diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,9597 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 1365,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.002197802197802198,
+      "grad_norm": 4.858581066131592,
+      "learning_rate": 0.0,
+      "loss": 8.543,
+      "step": 1
+    },
+    {
+      "epoch": 0.004395604395604396,
+      "grad_norm": 12.996786117553711,
+      "learning_rate": 2.18978102189781e-05,
+      "loss": 8.498,
+      "step": 2
+    },
+    {
+      "epoch": 0.006593406593406593,
+      "grad_norm": 5.062016010284424,
+      "learning_rate": 4.37956204379562e-05,
+      "loss": 8.2324,
+      "step": 3
+    },
+    {
+      "epoch": 0.008791208791208791,
+      "grad_norm": 13.022345542907715,
+      "learning_rate": 6.569343065693432e-05,
+      "loss": 7.875,
+      "step": 4
+    },
+    {
+      "epoch": 0.01098901098901099,
+      "grad_norm": 6.494384765625,
+      "learning_rate": 8.75912408759124e-05,
+      "loss": 7.8906,
+      "step": 5
+    },
+    {
+      "epoch": 0.013186813186813187,
+      "grad_norm": 5.402078151702881,
+      "learning_rate": 0.00010948905109489051,
+      "loss": 8.375,
+      "step": 6
+    },
+    {
+      "epoch": 0.015384615384615385,
+      "grad_norm": 28.781536102294922,
+      "learning_rate": 0.00013138686131386864,
+      "loss": 8.1406,
+      "step": 7
+    },
+    {
+      "epoch": 0.017582417582417582,
+      "grad_norm": 6.1622419357299805,
+      "learning_rate": 0.00015328467153284672,
+      "loss": 8.3027,
+      "step": 8
+    },
+    {
+      "epoch": 0.01978021978021978,
+      "grad_norm": 13.30415153503418,
+      "learning_rate": 0.0001751824817518248,
+      "loss": 7.8691,
+      "step": 9
+    },
+    {
+      "epoch": 0.02197802197802198,
+      "grad_norm": 21.408166885375977,
+      "learning_rate": 0.00019708029197080293,
+      "loss": 9.0117,
+      "step": 10
+    },
+    {
+      "epoch": 0.024175824175824177,
+      "grad_norm": 10.444605827331543,
+      "learning_rate": 0.00021897810218978101,
+      "loss": 9.0156,
+      "step": 11
+    },
+    {
+      "epoch": 0.026373626373626374,
+      "grad_norm": 11.979896545410156,
+      "learning_rate": 0.00024087591240875913,
+      "loss": 8.7109,
+      "step": 12
+    },
+    {
+      "epoch": 0.02857142857142857,
+      "grad_norm": 3.8272347450256348,
+      "learning_rate": 0.00026277372262773727,
+      "loss": 8.3633,
+      "step": 13
+    },
+    {
+      "epoch": 0.03076923076923077,
+      "grad_norm": 6.037533760070801,
+      "learning_rate": 0.00028467153284671533,
+      "loss": 8.5,
+      "step": 14
+    },
+    {
+      "epoch": 0.03296703296703297,
+      "grad_norm": 12.077258110046387,
+      "learning_rate": 0.00030656934306569345,
+      "loss": 9.0801,
+      "step": 15
+    },
+    {
+      "epoch": 0.035164835164835165,
+      "grad_norm": 4.213659286499023,
+      "learning_rate": 0.00032846715328467156,
+      "loss": 8.3984,
+      "step": 16
+    },
+    {
+      "epoch": 0.03736263736263736,
+      "grad_norm": 10.319239616394043,
+      "learning_rate": 0.0003503649635036496,
+      "loss": 8.959,
+      "step": 17
+    },
+    {
+      "epoch": 0.03956043956043956,
+      "grad_norm": 3.72821307182312,
+      "learning_rate": 0.00037226277372262774,
+      "loss": 8.3789,
+      "step": 18
+    },
+    {
+      "epoch": 0.041758241758241756,
+      "grad_norm": 6.265636444091797,
+      "learning_rate": 0.00039416058394160585,
+      "loss": 8.3184,
+      "step": 19
+    },
+    {
+      "epoch": 0.04395604395604396,
+      "grad_norm": 4.621541500091553,
+      "learning_rate": 0.00041605839416058397,
+      "loss": 8.252,
+      "step": 20
+    },
+    {
+      "epoch": 0.046153846153846156,
+      "grad_norm": 3.733940839767456,
+      "learning_rate": 0.00043795620437956203,
+      "loss": 8.0234,
+      "step": 21
+    },
+    {
+      "epoch": 0.04835164835164835,
+      "grad_norm": 11.054878234863281,
+      "learning_rate": 0.0004598540145985402,
+      "loss": 9.1758,
+      "step": 22
+    },
+    {
+      "epoch": 0.05054945054945055,
+      "grad_norm": 6.400592803955078,
+      "learning_rate": 0.00048175182481751826,
+      "loss": 8.6973,
+      "step": 23
+    },
+    {
+      "epoch": 0.05274725274725275,
+      "grad_norm": 6.150137901306152,
+      "learning_rate": 0.0005036496350364964,
+      "loss": 8.4805,
+      "step": 24
+    },
+    {
+      "epoch": 0.054945054945054944,
+      "grad_norm": 5.878363132476807,
+      "learning_rate": 0.0005255474452554745,
+      "loss": 7.8086,
+      "step": 25
+    },
+    {
+      "epoch": 0.05714285714285714,
+      "grad_norm": 6.445720672607422,
+      "learning_rate": 0.0005474452554744526,
+      "loss": 8.5684,
+      "step": 26
+    },
+    {
+      "epoch": 0.05934065934065934,
+      "grad_norm": 6.152160167694092,
+      "learning_rate": 0.0005693430656934307,
+      "loss": 7.6484,
+      "step": 27
+    },
+    {
+      "epoch": 0.06153846153846154,
+      "grad_norm": 12.085118293762207,
+      "learning_rate": 0.0005912408759124088,
+      "loss": 8.6816,
+      "step": 28
+    },
+    {
+      "epoch": 0.06373626373626373,
+      "grad_norm": 5.764835357666016,
+      "learning_rate": 0.0006131386861313869,
+      "loss": 7.5371,
+      "step": 29
+    },
+    {
+      "epoch": 0.06593406593406594,
+      "grad_norm": 4.375746250152588,
+      "learning_rate": 0.000635036496350365,
+      "loss": 7.998,
+      "step": 30
+    },
+    {
+      "epoch": 0.06813186813186813,
+      "grad_norm": 22.013959884643555,
+      "learning_rate": 0.0006569343065693431,
+      "loss": 9.0684,
+      "step": 31
+    },
+    {
+      "epoch": 0.07032967032967033,
+      "grad_norm": 10.12054443359375,
+      "learning_rate": 0.0006788321167883212,
+      "loss": 8.5527,
+      "step": 32
+    },
+    {
+      "epoch": 0.07252747252747253,
+      "grad_norm": 6.671593189239502,
+      "learning_rate": 0.0007007299270072992,
+      "loss": 7.9355,
+      "step": 33
+    },
+    {
+      "epoch": 0.07472527472527472,
+      "grad_norm": 6.172004699707031,
+      "learning_rate": 0.0007226277372262774,
+      "loss": 7.9648,
+      "step": 34
+    },
+    {
+      "epoch": 0.07692307692307693,
+      "grad_norm": 5.27730131149292,
+      "learning_rate": 0.0007445255474452555,
+      "loss": 8.0723,
+      "step": 35
+    },
+    {
+      "epoch": 0.07912087912087912,
+      "grad_norm": 5.107365608215332,
+      "learning_rate": 0.0007664233576642336,
+      "loss": 8.3008,
+      "step": 36
+    },
+    {
+      "epoch": 0.08131868131868132,
+      "grad_norm": 15.500955581665039,
+      "learning_rate": 0.0007883211678832117,
+      "loss": 7.9629,
+      "step": 37
+    },
+    {
+      "epoch": 0.08351648351648351,
+      "grad_norm": 14.203560829162598,
+      "learning_rate": 0.0008102189781021898,
+      "loss": 8.9434,
+      "step": 38
+    },
+    {
+      "epoch": 0.08571428571428572,
+      "grad_norm": 4.801837921142578,
+      "learning_rate": 0.0008321167883211679,
+      "loss": 8.3301,
+      "step": 39
+    },
+    {
+      "epoch": 0.08791208791208792,
+      "grad_norm": 5.906317234039307,
+      "learning_rate": 0.000854014598540146,
+      "loss": 8.0156,
+      "step": 40
+    },
+    {
+      "epoch": 0.09010989010989011,
+      "grad_norm": 38.40510177612305,
+      "learning_rate": 0.0008759124087591241,
+      "loss": 7.9961,
+      "step": 41
+    },
+    {
+      "epoch": 0.09230769230769231,
+      "grad_norm": 6.164946556091309,
+      "learning_rate": 0.0008978102189781023,
+      "loss": 7.9961,
+      "step": 42
+    },
+    {
+      "epoch": 0.0945054945054945,
+      "grad_norm": 6.462623119354248,
+      "learning_rate": 0.0009197080291970804,
+      "loss": 7.7305,
+      "step": 43
+    },
+    {
+      "epoch": 0.0967032967032967,
+      "grad_norm": 10.502652168273926,
+      "learning_rate": 0.0009416058394160585,
+      "loss": 7.9297,
+      "step": 44
+    },
+    {
+      "epoch": 0.0989010989010989,
+      "grad_norm": 5.193427085876465,
+      "learning_rate": 0.0009635036496350365,
+      "loss": 9.0039,
+      "step": 45
+    },
+    {
+      "epoch": 0.1010989010989011,
+      "grad_norm": 5.1076226234436035,
+      "learning_rate": 0.0009854014598540146,
+      "loss": 7.9512,
+      "step": 46
+    },
+    {
+      "epoch": 0.10329670329670329,
+      "grad_norm": 6.184417724609375,
+      "learning_rate": 0.0010072992700729927,
+      "loss": 7.5312,
+      "step": 47
+    },
+    {
+      "epoch": 0.1054945054945055,
+      "grad_norm": 7.716291904449463,
+      "learning_rate": 0.001029197080291971,
+      "loss": 8.3613,
+      "step": 48
+    },
+    {
+      "epoch": 0.1076923076923077,
+      "grad_norm": 7.284091949462891,
+      "learning_rate": 0.001051094890510949,
+      "loss": 8.0898,
+      "step": 49
+    },
+    {
+      "epoch": 0.10989010989010989,
+      "grad_norm": 3.336089611053467,
+      "learning_rate": 0.001072992700729927,
+      "loss": 8.709,
+      "step": 50
+    },
+    {
+      "epoch": 0.11208791208791209,
+      "grad_norm": 13.434599876403809,
+      "learning_rate": 0.0010948905109489052,
+      "loss": 8.1211,
+      "step": 51
+    },
+    {
+      "epoch": 0.11428571428571428,
+      "grad_norm": 3.872445821762085,
+      "learning_rate": 0.0011167883211678832,
+      "loss": 7.748,
+      "step": 52
+    },
+    {
+      "epoch": 0.11648351648351649,
+      "grad_norm": 2.639319658279419,
+      "learning_rate": 0.0011386861313868613,
+      "loss": 7.1836,
+      "step": 53
+    },
+    {
+      "epoch": 0.11868131868131868,
+      "grad_norm": 20.89971160888672,
+      "learning_rate": 0.0011605839416058395,
+      "loss": 7.9844,
+      "step": 54
+    },
+    {
+      "epoch": 0.12087912087912088,
+      "grad_norm": 2.698054552078247,
+      "learning_rate": 0.0011824817518248177,
+      "loss": 7.4707,
+      "step": 55
+    },
+    {
+      "epoch": 0.12307692307692308,
+      "grad_norm": 7.622622489929199,
+      "learning_rate": 0.0012043795620437956,
+      "loss": 7.252,
+      "step": 56
+    },
+    {
+      "epoch": 0.12527472527472527,
+      "grad_norm": 4.694027423858643,
+      "learning_rate": 0.0012262773722627738,
+      "loss": 7.9629,
+      "step": 57
+    },
+    {
+      "epoch": 0.12747252747252746,
+      "grad_norm": 5.863952159881592,
+      "learning_rate": 0.0012481751824817517,
+      "loss": 8.1895,
+      "step": 58
+    },
+    {
+      "epoch": 0.12967032967032968,
+      "grad_norm": 5.521010398864746,
+      "learning_rate": 0.00127007299270073,
+      "loss": 7.8125,
+      "step": 59
+    },
+    {
+      "epoch": 0.13186813186813187,
+      "grad_norm": 3.25895619392395,
+      "learning_rate": 0.001291970802919708,
+      "loss": 7.6133,
+      "step": 60
+    },
+    {
+      "epoch": 0.13406593406593406,
+      "grad_norm": 6.815740585327148,
+      "learning_rate": 0.0013138686131386862,
+      "loss": 7.4512,
+      "step": 61
+    },
+    {
+      "epoch": 0.13626373626373625,
+      "grad_norm": 5.058718681335449,
+      "learning_rate": 0.0013357664233576642,
+      "loss": 7.6348,
+      "step": 62
+    },
+    {
+      "epoch": 0.13846153846153847,
+      "grad_norm": 3.2182068824768066,
+      "learning_rate": 0.0013576642335766424,
+      "loss": 7.7832,
+      "step": 63
+    },
+    {
+      "epoch": 0.14065934065934066,
+      "grad_norm": 5.2666497230529785,
+      "learning_rate": 0.0013795620437956205,
+      "loss": 7.6035,
+      "step": 64
+    },
+    {
+      "epoch": 0.14285714285714285,
+      "grad_norm": 6.392567157745361,
+      "learning_rate": 0.0014014598540145985,
+      "loss": 7.7109,
+      "step": 65
+    },
+    {
+      "epoch": 0.14505494505494507,
+      "grad_norm": 5.242607116699219,
+      "learning_rate": 0.0014233576642335767,
+      "loss": 7.9375,
+      "step": 66
+    },
+    {
+      "epoch": 0.14725274725274726,
+      "grad_norm": 7.719801425933838,
+      "learning_rate": 0.0014452554744525548,
+      "loss": 7.7969,
+      "step": 67
+    },
+    {
+      "epoch": 0.14945054945054945,
+      "grad_norm": 3.7887346744537354,
+      "learning_rate": 0.001467153284671533,
+      "loss": 7.8457,
+      "step": 68
+    },
+    {
+      "epoch": 0.15164835164835164,
+      "grad_norm": 4.152592658996582,
+      "learning_rate": 0.001489051094890511,
+      "loss": 7.6367,
+      "step": 69
+    },
+    {
+      "epoch": 0.15384615384615385,
+      "grad_norm": 4.131302356719971,
+      "learning_rate": 0.0015109489051094893,
+      "loss": 7.5156,
+      "step": 70
+    },
+    {
+      "epoch": 0.15604395604395604,
+      "grad_norm": 12.243279457092285,
+      "learning_rate": 0.0015328467153284673,
+      "loss": 7.7441,
+      "step": 71
+    },
+    {
+      "epoch": 0.15824175824175823,
+      "grad_norm": 5.938039779663086,
+      "learning_rate": 0.0015547445255474455,
+      "loss": 7.5625,
+      "step": 72
+    },
+    {
+      "epoch": 0.16043956043956045,
+      "grad_norm": 7.562601089477539,
+      "learning_rate": 0.0015766423357664234,
+      "loss": 7.5859,
+      "step": 73
+    },
+    {
+      "epoch": 0.16263736263736264,
+      "grad_norm": 5.672028541564941,
+      "learning_rate": 0.0015985401459854016,
+      "loss": 7.7734,
+      "step": 74
+    },
+    {
+      "epoch": 0.16483516483516483,
+      "grad_norm": 4.41909646987915,
+      "learning_rate": 0.0016204379562043795,
+      "loss": 7.8574,
+      "step": 75
+    },
+    {
+      "epoch": 0.16703296703296702,
+      "grad_norm": 6.357317924499512,
+      "learning_rate": 0.0016423357664233577,
+      "loss": 7.3887,
+      "step": 76
+    },
+    {
+      "epoch": 0.16923076923076924,
+      "grad_norm": 4.766127109527588,
+      "learning_rate": 0.0016642335766423359,
+      "loss": 7.6016,
+      "step": 77
+    },
+    {
+      "epoch": 0.17142857142857143,
+      "grad_norm": 7.695120811462402,
+      "learning_rate": 0.0016861313868613138,
+      "loss": 7.5918,
+      "step": 78
+    },
+    {
+      "epoch": 0.17362637362637362,
+      "grad_norm": 3.586937427520752,
+      "learning_rate": 0.001708029197080292,
+      "loss": 7.4766,
+      "step": 79
+    },
+    {
+      "epoch": 0.17582417582417584,
+      "grad_norm": 2.6052229404449463,
+      "learning_rate": 0.00172992700729927,
+      "loss": 7.4277,
+      "step": 80
+    },
+    {
+      "epoch": 0.17802197802197803,
+      "grad_norm": 11.795882225036621,
+      "learning_rate": 0.0017518248175182481,
+      "loss": 7.5547,
+      "step": 81
+    },
+    {
+      "epoch": 0.18021978021978022,
+      "grad_norm": 7.216474533081055,
+      "learning_rate": 0.0017737226277372265,
+      "loss": 6.9551,
+      "step": 82
+    },
+    {
+      "epoch": 0.1824175824175824,
+      "grad_norm": 5.592651844024658,
+      "learning_rate": 0.0017956204379562047,
+      "loss": 7.3438,
+      "step": 83
+    },
+    {
+      "epoch": 0.18461538461538463,
+      "grad_norm": 3.296220302581787,
+      "learning_rate": 0.0018175182481751826,
+      "loss": 7.2324,
+      "step": 84
+    },
+    {
+      "epoch": 0.18681318681318682,
+      "grad_norm": 6.335390090942383,
+      "learning_rate": 0.0018394160583941608,
+      "loss": 7.252,
+      "step": 85
+    },
+    {
+      "epoch": 0.189010989010989,
+      "grad_norm": 22.90656852722168,
+      "learning_rate": 0.0018613138686131387,
+      "loss": 6.9277,
+      "step": 86
+    },
+    {
+      "epoch": 0.1912087912087912,
+      "grad_norm": 3.1930649280548096,
+      "learning_rate": 0.001883211678832117,
+      "loss": 7.0293,
+      "step": 87
+    },
+    {
+      "epoch": 0.1934065934065934,
+      "grad_norm": 3.647556781768799,
+      "learning_rate": 0.0019051094890510949,
+      "loss": 7.4844,
+      "step": 88
+    },
+    {
+      "epoch": 0.1956043956043956,
+      "grad_norm": 6.325287818908691,
+      "learning_rate": 0.001927007299270073,
+      "loss": 7.3516,
+      "step": 89
+    },
+    {
+      "epoch": 0.1978021978021978,
+      "grad_norm": 7.5607123374938965,
+      "learning_rate": 0.001948905109489051,
+      "loss": 7.1885,
+      "step": 90
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.488003730773926,
+      "learning_rate": 0.001970802919708029,
+      "loss": 6.9902,
+      "step": 91
+    },
+    {
+      "epoch": 0.2021978021978022,
+      "grad_norm": 6.433924674987793,
+      "learning_rate": 0.0019927007299270073,
+      "loss": 7.4248,
+      "step": 92
+    },
+    {
+      "epoch": 0.2043956043956044,
+      "grad_norm": 4.302730083465576,
+      "learning_rate": 0.0020145985401459855,
+      "loss": 6.8125,
+      "step": 93
+    },
+    {
+      "epoch": 0.20659340659340658,
+      "grad_norm": 5.370587348937988,
+      "learning_rate": 0.0020364963503649632,
+      "loss": 7.2969,
+      "step": 94
+    },
+    {
+      "epoch": 0.2087912087912088,
+      "grad_norm": 5.8809895515441895,
+      "learning_rate": 0.002058394160583942,
+      "loss": 6.3799,
+      "step": 95
+    },
+    {
+      "epoch": 0.210989010989011,
+      "grad_norm": 7.008445739746094,
+      "learning_rate": 0.00208029197080292,
+      "loss": 6.9043,
+      "step": 96
+    },
+    {
+      "epoch": 0.21318681318681318,
+      "grad_norm": 7.289389133453369,
+      "learning_rate": 0.002102189781021898,
+      "loss": 6.6035,
+      "step": 97
+    },
+    {
+      "epoch": 0.2153846153846154,
+      "grad_norm": 5.653728008270264,
+      "learning_rate": 0.002124087591240876,
+      "loss": 5.5186,
+      "step": 98
+    },
+    {
+      "epoch": 0.2175824175824176,
+      "grad_norm": 6.315726280212402,
+      "learning_rate": 0.002145985401459854,
+      "loss": 6.5156,
+      "step": 99
+    },
+    {
+      "epoch": 0.21978021978021978,
+      "grad_norm": 6.961355686187744,
+      "learning_rate": 0.0021678832116788322,
+      "loss": 5.7285,
+      "step": 100
+    },
+    {
+      "epoch": 0.22197802197802197,
+      "grad_norm": 4.219401836395264,
+      "learning_rate": 0.0021897810218978104,
+      "loss": 6.125,
+      "step": 101
+    },
+    {
+      "epoch": 0.22417582417582418,
+      "grad_norm": 2.284884452819824,
+      "learning_rate": 0.002211678832116788,
+      "loss": 6.1484,
+      "step": 102
+    },
+    {
+      "epoch": 0.22637362637362637,
+      "grad_norm": 3.7214412689208984,
+      "learning_rate": 0.0022335766423357663,
+      "loss": 6.2598,
+      "step": 103
+    },
+    {
+      "epoch": 0.22857142857142856,
+      "grad_norm": 8.243475914001465,
+      "learning_rate": 0.0022554744525547445,
+      "loss": 5.957,
+      "step": 104
+    },
+    {
+      "epoch": 0.23076923076923078,
+      "grad_norm": 67.06355285644531,
+      "learning_rate": 0.0022773722627737226,
+      "loss": 6.2656,
+      "step": 105
+    },
+    {
+      "epoch": 0.23296703296703297,
+      "grad_norm": 4.337093830108643,
+      "learning_rate": 0.002299270072992701,
+      "loss": 5.874,
+      "step": 106
+    },
+    {
+      "epoch": 0.23516483516483516,
+      "grad_norm": 3.00150990486145,
+      "learning_rate": 0.002321167883211679,
+      "loss": 5.8652,
+      "step": 107
+    },
+    {
+      "epoch": 0.23736263736263735,
+      "grad_norm": 4.661133289337158,
+      "learning_rate": 0.002343065693430657,
+      "loss": 5.5146,
+      "step": 108
+    },
+    {
+      "epoch": 0.23956043956043957,
+      "grad_norm": 3.5382633209228516,
+      "learning_rate": 0.0023649635036496353,
+      "loss": 4.6787,
+      "step": 109
+    },
+    {
+      "epoch": 0.24175824175824176,
+      "grad_norm": 2.4006574153900146,
+      "learning_rate": 0.002386861313868613,
+      "loss": 5.1025,
+      "step": 110
+    },
+    {
+      "epoch": 0.24395604395604395,
+      "grad_norm": 2.6122395992279053,
+      "learning_rate": 0.0024087591240875912,
+      "loss": 4.9385,
+      "step": 111
+    },
+    {
+      "epoch": 0.24615384615384617,
+      "grad_norm": 4.556558609008789,
+      "learning_rate": 0.0024306569343065694,
+      "loss": 5.5547,
+      "step": 112
+    },
+    {
+      "epoch": 0.24835164835164836,
+      "grad_norm": 2.098949670791626,
+      "learning_rate": 0.0024525547445255476,
+      "loss": 4.5801,
+      "step": 113
+    },
+    {
+      "epoch": 0.25054945054945055,
+      "grad_norm": 2.0069708824157715,
+      "learning_rate": 0.0024744525547445257,
+      "loss": 4.9141,
+      "step": 114
+    },
+    {
+      "epoch": 0.25274725274725274,
+      "grad_norm": 3.186556816101074,
+      "learning_rate": 0.0024963503649635035,
+      "loss": 4.7881,
+      "step": 115
+    },
+    {
+      "epoch": 0.2549450549450549,
+      "grad_norm": 2.8131697177886963,
+      "learning_rate": 0.0025182481751824816,
+      "loss": 4.9316,
+      "step": 116
+    },
+    {
+      "epoch": 0.2571428571428571,
+      "grad_norm": 17.2800350189209,
+      "learning_rate": 0.00254014598540146,
+      "loss": 4.3604,
+      "step": 117
+    },
+    {
+      "epoch": 0.25934065934065936,
+      "grad_norm": 2.1800217628479004,
+      "learning_rate": 0.002562043795620438,
+      "loss": 4.2949,
+      "step": 118
+    },
+    {
+      "epoch": 0.26153846153846155,
+      "grad_norm": 2.6630020141601562,
+      "learning_rate": 0.002583941605839416,
+      "loss": 4.5039,
+      "step": 119
+    },
+    {
+      "epoch": 0.26373626373626374,
+      "grad_norm": 1.2198642492294312,
+      "learning_rate": 0.0026058394160583943,
+      "loss": 4.3701,
+      "step": 120
+    },
+    {
+      "epoch": 0.26593406593406593,
+      "grad_norm": 1.7308375835418701,
+      "learning_rate": 0.0026277372262773725,
+      "loss": 4.2598,
+      "step": 121
+    },
+    {
+      "epoch": 0.2681318681318681,
+      "grad_norm": 3.3611032962799072,
+      "learning_rate": 0.0026496350364963507,
+      "loss": 4.167,
+      "step": 122
+    },
+    {
+      "epoch": 0.2703296703296703,
+      "grad_norm": 3.4361531734466553,
+      "learning_rate": 0.0026715328467153284,
+      "loss": 4.6631,
+      "step": 123
+    },
+    {
+      "epoch": 0.2725274725274725,
+      "grad_norm": 2.1338741779327393,
+      "learning_rate": 0.0026934306569343066,
+      "loss": 4.1348,
+      "step": 124
+    },
+    {
+      "epoch": 0.27472527472527475,
+      "grad_norm": 1.3278894424438477,
+      "learning_rate": 0.0027153284671532847,
+      "loss": 4.2549,
+      "step": 125
+    },
+    {
+      "epoch": 0.27692307692307694,
+      "grad_norm": 1.5529205799102783,
+      "learning_rate": 0.002737226277372263,
+      "loss": 4.415,
+      "step": 126
+    },
+    {
+      "epoch": 0.27912087912087913,
+      "grad_norm": 3.3766613006591797,
+      "learning_rate": 0.002759124087591241,
+      "loss": 4.1182,
+      "step": 127
+    },
+    {
+      "epoch": 0.2813186813186813,
+      "grad_norm": 1.2177053689956665,
+      "learning_rate": 0.002781021897810219,
+      "loss": 3.8438,
+      "step": 128
+    },
+    {
+      "epoch": 0.2835164835164835,
+      "grad_norm": 1.4136781692504883,
+      "learning_rate": 0.002802919708029197,
+      "loss": 4.0371,
+      "step": 129
+    },
+    {
+      "epoch": 0.2857142857142857,
+      "grad_norm": 1.3534694910049438,
+      "learning_rate": 0.002824817518248175,
+      "loss": 4.0898,
+      "step": 130
+    },
+    {
+      "epoch": 0.2879120879120879,
+      "grad_norm": 0.9128909111022949,
+      "learning_rate": 0.0028467153284671533,
+      "loss": 3.6504,
+      "step": 131
+    },
+    {
+      "epoch": 0.29010989010989013,
+      "grad_norm": 1.8108292818069458,
+      "learning_rate": 0.0028686131386861315,
+      "loss": 3.8223,
+      "step": 132
+    },
+    {
+      "epoch": 0.2923076923076923,
+      "grad_norm": 0.7729389667510986,
+      "learning_rate": 0.0028905109489051097,
+      "loss": 3.4414,
+      "step": 133
+    },
+    {
+      "epoch": 0.2945054945054945,
+      "grad_norm": 1.5539271831512451,
+      "learning_rate": 0.002912408759124088,
+      "loss": 3.2534,
+      "step": 134
+    },
+    {
+      "epoch": 0.2967032967032967,
+      "grad_norm": 2.109830141067505,
+      "learning_rate": 0.002934306569343066,
+      "loss": 3.5684,
+      "step": 135
+    },
+    {
+      "epoch": 0.2989010989010989,
+      "grad_norm": 0.9735444784164429,
+      "learning_rate": 0.0029562043795620437,
+      "loss": 3.6953,
+      "step": 136
+    },
+    {
+      "epoch": 0.3010989010989011,
+      "grad_norm": 2.182887315750122,
+      "learning_rate": 0.002978102189781022,
+      "loss": 3.9834,
+      "step": 137
+    },
+    {
+      "epoch": 0.3032967032967033,
+      "grad_norm": 0.8103097081184387,
+      "learning_rate": 0.003,
+      "loss": 3.2188,
+      "step": 138
+    },
+    {
+      "epoch": 0.3054945054945055,
+      "grad_norm": 2.1591813564300537,
+      "learning_rate": 0.002997557003257329,
+      "loss": 3.2012,
+      "step": 139
+    },
+    {
+      "epoch": 0.3076923076923077,
+      "grad_norm": 1.3161628246307373,
+      "learning_rate": 0.002995114006514658,
+      "loss": 2.9814,
+      "step": 140
+    },
+    {
+      "epoch": 0.3098901098901099,
+      "grad_norm": 1.1639753580093384,
+      "learning_rate": 0.002992671009771987,
+      "loss": 3.3828,
+      "step": 141
+    },
+    {
+      "epoch": 0.3120879120879121,
+      "grad_norm": 0.8245417475700378,
+      "learning_rate": 0.002990228013029316,
+      "loss": 3.0254,
+      "step": 142
+    },
+    {
+      "epoch": 0.3142857142857143,
+      "grad_norm": 1.747742772102356,
+      "learning_rate": 0.002987785016286645,
+      "loss": 2.8037,
+      "step": 143
+    },
+    {
+      "epoch": 0.31648351648351647,
+      "grad_norm": 0.6099988222122192,
+      "learning_rate": 0.002985342019543974,
+      "loss": 2.6646,
+      "step": 144
+    },
+    {
+      "epoch": 0.31868131868131866,
+      "grad_norm": 0.8328338861465454,
+      "learning_rate": 0.002982899022801303,
+      "loss": 2.7026,
+      "step": 145
+    },
+    {
+      "epoch": 0.3208791208791209,
+      "grad_norm": 1.3453677892684937,
+      "learning_rate": 0.002980456026058632,
+      "loss": 2.8062,
+      "step": 146
+    },
+    {
+      "epoch": 0.3230769230769231,
+      "grad_norm": 0.7497466206550598,
+      "learning_rate": 0.002978013029315961,
+      "loss": 2.2327,
+      "step": 147
+    },
+    {
+      "epoch": 0.3252747252747253,
+      "grad_norm": 0.7561153173446655,
+      "learning_rate": 0.00297557003257329,
+      "loss": 2.0457,
+      "step": 148
+    },
+    {
+      "epoch": 0.3274725274725275,
+      "grad_norm": 8.744389533996582,
+      "learning_rate": 0.0029731270358306187,
+      "loss": 2.6938,
+      "step": 149
+    },
+    {
+      "epoch": 0.32967032967032966,
+      "grad_norm": 0.6117238998413086,
+      "learning_rate": 0.002970684039087948,
+      "loss": 2.0093,
+      "step": 150
+    },
+    {
+      "epoch": 0.33186813186813185,
+      "grad_norm": 0.5558815598487854,
+      "learning_rate": 0.002968241042345277,
+      "loss": 1.832,
+      "step": 151
+    },
+    {
+      "epoch": 0.33406593406593404,
+      "grad_norm": 0.7854732275009155,
+      "learning_rate": 0.002965798045602606,
+      "loss": 1.9392,
+      "step": 152
+    },
+    {
+      "epoch": 0.3362637362637363,
+      "grad_norm": 0.6039825081825256,
+      "learning_rate": 0.0029633550488599347,
+      "loss": 2.1001,
+      "step": 153
+    },
+    {
+      "epoch": 0.3384615384615385,
+      "grad_norm": 1.6128307580947876,
+      "learning_rate": 0.002960912052117264,
+      "loss": 2.248,
+      "step": 154
+    },
+    {
+      "epoch": 0.34065934065934067,
+      "grad_norm": 3.6781067848205566,
+      "learning_rate": 0.002958469055374593,
+      "loss": 1.916,
+      "step": 155
+    },
+    {
+      "epoch": 0.34285714285714286,
+      "grad_norm": 1.9110193252563477,
+      "learning_rate": 0.002956026058631922,
+      "loss": 1.9104,
+      "step": 156
+    },
+    {
+      "epoch": 0.34505494505494505,
+      "grad_norm": 3.1486947536468506,
+      "learning_rate": 0.0029535830618892507,
+      "loss": 1.8477,
+      "step": 157
+    },
+    {
+      "epoch": 0.34725274725274724,
+      "grad_norm": 0.7627822160720825,
+      "learning_rate": 0.0029511400651465796,
+      "loss": 1.5469,
+      "step": 158
+    },
+    {
+      "epoch": 0.34945054945054943,
+      "grad_norm": 2.991727828979492,
+      "learning_rate": 0.002948697068403909,
+      "loss": 1.7896,
+      "step": 159
+    },
+    {
+      "epoch": 0.3516483516483517,
+      "grad_norm": 1.336944341659546,
+      "learning_rate": 0.002946254071661238,
+      "loss": 1.9343,
+      "step": 160
+    },
+    {
+      "epoch": 0.35384615384615387,
+      "grad_norm": 0.6944097876548767,
+      "learning_rate": 0.0029438110749185667,
+      "loss": 1.5403,
+      "step": 161
+    },
+    {
+      "epoch": 0.35604395604395606,
+      "grad_norm": 9.288591384887695,
+      "learning_rate": 0.0029413680781758956,
+      "loss": 1.6082,
+      "step": 162
+    },
+    {
+      "epoch": 0.35824175824175825,
+      "grad_norm": 0.5690299868583679,
+      "learning_rate": 0.002938925081433225,
+      "loss": 1.6816,
+      "step": 163
+    },
+    {
+      "epoch": 0.36043956043956044,
+      "grad_norm": 0.4263255000114441,
+      "learning_rate": 0.002936482084690554,
+      "loss": 1.4851,
+      "step": 164
+    },
+    {
+      "epoch": 0.3626373626373626,
+      "grad_norm": 0.38679444789886475,
+      "learning_rate": 0.0029340390879478827,
+      "loss": 1.2087,
+      "step": 165
+    },
+    {
+      "epoch": 0.3648351648351648,
+      "grad_norm": 0.46679607033729553,
+      "learning_rate": 0.0029315960912052116,
+      "loss": 1.521,
+      "step": 166
+    },
+    {
+      "epoch": 0.367032967032967,
+      "grad_norm": 0.582525908946991,
+      "learning_rate": 0.0029291530944625405,
+      "loss": 1.3997,
+      "step": 167
+    },
+    {
+      "epoch": 0.36923076923076925,
+      "grad_norm": 0.3757668137550354,
+      "learning_rate": 0.00292671009771987,
+      "loss": 1.2198,
+      "step": 168
+    },
+    {
+      "epoch": 0.37142857142857144,
+      "grad_norm": 0.26723504066467285,
+      "learning_rate": 0.0029242671009771988,
+      "loss": 1.0557,
+      "step": 169
+    },
+    {
+      "epoch": 0.37362637362637363,
+      "grad_norm": 0.37234529852867126,
+      "learning_rate": 0.0029218241042345276,
+      "loss": 1.0786,
+      "step": 170
+    },
+    {
+      "epoch": 0.3758241758241758,
+      "grad_norm": 0.371770977973938,
+      "learning_rate": 0.0029193811074918565,
+      "loss": 1.2463,
+      "step": 171
+    },
+    {
+      "epoch": 0.378021978021978,
+      "grad_norm": 1.8547030687332153,
+      "learning_rate": 0.002916938110749186,
