diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,9143 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.0,
+  "eval_steps": 500,
+  "global_step": 1300,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0015384615384615385,
+      "grad_norm": 8.719981218723293,
+      "learning_rate": 2e-05,
+      "loss": 1.1094,
+      "step": 1
+    },
+    {
+      "epoch": 0.003076923076923077,
+      "grad_norm": 2.857634926149163,
+      "learning_rate": 1.999997080000119e-05,
+      "loss": 0.8812,
+      "step": 2
+    },
+    {
+      "epoch": 0.004615384615384616,
+      "grad_norm": 3.119706197426241,
+      "learning_rate": 1.9999883200175286e-05,
+      "loss": 0.7875,
+      "step": 3
+    },
+    {
+      "epoch": 0.006153846153846154,
+      "grad_norm": 1.786007994677646,
+      "learning_rate": 1.9999737201033877e-05,
+      "loss": 0.7231,
+      "step": 4
+    },
+    {
+      "epoch": 0.007692307692307693,
+      "grad_norm": 2.00918151690818,
+      "learning_rate": 1.999953280342959e-05,
+      "loss": 0.6921,
+      "step": 5
+    },
+    {
+      "epoch": 0.009230769230769232,
+      "grad_norm": 1.1243353749850378,
+      "learning_rate": 1.9999270008556108e-05,
+      "loss": 0.6443,
+      "step": 6
+    },
+    {
+      "epoch": 0.010769230769230769,
+      "grad_norm": 0.8661186793152397,
+      "learning_rate": 1.9998948817948157e-05,
+      "loss": 0.582,
+      "step": 7
+    },
+    {
+      "epoch": 0.012307692307692308,
+      "grad_norm": 0.931762068207405,
+      "learning_rate": 1.999856923348149e-05,
+      "loss": 0.5741,
+      "step": 8
+    },
+    {
+      "epoch": 0.013846153846153847,
+      "grad_norm": 0.71805508796234,
+      "learning_rate": 1.9998131257372878e-05,
+      "loss": 0.5624,
+      "step": 9
+    },
+    {
+      "epoch": 0.015384615384615385,
+      "grad_norm": 0.7271352074162678,
+      "learning_rate": 1.99976348921801e-05,
+      "loss": 0.554,
+      "step": 10
+    },
+    {
+      "epoch": 0.016923076923076923,
+      "grad_norm": 0.691298121896311,
+      "learning_rate": 1.9997080140801932e-05,
+      "loss": 0.5468,
+      "step": 11
+    },
+    {
+      "epoch": 0.018461538461538463,
+      "grad_norm": 0.7656378184557648,
+      "learning_rate": 1.999646700647812e-05,
+      "loss": 0.5343,
+      "step": 12
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6164397568987413,
+      "learning_rate": 1.9995795492789368e-05,
+      "loss": 0.5194,
+      "step": 13
+    },
+    {
+      "epoch": 0.021538461538461538,
+      "grad_norm": 0.5546285387197363,
+      "learning_rate": 1.9995065603657317e-05,
+      "loss": 0.5007,
+      "step": 14
+    },
+    {
+      "epoch": 0.023076923076923078,
+      "grad_norm": 0.6008331537546505,
+      "learning_rate": 1.999427734334452e-05,
+      "loss": 0.5051,
+      "step": 15
+    },
+    {
+      "epoch": 0.024615384615384615,
+      "grad_norm": 0.5500757655170632,
+      "learning_rate": 1.9993430716454415e-05,
+      "loss": 0.4964,
+      "step": 16
+    },
+    {
+      "epoch": 0.026153846153846153,
+      "grad_norm": 0.5496490725348775,
+      "learning_rate": 1.9992525727931303e-05,
+      "loss": 0.4886,
+      "step": 17
+    },
+    {
+      "epoch": 0.027692307692307693,
+      "grad_norm": 0.48860119209634406,
+      "learning_rate": 1.9991562383060316e-05,
+      "loss": 0.4978,
+      "step": 18
+    },
+    {
+      "epoch": 0.02923076923076923,
+      "grad_norm": 0.5213604708312909,
+      "learning_rate": 1.9990540687467394e-05,
+      "loss": 0.4918,
+      "step": 19
+    },
+    {
+      "epoch": 0.03076923076923077,
+      "grad_norm": 0.4907733080986667,
+      "learning_rate": 1.9989460647119232e-05,
+      "loss": 0.4734,
+      "step": 20
+    },
+    {
+      "epoch": 0.03230769230769231,
+      "grad_norm": 0.45390452831580014,
+      "learning_rate": 1.998832226832327e-05,
+      "loss": 0.4696,
+      "step": 21
+    },
+    {
+      "epoch": 0.033846153846153845,
+      "grad_norm": 0.480861761744533,
+      "learning_rate": 1.9987125557727633e-05,
+      "loss": 0.4609,
+      "step": 22
+    },
+    {
+      "epoch": 0.03538461538461538,
+      "grad_norm": 0.46459703572766675,
+      "learning_rate": 1.9985870522321118e-05,
+      "loss": 0.488,
+      "step": 23
+    },
+    {
+      "epoch": 0.036923076923076927,
+      "grad_norm": 0.4395906044390323,
+      "learning_rate": 1.9984557169433126e-05,
+      "loss": 0.4708,
+      "step": 24
+    },
+    {
+      "epoch": 0.038461538461538464,
+      "grad_norm": 0.45070764926355616,
+      "learning_rate": 1.9983185506733643e-05,
+      "loss": 0.4625,
+      "step": 25
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4341431642659233,
+      "learning_rate": 1.9981755542233175e-05,
+      "loss": 0.459,
+      "step": 26
+    },
+    {
+      "epoch": 0.04153846153846154,
+      "grad_norm": 0.5062969956581659,
+      "learning_rate": 1.9980267284282718e-05,
+      "loss": 0.4559,
+      "step": 27
+    },
+    {
+      "epoch": 0.043076923076923075,
+      "grad_norm": 0.47133127732951885,
+      "learning_rate": 1.9978720741573693e-05,
+      "loss": 0.471,
+      "step": 28
+    },
+    {
+      "epoch": 0.04461538461538461,
+      "grad_norm": 0.44673748680895614,
+      "learning_rate": 1.9977115923137912e-05,
+      "loss": 0.4696,
+      "step": 29
+    },
+    {
+      "epoch": 0.046153846153846156,
+      "grad_norm": 0.43779008750875575,
+      "learning_rate": 1.9975452838347513e-05,
+      "loss": 0.4667,
+      "step": 30
+    },
+    {
+      "epoch": 0.047692307692307694,
+      "grad_norm": 0.537808204198567,
+      "learning_rate": 1.9973731496914914e-05,
+      "loss": 0.4474,
+      "step": 31
+    },
+    {
+      "epoch": 0.04923076923076923,
+      "grad_norm": 0.5029909772241813,
+      "learning_rate": 1.9971951908892743e-05,
+      "loss": 0.4359,
+      "step": 32
+    },
+    {
+      "epoch": 0.05076923076923077,
+      "grad_norm": 0.5086116827393183,
+      "learning_rate": 1.9970114084673796e-05,
+      "loss": 0.4425,
+      "step": 33
+    },
+    {
+      "epoch": 0.052307692307692305,
+      "grad_norm": 0.4071622609137339,
+      "learning_rate": 1.996821803499097e-05,
+      "loss": 0.4564,
+      "step": 34
+    },
+    {
+      "epoch": 0.05384615384615385,
+      "grad_norm": 0.4302119670716149,
+      "learning_rate": 1.9966263770917192e-05,
+      "loss": 0.4381,
+      "step": 35
+    },
+    {
+      "epoch": 0.055384615384615386,
+      "grad_norm": 0.422589290081668,
+      "learning_rate": 1.9964251303865362e-05,
+      "loss": 0.443,
+      "step": 36
+    },
+    {
+      "epoch": 0.05692307692307692,
+      "grad_norm": 0.4890614311173147,
+      "learning_rate": 1.996218064558829e-05,
+      "loss": 0.4364,
+      "step": 37
+    },
+    {
+      "epoch": 0.05846153846153846,
+      "grad_norm": 0.44871932965848343,
+      "learning_rate": 1.9960051808178616e-05,
+      "loss": 0.4315,
+      "step": 38
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.46394560279566827,
+      "learning_rate": 1.9957864804068752e-05,
+      "loss": 0.442,
+      "step": 39
+    },
+    {
+      "epoch": 0.06153846153846154,
+      "grad_norm": 0.4853484602516898,
+      "learning_rate": 1.99556196460308e-05,
+      "loss": 0.4268,
+      "step": 40
+    },
+    {
+      "epoch": 0.06307692307692307,
+      "grad_norm": 0.4686555271366835,
+      "learning_rate": 1.995331634717649e-05,
+      "loss": 0.4378,
+      "step": 41
+    },
+    {
+      "epoch": 0.06461538461538462,
+      "grad_norm": 0.4251309519127146,
+      "learning_rate": 1.9950954920957074e-05,
+      "loss": 0.4382,
+      "step": 42
+    },
+    {
+      "epoch": 0.06615384615384616,
+      "grad_norm": 0.47033955140744915,
+      "learning_rate": 1.994853538116329e-05,
+      "loss": 0.443,
+      "step": 43
+    },
+    {
+      "epoch": 0.06769230769230769,
+      "grad_norm": 0.43873749888912006,
+      "learning_rate": 1.994605774192525e-05,
+      "loss": 0.4346,
+      "step": 44
+    },
+    {
+      "epoch": 0.06923076923076923,
+      "grad_norm": 0.41247310971108225,
+      "learning_rate": 1.994352201771236e-05,
+      "loss": 0.4307,
+      "step": 45
+    },
+    {
+      "epoch": 0.07076923076923076,
+      "grad_norm": 0.46471575806556253,
+      "learning_rate": 1.9940928223333254e-05,
+      "loss": 0.419,
+      "step": 46
+    },
+    {
+      "epoch": 0.07230769230769231,
+      "grad_norm": 0.49920986157774244,
+      "learning_rate": 1.9938276373935688e-05,
+      "loss": 0.4309,
+      "step": 47
+    },
+    {
+      "epoch": 0.07384615384615385,
+      "grad_norm": 0.5165106469583145,
+      "learning_rate": 1.9935566485006464e-05,
+      "loss": 0.4106,
+      "step": 48
+    },
+    {
+      "epoch": 0.07538461538461538,
+      "grad_norm": 0.42000776549500274,
+      "learning_rate": 1.993279857237133e-05,
+      "loss": 0.4297,
+      "step": 49
+    },
+    {
+      "epoch": 0.07692307692307693,
+      "grad_norm": 0.4456798761609307,
+      "learning_rate": 1.99299726521949e-05,
+      "loss": 0.4301,
+      "step": 50
+    },
+    {
+      "epoch": 0.07846153846153846,
+      "grad_norm": 0.43041312016240524,
+      "learning_rate": 1.992708874098054e-05,
+      "loss": 0.4354,
+      "step": 51
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4865995223799656,
+      "learning_rate": 1.9924146855570298e-05,
+      "loss": 0.4177,
+      "step": 52
+    },
+    {
+      "epoch": 0.08153846153846153,
+      "grad_norm": 0.49386735818058997,
+      "learning_rate": 1.9921147013144782e-05,
+      "loss": 0.422,
+      "step": 53
+    },
+    {
+      "epoch": 0.08307692307692308,
+      "grad_norm": 0.461649928406544,
+      "learning_rate": 1.9918089231223066e-05,
+      "loss": 0.4216,
+      "step": 54
+    },
+    {
+      "epoch": 0.08461538461538462,
+      "grad_norm": 0.4703667570538988,
+      "learning_rate": 1.99149735276626e-05,
+      "loss": 0.4231,
+      "step": 55
+    },
+    {
+      "epoch": 0.08615384615384615,
+      "grad_norm": 0.45703810516219306,
+      "learning_rate": 1.9911799920659093e-05,
+      "loss": 0.4138,
+      "step": 56
+    },
+    {
+      "epoch": 0.0876923076923077,
+      "grad_norm": 0.41363702902939065,
+      "learning_rate": 1.9908568428746408e-05,
+      "loss": 0.4332,
+      "step": 57
+    },
+    {
+      "epoch": 0.08923076923076922,
+      "grad_norm": 0.4408788614860877,
+      "learning_rate": 1.9905279070796454e-05,
+      "loss": 0.4229,
+      "step": 58
+    },
+    {
+      "epoch": 0.09076923076923077,
+      "grad_norm": 0.4470958998878198,
+      "learning_rate": 1.9901931866019087e-05,
+      "loss": 0.4136,
+      "step": 59
+    },
+    {
+      "epoch": 0.09230769230769231,
+      "grad_norm": 0.4631914715416953,
+      "learning_rate": 1.989852683396198e-05,
+      "loss": 0.4117,
+      "step": 60
+    },
+    {
+      "epoch": 0.09384615384615384,
+      "grad_norm": 0.43674476292646275,
+      "learning_rate": 1.9895063994510512e-05,
+      "loss": 0.4264,
+      "step": 61
+    },
+    {
+      "epoch": 0.09538461538461539,
+      "grad_norm": 0.44678131236132346,
+      "learning_rate": 1.9891543367887675e-05,
+      "loss": 0.4209,
+      "step": 62
+    },
+    {
+      "epoch": 0.09692307692307692,
+      "grad_norm": 0.4100670081436192,
+      "learning_rate": 1.988796497465392e-05,
+      "loss": 0.4256,
+      "step": 63
+    },
+    {
+      "epoch": 0.09846153846153846,
+      "grad_norm": 0.4423525109689805,
+      "learning_rate": 1.988432883570707e-05,
+      "loss": 0.4159,
+      "step": 64
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4718279672175279,
+      "learning_rate": 1.9880634972282168e-05,
+      "loss": 0.4112,
+      "step": 65
+    },
+    {
+      "epoch": 0.10153846153846154,
+      "grad_norm": 0.45663009818729866,
+      "learning_rate": 1.9876883405951378e-05,
+      "loss": 0.4217,
+      "step": 66
+    },
+    {
+      "epoch": 0.10307692307692308,
+      "grad_norm": 0.46198665198139227,
+      "learning_rate": 1.987307415862385e-05,
+      "loss": 0.4082,
+      "step": 67
+    },
+    {
+      "epoch": 0.10461538461538461,
+      "grad_norm": 0.4395530590265812,
+      "learning_rate": 1.9869207252545582e-05,
+      "loss": 0.4126,
+      "step": 68
+    },
+    {
+      "epoch": 0.10615384615384615,
+      "grad_norm": 0.41316450326373205,
+      "learning_rate": 1.986528271029931e-05,
+      "loss": 0.4241,
+      "step": 69
+    },
+    {
+      "epoch": 0.1076923076923077,
+      "grad_norm": 0.4637796779791821,
+      "learning_rate": 1.9861300554804357e-05,
+      "loss": 0.4096,
+      "step": 70
+    },
+    {
+      "epoch": 0.10923076923076923,
+      "grad_norm": 0.4165786229490775,
+      "learning_rate": 1.985726080931651e-05,
+      "loss": 0.4092,
+      "step": 71
+    },
+    {
+      "epoch": 0.11076923076923077,
+      "grad_norm": 0.4493255993354414,
+      "learning_rate": 1.9853163497427885e-05,
+      "loss": 0.4189,
+      "step": 72
+    },
+    {
+      "epoch": 0.1123076923076923,
+      "grad_norm": 0.4524687959375285,
+      "learning_rate": 1.9849008643066774e-05,
+      "loss": 0.4013,
+      "step": 73
+    },
+    {
+      "epoch": 0.11384615384615385,
+      "grad_norm": 0.4382535298994216,
+      "learning_rate": 1.984479627049753e-05,
+      "loss": 0.4198,
+      "step": 74
+    },
+    {
+      "epoch": 0.11538461538461539,
+      "grad_norm": 0.448254695336545,
+      "learning_rate": 1.9840526404320415e-05,
+      "loss": 0.4052,
+      "step": 75
+    },
+    {
+      "epoch": 0.11692307692307692,
+      "grad_norm": 0.46210841061993035,
+      "learning_rate": 1.983619906947144e-05,
+      "loss": 0.391,
+      "step": 76
+    },
+    {
+      "epoch": 0.11846153846153847,
+      "grad_norm": 0.45410549904482356,
+      "learning_rate": 1.9831814291222233e-05,
+      "loss": 0.4212,
+      "step": 77
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4619653756093457,
+      "learning_rate": 1.982737209517991e-05,
+      "loss": 0.408,
+      "step": 78
+    },
+    {
+      "epoch": 0.12153846153846154,
+      "grad_norm": 0.45914836450254387,
+      "learning_rate": 1.982287250728689e-05,
+      "loss": 0.4025,
+      "step": 79
+    },
+    {
+      "epoch": 0.12307692307692308,
+      "grad_norm": 0.47155262949088383,
+      "learning_rate": 1.981831555382076e-05,
+      "loss": 0.4183,
+      "step": 80
+    },
+    {
+      "epoch": 0.12461538461538461,
+      "grad_norm": 0.4722198730416428,
+      "learning_rate": 1.9813701261394136e-05,
+      "loss": 0.4006,
+      "step": 81
+    },
+    {
+      "epoch": 0.12615384615384614,
+      "grad_norm": 0.5210721615975413,
+      "learning_rate": 1.980902965695448e-05,
+      "loss": 0.3905,
+      "step": 82
+    },
+    {
+      "epoch": 0.1276923076923077,
+      "grad_norm": 0.47253752161225565,
+      "learning_rate": 1.9804300767783958e-05,
+      "loss": 0.4126,
+      "step": 83
+    },
+    {
+      "epoch": 0.12923076923076923,
+      "grad_norm": 0.48042588417599785,
+      "learning_rate": 1.979951462149929e-05,
+      "loss": 0.4091,
+      "step": 84
+    },
+    {
+      "epoch": 0.13076923076923078,
+      "grad_norm": 0.4196254833427212,
+      "learning_rate": 1.979467124605156e-05,
+      "loss": 0.4099,
+      "step": 85
+    },
+    {
+      "epoch": 0.13230769230769232,
+      "grad_norm": 0.5042648190839037,
+      "learning_rate": 1.9789770669726088e-05,
+      "loss": 0.4128,
+      "step": 86
+    },
+    {
+      "epoch": 0.13384615384615384,
+      "grad_norm": 0.48318808152888754,
+      "learning_rate": 1.9784812921142232e-05,
+      "loss": 0.4044,
+      "step": 87
+    },
+    {
+      "epoch": 0.13538461538461538,
+      "grad_norm": 0.44092585053098554,
+      "learning_rate": 1.977979802925324e-05,
+      "loss": 0.4131,
+      "step": 88
+    },
+    {
+      "epoch": 0.13692307692307693,
+      "grad_norm": 0.539732363228096,
+      "learning_rate": 1.977472602334609e-05,
+      "loss": 0.4067,
+      "step": 89
+    },
+    {
+      "epoch": 0.13846153846153847,
+      "grad_norm": 0.4902058273513082,
+      "learning_rate": 1.976959693304129e-05,
+      "loss": 0.3852,
+      "step": 90
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.46264438444120237,
+      "learning_rate": 1.9764410788292724e-05,
+      "loss": 0.4069,
+      "step": 91
+    },
+    {
+      "epoch": 0.14153846153846153,
+      "grad_norm": 0.459869174959949,
+      "learning_rate": 1.9759167619387474e-05,
+      "loss": 0.3964,
+      "step": 92
+    },
+    {
+      "epoch": 0.14307692307692307,
+      "grad_norm": 0.45536657171147943,
+      "learning_rate": 1.9753867456945653e-05,
+      "loss": 0.3721,
+      "step": 93
+    },
+    {
+      "epoch": 0.14461538461538462,
+      "grad_norm": 0.45122524505775485,
+      "learning_rate": 1.9748510331920204e-05,
+      "loss": 0.4085,
+      "step": 94
+    },
+    {
+      "epoch": 0.14615384615384616,
+      "grad_norm": 0.4980488929391611,
+      "learning_rate": 1.9743096275596735e-05,
+      "loss": 0.418,
+      "step": 95
+    },
+    {
+      "epoch": 0.1476923076923077,
+      "grad_norm": 0.3909703378516853,
+      "learning_rate": 1.9737625319593338e-05,
+      "loss": 0.4123,
+      "step": 96
+    },
+    {
+      "epoch": 0.14923076923076922,
+      "grad_norm": 0.459147539574416,
+      "learning_rate": 1.9732097495860388e-05,
+      "loss": 0.4003,
+      "step": 97
+    },
+    {
+      "epoch": 0.15076923076923077,
+      "grad_norm": 0.4407198489664392,
+      "learning_rate": 1.972651283668038e-05,
+      "loss": 0.3862,
+      "step": 98
+    },
+    {
+      "epoch": 0.1523076923076923,
+      "grad_norm": 0.4457250563260398,
+      "learning_rate": 1.9720871374667714e-05,
+      "loss": 0.3952,
+      "step": 99
+    },
+    {
+      "epoch": 0.15384615384615385,
+      "grad_norm": 0.48709061128785947,
+      "learning_rate": 1.971517314276854e-05,
+      "loss": 0.4072,
+      "step": 100
+    },
+    {
+      "epoch": 0.15538461538461537,
+      "grad_norm": 0.4292633067077418,
+      "learning_rate": 1.9709418174260523e-05,
+      "loss": 0.4089,
+      "step": 101
+    },
+    {
+      "epoch": 0.15692307692307692,
+      "grad_norm": 0.41891705808866225,
+      "learning_rate": 1.9703606502752674e-05,
+      "loss": 0.4054,
+      "step": 102
+    },
+    {
+      "epoch": 0.15846153846153846,
+      "grad_norm": 0.42654949290246036,
+      "learning_rate": 1.9697738162185163e-05,
+      "loss": 0.3964,
+      "step": 103
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.45425372471635195,
+      "learning_rate": 1.969181318682909e-05,
+      "loss": 0.3967,
+      "step": 104
+    },
+    {
+      "epoch": 0.16153846153846155,
+      "grad_norm": 0.45288807575933704,
+      "learning_rate": 1.9685831611286312e-05,
+      "loss": 0.3967,
+      "step": 105
+    },
+    {
+      "epoch": 0.16307692307692306,
+      "grad_norm": 0.43112422397486094,
+      "learning_rate": 1.967979347048923e-05,
+      "loss": 0.4015,
+      "step": 106
+    },
+    {
+      "epoch": 0.1646153846153846,
+      "grad_norm": 0.4614379965358889,
+      "learning_rate": 1.9673698799700582e-05,
+      "loss": 0.4043,
+      "step": 107
+    },
+    {
+      "epoch": 0.16615384615384615,
+      "grad_norm": 0.4234267683966117,
+      "learning_rate": 1.9667547634513248e-05,
+      "loss": 0.3849,
+      "step": 108
+    },
+    {
+      "epoch": 0.1676923076923077,
+      "grad_norm": 0.42171008514818914,
+      "learning_rate": 1.9661340010850025e-05,
+      "loss": 0.3829,
+      "step": 109
+    },
+    {
+      "epoch": 0.16923076923076924,
+      "grad_norm": 0.4575756568936066,
+      "learning_rate": 1.9655075964963443e-05,
+      "loss": 0.3967,
+      "step": 110
+    },
+    {
+      "epoch": 0.17076923076923076,
+      "grad_norm": 0.4120262760619657,
+      "learning_rate": 1.9648755533435517e-05,
+      "loss": 0.4059,
+      "step": 111
+    },
+    {
+      "epoch": 0.1723076923076923,
+      "grad_norm": 0.4587196344523497,
+      "learning_rate": 1.9642378753177573e-05,
+      "loss": 0.3924,
+      "step": 112
+    },
+    {
+      "epoch": 0.17384615384615384,
+      "grad_norm": 0.4340056775480998,
+      "learning_rate": 1.9635945661430006e-05,
+      "loss": 0.3964,
+      "step": 113
+    },
+    {
+      "epoch": 0.1753846153846154,
+      "grad_norm": 0.42987114770818896,
+      "learning_rate": 1.9629456295762067e-05,
+      "loss": 0.3928,
+      "step": 114
+    },
+    {
+      "epoch": 0.17692307692307693,
+      "grad_norm": 0.4229103193412076,
+      "learning_rate": 1.9622910694071654e-05,
+      "loss": 0.4009,
+      "step": 115
+    },
+    {
+      "epoch": 0.17846153846153845,
+      "grad_norm": 0.45675896237179386,
+      "learning_rate": 1.9616308894585078e-05,
+      "loss": 0.3995,
+      "step": 116
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.42891991435511145,
+      "learning_rate": 1.9609650935856847e-05,
+      "loss": 0.4038,
+      "step": 117
+    },
+    {
+      "epoch": 0.18153846153846154,
+      "grad_norm": 0.4490370243808544,
+      "learning_rate": 1.9602936856769432e-05,
+      "loss": 0.3961,
+      "step": 118
+    },
+    {
+      "epoch": 0.18307692307692308,
+      "grad_norm": 0.42523169920785225,
+      "learning_rate": 1.9596166696533062e-05,
+      "loss": 0.3817,
+      "step": 119
+    },
+    {
+      "epoch": 0.18461538461538463,
+      "grad_norm": 0.4390191218237474,
+      "learning_rate": 1.9589340494685464e-05,
+      "loss": 0.3946,
+      "step": 120
+    },
+    {
+      "epoch": 0.18615384615384614,
+      "grad_norm": 0.45915297027117175,
+      "learning_rate": 1.9582458291091664e-05,
+      "loss": 0.3679,
+      "step": 121
+    },
+    {
+      "epoch": 0.18769230769230769,
+      "grad_norm": 0.41956234857822733,
+      "learning_rate": 1.957552012594372e-05,
+      "loss": 0.3962,
+      "step": 122
+    },
+    {
+      "epoch": 0.18923076923076923,
+      "grad_norm": 0.4509203383013446,
+      "learning_rate": 1.956852603976052e-05,
+      "loss": 0.4,
+      "step": 123
+    },
+    {
+      "epoch": 0.19076923076923077,
+      "grad_norm": 0.42970024830432557,
+      "learning_rate": 1.9561476073387527e-05,
+      "loss": 0.3899,
+      "step": 124
+    },
+    {
+      "epoch": 0.19230769230769232,
+      "grad_norm": 0.4161963645832618,
+      "learning_rate": 1.9554370267996537e-05,
+      "loss": 0.3899,
+      "step": 125
+    },
+    {
+      "epoch": 0.19384615384615383,
+      "grad_norm": 0.45816126154800635,
+      "learning_rate": 1.954720866508546e-05,
+      "loss": 0.3932,
+      "step": 126
+    },
+    {
+      "epoch": 0.19538461538461538,
+      "grad_norm": 0.4537803299501236,
+      "learning_rate": 1.9539991306478046e-05,
+      "loss": 0.3977,
+      "step": 127
+    },
+    {
+      "epoch": 0.19692307692307692,
+      "grad_norm": 0.4039814930168589,
+      "learning_rate": 1.953271823432367e-05,
+      "loss": 0.3973,
+      "step": 128
+    },
+    {
+      "epoch": 0.19846153846153847,
+      "grad_norm": 0.4900092310878065,
+      "learning_rate": 1.952538949109708e-05,
+      "loss": 0.4019,
+      "step": 129
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.39754923725562885,
+      "learning_rate": 1.9518005119598124e-05,
+      "loss": 0.4184,
+      "step": 130
+    },
+    {
+      "epoch": 0.20153846153846153,
+      "grad_norm": 0.45798884585132515,
+      "learning_rate": 1.9510565162951538e-05,
+      "loss": 0.4032,
+      "step": 131
+    },
+    {
+      "epoch": 0.20307692307692307,
+      "grad_norm": 0.4260303732298482,
+      "learning_rate": 1.9503069664606663e-05,
+      "loss": 0.4024,
+      "step": 132
+    },
+    {
+      "epoch": 0.20461538461538462,
+      "grad_norm": 0.3947111108441383,
+      "learning_rate": 1.9495518668337204e-05,
+      "loss": 0.3959,
+      "step": 133
+    },
+    {
+      "epoch": 0.20615384615384616,
+      "grad_norm": 0.4642553250131546,
+      "learning_rate": 1.9487912218240983e-05,
+      "loss": 0.3962,
+      "step": 134
+    },
+    {
+      "epoch": 0.2076923076923077,
+      "grad_norm": 0.44067853929295847,
+      "learning_rate": 1.9480250358739667e-05,
+      "loss": 0.3764,
+      "step": 135
+    },
+    {
+      "epoch": 0.20923076923076922,
+      "grad_norm": 0.4639861470416291,
+      "learning_rate": 1.947253313457851e-05,
+      "loss": 0.3794,
+      "step": 136
+    },
+    {
+      "epoch": 0.21076923076923076,
+      "grad_norm": 0.4345508766781817,
+      "learning_rate": 1.94647605908261e-05,
+      "loss": 0.3896,
+      "step": 137
+    },
+    {
+      "epoch": 0.2123076923076923,
+      "grad_norm": 0.4354963590288285,
+      "learning_rate": 1.9456932772874092e-05,
+      "loss": 0.3966,
+      "step": 138
+    },
+    {
+      "epoch": 0.21384615384615385,
+      "grad_norm": 0.4667014539424043,
+      "learning_rate": 1.944904972643694e-05,
+      "loss": 0.3899,
+      "step": 139
+    },
+    {
+      "epoch": 0.2153846153846154,
+      "grad_norm": 0.4518242044619916,
+      "learning_rate": 1.944111149755164e-05,
+      "loss": 0.3871,
+      "step": 140
+    },
+    {
+      "epoch": 0.2169230769230769,
+      "grad_norm": 0.4328103210654101,
+      "learning_rate": 1.9433118132577432e-05,
+      "loss": 0.3762,
+      "step": 141
+    },
+    {
+      "epoch": 0.21846153846153846,
+      "grad_norm": 0.4255179604535429,
+      "learning_rate": 1.9425069678195577e-05,
+      "loss": 0.3891,
+      "step": 142
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.47623799922161536,
+      "learning_rate": 1.9416966181409047e-05,
+      "loss": 0.3909,
+      "step": 143
+    },
+    {
+      "epoch": 0.22153846153846155,
+      "grad_norm": 0.4861870367525958,
+      "learning_rate": 1.9408807689542257e-05,
+      "loss": 0.409,
+      "step": 144
+    },
+    {
+      "epoch": 0.2230769230769231,
+      "grad_norm": 0.4647223437701761,
+      "learning_rate": 1.94005942502408e-05,
+      "loss": 0.4002,
+      "step": 145
+    },
+    {
+      "epoch": 0.2246153846153846,
+      "grad_norm": 0.46548716293562825,
+      "learning_rate": 1.9392325911471154e-05,
+      "loss": 0.3967,
+      "step": 146
+    },
+    {
+      "epoch": 0.22615384615384615,
+      "grad_norm": 0.4281960631413778,
+      "learning_rate": 1.9384002721520423e-05,
+      "loss": 0.3819,
+      "step": 147
+    },
+    {
+      "epoch": 0.2276923076923077,
+      "grad_norm": 0.42213942963661416,
+      "learning_rate": 1.937562472899603e-05,
+      "loss": 0.3802,
+      "step": 148
+    },
+    {
+      "epoch": 0.22923076923076924,
+      "grad_norm": 0.39617908384741596,
+      "learning_rate": 1.936719198282545e-05,
+      "loss": 0.3946,
+      "step": 149
+    },
+    {
+      "epoch": 0.23076923076923078,
+      "grad_norm": 0.38266742009553123,
+      "learning_rate": 1.935870453225592e-05,
+      "loss": 0.3919,
+      "step": 150
+    },
+    {
+      "epoch": 0.2323076923076923,
+      "grad_norm": 0.43012539629986213,
+      "learning_rate": 1.9350162426854152e-05,
+      "loss": 0.3925,
+      "step": 151
+    },
+    {
+      "epoch": 0.23384615384615384,
+      "grad_norm": 0.40699797565666346,
+      "learning_rate": 1.934156571650603e-05,
+      "loss": 0.3837,
+      "step": 152
+    },
+    {
+      "epoch": 0.2353846153846154,
+      "grad_norm": 0.4178060531661853,
+      "learning_rate": 1.933291445141635e-05,
+      "loss": 0.3907,
+      "step": 153
+    },
+    {
+      "epoch": 0.23692307692307693,
+      "grad_norm": 0.42409932452834603,
+      "learning_rate": 1.9324208682108493e-05,
+      "loss": 0.384,
+      "step": 154
+    },
+    {
+      "epoch": 0.23846153846153847,
+      "grad_norm": 0.45125956868378864,
+      "learning_rate": 1.931544845942415e-05,
+      "loss": 0.3764,
+      "step": 155
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3633838903359922,
+      "learning_rate": 1.9306633834523022e-05,
+      "loss": 0.3823,
+      "step": 156
+    },
+    {
+      "epoch": 0.24153846153846154,
+      "grad_norm": 0.411975244203751,
+      "learning_rate": 1.9297764858882516e-05,
+      "loss": 0.3782,
+      "step": 157
+    },
+    {
+      "epoch": 0.24307692307692308,
+      "grad_norm": 0.4194875151750598,
+      "learning_rate": 1.9288841584297445e-05,
+      "loss": 0.38,
+      "step": 158
+    },
+    {
+      "epoch": 0.24461538461538462,
+      "grad_norm": 0.3998291172807816,
+      "learning_rate": 1.927986406287973e-05,
+      "loss": 0.3971,
+      "step": 159
+    },
+    {
+      "epoch": 0.24615384615384617,
+      "grad_norm": 0.42912053159246866,
+      "learning_rate": 1.92708323470581e-05,
+      "loss": 0.3965,
+      "step": 160
+    },
+    {
+      "epoch": 0.24769230769230768,
+      "grad_norm": 0.40882396251520026,
+      "learning_rate": 1.9261746489577767e-05,
+      "loss": 0.3989,
+      "step": 161
+    },
+    {
+      "epoch": 0.24923076923076923,
+      "grad_norm": 0.4723540880687314,
+      "learning_rate": 1.925260654350014e-05,
+      "loss": 0.3794,
+      "step": 162
+    },
