diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,35869 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 26.393548387096775,
+  "eval_steps": 290,
+  "global_step": 5100,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.005161290322580645,
+      "grad_norm": 10.646817207336426,
+      "learning_rate": 4e-05,
+      "loss": 2.2469,
+      "step": 1
+    },
+    {
+      "epoch": 0.01032258064516129,
+      "grad_norm": 3.2345433235168457,
+      "learning_rate": 8e-05,
+      "loss": 2.408,
+      "step": 2
+    },
+    {
+      "epoch": 0.015483870967741935,
+      "grad_norm": 5.629693031311035,
+      "learning_rate": 0.00012,
+      "loss": 2.3563,
+      "step": 3
+    },
+    {
+      "epoch": 0.02064516129032258,
+      "grad_norm": 3.63735294342041,
+      "learning_rate": 0.00016,
+      "loss": 2.0797,
+      "step": 4
+    },
+    {
+      "epoch": 0.025806451612903226,
+      "grad_norm": 2.210094451904297,
+      "learning_rate": 0.0002,
+      "loss": 1.9625,
+      "step": 5
+    },
+    {
+      "epoch": 0.03096774193548387,
+      "grad_norm": 10.694299697875977,
+      "learning_rate": 0.00019979166666666667,
+      "loss": 1.7508,
+      "step": 6
+    },
+    {
+      "epoch": 0.03612903225806452,
+      "grad_norm": 0.8651490211486816,
+      "learning_rate": 0.00019958333333333335,
+      "loss": 1.774,
+      "step": 7
+    },
+    {
+      "epoch": 0.04129032258064516,
+      "grad_norm": 0.8025176525115967,
+      "learning_rate": 0.000199375,
+      "loss": 1.6877,
+      "step": 8
+    },
+    {
+      "epoch": 0.04645161290322581,
+      "grad_norm": 2.0353169441223145,
+      "learning_rate": 0.0001991666666666667,
+      "loss": 1.6632,
+      "step": 9
+    },
+    {
+      "epoch": 0.05161290322580645,
+      "grad_norm": 0.7836020588874817,
+      "learning_rate": 0.00019895833333333332,
+      "loss": 1.6797,
+      "step": 10
+    },
+    {
+      "epoch": 0.0567741935483871,
+      "grad_norm": 0.7764881253242493,
+      "learning_rate": 0.00019875,
+      "loss": 1.6869,
+      "step": 11
+    },
+    {
+      "epoch": 0.06193548387096774,
+      "grad_norm": 1.959869384765625,
+      "learning_rate": 0.00019854166666666666,
+      "loss": 1.6505,
+      "step": 12
+    },
+    {
+      "epoch": 0.06709677419354838,
+      "grad_norm": 0.852027952671051,
+      "learning_rate": 0.00019833333333333335,
+      "loss": 1.6088,
+      "step": 13
+    },
+    {
+      "epoch": 0.07225806451612904,
+      "grad_norm": 0.939630389213562,
+      "learning_rate": 0.000198125,
+      "loss": 1.6021,
+      "step": 14
+    },
+    {
+      "epoch": 0.07741935483870968,
+      "grad_norm": 1.0366936922073364,
+      "learning_rate": 0.0001979166666666667,
+      "loss": 1.6664,
+      "step": 15
+    },
+    {
+      "epoch": 0.08258064516129032,
+      "grad_norm": 3.928151845932007,
+      "learning_rate": 0.00019770833333333335,
+      "loss": 1.6239,
+      "step": 16
+    },
+    {
+      "epoch": 0.08774193548387096,
+      "grad_norm": 1.4234848022460938,
+      "learning_rate": 0.00019750000000000003,
+      "loss": 1.5527,
+      "step": 17
+    },
+    {
+      "epoch": 0.09290322580645162,
+      "grad_norm": 1.2283973693847656,
+      "learning_rate": 0.0001972916666666667,
+      "loss": 1.6029,
+      "step": 18
+    },
+    {
+      "epoch": 0.09806451612903226,
+      "grad_norm": 1.3136969804763794,
+      "learning_rate": 0.00019708333333333334,
+      "loss": 1.5547,
+      "step": 19
+    },
+    {
+      "epoch": 0.1032258064516129,
+      "grad_norm": 0.7701082229614258,
+      "learning_rate": 0.000196875,
+      "loss": 1.5702,
+      "step": 20
+    },
+    {
+      "epoch": 0.10838709677419354,
+      "grad_norm": 0.5927395224571228,
+      "learning_rate": 0.00019666666666666666,
+      "loss": 1.5292,
+      "step": 21
+    },
+    {
+      "epoch": 0.1135483870967742,
+      "grad_norm": 0.5584335923194885,
+      "learning_rate": 0.00019645833333333334,
+      "loss": 1.5661,
+      "step": 22
+    },
+    {
+      "epoch": 0.11870967741935484,
+      "grad_norm": 0.8166844248771667,
+      "learning_rate": 0.00019625,
+      "loss": 1.5316,
+      "step": 23
+    },
+    {
+      "epoch": 0.12387096774193548,
+      "grad_norm": 0.4093015491962433,
+      "learning_rate": 0.00019604166666666668,
+      "loss": 1.5593,
+      "step": 24
+    },
+    {
+      "epoch": 0.12903225806451613,
+      "grad_norm": 0.3005596697330475,
+      "learning_rate": 0.00019583333333333334,
+      "loss": 1.5016,
+      "step": 25
+    },
+    {
+      "epoch": 0.13419354838709677,
+      "grad_norm": 0.36477768421173096,
+      "learning_rate": 0.00019562500000000003,
+      "loss": 1.4857,
+      "step": 26
+    },
+    {
+      "epoch": 0.1393548387096774,
+      "grad_norm": 0.4108060896396637,
+      "learning_rate": 0.00019541666666666668,
+      "loss": 1.563,
+      "step": 27
+    },
+    {
+      "epoch": 0.14451612903225808,
+      "grad_norm": 0.3080444037914276,
+      "learning_rate": 0.00019520833333333334,
+      "loss": 1.4811,
+      "step": 28
+    },
+    {
+      "epoch": 0.14967741935483872,
+      "grad_norm": 0.4420616626739502,
+      "learning_rate": 0.000195,
+      "loss": 1.5123,
+      "step": 29
+    },
+    {
+      "epoch": 0.15483870967741936,
+      "grad_norm": 0.7872723340988159,
+      "learning_rate": 0.00019479166666666668,
+      "loss": 1.514,
+      "step": 30
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.32818853855133057,
+      "learning_rate": 0.00019458333333333334,
+      "loss": 1.4698,
+      "step": 31
+    },
+    {
+      "epoch": 0.16516129032258065,
+      "grad_norm": 0.6023102402687073,
+      "learning_rate": 0.00019437500000000002,
+      "loss": 1.5917,
+      "step": 32
+    },
+    {
+      "epoch": 0.1703225806451613,
+      "grad_norm": 0.3983575999736786,
+      "learning_rate": 0.00019416666666666668,
+      "loss": 1.5054,
+      "step": 33
+    },
+    {
+      "epoch": 0.17548387096774193,
+      "grad_norm": 0.46858328580856323,
+      "learning_rate": 0.00019395833333333336,
+      "loss": 1.5075,
+      "step": 34
+    },
+    {
+      "epoch": 0.18064516129032257,
+      "grad_norm": 0.40511757135391235,
+      "learning_rate": 0.00019375000000000002,
+      "loss": 1.5019,
+      "step": 35
+    },
+    {
+      "epoch": 0.18580645161290324,
+      "grad_norm": 0.34969276189804077,
+      "learning_rate": 0.00019354166666666668,
+      "loss": 1.4614,
+      "step": 36
+    },
+    {
+      "epoch": 0.19096774193548388,
+      "grad_norm": 0.47907140851020813,
+      "learning_rate": 0.00019333333333333333,
+      "loss": 1.5661,
+      "step": 37
+    },
+    {
+      "epoch": 0.19612903225806452,
+      "grad_norm": 0.3557140529155731,
+      "learning_rate": 0.000193125,
+      "loss": 1.3981,
+      "step": 38
+    },
+    {
+      "epoch": 0.20129032258064516,
+      "grad_norm": 0.3491615355014801,
+      "learning_rate": 0.00019291666666666668,
+      "loss": 1.5651,
+      "step": 39
+    },
+    {
+      "epoch": 0.2064516129032258,
+      "grad_norm": 0.3470868766307831,
+      "learning_rate": 0.00019270833333333333,
+      "loss": 1.5024,
+      "step": 40
+    },
+    {
+      "epoch": 0.21161290322580645,
+      "grad_norm": 0.3585146367549896,
+      "learning_rate": 0.00019250000000000002,
+      "loss": 1.4091,
+      "step": 41
+    },
+    {
+      "epoch": 0.2167741935483871,
+      "grad_norm": 0.3007749915122986,
+      "learning_rate": 0.00019229166666666667,
+      "loss": 1.51,
+      "step": 42
+    },
+    {
+      "epoch": 0.22193548387096773,
+      "grad_norm": 0.3476178050041199,
+      "learning_rate": 0.00019208333333333336,
+      "loss": 1.5569,
+      "step": 43
+    },
+    {
+      "epoch": 0.2270967741935484,
+      "grad_norm": 0.30271115899086,
+      "learning_rate": 0.00019187500000000002,
+      "loss": 1.4381,
+      "step": 44
+    },
+    {
+      "epoch": 0.23225806451612904,
+      "grad_norm": 0.298195481300354,
+      "learning_rate": 0.00019166666666666667,
+      "loss": 1.5345,
+      "step": 45
+    },
+    {
+      "epoch": 0.23741935483870968,
+      "grad_norm": 1.3948817253112793,
+      "learning_rate": 0.00019145833333333333,
+      "loss": 1.4832,
+      "step": 46
+    },
+    {
+      "epoch": 0.24258064516129033,
+      "grad_norm": 0.28374552726745605,
+      "learning_rate": 0.00019125000000000001,
+      "loss": 1.5175,
+      "step": 47
+    },
+    {
+      "epoch": 0.24774193548387097,
+      "grad_norm": 0.27413395047187805,
+      "learning_rate": 0.00019104166666666667,
+      "loss": 1.5152,
+      "step": 48
+    },
+    {
+      "epoch": 0.25290322580645164,
+      "grad_norm": 0.3088136613368988,
+      "learning_rate": 0.00019083333333333336,
+      "loss": 1.4983,
+      "step": 49
+    },
+    {
+      "epoch": 0.25806451612903225,
+      "grad_norm": 0.38825154304504395,
+      "learning_rate": 0.000190625,
+      "loss": 1.431,
+      "step": 50
+    },
+    {
+      "epoch": 0.2632258064516129,
+      "grad_norm": 0.34019070863723755,
+      "learning_rate": 0.00019041666666666667,
+      "loss": 1.5436,
+      "step": 51
+    },
+    {
+      "epoch": 0.26838709677419353,
+      "grad_norm": 0.3067444860935211,
+      "learning_rate": 0.00019020833333333335,
+      "loss": 1.5228,
+      "step": 52
+    },
+    {
+      "epoch": 0.2735483870967742,
+      "grad_norm": 0.4235207140445709,
+      "learning_rate": 0.00019,
+      "loss": 1.5093,
+      "step": 53
+    },
+    {
+      "epoch": 0.2787096774193548,
+      "grad_norm": 0.32013460993766785,
+      "learning_rate": 0.00018979166666666667,
+      "loss": 1.5148,
+      "step": 54
+    },
+    {
+      "epoch": 0.2838709677419355,
+      "grad_norm": 0.304765909910202,
+      "learning_rate": 0.00018958333333333332,
+      "loss": 1.4818,
+      "step": 55
+    },
+    {
+      "epoch": 0.28903225806451616,
+      "grad_norm": 0.3399871587753296,
+      "learning_rate": 0.000189375,
+      "loss": 1.4296,
+      "step": 56
+    },
+    {
+      "epoch": 0.29419354838709677,
+      "grad_norm": 0.29919394850730896,
+      "learning_rate": 0.00018916666666666667,
+      "loss": 1.3722,
+      "step": 57
+    },
+    {
+      "epoch": 0.29935483870967744,
+      "grad_norm": 1.1093330383300781,
+      "learning_rate": 0.00018895833333333335,
+      "loss": 1.5098,
+      "step": 58
+    },
+    {
+      "epoch": 0.30451612903225805,
+      "grad_norm": 0.26281917095184326,
+      "learning_rate": 0.00018875,
+      "loss": 1.5859,
+      "step": 59
+    },
+    {
+      "epoch": 0.3096774193548387,
+      "grad_norm": 0.2736188471317291,
+      "learning_rate": 0.0001885416666666667,
+      "loss": 1.4795,
+      "step": 60
+    },
+    {
+      "epoch": 0.31483870967741934,
+      "grad_norm": 0.3001708388328552,
+      "learning_rate": 0.00018833333333333335,
+      "loss": 1.4558,
+      "step": 61
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.36353799700737,
+      "learning_rate": 0.000188125,
+      "loss": 1.5026,
+      "step": 62
+    },
+    {
+      "epoch": 0.3251612903225806,
+      "grad_norm": 0.47671204805374146,
+      "learning_rate": 0.00018791666666666666,
+      "loss": 1.4526,
+      "step": 63
+    },
+    {
+      "epoch": 0.3303225806451613,
+      "grad_norm": 0.25515216588974,
+      "learning_rate": 0.00018770833333333335,
+      "loss": 1.4699,
+      "step": 64
+    },
+    {
+      "epoch": 0.33548387096774196,
+      "grad_norm": 0.30145615339279175,
+      "learning_rate": 0.0001875,
+      "loss": 1.5093,
+      "step": 65
+    },
+    {
+      "epoch": 0.3406451612903226,
+      "grad_norm": 0.46390941739082336,
+      "learning_rate": 0.00018729166666666666,
+      "loss": 1.5151,
+      "step": 66
+    },
+    {
+      "epoch": 0.34580645161290324,
+      "grad_norm": 0.3254154324531555,
+      "learning_rate": 0.00018708333333333335,
+      "loss": 1.5193,
+      "step": 67
+    },
+    {
+      "epoch": 0.35096774193548386,
+      "grad_norm": 0.2633412480354309,
+      "learning_rate": 0.000186875,
+      "loss": 1.4702,
+      "step": 68
+    },
+    {
+      "epoch": 0.3561290322580645,
+      "grad_norm": 0.30206188559532166,
+      "learning_rate": 0.0001866666666666667,
+      "loss": 1.4848,
+      "step": 69
+    },
+    {
+      "epoch": 0.36129032258064514,
+      "grad_norm": 0.32650482654571533,
+      "learning_rate": 0.00018645833333333334,
+      "loss": 1.4522,
+      "step": 70
+    },
+    {
+      "epoch": 0.3664516129032258,
+      "grad_norm": 0.8271152973175049,
+      "learning_rate": 0.00018625,
+      "loss": 1.4305,
+      "step": 71
+    },
+    {
+      "epoch": 0.3716129032258065,
+      "grad_norm": 0.40941306948661804,
+      "learning_rate": 0.00018604166666666666,
+      "loss": 1.4455,
+      "step": 72
+    },
+    {
+      "epoch": 0.3767741935483871,
+      "grad_norm": 0.28478285670280457,
+      "learning_rate": 0.00018583333333333334,
+      "loss": 1.4806,
+      "step": 73
+    },
+    {
+      "epoch": 0.38193548387096776,
+      "grad_norm": 0.2684497833251953,
+      "learning_rate": 0.000185625,
+      "loss": 1.506,
+      "step": 74
+    },
+    {
+      "epoch": 0.3870967741935484,
+      "grad_norm": 0.32135680317878723,
+      "learning_rate": 0.00018541666666666668,
+      "loss": 1.4822,
+      "step": 75
+    },
+    {
+      "epoch": 0.39225806451612905,
+      "grad_norm": 0.4313993453979492,
+      "learning_rate": 0.00018520833333333334,
+      "loss": 1.3933,
+      "step": 76
+    },
+    {
+      "epoch": 0.39741935483870966,
+      "grad_norm": 0.3326885402202606,
+      "learning_rate": 0.00018500000000000002,
+      "loss": 1.5499,
+      "step": 77
+    },
+    {
+      "epoch": 0.40258064516129033,
+      "grad_norm": 0.3104792833328247,
+      "learning_rate": 0.00018479166666666668,
+      "loss": 1.4579,
+      "step": 78
+    },
+    {
+      "epoch": 0.40774193548387094,
+      "grad_norm": 0.2651408910751343,
+      "learning_rate": 0.00018458333333333337,
+      "loss": 1.4354,
+      "step": 79
+    },
+    {
+      "epoch": 0.4129032258064516,
+      "grad_norm": 0.2535545825958252,
+      "learning_rate": 0.000184375,
+      "loss": 1.4148,
+      "step": 80
+    },
+    {
+      "epoch": 0.4180645161290323,
+      "grad_norm": 0.25613802671432495,
+      "learning_rate": 0.00018416666666666665,
+      "loss": 1.4814,
+      "step": 81
+    },
+    {
+      "epoch": 0.4232258064516129,
+      "grad_norm": 0.3124407231807709,
+      "learning_rate": 0.00018395833333333334,
+      "loss": 1.4618,
+      "step": 82
+    },
+    {
+      "epoch": 0.42838709677419357,
+      "grad_norm": 0.28805267810821533,
+      "learning_rate": 0.00018375,
+      "loss": 1.4845,
+      "step": 83
+    },
+    {
+      "epoch": 0.4335483870967742,
+      "grad_norm": 0.39042362570762634,
+      "learning_rate": 0.00018354166666666668,
+      "loss": 1.4036,
+      "step": 84
+    },
+    {
+      "epoch": 0.43870967741935485,
+      "grad_norm": 0.28015458583831787,
+      "learning_rate": 0.00018333333333333334,
+      "loss": 1.4294,
+      "step": 85
+    },
+    {
+      "epoch": 0.44387096774193546,
+      "grad_norm": 0.2643658518791199,
+      "learning_rate": 0.00018312500000000002,
+      "loss": 1.459,
+      "step": 86
+    },
+    {
+      "epoch": 0.44903225806451613,
+      "grad_norm": 0.2695503234863281,
+      "learning_rate": 0.00018291666666666668,
+      "loss": 1.4856,
+      "step": 87
+    },
+    {
+      "epoch": 0.4541935483870968,
+      "grad_norm": 0.28629231452941895,
+      "learning_rate": 0.00018270833333333336,
+      "loss": 1.4686,
+      "step": 88
+    },
+    {
+      "epoch": 0.4593548387096774,
+      "grad_norm": 0.32034188508987427,
+      "learning_rate": 0.0001825,
+      "loss": 1.44,
+      "step": 89
+    },
+    {
+      "epoch": 0.4645161290322581,
+      "grad_norm": 0.27616673707962036,
+      "learning_rate": 0.00018229166666666667,
+      "loss": 1.4825,
+      "step": 90
+    },
+    {
+      "epoch": 0.4696774193548387,
+      "grad_norm": 0.2740142047405243,
+      "learning_rate": 0.00018208333333333333,
+      "loss": 1.4521,
+      "step": 91
+    },
+    {
+      "epoch": 0.47483870967741937,
+      "grad_norm": 0.28214284777641296,
+      "learning_rate": 0.00018187500000000002,
+      "loss": 1.4686,
+      "step": 92
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.25340214371681213,
+      "learning_rate": 0.00018166666666666667,
+      "loss": 1.4802,
+      "step": 93
+    },
+    {
+      "epoch": 0.48516129032258065,
+      "grad_norm": 0.2903454899787903,
+      "learning_rate": 0.00018145833333333336,
+      "loss": 1.4289,
+      "step": 94
+    },
+    {
+      "epoch": 0.49032258064516127,
+      "grad_norm": 0.2612853944301605,
+      "learning_rate": 0.00018125000000000001,
+      "loss": 1.4682,
+      "step": 95
+    },
+    {
+      "epoch": 0.49548387096774194,
+      "grad_norm": 0.2892136871814728,
+      "learning_rate": 0.00018104166666666667,
+      "loss": 1.4487,
+      "step": 96
+    },
+    {
+      "epoch": 0.5006451612903225,
+      "grad_norm": 0.2713925242424011,
+      "learning_rate": 0.00018083333333333336,
+      "loss": 1.4538,
+      "step": 97
+    },
+    {
+      "epoch": 0.5058064516129033,
+      "grad_norm": 0.2478211671113968,
+      "learning_rate": 0.000180625,
+      "loss": 1.4319,
+      "step": 98
+    },
+    {
+      "epoch": 0.5109677419354839,
+      "grad_norm": 0.2676670551300049,
+      "learning_rate": 0.00018041666666666667,
+      "loss": 1.5061,
+      "step": 99
+    },
+    {
+      "epoch": 0.5161290322580645,
+      "grad_norm": 0.358840674161911,
+      "learning_rate": 0.00018020833333333333,
+      "loss": 1.4773,
+      "step": 100
+    },
+    {
+      "epoch": 0.5212903225806451,
+      "grad_norm": 0.24543696641921997,
+      "learning_rate": 0.00018,
+      "loss": 1.5181,
+      "step": 101
+    },
+    {
+      "epoch": 0.5264516129032258,
+      "grad_norm": 0.2876887023448944,
+      "learning_rate": 0.00017979166666666667,
+      "loss": 1.4734,
+      "step": 102
+    },
+    {
+      "epoch": 0.5316129032258065,
+      "grad_norm": 0.26805076003074646,
+      "learning_rate": 0.00017958333333333335,
+      "loss": 1.4446,
+      "step": 103
+    },
+    {
+      "epoch": 0.5367741935483871,
+      "grad_norm": 0.2595707178115845,
+      "learning_rate": 0.000179375,
+      "loss": 1.3896,
+      "step": 104
+    },
+    {
+      "epoch": 0.5419354838709678,
+      "grad_norm": 0.2507528066635132,
+      "learning_rate": 0.0001791666666666667,
+      "loss": 1.4899,
+      "step": 105
+    },
+    {
+      "epoch": 0.5470967741935484,
+      "grad_norm": 0.2564782202243805,
+      "learning_rate": 0.00017895833333333335,
+      "loss": 1.4406,
+      "step": 106
+    },
+    {
+      "epoch": 0.552258064516129,
+      "grad_norm": 0.24977561831474304,
+      "learning_rate": 0.00017875,
+      "loss": 1.4383,
+      "step": 107
+    },
+    {
+      "epoch": 0.5574193548387096,
+      "grad_norm": 0.2598692774772644,
+      "learning_rate": 0.00017854166666666667,
+      "loss": 1.4169,
+      "step": 108
+    },
+    {
+      "epoch": 0.5625806451612904,
+      "grad_norm": 0.2826281785964966,
+      "learning_rate": 0.00017833333333333335,
+      "loss": 1.4823,
+      "step": 109
+    },
+    {
+      "epoch": 0.567741935483871,
+      "grad_norm": 0.5479724407196045,
+      "learning_rate": 0.000178125,
+      "loss": 1.4915,
+      "step": 110
+    },
+    {
+      "epoch": 0.5729032258064516,
+      "grad_norm": 0.25376367568969727,
+      "learning_rate": 0.00017791666666666666,
+      "loss": 1.4825,
+      "step": 111
+    },
+    {
+      "epoch": 0.5780645161290323,
+      "grad_norm": 0.27460986375808716,
+      "learning_rate": 0.00017770833333333335,
+      "loss": 1.443,
+      "step": 112
+    },
+    {
+      "epoch": 0.5832258064516129,
+      "grad_norm": 0.3225977122783661,
+      "learning_rate": 0.0001775,
+      "loss": 1.3896,
+      "step": 113
+    },
+    {
+      "epoch": 0.5883870967741935,
+      "grad_norm": 0.26176726818084717,
+      "learning_rate": 0.0001772916666666667,
+      "loss": 1.3732,
+      "step": 114
+    },
+    {
+      "epoch": 0.5935483870967742,
+      "grad_norm": 0.4086163341999054,
+      "learning_rate": 0.00017708333333333335,
+      "loss": 1.44,
+      "step": 115
+    },
+    {
+      "epoch": 0.5987096774193549,
+      "grad_norm": 0.25746962428092957,
+      "learning_rate": 0.000176875,
+      "loss": 1.4817,
+      "step": 116
+    },
+    {
+      "epoch": 0.6038709677419355,
+      "grad_norm": 0.28546246886253357,
+      "learning_rate": 0.00017666666666666666,
+      "loss": 1.4969,
+      "step": 117
+    },
+    {
+      "epoch": 0.6090322580645161,
+      "grad_norm": 0.2487628012895584,
+      "learning_rate": 0.00017645833333333334,
+      "loss": 1.4595,
+      "step": 118
+    },
+    {
+      "epoch": 0.6141935483870967,
+      "grad_norm": 0.26717832684516907,
+      "learning_rate": 0.00017625,
+      "loss": 1.4658,
+      "step": 119
+    },
+    {
+      "epoch": 0.6193548387096774,
+      "grad_norm": 0.25558313727378845,
+      "learning_rate": 0.00017604166666666669,
+      "loss": 1.4667,
+      "step": 120
+    },
+    {
+      "epoch": 0.6245161290322581,
+      "grad_norm": 0.263571560382843,
+      "learning_rate": 0.00017583333333333334,
+      "loss": 1.4602,
+      "step": 121
+    },
+    {
+      "epoch": 0.6296774193548387,
+      "grad_norm": 0.3080665171146393,
+      "learning_rate": 0.00017562500000000003,
+      "loss": 1.3268,
+      "step": 122
+    },
+    {
+      "epoch": 0.6348387096774194,
+      "grad_norm": 0.26590460538864136,
+      "learning_rate": 0.00017541666666666668,
+      "loss": 1.4909,
+      "step": 123
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7243748903274536,
+      "learning_rate": 0.00017520833333333334,
+      "loss": 1.4944,
+      "step": 124
+    },
+    {
+      "epoch": 0.6451612903225806,
+      "grad_norm": 0.5197208523750305,
+      "learning_rate": 0.000175,
+      "loss": 1.4483,
+      "step": 125
+    },
+    {
+      "epoch": 0.6503225806451612,
+      "grad_norm": 0.2566022276878357,
+      "learning_rate": 0.00017479166666666666,
+      "loss": 1.4818,
+      "step": 126
+    },
+    {
+      "epoch": 0.655483870967742,
+      "grad_norm": 0.4484021067619324,
+      "learning_rate": 0.00017458333333333334,
+      "loss": 1.4981,
+      "step": 127
+    },
+    {
+      "epoch": 0.6606451612903226,
+      "grad_norm": 0.26064327359199524,
+      "learning_rate": 0.000174375,
+      "loss": 1.4873,
+      "step": 128
+    },
+    {
+      "epoch": 0.6658064516129032,
+      "grad_norm": 0.2629943788051605,
+      "learning_rate": 0.00017416666666666668,
+      "loss": 1.5315,
+      "step": 129
+    },
+    {
+      "epoch": 0.6709677419354839,
+      "grad_norm": 0.26523739099502563,
+      "learning_rate": 0.00017395833333333334,
+      "loss": 1.4068,
+      "step": 130
+    },
+    {
+      "epoch": 0.6761290322580645,
+      "grad_norm": 0.2996821403503418,
+      "learning_rate": 0.00017375000000000002,
+      "loss": 1.438,
+      "step": 131
+    },
+    {
+      "epoch": 0.6812903225806451,
+      "grad_norm": 0.3735979497432709,
+      "learning_rate": 0.00017354166666666668,
+      "loss": 1.3826,
+      "step": 132
+    },
+    {
+      "epoch": 0.6864516129032258,
+      "grad_norm": 0.2836158573627472,
+      "learning_rate": 0.00017333333333333334,
+      "loss": 1.4757,
+      "step": 133
+    },
+    {
+      "epoch": 0.6916129032258065,
+      "grad_norm": 0.34014788269996643,
+      "learning_rate": 0.000173125,
+      "loss": 1.4641,
+      "step": 134
+    },
+    {
+      "epoch": 0.6967741935483871,
+      "grad_norm": 0.2906434237957001,
+      "learning_rate": 0.00017291666666666668,
+      "loss": 1.4734,
+      "step": 135
+    },
+    {
+      "epoch": 0.7019354838709677,
+      "grad_norm": 0.2812221348285675,
+      "learning_rate": 0.00017270833333333333,
+      "loss": 1.463,
+      "step": 136
+    },
+    {
+      "epoch": 0.7070967741935484,
+      "grad_norm": 0.23764242231845856,
+      "learning_rate": 0.00017250000000000002,
+      "loss": 1.4635,
+      "step": 137
+    },
+    {
+      "epoch": 0.712258064516129,
+      "grad_norm": 0.29817283153533936,
+      "learning_rate": 0.00017229166666666668,
+      "loss": 1.404,
+      "step": 138
+    },
+    {
+      "epoch": 0.7174193548387097,
+      "grad_norm": 0.3388894498348236,
+      "learning_rate": 0.00017208333333333336,
+      "loss": 1.4405,
+      "step": 139
+    },
+    {
+      "epoch": 0.7225806451612903,
+      "grad_norm": 0.26932695508003235,
+      "learning_rate": 0.00017187500000000002,
+      "loss": 1.4976,
+      "step": 140
+    },
+    {
+      "epoch": 0.727741935483871,
+      "grad_norm": 0.2872125804424286,
+      "learning_rate": 0.00017166666666666667,
+      "loss": 1.4831,
+      "step": 141
+    },
+    {
+      "epoch": 0.7329032258064516,
+      "grad_norm": 0.29572561383247375,
+      "learning_rate": 0.00017145833333333333,
+      "loss": 1.4303,
+      "step": 142
+    },
+    {
+      "epoch": 0.7380645161290322,
+      "grad_norm": 0.25724682211875916,
+      "learning_rate": 0.00017125,
+      "loss": 1.454,
+      "step": 143
+    },
+    {
+      "epoch": 0.743225806451613,
+      "grad_norm": 0.4314388930797577,
+      "learning_rate": 0.00017104166666666667,
+      "loss": 1.399,
+      "step": 144
+    },
+    {
+      "epoch": 0.7483870967741936,
+      "grad_norm": 0.26073867082595825,
+      "learning_rate": 0.00017083333333333333,
+      "loss": 1.3921,
+      "step": 145
+    },
+    {
+      "epoch": 0.7535483870967742,
+      "grad_norm": 0.3001374900341034,
+      "learning_rate": 0.00017062500000000001,
+      "loss": 1.4312,
+      "step": 146
+    },
+    {
+      "epoch": 0.7587096774193548,
+      "grad_norm": 0.43840205669403076,
+      "learning_rate": 0.00017041666666666667,
+      "loss": 1.4462,
+      "step": 147
+    },
+    {
+      "epoch": 0.7638709677419355,
+      "grad_norm": 0.2748488783836365,
+      "learning_rate": 0.00017020833333333336,
+      "loss": 1.3283,
+      "step": 148
+    },
+    {
+      "epoch": 0.7690322580645161,
+      "grad_norm": 0.29019895195961,
+      "learning_rate": 0.00017,
+      "loss": 1.3823,
+      "step": 149
+    },
+    {
+      "epoch": 0.7741935483870968,
+      "grad_norm": 0.25083813071250916,
+      "learning_rate": 0.00016979166666666667,
+      "loss": 1.3976,
+      "step": 150
+    },
+    {
+      "epoch": 0.7793548387096774,
+      "grad_norm": 0.22703422605991364,
+      "learning_rate": 0.00016958333333333333,
+      "loss": 1.4609,
+      "step": 151
+    },
+    {
+      "epoch": 0.7845161290322581,
+      "grad_norm": 0.3423355519771576,
+      "learning_rate": 0.000169375,
+      "loss": 1.4924,
+      "step": 152
+    },
+    {
+      "epoch": 0.7896774193548387,
+      "grad_norm": 0.27506953477859497,
+      "learning_rate": 0.00016916666666666667,
+      "loss": 1.4535,
+      "step": 153
+    },
+    {
+      "epoch": 0.7948387096774193,
+      "grad_norm": 0.2847161889076233,
+      "learning_rate": 0.00016895833333333335,
+      "loss": 1.3699,
+      "step": 154
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.25352516770362854,
+      "learning_rate": 0.00016875,
+      "loss": 1.4519,
+      "step": 155
+    },
+    {
+      "epoch": 0.8051612903225807,
+      "grad_norm": 0.251001238822937,
+      "learning_rate": 0.00016854166666666667,
+      "loss": 1.4398,
+      "step": 156
+    },
+    {
+      "epoch": 0.8103225806451613,
+      "grad_norm": 0.32508385181427,
+      "learning_rate": 0.00016833333333333335,
+      "loss": 1.5424,
+      "step": 157
+    },
+    {
+      "epoch": 0.8154838709677419,
+      "grad_norm": 0.23881129920482635,
+      "learning_rate": 0.000168125,
+      "loss": 1.4728,
+      "step": 158
+    },
+    {
+      "epoch": 0.8206451612903226,
+      "grad_norm": 0.42169275879859924,
+      "learning_rate": 0.00016791666666666666,
+      "loss": 1.3924,
+      "step": 159
+    },
+    {
+      "epoch": 0.8258064516129032,
+      "grad_norm": 0.31452420353889465,
+      "learning_rate": 0.00016770833333333332,
+      "loss": 1.463,
+      "step": 160
+    },
+    {
+      "epoch": 0.8309677419354838,
+      "grad_norm": 0.24824711680412292,
+      "learning_rate": 0.0001675,
+      "loss": 1.4583,
+      "step": 161
+    },
+    {
+      "epoch": 0.8361290322580646,
+      "grad_norm": 0.26430845260620117,
+      "learning_rate": 0.00016729166666666666,
+      "loss": 1.4791,
+      "step": 162
+    },
+    {
+      "epoch": 0.8412903225806452,
+      "grad_norm": 0.25048938393592834,
+      "learning_rate": 0.00016708333333333335,
+      "loss": 1.3997,
+      "step": 163
+    },
+    {
+      "epoch": 0.8464516129032258,
+      "grad_norm": 0.23190194368362427,
+      "learning_rate": 0.000166875,
+      "loss": 1.4007,
+      "step": 164
+    },
+    {
+      "epoch": 0.8516129032258064,
+      "grad_norm": 0.25311771035194397,
+      "learning_rate": 0.0001666666666666667,
+      "loss": 1.4117,
+      "step": 165
+    },
+    {
+      "epoch": 0.8567741935483871,
+      "grad_norm": 0.24602946639060974,
+      "learning_rate": 0.00016645833333333335,
+      "loss": 1.4669,
+      "step": 166
+    },
+    {
+      "epoch": 0.8619354838709677,
+      "grad_norm": 0.31980133056640625,
+      "learning_rate": 0.00016625000000000003,
+      "loss": 1.2822,
+      "step": 167
+    },
+    {
+      "epoch": 0.8670967741935484,
+      "grad_norm": 0.25325366854667664,
+      "learning_rate": 0.0001660416666666667,
+      "loss": 1.437,
+      "step": 168
+    },
+    {
+      "epoch": 0.8722580645161291,
+      "grad_norm": 0.26424503326416016,
+      "learning_rate": 0.00016583333333333334,
+      "loss": 1.4439,
+      "step": 169
+    },
+    {
+      "epoch": 0.8774193548387097,
+      "grad_norm": 0.3035740554332733,
+      "learning_rate": 0.000165625,
+      "loss": 1.3533,
+      "step": 170
+    },
+    {
+      "epoch": 0.8825806451612903,
+      "grad_norm": 0.25016552209854126,
+      "learning_rate": 0.00016541666666666666,
+      "loss": 1.4572,
+      "step": 171
+    },
+    {
+      "epoch": 0.8877419354838709,
+      "grad_norm": 0.27196910977363586,
+      "learning_rate": 0.00016520833333333334,
+      "loss": 1.4935,
+      "step": 172
+    },
+    {
+      "epoch": 0.8929032258064517,
+      "grad_norm": 0.23871557414531708,
+      "learning_rate": 0.000165,
+      "loss": 1.4389,
+      "step": 173
+    },
+    {
+      "epoch": 0.8980645161290323,
+      "grad_norm": 0.4004731774330139,
+      "learning_rate": 0.00016479166666666668,
+      "loss": 1.3672,
+      "step": 174
+    },
+    {
+      "epoch": 0.9032258064516129,
+      "grad_norm": 0.27934232354164124,
+      "learning_rate": 0.00016458333333333334,
+      "loss": 1.3995,
+      "step": 175
+    },
+    {
+      "epoch": 0.9083870967741936,
+      "grad_norm": 0.35883888602256775,
+      "learning_rate": 0.00016437500000000002,
+      "loss": 1.2908,
+      "step": 176
+    },
+    {
+      "epoch": 0.9135483870967742,
+      "grad_norm": 0.2548966705799103,
+      "learning_rate": 0.00016416666666666668,
+      "loss": 1.4714,
+      "step": 177
+    },
+    {
+      "epoch": 0.9187096774193548,
+      "grad_norm": 0.2585020661354065,
+      "learning_rate": 0.00016395833333333334,
+      "loss": 1.4843,
+      "step": 178
+    },
+    {
+      "epoch": 0.9238709677419354,
+      "grad_norm": 0.2773195803165436,
+      "learning_rate": 0.00016375,
+      "loss": 1.4401,
+      "step": 179
+    },
+    {
+      "epoch": 0.9290322580645162,
+      "grad_norm": 0.2668519914150238,
+      "learning_rate": 0.00016354166666666668,
+      "loss": 1.4806,
+      "step": 180
+    },
+    {
+      "epoch": 0.9341935483870968,
+      "grad_norm": 0.2520642578601837,
+      "learning_rate": 0.00016333333333333334,
+      "loss": 1.487,
+      "step": 181
+    },
+    {
+      "epoch": 0.9393548387096774,
+      "grad_norm": 0.2447890341281891,
+      "learning_rate": 0.00016312500000000002,
+      "loss": 1.4299,
+      "step": 182
+    },
+    {
+      "epoch": 0.944516129032258,
+      "grad_norm": 0.2460276186466217,
+      "learning_rate": 0.00016291666666666668,
+      "loss": 1.4403,
+      "step": 183
+    },
+    {
+      "epoch": 0.9496774193548387,
+      "grad_norm": 0.2651495337486267,
+      "learning_rate": 0.00016270833333333336,
+      "loss": 1.4468,
+      "step": 184
+    },
+    {
+      "epoch": 0.9548387096774194,
+      "grad_norm": 0.24191385507583618,
+      "learning_rate": 0.00016250000000000002,
+      "loss": 1.4421,
+      "step": 185
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.24822644889354706,
+      "learning_rate": 0.00016229166666666668,
+      "loss": 1.4724,
+      "step": 186
+    },
+    {
+      "epoch": 0.9651612903225807,
+      "grad_norm": 0.3316594064235687,
+      "learning_rate": 0.00016208333333333333,
+      "loss": 1.431,
+      "step": 187
+    },
+    {
+      "epoch": 0.9703225806451613,
+      "grad_norm": 0.24073320627212524,
+      "learning_rate": 0.000161875,
+      "loss": 1.4648,
+      "step": 188
+    },
+    {
+      "epoch": 0.9754838709677419,
+      "grad_norm": 0.2788798213005066,
+      "learning_rate": 0.00016166666666666668,
+      "loss": 1.4203,
+      "step": 189
+    },
+    {
+      "epoch": 0.9806451612903225,
+      "grad_norm": 0.24988716840744019,
+      "learning_rate": 0.00016145833333333333,
+      "loss": 1.4586,
+      "step": 190
+    },
+    {
+      "epoch": 0.9858064516129033,
+      "grad_norm": 0.25704294443130493,
+      "learning_rate": 0.00016125000000000002,
+      "loss": 1.4066,
+      "step": 191
+    },
+    {
+      "epoch": 0.9909677419354839,
+      "grad_norm": 0.2555101811885834,
+      "learning_rate": 0.00016104166666666667,
+      "loss": 1.4698,
+      "step": 192
+    },
+    {
+      "epoch": 0.9961290322580645,
+      "grad_norm": 0.32110294699668884,
+      "learning_rate": 0.00016083333333333336,
+      "loss": 1.4582,
+      "step": 193
+    },
+    {
+      "epoch": 1.0025806451612904,
+      "grad_norm": 0.38127434253692627,
+      "learning_rate": 0.00016062500000000001,
+      "loss": 2.0881,
+      "step": 194
+    },
+    {
+      "epoch": 1.007741935483871,
+      "grad_norm": 0.2512829601764679,
+      "learning_rate": 0.00016041666666666667,
+      "loss": 1.2625,
+      "step": 195
+    },
+    {
+      "epoch": 1.0129032258064516,
+      "grad_norm": 0.27245259284973145,
+      "learning_rate": 0.00016020833333333333,
+      "loss": 1.2867,
+      "step": 196
+    },
+    {
+      "epoch": 1.0180645161290323,
+      "grad_norm": 0.2516975402832031,
+      "learning_rate": 0.00016,
+      "loss": 1.3286,
+      "step": 197
+    },
+    {
+      "epoch": 1.0232258064516129,
+      "grad_norm": 0.31121954321861267,
+      "learning_rate": 0.00015979166666666667,
+      "loss": 1.3711,
+      "step": 198
+    },
+    {
+      "epoch": 1.0283870967741935,
+      "grad_norm": 0.30174267292022705,
+      "learning_rate": 0.00015958333333333335,
+      "loss": 1.3604,
+      "step": 199
+    },
+    {
+      "epoch": 1.033548387096774,
+      "grad_norm": 0.2846987843513489,
+      "learning_rate": 0.000159375,
+      "loss": 1.354,
+      "step": 200
+    },
+    {
+      "epoch": 1.038709677419355,
+      "grad_norm": 0.29153668880462646,
+      "learning_rate": 0.00015916666666666667,
+      "loss": 1.3471,
+      "step": 201
+    },
+    {
+      "epoch": 1.0438709677419356,
+      "grad_norm": 0.30284610390663147,
+      "learning_rate": 0.00015895833333333335,
+      "loss": 1.397,
+      "step": 202
+    },
+    {
+      "epoch": 1.0490322580645162,
+      "grad_norm": 0.3039468228816986,
+      "learning_rate": 0.00015875,
+      "loss": 1.352,
+      "step": 203
+    },
+    {
+      "epoch": 1.0541935483870968,
+      "grad_norm": 0.3008759021759033,
+      "learning_rate": 0.00015854166666666667,
+      "loss": 1.3703,
+      "step": 204
+    },
+    {
+      "epoch": 1.0593548387096774,
+      "grad_norm": 0.30333173274993896,
+      "learning_rate": 0.00015833333333333332,
+      "loss": 1.4017,
+      "step": 205
+    },
+    {
+      "epoch": 1.064516129032258,
+      "grad_norm": 0.30104291439056396,
+      "learning_rate": 0.000158125,
+      "loss": 1.3435,
+      "step": 206
+    },
+    {
+      "epoch": 1.0696774193548386,
+      "grad_norm": 0.330991268157959,
+      "learning_rate": 0.00015791666666666667,
+      "loss": 1.3576,
+      "step": 207
+    },
+    {
+      "epoch": 1.0748387096774195,
+      "grad_norm": 0.29556649923324585,
+      "learning_rate": 0.00015770833333333335,
+      "loss": 1.3096,
+      "step": 208
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.2945300042629242,
+      "learning_rate": 0.0001575,
+      "loss": 1.2997,
+      "step": 209
+    },
+    {
+      "epoch": 1.0851612903225807,
+      "grad_norm": 0.3147192597389221,
+      "learning_rate": 0.0001572916666666667,
+      "loss": 1.3169,
+      "step": 210
+    },
+    {
+      "epoch": 1.0903225806451613,
+      "grad_norm": 0.32288652658462524,
+      "learning_rate": 0.00015708333333333335,
+      "loss": 1.3542,
+      "step": 211
+    },
+    {
+      "epoch": 1.095483870967742,
+      "grad_norm": 0.32715293765068054,
+      "learning_rate": 0.000156875,
+      "loss": 1.2741,
+      "step": 212
+    },
+    {
+      "epoch": 1.1006451612903225,
+      "grad_norm": 0.34655511379241943,
+      "learning_rate": 0.00015666666666666666,
+      "loss": 1.3624,
+      "step": 213
+    },
+    {
+      "epoch": 1.1058064516129031,
+      "grad_norm": 0.33325520157814026,
+      "learning_rate": 0.00015645833333333335,
+      "loss": 1.2963,
+      "step": 214
+    },
+    {
+      "epoch": 1.1109677419354838,
+      "grad_norm": 0.3129732012748718,
+      "learning_rate": 0.00015625,
+      "loss": 1.2475,
+      "step": 215
+    },
+    {
+      "epoch": 1.1161290322580646,
+      "grad_norm": 0.31625816226005554,
+      "learning_rate": 0.00015604166666666666,
+      "loss": 1.3668,
+      "step": 216
+    },
+    {
+      "epoch": 1.1212903225806452,
+      "grad_norm": 0.40074801445007324,
+      "learning_rate": 0.00015583333333333334,
+      "loss": 1.373,
+      "step": 217
+    },
+    {
+      "epoch": 1.1264516129032258,
+      "grad_norm": 0.34948697686195374,
+      "learning_rate": 0.000155625,
+      "loss": 1.3161,
+      "step": 218
+    },
+    {
+      "epoch": 1.1316129032258064,
+      "grad_norm": 0.3386715054512024,
+      "learning_rate": 0.00015541666666666669,
+      "loss": 1.3309,
+      "step": 219
+    },
+    {
+      "epoch": 1.136774193548387,
+      "grad_norm": 0.36897939443588257,
+      "learning_rate": 0.00015520833333333334,
+      "loss": 1.3476,
+      "step": 220
+    },
+    {
+      "epoch": 1.1419354838709677,
+      "grad_norm": 0.32543185353279114,
+      "learning_rate": 0.000155,
+      "loss": 1.2557,
+      "step": 221
+    },
+    {
+      "epoch": 1.1470967741935483,
+      "grad_norm": 0.32470881938934326,
+      "learning_rate": 0.00015479166666666666,
+      "loss": 1.2733,
+      "step": 222
+    },
+    {
+      "epoch": 1.152258064516129,
+      "grad_norm": 0.3357822000980377,
+      "learning_rate": 0.00015458333333333334,
+      "loss": 1.3103,
+      "step": 223
+    },
+    {
+      "epoch": 1.1574193548387097,
+      "grad_norm": 0.3299640119075775,
+      "learning_rate": 0.000154375,
+      "loss": 1.2753,
+      "step": 224
+    },
+    {
+      "epoch": 1.1625806451612903,
+      "grad_norm": 0.3236103951931,
+      "learning_rate": 0.00015416666666666668,
+      "loss": 1.2911,
+      "step": 225
+    },
+    {
+      "epoch": 1.167741935483871,
+      "grad_norm": 0.3308989703655243,
+      "learning_rate": 0.00015395833333333334,
+      "loss": 1.3789,
+      "step": 226
+    },
+    {
+      "epoch": 1.1729032258064516,
+      "grad_norm": 0.35100898146629333,
+      "learning_rate": 0.00015375000000000002,
+      "loss": 1.3381,
+      "step": 227
+    },
+    {
+      "epoch": 1.1780645161290322,
+      "grad_norm": 0.34906187653541565,
+      "learning_rate": 0.00015354166666666668,
+      "loss": 1.3911,
+      "step": 228
+    },
+    {
+      "epoch": 1.1832258064516128,
+      "grad_norm": 0.3610292077064514,
+      "learning_rate": 0.00015333333333333334,
+      "loss": 1.3375,
+      "step": 229
+    },
+    {
+      "epoch": 1.1883870967741936,
+      "grad_norm": 0.45614510774612427,
+      "learning_rate": 0.000153125,
+      "loss": 1.3852,
+      "step": 230
+    },
+    {
+      "epoch": 1.1935483870967742,
+      "grad_norm": 0.3600503206253052,
+      "learning_rate": 0.00015291666666666665,
+      "loss": 1.3558,
+      "step": 231
+    },
+    {
+      "epoch": 1.1987096774193549,
+      "grad_norm": 0.3356868326663971,
+      "learning_rate": 0.00015270833333333334,
+      "loss": 1.3544,
+      "step": 232
+    },
+    {
+      "epoch": 1.2038709677419355,
+      "grad_norm": 0.3423336148262024,
+      "learning_rate": 0.0001525,
+      "loss": 1.2441,
+      "step": 233
+    },
+    {
+      "epoch": 1.209032258064516,
+      "grad_norm": 0.3377212584018707,
+      "learning_rate": 0.00015229166666666668,
+      "loss": 1.3516,
+      "step": 234
+    },
+    {
+      "epoch": 1.2141935483870967,
+      "grad_norm": 0.340678334236145,
+      "learning_rate": 0.00015208333333333333,
+      "loss": 1.3069,
+      "step": 235
+    },
+    {
+      "epoch": 1.2193548387096773,
+      "grad_norm": 0.3946521580219269,
+      "learning_rate": 0.00015187500000000002,
+      "loss": 1.282,
+      "step": 236
+    },
+    {
+      "epoch": 1.2245161290322581,
+      "grad_norm": 0.40371814370155334,
+      "learning_rate": 0.00015166666666666668,
+      "loss": 1.408,
+      "step": 237
+    },
+    {
+      "epoch": 1.2296774193548388,
+      "grad_norm": 0.33231934905052185,
+      "learning_rate": 0.00015145833333333333,
+      "loss": 1.3046,
+      "step": 238
+    },
+    {
+      "epoch": 1.2348387096774194,
+      "grad_norm": 0.3452011048793793,
+      "learning_rate": 0.00015125,
+      "loss": 1.2823,
+      "step": 239
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.33732855319976807,
+      "learning_rate": 0.00015104166666666667,
+      "loss": 1.3071,
+      "step": 240
+    },
+    {
+      "epoch": 1.2451612903225806,
+      "grad_norm": 0.3544975817203522,
+      "learning_rate": 0.00015083333333333333,
+      "loss": 1.3537,
+      "step": 241
+    },
+    {
+      "epoch": 1.2503225806451612,
+      "grad_norm": 0.3354959785938263,
+      "learning_rate": 0.00015062500000000002,
+      "loss": 1.2918,
+      "step": 242
+    },
+    {
+      "epoch": 1.2554838709677418,
+      "grad_norm": 0.35639578104019165,
+      "learning_rate": 0.00015041666666666667,
+      "loss": 1.3281,
+      "step": 243
+    },
+    {
+      "epoch": 1.2606451612903227,
+      "grad_norm": 0.34774360060691833,
+      "learning_rate": 0.00015020833333333336,
+      "loss": 1.351,
+      "step": 244
+    },
+    {
+      "epoch": 1.2658064516129033,
+      "grad_norm": 0.32612374424934387,
+      "learning_rate": 0.00015000000000000001,
+      "loss": 1.2595,
+      "step": 245
+    },
+    {
+      "epoch": 1.270967741935484,
+      "grad_norm": 0.38271042704582214,
+      "learning_rate": 0.00014979166666666667,
+      "loss": 1.1875,
+      "step": 246
+    },
+    {
+      "epoch": 1.2761290322580645,
+      "grad_norm": 0.3386516571044922,
+      "learning_rate": 0.00014958333333333336,
+      "loss": 1.334,
+      "step": 247
+    },
+    {
+      "epoch": 1.2812903225806451,
+      "grad_norm": 0.371288001537323,
+      "learning_rate": 0.00014937499999999999,
+      "loss": 1.318,
+      "step": 248
+    },
+    {
+      "epoch": 1.2864516129032257,
+      "grad_norm": 0.35285741090774536,
+      "learning_rate": 0.00014916666666666667,
+      "loss": 1.3182,
+      "step": 249
+    },
+    {
+      "epoch": 1.2916129032258064,
+      "grad_norm": 0.3875463902950287,
+      "learning_rate": 0.00014895833333333333,
+      "loss": 1.385,
+      "step": 250
+    },
+    {
+      "epoch": 1.2967741935483872,
+      "grad_norm": 0.3309260308742523,
+      "learning_rate": 0.00014875,
+      "loss": 1.3276,
+      "step": 251
+    },
+    {
+      "epoch": 1.3019354838709678,
+      "grad_norm": 0.3166562616825104,
+      "learning_rate": 0.00014854166666666667,
+      "loss": 1.3561,
+      "step": 252
+    },
+    {
+      "epoch": 1.3070967741935484,
+      "grad_norm": 0.30836954712867737,
+      "learning_rate": 0.00014833333333333335,
+      "loss": 1.3112,
+      "step": 253
+    },
+    {
+      "epoch": 1.312258064516129,
+      "grad_norm": 0.35156041383743286,
+      "learning_rate": 0.000148125,
+      "loss": 1.2821,
+      "step": 254
+    },
+    {
+      "epoch": 1.3174193548387096,
+      "grad_norm": 0.3397453725337982,
+      "learning_rate": 0.0001479166666666667,
+      "loss": 1.327,
+      "step": 255
+    },
+    {
+      "epoch": 1.3225806451612903,
+      "grad_norm": 1.1283373832702637,
+      "learning_rate": 0.00014770833333333335,
+      "loss": 1.2821,
+      "step": 256
+    },
+    {
+      "epoch": 1.3277419354838709,
+      "grad_norm": 0.33926060795783997,
+      "learning_rate": 0.0001475,
+      "loss": 1.3943,
+      "step": 257
+    },
+    {
+      "epoch": 1.3329032258064517,
+      "grad_norm": 0.3800329864025116,
+      "learning_rate": 0.00014729166666666666,
+      "loss": 1.2681,
+      "step": 258
+    },
+    {
+      "epoch": 1.3380645161290323,
+      "grad_norm": 0.3152649998664856,
+      "learning_rate": 0.00014708333333333335,
+      "loss": 1.3031,
+      "step": 259
+    },
+    {
+      "epoch": 1.343225806451613,
+      "grad_norm": 0.35031044483184814,
+      "learning_rate": 0.000146875,
+      "loss": 1.306,
+      "step": 260
+    },
+    {
+      "epoch": 1.3483870967741935,
+      "grad_norm": 0.3241156041622162,
+      "learning_rate": 0.00014666666666666666,
+      "loss": 1.3768,
+      "step": 261
+    },
+    {
+      "epoch": 1.3535483870967742,
+      "grad_norm": 0.35180339217185974,
+      "learning_rate": 0.00014645833333333335,
+      "loss": 1.3108,
+      "step": 262
+    },
+    {
+      "epoch": 1.3587096774193548,
+      "grad_norm": 0.33626481890678406,
+      "learning_rate": 0.00014625,
+      "loss": 1.3053,
+      "step": 263
+    },
+    {
+      "epoch": 1.3638709677419354,
+      "grad_norm": 0.3846721351146698,
+      "learning_rate": 0.0001460416666666667,
+      "loss": 1.3466,
+      "step": 264
+    },
+    {
+      "epoch": 1.3690322580645162,
+      "grad_norm": 0.39349740743637085,
+      "learning_rate": 0.00014583333333333335,
+      "loss": 1.2446,
+      "step": 265
+    },
+    {
+      "epoch": 1.3741935483870968,
+      "grad_norm": 0.3652196526527405,
+      "learning_rate": 0.000145625,
+      "loss": 1.3862,
+      "step": 266
+    },
+    {
+      "epoch": 1.3793548387096775,
+      "grad_norm": 0.3696790635585785,
+      "learning_rate": 0.00014541666666666666,
+      "loss": 1.3908,
+      "step": 267
+    },
+    {
+      "epoch": 1.384516129032258,
+      "grad_norm": 0.3426453173160553,
+      "learning_rate": 0.00014520833333333334,
+      "loss": 1.227,
+      "step": 268
+    },
+    {
+      "epoch": 1.3896774193548387,
+      "grad_norm": 0.34999752044677734,
+      "learning_rate": 0.000145,
+      "loss": 1.3119,
+      "step": 269
+    },
+    {
+      "epoch": 1.3948387096774193,
+      "grad_norm": 0.32980069518089294,
+      "learning_rate": 0.00014479166666666669,
+      "loss": 1.2531,
+      "step": 270
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.48813432455062866,
+      "learning_rate": 0.00014458333333333334,
+      "loss": 1.2712,
+      "step": 271
+    },
+    {
+      "epoch": 1.4051612903225807,
+      "grad_norm": 0.34689202904701233,
+      "learning_rate": 0.00014437500000000003,
+      "loss": 1.3685,
+      "step": 272
+    },
+    {
+      "epoch": 1.4103225806451614,
+      "grad_norm": 0.36113840341567993,
+      "learning_rate": 0.00014416666666666668,
+      "loss": 1.2531,
+      "step": 273
+    },
+    {
+      "epoch": 1.415483870967742,
+      "grad_norm": 0.3345704972743988,
+      "learning_rate": 0.00014395833333333334,
+      "loss": 1.3315,
+      "step": 274
+    },
+    {
+      "epoch": 1.4206451612903226,
+      "grad_norm": 0.41671687364578247,
+      "learning_rate": 0.00014375,
+      "loss": 1.3168,
+      "step": 275
+    },
+    {
+      "epoch": 1.4258064516129032,
+      "grad_norm": 0.3431110680103302,
+      "learning_rate": 0.00014354166666666665,
+      "loss": 1.2936,
+      "step": 276
+    },
+    {
+      "epoch": 1.4309677419354838,
+      "grad_norm": 0.36257949471473694,
+      "learning_rate": 0.00014333333333333334,
+      "loss": 1.3312,
+      "step": 277
+    },
+    {
+      "epoch": 1.4361290322580644,
+      "grad_norm": 0.3584994971752167,
+      "learning_rate": 0.000143125,
+      "loss": 1.3274,
+      "step": 278
+    },
+    {
+      "epoch": 1.4412903225806453,
+      "grad_norm": 0.37135449051856995,
+      "learning_rate": 0.00014291666666666668,
+      "loss": 1.3329,
+      "step": 279
+    },
+    {
+      "epoch": 1.4464516129032259,
+      "grad_norm": 0.3631846308708191,
+      "learning_rate": 0.00014270833333333334,
+      "loss": 1.3609,
+      "step": 280
+    },
+    {
+      "epoch": 1.4516129032258065,
+      "grad_norm": 0.39546501636505127,
+      "learning_rate": 0.00014250000000000002,
+      "loss": 1.3477,
+      "step": 281
+    },
+    {
+      "epoch": 1.456774193548387,
+      "grad_norm": 0.3552427589893341,
+      "learning_rate": 0.00014229166666666668,
+      "loss": 1.32,
+      "step": 282
+    },
+    {
+      "epoch": 1.4619354838709677,
+      "grad_norm": 0.38950273394584656,
+      "learning_rate": 0.00014208333333333334,
+      "loss": 1.3553,
+      "step": 283
+    },
+    {
+      "epoch": 1.4670967741935483,
+      "grad_norm": 0.33400124311447144,
+      "learning_rate": 0.000141875,
+      "loss": 1.3552,
+      "step": 284
+    },
+    {
+      "epoch": 1.472258064516129,
+      "grad_norm": 0.34412115812301636,
+      "learning_rate": 0.00014166666666666668,
+      "loss": 1.3003,
+      "step": 285
+    },
+    {
+      "epoch": 1.4774193548387098,
+      "grad_norm": 0.3470352590084076,
+      "learning_rate": 0.00014145833333333333,
+      "loss": 1.3113,
+      "step": 286
+    },
+    {
+      "epoch": 1.4825806451612904,
+      "grad_norm": 0.7647251486778259,
+      "learning_rate": 0.00014125000000000002,
+      "loss": 1.1627,
+      "step": 287
+    },
+    {
+      "epoch": 1.487741935483871,
+      "grad_norm": 0.3180231750011444,
+      "learning_rate": 0.00014104166666666668,
+      "loss": 1.3094,
+      "step": 288
+    },
+    {
+      "epoch": 1.4929032258064516,
+      "grad_norm": 0.358530193567276,
+      "learning_rate": 0.00014083333333333336,
+      "loss": 1.3224,
+      "step": 289
+    },
+    {
+      "epoch": 1.4980645161290322,
+      "grad_norm": 0.36877402663230896,
+      "learning_rate": 0.00014062500000000002,
+      "loss": 1.2069,
+      "step": 290
+    },
+    {
+      "epoch": 1.4980645161290322,
+      "eval_loss": 1.4567794799804688,
+      "eval_runtime": 21.5535,
+      "eval_samples_per_second": 3.665,
+      "eval_steps_per_second": 0.464,
+      "step": 290
+    },
+    {
+      "epoch": 1.5032258064516129,
+      "grad_norm": 0.36623072624206543,
+      "learning_rate": 0.00014041666666666667,
+      "loss": 1.3277,
+      "step": 291
+    },
+    {
+      "epoch": 1.5083870967741935,
+      "grad_norm": 0.367258220911026,
+      "learning_rate": 0.00014020833333333333,
+      "loss": 1.2029,
+      "step": 292
+    },
+    {
+      "epoch": 1.5135483870967743,
+      "grad_norm": 0.3648794889450073,
+      "learning_rate": 0.00014,
+      "loss": 1.3565,
+      "step": 293
+    },
+    {
+      "epoch": 1.5187096774193547,
+      "grad_norm": 0.33040714263916016,
+      "learning_rate": 0.00013979166666666667,
+      "loss": 1.3583,
+      "step": 294
+    },
+    {
+      "epoch": 1.5238709677419355,
+      "grad_norm": 0.33568263053894043,
+      "learning_rate": 0.00013958333333333333,
+      "loss": 1.3264,
+      "step": 295
+    },
+    {
+      "epoch": 1.5290322580645161,
+      "grad_norm": 0.3435729146003723,
+      "learning_rate": 0.000139375,
+      "loss": 1.3458,
+      "step": 296
+    },
+    {
+      "epoch": 1.5341935483870968,
+      "grad_norm": 0.34406206011772156,
+      "learning_rate": 0.00013916666666666667,
+      "loss": 1.3851,
+      "step": 297
+    },
+    {
+      "epoch": 1.5393548387096774,
+      "grad_norm": 0.34225884079933167,
+      "learning_rate": 0.00013895833333333335,
+      "loss": 1.3058,
+      "step": 298
+    },
+    {
+      "epoch": 1.544516129032258,
+      "grad_norm": 0.34024402499198914,
+      "learning_rate": 0.00013875,
+      "loss": 1.2546,
+      "step": 299
+    },
+    {
+      "epoch": 1.5496774193548388,
+      "grad_norm": 0.359432578086853,
+      "learning_rate": 0.00013854166666666667,
+      "loss": 1.2083,
+      "step": 300
+    },
+    {
+      "epoch": 1.5548387096774192,
+      "grad_norm": 0.3488602340221405,
+      "learning_rate": 0.00013833333333333333,
+      "loss": 1.2802,
+      "step": 301
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.3424002528190613,
+      "learning_rate": 0.000138125,
+      "loss": 1.3004,
+      "step": 302
+    },
+    {
+      "epoch": 1.5651612903225807,
+      "grad_norm": 0.3500625193119049,
+      "learning_rate": 0.00013791666666666667,
+      "loss": 1.2455,
+      "step": 303
+    },
+    {
+      "epoch": 1.5703225806451613,
+      "grad_norm": 0.33178749680519104,
+      "learning_rate": 0.00013770833333333335,
+      "loss": 1.3174,
+      "step": 304
+    },
+    {
+      "epoch": 1.575483870967742,
+      "grad_norm": 0.329635351896286,
+      "learning_rate": 0.0001375,
+      "loss": 1.3551,
+      "step": 305
+    },
+    {
+      "epoch": 1.5806451612903225,
+      "grad_norm": 0.34704092144966125,
+      "learning_rate": 0.00013729166666666667,
+      "loss": 1.3841,
+      "step": 306
+    },
+    {
+      "epoch": 1.5858064516129033,
+      "grad_norm": 0.3591984510421753,
+      "learning_rate": 0.00013708333333333335,
+      "loss": 1.2913,
+      "step": 307
+    },
+    {
+      "epoch": 1.5909677419354837,
+      "grad_norm": 0.33542248606681824,
+      "learning_rate": 0.000136875,
+      "loss": 1.3756,
+      "step": 308
+    },
+    {
+      "epoch": 1.5961290322580646,
+      "grad_norm": 0.3401111364364624,
+      "learning_rate": 0.00013666666666666666,
+      "loss": 1.3456,
+      "step": 309
+    },
+    {
+      "epoch": 1.6012903225806452,
+      "grad_norm": 0.36281338334083557,
+      "learning_rate": 0.00013645833333333332,
+      "loss": 1.2436,
+      "step": 310
+    },
+    {
+      "epoch": 1.6064516129032258,
+      "grad_norm": 0.38459697365760803,
+      "learning_rate": 0.00013625,
+      "loss": 1.3186,
+      "step": 311
+    },
+    {
+      "epoch": 1.6116129032258064,
+      "grad_norm": 0.33517658710479736,
+      "learning_rate": 0.00013604166666666666,
+      "loss": 1.3631,
+      "step": 312
+    },
+    {
+      "epoch": 1.616774193548387,
+      "grad_norm": 0.3519268035888672,
+      "learning_rate": 0.00013583333333333335,
+      "loss": 1.3753,
+      "step": 313
+    },
+    {
+      "epoch": 1.6219354838709679,
+      "grad_norm": 0.3473758101463318,
+      "learning_rate": 0.000135625,
+      "loss": 1.3139,
+      "step": 314
+    },
+    {
+      "epoch": 1.6270967741935483,
+      "grad_norm": 0.33942532539367676,
+      "learning_rate": 0.0001354166666666667,
+      "loss": 1.2212,
+      "step": 315
+    },
+    {
+      "epoch": 1.632258064516129,
+      "grad_norm": 0.33516523241996765,
+      "learning_rate": 0.00013520833333333334,
+      "loss": 1.2754,
+      "step": 316
+    },
+    {
+      "epoch": 1.6374193548387097,
+      "grad_norm": 0.35950717329978943,
+      "learning_rate": 0.00013500000000000003,
+      "loss": 1.3181,
+      "step": 317
+    },
+    {
+      "epoch": 1.6425806451612903,
+      "grad_norm": 0.33711084723472595,
+      "learning_rate": 0.00013479166666666666,
+      "loss": 1.3093,
+      "step": 318
+    },
+    {
+      "epoch": 1.647741935483871,
+      "grad_norm": 0.34433484077453613,
+      "learning_rate": 0.00013458333333333334,
+      "loss": 1.3016,
+      "step": 319
+    },
+    {
+      "epoch": 1.6529032258064515,
+      "grad_norm": 0.34102678298950195,
+      "learning_rate": 0.000134375,
+      "loss": 1.3645,
+      "step": 320
+    },
+    {
+      "epoch": 1.6580645161290324,
+      "grad_norm": 0.5951990485191345,
+      "learning_rate": 0.00013416666666666666,
+      "loss": 1.3206,
+      "step": 321
+    },
+    {
+      "epoch": 1.6632258064516128,
+      "grad_norm": 0.35340815782546997,
+      "learning_rate": 0.00013395833333333334,
+      "loss": 1.3128,
+      "step": 322
+    },
+    {
+      "epoch": 1.6683870967741936,
+      "grad_norm": 0.361688494682312,
+      "learning_rate": 0.00013375,
+      "loss": 1.3147,
+      "step": 323
+    },
+    {
+      "epoch": 1.6735483870967742,
+      "grad_norm": 0.37303170561790466,
+      "learning_rate": 0.00013354166666666668,
+      "loss": 1.2384,
+      "step": 324
+    },
+    {
+      "epoch": 1.6787096774193548,
+      "grad_norm": 0.3616917133331299,
+      "learning_rate": 0.00013333333333333334,
