning_rate": 1.8042372337029455e-05,
+      "loss": 0.3516,
+      "step": 1136
+    },
+    {
+      "epoch": 0.7297817715019256,
+      "grad_norm": 0.011949255131185055,
+      "learning_rate": 1.7962489913095987e-05,
+      "loss": 0.375,
+      "step": 1137
+    },
+    {
+      "epoch": 0.730423620025674,
+      "grad_norm": 0.012383786030113697,
+      "learning_rate": 1.7882745982670573e-05,
+      "loss": 0.3984,
+      "step": 1138
+    },
+    {
+      "epoch": 0.7310654685494223,
+      "grad_norm": 0.011663233861327171,
+      "learning_rate": 1.7803140890474674e-05,
+      "loss": 0.375,
+      "step": 1139
+    },
+    {
+      "epoch": 0.7317073170731707,
+      "grad_norm": 0.014100736938416958,
+      "learning_rate": 1.7723674980629572e-05,
+      "loss": 0.3711,
+      "step": 1140
+    },
+    {
+      "epoch": 0.7323491655969191,
+      "grad_norm": 0.012323318980634212,
+      "learning_rate": 1.7644348596654837e-05,
+      "loss": 0.3828,
+      "step": 1141
+    },
+    {
+      "epoch": 0.7329910141206675,
+      "grad_norm": 0.012874532490968704,
+      "learning_rate": 1.756516208146693e-05,
+      "loss": 0.3672,
+      "step": 1142
+    },
+    {
+      "epoch": 0.733632862644416,
+      "grad_norm": 0.013383268378674984,
+      "learning_rate": 1.7486115777377667e-05,
+      "loss": 0.3711,
+      "step": 1143
+    },
+    {
+      "epoch": 0.7342747111681643,
+      "grad_norm": 0.012120241299271584,
+      "learning_rate": 1.740721002609275e-05,
+      "loss": 0.3789,
+      "step": 1144
+    },
+    {
+      "epoch": 0.7349165596919127,
+      "grad_norm": 0.014149006456136703,
+      "learning_rate": 1.7328445168710323e-05,
+      "loss": 0.3926,
+      "step": 1145
+    },
+    {
+      "epoch": 0.7355584082156611,
+      "grad_norm": 0.012970617972314358,
+      "learning_rate": 1.7249821545719387e-05,
+      "loss": 0.4062,
+      "step": 1146
+    },
+    {
+      "epoch": 0.7362002567394095,
+      "grad_norm": 0.013903571292757988,
+      "learning_rate": 1.717133949699849e-05,
+      "loss": 0.3945,
+      "step": 1147
+    },
+    {
+      "epoch": 0.7368421052631579,
+      "grad_norm": 0.012635747902095318,
+      "learning_rate": 1.7092999361814126e-05,
+      "loss": 0.3867,
+      "step": 1148
+    },
+    {
+      "epoch": 0.7374839537869063,
+      "grad_norm": 0.012244436889886856,
+      "learning_rate": 1.7014801478819347e-05,
+      "loss": 0.3926,
+      "step": 1149
+    },
+    {
+      "epoch": 0.7381258023106547,
+      "grad_norm": 0.014593473635613918,
+      "learning_rate": 1.693674618605227e-05,
+      "loss": 0.4141,
+      "step": 1150
+    },
+    {
+      "epoch": 0.7387676508344031,
+      "grad_norm": 0.014035548083484173,
+      "learning_rate": 1.6858833820934567e-05,
+      "loss": 0.3848,
+      "step": 1151
+    },
+    {
+      "epoch": 0.7394094993581515,
+      "grad_norm": 0.013262953609228134,
+      "learning_rate": 1.6781064720270105e-05,
+      "loss": 0.3965,
+      "step": 1152
+    },
+    {
+      "epoch": 0.7400513478818999,
+      "grad_norm": 0.01288782898336649,
+      "learning_rate": 1.6703439220243423e-05,
+      "loss": 0.4102,
+      "step": 1153
+    },
+    {
+      "epoch": 0.7406931964056482,
+      "grad_norm": 0.011735158041119576,
+      "learning_rate": 1.6625957656418295e-05,
+      "loss": 0.373,
+      "step": 1154
+    },
+    {
+      "epoch": 0.7413350449293966,
+      "grad_norm": 0.012648727744817734,
+      "learning_rate": 1.6548620363736294e-05,
+      "loss": 0.3652,
+      "step": 1155
+    },
+    {
+      "epoch": 0.7419768934531451,
+      "grad_norm": 0.013695376925170422,
+      "learning_rate": 1.6471427676515288e-05,
+      "loss": 0.4004,
+      "step": 1156
+    },
+    {
+      "epoch": 0.7426187419768935,
+      "grad_norm": 0.015565620735287666,
+      "learning_rate": 1.6394379928448083e-05,
+      "loss": 0.3984,
+      "step": 1157
+    },
+    {
+      "epoch": 0.7432605905006419,
+      "grad_norm": 0.012965993955731392,
+      "learning_rate": 1.6317477452600917e-05,
+      "loss": 0.3809,
+      "step": 1158
+    },
+    {
+      "epoch": 0.7439024390243902,
+      "grad_norm": 0.013348723761737347,
+      "learning_rate": 1.624072058141202e-05,
+      "loss": 0.3828,
+      "step": 1159
+    },
+    {
+      "epoch": 0.7445442875481386,
+      "grad_norm": 0.013615473173558712,
+      "learning_rate": 1.616410964669025e-05,
+      "loss": 0.3516,
+      "step": 1160
+    },
+    {
+      "epoch": 0.745186136071887,
+      "grad_norm": 0.012235503643751144,
+      "learning_rate": 1.6087644979613514e-05,
+      "loss": 0.3711,
+      "step": 1161
+    },
+    {
+      "epoch": 0.7458279845956355,
+      "grad_norm": 0.01184153463691473,
+      "learning_rate": 1.6011326910727492e-05,
+      "loss": 0.3574,
+      "step": 1162
+    },
+    {
+      "epoch": 0.7464698331193839,
+      "grad_norm": 0.012317166663706303,
+      "learning_rate": 1.593515576994415e-05,
+      "loss": 0.3516,
+      "step": 1163
+    },
+    {
+      "epoch": 0.7471116816431322,
+      "grad_norm": 0.013108180835843086,
+      "learning_rate": 1.5859131886540214e-05,
+      "loss": 0.4004,
+      "step": 1164
+    },
+    {
+      "epoch": 0.7477535301668806,
+      "grad_norm": 0.013911323621869087,
+      "learning_rate": 1.578325558915598e-05,
+      "loss": 0.4062,
+      "step": 1165
+    },
+    {
+      "epoch": 0.748395378690629,
+      "grad_norm": 0.014530111104249954,
+      "learning_rate": 1.570752720579362e-05,
+      "loss": 0.3926,
+      "step": 1166
+    },
+    {
+      "epoch": 0.7490372272143774,
+      "grad_norm": 0.014330373145639896,
+      "learning_rate": 1.563194706381597e-05,
+      "loss": 0.3789,
+      "step": 1167
+    },
+    {
+      "epoch": 0.7496790757381258,
+      "grad_norm": 0.013744387775659561,
+      "learning_rate": 1.555651548994503e-05,
+      "loss": 0.3867,
+      "step": 1168
+    },
+    {
+      "epoch": 0.7503209242618742,
+      "grad_norm": 0.01154696848243475,
+      "learning_rate": 1.548123281026051e-05,
+      "loss": 0.3789,
+      "step": 1169
+    },
+    {
+      "epoch": 0.7509627727856226,
+      "grad_norm": 0.01333354040980339,
+      "learning_rate": 1.5406099350198544e-05,
+      "loss": 0.3789,
+      "step": 1170
+    },
+    {
+      "epoch": 0.751604621309371,
+      "grad_norm": 0.01610654406249523,
+      "learning_rate": 1.533111543455017e-05,
+      "loss": 0.3691,
+      "step": 1171
+    },
+    {
+      "epoch": 0.7522464698331194,
+      "grad_norm": 0.012678599916398525,
+      "learning_rate": 1.5256281387459975e-05,
+      "loss": 0.3438,
+      "step": 1172
+    },
+    {
+      "epoch": 0.7528883183568678,
+      "grad_norm": 0.013093999586999416,
+      "learning_rate": 1.5181597532424713e-05,
+      "loss": 0.3633,
+      "step": 1173
+    },
+    {
+      "epoch": 0.7535301668806161,
+      "grad_norm": 0.014720738865435123,
+      "learning_rate": 1.5107064192291808e-05,
+      "loss": 0.3906,
+      "step": 1174
+    },
+    {
+      "epoch": 0.7541720154043645,
+      "grad_norm": 0.011939354240894318,
+      "learning_rate": 1.5032681689258105e-05,
+      "loss": 0.3984,
+      "step": 1175
+    },
+    {
+      "epoch": 0.754813863928113,
+      "grad_norm": 0.014355121180415154,
+      "learning_rate": 1.495845034486837e-05,
+      "loss": 0.3633,
+      "step": 1176
+    },
+    {
+      "epoch": 0.7554557124518614,
+      "grad_norm": 0.013115710578858852,
+      "learning_rate": 1.4884370480013931e-05,
+      "loss": 0.418,
+      "step": 1177
+    },
+    {
+      "epoch": 0.7560975609756098,
+      "grad_norm": 0.012193549424409866,
+      "learning_rate": 1.481044241493132e-05,
+      "loss": 0.3906,
+      "step": 1178
+    },
+    {
+      "epoch": 0.7567394094993581,
+      "grad_norm": 0.013995474204421043,
+      "learning_rate": 1.4736666469200793e-05,
+      "loss": 0.3848,
+      "step": 1179
+    },
+    {
+      "epoch": 0.7573812580231065,
+      "grad_norm": 0.01327692624181509,
+      "learning_rate": 1.4663042961745083e-05,
+      "loss": 0.3457,
+      "step": 1180
+    },
+    {
+      "epoch": 0.7580231065468549,
+      "grad_norm": 0.013241326436400414,
+      "learning_rate": 1.4589572210827934e-05,
+      "loss": 0.3926,
+      "step": 1181
+    },
+    {
+      "epoch": 0.7586649550706034,
+      "grad_norm": 0.014227569103240967,
+      "learning_rate": 1.4516254534052736e-05,
+      "loss": 0.3945,
+      "step": 1182
+    },
+    {
+      "epoch": 0.7593068035943518,
+      "grad_norm": 0.012828252278268337,
+      "learning_rate": 1.444309024836119e-05,
+      "loss": 0.3594,
+      "step": 1183
+    },
+    {
+      "epoch": 0.7599486521181001,
+      "grad_norm": 0.01393461786210537,
+      "learning_rate": 1.4370079670031854e-05,
+      "loss": 0.3867,
+      "step": 1184
+    },
+    {
+      "epoch": 0.7605905006418485,
+      "grad_norm": 0.013498412445187569,
+      "learning_rate": 1.4297223114678887e-05,
+      "loss": 0.4082,
+      "step": 1185
+    },
+    {
+      "epoch": 0.7612323491655969,
+      "grad_norm": 0.013050911016762257,
+      "learning_rate": 1.4224520897250599e-05,
+      "loss": 0.3711,
+      "step": 1186
+    },
+    {
+      "epoch": 0.7618741976893453,
+      "grad_norm": 0.012300627306103706,
+      "learning_rate": 1.4151973332028134e-05,
+      "loss": 0.3379,
+      "step": 1187
+    },
+    {
+      "epoch": 0.7625160462130937,
+      "grad_norm": 0.011986950412392616,
+      "learning_rate": 1.4079580732624104e-05,
+      "loss": 0.3555,
+      "step": 1188
+    },
+    {
+      "epoch": 0.7631578947368421,
+      "grad_norm": 0.013296863064169884,