+      "loss": 1.0751,
+      "step": 172
+    },
+    {
+      "epoch": 0.3802197802197802,
+      "grad_norm": 0.4806371033191681,
+      "learning_rate": 0.0029144951140065148,
+      "loss": 1.0597,
+      "step": 173
+    },
+    {
+      "epoch": 0.3824175824175824,
+      "grad_norm": 0.3849234879016876,
+      "learning_rate": 0.0029120521172638436,
+      "loss": 1.3528,
+      "step": 174
+    },
+    {
+      "epoch": 0.38461538461538464,
+      "grad_norm": 0.34507012367248535,
+      "learning_rate": 0.0029096091205211725,
+      "loss": 1.0276,
+      "step": 175
+    },
+    {
+      "epoch": 0.3868131868131868,
+      "grad_norm": 1.3840712308883667,
+      "learning_rate": 0.002907166123778502,
+      "loss": 1.3585,
+      "step": 176
+    },
+    {
+      "epoch": 0.389010989010989,
+      "grad_norm": 0.5014913082122803,
+      "learning_rate": 0.0029047231270358308,
+      "loss": 1.3337,
+      "step": 177
+    },
+    {
+      "epoch": 0.3912087912087912,
+      "grad_norm": 0.2523179352283478,
+      "learning_rate": 0.0029022801302931596,
+      "loss": 1.1243,
+      "step": 178
+    },
+    {
+      "epoch": 0.3934065934065934,
+      "grad_norm": 0.21229645609855652,
+      "learning_rate": 0.0028998371335504885,
+      "loss": 1.0381,
+      "step": 179
+    },
+    {
+      "epoch": 0.3956043956043956,
+      "grad_norm": 0.261942982673645,
+      "learning_rate": 0.0028973941368078174,
+      "loss": 1.3035,
+      "step": 180
+    },
+    {
+      "epoch": 0.3978021978021978,
+      "grad_norm": 1.0811759233474731,
+      "learning_rate": 0.0028949511400651468,
+      "loss": 1.2678,
+      "step": 181
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.573317050933838,
+      "learning_rate": 0.0028925081433224756,
+      "loss": 1.2937,
+      "step": 182
+    },
+    {
+      "epoch": 0.4021978021978022,
+      "grad_norm": 0.12203257530927658,
+      "learning_rate": 0.0028900651465798045,
+      "loss": 0.8251,
+      "step": 183
+    },
+    {
+      "epoch": 0.4043956043956044,
+      "grad_norm": 0.3833327293395996,
+      "learning_rate": 0.0028876221498371334,
+      "loss": 1.1165,
+      "step": 184
+    },
+    {
+      "epoch": 0.4065934065934066,
+      "grad_norm": 6.216048717498779,
+      "learning_rate": 0.0028851791530944628,
+      "loss": 1.4646,
+      "step": 185
+    },
+    {
+      "epoch": 0.4087912087912088,
+      "grad_norm": 0.17099091410636902,
+      "learning_rate": 0.0028827361563517916,
+      "loss": 1.0264,
+      "step": 186
+    },
+    {
+      "epoch": 0.41098901098901097,
+      "grad_norm": 0.17816264927387238,
+      "learning_rate": 0.0028802931596091205,
+      "loss": 1.0166,
+      "step": 187
+    },
+    {
+      "epoch": 0.41318681318681316,
+      "grad_norm": 0.24949267506599426,
+      "learning_rate": 0.0028778501628664494,
+      "loss": 1.269,
+      "step": 188
+    },
+    {
+      "epoch": 0.4153846153846154,
+      "grad_norm": 0.13837742805480957,
+      "learning_rate": 0.0028754071661237783,
+      "loss": 1.0591,
+      "step": 189
+    },
+    {
+      "epoch": 0.4175824175824176,
+      "grad_norm": 0.18339362740516663,
+      "learning_rate": 0.0028729641693811077,
+      "loss": 1.126,
+      "step": 190
+    },
+    {
+      "epoch": 0.4197802197802198,
+      "grad_norm": 12.316235542297363,
+      "learning_rate": 0.0028705211726384365,
+      "loss": 1.6464,
+      "step": 191
+    },
+    {
+      "epoch": 0.421978021978022,
+      "grad_norm": 1.7392607927322388,
+      "learning_rate": 0.0028680781758957654,
+      "loss": 1.7419,
+      "step": 192
+    },
+    {
+      "epoch": 0.42417582417582417,
+      "grad_norm": 13.854021072387695,
+      "learning_rate": 0.0028656351791530943,
+      "loss": 3.8933,
+      "step": 193
+    },
+    {
+      "epoch": 0.42637362637362636,
+      "grad_norm": 292.6780700683594,
+      "learning_rate": 0.0028631921824104237,
+      "loss": 3.8348,
+      "step": 194
+    },
+    {
+      "epoch": 0.42857142857142855,
+      "grad_norm": 9.216120719909668,
+      "learning_rate": 0.0028607491856677525,
+      "loss": 4.0889,
+      "step": 195
+    },
+    {
+      "epoch": 0.4307692307692308,
+      "grad_norm": 22.34186553955078,
+      "learning_rate": 0.0028583061889250814,
+      "loss": 6.5884,
+      "step": 196
+    },
+    {
+      "epoch": 0.432967032967033,
+      "grad_norm": 137.80914306640625,
+      "learning_rate": 0.0028558631921824103,
+      "loss": 6.7974,
+      "step": 197
+    },
+    {
+      "epoch": 0.4351648351648352,
+      "grad_norm": 32.77785110473633,
+      "learning_rate": 0.0028534201954397392,
+      "loss": 4.1812,
+      "step": 198
+    },
+    {
+      "epoch": 0.43736263736263736,
+      "grad_norm": 64.56425476074219,
+      "learning_rate": 0.0028509771986970685,
+      "loss": 5.3972,
+      "step": 199
+    },
+    {
+      "epoch": 0.43956043956043955,
+      "grad_norm": 21.897268295288086,
+      "learning_rate": 0.0028485342019543974,
+      "loss": 5.0261,
+      "step": 200
+    },
+    {
+      "epoch": 0.44175824175824174,
+      "grad_norm": 2.520737886428833,
+      "learning_rate": 0.0028460912052117263,
+      "loss": 2.8824,
+      "step": 201
+    },
+    {
+      "epoch": 0.44395604395604393,
+      "grad_norm": 3.596407413482666,
+      "learning_rate": 0.0028436482084690552,
+      "loss": 3.2977,
+      "step": 202
+    },
+    {
+      "epoch": 0.4461538461538462,
+      "grad_norm": 26.778114318847656,
+      "learning_rate": 0.0028412052117263845,
+      "loss": 2.1417,
+      "step": 203
+    },
+    {
+      "epoch": 0.44835164835164837,
+      "grad_norm": 5.726716995239258,
+      "learning_rate": 0.0028387622149837134,
+      "loss": 1.6631,
+      "step": 204
+    },
+    {
+      "epoch": 0.45054945054945056,
+      "grad_norm": 0.23155227303504944,
+      "learning_rate": 0.0028363192182410423,
+      "loss": 0.9664,
+      "step": 205
+    },
+    {
+      "epoch": 0.45274725274725275,
+      "grad_norm": 120.6667709350586,
+      "learning_rate": 0.0028338762214983712,
+      "loss": 0.8973,
+      "step": 206
+    },
+    {
+      "epoch": 0.45494505494505494,
+      "grad_norm": 0.13380751013755798,
+      "learning_rate": 0.0028314332247557,
+      "loss": 1.0208,
+      "step": 207
+    },
+    {
+      "epoch": 0.45714285714285713,
+      "grad_norm": 0.1634182184934616,
+      "learning_rate": 0.0028289902280130294,
+      "loss": 0.902,
+      "step": 208
+    },
+    {
+      "epoch": 0.4593406593406593,
+      "grad_norm": 0.641819417476654,
+      "learning_rate": 0.0028265472312703583,
+      "loss": 1.01,
+      "step": 209
+    },
+    {
+      "epoch": 0.46153846153846156,
+      "grad_norm": 0.24499137699604034,
+      "learning_rate": 0.0028241042345276872,
+      "loss": 0.9456,
+      "step": 210
+    },
+    {
+      "epoch": 0.46373626373626375,
+      "grad_norm": 0.13139550387859344,
+      "learning_rate": 0.002821661237785016,
+      "loss": 0.8563,
+      "step": 211
+    },
+    {
+      "epoch": 0.46593406593406594,
+      "grad_norm": 1.0476970672607422,
+      "learning_rate": 0.0028192182410423454,
+      "loss": 1.0006,
+      "step": 212
+    },
+    {
+      "epoch": 0.46813186813186813,
+      "grad_norm": 0.11675205081701279,
+      "learning_rate": 0.0028167752442996743,
+      "loss": 0.8749,
+      "step": 213
+    },
+    {
+      "epoch": 0.4703296703296703,
+      "grad_norm": 0.16503529250621796,
+      "learning_rate": 0.0028143322475570032,
+      "loss": 0.9617,
+      "step": 214
+    },
+    {
+      "epoch": 0.4725274725274725,
+      "grad_norm": 0.4273809790611267,
+      "learning_rate": 0.002811889250814332,
+      "loss": 1.0109,
+      "step": 215
+    },
+    {
+      "epoch": 0.4747252747252747,
+      "grad_norm": 0.15598806738853455,
+      "learning_rate": 0.002809446254071661,
+      "loss": 0.9204,
+      "step": 216
+    },
+    {
+      "epoch": 0.47692307692307695,
+      "grad_norm": 0.09496546536684036,
+      "learning_rate": 0.0028070032573289903,
+      "loss": 0.8389,
+      "step": 217
+    },
+    {
+      "epoch": 0.47912087912087914,
+      "grad_norm": 0.19012843072414398,
+      "learning_rate": 0.0028045602605863192,
+      "loss": 0.7574,
+      "step": 218
+    },
+    {
+      "epoch": 0.48131868131868133,
+      "grad_norm": 0.13629943132400513,
+      "learning_rate": 0.002802117263843648,
+      "loss": 0.7666,
+      "step": 219
+    },
+    {
+      "epoch": 0.4835164835164835,
+      "grad_norm": 0.29692134261131287,
+      "learning_rate": 0.002799674267100977,
+      "loss": 1.0735,
+      "step": 220
+    },
+    {
+      "epoch": 0.4857142857142857,
+      "grad_norm": 0.08060973882675171,
+      "learning_rate": 0.0027972312703583063,
+      "loss": 1.0283,
+      "step": 221
+    },
+    {
+      "epoch": 0.4879120879120879,
+      "grad_norm": 0.10856381058692932,
+      "learning_rate": 0.0027947882736156352,
+      "loss": 0.9355,
+      "step": 222
+    },
+    {
+      "epoch": 0.4901098901098901,
+      "grad_norm": 0.10103047639131546,
+      "learning_rate": 0.002792345276872964,
+      "loss": 0.7602,
+      "step": 223
+    },
+    {
+      "epoch": 0.49230769230769234,
+      "grad_norm": 0.1244843378663063,
+      "learning_rate": 0.002789902280130293,
+      "loss": 0.9337,
+      "step": 224
+    },
+    {
+      "epoch": 0.4945054945054945,
+      "grad_norm": 1.0263439416885376,
+      "learning_rate": 0.002787459283387622,
+      "loss": 1.3448,
+      "step": 225
+    },
+    {
+      "epoch": 0.4967032967032967,
+      "grad_norm": 0.12465494126081467,
+      "learning_rate": 0.0027850162866449512,
+      "loss": 0.9543,
+      "step": 226
+    },
+    {
+      "epoch": 0.4989010989010989,
+      "grad_norm": 0.1810968518257141,
+      "learning_rate": 0.00278257328990228,
+      "loss": 1.2026,
+      "step": 227
+    },
+    {
+      "epoch": 0.5010989010989011,
+      "grad_norm": 0.5792099237442017,
+      "learning_rate": 0.002780130293159609,
+      "loss": 1.0861,
+      "step": 228
+    },
+    {
+      "epoch": 0.5032967032967033,
+      "grad_norm": 0.5234378576278687,
+      "learning_rate": 0.002777687296416938,
+      "loss": 0.8916,
+      "step": 229
+    },
+    {
+      "epoch": 0.5054945054945055,
+      "grad_norm": 0.1531374305486679,
+      "learning_rate": 0.0027752442996742672,
+      "loss": 0.9263,
+      "step": 230
+    },
+    {
+      "epoch": 0.5076923076923077,
+      "grad_norm": 0.051747970283031464,
+      "learning_rate": 0.002772801302931596,
+      "loss": 0.8783,
+      "step": 231
+    },
+    {
+      "epoch": 0.5098901098901099,
+      "grad_norm": 0.15211984515190125,
+      "learning_rate": 0.002770358306188925,
+      "loss": 0.9636,
+      "step": 232
+    },
+    {
+      "epoch": 0.512087912087912,
+      "grad_norm": 0.6442863941192627,
+      "learning_rate": 0.002767915309446254,
+      "loss": 0.9253,
+      "step": 233
+    },
+    {
+      "epoch": 0.5142857142857142,
+      "grad_norm": 0.14286652207374573,
+      "learning_rate": 0.002765472312703583,
+      "loss": 0.8801,
+      "step": 234
+    },
+    {
+      "epoch": 0.5164835164835165,
+      "grad_norm": 0.18937940895557404,
+      "learning_rate": 0.002763029315960912,
+      "loss": 1.0144,
+      "step": 235
+    },
+    {
+      "epoch": 0.5186813186813187,
+      "grad_norm": 0.09786862134933472,
+      "learning_rate": 0.002760586319218241,
+      "loss": 0.8003,
+      "step": 236
+    },
+    {
+      "epoch": 0.5208791208791209,
+      "grad_norm": 0.1945120096206665,
+      "learning_rate": 0.00275814332247557,
+      "loss": 1.0186,
+      "step": 237
+    },
+    {
+      "epoch": 0.5230769230769231,
+      "grad_norm": 0.08694169670343399,
+      "learning_rate": 0.002755700325732899,
+      "loss": 0.8834,
+      "step": 238
+    },
+    {
+      "epoch": 0.5252747252747253,
+      "grad_norm": 0.09607340395450592,
+      "learning_rate": 0.002753257328990228,
+      "loss": 0.8754,
+      "step": 239
+    },
+    {
+      "epoch": 0.5274725274725275,
+      "grad_norm": 0.479829341173172,
+      "learning_rate": 0.002750814332247557,
+      "loss": 0.9479,
+      "step": 240
+    },
+    {
+      "epoch": 0.5296703296703297,
+      "grad_norm": 0.10453948378562927,
+      "learning_rate": 0.002748371335504886,
+      "loss": 0.8915,
+      "step": 241
+    },
+    {
+      "epoch": 0.5318681318681319,
+      "grad_norm": 0.21772126853466034,
+      "learning_rate": 0.002745928338762215,
+      "loss": 0.9243,
+      "step": 242
+    },
+    {
+      "epoch": 0.5340659340659341,
+      "grad_norm": 0.06437736749649048,
+      "learning_rate": 0.002743485342019544,
+      "loss": 0.9042,
+      "step": 243
+    },
+    {
+      "epoch": 0.5362637362637362,
+      "grad_norm": 0.08140173554420471,
+      "learning_rate": 0.002741042345276873,
+      "loss": 0.894,
+      "step": 244
+    },
+    {
+      "epoch": 0.5384615384615384,
+      "grad_norm": 0.08294742554426193,
+      "learning_rate": 0.002738599348534202,
+      "loss": 0.8397,
+      "step": 245
+    },
+    {
+      "epoch": 0.5406593406593406,
+      "grad_norm": 4.109051704406738,
+      "learning_rate": 0.002736156351791531,
+      "loss": 1.3635,
+      "step": 246
+    },
+    {
+      "epoch": 0.5428571428571428,
+      "grad_norm": 0.07352132350206375,
+      "learning_rate": 0.0027337133550488597,
+      "loss": 0.8215,
+      "step": 247
+    },
+    {
+      "epoch": 0.545054945054945,
+      "grad_norm": 0.10033997893333435,
+      "learning_rate": 0.002731270358306189,
+      "loss": 0.9209,
+      "step": 248
+    },
+    {
+      "epoch": 0.5472527472527473,
+      "grad_norm": 0.053501419723033905,
+      "learning_rate": 0.002728827361563518,
+      "loss": 0.7498,
+      "step": 249
+    },
+    {
+      "epoch": 0.5494505494505495,
+      "grad_norm": 0.08118750154972076,
+      "learning_rate": 0.002726384364820847,
+      "loss": 1.0005,
+      "step": 250
+    },
+    {
+      "epoch": 0.5516483516483517,
+      "grad_norm": 0.23336473107337952,
+      "learning_rate": 0.0027239413680781757,
+      "loss": 0.9534,
+      "step": 251
+    },
+    {
+      "epoch": 0.5538461538461539,
+      "grad_norm": 0.20108911395072937,
+      "learning_rate": 0.002721498371335505,
+      "loss": 0.9565,
+      "step": 252
+    },
+    {
+      "epoch": 0.5560439560439561,
+      "grad_norm": 0.09861844033002853,
+      "learning_rate": 0.002719055374592834,
+      "loss": 0.6951,
+      "step": 253
+    },
+    {
+      "epoch": 0.5582417582417583,
+      "grad_norm": 0.08034469187259674,
+      "learning_rate": 0.002716612377850163,
+      "loss": 0.9739,
+      "step": 254
+    },
+    {
+      "epoch": 0.5604395604395604,
+      "grad_norm": 0.056816838681697845,
+      "learning_rate": 0.0027141693811074917,
+      "loss": 0.8843,
+      "step": 255
+    },
+    {
+      "epoch": 0.5626373626373626,
+      "grad_norm": 0.0656348466873169,
+      "learning_rate": 0.0027117263843648206,
+      "loss": 0.9028,
+      "step": 256
+    },
+    {
+      "epoch": 0.5648351648351648,
+      "grad_norm": 0.1130959764122963,
+      "learning_rate": 0.00270928338762215,
+      "loss": 0.8822,
+      "step": 257
+    },
+    {
+      "epoch": 0.567032967032967,
+      "grad_norm": 0.07569181174039841,
+      "learning_rate": 0.002706840390879479,
+      "loss": 0.9138,
+      "step": 258
+    },
+    {
+      "epoch": 0.5692307692307692,
+      "grad_norm": 0.0416770838201046,
+      "learning_rate": 0.0027043973941368077,
+      "loss": 0.8055,
+      "step": 259
+    },
+    {
+      "epoch": 0.5714285714285714,
+      "grad_norm": 0.1100807785987854,
+      "learning_rate": 0.0027019543973941366,
+      "loss": 1.0137,
+      "step": 260
+    },
+    {
+      "epoch": 0.5736263736263736,
+      "grad_norm": 0.1475360095500946,
+      "learning_rate": 0.002699511400651466,
+      "loss": 1.1345,
+      "step": 261
+    },
+    {
+      "epoch": 0.5758241758241758,
+      "grad_norm": 0.09156166762113571,
+      "learning_rate": 0.002697068403908795,
+      "loss": 0.9169,
+      "step": 262
+    },
+    {
+      "epoch": 0.578021978021978,
+      "grad_norm": 0.08868283778429031,
+      "learning_rate": 0.0026946254071661237,
+      "loss": 0.912,
+      "step": 263
+    },
+    {
+      "epoch": 0.5802197802197803,
+      "grad_norm": 0.0911644920706749,
+      "learning_rate": 0.0026921824104234526,
+      "loss": 1.0098,
+      "step": 264
+    },
+    {
+      "epoch": 0.5824175824175825,
+      "grad_norm": 0.1400439739227295,
+      "learning_rate": 0.0026897394136807815,
+      "loss": 0.9406,
+      "step": 265
+    },
+    {
+      "epoch": 0.5846153846153846,
+      "grad_norm": 0.2803069055080414,
+      "learning_rate": 0.002687296416938111,
+      "loss": 0.9487,
+      "step": 266
+    },
+    {
+      "epoch": 0.5868131868131868,
+      "grad_norm": 0.08409854769706726,
+      "learning_rate": 0.0026848534201954397,
+      "loss": 0.7949,
+      "step": 267
+    },
+    {
+      "epoch": 0.589010989010989,
+      "grad_norm": 1.8260517120361328,
+      "learning_rate": 0.0026824104234527686,
+      "loss": 0.8214,
+      "step": 268
+    },
+    {
+      "epoch": 0.5912087912087912,
+      "grad_norm": 0.11200416088104248,
+      "learning_rate": 0.0026799674267100975,
+      "loss": 1.1094,
+      "step": 269
+    },
+    {
+      "epoch": 0.5934065934065934,
+      "grad_norm": 0.26965561509132385,
+      "learning_rate": 0.002677524429967427,
+      "loss": 0.7644,
+      "step": 270
+    },
+    {
+      "epoch": 0.5956043956043956,
+      "grad_norm": 0.052464816719293594,
+      "learning_rate": 0.0026750814332247557,
+      "loss": 0.7999,
+      "step": 271
+    },
+    {
+      "epoch": 0.5978021978021978,
+      "grad_norm": 0.13794219493865967,
+      "learning_rate": 0.0026726384364820846,
+      "loss": 0.9431,
+      "step": 272
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.10606757551431656,
+      "learning_rate": 0.0026701954397394135,
+      "loss": 1.376,
+      "step": 273
+    },
+    {
+      "epoch": 0.6021978021978022,
+      "grad_norm": 0.0592002309858799,
+      "learning_rate": 0.0026677524429967424,
+      "loss": 0.8944,
+      "step": 274
+    },
+    {
+      "epoch": 0.6043956043956044,
+      "grad_norm": 0.06424269825220108,
+      "learning_rate": 0.0026653094462540717,
+      "loss": 1.0527,
+      "step": 275
+    },
+    {
+      "epoch": 0.6065934065934065,
+      "grad_norm": 0.12087871134281158,
+      "learning_rate": 0.0026628664495114006,
+      "loss": 0.9998,
+      "step": 276
+    },
+    {
+      "epoch": 0.6087912087912087,
+      "grad_norm": 0.04757562279701233,
+      "learning_rate": 0.0026604234527687295,
+      "loss": 0.8324,
+      "step": 277
+    },
+    {
+      "epoch": 0.610989010989011,
+      "grad_norm": 0.061450786888599396,
+      "learning_rate": 0.0026579804560260584,
+      "loss": 0.9735,
+      "step": 278
+    },
+    {
+      "epoch": 0.6131868131868132,
+      "grad_norm": 0.08181532472372055,
+      "learning_rate": 0.0026555374592833877,
+      "loss": 1.0676,
+      "step": 279
+    },
+    {
+      "epoch": 0.6153846153846154,
+      "grad_norm": 0.04907115548849106,
+      "learning_rate": 0.0026530944625407166,
+      "loss": 0.8545,
+      "step": 280
+    },
+    {
+      "epoch": 0.6175824175824176,
+      "grad_norm": 0.05714426934719086,
+      "learning_rate": 0.0026506514657980455,
+      "loss": 0.8335,
+      "step": 281
+    },
+    {
+      "epoch": 0.6197802197802198,
+      "grad_norm": 0.13850869238376617,
+      "learning_rate": 0.0026482084690553744,
+      "loss": 0.9478,
+      "step": 282
+    },
+    {
+      "epoch": 0.621978021978022,
+      "grad_norm": 0.08520478010177612,
+      "learning_rate": 0.0026457654723127033,
+      "loss": 1.0957,
+      "step": 283
+    },
+    {
+      "epoch": 0.6241758241758242,
+      "grad_norm": 0.05593477934598923,
+      "learning_rate": 0.0026433224755700326,
+      "loss": 0.8075,
+      "step": 284
+    },
+    {
+      "epoch": 0.6263736263736264,
+      "grad_norm": 0.05911228433251381,
+      "learning_rate": 0.0026408794788273615,
+      "loss": 0.9491,
+      "step": 285
+    },
+    {
+      "epoch": 0.6285714285714286,
+      "grad_norm": 0.05830894410610199,
+      "learning_rate": 0.0026384364820846904,
+      "loss": 0.8815,
+      "step": 286
+    },
+    {
+      "epoch": 0.6307692307692307,
+      "grad_norm": 0.08350221067667007,
+      "learning_rate": 0.0026359934853420193,
+      "loss": 0.8912,
+      "step": 287
+    },
+    {
+      "epoch": 0.6329670329670329,
+      "grad_norm": 0.07531853765249252,
+      "learning_rate": 0.0026335504885993486,
+      "loss": 0.8536,
+      "step": 288
+    },
+    {
+      "epoch": 0.6351648351648351,
+      "grad_norm": 0.055327270179986954,
+      "learning_rate": 0.0026311074918566775,
+      "loss": 0.9855,
+      "step": 289
+    },
+    {
+      "epoch": 0.6373626373626373,
+      "grad_norm": 0.04534522071480751,
+      "learning_rate": 0.0026286644951140064,
+      "loss": 1.0286,
+      "step": 290
+    },
+    {
+      "epoch": 0.6395604395604395,
+      "grad_norm": 0.057625748217105865,
+      "learning_rate": 0.0026262214983713353,
+      "loss": 0.8945,
+      "step": 291
+    },
+    {
+      "epoch": 0.6417582417582418,
+      "grad_norm": 0.04673678055405617,
+      "learning_rate": 0.0026237785016286646,
+      "loss": 0.6147,
+      "step": 292
+    },
+    {
+      "epoch": 0.643956043956044,
+      "grad_norm": 2.661560535430908,
+      "learning_rate": 0.0026213355048859935,
+      "loss": 1.2194,
+      "step": 293
+    },
+    {
+      "epoch": 0.6461538461538462,
+      "grad_norm": 0.06859428435564041,
+      "learning_rate": 0.002618892508143323,
+      "loss": 0.9768,
+      "step": 294
+    },
+    {
+      "epoch": 0.6483516483516484,
+      "grad_norm": 0.07032730430364609,
+      "learning_rate": 0.0026164495114006517,
+      "loss": 0.8418,
+      "step": 295
+    },
+    {
+      "epoch": 0.6505494505494506,
+      "grad_norm": 0.08709180355072021,
+      "learning_rate": 0.0026140065146579806,
+      "loss": 0.7837,
+      "step": 296
+    },
+    {
+      "epoch": 0.6527472527472528,
+      "grad_norm": 0.06401631981134415,
+      "learning_rate": 0.0026115635179153095,
+      "loss": 0.9283,
+      "step": 297
+    },
+    {
+      "epoch": 0.654945054945055,
+      "grad_norm": 0.08315717428922653,
+      "learning_rate": 0.002609120521172639,
+      "loss": 0.896,
+      "step": 298
+    },
+    {
+      "epoch": 0.6571428571428571,
+      "grad_norm": 0.06237190589308739,
+      "learning_rate": 0.0026066775244299677,
+      "loss": 0.856,
+      "step": 299
+    },
+    {
+      "epoch": 0.6593406593406593,
+      "grad_norm": 0.04170721769332886,
+      "learning_rate": 0.0026042345276872966,
+      "loss": 0.9714,
+      "step": 300
+    },
+    {
+      "epoch": 0.6615384615384615,
+      "grad_norm": 0.17676624655723572,
+      "learning_rate": 0.0026017915309446255,
+      "loss": 0.8474,
+      "step": 301
+    },
+    {
+      "epoch": 0.6637362637362637,
+      "grad_norm": 0.09068896621465683,
+      "learning_rate": 0.0025993485342019544,
+      "loss": 0.8967,
+      "step": 302
+    },
+    {
+      "epoch": 0.6659340659340659,
+      "grad_norm": 0.05594249442219734,
+      "learning_rate": 0.0025969055374592837,
+      "loss": 0.9972,
+      "step": 303
+    },
+    {
+      "epoch": 0.6681318681318681,
+      "grad_norm": 0.07632527500391006,
+      "learning_rate": 0.0025944625407166126,
+      "loss": 0.8208,
+      "step": 304
+    },
+    {
+      "epoch": 0.6703296703296703,
+      "grad_norm": 0.473337858915329,
+      "learning_rate": 0.0025920195439739415,
+      "loss": 1.0291,
+      "step": 305
+    },
+    {
+      "epoch": 0.6725274725274726,
+      "grad_norm": 0.054862577468156815,
+      "learning_rate": 0.0025895765472312704,
+      "loss": 0.8691,
+      "step": 306
+    },
+    {
+      "epoch": 0.6747252747252748,
+      "grad_norm": 0.06608554720878601,
+      "learning_rate": 0.0025871335504885997,
+      "loss": 0.8525,
+      "step": 307
+    },
+    {
+      "epoch": 0.676923076923077,
+      "grad_norm": 0.07392515987157822,
+      "learning_rate": 0.0025846905537459286,
+      "loss": 0.7842,
+      "step": 308
+    },
+    {
+      "epoch": 0.6791208791208792,
+      "grad_norm": 0.09137177467346191,
+      "learning_rate": 0.0025822475570032575,
+      "loss": 0.8604,
+      "step": 309
+    },
+    {
+      "epoch": 0.6813186813186813,
+      "grad_norm": 0.05838591977953911,
+      "learning_rate": 0.0025798045602605864,
+      "loss": 0.9864,
+      "step": 310
+    },
+    {
+      "epoch": 0.6835164835164835,
+      "grad_norm": 0.07160185277462006,
+      "learning_rate": 0.0025773615635179153,
+      "loss": 0.8982,
+      "step": 311
+    },
+    {
+      "epoch": 0.6857142857142857,
+      "grad_norm": 0.06732136011123657,
+      "learning_rate": 0.0025749185667752446,
+      "loss": 0.9707,
+      "step": 312
+    },
+    {
+      "epoch": 0.6879120879120879,
+      "grad_norm": 0.4440630376338959,
+      "learning_rate": 0.0025724755700325735,
+      "loss": 1.0743,
+      "step": 313
+    },
+    {
+      "epoch": 0.6901098901098901,
+      "grad_norm": 0.16018560528755188,
+      "learning_rate": 0.0025700325732899024,
+      "loss": 0.9472,
+      "step": 314
+    },
+    {
+      "epoch": 0.6923076923076923,
+      "grad_norm": 0.120985247194767,
+      "learning_rate": 0.0025675895765472313,
+      "loss": 0.8809,
+      "step": 315
+    },
+    {
+      "epoch": 0.6945054945054945,
+      "grad_norm": 0.06684579700231552,
+      "learning_rate": 0.0025651465798045606,
+      "loss": 1.1101,
+      "step": 316
+    },
+    {
+      "epoch": 0.6967032967032967,
+      "grad_norm": 0.07874082028865814,
+      "learning_rate": 0.0025627035830618895,
+      "loss": 0.9469,
+      "step": 317
+    },
+    {
+      "epoch": 0.6989010989010989,
+      "grad_norm": 0.0532657653093338,
+      "learning_rate": 0.0025602605863192184,
+      "loss": 0.8217,
+      "step": 318
+    },
+    {
+      "epoch": 0.701098901098901,
+      "grad_norm": 0.05162207409739494,
+      "learning_rate": 0.0025578175895765473,
+      "loss": 1.1335,
+      "step": 319
+    },
+    {
+      "epoch": 0.7032967032967034,
+      "grad_norm": 0.1402101069688797,
+      "learning_rate": 0.0025553745928338766,
+      "loss": 0.8317,
+      "step": 320
+    },
+    {
+      "epoch": 0.7054945054945055,
+      "grad_norm": 0.14360734820365906,
+      "learning_rate": 0.0025529315960912055,
+      "loss": 0.9983,
+      "step": 321
+    },
+    {
+      "epoch": 0.7076923076923077,
+      "grad_norm": 0.04879635199904442,
+      "learning_rate": 0.0025504885993485344,
+      "loss": 0.7789,
+      "step": 322
+    },
+    {
+      "epoch": 0.7098901098901099,
+      "grad_norm": 0.10033449530601501,
+      "learning_rate": 0.0025480456026058633,
+      "loss": 1.0121,
+      "step": 323
+    },
+    {
+      "epoch": 0.7120879120879121,
+      "grad_norm": 0.18021267652511597,
+      "learning_rate": 0.002545602605863192,
+      "loss": 0.9443,
+      "step": 324
+    },
+    {
+      "epoch": 0.7142857142857143,
+      "grad_norm": 0.0704277977347374,
+      "learning_rate": 0.0025431596091205215,
+      "loss": 0.8751,
+      "step": 325
+    },
+    {
+      "epoch": 0.7164835164835165,
+      "grad_norm": 0.07012510299682617,
+      "learning_rate": 0.0025407166123778504,
+      "loss": 0.8144,
+      "step": 326
+    },
+    {
+      "epoch": 0.7186813186813187,
+      "grad_norm": 0.09309165924787521,
+      "learning_rate": 0.0025382736156351793,
+      "loss": 0.8489,
+      "step": 327
+    },
+    {
+      "epoch": 0.7208791208791209,
+      "grad_norm": 0.02774563990533352,
+      "learning_rate": 0.002535830618892508,
+      "loss": 0.8301,
+      "step": 328
+    },
+    {
+      "epoch": 0.7230769230769231,
+      "grad_norm": 0.0837404802441597,
+      "learning_rate": 0.0025333876221498375,
+      "loss": 0.8846,
+      "step": 329
+    },
+    {
+      "epoch": 0.7252747252747253,
+      "grad_norm": 0.033590421080589294,
+      "learning_rate": 0.0025309446254071664,
+      "loss": 0.996,
+      "step": 330
+    },
+    {
+      "epoch": 0.7274725274725274,
+      "grad_norm": 0.15004494786262512,
+      "learning_rate": 0.0025285016286644953,
+      "loss": 0.8488,
+      "step": 331
+    },
+    {
+      "epoch": 0.7296703296703296,
+      "grad_norm": 0.05448528379201889,
+      "learning_rate": 0.002526058631921824,
+      "loss": 1.0398,
+      "step": 332
+    },
+    {
+      "epoch": 0.7318681318681318,
+      "grad_norm": 0.041912950575351715,
+      "learning_rate": 0.002523615635179153,
+      "loss": 0.7956,
+      "step": 333
+    },
+    {
+      "epoch": 0.734065934065934,
+      "grad_norm": 0.07482337951660156,
+      "learning_rate": 0.0025211726384364824,
+      "loss": 0.8518,
+      "step": 334
+    },
+    {
+      "epoch": 0.7362637362637363,
+      "grad_norm": 0.05419366806745529,
+      "learning_rate": 0.0025187296416938113,
+      "loss": 0.9319,
+      "step": 335
+    },
+    {
+      "epoch": 0.7384615384615385,
+      "grad_norm": 0.16014552116394043,
+      "learning_rate": 0.00251628664495114,
+      "loss": 1.1697,
+      "step": 336
+    },
+    {
+      "epoch": 0.7406593406593407,
+      "grad_norm": 0.03912324830889702,
+      "learning_rate": 0.002513843648208469,
+      "loss": 0.9458,
+      "step": 337
+    },
+    {
+      "epoch": 0.7428571428571429,
+      "grad_norm": 0.027873236685991287,
+      "learning_rate": 0.0025114006514657984,
+      "loss": 0.9165,
+      "step": 338
+    },
+    {
+      "epoch": 0.7450549450549451,
+      "grad_norm": 0.04961889237165451,
+      "learning_rate": 0.0025089576547231273,
+      "loss": 0.8336,
+      "step": 339
+    },
+    {
+      "epoch": 0.7472527472527473,
+      "grad_norm": 0.13535408675670624,
+      "learning_rate": 0.002506514657980456,
+      "loss": 1.1465,
+      "step": 340
+    },
+    {
+      "epoch": 0.7494505494505495,
+      "grad_norm": 0.08348392695188522,
+      "learning_rate": 0.002504071661237785,
+      "loss": 0.9365,
+      "step": 341
+    },
+    {
+      "epoch": 0.7516483516483516,
+      "grad_norm": 0.048418886959552765,
+      "learning_rate": 0.002501628664495114,
+      "loss": 0.8196,
+      "step": 342
+    },
+    {
+      "epoch": 0.7538461538461538,
+      "grad_norm": 0.28499606251716614,
+      "learning_rate": 0.0024991856677524433,
+      "loss": 0.8823,
+      "step": 343
+    },
+    {