+    {
+      "epoch": 0.25076923076923074,
+      "grad_norm": 0.384321306204413,
+      "learning_rate": 1.92434125622025e-05,
+      "loss": 0.3767,
+      "step": 163
+    },
+    {
+      "epoch": 0.2523076923076923,
+      "grad_norm": 0.5074427686560539,
+      "learning_rate": 1.9234164599377692e-05,
+      "loss": 0.3717,
+      "step": 164
+    },
+    {
+      "epoch": 0.25384615384615383,
+      "grad_norm": 0.4272226637226024,
+      "learning_rate": 1.9224862709033823e-05,
+      "loss": 0.3794,
+      "step": 165
+    },
+    {
+      "epoch": 0.2553846153846154,
+      "grad_norm": 0.4335469412249708,
+      "learning_rate": 1.9215506945493933e-05,
+      "loss": 0.3837,
+      "step": 166
+    },
+    {
+      "epoch": 0.2569230769230769,
+      "grad_norm": 0.4804945789975517,
+      "learning_rate": 1.9206097363395668e-05,
+      "loss": 0.3932,
+      "step": 167
+    },
+    {
+      "epoch": 0.25846153846153846,
+      "grad_norm": 0.44904255093202744,
+      "learning_rate": 1.9196634017690993e-05,
+      "loss": 0.3788,
+      "step": 168
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.47237360699091735,
+      "learning_rate": 1.9187116963645845e-05,
+      "loss": 0.3915,
+      "step": 169
+    },
+    {
+      "epoch": 0.26153846153846155,
+      "grad_norm": 0.4540044958139465,
+      "learning_rate": 1.9177546256839814e-05,
+      "loss": 0.3681,
+      "step": 170
+    },
+    {
+      "epoch": 0.2630769230769231,
+      "grad_norm": 0.4916614062406827,
+      "learning_rate": 1.9167921953165827e-05,
+      "loss": 0.3896,
+      "step": 171
+    },
+    {
+      "epoch": 0.26461538461538464,
+      "grad_norm": 0.4101182799624699,
+      "learning_rate": 1.9158244108829815e-05,
+      "loss": 0.3848,
+      "step": 172
+    },
+    {
+      "epoch": 0.26615384615384613,
+      "grad_norm": 0.4624296070826894,
+      "learning_rate": 1.9148512780350384e-05,
+      "loss": 0.3859,
+      "step": 173
+    },
+    {
+      "epoch": 0.2676923076923077,
+      "grad_norm": 0.42658888223986424,
+      "learning_rate": 1.9138728024558494e-05,
+      "loss": 0.3963,
+      "step": 174
+    },
+    {
+      "epoch": 0.2692307692307692,
+      "grad_norm": 0.41533743324896594,
+      "learning_rate": 1.9128889898597117e-05,
+      "loss": 0.3871,
+      "step": 175
+    },
+    {
+      "epoch": 0.27076923076923076,
+      "grad_norm": 0.38733094157023334,
+      "learning_rate": 1.91189984599209e-05,
+      "loss": 0.3821,
+      "step": 176
+    },
+    {
+      "epoch": 0.2723076923076923,
+      "grad_norm": 0.40907138544855687,
+      "learning_rate": 1.910905376629585e-05,
+      "loss": 0.3711,
+      "step": 177
+    },
+    {
+      "epoch": 0.27384615384615385,
+      "grad_norm": 0.4086668136967447,
+      "learning_rate": 1.9099055875798974e-05,
+      "loss": 0.3717,
+      "step": 178
+    },
+    {
+      "epoch": 0.2753846153846154,
+      "grad_norm": 0.40042572748262395,
+      "learning_rate": 1.9089004846817947e-05,
+      "loss": 0.3756,
+      "step": 179
+    },
+    {
+      "epoch": 0.27692307692307694,
+      "grad_norm": 0.4409557791264853,
+      "learning_rate": 1.9078900738050776e-05,
+      "loss": 0.3754,
+      "step": 180
+    },
+    {
+      "epoch": 0.2784615384615385,
+      "grad_norm": 0.37915178868737887,
+      "learning_rate": 1.9068743608505454e-05,
+      "loss": 0.3962,
+      "step": 181
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4396504431348858,
+      "learning_rate": 1.905853351749962e-05,
+      "loss": 0.3716,
+      "step": 182
+    },
+    {
+      "epoch": 0.2815384615384615,
+      "grad_norm": 0.3929750004198895,
+      "learning_rate": 1.9048270524660197e-05,
+      "loss": 0.4084,
+      "step": 183
+    },
+    {
+      "epoch": 0.28307692307692306,
+      "grad_norm": 0.43687233367491674,
+      "learning_rate": 1.903795468992306e-05,
+      "loss": 0.3803,
+      "step": 184
+    },
+    {
+      "epoch": 0.2846153846153846,
+      "grad_norm": 0.41326035092763636,
+      "learning_rate": 1.902758607353269e-05,
+      "loss": 0.3853,
+      "step": 185
+    },
+    {
+      "epoch": 0.28615384615384615,
+      "grad_norm": 0.3999146480513722,
+      "learning_rate": 1.9017164736041795e-05,
+      "loss": 0.3819,
+      "step": 186
+    },
+    {
+      "epoch": 0.2876923076923077,
+      "grad_norm": 0.42335995803261006,
+      "learning_rate": 1.9006690738310988e-05,
+      "loss": 0.3789,
+      "step": 187
+    },
+    {
+      "epoch": 0.28923076923076924,
+      "grad_norm": 0.39794892598800413,
+      "learning_rate": 1.8996164141508412e-05,
+      "loss": 0.3688,
+      "step": 188
+    },
+    {
+      "epoch": 0.2907692307692308,
+      "grad_norm": 0.4135897671699664,
+      "learning_rate": 1.898558500710939e-05,
+      "loss": 0.3848,
+      "step": 189
+    },
+    {
+      "epoch": 0.2923076923076923,
+      "grad_norm": 0.3843262612378958,
+      "learning_rate": 1.8974953396896066e-05,
+      "loss": 0.3538,
+      "step": 190
+    },
+    {
+      "epoch": 0.29384615384615387,
+      "grad_norm": 0.47092394810543503,
+      "learning_rate": 1.896426937295704e-05,
+      "loss": 0.376,
+      "step": 191
+    },
+    {
+      "epoch": 0.2953846153846154,
+      "grad_norm": 0.3879305717589152,
+      "learning_rate": 1.8953532997687008e-05,
+      "loss": 0.3725,
+      "step": 192
+    },
+    {
+      "epoch": 0.2969230769230769,
+      "grad_norm": 0.41590458489361176,
+      "learning_rate": 1.89427443337864e-05,
+      "loss": 0.3871,
+      "step": 193
+    },
+    {
+      "epoch": 0.29846153846153844,
+      "grad_norm": 0.3879527137700781,
+      "learning_rate": 1.8931903444261007e-05,
+      "loss": 0.3847,
+      "step": 194
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4025235904243639,
+      "learning_rate": 1.8921010392421628e-05,
+      "loss": 0.3811,
+      "step": 195
+    },
+    {
+      "epoch": 0.30153846153846153,
+      "grad_norm": 0.41095845301346456,
+      "learning_rate": 1.891006524188368e-05,
+      "loss": 0.39,
+      "step": 196
+    },
+    {
+      "epoch": 0.3030769230769231,
+      "grad_norm": 0.37705794038575385,
+      "learning_rate": 1.889906805656684e-05,
+      "loss": 0.381,
+      "step": 197
+    },
+    {
+      "epoch": 0.3046153846153846,
+      "grad_norm": 0.43702054483804265,
+      "learning_rate": 1.888801890069467e-05,
+      "loss": 0.3527,
+      "step": 198
+    },
+    {
+      "epoch": 0.30615384615384617,
+      "grad_norm": 0.3792774587926141,
+      "learning_rate": 1.8876917838794226e-05,
+      "loss": 0.3689,
+      "step": 199
+    },
+    {
+      "epoch": 0.3076923076923077,
+      "grad_norm": 0.3956973889881983,
+      "learning_rate": 1.886576493569572e-05,
+      "loss": 0.3756,
+      "step": 200
+    },
+    {
+      "epoch": 0.30923076923076925,
+      "grad_norm": 0.39847179394020243,
+      "learning_rate": 1.8854560256532098e-05,
+      "loss": 0.3608,
+      "step": 201
+    },
+    {
+      "epoch": 0.31076923076923074,
+      "grad_norm": 0.38291825513309125,
+      "learning_rate": 1.884330386673869e-05,
+      "loss": 0.3778,
+      "step": 202
+    },
+    {
+      "epoch": 0.3123076923076923,
+      "grad_norm": 0.3836369231633009,
+      "learning_rate": 1.8831995832052802e-05,
+      "loss": 0.3784,
+      "step": 203
+    },
+    {
+      "epoch": 0.31384615384615383,
+      "grad_norm": 0.357176637420167,
+      "learning_rate": 1.8820636218513354e-05,
+      "loss": 0.3687,
+      "step": 204
+    },
+    {
+      "epoch": 0.3153846153846154,
+      "grad_norm": 0.38895215556350554,
+      "learning_rate": 1.8809225092460488e-05,
+      "loss": 0.3821,
+      "step": 205
+    },
+    {
+      "epoch": 0.3169230769230769,
+      "grad_norm": 0.40869727785874316,
+      "learning_rate": 1.8797762520535178e-05,
+      "loss": 0.3723,
+      "step": 206
+    },
+    {
+      "epoch": 0.31846153846153846,
+      "grad_norm": 0.3504806444347977,
+      "learning_rate": 1.8786248569678847e-05,
+      "loss": 0.3751,
+      "step": 207
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4091856994188944,
+      "learning_rate": 1.8774683307132956e-05,
+      "loss": 0.3819,
+      "step": 208
+    },
+    {
+      "epoch": 0.32153846153846155,
+      "grad_norm": 0.36084489955312327,
+      "learning_rate": 1.8763066800438638e-05,
+      "loss": 0.3769,
+      "step": 209
+    },
+    {
+      "epoch": 0.3230769230769231,
+      "grad_norm": 0.3683966793046095,
+      "learning_rate": 1.8751399117436292e-05,
+      "loss": 0.3715,
+      "step": 210
+    },
+    {
+      "epoch": 0.32461538461538464,
+      "grad_norm": 0.3783538129221332,
+      "learning_rate": 1.873968032626518e-05,
+      "loss": 0.3791,
+      "step": 211
+    },
+    {
+      "epoch": 0.3261538461538461,
+      "grad_norm": 0.4194100108759895,
+      "learning_rate": 1.8727910495363043e-05,
+      "loss": 0.3769,
+      "step": 212
+    },
+    {
+      "epoch": 0.32769230769230767,
+      "grad_norm": 0.3720277420480326,
+      "learning_rate": 1.8716089693465696e-05,
+      "loss": 0.3644,
+      "step": 213
+    },
+    {
+      "epoch": 0.3292307692307692,
+      "grad_norm": 0.45815837742977156,
+      "learning_rate": 1.8704217989606606e-05,
+      "loss": 0.3794,
+      "step": 214
+    },
+    {
+      "epoch": 0.33076923076923076,
+      "grad_norm": 0.3937738257961415,
+      "learning_rate": 1.869229545311653e-05,
+      "loss": 0.3769,
+      "step": 215
+    },
+    {
+      "epoch": 0.3323076923076923,
+      "grad_norm": 0.39161147744491276,
+      "learning_rate": 1.8680322153623077e-05,
+      "loss": 0.3691,
+      "step": 216
+    },
+    {
+      "epoch": 0.33384615384615385,
+      "grad_norm": 0.44681195919288336,
+      "learning_rate": 1.8668298161050308e-05,
+      "loss": 0.382,
+      "step": 217
+    },
+    {
+      "epoch": 0.3353846153846154,
+      "grad_norm": 0.37456194173188356,
+      "learning_rate": 1.8656223545618345e-05,
+      "loss": 0.3713,
+      "step": 218
+    },
+    {
+      "epoch": 0.33692307692307694,
+      "grad_norm": 0.4024250694855577,
+      "learning_rate": 1.8644098377842934e-05,
+      "loss": 0.3715,
+      "step": 219
+    },
+    {
+      "epoch": 0.3384615384615385,
+      "grad_norm": 0.4101469310833317,
+      "learning_rate": 1.8631922728535054e-05,
+      "loss": 0.3749,
+      "step": 220
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3626402021647864,
+      "learning_rate": 1.8619696668800494e-05,
+      "loss": 0.3717,
+      "step": 221
+    },
+    {
+      "epoch": 0.3415384615384615,
+      "grad_norm": 0.40803246104051527,
+      "learning_rate": 1.860742027003944e-05,
+      "loss": 0.3649,
+      "step": 222
+    },
+    {
+      "epoch": 0.34307692307692306,
+      "grad_norm": 0.37304666136452697,
+      "learning_rate": 1.8595093603946053e-05,
+      "loss": 0.368,
+      "step": 223
+    },
+    {
+      "epoch": 0.3446153846153846,
+      "grad_norm": 0.38153823696500755,
+      "learning_rate": 1.8582716742508066e-05,
+      "loss": 0.3715,
+      "step": 224
+    },
+    {
+      "epoch": 0.34615384615384615,
+      "grad_norm": 0.3889642919120175,
+      "learning_rate": 1.8570289758006346e-05,
+      "loss": 0.3718,
+      "step": 225
+    },
+    {
+      "epoch": 0.3476923076923077,
+      "grad_norm": 0.4045162153808238,
+      "learning_rate": 1.8557812723014476e-05,
+      "loss": 0.3713,
+      "step": 226
+    },
+    {
+      "epoch": 0.34923076923076923,
+      "grad_norm": 0.4218933050125384,
+      "learning_rate": 1.8545285710398343e-05,
+      "loss": 0.3709,
+      "step": 227
+    },
+    {
+      "epoch": 0.3507692307692308,
+      "grad_norm": 0.37857910705496106,
+      "learning_rate": 1.853270879331569e-05,
+      "loss": 0.3734,
+      "step": 228
+    },
+    {
+      "epoch": 0.3523076923076923,
+      "grad_norm": 0.42090559910885333,
+      "learning_rate": 1.852008204521572e-05,
+      "loss": 0.3604,
+      "step": 229
+    },
+    {
+      "epoch": 0.35384615384615387,
+      "grad_norm": 0.40221890351118744,
+      "learning_rate": 1.850740553983863e-05,
+      "loss": 0.3711,
+      "step": 230
+    },
+    {
+      "epoch": 0.3553846153846154,
+      "grad_norm": 0.3831927418726805,
+      "learning_rate": 1.8494679351215212e-05,
+      "loss": 0.3651,
+      "step": 231
+    },
+    {
+      "epoch": 0.3569230769230769,
+      "grad_norm": 0.39805114519231477,
+      "learning_rate": 1.8481903553666405e-05,
+      "loss": 0.3861,
+      "step": 232
+    },
+    {
+      "epoch": 0.35846153846153844,
+      "grad_norm": 0.4268531301053383,
+      "learning_rate": 1.846907822180286e-05,
+      "loss": 0.3796,
+      "step": 233
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3882909628821941,
+      "learning_rate": 1.845620343052452e-05,
+      "loss": 0.3694,
+      "step": 234
+    },
+    {
+      "epoch": 0.36153846153846153,
+      "grad_norm": 0.41238671116025616,
+      "learning_rate": 1.8443279255020153e-05,
+      "loss": 0.3567,
+      "step": 235
+    },
+    {
+      "epoch": 0.3630769230769231,
+      "grad_norm": 0.40874687018717815,
+      "learning_rate": 1.8430305770766947e-05,
+      "loss": 0.3725,
+      "step": 236
+    },
+    {
+      "epoch": 0.3646153846153846,
+      "grad_norm": 0.3919623738207391,
+      "learning_rate": 1.8417283053530047e-05,
+      "loss": 0.3724,
+      "step": 237
+    },
+    {
+      "epoch": 0.36615384615384616,
+      "grad_norm": 0.38426088368672884,
+      "learning_rate": 1.8404211179362116e-05,
+      "loss": 0.364,
+      "step": 238
+    },
+    {
+      "epoch": 0.3676923076923077,
+      "grad_norm": 0.3970759740223237,
+      "learning_rate": 1.8391090224602895e-05,
+      "loss": 0.3749,
+      "step": 239
+    },
+    {
+      "epoch": 0.36923076923076925,
+      "grad_norm": 0.36677393144196546,
+      "learning_rate": 1.837792026587876e-05,
+      "loss": 0.3605,
+      "step": 240
+    },
+    {
+      "epoch": 0.3707692307692308,
+      "grad_norm": 0.4213334994808569,
+      "learning_rate": 1.8364701380102267e-05,
+      "loss": 0.3645,
+      "step": 241
+    },
+    {
+      "epoch": 0.3723076923076923,
+      "grad_norm": 0.3937438051197529,
+      "learning_rate": 1.8351433644471708e-05,
+      "loss": 0.3567,
+      "step": 242
+    },
+    {
+      "epoch": 0.37384615384615383,
+      "grad_norm": 0.4373284356181688,
+      "learning_rate": 1.8338117136470648e-05,
+      "loss": 0.3954,
+      "step": 243
+    },
+    {
+      "epoch": 0.37538461538461537,
+      "grad_norm": 0.3808473799670725,
+      "learning_rate": 1.8324751933867496e-05,
+      "loss": 0.3761,
+      "step": 244
+    },
+    {
+      "epoch": 0.3769230769230769,
+      "grad_norm": 0.4092145904169988,
+      "learning_rate": 1.831133811471503e-05,
+      "loss": 0.3766,
+      "step": 245
+    },
+    {
+      "epoch": 0.37846153846153846,
+      "grad_norm": 0.36968721732711635,
+      "learning_rate": 1.829787575734995e-05,
+      "loss": 0.3728,
+      "step": 246
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3717174342060678,
+      "learning_rate": 1.8284364940392426e-05,
+      "loss": 0.3703,
+      "step": 247
+    },
+    {
+      "epoch": 0.38153846153846155,
+      "grad_norm": 0.375042860567386,
+      "learning_rate": 1.827080574274562e-05,
+      "loss": 0.3714,
+      "step": 248
+    },
+    {
+      "epoch": 0.3830769230769231,
+      "grad_norm": 0.3907920587288226,
+      "learning_rate": 1.825719824359524e-05,
+      "loss": 0.3756,
+      "step": 249
+    },
+    {
+      "epoch": 0.38461538461538464,
+      "grad_norm": 0.3982540672258603,
+      "learning_rate": 1.824354252240909e-05,
+      "loss": 0.3738,
+      "step": 250
+    },
+    {
+      "epoch": 0.3861538461538462,
+      "grad_norm": 0.361205484939361,
+      "learning_rate": 1.8229838658936566e-05,
+      "loss": 0.3632,
+      "step": 251
+    },
+    {
+      "epoch": 0.38769230769230767,
+      "grad_norm": 0.3890112620094363,
+      "learning_rate": 1.821608673320823e-05,
+      "loss": 0.3578,
+      "step": 252
+    },
+    {
+      "epoch": 0.3892307692307692,
+      "grad_norm": 0.3713523130300497,
+      "learning_rate": 1.820228682553533e-05,
+      "loss": 0.3567,
+      "step": 253
+    },
+    {
+      "epoch": 0.39076923076923076,
+      "grad_norm": 0.405654695757502,
+      "learning_rate": 1.818843901650932e-05,
+      "loss": 0.3722,
+      "step": 254
+    },
+    {
+      "epoch": 0.3923076923076923,
+      "grad_norm": 0.37446852915147966,
+      "learning_rate": 1.8174543387001403e-05,
+      "loss": 0.3705,
+      "step": 255
+    },
+    {
+      "epoch": 0.39384615384615385,
+      "grad_norm": 0.36557641948445174,
+      "learning_rate": 1.816060001816205e-05,
+      "loss": 0.3724,
+      "step": 256
+    },
+    {
+      "epoch": 0.3953846153846154,
+      "grad_norm": 0.41409635686326385,
+      "learning_rate": 1.8146608991420533e-05,
+      "loss": 0.3684,
+      "step": 257
+    },
+    {
+      "epoch": 0.39692307692307693,
+      "grad_norm": 0.4095941216657584,
+      "learning_rate": 1.8132570388484442e-05,
+      "loss": 0.377,
+      "step": 258
+    },
+    {
+      "epoch": 0.3984615384615385,
+      "grad_norm": 0.37105953110806256,
+      "learning_rate": 1.811848429133922e-05,
+      "loss": 0.3643,
+      "step": 259
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.44205163106595546,
+      "learning_rate": 1.810435078224767e-05,
+      "loss": 0.3669,
+      "step": 260
+    },
+    {
+      "epoch": 0.4015384615384615,
+      "grad_norm": 0.38336389303729085,
+      "learning_rate": 1.8090169943749477e-05,
+      "loss": 0.3856,
+      "step": 261
+    },
+    {
+      "epoch": 0.40307692307692305,
+      "grad_norm": 0.464235800351049,
+      "learning_rate": 1.8075941858660737e-05,
+      "loss": 0.3638,
+      "step": 262
+    },
+    {
+      "epoch": 0.4046153846153846,
+      "grad_norm": 0.4152673795017994,
+      "learning_rate": 1.8061666610073465e-05,
+      "loss": 0.3693,
+      "step": 263
+    },
+    {
+      "epoch": 0.40615384615384614,
+      "grad_norm": 0.43489061973945825,
+      "learning_rate": 1.8047344281355112e-05,
+      "loss": 0.3724,
+      "step": 264
+    },
+    {
+      "epoch": 0.4076923076923077,
+      "grad_norm": 0.4150301949054034,
+      "learning_rate": 1.8032974956148064e-05,
+      "loss": 0.3674,
+      "step": 265
+    },
+    {
+      "epoch": 0.40923076923076923,
+      "grad_norm": 0.4497060976787947,
+      "learning_rate": 1.8018558718369187e-05,
+      "loss": 0.3684,
+      "step": 266
+    },
+    {
+      "epoch": 0.4107692307692308,
+      "grad_norm": 0.3902334643302586,
+      "learning_rate": 1.8004095652209304e-05,
+      "loss": 0.3556,
+      "step": 267
+    },
+    {
+      "epoch": 0.4123076923076923,
+      "grad_norm": 0.42095187141286594,
+      "learning_rate": 1.7989585842132713e-05,
+      "loss": 0.3731,
+      "step": 268
+    },
+    {
+      "epoch": 0.41384615384615386,
+      "grad_norm": 0.3818670731000476,
+      "learning_rate": 1.7975029372876706e-05,
+      "loss": 0.3601,
+      "step": 269
+    },
+    {
+      "epoch": 0.4153846153846154,
+      "grad_norm": 0.4103882136018874,
+      "learning_rate": 1.7960426329451062e-05,
+      "loss": 0.3642,
+      "step": 270
+    },
+    {
+      "epoch": 0.4169230769230769,
+      "grad_norm": 0.39617920284419667,
+      "learning_rate": 1.7945776797137544e-05,
+      "loss": 0.38,
+      "step": 271
+    },
+    {
+      "epoch": 0.41846153846153844,
+      "grad_norm": 0.4410136640970846,
+      "learning_rate": 1.7931080861489425e-05,
+      "loss": 0.372,
+      "step": 272
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3879162963512455,
+      "learning_rate": 1.791633860833096e-05,
+      "loss": 0.3631,
+      "step": 273
+    },
+    {
+      "epoch": 0.42153846153846153,
+      "grad_norm": 0.40257973390034796,
+      "learning_rate": 1.7901550123756906e-05,
+      "loss": 0.379,
+      "step": 274
+    },
+    {
+      "epoch": 0.4230769230769231,
+      "grad_norm": 0.3585833578527905,
+      "learning_rate": 1.7886715494132008e-05,
+      "loss": 0.3595,
+      "step": 275
+    },
+    {
+      "epoch": 0.4246153846153846,
+      "grad_norm": 0.3848832106066258,
+      "learning_rate": 1.7871834806090502e-05,
+      "loss": 0.3563,
+      "step": 276
+    },
+    {
+      "epoch": 0.42615384615384616,
+      "grad_norm": 0.40097532549068327,
+      "learning_rate": 1.7856908146535602e-05,
+      "loss": 0.3654,
+      "step": 277
+    },
+    {
+      "epoch": 0.4276923076923077,
+      "grad_norm": 0.3970403761299138,
+      "learning_rate": 1.7841935602638997e-05,
+      "loss": 0.366,
+      "step": 278
+    },
+    {
+      "epoch": 0.42923076923076925,
+      "grad_norm": 0.40596536626574137,
+      "learning_rate": 1.7826917261840337e-05,
+      "loss": 0.3669,
+      "step": 279
+    },
+    {
+      "epoch": 0.4307692307692308,
+      "grad_norm": 0.3859590309083557,
+      "learning_rate": 1.781185321184673e-05,
+      "loss": 0.3623,
+      "step": 280
+    },
+    {
+      "epoch": 0.4323076923076923,
+      "grad_norm": 0.3881945336753292,
+      "learning_rate": 1.7796743540632226e-05,
+      "loss": 0.3666,
+      "step": 281
+    },
+    {
+      "epoch": 0.4338461538461538,
+      "grad_norm": 0.3990409977557076,
+      "learning_rate": 1.77815883364373e-05,
+      "loss": 0.3599,
+      "step": 282
+    },
+    {
+      "epoch": 0.43538461538461537,
+      "grad_norm": 0.3828911893070262,
+      "learning_rate": 1.7766387687768338e-05,
+      "loss": 0.3675,
+      "step": 283
+    },
+    {
+      "epoch": 0.4369230769230769,
+      "grad_norm": 0.36751683518870193,
+      "learning_rate": 1.7751141683397128e-05,
+      "loss": 0.3626,
+      "step": 284
+    },
+    {
+      "epoch": 0.43846153846153846,
+      "grad_norm": 0.3763048147079376,
+      "learning_rate": 1.7735850412360332e-05,
+      "loss": 0.3663,
+      "step": 285
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3730587060578288,
+      "learning_rate": 1.772051396395897e-05,
+      "loss": 0.373,
+      "step": 286
+    },
+    {
+      "epoch": 0.44153846153846155,
+      "grad_norm": 0.39677615448565345,
+      "learning_rate": 1.7705132427757895e-05,
+      "loss": 0.3651,
+      "step": 287
+    },
+    {
+      "epoch": 0.4430769230769231,
+      "grad_norm": 0.3631793887891446,
+      "learning_rate": 1.7689705893585273e-05,
+      "loss": 0.37,
+      "step": 288
+    },
+    {
+      "epoch": 0.44461538461538463,
+      "grad_norm": 0.3760465874622631,
+      "learning_rate": 1.7674234451532065e-05,
+      "loss": 0.3629,
+      "step": 289
+    },
+    {
+      "epoch": 0.4461538461538462,
+      "grad_norm": 0.41986105866909235,
+      "learning_rate": 1.7658718191951483e-05,
+      "loss": 0.372,
+      "step": 290
+    },
+    {
+      "epoch": 0.44769230769230767,
+      "grad_norm": 0.3691763384447798,
+      "learning_rate": 1.7643157205458483e-05,
+      "loss": 0.3699,
+      "step": 291
+    },
+    {
+      "epoch": 0.4492307692307692,
+      "grad_norm": 0.41926969701030004,
+      "learning_rate": 1.7627551582929223e-05,
+      "loss": 0.3834,
+      "step": 292
+    },
+    {
+      "epoch": 0.45076923076923076,
+      "grad_norm": 0.4078689333656551,
+      "learning_rate": 1.7611901415500536e-05,
+      "loss": 0.343,
+      "step": 293
+    },
+    {
+      "epoch": 0.4523076923076923,
+      "grad_norm": 0.43487539817970144,
+      "learning_rate": 1.759620679456939e-05,
+      "loss": 0.3652,
+      "step": 294
+    },
+    {
+      "epoch": 0.45384615384615384,
+      "grad_norm": 0.42514041307688816,
+      "learning_rate": 1.7580467811792374e-05,
+      "loss": 0.3705,
+      "step": 295
+    },
+    {
+      "epoch": 0.4553846153846154,
+      "grad_norm": 0.46665263341350904,
+      "learning_rate": 1.7564684559085138e-05,
+      "loss": 0.3583,
+      "step": 296
+    },
+    {
+      "epoch": 0.45692307692307693,
+      "grad_norm": 0.42300191209413235,
+      "learning_rate": 1.7548857128621878e-05,
+      "loss": 0.3605,
+      "step": 297
+    },
+    {
+      "epoch": 0.4584615384615385,
+      "grad_norm": 0.3680475920181223,
+      "learning_rate": 1.753298561283478e-05,
+      "loss": 0.3607,
+      "step": 298
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.45131611607895433,
+      "learning_rate": 1.7517070104413497e-05,
+      "loss": 0.3624,
+      "step": 299
+    },
+    {
+      "epoch": 0.46153846153846156,
+      "grad_norm": 0.4276973721446335,
+      "learning_rate": 1.7501110696304598e-05,
+      "loss": 0.3754,
+      "step": 300
+    },
+    {
+      "epoch": 0.46307692307692305,
+      "grad_norm": 0.3745387735880758,
+      "learning_rate": 1.7485107481711014e-05,
+      "loss": 0.3579,
+      "step": 301
+    },
+    {
+      "epoch": 0.4646153846153846,
+      "grad_norm": 0.4338815397483956,
+      "learning_rate": 1.7469060554091518e-05,
+      "loss": 0.3752,
+      "step": 302
+    },
+    {
+      "epoch": 0.46615384615384614,
+      "grad_norm": 0.39818355741157024,
+      "learning_rate": 1.745297000716016e-05,
+      "loss": 0.3586,
+      "step": 303
+    },
+    {
+      "epoch": 0.4676923076923077,
+      "grad_norm": 0.3832231313721789,
+      "learning_rate": 1.7436835934885735e-05,
+      "loss": 0.3617,
+      "step": 304
+    },
+    {
+      "epoch": 0.46923076923076923,
+      "grad_norm": 0.4347588091129578,
+      "learning_rate": 1.7420658431491224e-05,
+      "loss": 0.3528,
+      "step": 305
+    },
+    {
+      "epoch": 0.4707692307692308,
+      "grad_norm": 0.38421901062565267,
+      "learning_rate": 1.7404437591453237e-05,
+      "loss": 0.3698,
+      "step": 306
+    },
+    {
+      "epoch": 0.4723076923076923,
+      "grad_norm": 0.4228476161532909,
+      "learning_rate": 1.7388173509501475e-05,
+      "loss": 0.3624,
+      "step": 307
+    },
+    {
+      "epoch": 0.47384615384615386,
+      "grad_norm": 0.3816385472742478,
+      "learning_rate": 1.7371866280618176e-05,
+      "loss": 0.3843,
+      "step": 308
+    },
+    {
+      "epoch": 0.4753846153846154,
+      "grad_norm": 0.40985141023350435,
+      "learning_rate": 1.7355516000037555e-05,
+      "loss": 0.3596,
+      "step": 309
+    },
+    {
+      "epoch": 0.47692307692307695,
+      "grad_norm": 0.3934905369941904,
+      "learning_rate": 1.733912276324524e-05,
+      "loss": 0.3656,
+      "step": 310
+    },
+    {
+      "epoch": 0.47846153846153844,
+      "grad_norm": 0.41809523439614077,
+      "learning_rate": 1.7322686665977738e-05,
+      "loss": 0.3681,
+      "step": 311
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3706461422689419,
+      "learning_rate": 1.7306207804221845e-05,
+      "loss": 0.3485,
+      "step": 312
+    },
+    {
+      "epoch": 0.4815384615384615,
+      "grad_norm": 0.39333779244815203,
+      "learning_rate": 1.7289686274214116e-05,
+      "loss": 0.3473,
+      "step": 313
+    },
+    {
+      "epoch": 0.48307692307692307,
+      "grad_norm": 0.46509841987458855,
+      "learning_rate": 1.727312217244028e-05,
+      "loss": 0.353,
+      "step": 314
+    },
+    {
+      "epoch": 0.4846153846153846,
+      "grad_norm": 0.4146022141801263,
+      "learning_rate": 1.7256515595634688e-05,
+      "loss": 0.3623,
+      "step": 315
+    },
+    {
+      "epoch": 0.48615384615384616,
+      "grad_norm": 0.3928106419082748,
+      "learning_rate": 1.7239866640779745e-05,
+      "loss": 0.3643,
+      "step": 316
+    },
+    {
+      "epoch": 0.4876923076923077,
+      "grad_norm": 0.39057388107462493,
+      "learning_rate": 1.722317540510534e-05,
+      "loss": 0.3591,
+      "step": 317
+    },
+    {
+      "epoch": 0.48923076923076925,
+      "grad_norm": 0.3643026764737496,
+      "learning_rate": 1.720644198608829e-05,
+      "loss": 0.364,
+      "step": 318
+    },
+    {
+      "epoch": 0.4907692307692308,
+      "grad_norm": 0.38549315868399825,
+      "learning_rate": 1.7189666481451755e-05,
+      "loss": 0.3811,
+      "step": 319
+    },
+    {
+      "epoch": 0.49230769230769234,
+      "grad_norm": 0.3564779772944027,
+      "learning_rate": 1.717284898916468e-05,
+      "loss": 0.3592,
+      "step": 320
+    },
+    {
+      "epoch": 0.4938461538461538,
+      "grad_norm": 0.3623249043958133,
+      "learning_rate": 1.715598960744121e-05,
+      "loss": 0.3741,
+      "step": 321
+    },
+    {
+      "epoch": 0.49538461538461537,
+      "grad_norm": 0.38297047290608954,
+      "learning_rate": 1.7139088434740142e-05,
+      "loss": 0.3475,
+      "step": 322
+    },
+    {
+      "epoch": 0.4969230769230769,
+      "grad_norm": 0.34854518066141543,
+      "learning_rate": 1.712214556976431e-05,
+      "loss": 0.3536,
+      "step": 323
+    },
+    {
+      "epoch": 0.49846153846153846,
+      "grad_norm": 0.37993428050784794,
+      "learning_rate": 1.7105161111460046e-05,
+      "loss": 0.3609,
+      "step": 324
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3729380788711938,
+      "learning_rate": 1.7088135159016584e-05,
