diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,9492 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.9991668980838657,
+  "eval_steps": 500,
+  "global_step": 1350,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0022216051096917524,
+      "grad_norm": 0.06047070026397705,
+      "learning_rate": 2.2222222222222223e-05,
+      "loss": 10.6079,
+      "step": 1
+    },
+    {
+      "epoch": 0.004443210219383505,
+      "grad_norm": 0.04731635004281998,
+      "learning_rate": 4.4444444444444447e-05,
+      "loss": 10.247,
+      "step": 2
+    },
+    {
+      "epoch": 0.006664815329075257,
+      "grad_norm": 0.049777183681726456,
+      "learning_rate": 6.666666666666667e-05,
+      "loss": 10.4556,
+      "step": 3
+    },
+    {
+      "epoch": 0.00888642043876701,
+      "grad_norm": 0.04873029887676239,
+      "learning_rate": 8.888888888888889e-05,
+      "loss": 10.7528,
+      "step": 4
+    },
+    {
+      "epoch": 0.011108025548458762,
+      "grad_norm": 0.05402935668826103,
+      "learning_rate": 0.0001111111111111111,
+      "loss": 10.7387,
+      "step": 5
+    },
+    {
+      "epoch": 0.013329630658150514,
+      "grad_norm": 0.050039034336805344,
+      "learning_rate": 0.00013333333333333334,
+      "loss": 10.3825,
+      "step": 6
+    },
+    {
+      "epoch": 0.015551235767842266,
+      "grad_norm": 0.05475683882832527,
+      "learning_rate": 0.00015555555555555556,
+      "loss": 10.7209,
+      "step": 7
+    },
+    {
+      "epoch": 0.01777284087753402,
+      "grad_norm": 0.05116181820631027,
+      "learning_rate": 0.00017777777777777779,
+      "loss": 10.5757,
+      "step": 8
+    },
+    {
+      "epoch": 0.01999444598722577,
+      "grad_norm": 0.05796043574810028,
+      "learning_rate": 0.0002,
+      "loss": 10.7621,
+      "step": 9
+    },
+    {
+      "epoch": 0.022216051096917523,
+      "grad_norm": 0.050162676721811295,
+      "learning_rate": 0.0002222222222222222,
+      "loss": 10.6925,
+      "step": 10
+    },
+    {
+      "epoch": 0.024437656206609277,
+      "grad_norm": 0.050715889781713486,
+      "learning_rate": 0.0002444444444444445,
+      "loss": 10.5473,
+      "step": 11
+    },
+    {
+      "epoch": 0.026659261316301027,
+      "grad_norm": 0.0544884130358696,
+      "learning_rate": 0.0002666666666666667,
+      "loss": 10.5229,
+      "step": 12
+    },
+    {
+      "epoch": 0.02888086642599278,
+      "grad_norm": 0.05708073824644089,
+      "learning_rate": 0.0002888888888888889,
+      "loss": 10.4162,
+      "step": 13
+    },
+    {
+      "epoch": 0.03110247153568453,
+      "grad_norm": 0.05594116821885109,
+      "learning_rate": 0.0003111111111111111,
+      "loss": 10.8703,
+      "step": 14
+    },
+    {
+      "epoch": 0.03332407664537628,
+      "grad_norm": 0.05033649876713753,
+      "learning_rate": 0.0003333333333333333,
+      "loss": 10.3323,
+      "step": 15
+    },
+    {
+      "epoch": 0.03554568175506804,
+      "grad_norm": 0.053412653505802155,
+      "learning_rate": 0.00035555555555555557,
+      "loss": 10.4398,
+      "step": 16
+    },
+    {
+      "epoch": 0.03776728686475979,
+      "grad_norm": 0.054381825029850006,
+      "learning_rate": 0.00037777777777777777,
+      "loss": 10.3598,
+      "step": 17
+    },
+    {
+      "epoch": 0.03998889197445154,
+      "grad_norm": 0.055913932621479034,
+      "learning_rate": 0.0004,
+      "loss": 10.3756,
+      "step": 18
+    },
+    {
+      "epoch": 0.042210497084143296,
+      "grad_norm": 0.056334175169467926,
+      "learning_rate": 0.00042222222222222227,
+      "loss": 10.7314,
+      "step": 19
+    },
+    {
+      "epoch": 0.044432102193835046,
+      "grad_norm": 0.05052949860692024,
+      "learning_rate": 0.0004444444444444444,
+      "loss": 10.6529,
+      "step": 20
+    },
+    {
+      "epoch": 0.0466537073035268,
+      "grad_norm": 0.05472550168633461,
+      "learning_rate": 0.00046666666666666666,
+      "loss": 10.4808,
+      "step": 21
+    },
+    {
+      "epoch": 0.048875312413218554,
+      "grad_norm": 0.04639170691370964,
+      "learning_rate": 0.000488888888888889,
+      "loss": 10.4715,
+      "step": 22
+    },
+    {
+      "epoch": 0.051096917522910304,
+      "grad_norm": 0.05662548914551735,
+      "learning_rate": 0.0005111111111111112,
+      "loss": 10.3969,
+      "step": 23
+    },
+    {
+      "epoch": 0.053318522632602054,
+      "grad_norm": 0.04989445582032204,
+      "learning_rate": 0.0005333333333333334,
+      "loss": 10.3438,
+      "step": 24
+    },
+    {
+      "epoch": 0.055540127742293804,
+      "grad_norm": 0.05659613013267517,
+      "learning_rate": 0.0005555555555555556,
+      "loss": 10.2444,
+      "step": 25
+    },
+    {
+      "epoch": 0.05776173285198556,
+      "grad_norm": 0.05034107714891434,
+      "learning_rate": 0.0005777777777777778,
+      "loss": 10.47,
+      "step": 26
+    },
+    {
+      "epoch": 0.05998333796167731,
+      "grad_norm": 0.053235191851854324,
+      "learning_rate": 0.0006000000000000001,
+      "loss": 10.4771,
+      "step": 27
+    },
+    {
+      "epoch": 0.06220494307136906,
+      "grad_norm": 0.052242521196603775,
+      "learning_rate": 0.0006222222222222223,
+      "loss": 10.6162,
+      "step": 28
+    },
+    {
+      "epoch": 0.06442654818106082,
+      "grad_norm": 0.051469966769218445,
+      "learning_rate": 0.0006444444444444444,
+      "loss": 10.6046,
+      "step": 29
+    },
+    {
+      "epoch": 0.06664815329075256,
+      "grad_norm": 0.048590026795864105,
+      "learning_rate": 0.0006666666666666666,
+      "loss": 10.5241,
+      "step": 30
+    },
+    {
+      "epoch": 0.06886975840044432,
+      "grad_norm": 0.050730764865875244,
+      "learning_rate": 0.000688888888888889,
+      "loss": 10.3362,
+      "step": 31
+    },
+    {
+      "epoch": 0.07109136351013608,
+      "grad_norm": 0.05845819041132927,
+      "learning_rate": 0.0007111111111111111,
+      "loss": 10.4505,
+      "step": 32
+    },
+    {
+      "epoch": 0.07331296861982782,
+      "grad_norm": 0.053127389401197433,
+      "learning_rate": 0.0007333333333333333,
+      "loss": 10.6673,
+      "step": 33
+    },
+    {
+      "epoch": 0.07553457372951958,
+      "grad_norm": 0.05221354588866234,
+      "learning_rate": 0.0007555555555555555,
+      "loss": 10.4441,
+      "step": 34
+    },
+    {
+      "epoch": 0.07775617883921133,
+      "grad_norm": 0.052416905760765076,
+      "learning_rate": 0.0007777777777777777,
+      "loss": 10.5406,
+      "step": 35
+    },
+    {
+      "epoch": 0.07997778394890308,
+      "grad_norm": 0.055885300040245056,
+      "learning_rate": 0.0008,
+      "loss": 10.6719,
+      "step": 36
+    },
+    {
+      "epoch": 0.08219938905859484,
+      "grad_norm": 0.05125787854194641,
+      "learning_rate": 0.0008222222222222222,
+      "loss": 10.3831,
+      "step": 37
+    },
+    {
+      "epoch": 0.08442099416828659,
+      "grad_norm": 0.05321040004491806,
+      "learning_rate": 0.0008444444444444445,
+      "loss": 10.5362,
+      "step": 38
+    },
+    {
+      "epoch": 0.08664259927797834,
+      "grad_norm": 0.048330310732126236,
+      "learning_rate": 0.0008666666666666666,
+      "loss": 10.618,
+      "step": 39
+    },
+    {
+      "epoch": 0.08886420438767009,
+      "grad_norm": 0.05690504238009453,
+      "learning_rate": 0.0008888888888888888,
+      "loss": 10.3795,
+      "step": 40
+    },
+    {
+      "epoch": 0.09108580949736185,
+      "grad_norm": 0.050180643796920776,
+      "learning_rate": 0.0009111111111111111,
+      "loss": 10.5554,
+      "step": 41
+    },
+    {
+      "epoch": 0.0933074146070536,
+      "grad_norm": 0.05029109865427017,
+      "learning_rate": 0.0009333333333333333,
+      "loss": 10.3326,
+      "step": 42
+    },
+    {
+      "epoch": 0.09552901971674535,
+      "grad_norm": 0.05197408050298691,
+      "learning_rate": 0.0009555555555555556,
+      "loss": 10.2956,
+      "step": 43
+    },
+    {
+      "epoch": 0.09775062482643711,
+      "grad_norm": 0.0569787472486496,
+      "learning_rate": 0.000977777777777778,
+      "loss": 10.3524,
+      "step": 44
+    },
+    {
+      "epoch": 0.09997222993612885,
+      "grad_norm": 0.051295869052410126,
+      "learning_rate": 0.001,
+      "loss": 10.6064,
+      "step": 45
+    },
+    {
+      "epoch": 0.10219383504582061,
+      "grad_norm": 0.043966952711343765,
+      "learning_rate": 0.0010222222222222223,
+      "loss": 10.4717,
+      "step": 46
+    },
+    {
+      "epoch": 0.10441544015551235,
+      "grad_norm": 0.05129879713058472,
+      "learning_rate": 0.0010444444444444444,
+      "loss": 10.3595,
+      "step": 47
+    },
+    {
+      "epoch": 0.10663704526520411,
+      "grad_norm": 0.053009241819381714,
+      "learning_rate": 0.0010666666666666667,
+      "loss": 10.5078,
+      "step": 48
+    },
+    {
+      "epoch": 0.10885865037489587,
+      "grad_norm": 0.04631511867046356,
+      "learning_rate": 0.001088888888888889,
+      "loss": 10.456,
+      "step": 49
+    },
+    {
+      "epoch": 0.11108025548458761,
+      "grad_norm": 0.055996429175138474,
+      "learning_rate": 0.0011111111111111111,
+      "loss": 10.1887,
+      "step": 50
+    },
+    {
+      "epoch": 0.11330186059427937,
+      "grad_norm": 0.0516912117600441,
+      "learning_rate": 0.0011333333333333334,
+      "loss": 10.0742,
+      "step": 51
+    },
+    {
+      "epoch": 0.11552346570397112,
+      "grad_norm": 0.053418293595314026,
+      "learning_rate": 0.0011555555555555555,
+      "loss": 10.4041,
+      "step": 52
+    },
+    {
+      "epoch": 0.11774507081366287,
+      "grad_norm": 0.05349532887339592,
+      "learning_rate": 0.0011777777777777778,
+      "loss": 10.1057,
+      "step": 53
+    },
+    {
+      "epoch": 0.11996667592335462,
+      "grad_norm": 0.05413784086704254,
+      "learning_rate": 0.0012000000000000001,
+      "loss": 10.2958,
+      "step": 54
+    },
+    {
+      "epoch": 0.12218828103304638,
+      "grad_norm": 0.04732006415724754,
+      "learning_rate": 0.0012222222222222222,
+      "loss": 10.229,
+      "step": 55
+    },
+    {
+      "epoch": 0.12440988614273812,
+      "grad_norm": 0.06012426316738129,
+      "learning_rate": 0.0012444444444444445,
+      "loss": 10.0901,
+      "step": 56
+    },
+    {
+      "epoch": 0.12663149125242987,
+      "grad_norm": 0.05074010044336319,
+      "learning_rate": 0.0012666666666666666,
+      "loss": 10.273,
+      "step": 57
+    },
+    {
+      "epoch": 0.12885309636212164,
+      "grad_norm": 0.048709437251091,
+      "learning_rate": 0.001288888888888889,
+      "loss": 10.1621,
+      "step": 58
+    },
+    {
+      "epoch": 0.13107470147181338,
+      "grad_norm": 0.05393685773015022,
+      "learning_rate": 0.0013111111111111112,
+      "loss": 10.0042,
+      "step": 59
+    },
+    {
+      "epoch": 0.13329630658150513,
+      "grad_norm": 0.04674612730741501,
+      "learning_rate": 0.0013333333333333333,
+      "loss": 10.2969,
+      "step": 60
+    },
+    {
+      "epoch": 0.1355179116911969,
+      "grad_norm": 0.05694056674838066,
+      "learning_rate": 0.0013555555555555556,
+      "loss": 10.151,
+      "step": 61
+    },
+    {
+      "epoch": 0.13773951680088864,
+      "grad_norm": 0.057547781616449356,
+      "learning_rate": 0.001377777777777778,
+      "loss": 10.5158,
+      "step": 62
+    },
+    {
+      "epoch": 0.13996112191058038,
+      "grad_norm": 0.05202538147568703,
+      "learning_rate": 0.0014,
+      "loss": 10.0441,
+      "step": 63
+    },
+    {
+      "epoch": 0.14218272702027215,
+      "grad_norm": 0.05227857083082199,
+      "learning_rate": 0.0014222222222222223,
+      "loss": 10.0009,
+      "step": 64
+    },
+    {
+      "epoch": 0.1444043321299639,
+      "grad_norm": 0.05450652912259102,
+      "learning_rate": 0.0014444444444444444,
+      "loss": 10.1787,
+      "step": 65
+    },
+    {
+      "epoch": 0.14662593723965564,
+      "grad_norm": 0.04978035390377045,
+      "learning_rate": 0.0014666666666666667,
+      "loss": 10.6574,
+      "step": 66
+    },
+    {
+      "epoch": 0.1488475423493474,
+      "grad_norm": 0.04803209751844406,
+      "learning_rate": 0.001488888888888889,
+      "loss": 10.1993,
+      "step": 67
+    },
+    {
+      "epoch": 0.15106914745903915,
+      "grad_norm": 0.04095118120312691,
+      "learning_rate": 0.001511111111111111,
+      "loss": 9.9589,
+      "step": 68
+    },
+    {
+      "epoch": 0.1532907525687309,
+      "grad_norm": 0.049227792769670486,
+      "learning_rate": 0.0015333333333333332,
+      "loss": 10.1613,
+      "step": 69
+    },
+    {
+      "epoch": 0.15551235767842267,
+      "grad_norm": 0.056645046919584274,
+      "learning_rate": 0.0015555555555555555,
+      "loss": 10.157,
+      "step": 70
+    },
+    {
+      "epoch": 0.1577339627881144,
+      "grad_norm": 0.04908820986747742,
+      "learning_rate": 0.0015777777777777778,
+      "loss": 10.3338,
+      "step": 71
+    },
+    {
+      "epoch": 0.15995556789780616,
+      "grad_norm": 0.048528026789426804,
+      "learning_rate": 0.0016,
+      "loss": 10.1877,
+      "step": 72
+    },
+    {
+      "epoch": 0.16217717300749793,
+      "grad_norm": 0.0497007817029953,
+      "learning_rate": 0.0016222222222222222,
+      "loss": 10.0367,
+      "step": 73
+    },
+    {
+      "epoch": 0.16439877811718967,
+      "grad_norm": 0.05009305849671364,
+      "learning_rate": 0.0016444444444444445,
+      "loss": 9.9939,
+      "step": 74
+    },
+    {
+      "epoch": 0.1666203832268814,
+      "grad_norm": 0.04887579008936882,
+      "learning_rate": 0.0016666666666666668,
+      "loss": 10.2447,
+      "step": 75
+    },
+    {
+      "epoch": 0.16884198833657318,
+      "grad_norm": 0.04899114370346069,
+      "learning_rate": 0.001688888888888889,
+      "loss": 10.0718,
+      "step": 76
+    },
+    {
+      "epoch": 0.17106359344626493,
+      "grad_norm": 0.05017256736755371,
+      "learning_rate": 0.0017111111111111114,
+      "loss": 9.7071,
+      "step": 77
+    },
+    {
+      "epoch": 0.17328519855595667,
+      "grad_norm": 0.05062386393547058,
+      "learning_rate": 0.0017333333333333333,
+      "loss": 10.3495,
+      "step": 78
+    },
+    {
+      "epoch": 0.17550680366564844,
+      "grad_norm": 0.054166313260793686,
+      "learning_rate": 0.0017555555555555556,
+      "loss": 9.8393,
+      "step": 79
+    },
+    {
+      "epoch": 0.17772840877534019,
+      "grad_norm": 0.04704143479466438,
+      "learning_rate": 0.0017777777777777776,
+      "loss": 9.8577,
+      "step": 80
+    },
+    {
+      "epoch": 0.17995001388503193,
+      "grad_norm": 0.0470828153192997,
+      "learning_rate": 0.0018,
+      "loss": 10.2248,
+      "step": 81
+    },
+    {
+      "epoch": 0.1821716189947237,
+      "grad_norm": 0.046098653227090836,
+      "learning_rate": 0.0018222222222222223,
+      "loss": 10.0535,
+      "step": 82
+    },
+    {
+      "epoch": 0.18439322410441544,
+      "grad_norm": 0.0534447506070137,
+      "learning_rate": 0.0018444444444444446,
+      "loss": 9.9007,
+      "step": 83
+    },
+    {
+      "epoch": 0.1866148292141072,
+      "grad_norm": 0.05775776132941246,
+      "learning_rate": 0.0018666666666666666,
+      "loss": 10.0096,
+      "step": 84
+    },
+    {
+      "epoch": 0.18883643432379896,
+      "grad_norm": 0.0538940355181694,
+      "learning_rate": 0.001888888888888889,
+      "loss": 9.8869,
+      "step": 85
+    },
+    {
+      "epoch": 0.1910580394334907,
+      "grad_norm": 0.055318545550107956,
+      "learning_rate": 0.0019111111111111113,
+      "loss": 10.0082,
+      "step": 86
+    },
+    {
+      "epoch": 0.19327964454318244,
+      "grad_norm": 0.04999914392828941,
+      "learning_rate": 0.0019333333333333336,
+      "loss": 9.9907,
+      "step": 87
+    },
+    {
+      "epoch": 0.19550124965287421,
+      "grad_norm": 0.04839245602488518,
+      "learning_rate": 0.001955555555555556,
+      "loss": 9.6986,
+      "step": 88
+    },
+    {
+      "epoch": 0.19772285476256596,
+      "grad_norm": 0.051004353910684586,
+      "learning_rate": 0.0019777777777777775,
+      "loss": 10.0626,
+      "step": 89
+    },
+    {
+      "epoch": 0.1999444598722577,
+      "grad_norm": 0.050714798271656036,
+      "learning_rate": 0.002,
+      "loss": 9.4952,
+      "step": 90
+    },
+    {
+      "epoch": 0.20216606498194944,
+      "grad_norm": 0.046029288321733475,
+      "learning_rate": 0.002022222222222222,
+      "loss": 10.0677,
+      "step": 91
+    },
+    {
+      "epoch": 0.20438767009164122,
+      "grad_norm": 0.052206166088581085,
+      "learning_rate": 0.0020444444444444447,
+      "loss": 9.7142,
+      "step": 92
+    },
+    {
+      "epoch": 0.20660927520133296,
+      "grad_norm": 0.05498271808028221,
+      "learning_rate": 0.0020666666666666667,
+      "loss": 10.1042,
+      "step": 93
+    },
+    {
+      "epoch": 0.2088308803110247,
+      "grad_norm": 0.05288155749440193,
+      "learning_rate": 0.002088888888888889,
+      "loss": 9.6725,
+      "step": 94
+    },
+    {
+      "epoch": 0.21105248542071647,
+      "grad_norm": 0.04739094898104668,
+      "learning_rate": 0.0021111111111111113,
+      "loss": 9.8394,
+      "step": 95
+    },
+    {
+      "epoch": 0.21327409053040822,
+      "grad_norm": 0.05595097690820694,
+      "learning_rate": 0.0021333333333333334,
+      "loss": 9.8368,
+      "step": 96
+    },
+    {
+      "epoch": 0.21549569564009996,
+      "grad_norm": 0.052203740924596786,
+      "learning_rate": 0.0021555555555555555,
+      "loss": 9.5522,
+      "step": 97
+    },
+    {
+      "epoch": 0.21771730074979173,
+      "grad_norm": 0.04803293198347092,
+      "learning_rate": 0.002177777777777778,
+      "loss": 9.7696,
+      "step": 98
+    },
+    {
+      "epoch": 0.21993890585948347,
+      "grad_norm": 0.054248202592134476,
+      "learning_rate": 0.0021999999999999997,
+      "loss": 9.7664,
+      "step": 99
+    },
+    {
+      "epoch": 0.22216051096917522,
+      "grad_norm": 0.04804198816418648,
+      "learning_rate": 0.0022222222222222222,
+      "loss": 9.6602,
+      "step": 100
+    },
+    {
+      "epoch": 0.224382116078867,
+      "grad_norm": 0.0431906133890152,
+      "learning_rate": 0.0022444444444444443,
+      "loss": 9.4522,
+      "step": 101
+    },
+    {
+      "epoch": 0.22660372118855873,
+      "grad_norm": 0.05638457089662552,
+      "learning_rate": 0.002266666666666667,
+      "loss": 9.5481,
+      "step": 102
+    },
+    {
+      "epoch": 0.22882532629825048,
+      "grad_norm": 0.053331535309553146,
+      "learning_rate": 0.002288888888888889,
+      "loss": 9.6493,
+      "step": 103
+    },
+    {
+      "epoch": 0.23104693140794225,
+      "grad_norm": 0.05476195365190506,
+      "learning_rate": 0.002311111111111111,
+      "loss": 9.536,
+      "step": 104
+    },
+    {
+      "epoch": 0.233268536517634,
+      "grad_norm": 0.056775737553834915,
+      "learning_rate": 0.0023333333333333335,
+      "loss": 9.3876,
+      "step": 105
+    },
+    {
+      "epoch": 0.23549014162732573,
+      "grad_norm": 0.04439986124634743,
+      "learning_rate": 0.0023555555555555556,
+      "loss": 9.671,
+      "step": 106
+    },
+    {
+      "epoch": 0.2377117467370175,
+      "grad_norm": 0.04525638371706009,
+      "learning_rate": 0.002377777777777778,
+      "loss": 9.5923,
+      "step": 107
+    },
+    {
+      "epoch": 0.23993335184670925,
+      "grad_norm": 0.04767855256795883,
+      "learning_rate": 0.0024000000000000002,
+      "loss": 9.6033,
+      "step": 108
+    },
+    {
+      "epoch": 0.242154956956401,
+      "grad_norm": 0.047220680862665176,
+      "learning_rate": 0.0024222222222222223,
+      "loss": 9.5502,
+      "step": 109
+    },
+    {
+      "epoch": 0.24437656206609276,
+      "grad_norm": 0.04153040051460266,
+      "learning_rate": 0.0024444444444444444,
+      "loss": 9.5775,
+      "step": 110
+    },
+    {
+      "epoch": 0.2465981671757845,
+      "grad_norm": 0.050917111337184906,
+      "learning_rate": 0.0024666666666666665,
+      "loss": 9.5931,
+      "step": 111
+    },
+    {
+      "epoch": 0.24881977228547625,
+      "grad_norm": 0.0504818893969059,
+      "learning_rate": 0.002488888888888889,
+      "loss": 9.4467,
+      "step": 112
+    },
+    {
+      "epoch": 0.251041377395168,
+      "grad_norm": 0.0438624769449234,
+      "learning_rate": 0.002511111111111111,
+      "loss": 9.2971,
+      "step": 113
+    },
+    {
+      "epoch": 0.25326298250485973,
+      "grad_norm": 0.04707180708646774,
+      "learning_rate": 0.002533333333333333,
+      "loss": 9.3382,
+      "step": 114
+    },
+    {
+      "epoch": 0.2554845876145515,
+      "grad_norm": 0.049067236483097076,
+      "learning_rate": 0.0025555555555555557,
+      "loss": 9.4641,
+      "step": 115
+    },
+    {
+      "epoch": 0.2577061927242433,
+      "grad_norm": 0.046793680638074875,
+      "learning_rate": 0.002577777777777778,
+      "loss": 9.6351,
+      "step": 116
+    },
+    {
+      "epoch": 0.259927797833935,
+      "grad_norm": 0.048384182155132294,
+      "learning_rate": 0.0026000000000000003,
+      "loss": 9.3845,
+      "step": 117
+    },
+    {
+      "epoch": 0.26214940294362676,
+      "grad_norm": 0.051988836377859116,
+      "learning_rate": 0.0026222222222222224,
+      "loss": 9.5515,
+      "step": 118
+    },
+    {
+      "epoch": 0.26437100805331853,
+      "grad_norm": 0.04786220192909241,
+      "learning_rate": 0.0026444444444444445,
+      "loss": 9.3949,
+      "step": 119
+    },
+    {
+      "epoch": 0.26659261316301025,
+      "grad_norm": 0.05733378604054451,
+      "learning_rate": 0.0026666666666666666,
+      "loss": 9.0427,
+      "step": 120
+    },
+    {
+      "epoch": 0.268814218272702,
+      "grad_norm": 0.05151943117380142,
+      "learning_rate": 0.0026888888888888887,
+      "loss": 9.4388,
+      "step": 121
+    },
+    {
+      "epoch": 0.2710358233823938,
+      "grad_norm": 0.04847846180200577,
+      "learning_rate": 0.002711111111111111,
+      "loss": 9.3863,
+      "step": 122
+    },
+    {
+      "epoch": 0.2732574284920855,
+      "grad_norm": 0.052508387714624405,
+      "learning_rate": 0.0027333333333333333,
+      "loss": 9.2321,
+      "step": 123
+    },
+    {
+      "epoch": 0.2754790336017773,
+      "grad_norm": 0.04624883830547333,
+      "learning_rate": 0.002755555555555556,
+      "loss": 9.5099,
+      "step": 124
+    },
+    {
+      "epoch": 0.27770063871146905,
+      "grad_norm": 0.05091821402311325,
+      "learning_rate": 0.002777777777777778,
+      "loss": 9.1828,
+      "step": 125
+    },
+    {
+      "epoch": 0.27992224382116077,
+      "grad_norm": 0.05264339596033096,
+      "learning_rate": 0.0028,
+      "loss": 9.055,
+      "step": 126
+    },
+    {
+      "epoch": 0.28214384893085254,
+      "grad_norm": 0.05629086494445801,
+      "learning_rate": 0.0028222222222222225,
+      "loss": 8.9799,
+      "step": 127
+    },
+    {
+      "epoch": 0.2843654540405443,
+      "grad_norm": 0.051374807953834534,
+      "learning_rate": 0.0028444444444444446,
+      "loss": 9.0618,
+      "step": 128
+    },
+    {
+      "epoch": 0.286587059150236,
+      "grad_norm": 0.048084281384944916,
+      "learning_rate": 0.0028666666666666667,
+      "loss": 9.1234,
+      "step": 129
+    },
+    {
+      "epoch": 0.2888086642599278,
+      "grad_norm": 0.05345724895596504,
+      "learning_rate": 0.0028888888888888888,
+      "loss": 8.9961,
+      "step": 130
+    },
+    {
+      "epoch": 0.29103026936961957,
+      "grad_norm": 0.04753989726305008,
+      "learning_rate": 0.002911111111111111,
+      "loss": 9.2071,
+      "step": 131
+    },
+    {
+      "epoch": 0.2932518744793113,
+      "grad_norm": 0.042281683534383774,
+      "learning_rate": 0.0029333333333333334,
+      "loss": 9.0905,
+      "step": 132
+    },
+    {
+      "epoch": 0.29547347958900305,
+      "grad_norm": 0.04851217195391655,
+      "learning_rate": 0.0029555555555555555,
+      "loss": 8.7241,
+      "step": 133
+    },
+    {
+      "epoch": 0.2976950846986948,
+      "grad_norm": 0.04813491553068161,
+      "learning_rate": 0.002977777777777778,
+      "loss": 9.0812,
+      "step": 134
+    },
+    {
+      "epoch": 0.29991668980838654,
+      "grad_norm": 0.052904680371284485,
+      "learning_rate": 0.003,
+      "loss": 8.6711,
+      "step": 135
+    },
+    {
+      "epoch": 0.3021382949180783,
+      "grad_norm": 0.05275777354836464,
+      "learning_rate": 0.0029975308641975312,
+      "loss": 8.6604,
+      "step": 136
+    },
+    {
+      "epoch": 0.3043599000277701,
+      "grad_norm": 0.04787842929363251,
+      "learning_rate": 0.0029950617283950615,
+      "loss": 8.8263,
+      "step": 137
+    },
+    {
+      "epoch": 0.3065815051374618,
+      "grad_norm": 0.05127862095832825,
+      "learning_rate": 0.0029925925925925927,
+      "loss": 8.8265,
+      "step": 138
+    },
+    {
+      "epoch": 0.30880311024715357,
+      "grad_norm": 0.04686402529478073,
+      "learning_rate": 0.0029901234567901234,
+      "loss": 9.059,
+      "step": 139
+    },
+    {
+      "epoch": 0.31102471535684534,
+      "grad_norm": 0.04877879098057747,
+      "learning_rate": 0.0029876543209876546,
+      "loss": 8.9707,
+      "step": 140
+    },
+    {
+      "epoch": 0.31324632046653705,
+      "grad_norm": 0.0502287782728672,
+      "learning_rate": 0.002985185185185185,
+      "loss": 8.7495,
+      "step": 141
+    },
+    {
+      "epoch": 0.3154679255762288,
+      "grad_norm": 0.044707659631967545,
+      "learning_rate": 0.002982716049382716,
+      "loss": 8.9677,
+      "step": 142
+    },
+    {
+      "epoch": 0.3176895306859206,
+      "grad_norm": 0.052578676491975784,
+      "learning_rate": 0.0029802469135802472,
+      "loss": 8.6041,
+      "step": 143
+    },
+    {
+      "epoch": 0.3199111357956123,
+      "grad_norm": 0.04367254674434662,
+      "learning_rate": 0.002977777777777778,
+      "loss": 8.907,
+      "step": 144
+    },
+    {
+      "epoch": 0.3221327409053041,
+      "grad_norm": 0.048488833010196686,
+      "learning_rate": 0.0029753086419753087,
+      "loss": 8.5004,
+      "step": 145
+    },
+    {
+      "epoch": 0.32435434601499585,
+      "grad_norm": 0.04960470646619797,
+      "learning_rate": 0.0029728395061728394,
+      "loss": 8.5826,
+      "step": 146
+    },
+    {
+      "epoch": 0.32657595112468757,
+      "grad_norm": 0.05014924705028534,
+      "learning_rate": 0.0029703703703703706,
+      "loss": 8.6261,
+      "step": 147
+    },
+    {
+      "epoch": 0.32879755623437934,
+      "grad_norm": 0.04689573496580124,
+      "learning_rate": 0.0029679012345679013,
+      "loss": 8.7356,
+      "step": 148
+    },
+    {
+      "epoch": 0.3310191613440711,
+      "grad_norm": 0.051323119550943375,
+      "learning_rate": 0.002965432098765432,
+      "loss": 8.7201,
+      "step": 149
+    },
+    {
+      "epoch": 0.3332407664537628,
+      "grad_norm": 0.047252483665943146,
+      "learning_rate": 0.002962962962962963,
+      "loss": 8.8438,
+      "step": 150
+    },
+    {
+      "epoch": 0.3354623715634546,
+      "grad_norm": 0.04643348976969719,
+      "learning_rate": 0.002960493827160494,
+      "loss": 8.6791,
+      "step": 151
+    },
+    {
+      "epoch": 0.33768397667314637,
+      "grad_norm": 0.05457242578268051,
+      "learning_rate": 0.0029580246913580247,
+      "loss": 8.4789,
+      "step": 152
+    },
+    {
+      "epoch": 0.3399055817828381,
+      "grad_norm": 0.050342462956905365,
+      "learning_rate": 0.0029555555555555555,
+      "loss": 8.6765,
+      "step": 153
+    },
+    {
+      "epoch": 0.34212718689252986,
+      "grad_norm": 0.047568775713443756,
+      "learning_rate": 0.0029530864197530866,
+      "loss": 8.4614,
+      "step": 154
+    },
+    {
+      "epoch": 0.3443487920022216,
+      "grad_norm": 0.049684301018714905,
+      "learning_rate": 0.0029506172839506174,
+      "loss": 8.7243,
+      "step": 155
+    },
+    {
+      "epoch": 0.34657039711191334,
+      "grad_norm": 0.04956204071640968,
+      "learning_rate": 0.002948148148148148,
+      "loss": 8.3286,
+      "step": 156
+    },
+    {
+      "epoch": 0.3487920022216051,
+      "grad_norm": 0.05027199909090996,
+      "learning_rate": 0.002945679012345679,
+      "loss": 8.4557,
+      "step": 157
+    },
+    {
+      "epoch": 0.3510136073312969,
+      "grad_norm": 0.049550533294677734,
+      "learning_rate": 0.00294320987654321,
+      "loss": 8.5551,
+      "step": 158
+    },
+    {
+      "epoch": 0.3532352124409886,
+      "grad_norm": 0.0448816642165184,
+      "learning_rate": 0.0029407407407407407,
+      "loss": 8.5497,
+      "step": 159
+    },
+    {
+      "epoch": 0.35545681755068037,
+      "grad_norm": 0.0545976497232914,
+      "learning_rate": 0.0029382716049382715,
+      "loss": 8.3274,
+      "step": 160
+    },
+    {
+      "epoch": 0.35767842266037214,
+      "grad_norm": 0.04282578080892563,
+      "learning_rate": 0.0029358024691358026,
+      "loss": 8.2482,
+      "step": 161
+    },
+    {
+      "epoch": 0.35990002777006386,
+      "grad_norm": 0.04906844720244408,
+      "learning_rate": 0.0029333333333333334,
+      "loss": 8.4081,
+      "step": 162
+    },
+    {
+      "epoch": 0.36212163287975563,
+      "grad_norm": 0.05085057392716408,
+      "learning_rate": 0.002930864197530864,
+      "loss": 8.1858,
+      "step": 163
+    },
+    {
+      "epoch": 0.3643432379894474,
+      "grad_norm": 0.042417705059051514,
+      "learning_rate": 0.002928395061728395,
+      "loss": 8.3901,
+      "step": 164
+    },
+    {
+      "epoch": 0.3665648430991391,
+      "grad_norm": 0.041248828172683716,
+      "learning_rate": 0.002925925925925926,
+      "loss": 8.1118,
+      "step": 165
+    },
+    {
+      "epoch": 0.3687864482088309,
+      "grad_norm": 0.04611312597990036,
+      "learning_rate": 0.002923456790123457,
+      "loss": 8.3726,
+      "step": 166
+    },
+    {
+      "epoch": 0.37100805331852266,
+      "grad_norm": 0.04500536993145943,
+      "learning_rate": 0.0029209876543209875,
+      "loss": 8.3331,
+      "step": 167
+    },
+    {
+      "epoch": 0.3732296584282144,
+      "grad_norm": 0.049214959144592285,
+      "learning_rate": 0.0029185185185185186,
+      "loss": 8.1056,
+      "step": 168
+    },
+    {
+      "epoch": 0.37545126353790614,