+      "epoch": 0.756043956043956,
+      "grad_norm": 0.1474222093820572,
+      "learning_rate": 0.002496742671009772,
+      "loss": 0.7377,
+      "step": 344
+    },
+    {
+      "epoch": 0.7582417582417582,
+      "grad_norm": 0.0787111297249794,
+      "learning_rate": 0.002494299674267101,
+      "loss": 0.9211,
+      "step": 345
+    },
+    {
+      "epoch": 0.7604395604395604,
+      "grad_norm": 0.10333354026079178,
+      "learning_rate": 0.00249185667752443,
+      "loss": 0.816,
+      "step": 346
+    },
+    {
+      "epoch": 0.7626373626373626,
+      "grad_norm": 0.06213449686765671,
+      "learning_rate": 0.0024894136807817593,
+      "loss": 0.7897,
+      "step": 347
+    },
+    {
+      "epoch": 0.7648351648351648,
+      "grad_norm": 0.12437117099761963,
+      "learning_rate": 0.002486970684039088,
+      "loss": 1.0428,
+      "step": 348
+    },
+    {
+      "epoch": 0.7670329670329671,
+      "grad_norm": 0.044846177101135254,
+      "learning_rate": 0.002484527687296417,
+      "loss": 0.9299,
+      "step": 349
+    },
+    {
+      "epoch": 0.7692307692307693,
+      "grad_norm": 0.10742048919200897,
+      "learning_rate": 0.002482084690553746,
+      "loss": 0.8613,
+      "step": 350
+    },
+    {
+      "epoch": 0.7714285714285715,
+      "grad_norm": 0.053234536200761795,
+      "learning_rate": 0.002479641693811075,
+      "loss": 0.8037,
+      "step": 351
+    },
+    {
+      "epoch": 0.7736263736263737,
+      "grad_norm": 0.02632087655365467,
+      "learning_rate": 0.002477198697068404,
+      "loss": 0.8271,
+      "step": 352
+    },
+    {
+      "epoch": 0.7758241758241758,
+      "grad_norm": 0.08522136509418488,
+      "learning_rate": 0.002474755700325733,
+      "loss": 0.8401,
+      "step": 353
+    },
+    {
+      "epoch": 0.778021978021978,
+      "grad_norm": 0.06096068397164345,
+      "learning_rate": 0.002472312703583062,
+      "loss": 1.0178,
+      "step": 354
+    },
+    {
+      "epoch": 0.7802197802197802,
+      "grad_norm": 0.048374224454164505,
+      "learning_rate": 0.002469869706840391,
+      "loss": 0.7733,
+      "step": 355
+    },
+    {
+      "epoch": 0.7824175824175824,
+      "grad_norm": 0.045180805027484894,
+      "learning_rate": 0.00246742671009772,
+      "loss": 0.8313,
+      "step": 356
+    },
+    {
+      "epoch": 0.7846153846153846,
+      "grad_norm": 0.02626035176217556,
+      "learning_rate": 0.002464983713355049,
+      "loss": 0.7465,
+      "step": 357
+    },
+    {
+      "epoch": 0.7868131868131868,
+      "grad_norm": 0.0593128576874733,
+      "learning_rate": 0.002462540716612378,
+      "loss": 0.8256,
+      "step": 358
+    },
+    {
+      "epoch": 0.789010989010989,
+      "grad_norm": 0.20773394405841827,
+      "learning_rate": 0.002460097719869707,
+      "loss": 0.85,
+      "step": 359
+    },
+    {
+      "epoch": 0.7912087912087912,
+      "grad_norm": 0.08861587196588516,
+      "learning_rate": 0.0024576547231270358,
+      "loss": 0.9219,
+      "step": 360
+    },
+    {
+      "epoch": 0.7934065934065934,
+      "grad_norm": 0.04488822817802429,
+      "learning_rate": 0.002455211726384365,
+      "loss": 0.779,
+      "step": 361
+    },
+    {
+      "epoch": 0.7956043956043956,
+      "grad_norm": 0.04695796221494675,
+      "learning_rate": 0.002452768729641694,
+      "loss": 0.993,
+      "step": 362
+    },
+    {
+      "epoch": 0.7978021978021979,
+      "grad_norm": 0.05028370022773743,
+      "learning_rate": 0.002450325732899023,
+      "loss": 0.8445,
+      "step": 363
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.11178960651159286,
+      "learning_rate": 0.0024478827361563518,
+      "loss": 1.0978,
+      "step": 364
+    },
+    {
+      "epoch": 0.8021978021978022,
+      "grad_norm": 0.038159411400556564,
+      "learning_rate": 0.002445439739413681,
+      "loss": 0.9092,
+      "step": 365
+    },
+    {
+      "epoch": 0.8043956043956044,
+      "grad_norm": 0.09004521369934082,
+      "learning_rate": 0.00244299674267101,
+      "loss": 0.8896,
+      "step": 366
+    },
+    {
+      "epoch": 0.8065934065934066,
+      "grad_norm": 0.05131766200065613,
+      "learning_rate": 0.002440553745928339,
+      "loss": 0.8475,
+      "step": 367
+    },
+    {
+      "epoch": 0.8087912087912088,
+      "grad_norm": 0.05716368556022644,
+      "learning_rate": 0.0024381107491856678,
+      "loss": 0.9553,
+      "step": 368
+    },
+    {
+      "epoch": 0.810989010989011,
+      "grad_norm": 0.08233657479286194,
+      "learning_rate": 0.0024356677524429967,
+      "loss": 0.7876,
+      "step": 369
+    },
+    {
+      "epoch": 0.8131868131868132,
+      "grad_norm": 0.4040449857711792,
+      "learning_rate": 0.002433224755700326,
+      "loss": 0.9941,
+      "step": 370
+    },
+    {
+      "epoch": 0.8153846153846154,
+      "grad_norm": 0.14026129245758057,
+      "learning_rate": 0.002430781758957655,
+      "loss": 1.2629,
+      "step": 371
+    },
+    {
+      "epoch": 0.8175824175824176,
+      "grad_norm": 0.03528503328561783,
+      "learning_rate": 0.0024283387622149838,
+      "loss": 0.8115,
+      "step": 372
+    },
+    {
+      "epoch": 0.8197802197802198,
+      "grad_norm": 0.10299558937549591,
+      "learning_rate": 0.0024258957654723127,
+      "loss": 0.9766,
+      "step": 373
+    },
+    {
+      "epoch": 0.8219780219780219,
+      "grad_norm": 0.3495822250843048,
+      "learning_rate": 0.002423452768729642,
+      "loss": 0.9706,
+      "step": 374
+    },
+    {
+      "epoch": 0.8241758241758241,
+      "grad_norm": 4.7798333168029785,
+      "learning_rate": 0.002421009771986971,
+      "loss": 0.9772,
+      "step": 375
+    },
+    {
+      "epoch": 0.8263736263736263,
+      "grad_norm": 0.03864486888051033,
+      "learning_rate": 0.0024185667752442998,
+      "loss": 0.7987,
+      "step": 376
+    },
+    {
+      "epoch": 0.8285714285714286,
+      "grad_norm": 0.05099165439605713,
+      "learning_rate": 0.0024161237785016287,
+      "loss": 0.8459,
+      "step": 377
+    },
+    {
+      "epoch": 0.8307692307692308,
+      "grad_norm": 0.12386634945869446,
+      "learning_rate": 0.0024136807817589576,
+      "loss": 0.8896,
+      "step": 378
+    },
+    {
+      "epoch": 0.832967032967033,
+      "grad_norm": 0.041459277272224426,
+      "learning_rate": 0.002411237785016287,
+      "loss": 0.8643,
+      "step": 379
+    },
+    {
+      "epoch": 0.8351648351648352,
+      "grad_norm": 0.0727735310792923,
+      "learning_rate": 0.0024087947882736158,
+      "loss": 0.9993,
+      "step": 380
+    },
+    {
+      "epoch": 0.8373626373626374,
+      "grad_norm": 0.13729000091552734,
+      "learning_rate": 0.0024063517915309447,
+      "loss": 0.8708,
+      "step": 381
+    },
+    {
+      "epoch": 0.8395604395604396,
+      "grad_norm": 0.6980714797973633,
+      "learning_rate": 0.0024039087947882736,
+      "loss": 0.8368,
+      "step": 382
+    },
+    {
+      "epoch": 0.8417582417582418,
+      "grad_norm": 0.04392155259847641,
+      "learning_rate": 0.002401465798045603,
+      "loss": 0.8632,
+      "step": 383
+    },
+    {
+      "epoch": 0.843956043956044,
+      "grad_norm": 1.0243895053863525,
+      "learning_rate": 0.0023990228013029318,
+      "loss": 0.8837,
+      "step": 384
+    },
+    {
+      "epoch": 0.8461538461538461,
+      "grad_norm": 0.049360133707523346,
+      "learning_rate": 0.0023965798045602607,
+      "loss": 1.0476,
+      "step": 385
+    },
+    {
+      "epoch": 0.8483516483516483,
+      "grad_norm": 0.04752640798687935,
+      "learning_rate": 0.0023941368078175896,
+      "loss": 1.0436,
+      "step": 386
+    },
+    {
+      "epoch": 0.8505494505494505,
+      "grad_norm": 0.0454024076461792,
+      "learning_rate": 0.002391693811074919,
+      "loss": 0.7708,
+      "step": 387
+    },
+    {
+      "epoch": 0.8527472527472527,
+      "grad_norm": 0.10594017803668976,
+      "learning_rate": 0.0023892508143322478,
+      "loss": 0.9608,
+      "step": 388
+    },
+    {
+      "epoch": 0.8549450549450549,
+      "grad_norm": 0.035395313054323196,
+      "learning_rate": 0.0023868078175895767,
+      "loss": 0.8184,
+      "step": 389
+    },
+    {
+      "epoch": 0.8571428571428571,
+      "grad_norm": 0.08306609839200974,
+      "learning_rate": 0.0023843648208469056,
+      "loss": 0.9841,
+      "step": 390
+    },
+    {
+      "epoch": 0.8593406593406593,
+      "grad_norm": 0.04550642520189285,
+      "learning_rate": 0.0023819218241042345,
+      "loss": 1.0652,
+      "step": 391
+    },
+    {
+      "epoch": 0.8615384615384616,
+      "grad_norm": 0.04761195927858353,
+      "learning_rate": 0.002379478827361564,
+      "loss": 0.8301,
+      "step": 392
+    },
+    {
+      "epoch": 0.8637362637362638,
+      "grad_norm": 0.052182573825120926,
+      "learning_rate": 0.0023770358306188927,
+      "loss": 0.963,
+      "step": 393
+    },
+    {
+      "epoch": 0.865934065934066,
+      "grad_norm": 0.02773110195994377,
+      "learning_rate": 0.0023745928338762216,
+      "loss": 0.845,
+      "step": 394
+    },
+    {
+      "epoch": 0.8681318681318682,
+      "grad_norm": 0.05920000001788139,
+      "learning_rate": 0.0023721498371335505,
+      "loss": 0.7963,
+      "step": 395
+    },
+    {
+      "epoch": 0.8703296703296703,
+      "grad_norm": 0.05519743263721466,
+      "learning_rate": 0.00236970684039088,
+      "loss": 0.7159,
+      "step": 396
+    },
+    {
+      "epoch": 0.8725274725274725,
+      "grad_norm": 0.21729148924350739,
+      "learning_rate": 0.0023672638436482087,
+      "loss": 0.7501,
+      "step": 397
+    },
+    {
+      "epoch": 0.8747252747252747,
+      "grad_norm": 0.05414717644453049,
+      "learning_rate": 0.0023648208469055376,
+      "loss": 0.8683,
+      "step": 398
+    },
+    {
+      "epoch": 0.8769230769230769,
+      "grad_norm": 0.1997588872909546,
+      "learning_rate": 0.0023623778501628665,
+      "loss": 0.8903,
+      "step": 399
+    },
+    {
+      "epoch": 0.8791208791208791,
+      "grad_norm": 0.04832925274968147,
+      "learning_rate": 0.0023599348534201954,
+      "loss": 0.6782,
+      "step": 400
+    },
+    {
+      "epoch": 0.8813186813186813,
+      "grad_norm": 0.0891360342502594,
+      "learning_rate": 0.0023574918566775247,
+      "loss": 0.8201,
+      "step": 401
+    },
+    {
+      "epoch": 0.8835164835164835,
+      "grad_norm": 1.9886890649795532,
+      "learning_rate": 0.0023550488599348536,
+      "loss": 1.3708,
+      "step": 402
+    },
+    {
+      "epoch": 0.8857142857142857,
+      "grad_norm": 0.05694957077503204,
+      "learning_rate": 0.0023526058631921825,
+      "loss": 0.9302,
+      "step": 403
+    },
+    {
+      "epoch": 0.8879120879120879,
+      "grad_norm": 0.11793792992830276,
+      "learning_rate": 0.0023501628664495114,
+      "loss": 0.8701,
+      "step": 404
+    },
+    {
+      "epoch": 0.8901098901098901,
+      "grad_norm": 0.28880926966667175,
+      "learning_rate": 0.0023477198697068407,
+      "loss": 1.1316,
+      "step": 405
+    },
+    {
+      "epoch": 0.8923076923076924,
+      "grad_norm": 0.0657578855752945,
+      "learning_rate": 0.0023452768729641696,
+      "loss": 0.8546,
+      "step": 406
+    },
+    {
+      "epoch": 0.8945054945054945,
+      "grad_norm": 0.0305941179394722,
+      "learning_rate": 0.0023428338762214985,
+      "loss": 0.9834,
+      "step": 407
+    },
+    {
+      "epoch": 0.8967032967032967,
+      "grad_norm": 0.059357304126024246,
+      "learning_rate": 0.0023403908794788274,
+      "loss": 0.9172,
+      "step": 408
+    },
+    {
+      "epoch": 0.8989010989010989,
+      "grad_norm": 0.024256514385342598,
+      "learning_rate": 0.0023379478827361562,
+      "loss": 0.8335,
+      "step": 409
+    },
+    {
+      "epoch": 0.9010989010989011,
+      "grad_norm": 0.054907407611608505,
+      "learning_rate": 0.0023355048859934856,
+      "loss": 0.9088,
+      "step": 410
+    },
+    {
+      "epoch": 0.9032967032967033,
+      "grad_norm": 0.07981804013252258,
+      "learning_rate": 0.0023330618892508145,
+      "loss": 0.8169,
+      "step": 411
+    },
+    {
+      "epoch": 0.9054945054945055,
+      "grad_norm": 0.07843688875436783,
+      "learning_rate": 0.0023306188925081434,
+      "loss": 0.9562,
+      "step": 412
+    },
+    {
+      "epoch": 0.9076923076923077,
+      "grad_norm": 0.07083965092897415,
+      "learning_rate": 0.0023281758957654723,
+      "loss": 1.0081,
+      "step": 413
+    },
+    {
+      "epoch": 0.9098901098901099,
+      "grad_norm": 0.0345904715359211,
+      "learning_rate": 0.0023257328990228016,
+      "loss": 0.9854,
+      "step": 414
+    },
+    {
+      "epoch": 0.9120879120879121,
+      "grad_norm": 0.0834261029958725,
+      "learning_rate": 0.0023232899022801305,
+      "loss": 1.2914,
+      "step": 415
+    },
+    {
+      "epoch": 0.9142857142857143,
+      "grad_norm": 0.06423535197973251,
+      "learning_rate": 0.0023208469055374594,
+      "loss": 0.9758,
+      "step": 416
+    },
+    {
+      "epoch": 0.9164835164835164,
+      "grad_norm": 0.050010014325380325,
+      "learning_rate": 0.0023184039087947883,
+      "loss": 0.9546,
+      "step": 417
+    },
+    {
+      "epoch": 0.9186813186813186,
+      "grad_norm": 0.045019909739494324,
+      "learning_rate": 0.002315960912052117,
+      "loss": 0.7594,
+      "step": 418
+    },
+    {
+      "epoch": 0.9208791208791208,
+      "grad_norm": 0.05144082009792328,
+      "learning_rate": 0.0023135179153094465,
+      "loss": 1.0704,
+      "step": 419
+    },
+    {
+      "epoch": 0.9230769230769231,
+      "grad_norm": 0.04120473563671112,
+      "learning_rate": 0.0023110749185667754,
+      "loss": 0.7784,
+      "step": 420
+    },
+    {
+      "epoch": 0.9252747252747253,
+      "grad_norm": 0.08961661905050278,
+      "learning_rate": 0.0023086319218241043,
+      "loss": 0.897,
+      "step": 421
+    },
+    {
+      "epoch": 0.9274725274725275,
+      "grad_norm": 0.07543803006410599,
+      "learning_rate": 0.002306188925081433,
+      "loss": 0.9603,
+      "step": 422
+    },
+    {
+      "epoch": 0.9296703296703297,
+      "grad_norm": 0.059249624609947205,
+      "learning_rate": 0.0023037459283387625,
+      "loss": 0.953,
+      "step": 423
+    },
+    {
+      "epoch": 0.9318681318681319,
+      "grad_norm": 0.061665166169404984,
+      "learning_rate": 0.0023013029315960914,
+      "loss": 0.7479,
+      "step": 424
+    },
+    {
+      "epoch": 0.9340659340659341,
+      "grad_norm": 0.0700431615114212,
+      "learning_rate": 0.0022988599348534203,
+      "loss": 0.9089,
+      "step": 425
+    },
+    {
+      "epoch": 0.9362637362637363,
+      "grad_norm": 0.09456595033407211,
+      "learning_rate": 0.002296416938110749,
+      "loss": 0.9736,
+      "step": 426
+    },
+    {
+      "epoch": 0.9384615384615385,
+      "grad_norm": 0.04962339624762535,
+      "learning_rate": 0.002293973941368078,
+      "loss": 0.9771,
+      "step": 427
+    },
+    {
+      "epoch": 0.9406593406593406,
+      "grad_norm": 0.22272050380706787,
+      "learning_rate": 0.0022915309446254074,
+      "loss": 0.8767,
+      "step": 428
+    },
+    {
+      "epoch": 0.9428571428571428,
+      "grad_norm": 0.0336921289563179,
+      "learning_rate": 0.0022890879478827363,
+      "loss": 0.9146,
+      "step": 429
+    },
+    {
+      "epoch": 0.945054945054945,
+      "grad_norm": 0.10658980906009674,
+      "learning_rate": 0.002286644951140065,
+      "loss": 0.7424,
+      "step": 430
+    },
+    {
+      "epoch": 0.9472527472527472,
+      "grad_norm": 0.03370064124464989,
+      "learning_rate": 0.002284201954397394,
+      "loss": 0.8541,
+      "step": 431
+    },
+    {
+      "epoch": 0.9494505494505494,
+      "grad_norm": 0.03657378628849983,
+      "learning_rate": 0.0022817589576547234,
+      "loss": 0.9678,
+      "step": 432
+    },
+    {
+      "epoch": 0.9516483516483516,
+      "grad_norm": 0.0981496199965477,
+      "learning_rate": 0.0022793159609120523,
+      "loss": 0.9065,
+      "step": 433
+    },
+    {
+      "epoch": 0.9538461538461539,
+      "grad_norm": 0.06817454844713211,
+      "learning_rate": 0.002276872964169381,
+      "loss": 0.7096,
+      "step": 434
+    },
+    {
+      "epoch": 0.9560439560439561,
+      "grad_norm": 0.8452592492103577,
+      "learning_rate": 0.00227442996742671,
+      "loss": 1.1313,
+      "step": 435
+    },
+    {
+      "epoch": 0.9582417582417583,
+      "grad_norm": 0.023103436455130577,
+      "learning_rate": 0.002271986970684039,
+      "loss": 0.9307,
+      "step": 436
+    },
+    {
+      "epoch": 0.9604395604395605,
+      "grad_norm": 0.12862806022167206,
+      "learning_rate": 0.0022695439739413683,
+      "loss": 0.9483,
+      "step": 437
+    },
+    {
+      "epoch": 0.9626373626373627,
+      "grad_norm": 0.03854559361934662,
+      "learning_rate": 0.002267100977198697,
+      "loss": 0.97,
+      "step": 438
+    },
+    {
+      "epoch": 0.9648351648351648,
+      "grad_norm": 0.08664513379335403,
+      "learning_rate": 0.002264657980456026,
+      "loss": 0.8536,
+      "step": 439
+    },
+    {
+      "epoch": 0.967032967032967,
+      "grad_norm": 0.049822840839624405,
+      "learning_rate": 0.002262214983713355,
+      "loss": 0.8947,
+      "step": 440
+    },
+    {
+      "epoch": 0.9692307692307692,
+      "grad_norm": 0.03493356332182884,
+      "learning_rate": 0.0022597719869706843,
+      "loss": 1.066,
+      "step": 441
+    },
+    {
+      "epoch": 0.9714285714285714,
+      "grad_norm": 0.09576951712369919,
+      "learning_rate": 0.002257328990228013,
+      "loss": 0.8679,
+      "step": 442
+    },
+    {
+      "epoch": 0.9736263736263736,
+      "grad_norm": 0.04924570769071579,
+      "learning_rate": 0.002254885993485342,
+      "loss": 0.6763,
+      "step": 443
+    },
+    {
+      "epoch": 0.9758241758241758,
+      "grad_norm": 0.11691728234291077,
+      "learning_rate": 0.002252442996742671,
+      "loss": 0.8409,
+      "step": 444
+    },
+    {
+      "epoch": 0.978021978021978,
+      "grad_norm": 0.10364262014627457,
+      "learning_rate": 0.0022500000000000003,
+      "loss": 0.9213,
+      "step": 445
+    },
+    {
+      "epoch": 0.9802197802197802,
+      "grad_norm": 0.048065803945064545,
+      "learning_rate": 0.002247557003257329,
+      "loss": 0.7175,
+      "step": 446
+    },
+    {
+      "epoch": 0.9824175824175824,
+      "grad_norm": 0.04745359718799591,
+      "learning_rate": 0.002245114006514658,
+      "loss": 0.9177,
+      "step": 447
+    },
+    {
+      "epoch": 0.9846153846153847,
+      "grad_norm": 0.04276353120803833,
+      "learning_rate": 0.002242671009771987,
+      "loss": 0.9462,
+      "step": 448
+    },
+    {
+      "epoch": 0.9868131868131869,
+      "grad_norm": 0.04319168999791145,
+      "learning_rate": 0.002240228013029316,
+      "loss": 0.8923,
+      "step": 449
+    },
+    {
+      "epoch": 0.989010989010989,
+      "grad_norm": 0.04454359784722328,
+      "learning_rate": 0.002237785016286645,
+      "loss": 0.9423,
+      "step": 450
+    },
+    {
+      "epoch": 0.9912087912087912,
+      "grad_norm": 0.06027473881840706,
+      "learning_rate": 0.002235342019543974,
+      "loss": 0.8748,
+      "step": 451
+    },
+    {
+      "epoch": 0.9934065934065934,
+      "grad_norm": 0.04835286736488342,
+      "learning_rate": 0.002232899022801303,
+      "loss": 0.8799,
+      "step": 452
+    },
+    {
+      "epoch": 0.9956043956043956,
+      "grad_norm": 0.035824671387672424,
+      "learning_rate": 0.002230456026058632,
+      "loss": 0.7589,
+      "step": 453
+    },
+    {
+      "epoch": 0.9978021978021978,
+      "grad_norm": 0.09445648640394211,
+      "learning_rate": 0.002228013029315961,
+      "loss": 0.936,
+      "step": 454
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.13734450936317444,
+      "learning_rate": 0.00222557003257329,
+      "loss": 0.9591,
+      "step": 455
+    },
+    {
+      "epoch": 1.0021978021978022,
+      "grad_norm": 0.0319046787917614,
+      "learning_rate": 0.002223127035830619,
+      "loss": 0.6969,
+      "step": 456
+    },
+    {
+      "epoch": 1.0043956043956044,
+      "grad_norm": 0.04424639046192169,
+      "learning_rate": 0.002220684039087948,
+      "loss": 0.7051,
+      "step": 457
+    },
+    {
+      "epoch": 1.0065934065934066,
+      "grad_norm": 0.1892203688621521,
+      "learning_rate": 0.0022182410423452767,
+      "loss": 1.1239,
+      "step": 458
+    },
+    {
+      "epoch": 1.0087912087912088,
+      "grad_norm": 0.14804774522781372,
+      "learning_rate": 0.002215798045602606,
+      "loss": 0.817,
+      "step": 459
+    },
+    {
+      "epoch": 1.010989010989011,
+      "grad_norm": 0.13918998837471008,
+      "learning_rate": 0.002213355048859935,
+      "loss": 1.0336,
+      "step": 460
+    },
+    {
+      "epoch": 1.0131868131868131,
+      "grad_norm": 0.059844065457582474,
+      "learning_rate": 0.002210912052117264,
+      "loss": 0.8552,
+      "step": 461
+    },
+    {
+      "epoch": 1.0153846153846153,
+      "grad_norm": 0.03513557091355324,
+      "learning_rate": 0.0022084690553745927,
+      "loss": 0.7459,
+      "step": 462
+    },
+    {
+      "epoch": 1.0175824175824175,
+      "grad_norm": 0.04752732440829277,
+      "learning_rate": 0.002206026058631922,
+      "loss": 1.0452,
+      "step": 463
+    },
+    {
+      "epoch": 1.0197802197802197,
+      "grad_norm": 0.0643305554986,
+      "learning_rate": 0.002203583061889251,
+      "loss": 0.9969,
+      "step": 464
+    },
+    {
+      "epoch": 1.021978021978022,
+      "grad_norm": 0.10113230347633362,
+      "learning_rate": 0.00220114006514658,
+      "loss": 0.8162,
+      "step": 465
+    },
+    {
+      "epoch": 1.024175824175824,
+      "grad_norm": 0.06583012640476227,
+      "learning_rate": 0.0021986970684039087,
+      "loss": 0.9064,
+      "step": 466
+    },
+    {
+      "epoch": 1.0263736263736263,
+      "grad_norm": 0.040902573615312576,
+      "learning_rate": 0.0021962540716612376,
+      "loss": 0.8978,
+      "step": 467
+    },
+    {
+      "epoch": 1.0285714285714285,
+      "grad_norm": 0.06204291805624962,
+      "learning_rate": 0.002193811074918567,
+      "loss": 0.9303,
+      "step": 468
+    },
+    {
+      "epoch": 1.0307692307692307,
+      "grad_norm": 0.07494299113750458,
+      "learning_rate": 0.002191368078175896,
+      "loss": 0.9539,
+      "step": 469
+    },
+    {
+      "epoch": 1.032967032967033,
+      "grad_norm": 0.07009342312812805,
+      "learning_rate": 0.0021889250814332247,
+      "loss": 0.7964,
+      "step": 470
+    },
+    {
+      "epoch": 1.0351648351648353,
+      "grad_norm": 0.09100809693336487,
+      "learning_rate": 0.0021864820846905536,
+      "loss": 0.7511,
+      "step": 471
+    },
+    {
+      "epoch": 1.0373626373626375,
+      "grad_norm": 0.059530071914196014,
+      "learning_rate": 0.002184039087947883,
+      "loss": 0.9158,
+      "step": 472
+    },
+    {
+      "epoch": 1.0395604395604396,
+      "grad_norm": 0.05729622021317482,
+      "learning_rate": 0.002181596091205212,
+      "loss": 0.8225,
+      "step": 473
+    },
+    {
+      "epoch": 1.0417582417582418,
+      "grad_norm": 0.08491063863039017,
+      "learning_rate": 0.0021791530944625407,
+      "loss": 1.033,
+      "step": 474
+    },
+    {
+      "epoch": 1.043956043956044,
+      "grad_norm": 0.06388948857784271,
+      "learning_rate": 0.0021767100977198696,
+      "loss": 0.7809,
+      "step": 475
+    },
+    {
+      "epoch": 1.0461538461538462,
+      "grad_norm": 0.055855732411146164,
+      "learning_rate": 0.0021742671009771985,
+      "loss": 0.8998,
+      "step": 476
+    },
+    {
+      "epoch": 1.0483516483516484,
+      "grad_norm": 0.047501202672719955,
+      "learning_rate": 0.002171824104234528,
+      "loss": 1.0957,
+      "step": 477
+    },
+    {
+      "epoch": 1.0505494505494506,
+      "grad_norm": 0.06009456887841225,
+      "learning_rate": 0.0021693811074918567,
+      "loss": 0.8314,
+      "step": 478
+    },
+    {
+      "epoch": 1.0527472527472528,
+      "grad_norm": 0.023379506543278694,
+      "learning_rate": 0.0021669381107491856,
+      "loss": 0.7305,
+      "step": 479
+    },
+    {
+      "epoch": 1.054945054945055,
+      "grad_norm": 0.14162345230579376,
+      "learning_rate": 0.0021644951140065145,
+      "loss": 0.9821,
+      "step": 480
+    },
+    {
+      "epoch": 1.0571428571428572,
+      "grad_norm": 0.031201224774122238,
+      "learning_rate": 0.002162052117263844,
+      "loss": 0.9041,
+      "step": 481
+    },
+    {
+      "epoch": 1.0593406593406594,
+      "grad_norm": 0.10400353372097015,
+      "learning_rate": 0.0021596091205211727,
+      "loss": 0.8577,
+      "step": 482
+    },
+    {
+      "epoch": 1.0615384615384615,
+      "grad_norm": 0.07290414720773697,
+      "learning_rate": 0.0021571661237785016,
+      "loss": 0.8071,
+      "step": 483
+    },
+    {
+      "epoch": 1.0637362637362637,
+      "grad_norm": 0.025051966309547424,
+      "learning_rate": 0.0021547231270358305,
+      "loss": 0.905,
+      "step": 484
+    },
+    {
+      "epoch": 1.065934065934066,
+      "grad_norm": 0.07228747010231018,
+      "learning_rate": 0.0021522801302931594,
+      "loss": 0.8405,
+      "step": 485
+    },
+    {
+      "epoch": 1.0681318681318681,
+      "grad_norm": 0.08840841799974442,
+      "learning_rate": 0.0021498371335504887,
+      "loss": 1.0314,
+      "step": 486
+    },
+    {
+      "epoch": 1.0703296703296703,
+      "grad_norm": 0.07020457088947296,
+      "learning_rate": 0.0021473941368078176,
+      "loss": 1.1058,
+      "step": 487
+    },
+    {
+      "epoch": 1.0725274725274725,
+      "grad_norm": 0.029913805425167084,
+      "learning_rate": 0.0021449511400651465,
+      "loss": 0.6871,
+      "step": 488
+    },
+    {
+      "epoch": 1.0747252747252747,
+      "grad_norm": 0.05238842964172363,
+      "learning_rate": 0.0021425081433224754,
+      "loss": 0.9225,
+      "step": 489
+    },
+    {
+      "epoch": 1.0769230769230769,
+      "grad_norm": 0.053796958178281784,
+      "learning_rate": 0.0021400651465798047,
+      "loss": 0.9183,
+      "step": 490
+    },
+    {
+      "epoch": 1.079120879120879,
+      "grad_norm": 0.06680779159069061,
+      "learning_rate": 0.0021376221498371336,
+      "loss": 0.9569,
+      "step": 491
+    },
+    {
+      "epoch": 1.0813186813186813,
+      "grad_norm": 0.04390100762248039,
+      "learning_rate": 0.0021351791530944625,
+      "loss": 0.864,
+      "step": 492
+    },
+    {
+      "epoch": 1.0835164835164834,
+      "grad_norm": 0.04362115636467934,
+      "learning_rate": 0.0021327361563517914,
+      "loss": 0.8671,
+      "step": 493
+    },
+    {
+      "epoch": 1.0857142857142856,
+      "grad_norm": 0.05947096645832062,
+      "learning_rate": 0.0021302931596091203,
+      "loss": 0.978,
+      "step": 494
+    },
+    {
+      "epoch": 1.0879120879120878,
+      "grad_norm": 0.12668929994106293,
+      "learning_rate": 0.0021278501628664496,
+      "loss": 0.7484,
+      "step": 495
+    },
+    {
+      "epoch": 1.09010989010989,
+      "grad_norm": 0.05035700276494026,
+      "learning_rate": 0.0021254071661237785,
+      "loss": 1.01,
+      "step": 496
+    },
+    {
+      "epoch": 1.0923076923076924,
+      "grad_norm": 0.054348185658454895,
+      "learning_rate": 0.0021229641693811074,
+      "loss": 0.7815,
+      "step": 497
+    },
+    {
+      "epoch": 1.0945054945054946,
+      "grad_norm": 0.0411369726061821,
+      "learning_rate": 0.0021205211726384363,
+      "loss": 0.8517,
+      "step": 498
+    },
+    {
+      "epoch": 1.0967032967032968,
+      "grad_norm": 0.0687478557229042,
+      "learning_rate": 0.0021180781758957656,
+      "loss": 0.8744,
+      "step": 499
+    },
+    {
+      "epoch": 1.098901098901099,
+      "grad_norm": 0.04022733494639397,
+      "learning_rate": 0.0021156351791530945,
+      "loss": 0.7409,
+      "step": 500
+    },
+    {
+      "epoch": 1.1010989010989012,
+      "grad_norm": 0.07496348768472672,
+      "learning_rate": 0.0021131921824104234,
+      "loss": 0.7494,
+      "step": 501
+    },
+    {
+      "epoch": 1.1032967032967034,
+      "grad_norm": 0.03471522778272629,
+      "learning_rate": 0.0021107491856677523,
+      "loss": 0.8495,
+      "step": 502
+    },
+    {
+      "epoch": 1.1054945054945056,
+      "grad_norm": 0.10526181757450104,
+      "learning_rate": 0.002108306188925081,
+      "loss": 0.7972,
+      "step": 503
+    },
+    {
+      "epoch": 1.1076923076923078,
+      "grad_norm": 0.08372342586517334,
+      "learning_rate": 0.0021058631921824105,
+      "loss": 1.0791,
+      "step": 504
+    },
+    {
+      "epoch": 1.10989010989011,
+      "grad_norm": 0.12276377528905869,
+      "learning_rate": 0.0021034201954397394,
+      "loss": 0.8461,
+      "step": 505
+    },
+    {
+      "epoch": 1.1120879120879121,
+      "grad_norm": 0.03935771808028221,
+      "learning_rate": 0.0021009771986970683,
+      "loss": 0.8154,
+      "step": 506
+    },
+    {
+      "epoch": 1.1142857142857143,
+      "grad_norm": 0.11993210762739182,
+      "learning_rate": 0.002098534201954397,
+      "loss": 0.8524,
+      "step": 507
+    },
+    {
+      "epoch": 1.1164835164835165,
+      "grad_norm": 0.163649782538414,
+      "learning_rate": 0.0020960912052117265,
+      "loss": 0.8444,
+      "step": 508
+    },
+    {
+      "epoch": 1.1186813186813187,
+      "grad_norm": 0.046418070793151855,
+      "learning_rate": 0.0020936482084690554,
+      "loss": 0.881,
+      "step": 509
+    },
+    {
+      "epoch": 1.120879120879121,
+      "grad_norm": 0.11414097249507904,
+      "learning_rate": 0.0020912052117263843,
+      "loss": 0.9534,
+      "step": 510
+    },
+    {
+      "epoch": 1.123076923076923,
+      "grad_norm": 0.04443992301821709,
+      "learning_rate": 0.002088762214983713,
+      "loss": 0.8815,
+      "step": 511
+    },
+    {
+      "epoch": 1.1252747252747253,
+      "grad_norm": 0.1559406816959381,
+      "learning_rate": 0.0020863192182410425,
+      "loss": 0.8993,
+      "step": 512
+    },
+    {
+      "epoch": 1.1274725274725275,
+      "grad_norm": 0.06435513496398926,
+      "learning_rate": 0.0020838762214983714,
+      "loss": 0.8472,
+      "step": 513
+    },
+    {
+      "epoch": 1.1296703296703297,
+      "grad_norm": 0.05423581600189209,
+      "learning_rate": 0.0020814332247557003,
+      "loss": 0.9839,
+      "step": 514
+    },
+    {
+      "epoch": 1.1318681318681318,
+      "grad_norm": 0.06848083436489105,
