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": 5.063921474779319,
+      "learning_rate": 2e-05,
+      "loss": 0.9605,
+      "step": 1
+    },
+    {
+      "epoch": 0.003076923076923077,
+      "grad_norm": 3.369623660244633,
+      "learning_rate": 1.999997080000119e-05,
+      "loss": 0.8245,
+      "step": 2
+    },
+    {
+      "epoch": 0.004615384615384616,
+      "grad_norm": 2.686003070114147,
+      "learning_rate": 1.9999883200175286e-05,
+      "loss": 0.7138,
+      "step": 3
+    },
+    {
+      "epoch": 0.006153846153846154,
+      "grad_norm": 1.5234580549826773,
+      "learning_rate": 1.9999737201033877e-05,
+      "loss": 0.6496,
+      "step": 4
+    },
+    {
+      "epoch": 0.007692307692307693,
+      "grad_norm": 2.609918856467969,
+      "learning_rate": 1.999953280342959e-05,
+      "loss": 0.6476,
+      "step": 5
+    },
+    {
+      "epoch": 0.009230769230769232,
+      "grad_norm": 1.6701564603819716,
+      "learning_rate": 1.9999270008556108e-05,
+      "loss": 0.6071,
+      "step": 6
+    },
+    {
+      "epoch": 0.010769230769230769,
+      "grad_norm": 0.9106505494488014,
+      "learning_rate": 1.9998948817948157e-05,
+      "loss": 0.545,
+      "step": 7
+    },
+    {
+      "epoch": 0.012307692307692308,
+      "grad_norm": 1.018082854225675,
+      "learning_rate": 1.999856923348149e-05,
+      "loss": 0.5326,
+      "step": 8
+    },
+    {
+      "epoch": 0.013846153846153847,
+      "grad_norm": 0.7909606612418608,
+      "learning_rate": 1.9998131257372878e-05,
+      "loss": 0.524,
+      "step": 9
+    },
+    {
+      "epoch": 0.015384615384615385,
+      "grad_norm": 0.7555909222636807,
+      "learning_rate": 1.99976348921801e-05,
+      "loss": 0.5163,
+      "step": 10
+    },
+    {
+      "epoch": 0.016923076923076923,
+      "grad_norm": 0.7450235308921992,
+      "learning_rate": 1.9997080140801932e-05,
+      "loss": 0.5088,
+      "step": 11
+    },
+    {
+      "epoch": 0.018461538461538463,
+      "grad_norm": 0.7134021367422554,
+      "learning_rate": 1.999646700647812e-05,
+      "loss": 0.4993,
+      "step": 12
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5759662130624531,
+      "learning_rate": 1.9995795492789368e-05,
+      "loss": 0.4851,
+      "step": 13
+    },
+    {
+      "epoch": 0.021538461538461538,
+      "grad_norm": 0.5606778035425145,
+      "learning_rate": 1.9995065603657317e-05,
+      "loss": 0.4666,
+      "step": 14
+    },
+    {
+      "epoch": 0.023076923076923078,
+      "grad_norm": 0.5398744265890032,
+      "learning_rate": 1.999427734334452e-05,
+      "loss": 0.4691,
+      "step": 15
+    },
+    {
+      "epoch": 0.024615384615384615,
+      "grad_norm": 0.5156262642068832,
+      "learning_rate": 1.9993430716454415e-05,
+      "loss": 0.4605,
+      "step": 16
+    },
+    {
+      "epoch": 0.026153846153846153,
+      "grad_norm": 0.5268519764536594,
+      "learning_rate": 1.9992525727931303e-05,
+      "loss": 0.4564,
+      "step": 17
+    },
+    {
+      "epoch": 0.027692307692307693,
+      "grad_norm": 0.5454016980521769,
+      "learning_rate": 1.9991562383060316e-05,
+      "loss": 0.4659,
+      "step": 18
+    },
+    {
+      "epoch": 0.02923076923076923,
+      "grad_norm": 0.4816040745018356,
+      "learning_rate": 1.9990540687467394e-05,
+      "loss": 0.4565,
+      "step": 19
+    },
+    {
+      "epoch": 0.03076923076923077,
+      "grad_norm": 0.4590572903497101,
+      "learning_rate": 1.9989460647119232e-05,
+      "loss": 0.4407,
+      "step": 20
+    },
+    {
+      "epoch": 0.03230769230769231,
+      "grad_norm": 0.4226998877202937,
+      "learning_rate": 1.998832226832327e-05,
+      "loss": 0.4373,
+      "step": 21
+    },
+    {
+      "epoch": 0.033846153846153845,
+      "grad_norm": 0.4536363868403153,
+      "learning_rate": 1.9987125557727633e-05,
+      "loss": 0.4313,
+      "step": 22
+    },
+    {
+      "epoch": 0.03538461538461538,
+      "grad_norm": 0.42868619210170616,
+      "learning_rate": 1.9985870522321118e-05,
+      "loss": 0.4553,
+      "step": 23
+    },
+    {
+      "epoch": 0.036923076923076927,
+      "grad_norm": 0.45896932342495994,
+      "learning_rate": 1.9984557169433126e-05,
+      "loss": 0.441,
+      "step": 24
+    },
+    {
+      "epoch": 0.038461538461538464,
+      "grad_norm": 0.4294926658833097,
+      "learning_rate": 1.9983185506733643e-05,
+      "loss": 0.4318,
+      "step": 25
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.44540317006888985,
+      "learning_rate": 1.9981755542233175e-05,
+      "loss": 0.4308,
+      "step": 26
+    },
+    {
+      "epoch": 0.04153846153846154,
+      "grad_norm": 0.4559389600389272,
+      "learning_rate": 1.9980267284282718e-05,
+      "loss": 0.4278,
+      "step": 27
+    },
+    {
+      "epoch": 0.043076923076923075,
+      "grad_norm": 0.37690300997098264,
+      "learning_rate": 1.9978720741573693e-05,
+      "loss": 0.4405,
+      "step": 28
+    },
+    {
+      "epoch": 0.04461538461538461,
+      "grad_norm": 0.439464993843592,
+      "learning_rate": 1.9977115923137912e-05,
+      "loss": 0.4387,
+      "step": 29
+    },
+    {
+      "epoch": 0.046153846153846156,
+      "grad_norm": 0.43264990079260623,
+      "learning_rate": 1.9975452838347513e-05,
+      "loss": 0.4362,
+      "step": 30
+    },
+    {
+      "epoch": 0.047692307692307694,
+      "grad_norm": 0.4644507012498486,
+      "learning_rate": 1.9973731496914914e-05,
+      "loss": 0.4179,
+      "step": 31
+    },
+    {
+      "epoch": 0.04923076923076923,
+      "grad_norm": 0.4572720391818459,
+      "learning_rate": 1.9971951908892743e-05,
+      "loss": 0.4092,
+      "step": 32
+    },
+    {
+      "epoch": 0.05076923076923077,
+      "grad_norm": 0.46432612449175953,
+      "learning_rate": 1.9970114084673796e-05,
+      "loss": 0.4146,
+      "step": 33
+    },
+    {
+      "epoch": 0.052307692307692305,
+      "grad_norm": 0.4129940856967706,
+      "learning_rate": 1.996821803499097e-05,
+      "loss": 0.4297,
+      "step": 34
+    },
+    {
+      "epoch": 0.05384615384615385,
+      "grad_norm": 0.4042181784594487,
+      "learning_rate": 1.9966263770917192e-05,
+      "loss": 0.4112,
+      "step": 35
+    },
+    {
+      "epoch": 0.055384615384615386,
+      "grad_norm": 0.3906521043350921,
+      "learning_rate": 1.9964251303865362e-05,
+      "loss": 0.4161,
+      "step": 36
+    },
+    {
+      "epoch": 0.05692307692307692,
+      "grad_norm": 0.4418884612703507,
+      "learning_rate": 1.996218064558829e-05,
+      "loss": 0.4101,
+      "step": 37
+    },
+    {
+      "epoch": 0.05846153846153846,
+      "grad_norm": 0.42008732000288845,
+      "learning_rate": 1.9960051808178616e-05,
+      "loss": 0.4043,
+      "step": 38
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.43296733708514057,
+      "learning_rate": 1.9957864804068752e-05,
+      "loss": 0.4128,
+      "step": 39
+    },
+    {
+      "epoch": 0.06153846153846154,
+      "grad_norm": 0.45169856277574605,
+      "learning_rate": 1.99556196460308e-05,
+      "loss": 0.3995,
+      "step": 40
+    },
+    {
+      "epoch": 0.06307692307692307,
+      "grad_norm": 0.4164452322138832,
+      "learning_rate": 1.995331634717649e-05,
+      "loss": 0.4112,
+      "step": 41
+    },
+    {
+      "epoch": 0.06461538461538462,
+      "grad_norm": 0.4099581368464527,
+      "learning_rate": 1.9950954920957074e-05,
+      "loss": 0.4119,
+      "step": 42
+    },
+    {
+      "epoch": 0.06615384615384616,
+      "grad_norm": 0.44696899014004116,
+      "learning_rate": 1.994853538116329e-05,
+      "loss": 0.4146,
+      "step": 43
+    },
+    {
+      "epoch": 0.06769230769230769,
+      "grad_norm": 0.4055013281834995,
+      "learning_rate": 1.994605774192525e-05,
+      "loss": 0.4093,
+      "step": 44
+    },
+    {
+      "epoch": 0.06923076923076923,
+      "grad_norm": 0.40395767432429797,
+      "learning_rate": 1.994352201771236e-05,
+      "loss": 0.4063,
+      "step": 45
+    },
+    {
+      "epoch": 0.07076923076923076,
+      "grad_norm": 0.4116346989910219,
+      "learning_rate": 1.9940928223333254e-05,
+      "loss": 0.393,
+      "step": 46
+    },
+    {
+      "epoch": 0.07230769230769231,
+      "grad_norm": 0.3591126773969262,
+      "learning_rate": 1.9938276373935688e-05,
+      "loss": 0.4039,
+      "step": 47
+    },
+    {
+      "epoch": 0.07384615384615385,
+      "grad_norm": 0.46456735467456856,
+      "learning_rate": 1.9935566485006464e-05,
+      "loss": 0.3851,
+      "step": 48
+    },
+    {
+      "epoch": 0.07538461538461538,
+      "grad_norm": 0.4268493349476963,
+      "learning_rate": 1.993279857237133e-05,
+      "loss": 0.4044,
+      "step": 49
+    },
+    {
+      "epoch": 0.07692307692307693,
+      "grad_norm": 0.38063003139174306,
+      "learning_rate": 1.99299726521949e-05,
+      "loss": 0.404,
+      "step": 50
+    },
+    {
+      "epoch": 0.07846153846153846,
+      "grad_norm": 0.4497110041221097,
+      "learning_rate": 1.992708874098054e-05,
+      "loss": 0.41,
+      "step": 51
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4157742963507939,
+      "learning_rate": 1.9924146855570298e-05,
+      "loss": 0.3941,
+      "step": 52
+    },
+    {
+      "epoch": 0.08153846153846153,
+      "grad_norm": 0.4414354437213974,
+      "learning_rate": 1.9921147013144782e-05,
+      "loss": 0.3993,
+      "step": 53
+    },
+    {
+      "epoch": 0.08307692307692308,
+      "grad_norm": 0.43571701128684,
+      "learning_rate": 1.9918089231223066e-05,
+      "loss": 0.3959,
+      "step": 54
+    },
+    {
+      "epoch": 0.08461538461538462,
+      "grad_norm": 0.4371043287551301,
+      "learning_rate": 1.99149735276626e-05,
+      "loss": 0.3986,
+      "step": 55
+    },
+    {
+      "epoch": 0.08615384615384615,
+      "grad_norm": 0.4137823130711333,
+      "learning_rate": 1.9911799920659093e-05,
+      "loss": 0.3891,
+      "step": 56
+    },
+    {
+      "epoch": 0.0876923076923077,
+      "grad_norm": 0.43271326063346954,
+      "learning_rate": 1.9908568428746408e-05,
+      "loss": 0.408,
+      "step": 57
+    },
+    {
+      "epoch": 0.08923076923076922,
+      "grad_norm": 0.3809934064411492,
+      "learning_rate": 1.9905279070796454e-05,
+      "loss": 0.3983,
+      "step": 58
+    },
+    {
+      "epoch": 0.09076923076923077,
+      "grad_norm": 0.4431656521313663,
+      "learning_rate": 1.9901931866019087e-05,
+      "loss": 0.3915,
+      "step": 59
+    },
+    {
+      "epoch": 0.09230769230769231,
+      "grad_norm": 0.4382895852653331,
+      "learning_rate": 1.989852683396198e-05,
+      "loss": 0.3896,
+      "step": 60
+    },
+    {
+      "epoch": 0.09384615384615384,
+      "grad_norm": 0.43423688769031565,
+      "learning_rate": 1.9895063994510512e-05,
+      "loss": 0.4006,
+      "step": 61
+    },
+    {
+      "epoch": 0.09538461538461539,
+      "grad_norm": 0.4564712962684036,
+      "learning_rate": 1.9891543367887675e-05,
+      "loss": 0.3936,
+      "step": 62
+    },
+    {
+      "epoch": 0.09692307692307692,
+      "grad_norm": 0.4079034300130273,
+      "learning_rate": 1.988796497465392e-05,
+      "loss": 0.4032,
+      "step": 63
+    },
+    {
+      "epoch": 0.09846153846153846,
+      "grad_norm": 0.4405646045189855,
+      "learning_rate": 1.988432883570707e-05,
+      "loss": 0.3913,
+      "step": 64
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.41717807318047573,
+      "learning_rate": 1.9880634972282168e-05,
+      "loss": 0.3876,
+      "step": 65
+    },
+    {
+      "epoch": 0.10153846153846154,
+      "grad_norm": 0.4684874071170093,
+      "learning_rate": 1.9876883405951378e-05,
+      "loss": 0.3989,
+      "step": 66
+    },
+    {
+      "epoch": 0.10307692307692308,
+      "grad_norm": 0.41741113705795535,
+      "learning_rate": 1.987307415862385e-05,
+      "loss": 0.3853,
+      "step": 67
+    },
+    {
+      "epoch": 0.10461538461538461,
+      "grad_norm": 0.42913698621302665,
+      "learning_rate": 1.9869207252545582e-05,
+      "loss": 0.389,
+      "step": 68
+    },
+    {
+      "epoch": 0.10615384615384615,
+      "grad_norm": 0.4135673359860317,
+      "learning_rate": 1.986528271029931e-05,
+      "loss": 0.3991,
+      "step": 69
+    },
+    {
+      "epoch": 0.1076923076923077,
+      "grad_norm": 0.3824388045772453,
+      "learning_rate": 1.9861300554804357e-05,
+      "loss": 0.3862,
+      "step": 70
+    },
+    {
+      "epoch": 0.10923076923076923,
+      "grad_norm": 0.4088649251683559,
+      "learning_rate": 1.985726080931651e-05,
+      "loss": 0.3861,
+      "step": 71
+    },
+    {
+      "epoch": 0.11076923076923077,
+      "grad_norm": 0.41188429042311264,
+      "learning_rate": 1.9853163497427885e-05,
+      "loss": 0.3933,
+      "step": 72
+    },
+    {
+      "epoch": 0.1123076923076923,
+      "grad_norm": 0.4042211920009733,
+      "learning_rate": 1.9849008643066774e-05,
+      "loss": 0.3801,
+      "step": 73
+    },
+    {
+      "epoch": 0.11384615384615385,
+      "grad_norm": 0.4349260424024348,
+      "learning_rate": 1.984479627049753e-05,
+      "loss": 0.3972,
+      "step": 74
+    },
+    {
+      "epoch": 0.11538461538461539,
+      "grad_norm": 0.42179852404385665,
+      "learning_rate": 1.9840526404320415e-05,
+      "loss": 0.3814,
+      "step": 75
+    },
+    {
+      "epoch": 0.11692307692307692,
+      "grad_norm": 0.4361915252579203,
+      "learning_rate": 1.983619906947144e-05,
+      "loss": 0.3682,
+      "step": 76
+    },
+    {
+      "epoch": 0.11846153846153847,
+      "grad_norm": 0.41443723031355584,
+      "learning_rate": 1.9831814291222233e-05,
+      "loss": 0.3974,
+      "step": 77
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4713684797848488,
+      "learning_rate": 1.982737209517991e-05,
+      "loss": 0.3853,
+      "step": 78
+    },
+    {
+      "epoch": 0.12153846153846154,
+      "grad_norm": 0.4319359759319244,
+      "learning_rate": 1.982287250728689e-05,
+      "loss": 0.382,
+      "step": 79
+    },
+    {
+      "epoch": 0.12307692307692308,
+      "grad_norm": 0.46173603703218374,
+      "learning_rate": 1.981831555382076e-05,
+      "loss": 0.3986,
+      "step": 80
+    },
+    {
+      "epoch": 0.12461538461538461,
+      "grad_norm": 0.47085778692524854,
+      "learning_rate": 1.9813701261394136e-05,
+      "loss": 0.3787,
+      "step": 81
+    },
+    {
+      "epoch": 0.12615384615384614,
+      "grad_norm": 0.48130879868276283,
+      "learning_rate": 1.980902965695448e-05,
+      "loss": 0.368,
+      "step": 82
+    },
+    {
+      "epoch": 0.1276923076923077,
+      "grad_norm": 0.4375676704104233,
+      "learning_rate": 1.9804300767783958e-05,
+      "loss": 0.3889,
+      "step": 83
+    },
+    {
+      "epoch": 0.12923076923076923,
+      "grad_norm": 0.47219005186638796,
+      "learning_rate": 1.979951462149929e-05,
+      "loss": 0.3875,
+      "step": 84
+    },
+    {
+      "epoch": 0.13076923076923078,
+      "grad_norm": 0.3831973173530425,
+      "learning_rate": 1.979467124605156e-05,
+      "loss": 0.3856,
+      "step": 85
+    },
+    {
+      "epoch": 0.13230769230769232,
+      "grad_norm": 0.4442260067203606,
+      "learning_rate": 1.9789770669726088e-05,
+      "loss": 0.3888,
+      "step": 86
+    },
+    {
+      "epoch": 0.13384615384615384,
+      "grad_norm": 0.5019617632101833,
+      "learning_rate": 1.9784812921142232e-05,
+      "loss": 0.3828,
+      "step": 87
+    },
+    {
+      "epoch": 0.13538461538461538,
+      "grad_norm": 0.43270240554403,
+      "learning_rate": 1.977979802925324e-05,
+      "loss": 0.3918,
+      "step": 88
+    },
+    {
+      "epoch": 0.13692307692307693,
+      "grad_norm": 0.5121748308328352,
+      "learning_rate": 1.977472602334609e-05,
+      "loss": 0.3842,
+      "step": 89
+    },
+    {
+      "epoch": 0.13846153846153847,
+      "grad_norm": 0.4285419146768387,
+      "learning_rate": 1.976959693304129e-05,
+      "loss": 0.3646,
+      "step": 90
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4443754327929979,
+      "learning_rate": 1.9764410788292724e-05,
+      "loss": 0.3839,
+      "step": 91
+    },
+    {
+      "epoch": 0.14153846153846153,
+      "grad_norm": 0.44335647718942606,
+      "learning_rate": 1.9759167619387474e-05,
+      "loss": 0.3755,
+      "step": 92
+    },
+    {
+      "epoch": 0.14307692307692307,
+      "grad_norm": 0.4485326069844085,
+      "learning_rate": 1.9753867456945653e-05,
+      "loss": 0.353,
+      "step": 93
+    },
+    {
+      "epoch": 0.14461538461538462,
+      "grad_norm": 0.41159403443114567,
+      "learning_rate": 1.9748510331920204e-05,
+      "loss": 0.3861,
+      "step": 94
+    },
+    {
+      "epoch": 0.14615384615384616,
+      "grad_norm": 0.4507383444906886,
+      "learning_rate": 1.9743096275596735e-05,
+      "loss": 0.3942,
+      "step": 95
+    },
+    {
+      "epoch": 0.1476923076923077,
+      "grad_norm": 0.39621991179394783,
+      "learning_rate": 1.9737625319593338e-05,
+      "loss": 0.3891,
+      "step": 96
+    },
+    {
+      "epoch": 0.14923076923076922,
+      "grad_norm": 0.40824533675854696,
+      "learning_rate": 1.9732097495860388e-05,
+      "loss": 0.3773,
+      "step": 97
+    },
+    {
+      "epoch": 0.15076923076923077,
+      "grad_norm": 0.4036362707197561,
+      "learning_rate": 1.972651283668038e-05,
+      "loss": 0.3673,
+      "step": 98
+    },
+    {
+      "epoch": 0.1523076923076923,
+      "grad_norm": 0.39196000792144625,
+      "learning_rate": 1.9720871374667714e-05,
+      "loss": 0.3742,
+      "step": 99
+    },
+    {
+      "epoch": 0.15384615384615385,
+      "grad_norm": 0.45342715962137886,
+      "learning_rate": 1.971517314276854e-05,
+      "loss": 0.3843,
+      "step": 100
+    },
+    {
+      "epoch": 0.15538461538461537,
+      "grad_norm": 0.39932262509282124,
+      "learning_rate": 1.9709418174260523e-05,
+      "loss": 0.3869,
+      "step": 101
+    },
+    {
+      "epoch": 0.15692307692307692,
+      "grad_norm": 0.37268042255848166,
+      "learning_rate": 1.9703606502752674e-05,
+      "loss": 0.3831,
+      "step": 102
+    },
+    {
+      "epoch": 0.15846153846153846,
+      "grad_norm": 0.3997668194053964,
+      "learning_rate": 1.9697738162185163e-05,
+      "loss": 0.3755,
+      "step": 103
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.40502233507726826,
+      "learning_rate": 1.969181318682909e-05,
+      "loss": 0.3764,
+      "step": 104
+    },
+    {
+      "epoch": 0.16153846153846155,
+      "grad_norm": 0.439443980401936,
+      "learning_rate": 1.9685831611286312e-05,
+      "loss": 0.374,
+      "step": 105
+    },
+    {
+      "epoch": 0.16307692307692306,
+      "grad_norm": 0.40103661213841674,
+      "learning_rate": 1.967979347048923e-05,
+      "loss": 0.3806,
+      "step": 106
+    },
+    {
+      "epoch": 0.1646153846153846,
+      "grad_norm": 0.4211147893705695,
+      "learning_rate": 1.9673698799700582e-05,
+      "loss": 0.3839,
+      "step": 107
+    },
+    {
+      "epoch": 0.16615384615384615,
+      "grad_norm": 0.38205382927654796,
+      "learning_rate": 1.9667547634513248e-05,
+      "loss": 0.3647,
+      "step": 108
+    },
+    {
+      "epoch": 0.1676923076923077,
+      "grad_norm": 0.37836543863414746,
+      "learning_rate": 1.9661340010850025e-05,
+      "loss": 0.361,
+      "step": 109
+    },
+    {
+      "epoch": 0.16923076923076924,
+      "grad_norm": 0.41111158527489766,
+      "learning_rate": 1.9655075964963443e-05,
+      "loss": 0.3766,
+      "step": 110
+    },
+    {
+      "epoch": 0.17076923076923076,
+      "grad_norm": 0.3885346262736236,
+      "learning_rate": 1.9648755533435517e-05,
+      "loss": 0.3845,
+      "step": 111
+    },
+    {
+      "epoch": 0.1723076923076923,
+      "grad_norm": 0.4039456045891746,
+      "learning_rate": 1.9642378753177573e-05,
+      "loss": 0.3715,
+      "step": 112
+    },
+    {
+      "epoch": 0.17384615384615384,
+      "grad_norm": 0.3951420777639365,
+      "learning_rate": 1.9635945661430006e-05,
+      "loss": 0.3753,
+      "step": 113
+    },
+    {
+      "epoch": 0.1753846153846154,
+      "grad_norm": 0.4173309904873508,
+      "learning_rate": 1.9629456295762067e-05,
+      "loss": 0.3725,
+      "step": 114
+    },
+    {
+      "epoch": 0.17692307692307693,
+      "grad_norm": 0.3972068234381968,
+      "learning_rate": 1.9622910694071654e-05,
+      "loss": 0.3783,
+      "step": 115
+    },
+    {
+      "epoch": 0.17846153846153845,
+      "grad_norm": 0.4396990125524382,
+      "learning_rate": 1.9616308894585078e-05,
+      "loss": 0.3787,
+      "step": 116
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3931184722966462,
+      "learning_rate": 1.9609650935856847e-05,
+      "loss": 0.3838,
+      "step": 117
+    },
+    {
+      "epoch": 0.18153846153846154,
+      "grad_norm": 0.42394116553785294,
+      "learning_rate": 1.9602936856769432e-05,
+      "loss": 0.3757,
+      "step": 118
+    },
+    {
+      "epoch": 0.18307692307692308,
+      "grad_norm": 0.3872711590188605,
+      "learning_rate": 1.9596166696533062e-05,
+      "loss": 0.3622,
+      "step": 119
+    },
+    {
+      "epoch": 0.18461538461538463,
+      "grad_norm": 0.43209247940136025,
+      "learning_rate": 1.9589340494685464e-05,
+      "loss": 0.3747,
+      "step": 120
+    },
+    {
+      "epoch": 0.18615384615384614,
+      "grad_norm": 0.378527901108427,
+      "learning_rate": 1.9582458291091664e-05,
+      "loss": 0.3501,
+      "step": 121
+    },
+    {
+      "epoch": 0.18769230769230769,
+      "grad_norm": 0.4372506951423749,
+      "learning_rate": 1.957552012594372e-05,
+      "loss": 0.3769,
+      "step": 122
+    },
+    {
+      "epoch": 0.18923076923076923,
+      "grad_norm": 0.42034408938336454,
+      "learning_rate": 1.956852603976052e-05,
+      "loss": 0.3799,
+      "step": 123
+    },
+    {
+      "epoch": 0.19076923076923077,
+      "grad_norm": 0.413661131437597,
+      "learning_rate": 1.9561476073387527e-05,
+      "loss": 0.3694,
+      "step": 124
+    },
+    {
+      "epoch": 0.19230769230769232,
+      "grad_norm": 0.4225946538123462,
+      "learning_rate": 1.9554370267996537e-05,
+      "loss": 0.3698,
+      "step": 125
+    },
+    {
+      "epoch": 0.19384615384615383,
+      "grad_norm": 0.4211824517053695,
+      "learning_rate": 1.954720866508546e-05,
+      "loss": 0.3723,
+      "step": 126
+    },
+    {
+      "epoch": 0.19538461538461538,
+      "grad_norm": 0.441349567520454,
+      "learning_rate": 1.9539991306478046e-05,
+      "loss": 0.3776,
+      "step": 127
+    },
+    {
+      "epoch": 0.19692307692307692,
+      "grad_norm": 0.38925256704611244,
+      "learning_rate": 1.953271823432367e-05,
+      "loss": 0.3771,
+      "step": 128
+    },
+    {
+      "epoch": 0.19846153846153847,
+      "grad_norm": 0.44208731465380763,
+      "learning_rate": 1.952538949109708e-05,
+      "loss": 0.3817,
+      "step": 129
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.37980861881568023,
+      "learning_rate": 1.9518005119598124e-05,
+      "loss": 0.3956,
+      "step": 130
+    },
+    {
+      "epoch": 0.20153846153846153,
+      "grad_norm": 0.43746564826656514,
+      "learning_rate": 1.9510565162951538e-05,
+      "loss": 0.383,
+      "step": 131
+    },
+    {
+      "epoch": 0.20307692307692307,
+      "grad_norm": 0.37554151931595203,
+      "learning_rate": 1.9503069664606663e-05,
+      "loss": 0.3812,
+      "step": 132
+    },
+    {
+      "epoch": 0.20461538461538462,
+      "grad_norm": 0.3938512517910276,
+      "learning_rate": 1.9495518668337204e-05,
+      "loss": 0.3755,
+      "step": 133
+    },
+    {
+      "epoch": 0.20615384615384616,
+      "grad_norm": 0.4315786747382278,
+      "learning_rate": 1.9487912218240983e-05,
+      "loss": 0.3769,
+      "step": 134
+    },
+    {
+      "epoch": 0.2076923076923077,
+      "grad_norm": 0.38364198110864106,
+      "learning_rate": 1.9480250358739667e-05,
+      "loss": 0.3562,
+      "step": 135
+    },
+    {
+      "epoch": 0.20923076923076922,
+      "grad_norm": 0.44967743929701914,
+      "learning_rate": 1.947253313457851e-05,
+      "loss": 0.36,
+      "step": 136
+    },
+    {
+      "epoch": 0.21076923076923076,
+      "grad_norm": 0.39255599974719213,
+      "learning_rate": 1.94647605908261e-05,
+      "loss": 0.3693,
+      "step": 137
+    },
+    {
+      "epoch": 0.2123076923076923,
+      "grad_norm": 0.43233859547333486,
+      "learning_rate": 1.9456932772874092e-05,
+      "loss": 0.3747,
+      "step": 138
+    },
+    {
+      "epoch": 0.21384615384615385,
+      "grad_norm": 0.4544360118997912,
+      "learning_rate": 1.944904972643694e-05,
+      "loss": 0.3705,
+      "step": 139
+    },
+    {
+      "epoch": 0.2153846153846154,
+      "grad_norm": 0.4569398610982909,
+      "learning_rate": 1.944111149755164e-05,
+      "loss": 0.3663,
+      "step": 140
+    },
+    {
+      "epoch": 0.2169230769230769,
+      "grad_norm": 0.4164699887840533,
+      "learning_rate": 1.9433118132577432e-05,
+      "loss": 0.3582,
+      "step": 141
+    },
+    {
+      "epoch": 0.21846153846153846,
+      "grad_norm": 0.4013888688937234,
+      "learning_rate": 1.9425069678195577e-05,
+      "loss": 0.3687,
+      "step": 142
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4505585773925497,
+      "learning_rate": 1.9416966181409047e-05,
+      "loss": 0.3715,
+      "step": 143
+    },
+    {
+      "epoch": 0.22153846153846155,
+      "grad_norm": 0.44354218271689405,
+      "learning_rate": 1.9408807689542257e-05,
+      "loss": 0.3863,
+      "step": 144
+    },
+    {
+      "epoch": 0.2230769230769231,
+      "grad_norm": 0.46031650050587875,
+      "learning_rate": 1.94005942502408e-05,
+      "loss": 0.376,
+      "step": 145
+    },
+    {
+      "epoch": 0.2246153846153846,
+      "grad_norm": 0.4248853811443392,
+      "learning_rate": 1.9392325911471154e-05,
+      "loss": 0.3753,
+      "step": 146
+    },
+    {
+      "epoch": 0.22615384615384615,
+      "grad_norm": 0.4185179395452762,
+      "learning_rate": 1.9384002721520423e-05,
+      "loss": 0.3631,
+      "step": 147
+    },
+    {
+      "epoch": 0.2276923076923077,
+      "grad_norm": 0.4008997188426561,
+      "learning_rate": 1.937562472899603e-05,
+      "loss": 0.3608,
+      "step": 148
+    },
+    {
+      "epoch": 0.22923076923076924,
+      "grad_norm": 0.34580398387061256,
+      "learning_rate": 1.936719198282545e-05,
+      "loss": 0.374,
+      "step": 149
+    },
+    {
+      "epoch": 0.23076923076923078,
+      "grad_norm": 0.38734751172931087,
+      "learning_rate": 1.935870453225592e-05,
+      "loss": 0.3717,
+      "step": 150
+    },
+    {
+      "epoch": 0.2323076923076923,
+      "grad_norm": 0.3960233199543072,
+      "learning_rate": 1.9350162426854152e-05,
+      "loss": 0.3723,
+      "step": 151
+    },
+    {
+      "epoch": 0.23384615384615384,
+      "grad_norm": 0.4002038595149067,
+      "learning_rate": 1.934156571650603e-05,
+      "loss": 0.3645,
+      "step": 152
+    },
+    {
+      "epoch": 0.2353846153846154,
+      "grad_norm": 0.40387269153898225,
+      "learning_rate": 1.933291445141635e-05,
+      "loss": 0.3693,
+      "step": 153
+    },
+    {
+      "epoch": 0.23692307692307693,
+      "grad_norm": 0.3948395913072729,
+      "learning_rate": 1.9324208682108493e-05,
+      "loss": 0.3638,
+      "step": 154
+    },
+    {
+      "epoch": 0.23846153846153847,
+      "grad_norm": 0.45422980198428947,
+      "learning_rate": 1.931544845942415e-05,
+      "loss": 0.3593,
+      "step": 155
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3944506938629949,
+      "learning_rate": 1.9306633834523022e-05,
+      "loss": 0.3637,
+      "step": 156
+    },
+    {
+      "epoch": 0.24153846153846154,
+      "grad_norm": 0.37482301609081387,
+      "learning_rate": 1.9297764858882516e-05,
+      "loss": 0.3577,
+      "step": 157
+    },
+    {
+      "epoch": 0.24307692307692308,
+      "grad_norm": 0.41539320799014184,
+      "learning_rate": 1.9288841584297445e-05,
+      "loss": 0.3603,
+      "step": 158
+    },
+    {
+      "epoch": 0.24461538461538462,
+      "grad_norm": 0.3812817138489635,
+      "learning_rate": 1.927986406287973e-05,
+      "loss": 0.3758,
+      "step": 159
+    },
+    {
+      "epoch": 0.24615384615384617,
+      "grad_norm": 0.4090759124668495,
+      "learning_rate": 1.92708323470581e-05,
+      "loss": 0.3758,
+      "step": 160
+    },
+    {
+      "epoch": 0.24769230769230768,
+      "grad_norm": 0.37892023888017984,
+      "learning_rate": 1.9261746489577767e-05,
+      "loss": 0.3776,
+      "step": 161
+    },
+    {
+      "epoch": 0.24923076923076923,
+      "grad_norm": 0.44478751408219885,
+      "learning_rate": 1.925260654350014e-05,
+      "loss": 0.3589,
+      "step": 162
+    },
+    {
+      "epoch": 0.25076923076923074,
+      "grad_norm": 0.3648405453963439,
+      "learning_rate": 1.92434125622025e-05,
+      "loss": 0.3577,