+      "learning_rate": 1.4007343411981189e-05,
+      "loss": 0.3438,
+      "step": 1189
+    },
+    {
+      "epoch": 0.7637997432605905,
+      "grad_norm": 0.014532539993524551,
+      "learning_rate": 1.3935261682370849e-05,
+      "loss": 0.3672,
+      "step": 1190
+    },
+    {
+      "epoch": 0.7644415917843389,
+      "grad_norm": 0.012205745093524456,
+      "learning_rate": 1.3863335855391967e-05,
+      "loss": 0.3711,
+      "step": 1191
+    },
+    {
+      "epoch": 0.7650834403080873,
+      "grad_norm": 0.012984980829060078,
+      "learning_rate": 1.379156624196941e-05,
+      "loss": 0.3984,
+      "step": 1192
+    },
+    {
+      "epoch": 0.7657252888318357,
+      "grad_norm": 0.012554211542010307,
+      "learning_rate": 1.371995315235287e-05,
+      "loss": 0.3477,
+      "step": 1193
+    },
+    {
+      "epoch": 0.766367137355584,
+      "grad_norm": 0.013144235126674175,
+      "learning_rate": 1.3648496896115292e-05,
+      "loss": 0.3535,
+      "step": 1194
+    },
+    {
+      "epoch": 0.7670089858793325,
+      "grad_norm": 0.012609812431037426,
+      "learning_rate": 1.3577197782151724e-05,
+      "loss": 0.3633,
+      "step": 1195
+    },
+    {
+      "epoch": 0.7676508344030809,
+      "grad_norm": 0.012705634348094463,
+      "learning_rate": 1.35060561186779e-05,
+      "loss": 0.3984,
+      "step": 1196
+    },
+    {
+      "epoch": 0.7682926829268293,
+      "grad_norm": 0.013995714485645294,
+      "learning_rate": 1.3435072213228889e-05,
+      "loss": 0.3711,
+      "step": 1197
+    },
+    {
+      "epoch": 0.7689345314505777,
+      "grad_norm": 0.0138862868770957,
+      "learning_rate": 1.3364246372657824e-05,
+      "loss": 0.3984,
+      "step": 1198
+    },
+    {
+      "epoch": 0.769576379974326,
+      "grad_norm": 0.013842416927218437,
+      "learning_rate": 1.3293578903134546e-05,
+      "loss": 0.3984,
+      "step": 1199
+    },
+    {
+      "epoch": 0.7702182284980744,
+      "grad_norm": 0.013586601242423058,
+      "learning_rate": 1.3223070110144265e-05,
+      "loss": 0.3906,
+      "step": 1200
+    },
+    {
+      "epoch": 0.7708600770218228,
+      "grad_norm": 0.014347325079143047,
+      "learning_rate": 1.3152720298486276e-05,
+      "loss": 0.3984,
+      "step": 1201
+    },
+    {
+      "epoch": 0.7715019255455713,
+      "grad_norm": 0.012852374464273453,
+      "learning_rate": 1.308252977227259e-05,
+      "loss": 0.3945,
+      "step": 1202
+    },
+    {
+      "epoch": 0.7721437740693197,
+      "grad_norm": 0.012284908443689346,
+      "learning_rate": 1.3012498834926662e-05,
+      "loss": 0.3672,
+      "step": 1203
+    },
+    {
+      "epoch": 0.772785622593068,
+      "grad_norm": 0.013469730503857136,
+      "learning_rate": 1.2942627789182121e-05,
+      "loss": 0.3633,
+      "step": 1204
+    },
+    {
+      "epoch": 0.7734274711168164,
+      "grad_norm": 0.014604169875383377,
+      "learning_rate": 1.2872916937081308e-05,
+      "loss": 0.3945,
+      "step": 1205
+    },
+    {
+      "epoch": 0.7740693196405648,
+      "grad_norm": 0.012190227396786213,
+      "learning_rate": 1.280336657997417e-05,
+      "loss": 0.3984,
+      "step": 1206
+    },
+    {
+      "epoch": 0.7747111681643132,
+      "grad_norm": 0.015167908743023872,
+      "learning_rate": 1.2733977018516758e-05,
+      "loss": 0.4238,
+      "step": 1207
+    },
+    {
+      "epoch": 0.7753530166880617,
+      "grad_norm": 0.01425865013152361,
+      "learning_rate": 1.2664748552670113e-05,
+      "loss": 0.3672,
+      "step": 1208
+    },
+    {
+      "epoch": 0.77599486521181,
+      "grad_norm": 0.01407993957400322,
+      "learning_rate": 1.259568148169884e-05,
+      "loss": 0.3945,
+      "step": 1209
+    },
+    {
+      "epoch": 0.7766367137355584,
+      "grad_norm": 0.013557854108512402,
+      "learning_rate": 1.2526776104169868e-05,
+      "loss": 0.3809,
+      "step": 1210
+    },
+    {
+      "epoch": 0.7772785622593068,
+      "grad_norm": 0.013125860132277012,
+      "learning_rate": 1.2458032717951163e-05,
+      "loss": 0.3594,
+      "step": 1211
+    },
+    {
+      "epoch": 0.7779204107830552,
+      "grad_norm": 0.012713338248431683,
+      "learning_rate": 1.238945162021038e-05,
+      "loss": 0.3906,
+      "step": 1212
+    },
+    {
+      "epoch": 0.7785622593068036,
+      "grad_norm": 0.013876782730221748,
+      "learning_rate": 1.2321033107413666e-05,
+      "loss": 0.3789,
+      "step": 1213
+    },
+    {
+      "epoch": 0.7792041078305519,
+      "grad_norm": 0.013413132168352604,
+      "learning_rate": 1.2252777475324334e-05,
+      "loss": 0.3633,
+      "step": 1214
+    },
+    {
+      "epoch": 0.7798459563543004,
+      "grad_norm": 0.012726066634058952,
+      "learning_rate": 1.2184685019001574e-05,
+      "loss": 0.3887,
+      "step": 1215
+    },
+    {
+      "epoch": 0.7804878048780488,
+      "grad_norm": 0.013412890955805779,
+      "learning_rate": 1.2116756032799193e-05,
+      "loss": 0.3945,
+      "step": 1216
+    },
+    {
+      "epoch": 0.7811296534017972,
+      "grad_norm": 0.012434509582817554,
+      "learning_rate": 1.2048990810364363e-05,
+      "loss": 0.3906,
+      "step": 1217
+    },
+    {
+      "epoch": 0.7817715019255456,
+      "grad_norm": 0.014923973940312862,
+      "learning_rate": 1.1981389644636276e-05,
+      "loss": 0.4062,
+      "step": 1218
+    },
+    {
+      "epoch": 0.7824133504492939,
+      "grad_norm": 0.012189259752631187,
+      "learning_rate": 1.1913952827844993e-05,
+      "loss": 0.3691,
+      "step": 1219
+    },
+    {
+      "epoch": 0.7830551989730423,
+      "grad_norm": 0.012858893722295761,
+      "learning_rate": 1.184668065151005e-05,
+      "loss": 0.3711,
+      "step": 1220
+    },
+    {
+      "epoch": 0.7836970474967908,
+      "grad_norm": 0.013458888977766037,
+      "learning_rate": 1.1779573406439343e-05,
+      "loss": 0.3555,
+      "step": 1221
+    },
+    {
+      "epoch": 0.7843388960205392,
+      "grad_norm": 0.014321232214570045,
+      "learning_rate": 1.1712631382727763e-05,
+      "loss": 0.3906,
+      "step": 1222
+    },
+    {
+      "epoch": 0.7849807445442876,
+      "grad_norm": 0.013449189253151417,
+      "learning_rate": 1.1645854869755928e-05,
+      "loss": 0.3711,
+      "step": 1223
+    },
+    {
+      "epoch": 0.785622593068036,
+      "grad_norm": 0.013625388965010643,
+      "learning_rate": 1.1579244156189057e-05,
+      "loss": 0.4023,
+      "step": 1224
+    },
+    {
+      "epoch": 0.7862644415917843,
+      "grad_norm": 0.012530328705906868,
+      "learning_rate": 1.151279952997556e-05,
+      "loss": 0.3711,
+      "step": 1225
+    },
+    {
+      "epoch": 0.7869062901155327,
+      "grad_norm": 0.013215388171374798,
+      "learning_rate": 1.1446521278345928e-05,
+      "loss": 0.375,
+      "step": 1226
+    },
+    {
+      "epoch": 0.7875481386392811,
+      "grad_norm": 0.011637991294264793,
+      "learning_rate": 1.1380409687811461e-05,
+      "loss": 0.3809,
+      "step": 1227
+    },
+    {
+      "epoch": 0.7881899871630296,
+      "grad_norm": 0.014323128387331963,
+      "learning_rate": 1.1314465044162932e-05,
+      "loss": 0.4062,
+      "step": 1228
+    },
+    {
+      "epoch": 0.788831835686778,
+      "grad_norm": 0.011796939186751842,
+      "learning_rate": 1.1248687632469485e-05,
+      "loss": 0.3672,
+      "step": 1229
+    },
+    {
+      "epoch": 0.7894736842105263,
+      "grad_norm": 0.01327999122440815,
+      "learning_rate": 1.1183077737077336e-05,
+      "loss": 0.373,
+      "step": 1230
+    },
+    {
+      "epoch": 0.7901155327342747,
+      "grad_norm": 0.012252957560122013,
+      "learning_rate": 1.1117635641608509e-05,
+      "loss": 0.3945,
+      "step": 1231
+    },
+    {
+      "epoch": 0.7907573812580231,
+      "grad_norm": 0.011731931939721107,
+      "learning_rate": 1.1052361628959745e-05,
+      "loss": 0.3633,
+      "step": 1232
+    },
+    {
+      "epoch": 0.7913992297817715,
+      "grad_norm": 0.013072323985397816,
+      "learning_rate": 1.098725598130108e-05,
+      "loss": 0.3555,
+      "step": 1233
+    },
+    {
+      "epoch": 0.79204107830552,
+      "grad_norm": 0.01314816065132618,
+      "learning_rate": 1.0922318980074808e-05,
+      "loss": 0.3594,
+      "step": 1234
+    },
+    {
+      "epoch": 0.7926829268292683,
+      "grad_norm": 0.012540193274617195,
+      "learning_rate": 1.0857550905994175e-05,
+      "loss": 0.3809,
+      "step": 1235
+    },
+    {
+      "epoch": 0.7933247753530167,
+      "grad_norm": 0.012471678666770458,
+      "learning_rate": 1.079295203904213e-05,
+      "loss": 0.3594,
+      "step": 1236
+    },
+    {
+      "epoch": 0.7939666238767651,
+      "grad_norm": 0.012217460200190544,
+      "learning_rate": 1.0728522658470231e-05,
+      "loss": 0.3789,
+      "step": 1237
+    },
+    {
+      "epoch": 0.7946084724005135,
+      "grad_norm": 0.0153912752866745,
+      "learning_rate": 1.0664263042797335e-05,
+      "loss": 0.3594,
+      "step": 1238
+    },
+    {
+      "epoch": 0.7952503209242618,
+      "grad_norm": 0.013098095543682575,
+      "learning_rate": 1.0600173469808444e-05,
+      "loss": 0.3516,
+      "step": 1239
+    },
+    {
+      "epoch": 0.7958921694480102,
+      "grad_norm": 0.0131600983440876,
+      "learning_rate": 1.0536254216553487e-05,
+      "loss": 0.3828,
+      "step": 1240
+    },
+    {
+      "epoch": 0.7965340179717587,
+      "grad_norm": 0.01303099375218153,
+      "learning_rate": 1.0472505559346102e-05,
+      "loss": 0.377,
+      "step": 1241
+    },
+    {
+      "epoch": 0.7971758664955071,
+      "grad_norm": 0.013537733815610409,