+      "learning_rate": 0.002078990228013029,
+      "loss": 0.9077,
+      "step": 515
+    },
+    {
+      "epoch": 1.134065934065934,
+      "grad_norm": 0.04161064699292183,
+      "learning_rate": 0.002076547231270358,
+      "loss": 0.7664,
+      "step": 516
+    },
+    {
+      "epoch": 1.1362637362637362,
+      "grad_norm": 0.032689645886421204,
+      "learning_rate": 0.0020741042345276874,
+      "loss": 0.8578,
+      "step": 517
+    },
+    {
+      "epoch": 1.1384615384615384,
+      "grad_norm": 0.05889417976140976,
+      "learning_rate": 0.0020716612377850163,
+      "loss": 0.8934,
+      "step": 518
+    },
+    {
+      "epoch": 1.1406593406593406,
+      "grad_norm": 0.08336044102907181,
+      "learning_rate": 0.002069218241042345,
+      "loss": 0.8611,
+      "step": 519
+    },
+    {
+      "epoch": 1.1428571428571428,
+      "grad_norm": 0.050386011600494385,
+      "learning_rate": 0.002066775244299674,
+      "loss": 1.15,
+      "step": 520
+    },
+    {
+      "epoch": 1.145054945054945,
+      "grad_norm": 0.12142748385667801,
+      "learning_rate": 0.0020643322475570034,
+      "loss": 1.0372,
+      "step": 521
+    },
+    {
+      "epoch": 1.1472527472527472,
+      "grad_norm": 0.024582121521234512,
+      "learning_rate": 0.0020618892508143323,
+      "loss": 0.8644,
+      "step": 522
+    },
+    {
+      "epoch": 1.1494505494505494,
+      "grad_norm": 0.02363600768148899,
+      "learning_rate": 0.002059446254071661,
+      "loss": 0.6759,
+      "step": 523
+    },
+    {
+      "epoch": 1.1516483516483516,
+      "grad_norm": 0.051005635410547256,
+      "learning_rate": 0.00205700325732899,
+      "loss": 0.6683,
+      "step": 524
+    },
+    {
+      "epoch": 1.1538461538461537,
+      "grad_norm": 0.027199657633900642,
+      "learning_rate": 0.002054560260586319,
+      "loss": 0.7732,
+      "step": 525
+    },
+    {
+      "epoch": 1.156043956043956,
+      "grad_norm": 0.04509895667433739,
+      "learning_rate": 0.0020521172638436483,
+      "loss": 0.8279,
+      "step": 526
+    },
+    {
+      "epoch": 1.1582417582417581,
+      "grad_norm": 0.051365822553634644,
+      "learning_rate": 0.002049674267100977,
+      "loss": 0.8961,
+      "step": 527
+    },
+    {
+      "epoch": 1.1604395604395605,
+      "grad_norm": 0.03640604764223099,
+      "learning_rate": 0.002047231270358306,
+      "loss": 0.9626,
+      "step": 528
+    },
+    {
+      "epoch": 1.1626373626373627,
+      "grad_norm": 0.11194271594285965,
+      "learning_rate": 0.002044788273615635,
+      "loss": 0.8087,
+      "step": 529
+    },
+    {
+      "epoch": 1.164835164835165,
+      "grad_norm": 0.032918985933065414,
+      "learning_rate": 0.0020423452768729643,
+      "loss": 0.7039,
+      "step": 530
+    },
+    {
+      "epoch": 1.167032967032967,
+      "grad_norm": 0.04779953137040138,
+      "learning_rate": 0.002039902280130293,
+      "loss": 0.875,
+      "step": 531
+    },
+    {
+      "epoch": 1.1692307692307693,
+      "grad_norm": 0.06199371814727783,
+      "learning_rate": 0.002037459283387622,
+      "loss": 0.7221,
+      "step": 532
+    },
+    {
+      "epoch": 1.1714285714285715,
+      "grad_norm": 0.08789156377315521,
+      "learning_rate": 0.002035016286644951,
+      "loss": 0.9734,
+      "step": 533
+    },
+    {
+      "epoch": 1.1736263736263737,
+      "grad_norm": 0.05173993110656738,
+      "learning_rate": 0.00203257328990228,
+      "loss": 0.957,
+      "step": 534
+    },
+    {
+      "epoch": 1.1758241758241759,
+      "grad_norm": 0.05407477915287018,
+      "learning_rate": 0.0020301302931596092,
+      "loss": 0.7372,
+      "step": 535
+    },
+    {
+      "epoch": 1.178021978021978,
+      "grad_norm": 0.07874805480241776,
+      "learning_rate": 0.002027687296416938,
+      "loss": 0.8505,
+      "step": 536
+    },
+    {
+      "epoch": 1.1802197802197802,
+      "grad_norm": 0.061152003705501556,
+      "learning_rate": 0.002025244299674267,
+      "loss": 0.7838,
+      "step": 537
+    },
+    {
+      "epoch": 1.1824175824175824,
+      "grad_norm": 0.05171409249305725,
+      "learning_rate": 0.002022801302931596,
+      "loss": 0.8376,
+      "step": 538
+    },
+    {
+      "epoch": 1.1846153846153846,
+      "grad_norm": 0.05985184758901596,
+      "learning_rate": 0.0020203583061889252,
+      "loss": 0.9905,
+      "step": 539
+    },
+    {
+      "epoch": 1.1868131868131868,
+      "grad_norm": 0.12323056906461716,
+      "learning_rate": 0.002017915309446254,
+      "loss": 0.8964,
+      "step": 540
+    },
+    {
+      "epoch": 1.189010989010989,
+      "grad_norm": 0.03720412403345108,
+      "learning_rate": 0.002015472312703583,
+      "loss": 0.8045,
+      "step": 541
+    },
+    {
+      "epoch": 1.1912087912087912,
+      "grad_norm": 0.05883677303791046,
+      "learning_rate": 0.002013029315960912,
+      "loss": 0.865,
+      "step": 542
+    },
+    {
+      "epoch": 1.1934065934065934,
+      "grad_norm": 0.15464888513088226,
+      "learning_rate": 0.002010586319218241,
+      "loss": 0.7717,
+      "step": 543
+    },
+    {
+      "epoch": 1.1956043956043956,
+      "grad_norm": 0.1098957359790802,
+      "learning_rate": 0.00200814332247557,
+      "loss": 1.027,
+      "step": 544
+    },
+    {
+      "epoch": 1.1978021978021978,
+      "grad_norm": 0.041204702109098434,
+      "learning_rate": 0.002005700325732899,
+      "loss": 0.8015,
+      "step": 545
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.04377687722444534,
+      "learning_rate": 0.002003257328990228,
+      "loss": 0.9133,
+      "step": 546
+    },
+    {
+      "epoch": 1.2021978021978021,
+      "grad_norm": 0.034748874604701996,
+      "learning_rate": 0.002000814332247557,
+      "loss": 0.8895,
+      "step": 547
+    },
+    {
+      "epoch": 1.2043956043956043,
+      "grad_norm": 0.042624615132808685,
+      "learning_rate": 0.001998371335504886,
+      "loss": 0.8239,
+      "step": 548
+    },
+    {
+      "epoch": 1.2065934065934065,
+      "grad_norm": 0.09426282346248627,
+      "learning_rate": 0.001995928338762215,
+      "loss": 1.0433,
+      "step": 549
+    },
+    {
+      "epoch": 1.2087912087912087,
+      "grad_norm": 0.04767945036292076,
+      "learning_rate": 0.001993485342019544,
+      "loss": 0.9648,
+      "step": 550
+    },
+    {
+      "epoch": 1.210989010989011,
+      "grad_norm": 0.0780244991183281,
+      "learning_rate": 0.001991042345276873,
+      "loss": 0.8964,
+      "step": 551
+    },
+    {
+      "epoch": 1.213186813186813,
+      "grad_norm": 0.04296761006116867,
+      "learning_rate": 0.0019885993485342017,
+      "loss": 0.9178,
+      "step": 552
+    },
+    {
+      "epoch": 1.2153846153846155,
+      "grad_norm": 0.0720037892460823,
+      "learning_rate": 0.001986156351791531,
+      "loss": 0.8148,
+      "step": 553
+    },
+    {
+      "epoch": 1.2175824175824177,
+      "grad_norm": 0.07130163908004761,
+      "learning_rate": 0.00198371335504886,
+      "loss": 0.9156,
+      "step": 554
+    },
+    {
+      "epoch": 1.2197802197802199,
+      "grad_norm": 0.044256389141082764,
+      "learning_rate": 0.001981270358306189,
+      "loss": 0.7921,
+      "step": 555
+    },
+    {
+      "epoch": 1.221978021978022,
+      "grad_norm": 0.046621907502412796,
+      "learning_rate": 0.0019788273615635177,
+      "loss": 0.8549,
+      "step": 556
+    },
+    {
+      "epoch": 1.2241758241758243,
+      "grad_norm": 0.07500282675027847,
+      "learning_rate": 0.001976384364820847,
+      "loss": 1.0273,
+      "step": 557
+    },
+    {
+      "epoch": 1.2263736263736265,
+      "grad_norm": 0.18146595358848572,
+      "learning_rate": 0.001973941368078176,
+      "loss": 0.8335,
+      "step": 558
+    },
+    {
+      "epoch": 1.2285714285714286,
+      "grad_norm": 0.042887698858976364,
+      "learning_rate": 0.001971498371335505,
+      "loss": 0.8923,
+      "step": 559
+    },
+    {
+      "epoch": 1.2307692307692308,
+      "grad_norm": 0.027610143646597862,
+      "learning_rate": 0.0019690553745928337,
+      "loss": 0.7842,
+      "step": 560
+    },
+    {
+      "epoch": 1.232967032967033,
+      "grad_norm": 0.06510062515735626,
+      "learning_rate": 0.0019666123778501626,
+      "loss": 0.8435,
+      "step": 561
+    },
+    {
+      "epoch": 1.2351648351648352,
+      "grad_norm": 0.05440305918455124,
+      "learning_rate": 0.001964169381107492,
+      "loss": 1.0525,
+      "step": 562
+    },
+    {
+      "epoch": 1.2373626373626374,
+      "grad_norm": 0.0678519606590271,
+      "learning_rate": 0.001961726384364821,
+      "loss": 0.8947,
+      "step": 563
+    },
+    {
+      "epoch": 1.2395604395604396,
+      "grad_norm": 0.1181211769580841,
+      "learning_rate": 0.0019592833876221497,
+      "loss": 0.7141,
+      "step": 564
+    },
+    {
+      "epoch": 1.2417582417582418,
+      "grad_norm": 0.06664780527353287,
+      "learning_rate": 0.0019568403908794786,
+      "loss": 0.8073,
+      "step": 565
+    },
+    {
+      "epoch": 1.243956043956044,
+      "grad_norm": 0.04273368418216705,
+      "learning_rate": 0.001954397394136808,
+      "loss": 0.7919,
+      "step": 566
+    },
+    {
+      "epoch": 1.2461538461538462,
+      "grad_norm": 0.03139184042811394,
+      "learning_rate": 0.0019519543973941368,
+      "loss": 0.7657,
+      "step": 567
+    },
+    {
+      "epoch": 1.2483516483516484,
+      "grad_norm": 0.06013927981257439,
+      "learning_rate": 0.0019495114006514657,
+      "loss": 1.0389,
+      "step": 568
+    },
+    {
+      "epoch": 1.2505494505494505,
+      "grad_norm": 0.037411000579595566,
+      "learning_rate": 0.0019470684039087948,
+      "loss": 0.9003,
+      "step": 569
+    },
+    {
+      "epoch": 1.2527472527472527,
+      "grad_norm": 0.04453635960817337,
+      "learning_rate": 0.0019446254071661237,
+      "loss": 0.7938,
+      "step": 570
+    },
+    {
+      "epoch": 1.254945054945055,
+      "grad_norm": 0.027140459045767784,
+      "learning_rate": 0.0019421824104234526,
+      "loss": 0.9043,
+      "step": 571
+    },
+    {
+      "epoch": 1.2571428571428571,
+      "grad_norm": 0.028384899720549583,
+      "learning_rate": 0.0019397394136807817,
+      "loss": 0.7299,
+      "step": 572
+    },
+    {
+      "epoch": 1.2593406593406593,
+      "grad_norm": 0.060742396861314774,
+      "learning_rate": 0.0019372964169381106,
+      "loss": 0.7904,
+      "step": 573
+    },
+    {
+      "epoch": 1.2615384615384615,
+      "grad_norm": 0.09768784046173096,
+      "learning_rate": 0.0019348534201954397,
+      "loss": 1.0343,
+      "step": 574
+    },
+    {
+      "epoch": 1.2637362637362637,
+      "grad_norm": 0.026327388361096382,
+      "learning_rate": 0.0019324104234527686,
+      "loss": 0.9014,
+      "step": 575
+    },
+    {
+      "epoch": 1.2659340659340659,
+      "grad_norm": 0.4757933020591736,
+      "learning_rate": 0.0019299674267100977,
+      "loss": 0.8463,
+      "step": 576
+    },
+    {
+      "epoch": 1.268131868131868,
+      "grad_norm": 0.05362511798739433,
+      "learning_rate": 0.0019275244299674266,
+      "loss": 0.9303,
+      "step": 577
+    },
+    {
+      "epoch": 1.2703296703296703,
+      "grad_norm": 0.07107258588075638,
+      "learning_rate": 0.0019250814332247557,
+      "loss": 0.9485,
+      "step": 578
+    },
+    {
+      "epoch": 1.2725274725274724,
+      "grad_norm": 0.07357155531644821,
+      "learning_rate": 0.0019226384364820846,
+      "loss": 0.6658,
+      "step": 579
+    },
+    {
+      "epoch": 1.2747252747252746,
+      "grad_norm": 0.03434056043624878,
+      "learning_rate": 0.0019201954397394135,
+      "loss": 0.8436,
+      "step": 580
+    },
+    {
+      "epoch": 1.2769230769230768,
+      "grad_norm": 0.02289126068353653,
+      "learning_rate": 0.0019177524429967426,
+      "loss": 0.9347,
+      "step": 581
+    },
+    {
+      "epoch": 1.279120879120879,
+      "grad_norm": 0.03093271516263485,
+      "learning_rate": 0.0019153094462540715,
+      "loss": 0.726,
+      "step": 582
+    },
+    {
+      "epoch": 1.2813186813186812,
+      "grad_norm": 0.05604797601699829,
+      "learning_rate": 0.0019128664495114006,
+      "loss": 0.7935,
+      "step": 583
+    },
+    {
+      "epoch": 1.2835164835164834,
+      "grad_norm": 0.054815225303173065,
+      "learning_rate": 0.0019104234527687295,
+      "loss": 1.0261,
+      "step": 584
+    },
+    {
+      "epoch": 1.2857142857142856,
+      "grad_norm": 0.04914272949099541,
+      "learning_rate": 0.0019079804560260586,
+      "loss": 0.9114,
+      "step": 585
+    },
+    {
+      "epoch": 1.2879120879120878,
+      "grad_norm": 0.030239354819059372,
+      "learning_rate": 0.0019055374592833875,
+      "loss": 0.9067,
+      "step": 586
+    },
+    {
+      "epoch": 1.2901098901098902,
+      "grad_norm": 0.05694786459207535,
+      "learning_rate": 0.0019030944625407166,
+      "loss": 1.0383,
+      "step": 587
+    },
+    {
+      "epoch": 1.2923076923076924,
+      "grad_norm": 0.043653544038534164,
+      "learning_rate": 0.0019006514657980455,
+      "loss": 0.67,
+      "step": 588
+    },
+    {
+      "epoch": 1.2945054945054946,
+      "grad_norm": 0.03811546415090561,
+      "learning_rate": 0.0018982084690553744,
+      "loss": 0.7412,
+      "step": 589
+    },
+    {
+      "epoch": 1.2967032967032968,
+      "grad_norm": 0.05890112370252609,
+      "learning_rate": 0.0018957654723127035,
+      "loss": 0.856,
+      "step": 590
+    },
+    {
+      "epoch": 1.298901098901099,
+      "grad_norm": 0.0598793663084507,
+      "learning_rate": 0.0018933224755700324,
+      "loss": 0.8633,
+      "step": 591
+    },
+    {
+      "epoch": 1.3010989010989011,
+      "grad_norm": 0.11402341723442078,
+      "learning_rate": 0.0018908794788273615,
+      "loss": 0.8727,
+      "step": 592
+    },
+    {
+      "epoch": 1.3032967032967033,
+      "grad_norm": 0.11978960782289505,
+      "learning_rate": 0.0018884364820846904,
+      "loss": 0.9443,
+      "step": 593
+    },
+    {
+      "epoch": 1.3054945054945055,
+      "grad_norm": 0.05446099117398262,
+      "learning_rate": 0.0018859934853420195,
+      "loss": 1.021,
+      "step": 594
+    },
+    {
+      "epoch": 1.3076923076923077,
+      "grad_norm": 0.05832438915967941,
+      "learning_rate": 0.0018835504885993484,
+      "loss": 1.1145,
+      "step": 595
+    },
+    {
+      "epoch": 1.30989010989011,
+      "grad_norm": 0.05927540734410286,
+      "learning_rate": 0.0018811074918566775,
+      "loss": 0.8286,
+      "step": 596
+    },
+    {
+      "epoch": 1.312087912087912,
+      "grad_norm": 0.024790456518530846,
+      "learning_rate": 0.0018786644951140064,
+      "loss": 0.7817,
+      "step": 597
+    },
+    {
+      "epoch": 1.3142857142857143,
+      "grad_norm": 0.22037923336029053,
+      "learning_rate": 0.0018762214983713355,
+      "loss": 0.8688,
+      "step": 598
+    },
+    {
+      "epoch": 1.3164835164835165,
+      "grad_norm": 0.03462598845362663,
+      "learning_rate": 0.0018737785016286646,
+      "loss": 0.9163,
+      "step": 599
+    },
+    {
+      "epoch": 1.3186813186813187,
+      "grad_norm": 0.10367460548877716,
+      "learning_rate": 0.0018713355048859937,
+      "loss": 0.897,
+      "step": 600
+    },
+    {
+      "epoch": 1.3208791208791208,
+      "grad_norm": 9.048579216003418,
+      "learning_rate": 0.0018688925081433226,
+      "loss": 0.9803,
+      "step": 601
+    },
+    {
+      "epoch": 1.323076923076923,
+      "grad_norm": 0.03491645306348801,
+      "learning_rate": 0.0018664495114006517,
+      "loss": 0.864,
+      "step": 602
+    },
+    {
+      "epoch": 1.3252747252747252,
+      "grad_norm": 0.04878610745072365,
+      "learning_rate": 0.0018640065146579806,
+      "loss": 0.8414,
+      "step": 603
+    },
+    {
+      "epoch": 1.3274725274725274,
+      "grad_norm": 0.06004871800541878,
+      "learning_rate": 0.0018615635179153097,
+      "loss": 0.8719,
+      "step": 604
+    },
+    {
+      "epoch": 1.3296703296703296,
+      "grad_norm": 0.029605960473418236,
+      "learning_rate": 0.0018591205211726386,
+      "loss": 0.8321,
+      "step": 605
+    },
+    {
+      "epoch": 1.3318681318681318,
+      "grad_norm": 0.08186768740415573,
+      "learning_rate": 0.0018566775244299677,
+      "loss": 0.9423,
+      "step": 606
+    },
+    {
+      "epoch": 1.334065934065934,
+      "grad_norm": 0.029842093586921692,
+      "learning_rate": 0.0018542345276872966,
+      "loss": 0.8036,
+      "step": 607
+    },
+    {
+      "epoch": 1.3362637362637364,
+      "grad_norm": 0.028520282357931137,
+      "learning_rate": 0.0018517915309446257,
+      "loss": 0.7925,
+      "step": 608
+    },
+    {
+      "epoch": 1.3384615384615386,
+      "grad_norm": 0.03816806152462959,
+      "learning_rate": 0.0018493485342019546,
+      "loss": 0.7349,
+      "step": 609
+    },
+    {
+      "epoch": 1.3406593406593408,
+      "grad_norm": 0.035781629383563995,
+      "learning_rate": 0.0018469055374592835,
+      "loss": 0.7688,
+      "step": 610
+    },
+    {
+      "epoch": 1.342857142857143,
+      "grad_norm": 0.03208281844854355,
+      "learning_rate": 0.0018444625407166126,
+      "loss": 0.7004,
+      "step": 611
+    },
+    {
+      "epoch": 1.3450549450549452,
+      "grad_norm": 0.026422608643770218,
+      "learning_rate": 0.0018420195439739415,
+      "loss": 0.7537,
+      "step": 612
+    },
+    {
+      "epoch": 1.3472527472527474,
+      "grad_norm": 0.07194595038890839,
+      "learning_rate": 0.0018395765472312706,
+      "loss": 0.8617,
+      "step": 613
+    },
+    {
+      "epoch": 1.3494505494505495,
+      "grad_norm": 0.042611271142959595,
+      "learning_rate": 0.0018371335504885995,
+      "loss": 1.0392,
+      "step": 614
+    },
+    {
+      "epoch": 1.3516483516483517,
+      "grad_norm": 0.04559699818491936,
+      "learning_rate": 0.0018346905537459286,
+      "loss": 0.8586,
+      "step": 615
+    },
+    {
+      "epoch": 1.353846153846154,
+      "grad_norm": 0.0343613438308239,
+      "learning_rate": 0.0018322475570032575,
+      "loss": 0.9421,
+      "step": 616
+    },
+    {
+      "epoch": 1.3560439560439561,
+      "grad_norm": 0.04783899337053299,
+      "learning_rate": 0.0018298045602605866,
+      "loss": 0.8875,
+      "step": 617
+    },
+    {
+      "epoch": 1.3582417582417583,
+      "grad_norm": 0.06508871912956238,
+      "learning_rate": 0.0018273615635179155,
+      "loss": 0.8341,
+      "step": 618
+    },
+    {
+      "epoch": 1.3604395604395605,
+      "grad_norm": 0.05235806107521057,
+      "learning_rate": 0.0018249185667752444,
+      "loss": 0.9222,
+      "step": 619
+    },
+    {
+      "epoch": 1.3626373626373627,
+      "grad_norm": 0.07014351338148117,
+      "learning_rate": 0.0018224755700325735,
+      "loss": 0.8481,
+      "step": 620
+    },
+    {
+      "epoch": 1.3648351648351649,
+      "grad_norm": 0.03672454506158829,
+      "learning_rate": 0.0018200325732899024,
+      "loss": 0.8693,
+      "step": 621
+    },
+    {
+      "epoch": 1.367032967032967,
+      "grad_norm": 0.04049781337380409,
+      "learning_rate": 0.0018175895765472315,
+      "loss": 0.7762,
+      "step": 622
+    },
+    {
+      "epoch": 1.3692307692307693,
+      "grad_norm": 0.05781969800591469,
+      "learning_rate": 0.0018151465798045604,
+      "loss": 0.7537,
+      "step": 623
+    },
+    {
+      "epoch": 1.3714285714285714,
+      "grad_norm": 0.04968271404504776,
+      "learning_rate": 0.0018127035830618895,
+      "loss": 0.842,
+      "step": 624
+    },
+    {
+      "epoch": 1.3736263736263736,
+      "grad_norm": 0.035343658179044724,
+      "learning_rate": 0.0018102605863192184,
+      "loss": 0.9314,
+      "step": 625
+    },
+    {
+      "epoch": 1.3758241758241758,
+      "grad_norm": 0.04070547595620155,
+      "learning_rate": 0.0018078175895765475,
+      "loss": 0.9304,
+      "step": 626
+    },
+    {
+      "epoch": 1.378021978021978,
+      "grad_norm": 0.0533272884786129,
+      "learning_rate": 0.0018053745928338764,
+      "loss": 0.7939,
+      "step": 627
+    },
+    {
+      "epoch": 1.3802197802197802,
+      "grad_norm": 0.04389304667711258,
+      "learning_rate": 0.0018029315960912053,
+      "loss": 0.923,
+      "step": 628
+    },
+    {
+      "epoch": 1.3824175824175824,
+      "grad_norm": 0.06306538730859756,
+      "learning_rate": 0.0018004885993485344,
+      "loss": 0.9923,
+      "step": 629
+    },
+    {
+      "epoch": 1.3846153846153846,
+      "grad_norm": 0.06568475067615509,
+      "learning_rate": 0.0017980456026058633,
+      "loss": 1.0732,
+      "step": 630
+    },
+    {
+      "epoch": 1.3868131868131868,
+      "grad_norm": 0.11930973827838898,
+      "learning_rate": 0.0017956026058631924,
+      "loss": 0.8578,
+      "step": 631
+    },
+    {
+      "epoch": 1.389010989010989,
+      "grad_norm": 0.0629403367638588,
+      "learning_rate": 0.0017931596091205213,
+      "loss": 0.8986,
+      "step": 632
+    },
+    {
+      "epoch": 1.3912087912087912,
+      "grad_norm": 0.049731504172086716,
+      "learning_rate": 0.0017907166123778504,
+      "loss": 1.1664,
+      "step": 633
+    },
+    {
+      "epoch": 1.3934065934065933,
+      "grad_norm": 0.033620987087488174,
+      "learning_rate": 0.0017882736156351793,
+      "loss": 0.9437,
+      "step": 634
+    },
+    {
+      "epoch": 1.3956043956043955,
+      "grad_norm": 0.06671487540006638,
+      "learning_rate": 0.0017858306188925084,
+      "loss": 0.8788,
+      "step": 635
+    },
+    {
+      "epoch": 1.3978021978021977,
+      "grad_norm": 0.03259884566068649,
+      "learning_rate": 0.0017833876221498373,
+      "loss": 0.8463,
+      "step": 636
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.07015585154294968,
+      "learning_rate": 0.0017809446254071664,
+      "loss": 0.7203,
+      "step": 637
+    },
+    {
+      "epoch": 1.402197802197802,
+      "grad_norm": 0.03657226264476776,
+      "learning_rate": 0.0017785016286644953,
+      "loss": 0.8977,
+      "step": 638
+    },
+    {
+      "epoch": 1.4043956043956043,
+      "grad_norm": 0.058222655206918716,
+      "learning_rate": 0.0017760586319218242,
+      "loss": 0.7385,
+      "step": 639
+    },
+    {
+      "epoch": 1.4065934065934065,
+      "grad_norm": 0.04094264656305313,
+      "learning_rate": 0.0017736156351791533,
+      "loss": 0.894,
+      "step": 640
+    },
+    {
+      "epoch": 1.4087912087912087,
+      "grad_norm": 0.0629672184586525,
+      "learning_rate": 0.0017711726384364822,
+      "loss": 0.7698,
+      "step": 641
+    },
+    {
+      "epoch": 1.4109890109890109,
+      "grad_norm": 0.08818002045154572,
+      "learning_rate": 0.0017687296416938113,
+      "loss": 0.7784,
+      "step": 642
+    },
+    {
+      "epoch": 1.413186813186813,
+      "grad_norm": 0.2597349286079407,
+      "learning_rate": 0.0017662866449511402,
+      "loss": 0.8903,
+      "step": 643
+    },
+    {
+      "epoch": 1.4153846153846155,
+      "grad_norm": 0.040096838027238846,
+      "learning_rate": 0.0017638436482084693,
+      "loss": 0.7466,
+      "step": 644
+    },
+    {
+      "epoch": 1.4175824175824177,
+      "grad_norm": 0.0966111272573471,
+      "learning_rate": 0.0017614006514657982,
+      "loss": 0.9355,
+      "step": 645
+    },
+    {
+      "epoch": 1.4197802197802198,
+      "grad_norm": 0.0611172579228878,
+      "learning_rate": 0.0017589576547231273,
+      "loss": 0.8871,
+      "step": 646
+    },
+    {
+      "epoch": 1.421978021978022,
+      "grad_norm": 0.09354828298091888,
+      "learning_rate": 0.0017565146579804562,
+      "loss": 0.83,
+      "step": 647
+    },
+    {
+      "epoch": 1.4241758241758242,
+      "grad_norm": 0.043614018708467484,
+      "learning_rate": 0.001754071661237785,
+      "loss": 0.8489,
+      "step": 648
+    },
+    {
+      "epoch": 1.4263736263736264,
+      "grad_norm": 0.022785339504480362,
+      "learning_rate": 0.0017516286644951142,
+      "loss": 0.8303,
+      "step": 649
+    },
+    {
+      "epoch": 1.4285714285714286,
+      "grad_norm": 0.11944230645895004,
+      "learning_rate": 0.001749185667752443,
+      "loss": 0.9763,
+      "step": 650
+    },
+    {
+      "epoch": 1.4307692307692308,
+      "grad_norm": 0.02540021389722824,
+      "learning_rate": 0.0017467426710097722,
+      "loss": 0.8285,
+      "step": 651
+    },
+    {
+      "epoch": 1.432967032967033,
+      "grad_norm": 0.049928437918424606,
+      "learning_rate": 0.001744299674267101,
+      "loss": 0.9934,
+      "step": 652
+    },
+    {
+      "epoch": 1.4351648351648352,
+      "grad_norm": 0.08238594979047775,
+      "learning_rate": 0.0017418566775244302,
+      "loss": 0.8184,
+      "step": 653
+    },
+    {
+      "epoch": 1.4373626373626374,
+      "grad_norm": 0.025115888565778732,
+      "learning_rate": 0.001739413680781759,
+      "loss": 0.796,
+      "step": 654
+    },
+    {
+      "epoch": 1.4395604395604396,
+      "grad_norm": 0.039658959954977036,
+      "learning_rate": 0.0017369706840390882,
+      "loss": 0.9478,
+      "step": 655
+    },
+    {
+      "epoch": 1.4417582417582417,
+      "grad_norm": 0.0937625840306282,
+      "learning_rate": 0.001734527687296417,
+      "loss": 0.6744,
+      "step": 656
+    },
+    {
+      "epoch": 1.443956043956044,
+      "grad_norm": 0.027652230113744736,
+      "learning_rate": 0.001732084690553746,
+      "loss": 0.8684,
+      "step": 657
+    },
+    {
+      "epoch": 1.4461538461538461,
+      "grad_norm": 0.03960264101624489,
+      "learning_rate": 0.001729641693811075,
+      "loss": 0.8947,
+      "step": 658
+    },
+    {
+      "epoch": 1.4483516483516483,
+      "grad_norm": 0.04472724720835686,
+      "learning_rate": 0.001727198697068404,
+      "loss": 0.7986,
+      "step": 659
+    },
+    {
+      "epoch": 1.4505494505494505,
+      "grad_norm": 0.04149986803531647,
+      "learning_rate": 0.001724755700325733,
+      "loss": 0.9487,
+      "step": 660
+    },
+    {
+      "epoch": 1.4527472527472527,
+      "grad_norm": 0.036768488585948944,
+      "learning_rate": 0.001722312703583062,
+      "loss": 0.9244,
+      "step": 661
+    },
+    {
+      "epoch": 1.4549450549450549,
+      "grad_norm": 0.3078508973121643,
+      "learning_rate": 0.001719869706840391,
+      "loss": 0.9244,
+      "step": 662
+    },
+    {
+      "epoch": 1.457142857142857,
+      "grad_norm": 0.05046495050191879,
+      "learning_rate": 0.00171742671009772,
+      "loss": 0.8336,
+      "step": 663
+    },
+    {
+      "epoch": 1.4593406593406593,
+      "grad_norm": 0.07425173372030258,
+      "learning_rate": 0.001714983713355049,
+      "loss": 0.7726,
+      "step": 664
+    },
+    {
+      "epoch": 1.4615384615384617,
+      "grad_norm": 0.0336199551820755,
+      "learning_rate": 0.001712540716612378,
+      "loss": 0.9484,
+      "step": 665
+    },
+    {
+      "epoch": 1.4637362637362639,
+      "grad_norm": 0.042435843497514725,
+      "learning_rate": 0.0017100977198697069,
+      "loss": 0.9543,
+      "step": 666
+    },
+    {
+      "epoch": 1.465934065934066,
+      "grad_norm": 0.046244941651821136,
+      "learning_rate": 0.001707654723127036,
+      "loss": 0.9664,
+      "step": 667
+    },
+    {
+      "epoch": 1.4681318681318682,
+      "grad_norm": 0.028900867328047752,
+      "learning_rate": 0.0017052117263843649,
+      "loss": 0.7365,
+      "step": 668
+    },
+    {
+      "epoch": 1.4703296703296704,
+      "grad_norm": 0.036495063453912735,
+      "learning_rate": 0.001702768729641694,
+      "loss": 0.7271,
+      "step": 669
+    },
+    {
+      "epoch": 1.4725274725274726,
+      "grad_norm": 0.018637124449014664,
+      "learning_rate": 0.0017003257328990229,
+      "loss": 0.7472,
+      "step": 670
+    },
+    {
+      "epoch": 1.4747252747252748,
+      "grad_norm": 0.04221239313483238,
+      "learning_rate": 0.001697882736156352,
+      "loss": 0.7635,
+      "step": 671
+    },
+    {
+      "epoch": 1.476923076923077,
+      "grad_norm": 0.03966229408979416,
+      "learning_rate": 0.0016954397394136809,
+      "loss": 0.9103,
+      "step": 672
+    },
+    {
+      "epoch": 1.4791208791208792,
+      "grad_norm": 0.046825431287288666,
+      "learning_rate": 0.00169299674267101,
+      "loss": 1.0641,
+      "step": 673
+    },
+    {
+      "epoch": 1.4813186813186814,
+      "grad_norm": 0.04803401231765747,
+      "learning_rate": 0.0016905537459283389,
+      "loss": 0.9167,
+      "step": 674
+    },
+    {
+      "epoch": 1.4835164835164836,
+      "grad_norm": 0.0367858000099659,
+      "learning_rate": 0.001688110749185668,
+      "loss": 0.8032,
+      "step": 675
+    },
+    {
+      "epoch": 1.4857142857142858,
+      "grad_norm": 0.06937430053949356,
+      "learning_rate": 0.0016856677524429969,
+      "loss": 0.8396,
+      "step": 676
+    },
+    {
+      "epoch": 1.487912087912088,
+      "grad_norm": 0.12049335241317749,
+      "learning_rate": 0.0016832247557003258,
+      "loss": 0.8721,
+      "step": 677
+    },
+    {
+      "epoch": 1.4901098901098901,
+      "grad_norm": 0.08494982868432999,
+      "learning_rate": 0.0016807817589576549,
+      "loss": 1.0336,
+      "step": 678
+    },
+    {
+      "epoch": 1.4923076923076923,
+      "grad_norm": 0.03498467057943344,
+      "learning_rate": 0.0016783387622149838,
+      "loss": 0.7971,
+      "step": 679
+    },
+    {
+      "epoch": 1.4945054945054945,
+      "grad_norm": 0.058537233620882034,
+      "learning_rate": 0.0016758957654723129,
+      "loss": 1.0681,
+      "step": 680
+    },
+    {
+      "epoch": 1.4967032967032967,
+      "grad_norm": 0.2009536176919937,
+      "learning_rate": 0.0016734527687296418,
+      "loss": 0.8621,
+      "step": 681
+    },
+    {
+      "epoch": 1.498901098901099,
+      "grad_norm": 0.16682758927345276,
+      "learning_rate": 0.0016710097719869709,
+      "loss": 0.7328,
+      "step": 682
+    },
+    {
+      "epoch": 1.501098901098901,
+      "grad_norm": 0.043158505111932755,
+      "learning_rate": 0.0016685667752442998,
+      "loss": 0.7841,
+      "step": 683
+    },
+    {
+      "epoch": 1.5032967032967033,
+      "grad_norm": 0.22982844710350037,
+      "learning_rate": 0.0016661237785016289,
+      "loss": 0.9509,
+      "step": 684
+    },
+    {
+      "epoch": 1.5054945054945055,
+      "grad_norm": 0.018600644543766975,
+      "learning_rate": 0.0016636807817589578,
+      "loss": 0.8185,
+      "step": 685
+    },
+    {
+      "epoch": 1.5076923076923077,
+      "grad_norm": 0.17662671208381653,