+      "step": 163
+    },
+    {
+      "epoch": 0.2523076923076923,
+      "grad_norm": 0.4879484783466178,
+      "learning_rate": 1.9234164599377692e-05,
+      "loss": 0.3531,
+      "step": 164
+    },
+    {
+      "epoch": 0.25384615384615383,
+      "grad_norm": 0.3731118065527373,
+      "learning_rate": 1.9224862709033823e-05,
+      "loss": 0.3607,
+      "step": 165
+    },
+    {
+      "epoch": 0.2553846153846154,
+      "grad_norm": 0.4097056211741247,
+      "learning_rate": 1.9215506945493933e-05,
+      "loss": 0.3637,
+      "step": 166
+    },
+    {
+      "epoch": 0.2569230769230769,
+      "grad_norm": 0.45284781505380334,
+      "learning_rate": 1.9206097363395668e-05,
+      "loss": 0.3732,
+      "step": 167
+    },
+    {
+      "epoch": 0.25846153846153846,
+      "grad_norm": 0.4062016992258126,
+      "learning_rate": 1.9196634017690993e-05,
+      "loss": 0.361,
+      "step": 168
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4282966491275036,
+      "learning_rate": 1.9187116963645845e-05,
+      "loss": 0.3715,
+      "step": 169
+    },
+    {
+      "epoch": 0.26153846153846155,
+      "grad_norm": 0.4242618592829376,
+      "learning_rate": 1.9177546256839814e-05,
+      "loss": 0.3491,
+      "step": 170
+    },
+    {
+      "epoch": 0.2630769230769231,
+      "grad_norm": 0.46549729776801385,
+      "learning_rate": 1.9167921953165827e-05,
+      "loss": 0.3708,
+      "step": 171
+    },
+    {
+      "epoch": 0.26461538461538464,
+      "grad_norm": 0.39329336860390796,
+      "learning_rate": 1.9158244108829815e-05,
+      "loss": 0.3653,
+      "step": 172
+    },
+    {
+      "epoch": 0.26615384615384613,
+      "grad_norm": 0.42072110375926625,
+      "learning_rate": 1.9148512780350384e-05,
+      "loss": 0.3658,
+      "step": 173
+    },
+    {
+      "epoch": 0.2676923076923077,
+      "grad_norm": 0.40327284519428425,
+      "learning_rate": 1.9138728024558494e-05,
+      "loss": 0.375,
+      "step": 174
+    },
+    {
+      "epoch": 0.2692307692307692,
+      "grad_norm": 0.4008125375412133,
+      "learning_rate": 1.9128889898597117e-05,
+      "loss": 0.3681,
+      "step": 175
+    },
+    {
+      "epoch": 0.27076923076923076,
+      "grad_norm": 0.3734692274716112,
+      "learning_rate": 1.91189984599209e-05,
+      "loss": 0.3617,
+      "step": 176
+    },
+    {
+      "epoch": 0.2723076923076923,
+      "grad_norm": 0.37325469734795563,
+      "learning_rate": 1.910905376629585e-05,
+      "loss": 0.3521,
+      "step": 177
+    },
+    {
+      "epoch": 0.27384615384615385,
+      "grad_norm": 0.3856667934315571,
+      "learning_rate": 1.9099055875798974e-05,
+      "loss": 0.3531,
+      "step": 178
+    },
+    {
+      "epoch": 0.2753846153846154,
+      "grad_norm": 0.3808888210552385,
+      "learning_rate": 1.9089004846817947e-05,
+      "loss": 0.3595,
+      "step": 179
+    },
+    {
+      "epoch": 0.27692307692307694,
+      "grad_norm": 0.3745173965711979,
+      "learning_rate": 1.9078900738050776e-05,
+      "loss": 0.3564,
+      "step": 180
+    },
+    {
+      "epoch": 0.2784615384615385,
+      "grad_norm": 0.34898266841145875,
+      "learning_rate": 1.9068743608505454e-05,
+      "loss": 0.3753,
+      "step": 181
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3644769885231561,
+      "learning_rate": 1.905853351749962e-05,
+      "loss": 0.3512,
+      "step": 182
+    },
+    {
+      "epoch": 0.2815384615384615,
+      "grad_norm": 0.3858389082782684,
+      "learning_rate": 1.9048270524660197e-05,
+      "loss": 0.3876,
+      "step": 183
+    },
+    {
+      "epoch": 0.28307692307692306,
+      "grad_norm": 0.38818683800074927,
+      "learning_rate": 1.903795468992306e-05,
+      "loss": 0.361,
+      "step": 184
+    },
+    {
+      "epoch": 0.2846153846153846,
+      "grad_norm": 0.36695466265129156,
+      "learning_rate": 1.902758607353269e-05,
+      "loss": 0.3654,
+      "step": 185
+    },
+    {
+      "epoch": 0.28615384615384615,
+      "grad_norm": 0.39570005961028404,
+      "learning_rate": 1.9017164736041795e-05,
+      "loss": 0.364,
+      "step": 186
+    },
+    {
+      "epoch": 0.2876923076923077,
+      "grad_norm": 0.3771366863278202,
+      "learning_rate": 1.9006690738310988e-05,
+      "loss": 0.3603,
+      "step": 187
+    },
+    {
+      "epoch": 0.28923076923076924,
+      "grad_norm": 0.38878503899522604,
+      "learning_rate": 1.8996164141508412e-05,
+      "loss": 0.3507,
+      "step": 188
+    },
+    {
+      "epoch": 0.2907692307692308,
+      "grad_norm": 0.3956545304159193,
+      "learning_rate": 1.898558500710939e-05,
+      "loss": 0.3667,
+      "step": 189
+    },
+    {
+      "epoch": 0.2923076923076923,
+      "grad_norm": 0.35381083702922,
+      "learning_rate": 1.8974953396896066e-05,
+      "loss": 0.3376,
+      "step": 190
+    },
+    {
+      "epoch": 0.29384615384615387,
+      "grad_norm": 0.4170829328048133,
+      "learning_rate": 1.896426937295704e-05,
+      "loss": 0.3572,
+      "step": 191
+    },
+    {
+      "epoch": 0.2953846153846154,
+      "grad_norm": 0.3559172538176691,
+      "learning_rate": 1.8953532997687008e-05,
+      "loss": 0.3541,
+      "step": 192
+    },
+    {
+      "epoch": 0.2969230769230769,
+      "grad_norm": 0.3869731893258225,
+      "learning_rate": 1.89427443337864e-05,
+      "loss": 0.3678,
+      "step": 193
+    },
+    {
+      "epoch": 0.29846153846153844,
+      "grad_norm": 0.3594161161799857,
+      "learning_rate": 1.8931903444261007e-05,
+      "loss": 0.3649,
+      "step": 194
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3613700145720469,
+      "learning_rate": 1.8921010392421628e-05,
+      "loss": 0.3622,
+      "step": 195
+    },
+    {
+      "epoch": 0.30153846153846153,
+      "grad_norm": 0.37149126110892394,
+      "learning_rate": 1.891006524188368e-05,
+      "loss": 0.37,
+      "step": 196
+    },
+    {
+      "epoch": 0.3030769230769231,
+      "grad_norm": 0.40590505685938155,
+      "learning_rate": 1.889906805656684e-05,
+      "loss": 0.3625,
+      "step": 197
+    },
+    {
+      "epoch": 0.3046153846153846,
+      "grad_norm": 0.36776799202272653,
+      "learning_rate": 1.888801890069467e-05,
+      "loss": 0.3348,
+      "step": 198
+    },
+    {
+      "epoch": 0.30615384615384617,
+      "grad_norm": 0.38714355877029916,
+      "learning_rate": 1.8876917838794226e-05,
+      "loss": 0.3518,
+      "step": 199
+    },
+    {
+      "epoch": 0.3076923076923077,
+      "grad_norm": 0.39132293890378467,
+      "learning_rate": 1.886576493569572e-05,
+      "loss": 0.3567,
+      "step": 200
+    },
+    {
+      "epoch": 0.30923076923076925,
+      "grad_norm": 0.37037393157233806,
+      "learning_rate": 1.8854560256532098e-05,
+      "loss": 0.3433,
+      "step": 201
+    },
+    {
+      "epoch": 0.31076923076923074,
+      "grad_norm": 0.3742182134623614,
+      "learning_rate": 1.884330386673869e-05,
+      "loss": 0.3602,
+      "step": 202
+    },
+    {
+      "epoch": 0.3123076923076923,
+      "grad_norm": 0.3745982088926344,
+      "learning_rate": 1.8831995832052802e-05,
+      "loss": 0.3598,
+      "step": 203
+    },
+    {
+      "epoch": 0.31384615384615383,
+      "grad_norm": 0.3148505000859751,
+      "learning_rate": 1.8820636218513354e-05,
+      "loss": 0.3514,
+      "step": 204
+    },
+    {
+      "epoch": 0.3153846153846154,
+      "grad_norm": 0.3777147146126031,
+      "learning_rate": 1.8809225092460488e-05,
+      "loss": 0.3624,
+      "step": 205
+    },
+    {
+      "epoch": 0.3169230769230769,
+      "grad_norm": 0.3620744024651921,
+      "learning_rate": 1.8797762520535178e-05,
+      "loss": 0.3545,
+      "step": 206
+    },
+    {
+      "epoch": 0.31846153846153846,
+      "grad_norm": 0.3275337617591835,
+      "learning_rate": 1.8786248569678847e-05,
+      "loss": 0.3549,
+      "step": 207
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.37980267084252867,
+      "learning_rate": 1.8774683307132956e-05,
+      "loss": 0.3632,
+      "step": 208
+    },
+    {
+      "epoch": 0.32153846153846155,
+      "grad_norm": 0.34585621541822953,
+      "learning_rate": 1.8763066800438638e-05,
+      "loss": 0.36,
+      "step": 209
+    },
+    {
+      "epoch": 0.3230769230769231,
+      "grad_norm": 0.3480579524545715,
+      "learning_rate": 1.8751399117436292e-05,
+      "loss": 0.3531,
+      "step": 210
+    },
+    {
+      "epoch": 0.32461538461538464,
+      "grad_norm": 0.3496509737199204,
+      "learning_rate": 1.873968032626518e-05,
+      "loss": 0.3605,
+      "step": 211
+    },
+    {
+      "epoch": 0.3261538461538461,
+      "grad_norm": 0.40524163043153,
+      "learning_rate": 1.8727910495363043e-05,
+      "loss": 0.3583,
+      "step": 212
+    },
+    {
+      "epoch": 0.32769230769230767,
+      "grad_norm": 0.34877588777050383,
+      "learning_rate": 1.8716089693465696e-05,
+      "loss": 0.347,
+      "step": 213
+    },
+    {
+      "epoch": 0.3292307692307692,
+      "grad_norm": 0.4592738818592951,
+      "learning_rate": 1.8704217989606606e-05,
+      "loss": 0.3598,
+      "step": 214
+    },
+    {
+      "epoch": 0.33076923076923076,
+      "grad_norm": 0.3783649551999555,
+      "learning_rate": 1.869229545311653e-05,
+      "loss": 0.3583,
+      "step": 215
+    },
+    {
+      "epoch": 0.3323076923076923,
+      "grad_norm": 0.3536547509407451,
+      "learning_rate": 1.8680322153623077e-05,
+      "loss": 0.3498,
+      "step": 216
+    },
+    {
+      "epoch": 0.33384615384615385,
+      "grad_norm": 0.4079948196470235,
+      "learning_rate": 1.8668298161050308e-05,
+      "loss": 0.3622,
+      "step": 217
+    },
+    {
+      "epoch": 0.3353846153846154,
+      "grad_norm": 0.36942640741898625,
+      "learning_rate": 1.8656223545618345e-05,
+      "loss": 0.353,
+      "step": 218
+    },
+    {
+      "epoch": 0.33692307692307694,
+      "grad_norm": 0.39196579363565714,
+      "learning_rate": 1.8644098377842934e-05,
+      "loss": 0.3532,
+      "step": 219
+    },
+    {
+      "epoch": 0.3384615384615385,
+      "grad_norm": 0.3564332628975117,
+      "learning_rate": 1.8631922728535054e-05,
+      "loss": 0.3551,
+      "step": 220
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.34786080023934873,
+      "learning_rate": 1.8619696668800494e-05,
+      "loss": 0.3531,
+      "step": 221
+    },
+    {
+      "epoch": 0.3415384615384615,
+      "grad_norm": 0.37096480013302563,
+      "learning_rate": 1.860742027003944e-05,
+      "loss": 0.3472,
+      "step": 222
+    },
+    {
+      "epoch": 0.34307692307692306,
+      "grad_norm": 0.33695890031849024,
+      "learning_rate": 1.8595093603946053e-05,
+      "loss": 0.3489,
+      "step": 223
+    },
+    {
+      "epoch": 0.3446153846153846,
+      "grad_norm": 0.36025354068440185,
+      "learning_rate": 1.8582716742508066e-05,
+      "loss": 0.3539,
+      "step": 224
+    },
+    {
+      "epoch": 0.34615384615384615,
+      "grad_norm": 0.3410672872818961,
+      "learning_rate": 1.8570289758006346e-05,
+      "loss": 0.352,
+      "step": 225
+    },
+    {
+      "epoch": 0.3476923076923077,
+      "grad_norm": 0.3598105304326528,
+      "learning_rate": 1.8557812723014476e-05,
+      "loss": 0.3538,
+      "step": 226
+    },
+    {
+      "epoch": 0.34923076923076923,
+      "grad_norm": 0.36804240330611765,
+      "learning_rate": 1.8545285710398343e-05,
+      "loss": 0.3519,
+      "step": 227
+    },
+    {
+      "epoch": 0.3507692307692308,
+      "grad_norm": 0.32813391778641043,
+      "learning_rate": 1.853270879331569e-05,
+      "loss": 0.3539,
+      "step": 228
+    },
+    {
+      "epoch": 0.3523076923076923,
+      "grad_norm": 0.3846644370625242,
+      "learning_rate": 1.852008204521572e-05,
+      "loss": 0.3429,
+      "step": 229
+    },
+    {
+      "epoch": 0.35384615384615387,
+      "grad_norm": 0.36227501037133736,
+      "learning_rate": 1.850740553983863e-05,
+      "loss": 0.3524,
+      "step": 230
+    },
+    {
+      "epoch": 0.3553846153846154,
+      "grad_norm": 0.35547652422353576,
+      "learning_rate": 1.8494679351215212e-05,
+      "loss": 0.3459,
+      "step": 231
+    },
+    {
+      "epoch": 0.3569230769230769,
+      "grad_norm": 0.3439467119964199,
+      "learning_rate": 1.8481903553666405e-05,
+      "loss": 0.3635,
+      "step": 232
+    },
+    {
+      "epoch": 0.35846153846153844,
+      "grad_norm": 0.38495422393809375,
+      "learning_rate": 1.846907822180286e-05,
+      "loss": 0.3583,
+      "step": 233
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3468097978390537,
+      "learning_rate": 1.845620343052452e-05,
+      "loss": 0.3519,
+      "step": 234
+    },
+    {
+      "epoch": 0.36153846153846153,
+      "grad_norm": 0.3939392362586235,
+      "learning_rate": 1.8443279255020153e-05,
+      "loss": 0.3389,
+      "step": 235
+    },
+    {
+      "epoch": 0.3630769230769231,
+      "grad_norm": 0.3557187005044313,
+      "learning_rate": 1.8430305770766947e-05,
+      "loss": 0.3538,
+      "step": 236
+    },
+    {
+      "epoch": 0.3646153846153846,
+      "grad_norm": 0.3613084285594687,
+      "learning_rate": 1.8417283053530047e-05,
+      "loss": 0.3544,
+      "step": 237
+    },
+    {
+      "epoch": 0.36615384615384616,
+      "grad_norm": 0.37263843055238405,
+      "learning_rate": 1.8404211179362116e-05,
+      "loss": 0.3467,
+      "step": 238
+    },
+    {
+      "epoch": 0.3676923076923077,
+      "grad_norm": 0.3562469357331947,
+      "learning_rate": 1.8391090224602895e-05,
+      "loss": 0.3557,
+      "step": 239
+    },
+    {
+      "epoch": 0.36923076923076925,
+      "grad_norm": 0.3409376089014781,
+      "learning_rate": 1.837792026587876e-05,
+      "loss": 0.3423,
+      "step": 240
+    },
+    {
+      "epoch": 0.3707692307692308,
+      "grad_norm": 0.42464049754075456,
+      "learning_rate": 1.8364701380102267e-05,
+      "loss": 0.3463,
+      "step": 241
+    },
+    {
+      "epoch": 0.3723076923076923,
+      "grad_norm": 0.37801548493364734,
+      "learning_rate": 1.8351433644471708e-05,
+      "loss": 0.3409,
+      "step": 242
+    },
+    {
+      "epoch": 0.37384615384615383,
+      "grad_norm": 0.408672340900871,
+      "learning_rate": 1.8338117136470648e-05,
+      "loss": 0.3759,
+      "step": 243
+    },
+    {
+      "epoch": 0.37538461538461537,
+      "grad_norm": 0.37546504595179914,
+      "learning_rate": 1.8324751933867496e-05,
+      "loss": 0.3563,
+      "step": 244
+    },
+    {
+      "epoch": 0.3769230769230769,
+      "grad_norm": 0.36631809004670857,
+      "learning_rate": 1.831133811471503e-05,
+      "loss": 0.3578,
+      "step": 245
+    },
+    {
+      "epoch": 0.37846153846153846,
+      "grad_norm": 0.3736994550757926,
+      "learning_rate": 1.829787575734995e-05,
+      "loss": 0.3538,
+      "step": 246
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.37463686730232026,
+      "learning_rate": 1.8284364940392426e-05,
+      "loss": 0.3544,
+      "step": 247
+    },
+    {
+      "epoch": 0.38153846153846155,
+      "grad_norm": 0.3725028314834782,
+      "learning_rate": 1.827080574274562e-05,
+      "loss": 0.3553,
+      "step": 248
+    },
+    {
+      "epoch": 0.3830769230769231,
+      "grad_norm": 0.39321544749534515,
+      "learning_rate": 1.825719824359524e-05,
+      "loss": 0.3572,
+      "step": 249
+    },
+    {
+      "epoch": 0.38461538461538464,
+      "grad_norm": 0.39648989139486013,
+      "learning_rate": 1.824354252240909e-05,
+      "loss": 0.356,
+      "step": 250
+    },
+    {
+      "epoch": 0.3861538461538462,
+      "grad_norm": 0.34380776676973107,
+      "learning_rate": 1.8229838658936566e-05,
+      "loss": 0.3452,
+      "step": 251
+    },
+    {
+      "epoch": 0.38769230769230767,
+      "grad_norm": 0.36601585597171354,
+      "learning_rate": 1.821608673320823e-05,
+      "loss": 0.3396,
+      "step": 252
+    },
+    {
+      "epoch": 0.3892307692307692,
+      "grad_norm": 0.3616253511776755,
+      "learning_rate": 1.820228682553533e-05,
+      "loss": 0.3396,
+      "step": 253
+    },
+    {
+      "epoch": 0.39076923076923076,
+      "grad_norm": 0.3516180682215301,
+      "learning_rate": 1.818843901650932e-05,
+      "loss": 0.3533,
+      "step": 254
+    },
+    {
+      "epoch": 0.3923076923076923,
+      "grad_norm": 0.35992255609326845,
+      "learning_rate": 1.8174543387001403e-05,
+      "loss": 0.353,
+      "step": 255
+    },
+    {
+      "epoch": 0.39384615384615385,
+      "grad_norm": 0.359104051539393,
+      "learning_rate": 1.816060001816205e-05,
+      "loss": 0.3528,
+      "step": 256
+    },
+    {
+      "epoch": 0.3953846153846154,
+      "grad_norm": 0.3839196503752293,
+      "learning_rate": 1.8146608991420533e-05,
+      "loss": 0.3528,
+      "step": 257
+    },
+    {
+      "epoch": 0.39692307692307693,
+      "grad_norm": 0.3945898720253854,
+      "learning_rate": 1.8132570388484442e-05,
+      "loss": 0.3591,
+      "step": 258
+    },
+    {
+      "epoch": 0.3984615384615385,
+      "grad_norm": 0.35326375322997333,
+      "learning_rate": 1.811848429133922e-05,
+      "loss": 0.3476,
+      "step": 259
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3987685904984489,
+      "learning_rate": 1.810435078224767e-05,
+      "loss": 0.349,
+      "step": 260
+    },
+    {
+      "epoch": 0.4015384615384615,
+      "grad_norm": 0.36852779510455175,
+      "learning_rate": 1.8090169943749477e-05,
+      "loss": 0.3651,
+      "step": 261
+    },
+    {
+      "epoch": 0.40307692307692305,
+      "grad_norm": 0.4086515779440857,
+      "learning_rate": 1.8075941858660737e-05,
+      "loss": 0.3461,
+      "step": 262
+    },
+    {
+      "epoch": 0.4046153846153846,
+      "grad_norm": 0.3793557979593047,
+      "learning_rate": 1.8061666610073465e-05,
+      "loss": 0.3522,
+      "step": 263
+    },
+    {
+      "epoch": 0.40615384615384614,
+      "grad_norm": 0.4119306737647016,
+      "learning_rate": 1.8047344281355112e-05,
+      "loss": 0.3541,
+      "step": 264
+    },
+    {
+      "epoch": 0.4076923076923077,
+      "grad_norm": 0.3808370864185399,
+      "learning_rate": 1.8032974956148064e-05,
+      "loss": 0.3501,
+      "step": 265
+    },
+    {
+      "epoch": 0.40923076923076923,
+      "grad_norm": 0.3827615753077048,
+      "learning_rate": 1.8018558718369187e-05,
+      "loss": 0.3507,
+      "step": 266
+    },
+    {
+      "epoch": 0.4107692307692308,
+      "grad_norm": 0.3876568030885218,
+      "learning_rate": 1.8004095652209304e-05,
+      "loss": 0.3386,
+      "step": 267
+    },
+    {
+      "epoch": 0.4123076923076923,
+      "grad_norm": 0.3738590685818937,
+      "learning_rate": 1.7989585842132713e-05,
+      "loss": 0.3547,
+      "step": 268
+    },
+    {
+      "epoch": 0.41384615384615386,
+      "grad_norm": 0.34897180553671925,
+      "learning_rate": 1.7975029372876706e-05,
+      "loss": 0.3427,
+      "step": 269
+    },
+    {
+      "epoch": 0.4153846153846154,
+      "grad_norm": 0.3792529674837759,
+      "learning_rate": 1.7960426329451062e-05,
+      "loss": 0.347,
+      "step": 270
+    },
+    {
+      "epoch": 0.4169230769230769,
+      "grad_norm": 0.3699278879810063,
+      "learning_rate": 1.7945776797137544e-05,
+      "loss": 0.36,
+      "step": 271
+    },
+    {
+      "epoch": 0.41846153846153844,
+      "grad_norm": 0.37408068508210207,
+      "learning_rate": 1.7931080861489425e-05,
+      "loss": 0.353,
+      "step": 272
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3470876802937946,
+      "learning_rate": 1.791633860833096e-05,
+      "loss": 0.3452,
+      "step": 273
+    },
+    {
+      "epoch": 0.42153846153846153,
+      "grad_norm": 0.32277614204114535,
+      "learning_rate": 1.7901550123756906e-05,
+      "loss": 0.3597,
+      "step": 274
+    },
+    {
+      "epoch": 0.4230769230769231,
+      "grad_norm": 0.3122099562281799,
+      "learning_rate": 1.7886715494132008e-05,
+      "loss": 0.343,
+      "step": 275
+    },
+    {
+      "epoch": 0.4246153846153846,
+      "grad_norm": 0.35093054267366997,
+      "learning_rate": 1.7871834806090502e-05,
+      "loss": 0.3402,
+      "step": 276
+    },
+    {
+      "epoch": 0.42615384615384616,
+      "grad_norm": 0.3526844675215035,
+      "learning_rate": 1.7856908146535602e-05,
+      "loss": 0.3471,
+      "step": 277
+    },
+    {
+      "epoch": 0.4276923076923077,
+      "grad_norm": 0.36989716444938414,
+      "learning_rate": 1.7841935602638997e-05,
+      "loss": 0.3493,
+      "step": 278
+    },
+    {
+      "epoch": 0.42923076923076925,
+      "grad_norm": 0.34659497432637826,
+      "learning_rate": 1.7826917261840337e-05,
+      "loss": 0.3476,
+      "step": 279
+    },
+    {
+      "epoch": 0.4307692307692308,
+      "grad_norm": 0.36942940048352685,
+      "learning_rate": 1.781185321184673e-05,
+      "loss": 0.3465,
+      "step": 280
+    },
+    {
+      "epoch": 0.4323076923076923,
+      "grad_norm": 0.33930299087150695,
+      "learning_rate": 1.7796743540632226e-05,
+      "loss": 0.3488,
+      "step": 281
+    },
+    {
+      "epoch": 0.4338461538461538,
+      "grad_norm": 0.3532264238998626,
+      "learning_rate": 1.77815883364373e-05,
+      "loss": 0.3412,
+      "step": 282
+    },
+    {
+      "epoch": 0.43538461538461537,
+      "grad_norm": 0.34563607097263,
+      "learning_rate": 1.7766387687768338e-05,
+      "loss": 0.3492,
+      "step": 283
+    },
+    {
+      "epoch": 0.4369230769230769,
+      "grad_norm": 0.3359455576897893,
+      "learning_rate": 1.7751141683397128e-05,
+      "loss": 0.3453,
+      "step": 284
+    },
+    {
+      "epoch": 0.43846153846153846,
+      "grad_norm": 0.33525285705398855,
+      "learning_rate": 1.7735850412360332e-05,
+      "loss": 0.3493,
+      "step": 285
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3349856346616677,
+      "learning_rate": 1.772051396395897e-05,
+      "loss": 0.3534,
+      "step": 286
+    },
+    {
+      "epoch": 0.44153846153846155,
+      "grad_norm": 0.3392621630571748,
+      "learning_rate": 1.7705132427757895e-05,
+      "loss": 0.3487,
+      "step": 287
+    },
+    {
+      "epoch": 0.4430769230769231,
+      "grad_norm": 0.3367417477753544,
+      "learning_rate": 1.7689705893585273e-05,
+      "loss": 0.3517,
+      "step": 288
+    },
+    {
+      "epoch": 0.44461538461538463,
+      "grad_norm": 0.3451957771468646,
+      "learning_rate": 1.7674234451532065e-05,
+      "loss": 0.3452,
+      "step": 289
+    },
+    {
+      "epoch": 0.4461538461538462,
+      "grad_norm": 0.3685834045166528,
+      "learning_rate": 1.7658718191951483e-05,
+      "loss": 0.3541,
+      "step": 290
+    },
+    {
+      "epoch": 0.44769230769230767,
+      "grad_norm": 0.35075879867972815,
+      "learning_rate": 1.7643157205458483e-05,
+      "loss": 0.3517,
+      "step": 291
+    },
+    {
+      "epoch": 0.4492307692307692,
+      "grad_norm": 0.363045183643031,
+      "learning_rate": 1.7627551582929223e-05,
+      "loss": 0.3637,
+      "step": 292
+    },
+    {
+      "epoch": 0.45076923076923076,
+      "grad_norm": 0.3685236151251217,
+      "learning_rate": 1.7611901415500536e-05,
+      "loss": 0.3257,
+      "step": 293
+    },
+    {
+      "epoch": 0.4523076923076923,
+      "grad_norm": 0.3538775664927815,
+      "learning_rate": 1.759620679456939e-05,
+      "loss": 0.3466,
+      "step": 294
+    },
+    {
+      "epoch": 0.45384615384615384,
+      "grad_norm": 0.37169949722464124,
+      "learning_rate": 1.7580467811792374e-05,
+      "loss": 0.3522,
+      "step": 295
+    },
+    {
+      "epoch": 0.4553846153846154,
+      "grad_norm": 0.40012856967527305,
+      "learning_rate": 1.7564684559085138e-05,
+      "loss": 0.3408,
+      "step": 296
+    },
+    {
+      "epoch": 0.45692307692307693,
+      "grad_norm": 0.3666356782056923,
+      "learning_rate": 1.7548857128621878e-05,
+      "loss": 0.3452,
+      "step": 297
+    },
+    {
+      "epoch": 0.4584615384615385,
+      "grad_norm": 0.3272969997669924,
+      "learning_rate": 1.753298561283478e-05,
+      "loss": 0.3434,
+      "step": 298
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.38831742649553147,
+      "learning_rate": 1.7517070104413497e-05,
+      "loss": 0.3452,
+      "step": 299
+    },
+    {
+      "epoch": 0.46153846153846156,
+      "grad_norm": 0.3602246989660254,
+      "learning_rate": 1.7501110696304598e-05,
+      "loss": 0.3561,
+      "step": 300
+    },
+    {
+      "epoch": 0.46307692307692305,
+      "grad_norm": 0.32751426562964475,
+      "learning_rate": 1.7485107481711014e-05,
+      "loss": 0.3422,
+      "step": 301
+    },
+    {
+      "epoch": 0.4646153846153846,
+      "grad_norm": 0.3775562040944554,
+      "learning_rate": 1.7469060554091518e-05,
+      "loss": 0.3567,
+      "step": 302
+    },
+    {
+      "epoch": 0.46615384615384614,
+      "grad_norm": 0.3746811301380941,
+      "learning_rate": 1.745297000716016e-05,
+      "loss": 0.3417,
+      "step": 303
+    },
+    {
+      "epoch": 0.4676923076923077,
+      "grad_norm": 0.32919868953567194,
+      "learning_rate": 1.7436835934885735e-05,
+      "loss": 0.3442,
+      "step": 304
+    },
+    {
+      "epoch": 0.46923076923076923,
+      "grad_norm": 0.3862876725431514,
+      "learning_rate": 1.7420658431491224e-05,
+      "loss": 0.3361,
+      "step": 305
+    },
+    {
+      "epoch": 0.4707692307692308,
+      "grad_norm": 0.34901646771529365,
+      "learning_rate": 1.7404437591453237e-05,
+      "loss": 0.3508,
+      "step": 306
+    },
+    {
+      "epoch": 0.4723076923076923,
+      "grad_norm": 0.3675088015884468,
+      "learning_rate": 1.7388173509501475e-05,
+      "loss": 0.3457,
+      "step": 307
+    },
+    {
+      "epoch": 0.47384615384615386,
+      "grad_norm": 0.3515874155159066,
+      "learning_rate": 1.7371866280618176e-05,
+      "loss": 0.3643,
+      "step": 308
+    },
+    {
+      "epoch": 0.4753846153846154,
+      "grad_norm": 0.36516295839622703,
+      "learning_rate": 1.7355516000037555e-05,
+      "loss": 0.3434,
+      "step": 309
+    },
+    {
+      "epoch": 0.47692307692307695,
+      "grad_norm": 0.36215517755761173,
+      "learning_rate": 1.733912276324524e-05,
+      "loss": 0.3472,
+      "step": 310
+    },
+    {
+      "epoch": 0.47846153846153844,
+      "grad_norm": 0.36430798311353485,
+      "learning_rate": 1.7322686665977738e-05,
+      "loss": 0.35,
+      "step": 311
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3363708591416349,
+      "learning_rate": 1.7306207804221845e-05,
+      "loss": 0.3327,
+      "step": 312
+    },
+    {
+      "epoch": 0.4815384615384615,
+      "grad_norm": 0.3464405667250271,
+      "learning_rate": 1.7289686274214116e-05,
+      "loss": 0.3309,
+      "step": 313
+    },
+    {
+      "epoch": 0.48307692307692307,
+      "grad_norm": 0.41057799162040165,
+      "learning_rate": 1.727312217244028e-05,
+      "loss": 0.3362,
+      "step": 314
+    },
+    {
+      "epoch": 0.4846153846153846,
+      "grad_norm": 0.3508212343362318,
+      "learning_rate": 1.7256515595634688e-05,
+      "loss": 0.3445,
+      "step": 315
+    },
+    {
+      "epoch": 0.48615384615384616,
+      "grad_norm": 0.36827996807255337,
+      "learning_rate": 1.7239866640779745e-05,
+      "loss": 0.3463,
+      "step": 316
+    },
+    {
+      "epoch": 0.4876923076923077,
+      "grad_norm": 0.34115993415767315,
+      "learning_rate": 1.722317540510534e-05,
+      "loss": 0.3426,
+      "step": 317
+    },
+    {
+      "epoch": 0.48923076923076925,
+      "grad_norm": 0.3312349293414821,
+      "learning_rate": 1.720644198608829e-05,
+      "loss": 0.347,
+      "step": 318
+    },
+    {
+      "epoch": 0.4907692307692308,
+      "grad_norm": 0.349340239732704,
+      "learning_rate": 1.7189666481451755e-05,
+      "loss": 0.3615,
+      "step": 319
+    },
+    {
+      "epoch": 0.49230769230769234,
+      "grad_norm": 0.3330797470660715,
+      "learning_rate": 1.717284898916468e-05,
+      "loss": 0.3421,
+      "step": 320
+    },
+    {
+      "epoch": 0.4938461538461538,
+      "grad_norm": 0.32038359438240244,
+      "learning_rate": 1.715598960744121e-05,
+      "loss": 0.356,
+      "step": 321
+    },
+    {
+      "epoch": 0.49538461538461537,
+      "grad_norm": 0.3429307183460945,
+      "learning_rate": 1.7139088434740142e-05,
+      "loss": 0.3315,
+      "step": 322
+    },
+    {
+      "epoch": 0.4969230769230769,
+      "grad_norm": 0.3329575955338515,
+      "learning_rate": 1.712214556976431e-05,
+      "loss": 0.3374,
+      "step": 323
+    },
+    {
+      "epoch": 0.49846153846153846,
+      "grad_norm": 0.33173791143793213,
+      "learning_rate": 1.7105161111460046e-05,
+      "loss": 0.3416,
+      "step": 324
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.35973140787573293,
+      "learning_rate": 1.7088135159016584e-05,
+      "loss": 0.3557,
+      "step": 325
+    },
+    {
+      "epoch": 0.5015384615384615,
+      "grad_norm": 0.32789280511095725,