+      "grad_norm": 0.042415715754032135,
+      "learning_rate": 0.0029160493827160494,
+      "loss": 8.3253,
+      "step": 169
+    },
+    {
+      "epoch": 0.3776728686475979,
+      "grad_norm": 0.04630446061491966,
+      "learning_rate": 0.0029135802469135805,
+      "loss": 8.3159,
+      "step": 170
+    },
+    {
+      "epoch": 0.37989447375728963,
+      "grad_norm": 0.04075557366013527,
+      "learning_rate": 0.002911111111111111,
+      "loss": 8.1875,
+      "step": 171
+    },
+    {
+      "epoch": 0.3821160788669814,
+      "grad_norm": 0.04644331336021423,
+      "learning_rate": 0.002908641975308642,
+      "loss": 8.4214,
+      "step": 172
+    },
+    {
+      "epoch": 0.3843376839766732,
+      "grad_norm": 0.04412577301263809,
+      "learning_rate": 0.002906172839506173,
+      "loss": 8.072,
+      "step": 173
+    },
+    {
+      "epoch": 0.3865592890863649,
+      "grad_norm": 0.04354352876543999,
+      "learning_rate": 0.002903703703703704,
+      "loss": 8.0577,
+      "step": 174
+    },
+    {
+      "epoch": 0.38878089419605666,
+      "grad_norm": 0.04611154645681381,
+      "learning_rate": 0.0029012345679012346,
+      "loss": 8.3131,
+      "step": 175
+    },
+    {
+      "epoch": 0.39100249930574843,
+      "grad_norm": 0.040660127997398376,
+      "learning_rate": 0.0028987654320987654,
+      "loss": 8.2307,
+      "step": 176
+    },
+    {
+      "epoch": 0.39322410441544015,
+      "grad_norm": 0.04646790400147438,
+      "learning_rate": 0.0028962962962962966,
+      "loss": 8.2717,
+      "step": 177
+    },
+    {
+      "epoch": 0.3954457095251319,
+      "grad_norm": 0.04254138097167015,
+      "learning_rate": 0.0028938271604938273,
+      "loss": 8.0337,
+      "step": 178
+    },
+    {
+      "epoch": 0.3976673146348237,
+      "grad_norm": 0.03916044905781746,
+      "learning_rate": 0.002891358024691358,
+      "loss": 7.8892,
+      "step": 179
+    },
+    {
+      "epoch": 0.3998889197445154,
+      "grad_norm": 0.0402020625770092,
+      "learning_rate": 0.0028888888888888888,
+      "loss": 8.059,
+      "step": 180
+    },
+    {
+      "epoch": 0.4021105248542072,
+      "grad_norm": 0.04283468797802925,
+      "learning_rate": 0.00288641975308642,
+      "loss": 8.1629,
+      "step": 181
+    },
+    {
+      "epoch": 0.4043321299638989,
+      "grad_norm": 0.03644024208188057,
+      "learning_rate": 0.0028839506172839507,
+      "loss": 7.9251,
+      "step": 182
+    },
+    {
+      "epoch": 0.40655373507359066,
+      "grad_norm": 0.03857982158660889,
+      "learning_rate": 0.0028814814814814814,
+      "loss": 7.957,
+      "step": 183
+    },
+    {
+      "epoch": 0.40877534018328243,
+      "grad_norm": 0.03826620802283287,
+      "learning_rate": 0.0028790123456790126,
+      "loss": 8.0753,
+      "step": 184
+    },
+    {
+      "epoch": 0.41099694529297415,
+      "grad_norm": 0.03651922941207886,
+      "learning_rate": 0.0028765432098765433,
+      "loss": 7.9539,
+      "step": 185
+    },
+    {
+      "epoch": 0.4132185504026659,
+      "grad_norm": 0.04460339993238449,
+      "learning_rate": 0.002874074074074074,
+      "loss": 7.6859,
+      "step": 186
+    },
+    {
+      "epoch": 0.4154401555123577,
+      "grad_norm": 0.03517116606235504,
+      "learning_rate": 0.0028716049382716048,
+      "loss": 7.9726,
+      "step": 187
+    },
+    {
+      "epoch": 0.4176617606220494,
+      "grad_norm": 0.036879803985357285,
+      "learning_rate": 0.002869135802469136,
+      "loss": 8.0059,
+      "step": 188
+    },
+    {
+      "epoch": 0.4198833657317412,
+      "grad_norm": 0.036895960569381714,
+      "learning_rate": 0.0028666666666666667,
+      "loss": 7.7974,
+      "step": 189
+    },
+    {
+      "epoch": 0.42210497084143295,
+      "grad_norm": 0.03512522578239441,
+      "learning_rate": 0.0028641975308641974,
+      "loss": 7.9832,
+      "step": 190
+    },
+    {
+      "epoch": 0.42432657595112466,
+      "grad_norm": 0.043111640959978104,
+      "learning_rate": 0.0028617283950617286,
+      "loss": 7.7488,
+      "step": 191
+    },
+    {
+      "epoch": 0.42654818106081643,
+      "grad_norm": 0.03965785354375839,
+      "learning_rate": 0.0028592592592592593,
+      "loss": 7.6362,
+      "step": 192
+    },
+    {
+      "epoch": 0.4287697861705082,
+      "grad_norm": 0.04848213866353035,
+      "learning_rate": 0.00285679012345679,
+      "loss": 7.5347,
+      "step": 193
+    },
+    {
+      "epoch": 0.4309913912801999,
+      "grad_norm": 0.0401473306119442,
+      "learning_rate": 0.0028543209876543208,
+      "loss": 7.6057,
+      "step": 194
+    },
+    {
+      "epoch": 0.4332129963898917,
+      "grad_norm": 0.03666435182094574,
+      "learning_rate": 0.002851851851851852,
+      "loss": 7.585,
+      "step": 195
+    },
+    {
+      "epoch": 0.43543460149958346,
+      "grad_norm": 0.033364903181791306,
+      "learning_rate": 0.002849382716049383,
+      "loss": 7.9137,
+      "step": 196
+    },
+    {
+      "epoch": 0.4376562066092752,
+      "grad_norm": 0.036309197545051575,
+      "learning_rate": 0.0028469135802469134,
+      "loss": 7.5144,
+      "step": 197
+    },
+    {
+      "epoch": 0.43987781171896695,
+      "grad_norm": 0.03403438627719879,
+      "learning_rate": 0.0028444444444444446,
+      "loss": 7.884,
+      "step": 198
+    },
+    {
+      "epoch": 0.4420994168286587,
+      "grad_norm": 0.033892132341861725,
+      "learning_rate": 0.0028419753086419753,
+      "loss": 7.9242,
+      "step": 199
+    },
+    {
+      "epoch": 0.44432102193835044,
+      "grad_norm": 0.034154847264289856,
+      "learning_rate": 0.0028395061728395065,
+      "loss": 7.817,
+      "step": 200
+    },
+    {
+      "epoch": 0.4465426270480422,
+      "grad_norm": 0.03998449072241783,
+      "learning_rate": 0.002837037037037037,
+      "loss": 7.677,
+      "step": 201
+    },
+    {
+      "epoch": 0.448764232157734,
+      "grad_norm": 0.038014624267816544,
+      "learning_rate": 0.002834567901234568,
+      "loss": 7.6986,
+      "step": 202
+    },
+    {
+      "epoch": 0.4509858372674257,
+      "grad_norm": 0.03981682285666466,
+      "learning_rate": 0.002832098765432099,
+      "loss": 7.5149,
+      "step": 203
+    },
+    {
+      "epoch": 0.45320744237711746,
+      "grad_norm": 0.03799785301089287,
+      "learning_rate": 0.00282962962962963,
+      "loss": 7.3572,
+      "step": 204
+    },
+    {
+      "epoch": 0.45542904748680924,
+      "grad_norm": 0.04427637159824371,
+      "learning_rate": 0.0028271604938271606,
+      "loss": 7.1368,
+      "step": 205
+    },
+    {
+      "epoch": 0.45765065259650095,
+      "grad_norm": 0.03394748643040657,
+      "learning_rate": 0.0028246913580246913,
+      "loss": 7.5763,
+      "step": 206
+    },
+    {
+      "epoch": 0.4598722577061927,
+      "grad_norm": 0.03583474084734917,
+      "learning_rate": 0.0028222222222222225,
+      "loss": 7.7005,
+      "step": 207
+    },
+    {
+      "epoch": 0.4620938628158845,
+      "grad_norm": 0.033851176500320435,
+      "learning_rate": 0.0028197530864197532,
+      "loss": 7.7524,
+      "step": 208
+    },
+    {
+      "epoch": 0.4643154679255762,
+      "grad_norm": 0.03364001214504242,
+      "learning_rate": 0.002817283950617284,
+      "loss": 7.6403,
+      "step": 209
+    },
+    {
+      "epoch": 0.466537073035268,
+      "grad_norm": 0.039782341569662094,
+      "learning_rate": 0.0028148148148148147,
+      "loss": 7.4501,
+      "step": 210
+    },
+    {
+      "epoch": 0.46875867814495975,
+      "grad_norm": 0.033960845321416855,
+      "learning_rate": 0.002812345679012346,
+      "loss": 7.5254,
+      "step": 211
+    },
+    {
+      "epoch": 0.47098028325465147,
+      "grad_norm": 0.035350967198610306,
+      "learning_rate": 0.0028098765432098766,
+      "loss": 7.4594,
+      "step": 212
+    },
+    {
+      "epoch": 0.47320188836434324,
+      "grad_norm": 0.030056284740567207,
+      "learning_rate": 0.0028074074074074073,
+      "loss": 7.5952,
+      "step": 213
+    },
+    {
+      "epoch": 0.475423493474035,
+      "grad_norm": 0.03404721990227699,
+      "learning_rate": 0.0028049382716049385,
+      "loss": 7.26,
+      "step": 214
+    },
+    {
+      "epoch": 0.4776450985837267,
+      "grad_norm": 0.03368282690644264,
+      "learning_rate": 0.0028024691358024692,
+      "loss": 7.5807,
+      "step": 215
+    },
+    {
+      "epoch": 0.4798667036934185,
+      "grad_norm": 0.03246147185564041,
+      "learning_rate": 0.0028,
+      "loss": 7.4564,
+      "step": 216
+    },
+    {
+      "epoch": 0.48208830880311027,
+      "grad_norm": 0.03420104831457138,
+      "learning_rate": 0.0027975308641975307,
+      "loss": 7.413,
+      "step": 217
+    },
+    {
+      "epoch": 0.484309913912802,
+      "grad_norm": 0.03454587236046791,
+      "learning_rate": 0.002795061728395062,
+      "loss": 7.5475,
+      "step": 218
+    },
+    {
+      "epoch": 0.48653151902249375,
+      "grad_norm": 0.030751582235097885,
+      "learning_rate": 0.0027925925925925926,
+      "loss": 7.4298,
+      "step": 219
+    },
+    {
+      "epoch": 0.4887531241321855,
+      "grad_norm": 0.03355126827955246,
+      "learning_rate": 0.0027901234567901233,
+      "loss": 7.633,
+      "step": 220
+    },
+    {
+      "epoch": 0.49097472924187724,
+      "grad_norm": 0.03314507007598877,
+      "learning_rate": 0.0027876543209876545,
+      "loss": 7.5276,
+      "step": 221
+    },
+    {
+      "epoch": 0.493196334351569,
+      "grad_norm": 0.03196662664413452,
+      "learning_rate": 0.0027851851851851852,
+      "loss": 7.8203,
+      "step": 222
+    },
+    {
+      "epoch": 0.4954179394612608,
+      "grad_norm": 0.03462420031428337,
+      "learning_rate": 0.002782716049382716,
+      "loss": 7.2141,
+      "step": 223
+    },
+    {
+      "epoch": 0.4976395445709525,
+      "grad_norm": 0.027754031121730804,
+      "learning_rate": 0.0027802469135802467,
+      "loss": 7.2748,
+      "step": 224
+    },
+    {
+      "epoch": 0.49986114968064427,
+      "grad_norm": 0.040720872581005096,
+      "learning_rate": 0.002777777777777778,
+      "loss": 7.1818,
+      "step": 225
+    },
+    {
+      "epoch": 0.502082754790336,
+      "grad_norm": 0.04294149950146675,
+      "learning_rate": 0.002775308641975309,
+      "loss": 7.3151,
+      "step": 226
+    },
+    {
+      "epoch": 0.5043043599000278,
+      "grad_norm": 0.03208593651652336,
+      "learning_rate": 0.0027728395061728394,
+      "loss": 7.3898,
+      "step": 227
+    },
+    {
+      "epoch": 0.5065259650097195,
+      "grad_norm": 0.03241240605711937,
+      "learning_rate": 0.0027703703703703705,
+      "loss": 7.5746,
+      "step": 228
+    },
+    {
+      "epoch": 0.5087475701194113,
+      "grad_norm": 0.026224076747894287,
+      "learning_rate": 0.0027679012345679013,
+      "loss": 7.5724,
+      "step": 229
+    },
+    {
+      "epoch": 0.510969175229103,
+      "grad_norm": 0.02820778824388981,
+      "learning_rate": 0.0027654320987654324,
+      "loss": 7.3347,
+      "step": 230
+    },
+    {
+      "epoch": 0.5131907803387947,
+      "grad_norm": 0.030725786462426186,
+      "learning_rate": 0.0027629629629629627,
+      "loss": 7.0236,
+      "step": 231
+    },
+    {
+      "epoch": 0.5154123854484866,
+      "grad_norm": 0.03701042756438255,
+      "learning_rate": 0.002760493827160494,
+      "loss": 7.4099,
+      "step": 232
+    },
+    {
+      "epoch": 0.5176339905581783,
+      "grad_norm": 0.028012625873088837,
+      "learning_rate": 0.0027580246913580246,
+      "loss": 7.3943,
+      "step": 233
+    },
+    {
+      "epoch": 0.51985559566787,
+      "grad_norm": 0.03428445756435394,
+      "learning_rate": 0.002755555555555556,
+      "loss": 7.4596,
+      "step": 234
+    },
+    {
+      "epoch": 0.5220772007775618,
+      "grad_norm": 0.03811949864029884,
+      "learning_rate": 0.0027530864197530865,
+      "loss": 7.3027,
+      "step": 235
+    },
+    {
+      "epoch": 0.5242988058872535,
+      "grad_norm": 0.031404219567775726,
+      "learning_rate": 0.0027506172839506173,
+      "loss": 7.5975,
+      "step": 236
+    },
+    {
+      "epoch": 0.5265204109969452,
+      "grad_norm": 0.029649903997778893,
+      "learning_rate": 0.0027481481481481484,
+      "loss": 7.3965,
+      "step": 237
+    },
+    {
+      "epoch": 0.5287420161066371,
+      "grad_norm": 0.037028517574071884,
+      "learning_rate": 0.002745679012345679,
+      "loss": 7.1045,
+      "step": 238
+    },
+    {
+      "epoch": 0.5309636212163288,
+      "grad_norm": 0.03681131452322006,
+      "learning_rate": 0.00274320987654321,
+      "loss": 7.0011,
+      "step": 239
+    },
+    {
+      "epoch": 0.5331852263260205,
+      "grad_norm": 0.02837192639708519,
+      "learning_rate": 0.0027407407407407406,
+      "loss": 7.3327,
+      "step": 240
+    },
+    {
+      "epoch": 0.5354068314357123,
+      "grad_norm": 0.03215103596448898,
+      "learning_rate": 0.002738271604938272,
+      "loss": 7.2435,
+      "step": 241
+    },
+    {
+      "epoch": 0.537628436545404,
+      "grad_norm": 0.027541397139430046,
+      "learning_rate": 0.0027358024691358025,
+      "loss": 7.257,
+      "step": 242
+    },
+    {
+      "epoch": 0.5398500416550958,
+      "grad_norm": 0.02753252163529396,
+      "learning_rate": 0.0027333333333333333,
+      "loss": 7.257,
+      "step": 243
+    },
+    {
+      "epoch": 0.5420716467647876,
+      "grad_norm": 0.03283260762691498,
+      "learning_rate": 0.0027308641975308644,
+      "loss": 7.0467,
+      "step": 244
+    },
+    {
+      "epoch": 0.5442932518744793,
+      "grad_norm": 0.030674073845148087,
+      "learning_rate": 0.002728395061728395,
+      "loss": 7.157,
+      "step": 245
+    },
+    {
+      "epoch": 0.546514856984171,
+      "grad_norm": 0.030835680663585663,
+      "learning_rate": 0.002725925925925926,
+      "loss": 7.2868,
+      "step": 246
+    },
+    {
+      "epoch": 0.5487364620938628,
+      "grad_norm": 0.03128199651837349,
+      "learning_rate": 0.0027234567901234566,
+      "loss": 7.3337,
+      "step": 247
+    },
+    {
+      "epoch": 0.5509580672035546,
+      "grad_norm": 0.033430133014917374,
+      "learning_rate": 0.002720987654320988,
+      "loss": 7.6031,
+      "step": 248
+    },
+    {
+      "epoch": 0.5531796723132463,
+      "grad_norm": 0.0288421381264925,
+      "learning_rate": 0.0027185185185185185,
+      "loss": 6.896,
+      "step": 249
+    },
+    {
+      "epoch": 0.5554012774229381,
+      "grad_norm": 0.029844773933291435,
+      "learning_rate": 0.0027160493827160493,
+      "loss": 6.7998,
+      "step": 250
+    },
+    {
+      "epoch": 0.5576228825326298,
+      "grad_norm": 0.02905687503516674,
+      "learning_rate": 0.0027135802469135805,
+      "loss": 6.8452,
+      "step": 251
+    },
+    {
+      "epoch": 0.5598444876423215,
+      "grad_norm": 0.028612490743398666,
+      "learning_rate": 0.002711111111111111,
+      "loss": 7.3031,
+      "step": 252
+    },
+    {
+      "epoch": 0.5620660927520134,
+      "grad_norm": 0.03148072585463524,
+      "learning_rate": 0.002708641975308642,
+      "loss": 6.9144,
+      "step": 253
+    },
+    {
+      "epoch": 0.5642876978617051,
+      "grad_norm": 0.02918795496225357,
+      "learning_rate": 0.0027061728395061727,
+      "loss": 7.1563,
+      "step": 254
+    },
+    {
+      "epoch": 0.5665093029713968,
+      "grad_norm": 0.025865094736218452,
+      "learning_rate": 0.002703703703703704,
+      "loss": 7.1238,
+      "step": 255
+    },
+    {
+      "epoch": 0.5687309080810886,
+      "grad_norm": 0.03287065774202347,
+      "learning_rate": 0.002701234567901235,
+      "loss": 7.3028,
+      "step": 256
+    },
+    {
+      "epoch": 0.5709525131907803,
+      "grad_norm": 0.027668677270412445,
+      "learning_rate": 0.0026987654320987653,
+      "loss": 7.5821,
+      "step": 257
+    },
+    {
+      "epoch": 0.573174118300472,
+      "grad_norm": 0.031455788761377335,
+      "learning_rate": 0.0026962962962962965,
+      "loss": 7.4093,
+      "step": 258
+    },
+    {
+      "epoch": 0.5753957234101639,
+      "grad_norm": 0.031509652733802795,
+      "learning_rate": 0.002693827160493827,
+      "loss": 7.4072,
+      "step": 259
+    },
+    {
+      "epoch": 0.5776173285198556,
+      "grad_norm": 0.025732247158885002,
+      "learning_rate": 0.0026913580246913584,
+      "loss": 7.1182,
+      "step": 260
+    },
+    {
+      "epoch": 0.5798389336295473,
+      "grad_norm": 0.030423138290643692,
+      "learning_rate": 0.0026888888888888887,
+      "loss": 7.0401,
+      "step": 261
+    },
+    {
+      "epoch": 0.5820605387392391,
+      "grad_norm": 0.03525110334157944,
+      "learning_rate": 0.00268641975308642,
+      "loss": 7.2664,
+      "step": 262
+    },
+    {
+      "epoch": 0.5842821438489308,
+      "grad_norm": 0.02785809151828289,
+      "learning_rate": 0.0026839506172839506,
+      "loss": 7.1201,
+      "step": 263
+    },
+    {
+      "epoch": 0.5865037489586226,
+      "grad_norm": 0.031698115170001984,
+      "learning_rate": 0.0026814814814814817,
+      "loss": 6.9975,
+      "step": 264
+    },
+    {
+      "epoch": 0.5887253540683144,
+      "grad_norm": 0.031081918627023697,
+      "learning_rate": 0.0026790123456790125,
+      "loss": 6.9444,
+      "step": 265
+    },
+    {
+      "epoch": 0.5909469591780061,
+      "grad_norm": 0.0324559286236763,
+      "learning_rate": 0.002676543209876543,
+      "loss": 7.1328,
+      "step": 266
+    },
+    {
+      "epoch": 0.5931685642876978,
+      "grad_norm": 0.03233259543776512,
+      "learning_rate": 0.0026740740740740744,
+      "loss": 6.9415,
+      "step": 267
+    },
+    {
+      "epoch": 0.5953901693973896,
+      "grad_norm": 0.04220357909798622,
+      "learning_rate": 0.0026716049382716047,
+      "loss": 6.8164,
+      "step": 268
+    },
+    {
+      "epoch": 0.5976117745070814,
+      "grad_norm": 0.034913282841444016,
+      "learning_rate": 0.002669135802469136,
+      "loss": 7.2838,
+      "step": 269
+    },
+    {
+      "epoch": 0.5998333796167731,
+      "grad_norm": 0.02642977610230446,
+      "learning_rate": 0.0026666666666666666,
+      "loss": 7.3645,
+      "step": 270
+    },
+    {
+      "epoch": 0.6020549847264649,
+      "grad_norm": 0.030156614258885384,
+      "learning_rate": 0.0026641975308641977,
+      "loss": 7.0999,
+      "step": 271
+    },
+    {
+      "epoch": 0.6042765898361566,
+      "grad_norm": 0.03368644416332245,
+      "learning_rate": 0.0026617283950617285,
+      "loss": 7.2405,
+      "step": 272
+    },
+    {
+      "epoch": 0.6064981949458483,
+      "grad_norm": 0.03504369780421257,
+      "learning_rate": 0.002659259259259259,
+      "loss": 7.0479,
+      "step": 273
+    },
+    {
+      "epoch": 0.6087198000555402,
+      "grad_norm": 0.029628051444888115,
+      "learning_rate": 0.0026567901234567904,
+      "loss": 6.7508,
+      "step": 274
+    },
+    {
+      "epoch": 0.6109414051652319,
+      "grad_norm": 0.029622158035635948,
+      "learning_rate": 0.002654320987654321,
+      "loss": 7.2015,
+      "step": 275
+    },
+    {
+      "epoch": 0.6131630102749236,
+      "grad_norm": 0.032323166728019714,
+      "learning_rate": 0.002651851851851852,
+      "loss": 6.9056,
+      "step": 276
+    },
+    {
+      "epoch": 0.6153846153846154,
+      "grad_norm": 0.026018457487225533,
+      "learning_rate": 0.0026493827160493826,
+      "loss": 7.2168,
+      "step": 277
+    },
+    {
+      "epoch": 0.6176062204943071,
+      "grad_norm": 0.030117856338620186,
+      "learning_rate": 0.0026469135802469138,
+      "loss": 7.0721,
+      "step": 278
+    },
+    {
+      "epoch": 0.6198278256039988,
+      "grad_norm": 0.030448278412222862,
+      "learning_rate": 0.0026444444444444445,
+      "loss": 6.9572,
+      "step": 279
+    },
+    {
+      "epoch": 0.6220494307136907,
+      "grad_norm": 0.028298810124397278,
+      "learning_rate": 0.0026419753086419752,
+      "loss": 7.2107,
+      "step": 280
+    },
+    {
+      "epoch": 0.6242710358233824,
+      "grad_norm": 0.027341045439243317,
+      "learning_rate": 0.0026395061728395064,
+      "loss": 7.1902,
+      "step": 281
+    },
+    {
+      "epoch": 0.6264926409330741,
+      "grad_norm": 0.03337961807847023,
+      "learning_rate": 0.002637037037037037,
+      "loss": 7.1904,
+      "step": 282
+    },
+    {
+      "epoch": 0.6287142460427659,
+      "grad_norm": 0.030440054833889008,
+      "learning_rate": 0.002634567901234568,
+      "loss": 7.4463,
+      "step": 283
+    },
+    {
+      "epoch": 0.6309358511524576,
+      "grad_norm": 0.03126044571399689,
+      "learning_rate": 0.0026320987654320986,
+      "loss": 7.4159,
+      "step": 284
+    },
+    {
+      "epoch": 0.6331574562621494,
+      "grad_norm": 0.031162705272436142,
+      "learning_rate": 0.0026296296296296298,
+      "loss": 7.1282,
+      "step": 285
+    },
+    {
+      "epoch": 0.6353790613718412,
+      "grad_norm": 0.026622280478477478,
+      "learning_rate": 0.0026271604938271605,
+      "loss": 7.1215,
+      "step": 286
+    },
+    {
+      "epoch": 0.6376006664815329,
+      "grad_norm": 0.03224776312708855,
+      "learning_rate": 0.0026246913580246912,
+      "loss": 7.0815,
+      "step": 287
+    },
+    {
+      "epoch": 0.6398222715912246,
+      "grad_norm": 0.030211608856916428,
+      "learning_rate": 0.0026222222222222224,
+      "loss": 7.1783,
+      "step": 288
+    },
+    {
+      "epoch": 0.6420438767009164,
+      "grad_norm": 0.036949314177036285,
+      "learning_rate": 0.002619753086419753,
+      "loss": 7.2292,
+      "step": 289
+    },
+    {
+      "epoch": 0.6442654818106082,
+      "grad_norm": 0.030840424820780754,
+      "learning_rate": 0.002617283950617284,
+      "loss": 7.2447,
+      "step": 290
+    },
+    {
+      "epoch": 0.6464870869202999,
+      "grad_norm": 0.03116264007985592,
+      "learning_rate": 0.0026148148148148146,
+      "loss": 6.9522,
+      "step": 291
+    },
+    {
+      "epoch": 0.6487086920299917,
+      "grad_norm": 0.03676169738173485,
+      "learning_rate": 0.0026123456790123458,
+      "loss": 6.8615,
+      "step": 292
+    },
+    {
+      "epoch": 0.6509302971396834,
+      "grad_norm": 0.03387770429253578,
+      "learning_rate": 0.0026098765432098765,
+      "loss": 7.0317,
+      "step": 293
+    },
+    {
+      "epoch": 0.6531519022493751,
+      "grad_norm": 0.03344965726137161,
+      "learning_rate": 0.0026074074074074072,
+      "loss": 6.9111,
+      "step": 294
+    },
+    {
+      "epoch": 0.655373507359067,
+      "grad_norm": 0.03378414735198021,
+      "learning_rate": 0.0026049382716049384,
+      "loss": 6.8334,
+      "step": 295
+    },
+    {
+      "epoch": 0.6575951124687587,
+      "grad_norm": 0.028000645339488983,
+      "learning_rate": 0.002602469135802469,
+      "loss": 7.2455,
+      "step": 296
+    },
+    {
+      "epoch": 0.6598167175784504,
+      "grad_norm": 0.029027551412582397,
+      "learning_rate": 0.0026000000000000003,
+      "loss": 6.7113,
+      "step": 297
+    },
+    {
+      "epoch": 0.6620383226881422,
+      "grad_norm": 0.030199337750673294,
+      "learning_rate": 0.0025975308641975306,
+      "loss": 6.9931,
+      "step": 298
+    },
+    {
+      "epoch": 0.6642599277978339,
+      "grad_norm": 0.027782822027802467,
+      "learning_rate": 0.002595061728395062,
+      "loss": 7.0203,
+      "step": 299
+    },
+    {
+      "epoch": 0.6664815329075257,
+      "grad_norm": 0.027012750506401062,
+      "learning_rate": 0.0025925925925925925,
+      "loss": 7.0655,
+      "step": 300
+    },
+    {
+      "epoch": 0.6687031380172175,
+      "grad_norm": 0.026889488101005554,
+      "learning_rate": 0.0025901234567901237,
+      "loss": 7.0643,
+      "step": 301
+    },
+    {
+      "epoch": 0.6709247431269092,
+      "grad_norm": 0.03029460459947586,
+      "learning_rate": 0.0025876543209876544,
+      "loss": 7.0144,
+      "step": 302
+    },
+    {
+      "epoch": 0.6731463482366009,
+      "grad_norm": 0.026146529242396355,
+      "learning_rate": 0.002585185185185185,
+      "loss": 7.2646,
+      "step": 303
+    },
+    {
+      "epoch": 0.6753679533462927,
+      "grad_norm": 0.03537617623806,
+      "learning_rate": 0.0025827160493827163,
+      "loss": 6.8476,
+      "step": 304
+    },
+    {
+      "epoch": 0.6775895584559845,
+      "grad_norm": 0.03081704117357731,
+      "learning_rate": 0.002580246913580247,
+      "loss": 7.1906,
+      "step": 305
+    },
+    {
+      "epoch": 0.6798111635656762,
+      "grad_norm": 0.033128634095191956,
+      "learning_rate": 0.002577777777777778,
+      "loss": 7.0696,
+      "step": 306
+    },
+    {
+      "epoch": 0.682032768675368,
+      "grad_norm": 0.03375165909528732,
+      "learning_rate": 0.0025753086419753085,
+      "loss": 6.6613,
+      "step": 307
+    },
+    {
+      "epoch": 0.6842543737850597,
+      "grad_norm": 0.029131779447197914,
+      "learning_rate": 0.0025728395061728397,
+      "loss": 7.2572,
+      "step": 308
+    },
+    {
+      "epoch": 0.6864759788947514,
+      "grad_norm": 0.026389723643660545,
+      "learning_rate": 0.0025703703703703704,
+      "loss": 7.0176,
+      "step": 309
+    },
+    {
+      "epoch": 0.6886975840044433,
+      "grad_norm": 0.03170786425471306,
+      "learning_rate": 0.002567901234567901,
+      "loss": 6.8322,
+      "step": 310
+    },
+    {
+      "epoch": 0.690919189114135,
+      "grad_norm": 0.02895709127187729,
+      "learning_rate": 0.0025654320987654323,
+      "loss": 7.0894,
+      "step": 311
+    },
+    {
+      "epoch": 0.6931407942238267,
+      "grad_norm": 0.025471147149801254,
+      "learning_rate": 0.002562962962962963,
+      "loss": 7.0324,
+      "step": 312
+    },
+    {
+      "epoch": 0.6953623993335185,
+      "grad_norm": 0.029405398294329643,
+      "learning_rate": 0.002560493827160494,
+      "loss": 6.8744,
+      "step": 313
+    },
+    {
+      "epoch": 0.6975840044432102,
+      "grad_norm": 0.02752537652850151,
+      "learning_rate": 0.0025580246913580245,
+      "loss": 6.8055,
+      "step": 314
+    },
+    {
+      "epoch": 0.6998056095529019,
+      "grad_norm": 0.024177823215723038,
+      "learning_rate": 0.0025555555555555557,
+      "loss": 7.3241,
+      "step": 315
+    },
+    {
+      "epoch": 0.7020272146625938,
+      "grad_norm": 0.02782405912876129,
+      "learning_rate": 0.0025530864197530864,
+      "loss": 7.0388,
+      "step": 316
+    },
+    {
+      "epoch": 0.7042488197722855,
+      "grad_norm": 0.026573065668344498,
+      "learning_rate": 0.002550617283950617,
+      "loss": 7.0159,
+      "step": 317
+    },
+    {
+      "epoch": 0.7064704248819772,
+      "grad_norm": 0.03358136862516403,
+      "learning_rate": 0.0025481481481481483,
+      "loss": 7.1215,
+      "step": 318
+    },
+    {
+      "epoch": 0.708692029991669,
+      "grad_norm": 0.03143709897994995,
+      "learning_rate": 0.002545679012345679,
+      "loss": 6.6386,
+      "step": 319
+    },
+    {
+      "epoch": 0.7109136351013607,
+      "grad_norm": 0.033959418535232544,
+      "learning_rate": 0.00254320987654321,
+      "loss": 6.8983,
+      "step": 320
+    },
+    {
+      "epoch": 0.7131352402110525,
+      "grad_norm": 0.024646824225783348,
+      "learning_rate": 0.0025407407407407405,
+      "loss": 6.8726,
+      "step": 321
+    },
+    {
+      "epoch": 0.7153568453207443,
+      "grad_norm": 0.02922394871711731,
+      "learning_rate": 0.0025382716049382717,
+      "loss": 6.9099,
+      "step": 322
+    },
+    {
+      "epoch": 0.717578450430436,
+      "grad_norm": 0.030439792200922966,
+      "learning_rate": 0.0025358024691358024,
+      "loss": 6.9323,
+      "step": 323
+    },
+    {
+      "epoch": 0.7198000555401277,
+      "grad_norm": 0.04333483427762985,
+      "learning_rate": 0.002533333333333333,
+      "loss": 6.9007,
+      "step": 324
+    },
+    {
+      "epoch": 0.7220216606498195,
+      "grad_norm": 0.03616077080368996,
+      "learning_rate": 0.0025308641975308644,
+      "loss": 6.8624,
+      "step": 325
+    },
+    {
+      "epoch": 0.7242432657595113,
+      "grad_norm": 0.033735282719135284,
+      "learning_rate": 0.002528395061728395,
+      "loss": 6.8608,
+      "step": 326
+    },
+    {
+      "epoch": 0.726464870869203,
+      "grad_norm": 0.03077850490808487,
+      "learning_rate": 0.0025259259259259263,
+      "loss": 7.1149,
+      "step": 327
+    },
+    {
+      "epoch": 0.7286864759788948,
+      "grad_norm": 0.03490167856216431,
+      "learning_rate": 0.0025234567901234566,
+      "loss": 6.9609,
+      "step": 328
+    },
+    {
+      "epoch": 0.7309080810885865,
+      "grad_norm": 0.029763774946331978,
+      "learning_rate": 0.0025209876543209877,
+      "loss": 6.9148,
+      "step": 329
+    },
+    {
+      "epoch": 0.7331296861982782,
+      "grad_norm": 0.029299907386302948,
+      "learning_rate": 0.0025185185185185185,
+      "loss": 6.9816,
+      "step": 330
+    },
+    {
+      "epoch": 0.7353512913079701,
+      "grad_norm": 0.02755190059542656,
+      "learning_rate": 0.0025160493827160496,
+      "loss": 6.724,
+      "step": 331
+    },
+    {
+      "epoch": 0.7375728964176618,
+      "grad_norm": 0.02832108922302723,
+      "learning_rate": 0.0025135802469135804,
+      "loss": 6.8974,
+      "step": 332
+    },
+    {
+      "epoch": 0.7397945015273535,
+      "grad_norm": 0.02688843198120594,
+      "learning_rate": 0.002511111111111111,
+      "loss": 7.1274,
+      "step": 333
+    },
+    {
+      "epoch": 0.7420161066370453,
+      "grad_norm": 0.030441652983427048,
+      "learning_rate": 0.0025086419753086423,
+      "loss": 6.9542,
+      "step": 334
+    },
+    {
+      "epoch": 0.744237711746737,
+      "grad_norm": 0.0377890020608902,
+      "learning_rate": 0.002506172839506173,
+      "loss": 7.0826,
+      "step": 335
+    },
+    {
+      "epoch": 0.7464593168564287,
+      "grad_norm": 0.03879869356751442,
+      "learning_rate": 0.0025037037037037037,
+      "loss": 7.051,
+      "step": 336
+    },
+    {
+      "epoch": 0.7486809219661206,
+      "grad_norm": 0.022517995908856392,
+      "learning_rate": 0.0025012345679012345,
+      "loss": 7.0527,
+      "step": 337
+    },
+    {
+      "epoch": 0.7509025270758123,