+      "learning_rate": 0.0016612377850162867,
+      "loss": 0.9172,
+      "step": 686
+    },
+    {
+      "epoch": 1.5098901098901099,
+      "grad_norm": 0.033524177968502045,
+      "learning_rate": 0.0016587947882736158,
+      "loss": 0.8433,
+      "step": 687
+    },
+    {
+      "epoch": 1.512087912087912,
+      "grad_norm": 0.038456737995147705,
+      "learning_rate": 0.0016563517915309447,
+      "loss": 0.7667,
+      "step": 688
+    },
+    {
+      "epoch": 1.5142857142857142,
+      "grad_norm": 0.05938485637307167,
+      "learning_rate": 0.0016539087947882738,
+      "loss": 0.7913,
+      "step": 689
+    },
+    {
+      "epoch": 1.5164835164835164,
+      "grad_norm": 0.05964992940425873,
+      "learning_rate": 0.0016514657980456027,
+      "loss": 1.0393,
+      "step": 690
+    },
+    {
+      "epoch": 1.5186813186813186,
+      "grad_norm": 0.03610994666814804,
+      "learning_rate": 0.0016490228013029318,
+      "loss": 0.7123,
+      "step": 691
+    },
+    {
+      "epoch": 1.5208791208791208,
+      "grad_norm": 0.1029958575963974,
+      "learning_rate": 0.0016465798045602607,
+      "loss": 0.8756,
+      "step": 692
+    },
+    {
+      "epoch": 1.523076923076923,
+      "grad_norm": 0.045039985328912735,
+      "learning_rate": 0.0016441368078175898,
+      "loss": 0.8407,
+      "step": 693
+    },
+    {
+      "epoch": 1.5252747252747252,
+      "grad_norm": 0.0545758418738842,
+      "learning_rate": 0.0016416938110749187,
+      "loss": 0.8225,
+      "step": 694
+    },
+    {
+      "epoch": 1.5274725274725274,
+      "grad_norm": 0.05336428061127663,
+      "learning_rate": 0.0016392508143322475,
+      "loss": 0.9851,
+      "step": 695
+    },
+    {
+      "epoch": 1.5296703296703296,
+      "grad_norm": 0.026310784742236137,
+      "learning_rate": 0.0016368078175895767,
+      "loss": 0.6662,
+      "step": 696
+    },
+    {
+      "epoch": 1.5318681318681318,
+      "grad_norm": 0.03884543851017952,
+      "learning_rate": 0.0016343648208469056,
+      "loss": 1.0976,
+      "step": 697
+    },
+    {
+      "epoch": 1.534065934065934,
+      "grad_norm": 0.10137592256069183,
+      "learning_rate": 0.0016319218241042347,
+      "loss": 0.9607,
+      "step": 698
+    },
+    {
+      "epoch": 1.5362637362637361,
+      "grad_norm": 0.06148463115096092,
+      "learning_rate": 0.0016294788273615636,
+      "loss": 0.9457,
+      "step": 699
+    },
+    {
+      "epoch": 1.5384615384615383,
+      "grad_norm": 0.05321504548192024,
+      "learning_rate": 0.0016270358306188927,
+      "loss": 0.8822,
+      "step": 700
+    },
+    {
+      "epoch": 1.5406593406593405,
+      "grad_norm": 0.052986450493335724,
+      "learning_rate": 0.0016245928338762216,
+      "loss": 1.1144,
+      "step": 701
+    },
+    {
+      "epoch": 1.5428571428571427,
+      "grad_norm": 0.038553860038518906,
+      "learning_rate": 0.0016221498371335507,
+      "loss": 1.1331,
+      "step": 702
+    },
+    {
+      "epoch": 1.545054945054945,
+      "grad_norm": 0.09367532283067703,
+      "learning_rate": 0.0016197068403908796,
+      "loss": 0.9779,
+      "step": 703
+    },
+    {
+      "epoch": 1.5472527472527473,
+      "grad_norm": 1.5256266593933105,
+      "learning_rate": 0.0016172638436482087,
+      "loss": 0.8414,
+      "step": 704
+    },
+    {
+      "epoch": 1.5494505494505495,
+      "grad_norm": 0.9524049162864685,
+      "learning_rate": 0.0016148208469055376,
+      "loss": 0.8641,
+      "step": 705
+    },
+    {
+      "epoch": 1.5516483516483517,
+      "grad_norm": 0.027717284858226776,
+      "learning_rate": 0.0016123778501628664,
+      "loss": 0.964,
+      "step": 706
+    },
+    {
+      "epoch": 1.5538461538461539,
+      "grad_norm": 0.02728918381035328,
+      "learning_rate": 0.0016099348534201956,
+      "loss": 0.9938,
+      "step": 707
+    },
+    {
+      "epoch": 1.556043956043956,
+      "grad_norm": 0.09604862332344055,
+      "learning_rate": 0.0016074918566775244,
+      "loss": 0.8383,
+      "step": 708
+    },
+    {
+      "epoch": 1.5582417582417583,
+      "grad_norm": 0.023234959691762924,
+      "learning_rate": 0.0016050488599348536,
+      "loss": 0.6583,
+      "step": 709
+    },
+    {
+      "epoch": 1.5604395604395604,
+      "grad_norm": 0.04871445521712303,
+      "learning_rate": 0.0016026058631921824,
+      "loss": 0.795,
+      "step": 710
+    },
+    {
+      "epoch": 1.5626373626373626,
+      "grad_norm": 0.03740822523832321,
+      "learning_rate": 0.0016001628664495116,
+      "loss": 0.9315,
+      "step": 711
+    },
+    {
+      "epoch": 1.5648351648351648,
+      "grad_norm": 0.12605534493923187,
+      "learning_rate": 0.0015977198697068404,
+      "loss": 0.8666,
+      "step": 712
+    },
+    {
+      "epoch": 1.567032967032967,
+      "grad_norm": 0.17427438497543335,
+      "learning_rate": 0.0015952768729641696,
+      "loss": 0.8339,
+      "step": 713
+    },
+    {
+      "epoch": 1.5692307692307692,
+      "grad_norm": 0.0496843196451664,
+      "learning_rate": 0.0015928338762214984,
+      "loss": 0.8956,
+      "step": 714
+    },
+    {
+      "epoch": 1.5714285714285714,
+      "grad_norm": 0.07286012172698975,
+      "learning_rate": 0.0015903908794788273,
+      "loss": 0.8467,
+      "step": 715
+    },
+    {
+      "epoch": 1.5736263736263736,
+      "grad_norm": 0.05186188220977783,
+      "learning_rate": 0.0015879478827361564,
+      "loss": 1.0775,
+      "step": 716
+    },
+    {
+      "epoch": 1.5758241758241758,
+      "grad_norm": 0.03404043987393379,
+      "learning_rate": 0.0015855048859934853,
+      "loss": 0.7686,
+      "step": 717
+    },
+    {
+      "epoch": 1.578021978021978,
+      "grad_norm": 0.028001457452774048,
+      "learning_rate": 0.0015830618892508145,
+      "loss": 0.938,
+      "step": 718
+    },
+    {
+      "epoch": 1.5802197802197804,
+      "grad_norm": 0.09142988175153732,
+      "learning_rate": 0.0015806188925081433,
+      "loss": 1.0155,
+      "step": 719
+    },
+    {
+      "epoch": 1.5824175824175826,
+      "grad_norm": 0.04244792461395264,
+      "learning_rate": 0.0015781758957654725,
+      "loss": 0.7594,
+      "step": 720
+    },
+    {
+      "epoch": 1.5846153846153848,
+      "grad_norm": 0.023329678922891617,
+      "learning_rate": 0.0015757328990228013,
+      "loss": 0.8639,
+      "step": 721
+    },
+    {
+      "epoch": 1.586813186813187,
+      "grad_norm": 0.04569568112492561,
+      "learning_rate": 0.0015732899022801305,
+      "loss": 0.8318,
+      "step": 722
+    },
+    {
+      "epoch": 1.5890109890109891,
+      "grad_norm": 0.11132006347179413,
+      "learning_rate": 0.0015708469055374593,
+      "loss": 0.9203,
+      "step": 723
+    },
+    {
+      "epoch": 1.5912087912087913,
+      "grad_norm": 0.13975942134857178,
+      "learning_rate": 0.0015684039087947882,
+      "loss": 1.0544,
+      "step": 724
+    },
+    {
+      "epoch": 1.5934065934065935,
+      "grad_norm": 0.04746357351541519,
+      "learning_rate": 0.0015659609120521173,
+      "loss": 0.9164,
+      "step": 725
+    },
+    {
+      "epoch": 1.5956043956043957,
+      "grad_norm": 0.07409714162349701,
+      "learning_rate": 0.0015635179153094462,
+      "loss": 0.8659,
+      "step": 726
+    },
+    {
+      "epoch": 1.597802197802198,
+      "grad_norm": 0.25153887271881104,
+      "learning_rate": 0.0015610749185667753,
+      "loss": 0.7706,
+      "step": 727
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.04193171486258507,
+      "learning_rate": 0.0015586319218241042,
+      "loss": 1.0364,
+      "step": 728
+    },
+    {
+      "epoch": 1.6021978021978023,
+      "grad_norm": 0.052855994552373886,
+      "learning_rate": 0.0015561889250814333,
+      "loss": 0.8324,
+      "step": 729
+    },
+    {
+      "epoch": 1.6043956043956045,
+      "grad_norm": 0.06718911975622177,
+      "learning_rate": 0.0015537459283387622,
+      "loss": 0.9218,
+      "step": 730
+    },
+    {
+      "epoch": 1.6065934065934067,
+      "grad_norm": 0.02175552397966385,
+      "learning_rate": 0.0015513029315960913,
+      "loss": 0.8326,
+      "step": 731
+    },
+    {
+      "epoch": 1.6087912087912088,
+      "grad_norm": 0.0408305898308754,
+      "learning_rate": 0.0015488599348534202,
+      "loss": 0.9702,
+      "step": 732
+    },
+    {
+      "epoch": 1.610989010989011,
+      "grad_norm": 0.033620625734329224,
+      "learning_rate": 0.0015464169381107491,
+      "loss": 0.8921,
+      "step": 733
+    },
+    {
+      "epoch": 1.6131868131868132,
+      "grad_norm": 0.05553550645709038,
+      "learning_rate": 0.0015439739413680782,
+      "loss": 0.7432,
+      "step": 734
+    },
+    {
+      "epoch": 1.6153846153846154,
+      "grad_norm": 0.13387341797351837,
+      "learning_rate": 0.0015415309446254071,
+      "loss": 0.8995,
+      "step": 735
+    },
+    {
+      "epoch": 1.6175824175824176,
+      "grad_norm": 0.04314985126256943,
+      "learning_rate": 0.0015390879478827362,
+      "loss": 1.05,
+      "step": 736
+    },
+    {
+      "epoch": 1.6197802197802198,
+      "grad_norm": 0.047170042991638184,
+      "learning_rate": 0.0015366449511400651,
+      "loss": 0.8987,
+      "step": 737
+    },
+    {
+      "epoch": 1.621978021978022,
+      "grad_norm": 0.03407232090830803,
+      "learning_rate": 0.0015342019543973942,
+      "loss": 0.7666,
+      "step": 738
+    },
+    {
+      "epoch": 1.6241758241758242,
+      "grad_norm": 0.058059509843587875,
+      "learning_rate": 0.0015317589576547231,
+      "loss": 0.9807,
+      "step": 739
+    },
+    {
+      "epoch": 1.6263736263736264,
+      "grad_norm": 0.05226845666766167,
+      "learning_rate": 0.0015293159609120522,
+      "loss": 1.1343,
+      "step": 740
+    },
+    {
+      "epoch": 1.6285714285714286,
+      "grad_norm": 0.02792464941740036,
+      "learning_rate": 0.0015268729641693811,
+      "loss": 0.6925,
+      "step": 741
+    },
+    {
+      "epoch": 1.6307692307692307,
+      "grad_norm": 0.03169325366616249,
+      "learning_rate": 0.0015244299674267102,
+      "loss": 0.9382,
+      "step": 742
+    },
+    {
+      "epoch": 1.632967032967033,
+      "grad_norm": 0.05978107824921608,
+      "learning_rate": 0.0015219869706840391,
+      "loss": 0.9164,
+      "step": 743
+    },
+    {
+      "epoch": 1.6351648351648351,
+      "grad_norm": 0.07863892614841461,
+      "learning_rate": 0.001519543973941368,
+      "loss": 0.8905,
+      "step": 744
+    },
+    {
+      "epoch": 1.6373626373626373,
+      "grad_norm": 0.05766276642680168,
+      "learning_rate": 0.0015171009771986971,
+      "loss": 1.1193,
+      "step": 745
+    },
+    {
+      "epoch": 1.6395604395604395,
+      "grad_norm": 0.04148796945810318,
+      "learning_rate": 0.001514657980456026,
+      "loss": 0.8714,
+      "step": 746
+    },
+    {
+      "epoch": 1.6417582417582417,
+      "grad_norm": 0.04507949575781822,
+      "learning_rate": 0.0015122149837133551,
+      "loss": 0.9801,
+      "step": 747
+    },
+    {
+      "epoch": 1.6439560439560439,
+      "grad_norm": 0.02600765973329544,
+      "learning_rate": 0.001509771986970684,
+      "loss": 0.8549,
+      "step": 748
+    },
+    {
+      "epoch": 1.646153846153846,
+      "grad_norm": 0.04189063981175423,
+      "learning_rate": 0.0015073289902280131,
+      "loss": 0.8535,
+      "step": 749
+    },
+    {
+      "epoch": 1.6483516483516483,
+      "grad_norm": 0.03792524337768555,
+      "learning_rate": 0.001504885993485342,
+      "loss": 1.0411,
+      "step": 750
+    },
+    {
+      "epoch": 1.6505494505494505,
+      "grad_norm": 0.052046824246644974,
+      "learning_rate": 0.0015024429967426711,
+      "loss": 0.7329,
+      "step": 751
+    },
+    {
+      "epoch": 1.6527472527472526,
+      "grad_norm": 0.05570438504219055,
+      "learning_rate": 0.0015,
+      "loss": 0.812,
+      "step": 752
+    },
+    {
+      "epoch": 1.6549450549450548,
+      "grad_norm": 0.03411942347884178,
+      "learning_rate": 0.001497557003257329,
+      "loss": 0.9355,
+      "step": 753
+    },
+    {
+      "epoch": 1.657142857142857,
+      "grad_norm": 0.04974370449781418,
+      "learning_rate": 0.001495114006514658,
+      "loss": 0.8477,
+      "step": 754
+    },
+    {
+      "epoch": 1.6593406593406592,
+      "grad_norm": 0.024112170562148094,
+      "learning_rate": 0.001492671009771987,
+      "loss": 0.8453,
+      "step": 755
+    },
+    {
+      "epoch": 1.6615384615384614,
+      "grad_norm": 0.03615124151110649,
+      "learning_rate": 0.001490228013029316,
+      "loss": 0.8184,
+      "step": 756
+    },
+    {
+      "epoch": 1.6637362637362636,
+      "grad_norm": 0.9950895309448242,
+      "learning_rate": 0.001487785016286645,
+      "loss": 0.748,
+      "step": 757
+    },
+    {
+      "epoch": 1.6659340659340658,
+      "grad_norm": 0.08055254071950912,
+      "learning_rate": 0.001485342019543974,
+      "loss": 1.0105,
+      "step": 758
+    },
+    {
+      "epoch": 1.668131868131868,
+      "grad_norm": 0.04676898568868637,
+      "learning_rate": 0.001482899022801303,
+      "loss": 0.8325,
+      "step": 759
+    },
+    {
+      "epoch": 1.6703296703296702,
+      "grad_norm": 0.0818711593747139,
+      "learning_rate": 0.001480456026058632,
+      "loss": 0.8452,
+      "step": 760
+    },
+    {
+      "epoch": 1.6725274725274726,
+      "grad_norm": 0.04235965386033058,
+      "learning_rate": 0.001478013029315961,
+      "loss": 0.6762,
+      "step": 761
+    },
+    {
+      "epoch": 1.6747252747252748,
+      "grad_norm": 0.0398535281419754,
+      "learning_rate": 0.0014755700325732898,
+      "loss": 0.7734,
+      "step": 762
+    },
+    {
+      "epoch": 1.676923076923077,
+      "grad_norm": 0.06801621615886688,
+      "learning_rate": 0.001473127035830619,
+      "loss": 0.8533,
+      "step": 763
+    },
+    {
+      "epoch": 1.6791208791208792,
+      "grad_norm": 0.05141564458608627,
+      "learning_rate": 0.0014706840390879478,
+      "loss": 0.8955,
+      "step": 764
+    },
+    {
+      "epoch": 1.6813186813186813,
+      "grad_norm": 0.042457837611436844,
+      "learning_rate": 0.001468241042345277,
+      "loss": 0.9136,
+      "step": 765
+    },
+    {
+      "epoch": 1.6835164835164835,
+      "grad_norm": 0.05894291773438454,
+      "learning_rate": 0.0014657980456026058,
+      "loss": 0.8953,
+      "step": 766
+    },
+    {
+      "epoch": 1.6857142857142857,
+      "grad_norm": 0.08279906958341599,
+      "learning_rate": 0.001463355048859935,
+      "loss": 0.8816,
+      "step": 767
+    },
+    {
+      "epoch": 1.687912087912088,
+      "grad_norm": 0.03641095384955406,
+      "learning_rate": 0.0014609120521172638,
+      "loss": 0.7377,
+      "step": 768
+    },
+    {
+      "epoch": 1.69010989010989,
+      "grad_norm": 0.04417746886610985,
+      "learning_rate": 0.001458469055374593,
+      "loss": 0.9609,
+      "step": 769
+    },
+    {
+      "epoch": 1.6923076923076923,
+      "grad_norm": 0.02962510474026203,
+      "learning_rate": 0.0014560260586319218,
+      "loss": 0.8949,
+      "step": 770
+    },
+    {
+      "epoch": 1.6945054945054945,
+      "grad_norm": 0.05673578381538391,
+      "learning_rate": 0.001453583061889251,
+      "loss": 0.8103,
+      "step": 771
+    },
+    {
+      "epoch": 1.6967032967032967,
+      "grad_norm": 0.30835822224617004,
+      "learning_rate": 0.0014511400651465798,
+      "loss": 1.0289,
+      "step": 772
+    },
+    {
+      "epoch": 1.6989010989010989,
+      "grad_norm": 0.03898535296320915,
+      "learning_rate": 0.0014486970684039087,
+      "loss": 0.7971,
+      "step": 773
+    },
+    {
+      "epoch": 1.701098901098901,
+      "grad_norm": 0.04155929759144783,
+      "learning_rate": 0.0014462540716612378,
+      "loss": 0.8857,
+      "step": 774
+    },
+    {
+      "epoch": 1.7032967032967035,
+      "grad_norm": 0.06282802671194077,
+      "learning_rate": 0.0014438110749185667,
+      "loss": 0.9724,
+      "step": 775
+    },
+    {
+      "epoch": 1.7054945054945057,
+      "grad_norm": 0.07795851677656174,
+      "learning_rate": 0.0014413680781758958,
+      "loss": 0.8815,
+      "step": 776
+    },
+    {
+      "epoch": 1.7076923076923078,
+      "grad_norm": 0.03403620794415474,
+      "learning_rate": 0.0014389250814332247,
+      "loss": 0.9508,
+      "step": 777
+    },
+    {
+      "epoch": 1.70989010989011,
+      "grad_norm": 0.030924219638109207,
+      "learning_rate": 0.0014364820846905538,
+      "loss": 0.8354,
+      "step": 778
+    },
+    {
+      "epoch": 1.7120879120879122,
+      "grad_norm": 0.03887954354286194,
+      "learning_rate": 0.0014340390879478827,
+      "loss": 0.7964,
+      "step": 779
+    },
+    {
+      "epoch": 1.7142857142857144,
+      "grad_norm": 0.06878355890512466,
+      "learning_rate": 0.0014315960912052118,
+      "loss": 0.9886,
+      "step": 780
+    },
+    {
+      "epoch": 1.7164835164835166,
+      "grad_norm": 0.03642389178276062,
+      "learning_rate": 0.0014291530944625407,
+      "loss": 1.0339,
+      "step": 781
+    },
+    {
+      "epoch": 1.7186813186813188,
+      "grad_norm": 0.06248272955417633,
+      "learning_rate": 0.0014267100977198696,
+      "loss": 0.9563,
+      "step": 782
+    },
+    {
+      "epoch": 1.720879120879121,
+      "grad_norm": 0.02264326997101307,
+      "learning_rate": 0.0014242671009771987,
+      "loss": 0.801,
+      "step": 783
+    },
+    {
+      "epoch": 1.7230769230769232,
+      "grad_norm": 0.026206867769360542,
+      "learning_rate": 0.0014218241042345276,
+      "loss": 0.6619,
+      "step": 784
+    },
+    {
+      "epoch": 1.7252747252747254,
+      "grad_norm": 0.08069222420454025,
+      "learning_rate": 0.0014193811074918567,
+      "loss": 0.6964,
+      "step": 785
+    },
+    {
+      "epoch": 1.7274725274725276,
+      "grad_norm": 0.054062891751527786,
+      "learning_rate": 0.0014169381107491856,
+      "loss": 1.0295,
+      "step": 786
+    },
+    {
+      "epoch": 1.7296703296703297,
+      "grad_norm": 0.08005525171756744,
+      "learning_rate": 0.0014144951140065147,
+      "loss": 0.8562,
+      "step": 787
+    },
+    {
+      "epoch": 1.731868131868132,
+      "grad_norm": 0.04356931149959564,
+      "learning_rate": 0.0014120521172638436,
+      "loss": 0.7499,
+      "step": 788
+    },
+    {
+      "epoch": 1.7340659340659341,
+      "grad_norm": 0.0360998697578907,
+      "learning_rate": 0.0014096091205211727,
+      "loss": 0.9376,
+      "step": 789
+    },
+    {
+      "epoch": 1.7362637362637363,
+      "grad_norm": 0.16316482424736023,
+      "learning_rate": 0.0014071661237785016,
+      "loss": 0.926,
+      "step": 790
+    },
+    {
+      "epoch": 1.7384615384615385,
+      "grad_norm": 0.037409327924251556,
+      "learning_rate": 0.0014047231270358305,
+      "loss": 0.9421,
+      "step": 791
+    },
+    {
+      "epoch": 1.7406593406593407,
+      "grad_norm": 0.056520234793424606,
+      "learning_rate": 0.0014022801302931596,
+      "loss": 0.8433,
+      "step": 792
+    },
+    {
+      "epoch": 1.7428571428571429,
+      "grad_norm": 0.09042045474052429,
+      "learning_rate": 0.0013998371335504885,
+      "loss": 0.7579,
+      "step": 793
+    },
+    {
+      "epoch": 1.745054945054945,
+      "grad_norm": 0.0512566901743412,
+      "learning_rate": 0.0013973941368078176,
+      "loss": 0.8395,
+      "step": 794
+    },
+    {
+      "epoch": 1.7472527472527473,
+      "grad_norm": 0.0397796705365181,
+      "learning_rate": 0.0013949511400651465,
+      "loss": 0.8018,
+      "step": 795
+    },
+    {
+      "epoch": 1.7494505494505495,
+      "grad_norm": 0.02993011847138405,
+      "learning_rate": 0.0013925081433224756,
+      "loss": 0.9927,
+      "step": 796
+    },
+    {
+      "epoch": 1.7516483516483516,
+      "grad_norm": 0.030108636245131493,
+      "learning_rate": 0.0013900651465798045,
+      "loss": 0.7938,
+      "step": 797
+    },
+    {
+      "epoch": 1.7538461538461538,
+      "grad_norm": 0.045820556581020355,
+      "learning_rate": 0.0013876221498371336,
+      "loss": 1.0039,
+      "step": 798
+    },
+    {
+      "epoch": 1.756043956043956,
+      "grad_norm": 0.023320458829402924,
+      "learning_rate": 0.0013851791530944625,
+      "loss": 0.993,
+      "step": 799
+    },
+    {
+      "epoch": 1.7582417582417582,
+      "grad_norm": 0.03128175809979439,
+      "learning_rate": 0.0013827361563517914,
+      "loss": 0.8632,
+      "step": 800
+    },
+    {
+      "epoch": 1.7604395604395604,
+      "grad_norm": 0.046687088906764984,
+      "learning_rate": 0.0013802931596091205,
+      "loss": 0.894,
+      "step": 801
+    },
+    {
+      "epoch": 1.7626373626373626,
+      "grad_norm": 0.06110764667391777,
+      "learning_rate": 0.0013778501628664494,
+      "loss": 0.8744,
+      "step": 802
+    },
+    {
+      "epoch": 1.7648351648351648,
+      "grad_norm": 0.06707638502120972,
+      "learning_rate": 0.0013754071661237785,
+      "loss": 0.9609,
+      "step": 803
+    },
+    {
+      "epoch": 1.767032967032967,
+      "grad_norm": 0.07968633621931076,
+      "learning_rate": 0.0013729641693811074,
+      "loss": 0.9474,
+      "step": 804
+    },
+    {
+      "epoch": 1.7692307692307692,
+      "grad_norm": 0.025223134085536003,
+      "learning_rate": 0.0013705211726384365,
+      "loss": 0.844,
+      "step": 805
+    },
+    {
+      "epoch": 1.7714285714285714,
+      "grad_norm": 0.04119432717561722,
+      "learning_rate": 0.0013680781758957654,
+      "loss": 0.8221,
+      "step": 806
+    },
+    {
+      "epoch": 1.7736263736263735,
+      "grad_norm": 0.09176324307918549,
+      "learning_rate": 0.0013656351791530945,
+      "loss": 0.938,
+      "step": 807
+    },
+    {
+      "epoch": 1.7758241758241757,
+      "grad_norm": 0.0285447146743536,
+      "learning_rate": 0.0013631921824104234,
+      "loss": 0.7893,
+      "step": 808
+    },
+    {
+      "epoch": 1.778021978021978,
+      "grad_norm": 0.16660016775131226,
+      "learning_rate": 0.0013607491856677525,
+      "loss": 0.6508,
+      "step": 809
+    },
+    {
+      "epoch": 1.7802197802197801,
+      "grad_norm": 0.047678831964731216,
+      "learning_rate": 0.0013583061889250814,
+      "loss": 0.9113,
+      "step": 810
+    },
+    {
+      "epoch": 1.7824175824175823,
+      "grad_norm": 0.05949613079428673,
+      "learning_rate": 0.0013558631921824103,
+      "loss": 0.8055,
+      "step": 811
+    },
+    {
+      "epoch": 1.7846153846153845,
+      "grad_norm": 0.045708391815423965,
+      "learning_rate": 0.0013534201954397394,
+      "loss": 0.9279,
+      "step": 812
+    },
+    {
+      "epoch": 1.7868131868131867,
+      "grad_norm": 0.03344249352812767,
+      "learning_rate": 0.0013509771986970683,
+      "loss": 0.8698,
+      "step": 813
+    },
+    {
+      "epoch": 1.7890109890109889,
+      "grad_norm": 0.03435906022787094,
+      "learning_rate": 0.0013485342019543974,
+      "loss": 0.8029,
+      "step": 814
+    },
+    {
+      "epoch": 1.791208791208791,
+      "grad_norm": 0.2527799606323242,
+      "learning_rate": 0.0013460912052117263,
+      "loss": 0.9657,
+      "step": 815
+    },
+    {
+      "epoch": 1.7934065934065933,
+      "grad_norm": 0.043608445674180984,
+      "learning_rate": 0.0013436482084690554,
+      "loss": 0.8065,
+      "step": 816
+    },
+    {
+      "epoch": 1.7956043956043954,
+      "grad_norm": 0.1736743450164795,
+      "learning_rate": 0.0013412052117263843,
+      "loss": 0.8726,
+      "step": 817
+    },
+    {
+      "epoch": 1.7978021978021979,
+      "grad_norm": 0.03881852328777313,
+      "learning_rate": 0.0013387622149837134,
+      "loss": 0.7592,
+      "step": 818
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.0216361116617918,
+      "learning_rate": 0.0013363192182410423,
+      "loss": 0.8232,
+      "step": 819
+    },
+    {
+      "epoch": 1.8021978021978022,
+      "grad_norm": 0.03205220401287079,
+      "learning_rate": 0.0013338762214983712,
+      "loss": 0.8829,
+      "step": 820
+    },
+    {
+      "epoch": 1.8043956043956044,
+      "grad_norm": 0.038091324269771576,
+      "learning_rate": 0.0013314332247557003,
+      "loss": 0.6986,
+      "step": 821
+    },
+    {
+      "epoch": 1.8065934065934066,
+      "grad_norm": 0.0500926747918129,
+      "learning_rate": 0.0013289902280130292,
+      "loss": 0.9489,
+      "step": 822
+    },
+    {
+      "epoch": 1.8087912087912088,
+      "grad_norm": 0.050895363092422485,
+      "learning_rate": 0.0013265472312703583,
+      "loss": 0.88,
+      "step": 823
+    },
+    {
+      "epoch": 1.810989010989011,
+      "grad_norm": 0.16274334490299225,
+      "learning_rate": 0.0013241042345276872,
+      "loss": 0.887,
+      "step": 824
+    },
+    {
+      "epoch": 1.8131868131868132,
+      "grad_norm": 0.06001630052924156,
+      "learning_rate": 0.0013216612377850163,
+      "loss": 0.8405,
+      "step": 825
+    },
+    {
+      "epoch": 1.8153846153846154,
+      "grad_norm": 0.0315244160592556,
+      "learning_rate": 0.0013192182410423452,
+      "loss": 0.7231,
+      "step": 826
+    },
+    {
+      "epoch": 1.8175824175824176,
+      "grad_norm": 0.07712877541780472,
+      "learning_rate": 0.0013167752442996743,
+      "loss": 0.8188,
+      "step": 827
+    },
+    {
+      "epoch": 1.8197802197802198,
+      "grad_norm": 0.03803786262869835,
+      "learning_rate": 0.0013143322475570032,
+      "loss": 0.8544,
+      "step": 828
+    },
+    {
+      "epoch": 1.821978021978022,
+      "grad_norm": 0.04209129512310028,
+      "learning_rate": 0.0013118892508143323,
+      "loss": 0.8981,
+      "step": 829
+    },
+    {
+      "epoch": 1.8241758241758241,
+      "grad_norm": 0.0571487620472908,
+      "learning_rate": 0.0013094462540716614,
+      "loss": 0.7877,
+      "step": 830
+    },
+    {
+      "epoch": 1.8263736263736263,
+      "grad_norm": 0.04091283679008484,
+      "learning_rate": 0.0013070032573289903,
+      "loss": 0.8904,
+      "step": 831
+    },
+    {
+      "epoch": 1.8285714285714287,
+      "grad_norm": 0.03143646568059921,
+      "learning_rate": 0.0013045602605863194,
+      "loss": 0.8381,
+      "step": 832
+    },
+    {
+      "epoch": 1.830769230769231,
+      "grad_norm": 0.06479953974485397,
+      "learning_rate": 0.0013021172638436483,
+      "loss": 0.9414,
+      "step": 833
+    },
+    {
+      "epoch": 1.8329670329670331,
+      "grad_norm": 0.0165408868342638,
+      "learning_rate": 0.0012996742671009772,
+      "loss": 0.7736,
+      "step": 834
+    },
+    {
+      "epoch": 1.8351648351648353,
+      "grad_norm": 0.048045799136161804,
+      "learning_rate": 0.0012972312703583063,
+      "loss": 1.0549,
+      "step": 835
+    },
+    {
+      "epoch": 1.8373626373626375,
+      "grad_norm": 0.052748050540685654,
+      "learning_rate": 0.0012947882736156352,
+      "loss": 1.1913,
+      "step": 836
+    },
+    {
+      "epoch": 1.8395604395604397,
+      "grad_norm": 0.13713949918746948,
+      "learning_rate": 0.0012923452768729643,
+      "loss": 1.045,
+      "step": 837
+    },
+    {
+      "epoch": 1.8417582417582419,
+      "grad_norm": 0.04602542147040367,
+      "learning_rate": 0.0012899022801302932,
+      "loss": 0.8879,
+      "step": 838
+    },
+    {
+      "epoch": 1.843956043956044,
+      "grad_norm": 0.03842175006866455,
+      "learning_rate": 0.0012874592833876223,
+      "loss": 0.9351,
+      "step": 839
+    },
+    {
+      "epoch": 1.8461538461538463,
+      "grad_norm": 0.040805261582136154,
+      "learning_rate": 0.0012850162866449512,
+      "loss": 0.8306,
+      "step": 840
+    },
+    {
+      "epoch": 1.8483516483516484,
+      "grad_norm": 0.04454004764556885,
+      "learning_rate": 0.0012825732899022803,
+      "loss": 0.8062,
+      "step": 841
+    },
+    {
+      "epoch": 1.8505494505494506,
+      "grad_norm": 0.054268281906843185,
+      "learning_rate": 0.0012801302931596092,
+      "loss": 0.7186,
+      "step": 842
+    },
+    {
+      "epoch": 1.8527472527472528,
+      "grad_norm": 0.03111185133457184,
+      "learning_rate": 0.0012776872964169383,
+      "loss": 0.942,
+      "step": 843
+    },
+    {
+      "epoch": 1.854945054945055,
+      "grad_norm": 0.06069261208176613,
+      "learning_rate": 0.0012752442996742672,
+      "loss": 0.874,
+      "step": 844
+    },
+    {
+      "epoch": 1.8571428571428572,
+      "grad_norm": 0.0688580647110939,
+      "learning_rate": 0.001272801302931596,
+      "loss": 0.7474,
+      "step": 845
+    },
+    {
+      "epoch": 1.8593406593406594,
+      "grad_norm": 0.025120245292782784,
+      "learning_rate": 0.0012703583061889252,
+      "loss": 0.9117,
+      "step": 846
+    },
+    {
+      "epoch": 1.8615384615384616,
+      "grad_norm": 0.04289273917675018,
+      "learning_rate": 0.001267915309446254,
+      "loss": 0.7612,
+      "step": 847
+    },
+    {
+      "epoch": 1.8637362637362638,
+      "grad_norm": 0.034117285162210464,
+      "learning_rate": 0.0012654723127035832,
+      "loss": 0.8127,
+      "step": 848
+    },
+    {
+      "epoch": 1.865934065934066,
+      "grad_norm": 0.044596754014492035,
+      "learning_rate": 0.001263029315960912,
+      "loss": 0.856,
+      "step": 849
+    },
+    {
+      "epoch": 1.8681318681318682,
+      "grad_norm": 0.0651300922036171,
+      "learning_rate": 0.0012605863192182412,
+      "loss": 0.7172,
+      "step": 850
+    },
+    {
+      "epoch": 1.8703296703296703,
+      "grad_norm": 0.03420322760939598,
+      "learning_rate": 0.00125814332247557,
+      "loss": 0.8497,
+      "step": 851
+    },
+    {
+      "epoch": 1.8725274725274725,
+      "grad_norm": 0.03500472381711006,
+      "learning_rate": 0.0012557003257328992,
+      "loss": 0.9012,
+      "step": 852
+    },
+    {
+      "epoch": 1.8747252747252747,
+      "grad_norm": 0.06498914211988449,
+      "learning_rate": 0.001253257328990228,
+      "loss": 0.9229,
+      "step": 853
+    },
+    {
+      "epoch": 1.876923076923077,
+      "grad_norm": 0.15596267580986023,
+      "learning_rate": 0.001250814332247557,
+      "loss": 0.6895,
+      "step": 854
+    },
+    {
+      "epoch": 1.879120879120879,
+      "grad_norm": 0.06853580474853516,
+      "learning_rate": 0.001248371335504886,
+      "loss": 0.7207,
+      "step": 855
+    },
+    {
+      "epoch": 1.8813186813186813,
+      "grad_norm": 0.02783915214240551,
+      "learning_rate": 0.001245928338762215,
+      "loss": 0.7655,
+      "step": 856
+    },
+    {
+      "epoch": 1.8835164835164835,
+      "grad_norm": 0.02463470958173275,
+      "learning_rate": 0.001243485342019544,