+      "loss": 0.3723,
+      "step": 325
+    },
+    {
+      "epoch": 0.5015384615384615,
+      "grad_norm": 0.3971735990362918,
+      "learning_rate": 1.7071067811865477e-05,
+      "loss": 0.3717,
+      "step": 326
+    },
+    {
+      "epoch": 0.5030769230769231,
+      "grad_norm": 0.36609792821629655,
+      "learning_rate": 1.7053959169680033e-05,
+      "loss": 0.3679,
+      "step": 327
+    },
+    {
+      "epoch": 0.5046153846153846,
+      "grad_norm": 0.35627566288097,
+      "learning_rate": 1.7036809332374713e-05,
+      "loss": 0.3553,
+      "step": 328
+    },
+    {
+      "epoch": 0.5061538461538462,
+      "grad_norm": 0.420040314834087,
+      "learning_rate": 1.7019618400104572e-05,
+      "loss": 0.3534,
+      "step": 329
+    },
+    {
+      "epoch": 0.5076923076923077,
+      "grad_norm": 0.4174611880156514,
+      "learning_rate": 1.700238647326464e-05,
+      "loss": 0.3535,
+      "step": 330
+    },
+    {
+      "epoch": 0.5092307692307693,
+      "grad_norm": 0.4119343430838273,
+      "learning_rate": 1.6985113652489374e-05,
+      "loss": 0.3625,
+      "step": 331
+    },
+    {
+      "epoch": 0.5107692307692308,
+      "grad_norm": 0.41215189797488533,
+      "learning_rate": 1.6967800038652035e-05,
+      "loss": 0.3637,
+      "step": 332
+    },
+    {
+      "epoch": 0.5123076923076924,
+      "grad_norm": 0.3570445396851198,
+      "learning_rate": 1.695044573286413e-05,
+      "loss": 0.3561,
+      "step": 333
+    },
+    {
+      "epoch": 0.5138461538461538,
+      "grad_norm": 0.3863353465569076,
+      "learning_rate": 1.69330508364748e-05,
+      "loss": 0.3638,
+      "step": 334
+    },
+    {
+      "epoch": 0.5153846153846153,
+      "grad_norm": 0.3654544287272041,
+      "learning_rate": 1.6915615451070234e-05,
+      "loss": 0.382,
+      "step": 335
+    },
+    {
+      "epoch": 0.5169230769230769,
+      "grad_norm": 0.36912345175040406,
+      "learning_rate": 1.689813967847308e-05,
+      "loss": 0.3555,
+      "step": 336
+    },
+    {
+      "epoch": 0.5184615384615384,
+      "grad_norm": 0.41053868236746205,
+      "learning_rate": 1.6880623620741843e-05,
+      "loss": 0.355,
+      "step": 337
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3876075676428127,
+      "learning_rate": 1.68630673801703e-05,
+      "loss": 0.3496,
+      "step": 338
+    },
+    {
+      "epoch": 0.5215384615384615,
+      "grad_norm": 0.4713652353964082,
+      "learning_rate": 1.684547105928689e-05,
+      "loss": 0.3489,
+      "step": 339
+    },
+    {
+      "epoch": 0.5230769230769231,
+      "grad_norm": 0.386022064964522,
+      "learning_rate": 1.682783476085412e-05,
+      "loss": 0.3634,
+      "step": 340
+    },
+    {
+      "epoch": 0.5246153846153846,
+      "grad_norm": 0.41801072255768673,
+      "learning_rate": 1.6810158587867973e-05,
+      "loss": 0.3616,
+      "step": 341
+    },
+    {
+      "epoch": 0.5261538461538462,
+      "grad_norm": 0.42104036433759773,
+      "learning_rate": 1.679244264355729e-05,
+      "loss": 0.3673,
+      "step": 342
+    },
+    {
+      "epoch": 0.5276923076923077,
+      "grad_norm": 0.36242125371991196,
+      "learning_rate": 1.677468703138319e-05,
+      "loss": 0.3778,
+      "step": 343
+    },
+    {
+      "epoch": 0.5292307692307693,
+      "grad_norm": 0.41800015601158497,
+      "learning_rate": 1.6756891855038436e-05,
+      "loss": 0.3586,
+      "step": 344
+    },
+    {
+      "epoch": 0.5307692307692308,
+      "grad_norm": 0.4103509913549403,
+      "learning_rate": 1.673905721844686e-05,
+      "loss": 0.3563,
+      "step": 345
+    },
+    {
+      "epoch": 0.5323076923076923,
+      "grad_norm": 0.39231636301729744,
+      "learning_rate": 1.6721183225762726e-05,
+      "loss": 0.3582,
+      "step": 346
+    },
+    {
+      "epoch": 0.5338461538461539,
+      "grad_norm": 0.41035920838133416,
+      "learning_rate": 1.670326998137016e-05,
+      "loss": 0.3599,
+      "step": 347
+    },
+    {
+      "epoch": 0.5353846153846153,
+      "grad_norm": 0.3817754200468875,
+      "learning_rate": 1.668531758988249e-05,
+      "loss": 0.3572,
+      "step": 348
+    },
+    {
+      "epoch": 0.536923076923077,
+      "grad_norm": 0.39104207433625826,
+      "learning_rate": 1.666732615614169e-05,
+      "loss": 0.374,
+      "step": 349
+    },
+    {
+      "epoch": 0.5384615384615384,
+      "grad_norm": 0.3997257237066698,
+      "learning_rate": 1.6649295785217722e-05,
+      "loss": 0.364,
+      "step": 350
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3983777557756419,
+      "learning_rate": 1.6631226582407954e-05,
+      "loss": 0.3638,
+      "step": 351
+    },
+    {
+      "epoch": 0.5415384615384615,
+      "grad_norm": 0.41299399917390506,
+      "learning_rate": 1.661311865323652e-05,
+      "loss": 0.3662,
+      "step": 352
+    },
+    {
+      "epoch": 0.5430769230769231,
+      "grad_norm": 0.37642504354251777,
+      "learning_rate": 1.6594972103453727e-05,
+      "loss": 0.3683,
+      "step": 353
+    },
+    {
+      "epoch": 0.5446153846153846,
+      "grad_norm": 0.387048345643414,
+      "learning_rate": 1.6576787039035417e-05,
+      "loss": 0.3592,
+      "step": 354
+    },
+    {
+      "epoch": 0.5461538461538461,
+      "grad_norm": 0.4159632032485858,
+      "learning_rate": 1.6558563566182365e-05,
+      "loss": 0.3545,
+      "step": 355
+    },
+    {
+      "epoch": 0.5476923076923077,
+      "grad_norm": 0.3903924199198002,
+      "learning_rate": 1.6540301791319647e-05,
+      "loss": 0.3519,
+      "step": 356
+    },
+    {
+      "epoch": 0.5492307692307692,
+      "grad_norm": 0.3761909674801056,
+      "learning_rate": 1.652200182109602e-05,
+      "loss": 0.3576,
+      "step": 357
+    },
+    {
+      "epoch": 0.5507692307692308,
+      "grad_norm": 0.42728952189676916,
+      "learning_rate": 1.6503663762383312e-05,
+      "loss": 0.3587,
+      "step": 358
+    },
+    {
+      "epoch": 0.5523076923076923,
+      "grad_norm": 0.38286787118485327,
+      "learning_rate": 1.6485287722275783e-05,
+      "loss": 0.3649,
+      "step": 359
+    },
+    {
+      "epoch": 0.5538461538461539,
+      "grad_norm": 0.4252290614548713,
+      "learning_rate": 1.6466873808089496e-05,
+      "loss": 0.3752,
+      "step": 360
+    },
+    {
+      "epoch": 0.5553846153846154,
+      "grad_norm": 0.36603802256061513,
+      "learning_rate": 1.6448422127361707e-05,
+      "loss": 0.3615,
+      "step": 361
+    },
+    {
+      "epoch": 0.556923076923077,
+      "grad_norm": 0.38272811024959796,
+      "learning_rate": 1.642993278785023e-05,
+      "loss": 0.364,
+      "step": 362
+    },
+    {
+      "epoch": 0.5584615384615385,
+      "grad_norm": 0.37121604621728366,
+      "learning_rate": 1.64114058975328e-05,
+      "loss": 0.3576,
+      "step": 363
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3809356299942988,
+      "learning_rate": 1.639284156460646e-05,
+      "loss": 0.3618,
+      "step": 364
+    },
+    {
+      "epoch": 0.5615384615384615,
+      "grad_norm": 0.36442437596578714,
+      "learning_rate": 1.63742398974869e-05,
+      "loss": 0.376,
+      "step": 365
+    },
+    {
+      "epoch": 0.563076923076923,
+      "grad_norm": 0.3726526401264163,
+      "learning_rate": 1.6355601004807856e-05,
+      "loss": 0.367,
+      "step": 366
+    },
+    {
+      "epoch": 0.5646153846153846,
+      "grad_norm": 0.3892076603518679,
+      "learning_rate": 1.6336924995420453e-05,
+      "loss": 0.3582,
+      "step": 367
+    },
+    {
+      "epoch": 0.5661538461538461,
+      "grad_norm": 0.3544147696005193,
+      "learning_rate": 1.6318211978392588e-05,
+      "loss": 0.3639,
+      "step": 368
+    },
+    {
+      "epoch": 0.5676923076923077,
+      "grad_norm": 0.38662679912588155,
+      "learning_rate": 1.6299462063008272e-05,
+      "loss": 0.3621,
+      "step": 369
+    },
+    {
+      "epoch": 0.5692307692307692,
+      "grad_norm": 0.36111330477142,
+      "learning_rate": 1.6280675358767005e-05,
+      "loss": 0.3505,
+      "step": 370
+    },
+    {
+      "epoch": 0.5707692307692308,
+      "grad_norm": 0.34795958415202694,
+      "learning_rate": 1.626185197538314e-05,
+      "loss": 0.3688,
+      "step": 371
+    },
+    {
+      "epoch": 0.5723076923076923,
+      "grad_norm": 0.36954111921569166,
+      "learning_rate": 1.6242992022785225e-05,
+      "loss": 0.3561,
+      "step": 372
+    },
+    {
+      "epoch": 0.5738461538461539,
+      "grad_norm": 0.3559865926724835,
+      "learning_rate": 1.6224095611115385e-05,
+      "loss": 0.3599,
+      "step": 373
+    },
+    {
+      "epoch": 0.5753846153846154,
+      "grad_norm": 0.37753470778136516,
+      "learning_rate": 1.620516285072866e-05,
+      "loss": 0.3625,
+      "step": 374
+    },
+    {
+      "epoch": 0.5769230769230769,
+      "grad_norm": 0.37826053195813236,
+      "learning_rate": 1.6186193852192356e-05,
+      "loss": 0.3459,
+      "step": 375
+    },
+    {
+      "epoch": 0.5784615384615385,
+      "grad_norm": 0.3673975709261409,
+      "learning_rate": 1.6167188726285433e-05,
+      "loss": 0.3582,
+      "step": 376
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3660308610576506,
+      "learning_rate": 1.6148147583997813e-05,
+      "loss": 0.3684,
+      "step": 377
+    },
+    {
+      "epoch": 0.5815384615384616,
+      "grad_norm": 0.3950059500020442,
+      "learning_rate": 1.6129070536529767e-05,
+      "loss": 0.356,
+      "step": 378
+    },
+    {
+      "epoch": 0.583076923076923,
+      "grad_norm": 0.36353658366078445,
+      "learning_rate": 1.6109957695291246e-05,
+      "loss": 0.356,
+      "step": 379
+    },
+    {
+      "epoch": 0.5846153846153846,
+      "grad_norm": 0.3774949701161337,
+      "learning_rate": 1.6090809171901237e-05,
+      "loss": 0.3382,
+      "step": 380
+    },
+    {
+      "epoch": 0.5861538461538461,
+      "grad_norm": 0.38616944560142397,
+      "learning_rate": 1.6071625078187113e-05,
+      "loss": 0.361,
+      "step": 381
+    },
+    {
+      "epoch": 0.5876923076923077,
+      "grad_norm": 0.3561601113387756,
+      "learning_rate": 1.605240552618398e-05,
+      "loss": 0.3678,
+      "step": 382
+    },
+    {
+      "epoch": 0.5892307692307692,
+      "grad_norm": 0.3567228172902658,
+      "learning_rate": 1.603315062813401e-05,
+      "loss": 0.3573,
+      "step": 383
+    },
+    {
+      "epoch": 0.5907692307692308,
+      "grad_norm": 0.35209150938255385,
+      "learning_rate": 1.601386049648581e-05,
+      "loss": 0.3643,
+      "step": 384
+    },
+    {
+      "epoch": 0.5923076923076923,
+      "grad_norm": 0.36175594810149636,
+      "learning_rate": 1.5994535243893742e-05,
+      "loss": 0.3444,
+      "step": 385
+    },
+    {
+      "epoch": 0.5938461538461538,
+      "grad_norm": 0.34297219244708804,
+      "learning_rate": 1.5975174983217273e-05,
+      "loss": 0.3556,
+      "step": 386
+    },
+    {
+      "epoch": 0.5953846153846154,
+      "grad_norm": 0.3421636304731148,
+      "learning_rate": 1.5955779827520327e-05,
+      "loss": 0.3442,
+      "step": 387
+    },
+    {
+      "epoch": 0.5969230769230769,
+      "grad_norm": 0.33771797801592485,
+      "learning_rate": 1.5936349890070602e-05,
+      "loss": 0.3614,
+      "step": 388
+    },
+    {
+      "epoch": 0.5984615384615385,
+      "grad_norm": 0.37068105648921745,
+      "learning_rate": 1.5916885284338937e-05,
+      "loss": 0.3674,
+      "step": 389
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3746975557051116,
+      "learning_rate": 1.5897386123998613e-05,
+      "loss": 0.3564,
+      "step": 390
+    },
+    {
+      "epoch": 0.6015384615384616,
+      "grad_norm": 0.3493416644101713,
+      "learning_rate": 1.5877852522924733e-05,
+      "loss": 0.3556,
+      "step": 391
+    },
+    {
+      "epoch": 0.6030769230769231,
+      "grad_norm": 0.35773132127287827,
+      "learning_rate": 1.5858284595193514e-05,
+      "loss": 0.3476,
+      "step": 392
+    },
+    {
+      "epoch": 0.6046153846153847,
+      "grad_norm": 0.36517959964440383,
+      "learning_rate": 1.5838682455081657e-05,
+      "loss": 0.3658,
+      "step": 393
+    },
+    {
+      "epoch": 0.6061538461538462,
+      "grad_norm": 0.35429117734908727,
+      "learning_rate": 1.581904621706565e-05,
+      "loss": 0.3584,
+      "step": 394
+    },
+    {
+      "epoch": 0.6076923076923076,
+      "grad_norm": 0.3527979306903834,
+      "learning_rate": 1.5799375995821116e-05,
+      "loss": 0.3573,
+      "step": 395
+    },
+    {
+      "epoch": 0.6092307692307692,
+      "grad_norm": 0.37121593614097886,
+      "learning_rate": 1.577967190622215e-05,
+      "loss": 0.359,
+      "step": 396
+    },
+    {
+      "epoch": 0.6107692307692307,
+      "grad_norm": 0.353314357317465,
+      "learning_rate": 1.5759934063340627e-05,
+      "loss": 0.3441,
+      "step": 397
+    },
+    {
+      "epoch": 0.6123076923076923,
+      "grad_norm": 0.3332247310135152,
+      "learning_rate": 1.5740162582445545e-05,
+      "loss": 0.351,
+      "step": 398
+    },
+    {
+      "epoch": 0.6138461538461538,
+      "grad_norm": 0.34771339059491785,
+      "learning_rate": 1.5720357579002346e-05,
+      "loss": 0.3471,
+      "step": 399
+    },
+    {
+      "epoch": 0.6153846153846154,
+      "grad_norm": 0.38330551005394103,
+      "learning_rate": 1.5700519168672248e-05,
+      "loss": 0.3392,
+      "step": 400
+    },
+    {
+      "epoch": 0.6169230769230769,
+      "grad_norm": 0.3503070146055947,
+      "learning_rate": 1.568064746731156e-05,
+      "loss": 0.3564,
+      "step": 401
+    },
+    {
+      "epoch": 0.6184615384615385,
+      "grad_norm": 0.3763447587722038,
+      "learning_rate": 1.5660742590971014e-05,
+      "loss": 0.3511,
+      "step": 402
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.39319321598630663,
+      "learning_rate": 1.5640804655895086e-05,
+      "loss": 0.3607,
+      "step": 403
+    },
+    {
+      "epoch": 0.6215384615384615,
+      "grad_norm": 0.31840850434090867,
+      "learning_rate": 1.5620833778521306e-05,
+      "loss": 0.358,
+      "step": 404
+    },
+    {
+      "epoch": 0.6230769230769231,
+      "grad_norm": 0.3580537475543801,
+      "learning_rate": 1.5600830075479604e-05,
+      "loss": 0.3721,
+      "step": 405
+    },
+    {
+      "epoch": 0.6246153846153846,
+      "grad_norm": 0.35658579017575714,
+      "learning_rate": 1.5580793663591583e-05,
+      "loss": 0.3559,
+      "step": 406
+    },
+    {
+      "epoch": 0.6261538461538462,
+      "grad_norm": 0.35290033279176036,
+      "learning_rate": 1.5560724659869905e-05,
+      "loss": 0.3661,
+      "step": 407
+    },
+    {
+      "epoch": 0.6276923076923077,
+      "grad_norm": 0.3588412679532926,
+      "learning_rate": 1.5540623181517532e-05,
+      "loss": 0.3459,
+      "step": 408
+    },
+    {
+      "epoch": 0.6292307692307693,
+      "grad_norm": 0.34002214619484167,
+      "learning_rate": 1.5520489345927095e-05,
+      "loss": 0.343,
+      "step": 409
+    },
+    {
+      "epoch": 0.6307692307692307,
+      "grad_norm": 0.36308802753849084,
+      "learning_rate": 1.5500323270680194e-05,
+      "loss": 0.3535,
+      "step": 410
+    },
+    {
+      "epoch": 0.6323076923076923,
+      "grad_norm": 0.35004744098628765,
+      "learning_rate": 1.5480125073546705e-05,
+      "loss": 0.3553,
+      "step": 411
+    },
+    {
+      "epoch": 0.6338461538461538,
+      "grad_norm": 0.3691084110432458,
+      "learning_rate": 1.5459894872484083e-05,
+      "loss": 0.3379,
+      "step": 412
+    },
+    {
+      "epoch": 0.6353846153846154,
+      "grad_norm": 0.35297107037054587,
+      "learning_rate": 1.5439632785636707e-05,
+      "loss": 0.362,
+      "step": 413
+    },
+    {
+      "epoch": 0.6369230769230769,
+      "grad_norm": 0.40172027445266095,
+      "learning_rate": 1.5419338931335155e-05,
+      "loss": 0.3595,
+      "step": 414
+    },
+    {
+      "epoch": 0.6384615384615384,
+      "grad_norm": 0.39126279287077753,
+      "learning_rate": 1.539901342809554e-05,
+      "loss": 0.3557,
+      "step": 415
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.38346870946827183,
+      "learning_rate": 1.5378656394618788e-05,
+      "loss": 0.3567,
+      "step": 416
+    },
+    {
+      "epoch": 0.6415384615384615,
+      "grad_norm": 0.37232301943028345,
+      "learning_rate": 1.5358267949789968e-05,
+      "loss": 0.3493,
+      "step": 417
+    },
+    {
+      "epoch": 0.6430769230769231,
+      "grad_norm": 0.3945462684866295,
+      "learning_rate": 1.53378482126776e-05,
+      "loss": 0.3653,
+      "step": 418
+    },
+    {
+      "epoch": 0.6446153846153846,
+      "grad_norm": 0.37108379794189106,
+      "learning_rate": 1.5317397302532933e-05,
+      "loss": 0.3544,
+      "step": 419
+    },
+    {
+      "epoch": 0.6461538461538462,
+      "grad_norm": 0.37311737090932806,
+      "learning_rate": 1.529691533878929e-05,
+      "loss": 0.3509,
+      "step": 420
+    },
+    {
+      "epoch": 0.6476923076923077,
+      "grad_norm": 0.3360949667924758,
+      "learning_rate": 1.527640244106133e-05,
+      "loss": 0.344,
+      "step": 421
+    },
+    {
+      "epoch": 0.6492307692307693,
+      "grad_norm": 0.3875719414280799,
+      "learning_rate": 1.5255858729144368e-05,
+      "loss": 0.3467,
+      "step": 422
+    },
+    {
+      "epoch": 0.6507692307692308,
+      "grad_norm": 0.3657460956884447,
+      "learning_rate": 1.5235284323013674e-05,
+      "loss": 0.3575,
+      "step": 423
+    },
+    {
+      "epoch": 0.6523076923076923,
+      "grad_norm": 0.348905206534666,
+      "learning_rate": 1.5214679342823786e-05,
+      "loss": 0.3637,
+      "step": 424
+    },
+    {
+      "epoch": 0.6538461538461539,
+      "grad_norm": 0.39201965387777377,
+      "learning_rate": 1.5194043908907774e-05,
+      "loss": 0.348,
+      "step": 425
+    },
+    {
+      "epoch": 0.6553846153846153,
+      "grad_norm": 0.3520649547936933,
+      "learning_rate": 1.5173378141776569e-05,
+      "loss": 0.3485,
+      "step": 426
+    },
+    {
+      "epoch": 0.6569230769230769,
+      "grad_norm": 0.36244044595360625,
+      "learning_rate": 1.515268216211825e-05,
+      "loss": 0.3549,
+      "step": 427
+    },
+    {
+      "epoch": 0.6584615384615384,
+      "grad_norm": 0.35748262166766326,
+      "learning_rate": 1.5131956090797326e-05,
+      "loss": 0.3428,
+      "step": 428
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.35884535535220013,
+      "learning_rate": 1.5111200048854055e-05,
+      "loss": 0.3532,
+      "step": 429
+    },
+    {
+      "epoch": 0.6615384615384615,
+      "grad_norm": 0.38487103815147633,
+      "learning_rate": 1.5090414157503715e-05,
+      "loss": 0.359,
+      "step": 430
+    },
+    {
+      "epoch": 0.6630769230769231,
+      "grad_norm": 0.3455309729257337,
+      "learning_rate": 1.5069598538135905e-05,
+      "loss": 0.3537,
+      "step": 431
+    },
+    {
+      "epoch": 0.6646153846153846,
+      "grad_norm": 0.3488707267356114,
+      "learning_rate": 1.504875331231384e-05,
+      "loss": 0.3464,
+      "step": 432
+    },
+    {
+      "epoch": 0.6661538461538462,
+      "grad_norm": 0.36052575798716374,
+      "learning_rate": 1.5027878601773633e-05,
+      "loss": 0.3552,
+      "step": 433
+    },
+    {
+      "epoch": 0.6676923076923077,
+      "grad_norm": 0.3584913040673912,
+      "learning_rate": 1.5006974528423585e-05,
+      "loss": 0.3654,
+      "step": 434
+    },
+    {
+      "epoch": 0.6692307692307692,
+      "grad_norm": 0.34859871457141756,
+      "learning_rate": 1.4986041214343487e-05,
+      "loss": 0.3452,
+      "step": 435
+    },
+    {
+      "epoch": 0.6707692307692308,
+      "grad_norm": 0.3570791899012616,
+      "learning_rate": 1.4965078781783882e-05,
+      "loss": 0.3515,
+      "step": 436
+    },
+    {
+      "epoch": 0.6723076923076923,
+      "grad_norm": 0.3482171625407772,
+      "learning_rate": 1.494408735316537e-05,
+      "loss": 0.3343,
+      "step": 437
+    },
+    {
+      "epoch": 0.6738461538461539,
+      "grad_norm": 0.35354273819215465,
+      "learning_rate": 1.4923067051077893e-05,
+      "loss": 0.3517,
+      "step": 438
+    },
+    {
+      "epoch": 0.6753846153846154,
+      "grad_norm": 0.38831135587524834,
+      "learning_rate": 1.490201799828001e-05,
+      "loss": 0.3443,
+      "step": 439
+    },
+    {
+      "epoch": 0.676923076923077,
+      "grad_norm": 0.3430730062172276,
+      "learning_rate": 1.4880940317698182e-05,
+      "loss": 0.348,
+      "step": 440
+    },
+    {
+      "epoch": 0.6784615384615384,
+      "grad_norm": 0.3653001069982685,
+      "learning_rate": 1.485983413242606e-05,
+      "loss": 0.3519,
+      "step": 441
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.32885334473740424,
+      "learning_rate": 1.4838699565723764e-05,
+      "loss": 0.3628,
+      "step": 442
+    },
+    {
+      "epoch": 0.6815384615384615,
+      "grad_norm": 0.3454416094166431,
+      "learning_rate": 1.4817536741017153e-05,
+      "loss": 0.3585,
+      "step": 443
+    },
+    {
+      "epoch": 0.683076923076923,
+      "grad_norm": 0.34256485922716234,
+      "learning_rate": 1.479634578189712e-05,
+      "loss": 0.3468,
+      "step": 444
+    },
+    {
+      "epoch": 0.6846153846153846,
+      "grad_norm": 0.3744758019962427,
+      "learning_rate": 1.4775126812118865e-05,
+      "loss": 0.3583,
+      "step": 445
+    },
+    {
+      "epoch": 0.6861538461538461,
+      "grad_norm": 0.39796739258118163,
+      "learning_rate": 1.4753879955601162e-05,
+      "loss": 0.3454,
+      "step": 446
+    },
+    {
+      "epoch": 0.6876923076923077,
+      "grad_norm": 0.34010408858290997,
+      "learning_rate": 1.4732605336425651e-05,
+      "loss": 0.3548,
+      "step": 447
+    },
+    {
+      "epoch": 0.6892307692307692,
+      "grad_norm": 0.355844050078347,
+      "learning_rate": 1.4711303078836098e-05,
+      "loss": 0.3503,
+      "step": 448
+    },
+    {
+      "epoch": 0.6907692307692308,
+      "grad_norm": 0.3600060452448229,
+      "learning_rate": 1.4689973307237687e-05,
+      "loss": 0.3419,
+      "step": 449
+    },
+    {
+      "epoch": 0.6923076923076923,
+      "grad_norm": 0.3734313574333026,
+      "learning_rate": 1.466861614619628e-05,
+      "loss": 0.3535,
+      "step": 450
+    },
+    {
+      "epoch": 0.6938461538461539,
+      "grad_norm": 0.3529037241441457,
+      "learning_rate": 1.4647231720437687e-05,
+      "loss": 0.3421,
+      "step": 451
+    },
+    {
+      "epoch": 0.6953846153846154,
+      "grad_norm": 0.3695899703890885,
+      "learning_rate": 1.4625820154846953e-05,
+      "loss": 0.3521,
+      "step": 452
+    },
+    {
+      "epoch": 0.696923076923077,
+      "grad_norm": 0.3633866986704654,
+      "learning_rate": 1.4604381574467616e-05,
+      "loss": 0.3462,
+      "step": 453
+    },
+    {
+      "epoch": 0.6984615384615385,
+      "grad_norm": 0.3441590924732329,
+      "learning_rate": 1.4582916104500977e-05,
+      "loss": 0.3575,
+      "step": 454
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3852899740822867,
+      "learning_rate": 1.4561423870305383e-05,
+      "loss": 0.3642,
+      "step": 455
+    },
+    {
+      "epoch": 0.7015384615384616,
+      "grad_norm": 0.35295351423293775,
+      "learning_rate": 1.4539904997395468e-05,
+      "loss": 0.3603,
+      "step": 456
+    },
+    {
+      "epoch": 0.703076923076923,
+      "grad_norm": 0.33695364956508445,
+      "learning_rate": 1.4518359611441452e-05,
+      "loss": 0.3422,
+      "step": 457
+    },
+    {
+      "epoch": 0.7046153846153846,
+      "grad_norm": 0.355466003310791,
+      "learning_rate": 1.4496787838268378e-05,
+      "loss": 0.3659,
+      "step": 458
+    },
+    {
+      "epoch": 0.7061538461538461,
+      "grad_norm": 0.33637378495889547,
+      "learning_rate": 1.4475189803855399e-05,
+      "loss": 0.3462,
+      "step": 459
+    },
+    {
+      "epoch": 0.7076923076923077,
+      "grad_norm": 0.34678314469018473,
+      "learning_rate": 1.445356563433503e-05,
+      "loss": 0.3406,
+      "step": 460
+    },
+    {
+      "epoch": 0.7092307692307692,
+      "grad_norm": 0.34693573035489583,
+      "learning_rate": 1.4431915455992416e-05,
+      "loss": 0.3417,
+      "step": 461
+    },
+    {
+      "epoch": 0.7107692307692308,
+      "grad_norm": 0.3755760184343091,
+      "learning_rate": 1.4410239395264594e-05,
+      "loss": 0.3666,
+      "step": 462
+    },
+    {
+      "epoch": 0.7123076923076923,
+      "grad_norm": 0.34694634768245153,
+      "learning_rate": 1.438853757873975e-05,
+      "loss": 0.3617,
+      "step": 463
+    },
+    {
+      "epoch": 0.7138461538461538,
+      "grad_norm": 0.35885372749070094,
+      "learning_rate": 1.4366810133156495e-05,
+      "loss": 0.3498,
+      "step": 464
+    },
+    {
+      "epoch": 0.7153846153846154,
+      "grad_norm": 0.3621815063677923,
+      "learning_rate": 1.43450571854031e-05,
+      "loss": 0.3483,
+      "step": 465
+    },
+    {
+      "epoch": 0.7169230769230769,
+      "grad_norm": 0.3790882390066753,
+      "learning_rate": 1.4323278862516774e-05,
+      "loss": 0.3432,
+      "step": 466
+    },
+    {
+      "epoch": 0.7184615384615385,
+      "grad_norm": 0.365764679461306,
+      "learning_rate": 1.430147529168292e-05,
+      "loss": 0.3439,
+      "step": 467
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.36564698796499523,
+      "learning_rate": 1.4279646600234388e-05,
+      "loss": 0.348,
+      "step": 468
+    },
+    {
+      "epoch": 0.7215384615384616,
+      "grad_norm": 0.3986186526610047,
+      "learning_rate": 1.4257792915650728e-05,
+      "loss": 0.3441,
+      "step": 469
+    },
+    {
+      "epoch": 0.7230769230769231,
+      "grad_norm": 0.35796515815199853,
+      "learning_rate": 1.4235914365557455e-05,
+      "loss": 0.3389,
+      "step": 470
+    },
+    {
+      "epoch": 0.7246153846153847,
+      "grad_norm": 0.38322662824975406,
+      "learning_rate": 1.4214011077725293e-05,
+      "loss": 0.3583,
+      "step": 471
+    },
+    {
+      "epoch": 0.7261538461538461,
+      "grad_norm": 0.3875474353135227,
+      "learning_rate": 1.4192083180069441e-05,
+      "loss": 0.3616,
+      "step": 472
+    },
+    {
+      "epoch": 0.7276923076923076,
+      "grad_norm": 0.38829914667930143,
+      "learning_rate": 1.4170130800648814e-05,
+      "loss": 0.356,
+      "step": 473
+    },
+    {
+      "epoch": 0.7292307692307692,
+      "grad_norm": 0.3782542137652294,
+      "learning_rate": 1.4148154067665305e-05,
+      "loss": 0.3582,
+      "step": 474
+    },
+    {
+      "epoch": 0.7307692307692307,
+      "grad_norm": 0.3933783191605548,
+      "learning_rate": 1.4126153109463025e-05,
+      "loss": 0.3422,
+      "step": 475
+    },
+    {
+      "epoch": 0.7323076923076923,
+      "grad_norm": 0.39980162363393457,
+      "learning_rate": 1.410412805452757e-05,
+      "loss": 0.3376,
+      "step": 476
+    },
+    {
+      "epoch": 0.7338461538461538,
+      "grad_norm": 0.3931752301957426,
+      "learning_rate": 1.4082079031485253e-05,
+      "loss": 0.3405,
+      "step": 477
+    },
+    {
+      "epoch": 0.7353846153846154,
+      "grad_norm": 0.38549986736444253,
+      "learning_rate": 1.4060006169102363e-05,
+      "loss": 0.3604,
+      "step": 478
+    },
+    {
+      "epoch": 0.7369230769230769,
+      "grad_norm": 0.38549513066381924,
+      "learning_rate": 1.403790959628441e-05,
+      "loss": 0.363,
+      "step": 479
+    },
+    {
+      "epoch": 0.7384615384615385,
+      "grad_norm": 0.392485531939748,
+      "learning_rate": 1.4015789442075376e-05,
+      "loss": 0.3535,
+      "step": 480
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.36721477684914133,
+      "learning_rate": 1.3993645835656955e-05,
+      "loss": 0.3552,
+      "step": 481
+    },
+    {
+      "epoch": 0.7415384615384616,
+      "grad_norm": 0.3662546158542925,
+      "learning_rate": 1.3971478906347806e-05,
+      "loss": 0.3419,
+      "step": 482
+    },
+    {
+      "epoch": 0.7430769230769231,
+      "grad_norm": 0.3510410161452055,
+      "learning_rate": 1.394928878360279e-05,
+      "loss": 0.3492,
+      "step": 483
+    },
+    {
+      "epoch": 0.7446153846153846,
+      "grad_norm": 0.37521970298000795,
+      "learning_rate": 1.3927075597012215e-05,
+      "loss": 0.3572,
+      "step": 484
+    },
+    {
+      "epoch": 0.7461538461538462,
+      "grad_norm": 0.37116939985280545,
+      "learning_rate": 1.3904839476301091e-05,
+      "loss": 0.3518,
+      "step": 485
+    },
+    {
+      "epoch": 0.7476923076923077,
+      "grad_norm": 0.33881096934542837,
+      "learning_rate": 1.388258055132835e-05,
+      "loss": 0.3498,
+      "step": 486
+    },
+    {
+      "epoch": 0.7492307692307693,
+      "grad_norm": 0.36644333022307424,
+      "learning_rate": 1.3860298952086118e-05,
+      "loss": 0.3465,
+      "step": 487
+    },
+    {
+      "epoch": 0.7507692307692307,
+      "grad_norm": 0.38790855979634725,
+      "learning_rate": 1.383799480869892e-05,
+      "loss": 0.345,