+      "learning_rate": 1.04089277737625e-05,
+      "loss": 0.3828,
+      "step": 1242
+    },
+    {
+      "epoch": 0.7978177150192555,
+      "grad_norm": 0.016070852056145668,
+      "learning_rate": 1.0345521134640224e-05,
+      "loss": 0.373,
+      "step": 1243
+    },
+    {
+      "epoch": 0.7984595635430038,
+      "grad_norm": 0.012657620012760162,
+      "learning_rate": 1.0282285916076973e-05,
+      "loss": 0.3633,
+      "step": 1244
+    },
+    {
+      "epoch": 0.7991014120667522,
+      "grad_norm": 0.013497685082256794,
+      "learning_rate": 1.021922239142944e-05,
+      "loss": 0.3945,
+      "step": 1245
+    },
+    {
+      "epoch": 0.7997432605905006,
+      "grad_norm": 0.014096570201218128,
+      "learning_rate": 1.015633083331206e-05,
+      "loss": 0.3809,
+      "step": 1246
+    },
+    {
+      "epoch": 0.800385109114249,
+      "grad_norm": 0.0128650376573205,
+      "learning_rate": 1.009361151359593e-05,
+      "loss": 0.3848,
+      "step": 1247
+    },
+    {
+      "epoch": 0.8010269576379975,
+      "grad_norm": 0.012854866683483124,
+      "learning_rate": 1.0031064703407572e-05,
+      "loss": 0.3555,
+      "step": 1248
+    },
+    {
+      "epoch": 0.8016688061617459,
+      "grad_norm": 0.012545584701001644,
+      "learning_rate": 9.968690673127774e-06,
+      "loss": 0.3594,
+      "step": 1249
+    },
+    {
+      "epoch": 0.8023106546854942,
+      "grad_norm": 0.013711373321712017,
+      "learning_rate": 9.906489692390426e-06,
+      "loss": 0.3945,
+      "step": 1250
+    },
+    {
+      "epoch": 0.8029525032092426,
+      "grad_norm": 0.012173696421086788,
+      "learning_rate": 9.84446203008132e-06,
+      "loss": 0.3477,
+      "step": 1251
+    },
+    {
+      "epoch": 0.803594351732991,
+      "grad_norm": 0.013947328552603722,
+      "learning_rate": 9.782607954337059e-06,
+      "loss": 0.3652,
+      "step": 1252
+    },
+    {
+      "epoch": 0.8042362002567394,
+      "grad_norm": 0.012948707677423954,
+      "learning_rate": 9.720927732543845e-06,
+      "loss": 0.3984,
+      "step": 1253
+    },
+    {
+      "epoch": 0.8048780487804879,
+      "grad_norm": 0.013634390197694302,
+      "learning_rate": 9.659421631336295e-06,
+      "loss": 0.3535,
+      "step": 1254
+    },
+    {
+      "epoch": 0.8055198973042362,
+      "grad_norm": 0.012434390373528004,
+      "learning_rate": 9.59808991659641e-06,
+      "loss": 0.3555,
+      "step": 1255
+    },
+    {
+      "epoch": 0.8061617458279846,
+      "grad_norm": 0.013895786367356777,
+      "learning_rate": 9.536932853452252e-06,
+      "loss": 0.375,
+      "step": 1256
+    },
+    {
+      "epoch": 0.806803594351733,
+      "grad_norm": 0.014669674448668957,
+      "learning_rate": 9.475950706276948e-06,
+      "loss": 0.3984,
+      "step": 1257
+    },
+    {
+      "epoch": 0.8074454428754814,
+      "grad_norm": 0.013176476582884789,
+      "learning_rate": 9.415143738687493e-06,
+      "loss": 0.3281,
+      "step": 1258
+    },
+    {
+      "epoch": 0.8080872913992297,
+      "grad_norm": 0.014073810540139675,
+      "learning_rate": 9.354512213543538e-06,
+      "loss": 0.3848,
+      "step": 1259
+    },
+    {
+      "epoch": 0.8087291399229781,
+      "grad_norm": 0.013018610887229443,
+      "learning_rate": 9.294056392946427e-06,
+      "loss": 0.3906,
+      "step": 1260
+    },
+    {
+      "epoch": 0.8093709884467266,
+      "grad_norm": 0.013114540837705135,
+      "learning_rate": 9.233776538237854e-06,
+      "loss": 0.373,
+      "step": 1261
+    },
+    {
+      "epoch": 0.810012836970475,
+      "grad_norm": 0.013796181418001652,
+      "learning_rate": 9.17367290999891e-06,
+      "loss": 0.373,
+      "step": 1262
+    },
+    {
+      "epoch": 0.8106546854942234,
+      "grad_norm": 0.01330437883734703,
+      "learning_rate": 9.113745768048865e-06,
+      "loss": 0.3574,
+      "step": 1263
+    },
+    {
+      "epoch": 0.8112965340179717,
+      "grad_norm": 0.01230168342590332,
+      "learning_rate": 9.053995371444035e-06,
+      "loss": 0.3809,
+      "step": 1264
+    },
+    {
+      "epoch": 0.8119383825417201,
+      "grad_norm": 0.01638161949813366,
+      "learning_rate": 8.994421978476735e-06,
+      "loss": 0.3574,
+      "step": 1265
+    },
+    {
+      "epoch": 0.8125802310654685,
+      "grad_norm": 0.014375623315572739,
+      "learning_rate": 8.935025846674088e-06,
+      "loss": 0.4043,
+      "step": 1266
+    },
+    {
+      "epoch": 0.813222079589217,
+      "grad_norm": 0.012246549129486084,
+      "learning_rate": 8.875807232796968e-06,
+      "loss": 0.3711,
+      "step": 1267
+    },
+    {
+      "epoch": 0.8138639281129654,
+      "grad_norm": 0.01342903170734644,
+      "learning_rate": 8.816766392838855e-06,
+      "loss": 0.3711,
+      "step": 1268
+    },
+    {
+      "epoch": 0.8145057766367138,
+      "grad_norm": 0.012210494838654995,
+      "learning_rate": 8.757903582024706e-06,
+      "loss": 0.3672,
+      "step": 1269
+    },
+    {
+      "epoch": 0.8151476251604621,
+      "grad_norm": 0.012337923049926758,
+      "learning_rate": 8.699219054809937e-06,
+      "loss": 0.3438,
+      "step": 1270
+    },
+    {
+      "epoch": 0.8157894736842105,
+      "grad_norm": 0.01411160733550787,
+      "learning_rate": 8.640713064879236e-06,
+      "loss": 0.4023,
+      "step": 1271
+    },
+    {
+      "epoch": 0.8164313222079589,
+      "grad_norm": 0.013467960990965366,
+      "learning_rate": 8.582385865145508e-06,
+      "loss": 0.3867,
+      "step": 1272
+    },
+    {
+      "epoch": 0.8170731707317073,
+      "grad_norm": 0.012686812318861485,
+      "learning_rate": 8.5242377077488e-06,
+      "loss": 0.375,
+      "step": 1273
+    },
+    {
+      "epoch": 0.8177150192554558,
+      "grad_norm": 0.01201626192778349,
+      "learning_rate": 8.46626884405512e-06,
+      "loss": 0.3809,
+      "step": 1274
+    },
+    {
+      "epoch": 0.8183568677792041,
+      "grad_norm": 0.012600788846611977,
+      "learning_rate": 8.408479524655477e-06,
+      "loss": 0.3828,
+      "step": 1275
+    },
+    {
+      "epoch": 0.8189987163029525,
+      "grad_norm": 0.012168734334409237,
+      "learning_rate": 8.350869999364713e-06,
+      "loss": 0.373,
+      "step": 1276
+    },
+    {
+      "epoch": 0.8196405648267009,
+      "grad_norm": 0.013481889851391315,
+      "learning_rate": 8.293440517220446e-06,
+      "loss": 0.3613,
+      "step": 1277
+    },
+    {
+      "epoch": 0.8202824133504493,
+      "grad_norm": 0.0141798360273242,
+      "learning_rate": 8.236191326482007e-06,
+      "loss": 0.3906,
+      "step": 1278
+    },
+    {
+      "epoch": 0.8209242618741976,
+      "grad_norm": 0.012813771143555641,
+      "learning_rate": 8.179122674629324e-06,
+      "loss": 0.375,
+      "step": 1279
+    },
+    {
+      "epoch": 0.8215661103979461,
+      "grad_norm": 0.014093801379203796,
+      "learning_rate": 8.122234808361907e-06,
+      "loss": 0.3945,
+      "step": 1280
+    },
+    {
+      "epoch": 0.8222079589216945,
+      "grad_norm": 0.013460157439112663,
+      "learning_rate": 8.065527973597742e-06,
+      "loss": 0.4004,
+      "step": 1281
+    },
+    {
+      "epoch": 0.8228498074454429,
+      "grad_norm": 0.01379771064966917,
+      "learning_rate": 8.00900241547226e-06,
+      "loss": 0.3867,
+      "step": 1282
+    },
+    {
+      "epoch": 0.8234916559691913,
+      "grad_norm": 0.013519086875021458,
+      "learning_rate": 7.952658378337252e-06,
+      "loss": 0.4043,
+      "step": 1283
+    },
+    {
+      "epoch": 0.8241335044929397,
+      "grad_norm": 0.012356378138065338,
+      "learning_rate": 7.896496105759799e-06,
+      "loss": 0.3906,
+      "step": 1284
+    },
+    {
+      "epoch": 0.824775353016688,
+      "grad_norm": 0.013102208264172077,
+      "learning_rate": 7.840515840521263e-06,
+      "loss": 0.3418,
+      "step": 1285
+    },
+    {
+      "epoch": 0.8254172015404364,
+      "grad_norm": 0.012744414620101452,
+      "learning_rate": 7.784717824616222e-06,
+      "loss": 0.3711,
+      "step": 1286
+    },
+    {
+      "epoch": 0.8260590500641849,
+      "grad_norm": 0.01288849301636219,
+      "learning_rate": 7.729102299251367e-06,
+      "loss": 0.3906,
+      "step": 1287
+    },
+    {
+      "epoch": 0.8267008985879333,
+      "grad_norm": 0.012660497799515724,
+      "learning_rate": 7.673669504844599e-06,
+      "loss": 0.3906,
+      "step": 1288
+    },
+    {
+      "epoch": 0.8273427471116817,
+      "grad_norm": 0.011713840067386627,
+      "learning_rate": 7.618419681023809e-06,
+      "loss": 0.3594,
+      "step": 1289
+    },
+    {
+      "epoch": 0.82798459563543,
+      "grad_norm": 0.013892716728150845,
+      "learning_rate": 7.563353066625972e-06,
+      "loss": 0.3633,
+      "step": 1290
+    },
+    {
+      "epoch": 0.8286264441591784,
+      "grad_norm": 0.01375551987439394,
+      "learning_rate": 7.508469899696096e-06,
+      "loss": 0.3789,
+      "step": 1291
+    },
+    {
+      "epoch": 0.8292682926829268,
+      "grad_norm": 0.013938702642917633,
+      "learning_rate": 7.453770417486123e-06,
+      "loss": 0.3652,
+      "step": 1292
+    },
+    {
+      "epoch": 0.8299101412066753,
+      "grad_norm": 0.012904335744678974,
+      "learning_rate": 7.399254856453985e-06,
+      "loss": 0.3848,
+      "step": 1293
+    },
+    {
+      "epoch": 0.8305519897304237,
+      "grad_norm": 0.01319153606891632,
+      "learning_rate": 7.344923452262548e-06,
+      "loss": 0.3789,
+      "step": 1294
+    },
+    {
+      "epoch": 0.831193838254172,
+      "grad_norm": 0.0127947349101305,
+      "learning_rate": 7.2907764397785845e-06,
+      "loss": 0.3672,