+      "loss": 0.8348,
+      "step": 857
+    },
+    {
+      "epoch": 1.8857142857142857,
+      "grad_norm": 0.030513158068060875,
+      "learning_rate": 0.001241042345276873,
+      "loss": 0.8977,
+      "step": 858
+    },
+    {
+      "epoch": 1.8879120879120879,
+      "grad_norm": 0.08373177796602249,
+      "learning_rate": 0.001238599348534202,
+      "loss": 0.7158,
+      "step": 859
+    },
+    {
+      "epoch": 1.89010989010989,
+      "grad_norm": 0.1106758713722229,
+      "learning_rate": 0.001236156351791531,
+      "loss": 0.8221,
+      "step": 860
+    },
+    {
+      "epoch": 1.8923076923076922,
+      "grad_norm": 0.0783909484744072,
+      "learning_rate": 0.00123371335504886,
+      "loss": 0.9052,
+      "step": 861
+    },
+    {
+      "epoch": 1.8945054945054944,
+      "grad_norm": 0.04555298760533333,
+      "learning_rate": 0.001231270358306189,
+      "loss": 0.7931,
+      "step": 862
+    },
+    {
+      "epoch": 1.8967032967032966,
+      "grad_norm": 0.10880576074123383,
+      "learning_rate": 0.0012288273615635179,
+      "loss": 0.9585,
+      "step": 863
+    },
+    {
+      "epoch": 1.8989010989010988,
+      "grad_norm": 0.44529005885124207,
+      "learning_rate": 0.001226384364820847,
+      "loss": 0.8861,
+      "step": 864
+    },
+    {
+      "epoch": 1.901098901098901,
+      "grad_norm": 0.11483030021190643,
+      "learning_rate": 0.0012239413680781759,
+      "loss": 0.8553,
+      "step": 865
+    },
+    {
+      "epoch": 1.9032967032967032,
+      "grad_norm": 0.06093232333660126,
+      "learning_rate": 0.001221498371335505,
+      "loss": 0.969,
+      "step": 866
+    },
+    {
+      "epoch": 1.9054945054945054,
+      "grad_norm": 0.028702178969979286,
+      "learning_rate": 0.0012190553745928339,
+      "loss": 0.7285,
+      "step": 867
+    },
+    {
+      "epoch": 1.9076923076923076,
+      "grad_norm": 0.29810142517089844,
+      "learning_rate": 0.001216612377850163,
+      "loss": 0.8402,
+      "step": 868
+    },
+    {
+      "epoch": 1.9098901098901098,
+      "grad_norm": 0.03954026848077774,
+      "learning_rate": 0.0012141693811074919,
+      "loss": 0.9407,
+      "step": 869
+    },
+    {
+      "epoch": 1.912087912087912,
+      "grad_norm": 0.07889142632484436,
+      "learning_rate": 0.001211726384364821,
+      "loss": 0.8472,
+      "step": 870
+    },
+    {
+      "epoch": 1.9142857142857141,
+      "grad_norm": 0.034086924046278,
+      "learning_rate": 0.0012092833876221499,
+      "loss": 0.8385,
+      "step": 871
+    },
+    {
+      "epoch": 1.9164835164835163,
+      "grad_norm": 0.03281879425048828,
+      "learning_rate": 0.0012068403908794788,
+      "loss": 0.8425,
+      "step": 872
+    },
+    {
+      "epoch": 1.9186813186813185,
+      "grad_norm": 0.06810958683490753,
+      "learning_rate": 0.0012043973941368079,
+      "loss": 1.0217,
+      "step": 873
+    },
+    {
+      "epoch": 1.9208791208791207,
+      "grad_norm": 0.029852233827114105,
+      "learning_rate": 0.0012019543973941368,
+      "loss": 0.7875,
+      "step": 874
+    },
+    {
+      "epoch": 1.9230769230769231,
+      "grad_norm": 0.07168880850076675,
+      "learning_rate": 0.0011995114006514659,
+      "loss": 1.0017,
+      "step": 875
+    },
+    {
+      "epoch": 1.9252747252747253,
+      "grad_norm": 0.0436844602227211,
+      "learning_rate": 0.0011970684039087948,
+      "loss": 0.8109,
+      "step": 876
+    },
+    {
+      "epoch": 1.9274725274725275,
+      "grad_norm": 0.057308297604322433,
+      "learning_rate": 0.0011946254071661239,
+      "loss": 0.9934,
+      "step": 877
+    },
+    {
+      "epoch": 1.9296703296703297,
+      "grad_norm": 0.4405398964881897,
+      "learning_rate": 0.0011921824104234528,
+      "loss": 0.9901,
+      "step": 878
+    },
+    {
+      "epoch": 1.9318681318681319,
+      "grad_norm": 0.07887808233499527,
+      "learning_rate": 0.001189739413680782,
+      "loss": 0.9744,
+      "step": 879
+    },
+    {
+      "epoch": 1.934065934065934,
+      "grad_norm": 0.06607577949762344,
+      "learning_rate": 0.0011872964169381108,
+      "loss": 0.7361,
+      "step": 880
+    },
+    {
+      "epoch": 1.9362637362637363,
+      "grad_norm": 0.04684818163514137,
+      "learning_rate": 0.00118485342019544,
+      "loss": 0.8572,
+      "step": 881
+    },
+    {
+      "epoch": 1.9384615384615385,
+      "grad_norm": 0.028856562450528145,
+      "learning_rate": 0.0011824104234527688,
+      "loss": 0.7849,
+      "step": 882
+    },
+    {
+      "epoch": 1.9406593406593406,
+      "grad_norm": 0.11375346034765244,
+      "learning_rate": 0.0011799674267100977,
+      "loss": 0.9354,
+      "step": 883
+    },
+    {
+      "epoch": 1.9428571428571428,
+      "grad_norm": 0.028785506263375282,
+      "learning_rate": 0.0011775244299674268,
+      "loss": 0.7446,
+      "step": 884
+    },
+    {
+      "epoch": 1.945054945054945,
+      "grad_norm": 0.04495609179139137,
+      "learning_rate": 0.0011750814332247557,
+      "loss": 0.8294,
+      "step": 885
+    },
+    {
+      "epoch": 1.9472527472527472,
+      "grad_norm": 0.04621626436710358,
+      "learning_rate": 0.0011726384364820848,
+      "loss": 0.8309,
+      "step": 886
+    },
+    {
+      "epoch": 1.9494505494505494,
+      "grad_norm": 0.07226962596178055,
+      "learning_rate": 0.0011701954397394137,
+      "loss": 1.0671,
+      "step": 887
+    },
+    {
+      "epoch": 1.9516483516483516,
+      "grad_norm": 0.05816445127129555,
+      "learning_rate": 0.0011677524429967428,
+      "loss": 0.8054,
+      "step": 888
+    },
+    {
+      "epoch": 1.953846153846154,
+      "grad_norm": 0.023325450718402863,
+      "learning_rate": 0.0011653094462540717,
+      "loss": 0.6833,
+      "step": 889
+    },
+    {
+      "epoch": 1.9560439560439562,
+      "grad_norm": 0.05415589362382889,
+      "learning_rate": 0.0011628664495114008,
+      "loss": 0.7779,
+      "step": 890
+    },
+    {
+      "epoch": 1.9582417582417584,
+      "grad_norm": 0.04167710617184639,
+      "learning_rate": 0.0011604234527687297,
+      "loss": 0.8656,
+      "step": 891
+    },
+    {
+      "epoch": 1.9604395604395606,
+      "grad_norm": 0.06591608375310898,
+      "learning_rate": 0.0011579804560260586,
+      "loss": 0.887,
+      "step": 892
+    },
+    {
+      "epoch": 1.9626373626373628,
+      "grad_norm": 0.03845475986599922,
+      "learning_rate": 0.0011555374592833877,
+      "loss": 0.9391,
+      "step": 893
+    },
+    {
+      "epoch": 1.964835164835165,
+      "grad_norm": 0.02555324137210846,
+      "learning_rate": 0.0011530944625407166,
+      "loss": 0.9041,
+      "step": 894
+    },
+    {
+      "epoch": 1.9670329670329672,
+      "grad_norm": 0.027588456869125366,
+      "learning_rate": 0.0011506514657980457,
+      "loss": 0.8258,
+      "step": 895
+    },
+    {
+      "epoch": 1.9692307692307693,
+      "grad_norm": 0.0401766411960125,
+      "learning_rate": 0.0011482084690553746,
+      "loss": 0.8247,
+      "step": 896
+    },
+    {
+      "epoch": 1.9714285714285715,
+      "grad_norm": 0.032968584448099136,
+      "learning_rate": 0.0011457654723127037,
+      "loss": 0.8208,
+      "step": 897
+    },
+    {
+      "epoch": 1.9736263736263737,
+      "grad_norm": 0.04281003773212433,
+      "learning_rate": 0.0011433224755700326,
+      "loss": 0.8551,
+      "step": 898
+    },
+    {
+      "epoch": 1.975824175824176,
+      "grad_norm": 0.03784700855612755,
+      "learning_rate": 0.0011408794788273617,
+      "loss": 0.8307,
+      "step": 899
+    },
+    {
+      "epoch": 1.978021978021978,
+      "grad_norm": 0.03873506933450699,
+      "learning_rate": 0.0011384364820846906,
+      "loss": 0.7418,
+      "step": 900
+    },
+    {
+      "epoch": 1.9802197802197803,
+      "grad_norm": 0.0607513003051281,
+      "learning_rate": 0.0011359934853420195,
+      "loss": 0.8923,
+      "step": 901
+    },
+    {
+      "epoch": 1.9824175824175825,
+      "grad_norm": 0.04378557950258255,
+      "learning_rate": 0.0011335504885993486,
+      "loss": 0.8419,
+      "step": 902
+    },
+    {
+      "epoch": 1.9846153846153847,
+      "grad_norm": 0.05366958677768707,
+      "learning_rate": 0.0011311074918566775,
+      "loss": 0.7059,
+      "step": 903
+    },
+    {
+      "epoch": 1.9868131868131869,
+      "grad_norm": 0.04219432920217514,
+      "learning_rate": 0.0011286644951140066,
+      "loss": 0.7755,
+      "step": 904
+    },
+    {
+      "epoch": 1.989010989010989,
+      "grad_norm": 0.10763835906982422,
+      "learning_rate": 0.0011262214983713355,
+      "loss": 0.8505,
+      "step": 905
+    },
+    {
+      "epoch": 1.9912087912087912,
+      "grad_norm": 0.03905473276972771,
+      "learning_rate": 0.0011237785016286646,
+      "loss": 0.8668,
+      "step": 906
+    },
+    {
+      "epoch": 1.9934065934065934,
+      "grad_norm": 0.03982299566268921,
+      "learning_rate": 0.0011213355048859935,
+      "loss": 0.8893,
+      "step": 907
+    },
+    {
+      "epoch": 1.9956043956043956,
+      "grad_norm": 0.03761680796742439,
+      "learning_rate": 0.0011188925081433226,
+      "loss": 0.782,
+      "step": 908
+    },
+    {
+      "epoch": 1.9978021978021978,
+      "grad_norm": 0.06367536634206772,
+      "learning_rate": 0.0011164495114006515,
+      "loss": 1.0983,
+      "step": 909
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.4264332056045532,
+      "learning_rate": 0.0011140065146579806,
+      "loss": 0.8666,
+      "step": 910
+    },
+    {
+      "epoch": 2.002197802197802,
+      "grad_norm": 0.04869040474295616,
+      "learning_rate": 0.0011115635179153095,
+      "loss": 0.8231,
+      "step": 911
+    },
+    {
+      "epoch": 2.0043956043956044,
+      "grad_norm": 0.046542514115571976,
+      "learning_rate": 0.0011091205211726384,
+      "loss": 0.8846,
+      "step": 912
+    },
+    {
+      "epoch": 2.0065934065934066,
+      "grad_norm": 0.036480579525232315,
+      "learning_rate": 0.0011066775244299675,
+      "loss": 0.6804,
+      "step": 913
+    },
+    {
+      "epoch": 2.0087912087912088,
+      "grad_norm": 0.057561203837394714,
+      "learning_rate": 0.0011042345276872964,
+      "loss": 0.942,
+      "step": 914
+    },
+    {
+      "epoch": 2.010989010989011,
+      "grad_norm": 0.020143093541264534,
+      "learning_rate": 0.0011017915309446255,
+      "loss": 0.854,
+      "step": 915
+    },
+    {
+      "epoch": 2.013186813186813,
+      "grad_norm": 0.04090616852045059,
+      "learning_rate": 0.0010993485342019544,
+      "loss": 0.7667,
+      "step": 916
+    },
+    {
+      "epoch": 2.0153846153846153,
+      "grad_norm": 0.0554121732711792,
+      "learning_rate": 0.0010969055374592835,
+      "loss": 0.9576,
+      "step": 917
+    },
+    {
+      "epoch": 2.0175824175824175,
+      "grad_norm": 0.09384334087371826,
+      "learning_rate": 0.0010944625407166124,
+      "loss": 0.9948,
+      "step": 918
+    },
+    {
+      "epoch": 2.0197802197802197,
+      "grad_norm": 0.027293479070067406,
+      "learning_rate": 0.0010920195439739415,
+      "loss": 0.6964,
+      "step": 919
+    },
+    {
+      "epoch": 2.021978021978022,
+      "grad_norm": 0.05100255832076073,
+      "learning_rate": 0.0010895765472312704,
+      "loss": 0.8148,
+      "step": 920
+    },
+    {
+      "epoch": 2.024175824175824,
+      "grad_norm": 0.048554129898548126,
+      "learning_rate": 0.0010871335504885993,
+      "loss": 0.8376,
+      "step": 921
+    },
+    {
+      "epoch": 2.0263736263736263,
+      "grad_norm": 0.07291349768638611,
+      "learning_rate": 0.0010846905537459284,
+      "loss": 0.782,
+      "step": 922
+    },
+    {
+      "epoch": 2.0285714285714285,
+      "grad_norm": 0.051710668951272964,
+      "learning_rate": 0.0010822475570032573,
+      "loss": 0.7931,
+      "step": 923
+    },
+    {
+      "epoch": 2.0307692307692307,
+      "grad_norm": 0.038739949464797974,
+      "learning_rate": 0.0010798045602605864,
+      "loss": 0.8657,
+      "step": 924
+    },
+    {
+      "epoch": 2.032967032967033,
+      "grad_norm": 0.03753436356782913,
+      "learning_rate": 0.0010773615635179153,
+      "loss": 0.7821,
+      "step": 925
+    },
+    {
+      "epoch": 2.035164835164835,
+      "grad_norm": 0.04932479187846184,
+      "learning_rate": 0.0010749185667752444,
+      "loss": 1.0826,
+      "step": 926
+    },
+    {
+      "epoch": 2.0373626373626372,
+      "grad_norm": 0.10080870985984802,
+      "learning_rate": 0.0010724755700325733,
+      "loss": 0.8098,
+      "step": 927
+    },
+    {
+      "epoch": 2.0395604395604394,
+      "grad_norm": 0.031826019287109375,
+      "learning_rate": 0.0010700325732899024,
+      "loss": 0.7756,
+      "step": 928
+    },
+    {
+      "epoch": 2.0417582417582416,
+      "grad_norm": 0.10890044271945953,
+      "learning_rate": 0.0010675895765472313,
+      "loss": 1.03,
+      "step": 929
+    },
+    {
+      "epoch": 2.043956043956044,
+      "grad_norm": 0.0653642788529396,
+      "learning_rate": 0.0010651465798045602,
+      "loss": 0.854,
+      "step": 930
+    },
+    {
+      "epoch": 2.046153846153846,
+      "grad_norm": 0.04255935922265053,
+      "learning_rate": 0.0010627035830618893,
+      "loss": 0.9427,
+      "step": 931
+    },
+    {
+      "epoch": 2.048351648351648,
+      "grad_norm": 0.05656895786523819,
+      "learning_rate": 0.0010602605863192182,
+      "loss": 1.1377,
+      "step": 932
+    },
+    {
+      "epoch": 2.0505494505494504,
+      "grad_norm": 0.0289339330047369,
+      "learning_rate": 0.0010578175895765473,
+      "loss": 0.8788,
+      "step": 933
+    },
+    {
+      "epoch": 2.0527472527472526,
+      "grad_norm": 0.05222625285387039,
+      "learning_rate": 0.0010553745928338762,
+      "loss": 0.8029,
+      "step": 934
+    },
+    {
+      "epoch": 2.0549450549450547,
+      "grad_norm": 0.03593906760215759,
+      "learning_rate": 0.0010529315960912053,
+      "loss": 0.7533,
+      "step": 935
+    },
+    {
+      "epoch": 2.057142857142857,
+      "grad_norm": 0.024449577555060387,
+      "learning_rate": 0.0010504885993485342,
+      "loss": 0.6134,
+      "step": 936
+    },
+    {
+      "epoch": 2.059340659340659,
+      "grad_norm": 0.02460574172437191,
+      "learning_rate": 0.0010480456026058633,
+      "loss": 0.9216,
+      "step": 937
+    },
+    {
+      "epoch": 2.0615384615384613,
+      "grad_norm": 0.05404870584607124,
+      "learning_rate": 0.0010456026058631922,
+      "loss": 1.0375,
+      "step": 938
+    },
+    {
+      "epoch": 2.0637362637362635,
+      "grad_norm": 0.04211109131574631,
+      "learning_rate": 0.0010431596091205213,
+      "loss": 0.7824,
+      "step": 939
+    },
+    {
+      "epoch": 2.065934065934066,
+      "grad_norm": 0.03853631019592285,
+      "learning_rate": 0.0010407166123778502,
+      "loss": 0.8348,
+      "step": 940
+    },
+    {
+      "epoch": 2.0681318681318683,
+      "grad_norm": 0.05084240809082985,
+      "learning_rate": 0.001038273615635179,
+      "loss": 0.8772,
+      "step": 941
+    },
+    {
+      "epoch": 2.0703296703296705,
+      "grad_norm": 0.06858860701322556,
+      "learning_rate": 0.0010358306188925082,
+      "loss": 0.7902,
+      "step": 942
+    },
+    {
+      "epoch": 2.0725274725274727,
+      "grad_norm": 0.05968482047319412,
+      "learning_rate": 0.001033387622149837,
+      "loss": 0.954,
+      "step": 943
+    },
+    {
+      "epoch": 2.074725274725275,
+      "grad_norm": 0.07145629823207855,
+      "learning_rate": 0.0010309446254071662,
+      "loss": 0.8204,
+      "step": 944
+    },
+    {
+      "epoch": 2.076923076923077,
+      "grad_norm": 0.060320887714624405,
+      "learning_rate": 0.001028501628664495,
+      "loss": 0.7189,
+      "step": 945
+    },
+    {
+      "epoch": 2.0791208791208793,
+      "grad_norm": 0.17356663942337036,
+      "learning_rate": 0.0010260586319218242,
+      "loss": 0.8864,
+      "step": 946
+    },
+    {
+      "epoch": 2.0813186813186815,
+      "grad_norm": 0.12528428435325623,
+      "learning_rate": 0.001023615635179153,
+      "loss": 0.9163,
+      "step": 947
+    },
+    {
+      "epoch": 2.0835164835164837,
+      "grad_norm": 0.10495130717754364,
+      "learning_rate": 0.0010211726384364822,
+      "loss": 0.921,
+      "step": 948
+    },
+    {
+      "epoch": 2.085714285714286,
+      "grad_norm": 0.1029607281088829,
+      "learning_rate": 0.001018729641693811,
+      "loss": 0.6625,
+      "step": 949
+    },
+    {
+      "epoch": 2.087912087912088,
+      "grad_norm": 0.03561295568943024,
+      "learning_rate": 0.00101628664495114,
+      "loss": 1.006,
+      "step": 950
+    },
+    {
+      "epoch": 2.0901098901098902,
+      "grad_norm": 0.07599736005067825,
+      "learning_rate": 0.001013843648208469,
+      "loss": 0.9087,
+      "step": 951
+    },
+    {
+      "epoch": 2.0923076923076924,
+      "grad_norm": 0.03531849384307861,
+      "learning_rate": 0.001011400651465798,
+      "loss": 0.9161,
+      "step": 952
+    },
+    {
+      "epoch": 2.0945054945054946,
+      "grad_norm": 0.027006734162569046,
+      "learning_rate": 0.001008957654723127,
+      "loss": 0.8748,
+      "step": 953
+    },
+    {
+      "epoch": 2.096703296703297,
+      "grad_norm": 0.10955226421356201,
+      "learning_rate": 0.001006514657980456,
+      "loss": 0.9703,
+      "step": 954
+    },
+    {
+      "epoch": 2.098901098901099,
+      "grad_norm": 0.16619734466075897,
+      "learning_rate": 0.001004071661237785,
+      "loss": 0.8031,
+      "step": 955
+    },
+    {
+      "epoch": 2.101098901098901,
+      "grad_norm": 0.072159044444561,
+      "learning_rate": 0.001001628664495114,
+      "loss": 0.756,
+      "step": 956
+    },
+    {
+      "epoch": 2.1032967032967034,
+      "grad_norm": 0.04266434162855148,
+      "learning_rate": 0.000999185667752443,
+      "loss": 0.7678,
+      "step": 957
+    },
+    {
+      "epoch": 2.1054945054945056,
+      "grad_norm": 0.050206009298563004,
+      "learning_rate": 0.000996742671009772,
+      "loss": 0.9131,
+      "step": 958
+    },
+    {
+      "epoch": 2.1076923076923078,
+      "grad_norm": 0.07007177174091339,
+      "learning_rate": 0.0009942996742671008,
+      "loss": 0.9445,
+      "step": 959
+    },
+    {
+      "epoch": 2.10989010989011,
+      "grad_norm": 0.2698993384838104,
+      "learning_rate": 0.00099185667752443,
+      "loss": 0.9663,
+      "step": 960
+    },
+    {
+      "epoch": 2.112087912087912,
+      "grad_norm": 0.2064468264579773,
+      "learning_rate": 0.0009894136807817588,
+      "loss": 0.9282,
+      "step": 961
+    },
+    {
+      "epoch": 2.1142857142857143,
+      "grad_norm": 0.023735875263810158,
+      "learning_rate": 0.000986970684039088,
+      "loss": 0.8177,
+      "step": 962
+    },
+    {
+      "epoch": 2.1164835164835165,
+      "grad_norm": 0.03540763631463051,
+      "learning_rate": 0.0009845276872964168,
+      "loss": 0.861,
+      "step": 963
+    },
+    {
+      "epoch": 2.1186813186813187,
+      "grad_norm": 0.03265935182571411,
+      "learning_rate": 0.000982084690553746,
+      "loss": 0.767,
+      "step": 964
+    },
+    {
+      "epoch": 2.120879120879121,
+      "grad_norm": 0.08144727349281311,
+      "learning_rate": 0.0009796416938110748,
+      "loss": 0.6914,
+      "step": 965
+    },
+    {
+      "epoch": 2.123076923076923,
+      "grad_norm": 0.03306640312075615,
+      "learning_rate": 0.000977198697068404,
+      "loss": 0.8337,
+      "step": 966
+    },
+    {
+      "epoch": 2.1252747252747253,
+      "grad_norm": 0.293979287147522,
+      "learning_rate": 0.0009747557003257328,
+      "loss": 0.8971,
+      "step": 967
+    },
+    {
+      "epoch": 2.1274725274725275,
+      "grad_norm": 0.09393005818128586,
+      "learning_rate": 0.0009723127035830618,
+      "loss": 1.0027,
+      "step": 968
+    },
+    {
+      "epoch": 2.1296703296703297,
+      "grad_norm": 0.05901549011468887,
+      "learning_rate": 0.0009698697068403908,
+      "loss": 0.8013,
+      "step": 969
+    },
+    {
+      "epoch": 2.131868131868132,
+      "grad_norm": 0.11195430904626846,
+      "learning_rate": 0.0009674267100977198,
+      "loss": 0.9277,
+      "step": 970
+    },
+    {
+      "epoch": 2.134065934065934,
+      "grad_norm": 0.0625724196434021,
+      "learning_rate": 0.0009649837133550488,
+      "loss": 0.7569,
+      "step": 971
+    },
+    {
+      "epoch": 2.1362637362637362,
+      "grad_norm": 0.023438269272446632,
+      "learning_rate": 0.0009625407166123778,
+      "loss": 0.7258,
+      "step": 972
+    },
+    {
+      "epoch": 2.1384615384615384,
+      "grad_norm": 0.031200572848320007,
+      "learning_rate": 0.0009600977198697067,
+      "loss": 0.7449,
+      "step": 973
+    },
+    {
+      "epoch": 2.1406593406593406,
+      "grad_norm": 0.04762737452983856,
+      "learning_rate": 0.0009576547231270357,
+      "loss": 0.7981,
+      "step": 974
+    },
+    {
+      "epoch": 2.142857142857143,
+      "grad_norm": 0.042736753821372986,
+      "learning_rate": 0.0009552117263843647,
+      "loss": 0.8629,
+      "step": 975
+    },
+    {
+      "epoch": 2.145054945054945,
+      "grad_norm": 0.02684307098388672,
+      "learning_rate": 0.0009527687296416937,
+      "loss": 0.6771,
+      "step": 976
+    },
+    {
+      "epoch": 2.147252747252747,
+      "grad_norm": 0.13796961307525635,
+      "learning_rate": 0.0009503257328990227,
+      "loss": 0.8339,
+      "step": 977
+    },
+    {
+      "epoch": 2.1494505494505494,
+      "grad_norm": 0.036969803273677826,
+      "learning_rate": 0.0009478827361563517,
+      "loss": 0.9153,
+      "step": 978
+    },
+    {
+      "epoch": 2.1516483516483516,
+      "grad_norm": 0.0828632190823555,
+      "learning_rate": 0.0009454397394136807,
+      "loss": 0.9517,
+      "step": 979
+    },
+    {
+      "epoch": 2.1538461538461537,
+      "grad_norm": 0.021827332675457,
+      "learning_rate": 0.0009429967426710097,
+      "loss": 0.7293,
+      "step": 980
+    },
+    {
+      "epoch": 2.156043956043956,
+      "grad_norm": 0.07524440437555313,
+      "learning_rate": 0.0009405537459283387,
+      "loss": 0.8766,
+      "step": 981
+    },
+    {
+      "epoch": 2.158241758241758,
+      "grad_norm": 0.03788432106375694,
+      "learning_rate": 0.0009381107491856677,
+      "loss": 0.8301,
+      "step": 982
+    },
+    {
+      "epoch": 2.1604395604395603,
+      "grad_norm": 0.04614375904202461,
+      "learning_rate": 0.0009356677524429969,
+      "loss": 0.8868,
+      "step": 983
+    },
+    {
+      "epoch": 2.1626373626373625,
+      "grad_norm": 0.03925676271319389,
+      "learning_rate": 0.0009332247557003259,
+      "loss": 0.8926,
+      "step": 984
+    },
+    {
+      "epoch": 2.1648351648351647,
+      "grad_norm": 0.06194640323519707,
+      "learning_rate": 0.0009307817589576549,
+      "loss": 0.9268,
+      "step": 985
+    },
+    {
+      "epoch": 2.167032967032967,
+      "grad_norm": 0.07021873444318771,
+      "learning_rate": 0.0009283387622149839,
+      "loss": 0.8862,
+      "step": 986
+    },
+    {
+      "epoch": 2.169230769230769,
+      "grad_norm": 0.02965611219406128,
+      "learning_rate": 0.0009258957654723129,
+      "loss": 1.011,
+      "step": 987
+    },
+    {
+      "epoch": 2.1714285714285713,
+      "grad_norm": 0.057836856693029404,
+      "learning_rate": 0.0009234527687296417,
+      "loss": 1.0048,
+      "step": 988
+    },
+    {
+      "epoch": 2.1736263736263735,
+      "grad_norm": 0.10224360972642899,
+      "learning_rate": 0.0009210097719869707,
+      "loss": 0.821,
+      "step": 989
+    },
+    {
+      "epoch": 2.1758241758241756,
+      "grad_norm": 0.024154750630259514,
+      "learning_rate": 0.0009185667752442997,
+      "loss": 0.7692,
+      "step": 990
+    },
+    {
+      "epoch": 2.178021978021978,
+      "grad_norm": 0.04337792471051216,
+      "learning_rate": 0.0009161237785016287,
+      "loss": 0.9121,
+      "step": 991
+    },
+    {
+      "epoch": 2.18021978021978,
+      "grad_norm": 0.14515343308448792,
+      "learning_rate": 0.0009136807817589577,
+      "loss": 0.8069,
+      "step": 992
+    },
+    {
+      "epoch": 2.182417582417582,
+      "grad_norm": 0.030815977603197098,
+      "learning_rate": 0.0009112377850162867,
+      "loss": 0.8569,
+      "step": 993
+    },
+    {
+      "epoch": 2.184615384615385,
+      "grad_norm": 0.05467939004302025,
+      "learning_rate": 0.0009087947882736157,
+      "loss": 1.118,
+      "step": 994
+    },
+    {
+      "epoch": 2.186813186813187,
+      "grad_norm": 0.16391274333000183,
+      "learning_rate": 0.0009063517915309447,
+      "loss": 0.8405,
+      "step": 995
+    },
+    {
+      "epoch": 2.1890109890109892,
+      "grad_norm": 0.08046643435955048,
+      "learning_rate": 0.0009039087947882737,
+      "loss": 0.7395,
+      "step": 996
+    },
+    {
+      "epoch": 2.1912087912087914,
+      "grad_norm": 0.0754573792219162,
+      "learning_rate": 0.0009014657980456026,
+      "loss": 0.7854,
+      "step": 997
+    },
+    {
+      "epoch": 2.1934065934065936,
+      "grad_norm": 0.052553098648786545,
+      "learning_rate": 0.0008990228013029316,
+      "loss": 1.0121,
+      "step": 998
+    },
+    {
+      "epoch": 2.195604395604396,
+      "grad_norm": 0.026032844558358192,
+      "learning_rate": 0.0008965798045602606,
+      "loss": 0.764,
+      "step": 999
+    },
+    {
+      "epoch": 2.197802197802198,
+      "grad_norm": 0.044275086373090744,
+      "learning_rate": 0.0008941368078175896,
+      "loss": 0.9257,
+      "step": 1000
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.02454984374344349,
+      "learning_rate": 0.0008916938110749186,
+      "loss": 0.8286,
+      "step": 1001
+    },
+    {
+      "epoch": 2.2021978021978024,
+      "grad_norm": 0.15191242098808289,
+      "learning_rate": 0.0008892508143322476,
+      "loss": 0.9993,
+      "step": 1002
+    },
+    {
+      "epoch": 2.2043956043956046,
+      "grad_norm": 0.0429522730410099,
+      "learning_rate": 0.0008868078175895766,
+      "loss": 0.7462,
+      "step": 1003
+    },
+    {
+      "epoch": 2.2065934065934067,
+      "grad_norm": 0.08294845372438431,
+      "learning_rate": 0.0008843648208469056,
+      "loss": 0.9788,
+      "step": 1004
+    },
+    {
+      "epoch": 2.208791208791209,
+      "grad_norm": 0.10757332295179367,
+      "learning_rate": 0.0008819218241042346,
+      "loss": 0.9253,
+      "step": 1005
+    },
+    {
+      "epoch": 2.210989010989011,
+      "grad_norm": 0.03667830675840378,
+      "learning_rate": 0.0008794788273615636,
+      "loss": 0.863,
+      "step": 1006
+    },
+    {
+      "epoch": 2.2131868131868133,
+      "grad_norm": 0.04648243263363838,
+      "learning_rate": 0.0008770358306188925,
+      "loss": 0.7762,
+      "step": 1007
+    },
+    {
+      "epoch": 2.2153846153846155,
+      "grad_norm": 0.04681772366166115,
+      "learning_rate": 0.0008745928338762215,
+      "loss": 0.7675,
+      "step": 1008
+    },
+    {
+      "epoch": 2.2175824175824177,
+      "grad_norm": 0.02563762664794922,
+      "learning_rate": 0.0008721498371335505,
+      "loss": 0.7769,
+      "step": 1009
+    },
+    {
+      "epoch": 2.21978021978022,
+      "grad_norm": 0.05033525824546814,
+      "learning_rate": 0.0008697068403908795,
+      "loss": 0.7828,
+      "step": 1010
+    },
+    {
+      "epoch": 2.221978021978022,
+      "grad_norm": 0.08197668194770813,
+      "learning_rate": 0.0008672638436482085,
+      "loss": 0.908,
+      "step": 1011
+    },
+    {
+      "epoch": 2.2241758241758243,
+      "grad_norm": 0.3260814845561981,
+      "learning_rate": 0.0008648208469055375,
+      "loss": 0.9025,
+      "step": 1012
+    },
+    {
+      "epoch": 2.2263736263736265,
+      "grad_norm": 0.02843194641172886,
+      "learning_rate": 0.0008623778501628665,
+      "loss": 0.8806,
+      "step": 1013
+    },
+    {
+      "epoch": 2.2285714285714286,
+      "grad_norm": 0.022218836471438408,
+      "learning_rate": 0.0008599348534201955,
+      "loss": 0.843,
+      "step": 1014
+    },
+    {
+      "epoch": 2.230769230769231,
+      "grad_norm": 0.04393913969397545,
+      "learning_rate": 0.0008574918566775245,
+      "loss": 0.8741,
+      "step": 1015
+    },
+    {
+      "epoch": 2.232967032967033,
+      "grad_norm": 0.022267989814281464,
+      "learning_rate": 0.0008550488599348534,
+      "loss": 0.8752,
+      "step": 1016
+    },
+    {
+      "epoch": 2.235164835164835,
+      "grad_norm": 0.05330319330096245,
+      "learning_rate": 0.0008526058631921824,
+      "loss": 0.7616,
+      "step": 1017
+    },
+    {
+      "epoch": 2.2373626373626374,
+      "grad_norm": 0.038530874997377396,
+      "learning_rate": 0.0008501628664495114,
+      "loss": 0.8903,
+      "step": 1018
+    },
+    {
+      "epoch": 2.2395604395604396,
+      "grad_norm": 0.027849892154335976,
+      "learning_rate": 0.0008477198697068404,
+      "loss": 0.7509,
+      "step": 1019
+    },
+    {
+      "epoch": 2.241758241758242,
+      "grad_norm": 0.04720039293169975,
+      "learning_rate": 0.0008452768729641694,
+      "loss": 0.9113,
+      "step": 1020
+    },
+    {
+      "epoch": 2.243956043956044,
+      "grad_norm": 0.019380079582333565,
+      "learning_rate": 0.0008428338762214984,
+      "loss": 0.8444,
+      "step": 1021
+    },
+    {
+      "epoch": 2.246153846153846,
+      "grad_norm": 0.0780884325504303,
+      "learning_rate": 0.0008403908794788274,
+      "loss": 0.8821,
+      "step": 1022
+    },
+    {
+      "epoch": 2.2483516483516484,
+      "grad_norm": 0.048055682331323624,
+      "learning_rate": 0.0008379478827361564,
+      "loss": 0.7898,
+      "step": 1023
+    },
+    {
+      "epoch": 2.2505494505494505,
+      "grad_norm": 0.030040672048926353,
+      "learning_rate": 0.0008355048859934854,
+      "loss": 0.7622,
+      "step": 1024
+    },
+    {
+      "epoch": 2.2527472527472527,
+      "grad_norm": 0.14451251924037933,
+      "learning_rate": 0.0008330618892508144,
+      "loss": 0.8157,
+      "step": 1025
+    },
+    {
+      "epoch": 2.254945054945055,
+      "grad_norm": 0.06233696639537811,
+      "learning_rate": 0.0008306188925081433,
+      "loss": 0.8602,
+      "step": 1026
+    },
+    {
+      "epoch": 2.257142857142857,
+      "grad_norm": 0.04990050941705704,
+      "learning_rate": 0.0008281758957654723,
+      "loss": 0.9462,
+      "step": 1027
+    },
+    {
+      "epoch": 2.2593406593406593,
+      "grad_norm": 0.03543087840080261,
+      "learning_rate": 0.0008257328990228013,