+      "step": 488
+    },
+    {
+      "epoch": 0.7523076923076923,
+      "grad_norm": 0.37472166698238596,
+      "learning_rate": 1.3815668251422953e-05,
+      "loss": 0.3502,
+      "step": 489
+    },
+    {
+      "epoch": 0.7538461538461538,
+      "grad_norm": 0.3427542603425603,
+      "learning_rate": 1.3793319410645307e-05,
+      "loss": 0.3435,
+      "step": 490
+    },
+    {
+      "epoch": 0.7553846153846154,
+      "grad_norm": 0.35112856255783875,
+      "learning_rate": 1.3770948416883205e-05,
+      "loss": 0.3469,
+      "step": 491
+    },
+    {
+      "epoch": 0.7569230769230769,
+      "grad_norm": 0.36959741275757085,
+      "learning_rate": 1.3748555400783245e-05,
+      "loss": 0.3559,
+      "step": 492
+    },
+    {
+      "epoch": 0.7584615384615384,
+      "grad_norm": 0.3302928924698725,
+      "learning_rate": 1.3726140493120639e-05,
+      "loss": 0.3442,
+      "step": 493
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.36854942260784,
+      "learning_rate": 1.3703703824798438e-05,
+      "loss": 0.3431,
+      "step": 494
+    },
+    {
+      "epoch": 0.7615384615384615,
+      "grad_norm": 0.3458940271863601,
+      "learning_rate": 1.3681245526846782e-05,
+      "loss": 0.3513,
+      "step": 495
+    },
+    {
+      "epoch": 0.7630769230769231,
+      "grad_norm": 0.3353051139537092,
+      "learning_rate": 1.3658765730422126e-05,
+      "loss": 0.3597,
+      "step": 496
+    },
+    {
+      "epoch": 0.7646153846153846,
+      "grad_norm": 0.36520927014869203,
+      "learning_rate": 1.3636264566806473e-05,
+      "loss": 0.3576,
+      "step": 497
+    },
+    {
+      "epoch": 0.7661538461538462,
+      "grad_norm": 0.33771512840348084,
+      "learning_rate": 1.3613742167406614e-05,
+      "loss": 0.3536,
+      "step": 498
+    },
+    {
+      "epoch": 0.7676923076923077,
+      "grad_norm": 0.37034650663955004,
+      "learning_rate": 1.3591198663753358e-05,
+      "loss": 0.3381,
+      "step": 499
+    },
+    {
+      "epoch": 0.7692307692307693,
+      "grad_norm": 0.3337216245384059,
+      "learning_rate": 1.3568634187500762e-05,
+      "loss": 0.3456,
+      "step": 500
+    },
+    {
+      "epoch": 0.7707692307692308,
+      "grad_norm": 0.35429837338197046,
+      "learning_rate": 1.3546048870425356e-05,
+      "loss": 0.3387,
+      "step": 501
+    },
+    {
+      "epoch": 0.7723076923076924,
+      "grad_norm": 0.3580341348262144,
+      "learning_rate": 1.3523442844425393e-05,
+      "loss": 0.3576,
+      "step": 502
+    },
+    {
+      "epoch": 0.7738461538461539,
+      "grad_norm": 0.35891519952138096,
+      "learning_rate": 1.3500816241520059e-05,
+      "loss": 0.3472,
+      "step": 503
+    },
+    {
+      "epoch": 0.7753846153846153,
+      "grad_norm": 0.37027324684657026,
+      "learning_rate": 1.3478169193848705e-05,
+      "loss": 0.345,
+      "step": 504
+    },
+    {
+      "epoch": 0.7769230769230769,
+      "grad_norm": 0.36948142626898645,
+      "learning_rate": 1.3455501833670089e-05,
+      "loss": 0.3486,
+      "step": 505
+    },
+    {
+      "epoch": 0.7784615384615384,
+      "grad_norm": 0.3527158306718485,
+      "learning_rate": 1.3432814293361585e-05,
+      "loss": 0.3512,
+      "step": 506
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.34415749552151537,
+      "learning_rate": 1.3410106705418424e-05,
+      "loss": 0.3421,
+      "step": 507
+    },
+    {
+      "epoch": 0.7815384615384615,
+      "grad_norm": 0.3619013684882269,
+      "learning_rate": 1.3387379202452917e-05,
+      "loss": 0.3539,
+      "step": 508
+    },
+    {
+      "epoch": 0.7830769230769231,
+      "grad_norm": 0.3656945870117269,
+      "learning_rate": 1.3364631917193671e-05,
+      "loss": 0.3469,
+      "step": 509
+    },
+    {
+      "epoch": 0.7846153846153846,
+      "grad_norm": 0.3429688549948117,
+      "learning_rate": 1.3341864982484828e-05,
+      "loss": 0.3547,
+      "step": 510
+    },
+    {
+      "epoch": 0.7861538461538462,
+      "grad_norm": 0.36349927736283777,
+      "learning_rate": 1.3319078531285286e-05,
+      "loss": 0.3562,
+      "step": 511
+    },
+    {
+      "epoch": 0.7876923076923077,
+      "grad_norm": 0.3713709942520237,
+      "learning_rate": 1.329627269666791e-05,
+      "loss": 0.3393,
+      "step": 512
+    },
+    {
+      "epoch": 0.7892307692307692,
+      "grad_norm": 0.35076269577930497,
+      "learning_rate": 1.3273447611818768e-05,
+      "loss": 0.3491,
+      "step": 513
+    },
+    {
+      "epoch": 0.7907692307692308,
+      "grad_norm": 0.363674932935047,
+      "learning_rate": 1.3250603410036356e-05,
+      "loss": 0.3432,
+      "step": 514
+    },
+    {
+      "epoch": 0.7923076923076923,
+      "grad_norm": 0.33449536173825667,
+      "learning_rate": 1.3227740224730799e-05,
+      "loss": 0.3389,
+      "step": 515
+    },
+    {
+      "epoch": 0.7938461538461539,
+      "grad_norm": 0.37647494155136074,
+      "learning_rate": 1.3204858189423097e-05,
+      "loss": 0.3493,
+      "step": 516
+    },
+    {
+      "epoch": 0.7953846153846154,
+      "grad_norm": 0.34795941342094305,
+      "learning_rate": 1.3181957437744333e-05,
+      "loss": 0.3565,
+      "step": 517
+    },
+    {
+      "epoch": 0.796923076923077,
+      "grad_norm": 0.3655290189743229,
+      "learning_rate": 1.3159038103434889e-05,
+      "loss": 0.3657,
+      "step": 518
+    },
+    {
+      "epoch": 0.7984615384615384,
+      "grad_norm": 0.353460731339934,
+      "learning_rate": 1.3136100320343674e-05,
+      "loss": 0.3577,
+      "step": 519
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3451195837223336,
+      "learning_rate": 1.3113144222427334e-05,
+      "loss": 0.3323,
+      "step": 520
+    },
+    {
+      "epoch": 0.8015384615384615,
+      "grad_norm": 0.36161080986323013,
+      "learning_rate": 1.3090169943749475e-05,
+      "loss": 0.3525,
+      "step": 521
+    },
+    {
+      "epoch": 0.803076923076923,
+      "grad_norm": 0.3155647623819213,
+      "learning_rate": 1.3067177618479883e-05,
+      "loss": 0.3574,
+      "step": 522
+    },
+    {
+      "epoch": 0.8046153846153846,
+      "grad_norm": 0.3551322155176124,
+      "learning_rate": 1.3044167380893726e-05,
+      "loss": 0.3434,
+      "step": 523
+    },
+    {
+      "epoch": 0.8061538461538461,
+      "grad_norm": 0.35939455893572503,
+      "learning_rate": 1.3021139365370787e-05,
+      "loss": 0.3633,
+      "step": 524
+    },
+    {
+      "epoch": 0.8076923076923077,
+      "grad_norm": 0.36548262735704823,
+      "learning_rate": 1.2998093706394676e-05,
+      "loss": 0.3656,
+      "step": 525
+    },
+    {
+      "epoch": 0.8092307692307692,
+      "grad_norm": 0.3660577514598494,
+      "learning_rate": 1.297503053855203e-05,
+      "loss": 0.3531,
+      "step": 526
+    },
+    {
+      "epoch": 0.8107692307692308,
+      "grad_norm": 0.383634724015809,
+      "learning_rate": 1.295194999653175e-05,
+      "loss": 0.3579,
+      "step": 527
+    },
+    {
+      "epoch": 0.8123076923076923,
+      "grad_norm": 0.35010457369054804,
+      "learning_rate": 1.292885221512419e-05,
+      "loss": 0.3425,
+      "step": 528
+    },
+    {
+      "epoch": 0.8138461538461539,
+      "grad_norm": 0.41440685302820945,
+      "learning_rate": 1.2905737329220394e-05,
+      "loss": 0.3301,
+      "step": 529
+    },
+    {
+      "epoch": 0.8153846153846154,
+      "grad_norm": 0.3168491979178069,
+      "learning_rate": 1.2882605473811282e-05,
+      "loss": 0.3491,
+      "step": 530
+    },
+    {
+      "epoch": 0.816923076923077,
+      "grad_norm": 0.3466892491312682,
+      "learning_rate": 1.2859456783986892e-05,
+      "loss": 0.3441,
+      "step": 531
+    },
+    {
+      "epoch": 0.8184615384615385,
+      "grad_norm": 0.34659205331605203,
+      "learning_rate": 1.2836291394935568e-05,
+      "loss": 0.347,
+      "step": 532
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3294814934478504,
+      "learning_rate": 1.2813109441943166e-05,
+      "loss": 0.3504,
+      "step": 533
+    },
+    {
+      "epoch": 0.8215384615384616,
+      "grad_norm": 0.3353155021431086,
+      "learning_rate": 1.2789911060392295e-05,
+      "loss": 0.3498,
+      "step": 534
+    },
+    {
+      "epoch": 0.823076923076923,
+      "grad_norm": 0.35452675103242876,
+      "learning_rate": 1.2766696385761494e-05,
+      "loss": 0.3448,
+      "step": 535
+    },
+    {
+      "epoch": 0.8246153846153846,
+      "grad_norm": 0.3454011959188025,
+      "learning_rate": 1.274346555362446e-05,
+      "loss": 0.3482,
+      "step": 536
+    },
+    {
+      "epoch": 0.8261538461538461,
+      "grad_norm": 0.3489630238448088,
+      "learning_rate": 1.2720218699649243e-05,
+      "loss": 0.3629,
+      "step": 537
+    },
+    {
+      "epoch": 0.8276923076923077,
+      "grad_norm": 0.3314352804240403,
+      "learning_rate": 1.269695595959747e-05,
+      "loss": 0.3514,
+      "step": 538
+    },
+    {
+      "epoch": 0.8292307692307692,
+      "grad_norm": 0.3309499068989567,
+      "learning_rate": 1.2673677469323532e-05,
+      "loss": 0.3451,
+      "step": 539
+    },
+    {
+      "epoch": 0.8307692307692308,
+      "grad_norm": 0.3575040159510276,
+      "learning_rate": 1.2650383364773812e-05,
+      "loss": 0.3368,
+      "step": 540
+    },
+    {
+      "epoch": 0.8323076923076923,
+      "grad_norm": 0.3571920506581653,
+      "learning_rate": 1.262707378198587e-05,
+      "loss": 0.3328,
+      "step": 541
+    },
+    {
+      "epoch": 0.8338461538461538,
+      "grad_norm": 0.31647257594523176,
+      "learning_rate": 1.2603748857087668e-05,
+      "loss": 0.3555,
+      "step": 542
+    },
+    {
+      "epoch": 0.8353846153846154,
+      "grad_norm": 0.35872309297602645,
+      "learning_rate": 1.258040872629676e-05,
+      "loss": 0.3423,
+      "step": 543
+    },
+    {
+      "epoch": 0.8369230769230769,
+      "grad_norm": 0.34916630594780695,
+      "learning_rate": 1.2557053525919503e-05,
+      "loss": 0.3324,
+      "step": 544
+    },
+    {
+      "epoch": 0.8384615384615385,
+      "grad_norm": 0.38411688464812216,
+      "learning_rate": 1.2533683392350264e-05,
+      "loss": 0.3522,
+      "step": 545
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.34722786376449755,
+      "learning_rate": 1.2510298462070619e-05,
+      "loss": 0.3517,
+      "step": 546
+    },
+    {
+      "epoch": 0.8415384615384616,
+      "grad_norm": 0.33759772731430543,
+      "learning_rate": 1.2486898871648552e-05,
+      "loss": 0.3532,
+      "step": 547
+    },
+    {
+      "epoch": 0.8430769230769231,
+      "grad_norm": 0.34584883605483174,
+      "learning_rate": 1.2463484757737663e-05,
+      "loss": 0.3594,
+      "step": 548
+    },
+    {
+      "epoch": 0.8446153846153847,
+      "grad_norm": 0.3527014422213075,
+      "learning_rate": 1.2440056257076376e-05,
+      "loss": 0.3431,
+      "step": 549
+    },
+    {
+      "epoch": 0.8461538461538461,
+      "grad_norm": 0.33332102329961916,
+      "learning_rate": 1.241661350648713e-05,
+      "loss": 0.3468,
+      "step": 550
+    },
+    {
+      "epoch": 0.8476923076923077,
+      "grad_norm": 0.33805518422743924,
+      "learning_rate": 1.2393156642875579e-05,
+      "loss": 0.3465,
+      "step": 551
+    },
+    {
+      "epoch": 0.8492307692307692,
+      "grad_norm": 0.35378802513994373,
+      "learning_rate": 1.2369685803229802e-05,
+      "loss": 0.3523,
+      "step": 552
+    },
+    {
+      "epoch": 0.8507692307692307,
+      "grad_norm": 0.3534056077972205,
+      "learning_rate": 1.2346201124619502e-05,
+      "loss": 0.3437,
+      "step": 553
+    },
+    {
+      "epoch": 0.8523076923076923,
+      "grad_norm": 0.3661736520283975,
+      "learning_rate": 1.2322702744195192e-05,
+      "loss": 0.3504,
+      "step": 554
+    },
+    {
+      "epoch": 0.8538461538461538,
+      "grad_norm": 0.35447236142528954,
+      "learning_rate": 1.2299190799187405e-05,
+      "loss": 0.3436,
+      "step": 555
+    },
+    {
+      "epoch": 0.8553846153846154,
+      "grad_norm": 0.3466102892973438,
+      "learning_rate": 1.22756654269059e-05,
+      "loss": 0.3419,
+      "step": 556
+    },
+    {
+      "epoch": 0.8569230769230769,
+      "grad_norm": 0.3538602085306258,
+      "learning_rate": 1.2252126764738845e-05,
+      "loss": 0.3386,
+      "step": 557
+    },
+    {
+      "epoch": 0.8584615384615385,
+      "grad_norm": 0.3413042684331384,
+      "learning_rate": 1.222857495015202e-05,
+      "loss": 0.3505,
+      "step": 558
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.34618554289093917,
+      "learning_rate": 1.2205010120688012e-05,
+      "loss": 0.3468,
+      "step": 559
+    },
+    {
+      "epoch": 0.8615384615384616,
+      "grad_norm": 0.35492417094928075,
+      "learning_rate": 1.2181432413965428e-05,
+      "loss": 0.3677,
+      "step": 560
+    },
+    {
+      "epoch": 0.8630769230769231,
+      "grad_norm": 0.3505090771597307,
+      "learning_rate": 1.2157841967678064e-05,
+      "loss": 0.3471,
+      "step": 561
+    },
+    {
+      "epoch": 0.8646153846153846,
+      "grad_norm": 0.34322213549644404,
+      "learning_rate": 1.2134238919594122e-05,
+      "loss": 0.3409,
+      "step": 562
+    },
+    {
+      "epoch": 0.8661538461538462,
+      "grad_norm": 0.34938441253892444,
+      "learning_rate": 1.2110623407555398e-05,
+      "loss": 0.3544,
+      "step": 563
+    },
+    {
+      "epoch": 0.8676923076923077,
+      "grad_norm": 0.3559310998830923,
+      "learning_rate": 1.2086995569476474e-05,
+      "loss": 0.3445,
+      "step": 564
+    },
+    {
+      "epoch": 0.8692307692307693,
+      "grad_norm": 0.32741970143833365,
+      "learning_rate": 1.2063355543343925e-05,
+      "loss": 0.3496,
+      "step": 565
+    },
+    {
+      "epoch": 0.8707692307692307,
+      "grad_norm": 0.3543356528034113,
+      "learning_rate": 1.2039703467215489e-05,
+      "loss": 0.3431,
+      "step": 566
+    },
+    {
+      "epoch": 0.8723076923076923,
+      "grad_norm": 0.3359484538958532,
+      "learning_rate": 1.2016039479219293e-05,
+      "loss": 0.3384,
+      "step": 567
+    },
+    {
+      "epoch": 0.8738461538461538,
+      "grad_norm": 0.3296540762563698,
+      "learning_rate": 1.1992363717553015e-05,
+      "loss": 0.3497,
+      "step": 568
+    },
+    {
+      "epoch": 0.8753846153846154,
+      "grad_norm": 0.33918077141863096,
+      "learning_rate": 1.1968676320483103e-05,
+      "loss": 0.3365,
+      "step": 569
+    },
+    {
+      "epoch": 0.8769230769230769,
+      "grad_norm": 0.35531014163338004,
+      "learning_rate": 1.194497742634395e-05,
+      "loss": 0.344,
+      "step": 570
+    },
+    {
+      "epoch": 0.8784615384615385,
+      "grad_norm": 0.32438629775383093,
+      "learning_rate": 1.1921267173537085e-05,
+      "loss": 0.3401,
+      "step": 571
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3447007370556632,
+      "learning_rate": 1.1897545700530387e-05,
+      "loss": 0.3575,
+      "step": 572
+    },
+    {
+      "epoch": 0.8815384615384615,
+      "grad_norm": 0.34256966113916854,
+      "learning_rate": 1.187381314585725e-05,
+      "loss": 0.3349,
+      "step": 573
+    },
+    {
+      "epoch": 0.8830769230769231,
+      "grad_norm": 0.3722782399770237,
+      "learning_rate": 1.1850069648115785e-05,
+      "loss": 0.3566,
+      "step": 574
+    },
+    {
+      "epoch": 0.8846153846153846,
+      "grad_norm": 0.3594904105249484,
+      "learning_rate": 1.1826315345968014e-05,
+      "loss": 0.3359,
+      "step": 575
+    },
+    {
+      "epoch": 0.8861538461538462,
+      "grad_norm": 0.37788439139952407,
+      "learning_rate": 1.180255037813906e-05,
+      "loss": 0.3476,
+      "step": 576
+    },
+    {
+      "epoch": 0.8876923076923077,
+      "grad_norm": 0.35392483059482593,
+      "learning_rate": 1.1778774883416325e-05,
+      "loss": 0.3369,
+      "step": 577
+    },
+    {
+      "epoch": 0.8892307692307693,
+      "grad_norm": 0.3539767933125652,
+      "learning_rate": 1.1754989000648693e-05,
+      "loss": 0.3469,
+      "step": 578
+    },
+    {
+      "epoch": 0.8907692307692308,
+      "grad_norm": 0.3390764124154708,
+      "learning_rate": 1.1731192868745716e-05,
+      "loss": 0.3434,
+      "step": 579
+    },
+    {
+      "epoch": 0.8923076923076924,
+      "grad_norm": 0.32360695576046367,
+      "learning_rate": 1.1707386626676798e-05,
+      "loss": 0.3513,
+      "step": 580
+    },
+    {
+      "epoch": 0.8938461538461538,
+      "grad_norm": 0.3493125197785865,
+      "learning_rate": 1.1683570413470384e-05,
+      "loss": 0.354,
+      "step": 581
+    },
+    {
+      "epoch": 0.8953846153846153,
+      "grad_norm": 0.33816840237907714,
+      "learning_rate": 1.1659744368213159e-05,
+      "loss": 0.3518,
+      "step": 582
+    },
+    {
+      "epoch": 0.8969230769230769,
+      "grad_norm": 0.3184016605833142,
+      "learning_rate": 1.163590863004922e-05,
+      "loss": 0.3242,
+      "step": 583
+    },
+    {
+      "epoch": 0.8984615384615384,
+      "grad_norm": 0.3566487092290581,
+      "learning_rate": 1.1612063338179269e-05,
+      "loss": 0.3534,
+      "step": 584
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3560701326157053,
+      "learning_rate": 1.1588208631859808e-05,
+      "loss": 0.3437,
+      "step": 585
+    },
+    {
+      "epoch": 0.9015384615384615,
+      "grad_norm": 0.36188986635973247,
+      "learning_rate": 1.156434465040231e-05,
+      "loss": 0.3374,
+      "step": 586
+    },
+    {
+      "epoch": 0.9030769230769231,
+      "grad_norm": 0.34409952856058285,
+      "learning_rate": 1.154047153317243e-05,
+      "loss": 0.3491,
+      "step": 587
+    },
+    {
+      "epoch": 0.9046153846153846,
+      "grad_norm": 0.3939168357449172,
+      "learning_rate": 1.1516589419589159e-05,
+      "loss": 0.354,
+      "step": 588
+    },
+    {
+      "epoch": 0.9061538461538462,
+      "grad_norm": 0.3777832689447866,
+      "learning_rate": 1.1492698449124042e-05,
+      "loss": 0.3398,
+      "step": 589
+    },
+    {
+      "epoch": 0.9076923076923077,
+      "grad_norm": 0.3445223297314293,
+      "learning_rate": 1.1468798761300335e-05,
+      "loss": 0.345,
+      "step": 590
+    },
+    {
+      "epoch": 0.9092307692307692,
+      "grad_norm": 0.3740149056152916,
+      "learning_rate": 1.1444890495692214e-05,
+      "loss": 0.3356,
+      "step": 591
+    },
+    {
+      "epoch": 0.9107692307692308,
+      "grad_norm": 0.37837242377388475,
+      "learning_rate": 1.1420973791923941e-05,
+      "loss": 0.3336,
+      "step": 592
+    },
+    {
+      "epoch": 0.9123076923076923,
+      "grad_norm": 0.38189414219722784,
+      "learning_rate": 1.1397048789669061e-05,
+      "loss": 0.3365,
+      "step": 593
+    },
+    {
+      "epoch": 0.9138461538461539,
+      "grad_norm": 0.3593021584475916,
+      "learning_rate": 1.1373115628649582e-05,
+      "loss": 0.3545,
+      "step": 594
+    },
+    {
+      "epoch": 0.9153846153846154,
+      "grad_norm": 0.3526439037811398,
+      "learning_rate": 1.1349174448635158e-05,
+      "loss": 0.3471,
+      "step": 595
+    },
+    {
+      "epoch": 0.916923076923077,
+      "grad_norm": 0.3625740674710596,
+      "learning_rate": 1.1325225389442278e-05,
+      "loss": 0.3586,
+      "step": 596
+    },
+    {
+      "epoch": 0.9184615384615384,
+      "grad_norm": 0.34507736342174916,
+      "learning_rate": 1.1301268590933434e-05,
+      "loss": 0.3505,
+      "step": 597
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.34257542353995907,
+      "learning_rate": 1.1277304193016332e-05,
+      "loss": 0.3378,
+      "step": 598
+    },
+    {
+      "epoch": 0.9215384615384615,
+      "grad_norm": 0.3689595913126189,
+      "learning_rate": 1.1253332335643043e-05,
+      "loss": 0.3443,
+      "step": 599
+    },
+    {
+      "epoch": 0.9230769230769231,
+      "grad_norm": 0.3829314362398453,
+      "learning_rate": 1.1229353158809216e-05,
+      "loss": 0.3386,
+      "step": 600
+    },
+    {
+      "epoch": 0.9246153846153846,
+      "grad_norm": 0.338780603891905,
+      "learning_rate": 1.1205366802553231e-05,
+      "loss": 0.3505,
+      "step": 601
+    },
+    {
+      "epoch": 0.9261538461538461,
+      "grad_norm": 0.3593594093111377,
+      "learning_rate": 1.118137340695541e-05,
+      "loss": 0.3542,
+      "step": 602
+    },
+    {
+      "epoch": 0.9276923076923077,
+      "grad_norm": 0.3899570986553566,
+      "learning_rate": 1.1157373112137171e-05,
+      "loss": 0.3285,
+      "step": 603
+    },
+    {
+      "epoch": 0.9292307692307692,
+      "grad_norm": 0.3524615219017636,
+      "learning_rate": 1.1133366058260232e-05,
+      "loss": 0.3528,
+      "step": 604
+    },
+    {
+      "epoch": 0.9307692307692308,
+      "grad_norm": 0.3462143256238977,
+      "learning_rate": 1.1109352385525782e-05,
+      "loss": 0.3579,
+      "step": 605
+    },
+    {
+      "epoch": 0.9323076923076923,
+      "grad_norm": 0.35288326812215876,
+      "learning_rate": 1.1085332234173664e-05,
+      "loss": 0.3387,
+      "step": 606
+    },
+    {
+      "epoch": 0.9338461538461539,
+      "grad_norm": 0.37335112680493865,
+      "learning_rate": 1.106130574448156e-05,
+      "loss": 0.3298,
+      "step": 607
+    },
+    {
+      "epoch": 0.9353846153846154,
+      "grad_norm": 0.3326154088699352,
+      "learning_rate": 1.1037273056764157e-05,
+      "loss": 0.3482,
+      "step": 608
+    },
+    {
+      "epoch": 0.936923076923077,
+      "grad_norm": 0.34303809897316306,
+      "learning_rate": 1.1013234311372353e-05,
+      "loss": 0.3387,
+      "step": 609
+    },
+    {
+      "epoch": 0.9384615384615385,
+      "grad_norm": 0.34940035572980127,
+      "learning_rate": 1.0989189648692408e-05,
+      "loss": 0.3432,
+      "step": 610
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.33350776781381053,
+      "learning_rate": 1.0965139209145153e-05,
+      "loss": 0.3337,
+      "step": 611
+    },
+    {
+      "epoch": 0.9415384615384615,
+      "grad_norm": 0.348017444275996,
+      "learning_rate": 1.0941083133185146e-05,
+      "loss": 0.3337,
+      "step": 612
+    },
+    {
+      "epoch": 0.943076923076923,
+      "grad_norm": 0.34458074693378016,
+      "learning_rate": 1.0917021561299864e-05,
+      "loss": 0.345,
+      "step": 613
+    },
+    {
+      "epoch": 0.9446153846153846,
+      "grad_norm": 0.32103679769954235,
+      "learning_rate": 1.089295463400888e-05,
+      "loss": 0.3462,
+      "step": 614
+    },
+    {
+      "epoch": 0.9461538461538461,
+      "grad_norm": 0.33201319528048107,
+      "learning_rate": 1.0868882491863048e-05,
+      "loss": 0.3436,
+      "step": 615
+    },
+    {
+      "epoch": 0.9476923076923077,
+      "grad_norm": 0.32732352522025615,
+      "learning_rate": 1.0844805275443673e-05,
+      "loss": 0.347,
+      "step": 616
+    },
+    {
+      "epoch": 0.9492307692307692,
+      "grad_norm": 0.32188817162482686,
+      "learning_rate": 1.0820723125361685e-05,
+      "loss": 0.3435,
+      "step": 617
+    },
+    {
+      "epoch": 0.9507692307692308,
+      "grad_norm": 0.3376610743006626,
+      "learning_rate": 1.0796636182256846e-05,
+      "loss": 0.3428,
+      "step": 618
+    },
+    {
+      "epoch": 0.9523076923076923,
+      "grad_norm": 0.33984293740217986,
+      "learning_rate": 1.077254458679689e-05,
+      "loss": 0.3387,
+      "step": 619
+    },
+    {
+      "epoch": 0.9538461538461539,
+      "grad_norm": 0.340831435489129,
+      "learning_rate": 1.074844847967673e-05,
+      "loss": 0.3531,
+      "step": 620
+    },
+    {
+      "epoch": 0.9553846153846154,
+      "grad_norm": 0.31999300537272357,
+      "learning_rate": 1.0724348001617626e-05,
+      "loss": 0.3409,
+      "step": 621
+    },
+    {
+      "epoch": 0.9569230769230769,
+      "grad_norm": 0.3558209795432548,
+      "learning_rate": 1.0700243293366365e-05,
+      "loss": 0.359,
+      "step": 622
+    },
+    {
+      "epoch": 0.9584615384615385,
+      "grad_norm": 0.3602203142260847,
+      "learning_rate": 1.0676134495694439e-05,
+      "loss": 0.3281,
+      "step": 623
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.335863396796157,
+      "learning_rate": 1.0652021749397216e-05,
+      "loss": 0.3293,
+      "step": 624
+    },
+    {
+      "epoch": 0.9615384615384616,
+      "grad_norm": 0.3297032096427006,
+      "learning_rate": 1.0627905195293135e-05,
+      "loss": 0.3348,
+      "step": 625
+    },
+    {
+      "epoch": 0.963076923076923,
+      "grad_norm": 0.3400281270117834,
+      "learning_rate": 1.0603784974222862e-05,
+      "loss": 0.348,
+      "step": 626
+    },
+    {
+      "epoch": 0.9646153846153847,
+      "grad_norm": 0.33482256024214785,
+      "learning_rate": 1.0579661227048484e-05,
+      "loss": 0.3476,
+      "step": 627
+    },
+    {
+      "epoch": 0.9661538461538461,
+      "grad_norm": 0.3360076177902745,
+      "learning_rate": 1.0555534094652675e-05,
+      "loss": 0.3337,
+      "step": 628
+    },
+    {
+      "epoch": 0.9676923076923077,
+      "grad_norm": 0.3415880066145126,
+      "learning_rate": 1.0531403717937888e-05,
+      "loss": 0.3453,
+      "step": 629
+    },
+    {
+      "epoch": 0.9692307692307692,
+      "grad_norm": 0.31224262819412874,
+      "learning_rate": 1.0507270237825513e-05,
+      "loss": 0.3519,
+      "step": 630
+    },
+    {
+      "epoch": 0.9707692307692307,
+      "grad_norm": 0.3532661213357234,
+      "learning_rate": 1.0483133795255072e-05,
+      "loss": 0.3289,
+      "step": 631
+    },
+    {
+      "epoch": 0.9723076923076923,
+      "grad_norm": 0.31941447465042055,
+      "learning_rate": 1.045899453118338e-05,
+      "loss": 0.3438,
+      "step": 632
+    },
+    {
+      "epoch": 0.9738461538461538,
+      "grad_norm": 0.3361925498662374,
+      "learning_rate": 1.0434852586583737e-05,
+      "loss": 0.3475,
+      "step": 633
+    },
+    {
+      "epoch": 0.9753846153846154,
+      "grad_norm": 0.33752629556625047,
+      "learning_rate": 1.0410708102445091e-05,
+      "loss": 0.3481,
+      "step": 634
+    },
+    {
+      "epoch": 0.9769230769230769,
+      "grad_norm": 0.3319317450622859,
+      "learning_rate": 1.0386561219771222e-05,
+      "loss": 0.3479,
+      "step": 635
+    },
+    {
+      "epoch": 0.9784615384615385,
+      "grad_norm": 0.35351757256709754,
+      "learning_rate": 1.0362412079579925e-05,
+      "loss": 0.3423,
+      "step": 636
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.33323249680097045,
+      "learning_rate": 1.0338260822902166e-05,
+      "loss": 0.3206,
+      "step": 637
+    },
+    {
+      "epoch": 0.9815384615384616,
+      "grad_norm": 0.3353745062775181,
+      "learning_rate": 1.0314107590781284e-05,
+      "loss": 0.3289,
+      "step": 638
+    },
+    {
+      "epoch": 0.9830769230769231,
+      "grad_norm": 0.36725648761545954,
+      "learning_rate": 1.0289952524272147e-05,
+      "loss": 0.3431,
+      "step": 639
+    },
+    {
+      "epoch": 0.9846153846153847,
+      "grad_norm": 0.3517787986756908,
+      "learning_rate": 1.0265795764440335e-05,
+      "loss": 0.3471,
+      "step": 640
+    },
+    {
+      "epoch": 0.9861538461538462,
+      "grad_norm": 0.3429810760182684,
+      "learning_rate": 1.0241637452361323e-05,
+      "loss": 0.3315,
+      "step": 641
+    },
+    {
+      "epoch": 0.9876923076923076,
+      "grad_norm": 0.30885118309581727,
+      "learning_rate": 1.0217477729119648e-05,
+      "loss": 0.3229,
+      "step": 642
+    },
+    {
+      "epoch": 0.9892307692307692,
+      "grad_norm": 0.34737239468969927,
+      "learning_rate": 1.0193316735808085e-05,
+      "loss": 0.3377,
+      "step": 643
+    },
+    {
+      "epoch": 0.9907692307692307,
+      "grad_norm": 0.35843069481643314,
+      "learning_rate": 1.0169154613526831e-05,
+      "loss": 0.3375,
+      "step": 644
+    },
+    {
+      "epoch": 0.9923076923076923,
+      "grad_norm": 0.34378862978883884,
+      "learning_rate": 1.0144991503382676e-05,
+      "loss": 0.3425,
+      "step": 645
+    },
+    {
+      "epoch": 0.9938461538461538,
+      "grad_norm": 0.3340481201752787,
+      "learning_rate": 1.0120827546488175e-05,
+      "loss": 0.3445,
+      "step": 646
+    },
+    {
+      "epoch": 0.9953846153846154,
+      "grad_norm": 0.3674469423573523,
+      "learning_rate": 1.0096662883960833e-05,
+      "loss": 0.3349,
+      "step": 647
+    },
+    {
+      "epoch": 0.9969230769230769,
+      "grad_norm": 0.3353084737215694,
+      "learning_rate": 1.0072497656922266e-05,
+      "loss": 0.3472,
+      "step": 648
+    },
+    {
+      "epoch": 0.9984615384615385,
+      "grad_norm": 0.35369563263056164,
+      "learning_rate": 1.0048332006497406e-05,
+      "loss": 0.3319,
+      "step": 649
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.3369224805068213,
+      "learning_rate": 1.0024166073813634e-05,
+      "loss": 0.3344,
+      "step": 650
+    },
+    {
+      "epoch": 1.0015384615384615,
+      "grad_norm": 0.4028597131542619,
+      "learning_rate": 1e-05,
+      "loss": 0.3005,