+      "learning_rate": 1.7071067811865477e-05,
+      "loss": 0.353,
+      "step": 326
+    },
+    {
+      "epoch": 0.5030769230769231,
+      "grad_norm": 0.40991419155641984,
+      "learning_rate": 1.7053959169680033e-05,
+      "loss": 0.3488,
+      "step": 327
+    },
+    {
+      "epoch": 0.5046153846153846,
+      "grad_norm": 0.31432251880938783,
+      "learning_rate": 1.7036809332374713e-05,
+      "loss": 0.3377,
+      "step": 328
+    },
+    {
+      "epoch": 0.5061538461538462,
+      "grad_norm": 0.3665523546008641,
+      "learning_rate": 1.7019618400104572e-05,
+      "loss": 0.3378,
+      "step": 329
+    },
+    {
+      "epoch": 0.5076923076923077,
+      "grad_norm": 0.3320078438985165,
+      "learning_rate": 1.700238647326464e-05,
+      "loss": 0.3361,
+      "step": 330
+    },
+    {
+      "epoch": 0.5092307692307693,
+      "grad_norm": 0.35680941172132136,
+      "learning_rate": 1.6985113652489374e-05,
+      "loss": 0.3453,
+      "step": 331
+    },
+    {
+      "epoch": 0.5107692307692308,
+      "grad_norm": 0.3654902422894205,
+      "learning_rate": 1.6967800038652035e-05,
+      "loss": 0.3467,
+      "step": 332
+    },
+    {
+      "epoch": 0.5123076923076924,
+      "grad_norm": 0.3332106421189731,
+      "learning_rate": 1.695044573286413e-05,
+      "loss": 0.3395,
+      "step": 333
+    },
+    {
+      "epoch": 0.5138461538461538,
+      "grad_norm": 0.3576528257798259,
+      "learning_rate": 1.69330508364748e-05,
+      "loss": 0.3454,
+      "step": 334
+    },
+    {
+      "epoch": 0.5153846153846153,
+      "grad_norm": 0.3154421189547101,
+      "learning_rate": 1.6915615451070234e-05,
+      "loss": 0.3627,
+      "step": 335
+    },
+    {
+      "epoch": 0.5169230769230769,
+      "grad_norm": 0.33079254546716974,
+      "learning_rate": 1.689813967847308e-05,
+      "loss": 0.3385,
+      "step": 336
+    },
+    {
+      "epoch": 0.5184615384615384,
+      "grad_norm": 0.3718374242713586,
+      "learning_rate": 1.6880623620741843e-05,
+      "loss": 0.3386,
+      "step": 337
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.35919342674204147,
+      "learning_rate": 1.68630673801703e-05,
+      "loss": 0.3347,
+      "step": 338
+    },
+    {
+      "epoch": 0.5215384615384615,
+      "grad_norm": 0.4145660600161487,
+      "learning_rate": 1.684547105928689e-05,
+      "loss": 0.3326,
+      "step": 339
+    },
+    {
+      "epoch": 0.5230769230769231,
+      "grad_norm": 0.35262102463458106,
+      "learning_rate": 1.682783476085412e-05,
+      "loss": 0.3468,
+      "step": 340
+    },
+    {
+      "epoch": 0.5246153846153846,
+      "grad_norm": 0.3969110582273147,
+      "learning_rate": 1.6810158587867973e-05,
+      "loss": 0.345,
+      "step": 341
+    },
+    {
+      "epoch": 0.5261538461538462,
+      "grad_norm": 0.36729152355342054,
+      "learning_rate": 1.679244264355729e-05,
+      "loss": 0.3494,
+      "step": 342
+    },
+    {
+      "epoch": 0.5276923076923077,
+      "grad_norm": 0.34584656489173654,
+      "learning_rate": 1.677468703138319e-05,
+      "loss": 0.3598,
+      "step": 343
+    },
+    {
+      "epoch": 0.5292307692307693,
+      "grad_norm": 0.3769779959966056,
+      "learning_rate": 1.6756891855038436e-05,
+      "loss": 0.3428,
+      "step": 344
+    },
+    {
+      "epoch": 0.5307692307692308,
+      "grad_norm": 0.3870397539579194,
+      "learning_rate": 1.673905721844686e-05,
+      "loss": 0.3399,
+      "step": 345
+    },
+    {
+      "epoch": 0.5323076923076923,
+      "grad_norm": 0.3803105690557983,
+      "learning_rate": 1.6721183225762726e-05,
+      "loss": 0.3407,
+      "step": 346
+    },
+    {
+      "epoch": 0.5338461538461539,
+      "grad_norm": 0.3811961355506509,
+      "learning_rate": 1.670326998137016e-05,
+      "loss": 0.3435,
+      "step": 347
+    },
+    {
+      "epoch": 0.5353846153846153,
+      "grad_norm": 0.355424944231993,
+      "learning_rate": 1.668531758988249e-05,
+      "loss": 0.3419,
+      "step": 348
+    },
+    {
+      "epoch": 0.536923076923077,
+      "grad_norm": 0.3697051964035359,
+      "learning_rate": 1.666732615614169e-05,
+      "loss": 0.3558,
+      "step": 349
+    },
+    {
+      "epoch": 0.5384615384615384,
+      "grad_norm": 0.3724554786238689,
+      "learning_rate": 1.6649295785217722e-05,
+      "loss": 0.3465,
+      "step": 350
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3713258592311073,
+      "learning_rate": 1.6631226582407954e-05,
+      "loss": 0.3472,
+      "step": 351
+    },
+    {
+      "epoch": 0.5415384615384615,
+      "grad_norm": 0.37379815805633365,
+      "learning_rate": 1.661311865323652e-05,
+      "loss": 0.3476,
+      "step": 352
+    },
+    {
+      "epoch": 0.5430769230769231,
+      "grad_norm": 0.3574918037766584,
+      "learning_rate": 1.6594972103453727e-05,
+      "loss": 0.3505,
+      "step": 353
+    },
+    {
+      "epoch": 0.5446153846153846,
+      "grad_norm": 0.36735856136671163,
+      "learning_rate": 1.6576787039035417e-05,
+      "loss": 0.3437,
+      "step": 354
+    },
+    {
+      "epoch": 0.5461538461538461,
+      "grad_norm": 0.39795420244333224,
+      "learning_rate": 1.6558563566182365e-05,
+      "loss": 0.3387,
+      "step": 355
+    },
+    {
+      "epoch": 0.5476923076923077,
+      "grad_norm": 0.35332927485568466,
+      "learning_rate": 1.6540301791319647e-05,
+      "loss": 0.3351,
+      "step": 356
+    },
+    {
+      "epoch": 0.5492307692307692,
+      "grad_norm": 0.3405051238522205,
+      "learning_rate": 1.652200182109602e-05,
+      "loss": 0.3388,
+      "step": 357
+    },
+    {
+      "epoch": 0.5507692307692308,
+      "grad_norm": 0.3833225434413435,
+      "learning_rate": 1.6503663762383312e-05,
+      "loss": 0.3419,
+      "step": 358
+    },
+    {
+      "epoch": 0.5523076923076923,
+      "grad_norm": 0.3462506998511973,
+      "learning_rate": 1.6485287722275783e-05,
+      "loss": 0.3475,
+      "step": 359
+    },
+    {
+      "epoch": 0.5538461538461539,
+      "grad_norm": 0.391689506628609,
+      "learning_rate": 1.6466873808089496e-05,
+      "loss": 0.357,
+      "step": 360
+    },
+    {
+      "epoch": 0.5553846153846154,
+      "grad_norm": 0.3270791225847661,
+      "learning_rate": 1.6448422127361707e-05,
+      "loss": 0.3443,
+      "step": 361
+    },
+    {
+      "epoch": 0.556923076923077,
+      "grad_norm": 0.37696990145987513,
+      "learning_rate": 1.642993278785023e-05,
+      "loss": 0.3461,
+      "step": 362
+    },
+    {
+      "epoch": 0.5584615384615385,
+      "grad_norm": 0.3489276120822278,
+      "learning_rate": 1.64114058975328e-05,
+      "loss": 0.3416,
+      "step": 363
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.35545398572334047,
+      "learning_rate": 1.639284156460646e-05,
+      "loss": 0.3451,
+      "step": 364
+    },
+    {
+      "epoch": 0.5615384615384615,
+      "grad_norm": 0.3496864708346177,
+      "learning_rate": 1.63742398974869e-05,
+      "loss": 0.3582,
+      "step": 365
+    },
+    {
+      "epoch": 0.563076923076923,
+      "grad_norm": 0.3433728900258887,
+      "learning_rate": 1.6355601004807856e-05,
+      "loss": 0.35,
+      "step": 366
+    },
+    {
+      "epoch": 0.5646153846153846,
+      "grad_norm": 0.363448160242182,
+      "learning_rate": 1.6336924995420453e-05,
+      "loss": 0.3412,
+      "step": 367
+    },
+    {
+      "epoch": 0.5661538461538461,
+      "grad_norm": 0.3358514480354539,
+      "learning_rate": 1.6318211978392588e-05,
+      "loss": 0.3462,
+      "step": 368
+    },
+    {
+      "epoch": 0.5676923076923077,
+      "grad_norm": 0.3529963835277034,
+      "learning_rate": 1.6299462063008272e-05,
+      "loss": 0.3454,
+      "step": 369
+    },
+    {
+      "epoch": 0.5692307692307692,
+      "grad_norm": 0.35264544772660966,
+      "learning_rate": 1.6280675358767005e-05,
+      "loss": 0.3335,
+      "step": 370
+    },
+    {
+      "epoch": 0.5707692307692308,
+      "grad_norm": 0.34118145221992047,
+      "learning_rate": 1.626185197538314e-05,
+      "loss": 0.3506,
+      "step": 371
+    },
+    {
+      "epoch": 0.5723076923076923,
+      "grad_norm": 0.33861364933452004,
+      "learning_rate": 1.6242992022785225e-05,
+      "loss": 0.3394,
+      "step": 372
+    },
+    {
+      "epoch": 0.5738461538461539,
+      "grad_norm": 0.34114296179605763,
+      "learning_rate": 1.6224095611115385e-05,
+      "loss": 0.342,
+      "step": 373
+    },
+    {
+      "epoch": 0.5753846153846154,
+      "grad_norm": 0.35031645242644793,
+      "learning_rate": 1.620516285072866e-05,
+      "loss": 0.3446,
+      "step": 374
+    },
+    {
+      "epoch": 0.5769230769230769,
+      "grad_norm": 0.33369847070376424,
+      "learning_rate": 1.6186193852192356e-05,
+      "loss": 0.3283,
+      "step": 375
+    },
+    {
+      "epoch": 0.5784615384615385,
+      "grad_norm": 0.35653923352092576,
+      "learning_rate": 1.6167188726285433e-05,
+      "loss": 0.3413,
+      "step": 376
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.33404971938811645,
+      "learning_rate": 1.6148147583997813e-05,
+      "loss": 0.3501,
+      "step": 377
+    },
+    {
+      "epoch": 0.5815384615384616,
+      "grad_norm": 0.37355957953604907,
+      "learning_rate": 1.6129070536529767e-05,
+      "loss": 0.3381,
+      "step": 378
+    },
+    {
+      "epoch": 0.583076923076923,
+      "grad_norm": 0.3503877045317371,
+      "learning_rate": 1.6109957695291246e-05,
+      "loss": 0.3394,
+      "step": 379
+    },
+    {
+      "epoch": 0.5846153846153846,
+      "grad_norm": 0.3458427753639511,
+      "learning_rate": 1.6090809171901237e-05,
+      "loss": 0.3216,
+      "step": 380
+    },
+    {
+      "epoch": 0.5861538461538461,
+      "grad_norm": 0.3602325778928683,
+      "learning_rate": 1.6071625078187113e-05,
+      "loss": 0.3453,
+      "step": 381
+    },
+    {
+      "epoch": 0.5876923076923077,
+      "grad_norm": 0.3193438829573852,
+      "learning_rate": 1.605240552618398e-05,
+      "loss": 0.3495,
+      "step": 382
+    },
+    {
+      "epoch": 0.5892307692307692,
+      "grad_norm": 0.33548527878831463,
+      "learning_rate": 1.603315062813401e-05,
+      "loss": 0.3417,
+      "step": 383
+    },
+    {
+      "epoch": 0.5907692307692308,
+      "grad_norm": 0.32840411596842495,
+      "learning_rate": 1.601386049648581e-05,
+      "loss": 0.347,
+      "step": 384
+    },
+    {
+      "epoch": 0.5923076923076923,
+      "grad_norm": 0.3298738925919233,
+      "learning_rate": 1.5994535243893742e-05,
+      "loss": 0.329,
+      "step": 385
+    },
+    {
+      "epoch": 0.5938461538461538,
+      "grad_norm": 0.3316001678941656,
+      "learning_rate": 1.5975174983217273e-05,
+      "loss": 0.3374,
+      "step": 386
+    },
+    {
+      "epoch": 0.5953846153846154,
+      "grad_norm": 0.31506047903749107,
+      "learning_rate": 1.5955779827520327e-05,
+      "loss": 0.3286,
+      "step": 387
+    },
+    {
+      "epoch": 0.5969230769230769,
+      "grad_norm": 0.3205561551558476,
+      "learning_rate": 1.5936349890070602e-05,
+      "loss": 0.3439,
+      "step": 388
+    },
+    {
+      "epoch": 0.5984615384615385,
+      "grad_norm": 0.34332344122828284,
+      "learning_rate": 1.5916885284338937e-05,
+      "loss": 0.3498,
+      "step": 389
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3378431634660829,
+      "learning_rate": 1.5897386123998613e-05,
+      "loss": 0.3411,
+      "step": 390
+    },
+    {
+      "epoch": 0.6015384615384616,
+      "grad_norm": 0.3308198558930134,
+      "learning_rate": 1.5877852522924733e-05,
+      "loss": 0.3401,
+      "step": 391
+    },
+    {
+      "epoch": 0.6030769230769231,
+      "grad_norm": 0.3382324720181038,
+      "learning_rate": 1.5858284595193514e-05,
+      "loss": 0.3313,
+      "step": 392
+    },
+    {
+      "epoch": 0.6046153846153847,
+      "grad_norm": 0.3115119958797461,
+      "learning_rate": 1.5838682455081657e-05,
+      "loss": 0.3473,
+      "step": 393
+    },
+    {
+      "epoch": 0.6061538461538462,
+      "grad_norm": 0.345775349619465,
+      "learning_rate": 1.581904621706565e-05,
+      "loss": 0.3416,
+      "step": 394
+    },
+    {
+      "epoch": 0.6076923076923076,
+      "grad_norm": 0.3278035432263032,
+      "learning_rate": 1.5799375995821116e-05,
+      "loss": 0.3402,
+      "step": 395
+    },
+    {
+      "epoch": 0.6092307692307692,
+      "grad_norm": 0.3091426767370126,
+      "learning_rate": 1.577967190622215e-05,
+      "loss": 0.3419,
+      "step": 396
+    },
+    {
+      "epoch": 0.6107692307692307,
+      "grad_norm": 0.32700056228623475,
+      "learning_rate": 1.5759934063340627e-05,
+      "loss": 0.3286,
+      "step": 397
+    },
+    {
+      "epoch": 0.6123076923076923,
+      "grad_norm": 0.31000502760088167,
+      "learning_rate": 1.5740162582445545e-05,
+      "loss": 0.3359,
+      "step": 398
+    },
+    {
+      "epoch": 0.6138461538461538,
+      "grad_norm": 0.315042243493686,
+      "learning_rate": 1.5720357579002346e-05,
+      "loss": 0.3302,
+      "step": 399
+    },
+    {
+      "epoch": 0.6153846153846154,
+      "grad_norm": 0.3562010310358661,
+      "learning_rate": 1.5700519168672248e-05,
+      "loss": 0.3241,
+      "step": 400
+    },
+    {
+      "epoch": 0.6169230769230769,
+      "grad_norm": 0.32548257408496123,
+      "learning_rate": 1.568064746731156e-05,
+      "loss": 0.3396,
+      "step": 401
+    },
+    {
+      "epoch": 0.6184615384615385,
+      "grad_norm": 0.34393891513916103,
+      "learning_rate": 1.5660742590971014e-05,
+      "loss": 0.3344,
+      "step": 402
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3504882415860048,
+      "learning_rate": 1.5640804655895086e-05,
+      "loss": 0.3435,
+      "step": 403
+    },
+    {
+      "epoch": 0.6215384615384615,
+      "grad_norm": 0.2977739667457439,
+      "learning_rate": 1.5620833778521306e-05,
+      "loss": 0.3406,
+      "step": 404
+    },
+    {
+      "epoch": 0.6230769230769231,
+      "grad_norm": 0.3465963236491581,
+      "learning_rate": 1.5600830075479604e-05,
+      "loss": 0.3541,
+      "step": 405
+    },
+    {
+      "epoch": 0.6246153846153846,
+      "grad_norm": 0.3278292046990623,
+      "learning_rate": 1.5580793663591583e-05,
+      "loss": 0.3399,
+      "step": 406
+    },
+    {
+      "epoch": 0.6261538461538462,
+      "grad_norm": 0.3178133040140898,
+      "learning_rate": 1.5560724659869905e-05,
+      "loss": 0.3487,
+      "step": 407
+    },
+    {
+      "epoch": 0.6276923076923077,
+      "grad_norm": 0.3189251567444249,
+      "learning_rate": 1.5540623181517532e-05,
+      "loss": 0.3316,
+      "step": 408
+    },
+    {
+      "epoch": 0.6292307692307693,
+      "grad_norm": 0.3082890284907204,
+      "learning_rate": 1.5520489345927095e-05,
+      "loss": 0.327,
+      "step": 409
+    },
+    {
+      "epoch": 0.6307692307692307,
+      "grad_norm": 0.32991631832198975,
+      "learning_rate": 1.5500323270680194e-05,
+      "loss": 0.3363,
+      "step": 410
+    },
+    {
+      "epoch": 0.6323076923076923,
+      "grad_norm": 0.2992131839708694,
+      "learning_rate": 1.5480125073546705e-05,
+      "loss": 0.338,
+      "step": 411
+    },
+    {
+      "epoch": 0.6338461538461538,
+      "grad_norm": 0.3325641611293239,
+      "learning_rate": 1.5459894872484083e-05,
+      "loss": 0.3227,
+      "step": 412
+    },
+    {
+      "epoch": 0.6353846153846154,
+      "grad_norm": 0.32566576056931756,
+      "learning_rate": 1.5439632785636707e-05,
+      "loss": 0.3455,
+      "step": 413
+    },
+    {
+      "epoch": 0.6369230769230769,
+      "grad_norm": 0.3523675990715312,
+      "learning_rate": 1.5419338931335155e-05,
+      "loss": 0.3427,
+      "step": 414
+    },
+    {
+      "epoch": 0.6384615384615384,
+      "grad_norm": 0.35998620466162407,
+      "learning_rate": 1.539901342809554e-05,
+      "loss": 0.3372,
+      "step": 415
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3446553956298386,
+      "learning_rate": 1.5378656394618788e-05,
+      "loss": 0.3394,
+      "step": 416
+    },
+    {
+      "epoch": 0.6415384615384615,
+      "grad_norm": 0.3616445233684353,
+      "learning_rate": 1.5358267949789968e-05,
+      "loss": 0.3322,
+      "step": 417
+    },
+    {
+      "epoch": 0.6430769230769231,
+      "grad_norm": 0.3639906664590491,
+      "learning_rate": 1.53378482126776e-05,
+      "loss": 0.3468,
+      "step": 418
+    },
+    {
+      "epoch": 0.6446153846153846,
+      "grad_norm": 0.3217847263901737,
+      "learning_rate": 1.5317397302532933e-05,
+      "loss": 0.3386,
+      "step": 419
+    },
+    {
+      "epoch": 0.6461538461538462,
+      "grad_norm": 0.3523419339077219,
+      "learning_rate": 1.529691533878929e-05,
+      "loss": 0.3345,
+      "step": 420
+    },
+    {
+      "epoch": 0.6476923076923077,
+      "grad_norm": 0.29498589810539044,
+      "learning_rate": 1.527640244106133e-05,
+      "loss": 0.3282,
+      "step": 421
+    },
+    {
+      "epoch": 0.6492307692307693,
+      "grad_norm": 0.35484962462726277,
+      "learning_rate": 1.5255858729144368e-05,
+      "loss": 0.3318,
+      "step": 422
+    },
+    {
+      "epoch": 0.6507692307692308,
+      "grad_norm": 0.32079171813409035,
+      "learning_rate": 1.5235284323013674e-05,
+      "loss": 0.3405,
+      "step": 423
+    },
+    {
+      "epoch": 0.6523076923076923,
+      "grad_norm": 0.3160074548522989,
+      "learning_rate": 1.5214679342823786e-05,
+      "loss": 0.347,
+      "step": 424
+    },
+    {
+      "epoch": 0.6538461538461539,
+      "grad_norm": 0.3539240389229003,
+      "learning_rate": 1.5194043908907774e-05,
+      "loss": 0.3323,
+      "step": 425
+    },
+    {
+      "epoch": 0.6553846153846153,
+      "grad_norm": 0.330852711681434,
+      "learning_rate": 1.5173378141776569e-05,
+      "loss": 0.3327,
+      "step": 426
+    },
+    {
+      "epoch": 0.6569230769230769,
+      "grad_norm": 0.322980956210938,
+      "learning_rate": 1.515268216211825e-05,
+      "loss": 0.3382,
+      "step": 427
+    },
+    {
+      "epoch": 0.6584615384615384,
+      "grad_norm": 0.3529073490405514,
+      "learning_rate": 1.5131956090797326e-05,
+      "loss": 0.3277,
+      "step": 428
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3535741007698121,
+      "learning_rate": 1.5111200048854055e-05,
+      "loss": 0.3378,
+      "step": 429
+    },
+    {
+      "epoch": 0.6615384615384615,
+      "grad_norm": 0.3443986070735677,
+      "learning_rate": 1.5090414157503715e-05,
+      "loss": 0.3434,
+      "step": 430
+    },
+    {
+      "epoch": 0.6630769230769231,
+      "grad_norm": 0.3190609653993291,
+      "learning_rate": 1.5069598538135905e-05,
+      "loss": 0.3368,
+      "step": 431
+    },
+    {
+      "epoch": 0.6646153846153846,
+      "grad_norm": 0.3208785969499383,
+      "learning_rate": 1.504875331231384e-05,
+      "loss": 0.3301,
+      "step": 432
+    },
+    {
+      "epoch": 0.6661538461538462,
+      "grad_norm": 0.3405536133260573,
+      "learning_rate": 1.5027878601773633e-05,
+      "loss": 0.3381,
+      "step": 433
+    },
+    {
+      "epoch": 0.6676923076923077,
+      "grad_norm": 0.3235610875812187,
+      "learning_rate": 1.5006974528423585e-05,
+      "loss": 0.348,
+      "step": 434
+    },
+    {
+      "epoch": 0.6692307692307692,
+      "grad_norm": 0.33286382213223575,
+      "learning_rate": 1.4986041214343487e-05,
+      "loss": 0.3298,
+      "step": 435
+    },
+    {
+      "epoch": 0.6707692307692308,
+      "grad_norm": 0.31311886460294547,
+      "learning_rate": 1.4965078781783882e-05,
+      "loss": 0.3347,
+      "step": 436
+    },
+    {
+      "epoch": 0.6723076923076923,
+      "grad_norm": 0.34048661944444236,
+      "learning_rate": 1.494408735316537e-05,
+      "loss": 0.3179,
+      "step": 437
+    },
+    {
+      "epoch": 0.6738461538461539,
+      "grad_norm": 0.30741869868907906,
+      "learning_rate": 1.4923067051077893e-05,
+      "loss": 0.3342,
+      "step": 438
+    },
+    {
+      "epoch": 0.6753846153846154,
+      "grad_norm": 0.384913969400681,
+      "learning_rate": 1.490201799828001e-05,
+      "loss": 0.328,
+      "step": 439
+    },
+    {
+      "epoch": 0.676923076923077,
+      "grad_norm": 0.3252037127931481,
+      "learning_rate": 1.4880940317698182e-05,
+      "loss": 0.3325,
+      "step": 440
+    },
+    {
+      "epoch": 0.6784615384615384,
+      "grad_norm": 0.3381708194506858,
+      "learning_rate": 1.485983413242606e-05,
+      "loss": 0.3347,
+      "step": 441
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3138690227925255,
+      "learning_rate": 1.4838699565723764e-05,
+      "loss": 0.3457,
+      "step": 442
+    },
+    {
+      "epoch": 0.6815384615384615,
+      "grad_norm": 0.31660415974825973,
+      "learning_rate": 1.4817536741017153e-05,
+      "loss": 0.341,
+      "step": 443
+    },
+    {
+      "epoch": 0.683076923076923,
+      "grad_norm": 0.32414107649278284,
+      "learning_rate": 1.479634578189712e-05,
+      "loss": 0.3318,
+      "step": 444
+    },
+    {
+      "epoch": 0.6846153846153846,
+      "grad_norm": 0.33198663699808667,
+      "learning_rate": 1.4775126812118865e-05,
+      "loss": 0.342,
+      "step": 445
+    },
+    {
+      "epoch": 0.6861538461538461,
+      "grad_norm": 0.3574753350792076,
+      "learning_rate": 1.4753879955601162e-05,
+      "loss": 0.3293,
+      "step": 446
+    },
+    {
+      "epoch": 0.6876923076923077,
+      "grad_norm": 0.3039010914154921,
+      "learning_rate": 1.4732605336425651e-05,
+      "loss": 0.3377,
+      "step": 447
+    },
+    {
+      "epoch": 0.6892307692307692,
+      "grad_norm": 0.331147958191396,
+      "learning_rate": 1.4711303078836098e-05,
+      "loss": 0.3341,
+      "step": 448
+    },
+    {
+      "epoch": 0.6907692307692308,
+      "grad_norm": 0.33253847273949044,
+      "learning_rate": 1.4689973307237687e-05,
+      "loss": 0.3246,
+      "step": 449
+    },
+    {
+      "epoch": 0.6923076923076923,
+      "grad_norm": 0.34714104465860907,
+      "learning_rate": 1.466861614619628e-05,
+      "loss": 0.3376,
+      "step": 450
+    },
+    {
+      "epoch": 0.6938461538461539,
+      "grad_norm": 0.3185908753912035,
+      "learning_rate": 1.4647231720437687e-05,
+      "loss": 0.3272,
+      "step": 451
+    },
+    {
+      "epoch": 0.6953846153846154,
+      "grad_norm": 0.3528975380407178,
+      "learning_rate": 1.4625820154846953e-05,
+      "loss": 0.336,
+      "step": 452
+    },
+    {
+      "epoch": 0.696923076923077,
+      "grad_norm": 0.3277896571363917,
+      "learning_rate": 1.4604381574467616e-05,
+      "loss": 0.3306,
+      "step": 453
+    },
+    {
+      "epoch": 0.6984615384615385,
+      "grad_norm": 0.3243374356110855,
+      "learning_rate": 1.4582916104500977e-05,
+      "loss": 0.3412,
+      "step": 454
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.34889136163694917,
+      "learning_rate": 1.4561423870305383e-05,
+      "loss": 0.3471,
+      "step": 455
+    },
+    {
+      "epoch": 0.7015384615384616,
+      "grad_norm": 0.32513754947818535,
+      "learning_rate": 1.4539904997395468e-05,
+      "loss": 0.3432,
+      "step": 456
+    },
+    {
+      "epoch": 0.703076923076923,
+      "grad_norm": 0.3067256440653582,
+      "learning_rate": 1.4518359611441452e-05,
+      "loss": 0.3254,
+      "step": 457
+    },
+    {
+      "epoch": 0.7046153846153846,
+      "grad_norm": 0.32317304945228514,
+      "learning_rate": 1.4496787838268378e-05,
+      "loss": 0.3471,
+      "step": 458
+    },
+    {
+      "epoch": 0.7061538461538461,
+      "grad_norm": 0.3201968707132547,
+      "learning_rate": 1.4475189803855399e-05,
+      "loss": 0.33,
+      "step": 459
+    },
+    {
+      "epoch": 0.7076923076923077,
+      "grad_norm": 0.31117563005728266,
+      "learning_rate": 1.445356563433503e-05,
+      "loss": 0.3241,
+      "step": 460
+    },
+    {
+      "epoch": 0.7092307692307692,
+      "grad_norm": 0.3231141370386228,
+      "learning_rate": 1.4431915455992416e-05,
+      "loss": 0.325,
+      "step": 461
+    },
+    {
+      "epoch": 0.7107692307692308,
+      "grad_norm": 0.3505400578573498,
+      "learning_rate": 1.4410239395264594e-05,
+      "loss": 0.3504,
+      "step": 462
+    },
+    {
+      "epoch": 0.7123076923076923,
+      "grad_norm": 0.3641284162614635,
+      "learning_rate": 1.438853757873975e-05,
+      "loss": 0.3451,
+      "step": 463
+    },
+    {
+      "epoch": 0.7138461538461538,
+      "grad_norm": 0.341362150690221,
+      "learning_rate": 1.4366810133156495e-05,
+      "loss": 0.334,
+      "step": 464
+    },
+    {
+      "epoch": 0.7153846153846154,
+      "grad_norm": 0.3366781323480242,
+      "learning_rate": 1.43450571854031e-05,
+      "loss": 0.3317,
+      "step": 465
+    },
+    {
+      "epoch": 0.7169230769230769,
+      "grad_norm": 0.33305216446331765,
+      "learning_rate": 1.4323278862516774e-05,
+      "loss": 0.3275,
+      "step": 466
+    },
+    {
+      "epoch": 0.7184615384615385,
+      "grad_norm": 0.3463239764810698,
+      "learning_rate": 1.430147529168292e-05,
+      "loss": 0.3275,
+      "step": 467
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.32846628027216124,
+      "learning_rate": 1.4279646600234388e-05,
+      "loss": 0.3319,
+      "step": 468
+    },
+    {
+      "epoch": 0.7215384615384616,
+      "grad_norm": 0.34238219227879857,
+      "learning_rate": 1.4257792915650728e-05,
+      "loss": 0.3293,
+      "step": 469
+    },
+    {
+      "epoch": 0.7230769230769231,
+      "grad_norm": 0.32976197302001725,
+      "learning_rate": 1.4235914365557455e-05,
+      "loss": 0.3237,
+      "step": 470
+    },
+    {
+      "epoch": 0.7246153846153847,
+      "grad_norm": 0.34893869828996976,
+      "learning_rate": 1.4214011077725293e-05,
+      "loss": 0.3419,
+      "step": 471
+    },
+    {
+      "epoch": 0.7261538461538461,
+      "grad_norm": 0.31966372573316476,
+      "learning_rate": 1.4192083180069441e-05,
+      "loss": 0.3427,
+      "step": 472
+    },
+    {
+      "epoch": 0.7276923076923076,
+      "grad_norm": 0.35132035791751987,
+      "learning_rate": 1.4170130800648814e-05,
+      "loss": 0.3395,
+      "step": 473
+    },
+    {
+      "epoch": 0.7292307692307692,
+      "grad_norm": 0.3664547092525681,
+      "learning_rate": 1.4148154067665305e-05,
+      "loss": 0.3426,
+      "step": 474
+    },
+    {
+      "epoch": 0.7307692307692307,
+      "grad_norm": 0.3368990568479692,
+      "learning_rate": 1.4126153109463025e-05,
+      "loss": 0.3276,
+      "step": 475
+    },
+    {
+      "epoch": 0.7323076923076923,
+      "grad_norm": 0.3651280080968852,
+      "learning_rate": 1.410412805452757e-05,
+      "loss": 0.3227,
+      "step": 476
+    },
+    {
+      "epoch": 0.7338461538461538,
+      "grad_norm": 0.34742282148815545,
+      "learning_rate": 1.4082079031485253e-05,
+      "loss": 0.3253,
+      "step": 477
+    },
+    {
+      "epoch": 0.7353846153846154,
+      "grad_norm": 0.32137945130339735,
+      "learning_rate": 1.4060006169102363e-05,
+      "loss": 0.3423,
+      "step": 478
+    },
+    {
+      "epoch": 0.7369230769230769,
+      "grad_norm": 0.34666033781216216,
+      "learning_rate": 1.403790959628441e-05,
+      "loss": 0.3475,
+      "step": 479
+    },
+    {
+      "epoch": 0.7384615384615385,
+      "grad_norm": 0.35142123558854127,
+      "learning_rate": 1.4015789442075376e-05,
+      "loss": 0.3385,
+      "step": 480
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3341424816484901,
+      "learning_rate": 1.3993645835656955e-05,
+      "loss": 0.3389,
+      "step": 481
+    },
+    {
+      "epoch": 0.7415384615384616,
+      "grad_norm": 0.32207944514331105,
+      "learning_rate": 1.3971478906347806e-05,
+      "loss": 0.3259,
+      "step": 482
+    },
+    {
+      "epoch": 0.7430769230769231,
+      "grad_norm": 0.3179744347055682,
+      "learning_rate": 1.394928878360279e-05,
+      "loss": 0.3306,
+      "step": 483
+    },
+    {
+      "epoch": 0.7446153846153846,
+      "grad_norm": 0.3299785758953266,
+      "learning_rate": 1.3927075597012215e-05,
+      "loss": 0.3403,
+      "step": 484
+    },
+    {
+      "epoch": 0.7461538461538462,
+      "grad_norm": 0.3479339850990576,
+      "learning_rate": 1.3904839476301091e-05,
+      "loss": 0.3363,
+      "step": 485
+    },
+    {
+      "epoch": 0.7476923076923077,
+      "grad_norm": 0.3160433072118054,
+      "learning_rate": 1.388258055132835e-05,
+      "loss": 0.3339,
+      "step": 486
+    },
+    {
+      "epoch": 0.7492307692307693,
+      "grad_norm": 0.3168391369647155,
+      "learning_rate": 1.3860298952086118e-05,
+      "loss": 0.3303,
+      "step": 487
+    },
+    {
+      "epoch": 0.7507692307692307,
+      "grad_norm": 0.3255877966163271,
+      "learning_rate": 1.383799480869892e-05,
+      "loss": 0.329,
+      "step": 488
+    },
+    {
+      "epoch": 0.7523076923076923,
+      "grad_norm": 0.33535362647277184,