+      "grad_norm": 0.02492484450340271,
+      "learning_rate": 0.0024987654320987656,
+      "loss": 6.867,
+      "step": 338
+    },
+    {
+      "epoch": 0.753124132185504,
+      "grad_norm": 0.0263162013143301,
+      "learning_rate": 0.0024962962962962964,
+      "loss": 6.9205,
+      "step": 339
+    },
+    {
+      "epoch": 0.7553457372951958,
+      "grad_norm": 0.03145473077893257,
+      "learning_rate": 0.002493827160493827,
+      "loss": 7.1822,
+      "step": 340
+    },
+    {
+      "epoch": 0.7575673424048875,
+      "grad_norm": 0.02679683268070221,
+      "learning_rate": 0.0024913580246913583,
+      "loss": 7.0945,
+      "step": 341
+    },
+    {
+      "epoch": 0.7597889475145793,
+      "grad_norm": 0.03611517325043678,
+      "learning_rate": 0.002488888888888889,
+      "loss": 6.5617,
+      "step": 342
+    },
+    {
+      "epoch": 0.7620105526242711,
+      "grad_norm": 0.028081588447093964,
+      "learning_rate": 0.0024864197530864197,
+      "loss": 7.0982,
+      "step": 343
+    },
+    {
+      "epoch": 0.7642321577339628,
+      "grad_norm": 0.03137503191828728,
+      "learning_rate": 0.0024839506172839505,
+      "loss": 6.6402,
+      "step": 344
+    },
+    {
+      "epoch": 0.7664537628436545,
+      "grad_norm": 0.03392868861556053,
+      "learning_rate": 0.0024814814814814816,
+      "loss": 7.1155,
+      "step": 345
+    },
+    {
+      "epoch": 0.7686753679533463,
+      "grad_norm": 0.030533727258443832,
+      "learning_rate": 0.0024790123456790124,
+      "loss": 7.0098,
+      "step": 346
+    },
+    {
+      "epoch": 0.7708969730630381,
+      "grad_norm": 0.03094147890806198,
+      "learning_rate": 0.002476543209876543,
+      "loss": 6.7761,
+      "step": 347
+    },
+    {
+      "epoch": 0.7731185781727298,
+      "grad_norm": 0.025714484974741936,
+      "learning_rate": 0.0024740740740740743,
+      "loss": 6.8854,
+      "step": 348
+    },
+    {
+      "epoch": 0.7753401832824216,
+      "grad_norm": 0.031463462859392166,
+      "learning_rate": 0.002471604938271605,
+      "loss": 6.8518,
+      "step": 349
+    },
+    {
+      "epoch": 0.7775617883921133,
+      "grad_norm": 0.025870738551020622,
+      "learning_rate": 0.0024691358024691358,
+      "loss": 6.9806,
+      "step": 350
+    },
+    {
+      "epoch": 0.779783393501805,
+      "grad_norm": 0.029403753578662872,
+      "learning_rate": 0.0024666666666666665,
+      "loss": 6.801,
+      "step": 351
+    },
+    {
+      "epoch": 0.7820049986114969,
+      "grad_norm": 0.029973311349749565,
+      "learning_rate": 0.0024641975308641977,
+      "loss": 7.143,
+      "step": 352
+    },
+    {
+      "epoch": 0.7842266037211886,
+      "grad_norm": 0.02538173645734787,
+      "learning_rate": 0.0024617283950617284,
+      "loss": 6.7702,
+      "step": 353
+    },
+    {
+      "epoch": 0.7864482088308803,
+      "grad_norm": 0.02790713682770729,
+      "learning_rate": 0.002459259259259259,
+      "loss": 6.6072,
+      "step": 354
+    },
+    {
+      "epoch": 0.7886698139405721,
+      "grad_norm": 0.028936631977558136,
+      "learning_rate": 0.0024567901234567903,
+      "loss": 6.8102,
+      "step": 355
+    },
+    {
+      "epoch": 0.7908914190502638,
+      "grad_norm": 0.029574889689683914,
+      "learning_rate": 0.002454320987654321,
+      "loss": 7.088,
+      "step": 356
+    },
+    {
+      "epoch": 0.7931130241599555,
+      "grad_norm": 0.03510890156030655,
+      "learning_rate": 0.002451851851851852,
+      "loss": 6.6048,
+      "step": 357
+    },
+    {
+      "epoch": 0.7953346292696474,
+      "grad_norm": 0.03128746896982193,
+      "learning_rate": 0.0024493827160493825,
+      "loss": 6.9001,
+      "step": 358
+    },
+    {
+      "epoch": 0.7975562343793391,
+      "grad_norm": 0.030788661912083626,
+      "learning_rate": 0.0024469135802469137,
+      "loss": 6.8839,
+      "step": 359
+    },
+    {
+      "epoch": 0.7997778394890308,
+      "grad_norm": 0.028769545257091522,
+      "learning_rate": 0.0024444444444444444,
+      "loss": 6.9937,
+      "step": 360
+    },
+    {
+      "epoch": 0.8019994445987225,
+      "grad_norm": 0.028774607926607132,
+      "learning_rate": 0.0024419753086419756,
+      "loss": 6.877,
+      "step": 361
+    },
+    {
+      "epoch": 0.8042210497084143,
+      "grad_norm": 0.03091941773891449,
+      "learning_rate": 0.0024395061728395063,
+      "loss": 6.8068,
+      "step": 362
+    },
+    {
+      "epoch": 0.8064426548181061,
+      "grad_norm": 0.03130682557821274,
+      "learning_rate": 0.002437037037037037,
+      "loss": 6.851,
+      "step": 363
+    },
+    {
+      "epoch": 0.8086642599277978,
+      "grad_norm": 0.026035597547888756,
+      "learning_rate": 0.002434567901234568,
+      "loss": 6.8659,
+      "step": 364
+    },
+    {
+      "epoch": 0.8108858650374896,
+      "grad_norm": 0.027542999014258385,
+      "learning_rate": 0.002432098765432099,
+      "loss": 7.1436,
+      "step": 365
+    },
+    {
+      "epoch": 0.8131074701471813,
+      "grad_norm": 0.030278557911515236,
+      "learning_rate": 0.0024296296296296297,
+      "loss": 7.2488,
+      "step": 366
+    },
+    {
+      "epoch": 0.815329075256873,
+      "grad_norm": 0.031210409477353096,
+      "learning_rate": 0.0024271604938271604,
+      "loss": 6.7226,
+      "step": 367
+    },
+    {
+      "epoch": 0.8175506803665649,
+      "grad_norm": 0.03228698670864105,
+      "learning_rate": 0.0024246913580246916,
+      "loss": 7.0787,
+      "step": 368
+    },
+    {
+      "epoch": 0.8197722854762566,
+      "grad_norm": 0.034072816371917725,
+      "learning_rate": 0.0024222222222222223,
+      "loss": 6.7884,
+      "step": 369
+    },
+    {
+      "epoch": 0.8219938905859483,
+      "grad_norm": 0.02897859923541546,
+      "learning_rate": 0.002419753086419753,
+      "loss": 6.685,
+      "step": 370
+    },
+    {
+      "epoch": 0.8242154956956401,
+      "grad_norm": 0.03470846638083458,
+      "learning_rate": 0.002417283950617284,
+      "loss": 6.8452,
+      "step": 371
+    },
+    {
+      "epoch": 0.8264371008053318,
+      "grad_norm": 0.039191946387290955,
+      "learning_rate": 0.002414814814814815,
+      "loss": 6.8654,
+      "step": 372
+    },
+    {
+      "epoch": 0.8286587059150236,
+      "grad_norm": 0.027019087225198746,
+      "learning_rate": 0.0024123456790123457,
+      "loss": 6.7975,
+      "step": 373
+    },
+    {
+      "epoch": 0.8308803110247154,
+      "grad_norm": 0.028417536988854408,
+      "learning_rate": 0.0024098765432098764,
+      "loss": 7.179,
+      "step": 374
+    },
+    {
+      "epoch": 0.8331019161344071,
+      "grad_norm": 0.031571824103593826,
+      "learning_rate": 0.0024074074074074076,
+      "loss": 6.6972,
+      "step": 375
+    },
+    {
+      "epoch": 0.8353235212440988,
+      "grad_norm": 0.025722427293658257,
+      "learning_rate": 0.0024049382716049383,
+      "loss": 6.7311,
+      "step": 376
+    },
+    {
+      "epoch": 0.8375451263537906,
+      "grad_norm": 0.03283727914094925,
+      "learning_rate": 0.002402469135802469,
+      "loss": 6.7464,
+      "step": 377
+    },
+    {
+      "epoch": 0.8397667314634824,
+      "grad_norm": 0.025677897036075592,
+      "learning_rate": 0.0024000000000000002,
+      "loss": 6.8963,
+      "step": 378
+    },
+    {
+      "epoch": 0.8419883365731741,
+      "grad_norm": 0.047866322100162506,
+      "learning_rate": 0.002397530864197531,
+      "loss": 6.7552,
+      "step": 379
+    },
+    {
+      "epoch": 0.8442099416828659,
+      "grad_norm": 0.025421174243092537,
+      "learning_rate": 0.0023950617283950617,
+      "loss": 6.8062,
+      "step": 380
+    },
+    {
+      "epoch": 0.8464315467925576,
+      "grad_norm": 0.03164597600698471,
+      "learning_rate": 0.0023925925925925924,
+      "loss": 6.7044,
+      "step": 381
+    },
+    {
+      "epoch": 0.8486531519022493,
+      "grad_norm": 0.03499810770153999,
+      "learning_rate": 0.0023901234567901236,
+      "loss": 6.8009,
+      "step": 382
+    },
+    {
+      "epoch": 0.8508747570119412,
+      "grad_norm": 0.03318408131599426,
+      "learning_rate": 0.0023876543209876543,
+      "loss": 6.861,
+      "step": 383
+    },
+    {
+      "epoch": 0.8530963621216329,
+      "grad_norm": 0.030897561460733414,
+      "learning_rate": 0.002385185185185185,
+      "loss": 6.7319,
+      "step": 384
+    },
+    {
+      "epoch": 0.8553179672313246,
+      "grad_norm": 0.035121310502290726,
+      "learning_rate": 0.0023827160493827162,
+      "loss": 6.9468,
+      "step": 385
+    },
+    {
+      "epoch": 0.8575395723410164,
+      "grad_norm": 0.02977081760764122,
+      "learning_rate": 0.002380246913580247,
+      "loss": 6.6821,
+      "step": 386
+    },
+    {
+      "epoch": 0.8597611774507081,
+      "grad_norm": 0.027311140671372414,
+      "learning_rate": 0.002377777777777778,
+      "loss": 6.8621,
+      "step": 387
+    },
+    {
+      "epoch": 0.8619827825603998,
+      "grad_norm": 0.027461953461170197,
+      "learning_rate": 0.0023753086419753084,
+      "loss": 7.0028,
+      "step": 388
+    },
+    {
+      "epoch": 0.8642043876700917,
+      "grad_norm": 0.02862822264432907,
+      "learning_rate": 0.0023728395061728396,
+      "loss": 6.726,
+      "step": 389
+    },
+    {
+      "epoch": 0.8664259927797834,
+      "grad_norm": 0.02961575612425804,
+      "learning_rate": 0.0023703703703703703,
+      "loss": 6.6011,
+      "step": 390
+    },
+    {
+      "epoch": 0.8686475978894751,
+      "grad_norm": 0.02995389699935913,
+      "learning_rate": 0.0023679012345679015,
+      "loss": 6.5965,
+      "step": 391
+    },
+    {
+      "epoch": 0.8708692029991669,
+      "grad_norm": 0.024569949135184288,
+      "learning_rate": 0.0023654320987654322,
+      "loss": 6.6053,
+      "step": 392
+    },
+    {
+      "epoch": 0.8730908081088586,
+      "grad_norm": 0.04112711548805237,
+      "learning_rate": 0.002362962962962963,
+      "loss": 6.7283,
+      "step": 393
+    },
+    {
+      "epoch": 0.8753124132185504,
+      "grad_norm": 0.03940947726368904,
+      "learning_rate": 0.002360493827160494,
+      "loss": 7.1214,
+      "step": 394
+    },
+    {
+      "epoch": 0.8775340183282422,
+      "grad_norm": 0.03772582858800888,
+      "learning_rate": 0.002358024691358025,
+      "loss": 6.7342,
+      "step": 395
+    },
+    {
+      "epoch": 0.8797556234379339,
+      "grad_norm": 0.026038886979222298,
+      "learning_rate": 0.0023555555555555556,
+      "loss": 6.7722,
+      "step": 396
+    },
+    {
+      "epoch": 0.8819772285476256,
+      "grad_norm": 0.02900727652013302,
+      "learning_rate": 0.0023530864197530863,
+      "loss": 6.6497,
+      "step": 397
+    },
+    {
+      "epoch": 0.8841988336573174,
+      "grad_norm": 0.03674020990729332,
+      "learning_rate": 0.0023506172839506175,
+      "loss": 6.4395,
+      "step": 398
+    },
+    {
+      "epoch": 0.8864204387670092,
+      "grad_norm": 0.030998943373560905,
+      "learning_rate": 0.002348148148148148,
+      "loss": 6.6632,
+      "step": 399
+    },
+    {
+      "epoch": 0.8886420438767009,
+      "grad_norm": 0.03351856395602226,
+      "learning_rate": 0.002345679012345679,
+      "loss": 6.6518,
+      "step": 400
+    },
+    {
+      "epoch": 0.8908636489863927,
+      "grad_norm": 0.030786389485001564,
+      "learning_rate": 0.00234320987654321,
+      "loss": 6.7591,
+      "step": 401
+    },
+    {
+      "epoch": 0.8930852540960844,
+      "grad_norm": 0.03690814599394798,
+      "learning_rate": 0.002340740740740741,
+      "loss": 6.5874,
+      "step": 402
+    },
+    {
+      "epoch": 0.8953068592057761,
+      "grad_norm": 0.030751686543226242,
+      "learning_rate": 0.0023382716049382716,
+      "loss": 6.6858,
+      "step": 403
+    },
+    {
+      "epoch": 0.897528464315468,
+      "grad_norm": 0.03505620360374451,
+      "learning_rate": 0.0023358024691358024,
+      "loss": 6.6377,
+      "step": 404
+    },
+    {
+      "epoch": 0.8997500694251597,
+      "grad_norm": 0.034183319658041,
+      "learning_rate": 0.0023333333333333335,
+      "loss": 6.6526,
+      "step": 405
+    },
+    {
+      "epoch": 0.9019716745348514,
+      "grad_norm": 0.03079032339155674,
+      "learning_rate": 0.0023308641975308643,
+      "loss": 6.7385,
+      "step": 406
+    },
+    {
+      "epoch": 0.9041932796445432,
+      "grad_norm": 0.03357497230172157,
+      "learning_rate": 0.002328395061728395,
+      "loss": 6.9186,
+      "step": 407
+    },
+    {
+      "epoch": 0.9064148847542349,
+      "grad_norm": 0.030285147950053215,
+      "learning_rate": 0.002325925925925926,
+      "loss": 6.8228,
+      "step": 408
+    },
+    {
+      "epoch": 0.9086364898639266,
+      "grad_norm": 0.03262084349989891,
+      "learning_rate": 0.002323456790123457,
+      "loss": 6.6384,
+      "step": 409
+    },
+    {
+      "epoch": 0.9108580949736185,
+      "grad_norm": 0.03636816889047623,
+      "learning_rate": 0.0023209876543209876,
+      "loss": 7.0531,
+      "step": 410
+    },
+    {
+      "epoch": 0.9130797000833102,
+      "grad_norm": 0.04130619764328003,
+      "learning_rate": 0.0023185185185185184,
+      "loss": 6.5445,
+      "step": 411
+    },
+    {
+      "epoch": 0.9153013051930019,
+      "grad_norm": 0.03653937578201294,
+      "learning_rate": 0.0023160493827160495,
+      "loss": 6.7057,
+      "step": 412
+    },
+    {
+      "epoch": 0.9175229103026937,
+      "grad_norm": 0.032455723732709885,
+      "learning_rate": 0.0023135802469135803,
+      "loss": 6.5941,
+      "step": 413
+    },
+    {
+      "epoch": 0.9197445154123854,
+      "grad_norm": 0.03412288427352905,
+      "learning_rate": 0.002311111111111111,
+      "loss": 6.9319,
+      "step": 414
+    },
+    {
+      "epoch": 0.9219661205220772,
+      "grad_norm": 0.035218141973018646,
+      "learning_rate": 0.002308641975308642,
+      "loss": 6.3276,
+      "step": 415
+    },
+    {
+      "epoch": 0.924187725631769,
+      "grad_norm": 0.03169059380888939,
+      "learning_rate": 0.002306172839506173,
+      "loss": 6.7758,
+      "step": 416
+    },
+    {
+      "epoch": 0.9264093307414607,
+      "grad_norm": 0.024801066145300865,
+      "learning_rate": 0.0023037037037037036,
+      "loss": 6.7239,
+      "step": 417
+    },
+    {
+      "epoch": 0.9286309358511524,
+      "grad_norm": 0.02555897831916809,
+      "learning_rate": 0.0023012345679012344,
+      "loss": 6.6137,
+      "step": 418
+    },
+    {
+      "epoch": 0.9308525409608442,
+      "grad_norm": 0.031776994466781616,
+      "learning_rate": 0.0022987654320987655,
+      "loss": 6.7518,
+      "step": 419
+    },
+    {
+      "epoch": 0.933074146070536,
+      "grad_norm": 0.03202354535460472,
+      "learning_rate": 0.0022962962962962963,
+      "loss": 6.9085,
+      "step": 420
+    },
+    {
+      "epoch": 0.9352957511802277,
+      "grad_norm": 0.029564661905169487,
+      "learning_rate": 0.002293827160493827,
+      "loss": 6.7439,
+      "step": 421
+    },
+    {
+      "epoch": 0.9375173562899195,
+      "grad_norm": 0.02673780918121338,
+      "learning_rate": 0.002291358024691358,
+      "loss": 6.8067,
+      "step": 422
+    },
+    {
+      "epoch": 0.9397389613996112,
+      "grad_norm": 0.031321026384830475,
+      "learning_rate": 0.002288888888888889,
+      "loss": 6.5093,
+      "step": 423
+    },
+    {
+      "epoch": 0.9419605665093029,
+      "grad_norm": 0.027808304876089096,
+      "learning_rate": 0.00228641975308642,
+      "loss": 6.4626,
+      "step": 424
+    },
+    {
+      "epoch": 0.9441821716189948,
+      "grad_norm": 0.025776829570531845,
+      "learning_rate": 0.0022839506172839504,
+      "loss": 6.7953,
+      "step": 425
+    },
+    {
+      "epoch": 0.9464037767286865,
+      "grad_norm": 0.03336632251739502,
+      "learning_rate": 0.0022814814814814816,
+      "loss": 6.5457,
+      "step": 426
+    },
+    {
+      "epoch": 0.9486253818383782,
+      "grad_norm": 0.028170697391033173,
+      "learning_rate": 0.0022790123456790123,
+      "loss": 6.785,
+      "step": 427
+    },
+    {
+      "epoch": 0.95084698694807,
+      "grad_norm": 0.03778989240527153,
+      "learning_rate": 0.0022765432098765435,
+      "loss": 6.6185,
+      "step": 428
+    },
+    {
+      "epoch": 0.9530685920577617,
+      "grad_norm": 0.03384809568524361,
+      "learning_rate": 0.0022740740740740738,
+      "loss": 6.6729,
+      "step": 429
+    },
+    {
+      "epoch": 0.9552901971674534,
+      "grad_norm": 0.03491503372788429,
+      "learning_rate": 0.002271604938271605,
+      "loss": 6.506,
+      "step": 430
+    },
+    {
+      "epoch": 0.9575118022771453,
+      "grad_norm": 0.028737282380461693,
+      "learning_rate": 0.002269135802469136,
+      "loss": 6.5785,
+      "step": 431
+    },
+    {
+      "epoch": 0.959733407386837,
+      "grad_norm": 0.029536008834838867,
+      "learning_rate": 0.002266666666666667,
+      "loss": 6.5539,
+      "step": 432
+    },
+    {
+      "epoch": 0.9619550124965287,
+      "grad_norm": 0.029247025027871132,
+      "learning_rate": 0.0022641975308641976,
+      "loss": 6.4651,
+      "step": 433
+    },
+    {
+      "epoch": 0.9641766176062205,
+      "grad_norm": 0.037673089653253555,
+      "learning_rate": 0.0022617283950617283,
+      "loss": 6.3081,
+      "step": 434
+    },
+    {
+      "epoch": 0.9663982227159122,
+      "grad_norm": 0.031512826681137085,
+      "learning_rate": 0.0022592592592592595,
+      "loss": 6.8234,
+      "step": 435
+    },
+    {
+      "epoch": 0.968619827825604,
+      "grad_norm": 0.03206060081720352,
+      "learning_rate": 0.00225679012345679,
+      "loss": 6.5044,
+      "step": 436
+    },
+    {
+      "epoch": 0.9708414329352958,
+      "grad_norm": 0.02888185903429985,
+      "learning_rate": 0.002254320987654321,
+      "loss": 6.301,
+      "step": 437
+    },
+    {
+      "epoch": 0.9730630380449875,
+      "grad_norm": 0.03852174058556557,
+      "learning_rate": 0.002251851851851852,
+      "loss": 6.614,
+      "step": 438
+    },
+    {
+      "epoch": 0.9752846431546792,
+      "grad_norm": 0.02977623976767063,
+      "learning_rate": 0.002249382716049383,
+      "loss": 6.8427,
+      "step": 439
+    },
+    {
+      "epoch": 0.977506248264371,
+      "grad_norm": 0.036224476993083954,
+      "learning_rate": 0.0022469135802469136,
+      "loss": 6.6819,
+      "step": 440
+    },
+    {
+      "epoch": 0.9797278533740628,
+      "grad_norm": 0.028337834402918816,
+      "learning_rate": 0.0022444444444444443,
+      "loss": 6.9447,
+      "step": 441
+    },
+    {
+      "epoch": 0.9819494584837545,
+      "grad_norm": 0.032608017325401306,
+      "learning_rate": 0.0022419753086419755,
+      "loss": 6.521,
+      "step": 442
+    },
+    {
+      "epoch": 0.9841710635934463,
+      "grad_norm": 0.030470674857497215,
+      "learning_rate": 0.002239506172839506,
+      "loss": 6.3275,
+      "step": 443
+    },
+    {
+      "epoch": 0.986392668703138,
+      "grad_norm": 0.045779451727867126,
+      "learning_rate": 0.002237037037037037,
+      "loss": 6.4221,
+      "step": 444
+    },
+    {
+      "epoch": 0.9886142738128297,
+      "grad_norm": 0.03529239073395729,
+      "learning_rate": 0.002234567901234568,
+      "loss": 6.1334,
+      "step": 445
+    },
+    {
+      "epoch": 0.9908358789225216,
+      "grad_norm": 0.04795719310641289,
+      "learning_rate": 0.002232098765432099,
+      "loss": 6.6265,
+      "step": 446
+    },
+    {
+      "epoch": 0.9930574840322133,
+      "grad_norm": 0.0322149284183979,
+      "learning_rate": 0.0022296296296296296,
+      "loss": 6.3979,
+      "step": 447
+    },
+    {
+      "epoch": 0.995279089141905,
+      "grad_norm": 0.030884549021720886,
+      "learning_rate": 0.0022271604938271603,
+      "loss": 6.5888,
+      "step": 448
+    },
+    {
+      "epoch": 0.9975006942515968,
+      "grad_norm": 0.03471051901578903,
+      "learning_rate": 0.0022246913580246915,
+      "loss": 6.4639,
+      "step": 449
+    },
+    {
+      "epoch": 0.9997222993612885,
+      "grad_norm": 0.029371246695518494,
+      "learning_rate": 0.0022222222222222222,
+      "loss": 6.6529,
+      "step": 450
+    },
+    {
+      "epoch": 1.0019439044709804,
+      "grad_norm": 0.02985074557363987,
+      "learning_rate": 0.002219753086419753,
+      "loss": 6.3826,
+      "step": 451
+    },
+    {
+      "epoch": 1.004165509580672,
+      "grad_norm": 0.03519700467586517,
+      "learning_rate": 0.002217283950617284,
+      "loss": 6.4751,
+      "step": 452
+    },
+    {
+      "epoch": 1.0063871146903638,
+      "grad_norm": 0.037781670689582825,
+      "learning_rate": 0.002214814814814815,
+      "loss": 6.4896,
+      "step": 453
+    },
+    {
+      "epoch": 1.0086087198000555,
+      "grad_norm": 0.027791369706392288,
+      "learning_rate": 0.002212345679012346,
+      "loss": 6.6676,
+      "step": 454
+    },
+    {
+      "epoch": 1.0108303249097472,
+      "grad_norm": 0.0314226932823658,
+      "learning_rate": 0.0022098765432098763,
+      "loss": 6.4966,
+      "step": 455
+    },
+    {
+      "epoch": 1.013051930019439,
+      "grad_norm": 0.03207942843437195,
+      "learning_rate": 0.0022074074074074075,
+      "loss": 6.7773,
+      "step": 456
+    },
+    {
+      "epoch": 1.0152735351291309,
+      "grad_norm": 0.03154020011425018,
+      "learning_rate": 0.0022049382716049382,
+      "loss": 6.5173,
+      "step": 457
+    },
+    {
+      "epoch": 1.0174951402388226,
+      "grad_norm": 0.029535001143813133,
+      "learning_rate": 0.0022024691358024694,
+      "loss": 6.6344,
+      "step": 458
+    },
+    {
+      "epoch": 1.0197167453485143,
+      "grad_norm": 0.0366988480091095,
+      "learning_rate": 0.0021999999999999997,
+      "loss": 5.9945,
+      "step": 459
+    },
+    {
+      "epoch": 1.021938350458206,
+      "grad_norm": 0.03053160198032856,
+      "learning_rate": 0.002197530864197531,
+      "loss": 6.4279,
+      "step": 460
+    },
+    {
+      "epoch": 1.0241599555678977,
+      "grad_norm": 0.02893592044711113,
+      "learning_rate": 0.002195061728395062,
+      "loss": 6.5733,
+      "step": 461
+    },
+    {
+      "epoch": 1.0263815606775895,
+      "grad_norm": 0.031111309304833412,
+      "learning_rate": 0.0021925925925925928,
+      "loss": 6.6382,
+      "step": 462
+    },
+    {
+      "epoch": 1.0286031657872814,
+      "grad_norm": 0.03582914173603058,
+      "learning_rate": 0.0021901234567901235,
+      "loss": 6.3495,
+      "step": 463
+    },
+    {
+      "epoch": 1.030824770896973,
+      "grad_norm": 0.029999269172549248,
+      "learning_rate": 0.0021876543209876542,
+      "loss": 6.8815,
+      "step": 464
+    },
+    {
+      "epoch": 1.0330463760066648,
+      "grad_norm": 0.03019753284752369,
+      "learning_rate": 0.0021851851851851854,
+      "loss": 6.7338,
+      "step": 465
+    },
+    {
+      "epoch": 1.0352679811163565,
+      "grad_norm": 0.03386778384447098,
+      "learning_rate": 0.002182716049382716,
+      "loss": 6.3456,
+      "step": 466
+    },
+    {
+      "epoch": 1.0374895862260483,
+      "grad_norm": 0.033411286771297455,
+      "learning_rate": 0.002180246913580247,
+      "loss": 6.6192,
+      "step": 467
+    },
+    {
+      "epoch": 1.03971119133574,
+      "grad_norm": 0.03734864294528961,
+      "learning_rate": 0.002177777777777778,
+      "loss": 6.5337,
+      "step": 468
+    },
+    {
+      "epoch": 1.041932796445432,
+      "grad_norm": 0.034468065947294235,
+      "learning_rate": 0.0021753086419753088,
+      "loss": 6.4662,
+      "step": 469
+    },
+    {
+      "epoch": 1.0441544015551236,
+      "grad_norm": 0.029232153668999672,
+      "learning_rate": 0.0021728395061728395,
+      "loss": 6.603,
+      "step": 470
+    },
+    {
+      "epoch": 1.0463760066648153,
+      "grad_norm": 0.035031069070100784,
+      "learning_rate": 0.0021703703703703702,
+      "loss": 6.4214,
+      "step": 471
+    },
+    {
+      "epoch": 1.048597611774507,
+      "grad_norm": 0.05113089084625244,
+      "learning_rate": 0.0021679012345679014,
+      "loss": 6.6573,
+      "step": 472
+    },
+    {
+      "epoch": 1.0508192168841988,
+      "grad_norm": 0.03398042917251587,
+      "learning_rate": 0.002165432098765432,
+      "loss": 6.4512,
+      "step": 473
+    },
+    {
+      "epoch": 1.0530408219938905,
+      "grad_norm": 0.032400310039520264,
+      "learning_rate": 0.002162962962962963,
+      "loss": 6.6219,
+      "step": 474
+    },
+    {
+      "epoch": 1.0552624271035824,
+      "grad_norm": 0.0310080386698246,
+      "learning_rate": 0.002160493827160494,
+      "loss": 6.7245,
+      "step": 475
+    },
+    {
+      "epoch": 1.0574840322132741,
+      "grad_norm": 0.02772977575659752,
+      "learning_rate": 0.002158024691358025,
+      "loss": 6.3142,
+      "step": 476
+    },
+    {
+      "epoch": 1.0597056373229659,
+      "grad_norm": 0.029645215719938278,
+      "learning_rate": 0.0021555555555555555,
+      "loss": 6.3477,
+      "step": 477
+    },
+    {
+      "epoch": 1.0619272424326576,
+      "grad_norm": 0.03764863312244415,
+      "learning_rate": 0.0021530864197530863,
+      "loss": 6.0773,
+      "step": 478
+    },
+    {
+      "epoch": 1.0641488475423493,
+      "grad_norm": 0.034320443868637085,
+      "learning_rate": 0.0021506172839506174,
+      "loss": 6.0489,
+      "step": 479
+    },
+    {
+      "epoch": 1.066370452652041,
+      "grad_norm": 0.03357730433344841,
+      "learning_rate": 0.002148148148148148,
+      "loss": 6.6008,
+      "step": 480
+    },
+    {
+      "epoch": 1.068592057761733,
+      "grad_norm": 0.030703729018568993,
+      "learning_rate": 0.002145679012345679,
+      "loss": 6.3549,
+      "step": 481
+    },
+    {
+      "epoch": 1.0708136628714247,
+      "grad_norm": 0.029614439234137535,
+      "learning_rate": 0.0021432098765432096,
+      "loss": 6.4135,
+      "step": 482
+    },
+    {
+      "epoch": 1.0730352679811164,
+      "grad_norm": 0.030537104234099388,
+      "learning_rate": 0.002140740740740741,
+      "loss": 6.3719,
+      "step": 483
+    },
+    {
+      "epoch": 1.075256873090808,
+      "grad_norm": 0.028199033811688423,
+      "learning_rate": 0.002138271604938272,
+      "loss": 6.6597,
+      "step": 484
+    },
+    {
+      "epoch": 1.0774784782004998,
+      "grad_norm": 0.03533554822206497,
+      "learning_rate": 0.0021358024691358023,
+      "loss": 6.654,
+      "step": 485
+    },
+    {
+      "epoch": 1.0797000833101915,
+      "grad_norm": 0.03641669452190399,
+      "learning_rate": 0.0021333333333333334,
+      "loss": 6.4385,
+      "step": 486
+    },
+    {
+      "epoch": 1.0819216884198835,
+      "grad_norm": 0.03398216515779495,
+      "learning_rate": 0.002130864197530864,
+      "loss": 6.6636,
+      "step": 487
+    },
+    {
+      "epoch": 1.0841432935295752,
+      "grad_norm": 0.041564710438251495,
+      "learning_rate": 0.0021283950617283953,
+      "loss": 6.8422,
+      "step": 488
+    },
+    {
+      "epoch": 1.0863648986392669,
+      "grad_norm": 0.038841165602207184,
+      "learning_rate": 0.0021259259259259256,
+      "loss": 6.3757,
+      "step": 489
+    },
+    {
+      "epoch": 1.0885865037489586,
+      "grad_norm": 0.033147819340229034,
+      "learning_rate": 0.002123456790123457,
+      "loss": 6.5191,
+      "step": 490
+    },
+    {
+      "epoch": 1.0908081088586503,
+      "grad_norm": 0.02892182767391205,
+      "learning_rate": 0.002120987654320988,
+      "loss": 6.4549,
+      "step": 491
+    },
+    {
+      "epoch": 1.093029713968342,
+      "grad_norm": 0.029515832662582397,
+      "learning_rate": 0.0021185185185185187,
+      "loss": 6.2039,
+      "step": 492
+    },
+    {
+      "epoch": 1.095251319078034,
+      "grad_norm": 0.03112645447254181,
+      "learning_rate": 0.0021160493827160494,
+      "loss": 6.5373,
+      "step": 493
+    },
+    {
+      "epoch": 1.0974729241877257,
+      "grad_norm": 0.03828863427042961,
+      "learning_rate": 0.00211358024691358,
+      "loss": 6.6497,
+      "step": 494
+    },
+    {
+      "epoch": 1.0996945292974174,
+      "grad_norm": 0.033057838678359985,
+      "learning_rate": 0.0021111111111111113,
+      "loss": 6.3881,
+      "step": 495
+    },
+    {
+      "epoch": 1.1019161344071091,
+      "grad_norm": 0.027790391817688942,
+      "learning_rate": 0.002108641975308642,
+      "loss": 6.4662,
+      "step": 496
+    },
+    {
+      "epoch": 1.1041377395168008,
+      "grad_norm": 0.03252388536930084,
+      "learning_rate": 0.002106172839506173,
+      "loss": 6.5765,
+      "step": 497
+    },
+    {
+      "epoch": 1.1063593446264925,
+      "grad_norm": 0.03879280760884285,
+      "learning_rate": 0.002103703703703704,
+      "loss": 6.2594,
+      "step": 498
+    },
+    {
+      "epoch": 1.1085809497361845,
+      "grad_norm": 0.04156181216239929,
+      "learning_rate": 0.0021012345679012347,
+      "loss": 7.0039,
+      "step": 499
+    },
+    {
+      "epoch": 1.1108025548458762,
+      "grad_norm": 0.03543870523571968,
+      "learning_rate": 0.0020987654320987655,
+      "loss": 6.3444,
+      "step": 500
+    },
+    {
+      "epoch": 1.113024159955568,
+      "grad_norm": 0.05665072053670883,
+      "learning_rate": 0.002096296296296296,
+      "loss": 6.3474,
+      "step": 501
+    },
+    {
+      "epoch": 1.1152457650652596,
+      "grad_norm": 0.0283832885324955,
+      "learning_rate": 0.0020938271604938274,
+      "loss": 6.5432,
+      "step": 502
+    },
+    {
+      "epoch": 1.1174673701749513,
+      "grad_norm": 0.03598196804523468,
+      "learning_rate": 0.002091358024691358,
+      "loss": 6.3492,
+      "step": 503
+    },
+    {
+      "epoch": 1.119688975284643,
+      "grad_norm": 0.03885941207408905,
+      "learning_rate": 0.002088888888888889,
+      "loss": 6.5366,
+      "step": 504
+    },
+    {
+      "epoch": 1.121910580394335,
+      "grad_norm": 0.0317206010222435,
+      "learning_rate": 0.00208641975308642,
+      "loss": 6.4539,
+      "step": 505
+    },
+    {
+      "epoch": 1.1241321855040267,
+      "grad_norm": 0.040018461644649506,
+      "learning_rate": 0.0020839506172839507,
+      "loss": 6.2432,
+      "step": 506
+    },
+    {
+      "epoch": 1.1263537906137184,
+      "grad_norm": 0.04189518094062805,
+      "learning_rate": 0.0020814814814814815,