+      "step": 1295
+    },
+    {
+      "epoch": 0.8318356867779204,
+      "grad_norm": 0.014471789821982384,
+      "learning_rate": 7.2368140530717945e-06,
+      "loss": 0.3828,
+      "step": 1296
+    },
+    {
+      "epoch": 0.8324775353016688,
+      "grad_norm": 0.011869586072862148,
+      "learning_rate": 7.183036525413716e-06,
+      "loss": 0.375,
+      "step": 1297
+    },
+    {
+      "epoch": 0.8331193838254172,
+      "grad_norm": 0.012698578648269176,
+      "learning_rate": 7.1294440892768155e-06,
+      "loss": 0.3457,
+      "step": 1298
+    },
+    {
+      "epoch": 0.8337612323491655,
+      "grad_norm": 0.011361399665474892,
+      "learning_rate": 7.076036976333417e-06,
+      "loss": 0.3945,
+      "step": 1299
+    },
+    {
+      "epoch": 0.834403080872914,
+      "grad_norm": 0.01272041629999876,
+      "learning_rate": 7.02281541745472e-06,
+      "loss": 0.3652,
+      "step": 1300
+    },
+    {
+      "epoch": 0.8350449293966624,
+      "grad_norm": 0.012732606381177902,
+      "learning_rate": 6.969779642709817e-06,
+      "loss": 0.3789,
+      "step": 1301
+    },
+    {
+      "epoch": 0.8356867779204108,
+      "grad_norm": 0.012499982491135597,
+      "learning_rate": 6.916929881364642e-06,
+      "loss": 0.3906,
+      "step": 1302
+    },
+    {
+      "epoch": 0.8363286264441592,
+      "grad_norm": 0.01187952607870102,
+      "learning_rate": 6.864266361881056e-06,
+      "loss": 0.3789,
+      "step": 1303
+    },
+    {
+      "epoch": 0.8369704749679076,
+      "grad_norm": 0.011955213733017445,
+      "learning_rate": 6.8117893119158176e-06,
+      "loss": 0.3906,
+      "step": 1304
+    },
+    {
+      "epoch": 0.8376123234916559,
+      "grad_norm": 0.011965644545853138,
+      "learning_rate": 6.759498958319599e-06,
+      "loss": 0.3555,
+      "step": 1305
+    },
+    {
+      "epoch": 0.8382541720154044,
+      "grad_norm": 0.012529222294688225,
+      "learning_rate": 6.707395527136018e-06,
+      "loss": 0.3633,
+      "step": 1306
+    },
+    {
+      "epoch": 0.8388960205391528,
+      "grad_norm": 0.012780443765223026,
+      "learning_rate": 6.655479243600632e-06,
+      "loss": 0.3789,
+      "step": 1307
+    },
+    {
+      "epoch": 0.8395378690629012,
+      "grad_norm": 0.014528384432196617,
+      "learning_rate": 6.603750332140019e-06,
+      "loss": 0.4121,
+      "step": 1308
+    },
+    {
+      "epoch": 0.8401797175866496,
+      "grad_norm": 0.012622330337762833,
+      "learning_rate": 6.5522090163707415e-06,
+      "loss": 0.375,
+      "step": 1309
+    },
+    {
+      "epoch": 0.8408215661103979,
+      "grad_norm": 0.013054592534899712,
+      "learning_rate": 6.500855519098448e-06,
+      "loss": 0.3809,
+      "step": 1310
+    },
+    {
+      "epoch": 0.8414634146341463,
+      "grad_norm": 0.011494606733322144,
+      "learning_rate": 6.44969006231686e-06,
+      "loss": 0.3398,
+      "step": 1311
+    },
+    {
+      "epoch": 0.8421052631578947,
+      "grad_norm": 0.011829941533505917,
+      "learning_rate": 6.398712867206797e-06,
+      "loss": 0.3652,
+      "step": 1312
+    },
+    {
+      "epoch": 0.8427471116816432,
+      "grad_norm": 0.012617502361536026,
+      "learning_rate": 6.347924154135298e-06,
+      "loss": 0.3828,
+      "step": 1313
+    },
+    {
+      "epoch": 0.8433889602053916,
+      "grad_norm": 0.013350032269954681,
+      "learning_rate": 6.2973241426546095e-06,
+      "loss": 0.3594,
+      "step": 1314
+    },
+    {
+      "epoch": 0.8440308087291399,
+      "grad_norm": 0.012515634298324585,
+      "learning_rate": 6.246913051501202e-06,
+      "loss": 0.3828,
+      "step": 1315
+    },
+    {
+      "epoch": 0.8446726572528883,
+      "grad_norm": 0.012150679714977741,
+      "learning_rate": 6.196691098594953e-06,
+      "loss": 0.3516,
+      "step": 1316
+    },
+    {
+      "epoch": 0.8453145057766367,
+      "grad_norm": 0.012245026417076588,
+      "learning_rate": 6.146658501038055e-06,
+      "loss": 0.3945,
+      "step": 1317
+    },
+    {
+      "epoch": 0.8459563543003851,
+      "grad_norm": 0.012508809566497803,
+      "learning_rate": 6.09681547511417e-06,
+      "loss": 0.3848,
+      "step": 1318
+    },
+    {
+      "epoch": 0.8465982028241335,
+      "grad_norm": 0.014197138138115406,
+      "learning_rate": 6.047162236287485e-06,
+      "loss": 0.3809,
+      "step": 1319
+    },
+    {
+      "epoch": 0.8472400513478819,
+      "grad_norm": 0.011651071719825268,
+      "learning_rate": 5.997698999201723e-06,
+      "loss": 0.3906,
+      "step": 1320
+    },
+    {
+      "epoch": 0.8478818998716303,
+      "grad_norm": 0.012558822520077229,
+      "learning_rate": 5.948425977679289e-06,
+      "loss": 0.3945,
+      "step": 1321
+    },
+    {
+      "epoch": 0.8485237483953787,
+      "grad_norm": 0.01349442545324564,
+      "learning_rate": 5.899343384720313e-06,
+      "loss": 0.3789,
+      "step": 1322
+    },
+    {
+      "epoch": 0.8491655969191271,
+      "grad_norm": 0.013269457034766674,
+      "learning_rate": 5.850451432501725e-06,
+      "loss": 0.3594,
+      "step": 1323
+    },
+    {
+      "epoch": 0.8498074454428755,
+      "grad_norm": 0.01424845214933157,
+      "learning_rate": 5.8017503323763375e-06,
+      "loss": 0.3984,
+      "step": 1324
+    },
+    {
+      "epoch": 0.8504492939666238,
+      "grad_norm": 0.012251721695065498,
+      "learning_rate": 5.753240294871937e-06,
+      "loss": 0.3633,
+      "step": 1325
+    },
+    {
+      "epoch": 0.8510911424903723,
+      "grad_norm": 0.01245789136737585,
+      "learning_rate": 5.704921529690377e-06,
+      "loss": 0.3555,
+      "step": 1326
+    },
+    {
+      "epoch": 0.8517329910141207,
+      "grad_norm": 0.012180859223008156,
+      "learning_rate": 5.656794245706676e-06,
+      "loss": 0.3848,
+      "step": 1327
+    },
+    {
+      "epoch": 0.8523748395378691,
+      "grad_norm": 0.011467906646430492,
+      "learning_rate": 5.608858650968096e-06,
+      "loss": 0.3438,
+      "step": 1328
+    },
+    {
+      "epoch": 0.8530166880616175,
+      "grad_norm": 0.01379603985697031,
+      "learning_rate": 5.561114952693269e-06,
+      "loss": 0.3672,
+      "step": 1329
+    },
+    {
+      "epoch": 0.8536585365853658,
+      "grad_norm": 0.013673442415893078,
+      "learning_rate": 5.513563357271256e-06,
+      "loss": 0.3516,
+      "step": 1330
+    },
+    {
+      "epoch": 0.8543003851091142,
+      "grad_norm": 0.012737073935568333,
+      "learning_rate": 5.466204070260717e-06,
+      "loss": 0.4004,
+      "step": 1331
+    },
+    {
+      "epoch": 0.8549422336328626,
+      "grad_norm": 0.011530360206961632,
+      "learning_rate": 5.4190372963889794e-06,
+      "loss": 0.3594,
+      "step": 1332
+    },
+    {
+      "epoch": 0.8555840821566111,
+      "grad_norm": 0.01489552017301321,
+      "learning_rate": 5.372063239551162e-06,
+      "loss": 0.3691,
+      "step": 1333
+    },
+    {
+      "epoch": 0.8562259306803595,
+      "grad_norm": 0.01377855334430933,
+      "learning_rate": 5.325282102809304e-06,
+      "loss": 0.3672,
+      "step": 1334
+    },
+    {
+      "epoch": 0.8568677792041078,
+      "grad_norm": 0.012593116611242294,
+      "learning_rate": 5.278694088391462e-06,
+      "loss": 0.4102,
+      "step": 1335
+    },
+    {
+      "epoch": 0.8575096277278562,
+      "grad_norm": 0.011791418306529522,
+      "learning_rate": 5.23229939769086e-06,
+      "loss": 0.3867,
+      "step": 1336
+    },
+    {
+      "epoch": 0.8581514762516046,
+      "grad_norm": 0.01504043210297823,
+      "learning_rate": 5.186098231265024e-06,
+      "loss": 0.3984,
+      "step": 1337
+    },
+    {
+      "epoch": 0.858793324775353,
+      "grad_norm": 0.013028615154325962,
+      "learning_rate": 5.140090788834895e-06,
+      "loss": 0.3516,
+      "step": 1338
+    },
+    {
+      "epoch": 0.8594351732991015,
+      "grad_norm": 0.01274171657860279,
+      "learning_rate": 5.0942772692839755e-06,
+      "loss": 0.377,
+      "step": 1339
+    },
+    {
+      "epoch": 0.8600770218228498,
+      "grad_norm": 0.015606660395860672,
+      "learning_rate": 5.048657870657447e-06,
+      "loss": 0.4062,
+      "step": 1340
+    },
+    {
+      "epoch": 0.8607188703465982,
+      "grad_norm": 0.01197892427444458,
+      "learning_rate": 5.003232790161366e-06,
+      "loss": 0.3496,
+      "step": 1341
+    },
+    {
+      "epoch": 0.8613607188703466,
+      "grad_norm": 0.013816019520163536,
+      "learning_rate": 4.9580022241617674e-06,
+      "loss": 0.377,
+      "step": 1342
+    },
+    {
+      "epoch": 0.862002567394095,
+      "grad_norm": 0.012484337203204632,
+      "learning_rate": 4.912966368183797e-06,
+      "loss": 0.3516,
+      "step": 1343
+    },
+    {
+      "epoch": 0.8626444159178434,
+      "grad_norm": 0.012311607599258423,
+      "learning_rate": 4.868125416910957e-06,
+      "loss": 0.375,
+      "step": 1344
+    },
+    {
+      "epoch": 0.8632862644415917,
+      "grad_norm": 0.013449694029986858,
+      "learning_rate": 4.82347956418418e-06,
+      "loss": 0.418,
+      "step": 1345
+    },
+    {
+      "epoch": 0.8639281129653402,
+      "grad_norm": 0.01339624635875225,
+      "learning_rate": 4.779029003000979e-06,
+      "loss": 0.3477,
+      "step": 1346
+    },
+    {
+      "epoch": 0.8645699614890886,
+      "grad_norm": 0.012256176210939884,
+      "learning_rate": 4.734773925514712e-06,
+      "loss": 0.3789,
+      "step": 1347
+    },
+    {
+      "epoch": 0.865211810012837,
+      "grad_norm": 0.012527239508926868,
+      "learning_rate": 4.690714523033635e-06,
+      "loss": 0.3613,
+      "step": 1348
+    },
+    {