+      "step": 651
+    },
+    {
+      "epoch": 1.003076923076923,
+      "grad_norm": 0.3775850459249221,
+      "learning_rate": 9.975833926186367e-06,
+      "loss": 0.305,
+      "step": 652
+    },
+    {
+      "epoch": 1.0046153846153847,
+      "grad_norm": 0.36045052329330923,
+      "learning_rate": 9.951667993502599e-06,
+      "loss": 0.3161,
+      "step": 653
+    },
+    {
+      "epoch": 1.0061538461538462,
+      "grad_norm": 0.38095412887911145,
+      "learning_rate": 9.927502343077732e-06,
+      "loss": 0.2937,
+      "step": 654
+    },
+    {
+      "epoch": 1.0076923076923077,
+      "grad_norm": 0.40193682268255326,
+      "learning_rate": 9.903337116039172e-06,
+      "loss": 0.3111,
+      "step": 655
+    },
+    {
+      "epoch": 1.0092307692307692,
+      "grad_norm": 0.3902444157296882,
+      "learning_rate": 9.879172453511827e-06,
+      "loss": 0.3114,
+      "step": 656
+    },
+    {
+      "epoch": 1.0107692307692309,
+      "grad_norm": 0.3707405329042767,
+      "learning_rate": 9.855008496617326e-06,
+      "loss": 0.3003,
+      "step": 657
+    },
+    {
+      "epoch": 1.0123076923076924,
+      "grad_norm": 0.3698108757486004,
+      "learning_rate": 9.830845386473169e-06,
+      "loss": 0.3086,
+      "step": 658
+    },
+    {
+      "epoch": 1.0138461538461538,
+      "grad_norm": 0.3725237820755244,
+      "learning_rate": 9.806683264191916e-06,
+      "loss": 0.2948,
+      "step": 659
+    },
+    {
+      "epoch": 1.0153846153846153,
+      "grad_norm": 0.3626535937433891,
+      "learning_rate": 9.782522270880354e-06,
+      "loss": 0.3088,
+      "step": 660
+    },
+    {
+      "epoch": 1.0169230769230768,
+      "grad_norm": 0.3661033120711208,
+      "learning_rate": 9.75836254763868e-06,
+      "loss": 0.3073,
+      "step": 661
+    },
+    {
+      "epoch": 1.0184615384615385,
+      "grad_norm": 0.37150318723384673,
+      "learning_rate": 9.73420423555967e-06,
+      "loss": 0.3041,
+      "step": 662
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.3317628833691345,
+      "learning_rate": 9.710047475727854e-06,
+      "loss": 0.3046,
+      "step": 663
+    },
+    {
+      "epoch": 1.0215384615384615,
+      "grad_norm": 0.3457661198631839,
+      "learning_rate": 9.685892409218718e-06,
+      "loss": 0.3149,
+      "step": 664
+    },
+    {
+      "epoch": 1.023076923076923,
+      "grad_norm": 0.3841069391953467,
+      "learning_rate": 9.661739177097836e-06,
+      "loss": 0.3007,
+      "step": 665
+    },
+    {
+      "epoch": 1.0246153846153847,
+      "grad_norm": 0.3709251555096877,
+      "learning_rate": 9.63758792042008e-06,
+      "loss": 0.3075,
+      "step": 666
+    },
+    {
+      "epoch": 1.0261538461538462,
+      "grad_norm": 0.37912443840426163,
+      "learning_rate": 9.613438780228777e-06,
+      "loss": 0.3031,
+      "step": 667
+    },
+    {
+      "epoch": 1.0276923076923077,
+      "grad_norm": 0.37469870223964,
+      "learning_rate": 9.589291897554912e-06,
+      "loss": 0.2987,
+      "step": 668
+    },
+    {
+      "epoch": 1.0292307692307692,
+      "grad_norm": 0.34569077659683767,
+      "learning_rate": 9.565147413416266e-06,
+      "loss": 0.3147,
+      "step": 669
+    },
+    {
+      "epoch": 1.0307692307692307,
+      "grad_norm": 0.38868044207828484,
+      "learning_rate": 9.541005468816622e-06,
+      "loss": 0.3064,
+      "step": 670
+    },
+    {
+      "epoch": 1.0323076923076924,
+      "grad_norm": 0.3456339215688254,
+      "learning_rate": 9.516866204744932e-06,
+      "loss": 0.3062,
+      "step": 671
+    },
+    {
+      "epoch": 1.0338461538461539,
+      "grad_norm": 0.3345779108281978,
+      "learning_rate": 9.492729762174489e-06,
+      "loss": 0.3148,
+      "step": 672
+    },
+    {
+      "epoch": 1.0353846153846153,
+      "grad_norm": 0.35272249812685313,
+      "learning_rate": 9.468596282062114e-06,
+      "loss": 0.3047,
+      "step": 673
+    },
+    {
+      "epoch": 1.0369230769230768,
+      "grad_norm": 0.34480812202532446,
+      "learning_rate": 9.444465905347327e-06,
+      "loss": 0.303,
+      "step": 674
+    },
+    {
+      "epoch": 1.0384615384615385,
+      "grad_norm": 0.3344521747967752,
+      "learning_rate": 9.420338772951521e-06,
+      "loss": 0.3048,
+      "step": 675
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.35874892066843317,
+      "learning_rate": 9.39621502577714e-06,
+      "loss": 0.3086,
+      "step": 676
+    },
+    {
+      "epoch": 1.0415384615384615,
+      "grad_norm": 0.34744335314802705,
+      "learning_rate": 9.372094804706867e-06,
+      "loss": 0.3071,
+      "step": 677
+    },
+    {
+      "epoch": 1.043076923076923,
+      "grad_norm": 0.3377026249867808,
+      "learning_rate": 9.347978250602786e-06,
+      "loss": 0.3131,
+      "step": 678
+    },
+    {
+      "epoch": 1.0446153846153847,
+      "grad_norm": 0.34758725836641163,
+      "learning_rate": 9.323865504305566e-06,
+      "loss": 0.3087,
+      "step": 679
+    },
+    {
+      "epoch": 1.0461538461538462,
+      "grad_norm": 0.3451402363074388,
+      "learning_rate": 9.299756706633636e-06,
+      "loss": 0.3056,
+      "step": 680
+    },
+    {
+      "epoch": 1.0476923076923077,
+      "grad_norm": 0.3465731760615777,
+      "learning_rate": 9.275651998382377e-06,
+      "loss": 0.3072,
+      "step": 681
+    },
+    {
+      "epoch": 1.0492307692307692,
+      "grad_norm": 0.35245973393380825,
+      "learning_rate": 9.251551520323273e-06,
+      "loss": 0.3115,
+      "step": 682
+    },
+    {
+      "epoch": 1.0507692307692307,
+      "grad_norm": 0.3404317492695898,
+      "learning_rate": 9.227455413203115e-06,
+      "loss": 0.296,
+      "step": 683
+    },
+    {
+      "epoch": 1.0523076923076924,
+      "grad_norm": 0.3684627970623305,
+      "learning_rate": 9.203363817743159e-06,
+      "loss": 0.3008,
+      "step": 684
+    },
+    {
+      "epoch": 1.0538461538461539,
+      "grad_norm": 0.3446734731834438,
+      "learning_rate": 9.179276874638315e-06,
+      "loss": 0.2988,
+      "step": 685
+    },
+    {
+      "epoch": 1.0553846153846154,
+      "grad_norm": 0.3050730974261936,
+      "learning_rate": 9.15519472455633e-06,
+      "loss": 0.3212,
+      "step": 686
+    },
+    {
+      "epoch": 1.0569230769230769,
+      "grad_norm": 0.3715021930088381,
+      "learning_rate": 9.131117508136953e-06,
+      "loss": 0.3119,
+      "step": 687
+    },
+    {
+      "epoch": 1.0584615384615386,
+      "grad_norm": 0.36567888179065444,
+      "learning_rate": 9.107045365991123e-06,
+      "loss": 0.2934,
+      "step": 688
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.32585563102313647,
+      "learning_rate": 9.082978438700138e-06,
+      "loss": 0.3037,
+      "step": 689
+    },
+    {
+      "epoch": 1.0615384615384615,
+      "grad_norm": 0.33498395115376667,
+      "learning_rate": 9.058916866814857e-06,
+      "loss": 0.2947,
+      "step": 690
+    },
+    {
+      "epoch": 1.063076923076923,
+      "grad_norm": 0.3390311059305257,
+      "learning_rate": 9.034860790854848e-06,
+      "loss": 0.3066,
+      "step": 691
+    },
+    {
+      "epoch": 1.0646153846153845,
+      "grad_norm": 0.3397621976049178,
+      "learning_rate": 9.010810351307593e-06,
+      "loss": 0.3136,
+      "step": 692
+    },
+    {
+      "epoch": 1.0661538461538462,
+      "grad_norm": 0.31495671802773795,
+      "learning_rate": 8.986765688627652e-06,
+      "loss": 0.2987,
+      "step": 693
+    },
+    {
+      "epoch": 1.0676923076923077,
+      "grad_norm": 0.32373621675852543,
+      "learning_rate": 8.962726943235845e-06,
+      "loss": 0.3038,
+      "step": 694
+    },
+    {
+      "epoch": 1.0692307692307692,
+      "grad_norm": 0.3213235403437696,
+      "learning_rate": 8.938694255518444e-06,
+      "loss": 0.3136,
+      "step": 695
+    },
+    {
+      "epoch": 1.0707692307692307,
+      "grad_norm": 0.3380929187622892,
+      "learning_rate": 8.91466776582634e-06,
+      "loss": 0.3149,
+      "step": 696
+    },
+    {
+      "epoch": 1.0723076923076924,
+      "grad_norm": 0.33006747280076376,
+      "learning_rate": 8.890647614474223e-06,
+      "loss": 0.3064,
+      "step": 697
+    },
+    {
+      "epoch": 1.073846153846154,
+      "grad_norm": 0.32042612599866577,
+      "learning_rate": 8.86663394173977e-06,
+      "loss": 0.2976,
+      "step": 698
+    },
+    {
+      "epoch": 1.0753846153846154,
+      "grad_norm": 0.3358613996089108,
+      "learning_rate": 8.842626887862832e-06,
+      "loss": 0.3099,
+      "step": 699
+    },
+    {
+      "epoch": 1.0769230769230769,
+      "grad_norm": 0.33773300355373426,
+      "learning_rate": 8.818626593044595e-06,
+      "loss": 0.3125,
+      "step": 700
+    },
+    {
+      "epoch": 1.0784615384615384,
+      "grad_norm": 0.3129206659809701,
+      "learning_rate": 8.79463319744677e-06,
+      "loss": 0.307,
+      "step": 701
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.3363154857551068,
+      "learning_rate": 8.77064684119079e-06,
+      "loss": 0.3055,
+      "step": 702
+    },
+    {
+      "epoch": 1.0815384615384616,
+      "grad_norm": 0.3538029669742107,
+      "learning_rate": 8.746667664356957e-06,
+      "loss": 0.3175,
+      "step": 703
+    },
+    {
+      "epoch": 1.083076923076923,
+      "grad_norm": 0.3302913356561542,
+      "learning_rate": 8.722695806983673e-06,
+      "loss": 0.3022,
+      "step": 704
+    },
+    {
+      "epoch": 1.0846153846153845,
+      "grad_norm": 0.33899865967006676,
+      "learning_rate": 8.69873140906657e-06,
+      "loss": 0.2997,
+      "step": 705
+    },
+    {
+      "epoch": 1.0861538461538462,
+      "grad_norm": 0.33108319948923426,
+      "learning_rate": 8.674774610557728e-06,
+      "loss": 0.3008,
+      "step": 706
+    },
+    {
+      "epoch": 1.0876923076923077,
+      "grad_norm": 0.3695678503981211,
+      "learning_rate": 8.650825551364844e-06,
+      "loss": 0.3073,
+      "step": 707
+    },
+    {
+      "epoch": 1.0892307692307692,
+      "grad_norm": 0.32658484700937074,
+      "learning_rate": 8.626884371350421e-06,
+      "loss": 0.3091,
+      "step": 708
+    },
+    {
+      "epoch": 1.0907692307692307,
+      "grad_norm": 0.3468408256171113,
+      "learning_rate": 8.602951210330942e-06,
+      "loss": 0.309,
+      "step": 709
+    },
+    {
+      "epoch": 1.0923076923076924,
+      "grad_norm": 0.33832777282957543,
+      "learning_rate": 8.579026208076064e-06,
+      "loss": 0.307,
+      "step": 710
+    },
+    {
+      "epoch": 1.093846153846154,
+      "grad_norm": 0.35644644047228424,
+      "learning_rate": 8.55510950430779e-06,
+      "loss": 0.3033,
+      "step": 711
+    },
+    {
+      "epoch": 1.0953846153846154,
+      "grad_norm": 0.371212082952088,
+      "learning_rate": 8.531201238699665e-06,
+      "loss": 0.2965,
+      "step": 712
+    },
+    {
+      "epoch": 1.096923076923077,
+      "grad_norm": 0.35471491772598285,
+      "learning_rate": 8.50730155087596e-06,
+      "loss": 0.2968,
+      "step": 713
+    },
+    {
+      "epoch": 1.0984615384615384,
+      "grad_norm": 0.35165840630384365,
+      "learning_rate": 8.483410580410843e-06,
+      "loss": 0.3149,
+      "step": 714
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.3582390326143127,
+      "learning_rate": 8.459528466827576e-06,
+      "loss": 0.3151,
+      "step": 715
+    },
+    {
+      "epoch": 1.1015384615384616,
+      "grad_norm": 0.3753909204006332,
+      "learning_rate": 8.43565534959769e-06,
+      "loss": 0.3048,
+      "step": 716
+    },
+    {
+      "epoch": 1.103076923076923,
+      "grad_norm": 0.3561340557866992,
+      "learning_rate": 8.411791368140197e-06,
+      "loss": 0.2993,
+      "step": 717
+    },
+    {
+      "epoch": 1.1046153846153846,
+      "grad_norm": 0.3446754836254301,
+      "learning_rate": 8.387936661820733e-06,
+      "loss": 0.3112,
+      "step": 718
+    },
+    {
+      "epoch": 1.106153846153846,
+      "grad_norm": 0.34463937200617534,
+      "learning_rate": 8.364091369950783e-06,
+      "loss": 0.3018,
+      "step": 719
+    },
+    {
+      "epoch": 1.1076923076923078,
+      "grad_norm": 0.33355995760153945,
+      "learning_rate": 8.340255631786843e-06,
+      "loss": 0.3033,
+      "step": 720
+    },
+    {
+      "epoch": 1.1092307692307692,
+      "grad_norm": 0.3334734895069594,
+      "learning_rate": 8.316429586529616e-06,
+      "loss": 0.2986,
+      "step": 721
+    },
+    {
+      "epoch": 1.1107692307692307,
+      "grad_norm": 0.3262181835683419,
+      "learning_rate": 8.292613373323203e-06,
+      "loss": 0.3048,
+      "step": 722
+    },
+    {
+      "epoch": 1.1123076923076922,
+      "grad_norm": 0.3409118579801342,
+      "learning_rate": 8.268807131254288e-06,
+      "loss": 0.2998,
+      "step": 723
+    },
+    {
+      "epoch": 1.113846153846154,
+      "grad_norm": 0.3286141191233459,
+      "learning_rate": 8.24501099935131e-06,
+      "loss": 0.3134,
+      "step": 724
+    },
+    {
+      "epoch": 1.1153846153846154,
+      "grad_norm": 0.34396314368175124,
+      "learning_rate": 8.221225116583677e-06,
+      "loss": 0.3078,
+      "step": 725
+    },
+    {
+      "epoch": 1.116923076923077,
+      "grad_norm": 0.34085254081559146,
+      "learning_rate": 8.197449621860944e-06,
+      "loss": 0.3017,
+      "step": 726
+    },
+    {
+      "epoch": 1.1184615384615384,
+      "grad_norm": 0.34648455956802243,
+      "learning_rate": 8.17368465403199e-06,
+      "loss": 0.3062,
+      "step": 727
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.3419722452119597,
+      "learning_rate": 8.14993035188422e-06,
+      "loss": 0.3086,
+      "step": 728
+    },
+    {
+      "epoch": 1.1215384615384616,
+      "grad_norm": 0.3379251683892563,
+      "learning_rate": 8.126186854142752e-06,
+      "loss": 0.3041,
+      "step": 729
+    },
+    {
+      "epoch": 1.123076923076923,
+      "grad_norm": 0.3533148220681155,
+      "learning_rate": 8.102454299469615e-06,
+      "loss": 0.3056,
+      "step": 730
+    },
+    {
+      "epoch": 1.1246153846153846,
+      "grad_norm": 0.3318022575623678,
+      "learning_rate": 8.078732826462917e-06,
+      "loss": 0.3179,
+      "step": 731
+    },
+    {
+      "epoch": 1.126153846153846,
+      "grad_norm": 0.35410783155201925,
+      "learning_rate": 8.055022573656055e-06,
+      "loss": 0.2946,
+      "step": 732
+    },
+    {
+      "epoch": 1.1276923076923078,
+      "grad_norm": 0.34539874345562344,
+      "learning_rate": 8.0313236795169e-06,
+      "loss": 0.2944,
+      "step": 733
+    },
+    {
+      "epoch": 1.1292307692307693,
+      "grad_norm": 0.31283719825398615,
+      "learning_rate": 8.007636282446986e-06,
+      "loss": 0.2941,
+      "step": 734
+    },
+    {
+      "epoch": 1.1307692307692307,
+      "grad_norm": 0.33533448624684864,
+      "learning_rate": 7.983960520780712e-06,
+      "loss": 0.3085,
+      "step": 735
+    },
+    {
+      "epoch": 1.1323076923076922,
+      "grad_norm": 0.3655899593672177,
+      "learning_rate": 7.960296532784515e-06,
+      "loss": 0.3076,
+      "step": 736
+    },
+    {
+      "epoch": 1.1338461538461537,
+      "grad_norm": 0.3359402942165652,
+      "learning_rate": 7.936644456656082e-06,
+      "loss": 0.3155,
+      "step": 737
+    },
+    {
+      "epoch": 1.1353846153846154,
+      "grad_norm": 0.39769037411914804,
+      "learning_rate": 7.913004430523526e-06,
+      "loss": 0.2981,
+      "step": 738
+    },
+    {
+      "epoch": 1.136923076923077,
+      "grad_norm": 0.36441887987288324,
+      "learning_rate": 7.889376592444605e-06,
+      "loss": 0.3078,
+      "step": 739
+    },
+    {
+      "epoch": 1.1384615384615384,
+      "grad_norm": 0.32365717196882854,
+      "learning_rate": 7.865761080405882e-06,
+      "loss": 0.3143,
+      "step": 740
+    },
+    {
+      "epoch": 1.1400000000000001,
+      "grad_norm": 0.3037037185521183,
+      "learning_rate": 7.84215803232194e-06,
+      "loss": 0.3023,
+      "step": 741
+    },
+    {
+      "epoch": 1.1415384615384616,
+      "grad_norm": 0.3465161890883791,
+      "learning_rate": 7.818567586034578e-06,
+      "loss": 0.3079,
+      "step": 742
+    },
+    {
+      "epoch": 1.143076923076923,
+      "grad_norm": 0.3417358865342288,
+      "learning_rate": 7.794989879311991e-06,
+      "loss": 0.3084,
+      "step": 743
+    },
+    {
+      "epoch": 1.1446153846153846,
+      "grad_norm": 0.32353042713035,
+      "learning_rate": 7.771425049847984e-06,
+      "loss": 0.3065,
+      "step": 744
+    },
+    {
+      "epoch": 1.146153846153846,
+      "grad_norm": 0.3409712260412258,
+      "learning_rate": 7.747873235261157e-06,
+      "loss": 0.3163,
+      "step": 745
+    },
+    {
+      "epoch": 1.1476923076923078,
+      "grad_norm": 0.3326353140393213,
+      "learning_rate": 7.724334573094101e-06,
+      "loss": 0.3105,
+      "step": 746
+    },
+    {
+      "epoch": 1.1492307692307693,
+      "grad_norm": 0.3402037220424562,
+      "learning_rate": 7.700809200812596e-06,
+      "loss": 0.3128,
+      "step": 747
+    },
+    {
+      "epoch": 1.1507692307692308,
+      "grad_norm": 0.3543906080366547,
+      "learning_rate": 7.677297255804811e-06,
+      "loss": 0.3091,
+      "step": 748
+    },
+    {
+      "epoch": 1.1523076923076923,
+      "grad_norm": 0.3175774615570522,
+      "learning_rate": 7.6537988753805e-06,
+      "loss": 0.3118,
+      "step": 749
+    },
+    {
+      "epoch": 1.1538461538461537,
+      "grad_norm": 0.34111214335775897,
+      "learning_rate": 7.6303141967702e-06,
+      "loss": 0.3164,
+      "step": 750
+    },
+    {
+      "epoch": 1.1553846153846155,
+      "grad_norm": 0.3521859791272043,
+      "learning_rate": 7.606843357124426e-06,
+      "loss": 0.3094,
+      "step": 751
+    },
+    {
+      "epoch": 1.156923076923077,
+      "grad_norm": 0.3148332782385486,
+      "learning_rate": 7.583386493512872e-06,
+      "loss": 0.3052,
+      "step": 752
+    },
+    {
+      "epoch": 1.1584615384615384,
+      "grad_norm": 0.3451026513884424,
+      "learning_rate": 7.559943742923626e-06,
+      "loss": 0.3025,
+      "step": 753
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.30001064276901357,
+      "learning_rate": 7.536515242262341e-06,
+      "loss": 0.2963,
+      "step": 754
+    },
+    {
+      "epoch": 1.1615384615384616,
+      "grad_norm": 0.3549730486566442,
+      "learning_rate": 7.513101128351454e-06,
+      "loss": 0.3069,
+      "step": 755
+    },
+    {
+      "epoch": 1.1630769230769231,
+      "grad_norm": 0.3527796783803592,
+      "learning_rate": 7.489701537929384e-06,
+      "loss": 0.3122,
+      "step": 756
+    },
+    {
+      "epoch": 1.1646153846153846,
+      "grad_norm": 0.3180649127200816,
+      "learning_rate": 7.4663166076497376e-06,
+      "loss": 0.309,
+      "step": 757
+    },
+    {
+      "epoch": 1.166153846153846,
+      "grad_norm": 0.33504286764744823,
+      "learning_rate": 7.442946474080499e-06,
+      "loss": 0.3027,
+      "step": 758
+    },
+    {
+      "epoch": 1.1676923076923078,
+      "grad_norm": 0.36644618001205087,
+      "learning_rate": 7.419591273703245e-06,
+      "loss": 0.3062,
+      "step": 759
+    },
+    {
+      "epoch": 1.1692307692307693,
+      "grad_norm": 0.3196909647006109,
+      "learning_rate": 7.396251142912337e-06,
+      "loss": 0.3091,
+      "step": 760
+    },
+    {
+      "epoch": 1.1707692307692308,
+      "grad_norm": 0.3283684576636401,
+      "learning_rate": 7.372926218014131e-06,
+      "loss": 0.2995,
+      "step": 761
+    },
+    {
+      "epoch": 1.1723076923076923,
+      "grad_norm": 0.3236434186585492,
+      "learning_rate": 7.349616635226191e-06,
+      "loss": 0.2994,
+      "step": 762
+    },
+    {
+      "epoch": 1.1738461538461538,
+      "grad_norm": 0.3251120275356546,
+      "learning_rate": 7.326322530676471e-06,
+      "loss": 0.3118,
+      "step": 763
+    },
+    {
+      "epoch": 1.1753846153846155,
+      "grad_norm": 0.34202361396071007,
+      "learning_rate": 7.303044040402536e-06,
+      "loss": 0.3113,
+      "step": 764
+    },
+    {
+      "epoch": 1.176923076923077,
+      "grad_norm": 0.28971407329970267,
+      "learning_rate": 7.279781300350758e-06,
+      "loss": 0.2978,
+      "step": 765
+    },
+    {
+      "epoch": 1.1784615384615384,
+      "grad_norm": 0.3263340332124883,
+      "learning_rate": 7.256534446375543e-06,
+      "loss": 0.3021,
+      "step": 766
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.3362810748983171,
+      "learning_rate": 7.23330361423851e-06,
+      "loss": 0.3126,
+      "step": 767
+    },
+    {
+      "epoch": 1.1815384615384614,
+      "grad_norm": 0.33000561089633407,
+      "learning_rate": 7.210088939607709e-06,
+      "loss": 0.3076,
+      "step": 768
+    },
+    {
+      "epoch": 1.1830769230769231,
+      "grad_norm": 0.31584659613512245,
+      "learning_rate": 7.186890558056836e-06,
+      "loss": 0.3015,
+      "step": 769
+    },
+    {
+      "epoch": 1.1846153846153846,
+      "grad_norm": 0.3230244494845988,
+      "learning_rate": 7.163708605064437e-06,
+      "loss": 0.3135,
+      "step": 770
+    },
+    {
+      "epoch": 1.1861538461538461,
+      "grad_norm": 0.3227128520541019,
+      "learning_rate": 7.140543216013109e-06,
+      "loss": 0.3015,
+      "step": 771
+    },
+    {
+      "epoch": 1.1876923076923076,
+      "grad_norm": 0.3398838111565539,
+      "learning_rate": 7.117394526188719e-06,
+      "loss": 0.299,
+      "step": 772
+    },
+    {
+      "epoch": 1.1892307692307693,
+      "grad_norm": 0.3191494391287238,
+      "learning_rate": 7.094262670779611e-06,
+      "loss": 0.31,
+      "step": 773
+    },
+    {
+      "epoch": 1.1907692307692308,
+      "grad_norm": 0.3163509565668907,
+      "learning_rate": 7.071147784875809e-06,
+      "loss": 0.3058,
+      "step": 774
+    },
+    {
+      "epoch": 1.1923076923076923,
+      "grad_norm": 0.31683743515779933,
+      "learning_rate": 7.048050003468252e-06,
+      "loss": 0.3018,
+      "step": 775
+    },
+    {
+      "epoch": 1.1938461538461538,
+      "grad_norm": 0.3324792580728335,
+      "learning_rate": 7.024969461447973e-06,
+      "loss": 0.3168,
+      "step": 776
+    },
+    {
+      "epoch": 1.1953846153846155,
+      "grad_norm": 0.3134978102972266,
+      "learning_rate": 7.001906293605329e-06,
+      "loss": 0.2928,
+      "step": 777
+    },
+    {
+      "epoch": 1.196923076923077,
+      "grad_norm": 0.33084215934405464,
+      "learning_rate": 6.978860634629213e-06,
+      "loss": 0.2834,
+      "step": 778
+    },
+    {
+      "epoch": 1.1984615384615385,
+      "grad_norm": 0.3247253681197987,
+      "learning_rate": 6.9558326191062775e-06,
+      "loss": 0.2945,
+      "step": 779
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.3306577049343056,
+      "learning_rate": 6.932822381520121e-06,
+      "loss": 0.2915,
+      "step": 780
+    },
+    {
+      "epoch": 1.2015384615384614,
+      "grad_norm": 0.3407912418079179,
+      "learning_rate": 6.909830056250527e-06,
+      "loss": 0.3068,
+      "step": 781
+    },
+    {
+      "epoch": 1.2030769230769232,
+      "grad_norm": 0.32548273752327034,
+      "learning_rate": 6.88685577757267e-06,
+      "loss": 0.3017,
+      "step": 782
+    },
+    {
+      "epoch": 1.2046153846153846,
+      "grad_norm": 0.335519503395646,
+      "learning_rate": 6.8638996796563275e-06,
+      "loss": 0.3124,
+      "step": 783
+    },
+    {
+      "epoch": 1.2061538461538461,
+      "grad_norm": 0.3501864142437491,
+      "learning_rate": 6.8409618965651125e-06,
+      "loss": 0.2976,
+      "step": 784
+    },
+    {
+      "epoch": 1.2076923076923076,
+      "grad_norm": 0.31101023902365416,
+      "learning_rate": 6.81804256225567e-06,
+      "loss": 0.3112,
+      "step": 785
+    },
+    {
+      "epoch": 1.209230769230769,
+      "grad_norm": 0.36347305328084234,
+      "learning_rate": 6.795141810576906e-06,
+      "loss": 0.3062,
+      "step": 786
+    },
+    {
+      "epoch": 1.2107692307692308,
+      "grad_norm": 0.34994293873835053,
+      "learning_rate": 6.772259775269203e-06,
+      "loss": 0.3002,
+      "step": 787
+    },
+    {
+      "epoch": 1.2123076923076923,
+      "grad_norm": 0.33164195400888197,
+      "learning_rate": 6.7493965899636486e-06,
+      "loss": 0.3027,
+      "step": 788
+    },
+    {
+      "epoch": 1.2138461538461538,
+      "grad_norm": 0.3074524045746076,
+      "learning_rate": 6.7265523881812335e-06,
+      "loss": 0.3034,
+      "step": 789
+    },
+    {
+      "epoch": 1.2153846153846155,
+      "grad_norm": 0.34845264769510725,
+      "learning_rate": 6.703727303332094e-06,
+      "loss": 0.3015,
+      "step": 790
+    },
+    {
+      "epoch": 1.216923076923077,
+      "grad_norm": 0.31401391228845343,
+      "learning_rate": 6.680921468714718e-06,
+      "loss": 0.2971,
+      "step": 791
+    },
+    {
+      "epoch": 1.2184615384615385,
+      "grad_norm": 0.3344738184510328,
+      "learning_rate": 6.6581350175151715e-06,
+      "loss": 0.3035,
+      "step": 792
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.3223383527746208,
+      "learning_rate": 6.6353680828063306e-06,
+      "loss": 0.2975,
+      "step": 793
+    },
+    {
+      "epoch": 1.2215384615384615,
+      "grad_norm": 0.3084349374550551,
+      "learning_rate": 6.612620797547087e-06,
+      "loss": 0.2895,
+      "step": 794
+    },
+    {
+      "epoch": 1.2230769230769232,
+      "grad_norm": 0.3234334066539162,
+      "learning_rate": 6.589893294581579e-06,
+      "loss": 0.3057,
+      "step": 795
+    },
+    {
+      "epoch": 1.2246153846153847,
+      "grad_norm": 0.3140355556165583,
+      "learning_rate": 6.567185706638417e-06,
+      "loss": 0.3031,
+      "step": 796
+    },
+    {
+      "epoch": 1.2261538461538461,
+      "grad_norm": 0.33153851295758796,
+      "learning_rate": 6.5444981663299135e-06,
+      "loss": 0.2932,
+      "step": 797
+    },
+    {
+      "epoch": 1.2276923076923076,
+      "grad_norm": 0.30377258235888455,
+      "learning_rate": 6.521830806151297e-06,
+      "loss": 0.3154,
+      "step": 798
+    },
+    {
+      "epoch": 1.2292307692307691,
+      "grad_norm": 0.3241148494008968,
+      "learning_rate": 6.499183758479944e-06,
+      "loss": 0.3114,
+      "step": 799
+    },
+    {
+      "epoch": 1.2307692307692308,
+      "grad_norm": 0.3136486572122067,
+      "learning_rate": 6.47655715557461e-06,
+      "loss": 0.3043,
+      "step": 800
+    },
+    {
+      "epoch": 1.2323076923076923,
+      "grad_norm": 0.3222516306791619,
+      "learning_rate": 6.453951129574644e-06,
+      "loss": 0.3031,
+      "step": 801
+    },
+    {
+      "epoch": 1.2338461538461538,
+      "grad_norm": 0.32336239179170656,
+      "learning_rate": 6.431365812499242e-06,
+      "loss": 0.319,
+      "step": 802
+    },
+    {
+      "epoch": 1.2353846153846153,
+      "grad_norm": 0.3302606676399345,
+      "learning_rate": 6.408801336246645e-06,
+      "loss": 0.3096,
+      "step": 803
+    },
+    {
+      "epoch": 1.236923076923077,
+      "grad_norm": 0.3063288390807676,
+      "learning_rate": 6.38625783259339e-06,
+      "loss": 0.3081,
+      "step": 804
+    },
+    {
+      "epoch": 1.2384615384615385,
+      "grad_norm": 0.3518190513124168,
+      "learning_rate": 6.36373543319353e-06,
+      "loss": 0.3046,
+      "step": 805
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.32946440261639925,
+      "learning_rate": 6.341234269577878e-06,
+      "loss": 0.2908,
+      "step": 806
+    },
+    {
+      "epoch": 1.2415384615384615,
+      "grad_norm": 0.3098657958913532,
+      "learning_rate": 6.318754473153221e-06,
+      "loss": 0.306,
+      "step": 807
+    },
+    {
+      "epoch": 1.2430769230769232,
+      "grad_norm": 0.34957369863007254,
+      "learning_rate": 6.296296175201565e-06,
+      "loss": 0.302,
+      "step": 808
+    },
+    {
+      "epoch": 1.2446153846153847,
+      "grad_norm": 0.3522356245737633,
+      "learning_rate": 6.273859506879365e-06,
+      "loss": 0.2948,
+      "step": 809
+    },
+    {
+      "epoch": 1.2461538461538462,
+      "grad_norm": 0.35700241527702703,
+      "learning_rate": 6.251444599216756e-06,
+      "loss": 0.2989,
+      "step": 810
+    },
+    {
+      "epoch": 1.2476923076923077,
+      "grad_norm": 0.32690336198700204,
+      "learning_rate": 6.229051583116796e-06,
+      "loss": 0.3109,
+      "step": 811
+    },
+    {
+      "epoch": 1.2492307692307691,
+      "grad_norm": 0.35460320797864253,
+      "learning_rate": 6.206680589354696e-06,
+      "loss": 0.2975,
+      "step": 812
+    },
+    {
+      "epoch": 1.2507692307692309,
+      "grad_norm": 0.31587492917639964,
+      "learning_rate": 6.18433174857705e-06,
+      "loss": 0.3124,
+      "step": 813
+    },
+    {
+      "epoch": 1.2523076923076923,
+      "grad_norm": 0.33794597274977345,
+      "learning_rate": 6.162005191301082e-06,
+      "loss": 0.302,
+      "step": 814
+    },
+    {
+      "epoch": 1.2538461538461538,
+      "grad_norm": 0.31444012451990255,
+      "learning_rate": 6.139701047913885e-06,