+      "learning_rate": 1.3815668251422953e-05,
+      "loss": 0.3353,
+      "step": 489
+    },
+    {
+      "epoch": 0.7538461538461538,
+      "grad_norm": 0.3271690638955483,
+      "learning_rate": 1.3793319410645307e-05,
+      "loss": 0.327,
+      "step": 490
+    },
+    {
+      "epoch": 0.7553846153846154,
+      "grad_norm": 0.29308835135249056,
+      "learning_rate": 1.3770948416883205e-05,
+      "loss": 0.3301,
+      "step": 491
+    },
+    {
+      "epoch": 0.7569230769230769,
+      "grad_norm": 0.33463441859903204,
+      "learning_rate": 1.3748555400783245e-05,
+      "loss": 0.3372,
+      "step": 492
+    },
+    {
+      "epoch": 0.7584615384615384,
+      "grad_norm": 0.30324390052773076,
+      "learning_rate": 1.3726140493120639e-05,
+      "loss": 0.3284,
+      "step": 493
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3233241376039668,
+      "learning_rate": 1.3703703824798438e-05,
+      "loss": 0.3263,
+      "step": 494
+    },
+    {
+      "epoch": 0.7615384615384615,
+      "grad_norm": 0.31728343223649114,
+      "learning_rate": 1.3681245526846782e-05,
+      "loss": 0.3352,
+      "step": 495
+    },
+    {
+      "epoch": 0.7630769230769231,
+      "grad_norm": 0.3214067215862394,
+      "learning_rate": 1.3658765730422126e-05,
+      "loss": 0.3425,
+      "step": 496
+    },
+    {
+      "epoch": 0.7646153846153846,
+      "grad_norm": 0.31943063401432914,
+      "learning_rate": 1.3636264566806473e-05,
+      "loss": 0.3401,
+      "step": 497
+    },
+    {
+      "epoch": 0.7661538461538462,
+      "grad_norm": 0.30418243324485683,
+      "learning_rate": 1.3613742167406614e-05,
+      "loss": 0.3372,
+      "step": 498
+    },
+    {
+      "epoch": 0.7676923076923077,
+      "grad_norm": 0.34821005797159027,
+      "learning_rate": 1.3591198663753358e-05,
+      "loss": 0.3217,
+      "step": 499
+    },
+    {
+      "epoch": 0.7692307692307693,
+      "grad_norm": 0.3087089373776307,
+      "learning_rate": 1.3568634187500762e-05,
+      "loss": 0.3283,
+      "step": 500
+    },
+    {
+      "epoch": 0.7707692307692308,
+      "grad_norm": 0.30635526513225797,
+      "learning_rate": 1.3546048870425356e-05,
+      "loss": 0.3233,
+      "step": 501
+    },
+    {
+      "epoch": 0.7723076923076924,
+      "grad_norm": 0.3276774808261006,
+      "learning_rate": 1.3523442844425393e-05,
+      "loss": 0.3413,
+      "step": 502
+    },
+    {
+      "epoch": 0.7738461538461539,
+      "grad_norm": 0.3148280444491086,
+      "learning_rate": 1.3500816241520059e-05,
+      "loss": 0.3315,
+      "step": 503
+    },
+    {
+      "epoch": 0.7753846153846153,
+      "grad_norm": 0.32807106888422827,
+      "learning_rate": 1.3478169193848705e-05,
+      "loss": 0.327,
+      "step": 504
+    },
+    {
+      "epoch": 0.7769230769230769,
+      "grad_norm": 0.3292383482401427,
+      "learning_rate": 1.3455501833670089e-05,
+      "loss": 0.3336,
+      "step": 505
+    },
+    {
+      "epoch": 0.7784615384615384,
+      "grad_norm": 0.3121711347510858,
+      "learning_rate": 1.3432814293361585e-05,
+      "loss": 0.3343,
+      "step": 506
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.31726922745700675,
+      "learning_rate": 1.3410106705418424e-05,
+      "loss": 0.3268,
+      "step": 507
+    },
+    {
+      "epoch": 0.7815384615384615,
+      "grad_norm": 0.3369144300435686,
+      "learning_rate": 1.3387379202452917e-05,
+      "loss": 0.3379,
+      "step": 508
+    },
+    {
+      "epoch": 0.7830769230769231,
+      "grad_norm": 0.3420593507280525,
+      "learning_rate": 1.3364631917193671e-05,
+      "loss": 0.3312,
+      "step": 509
+    },
+    {
+      "epoch": 0.7846153846153846,
+      "grad_norm": 0.3088403155613573,
+      "learning_rate": 1.3341864982484828e-05,
+      "loss": 0.3376,
+      "step": 510
+    },
+    {
+      "epoch": 0.7861538461538462,
+      "grad_norm": 0.3372304602489395,
+      "learning_rate": 1.3319078531285286e-05,
+      "loss": 0.3396,
+      "step": 511
+    },
+    {
+      "epoch": 0.7876923076923077,
+      "grad_norm": 0.3556448458865266,
+      "learning_rate": 1.329627269666791e-05,
+      "loss": 0.3259,
+      "step": 512
+    },
+    {
+      "epoch": 0.7892307692307692,
+      "grad_norm": 0.32239867500588243,
+      "learning_rate": 1.3273447611818768e-05,
+      "loss": 0.3334,
+      "step": 513
+    },
+    {
+      "epoch": 0.7907692307692308,
+      "grad_norm": 0.34313378935169225,
+      "learning_rate": 1.3250603410036356e-05,
+      "loss": 0.3278,
+      "step": 514
+    },
+    {
+      "epoch": 0.7923076923076923,
+      "grad_norm": 0.3130455306524999,
+      "learning_rate": 1.3227740224730799e-05,
+      "loss": 0.3227,
+      "step": 515
+    },
+    {
+      "epoch": 0.7938461538461539,
+      "grad_norm": 0.35119933240152384,
+      "learning_rate": 1.3204858189423097e-05,
+      "loss": 0.3328,
+      "step": 516
+    },
+    {
+      "epoch": 0.7953846153846154,
+      "grad_norm": 0.31600394128037407,
+      "learning_rate": 1.3181957437744333e-05,
+      "loss": 0.3395,
+      "step": 517
+    },
+    {
+      "epoch": 0.796923076923077,
+      "grad_norm": 0.3479657346532963,
+      "learning_rate": 1.3159038103434889e-05,
+      "loss": 0.3483,
+      "step": 518
+    },
+    {
+      "epoch": 0.7984615384615384,
+      "grad_norm": 0.3206870512019153,
+      "learning_rate": 1.3136100320343674e-05,
+      "loss": 0.3402,
+      "step": 519
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.31675024233834703,
+      "learning_rate": 1.3113144222427334e-05,
+      "loss": 0.3163,
+      "step": 520
+    },
+    {
+      "epoch": 0.8015384615384615,
+      "grad_norm": 0.34122270739362187,
+      "learning_rate": 1.3090169943749475e-05,
+      "loss": 0.3352,
+      "step": 521
+    },
+    {
+      "epoch": 0.803076923076923,
+      "grad_norm": 0.28611701873915973,
+      "learning_rate": 1.3067177618479883e-05,
+      "loss": 0.3405,
+      "step": 522
+    },
+    {
+      "epoch": 0.8046153846153846,
+      "grad_norm": 0.327896071980244,
+      "learning_rate": 1.3044167380893726e-05,
+      "loss": 0.3276,
+      "step": 523
+    },
+    {
+      "epoch": 0.8061538461538461,
+      "grad_norm": 0.31845280215584937,
+      "learning_rate": 1.3021139365370787e-05,
+      "loss": 0.348,
+      "step": 524
+    },
+    {
+      "epoch": 0.8076923076923077,
+      "grad_norm": 0.32071766074021624,
+      "learning_rate": 1.2998093706394676e-05,
+      "loss": 0.3491,
+      "step": 525
+    },
+    {
+      "epoch": 0.8092307692307692,
+      "grad_norm": 0.3233443240330192,
+      "learning_rate": 1.297503053855203e-05,
+      "loss": 0.336,
+      "step": 526
+    },
+    {
+      "epoch": 0.8107692307692308,
+      "grad_norm": 0.35467566330457945,
+      "learning_rate": 1.295194999653175e-05,
+      "loss": 0.3409,
+      "step": 527
+    },
+    {
+      "epoch": 0.8123076923076923,
+      "grad_norm": 0.316307679891141,
+      "learning_rate": 1.292885221512419e-05,
+      "loss": 0.326,
+      "step": 528
+    },
+    {
+      "epoch": 0.8138461538461539,
+      "grad_norm": 0.314005066699375,
+      "learning_rate": 1.2905737329220394e-05,
+      "loss": 0.3151,
+      "step": 529
+    },
+    {
+      "epoch": 0.8153846153846154,
+      "grad_norm": 0.3077830345752374,
+      "learning_rate": 1.2882605473811282e-05,
+      "loss": 0.3329,
+      "step": 530
+    },
+    {
+      "epoch": 0.816923076923077,
+      "grad_norm": 0.32958916574455754,
+      "learning_rate": 1.2859456783986892e-05,
+      "loss": 0.3287,
+      "step": 531
+    },
+    {
+      "epoch": 0.8184615384615385,
+      "grad_norm": 0.3067485099829722,
+      "learning_rate": 1.2836291394935568e-05,
+      "loss": 0.3309,
+      "step": 532
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3104385024448171,
+      "learning_rate": 1.2813109441943166e-05,
+      "loss": 0.334,
+      "step": 533
+    },
+    {
+      "epoch": 0.8215384615384616,
+      "grad_norm": 0.31618015424262574,
+      "learning_rate": 1.2789911060392295e-05,
+      "loss": 0.3337,
+      "step": 534
+    },
+    {
+      "epoch": 0.823076923076923,
+      "grad_norm": 0.32038146009037033,
+      "learning_rate": 1.2766696385761494e-05,
+      "loss": 0.3304,
+      "step": 535
+    },
+    {
+      "epoch": 0.8246153846153846,
+      "grad_norm": 0.30932611044623504,
+      "learning_rate": 1.274346555362446e-05,
+      "loss": 0.3314,
+      "step": 536
+    },
+    {
+      "epoch": 0.8261538461538461,
+      "grad_norm": 0.3153016831527301,
+      "learning_rate": 1.2720218699649243e-05,
+      "loss": 0.3439,
+      "step": 537
+    },
+    {
+      "epoch": 0.8276923076923077,
+      "grad_norm": 0.3019690928384475,
+      "learning_rate": 1.269695595959747e-05,
+      "loss": 0.3338,
+      "step": 538
+    },
+    {
+      "epoch": 0.8292307692307692,
+      "grad_norm": 0.29367927565138535,
+      "learning_rate": 1.2673677469323532e-05,
+      "loss": 0.3287,
+      "step": 539
+    },
+    {
+      "epoch": 0.8307692307692308,
+      "grad_norm": 0.3216958588833604,
+      "learning_rate": 1.2650383364773812e-05,
+      "loss": 0.32,
+      "step": 540
+    },
+    {
+      "epoch": 0.8323076923076923,
+      "grad_norm": 0.33236242408977856,
+      "learning_rate": 1.262707378198587e-05,
+      "loss": 0.3176,
+      "step": 541
+    },
+    {
+      "epoch": 0.8338461538461538,
+      "grad_norm": 0.29093819677752464,
+      "learning_rate": 1.2603748857087668e-05,
+      "loss": 0.337,
+      "step": 542
+    },
+    {
+      "epoch": 0.8353846153846154,
+      "grad_norm": 0.30218052941466567,
+      "learning_rate": 1.258040872629676e-05,
+      "loss": 0.3268,
+      "step": 543
+    },
+    {
+      "epoch": 0.8369230769230769,
+      "grad_norm": 0.31560988459280653,
+      "learning_rate": 1.2557053525919503e-05,
+      "loss": 0.3177,
+      "step": 544
+    },
+    {
+      "epoch": 0.8384615384615385,
+      "grad_norm": 0.3396123824016913,
+      "learning_rate": 1.2533683392350264e-05,
+      "loss": 0.3337,
+      "step": 545
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.30587015627634234,
+      "learning_rate": 1.2510298462070619e-05,
+      "loss": 0.3357,
+      "step": 546
+    },
+    {
+      "epoch": 0.8415384615384616,
+      "grad_norm": 0.314387767986292,
+      "learning_rate": 1.2486898871648552e-05,
+      "loss": 0.3365,
+      "step": 547
+    },
+    {
+      "epoch": 0.8430769230769231,
+      "grad_norm": 0.31389305994558653,
+      "learning_rate": 1.2463484757737663e-05,
+      "loss": 0.3412,
+      "step": 548
+    },
+    {
+      "epoch": 0.8446153846153847,
+      "grad_norm": 0.3060685974378655,
+      "learning_rate": 1.2440056257076376e-05,
+      "loss": 0.3272,
+      "step": 549
+    },
+    {
+      "epoch": 0.8461538461538461,
+      "grad_norm": 0.30434307224961593,
+      "learning_rate": 1.241661350648713e-05,
+      "loss": 0.3284,
+      "step": 550
+    },
+    {
+      "epoch": 0.8476923076923077,
+      "grad_norm": 0.3132509005169485,
+      "learning_rate": 1.2393156642875579e-05,
+      "loss": 0.3304,
+      "step": 551
+    },
+    {
+      "epoch": 0.8492307692307692,
+      "grad_norm": 0.31509089282351654,
+      "learning_rate": 1.2369685803229802e-05,
+      "loss": 0.3352,
+      "step": 552
+    },
+    {
+      "epoch": 0.8507692307692307,
+      "grad_norm": 0.3285296432652221,
+      "learning_rate": 1.2346201124619502e-05,
+      "loss": 0.3278,
+      "step": 553
+    },
+    {
+      "epoch": 0.8523076923076923,
+      "grad_norm": 0.31718979342496284,
+      "learning_rate": 1.2322702744195192e-05,
+      "loss": 0.3335,
+      "step": 554
+    },
+    {
+      "epoch": 0.8538461538461538,
+      "grad_norm": 0.31312636222764384,
+      "learning_rate": 1.2299190799187405e-05,
+      "loss": 0.3268,
+      "step": 555
+    },
+    {
+      "epoch": 0.8553846153846154,
+      "grad_norm": 0.3112656909831258,
+      "learning_rate": 1.22756654269059e-05,
+      "loss": 0.3245,
+      "step": 556
+    },
+    {
+      "epoch": 0.8569230769230769,
+      "grad_norm": 0.32161444094635033,
+      "learning_rate": 1.2252126764738845e-05,
+      "loss": 0.3244,
+      "step": 557
+    },
+    {
+      "epoch": 0.8584615384615385,
+      "grad_norm": 0.3218845643109017,
+      "learning_rate": 1.222857495015202e-05,
+      "loss": 0.3342,
+      "step": 558
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.31065573649293504,
+      "learning_rate": 1.2205010120688012e-05,
+      "loss": 0.3302,
+      "step": 559
+    },
+    {
+      "epoch": 0.8615384615384616,
+      "grad_norm": 0.324391159225024,
+      "learning_rate": 1.2181432413965428e-05,
+      "loss": 0.3497,
+      "step": 560
+    },
+    {
+      "epoch": 0.8630769230769231,
+      "grad_norm": 0.3043829743503028,
+      "learning_rate": 1.2157841967678064e-05,
+      "loss": 0.3305,
+      "step": 561
+    },
+    {
+      "epoch": 0.8646153846153846,
+      "grad_norm": 0.31249008455235144,
+      "learning_rate": 1.2134238919594122e-05,
+      "loss": 0.3249,
+      "step": 562
+    },
+    {
+      "epoch": 0.8661538461538462,
+      "grad_norm": 0.3196330879564442,
+      "learning_rate": 1.2110623407555398e-05,
+      "loss": 0.3369,
+      "step": 563
+    },
+    {
+      "epoch": 0.8676923076923077,
+      "grad_norm": 0.3185290427871534,
+      "learning_rate": 1.2086995569476474e-05,
+      "loss": 0.3288,
+      "step": 564
+    },
+    {
+      "epoch": 0.8692307692307693,
+      "grad_norm": 0.2870092449390502,
+      "learning_rate": 1.2063355543343925e-05,
+      "loss": 0.3346,
+      "step": 565
+    },
+    {
+      "epoch": 0.8707692307692307,
+      "grad_norm": 0.3257019255068011,
+      "learning_rate": 1.2039703467215489e-05,
+      "loss": 0.3262,
+      "step": 566
+    },
+    {
+      "epoch": 0.8723076923076923,
+      "grad_norm": 0.30786304302506623,
+      "learning_rate": 1.2016039479219293e-05,
+      "loss": 0.3241,
+      "step": 567
+    },
+    {
+      "epoch": 0.8738461538461538,
+      "grad_norm": 0.2871242511667203,
+      "learning_rate": 1.1992363717553015e-05,
+      "loss": 0.3338,
+      "step": 568
+    },
+    {
+      "epoch": 0.8753846153846154,
+      "grad_norm": 0.31399555825061976,
+      "learning_rate": 1.1968676320483103e-05,
+      "loss": 0.3214,
+      "step": 569
+    },
+    {
+      "epoch": 0.8769230769230769,
+      "grad_norm": 0.3206608664636182,
+      "learning_rate": 1.194497742634395e-05,
+      "loss": 0.327,
+      "step": 570
+    },
+    {
+      "epoch": 0.8784615384615385,
+      "grad_norm": 0.29015010647347106,
+      "learning_rate": 1.1921267173537085e-05,
+      "loss": 0.3236,
+      "step": 571
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3063448933173235,
+      "learning_rate": 1.1897545700530387e-05,
+      "loss": 0.3397,
+      "step": 572
+    },
+    {
+      "epoch": 0.8815384615384615,
+      "grad_norm": 0.3110184425830958,
+      "learning_rate": 1.187381314585725e-05,
+      "loss": 0.3213,
+      "step": 573
+    },
+    {
+      "epoch": 0.8830769230769231,
+      "grad_norm": 0.3457162836672877,
+      "learning_rate": 1.1850069648115785e-05,
+      "loss": 0.3397,
+      "step": 574
+    },
+    {
+      "epoch": 0.8846153846153846,
+      "grad_norm": 0.331003030927215,
+      "learning_rate": 1.1826315345968014e-05,
+      "loss": 0.3206,
+      "step": 575
+    },
+    {
+      "epoch": 0.8861538461538462,
+      "grad_norm": 0.315149509773091,
+      "learning_rate": 1.180255037813906e-05,
+      "loss": 0.33,
+      "step": 576
+    },
+    {
+      "epoch": 0.8876923076923077,
+      "grad_norm": 0.340006308605447,
+      "learning_rate": 1.1778774883416325e-05,
+      "loss": 0.3231,
+      "step": 577
+    },
+    {
+      "epoch": 0.8892307692307693,
+      "grad_norm": 0.3169393379879127,
+      "learning_rate": 1.1754989000648693e-05,
+      "loss": 0.3312,
+      "step": 578
+    },
+    {
+      "epoch": 0.8907692307692308,
+      "grad_norm": 0.30576326286204264,
+      "learning_rate": 1.1731192868745716e-05,
+      "loss": 0.3292,
+      "step": 579
+    },
+    {
+      "epoch": 0.8923076923076924,
+      "grad_norm": 0.29835580565844005,
+      "learning_rate": 1.1707386626676798e-05,
+      "loss": 0.3341,
+      "step": 580
+    },
+    {
+      "epoch": 0.8938461538461538,
+      "grad_norm": 0.30635228645222695,
+      "learning_rate": 1.1683570413470384e-05,
+      "loss": 0.3354,
+      "step": 581
+    },
+    {
+      "epoch": 0.8953846153846153,
+      "grad_norm": 0.3122003041641902,
+      "learning_rate": 1.1659744368213159e-05,
+      "loss": 0.3355,
+      "step": 582
+    },
+    {
+      "epoch": 0.8969230769230769,
+      "grad_norm": 0.29808807782765917,
+      "learning_rate": 1.163590863004922e-05,
+      "loss": 0.309,
+      "step": 583
+    },
+    {
+      "epoch": 0.8984615384615384,
+      "grad_norm": 0.31722094162289693,
+      "learning_rate": 1.1612063338179269e-05,
+      "loss": 0.3345,
+      "step": 584
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.34343706386507955,
+      "learning_rate": 1.1588208631859808e-05,
+      "loss": 0.3264,
+      "step": 585
+    },
+    {
+      "epoch": 0.9015384615384615,
+      "grad_norm": 0.3561421859445545,
+      "learning_rate": 1.156434465040231e-05,
+      "loss": 0.3221,
+      "step": 586
+    },
+    {
+      "epoch": 0.9030769230769231,
+      "grad_norm": 0.3122333077895763,
+      "learning_rate": 1.154047153317243e-05,
+      "loss": 0.334,
+      "step": 587
+    },
+    {
+      "epoch": 0.9046153846153846,
+      "grad_norm": 0.36462307583103276,
+      "learning_rate": 1.1516589419589159e-05,
+      "loss": 0.3383,
+      "step": 588
+    },
+    {
+      "epoch": 0.9061538461538462,
+      "grad_norm": 0.3787798255981821,
+      "learning_rate": 1.1492698449124042e-05,
+      "loss": 0.3236,
+      "step": 589
+    },
+    {
+      "epoch": 0.9076923076923077,
+      "grad_norm": 0.31052416517739034,
+      "learning_rate": 1.1468798761300335e-05,
+      "loss": 0.328,
+      "step": 590
+    },
+    {
+      "epoch": 0.9092307692307692,
+      "grad_norm": 0.34670105196219847,
+      "learning_rate": 1.1444890495692214e-05,
+      "loss": 0.32,
+      "step": 591
+    },
+    {
+      "epoch": 0.9107692307692308,
+      "grad_norm": 0.3435408690029312,
+      "learning_rate": 1.1420973791923941e-05,
+      "loss": 0.3179,
+      "step": 592
+    },
+    {
+      "epoch": 0.9123076923076923,
+      "grad_norm": 0.34535921979706546,
+      "learning_rate": 1.1397048789669061e-05,
+      "loss": 0.3202,
+      "step": 593
+    },
+    {
+      "epoch": 0.9138461538461539,
+      "grad_norm": 0.34135586974992327,
+      "learning_rate": 1.1373115628649582e-05,
+      "loss": 0.3377,
+      "step": 594
+    },
+    {
+      "epoch": 0.9153846153846154,
+      "grad_norm": 0.3340598746124593,
+      "learning_rate": 1.1349174448635158e-05,
+      "loss": 0.3301,
+      "step": 595
+    },
+    {
+      "epoch": 0.916923076923077,
+      "grad_norm": 0.31048990721325737,
+      "learning_rate": 1.1325225389442278e-05,
+      "loss": 0.3409,
+      "step": 596
+    },
+    {
+      "epoch": 0.9184615384615384,
+      "grad_norm": 0.32319816702516574,
+      "learning_rate": 1.1301268590933434e-05,
+      "loss": 0.3334,
+      "step": 597
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.32377592283674955,
+      "learning_rate": 1.1277304193016332e-05,
+      "loss": 0.3223,
+      "step": 598
+    },
+    {
+      "epoch": 0.9215384615384615,
+      "grad_norm": 0.32788798640022937,
+      "learning_rate": 1.1253332335643043e-05,
+      "loss": 0.3277,
+      "step": 599
+    },
+    {
+      "epoch": 0.9230769230769231,
+      "grad_norm": 0.3645233665166125,
+      "learning_rate": 1.1229353158809216e-05,
+      "loss": 0.3224,
+      "step": 600
+    },
+    {
+      "epoch": 0.9246153846153846,
+      "grad_norm": 0.31503479960191505,
+      "learning_rate": 1.1205366802553231e-05,
+      "loss": 0.3339,
+      "step": 601
+    },
+    {
+      "epoch": 0.9261538461538461,
+      "grad_norm": 0.331938769447499,
+      "learning_rate": 1.118137340695541e-05,
+      "loss": 0.3365,
+      "step": 602
+    },
+    {
+      "epoch": 0.9276923076923077,
+      "grad_norm": 0.36305216542449914,
+      "learning_rate": 1.1157373112137171e-05,
+      "loss": 0.3135,
+      "step": 603
+    },
+    {
+      "epoch": 0.9292307692307692,
+      "grad_norm": 0.29524307547847034,
+      "learning_rate": 1.1133366058260232e-05,
+      "loss": 0.3342,
+      "step": 604
+    },
+    {
+      "epoch": 0.9307692307692308,
+      "grad_norm": 0.3349025908800711,
+      "learning_rate": 1.1109352385525782e-05,
+      "loss": 0.3406,
+      "step": 605
+    },
+    {
+      "epoch": 0.9323076923076923,
+      "grad_norm": 0.3016000039609564,
+      "learning_rate": 1.1085332234173664e-05,
+      "loss": 0.3224,
+      "step": 606
+    },
+    {
+      "epoch": 0.9338461538461539,
+      "grad_norm": 0.334743615461434,
+      "learning_rate": 1.106130574448156e-05,
+      "loss": 0.3147,
+      "step": 607
+    },
+    {
+      "epoch": 0.9353846153846154,
+      "grad_norm": 0.3102917595170546,
+      "learning_rate": 1.1037273056764157e-05,
+      "loss": 0.3309,
+      "step": 608
+    },
+    {
+      "epoch": 0.936923076923077,
+      "grad_norm": 0.3102194233828966,
+      "learning_rate": 1.1013234311372353e-05,
+      "loss": 0.3217,
+      "step": 609
+    },
+    {
+      "epoch": 0.9384615384615385,
+      "grad_norm": 0.3250678456241659,
+      "learning_rate": 1.0989189648692408e-05,
+      "loss": 0.3279,
+      "step": 610
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2966650578987603,
+      "learning_rate": 1.0965139209145153e-05,
+      "loss": 0.3183,
+      "step": 611
+    },
+    {
+      "epoch": 0.9415384615384615,
+      "grad_norm": 0.3251521973725705,
+      "learning_rate": 1.0941083133185146e-05,
+      "loss": 0.3186,
+      "step": 612
+    },
+    {
+      "epoch": 0.943076923076923,
+      "grad_norm": 0.31455203366608475,
+      "learning_rate": 1.0917021561299864e-05,
+      "loss": 0.3278,
+      "step": 613
+    },
+    {
+      "epoch": 0.9446153846153846,
+      "grad_norm": 0.296941621015581,
+      "learning_rate": 1.089295463400888e-05,
+      "loss": 0.3298,
+      "step": 614
+    },
+    {
+      "epoch": 0.9461538461538461,
+      "grad_norm": 0.3103348965824328,
+      "learning_rate": 1.0868882491863048e-05,
+      "loss": 0.327,
+      "step": 615
+    },
+    {
+      "epoch": 0.9476923076923077,
+      "grad_norm": 0.3046972829370567,
+      "learning_rate": 1.0844805275443673e-05,
+      "loss": 0.3314,
+      "step": 616
+    },
+    {
+      "epoch": 0.9492307692307692,
+      "grad_norm": 0.3098146820614246,
+      "learning_rate": 1.0820723125361685e-05,
+      "loss": 0.3278,
+      "step": 617
+    },
+    {
+      "epoch": 0.9507692307692308,
+      "grad_norm": 0.30699825622994825,
+      "learning_rate": 1.0796636182256846e-05,
+      "loss": 0.3263,
+      "step": 618
+    },
+    {
+      "epoch": 0.9523076923076923,
+      "grad_norm": 0.31333681919188466,
+      "learning_rate": 1.077254458679689e-05,
+      "loss": 0.3219,
+      "step": 619
+    },
+    {
+      "epoch": 0.9538461538461539,
+      "grad_norm": 0.32070135634319324,
+      "learning_rate": 1.074844847967673e-05,
+      "loss": 0.3363,
+      "step": 620
+    },
+    {
+      "epoch": 0.9553846153846154,
+      "grad_norm": 0.28389243504267636,
+      "learning_rate": 1.0724348001617626e-05,
+      "loss": 0.3235,
+      "step": 621
+    },
+    {
+      "epoch": 0.9569230769230769,
+      "grad_norm": 0.31963587873691246,
+      "learning_rate": 1.0700243293366365e-05,
+      "loss": 0.3415,
+      "step": 622
+    },
+    {
+      "epoch": 0.9584615384615385,
+      "grad_norm": 0.32359664500198776,
+      "learning_rate": 1.0676134495694439e-05,
+      "loss": 0.3139,
+      "step": 623
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3031597251286149,
+      "learning_rate": 1.0652021749397216e-05,
+      "loss": 0.3152,
+      "step": 624
+    },
+    {
+      "epoch": 0.9615384615384616,
+      "grad_norm": 0.29998909574751775,
+      "learning_rate": 1.0627905195293135e-05,
+      "loss": 0.3201,
+      "step": 625
+    },
+    {
+      "epoch": 0.963076923076923,
+      "grad_norm": 0.3030414628868554,
+      "learning_rate": 1.0603784974222862e-05,
+      "loss": 0.3309,
+      "step": 626
+    },
+    {
+      "epoch": 0.9646153846153847,
+      "grad_norm": 0.31264131381498866,
+      "learning_rate": 1.0579661227048484e-05,
+      "loss": 0.3313,
+      "step": 627
+    },
+    {
+      "epoch": 0.9661538461538461,
+      "grad_norm": 0.3013874384696025,
+      "learning_rate": 1.0555534094652675e-05,
+      "loss": 0.3178,
+      "step": 628
+    },
+    {
+      "epoch": 0.9676923076923077,
+      "grad_norm": 0.3248209351600733,
+      "learning_rate": 1.0531403717937888e-05,
+      "loss": 0.3285,
+      "step": 629
+    },
+    {
+      "epoch": 0.9692307692307692,
+      "grad_norm": 0.29044342641824206,
+      "learning_rate": 1.0507270237825513e-05,
+      "loss": 0.3359,
+      "step": 630
+    },
+    {
+      "epoch": 0.9707692307692307,
+      "grad_norm": 0.32732965806122305,
+      "learning_rate": 1.0483133795255072e-05,
+      "loss": 0.3128,
+      "step": 631
+    },
+    {
+      "epoch": 0.9723076923076923,
+      "grad_norm": 0.29247913151654736,
+      "learning_rate": 1.045899453118338e-05,
+      "loss": 0.3268,
+      "step": 632
+    },
+    {
+      "epoch": 0.9738461538461538,
+      "grad_norm": 0.3127738976329768,
+      "learning_rate": 1.0434852586583737e-05,
+      "loss": 0.3311,
+      "step": 633
+    },
+    {
+      "epoch": 0.9753846153846154,
+      "grad_norm": 0.32245069255148034,
+      "learning_rate": 1.0410708102445091e-05,
+      "loss": 0.3329,
+      "step": 634
+    },
+    {
+      "epoch": 0.9769230769230769,
+      "grad_norm": 0.2961723521203939,
+      "learning_rate": 1.0386561219771222e-05,
+      "loss": 0.3313,
+      "step": 635
+    },
+    {
+      "epoch": 0.9784615384615385,
+      "grad_norm": 0.3164765116267267,
+      "learning_rate": 1.0362412079579925e-05,
+      "loss": 0.3271,
+      "step": 636
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.30693337679851157,
+      "learning_rate": 1.0338260822902166e-05,
+      "loss": 0.3068,
+      "step": 637
+    },
+    {
+      "epoch": 0.9815384615384616,
+      "grad_norm": 0.309847879496408,
+      "learning_rate": 1.0314107590781284e-05,
+      "loss": 0.3146,
+      "step": 638
+    },
+    {
+      "epoch": 0.9830769230769231,
+      "grad_norm": 0.3441561954467763,
+      "learning_rate": 1.0289952524272147e-05,
+      "loss": 0.3275,
+      "step": 639
+    },
+    {
+      "epoch": 0.9846153846153847,
+      "grad_norm": 0.32734844589039386,
+      "learning_rate": 1.0265795764440335e-05,
+      "loss": 0.3307,
+      "step": 640
+    },
+    {
+      "epoch": 0.9861538461538462,
+      "grad_norm": 0.29131479550823436,
+      "learning_rate": 1.0241637452361323e-05,
+      "loss": 0.3153,
+      "step": 641
+    },
+    {
+      "epoch": 0.9876923076923076,
+      "grad_norm": 0.2927444977942452,
+      "learning_rate": 1.0217477729119648e-05,
+      "loss": 0.308,
+      "step": 642
+    },
+    {
+      "epoch": 0.9892307692307692,
+      "grad_norm": 0.31859817571165966,
+      "learning_rate": 1.0193316735808085e-05,
+      "loss": 0.3237,
+      "step": 643
+    },
+    {
+      "epoch": 0.9907692307692307,
+      "grad_norm": 0.3143106412699962,
+      "learning_rate": 1.0169154613526831e-05,
+      "loss": 0.3211,
+      "step": 644
+    },
+    {
+      "epoch": 0.9923076923076923,
+      "grad_norm": 0.30834730776946945,
+      "learning_rate": 1.0144991503382676e-05,
+      "loss": 0.3246,
+      "step": 645
+    },
+    {
+      "epoch": 0.9938461538461538,
+      "grad_norm": 0.2967696610141374,
+      "learning_rate": 1.0120827546488175e-05,
+      "loss": 0.3279,
+      "step": 646
+    },
+    {
+      "epoch": 0.9953846153846154,
+      "grad_norm": 0.3197694465778131,
+      "learning_rate": 1.0096662883960833e-05,
+      "loss": 0.3188,
+      "step": 647
+    },
+    {
+      "epoch": 0.9969230769230769,
+      "grad_norm": 0.3033818619276967,
+      "learning_rate": 1.0072497656922266e-05,
+      "loss": 0.3303,
+      "step": 648
+    },
+    {
+      "epoch": 0.9984615384615385,
+      "grad_norm": 0.3239836485410687,
+      "learning_rate": 1.0048332006497406e-05,
+      "loss": 0.3165,
+      "step": 649
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.3016364748263467,
+      "learning_rate": 1.0024166073813634e-05,
+      "loss": 0.3191,
+      "step": 650
+    },
+    {
+      "epoch": 1.0015384615384615,
+      "grad_norm": 0.34732331187804666,
+      "learning_rate": 1e-05,
+      "loss": 0.28,
+      "step": 651
+    },
+    {
+      "epoch": 1.003076923076923,