+      "epoch": 0.8658536585365854,
+      "grad_norm": 0.011592155322432518,
+      "learning_rate": 4.646850986020151e-06,
+      "loss": 0.3945,
+      "step": 1349
+    },
+    {
+      "epoch": 0.8664955070603337,
+      "grad_norm": 0.012507716193795204,
+      "learning_rate": 4.603183504089997e-06,
+      "loss": 0.3828,
+      "step": 1350
+    },
+    {
+      "epoch": 0.8671373555840821,
+      "grad_norm": 0.012843828648328781,
+      "learning_rate": 4.559712266011329e-06,
+      "loss": 0.3984,
+      "step": 1351
+    },
+    {
+      "epoch": 0.8677792041078306,
+      "grad_norm": 0.0132277337834239,
+      "learning_rate": 4.516437459704032e-06,
+      "loss": 0.3887,
+      "step": 1352
+    },
+    {
+      "epoch": 0.868421052631579,
+      "grad_norm": 0.01294110156595707,
+      "learning_rate": 4.473359272238786e-06,
+      "loss": 0.3828,
+      "step": 1353
+    },
+    {
+      "epoch": 0.8690629011553274,
+      "grad_norm": 0.012413974851369858,
+      "learning_rate": 4.430477889836348e-06,
+      "loss": 0.3594,
+      "step": 1354
+    },
+    {
+      "epoch": 0.8697047496790757,
+      "grad_norm": 0.013880117796361446,
+      "learning_rate": 4.387793497866744e-06,
+      "loss": 0.377,
+      "step": 1355
+    },
+    {
+      "epoch": 0.8703465982028241,
+      "grad_norm": 0.013049155473709106,
+      "learning_rate": 4.345306280848377e-06,
+      "loss": 0.3906,
+      "step": 1356
+    },
+    {
+      "epoch": 0.8709884467265725,
+      "grad_norm": 0.013583934865891933,
+      "learning_rate": 4.3030164224473365e-06,
+      "loss": 0.3496,
+      "step": 1357
+    },
+    {
+      "epoch": 0.8716302952503209,
+      "grad_norm": 0.0125590730458498,
+      "learning_rate": 4.260924105476544e-06,
+      "loss": 0.4023,
+      "step": 1358
+    },
+    {
+      "epoch": 0.8722721437740694,
+      "grad_norm": 0.014681275002658367,
+      "learning_rate": 4.219029511894973e-06,
+      "loss": 0.3789,
+      "step": 1359
+    },
+    {
+      "epoch": 0.8729139922978177,
+      "grad_norm": 0.015730444341897964,
+      "learning_rate": 4.177332822806873e-06,
+      "loss": 0.3633,
+      "step": 1360
+    },
+    {
+      "epoch": 0.8735558408215661,
+      "grad_norm": 0.0119576221331954,
+      "learning_rate": 4.135834218460982e-06,
+      "loss": 0.3672,
+      "step": 1361
+    },
+    {
+      "epoch": 0.8741976893453145,
+      "grad_norm": 0.013510183431208134,
+      "learning_rate": 4.094533878249751e-06,
+      "loss": 0.3555,
+      "step": 1362
+    },
+    {
+      "epoch": 0.8748395378690629,
+      "grad_norm": 0.014038784429430962,
+      "learning_rate": 4.053431980708566e-06,
+      "loss": 0.3711,
+      "step": 1363
+    },
+    {
+      "epoch": 0.8754813863928113,
+      "grad_norm": 0.013317780569195747,
+      "learning_rate": 4.012528703514951e-06,
+      "loss": 0.3574,
+      "step": 1364
+    },
+    {
+      "epoch": 0.8761232349165597,
+      "grad_norm": 0.012826112098991871,
+      "learning_rate": 3.971824223487841e-06,
+      "loss": 0.3711,
+      "step": 1365
+    },
+    {
+      "epoch": 0.8767650834403081,
+      "grad_norm": 0.012482481077313423,
+      "learning_rate": 3.931318716586807e-06,
+      "loss": 0.373,
+      "step": 1366
+    },
+    {
+      "epoch": 0.8774069319640565,
+      "grad_norm": 0.01242758333683014,
+      "learning_rate": 3.891012357911272e-06,
+      "loss": 0.3906,
+      "step": 1367
+    },
+    {
+      "epoch": 0.8780487804878049,
+      "grad_norm": 0.011938832700252533,
+      "learning_rate": 3.850905321699788e-06,
+      "loss": 0.3535,
+      "step": 1368
+    },
+    {
+      "epoch": 0.8786906290115533,
+      "grad_norm": 0.015014137141406536,
+      "learning_rate": 3.810997781329234e-06,
+      "loss": 0.3945,
+      "step": 1369
+    },
+    {
+      "epoch": 0.8793324775353016,
+      "grad_norm": 0.01675754226744175,
+      "learning_rate": 3.7712899093141407e-06,
+      "loss": 0.3652,
+      "step": 1370
+    },
+    {
+      "epoch": 0.87997432605905,
+      "grad_norm": 0.013484257273375988,
+      "learning_rate": 3.7317818773058456e-06,
+      "loss": 0.377,
+      "step": 1371
+    },
+    {
+      "epoch": 0.8806161745827985,
+      "grad_norm": 0.012842392548918724,
+      "learning_rate": 3.692473856091866e-06,
+      "loss": 0.3789,
+      "step": 1372
+    },
+    {
+      "epoch": 0.8812580231065469,
+      "grad_norm": 0.012742377817630768,
+      "learning_rate": 3.6533660155950776e-06,
+      "loss": 0.3672,
+      "step": 1373
+    },
+    {
+      "epoch": 0.8818998716302953,
+      "grad_norm": 0.012952295131981373,
+      "learning_rate": 3.6144585248729824e-06,
+      "loss": 0.3906,
+      "step": 1374
+    },
+    {
+      "epoch": 0.8825417201540436,
+      "grad_norm": 0.012831165455281734,
+      "learning_rate": 3.575751552117029e-06,
+      "loss": 0.3848,
+      "step": 1375
+    },
+    {
+      "epoch": 0.883183568677792,
+      "grad_norm": 0.013856924138963223,
+      "learning_rate": 3.537245264651845e-06,
+      "loss": 0.3965,
+      "step": 1376
+    },
+    {
+      "epoch": 0.8838254172015404,
+      "grad_norm": 0.014107098802924156,
+      "learning_rate": 3.498939828934511e-06,
+      "loss": 0.3828,
+      "step": 1377
+    },
+    {
+      "epoch": 0.8844672657252889,
+      "grad_norm": 0.012293580919504166,
+      "learning_rate": 3.4608354105538976e-06,
+      "loss": 0.373,
+      "step": 1378
+    },
+    {
+      "epoch": 0.8851091142490373,
+      "grad_norm": 0.012825761921703815,
+      "learning_rate": 3.4229321742298427e-06,
+      "loss": 0.3613,
+      "step": 1379
+    },
+    {
+      "epoch": 0.8857509627727856,
+      "grad_norm": 0.015361820347607136,
+      "learning_rate": 3.3852302838125626e-06,
+      "loss": 0.3984,
+      "step": 1380
+    },
+    {
+      "epoch": 0.886392811296534,
+      "grad_norm": 0.012380460277199745,
+      "learning_rate": 3.3477299022818544e-06,
+      "loss": 0.3555,
+      "step": 1381
+    },
+    {
+      "epoch": 0.8870346598202824,
+      "grad_norm": 0.012735201977193356,
+      "learning_rate": 3.3104311917464104e-06,
+      "loss": 0.3633,
+      "step": 1382
+    },
+    {
+      "epoch": 0.8876765083440308,
+      "grad_norm": 0.013652042485773563,
+      "learning_rate": 3.2733343134431714e-06,
+      "loss": 0.3867,
+      "step": 1383
+    },
+    {
+      "epoch": 0.8883183568677792,
+      "grad_norm": 0.013185295276343822,
+      "learning_rate": 3.236439427736543e-06,
+      "loss": 0.3867,
+      "step": 1384
+    },
+    {
+      "epoch": 0.8889602053915276,
+      "grad_norm": 0.012528514489531517,
+      "learning_rate": 3.1997466941177666e-06,
+      "loss": 0.3711,
+      "step": 1385
+    },
+    {
+      "epoch": 0.889602053915276,
+      "grad_norm": 0.013030996546149254,
+      "learning_rate": 3.163256271204218e-06,
+      "loss": 0.3672,
+      "step": 1386
+    },
+    {
+      "epoch": 0.8902439024390244,
+      "grad_norm": 0.011969957500696182,
+      "learning_rate": 3.1269683167386764e-06,
+      "loss": 0.3867,
+      "step": 1387
+    },
+    {
+      "epoch": 0.8908857509627728,
+      "grad_norm": 0.014140266925096512,
+      "learning_rate": 3.0908829875887223e-06,
+      "loss": 0.3809,
+      "step": 1388
+    },
+    {
+      "epoch": 0.8915275994865212,
+      "grad_norm": 0.01246250607073307,
+      "learning_rate": 3.0550004397459976e-06,
+      "loss": 0.3672,
+      "step": 1389
+    },
+    {
+      "epoch": 0.8921694480102695,
+      "grad_norm": 0.012091501615941525,
+      "learning_rate": 3.019320828325539e-06,
+      "loss": 0.3711,
+      "step": 1390
+    },
+    {
+      "epoch": 0.8928112965340179,
+      "grad_norm": 0.012726930901408195,
+      "learning_rate": 2.9838443075651535e-06,
+      "loss": 0.3945,
+      "step": 1391
+    },
+    {
+      "epoch": 0.8934531450577664,
+      "grad_norm": 0.01227556075900793,
+      "learning_rate": 2.9485710308246706e-06,
+      "loss": 0.3789,
+      "step": 1392
+    },
+    {
+      "epoch": 0.8940949935815148,
+      "grad_norm": 0.013887990266084671,
+      "learning_rate": 2.9135011505853572e-06,
+      "loss": 0.3828,
+      "step": 1393
+    },
+    {
+      "epoch": 0.8947368421052632,
+      "grad_norm": 0.01397820096462965,
+      "learning_rate": 2.8786348184492105e-06,
+      "loss": 0.3867,
+      "step": 1394
+    },
+    {
+      "epoch": 0.8953786906290115,
+      "grad_norm": 0.012721987441182137,
+      "learning_rate": 2.8439721851383383e-06,
+      "loss": 0.3984,
+      "step": 1395
+    },
+    {
+      "epoch": 0.8960205391527599,
+      "grad_norm": 0.012987341731786728,
+      "learning_rate": 2.8095134004942737e-06,
+      "loss": 0.373,
+      "step": 1396
+    },
+    {
+      "epoch": 0.8966623876765083,
+      "grad_norm": 0.013185744173824787,
+      "learning_rate": 2.7752586134773327e-06,
+      "loss": 0.3828,
+      "step": 1397
+    },
+    {
+      "epoch": 0.8973042362002568,
+      "grad_norm": 0.012743929401040077,
+      "learning_rate": 2.7412079721659988e-06,
+      "loss": 0.3789,
+      "step": 1398
+    },
+    {
+      "epoch": 0.8979460847240052,
+      "grad_norm": 0.014703886583447456,
+      "learning_rate": 2.707361623756249e-06,
+      "loss": 0.3711,
+      "step": 1399
+    },
+    {
+      "epoch": 0.8985879332477535,
+      "grad_norm": 0.011874165385961533,
+      "learning_rate": 2.6737197145609404e-06,
+      "loss": 0.3867,
+      "step": 1400
+    },
+    {
+      "epoch": 0.8992297817715019,
+      "grad_norm": 0.013200345449149609,
+      "learning_rate": 2.6402823900091744e-06,
+      "loss": 0.3555,
+      "step": 1401
+    },
+    {
+      "epoch": 0.8998716302952503,