+      "loss": 1.3155,
+      "step": 325
+    },
+    {
+      "epoch": 1.6838709677419355,
+      "grad_norm": 0.37222525477409363,
+      "learning_rate": 0.00013312500000000002,
+      "loss": 1.251,
+      "step": 326
+    },
+    {
+      "epoch": 1.689032258064516,
+      "grad_norm": 0.36939021944999695,
+      "learning_rate": 0.00013291666666666665,
+      "loss": 1.353,
+      "step": 327
+    },
+    {
+      "epoch": 1.694193548387097,
+      "grad_norm": 0.350346177816391,
+      "learning_rate": 0.00013270833333333334,
+      "loss": 1.2388,
+      "step": 328
+    },
+    {
+      "epoch": 1.6993548387096773,
+      "grad_norm": 0.37214037775993347,
+      "learning_rate": 0.0001325,
+      "loss": 1.351,
+      "step": 329
+    },
+    {
+      "epoch": 1.7045161290322581,
+      "grad_norm": 0.3510715663433075,
+      "learning_rate": 0.00013229166666666668,
+      "loss": 1.3355,
+      "step": 330
+    },
+    {
+      "epoch": 1.7096774193548387,
+      "grad_norm": 0.3241048753261566,
+      "learning_rate": 0.00013208333333333334,
+      "loss": 1.3008,
+      "step": 331
+    },
+    {
+      "epoch": 1.7148387096774194,
+      "grad_norm": 0.3297675848007202,
+      "learning_rate": 0.00013187500000000002,
+      "loss": 1.291,
+      "step": 332
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.3625437617301941,
+      "learning_rate": 0.00013166666666666668,
+      "loss": 1.3254,
+      "step": 333
+    },
+    {
+      "epoch": 1.7251612903225806,
+      "grad_norm": 0.35241106152534485,
+      "learning_rate": 0.00013145833333333336,
+      "loss": 1.2829,
+      "step": 334
+    },
+    {
+      "epoch": 1.7303225806451614,
+      "grad_norm": 0.4532613158226013,
+      "learning_rate": 0.00013125000000000002,
+      "loss": 1.3189,
+      "step": 335
+    },
+    {
+      "epoch": 1.7354838709677418,
+      "grad_norm": 0.3502637445926666,
+      "learning_rate": 0.00013104166666666668,
+      "loss": 1.3774,
+      "step": 336
+    },
+    {
+      "epoch": 1.7406451612903227,
+      "grad_norm": 0.35347217321395874,
+      "learning_rate": 0.00013083333333333333,
+      "loss": 1.3082,
+      "step": 337
+    },
+    {
+      "epoch": 1.7458064516129033,
+      "grad_norm": 0.33776408433914185,
+      "learning_rate": 0.000130625,
+      "loss": 1.2691,
+      "step": 338
+    },
+    {
+      "epoch": 1.7509677419354839,
+      "grad_norm": 0.33636176586151123,
+      "learning_rate": 0.00013041666666666667,
+      "loss": 1.3143,
+      "step": 339
+    },
+    {
+      "epoch": 1.7561290322580645,
+      "grad_norm": 0.3489387631416321,
+      "learning_rate": 0.00013020833333333333,
+      "loss": 1.3357,
+      "step": 340
+    },
+    {
+      "epoch": 1.761290322580645,
+      "grad_norm": 0.3498457968235016,
+      "learning_rate": 0.00013000000000000002,
+      "loss": 1.331,
+      "step": 341
+    },
+    {
+      "epoch": 1.766451612903226,
+      "grad_norm": 0.3666294515132904,
+      "learning_rate": 0.00012979166666666667,
+      "loss": 1.3642,
+      "step": 342
+    },
+    {
+      "epoch": 1.7716129032258063,
+      "grad_norm": 0.34596627950668335,
+      "learning_rate": 0.00012958333333333336,
+      "loss": 1.3275,
+      "step": 343
+    },
+    {
+      "epoch": 1.7767741935483872,
+      "grad_norm": 0.35660508275032043,
+      "learning_rate": 0.00012937500000000001,
+      "loss": 1.3291,
+      "step": 344
+    },
+    {
+      "epoch": 1.7819354838709678,
+      "grad_norm": 0.34041476249694824,
+      "learning_rate": 0.00012916666666666667,
+      "loss": 1.3316,
+      "step": 345
+    },
+    {
+      "epoch": 1.7870967741935484,
+      "grad_norm": 0.3493117690086365,
+      "learning_rate": 0.00012895833333333333,
+      "loss": 1.2594,
+      "step": 346
+    },
+    {
+      "epoch": 1.792258064516129,
+      "grad_norm": 0.35230782628059387,
+      "learning_rate": 0.00012875,
+      "loss": 1.3184,
+      "step": 347
+    },
+    {
+      "epoch": 1.7974193548387096,
+      "grad_norm": 0.37954822182655334,
+      "learning_rate": 0.00012854166666666667,
+      "loss": 1.2128,
+      "step": 348
+    },
+    {
+      "epoch": 1.8025806451612905,
+      "grad_norm": 0.3440784811973572,
+      "learning_rate": 0.00012833333333333335,
+      "loss": 1.324,
+      "step": 349
+    },
+    {
+      "epoch": 1.8077419354838709,
+      "grad_norm": 0.3344900608062744,
+      "learning_rate": 0.000128125,
+      "loss": 1.3471,
+      "step": 350
+    },
+    {
+      "epoch": 1.8129032258064517,
+      "grad_norm": 0.33594515919685364,
+      "learning_rate": 0.00012791666666666667,
+      "loss": 1.3139,
+      "step": 351
+    },
+    {
+      "epoch": 1.8180645161290323,
+      "grad_norm": 0.35635945200920105,
+      "learning_rate": 0.00012770833333333335,
+      "loss": 1.3769,
+      "step": 352
+    },
+    {
+      "epoch": 1.823225806451613,
+      "grad_norm": 0.4343617558479309,
+      "learning_rate": 0.0001275,
+      "loss": 1.3983,
+      "step": 353
+    },
+    {
+      "epoch": 1.8283870967741935,
+      "grad_norm": 0.3398101329803467,
+      "learning_rate": 0.00012729166666666667,
+      "loss": 1.3725,
+      "step": 354
+    },
+    {
+      "epoch": 1.8335483870967741,
+      "grad_norm": 0.3443765342235565,
+      "learning_rate": 0.00012708333333333332,
+      "loss": 1.3573,
+      "step": 355
+    },
+    {
+      "epoch": 1.838709677419355,
+      "grad_norm": 0.3410966694355011,
+      "learning_rate": 0.000126875,
+      "loss": 1.3391,
+      "step": 356
+    },
+    {
+      "epoch": 1.8438709677419354,
+      "grad_norm": 0.35767605900764465,
+      "learning_rate": 0.00012666666666666666,
+      "loss": 1.301,
+      "step": 357
+    },
+    {
+      "epoch": 1.8490322580645162,
+      "grad_norm": 0.39232590794563293,
+      "learning_rate": 0.00012645833333333335,
+      "loss": 1.3339,
+      "step": 358
+    },
+    {
+      "epoch": 1.8541935483870968,
+      "grad_norm": 0.3309646248817444,
+      "learning_rate": 0.00012625,
+      "loss": 1.3052,
+      "step": 359
+    },
+    {
+      "epoch": 1.8593548387096774,
+      "grad_norm": 0.37598055601119995,
+      "learning_rate": 0.0001260416666666667,
+      "loss": 1.3024,
+      "step": 360
+    },
+    {
+      "epoch": 1.864516129032258,
+      "grad_norm": 0.3436594307422638,
+      "learning_rate": 0.00012583333333333335,
+      "loss": 1.2505,
+      "step": 361
+    },
+    {
+      "epoch": 1.8696774193548387,
+      "grad_norm": 0.41823405027389526,
+      "learning_rate": 0.000125625,
+      "loss": 1.2629,
+      "step": 362
+    },
+    {
+      "epoch": 1.8748387096774195,
+      "grad_norm": 0.3592495918273926,
+      "learning_rate": 0.00012541666666666666,
+      "loss": 1.3821,
+      "step": 363
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.3441467583179474,
+      "learning_rate": 0.00012520833333333335,
+      "loss": 1.3711,
+      "step": 364
+    },
+    {
+      "epoch": 1.8851612903225807,
+      "grad_norm": 0.3563661575317383,
+      "learning_rate": 0.000125,
+      "loss": 1.3406,
+      "step": 365
+    },
+    {
+      "epoch": 1.8903225806451613,
+      "grad_norm": 0.35218170285224915,
+      "learning_rate": 0.00012479166666666666,
+      "loss": 1.3675,
+      "step": 366
+    },
+    {
+      "epoch": 1.895483870967742,
+      "grad_norm": 0.3334798216819763,
+      "learning_rate": 0.00012458333333333334,
+      "loss": 1.328,
+      "step": 367
+    },
+    {
+      "epoch": 1.9006451612903226,
+      "grad_norm": 0.3432958424091339,
+      "learning_rate": 0.000124375,
+      "loss": 1.3596,
+      "step": 368
+    },
+    {
+      "epoch": 1.9058064516129032,
+      "grad_norm": 0.35067862272262573,
+      "learning_rate": 0.00012416666666666669,
+      "loss": 1.3289,
+      "step": 369
+    },
+    {
+      "epoch": 1.910967741935484,
+      "grad_norm": 0.3745068609714508,
+      "learning_rate": 0.00012395833333333334,
+      "loss": 1.2062,
+      "step": 370
+    },
+    {
+      "epoch": 1.9161290322580644,
+      "grad_norm": 0.3239261209964752,
+      "learning_rate": 0.00012375,
+      "loss": 1.3046,
+      "step": 371
+    },
+    {
+      "epoch": 1.9212903225806452,
+      "grad_norm": 0.3422105014324188,
+      "learning_rate": 0.00012354166666666666,
+      "loss": 1.2741,
+      "step": 372
+    },
+    {
+      "epoch": 1.9264516129032256,
+      "grad_norm": 0.3350420296192169,
+      "learning_rate": 0.00012333333333333334,
+      "loss": 1.3163,
+      "step": 373
+    },
+    {
+      "epoch": 1.9316129032258065,
+      "grad_norm": 0.33306682109832764,
+      "learning_rate": 0.000123125,
+      "loss": 1.3085,
+      "step": 374
+    },
+    {
+      "epoch": 1.936774193548387,
+      "grad_norm": 0.3442312777042389,
+      "learning_rate": 0.00012291666666666668,
+      "loss": 1.3813,
+      "step": 375
+    },
+    {
+      "epoch": 1.9419354838709677,
+      "grad_norm": 0.4310455918312073,
+      "learning_rate": 0.00012270833333333334,
+      "loss": 1.2402,
+      "step": 376
+    },
+    {
+      "epoch": 1.9470967741935485,
+      "grad_norm": 0.37676361203193665,
+      "learning_rate": 0.00012250000000000002,
+      "loss": 1.2373,
+      "step": 377
+    },
+    {
+      "epoch": 1.952258064516129,
+      "grad_norm": 0.3456939458847046,
+      "learning_rate": 0.00012229166666666668,
+      "loss": 1.2852,
+      "step": 378
+    },
+    {
+      "epoch": 1.9574193548387098,
+      "grad_norm": 0.3331674635410309,
+      "learning_rate": 0.00012208333333333334,
+      "loss": 1.3006,
+      "step": 379
+    },
+    {
+      "epoch": 1.9625806451612902,
+      "grad_norm": 0.35660937428474426,
+      "learning_rate": 0.00012187500000000001,
+      "loss": 1.3282,
+      "step": 380
+    },
+    {
+      "epoch": 1.967741935483871,
+      "grad_norm": 0.34395772218704224,
+      "learning_rate": 0.00012166666666666667,
+      "loss": 1.2875,
+      "step": 381
+    },
+    {
+      "epoch": 1.9729032258064516,
+      "grad_norm": 0.42433151602745056,
+      "learning_rate": 0.00012145833333333334,
+      "loss": 1.4105,
+      "step": 382
+    },
+    {
+      "epoch": 1.9780645161290322,
+      "grad_norm": 0.33939099311828613,
+      "learning_rate": 0.00012124999999999999,
+      "loss": 1.3212,
+      "step": 383
+    },
+    {
+      "epoch": 1.983225806451613,
+      "grad_norm": 0.3406873643398285,
+      "learning_rate": 0.00012104166666666668,
+      "loss": 1.312,
+      "step": 384
+    },
+    {
+      "epoch": 1.9883870967741935,
+      "grad_norm": 0.3459448218345642,
+      "learning_rate": 0.00012083333333333333,
+      "loss": 1.2441,
+      "step": 385
+    },
+    {
+      "epoch": 1.9935483870967743,
+      "grad_norm": 0.33860331773757935,
+      "learning_rate": 0.000120625,
+      "loss": 1.3732,
+      "step": 386
+    },
+    {
+      "epoch": 1.9987096774193547,
+      "grad_norm": 0.5217887759208679,
+      "learning_rate": 0.00012041666666666666,
+      "loss": 2.0209,
+      "step": 387
+    },
+    {
+      "epoch": 2.005161290322581,
+      "grad_norm": 0.422323614358902,
+      "learning_rate": 0.00012020833333333335,
+      "loss": 1.0769,
+      "step": 388
+    },
+    {
+      "epoch": 2.0103225806451612,
+      "grad_norm": 0.41807821393013,
+      "learning_rate": 0.00012,
+      "loss": 1.0968,
+      "step": 389
+    },
+    {
+      "epoch": 2.015483870967742,
+      "grad_norm": 0.42648643255233765,
+      "learning_rate": 0.00011979166666666667,
+      "loss": 1.0903,
+      "step": 390
+    },
+    {
+      "epoch": 2.0206451612903225,
+      "grad_norm": 0.5991876125335693,
+      "learning_rate": 0.00011958333333333333,
+      "loss": 1.1449,
+      "step": 391
+    },
+    {
+      "epoch": 2.0258064516129033,
+      "grad_norm": 0.585296630859375,
+      "learning_rate": 0.00011937500000000001,
+      "loss": 1.0182,
+      "step": 392
+    },
+    {
+      "epoch": 2.0309677419354837,
+      "grad_norm": 0.4932289719581604,
+      "learning_rate": 0.00011916666666666667,
+      "loss": 1.0938,
+      "step": 393
+    },
+    {
+      "epoch": 2.0361290322580645,
+      "grad_norm": 0.4801351726055145,
+      "learning_rate": 0.00011895833333333334,
+      "loss": 0.9956,
+      "step": 394
+    },
+    {
+      "epoch": 2.0412903225806454,
+      "grad_norm": 0.47499120235443115,
+      "learning_rate": 0.00011875,
+      "loss": 1.0633,
+      "step": 395
+    },
+    {
+      "epoch": 2.0464516129032257,
+      "grad_norm": 0.5020606517791748,
+      "learning_rate": 0.00011854166666666666,
+      "loss": 0.9862,
+      "step": 396
+    },
+    {
+      "epoch": 2.0516129032258066,
+      "grad_norm": 0.48372411727905273,
+      "learning_rate": 0.00011833333333333334,
+      "loss": 1.0675,
+      "step": 397
+    },
+    {
+      "epoch": 2.056774193548387,
+      "grad_norm": 1.8128656148910522,
+      "learning_rate": 0.000118125,
+      "loss": 1.0993,
+      "step": 398
+    },
+    {
+      "epoch": 2.061935483870968,
+      "grad_norm": 0.5115732550621033,
+      "learning_rate": 0.00011791666666666667,
+      "loss": 1.0833,
+      "step": 399
+    },
+    {
+      "epoch": 2.067096774193548,
+      "grad_norm": 0.46960705518722534,
+      "learning_rate": 0.00011770833333333333,
+      "loss": 1.0206,
+      "step": 400
+    },
+    {
+      "epoch": 2.072258064516129,
+      "grad_norm": 0.5921757817268372,
+      "learning_rate": 0.00011750000000000001,
+      "loss": 1.1173,
+      "step": 401
+    },
+    {
+      "epoch": 2.07741935483871,
+      "grad_norm": 0.4667699635028839,
+      "learning_rate": 0.00011729166666666667,
+      "loss": 1.1177,
+      "step": 402
+    },
+    {
+      "epoch": 2.0825806451612903,
+      "grad_norm": 0.4945167899131775,
+      "learning_rate": 0.00011708333333333335,
+      "loss": 1.1264,
+      "step": 403
+    },
+    {
+      "epoch": 2.087741935483871,
+      "grad_norm": 0.5206171274185181,
+      "learning_rate": 0.000116875,
+      "loss": 1.078,
+      "step": 404
+    },
+    {
+      "epoch": 2.0929032258064515,
+      "grad_norm": 0.50333172082901,
+      "learning_rate": 0.00011666666666666668,
+      "loss": 1.0803,
+      "step": 405
+    },
+    {
+      "epoch": 2.0980645161290323,
+      "grad_norm": 0.49614614248275757,
+      "learning_rate": 0.00011645833333333334,
+      "loss": 1.1437,
+      "step": 406
+    },
+    {
+      "epoch": 2.1032258064516127,
+      "grad_norm": 0.4997609257698059,
+      "learning_rate": 0.00011625000000000002,
+      "loss": 1.0346,
+      "step": 407
+    },
+    {
+      "epoch": 2.1083870967741936,
+      "grad_norm": 0.5245897769927979,
+      "learning_rate": 0.00011604166666666668,
+      "loss": 1.0928,
+      "step": 408
+    },
+    {
+      "epoch": 2.1135483870967744,
+      "grad_norm": 0.5079848766326904,
+      "learning_rate": 0.00011583333333333335,
+      "loss": 1.0785,
+      "step": 409
+    },
+    {
+      "epoch": 2.118709677419355,
+      "grad_norm": 0.5305972099304199,
+      "learning_rate": 0.000115625,
+      "loss": 1.0578,
+      "step": 410
+    },
+    {
+      "epoch": 2.1238709677419356,
+      "grad_norm": 0.6335098743438721,
+      "learning_rate": 0.00011541666666666666,
+      "loss": 1.0933,
+      "step": 411
+    },
+    {
+      "epoch": 2.129032258064516,
+      "grad_norm": 0.5254077315330505,
+      "learning_rate": 0.00011520833333333335,
+      "loss": 1.153,
+      "step": 412
+    },
+    {
+      "epoch": 2.134193548387097,
+      "grad_norm": 0.5157703161239624,
+      "learning_rate": 0.00011499999999999999,
+      "loss": 1.1176,
+      "step": 413
+    },
+    {
+      "epoch": 2.1393548387096772,
+      "grad_norm": 0.5097150802612305,
+      "learning_rate": 0.00011479166666666667,
+      "loss": 1.1314,
+      "step": 414
+    },
+    {
+      "epoch": 2.144516129032258,
+      "grad_norm": 0.5170332193374634,
+      "learning_rate": 0.00011458333333333333,
+      "loss": 1.0193,
+      "step": 415
+    },
+    {
+      "epoch": 2.149677419354839,
+      "grad_norm": 0.5179755091667175,
+      "learning_rate": 0.00011437500000000002,
+      "loss": 1.1449,
+      "step": 416
+    },
+    {
+      "epoch": 2.1548387096774193,
+      "grad_norm": 0.5507645606994629,
+      "learning_rate": 0.00011416666666666667,
+      "loss": 1.1232,
+      "step": 417
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.5016111135482788,
+      "learning_rate": 0.00011395833333333334,
+      "loss": 1.1816,
+      "step": 418
+    },
+    {
+      "epoch": 2.1651612903225805,
+      "grad_norm": 0.5098230838775635,
+      "learning_rate": 0.00011375,
+      "loss": 1.0991,
+      "step": 419
+    },
+    {
+      "epoch": 2.1703225806451614,
+      "grad_norm": 0.5694328546524048,
+      "learning_rate": 0.00011354166666666668,
+      "loss": 1.0125,
+      "step": 420
+    },
+    {
+      "epoch": 2.1754838709677418,
+      "grad_norm": 0.529375433921814,
+      "learning_rate": 0.00011333333333333334,
+      "loss": 1.0825,
+      "step": 421
+    },
+    {
+      "epoch": 2.1806451612903226,
+      "grad_norm": 0.5008748173713684,
+      "learning_rate": 0.00011312500000000001,
+      "loss": 1.0784,
+      "step": 422
+    },
+    {
+      "epoch": 2.1858064516129034,
+      "grad_norm": 0.5480524897575378,
+      "learning_rate": 0.00011291666666666667,
+      "loss": 1.1233,
+      "step": 423
+    },
+    {
+      "epoch": 2.190967741935484,
+      "grad_norm": 0.6432352662086487,
+      "learning_rate": 0.00011270833333333335,
+      "loss": 1.0234,
+      "step": 424
+    },
+    {
+      "epoch": 2.1961290322580647,
+      "grad_norm": 0.5089468359947205,
+      "learning_rate": 0.00011250000000000001,
+      "loss": 1.1064,
+      "step": 425
+    },
+    {
+      "epoch": 2.201290322580645,
+      "grad_norm": 0.4957018494606018,
+      "learning_rate": 0.00011229166666666667,
+      "loss": 1.0461,
+      "step": 426
+    },
+    {
+      "epoch": 2.206451612903226,
+      "grad_norm": 0.49691277742385864,
+      "learning_rate": 0.00011208333333333334,
+      "loss": 1.0221,
+      "step": 427
+    },
+    {
+      "epoch": 2.2116129032258063,
+      "grad_norm": 0.5094979405403137,
+      "learning_rate": 0.000111875,
+      "loss": 1.0683,
+      "step": 428
+    },
+    {
+      "epoch": 2.216774193548387,
+      "grad_norm": 0.5043177604675293,
+      "learning_rate": 0.00011166666666666668,
+      "loss": 1.0716,
+      "step": 429
+    },
+    {
+      "epoch": 2.2219354838709675,
+      "grad_norm": 0.5305630564689636,
+      "learning_rate": 0.00011145833333333334,
+      "loss": 1.0906,
+      "step": 430
+    },
+    {
+      "epoch": 2.2270967741935483,
+      "grad_norm": 0.5283196568489075,
+      "learning_rate": 0.00011125000000000001,
+      "loss": 1.0378,
+      "step": 431
+    },
+    {
+      "epoch": 2.232258064516129,
+      "grad_norm": 0.5894606709480286,
+      "learning_rate": 0.00011104166666666666,
+      "loss": 1.0964,
+      "step": 432
+    },
+    {
+      "epoch": 2.2374193548387096,
+      "grad_norm": 0.5202419757843018,
+      "learning_rate": 0.00011083333333333335,
+      "loss": 1.1171,
+      "step": 433
+    },
+    {
+      "epoch": 2.2425806451612904,
+      "grad_norm": 0.530718982219696,
+      "learning_rate": 0.000110625,
+      "loss": 1.0768,
+      "step": 434
+    },
+    {
+      "epoch": 2.247741935483871,
+      "grad_norm": 0.4953882098197937,
+      "learning_rate": 0.00011041666666666668,
+      "loss": 1.073,
+      "step": 435
+    },
+    {
+      "epoch": 2.2529032258064516,
+      "grad_norm": 0.531024158000946,
+      "learning_rate": 0.00011020833333333333,
+      "loss": 1.0476,
+      "step": 436
+    },
+    {
+      "epoch": 2.258064516129032,
+      "grad_norm": 0.530388593673706,
+      "learning_rate": 0.00011000000000000002,
+      "loss": 1.1418,
+      "step": 437
+    },
+    {
+      "epoch": 2.263225806451613,
+      "grad_norm": 0.5266593098640442,
+      "learning_rate": 0.00010979166666666667,
+      "loss": 1.0567,
+      "step": 438
+    },
+    {
+      "epoch": 2.2683870967741937,
+      "grad_norm": 0.5251802206039429,
+      "learning_rate": 0.00010958333333333335,
+      "loss": 1.1248,
+      "step": 439
+    },
+    {
+      "epoch": 2.273548387096774,
+      "grad_norm": 0.5153113603591919,
+      "learning_rate": 0.000109375,
+      "loss": 1.0723,
+      "step": 440
+    },
+    {
+      "epoch": 2.278709677419355,
+      "grad_norm": 0.5247308611869812,
+      "learning_rate": 0.00010916666666666666,
+      "loss": 1.1295,
+      "step": 441
+    },
+    {
+      "epoch": 2.2838709677419353,
+      "grad_norm": 0.5306227803230286,
+      "learning_rate": 0.00010895833333333334,
+      "loss": 1.1562,
+      "step": 442
+    },
+    {
+      "epoch": 2.289032258064516,
+      "grad_norm": 0.5346123576164246,
+      "learning_rate": 0.00010875,
+      "loss": 1.145,
+      "step": 443
+    },
+    {
+      "epoch": 2.2941935483870965,
+      "grad_norm": 0.5081747770309448,
+      "learning_rate": 0.00010854166666666667,
+      "loss": 1.1314,
+      "step": 444
+    },
+    {
+      "epoch": 2.2993548387096774,
+      "grad_norm": 0.524278461933136,
+      "learning_rate": 0.00010833333333333333,
+      "loss": 1.0951,
+      "step": 445
+    },
+    {
+      "epoch": 2.304516129032258,
+      "grad_norm": 0.5242152810096741,
+      "learning_rate": 0.00010812500000000001,
+      "loss": 1.0702,
+      "step": 446
+    },
+    {
+      "epoch": 2.3096774193548386,
+      "grad_norm": 0.5393962860107422,
+      "learning_rate": 0.00010791666666666667,
+      "loss": 1.1569,
+      "step": 447
+    },
+    {
+      "epoch": 2.3148387096774194,
+      "grad_norm": 0.5256329774856567,
+      "learning_rate": 0.00010770833333333334,
+      "loss": 1.1024,
+      "step": 448
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.5485777258872986,
+      "learning_rate": 0.0001075,
+      "loss": 1.1344,
+      "step": 449
+    },
+    {
+      "epoch": 2.3251612903225807,
+      "grad_norm": 0.5174791812896729,
+      "learning_rate": 0.00010729166666666668,
+      "loss": 1.1219,
+      "step": 450
+    },
+    {
+      "epoch": 2.330322580645161,
+      "grad_norm": 0.5283759832382202,
+      "learning_rate": 0.00010708333333333334,
+      "loss": 1.0838,
+      "step": 451
+    },
+    {
+      "epoch": 2.335483870967742,
+      "grad_norm": 0.5164680480957031,
+      "learning_rate": 0.00010687500000000001,
+      "loss": 1.1595,
+      "step": 452
+    },
+    {
+      "epoch": 2.3406451612903227,
+      "grad_norm": 0.5320565700531006,
+      "learning_rate": 0.00010666666666666667,
+      "loss": 1.1161,
+      "step": 453
+    },
+    {
+      "epoch": 2.345806451612903,
+      "grad_norm": 0.5368756651878357,
+      "learning_rate": 0.00010645833333333335,
+      "loss": 1.1101,
+      "step": 454
+    },
+    {
+      "epoch": 2.350967741935484,
+      "grad_norm": 0.5115728378295898,
+      "learning_rate": 0.00010625000000000001,
+      "loss": 1.052,
+      "step": 455
+    },
+    {
+      "epoch": 2.3561290322580644,
+      "grad_norm": 0.5163736939430237,
+      "learning_rate": 0.00010604166666666666,
+      "loss": 1.1405,
+      "step": 456
+    },
+    {
+      "epoch": 2.361290322580645,
+      "grad_norm": 0.5361839532852173,
+      "learning_rate": 0.00010583333333333334,
+      "loss": 1.0902,
+      "step": 457
+    },
+    {
+      "epoch": 2.3664516129032256,
+      "grad_norm": 0.5226168036460876,
+      "learning_rate": 0.00010562499999999999,
+      "loss": 1.0733,
+      "step": 458
+    },
+    {
+      "epoch": 2.3716129032258064,
+      "grad_norm": 0.8021620512008667,
+      "learning_rate": 0.00010541666666666668,
+      "loss": 1.0413,
+      "step": 459
+    },
+    {
+      "epoch": 2.3767741935483873,
+      "grad_norm": 0.5543711185455322,
+      "learning_rate": 0.00010520833333333333,
+      "loss": 1.0811,
+      "step": 460
+    },
+    {
+      "epoch": 2.3819354838709677,
+      "grad_norm": 0.5531166791915894,
+      "learning_rate": 0.000105,
+      "loss": 1.1064,
+      "step": 461
+    },
+    {
+      "epoch": 2.3870967741935485,
+      "grad_norm": 0.5170143842697144,
+      "learning_rate": 0.00010479166666666666,
+      "loss": 1.0146,
+      "step": 462
+    },
+    {
+      "epoch": 2.392258064516129,
+      "grad_norm": 0.5793948173522949,
+      "learning_rate": 0.00010458333333333335,
+      "loss": 1.0394,
+      "step": 463
+    },
+    {
+      "epoch": 2.3974193548387097,
+      "grad_norm": 0.5288783311843872,
+      "learning_rate": 0.000104375,
+      "loss": 1.056,
+      "step": 464
+    },
+    {
+      "epoch": 2.40258064516129,
+      "grad_norm": 0.5502579808235168,
+      "learning_rate": 0.00010416666666666667,
+      "loss": 0.97,
+      "step": 465
+    },
+    {
+      "epoch": 2.407741935483871,
+      "grad_norm": 0.5317333936691284,
+      "learning_rate": 0.00010395833333333333,
+      "loss": 1.0706,
+      "step": 466
+    },
+    {
+      "epoch": 2.412903225806452,
+      "grad_norm": 0.5385299324989319,
+      "learning_rate": 0.00010375000000000001,
+      "loss": 1.1993,
+      "step": 467
+    },
+    {
+      "epoch": 2.418064516129032,
+      "grad_norm": 0.6048040986061096,
+      "learning_rate": 0.00010354166666666667,
+      "loss": 1.063,
+      "step": 468
+    },
+    {
+      "epoch": 2.423225806451613,
+      "grad_norm": 0.5092827081680298,
+      "learning_rate": 0.00010333333333333334,
+      "loss": 1.0326,
+      "step": 469
+    },
+    {
+      "epoch": 2.4283870967741934,
+      "grad_norm": 0.5229476094245911,
+      "learning_rate": 0.000103125,
+      "loss": 1.1049,
+      "step": 470
+    },
+    {
+      "epoch": 2.4335483870967742,
+      "grad_norm": 0.5445364117622375,
+      "learning_rate": 0.00010291666666666666,
+      "loss": 1.0951,
+      "step": 471
+    },
+    {
+      "epoch": 2.4387096774193546,
+      "grad_norm": 0.5694670081138611,
+      "learning_rate": 0.00010270833333333334,
+      "loss": 1.0766,
+      "step": 472
+    },
+    {
+      "epoch": 2.4438709677419355,
+      "grad_norm": 0.5460540652275085,
+      "learning_rate": 0.0001025,
+      "loss": 1.0783,
+      "step": 473
+    },
+    {
+      "epoch": 2.4490322580645163,
+      "grad_norm": 0.5443238019943237,
+      "learning_rate": 0.00010229166666666667,
+      "loss": 1.0425,
+      "step": 474
+    },
+    {
+      "epoch": 2.4541935483870967,
+      "grad_norm": 0.5531854033470154,
+      "learning_rate": 0.00010208333333333333,
+      "loss": 1.116,
+      "step": 475
+    },
+    {
+      "epoch": 2.4593548387096775,
+      "grad_norm": 0.8330934643745422,
+      "learning_rate": 0.00010187500000000001,
+      "loss": 1.1445,
+      "step": 476
+    },
+    {
+      "epoch": 2.464516129032258,
+      "grad_norm": 0.5395287871360779,
+      "learning_rate": 0.00010166666666666667,
+      "loss": 1.0777,
+      "step": 477
+    },
+    {
+      "epoch": 2.4696774193548388,
+      "grad_norm": 0.5552036166191101,
+      "learning_rate": 0.00010145833333333334,
+      "loss": 1.173,
+      "step": 478
+    },
+    {
+      "epoch": 2.474838709677419,
+      "grad_norm": 0.5612345337867737,
+      "learning_rate": 0.00010125,
+      "loss": 1.1134,
+      "step": 479
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.5283383131027222,
+      "learning_rate": 0.00010104166666666668,
+      "loss": 1.1286,
+      "step": 480
+    },
+    {
+      "epoch": 2.485161290322581,
+      "grad_norm": 0.5075555443763733,
+      "learning_rate": 0.00010083333333333334,
+      "loss": 1.1266,
+      "step": 481
+    },
+    {
+      "epoch": 2.490322580645161,
+      "grad_norm": 0.5310274958610535,
+      "learning_rate": 0.00010062500000000002,
+      "loss": 1.1166,
+      "step": 482
+    },
+    {
+      "epoch": 2.495483870967742,
+      "grad_norm": 0.533200740814209,
+      "learning_rate": 0.00010041666666666666,
+      "loss": 1.1343,
+      "step": 483
+    },
+    {
+      "epoch": 2.5006451612903224,
+      "grad_norm": 0.5423817038536072,
+      "learning_rate": 0.00010020833333333335,
+      "loss": 1.1397,
+      "step": 484
+    },
+    {
+      "epoch": 2.5058064516129033,
+      "grad_norm": 0.5147575736045837,
+      "learning_rate": 0.0001,
+      "loss": 1.0116,
+      "step": 485
+    },
+    {
+      "epoch": 2.5109677419354837,
+      "grad_norm": 0.520893394947052,
+      "learning_rate": 9.979166666666668e-05,
+      "loss": 1.1173,
+      "step": 486
+    },
+    {
+      "epoch": 2.5161290322580645,
+      "grad_norm": 8.175589561462402,
+      "learning_rate": 9.958333333333335e-05,
+      "loss": 1.0945,
+      "step": 487
+    },
+    {
+      "epoch": 2.5212903225806453,
+      "grad_norm": 0.5501712560653687,
+      "learning_rate": 9.9375e-05,
+      "loss": 0.9821,
+      "step": 488
+    },
+    {
+      "epoch": 2.5264516129032257,
+      "grad_norm": 0.5551290512084961,
+      "learning_rate": 9.916666666666667e-05,
+      "loss": 1.1304,
+      "step": 489
+    },
+    {
+      "epoch": 2.5316129032258066,
+      "grad_norm": 0.5605106353759766,
+      "learning_rate": 9.895833333333334e-05,
+      "loss": 1.1542,
+      "step": 490
+    },
+    {
+      "epoch": 2.536774193548387,
+      "grad_norm": 0.5552902221679688,
+      "learning_rate": 9.875000000000002e-05,
+      "loss": 1.0832,
+      "step": 491
+    },
+    {
+      "epoch": 2.541935483870968,
+      "grad_norm": 0.5530599355697632,
+      "learning_rate": 9.854166666666667e-05,
+      "loss": 1.1286,
+      "step": 492
+    },
+    {
+      "epoch": 2.547096774193548,
+      "grad_norm": 0.5568006038665771,
+      "learning_rate": 9.833333333333333e-05,
+      "loss": 1.1299,
+      "step": 493
+    },
+    {
+      "epoch": 2.552258064516129,
+      "grad_norm": 0.5466172099113464,
+      "learning_rate": 9.8125e-05,
+      "loss": 1.1088,
+      "step": 494
+    },
+    {
+      "epoch": 2.55741935483871,
+      "grad_norm": 0.5217697620391846,
+      "learning_rate": 9.791666666666667e-05,
+      "loss": 1.0772,
+      "step": 495
+    },
+    {
+      "epoch": 2.5625806451612902,
+      "grad_norm": 0.5277625322341919,
+      "learning_rate": 9.770833333333334e-05,
+      "loss": 1.0664,
+      "step": 496
+    },
+    {
+      "epoch": 2.567741935483871,
+      "grad_norm": 0.6777426600456238,
+      "learning_rate": 9.75e-05,
+      "loss": 1.0871,
+      "step": 497
+    },
+    {
+      "epoch": 2.5729032258064515,
+      "grad_norm": 0.6249215006828308,
+      "learning_rate": 9.729166666666667e-05,
+      "loss": 1.1147,
+      "step": 498
+    },
+    {
+      "epoch": 2.5780645161290323,
+      "grad_norm": 0.5539170503616333,
+      "learning_rate": 9.708333333333334e-05,
+      "loss": 1.0683,
+      "step": 499
+    },
+    {
+      "epoch": 2.5832258064516127,
+      "grad_norm": 0.5429412126541138,
+      "learning_rate": 9.687500000000001e-05,
+      "loss": 1.1383,
+      "step": 500
+    },
+    {
+      "epoch": 2.5883870967741935,
+      "grad_norm": 0.537137508392334,
+      "learning_rate": 9.666666666666667e-05,
+      "loss": 1.1763,
+      "step": 501
+    },
+    {
+      "epoch": 2.5935483870967744,
+      "grad_norm": 0.5279580950737,
+      "learning_rate": 9.645833333333334e-05,
+      "loss": 1.0791,
+      "step": 502
+    },
+    {
+      "epoch": 2.5987096774193548,
+      "grad_norm": 0.535622239112854,
+      "learning_rate": 9.625000000000001e-05,
+      "loss": 1.1742,
+      "step": 503
+    },
+    {
+      "epoch": 2.6038709677419356,
+      "grad_norm": 0.5258580446243286,
+      "learning_rate": 9.604166666666668e-05,
+      "loss": 1.0989,
+      "step": 504
+    },
+    {
+      "epoch": 2.609032258064516,
+      "grad_norm": 0.5222040414810181,
+      "learning_rate": 9.583333333333334e-05,
+      "loss": 1.1326,
+      "step": 505
+    },
+    {
+      "epoch": 2.614193548387097,
+      "grad_norm": 0.6634904742240906,
+      "learning_rate": 9.562500000000001e-05,
+      "loss": 1.1134,
+      "step": 506
+    },
+    {
+      "epoch": 2.6193548387096772,
+      "grad_norm": 0.5135465860366821,
+      "learning_rate": 9.541666666666668e-05,
+      "loss": 1.08,
+      "step": 507
+    },
+    {
+      "epoch": 2.624516129032258,
+      "grad_norm": 0.5633925795555115,
+      "learning_rate": 9.520833333333333e-05,
+      "loss": 1.1251,
+      "step": 508
+    },
+    {
+      "epoch": 2.629677419354839,
+      "grad_norm": 0.5304936170578003,
+      "learning_rate": 9.5e-05,
+      "loss": 1.1261,
+      "step": 509
+    },
+    {
+      "epoch": 2.6348387096774193,
+      "grad_norm": 0.5506126880645752,
+      "learning_rate": 9.479166666666666e-05,
+      "loss": 1.1066,
+      "step": 510
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.6438670754432678,
+      "learning_rate": 9.458333333333333e-05,
+      "loss": 0.9764,
+      "step": 511
+    },
+    {
+      "epoch": 2.6451612903225805,
+      "grad_norm": 0.5361579656600952,
+      "learning_rate": 9.4375e-05,
+      "loss": 1.0869,
+      "step": 512
+    },
+    {
+      "epoch": 2.6503225806451614,
+      "grad_norm": 0.5298143029212952,
+      "learning_rate": 9.416666666666667e-05,
+      "loss": 1.1291,
+      "step": 513
+    },
+    {
+      "epoch": 2.6554838709677417,
+      "grad_norm": 0.524237334728241,
+      "learning_rate": 9.395833333333333e-05,
+      "loss": 1.1035,
+      "step": 514
+    },
+    {
+      "epoch": 2.6606451612903226,
+      "grad_norm": 0.5221346616744995,
+      "learning_rate": 9.375e-05,
+      "loss": 1.0087,
+      "step": 515
+    },
+    {
+      "epoch": 2.6658064516129034,
+      "grad_norm": 0.6549589037895203,
+      "learning_rate": 9.354166666666667e-05,
+      "loss": 1.0316,
+      "step": 516
+    },
+    {
+      "epoch": 2.670967741935484,
+      "grad_norm": 0.5298157930374146,
+      "learning_rate": 9.333333333333334e-05,
+      "loss": 1.1312,
+      "step": 517
+    },
+    {
+      "epoch": 2.6761290322580646,
+      "grad_norm": 0.5280676484107971,
+      "learning_rate": 9.3125e-05,
+      "loss": 1.1342,
+      "step": 518
+    },
+    {
+      "epoch": 2.681290322580645,
+      "grad_norm": 0.5192847847938538,
+      "learning_rate": 9.291666666666667e-05,
+      "loss": 1.1058,
+      "step": 519
+    },
+    {
+      "epoch": 2.686451612903226,
+      "grad_norm": 0.5235516428947449,
+      "learning_rate": 9.270833333333334e-05,
+      "loss": 1.0251,
+      "step": 520
+    },
+    {
+      "epoch": 2.6916129032258063,
+      "grad_norm": 0.5653045177459717,
+      "learning_rate": 9.250000000000001e-05,
+      "loss": 1.166,
+      "step": 521
+    },
+    {
+      "epoch": 2.696774193548387,
+      "grad_norm": 0.532331645488739,
+      "learning_rate": 9.229166666666668e-05,
+      "loss": 1.0877,
+      "step": 522
+    },
+    {
+      "epoch": 2.701935483870968,
+      "grad_norm": 0.5391806960105896,
+      "learning_rate": 9.208333333333333e-05,
+      "loss": 1.1116,
+      "step": 523
+    },
+    {
+      "epoch": 2.7070967741935483,
+      "grad_norm": 0.5263025164604187,
+      "learning_rate": 9.1875e-05,
+      "loss": 1.0716,
+      "step": 524
+    },
+    {
+      "epoch": 2.712258064516129,
+      "grad_norm": 0.5362182855606079,
+      "learning_rate": 9.166666666666667e-05,
+      "loss": 1.1132,
+      "step": 525
+    },
+    {
+      "epoch": 2.7174193548387096,
+      "grad_norm": 0.5260269045829773,
+      "learning_rate": 9.145833333333334e-05,
+      "loss": 1.1431,
+      "step": 526
+    },
+    {
+      "epoch": 2.7225806451612904,
+      "grad_norm": 0.5512535572052002,
+      "learning_rate": 9.125e-05,
+      "loss": 1.1082,
+      "step": 527
+    },
+    {
+      "epoch": 2.727741935483871,
+      "grad_norm": 0.5308423042297363,
+      "learning_rate": 9.104166666666667e-05,
+      "loss": 1.1022,
+      "step": 528
+    },
+    {
+      "epoch": 2.7329032258064516,
+      "grad_norm": 0.5448559522628784,
+      "learning_rate": 9.083333333333334e-05,
+      "loss": 1.1096,
+      "step": 529
+    },
+    {
+      "epoch": 2.7380645161290325,
+      "grad_norm": 0.5332733392715454,
+      "learning_rate": 9.062500000000001e-05,
+      "loss": 1.1599,
+      "step": 530
+    },
+    {
+      "epoch": 2.743225806451613,
+      "grad_norm": 0.541897714138031,
+      "learning_rate": 9.041666666666668e-05,
+      "loss": 1.1064,
+      "step": 531
+    },
+    {
+      "epoch": 2.7483870967741937,
+      "grad_norm": 0.556448757648468,
+      "learning_rate": 9.020833333333334e-05,
+      "loss": 1.1064,
+      "step": 532
+    },
+    {
+      "epoch": 2.753548387096774,
+      "grad_norm": 0.5427480340003967,
+      "learning_rate": 9e-05,
+      "loss": 1.0282,
+      "step": 533
+    },
+    {
+      "epoch": 2.758709677419355,
+      "grad_norm": 0.5297014117240906,
+      "learning_rate": 8.979166666666668e-05,
+      "loss": 1.1002,
+      "step": 534
+    },
+    {
+      "epoch": 2.7638709677419353,
+      "grad_norm": 0.5056309700012207,
+      "learning_rate": 8.958333333333335e-05,
+      "loss": 1.011,
+      "step": 535
+    },
+    {
+      "epoch": 2.769032258064516,
+      "grad_norm": 0.5249669551849365,
+      "learning_rate": 8.9375e-05,
+      "loss": 1.0477,
+      "step": 536
+    },
+    {
+      "epoch": 2.774193548387097,
+      "grad_norm": 0.543459415435791,
+      "learning_rate": 8.916666666666667e-05,
+      "loss": 1.1512,
+      "step": 537
+    },
+    {
+      "epoch": 2.7793548387096774,
+      "grad_norm": 0.5250842571258545,
+      "learning_rate": 8.895833333333333e-05,
+      "loss": 1.1466,
+      "step": 538
+    },
+    {
+      "epoch": 2.784516129032258,
+      "grad_norm": 0.5290008187294006,
+      "learning_rate": 8.875e-05,
+      "loss": 1.1048,
+      "step": 539
+    },
+    {
+      "epoch": 2.7896774193548386,
+      "grad_norm": 0.5875825881958008,
+      "learning_rate": 8.854166666666667e-05,
+      "loss": 1.1055,
+      "step": 540
+    },
+    {
+      "epoch": 2.7948387096774194,
+      "grad_norm": 0.5002789497375488,
+      "learning_rate": 8.833333333333333e-05,
+      "loss": 1.1314,
+      "step": 541
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.535954475402832,
+      "learning_rate": 8.8125e-05,
+      "loss": 1.1187,
+      "step": 542
+    },
+    {
+      "epoch": 2.8051612903225807,
+      "grad_norm": 0.5374142527580261,
+      "learning_rate": 8.791666666666667e-05,
+      "loss": 1.1367,
+      "step": 543
+    },
+    {
+      "epoch": 2.8103225806451615,
+      "grad_norm": 0.5354551672935486,
+      "learning_rate": 8.770833333333334e-05,
+      "loss": 1.0295,
+      "step": 544
+    },
+    {
+      "epoch": 2.815483870967742,
+      "grad_norm": 0.5401291847229004,
+      "learning_rate": 8.75e-05,
+      "loss": 1.1087,
+      "step": 545
+    },
+    {
+      "epoch": 2.8206451612903227,
+      "grad_norm": 0.5419206619262695,
+      "learning_rate": 8.729166666666667e-05,
+      "loss": 1.1592,
+      "step": 546
+    },
+    {
+      "epoch": 2.825806451612903,
+      "grad_norm": 0.6135550737380981,
+      "learning_rate": 8.708333333333334e-05,
+      "loss": 1.0896,
+      "step": 547
+    },
+    {
+      "epoch": 2.830967741935484,
+      "grad_norm": 0.5830162763595581,
+      "learning_rate": 8.687500000000001e-05,
+      "loss": 1.1634,
+      "step": 548
+    },
+    {
+      "epoch": 2.8361290322580643,
+      "grad_norm": 0.5492197871208191,
+      "learning_rate": 8.666666666666667e-05,
+      "loss": 1.1052,
+      "step": 549
+    },
+    {
+      "epoch": 2.841290322580645,
+      "grad_norm": 0.5553468465805054,
+      "learning_rate": 8.645833333333334e-05,
+      "loss": 1.0798,
+      "step": 550
+    },
+    {
+      "epoch": 2.846451612903226,
+      "grad_norm": 0.5313032865524292,
+      "learning_rate": 8.625000000000001e-05,
+      "loss": 1.1752,
+      "step": 551
+    },
+    {
+      "epoch": 2.8516129032258064,
+      "grad_norm": 0.5165290832519531,
+      "learning_rate": 8.604166666666668e-05,
+      "loss": 1.0594,
+      "step": 552
+    },
+    {
+      "epoch": 2.8567741935483872,
+      "grad_norm": 0.5348519682884216,
+      "learning_rate": 8.583333333333334e-05,
+      "loss": 1.1082,
+      "step": 553
+    },
+    {
+      "epoch": 2.8619354838709676,
+      "grad_norm": 0.5292428731918335,
+      "learning_rate": 8.5625e-05,
+      "loss": 1.1126,
+      "step": 554
+    },
+    {
+      "epoch": 2.8670967741935485,
+      "grad_norm": 0.5344215631484985,
+      "learning_rate": 8.541666666666666e-05,
+      "loss": 1.1077,
+      "step": 555
+    },
+    {
+      "epoch": 2.872258064516129,
+      "grad_norm": 0.5346764922142029,
+      "learning_rate": 8.520833333333334e-05,
+      "loss": 1.127,
+      "step": 556
+    },
+    {
+      "epoch": 2.8774193548387097,
+      "grad_norm": 0.5359936356544495,
+      "learning_rate": 8.5e-05,
+      "loss": 1.0589,
+      "step": 557
+    },
+    {
+      "epoch": 2.8825806451612905,
+      "grad_norm": 0.5293276309967041,
+      "learning_rate": 8.479166666666666e-05,
+      "loss": 1.1546,
+      "step": 558
+    },
+    {
+      "epoch": 2.887741935483871,
+      "grad_norm": 0.5357221961021423,
+      "learning_rate": 8.458333333333333e-05,
+      "loss": 1.1276,
+      "step": 559
+    },
+    {
+      "epoch": 2.8929032258064518,
+      "grad_norm": 0.7441781163215637,
+      "learning_rate": 8.4375e-05,
+      "loss": 1.1023,
+      "step": 560
+    },
+    {
+      "epoch": 2.898064516129032,
+      "grad_norm": 0.5698540210723877,
+      "learning_rate": 8.416666666666668e-05,
+      "loss": 1.0875,
+      "step": 561
+    },
+    {
+      "epoch": 2.903225806451613,
+      "grad_norm": 0.5905174016952515,
+      "learning_rate": 8.395833333333333e-05,
+      "loss": 1.1383,
+      "step": 562
+    },
+    {
+      "epoch": 2.9083870967741934,
+      "grad_norm": 0.5240707397460938,
+      "learning_rate": 8.375e-05,
+      "loss": 1.0927,
+      "step": 563
+    },
+    {
+      "epoch": 2.913548387096774,
+      "grad_norm": 0.555675745010376,
+      "learning_rate": 8.354166666666667e-05,
+      "loss": 1.0311,
+      "step": 564
+    },
+    {
+      "epoch": 2.918709677419355,
+      "grad_norm": 0.5716830492019653,
+      "learning_rate": 8.333333333333334e-05,
+      "loss": 1.0964,
+      "step": 565
+    },
+    {
+      "epoch": 2.9238709677419354,
+      "grad_norm": 0.5505912899971008,
+      "learning_rate": 8.312500000000001e-05,
+      "loss": 1.1227,
+      "step": 566
+    },
+    {
+      "epoch": 2.9290322580645163,
+      "grad_norm": 0.5224789977073669,
+      "learning_rate": 8.291666666666667e-05,
+      "loss": 1.0901,
+      "step": 567
+    },
+    {
+      "epoch": 2.9341935483870967,
+      "grad_norm": 0.5665974020957947,
+      "learning_rate": 8.270833333333333e-05,
+      "loss": 1.152,
+      "step": 568
+    },
+    {
+      "epoch": 2.9393548387096775,
+      "grad_norm": 0.5928233861923218,
+      "learning_rate": 8.25e-05,
+      "loss": 1.0426,
+      "step": 569
+    },
+    {
+      "epoch": 2.944516129032258,
+      "grad_norm": 0.5530390739440918,
+      "learning_rate": 8.229166666666667e-05,
+      "loss": 1.0963,
+      "step": 570
+    },
+    {
+      "epoch": 2.9496774193548387,
+      "grad_norm": 0.5208332538604736,
+      "learning_rate": 8.208333333333334e-05,
+      "loss": 0.9963,
+      "step": 571
+    },
+    {
+      "epoch": 2.9548387096774196,
+      "grad_norm": 0.5431433320045471,
+      "learning_rate": 8.1875e-05,
+      "loss": 1.0447,
+      "step": 572
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.5688531994819641,
+      "learning_rate": 8.166666666666667e-05,
+      "loss": 1.1455,
+      "step": 573
+    },
+    {
+      "epoch": 2.965161290322581,
+      "grad_norm": 0.6022582054138184,
+      "learning_rate": 8.145833333333334e-05,
+      "loss": 1.0892,
+      "step": 574
+    },
+    {
+      "epoch": 2.970322580645161,
+      "grad_norm": 0.5424197316169739,
+      "learning_rate": 8.125000000000001e-05,
+      "loss": 1.1381,
+      "step": 575
+    },
+    {
+      "epoch": 2.975483870967742,
+      "grad_norm": 0.6352680921554565,
+      "learning_rate": 8.104166666666667e-05,
+      "loss": 1.1292,
+      "step": 576
+    },
+    {
+      "epoch": 2.9806451612903224,
+      "grad_norm": 0.543643057346344,
+      "learning_rate": 8.083333333333334e-05,
+      "loss": 1.1045,
+      "step": 577
+    },
+    {
+      "epoch": 2.9858064516129033,
+      "grad_norm": 0.5401563048362732,
+      "learning_rate": 8.062500000000001e-05,
+      "loss": 1.0984,
+      "step": 578
+    },
+    {
+      "epoch": 2.990967741935484,
+      "grad_norm": 0.5305353403091431,
+      "learning_rate": 8.041666666666668e-05,
+      "loss": 1.1026,
+      "step": 579
+    },
+    {
+      "epoch": 2.9961290322580645,
+      "grad_norm": 0.5655547380447388,
+      "learning_rate": 8.020833333333334e-05,
+      "loss": 1.1214,
+      "step": 580
+    },
+    {
+      "epoch": 2.9961290322580645,
+      "eval_loss": 1.5244600772857666,
+      "eval_runtime": 21.2384,
+      "eval_samples_per_second": 3.72,
+      "eval_steps_per_second": 0.471,
+      "step": 580
+    },
+    {
+      "epoch": 3.00258064516129,
+      "grad_norm": 0.8791653513908386,
+      "learning_rate": 8e-05,
+      "loss": 1.5039,
+      "step": 581
+    },
+    {
+      "epoch": 3.007741935483871,
+      "grad_norm": 0.7113364934921265,
+      "learning_rate": 7.979166666666668e-05,
+      "loss": 0.9274,
+      "step": 582
+    },
+    {
+      "epoch": 3.0129032258064514,
+      "grad_norm": 0.6806617379188538,
+      "learning_rate": 7.958333333333333e-05,
+      "loss": 0.8455,
+      "step": 583
+    },
+    {
+      "epoch": 3.0180645161290323,
+      "grad_norm": 0.7928464412689209,
+      "learning_rate": 7.9375e-05,
+      "loss": 0.8011,
+      "step": 584
+    },
+    {
+      "epoch": 3.023225806451613,
+      "grad_norm": 1.2611262798309326,
+      "learning_rate": 7.916666666666666e-05,
+      "loss": 0.8022,
+      "step": 585
+    },
+    {
+      "epoch": 3.0283870967741935,
+      "grad_norm": 1.5146794319152832,
+      "learning_rate": 7.895833333333333e-05,
+      "loss": 0.8739,
+      "step": 586
+    },
+    {
+      "epoch": 3.0335483870967743,
+      "grad_norm": 1.0011584758758545,
+      "learning_rate": 7.875e-05,
+      "loss": 0.889,
+      "step": 587
+    },
+    {
+      "epoch": 3.0387096774193547,
+      "grad_norm": 0.9589004516601562,
+      "learning_rate": 7.854166666666667e-05,
+      "loss": 0.7802,
+      "step": 588
+    },
+    {
+      "epoch": 3.0438709677419356,
+      "grad_norm": 0.7189547419548035,
+      "learning_rate": 7.833333333333333e-05,
+      "loss": 0.8272,
+      "step": 589
+    },
+    {
+      "epoch": 3.049032258064516,
+      "grad_norm": 0.712325930595398,
+      "learning_rate": 7.8125e-05,
+      "loss": 0.833,
+      "step": 590
+    },
+    {
+      "epoch": 3.054193548387097,
+      "grad_norm": 0.610262930393219,
+      "learning_rate": 7.791666666666667e-05,
+      "loss": 0.7837,
+      "step": 591
+    },
+    {
+      "epoch": 3.0593548387096776,
+      "grad_norm": 0.6256819367408752,
+      "learning_rate": 7.770833333333334e-05,
+      "loss": 0.8324,
+      "step": 592
+    },
+    {
+      "epoch": 3.064516129032258,
+      "grad_norm": 0.7267535924911499,
+      "learning_rate": 7.75e-05,
+      "loss": 0.8243,
+      "step": 593
+    },
+    {
+      "epoch": 3.069677419354839,
+      "grad_norm": 0.7415085434913635,
+      "learning_rate": 7.729166666666667e-05,
+      "loss": 0.788,
+      "step": 594
+    },
+    {
+      "epoch": 3.0748387096774192,
+      "grad_norm": 0.840854287147522,
+      "learning_rate": 7.708333333333334e-05,
+      "loss": 0.7871,
+      "step": 595
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 0.9644313454627991,
+      "learning_rate": 7.687500000000001e-05,
+      "loss": 0.8017,
+      "step": 596
+    },
+    {
+      "epoch": 3.0851612903225805,
+      "grad_norm": 0.8267636299133301,
+      "learning_rate": 7.666666666666667e-05,
+      "loss": 0.799,
+      "step": 597
+    },
+    {
+      "epoch": 3.0903225806451613,
+      "grad_norm": 0.7972861528396606,
+      "learning_rate": 7.645833333333333e-05,
+      "loss": 0.7899,
+      "step": 598
+    },
+    {
+      "epoch": 3.095483870967742,
+      "grad_norm": 0.6854184865951538,
+      "learning_rate": 7.625e-05,
+      "loss": 0.8196,
+      "step": 599
+    },
+    {
+      "epoch": 3.1006451612903225,
+      "grad_norm": 0.6697509288787842,
+      "learning_rate": 7.604166666666667e-05,
+      "loss": 0.8292,
+      "step": 600
+    },
+    {
+      "epoch": 3.1058064516129034,
+      "grad_norm": 0.8080317974090576,
+      "learning_rate": 7.583333333333334e-05,
+      "loss": 0.8464,
+      "step": 601
+    },
+    {
+      "epoch": 3.1109677419354838,
+      "grad_norm": 0.6825260519981384,
+      "learning_rate": 7.5625e-05,
+      "loss": 0.7883,
+      "step": 602
+    },
+    {
+      "epoch": 3.1161290322580646,
+      "grad_norm": 0.7252619862556458,
+      "learning_rate": 7.541666666666667e-05,
+      "loss": 0.851,
+      "step": 603
+    },
+    {
+      "epoch": 3.121290322580645,
+      "grad_norm": 0.703316867351532,
+      "learning_rate": 7.520833333333334e-05,
+      "loss": 0.7822,
+      "step": 604
+    },
+    {
+      "epoch": 3.126451612903226,
+      "grad_norm": 0.7531264424324036,
+      "learning_rate": 7.500000000000001e-05,
+      "loss": 0.8329,
+      "step": 605
+    },
+    {
+      "epoch": 3.1316129032258067,
+      "grad_norm": 0.8428468704223633,
+      "learning_rate": 7.479166666666668e-05,
+      "loss": 0.8312,
+      "step": 606
+    },
+    {
+      "epoch": 3.136774193548387,
+      "grad_norm": 0.782310426235199,
+      "learning_rate": 7.458333333333333e-05,
+      "loss": 0.7561,
+      "step": 607
+    },
+    {
+      "epoch": 3.141935483870968,
+      "grad_norm": 1.0301823616027832,
+      "learning_rate": 7.4375e-05,
+      "loss": 0.7823,
+      "step": 608
+    },
+    {
+      "epoch": 3.1470967741935483,
+      "grad_norm": 0.7504616975784302,
+      "learning_rate": 7.416666666666668e-05,
+      "loss": 0.807,
+      "step": 609
+    },
+    {
+      "epoch": 3.152258064516129,
+      "grad_norm": 0.7420417666435242,
+      "learning_rate": 7.395833333333335e-05,
+      "loss": 0.7864,
+      "step": 610
+    },
+    {
+      "epoch": 3.1574193548387095,
+      "grad_norm": 0.7365292906761169,
+      "learning_rate": 7.375e-05,
+      "loss": 0.7724,
+      "step": 611
+    },
+    {
+      "epoch": 3.1625806451612903,
+      "grad_norm": 0.7809944748878479,
+      "learning_rate": 7.354166666666667e-05,
+      "loss": 0.8116,
+      "step": 612
+    },
+    {
+      "epoch": 3.167741935483871,
+      "grad_norm": 0.7591716051101685,
+      "learning_rate": 7.333333333333333e-05,
+      "loss": 0.784,
+      "step": 613
+    },
+    {
+      "epoch": 3.1729032258064516,
+      "grad_norm": 0.742615282535553,
+      "learning_rate": 7.3125e-05,
+      "loss": 0.7639,
+      "step": 614
+    },
+    {
+      "epoch": 3.1780645161290324,
+      "grad_norm": 1.0104111433029175,
+      "learning_rate": 7.291666666666667e-05,
+      "loss": 0.8074,
+      "step": 615
+    },
+    {
+      "epoch": 3.183225806451613,
+      "grad_norm": 0.7318562865257263,
+      "learning_rate": 7.270833333333333e-05,
+      "loss": 0.8239,
+      "step": 616
+    },
+    {
+      "epoch": 3.1883870967741936,
+      "grad_norm": 0.7688681483268738,
+      "learning_rate": 7.25e-05,
+      "loss": 0.8531,
+      "step": 617
+    },
+    {
+      "epoch": 3.193548387096774,
+      "grad_norm": 0.7818340063095093,
+      "learning_rate": 7.229166666666667e-05,
+      "loss": 0.8858,
+      "step": 618
+    },
+    {
+      "epoch": 3.198709677419355,
+      "grad_norm": 0.7311562299728394,
+      "learning_rate": 7.208333333333334e-05,
+      "loss": 0.8473,
+      "step": 619
+    },
+    {
+      "epoch": 3.2038709677419357,
+      "grad_norm": 0.7646205425262451,
+      "learning_rate": 7.1875e-05,
+      "loss": 0.8239,
+      "step": 620
+    },
+    {
+      "epoch": 3.209032258064516,
+      "grad_norm": 0.7341886758804321,
+      "learning_rate": 7.166666666666667e-05,
+      "loss": 0.848,
+      "step": 621
+    },
+    {
+      "epoch": 3.214193548387097,
+      "grad_norm": 0.7330101132392883,
+      "learning_rate": 7.145833333333334e-05,
+      "loss": 0.7241,
+      "step": 622
+    },
+    {
+      "epoch": 3.2193548387096773,
+      "grad_norm": 0.7430225014686584,
+      "learning_rate": 7.125000000000001e-05,
+      "loss": 0.8396,
+      "step": 623
+    },
+    {
+      "epoch": 3.224516129032258,
+      "grad_norm": 0.8028465509414673,
+      "learning_rate": 7.104166666666667e-05,
+      "loss": 0.8739,
+      "step": 624
+    },
+    {
+      "epoch": 3.2296774193548385,
+      "grad_norm": 0.7185546159744263,
+      "learning_rate": 7.083333333333334e-05,
+      "loss": 0.8071,
+      "step": 625
+    },
+    {
+      "epoch": 3.2348387096774194,
+      "grad_norm": 0.7849711775779724,
+      "learning_rate": 7.062500000000001e-05,
+      "loss": 0.8192,
+      "step": 626
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 0.7927383780479431,
+      "learning_rate": 7.041666666666668e-05,
+      "loss": 0.8795,
+      "step": 627
+    },
+    {
+      "epoch": 3.2451612903225806,
+      "grad_norm": 0.691220223903656,
+      "learning_rate": 7.020833333333334e-05,
+      "loss": 0.7745,
+      "step": 628
+    },
+    {
+      "epoch": 3.2503225806451614,
+      "grad_norm": 0.7085890769958496,
+      "learning_rate": 7e-05,
+      "loss": 0.813,
+      "step": 629
+    },
+    {
+      "epoch": 3.255483870967742,
+      "grad_norm": 0.7118912935256958,
+      "learning_rate": 6.979166666666666e-05,
+      "loss": 0.8351,
+      "step": 630
+    },
+    {
+      "epoch": 3.2606451612903227,
+      "grad_norm": 0.7210044264793396,
+      "learning_rate": 6.958333333333334e-05,
+      "loss": 0.8374,
+      "step": 631
+    },
+    {
+      "epoch": 3.265806451612903,
+      "grad_norm": 0.7748726010322571,
+      "learning_rate": 6.9375e-05,
+      "loss": 0.8182,
+      "step": 632
+    },
+    {
+      "epoch": 3.270967741935484,
+      "grad_norm": 0.7490830421447754,
+      "learning_rate": 6.916666666666666e-05,
+      "loss": 0.8061,
+      "step": 633
+    },
+    {
+      "epoch": 3.2761290322580647,
+      "grad_norm": 0.7532171010971069,
+      "learning_rate": 6.895833333333333e-05,
+      "loss": 0.8027,
+      "step": 634
+    },
+    {
+      "epoch": 3.281290322580645,
+      "grad_norm": 6.004159450531006,
+      "learning_rate": 6.875e-05,
+      "loss": 0.8425,
+      "step": 635
+    },
+    {
+      "epoch": 3.286451612903226,
+      "grad_norm": 0.7662097215652466,
+      "learning_rate": 6.854166666666667e-05,
+      "loss": 0.7501,
+      "step": 636
+    },
+    {
+      "epoch": 3.2916129032258064,
+      "grad_norm": 0.804734468460083,
+      "learning_rate": 6.833333333333333e-05,
+      "loss": 0.7978,
+      "step": 637
+    },
+    {
+      "epoch": 3.296774193548387,
+      "grad_norm": 0.7840864658355713,
+      "learning_rate": 6.8125e-05,
+      "loss": 0.8172,
+      "step": 638
+    },
+    {
+      "epoch": 3.3019354838709676,
+      "grad_norm": 0.7220107316970825,
+      "learning_rate": 6.791666666666667e-05,
+      "loss": 0.7278,
+      "step": 639
+    },
+    {
+      "epoch": 3.3070967741935484,
+      "grad_norm": 0.7706899046897888,
+      "learning_rate": 6.770833333333334e-05,
+      "loss": 0.7719,
+      "step": 640
+    },
+    {
+      "epoch": 3.3122580645161293,
+      "grad_norm": 0.7455759644508362,
+      "learning_rate": 6.750000000000001e-05,
+      "loss": 0.6809,
+      "step": 641
+    },
+    {
+      "epoch": 3.3174193548387096,
+      "grad_norm": 0.7620101571083069,
+      "learning_rate": 6.729166666666667e-05,
+      "loss": 0.8462,
+      "step": 642
+    },
+    {
+      "epoch": 3.3225806451612905,
+      "grad_norm": 0.7075409293174744,
+      "learning_rate": 6.708333333333333e-05,
+      "loss": 0.7908,
+      "step": 643
+    },
+    {
+      "epoch": 3.327741935483871,
+      "grad_norm": 0.73841392993927,
+      "learning_rate": 6.6875e-05,
+      "loss": 0.8048,
+      "step": 644
+    },
+    {
+      "epoch": 3.3329032258064517,
+      "grad_norm": 0.6862002015113831,
+      "learning_rate": 6.666666666666667e-05,
+      "loss": 0.7662,
+      "step": 645
+    },
+    {
+      "epoch": 3.338064516129032,
+      "grad_norm": 0.7678553462028503,
+      "learning_rate": 6.645833333333333e-05,
+      "loss": 0.7746,
+      "step": 646
+    },
+    {
+      "epoch": 3.343225806451613,
+      "grad_norm": 0.7641354203224182,
+      "learning_rate": 6.625e-05,
+      "loss": 0.7957,
+      "step": 647
+    },
+    {
+      "epoch": 3.3483870967741938,
+      "grad_norm": 0.797936737537384,
+      "learning_rate": 6.604166666666667e-05,
+      "loss": 0.8438,
+      "step": 648
+    },
+    {
+      "epoch": 3.353548387096774,
+      "grad_norm": 0.7911980152130127,
+      "learning_rate": 6.583333333333334e-05,
+      "loss": 0.8293,
+      "step": 649
+    },
+    {
+      "epoch": 3.358709677419355,
+      "grad_norm": 0.7442024350166321,
+      "learning_rate": 6.562500000000001e-05,
+      "loss": 0.8371,
+      "step": 650
+    },
+    {
+      "epoch": 3.3638709677419354,
+      "grad_norm": 0.7448225021362305,
+      "learning_rate": 6.541666666666667e-05,
+      "loss": 0.8579,
+      "step": 651
+    },
+    {
+      "epoch": 3.3690322580645162,
+      "grad_norm": 0.7056671380996704,