+      "loss": 0.2873,
+      "step": 815
+    },
+    {
+      "epoch": 1.2553846153846153,
+      "grad_norm": 0.33926177643729094,
+      "learning_rate": 6.117419448671651e-06,
+      "loss": 0.3152,
+      "step": 816
+    },
+    {
+      "epoch": 1.2569230769230768,
+      "grad_norm": 0.3532188842185072,
+      "learning_rate": 6.095160523698913e-06,
+      "loss": 0.3031,
+      "step": 817
+    },
+    {
+      "epoch": 1.2584615384615385,
+      "grad_norm": 0.349796554156469,
+      "learning_rate": 6.072924402987785e-06,
+      "loss": 0.2972,
+      "step": 818
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.3257039049930201,
+      "learning_rate": 6.050711216397212e-06,
+      "loss": 0.3002,
+      "step": 819
+    },
+    {
+      "epoch": 1.2615384615384615,
+      "grad_norm": 0.3264391455544534,
+      "learning_rate": 6.028521093652195e-06,
+      "loss": 0.3175,
+      "step": 820
+    },
+    {
+      "epoch": 1.2630769230769232,
+      "grad_norm": 0.38077886792406285,
+      "learning_rate": 6.006354164343047e-06,
+      "loss": 0.2943,
+      "step": 821
+    },
+    {
+      "epoch": 1.2646153846153847,
+      "grad_norm": 0.325790124722129,
+      "learning_rate": 5.984210557924628e-06,
+      "loss": 0.3043,
+      "step": 822
+    },
+    {
+      "epoch": 1.2661538461538462,
+      "grad_norm": 0.3216794342475427,
+      "learning_rate": 5.962090403715592e-06,
+      "loss": 0.3121,
+      "step": 823
+    },
+    {
+      "epoch": 1.2676923076923077,
+      "grad_norm": 0.3250466188941678,
+      "learning_rate": 5.939993830897641e-06,
+      "loss": 0.3031,
+      "step": 824
+    },
+    {
+      "epoch": 1.2692307692307692,
+      "grad_norm": 0.3544067370440847,
+      "learning_rate": 5.9179209685147525e-06,
+      "loss": 0.3029,
+      "step": 825
+    },
+    {
+      "epoch": 1.2707692307692309,
+      "grad_norm": 0.3210547589604769,
+      "learning_rate": 5.895871945472434e-06,
+      "loss": 0.3083,
+      "step": 826
+    },
+    {
+      "epoch": 1.2723076923076924,
+      "grad_norm": 0.33230024230404,
+      "learning_rate": 5.873846890536976e-06,
+      "loss": 0.3054,
+      "step": 827
+    },
+    {
+      "epoch": 1.2738461538461539,
+      "grad_norm": 0.32864274399605803,
+      "learning_rate": 5.851845932334698e-06,
+      "loss": 0.3093,
+      "step": 828
+    },
+    {
+      "epoch": 1.2753846153846153,
+      "grad_norm": 0.3071256746554175,
+      "learning_rate": 5.829869199351188e-06,
+      "loss": 0.3103,
+      "step": 829
+    },
+    {
+      "epoch": 1.2769230769230768,
+      "grad_norm": 0.32938855939931677,
+      "learning_rate": 5.807916819930563e-06,
+      "loss": 0.3044,
+      "step": 830
+    },
+    {
+      "epoch": 1.2784615384615385,
+      "grad_norm": 0.33716840349798904,
+      "learning_rate": 5.785988922274711e-06,
+      "loss": 0.3001,
+      "step": 831
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.30659755101668507,
+      "learning_rate": 5.7640856344425465e-06,
+      "loss": 0.2979,
+      "step": 832
+    },
+    {
+      "epoch": 1.2815384615384615,
+      "grad_norm": 0.31745928277952257,
+      "learning_rate": 5.742207084349274e-06,
+      "loss": 0.3066,
+      "step": 833
+    },
+    {
+      "epoch": 1.283076923076923,
+      "grad_norm": 0.33789016928259225,
+      "learning_rate": 5.720353399765615e-06,
+      "loss": 0.3093,
+      "step": 834
+    },
+    {
+      "epoch": 1.2846153846153845,
+      "grad_norm": 0.33796655405326004,
+      "learning_rate": 5.698524708317082e-06,
+      "loss": 0.3121,
+      "step": 835
+    },
+    {
+      "epoch": 1.2861538461538462,
+      "grad_norm": 0.31041992405960966,
+      "learning_rate": 5.676721137483226e-06,
+      "loss": 0.3022,
+      "step": 836
+    },
+    {
+      "epoch": 1.2876923076923077,
+      "grad_norm": 0.315518487863336,
+      "learning_rate": 5.654942814596902e-06,
+      "loss": 0.3198,
+      "step": 837
+    },
+    {
+      "epoch": 1.2892307692307692,
+      "grad_norm": 0.33934717586798074,
+      "learning_rate": 5.633189866843507e-06,
+      "loss": 0.3115,
+      "step": 838
+    },
+    {
+      "epoch": 1.290769230769231,
+      "grad_norm": 0.31675740029525207,
+      "learning_rate": 5.611462421260251e-06,
+      "loss": 0.3085,
+      "step": 839
+    },
+    {
+      "epoch": 1.2923076923076924,
+      "grad_norm": 0.29636621542877845,
+      "learning_rate": 5.58976060473541e-06,
+      "loss": 0.3108,
+      "step": 840
+    },
+    {
+      "epoch": 1.2938461538461539,
+      "grad_norm": 0.30523439903853045,
+      "learning_rate": 5.5680845440075885e-06,
+      "loss": 0.3151,
+      "step": 841
+    },
+    {
+      "epoch": 1.2953846153846154,
+      "grad_norm": 0.33692392468673477,
+      "learning_rate": 5.546434365664974e-06,
+      "loss": 0.2905,
+      "step": 842
+    },
+    {
+      "epoch": 1.2969230769230768,
+      "grad_norm": 0.33547664436771796,
+      "learning_rate": 5.5248101961446065e-06,
+      "loss": 0.3054,
+      "step": 843
+    },
+    {
+      "epoch": 1.2984615384615386,
+      "grad_norm": 0.32230952967894777,
+      "learning_rate": 5.503212161731628e-06,
+      "loss": 0.3079,
+      "step": 844
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.32373697652558836,
+      "learning_rate": 5.481640388558551e-06,
+      "loss": 0.3137,
+      "step": 845
+    },
+    {
+      "epoch": 1.3015384615384615,
+      "grad_norm": 0.3513367614485366,
+      "learning_rate": 5.460095002604533e-06,
+      "loss": 0.2962,
+      "step": 846
+    },
+    {
+      "epoch": 1.303076923076923,
+      "grad_norm": 0.3114599259136889,
+      "learning_rate": 5.43857612969462e-06,
+      "loss": 0.3068,
+      "step": 847
+    },
+    {
+      "epoch": 1.3046153846153845,
+      "grad_norm": 0.3177326490401273,
+      "learning_rate": 5.417083895499024e-06,
+      "loss": 0.3092,
+      "step": 848
+    },
+    {
+      "epoch": 1.3061538461538462,
+      "grad_norm": 0.3401565621302886,
+      "learning_rate": 5.39561842553239e-06,
+      "loss": 0.2962,
+      "step": 849
+    },
+    {
+      "epoch": 1.3076923076923077,
+      "grad_norm": 0.3287447592695306,
+      "learning_rate": 5.374179845153048e-06,
+      "loss": 0.2953,
+      "step": 850
+    },
+    {
+      "epoch": 1.3092307692307692,
+      "grad_norm": 0.33635860028856784,
+      "learning_rate": 5.352768279562315e-06,
+      "loss": 0.3023,
+      "step": 851
+    },
+    {
+      "epoch": 1.3107692307692307,
+      "grad_norm": 0.35512549313238523,
+      "learning_rate": 5.331383853803724e-06,
+      "loss": 0.3015,
+      "step": 852
+    },
+    {
+      "epoch": 1.3123076923076922,
+      "grad_norm": 0.3210988691917933,
+      "learning_rate": 5.310026692762316e-06,
+      "loss": 0.2894,
+      "step": 853
+    },
+    {
+      "epoch": 1.3138461538461539,
+      "grad_norm": 0.33747355116226485,
+      "learning_rate": 5.288696921163902e-06,
+      "loss": 0.2985,
+      "step": 854
+    },
+    {
+      "epoch": 1.3153846153846154,
+      "grad_norm": 0.34559539244281745,
+      "learning_rate": 5.267394663574351e-06,
+      "loss": 0.2835,
+      "step": 855
+    },
+    {
+      "epoch": 1.3169230769230769,
+      "grad_norm": 0.3123905203959201,
+      "learning_rate": 5.246120044398839e-06,
+      "loss": 0.3184,
+      "step": 856
+    },
+    {
+      "epoch": 1.3184615384615386,
+      "grad_norm": 0.31089902301048383,
+      "learning_rate": 5.224873187881136e-06,
+      "loss": 0.2972,
+      "step": 857
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.31463633217223824,
+      "learning_rate": 5.20365421810288e-06,
+      "loss": 0.3135,
+      "step": 858
+    },
+    {
+      "epoch": 1.3215384615384616,
+      "grad_norm": 0.310104652798278,
+      "learning_rate": 5.1824632589828465e-06,
+      "loss": 0.2895,
+      "step": 859
+    },
+    {
+      "epoch": 1.323076923076923,
+      "grad_norm": 0.3304042493468375,
+      "learning_rate": 5.161300434276237e-06,
+      "loss": 0.294,
+      "step": 860
+    },
+    {
+      "epoch": 1.3246153846153845,
+      "grad_norm": 0.3377043914563407,
+      "learning_rate": 5.14016586757394e-06,
+      "loss": 0.2999,
+      "step": 861
+    },
+    {
+      "epoch": 1.3261538461538462,
+      "grad_norm": 0.3128306146054052,
+      "learning_rate": 5.119059682301819e-06,
+      "loss": 0.3042,
+      "step": 862
+    },
+    {
+      "epoch": 1.3276923076923077,
+      "grad_norm": 0.3132704388548915,
+      "learning_rate": 5.097982001719994e-06,
+      "loss": 0.3037,
+      "step": 863
+    },
+    {
+      "epoch": 1.3292307692307692,
+      "grad_norm": 0.3276050165378912,
+      "learning_rate": 5.076932948922111e-06,
+      "loss": 0.3048,
+      "step": 864
+    },
+    {
+      "epoch": 1.3307692307692307,
+      "grad_norm": 0.3240308081497042,
+      "learning_rate": 5.0559126468346354e-06,
+      "loss": 0.3033,
+      "step": 865
+    },
+    {
+      "epoch": 1.3323076923076922,
+      "grad_norm": 0.31159878743700903,
+      "learning_rate": 5.034921218216126e-06,
+      "loss": 0.3002,
+      "step": 866
+    },
+    {
+      "epoch": 1.333846153846154,
+      "grad_norm": 0.32365175997561196,
+      "learning_rate": 5.013958785656516e-06,
+      "loss": 0.2928,
+      "step": 867
+    },
+    {
+      "epoch": 1.3353846153846154,
+      "grad_norm": 0.342989716874584,
+      "learning_rate": 4.993025471576417e-06,
+      "loss": 0.2956,
+      "step": 868
+    },
+    {
+      "epoch": 1.3369230769230769,
+      "grad_norm": 0.3174392737024924,
+      "learning_rate": 4.972121398226371e-06,
+      "loss": 0.3093,
+      "step": 869
+    },
+    {
+      "epoch": 1.3384615384615386,
+      "grad_norm": 0.30616032362229467,
+      "learning_rate": 4.951246687686164e-06,
+      "loss": 0.3085,
+      "step": 870
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.323097177879886,
+      "learning_rate": 4.930401461864099e-06,
+      "loss": 0.3067,
+      "step": 871
+    },
+    {
+      "epoch": 1.3415384615384616,
+      "grad_norm": 0.3202646930421606,
+      "learning_rate": 4.909585842496287e-06,
+      "loss": 0.3128,
+      "step": 872
+    },
+    {
+      "epoch": 1.343076923076923,
+      "grad_norm": 0.320354071438146,
+      "learning_rate": 4.888799951145948e-06,
+      "loss": 0.2991,
+      "step": 873
+    },
+    {
+      "epoch": 1.3446153846153845,
+      "grad_norm": 0.3250229035153826,
+      "learning_rate": 4.868043909202678e-06,
+      "loss": 0.3002,
+      "step": 874
+    },
+    {
+      "epoch": 1.3461538461538463,
+      "grad_norm": 0.30651660783074586,
+      "learning_rate": 4.847317837881757e-06,
+      "loss": 0.3028,
+      "step": 875
+    },
+    {
+      "epoch": 1.3476923076923077,
+      "grad_norm": 0.31667644641307485,
+      "learning_rate": 4.826621858223431e-06,
+      "loss": 0.3005,
+      "step": 876
+    },
+    {
+      "epoch": 1.3492307692307692,
+      "grad_norm": 0.3155312200914733,
+      "learning_rate": 4.805956091092228e-06,
+      "loss": 0.3065,
+      "step": 877
+    },
+    {
+      "epoch": 1.3507692307692307,
+      "grad_norm": 0.320601047307546,
+      "learning_rate": 4.785320657176216e-06,
+      "loss": 0.2954,
+      "step": 878
+    },
+    {
+      "epoch": 1.3523076923076922,
+      "grad_norm": 0.3418589068027817,
+      "learning_rate": 4.764715676986327e-06,
+      "loss": 0.3038,
+      "step": 879
+    },
+    {
+      "epoch": 1.353846153846154,
+      "grad_norm": 0.2974571723314658,
+      "learning_rate": 4.744141270855638e-06,
+      "loss": 0.2916,
+      "step": 880
+    },
+    {
+      "epoch": 1.3553846153846154,
+      "grad_norm": 0.2936378627037304,
+      "learning_rate": 4.7235975589386715e-06,
+      "loss": 0.3093,
+      "step": 881
+    },
+    {
+      "epoch": 1.356923076923077,
+      "grad_norm": 0.30718538802190726,
+      "learning_rate": 4.7030846612107105e-06,
+      "loss": 0.2897,
+      "step": 882
+    },
+    {
+      "epoch": 1.3584615384615384,
+      "grad_norm": 0.31964002160039096,
+      "learning_rate": 4.6826026974670665e-06,
+      "loss": 0.3097,
+      "step": 883
+    },
+    {
+      "epoch": 1.3599999999999999,
+      "grad_norm": 0.32434002264102607,
+      "learning_rate": 4.662151787322405e-06,
+      "loss": 0.3191,
+      "step": 884
+    },
+    {
+      "epoch": 1.3615384615384616,
+      "grad_norm": 0.3068250385914274,
+      "learning_rate": 4.641732050210032e-06,
+      "loss": 0.2988,
+      "step": 885
+    },
+    {
+      "epoch": 1.363076923076923,
+      "grad_norm": 0.29635729297375196,
+      "learning_rate": 4.621343605381215e-06,
+      "loss": 0.2955,
+      "step": 886
+    },
+    {
+      "epoch": 1.3646153846153846,
+      "grad_norm": 0.3135843534004588,
+      "learning_rate": 4.600986571904461e-06,
+      "loss": 0.2948,
+      "step": 887
+    },
+    {
+      "epoch": 1.3661538461538463,
+      "grad_norm": 0.29508851633969135,
+      "learning_rate": 4.580661068664844e-06,
+      "loss": 0.3016,
+      "step": 888
+    },
+    {
+      "epoch": 1.3676923076923078,
+      "grad_norm": 0.28653782904697656,
+      "learning_rate": 4.560367214363295e-06,
+      "loss": 0.3077,
+      "step": 889
+    },
+    {
+      "epoch": 1.3692307692307693,
+      "grad_norm": 0.29549152334783624,
+      "learning_rate": 4.540105127515921e-06,
+      "loss": 0.3146,
+      "step": 890
+    },
+    {
+      "epoch": 1.3707692307692307,
+      "grad_norm": 0.3114737881677389,
+      "learning_rate": 4.519874926453303e-06,
+      "loss": 0.2969,
+      "step": 891
+    },
+    {
+      "epoch": 1.3723076923076922,
+      "grad_norm": 0.3009306224087917,
+      "learning_rate": 4.499676729319809e-06,
+      "loss": 0.3007,
+      "step": 892
+    },
+    {
+      "epoch": 1.373846153846154,
+      "grad_norm": 0.30386215795990473,
+      "learning_rate": 4.479510654072909e-06,
+      "loss": 0.3037,
+      "step": 893
+    },
+    {
+      "epoch": 1.3753846153846154,
+      "grad_norm": 0.3237456994129661,
+      "learning_rate": 4.459376818482471e-06,
+      "loss": 0.2896,
+      "step": 894
+    },
+    {
+      "epoch": 1.376923076923077,
+      "grad_norm": 0.31574782281108643,
+      "learning_rate": 4.439275340130099e-06,
+      "loss": 0.2901,
+      "step": 895
+    },
+    {
+      "epoch": 1.3784615384615384,
+      "grad_norm": 0.30206776006851976,
+      "learning_rate": 4.419206336408418e-06,
+      "loss": 0.3006,
+      "step": 896
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.32162754120633813,
+      "learning_rate": 4.399169924520403e-06,
+      "loss": 0.3071,
+      "step": 897
+    },
+    {
+      "epoch": 1.3815384615384616,
+      "grad_norm": 0.3310811905991917,
+      "learning_rate": 4.379166221478697e-06,
+      "loss": 0.2969,
+      "step": 898
+    },
+    {
+      "epoch": 1.383076923076923,
+      "grad_norm": 0.31107496181714434,
+      "learning_rate": 4.359195344104916e-06,
+      "loss": 0.2904,
+      "step": 899
+    },
+    {
+      "epoch": 1.3846153846153846,
+      "grad_norm": 0.32263015481078,
+      "learning_rate": 4.339257409028987e-06,
+      "loss": 0.3031,
+      "step": 900
+    },
+    {
+      "epoch": 1.3861538461538463,
+      "grad_norm": 0.2974317077013365,
+      "learning_rate": 4.319352532688444e-06,
+      "loss": 0.3122,
+      "step": 901
+    },
+    {
+      "epoch": 1.3876923076923076,
+      "grad_norm": 0.33162802201526975,
+      "learning_rate": 4.2994808313277565e-06,
+      "loss": 0.3043,
+      "step": 902
+    },
+    {
+      "epoch": 1.3892307692307693,
+      "grad_norm": 0.33023266137768315,
+      "learning_rate": 4.279642420997655e-06,
+      "loss": 0.299,
+      "step": 903
+    },
+    {
+      "epoch": 1.3907692307692308,
+      "grad_norm": 0.33598474231424635,
+      "learning_rate": 4.259837417554457e-06,
+      "loss": 0.297,
+      "step": 904
+    },
+    {
+      "epoch": 1.3923076923076922,
+      "grad_norm": 0.29751491756642273,
+      "learning_rate": 4.240065936659374e-06,
+      "loss": 0.3107,
+      "step": 905
+    },
+    {
+      "epoch": 1.393846153846154,
+      "grad_norm": 0.34236456688614075,
+      "learning_rate": 4.220328093777851e-06,
+      "loss": 0.2952,
+      "step": 906
+    },
+    {
+      "epoch": 1.3953846153846154,
+      "grad_norm": 0.3263775617915975,
+      "learning_rate": 4.200624004178883e-06,
+      "loss": 0.3065,
+      "step": 907
+    },
+    {
+      "epoch": 1.396923076923077,
+      "grad_norm": 0.30186389177288103,
+      "learning_rate": 4.180953782934352e-06,
+      "loss": 0.3034,
+      "step": 908
+    },
+    {
+      "epoch": 1.3984615384615384,
+      "grad_norm": 0.30662047730658826,
+      "learning_rate": 4.161317544918345e-06,
+      "loss": 0.2956,
+      "step": 909
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.3161645973913341,
+      "learning_rate": 4.141715404806486e-06,
+      "loss": 0.2938,
+      "step": 910
+    },
+    {
+      "epoch": 1.4015384615384616,
+      "grad_norm": 0.31958608558099133,
+      "learning_rate": 4.12214747707527e-06,
+      "loss": 0.3025,
+      "step": 911
+    },
+    {
+      "epoch": 1.403076923076923,
+      "grad_norm": 0.3046866286301989,
+      "learning_rate": 4.1026138760013886e-06,
+      "loss": 0.3021,
+      "step": 912
+    },
+    {
+      "epoch": 1.4046153846153846,
+      "grad_norm": 0.301854494582748,
+      "learning_rate": 4.083114715661069e-06,
+      "loss": 0.3083,
+      "step": 913
+    },
+    {
+      "epoch": 1.406153846153846,
+      "grad_norm": 0.317089983486272,
+      "learning_rate": 4.0636501099294e-06,
+      "loss": 0.3036,
+      "step": 914
+    },
+    {
+      "epoch": 1.4076923076923076,
+      "grad_norm": 0.3256436150728904,
+      "learning_rate": 4.044220172479675e-06,
+      "loss": 0.302,
+      "step": 915
+    },
+    {
+      "epoch": 1.4092307692307693,
+      "grad_norm": 0.32398564113427775,
+      "learning_rate": 4.024825016782727e-06,
+      "loss": 0.3031,
+      "step": 916
+    },
+    {
+      "epoch": 1.4107692307692308,
+      "grad_norm": 0.31954413117671754,
+      "learning_rate": 4.0054647561062625e-06,
+      "loss": 0.2891,
+      "step": 917
+    },
+    {
+      "epoch": 1.4123076923076923,
+      "grad_norm": 0.3154553099120394,
+      "learning_rate": 3.9861395035141936e-06,
+      "loss": 0.3106,
+      "step": 918
+    },
+    {
+      "epoch": 1.413846153846154,
+      "grad_norm": 0.31073749875407,
+      "learning_rate": 3.9668493718659924e-06,
+      "loss": 0.2849,
+      "step": 919
+    },
+    {
+      "epoch": 1.4153846153846155,
+      "grad_norm": 0.32829577186678083,
+      "learning_rate": 3.947594473816026e-06,
+      "loss": 0.3063,
+      "step": 920
+    },
+    {
+      "epoch": 1.416923076923077,
+      "grad_norm": 0.3047664322705856,
+      "learning_rate": 3.9283749218128885e-06,
+      "loss": 0.2985,
+      "step": 921
+    },
+    {
+      "epoch": 1.4184615384615384,
+      "grad_norm": 0.3090276583488756,
+      "learning_rate": 3.909190828098766e-06,
+      "loss": 0.3047,
+      "step": 922
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.3129680154216449,
+      "learning_rate": 3.890042304708758e-06,
+      "loss": 0.3012,
+      "step": 923
+    },
+    {
+      "epoch": 1.4215384615384616,
+      "grad_norm": 0.32748964356418125,
+      "learning_rate": 3.8709294634702374e-06,
+      "loss": 0.2957,
+      "step": 924
+    },
+    {
+      "epoch": 1.4230769230769231,
+      "grad_norm": 0.31474454671188556,
+      "learning_rate": 3.8518524160021876e-06,
+      "loss": 0.2965,
+      "step": 925
+    },
+    {
+      "epoch": 1.4246153846153846,
+      "grad_norm": 0.3232749514789089,
+      "learning_rate": 3.832811273714569e-06,
+      "loss": 0.3026,
+      "step": 926
+    },
+    {
+      "epoch": 1.426153846153846,
+      "grad_norm": 0.31419583272769863,
+      "learning_rate": 3.813806147807645e-06,
+      "loss": 0.2955,
+      "step": 927
+    },
+    {
+      "epoch": 1.4276923076923076,
+      "grad_norm": 0.4962966308900197,
+      "learning_rate": 3.7948371492713454e-06,
+      "loss": 0.291,
+      "step": 928
+    },
+    {
+      "epoch": 1.4292307692307693,
+      "grad_norm": 0.3008550879462397,
+      "learning_rate": 3.775904388884618e-06,
+      "loss": 0.2975,
+      "step": 929
+    },
+    {
+      "epoch": 1.4307692307692308,
+      "grad_norm": 0.3082766000776801,
+      "learning_rate": 3.7570079772147748e-06,
+      "loss": 0.2912,
+      "step": 930
+    },
+    {
+      "epoch": 1.4323076923076923,
+      "grad_norm": 0.33299155606740327,
+      "learning_rate": 3.738148024616863e-06,
+      "loss": 0.3125,
+      "step": 931
+    },
+    {
+      "epoch": 1.4338461538461538,
+      "grad_norm": 0.30799445597383024,
+      "learning_rate": 3.7193246412329976e-06,
+      "loss": 0.3055,
+      "step": 932
+    },
+    {
+      "epoch": 1.4353846153846153,
+      "grad_norm": 0.31389916774676485,
+      "learning_rate": 3.700537936991733e-06,
+      "loss": 0.287,
+      "step": 933
+    },
+    {
+      "epoch": 1.436923076923077,
+      "grad_norm": 0.29782679605700335,
+      "learning_rate": 3.681788021607413e-06,
+      "loss": 0.3055,
+      "step": 934
+    },
+    {
+      "epoch": 1.4384615384615385,
+      "grad_norm": 0.3354524364783457,
+      "learning_rate": 3.6630750045795472e-06,
+      "loss": 0.3033,
+      "step": 935
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.30086703100503803,
+      "learning_rate": 3.6443989951921478e-06,
+      "loss": 0.2925,
+      "step": 936
+    },
+    {
+      "epoch": 1.4415384615384617,
+      "grad_norm": 0.30453041571413564,
+      "learning_rate": 3.625760102513103e-06,
+      "loss": 0.2972,
+      "step": 937
+    },
+    {
+      "epoch": 1.4430769230769231,
+      "grad_norm": 0.32054064800637316,
+      "learning_rate": 3.607158435393544e-06,
+      "loss": 0.3012,
+      "step": 938
+    },
+    {
+      "epoch": 1.4446153846153846,
+      "grad_norm": 0.3353206405738868,
+      "learning_rate": 3.5885941024672e-06,
+      "loss": 0.3071,
+      "step": 939
+    },
+    {
+      "epoch": 1.4461538461538461,
+      "grad_norm": 0.31848309742815456,
+      "learning_rate": 3.5700672121497728e-06,
+      "loss": 0.2923,
+      "step": 940
+    },
+    {
+      "epoch": 1.4476923076923076,
+      "grad_norm": 0.2968755624403521,
+      "learning_rate": 3.5515778726382967e-06,
+      "loss": 0.3028,
+      "step": 941
+    },
+    {
+      "epoch": 1.4492307692307693,
+      "grad_norm": 0.3412206347642324,
+      "learning_rate": 3.53312619191051e-06,
+      "loss": 0.3007,
+      "step": 942
+    },
+    {
+      "epoch": 1.4507692307692308,
+      "grad_norm": 0.3097376165627181,
+      "learning_rate": 3.5147122777242203e-06,
+      "loss": 0.305,
+      "step": 943
+    },
+    {
+      "epoch": 1.4523076923076923,
+      "grad_norm": 0.31633706056592903,
+      "learning_rate": 3.4963362376166886e-06,
+      "loss": 0.2996,
+      "step": 944
+    },
+    {
+      "epoch": 1.4538461538461538,
+      "grad_norm": 0.3104683767388432,
+      "learning_rate": 3.477998178903982e-06,
+      "loss": 0.3017,
+      "step": 945
+    },
+    {
+      "epoch": 1.4553846153846153,
+      "grad_norm": 0.3037086889007336,
+      "learning_rate": 3.459698208680359e-06,
+      "loss": 0.2989,
+      "step": 946
+    },
+    {
+      "epoch": 1.456923076923077,
+      "grad_norm": 0.31143542337111135,
+      "learning_rate": 3.441436433817641e-06,
+      "loss": 0.2936,
+      "step": 947
+    },
+    {
+      "epoch": 1.4584615384615385,
+      "grad_norm": 0.30765748498539663,
+      "learning_rate": 3.423212960964586e-06,
+      "loss": 0.298,
+      "step": 948
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.29811397523616656,
+      "learning_rate": 3.405027896546277e-06,
+      "loss": 0.3033,
+      "step": 949
+    },
+    {
+      "epoch": 1.4615384615384617,
+      "grad_norm": 0.29888506298514994,
+      "learning_rate": 3.3868813467634833e-06,
+      "loss": 0.302,
+      "step": 950
+    },
+    {
+      "epoch": 1.463076923076923,
+      "grad_norm": 0.3053162977493003,
+      "learning_rate": 3.3687734175920505e-06,
+      "loss": 0.2925,
+      "step": 951
+    },
+    {
+      "epoch": 1.4646153846153847,
+      "grad_norm": 0.2971791660430732,
+      "learning_rate": 3.350704214782278e-06,
+      "loss": 0.2835,
+      "step": 952
+    },
+    {
+      "epoch": 1.4661538461538461,
+      "grad_norm": 0.29770470644577074,
+      "learning_rate": 3.3326738438583116e-06,
+      "loss": 0.289,
+      "step": 953
+    },
+    {
+      "epoch": 1.4676923076923076,
+      "grad_norm": 0.29910643183641306,
+      "learning_rate": 3.314682410117511e-06,
+      "loss": 0.3046,
+      "step": 954
+    },
+    {
+      "epoch": 1.4692307692307693,
+      "grad_norm": 0.2919763104253795,
+      "learning_rate": 3.2967300186298456e-06,
+      "loss": 0.304,
+      "step": 955
+    },
+    {
+      "epoch": 1.4707692307692308,
+      "grad_norm": 0.32117968302916566,
+      "learning_rate": 3.2788167742372725e-06,
+      "loss": 0.2931,
+      "step": 956
+    },
+    {
+      "epoch": 1.4723076923076923,
+      "grad_norm": 0.3276844856461681,
+      "learning_rate": 3.2609427815531426e-06,
+      "loss": 0.3056,
+      "step": 957
+    },
+    {
+      "epoch": 1.4738461538461538,
+      "grad_norm": 0.30596875656090344,
+      "learning_rate": 3.243108144961563e-06,
+      "loss": 0.2976,
+      "step": 958
+    },
+    {
+      "epoch": 1.4753846153846153,
+      "grad_norm": 0.32116191589935694,
+      "learning_rate": 3.2253129686168105e-06,
+      "loss": 0.2998,
+      "step": 959
+    },
+    {
+      "epoch": 1.476923076923077,
+      "grad_norm": 0.3198590276627668,
+      "learning_rate": 3.2075573564427097e-06,
+      "loss": 0.2981,
+      "step": 960
+    },
+    {
+      "epoch": 1.4784615384615385,
+      "grad_norm": 0.3262658725554805,
+      "learning_rate": 3.1898414121320277e-06,
+      "loss": 0.2973,
+      "step": 961
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.32016365040939243,
+      "learning_rate": 3.1721652391458804e-06,
+      "loss": 0.2962,
+      "step": 962
+    },
+    {
+      "epoch": 1.4815384615384615,
+      "grad_norm": 0.324542860444357,
+      "learning_rate": 3.1545289407131128e-06,
+      "loss": 0.3084,
+      "step": 963
+    },
+    {
+      "epoch": 1.483076923076923,
+      "grad_norm": 0.2950912718170576,
+      "learning_rate": 3.1369326198297025e-06,
+      "loss": 0.2932,
+      "step": 964
+    },
+    {
+      "epoch": 1.4846153846153847,
+      "grad_norm": 0.2981652690481351,
+      "learning_rate": 3.11937637925816e-06,
+      "loss": 0.3166,
+      "step": 965
+    },
+    {
+      "epoch": 1.4861538461538462,
+      "grad_norm": 0.3154177218611572,
+      "learning_rate": 3.101860321526924e-06,
+      "loss": 0.3009,
+      "step": 966
+    },
+    {
+      "epoch": 1.4876923076923076,
+      "grad_norm": 0.31531993863084484,
+      "learning_rate": 3.0843845489297698e-06,
+      "loss": 0.2995,
+      "step": 967
+    },
+    {
+      "epoch": 1.4892307692307694,
+      "grad_norm": 0.29776367188667724,
+      "learning_rate": 3.066949163525205e-06,
+      "loss": 0.3128,
+      "step": 968
+    },
+    {
+      "epoch": 1.4907692307692308,
+      "grad_norm": 0.3006711490044736,
+      "learning_rate": 3.0495542671358745e-06,
+      "loss": 0.3011,
+      "step": 969
+    },
+    {
+      "epoch": 1.4923076923076923,
+      "grad_norm": 0.29334905949102336,
+      "learning_rate": 3.0321999613479668e-06,
+      "loss": 0.3002,
+      "step": 970
+    },
+    {
+      "epoch": 1.4938461538461538,
+      "grad_norm": 0.30793210955576666,
+      "learning_rate": 3.0148863475106315e-06,
+      "loss": 0.2942,
+      "step": 971
+    },
+    {
+      "epoch": 1.4953846153846153,
+      "grad_norm": 0.3211222025363804,
+      "learning_rate": 2.9976135267353636e-06,
+      "loss": 0.2929,
+      "step": 972
+    },
+    {
+      "epoch": 1.496923076923077,
+      "grad_norm": 0.31320288128174284,
+      "learning_rate": 2.9803815998954334e-06,
+      "loss": 0.2938,
+      "step": 973
+    },
+    {
+      "epoch": 1.4984615384615385,
+      "grad_norm": 0.31000689011337645,
+      "learning_rate": 2.9631906676252865e-06,
+      "loss": 0.291,
+      "step": 974
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.3108667388625751,
+      "learning_rate": 2.9460408303199696e-06,
+      "loss": 0.2998,
+      "step": 975
+    },
+    {
+      "epoch": 1.5015384615384615,
+      "grad_norm": 0.2939897367629014,
+      "learning_rate": 2.9289321881345257e-06,
+      "loss": 0.3149,
+      "step": 976
+    },
+    {
+      "epoch": 1.503076923076923,
+      "grad_norm": 0.33687035445432856,
+      "learning_rate": 2.9118648409834205e-06,
+      "loss": 0.3043,
+      "step": 977
+    },
+    {
+      "epoch": 1.5046153846153847,
+      "grad_norm": 0.3007783981264154,
+      "learning_rate": 2.894838888539957e-06,
+      "loss": 0.3038,
+      "step": 978
+    },
+    {
+      "epoch": 1.5061538461538462,