+      "loss": 6.5315,
+      "step": 507
+    },
+    {
+      "epoch": 1.1285753957234101,
+      "grad_norm": 0.02969651110470295,
+      "learning_rate": 0.002079012345679012,
+      "loss": 6.2752,
+      "step": 508
+    },
+    {
+      "epoch": 1.1307970008331019,
+      "grad_norm": 0.037640877068042755,
+      "learning_rate": 0.0020765432098765434,
+      "loss": 6.667,
+      "step": 509
+    },
+    {
+      "epoch": 1.1330186059427936,
+      "grad_norm": 0.035663068294525146,
+      "learning_rate": 0.002074074074074074,
+      "loss": 6.5661,
+      "step": 510
+    },
+    {
+      "epoch": 1.1352402110524855,
+      "grad_norm": 0.04258381202816963,
+      "learning_rate": 0.002071604938271605,
+      "loss": 6.4601,
+      "step": 511
+    },
+    {
+      "epoch": 1.1374618161621772,
+      "grad_norm": 0.0322837308049202,
+      "learning_rate": 0.0020691358024691356,
+      "loss": 6.2217,
+      "step": 512
+    },
+    {
+      "epoch": 1.139683421271869,
+      "grad_norm": 0.0373774953186512,
+      "learning_rate": 0.0020666666666666667,
+      "loss": 6.1486,
+      "step": 513
+    },
+    {
+      "epoch": 1.1419050263815607,
+      "grad_norm": 0.03975836560130119,
+      "learning_rate": 0.002064197530864198,
+      "loss": 6.3465,
+      "step": 514
+    },
+    {
+      "epoch": 1.1441266314912524,
+      "grad_norm": 0.029985908418893814,
+      "learning_rate": 0.002061728395061728,
+      "loss": 6.6245,
+      "step": 515
+    },
+    {
+      "epoch": 1.146348236600944,
+      "grad_norm": 0.03610779717564583,
+      "learning_rate": 0.0020592592592592594,
+      "loss": 6.0818,
+      "step": 516
+    },
+    {
+      "epoch": 1.148569841710636,
+      "grad_norm": 0.03605785220861435,
+      "learning_rate": 0.00205679012345679,
+      "loss": 6.6951,
+      "step": 517
+    },
+    {
+      "epoch": 1.1507914468203277,
+      "grad_norm": 0.03263959288597107,
+      "learning_rate": 0.0020543209876543213,
+      "loss": 6.3446,
+      "step": 518
+    },
+    {
+      "epoch": 1.1530130519300195,
+      "grad_norm": 0.0429079495370388,
+      "learning_rate": 0.0020518518518518516,
+      "loss": 6.4425,
+      "step": 519
+    },
+    {
+      "epoch": 1.1552346570397112,
+      "grad_norm": 0.04839729890227318,
+      "learning_rate": 0.0020493827160493827,
+      "loss": 6.3836,
+      "step": 520
+    },
+    {
+      "epoch": 1.157456262149403,
+      "grad_norm": 0.03985445573925972,
+      "learning_rate": 0.002046913580246914,
+      "loss": 6.3422,
+      "step": 521
+    },
+    {
+      "epoch": 1.1596778672590946,
+      "grad_norm": 0.03695117309689522,
+      "learning_rate": 0.0020444444444444447,
+      "loss": 6.5901,
+      "step": 522
+    },
+    {
+      "epoch": 1.1618994723687863,
+      "grad_norm": 0.03463996574282646,
+      "learning_rate": 0.0020419753086419754,
+      "loss": 6.4557,
+      "step": 523
+    },
+    {
+      "epoch": 1.1641210774784783,
+      "grad_norm": 0.037514619529247284,
+      "learning_rate": 0.002039506172839506,
+      "loss": 6.0191,
+      "step": 524
+    },
+    {
+      "epoch": 1.16634268258817,
+      "grad_norm": 0.03486974164843559,
+      "learning_rate": 0.0020370370370370373,
+      "loss": 6.487,
+      "step": 525
+    },
+    {
+      "epoch": 1.1685642876978617,
+      "grad_norm": 0.05132891237735748,
+      "learning_rate": 0.002034567901234568,
+      "loss": 6.6569,
+      "step": 526
+    },
+    {
+      "epoch": 1.1707858928075534,
+      "grad_norm": 0.040708813816308975,
+      "learning_rate": 0.0020320987654320988,
+      "loss": 6.056,
+      "step": 527
+    },
+    {
+      "epoch": 1.1730074979172451,
+      "grad_norm": 0.044833164662122726,
+      "learning_rate": 0.00202962962962963,
+      "loss": 6.388,
+      "step": 528
+    },
+    {
+      "epoch": 1.175229103026937,
+      "grad_norm": 0.027963142842054367,
+      "learning_rate": 0.0020271604938271607,
+      "loss": 6.5797,
+      "step": 529
+    },
+    {
+      "epoch": 1.1774507081366288,
+      "grad_norm": 0.0402199923992157,
+      "learning_rate": 0.0020246913580246914,
+      "loss": 6.7003,
+      "step": 530
+    },
+    {
+      "epoch": 1.1796723132463205,
+      "grad_norm": 0.04106440767645836,
+      "learning_rate": 0.002022222222222222,
+      "loss": 6.3391,
+      "step": 531
+    },
+    {
+      "epoch": 1.1818939183560122,
+      "grad_norm": 0.03796447440981865,
+      "learning_rate": 0.0020197530864197533,
+      "loss": 6.4775,
+      "step": 532
+    },
+    {
+      "epoch": 1.184115523465704,
+      "grad_norm": 0.0440424308180809,
+      "learning_rate": 0.002017283950617284,
+      "loss": 6.4393,
+      "step": 533
+    },
+    {
+      "epoch": 1.1863371285753956,
+      "grad_norm": 0.041621048003435135,
+      "learning_rate": 0.0020148148148148148,
+      "loss": 6.1435,
+      "step": 534
+    },
+    {
+      "epoch": 1.1885587336850874,
+      "grad_norm": 0.0351857990026474,
+      "learning_rate": 0.002012345679012346,
+      "loss": 6.2345,
+      "step": 535
+    },
+    {
+      "epoch": 1.1907803387947793,
+      "grad_norm": 0.04070345312356949,
+      "learning_rate": 0.0020098765432098767,
+      "loss": 6.0336,
+      "step": 536
+    },
+    {
+      "epoch": 1.193001943904471,
+      "grad_norm": 0.03114120475947857,
+      "learning_rate": 0.0020074074074074074,
+      "loss": 6.4161,
+      "step": 537
+    },
+    {
+      "epoch": 1.1952235490141627,
+      "grad_norm": 0.04014582931995392,
+      "learning_rate": 0.002004938271604938,
+      "loss": 6.3247,
+      "step": 538
+    },
+    {
+      "epoch": 1.1974451541238544,
+      "grad_norm": 0.03971127048134804,
+      "learning_rate": 0.0020024691358024693,
+      "loss": 6.4709,
+      "step": 539
+    },
+    {
+      "epoch": 1.1996667592335462,
+      "grad_norm": 0.03852924704551697,
+      "learning_rate": 0.002,
+      "loss": 6.3638,
+      "step": 540
+    },
+    {
+      "epoch": 1.201888364343238,
+      "grad_norm": 0.03519389033317566,
+      "learning_rate": 0.0019975308641975308,
+      "loss": 6.6442,
+      "step": 541
+    },
+    {
+      "epoch": 1.2041099694529298,
+      "grad_norm": 0.04014746472239494,
+      "learning_rate": 0.0019950617283950615,
+      "loss": 6.3269,
+      "step": 542
+    },
+    {
+      "epoch": 1.2063315745626215,
+      "grad_norm": 0.03208151459693909,
+      "learning_rate": 0.0019925925925925927,
+      "loss": 6.2944,
+      "step": 543
+    },
+    {
+      "epoch": 1.2085531796723132,
+      "grad_norm": 0.03822821378707886,
+      "learning_rate": 0.001990123456790124,
+      "loss": 6.6407,
+      "step": 544
+    },
+    {
+      "epoch": 1.210774784782005,
+      "grad_norm": 0.046587634831666946,
+      "learning_rate": 0.001987654320987654,
+      "loss": 6.593,
+      "step": 545
+    },
+    {
+      "epoch": 1.2129963898916967,
+      "grad_norm": 0.042928680777549744,
+      "learning_rate": 0.0019851851851851853,
+      "loss": 6.3402,
+      "step": 546
+    },
+    {
+      "epoch": 1.2152179950013884,
+      "grad_norm": 0.03751085326075554,
+      "learning_rate": 0.001982716049382716,
+      "loss": 6.3294,
+      "step": 547
+    },
+    {
+      "epoch": 1.2174396001110803,
+      "grad_norm": 0.03686273843050003,
+      "learning_rate": 0.001980246913580247,
+      "loss": 6.4659,
+      "step": 548
+    },
+    {
+      "epoch": 1.219661205220772,
+      "grad_norm": 0.03327074274420738,
+      "learning_rate": 0.0019777777777777775,
+      "loss": 6.2969,
+      "step": 549
+    },
+    {
+      "epoch": 1.2218828103304638,
+      "grad_norm": 0.032279789447784424,
+      "learning_rate": 0.0019753086419753087,
+      "loss": 6.1469,
+      "step": 550
+    },
+    {
+      "epoch": 1.2241044154401555,
+      "grad_norm": 0.051666878163814545,
+      "learning_rate": 0.00197283950617284,
+      "loss": 6.2377,
+      "step": 551
+    },
+    {
+      "epoch": 1.2263260205498472,
+      "grad_norm": 0.0387309193611145,
+      "learning_rate": 0.00197037037037037,
+      "loss": 6.0148,
+      "step": 552
+    },
+    {
+      "epoch": 1.2285476256595391,
+      "grad_norm": 0.04013807326555252,
+      "learning_rate": 0.0019679012345679013,
+      "loss": 6.3446,
+      "step": 553
+    },
+    {
+      "epoch": 1.2307692307692308,
+      "grad_norm": 0.03821302205324173,
+      "learning_rate": 0.001965432098765432,
+      "loss": 6.666,
+      "step": 554
+    },
+    {
+      "epoch": 1.2329908358789226,
+      "grad_norm": 0.03850986063480377,
+      "learning_rate": 0.0019629629629629632,
+      "loss": 6.3285,
+      "step": 555
+    },
+    {
+      "epoch": 1.2352124409886143,
+      "grad_norm": 0.036970287561416626,
+      "learning_rate": 0.0019604938271604935,
+      "loss": 6.5845,
+      "step": 556
+    },
+    {
+      "epoch": 1.237434046098306,
+      "grad_norm": 0.03782539442181587,
+      "learning_rate": 0.0019580246913580247,
+      "loss": 6.3414,
+      "step": 557
+    },
+    {
+      "epoch": 1.2396556512079977,
+      "grad_norm": 0.038746099919080734,
+      "learning_rate": 0.001955555555555556,
+      "loss": 6.5084,
+      "step": 558
+    },
+    {
+      "epoch": 1.2418772563176894,
+      "grad_norm": 0.03104456327855587,
+      "learning_rate": 0.0019530864197530864,
+      "loss": 6.1361,
+      "step": 559
+    },
+    {
+      "epoch": 1.2440988614273814,
+      "grad_norm": 0.039266884326934814,
+      "learning_rate": 0.0019506172839506173,
+      "loss": 6.5168,
+      "step": 560
+    },
+    {
+      "epoch": 1.246320466537073,
+      "grad_norm": 0.04455557093024254,
+      "learning_rate": 0.001948148148148148,
+      "loss": 6.4887,
+      "step": 561
+    },
+    {
+      "epoch": 1.2485420716467648,
+      "grad_norm": 0.03722786903381348,
+      "learning_rate": 0.001945679012345679,
+      "loss": 6.3669,
+      "step": 562
+    },
+    {
+      "epoch": 1.2507636767564565,
+      "grad_norm": 0.06070278212428093,
+      "learning_rate": 0.0019432098765432098,
+      "loss": 6.1768,
+      "step": 563
+    },
+    {
+      "epoch": 1.2529852818661482,
+      "grad_norm": 0.03736662119626999,
+      "learning_rate": 0.0019407407407407407,
+      "loss": 6.2368,
+      "step": 564
+    },
+    {
+      "epoch": 1.2552068869758402,
+      "grad_norm": 0.04633970558643341,
+      "learning_rate": 0.0019382716049382714,
+      "loss": 6.4326,
+      "step": 565
+    },
+    {
+      "epoch": 1.2574284920855319,
+      "grad_norm": 0.03372138366103172,
+      "learning_rate": 0.0019358024691358024,
+      "loss": 6.2443,
+      "step": 566
+    },
+    {
+      "epoch": 1.2596500971952236,
+      "grad_norm": 0.0386214554309845,
+      "learning_rate": 0.0019333333333333336,
+      "loss": 6.2969,
+      "step": 567
+    },
+    {
+      "epoch": 1.2618717023049153,
+      "grad_norm": 0.05762870982289314,
+      "learning_rate": 0.001930864197530864,
+      "loss": 6.321,
+      "step": 568
+    },
+    {
+      "epoch": 1.264093307414607,
+      "grad_norm": 0.03465733677148819,
+      "learning_rate": 0.0019283950617283952,
+      "loss": 6.2286,
+      "step": 569
+    },
+    {
+      "epoch": 1.2663149125242987,
+      "grad_norm": 0.02819642797112465,
+      "learning_rate": 0.0019259259259259258,
+      "loss": 6.0266,
+      "step": 570
+    },
+    {
+      "epoch": 1.2685365176339904,
+      "grad_norm": 0.033979251980781555,
+      "learning_rate": 0.001923456790123457,
+      "loss": 6.6357,
+      "step": 571
+    },
+    {
+      "epoch": 1.2707581227436824,
+      "grad_norm": 0.033352021127939224,
+      "learning_rate": 0.0019209876543209875,
+      "loss": 6.1982,
+      "step": 572
+    },
+    {
+      "epoch": 1.272979727853374,
+      "grad_norm": 0.0427844263613224,
+      "learning_rate": 0.0019185185185185186,
+      "loss": 6.2508,
+      "step": 573
+    },
+    {
+      "epoch": 1.2752013329630658,
+      "grad_norm": 0.03688507154583931,
+      "learning_rate": 0.0019160493827160496,
+      "loss": 6.3659,
+      "step": 574
+    },
+    {
+      "epoch": 1.2774229380727575,
+      "grad_norm": 0.04512680694460869,
+      "learning_rate": 0.0019135802469135803,
+      "loss": 6.5339,
+      "step": 575
+    },
+    {
+      "epoch": 1.2796445431824492,
+      "grad_norm": 0.03090626187622547,
+      "learning_rate": 0.0019111111111111113,
+      "loss": 6.2844,
+      "step": 576
+    },
+    {
+      "epoch": 1.2818661482921412,
+      "grad_norm": 0.034483786672353745,
+      "learning_rate": 0.001908641975308642,
+      "loss": 6.3264,
+      "step": 577
+    },
+    {
+      "epoch": 1.284087753401833,
+      "grad_norm": 0.031084293499588966,
+      "learning_rate": 0.001906172839506173,
+      "loss": 6.1903,
+      "step": 578
+    },
+    {
+      "epoch": 1.2863093585115246,
+      "grad_norm": 0.033466219902038574,
+      "learning_rate": 0.0019037037037037037,
+      "loss": 6.0493,
+      "step": 579
+    },
+    {
+      "epoch": 1.2885309636212163,
+      "grad_norm": 0.053052324801683426,
+      "learning_rate": 0.0019012345679012346,
+      "loss": 6.2936,
+      "step": 580
+    },
+    {
+      "epoch": 1.290752568730908,
+      "grad_norm": 0.03073807992041111,
+      "learning_rate": 0.0018987654320987656,
+      "loss": 6.1623,
+      "step": 581
+    },
+    {
+      "epoch": 1.2929741738405998,
+      "grad_norm": 0.029447197914123535,
+      "learning_rate": 0.0018962962962962963,
+      "loss": 6.1734,
+      "step": 582
+    },
+    {
+      "epoch": 1.2951957789502915,
+      "grad_norm": 0.03545555844902992,
+      "learning_rate": 0.0018938271604938273,
+      "loss": 6.067,
+      "step": 583
+    },
+    {
+      "epoch": 1.2974173840599834,
+      "grad_norm": 0.030690601095557213,
+      "learning_rate": 0.001891358024691358,
+      "loss": 6.3609,
+      "step": 584
+    },
+    {
+      "epoch": 1.2996389891696751,
+      "grad_norm": 0.031808044761419296,
+      "learning_rate": 0.001888888888888889,
+      "loss": 6.5173,
+      "step": 585
+    },
+    {
+      "epoch": 1.3018605942793668,
+      "grad_norm": 0.05404243990778923,
+      "learning_rate": 0.0018864197530864197,
+      "loss": 6.6066,
+      "step": 586
+    },
+    {
+      "epoch": 1.3040821993890586,
+      "grad_norm": 0.03418256714940071,
+      "learning_rate": 0.0018839506172839506,
+      "loss": 6.3429,
+      "step": 587
+    },
+    {
+      "epoch": 1.3063038044987503,
+      "grad_norm": 0.0333407036960125,
+      "learning_rate": 0.0018814814814814816,
+      "loss": 6.542,
+      "step": 588
+    },
+    {
+      "epoch": 1.3085254096084422,
+      "grad_norm": 0.04213091358542442,
+      "learning_rate": 0.0018790123456790123,
+      "loss": 5.9407,
+      "step": 589
+    },
+    {
+      "epoch": 1.310747014718134,
+      "grad_norm": 0.03965051472187042,
+      "learning_rate": 0.0018765432098765433,
+      "loss": 6.323,
+      "step": 590
+    },
+    {
+      "epoch": 1.3129686198278256,
+      "grad_norm": 0.041978783905506134,
+      "learning_rate": 0.001874074074074074,
+      "loss": 6.1851,
+      "step": 591
+    },
+    {
+      "epoch": 1.3151902249375174,
+      "grad_norm": 0.030312292277812958,
+      "learning_rate": 0.001871604938271605,
+      "loss": 6.5231,
+      "step": 592
+    },
+    {
+      "epoch": 1.317411830047209,
+      "grad_norm": 0.04681335762143135,
+      "learning_rate": 0.0018691358024691357,
+      "loss": 6.3259,
+      "step": 593
+    },
+    {
+      "epoch": 1.3196334351569008,
+      "grad_norm": 0.04509252309799194,
+      "learning_rate": 0.0018666666666666666,
+      "loss": 6.208,
+      "step": 594
+    },
+    {
+      "epoch": 1.3218550402665925,
+      "grad_norm": 0.03200953081250191,
+      "learning_rate": 0.0018641975308641974,
+      "loss": 6.1602,
+      "step": 595
+    },
+    {
+      "epoch": 1.3240766453762844,
+      "grad_norm": 0.032900337129831314,
+      "learning_rate": 0.0018617283950617283,
+      "loss": 6.129,
+      "step": 596
+    },
+    {
+      "epoch": 1.3262982504859762,
+      "grad_norm": 0.03652816638350487,
+      "learning_rate": 0.0018592592592592595,
+      "loss": 6.3783,
+      "step": 597
+    },
+    {
+      "epoch": 1.3285198555956679,
+      "grad_norm": 0.040998030453920364,
+      "learning_rate": 0.00185679012345679,
+      "loss": 6.869,
+      "step": 598
+    },
+    {
+      "epoch": 1.3307414607053596,
+      "grad_norm": 0.036215465515851974,
+      "learning_rate": 0.0018543209876543212,
+      "loss": 6.324,
+      "step": 599
+    },
+    {
+      "epoch": 1.3329630658150513,
+      "grad_norm": 0.04111867398023605,
+      "learning_rate": 0.0018518518518518517,
+      "loss": 6.2809,
+      "step": 600
+    },
+    {
+      "epoch": 1.3351846709247432,
+      "grad_norm": 0.041931070387363434,
+      "learning_rate": 0.0018493827160493829,
+      "loss": 6.5084,
+      "step": 601
+    },
+    {
+      "epoch": 1.337406276034435,
+      "grad_norm": 0.03920350596308708,
+      "learning_rate": 0.0018469135802469134,
+      "loss": 6.7175,
+      "step": 602
+    },
+    {
+      "epoch": 1.3396278811441267,
+      "grad_norm": 0.03189566358923912,
+      "learning_rate": 0.0018444444444444446,
+      "loss": 6.4014,
+      "step": 603
+    },
+    {
+      "epoch": 1.3418494862538184,
+      "grad_norm": 0.04909387603402138,
+      "learning_rate": 0.0018419753086419755,
+      "loss": 6.4304,
+      "step": 604
+    },
+    {
+      "epoch": 1.34407109136351,
+      "grad_norm": 0.037338417023420334,
+      "learning_rate": 0.0018395061728395062,
+      "loss": 6.5664,
+      "step": 605
+    },
+    {
+      "epoch": 1.3462926964732018,
+      "grad_norm": 0.040651820600032806,
+      "learning_rate": 0.0018370370370370372,
+      "loss": 6.4001,
+      "step": 606
+    },
+    {
+      "epoch": 1.3485143015828935,
+      "grad_norm": 0.0633411705493927,
+      "learning_rate": 0.001834567901234568,
+      "loss": 6.1285,
+      "step": 607
+    },
+    {
+      "epoch": 1.3507359066925855,
+      "grad_norm": 0.045840244740247726,
+      "learning_rate": 0.0018320987654320989,
+      "loss": 6.5343,
+      "step": 608
+    },
+    {
+      "epoch": 1.3529575118022772,
+      "grad_norm": 0.031113650649785995,
+      "learning_rate": 0.0018296296296296296,
+      "loss": 6.5995,
+      "step": 609
+    },
+    {
+      "epoch": 1.355179116911969,
+      "grad_norm": 0.03467819094657898,
+      "learning_rate": 0.0018271604938271606,
+      "loss": 6.3797,
+      "step": 610
+    },
+    {
+      "epoch": 1.3574007220216606,
+      "grad_norm": 0.05771856755018234,
+      "learning_rate": 0.0018246913580246915,
+      "loss": 6.1617,
+      "step": 611
+    },
+    {
+      "epoch": 1.3596223271313523,
+      "grad_norm": 0.0407741442322731,
+      "learning_rate": 0.0018222222222222223,
+      "loss": 6.3389,
+      "step": 612
+    },
+    {
+      "epoch": 1.3618439322410443,
+      "grad_norm": 0.04614827781915665,
+      "learning_rate": 0.0018197530864197532,
+      "loss": 6.4762,
+      "step": 613
+    },
+    {
+      "epoch": 1.364065537350736,
+      "grad_norm": 0.04630822315812111,
+      "learning_rate": 0.001817283950617284,
+      "loss": 6.1302,
+      "step": 614
+    },
+    {
+      "epoch": 1.3662871424604277,
+      "grad_norm": 0.04013388976454735,
+      "learning_rate": 0.001814814814814815,
+      "loss": 6.2063,
+      "step": 615
+    },
+    {
+      "epoch": 1.3685087475701194,
+      "grad_norm": 0.04549866542220116,
+      "learning_rate": 0.0018123456790123456,
+      "loss": 6.4473,
+      "step": 616
+    },
+    {
+      "epoch": 1.3707303526798111,
+      "grad_norm": 0.04472465440630913,
+      "learning_rate": 0.0018098765432098766,
+      "loss": 6.1478,
+      "step": 617
+    },
+    {
+      "epoch": 1.3729519577895029,
+      "grad_norm": 0.04715588688850403,
+      "learning_rate": 0.0018074074074074075,
+      "loss": 6.171,
+      "step": 618
+    },
+    {
+      "epoch": 1.3751735628991946,
+      "grad_norm": 0.04618202894926071,
+      "learning_rate": 0.0018049382716049383,
+      "loss": 5.8558,
+      "step": 619
+    },
+    {
+      "epoch": 1.3773951680088863,
+      "grad_norm": 0.05153551697731018,
+      "learning_rate": 0.0018024691358024692,
+      "loss": 6.2804,
+      "step": 620
+    },
+    {
+      "epoch": 1.3796167731185782,
+      "grad_norm": 0.04584895446896553,
+      "learning_rate": 0.0018,
+      "loss": 6.3909,
+      "step": 621
+    },
+    {
+      "epoch": 1.38183837822827,
+      "grad_norm": 0.03741832450032234,
+      "learning_rate": 0.001797530864197531,
+      "loss": 6.2864,
+      "step": 622
+    },
+    {
+      "epoch": 1.3840599833379617,
+      "grad_norm": 0.05302848666906357,
+      "learning_rate": 0.0017950617283950616,
+      "loss": 6.1101,
+      "step": 623
+    },
+    {
+      "epoch": 1.3862815884476534,
+      "grad_norm": 0.05580468848347664,
+      "learning_rate": 0.0017925925925925926,
+      "loss": 6.4619,
+      "step": 624
+    },
+    {
+      "epoch": 1.3885031935573453,
+      "grad_norm": 0.03729956969618797,
+      "learning_rate": 0.0017901234567901233,
+      "loss": 6.3855,
+      "step": 625
+    },
+    {
+      "epoch": 1.390724798667037,
+      "grad_norm": 0.03489207103848457,
+      "learning_rate": 0.0017876543209876543,
+      "loss": 6.1936,
+      "step": 626
+    },
+    {
+      "epoch": 1.3929464037767287,
+      "grad_norm": 0.035313766449689865,
+      "learning_rate": 0.0017851851851851854,
+      "loss": 6.009,
+      "step": 627
+    },
+    {
+      "epoch": 1.3951680088864205,
+      "grad_norm": 0.04091987758874893,
+      "learning_rate": 0.001782716049382716,
+      "loss": 6.1428,
+      "step": 628
+    },
+    {
+      "epoch": 1.3973896139961122,
+      "grad_norm": 0.03116190992295742,
+      "learning_rate": 0.0017802469135802471,
+      "loss": 6.0366,
+      "step": 629
+    },
+    {
+      "epoch": 1.3996112191058039,
+      "grad_norm": 0.03686531260609627,
+      "learning_rate": 0.0017777777777777776,
+      "loss": 6.3959,
+      "step": 630
+    },
+    {
+      "epoch": 1.4018328242154956,
+      "grad_norm": 0.03318272531032562,
+      "learning_rate": 0.0017753086419753088,
+      "loss": 6.4128,
+      "step": 631
+    },
+    {
+      "epoch": 1.4040544293251873,
+      "grad_norm": 0.041839905083179474,
+      "learning_rate": 0.0017728395061728393,
+      "loss": 6.5429,
+      "step": 632
+    },
+    {
+      "epoch": 1.4062760344348793,
+      "grad_norm": 0.030093267560005188,
+      "learning_rate": 0.0017703703703703705,
+      "loss": 6.2587,
+      "step": 633
+    },
+    {
+      "epoch": 1.408497639544571,
+      "grad_norm": 0.05865675210952759,
+      "learning_rate": 0.0017679012345679015,
+      "loss": 6.2798,
+      "step": 634
+    },
+    {
+      "epoch": 1.4107192446542627,
+      "grad_norm": 0.0343112088739872,
+      "learning_rate": 0.0017654320987654322,
+      "loss": 6.1428,
+      "step": 635
+    },
+    {
+      "epoch": 1.4129408497639544,
+      "grad_norm": 0.03351147472858429,
+      "learning_rate": 0.0017629629629629631,
+      "loss": 6.2343,
+      "step": 636
+    },
+    {
+      "epoch": 1.4151624548736463,
+      "grad_norm": 0.034447863698005676,
+      "learning_rate": 0.0017604938271604939,
+      "loss": 6.3156,
+      "step": 637
+    },
+    {
+      "epoch": 1.417384059983338,
+      "grad_norm": 0.03670838847756386,
+      "learning_rate": 0.0017580246913580248,
+      "loss": 6.3385,
+      "step": 638
+    },
+    {
+      "epoch": 1.4196056650930298,
+      "grad_norm": 0.03930442035198212,
+      "learning_rate": 0.0017555555555555556,
+      "loss": 6.1479,
+      "step": 639
+    },
+    {
+      "epoch": 1.4218272702027215,
+      "grad_norm": 0.03430639207363129,
+      "learning_rate": 0.0017530864197530865,
+      "loss": 6.1624,
+      "step": 640
+    },
+    {
+      "epoch": 1.4240488753124132,
+      "grad_norm": 0.04005846008658409,
+      "learning_rate": 0.0017506172839506175,
+      "loss": 6.3732,
+      "step": 641
+    },
+    {
+      "epoch": 1.426270480422105,
+      "grad_norm": 0.041829563677310944,
+      "learning_rate": 0.0017481481481481482,
+      "loss": 6.283,
+      "step": 642
+    },
+    {
+      "epoch": 1.4284920855317966,
+      "grad_norm": 0.03966071456670761,
+      "learning_rate": 0.0017456790123456791,
+      "loss": 6.3854,
+      "step": 643
+    },
+    {
+      "epoch": 1.4307136906414883,
+      "grad_norm": 0.04266310855746269,
+      "learning_rate": 0.0017432098765432099,
+      "loss": 5.8232,
+      "step": 644
+    },
+    {
+      "epoch": 1.4329352957511803,
+      "grad_norm": 0.034664738923311234,
+      "learning_rate": 0.0017407407407407408,
+      "loss": 6.1072,
+      "step": 645
+    },
+    {
+      "epoch": 1.435156900860872,
+      "grad_norm": 0.03348352760076523,
+      "learning_rate": 0.0017382716049382716,
+      "loss": 6.2337,
+      "step": 646
+    },
+    {
+      "epoch": 1.4373785059705637,
+      "grad_norm": 0.04647333547472954,
+      "learning_rate": 0.0017358024691358025,
+      "loss": 6.2249,
+      "step": 647
+    },
+    {
+      "epoch": 1.4396001110802554,
+      "grad_norm": 0.03087199293076992,
+      "learning_rate": 0.0017333333333333333,
+      "loss": 5.9,
+      "step": 648
+    },
+    {
+      "epoch": 1.4418217161899474,
+      "grad_norm": 0.031058957800269127,
+      "learning_rate": 0.0017308641975308642,
+      "loss": 6.2423,
+      "step": 649
+    },
+    {
+      "epoch": 1.444043321299639,
+      "grad_norm": 0.032819394022226334,
+      "learning_rate": 0.0017283950617283952,
+      "loss": 6.3537,
+      "step": 650
+    },
+    {
+      "epoch": 1.4462649264093308,
+      "grad_norm": 0.047803789377212524,
+      "learning_rate": 0.0017259259259259259,
+      "loss": 6.1019,
+      "step": 651
+    },
+    {
+      "epoch": 1.4484865315190225,
+      "grad_norm": 0.03795769438147545,
+      "learning_rate": 0.0017234567901234568,
+      "loss": 6.037,
+      "step": 652
+    },
+    {
+      "epoch": 1.4507081366287142,
+      "grad_norm": 0.042528439313173294,
+      "learning_rate": 0.0017209876543209876,
+      "loss": 6.3609,
+      "step": 653
+    },
+    {
+      "epoch": 1.452929741738406,
+      "grad_norm": 0.04530796408653259,
+      "learning_rate": 0.0017185185185185185,
+      "loss": 6.4745,
+      "step": 654
+    },
+    {
+      "epoch": 1.4551513468480977,
+      "grad_norm": 0.03794233873486519,
+      "learning_rate": 0.0017160493827160493,
+      "loss": 6.191,
+      "step": 655
+    },
+    {
+      "epoch": 1.4573729519577894,
+      "grad_norm": 0.045053884387016296,
+      "learning_rate": 0.0017135802469135802,
+      "loss": 6.085,
+      "step": 656
+    },
+    {
+      "epoch": 1.4595945570674813,
+      "grad_norm": 0.05708610638976097,
+      "learning_rate": 0.0017111111111111114,
+      "loss": 6.0017,
+      "step": 657
+    },
+    {
+      "epoch": 1.461816162177173,
+      "grad_norm": 0.03785408288240433,
+      "learning_rate": 0.001708641975308642,
+      "loss": 6.3146,
+      "step": 658
+    },
+    {
+      "epoch": 1.4640377672868647,
+      "grad_norm": 0.05379173904657364,
+      "learning_rate": 0.001706172839506173,
+      "loss": 5.7694,
+      "step": 659
+    },
+    {
+      "epoch": 1.4662593723965565,
+      "grad_norm": 0.043331608176231384,
+      "learning_rate": 0.0017037037037037036,
+      "loss": 6.3276,
+      "step": 660
+    },
+    {
+      "epoch": 1.4684809775062484,
+      "grad_norm": 0.03599626198410988,
+      "learning_rate": 0.0017012345679012348,
+      "loss": 6.3605,
+      "step": 661
+    },
+    {
+      "epoch": 1.4707025826159401,
+      "grad_norm": 0.0370354950428009,
+      "learning_rate": 0.0016987654320987653,
+      "loss": 6.2993,
+      "step": 662
+    },
+    {
+      "epoch": 1.4729241877256318,
+      "grad_norm": 0.03861134499311447,
+      "learning_rate": 0.0016962962962962964,
+      "loss": 6.4733,
+      "step": 663
+    },
+    {
+      "epoch": 1.4751457928353235,
+      "grad_norm": 0.03696829080581665,
+      "learning_rate": 0.0016938271604938274,
+      "loss": 6.3908,
+      "step": 664
+    },
+    {
+      "epoch": 1.4773673979450153,
+      "grad_norm": 0.03327971696853638,
+      "learning_rate": 0.0016913580246913581,
+      "loss": 5.9808,
+      "step": 665
+    },
+    {
+      "epoch": 1.479589003054707,
+      "grad_norm": 0.0478159636259079,
+      "learning_rate": 0.001688888888888889,
+      "loss": 6.1371,
+      "step": 666
+    },
+    {
+      "epoch": 1.4818106081643987,
+      "grad_norm": 0.03853146731853485,
+      "learning_rate": 0.0016864197530864198,
+      "loss": 6.1711,
+      "step": 667
+    },
+    {
+      "epoch": 1.4840322132740904,
+      "grad_norm": 0.041707154363393784,
+      "learning_rate": 0.0016839506172839508,
+      "loss": 6.2316,
+      "step": 668
+    },
+    {
+      "epoch": 1.4862538183837823,
+      "grad_norm": 0.031753845512866974,
+      "learning_rate": 0.0016814814814814813,
+      "loss": 6.3248,
+      "step": 669
+    },
+    {
+      "epoch": 1.488475423493474,
+      "grad_norm": 0.03383229672908783,
+      "learning_rate": 0.0016790123456790125,
+      "loss": 6.3733,
+      "step": 670
+    },
+    {
+      "epoch": 1.4906970286031658,
+      "grad_norm": 0.05415927246212959,
+      "learning_rate": 0.0016765432098765434,
+      "loss": 6.554,
+      "step": 671
+    },
+    {
+      "epoch": 1.4929186337128575,
+      "grad_norm": 0.05383379012346268,
+      "learning_rate": 0.0016740740740740741,
+      "loss": 6.3365,
+      "step": 672
+    },
+    {
+      "epoch": 1.4951402388225494,
+      "grad_norm": 0.03220098838210106,
+      "learning_rate": 0.001671604938271605,
+      "loss": 6.2972,
+      "step": 673
+    },
+    {
+      "epoch": 1.4973618439322411,
+      "grad_norm": 0.04361923038959503,
+      "learning_rate": 0.0016691358024691358,
+      "loss": 6.1342,
+      "step": 674
+    },
+    {
+      "epoch": 1.4995834490419329,
+      "grad_norm": 0.034433454275131226,
+      "learning_rate": 0.0016666666666666668,
+      "loss": 6.3875,
+      "step": 675
+    },
+    {
+      "epoch": 1.5018050541516246,
+      "grad_norm": 0.03784746676683426,
+      "learning_rate": 0.0016641975308641975,
+      "loss": 6.2219,
+      "step": 676
+    },
+    {
+      "epoch": 1.5040266592613163,