+      "grad_norm": 0.32444904819206183,
+      "learning_rate": 9.975833926186367e-06,
+      "loss": 0.2838,
+      "step": 652
+    },
+    {
+      "epoch": 1.0046153846153847,
+      "grad_norm": 0.3422552839442046,
+      "learning_rate": 9.951667993502599e-06,
+      "loss": 0.2926,
+      "step": 653
+    },
+    {
+      "epoch": 1.0061538461538462,
+      "grad_norm": 0.3320499383753382,
+      "learning_rate": 9.927502343077732e-06,
+      "loss": 0.2734,
+      "step": 654
+    },
+    {
+      "epoch": 1.0076923076923077,
+      "grad_norm": 0.3604432577698012,
+      "learning_rate": 9.903337116039172e-06,
+      "loss": 0.2867,
+      "step": 655
+    },
+    {
+      "epoch": 1.0092307692307692,
+      "grad_norm": 0.34312310604563834,
+      "learning_rate": 9.879172453511827e-06,
+      "loss": 0.2872,
+      "step": 656
+    },
+    {
+      "epoch": 1.0107692307692309,
+      "grad_norm": 0.33513487457394814,
+      "learning_rate": 9.855008496617326e-06,
+      "loss": 0.2784,
+      "step": 657
+    },
+    {
+      "epoch": 1.0123076923076924,
+      "grad_norm": 0.33806408587135317,
+      "learning_rate": 9.830845386473169e-06,
+      "loss": 0.2873,
+      "step": 658
+    },
+    {
+      "epoch": 1.0138461538461538,
+      "grad_norm": 0.3377539416935379,
+      "learning_rate": 9.806683264191916e-06,
+      "loss": 0.2759,
+      "step": 659
+    },
+    {
+      "epoch": 1.0153846153846153,
+      "grad_norm": 0.33051105629103283,
+      "learning_rate": 9.782522270880354e-06,
+      "loss": 0.285,
+      "step": 660
+    },
+    {
+      "epoch": 1.0169230769230768,
+      "grad_norm": 0.33144527590601386,
+      "learning_rate": 9.75836254763868e-06,
+      "loss": 0.2845,
+      "step": 661
+    },
+    {
+      "epoch": 1.0184615384615385,
+      "grad_norm": 0.3401441522675921,
+      "learning_rate": 9.73420423555967e-06,
+      "loss": 0.2825,
+      "step": 662
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.302990505339083,
+      "learning_rate": 9.710047475727854e-06,
+      "loss": 0.2817,
+      "step": 663
+    },
+    {
+      "epoch": 1.0215384615384615,
+      "grad_norm": 0.3287454011663842,
+      "learning_rate": 9.685892409218718e-06,
+      "loss": 0.2913,
+      "step": 664
+    },
+    {
+      "epoch": 1.023076923076923,
+      "grad_norm": 0.35566618631853003,
+      "learning_rate": 9.661739177097836e-06,
+      "loss": 0.2791,
+      "step": 665
+    },
+    {
+      "epoch": 1.0246153846153847,
+      "grad_norm": 0.3228421290063285,
+      "learning_rate": 9.63758792042008e-06,
+      "loss": 0.2845,
+      "step": 666
+    },
+    {
+      "epoch": 1.0261538461538462,
+      "grad_norm": 0.3437761267951725,
+      "learning_rate": 9.613438780228777e-06,
+      "loss": 0.281,
+      "step": 667
+    },
+    {
+      "epoch": 1.0276923076923077,
+      "grad_norm": 0.3453200532176201,
+      "learning_rate": 9.589291897554912e-06,
+      "loss": 0.2758,
+      "step": 668
+    },
+    {
+      "epoch": 1.0292307692307692,
+      "grad_norm": 0.3074205348443002,
+      "learning_rate": 9.565147413416266e-06,
+      "loss": 0.2927,
+      "step": 669
+    },
+    {
+      "epoch": 1.0307692307692307,
+      "grad_norm": 0.33000167975061806,
+      "learning_rate": 9.541005468816622e-06,
+      "loss": 0.2844,
+      "step": 670
+    },
+    {
+      "epoch": 1.0323076923076924,
+      "grad_norm": 0.3157400495137632,
+      "learning_rate": 9.516866204744932e-06,
+      "loss": 0.2827,
+      "step": 671
+    },
+    {
+      "epoch": 1.0338461538461539,
+      "grad_norm": 0.2994542642176147,
+      "learning_rate": 9.492729762174489e-06,
+      "loss": 0.2906,
+      "step": 672
+    },
+    {
+      "epoch": 1.0353846153846153,
+      "grad_norm": 0.31402805230733577,
+      "learning_rate": 9.468596282062114e-06,
+      "loss": 0.2819,
+      "step": 673
+    },
+    {
+      "epoch": 1.0369230769230768,
+      "grad_norm": 0.3100325537805443,
+      "learning_rate": 9.444465905347327e-06,
+      "loss": 0.2811,
+      "step": 674
+    },
+    {
+      "epoch": 1.0384615384615385,
+      "grad_norm": 0.3059953095706399,
+      "learning_rate": 9.420338772951521e-06,
+      "loss": 0.2828,
+      "step": 675
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.3133760427383818,
+      "learning_rate": 9.39621502577714e-06,
+      "loss": 0.2864,
+      "step": 676
+    },
+    {
+      "epoch": 1.0415384615384615,
+      "grad_norm": 0.3024464286164455,
+      "learning_rate": 9.372094804706867e-06,
+      "loss": 0.2818,
+      "step": 677
+    },
+    {
+      "epoch": 1.043076923076923,
+      "grad_norm": 0.30541483235210354,
+      "learning_rate": 9.347978250602786e-06,
+      "loss": 0.2899,
+      "step": 678
+    },
+    {
+      "epoch": 1.0446153846153847,
+      "grad_norm": 0.3248165537366672,
+      "learning_rate": 9.323865504305566e-06,
+      "loss": 0.2853,
+      "step": 679
+    },
+    {
+      "epoch": 1.0461538461538462,
+      "grad_norm": 0.3043409466949707,
+      "learning_rate": 9.299756706633636e-06,
+      "loss": 0.2836,
+      "step": 680
+    },
+    {
+      "epoch": 1.0476923076923077,
+      "grad_norm": 0.30670835427245857,
+      "learning_rate": 9.275651998382377e-06,
+      "loss": 0.2835,
+      "step": 681
+    },
+    {
+      "epoch": 1.0492307692307692,
+      "grad_norm": 0.3311046500288093,
+      "learning_rate": 9.251551520323273e-06,
+      "loss": 0.2882,
+      "step": 682
+    },
+    {
+      "epoch": 1.0507692307692307,
+      "grad_norm": 0.2988763207191971,
+      "learning_rate": 9.227455413203115e-06,
+      "loss": 0.2743,
+      "step": 683
+    },
+    {
+      "epoch": 1.0523076923076924,
+      "grad_norm": 0.3373397840286235,
+      "learning_rate": 9.203363817743159e-06,
+      "loss": 0.2792,
+      "step": 684
+    },
+    {
+      "epoch": 1.0538461538461539,
+      "grad_norm": 0.3009723950259022,
+      "learning_rate": 9.179276874638315e-06,
+      "loss": 0.2777,
+      "step": 685
+    },
+    {
+      "epoch": 1.0553846153846154,
+      "grad_norm": 0.27068925158047413,
+      "learning_rate": 9.15519472455633e-06,
+      "loss": 0.2951,
+      "step": 686
+    },
+    {
+      "epoch": 1.0569230769230769,
+      "grad_norm": 0.3317453704078991,
+      "learning_rate": 9.131117508136953e-06,
+      "loss": 0.2883,
+      "step": 687
+    },
+    {
+      "epoch": 1.0584615384615386,
+      "grad_norm": 0.32194852853215894,
+      "learning_rate": 9.107045365991123e-06,
+      "loss": 0.2705,
+      "step": 688
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.29148481095050166,
+      "learning_rate": 9.082978438700138e-06,
+      "loss": 0.2811,
+      "step": 689
+    },
+    {
+      "epoch": 1.0615384615384615,
+      "grad_norm": 0.3142038270601785,
+      "learning_rate": 9.058916866814857e-06,
+      "loss": 0.2736,
+      "step": 690
+    },
+    {
+      "epoch": 1.063076923076923,
+      "grad_norm": 0.31913692752332856,
+      "learning_rate": 9.034860790854848e-06,
+      "loss": 0.2825,
+      "step": 691
+    },
+    {
+      "epoch": 1.0646153846153845,
+      "grad_norm": 0.30108394301414343,
+      "learning_rate": 9.010810351307593e-06,
+      "loss": 0.2885,
+      "step": 692
+    },
+    {
+      "epoch": 1.0661538461538462,
+      "grad_norm": 0.2868490120035731,
+      "learning_rate": 8.986765688627652e-06,
+      "loss": 0.2772,
+      "step": 693
+    },
+    {
+      "epoch": 1.0676923076923077,
+      "grad_norm": 0.31221108777213336,
+      "learning_rate": 8.962726943235845e-06,
+      "loss": 0.2821,
+      "step": 694
+    },
+    {
+      "epoch": 1.0692307692307692,
+      "grad_norm": 0.29716262144245537,
+      "learning_rate": 8.938694255518444e-06,
+      "loss": 0.2881,
+      "step": 695
+    },
+    {
+      "epoch": 1.0707692307692307,
+      "grad_norm": 0.30277698505911427,
+      "learning_rate": 8.91466776582634e-06,
+      "loss": 0.2898,
+      "step": 696
+    },
+    {
+      "epoch": 1.0723076923076924,
+      "grad_norm": 0.29732736704215773,
+      "learning_rate": 8.890647614474223e-06,
+      "loss": 0.2836,
+      "step": 697
+    },
+    {
+      "epoch": 1.073846153846154,
+      "grad_norm": 0.3020279299659443,
+      "learning_rate": 8.86663394173977e-06,
+      "loss": 0.2755,
+      "step": 698
+    },
+    {
+      "epoch": 1.0753846153846154,
+      "grad_norm": 0.3037592281364952,
+      "learning_rate": 8.842626887862832e-06,
+      "loss": 0.2862,
+      "step": 699
+    },
+    {
+      "epoch": 1.0769230769230769,
+      "grad_norm": 0.31444223489573714,
+      "learning_rate": 8.818626593044595e-06,
+      "loss": 0.2889,
+      "step": 700
+    },
+    {
+      "epoch": 1.0784615384615384,
+      "grad_norm": 0.28879372525106306,
+      "learning_rate": 8.79463319744677e-06,
+      "loss": 0.2842,
+      "step": 701
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.31140195110522306,
+      "learning_rate": 8.77064684119079e-06,
+      "loss": 0.2832,
+      "step": 702
+    },
+    {
+      "epoch": 1.0815384615384616,
+      "grad_norm": 0.28956637297734567,
+      "learning_rate": 8.746667664356957e-06,
+      "loss": 0.2916,
+      "step": 703
+    },
+    {
+      "epoch": 1.083076923076923,
+      "grad_norm": 0.3143889238176807,
+      "learning_rate": 8.722695806983673e-06,
+      "loss": 0.2792,
+      "step": 704
+    },
+    {
+      "epoch": 1.0846153846153845,
+      "grad_norm": 0.31601279314706854,
+      "learning_rate": 8.69873140906657e-06,
+      "loss": 0.2779,
+      "step": 705
+    },
+    {
+      "epoch": 1.0861538461538462,
+      "grad_norm": 0.30658503317889174,
+      "learning_rate": 8.674774610557728e-06,
+      "loss": 0.2766,
+      "step": 706
+    },
+    {
+      "epoch": 1.0876923076923077,
+      "grad_norm": 0.34147915924162614,
+      "learning_rate": 8.650825551364844e-06,
+      "loss": 0.2849,
+      "step": 707
+    },
+    {
+      "epoch": 1.0892307692307692,
+      "grad_norm": 0.30821824160641453,
+      "learning_rate": 8.626884371350421e-06,
+      "loss": 0.2877,
+      "step": 708
+    },
+    {
+      "epoch": 1.0907692307692307,
+      "grad_norm": 0.31875400232628437,
+      "learning_rate": 8.602951210330942e-06,
+      "loss": 0.2861,
+      "step": 709
+    },
+    {
+      "epoch": 1.0923076923076924,
+      "grad_norm": 0.3192051063313104,
+      "learning_rate": 8.579026208076064e-06,
+      "loss": 0.2821,
+      "step": 710
+    },
+    {
+      "epoch": 1.093846153846154,
+      "grad_norm": 0.33251206019330676,
+      "learning_rate": 8.55510950430779e-06,
+      "loss": 0.2809,
+      "step": 711
+    },
+    {
+      "epoch": 1.0953846153846154,
+      "grad_norm": 0.32505900793848264,
+      "learning_rate": 8.531201238699665e-06,
+      "loss": 0.2752,
+      "step": 712
+    },
+    {
+      "epoch": 1.096923076923077,
+      "grad_norm": 0.3223453889854819,
+      "learning_rate": 8.50730155087596e-06,
+      "loss": 0.2757,
+      "step": 713
+    },
+    {
+      "epoch": 1.0984615384615384,
+      "grad_norm": 0.3336165182216915,
+      "learning_rate": 8.483410580410843e-06,
+      "loss": 0.2922,
+      "step": 714
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.3296284161675168,
+      "learning_rate": 8.459528466827576e-06,
+      "loss": 0.2903,
+      "step": 715
+    },
+    {
+      "epoch": 1.1015384615384616,
+      "grad_norm": 0.3187313302623022,
+      "learning_rate": 8.43565534959769e-06,
+      "loss": 0.2822,
+      "step": 716
+    },
+    {
+      "epoch": 1.103076923076923,
+      "grad_norm": 0.30974612871731544,
+      "learning_rate": 8.411791368140197e-06,
+      "loss": 0.2762,
+      "step": 717
+    },
+    {
+      "epoch": 1.1046153846153846,
+      "grad_norm": 0.3175016265965753,
+      "learning_rate": 8.387936661820733e-06,
+      "loss": 0.2872,
+      "step": 718
+    },
+    {
+      "epoch": 1.106153846153846,
+      "grad_norm": 0.30779535668335356,
+      "learning_rate": 8.364091369950783e-06,
+      "loss": 0.2804,
+      "step": 719
+    },
+    {
+      "epoch": 1.1076923076923078,
+      "grad_norm": 0.2920330333132833,
+      "learning_rate": 8.340255631786843e-06,
+      "loss": 0.28,
+      "step": 720
+    },
+    {
+      "epoch": 1.1092307692307692,
+      "grad_norm": 0.3111369113787312,
+      "learning_rate": 8.316429586529616e-06,
+      "loss": 0.2763,
+      "step": 721
+    },
+    {
+      "epoch": 1.1107692307692307,
+      "grad_norm": 0.32761302966957634,
+      "learning_rate": 8.292613373323203e-06,
+      "loss": 0.2826,
+      "step": 722
+    },
+    {
+      "epoch": 1.1123076923076922,
+      "grad_norm": 0.35928711060393653,
+      "learning_rate": 8.268807131254288e-06,
+      "loss": 0.2765,
+      "step": 723
+    },
+    {
+      "epoch": 1.113846153846154,
+      "grad_norm": 0.3086406308767895,
+      "learning_rate": 8.24501099935131e-06,
+      "loss": 0.2917,
+      "step": 724
+    },
+    {
+      "epoch": 1.1153846153846154,
+      "grad_norm": 0.3109545610843942,
+      "learning_rate": 8.221225116583677e-06,
+      "loss": 0.2847,
+      "step": 725
+    },
+    {
+      "epoch": 1.116923076923077,
+      "grad_norm": 0.30456642359778996,
+      "learning_rate": 8.197449621860944e-06,
+      "loss": 0.2801,
+      "step": 726
+    },
+    {
+      "epoch": 1.1184615384615384,
+      "grad_norm": 0.3003688527787175,
+      "learning_rate": 8.17368465403199e-06,
+      "loss": 0.2837,
+      "step": 727
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.3100907093038021,
+      "learning_rate": 8.14993035188422e-06,
+      "loss": 0.2861,
+      "step": 728
+    },
+    {
+      "epoch": 1.1215384615384616,
+      "grad_norm": 0.3132232125379502,
+      "learning_rate": 8.126186854142752e-06,
+      "loss": 0.284,
+      "step": 729
+    },
+    {
+      "epoch": 1.123076923076923,
+      "grad_norm": 0.31065863337930283,
+      "learning_rate": 8.102454299469615e-06,
+      "loss": 0.2826,
+      "step": 730
+    },
+    {
+      "epoch": 1.1246153846153846,
+      "grad_norm": 0.323662181174654,
+      "learning_rate": 8.078732826462917e-06,
+      "loss": 0.2926,
+      "step": 731
+    },
+    {
+      "epoch": 1.126153846153846,
+      "grad_norm": 0.3134358033047369,
+      "learning_rate": 8.055022573656055e-06,
+      "loss": 0.2731,
+      "step": 732
+    },
+    {
+      "epoch": 1.1276923076923078,
+      "grad_norm": 0.32361693876034364,
+      "learning_rate": 8.0313236795169e-06,
+      "loss": 0.2741,
+      "step": 733
+    },
+    {
+      "epoch": 1.1292307692307693,
+      "grad_norm": 0.29101903327880213,
+      "learning_rate": 8.007636282446986e-06,
+      "loss": 0.2721,
+      "step": 734
+    },
+    {
+      "epoch": 1.1307692307692307,
+      "grad_norm": 0.3018397285643358,
+      "learning_rate": 7.983960520780712e-06,
+      "loss": 0.2868,
+      "step": 735
+    },
+    {
+      "epoch": 1.1323076923076922,
+      "grad_norm": 0.3105308000272363,
+      "learning_rate": 7.960296532784515e-06,
+      "loss": 0.2842,
+      "step": 736
+    },
+    {
+      "epoch": 1.1338461538461537,
+      "grad_norm": 0.30612415691801403,
+      "learning_rate": 7.936644456656082e-06,
+      "loss": 0.2914,
+      "step": 737
+    },
+    {
+      "epoch": 1.1353846153846154,
+      "grad_norm": 0.313629256201062,
+      "learning_rate": 7.913004430523526e-06,
+      "loss": 0.2771,
+      "step": 738
+    },
+    {
+      "epoch": 1.136923076923077,
+      "grad_norm": 0.3240648431845592,
+      "learning_rate": 7.889376592444605e-06,
+      "loss": 0.2855,
+      "step": 739
+    },
+    {
+      "epoch": 1.1384615384615384,
+      "grad_norm": 0.3044134814557973,
+      "learning_rate": 7.865761080405882e-06,
+      "loss": 0.291,
+      "step": 740
+    },
+    {
+      "epoch": 1.1400000000000001,
+      "grad_norm": 0.27811641502889545,
+      "learning_rate": 7.84215803232194e-06,
+      "loss": 0.2794,
+      "step": 741
+    },
+    {
+      "epoch": 1.1415384615384616,
+      "grad_norm": 0.32218417728467347,
+      "learning_rate": 7.818567586034578e-06,
+      "loss": 0.2844,
+      "step": 742
+    },
+    {
+      "epoch": 1.143076923076923,
+      "grad_norm": 0.3067532071100499,
+      "learning_rate": 7.794989879311991e-06,
+      "loss": 0.285,
+      "step": 743
+    },
+    {
+      "epoch": 1.1446153846153846,
+      "grad_norm": 0.2991862411303418,
+      "learning_rate": 7.771425049847984e-06,
+      "loss": 0.2841,
+      "step": 744
+    },
+    {
+      "epoch": 1.146153846153846,
+      "grad_norm": 0.32247577406668865,
+      "learning_rate": 7.747873235261157e-06,
+      "loss": 0.2909,
+      "step": 745
+    },
+    {
+      "epoch": 1.1476923076923078,
+      "grad_norm": 0.31954229399729445,
+      "learning_rate": 7.724334573094101e-06,
+      "loss": 0.2878,
+      "step": 746
+    },
+    {
+      "epoch": 1.1492307692307693,
+      "grad_norm": 0.30083806921436773,
+      "learning_rate": 7.700809200812596e-06,
+      "loss": 0.2891,
+      "step": 747
+    },
+    {
+      "epoch": 1.1507692307692308,
+      "grad_norm": 0.3142510325234358,
+      "learning_rate": 7.677297255804811e-06,
+      "loss": 0.2856,
+      "step": 748
+    },
+    {
+      "epoch": 1.1523076923076923,
+      "grad_norm": 0.2870832473008288,
+      "learning_rate": 7.6537988753805e-06,
+      "loss": 0.2885,
+      "step": 749
+    },
+    {
+      "epoch": 1.1538461538461537,
+      "grad_norm": 0.3007062802890109,
+      "learning_rate": 7.6303141967702e-06,
+      "loss": 0.2934,
+      "step": 750
+    },
+    {
+      "epoch": 1.1553846153846155,
+      "grad_norm": 0.31805936290046793,
+      "learning_rate": 7.606843357124426e-06,
+      "loss": 0.2859,
+      "step": 751
+    },
+    {
+      "epoch": 1.156923076923077,
+      "grad_norm": 0.2888594426489485,
+      "learning_rate": 7.583386493512872e-06,
+      "loss": 0.2823,
+      "step": 752
+    },
+    {
+      "epoch": 1.1584615384615384,
+      "grad_norm": 0.2918277431569358,
+      "learning_rate": 7.559943742923626e-06,
+      "loss": 0.2806,
+      "step": 753
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.27255929440743293,
+      "learning_rate": 7.536515242262341e-06,
+      "loss": 0.2739,
+      "step": 754
+    },
+    {
+      "epoch": 1.1615384615384616,
+      "grad_norm": 0.33173787979163183,
+      "learning_rate": 7.513101128351454e-06,
+      "loss": 0.2834,
+      "step": 755
+    },
+    {
+      "epoch": 1.1630769230769231,
+      "grad_norm": 0.3238320694495885,
+      "learning_rate": 7.489701537929384e-06,
+      "loss": 0.289,
+      "step": 756
+    },
+    {
+      "epoch": 1.1646153846153846,
+      "grad_norm": 0.2809829073539621,
+      "learning_rate": 7.4663166076497376e-06,
+      "loss": 0.2838,
+      "step": 757
+    },
+    {
+      "epoch": 1.166153846153846,
+      "grad_norm": 0.2947202247452326,
+      "learning_rate": 7.442946474080499e-06,
+      "loss": 0.2787,
+      "step": 758
+    },
+    {
+      "epoch": 1.1676923076923078,
+      "grad_norm": 0.319228719886482,
+      "learning_rate": 7.419591273703245e-06,
+      "loss": 0.2828,
+      "step": 759
+    },
+    {
+      "epoch": 1.1692307692307693,
+      "grad_norm": 0.2847995636733814,
+      "learning_rate": 7.396251142912337e-06,
+      "loss": 0.2857,
+      "step": 760
+    },
+    {
+      "epoch": 1.1707692307692308,
+      "grad_norm": 0.29399040893561257,
+      "learning_rate": 7.372926218014131e-06,
+      "loss": 0.2776,
+      "step": 761
+    },
+    {
+      "epoch": 1.1723076923076923,
+      "grad_norm": 0.2955780863612147,
+      "learning_rate": 7.349616635226191e-06,
+      "loss": 0.2782,
+      "step": 762
+    },
+    {
+      "epoch": 1.1738461538461538,
+      "grad_norm": 0.28775904936662283,
+      "learning_rate": 7.326322530676471e-06,
+      "loss": 0.2873,
+      "step": 763
+    },
+    {
+      "epoch": 1.1753846153846155,
+      "grad_norm": 0.30879780507027793,
+      "learning_rate": 7.303044040402536e-06,
+      "loss": 0.2874,
+      "step": 764
+    },
+    {
+      "epoch": 1.176923076923077,
+      "grad_norm": 0.27411031230591126,
+      "learning_rate": 7.279781300350758e-06,
+      "loss": 0.2746,
+      "step": 765
+    },
+    {
+      "epoch": 1.1784615384615384,
+      "grad_norm": 0.3025701791637634,
+      "learning_rate": 7.256534446375543e-06,
+      "loss": 0.2798,
+      "step": 766
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.3047469460566567,
+      "learning_rate": 7.23330361423851e-06,
+      "loss": 0.2892,
+      "step": 767
+    },
+    {
+      "epoch": 1.1815384615384614,
+      "grad_norm": 0.2957662470836879,
+      "learning_rate": 7.210088939607709e-06,
+      "loss": 0.2839,
+      "step": 768
+    },
+    {
+      "epoch": 1.1830769230769231,
+      "grad_norm": 0.2849127131802998,
+      "learning_rate": 7.186890558056836e-06,
+      "loss": 0.2797,
+      "step": 769
+    },
+    {
+      "epoch": 1.1846153846153846,
+      "grad_norm": 0.29518177363624887,
+      "learning_rate": 7.163708605064437e-06,
+      "loss": 0.289,
+      "step": 770
+    },
+    {
+      "epoch": 1.1861538461538461,
+      "grad_norm": 0.28630707314852843,
+      "learning_rate": 7.140543216013109e-06,
+      "loss": 0.278,
+      "step": 771
+    },
+    {
+      "epoch": 1.1876923076923076,
+      "grad_norm": 0.2947200274521145,
+      "learning_rate": 7.117394526188719e-06,
+      "loss": 0.2762,
+      "step": 772
+    },
+    {
+      "epoch": 1.1892307692307693,
+      "grad_norm": 0.2871611641306348,
+      "learning_rate": 7.094262670779611e-06,
+      "loss": 0.2864,
+      "step": 773
+    },
+    {
+      "epoch": 1.1907692307692308,
+      "grad_norm": 0.2905509938001612,
+      "learning_rate": 7.071147784875809e-06,
+      "loss": 0.2811,
+      "step": 774
+    },
+    {
+      "epoch": 1.1923076923076923,
+      "grad_norm": 0.28852886723854776,
+      "learning_rate": 7.048050003468252e-06,
+      "loss": 0.2788,
+      "step": 775
+    },
+    {
+      "epoch": 1.1938461538461538,
+      "grad_norm": 0.2862816806354988,
+      "learning_rate": 7.024969461447973e-06,
+      "loss": 0.2921,
+      "step": 776
+    },
+    {
+      "epoch": 1.1953846153846155,
+      "grad_norm": 0.28035417832956877,
+      "learning_rate": 7.001906293605329e-06,
+      "loss": 0.27,
+      "step": 777
+    },
+    {
+      "epoch": 1.196923076923077,
+      "grad_norm": 0.3081549365320152,
+      "learning_rate": 6.978860634629213e-06,
+      "loss": 0.2628,
+      "step": 778
+    },
+    {
+      "epoch": 1.1984615384615385,
+      "grad_norm": 0.29646626529473624,
+      "learning_rate": 6.9558326191062775e-06,
+      "loss": 0.2739,
+      "step": 779
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.30203294671651026,
+      "learning_rate": 6.932822381520121e-06,
+      "loss": 0.268,
+      "step": 780
+    },
+    {
+      "epoch": 1.2015384615384614,
+      "grad_norm": 0.3093392278102877,
+      "learning_rate": 6.909830056250527e-06,
+      "loss": 0.2838,
+      "step": 781
+    },
+    {
+      "epoch": 1.2030769230769232,
+      "grad_norm": 0.29792246798832145,
+      "learning_rate": 6.88685577757267e-06,
+      "loss": 0.2803,
+      "step": 782
+    },
+    {
+      "epoch": 1.2046153846153846,
+      "grad_norm": 0.2977596725066236,
+      "learning_rate": 6.8638996796563275e-06,
+      "loss": 0.2875,
+      "step": 783
+    },
+    {
+      "epoch": 1.2061538461538461,
+      "grad_norm": 0.3043730929932303,
+      "learning_rate": 6.8409618965651125e-06,
+      "loss": 0.275,
+      "step": 784
+    },
+    {
+      "epoch": 1.2076923076923076,
+      "grad_norm": 0.2813402340518155,
+      "learning_rate": 6.81804256225567e-06,
+      "loss": 0.2862,
+      "step": 785
+    },
+    {
+      "epoch": 1.209230769230769,
+      "grad_norm": 0.32350615987131526,
+      "learning_rate": 6.795141810576906e-06,
+      "loss": 0.2846,
+      "step": 786
+    },
+    {
+      "epoch": 1.2107692307692308,
+      "grad_norm": 0.2861412130641737,
+      "learning_rate": 6.772259775269203e-06,
+      "loss": 0.2777,
+      "step": 787
+    },
+    {
+      "epoch": 1.2123076923076923,
+      "grad_norm": 0.2948577774888858,
+      "learning_rate": 6.7493965899636486e-06,
+      "loss": 0.2815,
+      "step": 788
+    },
+    {
+      "epoch": 1.2138461538461538,
+      "grad_norm": 0.28655075068199576,
+      "learning_rate": 6.7265523881812335e-06,
+      "loss": 0.2809,
+      "step": 789
+    },
+    {
+      "epoch": 1.2153846153846155,
+      "grad_norm": 0.2954405673613976,
+      "learning_rate": 6.703727303332094e-06,
+      "loss": 0.2793,
+      "step": 790
+    },
+    {
+      "epoch": 1.216923076923077,
+      "grad_norm": 0.285777124696368,
+      "learning_rate": 6.680921468714718e-06,
+      "loss": 0.277,
+      "step": 791
+    },
+    {
+      "epoch": 1.2184615384615385,
+      "grad_norm": 0.28801215245271594,
+      "learning_rate": 6.6581350175151715e-06,
+      "loss": 0.2817,
+      "step": 792
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.29358260711212547,
+      "learning_rate": 6.6353680828063306e-06,
+      "loss": 0.2761,
+      "step": 793
+    },
+    {
+      "epoch": 1.2215384615384615,
+      "grad_norm": 0.2732920910463464,
+      "learning_rate": 6.612620797547087e-06,
+      "loss": 0.2688,
+      "step": 794
+    },
+    {
+      "epoch": 1.2230769230769232,
+      "grad_norm": 0.2823232953579946,
+      "learning_rate": 6.589893294581579e-06,
+      "loss": 0.283,
+      "step": 795
+    },
+    {
+      "epoch": 1.2246153846153847,
+      "grad_norm": 0.27802707917977293,
+      "learning_rate": 6.567185706638417e-06,
+      "loss": 0.2812,
+      "step": 796
+    },
+    {
+      "epoch": 1.2261538461538461,
+      "grad_norm": 0.29195760944835975,
+      "learning_rate": 6.5444981663299135e-06,
+      "loss": 0.272,
+      "step": 797
+    },
+    {
+      "epoch": 1.2276923076923076,
+      "grad_norm": 0.271685328313952,
+      "learning_rate": 6.521830806151297e-06,
+      "loss": 0.2902,
+      "step": 798
+    },
+    {
+      "epoch": 1.2292307692307691,
+      "grad_norm": 0.28460765234355695,
+      "learning_rate": 6.499183758479944e-06,
+      "loss": 0.2882,
+      "step": 799
+    },
+    {
+      "epoch": 1.2307692307692308,
+      "grad_norm": 0.2863072902666916,
+      "learning_rate": 6.47655715557461e-06,
+      "loss": 0.2812,
+      "step": 800
+    },
+    {
+      "epoch": 1.2323076923076923,
+      "grad_norm": 0.28377583009326257,
+      "learning_rate": 6.453951129574644e-06,
+      "loss": 0.2795,
+      "step": 801
+    },
+    {
+      "epoch": 1.2338461538461538,
+      "grad_norm": 0.29333544616487517,
+      "learning_rate": 6.431365812499242e-06,
+      "loss": 0.2936,
+      "step": 802
+    },
+    {
+      "epoch": 1.2353846153846153,
+      "grad_norm": 0.29953315694101823,
+      "learning_rate": 6.408801336246645e-06,
+      "loss": 0.2867,
+      "step": 803
+    },
+    {
+      "epoch": 1.236923076923077,
+      "grad_norm": 0.27386702982109357,
+      "learning_rate": 6.38625783259339e-06,
+      "loss": 0.2837,
+      "step": 804
+    },
+    {
+      "epoch": 1.2384615384615385,
+      "grad_norm": 0.310898071421186,
+      "learning_rate": 6.36373543319353e-06,
+      "loss": 0.2826,
+      "step": 805
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.2944810008817099,
+      "learning_rate": 6.341234269577878e-06,
+      "loss": 0.2685,
+      "step": 806
+    },
+    {
+      "epoch": 1.2415384615384615,
+      "grad_norm": 0.2791213720495511,
+      "learning_rate": 6.318754473153221e-06,
+      "loss": 0.2824,
+      "step": 807
+    },
+    {
+      "epoch": 1.2430769230769232,
+      "grad_norm": 0.29365028663555504,
+      "learning_rate": 6.296296175201565e-06,
+      "loss": 0.2799,
+      "step": 808
+    },
+    {
+      "epoch": 1.2446153846153847,
+      "grad_norm": 0.3150248889907267,
+      "learning_rate": 6.273859506879365e-06,
+      "loss": 0.2744,
+      "step": 809
+    },
+    {
+      "epoch": 1.2461538461538462,
+      "grad_norm": 0.32842309193065544,
+      "learning_rate": 6.251444599216756e-06,
+      "loss": 0.2769,
+      "step": 810
+    },
+    {
+      "epoch": 1.2476923076923077,
+      "grad_norm": 0.29180890811914484,
+      "learning_rate": 6.229051583116796e-06,
+      "loss": 0.2863,
+      "step": 811
+    },
+    {
+      "epoch": 1.2492307692307691,
+      "grad_norm": 0.2887378998610124,
+      "learning_rate": 6.206680589354696e-06,
+      "loss": 0.2748,
+      "step": 812
+    },
+    {
+      "epoch": 1.2507692307692309,
+      "grad_norm": 0.2777487183276391,
+      "learning_rate": 6.18433174857705e-06,
+      "loss": 0.288,
+      "step": 813
+    },
+    {
+      "epoch": 1.2523076923076923,
+      "grad_norm": 0.30945368617265023,
+      "learning_rate": 6.162005191301082e-06,
+      "loss": 0.2788,
+      "step": 814
+    },
+    {
+      "epoch": 1.2538461538461538,
+      "grad_norm": 0.27052743875684937,
+      "learning_rate": 6.139701047913885e-06,
+      "loss": 0.2653,