+      "grad_norm": 0.2824676002667487,
+      "learning_rate": 2.8778544302356904e-06,
+      "loss": 0.2874,
+      "step": 979
+    },
+    {
+      "epoch": 1.5076923076923077,
+      "grad_norm": 0.31633903440629485,
+      "learning_rate": 2.8609115652598595e-06,
+      "loss": 0.298,
+      "step": 980
+    },
+    {
+      "epoch": 1.5092307692307694,
+      "grad_norm": 0.3159917038664522,
+      "learning_rate": 2.8440103925587904e-06,
+      "loss": 0.2949,
+      "step": 981
+    },
+    {
+      "epoch": 1.5107692307692306,
+      "grad_norm": 0.3021575910385271,
+      "learning_rate": 2.8271510108353237e-06,
+      "loss": 0.304,
+      "step": 982
+    },
+    {
+      "epoch": 1.5123076923076924,
+      "grad_norm": 0.3041730942371279,
+      "learning_rate": 2.810333518548246e-06,
+      "loss": 0.299,
+      "step": 983
+    },
+    {
+      "epoch": 1.5138461538461538,
+      "grad_norm": 0.2926158325878474,
+      "learning_rate": 2.7935580139117114e-06,
+      "loss": 0.3052,
+      "step": 984
+    },
+    {
+      "epoch": 1.5153846153846153,
+      "grad_norm": 0.3083314057111434,
+      "learning_rate": 2.7768245948946615e-06,
+      "loss": 0.3001,
+      "step": 985
+    },
+    {
+      "epoch": 1.516923076923077,
+      "grad_norm": 0.3164728374661138,
+      "learning_rate": 2.7601333592202583e-06,
+      "loss": 0.3021,
+      "step": 986
+    },
+    {
+      "epoch": 1.5184615384615383,
+      "grad_norm": 0.3051860274707252,
+      "learning_rate": 2.743484404365314e-06,
+      "loss": 0.3073,
+      "step": 987
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.4127905457453725,
+      "learning_rate": 2.7268778275597217e-06,
+      "loss": 0.2974,
+      "step": 988
+    },
+    {
+      "epoch": 1.5215384615384615,
+      "grad_norm": 0.30081816370891307,
+      "learning_rate": 2.7103137257858867e-06,
+      "loss": 0.2981,
+      "step": 989
+    },
+    {
+      "epoch": 1.523076923076923,
+      "grad_norm": 0.29498926865493064,
+      "learning_rate": 2.6937921957781587e-06,
+      "loss": 0.2966,
+      "step": 990
+    },
+    {
+      "epoch": 1.5246153846153847,
+      "grad_norm": 0.2935237562870546,
+      "learning_rate": 2.6773133340222677e-06,
+      "loss": 0.2972,
+      "step": 991
+    },
+    {
+      "epoch": 1.5261538461538462,
+      "grad_norm": 0.33621080873516174,
+      "learning_rate": 2.660877236754762e-06,
+      "loss": 0.2998,
+      "step": 992
+    },
+    {
+      "epoch": 1.5276923076923077,
+      "grad_norm": 0.3294824964855458,
+      "learning_rate": 2.6444839999624496e-06,
+      "loss": 0.2902,
+      "step": 993
+    },
+    {
+      "epoch": 1.5292307692307694,
+      "grad_norm": 0.3065386601276519,
+      "learning_rate": 2.6281337193818267e-06,
+      "loss": 0.3058,
+      "step": 994
+    },
+    {
+      "epoch": 1.5307692307692307,
+      "grad_norm": 0.3040209272698475,
+      "learning_rate": 2.61182649049853e-06,
+      "loss": 0.2862,
+      "step": 995
+    },
+    {
+      "epoch": 1.5323076923076924,
+      "grad_norm": 0.3094142360855812,
+      "learning_rate": 2.59556240854677e-06,
+      "loss": 0.2928,
+      "step": 996
+    },
+    {
+      "epoch": 1.5338461538461539,
+      "grad_norm": 0.3012445004355704,
+      "learning_rate": 2.5793415685087797e-06,
+      "loss": 0.3007,
+      "step": 997
+    },
+    {
+      "epoch": 1.5353846153846153,
+      "grad_norm": 0.28762674633831675,
+      "learning_rate": 2.5631640651142654e-06,
+      "loss": 0.3005,
+      "step": 998
+    },
+    {
+      "epoch": 1.536923076923077,
+      "grad_norm": 0.3061516270687886,
+      "learning_rate": 2.5470299928398424e-06,
+      "loss": 0.3024,
+      "step": 999
+    },
+    {
+      "epoch": 1.5384615384615383,
+      "grad_norm": 0.2873785092087951,
+      "learning_rate": 2.5309394459084878e-06,
+      "loss": 0.2871,
+      "step": 1000
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.2924924604142306,
+      "learning_rate": 2.514892518288988e-06,
+      "loss": 0.2882,
+      "step": 1001
+    },
+    {
+      "epoch": 1.5415384615384615,
+      "grad_norm": 0.28384576763366215,
+      "learning_rate": 2.4988893036954045e-06,
+      "loss": 0.2911,
+      "step": 1002
+    },
+    {
+      "epoch": 1.543076923076923,
+      "grad_norm": 0.30415796578717974,
+      "learning_rate": 2.4829298955865022e-06,
+      "loss": 0.2944,
+      "step": 1003
+    },
+    {
+      "epoch": 1.5446153846153847,
+      "grad_norm": 0.2971693194460424,
+      "learning_rate": 2.467014387165222e-06,
+      "loss": 0.3065,
+      "step": 1004
+    },
+    {
+      "epoch": 1.546153846153846,
+      "grad_norm": 0.31287338855943536,
+      "learning_rate": 2.451142871378124e-06,
+      "loss": 0.3091,
+      "step": 1005
+    },
+    {
+      "epoch": 1.5476923076923077,
+      "grad_norm": 0.30394746336813816,
+      "learning_rate": 2.4353154409148637e-06,
+      "loss": 0.2998,
+      "step": 1006
+    },
+    {
+      "epoch": 1.5492307692307692,
+      "grad_norm": 0.2955541437963843,
+      "learning_rate": 2.4195321882076295e-06,
+      "loss": 0.3005,
+      "step": 1007
+    },
+    {
+      "epoch": 1.5507692307692307,
+      "grad_norm": 0.3021488673120061,
+      "learning_rate": 2.4037932054306125e-06,
+      "loss": 0.2991,
+      "step": 1008
+    },
+    {
+      "epoch": 1.5523076923076924,
+      "grad_norm": 0.30648168671124143,
+      "learning_rate": 2.3880985844994674e-06,
+      "loss": 0.2997,
+      "step": 1009
+    },
+    {
+      "epoch": 1.5538461538461539,
+      "grad_norm": 0.3547032972739633,
+      "learning_rate": 2.372448417070776e-06,
+      "loss": 0.2932,
+      "step": 1010
+    },
+    {
+      "epoch": 1.5553846153846154,
+      "grad_norm": 0.3216079686771892,
+      "learning_rate": 2.3568427945415163e-06,
+      "loss": 0.2969,
+      "step": 1011
+    },
+    {
+      "epoch": 1.556923076923077,
+      "grad_norm": 0.2975979180003544,
+      "learning_rate": 2.3412818080485176e-06,
+      "loss": 0.3027,
+      "step": 1012
+    },
+    {
+      "epoch": 1.5584615384615383,
+      "grad_norm": 0.2889462660599791,
+      "learning_rate": 2.3257655484679376e-06,
+      "loss": 0.2849,
+      "step": 1013
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.3180478292364749,
+      "learning_rate": 2.3102941064147287e-06,
+      "loss": 0.3016,
+      "step": 1014
+    },
+    {
+      "epoch": 1.5615384615384615,
+      "grad_norm": 0.3178408229378375,
+      "learning_rate": 2.2948675722421086e-06,
+      "loss": 0.2929,
+      "step": 1015
+    },
+    {
+      "epoch": 1.563076923076923,
+      "grad_norm": 0.3042081351027437,
+      "learning_rate": 2.279486036041034e-06,
+      "loss": 0.2975,
+      "step": 1016
+    },
+    {
+      "epoch": 1.5646153846153847,
+      "grad_norm": 0.3174012708546431,
+      "learning_rate": 2.2641495876396713e-06,
+      "loss": 0.3167,
+      "step": 1017
+    },
+    {
+      "epoch": 1.566153846153846,
+      "grad_norm": 0.31937180451780706,
+      "learning_rate": 2.2488583166028754e-06,
+      "loss": 0.2993,
+      "step": 1018
+    },
+    {
+      "epoch": 1.5676923076923077,
+      "grad_norm": 0.3036032085718437,
+      "learning_rate": 2.2336123122316642e-06,
+      "loss": 0.2943,
+      "step": 1019
+    },
+    {
+      "epoch": 1.5692307692307692,
+      "grad_norm": 0.2896487254529284,
+      "learning_rate": 2.218411663562704e-06,
+      "loss": 0.3023,
+      "step": 1020
+    },
+    {
+      "epoch": 1.5707692307692307,
+      "grad_norm": 0.3129595839968114,
+      "learning_rate": 2.2032564593677773e-06,
+      "loss": 0.3007,
+      "step": 1021
+    },
+    {
+      "epoch": 1.5723076923076924,
+      "grad_norm": 0.306812556457923,
+      "learning_rate": 2.1881467881532737e-06,
+      "loss": 0.293,
+      "step": 1022
+    },
+    {
+      "epoch": 1.573846153846154,
+      "grad_norm": 0.30000813552544126,
+      "learning_rate": 2.1730827381596643e-06,
+      "loss": 0.2912,
+      "step": 1023
+    },
+    {
+      "epoch": 1.5753846153846154,
+      "grad_norm": 0.3280891494003774,
+      "learning_rate": 2.158064397361005e-06,
+      "loss": 0.2953,
+      "step": 1024
+    },
+    {
+      "epoch": 1.5769230769230769,
+      "grad_norm": 0.30132133832278324,
+      "learning_rate": 2.1430918534643996e-06,
+      "loss": 0.3021,
+      "step": 1025
+    },
+    {
+      "epoch": 1.5784615384615384,
+      "grad_norm": 0.3066966759651527,
+      "learning_rate": 2.1281651939094996e-06,
+      "loss": 0.3009,
+      "step": 1026
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.29478629136919327,
+      "learning_rate": 2.1132845058679942e-06,
+      "loss": 0.308,
+      "step": 1027
+    },
+    {
+      "epoch": 1.5815384615384616,
+      "grad_norm": 0.29005328878682973,
+      "learning_rate": 2.098449876243096e-06,
+      "loss": 0.3003,
+      "step": 1028
+    },
+    {
+      "epoch": 1.583076923076923,
+      "grad_norm": 0.30395913737658703,
+      "learning_rate": 2.083661391669043e-06,
+      "loss": 0.3091,
+      "step": 1029
+    },
+    {
+      "epoch": 1.5846153846153848,
+      "grad_norm": 0.29406096777127433,
+      "learning_rate": 2.0689191385105787e-06,
+      "loss": 0.2912,
+      "step": 1030
+    },
+    {
+      "epoch": 1.586153846153846,
+      "grad_norm": 0.2914236582614912,
+      "learning_rate": 2.0542232028624585e-06,
+      "loss": 0.2923,
+      "step": 1031
+    },
+    {
+      "epoch": 1.5876923076923077,
+      "grad_norm": 0.2850251017710342,
+      "learning_rate": 2.03957367054894e-06,
+      "loss": 0.2955,
+      "step": 1032
+    },
+    {
+      "epoch": 1.5892307692307692,
+      "grad_norm": 0.30652180735368845,
+      "learning_rate": 2.024970627123295e-06,
+      "loss": 0.2969,
+      "step": 1033
+    },
+    {
+      "epoch": 1.5907692307692307,
+      "grad_norm": 0.2893031895711992,
+      "learning_rate": 2.0104141578672887e-06,
+      "loss": 0.2948,
+      "step": 1034
+    },
+    {
+      "epoch": 1.5923076923076924,
+      "grad_norm": 0.3044041898577739,
+      "learning_rate": 1.9959043477907e-06,
+      "loss": 0.2914,
+      "step": 1035
+    },
+    {
+      "epoch": 1.5938461538461537,
+      "grad_norm": 0.29290920454426944,
+      "learning_rate": 1.981441281630816e-06,
+      "loss": 0.2905,
+      "step": 1036
+    },
+    {
+      "epoch": 1.5953846153846154,
+      "grad_norm": 0.3198464680313194,
+      "learning_rate": 1.967025043851939e-06,
+      "loss": 0.2862,
+      "step": 1037
+    },
+    {
+      "epoch": 1.596923076923077,
+      "grad_norm": 0.2927921230738868,
+      "learning_rate": 1.9526557186448924e-06,
+      "loss": 0.3072,
+      "step": 1038
+    },
+    {
+      "epoch": 1.5984615384615384,
+      "grad_norm": 0.30236454361586024,
+      "learning_rate": 1.9383333899265368e-06,
+      "loss": 0.306,
+      "step": 1039
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.31181650953152307,
+      "learning_rate": 1.9240581413392647e-06,
+      "loss": 0.292,
+      "step": 1040
+    },
+    {
+      "epoch": 1.6015384615384616,
+      "grad_norm": 0.2913258377428407,
+      "learning_rate": 1.9098300562505266e-06,
+      "loss": 0.308,
+      "step": 1041
+    },
+    {
+      "epoch": 1.603076923076923,
+      "grad_norm": 0.31131969180141933,
+      "learning_rate": 1.8956492177523345e-06,
+      "loss": 0.3011,
+      "step": 1042
+    },
+    {
+      "epoch": 1.6046153846153848,
+      "grad_norm": 0.30223958128816997,
+      "learning_rate": 1.8815157086607826e-06,
+      "loss": 0.304,
+      "step": 1043
+    },
+    {
+      "epoch": 1.606153846153846,
+      "grad_norm": 0.2982624818189629,
+      "learning_rate": 1.86742961151556e-06,
+      "loss": 0.2919,
+      "step": 1044
+    },
+    {
+      "epoch": 1.6076923076923078,
+      "grad_norm": 0.30607353070619775,
+      "learning_rate": 1.8533910085794714e-06,
+      "loss": 0.2954,
+      "step": 1045
+    },
+    {
+      "epoch": 1.6092307692307692,
+      "grad_norm": 0.3113699970353025,
+      "learning_rate": 1.8393999818379527e-06,
+      "loss": 0.2852,
+      "step": 1046
+    },
+    {
+      "epoch": 1.6107692307692307,
+      "grad_norm": 0.3048424135365832,
+      "learning_rate": 1.8254566129985996e-06,
+      "loss": 0.2994,
+      "step": 1047
+    },
+    {
+      "epoch": 1.6123076923076924,
+      "grad_norm": 0.28766837251724475,
+      "learning_rate": 1.8115609834906821e-06,
+      "loss": 0.3065,
+      "step": 1048
+    },
+    {
+      "epoch": 1.6138461538461537,
+      "grad_norm": 0.3117176531273534,
+      "learning_rate": 1.7977131744646724e-06,
+      "loss": 0.3037,
+      "step": 1049
+    },
+    {
+      "epoch": 1.6153846153846154,
+      "grad_norm": 0.2887362639331083,
+      "learning_rate": 1.7839132667917692e-06,
+      "loss": 0.3115,
+      "step": 1050
+    },
+    {
+      "epoch": 1.616923076923077,
+      "grad_norm": 0.292432776537684,
+      "learning_rate": 1.7701613410634367e-06,
+      "loss": 0.2914,
+      "step": 1051
+    },
+    {
+      "epoch": 1.6184615384615384,
+      "grad_norm": 0.3043909373848769,
+      "learning_rate": 1.7564574775909127e-06,
+      "loss": 0.2887,
+      "step": 1052
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.29547665624342945,
+      "learning_rate": 1.7428017564047594e-06,
+      "loss": 0.2758,
+      "step": 1053
+    },
+    {
+      "epoch": 1.6215384615384614,
+      "grad_norm": 0.2937163693100437,
+      "learning_rate": 1.7291942572543806e-06,
+      "loss": 0.315,
+      "step": 1054
+    },
+    {
+      "epoch": 1.623076923076923,
+      "grad_norm": 0.29567117123952397,
+      "learning_rate": 1.7156350596075743e-06,
+      "loss": 0.3009,
+      "step": 1055
+    },
+    {
+      "epoch": 1.6246153846153846,
+      "grad_norm": 0.3057036556316489,
+      "learning_rate": 1.7021242426500495e-06,
+      "loss": 0.2978,
+      "step": 1056
+    },
+    {
+      "epoch": 1.626153846153846,
+      "grad_norm": 0.2974014132020775,
+      "learning_rate": 1.6886618852849723e-06,
+      "loss": 0.294,
+      "step": 1057
+    },
+    {
+      "epoch": 1.6276923076923078,
+      "grad_norm": 0.3086904981880302,
+      "learning_rate": 1.6752480661325077e-06,
+      "loss": 0.3006,
+      "step": 1058
+    },
+    {
+      "epoch": 1.6292307692307693,
+      "grad_norm": 0.29452454472572687,
+      "learning_rate": 1.6618828635293538e-06,
+      "loss": 0.2841,
+      "step": 1059
+    },
+    {
+      "epoch": 1.6307692307692307,
+      "grad_norm": 0.2985420856067674,
+      "learning_rate": 1.6485663555282949e-06,
+      "loss": 0.2951,
+      "step": 1060
+    },
+    {
+      "epoch": 1.6323076923076925,
+      "grad_norm": 0.30107080292385824,
+      "learning_rate": 1.6352986198977327e-06,
+      "loss": 0.2978,
+      "step": 1061
+    },
+    {
+      "epoch": 1.6338461538461537,
+      "grad_norm": 0.3093685220433861,
+      "learning_rate": 1.6220797341212401e-06,
+      "loss": 0.2886,
+      "step": 1062
+    },
+    {
+      "epoch": 1.6353846153846154,
+      "grad_norm": 0.2954396029299978,
+      "learning_rate": 1.6089097753971061e-06,
+      "loss": 0.2905,
+      "step": 1063
+    },
+    {
+      "epoch": 1.636923076923077,
+      "grad_norm": 0.30558641966165406,
+      "learning_rate": 1.595788820637888e-06,
+      "loss": 0.2864,
+      "step": 1064
+    },
+    {
+      "epoch": 1.6384615384615384,
+      "grad_norm": 0.2816114969844964,
+      "learning_rate": 1.5827169464699576e-06,
+      "loss": 0.3037,
+      "step": 1065
+    },
+    {
+      "epoch": 1.6400000000000001,
+      "grad_norm": 0.28528914477241396,
+      "learning_rate": 1.5696942292330574e-06,
+      "loss": 0.3113,
+      "step": 1066
+    },
+    {
+      "epoch": 1.6415384615384614,
+      "grad_norm": 0.2961725860976731,
+      "learning_rate": 1.5567207449798517e-06,
+      "loss": 0.2976,
+      "step": 1067
+    },
+    {
+      "epoch": 1.643076923076923,
+      "grad_norm": 0.3369369798610147,
+      "learning_rate": 1.5437965694754842e-06,
+      "loss": 0.2973,
+      "step": 1068
+    },
+    {
+      "epoch": 1.6446153846153846,
+      "grad_norm": 0.31277693532314743,
+      "learning_rate": 1.5309217781971419e-06,
+      "loss": 0.3016,
+      "step": 1069
+    },
+    {
+      "epoch": 1.646153846153846,
+      "grad_norm": 0.30874906442343925,
+      "learning_rate": 1.518096446333599e-06,
+      "loss": 0.3025,
+      "step": 1070
+    },
+    {
+      "epoch": 1.6476923076923078,
+      "grad_norm": 0.30721013699733996,
+      "learning_rate": 1.5053206487847916e-06,
+      "loss": 0.2932,
+      "step": 1071
+    },
+    {
+      "epoch": 1.6492307692307693,
+      "grad_norm": 0.30940175578992224,
+      "learning_rate": 1.4925944601613718e-06,
+      "loss": 0.2974,
+      "step": 1072
+    },
+    {
+      "epoch": 1.6507692307692308,
+      "grad_norm": 0.2660850113801858,
+      "learning_rate": 1.4799179547842823e-06,
+      "loss": 0.2987,
+      "step": 1073
+    },
+    {
+      "epoch": 1.6523076923076923,
+      "grad_norm": 0.2788205661101268,
+      "learning_rate": 1.4672912066843103e-06,
+      "loss": 0.2923,
+      "step": 1074
+    },
+    {
+      "epoch": 1.6538461538461537,
+      "grad_norm": 0.3013785373926941,
+      "learning_rate": 1.454714289601661e-06,
+      "loss": 0.2977,
+      "step": 1075
+    },
+    {
+      "epoch": 1.6553846153846155,
+      "grad_norm": 0.28426980236184257,
+      "learning_rate": 1.4421872769855262e-06,
+      "loss": 0.3019,
+      "step": 1076
+    },
+    {
+      "epoch": 1.656923076923077,
+      "grad_norm": 0.2971426871587678,
+      "learning_rate": 1.4297102419936559e-06,
+      "loss": 0.2989,
+      "step": 1077
+    },
+    {
+      "epoch": 1.6584615384615384,
+      "grad_norm": 0.27969037920131673,
+      "learning_rate": 1.4172832574919359e-06,
+      "loss": 0.293,
+      "step": 1078
+    },
+    {
+      "epoch": 1.6600000000000001,
+      "grad_norm": 0.3013651067208784,
+      "learning_rate": 1.4049063960539488e-06,
+      "loss": 0.3025,
+      "step": 1079
+    },
+    {
+      "epoch": 1.6615384615384614,
+      "grad_norm": 0.27945051114194774,
+      "learning_rate": 1.3925797299605649e-06,
+      "loss": 0.2985,
+      "step": 1080
+    },
+    {
+      "epoch": 1.6630769230769231,
+      "grad_norm": 0.295937531579909,
+      "learning_rate": 1.3803033311995072e-06,
+      "loss": 0.2917,
+      "step": 1081
+    },
+    {
+      "epoch": 1.6646153846153846,
+      "grad_norm": 0.3228482401143072,
+      "learning_rate": 1.368077271464946e-06,
+      "loss": 0.2802,
+      "step": 1082
+    },
+    {
+      "epoch": 1.666153846153846,
+      "grad_norm": 0.29118424277549815,
+      "learning_rate": 1.3559016221570663e-06,
+      "loss": 0.3061,
+      "step": 1083
+    },
+    {
+      "epoch": 1.6676923076923078,
+      "grad_norm": 0.29327410105679824,
+      "learning_rate": 1.3437764543816556e-06,
+      "loss": 0.2988,
+      "step": 1084
+    },
+    {
+      "epoch": 1.669230769230769,
+      "grad_norm": 0.2904193477394739,
+      "learning_rate": 1.3317018389496927e-06,
+      "loss": 0.3127,
+      "step": 1085
+    },
+    {
+      "epoch": 1.6707692307692308,
+      "grad_norm": 0.2963639025499214,
+      "learning_rate": 1.3196778463769256e-06,
+      "loss": 0.2933,
+      "step": 1086
+    },
+    {
+      "epoch": 1.6723076923076923,
+      "grad_norm": 0.3079345350364671,
+      "learning_rate": 1.3077045468834714e-06,
+      "loss": 0.2814,
+      "step": 1087
+    },
+    {
+      "epoch": 1.6738461538461538,
+      "grad_norm": 0.2818776921772326,
+      "learning_rate": 1.295782010393396e-06,
+      "loss": 0.303,
+      "step": 1088
+    },
+    {
+      "epoch": 1.6753846153846155,
+      "grad_norm": 0.2869928793803104,
+      "learning_rate": 1.2839103065343084e-06,
+      "loss": 0.3015,
+      "step": 1089
+    },
+    {
+      "epoch": 1.676923076923077,
+      "grad_norm": 0.297976687750705,
+      "learning_rate": 1.2720895046369564e-06,
+      "loss": 0.2905,
+      "step": 1090
+    },
+    {
+      "epoch": 1.6784615384615384,
+      "grad_norm": 0.2812140611264347,
+      "learning_rate": 1.2603196737348211e-06,
+      "loss": 0.2987,
+      "step": 1091
+    },
+    {
+      "epoch": 1.6800000000000002,
+      "grad_norm": 0.2804335556384957,
+      "learning_rate": 1.2486008825637119e-06,
+      "loss": 0.2825,
+      "step": 1092
+    },
+    {
+      "epoch": 1.6815384615384614,
+      "grad_norm": 0.3071782818760706,
+      "learning_rate": 1.2369331995613664e-06,
+      "loss": 0.2956,
+      "step": 1093
+    },
+    {
+      "epoch": 1.6830769230769231,
+      "grad_norm": 0.2844997596345823,
+      "learning_rate": 1.2253166928670478e-06,
+      "loss": 0.2975,
+      "step": 1094
+    },
+    {
+      "epoch": 1.6846153846153846,
+      "grad_norm": 0.28683506331008113,
+      "learning_rate": 1.213751430321156e-06,
+      "loss": 0.294,
+      "step": 1095
+    },
+    {
+      "epoch": 1.6861538461538461,
+      "grad_norm": 0.2849064391557187,
+      "learning_rate": 1.2022374794648229e-06,
+      "loss": 0.298,
+      "step": 1096
+    },
+    {
+      "epoch": 1.6876923076923078,
+      "grad_norm": 0.29089888768154437,
+      "learning_rate": 1.1907749075395147e-06,
+      "loss": 0.2915,
+      "step": 1097
+    },
+    {
+      "epoch": 1.689230769230769,
+      "grad_norm": 0.28800575022567315,
+      "learning_rate": 1.179363781486651e-06,
+      "loss": 0.2895,
+      "step": 1098
+    },
+    {
+      "epoch": 1.6907692307692308,
+      "grad_norm": 0.31008251483746824,
+      "learning_rate": 1.168004167947202e-06,
+      "loss": 0.3037,
+      "step": 1099
+    },
+    {
+      "epoch": 1.6923076923076923,
+      "grad_norm": 0.312236160414082,
+      "learning_rate": 1.1566961332613136e-06,
+      "loss": 0.3096,
+      "step": 1100
+    },
+    {
+      "epoch": 1.6938461538461538,
+      "grad_norm": 0.2876360590008567,
+      "learning_rate": 1.1454397434679022e-06,
+      "loss": 0.297,
+      "step": 1101
+    },
+    {
+      "epoch": 1.6953846153846155,
+      "grad_norm": 0.31995321703395463,
+      "learning_rate": 1.1342350643042822e-06,
+      "loss": 0.2919,
+      "step": 1102
+    },
+    {
+      "epoch": 1.696923076923077,
+      "grad_norm": 0.3002903152211252,
+      "learning_rate": 1.123082161205775e-06,
+      "loss": 0.3051,
+      "step": 1103
+    },
+    {
+      "epoch": 1.6984615384615385,
+      "grad_norm": 0.2702402306373896,
+      "learning_rate": 1.111981099305336e-06,
+      "loss": 0.2932,
+      "step": 1104
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.3076912695270031,
+      "learning_rate": 1.1009319434331623e-06,
+      "loss": 0.293,
+      "step": 1105
+    },
+    {
+      "epoch": 1.7015384615384614,
+      "grad_norm": 0.27934320254587486,
+      "learning_rate": 1.0899347581163222e-06,
+      "loss": 0.308,
+      "step": 1106
+    },
+    {
+      "epoch": 1.7030769230769232,
+      "grad_norm": 0.28600221103923024,
+      "learning_rate": 1.0789896075783734e-06,
+      "loss": 0.2883,
+      "step": 1107
+    },
+    {
+      "epoch": 1.7046153846153846,
+      "grad_norm": 0.30723721814103333,
+      "learning_rate": 1.0680965557389934e-06,
+      "loss": 0.3025,
+      "step": 1108
+    },
+    {
+      "epoch": 1.7061538461538461,
+      "grad_norm": 0.29048566891002353,
+      "learning_rate": 1.0572556662136036e-06,
+      "loss": 0.2999,
+      "step": 1109
+    },
+    {
+      "epoch": 1.7076923076923078,
+      "grad_norm": 0.288359119940675,
+      "learning_rate": 1.0464670023129952e-06,
+      "loss": 0.2955,
+      "step": 1110
+    },
+    {
+      "epoch": 1.709230769230769,
+      "grad_norm": 0.2989471025652296,
+      "learning_rate": 1.0357306270429623e-06,
+      "loss": 0.2917,
+      "step": 1111
+    },
+    {
+      "epoch": 1.7107692307692308,
+      "grad_norm": 0.2880092274808718,
+      "learning_rate": 1.0250466031039353e-06,
+      "loss": 0.2954,
+      "step": 1112
+    },
+    {
+      "epoch": 1.7123076923076923,
+      "grad_norm": 0.2824338657390046,
+      "learning_rate": 1.014414992890611e-06,
+      "loss": 0.2952,
+      "step": 1113
+    },
+    {
+      "epoch": 1.7138461538461538,
+      "grad_norm": 0.3197516670934666,
+      "learning_rate": 1.0038358584915896e-06,
+      "loss": 0.2944,
+      "step": 1114
+    },
+    {
+      "epoch": 1.7153846153846155,
+      "grad_norm": 0.29761535660719984,
+      "learning_rate": 9.93309261689015e-07,
+      "loss": 0.3009,
+      "step": 1115
+    },
+    {
+      "epoch": 1.7169230769230768,
+      "grad_norm": 0.28756032290839234,
+      "learning_rate": 9.828352639582073e-07,
+      "loss": 0.2989,
+      "step": 1116
+    },
+    {
+      "epoch": 1.7184615384615385,
+      "grad_norm": 0.2846963660891568,
+      "learning_rate": 9.724139264673116e-07,
+      "loss": 0.3002,
+      "step": 1117
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.307418445752309,
+      "learning_rate": 9.6204531007694e-07,
+      "loss": 0.2951,
+      "step": 1118
+    },
+    {
+      "epoch": 1.7215384615384615,
+      "grad_norm": 0.30343032467654196,
+      "learning_rate": 9.517294753398066e-07,
+      "loss": 0.3063,
+      "step": 1119
+    },
+    {
+      "epoch": 1.7230769230769232,
+      "grad_norm": 0.3005301248581193,
+      "learning_rate": 9.414664825003838e-07,
+      "loss": 0.3003,
+      "step": 1120
+    },
+    {
+      "epoch": 1.7246153846153847,
+      "grad_norm": 0.28894141518022093,
+      "learning_rate": 9.312563914945461e-07,
+      "loss": 0.2864,
+      "step": 1121
+    },
+    {
+      "epoch": 1.7261538461538461,
+      "grad_norm": 0.2953804176655424,
+      "learning_rate": 9.210992619492254e-07,
+      "loss": 0.3007,
+      "step": 1122
+    },
+    {
+      "epoch": 1.7276923076923076,
+      "grad_norm": 0.283024496224878,
+      "learning_rate": 9.10995153182056e-07,
+      "loss": 0.3063,
+      "step": 1123
+    },
+    {
+      "epoch": 1.7292307692307691,
+      "grad_norm": 0.2710191054723531,
+      "learning_rate": 9.009441242010287e-07,
+      "loss": 0.2933,
+      "step": 1124
+    },
+    {
+      "epoch": 1.7307692307692308,
+      "grad_norm": 0.3024069321977267,
+      "learning_rate": 8.909462337041508e-07,
+      "loss": 0.2968,
+      "step": 1125
+    },
+    {
+      "epoch": 1.7323076923076923,
+      "grad_norm": 0.30865345828431323,
+      "learning_rate": 8.810015400790994e-07,
+      "loss": 0.2912,
+      "step": 1126
+    },
+    {
+      "epoch": 1.7338461538461538,
+      "grad_norm": 0.2894013192650705,
+      "learning_rate": 8.711101014028855e-07,
+      "loss": 0.2964,
+      "step": 1127
+    },
+    {
+      "epoch": 1.7353846153846155,
+      "grad_norm": 0.2925729981240855,
+      "learning_rate": 8.612719754415078e-07,
+      "loss": 0.3016,
+      "step": 1128
+    },
+    {
+      "epoch": 1.7369230769230768,
+      "grad_norm": 0.30193418891805185,
+      "learning_rate": 8.514872196496182e-07,
+      "loss": 0.2946,
+      "step": 1129
+    },
+    {
+      "epoch": 1.7384615384615385,
+      "grad_norm": 0.2919900192366226,
+      "learning_rate": 8.417558911701884e-07,
+      "loss": 0.292,
+      "step": 1130
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.2909252007648698,
+      "learning_rate": 8.320780468341761e-07,
+      "loss": 0.3005,
+      "step": 1131
+    },
+    {
+      "epoch": 1.7415384615384615,
+      "grad_norm": 0.28298888616991724,
+      "learning_rate": 8.224537431601886e-07,
+      "loss": 0.299,
+      "step": 1132
+    },
+    {
+      "epoch": 1.7430769230769232,
+      "grad_norm": 0.28368571925958197,
+      "learning_rate": 8.128830363541574e-07,
+      "loss": 0.2924,
+      "step": 1133
+    },
+    {
+      "epoch": 1.7446153846153845,
+      "grad_norm": 0.30745685144171864,
+      "learning_rate": 8.03365982309009e-07,
+      "loss": 0.2979,
+      "step": 1134
+    },
+    {
+      "epoch": 1.7461538461538462,
+      "grad_norm": 0.3158006104040895,
+      "learning_rate": 7.939026366043323e-07,
+      "loss": 0.2981,
+      "step": 1135
+    },
+    {
+      "epoch": 1.7476923076923077,
+      "grad_norm": 0.28918439926167333,
+      "learning_rate": 7.844930545060703e-07,
+      "loss": 0.2884,
+      "step": 1136
+    },
+    {
+      "epoch": 1.7492307692307691,
+      "grad_norm": 0.28730498350409545,
+      "learning_rate": 7.75137290966177e-07,
+      "loss": 0.294,
+      "step": 1137
+    },
+    {
+      "epoch": 1.7507692307692309,
+      "grad_norm": 0.2885652282032739,
+      "learning_rate": 7.65835400622309e-07,
+      "loss": 0.2916,
+      "step": 1138
+    },
+    {
+      "epoch": 1.7523076923076923,
+      "grad_norm": 0.30912721008745997,
+      "learning_rate": 7.565874377975046e-07,
+      "loss": 0.2942,
+      "step": 1139
+    },
+    {
+      "epoch": 1.7538461538461538,
+      "grad_norm": 0.2964238973443476,
+      "learning_rate": 7.473934564998641e-07,
+      "loss": 0.2889,
+      "step": 1140
+    },
+    {
+      "epoch": 1.7553846153846155,
+      "grad_norm": 0.29113108892167544,