+      "grad_norm": 0.03338635712862015,
+      "learning_rate": 0.0016617283950617285,
+      "loss": 6.5494,
+      "step": 677
+    },
+    {
+      "epoch": 1.506248264371008,
+      "grad_norm": 0.03723134845495224,
+      "learning_rate": 0.0016592592592592592,
+      "loss": 6.1127,
+      "step": 678
+    },
+    {
+      "epoch": 1.5084698694806997,
+      "grad_norm": 0.04377841576933861,
+      "learning_rate": 0.0016567901234567901,
+      "loss": 6.3558,
+      "step": 679
+    },
+    {
+      "epoch": 1.5106914745903914,
+      "grad_norm": 0.05502425879240036,
+      "learning_rate": 0.001654320987654321,
+      "loss": 6.0453,
+      "step": 680
+    },
+    {
+      "epoch": 1.5129130797000832,
+      "grad_norm": 0.06568772345781326,
+      "learning_rate": 0.0016518518518518518,
+      "loss": 6.0028,
+      "step": 681
+    },
+    {
+      "epoch": 1.515134684809775,
+      "grad_norm": 0.033911559730768204,
+      "learning_rate": 0.0016493827160493828,
+      "loss": 6.1029,
+      "step": 682
+    },
+    {
+      "epoch": 1.5173562899194668,
+      "grad_norm": 0.038924045860767365,
+      "learning_rate": 0.0016469135802469135,
+      "loss": 5.99,
+      "step": 683
+    },
+    {
+      "epoch": 1.5195778950291585,
+      "grad_norm": 0.07635977864265442,
+      "learning_rate": 0.0016444444444444445,
+      "loss": 6.0661,
+      "step": 684
+    },
+    {
+      "epoch": 1.5217995001388505,
+      "grad_norm": 0.04160960763692856,
+      "learning_rate": 0.0016419753086419752,
+      "loss": 6.0989,
+      "step": 685
+    },
+    {
+      "epoch": 1.5240211052485422,
+      "grad_norm": 0.04776043817400932,
+      "learning_rate": 0.0016395061728395062,
+      "loss": 6.38,
+      "step": 686
+    },
+    {
+      "epoch": 1.526242710358234,
+      "grad_norm": 0.038796450942754745,
+      "learning_rate": 0.001637037037037037,
+      "loss": 6.2388,
+      "step": 687
+    },
+    {
+      "epoch": 1.5284643154679256,
+      "grad_norm": 0.044839583337306976,
+      "learning_rate": 0.0016345679012345678,
+      "loss": 6.5456,
+      "step": 688
+    },
+    {
+      "epoch": 1.5306859205776173,
+      "grad_norm": 0.038193706423044205,
+      "learning_rate": 0.0016320987654320988,
+      "loss": 6.1368,
+      "step": 689
+    },
+    {
+      "epoch": 1.532907525687309,
+      "grad_norm": 0.03954074904322624,
+      "learning_rate": 0.0016296296296296295,
+      "loss": 6.209,
+      "step": 690
+    },
+    {
+      "epoch": 1.5351291307970008,
+      "grad_norm": 0.04989338293671608,
+      "learning_rate": 0.0016271604938271605,
+      "loss": 6.11,
+      "step": 691
+    },
+    {
+      "epoch": 1.5373507359066925,
+      "grad_norm": 0.03306099399924278,
+      "learning_rate": 0.0016246913580246912,
+      "loss": 6.079,
+      "step": 692
+    },
+    {
+      "epoch": 1.5395723410163842,
+      "grad_norm": 0.049217116087675095,
+      "learning_rate": 0.0016222222222222222,
+      "loss": 6.0589,
+      "step": 693
+    },
+    {
+      "epoch": 1.5417939461260761,
+      "grad_norm": 0.03254625201225281,
+      "learning_rate": 0.0016197530864197533,
+      "loss": 6.3458,
+      "step": 694
+    },
+    {
+      "epoch": 1.5440155512357678,
+      "grad_norm": 0.03594857081770897,
+      "learning_rate": 0.0016172839506172839,
+      "loss": 6.195,
+      "step": 695
+    },
+    {
+      "epoch": 1.5462371563454596,
+      "grad_norm": 0.07235965877771378,
+      "learning_rate": 0.001614814814814815,
+      "loss": 6.005,
+      "step": 696
+    },
+    {
+      "epoch": 1.5484587614551515,
+      "grad_norm": 0.06128450110554695,
+      "learning_rate": 0.0016123456790123455,
+      "loss": 6.5062,
+      "step": 697
+    },
+    {
+      "epoch": 1.5506803665648432,
+      "grad_norm": 0.03547682240605354,
+      "learning_rate": 0.0016098765432098767,
+      "loss": 6.0139,
+      "step": 698
+    },
+    {
+      "epoch": 1.552901971674535,
+      "grad_norm": 0.057652879506349564,
+      "learning_rate": 0.0016074074074074072,
+      "loss": 5.8171,
+      "step": 699
+    },
+    {
+      "epoch": 1.5551235767842266,
+      "grad_norm": 0.04588836058974266,
+      "learning_rate": 0.0016049382716049384,
+      "loss": 5.9613,
+      "step": 700
+    },
+    {
+      "epoch": 1.5573451818939184,
+      "grad_norm": 0.03351445123553276,
+      "learning_rate": 0.0016024691358024693,
+      "loss": 6.394,
+      "step": 701
+    },
+    {
+      "epoch": 1.55956678700361,
+      "grad_norm": 0.037374623119831085,
+      "learning_rate": 0.0016,
+      "loss": 5.9167,
+      "step": 702
+    },
+    {
+      "epoch": 1.5617883921133018,
+      "grad_norm": 0.04311612620949745,
+      "learning_rate": 0.001597530864197531,
+      "loss": 5.935,
+      "step": 703
+    },
+    {
+      "epoch": 1.5640099972229935,
+      "grad_norm": 0.04786458984017372,
+      "learning_rate": 0.0015950617283950618,
+      "loss": 6.3014,
+      "step": 704
+    },
+    {
+      "epoch": 1.5662316023326852,
+      "grad_norm": 0.0439060777425766,
+      "learning_rate": 0.0015925925925925927,
+      "loss": 6.1564,
+      "step": 705
+    },
+    {
+      "epoch": 1.5684532074423772,
+      "grad_norm": 0.04528342932462692,
+      "learning_rate": 0.0015901234567901234,
+      "loss": 6.2511,
+      "step": 706
+    },
+    {
+      "epoch": 1.5706748125520689,
+      "grad_norm": 0.04323355108499527,
+      "learning_rate": 0.0015876543209876544,
+      "loss": 6.2638,
+      "step": 707
+    },
+    {
+      "epoch": 1.5728964176617606,
+      "grad_norm": 0.04894525185227394,
+      "learning_rate": 0.0015851851851851851,
+      "loss": 6.0646,
+      "step": 708
+    },
+    {
+      "epoch": 1.5751180227714525,
+      "grad_norm": 0.044189661741256714,
+      "learning_rate": 0.001582716049382716,
+      "loss": 5.9077,
+      "step": 709
+    },
+    {
+      "epoch": 1.5773396278811442,
+      "grad_norm": 0.05571754276752472,
+      "learning_rate": 0.001580246913580247,
+      "loss": 6.4761,
+      "step": 710
+    },
+    {
+      "epoch": 1.579561232990836,
+      "grad_norm": 0.055017706006765366,
+      "learning_rate": 0.0015777777777777778,
+      "loss": 6.1056,
+      "step": 711
+    },
+    {
+      "epoch": 1.5817828381005277,
+      "grad_norm": 0.03247570991516113,
+      "learning_rate": 0.0015753086419753087,
+      "loss": 6.2023,
+      "step": 712
+    },
+    {
+      "epoch": 1.5840044432102194,
+      "grad_norm": 0.03337739408016205,
+      "learning_rate": 0.0015728395061728395,
+      "loss": 6.2494,
+      "step": 713
+    },
+    {
+      "epoch": 1.586226048319911,
+      "grad_norm": 0.03905110061168671,
+      "learning_rate": 0.0015703703703703704,
+      "loss": 6.1822,
+      "step": 714
+    },
+    {
+      "epoch": 1.5884476534296028,
+      "grad_norm": 0.050255563110113144,
+      "learning_rate": 0.0015679012345679011,
+      "loss": 5.9656,
+      "step": 715
+    },
+    {
+      "epoch": 1.5906692585392945,
+      "grad_norm": 0.040939997881650925,
+      "learning_rate": 0.001565432098765432,
+      "loss": 6.2584,
+      "step": 716
+    },
+    {
+      "epoch": 1.5928908636489862,
+      "grad_norm": 0.04037930443882942,
+      "learning_rate": 0.001562962962962963,
+      "loss": 6.2741,
+      "step": 717
+    },
+    {
+      "epoch": 1.5951124687586782,
+      "grad_norm": 0.04207153990864754,
+      "learning_rate": 0.0015604938271604938,
+      "loss": 6.4213,
+      "step": 718
+    },
+    {
+      "epoch": 1.59733407386837,
+      "grad_norm": 0.03851177915930748,
+      "learning_rate": 0.0015580246913580247,
+      "loss": 6.3191,
+      "step": 719
+    },
+    {
+      "epoch": 1.5995556789780616,
+      "grad_norm": 0.038728807121515274,
+      "learning_rate": 0.0015555555555555555,
+      "loss": 6.2651,
+      "step": 720
+    },
+    {
+      "epoch": 1.6017772840877536,
+      "grad_norm": 0.053951799869537354,
+      "learning_rate": 0.0015530864197530864,
+      "loss": 5.9007,
+      "step": 721
+    },
+    {
+      "epoch": 1.6039988891974453,
+      "grad_norm": 0.03671761229634285,
+      "learning_rate": 0.0015506172839506172,
+      "loss": 6.1562,
+      "step": 722
+    },
+    {
+      "epoch": 1.606220494307137,
+      "grad_norm": 0.04554244130849838,
+      "learning_rate": 0.001548148148148148,
+      "loss": 6.0253,
+      "step": 723
+    },
+    {
+      "epoch": 1.6084420994168287,
+      "grad_norm": 0.053976450115442276,
+      "learning_rate": 0.0015456790123456793,
+      "loss": 6.3573,
+      "step": 724
+    },
+    {
+      "epoch": 1.6106637045265204,
+      "grad_norm": 0.03967395797371864,
+      "learning_rate": 0.0015432098765432098,
+      "loss": 6.2846,
+      "step": 725
+    },
+    {
+      "epoch": 1.6128853096362121,
+      "grad_norm": 0.03604193776845932,
+      "learning_rate": 0.001540740740740741,
+      "loss": 6.2174,
+      "step": 726
+    },
+    {
+      "epoch": 1.6151069147459038,
+      "grad_norm": 0.040343984961509705,
+      "learning_rate": 0.0015382716049382715,
+      "loss": 6.4531,
+      "step": 727
+    },
+    {
+      "epoch": 1.6173285198555956,
+      "grad_norm": 0.04057557135820389,
+      "learning_rate": 0.0015358024691358026,
+      "loss": 6.0287,
+      "step": 728
+    },
+    {
+      "epoch": 1.6195501249652873,
+      "grad_norm": 0.049208082258701324,
+      "learning_rate": 0.0015333333333333332,
+      "loss": 6.3208,
+      "step": 729
+    },
+    {
+      "epoch": 1.6217717300749792,
+      "grad_norm": 0.05058284103870392,
+      "learning_rate": 0.0015308641975308643,
+      "loss": 6.2669,
+      "step": 730
+    },
+    {
+      "epoch": 1.623993335184671,
+      "grad_norm": 0.03900137171149254,
+      "learning_rate": 0.0015283950617283948,
+      "loss": 6.215,
+      "step": 731
+    },
+    {
+      "epoch": 1.6262149402943626,
+      "grad_norm": 0.028513478115200996,
+      "learning_rate": 0.001525925925925926,
+      "loss": 6.3372,
+      "step": 732
+    },
+    {
+      "epoch": 1.6284365454040546,
+      "grad_norm": 0.04768927022814751,
+      "learning_rate": 0.001523456790123457,
+      "loss": 6.4372,
+      "step": 733
+    },
+    {
+      "epoch": 1.6306581505137463,
+      "grad_norm": 0.05876043811440468,
+      "learning_rate": 0.0015209876543209877,
+      "loss": 6.3986,
+      "step": 734
+    },
+    {
+      "epoch": 1.632879755623438,
+      "grad_norm": 0.03884493187069893,
+      "learning_rate": 0.0015185185185185187,
+      "loss": 6.0089,
+      "step": 735
+    },
+    {
+      "epoch": 1.6351013607331297,
+      "grad_norm": 0.044803451746702194,
+      "learning_rate": 0.0015160493827160494,
+      "loss": 6.1839,
+      "step": 736
+    },
+    {
+      "epoch": 1.6373229658428214,
+      "grad_norm": 0.03137345612049103,
+      "learning_rate": 0.0015135802469135803,
+      "loss": 6.3144,
+      "step": 737
+    },
+    {
+      "epoch": 1.6395445709525132,
+      "grad_norm": 0.03509042412042618,
+      "learning_rate": 0.001511111111111111,
+      "loss": 6.5495,
+      "step": 738
+    },
+    {
+      "epoch": 1.6417661760622049,
+      "grad_norm": 0.043359674513339996,
+      "learning_rate": 0.001508641975308642,
+      "loss": 6.0988,
+      "step": 739
+    },
+    {
+      "epoch": 1.6439877811718966,
+      "grad_norm": 0.05285673215985298,
+      "learning_rate": 0.001506172839506173,
+      "loss": 6.4118,
+      "step": 740
+    },
+    {
+      "epoch": 1.6462093862815883,
+      "grad_norm": 0.051178302615880966,
+      "learning_rate": 0.0015037037037037037,
+      "loss": 5.986,
+      "step": 741
+    },
+    {
+      "epoch": 1.6484309913912802,
+      "grad_norm": 0.03230026364326477,
+      "learning_rate": 0.0015012345679012347,
+      "loss": 6.0246,
+      "step": 742
+    },
+    {
+      "epoch": 1.650652596500972,
+      "grad_norm": 0.04651426896452904,
+      "learning_rate": 0.0014987654320987656,
+      "loss": 6.0249,
+      "step": 743
+    },
+    {
+      "epoch": 1.6528742016106637,
+      "grad_norm": 0.05399320647120476,
+      "learning_rate": 0.0014962962962962963,
+      "loss": 6.1514,
+      "step": 744
+    },
+    {
+      "epoch": 1.6550958067203556,
+      "grad_norm": 0.05591241270303726,
+      "learning_rate": 0.0014938271604938273,
+      "loss": 6.3867,
+      "step": 745
+    },
+    {
+      "epoch": 1.6573174118300473,
+      "grad_norm": 0.042732879519462585,
+      "learning_rate": 0.001491358024691358,
+      "loss": 6.1903,
+      "step": 746
+    },
+    {
+      "epoch": 1.659539016939739,
+      "grad_norm": 0.04537397250533104,
+      "learning_rate": 0.001488888888888889,
+      "loss": 6.1143,
+      "step": 747
+    },
+    {
+      "epoch": 1.6617606220494308,
+      "grad_norm": 0.04793582111597061,
+      "learning_rate": 0.0014864197530864197,
+      "loss": 6.221,
+      "step": 748
+    },
+    {
+      "epoch": 1.6639822271591225,
+      "grad_norm": 0.03770630434155464,
+      "learning_rate": 0.0014839506172839507,
+      "loss": 6.183,
+      "step": 749
+    },
+    {
+      "epoch": 1.6662038322688142,
+      "grad_norm": 0.04926915839314461,
+      "learning_rate": 0.0014814814814814814,
+      "loss": 5.9252,
+      "step": 750
+    },
+    {
+      "epoch": 1.668425437378506,
+      "grad_norm": 0.029345575720071793,
+      "learning_rate": 0.0014790123456790124,
+      "loss": 6.0441,
+      "step": 751
+    },
+    {
+      "epoch": 1.6706470424881976,
+      "grad_norm": 0.04108971729874611,
+      "learning_rate": 0.0014765432098765433,
+      "loss": 6.3289,
+      "step": 752
+    },
+    {
+      "epoch": 1.6728686475978893,
+      "grad_norm": 0.052369631826877594,
+      "learning_rate": 0.001474074074074074,
+      "loss": 6.1526,
+      "step": 753
+    },
+    {
+      "epoch": 1.6750902527075813,
+      "grad_norm": 0.059364162385463715,
+      "learning_rate": 0.001471604938271605,
+      "loss": 6.1037,
+      "step": 754
+    },
+    {
+      "epoch": 1.677311857817273,
+      "grad_norm": 0.04138586297631264,
+      "learning_rate": 0.0014691358024691357,
+      "loss": 5.9273,
+      "step": 755
+    },
+    {
+      "epoch": 1.6795334629269647,
+      "grad_norm": 0.0487748347222805,
+      "learning_rate": 0.0014666666666666667,
+      "loss": 5.9145,
+      "step": 756
+    },
+    {
+      "epoch": 1.6817550680366566,
+      "grad_norm": 0.036047063767910004,
+      "learning_rate": 0.0014641975308641974,
+      "loss": 6.3341,
+      "step": 757
+    },
+    {
+      "epoch": 1.6839766731463484,
+      "grad_norm": 0.05469675362110138,
+      "learning_rate": 0.0014617283950617286,
+      "loss": 6.0284,
+      "step": 758
+    },
+    {
+      "epoch": 1.68619827825604,
+      "grad_norm": 0.04500849172472954,
+      "learning_rate": 0.0014592592592592593,
+      "loss": 6.2335,
+      "step": 759
+    },
+    {
+      "epoch": 1.6884198833657318,
+      "grad_norm": 0.04404061287641525,
+      "learning_rate": 0.0014567901234567903,
+      "loss": 6.2097,
+      "step": 760
+    },
+    {
+      "epoch": 1.6906414884754235,
+      "grad_norm": 0.04719531536102295,
+      "learning_rate": 0.001454320987654321,
+      "loss": 5.7834,
+      "step": 761
+    },
+    {
+      "epoch": 1.6928630935851152,
+      "grad_norm": 0.04584319889545441,
+      "learning_rate": 0.001451851851851852,
+      "loss": 6.0973,
+      "step": 762
+    },
+    {
+      "epoch": 1.695084698694807,
+      "grad_norm": 0.046094540506601334,
+      "learning_rate": 0.0014493827160493827,
+      "loss": 5.8519,
+      "step": 763
+    },
+    {
+      "epoch": 1.6973063038044987,
+      "grad_norm": 0.04250643029808998,
+      "learning_rate": 0.0014469135802469136,
+      "loss": 6.2503,
+      "step": 764
+    },
+    {
+      "epoch": 1.6995279089141904,
+      "grad_norm": 0.0384419821202755,
+      "learning_rate": 0.0014444444444444444,
+      "loss": 6.0693,
+      "step": 765
+    },
+    {
+      "epoch": 1.7017495140238823,
+      "grad_norm": 0.050014935433864594,
+      "learning_rate": 0.0014419753086419753,
+      "loss": 6.383,
+      "step": 766
+    },
+    {
+      "epoch": 1.703971119133574,
+      "grad_norm": 0.045423299074172974,
+      "learning_rate": 0.0014395061728395063,
+      "loss": 5.9695,
+      "step": 767
+    },
+    {
+      "epoch": 1.7061927242432657,
+      "grad_norm": 0.052475545555353165,
+      "learning_rate": 0.001437037037037037,
+      "loss": 6.1188,
+      "step": 768
+    },
+    {
+      "epoch": 1.7084143293529577,
+      "grad_norm": 0.04825114831328392,
+      "learning_rate": 0.001434567901234568,
+      "loss": 5.7932,
+      "step": 769
+    },
+    {
+      "epoch": 1.7106359344626494,
+      "grad_norm": 0.041027557104825974,
+      "learning_rate": 0.0014320987654320987,
+      "loss": 6.0139,
+      "step": 770
+    },
+    {
+      "epoch": 1.712857539572341,
+      "grad_norm": 0.054573532193899155,
+      "learning_rate": 0.0014296296296296297,
+      "loss": 6.2518,
+      "step": 771
+    },
+    {
+      "epoch": 1.7150791446820328,
+      "grad_norm": 0.03849795088171959,
+      "learning_rate": 0.0014271604938271604,
+      "loss": 6.1488,
+      "step": 772
+    },
+    {
+      "epoch": 1.7173007497917245,
+      "grad_norm": 0.038570936769247055,
+      "learning_rate": 0.0014246913580246916,
+      "loss": 6.1516,
+      "step": 773
+    },
+    {
+      "epoch": 1.7195223549014163,
+      "grad_norm": 0.03841586410999298,
+      "learning_rate": 0.0014222222222222223,
+      "loss": 6.0528,
+      "step": 774
+    },
+    {
+      "epoch": 1.721743960011108,
+      "grad_norm": 0.06937354803085327,
+      "learning_rate": 0.0014197530864197532,
+      "loss": 6.3445,
+      "step": 775
+    },
+    {
+      "epoch": 1.7239655651207997,
+      "grad_norm": 0.04137590900063515,
+      "learning_rate": 0.001417283950617284,
+      "loss": 5.8847,
+      "step": 776
+    },
+    {
+      "epoch": 1.7261871702304914,
+      "grad_norm": 0.0639619529247284,
+      "learning_rate": 0.001414814814814815,
+      "loss": 6.2012,
+      "step": 777
+    },
+    {
+      "epoch": 1.7284087753401833,
+      "grad_norm": 0.045112211257219315,
+      "learning_rate": 0.0014123456790123457,
+      "loss": 6.0829,
+      "step": 778
+    },
+    {
+      "epoch": 1.730630380449875,
+      "grad_norm": 0.04505884647369385,
+      "learning_rate": 0.0014098765432098766,
+      "loss": 6.307,
+      "step": 779
+    },
+    {
+      "epoch": 1.7328519855595668,
+      "grad_norm": 0.04798652231693268,
+      "learning_rate": 0.0014074074074074073,
+      "loss": 5.7065,
+      "step": 780
+    },
+    {
+      "epoch": 1.7350735906692587,
+      "grad_norm": 0.040401238948106766,
+      "learning_rate": 0.0014049382716049383,
+      "loss": 6.0646,
+      "step": 781
+    },
+    {
+      "epoch": 1.7372951957789504,
+      "grad_norm": 0.043497681617736816,
+      "learning_rate": 0.0014024691358024693,
+      "loss": 6.0617,
+      "step": 782
+    },
+    {
+      "epoch": 1.7395168008886421,
+      "grad_norm": 0.04268964007496834,
+      "learning_rate": 0.0014,
+      "loss": 5.9729,
+      "step": 783
+    },
+    {
+      "epoch": 1.7417384059983338,
+      "grad_norm": 0.042066771537065506,
+      "learning_rate": 0.001397530864197531,
+      "loss": 5.9399,
+      "step": 784
+    },
+    {
+      "epoch": 1.7439600111080256,
+      "grad_norm": 0.04098956659436226,
+      "learning_rate": 0.0013950617283950617,
+      "loss": 6.1768,
+      "step": 785
+    },
+    {
+      "epoch": 1.7461816162177173,
+      "grad_norm": 0.055857788771390915,
+      "learning_rate": 0.0013925925925925926,
+      "loss": 5.9895,
+      "step": 786
+    },
+    {
+      "epoch": 1.748403221327409,
+      "grad_norm": 0.05863477289676666,
+      "learning_rate": 0.0013901234567901234,
+      "loss": 5.966,
+      "step": 787
+    },
+    {
+      "epoch": 1.7506248264371007,
+      "grad_norm": 0.06156482547521591,
+      "learning_rate": 0.0013876543209876545,
+      "loss": 5.9458,
+      "step": 788
+    },
+    {
+      "epoch": 1.7528464315467924,
+      "grad_norm": 0.03802463412284851,
+      "learning_rate": 0.0013851851851851853,
+      "loss": 6.0276,
+      "step": 789
+    },
+    {
+      "epoch": 1.7550680366564844,
+      "grad_norm": 0.04245764762163162,
+      "learning_rate": 0.0013827160493827162,
+      "loss": 6.0092,
+      "step": 790
+    },
+    {
+      "epoch": 1.757289641766176,
+      "grad_norm": 0.04828885570168495,
+      "learning_rate": 0.001380246913580247,
+      "loss": 6.1139,
+      "step": 791
+    },
+    {
+      "epoch": 1.7595112468758678,
+      "grad_norm": 0.04812488704919815,
+      "learning_rate": 0.001377777777777778,
+      "loss": 6.4833,
+      "step": 792
+    },
+    {
+      "epoch": 1.7617328519855595,
+      "grad_norm": 0.0436619408428669,
+      "learning_rate": 0.0013753086419753086,
+      "loss": 6.0711,
+      "step": 793
+    },
+    {
+      "epoch": 1.7639544570952514,
+      "grad_norm": 0.04308826103806496,
+      "learning_rate": 0.0013728395061728396,
+      "loss": 6.2644,
+      "step": 794
+    },
+    {
+      "epoch": 1.7661760622049432,
+      "grad_norm": 0.05595472455024719,
+      "learning_rate": 0.0013703703703703703,
+      "loss": 6.1844,
+      "step": 795
+    },
+    {
+      "epoch": 1.7683976673146349,
+      "grad_norm": 0.03139431029558182,
+      "learning_rate": 0.0013679012345679013,
+      "loss": 6.2593,
+      "step": 796
+    },
+    {
+      "epoch": 1.7706192724243266,
+      "grad_norm": 0.04515387490391731,
+      "learning_rate": 0.0013654320987654322,
+      "loss": 6.0786,
+      "step": 797
+    },
+    {
+      "epoch": 1.7728408775340183,
+      "grad_norm": 0.03661363571882248,
+      "learning_rate": 0.001362962962962963,
+      "loss": 6.3232,
+      "step": 798
+    },
+    {
+      "epoch": 1.77506248264371,
+      "grad_norm": 0.030566265806555748,
+      "learning_rate": 0.001360493827160494,
+      "loss": 6.0365,
+      "step": 799
+    },
+    {
+      "epoch": 1.7772840877534017,
+      "grad_norm": 0.0449000783264637,
+      "learning_rate": 0.0013580246913580246,
+      "loss": 6.3158,
+      "step": 800
+    },
+    {
+      "epoch": 1.7795056928630935,
+      "grad_norm": 0.04717319831252098,
+      "learning_rate": 0.0013555555555555556,
+      "loss": 6.0393,
+      "step": 801
+    },
+    {
+      "epoch": 1.7817272979727852,
+      "grad_norm": 0.06261464208364487,
+      "learning_rate": 0.0013530864197530863,
+      "loss": 6.0147,
+      "step": 802
+    },
+    {
+      "epoch": 1.783948903082477,
+      "grad_norm": 0.044654421508312225,
+      "learning_rate": 0.0013506172839506175,
+      "loss": 6.1854,
+      "step": 803
+    },
+    {
+      "epoch": 1.7861705081921688,
+      "grad_norm": 0.04214232787489891,
+      "learning_rate": 0.0013481481481481482,
+      "loss": 6.1411,
+      "step": 804
+    },
+    {
+      "epoch": 1.7883921133018605,
+      "grad_norm": 0.06957382708787918,
+      "learning_rate": 0.0013456790123456792,
+      "loss": 5.9172,
+      "step": 805
+    },
+    {
+      "epoch": 1.7906137184115525,
+      "grad_norm": 0.035519879311323166,
+      "learning_rate": 0.00134320987654321,
+      "loss": 6.243,
+      "step": 806
+    },
+    {
+      "epoch": 1.7928353235212442,
+      "grad_norm": 0.04761626198887825,
+      "learning_rate": 0.0013407407407407409,
+      "loss": 6.2272,
+      "step": 807
+    },
+    {
+      "epoch": 1.795056928630936,
+      "grad_norm": 0.06698905676603317,
+      "learning_rate": 0.0013382716049382716,
+      "loss": 6.0552,
+      "step": 808
+    },
+    {
+      "epoch": 1.7972785337406276,
+      "grad_norm": 0.03679412230849266,
+      "learning_rate": 0.0013358024691358023,
+      "loss": 6.0877,
+      "step": 809
+    },
+    {
+      "epoch": 1.7995001388503193,
+      "grad_norm": 0.04155512899160385,
+      "learning_rate": 0.0013333333333333333,
+      "loss": 6.0179,
+      "step": 810
+    },
+    {
+      "epoch": 1.801721743960011,
+      "grad_norm": 0.044784773141145706,
+      "learning_rate": 0.0013308641975308642,
+      "loss": 6.2279,
+      "step": 811
+    },
+    {
+      "epoch": 1.8039433490697028,
+      "grad_norm": 0.04515000805258751,
+      "learning_rate": 0.0013283950617283952,
+      "loss": 6.0754,
+      "step": 812
+    },
+    {
+      "epoch": 1.8061649541793945,
+      "grad_norm": 0.06366923451423645,
+      "learning_rate": 0.001325925925925926,
+      "loss": 6.102,
+      "step": 813
+    },
+    {
+      "epoch": 1.8083865592890862,
+      "grad_norm": 0.04164503514766693,
+      "learning_rate": 0.0013234567901234569,
+      "loss": 5.9215,
+      "step": 814
+    },
+    {
+      "epoch": 1.8106081643987781,
+      "grad_norm": 0.059181466698646545,
+      "learning_rate": 0.0013209876543209876,
+      "loss": 6.1557,
+      "step": 815
+    },
+    {
+      "epoch": 1.8128297695084699,
+      "grad_norm": 0.04095742478966713,
+      "learning_rate": 0.0013185185185185186,
+      "loss": 5.9996,
+      "step": 816
+    },
+    {
+      "epoch": 1.8150513746181616,
+      "grad_norm": 0.057590506970882416,
+      "learning_rate": 0.0013160493827160493,
+      "loss": 5.9219,
+      "step": 817
+    },
+    {
+      "epoch": 1.8172729797278535,
+      "grad_norm": 0.039277397096157074,
+      "learning_rate": 0.0013135802469135802,
+      "loss": 5.9991,
+      "step": 818
+    },
+    {
+      "epoch": 1.8194945848375452,
+      "grad_norm": 0.0515102855861187,
+      "learning_rate": 0.0013111111111111112,
+      "loss": 5.9543,
+      "step": 819
+    },
+    {
+      "epoch": 1.821716189947237,
+      "grad_norm": 0.06030356138944626,
+      "learning_rate": 0.001308641975308642,
+      "loss": 6.1968,
+      "step": 820
+    },
+    {
+      "epoch": 1.8239377950569287,
+      "grad_norm": 0.042879413813352585,
+      "learning_rate": 0.0013061728395061729,
+      "loss": 6.0359,
+      "step": 821
+    },
+    {
+      "epoch": 1.8261594001666204,
+      "grad_norm": 0.049642011523246765,
+      "learning_rate": 0.0013037037037037036,
+      "loss": 6.2051,
+      "step": 822
+    },
+    {
+      "epoch": 1.828381005276312,
+      "grad_norm": 0.04189665988087654,
+      "learning_rate": 0.0013012345679012346,
+      "loss": 6.1738,
+      "step": 823
+    },
+    {
+      "epoch": 1.8306026103860038,
+      "grad_norm": 0.03751196712255478,
+      "learning_rate": 0.0012987654320987653,
+      "loss": 6.135,
+      "step": 824
+    },
+    {
+      "epoch": 1.8328242154956955,
+      "grad_norm": 0.03689640015363693,
+      "learning_rate": 0.0012962962962962963,
+      "loss": 6.1697,
+      "step": 825
+    },
+    {
+      "epoch": 1.8350458206053872,
+      "grad_norm": 0.059797253459692,
+      "learning_rate": 0.0012938271604938272,
+      "loss": 5.9132,
+      "step": 826
+    },
+    {
+      "epoch": 1.8372674257150792,
+      "grad_norm": 0.054130177944898605,
+      "learning_rate": 0.0012913580246913582,
+      "loss": 6.1891,
+      "step": 827
+    },
+    {
+      "epoch": 1.8394890308247709,
+      "grad_norm": 0.044893085956573486,
+      "learning_rate": 0.001288888888888889,
+      "loss": 5.7673,
+      "step": 828
+    },
+    {
+      "epoch": 1.8417106359344626,
+      "grad_norm": 0.060131531208753586,
+      "learning_rate": 0.0012864197530864198,
+      "loss": 5.9341,
+      "step": 829
+    },
+    {
+      "epoch": 1.8439322410441545,
+      "grad_norm": 0.06052136793732643,
+      "learning_rate": 0.0012839506172839506,
+      "loss": 5.9462,
+      "step": 830
+    },
+    {
+      "epoch": 1.8461538461538463,
+      "grad_norm": 0.04839181900024414,
+      "learning_rate": 0.0012814814814814815,
+      "loss": 5.9377,
+      "step": 831
+    },
+    {
+      "epoch": 1.848375451263538,
+      "grad_norm": 0.04253244027495384,
+      "learning_rate": 0.0012790123456790123,
+      "loss": 5.9653,
+      "step": 832
+    },
+    {
+      "epoch": 1.8505970563732297,
+      "grad_norm": 0.07292971014976501,
+      "learning_rate": 0.0012765432098765432,
+      "loss": 5.8236,
+      "step": 833
+    },
+    {
+      "epoch": 1.8528186614829214,
+      "grad_norm": 0.04111470654606819,
+      "learning_rate": 0.0012740740740740742,
+      "loss": 6.0388,
+      "step": 834
+    },
+    {
+      "epoch": 1.8550402665926131,
+      "grad_norm": 0.04093720018863678,
+      "learning_rate": 0.001271604938271605,
+      "loss": 5.8761,
+      "step": 835
+    },
+    {
+      "epoch": 1.8572618717023048,
+      "grad_norm": 0.04653589800000191,
+      "learning_rate": 0.0012691358024691359,
+      "loss": 5.9278,
+      "step": 836
+    },
+    {
+      "epoch": 1.8594834768119965,
+      "grad_norm": 0.05398708954453468,
+      "learning_rate": 0.0012666666666666666,
+      "loss": 5.9163,
+      "step": 837
+    },
+    {
+      "epoch": 1.8617050819216883,
+      "grad_norm": 0.038633935153484344,
+      "learning_rate": 0.0012641975308641975,
+      "loss": 5.8916,
+      "step": 838
+    },
+    {
+      "epoch": 1.8639266870313802,
+      "grad_norm": 0.04764648526906967,
+      "learning_rate": 0.0012617283950617283,
+      "loss": 6.2452,
+      "step": 839
+    },
+    {
+      "epoch": 1.866148292141072,
+      "grad_norm": 0.052322935312986374,
+      "learning_rate": 0.0012592592592592592,
+      "loss": 5.9655,
+      "step": 840
+    },
+    {
+      "epoch": 1.8683698972507636,
+      "grad_norm": 0.03198683634400368,
+      "learning_rate": 0.0012567901234567902,
+      "loss": 6.2927,
+      "step": 841
+    },
+    {
+      "epoch": 1.8705915023604556,
+      "grad_norm": 0.04638659209012985,
+      "learning_rate": 0.0012543209876543211,
+      "loss": 5.9515,
+      "step": 842
+    },
+    {
+      "epoch": 1.8728131074701473,
+      "grad_norm": 0.03224872052669525,
+      "learning_rate": 0.0012518518518518519,
+      "loss": 6.0,
+      "step": 843
+    },
+    {
+      "epoch": 1.875034712579839,
+      "grad_norm": 0.04067224636673927,
+      "learning_rate": 0.0012493827160493828,
+      "loss": 5.9006,
+      "step": 844
+    },
+    {
+      "epoch": 1.8772563176895307,
+      "grad_norm": 0.03820433095097542,
+      "learning_rate": 0.0012469135802469136,
+      "loss": 5.9431,
+      "step": 845
+    },
+    {
+      "epoch": 1.8794779227992224,
+      "grad_norm": 0.04805198311805725,
+      "learning_rate": 0.0012444444444444445,