+      "step": 815
+    },
+    {
+      "epoch": 1.2553846153846153,
+      "grad_norm": 0.30934036148283867,
+      "learning_rate": 6.117419448671651e-06,
+      "loss": 0.2892,
+      "step": 816
+    },
+    {
+      "epoch": 1.2569230769230768,
+      "grad_norm": 0.31001399577452204,
+      "learning_rate": 6.095160523698913e-06,
+      "loss": 0.2795,
+      "step": 817
+    },
+    {
+      "epoch": 1.2584615384615385,
+      "grad_norm": 0.306823491455034,
+      "learning_rate": 6.072924402987785e-06,
+      "loss": 0.2761,
+      "step": 818
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.2821450940838836,
+      "learning_rate": 6.050711216397212e-06,
+      "loss": 0.277,
+      "step": 819
+    },
+    {
+      "epoch": 1.2615384615384615,
+      "grad_norm": 0.29778869798743096,
+      "learning_rate": 6.028521093652195e-06,
+      "loss": 0.2938,
+      "step": 820
+    },
+    {
+      "epoch": 1.2630769230769232,
+      "grad_norm": 0.3355217690858126,
+      "learning_rate": 6.006354164343047e-06,
+      "loss": 0.2728,
+      "step": 821
+    },
+    {
+      "epoch": 1.2646153846153847,
+      "grad_norm": 0.289782832193456,
+      "learning_rate": 5.984210557924628e-06,
+      "loss": 0.2814,
+      "step": 822
+    },
+    {
+      "epoch": 1.2661538461538462,
+      "grad_norm": 0.2878090579578816,
+      "learning_rate": 5.962090403715592e-06,
+      "loss": 0.2866,
+      "step": 823
+    },
+    {
+      "epoch": 1.2676923076923077,
+      "grad_norm": 0.3019556593740292,
+      "learning_rate": 5.939993830897641e-06,
+      "loss": 0.2807,
+      "step": 824
+    },
+    {
+      "epoch": 1.2692307692307692,
+      "grad_norm": 0.3158326263040181,
+      "learning_rate": 5.9179209685147525e-06,
+      "loss": 0.2794,
+      "step": 825
+    },
+    {
+      "epoch": 1.2707692307692309,
+      "grad_norm": 0.28707432706520675,
+      "learning_rate": 5.895871945472434e-06,
+      "loss": 0.2845,
+      "step": 826
+    },
+    {
+      "epoch": 1.2723076923076924,
+      "grad_norm": 0.2963834535647672,
+      "learning_rate": 5.873846890536976e-06,
+      "loss": 0.2816,
+      "step": 827
+    },
+    {
+      "epoch": 1.2738461538461539,
+      "grad_norm": 0.31596386394483994,
+      "learning_rate": 5.851845932334698e-06,
+      "loss": 0.2855,
+      "step": 828
+    },
+    {
+      "epoch": 1.2753846153846153,
+      "grad_norm": 0.2828167486772763,
+      "learning_rate": 5.829869199351188e-06,
+      "loss": 0.2869,
+      "step": 829
+    },
+    {
+      "epoch": 1.2769230769230768,
+      "grad_norm": 0.3016058478748179,
+      "learning_rate": 5.807916819930563e-06,
+      "loss": 0.2812,
+      "step": 830
+    },
+    {
+      "epoch": 1.2784615384615385,
+      "grad_norm": 0.2955037144013558,
+      "learning_rate": 5.785988922274711e-06,
+      "loss": 0.2761,
+      "step": 831
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.28474944236281235,
+      "learning_rate": 5.7640856344425465e-06,
+      "loss": 0.2754,
+      "step": 832
+    },
+    {
+      "epoch": 1.2815384615384615,
+      "grad_norm": 0.2949274618459585,
+      "learning_rate": 5.742207084349274e-06,
+      "loss": 0.2829,
+      "step": 833
+    },
+    {
+      "epoch": 1.283076923076923,
+      "grad_norm": 0.3097048774109836,
+      "learning_rate": 5.720353399765615e-06,
+      "loss": 0.285,
+      "step": 834
+    },
+    {
+      "epoch": 1.2846153846153845,
+      "grad_norm": 0.3097721464332409,
+      "learning_rate": 5.698524708317082e-06,
+      "loss": 0.2894,
+      "step": 835
+    },
+    {
+      "epoch": 1.2861538461538462,
+      "grad_norm": 0.2766367492199567,
+      "learning_rate": 5.676721137483226e-06,
+      "loss": 0.2801,
+      "step": 836
+    },
+    {
+      "epoch": 1.2876923076923077,
+      "grad_norm": 0.29568337118433724,
+      "learning_rate": 5.654942814596902e-06,
+      "loss": 0.2955,
+      "step": 837
+    },
+    {
+      "epoch": 1.2892307692307692,
+      "grad_norm": 0.31699922367774136,
+      "learning_rate": 5.633189866843507e-06,
+      "loss": 0.288,
+      "step": 838
+    },
+    {
+      "epoch": 1.290769230769231,
+      "grad_norm": 0.28639959989906244,
+      "learning_rate": 5.611462421260251e-06,
+      "loss": 0.2843,
+      "step": 839
+    },
+    {
+      "epoch": 1.2923076923076924,
+      "grad_norm": 0.2678488804524567,
+      "learning_rate": 5.58976060473541e-06,
+      "loss": 0.2873,
+      "step": 840
+    },
+    {
+      "epoch": 1.2938461538461539,
+      "grad_norm": 0.2776902486642549,
+      "learning_rate": 5.5680845440075885e-06,
+      "loss": 0.2896,
+      "step": 841
+    },
+    {
+      "epoch": 1.2953846153846154,
+      "grad_norm": 0.3138059651765253,
+      "learning_rate": 5.546434365664974e-06,
+      "loss": 0.2701,
+      "step": 842
+    },
+    {
+      "epoch": 1.2969230769230768,
+      "grad_norm": 0.3146892839180133,
+      "learning_rate": 5.5248101961446065e-06,
+      "loss": 0.2817,
+      "step": 843
+    },
+    {
+      "epoch": 1.2984615384615386,
+      "grad_norm": 0.2939281513518946,
+      "learning_rate": 5.503212161731628e-06,
+      "loss": 0.2836,
+      "step": 844
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.3025062620250353,
+      "learning_rate": 5.481640388558551e-06,
+      "loss": 0.2902,
+      "step": 845
+    },
+    {
+      "epoch": 1.3015384615384615,
+      "grad_norm": 0.3162590193256863,
+      "learning_rate": 5.460095002604533e-06,
+      "loss": 0.2736,
+      "step": 846
+    },
+    {
+      "epoch": 1.303076923076923,
+      "grad_norm": 0.2844927223440658,
+      "learning_rate": 5.43857612969462e-06,
+      "loss": 0.2849,
+      "step": 847
+    },
+    {
+      "epoch": 1.3046153846153845,
+      "grad_norm": 0.29370463036029804,
+      "learning_rate": 5.417083895499024e-06,
+      "loss": 0.2842,
+      "step": 848
+    },
+    {
+      "epoch": 1.3061538461538462,
+      "grad_norm": 0.30408719752218816,
+      "learning_rate": 5.39561842553239e-06,
+      "loss": 0.2752,
+      "step": 849
+    },
+    {
+      "epoch": 1.3076923076923077,
+      "grad_norm": 0.29814878432273506,
+      "learning_rate": 5.374179845153048e-06,
+      "loss": 0.2725,
+      "step": 850
+    },
+    {
+      "epoch": 1.3092307692307692,
+      "grad_norm": 0.3090317868858365,
+      "learning_rate": 5.352768279562315e-06,
+      "loss": 0.2795,
+      "step": 851
+    },
+    {
+      "epoch": 1.3107692307692307,
+      "grad_norm": 0.3149051067033696,
+      "learning_rate": 5.331383853803724e-06,
+      "loss": 0.279,
+      "step": 852
+    },
+    {
+      "epoch": 1.3123076923076922,
+      "grad_norm": 0.29026031338987546,
+      "learning_rate": 5.310026692762316e-06,
+      "loss": 0.269,
+      "step": 853
+    },
+    {
+      "epoch": 1.3138461538461539,
+      "grad_norm": 0.3061058627098662,
+      "learning_rate": 5.288696921163902e-06,
+      "loss": 0.2755,
+      "step": 854
+    },
+    {
+      "epoch": 1.3153846153846154,
+      "grad_norm": 0.3067580088583177,
+      "learning_rate": 5.267394663574351e-06,
+      "loss": 0.2616,
+      "step": 855
+    },
+    {
+      "epoch": 1.3169230769230769,
+      "grad_norm": 0.28661946051764187,
+      "learning_rate": 5.246120044398839e-06,
+      "loss": 0.2949,
+      "step": 856
+    },
+    {
+      "epoch": 1.3184615384615386,
+      "grad_norm": 0.2862892658665391,
+      "learning_rate": 5.224873187881136e-06,
+      "loss": 0.275,
+      "step": 857
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.28534062120425885,
+      "learning_rate": 5.20365421810288e-06,
+      "loss": 0.2879,
+      "step": 858
+    },
+    {
+      "epoch": 1.3215384615384616,
+      "grad_norm": 0.27279911218675196,
+      "learning_rate": 5.1824632589828465e-06,
+      "loss": 0.2685,
+      "step": 859
+    },
+    {
+      "epoch": 1.323076923076923,
+      "grad_norm": 0.30170979115869934,
+      "learning_rate": 5.161300434276237e-06,
+      "loss": 0.2728,
+      "step": 860
+    },
+    {
+      "epoch": 1.3246153846153845,
+      "grad_norm": 0.31083997998278023,
+      "learning_rate": 5.14016586757394e-06,
+      "loss": 0.2765,
+      "step": 861
+    },
+    {
+      "epoch": 1.3261538461538462,
+      "grad_norm": 0.27275151442143564,
+      "learning_rate": 5.119059682301819e-06,
+      "loss": 0.2796,
+      "step": 862
+    },
+    {
+      "epoch": 1.3276923076923077,
+      "grad_norm": 0.28369026644045975,
+      "learning_rate": 5.097982001719994e-06,
+      "loss": 0.28,
+      "step": 863
+    },
+    {
+      "epoch": 1.3292307692307692,
+      "grad_norm": 0.304381409197318,
+      "learning_rate": 5.076932948922111e-06,
+      "loss": 0.2816,
+      "step": 864
+    },
+    {
+      "epoch": 1.3307692307692307,
+      "grad_norm": 0.2844124937713117,
+      "learning_rate": 5.0559126468346354e-06,
+      "loss": 0.2797,
+      "step": 865
+    },
+    {
+      "epoch": 1.3323076923076922,
+      "grad_norm": 0.2715699243242887,
+      "learning_rate": 5.034921218216126e-06,
+      "loss": 0.2767,
+      "step": 866
+    },
+    {
+      "epoch": 1.333846153846154,
+      "grad_norm": 0.2891583855016688,
+      "learning_rate": 5.013958785656516e-06,
+      "loss": 0.2722,
+      "step": 867
+    },
+    {
+      "epoch": 1.3353846153846154,
+      "grad_norm": 0.3086744282279073,
+      "learning_rate": 4.993025471576417e-06,
+      "loss": 0.2727,
+      "step": 868
+    },
+    {
+      "epoch": 1.3369230769230769,
+      "grad_norm": 0.2809894462210739,
+      "learning_rate": 4.972121398226371e-06,
+      "loss": 0.2843,
+      "step": 869
+    },
+    {
+      "epoch": 1.3384615384615386,
+      "grad_norm": 0.27436121104774197,
+      "learning_rate": 4.951246687686164e-06,
+      "loss": 0.2833,
+      "step": 870
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.283858740765341,
+      "learning_rate": 4.930401461864099e-06,
+      "loss": 0.2826,
+      "step": 871
+    },
+    {
+      "epoch": 1.3415384615384616,
+      "grad_norm": 0.2818450715432375,
+      "learning_rate": 4.909585842496287e-06,
+      "loss": 0.2887,
+      "step": 872
+    },
+    {
+      "epoch": 1.343076923076923,
+      "grad_norm": 0.28854298555763697,
+      "learning_rate": 4.888799951145948e-06,
+      "loss": 0.2768,
+      "step": 873
+    },
+    {
+      "epoch": 1.3446153846153845,
+      "grad_norm": 0.30065363915653354,
+      "learning_rate": 4.868043909202678e-06,
+      "loss": 0.278,
+      "step": 874
+    },
+    {
+      "epoch": 1.3461538461538463,
+      "grad_norm": 0.27408495965643176,
+      "learning_rate": 4.847317837881757e-06,
+      "loss": 0.28,
+      "step": 875
+    },
+    {
+      "epoch": 1.3476923076923077,
+      "grad_norm": 0.2873579092416445,
+      "learning_rate": 4.826621858223431e-06,
+      "loss": 0.2777,
+      "step": 876
+    },
+    {
+      "epoch": 1.3492307692307692,
+      "grad_norm": 0.2913491598278634,
+      "learning_rate": 4.805956091092228e-06,
+      "loss": 0.2829,
+      "step": 877
+    },
+    {
+      "epoch": 1.3507692307692307,
+      "grad_norm": 0.3000633859084866,
+      "learning_rate": 4.785320657176216e-06,
+      "loss": 0.2738,
+      "step": 878
+    },
+    {
+      "epoch": 1.3523076923076922,
+      "grad_norm": 0.2992709856059088,
+      "learning_rate": 4.764715676986327e-06,
+      "loss": 0.2797,
+      "step": 879
+    },
+    {
+      "epoch": 1.353846153846154,
+      "grad_norm": 0.2764988951323593,
+      "learning_rate": 4.744141270855638e-06,
+      "loss": 0.2696,
+      "step": 880
+    },
+    {
+      "epoch": 1.3553846153846154,
+      "grad_norm": 0.27731299592358694,
+      "learning_rate": 4.7235975589386715e-06,
+      "loss": 0.2838,
+      "step": 881
+    },
+    {
+      "epoch": 1.356923076923077,
+      "grad_norm": 0.27762040579303454,
+      "learning_rate": 4.7030846612107105e-06,
+      "loss": 0.2678,
+      "step": 882
+    },
+    {
+      "epoch": 1.3584615384615384,
+      "grad_norm": 0.3032540883629789,
+      "learning_rate": 4.6826026974670665e-06,
+      "loss": 0.2873,
+      "step": 883
+    },
+    {
+      "epoch": 1.3599999999999999,
+      "grad_norm": 0.28474295761604757,
+      "learning_rate": 4.662151787322405e-06,
+      "loss": 0.2931,
+      "step": 884
+    },
+    {
+      "epoch": 1.3615384615384616,
+      "grad_norm": 0.2767424780339405,
+      "learning_rate": 4.641732050210032e-06,
+      "loss": 0.2754,
+      "step": 885
+    },
+    {
+      "epoch": 1.363076923076923,
+      "grad_norm": 0.2661749973204445,
+      "learning_rate": 4.621343605381215e-06,
+      "loss": 0.2736,
+      "step": 886
+    },
+    {
+      "epoch": 1.3646153846153846,
+      "grad_norm": 0.2855111722858186,
+      "learning_rate": 4.600986571904461e-06,
+      "loss": 0.2724,
+      "step": 887
+    },
+    {
+      "epoch": 1.3661538461538463,
+      "grad_norm": 0.2670827052296132,
+      "learning_rate": 4.580661068664844e-06,
+      "loss": 0.2808,
+      "step": 888
+    },
+    {
+      "epoch": 1.3676923076923078,
+      "grad_norm": 0.26642196866945184,
+      "learning_rate": 4.560367214363295e-06,
+      "loss": 0.2832,
+      "step": 889
+    },
+    {
+      "epoch": 1.3692307692307693,
+      "grad_norm": 0.26447372975548733,
+      "learning_rate": 4.540105127515921e-06,
+      "loss": 0.2886,
+      "step": 890
+    },
+    {
+      "epoch": 1.3707692307692307,
+      "grad_norm": 0.27104430863461343,
+      "learning_rate": 4.519874926453303e-06,
+      "loss": 0.2732,
+      "step": 891
+    },
+    {
+      "epoch": 1.3723076923076922,
+      "grad_norm": 0.2639620175059038,
+      "learning_rate": 4.499676729319809e-06,
+      "loss": 0.2772,
+      "step": 892
+    },
+    {
+      "epoch": 1.373846153846154,
+      "grad_norm": 0.27441667168958817,
+      "learning_rate": 4.479510654072909e-06,
+      "loss": 0.2812,
+      "step": 893
+    },
+    {
+      "epoch": 1.3753846153846154,
+      "grad_norm": 0.29082870304085784,
+      "learning_rate": 4.459376818482471e-06,
+      "loss": 0.2682,
+      "step": 894
+    },
+    {
+      "epoch": 1.376923076923077,
+      "grad_norm": 0.2877630922254318,
+      "learning_rate": 4.439275340130099e-06,
+      "loss": 0.2681,
+      "step": 895
+    },
+    {
+      "epoch": 1.3784615384615384,
+      "grad_norm": 0.26989551556496555,
+      "learning_rate": 4.419206336408418e-06,
+      "loss": 0.279,
+      "step": 896
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.2869441676477556,
+      "learning_rate": 4.399169924520403e-06,
+      "loss": 0.2842,
+      "step": 897
+    },
+    {
+      "epoch": 1.3815384615384616,
+      "grad_norm": 0.30094163030214116,
+      "learning_rate": 4.379166221478697e-06,
+      "loss": 0.2747,
+      "step": 898
+    },
+    {
+      "epoch": 1.383076923076923,
+      "grad_norm": 0.2771588125935493,
+      "learning_rate": 4.359195344104916e-06,
+      "loss": 0.2685,
+      "step": 899
+    },
+    {
+      "epoch": 1.3846153846153846,
+      "grad_norm": 0.29507488075814775,
+      "learning_rate": 4.339257409028987e-06,
+      "loss": 0.2806,
+      "step": 900
+    },
+    {
+      "epoch": 1.3861538461538463,
+      "grad_norm": 0.26093386279230574,
+      "learning_rate": 4.319352532688444e-06,
+      "loss": 0.2888,
+      "step": 901
+    },
+    {
+      "epoch": 1.3876923076923076,
+      "grad_norm": 0.29971147233177703,
+      "learning_rate": 4.2994808313277565e-06,
+      "loss": 0.2806,
+      "step": 902
+    },
+    {
+      "epoch": 1.3892307692307693,
+      "grad_norm": 0.2925280942409716,
+      "learning_rate": 4.279642420997655e-06,
+      "loss": 0.2766,
+      "step": 903
+    },
+    {
+      "epoch": 1.3907692307692308,
+      "grad_norm": 0.310206982896478,
+      "learning_rate": 4.259837417554457e-06,
+      "loss": 0.274,
+      "step": 904
+    },
+    {
+      "epoch": 1.3923076923076922,
+      "grad_norm": 0.27515905727135725,
+      "learning_rate": 4.240065936659374e-06,
+      "loss": 0.2872,
+      "step": 905
+    },
+    {
+      "epoch": 1.393846153846154,
+      "grad_norm": 0.30847037362181456,
+      "learning_rate": 4.220328093777851e-06,
+      "loss": 0.2732,
+      "step": 906
+    },
+    {
+      "epoch": 1.3953846153846154,
+      "grad_norm": 0.29152540720862175,
+      "learning_rate": 4.200624004178883e-06,
+      "loss": 0.2815,
+      "step": 907
+    },
+    {
+      "epoch": 1.396923076923077,
+      "grad_norm": 0.2796021392011909,
+      "learning_rate": 4.180953782934352e-06,
+      "loss": 0.28,
+      "step": 908
+    },
+    {
+      "epoch": 1.3984615384615384,
+      "grad_norm": 0.2807257095406004,
+      "learning_rate": 4.161317544918345e-06,
+      "loss": 0.2733,
+      "step": 909
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.2772487280421376,
+      "learning_rate": 4.141715404806486e-06,
+      "loss": 0.2699,
+      "step": 910
+    },
+    {
+      "epoch": 1.4015384615384616,
+      "grad_norm": 0.2884138212191608,
+      "learning_rate": 4.12214747707527e-06,
+      "loss": 0.279,
+      "step": 911
+    },
+    {
+      "epoch": 1.403076923076923,
+      "grad_norm": 0.2680045744149855,
+      "learning_rate": 4.1026138760013886e-06,
+      "loss": 0.2798,
+      "step": 912
+    },
+    {
+      "epoch": 1.4046153846153846,
+      "grad_norm": 0.2692589868706836,
+      "learning_rate": 4.083114715661069e-06,
+      "loss": 0.2829,
+      "step": 913
+    },
+    {
+      "epoch": 1.406153846153846,
+      "grad_norm": 0.297498404277643,
+      "learning_rate": 4.0636501099294e-06,
+      "loss": 0.2797,
+      "step": 914
+    },
+    {
+      "epoch": 1.4076923076923076,
+      "grad_norm": 0.287113462303936,
+      "learning_rate": 4.044220172479675e-06,
+      "loss": 0.2787,
+      "step": 915
+    },
+    {
+      "epoch": 1.4092307692307693,
+      "grad_norm": 0.30033725563998004,
+      "learning_rate": 4.024825016782727e-06,
+      "loss": 0.2793,
+      "step": 916
+    },
+    {
+      "epoch": 1.4107692307692308,
+      "grad_norm": 0.2887459684418934,
+      "learning_rate": 4.0054647561062625e-06,
+      "loss": 0.2676,
+      "step": 917
+    },
+    {
+      "epoch": 1.4123076923076923,
+      "grad_norm": 0.28514681381668844,
+      "learning_rate": 3.9861395035141936e-06,
+      "loss": 0.2867,
+      "step": 918
+    },
+    {
+      "epoch": 1.413846153846154,
+      "grad_norm": 0.27472517359961973,
+      "learning_rate": 3.9668493718659924e-06,
+      "loss": 0.2636,
+      "step": 919
+    },
+    {
+      "epoch": 1.4153846153846155,
+      "grad_norm": 0.29407748344726986,
+      "learning_rate": 3.947594473816026e-06,
+      "loss": 0.2821,
+      "step": 920
+    },
+    {
+      "epoch": 1.416923076923077,
+      "grad_norm": 0.2704272597057848,
+      "learning_rate": 3.9283749218128885e-06,
+      "loss": 0.2749,
+      "step": 921
+    },
+    {
+      "epoch": 1.4184615384615384,
+      "grad_norm": 0.27420846427149403,
+      "learning_rate": 3.909190828098766e-06,
+      "loss": 0.2811,
+      "step": 922
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.2797656797273435,
+      "learning_rate": 3.890042304708758e-06,
+      "loss": 0.2773,
+      "step": 923
+    },
+    {
+      "epoch": 1.4215384615384616,
+      "grad_norm": 0.3078131735166291,
+      "learning_rate": 3.8709294634702374e-06,
+      "loss": 0.2727,
+      "step": 924
+    },
+    {
+      "epoch": 1.4230769230769231,
+      "grad_norm": 0.28470564287868144,
+      "learning_rate": 3.8518524160021876e-06,
+      "loss": 0.2739,
+      "step": 925
+    },
+    {
+      "epoch": 1.4246153846153846,
+      "grad_norm": 0.3020599963499438,
+      "learning_rate": 3.832811273714569e-06,
+      "loss": 0.2793,
+      "step": 926
+    },
+    {
+      "epoch": 1.426153846153846,
+      "grad_norm": 0.2849356503004256,
+      "learning_rate": 3.813806147807645e-06,
+      "loss": 0.2743,
+      "step": 927
+    },
+    {
+      "epoch": 1.4276923076923076,
+      "grad_norm": 0.2933219685932077,
+      "learning_rate": 3.7948371492713454e-06,
+      "loss": 0.2699,
+      "step": 928
+    },
+    {
+      "epoch": 1.4292307692307693,
+      "grad_norm": 0.26565134002528773,
+      "learning_rate": 3.775904388884618e-06,
+      "loss": 0.2733,
+      "step": 929
+    },
+    {
+      "epoch": 1.4307692307692308,
+      "grad_norm": 0.28025463474691675,
+      "learning_rate": 3.7570079772147748e-06,
+      "loss": 0.2706,
+      "step": 930
+    },
+    {
+      "epoch": 1.4323076923076923,
+      "grad_norm": 0.2977520733153107,
+      "learning_rate": 3.738148024616863e-06,
+      "loss": 0.2881,
+      "step": 931
+    },
+    {
+      "epoch": 1.4338461538461538,
+      "grad_norm": 0.28252543847339456,
+      "learning_rate": 3.7193246412329976e-06,
+      "loss": 0.2812,
+      "step": 932
+    },
+    {
+      "epoch": 1.4353846153846153,
+      "grad_norm": 0.2870050233029735,
+      "learning_rate": 3.700537936991733e-06,
+      "loss": 0.265,
+      "step": 933
+    },
+    {
+      "epoch": 1.436923076923077,
+      "grad_norm": 0.2652244026722324,
+      "learning_rate": 3.681788021607413e-06,
+      "loss": 0.2817,
+      "step": 934
+    },
+    {
+      "epoch": 1.4384615384615385,
+      "grad_norm": 0.31013959013850706,
+      "learning_rate": 3.6630750045795472e-06,
+      "loss": 0.2798,
+      "step": 935
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.2743987811436827,
+      "learning_rate": 3.6443989951921478e-06,
+      "loss": 0.2696,
+      "step": 936
+    },
+    {
+      "epoch": 1.4415384615384617,
+      "grad_norm": 0.27339074035174915,
+      "learning_rate": 3.625760102513103e-06,
+      "loss": 0.2733,
+      "step": 937
+    },
+    {
+      "epoch": 1.4430769230769231,
+      "grad_norm": 0.2839064463945384,
+      "learning_rate": 3.607158435393544e-06,
+      "loss": 0.2796,
+      "step": 938
+    },
+    {
+      "epoch": 1.4446153846153846,
+      "grad_norm": 0.295811608192796,
+      "learning_rate": 3.5885941024672e-06,
+      "loss": 0.2845,
+      "step": 939
+    },
+    {
+      "epoch": 1.4461538461538461,
+      "grad_norm": 0.29015867105530985,
+      "learning_rate": 3.5700672121497728e-06,
+      "loss": 0.2701,
+      "step": 940
+    },
+    {
+      "epoch": 1.4476923076923076,
+      "grad_norm": 0.2693689280307087,
+      "learning_rate": 3.5515778726382967e-06,
+      "loss": 0.2797,
+      "step": 941
+    },
+    {
+      "epoch": 1.4492307692307693,
+      "grad_norm": 0.2945264603333907,
+      "learning_rate": 3.53312619191051e-06,
+      "loss": 0.278,
+      "step": 942
+    },
+    {
+      "epoch": 1.4507692307692308,
+      "grad_norm": 0.2705312345116653,
+      "learning_rate": 3.5147122777242203e-06,
+      "loss": 0.28,
+      "step": 943
+    },
+    {
+      "epoch": 1.4523076923076923,
+      "grad_norm": 0.2847915292490043,
+      "learning_rate": 3.4963362376166886e-06,
+      "loss": 0.2777,
+      "step": 944
+    },
+    {
+      "epoch": 1.4538461538461538,
+      "grad_norm": 0.2753826997127367,
+      "learning_rate": 3.477998178903982e-06,
+      "loss": 0.2787,
+      "step": 945
+    },
+    {
+      "epoch": 1.4553846153846153,
+      "grad_norm": 0.27318410243587926,
+      "learning_rate": 3.459698208680359e-06,
+      "loss": 0.2765,
+      "step": 946
+    },
+    {
+      "epoch": 1.456923076923077,
+      "grad_norm": 0.28833597885945284,
+      "learning_rate": 3.441436433817641e-06,
+      "loss": 0.2708,
+      "step": 947
+    },
+    {
+      "epoch": 1.4584615384615385,
+      "grad_norm": 0.27620830508227745,
+      "learning_rate": 3.423212960964586e-06,
+      "loss": 0.2758,
+      "step": 948
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.271564199389485,
+      "learning_rate": 3.405027896546277e-06,
+      "loss": 0.2805,
+      "step": 949
+    },
+    {
+      "epoch": 1.4615384615384617,
+      "grad_norm": 0.26988340339307304,
+      "learning_rate": 3.3868813467634833e-06,
+      "loss": 0.2791,
+      "step": 950
+    },
+    {
+      "epoch": 1.463076923076923,
+      "grad_norm": 0.2838014425158747,
+      "learning_rate": 3.3687734175920505e-06,
+      "loss": 0.2716,
+      "step": 951
+    },
+    {
+      "epoch": 1.4646153846153847,
+      "grad_norm": 0.27584807857810434,
+      "learning_rate": 3.350704214782278e-06,
+      "loss": 0.2627,
+      "step": 952
+    },
+    {
+      "epoch": 1.4661538461538461,
+      "grad_norm": 0.2750476670562843,
+      "learning_rate": 3.3326738438583116e-06,
+      "loss": 0.2671,
+      "step": 953
+    },
+    {
+      "epoch": 1.4676923076923076,
+      "grad_norm": 0.27704107334718375,
+      "learning_rate": 3.314682410117511e-06,
+      "loss": 0.2802,
+      "step": 954
+    },
+    {
+      "epoch": 1.4692307692307693,
+      "grad_norm": 0.2649151011885699,
+      "learning_rate": 3.2967300186298456e-06,
+      "loss": 0.2817,
+      "step": 955
+    },
+    {
+      "epoch": 1.4707692307692308,
+      "grad_norm": 0.2815722650702043,
+      "learning_rate": 3.2788167742372725e-06,
+      "loss": 0.2706,
+      "step": 956
+    },
+    {
+      "epoch": 1.4723076923076923,
+      "grad_norm": 0.287669633361647,
+      "learning_rate": 3.2609427815531426e-06,
+      "loss": 0.2816,
+      "step": 957
+    },
+    {
+      "epoch": 1.4738461538461538,
+      "grad_norm": 0.2759316296845802,
+      "learning_rate": 3.243108144961563e-06,
+      "loss": 0.2745,
+      "step": 958
+    },
+    {
+      "epoch": 1.4753846153846153,
+      "grad_norm": 0.2928929578884538,
+      "learning_rate": 3.2253129686168105e-06,
+      "loss": 0.2781,
+      "step": 959
+    },
+    {
+      "epoch": 1.476923076923077,
+      "grad_norm": 0.28769583350600747,
+      "learning_rate": 3.2075573564427097e-06,
+      "loss": 0.275,
+      "step": 960
+    },
+    {
+      "epoch": 1.4784615384615385,
+      "grad_norm": 0.2950526179921246,
+      "learning_rate": 3.1898414121320277e-06,
+      "loss": 0.2746,
+      "step": 961
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.28082302779472623,
+      "learning_rate": 3.1721652391458804e-06,
+      "loss": 0.2731,
+      "step": 962
+    },
+    {
+      "epoch": 1.4815384615384615,
+      "grad_norm": 0.3037822234665378,
+      "learning_rate": 3.1545289407131128e-06,
+      "loss": 0.2846,
+      "step": 963
+    },
+    {
+      "epoch": 1.483076923076923,
+      "grad_norm": 0.26541552279778613,
+      "learning_rate": 3.1369326198297025e-06,
+      "loss": 0.2707,
+      "step": 964
+    },
+    {
+      "epoch": 1.4846153846153847,
+      "grad_norm": 0.27228670305719155,
+      "learning_rate": 3.11937637925816e-06,
+      "loss": 0.2907,
+      "step": 965
+    },
+    {
+      "epoch": 1.4861538461538462,
+      "grad_norm": 0.2898752439173941,
+      "learning_rate": 3.101860321526924e-06,
+      "loss": 0.2787,
+      "step": 966
+    },
+    {
+      "epoch": 1.4876923076923076,
+      "grad_norm": 0.28760668304946113,
+      "learning_rate": 3.0843845489297698e-06,
+      "loss": 0.2771,
+      "step": 967
+    },
+    {
+      "epoch": 1.4892307692307694,
+      "grad_norm": 0.26927083755746745,
+      "learning_rate": 3.066949163525205e-06,
+      "loss": 0.2884,
+      "step": 968
+    },
+    {
+      "epoch": 1.4907692307692308,
+      "grad_norm": 0.2722526618027186,
+      "learning_rate": 3.0495542671358745e-06,
+      "loss": 0.2776,
+      "step": 969
+    },
+    {
+      "epoch": 1.4923076923076923,
+      "grad_norm": 0.2628110690196813,
+      "learning_rate": 3.0321999613479668e-06,
+      "loss": 0.2756,
+      "step": 970
+    },
+    {
+      "epoch": 1.4938461538461538,
+      "grad_norm": 0.27947564803570074,
+      "learning_rate": 3.0148863475106315e-06,
+      "loss": 0.2716,
+      "step": 971
+    },
+    {
+      "epoch": 1.4953846153846153,
+      "grad_norm": 0.28847862238255323,
+      "learning_rate": 2.9976135267353636e-06,
+      "loss": 0.2705,
+      "step": 972
+    },
+    {
+      "epoch": 1.496923076923077,
+      "grad_norm": 0.2830722792497127,
+      "learning_rate": 2.9803815998954334e-06,
+      "loss": 0.2694,
+      "step": 973
+    },
+    {
+      "epoch": 1.4984615384615385,
+      "grad_norm": 0.2825543118881783,
+      "learning_rate": 2.9631906676252865e-06,
+      "loss": 0.2688,
+      "step": 974
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.2794572077398108,
+      "learning_rate": 2.9460408303199696e-06,
+      "loss": 0.2772,
+      "step": 975
+    },
+    {
+      "epoch": 1.5015384615384615,
+      "grad_norm": 0.26354437004415276,
+      "learning_rate": 2.9289321881345257e-06,
+      "loss": 0.289,
+      "step": 976
+    },
+    {
+      "epoch": 1.503076923076923,
+      "grad_norm": 0.2999748853854813,
+      "learning_rate": 2.9118648409834205e-06,
+      "loss": 0.2797,
+      "step": 977
+    },
+    {
+      "epoch": 1.5046153846153847,
+      "grad_norm": 0.26897393008418174,
+      "learning_rate": 2.894838888539957e-06,
+      "loss": 0.2799,
+      "step": 978
+    },
+    {
+      "epoch": 1.5061538461538462,