+      "grad_norm": 0.013502005487680435,
+      "learning_rate": 2.607049794645633e-06,
+      "loss": 0.3516,
+      "step": 1402
+    },
+    {
+      "epoch": 0.9005134788189987,
+      "grad_norm": 0.013549595139920712,
+      "learning_rate": 2.574022072130017e-06,
+      "loss": 0.375,
+      "step": 1403
+    },
+    {
+      "epoch": 0.9011553273427471,
+      "grad_norm": 0.013294607400894165,
+      "learning_rate": 2.541199365236391e-06,
+      "loss": 0.3789,
+      "step": 1404
+    },
+    {
+      "epoch": 0.9017971758664955,
+      "grad_norm": 0.01240833755582571,
+      "learning_rate": 2.508581815852523e-06,
+      "loss": 0.3711,
+      "step": 1405
+    },
+    {
+      "epoch": 0.9024390243902439,
+      "grad_norm": 0.013817439787089825,
+      "learning_rate": 2.4761695649793804e-06,
+      "loss": 0.3828,
+      "step": 1406
+    },
+    {
+      "epoch": 0.9030808729139923,
+      "grad_norm": 0.012335661798715591,
+      "learning_rate": 2.4439627527303997e-06,
+      "loss": 0.3789,
+      "step": 1407
+    },
+    {
+      "epoch": 0.9037227214377407,
+      "grad_norm": 0.011767784133553505,
+      "learning_rate": 2.4119615183309628e-06,
+      "loss": 0.375,
+      "step": 1408
+    },
+    {
+      "epoch": 0.9043645699614891,
+      "grad_norm": 0.012927521951496601,
+      "learning_rate": 2.3801660001177782e-06,
+      "loss": 0.3535,
+      "step": 1409
+    },
+    {
+      "epoch": 0.9050064184852374,
+      "grad_norm": 0.014147186651825905,
+      "learning_rate": 2.3485763355382273e-06,
+      "loss": 0.3672,
+      "step": 1410
+    },
+    {
+      "epoch": 0.9056482670089859,
+      "grad_norm": 0.01323013473302126,
+      "learning_rate": 2.3171926611498808e-06,
+      "loss": 0.3691,
+      "step": 1411
+    },
+    {
+      "epoch": 0.9062901155327343,
+      "grad_norm": 0.011728956364095211,
+      "learning_rate": 2.2860151126197825e-06,
+      "loss": 0.3516,
+      "step": 1412
+    },
+    {
+      "epoch": 0.9069319640564827,
+      "grad_norm": 0.014273487031459808,
+      "learning_rate": 2.255043824723968e-06,
+      "loss": 0.3984,
+      "step": 1413
+    },
+    {
+      "epoch": 0.9075738125802311,
+      "grad_norm": 0.011534382589161396,
+      "learning_rate": 2.224278931346824e-06,
+      "loss": 0.3555,
+      "step": 1414
+    },
+    {
+      "epoch": 0.9082156611039794,
+      "grad_norm": 0.01171374786645174,
+      "learning_rate": 2.1937205654805004e-06,
+      "loss": 0.377,
+      "step": 1415
+    },
+    {
+      "epoch": 0.9088575096277278,
+      "grad_norm": 0.012463386170566082,
+      "learning_rate": 2.1633688592244016e-06,
+      "loss": 0.3594,
+      "step": 1416
+    },
+    {
+      "epoch": 0.9094993581514762,
+      "grad_norm": 0.012697212398052216,
+      "learning_rate": 2.1332239437845345e-06,
+      "loss": 0.3809,
+      "step": 1417
+    },
+    {
+      "epoch": 0.9101412066752247,
+      "grad_norm": 0.012483914382755756,
+      "learning_rate": 2.1032859494730052e-06,
+      "loss": 0.3867,
+      "step": 1418
+    },
+    {
+      "epoch": 0.9107830551989731,
+      "grad_norm": 0.0118754543364048,
+      "learning_rate": 2.0735550057074225e-06,
+      "loss": 0.3672,
+      "step": 1419
+    },
+    {
+      "epoch": 0.9114249037227214,
+      "grad_norm": 0.013126706704497337,
+      "learning_rate": 2.04403124101033e-06,
+      "loss": 0.3711,
+      "step": 1420
+    },
+    {
+      "epoch": 0.9120667522464698,
+      "grad_norm": 0.013734589330852032,
+      "learning_rate": 2.0147147830086866e-06,
+      "loss": 0.3496,
+      "step": 1421
+    },
+    {
+      "epoch": 0.9127086007702182,
+      "grad_norm": 0.013323265127837658,
+      "learning_rate": 1.98560575843329e-06,
+      "loss": 0.3789,
+      "step": 1422
+    },
+    {
+      "epoch": 0.9133504492939666,
+      "grad_norm": 0.01396789401769638,
+      "learning_rate": 1.956704293118222e-06,
+      "loss": 0.3711,
+      "step": 1423
+    },
+    {
+      "epoch": 0.9139922978177151,
+      "grad_norm": 0.013047968968749046,
+      "learning_rate": 1.9280105120003333e-06,
+      "loss": 0.3691,
+      "step": 1424
+    },
+    {
+      "epoch": 0.9146341463414634,
+      "grad_norm": 0.01341619435697794,
+      "learning_rate": 1.8995245391186688e-06,
+      "loss": 0.3691,
+      "step": 1425
+    },
+    {
+      "epoch": 0.9152759948652118,
+      "grad_norm": 0.014094233512878418,
+      "learning_rate": 1.87124649761396e-06,
+      "loss": 0.3672,
+      "step": 1426
+    },
+    {
+      "epoch": 0.9159178433889602,
+      "grad_norm": 0.012613237835466862,
+      "learning_rate": 1.8431765097280784e-06,
+      "loss": 0.3633,
+      "step": 1427
+    },
+    {
+      "epoch": 0.9165596919127086,
+      "grad_norm": 0.013313335366547108,
+      "learning_rate": 1.8153146968035107e-06,
+      "loss": 0.3828,
+      "step": 1428
+    },
+    {
+      "epoch": 0.917201540436457,
+      "grad_norm": 0.0134997284039855,
+      "learning_rate": 1.7876611792828346e-06,
+      "loss": 0.4023,
+      "step": 1429
+    },
+    {
+      "epoch": 0.9178433889602053,
+      "grad_norm": 0.013145639561116695,
+      "learning_rate": 1.7602160767081822e-06,
+      "loss": 0.3535,
+      "step": 1430
+    },
+    {
+      "epoch": 0.9184852374839538,
+      "grad_norm": 0.014171122573316097,
+      "learning_rate": 1.7329795077207556e-06,
+      "loss": 0.375,
+      "step": 1431
+    },
+    {
+      "epoch": 0.9191270860077022,
+      "grad_norm": 0.015409087762236595,
+      "learning_rate": 1.7059515900603007e-06,
+      "loss": 0.373,
+      "step": 1432
+    },
+    {
+      "epoch": 0.9197689345314506,
+      "grad_norm": 0.013116035610437393,
+      "learning_rate": 1.6791324405645626e-06,
+      "loss": 0.3672,
+      "step": 1433
+    },
+    {
+      "epoch": 0.920410783055199,
+      "grad_norm": 0.012658379971981049,
+      "learning_rate": 1.6525221751688636e-06,
+      "loss": 0.3594,
+      "step": 1434
+    },
+    {
+      "epoch": 0.9210526315789473,
+      "grad_norm": 0.015524683520197868,
+      "learning_rate": 1.6261209089054986e-06,
+      "loss": 0.3828,
+      "step": 1435
+    },
+    {
+      "epoch": 0.9216944801026957,
+      "grad_norm": 0.013421234674751759,
+      "learning_rate": 1.5999287559033127e-06,
+      "loss": 0.3711,
+      "step": 1436
+    },
+    {
+      "epoch": 0.9223363286264442,
+      "grad_norm": 0.012796991504728794,
+      "learning_rate": 1.5739458293871968e-06,
+      "loss": 0.3633,
+      "step": 1437
+    },
+    {
+      "epoch": 0.9229781771501926,
+      "grad_norm": 0.01368678081780672,
+      "learning_rate": 1.5481722416775479e-06,
+      "loss": 0.3789,
+      "step": 1438
+    },
+    {
+      "epoch": 0.923620025673941,
+      "grad_norm": 0.014133045449852943,
+      "learning_rate": 1.5226081041898766e-06,
+      "loss": 0.3867,
+      "step": 1439
+    },
+    {
+      "epoch": 0.9242618741976893,
+      "grad_norm": 0.012482895515859127,
+      "learning_rate": 1.4972535274342225e-06,
+      "loss": 0.3809,
+      "step": 1440
+    },
+    {
+      "epoch": 0.9249037227214377,
+      "grad_norm": 0.011843664571642876,
+      "learning_rate": 1.472108621014745e-06,
+      "loss": 0.3535,
+      "step": 1441
+    },
+    {
+      "epoch": 0.9255455712451861,
+      "grad_norm": 0.013735570013523102,
+      "learning_rate": 1.4471734936292281e-06,
+      "loss": 0.3828,
+      "step": 1442
+    },
+    {
+      "epoch": 0.9261874197689345,
+      "grad_norm": 0.01336585357785225,
+      "learning_rate": 1.422448253068609e-06,
+      "loss": 0.3496,
+      "step": 1443
+    },
+    {
+      "epoch": 0.926829268292683,
+      "grad_norm": 0.01314416341483593,
+      "learning_rate": 1.3979330062165119e-06,
+      "loss": 0.3906,
+      "step": 1444
+    },
+    {
+      "epoch": 0.9274711168164314,
+      "grad_norm": 0.013479986228048801,
+      "learning_rate": 1.3736278590487927e-06,
+      "loss": 0.375,
+      "step": 1445
+    },
+    {
+      "epoch": 0.9281129653401797,
+      "grad_norm": 0.012583770789206028,
+      "learning_rate": 1.349532916633084e-06,
+      "loss": 0.375,
+      "step": 1446
+    },
+    {
+      "epoch": 0.9287548138639281,
+      "grad_norm": 0.012261179275810719,
+      "learning_rate": 1.3256482831283278e-06,
+      "loss": 0.3711,
+      "step": 1447
+    },
+    {
+      "epoch": 0.9293966623876765,
+      "grad_norm": 0.012957144528627396,
+      "learning_rate": 1.3019740617843167e-06,
+      "loss": 0.4043,
+      "step": 1448
+    },
+    {
+      "epoch": 0.9300385109114249,
+      "grad_norm": 0.012559046968817711,
+      "learning_rate": 1.2785103549412814e-06,
+      "loss": 0.3789,
+      "step": 1449
+    },
+    {
+      "epoch": 0.9306803594351734,
+      "grad_norm": 0.012377286329865456,
+      "learning_rate": 1.2552572640294247e-06,
+      "loss": 0.3711,
+      "step": 1450
+    },
+    {
+      "epoch": 0.9313222079589217,
+      "grad_norm": 0.012806874699890614,
+      "learning_rate": 1.2322148895684838e-06,
+      "loss": 0.3574,
+      "step": 1451
+    },
+    {
+      "epoch": 0.9319640564826701,
+      "grad_norm": 0.015327047556638718,
+      "learning_rate": 1.209383331167302e-06,
+      "loss": 0.3691,
+      "step": 1452
+    },
+    {
+      "epoch": 0.9326059050064185,
+      "grad_norm": 0.011319216340780258,
+      "learning_rate": 1.1867626875233851e-06,
+      "loss": 0.3613,
+      "step": 1453
+    },
+    {
+      "epoch": 0.9332477535301669,
+      "grad_norm": 0.012484993785619736,
+      "learning_rate": 1.1643530564225013e-06,
+      "loss": 0.3711,
+      "step": 1454
+    },
+    {
+      "epoch": 0.9338896020539152,