+      "learning_rate": 6.520833333333334e-05,
+      "loss": 0.8034,
+      "step": 652
+    },
+    {
+      "epoch": 3.3741935483870966,
+      "grad_norm": 0.7186213731765747,
+      "learning_rate": 6.500000000000001e-05,
+      "loss": 0.8532,
+      "step": 653
+    },
+    {
+      "epoch": 3.3793548387096775,
+      "grad_norm": 0.728285551071167,
+      "learning_rate": 6.479166666666668e-05,
+      "loss": 0.8351,
+      "step": 654
+    },
+    {
+      "epoch": 3.3845161290322583,
+      "grad_norm": 0.7283080816268921,
+      "learning_rate": 6.458333333333334e-05,
+      "loss": 0.8296,
+      "step": 655
+    },
+    {
+      "epoch": 3.3896774193548387,
+      "grad_norm": 0.7502827644348145,
+      "learning_rate": 6.4375e-05,
+      "loss": 0.8066,
+      "step": 656
+    },
+    {
+      "epoch": 3.3948387096774195,
+      "grad_norm": 0.7978790402412415,
+      "learning_rate": 6.416666666666668e-05,
+      "loss": 0.8248,
+      "step": 657
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 0.7639005780220032,
+      "learning_rate": 6.395833333333333e-05,
+      "loss": 0.8675,
+      "step": 658
+    },
+    {
+      "epoch": 3.4051612903225807,
+      "grad_norm": 0.7718410491943359,
+      "learning_rate": 6.375e-05,
+      "loss": 0.7257,
+      "step": 659
+    },
+    {
+      "epoch": 3.410322580645161,
+      "grad_norm": 0.7846454977989197,
+      "learning_rate": 6.354166666666666e-05,
+      "loss": 0.8744,
+      "step": 660
+    },
+    {
+      "epoch": 3.415483870967742,
+      "grad_norm": 0.7490540146827698,
+      "learning_rate": 6.333333333333333e-05,
+      "loss": 0.7866,
+      "step": 661
+    },
+    {
+      "epoch": 3.420645161290323,
+      "grad_norm": 0.7267460823059082,
+      "learning_rate": 6.3125e-05,
+      "loss": 0.7532,
+      "step": 662
+    },
+    {
+      "epoch": 3.425806451612903,
+      "grad_norm": 0.7694923281669617,
+      "learning_rate": 6.291666666666667e-05,
+      "loss": 0.8064,
+      "step": 663
+    },
+    {
+      "epoch": 3.430967741935484,
+      "grad_norm": 0.7657822370529175,
+      "learning_rate": 6.270833333333333e-05,
+      "loss": 0.8681,
+      "step": 664
+    },
+    {
+      "epoch": 3.4361290322580644,
+      "grad_norm": 0.7238038778305054,
+      "learning_rate": 6.25e-05,
+      "loss": 0.8426,
+      "step": 665
+    },
+    {
+      "epoch": 3.4412903225806453,
+      "grad_norm": 0.8218737244606018,
+      "learning_rate": 6.229166666666667e-05,
+      "loss": 0.8534,
+      "step": 666
+    },
+    {
+      "epoch": 3.4464516129032257,
+      "grad_norm": 0.7586380839347839,
+      "learning_rate": 6.208333333333334e-05,
+      "loss": 0.8511,
+      "step": 667
+    },
+    {
+      "epoch": 3.4516129032258065,
+      "grad_norm": 0.7779055237770081,
+      "learning_rate": 6.1875e-05,
+      "loss": 0.8381,
+      "step": 668
+    },
+    {
+      "epoch": 3.4567741935483873,
+      "grad_norm": 0.7387574911117554,
+      "learning_rate": 6.166666666666667e-05,
+      "loss": 0.849,
+      "step": 669
+    },
+    {
+      "epoch": 3.4619354838709677,
+      "grad_norm": 0.9027093648910522,
+      "learning_rate": 6.145833333333334e-05,
+      "loss": 0.8504,
+      "step": 670
+    },
+    {
+      "epoch": 3.4670967741935486,
+      "grad_norm": 0.7768834829330444,
+      "learning_rate": 6.125000000000001e-05,
+      "loss": 0.8166,
+      "step": 671
+    },
+    {
+      "epoch": 3.472258064516129,
+      "grad_norm": 0.8063477873802185,
+      "learning_rate": 6.104166666666667e-05,
+      "loss": 0.7208,
+      "step": 672
+    },
+    {
+      "epoch": 3.47741935483871,
+      "grad_norm": 0.8137701153755188,
+      "learning_rate": 6.083333333333333e-05,
+      "loss": 0.8816,
+      "step": 673
+    },
+    {
+      "epoch": 3.48258064516129,
+      "grad_norm": 0.7777495980262756,
+      "learning_rate": 6.0624999999999996e-05,
+      "loss": 0.8138,
+      "step": 674
+    },
+    {
+      "epoch": 3.487741935483871,
+      "grad_norm": 0.7630706429481506,
+      "learning_rate": 6.041666666666667e-05,
+      "loss": 0.8141,
+      "step": 675
+    },
+    {
+      "epoch": 3.492903225806452,
+      "grad_norm": 0.7368703484535217,
+      "learning_rate": 6.020833333333333e-05,
+      "loss": 0.8115,
+      "step": 676
+    },
+    {
+      "epoch": 3.4980645161290322,
+      "grad_norm": 0.7266126871109009,
+      "learning_rate": 6e-05,
+      "loss": 0.8618,
+      "step": 677
+    },
+    {
+      "epoch": 3.5032258064516126,
+      "grad_norm": 0.7643237113952637,
+      "learning_rate": 5.9791666666666665e-05,
+      "loss": 0.8781,
+      "step": 678
+    },
+    {
+      "epoch": 3.5083870967741935,
+      "grad_norm": 0.7601742148399353,
+      "learning_rate": 5.9583333333333336e-05,
+      "loss": 0.8585,
+      "step": 679
+    },
+    {
+      "epoch": 3.5135483870967743,
+      "grad_norm": 0.7264940738677979,
+      "learning_rate": 5.9375e-05,
+      "loss": 0.769,
+      "step": 680
+    },
+    {
+      "epoch": 3.5187096774193547,
+      "grad_norm": 0.7804604768753052,
+      "learning_rate": 5.916666666666667e-05,
+      "loss": 0.8203,
+      "step": 681
+    },
+    {
+      "epoch": 3.5238709677419355,
+      "grad_norm": 0.7801588773727417,
+      "learning_rate": 5.8958333333333334e-05,
+      "loss": 0.8154,
+      "step": 682
+    },
+    {
+      "epoch": 3.5290322580645164,
+      "grad_norm": 0.7485193610191345,
+      "learning_rate": 5.8750000000000005e-05,
+      "loss": 0.8066,
+      "step": 683
+    },
+    {
+      "epoch": 3.5341935483870968,
+      "grad_norm": 0.8037455081939697,
+      "learning_rate": 5.8541666666666676e-05,
+      "loss": 0.841,
+      "step": 684
+    },
+    {
+      "epoch": 3.539354838709677,
+      "grad_norm": 0.7463704347610474,
+      "learning_rate": 5.833333333333334e-05,
+      "loss": 0.8041,
+      "step": 685
+    },
+    {
+      "epoch": 3.544516129032258,
+      "grad_norm": 0.7943751215934753,
+      "learning_rate": 5.812500000000001e-05,
+      "loss": 0.8242,
+      "step": 686
+    },
+    {
+      "epoch": 3.549677419354839,
+      "grad_norm": 0.7888543009757996,
+      "learning_rate": 5.7916666666666674e-05,
+      "loss": 0.8881,
+      "step": 687
+    },
+    {
+      "epoch": 3.554838709677419,
+      "grad_norm": 0.7699112892150879,
+      "learning_rate": 5.770833333333333e-05,
+      "loss": 0.8045,
+      "step": 688
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 0.7419706583023071,
+      "learning_rate": 5.7499999999999995e-05,
+      "loss": 0.8216,
+      "step": 689
+    },
+    {
+      "epoch": 3.565161290322581,
+      "grad_norm": 0.7404356598854065,
+      "learning_rate": 5.7291666666666666e-05,
+      "loss": 0.8335,
+      "step": 690
+    },
+    {
+      "epoch": 3.5703225806451613,
+      "grad_norm": 0.7554022669792175,
+      "learning_rate": 5.7083333333333336e-05,
+      "loss": 0.8386,
+      "step": 691
+    },
+    {
+      "epoch": 3.5754838709677417,
+      "grad_norm": 0.7632009387016296,
+      "learning_rate": 5.6875e-05,
+      "loss": 0.8855,
+      "step": 692
+    },
+    {
+      "epoch": 3.5806451612903225,
+      "grad_norm": 0.796056866645813,
+      "learning_rate": 5.666666666666667e-05,
+      "loss": 0.846,
+      "step": 693
+    },
+    {
+      "epoch": 3.5858064516129033,
+      "grad_norm": 0.7852962017059326,
+      "learning_rate": 5.6458333333333335e-05,
+      "loss": 0.8863,
+      "step": 694
+    },
+    {
+      "epoch": 3.5909677419354837,
+      "grad_norm": 0.7232403755187988,
+      "learning_rate": 5.6250000000000005e-05,
+      "loss": 0.8464,
+      "step": 695
+    },
+    {
+      "epoch": 3.5961290322580646,
+      "grad_norm": 0.7412312030792236,
+      "learning_rate": 5.604166666666667e-05,
+      "loss": 0.7806,
+      "step": 696
+    },
+    {
+      "epoch": 3.6012903225806454,
+      "grad_norm": 0.7273991107940674,
+      "learning_rate": 5.583333333333334e-05,
+      "loss": 0.7558,
+      "step": 697
+    },
+    {
+      "epoch": 3.606451612903226,
+      "grad_norm": 0.7656369209289551,
+      "learning_rate": 5.5625000000000004e-05,
+      "loss": 0.8604,
+      "step": 698
+    },
+    {
+      "epoch": 3.611612903225806,
+      "grad_norm": 0.7474403977394104,
+      "learning_rate": 5.5416666666666674e-05,
+      "loss": 0.8502,
+      "step": 699
+    },
+    {
+      "epoch": 3.616774193548387,
+      "grad_norm": 0.9587624073028564,
+      "learning_rate": 5.520833333333334e-05,
+      "loss": 0.7549,
+      "step": 700
+    },
+    {
+      "epoch": 3.621935483870968,
+      "grad_norm": 0.752896785736084,
+      "learning_rate": 5.500000000000001e-05,
+      "loss": 0.7503,
+      "step": 701
+    },
+    {
+      "epoch": 3.6270967741935483,
+      "grad_norm": 0.7378438711166382,
+      "learning_rate": 5.479166666666667e-05,
+      "loss": 0.8194,
+      "step": 702
+    },
+    {
+      "epoch": 3.632258064516129,
+      "grad_norm": 0.7898781895637512,
+      "learning_rate": 5.458333333333333e-05,
+      "loss": 0.8779,
+      "step": 703
+    },
+    {
+      "epoch": 3.63741935483871,
+      "grad_norm": 0.7911474704742432,
+      "learning_rate": 5.4375e-05,
+      "loss": 0.8971,
+      "step": 704
+    },
+    {
+      "epoch": 3.6425806451612903,
+      "grad_norm": 0.9014036059379578,
+      "learning_rate": 5.4166666666666664e-05,
+      "loss": 0.8787,
+      "step": 705
+    },
+    {
+      "epoch": 3.6477419354838707,
+      "grad_norm": 0.787401556968689,
+      "learning_rate": 5.3958333333333335e-05,
+      "loss": 0.8458,
+      "step": 706
+    },
+    {
+      "epoch": 3.6529032258064515,
+      "grad_norm": 0.9262308478355408,
+      "learning_rate": 5.375e-05,
+      "loss": 0.81,
+      "step": 707
+    },
+    {
+      "epoch": 3.6580645161290324,
+      "grad_norm": 0.7629384994506836,
+      "learning_rate": 5.354166666666667e-05,
+      "loss": 0.8162,
+      "step": 708
+    },
+    {
+      "epoch": 3.6632258064516128,
+      "grad_norm": 0.8068932890892029,
+      "learning_rate": 5.333333333333333e-05,
+      "loss": 0.8406,
+      "step": 709
+    },
+    {
+      "epoch": 3.6683870967741936,
+      "grad_norm": 0.7606951594352722,
+      "learning_rate": 5.3125000000000004e-05,
+      "loss": 0.8157,
+      "step": 710
+    },
+    {
+      "epoch": 3.6735483870967744,
+      "grad_norm": 0.7581308484077454,
+      "learning_rate": 5.291666666666667e-05,
+      "loss": 0.8546,
+      "step": 711
+    },
+    {
+      "epoch": 3.678709677419355,
+      "grad_norm": 0.6999085545539856,
+      "learning_rate": 5.270833333333334e-05,
+      "loss": 0.7221,
+      "step": 712
+    },
+    {
+      "epoch": 3.6838709677419352,
+      "grad_norm": 0.7534416317939758,
+      "learning_rate": 5.25e-05,
+      "loss": 0.8407,
+      "step": 713
+    },
+    {
+      "epoch": 3.689032258064516,
+      "grad_norm": 0.773116409778595,
+      "learning_rate": 5.229166666666667e-05,
+      "loss": 0.8156,
+      "step": 714
+    },
+    {
+      "epoch": 3.694193548387097,
+      "grad_norm": 0.7262045741081238,
+      "learning_rate": 5.208333333333334e-05,
+      "loss": 0.7481,
+      "step": 715
+    },
+    {
+      "epoch": 3.6993548387096773,
+      "grad_norm": 0.7585551142692566,
+      "learning_rate": 5.187500000000001e-05,
+      "loss": 0.7861,
+      "step": 716
+    },
+    {
+      "epoch": 3.704516129032258,
+      "grad_norm": 0.7631360292434692,
+      "learning_rate": 5.166666666666667e-05,
+      "loss": 0.7976,
+      "step": 717
+    },
+    {
+      "epoch": 3.709677419354839,
+      "grad_norm": 0.809861958026886,
+      "learning_rate": 5.145833333333333e-05,
+      "loss": 0.8425,
+      "step": 718
+    },
+    {
+      "epoch": 3.7148387096774194,
+      "grad_norm": 0.8243117928504944,
+      "learning_rate": 5.125e-05,
+      "loss": 0.8017,
+      "step": 719
+    },
+    {
+      "epoch": 3.7199999999999998,
+      "grad_norm": 0.7856522798538208,
+      "learning_rate": 5.104166666666666e-05,
+      "loss": 0.8087,
+      "step": 720
+    },
+    {
+      "epoch": 3.7251612903225806,
+      "grad_norm": 0.7896134853363037,
+      "learning_rate": 5.0833333333333333e-05,
+      "loss": 0.8354,
+      "step": 721
+    },
+    {
+      "epoch": 3.7303225806451614,
+      "grad_norm": 0.7945472598075867,
+      "learning_rate": 5.0625e-05,
+      "loss": 0.9051,
+      "step": 722
+    },
+    {
+      "epoch": 3.735483870967742,
+      "grad_norm": 0.784895122051239,
+      "learning_rate": 5.041666666666667e-05,
+      "loss": 0.863,
+      "step": 723
+    },
+    {
+      "epoch": 3.7406451612903227,
+      "grad_norm": 0.8042978048324585,
+      "learning_rate": 5.020833333333333e-05,
+      "loss": 0.8132,
+      "step": 724
+    },
+    {
+      "epoch": 3.7458064516129035,
+      "grad_norm": 0.7665282487869263,
+      "learning_rate": 5e-05,
+      "loss": 0.8884,
+      "step": 725
+    },
+    {
+      "epoch": 3.750967741935484,
+      "grad_norm": 0.7703678011894226,
+      "learning_rate": 4.979166666666667e-05,
+      "loss": 0.8435,
+      "step": 726
+    },
+    {
+      "epoch": 3.7561290322580643,
+      "grad_norm": 1.0076745748519897,
+      "learning_rate": 4.958333333333334e-05,
+      "loss": 0.821,
+      "step": 727
+    },
+    {
+      "epoch": 3.761290322580645,
+      "grad_norm": 0.7281281352043152,
+      "learning_rate": 4.937500000000001e-05,
+      "loss": 0.8027,
+      "step": 728
+    },
+    {
+      "epoch": 3.766451612903226,
+      "grad_norm": 0.7235384583473206,
+      "learning_rate": 4.9166666666666665e-05,
+      "loss": 0.7137,
+      "step": 729
+    },
+    {
+      "epoch": 3.7716129032258063,
+      "grad_norm": 0.7476503849029541,
+      "learning_rate": 4.8958333333333335e-05,
+      "loss": 0.7255,
+      "step": 730
+    },
+    {
+      "epoch": 3.776774193548387,
+      "grad_norm": 0.8011372685432434,
+      "learning_rate": 4.875e-05,
+      "loss": 0.8743,
+      "step": 731
+    },
+    {
+      "epoch": 3.781935483870968,
+      "grad_norm": 0.7595112919807434,
+      "learning_rate": 4.854166666666667e-05,
+      "loss": 0.8311,
+      "step": 732
+    },
+    {
+      "epoch": 3.7870967741935484,
+      "grad_norm": 0.774946928024292,
+      "learning_rate": 4.8333333333333334e-05,
+      "loss": 0.7727,
+      "step": 733
+    },
+    {
+      "epoch": 3.792258064516129,
+      "grad_norm": 0.8403404355049133,
+      "learning_rate": 4.8125000000000004e-05,
+      "loss": 0.8339,
+      "step": 734
+    },
+    {
+      "epoch": 3.7974193548387096,
+      "grad_norm": 0.7683903574943542,
+      "learning_rate": 4.791666666666667e-05,
+      "loss": 0.8086,
+      "step": 735
+    },
+    {
+      "epoch": 3.8025806451612905,
+      "grad_norm": 0.802428126335144,
+      "learning_rate": 4.770833333333334e-05,
+      "loss": 0.7829,
+      "step": 736
+    },
+    {
+      "epoch": 3.807741935483871,
+      "grad_norm": 0.7652261853218079,
+      "learning_rate": 4.75e-05,
+      "loss": 0.8122,
+      "step": 737
+    },
+    {
+      "epoch": 3.8129032258064517,
+      "grad_norm": 0.7568336725234985,
+      "learning_rate": 4.7291666666666666e-05,
+      "loss": 0.7915,
+      "step": 738
+    },
+    {
+      "epoch": 3.8180645161290325,
+      "grad_norm": 0.7745715379714966,
+      "learning_rate": 4.708333333333334e-05,
+      "loss": 0.8808,
+      "step": 739
+    },
+    {
+      "epoch": 3.823225806451613,
+      "grad_norm": 0.7356999516487122,
+      "learning_rate": 4.6875e-05,
+      "loss": 0.8468,
+      "step": 740
+    },
+    {
+      "epoch": 3.8283870967741933,
+      "grad_norm": 0.7619341611862183,
+      "learning_rate": 4.666666666666667e-05,
+      "loss": 0.8158,
+      "step": 741
+    },
+    {
+      "epoch": 3.833548387096774,
+      "grad_norm": 0.7270290851593018,
+      "learning_rate": 4.6458333333333335e-05,
+      "loss": 0.8458,
+      "step": 742
+    },
+    {
+      "epoch": 3.838709677419355,
+      "grad_norm": 0.7820645570755005,
+      "learning_rate": 4.6250000000000006e-05,
+      "loss": 0.8576,
+      "step": 743
+    },
+    {
+      "epoch": 3.8438709677419354,
+      "grad_norm": 0.7956517934799194,
+      "learning_rate": 4.604166666666666e-05,
+      "loss": 0.8335,
+      "step": 744
+    },
+    {
+      "epoch": 3.849032258064516,
+      "grad_norm": 0.7973860502243042,
+      "learning_rate": 4.5833333333333334e-05,
+      "loss": 0.8477,
+      "step": 745
+    },
+    {
+      "epoch": 3.854193548387097,
+      "grad_norm": 0.7765873670578003,
+      "learning_rate": 4.5625e-05,
+      "loss": 0.7873,
+      "step": 746
+    },
+    {
+      "epoch": 3.8593548387096774,
+      "grad_norm": 0.7812808156013489,
+      "learning_rate": 4.541666666666667e-05,
+      "loss": 0.8404,
+      "step": 747
+    },
+    {
+      "epoch": 3.864516129032258,
+      "grad_norm": 0.7934351563453674,
+      "learning_rate": 4.520833333333334e-05,
+      "loss": 0.8587,
+      "step": 748
+    },
+    {
+      "epoch": 3.8696774193548387,
+      "grad_norm": 0.8019416928291321,
+      "learning_rate": 4.5e-05,
+      "loss": 0.8531,
+      "step": 749
+    },
+    {
+      "epoch": 3.8748387096774195,
+      "grad_norm": 0.8394574522972107,
+      "learning_rate": 4.4791666666666673e-05,
+      "loss": 0.8543,
+      "step": 750
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 0.7597457766532898,
+      "learning_rate": 4.458333333333334e-05,
+      "loss": 0.8694,
+      "step": 751
+    },
+    {
+      "epoch": 3.8851612903225807,
+      "grad_norm": 0.7650877237319946,
+      "learning_rate": 4.4375e-05,
+      "loss": 0.8577,
+      "step": 752
+    },
+    {
+      "epoch": 3.8903225806451616,
+      "grad_norm": 0.7785264849662781,
+      "learning_rate": 4.4166666666666665e-05,
+      "loss": 0.8267,
+      "step": 753
+    },
+    {
+      "epoch": 3.895483870967742,
+      "grad_norm": 0.7731680870056152,
+      "learning_rate": 4.3958333333333336e-05,
+      "loss": 0.8357,
+      "step": 754
+    },
+    {
+      "epoch": 3.9006451612903223,
+      "grad_norm": 0.7492013573646545,
+      "learning_rate": 4.375e-05,
+      "loss": 0.8196,
+      "step": 755
+    },
+    {
+      "epoch": 3.905806451612903,
+      "grad_norm": 0.7781587839126587,
+      "learning_rate": 4.354166666666667e-05,
+      "loss": 0.8356,
+      "step": 756
+    },
+    {
+      "epoch": 3.910967741935484,
+      "grad_norm": 0.7415270805358887,
+      "learning_rate": 4.3333333333333334e-05,
+      "loss": 0.7711,
+      "step": 757
+    },
+    {
+      "epoch": 3.9161290322580644,
+      "grad_norm": 0.7531086802482605,
+      "learning_rate": 4.3125000000000005e-05,
+      "loss": 0.8148,
+      "step": 758
+    },
+    {
+      "epoch": 3.9212903225806452,
+      "grad_norm": 0.766800582408905,
+      "learning_rate": 4.291666666666667e-05,
+      "loss": 0.8307,
+      "step": 759
+    },
+    {
+      "epoch": 3.9264516129032256,
+      "grad_norm": 0.7646375894546509,
+      "learning_rate": 4.270833333333333e-05,
+      "loss": 0.8149,
+      "step": 760
+    },
+    {
+      "epoch": 3.9316129032258065,
+      "grad_norm": 0.7876635789871216,
+      "learning_rate": 4.25e-05,
+      "loss": 0.8396,
+      "step": 761
+    },
+    {
+      "epoch": 3.936774193548387,
+      "grad_norm": 0.8202487826347351,
+      "learning_rate": 4.229166666666667e-05,
+      "loss": 0.9045,
+      "step": 762
+    },
+    {
+      "epoch": 3.9419354838709677,
+      "grad_norm": 0.8007482290267944,
+      "learning_rate": 4.208333333333334e-05,
+      "loss": 0.8316,
+      "step": 763
+    },
+    {
+      "epoch": 3.9470967741935485,
+      "grad_norm": 1.2683680057525635,
+      "learning_rate": 4.1875e-05,
+      "loss": 0.7341,
+      "step": 764
+    },
+    {
+      "epoch": 3.952258064516129,
+      "grad_norm": 0.7918137907981873,
+      "learning_rate": 4.166666666666667e-05,
+      "loss": 0.8732,
+      "step": 765
+    },
+    {
+      "epoch": 3.9574193548387098,
+      "grad_norm": 0.7748139500617981,
+      "learning_rate": 4.1458333333333336e-05,
+      "loss": 0.7292,
+      "step": 766
+    },
+    {
+      "epoch": 3.96258064516129,
+      "grad_norm": 0.785139262676239,
+      "learning_rate": 4.125e-05,
+      "loss": 0.8109,
+      "step": 767
+    },
+    {
+      "epoch": 3.967741935483871,
+      "grad_norm": 0.7704663872718811,
+      "learning_rate": 4.104166666666667e-05,
+      "loss": 0.7887,
+      "step": 768
+    },
+    {
+      "epoch": 3.9729032258064514,
+      "grad_norm": 0.799435555934906,
+      "learning_rate": 4.0833333333333334e-05,
+      "loss": 0.7726,
+      "step": 769
+    },
+    {
+      "epoch": 3.9780645161290322,
+      "grad_norm": 0.8685177564620972,
+      "learning_rate": 4.0625000000000005e-05,
+      "loss": 0.8884,
+      "step": 770
+    },
+    {
+      "epoch": 3.983225806451613,
+      "grad_norm": 0.7688182592391968,
+      "learning_rate": 4.041666666666667e-05,
+      "loss": 0.8073,
+      "step": 771
+    },
+    {
+      "epoch": 3.9883870967741935,
+      "grad_norm": 0.9119554758071899,
+      "learning_rate": 4.020833333333334e-05,
+      "loss": 0.8148,
+      "step": 772
+    },
+    {
+      "epoch": 3.9935483870967743,
+      "grad_norm": 0.791867733001709,
+      "learning_rate": 4e-05,
+      "loss": 0.8671,
+      "step": 773
+    },
+    {
+      "epoch": 3.9987096774193547,
+      "grad_norm": 1.1687339544296265,
+      "learning_rate": 3.979166666666667e-05,
+      "loss": 1.2081,
+      "step": 774
+    },
+    {
+      "epoch": 4.00516129032258,
+      "grad_norm": 0.750440239906311,
+      "learning_rate": 3.958333333333333e-05,
+      "loss": 0.6611,
+      "step": 775
+    },
+    {
+      "epoch": 4.010322580645162,
+      "grad_norm": 0.8063217997550964,
+      "learning_rate": 3.9375e-05,
+      "loss": 0.6704,
+      "step": 776
+    },
+    {
+      "epoch": 4.015483870967742,
+      "grad_norm": 0.6852978467941284,
+      "learning_rate": 3.9166666666666665e-05,
+      "loss": 0.6003,
+      "step": 777
+    },
+    {
+      "epoch": 4.0206451612903225,
+      "grad_norm": 0.6679868102073669,
+      "learning_rate": 3.8958333333333336e-05,
+      "loss": 0.5893,
+      "step": 778
+    },
+    {
+      "epoch": 4.025806451612903,
+      "grad_norm": 0.7534319162368774,
+      "learning_rate": 3.875e-05,
+      "loss": 0.592,
+      "step": 779
+    },
+    {
+      "epoch": 4.030967741935484,
+      "grad_norm": 0.9972567558288574,
+      "learning_rate": 3.854166666666667e-05,
+      "loss": 0.6203,
+      "step": 780
+    },
+    {
+      "epoch": 4.0361290322580645,
+      "grad_norm": 1.2949519157409668,
+      "learning_rate": 3.8333333333333334e-05,
+      "loss": 0.6352,
+      "step": 781
+    },
+    {
+      "epoch": 4.041290322580645,
+      "grad_norm": 1.2546395063400269,
+      "learning_rate": 3.8125e-05,
+      "loss": 0.637,
+      "step": 782
+    },
+    {
+      "epoch": 4.046451612903226,
+      "grad_norm": 1.274681806564331,
+      "learning_rate": 3.791666666666667e-05,
+      "loss": 0.6264,
+      "step": 783
+    },
+    {
+      "epoch": 4.051612903225807,
+      "grad_norm": 0.9987473487854004,
+      "learning_rate": 3.770833333333333e-05,
+      "loss": 0.5441,
+      "step": 784
+    },
+    {
+      "epoch": 4.056774193548387,
+      "grad_norm": 0.9196683168411255,
+      "learning_rate": 3.7500000000000003e-05,
+      "loss": 0.5717,
+      "step": 785
+    },
+    {
+      "epoch": 4.061935483870967,
+      "grad_norm": 0.8498789072036743,
+      "learning_rate": 3.729166666666667e-05,
+      "loss": 0.6121,
+      "step": 786
+    },
+    {
+      "epoch": 4.067096774193549,
+      "grad_norm": 0.7790414690971375,
+      "learning_rate": 3.708333333333334e-05,
+      "loss": 0.6403,
+      "step": 787
+    },
+    {
+      "epoch": 4.072258064516129,
+      "grad_norm": 0.7922559976577759,
+      "learning_rate": 3.6875e-05,
+      "loss": 0.62,
+      "step": 788
+    },
+    {
+      "epoch": 4.077419354838709,
+      "grad_norm": 0.7587729692459106,
+      "learning_rate": 3.6666666666666666e-05,
+      "loss": 0.605,
+      "step": 789
+    },
+    {
+      "epoch": 4.082580645161291,
+      "grad_norm": 0.7607321739196777,
+      "learning_rate": 3.6458333333333336e-05,
+      "loss": 0.6097,
+      "step": 790
+    },
+    {
+      "epoch": 4.087741935483871,
+      "grad_norm": 0.764533281326294,
+      "learning_rate": 3.625e-05,
+      "loss": 0.5912,
+      "step": 791
+    },
+    {
+      "epoch": 4.0929032258064515,
+      "grad_norm": 0.7470724582672119,
+      "learning_rate": 3.604166666666667e-05,
+      "loss": 0.619,
+      "step": 792
+    },
+    {
+      "epoch": 4.098064516129032,
+      "grad_norm": 0.7989603877067566,
+      "learning_rate": 3.5833333333333335e-05,
+      "loss": 0.6039,
+      "step": 793
+    },
+    {
+      "epoch": 4.103225806451613,
+      "grad_norm": 0.8116969466209412,
+      "learning_rate": 3.5625000000000005e-05,
+      "loss": 0.6019,
+      "step": 794
+    },
+    {
+      "epoch": 4.108387096774194,
+      "grad_norm": 0.7786195278167725,
+      "learning_rate": 3.541666666666667e-05,
+      "loss": 0.6002,
+      "step": 795
+    },
+    {
+      "epoch": 4.113548387096774,
+      "grad_norm": 0.84043288230896,
+      "learning_rate": 3.520833333333334e-05,
+      "loss": 0.5855,
+      "step": 796
+    },
+    {
+      "epoch": 4.118709677419355,
+      "grad_norm": 0.9318393468856812,
+      "learning_rate": 3.5e-05,
+      "loss": 0.6142,
+      "step": 797
+    },
+    {
+      "epoch": 4.123870967741936,
+      "grad_norm": 0.9443640112876892,
+      "learning_rate": 3.479166666666667e-05,
+      "loss": 0.5898,
+      "step": 798
+    },
+    {
+      "epoch": 4.129032258064516,
+      "grad_norm": 0.8349962830543518,
+      "learning_rate": 3.458333333333333e-05,
+      "loss": 0.5805,
+      "step": 799
+    },
+    {
+      "epoch": 4.134193548387096,
+      "grad_norm": 0.9303920269012451,
+      "learning_rate": 3.4375e-05,
+      "loss": 0.6077,
+      "step": 800
+    },
+    {
+      "epoch": 4.139354838709678,
+      "grad_norm": 0.8999093770980835,
+      "learning_rate": 3.4166666666666666e-05,
+      "loss": 0.6331,
+      "step": 801
+    },
+    {
+      "epoch": 4.144516129032258,
+      "grad_norm": 0.8509075045585632,
+      "learning_rate": 3.3958333333333337e-05,
+      "loss": 0.6006,
+      "step": 802
+    },
+    {
+      "epoch": 4.1496774193548385,
+      "grad_norm": 0.8161064386367798,
+      "learning_rate": 3.375000000000001e-05,
+      "loss": 0.5907,
+      "step": 803
+    },
+    {
+      "epoch": 4.15483870967742,
+      "grad_norm": 0.7541460394859314,
+      "learning_rate": 3.3541666666666664e-05,
+      "loss": 0.5877,
+      "step": 804
+    },
+    {
+      "epoch": 4.16,
+      "grad_norm": 0.7236863970756531,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 0.5732,
+      "step": 805
+    },
+    {
+      "epoch": 4.1651612903225805,
+      "grad_norm": 0.7611268758773804,
+      "learning_rate": 3.3125e-05,
+      "loss": 0.6233,
+      "step": 806
+    },
+    {
+      "epoch": 4.170322580645161,
+      "grad_norm": 0.7469662427902222,
+      "learning_rate": 3.291666666666667e-05,
+      "loss": 0.5925,
+      "step": 807
+    },
+    {
+      "epoch": 4.175483870967742,
+      "grad_norm": 0.8399537205696106,
+      "learning_rate": 3.270833333333333e-05,
+      "loss": 0.5773,
+      "step": 808
+    },
+    {
+      "epoch": 4.180645161290323,
+      "grad_norm": 0.8059346079826355,
+      "learning_rate": 3.2500000000000004e-05,
+      "loss": 0.6012,
+      "step": 809
+    },
+    {
+      "epoch": 4.185806451612903,
+      "grad_norm": 0.7592855095863342,
+      "learning_rate": 3.229166666666667e-05,
+      "loss": 0.5922,
+      "step": 810
+    },
+    {
+      "epoch": 4.190967741935484,
+      "grad_norm": 0.7669108510017395,
+      "learning_rate": 3.208333333333334e-05,
+      "loss": 0.556,
+      "step": 811
+    },
+    {
+      "epoch": 4.196129032258065,
+      "grad_norm": 0.8931100368499756,
+      "learning_rate": 3.1875e-05,
+      "loss": 0.6591,
+      "step": 812
+    },
+    {
+      "epoch": 4.201290322580645,
+      "grad_norm": 1.8222947120666504,
+      "learning_rate": 3.1666666666666666e-05,
+      "loss": 0.603,
+      "step": 813
+    },
+    {
+      "epoch": 4.2064516129032254,
+      "grad_norm": 0.8253236413002014,
+      "learning_rate": 3.145833333333334e-05,
+      "loss": 0.6121,
+      "step": 814
+    },
+    {
+      "epoch": 4.211612903225807,
+      "grad_norm": 0.7631076574325562,
+      "learning_rate": 3.125e-05,
+      "loss": 0.5689,
+      "step": 815
+    },
+    {
+      "epoch": 4.216774193548387,
+      "grad_norm": 0.818957507610321,
+      "learning_rate": 3.104166666666667e-05,
+      "loss": 0.5641,
+      "step": 816
+    },
+    {
+      "epoch": 4.2219354838709675,
+      "grad_norm": 0.8815581202507019,
+      "learning_rate": 3.0833333333333335e-05,
+      "loss": 0.5241,
+      "step": 817
+    },
+    {
+      "epoch": 4.227096774193549,
+      "grad_norm": 0.8291641473770142,
+      "learning_rate": 3.0625000000000006e-05,
+      "loss": 0.641,
+      "step": 818
+    },
+    {
+      "epoch": 4.232258064516129,
+      "grad_norm": 0.8939973711967468,
+      "learning_rate": 3.0416666666666666e-05,
+      "loss": 0.5443,
+      "step": 819
+    },
+    {
+      "epoch": 4.23741935483871,
+      "grad_norm": 0.8753037452697754,
+      "learning_rate": 3.0208333333333334e-05,
+      "loss": 0.6129,
+      "step": 820
+    },
+    {
+      "epoch": 4.24258064516129,
+      "grad_norm": 0.7449681758880615,
+      "learning_rate": 3e-05,
+      "loss": 0.5261,
+      "step": 821
+    },
+    {
+      "epoch": 4.247741935483871,
+      "grad_norm": 0.7956663966178894,
+      "learning_rate": 2.9791666666666668e-05,
+      "loss": 0.5896,
+      "step": 822
+    },
+    {
+      "epoch": 4.252903225806452,
+      "grad_norm": 0.8466067314147949,
+      "learning_rate": 2.9583333333333335e-05,
+      "loss": 0.558,
+      "step": 823
+    },
+    {
+      "epoch": 4.258064516129032,
+      "grad_norm": 0.7523233890533447,
+      "learning_rate": 2.9375000000000003e-05,
+      "loss": 0.5911,
+      "step": 824
+    },
+    {
+      "epoch": 4.263225806451613,
+      "grad_norm": 0.8161501884460449,
+      "learning_rate": 2.916666666666667e-05,
+      "loss": 0.6066,
+      "step": 825
+    },
+    {
+      "epoch": 4.268387096774194,
+      "grad_norm": 0.813704788684845,
+      "learning_rate": 2.8958333333333337e-05,
+      "loss": 0.5966,
+      "step": 826
+    },
+    {
+      "epoch": 4.273548387096774,
+      "grad_norm": 0.8174046277999878,
+      "learning_rate": 2.8749999999999997e-05,
+      "loss": 0.5265,
+      "step": 827
+    },
+    {
+      "epoch": 4.2787096774193545,
+      "grad_norm": 0.8048108816146851,
+      "learning_rate": 2.8541666666666668e-05,
+      "loss": 0.6069,
+      "step": 828
+    },
+    {
+      "epoch": 4.283870967741936,
+      "grad_norm": 0.7996951341629028,
+      "learning_rate": 2.8333333333333335e-05,
+      "loss": 0.6103,
+      "step": 829
+    },
+    {
+      "epoch": 4.289032258064516,
+      "grad_norm": 0.8226521015167236,
+      "learning_rate": 2.8125000000000003e-05,
+      "loss": 0.6079,
+      "step": 830
+    },
+    {
+      "epoch": 4.2941935483870965,
+      "grad_norm": 0.9158440232276917,
+      "learning_rate": 2.791666666666667e-05,
+      "loss": 0.6532,
+      "step": 831
+    },
+    {
+      "epoch": 4.299354838709678,
+      "grad_norm": 0.7324296832084656,
+      "learning_rate": 2.7708333333333337e-05,
+      "loss": 0.5473,
+      "step": 832
+    },
+    {
+      "epoch": 4.304516129032258,
+      "grad_norm": 0.9104575514793396,
+      "learning_rate": 2.7500000000000004e-05,
+      "loss": 0.664,
+      "step": 833
+    },
+    {
+      "epoch": 4.309677419354839,
+      "grad_norm": 0.8640415072441101,
+      "learning_rate": 2.7291666666666665e-05,
+      "loss": 0.5982,
+      "step": 834
+    },
+    {
+      "epoch": 4.314838709677419,
+      "grad_norm": 0.8377175331115723,
+      "learning_rate": 2.7083333333333332e-05,
+      "loss": 0.6323,
+      "step": 835
+    },
+    {
+      "epoch": 4.32,
+      "grad_norm": 0.8285635113716125,
+      "learning_rate": 2.6875e-05,
+      "loss": 0.5809,
+      "step": 836
+    },
+    {
+      "epoch": 4.325161290322581,
+      "grad_norm": 0.8491920828819275,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 0.5809,
+      "step": 837
+    },
+    {
+      "epoch": 4.330322580645161,
+      "grad_norm": 0.8195666670799255,
+      "learning_rate": 2.6458333333333334e-05,
+      "loss": 0.5565,
+      "step": 838
+    },
+    {
+      "epoch": 4.335483870967742,
+      "grad_norm": 0.8263185620307922,
+      "learning_rate": 2.625e-05,
+      "loss": 0.6303,
+      "step": 839
+    },
+    {
+      "epoch": 4.340645161290323,
+      "grad_norm": 0.8749952912330627,
+      "learning_rate": 2.604166666666667e-05,
+      "loss": 0.6308,
+      "step": 840
+    },
+    {
+      "epoch": 4.345806451612903,
+      "grad_norm": 0.8245086669921875,
+      "learning_rate": 2.5833333333333336e-05,
+      "loss": 0.622,
+      "step": 841
+    },
+    {
+      "epoch": 4.3509677419354835,
+      "grad_norm": 0.782157301902771,
+      "learning_rate": 2.5625e-05,
+      "loss": 0.5874,
+      "step": 842
+    },
+    {
+      "epoch": 4.356129032258065,
+      "grad_norm": 0.7814539670944214,
+      "learning_rate": 2.5416666666666667e-05,
+      "loss": 0.6185,
+      "step": 843
+    },
+    {
+      "epoch": 4.361290322580645,
+      "grad_norm": 0.7956915497779846,
+      "learning_rate": 2.5208333333333334e-05,
+      "loss": 0.6008,
+      "step": 844
+    },
+    {
+      "epoch": 4.366451612903226,
+      "grad_norm": 0.7877438068389893,
+      "learning_rate": 2.5e-05,
+      "loss": 0.5152,
+      "step": 845
+    },
+    {
+      "epoch": 4.371612903225807,
+      "grad_norm": 0.8702367544174194,
+      "learning_rate": 2.479166666666667e-05,
+      "loss": 0.5647,
+      "step": 846
+    },
+    {
+      "epoch": 4.376774193548387,
+      "grad_norm": 0.8250179886817932,
+      "learning_rate": 2.4583333333333332e-05,
+      "loss": 0.604,
+      "step": 847
+    },
+    {
+      "epoch": 4.381935483870968,
+      "grad_norm": 0.7943134903907776,
+      "learning_rate": 2.4375e-05,
+      "loss": 0.5615,
+      "step": 848
+    },
+    {
+      "epoch": 4.387096774193548,
+      "grad_norm": 0.8098966479301453,
+      "learning_rate": 2.4166666666666667e-05,
+      "loss": 0.6218,
+      "step": 849
+    },
+    {
+      "epoch": 4.392258064516129,
+      "grad_norm": 0.8729791045188904,
+      "learning_rate": 2.3958333333333334e-05,
+      "loss": 0.6122,
+      "step": 850
+    },
+    {
+      "epoch": 4.39741935483871,
+      "grad_norm": 0.8003780841827393,
+      "learning_rate": 2.375e-05,
+      "loss": 0.6132,
+      "step": 851
+    },
+    {
+      "epoch": 4.40258064516129,
+      "grad_norm": 0.7254167199134827,
+      "learning_rate": 2.354166666666667e-05,
+      "loss": 0.4762,
+      "step": 852
+    },
+    {
+      "epoch": 4.407741935483871,
+      "grad_norm": 0.8480260968208313,
+      "learning_rate": 2.3333333333333336e-05,
+      "loss": 0.6229,
+      "step": 853
+    },
+    {
+      "epoch": 4.412903225806452,
+      "grad_norm": 0.8236071467399597,
+      "learning_rate": 2.3125000000000003e-05,
+      "loss": 0.5863,
+      "step": 854
+    },
+    {
+      "epoch": 4.418064516129032,
+      "grad_norm": 0.8940121531486511,
+      "learning_rate": 2.2916666666666667e-05,
+      "loss": 0.5913,
+      "step": 855
+    },
+    {
+      "epoch": 4.423225806451613,
+      "grad_norm": 0.8161671757698059,
+      "learning_rate": 2.2708333333333334e-05,
+      "loss": 0.5422,
+      "step": 856
+    },
+    {
+      "epoch": 4.428387096774194,
+      "grad_norm": 0.8735789060592651,
+      "learning_rate": 2.25e-05,
+      "loss": 0.6146,
+      "step": 857
+    },
+    {
+      "epoch": 4.433548387096774,
+      "grad_norm": 0.8856715559959412,
+      "learning_rate": 2.229166666666667e-05,
+      "loss": 0.5712,
+      "step": 858
+    },
+    {
+      "epoch": 4.438709677419355,
+      "grad_norm": 0.9326929450035095,
+      "learning_rate": 2.2083333333333333e-05,
+      "loss": 0.6325,
+      "step": 859
+    },
+    {
+      "epoch": 4.443870967741935,
+      "grad_norm": 0.9000937342643738,
+      "learning_rate": 2.1875e-05,
+      "loss": 0.575,
+      "step": 860
+    },
+    {
+      "epoch": 4.449032258064516,
+      "grad_norm": 0.8427292704582214,
+      "learning_rate": 2.1666666666666667e-05,
+      "loss": 0.5844,
+      "step": 861
+    },
+    {
+      "epoch": 4.454193548387097,
+      "grad_norm": 0.8438614010810852,
+      "learning_rate": 2.1458333333333334e-05,
+      "loss": 0.6591,
+      "step": 862
+    },
+    {
+      "epoch": 4.459354838709677,
+      "grad_norm": 0.7900532484054565,
+      "learning_rate": 2.125e-05,
+      "loss": 0.6021,
+      "step": 863
+    },
+    {
+      "epoch": 4.464516129032258,
+      "grad_norm": 0.7891470193862915,
+      "learning_rate": 2.104166666666667e-05,
+      "loss": 0.6067,
+      "step": 864
+    },
+    {
+      "epoch": 4.469677419354839,
+      "grad_norm": 0.8239520788192749,
+      "learning_rate": 2.0833333333333336e-05,
+      "loss": 0.6052,
+      "step": 865
+    },
+    {
+      "epoch": 4.474838709677419,
+      "grad_norm": 0.9424383044242859,
+      "learning_rate": 2.0625e-05,
+      "loss": 0.6317,
+      "step": 866
+    },
+    {
+      "epoch": 4.48,
+      "grad_norm": 0.7990756034851074,
+      "learning_rate": 2.0416666666666667e-05,
+      "loss": 0.5959,
+      "step": 867
+    },
+    {
+      "epoch": 4.485161290322581,
+      "grad_norm": 0.8171214461326599,
+      "learning_rate": 2.0208333333333334e-05,
+      "loss": 0.64,
+      "step": 868
+    },
+    {
+      "epoch": 4.490322580645161,
+      "grad_norm": 0.874057948589325,
+      "learning_rate": 2e-05,
+      "loss": 0.6379,
+      "step": 869
+    },
+    {
+      "epoch": 4.495483870967742,
+      "grad_norm": 0.8302012085914612,
+      "learning_rate": 1.9791666666666665e-05,
+      "loss": 0.6163,
+      "step": 870
+    },
+    {
+      "epoch": 4.495483870967742,
+      "eval_loss": 1.8783903121948242,
+      "eval_runtime": 21.2871,
+      "eval_samples_per_second": 3.711,
+      "eval_steps_per_second": 0.47,
+      "step": 870
+    },
+    {
+      "epoch": 4.500645161290323,
+      "grad_norm": 0.7921658754348755,
+      "learning_rate": 1.9583333333333333e-05,
+      "loss": 0.5737,
+      "step": 871
+    },
+    {
+      "epoch": 4.505806451612903,
+      "grad_norm": 0.810171902179718,
+      "learning_rate": 1.9375e-05,
+      "loss": 0.6025,
+      "step": 872
+    },
+    {
+      "epoch": 4.510967741935484,
+      "grad_norm": 0.8471074104309082,
+      "learning_rate": 1.9166666666666667e-05,
+      "loss": 0.5951,
+      "step": 873
+    },
+    {
+      "epoch": 4.516129032258064,
+      "grad_norm": 0.824587345123291,
+      "learning_rate": 1.8958333333333334e-05,
+      "loss": 0.6245,
+      "step": 874
+    },
+    {
+      "epoch": 4.521290322580645,
+      "grad_norm": 0.8031266927719116,
+      "learning_rate": 1.8750000000000002e-05,
+      "loss": 0.6097,
+      "step": 875
+    },
+    {
+      "epoch": 4.526451612903226,
+      "grad_norm": 0.8272472620010376,
+      "learning_rate": 1.854166666666667e-05,
+      "loss": 0.5758,
+      "step": 876
+    },
+    {
+      "epoch": 4.531612903225806,
+      "grad_norm": 0.8287191390991211,
+      "learning_rate": 1.8333333333333333e-05,
+      "loss": 0.5804,
+      "step": 877
+    },
+    {
+      "epoch": 4.536774193548387,
+      "grad_norm": 0.8832491636276245,
+      "learning_rate": 1.8125e-05,
+      "loss": 0.5911,
+      "step": 878
+    },
+    {
+      "epoch": 4.541935483870968,
+      "grad_norm": 0.8580238819122314,
+      "learning_rate": 1.7916666666666667e-05,
+      "loss": 0.6071,
+      "step": 879
+    },
+    {
+      "epoch": 4.547096774193548,
+      "grad_norm": 0.7854690551757812,
+      "learning_rate": 1.7708333333333335e-05,
+      "loss": 0.5786,
+      "step": 880
+    },
+    {
+      "epoch": 4.5522580645161295,
+      "grad_norm": 0.9048441648483276,
+      "learning_rate": 1.75e-05,
+      "loss": 0.6036,
+      "step": 881
+    },
+    {
+      "epoch": 4.55741935483871,
+      "grad_norm": 0.7667028903961182,
+      "learning_rate": 1.7291666666666666e-05,
+      "loss": 0.5279,
+      "step": 882
+    },
+    {
+      "epoch": 4.56258064516129,
+      "grad_norm": 0.8021149039268494,
+      "learning_rate": 1.7083333333333333e-05,
+      "loss": 0.5538,
+      "step": 883
+    },
+    {
+      "epoch": 4.567741935483871,
+      "grad_norm": 0.8307644724845886,
+      "learning_rate": 1.6875000000000004e-05,
+      "loss": 0.5672,
+      "step": 884
+    },
+    {
+      "epoch": 4.572903225806452,
+      "grad_norm": 0.8047096729278564,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 0.561,
+      "step": 885
+    },
+    {
+      "epoch": 4.578064516129032,
+      "grad_norm": 0.8412072658538818,
+      "learning_rate": 1.6458333333333335e-05,
+      "loss": 0.5153,
+      "step": 886
+    },
+    {
+      "epoch": 4.583225806451613,
+      "grad_norm": 0.9264989495277405,
+      "learning_rate": 1.6250000000000002e-05,
+      "loss": 0.5846,
+      "step": 887
+    },
+    {
+      "epoch": 4.588387096774193,
+      "grad_norm": 0.9487907886505127,
+      "learning_rate": 1.604166666666667e-05,
+      "loss": 0.5986,
+      "step": 888
+    },
+    {
+      "epoch": 4.593548387096774,
+      "grad_norm": 0.8134511113166809,
+      "learning_rate": 1.5833333333333333e-05,
+      "loss": 0.6006,
+      "step": 889
+    },
+    {
+      "epoch": 4.598709677419355,
+      "grad_norm": 0.7911557555198669,
+      "learning_rate": 1.5625e-05,
+      "loss": 0.6034,
+      "step": 890
+    },
+    {
+      "epoch": 4.603870967741935,
+      "grad_norm": 0.9281388521194458,
+      "learning_rate": 1.5416666666666668e-05,
+      "loss": 0.605,
+      "step": 891
+    },
+    {
+      "epoch": 4.609032258064516,
+      "grad_norm": 0.8300231099128723,
+      "learning_rate": 1.5208333333333333e-05,
+      "loss": 0.5389,
+      "step": 892
+    },
+    {
+      "epoch": 4.614193548387097,
+      "grad_norm": 0.8294689655303955,
+      "learning_rate": 1.5e-05,
+      "loss": 0.6089,
+      "step": 893
+    },
+    {
+      "epoch": 4.619354838709677,
+      "grad_norm": 0.8062645196914673,
+      "learning_rate": 1.4791666666666668e-05,
+      "loss": 0.6033,
+      "step": 894
+    },
+    {
+      "epoch": 4.6245161290322585,
+      "grad_norm": 0.8522666096687317,
+      "learning_rate": 1.4583333333333335e-05,
+      "loss": 0.6023,
+      "step": 895
+    },
+    {
+      "epoch": 4.629677419354839,
+      "grad_norm": 0.809053897857666,
+      "learning_rate": 1.4374999999999999e-05,
+      "loss": 0.5612,
+      "step": 896
+    },
+    {
+      "epoch": 4.634838709677419,
+      "grad_norm": 0.7808062434196472,
+      "learning_rate": 1.4166666666666668e-05,
+      "loss": 0.5934,
+      "step": 897
+    },
+    {
+      "epoch": 4.64,
+      "grad_norm": 0.8294617533683777,
+      "learning_rate": 1.3958333333333335e-05,
+      "loss": 0.5886,
+      "step": 898
+    },
+    {
+      "epoch": 4.645161290322581,
+      "grad_norm": 0.7806969285011292,
+      "learning_rate": 1.3750000000000002e-05,
+      "loss": 0.5924,
+      "step": 899
+    },
+    {
+      "epoch": 4.650322580645161,
+      "grad_norm": 0.8098063468933105,
+      "learning_rate": 1.3541666666666666e-05,
+      "loss": 0.6281,
+      "step": 900
+    },
+    {
+      "epoch": 4.655483870967742,
+      "grad_norm": 0.7569921016693115,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.5425,
+      "step": 901
+    },
+    {
+      "epoch": 4.660645161290322,
+      "grad_norm": 0.8233391046524048,
+      "learning_rate": 1.3125e-05,
+      "loss": 0.5719,
+      "step": 902
+    },
+    {
+      "epoch": 4.665806451612903,
+      "grad_norm": 0.8115103840827942,
+      "learning_rate": 1.2916666666666668e-05,
+      "loss": 0.5819,
+      "step": 903
+    },
+    {
+      "epoch": 4.670967741935484,
+      "grad_norm": 0.8324194550514221,
+      "learning_rate": 1.2708333333333333e-05,
+      "loss": 0.573,
+      "step": 904
+    },
+    {
+      "epoch": 4.676129032258064,
+      "grad_norm": 0.9103081822395325,
+      "learning_rate": 1.25e-05,
+      "loss": 0.6276,
+      "step": 905
+    },
+    {
+      "epoch": 4.6812903225806455,
+      "grad_norm": 0.8303936719894409,
+      "learning_rate": 1.2291666666666666e-05,
+      "loss": 0.5962,
+      "step": 906
+    },
+    {
+      "epoch": 4.686451612903226,
+      "grad_norm": 0.814725399017334,
+      "learning_rate": 1.2083333333333333e-05,
+      "loss": 0.57,
+      "step": 907
+    },
+    {
+      "epoch": 4.691612903225806,
+      "grad_norm": 0.8399000763893127,
+      "learning_rate": 1.1875e-05,
+      "loss": 0.5888,
+      "step": 908
+    },
+    {
+      "epoch": 4.6967741935483875,
+      "grad_norm": 0.8585182428359985,
+      "learning_rate": 1.1666666666666668e-05,
+      "loss": 0.5976,
+      "step": 909
+    },
+    {
+      "epoch": 4.701935483870968,
+      "grad_norm": 0.8289327621459961,
+      "learning_rate": 1.1458333333333333e-05,
+      "loss": 0.5564,
+      "step": 910
+    },
+    {
+      "epoch": 4.707096774193548,
+      "grad_norm": 1.0039379596710205,
+      "learning_rate": 1.125e-05,
+      "loss": 0.6359,
+      "step": 911
+    },
+    {
+      "epoch": 4.712258064516129,
+      "grad_norm": 0.8319773077964783,
+      "learning_rate": 1.1041666666666666e-05,
+      "loss": 0.5852,
+      "step": 912
+    },
+    {
+      "epoch": 4.71741935483871,
+      "grad_norm": 0.838598906993866,
+      "learning_rate": 1.0833333333333334e-05,
+      "loss": 0.6005,
+      "step": 913
+    },
+    {
+      "epoch": 4.72258064516129,
+      "grad_norm": 0.8527039885520935,
+      "learning_rate": 1.0625e-05,
+      "loss": 0.6233,
+      "step": 914
+    },
+    {
+      "epoch": 4.727741935483871,
+      "grad_norm": 0.8852501511573792,
+      "learning_rate": 1.0416666666666668e-05,
+      "loss": 0.5913,
+      "step": 915
+    },
+    {
+      "epoch": 4.732903225806451,
+      "grad_norm": 0.8244696259498596,
+      "learning_rate": 1.0208333333333334e-05,
+      "loss": 0.6458,
+      "step": 916
+    },
+    {
+      "epoch": 4.7380645161290325,
+      "grad_norm": 0.8813903331756592,
+      "learning_rate": 1e-05,
+      "loss": 0.6293,
+      "step": 917
+    },
+    {
+      "epoch": 4.743225806451613,
+      "grad_norm": 0.8205885887145996,
+      "learning_rate": 9.791666666666666e-06,
+      "loss": 0.6214,
+      "step": 918
+    },
+    {
+      "epoch": 4.748387096774193,
+      "grad_norm": 0.8197274208068848,
+      "learning_rate": 9.583333333333334e-06,
+      "loss": 0.5808,
+      "step": 919
+    },
+    {
+      "epoch": 4.7535483870967745,
+      "grad_norm": 0.8159494400024414,
+      "learning_rate": 9.375000000000001e-06,
+      "loss": 0.5486,
+      "step": 920
+    },
+    {
+      "epoch": 4.758709677419355,
+      "grad_norm": 0.7876229882240295,
+      "learning_rate": 9.166666666666666e-06,
+      "loss": 0.6191,
+      "step": 921
+    },
+    {
+      "epoch": 4.763870967741935,
+      "grad_norm": 0.7964573502540588,
+      "learning_rate": 8.958333333333334e-06,
+      "loss": 0.568,
+      "step": 922
+    },
+    {
+      "epoch": 4.769032258064517,
+      "grad_norm": 0.9539830088615417,
+      "learning_rate": 8.75e-06,
+      "loss": 0.6531,
+      "step": 923
+    },
+    {
+      "epoch": 4.774193548387097,
+      "grad_norm": 0.8864449858665466,
+      "learning_rate": 8.541666666666666e-06,
+      "loss": 0.6329,
+      "step": 924
+    },
+    {
+      "epoch": 4.779354838709677,
+      "grad_norm": 0.8039810657501221,
+      "learning_rate": 8.333333333333334e-06,
+      "loss": 0.6192,
+      "step": 925
+    },
+    {
+      "epoch": 4.784516129032258,
+      "grad_norm": 0.9101413488388062,
+      "learning_rate": 8.125000000000001e-06,
+      "loss": 0.639,
+      "step": 926
+    },
+    {
+      "epoch": 4.789677419354839,
+      "grad_norm": 0.8219101428985596,
+      "learning_rate": 7.916666666666667e-06,
+      "loss": 0.6097,
+      "step": 927
+    },
+    {
+      "epoch": 4.794838709677419,
+      "grad_norm": 0.8232947587966919,
+      "learning_rate": 7.708333333333334e-06,
+      "loss": 0.5598,
+      "step": 928
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 0.7906395792961121,
+      "learning_rate": 7.5e-06,
+      "loss": 0.6011,
+      "step": 929
+    },
+    {
+      "epoch": 4.80516129032258,
+      "grad_norm": 0.7940424680709839,
+      "learning_rate": 7.2916666666666674e-06,
+      "loss": 0.624,
+      "step": 930
+    },
+    {
+      "epoch": 4.8103225806451615,
+      "grad_norm": 0.8177192211151123,
+      "learning_rate": 7.083333333333334e-06,
+      "loss": 0.5818,
+      "step": 931
+    },
+    {
+      "epoch": 4.815483870967742,
+      "grad_norm": 0.7968092560768127,
+      "learning_rate": 6.875000000000001e-06,
+      "loss": 0.5649,
+      "step": 932
+    },
+    {
+      "epoch": 4.820645161290322,
+      "grad_norm": 0.8273805379867554,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 0.5643,
+      "step": 933
+    },
+    {
+      "epoch": 4.825806451612904,
+      "grad_norm": 0.8781915903091431,
+      "learning_rate": 6.458333333333334e-06,
+      "loss": 0.6222,
+      "step": 934
+    },
+    {
+      "epoch": 4.830967741935484,
+      "grad_norm": 0.7829635739326477,
+      "learning_rate": 6.25e-06,
+      "loss": 0.5446,
+      "step": 935
+    },
+    {
+      "epoch": 4.836129032258064,
+      "grad_norm": 0.8310282826423645,
+      "learning_rate": 6.041666666666667e-06,
+      "loss": 0.6216,
+      "step": 936
+    },
+    {
+      "epoch": 4.841290322580646,
+      "grad_norm": 0.8109216690063477,
+      "learning_rate": 5.833333333333334e-06,
+      "loss": 0.5673,
+      "step": 937
+    },
+    {
+      "epoch": 4.846451612903226,
+      "grad_norm": 0.8850212097167969,
+      "learning_rate": 5.625e-06,
+      "loss": 0.6311,
+      "step": 938
+    },
+    {
+      "epoch": 4.851612903225806,
+      "grad_norm": 0.8448730707168579,
+      "learning_rate": 5.416666666666667e-06,
+      "loss": 0.6392,
+      "step": 939
+    },
+    {
+      "epoch": 4.856774193548387,
+      "grad_norm": 0.7985132336616516,
+      "learning_rate": 5.208333333333334e-06,
+      "loss": 0.6147,
+      "step": 940
+    },
+    {
+      "epoch": 4.861935483870968,
+      "grad_norm": 0.8426609039306641,
+      "learning_rate": 5e-06,
+      "loss": 0.6123,
+      "step": 941
+    },
+    {
+      "epoch": 4.8670967741935485,
+      "grad_norm": 0.8368214964866638,
+      "learning_rate": 4.791666666666667e-06,
+      "loss": 0.4831,
+      "step": 942
+    },
+    {
+      "epoch": 4.872258064516129,
+      "grad_norm": 0.7565503120422363,
+      "learning_rate": 4.583333333333333e-06,
+      "loss": 0.5219,
+      "step": 943
+    },
+    {
+      "epoch": 4.877419354838709,
+      "grad_norm": 0.8006893992424011,
+      "learning_rate": 4.375e-06,
+      "loss": 0.5173,
+      "step": 944
+    },
+    {
+      "epoch": 4.8825806451612905,
+      "grad_norm": 0.8365283012390137,
+      "learning_rate": 4.166666666666667e-06,
+      "loss": 0.6264,
+      "step": 945
+    },
+    {
+      "epoch": 4.887741935483871,
+      "grad_norm": 0.817327082157135,
+      "learning_rate": 3.958333333333333e-06,
+      "loss": 0.6199,
+      "step": 946
+    },
+    {
+      "epoch": 4.892903225806451,
+      "grad_norm": 0.8059232831001282,
+      "learning_rate": 3.75e-06,
+      "loss": 0.5594,
+      "step": 947
+    },
+    {
+      "epoch": 4.898064516129033,
+      "grad_norm": 0.836398184299469,
+      "learning_rate": 3.541666666666667e-06,
+      "loss": 0.6177,
+      "step": 948
+    },
+    {
+      "epoch": 4.903225806451613,
+      "grad_norm": 0.8153460621833801,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 0.5891,
+      "step": 949
+    },
+    {
+      "epoch": 4.908387096774193,
+      "grad_norm": 0.9284054040908813,
+      "learning_rate": 3.125e-06,
+      "loss": 0.6283,
+      "step": 950
+    },
+    {
+      "epoch": 4.913548387096775,
+      "grad_norm": 0.8782488703727722,
+      "learning_rate": 2.916666666666667e-06,
+      "loss": 0.5855,
+      "step": 951
+    },
+    {
+      "epoch": 4.918709677419355,
+      "grad_norm": 0.8501858115196228,
+      "learning_rate": 2.7083333333333334e-06,
+      "loss": 0.5643,
+      "step": 952
+    },
+    {
+      "epoch": 4.9238709677419354,
+      "grad_norm": 0.838408350944519,
+      "learning_rate": 2.5e-06,
+      "loss": 0.6152,
+      "step": 953
+    },
+    {
+      "epoch": 4.929032258064516,
+      "grad_norm": 0.8043311834335327,
+      "learning_rate": 2.2916666666666666e-06,
+      "loss": 0.5521,
+      "step": 954
+    },
+    {
+      "epoch": 4.934193548387097,
+      "grad_norm": 0.8513094782829285,
+      "learning_rate": 2.0833333333333334e-06,
+      "loss": 0.5155,
+      "step": 955
+    },
+    {
+      "epoch": 4.9393548387096775,
+      "grad_norm": 0.8407511115074158,
+      "learning_rate": 1.875e-06,
+      "loss": 0.5719,
+      "step": 956
+    },
+    {
+      "epoch": 4.944516129032258,
+      "grad_norm": 0.8200874924659729,
+      "learning_rate": 1.6666666666666667e-06,
+      "loss": 0.6213,
+      "step": 957
+    },
+    {
+      "epoch": 4.949677419354838,
+      "grad_norm": 0.8391759395599365,
+      "learning_rate": 1.4583333333333335e-06,
+      "loss": 0.6061,
+      "step": 958
+    },
+    {
+      "epoch": 4.95483870967742,
+      "grad_norm": 0.8139519095420837,
+      "learning_rate": 1.25e-06,
+      "loss": 0.6191,
+      "step": 959
+    },
+    {
+      "epoch": 4.96,
+      "grad_norm": 0.8107402324676514,
+      "learning_rate": 1.0416666666666667e-06,
+      "loss": 0.585,
+      "step": 960
+    },
+    {
+      "epoch": 4.96516129032258,
+      "grad_norm": 0.8392252326011658,
+      "learning_rate": 8.333333333333333e-07,
+      "loss": 0.6128,
+      "step": 961
+    },
+    {
+      "epoch": 4.970322580645162,
+      "grad_norm": 0.9094356298446655,
+      "learning_rate": 6.25e-07,
+      "loss": 0.5626,
+      "step": 962
+    },
+    {
+      "epoch": 4.975483870967742,
+      "grad_norm": 0.8394489884376526,
+      "learning_rate": 4.1666666666666667e-07,
+      "loss": 0.6437,
+      "step": 963
+    },
+    {
+      "epoch": 4.980645161290322,
+      "grad_norm": 0.7896241545677185,
+      "learning_rate": 2.0833333333333333e-07,
+      "loss": 0.6175,
+      "step": 964
+    },
+    {
+      "epoch": 4.985806451612904,
+      "grad_norm": 0.7650795578956604,
+      "learning_rate": 0.0,
+      "loss": 0.5858,
+      "step": 965
+    },
+    {
+      "epoch": 5.00516129032258,
+      "grad_norm": 0.6613453030586243,
+      "learning_rate": 0.00016677614520311151,
+      "loss": 0.5455,
+      "step": 966
+    },
+    {
+      "epoch": 5.010322580645162,
+      "grad_norm": 0.6841882467269897,
+      "learning_rate": 0.00016674157303370787,
+      "loss": 0.5692,
+      "step": 967
+    },
+    {
+      "epoch": 5.015483870967742,
+      "grad_norm": 0.8633980751037598,
+      "learning_rate": 0.00016670700086430425,
+      "loss": 0.5726,
+      "step": 968
+    },
+    {
+      "epoch": 5.0206451612903225,
+      "grad_norm": 1.2848694324493408,
+      "learning_rate": 0.0001666724286949006,
+      "loss": 0.5086,
+      "step": 969
+    },
+    {
+      "epoch": 5.025806451612903,
+      "grad_norm": 1.1172974109649658,
+      "learning_rate": 0.00016663785652549698,
+      "loss": 0.4798,
+      "step": 970
+    },
+    {
+      "epoch": 5.030967741935484,
+      "grad_norm": 0.9701982736587524,
+      "learning_rate": 0.00016660328435609336,
+      "loss": 0.5126,
+      "step": 971
+    },
+    {
+      "epoch": 5.0361290322580645,
+      "grad_norm": 0.7883008718490601,
+      "learning_rate": 0.00016656871218668972,
+      "loss": 0.532,
+      "step": 972
+    },
+    {
+      "epoch": 5.041290322580645,
+      "grad_norm": 0.7259140610694885,
+      "learning_rate": 0.0001665341400172861,
+      "loss": 0.5471,
+      "step": 973
+    },
+    {
+      "epoch": 5.046451612903226,
+      "grad_norm": 0.81528240442276,
+      "learning_rate": 0.00016649956784788245,
+      "loss": 0.5357,
+      "step": 974
+    },
+    {
+      "epoch": 5.051612903225807,
+      "grad_norm": 1.1066521406173706,
+      "learning_rate": 0.00016646499567847883,
+      "loss": 0.5835,
+      "step": 975
+    },
+    {
+      "epoch": 5.056774193548387,
+      "grad_norm": 0.9075446128845215,
+      "learning_rate": 0.0001664304235090752,
+      "loss": 0.4751,
+      "step": 976
+    },
+    {
+      "epoch": 5.061935483870967,
+      "grad_norm": 1.1995927095413208,
+      "learning_rate": 0.00016639585133967157,
+      "loss": 0.5776,
+      "step": 977
+    },
+    {
+      "epoch": 5.067096774193549,
+      "grad_norm": 1.0916688442230225,
+      "learning_rate": 0.00016636127917026795,
+      "loss": 0.5083,
+      "step": 978
+    },
+    {
+      "epoch": 5.072258064516129,
+      "grad_norm": 2.419167995452881,
+      "learning_rate": 0.0001663267070008643,
+      "loss": 0.6142,
+      "step": 979
+    },
+    {
+      "epoch": 5.077419354838709,
+      "grad_norm": 0.9902614951133728,
+      "learning_rate": 0.00016629213483146068,