+      "grad_norm": 0.2756670966010222,
+      "learning_rate": 2.8778544302356904e-06,
+      "loss": 0.2669,
+      "step": 979
+    },
+    {
+      "epoch": 1.5076923076923077,
+      "grad_norm": 0.28448485978171967,
+      "learning_rate": 2.8609115652598595e-06,
+      "loss": 0.2748,
+      "step": 980
+    },
+    {
+      "epoch": 1.5092307692307694,
+      "grad_norm": 0.2750366525351591,
+      "learning_rate": 2.8440103925587904e-06,
+      "loss": 0.2719,
+      "step": 981
+    },
+    {
+      "epoch": 1.5107692307692306,
+      "grad_norm": 0.2759733005813308,
+      "learning_rate": 2.8271510108353237e-06,
+      "loss": 0.2801,
+      "step": 982
+    },
+    {
+      "epoch": 1.5123076923076924,
+      "grad_norm": 0.2723203530361154,
+      "learning_rate": 2.810333518548246e-06,
+      "loss": 0.2772,
+      "step": 983
+    },
+    {
+      "epoch": 1.5138461538461538,
+      "grad_norm": 0.27696349217523963,
+      "learning_rate": 2.7935580139117114e-06,
+      "loss": 0.2807,
+      "step": 984
+    },
+    {
+      "epoch": 1.5153846153846153,
+      "grad_norm": 0.27417573211998586,
+      "learning_rate": 2.7768245948946615e-06,
+      "loss": 0.2768,
+      "step": 985
+    },
+    {
+      "epoch": 1.516923076923077,
+      "grad_norm": 0.2857478167106004,
+      "learning_rate": 2.7601333592202583e-06,
+      "loss": 0.2789,
+      "step": 986
+    },
+    {
+      "epoch": 1.5184615384615383,
+      "grad_norm": 0.27385986851994265,
+      "learning_rate": 2.743484404365314e-06,
+      "loss": 0.2829,
+      "step": 987
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.2719273680384911,
+      "learning_rate": 2.7268778275597217e-06,
+      "loss": 0.2735,
+      "step": 988
+    },
+    {
+      "epoch": 1.5215384615384615,
+      "grad_norm": 0.277913548440064,
+      "learning_rate": 2.7103137257858867e-06,
+      "loss": 0.2742,
+      "step": 989
+    },
+    {
+      "epoch": 1.523076923076923,
+      "grad_norm": 0.2655837420363689,
+      "learning_rate": 2.6937921957781587e-06,
+      "loss": 0.2737,
+      "step": 990
+    },
+    {
+      "epoch": 1.5246153846153847,
+      "grad_norm": 0.25939819853408147,
+      "learning_rate": 2.6773133340222677e-06,
+      "loss": 0.2753,
+      "step": 991
+    },
+    {
+      "epoch": 1.5261538461538462,
+      "grad_norm": 0.299768904794166,
+      "learning_rate": 2.660877236754762e-06,
+      "loss": 0.2774,
+      "step": 992
+    },
+    {
+      "epoch": 1.5276923076923077,
+      "grad_norm": 0.2918871574747949,
+      "learning_rate": 2.6444839999624496e-06,
+      "loss": 0.2698,
+      "step": 993
+    },
+    {
+      "epoch": 1.5292307692307694,
+      "grad_norm": 0.27965606944159893,
+      "learning_rate": 2.6281337193818267e-06,
+      "loss": 0.2813,
+      "step": 994
+    },
+    {
+      "epoch": 1.5307692307692307,
+      "grad_norm": 0.2699316433426434,
+      "learning_rate": 2.61182649049853e-06,
+      "loss": 0.264,
+      "step": 995
+    },
+    {
+      "epoch": 1.5323076923076924,
+      "grad_norm": 0.2674235361378631,
+      "learning_rate": 2.59556240854677e-06,
+      "loss": 0.2709,
+      "step": 996
+    },
+    {
+      "epoch": 1.5338461538461539,
+      "grad_norm": 0.2737319487789186,
+      "learning_rate": 2.5793415685087797e-06,
+      "loss": 0.2776,
+      "step": 997
+    },
+    {
+      "epoch": 1.5353846153846153,
+      "grad_norm": 0.2554653387177025,
+      "learning_rate": 2.5631640651142654e-06,
+      "loss": 0.2775,
+      "step": 998
+    },
+    {
+      "epoch": 1.536923076923077,
+      "grad_norm": 0.2805688350077011,
+      "learning_rate": 2.5470299928398424e-06,
+      "loss": 0.2777,
+      "step": 999
+    },
+    {
+      "epoch": 1.5384615384615383,
+      "grad_norm": 0.25830358096575184,
+      "learning_rate": 2.5309394459084878e-06,
+      "loss": 0.2661,
+      "step": 1000
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.26874499745479663,
+      "learning_rate": 2.514892518288988e-06,
+      "loss": 0.2671,
+      "step": 1001
+    },
+    {
+      "epoch": 1.5415384615384615,
+      "grad_norm": 0.25765810204174705,
+      "learning_rate": 2.4988893036954045e-06,
+      "loss": 0.2688,
+      "step": 1002
+    },
+    {
+      "epoch": 1.543076923076923,
+      "grad_norm": 0.2716791072454054,
+      "learning_rate": 2.4829298955865022e-06,
+      "loss": 0.2714,
+      "step": 1003
+    },
+    {
+      "epoch": 1.5446153846153847,
+      "grad_norm": 0.2765676929624752,
+      "learning_rate": 2.467014387165222e-06,
+      "loss": 0.2815,
+      "step": 1004
+    },
+    {
+      "epoch": 1.546153846153846,
+      "grad_norm": 0.2892228732777564,
+      "learning_rate": 2.451142871378124e-06,
+      "loss": 0.2854,
+      "step": 1005
+    },
+    {
+      "epoch": 1.5476923076923077,
+      "grad_norm": 0.279076193937244,
+      "learning_rate": 2.4353154409148637e-06,
+      "loss": 0.2767,
+      "step": 1006
+    },
+    {
+      "epoch": 1.5492307692307692,
+      "grad_norm": 0.25962663695514016,
+      "learning_rate": 2.4195321882076295e-06,
+      "loss": 0.2746,
+      "step": 1007
+    },
+    {
+      "epoch": 1.5507692307692307,
+      "grad_norm": 0.2727392844670763,
+      "learning_rate": 2.4037932054306125e-06,
+      "loss": 0.2737,
+      "step": 1008
+    },
+    {
+      "epoch": 1.5523076923076924,
+      "grad_norm": 0.26846928534735465,
+      "learning_rate": 2.3880985844994674e-06,
+      "loss": 0.2759,
+      "step": 1009
+    },
+    {
+      "epoch": 1.5538461538461539,
+      "grad_norm": 0.26462438345050504,
+      "learning_rate": 2.372448417070776e-06,
+      "loss": 0.2716,
+      "step": 1010
+    },
+    {
+      "epoch": 1.5553846153846154,
+      "grad_norm": 0.3037970471484453,
+      "learning_rate": 2.3568427945415163e-06,
+      "loss": 0.2744,
+      "step": 1011
+    },
+    {
+      "epoch": 1.556923076923077,
+      "grad_norm": 0.2753805507186117,
+      "learning_rate": 2.3412818080485176e-06,
+      "loss": 0.2777,
+      "step": 1012
+    },
+    {
+      "epoch": 1.5584615384615383,
+      "grad_norm": 0.26354274514511733,
+      "learning_rate": 2.3257655484679376e-06,
+      "loss": 0.2638,
+      "step": 1013
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.2883351270138847,
+      "learning_rate": 2.3102941064147287e-06,
+      "loss": 0.2769,
+      "step": 1014
+    },
+    {
+      "epoch": 1.5615384615384615,
+      "grad_norm": 0.2930116252923399,
+      "learning_rate": 2.2948675722421086e-06,
+      "loss": 0.2701,
+      "step": 1015
+    },
+    {
+      "epoch": 1.563076923076923,
+      "grad_norm": 0.27465138940322426,
+      "learning_rate": 2.279486036041034e-06,
+      "loss": 0.274,
+      "step": 1016
+    },
+    {
+      "epoch": 1.5646153846153847,
+      "grad_norm": 0.2785774804014197,
+      "learning_rate": 2.2641495876396713e-06,
+      "loss": 0.2909,
+      "step": 1017
+    },
+    {
+      "epoch": 1.566153846153846,
+      "grad_norm": 0.2957628352417127,
+      "learning_rate": 2.2488583166028754e-06,
+      "loss": 0.2765,
+      "step": 1018
+    },
+    {
+      "epoch": 1.5676923076923077,
+      "grad_norm": 0.2751446889954304,
+      "learning_rate": 2.2336123122316642e-06,
+      "loss": 0.2723,
+      "step": 1019
+    },
+    {
+      "epoch": 1.5692307692307692,
+      "grad_norm": 0.2636974738762224,
+      "learning_rate": 2.218411663562704e-06,
+      "loss": 0.2783,
+      "step": 1020
+    },
+    {
+      "epoch": 1.5707692307692307,
+      "grad_norm": 0.2830324109973893,
+      "learning_rate": 2.2032564593677773e-06,
+      "loss": 0.2781,
+      "step": 1021
+    },
+    {
+      "epoch": 1.5723076923076924,
+      "grad_norm": 0.2727444671070684,
+      "learning_rate": 2.1881467881532737e-06,
+      "loss": 0.2695,
+      "step": 1022
+    },
+    {
+      "epoch": 1.573846153846154,
+      "grad_norm": 0.27167201485030423,
+      "learning_rate": 2.1730827381596643e-06,
+      "loss": 0.2694,
+      "step": 1023
+    },
+    {
+      "epoch": 1.5753846153846154,
+      "grad_norm": 0.2912522223958155,
+      "learning_rate": 2.158064397361005e-06,
+      "loss": 0.2725,
+      "step": 1024
+    },
+    {
+      "epoch": 1.5769230769230769,
+      "grad_norm": 0.27334940402277946,
+      "learning_rate": 2.1430918534643996e-06,
+      "loss": 0.2778,
+      "step": 1025
+    },
+    {
+      "epoch": 1.5784615384615384,
+      "grad_norm": 0.28204943693802564,
+      "learning_rate": 2.1281651939094996e-06,
+      "loss": 0.2773,
+      "step": 1026
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.2675264102046118,
+      "learning_rate": 2.1132845058679942e-06,
+      "loss": 0.284,
+      "step": 1027
+    },
+    {
+      "epoch": 1.5815384615384616,
+      "grad_norm": 0.26218450323544323,
+      "learning_rate": 2.098449876243096e-06,
+      "loss": 0.2757,
+      "step": 1028
+    },
+    {
+      "epoch": 1.583076923076923,
+      "grad_norm": 0.27703152952301313,
+      "learning_rate": 2.083661391669043e-06,
+      "loss": 0.2834,
+      "step": 1029
+    },
+    {
+      "epoch": 1.5846153846153848,
+      "grad_norm": 0.2662129547997518,
+      "learning_rate": 2.0689191385105787e-06,
+      "loss": 0.2676,
+      "step": 1030
+    },
+    {
+      "epoch": 1.586153846153846,
+      "grad_norm": 0.26446456191661216,
+      "learning_rate": 2.0542232028624585e-06,
+      "loss": 0.2694,
+      "step": 1031
+    },
+    {
+      "epoch": 1.5876923076923077,
+      "grad_norm": 0.2627648312541549,
+      "learning_rate": 2.03957367054894e-06,
+      "loss": 0.2711,
+      "step": 1032
+    },
+    {
+      "epoch": 1.5892307692307692,
+      "grad_norm": 0.27395360055874696,
+      "learning_rate": 2.024970627123295e-06,
+      "loss": 0.2726,
+      "step": 1033
+    },
+    {
+      "epoch": 1.5907692307692307,
+      "grad_norm": 0.2697412149903341,
+      "learning_rate": 2.0104141578672887e-06,
+      "loss": 0.2735,
+      "step": 1034
+    },
+    {
+      "epoch": 1.5923076923076924,
+      "grad_norm": 0.2754963765655869,
+      "learning_rate": 1.9959043477907e-06,
+      "loss": 0.2691,
+      "step": 1035
+    },
+    {
+      "epoch": 1.5938461538461537,
+      "grad_norm": 0.262332165634561,
+      "learning_rate": 1.981441281630816e-06,
+      "loss": 0.2679,
+      "step": 1036
+    },
+    {
+      "epoch": 1.5953846153846154,
+      "grad_norm": 0.28477683389530106,
+      "learning_rate": 1.967025043851939e-06,
+      "loss": 0.2661,
+      "step": 1037
+    },
+    {
+      "epoch": 1.596923076923077,
+      "grad_norm": 0.26484391672500696,
+      "learning_rate": 1.9526557186448924e-06,
+      "loss": 0.2843,
+      "step": 1038
+    },
+    {
+      "epoch": 1.5984615384615384,
+      "grad_norm": 0.27334779504852813,
+      "learning_rate": 1.9383333899265368e-06,
+      "loss": 0.2818,
+      "step": 1039
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.2848366958654719,
+      "learning_rate": 1.9240581413392647e-06,
+      "loss": 0.2685,
+      "step": 1040
+    },
+    {
+      "epoch": 1.6015384615384616,
+      "grad_norm": 0.2544014207312848,
+      "learning_rate": 1.9098300562505266e-06,
+      "loss": 0.2827,
+      "step": 1041
+    },
+    {
+      "epoch": 1.603076923076923,
+      "grad_norm": 0.2810041947592908,
+      "learning_rate": 1.8956492177523345e-06,
+      "loss": 0.2774,
+      "step": 1042
+    },
+    {
+      "epoch": 1.6046153846153848,
+      "grad_norm": 0.27727230328919683,
+      "learning_rate": 1.8815157086607826e-06,
+      "loss": 0.2801,
+      "step": 1043
+    },
+    {
+      "epoch": 1.606153846153846,
+      "grad_norm": 0.27123994829514625,
+      "learning_rate": 1.86742961151556e-06,
+      "loss": 0.2694,
+      "step": 1044
+    },
+    {
+      "epoch": 1.6076923076923078,
+      "grad_norm": 0.2754553311610714,
+      "learning_rate": 1.8533910085794714e-06,
+      "loss": 0.2719,
+      "step": 1045
+    },
+    {
+      "epoch": 1.6092307692307692,
+      "grad_norm": 0.27863073636816,
+      "learning_rate": 1.8393999818379527e-06,
+      "loss": 0.2645,
+      "step": 1046
+    },
+    {
+      "epoch": 1.6107692307692307,
+      "grad_norm": 0.2671126481268661,
+      "learning_rate": 1.8254566129985996e-06,
+      "loss": 0.2761,
+      "step": 1047
+    },
+    {
+      "epoch": 1.6123076923076924,
+      "grad_norm": 0.26267406682118105,
+      "learning_rate": 1.8115609834906821e-06,
+      "loss": 0.2818,
+      "step": 1048
+    },
+    {
+      "epoch": 1.6138461538461537,
+      "grad_norm": 0.2671493987923625,
+      "learning_rate": 1.7977131744646724e-06,
+      "loss": 0.2789,
+      "step": 1049
+    },
+    {
+      "epoch": 1.6153846153846154,
+      "grad_norm": 0.2505733188596916,
+      "learning_rate": 1.7839132667917692e-06,
+      "loss": 0.2859,
+      "step": 1050
+    },
+    {
+      "epoch": 1.616923076923077,
+      "grad_norm": 0.263775565972645,
+      "learning_rate": 1.7701613410634367e-06,
+      "loss": 0.2693,
+      "step": 1051
+    },
+    {
+      "epoch": 1.6184615384615384,
+      "grad_norm": 0.2702206938718406,
+      "learning_rate": 1.7564574775909127e-06,
+      "loss": 0.2669,
+      "step": 1052
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.26997570628252127,
+      "learning_rate": 1.7428017564047594e-06,
+      "loss": 0.2555,
+      "step": 1053
+    },
+    {
+      "epoch": 1.6215384615384614,
+      "grad_norm": 0.2662750732164586,
+      "learning_rate": 1.7291942572543806e-06,
+      "loss": 0.289,
+      "step": 1054
+    },
+    {
+      "epoch": 1.623076923076923,
+      "grad_norm": 0.2639379570596132,
+      "learning_rate": 1.7156350596075743e-06,
+      "loss": 0.2769,
+      "step": 1055
+    },
+    {
+      "epoch": 1.6246153846153846,
+      "grad_norm": 0.27426927931869355,
+      "learning_rate": 1.7021242426500495e-06,
+      "loss": 0.2741,
+      "step": 1056
+    },
+    {
+      "epoch": 1.626153846153846,
+      "grad_norm": 0.2662514430923517,
+      "learning_rate": 1.6886618852849723e-06,
+      "loss": 0.2705,
+      "step": 1057
+    },
+    {
+      "epoch": 1.6276923076923078,
+      "grad_norm": 0.28223326673461585,
+      "learning_rate": 1.6752480661325077e-06,
+      "loss": 0.2763,
+      "step": 1058
+    },
+    {
+      "epoch": 1.6292307692307693,
+      "grad_norm": 0.2729469232622983,
+      "learning_rate": 1.6618828635293538e-06,
+      "loss": 0.2622,
+      "step": 1059
+    },
+    {
+      "epoch": 1.6307692307692307,
+      "grad_norm": 0.2709645746392681,
+      "learning_rate": 1.6485663555282949e-06,
+      "loss": 0.2725,
+      "step": 1060
+    },
+    {
+      "epoch": 1.6323076923076925,
+      "grad_norm": 0.26714503175176485,
+      "learning_rate": 1.6352986198977327e-06,
+      "loss": 0.2732,
+      "step": 1061
+    },
+    {
+      "epoch": 1.6338461538461537,
+      "grad_norm": 0.2772109598438791,
+      "learning_rate": 1.6220797341212401e-06,
+      "loss": 0.2662,
+      "step": 1062
+    },
+    {
+      "epoch": 1.6353846153846154,
+      "grad_norm": 0.2753978577038448,
+      "learning_rate": 1.6089097753971061e-06,
+      "loss": 0.2687,
+      "step": 1063
+    },
+    {
+      "epoch": 1.636923076923077,
+      "grad_norm": 0.27445806854845634,
+      "learning_rate": 1.595788820637888e-06,
+      "loss": 0.2647,
+      "step": 1064
+    },
+    {
+      "epoch": 1.6384615384615384,
+      "grad_norm": 0.2546044446500755,
+      "learning_rate": 1.5827169464699576e-06,
+      "loss": 0.2795,
+      "step": 1065
+    },
+    {
+      "epoch": 1.6400000000000001,
+      "grad_norm": 0.26446018640312374,
+      "learning_rate": 1.5696942292330574e-06,
+      "loss": 0.2855,
+      "step": 1066
+    },
+    {
+      "epoch": 1.6415384615384614,
+      "grad_norm": 0.2703615121193336,
+      "learning_rate": 1.5567207449798517e-06,
+      "loss": 0.2754,
+      "step": 1067
+    },
+    {
+      "epoch": 1.643076923076923,
+      "grad_norm": 0.25940079508718933,
+      "learning_rate": 1.5437965694754842e-06,
+      "loss": 0.2724,
+      "step": 1068
+    },
+    {
+      "epoch": 1.6446153846153846,
+      "grad_norm": 0.2938743036524522,
+      "learning_rate": 1.5309217781971419e-06,
+      "loss": 0.2788,
+      "step": 1069
+    },
+    {
+      "epoch": 1.646153846153846,
+      "grad_norm": 0.27539329200714313,
+      "learning_rate": 1.518096446333599e-06,
+      "loss": 0.279,
+      "step": 1070
+    },
+    {
+      "epoch": 1.6476923076923078,
+      "grad_norm": 0.27399731752015927,
+      "learning_rate": 1.5053206487847916e-06,
+      "loss": 0.2707,
+      "step": 1071
+    },
+    {
+      "epoch": 1.6492307692307693,
+      "grad_norm": 0.2932148175507058,
+      "learning_rate": 1.4925944601613718e-06,
+      "loss": 0.2742,
+      "step": 1072
+    },
+    {
+      "epoch": 1.6507692307692308,
+      "grad_norm": 0.2460474178500085,
+      "learning_rate": 1.4799179547842823e-06,
+      "loss": 0.2737,
+      "step": 1073
+    },
+    {
+      "epoch": 1.6523076923076923,
+      "grad_norm": 0.24752302212753835,
+      "learning_rate": 1.4672912066843103e-06,
+      "loss": 0.2684,
+      "step": 1074
+    },
+    {
+      "epoch": 1.6538461538461537,
+      "grad_norm": 0.27662182502454175,
+      "learning_rate": 1.454714289601661e-06,
+      "loss": 0.2768,
+      "step": 1075
+    },
+    {
+      "epoch": 1.6553846153846155,
+      "grad_norm": 0.25210921110835766,
+      "learning_rate": 1.4421872769855262e-06,
+      "loss": 0.2781,
+      "step": 1076
+    },
+    {
+      "epoch": 1.656923076923077,
+      "grad_norm": 0.2645704559801182,
+      "learning_rate": 1.4297102419936559e-06,
+      "loss": 0.2754,
+      "step": 1077
+    },
+    {
+      "epoch": 1.6584615384615384,
+      "grad_norm": 0.2583307882783813,
+      "learning_rate": 1.4172832574919359e-06,
+      "loss": 0.2716,
+      "step": 1078
+    },
+    {
+      "epoch": 1.6600000000000001,
+      "grad_norm": 0.2714938620970922,
+      "learning_rate": 1.4049063960539488e-06,
+      "loss": 0.2794,
+      "step": 1079
+    },
+    {
+      "epoch": 1.6615384615384614,
+      "grad_norm": 0.25378212618966794,
+      "learning_rate": 1.3925797299605649e-06,
+      "loss": 0.2748,
+      "step": 1080
+    },
+    {
+      "epoch": 1.6630769230769231,
+      "grad_norm": 0.26790623929740864,
+      "learning_rate": 1.3803033311995072e-06,
+      "loss": 0.2677,
+      "step": 1081
+    },
+    {
+      "epoch": 1.6646153846153846,
+      "grad_norm": 0.28686817516841456,
+      "learning_rate": 1.368077271464946e-06,
+      "loss": 0.2612,
+      "step": 1082
+    },
+    {
+      "epoch": 1.666153846153846,
+      "grad_norm": 0.2608494931055921,
+      "learning_rate": 1.3559016221570663e-06,
+      "loss": 0.283,
+      "step": 1083
+    },
+    {
+      "epoch": 1.6676923076923078,
+      "grad_norm": 0.2679051422710405,
+      "learning_rate": 1.3437764543816556e-06,
+      "loss": 0.2754,
+      "step": 1084
+    },
+    {
+      "epoch": 1.669230769230769,
+      "grad_norm": 0.25609847300489263,
+      "learning_rate": 1.3317018389496927e-06,
+      "loss": 0.286,
+      "step": 1085
+    },
+    {
+      "epoch": 1.6707692307692308,
+      "grad_norm": 0.26682787951436476,
+      "learning_rate": 1.3196778463769256e-06,
+      "loss": 0.2718,
+      "step": 1086
+    },
+    {
+      "epoch": 1.6723076923076923,
+      "grad_norm": 0.28056506397402664,
+      "learning_rate": 1.3077045468834714e-06,
+      "loss": 0.2601,
+      "step": 1087
+    },
+    {
+      "epoch": 1.6738461538461538,
+      "grad_norm": 0.24980681917796863,
+      "learning_rate": 1.295782010393396e-06,
+      "loss": 0.2785,
+      "step": 1088
+    },
+    {
+      "epoch": 1.6753846153846155,
+      "grad_norm": 0.25805836310849545,
+      "learning_rate": 1.2839103065343084e-06,
+      "loss": 0.2783,
+      "step": 1089
+    },
+    {
+      "epoch": 1.676923076923077,
+      "grad_norm": 0.274089796066655,
+      "learning_rate": 1.2720895046369564e-06,
+      "loss": 0.2695,
+      "step": 1090
+    },
+    {
+      "epoch": 1.6784615384615384,
+      "grad_norm": 0.26254269044134937,
+      "learning_rate": 1.2603196737348211e-06,
+      "loss": 0.2743,
+      "step": 1091
+    },
+    {
+      "epoch": 1.6800000000000002,
+      "grad_norm": 0.2617720717953636,
+      "learning_rate": 1.2486008825637119e-06,
+      "loss": 0.2612,
+      "step": 1092
+    },
+    {
+      "epoch": 1.6815384615384614,
+      "grad_norm": 0.2759743866219765,
+      "learning_rate": 1.2369331995613664e-06,
+      "loss": 0.2727,
+      "step": 1093
+    },
+    {
+      "epoch": 1.6830769230769231,
+      "grad_norm": 0.25653527827770223,
+      "learning_rate": 1.2253166928670478e-06,
+      "loss": 0.2747,
+      "step": 1094
+    },
+    {
+      "epoch": 1.6846153846153846,
+      "grad_norm": 0.25821672413208674,
+      "learning_rate": 1.213751430321156e-06,
+      "loss": 0.2704,
+      "step": 1095
+    },
+    {
+      "epoch": 1.6861538461538461,
+      "grad_norm": 0.2639461309608209,
+      "learning_rate": 1.2022374794648229e-06,
+      "loss": 0.2747,
+      "step": 1096
+    },
+    {
+      "epoch": 1.6876923076923078,
+      "grad_norm": 0.2684576261799393,
+      "learning_rate": 1.1907749075395147e-06,
+      "loss": 0.2677,
+      "step": 1097
+    },
+    {
+      "epoch": 1.689230769230769,
+      "grad_norm": 0.25985274031169525,
+      "learning_rate": 1.179363781486651e-06,
+      "loss": 0.2673,
+      "step": 1098
+    },
+    {
+      "epoch": 1.6907692307692308,
+      "grad_norm": 0.2688011611759083,
+      "learning_rate": 1.168004167947202e-06,
+      "loss": 0.2777,
+      "step": 1099
+    },
+    {
+      "epoch": 1.6923076923076923,
+      "grad_norm": 0.28355243170627664,
+      "learning_rate": 1.1566961332613136e-06,
+      "loss": 0.2846,
+      "step": 1100
+    },
+    {
+      "epoch": 1.6938461538461538,
+      "grad_norm": 0.2603109920930245,
+      "learning_rate": 1.1454397434679022e-06,
+      "loss": 0.2752,
+      "step": 1101
+    },
+    {
+      "epoch": 1.6953846153846155,
+      "grad_norm": 0.2895708885344929,
+      "learning_rate": 1.1342350643042822e-06,
+      "loss": 0.269,
+      "step": 1102
+    },
+    {
+      "epoch": 1.696923076923077,
+      "grad_norm": 0.2773699105634792,
+      "learning_rate": 1.123082161205775e-06,
+      "loss": 0.2821,
+      "step": 1103
+    },
+    {
+      "epoch": 1.6984615384615385,
+      "grad_norm": 0.2518089725672006,
+      "learning_rate": 1.111981099305336e-06,
+      "loss": 0.2708,
+      "step": 1104
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.27872001467564333,
+      "learning_rate": 1.1009319434331623e-06,
+      "loss": 0.2693,
+      "step": 1105
+    },
+    {
+      "epoch": 1.7015384615384614,
+      "grad_norm": 0.2605607686909919,
+      "learning_rate": 1.0899347581163222e-06,
+      "loss": 0.2837,
+      "step": 1106
+    },
+    {
+      "epoch": 1.7030769230769232,
+      "grad_norm": 0.26986358015400225,
+      "learning_rate": 1.0789896075783734e-06,
+      "loss": 0.2655,
+      "step": 1107
+    },
+    {
+      "epoch": 1.7046153846153846,
+      "grad_norm": 0.27609853188608596,
+      "learning_rate": 1.0680965557389934e-06,
+      "loss": 0.2793,
+      "step": 1108
+    },
+    {
+      "epoch": 1.7061538461538461,
+      "grad_norm": 0.27846279018096237,
+      "learning_rate": 1.0572556662136036e-06,
+      "loss": 0.2755,
+      "step": 1109
+    },
+    {
+      "epoch": 1.7076923076923078,
+      "grad_norm": 0.2573280880106174,
+      "learning_rate": 1.0464670023129952e-06,
+      "loss": 0.272,
+      "step": 1110
+    },
+    {
+      "epoch": 1.709230769230769,
+      "grad_norm": 0.26407674773187606,
+      "learning_rate": 1.0357306270429623e-06,
+      "loss": 0.2692,
+      "step": 1111
+    },
+    {
+      "epoch": 1.7107692307692308,
+      "grad_norm": 0.2637903458053623,
+      "learning_rate": 1.0250466031039353e-06,
+      "loss": 0.2729,
+      "step": 1112
+    },
+    {
+      "epoch": 1.7123076923076923,
+      "grad_norm": 0.2624748117136115,
+      "learning_rate": 1.014414992890611e-06,
+      "loss": 0.273,
+      "step": 1113
+    },
+    {
+      "epoch": 1.7138461538461538,
+      "grad_norm": 0.29472250276583384,
+      "learning_rate": 1.0038358584915896e-06,
+      "loss": 0.2704,
+      "step": 1114
+    },
+    {
+      "epoch": 1.7153846153846155,
+      "grad_norm": 0.26606886810615454,
+      "learning_rate": 9.93309261689015e-07,
+      "loss": 0.2762,
+      "step": 1115
+    },
+    {
+      "epoch": 1.7169230769230768,
+      "grad_norm": 0.2643526797569404,
+      "learning_rate": 9.828352639582073e-07,
+      "loss": 0.2762,
+      "step": 1116
+    },
+    {
+      "epoch": 1.7184615384615385,
+      "grad_norm": 0.26253274898605866,
+      "learning_rate": 9.724139264673116e-07,
+      "loss": 0.2785,
+      "step": 1117
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.28539607445052123,
+      "learning_rate": 9.6204531007694e-07,
+      "loss": 0.2723,
+      "step": 1118
+    },
+    {
+      "epoch": 1.7215384615384615,
+      "grad_norm": 0.2754770798892866,
+      "learning_rate": 9.517294753398066e-07,
+      "loss": 0.2821,
+      "step": 1119
+    },
+    {
+      "epoch": 1.7230769230769232,
+      "grad_norm": 0.27184406321376414,
+      "learning_rate": 9.414664825003838e-07,
+      "loss": 0.2768,
+      "step": 1120
+    },
+    {
+      "epoch": 1.7246153846153847,
+      "grad_norm": 0.259292322429861,
+      "learning_rate": 9.312563914945461e-07,
+      "loss": 0.2636,
+      "step": 1121
+    },
+    {
+      "epoch": 1.7261538461538461,
+      "grad_norm": 0.26805868801377686,
+      "learning_rate": 9.210992619492254e-07,
+      "loss": 0.2778,
+      "step": 1122
+    },
+    {
+      "epoch": 1.7276923076923076,
+      "grad_norm": 0.25735702425768975,
+      "learning_rate": 9.10995153182056e-07,
+      "loss": 0.2821,
+      "step": 1123
+    },
+    {
+      "epoch": 1.7292307692307691,
+      "grad_norm": 0.25299101144437586,
+      "learning_rate": 9.009441242010287e-07,
+      "loss": 0.2703,
+      "step": 1124
+    },
+    {
+      "epoch": 1.7307692307692308,
+      "grad_norm": 0.2765481093679949,
+      "learning_rate": 8.909462337041508e-07,
+      "loss": 0.2731,
+      "step": 1125
+    },
+    {
+      "epoch": 1.7323076923076923,
+      "grad_norm": 0.2755662572987114,
+      "learning_rate": 8.810015400790994e-07,
+      "loss": 0.2689,
+      "step": 1126
+    },
+    {
+      "epoch": 1.7338461538461538,
+      "grad_norm": 0.2625023002565323,
+      "learning_rate": 8.711101014028855e-07,
+      "loss": 0.2728,
+      "step": 1127
+    },
+    {
+      "epoch": 1.7353846153846155,
+      "grad_norm": 0.26985914746536965,
+      "learning_rate": 8.612719754415078e-07,
+      "loss": 0.2773,
+      "step": 1128
+    },
+    {
+      "epoch": 1.7369230769230768,
+      "grad_norm": 0.27204356751705255,
+      "learning_rate": 8.514872196496182e-07,
+      "loss": 0.2718,
+      "step": 1129
+    },
+    {
+      "epoch": 1.7384615384615385,
+      "grad_norm": 0.2589644443476737,
+      "learning_rate": 8.417558911701884e-07,
+      "loss": 0.2674,
+      "step": 1130
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.2619031108337043,
+      "learning_rate": 8.320780468341761e-07,
+      "loss": 0.2769,
+      "step": 1131
+    },
+    {
+      "epoch": 1.7415384615384615,
+      "grad_norm": 0.24663096862283984,
+      "learning_rate": 8.224537431601886e-07,
+      "loss": 0.2749,
+      "step": 1132
+    },
+    {
+      "epoch": 1.7430769230769232,
+      "grad_norm": 0.25329093409597475,
+      "learning_rate": 8.128830363541574e-07,
+      "loss": 0.2674,
+      "step": 1133
+    },
+    {
+      "epoch": 1.7446153846153845,
+      "grad_norm": 0.25855666495593144,
+      "learning_rate": 8.03365982309009e-07,
+      "loss": 0.2732,
+      "step": 1134
+    },
+    {
+      "epoch": 1.7461538461538462,
+      "grad_norm": 0.2575342547274432,
+      "learning_rate": 7.939026366043323e-07,
+      "loss": 0.2734,
+      "step": 1135
+    },
+    {
+      "epoch": 1.7476923076923077,
+      "grad_norm": 0.2572051505754092,
+      "learning_rate": 7.844930545060703e-07,
+      "loss": 0.2672,
+      "step": 1136
+    },
+    {
+      "epoch": 1.7492307692307691,
+      "grad_norm": 0.25659868456171797,
+      "learning_rate": 7.75137290966177e-07,
+      "loss": 0.2694,
+      "step": 1137
+    },
+    {
+      "epoch": 1.7507692307692309,
+      "grad_norm": 0.26701664877195225,
+      "learning_rate": 7.65835400622309e-07,
+      "loss": 0.2706,
+      "step": 1138
+    },
+    {
+      "epoch": 1.7523076923076923,
+      "grad_norm": 0.28286665007577344,
+      "learning_rate": 7.565874377975046e-07,
+      "loss": 0.272,
+      "step": 1139
+    },
+    {
+      "epoch": 1.7538461538461538,
+      "grad_norm": 0.27858205811414216,
+      "learning_rate": 7.473934564998641e-07,
+      "loss": 0.267,
+      "step": 1140
+    },
+    {
+      "epoch": 1.7553846153846155,