+      "loss": 6.2114,
+      "step": 846
+    },
+    {
+      "epoch": 1.8816995279089141,
+      "grad_norm": 0.060539450496435165,
+      "learning_rate": 0.0012419753086419752,
+      "loss": 6.051,
+      "step": 847
+    },
+    {
+      "epoch": 1.8839211330186059,
+      "grad_norm": 0.05590864643454552,
+      "learning_rate": 0.0012395061728395062,
+      "loss": 6.0311,
+      "step": 848
+    },
+    {
+      "epoch": 1.8861427381282976,
+      "grad_norm": 0.04005192592740059,
+      "learning_rate": 0.0012370370370370371,
+      "loss": 5.821,
+      "step": 849
+    },
+    {
+      "epoch": 1.8883643432379893,
+      "grad_norm": 0.04435478150844574,
+      "learning_rate": 0.0012345679012345679,
+      "loss": 6.0769,
+      "step": 850
+    },
+    {
+      "epoch": 1.8905859483476812,
+      "grad_norm": 0.05554912984371185,
+      "learning_rate": 0.0012320987654320988,
+      "loss": 6.1892,
+      "step": 851
+    },
+    {
+      "epoch": 1.892807553457373,
+      "grad_norm": 0.05200256034731865,
+      "learning_rate": 0.0012296296296296296,
+      "loss": 5.8477,
+      "step": 852
+    },
+    {
+      "epoch": 1.8950291585670647,
+      "grad_norm": 0.04454224184155464,
+      "learning_rate": 0.0012271604938271605,
+      "loss": 5.9329,
+      "step": 853
+    },
+    {
+      "epoch": 1.8972507636767566,
+      "grad_norm": 0.05036364868283272,
+      "learning_rate": 0.0012246913580246912,
+      "loss": 6.3992,
+      "step": 854
+    },
+    {
+      "epoch": 1.8994723687864483,
+      "grad_norm": 0.05385028198361397,
+      "learning_rate": 0.0012222222222222222,
+      "loss": 6.1398,
+      "step": 855
+    },
+    {
+      "epoch": 1.90169397389614,
+      "grad_norm": 0.04300863295793533,
+      "learning_rate": 0.0012197530864197532,
+      "loss": 5.894,
+      "step": 856
+    },
+    {
+      "epoch": 1.9039155790058317,
+      "grad_norm": 0.04779009521007538,
+      "learning_rate": 0.001217283950617284,
+      "loss": 5.9571,
+      "step": 857
+    },
+    {
+      "epoch": 1.9061371841155235,
+      "grad_norm": 0.03509556129574776,
+      "learning_rate": 0.0012148148148148148,
+      "loss": 6.0471,
+      "step": 858
+    },
+    {
+      "epoch": 1.9083587892252152,
+      "grad_norm": 0.044372253119945526,
+      "learning_rate": 0.0012123456790123458,
+      "loss": 5.7884,
+      "step": 859
+    },
+    {
+      "epoch": 1.910580394334907,
+      "grad_norm": 0.04641924053430557,
+      "learning_rate": 0.0012098765432098765,
+      "loss": 6.0532,
+      "step": 860
+    },
+    {
+      "epoch": 1.9128019994445986,
+      "grad_norm": 0.05756306275725365,
+      "learning_rate": 0.0012074074074074075,
+      "loss": 6.0295,
+      "step": 861
+    },
+    {
+      "epoch": 1.9150236045542903,
+      "grad_norm": 0.05073798447847366,
+      "learning_rate": 0.0012049382716049382,
+      "loss": 5.8082,
+      "step": 862
+    },
+    {
+      "epoch": 1.9172452096639823,
+      "grad_norm": 0.05037453770637512,
+      "learning_rate": 0.0012024691358024692,
+      "loss": 5.9952,
+      "step": 863
+    },
+    {
+      "epoch": 1.919466814773674,
+      "grad_norm": 0.047901567071676254,
+      "learning_rate": 0.0012000000000000001,
+      "loss": 6.0192,
+      "step": 864
+    },
+    {
+      "epoch": 1.9216884198833657,
+      "grad_norm": 0.05425766482949257,
+      "learning_rate": 0.0011975308641975308,
+      "loss": 5.9219,
+      "step": 865
+    },
+    {
+      "epoch": 1.9239100249930576,
+      "grad_norm": 0.05082131549715996,
+      "learning_rate": 0.0011950617283950618,
+      "loss": 5.8619,
+      "step": 866
+    },
+    {
+      "epoch": 1.9261316301027493,
+      "grad_norm": 0.04783160984516144,
+      "learning_rate": 0.0011925925925925925,
+      "loss": 6.1955,
+      "step": 867
+    },
+    {
+      "epoch": 1.928353235212441,
+      "grad_norm": 0.07691506296396255,
+      "learning_rate": 0.0011901234567901235,
+      "loss": 6.1067,
+      "step": 868
+    },
+    {
+      "epoch": 1.9305748403221328,
+      "grad_norm": 0.0328618660569191,
+      "learning_rate": 0.0011876543209876542,
+      "loss": 6.248,
+      "step": 869
+    },
+    {
+      "epoch": 1.9327964454318245,
+      "grad_norm": 0.0480203740298748,
+      "learning_rate": 0.0011851851851851852,
+      "loss": 5.823,
+      "step": 870
+    },
+    {
+      "epoch": 1.9350180505415162,
+      "grad_norm": 0.04856712743639946,
+      "learning_rate": 0.0011827160493827161,
+      "loss": 6.0644,
+      "step": 871
+    },
+    {
+      "epoch": 1.937239655651208,
+      "grad_norm": 0.06229954957962036,
+      "learning_rate": 0.001180246913580247,
+      "loss": 6.4556,
+      "step": 872
+    },
+    {
+      "epoch": 1.9394612607608996,
+      "grad_norm": 0.06782695651054382,
+      "learning_rate": 0.0011777777777777778,
+      "loss": 5.7211,
+      "step": 873
+    },
+    {
+      "epoch": 1.9416828658705914,
+      "grad_norm": 0.04558860510587692,
+      "learning_rate": 0.0011753086419753088,
+      "loss": 5.8684,
+      "step": 874
+    },
+    {
+      "epoch": 1.9439044709802833,
+      "grad_norm": 0.05876922979950905,
+      "learning_rate": 0.0011728395061728395,
+      "loss": 5.9356,
+      "step": 875
+    },
+    {
+      "epoch": 1.946126076089975,
+      "grad_norm": 0.05528077483177185,
+      "learning_rate": 0.0011703703703703704,
+      "loss": 6.09,
+      "step": 876
+    },
+    {
+      "epoch": 1.9483476811996667,
+      "grad_norm": 0.05866006761789322,
+      "learning_rate": 0.0011679012345679012,
+      "loss": 6.0989,
+      "step": 877
+    },
+    {
+      "epoch": 1.9505692863093587,
+      "grad_norm": 0.06259995698928833,
+      "learning_rate": 0.0011654320987654321,
+      "loss": 5.9589,
+      "step": 878
+    },
+    {
+      "epoch": 1.9527908914190504,
+      "grad_norm": 0.046732097864151,
+      "learning_rate": 0.001162962962962963,
+      "loss": 6.0792,
+      "step": 879
+    },
+    {
+      "epoch": 1.955012496528742,
+      "grad_norm": 0.04859071597456932,
+      "learning_rate": 0.0011604938271604938,
+      "loss": 5.9421,
+      "step": 880
+    },
+    {
+      "epoch": 1.9572341016384338,
+      "grad_norm": 0.07568682730197906,
+      "learning_rate": 0.0011580246913580248,
+      "loss": 6.0496,
+      "step": 881
+    },
+    {
+      "epoch": 1.9594557067481255,
+      "grad_norm": 0.07273954898118973,
+      "learning_rate": 0.0011555555555555555,
+      "loss": 6.1588,
+      "step": 882
+    },
+    {
+      "epoch": 1.9616773118578172,
+      "grad_norm": 0.048970118165016174,
+      "learning_rate": 0.0011530864197530865,
+      "loss": 5.9395,
+      "step": 883
+    },
+    {
+      "epoch": 1.963898916967509,
+      "grad_norm": 0.05090354010462761,
+      "learning_rate": 0.0011506172839506172,
+      "loss": 5.8694,
+      "step": 884
+    },
+    {
+      "epoch": 1.9661205220772007,
+      "grad_norm": 0.044794149696826935,
+      "learning_rate": 0.0011481481481481481,
+      "loss": 5.8916,
+      "step": 885
+    },
+    {
+      "epoch": 1.9683421271868924,
+      "grad_norm": 0.05274913087487221,
+      "learning_rate": 0.001145679012345679,
+      "loss": 6.1292,
+      "step": 886
+    },
+    {
+      "epoch": 1.9705637322965843,
+      "grad_norm": 0.05570974200963974,
+      "learning_rate": 0.00114320987654321,
+      "loss": 6.0561,
+      "step": 887
+    },
+    {
+      "epoch": 1.972785337406276,
+      "grad_norm": 0.059043627232313156,
+      "learning_rate": 0.0011407407407407408,
+      "loss": 6.1266,
+      "step": 888
+    },
+    {
+      "epoch": 1.9750069425159678,
+      "grad_norm": 0.051148053258657455,
+      "learning_rate": 0.0011382716049382717,
+      "loss": 6.0357,
+      "step": 889
+    },
+    {
+      "epoch": 1.9772285476256597,
+      "grad_norm": 0.08695517480373383,
+      "learning_rate": 0.0011358024691358025,
+      "loss": 5.7319,
+      "step": 890
+    },
+    {
+      "epoch": 1.9794501527353514,
+      "grad_norm": 0.05379338189959526,
+      "learning_rate": 0.0011333333333333334,
+      "loss": 5.6979,
+      "step": 891
+    },
+    {
+      "epoch": 1.9816717578450431,
+      "grad_norm": 0.044213637709617615,
+      "learning_rate": 0.0011308641975308641,
+      "loss": 6.0704,
+      "step": 892
+    },
+    {
+      "epoch": 1.9838933629547348,
+      "grad_norm": 0.04649046063423157,
+      "learning_rate": 0.001128395061728395,
+      "loss": 6.2296,
+      "step": 893
+    },
+    {
+      "epoch": 1.9861149680644266,
+      "grad_norm": 0.05491134151816368,
+      "learning_rate": 0.001125925925925926,
+      "loss": 6.1292,
+      "step": 894
+    },
+    {
+      "epoch": 1.9883365731741183,
+      "grad_norm": 0.07654581218957901,
+      "learning_rate": 0.0011234567901234568,
+      "loss": 6.1115,
+      "step": 895
+    },
+    {
+      "epoch": 1.99055817828381,
+      "grad_norm": 0.0780540183186531,
+      "learning_rate": 0.0011209876543209877,
+      "loss": 6.0279,
+      "step": 896
+    },
+    {
+      "epoch": 1.9927797833935017,
+      "grad_norm": 0.06326672434806824,
+      "learning_rate": 0.0011185185185185185,
+      "loss": 6.0289,
+      "step": 897
+    },
+    {
+      "epoch": 1.9950013885031934,
+      "grad_norm": 0.042030349373817444,
+      "learning_rate": 0.0011160493827160494,
+      "loss": 6.1248,
+      "step": 898
+    },
+    {
+      "epoch": 1.9972229936128854,
+      "grad_norm": 0.04940253868699074,
+      "learning_rate": 0.0011135802469135802,
+      "loss": 5.8414,
+      "step": 899
+    },
+    {
+      "epoch": 1.999444598722577,
+      "grad_norm": 0.04751197621226311,
+      "learning_rate": 0.0011111111111111111,
+      "loss": 5.9401,
+      "step": 900
+    },
+    {
+      "epoch": 2.001666203832269,
+      "grad_norm": 0.046355050057172775,
+      "learning_rate": 0.001108641975308642,
+      "loss": 6.3377,
+      "step": 901
+    },
+    {
+      "epoch": 2.0038878089419607,
+      "grad_norm": 0.05937094986438751,
+      "learning_rate": 0.001106172839506173,
+      "loss": 5.6215,
+      "step": 902
+    },
+    {
+      "epoch": 2.0061094140516524,
+      "grad_norm": 0.05452687293291092,
+      "learning_rate": 0.0011037037037037037,
+      "loss": 5.8461,
+      "step": 903
+    },
+    {
+      "epoch": 2.008331019161344,
+      "grad_norm": 0.058435868471860886,
+      "learning_rate": 0.0011012345679012347,
+      "loss": 6.185,
+      "step": 904
+    },
+    {
+      "epoch": 2.010552624271036,
+      "grad_norm": 0.053980663418769836,
+      "learning_rate": 0.0010987654320987654,
+      "loss": 5.6601,
+      "step": 905
+    },
+    {
+      "epoch": 2.0127742293807276,
+      "grad_norm": 0.06021531671285629,
+      "learning_rate": 0.0010962962962962964,
+      "loss": 6.1838,
+      "step": 906
+    },
+    {
+      "epoch": 2.0149958344904193,
+      "grad_norm": 0.06973876804113388,
+      "learning_rate": 0.0010938271604938271,
+      "loss": 6.1234,
+      "step": 907
+    },
+    {
+      "epoch": 2.017217439600111,
+      "grad_norm": 0.049507174640893936,
+      "learning_rate": 0.001091358024691358,
+      "loss": 6.0305,
+      "step": 908
+    },
+    {
+      "epoch": 2.0194390447098027,
+      "grad_norm": 0.04300854727625847,
+      "learning_rate": 0.001088888888888889,
+      "loss": 5.7729,
+      "step": 909
+    },
+    {
+      "epoch": 2.0216606498194944,
+      "grad_norm": 0.049711547791957855,
+      "learning_rate": 0.0010864197530864198,
+      "loss": 6.3389,
+      "step": 910
+    },
+    {
+      "epoch": 2.023882254929186,
+      "grad_norm": 0.061491284519433975,
+      "learning_rate": 0.0010839506172839507,
+      "loss": 6.0795,
+      "step": 911
+    },
+    {
+      "epoch": 2.026103860038878,
+      "grad_norm": 0.04442786052823067,
+      "learning_rate": 0.0010814814814814814,
+      "loss": 6.3198,
+      "step": 912
+    },
+    {
+      "epoch": 2.02832546514857,
+      "grad_norm": 0.05021671578288078,
+      "learning_rate": 0.0010790123456790124,
+      "loss": 6.0581,
+      "step": 913
+    },
+    {
+      "epoch": 2.0305470702582618,
+      "grad_norm": 0.06689216941595078,
+      "learning_rate": 0.0010765432098765431,
+      "loss": 5.8025,
+      "step": 914
+    },
+    {
+      "epoch": 2.0327686753679535,
+      "grad_norm": 0.03939792141318321,
+      "learning_rate": 0.001074074074074074,
+      "loss": 5.6586,
+      "step": 915
+    },
+    {
+      "epoch": 2.034990280477645,
+      "grad_norm": 0.035246171057224274,
+      "learning_rate": 0.0010716049382716048,
+      "loss": 5.8321,
+      "step": 916
+    },
+    {
+      "epoch": 2.037211885587337,
+      "grad_norm": 0.06485025584697723,
+      "learning_rate": 0.001069135802469136,
+      "loss": 6.0704,
+      "step": 917
+    },
+    {
+      "epoch": 2.0394334906970286,
+      "grad_norm": 0.04535793140530586,
+      "learning_rate": 0.0010666666666666667,
+      "loss": 6.3523,
+      "step": 918
+    },
+    {
+      "epoch": 2.0416550958067203,
+      "grad_norm": 0.03397076949477196,
+      "learning_rate": 0.0010641975308641977,
+      "loss": 5.8642,
+      "step": 919
+    },
+    {
+      "epoch": 2.043876700916412,
+      "grad_norm": 0.04708784073591232,
+      "learning_rate": 0.0010617283950617284,
+      "loss": 5.8981,
+      "step": 920
+    },
+    {
+      "epoch": 2.0460983060261038,
+      "grad_norm": 0.03657303377985954,
+      "learning_rate": 0.0010592592592592594,
+      "loss": 5.9779,
+      "step": 921
+    },
+    {
+      "epoch": 2.0483199111357955,
+      "grad_norm": 0.05296921357512474,
+      "learning_rate": 0.00105679012345679,
+      "loss": 5.8182,
+      "step": 922
+    },
+    {
+      "epoch": 2.050541516245487,
+      "grad_norm": 0.04714735597372055,
+      "learning_rate": 0.001054320987654321,
+      "loss": 5.9424,
+      "step": 923
+    },
+    {
+      "epoch": 2.052763121355179,
+      "grad_norm": 0.04999284818768501,
+      "learning_rate": 0.001051851851851852,
+      "loss": 6.2192,
+      "step": 924
+    },
+    {
+      "epoch": 2.054984726464871,
+      "grad_norm": 0.03968612477183342,
+      "learning_rate": 0.0010493827160493827,
+      "loss": 6.096,
+      "step": 925
+    },
+    {
+      "epoch": 2.057206331574563,
+      "grad_norm": 0.038185521960258484,
+      "learning_rate": 0.0010469135802469137,
+      "loss": 6.2274,
+      "step": 926
+    },
+    {
+      "epoch": 2.0594279366842545,
+      "grad_norm": 0.07575882226228714,
+      "learning_rate": 0.0010444444444444444,
+      "loss": 6.0759,
+      "step": 927
+    },
+    {
+      "epoch": 2.061649541793946,
+      "grad_norm": 0.05830669403076172,
+      "learning_rate": 0.0010419753086419754,
+      "loss": 5.8577,
+      "step": 928
+    },
+    {
+      "epoch": 2.063871146903638,
+      "grad_norm": 0.05522337928414345,
+      "learning_rate": 0.001039506172839506,
+      "loss": 5.9179,
+      "step": 929
+    },
+    {
+      "epoch": 2.0660927520133296,
+      "grad_norm": 0.04078824445605278,
+      "learning_rate": 0.001037037037037037,
+      "loss": 6.1919,
+      "step": 930
+    },
+    {
+      "epoch": 2.0683143571230214,
+      "grad_norm": 0.0531235933303833,
+      "learning_rate": 0.0010345679012345678,
+      "loss": 6.1473,
+      "step": 931
+    },
+    {
+      "epoch": 2.070535962232713,
+      "grad_norm": 0.06885974854230881,
+      "learning_rate": 0.001032098765432099,
+      "loss": 5.994,
+      "step": 932
+    },
+    {
+      "epoch": 2.072757567342405,
+      "grad_norm": 0.051051583141088486,
+      "learning_rate": 0.0010296296296296297,
+      "loss": 6.0154,
+      "step": 933
+    },
+    {
+      "epoch": 2.0749791724520965,
+      "grad_norm": 0.05040254071354866,
+      "learning_rate": 0.0010271604938271606,
+      "loss": 5.911,
+      "step": 934
+    },
+    {
+      "epoch": 2.0772007775617882,
+      "grad_norm": 0.048295751214027405,
+      "learning_rate": 0.0010246913580246914,
+      "loss": 6.2185,
+      "step": 935
+    },
+    {
+      "epoch": 2.07942238267148,
+      "grad_norm": 0.07192625105381012,
+      "learning_rate": 0.0010222222222222223,
+      "loss": 6.063,
+      "step": 936
+    },
+    {
+      "epoch": 2.081643987781172,
+      "grad_norm": 0.06737194210290909,
+      "learning_rate": 0.001019753086419753,
+      "loss": 5.9556,
+      "step": 937
+    },
+    {
+      "epoch": 2.083865592890864,
+      "grad_norm": 0.052287761121988297,
+      "learning_rate": 0.001017283950617284,
+      "loss": 5.891,
+      "step": 938
+    },
+    {
+      "epoch": 2.0860871980005555,
+      "grad_norm": 0.0601489432156086,
+      "learning_rate": 0.001014814814814815,
+      "loss": 5.9905,
+      "step": 939
+    },
+    {
+      "epoch": 2.0883088031102472,
+      "grad_norm": 0.04830304533243179,
+      "learning_rate": 0.0010123456790123457,
+      "loss": 5.9515,
+      "step": 940
+    },
+    {
+      "epoch": 2.090530408219939,
+      "grad_norm": 0.05534185469150543,
+      "learning_rate": 0.0010098765432098766,
+      "loss": 5.7164,
+      "step": 941
+    },
+    {
+      "epoch": 2.0927520133296307,
+      "grad_norm": 0.048778556287288666,
+      "learning_rate": 0.0010074074074074074,
+      "loss": 5.9678,
+      "step": 942
+    },
+    {
+      "epoch": 2.0949736184393224,
+      "grad_norm": 0.09851929545402527,
+      "learning_rate": 0.0010049382716049383,
+      "loss": 5.5509,
+      "step": 943
+    },
+    {
+      "epoch": 2.097195223549014,
+      "grad_norm": 0.05525709316134453,
+      "learning_rate": 0.001002469135802469,
+      "loss": 6.44,
+      "step": 944
+    },
+    {
+      "epoch": 2.099416828658706,
+      "grad_norm": 0.04130396246910095,
+      "learning_rate": 0.001,
+      "loss": 5.8139,
+      "step": 945
+    },
+    {
+      "epoch": 2.1016384337683975,
+      "grad_norm": 0.04005426913499832,
+      "learning_rate": 0.0009975308641975308,
+      "loss": 5.9008,
+      "step": 946
+    },
+    {
+      "epoch": 2.1038600388780893,
+      "grad_norm": 0.05195621773600578,
+      "learning_rate": 0.000995061728395062,
+      "loss": 5.9189,
+      "step": 947
+    },
+    {
+      "epoch": 2.106081643987781,
+      "grad_norm": 0.05769798904657364,
+      "learning_rate": 0.0009925925925925927,
+      "loss": 5.731,
+      "step": 948
+    },
+    {
+      "epoch": 2.108303249097473,
+      "grad_norm": 0.05508211627602577,
+      "learning_rate": 0.0009901234567901234,
+      "loss": 5.7188,
+      "step": 949
+    },
+    {
+      "epoch": 2.110524854207165,
+      "grad_norm": 0.044090636074543,
+      "learning_rate": 0.0009876543209876543,
+      "loss": 6.1145,
+      "step": 950
+    },
+    {
+      "epoch": 2.1127464593168566,
+      "grad_norm": 0.07029680907726288,
+      "learning_rate": 0.000985185185185185,
+      "loss": 6.0173,
+      "step": 951
+    },
+    {
+      "epoch": 2.1149680644265483,
+      "grad_norm": 0.051356032490730286,
+      "learning_rate": 0.000982716049382716,
+      "loss": 6.0828,
+      "step": 952
+    },
+    {
+      "epoch": 2.11718966953624,
+      "grad_norm": 0.06374996155500412,
+      "learning_rate": 0.0009802469135802468,
+      "loss": 5.9538,
+      "step": 953
+    },
+    {
+      "epoch": 2.1194112746459317,
+      "grad_norm": 0.05067945644259453,
+      "learning_rate": 0.000977777777777778,
+      "loss": 6.1191,
+      "step": 954
+    },
+    {
+      "epoch": 2.1216328797556234,
+      "grad_norm": 0.04271899163722992,
+      "learning_rate": 0.0009753086419753087,
+      "loss": 6.282,
+      "step": 955
+    },
+    {
+      "epoch": 2.123854484865315,
+      "grad_norm": 0.050401248037815094,
+      "learning_rate": 0.0009728395061728395,
+      "loss": 5.889,
+      "step": 956
+    },
+    {
+      "epoch": 2.126076089975007,
+      "grad_norm": 0.06142055243253708,
+      "learning_rate": 0.0009703703703703704,
+      "loss": 5.8915,
+      "step": 957
+    },
+    {
+      "epoch": 2.1282976950846986,
+      "grad_norm": 0.045352134853601456,
+      "learning_rate": 0.0009679012345679012,
+      "loss": 6.2175,
+      "step": 958
+    },
+    {
+      "epoch": 2.1305193001943903,
+      "grad_norm": 0.05124196782708168,
+      "learning_rate": 0.000965432098765432,
+      "loss": 6.108,
+      "step": 959
+    },
+    {
+      "epoch": 2.132740905304082,
+      "grad_norm": 0.07164105027914047,
+      "learning_rate": 0.0009629629629629629,
+      "loss": 5.7826,
+      "step": 960
+    },
+    {
+      "epoch": 2.134962510413774,
+      "grad_norm": 0.08650267869234085,
+      "learning_rate": 0.0009604938271604937,
+      "loss": 5.8555,
+      "step": 961
+    },
+    {
+      "epoch": 2.137184115523466,
+      "grad_norm": 0.09889843314886093,
+      "learning_rate": 0.0009580246913580248,
+      "loss": 5.9192,
+      "step": 962
+    },
+    {
+      "epoch": 2.1394057206331576,
+      "grad_norm": 0.085551917552948,
+      "learning_rate": 0.0009555555555555556,
+      "loss": 5.747,
+      "step": 963
+    },
+    {
+      "epoch": 2.1416273257428493,
+      "grad_norm": 0.0767153725028038,
+      "learning_rate": 0.0009530864197530865,
+      "loss": 6.2098,
+      "step": 964
+    },
+    {
+      "epoch": 2.143848930852541,
+      "grad_norm": 0.044550374150276184,
+      "learning_rate": 0.0009506172839506173,
+      "loss": 5.8646,
+      "step": 965
+    },
+    {
+      "epoch": 2.1460705359622327,
+      "grad_norm": 0.05971762165427208,
+      "learning_rate": 0.0009481481481481482,
+      "loss": 5.8602,
+      "step": 966
+    },
+    {
+      "epoch": 2.1482921410719245,
+      "grad_norm": 0.04591561481356621,
+      "learning_rate": 0.000945679012345679,
+      "loss": 5.8252,
+      "step": 967
+    },
+    {
+      "epoch": 2.150513746181616,
+      "grad_norm": 0.08381897956132889,
+      "learning_rate": 0.0009432098765432098,
+      "loss": 5.9186,
+      "step": 968
+    },
+    {
+      "epoch": 2.152735351291308,
+      "grad_norm": 0.0634213387966156,
+      "learning_rate": 0.0009407407407407408,
+      "loss": 5.8926,
+      "step": 969
+    },
+    {
+      "epoch": 2.1549569564009996,
+      "grad_norm": 0.05233628675341606,
+      "learning_rate": 0.0009382716049382716,
+      "loss": 5.6604,
+      "step": 970
+    },
+    {
+      "epoch": 2.1571785615106913,
+      "grad_norm": 0.05524004250764847,
+      "learning_rate": 0.0009358024691358025,
+      "loss": 6.064,
+      "step": 971
+    },
+    {
+      "epoch": 2.159400166620383,
+      "grad_norm": 0.047200750559568405,
+      "learning_rate": 0.0009333333333333333,
+      "loss": 5.7086,
+      "step": 972
+    },
+    {
+      "epoch": 2.1616217717300747,
+      "grad_norm": 0.044266339391469955,
+      "learning_rate": 0.0009308641975308642,
+      "loss": 5.8878,
+      "step": 973
+    },
+    {
+      "epoch": 2.163843376839767,
+      "grad_norm": 0.0752163678407669,
+      "learning_rate": 0.000928395061728395,
+      "loss": 5.9089,
+      "step": 974
+    },
+    {
+      "epoch": 2.1660649819494586,
+      "grad_norm": 0.049727752804756165,
+      "learning_rate": 0.0009259259259259259,
+      "loss": 6.0005,
+      "step": 975
+    },
+    {
+      "epoch": 2.1682865870591503,
+      "grad_norm": 0.05456957221031189,
+      "learning_rate": 0.0009234567901234567,
+      "loss": 6.0706,
+      "step": 976
+    },
+    {
+      "epoch": 2.170508192168842,
+      "grad_norm": 0.05987009406089783,
+      "learning_rate": 0.0009209876543209878,
+      "loss": 5.9255,
+      "step": 977
+    },
+    {
+      "epoch": 2.1727297972785338,
+      "grad_norm": 0.06586159765720367,
+      "learning_rate": 0.0009185185185185186,
+      "loss": 6.0354,
+      "step": 978
+    },
+    {
+      "epoch": 2.1749514023882255,
+      "grad_norm": 0.058533716946840286,
+      "learning_rate": 0.0009160493827160494,
+      "loss": 6.0168,
+      "step": 979
+    },
+    {
+      "epoch": 2.177173007497917,
+      "grad_norm": 0.0423620268702507,
+      "learning_rate": 0.0009135802469135803,
+      "loss": 5.6981,
+      "step": 980
+    },
+    {
+      "epoch": 2.179394612607609,
+      "grad_norm": 0.03982330858707428,
+      "learning_rate": 0.0009111111111111111,
+      "loss": 6.0324,
+      "step": 981
+    },
+    {
+      "epoch": 2.1816162177173006,
+      "grad_norm": 0.05875158682465553,
+      "learning_rate": 0.000908641975308642,
+      "loss": 5.7238,
+      "step": 982
+    },
+    {
+      "epoch": 2.1838378228269923,
+      "grad_norm": 0.05447810888290405,
+      "learning_rate": 0.0009061728395061728,
+      "loss": 5.9205,
+      "step": 983
+    },
+    {
+      "epoch": 2.186059427936684,
+      "grad_norm": 0.041009433567523956,
+      "learning_rate": 0.0009037037037037038,
+      "loss": 6.1676,
+      "step": 984
+    },
+    {
+      "epoch": 2.1882810330463762,
+      "grad_norm": 0.0804993212223053,
+      "learning_rate": 0.0009012345679012346,
+      "loss": 6.1122,
+      "step": 985
+    },
+    {
+      "epoch": 2.190502638156068,
+      "grad_norm": 0.04410282149910927,
+      "learning_rate": 0.0008987654320987655,
+      "loss": 5.8397,
+      "step": 986
+    },
+    {
+      "epoch": 2.1927242432657597,
+      "grad_norm": 0.047606438398361206,
+      "learning_rate": 0.0008962962962962963,
+      "loss": 6.1118,
+      "step": 987
+    },
+    {
+      "epoch": 2.1949458483754514,
+      "grad_norm": 0.04506830871105194,
+      "learning_rate": 0.0008938271604938271,
+      "loss": 5.7183,
+      "step": 988
+    },
+    {
+      "epoch": 2.197167453485143,
+      "grad_norm": 0.04143473878502846,
+      "learning_rate": 0.000891358024691358,
+      "loss": 5.5866,
+      "step": 989
+    },
+    {
+      "epoch": 2.199389058594835,
+      "grad_norm": 0.0455172173678875,
+      "learning_rate": 0.0008888888888888888,
+      "loss": 5.6809,
+      "step": 990
+    },
+    {
+      "epoch": 2.2016106637045265,
+      "grad_norm": 0.04900422319769859,
+      "learning_rate": 0.0008864197530864197,
+      "loss": 5.7769,
+      "step": 991
+    },
+    {
+      "epoch": 2.2038322688142182,
+      "grad_norm": 0.05145672708749771,
+      "learning_rate": 0.0008839506172839507,
+      "loss": 5.9521,
+      "step": 992
+    },
+    {
+      "epoch": 2.20605387392391,
+      "grad_norm": 0.0465843640267849,
+      "learning_rate": 0.0008814814814814816,
+      "loss": 5.9149,
+      "step": 993
+    },
+    {
+      "epoch": 2.2082754790336017,
+      "grad_norm": 0.07542437314987183,
+      "learning_rate": 0.0008790123456790124,
+      "loss": 6.0509,
+      "step": 994
+    },
+    {
+      "epoch": 2.2104970841432934,
+      "grad_norm": 0.05861465260386467,
+      "learning_rate": 0.0008765432098765433,
+      "loss": 5.7998,
+      "step": 995
+    },
+    {
+      "epoch": 2.212718689252985,
+      "grad_norm": 0.05698259547352791,
+      "learning_rate": 0.0008740740740740741,
+      "loss": 6.0667,
+      "step": 996
+    },
+    {
+      "epoch": 2.214940294362677,
+      "grad_norm": 0.054383549839258194,
+      "learning_rate": 0.0008716049382716049,
+      "loss": 5.8778,
+      "step": 997
+    },
+    {
+      "epoch": 2.217161899472369,
+      "grad_norm": 0.09017494320869446,
+      "learning_rate": 0.0008691358024691358,
+      "loss": 5.9201,
+      "step": 998
+    },
+    {
+      "epoch": 2.2193835045820607,
+      "grad_norm": 0.057505007833242416,
+      "learning_rate": 0.0008666666666666666,
+      "loss": 5.8489,
+      "step": 999
+    },
+    {
+      "epoch": 2.2216051096917524,
+      "grad_norm": 0.047291986644268036,
+      "learning_rate": 0.0008641975308641976,
+      "loss": 6.0737,
+      "step": 1000
+    },
+    {
+      "epoch": 2.223826714801444,
+      "grad_norm": 0.0538238063454628,
+      "learning_rate": 0.0008617283950617284,
+      "loss": 6.228,
+      "step": 1001
+    },
+    {
+      "epoch": 2.226048319911136,
+      "grad_norm": 0.04707043245434761,
+      "learning_rate": 0.0008592592592592593,
+      "loss": 6.1274,
+      "step": 1002
+    },
+    {
+      "epoch": 2.2282699250208275,
+      "grad_norm": 0.06756126880645752,
+      "learning_rate": 0.0008567901234567901,
+      "loss": 6.0493,
+      "step": 1003
+    },
+    {
+      "epoch": 2.2304915301305193,
+      "grad_norm": 0.05249028652906418,
+      "learning_rate": 0.000854320987654321,
+      "loss": 5.8115,
+      "step": 1004
+    },
+    {
+      "epoch": 2.232713135240211,
+      "grad_norm": 0.04003097116947174,
+      "learning_rate": 0.0008518518518518518,
+      "loss": 5.7495,
+      "step": 1005
+    },
+    {
+      "epoch": 2.2349347403499027,
+      "grad_norm": 0.04583067446947098,
+      "learning_rate": 0.0008493827160493826,
+      "loss": 6.0212,
+      "step": 1006
+    },
+    {
+      "epoch": 2.2371563454595944,
+      "grad_norm": 0.044572725892066956,
+      "learning_rate": 0.0008469135802469137,
+      "loss": 5.9738,
+      "step": 1007
+    },
+    {
+      "epoch": 2.239377950569286,
+      "grad_norm": 0.046149302273988724,
+      "learning_rate": 0.0008444444444444445,
+      "loss": 5.8659,
+      "step": 1008
+    },
+    {
+      "epoch": 2.2415995556789783,
+      "grad_norm": 0.06371072679758072,
+      "learning_rate": 0.0008419753086419754,
+      "loss": 5.7136,
+      "step": 1009
+    },
+    {
+      "epoch": 2.24382116078867,
+      "grad_norm": 0.06779153645038605,
+      "learning_rate": 0.0008395061728395062,
+      "loss": 5.7331,
+      "step": 1010
+    },
+    {
+      "epoch": 2.2460427658983617,
+      "grad_norm": 0.06367138028144836,
+      "learning_rate": 0.0008370370370370371,
+      "loss": 5.6489,
+      "step": 1011
+    },
+    {
+      "epoch": 2.2482643710080534,
+      "grad_norm": 0.05142154544591904,
+      "learning_rate": 0.0008345679012345679,
+      "loss": 5.8864,
+      "step": 1012
+    },
+    {
+      "epoch": 2.250485976117745,
+      "grad_norm": 0.057787828147411346,
+      "learning_rate": 0.0008320987654320988,
+      "loss": 5.9604,
+      "step": 1013
+    },
+    {
+      "epoch": 2.252707581227437,
+      "grad_norm": 0.0527784489095211,
+      "learning_rate": 0.0008296296296296296,
+      "loss": 5.7835,
+      "step": 1014
+    },
+    {
+      "epoch": 2.2549291863371286,
+      "grad_norm": 0.07328028976917267,
+      "learning_rate": 0.0008271604938271605,
+      "loss": 5.7806,
+      "step": 1015
+    },
+    {
+      "epoch": 2.2571507914468203,
+      "grad_norm": 0.04314279928803444,