+      "loss": 0.5909,
+      "step": 980
+    },
+    {
+      "epoch": 5.082580645161291,
+      "grad_norm": 1.182059407234192,
+      "learning_rate": 0.00016625756266205706,
+      "loss": 0.5566,
+      "step": 981
+    },
+    {
+      "epoch": 5.087741935483871,
+      "grad_norm": 0.8433648943901062,
+      "learning_rate": 0.00016622299049265341,
+      "loss": 0.563,
+      "step": 982
+    },
+    {
+      "epoch": 5.0929032258064515,
+      "grad_norm": 1.0695401430130005,
+      "learning_rate": 0.0001661884183232498,
+      "loss": 0.5864,
+      "step": 983
+    },
+    {
+      "epoch": 5.098064516129032,
+      "grad_norm": 1.0976457595825195,
+      "learning_rate": 0.00016615384615384617,
+      "loss": 0.5866,
+      "step": 984
+    },
+    {
+      "epoch": 5.103225806451613,
+      "grad_norm": 1.230617880821228,
+      "learning_rate": 0.00016611927398444253,
+      "loss": 0.5865,
+      "step": 985
+    },
+    {
+      "epoch": 5.108387096774194,
+      "grad_norm": 1.2860138416290283,
+      "learning_rate": 0.0001660847018150389,
+      "loss": 0.6081,
+      "step": 986
+    },
+    {
+      "epoch": 5.113548387096774,
+      "grad_norm": 1.040961742401123,
+      "learning_rate": 0.00016605012964563526,
+      "loss": 0.609,
+      "step": 987
+    },
+    {
+      "epoch": 5.118709677419355,
+      "grad_norm": 1.1050902605056763,
+      "learning_rate": 0.00016601555747623164,
+      "loss": 0.5969,
+      "step": 988
+    },
+    {
+      "epoch": 5.123870967741936,
+      "grad_norm": 0.993930995464325,
+      "learning_rate": 0.00016598098530682802,
+      "loss": 0.6348,
+      "step": 989
+    },
+    {
+      "epoch": 5.129032258064516,
+      "grad_norm": 0.9431015253067017,
+      "learning_rate": 0.00016594641313742438,
+      "loss": 0.5447,
+      "step": 990
+    },
+    {
+      "epoch": 5.134193548387096,
+      "grad_norm": 1.0567395687103271,
+      "learning_rate": 0.00016591184096802076,
+      "loss": 0.6176,
+      "step": 991
+    },
+    {
+      "epoch": 5.139354838709678,
+      "grad_norm": 1.1539596319198608,
+      "learning_rate": 0.0001658772687986171,
+      "loss": 0.6032,
+      "step": 992
+    },
+    {
+      "epoch": 5.144516129032258,
+      "grad_norm": 1.164534330368042,
+      "learning_rate": 0.0001658426966292135,
+      "loss": 0.5697,
+      "step": 993
+    },
+    {
+      "epoch": 5.1496774193548385,
+      "grad_norm": 1.1888781785964966,
+      "learning_rate": 0.00016580812445980987,
+      "loss": 0.6179,
+      "step": 994
+    },
+    {
+      "epoch": 5.15483870967742,
+      "grad_norm": 1.1361608505249023,
+      "learning_rate": 0.00016577355229040623,
+      "loss": 0.6296,
+      "step": 995
+    },
+    {
+      "epoch": 5.16,
+      "grad_norm": 0.9821083545684814,
+      "learning_rate": 0.0001657389801210026,
+      "loss": 0.5831,
+      "step": 996
+    },
+    {
+      "epoch": 5.1651612903225805,
+      "grad_norm": 1.1042648553848267,
+      "learning_rate": 0.00016570440795159896,
+      "loss": 0.6484,
+      "step": 997
+    },
+    {
+      "epoch": 5.170322580645161,
+      "grad_norm": 1.0881468057632446,
+      "learning_rate": 0.00016566983578219534,
+      "loss": 0.6139,
+      "step": 998
+    },
+    {
+      "epoch": 5.175483870967742,
+      "grad_norm": 1.0723835229873657,
+      "learning_rate": 0.00016563526361279172,
+      "loss": 0.62,
+      "step": 999
+    },
+    {
+      "epoch": 5.180645161290323,
+      "grad_norm": 1.0688388347625732,
+      "learning_rate": 0.00016560069144338807,
+      "loss": 0.5693,
+      "step": 1000
+    },
+    {
+      "epoch": 5.185806451612903,
+      "grad_norm": 1.0733652114868164,
+      "learning_rate": 0.00016556611927398445,
+      "loss": 0.5864,
+      "step": 1001
+    },
+    {
+      "epoch": 5.190967741935484,
+      "grad_norm": 1.1086695194244385,
+      "learning_rate": 0.00016553154710458083,
+      "loss": 0.6369,
+      "step": 1002
+    },
+    {
+      "epoch": 5.196129032258065,
+      "grad_norm": 1.0873360633850098,
+      "learning_rate": 0.0001654969749351772,
+      "loss": 0.5058,
+      "step": 1003
+    },
+    {
+      "epoch": 5.201290322580645,
+      "grad_norm": 1.0094376802444458,
+      "learning_rate": 0.00016546240276577357,
+      "loss": 0.4815,
+      "step": 1004
+    },
+    {
+      "epoch": 5.2064516129032254,
+      "grad_norm": 1.0067589282989502,
+      "learning_rate": 0.00016542783059636992,
+      "loss": 0.572,
+      "step": 1005
+    },
+    {
+      "epoch": 5.211612903225807,
+      "grad_norm": 1.2047669887542725,
+      "learning_rate": 0.0001653932584269663,
+      "loss": 0.667,
+      "step": 1006
+    },
+    {
+      "epoch": 5.216774193548387,
+      "grad_norm": 1.126724362373352,
+      "learning_rate": 0.00016535868625756268,
+      "loss": 0.6023,
+      "step": 1007
+    },
+    {
+      "epoch": 5.2219354838709675,
+      "grad_norm": 1.1412140130996704,
+      "learning_rate": 0.00016532411408815904,
+      "loss": 0.6512,
+      "step": 1008
+    },
+    {
+      "epoch": 5.227096774193549,
+      "grad_norm": 1.0202922821044922,
+      "learning_rate": 0.00016528954191875542,
+      "loss": 0.6175,
+      "step": 1009
+    },
+    {
+      "epoch": 5.232258064516129,
+      "grad_norm": 1.084920048713684,
+      "learning_rate": 0.00016525496974935177,
+      "loss": 0.642,
+      "step": 1010
+    },
+    {
+      "epoch": 5.23741935483871,
+      "grad_norm": 1.0586888790130615,
+      "learning_rate": 0.00016522039757994815,
+      "loss": 0.6622,
+      "step": 1011
+    },
+    {
+      "epoch": 5.24258064516129,
+      "grad_norm": 1.0660359859466553,
+      "learning_rate": 0.00016518582541054453,
+      "loss": 0.6367,
+      "step": 1012
+    },
+    {
+      "epoch": 5.247741935483871,
+      "grad_norm": 1.0708181858062744,
+      "learning_rate": 0.00016515125324114089,
+      "loss": 0.6731,
+      "step": 1013
+    },
+    {
+      "epoch": 5.252903225806452,
+      "grad_norm": 1.1763105392456055,
+      "learning_rate": 0.00016511668107173727,
+      "loss": 0.6071,
+      "step": 1014
+    },
+    {
+      "epoch": 5.258064516129032,
+      "grad_norm": 1.1695592403411865,
+      "learning_rate": 0.00016508210890233362,
+      "loss": 0.6458,
+      "step": 1015
+    },
+    {
+      "epoch": 5.263225806451613,
+      "grad_norm": 1.0554745197296143,
+      "learning_rate": 0.00016504753673293,
+      "loss": 0.5925,
+      "step": 1016
+    },
+    {
+      "epoch": 5.268387096774194,
+      "grad_norm": 1.1234837770462036,
+      "learning_rate": 0.00016501296456352638,
+      "loss": 0.6843,
+      "step": 1017
+    },
+    {
+      "epoch": 5.273548387096774,
+      "grad_norm": 1.105829119682312,
+      "learning_rate": 0.00016497839239412273,
+      "loss": 0.6502,
+      "step": 1018
+    },
+    {
+      "epoch": 5.2787096774193545,
+      "grad_norm": 1.0738614797592163,
+      "learning_rate": 0.00016494382022471911,
+      "loss": 0.6454,
+      "step": 1019
+    },
+    {
+      "epoch": 5.283870967741936,
+      "grad_norm": 1.0996984243392944,
+      "learning_rate": 0.00016490924805531547,
+      "loss": 0.6343,
+      "step": 1020
+    },
+    {
+      "epoch": 5.289032258064516,
+      "grad_norm": 1.1064165830612183,
+      "learning_rate": 0.00016487467588591185,
+      "loss": 0.6521,
+      "step": 1021
+    },
+    {
+      "epoch": 5.2941935483870965,
+      "grad_norm": 1.118908166885376,
+      "learning_rate": 0.00016484010371650823,
+      "loss": 0.6503,
+      "step": 1022
+    },
+    {
+      "epoch": 5.299354838709678,
+      "grad_norm": 1.119377613067627,
+      "learning_rate": 0.00016480553154710458,
+      "loss": 0.6791,
+      "step": 1023
+    },
+    {
+      "epoch": 5.304516129032258,
+      "grad_norm": 1.150528907775879,
+      "learning_rate": 0.00016477095937770096,
+      "loss": 0.6819,
+      "step": 1024
+    },
+    {
+      "epoch": 5.309677419354839,
+      "grad_norm": 1.0134869813919067,
+      "learning_rate": 0.00016473638720829734,
+      "loss": 0.653,
+      "step": 1025
+    },
+    {
+      "epoch": 5.314838709677419,
+      "grad_norm": 1.1313871145248413,
+      "learning_rate": 0.0001647018150388937,
+      "loss": 0.6981,
+      "step": 1026
+    },
+    {
+      "epoch": 5.32,
+      "grad_norm": 1.1238486766815186,
+      "learning_rate": 0.00016466724286949008,
+      "loss": 0.6913,
+      "step": 1027
+    },
+    {
+      "epoch": 5.325161290322581,
+      "grad_norm": 1.17030930519104,
+      "learning_rate": 0.00016463267070008643,
+      "loss": 0.6852,
+      "step": 1028
+    },
+    {
+      "epoch": 5.330322580645161,
+      "grad_norm": 1.2329025268554688,
+      "learning_rate": 0.0001645980985306828,
+      "loss": 0.7575,
+      "step": 1029
+    },
+    {
+      "epoch": 5.335483870967742,
+      "grad_norm": 1.0909053087234497,
+      "learning_rate": 0.0001645635263612792,
+      "loss": 0.6843,
+      "step": 1030
+    },
+    {
+      "epoch": 5.340645161290323,
+      "grad_norm": 1.0104080438613892,
+      "learning_rate": 0.00016452895419187555,
+      "loss": 0.6523,
+      "step": 1031
+    },
+    {
+      "epoch": 5.345806451612903,
+      "grad_norm": 1.0352752208709717,
+      "learning_rate": 0.00016449438202247193,
+      "loss": 0.6087,
+      "step": 1032
+    },
+    {
+      "epoch": 5.3509677419354835,
+      "grad_norm": 1.0280145406723022,
+      "learning_rate": 0.00016445980985306828,
+      "loss": 0.6941,
+      "step": 1033
+    },
+    {
+      "epoch": 5.356129032258065,
+      "grad_norm": 1.119673252105713,
+      "learning_rate": 0.00016442523768366466,
+      "loss": 0.687,
+      "step": 1034
+    },
+    {
+      "epoch": 5.361290322580645,
+      "grad_norm": 1.0995351076126099,
+      "learning_rate": 0.00016439066551426104,
+      "loss": 0.6944,
+      "step": 1035
+    },
+    {
+      "epoch": 5.366451612903226,
+      "grad_norm": 1.2425403594970703,
+      "learning_rate": 0.0001643560933448574,
+      "loss": 0.6825,
+      "step": 1036
+    },
+    {
+      "epoch": 5.371612903225807,
+      "grad_norm": 1.1850897073745728,
+      "learning_rate": 0.00016432152117545377,
+      "loss": 0.6579,
+      "step": 1037
+    },
+    {
+      "epoch": 5.376774193548387,
+      "grad_norm": 1.1279488801956177,
+      "learning_rate": 0.00016428694900605013,
+      "loss": 0.7124,
+      "step": 1038
+    },
+    {
+      "epoch": 5.381935483870968,
+      "grad_norm": 1.1537352800369263,
+      "learning_rate": 0.0001642523768366465,
+      "loss": 0.721,
+      "step": 1039
+    },
+    {
+      "epoch": 5.387096774193548,
+      "grad_norm": 1.1323246955871582,
+      "learning_rate": 0.0001642178046672429,
+      "loss": 0.6402,
+      "step": 1040
+    },
+    {
+      "epoch": 5.392258064516129,
+      "grad_norm": 1.098808765411377,
+      "learning_rate": 0.00016418323249783924,
+      "loss": 0.7014,
+      "step": 1041
+    },
+    {
+      "epoch": 5.39741935483871,
+      "grad_norm": 1.0692648887634277,
+      "learning_rate": 0.00016414866032843562,
+      "loss": 0.623,
+      "step": 1042
+    },
+    {
+      "epoch": 5.40258064516129,
+      "grad_norm": 0.9733750224113464,
+      "learning_rate": 0.000164114088159032,
+      "loss": 0.6211,
+      "step": 1043
+    },
+    {
+      "epoch": 5.407741935483871,
+      "grad_norm": 1.0590685606002808,
+      "learning_rate": 0.00016407951598962836,
+      "loss": 0.6934,
+      "step": 1044
+    },
+    {
+      "epoch": 5.412903225806452,
+      "grad_norm": 1.1508102416992188,
+      "learning_rate": 0.00016404494382022474,
+      "loss": 0.6818,
+      "step": 1045
+    },
+    {
+      "epoch": 5.418064516129032,
+      "grad_norm": 1.2017369270324707,
+      "learning_rate": 0.0001640103716508211,
+      "loss": 0.7115,
+      "step": 1046
+    },
+    {
+      "epoch": 5.423225806451613,
+      "grad_norm": 1.245505928993225,
+      "learning_rate": 0.00016397579948141747,
+      "loss": 0.745,
+      "step": 1047
+    },
+    {
+      "epoch": 5.428387096774194,
+      "grad_norm": 1.1033475399017334,
+      "learning_rate": 0.00016394122731201385,
+      "loss": 0.6836,
+      "step": 1048
+    },
+    {
+      "epoch": 5.433548387096774,
+      "grad_norm": 1.1161737442016602,
+      "learning_rate": 0.0001639066551426102,
+      "loss": 0.6606,
+      "step": 1049
+    },
+    {
+      "epoch": 5.438709677419355,
+      "grad_norm": 0.9725407958030701,
+      "learning_rate": 0.00016387208297320659,
+      "loss": 0.7215,
+      "step": 1050
+    },
+    {
+      "epoch": 5.443870967741935,
+      "grad_norm": 0.9560017585754395,
+      "learning_rate": 0.00016383751080380294,
+      "loss": 0.703,
+      "step": 1051
+    },
+    {
+      "epoch": 5.449032258064516,
+      "grad_norm": 1.063751459121704,
+      "learning_rate": 0.0001638029386343993,
+      "loss": 0.733,
+      "step": 1052
+    },
+    {
+      "epoch": 5.454193548387097,
+      "grad_norm": 1.1284129619598389,
+      "learning_rate": 0.0001637683664649957,
+      "loss": 0.687,
+      "step": 1053
+    },
+    {
+      "epoch": 5.459354838709677,
+      "grad_norm": 1.1844428777694702,
+      "learning_rate": 0.00016373379429559205,
+      "loss": 0.7247,
+      "step": 1054
+    },
+    {
+      "epoch": 5.464516129032258,
+      "grad_norm": 1.1811026334762573,
+      "learning_rate": 0.00016369922212618843,
+      "loss": 0.7291,
+      "step": 1055
+    },
+    {
+      "epoch": 5.469677419354839,
+      "grad_norm": 1.1146628856658936,
+      "learning_rate": 0.0001636646499567848,
+      "loss": 0.7205,
+      "step": 1056
+    },
+    {
+      "epoch": 5.474838709677419,
+      "grad_norm": 1.3297598361968994,
+      "learning_rate": 0.00016363007778738117,
+      "loss": 0.7317,
+      "step": 1057
+    },
+    {
+      "epoch": 5.48,
+      "grad_norm": 1.117520809173584,
+      "learning_rate": 0.00016359550561797755,
+      "loss": 0.81,
+      "step": 1058
+    },
+    {
+      "epoch": 5.485161290322581,
+      "grad_norm": 1.1416107416152954,
+      "learning_rate": 0.0001635609334485739,
+      "loss": 0.7547,
+      "step": 1059
+    },
+    {
+      "epoch": 5.490322580645161,
+      "grad_norm": 1.0028117895126343,
+      "learning_rate": 0.00016352636127917028,
+      "loss": 0.6181,
+      "step": 1060
+    },
+    {
+      "epoch": 5.495483870967742,
+      "grad_norm": 1.1248133182525635,
+      "learning_rate": 0.00016349178910976664,
+      "loss": 0.7285,
+      "step": 1061
+    },
+    {
+      "epoch": 5.500645161290323,
+      "grad_norm": 1.1524790525436401,
+      "learning_rate": 0.00016345721694036302,
+      "loss": 0.7147,
+      "step": 1062
+    },
+    {
+      "epoch": 5.505806451612903,
+      "grad_norm": 1.193189024925232,
+      "learning_rate": 0.0001634226447709594,
+      "loss": 0.7374,
+      "step": 1063
+    },
+    {
+      "epoch": 5.510967741935484,
+      "grad_norm": 1.1519007682800293,
+      "learning_rate": 0.00016338807260155575,
+      "loss": 0.8039,
+      "step": 1064
+    },
+    {
+      "epoch": 5.516129032258064,
+      "grad_norm": 1.0880852937698364,
+      "learning_rate": 0.00016335350043215213,
+      "loss": 0.7117,
+      "step": 1065
+    },
+    {
+      "epoch": 5.521290322580645,
+      "grad_norm": 1.0384647846221924,
+      "learning_rate": 0.0001633189282627485,
+      "loss": 0.7186,
+      "step": 1066
+    },
+    {
+      "epoch": 5.526451612903226,
+      "grad_norm": 0.9708914756774902,
+      "learning_rate": 0.00016328435609334486,
+      "loss": 0.688,
+      "step": 1067
+    },
+    {
+      "epoch": 5.531612903225806,
+      "grad_norm": 1.0140278339385986,
+      "learning_rate": 0.00016324978392394125,
+      "loss": 0.7159,
+      "step": 1068
+    },
+    {
+      "epoch": 5.536774193548387,
+      "grad_norm": 1.1154056787490845,
+      "learning_rate": 0.0001632152117545376,
+      "loss": 0.7368,
+      "step": 1069
+    },
+    {
+      "epoch": 5.541935483870968,
+      "grad_norm": 1.1336860656738281,
+      "learning_rate": 0.00016318063958513395,
+      "loss": 0.7403,
+      "step": 1070
+    },
+    {
+      "epoch": 5.547096774193548,
+      "grad_norm": 1.3124700784683228,
+      "learning_rate": 0.00016314606741573036,
+      "loss": 0.7456,
+      "step": 1071
+    },
+    {
+      "epoch": 5.5522580645161295,
+      "grad_norm": 1.2411500215530396,
+      "learning_rate": 0.0001631114952463267,
+      "loss": 0.7626,
+      "step": 1072
+    },
+    {
+      "epoch": 5.55741935483871,
+      "grad_norm": 1.1269874572753906,
+      "learning_rate": 0.0001630769230769231,
+      "loss": 0.654,
+      "step": 1073
+    },
+    {
+      "epoch": 5.56258064516129,
+      "grad_norm": 1.1101837158203125,
+      "learning_rate": 0.00016304235090751945,
+      "loss": 0.6944,
+      "step": 1074
+    },
+    {
+      "epoch": 5.567741935483871,
+      "grad_norm": 1.1272631883621216,
+      "learning_rate": 0.00016300777873811583,
+      "loss": 0.7332,
+      "step": 1075
+    },
+    {
+      "epoch": 5.572903225806452,
+      "grad_norm": 1.1115466356277466,
+      "learning_rate": 0.0001629732065687122,
+      "loss": 0.7699,
+      "step": 1076
+    },
+    {
+      "epoch": 5.578064516129032,
+      "grad_norm": 1.0561658143997192,
+      "learning_rate": 0.00016293863439930856,
+      "loss": 0.7115,
+      "step": 1077
+    },
+    {
+      "epoch": 5.583225806451613,
+      "grad_norm": 1.1047935485839844,
+      "learning_rate": 0.00016290406222990494,
+      "loss": 0.7695,
+      "step": 1078
+    },
+    {
+      "epoch": 5.588387096774193,
+      "grad_norm": 1.0621674060821533,
+      "learning_rate": 0.0001628694900605013,
+      "loss": 0.7723,
+      "step": 1079
+    },
+    {
+      "epoch": 5.593548387096774,
+      "grad_norm": 1.0361956357955933,
+      "learning_rate": 0.00016283491789109768,
+      "loss": 0.6491,
+      "step": 1080
+    },
+    {
+      "epoch": 5.598709677419355,
+      "grad_norm": 1.1300548315048218,
+      "learning_rate": 0.00016280034572169406,
+      "loss": 0.7221,
+      "step": 1081
+    },
+    {
+      "epoch": 5.603870967741935,
+      "grad_norm": 1.1098562479019165,
+      "learning_rate": 0.0001627657735522904,
+      "loss": 0.7679,
+      "step": 1082
+    },
+    {
+      "epoch": 5.609032258064516,
+      "grad_norm": 1.1171636581420898,
+      "learning_rate": 0.0001627312013828868,
+      "loss": 0.7609,
+      "step": 1083
+    },
+    {
+      "epoch": 5.614193548387097,
+      "grad_norm": 1.008507251739502,
+      "learning_rate": 0.00016269662921348317,
+      "loss": 0.6712,
+      "step": 1084
+    },
+    {
+      "epoch": 5.619354838709677,
+      "grad_norm": 1.0691373348236084,
+      "learning_rate": 0.00016266205704407952,
+      "loss": 0.7313,
+      "step": 1085
+    },
+    {
+      "epoch": 5.6245161290322585,
+      "grad_norm": 1.1126989126205444,
+      "learning_rate": 0.0001626274848746759,
+      "loss": 0.6906,
+      "step": 1086
+    },
+    {
+      "epoch": 5.629677419354839,
+      "grad_norm": 1.0986865758895874,
+      "learning_rate": 0.00016259291270527226,
+      "loss": 0.7541,
+      "step": 1087
+    },
+    {
+      "epoch": 5.634838709677419,
+      "grad_norm": 1.1169289350509644,
+      "learning_rate": 0.0001625583405358686,
+      "loss": 0.6969,
+      "step": 1088
+    },
+    {
+      "epoch": 5.64,
+      "grad_norm": 1.115019679069519,
+      "learning_rate": 0.00016252376836646502,
+      "loss": 0.7417,
+      "step": 1089
+    },
+    {
+      "epoch": 5.645161290322581,
+      "grad_norm": 0.9654311537742615,
+      "learning_rate": 0.00016248919619706137,
+      "loss": 0.6584,
+      "step": 1090
+    },
+    {
+      "epoch": 5.650322580645161,
+      "grad_norm": 1.089620590209961,
+      "learning_rate": 0.00016245462402765775,
+      "loss": 0.7944,
+      "step": 1091
+    },
+    {
+      "epoch": 5.655483870967742,
+      "grad_norm": 1.0662012100219727,
+      "learning_rate": 0.0001624200518582541,
+      "loss": 0.7704,
+      "step": 1092
+    },
+    {
+      "epoch": 5.660645161290322,
+      "grad_norm": 1.125723123550415,
+      "learning_rate": 0.00016238547968885046,
+      "loss": 0.7793,
+      "step": 1093
+    },
+    {
+      "epoch": 5.665806451612903,
+      "grad_norm": 1.0247701406478882,
+      "learning_rate": 0.00016235090751944687,
+      "loss": 0.688,
+      "step": 1094
+    },
+    {
+      "epoch": 5.670967741935484,
+      "grad_norm": 1.0904461145401,
+      "learning_rate": 0.00016231633535004322,
+      "loss": 0.8022,
+      "step": 1095
+    },
+    {
+      "epoch": 5.676129032258064,
+      "grad_norm": 1.0644805431365967,
+      "learning_rate": 0.0001622817631806396,
+      "loss": 0.7715,
+      "step": 1096
+    },
+    {
+      "epoch": 5.6812903225806455,
+      "grad_norm": 1.0406841039657593,
+      "learning_rate": 0.00016224719101123596,
+      "loss": 0.7379,
+      "step": 1097
+    },
+    {
+      "epoch": 5.686451612903226,
+      "grad_norm": 1.106176733970642,
+      "learning_rate": 0.00016221261884183234,
+      "loss": 0.8054,
+      "step": 1098
+    },
+    {
+      "epoch": 5.691612903225806,
+      "grad_norm": 1.062544822692871,
+      "learning_rate": 0.00016217804667242872,
+      "loss": 0.7539,
+      "step": 1099
+    },
+    {
+      "epoch": 5.6967741935483875,
+      "grad_norm": 1.0066852569580078,
+      "learning_rate": 0.00016214347450302507,
+      "loss": 0.7212,
+      "step": 1100
+    },
+    {
+      "epoch": 5.701935483870968,
+      "grad_norm": 1.1381007432937622,
+      "learning_rate": 0.00016210890233362145,
+      "loss": 0.7971,
+      "step": 1101
+    },
+    {
+      "epoch": 5.707096774193548,
+      "grad_norm": 1.1032809019088745,
+      "learning_rate": 0.0001620743301642178,
+      "loss": 0.7315,
+      "step": 1102
+    },
+    {
+      "epoch": 5.712258064516129,
+      "grad_norm": 1.061509132385254,
+      "learning_rate": 0.00016203975799481418,
+      "loss": 0.7244,
+      "step": 1103
+    },
+    {
+      "epoch": 5.71741935483871,
+      "grad_norm": 1.114195704460144,
+      "learning_rate": 0.00016200518582541057,
+      "loss": 0.7812,
+      "step": 1104
+    },
+    {
+      "epoch": 5.72258064516129,
+      "grad_norm": 1.1118159294128418,
+      "learning_rate": 0.00016197061365600692,
+      "loss": 0.789,
+      "step": 1105
+    },
+    {
+      "epoch": 5.727741935483871,
+      "grad_norm": 1.072842001914978,
+      "learning_rate": 0.00016193604148660327,
+      "loss": 0.7605,
+      "step": 1106
+    },
+    {
+      "epoch": 5.732903225806451,
+      "grad_norm": 1.0300045013427734,
+      "learning_rate": 0.00016190146931719968,
+      "loss": 0.7422,
+      "step": 1107
+    },
+    {
+      "epoch": 5.7380645161290325,
+      "grad_norm": 1.0478154420852661,
+      "learning_rate": 0.00016186689714779603,
+      "loss": 0.7848,
+      "step": 1108
+    },
+    {
+      "epoch": 5.743225806451613,
+      "grad_norm": 1.116895079612732,
+      "learning_rate": 0.00016183232497839241,
+      "loss": 0.7222,
+      "step": 1109
+    },
+    {
+      "epoch": 5.748387096774193,
+      "grad_norm": 1.0723727941513062,
+      "learning_rate": 0.00016179775280898877,
+      "loss": 0.6935,
+      "step": 1110
+    },
+    {
+      "epoch": 5.7535483870967745,
+      "grad_norm": 1.2100520133972168,
+      "learning_rate": 0.00016176318063958512,
+      "loss": 0.7686,
+      "step": 1111
+    },
+    {
+      "epoch": 5.758709677419355,
+      "grad_norm": 1.1042420864105225,
+      "learning_rate": 0.00016172860847018153,
+      "loss": 0.762,
+      "step": 1112
+    },
+    {
+      "epoch": 5.763870967741935,
+      "grad_norm": 1.117038369178772,
+      "learning_rate": 0.00016169403630077788,
+      "loss": 0.6972,
+      "step": 1113
+    },
+    {
+      "epoch": 5.769032258064517,
+      "grad_norm": 1.0118979215621948,
+      "learning_rate": 0.00016165946413137426,
+      "loss": 0.6983,
+      "step": 1114
+    },
+    {
+      "epoch": 5.774193548387097,
+      "grad_norm": 1.5192986726760864,
+      "learning_rate": 0.00016162489196197062,
+      "loss": 0.7625,
+      "step": 1115
+    },
+    {
+      "epoch": 5.779354838709677,
+      "grad_norm": 1.0850465297698975,
+      "learning_rate": 0.000161590319792567,
+      "loss": 0.7423,
+      "step": 1116
+    },
+    {
+      "epoch": 5.784516129032258,
+      "grad_norm": 1.0667920112609863,
+      "learning_rate": 0.00016155574762316338,
+      "loss": 0.7622,
+      "step": 1117
+    },
+    {
+      "epoch": 5.789677419354839,
+      "grad_norm": 1.0413289070129395,
+      "learning_rate": 0.00016152117545375973,
+      "loss": 0.7372,
+      "step": 1118
+    },
+    {
+      "epoch": 5.794838709677419,
+      "grad_norm": 1.0257551670074463,
+      "learning_rate": 0.00016148660328435608,
+      "loss": 0.7129,
+      "step": 1119
+    },
+    {
+      "epoch": 5.8,
+      "grad_norm": 1.0711652040481567,
+      "learning_rate": 0.00016145203111495246,
+      "loss": 0.8223,
+      "step": 1120
+    },
+    {
+      "epoch": 5.80516129032258,
+      "grad_norm": 1.04441237449646,
+      "learning_rate": 0.00016141745894554884,
+      "loss": 0.7417,
+      "step": 1121
+    },
+    {
+      "epoch": 5.8103225806451615,
+      "grad_norm": 1.030312418937683,
+      "learning_rate": 0.00016138288677614523,
+      "loss": 0.7687,
+      "step": 1122
+    },
+    {
+      "epoch": 5.815483870967742,
+      "grad_norm": 1.0566182136535645,
+      "learning_rate": 0.00016134831460674158,
+      "loss": 0.7841,
+      "step": 1123
+    },
+    {
+      "epoch": 5.820645161290322,
+      "grad_norm": 1.0681883096694946,
+      "learning_rate": 0.00016131374243733793,
+      "loss": 0.7083,
+      "step": 1124
+    },
+    {
+      "epoch": 5.825806451612904,
+      "grad_norm": 1.0629135370254517,
+      "learning_rate": 0.0001612791702679343,
+      "loss": 0.7768,
+      "step": 1125
+    },
+    {
+      "epoch": 5.830967741935484,
+      "grad_norm": 1.1138688325881958,
+      "learning_rate": 0.0001612445980985307,
+      "loss": 0.7952,
+      "step": 1126
+    },
+    {
+      "epoch": 5.836129032258064,
+      "grad_norm": 1.055606484413147,
+      "learning_rate": 0.00016121002592912707,
+      "loss": 0.7925,
+      "step": 1127
+    },
+    {
+      "epoch": 5.841290322580646,
+      "grad_norm": 1.0343049764633179,
+      "learning_rate": 0.00016117545375972343,
+      "loss": 0.7708,
+      "step": 1128
+    },
+    {
+      "epoch": 5.846451612903226,
+      "grad_norm": 0.9960595369338989,
+      "learning_rate": 0.00016114088159031978,
+      "loss": 0.7546,
+      "step": 1129
+    },
+    {
+      "epoch": 5.851612903225806,
+      "grad_norm": 0.9953485727310181,
+      "learning_rate": 0.0001611063094209162,
+      "loss": 0.7504,
+      "step": 1130
+    },
+    {
+      "epoch": 5.856774193548387,
+      "grad_norm": 1.0808544158935547,
+      "learning_rate": 0.00016107173725151254,
+      "loss": 0.7793,
+      "step": 1131
+    },
+    {
+      "epoch": 5.861935483870968,
+      "grad_norm": 1.1002646684646606,
+      "learning_rate": 0.00016103716508210892,
+      "loss": 0.7505,
+      "step": 1132
+    },
+    {
+      "epoch": 5.8670967741935485,
+      "grad_norm": 1.1822736263275146,
+      "learning_rate": 0.00016100259291270528,
+      "loss": 0.7208,
+      "step": 1133
+    },
+    {
+      "epoch": 5.872258064516129,
+      "grad_norm": 1.0634794235229492,
+      "learning_rate": 0.00016096802074330163,
+      "loss": 0.7105,
+      "step": 1134
+    },
+    {
+      "epoch": 5.877419354838709,
+      "grad_norm": 1.083084225654602,
+      "learning_rate": 0.00016093344857389804,
+      "loss": 0.7822,
+      "step": 1135
+    },
+    {
+      "epoch": 5.8825806451612905,
+      "grad_norm": 1.0536597967147827,
+      "learning_rate": 0.0001608988764044944,
+      "loss": 0.8485,
+      "step": 1136
+    },
+    {
+      "epoch": 5.887741935483871,
+      "grad_norm": 1.038663625717163,
+      "learning_rate": 0.00016086430423509074,
+      "loss": 0.8155,
+      "step": 1137
+    },
+    {
+      "epoch": 5.892903225806451,
+      "grad_norm": 0.9630552530288696,
+      "learning_rate": 0.00016082973206568712,
+      "loss": 0.6217,
+      "step": 1138
+    },
+    {
+      "epoch": 5.898064516129033,
+      "grad_norm": 1.0534899234771729,
+      "learning_rate": 0.0001607951598962835,
+      "loss": 0.8034,
+      "step": 1139
+    },
+    {
+      "epoch": 5.903225806451613,
+      "grad_norm": 1.047546148300171,
+      "learning_rate": 0.00016076058772687989,
+      "loss": 0.8154,
+      "step": 1140
+    },
+    {
+      "epoch": 5.908387096774193,
+      "grad_norm": 1.0899145603179932,
+      "learning_rate": 0.00016072601555747624,
+      "loss": 0.7634,
+      "step": 1141
+    },
+    {
+      "epoch": 5.913548387096775,
+      "grad_norm": 1.1362152099609375,
+      "learning_rate": 0.0001606914433880726,
+      "loss": 0.8331,
+      "step": 1142
+    },
+    {
+      "epoch": 5.918709677419355,
+      "grad_norm": 1.10222327709198,
+      "learning_rate": 0.00016065687121866897,
+      "loss": 0.7926,
+      "step": 1143
+    },
+    {
+      "epoch": 5.9238709677419354,
+      "grad_norm": 1.0164943933486938,
+      "learning_rate": 0.00016062229904926535,
+      "loss": 0.7883,
+      "step": 1144
+    },
+    {
+      "epoch": 5.929032258064516,
+      "grad_norm": 0.981583297252655,
+      "learning_rate": 0.00016058772687986173,
+      "loss": 0.7244,
+      "step": 1145
+    },
+    {
+      "epoch": 5.934193548387097,
+      "grad_norm": 1.0531173944473267,
+      "learning_rate": 0.0001605531547104581,
+      "loss": 0.7872,
+      "step": 1146
+    },
+    {
+      "epoch": 5.9393548387096775,
+      "grad_norm": 1.2195733785629272,
+      "learning_rate": 0.00016051858254105444,
+      "loss": 0.7196,
+      "step": 1147
+    },
+    {
+      "epoch": 5.944516129032258,
+      "grad_norm": 1.1409739255905151,
+      "learning_rate": 0.00016048401037165085,
+      "loss": 0.7972,
+      "step": 1148
+    },
+    {
+      "epoch": 5.949677419354838,
+      "grad_norm": 1.1853970289230347,
+      "learning_rate": 0.0001604494382022472,
+      "loss": 0.8263,
+      "step": 1149
+    },
+    {
+      "epoch": 5.95483870967742,
+      "grad_norm": 1.1241440773010254,
+      "learning_rate": 0.00016041486603284358,
+      "loss": 0.8145,
+      "step": 1150
+    },
+    {
+      "epoch": 5.96,
+      "grad_norm": 1.0923287868499756,
+      "learning_rate": 0.00016038029386343994,
+      "loss": 0.7482,
+      "step": 1151
+    },
+    {
+      "epoch": 5.96516129032258,
+      "grad_norm": 1.0010133981704712,
+      "learning_rate": 0.0001603457216940363,
+      "loss": 0.7282,
+      "step": 1152
+    },
+    {
+      "epoch": 5.970322580645162,
+      "grad_norm": 1.0343108177185059,
+      "learning_rate": 0.0001603111495246327,
+      "loss": 0.8204,
+      "step": 1153
+    },
+    {
+      "epoch": 5.975483870967742,
+      "grad_norm": 0.9813974499702454,
+      "learning_rate": 0.00016027657735522905,
+      "loss": 0.6954,
+      "step": 1154
+    },
+    {
+      "epoch": 5.980645161290322,
+      "grad_norm": 1.1140379905700684,
+      "learning_rate": 0.0001602420051858254,
+      "loss": 0.8183,
+      "step": 1155
+    },
+    {
+      "epoch": 5.985806451612904,
+      "grad_norm": 1.087619662284851,
+      "learning_rate": 0.00016020743301642178,
+      "loss": 0.7155,
+      "step": 1156
+    },
+    {
+      "epoch": 5.990967741935484,
+      "grad_norm": 1.0367388725280762,
+      "learning_rate": 0.00016017286084701816,
+      "loss": 0.7739,
+      "step": 1157
+    },
+    {
+      "epoch": 5.9961290322580645,
+      "grad_norm": 0.9448538422584534,
+      "learning_rate": 0.00016013828867761455,
+      "loss": 0.6892,
+      "step": 1158
+    },
+    {
+      "epoch": 6.002580645161291,
+      "grad_norm": 1.606164574623108,
+      "learning_rate": 0.0001601037165082109,
+      "loss": 1.0458,
+      "step": 1159
+    },
+    {
+      "epoch": 6.007741935483871,
+      "grad_norm": 0.9125086069107056,
+      "learning_rate": 0.00016006914433880725,
+      "loss": 0.5868,
+      "step": 1160
+    },
+    {
+      "epoch": 6.007741935483871,
+      "eval_loss": 1.823500633239746,
+      "eval_runtime": 21.3282,
+      "eval_samples_per_second": 3.704,
+      "eval_steps_per_second": 0.469,
+      "step": 1160
+    },
+    {
+      "epoch": 6.012903225806451,
+      "grad_norm": 0.9444449543952942,
+      "learning_rate": 0.00016003457216940363,
+      "loss": 0.5566,
+      "step": 1161
+    },
+    {
+      "epoch": 6.018064516129032,
+      "grad_norm": 1.081428050994873,
+      "learning_rate": 0.00016,
+      "loss": 0.5202,
+      "step": 1162
+    },
+    {
+      "epoch": 6.023225806451613,
+      "grad_norm": 1.088323712348938,
+      "learning_rate": 0.0001599654278305964,
+      "loss": 0.5583,
+      "step": 1163
+    },
+    {
+      "epoch": 6.0283870967741935,
+      "grad_norm": 1.236128807067871,
+      "learning_rate": 0.00015993085566119275,
+      "loss": 0.5756,
+      "step": 1164
+    },
+    {
+      "epoch": 6.033548387096774,
+      "grad_norm": 1.2892690896987915,
+      "learning_rate": 0.0001598962834917891,
+      "loss": 0.6258,
+      "step": 1165
+    },
+    {
+      "epoch": 6.038709677419355,
+      "grad_norm": 1.1050348281860352,
+      "learning_rate": 0.00015986171132238548,
+      "loss": 0.5945,
+      "step": 1166
+    },
+    {
+      "epoch": 6.0438709677419356,
+      "grad_norm": 0.9840640425682068,
+      "learning_rate": 0.00015982713915298186,
+      "loss": 0.5637,
+      "step": 1167
+    },
+    {
+      "epoch": 6.049032258064516,
+      "grad_norm": 0.9715596437454224,
+      "learning_rate": 0.00015979256698357824,
+      "loss": 0.6198,
+      "step": 1168
+    },
+    {
+      "epoch": 6.054193548387096,
+      "grad_norm": 0.9735950231552124,
+      "learning_rate": 0.0001597579948141746,
+      "loss": 0.655,
+      "step": 1169
+    },
+    {
+      "epoch": 6.059354838709678,
+      "grad_norm": 0.8937168717384338,
+      "learning_rate": 0.00015972342264477095,
+      "loss": 0.5331,
+      "step": 1170
+    },
+    {
+      "epoch": 6.064516129032258,
+      "grad_norm": 1.16604745388031,
+      "learning_rate": 0.00015968885047536736,
+      "loss": 0.6187,
+      "step": 1171
+    },
+    {
+      "epoch": 6.069677419354838,
+      "grad_norm": 1.055781364440918,
+      "learning_rate": 0.0001596542783059637,
+      "loss": 0.6146,
+      "step": 1172
+    },
+    {
+      "epoch": 6.07483870967742,
+      "grad_norm": 1.5964897871017456,
+      "learning_rate": 0.00015961970613656006,
+      "loss": 0.5604,
+      "step": 1173
+    },
+    {
+      "epoch": 6.08,
+      "grad_norm": 1.0792604684829712,
+      "learning_rate": 0.00015958513396715644,
+      "loss": 0.5415,
+      "step": 1174
+    },
+    {
+      "epoch": 6.0851612903225805,
+      "grad_norm": 1.2483307123184204,
+      "learning_rate": 0.0001595505617977528,
+      "loss": 0.5956,
+      "step": 1175
+    },
+    {
+      "epoch": 6.090322580645161,
+      "grad_norm": 1.2184582948684692,
+      "learning_rate": 0.0001595159896283492,
+      "loss": 0.6139,
+      "step": 1176
+    },
+    {
+      "epoch": 6.095483870967742,
+      "grad_norm": 0.9432593584060669,
+      "learning_rate": 0.00015948141745894556,
+      "loss": 0.5991,
+      "step": 1177
+    },
+    {
+      "epoch": 6.1006451612903225,
+      "grad_norm": 1.0299996137619019,
+      "learning_rate": 0.0001594468452895419,
+      "loss": 0.5782,
+      "step": 1178
+    },
+    {
+      "epoch": 6.105806451612903,
+      "grad_norm": 1.0478683710098267,
+      "learning_rate": 0.0001594122731201383,
+      "loss": 0.6121,
+      "step": 1179
+    },
+    {
+      "epoch": 6.110967741935484,
+      "grad_norm": 1.0309667587280273,
+      "learning_rate": 0.00015937770095073467,
+      "loss": 0.5363,
+      "step": 1180
+    },
+    {
+      "epoch": 6.116129032258065,
+      "grad_norm": 1.0965996980667114,
+      "learning_rate": 0.00015934312878133105,
+      "loss": 0.5802,
+      "step": 1181
+    },
+    {
+      "epoch": 6.121290322580645,
+      "grad_norm": 0.9242439270019531,
+      "learning_rate": 0.0001593085566119274,
+      "loss": 0.5619,
+      "step": 1182
+    },
+    {
+      "epoch": 6.126451612903226,
+      "grad_norm": 1.094309687614441,
+      "learning_rate": 0.00015927398444252376,
+      "loss": 0.5604,
+      "step": 1183
+    },
+    {
+      "epoch": 6.131612903225807,
+      "grad_norm": 1.0809850692749023,
+      "learning_rate": 0.00015923941227312014,
+      "loss": 0.6067,
+      "step": 1184
+    },
+    {
+      "epoch": 6.136774193548387,
+      "grad_norm": 1.1499171257019043,
+      "learning_rate": 0.00015920484010371652,
+      "loss": 0.581,
+      "step": 1185
+    },
+    {
+      "epoch": 6.141935483870967,
+      "grad_norm": 1.1068612337112427,
+      "learning_rate": 0.0001591702679343129,
+      "loss": 0.6314,
+      "step": 1186
+    },
+    {
+      "epoch": 6.147096774193549,
+      "grad_norm": 1.0385297536849976,
+      "learning_rate": 0.00015913569576490926,
+      "loss": 0.5937,
+      "step": 1187
+    },
+    {
+      "epoch": 6.152258064516129,
+      "grad_norm": 1.013228178024292,
+      "learning_rate": 0.0001591011235955056,
+      "loss": 0.5629,
+      "step": 1188
+    },
+    {
+      "epoch": 6.1574193548387095,
+      "grad_norm": 1.0152748823165894,
+      "learning_rate": 0.00015906655142610202,
+      "loss": 0.5979,
+      "step": 1189
+    },
+    {
+      "epoch": 6.16258064516129,
+      "grad_norm": 1.0117329359054565,
+      "learning_rate": 0.00015903197925669837,
+      "loss": 0.6118,
+      "step": 1190
+    },
+    {
+      "epoch": 6.167741935483871,
+      "grad_norm": 1.0061877965927124,
+      "learning_rate": 0.00015899740708729472,
+      "loss": 0.5866,
+      "step": 1191
+    },
+    {
+      "epoch": 6.172903225806452,
+      "grad_norm": 0.9518502354621887,
+      "learning_rate": 0.0001589628349178911,
+      "loss": 0.5973,
+      "step": 1192
+    },
+    {
+      "epoch": 6.178064516129032,
+      "grad_norm": 1.1258810758590698,
+      "learning_rate": 0.00015892826274848746,
+      "loss": 0.5958,
+      "step": 1193
+    },
+    {
+      "epoch": 6.183225806451613,
+      "grad_norm": 0.9577953815460205,
+      "learning_rate": 0.00015889369057908387,
+      "loss": 0.5645,
+      "step": 1194
+    },
+    {
+      "epoch": 6.188387096774194,
+      "grad_norm": 1.0703774690628052,
+      "learning_rate": 0.00015885911840968022,
+      "loss": 0.6164,
+      "step": 1195
+    },
+    {
+      "epoch": 6.193548387096774,
+      "grad_norm": 1.1539541482925415,
+      "learning_rate": 0.00015882454624027657,
+      "loss": 0.6459,
+      "step": 1196
+    },
+    {
+      "epoch": 6.198709677419354,
+      "grad_norm": 1.0367482900619507,
+      "learning_rate": 0.00015878997407087295,
+      "loss": 0.5429,
+      "step": 1197
+    },
+    {
+      "epoch": 6.203870967741936,
+      "grad_norm": 0.9971298575401306,
+      "learning_rate": 0.0001587554019014693,
+      "loss": 0.6144,
+      "step": 1198
+    },
+    {
+      "epoch": 6.209032258064516,
+      "grad_norm": 0.9817131161689758,
+      "learning_rate": 0.0001587208297320657,
+      "loss": 0.567,
+      "step": 1199
+    },
+    {
+      "epoch": 6.2141935483870965,
+      "grad_norm": 0.9563860297203064,
+      "learning_rate": 0.00015868625756266207,
+      "loss": 0.5661,
+      "step": 1200
+    },
+    {
+      "epoch": 6.219354838709678,
+      "grad_norm": 1.0225104093551636,
+      "learning_rate": 0.00015865168539325842,
+      "loss": 0.5842,
+      "step": 1201
+    },
+    {
+      "epoch": 6.224516129032258,
+      "grad_norm": 1.1954907178878784,
+      "learning_rate": 0.0001586171132238548,
+      "loss": 0.5425,
+      "step": 1202
+    },
+    {
+      "epoch": 6.2296774193548385,
+      "grad_norm": 1.0410892963409424,
+      "learning_rate": 0.00015858254105445118,
+      "loss": 0.599,
+      "step": 1203
+    },
+    {
+      "epoch": 6.234838709677419,
+      "grad_norm": 1.0761679410934448,
+      "learning_rate": 0.00015854796888504756,
+      "loss": 0.6016,
+      "step": 1204
+    },
+    {
+      "epoch": 6.24,
+      "grad_norm": 1.0838356018066406,
+      "learning_rate": 0.00015851339671564392,
+      "loss": 0.6366,
+      "step": 1205
+    },
+    {
+      "epoch": 6.245161290322581,
+      "grad_norm": 1.0902661085128784,
+      "learning_rate": 0.00015847882454624027,
+      "loss": 0.6582,
+      "step": 1206
+    },
+    {
+      "epoch": 6.250322580645161,
+      "grad_norm": 1.005958080291748,
+      "learning_rate": 0.00015844425237683665,
+      "loss": 0.5587,
+      "step": 1207
+    },
+    {
+      "epoch": 6.255483870967742,
+      "grad_norm": 0.9429857730865479,
+      "learning_rate": 0.00015840968020743303,
+      "loss": 0.5807,
+      "step": 1208
+    },
+    {
+      "epoch": 6.260645161290323,
+      "grad_norm": 1.0042994022369385,
+      "learning_rate": 0.00015837510803802938,
+      "loss": 0.5658,
+      "step": 1209
+    },
+    {
+      "epoch": 6.265806451612903,
+      "grad_norm": 1.1328394412994385,
+      "learning_rate": 0.00015834053586862576,
+      "loss": 0.6001,
+      "step": 1210
+    },
+    {
+      "epoch": 6.270967741935484,
+      "grad_norm": 0.992401659488678,
+      "learning_rate": 0.00015830596369922212,
+      "loss": 0.5637,
+      "step": 1211
+    },
+    {
+      "epoch": 6.276129032258065,
+      "grad_norm": 1.0163962841033936,
+      "learning_rate": 0.00015827139152981852,
+      "loss": 0.593,
+      "step": 1212
+    },
+    {
+      "epoch": 6.281290322580645,
+      "grad_norm": 1.0893526077270508,
+      "learning_rate": 0.00015823681936041488,
+      "loss": 0.6339,
+      "step": 1213
+    },
+    {
+      "epoch": 6.2864516129032255,
+      "grad_norm": 1.053652286529541,
+      "learning_rate": 0.00015820224719101123,
+      "loss": 0.5966,
+      "step": 1214
+    },
+    {
+      "epoch": 6.291612903225807,
+      "grad_norm": 1.154402256011963,
+      "learning_rate": 0.0001581676750216076,
+      "loss": 0.6388,
+      "step": 1215
+    },
+    {
+      "epoch": 6.296774193548387,
+      "grad_norm": 1.003736972808838,
+      "learning_rate": 0.00015813310285220397,
+      "loss": 0.5708,
+      "step": 1216
+    },
+    {
+      "epoch": 6.301935483870968,
+      "grad_norm": 0.9400631785392761,
+      "learning_rate": 0.00015809853068280037,
+      "loss": 0.5991,
+      "step": 1217
+    },
+    {
+      "epoch": 6.307096774193548,
+      "grad_norm": 1.0645462274551392,
+      "learning_rate": 0.00015806395851339673,
+      "loss": 0.6089,
+      "step": 1218
+    },
+    {
+      "epoch": 6.312258064516129,
+      "grad_norm": 0.9896891713142395,
+      "learning_rate": 0.00015802938634399308,
+      "loss": 0.5706,
+      "step": 1219
+    },
+    {
+      "epoch": 6.31741935483871,
+      "grad_norm": 1.02813720703125,
+      "learning_rate": 0.00015799481417458946,
+      "loss": 0.6003,
+      "step": 1220
+    },
+    {
+      "epoch": 6.32258064516129,
+      "grad_norm": 1.101110816001892,
+      "learning_rate": 0.00015796024200518584,
+      "loss": 0.5899,
+      "step": 1221
+    },
+    {
+      "epoch": 6.327741935483871,
+      "grad_norm": 1.1690417528152466,
+      "learning_rate": 0.00015792566983578222,
+      "loss": 0.6008,
+      "step": 1222
+    },
+    {
+      "epoch": 6.332903225806452,
+      "grad_norm": 1.0852603912353516,
+      "learning_rate": 0.00015789109766637858,
+      "loss": 0.638,
+      "step": 1223
+    },
+    {
+      "epoch": 6.338064516129032,
+      "grad_norm": 1.1378023624420166,
+      "learning_rate": 0.00015785652549697493,
+      "loss": 0.6379,
+      "step": 1224
+    },
+    {
+      "epoch": 6.3432258064516125,
+      "grad_norm": 0.9421341419219971,
+      "learning_rate": 0.0001578219533275713,
+      "loss": 0.5968,
+      "step": 1225
+    },
+    {
+      "epoch": 6.348387096774194,
+      "grad_norm": 1.0730657577514648,
+      "learning_rate": 0.0001577873811581677,
+      "loss": 0.6416,
+      "step": 1226
+    },
+    {
+      "epoch": 6.353548387096774,
+      "grad_norm": 1.0939288139343262,
+      "learning_rate": 0.00015775280898876404,
+      "loss": 0.6338,
+      "step": 1227
+    },
+    {
+      "epoch": 6.3587096774193546,
+      "grad_norm": 1.0840965509414673,
+      "learning_rate": 0.00015771823681936042,
+      "loss": 0.6359,
+      "step": 1228
+    },
+    {
+      "epoch": 6.363870967741936,
+      "grad_norm": 0.9736648797988892,
+      "learning_rate": 0.00015768366464995678,
+      "loss": 0.6125,
+      "step": 1229
+    },
+    {
+      "epoch": 6.369032258064516,
+      "grad_norm": 1.012280821800232,
+      "learning_rate": 0.00015764909248055318,
+      "loss": 0.6014,
+      "step": 1230
+    },
+    {
+      "epoch": 6.374193548387097,
+      "grad_norm": 1.0381951332092285,
+      "learning_rate": 0.00015761452031114954,
+      "loss": 0.6205,
+      "step": 1231
+    },
+    {
+      "epoch": 6.379354838709677,
+      "grad_norm": 1.0712312459945679,
+      "learning_rate": 0.0001575799481417459,
+      "loss": 0.6454,
+      "step": 1232
+    },
+    {
+      "epoch": 6.384516129032258,
+      "grad_norm": 1.1191792488098145,
+      "learning_rate": 0.00015754537597234227,
+      "loss": 0.5945,
+      "step": 1233
+    },
+    {
+      "epoch": 6.389677419354839,
+      "grad_norm": 1.0398106575012207,
+      "learning_rate": 0.00015751080380293863,
+      "loss": 0.6494,
+      "step": 1234
+    },
+    {
+      "epoch": 6.394838709677419,
+      "grad_norm": 1.157550573348999,
+      "learning_rate": 0.00015747623163353503,
+      "loss": 0.6275,
+      "step": 1235
+    },
+    {
+      "epoch": 6.4,
+      "grad_norm": 1.0539355278015137,
+      "learning_rate": 0.0001574416594641314,
+      "loss": 0.6006,
+      "step": 1236
+    },
+    {
+      "epoch": 6.405161290322581,
+      "grad_norm": 2.192272424697876,
+      "learning_rate": 0.00015740708729472774,
+      "loss": 0.5753,
+      "step": 1237
+    },
+    {
+      "epoch": 6.410322580645161,
+      "grad_norm": 1.0441151857376099,
+      "learning_rate": 0.00015737251512532412,
+      "loss": 0.6396,
+      "step": 1238
+    },
+    {
+      "epoch": 6.4154838709677415,
+      "grad_norm": 1.0677762031555176,
+      "learning_rate": 0.00015733794295592047,
+      "loss": 0.6402,
+      "step": 1239
+    },
+    {
+      "epoch": 6.420645161290323,
+      "grad_norm": 1.0464727878570557,
+      "learning_rate": 0.00015730337078651685,
+      "loss": 0.634,
+      "step": 1240
+    },
+    {
+      "epoch": 6.425806451612903,
+      "grad_norm": 1.1254152059555054,
+      "learning_rate": 0.00015726879861711324,
+      "loss": 0.6718,
+      "step": 1241
+    },
+    {
+      "epoch": 6.430967741935484,
+      "grad_norm": 0.9842066764831543,
+      "learning_rate": 0.0001572342264477096,
+      "loss": 0.5329,
+      "step": 1242
+    },
+    {
+      "epoch": 6.436129032258065,
+      "grad_norm": 1.1525213718414307,
+      "learning_rate": 0.00015719965427830597,
+      "loss": 0.6087,
+      "step": 1243
+    },
+    {
+      "epoch": 6.441290322580645,
+      "grad_norm": 1.07542085647583,
+      "learning_rate": 0.00015716508210890235,
+      "loss": 0.5874,
+      "step": 1244
+    },
+    {
+      "epoch": 6.446451612903226,
+      "grad_norm": 1.2360172271728516,
+      "learning_rate": 0.0001571305099394987,
+      "loss": 0.656,
+      "step": 1245
+    },
+    {
+      "epoch": 6.451612903225806,
+      "grad_norm": 1.1499998569488525,
+      "learning_rate": 0.00015709593777009508,
+      "loss": 0.6343,
+      "step": 1246
+    },
+    {
+      "epoch": 6.456774193548387,
+      "grad_norm": 1.085065245628357,
+      "learning_rate": 0.00015706136560069144,
+      "loss": 0.6028,
+      "step": 1247
+    },
+    {
+      "epoch": 6.461935483870968,
+      "grad_norm": 1.0148816108703613,
+      "learning_rate": 0.00015702679343128782,
+      "loss": 0.6045,
+      "step": 1248
+    },
+    {
+      "epoch": 6.467096774193548,
+      "grad_norm": 1.0214478969573975,
+      "learning_rate": 0.0001569922212618842,
+      "loss": 0.6368,
+      "step": 1249
+    },
+    {
+      "epoch": 6.472258064516129,
+      "grad_norm": 1.0721983909606934,
+      "learning_rate": 0.00015695764909248055,
+      "loss": 0.6245,
+      "step": 1250
+    },
+    {
+      "epoch": 6.47741935483871,
+      "grad_norm": 1.066588044166565,
+      "learning_rate": 0.00015692307692307693,
+      "loss": 0.5774,
+      "step": 1251
+    },
+    {
+      "epoch": 6.48258064516129,
+      "grad_norm": 1.1868442296981812,
+      "learning_rate": 0.00015688850475367329,
+      "loss": 0.6738,
+      "step": 1252
+    },
+    {
+      "epoch": 6.487741935483871,
+      "grad_norm": 1.0562095642089844,
+      "learning_rate": 0.0001568539325842697,
+      "loss": 0.635,
+      "step": 1253
+    },
+    {
+      "epoch": 6.492903225806452,
+      "grad_norm": 0.9364386796951294,
+      "learning_rate": 0.00015681936041486605,
+      "loss": 0.5824,
+      "step": 1254
+    },
+    {
+      "epoch": 6.498064516129032,
+      "grad_norm": 1.0817921161651611,
+      "learning_rate": 0.0001567847882454624,
+      "loss": 0.5913,
+      "step": 1255
+    },
+    {
+      "epoch": 6.503225806451613,
+      "grad_norm": 1.1063023805618286,
+      "learning_rate": 0.00015675021607605878,
+      "loss": 0.6091,
+      "step": 1256
+    },
+    {
+      "epoch": 6.508387096774194,
+      "grad_norm": 1.1734906435012817,
+      "learning_rate": 0.00015671564390665513,
+      "loss": 0.6687,
+      "step": 1257
+    },
+    {
+      "epoch": 6.513548387096774,
+      "grad_norm": 1.050191044807434,
+      "learning_rate": 0.00015668107173725151,
+      "loss": 0.6591,
+      "step": 1258
+    },
+    {
+      "epoch": 6.518709677419355,
+      "grad_norm": 1.0617167949676514,
+      "learning_rate": 0.0001566464995678479,
+      "loss": 0.6214,
+      "step": 1259
+    },
+    {
+      "epoch": 6.523870967741935,
+      "grad_norm": 1.0583380460739136,
+      "learning_rate": 0.00015661192739844425,
+      "loss": 0.6057,
+      "step": 1260
+    },
+    {
+      "epoch": 6.529032258064516,
+      "grad_norm": 1.022868037223816,
+      "learning_rate": 0.00015657735522904063,
+      "loss": 0.628,
+      "step": 1261
+    },
+    {
+      "epoch": 6.534193548387097,
+      "grad_norm": 1.067391276359558,
+      "learning_rate": 0.000156542783059637,
+      "loss": 0.6114,
+      "step": 1262
+    },
+    {
+      "epoch": 6.539354838709677,
+      "grad_norm": 1.138838768005371,
+      "learning_rate": 0.00015650821089023336,
+      "loss": 0.621,
+      "step": 1263
+    },
+    {
+      "epoch": 6.544516129032258,
+      "grad_norm": 1.1029702425003052,
+      "learning_rate": 0.00015647363872082974,
+      "loss": 0.592,
+      "step": 1264
+    },
+    {
+      "epoch": 6.549677419354839,
+      "grad_norm": 1.0807851552963257,
+      "learning_rate": 0.0001564390665514261,
+      "loss": 0.639,
+      "step": 1265
+    },
+    {
+      "epoch": 6.554838709677419,
+      "grad_norm": 0.9720795154571533,
+      "learning_rate": 0.00015640449438202248,
+      "loss": 0.6252,
+      "step": 1266
+    },
+    {
+      "epoch": 6.5600000000000005,
+      "grad_norm": 1.0202288627624512,
+      "learning_rate": 0.00015636992221261886,
+      "loss": 0.6578,
+      "step": 1267
+    },
+    {
+      "epoch": 6.565161290322581,
+      "grad_norm": 0.9742851257324219,
+      "learning_rate": 0.0001563353500432152,
+      "loss": 0.5775,
+      "step": 1268
+    },
+    {
+      "epoch": 6.570322580645161,
+      "grad_norm": 1.016922950744629,
+      "learning_rate": 0.0001563007778738116,
+      "loss": 0.5631,
+      "step": 1269
+    },
+    {
+      "epoch": 6.575483870967742,
+      "grad_norm": 0.9774826169013977,
+      "learning_rate": 0.00015626620570440795,
+      "loss": 0.5682,
+      "step": 1270
+    },
+    {
+      "epoch": 6.580645161290323,
+      "grad_norm": 1.1458595991134644,
+      "learning_rate": 0.00015623163353500433,
+      "loss": 0.6459,
+      "step": 1271
+    },
+    {
+      "epoch": 6.585806451612903,
+      "grad_norm": 1.07783842086792,
+      "learning_rate": 0.0001561970613656007,
+      "loss": 0.6215,
+      "step": 1272
+    },
+    {
+      "epoch": 6.590967741935484,
+      "grad_norm": 1.1845786571502686,
+      "learning_rate": 0.00015616248919619706,
+      "loss": 0.6863,
+      "step": 1273
+    },
+    {
+      "epoch": 6.596129032258064,
+      "grad_norm": 1.1961218118667603,
+      "learning_rate": 0.00015612791702679344,
+      "loss": 0.5924,
+      "step": 1274
+    },
+    {
+      "epoch": 6.601290322580645,
+      "grad_norm": 1.1342319250106812,
+      "learning_rate": 0.0001560933448573898,
+      "loss": 0.6409,
+      "step": 1275
+    },
+    {
+      "epoch": 6.606451612903226,
+      "grad_norm": 1.0397125482559204,
+      "learning_rate": 0.00015605877268798617,
+      "loss": 0.5708,
+      "step": 1276
+    },
+    {
+      "epoch": 6.611612903225806,
+      "grad_norm": 1.132054090499878,
+      "learning_rate": 0.00015602420051858256,
+      "loss": 0.6586,
+      "step": 1277
+    },
+    {
+      "epoch": 6.6167741935483875,
+      "grad_norm": 1.041383147239685,
+      "learning_rate": 0.0001559896283491789,
+      "loss": 0.5698,
+      "step": 1278
+    },
+    {
+      "epoch": 6.621935483870968,
+      "grad_norm": 1.0820245742797852,
+      "learning_rate": 0.0001559550561797753,
+      "loss": 0.6489,
+      "step": 1279
+    },
+    {
+      "epoch": 6.627096774193548,
+      "grad_norm": 1.0891035795211792,
+      "learning_rate": 0.00015592048401037164,
+      "loss": 0.6732,
+      "step": 1280
+    },
+    {
+      "epoch": 6.632258064516129,
+      "grad_norm": 1.0189669132232666,
+      "learning_rate": 0.00015588591184096802,
+      "loss": 0.6267,
+      "step": 1281
+    },
+    {
+      "epoch": 6.63741935483871,
+      "grad_norm": 1.0899899005889893,
+      "learning_rate": 0.0001558513396715644,
+      "loss": 0.5471,
+      "step": 1282
+    },
+    {
+      "epoch": 6.64258064516129,
+      "grad_norm": 1.0849249362945557,
+      "learning_rate": 0.00015581676750216076,
+      "loss": 0.6124,
+      "step": 1283
+    },
+    {
+      "epoch": 6.647741935483871,
+      "grad_norm": 1.090311884880066,
+      "learning_rate": 0.00015578219533275714,
+      "loss": 0.6406,
+      "step": 1284
+    },
+    {
+      "epoch": 6.652903225806452,
+      "grad_norm": 1.0468347072601318,
+      "learning_rate": 0.00015574762316335352,
+      "loss": 0.6005,
+      "step": 1285
+    },
+    {
+      "epoch": 6.658064516129032,
+      "grad_norm": 1.107105016708374,
+      "learning_rate": 0.00015571305099394987,
+      "loss": 0.6545,
+      "step": 1286
+    },
+    {
+      "epoch": 6.663225806451613,
+      "grad_norm": 1.1085487604141235,
+      "learning_rate": 0.00015567847882454625,
+      "loss": 0.6583,
+      "step": 1287
+    },
+    {
+      "epoch": 6.668387096774193,
+      "grad_norm": 1.3117038011550903,
+      "learning_rate": 0.0001556439066551426,
+      "loss": 0.6661,
+      "step": 1288
+    },
+    {
+      "epoch": 6.6735483870967744,
+      "grad_norm": 1.0163108110427856,
+      "learning_rate": 0.00015560933448573899,
+      "loss": 0.6362,
+      "step": 1289
+    },
+    {
+      "epoch": 6.678709677419355,
+      "grad_norm": 1.0256747007369995,
+      "learning_rate": 0.00015557476231633537,
+      "loss": 0.6494,
+      "step": 1290
+    },
+    {
+      "epoch": 6.683870967741935,
+      "grad_norm": 1.0607695579528809,
+      "learning_rate": 0.00015554019014693172,
+      "loss": 0.6352,
+      "step": 1291
+    },
+    {
+      "epoch": 6.6890322580645165,
+      "grad_norm": 1.1411441564559937,
+      "learning_rate": 0.0001555056179775281,
+      "loss": 0.6343,
+      "step": 1292
+    },
+    {
+      "epoch": 6.694193548387097,
+      "grad_norm": 1.1059685945510864,
+      "learning_rate": 0.00015547104580812445,
+      "loss": 0.6067,
+      "step": 1293
+    },
+    {
+      "epoch": 6.699354838709677,
+      "grad_norm": 1.1858528852462769,
+      "learning_rate": 0.00015543647363872083,
+      "loss": 0.6742,
+      "step": 1294
+    },
+    {
+      "epoch": 6.704516129032258,
+      "grad_norm": 0.9606258273124695,
+      "learning_rate": 0.00015540190146931722,
+      "loss": 0.6349,
+      "step": 1295
+    },
+    {
+      "epoch": 6.709677419354839,
+      "grad_norm": 1.1092184782028198,
+      "learning_rate": 0.00015536732929991357,
+      "loss": 0.7033,
+      "step": 1296
+    },
+    {
+      "epoch": 6.714838709677419,
+      "grad_norm": 1.059554934501648,
+      "learning_rate": 0.00015533275713050995,
+      "loss": 0.6204,
+      "step": 1297
+    },
+    {
+      "epoch": 6.72,
+      "grad_norm": 1.1530689001083374,
+      "learning_rate": 0.0001552981849611063,
+      "loss": 0.6735,
+      "step": 1298
+    },
+    {
+      "epoch": 6.725161290322581,
+      "grad_norm": 1.0371016263961792,
+      "learning_rate": 0.00015526361279170268,
+      "loss": 0.5281,
+      "step": 1299
+    },
+    {
+      "epoch": 6.730322580645161,
+      "grad_norm": 1.1626323461532593,
+      "learning_rate": 0.00015522904062229906,
+      "loss": 0.6725,
+      "step": 1300
+    },
+    {
+      "epoch": 6.735483870967742,
+      "grad_norm": 1.006343126296997,
+      "learning_rate": 0.00015519446845289542,
+      "loss": 0.6244,
+      "step": 1301
+    },
+    {