+      "loss": 0.7871,
+      "step": 1028
+    },
+    {
+      "epoch": 2.2615384615384615,
+      "grad_norm": 0.028012089431285858,
+      "learning_rate": 0.0008232899022801303,
+      "loss": 0.7938,
+      "step": 1029
+    },
+    {
+      "epoch": 2.2637362637362637,
+      "grad_norm": 0.04357071965932846,
+      "learning_rate": 0.0008208469055374593,
+      "loss": 0.8529,
+      "step": 1030
+    },
+    {
+      "epoch": 2.265934065934066,
+      "grad_norm": 0.030966050922870636,
+      "learning_rate": 0.0008184039087947883,
+      "loss": 0.9758,
+      "step": 1031
+    },
+    {
+      "epoch": 2.268131868131868,
+      "grad_norm": 0.03538753464818001,
+      "learning_rate": 0.0008159609120521173,
+      "loss": 0.8232,
+      "step": 1032
+    },
+    {
+      "epoch": 2.2703296703296703,
+      "grad_norm": 0.038866158574819565,
+      "learning_rate": 0.0008135179153094463,
+      "loss": 0.9777,
+      "step": 1033
+    },
+    {
+      "epoch": 2.2725274725274724,
+      "grad_norm": 0.026308350265026093,
+      "learning_rate": 0.0008110749185667753,
+      "loss": 0.8613,
+      "step": 1034
+    },
+    {
+      "epoch": 2.2747252747252746,
+      "grad_norm": 0.029214028269052505,
+      "learning_rate": 0.0008086319218241043,
+      "loss": 0.8735,
+      "step": 1035
+    },
+    {
+      "epoch": 2.276923076923077,
+      "grad_norm": 0.03896724060177803,
+      "learning_rate": 0.0008061889250814332,
+      "loss": 0.7012,
+      "step": 1036
+    },
+    {
+      "epoch": 2.279120879120879,
+      "grad_norm": 0.04351063817739487,
+      "learning_rate": 0.0008037459283387622,
+      "loss": 0.7107,
+      "step": 1037
+    },
+    {
+      "epoch": 2.281318681318681,
+      "grad_norm": 0.13533365726470947,
+      "learning_rate": 0.0008013029315960912,
+      "loss": 0.8096,
+      "step": 1038
+    },
+    {
+      "epoch": 2.2835164835164834,
+      "grad_norm": 0.07058882713317871,
+      "learning_rate": 0.0007988599348534202,
+      "loss": 0.945,
+      "step": 1039
+    },
+    {
+      "epoch": 2.2857142857142856,
+      "grad_norm": 0.027355186641216278,
+      "learning_rate": 0.0007964169381107492,
+      "loss": 0.8354,
+      "step": 1040
+    },
+    {
+      "epoch": 2.2879120879120878,
+      "grad_norm": 0.027744503691792488,
+      "learning_rate": 0.0007939739413680782,
+      "loss": 0.7826,
+      "step": 1041
+    },
+    {
+      "epoch": 2.29010989010989,
+      "grad_norm": 0.06831622868776321,
+      "learning_rate": 0.0007915309446254072,
+      "loss": 0.8871,
+      "step": 1042
+    },
+    {
+      "epoch": 2.292307692307692,
+      "grad_norm": 0.04831789806485176,
+      "learning_rate": 0.0007890879478827362,
+      "loss": 0.7789,
+      "step": 1043
+    },
+    {
+      "epoch": 2.2945054945054943,
+      "grad_norm": 0.02339034341275692,
+      "learning_rate": 0.0007866449511400652,
+      "loss": 0.6735,
+      "step": 1044
+    },
+    {
+      "epoch": 2.2967032967032965,
+      "grad_norm": 0.037956297397613525,
+      "learning_rate": 0.0007842019543973941,
+      "loss": 0.8263,
+      "step": 1045
+    },
+    {
+      "epoch": 2.2989010989010987,
+      "grad_norm": 0.04281678423285484,
+      "learning_rate": 0.0007817589576547231,
+      "loss": 0.7632,
+      "step": 1046
+    },
+    {
+      "epoch": 2.301098901098901,
+      "grad_norm": 0.1776205152273178,
+      "learning_rate": 0.0007793159609120521,
+      "loss": 0.8271,
+      "step": 1047
+    },
+    {
+      "epoch": 2.303296703296703,
+      "grad_norm": 0.02817639335989952,
+      "learning_rate": 0.0007768729641693811,
+      "loss": 0.6427,
+      "step": 1048
+    },
+    {
+      "epoch": 2.3054945054945053,
+      "grad_norm": 0.12283371388912201,
+      "learning_rate": 0.0007744299674267101,
+      "loss": 0.637,
+      "step": 1049
+    },
+    {
+      "epoch": 2.3076923076923075,
+      "grad_norm": 0.032688308507204056,
+      "learning_rate": 0.0007719869706840391,
+      "loss": 0.7786,
+      "step": 1050
+    },
+    {
+      "epoch": 2.3098901098901097,
+      "grad_norm": 0.028588132932782173,
+      "learning_rate": 0.0007695439739413681,
+      "loss": 0.84,
+      "step": 1051
+    },
+    {
+      "epoch": 2.312087912087912,
+      "grad_norm": 0.03618638962507248,
+      "learning_rate": 0.0007671009771986971,
+      "loss": 0.6893,
+      "step": 1052
+    },
+    {
+      "epoch": 2.314285714285714,
+      "grad_norm": 0.04124718904495239,
+      "learning_rate": 0.0007646579804560261,
+      "loss": 0.7963,
+      "step": 1053
+    },
+    {
+      "epoch": 2.3164835164835162,
+      "grad_norm": 0.07241242378950119,
+      "learning_rate": 0.0007622149837133551,
+      "loss": 0.9066,
+      "step": 1054
+    },
+    {
+      "epoch": 2.3186813186813184,
+      "grad_norm": 0.03284353390336037,
+      "learning_rate": 0.000759771986970684,
+      "loss": 0.7314,
+      "step": 1055
+    },
+    {
+      "epoch": 2.320879120879121,
+      "grad_norm": 0.08500389009714127,
+      "learning_rate": 0.000757328990228013,
+      "loss": 0.8107,
+      "step": 1056
+    },
+    {
+      "epoch": 2.3230769230769233,
+      "grad_norm": 0.06767614930868149,
+      "learning_rate": 0.000754885993485342,
+      "loss": 0.6744,
+      "step": 1057
+    },
+    {
+      "epoch": 2.3252747252747255,
+      "grad_norm": 0.04953731596469879,
+      "learning_rate": 0.000752442996742671,
+      "loss": 0.7306,
+      "step": 1058
+    },
+    {
+      "epoch": 2.3274725274725276,
+      "grad_norm": 0.019502393901348114,
+      "learning_rate": 0.00075,
+      "loss": 0.7145,
+      "step": 1059
+    },
+    {
+      "epoch": 2.32967032967033,
+      "grad_norm": 0.02289710007607937,
+      "learning_rate": 0.000747557003257329,
+      "loss": 0.7544,
+      "step": 1060
+    },
+    {
+      "epoch": 2.331868131868132,
+      "grad_norm": 0.03434208035469055,
+      "learning_rate": 0.000745114006514658,
+      "loss": 0.7272,
+      "step": 1061
+    },
+    {
+      "epoch": 2.334065934065934,
+      "grad_norm": 0.07401902973651886,
+      "learning_rate": 0.000742671009771987,
+      "loss": 0.7821,
+      "step": 1062
+    },
+    {
+      "epoch": 2.3362637362637364,
+      "grad_norm": 0.03183607757091522,
+      "learning_rate": 0.000740228013029316,
+      "loss": 0.9612,
+      "step": 1063
+    },
+    {
+      "epoch": 2.3384615384615386,
+      "grad_norm": 0.03336568921804428,
+      "learning_rate": 0.0007377850162866449,
+      "loss": 0.8182,
+      "step": 1064
+    },
+    {
+      "epoch": 2.340659340659341,
+      "grad_norm": 0.07736802846193314,
+      "learning_rate": 0.0007353420195439739,
+      "loss": 0.9117,
+      "step": 1065
+    },
+    {
+      "epoch": 2.342857142857143,
+      "grad_norm": 0.1052742674946785,
+      "learning_rate": 0.0007328990228013029,
+      "loss": 0.8114,
+      "step": 1066
+    },
+    {
+      "epoch": 2.345054945054945,
+      "grad_norm": 0.022170396521687508,
+      "learning_rate": 0.0007304560260586319,
+      "loss": 0.8133,
+      "step": 1067
+    },
+    {
+      "epoch": 2.3472527472527474,
+      "grad_norm": 0.03888809308409691,
+      "learning_rate": 0.0007280130293159609,
+      "loss": 0.8745,
+      "step": 1068
+    },
+    {
+      "epoch": 2.3494505494505495,
+      "grad_norm": 0.034501928836107254,
+      "learning_rate": 0.0007255700325732899,
+      "loss": 0.9215,
+      "step": 1069
+    },
+    {
+      "epoch": 2.3516483516483517,
+      "grad_norm": 0.059524763375520706,
+      "learning_rate": 0.0007231270358306189,
+      "loss": 0.9247,
+      "step": 1070
+    },
+    {
+      "epoch": 2.353846153846154,
+      "grad_norm": 0.061633676290512085,
+      "learning_rate": 0.0007206840390879479,
+      "loss": 0.9112,
+      "step": 1071
+    },
+    {
+      "epoch": 2.356043956043956,
+      "grad_norm": 0.05481690540909767,
+      "learning_rate": 0.0007182410423452769,
+      "loss": 0.8023,
+      "step": 1072
+    },
+    {
+      "epoch": 2.3582417582417583,
+      "grad_norm": 0.03949303552508354,
+      "learning_rate": 0.0007157980456026059,
+      "loss": 0.7084,
+      "step": 1073
+    },
+    {
+      "epoch": 2.3604395604395605,
+      "grad_norm": 0.06559166312217712,
+      "learning_rate": 0.0007133550488599348,
+      "loss": 0.8097,
+      "step": 1074
+    },
+    {
+      "epoch": 2.3626373626373627,
+      "grad_norm": 0.04862130805850029,
+      "learning_rate": 0.0007109120521172638,
+      "loss": 0.8805,
+      "step": 1075
+    },
+    {
+      "epoch": 2.364835164835165,
+      "grad_norm": 0.037911705672740936,
+      "learning_rate": 0.0007084690553745928,
+      "loss": 0.8593,
+      "step": 1076
+    },
+    {
+      "epoch": 2.367032967032967,
+      "grad_norm": 0.05788146331906319,
+      "learning_rate": 0.0007060260586319218,
+      "loss": 0.654,
+      "step": 1077
+    },
+    {
+      "epoch": 2.3692307692307693,
+      "grad_norm": 0.06548113375902176,
+      "learning_rate": 0.0007035830618892508,
+      "loss": 0.8776,
+      "step": 1078
+    },
+    {
+      "epoch": 2.3714285714285714,
+      "grad_norm": 0.10516357421875,
+      "learning_rate": 0.0007011400651465798,
+      "loss": 0.7225,
+      "step": 1079
+    },
+    {
+      "epoch": 2.3736263736263736,
+      "grad_norm": 0.03788142278790474,
+      "learning_rate": 0.0006986970684039088,
+      "loss": 0.8585,
+      "step": 1080
+    },
+    {
+      "epoch": 2.375824175824176,
+      "grad_norm": 0.03583267331123352,
+      "learning_rate": 0.0006962540716612378,
+      "loss": 0.8773,
+      "step": 1081
+    },
+    {
+      "epoch": 2.378021978021978,
+      "grad_norm": 0.038455914705991745,
+      "learning_rate": 0.0006938110749185668,
+      "loss": 0.8285,
+      "step": 1082
+    },
+    {
+      "epoch": 2.38021978021978,
+      "grad_norm": 0.039770398288965225,
+      "learning_rate": 0.0006913680781758957,
+      "loss": 1.0342,
+      "step": 1083
+    },
+    {
+      "epoch": 2.3824175824175824,
+      "grad_norm": 0.045302752405405045,
+      "learning_rate": 0.0006889250814332247,
+      "loss": 0.8802,
+      "step": 1084
+    },
+    {
+      "epoch": 2.3846153846153846,
+      "grad_norm": 0.08415472507476807,
+      "learning_rate": 0.0006864820846905537,
+      "loss": 0.8513,
+      "step": 1085
+    },
+    {
+      "epoch": 2.3868131868131868,
+      "grad_norm": 0.02417987398803234,
+      "learning_rate": 0.0006840390879478827,
+      "loss": 0.8607,
+      "step": 1086
+    },
+    {
+      "epoch": 2.389010989010989,
+      "grad_norm": 0.06358985602855682,
+      "learning_rate": 0.0006815960912052117,
+      "loss": 0.8242,
+      "step": 1087
+    },
+    {
+      "epoch": 2.391208791208791,
+      "grad_norm": 0.05822953209280968,
+      "learning_rate": 0.0006791530944625407,
+      "loss": 0.9501,
+      "step": 1088
+    },
+    {
+      "epoch": 2.3934065934065933,
+      "grad_norm": 0.033758360892534256,
+      "learning_rate": 0.0006767100977198697,
+      "loss": 0.8581,
+      "step": 1089
+    },
+    {
+      "epoch": 2.3956043956043955,
+      "grad_norm": 0.025544611737132072,
+      "learning_rate": 0.0006742671009771987,
+      "loss": 0.7891,
+      "step": 1090
+    },
+    {
+      "epoch": 2.3978021978021977,
+      "grad_norm": 0.024591553956270218,
+      "learning_rate": 0.0006718241042345277,
+      "loss": 0.794,
+      "step": 1091
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.025485556572675705,
+      "learning_rate": 0.0006693811074918567,
+      "loss": 0.7191,
+      "step": 1092
+    },
+    {
+      "epoch": 2.402197802197802,
+      "grad_norm": 0.038549426943063736,
+      "learning_rate": 0.0006669381107491856,
+      "loss": 0.6846,
+      "step": 1093
+    },
+    {
+      "epoch": 2.4043956043956043,
+      "grad_norm": 0.02251487784087658,
+      "learning_rate": 0.0006644951140065146,
+      "loss": 0.7523,
+      "step": 1094
+    },
+    {
+      "epoch": 2.4065934065934065,
+      "grad_norm": 0.0556732602417469,
+      "learning_rate": 0.0006620521172638436,
+      "loss": 0.7845,
+      "step": 1095
+    },
+    {
+      "epoch": 2.4087912087912087,
+      "grad_norm": 0.033649053424596786,
+      "learning_rate": 0.0006596091205211726,
+      "loss": 0.9977,
+      "step": 1096
+    },
+    {
+      "epoch": 2.410989010989011,
+      "grad_norm": 0.15322090685367584,
+      "learning_rate": 0.0006571661237785016,
+      "loss": 0.9648,
+      "step": 1097
+    },
+    {
+      "epoch": 2.413186813186813,
+      "grad_norm": 0.09284571558237076,
+      "learning_rate": 0.0006547231270358307,
+      "loss": 0.7599,
+      "step": 1098
+    },
+    {
+      "epoch": 2.4153846153846152,
+      "grad_norm": 0.05383903160691261,
+      "learning_rate": 0.0006522801302931597,
+      "loss": 0.9291,
+      "step": 1099
+    },
+    {
+      "epoch": 2.4175824175824174,
+      "grad_norm": 0.04521678015589714,
+      "learning_rate": 0.0006498371335504886,
+      "loss": 1.1029,
+      "step": 1100
+    },
+    {
+      "epoch": 2.4197802197802196,
+      "grad_norm": 0.03905361890792847,
+      "learning_rate": 0.0006473941368078176,
+      "loss": 0.8402,
+      "step": 1101
+    },
+    {
+      "epoch": 2.421978021978022,
+      "grad_norm": 0.04434172064065933,
+      "learning_rate": 0.0006449511400651466,
+      "loss": 0.9532,
+      "step": 1102
+    },
+    {
+      "epoch": 2.424175824175824,
+      "grad_norm": 0.03799767047166824,
+      "learning_rate": 0.0006425081433224756,
+      "loss": 0.8296,
+      "step": 1103
+    },
+    {
+      "epoch": 2.426373626373626,
+      "grad_norm": 0.0398729145526886,
+      "learning_rate": 0.0006400651465798046,
+      "loss": 0.6865,
+      "step": 1104
+    },
+    {
+      "epoch": 2.4285714285714284,
+      "grad_norm": 0.03377635404467583,
+      "learning_rate": 0.0006376221498371336,
+      "loss": 0.7873,
+      "step": 1105
+    },
+    {
+      "epoch": 2.430769230769231,
+      "grad_norm": 0.03745070844888687,
+      "learning_rate": 0.0006351791530944626,
+      "loss": 0.7477,
+      "step": 1106
+    },
+    {
+      "epoch": 2.432967032967033,
+      "grad_norm": 0.0686022937297821,
+      "learning_rate": 0.0006327361563517916,
+      "loss": 0.7983,
+      "step": 1107
+    },
+    {
+      "epoch": 2.4351648351648354,
+      "grad_norm": 0.034050170332193375,
+      "learning_rate": 0.0006302931596091206,
+      "loss": 0.8252,
+      "step": 1108
+    },
+    {
+      "epoch": 2.4373626373626376,
+      "grad_norm": 0.058105502277612686,
+      "learning_rate": 0.0006278501628664496,
+      "loss": 0.9314,
+      "step": 1109
+    },
+    {
+      "epoch": 2.4395604395604398,
+      "grad_norm": 0.023845497518777847,
+      "learning_rate": 0.0006254071661237785,
+      "loss": 0.8002,
+      "step": 1110
+    },
+    {
+      "epoch": 2.441758241758242,
+      "grad_norm": 0.0604868121445179,
+      "learning_rate": 0.0006229641693811075,
+      "loss": 0.7939,
+      "step": 1111
+    },
+    {
+      "epoch": 2.443956043956044,
+      "grad_norm": 0.021948397159576416,
+      "learning_rate": 0.0006205211726384365,
+      "loss": 0.7332,
+      "step": 1112
+    },
+    {
+      "epoch": 2.4461538461538463,
+      "grad_norm": 0.05694042146205902,
+      "learning_rate": 0.0006180781758957655,
+      "loss": 1.0044,
+      "step": 1113
+    },
+    {
+      "epoch": 2.4483516483516485,
+      "grad_norm": 0.024171890690922737,
+      "learning_rate": 0.0006156351791530945,
+      "loss": 0.8956,
+      "step": 1114
+    },
+    {
+      "epoch": 2.4505494505494507,
+      "grad_norm": 0.07226133346557617,
+      "learning_rate": 0.0006131921824104235,
+      "loss": 0.9913,
+      "step": 1115
+    },
+    {
+      "epoch": 2.452747252747253,
+      "grad_norm": 0.033379461616277695,
+      "learning_rate": 0.0006107491856677525,
+      "loss": 0.8414,
+      "step": 1116
+    },
+    {
+      "epoch": 2.454945054945055,
+      "grad_norm": 0.02720930054783821,
+      "learning_rate": 0.0006083061889250815,
+      "loss": 0.8799,
+      "step": 1117
+    },
+    {
+      "epoch": 2.4571428571428573,
+      "grad_norm": 0.05313517525792122,
+      "learning_rate": 0.0006058631921824105,
+      "loss": 0.8248,
+      "step": 1118
+    },
+    {
+      "epoch": 2.4593406593406595,
+      "grad_norm": 0.04378783702850342,
+      "learning_rate": 0.0006034201954397394,
+      "loss": 0.7256,
+      "step": 1119
+    },
+    {
+      "epoch": 2.4615384615384617,
+      "grad_norm": 0.06424816697835922,
+      "learning_rate": 0.0006009771986970684,
+      "loss": 0.9927,
+      "step": 1120
+    },
+    {
+      "epoch": 2.463736263736264,
+      "grad_norm": 0.031644824892282486,
+      "learning_rate": 0.0005985342019543974,
+      "loss": 0.9691,
+      "step": 1121
+    },
+    {
+      "epoch": 2.465934065934066,
+      "grad_norm": 0.05414837971329689,
+      "learning_rate": 0.0005960912052117264,
+      "loss": 0.7972,
+      "step": 1122
+    },
+    {
+      "epoch": 2.4681318681318682,
+      "grad_norm": 0.03846507519483566,
+      "learning_rate": 0.0005936482084690554,
+      "loss": 0.6849,
+      "step": 1123
+    },
+    {
+      "epoch": 2.4703296703296704,
+      "grad_norm": 0.05029500275850296,
+      "learning_rate": 0.0005912052117263844,
+      "loss": 0.8689,
+      "step": 1124
+    },
+    {
+      "epoch": 2.4725274725274726,
+      "grad_norm": 0.027290821075439453,
+      "learning_rate": 0.0005887622149837134,
+      "loss": 0.8147,
+      "step": 1125
+    },
+    {
+      "epoch": 2.474725274725275,
+      "grad_norm": 0.2626986503601074,
+      "learning_rate": 0.0005863192182410424,
+      "loss": 0.8116,
+      "step": 1126
+    },
+    {
+      "epoch": 2.476923076923077,
+      "grad_norm": 0.03862101212143898,
+      "learning_rate": 0.0005838762214983714,
+      "loss": 0.7262,
+      "step": 1127
+    },
+    {
+      "epoch": 2.479120879120879,
+      "grad_norm": 0.08320759236812592,
+      "learning_rate": 0.0005814332247557004,
+      "loss": 0.7885,
+      "step": 1128
+    },
+    {
+      "epoch": 2.4813186813186814,
+      "grad_norm": 0.0410023033618927,
+      "learning_rate": 0.0005789902280130293,
+      "loss": 0.8613,
+      "step": 1129
+    },
+    {
+      "epoch": 2.4835164835164836,
+      "grad_norm": 0.04089837521314621,
+      "learning_rate": 0.0005765472312703583,
+      "loss": 0.9518,
+      "step": 1130
+    },
+    {
+      "epoch": 2.4857142857142858,
+      "grad_norm": 0.06741631776094437,
+      "learning_rate": 0.0005741042345276873,
+      "loss": 0.8198,
+      "step": 1131
+    },
+    {
+      "epoch": 2.487912087912088,
+      "grad_norm": 0.10417001694440842,
+      "learning_rate": 0.0005716612377850163,
+      "loss": 0.8202,
+      "step": 1132
+    },
+    {
+      "epoch": 2.49010989010989,
+      "grad_norm": 0.08746853470802307,
+      "learning_rate": 0.0005692182410423453,
+      "loss": 0.8427,
+      "step": 1133
+    },
+    {
+      "epoch": 2.4923076923076923,
+      "grad_norm": 0.039591509848833084,
+      "learning_rate": 0.0005667752442996743,
+      "loss": 0.8857,
+      "step": 1134
+    },
+    {
+      "epoch": 2.4945054945054945,
+      "grad_norm": 0.05287199094891548,
+      "learning_rate": 0.0005643322475570033,
+      "loss": 1.0376,
+      "step": 1135
+    },
+    {
+      "epoch": 2.4967032967032967,
+      "grad_norm": 0.06744746118783951,
+      "learning_rate": 0.0005618892508143323,
+      "loss": 0.9911,
+      "step": 1136
+    },
+    {
+      "epoch": 2.498901098901099,
+      "grad_norm": 0.08682337403297424,
+      "learning_rate": 0.0005594462540716613,
+      "loss": 0.6674,
+      "step": 1137
+    },
+    {
+      "epoch": 2.501098901098901,
+      "grad_norm": 0.028393127024173737,
+      "learning_rate": 0.0005570032573289903,
+      "loss": 0.8352,
+      "step": 1138
+    },
+    {
+      "epoch": 2.5032967032967033,
+      "grad_norm": 0.05877330154180527,
+      "learning_rate": 0.0005545602605863192,
+      "loss": 1.1436,
+      "step": 1139
+    },
+    {
+      "epoch": 2.5054945054945055,
+      "grad_norm": 0.061341069638729095,
+      "learning_rate": 0.0005521172638436482,
+      "loss": 0.7913,
+      "step": 1140
+    },
+    {
+      "epoch": 2.5076923076923077,
+      "grad_norm": 0.03545743599534035,
+      "learning_rate": 0.0005496742671009772,
+      "loss": 0.9861,
+      "step": 1141
+    },
+    {
+      "epoch": 2.50989010989011,
+      "grad_norm": 0.05896063894033432,
+      "learning_rate": 0.0005472312703583062,
+      "loss": 0.8264,
+      "step": 1142
+    },
+    {
+      "epoch": 2.512087912087912,
+      "grad_norm": 0.0333087183535099,
+      "learning_rate": 0.0005447882736156352,
+      "loss": 0.6563,
+      "step": 1143
+    },
+    {
+      "epoch": 2.5142857142857142,
+      "grad_norm": 0.050718266516923904,
+      "learning_rate": 0.0005423452768729642,
+      "loss": 0.7675,
+      "step": 1144
+    },
+    {
+      "epoch": 2.5164835164835164,
+      "grad_norm": 0.05007106065750122,
+      "learning_rate": 0.0005399022801302932,
+      "loss": 0.8342,
+      "step": 1145
+    },
+    {
+      "epoch": 2.5186813186813186,
+      "grad_norm": 0.050470057874917984,
+      "learning_rate": 0.0005374592833876222,
+      "loss": 0.833,
+      "step": 1146
+    },
+    {
+      "epoch": 2.520879120879121,
+      "grad_norm": 0.05283505469560623,
+      "learning_rate": 0.0005350162866449512,
+      "loss": 0.7697,
+      "step": 1147
+    },
+    {
+      "epoch": 2.523076923076923,
+      "grad_norm": 0.03968755155801773,
+      "learning_rate": 0.0005325732899022801,
+      "loss": 0.8342,
+      "step": 1148
+    },
+    {
+      "epoch": 2.525274725274725,
+      "grad_norm": 0.04729480296373367,
+      "learning_rate": 0.0005301302931596091,
+      "loss": 0.7335,
+      "step": 1149
+    },
+    {
+      "epoch": 2.5274725274725274,
+      "grad_norm": 0.03558095917105675,
+      "learning_rate": 0.0005276872964169381,
+      "loss": 0.813,
+      "step": 1150
+    },
+    {
+      "epoch": 2.5296703296703296,
+      "grad_norm": 0.05578270182013512,
+      "learning_rate": 0.0005252442996742671,
+      "loss": 0.8749,
+      "step": 1151
+    },
+    {
+      "epoch": 2.5318681318681318,
+      "grad_norm": 0.0515776127576828,
+      "learning_rate": 0.0005228013029315961,
+      "loss": 0.8883,
+      "step": 1152
+    },
+    {
+      "epoch": 2.534065934065934,
+      "grad_norm": 0.04868028312921524,
+      "learning_rate": 0.0005203583061889251,
+      "loss": 0.8863,
+      "step": 1153
+    },
+    {
+      "epoch": 2.536263736263736,
+      "grad_norm": 0.10975392907857895,
+      "learning_rate": 0.0005179153094462541,
+      "loss": 0.8121,
+      "step": 1154
+    },
+    {
+      "epoch": 2.5384615384615383,
+      "grad_norm": 0.045733317732810974,
+      "learning_rate": 0.0005154723127035831,
+      "loss": 0.8911,
+      "step": 1155
+    },
+    {
+      "epoch": 2.5406593406593405,
+      "grad_norm": 0.08059952408075333,
+      "learning_rate": 0.0005130293159609121,
+      "loss": 0.7579,
+      "step": 1156
+    },
+    {
+      "epoch": 2.5428571428571427,
+      "grad_norm": 0.034604571759700775,
+      "learning_rate": 0.0005105863192182411,
+      "loss": 0.8721,
+      "step": 1157
+    },
+    {
+      "epoch": 2.545054945054945,
+      "grad_norm": 0.06737070530653,
+      "learning_rate": 0.00050814332247557,
+      "loss": 0.9207,
+      "step": 1158
+    },
+    {
+      "epoch": 2.547252747252747,
+      "grad_norm": 0.013713842257857323,
+      "learning_rate": 0.000505700325732899,
+      "loss": 0.8446,
+      "step": 1159
+    },
+    {
+      "epoch": 2.5494505494505493,
+      "grad_norm": 0.031596675515174866,
+      "learning_rate": 0.000503257328990228,
+      "loss": 0.8076,
+      "step": 1160
+    },
+    {
+      "epoch": 2.5516483516483515,
+      "grad_norm": 0.05692550912499428,
+      "learning_rate": 0.000500814332247557,
+      "loss": 1.1024,
+      "step": 1161
+    },
+    {
+      "epoch": 2.5538461538461537,
+      "grad_norm": 0.033143773674964905,
+      "learning_rate": 0.000498371335504886,
+      "loss": 0.7517,
+      "step": 1162
+    },
+    {
+      "epoch": 2.556043956043956,
+      "grad_norm": 0.04348577558994293,
+      "learning_rate": 0.000495928338762215,
+      "loss": 0.7334,
+      "step": 1163
+    },
+    {
+      "epoch": 2.558241758241758,
+      "grad_norm": 0.13318553566932678,
+      "learning_rate": 0.000493485342019544,
+      "loss": 0.6398,
+      "step": 1164
+    },
+    {
+      "epoch": 2.5604395604395602,
+      "grad_norm": 0.033309467136859894,
+      "learning_rate": 0.000491042345276873,
+      "loss": 0.7888,
+      "step": 1165
+    },
+    {
+      "epoch": 2.5626373626373624,
+      "grad_norm": 0.06921177357435226,
+      "learning_rate": 0.000488599348534202,
+      "loss": 0.8113,
+      "step": 1166
+    },
+    {
+      "epoch": 2.5648351648351646,
+      "grad_norm": 0.04111246392130852,
+      "learning_rate": 0.0004861563517915309,
+      "loss": 1.0645,
+      "step": 1167
+    },
+    {
+      "epoch": 2.567032967032967,
+      "grad_norm": 0.04049806669354439,
+      "learning_rate": 0.0004837133550488599,
+      "loss": 0.8799,
+      "step": 1168
+    },
+    {
+      "epoch": 2.569230769230769,
+      "grad_norm": 0.06337505578994751,
+      "learning_rate": 0.0004812703583061889,
+      "loss": 0.7446,
+      "step": 1169
+    },
+    {
+      "epoch": 2.571428571428571,
+      "grad_norm": 0.05321670323610306,
+      "learning_rate": 0.00047882736156351787,
+      "loss": 0.8164,
+      "step": 1170
+    },
+    {
+      "epoch": 2.5736263736263734,
+      "grad_norm": 0.045682184398174286,
+      "learning_rate": 0.00047638436482084687,
+      "loss": 0.8701,
+      "step": 1171
+    },
+    {
+      "epoch": 2.5758241758241756,
+      "grad_norm": 0.03024928830564022,
+      "learning_rate": 0.00047394136807817587,
+      "loss": 0.8954,
+      "step": 1172
+    },
+    {
+      "epoch": 2.5780219780219777,
+      "grad_norm": 0.08092090487480164,
+      "learning_rate": 0.00047149837133550487,
+      "loss": 0.9333,
+      "step": 1173
+    },
+    {
+      "epoch": 2.5802197802197804,
+      "grad_norm": 0.0415789894759655,
+      "learning_rate": 0.00046905537459283387,
+      "loss": 0.8782,
+      "step": 1174
+    },
+    {
+      "epoch": 2.5824175824175826,
+      "grad_norm": 0.0640019103884697,
+      "learning_rate": 0.0004666123778501629,
+      "loss": 1.0127,
+      "step": 1175
+    },
+    {
+      "epoch": 2.5846153846153848,
+      "grad_norm": 0.026831114664673805,
+      "learning_rate": 0.0004641693811074919,
+      "loss": 0.7675,
+      "step": 1176
+    },
+    {
+      "epoch": 2.586813186813187,
+      "grad_norm": 0.04181014001369476,
+      "learning_rate": 0.00046172638436482087,
+      "loss": 0.8774,
+      "step": 1177
+    },
+    {
+      "epoch": 2.589010989010989,
+      "grad_norm": 0.04630782827734947,
+      "learning_rate": 0.00045928338762214987,
+      "loss": 0.7349,
+      "step": 1178
+    },
+    {
+      "epoch": 2.5912087912087913,
+      "grad_norm": 0.03895869851112366,
+      "learning_rate": 0.0004568403908794789,
+      "loss": 0.6962,
+      "step": 1179
+    },
+    {
+      "epoch": 2.5934065934065935,
+      "grad_norm": 0.060818061232566833,
+      "learning_rate": 0.0004543973941368079,
+      "loss": 1.007,
+      "step": 1180
+    },
+    {
+      "epoch": 2.5956043956043957,
+      "grad_norm": 0.03854983672499657,
+      "learning_rate": 0.0004519543973941369,
+      "loss": 0.693,
+      "step": 1181
+    },
+    {
+      "epoch": 2.597802197802198,
+      "grad_norm": 0.024709226563572884,
+      "learning_rate": 0.0004495114006514658,
+      "loss": 0.7278,
+      "step": 1182
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.04345419257879257,
+      "learning_rate": 0.0004470684039087948,
+      "loss": 0.7476,
+      "step": 1183
+    },
+    {
+      "epoch": 2.6021978021978023,
+      "grad_norm": 0.02433425560593605,
+      "learning_rate": 0.0004446254071661238,
+      "loss": 0.7563,
+      "step": 1184
+    },
+    {
+      "epoch": 2.6043956043956045,
+      "grad_norm": 0.026963217183947563,
+      "learning_rate": 0.0004421824104234528,
+      "loss": 0.7241,
+      "step": 1185
+    },
+    {
+      "epoch": 2.6065934065934067,
+      "grad_norm": 0.024603791534900665,
+      "learning_rate": 0.0004397394136807818,
+      "loss": 0.8947,
+      "step": 1186
+    },
+    {
+      "epoch": 2.608791208791209,
+      "grad_norm": 0.03858330473303795,
+      "learning_rate": 0.00043729641693811077,
+      "loss": 0.8714,
+      "step": 1187
+    },
+    {
+      "epoch": 2.610989010989011,
+      "grad_norm": 0.0851096585392952,
+      "learning_rate": 0.00043485342019543977,
+      "loss": 0.7924,
+      "step": 1188
+    },
+    {
+      "epoch": 2.6131868131868132,
+      "grad_norm": 0.08407802879810333,
+      "learning_rate": 0.00043241042345276877,
+      "loss": 0.9482,
+      "step": 1189
+    },
+    {
+      "epoch": 2.6153846153846154,
+      "grad_norm": 0.0334550179541111,
+      "learning_rate": 0.00042996742671009777,
+      "loss": 0.9741,
+      "step": 1190
+    },
+    {
+      "epoch": 2.6175824175824176,
+      "grad_norm": 0.03994135558605194,
+      "learning_rate": 0.0004275244299674267,
+      "loss": 0.8008,
+      "step": 1191
+    },
+    {
+      "epoch": 2.61978021978022,
+      "grad_norm": 0.027640996500849724,
+      "learning_rate": 0.0004250814332247557,
+      "loss": 0.7552,
+      "step": 1192
+    },
+    {
+      "epoch": 2.621978021978022,
+      "grad_norm": 0.043485235422849655,
+      "learning_rate": 0.0004226384364820847,
+      "loss": 0.7406,
+      "step": 1193
+    },
+    {
+      "epoch": 2.624175824175824,
+      "grad_norm": 0.04681190475821495,
+      "learning_rate": 0.0004201954397394137,
+      "loss": 0.7832,
+      "step": 1194
+    },
+    {
+      "epoch": 2.6263736263736264,
+      "grad_norm": 0.0703037679195404,
+      "learning_rate": 0.0004177524429967427,
+      "loss": 0.8367,
+      "step": 1195
+    },
+    {
+      "epoch": 2.6285714285714286,
+      "grad_norm": 0.03132481127977371,
+      "learning_rate": 0.00041530944625407166,
+      "loss": 0.7767,
+      "step": 1196
+    },
+    {
+      "epoch": 2.6307692307692307,
+      "grad_norm": 0.02826203592121601,
+      "learning_rate": 0.00041286644951140066,
+      "loss": 0.8141,
+      "step": 1197
+    },
+    {
+      "epoch": 2.632967032967033,
+      "grad_norm": 0.02244557812809944,
+      "learning_rate": 0.00041042345276872966,
+      "loss": 0.7173,