+      "learning_rate": 0.0008246913580246914,
+      "loss": 6.1362,
+      "step": 1016
+    },
+    {
+      "epoch": 2.259372396556512,
+      "grad_norm": 0.05321420729160309,
+      "learning_rate": 0.0008222222222222222,
+      "loss": 5.876,
+      "step": 1017
+    },
+    {
+      "epoch": 2.2615940016662037,
+      "grad_norm": 0.051915451884269714,
+      "learning_rate": 0.0008197530864197531,
+      "loss": 5.952,
+      "step": 1018
+    },
+    {
+      "epoch": 2.2638156067758954,
+      "grad_norm": 0.04124360531568527,
+      "learning_rate": 0.0008172839506172839,
+      "loss": 5.9982,
+      "step": 1019
+    },
+    {
+      "epoch": 2.266037211885587,
+      "grad_norm": 0.05607706680893898,
+      "learning_rate": 0.0008148148148148148,
+      "loss": 5.6548,
+      "step": 1020
+    },
+    {
+      "epoch": 2.268258816995279,
+      "grad_norm": 0.04974445328116417,
+      "learning_rate": 0.0008123456790123456,
+      "loss": 5.9214,
+      "step": 1021
+    },
+    {
+      "epoch": 2.270480422104971,
+      "grad_norm": 0.054538171738386154,
+      "learning_rate": 0.0008098765432098767,
+      "loss": 5.8079,
+      "step": 1022
+    },
+    {
+      "epoch": 2.2727020272146627,
+      "grad_norm": 0.07355673611164093,
+      "learning_rate": 0.0008074074074074075,
+      "loss": 5.7347,
+      "step": 1023
+    },
+    {
+      "epoch": 2.2749236323243545,
+      "grad_norm": 0.04785652458667755,
+      "learning_rate": 0.0008049382716049384,
+      "loss": 5.8351,
+      "step": 1024
+    },
+    {
+      "epoch": 2.277145237434046,
+      "grad_norm": 0.0509219616651535,
+      "learning_rate": 0.0008024691358024692,
+      "loss": 5.9801,
+      "step": 1025
+    },
+    {
+      "epoch": 2.279366842543738,
+      "grad_norm": 0.06415324658155441,
+      "learning_rate": 0.0008,
+      "loss": 5.9359,
+      "step": 1026
+    },
+    {
+      "epoch": 2.2815884476534296,
+      "grad_norm": 0.05211460962891579,
+      "learning_rate": 0.0007975308641975309,
+      "loss": 5.9854,
+      "step": 1027
+    },
+    {
+      "epoch": 2.2838100527631213,
+      "grad_norm": 0.07868483662605286,
+      "learning_rate": 0.0007950617283950617,
+      "loss": 6.066,
+      "step": 1028
+    },
+    {
+      "epoch": 2.286031657872813,
+      "grad_norm": 0.056249335408210754,
+      "learning_rate": 0.0007925925925925926,
+      "loss": 5.9595,
+      "step": 1029
+    },
+    {
+      "epoch": 2.2882532629825048,
+      "grad_norm": 0.034571342170238495,
+      "learning_rate": 0.0007901234567901235,
+      "loss": 6.1133,
+      "step": 1030
+    },
+    {
+      "epoch": 2.2904748680921965,
+      "grad_norm": 0.06290554255247116,
+      "learning_rate": 0.0007876543209876544,
+      "loss": 6.0943,
+      "step": 1031
+    },
+    {
+      "epoch": 2.292696473201888,
+      "grad_norm": 0.05772720277309418,
+      "learning_rate": 0.0007851851851851852,
+      "loss": 5.7976,
+      "step": 1032
+    },
+    {
+      "epoch": 2.2949180783115803,
+      "grad_norm": 0.0646650418639183,
+      "learning_rate": 0.000782716049382716,
+      "loss": 6.0501,
+      "step": 1033
+    },
+    {
+      "epoch": 2.297139683421272,
+      "grad_norm": 0.06781128793954849,
+      "learning_rate": 0.0007802469135802469,
+      "loss": 6.1345,
+      "step": 1034
+    },
+    {
+      "epoch": 2.2993612885309638,
+      "grad_norm": 0.05525821074843407,
+      "learning_rate": 0.0007777777777777777,
+      "loss": 5.9788,
+      "step": 1035
+    },
+    {
+      "epoch": 2.3015828936406555,
+      "grad_norm": 0.0514153353869915,
+      "learning_rate": 0.0007753086419753086,
+      "loss": 5.6737,
+      "step": 1036
+    },
+    {
+      "epoch": 2.303804498750347,
+      "grad_norm": 0.07104574888944626,
+      "learning_rate": 0.0007728395061728396,
+      "loss": 5.8728,
+      "step": 1037
+    },
+    {
+      "epoch": 2.306026103860039,
+      "grad_norm": 0.07069091498851776,
+      "learning_rate": 0.0007703703703703705,
+      "loss": 5.9203,
+      "step": 1038
+    },
+    {
+      "epoch": 2.3082477089697306,
+      "grad_norm": 0.07263226062059402,
+      "learning_rate": 0.0007679012345679013,
+      "loss": 5.8229,
+      "step": 1039
+    },
+    {
+      "epoch": 2.3104693140794224,
+      "grad_norm": 0.04691462963819504,
+      "learning_rate": 0.0007654320987654322,
+      "loss": 6.0475,
+      "step": 1040
+    },
+    {
+      "epoch": 2.312690919189114,
+      "grad_norm": 0.049608971923589706,
+      "learning_rate": 0.000762962962962963,
+      "loss": 6.0244,
+      "step": 1041
+    },
+    {
+      "epoch": 2.314912524298806,
+      "grad_norm": 0.04370072856545448,
+      "learning_rate": 0.0007604938271604939,
+      "loss": 6.0077,
+      "step": 1042
+    },
+    {
+      "epoch": 2.3171341294084975,
+      "grad_norm": 0.07816038280725479,
+      "learning_rate": 0.0007580246913580247,
+      "loss": 5.7582,
+      "step": 1043
+    },
+    {
+      "epoch": 2.319355734518189,
+      "grad_norm": 0.04159904271364212,
+      "learning_rate": 0.0007555555555555555,
+      "loss": 5.9899,
+      "step": 1044
+    },
+    {
+      "epoch": 2.321577339627881,
+      "grad_norm": 0.06587567180395126,
+      "learning_rate": 0.0007530864197530865,
+      "loss": 6.055,
+      "step": 1045
+    },
+    {
+      "epoch": 2.3237989447375726,
+      "grad_norm": 0.049506593495607376,
+      "learning_rate": 0.0007506172839506173,
+      "loss": 6.0308,
+      "step": 1046
+    },
+    {
+      "epoch": 2.326020549847265,
+      "grad_norm": 0.05315249413251877,
+      "learning_rate": 0.0007481481481481482,
+      "loss": 5.6239,
+      "step": 1047
+    },
+    {
+      "epoch": 2.3282421549569565,
+      "grad_norm": 0.0802527591586113,
+      "learning_rate": 0.000745679012345679,
+      "loss": 5.8756,
+      "step": 1048
+    },
+    {
+      "epoch": 2.3304637600666482,
+      "grad_norm": 0.05673157796263695,
+      "learning_rate": 0.0007432098765432099,
+      "loss": 5.8657,
+      "step": 1049
+    },
+    {
+      "epoch": 2.33268536517634,
+      "grad_norm": 0.03893468156456947,
+      "learning_rate": 0.0007407407407407407,
+      "loss": 5.9952,
+      "step": 1050
+    },
+    {
+      "epoch": 2.3349069702860317,
+      "grad_norm": 0.0511779710650444,
+      "learning_rate": 0.0007382716049382717,
+      "loss": 5.493,
+      "step": 1051
+    },
+    {
+      "epoch": 2.3371285753957234,
+      "grad_norm": 0.038521114736795425,
+      "learning_rate": 0.0007358024691358025,
+      "loss": 5.7782,
+      "step": 1052
+    },
+    {
+      "epoch": 2.339350180505415,
+      "grad_norm": 0.08118283748626709,
+      "learning_rate": 0.0007333333333333333,
+      "loss": 5.9379,
+      "step": 1053
+    },
+    {
+      "epoch": 2.341571785615107,
+      "grad_norm": 0.04421069100499153,
+      "learning_rate": 0.0007308641975308643,
+      "loss": 5.9435,
+      "step": 1054
+    },
+    {
+      "epoch": 2.3437933907247985,
+      "grad_norm": 0.051811590790748596,
+      "learning_rate": 0.0007283950617283951,
+      "loss": 6.0192,
+      "step": 1055
+    },
+    {
+      "epoch": 2.3460149958344902,
+      "grad_norm": 0.0796356350183487,
+      "learning_rate": 0.000725925925925926,
+      "loss": 5.628,
+      "step": 1056
+    },
+    {
+      "epoch": 2.3482366009441824,
+      "grad_norm": 0.06779766827821732,
+      "learning_rate": 0.0007234567901234568,
+      "loss": 5.8107,
+      "step": 1057
+    },
+    {
+      "epoch": 2.350458206053874,
+      "grad_norm": 0.07066726684570312,
+      "learning_rate": 0.0007209876543209877,
+      "loss": 5.8353,
+      "step": 1058
+    },
+    {
+      "epoch": 2.352679811163566,
+      "grad_norm": 0.06709866970777512,
+      "learning_rate": 0.0007185185185185185,
+      "loss": 5.8774,
+      "step": 1059
+    },
+    {
+      "epoch": 2.3549014162732576,
+      "grad_norm": 0.061553098261356354,
+      "learning_rate": 0.0007160493827160494,
+      "loss": 5.5262,
+      "step": 1060
+    },
+    {
+      "epoch": 2.3571230213829493,
+      "grad_norm": 0.04372899979352951,
+      "learning_rate": 0.0007135802469135802,
+      "loss": 5.9394,
+      "step": 1061
+    },
+    {
+      "epoch": 2.359344626492641,
+      "grad_norm": 0.04984388127923012,
+      "learning_rate": 0.0007111111111111111,
+      "loss": 6.1919,
+      "step": 1062
+    },
+    {
+      "epoch": 2.3615662316023327,
+      "grad_norm": 0.05325157940387726,
+      "learning_rate": 0.000708641975308642,
+      "loss": 5.8512,
+      "step": 1063
+    },
+    {
+      "epoch": 2.3637878367120244,
+      "grad_norm": 0.06130823865532875,
+      "learning_rate": 0.0007061728395061728,
+      "loss": 5.8554,
+      "step": 1064
+    },
+    {
+      "epoch": 2.366009441821716,
+      "grad_norm": 0.043308887630701065,
+      "learning_rate": 0.0007037037037037037,
+      "loss": 5.8488,
+      "step": 1065
+    },
+    {
+      "epoch": 2.368231046931408,
+      "grad_norm": 0.0811159536242485,
+      "learning_rate": 0.0007012345679012346,
+      "loss": 5.781,
+      "step": 1066
+    },
+    {
+      "epoch": 2.3704526520410996,
+      "grad_norm": 0.056953951716423035,
+      "learning_rate": 0.0006987654320987655,
+      "loss": 5.7746,
+      "step": 1067
+    },
+    {
+      "epoch": 2.3726742571507913,
+      "grad_norm": 0.06400550156831741,
+      "learning_rate": 0.0006962962962962963,
+      "loss": 5.9784,
+      "step": 1068
+    },
+    {
+      "epoch": 2.374895862260483,
+      "grad_norm": 0.07541897892951965,
+      "learning_rate": 0.0006938271604938273,
+      "loss": 5.804,
+      "step": 1069
+    },
+    {
+      "epoch": 2.3771174673701747,
+      "grad_norm": 0.05397675186395645,
+      "learning_rate": 0.0006913580246913581,
+      "loss": 6.0465,
+      "step": 1070
+    },
+    {
+      "epoch": 2.379339072479867,
+      "grad_norm": 0.06133056432008743,
+      "learning_rate": 0.000688888888888889,
+      "loss": 5.7814,
+      "step": 1071
+    },
+    {
+      "epoch": 2.3815606775895586,
+      "grad_norm": 0.04709472507238388,
+      "learning_rate": 0.0006864197530864198,
+      "loss": 5.9838,
+      "step": 1072
+    },
+    {
+      "epoch": 2.3837822826992503,
+      "grad_norm": 0.07361142337322235,
+      "learning_rate": 0.0006839506172839506,
+      "loss": 5.9829,
+      "step": 1073
+    },
+    {
+      "epoch": 2.386003887808942,
+      "grad_norm": 0.05288068205118179,
+      "learning_rate": 0.0006814814814814815,
+      "loss": 5.4474,
+      "step": 1074
+    },
+    {
+      "epoch": 2.3882254929186337,
+      "grad_norm": 0.060563940554857254,
+      "learning_rate": 0.0006790123456790123,
+      "loss": 6.3124,
+      "step": 1075
+    },
+    {
+      "epoch": 2.3904470980283254,
+      "grad_norm": 0.04115691035985947,
+      "learning_rate": 0.0006765432098765432,
+      "loss": 5.8899,
+      "step": 1076
+    },
+    {
+      "epoch": 2.392668703138017,
+      "grad_norm": 0.07105467468500137,
+      "learning_rate": 0.0006740740740740741,
+      "loss": 5.896,
+      "step": 1077
+    },
+    {
+      "epoch": 2.394890308247709,
+      "grad_norm": 0.06776075065135956,
+      "learning_rate": 0.000671604938271605,
+      "loss": 5.9926,
+      "step": 1078
+    },
+    {
+      "epoch": 2.3971119133574006,
+      "grad_norm": 0.04716327413916588,
+      "learning_rate": 0.0006691358024691358,
+      "loss": 5.8439,
+      "step": 1079
+    },
+    {
+      "epoch": 2.3993335184670923,
+      "grad_norm": 0.04947780445218086,
+      "learning_rate": 0.0006666666666666666,
+      "loss": 5.8964,
+      "step": 1080
+    },
+    {
+      "epoch": 2.4015551235767845,
+      "grad_norm": 0.08443167805671692,
+      "learning_rate": 0.0006641975308641976,
+      "loss": 5.9658,
+      "step": 1081
+    },
+    {
+      "epoch": 2.403776728686476,
+      "grad_norm": 0.05708467215299606,
+      "learning_rate": 0.0006617283950617284,
+      "loss": 5.5061,
+      "step": 1082
+    },
+    {
+      "epoch": 2.405998333796168,
+      "grad_norm": 0.08224623650312424,
+      "learning_rate": 0.0006592592592592593,
+      "loss": 5.6903,
+      "step": 1083
+    },
+    {
+      "epoch": 2.4082199389058596,
+      "grad_norm": 0.059096187353134155,
+      "learning_rate": 0.0006567901234567901,
+      "loss": 5.8495,
+      "step": 1084
+    },
+    {
+      "epoch": 2.4104415440155513,
+      "grad_norm": 0.08303750306367874,
+      "learning_rate": 0.000654320987654321,
+      "loss": 5.9612,
+      "step": 1085
+    },
+    {
+      "epoch": 2.412663149125243,
+      "grad_norm": 0.06013647839426994,
+      "learning_rate": 0.0006518518518518518,
+      "loss": 5.8311,
+      "step": 1086
+    },
+    {
+      "epoch": 2.4148847542349348,
+      "grad_norm": 0.06758897751569748,
+      "learning_rate": 0.0006493827160493827,
+      "loss": 6.0291,
+      "step": 1087
+    },
+    {
+      "epoch": 2.4171063593446265,
+      "grad_norm": 0.06581225246191025,
+      "learning_rate": 0.0006469135802469136,
+      "loss": 6.1888,
+      "step": 1088
+    },
+    {
+      "epoch": 2.419327964454318,
+      "grad_norm": 0.04392261058092117,
+      "learning_rate": 0.0006444444444444444,
+      "loss": 5.7335,
+      "step": 1089
+    },
+    {
+      "epoch": 2.42154956956401,
+      "grad_norm": 0.05450277030467987,
+      "learning_rate": 0.0006419753086419753,
+      "loss": 5.7078,
+      "step": 1090
+    },
+    {
+      "epoch": 2.4237711746737016,
+      "grad_norm": 0.054721567779779434,
+      "learning_rate": 0.0006395061728395061,
+      "loss": 6.0187,
+      "step": 1091
+    },
+    {
+      "epoch": 2.4259927797833933,
+      "grad_norm": 0.06465310603380203,
+      "learning_rate": 0.0006370370370370371,
+      "loss": 6.0899,
+      "step": 1092
+    },
+    {
+      "epoch": 2.428214384893085,
+      "grad_norm": 0.04651114344596863,
+      "learning_rate": 0.0006345679012345679,
+      "loss": 5.844,
+      "step": 1093
+    },
+    {
+      "epoch": 2.4304359900027768,
+      "grad_norm": 0.07651801407337189,
+      "learning_rate": 0.0006320987654320988,
+      "loss": 5.6907,
+      "step": 1094
+    },
+    {
+      "epoch": 2.432657595112469,
+      "grad_norm": 0.04985710605978966,
+      "learning_rate": 0.0006296296296296296,
+      "loss": 5.9672,
+      "step": 1095
+    },
+    {
+      "epoch": 2.4348792002221606,
+      "grad_norm": 0.05078880116343498,
+      "learning_rate": 0.0006271604938271606,
+      "loss": 5.8864,
+      "step": 1096
+    },
+    {
+      "epoch": 2.4371008053318524,
+      "grad_norm": 0.05990299582481384,
+      "learning_rate": 0.0006246913580246914,
+      "loss": 6.0482,
+      "step": 1097
+    },
+    {
+      "epoch": 2.439322410441544,
+      "grad_norm": 0.06744808703660965,
+      "learning_rate": 0.0006222222222222223,
+      "loss": 5.7663,
+      "step": 1098
+    },
+    {
+      "epoch": 2.441544015551236,
+      "grad_norm": 0.06793643534183502,
+      "learning_rate": 0.0006197530864197531,
+      "loss": 6.1515,
+      "step": 1099
+    },
+    {
+      "epoch": 2.4437656206609275,
+      "grad_norm": 0.054759904742240906,
+      "learning_rate": 0.0006172839506172839,
+      "loss": 5.8285,
+      "step": 1100
+    },
+    {
+      "epoch": 2.4459872257706192,
+      "grad_norm": 0.052103910595178604,
+      "learning_rate": 0.0006148148148148148,
+      "loss": 5.8897,
+      "step": 1101
+    },
+    {
+      "epoch": 2.448208830880311,
+      "grad_norm": 0.05217638239264488,
+      "learning_rate": 0.0006123456790123456,
+      "loss": 6.143,
+      "step": 1102
+    },
+    {
+      "epoch": 2.4504304359900027,
+      "grad_norm": 0.07950019836425781,
+      "learning_rate": 0.0006098765432098766,
+      "loss": 5.947,
+      "step": 1103
+    },
+    {
+      "epoch": 2.4526520410996944,
+      "grad_norm": 0.04472443833947182,
+      "learning_rate": 0.0006074074074074074,
+      "loss": 5.876,
+      "step": 1104
+    },
+    {
+      "epoch": 2.4548736462093865,
+      "grad_norm": 0.04999097064137459,
+      "learning_rate": 0.0006049382716049383,
+      "loss": 5.8068,
+      "step": 1105
+    },
+    {
+      "epoch": 2.4570952513190782,
+      "grad_norm": 0.043887995183467865,
+      "learning_rate": 0.0006024691358024691,
+      "loss": 5.6662,
+      "step": 1106
+    },
+    {
+      "epoch": 2.45931685642877,
+      "grad_norm": 0.053057532757520676,
+      "learning_rate": 0.0006000000000000001,
+      "loss": 5.7395,
+      "step": 1107
+    },
+    {
+      "epoch": 2.4615384615384617,
+      "grad_norm": 0.09846268594264984,
+      "learning_rate": 0.0005975308641975309,
+      "loss": 6.1488,
+      "step": 1108
+    },
+    {
+      "epoch": 2.4637600666481534,
+      "grad_norm": 0.058038853108882904,
+      "learning_rate": 0.0005950617283950617,
+      "loss": 5.6383,
+      "step": 1109
+    },
+    {
+      "epoch": 2.465981671757845,
+      "grad_norm": 0.038222573697566986,
+      "learning_rate": 0.0005925925925925926,
+      "loss": 5.9628,
+      "step": 1110
+    },
+    {
+      "epoch": 2.468203276867537,
+      "grad_norm": 0.06453481316566467,
+      "learning_rate": 0.0005901234567901235,
+      "loss": 5.7804,
+      "step": 1111
+    },
+    {
+      "epoch": 2.4704248819772285,
+      "grad_norm": 0.05904708802700043,
+      "learning_rate": 0.0005876543209876544,
+      "loss": 5.9227,
+      "step": 1112
+    },
+    {
+      "epoch": 2.4726464870869203,
+      "grad_norm": 0.04702883958816528,
+      "learning_rate": 0.0005851851851851852,
+      "loss": 5.8564,
+      "step": 1113
+    },
+    {
+      "epoch": 2.474868092196612,
+      "grad_norm": 0.07017482072114944,
+      "learning_rate": 0.0005827160493827161,
+      "loss": 5.9593,
+      "step": 1114
+    },
+    {
+      "epoch": 2.4770896973063037,
+      "grad_norm": 0.05562103912234306,
+      "learning_rate": 0.0005802469135802469,
+      "loss": 5.6843,
+      "step": 1115
+    },
+    {
+      "epoch": 2.4793113024159954,
+      "grad_norm": 0.0637110024690628,
+      "learning_rate": 0.0005777777777777778,
+      "loss": 5.7582,
+      "step": 1116
+    },
+    {
+      "epoch": 2.481532907525687,
+      "grad_norm": 0.05489303171634674,
+      "learning_rate": 0.0005753086419753086,
+      "loss": 5.6139,
+      "step": 1117
+    },
+    {
+      "epoch": 2.483754512635379,
+      "grad_norm": 0.05454476550221443,
+      "learning_rate": 0.0005728395061728395,
+      "loss": 5.7242,
+      "step": 1118
+    },
+    {
+      "epoch": 2.485976117745071,
+      "grad_norm": 0.04765939712524414,
+      "learning_rate": 0.0005703703703703704,
+      "loss": 6.0174,
+      "step": 1119
+    },
+    {
+      "epoch": 2.4881977228547627,
+      "grad_norm": 0.04585859924554825,
+      "learning_rate": 0.0005679012345679012,
+      "loss": 5.9768,
+      "step": 1120
+    },
+    {
+      "epoch": 2.4904193279644544,
+      "grad_norm": 0.05365516245365143,
+      "learning_rate": 0.0005654320987654321,
+      "loss": 6.0696,
+      "step": 1121
+    },
+    {
+      "epoch": 2.492640933074146,
+      "grad_norm": 0.04974067583680153,
+      "learning_rate": 0.000562962962962963,
+      "loss": 5.9656,
+      "step": 1122
+    },
+    {
+      "epoch": 2.494862538183838,
+      "grad_norm": 0.056851621717214584,
+      "learning_rate": 0.0005604938271604939,
+      "loss": 6.3004,
+      "step": 1123
+    },
+    {
+      "epoch": 2.4970841432935296,
+      "grad_norm": 0.055615637451410294,
+      "learning_rate": 0.0005580246913580247,
+      "loss": 5.9742,
+      "step": 1124
+    },
+    {
+      "epoch": 2.4993057484032213,
+      "grad_norm": 0.05103308707475662,
+      "learning_rate": 0.0005555555555555556,
+      "loss": 5.7587,
+      "step": 1125
+    },
+    {
+      "epoch": 2.501527353512913,
+      "grad_norm": 0.052293259650468826,
+      "learning_rate": 0.0005530864197530865,
+      "loss": 5.8624,
+      "step": 1126
+    },
+    {
+      "epoch": 2.5037489586226047,
+      "grad_norm": 0.05106515437364578,
+      "learning_rate": 0.0005506172839506173,
+      "loss": 5.8215,
+      "step": 1127
+    },
+    {
+      "epoch": 2.5059705637322964,
+      "grad_norm": 0.05350141227245331,
+      "learning_rate": 0.0005481481481481482,
+      "loss": 5.491,
+      "step": 1128
+    },
+    {
+      "epoch": 2.5081921688419886,
+      "grad_norm": 0.048063699156045914,
+      "learning_rate": 0.000545679012345679,
+      "loss": 5.618,
+      "step": 1129
+    },
+    {
+      "epoch": 2.5104137739516803,
+      "grad_norm": 0.041624557226896286,
+      "learning_rate": 0.0005432098765432099,
+      "loss": 5.872,
+      "step": 1130
+    },
+    {
+      "epoch": 2.512635379061372,
+      "grad_norm": 0.05397861450910568,
+      "learning_rate": 0.0005407407407407407,
+      "loss": 5.5568,
+      "step": 1131
+    },
+    {
+      "epoch": 2.5148569841710637,
+      "grad_norm": 0.04723358154296875,
+      "learning_rate": 0.0005382716049382716,
+      "loss": 5.7521,
+      "step": 1132
+    },
+    {
+      "epoch": 2.5170785892807555,
+      "grad_norm": 0.04932169243693352,
+      "learning_rate": 0.0005358024691358024,
+      "loss": 5.6926,
+      "step": 1133
+    },
+    {
+      "epoch": 2.519300194390447,
+      "grad_norm": 0.049961794167757034,
+      "learning_rate": 0.0005333333333333334,
+      "loss": 5.8031,
+      "step": 1134
+    },
+    {
+      "epoch": 2.521521799500139,
+      "grad_norm": 0.06546290963888168,
+      "learning_rate": 0.0005308641975308642,
+      "loss": 5.948,
+      "step": 1135
+    },
+    {
+      "epoch": 2.5237434046098306,
+      "grad_norm": 0.05172509700059891,
+      "learning_rate": 0.000528395061728395,
+      "loss": 5.9811,
+      "step": 1136
+    },
+    {
+      "epoch": 2.5259650097195223,
+      "grad_norm": 0.05558503046631813,
+      "learning_rate": 0.000525925925925926,
+      "loss": 5.5325,
+      "step": 1137
+    },
+    {
+      "epoch": 2.528186614829214,
+      "grad_norm": 0.04770063981413841,
+      "learning_rate": 0.0005234567901234568,
+      "loss": 6.0555,
+      "step": 1138
+    },
+    {
+      "epoch": 2.5304082199389057,
+      "grad_norm": 0.05725065991282463,
+      "learning_rate": 0.0005209876543209877,
+      "loss": 6.4155,
+      "step": 1139
+    },
+    {
+      "epoch": 2.5326298250485975,
+      "grad_norm": 0.056243401020765305,
+      "learning_rate": 0.0005185185185185185,
+      "loss": 5.6629,
+      "step": 1140
+    },
+    {
+      "epoch": 2.534851430158289,
+      "grad_norm": 0.05091274157166481,
+      "learning_rate": 0.0005160493827160495,
+      "loss": 6.0942,
+      "step": 1141
+    },
+    {
+      "epoch": 2.537073035267981,
+      "grad_norm": 0.08415589481592178,
+      "learning_rate": 0.0005135802469135803,
+      "loss": 6.0429,
+      "step": 1142
+    },
+    {
+      "epoch": 2.5392946403776726,
+      "grad_norm": 0.05046810582280159,
+      "learning_rate": 0.0005111111111111112,
+      "loss": 6.0881,
+      "step": 1143
+    },
+    {
+      "epoch": 2.5415162454873648,
+      "grad_norm": 0.05902523174881935,
+      "learning_rate": 0.000508641975308642,
+      "loss": 5.7132,
+      "step": 1144
+    },
+    {
+      "epoch": 2.5437378505970565,
+      "grad_norm": 0.057459522038698196,
+      "learning_rate": 0.0005061728395061728,
+      "loss": 5.9,
+      "step": 1145
+    },
+    {
+      "epoch": 2.545959455706748,
+      "grad_norm": 0.052608080208301544,
+      "learning_rate": 0.0005037037037037037,
+      "loss": 6.0521,
+      "step": 1146
+    },
+    {
+      "epoch": 2.54818106081644,
+      "grad_norm": 0.03979867696762085,
+      "learning_rate": 0.0005012345679012345,
+      "loss": 5.5871,
+      "step": 1147
+    },
+    {
+      "epoch": 2.5504026659261316,
+      "grad_norm": 0.04679068177938461,
+      "learning_rate": 0.0004987654320987654,
+      "loss": 5.9721,
+      "step": 1148
+    },
+    {
+      "epoch": 2.5526242710358233,
+      "grad_norm": 0.07402826845645905,
+      "learning_rate": 0.0004962962962962963,
+      "loss": 5.858,
+      "step": 1149
+    },
+    {
+      "epoch": 2.554845876145515,
+      "grad_norm": 0.054051779210567474,
+      "learning_rate": 0.0004938271604938272,
+      "loss": 5.8653,
+      "step": 1150
+    },
+    {
+      "epoch": 2.5570674812552068,
+      "grad_norm": 0.06473316252231598,
+      "learning_rate": 0.000491358024691358,
+      "loss": 5.7465,
+      "step": 1151
+    },
+    {
+      "epoch": 2.5592890863648985,
+      "grad_norm": 0.04736412316560745,
+      "learning_rate": 0.000488888888888889,
+      "loss": 5.8293,
+      "step": 1152
+    },
+    {
+      "epoch": 2.5615106914745907,
+      "grad_norm": 0.04081957787275314,
+      "learning_rate": 0.00048641975308641976,
+      "loss": 5.8449,
+      "step": 1153
+    },
+    {
+      "epoch": 2.5637322965842824,
+      "grad_norm": 0.04959014803171158,
+      "learning_rate": 0.0004839506172839506,
+      "loss": 5.729,
+      "step": 1154
+    },
+    {
+      "epoch": 2.565953901693974,
+      "grad_norm": 0.09538524597883224,
+      "learning_rate": 0.00048148148148148144,
+      "loss": 5.7509,
+      "step": 1155
+    },
+    {
+      "epoch": 2.568175506803666,
+      "grad_norm": 0.05531400814652443,
+      "learning_rate": 0.0004790123456790124,
+      "loss": 5.8802,
+      "step": 1156
+    },
+    {
+      "epoch": 2.5703971119133575,
+      "grad_norm": 0.06068155914545059,
+      "learning_rate": 0.00047654320987654324,
+      "loss": 6.134,
+      "step": 1157
+    },
+    {
+      "epoch": 2.5726187170230492,
+      "grad_norm": 0.05842263251543045,
+      "learning_rate": 0.0004740740740740741,
+      "loss": 5.8351,
+      "step": 1158
+    },
+    {
+      "epoch": 2.574840322132741,
+      "grad_norm": 0.06507426500320435,
+      "learning_rate": 0.0004716049382716049,
+      "loss": 5.5752,
+      "step": 1159
+    },
+    {
+      "epoch": 2.5770619272424327,
+      "grad_norm": 0.04701032117009163,
+      "learning_rate": 0.0004691358024691358,
+      "loss": 6.0044,
+      "step": 1160
+    },
+    {
+      "epoch": 2.5792835323521244,
+      "grad_norm": 0.04070613533258438,
+      "learning_rate": 0.00046666666666666666,
+      "loss": 6.0465,
+      "step": 1161
+    },
+    {
+      "epoch": 2.581505137461816,
+      "grad_norm": 0.04922967404127121,
+      "learning_rate": 0.0004641975308641975,
+      "loss": 5.9557,
+      "step": 1162
+    },
+    {
+      "epoch": 2.583726742571508,
+      "grad_norm": 0.055332690477371216,
+      "learning_rate": 0.00046172839506172835,
+      "loss": 5.9638,
+      "step": 1163
+    },
+    {
+      "epoch": 2.5859483476811995,
+      "grad_norm": 0.054886020720005035,
+      "learning_rate": 0.0004592592592592593,
+      "loss": 6.1071,
+      "step": 1164
+    },
+    {
+      "epoch": 2.5881699527908912,
+      "grad_norm": 0.0459747239947319,
+      "learning_rate": 0.00045679012345679014,
+      "loss": 5.9422,
+      "step": 1165
+    },
+    {
+      "epoch": 2.590391557900583,
+      "grad_norm": 0.0674658864736557,
+      "learning_rate": 0.000454320987654321,
+      "loss": 5.9015,
+      "step": 1166
+    },
+    {
+      "epoch": 2.5926131630102747,
+      "grad_norm": 0.06457730382680893,
+      "learning_rate": 0.0004518518518518519,
+      "loss": 6.177,
+      "step": 1167
+    },
+    {
+      "epoch": 2.594834768119967,
+      "grad_norm": 0.05291493982076645,
+      "learning_rate": 0.0004493827160493827,
+      "loss": 6.1573,
+      "step": 1168
+    },
+    {
+      "epoch": 2.5970563732296585,
+      "grad_norm": 0.0366109199821949,
+      "learning_rate": 0.00044691358024691357,
+      "loss": 5.9545,
+      "step": 1169
+    },
+    {
+      "epoch": 2.5992779783393503,
+      "grad_norm": 0.07245033979415894,
+      "learning_rate": 0.0004444444444444444,
+      "loss": 5.7638,
+      "step": 1170
+    },
+    {
+      "epoch": 2.601499583449042,
+      "grad_norm": 0.06531214714050293,
+      "learning_rate": 0.00044197530864197536,
+      "loss": 5.6332,
+      "step": 1171
+    },
+    {
+      "epoch": 2.6037211885587337,
+      "grad_norm": 0.05225894972681999,
+      "learning_rate": 0.0004395061728395062,
+      "loss": 6.0085,
+      "step": 1172
+    },
+    {
+      "epoch": 2.6059427936684254,
+      "grad_norm": 0.0548102930188179,
+      "learning_rate": 0.00043703703703703705,
+      "loss": 5.6063,
+      "step": 1173
+    },
+    {
+      "epoch": 2.608164398778117,
+      "grad_norm": 0.05640947446227074,
+      "learning_rate": 0.0004345679012345679,
+      "loss": 5.7361,
+      "step": 1174
+    },
+    {
+      "epoch": 2.610386003887809,
+      "grad_norm": 0.06905799359083176,
+      "learning_rate": 0.0004320987654320988,
+      "loss": 5.7613,
+      "step": 1175
+    },
+    {
+      "epoch": 2.6126076089975006,
+      "grad_norm": 0.04869680106639862,
+      "learning_rate": 0.00042962962962962963,
+      "loss": 5.8906,
+      "step": 1176
+    },
+    {
+      "epoch": 2.6148292141071927,
+      "grad_norm": 0.04864457622170448,
+      "learning_rate": 0.0004271604938271605,
+      "loss": 5.3902,
+      "step": 1177
+    },
+    {
+      "epoch": 2.6170508192168844,
+      "grad_norm": 0.05964788794517517,
+      "learning_rate": 0.0004246913580246913,
+      "loss": 5.5457,
+      "step": 1178
+    },
+    {
+      "epoch": 2.619272424326576,
+      "grad_norm": 0.04340321198105812,
+      "learning_rate": 0.00042222222222222227,
+      "loss": 5.9072,
+      "step": 1179
+    },
+    {
+      "epoch": 2.621494029436268,
+      "grad_norm": 0.04552336782217026,
+      "learning_rate": 0.0004197530864197531,
+      "loss": 6.0247,
+      "step": 1180
+    },
+    {
+      "epoch": 2.6237156345459596,
+      "grad_norm": 0.04501795768737793,
+      "learning_rate": 0.00041728395061728396,
+      "loss": 5.7657,
+      "step": 1181
+    },
+    {
+      "epoch": 2.6259372396556513,
+      "grad_norm": 0.09421583265066147,
+      "learning_rate": 0.0004148148148148148,
+      "loss": 6.128,
+      "step": 1182
+    },
+    {
+      "epoch": 2.628158844765343,
+      "grad_norm": 0.04200819879770279,
+      "learning_rate": 0.0004123456790123457,
+      "loss": 6.1594,
+      "step": 1183
+    },
+    {
+      "epoch": 2.6303804498750347,
+      "grad_norm": 0.05063186213374138,
+      "learning_rate": 0.00040987654320987654,
+      "loss": 5.5544,
+      "step": 1184
+    },
+    {