+      "learning_rate": 7.382535104222366e-07,
+      "loss": 0.2971,
+      "step": 1141
+    },
+    {
+      "epoch": 1.7569230769230768,
+      "grad_norm": 0.275104358352756,
+      "learning_rate": 7.291676529419034e-07,
+      "loss": 0.2955,
+      "step": 1142
+    },
+    {
+      "epoch": 1.7584615384615385,
+      "grad_norm": 0.2983472777968824,
+      "learning_rate": 7.201359371202698e-07,
+      "loss": 0.2919,
+      "step": 1143
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.28799659758737967,
+      "learning_rate": 7.111584157025575e-07,
+      "loss": 0.3015,
+      "step": 1144
+    },
+    {
+      "epoch": 1.7615384615384615,
+      "grad_norm": 0.27764543668710895,
+      "learning_rate": 7.022351411174866e-07,
+      "loss": 0.2986,
+      "step": 1145
+    },
+    {
+      "epoch": 1.7630769230769232,
+      "grad_norm": 0.2802317982709836,
+      "learning_rate": 6.933661654769797e-07,
+      "loss": 0.3019,
+      "step": 1146
+    },
+    {
+      "epoch": 1.7646153846153845,
+      "grad_norm": 0.27986390652380927,
+      "learning_rate": 6.845515405758518e-07,
+      "loss": 0.2971,
+      "step": 1147
+    },
+    {
+      "epoch": 1.7661538461538462,
+      "grad_norm": 0.30385368076049846,
+      "learning_rate": 6.757913178915087e-07,
+      "loss": 0.2992,
+      "step": 1148
+    },
+    {
+      "epoch": 1.7676923076923077,
+      "grad_norm": 0.3162808669727755,
+      "learning_rate": 6.670855485836525e-07,
+      "loss": 0.2899,
+      "step": 1149
+    },
+    {
+      "epoch": 1.7692307692307692,
+      "grad_norm": 0.2855493712217652,
+      "learning_rate": 6.584342834939717e-07,
+      "loss": 0.2942,
+      "step": 1150
+    },
+    {
+      "epoch": 1.7707692307692309,
+      "grad_norm": 0.31071680682313796,
+      "learning_rate": 6.498375731458529e-07,
+      "loss": 0.286,
+      "step": 1151
+    },
+    {
+      "epoch": 1.7723076923076924,
+      "grad_norm": 0.2785945275534163,
+      "learning_rate": 6.412954677440797e-07,
+      "loss": 0.3003,
+      "step": 1152
+    },
+    {
+      "epoch": 1.7738461538461539,
+      "grad_norm": 0.26908931497920174,
+      "learning_rate": 6.32808017174551e-07,
+      "loss": 0.3056,
+      "step": 1153
+    },
+    {
+      "epoch": 1.7753846153846153,
+      "grad_norm": 0.27792172830373735,
+      "learning_rate": 6.243752710039719e-07,
+      "loss": 0.2931,
+      "step": 1154
+    },
+    {
+      "epoch": 1.7769230769230768,
+      "grad_norm": 0.27817702507789704,
+      "learning_rate": 6.159972784795798e-07,
+      "loss": 0.2907,
+      "step": 1155
+    },
+    {
+      "epoch": 1.7784615384615385,
+      "grad_norm": 0.29224151451364866,
+      "learning_rate": 6.076740885288479e-07,
+      "loss": 0.2878,
+      "step": 1156
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.2789677930224576,
+      "learning_rate": 5.994057497592032e-07,
+      "loss": 0.2925,
+      "step": 1157
+    },
+    {
+      "epoch": 1.7815384615384615,
+      "grad_norm": 0.2858438784150728,
+      "learning_rate": 5.911923104577455e-07,
+      "loss": 0.29,
+      "step": 1158
+    },
+    {
+      "epoch": 1.7830769230769232,
+      "grad_norm": 0.2995827639958211,
+      "learning_rate": 5.830338185909545e-07,
+      "loss": 0.293,
+      "step": 1159
+    },
+    {
+      "epoch": 1.7846153846153845,
+      "grad_norm": 0.28992097953036083,
+      "learning_rate": 5.749303218044234e-07,
+      "loss": 0.2918,
+      "step": 1160
+    },
+    {
+      "epoch": 1.7861538461538462,
+      "grad_norm": 0.2845454485535801,
+      "learning_rate": 5.668818674225684e-07,
+      "loss": 0.2944,
+      "step": 1161
+    },
+    {
+      "epoch": 1.7876923076923077,
+      "grad_norm": 0.28657029253363236,
+      "learning_rate": 5.588885024483648e-07,
+      "loss": 0.2907,
+      "step": 1162
+    },
+    {
+      "epoch": 1.7892307692307692,
+      "grad_norm": 0.2814747550101319,
+      "learning_rate": 5.509502735630601e-07,
+      "loss": 0.2987,
+      "step": 1163
+    },
+    {
+      "epoch": 1.790769230769231,
+      "grad_norm": 0.293869035289976,
+      "learning_rate": 5.430672271259096e-07,
+      "loss": 0.2859,
+      "step": 1164
+    },
+    {
+      "epoch": 1.7923076923076922,
+      "grad_norm": 0.27711145533788184,
+      "learning_rate": 5.352394091739022e-07,
+      "loss": 0.2864,
+      "step": 1165
+    },
+    {
+      "epoch": 1.7938461538461539,
+      "grad_norm": 0.30047756669775866,
+      "learning_rate": 5.274668654214931e-07,
+      "loss": 0.29,
+      "step": 1166
+    },
+    {
+      "epoch": 1.7953846153846154,
+      "grad_norm": 0.28716797072211264,
+      "learning_rate": 5.197496412603365e-07,
+      "loss": 0.3057,
+      "step": 1167
+    },
+    {
+      "epoch": 1.7969230769230768,
+      "grad_norm": 0.31516859702180683,
+      "learning_rate": 5.120877817590197e-07,
+      "loss": 0.2978,
+      "step": 1168
+    },
+    {
+      "epoch": 1.7984615384615386,
+      "grad_norm": 0.28828223920951385,
+      "learning_rate": 5.044813316627994e-07,
+      "loss": 0.2902,
+      "step": 1169
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.3139702765239355,
+      "learning_rate": 4.969303353933408e-07,
+      "loss": 0.2972,
+      "step": 1170
+    },
+    {
+      "epoch": 1.8015384615384615,
+      "grad_norm": 0.3166981530068666,
+      "learning_rate": 4.894348370484648e-07,
+      "loss": 0.2958,
+      "step": 1171
+    },
+    {
+      "epoch": 1.803076923076923,
+      "grad_norm": 0.28132094323432094,
+      "learning_rate": 4.819948804018771e-07,
+      "loss": 0.2939,
+      "step": 1172
+    },
+    {
+      "epoch": 1.8046153846153845,
+      "grad_norm": 0.28635840715466815,
+      "learning_rate": 4.746105089029229e-07,
+      "loss": 0.3033,
+      "step": 1173
+    },
+    {
+      "epoch": 1.8061538461538462,
+      "grad_norm": 0.2874134343844696,
+      "learning_rate": 4.6728176567633065e-07,
+      "loss": 0.2973,
+      "step": 1174
+    },
+    {
+      "epoch": 1.8076923076923077,
+      "grad_norm": 0.28915641288921606,
+      "learning_rate": 4.6000869352195607e-07,
+      "loss": 0.3037,
+      "step": 1175
+    },
+    {
+      "epoch": 1.8092307692307692,
+      "grad_norm": 0.28873978460490446,
+      "learning_rate": 4.5279133491454406e-07,
+      "loss": 0.3025,
+      "step": 1176
+    },
+    {
+      "epoch": 1.810769230769231,
+      "grad_norm": 0.2883150702012467,
+      "learning_rate": 4.4562973200346413e-07,
+      "loss": 0.309,
+      "step": 1177
+    },
+    {
+      "epoch": 1.8123076923076922,
+      "grad_norm": 0.2846566056818834,
+      "learning_rate": 4.385239266124752e-07,
+      "loss": 0.2922,
+      "step": 1178
+    },
+    {
+      "epoch": 1.8138461538461539,
+      "grad_norm": 0.29231876819507663,
+      "learning_rate": 4.314739602394791e-07,
+      "loss": 0.2981,
+      "step": 1179
+    },
+    {
+      "epoch": 1.8153846153846154,
+      "grad_norm": 0.29109111624604317,
+      "learning_rate": 4.2447987405628054e-07,
+      "loss": 0.3025,
+      "step": 1180
+    },
+    {
+      "epoch": 1.8169230769230769,
+      "grad_norm": 0.2895427581153819,
+      "learning_rate": 4.1754170890833777e-07,
+      "loss": 0.2873,
+      "step": 1181
+    },
+    {
+      "epoch": 1.8184615384615386,
+      "grad_norm": 0.3099472871182504,
+      "learning_rate": 4.106595053145357e-07,
+      "loss": 0.291,
+      "step": 1182
+    },
+    {
+      "epoch": 1.8199999999999998,
+      "grad_norm": 0.2844681037434324,
+      "learning_rate": 4.038333034669406e-07,
+      "loss": 0.2876,
+      "step": 1183
+    },
+    {
+      "epoch": 1.8215384615384616,
+      "grad_norm": 0.2941880091590556,
+      "learning_rate": 3.9706314323056936e-07,
+      "loss": 0.2981,
+      "step": 1184
+    },
+    {
+      "epoch": 1.823076923076923,
+      "grad_norm": 0.2773856619973786,
+      "learning_rate": 3.903490641431573e-07,
+      "loss": 0.3068,
+      "step": 1185
+    },
+    {
+      "epoch": 1.8246153846153845,
+      "grad_norm": 0.29007851821611963,
+      "learning_rate": 3.8369110541492396e-07,
+      "loss": 0.3016,
+      "step": 1186
+    },
+    {
+      "epoch": 1.8261538461538462,
+      "grad_norm": 0.2868719655853767,
+      "learning_rate": 3.770893059283465e-07,
+      "loss": 0.308,
+      "step": 1187
+    },
+    {
+      "epoch": 1.8276923076923077,
+      "grad_norm": 0.2806687124089037,
+      "learning_rate": 3.705437042379334e-07,
+      "loss": 0.291,
+      "step": 1188
+    },
+    {
+      "epoch": 1.8292307692307692,
+      "grad_norm": 0.2984253292754862,
+      "learning_rate": 3.6405433856999684e-07,
+      "loss": 0.3004,
+      "step": 1189
+    },
+    {
+      "epoch": 1.830769230769231,
+      "grad_norm": 0.29565855952776743,
+      "learning_rate": 3.5762124682242936e-07,
+      "loss": 0.2965,
+      "step": 1190
+    },
+    {
+      "epoch": 1.8323076923076922,
+      "grad_norm": 0.2936559870704328,
+      "learning_rate": 3.5124446656448654e-07,
+      "loss": 0.2881,
+      "step": 1191
+    },
+    {
+      "epoch": 1.833846153846154,
+      "grad_norm": 0.29198185579387714,
+      "learning_rate": 3.4492403503656236e-07,
+      "loss": 0.2968,
+      "step": 1192
+    },
+    {
+      "epoch": 1.8353846153846154,
+      "grad_norm": 0.2799051648611265,
+      "learning_rate": 3.3865998914997645e-07,
+      "loss": 0.2896,
+      "step": 1193
+    },
+    {
+      "epoch": 1.8369230769230769,
+      "grad_norm": 0.2948875289026568,
+      "learning_rate": 3.324523654867551e-07,
+      "loss": 0.3044,
+      "step": 1194
+    },
+    {
+      "epoch": 1.8384615384615386,
+      "grad_norm": 0.2870396431180434,
+      "learning_rate": 3.2630120029942034e-07,
+      "loss": 0.2967,
+      "step": 1195
+    },
+    {
+      "epoch": 1.8399999999999999,
+      "grad_norm": 0.28422327849635326,
+      "learning_rate": 3.2020652951077256e-07,
+      "loss": 0.2964,
+      "step": 1196
+    },
+    {
+      "epoch": 1.8415384615384616,
+      "grad_norm": 0.2843250331695583,
+      "learning_rate": 3.1416838871368925e-07,
+      "loss": 0.2903,
+      "step": 1197
+    },
+    {
+      "epoch": 1.843076923076923,
+      "grad_norm": 0.29045638459864764,
+      "learning_rate": 3.081868131709109e-07,
+      "loss": 0.2962,
+      "step": 1198
+    },
+    {
+      "epoch": 1.8446153846153845,
+      "grad_norm": 0.29778044383408825,
+      "learning_rate": 3.0226183781483897e-07,
+      "loss": 0.2948,
+      "step": 1199
+    },
+    {
+      "epoch": 1.8461538461538463,
+      "grad_norm": 0.29745832139833833,
+      "learning_rate": 2.963934972473259e-07,
+      "loss": 0.3042,
+      "step": 1200
+    },
+    {
+      "epoch": 1.8476923076923077,
+      "grad_norm": 0.27071527831325026,
+      "learning_rate": 2.905818257394799e-07,
+      "loss": 0.3023,
+      "step": 1201
+    },
+    {
+      "epoch": 1.8492307692307692,
+      "grad_norm": 0.2829295834534734,
+      "learning_rate": 2.848268572314616e-07,
+      "loss": 0.3026,
+      "step": 1202
+    },
+    {
+      "epoch": 1.8507692307692307,
+      "grad_norm": 0.2819444333760147,
+      "learning_rate": 2.791286253322856e-07,
+      "loss": 0.2855,
+      "step": 1203
+    },
+    {
+      "epoch": 1.8523076923076922,
+      "grad_norm": 0.29192348543413565,
+      "learning_rate": 2.734871633196246e-07,
+      "loss": 0.3047,
+      "step": 1204
+    },
+    {
+      "epoch": 1.853846153846154,
+      "grad_norm": 0.3040062200985359,
+      "learning_rate": 2.679025041396155e-07,
+      "loss": 0.2874,
+      "step": 1205
+    },
+    {
+      "epoch": 1.8553846153846154,
+      "grad_norm": 0.2906945754403228,
+      "learning_rate": 2.6237468040666515e-07,
+      "loss": 0.3023,
+      "step": 1206
+    },
+    {
+      "epoch": 1.856923076923077,
+      "grad_norm": 0.2774096205598689,
+      "learning_rate": 2.569037244032657e-07,
+      "loss": 0.2857,
+      "step": 1207
+    },
+    {
+      "epoch": 1.8584615384615386,
+      "grad_norm": 0.29437106511484507,
+      "learning_rate": 2.5148966807979733e-07,
+      "loss": 0.297,
+      "step": 1208
+    },
+    {
+      "epoch": 1.8599999999999999,
+      "grad_norm": 0.2817655480894342,
+      "learning_rate": 2.461325430543482e-07,
+      "loss": 0.2841,
+      "step": 1209
+    },
+    {
+      "epoch": 1.8615384615384616,
+      "grad_norm": 0.3019684935537153,
+      "learning_rate": 2.4083238061252565e-07,
+      "loss": 0.2894,
+      "step": 1210
+    },
+    {
+      "epoch": 1.863076923076923,
+      "grad_norm": 0.28758512946790726,
+      "learning_rate": 2.355892117072789e-07,
+      "loss": 0.2986,
+      "step": 1211
+    },
+    {
+      "epoch": 1.8646153846153846,
+      "grad_norm": 0.2834712454078908,
+      "learning_rate": 2.3040306695871319e-07,
+      "loss": 0.3003,
+      "step": 1212
+    },
+    {
+      "epoch": 1.8661538461538463,
+      "grad_norm": 0.2928744336917785,
+      "learning_rate": 2.2527397665391026e-07,
+      "loss": 0.3017,
+      "step": 1213
+    },
+    {
+      "epoch": 1.8676923076923075,
+      "grad_norm": 0.2722793316592535,
+      "learning_rate": 2.2020197074675952e-07,
+      "loss": 0.2769,
+      "step": 1214
+    },
+    {
+      "epoch": 1.8692307692307693,
+      "grad_norm": 0.30003418399195037,
+      "learning_rate": 2.1518707885777147e-07,
+      "loss": 0.301,
+      "step": 1215
+    },
+    {
+      "epoch": 1.8707692307692307,
+      "grad_norm": 0.27960941370533937,
+      "learning_rate": 2.1022933027391555e-07,
+      "loss": 0.3057,
+      "step": 1216
+    },
+    {
+      "epoch": 1.8723076923076922,
+      "grad_norm": 0.2869179241164619,
+      "learning_rate": 2.0532875394844053e-07,
+      "loss": 0.2905,
+      "step": 1217
+    },
+    {
+      "epoch": 1.873846153846154,
+      "grad_norm": 0.292207810108955,
+      "learning_rate": 2.0048537850071326e-07,
+      "loss": 0.2846,
+      "step": 1218
+    },
+    {
+      "epoch": 1.8753846153846154,
+      "grad_norm": 0.3075244833170072,
+      "learning_rate": 1.9569923221604224e-07,
+      "loss": 0.294,
+      "step": 1219
+    },
+    {
+      "epoch": 1.876923076923077,
+      "grad_norm": 0.2939909565303583,
+      "learning_rate": 1.909703430455223e-07,
+      "loss": 0.293,
+      "step": 1220
+    },
+    {
+      "epoch": 1.8784615384615386,
+      "grad_norm": 0.29480655255101934,
+      "learning_rate": 1.8629873860586567e-07,
+      "loss": 0.285,
+      "step": 1221
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.2935702658529837,
+      "learning_rate": 1.8168444617924107e-07,
+      "loss": 0.2951,
+      "step": 1222
+    },
+    {
+      "epoch": 1.8815384615384616,
+      "grad_norm": 0.29667447339509234,
+      "learning_rate": 1.7712749271311392e-07,
+      "loss": 0.2932,
+      "step": 1223
+    },
+    {
+      "epoch": 1.883076923076923,
+      "grad_norm": 0.2862920537783253,
+      "learning_rate": 1.7262790482009184e-07,
+      "loss": 0.2939,
+      "step": 1224
+    },
+    {
+      "epoch": 1.8846153846153846,
+      "grad_norm": 0.2810180438431519,
+      "learning_rate": 1.681857087777672e-07,
+      "loss": 0.2906,
+      "step": 1225
+    },
+    {
+      "epoch": 1.8861538461538463,
+      "grad_norm": 0.2891424309233074,
+      "learning_rate": 1.6380093052856482e-07,
+      "loss": 0.2934,
+      "step": 1226
+    },
+    {
+      "epoch": 1.8876923076923076,
+      "grad_norm": 0.28711438496084785,
+      "learning_rate": 1.5947359567958677e-07,
+      "loss": 0.2992,
+      "step": 1227
+    },
+    {
+      "epoch": 1.8892307692307693,
+      "grad_norm": 0.28838615803535156,
+      "learning_rate": 1.5520372950246888e-07,
+      "loss": 0.2947,
+      "step": 1228
+    },
+    {
+      "epoch": 1.8907692307692308,
+      "grad_norm": 0.29855832389108056,
+      "learning_rate": 1.5099135693322776e-07,
+      "loss": 0.2973,
+      "step": 1229
+    },
+    {
+      "epoch": 1.8923076923076922,
+      "grad_norm": 0.2816177472561755,
+      "learning_rate": 1.4683650257211967e-07,
+      "loss": 0.2919,
+      "step": 1230
+    },
+    {
+      "epoch": 1.893846153846154,
+      "grad_norm": 0.2851220861170624,
+      "learning_rate": 1.4273919068349184e-07,
+      "loss": 0.2924,
+      "step": 1231
+    },
+    {
+      "epoch": 1.8953846153846152,
+      "grad_norm": 0.29468012946037214,
+      "learning_rate": 1.386994451956436e-07,
+      "loss": 0.2906,
+      "step": 1232
+    },
+    {
+      "epoch": 1.896923076923077,
+      "grad_norm": 0.29928273386454685,
+      "learning_rate": 1.3471728970068986e-07,
+      "loss": 0.2939,
+      "step": 1233
+    },
+    {
+      "epoch": 1.8984615384615384,
+      "grad_norm": 0.29142659164043155,
+      "learning_rate": 1.3079274745441794e-07,
+      "loss": 0.2874,
+      "step": 1234
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.2752984862928927,
+      "learning_rate": 1.2692584137615205e-07,
+      "loss": 0.2939,
+      "step": 1235
+    },
+    {
+      "epoch": 1.9015384615384616,
+      "grad_norm": 0.2974512527197222,
+      "learning_rate": 1.231165940486234e-07,
+      "loss": 0.2941,
+      "step": 1236
+    },
+    {
+      "epoch": 1.903076923076923,
+      "grad_norm": 0.30378797478609865,
+      "learning_rate": 1.1936502771783488e-07,
+      "loss": 0.2955,
+      "step": 1237
+    },
+    {
+      "epoch": 1.9046153846153846,
+      "grad_norm": 0.28477397879781885,
+      "learning_rate": 1.1567116429293424e-07,
+      "loss": 0.3024,
+      "step": 1238
+    },
+    {
+      "epoch": 1.9061538461538463,
+      "grad_norm": 0.2985434368623264,
+      "learning_rate": 1.1203502534608113e-07,
+      "loss": 0.2887,
+      "step": 1239
+    },
+    {
+      "epoch": 1.9076923076923076,
+      "grad_norm": 0.28168974546129005,
+      "learning_rate": 1.0845663211232704e-07,
+      "loss": 0.2936,
+      "step": 1240
+    },
+    {
+      "epoch": 1.9092307692307693,
+      "grad_norm": 0.2948638535745993,
+      "learning_rate": 1.0493600548948879e-07,
+      "loss": 0.3047,
+      "step": 1241
+    },
+    {
+      "epoch": 1.9107692307692308,
+      "grad_norm": 0.2666238065946024,
+      "learning_rate": 1.0147316603802415e-07,
+      "loss": 0.2926,
+      "step": 1242
+    },
+    {
+      "epoch": 1.9123076923076923,
+      "grad_norm": 0.2858994691151879,
+      "learning_rate": 9.806813398091419e-08,
+      "loss": 0.2914,
+      "step": 1243
+    },
+    {
+      "epoch": 1.913846153846154,
+      "grad_norm": 0.275267029168629,
+      "learning_rate": 9.47209292035467e-08,
+      "loss": 0.2978,
+      "step": 1244
+    },
+    {
+      "epoch": 1.9153846153846152,
+      "grad_norm": 0.28734123984025134,
+      "learning_rate": 9.143157125359514e-08,
+      "loss": 0.296,
+      "step": 1245
+    },
+    {
+      "epoch": 1.916923076923077,
+      "grad_norm": 0.2756222757142002,
+      "learning_rate": 8.82000793409088e-08,
+      "loss": 0.3007,
+      "step": 1246
+    },
+    {
+      "epoch": 1.9184615384615384,
+      "grad_norm": 0.28565771596063383,
+      "learning_rate": 8.502647233740169e-08,
+      "loss": 0.2995,
+      "step": 1247
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.2975809575409654,
+      "learning_rate": 8.191076877693605e-08,
+      "loss": 0.2962,
+      "step": 1248
+    },
+    {
+      "epoch": 1.9215384615384616,
+      "grad_norm": 0.2925530387252898,
+      "learning_rate": 7.885298685522235e-08,
+      "loss": 0.2956,
+      "step": 1249
+    },
+    {
+      "epoch": 1.9230769230769231,
+      "grad_norm": 0.312559325042889,
+      "learning_rate": 7.585314442970282e-08,
+      "loss": 0.2931,
+      "step": 1250
+    },
+    {
+      "epoch": 1.9246153846153846,
+      "grad_norm": 0.29566448870635603,
+      "learning_rate": 7.291125901946027e-08,
+      "loss": 0.2923,
+      "step": 1251
+    },
+    {
+      "epoch": 1.926153846153846,
+      "grad_norm": 0.29162174985850026,
+      "learning_rate": 7.002734780510279e-08,
+      "loss": 0.2955,
+      "step": 1252
+    },
+    {
+      "epoch": 1.9276923076923076,
+      "grad_norm": 0.2990373715685646,
+      "learning_rate": 6.720142762867032e-08,
+      "loss": 0.297,
+      "step": 1253
+    },
+    {
+      "epoch": 1.9292307692307693,
+      "grad_norm": 0.2922918111620488,
+      "learning_rate": 6.443351499353823e-08,
+      "loss": 0.2954,
+      "step": 1254
+    },
+    {
+      "epoch": 1.9307692307692308,
+      "grad_norm": 0.2896544639031337,
+      "learning_rate": 6.172362606431281e-08,
+      "loss": 0.2959,
+      "step": 1255
+    },
+    {
+      "epoch": 1.9323076923076923,
+      "grad_norm": 0.27803033460355636,
+      "learning_rate": 5.907177666674813e-08,
+      "loss": 0.3076,
+      "step": 1256
+    },
+    {
+      "epoch": 1.933846153846154,
+      "grad_norm": 0.2961598876438012,
+      "learning_rate": 5.647798228764156e-08,
+      "loss": 0.2925,
+      "step": 1257
+    },
+    {
+      "epoch": 1.9353846153846153,
+      "grad_norm": 0.2867619417977115,
+      "learning_rate": 5.394225807475284e-08,
+      "loss": 0.2895,
+      "step": 1258
+    },
+    {
+      "epoch": 1.936923076923077,
+      "grad_norm": 0.28377361680406005,
+      "learning_rate": 5.146461883671072e-08,
+      "loss": 0.2863,
+      "step": 1259
+    },
+    {
+      "epoch": 1.9384615384615385,
+      "grad_norm": 0.28669282539245955,
+      "learning_rate": 4.9045079042926434e-08,
+      "loss": 0.3041,
+      "step": 1260
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.2605785356032334,
+      "learning_rate": 4.6683652823513725e-08,
+      "loss": 0.3067,
+      "step": 1261
+    },
+    {
+      "epoch": 1.9415384615384617,
+      "grad_norm": 0.2845207812409248,
+      "learning_rate": 4.438035396920004e-08,
+      "loss": 0.2969,
+      "step": 1262
+    },
+    {
+      "epoch": 1.943076923076923,
+      "grad_norm": 0.2864341128215156,
+      "learning_rate": 4.2135195931249925e-08,
+      "loss": 0.2924,
+      "step": 1263
+    },
+    {
+      "epoch": 1.9446153846153846,
+      "grad_norm": 0.3045274166513475,
+      "learning_rate": 3.9948191821386205e-08,
+      "loss": 0.2878,
+      "step": 1264
+    },
+    {
+      "epoch": 1.9461538461538461,
+      "grad_norm": 0.29656148896716283,
+      "learning_rate": 3.7819354411713364e-08,
+      "loss": 0.2968,
+      "step": 1265
+    },
+    {
+      "epoch": 1.9476923076923076,
+      "grad_norm": 0.27150633586112766,
+      "learning_rate": 3.5748696134639825e-08,
+      "loss": 0.2889,
+      "step": 1266
+    },
+    {
+      "epoch": 1.9492307692307693,
+      "grad_norm": 0.2703843147413634,
+      "learning_rate": 3.373622908280916e-08,
+      "loss": 0.2913,
+      "step": 1267
+    },
+    {
+      "epoch": 1.9507692307692308,
+      "grad_norm": 0.2921919887496798,
+      "learning_rate": 3.178196500903008e-08,
+      "loss": 0.2945,
+      "step": 1268
+    },
+    {
+      "epoch": 1.9523076923076923,
+      "grad_norm": 0.2984122929536049,
+      "learning_rate": 2.988591532620322e-08,
+      "loss": 0.2851,
+      "step": 1269
+    },
+    {
+      "epoch": 1.953846153846154,
+      "grad_norm": 0.2869450373203219,
+      "learning_rate": 2.8048091107258925e-08,
+      "loss": 0.3013,
+      "step": 1270
+    },
+    {
+      "epoch": 1.9553846153846153,
+      "grad_norm": 0.2872567599385607,
+      "learning_rate": 2.6268503085089547e-08,
+      "loss": 0.2977,
+      "step": 1271
+    },
+    {
+      "epoch": 1.956923076923077,
+      "grad_norm": 0.2970826171524059,
+      "learning_rate": 2.4547161652488382e-08,
+      "loss": 0.3035,
+      "step": 1272
+    },
+    {
+      "epoch": 1.9584615384615385,
+      "grad_norm": 0.27476566148375065,
+      "learning_rate": 2.2884076862089712e-08,
+      "loss": 0.3008,
+      "step": 1273
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.28642867711108994,
+      "learning_rate": 2.1279258426308848e-08,
+      "loss": 0.2838,
+      "step": 1274
+    },
+    {
+      "epoch": 1.9615384615384617,
+      "grad_norm": 0.29752164693394895,
+      "learning_rate": 1.973271571728441e-08,
+      "loss": 0.2902,
+      "step": 1275
+    },
+    {
+      "epoch": 1.963076923076923,
+      "grad_norm": 0.30159234323348555,
+      "learning_rate": 1.824445776682504e-08,
+      "loss": 0.286,
+      "step": 1276
+    },
+    {
+      "epoch": 1.9646153846153847,
+      "grad_norm": 0.2986365051624472,
+      "learning_rate": 1.6814493266357202e-08,
+      "loss": 0.2946,
+      "step": 1277
+    },
+    {
+      "epoch": 1.9661538461538461,
+      "grad_norm": 0.2927008764556385,
+      "learning_rate": 1.5442830566874123e-08,
+      "loss": 0.2962,
+      "step": 1278
+    },
+    {
+      "epoch": 1.9676923076923076,
+      "grad_norm": 0.31794811909537735,
+      "learning_rate": 1.4129477678884728e-08,
+      "loss": 0.2969,
+      "step": 1279
+    },
+    {
+      "epoch": 1.9692307692307693,
+      "grad_norm": 0.29178996440141997,
+      "learning_rate": 1.2874442272369225e-08,
+      "loss": 0.2999,
+      "step": 1280
+    },
+    {
+      "epoch": 1.9707692307692306,
+      "grad_norm": 0.30083640729923533,
+      "learning_rate": 1.1677731676733584e-08,
+      "loss": 0.2909,
+      "step": 1281
+    },
+    {
+      "epoch": 1.9723076923076923,
+      "grad_norm": 0.27825233353564915,
+      "learning_rate": 1.053935288076957e-08,
+      "loss": 0.3009,
+      "step": 1282
+    },
+    {
+      "epoch": 1.9738461538461538,
+      "grad_norm": 0.29213371151213485,
+      "learning_rate": 9.459312532608122e-09,
+      "loss": 0.29,
+      "step": 1283
+    },
+    {
+      "epoch": 1.9753846153846153,
+      "grad_norm": 0.2877419396877038,
+      "learning_rate": 8.437616939683812e-09,
+      "loss": 0.292,
+      "step": 1284
+    },
+    {
+      "epoch": 1.976923076923077,
+      "grad_norm": 0.2877343885263764,
+      "learning_rate": 7.474272068698219e-09,
+      "loss": 0.3004,
+      "step": 1285
+    },
+    {
+      "epoch": 1.9784615384615385,
+      "grad_norm": 0.2817396722101799,
+      "learning_rate": 6.569283545587724e-09,
+      "loss": 0.294,
+      "step": 1286
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.2939115978628203,
+      "learning_rate": 5.722656655482439e-09,
+      "loss": 0.2789,
+      "step": 1287
+    },
+    {
+      "epoch": 1.9815384615384617,
+      "grad_norm": 0.2915889225812978,
+      "learning_rate": 4.9343963426840006e-09,
+      "loss": 0.2979,
+      "step": 1288
+    },
+    {
+      "epoch": 1.983076923076923,
+      "grad_norm": 0.27718339498875494,
+      "learning_rate": 4.204507210633368e-09,
+      "loss": 0.3054,
+      "step": 1289
+    },
+    {
+      "epoch": 1.9846153846153847,
+      "grad_norm": 0.2921513255474594,
+      "learning_rate": 3.5329935218819668e-09,
+      "loss": 0.3024,
+      "step": 1290
+    },
+    {
+      "epoch": 1.9861538461538462,
+      "grad_norm": 0.28112548556461614,
+      "learning_rate": 2.9198591980705847e-09,
+      "loss": 0.3041,
+      "step": 1291
+    },
+    {
+      "epoch": 1.9876923076923076,
+      "grad_norm": 0.28438961891728476,
+      "learning_rate": 2.3651078199016244e-09,
+      "loss": 0.2925,
+      "step": 1292
+    },
+    {
+      "epoch": 1.9892307692307694,
+      "grad_norm": 0.2811692715727547,
+      "learning_rate": 1.8687426271246646e-09,
+      "loss": 0.2945,
+      "step": 1293
+    },
+    {
+      "epoch": 1.9907692307692306,
+      "grad_norm": 0.2841074561924985,
+      "learning_rate": 1.430766518512039e-09,
+      "loss": 0.2935,
+      "step": 1294
+    },
+    {
+      "epoch": 1.9923076923076923,
+      "grad_norm": 0.29705816944075764,
+      "learning_rate": 1.0511820518432915e-09,
+      "loss": 0.2854,
+      "step": 1295
+    },
+    {
+      "epoch": 1.9938461538461538,
+      "grad_norm": 0.2919224539526079,
+      "learning_rate": 7.299914438929634e-10,
+      "loss": 0.2842,
+      "step": 1296
+    },
+    {
+      "epoch": 1.9953846153846153,
+      "grad_norm": 0.2970632425826639,
+      "learning_rate": 4.671965704128312e-10,
+      "loss": 0.3009,
+      "step": 1297
+    },
+    {
+      "epoch": 1.996923076923077,
+      "grad_norm": 0.2748541619910348,
+      "learning_rate": 2.6279896612524393e-10,
+      "loss": 0.2994,
+      "step": 1298
+    },
+    {
+      "epoch": 1.9984615384615385,
+      "grad_norm": 0.2949681819269235,
+      "learning_rate": 1.167998247131319e-10,
+      "loss": 0.2899,
+      "step": 1299
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.2973447897180648,
+      "learning_rate": 2.919999881001445e-11,
+      "loss": 0.2891,
+      "step": 1300
+    },
+    {
+      "epoch": 2.0,
+      "step": 1300,
+      "total_flos": 749986945269760.0,
+      "train_loss": 0.3381902449864608,
+      "train_runtime": 28530.4364,
+      "train_samples_per_second": 5.832,
+      "train_steps_per_second": 0.046
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1300,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 3000000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 749986945269760.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}