+      "grad_norm": 0.26001654197560764,
+      "learning_rate": 7.382535104222366e-07,
+      "loss": 0.2741,
+      "step": 1141
+    },
+    {
+      "epoch": 1.7569230769230768,
+      "grad_norm": 0.2649493247350367,
+      "learning_rate": 7.291676529419034e-07,
+      "loss": 0.2735,
+      "step": 1142
+    },
+    {
+      "epoch": 1.7584615384615385,
+      "grad_norm": 0.2543629832367694,
+      "learning_rate": 7.201359371202698e-07,
+      "loss": 0.2692,
+      "step": 1143
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.25355471214350306,
+      "learning_rate": 7.111584157025575e-07,
+      "loss": 0.2768,
+      "step": 1144
+    },
+    {
+      "epoch": 1.7615384615384615,
+      "grad_norm": 0.24548207312048848,
+      "learning_rate": 7.022351411174866e-07,
+      "loss": 0.2763,
+      "step": 1145
+    },
+    {
+      "epoch": 1.7630769230769232,
+      "grad_norm": 0.24529106469337025,
+      "learning_rate": 6.933661654769797e-07,
+      "loss": 0.277,
+      "step": 1146
+    },
+    {
+      "epoch": 1.7646153846153845,
+      "grad_norm": 0.2575076262302863,
+      "learning_rate": 6.845515405758518e-07,
+      "loss": 0.2743,
+      "step": 1147
+    },
+    {
+      "epoch": 1.7661538461538462,
+      "grad_norm": 0.26604894439329496,
+      "learning_rate": 6.757913178915087e-07,
+      "loss": 0.2749,
+      "step": 1148
+    },
+    {
+      "epoch": 1.7676923076923077,
+      "grad_norm": 0.28681485404621826,
+      "learning_rate": 6.670855485836525e-07,
+      "loss": 0.2686,
+      "step": 1149
+    },
+    {
+      "epoch": 1.7692307692307692,
+      "grad_norm": 0.2589106809589985,
+      "learning_rate": 6.584342834939717e-07,
+      "loss": 0.2704,
+      "step": 1150
+    },
+    {
+      "epoch": 1.7707692307692309,
+      "grad_norm": 0.28911580411212706,
+      "learning_rate": 6.498375731458529e-07,
+      "loss": 0.2652,
+      "step": 1151
+    },
+    {
+      "epoch": 1.7723076923076924,
+      "grad_norm": 0.24524281594408773,
+      "learning_rate": 6.412954677440797e-07,
+      "loss": 0.2769,
+      "step": 1152
+    },
+    {
+      "epoch": 1.7738461538461539,
+      "grad_norm": 0.2413224345249869,
+      "learning_rate": 6.32808017174551e-07,
+      "loss": 0.2824,
+      "step": 1153
+    },
+    {
+      "epoch": 1.7753846153846153,
+      "grad_norm": 0.2488148241737766,
+      "learning_rate": 6.243752710039719e-07,
+      "loss": 0.2707,
+      "step": 1154
+    },
+    {
+      "epoch": 1.7769230769230768,
+      "grad_norm": 0.2510817209486956,
+      "learning_rate": 6.159972784795798e-07,
+      "loss": 0.2669,
+      "step": 1155
+    },
+    {
+      "epoch": 1.7784615384615385,
+      "grad_norm": 0.26078478335979977,
+      "learning_rate": 6.076740885288479e-07,
+      "loss": 0.2644,
+      "step": 1156
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.24929356810208964,
+      "learning_rate": 5.994057497592032e-07,
+      "loss": 0.2688,
+      "step": 1157
+    },
+    {
+      "epoch": 1.7815384615384615,
+      "grad_norm": 0.25711900466450466,
+      "learning_rate": 5.911923104577455e-07,
+      "loss": 0.2692,
+      "step": 1158
+    },
+    {
+      "epoch": 1.7830769230769232,
+      "grad_norm": 0.26857644105403744,
+      "learning_rate": 5.830338185909545e-07,
+      "loss": 0.27,
+      "step": 1159
+    },
+    {
+      "epoch": 1.7846153846153845,
+      "grad_norm": 0.26355737383923983,
+      "learning_rate": 5.749303218044234e-07,
+      "loss": 0.2685,
+      "step": 1160
+    },
+    {
+      "epoch": 1.7861538461538462,
+      "grad_norm": 0.25433788202515617,
+      "learning_rate": 5.668818674225684e-07,
+      "loss": 0.2705,
+      "step": 1161
+    },
+    {
+      "epoch": 1.7876923076923077,
+      "grad_norm": 0.2540172872801744,
+      "learning_rate": 5.588885024483648e-07,
+      "loss": 0.2683,
+      "step": 1162
+    },
+    {
+      "epoch": 1.7892307692307692,
+      "grad_norm": 0.2557887226091531,
+      "learning_rate": 5.509502735630601e-07,
+      "loss": 0.2759,
+      "step": 1163
+    },
+    {
+      "epoch": 1.790769230769231,
+      "grad_norm": 0.265619907163393,
+      "learning_rate": 5.430672271259096e-07,
+      "loss": 0.2648,
+      "step": 1164
+    },
+    {
+      "epoch": 1.7923076923076922,
+      "grad_norm": 0.2623517115628094,
+      "learning_rate": 5.352394091739022e-07,
+      "loss": 0.2647,
+      "step": 1165
+    },
+    {
+      "epoch": 1.7938461538461539,
+      "grad_norm": 0.27799113214084537,
+      "learning_rate": 5.274668654214931e-07,
+      "loss": 0.2671,
+      "step": 1166
+    },
+    {
+      "epoch": 1.7953846153846154,
+      "grad_norm": 0.25386965217553553,
+      "learning_rate": 5.197496412603365e-07,
+      "loss": 0.2813,
+      "step": 1167
+    },
+    {
+      "epoch": 1.7969230769230768,
+      "grad_norm": 0.26600539307745497,
+      "learning_rate": 5.120877817590197e-07,
+      "loss": 0.2749,
+      "step": 1168
+    },
+    {
+      "epoch": 1.7984615384615386,
+      "grad_norm": 0.2598010369561241,
+      "learning_rate": 5.044813316627994e-07,
+      "loss": 0.2672,
+      "step": 1169
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.28224438868540486,
+      "learning_rate": 4.969303353933408e-07,
+      "loss": 0.2745,
+      "step": 1170
+    },
+    {
+      "epoch": 1.8015384615384615,
+      "grad_norm": 0.2805661553631755,
+      "learning_rate": 4.894348370484648e-07,
+      "loss": 0.2744,
+      "step": 1171
+    },
+    {
+      "epoch": 1.803076923076923,
+      "grad_norm": 0.25794334710550715,
+      "learning_rate": 4.819948804018771e-07,
+      "loss": 0.2699,
+      "step": 1172
+    },
+    {
+      "epoch": 1.8046153846153845,
+      "grad_norm": 0.25722518471927297,
+      "learning_rate": 4.746105089029229e-07,
+      "loss": 0.2774,
+      "step": 1173
+    },
+    {
+      "epoch": 1.8061538461538462,
+      "grad_norm": 0.25812083560921784,
+      "learning_rate": 4.6728176567633065e-07,
+      "loss": 0.2727,
+      "step": 1174
+    },
+    {
+      "epoch": 1.8076923076923077,
+      "grad_norm": 0.25551672554775057,
+      "learning_rate": 4.6000869352195607e-07,
+      "loss": 0.2814,
+      "step": 1175
+    },
+    {
+      "epoch": 1.8092307692307692,
+      "grad_norm": 0.261306292310136,
+      "learning_rate": 4.5279133491454406e-07,
+      "loss": 0.2773,
+      "step": 1176
+    },
+    {
+      "epoch": 1.810769230769231,
+      "grad_norm": 0.2612795334066712,
+      "learning_rate": 4.4562973200346413e-07,
+      "loss": 0.2844,
+      "step": 1177
+    },
+    {
+      "epoch": 1.8123076923076922,
+      "grad_norm": 0.2557868514983067,
+      "learning_rate": 4.385239266124752e-07,
+      "loss": 0.2684,
+      "step": 1178
+    },
+    {
+      "epoch": 1.8138461538461539,
+      "grad_norm": 0.2621733323408126,
+      "learning_rate": 4.314739602394791e-07,
+      "loss": 0.2752,
+      "step": 1179
+    },
+    {
+      "epoch": 1.8153846153846154,
+      "grad_norm": 0.25941984414296076,
+      "learning_rate": 4.2447987405628054e-07,
+      "loss": 0.2789,
+      "step": 1180
+    },
+    {
+      "epoch": 1.8169230769230769,
+      "grad_norm": 0.25937657312299867,
+      "learning_rate": 4.1754170890833777e-07,
+      "loss": 0.2644,
+      "step": 1181
+    },
+    {
+      "epoch": 1.8184615384615386,
+      "grad_norm": 0.2826036343635729,
+      "learning_rate": 4.106595053145357e-07,
+      "loss": 0.2688,
+      "step": 1182
+    },
+    {
+      "epoch": 1.8199999999999998,
+      "grad_norm": 0.25844365974175953,
+      "learning_rate": 4.038333034669406e-07,
+      "loss": 0.2642,
+      "step": 1183
+    },
+    {
+      "epoch": 1.8215384615384616,
+      "grad_norm": 0.26940689160820397,
+      "learning_rate": 3.9706314323056936e-07,
+      "loss": 0.2734,
+      "step": 1184
+    },
+    {
+      "epoch": 1.823076923076923,
+      "grad_norm": 0.25793667787420654,
+      "learning_rate": 3.903490641431573e-07,
+      "loss": 0.282,
+      "step": 1185
+    },
+    {
+      "epoch": 1.8246153846153845,
+      "grad_norm": 0.26494408613001746,
+      "learning_rate": 3.8369110541492396e-07,
+      "loss": 0.2784,
+      "step": 1186
+    },
+    {
+      "epoch": 1.8261538461538462,
+      "grad_norm": 0.26227052155564934,
+      "learning_rate": 3.770893059283465e-07,
+      "loss": 0.2819,
+      "step": 1187
+    },
+    {
+      "epoch": 1.8276923076923077,
+      "grad_norm": 0.2602766507292804,
+      "learning_rate": 3.705437042379334e-07,
+      "loss": 0.2682,
+      "step": 1188
+    },
+    {
+      "epoch": 1.8292307692307692,
+      "grad_norm": 0.2687663797936269,
+      "learning_rate": 3.6405433856999684e-07,
+      "loss": 0.2766,
+      "step": 1189
+    },
+    {
+      "epoch": 1.830769230769231,
+      "grad_norm": 0.2638930778071268,
+      "learning_rate": 3.5762124682242936e-07,
+      "loss": 0.274,
+      "step": 1190
+    },
+    {
+      "epoch": 1.8323076923076922,
+      "grad_norm": 0.26708725438168995,
+      "learning_rate": 3.5124446656448654e-07,
+      "loss": 0.2657,
+      "step": 1191
+    },
+    {
+      "epoch": 1.833846153846154,
+      "grad_norm": 0.26563786184808197,
+      "learning_rate": 3.4492403503656236e-07,
+      "loss": 0.2736,
+      "step": 1192
+    },
+    {
+      "epoch": 1.8353846153846154,
+      "grad_norm": 0.25267333891215393,
+      "learning_rate": 3.3865998914997645e-07,
+      "loss": 0.2669,
+      "step": 1193
+    },
+    {
+      "epoch": 1.8369230769230769,
+      "grad_norm": 0.2670140474295601,
+      "learning_rate": 3.324523654867551e-07,
+      "loss": 0.2823,
+      "step": 1194
+    },
+    {
+      "epoch": 1.8384615384615386,
+      "grad_norm": 0.25918910373856735,
+      "learning_rate": 3.2630120029942034e-07,
+      "loss": 0.2728,
+      "step": 1195
+    },
+    {
+      "epoch": 1.8399999999999999,
+      "grad_norm": 0.2605937731790264,
+      "learning_rate": 3.2020652951077256e-07,
+      "loss": 0.2738,
+      "step": 1196
+    },
+    {
+      "epoch": 1.8415384615384616,
+      "grad_norm": 0.2653165480130905,
+      "learning_rate": 3.1416838871368925e-07,
+      "loss": 0.2682,
+      "step": 1197
+    },
+    {
+      "epoch": 1.843076923076923,
+      "grad_norm": 0.25866129288975076,
+      "learning_rate": 3.081868131709109e-07,
+      "loss": 0.2711,
+      "step": 1198
+    },
+    {
+      "epoch": 1.8446153846153845,
+      "grad_norm": 0.2659815203267659,
+      "learning_rate": 3.0226183781483897e-07,
+      "loss": 0.2716,
+      "step": 1199
+    },
+    {
+      "epoch": 1.8461538461538463,
+      "grad_norm": 0.26552289033198906,
+      "learning_rate": 2.963934972473259e-07,
+      "loss": 0.2804,
+      "step": 1200
+    },
+    {
+      "epoch": 1.8476923076923077,
+      "grad_norm": 0.24900482766301524,
+      "learning_rate": 2.905818257394799e-07,
+      "loss": 0.2768,
+      "step": 1201
+    },
+    {
+      "epoch": 1.8492307692307692,
+      "grad_norm": 0.24947347437293466,
+      "learning_rate": 2.848268572314616e-07,
+      "loss": 0.2783,
+      "step": 1202
+    },
+    {
+      "epoch": 1.8507692307692307,
+      "grad_norm": 0.25930581415360415,
+      "learning_rate": 2.791286253322856e-07,
+      "loss": 0.2614,
+      "step": 1203
+    },
+    {
+      "epoch": 1.8523076923076922,
+      "grad_norm": 0.26542775008735076,
+      "learning_rate": 2.734871633196246e-07,
+      "loss": 0.28,
+      "step": 1204
+    },
+    {
+      "epoch": 1.853846153846154,
+      "grad_norm": 0.27067867703064946,
+      "learning_rate": 2.679025041396155e-07,
+      "loss": 0.2663,
+      "step": 1205
+    },
+    {
+      "epoch": 1.8553846153846154,
+      "grad_norm": 0.25825385919738536,
+      "learning_rate": 2.6237468040666515e-07,
+      "loss": 0.2789,
+      "step": 1206
+    },
+    {
+      "epoch": 1.856923076923077,
+      "grad_norm": 0.250639299715114,
+      "learning_rate": 2.569037244032657e-07,
+      "loss": 0.2642,
+      "step": 1207
+    },
+    {
+      "epoch": 1.8584615384615386,
+      "grad_norm": 0.2627722533489089,
+      "learning_rate": 2.5148966807979733e-07,
+      "loss": 0.2725,
+      "step": 1208
+    },
+    {
+      "epoch": 1.8599999999999999,
+      "grad_norm": 0.25698102394911715,
+      "learning_rate": 2.461325430543482e-07,
+      "loss": 0.2624,
+      "step": 1209
+    },
+    {
+      "epoch": 1.8615384615384616,
+      "grad_norm": 0.27170513125540613,
+      "learning_rate": 2.4083238061252565e-07,
+      "loss": 0.266,
+      "step": 1210
+    },
+    {
+      "epoch": 1.863076923076923,
+      "grad_norm": 0.26016401860293503,
+      "learning_rate": 2.355892117072789e-07,
+      "loss": 0.2754,
+      "step": 1211
+    },
+    {
+      "epoch": 1.8646153846153846,
+      "grad_norm": 0.2573951151273573,
+      "learning_rate": 2.3040306695871319e-07,
+      "loss": 0.2771,
+      "step": 1212
+    },
+    {
+      "epoch": 1.8661538461538463,
+      "grad_norm": 0.2578731976600654,
+      "learning_rate": 2.2527397665391026e-07,
+      "loss": 0.2766,
+      "step": 1213
+    },
+    {
+      "epoch": 1.8676923076923075,
+      "grad_norm": 0.24145426645886497,
+      "learning_rate": 2.2020197074675952e-07,
+      "loss": 0.2555,
+      "step": 1214
+    },
+    {
+      "epoch": 1.8692307692307693,
+      "grad_norm": 0.2651433209823767,
+      "learning_rate": 2.1518707885777147e-07,
+      "loss": 0.278,
+      "step": 1215
+    },
+    {
+      "epoch": 1.8707692307692307,
+      "grad_norm": 0.24881430546225117,
+      "learning_rate": 2.1022933027391555e-07,
+      "loss": 0.2798,
+      "step": 1216
+    },
+    {
+      "epoch": 1.8723076923076922,
+      "grad_norm": 0.25467344376243073,
+      "learning_rate": 2.0532875394844053e-07,
+      "loss": 0.2683,
+      "step": 1217
+    },
+    {
+      "epoch": 1.873846153846154,
+      "grad_norm": 0.26531380563696,
+      "learning_rate": 2.0048537850071326e-07,
+      "loss": 0.2622,
+      "step": 1218
+    },
+    {
+      "epoch": 1.8753846153846154,
+      "grad_norm": 0.2754942481670186,
+      "learning_rate": 1.9569923221604224e-07,
+      "loss": 0.2701,
+      "step": 1219
+    },
+    {
+      "epoch": 1.876923076923077,
+      "grad_norm": 0.26202066271597907,
+      "learning_rate": 1.909703430455223e-07,
+      "loss": 0.2701,
+      "step": 1220
+    },
+    {
+      "epoch": 1.8784615384615386,
+      "grad_norm": 0.26462179129135577,
+      "learning_rate": 1.8629873860586567e-07,
+      "loss": 0.2619,
+      "step": 1221
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.2599626125269344,
+      "learning_rate": 1.8168444617924107e-07,
+      "loss": 0.2714,
+      "step": 1222
+    },
+    {
+      "epoch": 1.8815384615384616,
+      "grad_norm": 0.2717438370705554,
+      "learning_rate": 1.7712749271311392e-07,
+      "loss": 0.2695,
+      "step": 1223
+    },
+    {
+      "epoch": 1.883076923076923,
+      "grad_norm": 0.2594326713284917,
+      "learning_rate": 1.7262790482009184e-07,
+      "loss": 0.2717,
+      "step": 1224
+    },
+    {
+      "epoch": 1.8846153846153846,
+      "grad_norm": 0.2511033817455497,
+      "learning_rate": 1.681857087777672e-07,
+      "loss": 0.2678,
+      "step": 1225
+    },
+    {
+      "epoch": 1.8861538461538463,
+      "grad_norm": 0.26691991336692045,
+      "learning_rate": 1.6380093052856482e-07,
+      "loss": 0.2691,
+      "step": 1226
+    },
+    {
+      "epoch": 1.8876923076923076,
+      "grad_norm": 0.2631060513470928,
+      "learning_rate": 1.5947359567958677e-07,
+      "loss": 0.2752,
+      "step": 1227
+    },
+    {
+      "epoch": 1.8892307692307693,
+      "grad_norm": 0.2551511842652523,
+      "learning_rate": 1.5520372950246888e-07,
+      "loss": 0.2703,
+      "step": 1228
+    },
+    {
+      "epoch": 1.8907692307692308,
+      "grad_norm": 0.2649291705796171,
+      "learning_rate": 1.5099135693322776e-07,
+      "loss": 0.2737,
+      "step": 1229
+    },
+    {
+      "epoch": 1.8923076923076922,
+      "grad_norm": 0.2641129353469266,
+      "learning_rate": 1.4683650257211967e-07,
+      "loss": 0.2696,
+      "step": 1230
+    },
+    {
+      "epoch": 1.893846153846154,
+      "grad_norm": 0.25763156609824617,
+      "learning_rate": 1.4273919068349184e-07,
+      "loss": 0.2681,
+      "step": 1231
+    },
+    {
+      "epoch": 1.8953846153846152,
+      "grad_norm": 0.2544835090162467,
+      "learning_rate": 1.386994451956436e-07,
+      "loss": 0.267,
+      "step": 1232
+    },
+    {
+      "epoch": 1.896923076923077,
+      "grad_norm": 0.2658293069336348,
+      "learning_rate": 1.3471728970068986e-07,
+      "loss": 0.2708,
+      "step": 1233
+    },
+    {
+      "epoch": 1.8984615384615384,
+      "grad_norm": 0.2667104526648449,
+      "learning_rate": 1.3079274745441794e-07,
+      "loss": 0.265,
+      "step": 1234
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.24963969373244324,
+      "learning_rate": 1.2692584137615205e-07,
+      "loss": 0.2709,
+      "step": 1235
+    },
+    {
+      "epoch": 1.9015384615384616,
+      "grad_norm": 0.26247731750631575,
+      "learning_rate": 1.231165940486234e-07,
+      "loss": 0.2701,
+      "step": 1236
+    },
+    {
+      "epoch": 1.903076923076923,
+      "grad_norm": 0.2725078089641997,
+      "learning_rate": 1.1936502771783488e-07,
+      "loss": 0.2728,
+      "step": 1237
+    },
+    {
+      "epoch": 1.9046153846153846,
+      "grad_norm": 0.25986490859938516,
+      "learning_rate": 1.1567116429293424e-07,
+      "loss": 0.2765,
+      "step": 1238
+    },
+    {
+      "epoch": 1.9061538461538463,
+      "grad_norm": 0.2730232582178096,
+      "learning_rate": 1.1203502534608113e-07,
+      "loss": 0.2663,
+      "step": 1239
+    },
+    {
+      "epoch": 1.9076923076923076,
+      "grad_norm": 0.25156013265415067,
+      "learning_rate": 1.0845663211232704e-07,
+      "loss": 0.2707,
+      "step": 1240
+    },
+    {
+      "epoch": 1.9092307692307693,
+      "grad_norm": 0.2590488093159212,
+      "learning_rate": 1.0493600548948879e-07,
+      "loss": 0.2811,
+      "step": 1241
+    },
+    {
+      "epoch": 1.9107692307692308,
+      "grad_norm": 0.24711167458443112,
+      "learning_rate": 1.0147316603802415e-07,
+      "loss": 0.27,
+      "step": 1242
+    },
+    {
+      "epoch": 1.9123076923076923,
+      "grad_norm": 0.25768794048943483,
+      "learning_rate": 9.806813398091419e-08,
+      "loss": 0.2687,
+      "step": 1243
+    },
+    {
+      "epoch": 1.913846153846154,
+      "grad_norm": 0.25199192932976877,
+      "learning_rate": 9.47209292035467e-08,
+      "loss": 0.2748,
+      "step": 1244
+    },
+    {
+      "epoch": 1.9153846153846152,
+      "grad_norm": 0.2647710191663874,
+      "learning_rate": 9.143157125359514e-08,
+      "loss": 0.2727,
+      "step": 1245
+    },
+    {
+      "epoch": 1.916923076923077,
+      "grad_norm": 0.24735239434412856,
+      "learning_rate": 8.82000793409088e-08,
+      "loss": 0.2768,
+      "step": 1246
+    },
+    {
+      "epoch": 1.9184615384615384,
+      "grad_norm": 0.2582071850232192,
+      "learning_rate": 8.502647233740169e-08,
+      "loss": 0.2758,
+      "step": 1247
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.26608807486798247,
+      "learning_rate": 8.191076877693605e-08,
+      "loss": 0.2726,
+      "step": 1248
+    },
+    {
+      "epoch": 1.9215384615384616,
+      "grad_norm": 0.2674053658937129,
+      "learning_rate": 7.885298685522235e-08,
+      "loss": 0.2719,
+      "step": 1249
+    },
+    {
+      "epoch": 1.9230769230769231,
+      "grad_norm": 0.28414767606911856,
+      "learning_rate": 7.585314442970282e-08,
+      "loss": 0.2706,
+      "step": 1250
+    },
+    {
+      "epoch": 1.9246153846153846,
+      "grad_norm": 0.26825434762357314,
+      "learning_rate": 7.291125901946027e-08,
+      "loss": 0.2694,
+      "step": 1251
+    },
+    {
+      "epoch": 1.926153846153846,
+      "grad_norm": 0.2618931427864471,
+      "learning_rate": 7.002734780510279e-08,
+      "loss": 0.2708,
+      "step": 1252
+    },
+    {
+      "epoch": 1.9276923076923076,
+      "grad_norm": 0.26165575374832156,
+      "learning_rate": 6.720142762867032e-08,
+      "loss": 0.2729,
+      "step": 1253
+    },
+    {
+      "epoch": 1.9292307692307693,
+      "grad_norm": 0.2618441624570343,
+      "learning_rate": 6.443351499353823e-08,
+      "loss": 0.2719,
+      "step": 1254
+    },
+    {
+      "epoch": 1.9307692307692308,
+      "grad_norm": 0.2587712650866823,
+      "learning_rate": 6.172362606431281e-08,
+      "loss": 0.2724,
+      "step": 1255
+    },
+    {
+      "epoch": 1.9323076923076923,
+      "grad_norm": 0.2486902402553849,
+      "learning_rate": 5.907177666674813e-08,
+      "loss": 0.2844,
+      "step": 1256
+    },
+    {
+      "epoch": 1.933846153846154,
+      "grad_norm": 0.26350210307540733,
+      "learning_rate": 5.647798228764156e-08,
+      "loss": 0.2701,
+      "step": 1257
+    },
+    {
+      "epoch": 1.9353846153846153,
+      "grad_norm": 0.2635749257649734,
+      "learning_rate": 5.394225807475284e-08,
+      "loss": 0.2663,
+      "step": 1258
+    },
+    {
+      "epoch": 1.936923076923077,
+      "grad_norm": 0.25076562478883774,
+      "learning_rate": 5.146461883671072e-08,
+      "loss": 0.2637,
+      "step": 1259
+    },
+    {
+      "epoch": 1.9384615384615385,
+      "grad_norm": 0.25547826529213463,
+      "learning_rate": 4.9045079042926434e-08,
+      "loss": 0.2787,
+      "step": 1260
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.2367090690776121,
+      "learning_rate": 4.6683652823513725e-08,
+      "loss": 0.2819,
+      "step": 1261
+    },
+    {
+      "epoch": 1.9415384615384617,
+      "grad_norm": 0.2512139791982206,
+      "learning_rate": 4.438035396920004e-08,
+      "loss": 0.2746,
+      "step": 1262
+    },
+    {
+      "epoch": 1.943076923076923,
+      "grad_norm": 0.2576870779313507,
+      "learning_rate": 4.2135195931249925e-08,
+      "loss": 0.2683,
+      "step": 1263
+    },
+    {
+      "epoch": 1.9446153846153846,
+      "grad_norm": 0.2742765354640819,
+      "learning_rate": 3.9948191821386205e-08,
+      "loss": 0.2651,
+      "step": 1264
+    },
+    {
+      "epoch": 1.9461538461538461,
+      "grad_norm": 0.2705148767586398,
+      "learning_rate": 3.7819354411713364e-08,
+      "loss": 0.2734,
+      "step": 1265
+    },
+    {
+      "epoch": 1.9476923076923076,
+      "grad_norm": 0.24567267475694976,
+      "learning_rate": 3.5748696134639825e-08,
+      "loss": 0.2665,
+      "step": 1266
+    },
+    {
+      "epoch": 1.9492307692307693,
+      "grad_norm": 0.24414183004306245,
+      "learning_rate": 3.373622908280916e-08,
+      "loss": 0.2681,
+      "step": 1267
+    },
+    {
+      "epoch": 1.9507692307692308,
+      "grad_norm": 0.2569335080497009,
+      "learning_rate": 3.178196500903008e-08,
+      "loss": 0.2719,
+      "step": 1268
+    },
+    {
+      "epoch": 1.9523076923076923,
+      "grad_norm": 0.27125049629442594,
+      "learning_rate": 2.988591532620322e-08,
+      "loss": 0.263,
+      "step": 1269
+    },
+    {
+      "epoch": 1.953846153846154,
+      "grad_norm": 0.2540792040831073,
+      "learning_rate": 2.8048091107258925e-08,
+      "loss": 0.2775,
+      "step": 1270
+    },
+    {
+      "epoch": 1.9553846153846153,
+      "grad_norm": 0.26516722010683846,
+      "learning_rate": 2.6268503085089547e-08,
+      "loss": 0.2749,
+      "step": 1271
+    },
+    {
+      "epoch": 1.956923076923077,
+      "grad_norm": 0.2555234354524852,
+      "learning_rate": 2.4547161652488382e-08,
+      "loss": 0.2781,
+      "step": 1272
+    },
+    {
+      "epoch": 1.9584615384615385,
+      "grad_norm": 0.25473215083860884,
+      "learning_rate": 2.2884076862089712e-08,
+      "loss": 0.2769,
+      "step": 1273
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.2579144478345488,
+      "learning_rate": 2.1279258426308848e-08,
+      "loss": 0.2618,
+      "step": 1274
+    },
+    {
+      "epoch": 1.9615384615384617,
+      "grad_norm": 0.26831411502080194,
+      "learning_rate": 1.973271571728441e-08,
+      "loss": 0.2683,
+      "step": 1275
+    },
+    {
+      "epoch": 1.963076923076923,
+      "grad_norm": 0.2699221453086357,
+      "learning_rate": 1.824445776682504e-08,
+      "loss": 0.2638,
+      "step": 1276
+    },
+    {
+      "epoch": 1.9646153846153847,
+      "grad_norm": 0.26636229804375944,
+      "learning_rate": 1.6814493266357202e-08,
+      "loss": 0.2719,
+      "step": 1277
+    },
+    {
+      "epoch": 1.9661538461538461,
+      "grad_norm": 0.25754559387076065,
+      "learning_rate": 1.5442830566874123e-08,
+      "loss": 0.2723,
+      "step": 1278
+    },
+    {
+      "epoch": 1.9676923076923076,
+      "grad_norm": 0.27746360300811806,
+      "learning_rate": 1.4129477678884728e-08,
+      "loss": 0.2748,
+      "step": 1279
+    },
+    {
+      "epoch": 1.9692307692307693,
+      "grad_norm": 0.26116574851475305,
+      "learning_rate": 1.2874442272369225e-08,
+      "loss": 0.2758,
+      "step": 1280
+    },
+    {
+      "epoch": 1.9707692307692306,
+      "grad_norm": 0.2738399417814919,
+      "learning_rate": 1.1677731676733584e-08,
+      "loss": 0.2693,
+      "step": 1281
+    },
+    {
+      "epoch": 1.9723076923076923,
+      "grad_norm": 0.24827608250119376,
+      "learning_rate": 1.053935288076957e-08,
+      "loss": 0.2763,
+      "step": 1282
+    },
+    {
+      "epoch": 1.9738461538461538,
+      "grad_norm": 0.264826230750588,
+      "learning_rate": 9.459312532608122e-09,
+      "loss": 0.2686,
+      "step": 1283
+    },
+    {
+      "epoch": 1.9753846153846153,
+      "grad_norm": 0.25756693518005147,
+      "learning_rate": 8.437616939683812e-09,
+      "loss": 0.2705,
+      "step": 1284
+    },
+    {
+      "epoch": 1.976923076923077,
+      "grad_norm": 0.26183486780931053,
+      "learning_rate": 7.474272068698219e-09,
+      "loss": 0.2765,
+      "step": 1285
+    },
+    {
+      "epoch": 1.9784615384615385,
+      "grad_norm": 0.25258521345991264,
+      "learning_rate": 6.569283545587724e-09,
+      "loss": 0.2699,
+      "step": 1286
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.2580148577432403,
+      "learning_rate": 5.722656655482439e-09,
+      "loss": 0.2581,
+      "step": 1287
+    },
+    {
+      "epoch": 1.9815384615384617,
+      "grad_norm": 0.2556746838024097,
+      "learning_rate": 4.9343963426840006e-09,
+      "loss": 0.2744,
+      "step": 1288
+    },
+    {
+      "epoch": 1.983076923076923,
+      "grad_norm": 0.25211453817832763,
+      "learning_rate": 4.204507210633368e-09,
+      "loss": 0.2794,
+      "step": 1289
+    },
+    {
+      "epoch": 1.9846153846153847,
+      "grad_norm": 0.2616628377298744,
+      "learning_rate": 3.5329935218819668e-09,
+      "loss": 0.2795,
+      "step": 1290
+    },
+    {
+      "epoch": 1.9861538461538462,
+      "grad_norm": 0.2569984403141055,
+      "learning_rate": 2.9198591980705847e-09,
+      "loss": 0.2806,
+      "step": 1291
+    },
+    {
+      "epoch": 1.9876923076923076,
+      "grad_norm": 0.2572173040806931,
+      "learning_rate": 2.3651078199016244e-09,
+      "loss": 0.2704,
+      "step": 1292
+    },
+    {
+      "epoch": 1.9892307692307694,
+      "grad_norm": 0.26173620976106304,
+      "learning_rate": 1.8687426271246646e-09,
+      "loss": 0.2719,
+      "step": 1293
+    },
+    {
+      "epoch": 1.9907692307692306,
+      "grad_norm": 0.24970039466414837,
+      "learning_rate": 1.430766518512039e-09,
+      "loss": 0.2694,
+      "step": 1294
+    },
+    {
+      "epoch": 1.9923076923076923,
+      "grad_norm": 0.28147304798836176,
+      "learning_rate": 1.0511820518432915e-09,
+      "loss": 0.2629,
+      "step": 1295
+    },
+    {
+      "epoch": 1.9938461538461538,
+      "grad_norm": 0.26120052432789,
+      "learning_rate": 7.299914438929634e-10,
+      "loss": 0.2619,
+      "step": 1296
+    },
+    {
+      "epoch": 1.9953846153846153,
+      "grad_norm": 0.26837836790650904,
+      "learning_rate": 4.671965704128312e-10,
+      "loss": 0.2773,
+      "step": 1297
+    },
+    {
+      "epoch": 1.996923076923077,
+      "grad_norm": 0.2474473453712718,
+      "learning_rate": 2.6279896612524393e-10,
+      "loss": 0.2753,
+      "step": 1298
+    },
+    {
+      "epoch": 1.9984615384615385,
+      "grad_norm": 0.2664258429360584,
+      "learning_rate": 1.167998247131319e-10,
+      "loss": 0.2671,
+      "step": 1299
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.27100229636351086,
+      "learning_rate": 2.919999881001445e-11,
+      "loss": 0.2669,
+      "step": 1300
+    },
+    {
+      "epoch": 2.0,
+      "step": 1300,
+      "total_flos": 1034035157860352.0,
+      "train_loss": 0.31700336213295277,
+      "train_runtime": 43306.5433,
+      "train_samples_per_second": 3.842,
+      "train_steps_per_second": 0.03
+    }
+  ],
+  "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": 1034035157860352.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}