+      "epoch": 2.6326020549847264,
+      "grad_norm": 0.04634084552526474,
+      "learning_rate": 0.0004074074074074074,
+      "loss": 5.7918,
+      "step": 1185
+    },
+    {
+      "epoch": 2.634823660094418,
+      "grad_norm": 0.05453971028327942,
+      "learning_rate": 0.00040493827160493833,
+      "loss": 5.9909,
+      "step": 1186
+    },
+    {
+      "epoch": 2.63704526520411,
+      "grad_norm": 0.0714196115732193,
+      "learning_rate": 0.0004024691358024692,
+      "loss": 5.9813,
+      "step": 1187
+    },
+    {
+      "epoch": 2.6392668703138016,
+      "grad_norm": 0.04862188175320625,
+      "learning_rate": 0.0004,
+      "loss": 5.8212,
+      "step": 1188
+    },
+    {
+      "epoch": 2.6414884754234933,
+      "grad_norm": 0.1108190268278122,
+      "learning_rate": 0.00039753086419753086,
+      "loss": 5.4844,
+      "step": 1189
+    },
+    {
+      "epoch": 2.643710080533185,
+      "grad_norm": 0.05084674805402756,
+      "learning_rate": 0.00039506172839506176,
+      "loss": 5.8944,
+      "step": 1190
+    },
+    {
+      "epoch": 2.6459316856428767,
+      "grad_norm": 0.04934384301304817,
+      "learning_rate": 0.0003925925925925926,
+      "loss": 6.0849,
+      "step": 1191
+    },
+    {
+      "epoch": 2.648153290752569,
+      "grad_norm": 0.06427028030157089,
+      "learning_rate": 0.00039012345679012345,
+      "loss": 6.0675,
+      "step": 1192
+    },
+    {
+      "epoch": 2.6503748958622606,
+      "grad_norm": 0.05401918292045593,
+      "learning_rate": 0.0003876543209876543,
+      "loss": 6.1425,
+      "step": 1193
+    },
+    {
+      "epoch": 2.6525965009719523,
+      "grad_norm": 0.08267608284950256,
+      "learning_rate": 0.00038518518518518524,
+      "loss": 5.8855,
+      "step": 1194
+    },
+    {
+      "epoch": 2.654818106081644,
+      "grad_norm": 0.05584852769970894,
+      "learning_rate": 0.0003827160493827161,
+      "loss": 5.8688,
+      "step": 1195
+    },
+    {
+      "epoch": 2.6570397111913358,
+      "grad_norm": 0.06895650923252106,
+      "learning_rate": 0.0003802469135802469,
+      "loss": 6.1148,
+      "step": 1196
+    },
+    {
+      "epoch": 2.6592613163010275,
+      "grad_norm": 0.08543485403060913,
+      "learning_rate": 0.00037777777777777777,
+      "loss": 5.877,
+      "step": 1197
+    },
+    {
+      "epoch": 2.661482921410719,
+      "grad_norm": 0.08173643052577972,
+      "learning_rate": 0.00037530864197530867,
+      "loss": 5.9988,
+      "step": 1198
+    },
+    {
+      "epoch": 2.663704526520411,
+      "grad_norm": 0.05837205797433853,
+      "learning_rate": 0.0003728395061728395,
+      "loss": 5.8937,
+      "step": 1199
+    },
+    {
+      "epoch": 2.6659261316301026,
+      "grad_norm": 0.04911995306611061,
+      "learning_rate": 0.00037037037037037035,
+      "loss": 5.7209,
+      "step": 1200
+    },
+    {
+      "epoch": 2.6681477367397948,
+      "grad_norm": 0.06738217920064926,
+      "learning_rate": 0.00036790123456790125,
+      "loss": 6.1993,
+      "step": 1201
+    },
+    {
+      "epoch": 2.6703693418494865,
+      "grad_norm": 0.07234985381364822,
+      "learning_rate": 0.00036543209876543215,
+      "loss": 5.493,
+      "step": 1202
+    },
+    {
+      "epoch": 2.672590946959178,
+      "grad_norm": 0.07034603506326675,
+      "learning_rate": 0.000362962962962963,
+      "loss": 5.4333,
+      "step": 1203
+    },
+    {
+      "epoch": 2.67481255206887,
+      "grad_norm": 0.04224281385540962,
+      "learning_rate": 0.00036049382716049383,
+      "loss": 6.0671,
+      "step": 1204
+    },
+    {
+      "epoch": 2.6770341571785616,
+      "grad_norm": 0.06850730627775192,
+      "learning_rate": 0.0003580246913580247,
+      "loss": 5.9359,
+      "step": 1205
+    },
+    {
+      "epoch": 2.6792557622882534,
+      "grad_norm": 0.06015927717089653,
+      "learning_rate": 0.00035555555555555557,
+      "loss": 5.7879,
+      "step": 1206
+    },
+    {
+      "epoch": 2.681477367397945,
+      "grad_norm": 0.06330190598964691,
+      "learning_rate": 0.0003530864197530864,
+      "loss": 5.9427,
+      "step": 1207
+    },
+    {
+      "epoch": 2.683698972507637,
+      "grad_norm": 0.06587096303701401,
+      "learning_rate": 0.0003506172839506173,
+      "loss": 5.8831,
+      "step": 1208
+    },
+    {
+      "epoch": 2.6859205776173285,
+      "grad_norm": 0.04329574853181839,
+      "learning_rate": 0.00034814814814814816,
+      "loss": 5.8498,
+      "step": 1209
+    },
+    {
+      "epoch": 2.68814218272702,
+      "grad_norm": 0.06754133105278015,
+      "learning_rate": 0.00034567901234567905,
+      "loss": 6.0184,
+      "step": 1210
+    },
+    {
+      "epoch": 2.690363787836712,
+      "grad_norm": 0.06105200573801994,
+      "learning_rate": 0.0003432098765432099,
+      "loss": 5.7446,
+      "step": 1211
+    },
+    {
+      "epoch": 2.6925853929464036,
+      "grad_norm": 0.06230635941028595,
+      "learning_rate": 0.00034074074074074074,
+      "loss": 6.0185,
+      "step": 1212
+    },
+    {
+      "epoch": 2.6948069980560954,
+      "grad_norm": 0.08427722752094269,
+      "learning_rate": 0.0003382716049382716,
+      "loss": 5.7514,
+      "step": 1213
+    },
+    {
+      "epoch": 2.697028603165787,
+      "grad_norm": 0.05446156859397888,
+      "learning_rate": 0.0003358024691358025,
+      "loss": 5.9268,
+      "step": 1214
+    },
+    {
+      "epoch": 2.699250208275479,
+      "grad_norm": 0.07585136592388153,
+      "learning_rate": 0.0003333333333333333,
+      "loss": 6.0322,
+      "step": 1215
+    },
+    {
+      "epoch": 2.701471813385171,
+      "grad_norm": 0.05803457275032997,
+      "learning_rate": 0.0003308641975308642,
+      "loss": 5.8755,
+      "step": 1216
+    },
+    {
+      "epoch": 2.7036934184948627,
+      "grad_norm": 0.06280191987752914,
+      "learning_rate": 0.00032839506172839506,
+      "loss": 5.6669,
+      "step": 1217
+    },
+    {
+      "epoch": 2.7059150236045544,
+      "grad_norm": 0.05328867584466934,
+      "learning_rate": 0.0003259259259259259,
+      "loss": 5.7171,
+      "step": 1218
+    },
+    {
+      "epoch": 2.708136628714246,
+      "grad_norm": 0.06539945304393768,
+      "learning_rate": 0.0003234567901234568,
+      "loss": 6.1191,
+      "step": 1219
+    },
+    {
+      "epoch": 2.710358233823938,
+      "grad_norm": 0.06002845987677574,
+      "learning_rate": 0.00032098765432098765,
+      "loss": 5.8588,
+      "step": 1220
+    },
+    {
+      "epoch": 2.7125798389336295,
+      "grad_norm": 0.055796269327402115,
+      "learning_rate": 0.00031851851851851854,
+      "loss": 5.7848,
+      "step": 1221
+    },
+    {
+      "epoch": 2.7148014440433212,
+      "grad_norm": 0.055544886738061905,
+      "learning_rate": 0.0003160493827160494,
+      "loss": 5.5401,
+      "step": 1222
+    },
+    {
+      "epoch": 2.717023049153013,
+      "grad_norm": 0.05194031074643135,
+      "learning_rate": 0.0003135802469135803,
+      "loss": 5.7283,
+      "step": 1223
+    },
+    {
+      "epoch": 2.7192446542627047,
+      "grad_norm": 0.057613443583250046,
+      "learning_rate": 0.0003111111111111111,
+      "loss": 5.8282,
+      "step": 1224
+    },
+    {
+      "epoch": 2.7214662593723964,
+      "grad_norm": 0.060236454010009766,
+      "learning_rate": 0.00030864197530864197,
+      "loss": 5.8772,
+      "step": 1225
+    },
+    {
+      "epoch": 2.7236878644820885,
+      "grad_norm": 0.0555168054997921,
+      "learning_rate": 0.0003061728395061728,
+      "loss": 5.9074,
+      "step": 1226
+    },
+    {
+      "epoch": 2.7259094695917803,
+      "grad_norm": 0.04928433895111084,
+      "learning_rate": 0.0003037037037037037,
+      "loss": 6.0044,
+      "step": 1227
+    },
+    {
+      "epoch": 2.728131074701472,
+      "grad_norm": 0.07700496912002563,
+      "learning_rate": 0.00030123456790123455,
+      "loss": 6.0433,
+      "step": 1228
+    },
+    {
+      "epoch": 2.7303526798111637,
+      "grad_norm": 0.04453560337424278,
+      "learning_rate": 0.00029876543209876545,
+      "loss": 5.7116,
+      "step": 1229
+    },
+    {
+      "epoch": 2.7325742849208554,
+      "grad_norm": 0.06810157746076584,
+      "learning_rate": 0.0002962962962962963,
+      "loss": 5.4879,
+      "step": 1230
+    },
+    {
+      "epoch": 2.734795890030547,
+      "grad_norm": 0.059812840074300766,
+      "learning_rate": 0.0002938271604938272,
+      "loss": 5.8898,
+      "step": 1231
+    },
+    {
+      "epoch": 2.737017495140239,
+      "grad_norm": 0.06308246403932571,
+      "learning_rate": 0.00029135802469135803,
+      "loss": 5.6131,
+      "step": 1232
+    },
+    {
+      "epoch": 2.7392391002499306,
+      "grad_norm": 0.04889903962612152,
+      "learning_rate": 0.0002888888888888889,
+      "loss": 5.5685,
+      "step": 1233
+    },
+    {
+      "epoch": 2.7414607053596223,
+      "grad_norm": 0.05145961046218872,
+      "learning_rate": 0.00028641975308641977,
+      "loss": 5.9801,
+      "step": 1234
+    },
+    {
+      "epoch": 2.743682310469314,
+      "grad_norm": 0.0958176925778389,
+      "learning_rate": 0.0002839506172839506,
+      "loss": 5.6457,
+      "step": 1235
+    },
+    {
+      "epoch": 2.7459039155790057,
+      "grad_norm": 0.049419160932302475,
+      "learning_rate": 0.0002814814814814815,
+      "loss": 5.8724,
+      "step": 1236
+    },
+    {
+      "epoch": 2.7481255206886974,
+      "grad_norm": 0.07048829644918442,
+      "learning_rate": 0.00027901234567901236,
+      "loss": 5.9755,
+      "step": 1237
+    },
+    {
+      "epoch": 2.750347125798389,
+      "grad_norm": 0.048005539923906326,
+      "learning_rate": 0.00027654320987654325,
+      "loss": 5.6778,
+      "step": 1238
+    },
+    {
+      "epoch": 2.752568730908081,
+      "grad_norm": 0.05804963409900665,
+      "learning_rate": 0.0002740740740740741,
+      "loss": 5.8681,
+      "step": 1239
+    },
+    {
+      "epoch": 2.7547903360177726,
+      "grad_norm": 0.05638793855905533,
+      "learning_rate": 0.00027160493827160494,
+      "loss": 6.0751,
+      "step": 1240
+    },
+    {
+      "epoch": 2.7570119411274647,
+      "grad_norm": 0.047091707587242126,
+      "learning_rate": 0.0002691358024691358,
+      "loss": 5.8273,
+      "step": 1241
+    },
+    {
+      "epoch": 2.7592335462371564,
+      "grad_norm": 0.07317803055047989,
+      "learning_rate": 0.0002666666666666667,
+      "loss": 5.8787,
+      "step": 1242
+    },
+    {
+      "epoch": 2.761455151346848,
+      "grad_norm": 0.05119254067540169,
+      "learning_rate": 0.0002641975308641975,
+      "loss": 5.8983,
+      "step": 1243
+    },
+    {
+      "epoch": 2.76367675645654,
+      "grad_norm": 0.05338507890701294,
+      "learning_rate": 0.0002617283950617284,
+      "loss": 5.8812,
+      "step": 1244
+    },
+    {
+      "epoch": 2.7658983615662316,
+      "grad_norm": 0.04814758151769638,
+      "learning_rate": 0.00025925925925925926,
+      "loss": 5.9619,
+      "step": 1245
+    },
+    {
+      "epoch": 2.7681199666759233,
+      "grad_norm": 0.06921735405921936,
+      "learning_rate": 0.00025679012345679016,
+      "loss": 5.4222,
+      "step": 1246
+    },
+    {
+      "epoch": 2.770341571785615,
+      "grad_norm": 0.05429822951555252,
+      "learning_rate": 0.000254320987654321,
+      "loss": 5.7432,
+      "step": 1247
+    },
+    {
+      "epoch": 2.7725631768953067,
+      "grad_norm": 0.07386202365159988,
+      "learning_rate": 0.00025185185185185185,
+      "loss": 5.8407,
+      "step": 1248
+    },
+    {
+      "epoch": 2.7747847820049985,
+      "grad_norm": 0.07264435291290283,
+      "learning_rate": 0.0002493827160493827,
+      "loss": 5.9617,
+      "step": 1249
+    },
+    {
+      "epoch": 2.7770063871146906,
+      "grad_norm": 0.05628679692745209,
+      "learning_rate": 0.0002469135802469136,
+      "loss": 6.1518,
+      "step": 1250
+    },
+    {
+      "epoch": 2.7792279922243823,
+      "grad_norm": 0.0481853187084198,
+      "learning_rate": 0.0002444444444444445,
+      "loss": 5.9351,
+      "step": 1251
+    },
+    {
+      "epoch": 2.781449597334074,
+      "grad_norm": 0.04323854297399521,
+      "learning_rate": 0.0002419753086419753,
+      "loss": 6.051,
+      "step": 1252
+    },
+    {
+      "epoch": 2.7836712024437658,
+      "grad_norm": 0.06301230937242508,
+      "learning_rate": 0.0002395061728395062,
+      "loss": 5.686,
+      "step": 1253
+    },
+    {
+      "epoch": 2.7858928075534575,
+      "grad_norm": 0.05843862518668175,
+      "learning_rate": 0.00023703703703703704,
+      "loss": 5.8487,
+      "step": 1254
+    },
+    {
+      "epoch": 2.788114412663149,
+      "grad_norm": 0.05112096294760704,
+      "learning_rate": 0.0002345679012345679,
+      "loss": 5.7456,
+      "step": 1255
+    },
+    {
+      "epoch": 2.790336017772841,
+      "grad_norm": 0.06781143695116043,
+      "learning_rate": 0.00023209876543209875,
+      "loss": 5.8486,
+      "step": 1256
+    },
+    {
+      "epoch": 2.7925576228825326,
+      "grad_norm": 0.04605870693922043,
+      "learning_rate": 0.00022962962962962965,
+      "loss": 5.8111,
+      "step": 1257
+    },
+    {
+      "epoch": 2.7947792279922243,
+      "grad_norm": 0.058881402015686035,
+      "learning_rate": 0.0002271604938271605,
+      "loss": 6.2209,
+      "step": 1258
+    },
+    {
+      "epoch": 2.797000833101916,
+      "grad_norm": 0.0564490482211113,
+      "learning_rate": 0.00022469135802469136,
+      "loss": 5.887,
+      "step": 1259
+    },
+    {
+      "epoch": 2.7992224382116078,
+      "grad_norm": 0.06960400193929672,
+      "learning_rate": 0.0002222222222222222,
+      "loss": 6.148,
+      "step": 1260
+    },
+    {
+      "epoch": 2.8014440433212995,
+      "grad_norm": 0.05847383290529251,
+      "learning_rate": 0.0002197530864197531,
+      "loss": 5.9894,
+      "step": 1261
+    },
+    {
+      "epoch": 2.803665648430991,
+      "grad_norm": 0.0737534686923027,
+      "learning_rate": 0.00021728395061728395,
+      "loss": 6.0911,
+      "step": 1262
+    },
+    {
+      "epoch": 2.805887253540683,
+      "grad_norm": 0.07557881623506546,
+      "learning_rate": 0.00021481481481481482,
+      "loss": 5.7048,
+      "step": 1263
+    },
+    {
+      "epoch": 2.8081088586503746,
+      "grad_norm": 0.06748920679092407,
+      "learning_rate": 0.00021234567901234566,
+      "loss": 6.051,
+      "step": 1264
+    },
+    {
+      "epoch": 2.810330463760067,
+      "grad_norm": 0.05918078124523163,
+      "learning_rate": 0.00020987654320987656,
+      "loss": 5.9721,
+      "step": 1265
+    },
+    {
+      "epoch": 2.8125520688697585,
+      "grad_norm": 0.06494231522083282,
+      "learning_rate": 0.0002074074074074074,
+      "loss": 5.8291,
+      "step": 1266
+    },
+    {
+      "epoch": 2.81477367397945,
+      "grad_norm": 0.05254039913415909,
+      "learning_rate": 0.00020493827160493827,
+      "loss": 5.9545,
+      "step": 1267
+    },
+    {
+      "epoch": 2.816995279089142,
+      "grad_norm": 0.05591006949543953,
+      "learning_rate": 0.00020246913580246917,
+      "loss": 5.7866,
+      "step": 1268
+    },
+    {
+      "epoch": 2.8192168841988337,
+      "grad_norm": 0.05167578533291817,
+      "learning_rate": 0.0002,
+      "loss": 5.8748,
+      "step": 1269
+    },
+    {
+      "epoch": 2.8214384893085254,
+      "grad_norm": 0.08402105420827866,
+      "learning_rate": 0.00019753086419753088,
+      "loss": 5.6071,
+      "step": 1270
+    },
+    {
+      "epoch": 2.823660094418217,
+      "grad_norm": 0.05030294507741928,
+      "learning_rate": 0.00019506172839506172,
+      "loss": 6.0468,
+      "step": 1271
+    },
+    {
+      "epoch": 2.825881699527909,
+      "grad_norm": 0.05387067794799805,
+      "learning_rate": 0.00019259259259259262,
+      "loss": 5.9581,
+      "step": 1272
+    },
+    {
+      "epoch": 2.8281033046376005,
+      "grad_norm": 0.04881101846694946,
+      "learning_rate": 0.00019012345679012346,
+      "loss": 5.703,
+      "step": 1273
+    },
+    {
+      "epoch": 2.8303249097472927,
+      "grad_norm": 0.05300934612751007,
+      "learning_rate": 0.00018765432098765433,
+      "loss": 5.7739,
+      "step": 1274
+    },
+    {
+      "epoch": 2.8325465148569844,
+      "grad_norm": 0.050771087408065796,
+      "learning_rate": 0.00018518518518518518,
+      "loss": 5.9551,
+      "step": 1275
+    },
+    {
+      "epoch": 2.834768119966676,
+      "grad_norm": 0.052648287266492844,
+      "learning_rate": 0.00018271604938271607,
+      "loss": 5.8995,
+      "step": 1276
+    },
+    {
+      "epoch": 2.836989725076368,
+      "grad_norm": 0.05061740055680275,
+      "learning_rate": 0.00018024691358024692,
+      "loss": 5.7913,
+      "step": 1277
+    },
+    {
+      "epoch": 2.8392113301860595,
+      "grad_norm": 0.04365115985274315,
+      "learning_rate": 0.00017777777777777779,
+      "loss": 5.823,
+      "step": 1278
+    },
+    {
+      "epoch": 2.8414329352957512,
+      "grad_norm": 0.06664296984672546,
+      "learning_rate": 0.00017530864197530866,
+      "loss": 6.2006,
+      "step": 1279
+    },
+    {
+      "epoch": 2.843654540405443,
+      "grad_norm": 0.055986661463975906,
+      "learning_rate": 0.00017283950617283953,
+      "loss": 5.8978,
+      "step": 1280
+    },
+    {
+      "epoch": 2.8458761455151347,
+      "grad_norm": 0.04974586144089699,
+      "learning_rate": 0.00017037037037037037,
+      "loss": 6.0127,
+      "step": 1281
+    },
+    {
+      "epoch": 2.8480977506248264,
+      "grad_norm": 0.06275002658367157,
+      "learning_rate": 0.00016790123456790124,
+      "loss": 6.3379,
+      "step": 1282
+    },
+    {
+      "epoch": 2.850319355734518,
+      "grad_norm": 0.05482464283704758,
+      "learning_rate": 0.0001654320987654321,
+      "loss": 5.7202,
+      "step": 1283
+    },
+    {
+      "epoch": 2.85254096084421,
+      "grad_norm": 0.056165117770433426,
+      "learning_rate": 0.00016296296296296295,
+      "loss": 5.7521,
+      "step": 1284
+    },
+    {
+      "epoch": 2.8547625659539015,
+      "grad_norm": 0.05828070640563965,
+      "learning_rate": 0.00016049382716049382,
+      "loss": 6.176,
+      "step": 1285
+    },
+    {
+      "epoch": 2.8569841710635933,
+      "grad_norm": 0.0738142803311348,
+      "learning_rate": 0.0001580246913580247,
+      "loss": 6.1232,
+      "step": 1286
+    },
+    {
+      "epoch": 2.859205776173285,
+      "grad_norm": 0.051991552114486694,
+      "learning_rate": 0.00015555555555555556,
+      "loss": 5.7182,
+      "step": 1287
+    },
+    {
+      "epoch": 2.8614273812829767,
+      "grad_norm": 0.06892450898885727,
+      "learning_rate": 0.0001530864197530864,
+      "loss": 6.0863,
+      "step": 1288
+    },
+    {
+      "epoch": 2.863648986392669,
+      "grad_norm": 0.04431318864226341,
+      "learning_rate": 0.00015061728395061728,
+      "loss": 6.0564,
+      "step": 1289
+    },
+    {
+      "epoch": 2.8658705915023606,
+      "grad_norm": 0.07410877197980881,
+      "learning_rate": 0.00014814814814814815,
+      "loss": 6.0874,
+      "step": 1290
+    },
+    {
+      "epoch": 2.8680921966120523,
+      "grad_norm": 0.05466211214661598,
+      "learning_rate": 0.00014567901234567902,
+      "loss": 5.9654,
+      "step": 1291
+    },
+    {
+      "epoch": 2.870313801721744,
+      "grad_norm": 0.045556023716926575,
+      "learning_rate": 0.00014320987654320989,
+      "loss": 5.8639,
+      "step": 1292
+    },
+    {
+      "epoch": 2.8725354068314357,
+      "grad_norm": 0.04874875396490097,
+      "learning_rate": 0.00014074074074074076,
+      "loss": 5.9451,
+      "step": 1293
+    },
+    {
+      "epoch": 2.8747570119411274,
+      "grad_norm": 0.05590863898396492,
+      "learning_rate": 0.00013827160493827163,
+      "loss": 6.0744,
+      "step": 1294
+    },
+    {
+      "epoch": 2.876978617050819,
+      "grad_norm": 0.05163506045937538,
+      "learning_rate": 0.00013580246913580247,
+      "loss": 5.6175,
+      "step": 1295
+    },
+    {
+      "epoch": 2.879200222160511,
+      "grad_norm": 0.058315426111221313,
+      "learning_rate": 0.00013333333333333334,
+      "loss": 5.7961,
+      "step": 1296
+    },
+    {
+      "epoch": 2.8814218272702026,
+      "grad_norm": 0.0603594146668911,
+      "learning_rate": 0.0001308641975308642,
+      "loss": 5.8225,
+      "step": 1297
+    },
+    {
+      "epoch": 2.8836434323798947,
+      "grad_norm": 0.06714392453432083,
+      "learning_rate": 0.00012839506172839508,
+      "loss": 5.9419,
+      "step": 1298
+    },
+    {
+      "epoch": 2.8858650374895864,
+      "grad_norm": 0.06128297373652458,
+      "learning_rate": 0.00012592592592592592,
+      "loss": 5.6193,
+      "step": 1299
+    },
+    {
+      "epoch": 2.888086642599278,
+      "grad_norm": 0.06052326411008835,
+      "learning_rate": 0.0001234567901234568,
+      "loss": 5.4785,
+      "step": 1300
+    },
+    {
+      "epoch": 2.89030824770897,
+      "grad_norm": 0.0529288649559021,
+      "learning_rate": 0.00012098765432098765,
+      "loss": 6.2424,
+      "step": 1301
+    },
+    {
+      "epoch": 2.8925298528186616,
+      "grad_norm": 0.05307260900735855,
+      "learning_rate": 0.00011851851851851852,
+      "loss": 5.9248,
+      "step": 1302
+    },
+    {
+      "epoch": 2.8947514579283533,
+      "grad_norm": 0.04504923149943352,
+      "learning_rate": 0.00011604938271604938,
+      "loss": 5.6298,
+      "step": 1303
+    },
+    {
+      "epoch": 2.896973063038045,
+      "grad_norm": 0.06184275075793266,
+      "learning_rate": 0.00011358024691358025,
+      "loss": 6.0494,
+      "step": 1304
+    },
+    {
+      "epoch": 2.8991946681477367,
+      "grad_norm": 0.06522642076015472,
+      "learning_rate": 0.0001111111111111111,
+      "loss": 5.9472,
+      "step": 1305
+    },
+    {
+      "epoch": 2.9014162732574285,
+      "grad_norm": 0.045491501688957214,
+      "learning_rate": 0.00010864197530864197,
+      "loss": 5.7783,
+      "step": 1306
+    },
+    {
+      "epoch": 2.90363787836712,
+      "grad_norm": 0.05264101177453995,
+      "learning_rate": 0.00010617283950617283,
+      "loss": 5.7188,
+      "step": 1307
+    },
+    {
+      "epoch": 2.905859483476812,
+      "grad_norm": 0.08518176525831223,
+      "learning_rate": 0.0001037037037037037,
+      "loss": 6.0351,
+      "step": 1308
+    },
+    {
+      "epoch": 2.9080810885865036,
+      "grad_norm": 0.06091070547699928,
+      "learning_rate": 0.00010123456790123458,
+      "loss": 5.5296,
+      "step": 1309
+    },
+    {
+      "epoch": 2.9103026936961953,
+      "grad_norm": 0.06074732169508934,
+      "learning_rate": 9.876543209876544e-05,
+      "loss": 5.6439,
+      "step": 1310
+    },
+    {
+      "epoch": 2.912524298805887,
+      "grad_norm": 0.04879772290587425,
+      "learning_rate": 9.629629629629631e-05,
+      "loss": 5.7684,
+      "step": 1311
+    },
+    {
+      "epoch": 2.9147459039155788,
+      "grad_norm": 0.09098317474126816,
+      "learning_rate": 9.382716049382717e-05,
+      "loss": 6.2529,
+      "step": 1312
+    },
+    {
+      "epoch": 2.916967509025271,
+      "grad_norm": 0.06702974438667297,
+      "learning_rate": 9.135802469135804e-05,
+      "loss": 5.9017,
+      "step": 1313
+    },
+    {
+      "epoch": 2.9191891141349626,
+      "grad_norm": 0.05466683581471443,
+      "learning_rate": 8.888888888888889e-05,
+      "loss": 5.8515,
+      "step": 1314
+    },
+    {
+      "epoch": 2.9214107192446543,
+      "grad_norm": 0.10034381598234177,
+      "learning_rate": 8.641975308641976e-05,
+      "loss": 5.572,
+      "step": 1315
+    },
+    {
+      "epoch": 2.923632324354346,
+      "grad_norm": 0.049551207572221756,
+      "learning_rate": 8.395061728395062e-05,
+      "loss": 5.622,
+      "step": 1316
+    },
+    {
+      "epoch": 2.9258539294640378,
+      "grad_norm": 0.06635282188653946,
+      "learning_rate": 8.148148148148148e-05,
+      "loss": 6.2147,
+      "step": 1317
+    },
+    {
+      "epoch": 2.9280755345737295,
+      "grad_norm": 0.0814097449183464,
+      "learning_rate": 7.901234567901235e-05,
+      "loss": 5.8872,
+      "step": 1318
+    },
+    {
+      "epoch": 2.930297139683421,
+      "grad_norm": 0.06352907419204712,
+      "learning_rate": 7.65432098765432e-05,
+      "loss": 5.875,
+      "step": 1319
+    },
+    {
+      "epoch": 2.932518744793113,
+      "grad_norm": 0.0544845350086689,
+      "learning_rate": 7.407407407407407e-05,
+      "loss": 5.9321,
+      "step": 1320
+    },
+    {
+      "epoch": 2.9347403499028046,
+      "grad_norm": 0.04383888095617294,
+      "learning_rate": 7.160493827160494e-05,
+      "loss": 6.0925,
+      "step": 1321
+    },
+    {
+      "epoch": 2.936961955012497,
+      "grad_norm": 0.06635189801454544,
+      "learning_rate": 6.913580246913581e-05,
+      "loss": 6.0666,
+      "step": 1322
+    },
+    {
+      "epoch": 2.9391835601221885,
+      "grad_norm": 0.06565090268850327,
+      "learning_rate": 6.666666666666667e-05,
+      "loss": 6.0449,
+      "step": 1323
+    },
+    {
+      "epoch": 2.9414051652318802,
+      "grad_norm": 0.04763912782073021,
+      "learning_rate": 6.419753086419754e-05,
+      "loss": 6.076,
+      "step": 1324
+    },
+    {
+      "epoch": 2.943626770341572,
+      "grad_norm": 0.07313438504934311,
+      "learning_rate": 6.17283950617284e-05,
+      "loss": 5.9441,
+      "step": 1325
+    },
+    {
+      "epoch": 2.9458483754512637,
+      "grad_norm": 0.07543905079364777,
+      "learning_rate": 5.925925925925926e-05,
+      "loss": 6.0067,
+      "step": 1326
+    },
+    {
+      "epoch": 2.9480699805609554,
+      "grad_norm": 0.057329267263412476,
+      "learning_rate": 5.679012345679012e-05,
+      "loss": 5.655,
+      "step": 1327
+    },
+    {
+      "epoch": 2.950291585670647,
+      "grad_norm": 0.04951634630560875,
+      "learning_rate": 5.4320987654320986e-05,
+      "loss": 5.9234,
+      "step": 1328
+    },
+    {
+      "epoch": 2.952513190780339,
+      "grad_norm": 0.06038312241435051,
+      "learning_rate": 5.185185185185185e-05,
+      "loss": 5.9723,
+      "step": 1329
+    },
+    {
+      "epoch": 2.9547347958900305,
+      "grad_norm": 0.05027232691645622,
+      "learning_rate": 4.938271604938272e-05,
+      "loss": 5.6912,
+      "step": 1330
+    },
+    {
+      "epoch": 2.9569564009997222,
+      "grad_norm": 0.056142378598451614,
+      "learning_rate": 4.691358024691358e-05,
+      "loss": 5.8446,
+      "step": 1331
+    },
+    {
+      "epoch": 2.959178006109414,
+      "grad_norm": 0.06986388564109802,
+      "learning_rate": 4.4444444444444447e-05,
+      "loss": 5.8146,
+      "step": 1332
+    },
+    {
+      "epoch": 2.9613996112191057,
+      "grad_norm": 0.059677302837371826,
+      "learning_rate": 4.197530864197531e-05,
+      "loss": 6.0107,
+      "step": 1333
+    },
+    {
+      "epoch": 2.9636212163287974,
+      "grad_norm": 0.06304765492677689,
+      "learning_rate": 3.950617283950617e-05,
+      "loss": 6.0351,
+      "step": 1334
+    },
+    {
+      "epoch": 2.965842821438489,
+      "grad_norm": 0.05687398463487625,
+      "learning_rate": 3.7037037037037037e-05,
+      "loss": 5.9451,
+      "step": 1335
+    },
+    {
+      "epoch": 2.968064426548181,
+      "grad_norm": 0.04591461271047592,
+      "learning_rate": 3.456790123456791e-05,
+      "loss": 6.1389,
+      "step": 1336
+    },
+    {
+      "epoch": 2.970286031657873,
+      "grad_norm": 0.06220898777246475,
+      "learning_rate": 3.209876543209877e-05,
+      "loss": 5.6631,
+      "step": 1337
+    },
+    {
+      "epoch": 2.9725076367675647,
+      "grad_norm": 0.04418789967894554,
+      "learning_rate": 2.962962962962963e-05,
+      "loss": 5.8768,
+      "step": 1338
+    },
+    {
+      "epoch": 2.9747292418772564,
+      "grad_norm": 0.05900184065103531,
+      "learning_rate": 2.7160493827160493e-05,
+      "loss": 5.945,
+      "step": 1339
+    },
+    {
+      "epoch": 2.976950846986948,
+      "grad_norm": 0.05201757326722145,
+      "learning_rate": 2.469135802469136e-05,
+      "loss": 5.9387,
+      "step": 1340
+    },
+    {
+      "epoch": 2.97917245209664,
+      "grad_norm": 0.05885840207338333,
+      "learning_rate": 2.2222222222222223e-05,
+      "loss": 5.904,
+      "step": 1341
+    },
+    {
+      "epoch": 2.9813940572063315,
+      "grad_norm": 0.05918167158961296,
+      "learning_rate": 1.9753086419753087e-05,
+      "loss": 5.8086,
+      "step": 1342
+    },
+    {
+      "epoch": 2.9836156623160233,
+      "grad_norm": 0.06754791736602783,
+      "learning_rate": 1.7283950617283953e-05,
+      "loss": 5.6532,
+      "step": 1343
+    },
+    {
+      "epoch": 2.985837267425715,
+      "grad_norm": 0.06648389250040054,
+      "learning_rate": 1.4814814814814815e-05,
+      "loss": 5.9772,
+      "step": 1344
+    },
+    {
+      "epoch": 2.9880588725354067,
+      "grad_norm": 0.05073931813240051,
+      "learning_rate": 1.234567901234568e-05,
+      "loss": 6.1126,
+      "step": 1345
+    },
+    {
+      "epoch": 2.990280477645099,
+      "grad_norm": 0.051233015954494476,
+      "learning_rate": 9.876543209876543e-06,
+      "loss": 5.8266,
+      "step": 1346
+    },
+    {
+      "epoch": 2.9925020827547906,
+      "grad_norm": 0.06040095537900925,
+      "learning_rate": 7.4074074074074075e-06,
+      "loss": 5.7803,
+      "step": 1347
+    },
+    {
+      "epoch": 2.9947236878644823,
+      "grad_norm": 0.04836644232273102,
+      "learning_rate": 4.938271604938272e-06,
+      "loss": 5.9451,
+      "step": 1348
+    },
+    {
+      "epoch": 2.996945292974174,
+      "grad_norm": 0.06376051902770996,
+      "learning_rate": 2.469135802469136e-06,
+      "loss": 5.7011,
+      "step": 1349
+    },
+    {
+      "epoch": 2.9991668980838657,
+      "grad_norm": 0.052688080817461014,
+      "learning_rate": 0.0,
+      "loss": 5.8671,
+      "step": 1350
+    },
+    {
+      "epoch": 2.9991668980838657,
+      "step": 1350,
+      "total_flos": 2.6133172282667827e+17,
+      "train_loss": 6.73684572007921,
+      "train_runtime": 2224.0851,
+      "train_samples_per_second": 9.715,
+      "train_steps_per_second": 0.607
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 1350,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.6133172282667827e+17,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}