+      "grad_norm": 0.011937733739614487,
+      "learning_rate": 1.1421545347382378e-06,
+      "loss": 0.3496,
+      "step": 1455
+    },
+    {
+      "epoch": 0.9345314505776636,
+      "grad_norm": 0.012133189477026463,
+      "learning_rate": 1.1201672184315726e-06,
+      "loss": 0.3496,
+      "step": 1456
+    },
+    {
+      "epoch": 0.9351732991014121,
+      "grad_norm": 0.013368112966418266,
+      "learning_rate": 1.0983912025505028e-06,
+      "loss": 0.3652,
+      "step": 1457
+    },
+    {
+      "epoch": 0.9358151476251605,
+      "grad_norm": 0.011588167399168015,
+      "learning_rate": 1.076826581229562e-06,
+      "loss": 0.3496,
+      "step": 1458
+    },
+    {
+      "epoch": 0.9364569961489089,
+      "grad_norm": 0.013561406172811985,
+      "learning_rate": 1.0554734476894923e-06,
+      "loss": 0.3555,
+      "step": 1459
+    },
+    {
+      "epoch": 0.9370988446726572,
+      "grad_norm": 0.012453819625079632,
+      "learning_rate": 1.0343318942367951e-06,
+      "loss": 0.3906,
+      "step": 1460
+    },
+    {
+      "epoch": 0.9377406931964056,
+      "grad_norm": 0.01237612683326006,
+      "learning_rate": 1.0134020122633313e-06,
+      "loss": 0.373,
+      "step": 1461
+    },
+    {
+      "epoch": 0.938382541720154,
+      "grad_norm": 0.013195016421377659,
+      "learning_rate": 9.926838922459492e-07,
+      "loss": 0.3867,
+      "step": 1462
+    },
+    {
+      "epoch": 0.9390243902439024,
+      "grad_norm": 0.014692552387714386,
+      "learning_rate": 9.721776237460734e-07,
+      "loss": 0.3672,
+      "step": 1463
+    },
+    {
+      "epoch": 0.9396662387676509,
+      "grad_norm": 0.012344533577561378,
+      "learning_rate": 9.518832954093282e-07,
+      "loss": 0.3789,
+      "step": 1464
+    },
+    {
+      "epoch": 0.9403080872913993,
+      "grad_norm": 0.012914840131998062,
+      "learning_rate": 9.31800994965154e-07,
+      "loss": 0.3711,
+      "step": 1465
+    },
+    {
+      "epoch": 0.9409499358151476,
+      "grad_norm": 0.012996658682823181,
+      "learning_rate": 9.119308092264078e-07,
+      "loss": 0.3633,
+      "step": 1466
+    },
+    {
+      "epoch": 0.941591784338896,
+      "grad_norm": 0.0120243513956666,
+      "learning_rate": 8.92272824089041e-07,
+      "loss": 0.3672,
+      "step": 1467
+    },
+    {
+      "epoch": 0.9422336328626444,
+      "grad_norm": 0.012632016092538834,
+      "learning_rate": 8.728271245316555e-07,
+      "loss": 0.3672,
+      "step": 1468
+    },
+    {
+      "epoch": 0.9428754813863928,
+      "grad_norm": 0.012309476733207703,
+      "learning_rate": 8.535937946151983e-07,
+      "loss": 0.3848,
+      "step": 1469
+    },
+    {
+      "epoch": 0.9435173299101413,
+      "grad_norm": 0.0127309774979949,
+      "learning_rate": 8.345729174825623e-07,
+      "loss": 0.375,
+      "step": 1470
+    },
+    {
+      "epoch": 0.9441591784338896,
+      "grad_norm": 0.013301864266395569,
+      "learning_rate": 8.157645753582299e-07,
+      "loss": 0.375,
+      "step": 1471
+    },
+    {
+      "epoch": 0.944801026957638,
+      "grad_norm": 0.012385779060423374,
+      "learning_rate": 7.971688495479468e-07,
+      "loss": 0.373,
+      "step": 1472
+    },
+    {
+      "epoch": 0.9454428754813864,
+      "grad_norm": 0.012698372825980186,
+      "learning_rate": 7.78785820438338e-07,
+      "loss": 0.3672,
+      "step": 1473
+    },
+    {
+      "epoch": 0.9460847240051348,
+      "grad_norm": 0.012753445655107498,
+      "learning_rate": 7.606155674965698e-07,
+      "loss": 0.375,
+      "step": 1474
+    },
+    {
+      "epoch": 0.9467265725288831,
+      "grad_norm": 0.01262593548744917,
+      "learning_rate": 7.426581692700052e-07,
+      "loss": 0.3711,
+      "step": 1475
+    },
+    {
+      "epoch": 0.9473684210526315,
+      "grad_norm": 0.012594195082783699,
+      "learning_rate": 7.24913703385871e-07,
+      "loss": 0.3789,
+      "step": 1476
+    },
+    {
+      "epoch": 0.94801026957638,
+      "grad_norm": 0.011902299709618092,
+      "learning_rate": 7.073822465509194e-07,
+      "loss": 0.377,
+      "step": 1477
+    },
+    {
+      "epoch": 0.9486521181001284,
+      "grad_norm": 0.013070139102637768,
+      "learning_rate": 6.900638745511057e-07,
+      "loss": 0.3789,
+      "step": 1478
+    },
+    {
+      "epoch": 0.9492939666238768,
+      "grad_norm": 0.013074691407382488,
+      "learning_rate": 6.729586622512274e-07,
+      "loss": 0.3926,
+      "step": 1479
+    },
+    {
+      "epoch": 0.9499358151476252,
+      "grad_norm": 0.012703335843980312,
+      "learning_rate": 6.560666835946416e-07,
+      "loss": 0.377,
+      "step": 1480
+    },
+    {
+      "epoch": 0.9505776636713735,
+      "grad_norm": 0.012635687366127968,
+      "learning_rate": 6.39388011602915e-07,
+      "loss": 0.4004,
+      "step": 1481
+    },
+    {
+      "epoch": 0.9512195121951219,
+      "grad_norm": 0.013130598701536655,
+      "learning_rate": 6.229227183755293e-07,
+      "loss": 0.3633,
+      "step": 1482
+    },
+    {
+      "epoch": 0.9518613607188704,
+      "grad_norm": 0.011977552436292171,
+      "learning_rate": 6.0667087508956e-07,
+      "loss": 0.3652,
+      "step": 1483
+    },
+    {
+      "epoch": 0.9525032092426188,
+      "grad_norm": 0.013489103876054287,
+      "learning_rate": 5.906325519993705e-07,
+      "loss": 0.3691,
+      "step": 1484
+    },
+    {
+      "epoch": 0.9531450577663672,
+      "grad_norm": 0.013404084369540215,
+      "learning_rate": 5.748078184363015e-07,
+      "loss": 0.3711,
+      "step": 1485
+    },
+    {
+      "epoch": 0.9537869062901155,
+      "grad_norm": 0.012206465937197208,
+      "learning_rate": 5.591967428083822e-07,
+      "loss": 0.3672,
+      "step": 1486
+    },
+    {
+      "epoch": 0.9544287548138639,
+      "grad_norm": 0.01376590970903635,
+      "learning_rate": 5.437993926000195e-07,
+      "loss": 0.3965,
+      "step": 1487
+    },
+    {
+      "epoch": 0.9550706033376123,
+      "grad_norm": 0.011835059151053429,
+      "learning_rate": 5.286158343717429e-07,
+      "loss": 0.3438,
+      "step": 1488
+    },
+    {
+      "epoch": 0.9557124518613607,
+      "grad_norm": 0.012890417128801346,
+      "learning_rate": 5.136461337598486e-07,
+      "loss": 0.3848,
+      "step": 1489
+    },
+    {
+      "epoch": 0.9563543003851092,
+      "grad_norm": 0.013126579113304615,
+      "learning_rate": 4.988903554761948e-07,
+      "loss": 0.4023,
+      "step": 1490
+    },
+    {
+      "epoch": 0.9569961489088575,
+      "grad_norm": 0.013360566459596157,
+      "learning_rate": 4.843485633078681e-07,
+      "loss": 0.3867,
+      "step": 1491
+    },
+    {
+      "epoch": 0.9576379974326059,
+      "grad_norm": 0.012557647190988064,
+      "learning_rate": 4.700208201169232e-07,
+      "loss": 0.3477,
+      "step": 1492
+    },
+    {
+      "epoch": 0.9582798459563543,
+      "grad_norm": 0.012240637093782425,
+      "learning_rate": 4.559071878401211e-07,
+      "loss": 0.3359,
+      "step": 1493
+    },
+    {
+      "epoch": 0.9589216944801027,
+      "grad_norm": 0.011939102783799171,
+      "learning_rate": 4.420077274886414e-07,
+      "loss": 0.3906,
+      "step": 1494
+    },
+    {
+      "epoch": 0.959563543003851,
+      "grad_norm": 0.013211379759013653,
+      "learning_rate": 4.283224991478374e-07,
+      "loss": 0.3809,
+      "step": 1495
+    },
+    {
+      "epoch": 0.9602053915275995,
+      "grad_norm": 0.012315848842263222,
+      "learning_rate": 4.148515619769644e-07,
+      "loss": 0.3809,
+      "step": 1496
+    },
+    {
+      "epoch": 0.9608472400513479,
+      "grad_norm": 0.013882388360798359,
+      "learning_rate": 4.0159497420892976e-07,
+      "loss": 0.3984,
+      "step": 1497
+    },
+    {
+      "epoch": 0.9614890885750963,
+      "grad_norm": 0.014947362244129181,
+      "learning_rate": 3.885527931500377e-07,
+      "loss": 0.3809,
+      "step": 1498
+    },
+    {
+      "epoch": 0.9621309370988447,
+      "grad_norm": 0.011844776570796967,
+      "learning_rate": 3.7572507517974475e-07,
+      "loss": 0.3945,
+      "step": 1499
+    },
+    {
+      "epoch": 0.962772785622593,
+      "grad_norm": 0.011926044709980488,
+      "learning_rate": 3.631118757504159e-07,
+      "loss": 0.3633,
+      "step": 1500
+    },
+    {
+      "epoch": 0.9634146341463414,
+      "grad_norm": 0.012529637664556503,
+      "learning_rate": 3.507132493870857e-07,
+      "loss": 0.3672,
+      "step": 1501
+    },
+    {
+      "epoch": 0.9640564826700898,
+      "grad_norm": 0.012636232189834118,
+      "learning_rate": 3.3852924968720834e-07,
+      "loss": 0.3711,
+      "step": 1502
+    },
+    {
+      "epoch": 0.9646983311938383,
+      "grad_norm": 0.01360595878213644,
+      "learning_rate": 3.2655992932045796e-07,
+      "loss": 0.3906,
+      "step": 1503
+    },
+    {
+      "epoch": 0.9653401797175867,
+      "grad_norm": 0.014324655756354332,
+      "learning_rate": 3.148053400284623e-07,
+      "loss": 0.3516,
+      "step": 1504
+    },
+    {
+      "epoch": 0.9659820282413351,
+      "grad_norm": 0.013343398459255695,
+      "learning_rate": 3.0326553262460255e-07,
+      "loss": 0.3711,
+      "step": 1505
+    },
+    {
+      "epoch": 0.9666238767650834,
+      "grad_norm": 0.01268340926617384,
+      "learning_rate": 2.9194055699380275e-07,
+      "loss": 0.3711,
+      "step": 1506
+    },
+    {
+      "epoch": 0.9672657252888318,
+      "grad_norm": 0.012414415366947651,
+      "learning_rate": 2.8083046209228523e-07,
+      "loss": 0.3867,
+      "step": 1507
+    },
+    {
+      "epoch": 0.9679075738125802,
+      "grad_norm": 0.012590642087161541,