+      "epoch": 6.740645161290322,
+      "grad_norm": 0.9943289160728455,
+      "learning_rate": 0.0001551598962834918,
+      "loss": 0.6148,
+      "step": 1302
+    },
+    {
+      "epoch": 6.7458064516129035,
+      "grad_norm": 1.0621330738067627,
+      "learning_rate": 0.00015512532411408818,
+      "loss": 0.598,
+      "step": 1303
+    },
+    {
+      "epoch": 6.750967741935484,
+      "grad_norm": 1.1347543001174927,
+      "learning_rate": 0.00015509075194468453,
+      "loss": 0.6538,
+      "step": 1304
+    },
+    {
+      "epoch": 6.756129032258064,
+      "grad_norm": 1.1031992435455322,
+      "learning_rate": 0.0001550561797752809,
+      "loss": 0.6518,
+      "step": 1305
+    },
+    {
+      "epoch": 6.7612903225806456,
+      "grad_norm": 1.1802830696105957,
+      "learning_rate": 0.00015502160760587727,
+      "loss": 0.671,
+      "step": 1306
+    },
+    {
+      "epoch": 6.766451612903226,
+      "grad_norm": 1.119118332862854,
+      "learning_rate": 0.00015498703543647365,
+      "loss": 0.6163,
+      "step": 1307
+    },
+    {
+      "epoch": 6.771612903225806,
+      "grad_norm": 1.1546950340270996,
+      "learning_rate": 0.00015495246326707003,
+      "loss": 0.6664,
+      "step": 1308
+    },
+    {
+      "epoch": 6.776774193548387,
+      "grad_norm": 1.0121320486068726,
+      "learning_rate": 0.00015491789109766638,
+      "loss": 0.5886,
+      "step": 1309
+    },
+    {
+      "epoch": 6.781935483870968,
+      "grad_norm": 1.098240613937378,
+      "learning_rate": 0.00015488331892826276,
+      "loss": 0.6559,
+      "step": 1310
+    },
+    {
+      "epoch": 6.787096774193548,
+      "grad_norm": 0.983754575252533,
+      "learning_rate": 0.00015484874675885911,
+      "loss": 0.6027,
+      "step": 1311
+    },
+    {
+      "epoch": 6.792258064516129,
+      "grad_norm": 1.1163573265075684,
+      "learning_rate": 0.0001548141745894555,
+      "loss": 0.6492,
+      "step": 1312
+    },
+    {
+      "epoch": 6.797419354838709,
+      "grad_norm": 1.0222346782684326,
+      "learning_rate": 0.00015477960242005188,
+      "loss": 0.5797,
+      "step": 1313
+    },
+    {
+      "epoch": 6.8025806451612905,
+      "grad_norm": 1.0159947872161865,
+      "learning_rate": 0.00015474503025064823,
+      "loss": 0.665,
+      "step": 1314
+    },
+    {
+      "epoch": 6.807741935483871,
+      "grad_norm": 0.9962977766990662,
+      "learning_rate": 0.0001547104580812446,
+      "loss": 0.6065,
+      "step": 1315
+    },
+    {
+      "epoch": 6.812903225806451,
+      "grad_norm": 1.015334963798523,
+      "learning_rate": 0.00015467588591184096,
+      "loss": 0.6257,
+      "step": 1316
+    },
+    {
+      "epoch": 6.8180645161290325,
+      "grad_norm": 1.1503382921218872,
+      "learning_rate": 0.00015464131374243734,
+      "loss": 0.6927,
+      "step": 1317
+    },
+    {
+      "epoch": 6.823225806451613,
+      "grad_norm": 1.0710428953170776,
+      "learning_rate": 0.00015460674157303372,
+      "loss": 0.6151,
+      "step": 1318
+    },
+    {
+      "epoch": 6.828387096774193,
+      "grad_norm": 1.1281623840332031,
+      "learning_rate": 0.00015457216940363008,
+      "loss": 0.5962,
+      "step": 1319
+    },
+    {
+      "epoch": 6.833548387096775,
+      "grad_norm": 1.1299762725830078,
+      "learning_rate": 0.00015453759723422646,
+      "loss": 0.6253,
+      "step": 1320
+    },
+    {
+      "epoch": 6.838709677419355,
+      "grad_norm": 1.177698016166687,
+      "learning_rate": 0.0001545030250648228,
+      "loss": 0.6608,
+      "step": 1321
+    },
+    {
+      "epoch": 6.843870967741935,
+      "grad_norm": 1.034962773323059,
+      "learning_rate": 0.0001544684528954192,
+      "loss": 0.6286,
+      "step": 1322
+    },
+    {
+      "epoch": 6.849032258064516,
+      "grad_norm": 1.1056431531906128,
+      "learning_rate": 0.00015443388072601557,
+      "loss": 0.6543,
+      "step": 1323
+    },
+    {
+      "epoch": 6.854193548387097,
+      "grad_norm": 1.4399502277374268,
+      "learning_rate": 0.00015439930855661193,
+      "loss": 0.679,
+      "step": 1324
+    },
+    {
+      "epoch": 6.859354838709677,
+      "grad_norm": 1.0068262815475464,
+      "learning_rate": 0.0001543647363872083,
+      "loss": 0.5575,
+      "step": 1325
+    },
+    {
+      "epoch": 6.864516129032258,
+      "grad_norm": 1.050834059715271,
+      "learning_rate": 0.0001543301642178047,
+      "loss": 0.6455,
+      "step": 1326
+    },
+    {
+      "epoch": 6.869677419354838,
+      "grad_norm": 1.1586737632751465,
+      "learning_rate": 0.00015429559204840104,
+      "loss": 0.6745,
+      "step": 1327
+    },
+    {
+      "epoch": 6.8748387096774195,
+      "grad_norm": 1.0993822813034058,
+      "learning_rate": 0.00015426101987899742,
+      "loss": 0.623,
+      "step": 1328
+    },
+    {
+      "epoch": 6.88,
+      "grad_norm": 1.4145764112472534,
+      "learning_rate": 0.00015422644770959377,
+      "loss": 0.6262,
+      "step": 1329
+    },
+    {
+      "epoch": 6.88516129032258,
+      "grad_norm": 1.106887936592102,
+      "learning_rate": 0.00015419187554019015,
+      "loss": 0.6535,
+      "step": 1330
+    },
+    {
+      "epoch": 6.890322580645162,
+      "grad_norm": 1.085783839225769,
+      "learning_rate": 0.00015415730337078654,
+      "loss": 0.5946,
+      "step": 1331
+    },
+    {
+      "epoch": 6.895483870967742,
+      "grad_norm": 1.1583377122879028,
+      "learning_rate": 0.0001541227312013829,
+      "loss": 0.6783,
+      "step": 1332
+    },
+    {
+      "epoch": 6.900645161290322,
+      "grad_norm": 1.0630621910095215,
+      "learning_rate": 0.00015408815903197927,
+      "loss": 0.6699,
+      "step": 1333
+    },
+    {
+      "epoch": 6.905806451612904,
+      "grad_norm": 1.0649069547653198,
+      "learning_rate": 0.00015405358686257562,
+      "loss": 0.6548,
+      "step": 1334
+    },
+    {
+      "epoch": 6.910967741935484,
+      "grad_norm": 1.1254299879074097,
+      "learning_rate": 0.000154019014693172,
+      "loss": 0.6455,
+      "step": 1335
+    },
+    {
+      "epoch": 6.916129032258064,
+      "grad_norm": 0.991092324256897,
+      "learning_rate": 0.00015398444252376838,
+      "loss": 0.6311,
+      "step": 1336
+    },
+    {
+      "epoch": 6.921290322580645,
+      "grad_norm": 1.0058887004852295,
+      "learning_rate": 0.00015394987035436474,
+      "loss": 0.6085,
+      "step": 1337
+    },
+    {
+      "epoch": 6.926451612903226,
+      "grad_norm": 1.0077121257781982,
+      "learning_rate": 0.00015391529818496112,
+      "loss": 0.5927,
+      "step": 1338
+    },
+    {
+      "epoch": 6.9316129032258065,
+      "grad_norm": 1.0935722589492798,
+      "learning_rate": 0.00015388072601555747,
+      "loss": 0.6687,
+      "step": 1339
+    },
+    {
+      "epoch": 6.936774193548387,
+      "grad_norm": 1.067617416381836,
+      "learning_rate": 0.00015384615384615385,
+      "loss": 0.6221,
+      "step": 1340
+    },
+    {
+      "epoch": 6.941935483870967,
+      "grad_norm": 1.1425668001174927,
+      "learning_rate": 0.00015381158167675023,
+      "loss": 0.6592,
+      "step": 1341
+    },
+    {
+      "epoch": 6.9470967741935485,
+      "grad_norm": 1.147863507270813,
+      "learning_rate": 0.00015377700950734659,
+      "loss": 0.6887,
+      "step": 1342
+    },
+    {
+      "epoch": 6.952258064516129,
+      "grad_norm": 1.0637719631195068,
+      "learning_rate": 0.00015374243733794297,
+      "loss": 0.685,
+      "step": 1343
+    },
+    {
+      "epoch": 6.957419354838709,
+      "grad_norm": 1.0223208665847778,
+      "learning_rate": 0.00015370786516853932,
+      "loss": 0.6076,
+      "step": 1344
+    },
+    {
+      "epoch": 6.962580645161291,
+      "grad_norm": 1.1109447479248047,
+      "learning_rate": 0.0001536732929991357,
+      "loss": 0.6624,
+      "step": 1345
+    },
+    {
+      "epoch": 6.967741935483871,
+      "grad_norm": 1.06330406665802,
+      "learning_rate": 0.00015363872082973208,
+      "loss": 0.6296,
+      "step": 1346
+    },
+    {
+      "epoch": 6.972903225806451,
+      "grad_norm": 1.2010246515274048,
+      "learning_rate": 0.00015360414866032843,
+      "loss": 0.7059,
+      "step": 1347
+    },
+    {
+      "epoch": 6.978064516129033,
+      "grad_norm": 1.0649375915527344,
+      "learning_rate": 0.00015356957649092481,
+      "loss": 0.656,
+      "step": 1348
+    },
+    {
+      "epoch": 6.983225806451613,
+      "grad_norm": 1.0661113262176514,
+      "learning_rate": 0.0001535350043215212,
+      "loss": 0.5752,
+      "step": 1349
+    },
+    {
+      "epoch": 6.9883870967741935,
+      "grad_norm": 1.136391282081604,
+      "learning_rate": 0.00015350043215211755,
+      "loss": 0.6751,
+      "step": 1350
+    },
+    {
+      "epoch": 6.993548387096774,
+      "grad_norm": 1.094062089920044,
+      "learning_rate": 0.00015346585998271393,
+      "loss": 0.6792,
+      "step": 1351
+    },
+    {
+      "epoch": 6.998709677419355,
+      "grad_norm": 1.7549867630004883,
+      "learning_rate": 0.00015343128781331028,
+      "loss": 0.9891,
+      "step": 1352
+    },
+    {
+      "epoch": 7.00516129032258,
+      "grad_norm": 0.6446539759635925,
+      "learning_rate": 0.00015339671564390666,
+      "loss": 0.4696,
+      "step": 1353
+    },
+    {
+      "epoch": 7.010322580645162,
+      "grad_norm": 0.6480467915534973,
+      "learning_rate": 0.00015336214347450304,
+      "loss": 0.4653,
+      "step": 1354
+    },
+    {
+      "epoch": 7.015483870967742,
+      "grad_norm": 0.7539472579956055,
+      "learning_rate": 0.0001533275713050994,
+      "loss": 0.4765,
+      "step": 1355
+    },
+    {
+      "epoch": 7.0206451612903225,
+      "grad_norm": 0.9261848330497742,
+      "learning_rate": 0.00015329299913569578,
+      "loss": 0.5183,
+      "step": 1356
+    },
+    {
+      "epoch": 7.025806451612903,
+      "grad_norm": 0.9499607682228088,
+      "learning_rate": 0.00015325842696629213,
+      "loss": 0.5006,
+      "step": 1357
+    },
+    {
+      "epoch": 7.030967741935484,
+      "grad_norm": 0.9341151118278503,
+      "learning_rate": 0.0001532238547968885,
+      "loss": 0.4963,
+      "step": 1358
+    },
+    {
+      "epoch": 7.0361290322580645,
+      "grad_norm": 0.7487500905990601,
+      "learning_rate": 0.0001531892826274849,
+      "loss": 0.4518,
+      "step": 1359
+    },
+    {
+      "epoch": 7.041290322580645,
+      "grad_norm": 1.0292613506317139,
+      "learning_rate": 0.00015315471045808125,
+      "loss": 0.4552,
+      "step": 1360
+    },
+    {
+      "epoch": 7.046451612903226,
+      "grad_norm": 1.007854700088501,
+      "learning_rate": 0.00015312013828867763,
+      "loss": 0.5171,
+      "step": 1361
+    },
+    {
+      "epoch": 7.051612903225807,
+      "grad_norm": 0.8299318552017212,
+      "learning_rate": 0.00015308556611927398,
+      "loss": 0.475,
+      "step": 1362
+    },
+    {
+      "epoch": 7.056774193548387,
+      "grad_norm": 0.7333000898361206,
+      "learning_rate": 0.00015305099394987036,
+      "loss": 0.4813,
+      "step": 1363
+    },
+    {
+      "epoch": 7.061935483870967,
+      "grad_norm": 0.8741991519927979,
+      "learning_rate": 0.00015301642178046674,
+      "loss": 0.5009,
+      "step": 1364
+    },
+    {
+      "epoch": 7.067096774193549,
+      "grad_norm": 0.8022140860557556,
+      "learning_rate": 0.0001529818496110631,
+      "loss": 0.4779,
+      "step": 1365
+    },
+    {
+      "epoch": 7.072258064516129,
+      "grad_norm": 0.7671019434928894,
+      "learning_rate": 0.00015294727744165947,
+      "loss": 0.4802,
+      "step": 1366
+    },
+    {
+      "epoch": 7.077419354838709,
+      "grad_norm": 0.8278824687004089,
+      "learning_rate": 0.00015291270527225585,
+      "loss": 0.4462,
+      "step": 1367
+    },
+    {
+      "epoch": 7.082580645161291,
+      "grad_norm": 0.8999887704849243,
+      "learning_rate": 0.0001528781331028522,
+      "loss": 0.4864,
+      "step": 1368
+    },
+    {
+      "epoch": 7.087741935483871,
+      "grad_norm": 0.9319064021110535,
+      "learning_rate": 0.0001528435609334486,
+      "loss": 0.521,
+      "step": 1369
+    },
+    {
+      "epoch": 7.0929032258064515,
+      "grad_norm": 0.8851963877677917,
+      "learning_rate": 0.00015280898876404494,
+      "loss": 0.4986,
+      "step": 1370
+    },
+    {
+      "epoch": 7.098064516129032,
+      "grad_norm": 1.044009804725647,
+      "learning_rate": 0.00015277441659464132,
+      "loss": 0.5216,
+      "step": 1371
+    },
+    {
+      "epoch": 7.103225806451613,
+      "grad_norm": 0.872248113155365,
+      "learning_rate": 0.0001527398444252377,
+      "loss": 0.4468,
+      "step": 1372
+    },
+    {
+      "epoch": 7.108387096774194,
+      "grad_norm": 0.8684067130088806,
+      "learning_rate": 0.00015270527225583406,
+      "loss": 0.5246,
+      "step": 1373
+    },
+    {
+      "epoch": 7.113548387096774,
+      "grad_norm": 0.8383293151855469,
+      "learning_rate": 0.00015267070008643044,
+      "loss": 0.4014,
+      "step": 1374
+    },
+    {
+      "epoch": 7.118709677419355,
+      "grad_norm": 1.0672733783721924,
+      "learning_rate": 0.0001526361279170268,
+      "loss": 0.474,
+      "step": 1375
+    },
+    {
+      "epoch": 7.123870967741936,
+      "grad_norm": 0.8337821364402771,
+      "learning_rate": 0.00015260155574762317,
+      "loss": 0.512,
+      "step": 1376
+    },
+    {
+      "epoch": 7.129032258064516,
+      "grad_norm": 0.8597740530967712,
+      "learning_rate": 0.00015256698357821955,
+      "loss": 0.527,
+      "step": 1377
+    },
+    {
+      "epoch": 7.134193548387096,
+      "grad_norm": 0.7548032999038696,
+      "learning_rate": 0.0001525324114088159,
+      "loss": 0.4355,
+      "step": 1378
+    },
+    {
+      "epoch": 7.139354838709678,
+      "grad_norm": 0.8028987050056458,
+      "learning_rate": 0.00015249783923941229,
+      "loss": 0.4022,
+      "step": 1379
+    },
+    {
+      "epoch": 7.144516129032258,
+      "grad_norm": 0.7867223024368286,
+      "learning_rate": 0.00015246326707000864,
+      "loss": 0.4412,
+      "step": 1380
+    },
+    {
+      "epoch": 7.1496774193548385,
+      "grad_norm": 0.8616959452629089,
+      "learning_rate": 0.00015242869490060502,
+      "loss": 0.4872,
+      "step": 1381
+    },
+    {
+      "epoch": 7.15483870967742,
+      "grad_norm": 0.846269428730011,
+      "learning_rate": 0.0001523941227312014,
+      "loss": 0.505,
+      "step": 1382
+    },
+    {
+      "epoch": 7.16,
+      "grad_norm": 0.8678281903266907,
+      "learning_rate": 0.00015235955056179775,
+      "loss": 0.4622,
+      "step": 1383
+    },
+    {
+      "epoch": 7.1651612903225805,
+      "grad_norm": 0.8200938701629639,
+      "learning_rate": 0.00015232497839239413,
+      "loss": 0.433,
+      "step": 1384
+    },
+    {
+      "epoch": 7.170322580645161,
+      "grad_norm": 0.7759093046188354,
+      "learning_rate": 0.0001522904062229905,
+      "loss": 0.4699,
+      "step": 1385
+    },
+    {
+      "epoch": 7.175483870967742,
+      "grad_norm": 0.9450991153717041,
+      "learning_rate": 0.00015225583405358687,
+      "loss": 0.5314,
+      "step": 1386
+    },
+    {
+      "epoch": 7.180645161290323,
+      "grad_norm": 0.7934908866882324,
+      "learning_rate": 0.00015222126188418325,
+      "loss": 0.5179,
+      "step": 1387
+    },
+    {
+      "epoch": 7.185806451612903,
+      "grad_norm": 0.966424822807312,
+      "learning_rate": 0.0001521866897147796,
+      "loss": 0.525,
+      "step": 1388
+    },
+    {
+      "epoch": 7.190967741935484,
+      "grad_norm": 0.8634904026985168,
+      "learning_rate": 0.00015215211754537598,
+      "loss": 0.5025,
+      "step": 1389
+    },
+    {
+      "epoch": 7.196129032258065,
+      "grad_norm": 0.9076803922653198,
+      "learning_rate": 0.00015211754537597236,
+      "loss": 0.5051,
+      "step": 1390
+    },
+    {
+      "epoch": 7.201290322580645,
+      "grad_norm": 0.8382192254066467,
+      "learning_rate": 0.00015208297320656872,
+      "loss": 0.486,
+      "step": 1391
+    },
+    {
+      "epoch": 7.2064516129032254,
+      "grad_norm": 0.903254508972168,
+      "learning_rate": 0.0001520484010371651,
+      "loss": 0.5401,
+      "step": 1392
+    },
+    {
+      "epoch": 7.211612903225807,
+      "grad_norm": 0.9082735180854797,
+      "learning_rate": 0.00015201382886776145,
+      "loss": 0.519,
+      "step": 1393
+    },
+    {
+      "epoch": 7.216774193548387,
+      "grad_norm": 0.8596190810203552,
+      "learning_rate": 0.00015197925669835783,
+      "loss": 0.4943,
+      "step": 1394
+    },
+    {
+      "epoch": 7.2219354838709675,
+      "grad_norm": 0.970220148563385,
+      "learning_rate": 0.0001519446845289542,
+      "loss": 0.4953,
+      "step": 1395
+    },
+    {
+      "epoch": 7.227096774193549,
+      "grad_norm": 0.9131587147712708,
+      "learning_rate": 0.00015191011235955057,
+      "loss": 0.5173,
+      "step": 1396
+    },
+    {
+      "epoch": 7.232258064516129,
+      "grad_norm": 0.8112829327583313,
+      "learning_rate": 0.00015187554019014695,
+      "loss": 0.4497,
+      "step": 1397
+    },
+    {
+      "epoch": 7.23741935483871,
+      "grad_norm": 1.036558747291565,
+      "learning_rate": 0.0001518409680207433,
+      "loss": 0.5122,
+      "step": 1398
+    },
+    {
+      "epoch": 7.24258064516129,
+      "grad_norm": 0.8115705847740173,
+      "learning_rate": 0.00015180639585133968,
+      "loss": 0.4503,
+      "step": 1399
+    },
+    {
+      "epoch": 7.247741935483871,
+      "grad_norm": 0.797034502029419,
+      "learning_rate": 0.00015177182368193606,
+      "loss": 0.4748,
+      "step": 1400
+    },
+    {
+      "epoch": 7.252903225806452,
+      "grad_norm": 1.1699554920196533,
+      "learning_rate": 0.00015173725151253241,
+      "loss": 0.5074,
+      "step": 1401
+    },
+    {
+      "epoch": 7.258064516129032,
+      "grad_norm": 0.8271426558494568,
+      "learning_rate": 0.0001517026793431288,
+      "loss": 0.4473,
+      "step": 1402
+    },
+    {
+      "epoch": 7.263225806451613,
+      "grad_norm": 0.9716576337814331,
+      "learning_rate": 0.00015166810717372515,
+      "loss": 0.5338,
+      "step": 1403
+    },
+    {
+      "epoch": 7.268387096774194,
+      "grad_norm": 0.8789535760879517,
+      "learning_rate": 0.00015163353500432153,
+      "loss": 0.5499,
+      "step": 1404
+    },
+    {
+      "epoch": 7.273548387096774,
+      "grad_norm": 0.8645904660224915,
+      "learning_rate": 0.0001515989628349179,
+      "loss": 0.4917,
+      "step": 1405
+    },
+    {
+      "epoch": 7.2787096774193545,
+      "grad_norm": 0.805229127407074,
+      "learning_rate": 0.00015156439066551426,
+      "loss": 0.4656,
+      "step": 1406
+    },
+    {
+      "epoch": 7.283870967741936,
+      "grad_norm": 0.8840547204017639,
+      "learning_rate": 0.00015152981849611064,
+      "loss": 0.4847,
+      "step": 1407
+    },
+    {
+      "epoch": 7.289032258064516,
+      "grad_norm": 0.8369616270065308,
+      "learning_rate": 0.00015149524632670702,
+      "loss": 0.4857,
+      "step": 1408
+    },
+    {
+      "epoch": 7.2941935483870965,
+      "grad_norm": 0.9104682207107544,
+      "learning_rate": 0.00015146067415730338,
+      "loss": 0.4842,
+      "step": 1409
+    },
+    {
+      "epoch": 7.299354838709678,
+      "grad_norm": 0.9500490427017212,
+      "learning_rate": 0.00015142610198789976,
+      "loss": 0.486,
+      "step": 1410
+    },
+    {
+      "epoch": 7.304516129032258,
+      "grad_norm": 0.9062269926071167,
+      "learning_rate": 0.0001513915298184961,
+      "loss": 0.476,
+      "step": 1411
+    },
+    {
+      "epoch": 7.309677419354839,
+      "grad_norm": 0.8756474256515503,
+      "learning_rate": 0.0001513569576490925,
+      "loss": 0.5105,
+      "step": 1412
+    },
+    {
+      "epoch": 7.314838709677419,
+      "grad_norm": 0.883924126625061,
+      "learning_rate": 0.00015132238547968887,
+      "loss": 0.5,
+      "step": 1413
+    },
+    {
+      "epoch": 7.32,
+      "grad_norm": 0.8824107050895691,
+      "learning_rate": 0.00015128781331028523,
+      "loss": 0.5084,
+      "step": 1414
+    },
+    {
+      "epoch": 7.325161290322581,
+      "grad_norm": 0.8614217638969421,
+      "learning_rate": 0.0001512532411408816,
+      "loss": 0.5021,
+      "step": 1415
+    },
+    {
+      "epoch": 7.330322580645161,
+      "grad_norm": 0.91167813539505,
+      "learning_rate": 0.00015121866897147796,
+      "loss": 0.5055,
+      "step": 1416
+    },
+    {
+      "epoch": 7.335483870967742,
+      "grad_norm": 0.837960422039032,
+      "learning_rate": 0.00015118409680207434,
+      "loss": 0.5182,
+      "step": 1417
+    },
+    {
+      "epoch": 7.340645161290323,
+      "grad_norm": 0.943089485168457,
+      "learning_rate": 0.00015114952463267072,
+      "loss": 0.5192,
+      "step": 1418
+    },
+    {
+      "epoch": 7.345806451612903,
+      "grad_norm": 0.8962975740432739,
+      "learning_rate": 0.00015111495246326707,
+      "loss": 0.4975,
+      "step": 1419
+    },
+    {
+      "epoch": 7.3509677419354835,
+      "grad_norm": 1.0125638246536255,
+      "learning_rate": 0.00015108038029386345,
+      "loss": 0.5113,
+      "step": 1420
+    },
+    {
+      "epoch": 7.356129032258065,
+      "grad_norm": 0.8467827439308167,
+      "learning_rate": 0.0001510458081244598,
+      "loss": 0.4884,
+      "step": 1421
+    },
+    {
+      "epoch": 7.361290322580645,
+      "grad_norm": 0.8722594976425171,
+      "learning_rate": 0.0001510112359550562,
+      "loss": 0.5024,
+      "step": 1422
+    },
+    {
+      "epoch": 7.366451612903226,
+      "grad_norm": 0.9010210633277893,
+      "learning_rate": 0.00015097666378565257,
+      "loss": 0.4978,
+      "step": 1423
+    },
+    {
+      "epoch": 7.371612903225807,
+      "grad_norm": 1.0606017112731934,
+      "learning_rate": 0.00015094209161624892,
+      "loss": 0.5184,
+      "step": 1424
+    },
+    {
+      "epoch": 7.376774193548387,
+      "grad_norm": 0.8984857797622681,
+      "learning_rate": 0.0001509075194468453,
+      "loss": 0.534,
+      "step": 1425
+    },
+    {
+      "epoch": 7.381935483870968,
+      "grad_norm": 0.7991496920585632,
+      "learning_rate": 0.00015087294727744166,
+      "loss": 0.4713,
+      "step": 1426
+    },
+    {
+      "epoch": 7.387096774193548,
+      "grad_norm": 0.8343316316604614,
+      "learning_rate": 0.00015083837510803804,
+      "loss": 0.5001,
+      "step": 1427
+    },
+    {
+      "epoch": 7.392258064516129,
+      "grad_norm": 0.9725926518440247,
+      "learning_rate": 0.00015080380293863442,
+      "loss": 0.5414,
+      "step": 1428
+    },
+    {
+      "epoch": 7.39741935483871,
+      "grad_norm": 1.4883674383163452,
+      "learning_rate": 0.00015076923076923077,
+      "loss": 0.4734,
+      "step": 1429
+    },
+    {
+      "epoch": 7.40258064516129,
+      "grad_norm": 0.8751910328865051,
+      "learning_rate": 0.00015073465859982715,
+      "loss": 0.5057,
+      "step": 1430
+    },
+    {
+      "epoch": 7.407741935483871,
+      "grad_norm": 0.780537486076355,
+      "learning_rate": 0.00015070008643042353,
+      "loss": 0.4353,
+      "step": 1431
+    },
+    {
+      "epoch": 7.412903225806452,
+      "grad_norm": 0.9572322964668274,
+      "learning_rate": 0.00015066551426101989,
+      "loss": 0.4898,
+      "step": 1432
+    },
+    {
+      "epoch": 7.418064516129032,
+      "grad_norm": 0.8475635647773743,
+      "learning_rate": 0.00015063094209161627,
+      "loss": 0.501,
+      "step": 1433
+    },
+    {
+      "epoch": 7.423225806451613,
+      "grad_norm": 0.8863414525985718,
+      "learning_rate": 0.00015059636992221262,
+      "loss": 0.4699,
+      "step": 1434
+    },
+    {
+      "epoch": 7.428387096774194,
+      "grad_norm": 0.8363776206970215,
+      "learning_rate": 0.000150561797752809,
+      "loss": 0.4578,
+      "step": 1435
+    },
+    {
+      "epoch": 7.433548387096774,
+      "grad_norm": 0.9768350124359131,
+      "learning_rate": 0.00015052722558340538,
+      "loss": 0.5565,
+      "step": 1436
+    },
+    {
+      "epoch": 7.438709677419355,
+      "grad_norm": 0.9602251052856445,
+      "learning_rate": 0.00015049265341400173,
+      "loss": 0.5463,
+      "step": 1437
+    },
+    {
+      "epoch": 7.443870967741935,
+      "grad_norm": 0.9805073142051697,
+      "learning_rate": 0.00015045808124459811,
+      "loss": 0.4635,
+      "step": 1438
+    },
+    {
+      "epoch": 7.449032258064516,
+      "grad_norm": 0.9675642848014832,
+      "learning_rate": 0.00015042350907519447,
+      "loss": 0.4847,
+      "step": 1439
+    },
+    {
+      "epoch": 7.454193548387097,
+      "grad_norm": 0.9387705326080322,
+      "learning_rate": 0.00015038893690579085,
+      "loss": 0.5264,
+      "step": 1440
+    },
+    {
+      "epoch": 7.459354838709677,
+      "grad_norm": 0.8858237862586975,
+      "learning_rate": 0.00015035436473638723,
+      "loss": 0.5161,
+      "step": 1441
+    },
+    {
+      "epoch": 7.464516129032258,
+      "grad_norm": 0.8779551386833191,
+      "learning_rate": 0.00015031979256698358,
+      "loss": 0.4934,
+      "step": 1442
+    },
+    {
+      "epoch": 7.469677419354839,
+      "grad_norm": 0.8100745677947998,
+      "learning_rate": 0.00015028522039757996,
+      "loss": 0.4868,
+      "step": 1443
+    },
+    {
+      "epoch": 7.474838709677419,
+      "grad_norm": 0.8881556987762451,
+      "learning_rate": 0.00015025064822817632,
+      "loss": 0.4787,
+      "step": 1444
+    },
+    {
+      "epoch": 7.48,
+      "grad_norm": 0.9295876026153564,
+      "learning_rate": 0.0001502160760587727,
+      "loss": 0.5253,
+      "step": 1445
+    },
+    {
+      "epoch": 7.485161290322581,
+      "grad_norm": 0.9915912747383118,
+      "learning_rate": 0.00015018150388936908,
+      "loss": 0.5177,
+      "step": 1446
+    },
+    {
+      "epoch": 7.490322580645161,
+      "grad_norm": 0.9058102965354919,
+      "learning_rate": 0.00015014693171996543,
+      "loss": 0.5226,
+      "step": 1447
+    },
+    {
+      "epoch": 7.495483870967742,
+      "grad_norm": 0.8928213715553284,
+      "learning_rate": 0.0001501123595505618,
+      "loss": 0.5394,
+      "step": 1448
+    },
+    {
+      "epoch": 7.500645161290323,
+      "grad_norm": 0.8370668888092041,
+      "learning_rate": 0.0001500777873811582,
+      "loss": 0.5174,
+      "step": 1449
+    },
+    {
+      "epoch": 7.505806451612903,
+      "grad_norm": 0.980941116809845,
+      "learning_rate": 0.00015004321521175455,
+      "loss": 0.5233,
+      "step": 1450
+    },
+    {
+      "epoch": 7.505806451612903,
+      "eval_loss": 2.013017177581787,
+      "eval_runtime": 21.3068,
+      "eval_samples_per_second": 3.708,
+      "eval_steps_per_second": 0.469,
+      "step": 1450
+    },
+    {
+      "epoch": 7.510967741935484,
+      "grad_norm": 0.9632622599601746,
+      "learning_rate": 0.00015000864304235093,
+      "loss": 0.536,
+      "step": 1451
+    },
+    {
+      "epoch": 7.516129032258064,
+      "grad_norm": 0.9036365747451782,
+      "learning_rate": 0.00014997407087294728,
+      "loss": 0.4524,
+      "step": 1452
+    },
+    {
+      "epoch": 7.521290322580645,
+      "grad_norm": 1.0012192726135254,
+      "learning_rate": 0.00014993949870354366,
+      "loss": 0.5011,
+      "step": 1453
+    },
+    {
+      "epoch": 7.526451612903226,
+      "grad_norm": 0.9224442839622498,
+      "learning_rate": 0.00014990492653414004,
+      "loss": 0.5184,
+      "step": 1454
+    },
+    {
+      "epoch": 7.531612903225806,
+      "grad_norm": 0.898026168346405,
+      "learning_rate": 0.0001498703543647364,
+      "loss": 0.5278,
+      "step": 1455
+    },
+    {
+      "epoch": 7.536774193548387,
+      "grad_norm": 0.931157112121582,
+      "learning_rate": 0.00014983578219533277,
+      "loss": 0.5549,
+      "step": 1456
+    },
+    {
+      "epoch": 7.541935483870968,
+      "grad_norm": 0.891901969909668,
+      "learning_rate": 0.00014980121002592913,
+      "loss": 0.5181,
+      "step": 1457
+    },
+    {
+      "epoch": 7.547096774193548,
+      "grad_norm": 0.9420850276947021,
+      "learning_rate": 0.00014976663785652548,
+      "loss": 0.5463,
+      "step": 1458
+    },
+    {
+      "epoch": 7.5522580645161295,
+      "grad_norm": 0.8835694789886475,
+      "learning_rate": 0.0001497320656871219,
+      "loss": 0.5388,
+      "step": 1459
+    },
+    {
+      "epoch": 7.55741935483871,
+      "grad_norm": 0.8482972383499146,
+      "learning_rate": 0.00014969749351771824,
+      "loss": 0.5153,
+      "step": 1460
+    },
+    {
+      "epoch": 7.56258064516129,
+      "grad_norm": 0.9354886412620544,
+      "learning_rate": 0.00014966292134831462,
+      "loss": 0.5463,
+      "step": 1461
+    },
+    {
+      "epoch": 7.567741935483871,
+      "grad_norm": 0.8587084412574768,
+      "learning_rate": 0.00014962834917891098,
+      "loss": 0.4777,
+      "step": 1462
+    },
+    {
+      "epoch": 7.572903225806452,
+      "grad_norm": 0.963616669178009,
+      "learning_rate": 0.00014959377700950736,
+      "loss": 0.5051,
+      "step": 1463
+    },
+    {
+      "epoch": 7.578064516129032,
+      "grad_norm": 0.9578056335449219,
+      "learning_rate": 0.00014955920484010374,
+      "loss": 0.5453,
+      "step": 1464
+    },
+    {
+      "epoch": 7.583225806451613,
+      "grad_norm": 0.9405192732810974,
+      "learning_rate": 0.0001495246326707001,
+      "loss": 0.5032,
+      "step": 1465
+    },
+    {
+      "epoch": 7.588387096774193,
+      "grad_norm": 0.9591034650802612,
+      "learning_rate": 0.00014949006050129647,
+      "loss": 0.5441,
+      "step": 1466
+    },
+    {
+      "epoch": 7.593548387096774,
+      "grad_norm": 0.9607383608818054,
+      "learning_rate": 0.00014945548833189282,
+      "loss": 0.4882,
+      "step": 1467
+    },
+    {
+      "epoch": 7.598709677419355,
+      "grad_norm": 0.9627270102500916,
+      "learning_rate": 0.0001494209161624892,
+      "loss": 0.529,
+      "step": 1468
+    },
+    {
+      "epoch": 7.603870967741935,
+      "grad_norm": 0.9902101755142212,
+      "learning_rate": 0.00014938634399308559,
+      "loss": 0.5645,
+      "step": 1469
+    },
+    {
+      "epoch": 7.609032258064516,
+      "grad_norm": 0.9811556339263916,
+      "learning_rate": 0.00014935177182368194,
+      "loss": 0.552,
+      "step": 1470
+    },
+    {
+      "epoch": 7.614193548387097,
+      "grad_norm": 0.9566178917884827,
+      "learning_rate": 0.0001493171996542783,
+      "loss": 0.5641,
+      "step": 1471
+    },
+    {
+      "epoch": 7.619354838709677,
+      "grad_norm": 0.9235225319862366,
+      "learning_rate": 0.0001492826274848747,
+      "loss": 0.5035,
+      "step": 1472
+    },
+    {
+      "epoch": 7.6245161290322585,
+      "grad_norm": 0.9713773727416992,
+      "learning_rate": 0.00014924805531547105,
+      "loss": 0.5441,
+      "step": 1473
+    },
+    {
+      "epoch": 7.629677419354839,
+      "grad_norm": 0.9252253174781799,
+      "learning_rate": 0.00014921348314606743,
+      "loss": 0.5231,
+      "step": 1474
+    },
+    {
+      "epoch": 7.634838709677419,
+      "grad_norm": 0.9318074584007263,
+      "learning_rate": 0.0001491789109766638,
+      "loss": 0.5471,
+      "step": 1475
+    },
+    {
+      "epoch": 7.64,
+      "grad_norm": 0.9392902851104736,
+      "learning_rate": 0.00014914433880726014,
+      "loss": 0.5404,
+      "step": 1476
+    },
+    {
+      "epoch": 7.645161290322581,
+      "grad_norm": 0.9631742835044861,
+      "learning_rate": 0.00014910976663785655,
+      "loss": 0.5409,
+      "step": 1477
+    },
+    {
+      "epoch": 7.650322580645161,
+      "grad_norm": 1.0287883281707764,
+      "learning_rate": 0.0001490751944684529,
+      "loss": 0.5651,
+      "step": 1478
+    },
+    {
+      "epoch": 7.655483870967742,
+      "grad_norm": 0.9384641051292419,
+      "learning_rate": 0.00014904062229904928,
+      "loss": 0.504,
+      "step": 1479
+    },
+    {
+      "epoch": 7.660645161290322,
+      "grad_norm": 0.9910617470741272,
+      "learning_rate": 0.00014900605012964564,
+      "loss": 0.6019,
+      "step": 1480
+    },
+    {
+      "epoch": 7.665806451612903,
+      "grad_norm": 0.988002359867096,
+      "learning_rate": 0.00014897147796024202,
+      "loss": 0.5351,
+      "step": 1481
+    },
+    {
+      "epoch": 7.670967741935484,
+      "grad_norm": 0.9243808388710022,
+      "learning_rate": 0.0001489369057908384,
+      "loss": 0.5218,
+      "step": 1482
+    },
+    {
+      "epoch": 7.676129032258064,
+      "grad_norm": 0.96613609790802,
+      "learning_rate": 0.00014890233362143475,
+      "loss": 0.5414,
+      "step": 1483
+    },
+    {
+      "epoch": 7.6812903225806455,
+      "grad_norm": 0.9242269992828369,
+      "learning_rate": 0.00014886776145203113,
+      "loss": 0.5338,
+      "step": 1484
+    },
+    {
+      "epoch": 7.686451612903226,
+      "grad_norm": 0.9686539173126221,
+      "learning_rate": 0.00014883318928262748,
+      "loss": 0.5418,
+      "step": 1485
+    },
+    {
+      "epoch": 7.691612903225806,
+      "grad_norm": 0.9492871165275574,
+      "learning_rate": 0.00014879861711322386,
+      "loss": 0.5225,
+      "step": 1486
+    },
+    {
+      "epoch": 7.6967741935483875,
+      "grad_norm": 1.0394287109375,
+      "learning_rate": 0.00014876404494382025,
+      "loss": 0.5142,
+      "step": 1487
+    },
+    {
+      "epoch": 7.701935483870968,
+      "grad_norm": 0.9882715344429016,
+      "learning_rate": 0.0001487294727744166,
+      "loss": 0.5366,
+      "step": 1488
+    },
+    {
+      "epoch": 7.707096774193548,
+      "grad_norm": 0.9601888060569763,
+      "learning_rate": 0.00014869490060501295,
+      "loss": 0.5266,
+      "step": 1489
+    },
+    {
+      "epoch": 7.712258064516129,
+      "grad_norm": 0.9324442744255066,
+      "learning_rate": 0.00014866032843560933,
+      "loss": 0.5192,
+      "step": 1490
+    },
+    {
+      "epoch": 7.71741935483871,
+      "grad_norm": 0.8981822729110718,
+      "learning_rate": 0.0001486257562662057,
+      "loss": 0.5418,
+      "step": 1491
+    },
+    {
+      "epoch": 7.72258064516129,
+      "grad_norm": 1.1587419509887695,
+      "learning_rate": 0.0001485911840968021,
+      "loss": 0.5127,
+      "step": 1492
+    },
+    {
+      "epoch": 7.727741935483871,
+      "grad_norm": 0.9748970866203308,
+      "learning_rate": 0.00014855661192739845,
+      "loss": 0.5266,
+      "step": 1493
+    },
+    {
+      "epoch": 7.732903225806451,
+      "grad_norm": 0.8825200200080872,
+      "learning_rate": 0.0001485220397579948,
+      "loss": 0.5454,
+      "step": 1494
+    },
+    {
+      "epoch": 7.7380645161290325,
+      "grad_norm": 0.9848824739456177,
+      "learning_rate": 0.0001484874675885912,
+      "loss": 0.5434,
+      "step": 1495
+    },
+    {
+      "epoch": 7.743225806451613,
+      "grad_norm": 1.0212171077728271,
+      "learning_rate": 0.00014845289541918756,
+      "loss": 0.5418,
+      "step": 1496
+    },
+    {
+      "epoch": 7.748387096774193,
+      "grad_norm": 0.9528943300247192,
+      "learning_rate": 0.00014841832324978394,
+      "loss": 0.5787,
+      "step": 1497
+    },
+    {
+      "epoch": 7.7535483870967745,
+      "grad_norm": 0.9677993059158325,
+      "learning_rate": 0.0001483837510803803,
+      "loss": 0.5551,
+      "step": 1498
+    },
+    {
+      "epoch": 7.758709677419355,
+      "grad_norm": 0.9510045051574707,
+      "learning_rate": 0.00014834917891097665,
+      "loss": 0.5282,
+      "step": 1499
+    },
+    {
+      "epoch": 7.763870967741935,
+      "grad_norm": 1.0408774614334106,
+      "learning_rate": 0.00014831460674157306,
+      "loss": 0.5831,
+      "step": 1500
+    },
+    {
+      "epoch": 7.769032258064517,
+      "grad_norm": 0.9247439503669739,
+      "learning_rate": 0.0001482800345721694,
+      "loss": 0.5459,
+      "step": 1501
+    },
+    {
+      "epoch": 7.774193548387097,
+      "grad_norm": 0.8859182000160217,
+      "learning_rate": 0.0001482454624027658,
+      "loss": 0.5019,
+      "step": 1502
+    },
+    {
+      "epoch": 7.779354838709677,
+      "grad_norm": 0.9574293494224548,
+      "learning_rate": 0.00014821089023336214,
+      "loss": 0.53,
+      "step": 1503
+    },
+    {
+      "epoch": 7.784516129032258,
+      "grad_norm": 0.9493100047111511,
+      "learning_rate": 0.00014817631806395852,
+      "loss": 0.5365,
+      "step": 1504
+    },
+    {
+      "epoch": 7.789677419354839,
+      "grad_norm": 0.8856602907180786,
+      "learning_rate": 0.0001481417458945549,
+      "loss": 0.5072,
+      "step": 1505
+    },
+    {
+      "epoch": 7.794838709677419,
+      "grad_norm": 0.9487787485122681,
+      "learning_rate": 0.00014810717372515126,
+      "loss": 0.4878,
+      "step": 1506
+    },
+    {
+      "epoch": 7.8,
+      "grad_norm": 1.0355792045593262,
+      "learning_rate": 0.0001480726015557476,
+      "loss": 0.4882,
+      "step": 1507
+    },
+    {
+      "epoch": 7.80516129032258,
+      "grad_norm": 0.9669519662857056,
+      "learning_rate": 0.000148038029386344,
+      "loss": 0.5286,
+      "step": 1508
+    },
+    {
+      "epoch": 7.8103225806451615,
+      "grad_norm": 0.9685948491096497,
+      "learning_rate": 0.00014800345721694037,
+      "loss": 0.5298,
+      "step": 1509
+    },
+    {
+      "epoch": 7.815483870967742,
+      "grad_norm": 1.0399237871170044,
+      "learning_rate": 0.00014796888504753675,
+      "loss": 0.5564,
+      "step": 1510
+    },
+    {
+      "epoch": 7.820645161290322,
+      "grad_norm": 0.9737111330032349,
+      "learning_rate": 0.0001479343128781331,
+      "loss": 0.5402,
+      "step": 1511
+    },
+    {
+      "epoch": 7.825806451612904,
+      "grad_norm": 1.036259651184082,
+      "learning_rate": 0.00014789974070872946,
+      "loss": 0.5403,
+      "step": 1512
+    },
+    {
+      "epoch": 7.830967741935484,
+      "grad_norm": 1.0763616561889648,
+      "learning_rate": 0.00014786516853932587,
+      "loss": 0.5744,
+      "step": 1513
+    },
+    {
+      "epoch": 7.836129032258064,
+      "grad_norm": 0.9492955803871155,
+      "learning_rate": 0.00014783059636992222,
+      "loss": 0.5363,
+      "step": 1514
+    },
+    {
+      "epoch": 7.841290322580646,
+      "grad_norm": 0.9868084192276001,
+      "learning_rate": 0.0001477960242005186,
+      "loss": 0.5392,
+      "step": 1515
+    },
+    {
+      "epoch": 7.846451612903226,
+      "grad_norm": 1.3577789068222046,
+      "learning_rate": 0.00014776145203111496,
+      "loss": 0.538,
+      "step": 1516
+    },
+    {
+      "epoch": 7.851612903225806,
+      "grad_norm": 0.8922302722930908,
+      "learning_rate": 0.0001477268798617113,
+      "loss": 0.5066,
+      "step": 1517
+    },
+    {
+      "epoch": 7.856774193548387,
+      "grad_norm": 0.9309819340705872,
+      "learning_rate": 0.00014769230769230772,
+      "loss": 0.5434,
+      "step": 1518
+    },
+    {
+      "epoch": 7.861935483870968,
+      "grad_norm": 0.9251433610916138,
+      "learning_rate": 0.00014765773552290407,
+      "loss": 0.5152,
+      "step": 1519
+    },
+    {
+      "epoch": 7.8670967741935485,
+      "grad_norm": 0.9834603071212769,
+      "learning_rate": 0.00014762316335350045,
+      "loss": 0.5374,
+      "step": 1520
+    },
+    {
+      "epoch": 7.872258064516129,
+      "grad_norm": 0.9278445243835449,
+      "learning_rate": 0.0001475885911840968,
+      "loss": 0.5179,
+      "step": 1521
+    },
+    {
+      "epoch": 7.877419354838709,
+      "grad_norm": 0.9892044067382812,
+      "learning_rate": 0.00014755401901469318,
+      "loss": 0.5613,
+      "step": 1522
+    },
+    {
+      "epoch": 7.8825806451612905,
+      "grad_norm": 1.0614113807678223,
+      "learning_rate": 0.00014751944684528957,
+      "loss": 0.5767,
+      "step": 1523
+    },
+    {
+      "epoch": 7.887741935483871,
+      "grad_norm": 0.8457578420639038,
+      "learning_rate": 0.00014748487467588592,
+      "loss": 0.441,
+      "step": 1524
+    },
+    {
+      "epoch": 7.892903225806451,
+      "grad_norm": 0.9680014252662659,
+      "learning_rate": 0.00014745030250648227,
+      "loss": 0.5376,
+      "step": 1525
+    },
+    {
+      "epoch": 7.898064516129033,
+      "grad_norm": 0.9269333481788635,
+      "learning_rate": 0.00014741573033707865,
+      "loss": 0.5489,
+      "step": 1526
+    },
+    {
+      "epoch": 7.903225806451613,
+      "grad_norm": 0.9175553321838379,
+      "learning_rate": 0.00014738115816767503,
+      "loss": 0.4972,
+      "step": 1527
+    },
+    {
+      "epoch": 7.908387096774193,
+      "grad_norm": 1.0178897380828857,
+      "learning_rate": 0.00014734658599827141,
+      "loss": 0.5299,
+      "step": 1528
+    },
+    {
+      "epoch": 7.913548387096775,
+      "grad_norm": 0.9134308099746704,
+      "learning_rate": 0.00014731201382886777,
+      "loss": 0.5264,
+      "step": 1529
+    },
+    {
+      "epoch": 7.918709677419355,
+      "grad_norm": 0.9683631062507629,
+      "learning_rate": 0.00014727744165946412,
+      "loss": 0.5628,
+      "step": 1530
+    },
+    {
+      "epoch": 7.9238709677419354,
+      "grad_norm": 0.9759796857833862,
+      "learning_rate": 0.0001472428694900605,
+      "loss": 0.545,
+      "step": 1531
+    },
+    {
+      "epoch": 7.929032258064516,
+      "grad_norm": 1.009869933128357,
+      "learning_rate": 0.00014720829732065688,
+      "loss": 0.5433,
+      "step": 1532
+    },
+    {
+      "epoch": 7.934193548387097,
+      "grad_norm": 1.186924695968628,
+      "learning_rate": 0.00014717372515125326,
+      "loss": 0.5514,
+      "step": 1533
+    },
+    {
+      "epoch": 7.9393548387096775,
+      "grad_norm": 0.9221572279930115,
+      "learning_rate": 0.00014713915298184962,
+      "loss": 0.5397,
+      "step": 1534
+    },
+    {
+      "epoch": 7.944516129032258,
+      "grad_norm": 0.9694724082946777,
+      "learning_rate": 0.00014710458081244597,
+      "loss": 0.553,
+      "step": 1535
+    },
+    {
+      "epoch": 7.949677419354838,
+      "grad_norm": 0.977093517780304,
+      "learning_rate": 0.00014707000864304238,
+      "loss": 0.4828,
+      "step": 1536
+    },
+    {
+      "epoch": 7.95483870967742,
+      "grad_norm": 0.9063377976417542,
+      "learning_rate": 0.00014703543647363873,
+      "loss": 0.4176,
+      "step": 1537
+    },
+    {
+      "epoch": 7.96,
+      "grad_norm": 1.5321789979934692,
+      "learning_rate": 0.0001470008643042351,
+      "loss": 0.5639,
+      "step": 1538
+    },
+    {
+      "epoch": 7.96516129032258,
+      "grad_norm": 0.8826190829277039,
+      "learning_rate": 0.00014696629213483146,
+      "loss": 0.5583,
+      "step": 1539
+    },
+    {
+      "epoch": 7.970322580645162,
+      "grad_norm": 0.9844704270362854,
+      "learning_rate": 0.00014693171996542782,
+      "loss": 0.5298,
+      "step": 1540
+    },
+    {
+      "epoch": 7.975483870967742,
+      "grad_norm": 1.0041615962982178,
+      "learning_rate": 0.00014689714779602423,
+      "loss": 0.5283,
+      "step": 1541
+    },
+    {
+      "epoch": 7.980645161290322,
+      "grad_norm": 1.1004135608673096,
+      "learning_rate": 0.00014686257562662058,
+      "loss": 0.5138,
+      "step": 1542
+    },
+    {
+      "epoch": 7.985806451612904,
+      "grad_norm": 1.0266633033752441,
+      "learning_rate": 0.00014682800345721693,
+      "loss": 0.5634,
+      "step": 1543
+    },
+    {
+      "epoch": 7.990967741935484,
+      "grad_norm": 1.1004083156585693,
+      "learning_rate": 0.0001467934312878133,
+      "loss": 0.5975,
+      "step": 1544
+    },
+    {
+      "epoch": 7.9961290322580645,
+      "grad_norm": 0.9985640048980713,
+      "learning_rate": 0.0001467588591184097,
+      "loss": 0.5202,
+      "step": 1545
+    },
+    {
+      "epoch": 8.00258064516129,
+      "grad_norm": 1.4235725402832031,
+      "learning_rate": 0.00014672428694900607,
+      "loss": 0.7292,
+      "step": 1546
+    },
+    {
+      "epoch": 8.007741935483871,
+      "grad_norm": 0.5662810206413269,
+      "learning_rate": 0.00014668971477960243,
+      "loss": 0.4063,
+      "step": 1547
+    },
+    {
+      "epoch": 8.012903225806452,
+      "grad_norm": 0.6014909744262695,
+      "learning_rate": 0.00014665514261019878,
+      "loss": 0.405,
+      "step": 1548
+    },
+    {
+      "epoch": 8.018064516129032,
+      "grad_norm": 0.7147842049598694,
+      "learning_rate": 0.00014662057044079516,
+      "loss": 0.4542,
+      "step": 1549
+    },
+    {
+      "epoch": 8.023225806451613,
+      "grad_norm": 0.7360666394233704,
+      "learning_rate": 0.00014658599827139154,
+      "loss": 0.4748,
+      "step": 1550
+    },
+    {
+      "epoch": 8.028387096774194,
+      "grad_norm": 0.8314833045005798,
+      "learning_rate": 0.00014655142610198792,
+      "loss": 0.4579,
+      "step": 1551
+    },
+    {
+      "epoch": 8.033548387096774,
+      "grad_norm": 0.9159157276153564,
+      "learning_rate": 0.00014651685393258428,
+      "loss": 0.458,
+      "step": 1552
+    },
+    {
+      "epoch": 8.038709677419355,
+      "grad_norm": 0.7821289896965027,
+      "learning_rate": 0.00014648228176318063,
+      "loss": 0.4537,
+      "step": 1553
+    },
+    {
+      "epoch": 8.043870967741935,
+      "grad_norm": 0.6763488054275513,
+      "learning_rate": 0.00014644770959377704,
+      "loss": 0.4403,
+      "step": 1554
+    },
+    {
+      "epoch": 8.049032258064516,
+      "grad_norm": 0.7865504622459412,
+      "learning_rate": 0.0001464131374243734,
+      "loss": 0.462,
+      "step": 1555
+    },
+    {
+      "epoch": 8.054193548387097,
+      "grad_norm": 0.7396550178527832,
+      "learning_rate": 0.00014637856525496977,
+      "loss": 0.4771,
+      "step": 1556
+    },
+    {
+      "epoch": 8.059354838709677,
+      "grad_norm": 0.7811912298202515,
+      "learning_rate": 0.00014634399308556612,
+      "loss": 0.4505,
+      "step": 1557
+    },
+    {
+      "epoch": 8.064516129032258,
+      "grad_norm": 0.8060530424118042,
+      "learning_rate": 0.00014630942091616248,
+      "loss": 0.4335,
+      "step": 1558
+    },
+    {
+      "epoch": 8.06967741935484,
+      "grad_norm": 0.6927765607833862,
+      "learning_rate": 0.00014627484874675889,
+      "loss": 0.4786,
+      "step": 1559
+    },
+    {
+      "epoch": 8.074838709677419,
+      "grad_norm": 0.6840382218360901,
+      "learning_rate": 0.00014624027657735524,
+      "loss": 0.4601,
+      "step": 1560
+    },
+    {
+      "epoch": 8.08,
+      "grad_norm": 0.7399626970291138,
+      "learning_rate": 0.0001462057044079516,
+      "loss": 0.4516,
+      "step": 1561
+    },
+    {
+      "epoch": 8.085161290322581,
+      "grad_norm": 0.7533694505691528,
+      "learning_rate": 0.00014617113223854797,
+      "loss": 0.4437,
+      "step": 1562
+    },
+    {
+      "epoch": 8.09032258064516,
+      "grad_norm": 0.7493161559104919,
+      "learning_rate": 0.00014613656006914433,
+      "loss": 0.4958,
+      "step": 1563
+    },
+    {
+      "epoch": 8.095483870967742,
+      "grad_norm": 0.8073078989982605,
+      "learning_rate": 0.00014610198789974073,
+      "loss": 0.4242,
+      "step": 1564
+    },
+    {
+      "epoch": 8.100645161290323,
+      "grad_norm": 0.7878143191337585,
+      "learning_rate": 0.0001460674157303371,
+      "loss": 0.4801,
+      "step": 1565
+    },
+    {
+      "epoch": 8.105806451612903,
+      "grad_norm": 0.7044801712036133,
+      "learning_rate": 0.00014603284356093344,
+      "loss": 0.4297,
+      "step": 1566
+    },
+    {
+      "epoch": 8.110967741935484,
+      "grad_norm": 0.7876698970794678,
+      "learning_rate": 0.00014599827139152982,
+      "loss": 0.463,
+      "step": 1567
+    },
+    {
+      "epoch": 8.116129032258064,
+      "grad_norm": 0.7920832633972168,
+      "learning_rate": 0.0001459636992221262,
+      "loss": 0.4739,
+      "step": 1568
+    },
+    {
+      "epoch": 8.121290322580645,
+      "grad_norm": 0.7470188736915588,
+      "learning_rate": 0.00014592912705272258,
+      "loss": 0.4903,
+      "step": 1569
+    },
+    {
+      "epoch": 8.126451612903226,
+      "grad_norm": 0.7946308851242065,
+      "learning_rate": 0.00014589455488331894,
+      "loss": 0.4874,
+      "step": 1570
+    },
+    {
+      "epoch": 8.131612903225806,
+      "grad_norm": 0.7578579783439636,
+      "learning_rate": 0.0001458599827139153,
+      "loss": 0.4594,
+      "step": 1571
+    },
+    {
+      "epoch": 8.136774193548387,
+      "grad_norm": 0.7009756565093994,
+      "learning_rate": 0.00014582541054451167,
+      "loss": 0.4669,
+      "step": 1572
+    },
+    {
+      "epoch": 8.141935483870968,
+      "grad_norm": 0.7506234645843506,
+      "learning_rate": 0.00014579083837510805,
+      "loss": 0.461,
+      "step": 1573
+    },
+    {
+      "epoch": 8.147096774193548,
+      "grad_norm": 0.8052283525466919,
+      "learning_rate": 0.0001457562662057044,
+      "loss": 0.4399,
+      "step": 1574
+    },
+    {
+      "epoch": 8.152258064516129,
+      "grad_norm": 0.7534711956977844,
+      "learning_rate": 0.00014572169403630078,
+      "loss": 0.4594,
+      "step": 1575
+    },
+    {
+      "epoch": 8.15741935483871,
+      "grad_norm": 0.8186956644058228,
+      "learning_rate": 0.00014568712186689714,
+      "loss": 0.4702,
+      "step": 1576
+    },
+    {
+      "epoch": 8.16258064516129,
+      "grad_norm": 0.7773800492286682,
+      "learning_rate": 0.00014565254969749355,
+      "loss": 0.4847,
+      "step": 1577
+    },
+    {
+      "epoch": 8.167741935483871,
+      "grad_norm": 0.7309033274650574,
+      "learning_rate": 0.0001456179775280899,
+      "loss": 0.4447,
+      "step": 1578
+    },
+    {
+      "epoch": 8.172903225806452,
+      "grad_norm": 0.6817725300788879,
+      "learning_rate": 0.00014558340535868625,
+      "loss": 0.4266,
+      "step": 1579
+    },
+    {
+      "epoch": 8.178064516129032,
+      "grad_norm": 0.8252871036529541,
+      "learning_rate": 0.00014554883318928263,
+      "loss": 0.4321,
+      "step": 1580
+    },
+    {
+      "epoch": 8.183225806451613,
+      "grad_norm": 0.790168285369873,
+      "learning_rate": 0.00014551426101987899,
+      "loss": 0.4721,
+      "step": 1581
+    },
+    {
+      "epoch": 8.188387096774193,
+      "grad_norm": 2.4513909816741943,
+      "learning_rate": 0.0001454796888504754,
+      "loss": 0.4618,
+      "step": 1582
+    },
+    {
+      "epoch": 8.193548387096774,
+      "grad_norm": 0.880634605884552,
+      "learning_rate": 0.00014544511668107175,
+      "loss": 0.4902,
+      "step": 1583
+    },
+    {
+      "epoch": 8.198709677419355,
+      "grad_norm": 0.7596932649612427,
+      "learning_rate": 0.0001454105445116681,
+      "loss": 0.3917,
+      "step": 1584
+    },
+    {
+      "epoch": 8.203870967741935,
+      "grad_norm": 0.8166360855102539,
+      "learning_rate": 0.00014537597234226448,
+      "loss": 0.4432,
+      "step": 1585
+    },
+    {
+      "epoch": 8.209032258064516,
+      "grad_norm": 0.8125788569450378,
+      "learning_rate": 0.00014534140017286086,
+      "loss": 0.4697,
+      "step": 1586
+    },
+    {
+      "epoch": 8.214193548387097,
+      "grad_norm": 0.802681565284729,
+      "learning_rate": 0.00014530682800345724,
+      "loss": 0.4689,
+      "step": 1587
+    },
+    {
+      "epoch": 8.219354838709677,
+      "grad_norm": 0.7260962128639221,
+      "learning_rate": 0.0001452722558340536,
+      "loss": 0.4776,
+      "step": 1588
+    },
+    {
+      "epoch": 8.224516129032258,
+      "grad_norm": 0.631947934627533,
+      "learning_rate": 0.00014523768366464995,
+      "loss": 0.4458,
+      "step": 1589
+    },
+    {
+      "epoch": 8.22967741935484,
+      "grad_norm": 0.7384147644042969,
+      "learning_rate": 0.00014520311149524633,
+      "loss": 0.4605,
+      "step": 1590
+    },
+    {
+      "epoch": 8.234838709677419,
+      "grad_norm": 0.7199890613555908,
+      "learning_rate": 0.0001451685393258427,
+      "loss": 0.4491,
+      "step": 1591
+    },
+    {
+      "epoch": 8.24,
+      "grad_norm": 0.7515553832054138,
+      "learning_rate": 0.00014513396715643906,
+      "loss": 0.4646,
+      "step": 1592
+    },
+    {
+      "epoch": 8.245161290322581,
+      "grad_norm": 0.7246177792549133,
+      "learning_rate": 0.00014509939498703544,
+      "loss": 0.4185,
+      "step": 1593
+    },
+    {
+      "epoch": 8.250322580645161,
+      "grad_norm": 0.7571049332618713,
+      "learning_rate": 0.0001450648228176318,
+      "loss": 0.4254,
+      "step": 1594
+    },
+    {
+      "epoch": 8.255483870967742,
+      "grad_norm": 0.7692784070968628,
+      "learning_rate": 0.0001450302506482282,
+      "loss": 0.415,
+      "step": 1595
+    },
+    {
+      "epoch": 8.260645161290322,
+      "grad_norm": 0.6954889297485352,
+      "learning_rate": 0.00014499567847882456,
+      "loss": 0.4583,
+      "step": 1596
+    },
+    {
+      "epoch": 8.265806451612903,
+      "grad_norm": 0.6973058581352234,
+      "learning_rate": 0.0001449611063094209,
+      "loss": 0.4907,
+      "step": 1597
+    },
+    {
+      "epoch": 8.270967741935484,
+      "grad_norm": 0.8663455247879028,
+      "learning_rate": 0.0001449265341400173,
+      "loss": 0.4809,
+      "step": 1598
+    },
+    {
+      "epoch": 8.276129032258064,
+      "grad_norm": 0.7677947282791138,
+      "learning_rate": 0.00014489196197061365,
+      "loss": 0.4771,
+      "step": 1599
+    },
+    {
+      "epoch": 8.281290322580645,
+      "grad_norm": 0.77374267578125,
+      "learning_rate": 0.00014485738980121005,
+      "loss": 0.4339,
+      "step": 1600
+    },
+    {
+      "epoch": 8.286451612903226,
+      "grad_norm": 0.7534428238868713,
+      "learning_rate": 0.0001448228176318064,
+      "loss": 0.4595,
+      "step": 1601
+    },
+    {
+      "epoch": 8.291612903225806,
+      "grad_norm": 0.7802786231040955,
+      "learning_rate": 0.00014478824546240276,
+      "loss": 0.4371,
+      "step": 1602
+    },
+    {
+      "epoch": 8.296774193548387,
+      "grad_norm": 0.7185215353965759,
+      "learning_rate": 0.00014475367329299914,
+      "loss": 0.4519,
+      "step": 1603
+    },
+    {
+      "epoch": 8.301935483870968,
+      "grad_norm": 0.7836042642593384,
+      "learning_rate": 0.0001447191011235955,
+      "loss": 0.4673,
+      "step": 1604
+    },
+    {
+      "epoch": 8.307096774193548,
+      "grad_norm": 0.7933247685432434,
+      "learning_rate": 0.0001446845289541919,
+      "loss": 0.4809,
+      "step": 1605
+    },
+    {
+      "epoch": 8.31225806451613,
+      "grad_norm": 0.7738546133041382,
+      "learning_rate": 0.00014464995678478826,
+      "loss": 0.3786,
+      "step": 1606
+    },
+    {
+      "epoch": 8.31741935483871,
+      "grad_norm": 0.7983633279800415,
+      "learning_rate": 0.0001446153846153846,
+      "loss": 0.4577,
+      "step": 1607
+    },
+    {
+      "epoch": 8.32258064516129,
+      "grad_norm": 0.8690028786659241,
+      "learning_rate": 0.000144580812445981,
+      "loss": 0.4503,
+      "step": 1608
+    },
+    {
+      "epoch": 8.327741935483871,
+      "grad_norm": 0.8182960152626038,
+      "learning_rate": 0.00014454624027657737,
+      "loss": 0.4602,
+      "step": 1609
+    },
+    {
+      "epoch": 8.33290322580645,
+      "grad_norm": 0.7107269763946533,
+      "learning_rate": 0.00014451166810717372,
+      "loss": 0.4422,
+      "step": 1610
+    },
+    {
+      "epoch": 8.338064516129032,
+      "grad_norm": 0.7924053072929382,
+      "learning_rate": 0.0001444770959377701,
+      "loss": 0.4639,
+      "step": 1611
+    },
+    {
+      "epoch": 8.343225806451613,
+      "grad_norm": 0.7608335614204407,
+      "learning_rate": 0.00014444252376836646,
+      "loss": 0.4805,
+      "step": 1612
+    },
+    {
+      "epoch": 8.348387096774193,
+      "grad_norm": 0.7586287260055542,
+      "learning_rate": 0.00014440795159896284,
+      "loss": 0.4453,
+      "step": 1613
+    },
+    {
+      "epoch": 8.353548387096774,
+      "grad_norm": 0.7487065196037292,
+      "learning_rate": 0.00014437337942955922,
+      "loss": 0.4705,
+      "step": 1614
+    },
+    {
+      "epoch": 8.358709677419355,
+      "grad_norm": 0.7204211950302124,
+      "learning_rate": 0.00014433880726015557,
+      "loss": 0.4495,
+      "step": 1615
+    },
+    {
+      "epoch": 8.363870967741935,
+      "grad_norm": 0.7663567066192627,
+      "learning_rate": 0.00014430423509075195,
+      "loss": 0.4594,
+      "step": 1616
+    },
+    {
+      "epoch": 8.369032258064516,
+      "grad_norm": 0.8133089542388916,
+      "learning_rate": 0.0001442696629213483,
+      "loss": 0.4722,
+      "step": 1617
+    },
+    {
+      "epoch": 8.374193548387098,
+      "grad_norm": 0.7820075154304504,
+      "learning_rate": 0.00014423509075194471,
+      "loss": 0.4603,
+      "step": 1618
+    },
+    {
+      "epoch": 8.379354838709677,
+      "grad_norm": 0.7656058669090271,
+      "learning_rate": 0.00014420051858254107,
+      "loss": 0.4478,
+      "step": 1619
+    },
+    {
+      "epoch": 8.384516129032258,
+      "grad_norm": 0.8208919763565063,
+      "learning_rate": 0.00014416594641313742,
+      "loss": 0.4275,
+      "step": 1620
+    },
+    {
+      "epoch": 8.38967741935484,
+      "grad_norm": 0.8575781583786011,
+      "learning_rate": 0.0001441313742437338,
+      "loss": 0.4868,
+      "step": 1621
+    },
+    {
+      "epoch": 8.394838709677419,
+      "grad_norm": 0.9061164259910583,
+      "learning_rate": 0.00014409680207433015,
+      "loss": 0.4585,
+      "step": 1622
+    },
+    {
+      "epoch": 8.4,