+      "step": 1198
+    },
+    {
+      "epoch": 2.635164835164835,
+      "grad_norm": 0.04779929667711258,
+      "learning_rate": 0.00040798045602605866,
+      "loss": 0.8556,
+      "step": 1199
+    },
+    {
+      "epoch": 2.6373626373626373,
+      "grad_norm": 0.04575258120894432,
+      "learning_rate": 0.00040553745928338767,
+      "loss": 0.817,
+      "step": 1200
+    },
+    {
+      "epoch": 2.6395604395604395,
+      "grad_norm": 0.10869526863098145,
+      "learning_rate": 0.0004030944625407166,
+      "loss": 0.8213,
+      "step": 1201
+    },
+    {
+      "epoch": 2.6417582417582417,
+      "grad_norm": 0.04469895362854004,
+      "learning_rate": 0.0004006514657980456,
+      "loss": 0.9783,
+      "step": 1202
+    },
+    {
+      "epoch": 2.643956043956044,
+      "grad_norm": 0.0492338128387928,
+      "learning_rate": 0.0003982084690553746,
+      "loss": 0.8673,
+      "step": 1203
+    },
+    {
+      "epoch": 2.646153846153846,
+      "grad_norm": 0.030009517446160316,
+      "learning_rate": 0.0003957654723127036,
+      "loss": 0.7666,
+      "step": 1204
+    },
+    {
+      "epoch": 2.6483516483516483,
+      "grad_norm": 0.03115699626505375,
+      "learning_rate": 0.0003933224755700326,
+      "loss": 1.0923,
+      "step": 1205
+    },
+    {
+      "epoch": 2.6505494505494505,
+      "grad_norm": 0.03402196243405342,
+      "learning_rate": 0.00039087947882736156,
+      "loss": 0.8734,
+      "step": 1206
+    },
+    {
+      "epoch": 2.6527472527472526,
+      "grad_norm": 0.04155919700860977,
+      "learning_rate": 0.00038843648208469056,
+      "loss": 0.8937,
+      "step": 1207
+    },
+    {
+      "epoch": 2.654945054945055,
+      "grad_norm": 0.15309584140777588,
+      "learning_rate": 0.00038599348534201956,
+      "loss": 0.8713,
+      "step": 1208
+    },
+    {
+      "epoch": 2.657142857142857,
+      "grad_norm": 0.06503726541996002,
+      "learning_rate": 0.00038355048859934856,
+      "loss": 0.7003,
+      "step": 1209
+    },
+    {
+      "epoch": 2.659340659340659,
+      "grad_norm": 0.0300603024661541,
+      "learning_rate": 0.00038110749185667756,
+      "loss": 0.9133,
+      "step": 1210
+    },
+    {
+      "epoch": 2.6615384615384614,
+      "grad_norm": 0.028050631284713745,
+      "learning_rate": 0.0003786644951140065,
+      "loss": 0.7175,
+      "step": 1211
+    },
+    {
+      "epoch": 2.6637362637362636,
+      "grad_norm": 0.035878293216228485,
+      "learning_rate": 0.0003762214983713355,
+      "loss": 0.8427,
+      "step": 1212
+    },
+    {
+      "epoch": 2.665934065934066,
+      "grad_norm": 0.08893197029829025,
+      "learning_rate": 0.0003737785016286645,
+      "loss": 0.8929,
+      "step": 1213
+    },
+    {
+      "epoch": 2.668131868131868,
+      "grad_norm": 0.04400551691651344,
+      "learning_rate": 0.0003713355048859935,
+      "loss": 0.8596,
+      "step": 1214
+    },
+    {
+      "epoch": 2.67032967032967,
+      "grad_norm": 0.07465469837188721,
+      "learning_rate": 0.00036889250814332245,
+      "loss": 0.8559,
+      "step": 1215
+    },
+    {
+      "epoch": 2.672527472527473,
+      "grad_norm": 0.06599140912294388,
+      "learning_rate": 0.00036644951140065146,
+      "loss": 0.8713,
+      "step": 1216
+    },
+    {
+      "epoch": 2.674725274725275,
+      "grad_norm": 0.05834246054291725,
+      "learning_rate": 0.00036400651465798046,
+      "loss": 0.9168,
+      "step": 1217
+    },
+    {
+      "epoch": 2.676923076923077,
+      "grad_norm": 0.03897550329566002,
+      "learning_rate": 0.00036156351791530946,
+      "loss": 0.7859,
+      "step": 1218
+    },
+    {
+      "epoch": 2.6791208791208794,
+      "grad_norm": 0.03801744803786278,
+      "learning_rate": 0.00035912052117263846,
+      "loss": 0.8344,
+      "step": 1219
+    },
+    {
+      "epoch": 2.6813186813186816,
+      "grad_norm": 0.05906485393643379,
+      "learning_rate": 0.0003566775244299674,
+      "loss": 0.693,
+      "step": 1220
+    },
+    {
+      "epoch": 2.6835164835164838,
+      "grad_norm": 0.02479812130331993,
+      "learning_rate": 0.0003542345276872964,
+      "loss": 0.6826,
+      "step": 1221
+    },
+    {
+      "epoch": 2.685714285714286,
+      "grad_norm": 0.04186400771141052,
+      "learning_rate": 0.0003517915309446254,
+      "loss": 0.9882,
+      "step": 1222
+    },
+    {
+      "epoch": 2.687912087912088,
+      "grad_norm": 0.02666168101131916,
+      "learning_rate": 0.0003493485342019544,
+      "loss": 0.8284,
+      "step": 1223
+    },
+    {
+      "epoch": 2.6901098901098903,
+      "grad_norm": 0.04419548436999321,
+      "learning_rate": 0.0003469055374592834,
+      "loss": 0.816,
+      "step": 1224
+    },
+    {
+      "epoch": 2.6923076923076925,
+      "grad_norm": 0.02876039408147335,
+      "learning_rate": 0.00034446254071661235,
+      "loss": 0.9474,
+      "step": 1225
+    },
+    {
+      "epoch": 2.6945054945054947,
+      "grad_norm": 0.02653048001229763,
+      "learning_rate": 0.00034201954397394135,
+      "loss": 0.8032,
+      "step": 1226
+    },
+    {
+      "epoch": 2.696703296703297,
+      "grad_norm": 0.042959775775671005,
+      "learning_rate": 0.00033957654723127035,
+      "loss": 0.6899,
+      "step": 1227
+    },
+    {
+      "epoch": 2.698901098901099,
+      "grad_norm": 0.033319804817438126,
+      "learning_rate": 0.00033713355048859935,
+      "loss": 0.952,
+      "step": 1228
+    },
+    {
+      "epoch": 2.7010989010989013,
+      "grad_norm": 0.05659925565123558,
+      "learning_rate": 0.00033469055374592835,
+      "loss": 0.7986,
+      "step": 1229
+    },
+    {
+      "epoch": 2.7032967032967035,
+      "grad_norm": 0.0663997083902359,
+      "learning_rate": 0.0003322475570032573,
+      "loss": 0.8722,
+      "step": 1230
+    },
+    {
+      "epoch": 2.7054945054945057,
+      "grad_norm": 0.04710668697953224,
+      "learning_rate": 0.0003298045602605863,
+      "loss": 0.8303,
+      "step": 1231
+    },
+    {
+      "epoch": 2.707692307692308,
+      "grad_norm": 0.026661334559321404,
+      "learning_rate": 0.00032736156351791535,
+      "loss": 0.8264,
+      "step": 1232
+    },
+    {
+      "epoch": 2.70989010989011,
+      "grad_norm": 0.040238332003355026,
+      "learning_rate": 0.0003249185667752443,
+      "loss": 0.776,
+      "step": 1233
+    },
+    {
+      "epoch": 2.7120879120879122,
+      "grad_norm": 0.059563782066106796,
+      "learning_rate": 0.0003224755700325733,
+      "loss": 0.8483,
+      "step": 1234
+    },
+    {
+      "epoch": 2.7142857142857144,
+      "grad_norm": 0.04587594419717789,
+      "learning_rate": 0.0003200325732899023,
+      "loss": 0.837,
+      "step": 1235
+    },
+    {
+      "epoch": 2.7164835164835166,
+      "grad_norm": 0.08958257734775543,
+      "learning_rate": 0.0003175895765472313,
+      "loss": 0.7969,
+      "step": 1236
+    },
+    {
+      "epoch": 2.718681318681319,
+      "grad_norm": 0.035689424723386765,
+      "learning_rate": 0.0003151465798045603,
+      "loss": 0.9386,
+      "step": 1237
+    },
+    {
+      "epoch": 2.720879120879121,
+      "grad_norm": 0.09529056400060654,
+      "learning_rate": 0.00031270358306188925,
+      "loss": 0.6157,
+      "step": 1238
+    },
+    {
+      "epoch": 2.723076923076923,
+      "grad_norm": 0.02590990997850895,
+      "learning_rate": 0.00031026058631921825,
+      "loss": 0.934,
+      "step": 1239
+    },
+    {
+      "epoch": 2.7252747252747254,
+      "grad_norm": 0.14555901288986206,
+      "learning_rate": 0.00030781758957654725,
+      "loss": 0.8538,
+      "step": 1240
+    },
+    {
+      "epoch": 2.7274725274725276,
+      "grad_norm": 0.0576401948928833,
+      "learning_rate": 0.00030537459283387625,
+      "loss": 0.9315,
+      "step": 1241
+    },
+    {
+      "epoch": 2.7296703296703297,
+      "grad_norm": 0.032129961997270584,
+      "learning_rate": 0.00030293159609120525,
+      "loss": 0.7089,
+      "step": 1242
+    },
+    {
+      "epoch": 2.731868131868132,
+      "grad_norm": 0.040267884731292725,
+      "learning_rate": 0.0003004885993485342,
+      "loss": 0.9554,
+      "step": 1243
+    },
+    {
+      "epoch": 2.734065934065934,
+      "grad_norm": 0.05133737996220589,
+      "learning_rate": 0.0002980456026058632,
+      "loss": 0.7845,
+      "step": 1244
+    },
+    {
+      "epoch": 2.7362637362637363,
+      "grad_norm": 0.034424956887960434,
+      "learning_rate": 0.0002956026058631922,
+      "loss": 0.801,
+      "step": 1245
+    },
+    {
+      "epoch": 2.7384615384615385,
+      "grad_norm": 0.04070351645350456,
+      "learning_rate": 0.0002931596091205212,
+      "loss": 0.9044,
+      "step": 1246
+    },
+    {
+      "epoch": 2.7406593406593407,
+      "grad_norm": 0.07627566903829575,
+      "learning_rate": 0.0002907166123778502,
+      "loss": 0.9366,
+      "step": 1247
+    },
+    {
+      "epoch": 2.742857142857143,
+      "grad_norm": 0.04275214672088623,
+      "learning_rate": 0.00028827361563517914,
+      "loss": 1.0009,
+      "step": 1248
+    },
+    {
+      "epoch": 2.745054945054945,
+      "grad_norm": 0.10861328989267349,
+      "learning_rate": 0.00028583061889250814,
+      "loss": 1.0178,
+      "step": 1249
+    },
+    {
+      "epoch": 2.7472527472527473,
+      "grad_norm": 0.018796581774950027,
+      "learning_rate": 0.00028338762214983714,
+      "loss": 0.6799,
+      "step": 1250
+    },
+    {
+      "epoch": 2.7494505494505495,
+      "grad_norm": 0.06933070719242096,
+      "learning_rate": 0.00028094462540716614,
+      "loss": 0.8915,
+      "step": 1251
+    },
+    {
+      "epoch": 2.7516483516483516,
+      "grad_norm": 0.05484369397163391,
+      "learning_rate": 0.00027850162866449514,
+      "loss": 0.7772,
+      "step": 1252
+    },
+    {
+      "epoch": 2.753846153846154,
+      "grad_norm": 0.028170472010970116,
+      "learning_rate": 0.0002760586319218241,
+      "loss": 0.7806,
+      "step": 1253
+    },
+    {
+      "epoch": 2.756043956043956,
+      "grad_norm": 0.02904343418776989,
+      "learning_rate": 0.0002736156351791531,
+      "loss": 0.9183,
+      "step": 1254
+    },
+    {
+      "epoch": 2.758241758241758,
+      "grad_norm": 0.11322277784347534,
+      "learning_rate": 0.0002711726384364821,
+      "loss": 0.8806,
+      "step": 1255
+    },
+    {
+      "epoch": 2.7604395604395604,
+      "grad_norm": 0.04259241372346878,
+      "learning_rate": 0.0002687296416938111,
+      "loss": 0.8536,
+      "step": 1256
+    },
+    {
+      "epoch": 2.7626373626373626,
+      "grad_norm": 0.10773425549268723,
+      "learning_rate": 0.00026628664495114004,
+      "loss": 0.8438,
+      "step": 1257
+    },
+    {
+      "epoch": 2.764835164835165,
+      "grad_norm": 0.07373189926147461,
+      "learning_rate": 0.00026384364820846904,
+      "loss": 0.937,
+      "step": 1258
+    },
+    {
+      "epoch": 2.767032967032967,
+      "grad_norm": 0.029525410383939743,
+      "learning_rate": 0.00026140065146579804,
+      "loss": 0.7947,
+      "step": 1259
+    },
+    {
+      "epoch": 2.769230769230769,
+      "grad_norm": 0.19173263013362885,
+      "learning_rate": 0.00025895765472312704,
+      "loss": 0.8199,
+      "step": 1260
+    },
+    {
+      "epoch": 2.7714285714285714,
+      "grad_norm": 0.013786066323518753,
+      "learning_rate": 0.00025651465798045604,
+      "loss": 0.7028,
+      "step": 1261
+    },
+    {
+      "epoch": 2.7736263736263735,
+      "grad_norm": 0.033264271914958954,
+      "learning_rate": 0.000254071661237785,
+      "loss": 0.6296,
+      "step": 1262
+    },
+    {
+      "epoch": 2.7758241758241757,
+      "grad_norm": 0.018916914239525795,
+      "learning_rate": 0.000251628664495114,
+      "loss": 0.7974,
+      "step": 1263
+    },
+    {
+      "epoch": 2.778021978021978,
+      "grad_norm": 0.02991408109664917,
+      "learning_rate": 0.000249185667752443,
+      "loss": 0.7045,
+      "step": 1264
+    },
+    {
+      "epoch": 2.78021978021978,
+      "grad_norm": 0.030526209622621536,
+      "learning_rate": 0.000246742671009772,
+      "loss": 0.6476,
+      "step": 1265
+    },
+    {
+      "epoch": 2.7824175824175823,
+      "grad_norm": 0.12691612541675568,
+      "learning_rate": 0.000244299674267101,
+      "loss": 0.7377,
+      "step": 1266
+    },
+    {
+      "epoch": 2.7846153846153845,
+      "grad_norm": 0.06126946210861206,
+      "learning_rate": 0.00024185667752442996,
+      "loss": 0.8761,
+      "step": 1267
+    },
+    {
+      "epoch": 2.7868131868131867,
+      "grad_norm": 0.04359165579080582,
+      "learning_rate": 0.00023941368078175893,
+      "loss": 0.9893,
+      "step": 1268
+    },
+    {
+      "epoch": 2.789010989010989,
+      "grad_norm": 0.031718455255031586,
+      "learning_rate": 0.00023697068403908794,
+      "loss": 0.7826,
+      "step": 1269
+    },
+    {
+      "epoch": 2.791208791208791,
+      "grad_norm": 0.03431783989071846,
+      "learning_rate": 0.00023452768729641694,
+      "loss": 0.9468,
+      "step": 1270
+    },
+    {
+      "epoch": 2.7934065934065933,
+      "grad_norm": 0.040179427713155746,
+      "learning_rate": 0.00023208469055374596,
+      "loss": 1.0704,
+      "step": 1271
+    },
+    {
+      "epoch": 2.7956043956043954,
+      "grad_norm": 0.023270566016435623,
+      "learning_rate": 0.00022964169381107494,
+      "loss": 0.7969,
+      "step": 1272
+    },
+    {
+      "epoch": 2.7978021978021976,
+      "grad_norm": 0.018464533612132072,
+      "learning_rate": 0.00022719869706840394,
+      "loss": 0.7997,
+      "step": 1273
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.04456184431910515,
+      "learning_rate": 0.0002247557003257329,
+      "loss": 0.9355,
+      "step": 1274
+    },
+    {
+      "epoch": 2.802197802197802,
+      "grad_norm": 0.04683202505111694,
+      "learning_rate": 0.0002223127035830619,
+      "loss": 0.9001,
+      "step": 1275
+    },
+    {
+      "epoch": 2.804395604395604,
+      "grad_norm": 0.052037205547094345,
+      "learning_rate": 0.0002198697068403909,
+      "loss": 0.7361,
+      "step": 1276
+    },
+    {
+      "epoch": 2.8065934065934064,
+      "grad_norm": 0.03237243369221687,
+      "learning_rate": 0.00021742671009771988,
+      "loss": 0.7725,
+      "step": 1277
+    },
+    {
+      "epoch": 2.8087912087912086,
+      "grad_norm": 0.049339499324560165,
+      "learning_rate": 0.00021498371335504888,
+      "loss": 0.8268,
+      "step": 1278
+    },
+    {
+      "epoch": 2.8109890109890108,
+      "grad_norm": 0.07552281767129898,
+      "learning_rate": 0.00021254071661237786,
+      "loss": 0.8204,
+      "step": 1279
+    },
+    {
+      "epoch": 2.813186813186813,
+      "grad_norm": 0.06387704610824585,
+      "learning_rate": 0.00021009771986970686,
+      "loss": 0.7958,
+      "step": 1280
+    },
+    {
+      "epoch": 2.815384615384615,
+      "grad_norm": 0.05347391590476036,
+      "learning_rate": 0.00020765472312703583,
+      "loss": 0.8616,
+      "step": 1281
+    },
+    {
+      "epoch": 2.8175824175824173,
+      "grad_norm": 0.10428684949874878,
+      "learning_rate": 0.00020521172638436483,
+      "loss": 0.7987,
+      "step": 1282
+    },
+    {
+      "epoch": 2.8197802197802195,
+      "grad_norm": 0.10005024820566177,
+      "learning_rate": 0.00020276872964169383,
+      "loss": 0.8893,
+      "step": 1283
+    },
+    {
+      "epoch": 2.8219780219780217,
+      "grad_norm": 0.07185282558202744,
+      "learning_rate": 0.0002003257328990228,
+      "loss": 0.8785,
+      "step": 1284
+    },
+    {
+      "epoch": 2.824175824175824,
+      "grad_norm": 0.04143763333559036,
+      "learning_rate": 0.0001978827361563518,
+      "loss": 0.7968,
+      "step": 1285
+    },
+    {
+      "epoch": 2.826373626373626,
+      "grad_norm": 0.03309214115142822,
+      "learning_rate": 0.00019543973941368078,
+      "loss": 0.6825,
+      "step": 1286
+    },
+    {
+      "epoch": 2.8285714285714287,
+      "grad_norm": 0.04666907712817192,
+      "learning_rate": 0.00019299674267100978,
+      "loss": 0.6885,
+      "step": 1287
+    },
+    {
+      "epoch": 2.830769230769231,
+      "grad_norm": 0.03849129006266594,
+      "learning_rate": 0.00019055374592833878,
+      "loss": 0.7848,
+      "step": 1288
+    },
+    {
+      "epoch": 2.832967032967033,
+      "grad_norm": 0.15401321649551392,
+      "learning_rate": 0.00018811074918566775,
+      "loss": 0.692,
+      "step": 1289
+    },
+    {
+      "epoch": 2.8351648351648353,
+      "grad_norm": 0.048793528228998184,
+      "learning_rate": 0.00018566775244299675,
+      "loss": 0.6702,
+      "step": 1290
+    },
+    {
+      "epoch": 2.8373626373626375,
+      "grad_norm": 0.05848393961787224,
+      "learning_rate": 0.00018322475570032573,
+      "loss": 0.7404,
+      "step": 1291
+    },
+    {
+      "epoch": 2.8395604395604397,
+      "grad_norm": 0.04924844577908516,
+      "learning_rate": 0.00018078175895765473,
+      "loss": 0.9426,
+      "step": 1292
+    },
+    {
+      "epoch": 2.841758241758242,
+      "grad_norm": 0.16004246473312378,
+      "learning_rate": 0.0001783387622149837,
+      "loss": 0.6646,
+      "step": 1293
+    },
+    {
+      "epoch": 2.843956043956044,
+      "grad_norm": 0.040902912616729736,
+      "learning_rate": 0.0001758957654723127,
+      "loss": 0.9736,
+      "step": 1294
+    },
+    {
+      "epoch": 2.8461538461538463,
+      "grad_norm": 0.11191163957118988,
+      "learning_rate": 0.0001734527687296417,
+      "loss": 0.6827,
+      "step": 1295
+    },
+    {
+      "epoch": 2.8483516483516484,
+      "grad_norm": 0.12588852643966675,
+      "learning_rate": 0.00017100977198697068,
+      "loss": 0.9279,
+      "step": 1296
+    },
+    {
+      "epoch": 2.8505494505494506,
+      "grad_norm": 0.026292411610484123,
+      "learning_rate": 0.00016856677524429968,
+      "loss": 0.969,
+      "step": 1297
+    },
+    {
+      "epoch": 2.852747252747253,
+      "grad_norm": 0.0520312637090683,
+      "learning_rate": 0.00016612377850162865,
+      "loss": 0.8558,
+      "step": 1298
+    },
+    {
+      "epoch": 2.854945054945055,
+      "grad_norm": 0.036654647439718246,
+      "learning_rate": 0.00016368078175895768,
+      "loss": 0.8004,
+      "step": 1299
+    },
+    {
+      "epoch": 2.857142857142857,
+      "grad_norm": 0.02587447501718998,
+      "learning_rate": 0.00016123778501628665,
+      "loss": 0.8184,
+      "step": 1300
+    },
+    {
+      "epoch": 2.8593406593406594,
+      "grad_norm": 0.07368086278438568,
+      "learning_rate": 0.00015879478827361565,
+      "loss": 0.9359,
+      "step": 1301
+    },
+    {
+      "epoch": 2.8615384615384616,
+      "grad_norm": 0.04154244810342789,
+      "learning_rate": 0.00015635179153094462,
+      "loss": 0.8258,
+      "step": 1302
+    },
+    {
+      "epoch": 2.8637362637362638,
+      "grad_norm": 0.05571328476071358,
+      "learning_rate": 0.00015390879478827362,
+      "loss": 0.8976,
+      "step": 1303
+    },
+    {
+      "epoch": 2.865934065934066,
+      "grad_norm": 0.020754875615239143,
+      "learning_rate": 0.00015146579804560262,
+      "loss": 0.6823,
+      "step": 1304
+    },
+    {
+      "epoch": 2.868131868131868,
+      "grad_norm": 0.05646703764796257,
+      "learning_rate": 0.0001490228013029316,
+      "loss": 0.8302,
+      "step": 1305
+    },
+    {
+      "epoch": 2.8703296703296703,
+      "grad_norm": 0.028053343296051025,
+      "learning_rate": 0.0001465798045602606,
+      "loss": 0.8186,
+      "step": 1306
+    },
+    {
+      "epoch": 2.8725274725274725,
+      "grad_norm": 0.032905153930187225,
+      "learning_rate": 0.00014413680781758957,
+      "loss": 0.7501,
+      "step": 1307
+    },
+    {
+      "epoch": 2.8747252747252747,
+      "grad_norm": 0.027730310335755348,
+      "learning_rate": 0.00014169381107491857,
+      "loss": 0.7836,
+      "step": 1308
+    },
+    {
+      "epoch": 2.876923076923077,
+      "grad_norm": 0.07724587619304657,
+      "learning_rate": 0.00013925081433224757,
+      "loss": 0.9606,
+      "step": 1309
+    },
+    {
+      "epoch": 2.879120879120879,
+      "grad_norm": 0.02385147102177143,
+      "learning_rate": 0.00013680781758957655,
+      "loss": 0.7465,
+      "step": 1310
+    },
+    {
+      "epoch": 2.8813186813186813,
+      "grad_norm": 0.03696766495704651,
+      "learning_rate": 0.00013436482084690555,
+      "loss": 0.7386,
+      "step": 1311
+    },
+    {
+      "epoch": 2.8835164835164835,
+      "grad_norm": 0.02561488002538681,
+      "learning_rate": 0.00013192182410423452,
+      "loss": 0.768,
+      "step": 1312
+    },
+    {
+      "epoch": 2.8857142857142857,
+      "grad_norm": 0.03198729082942009,
+      "learning_rate": 0.00012947882736156352,
+      "loss": 0.77,
+      "step": 1313
+    },
+    {
+      "epoch": 2.887912087912088,
+      "grad_norm": 0.043216221034526825,
+      "learning_rate": 0.0001270358306188925,
+      "loss": 0.8081,
+      "step": 1314
+    },
+    {
+      "epoch": 2.89010989010989,
+      "grad_norm": 0.05349990352988243,
+      "learning_rate": 0.0001245928338762215,
+      "loss": 0.6969,
+      "step": 1315
+    },
+    {
+      "epoch": 2.8923076923076922,
+      "grad_norm": 0.022735117003321648,
+      "learning_rate": 0.0001221498371335505,
+      "loss": 0.7771,
+      "step": 1316
+    },
+    {
+      "epoch": 2.8945054945054944,
+      "grad_norm": 0.034353192895650864,
+      "learning_rate": 0.00011970684039087947,
+      "loss": 0.7535,
+      "step": 1317
+    },
+    {
+      "epoch": 2.8967032967032966,
+      "grad_norm": 0.07207413762807846,
+      "learning_rate": 0.00011726384364820847,
+      "loss": 0.7683,
+      "step": 1318
+    },
+    {
+      "epoch": 2.898901098901099,
+      "grad_norm": 0.03422323241829872,
+      "learning_rate": 0.00011482084690553747,
+      "loss": 0.8262,
+      "step": 1319
+    },
+    {
+      "epoch": 2.901098901098901,
+      "grad_norm": 0.024574125185608864,
+      "learning_rate": 0.00011237785016286646,
+      "loss": 0.7222,
+      "step": 1320
+    },
+    {
+      "epoch": 2.903296703296703,
+      "grad_norm": 0.04520922526717186,
+      "learning_rate": 0.00010993485342019546,
+      "loss": 0.8351,
+      "step": 1321
+    },
+    {
+      "epoch": 2.9054945054945054,
+      "grad_norm": 0.03352086991071701,
+      "learning_rate": 0.00010749185667752444,
+      "loss": 0.7459,
+      "step": 1322
+    },
+    {
+      "epoch": 2.9076923076923076,
+      "grad_norm": 0.04798645153641701,
+      "learning_rate": 0.00010504885993485343,
+      "loss": 0.7695,
+      "step": 1323
+    },
+    {
+      "epoch": 2.9098901098901098,
+      "grad_norm": 0.044279519468545914,
+      "learning_rate": 0.00010260586319218242,
+      "loss": 0.9672,
+      "step": 1324
+    },
+    {
+      "epoch": 2.912087912087912,
+      "grad_norm": 0.42676275968551636,
+      "learning_rate": 0.0001001628664495114,
+      "loss": 0.9247,
+      "step": 1325
+    },
+    {
+      "epoch": 2.914285714285714,
+      "grad_norm": 0.05856403708457947,
+      "learning_rate": 9.771986970684039e-05,
+      "loss": 0.6903,
+      "step": 1326
+    },
+    {
+      "epoch": 2.9164835164835163,
+      "grad_norm": 0.03217325359582901,
+      "learning_rate": 9.527687296416939e-05,
+      "loss": 0.7039,
+      "step": 1327
+    },
+    {
+      "epoch": 2.9186813186813185,
+      "grad_norm": 0.02469526417553425,
+      "learning_rate": 9.283387622149838e-05,
+      "loss": 0.8396,
+      "step": 1328
+    },
+    {
+      "epoch": 2.9208791208791207,
+      "grad_norm": 0.18351314961910248,
+      "learning_rate": 9.039087947882736e-05,
+      "loss": 0.9792,
+      "step": 1329
+    },
+    {
+      "epoch": 2.9230769230769234,
+      "grad_norm": 0.04656409099698067,
+      "learning_rate": 8.794788273615635e-05,
+      "loss": 0.7045,
+      "step": 1330
+    },
+    {
+      "epoch": 2.9252747252747255,
+      "grad_norm": 0.046893879771232605,
+      "learning_rate": 8.550488599348534e-05,
+      "loss": 0.7605,
+      "step": 1331
+    },
+    {
+      "epoch": 2.9274725274725277,
+      "grad_norm": 0.026826288551092148,
+      "learning_rate": 8.306188925081432e-05,
+      "loss": 0.7491,
+      "step": 1332
+    },
+    {
+      "epoch": 2.92967032967033,
+      "grad_norm": 0.043118271976709366,
+      "learning_rate": 8.061889250814332e-05,
+      "loss": 0.7472,
+      "step": 1333
+    },
+    {
+      "epoch": 2.931868131868132,
+      "grad_norm": 0.0844450518488884,
+      "learning_rate": 7.817589576547231e-05,
+      "loss": 0.6837,
+      "step": 1334
+    },
+    {
+      "epoch": 2.9340659340659343,
+      "grad_norm": 0.042086366564035416,
+      "learning_rate": 7.573289902280131e-05,
+      "loss": 0.9816,
+      "step": 1335
+    },
+    {
+      "epoch": 2.9362637362637365,
+      "grad_norm": 0.021511703729629517,
+      "learning_rate": 7.32899022801303e-05,
+      "loss": 0.6949,
+      "step": 1336
+    },
+    {
+      "epoch": 2.9384615384615387,
+      "grad_norm": 0.05497022345662117,
+      "learning_rate": 7.084690553745929e-05,
+      "loss": 0.8197,
+      "step": 1337
+    },
+    {
+      "epoch": 2.940659340659341,
+      "grad_norm": 0.04968016967177391,
+      "learning_rate": 6.840390879478827e-05,
+      "loss": 0.8097,
+      "step": 1338
+    },
+    {
+      "epoch": 2.942857142857143,
+      "grad_norm": 0.030060557648539543,
+      "learning_rate": 6.596091205211726e-05,
+      "loss": 0.7716,
+      "step": 1339
+    },
+    {
+      "epoch": 2.9450549450549453,
+      "grad_norm": 0.03294103220105171,
+      "learning_rate": 6.351791530944625e-05,
+      "loss": 0.7085,
+      "step": 1340
+    },
+    {
+      "epoch": 2.9472527472527474,
+      "grad_norm": 0.18533386290073395,
+      "learning_rate": 6.107491856677525e-05,
+      "loss": 0.829,
+      "step": 1341
+    },
+    {
+      "epoch": 2.9494505494505496,
+      "grad_norm": 0.03592333570122719,
+      "learning_rate": 5.8631921824104234e-05,
+      "loss": 0.7161,
+      "step": 1342
+    },
+    {
+      "epoch": 2.951648351648352,
+      "grad_norm": 0.025183847174048424,
+      "learning_rate": 5.618892508143323e-05,
+      "loss": 0.774,
+      "step": 1343
+    },
+    {
+      "epoch": 2.953846153846154,
+      "grad_norm": 0.03486381843686104,
+      "learning_rate": 5.374592833876222e-05,
+      "loss": 0.7673,
+      "step": 1344
+    },
+    {
+      "epoch": 2.956043956043956,
+      "grad_norm": 0.03806711733341217,
+      "learning_rate": 5.130293159609121e-05,
+      "loss": 0.8133,
+      "step": 1345
+    },
+    {
+      "epoch": 2.9582417582417584,
+      "grad_norm": 0.06383506953716278,
+      "learning_rate": 4.8859934853420195e-05,
+      "loss": 0.9088,
+      "step": 1346
+    },
+    {
+      "epoch": 2.9604395604395606,
+      "grad_norm": 0.035278767347335815,
+      "learning_rate": 4.641693811074919e-05,
+      "loss": 0.9001,
+      "step": 1347
+    },
+    {
+      "epoch": 2.9626373626373628,
+      "grad_norm": 0.035755954682826996,
+      "learning_rate": 4.3973941368078175e-05,
+      "loss": 0.7512,
+      "step": 1348
+    },
+    {
+      "epoch": 2.964835164835165,
+      "grad_norm": 0.029191862791776657,
+      "learning_rate": 4.153094462540716e-05,
+      "loss": 0.8186,
+      "step": 1349
+    },
+    {
+      "epoch": 2.967032967032967,
+      "grad_norm": 0.022751852869987488,
+      "learning_rate": 3.9087947882736156e-05,
+      "loss": 0.9675,
+      "step": 1350
+    },
+    {
+      "epoch": 2.9692307692307693,
+      "grad_norm": 0.035929303616285324,
+      "learning_rate": 3.664495114006515e-05,
+      "loss": 0.8499,
+      "step": 1351
+    },
+    {
+      "epoch": 2.9714285714285715,
+      "grad_norm": 0.078434519469738,
+      "learning_rate": 3.4201954397394136e-05,
+      "loss": 0.6986,
+      "step": 1352
+    },
+    {
+      "epoch": 2.9736263736263737,
+      "grad_norm": 0.049849338829517365,
+      "learning_rate": 3.175895765472312e-05,
+      "loss": 0.7052,
+      "step": 1353
+    },
+    {
+      "epoch": 2.975824175824176,
+      "grad_norm": 0.03667932748794556,
+      "learning_rate": 2.9315960912052117e-05,
+      "loss": 0.7738,
+      "step": 1354
+    },
+    {
+      "epoch": 2.978021978021978,
+      "grad_norm": 0.04700959846377373,
+      "learning_rate": 2.687296416938111e-05,
+      "loss": 0.7048,
+      "step": 1355
+    },
+    {
+      "epoch": 2.9802197802197803,
+      "grad_norm": 0.030220799148082733,
+      "learning_rate": 2.4429967426710097e-05,
+      "loss": 0.7697,
+      "step": 1356
+    },
+    {
+      "epoch": 2.9824175824175825,
+      "grad_norm": 0.04600556939840317,
+      "learning_rate": 2.1986970684039088e-05,
+      "loss": 0.7263,
+      "step": 1357
+    },
+    {
+      "epoch": 2.9846153846153847,
+      "grad_norm": 0.04411328211426735,
+      "learning_rate": 1.9543973941368078e-05,
+      "loss": 0.8834,
+      "step": 1358
+    },
+    {
+      "epoch": 2.986813186813187,
+      "grad_norm": 0.08912625163793564,
+      "learning_rate": 1.7100977198697068e-05,
+      "loss": 1.0596,
+      "step": 1359
+    },
+    {
+      "epoch": 2.989010989010989,
+      "grad_norm": 0.15710528194904327,
+      "learning_rate": 1.4657980456026058e-05,
+      "loss": 0.8867,
+      "step": 1360
+    },
+    {
+      "epoch": 2.9912087912087912,
+      "grad_norm": 0.03181207925081253,
+      "learning_rate": 1.2214983713355049e-05,
+      "loss": 0.7307,
+      "step": 1361
+    },
+    {
+      "epoch": 2.9934065934065934,
+      "grad_norm": 0.039096519351005554,
+      "learning_rate": 9.771986970684039e-06,
+      "loss": 1.0054,
+      "step": 1362
+    },
+    {
+      "epoch": 2.9956043956043956,
+      "grad_norm": 0.0748264417052269,
+      "learning_rate": 7.328990228013029e-06,
+      "loss": 0.7658,
+      "step": 1363
+    },
+    {
+      "epoch": 2.997802197802198,
+      "grad_norm": 0.02484406717121601,
+      "learning_rate": 4.8859934853420195e-06,
+      "loss": 0.8184,
+      "step": 1364
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.031767744570970535,
+      "learning_rate": 2.4429967426710097e-06,
+      "loss": 0.7845,
+      "step": 1365
+    },
+    {
+      "epoch": 3.0,
+      "step": 1365,
+      "total_flos": 1.0021240006560973e+18,
+      "train_loss": 1.5448391757168614,
+      "train_runtime": 5139.738,
+      "train_samples_per_second": 4.249,
+      "train_steps_per_second": 0.266
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 1365,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.0021240006560973e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}