+      "learning_rate": 2.699352959473711e-07,
+      "loss": 0.3633,
+      "step": 1508
+    },
+    {
+      "epoch": 0.9685494223363287,
+      "grad_norm": 0.0130212577059865,
+      "learning_rate": 2.5925510565729115e-07,
+      "loss": 0.377,
+      "step": 1509
+    },
+    {
+      "epoch": 0.9691912708600771,
+      "grad_norm": 0.013702025637030602,
+      "learning_rate": 2.4878993739095857e-07,
+      "loss": 0.4102,
+      "step": 1510
+    },
+    {
+      "epoch": 0.9698331193838254,
+      "grad_norm": 0.013444257900118828,
+      "learning_rate": 2.385398363877689e-07,
+      "loss": 0.3672,
+      "step": 1511
+    },
+    {
+      "epoch": 0.9704749679075738,
+      "grad_norm": 0.011984091252088547,
+      "learning_rate": 2.2850484695743357e-07,
+      "loss": 0.373,
+      "step": 1512
+    },
+    {
+      "epoch": 0.9711168164313222,
+      "grad_norm": 0.01249521505087614,
+      "learning_rate": 2.1868501247974683e-07,
+      "loss": 0.3418,
+      "step": 1513
+    },
+    {
+      "epoch": 0.9717586649550706,
+      "grad_norm": 0.012439033947885036,
+      "learning_rate": 2.090803754044357e-07,
+      "loss": 0.3984,
+      "step": 1514
+    },
+    {
+      "epoch": 0.972400513478819,
+      "grad_norm": 0.012088153511285782,
+      "learning_rate": 1.9969097725094366e-07,
+      "loss": 0.3555,
+      "step": 1515
+    },
+    {
+      "epoch": 0.9730423620025674,
+      "grad_norm": 0.013228670693933964,
+      "learning_rate": 1.9051685860828062e-07,
+      "loss": 0.3828,
+      "step": 1516
+    },
+    {
+      "epoch": 0.9736842105263158,
+      "grad_norm": 0.013256936334073544,
+      "learning_rate": 1.8155805913483427e-07,
+      "loss": 0.3867,
+      "step": 1517
+    },
+    {
+      "epoch": 0.9743260590500642,
+      "grad_norm": 0.0124072739854455,
+      "learning_rate": 1.728146175581924e-07,
+      "loss": 0.3906,
+      "step": 1518
+    },
+    {
+      "epoch": 0.9749679075738126,
+      "grad_norm": 0.012670646421611309,
+      "learning_rate": 1.642865716749875e-07,
+      "loss": 0.3926,
+      "step": 1519
+    },
+    {
+      "epoch": 0.975609756097561,
+      "grad_norm": 0.011904159560799599,
+      "learning_rate": 1.5597395835071915e-07,
+      "loss": 0.3633,
+      "step": 1520
+    },
+    {
+      "epoch": 0.9762516046213093,
+      "grad_norm": 0.014184362255036831,
+      "learning_rate": 1.4787681351960959e-07,
+      "loss": 0.375,
+      "step": 1521
+    },
+    {
+      "epoch": 0.9768934531450578,
+      "grad_norm": 0.012831350788474083,
+      "learning_rate": 1.399951721844428e-07,
+      "loss": 0.3984,
+      "step": 1522
+    },
+    {
+      "epoch": 0.9775353016688062,
+      "grad_norm": 0.012582185678184032,
+      "learning_rate": 1.3232906841641469e-07,
+      "loss": 0.375,
+      "step": 1523
+    },
+    {
+      "epoch": 0.9781771501925546,
+      "grad_norm": 0.013945076614618301,
+      "learning_rate": 1.2487853535497197e-07,
+      "loss": 0.3574,
+      "step": 1524
+    },
+    {
+      "epoch": 0.978818998716303,
+      "grad_norm": 0.013039125129580498,
+      "learning_rate": 1.1764360520769568e-07,
+      "loss": 0.3672,
+      "step": 1525
+    },
+    {
+      "epoch": 0.9794608472400513,
+      "grad_norm": 0.013115467503666878,
+      "learning_rate": 1.1062430925012911e-07,
+      "loss": 0.3633,
+      "step": 1526
+    },
+    {
+      "epoch": 0.9801026957637997,
+      "grad_norm": 0.012725510634481907,
+      "learning_rate": 1.0382067782568338e-07,
+      "loss": 0.3867,
+      "step": 1527
+    },
+    {
+      "epoch": 0.9807445442875481,
+      "grad_norm": 0.013744517229497433,
+      "learning_rate": 9.723274034545981e-08,
+      "loss": 0.3672,
+      "step": 1528
+    },
+    {
+      "epoch": 0.9813863928112966,
+      "grad_norm": 0.01282016932964325,
+      "learning_rate": 9.086052528816113e-08,
+      "loss": 0.3672,
+      "step": 1529
+    },
+    {
+      "epoch": 0.982028241335045,
+      "grad_norm": 0.012451869435608387,
+      "learning_rate": 8.470406019994714e-08,
+      "loss": 0.3789,
+      "step": 1530
+    },
+    {
+      "epoch": 0.9826700898587933,
+      "grad_norm": 0.012845706194639206,
+      "learning_rate": 7.876337169432368e-08,
+      "loss": 0.3906,
+      "step": 1531
+    },
+    {
+      "epoch": 0.9833119383825417,
+      "grad_norm": 0.01277588028460741,
+      "learning_rate": 7.303848545202052e-08,
+      "loss": 0.3984,
+      "step": 1532
+    },
+    {
+      "epoch": 0.9839537869062901,
+      "grad_norm": 0.012288976460695267,
+      "learning_rate": 6.752942622089697e-08,
+      "loss": 0.3828,
+      "step": 1533
+    },
+    {
+      "epoch": 0.9845956354300385,
+      "grad_norm": 0.012748691253364086,
+      "learning_rate": 6.223621781581979e-08,
+      "loss": 0.3984,
+      "step": 1534
+    },
+    {
+      "epoch": 0.9852374839537869,
+      "grad_norm": 0.012774264439940453,
+      "learning_rate": 5.715888311855211e-08,
+      "loss": 0.3711,
+      "step": 1535
+    },
+    {
+      "epoch": 0.9858793324775353,
+      "grad_norm": 0.012290029786527157,
+      "learning_rate": 5.2297444077675784e-08,
+      "loss": 0.3867,
+      "step": 1536
+    },
+    {
+      "epoch": 0.9865211810012837,
+      "grad_norm": 0.012472493574023247,
+      "learning_rate": 4.765192170849697e-08,
+      "loss": 0.3516,
+      "step": 1537
+    },
+    {
+      "epoch": 0.9871630295250321,
+      "grad_norm": 0.012907999567687511,
+      "learning_rate": 4.322233609292403e-08,
+      "loss": 0.3926,
+      "step": 1538
+    },
+    {
+      "epoch": 0.9878048780487805,
+      "grad_norm": 0.013008086942136288,
+      "learning_rate": 3.9008706379412005e-08,
+      "loss": 0.3809,
+      "step": 1539
+    },
+    {
+      "epoch": 0.9884467265725289,
+      "grad_norm": 0.012015080079436302,
+      "learning_rate": 3.5011050782879364e-08,
+      "loss": 0.3691,
+      "step": 1540
+    },
+    {
+      "epoch": 0.9890885750962772,
+      "grad_norm": 0.012214963324368,
+      "learning_rate": 3.122938658460806e-08,
+      "loss": 0.3789,
+      "step": 1541
+    },
+    {
+      "epoch": 0.9897304236200257,
+      "grad_norm": 0.011233390308916569,
+      "learning_rate": 2.7663730132182485e-08,
+      "loss": 0.3516,
+      "step": 1542
+    },
+    {
+      "epoch": 0.9903722721437741,
+      "grad_norm": 0.011511021293699741,
+      "learning_rate": 2.4314096839417323e-08,
+      "loss": 0.3516,
+      "step": 1543
+    },
+    {
+      "epoch": 0.9910141206675225,
+      "grad_norm": 0.013254697434604168,
+      "learning_rate": 2.118050118629089e-08,
+      "loss": 0.3672,
+      "step": 1544
+    },
+    {
+      "epoch": 0.9916559691912709,
+      "grad_norm": 0.013225373812019825,
+      "learning_rate": 1.8262956718884117e-08,
+      "loss": 0.4023,
+      "step": 1545
+    },
+    {
+      "epoch": 0.9922978177150192,
+      "grad_norm": 0.012755938805639744,
+      "learning_rate": 1.5561476049325007e-08,
+      "loss": 0.3809,
+      "step": 1546
+    },
+    {
+      "epoch": 0.9929396662387676,
+      "grad_norm": 0.012606845237314701,
+      "learning_rate": 1.3076070855710942e-08,
+      "loss": 0.3555,
+      "step": 1547
+    },
+    {
+      "epoch": 0.993581514762516,
+      "grad_norm": 0.01340476330369711,
+      "learning_rate": 1.0806751882092014e-08,
+      "loss": 0.4023,
+      "step": 1548
+    },
+    {
+      "epoch": 0.9942233632862645,
+      "grad_norm": 0.013462042436003685,
+      "learning_rate": 8.753528938409972e-09,
+      "loss": 0.3516,
+      "step": 1549
+    },
+    {
+      "epoch": 0.9948652118100129,
+      "grad_norm": 0.01249738223850727,
+      "learning_rate": 6.9164109004427046e-09,
+      "loss": 0.3887,
+      "step": 1550
+    },
+    {
+      "epoch": 0.9955070603337612,
+      "grad_norm": 0.014013971202075481,
+      "learning_rate": 5.295405709787593e-09,
+      "loss": 0.4062,
+      "step": 1551
+    },
+    {
+      "epoch": 0.9961489088575096,
+      "grad_norm": 0.012625960633158684,
+      "learning_rate": 3.890520373817097e-09,
+      "loss": 0.3574,
+      "step": 1552
+    },
+    {
+      "epoch": 0.996790757381258,
+      "grad_norm": 0.012531465850770473,
+      "learning_rate": 2.7017609656454503e-09,
+      "loss": 0.3594,
+      "step": 1553
+    },
+    {
+      "epoch": 0.9974326059050064,
+      "grad_norm": 0.012438458390533924,
+      "learning_rate": 1.7291326241064553e-09,
+      "loss": 0.3711,
+      "step": 1554
+    },
+    {
+      "epoch": 0.9980744544287549,
+      "grad_norm": 0.013172676786780357,
+      "learning_rate": 9.726395537312806e-10,
+      "loss": 0.3613,
+      "step": 1555
+    },
+    {
+      "epoch": 0.9987163029525032,
+      "grad_norm": 0.013000345788896084,
+      "learning_rate": 4.3228502473180445e-10,
+      "loss": 0.3906,
+      "step": 1556
+    },
+    {
+      "epoch": 0.9993581514762516,
+      "grad_norm": 0.012555932626128197,
+      "learning_rate": 1.0807137297841329e-10,
+      "loss": 0.3828,
+      "step": 1557
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.013687052763998508,
+      "learning_rate": 0.0,
+      "loss": 0.373,
+      "step": 1558
+    },
+    {
+      "epoch": 1.0,
+      "step": 1558,
+      "total_flos": 0.0,
+      "train_loss": 0.16030793685815148,
+      "train_runtime": 10881.6807,
+      "train_samples_per_second": 586.271,
+      "train_steps_per_second": 0.143
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 1558,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 100,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 0.0,
+  "train_batch_size": 64,
+  "trial_name": null,
+  "trial_params": null
+}