+      "grad_norm": 0.8663834929466248,
+      "learning_rate": 0.00014406222990492656,
+      "loss": 0.46,
+      "step": 1623
+    },
+    {
+      "epoch": 8.40516129032258,
+      "grad_norm": 0.7779414653778076,
+      "learning_rate": 0.00014402765773552292,
+      "loss": 0.4687,
+      "step": 1624
+    },
+    {
+      "epoch": 8.410322580645161,
+      "grad_norm": 0.8122915029525757,
+      "learning_rate": 0.00014399308556611927,
+      "loss": 0.4949,
+      "step": 1625
+    },
+    {
+      "epoch": 8.415483870967742,
+      "grad_norm": 0.7776375412940979,
+      "learning_rate": 0.00014395851339671565,
+      "loss": 0.4646,
+      "step": 1626
+    },
+    {
+      "epoch": 8.420645161290322,
+      "grad_norm": 0.7504333853721619,
+      "learning_rate": 0.00014392394122731203,
+      "loss": 0.474,
+      "step": 1627
+    },
+    {
+      "epoch": 8.425806451612903,
+      "grad_norm": 0.7237344980239868,
+      "learning_rate": 0.00014388936905790838,
+      "loss": 0.4349,
+      "step": 1628
+    },
+    {
+      "epoch": 8.430967741935484,
+      "grad_norm": 0.7637043595314026,
+      "learning_rate": 0.00014385479688850476,
+      "loss": 0.4661,
+      "step": 1629
+    },
+    {
+      "epoch": 8.436129032258064,
+      "grad_norm": 0.7965600490570068,
+      "learning_rate": 0.00014382022471910112,
+      "loss": 0.4952,
+      "step": 1630
+    },
+    {
+      "epoch": 8.441290322580645,
+      "grad_norm": 0.8336679339408875,
+      "learning_rate": 0.0001437856525496975,
+      "loss": 0.4773,
+      "step": 1631
+    },
+    {
+      "epoch": 8.446451612903227,
+      "grad_norm": 0.8159080147743225,
+      "learning_rate": 0.00014375108038029388,
+      "loss": 0.4264,
+      "step": 1632
+    },
+    {
+      "epoch": 8.451612903225806,
+      "grad_norm": 0.8754715323448181,
+      "learning_rate": 0.00014371650821089023,
+      "loss": 0.4739,
+      "step": 1633
+    },
+    {
+      "epoch": 8.456774193548387,
+      "grad_norm": 0.7183371186256409,
+      "learning_rate": 0.0001436819360414866,
+      "loss": 0.4405,
+      "step": 1634
+    },
+    {
+      "epoch": 8.461935483870967,
+      "grad_norm": 0.8026530146598816,
+      "learning_rate": 0.00014364736387208297,
+      "loss": 0.4887,
+      "step": 1635
+    },
+    {
+      "epoch": 8.467096774193548,
+      "grad_norm": 0.8732632398605347,
+      "learning_rate": 0.00014361279170267935,
+      "loss": 0.4974,
+      "step": 1636
+    },
+    {
+      "epoch": 8.47225806451613,
+      "grad_norm": 0.7318917512893677,
+      "learning_rate": 0.00014357821953327573,
+      "loss": 0.4609,
+      "step": 1637
+    },
+    {
+      "epoch": 8.477419354838709,
+      "grad_norm": 0.8117645978927612,
+      "learning_rate": 0.00014354364736387208,
+      "loss": 0.4859,
+      "step": 1638
+    },
+    {
+      "epoch": 8.48258064516129,
+      "grad_norm": 0.8418295383453369,
+      "learning_rate": 0.00014350907519446846,
+      "loss": 0.494,
+      "step": 1639
+    },
+    {
+      "epoch": 8.487741935483871,
+      "grad_norm": 0.8272765278816223,
+      "learning_rate": 0.00014347450302506481,
+      "loss": 0.4821,
+      "step": 1640
+    },
+    {
+      "epoch": 8.492903225806451,
+      "grad_norm": 0.8535323739051819,
+      "learning_rate": 0.00014343993085566122,
+      "loss": 0.4828,
+      "step": 1641
+    },
+    {
+      "epoch": 8.498064516129032,
+      "grad_norm": 0.7667284607887268,
+      "learning_rate": 0.00014340535868625758,
+      "loss": 0.4622,
+      "step": 1642
+    },
+    {
+      "epoch": 8.503225806451614,
+      "grad_norm": 0.7598912119865417,
+      "learning_rate": 0.00014337078651685393,
+      "loss": 0.4018,
+      "step": 1643
+    },
+    {
+      "epoch": 8.508387096774193,
+      "grad_norm": 0.8022792935371399,
+      "learning_rate": 0.0001433362143474503,
+      "loss": 0.4748,
+      "step": 1644
+    },
+    {
+      "epoch": 8.513548387096774,
+      "grad_norm": 0.7985738515853882,
+      "learning_rate": 0.00014330164217804666,
+      "loss": 0.4451,
+      "step": 1645
+    },
+    {
+      "epoch": 8.518709677419356,
+      "grad_norm": 0.7468912601470947,
+      "learning_rate": 0.00014326707000864304,
+      "loss": 0.4409,
+      "step": 1646
+    },
+    {
+      "epoch": 8.523870967741935,
+      "grad_norm": 0.8290742635726929,
+      "learning_rate": 0.00014323249783923942,
+      "loss": 0.482,
+      "step": 1647
+    },
+    {
+      "epoch": 8.529032258064516,
+      "grad_norm": 0.9028515815734863,
+      "learning_rate": 0.00014319792566983578,
+      "loss": 0.5028,
+      "step": 1648
+    },
+    {
+      "epoch": 8.534193548387098,
+      "grad_norm": 0.855601966381073,
+      "learning_rate": 0.00014316335350043216,
+      "loss": 0.464,
+      "step": 1649
+    },
+    {
+      "epoch": 8.539354838709677,
+      "grad_norm": 0.8120416402816772,
+      "learning_rate": 0.00014312878133102854,
+      "loss": 0.4751,
+      "step": 1650
+    },
+    {
+      "epoch": 8.544516129032258,
+      "grad_norm": 0.8115382790565491,
+      "learning_rate": 0.0001430942091616249,
+      "loss": 0.4917,
+      "step": 1651
+    },
+    {
+      "epoch": 8.549677419354838,
+      "grad_norm": 0.7855458855628967,
+      "learning_rate": 0.00014305963699222127,
+      "loss": 0.4217,
+      "step": 1652
+    },
+    {
+      "epoch": 8.55483870967742,
+      "grad_norm": 0.8383639454841614,
+      "learning_rate": 0.00014302506482281763,
+      "loss": 0.4629,
+      "step": 1653
+    },
+    {
+      "epoch": 8.56,
+      "grad_norm": 0.7514215111732483,
+      "learning_rate": 0.000142990492653414,
+      "loss": 0.4855,
+      "step": 1654
+    },
+    {
+      "epoch": 8.56516129032258,
+      "grad_norm": 0.8071656227111816,
+      "learning_rate": 0.0001429559204840104,
+      "loss": 0.4556,
+      "step": 1655
+    },
+    {
+      "epoch": 8.570322580645161,
+      "grad_norm": 0.774469256401062,
+      "learning_rate": 0.00014292134831460674,
+      "loss": 0.4627,
+      "step": 1656
+    },
+    {
+      "epoch": 8.575483870967743,
+      "grad_norm": 0.7632143497467041,
+      "learning_rate": 0.00014288677614520312,
+      "loss": 0.4586,
+      "step": 1657
+    },
+    {
+      "epoch": 8.580645161290322,
+      "grad_norm": 0.7560074329376221,
+      "learning_rate": 0.00014285220397579947,
+      "loss": 0.4589,
+      "step": 1658
+    },
+    {
+      "epoch": 8.585806451612903,
+      "grad_norm": 0.7769772410392761,
+      "learning_rate": 0.00014281763180639588,
+      "loss": 0.4367,
+      "step": 1659
+    },
+    {
+      "epoch": 8.590967741935485,
+      "grad_norm": 0.748127281665802,
+      "learning_rate": 0.00014278305963699224,
+      "loss": 0.458,
+      "step": 1660
+    },
+    {
+      "epoch": 8.596129032258064,
+      "grad_norm": 0.8227567076683044,
+      "learning_rate": 0.0001427484874675886,
+      "loss": 0.4504,
+      "step": 1661
+    },
+    {
+      "epoch": 8.601290322580645,
+      "grad_norm": 0.7979136109352112,
+      "learning_rate": 0.00014271391529818497,
+      "loss": 0.4487,
+      "step": 1662
+    },
+    {
+      "epoch": 8.606451612903225,
+      "grad_norm": 0.7464467883110046,
+      "learning_rate": 0.00014267934312878132,
+      "loss": 0.5171,
+      "step": 1663
+    },
+    {
+      "epoch": 8.611612903225806,
+      "grad_norm": 0.7955998778343201,
+      "learning_rate": 0.0001426447709593777,
+      "loss": 0.4825,
+      "step": 1664
+    },
+    {
+      "epoch": 8.616774193548387,
+      "grad_norm": 0.7490065693855286,
+      "learning_rate": 0.00014261019878997408,
+      "loss": 0.3808,
+      "step": 1665
+    },
+    {
+      "epoch": 8.621935483870967,
+      "grad_norm": 0.9489019513130188,
+      "learning_rate": 0.00014257562662057044,
+      "loss": 0.4428,
+      "step": 1666
+    },
+    {
+      "epoch": 8.627096774193548,
+      "grad_norm": 0.8622388243675232,
+      "learning_rate": 0.00014254105445116682,
+      "loss": 0.4896,
+      "step": 1667
+    },
+    {
+      "epoch": 8.63225806451613,
+      "grad_norm": 0.843321681022644,
+      "learning_rate": 0.0001425064822817632,
+      "loss": 0.4949,
+      "step": 1668
+    },
+    {
+      "epoch": 8.637419354838709,
+      "grad_norm": 0.855689525604248,
+      "learning_rate": 0.00014247191011235955,
+      "loss": 0.4384,
+      "step": 1669
+    },
+    {
+      "epoch": 8.64258064516129,
+      "grad_norm": 0.8887670636177063,
+      "learning_rate": 0.00014243733794295593,
+      "loss": 0.458,
+      "step": 1670
+    },
+    {
+      "epoch": 8.647741935483872,
+      "grad_norm": 0.785417377948761,
+      "learning_rate": 0.00014240276577355229,
+      "loss": 0.4851,
+      "step": 1671
+    },
+    {
+      "epoch": 8.652903225806451,
+      "grad_norm": 0.8447320461273193,
+      "learning_rate": 0.00014236819360414867,
+      "loss": 0.4948,
+      "step": 1672
+    },
+    {
+      "epoch": 8.658064516129032,
+      "grad_norm": 0.7898286581039429,
+      "learning_rate": 0.00014233362143474505,
+      "loss": 0.4614,
+      "step": 1673
+    },
+    {
+      "epoch": 8.663225806451614,
+      "grad_norm": 0.7747907042503357,
+      "learning_rate": 0.0001422990492653414,
+      "loss": 0.4213,
+      "step": 1674
+    },
+    {
+      "epoch": 8.668387096774193,
+      "grad_norm": 0.8469463586807251,
+      "learning_rate": 0.00014226447709593778,
+      "loss": 0.5099,
+      "step": 1675
+    },
+    {
+      "epoch": 8.673548387096774,
+      "grad_norm": 0.8160892724990845,
+      "learning_rate": 0.00014222990492653413,
+      "loss": 0.4621,
+      "step": 1676
+    },
+    {
+      "epoch": 8.678709677419356,
+      "grad_norm": 0.8187229037284851,
+      "learning_rate": 0.00014219533275713051,
+      "loss": 0.4915,
+      "step": 1677
+    },
+    {
+      "epoch": 8.683870967741935,
+      "grad_norm": 0.7599018216133118,
+      "learning_rate": 0.0001421607605877269,
+      "loss": 0.3988,
+      "step": 1678
+    },
+    {
+      "epoch": 8.689032258064517,
+      "grad_norm": 0.9345723986625671,
+      "learning_rate": 0.00014212618841832325,
+      "loss": 0.4797,
+      "step": 1679
+    },
+    {
+      "epoch": 8.694193548387096,
+      "grad_norm": 0.8985433578491211,
+      "learning_rate": 0.00014209161624891963,
+      "loss": 0.4893,
+      "step": 1680
+    },
+    {
+      "epoch": 8.699354838709677,
+      "grad_norm": 0.8412796258926392,
+      "learning_rate": 0.00014205704407951598,
+      "loss": 0.4704,
+      "step": 1681
+    },
+    {
+      "epoch": 8.704516129032259,
+      "grad_norm": 0.8268060088157654,
+      "learning_rate": 0.00014202247191011236,
+      "loss": 0.4944,
+      "step": 1682
+    },
+    {
+      "epoch": 8.709677419354838,
+      "grad_norm": 0.845278799533844,
+      "learning_rate": 0.00014198789974070874,
+      "loss": 0.4871,
+      "step": 1683
+    },
+    {
+      "epoch": 8.71483870967742,
+      "grad_norm": 0.8307921290397644,
+      "learning_rate": 0.0001419533275713051,
+      "loss": 0.483,
+      "step": 1684
+    },
+    {
+      "epoch": 8.72,
+      "grad_norm": 0.9138636589050293,
+      "learning_rate": 0.00014191875540190148,
+      "loss": 0.4738,
+      "step": 1685
+    },
+    {
+      "epoch": 8.72516129032258,
+      "grad_norm": 0.8554360866546631,
+      "learning_rate": 0.00014188418323249783,
+      "loss": 0.5152,
+      "step": 1686
+    },
+    {
+      "epoch": 8.730322580645161,
+      "grad_norm": 0.7474501132965088,
+      "learning_rate": 0.0001418496110630942,
+      "loss": 0.4502,
+      "step": 1687
+    },
+    {
+      "epoch": 8.735483870967743,
+      "grad_norm": 0.7886642217636108,
+      "learning_rate": 0.0001418150388936906,
+      "loss": 0.4416,
+      "step": 1688
+    },
+    {
+      "epoch": 8.740645161290322,
+      "grad_norm": 0.8453039526939392,
+      "learning_rate": 0.00014178046672428695,
+      "loss": 0.5022,
+      "step": 1689
+    },
+    {
+      "epoch": 8.745806451612903,
+      "grad_norm": 0.8735306262969971,
+      "learning_rate": 0.00014174589455488333,
+      "loss": 0.4904,
+      "step": 1690
+    },
+    {
+      "epoch": 8.750967741935483,
+      "grad_norm": 0.8776302933692932,
+      "learning_rate": 0.0001417113223854797,
+      "loss": 0.4559,
+      "step": 1691
+    },
+    {
+      "epoch": 8.756129032258064,
+      "grad_norm": 0.8258122801780701,
+      "learning_rate": 0.00014167675021607606,
+      "loss": 0.4533,
+      "step": 1692
+    },
+    {
+      "epoch": 8.761290322580646,
+      "grad_norm": 0.8453822135925293,
+      "learning_rate": 0.00014164217804667244,
+      "loss": 0.4453,
+      "step": 1693
+    },
+    {
+      "epoch": 8.766451612903225,
+      "grad_norm": 0.7612801790237427,
+      "learning_rate": 0.0001416076058772688,
+      "loss": 0.4625,
+      "step": 1694
+    },
+    {
+      "epoch": 8.771612903225806,
+      "grad_norm": 0.8587717413902283,
+      "learning_rate": 0.00014157303370786517,
+      "loss": 0.5011,
+      "step": 1695
+    },
+    {
+      "epoch": 8.776774193548388,
+      "grad_norm": 0.8040095567703247,
+      "learning_rate": 0.00014153846153846156,
+      "loss": 0.5097,
+      "step": 1696
+    },
+    {
+      "epoch": 8.781935483870967,
+      "grad_norm": 0.782917320728302,
+      "learning_rate": 0.0001415038893690579,
+      "loss": 0.4299,
+      "step": 1697
+    },
+    {
+      "epoch": 8.787096774193548,
+      "grad_norm": 0.8529263138771057,
+      "learning_rate": 0.0001414693171996543,
+      "loss": 0.4862,
+      "step": 1698
+    },
+    {
+      "epoch": 8.79225806451613,
+      "grad_norm": 0.8536248207092285,
+      "learning_rate": 0.00014143474503025064,
+      "loss": 0.5312,
+      "step": 1699
+    },
+    {
+      "epoch": 8.79741935483871,
+      "grad_norm": 0.8290024399757385,
+      "learning_rate": 0.00014140017286084702,
+      "loss": 0.5027,
+      "step": 1700
+    },
+    {
+      "epoch": 8.80258064516129,
+      "grad_norm": 0.8671407699584961,
+      "learning_rate": 0.0001413656006914434,
+      "loss": 0.4151,
+      "step": 1701
+    },
+    {
+      "epoch": 8.807741935483872,
+      "grad_norm": 0.8628122210502625,
+      "learning_rate": 0.00014133102852203976,
+      "loss": 0.4292,
+      "step": 1702
+    },
+    {
+      "epoch": 8.812903225806451,
+      "grad_norm": 0.8464361429214478,
+      "learning_rate": 0.00014129645635263614,
+      "loss": 0.4845,
+      "step": 1703
+    },
+    {
+      "epoch": 8.818064516129033,
+      "grad_norm": 0.8388773202896118,
+      "learning_rate": 0.0001412618841832325,
+      "loss": 0.4747,
+      "step": 1704
+    },
+    {
+      "epoch": 8.823225806451614,
+      "grad_norm": 0.8786773681640625,
+      "learning_rate": 0.00014122731201382887,
+      "loss": 0.4749,
+      "step": 1705
+    },
+    {
+      "epoch": 8.828387096774193,
+      "grad_norm": 0.841955840587616,
+      "learning_rate": 0.00014119273984442525,
+      "loss": 0.496,
+      "step": 1706
+    },
+    {
+      "epoch": 8.833548387096775,
+      "grad_norm": 0.7751501202583313,
+      "learning_rate": 0.0001411581676750216,
+      "loss": 0.4726,
+      "step": 1707
+    },
+    {
+      "epoch": 8.838709677419354,
+      "grad_norm": 0.811134934425354,
+      "learning_rate": 0.00014112359550561799,
+      "loss": 0.4581,
+      "step": 1708
+    },
+    {
+      "epoch": 8.843870967741935,
+      "grad_norm": 0.8542317748069763,
+      "learning_rate": 0.00014108902333621434,
+      "loss": 0.4849,
+      "step": 1709
+    },
+    {
+      "epoch": 8.849032258064517,
+      "grad_norm": 0.8741058111190796,
+      "learning_rate": 0.00014105445116681072,
+      "loss": 0.4875,
+      "step": 1710
+    },
+    {
+      "epoch": 8.854193548387096,
+      "grad_norm": 0.8611905574798584,
+      "learning_rate": 0.0001410198789974071,
+      "loss": 0.4554,
+      "step": 1711
+    },
+    {
+      "epoch": 8.859354838709677,
+      "grad_norm": 2.9399664402008057,
+      "learning_rate": 0.00014098530682800345,
+      "loss": 0.4888,
+      "step": 1712
+    },
+    {
+      "epoch": 8.864516129032259,
+      "grad_norm": 0.8815257549285889,
+      "learning_rate": 0.00014095073465859983,
+      "loss": 0.4958,
+      "step": 1713
+    },
+    {
+      "epoch": 8.869677419354838,
+      "grad_norm": 0.8916751742362976,
+      "learning_rate": 0.00014091616248919622,
+      "loss": 0.4541,
+      "step": 1714
+    },
+    {
+      "epoch": 8.87483870967742,
+      "grad_norm": 0.8360041379928589,
+      "learning_rate": 0.00014088159031979257,
+      "loss": 0.5174,
+      "step": 1715
+    },
+    {
+      "epoch": 8.88,
+      "grad_norm": 0.8667903542518616,
+      "learning_rate": 0.00014084701815038895,
+      "loss": 0.4307,
+      "step": 1716
+    },
+    {
+      "epoch": 8.88516129032258,
+      "grad_norm": 0.796523928642273,
+      "learning_rate": 0.0001408124459809853,
+      "loss": 0.4826,
+      "step": 1717
+    },
+    {
+      "epoch": 8.890322580645162,
+      "grad_norm": 0.8221596479415894,
+      "learning_rate": 0.00014077787381158168,
+      "loss": 0.4288,
+      "step": 1718
+    },
+    {
+      "epoch": 8.895483870967741,
+      "grad_norm": 0.8943276405334473,
+      "learning_rate": 0.00014074330164217806,
+      "loss": 0.503,
+      "step": 1719
+    },
+    {
+      "epoch": 8.900645161290322,
+      "grad_norm": 0.9204633235931396,
+      "learning_rate": 0.00014070872947277442,
+      "loss": 0.4997,
+      "step": 1720
+    },
+    {
+      "epoch": 8.905806451612904,
+      "grad_norm": 0.8767181038856506,
+      "learning_rate": 0.0001406741573033708,
+      "loss": 0.5206,
+      "step": 1721
+    },
+    {
+      "epoch": 8.910967741935483,
+      "grad_norm": 0.9086239337921143,
+      "learning_rate": 0.00014063958513396715,
+      "loss": 0.4794,
+      "step": 1722
+    },
+    {
+      "epoch": 8.916129032258064,
+      "grad_norm": 0.8117924332618713,
+      "learning_rate": 0.00014060501296456353,
+      "loss": 0.4208,
+      "step": 1723
+    },
+    {
+      "epoch": 8.921290322580646,
+      "grad_norm": 4.038970947265625,
+      "learning_rate": 0.0001405704407951599,
+      "loss": 0.4674,
+      "step": 1724
+    },
+    {
+      "epoch": 8.926451612903225,
+      "grad_norm": 0.771162748336792,
+      "learning_rate": 0.00014053586862575627,
+      "loss": 0.434,
+      "step": 1725
+    },
+    {
+      "epoch": 8.931612903225806,
+      "grad_norm": 0.9108381867408752,
+      "learning_rate": 0.00014050129645635265,
+      "loss": 0.4519,
+      "step": 1726
+    },
+    {
+      "epoch": 8.936774193548388,
+      "grad_norm": 0.8793790936470032,
+      "learning_rate": 0.000140466724286949,
+      "loss": 0.5258,
+      "step": 1727
+    },
+    {
+      "epoch": 8.941935483870967,
+      "grad_norm": 0.9124688506126404,
+      "learning_rate": 0.00014043215211754538,
+      "loss": 0.4916,
+      "step": 1728
+    },
+    {
+      "epoch": 8.947096774193549,
+      "grad_norm": 0.9408337473869324,
+      "learning_rate": 0.00014039757994814176,
+      "loss": 0.4671,
+      "step": 1729
+    },
+    {
+      "epoch": 8.95225806451613,
+      "grad_norm": 0.8493264317512512,
+      "learning_rate": 0.00014036300777873811,
+      "loss": 0.5023,
+      "step": 1730
+    },
+    {
+      "epoch": 8.95741935483871,
+      "grad_norm": 0.8133382201194763,
+      "learning_rate": 0.0001403284356093345,
+      "loss": 0.4626,
+      "step": 1731
+    },
+    {
+      "epoch": 8.96258064516129,
+      "grad_norm": 0.8677392601966858,
+      "learning_rate": 0.00014029386343993088,
+      "loss": 0.4832,
+      "step": 1732
+    },
+    {
+      "epoch": 8.967741935483872,
+      "grad_norm": 0.8499205112457275,
+      "learning_rate": 0.00014025929127052723,
+      "loss": 0.4496,
+      "step": 1733
+    },
+    {
+      "epoch": 8.972903225806451,
+      "grad_norm": 0.7974291443824768,
+      "learning_rate": 0.0001402247191011236,
+      "loss": 0.4667,
+      "step": 1734
+    },
+    {
+      "epoch": 8.978064516129033,
+      "grad_norm": 0.8573561906814575,
+      "learning_rate": 0.00014019014693171996,
+      "loss": 0.426,
+      "step": 1735
+    },
+    {
+      "epoch": 8.983225806451612,
+      "grad_norm": 0.8910071849822998,
+      "learning_rate": 0.00014015557476231634,
+      "loss": 0.4968,
+      "step": 1736
+    },
+    {
+      "epoch": 8.988387096774193,
+      "grad_norm": 0.8981788754463196,
+      "learning_rate": 0.00014012100259291272,
+      "loss": 0.4257,
+      "step": 1737
+    },
+    {
+      "epoch": 8.993548387096775,
+      "grad_norm": 0.9378400444984436,
+      "learning_rate": 0.00014008643042350908,
+      "loss": 0.4935,
+      "step": 1738
+    },
+    {
+      "epoch": 8.998709677419354,
+      "grad_norm": 1.3971208333969116,
+      "learning_rate": 0.00014005185825410546,
+      "loss": 0.7725,
+      "step": 1739
+    },
+    {
+      "epoch": 9.005161290322581,
+      "grad_norm": 0.5530678629875183,
+      "learning_rate": 0.0001400172860847018,
+      "loss": 0.4522,
+      "step": 1740
+    },
+    {
+      "epoch": 9.005161290322581,
+      "eval_loss": 2.1547720432281494,
+      "eval_runtime": 21.3436,
+      "eval_samples_per_second": 3.701,
+      "eval_steps_per_second": 0.469,
+      "step": 1740
+    },
+    {
+      "epoch": 9.01032258064516,
+      "grad_norm": 0.5304129719734192,
+      "learning_rate": 0.0001399827139152982,
+      "loss": 0.4162,
+      "step": 1741
+    },
+    {
+      "epoch": 9.015483870967742,
+      "grad_norm": 0.4744911789894104,
+      "learning_rate": 0.00013994814174589457,
+      "loss": 0.4163,
+      "step": 1742
+    },
+    {
+      "epoch": 9.020645161290323,
+      "grad_norm": 0.6896321177482605,
+      "learning_rate": 0.00013991356957649093,
+      "loss": 0.4296,
+      "step": 1743
+    },
+    {
+      "epoch": 9.025806451612903,
+      "grad_norm": 0.6244247555732727,
+      "learning_rate": 0.0001398789974070873,
+      "loss": 0.4416,
+      "step": 1744
+    },
+    {
+      "epoch": 9.030967741935484,
+      "grad_norm": 0.715678870677948,
+      "learning_rate": 0.00013984442523768366,
+      "loss": 0.415,
+      "step": 1745
+    },
+    {
+      "epoch": 9.036129032258064,
+      "grad_norm": 0.7860274314880371,
+      "learning_rate": 0.00013980985306828004,
+      "loss": 0.4426,
+      "step": 1746
+    },
+    {
+      "epoch": 9.041290322580645,
+      "grad_norm": 0.6848698854446411,
+      "learning_rate": 0.00013977528089887642,
+      "loss": 0.4184,
+      "step": 1747
+    },
+    {
+      "epoch": 9.046451612903226,
+      "grad_norm": 0.6059648394584656,
+      "learning_rate": 0.00013974070872947277,
+      "loss": 0.4203,
+      "step": 1748
+    },
+    {
+      "epoch": 9.051612903225806,
+      "grad_norm": 0.819089412689209,
+      "learning_rate": 0.00013970613656006915,
+      "loss": 0.4414,
+      "step": 1749
+    },
+    {
+      "epoch": 9.056774193548387,
+      "grad_norm": 0.6701709628105164,
+      "learning_rate": 0.0001396715643906655,
+      "loss": 0.4516,
+      "step": 1750
+    },
+    {
+      "epoch": 9.061935483870968,
+      "grad_norm": 0.6554612517356873,
+      "learning_rate": 0.0001396369922212619,
+      "loss": 0.4237,
+      "step": 1751
+    },
+    {
+      "epoch": 9.067096774193548,
+      "grad_norm": 0.7074912190437317,
+      "learning_rate": 0.00013960242005185827,
+      "loss": 0.4363,
+      "step": 1752
+    },
+    {
+      "epoch": 9.072258064516129,
+      "grad_norm": 0.7133002281188965,
+      "learning_rate": 0.00013956784788245462,
+      "loss": 0.4569,
+      "step": 1753
+    },
+    {
+      "epoch": 9.07741935483871,
+      "grad_norm": 0.6331765651702881,
+      "learning_rate": 0.000139533275713051,
+      "loss": 0.4753,
+      "step": 1754
+    },
+    {
+      "epoch": 9.08258064516129,
+      "grad_norm": 0.6207808256149292,
+      "learning_rate": 0.00013949870354364738,
+      "loss": 0.4215,
+      "step": 1755
+    },
+    {
+      "epoch": 9.087741935483871,
+      "grad_norm": 0.5306448936462402,
+      "learning_rate": 0.00013946413137424374,
+      "loss": 0.4241,
+      "step": 1756
+    },
+    {
+      "epoch": 9.092903225806452,
+      "grad_norm": 0.75066077709198,
+      "learning_rate": 0.00013942955920484012,
+      "loss": 0.465,
+      "step": 1757
+    },
+    {
+      "epoch": 9.098064516129032,
+      "grad_norm": 0.6436359882354736,
+      "learning_rate": 0.00013939498703543647,
+      "loss": 0.3674,
+      "step": 1758
+    },
+    {
+      "epoch": 9.103225806451613,
+      "grad_norm": 0.6517727971076965,
+      "learning_rate": 0.00013936041486603285,
+      "loss": 0.4376,
+      "step": 1759
+    },
+    {
+      "epoch": 9.108387096774193,
+      "grad_norm": 0.6894286274909973,
+      "learning_rate": 0.00013932584269662923,
+      "loss": 0.4053,
+      "step": 1760
+    },
+    {
+      "epoch": 9.113548387096774,
+      "grad_norm": 0.7008318305015564,
+      "learning_rate": 0.00013929127052722559,
+      "loss": 0.4539,
+      "step": 1761
+    },
+    {
+      "epoch": 9.118709677419355,
+      "grad_norm": 0.7057509422302246,
+      "learning_rate": 0.00013925669835782197,
+      "loss": 0.4429,
+      "step": 1762
+    },
+    {
+      "epoch": 9.123870967741935,
+      "grad_norm": 0.7184132933616638,
+      "learning_rate": 0.00013922212618841832,
+      "loss": 0.4576,
+      "step": 1763
+    },
+    {
+      "epoch": 9.129032258064516,
+      "grad_norm": 0.676345705986023,
+      "learning_rate": 0.0001391875540190147,
+      "loss": 0.4285,
+      "step": 1764
+    },
+    {
+      "epoch": 9.134193548387097,
+      "grad_norm": 0.6346005201339722,
+      "learning_rate": 0.00013915298184961108,
+      "loss": 0.4651,
+      "step": 1765
+    },
+    {
+      "epoch": 9.139354838709677,
+      "grad_norm": 0.7055741548538208,
+      "learning_rate": 0.00013911840968020743,
+      "loss": 0.4078,
+      "step": 1766
+    },
+    {
+      "epoch": 9.144516129032258,
+      "grad_norm": 0.5901321768760681,
+      "learning_rate": 0.00013908383751080381,
+      "loss": 0.4219,
+      "step": 1767
+    },
+    {
+      "epoch": 9.14967741935484,
+      "grad_norm": 0.6489945650100708,
+      "learning_rate": 0.00013904926534140017,
+      "loss": 0.4194,
+      "step": 1768
+    },
+    {
+      "epoch": 9.154838709677419,
+      "grad_norm": 0.6097710132598877,
+      "learning_rate": 0.00013901469317199655,
+      "loss": 0.4097,
+      "step": 1769
+    },
+    {
+      "epoch": 9.16,
+      "grad_norm": 0.726087212562561,
+      "learning_rate": 0.00013898012100259293,
+      "loss": 0.426,
+      "step": 1770
+    },
+    {
+      "epoch": 9.165161290322581,
+      "grad_norm": 0.693672239780426,
+      "learning_rate": 0.00013894554883318928,
+      "loss": 0.4487,
+      "step": 1771
+    },
+    {
+      "epoch": 9.170322580645161,
+      "grad_norm": 0.6838711500167847,
+      "learning_rate": 0.00013891097666378566,
+      "loss": 0.4181,
+      "step": 1772
+    },
+    {
+      "epoch": 9.175483870967742,
+      "grad_norm": 0.7145890593528748,
+      "learning_rate": 0.00013887640449438204,
+      "loss": 0.454,
+      "step": 1773
+    },
+    {
+      "epoch": 9.180645161290322,
+      "grad_norm": 0.7231656908988953,
+      "learning_rate": 0.0001388418323249784,
+      "loss": 0.4431,
+      "step": 1774
+    },
+    {
+      "epoch": 9.185806451612903,
+      "grad_norm": 0.6733592748641968,
+      "learning_rate": 0.00013880726015557478,
+      "loss": 0.3934,
+      "step": 1775
+    },
+    {
+      "epoch": 9.190967741935484,
+      "grad_norm": 0.6744616627693176,
+      "learning_rate": 0.00013877268798617113,
+      "loss": 0.4063,
+      "step": 1776
+    },
+    {
+      "epoch": 9.196129032258064,
+      "grad_norm": 0.7792450189590454,
+      "learning_rate": 0.0001387381158167675,
+      "loss": 0.4187,
+      "step": 1777
+    },
+    {
+      "epoch": 9.201290322580645,
+      "grad_norm": 0.7441288232803345,
+      "learning_rate": 0.0001387035436473639,
+      "loss": 0.4369,
+      "step": 1778
+    },
+    {
+      "epoch": 9.206451612903226,
+      "grad_norm": 0.6719307899475098,
+      "learning_rate": 0.00013866897147796025,
+      "loss": 0.3679,
+      "step": 1779
+    },
+    {
+      "epoch": 9.211612903225806,
+      "grad_norm": 0.6938707232475281,
+      "learning_rate": 0.00013863439930855663,
+      "loss": 0.4484,
+      "step": 1780
+    },
+    {
+      "epoch": 9.216774193548387,
+      "grad_norm": 0.7516252994537354,
+      "learning_rate": 0.00013859982713915298,
+      "loss": 0.4155,
+      "step": 1781
+    },
+    {
+      "epoch": 9.221935483870968,
+      "grad_norm": 0.6943853497505188,
+      "learning_rate": 0.00013856525496974936,
+      "loss": 0.4628,
+      "step": 1782
+    },
+    {
+      "epoch": 9.227096774193548,
+      "grad_norm": 0.6411360502243042,
+      "learning_rate": 0.00013853068280034574,
+      "loss": 0.4462,
+      "step": 1783
+    },
+    {
+      "epoch": 9.23225806451613,
+      "grad_norm": 0.6068885922431946,
+      "learning_rate": 0.0001384961106309421,
+      "loss": 0.4513,
+      "step": 1784
+    },
+    {
+      "epoch": 9.23741935483871,
+      "grad_norm": 0.6016716957092285,
+      "learning_rate": 0.00013846153846153847,
+      "loss": 0.4482,
+      "step": 1785
+    },
+    {
+      "epoch": 9.24258064516129,
+      "grad_norm": 0.7177734375,
+      "learning_rate": 0.00013842696629213483,
+      "loss": 0.455,
+      "step": 1786
+    },
+    {
+      "epoch": 9.247741935483871,
+      "grad_norm": 0.688875675201416,
+      "learning_rate": 0.0001383923941227312,
+      "loss": 0.4466,
+      "step": 1787
+    },
+    {
+      "epoch": 9.252903225806453,
+      "grad_norm": 0.6392518877983093,
+      "learning_rate": 0.0001383578219533276,
+      "loss": 0.408,
+      "step": 1788
+    },
+    {
+      "epoch": 9.258064516129032,
+      "grad_norm": 0.6670936942100525,
+      "learning_rate": 0.00013832324978392394,
+      "loss": 0.459,
+      "step": 1789
+    },
+    {
+      "epoch": 9.263225806451613,
+      "grad_norm": 0.7178167104721069,
+      "learning_rate": 0.00013828867761452032,
+      "loss": 0.4784,
+      "step": 1790
+    },
+    {
+      "epoch": 9.268387096774193,
+      "grad_norm": 0.6738005876541138,
+      "learning_rate": 0.00013825410544511668,
+      "loss": 0.432,
+      "step": 1791
+    },
+    {
+      "epoch": 9.273548387096774,
+      "grad_norm": 0.6891224384307861,
+      "learning_rate": 0.00013821953327571306,
+      "loss": 0.4366,
+      "step": 1792
+    },
+    {
+      "epoch": 9.278709677419355,
+      "grad_norm": 0.5904545783996582,
+      "learning_rate": 0.00013818496110630944,
+      "loss": 0.3814,
+      "step": 1793
+    },
+    {
+      "epoch": 9.283870967741935,
+      "grad_norm": 0.7322163581848145,
+      "learning_rate": 0.0001381503889369058,
+      "loss": 0.4217,
+      "step": 1794
+    },
+    {
+      "epoch": 9.289032258064516,
+      "grad_norm": 0.7273543477058411,
+      "learning_rate": 0.00013811581676750217,
+      "loss": 0.4549,
+      "step": 1795
+    },
+    {
+      "epoch": 9.294193548387097,
+      "grad_norm": 0.6212940216064453,
+      "learning_rate": 0.00013808124459809855,
+      "loss": 0.3936,
+      "step": 1796
+    },
+    {
+      "epoch": 9.299354838709677,
+      "grad_norm": 0.7131409645080566,
+      "learning_rate": 0.0001380466724286949,
+      "loss": 0.4381,
+      "step": 1797
+    },
+    {
+      "epoch": 9.304516129032258,
+      "grad_norm": 0.8190418481826782,
+      "learning_rate": 0.00013801210025929129,
+      "loss": 0.4752,
+      "step": 1798
+    },
+    {
+      "epoch": 9.30967741935484,
+      "grad_norm": 0.669376015663147,
+      "learning_rate": 0.00013797752808988764,
+      "loss": 0.4426,
+      "step": 1799
+    },
+    {
+      "epoch": 9.314838709677419,
+      "grad_norm": 0.6486749053001404,
+      "learning_rate": 0.00013794295592048402,
+      "loss": 0.3718,
+      "step": 1800
+    },
+    {
+      "epoch": 9.32,
+      "grad_norm": 0.6617907881736755,
+      "learning_rate": 0.0001379083837510804,
+      "loss": 0.4829,
+      "step": 1801
+    },
+    {
+      "epoch": 9.32516129032258,
+      "grad_norm": 0.5981864333152771,
+      "learning_rate": 0.00013787381158167675,
+      "loss": 0.4432,
+      "step": 1802
+    },
+    {
+      "epoch": 9.330322580645161,
+      "grad_norm": 0.7223187685012817,
+      "learning_rate": 0.00013783923941227313,
+      "loss": 0.4128,
+      "step": 1803
+    },
+    {
+      "epoch": 9.335483870967742,
+      "grad_norm": 0.8082818984985352,
+      "learning_rate": 0.0001378046672428695,
+      "loss": 0.4403,
+      "step": 1804
+    },
+    {
+      "epoch": 9.340645161290322,
+      "grad_norm": 0.6313130855560303,
+      "learning_rate": 0.00013777009507346587,
+      "loss": 0.4678,
+      "step": 1805
+    },
+    {
+      "epoch": 9.345806451612903,
+      "grad_norm": 0.7852686047554016,
+      "learning_rate": 0.00013773552290406225,
+      "loss": 0.4589,
+      "step": 1806
+    },
+    {
+      "epoch": 9.350967741935484,
+      "grad_norm": 0.7433779239654541,
+      "learning_rate": 0.0001377009507346586,
+      "loss": 0.4872,
+      "step": 1807
+    },
+    {
+      "epoch": 9.356129032258064,
+      "grad_norm": 0.6377175450325012,
+      "learning_rate": 0.00013766637856525498,
+      "loss": 0.452,
+      "step": 1808
+    },
+    {
+      "epoch": 9.361290322580645,
+      "grad_norm": 0.7091626524925232,
+      "learning_rate": 0.00013763180639585134,
+      "loss": 0.4059,
+      "step": 1809
+    },
+    {
+      "epoch": 9.366451612903226,
+      "grad_norm": 0.6992821097373962,
+      "learning_rate": 0.00013759723422644772,
+      "loss": 0.435,
+      "step": 1810
+    },
+    {
+      "epoch": 9.371612903225806,
+      "grad_norm": 0.7191858291625977,
+      "learning_rate": 0.0001375626620570441,
+      "loss": 0.4058,
+      "step": 1811
+    },
+    {
+      "epoch": 9.376774193548387,
+      "grad_norm": 0.7320436239242554,
+      "learning_rate": 0.00013752808988764045,
+      "loss": 0.4611,
+      "step": 1812
+    },
+    {
+      "epoch": 9.381935483870969,
+      "grad_norm": 0.715164303779602,
+      "learning_rate": 0.00013749351771823683,
+      "loss": 0.4634,
+      "step": 1813
+    },
+    {
+      "epoch": 9.387096774193548,
+      "grad_norm": 0.6994943022727966,
+      "learning_rate": 0.0001374589455488332,
+      "loss": 0.4517,
+      "step": 1814
+    },
+    {
+      "epoch": 9.39225806451613,
+      "grad_norm": 0.7399652004241943,
+      "learning_rate": 0.00013742437337942957,
+      "loss": 0.3886,
+      "step": 1815
+    },
+    {
+      "epoch": 9.397419354838709,
+      "grad_norm": 0.7086983323097229,
+      "learning_rate": 0.00013738980121002595,
+      "loss": 0.4357,
+      "step": 1816
+    },
+    {
+      "epoch": 9.40258064516129,
+      "grad_norm": 0.6792604923248291,
+      "learning_rate": 0.0001373552290406223,
+      "loss": 0.408,
+      "step": 1817
+    },
+    {
+      "epoch": 9.407741935483871,
+      "grad_norm": 0.7420954704284668,
+      "learning_rate": 0.00013732065687121868,
+      "loss": 0.4441,
+      "step": 1818
+    },
+    {
+      "epoch": 9.412903225806451,
+      "grad_norm": 0.714895486831665,
+      "learning_rate": 0.00013728608470181506,
+      "loss": 0.4353,
+      "step": 1819
+    },
+    {
+      "epoch": 9.418064516129032,
+      "grad_norm": 0.6686902046203613,
+      "learning_rate": 0.00013725151253241141,
+      "loss": 0.4499,
+      "step": 1820
+    },
+    {
+      "epoch": 9.423225806451613,
+      "grad_norm": 0.65114426612854,
+      "learning_rate": 0.0001372169403630078,
+      "loss": 0.4505,
+      "step": 1821
+    },
+    {
+      "epoch": 9.428387096774193,
+      "grad_norm": 0.6974101066589355,
+      "learning_rate": 0.00013718236819360415,
+      "loss": 0.4549,
+      "step": 1822
+    },
+    {
+      "epoch": 9.433548387096774,
+      "grad_norm": 0.6582891941070557,
+      "learning_rate": 0.0001371477960242005,
+      "loss": 0.4457,
+      "step": 1823
+    },
+    {
+      "epoch": 9.438709677419356,
+      "grad_norm": 0.7576209306716919,
+      "learning_rate": 0.0001371132238547969,
+      "loss": 0.4329,
+      "step": 1824
+    },
+    {
+      "epoch": 9.443870967741935,
+      "grad_norm": 0.613598644733429,
+      "learning_rate": 0.00013707865168539326,
+      "loss": 0.3851,
+      "step": 1825
+    },
+    {
+      "epoch": 9.449032258064516,
+      "grad_norm": 0.7294527888298035,
+      "learning_rate": 0.00013704407951598964,
+      "loss": 0.4812,
+      "step": 1826
+    },
+    {
+      "epoch": 9.454193548387098,
+      "grad_norm": 0.6940247416496277,
+      "learning_rate": 0.000137009507346586,
+      "loss": 0.4228,
+      "step": 1827
+    },
+    {
+      "epoch": 9.459354838709677,
+      "grad_norm": 0.6420016884803772,
+      "learning_rate": 0.00013697493517718238,
+      "loss": 0.4175,
+      "step": 1828
+    },
+    {
+      "epoch": 9.464516129032258,
+      "grad_norm": 0.6380759477615356,
+      "learning_rate": 0.00013694036300777876,
+      "loss": 0.4772,
+      "step": 1829
+    },
+    {
+      "epoch": 9.469677419354838,
+      "grad_norm": 0.7091923952102661,
+      "learning_rate": 0.0001369057908383751,
+      "loss": 0.4612,
+      "step": 1830
+    },
+    {
+      "epoch": 9.47483870967742,
+      "grad_norm": 0.742102861404419,
+      "learning_rate": 0.0001368712186689715,
+      "loss": 0.4715,
+      "step": 1831
+    },
+    {
+      "epoch": 9.48,
+      "grad_norm": 0.6665038466453552,
+      "learning_rate": 0.00013683664649956784,
+      "loss": 0.419,
+      "step": 1832
+    },
+    {
+      "epoch": 9.48516129032258,
+      "grad_norm": 0.62575763463974,
+      "learning_rate": 0.00013680207433016423,
+      "loss": 0.3962,
+      "step": 1833
+    },
+    {
+      "epoch": 9.490322580645161,
+      "grad_norm": 0.7449005246162415,
+      "learning_rate": 0.0001367675021607606,
+      "loss": 0.436,
+      "step": 1834
+    },
+    {
+      "epoch": 9.495483870967742,
+      "grad_norm": 0.7571942806243896,
+      "learning_rate": 0.00013673292999135696,
+      "loss": 0.4189,
+      "step": 1835
+    },
+    {
+      "epoch": 9.500645161290322,
+      "grad_norm": 0.7013412117958069,
+      "learning_rate": 0.00013669835782195334,
+      "loss": 0.4433,
+      "step": 1836
+    },
+    {
+      "epoch": 9.505806451612903,
+      "grad_norm": 0.7367055416107178,
+      "learning_rate": 0.00013666378565254972,
+      "loss": 0.4121,
+      "step": 1837
+    },
+    {
+      "epoch": 9.510967741935485,
+      "grad_norm": 1.343169927597046,
+      "learning_rate": 0.00013662921348314607,
+      "loss": 0.4511,
+      "step": 1838
+    },
+    {
+      "epoch": 9.516129032258064,
+      "grad_norm": 0.7562336921691895,
+      "learning_rate": 0.00013659464131374245,
+      "loss": 0.4401,
+      "step": 1839
+    },
+    {
+      "epoch": 9.521290322580645,
+      "grad_norm": 0.7211841344833374,
+      "learning_rate": 0.0001365600691443388,
+      "loss": 0.4265,
+      "step": 1840
+    },
+    {
+      "epoch": 9.526451612903227,
+      "grad_norm": 0.682772159576416,
+      "learning_rate": 0.00013652549697493516,
+      "loss": 0.3821,
+      "step": 1841
+    },
+    {
+      "epoch": 9.531612903225806,
+      "grad_norm": 0.6546754837036133,
+      "learning_rate": 0.00013649092480553157,
+      "loss": 0.4309,
+      "step": 1842
+    },
+    {
+      "epoch": 9.536774193548387,
+      "grad_norm": 0.642796516418457,
+      "learning_rate": 0.00013645635263612792,
+      "loss": 0.4035,
+      "step": 1843
+    },
+    {
+      "epoch": 9.541935483870969,
+      "grad_norm": 0.7465437054634094,
+      "learning_rate": 0.0001364217804667243,
+      "loss": 0.4147,
+      "step": 1844
+    },
+    {
+      "epoch": 9.547096774193548,
+      "grad_norm": 0.7043260335922241,
+      "learning_rate": 0.00013638720829732066,
+      "loss": 0.4399,
+      "step": 1845
+    },
+    {
+      "epoch": 9.55225806451613,
+      "grad_norm": 0.6995072364807129,
+      "learning_rate": 0.00013635263612791704,
+      "loss": 0.4353,
+      "step": 1846
+    },
+    {
+      "epoch": 9.557419354838709,
+      "grad_norm": 0.7425726652145386,
+      "learning_rate": 0.00013631806395851342,
+      "loss": 0.4467,
+      "step": 1847
+    },
+    {
+      "epoch": 9.56258064516129,
+      "grad_norm": 0.7601122856140137,
+      "learning_rate": 0.00013628349178910977,
+      "loss": 0.4389,
+      "step": 1848
+    },
+    {
+      "epoch": 9.567741935483872,
+      "grad_norm": 0.7394837141036987,
+      "learning_rate": 0.00013624891961970615,
+      "loss": 0.4424,
+      "step": 1849
+    },
+    {
+      "epoch": 9.572903225806451,
+      "grad_norm": 0.7020159959793091,
+      "learning_rate": 0.0001362143474503025,
+      "loss": 0.3702,
+      "step": 1850
+    },
+    {
+      "epoch": 9.578064516129032,
+      "grad_norm": 0.7405157685279846,
+      "learning_rate": 0.00013617977528089889,
+      "loss": 0.4481,
+      "step": 1851
+    },
+    {
+      "epoch": 9.583225806451614,
+      "grad_norm": 0.7857433557510376,
+      "learning_rate": 0.00013614520311149527,
+      "loss": 0.4375,
+      "step": 1852
+    },
+    {
+      "epoch": 9.588387096774193,
+      "grad_norm": 0.7635135054588318,
+      "learning_rate": 0.00013611063094209162,
+      "loss": 0.4468,
+      "step": 1853
+    },
+    {
+      "epoch": 9.593548387096774,
+      "grad_norm": 0.7334104776382446,
+      "learning_rate": 0.000136076058772688,
+      "loss": 0.4534,
+      "step": 1854
+    },
+    {
+      "epoch": 9.598709677419356,
+      "grad_norm": 0.6774014234542847,
+      "learning_rate": 0.00013604148660328435,
+      "loss": 0.4207,
+      "step": 1855
+    },
+    {
+      "epoch": 9.603870967741935,
+      "grad_norm": 0.6980052590370178,
+      "learning_rate": 0.00013600691443388073,
+      "loss": 0.4234,
+      "step": 1856
+    },
+    {
+      "epoch": 9.609032258064516,
+      "grad_norm": 0.7015699744224548,
+      "learning_rate": 0.00013597234226447711,
+      "loss": 0.4528,
+      "step": 1857
+    },
+    {
+      "epoch": 9.614193548387096,
+      "grad_norm": 0.8216930031776428,
+      "learning_rate": 0.00013593777009507347,
+      "loss": 0.4685,
+      "step": 1858
+    },
+    {
+      "epoch": 9.619354838709677,
+      "grad_norm": 0.7389256954193115,
+      "learning_rate": 0.00013590319792566982,
+      "loss": 0.4296,
+      "step": 1859
+    },
+    {
+      "epoch": 9.624516129032259,
+      "grad_norm": 0.8392791152000427,
+      "learning_rate": 0.00013586862575626623,
+      "loss": 0.4326,
+      "step": 1860
+    },
+    {
+      "epoch": 9.629677419354838,
+      "grad_norm": 0.7449894547462463,
+      "learning_rate": 0.00013583405358686258,
+      "loss": 0.4576,
+      "step": 1861
+    },
+    {
+      "epoch": 9.63483870967742,
+      "grad_norm": 0.7563722729682922,
+      "learning_rate": 0.00013579948141745896,
+      "loss": 0.4501,
+      "step": 1862
+    },
+    {
+      "epoch": 9.64,
+      "grad_norm": 0.7838294506072998,
+      "learning_rate": 0.00013576490924805532,
+      "loss": 0.478,
+      "step": 1863
+    },
+    {
+      "epoch": 9.64516129032258,
+      "grad_norm": 0.7101895809173584,
+      "learning_rate": 0.00013573033707865167,
+      "loss": 0.4093,
+      "step": 1864
+    },
+    {
+      "epoch": 9.650322580645161,
+      "grad_norm": 0.8449984192848206,
+      "learning_rate": 0.00013569576490924808,
+      "loss": 0.465,
+      "step": 1865
+    },
+    {
+      "epoch": 9.655483870967743,
+      "grad_norm": 0.7697463035583496,
+      "learning_rate": 0.00013566119273984443,
+      "loss": 0.4256,
+      "step": 1866
+    },
+    {
+      "epoch": 9.660645161290322,
+      "grad_norm": 0.7203311920166016,
+      "learning_rate": 0.0001356266205704408,
+      "loss": 0.4578,
+      "step": 1867
+    },
+    {
+      "epoch": 9.665806451612903,
+      "grad_norm": 0.7207502126693726,
+      "learning_rate": 0.00013559204840103716,
+      "loss": 0.4366,
+      "step": 1868
+    },
+    {
+      "epoch": 9.670967741935485,
+      "grad_norm": 0.7068636417388916,
+      "learning_rate": 0.00013555747623163355,
+      "loss": 0.4224,
+      "step": 1869
+    },
+    {
+      "epoch": 9.676129032258064,
+      "grad_norm": 0.7015514373779297,
+      "learning_rate": 0.00013552290406222993,
+      "loss": 0.4699,
+      "step": 1870
+    },
+    {
+      "epoch": 9.681290322580645,
+      "grad_norm": 0.6848369240760803,
+      "learning_rate": 0.00013548833189282628,
+      "loss": 0.4064,
+      "step": 1871
+    },
+    {
+      "epoch": 9.686451612903225,
+      "grad_norm": 0.703742504119873,
+      "learning_rate": 0.00013545375972342266,
+      "loss": 0.4642,
+      "step": 1872
+    },
+    {
+      "epoch": 9.691612903225806,
+      "grad_norm": 0.7432376146316528,
+      "learning_rate": 0.000135419187554019,
+      "loss": 0.4746,
+      "step": 1873
+    },
+    {
+      "epoch": 9.696774193548388,
+      "grad_norm": 0.7278482913970947,
+      "learning_rate": 0.0001353846153846154,
+      "loss": 0.4432,
+      "step": 1874
+    },
+    {
+      "epoch": 9.701935483870967,
+      "grad_norm": 0.6945151090621948,
+      "learning_rate": 0.00013535004321521177,
+      "loss": 0.4567,
+      "step": 1875
+    },
+    {
+      "epoch": 9.707096774193548,
+      "grad_norm": 0.9178923964500427,
+      "learning_rate": 0.00013531547104580813,
+      "loss": 0.422,
+      "step": 1876
+    },
+    {
+      "epoch": 9.71225806451613,
+      "grad_norm": 0.6944664716720581,
+      "learning_rate": 0.00013528089887640448,
+      "loss": 0.4235,
+      "step": 1877
+    },
+    {
+      "epoch": 9.717419354838709,
+      "grad_norm": 0.6766967177391052,
+      "learning_rate": 0.0001352463267070009,
+      "loss": 0.4273,
+      "step": 1878
+    },
+    {
+      "epoch": 9.72258064516129,
+      "grad_norm": 0.837324857711792,
+      "learning_rate": 0.00013521175453759724,
+      "loss": 0.4875,
+      "step": 1879
+    },
+    {
+      "epoch": 9.727741935483872,
+      "grad_norm": 0.7537996172904968,
+      "learning_rate": 0.00013517718236819362,
+      "loss": 0.3871,
+      "step": 1880
+    },
+    {
+      "epoch": 9.732903225806451,
+      "grad_norm": 0.748978853225708,
+      "learning_rate": 0.00013514261019878998,
+      "loss": 0.4297,
+      "step": 1881
+    },
+    {
+      "epoch": 9.738064516129032,
+      "grad_norm": 0.7428856492042542,
+      "learning_rate": 0.00013510803802938633,
+      "loss": 0.4379,
+      "step": 1882
+    },
+    {
+      "epoch": 9.743225806451614,
+      "grad_norm": 0.778961181640625,
+      "learning_rate": 0.00013507346585998274,
+      "loss": 0.4405,
+      "step": 1883
+    },
+    {
+      "epoch": 9.748387096774193,
+      "grad_norm": 0.711621105670929,
+      "learning_rate": 0.0001350388936905791,
+      "loss": 0.4122,
+      "step": 1884
+    },
+    {
+      "epoch": 9.753548387096775,
+      "grad_norm": 0.6866260766983032,
+      "learning_rate": 0.00013500432152117547,
+      "loss": 0.3917,
+      "step": 1885
+    },
+    {
+      "epoch": 9.758709677419354,
+      "grad_norm": 0.7280053496360779,
+      "learning_rate": 0.00013496974935177182,
+      "loss": 0.4494,
+      "step": 1886
+    },
+    {
+      "epoch": 9.763870967741935,
+      "grad_norm": 0.7390021085739136,
+      "learning_rate": 0.0001349351771823682,
+      "loss": 0.4237,
+      "step": 1887
+    },
+    {
+      "epoch": 9.769032258064517,
+      "grad_norm": 0.7330762147903442,
+      "learning_rate": 0.00013490060501296459,
+      "loss": 0.4534,
+      "step": 1888
+    },
+    {
+      "epoch": 9.774193548387096,
+      "grad_norm": 0.7682686448097229,
+      "learning_rate": 0.00013486603284356094,
+      "loss": 0.4673,
+      "step": 1889
+    },
+    {
+      "epoch": 9.779354838709677,
+      "grad_norm": 0.7759533524513245,
+      "learning_rate": 0.00013483146067415732,
+      "loss": 0.4676,
+      "step": 1890
+    },
+    {
+      "epoch": 9.784516129032259,
+      "grad_norm": 0.7314127683639526,
+      "learning_rate": 0.00013479688850475367,
+      "loss": 0.45,
+      "step": 1891
+    },
+    {
+      "epoch": 9.789677419354838,
+      "grad_norm": 0.8055412173271179,
+      "learning_rate": 0.00013476231633535005,
+      "loss": 0.4422,
+      "step": 1892
+    },
+    {
+      "epoch": 9.79483870967742,
+      "grad_norm": 0.6766437292098999,
+      "learning_rate": 0.00013472774416594643,
+      "loss": 0.4113,
+      "step": 1893
+    },
+    {
+      "epoch": 9.8,
+      "grad_norm": 0.6854450106620789,
+      "learning_rate": 0.0001346931719965428,
+      "loss": 0.3912,
+      "step": 1894
+    },
+    {
+      "epoch": 9.80516129032258,
+      "grad_norm": 0.7668018937110901,
+      "learning_rate": 0.00013465859982713914,
+      "loss": 0.4316,
+      "step": 1895
+    },
+    {
+      "epoch": 9.810322580645161,
+      "grad_norm": 0.7431079149246216,
+      "learning_rate": 0.00013462402765773552,
+      "loss": 0.4701,
+      "step": 1896
+    },
+    {
+      "epoch": 9.815483870967743,
+      "grad_norm": 0.7447853684425354,
+      "learning_rate": 0.0001345894554883319,
+      "loss": 0.424,
+      "step": 1897
+    },
+    {
+      "epoch": 9.820645161290322,
+      "grad_norm": 0.6748610734939575,
+      "learning_rate": 0.00013455488331892828,
+      "loss": 0.3556,
+      "step": 1898
+    },
+    {
+      "epoch": 9.825806451612904,
+      "grad_norm": 0.741047739982605,
+      "learning_rate": 0.00013452031114952464,
+      "loss": 0.4255,
+      "step": 1899
+    },
+    {
+      "epoch": 9.830967741935483,
+      "grad_norm": 0.7065075635910034,
+      "learning_rate": 0.000134485738980121,
+      "loss": 0.4377,
+      "step": 1900
+    },
+    {
+      "epoch": 9.836129032258064,
+      "grad_norm": 0.7635223269462585,
+      "learning_rate": 0.0001344511668107174,
+      "loss": 0.4689,
+      "step": 1901
+    },
+    {
+      "epoch": 9.841290322580646,
+      "grad_norm": 0.7395762801170349,
+      "learning_rate": 0.00013441659464131375,
+      "loss": 0.4275,
+      "step": 1902
+    },
+    {
+      "epoch": 9.846451612903225,
+      "grad_norm": 0.8523423671722412,
+      "learning_rate": 0.00013438202247191013,
+      "loss": 0.4691,
+      "step": 1903
+    },
+    {
+      "epoch": 9.851612903225806,
+      "grad_norm": 0.7060533165931702,
+      "learning_rate": 0.00013434745030250648,
+      "loss": 0.4125,
+      "step": 1904
+    },
+    {
+      "epoch": 9.856774193548388,
+      "grad_norm": 0.7751105427742004,
+      "learning_rate": 0.00013431287813310284,
+      "loss": 0.4337,
+      "step": 1905
+    },
+    {
+      "epoch": 9.861935483870967,
+      "grad_norm": 0.720875084400177,
+      "learning_rate": 0.00013427830596369925,
+      "loss": 0.4814,
+      "step": 1906
+    },
+    {
+      "epoch": 9.867096774193548,
+      "grad_norm": 0.8009320497512817,
+      "learning_rate": 0.0001342437337942956,
+      "loss": 0.4557,
+      "step": 1907
+    },
+    {
+      "epoch": 9.87225806451613,
+      "grad_norm": 0.76982182264328,
+      "learning_rate": 0.00013420916162489195,
+      "loss": 0.4251,
+      "step": 1908
+    },
+    {
+      "epoch": 9.87741935483871,
+      "grad_norm": 0.7570178508758545,
+      "learning_rate": 0.00013417458945548833,
+      "loss": 0.4534,
+      "step": 1909
+    },
+    {
+      "epoch": 9.88258064516129,
+      "grad_norm": 0.7827386856079102,
+      "learning_rate": 0.0001341400172860847,
+      "loss": 0.4823,
+      "step": 1910
+    },
+    {
+      "epoch": 9.88774193548387,
+      "grad_norm": 0.7673949599266052,
+      "learning_rate": 0.0001341054451166811,
+      "loss": 0.4905,
+      "step": 1911
+    },
+    {
+      "epoch": 9.892903225806451,
+      "grad_norm": 0.7362082600593567,
+      "learning_rate": 0.00013407087294727745,
+      "loss": 0.4192,
+      "step": 1912
+    },
+    {
+      "epoch": 9.898064516129033,
+      "grad_norm": 0.8023606538772583,
+      "learning_rate": 0.0001340363007778738,
+      "loss": 0.4535,
+      "step": 1913
+    },
+    {
+      "epoch": 9.903225806451612,
+      "grad_norm": 0.7599285244941711,
+      "learning_rate": 0.00013400172860847018,
+      "loss": 0.4118,
+      "step": 1914
+    },
+    {
+      "epoch": 9.908387096774193,
+      "grad_norm": 0.6869081854820251,
+      "learning_rate": 0.00013396715643906656,
+      "loss": 0.4499,
+      "step": 1915
+    },
+    {
+      "epoch": 9.913548387096775,
+      "grad_norm": 0.7473227381706238,
+      "learning_rate": 0.00013393258426966294,
+      "loss": 0.4689,
+      "step": 1916
+    },
+    {
+      "epoch": 9.918709677419354,
+      "grad_norm": 0.7235974669456482,
+      "learning_rate": 0.0001338980121002593,
+      "loss": 0.4537,
+      "step": 1917
+    },
+    {
+      "epoch": 9.923870967741935,
+      "grad_norm": 0.7524259090423584,
+      "learning_rate": 0.00013386343993085565,
+      "loss": 0.4689,
+      "step": 1918
+    },
+    {
+      "epoch": 9.929032258064517,
+      "grad_norm": 0.7552942633628845,
+      "learning_rate": 0.00013382886776145206,
+      "loss": 0.4,
+      "step": 1919
+    },
+    {
+      "epoch": 9.934193548387096,
+      "grad_norm": 0.8120148777961731,
+      "learning_rate": 0.0001337942955920484,
+      "loss": 0.4675,
+      "step": 1920
+    },
+    {
+      "epoch": 9.939354838709678,
+      "grad_norm": 0.7683099508285522,
+      "learning_rate": 0.0001337597234226448,
+      "loss": 0.4389,
+      "step": 1921
+    },
+    {
+      "epoch": 9.944516129032259,
+      "grad_norm": 0.7912271618843079,
+      "learning_rate": 0.00013372515125324114,
+      "loss": 0.4411,
+      "step": 1922
+    },
+    {
+      "epoch": 9.949677419354838,
+      "grad_norm": 0.8148561716079712,
+      "learning_rate": 0.0001336905790838375,
+      "loss": 0.4976,
+      "step": 1923
+    },
+    {
+      "epoch": 9.95483870967742,
+      "grad_norm": 0.8190706372261047,
+      "learning_rate": 0.0001336560069144339,
+      "loss": 0.4422,
+      "step": 1924
+    },
+    {
+      "epoch": 9.96,
+      "grad_norm": 0.7836991548538208,
+      "learning_rate": 0.00013362143474503026,
+      "loss": 0.4428,
+      "step": 1925
+    },
+    {
+      "epoch": 9.96516129032258,
+      "grad_norm": 0.7812127470970154,
+      "learning_rate": 0.0001335868625756266,
+      "loss": 0.424,
+      "step": 1926
+    },
+    {
+      "epoch": 9.970322580645162,
+      "grad_norm": 0.7627947330474854,
+      "learning_rate": 0.000133552290406223,
+      "loss": 0.4605,
+      "step": 1927
+    },
+    {
+      "epoch": 9.975483870967741,
+      "grad_norm": 0.73377525806427,
+      "learning_rate": 0.00013351771823681935,
+      "loss": 0.4267,
+      "step": 1928
+    },
+    {
+      "epoch": 9.980645161290322,
+      "grad_norm": 0.7174935936927795,
+      "learning_rate": 0.00013348314606741575,
+      "loss": 0.4369,
+      "step": 1929
+    },
+    {
+      "epoch": 9.985806451612904,
+      "grad_norm": 0.7408297657966614,
+      "learning_rate": 0.0001334485738980121,
+      "loss": 0.4259,
+      "step": 1930
+    },
+    {
+      "epoch": 9.990967741935483,
+      "grad_norm": 0.8314993381500244,
+      "learning_rate": 0.00013341400172860846,
+      "loss": 0.3981,
+      "step": 1931
+    },
+    {
+      "epoch": 9.996129032258064,
+      "grad_norm": 0.7643855214118958,
+      "learning_rate": 0.00013337942955920484,
+      "loss": 0.4451,
+      "step": 1932
+    },
+    {
+      "epoch": 10.00258064516129,
+      "grad_norm": 1.1727362871170044,
+      "learning_rate": 0.00013334485738980122,
+      "loss": 0.6365,
+      "step": 1933
+    },
+    {
+      "epoch": 10.007741935483871,
+      "grad_norm": 0.5273740291595459,
+      "learning_rate": 0.0001333102852203976,
+      "loss": 0.4635,
+      "step": 1934
+    },
+    {
+      "epoch": 10.012903225806452,
+      "grad_norm": 0.5129649043083191,
+      "learning_rate": 0.00013327571305099396,
+      "loss": 0.4366,
+      "step": 1935
+    },
+    {
+      "epoch": 10.018064516129032,
+      "grad_norm": 0.5790820121765137,
+      "learning_rate": 0.0001332411408815903,
+      "loss": 0.4155,
+      "step": 1936
+    },
+    {
+      "epoch": 10.023225806451613,
+      "grad_norm": 0.5180301666259766,
+      "learning_rate": 0.0001332065687121867,
+      "loss": 0.3787,
+      "step": 1937
+    },
+    {
+      "epoch": 10.028387096774194,
+      "grad_norm": 0.5655651092529297,
+      "learning_rate": 0.00013317199654278307,
+      "loss": 0.3862,
+      "step": 1938
+    },
+    {
+      "epoch": 10.033548387096774,
+      "grad_norm": 0.6480773687362671,
+      "learning_rate": 0.00013313742437337945,
+      "loss": 0.4356,
+      "step": 1939
+    },
+    {
+      "epoch": 10.038709677419355,
+      "grad_norm": 0.6343084573745728,
+      "learning_rate": 0.0001331028522039758,
+      "loss": 0.3779,
+      "step": 1940
+    },
+    {
+      "epoch": 10.043870967741935,
+      "grad_norm": 0.7367557287216187,
+      "learning_rate": 0.00013306828003457216,
+      "loss": 0.4257,
+      "step": 1941
+    },
+    {
+      "epoch": 10.049032258064516,
+      "grad_norm": 0.6400299072265625,
+      "learning_rate": 0.00013303370786516857,
+      "loss": 0.42,
+      "step": 1942
+    },
+    {
+      "epoch": 10.054193548387097,
+      "grad_norm": 0.6952190399169922,
+      "learning_rate": 0.00013299913569576492,
+      "loss": 0.4359,
+      "step": 1943
+    },
+    {
+      "epoch": 10.059354838709677,
+      "grad_norm": 0.6141871809959412,
+      "learning_rate": 0.00013296456352636127,
+      "loss": 0.371,
+      "step": 1944
+    },
+    {
+      "epoch": 10.064516129032258,
+      "grad_norm": 0.5727283358573914,
+      "learning_rate": 0.00013292999135695765,
+      "loss": 0.3928,
+      "step": 1945
+    },
+    {
+      "epoch": 10.06967741935484,
+      "grad_norm": 0.5173968076705933,
+      "learning_rate": 0.000132895419187554,
+      "loss": 0.3874,
+      "step": 1946
+    },
+    {
+      "epoch": 10.074838709677419,
+      "grad_norm": 0.5824835896492004,
+      "learning_rate": 0.00013286084701815041,
+      "loss": 0.4179,
+      "step": 1947
+    },
+    {
+      "epoch": 10.08,
+      "grad_norm": 0.5003191232681274,
+      "learning_rate": 0.00013282627484874677,
+      "loss": 0.3722,
+      "step": 1948
+    },
+    {
+      "epoch": 10.085161290322581,
+      "grad_norm": 0.5444069504737854,
+      "learning_rate": 0.00013279170267934312,
+      "loss": 0.4146,
+      "step": 1949
+    },
+    {
+      "epoch": 10.09032258064516,
+      "grad_norm": 0.5744116306304932,
+      "learning_rate": 0.0001327571305099395,
+      "loss": 0.4086,
+      "step": 1950
+    },
+    {
+      "epoch": 10.095483870967742,
+      "grad_norm": 0.5888593196868896,
+      "learning_rate": 0.00013272255834053588,
+      "loss": 0.3936,
+      "step": 1951
+    },
+    {
+      "epoch": 10.100645161290323,
+      "grad_norm": 0.583817183971405,
+      "learning_rate": 0.00013268798617113226,
+      "loss": 0.4406,
+      "step": 1952
+    },
+    {
+      "epoch": 10.105806451612903,
+      "grad_norm": 0.6233108639717102,
+      "learning_rate": 0.00013265341400172862,
+      "loss": 0.3983,
+      "step": 1953
+    },
+    {
+      "epoch": 10.110967741935484,
+      "grad_norm": 0.6418638825416565,
+      "learning_rate": 0.00013261884183232497,
+      "loss": 0.4441,
+      "step": 1954
+    },
+    {
+      "epoch": 10.116129032258064,
+      "grad_norm": 0.5819984078407288,
+      "learning_rate": 0.00013258426966292135,
+      "loss": 0.376,
+      "step": 1955
+    },
+    {
+      "epoch": 10.121290322580645,
+      "grad_norm": 0.6008254885673523,
+      "learning_rate": 0.00013254969749351773,
+      "loss": 0.393,
+      "step": 1956
+    },
+    {
+      "epoch": 10.126451612903226,
+      "grad_norm": 0.602307915687561,
+      "learning_rate": 0.0001325151253241141,
+      "loss": 0.4038,
+      "step": 1957
+    },
+    {
+      "epoch": 10.131612903225806,
+      "grad_norm": 0.5595482587814331,
+      "learning_rate": 0.00013248055315471046,
+      "loss": 0.3822,
+      "step": 1958
+    },
+    {
+      "epoch": 10.136774193548387,
+      "grad_norm": 0.6016210317611694,
+      "learning_rate": 0.00013244598098530682,
+      "loss": 0.3744,
+      "step": 1959
+    },
+    {
+      "epoch": 10.141935483870968,
+      "grad_norm": 0.6553309559822083,
+      "learning_rate": 0.00013241140881590323,
+      "loss": 0.442,
+      "step": 1960
+    },
+    {
+      "epoch": 10.147096774193548,
+      "grad_norm": 0.6580355167388916,
+      "learning_rate": 0.00013237683664649958,
+      "loss": 0.4315,
+      "step": 1961
+    },
+    {
+      "epoch": 10.152258064516129,
+      "grad_norm": 0.5881802439689636,
+      "learning_rate": 0.00013234226447709593,
+      "loss": 0.4005,
+      "step": 1962
+    },
+    {
+      "epoch": 10.15741935483871,
+      "grad_norm": 0.49423128366470337,
+      "learning_rate": 0.0001323076923076923,
+      "loss": 0.3641,
+      "step": 1963
+    },
+    {
+      "epoch": 10.16258064516129,
+      "grad_norm": 0.712238609790802,
+      "learning_rate": 0.00013227312013828867,
+      "loss": 0.3955,
+      "step": 1964
+    },
+    {
+      "epoch": 10.167741935483871,
+      "grad_norm": 0.5915084481239319,
+      "learning_rate": 0.00013223854796888507,
+      "loss": 0.4027,
+      "step": 1965
+    },
+    {
+      "epoch": 10.172903225806452,
+      "grad_norm": 0.5979037880897522,
+      "learning_rate": 0.00013220397579948143,
+      "loss": 0.4391,
+      "step": 1966
+    },
+    {
+      "epoch": 10.178064516129032,
+      "grad_norm": 0.6753323078155518,
+      "learning_rate": 0.00013216940363007778,
+      "loss": 0.4194,
+      "step": 1967
+    },
+    {
+      "epoch": 10.183225806451613,
+      "grad_norm": 0.6329878568649292,
+      "learning_rate": 0.00013213483146067416,
+      "loss": 0.4045,
+      "step": 1968
+    },
+    {
+      "epoch": 10.188387096774193,
+      "grad_norm": 0.6254198551177979,
+      "learning_rate": 0.00013210025929127051,
+      "loss": 0.4083,
+      "step": 1969
+    },
+    {
+      "epoch": 10.193548387096774,
+      "grad_norm": 0.6360498070716858,
+      "learning_rate": 0.00013206568712186692,
+      "loss": 0.4361,
+      "step": 1970
+    },
+    {
+      "epoch": 10.198709677419355,
+      "grad_norm": 0.5992193222045898,
+      "learning_rate": 0.00013203111495246328,
+      "loss": 0.4216,
+      "step": 1971
+    },
+    {
+      "epoch": 10.203870967741935,
+      "grad_norm": 0.5562824606895447,
+      "learning_rate": 0.00013199654278305963,
+      "loss": 0.3977,
+      "step": 1972
+    },
+    {
+      "epoch": 10.209032258064516,
+      "grad_norm": 0.5793871879577637,
+      "learning_rate": 0.000131961970613656,
+      "loss": 0.3879,
+      "step": 1973
+    },
+    {
+      "epoch": 10.214193548387097,
+      "grad_norm": 0.6026304364204407,
+      "learning_rate": 0.0001319273984442524,
+      "loss": 0.4218,
+      "step": 1974
+    },
+    {
+      "epoch": 10.219354838709677,
+      "grad_norm": 0.5688220262527466,
+      "learning_rate": 0.00013189282627484877,
+      "loss": 0.4139,
+      "step": 1975
+    },
+    {
+      "epoch": 10.224516129032258,
+      "grad_norm": 0.5610377788543701,
+      "learning_rate": 0.00013185825410544512,
+      "loss": 0.4281,
+      "step": 1976
+    },
+    {
+      "epoch": 10.22967741935484,
+      "grad_norm": 0.6096415519714355,
+      "learning_rate": 0.00013182368193604148,
+      "loss": 0.433,
+      "step": 1977
+    },
+    {
+      "epoch": 10.234838709677419,
+      "grad_norm": 0.5812779664993286,
+      "learning_rate": 0.00013178910976663786,
+      "loss": 0.4269,
+      "step": 1978
+    },
+    {
+      "epoch": 10.24,
+      "grad_norm": 0.6062187552452087,
+      "learning_rate": 0.00013175453759723424,
+      "loss": 0.4408,
+      "step": 1979
+    },
+    {
+      "epoch": 10.245161290322581,
+      "grad_norm": 0.641016960144043,
+      "learning_rate": 0.0001317199654278306,
+      "loss": 0.4485,
+      "step": 1980
+    },
+    {
+      "epoch": 10.250322580645161,
+      "grad_norm": 0.6123722195625305,
+      "learning_rate": 0.00013168539325842697,
+      "loss": 0.4166,
+      "step": 1981
+    },
+    {
+      "epoch": 10.255483870967742,
+      "grad_norm": 0.5762332677841187,
+      "learning_rate": 0.00013165082108902333,
+      "loss": 0.407,
+      "step": 1982
+    },
+    {
+      "epoch": 10.260645161290322,
+      "grad_norm": 0.6724779009819031,
+      "learning_rate": 0.00013161624891961973,
+      "loss": 0.4267,
+      "step": 1983
+    },
+    {
+      "epoch": 10.265806451612903,
+      "grad_norm": 0.6221693158149719,
+      "learning_rate": 0.0001315816767502161,
+      "loss": 0.4055,
+      "step": 1984
+    },
+    {
+      "epoch": 10.270967741935484,
+      "grad_norm": 0.6238535642623901,
+      "learning_rate": 0.00013154710458081244,
+      "loss": 0.4296,
+      "step": 1985
+    },
+    {
+      "epoch": 10.276129032258064,
+      "grad_norm": 0.6058987975120544,
+      "learning_rate": 0.00013151253241140882,
+      "loss": 0.4484,
+      "step": 1986
+    },
+    {
+      "epoch": 10.281290322580645,
+      "grad_norm": 0.5579047203063965,
+      "learning_rate": 0.00013147796024200517,
+      "loss": 0.3891,
+      "step": 1987
+    },
+    {
+      "epoch": 10.286451612903226,
+      "grad_norm": 0.6153200268745422,
+      "learning_rate": 0.00013144338807260158,
+      "loss": 0.3958,
+      "step": 1988
+    },
+    {
+      "epoch": 10.291612903225806,
+      "grad_norm": 0.5433972477912903,
+      "learning_rate": 0.00013140881590319794,
+      "loss": 0.3871,
+      "step": 1989
+    },
+    {
+      "epoch": 10.296774193548387,
+      "grad_norm": 0.5915231108665466,
+      "learning_rate": 0.0001313742437337943,
+      "loss": 0.336,
+      "step": 1990
+    },
+    {
+      "epoch": 10.301935483870968,
+      "grad_norm": 0.6280999779701233,
+      "learning_rate": 0.00013133967156439067,
+      "loss": 0.3665,
+      "step": 1991
+    },
+    {
+      "epoch": 10.307096774193548,
+      "grad_norm": 0.6030144095420837,
+      "learning_rate": 0.00013130509939498705,
+      "loss": 0.386,
+      "step": 1992
+    },
+    {
+      "epoch": 10.31225806451613,
+      "grad_norm": 0.6012431979179382,
+      "learning_rate": 0.00013127052722558343,
+      "loss": 0.4191,
+      "step": 1993
+    },
+    {
+      "epoch": 10.31741935483871,
+      "grad_norm": 0.7044047117233276,
+      "learning_rate": 0.00013123595505617978,
+      "loss": 0.4123,
+      "step": 1994
+    },
+    {
+      "epoch": 10.32258064516129,
+      "grad_norm": 0.559673011302948,
+      "learning_rate": 0.00013120138288677614,
+      "loss": 0.3865,
+      "step": 1995
+    },
+    {
+      "epoch": 10.327741935483871,
+      "grad_norm": 0.7041928172111511,
+      "learning_rate": 0.00013116681071737252,
+      "loss": 0.4439,
+      "step": 1996
+    },
+    {
+      "epoch": 10.33290322580645,
+      "grad_norm": 0.668373703956604,
+      "learning_rate": 0.0001311322385479689,
+      "loss": 0.4234,
+      "step": 1997
+    },
+    {
+      "epoch": 10.338064516129032,
+      "grad_norm": 0.6147612929344177,
+      "learning_rate": 0.00013109766637856525,
+      "loss": 0.4196,
+      "step": 1998
+    },
+    {
+      "epoch": 10.343225806451613,
+      "grad_norm": 0.539064347743988,
+      "learning_rate": 0.00013106309420916163,
+      "loss": 0.4082,
+      "step": 1999
+    },
+    {
+      "epoch": 10.348387096774193,
+      "grad_norm": 0.5952771306037903,
+      "learning_rate": 0.00013102852203975799,
+      "loss": 0.4478,
+      "step": 2000
+    },
+    {
+      "epoch": 10.353548387096774,
+      "grad_norm": 0.6297308206558228,
+      "learning_rate": 0.00013099394987035437,
+      "loss": 0.3786,
+      "step": 2001
+    },
+    {
+      "epoch": 10.358709677419355,
+      "grad_norm": 0.630225419998169,
+      "learning_rate": 0.00013095937770095075,
+      "loss": 0.4477,
+      "step": 2002
+    },
+    {
+      "epoch": 10.363870967741935,
+      "grad_norm": 0.595138669013977,
+      "learning_rate": 0.0001309248055315471,
+      "loss": 0.4075,
+      "step": 2003
+    },
+    {
+      "epoch": 10.369032258064516,
+      "grad_norm": 0.6205843091011047,
+      "learning_rate": 0.00013089023336214348,
+      "loss": 0.4205,
+      "step": 2004
+    },
+    {
+      "epoch": 10.374193548387098,
+      "grad_norm": 0.6673692464828491,
+      "learning_rate": 0.00013085566119273983,
+      "loss": 0.4402,
+      "step": 2005
+    },
+    {
+      "epoch": 10.379354838709677,
+      "grad_norm": 0.6677766442298889,
+      "learning_rate": 0.00013082108902333624,
+      "loss": 0.4003,
+      "step": 2006
+    },
+    {
+      "epoch": 10.384516129032258,
+      "grad_norm": 0.6619275212287903,
+      "learning_rate": 0.0001307865168539326,
+      "loss": 0.4264,
+      "step": 2007
+    },
+    {
+      "epoch": 10.38967741935484,
+      "grad_norm": 0.5487556457519531,
+      "learning_rate": 0.00013075194468452895,
+      "loss": 0.4038,
+      "step": 2008
+    },
+    {
+      "epoch": 10.394838709677419,
+      "grad_norm": 0.6760810613632202,
+      "learning_rate": 0.00013071737251512533,
+      "loss": 0.4314,
+      "step": 2009
+    },
+    {
+      "epoch": 10.4,
+      "grad_norm": 0.618717610836029,
+      "learning_rate": 0.00013068280034572168,
+      "loss": 0.4419,
+      "step": 2010
+    },
+    {
+      "epoch": 10.40516129032258,
+      "grad_norm": 0.5799922347068787,
+      "learning_rate": 0.00013064822817631806,
+      "loss": 0.4206,
+      "step": 2011
+    },
+    {
+      "epoch": 10.410322580645161,
+      "grad_norm": 0.5809241533279419,
+      "learning_rate": 0.00013061365600691444,
+      "loss": 0.4213,
+      "step": 2012
+    },
+    {
+      "epoch": 10.415483870967742,
+      "grad_norm": 0.6215810179710388,
+      "learning_rate": 0.0001305790838375108,
+      "loss": 0.444,
+      "step": 2013
+    },
+    {
+      "epoch": 10.420645161290322,
+      "grad_norm": 0.6064811944961548,
+      "learning_rate": 0.00013054451166810718,
+      "loss": 0.4112,
+      "step": 2014
+    },
+    {
+      "epoch": 10.425806451612903,
+      "grad_norm": 0.6586175560951233,
+      "learning_rate": 0.00013050993949870356,
+      "loss": 0.4445,
+      "step": 2015
+    },
+    {
+      "epoch": 10.430967741935484,
+      "grad_norm": 0.5858752131462097,
+      "learning_rate": 0.0001304753673292999,
+      "loss": 0.4221,
+      "step": 2016
+    },
+    {
+      "epoch": 10.436129032258064,
+      "grad_norm": 0.6382436156272888,
+      "learning_rate": 0.0001304407951598963,
+      "loss": 0.4197,
+      "step": 2017
+    },
+    {
+      "epoch": 10.441290322580645,
+      "grad_norm": 0.6052298545837402,
+      "learning_rate": 0.00013040622299049265,
+      "loss": 0.4425,
+      "step": 2018
+    },
+    {
+      "epoch": 10.446451612903227,
+      "grad_norm": 0.5728411078453064,
+      "learning_rate": 0.00013037165082108903,
+      "loss": 0.4078,
+      "step": 2019
+    },
+    {
+      "epoch": 10.451612903225806,
+      "grad_norm": 0.686301052570343,
+      "learning_rate": 0.0001303370786516854,
+      "loss": 0.4381,
+      "step": 2020
+    },
+    {
+      "epoch": 10.456774193548387,
+      "grad_norm": 0.6365966200828552,
+      "learning_rate": 0.00013030250648228176,
+      "loss": 0.4465,
+      "step": 2021
+    },
+    {
+      "epoch": 10.461935483870967,
+      "grad_norm": 0.6815969347953796,
+      "learning_rate": 0.00013026793431287814,
+      "loss": 0.4135,
+      "step": 2022
+    },
+    {
+      "epoch": 10.467096774193548,
+      "grad_norm": 0.6123529076576233,
+      "learning_rate": 0.0001302333621434745,
+      "loss": 0.4536,
+      "step": 2023
+    },
+    {
+      "epoch": 10.47225806451613,
+      "grad_norm": 0.6690642237663269,
+      "learning_rate": 0.0001301987899740709,
+      "loss": 0.4404,
+      "step": 2024
+    },
+    {
+      "epoch": 10.477419354838709,
+      "grad_norm": 0.6295714974403381,
+      "learning_rate": 0.00013016421780466726,
+      "loss": 0.4214,
+      "step": 2025
+    },
+    {
+      "epoch": 10.48258064516129,
+      "grad_norm": 0.6704673171043396,
+      "learning_rate": 0.0001301296456352636,
+      "loss": 0.4291,
+      "step": 2026
+    },
+    {
+      "epoch": 10.487741935483871,
+      "grad_norm": 0.7165770530700684,
+      "learning_rate": 0.00013009507346586,
+      "loss": 0.4491,
+      "step": 2027
+    },
+    {
+      "epoch": 10.492903225806451,
+      "grad_norm": 0.6158996224403381,
+      "learning_rate": 0.00013006050129645634,
+      "loss": 0.441,
+      "step": 2028
+    },
+    {
+      "epoch": 10.498064516129032,
+      "grad_norm": 0.6307876110076904,
+      "learning_rate": 0.00013002592912705272,
+      "loss": 0.4271,
+      "step": 2029
+    },
+    {
+      "epoch": 10.503225806451614,
+      "grad_norm": 0.6207247376441956,
+      "learning_rate": 0.0001299913569576491,
+      "loss": 0.3854,
+      "step": 2030
+    },
+    {
+      "epoch": 10.503225806451614,
+      "eval_loss": 2.2759079933166504,
+      "eval_runtime": 21.3277,
+      "eval_samples_per_second": 3.704,
+      "eval_steps_per_second": 0.469,
+      "step": 2030
+    },
+    {
+      "epoch": 10.508387096774193,
+      "grad_norm": 0.6595866680145264,
+      "learning_rate": 0.00012995678478824546,
+      "loss": 0.438,
+      "step": 2031
+    },
+    {
+      "epoch": 10.513548387096774,
+      "grad_norm": 0.5970214009284973,
+      "learning_rate": 0.00012992221261884184,
+      "loss": 0.4304,
+      "step": 2032
+    },
+    {
+      "epoch": 10.518709677419356,
+      "grad_norm": 0.6391655206680298,
+      "learning_rate": 0.00012988764044943822,
+      "loss": 0.4328,
+      "step": 2033
+    },
+    {
+      "epoch": 10.523870967741935,
+      "grad_norm": 0.6348317265510559,
+      "learning_rate": 0.00012985306828003457,
+      "loss": 0.4375,
+      "step": 2034
+    },
+    {
+      "epoch": 10.529032258064516,
+      "grad_norm": 0.6792359352111816,
+      "learning_rate": 0.00012981849611063095,
+      "loss": 0.4534,
+      "step": 2035
+    },
+    {
+      "epoch": 10.534193548387098,
+      "grad_norm": 0.593288779258728,
+      "learning_rate": 0.0001297839239412273,
+      "loss": 0.3873,
+      "step": 2036
+    },
+    {
+      "epoch": 10.539354838709677,
+      "grad_norm": 0.6122384667396545,
+      "learning_rate": 0.0001297493517718237,
+      "loss": 0.4507,
+      "step": 2037
+    },
+    {
+      "epoch": 10.544516129032258,
+      "grad_norm": 0.6019447445869446,
+      "learning_rate": 0.00012971477960242007,
+      "loss": 0.4189,
+      "step": 2038
+    },
+    {
+      "epoch": 10.549677419354838,
+      "grad_norm": 0.6903626918792725,
+      "learning_rate": 0.00012968020743301642,
+      "loss": 0.4155,
+      "step": 2039
+    },
+    {
+      "epoch": 10.55483870967742,
+      "grad_norm": 0.5795913934707642,
+      "learning_rate": 0.0001296456352636128,
+      "loss": 0.4005,
+      "step": 2040
+    },
+    {
+      "epoch": 10.56,
+      "grad_norm": 0.6943153738975525,
+      "learning_rate": 0.00012961106309420915,
+      "loss": 0.4076,
+      "step": 2041
+    },
+    {
+      "epoch": 10.56516129032258,
+      "grad_norm": 0.6368845105171204,
+      "learning_rate": 0.00012957649092480554,
+      "loss": 0.4048,
+      "step": 2042
+    },
+    {
+      "epoch": 10.570322580645161,
+      "grad_norm": 0.6390585899353027,
+      "learning_rate": 0.00012954191875540192,
+      "loss": 0.4237,
+      "step": 2043
+    },
+    {
+      "epoch": 10.575483870967743,
+      "grad_norm": 0.6332088708877563,
+      "learning_rate": 0.00012950734658599827,
+      "loss": 0.4253,
+      "step": 2044
+    },
+    {
+      "epoch": 10.580645161290322,
+      "grad_norm": 0.6042909026145935,
+      "learning_rate": 0.00012947277441659465,
+      "loss": 0.3346,
+      "step": 2045
+    },
+    {
+      "epoch": 10.585806451612903,
+      "grad_norm": 0.7008855938911438,
+      "learning_rate": 0.000129438202247191,
+      "loss": 0.4312,
+      "step": 2046
+    },
+    {
+      "epoch": 10.590967741935485,
+      "grad_norm": 0.701012134552002,
+      "learning_rate": 0.00012940363007778738,
+      "loss": 0.4235,
+      "step": 2047
+    },
+    {
+      "epoch": 10.596129032258064,
+      "grad_norm": 0.6564736366271973,
+      "learning_rate": 0.00012936905790838376,
+      "loss": 0.4172,
+      "step": 2048
+    },
+    {
+      "epoch": 10.601290322580645,
+      "grad_norm": 0.635356605052948,
+      "learning_rate": 0.00012933448573898012,
+      "loss": 0.4509,
+      "step": 2049
+    },
+    {
+      "epoch": 10.606451612903225,
+      "grad_norm": 0.6131752133369446,
+      "learning_rate": 0.0001292999135695765,
+      "loss": 0.4185,
+      "step": 2050
+    },
+    {
+      "epoch": 10.611612903225806,
+      "grad_norm": 0.6390723586082458,
+      "learning_rate": 0.00012926534140017285,
+      "loss": 0.4337,
+      "step": 2051
+    },
+    {
+      "epoch": 10.616774193548387,
+      "grad_norm": 0.6386390328407288,
+      "learning_rate": 0.00012923076923076923,
+      "loss": 0.4164,
+      "step": 2052
+    },
+    {
+      "epoch": 10.621935483870967,
+      "grad_norm": 0.6543627381324768,
+      "learning_rate": 0.0001291961970613656,
+      "loss": 0.4141,
+      "step": 2053
+    },
+    {
+      "epoch": 10.627096774193548,
+      "grad_norm": 0.7052140831947327,
+      "learning_rate": 0.00012916162489196197,
+      "loss": 0.4234,
+      "step": 2054
+    },
+    {
+      "epoch": 10.63225806451613,
+      "grad_norm": 0.6348745822906494,
+      "learning_rate": 0.00012912705272255835,
+      "loss": 0.4344,
+      "step": 2055
+    },
+    {
+      "epoch": 10.637419354838709,
+      "grad_norm": 0.6717938780784607,
+      "learning_rate": 0.00012909248055315473,
+      "loss": 0.4093,
+      "step": 2056
+    },
+    {
+      "epoch": 10.64258064516129,
+      "grad_norm": 0.6630579829216003,
+      "learning_rate": 0.00012905790838375108,
+      "loss": 0.4151,
+      "step": 2057
+    },
+    {
+      "epoch": 10.647741935483872,
+      "grad_norm": 0.6930499076843262,
+      "learning_rate": 0.00012902333621434746,
+      "loss": 0.4131,
+      "step": 2058
+    },
+    {
+      "epoch": 10.652903225806451,
+      "grad_norm": 0.6215726137161255,
+      "learning_rate": 0.00012898876404494381,
+      "loss": 0.4057,
+      "step": 2059
+    },
+    {
+      "epoch": 10.658064516129032,
+      "grad_norm": 0.6012431383132935,
+      "learning_rate": 0.0001289541918755402,
+      "loss": 0.4136,
+      "step": 2060
+    },
+    {
+      "epoch": 10.663225806451614,
+      "grad_norm": 0.6120856404304504,
+      "learning_rate": 0.00012891961970613658,
+      "loss": 0.4363,
+      "step": 2061
+    },
+    {
+      "epoch": 10.668387096774193,
+      "grad_norm": 0.5731876492500305,
+      "learning_rate": 0.00012888504753673293,
+      "loss": 0.4345,
+      "step": 2062
+    },
+    {
+      "epoch": 10.673548387096774,
+      "grad_norm": 0.5980328321456909,
+      "learning_rate": 0.0001288504753673293,
+      "loss": 0.3678,
+      "step": 2063
+    },
+    {
+      "epoch": 10.678709677419356,
+      "grad_norm": 0.737721860408783,
+      "learning_rate": 0.00012881590319792566,
+      "loss": 0.4212,
+      "step": 2064
+    },
+    {
+      "epoch": 10.683870967741935,
+      "grad_norm": 0.6829202175140381,
+      "learning_rate": 0.00012878133102852204,
+      "loss": 0.4361,
+      "step": 2065
+    },
+    {
+      "epoch": 10.689032258064517,
+      "grad_norm": 0.6675949096679688,
+      "learning_rate": 0.00012874675885911842,
+      "loss": 0.4165,
+      "step": 2066
+    },
+    {
+      "epoch": 10.694193548387096,
+      "grad_norm": 0.6646575331687927,
+      "learning_rate": 0.00012871218668971478,
+      "loss": 0.4299,
+      "step": 2067
+    },
+    {
+      "epoch": 10.699354838709677,
+      "grad_norm": 0.6798309683799744,
+      "learning_rate": 0.00012867761452031116,
+      "loss": 0.4244,
+      "step": 2068
+    },
+    {
+      "epoch": 10.704516129032259,
+      "grad_norm": 0.7207285761833191,
+      "learning_rate": 0.0001286430423509075,
+      "loss": 0.4559,
+      "step": 2069
+    },
+    {
+      "epoch": 10.709677419354838,
+      "grad_norm": 0.6748340129852295,
+      "learning_rate": 0.0001286084701815039,
+      "loss": 0.4328,
+      "step": 2070
+    },
+    {
+      "epoch": 10.71483870967742,
+      "grad_norm": 0.6957266330718994,
+      "learning_rate": 0.00012857389801210027,
+      "loss": 0.4482,
+      "step": 2071
+    },
+    {
+      "epoch": 10.72,
+      "grad_norm": 0.7052022218704224,
+      "learning_rate": 0.00012853932584269663,
+      "loss": 0.4166,
+      "step": 2072
+    },
+    {
+      "epoch": 10.72516129032258,
+      "grad_norm": 0.6330098509788513,
+      "learning_rate": 0.000128504753673293,
+      "loss": 0.4211,
+      "step": 2073
+    },
+    {
+      "epoch": 10.730322580645161,
+      "grad_norm": 0.7416425347328186,
+      "learning_rate": 0.00012847018150388936,
+      "loss": 0.4306,
+      "step": 2074
+    },
+    {
+      "epoch": 10.735483870967743,
+      "grad_norm": 0.6034896969795227,
+      "learning_rate": 0.00012843560933448574,
+      "loss": 0.4171,
+      "step": 2075
+    },
+    {
+      "epoch": 10.740645161290322,
+      "grad_norm": 0.7180091142654419,
+      "learning_rate": 0.00012840103716508212,
+      "loss": 0.4003,
+      "step": 2076
+    },
+    {
+      "epoch": 10.745806451612903,
+      "grad_norm": 0.6948825716972351,
+      "learning_rate": 0.00012836646499567847,
+      "loss": 0.4094,
+      "step": 2077
+    },
+    {
+      "epoch": 10.750967741935483,
+      "grad_norm": 0.6762283444404602,
+      "learning_rate": 0.00012833189282627485,
+      "loss": 0.361,
+      "step": 2078
+    },
+    {
+      "epoch": 10.756129032258064,
+      "grad_norm": 0.6791343688964844,
+      "learning_rate": 0.00012829732065687124,
+      "loss": 0.3905,
+      "step": 2079
+    },
+    {
+      "epoch": 10.761290322580646,
+      "grad_norm": 0.7536624670028687,
+      "learning_rate": 0.0001282627484874676,
+      "loss": 0.4061,
+      "step": 2080
+    },
+    {
+      "epoch": 10.766451612903225,
+      "grad_norm": 0.7250949144363403,
+      "learning_rate": 0.00012822817631806397,
+      "loss": 0.4075,
+      "step": 2081
+    },
+    {
+      "epoch": 10.771612903225806,
+      "grad_norm": 0.6473909616470337,
+      "learning_rate": 0.00012819360414866032,
+      "loss": 0.3947,
+      "step": 2082
+    },
+    {
+      "epoch": 10.776774193548388,
+      "grad_norm": 0.6633297801017761,
+      "learning_rate": 0.0001281590319792567,
+      "loss": 0.431,
+      "step": 2083
+    },
+    {
+      "epoch": 10.781935483870967,
+      "grad_norm": 0.7178322076797485,
+      "learning_rate": 0.00012812445980985308,
+      "loss": 0.4247,
+      "step": 2084
+    },
+    {
+      "epoch": 10.787096774193548,
+      "grad_norm": 0.6894609928131104,
+      "learning_rate": 0.00012808988764044944,
+      "loss": 0.42,
+      "step": 2085
+    },
+    {
+      "epoch": 10.79225806451613,
+      "grad_norm": 0.6166192889213562,
+      "learning_rate": 0.00012805531547104582,
+      "loss": 0.3727,
+      "step": 2086
+    },
+    {
+      "epoch": 10.79741935483871,
+      "grad_norm": 0.7362133264541626,
+      "learning_rate": 0.00012802074330164217,
+      "loss": 0.4533,
+      "step": 2087
+    },
+    {
+      "epoch": 10.80258064516129,
+      "grad_norm": 0.7351956963539124,
+      "learning_rate": 0.00012798617113223855,
+      "loss": 0.4453,
+      "step": 2088
+    },
+    {
+      "epoch": 10.807741935483872,
+      "grad_norm": 0.6566995978355408,
+      "learning_rate": 0.00012795159896283493,
+      "loss": 0.4018,
+      "step": 2089
+    },
+    {
+      "epoch": 10.812903225806451,
+      "grad_norm": 0.6643356680870056,
+      "learning_rate": 0.00012791702679343129,
+      "loss": 0.4429,
+      "step": 2090
+    },
+    {
+      "epoch": 10.818064516129033,
+      "grad_norm": 0.6539942026138306,
+      "learning_rate": 0.00012788245462402767,
+      "loss": 0.4196,
+      "step": 2091
+    },
+    {
+      "epoch": 10.823225806451614,
+      "grad_norm": 0.6581376194953918,
+      "learning_rate": 0.00012784788245462402,
+      "loss": 0.444,
+      "step": 2092
+    },
+    {
+      "epoch": 10.828387096774193,
+      "grad_norm": 0.6877250671386719,
+      "learning_rate": 0.0001278133102852204,
+      "loss": 0.4332,
+      "step": 2093
+    },
+    {
+      "epoch": 10.833548387096775,
+      "grad_norm": 0.6926650404930115,
+      "learning_rate": 0.00012777873811581678,
+      "loss": 0.4171,
+      "step": 2094
+    },
+    {
+      "epoch": 10.838709677419354,
+      "grad_norm": 0.674882173538208,
+      "learning_rate": 0.00012774416594641313,
+      "loss": 0.3605,
+      "step": 2095
+    },
+    {
+      "epoch": 10.843870967741935,
+      "grad_norm": 0.6932137608528137,
+      "learning_rate": 0.00012770959377700951,
+      "loss": 0.422,
+      "step": 2096
+    },
+    {
+      "epoch": 10.849032258064517,
+      "grad_norm": 0.7092239856719971,
+      "learning_rate": 0.0001276750216076059,
+      "loss": 0.387,
+      "step": 2097
+    },
+    {
+      "epoch": 10.854193548387096,
+      "grad_norm": 0.6979272365570068,
+      "learning_rate": 0.00012764044943820225,
+      "loss": 0.4094,
+      "step": 2098
+    },
+    {
+      "epoch": 10.859354838709677,
+      "grad_norm": 0.7343307137489319,
+      "learning_rate": 0.00012760587726879863,
+      "loss": 0.4563,
+      "step": 2099
+    },
+    {
+      "epoch": 10.864516129032259,
+      "grad_norm": 0.7141764163970947,
+      "learning_rate": 0.00012757130509939498,
+      "loss": 0.4485,
+      "step": 2100
+    },
+    {
+      "epoch": 10.869677419354838,
+      "grad_norm": 0.7524417042732239,
+      "learning_rate": 0.00012753673292999136,
+      "loss": 0.3867,
+      "step": 2101
+    },
+    {
+      "epoch": 10.87483870967742,
+      "grad_norm": 0.6647288203239441,
+      "learning_rate": 0.00012750216076058774,
+      "loss": 0.4273,
+      "step": 2102
+    },
+    {
+      "epoch": 10.88,
+      "grad_norm": 0.7366842031478882,
+      "learning_rate": 0.0001274675885911841,
+      "loss": 0.4545,
+      "step": 2103
+    },
+    {
+      "epoch": 10.88516129032258,
+      "grad_norm": 0.6684737205505371,
+      "learning_rate": 0.00012743301642178048,
+      "loss": 0.427,
+      "step": 2104
+    },
+    {
+      "epoch": 10.890322580645162,
+      "grad_norm": 0.6502785682678223,
+      "learning_rate": 0.00012739844425237683,
+      "loss": 0.4276,
+      "step": 2105
+    },
+    {
+      "epoch": 10.895483870967741,
+      "grad_norm": 0.6435804963111877,
+      "learning_rate": 0.0001273638720829732,
+      "loss": 0.4104,
+      "step": 2106
+    },
+    {
+      "epoch": 10.900645161290322,
+      "grad_norm": 0.6841921210289001,
+      "learning_rate": 0.0001273292999135696,
+      "loss": 0.4536,
+      "step": 2107
+    },
+    {
+      "epoch": 10.905806451612904,
+      "grad_norm": 0.6715157628059387,
+      "learning_rate": 0.00012729472774416595,
+      "loss": 0.4133,
+      "step": 2108
+    },
+    {
+      "epoch": 10.910967741935483,
+      "grad_norm": 0.682794988155365,
+      "learning_rate": 0.00012726015557476233,
+      "loss": 0.4068,
+      "step": 2109
+    },
+    {
+      "epoch": 10.916129032258064,
+      "grad_norm": 0.7062206864356995,
+      "learning_rate": 0.00012722558340535868,
+      "loss": 0.3902,
+      "step": 2110
+    },
+    {
+      "epoch": 10.921290322580646,
+      "grad_norm": 0.6715657711029053,
+      "learning_rate": 0.00012719101123595506,
+      "loss": 0.4215,
+      "step": 2111
+    },
+    {
+      "epoch": 10.926451612903225,
+      "grad_norm": 0.7080001831054688,
+      "learning_rate": 0.00012715643906655144,
+      "loss": 0.4507,
+      "step": 2112
+    },
+    {
+      "epoch": 10.931612903225806,
+      "grad_norm": 0.7404335141181946,
+      "learning_rate": 0.0001271218668971478,
+      "loss": 0.4606,
+      "step": 2113
+    },
+    {
+      "epoch": 10.936774193548388,
+      "grad_norm": 0.7738038301467896,
+      "learning_rate": 0.00012708729472774417,
+      "loss": 0.4122,
+      "step": 2114
+    },
+    {
+      "epoch": 10.941935483870967,
+      "grad_norm": 0.8014647960662842,
+      "learning_rate": 0.00012705272255834053,
+      "loss": 0.4212,
+      "step": 2115
+    },
+    {
+      "epoch": 10.947096774193549,
+      "grad_norm": 0.711954653263092,
+      "learning_rate": 0.0001270181503889369,
+      "loss": 0.4508,
+      "step": 2116
+    },
+    {
+      "epoch": 10.95225806451613,
+      "grad_norm": 0.7111976146697998,
+      "learning_rate": 0.0001269835782195333,
+      "loss": 0.4127,
+      "step": 2117
+    },
+    {
+      "epoch": 10.95741935483871,
+      "grad_norm": 0.6944493651390076,
+      "learning_rate": 0.00012694900605012964,
+      "loss": 0.442,
+      "step": 2118
+    },
+    {
+      "epoch": 10.96258064516129,
+      "grad_norm": 0.6903209686279297,
+      "learning_rate": 0.00012691443388072602,
+      "loss": 0.3994,
+      "step": 2119
+    },
+    {
+      "epoch": 10.967741935483872,
+      "grad_norm": 0.6928601861000061,
+      "learning_rate": 0.0001268798617113224,
+      "loss": 0.4349,
+      "step": 2120
+    },
+    {
+      "epoch": 10.972903225806451,
+      "grad_norm": 0.7232123613357544,
+      "learning_rate": 0.00012684528954191876,
+      "loss": 0.4335,
+      "step": 2121
+    },
+    {
+      "epoch": 10.978064516129033,
+      "grad_norm": 0.6831758618354797,
+      "learning_rate": 0.00012681071737251514,
+      "loss": 0.4326,
+      "step": 2122
+    },
+    {
+      "epoch": 10.983225806451612,
+      "grad_norm": 0.6857551336288452,
+      "learning_rate": 0.0001267761452031115,
+      "loss": 0.3996,
+      "step": 2123
+    },
+    {
+      "epoch": 10.988387096774193,
+      "grad_norm": 0.7000252604484558,
+      "learning_rate": 0.00012674157303370787,
+      "loss": 0.394,
+      "step": 2124
+    },
+    {
+      "epoch": 10.993548387096775,
+      "grad_norm": 0.7301141023635864,
+      "learning_rate": 0.00012670700086430425,
+      "loss": 0.4643,
+      "step": 2125
+    },
+    {
+      "epoch": 10.998709677419354,
+      "grad_norm": 1.1058012247085571,
+      "learning_rate": 0.0001266724286949006,
+      "loss": 0.6684,
+      "step": 2126
+    },
+    {
+      "epoch": 11.005161290322581,
+      "grad_norm": 0.48096445202827454,
+      "learning_rate": 0.00012663785652549699,
+      "loss": 0.3989,
+      "step": 2127
+    },
+    {
+      "epoch": 11.01032258064516,
+      "grad_norm": 0.4272102117538452,
+      "learning_rate": 0.00012660328435609334,
+      "loss": 0.3875,
+      "step": 2128
+    },
+    {
+      "epoch": 11.015483870967742,
+      "grad_norm": 0.5112113356590271,
+      "learning_rate": 0.00012656871218668972,
+      "loss": 0.4113,
+      "step": 2129
+    },
+    {
+      "epoch": 11.020645161290323,
+      "grad_norm": 0.5110041499137878,
+      "learning_rate": 0.0001265341400172861,
+      "loss": 0.3765,
+      "step": 2130
+    },
+    {
+      "epoch": 11.025806451612903,
+      "grad_norm": 0.5860457420349121,
+      "learning_rate": 0.00012649956784788245,
+      "loss": 0.4041,
+      "step": 2131
+    },
+    {
+      "epoch": 11.030967741935484,
+      "grad_norm": 0.5408039093017578,
+      "learning_rate": 0.00012646499567847883,
+      "loss": 0.4184,
+      "step": 2132
+    },
+    {
+      "epoch": 11.036129032258064,
+      "grad_norm": 0.5557346940040588,
+      "learning_rate": 0.0001264304235090752,
+      "loss": 0.383,
+      "step": 2133
+    },
+    {
+      "epoch": 11.041290322580645,
+      "grad_norm": 0.4790826141834259,
+      "learning_rate": 0.00012639585133967157,
+      "loss": 0.366,
+      "step": 2134
+    },
+    {
+      "epoch": 11.046451612903226,
+      "grad_norm": 0.6097589135169983,
+      "learning_rate": 0.00012636127917026795,
+      "loss": 0.4266,
+      "step": 2135
+    },
+    {
+      "epoch": 11.051612903225806,
+      "grad_norm": 0.6016771197319031,
+      "learning_rate": 0.0001263267070008643,
+      "loss": 0.4286,
+      "step": 2136
+    },
+    {
+      "epoch": 11.056774193548387,
+      "grad_norm": 0.5937833786010742,
+      "learning_rate": 0.00012629213483146068,
+      "loss": 0.4301,
+      "step": 2137
+    },
+    {
+      "epoch": 11.061935483870968,
+      "grad_norm": 0.6466284990310669,
+      "learning_rate": 0.00012625756266205706,
+      "loss": 0.4129,
+      "step": 2138
+    },
+    {
+      "epoch": 11.067096774193548,
+      "grad_norm": 0.5362923741340637,
+      "learning_rate": 0.00012622299049265342,
+      "loss": 0.3935,
+      "step": 2139
+    },
+    {
+      "epoch": 11.072258064516129,
+      "grad_norm": 0.4633094370365143,
+      "learning_rate": 0.0001261884183232498,
+      "loss": 0.3908,
+      "step": 2140
+    },
+    {
+      "epoch": 11.07741935483871,
+      "grad_norm": 0.5564885139465332,
+      "learning_rate": 0.00012615384615384615,
+      "loss": 0.4001,
+      "step": 2141
+    },
+    {
+      "epoch": 11.08258064516129,
+      "grad_norm": 0.4776894152164459,
+      "learning_rate": 0.00012611927398444253,
+      "loss": 0.3556,
+      "step": 2142
+    },
+    {
+      "epoch": 11.087741935483871,
+      "grad_norm": 0.5452290773391724,
+      "learning_rate": 0.0001260847018150389,
+      "loss": 0.3875,
+      "step": 2143
+    },
+    {
+      "epoch": 11.092903225806452,
+      "grad_norm": 0.553068995475769,
+      "learning_rate": 0.00012605012964563527,
+      "loss": 0.4035,
+      "step": 2144
+    },
+    {
+      "epoch": 11.098064516129032,
+      "grad_norm": 0.5249651074409485,
+      "learning_rate": 0.00012601555747623165,
+      "loss": 0.3754,
+      "step": 2145
+    },
+    {
+      "epoch": 11.103225806451613,
+      "grad_norm": 0.5175422430038452,
+      "learning_rate": 0.000125980985306828,
+      "loss": 0.4344,
+      "step": 2146
+    },
+    {
+      "epoch": 11.108387096774193,
+      "grad_norm": 0.4917783737182617,
+      "learning_rate": 0.00012594641313742438,
+      "loss": 0.3864,
+      "step": 2147
+    },
+    {
+      "epoch": 11.113548387096774,
+      "grad_norm": 0.5799666047096252,
+      "learning_rate": 0.00012591184096802076,
+      "loss": 0.3839,
+      "step": 2148
+    },
+    {
+      "epoch": 11.118709677419355,
+      "grad_norm": 0.5378075242042542,
+      "learning_rate": 0.00012587726879861711,
+      "loss": 0.375,
+      "step": 2149
+    },
+    {
+      "epoch": 11.123870967741935,
+      "grad_norm": 0.5397888422012329,
+      "learning_rate": 0.0001258426966292135,
+      "loss": 0.3979,
+      "step": 2150
+    },
+    {
+      "epoch": 11.129032258064516,
+      "grad_norm": 0.5405542850494385,
+      "learning_rate": 0.00012580812445980985,
+      "loss": 0.3627,
+      "step": 2151
+    },
+    {
+      "epoch": 11.134193548387097,
+      "grad_norm": 0.5652140378952026,
+      "learning_rate": 0.00012577355229040623,
+      "loss": 0.387,
+      "step": 2152
+    },
+    {
+      "epoch": 11.139354838709677,
+      "grad_norm": 0.5445713400840759,
+      "learning_rate": 0.0001257389801210026,
+      "loss": 0.4194,
+      "step": 2153
+    },
+    {
+      "epoch": 11.144516129032258,
+      "grad_norm": 0.4748550355434418,
+      "learning_rate": 0.00012570440795159896,
+      "loss": 0.411,
+      "step": 2154
+    },
+    {
+      "epoch": 11.14967741935484,
+      "grad_norm": 0.5876363515853882,
+      "learning_rate": 0.00012566983578219534,
+      "loss": 0.4391,
+      "step": 2155
+    },
+    {
+      "epoch": 11.154838709677419,
+      "grad_norm": 0.5709888339042664,
+      "learning_rate": 0.0001256352636127917,
+      "loss": 0.3888,
+      "step": 2156
+    },
+    {
+      "epoch": 11.16,
+      "grad_norm": 0.5856212973594666,
+      "learning_rate": 0.00012560069144338808,
+      "loss": 0.386,
+      "step": 2157
+    },
+    {
+      "epoch": 11.165161290322581,
+      "grad_norm": 0.49595773220062256,
+      "learning_rate": 0.00012556611927398446,
+      "loss": 0.4195,
+      "step": 2158
+    },
+    {
+      "epoch": 11.170322580645161,
+      "grad_norm": 0.5901684165000916,
+      "learning_rate": 0.0001255315471045808,
+      "loss": 0.4146,
+      "step": 2159
+    },
+    {
+      "epoch": 11.175483870967742,
+      "grad_norm": 0.5453752875328064,
+      "learning_rate": 0.0001254969749351772,
+      "loss": 0.3774,
+      "step": 2160
+    },
+    {
+      "epoch": 11.180645161290322,
+      "grad_norm": 0.5405909419059753,
+      "learning_rate": 0.00012546240276577357,
+      "loss": 0.3861,
+      "step": 2161
+    },
+    {
+      "epoch": 11.185806451612903,
+      "grad_norm": 0.5745472311973572,
+      "learning_rate": 0.00012542783059636993,
+      "loss": 0.3778,
+      "step": 2162
+    },
+    {
+      "epoch": 11.190967741935484,
+      "grad_norm": 0.5713327527046204,
+      "learning_rate": 0.0001253932584269663,
+      "loss": 0.3892,
+      "step": 2163
+    },
+    {
+      "epoch": 11.196129032258064,
+      "grad_norm": 0.5434615612030029,
+      "learning_rate": 0.00012535868625756266,
+      "loss": 0.3988,
+      "step": 2164
+    },
+    {
+      "epoch": 11.201290322580645,
+      "grad_norm": 0.5741232633590698,
+      "learning_rate": 0.00012532411408815904,
+      "loss": 0.4271,
+      "step": 2165
+    },
+    {
+      "epoch": 11.206451612903226,
+      "grad_norm": 0.5185933113098145,
+      "learning_rate": 0.00012528954191875542,
+      "loss": 0.3468,
+      "step": 2166
+    },
+    {
+      "epoch": 11.211612903225806,
+      "grad_norm": 0.5403662919998169,
+      "learning_rate": 0.00012525496974935177,
+      "loss": 0.3825,
+      "step": 2167
+    },
+    {
+      "epoch": 11.216774193548387,
+      "grad_norm": 0.542273759841919,
+      "learning_rate": 0.00012522039757994815,
+      "loss": 0.378,
+      "step": 2168
+    },
+    {
+      "epoch": 11.221935483870968,
+      "grad_norm": 0.5427550077438354,
+      "learning_rate": 0.0001251858254105445,
+      "loss": 0.4206,
+      "step": 2169
+    },
+    {
+      "epoch": 11.227096774193548,
+      "grad_norm": 0.5867548584938049,
+      "learning_rate": 0.0001251512532411409,
+      "loss": 0.4317,
+      "step": 2170
+    },
+    {
+      "epoch": 11.23225806451613,
+      "grad_norm": 0.5206836462020874,
+      "learning_rate": 0.00012511668107173727,
+      "loss": 0.3952,
+      "step": 2171
+    },
+    {
+      "epoch": 11.23741935483871,
+      "grad_norm": 0.5485924482345581,
+      "learning_rate": 0.00012508210890233362,
+      "loss": 0.3528,
+      "step": 2172
+    },
+    {
+      "epoch": 11.24258064516129,
+      "grad_norm": 0.49900388717651367,
+      "learning_rate": 0.00012504753673293,
+      "loss": 0.43,
+      "step": 2173
+    },
+    {
+      "epoch": 11.247741935483871,
+      "grad_norm": 0.6302635669708252,
+      "learning_rate": 0.00012501296456352636,
+      "loss": 0.4297,
+      "step": 2174
+    },
+    {
+      "epoch": 11.252903225806453,
+      "grad_norm": 0.49692201614379883,
+      "learning_rate": 0.00012497839239412274,
+      "loss": 0.3959,
+      "step": 2175
+    },
+    {
+      "epoch": 11.258064516129032,
+      "grad_norm": 0.539385199546814,
+      "learning_rate": 0.00012494382022471912,
+      "loss": 0.363,
+      "step": 2176
+    },
+    {
+      "epoch": 11.263225806451613,
+      "grad_norm": 0.5539966821670532,
+      "learning_rate": 0.00012490924805531547,
+      "loss": 0.4018,
+      "step": 2177
+    },
+    {
+      "epoch": 11.268387096774193,
+      "grad_norm": 0.5479066371917725,
+      "learning_rate": 0.00012487467588591185,
+      "loss": 0.4108,
+      "step": 2178
+    },
+    {
+      "epoch": 11.273548387096774,
+      "grad_norm": 0.5514193773269653,
+      "learning_rate": 0.00012484010371650823,
+      "loss": 0.3698,
+      "step": 2179
+    },
+    {
+      "epoch": 11.278709677419355,
+      "grad_norm": 0.5458917617797852,
+      "learning_rate": 0.00012480553154710459,
+      "loss": 0.3365,
+      "step": 2180
+    },
+    {
+      "epoch": 11.283870967741935,
+      "grad_norm": 0.5534675121307373,
+      "learning_rate": 0.00012477095937770097,
+      "loss": 0.4113,
+      "step": 2181
+    },
+    {
+      "epoch": 11.289032258064516,
+      "grad_norm": 0.6429570913314819,
+      "learning_rate": 0.00012473638720829732,
+      "loss": 0.4102,
+      "step": 2182
+    },
+    {
+      "epoch": 11.294193548387097,
+      "grad_norm": 0.5661647915840149,
+      "learning_rate": 0.0001247018150388937,
+      "loss": 0.4143,
+      "step": 2183
+    },
+    {
+      "epoch": 11.299354838709677,
+      "grad_norm": 0.5607682466506958,
+      "learning_rate": 0.00012466724286949008,
+      "loss": 0.4335,
+      "step": 2184
+    },
+    {
+      "epoch": 11.304516129032258,
+      "grad_norm": 0.6359783411026001,
+      "learning_rate": 0.00012463267070008643,
+      "loss": 0.4061,
+      "step": 2185
+    },
+    {
+      "epoch": 11.30967741935484,
+      "grad_norm": 0.6276160478591919,
+      "learning_rate": 0.00012459809853068281,
+      "loss": 0.4046,
+      "step": 2186
+    },
+    {
+      "epoch": 11.314838709677419,
+      "grad_norm": 0.5340253114700317,
+      "learning_rate": 0.00012456352636127917,
+      "loss": 0.3947,
+      "step": 2187
+    },
+    {
+      "epoch": 11.32,
+      "grad_norm": 0.5886175036430359,
+      "learning_rate": 0.00012452895419187555,
+      "loss": 0.3997,
+      "step": 2188
+    },
+    {
+      "epoch": 11.32516129032258,
+      "grad_norm": 0.5621709227561951,
+      "learning_rate": 0.00012449438202247193,
+      "loss": 0.382,
+      "step": 2189
+    },
+    {
+      "epoch": 11.330322580645161,
+      "grad_norm": 0.6730782389640808,
+      "learning_rate": 0.00012445980985306828,
+      "loss": 0.3749,
+      "step": 2190
+    },
+    {
+      "epoch": 11.335483870967742,
+      "grad_norm": 0.5230964422225952,
+      "learning_rate": 0.00012442523768366466,
+      "loss": 0.3898,
+      "step": 2191
+    },
+    {
+      "epoch": 11.340645161290322,
+      "grad_norm": 0.5654788613319397,
+      "learning_rate": 0.00012439066551426102,
+      "loss": 0.3772,
+      "step": 2192
+    },
+    {
+      "epoch": 11.345806451612903,
+      "grad_norm": 0.5238366723060608,
+      "learning_rate": 0.0001243560933448574,
+      "loss": 0.3549,
+      "step": 2193
+    },
+    {
+      "epoch": 11.350967741935484,
+      "grad_norm": 0.6045211553573608,
+      "learning_rate": 0.00012432152117545378,
+      "loss": 0.4154,
+      "step": 2194
+    },
+    {
+      "epoch": 11.356129032258064,
+      "grad_norm": 0.5453572869300842,
+      "learning_rate": 0.00012428694900605013,
+      "loss": 0.3843,
+      "step": 2195
+    },
+    {
+      "epoch": 11.361290322580645,
+      "grad_norm": 0.5936123132705688,
+      "learning_rate": 0.0001242523768366465,
+      "loss": 0.3824,
+      "step": 2196
+    },
+    {
+      "epoch": 11.366451612903226,
+      "grad_norm": 0.621993899345398,
+      "learning_rate": 0.00012421780466724286,
+      "loss": 0.3833,
+      "step": 2197
+    },
+    {
+      "epoch": 11.371612903225806,
+      "grad_norm": 0.606462299823761,
+      "learning_rate": 0.00012418323249783925,
+      "loss": 0.413,
+      "step": 2198
+    },
+    {
+      "epoch": 11.376774193548387,
+      "grad_norm": 0.567048192024231,
+      "learning_rate": 0.00012414866032843563,
+      "loss": 0.3735,
+      "step": 2199
+    },
+    {
+      "epoch": 11.381935483870969,
+      "grad_norm": 0.5745161175727844,
+      "learning_rate": 0.00012411408815903198,
+      "loss": 0.3598,
+      "step": 2200
+    },
+    {
+      "epoch": 11.387096774193548,
+      "grad_norm": 0.5705851316452026,
+      "learning_rate": 0.00012407951598962836,
+      "loss": 0.3678,
+      "step": 2201
+    },
+    {
+      "epoch": 11.39225806451613,
+      "grad_norm": 0.6047068238258362,
+      "learning_rate": 0.00012404494382022474,
+      "loss": 0.4319,
+      "step": 2202
+    },
+    {
+      "epoch": 11.397419354838709,
+      "grad_norm": 0.555722177028656,
+      "learning_rate": 0.0001240103716508211,
+      "loss": 0.4265,
+      "step": 2203
+    },
+    {
+      "epoch": 11.40258064516129,
+      "grad_norm": 0.5797570943832397,
+      "learning_rate": 0.00012397579948141747,
+      "loss": 0.4127,
+      "step": 2204
+    },
+    {
+      "epoch": 11.407741935483871,
+      "grad_norm": 0.5483680963516235,
+      "learning_rate": 0.00012394122731201383,
+      "loss": 0.3708,
+      "step": 2205
+    },
+    {
+      "epoch": 11.412903225806451,
+      "grad_norm": 0.5110373497009277,
+      "learning_rate": 0.0001239066551426102,
+      "loss": 0.34,
+      "step": 2206
+    },
+    {
+      "epoch": 11.418064516129032,
+      "grad_norm": 0.5539288520812988,
+      "learning_rate": 0.0001238720829732066,
+      "loss": 0.3932,
+      "step": 2207
+    },
+    {
+      "epoch": 11.423225806451613,
+      "grad_norm": 0.5593680143356323,
+      "learning_rate": 0.00012383751080380294,
+      "loss": 0.4264,
+      "step": 2208
+    },
+    {
+      "epoch": 11.428387096774193,
+      "grad_norm": 0.6567428708076477,
+      "learning_rate": 0.00012380293863439932,
+      "loss": 0.3966,
+      "step": 2209
+    },
+    {
+      "epoch": 11.433548387096774,
+      "grad_norm": 0.6507887244224548,
+      "learning_rate": 0.00012376836646499568,
+      "loss": 0.3745,
+      "step": 2210
+    },
+    {
+      "epoch": 11.438709677419356,
+      "grad_norm": 0.647455096244812,
+      "learning_rate": 0.00012373379429559206,
+      "loss": 0.4269,
+      "step": 2211
+    },
+    {
+      "epoch": 11.443870967741935,
+      "grad_norm": 0.5858089923858643,
+      "learning_rate": 0.00012369922212618844,
+      "loss": 0.4172,
+      "step": 2212
+    },
+    {
+      "epoch": 11.449032258064516,
+      "grad_norm": 0.543190598487854,
+      "learning_rate": 0.0001236646499567848,
+      "loss": 0.3747,
+      "step": 2213
+    },
+    {
+      "epoch": 11.454193548387098,
+      "grad_norm": 0.6007044315338135,
+      "learning_rate": 0.00012363007778738117,
+      "loss": 0.421,
+      "step": 2214
+    },
+    {
+      "epoch": 11.459354838709677,
+      "grad_norm": 0.6070182919502258,
+      "learning_rate": 0.00012359550561797752,
+      "loss": 0.4145,
+      "step": 2215
+    },
+    {
+      "epoch": 11.464516129032258,
+      "grad_norm": 0.5984455943107605,
+      "learning_rate": 0.0001235609334485739,
+      "loss": 0.4057,
+      "step": 2216
+    },
+    {
+      "epoch": 11.469677419354838,
+      "grad_norm": 0.6070668697357178,
+      "learning_rate": 0.00012352636127917029,
+      "loss": 0.3821,
+      "step": 2217
+    },
+    {
+      "epoch": 11.47483870967742,
+      "grad_norm": 0.5519592761993408,
+      "learning_rate": 0.00012349178910976664,
+      "loss": 0.4416,
+      "step": 2218
+    },
+    {
+      "epoch": 11.48,
+      "grad_norm": 0.5762091875076294,
+      "learning_rate": 0.00012345721694036302,
+      "loss": 0.4267,
+      "step": 2219
+    },
+    {
+      "epoch": 11.48516129032258,
+      "grad_norm": 0.6189960241317749,
+      "learning_rate": 0.0001234226447709594,
+      "loss": 0.42,
+      "step": 2220
+    },
+    {
+      "epoch": 11.490322580645161,
+      "grad_norm": 0.5764818787574768,
+      "learning_rate": 0.00012338807260155575,
+      "loss": 0.3475,
+      "step": 2221
+    },
+    {
+      "epoch": 11.495483870967742,
+      "grad_norm": 0.5902151465415955,
+      "learning_rate": 0.00012335350043215213,
+      "loss": 0.4567,
+      "step": 2222
+    },
+    {
+      "epoch": 11.500645161290322,
+      "grad_norm": 0.6039356589317322,
+      "learning_rate": 0.0001233189282627485,
+      "loss": 0.4241,
+      "step": 2223
+    },
+    {
+      "epoch": 11.505806451612903,
+      "grad_norm": 0.5601852536201477,
+      "learning_rate": 0.00012328435609334487,
+      "loss": 0.4091,
+      "step": 2224
+    },
+    {
+      "epoch": 11.510967741935485,
+      "grad_norm": 0.6661171317100525,
+      "learning_rate": 0.00012324978392394125,
+      "loss": 0.421,
+      "step": 2225
+    },
+    {
+      "epoch": 11.516129032258064,
+      "grad_norm": 0.5608654022216797,
+      "learning_rate": 0.0001232152117545376,
+      "loss": 0.3859,
+      "step": 2226
+    },
+    {
+      "epoch": 11.521290322580645,
+      "grad_norm": 0.5995846390724182,
+      "learning_rate": 0.00012318063958513398,
+      "loss": 0.3944,
+      "step": 2227
+    },
+    {
+      "epoch": 11.526451612903227,
+      "grad_norm": 0.5613486766815186,
+      "learning_rate": 0.00012314606741573034,
+      "loss": 0.3854,
+      "step": 2228
+    },
+    {
+      "epoch": 11.531612903225806,
+      "grad_norm": 0.5987265110015869,
+      "learning_rate": 0.0001231114952463267,
+      "loss": 0.4321,
+      "step": 2229
+    },
+    {
+      "epoch": 11.536774193548387,
+      "grad_norm": 0.7773928642272949,
+      "learning_rate": 0.0001230769230769231,
+      "loss": 0.4262,
+      "step": 2230
+    },
+    {
+      "epoch": 11.541935483870969,
+      "grad_norm": 0.6019839644432068,
+      "learning_rate": 0.00012304235090751945,
+      "loss": 0.3736,
+      "step": 2231
+    },
+    {
+      "epoch": 11.547096774193548,
+      "grad_norm": 0.5686100125312805,
+      "learning_rate": 0.00012300777873811583,
+      "loss": 0.36,
+      "step": 2232
+    },
+    {
+      "epoch": 11.55225806451613,
+      "grad_norm": 0.636902928352356,
+      "learning_rate": 0.00012297320656871218,
+      "loss": 0.4146,
+      "step": 2233
+    },
+    {
+      "epoch": 11.557419354838709,
+      "grad_norm": 0.5216076374053955,
+      "learning_rate": 0.00012293863439930857,
+      "loss": 0.4296,
+      "step": 2234
+    },
+    {
+      "epoch": 11.56258064516129,
+      "grad_norm": 0.62120521068573,
+      "learning_rate": 0.00012290406222990495,
+      "loss": 0.4331,
+      "step": 2235
+    },
+    {
+      "epoch": 11.567741935483872,
+      "grad_norm": 0.5977622866630554,
+      "learning_rate": 0.0001228694900605013,
+      "loss": 0.4079,
+      "step": 2236
+    },
+    {
+      "epoch": 11.572903225806451,
+      "grad_norm": 0.5864602327346802,
+      "learning_rate": 0.00012283491789109768,
+      "loss": 0.4154,
+      "step": 2237
+    },
+    {
+      "epoch": 11.578064516129032,
+      "grad_norm": 0.652046799659729,
+      "learning_rate": 0.00012280034572169403,
+      "loss": 0.4228,
+      "step": 2238
+    },
+    {
+      "epoch": 11.583225806451614,
+      "grad_norm": 0.6422173976898193,
+      "learning_rate": 0.00012276577355229041,
+      "loss": 0.3586,
+      "step": 2239
+    },
+    {
+      "epoch": 11.588387096774193,
+      "grad_norm": 0.6424571871757507,
+      "learning_rate": 0.0001227312013828868,
+      "loss": 0.4504,
+      "step": 2240
+    },
+    {
+      "epoch": 11.593548387096774,
+      "grad_norm": 0.5985169410705566,
+      "learning_rate": 0.00012269662921348315,
+      "loss": 0.3971,
+      "step": 2241
+    },
+    {
+      "epoch": 11.598709677419356,
+      "grad_norm": 0.5461825132369995,
+      "learning_rate": 0.0001226620570440795,
+      "loss": 0.3516,
+      "step": 2242
+    },
+    {
+      "epoch": 11.603870967741935,
+      "grad_norm": 0.5376850962638855,
+      "learning_rate": 0.0001226274848746759,
+      "loss": 0.3966,
+      "step": 2243
+    },
+    {
+      "epoch": 11.609032258064516,
+      "grad_norm": 0.5444831252098083,
+      "learning_rate": 0.00012259291270527226,
+      "loss": 0.3858,
+      "step": 2244
+    },
+    {
+      "epoch": 11.614193548387096,
+      "grad_norm": 0.6107810139656067,
+      "learning_rate": 0.00012255834053586864,
+      "loss": 0.4255,
+      "step": 2245
+    },
+    {
+      "epoch": 11.619354838709677,
+      "grad_norm": 0.6298990845680237,
+      "learning_rate": 0.000122523768366465,
+      "loss": 0.4349,
+      "step": 2246
+    },
+    {
+      "epoch": 11.624516129032259,
+      "grad_norm": 0.6217508316040039,
+      "learning_rate": 0.00012248919619706135,
+      "loss": 0.3915,
+      "step": 2247
+    },
+    {
+      "epoch": 11.629677419354838,
+      "grad_norm": 0.6392242312431335,
+      "learning_rate": 0.00012245462402765776,
+      "loss": 0.4303,
+      "step": 2248
+    },
+    {
+      "epoch": 11.63483870967742,
+      "grad_norm": 0.5903552174568176,
+      "learning_rate": 0.0001224200518582541,
+      "loss": 0.4483,
+      "step": 2249
+    },
+    {
+      "epoch": 11.64,
+      "grad_norm": 0.5759409070014954,
+      "learning_rate": 0.0001223854796888505,
+      "loss": 0.3407,
+      "step": 2250
+    },
+    {
+      "epoch": 11.64516129032258,
+      "grad_norm": 0.5347396731376648,
+      "learning_rate": 0.00012235090751944684,
+      "loss": 0.4031,
+      "step": 2251
+    },
+    {
+      "epoch": 11.650322580645161,
+      "grad_norm": 0.6101537942886353,
+      "learning_rate": 0.00012231633535004323,
+      "loss": 0.3916,
+      "step": 2252
+    },
+    {
+      "epoch": 11.655483870967743,
+      "grad_norm": 0.600766122341156,
+      "learning_rate": 0.0001222817631806396,
+      "loss": 0.3771,
+      "step": 2253
+    },
+    {
+      "epoch": 11.660645161290322,
+      "grad_norm": 0.6975557804107666,
+      "learning_rate": 0.00012224719101123596,
+      "loss": 0.4127,
+      "step": 2254
+    },
+    {
+      "epoch": 11.665806451612903,
+      "grad_norm": 0.6188974380493164,
+      "learning_rate": 0.00012221261884183234,
+      "loss": 0.431,
+      "step": 2255
+    },
+    {
+      "epoch": 11.670967741935485,
+      "grad_norm": 0.6419153809547424,
+      "learning_rate": 0.0001221780466724287,
+      "loss": 0.4274,
+      "step": 2256
+    },
+    {
+      "epoch": 11.676129032258064,
+      "grad_norm": 0.648235023021698,
+      "learning_rate": 0.00012214347450302507,
+      "loss": 0.4378,
+      "step": 2257
+    },
+    {
+      "epoch": 11.681290322580645,
+      "grad_norm": 0.5980153679847717,
+      "learning_rate": 0.00012210890233362145,
+      "loss": 0.4004,
+      "step": 2258
+    },
+    {
+      "epoch": 11.686451612903225,
+      "grad_norm": 0.6128827929496765,
+      "learning_rate": 0.0001220743301642178,
+      "loss": 0.4194,
+      "step": 2259
+    },
+    {
+      "epoch": 11.691612903225806,
+      "grad_norm": 0.6508981585502625,
+      "learning_rate": 0.00012203975799481417,
+      "loss": 0.4263,
+      "step": 2260
+    },
+    {
+      "epoch": 11.696774193548388,
+      "grad_norm": 0.6369768977165222,
+      "learning_rate": 0.00012200518582541054,
+      "loss": 0.4298,
+      "step": 2261
+    },
+    {
+      "epoch": 11.701935483870967,
+      "grad_norm": 0.7224717736244202,
+      "learning_rate": 0.00012197061365600692,
+      "loss": 0.4414,
+      "step": 2262
+    },
+    {
+      "epoch": 11.707096774193548,
+      "grad_norm": 0.5966154336929321,
+      "learning_rate": 0.00012193604148660329,
+      "loss": 0.3514,
+      "step": 2263
+    },
+    {
+      "epoch": 11.71225806451613,
+      "grad_norm": 0.720032274723053,
+      "learning_rate": 0.00012190146931719966,
+      "loss": 0.4262,
+      "step": 2264
+    },
+    {
+      "epoch": 11.717419354838709,
+      "grad_norm": 0.6187936067581177,
+      "learning_rate": 0.00012186689714779602,
+      "loss": 0.3834,
+      "step": 2265
+    },
+    {
+      "epoch": 11.72258064516129,
+      "grad_norm": 0.5973230004310608,
+      "learning_rate": 0.0001218323249783924,
+      "loss": 0.4266,
+      "step": 2266
+    },
+    {
+      "epoch": 11.727741935483872,
+      "grad_norm": 0.6463858485221863,
+      "learning_rate": 0.00012179775280898877,
+      "loss": 0.4414,
+      "step": 2267
+    },
+    {
+      "epoch": 11.732903225806451,
+      "grad_norm": 0.6448946595191956,
+      "learning_rate": 0.00012176318063958514,
+      "loss": 0.426,
+      "step": 2268
+    },
+    {
+      "epoch": 11.738064516129032,
+      "grad_norm": 0.6711769700050354,
+      "learning_rate": 0.0001217286084701815,
+      "loss": 0.4197,
+      "step": 2269
+    },
+    {
+      "epoch": 11.743225806451614,
+      "grad_norm": 0.6486469507217407,
+      "learning_rate": 0.00012169403630077787,
+      "loss": 0.427,
+      "step": 2270
+    },
+    {
+      "epoch": 11.748387096774193,
+      "grad_norm": 0.6407445073127747,
+      "learning_rate": 0.00012165946413137425,
+      "loss": 0.4118,
+      "step": 2271
+    },
+    {
+      "epoch": 11.753548387096775,
+      "grad_norm": 0.6526361107826233,
+      "learning_rate": 0.00012162489196197062,
+      "loss": 0.3888,
+      "step": 2272
+    },
+    {
+      "epoch": 11.758709677419354,
+      "grad_norm": 0.6114718317985535,
+      "learning_rate": 0.00012159031979256699,
+      "loss": 0.4197,
+      "step": 2273
+    },
+    {
+      "epoch": 11.763870967741935,
+      "grad_norm": 0.6784541010856628,
+      "learning_rate": 0.00012155574762316335,
+      "loss": 0.388,
+      "step": 2274
+    },
+    {
+      "epoch": 11.769032258064517,
+      "grad_norm": 0.6461766362190247,
+      "learning_rate": 0.00012152117545375973,
+      "loss": 0.4114,
+      "step": 2275
+    },
+    {
+      "epoch": 11.774193548387096,
+      "grad_norm": 0.5766343474388123,
+      "learning_rate": 0.0001214866032843561,
+      "loss": 0.3841,
+      "step": 2276
+    },
+    {
+      "epoch": 11.779354838709677,
+      "grad_norm": 0.6033735275268555,
+      "learning_rate": 0.00012145203111495247,
+      "loss": 0.3928,
+      "step": 2277
+    },
+    {
+      "epoch": 11.784516129032259,
+      "grad_norm": 0.6223620772361755,
+      "learning_rate": 0.00012141745894554883,
+      "loss": 0.407,
+      "step": 2278
+    },
+    {
+      "epoch": 11.789677419354838,
+      "grad_norm": 0.6550286412239075,
+      "learning_rate": 0.0001213828867761452,
+      "loss": 0.4454,
+      "step": 2279
+    },
+    {
+      "epoch": 11.79483870967742,
+      "grad_norm": 0.6335977911949158,
+      "learning_rate": 0.00012134831460674158,
+      "loss": 0.3888,
+      "step": 2280
+    },
+    {
+      "epoch": 11.8,
+      "grad_norm": 0.6886581182479858,
+      "learning_rate": 0.00012131374243733795,
+      "loss": 0.3662,
+      "step": 2281
+    },
+    {
+      "epoch": 11.80516129032258,
+      "grad_norm": 0.6327982544898987,
+      "learning_rate": 0.00012127917026793432,
+      "loss": 0.4297,
+      "step": 2282
+    },
+    {
+      "epoch": 11.810322580645161,
+      "grad_norm": 0.6746610999107361,
+      "learning_rate": 0.00012124459809853068,
+      "loss": 0.4325,
+      "step": 2283
+    },
+    {
+      "epoch": 11.815483870967743,
+      "grad_norm": 0.6002910137176514,
+      "learning_rate": 0.00012121002592912706,
+      "loss": 0.4131,
+      "step": 2284
+    },
+    {
+      "epoch": 11.820645161290322,
+      "grad_norm": 0.6444786787033081,
+      "learning_rate": 0.00012117545375972343,
+      "loss": 0.4172,
+      "step": 2285
+    },
+    {
+      "epoch": 11.825806451612904,
+      "grad_norm": 0.6081560850143433,
+      "learning_rate": 0.0001211408815903198,
+      "loss": 0.4126,
+      "step": 2286
+    },
+    {
+      "epoch": 11.830967741935483,
+      "grad_norm": 0.6242349147796631,
+      "learning_rate": 0.00012110630942091616,
+      "loss": 0.3632,
+      "step": 2287
+    },
+    {
+      "epoch": 11.836129032258064,
+      "grad_norm": 0.648755669593811,
+      "learning_rate": 0.00012107173725151253,
+      "loss": 0.3896,
+      "step": 2288
+    },
+    {
+      "epoch": 11.841290322580646,
+      "grad_norm": 0.6389471292495728,
+      "learning_rate": 0.00012103716508210891,
+      "loss": 0.3949,
+      "step": 2289
+    },
+    {
+      "epoch": 11.846451612903225,
+      "grad_norm": 0.6827386021614075,
+      "learning_rate": 0.00012100259291270528,
+      "loss": 0.3907,
+      "step": 2290
+    },
+    {
+      "epoch": 11.851612903225806,
+      "grad_norm": 0.5789780020713806,
+      "learning_rate": 0.00012096802074330165,
+      "loss": 0.4331,
+      "step": 2291
+    },
+    {
+      "epoch": 11.856774193548388,
+      "grad_norm": 0.6560531258583069,
+      "learning_rate": 0.00012093344857389801,
+      "loss": 0.4394,
+      "step": 2292
+    },
+    {
+      "epoch": 11.861935483870967,
+      "grad_norm": 0.5691995024681091,
+      "learning_rate": 0.0001208988764044944,
+      "loss": 0.4058,
+      "step": 2293
+    },
+    {
+      "epoch": 11.867096774193548,
+      "grad_norm": 0.6605562567710876,
+      "learning_rate": 0.00012086430423509076,
+      "loss": 0.4298,
+      "step": 2294
+    },
+    {
+      "epoch": 11.87225806451613,
+      "grad_norm": 0.6542829871177673,
+      "learning_rate": 0.00012082973206568713,
+      "loss": 0.3792,
+      "step": 2295
+    },
+    {
+      "epoch": 11.87741935483871,
+      "grad_norm": 0.6704761981964111,
+      "learning_rate": 0.0001207951598962835,
+      "loss": 0.4239,
+      "step": 2296
+    },
+    {
+      "epoch": 11.88258064516129,
+      "grad_norm": 0.632278323173523,
+      "learning_rate": 0.00012076058772687986,
+      "loss": 0.4282,
+      "step": 2297
+    },
+    {
+      "epoch": 11.88774193548387,
+      "grad_norm": 0.6744612455368042,
+      "learning_rate": 0.00012072601555747624,
+      "loss": 0.4123,
+      "step": 2298
+    },
+    {
+      "epoch": 11.892903225806451,
+      "grad_norm": 0.5981385707855225,
+      "learning_rate": 0.00012069144338807261,
+      "loss": 0.4106,
+      "step": 2299
+    },
+    {
+      "epoch": 11.898064516129033,
+      "grad_norm": 0.7206805944442749,
+      "learning_rate": 0.00012065687121866898,
+      "loss": 0.4171,
+      "step": 2300
+    },
+    {
+      "epoch": 11.903225806451612,
+      "grad_norm": 0.6514812707901001,
+      "learning_rate": 0.00012062229904926534,
+      "loss": 0.4171,
+      "step": 2301
+    },
+    {
+      "epoch": 11.908387096774193,
+      "grad_norm": 0.6302539110183716,
+      "learning_rate": 0.00012058772687986171,
+      "loss": 0.3903,
+      "step": 2302
+    },
+    {
+      "epoch": 11.913548387096775,
+      "grad_norm": 0.6798263788223267,
+      "learning_rate": 0.00012055315471045809,
+      "loss": 0.4349,
+      "step": 2303
+    },
+    {
+      "epoch": 11.918709677419354,
+      "grad_norm": 0.6544688940048218,
+      "learning_rate": 0.00012051858254105446,
+      "loss": 0.4426,
+      "step": 2304
+    },
+    {
+      "epoch": 11.923870967741935,
+      "grad_norm": 0.6104977130889893,
+      "learning_rate": 0.00012048401037165082,
+      "loss": 0.388,
+      "step": 2305
+    },
+    {
+      "epoch": 11.929032258064517,
+      "grad_norm": 0.6034760475158691,
+      "learning_rate": 0.00012044943820224719,
+      "loss": 0.4096,
+      "step": 2306
+    },
+    {
+      "epoch": 11.934193548387096,
+      "grad_norm": 0.6883524060249329,
+      "learning_rate": 0.00012041486603284357,
+      "loss": 0.4074,
+      "step": 2307
+    },
+    {
+      "epoch": 11.939354838709678,
+      "grad_norm": 0.6273317933082581,
+      "learning_rate": 0.00012038029386343994,
+      "loss": 0.3619,
+      "step": 2308
+    },
+    {
+      "epoch": 11.944516129032259,
+      "grad_norm": 0.5860844254493713,
+      "learning_rate": 0.0001203457216940363,
+      "loss": 0.3986,
+      "step": 2309
+    },
+    {
+      "epoch": 11.949677419354838,
+      "grad_norm": 0.6864508986473083,
+      "learning_rate": 0.00012031114952463267,
+      "loss": 0.4285,
+      "step": 2310
+    },
+    {
+      "epoch": 11.95483870967742,
+      "grad_norm": 0.6759065985679626,
+      "learning_rate": 0.00012027657735522904,
+      "loss": 0.3915,
+      "step": 2311
+    },
+    {
+      "epoch": 11.96,
+      "grad_norm": 0.6529413461685181,
+      "learning_rate": 0.00012024200518582542,
+      "loss": 0.4136,
+      "step": 2312
+    },
+    {
+      "epoch": 11.96516129032258,
+      "grad_norm": 0.7011410593986511,
+      "learning_rate": 0.00012020743301642179,
+      "loss": 0.4237,
+      "step": 2313
+    },
+    {
+      "epoch": 11.970322580645162,
+      "grad_norm": 0.8100919127464294,
+      "learning_rate": 0.00012017286084701815,
+      "loss": 0.3794,
+      "step": 2314
+    },
+    {
+      "epoch": 11.975483870967741,
+      "grad_norm": 0.665132999420166,
+      "learning_rate": 0.00012013828867761452,
+      "loss": 0.4561,
+      "step": 2315
+    },
+    {
+      "epoch": 11.980645161290322,
+      "grad_norm": 0.6105929017066956,
+      "learning_rate": 0.0001201037165082109,
+      "loss": 0.4056,
+      "step": 2316
+    },
+    {
+      "epoch": 11.985806451612904,
+      "grad_norm": 0.658210039138794,
+      "learning_rate": 0.00012006914433880727,
+      "loss": 0.4144,
+      "step": 2317
+    },
+    {
+      "epoch": 11.990967741935483,
+      "grad_norm": 0.592808723449707,
+      "learning_rate": 0.00012003457216940364,
+      "loss": 0.3566,
+      "step": 2318
+    },
+    {
+      "epoch": 11.996129032258064,
+      "grad_norm": 0.6468664407730103,
+      "learning_rate": 0.00012,
+      "loss": 0.3813,
+      "step": 2319
+    },
+    {
+      "epoch": 12.00258064516129,
+      "grad_norm": 0.9621816873550415,
+      "learning_rate": 0.00011996542783059637,
+      "loss": 0.6506,
+      "step": 2320
+    },
+    {
+      "epoch": 12.00258064516129,
+      "eval_loss": 2.375605583190918,
+      "eval_runtime": 21.3083,
+      "eval_samples_per_second": 3.707,
+      "eval_steps_per_second": 0.469,
+      "step": 2320
+    },
+    {
+      "epoch": 12.007741935483871,
+      "grad_norm": 0.44612178206443787,
+      "learning_rate": 0.00011993085566119275,
+      "loss": 0.36,
+      "step": 2321
+    },
+    {
+      "epoch": 12.012903225806452,
+      "grad_norm": 0.43981918692588806,
+      "learning_rate": 0.00011989628349178912,
+      "loss": 0.3681,
+      "step": 2322
+    },
+    {
+      "epoch": 12.018064516129032,
+      "grad_norm": 0.5119754672050476,
+      "learning_rate": 0.00011986171132238548,
+      "loss": 0.41,
+      "step": 2323
+    },
+    {
+      "epoch": 12.023225806451613,
+      "grad_norm": 0.5204293727874756,
+      "learning_rate": 0.00011982713915298185,
+      "loss": 0.4121,
+      "step": 2324
+    },
+    {
+      "epoch": 12.028387096774194,
+      "grad_norm": 0.4636760950088501,
+      "learning_rate": 0.00011979256698357823,
+      "loss": 0.371,
+      "step": 2325
+    },
+    {
+      "epoch": 12.033548387096774,
+      "grad_norm": 0.5979949235916138,
+      "learning_rate": 0.0001197579948141746,
+      "loss": 0.4035,
+      "step": 2326
+    },
+    {
+      "epoch": 12.038709677419355,
+      "grad_norm": 0.6185718178749084,
+      "learning_rate": 0.00011972342264477097,
+      "loss": 0.3923,
+      "step": 2327
+    },
+    {
+      "epoch": 12.043870967741935,
+      "grad_norm": 0.6728377342224121,
+      "learning_rate": 0.00011968885047536733,
+      "loss": 0.3958,
+      "step": 2328
+    },
+    {
+      "epoch": 12.049032258064516,
+      "grad_norm": 0.6238680481910706,
+      "learning_rate": 0.0001196542783059637,
+      "loss": 0.3863,
+      "step": 2329
+    },
+    {
+      "epoch": 12.054193548387097,
+      "grad_norm": 0.5559340119361877,
+      "learning_rate": 0.00011961970613656008,
+      "loss": 0.3742,
+      "step": 2330
+    },
+    {
+      "epoch": 12.059354838709677,
+      "grad_norm": 0.6345334649085999,
+      "learning_rate": 0.00011958513396715645,
+      "loss": 0.4227,
+      "step": 2331
+    },
+    {
+      "epoch": 12.064516129032258,
+      "grad_norm": 0.4302883744239807,
+      "learning_rate": 0.00011955056179775281,
+      "loss": 0.3593,
+      "step": 2332
+    },
+    {
+      "epoch": 12.06967741935484,
+      "grad_norm": 0.52857506275177,
+      "learning_rate": 0.00011951598962834918,
+      "loss": 0.3688,
+      "step": 2333
+    },
+    {
+      "epoch": 12.074838709677419,
+      "grad_norm": 0.5246421694755554,
+      "learning_rate": 0.00011948141745894555,
+      "loss": 0.377,
+      "step": 2334
+    },
+    {
+      "epoch": 12.08,
+      "grad_norm": 0.5018008947372437,
+      "learning_rate": 0.00011944684528954193,
+      "loss": 0.3926,
+      "step": 2335
+    },
+    {
+      "epoch": 12.085161290322581,
+      "grad_norm": 0.4689622223377228,
+      "learning_rate": 0.0001194122731201383,
+      "loss": 0.3701,
+      "step": 2336
+    },
+    {
+      "epoch": 12.09032258064516,
+      "grad_norm": 0.5142914056777954,
+      "learning_rate": 0.00011937770095073466,
+      "loss": 0.356,
+      "step": 2337
+    },
+    {
+      "epoch": 12.095483870967742,
+      "grad_norm": 0.5583852529525757,
+      "learning_rate": 0.00011934312878133103,
+      "loss": 0.3654,
+      "step": 2338
+    },
+    {
+      "epoch": 12.100645161290323,
+      "grad_norm": 0.658481240272522,
+      "learning_rate": 0.00011930855661192741,
+      "loss": 0.3851,
+      "step": 2339
+    },
+    {
+      "epoch": 12.105806451612903,
+      "grad_norm": 0.506711483001709,
+      "learning_rate": 0.00011927398444252378,
+      "loss": 0.4073,
+      "step": 2340
+    },
+    {
+      "epoch": 12.110967741935484,
+      "grad_norm": 1.5528112649917603,
+      "learning_rate": 0.00011923941227312014,
+      "loss": 0.3853,
+      "step": 2341
+    },
+    {
+      "epoch": 12.116129032258064,
+      "grad_norm": 0.5187411904335022,
+      "learning_rate": 0.00011920484010371651,
+      "loss": 0.3696,
+      "step": 2342
+    },
+    {
+      "epoch": 12.121290322580645,
+      "grad_norm": 0.5326524972915649,
+      "learning_rate": 0.00011917026793431288,
+      "loss": 0.3818,
+      "step": 2343
+    },
+    {
+      "epoch": 12.126451612903226,
+      "grad_norm": 0.6194540858268738,
+      "learning_rate": 0.00011913569576490926,
+      "loss": 0.42,
+      "step": 2344
+    },
+    {
+      "epoch": 12.131612903225806,
+      "grad_norm": 0.5322785973548889,
+      "learning_rate": 0.00011910112359550563,
+      "loss": 0.3952,
+      "step": 2345
+    },
+    {
+      "epoch": 12.136774193548387,
+      "grad_norm": 0.4859057068824768,
+      "learning_rate": 0.00011906655142610199,
+      "loss": 0.3355,
+      "step": 2346
+    },
+    {
+      "epoch": 12.141935483870968,
+      "grad_norm": 0.5154667496681213,
+      "learning_rate": 0.00011903197925669836,
+      "loss": 0.3563,
+      "step": 2347
+    },
+    {
+      "epoch": 12.147096774193548,
+      "grad_norm": 0.5266841053962708,
+      "learning_rate": 0.00011899740708729474,
+      "loss": 0.389,
+      "step": 2348
+    },
+    {
+      "epoch": 12.152258064516129,
+      "grad_norm": 0.5085238218307495,
+      "learning_rate": 0.00011896283491789111,
+      "loss": 0.3986,
+      "step": 2349
+    },
+    {
+      "epoch": 12.15741935483871,
+      "grad_norm": 0.6228361129760742,
+      "learning_rate": 0.00011892826274848747,
+      "loss": 0.3893,
+      "step": 2350
+    },
+    {
+      "epoch": 12.16258064516129,
+      "grad_norm": 0.5671336650848389,
+      "learning_rate": 0.00011889369057908384,
+      "loss": 0.4053,
+      "step": 2351
+    },
+    {
+      "epoch": 12.167741935483871,
+      "grad_norm": 0.48899152874946594,
+      "learning_rate": 0.00011885911840968021,
+      "loss": 0.4144,
+      "step": 2352
+    },
+    {
+      "epoch": 12.172903225806452,
+      "grad_norm": 0.667559802532196,
+      "learning_rate": 0.00011882454624027659,
+      "loss": 0.3808,
+      "step": 2353
+    },
+    {
+      "epoch": 12.178064516129032,
+      "grad_norm": 0.5053604245185852,
+      "learning_rate": 0.00011878997407087296,
+      "loss": 0.3914,
+      "step": 2354
+    },
+    {
+      "epoch": 12.183225806451613,
+      "grad_norm": 0.5779446363449097,
+      "learning_rate": 0.00011875540190146932,
+      "loss": 0.387,
+      "step": 2355
+    },
+    {
+      "epoch": 12.188387096774193,
+      "grad_norm": 0.6320012211799622,
+      "learning_rate": 0.00011872082973206569,
+      "loss": 0.4089,
+      "step": 2356
+    },
+    {
+      "epoch": 12.193548387096774,
+      "grad_norm": 0.5122958421707153,
+      "learning_rate": 0.00011868625756266207,
+      "loss": 0.361,
+      "step": 2357
+    },
+    {
+      "epoch": 12.198709677419355,
+      "grad_norm": 0.4526745676994324,
+      "learning_rate": 0.00011865168539325844,
+      "loss": 0.3674,
+      "step": 2358
+    },
+    {
+      "epoch": 12.203870967741935,
+      "grad_norm": 0.5370525121688843,
+      "learning_rate": 0.0001186171132238548,
+      "loss": 0.4267,
+      "step": 2359
+    },
+    {
+      "epoch": 12.209032258064516,
+      "grad_norm": 0.5748568177223206,
+      "learning_rate": 0.00011858254105445117,
+      "loss": 0.3562,
+      "step": 2360
+    },
+    {
+      "epoch": 12.214193548387097,
+      "grad_norm": 0.5398175716400146,
+      "learning_rate": 0.00011854796888504754,
+      "loss": 0.3392,
+      "step": 2361
+    },
+    {
+      "epoch": 12.219354838709677,
+      "grad_norm": 0.6187689900398254,
+      "learning_rate": 0.00011851339671564392,
+      "loss": 0.372,
+      "step": 2362
+    },
+    {
+      "epoch": 12.224516129032258,
+      "grad_norm": 0.5433279275894165,
+      "learning_rate": 0.00011847882454624029,
+      "loss": 0.3671,
+      "step": 2363
+    },
+    {
+      "epoch": 12.22967741935484,
+      "grad_norm": 0.5176102519035339,
+      "learning_rate": 0.00011844425237683665,
+      "loss": 0.3719,
+      "step": 2364
+    },
+    {
+      "epoch": 12.234838709677419,
+      "grad_norm": 0.5345560908317566,
+      "learning_rate": 0.00011840968020743302,
+      "loss": 0.3977,
+      "step": 2365
+    },
+    {
+      "epoch": 12.24,
+      "grad_norm": 0.5357674956321716,
+      "learning_rate": 0.0001183751080380294,
+      "loss": 0.3916,
+      "step": 2366
+    },
+    {
+      "epoch": 12.245161290322581,
+      "grad_norm": 0.6013228297233582,
+      "learning_rate": 0.00011834053586862577,
+      "loss": 0.4218,
+      "step": 2367
+    },
+    {
+      "epoch": 12.250322580645161,
+      "grad_norm": 0.5482919216156006,
+      "learning_rate": 0.00011830596369922213,
+      "loss": 0.4042,
+      "step": 2368
+    },
+    {
+      "epoch": 12.255483870967742,
+      "grad_norm": 0.5560780763626099,
+      "learning_rate": 0.0001182713915298185,
+      "loss": 0.3984,
+      "step": 2369
+    },
+    {
+      "epoch": 12.260645161290322,
+      "grad_norm": 0.49483537673950195,
+      "learning_rate": 0.00011823681936041487,
+      "loss": 0.3696,
+      "step": 2370
+    },
+    {
+      "epoch": 12.265806451612903,
+      "grad_norm": 0.4803698658943176,
+      "learning_rate": 0.00011820224719101125,
+      "loss": 0.3151,
+      "step": 2371
+    },
+    {
+      "epoch": 12.270967741935484,
+      "grad_norm": 0.500063955783844,
+      "learning_rate": 0.00011816767502160762,
+      "loss": 0.3678,
+      "step": 2372
+    },
+    {
+      "epoch": 12.276129032258064,
+      "grad_norm": 0.5557253360748291,
+      "learning_rate": 0.00011813310285220398,
+      "loss": 0.3806,
+      "step": 2373
+    },
+    {
+      "epoch": 12.281290322580645,
+      "grad_norm": 0.5334653854370117,
+      "learning_rate": 0.00011809853068280035,
+      "loss": 0.391,
+      "step": 2374
+    },
+    {
+      "epoch": 12.286451612903226,
+      "grad_norm": 0.526862621307373,
+      "learning_rate": 0.00011806395851339672,
+      "loss": 0.4004,
+      "step": 2375
+    },
+    {
+      "epoch": 12.291612903225806,
+      "grad_norm": 0.5489553809165955,
+      "learning_rate": 0.0001180293863439931,
+      "loss": 0.3787,
+      "step": 2376
+    },
+    {
+      "epoch": 12.296774193548387,
+      "grad_norm": 0.7950393557548523,
+      "learning_rate": 0.00011799481417458946,
+      "loss": 0.4059,
+      "step": 2377
+    },
+    {
+      "epoch": 12.301935483870968,
+      "grad_norm": 0.49101442098617554,
+      "learning_rate": 0.00011796024200518583,
+      "loss": 0.3467,
+      "step": 2378
+    },
+    {
+      "epoch": 12.307096774193548,
+      "grad_norm": 0.5536491870880127,
+      "learning_rate": 0.0001179256698357822,
+      "loss": 0.3951,
+      "step": 2379
+    },
+    {
+      "epoch": 12.31225806451613,
+      "grad_norm": 0.4751553237438202,
+      "learning_rate": 0.00011789109766637858,
+      "loss": 0.3759,
+      "step": 2380
+    },
+    {
+      "epoch": 12.31741935483871,
+      "grad_norm": 0.5717628002166748,
+      "learning_rate": 0.00011785652549697495,
+      "loss": 0.4084,
+      "step": 2381
+    },
+    {
+      "epoch": 12.32258064516129,
+      "grad_norm": 0.5907983183860779,
+      "learning_rate": 0.00011782195332757131,
+      "loss": 0.4151,
+      "step": 2382
+    },
+    {
+      "epoch": 12.327741935483871,
+      "grad_norm": 0.5947985649108887,
+      "learning_rate": 0.00011778738115816768,
+      "loss": 0.4074,
+      "step": 2383
+    },
+    {
+      "epoch": 12.33290322580645,
+      "grad_norm": 0.5846248865127563,
+      "learning_rate": 0.00011775280898876405,
+      "loss": 0.4026,
+      "step": 2384
+    },
+    {
+      "epoch": 12.338064516129032,
+      "grad_norm": 0.6320284605026245,
+      "learning_rate": 0.00011771823681936043,
+      "loss": 0.4114,
+      "step": 2385
+    },
+    {
+      "epoch": 12.343225806451613,
+      "grad_norm": 0.5035812854766846,
+      "learning_rate": 0.0001176836646499568,
+      "loss": 0.3946,
+      "step": 2386
+    },
+    {
+      "epoch": 12.348387096774193,
+      "grad_norm": 0.5838691592216492,
+      "learning_rate": 0.00011764909248055316,
+      "loss": 0.3824,
+      "step": 2387
+    },
+    {
+      "epoch": 12.353548387096774,
+      "grad_norm": 0.5940775275230408,
+      "learning_rate": 0.00011761452031114953,
+      "loss": 0.4084,
+      "step": 2388
+    },
+    {
+      "epoch": 12.358709677419355,
+      "grad_norm": 0.5215787291526794,
+      "learning_rate": 0.00011757994814174591,
+      "loss": 0.3721,
+      "step": 2389
+    },
+    {
+      "epoch": 12.363870967741935,
+      "grad_norm": 0.5472300052642822,
+      "learning_rate": 0.00011754537597234228,
+      "loss": 0.3642,
+      "step": 2390
+    },
+    {
+      "epoch": 12.369032258064516,
+      "grad_norm": 0.5600304007530212,
+      "learning_rate": 0.00011751080380293864,
+      "loss": 0.3562,
+      "step": 2391
+    },
+    {
+      "epoch": 12.374193548387098,
+      "grad_norm": 0.620419979095459,
+      "learning_rate": 0.00011747623163353501,
+      "loss": 0.4397,
+      "step": 2392
+    },
+    {
+      "epoch": 12.379354838709677,
+      "grad_norm": 0.5490419864654541,
+      "learning_rate": 0.00011744165946413138,
+      "loss": 0.32,
+      "step": 2393
+    },
+    {
+      "epoch": 12.384516129032258,
+      "grad_norm": 0.5533275604248047,
+      "learning_rate": 0.00011740708729472776,
+      "loss": 0.3843,
+      "step": 2394
+    },
+    {
+      "epoch": 12.38967741935484,
+      "grad_norm": 0.5564206838607788,
+      "learning_rate": 0.00011737251512532412,
+      "loss": 0.4237,
+      "step": 2395
+    },
+    {
+      "epoch": 12.394838709677419,
+      "grad_norm": 0.5394004583358765,
+      "learning_rate": 0.00011733794295592049,
+      "loss": 0.3563,
+      "step": 2396
+    },
+    {
+      "epoch": 12.4,
+      "grad_norm": 0.5873191952705383,
+      "learning_rate": 0.00011730337078651686,
+      "loss": 0.382,
+      "step": 2397
+    },
+    {
+      "epoch": 12.40516129032258,
+      "grad_norm": 0.5913571119308472,
+      "learning_rate": 0.00011726879861711324,
+      "loss": 0.3968,
+      "step": 2398
+    },
+    {
+      "epoch": 12.410322580645161,
+      "grad_norm": 0.6425137519836426,
+      "learning_rate": 0.0001172342264477096,
+      "loss": 0.4141,
+      "step": 2399
+    },
+    {
+      "epoch": 12.415483870967742,
+      "grad_norm": 0.5675520300865173,
+      "learning_rate": 0.00011719965427830597,
+      "loss": 0.4093,
+      "step": 2400
+    },
+    {
+      "epoch": 12.420645161290322,
+      "grad_norm": 0.5191066861152649,
+      "learning_rate": 0.00011716508210890234,
+      "loss": 0.4249,
+      "step": 2401
+    },
+    {
+      "epoch": 12.425806451612903,
+      "grad_norm": 0.6223769783973694,
+      "learning_rate": 0.0001171305099394987,
+      "loss": 0.4139,
+      "step": 2402
+    },
+    {
+      "epoch": 12.430967741935484,
+      "grad_norm": 0.6177443861961365,
+      "learning_rate": 0.00011709593777009509,
+      "loss": 0.395,
+      "step": 2403
+    },
+    {
+      "epoch": 12.436129032258064,
+      "grad_norm": 0.6284211277961731,
+      "learning_rate": 0.00011706136560069145,
+      "loss": 0.3994,
+      "step": 2404
+    },
+    {
+      "epoch": 12.441290322580645,
+      "grad_norm": 0.49472999572753906,
+      "learning_rate": 0.00011702679343128782,
+      "loss": 0.3468,
+      "step": 2405
+    },
+    {
+      "epoch": 12.446451612903227,
+      "grad_norm": 0.5472164154052734,
+      "learning_rate": 0.00011699222126188419,
+      "loss": 0.3938,
+      "step": 2406
+    },
+    {
+      "epoch": 12.451612903225806,
+      "grad_norm": 0.6204166412353516,
+      "learning_rate": 0.00011695764909248054,
+      "loss": 0.4103,
+      "step": 2407
+    },
+    {
+      "epoch": 12.456774193548387,
+      "grad_norm": 0.540561854839325,
+      "learning_rate": 0.00011692307692307694,
+      "loss": 0.3253,
+      "step": 2408
+    },
+    {
+      "epoch": 12.461935483870967,
+      "grad_norm": 0.46690133213996887,
+      "learning_rate": 0.0001168885047536733,
+      "loss": 0.3117,
+      "step": 2409
+    },
+    {
+      "epoch": 12.467096774193548,
+      "grad_norm": 0.5706825852394104,
+      "learning_rate": 0.00011685393258426967,
+      "loss": 0.3517,
+      "step": 2410
+    },
+    {
+      "epoch": 12.47225806451613,
+      "grad_norm": 0.5669806599617004,
+      "learning_rate": 0.00011681936041486604,
+      "loss": 0.3906,
+      "step": 2411
+    },
+    {
+      "epoch": 12.477419354838709,
+      "grad_norm": 0.5264591574668884,
+      "learning_rate": 0.00011678478824546242,
+      "loss": 0.4005,
+      "step": 2412
+    },
+    {
+      "epoch": 12.48258064516129,
+      "grad_norm": 0.5995815992355347,
+      "learning_rate": 0.00011675021607605878,
+      "loss": 0.384,
+      "step": 2413
+    },
+    {
+      "epoch": 12.487741935483871,
+      "grad_norm": 0.5538098812103271,
+      "learning_rate": 0.00011671564390665515,
+      "loss": 0.3901,
+      "step": 2414
+    },
+    {
+      "epoch": 12.492903225806451,
+      "grad_norm": 0.5556499361991882,
+      "learning_rate": 0.00011668107173725152,
+      "loss": 0.3785,
+      "step": 2415
+    },
+    {
+      "epoch": 12.498064516129032,
+      "grad_norm": 0.5480387210845947,
+      "learning_rate": 0.00011664649956784787,
+      "loss": 0.3729,
+      "step": 2416
+    },
+    {
+      "epoch": 12.503225806451614,
+      "grad_norm": 0.6020554900169373,
+      "learning_rate": 0.00011661192739844427,
+      "loss": 0.4024,
+      "step": 2417
+    },
+    {
+      "epoch": 12.508387096774193,
+      "grad_norm": 0.583176851272583,
+      "learning_rate": 0.00011657735522904063,
+      "loss": 0.3843,
+      "step": 2418
+    },
+    {
+      "epoch": 12.513548387096774,
+      "grad_norm": 0.6603333353996277,
+      "learning_rate": 0.000116542783059637,
+      "loss": 0.4057,
+      "step": 2419
+    },
+    {
+      "epoch": 12.518709677419356,
+      "grad_norm": 0.538905680179596,
+      "learning_rate": 0.00011650821089023337,
+      "loss": 0.4042,
+      "step": 2420
+    },
+    {
+      "epoch": 12.523870967741935,
+      "grad_norm": 0.5485799312591553,
+      "learning_rate": 0.00011647363872082975,
+      "loss": 0.4072,
+      "step": 2421
+    },
+    {
+      "epoch": 12.529032258064516,
+      "grad_norm": 0.58696448802948,
+      "learning_rate": 0.00011643906655142611,
+      "loss": 0.4139,
+      "step": 2422
+    },
+    {
+      "epoch": 12.534193548387098,
+      "grad_norm": 0.5847681164741516,
+      "learning_rate": 0.00011640449438202248,
+      "loss": 0.3894,
+      "step": 2423
+    },
+    {
+      "epoch": 12.539354838709677,
+      "grad_norm": 0.6083982586860657,
+      "learning_rate": 0.00011636992221261885,
+      "loss": 0.3858,
+      "step": 2424
+    },
+    {
+      "epoch": 12.544516129032258,
+      "grad_norm": 0.6072269082069397,
+      "learning_rate": 0.0001163353500432152,
+      "loss": 0.3958,
+      "step": 2425
+    },
+    {
+      "epoch": 12.549677419354838,
+      "grad_norm": 0.5206655859947205,
+      "learning_rate": 0.0001163007778738116,
+      "loss": 0.3246,
+      "step": 2426
+    },
+    {
+      "epoch": 12.55483870967742,
+      "grad_norm": 0.5609920620918274,
+      "learning_rate": 0.00011626620570440796,
+      "loss": 0.3947,
+      "step": 2427
+    },
+    {
+      "epoch": 12.56,
+      "grad_norm": 1.124870777130127,
+      "learning_rate": 0.00011623163353500433,
+      "loss": 0.3657,
+      "step": 2428
+    },
+    {
+      "epoch": 12.56516129032258,
+      "grad_norm": 0.5994611978530884,
+      "learning_rate": 0.0001161970613656007,
+      "loss": 0.3907,
+      "step": 2429
+    },
+    {
+      "epoch": 12.570322580645161,
+      "grad_norm": 0.5956276059150696,
+      "learning_rate": 0.00011616248919619708,
+      "loss": 0.3794,
+      "step": 2430
+    },
+    {
+      "epoch": 12.575483870967743,
+      "grad_norm": 0.5408291816711426,
+      "learning_rate": 0.00011612791702679344,
+      "loss": 0.3845,
+      "step": 2431
+    },
+    {
+      "epoch": 12.580645161290322,
+      "grad_norm": 0.6004382371902466,
+      "learning_rate": 0.00011609334485738981,
+      "loss": 0.411,
+      "step": 2432
+    },
+    {
+      "epoch": 12.585806451612903,
+      "grad_norm": 0.6426616907119751,
+      "learning_rate": 0.00011605877268798618,
+      "loss": 0.4019,
+      "step": 2433
+    },
+    {
+      "epoch": 12.590967741935485,
+      "grad_norm": 0.6926738023757935,
+      "learning_rate": 0.00011602420051858253,
+      "loss": 0.3842,
+      "step": 2434
+    },
+    {
+      "epoch": 12.596129032258064,
+      "grad_norm": 0.6545553207397461,
+      "learning_rate": 0.00011598962834917893,
+      "loss": 0.3903,
+      "step": 2435
+    },
+    {
+      "epoch": 12.601290322580645,
+      "grad_norm": 0.5481811165809631,
+      "learning_rate": 0.00011595505617977529,
+      "loss": 0.3838,
+      "step": 2436
+    },
+    {
+      "epoch": 12.606451612903225,
+      "grad_norm": 0.638611376285553,
+      "learning_rate": 0.00011592048401037166,
+      "loss": 0.3763,
+      "step": 2437
+    },
+    {
+      "epoch": 12.611612903225806,
+      "grad_norm": 0.7013685703277588,
+      "learning_rate": 0.00011588591184096803,
+      "loss": 0.4138,
+      "step": 2438
+    },
+    {
+      "epoch": 12.616774193548387,
+      "grad_norm": 0.5939841866493225,
+      "learning_rate": 0.00011585133967156441,
+      "loss": 0.4157,
+      "step": 2439
+    },
+    {
+      "epoch": 12.621935483870967,
+      "grad_norm": 0.6315990090370178,
+      "learning_rate": 0.00011581676750216077,
+      "loss": 0.403,
+      "step": 2440
+    },
+    {
+      "epoch": 12.627096774193548,
+      "grad_norm": 0.663118302822113,
+      "learning_rate": 0.00011578219533275714,
+      "loss": 0.423,
+      "step": 2441
+    },
+    {
+      "epoch": 12.63225806451613,
+      "grad_norm": 0.6356132626533508,
+      "learning_rate": 0.00011574762316335351,
+      "loss": 0.415,
+      "step": 2442
+    },
+    {
+      "epoch": 12.637419354838709,
+      "grad_norm": 0.5745171308517456,
+      "learning_rate": 0.00011571305099394986,
+      "loss": 0.4126,
+      "step": 2443
+    },
+    {
+      "epoch": 12.64258064516129,
+      "grad_norm": 0.6649137139320374,
+      "learning_rate": 0.00011567847882454626,
+      "loss": 0.3985,
+      "step": 2444
+    },
+    {
+      "epoch": 12.647741935483872,
+      "grad_norm": 0.6353330016136169,
+      "learning_rate": 0.00011564390665514262,
+      "loss": 0.3997,
+      "step": 2445
+    },
+    {
+      "epoch": 12.652903225806451,
+      "grad_norm": 0.6020027995109558,
+      "learning_rate": 0.00011560933448573899,
+      "loss": 0.3489,
+      "step": 2446
+    },
+    {
+      "epoch": 12.658064516129032,
+      "grad_norm": 0.6462099552154541,
+      "learning_rate": 0.00011557476231633536,
+      "loss": 0.4318,
+      "step": 2447
+    },
+    {
+      "epoch": 12.663225806451614,
+      "grad_norm": 0.5892651677131653,
+      "learning_rate": 0.00011554019014693171,
+      "loss": 0.381,
+      "step": 2448
+    },
+    {
+      "epoch": 12.668387096774193,
+      "grad_norm": 0.6480687260627747,
+      "learning_rate": 0.0001155056179775281,
+      "loss": 0.3982,
+      "step": 2449
+    },
+    {
+      "epoch": 12.673548387096774,
+      "grad_norm": 0.6642018556594849,
+      "learning_rate": 0.00011547104580812447,
+      "loss": 0.4093,
+      "step": 2450
+    },
+    {
+      "epoch": 12.678709677419356,
+      "grad_norm": 0.54340660572052,
+      "learning_rate": 0.00011543647363872084,
+      "loss": 0.3972,
+      "step": 2451
+    },
+    {
+      "epoch": 12.683870967741935,
+      "grad_norm": 0.5970606803894043,
+      "learning_rate": 0.00011540190146931719,
+      "loss": 0.3873,
+      "step": 2452
+    },
+    {
+      "epoch": 12.689032258064517,
+      "grad_norm": 0.65812748670578,
+      "learning_rate": 0.00011536732929991359,
+      "loss": 0.412,
+      "step": 2453
+    },
+    {
+      "epoch": 12.694193548387096,
+      "grad_norm": 0.5064092874526978,
+      "learning_rate": 0.00011533275713050995,
+      "loss": 0.3644,
+      "step": 2454
+    },
+    {
+      "epoch": 12.699354838709677,
+      "grad_norm": 0.6443904638290405,
+      "learning_rate": 0.00011529818496110632,
+      "loss": 0.4071,
+      "step": 2455
+    },
+    {
+      "epoch": 12.704516129032259,
+      "grad_norm": 0.6286991834640503,
+      "learning_rate": 0.00011526361279170267,
+      "loss": 0.3897,
+      "step": 2456
+    },
+    {
+      "epoch": 12.709677419354838,
+      "grad_norm": 0.6404409408569336,
+      "learning_rate": 0.00011522904062229904,
+      "loss": 0.3931,
+      "step": 2457
+    },
+    {
+      "epoch": 12.71483870967742,
+      "grad_norm": 0.5889267921447754,
+      "learning_rate": 0.00011519446845289543,
+      "loss": 0.3605,
+      "step": 2458
+    },
+    {
+      "epoch": 12.72,
+      "grad_norm": 0.6314874887466431,
+      "learning_rate": 0.0001151598962834918,
+      "loss": 0.3778,
+      "step": 2459
+    },
+    {
+      "epoch": 12.72516129032258,
+      "grad_norm": 0.6305519938468933,
+      "learning_rate": 0.00011512532411408817,
+      "loss": 0.38,
+      "step": 2460
+    },
+    {
+      "epoch": 12.730322580645161,
+      "grad_norm": 0.6364988088607788,
+      "learning_rate": 0.00011509075194468452,
+      "loss": 0.3934,
+      "step": 2461
+    },
+    {
+      "epoch": 12.735483870967743,
+      "grad_norm": 0.6032713055610657,
+      "learning_rate": 0.00011505617977528092,
+      "loss": 0.3956,
+      "step": 2462
+    },
+    {
+      "epoch": 12.740645161290322,
+      "grad_norm": 0.5983800888061523,
+      "learning_rate": 0.00011502160760587728,
+      "loss": 0.3916,
+      "step": 2463
+    },
+    {
+      "epoch": 12.745806451612903,
+      "grad_norm": 0.6473186612129211,
+      "learning_rate": 0.00011498703543647365,
+      "loss": 0.4015,
+      "step": 2464
+    },
+    {
+      "epoch": 12.750967741935483,
+      "grad_norm": 2.975687265396118,
+      "learning_rate": 0.00011495246326707,
+      "loss": 0.3642,
+      "step": 2465
+    },
+    {
+      "epoch": 12.756129032258064,
+      "grad_norm": 0.5789406895637512,
+      "learning_rate": 0.00011491789109766637,
+      "loss": 0.3078,
+      "step": 2466
+    },
+    {
+      "epoch": 12.761290322580646,
+      "grad_norm": 0.603856086730957,
+      "learning_rate": 0.00011488331892826276,
+      "loss": 0.3613,
+      "step": 2467
+    },
+    {
+      "epoch": 12.766451612903225,
+      "grad_norm": 0.5685442686080933,
+      "learning_rate": 0.00011484874675885913,
+      "loss": 0.4048,
+      "step": 2468
+    },
+    {
+      "epoch": 12.771612903225806,
+      "grad_norm": 0.5826041102409363,
+      "learning_rate": 0.0001148141745894555,
+      "loss": 0.3657,
+      "step": 2469
+    },
+    {
+      "epoch": 12.776774193548388,
+      "grad_norm": 0.6135759949684143,
+      "learning_rate": 0.00011477960242005185,
+      "loss": 0.3904,
+      "step": 2470
+    },
+    {
+      "epoch": 12.781935483870967,
+      "grad_norm": 0.7327951788902283,
+      "learning_rate": 0.00011474503025064825,
+      "loss": 0.4105,
+      "step": 2471
+    },
+    {
+      "epoch": 12.787096774193548,
+      "grad_norm": 0.6851863265037537,
+      "learning_rate": 0.00011471045808124461,
+      "loss": 0.3996,
+      "step": 2472
+    },
+    {
+      "epoch": 12.79225806451613,
+      "grad_norm": 0.6059955358505249,
+      "learning_rate": 0.00011467588591184098,
+      "loss": 0.4172,
+      "step": 2473
+    },
+    {
+      "epoch": 12.79741935483871,
+      "grad_norm": 0.6543750166893005,
+      "learning_rate": 0.00011464131374243733,
+      "loss": 0.3786,
+      "step": 2474
+    },
+    {
+      "epoch": 12.80258064516129,
+      "grad_norm": 0.6433500647544861,
+      "learning_rate": 0.0001146067415730337,
+      "loss": 0.3954,
+      "step": 2475
+    },
+    {
+      "epoch": 12.807741935483872,
+      "grad_norm": 1.6961069107055664,
+      "learning_rate": 0.0001145721694036301,
+      "loss": 0.4149,
+      "step": 2476
+    },
+    {
+      "epoch": 12.812903225806451,
+      "grad_norm": 0.6527287364006042,
+      "learning_rate": 0.00011453759723422646,
+      "loss": 0.4279,
+      "step": 2477
+    },
+    {
+      "epoch": 12.818064516129033,
+      "grad_norm": 0.699590802192688,
+      "learning_rate": 0.00011450302506482283,
+      "loss": 0.4296,
+      "step": 2478
+    },
+    {
+      "epoch": 12.823225806451614,
+      "grad_norm": 0.694675087928772,
+      "learning_rate": 0.00011446845289541918,
+      "loss": 0.3931,
+      "step": 2479
+    },
+    {
+      "epoch": 12.828387096774193,
+      "grad_norm": 0.6256546974182129,
+      "learning_rate": 0.00011443388072601555,
+      "loss": 0.4106,
+      "step": 2480
+    },
+    {
+      "epoch": 12.833548387096775,
+      "grad_norm": 0.6571013927459717,
+      "learning_rate": 0.00011439930855661194,
+      "loss": 0.3573,
+      "step": 2481
+    },
+    {
+      "epoch": 12.838709677419354,
+      "grad_norm": 0.6555888652801514,
+      "learning_rate": 0.00011436473638720831,
+      "loss": 0.3983,
+      "step": 2482
+    },
+    {
+      "epoch": 12.843870967741935,
+      "grad_norm": 0.671787679195404,
+      "learning_rate": 0.00011433016421780466,
+      "loss": 0.3655,
+      "step": 2483
+    },
+    {
+      "epoch": 12.849032258064517,
+      "grad_norm": 0.6500824689865112,
+      "learning_rate": 0.00011429559204840103,
+      "loss": 0.403,
+      "step": 2484
+    },
+    {
+      "epoch": 12.854193548387096,
+      "grad_norm": 0.6709276437759399,
+      "learning_rate": 0.00011426101987899742,
+      "loss": 0.3899,
+      "step": 2485
+    },
+    {
+      "epoch": 12.859354838709677,
+      "grad_norm": 0.7322316765785217,
+      "learning_rate": 0.00011422644770959379,
+      "loss": 0.3892,
+      "step": 2486
+    },
+    {
+      "epoch": 12.864516129032259,
+      "grad_norm": 0.7675609588623047,
+      "learning_rate": 0.00011419187554019016,
+      "loss": 0.3984,
+      "step": 2487
+    },
+    {
+      "epoch": 12.869677419354838,
+      "grad_norm": 0.6345077753067017,
+      "learning_rate": 0.00011415730337078651,
+      "loss": 0.3645,
+      "step": 2488
+    },
+    {
+      "epoch": 12.87483870967742,
+      "grad_norm": 0.6407454609870911,
+      "learning_rate": 0.00011412273120138288,
+      "loss": 0.4177,
+      "step": 2489
+    },
+    {
+      "epoch": 12.88,
+      "grad_norm": 0.7305061221122742,
+      "learning_rate": 0.00011408815903197927,
+      "loss": 0.4035,
+      "step": 2490
+    },
+    {
+      "epoch": 12.88516129032258,
+      "grad_norm": 0.6148202419281006,
+      "learning_rate": 0.00011405358686257564,
+      "loss": 0.3978,
+      "step": 2491
+    },
+    {
+      "epoch": 12.890322580645162,
+      "grad_norm": 0.6016846895217896,
+      "learning_rate": 0.00011401901469317199,
+      "loss": 0.3752,
+      "step": 2492
+    },
+    {
+      "epoch": 12.895483870967741,
+      "grad_norm": 0.6535977125167847,
+      "learning_rate": 0.00011398444252376836,
+      "loss": 0.4285,
+      "step": 2493
+    },
+    {
+      "epoch": 12.900645161290322,
+      "grad_norm": 0.6632303595542908,
+      "learning_rate": 0.00011394987035436475,
+      "loss": 0.4023,
+      "step": 2494
+    },
+    {
+      "epoch": 12.905806451612904,
+      "grad_norm": 0.5986599922180176,
+      "learning_rate": 0.00011391529818496112,
+      "loss": 0.3589,
+      "step": 2495
+    },
+    {
+      "epoch": 12.910967741935483,
+      "grad_norm": 0.6673604846000671,
+      "learning_rate": 0.00011388072601555749,
+      "loss": 0.4095,
+      "step": 2496
+    },
+    {
+      "epoch": 12.916129032258064,
+      "grad_norm": 0.6250585913658142,
+      "learning_rate": 0.00011384615384615384,
+      "loss": 0.3886,
+      "step": 2497
+    },
+    {
+      "epoch": 12.921290322580646,
+      "grad_norm": 0.7693644165992737,
+      "learning_rate": 0.00011381158167675021,
+      "loss": 0.3918,
+      "step": 2498
+    },
+    {
+      "epoch": 12.926451612903225,
+      "grad_norm": 0.7224003672599792,
+      "learning_rate": 0.0001137770095073466,
+      "loss": 0.4018,
+      "step": 2499
+    },
+    {
+      "epoch": 12.931612903225806,
+      "grad_norm": 0.6882739067077637,
+      "learning_rate": 0.00011374243733794297,
+      "loss": 0.3895,
+      "step": 2500
+    },
+    {
+      "epoch": 12.936774193548388,
+      "grad_norm": 0.6337284445762634,
+      "learning_rate": 0.00011370786516853932,
+      "loss": 0.3742,
+      "step": 2501
+    },
+    {
+      "epoch": 12.941935483870967,
+      "grad_norm": 0.5995379686355591,
+      "learning_rate": 0.00011367329299913569,
+      "loss": 0.3518,
+      "step": 2502
+    },
+    {
+      "epoch": 12.947096774193549,
+      "grad_norm": 0.7505373358726501,
+      "learning_rate": 0.00011363872082973208,
+      "loss": 0.423,
+      "step": 2503
+    },
+    {
+      "epoch": 12.95225806451613,
+      "grad_norm": 0.6821178197860718,
+      "learning_rate": 0.00011360414866032845,
+      "loss": 0.3919,
+      "step": 2504
+    },
+    {
+      "epoch": 12.95741935483871,
+      "grad_norm": 0.7154505252838135,
+      "learning_rate": 0.00011356957649092482,
+      "loss": 0.4066,
+      "step": 2505
+    },
+    {
+      "epoch": 12.96258064516129,
+      "grad_norm": 0.6975933313369751,
+      "learning_rate": 0.00011353500432152117,
+      "loss": 0.3902,
+      "step": 2506
+    },
+    {
+      "epoch": 12.967741935483872,
+      "grad_norm": 0.5885283350944519,
+      "learning_rate": 0.00011350043215211754,
+      "loss": 0.3904,
+      "step": 2507
+    },
+    {
+      "epoch": 12.972903225806451,
+      "grad_norm": 0.6269501447677612,
+      "learning_rate": 0.00011346585998271393,
+      "loss": 0.4102,
+      "step": 2508
+    },
+    {
+      "epoch": 12.978064516129033,
+      "grad_norm": 0.6816064119338989,
+      "learning_rate": 0.0001134312878133103,
+      "loss": 0.4341,
+      "step": 2509
+    },
+    {
+      "epoch": 12.983225806451612,
+      "grad_norm": 0.6691199541091919,
+      "learning_rate": 0.00011339671564390665,
+      "loss": 0.3855,
+      "step": 2510
+    },
+    {
+      "epoch": 12.988387096774193,
+      "grad_norm": 0.6491720676422119,
+      "learning_rate": 0.00011336214347450302,
+      "loss": 0.3873,
+      "step": 2511
+    },
+    {
+      "epoch": 12.993548387096775,
+      "grad_norm": 0.6963995099067688,
+      "learning_rate": 0.00011332757130509941,
+      "loss": 0.415,
+      "step": 2512
+    },
+    {
+      "epoch": 12.998709677419354,
+      "grad_norm": 1.0249119997024536,
+      "learning_rate": 0.00011329299913569578,
+      "loss": 0.5958,
+      "step": 2513
+    },
+    {
+      "epoch": 13.005161290322581,
+      "grad_norm": 0.40266481041908264,
+      "learning_rate": 0.00011325842696629215,
+      "loss": 0.3503,
+      "step": 2514
+    },
+    {
+      "epoch": 13.01032258064516,
+      "grad_norm": 0.7908182740211487,
+      "learning_rate": 0.0001132238547968885,
+      "loss": 0.3835,
+      "step": 2515
+    },
+    {
+      "epoch": 13.015483870967742,
+      "grad_norm": 0.44175201654434204,
+      "learning_rate": 0.00011318928262748487,
+      "loss": 0.3789,
+      "step": 2516
+    },
+    {
+      "epoch": 13.020645161290323,
+      "grad_norm": 0.5299575924873352,
+      "learning_rate": 0.00011315471045808126,
+      "loss": 0.3715,
+      "step": 2517
+    },
+    {
+      "epoch": 13.025806451612903,
+      "grad_norm": 0.5370601415634155,
+      "learning_rate": 0.00011312013828867763,
+      "loss": 0.3773,
+      "step": 2518
+    },
+    {
+      "epoch": 13.030967741935484,
+      "grad_norm": 0.6225398778915405,
+      "learning_rate": 0.00011308556611927398,
+      "loss": 0.3515,
+      "step": 2519
+    },
+    {
+      "epoch": 13.036129032258064,
+      "grad_norm": 0.6598086357116699,
+      "learning_rate": 0.00011305099394987035,
+      "loss": 0.3721,
+      "step": 2520
+    },
+    {
+      "epoch": 13.041290322580645,
+      "grad_norm": 0.5911020636558533,
+      "learning_rate": 0.00011301642178046672,
+      "loss": 0.3432,
+      "step": 2521
+    },
+    {
+      "epoch": 13.046451612903226,
+      "grad_norm": 0.5885916352272034,
+      "learning_rate": 0.00011298184961106311,
+      "loss": 0.3598,
+      "step": 2522
+    },
+    {
+      "epoch": 13.051612903225806,
+      "grad_norm": 0.6376885175704956,
+      "learning_rate": 0.00011294727744165948,
+      "loss": 0.388,
+      "step": 2523
+    },
+    {
+      "epoch": 13.056774193548387,
+      "grad_norm": 0.548151433467865,
+      "learning_rate": 0.00011291270527225583,
+      "loss": 0.3575,
+      "step": 2524
+    },
+    {
+      "epoch": 13.061935483870968,
+      "grad_norm": 0.5196736454963684,
+      "learning_rate": 0.0001128781331028522,
+      "loss": 0.3795,
+      "step": 2525
+    },
+    {
+      "epoch": 13.067096774193548,
+      "grad_norm": 0.5900920033454895,
+      "learning_rate": 0.00011284356093344859,
+      "loss": 0.3793,
+      "step": 2526
+    },
+    {
+      "epoch": 13.072258064516129,
+      "grad_norm": 0.5807023644447327,
+      "learning_rate": 0.00011280898876404496,
+      "loss": 0.3624,
+      "step": 2527
+    },
+    {
+      "epoch": 13.07741935483871,
+      "grad_norm": 0.5367271304130554,
+      "learning_rate": 0.00011277441659464131,
+      "loss": 0.3887,
+      "step": 2528
+    },
+    {
+      "epoch": 13.08258064516129,
+      "grad_norm": 0.5641703009605408,
+      "learning_rate": 0.00011273984442523768,
+      "loss": 0.4134,
+      "step": 2529
+    },
+    {
+      "epoch": 13.087741935483871,
+      "grad_norm": 0.5918639302253723,
+      "learning_rate": 0.00011270527225583405,
+      "loss": 0.3873,
+      "step": 2530
+    },
+    {
+      "epoch": 13.092903225806452,
+      "grad_norm": 0.526766836643219,
+      "learning_rate": 0.00011267070008643044,
+      "loss": 0.3661,
+      "step": 2531
+    },
+    {
+      "epoch": 13.098064516129032,
+      "grad_norm": 0.5986608266830444,
+      "learning_rate": 0.00011263612791702681,
+      "loss": 0.4,
+      "step": 2532
+    },
+    {
+      "epoch": 13.103225806451613,
+      "grad_norm": 0.5225208401679993,
+      "learning_rate": 0.00011260155574762316,
+      "loss": 0.3856,
+      "step": 2533
+    },
+    {
+      "epoch": 13.108387096774193,
+      "grad_norm": 0.5107784271240234,
+      "learning_rate": 0.00011256698357821953,
+      "loss": 0.3681,
+      "step": 2534
+    },
+    {
+      "epoch": 13.113548387096774,
+      "grad_norm": 0.5421147346496582,
+      "learning_rate": 0.00011253241140881592,
+      "loss": 0.3444,
+      "step": 2535
+    },
+    {
+      "epoch": 13.118709677419355,
+      "grad_norm": 0.6043881773948669,
+      "learning_rate": 0.00011249783923941229,
+      "loss": 0.3644,
+      "step": 2536
+    },
+    {
+      "epoch": 13.123870967741935,
+      "grad_norm": 0.5405078530311584,
+      "learning_rate": 0.00011246326707000864,
+      "loss": 0.3716,
+      "step": 2537
+    },
+    {
+      "epoch": 13.129032258064516,
+      "grad_norm": 0.5122734904289246,
+      "learning_rate": 0.00011242869490060501,
+      "loss": 0.3378,
+      "step": 2538
+    },
+    {
+      "epoch": 13.134193548387097,
+      "grad_norm": 0.6231010556221008,
+      "learning_rate": 0.00011239412273120138,
+      "loss": 0.385,
+      "step": 2539
+    },
+    {
+      "epoch": 13.139354838709677,
+      "grad_norm": 0.599037230014801,
+      "learning_rate": 0.00011235955056179777,
+      "loss": 0.3843,
+      "step": 2540
+    },
+    {
+      "epoch": 13.144516129032258,
+      "grad_norm": 0.6373411417007446,
+      "learning_rate": 0.00011232497839239414,
+      "loss": 0.3747,
+      "step": 2541
+    },
+    {
+      "epoch": 13.14967741935484,
+      "grad_norm": 0.6157592535018921,
+      "learning_rate": 0.00011229040622299049,
+      "loss": 0.3683,
+      "step": 2542
+    },
+    {
+      "epoch": 13.154838709677419,
+      "grad_norm": 0.5772956609725952,
+      "learning_rate": 0.00011225583405358686,
+      "loss": 0.3843,
+      "step": 2543
+    },
+    {
+      "epoch": 13.16,
+      "grad_norm": 0.5116007328033447,
+      "learning_rate": 0.00011222126188418325,
+      "loss": 0.3802,
+      "step": 2544
+    },
+    {
+      "epoch": 13.165161290322581,
+      "grad_norm": 0.6959611773490906,
+      "learning_rate": 0.00011218668971477962,
+      "loss": 0.3379,
+      "step": 2545
+    },
+    {
+      "epoch": 13.170322580645161,
+      "grad_norm": 0.5233417749404907,
+      "learning_rate": 0.00011215211754537597,
+      "loss": 0.3455,
+      "step": 2546
+    },
+    {
+      "epoch": 13.175483870967742,
+      "grad_norm": 0.5422375202178955,
+      "learning_rate": 0.00011211754537597234,
+      "loss": 0.3789,
+      "step": 2547
+    },
+    {
+      "epoch": 13.180645161290322,
+      "grad_norm": 0.5277494788169861,
+      "learning_rate": 0.0001120829732065687,
+      "loss": 0.3549,
+      "step": 2548
+    },
+    {
+      "epoch": 13.185806451612903,
+      "grad_norm": 0.6404826045036316,
+      "learning_rate": 0.0001120484010371651,
+      "loss": 0.3807,
+      "step": 2549
+    },
+    {
+      "epoch": 13.190967741935484,
+      "grad_norm": 0.5980280637741089,
+      "learning_rate": 0.00011201382886776147,
+      "loss": 0.3926,
+      "step": 2550
+    },
+    {
+      "epoch": 13.196129032258064,
+      "grad_norm": 0.5758792161941528,
+      "learning_rate": 0.00011197925669835782,
+      "loss": 0.3888,
+      "step": 2551
+    },
+    {
+      "epoch": 13.201290322580645,
+      "grad_norm": 0.59283846616745,
+      "learning_rate": 0.00011194468452895419,
+      "loss": 0.3434,
+      "step": 2552
+    },
+    {
+      "epoch": 13.206451612903226,
+      "grad_norm": 0.5428313612937927,
+      "learning_rate": 0.00011191011235955056,
+      "loss": 0.4115,
+      "step": 2553
+    },
+    {
+      "epoch": 13.211612903225806,
+      "grad_norm": 0.5652146935462952,
+      "learning_rate": 0.00011187554019014695,
+      "loss": 0.3671,
+      "step": 2554
+    },
+    {
+      "epoch": 13.216774193548387,
+      "grad_norm": 0.5134239792823792,
+      "learning_rate": 0.0001118409680207433,
+      "loss": 0.3444,
+      "step": 2555
+    },
+    {
+      "epoch": 13.221935483870968,
+      "grad_norm": 0.5365434885025024,
+      "learning_rate": 0.00011180639585133967,
+      "loss": 0.3712,
+      "step": 2556
+    },
+    {
+      "epoch": 13.227096774193548,
+      "grad_norm": 0.5070716142654419,
+      "learning_rate": 0.00011177182368193604,
+      "loss": 0.3129,
+      "step": 2557
+    },
+    {
+      "epoch": 13.23225806451613,
+      "grad_norm": 0.6349546313285828,
+      "learning_rate": 0.00011173725151253243,
+      "loss": 0.369,
+      "step": 2558
+    },
+    {
+      "epoch": 13.23741935483871,
+      "grad_norm": 0.6009499430656433,
+      "learning_rate": 0.0001117026793431288,
+      "loss": 0.3887,
+      "step": 2559
+    },
+    {
+      "epoch": 13.24258064516129,
+      "grad_norm": 0.5744321942329407,
+      "learning_rate": 0.00011166810717372515,
+      "loss": 0.3819,
+      "step": 2560
+    },
+    {
+      "epoch": 13.247741935483871,
+      "grad_norm": 0.5499552488327026,
+      "learning_rate": 0.00011163353500432152,
+      "loss": 0.323,
+      "step": 2561
+    },
+    {
+      "epoch": 13.252903225806453,
+      "grad_norm": 0.5839877128601074,
+      "learning_rate": 0.00011159896283491789,
+      "loss": 0.3375,
+      "step": 2562
+    },
+    {
+      "epoch": 13.258064516129032,
+      "grad_norm": 0.6078647971153259,
+      "learning_rate": 0.00011156439066551428,
+      "loss": 0.3928,
+      "step": 2563
+    },
+    {
+      "epoch": 13.263225806451613,
+      "grad_norm": 0.5815805196762085,
+      "learning_rate": 0.00011152981849611063,
+      "loss": 0.3582,
+      "step": 2564
+    },
+    {
+      "epoch": 13.268387096774193,
+      "grad_norm": 0.5452626943588257,
+      "learning_rate": 0.000111495246326707,
+      "loss": 0.3451,
+      "step": 2565
+    },
+    {
+      "epoch": 13.273548387096774,
+      "grad_norm": 0.4948217570781708,
+      "learning_rate": 0.00011146067415730337,
+      "loss": 0.377,
+      "step": 2566
+    },
+    {
+      "epoch": 13.278709677419355,
+      "grad_norm": 0.5101603865623474,
+      "learning_rate": 0.00011142610198789976,
+      "loss": 0.332,
+      "step": 2567
+    },
+    {
+      "epoch": 13.283870967741935,
+      "grad_norm": 0.5704203844070435,
+      "learning_rate": 0.00011139152981849611,
+      "loss": 0.367,
+      "step": 2568
+    },
+    {
+      "epoch": 13.289032258064516,
+      "grad_norm": 0.6869714260101318,
+      "learning_rate": 0.00011135695764909248,
+      "loss": 0.4011,
+      "step": 2569
+    },
+    {
+      "epoch": 13.294193548387097,
+      "grad_norm": 0.5994212031364441,
+      "learning_rate": 0.00011132238547968885,
+      "loss": 0.3667,
+      "step": 2570
+    },
+    {
+      "epoch": 13.299354838709677,
+      "grad_norm": 0.5831963419914246,
+      "learning_rate": 0.00011128781331028522,
+      "loss": 0.3518,
+      "step": 2571
+    },
+    {
+      "epoch": 13.304516129032258,
+      "grad_norm": 0.5146107077598572,
+      "learning_rate": 0.00011125324114088161,
+      "loss": 0.3537,
+      "step": 2572
+    },
+    {
+      "epoch": 13.30967741935484,
+      "grad_norm": 0.5488823652267456,
+      "learning_rate": 0.00011121866897147796,
+      "loss": 0.3215,
+      "step": 2573
+    },
+    {
+      "epoch": 13.314838709677419,
+      "grad_norm": 0.5881301164627075,
+      "learning_rate": 0.00011118409680207433,
+      "loss": 0.3847,
+      "step": 2574
+    },
+    {
+      "epoch": 13.32,
+      "grad_norm": 0.594140350818634,
+      "learning_rate": 0.0001111495246326707,
+      "loss": 0.3814,
+      "step": 2575
+    },
+    {
+      "epoch": 13.32516129032258,
+      "grad_norm": 0.6266553997993469,
+      "learning_rate": 0.00011111495246326709,
+      "loss": 0.3981,
+      "step": 2576
+    },
+    {
+      "epoch": 13.330322580645161,
+      "grad_norm": 0.5200210809707642,
+      "learning_rate": 0.00011108038029386344,
+      "loss": 0.332,
+      "step": 2577
+    },
+    {
+      "epoch": 13.335483870967742,
+      "grad_norm": 0.5554406642913818,
+      "learning_rate": 0.00011104580812445981,
+      "loss": 0.3681,
+      "step": 2578
+    },
+    {
+      "epoch": 13.340645161290322,
+      "grad_norm": 0.5384899377822876,
+      "learning_rate": 0.00011101123595505618,
+      "loss": 0.3796,
+      "step": 2579
+    },
+    {
+      "epoch": 13.345806451612903,
+      "grad_norm": 0.5564590692520142,
+      "learning_rate": 0.00011097666378565255,
+      "loss": 0.3731,
+      "step": 2580
+    },
+    {
+      "epoch": 13.350967741935484,
+      "grad_norm": 0.5017397999763489,
+      "learning_rate": 0.00011094209161624894,
+      "loss": 0.3585,
+      "step": 2581
+    },
+    {
+      "epoch": 13.356129032258064,
+      "grad_norm": 0.5645277500152588,
+      "learning_rate": 0.00011090751944684529,
+      "loss": 0.3965,
+      "step": 2582
+    },
+    {
+      "epoch": 13.361290322580645,
+      "grad_norm": 0.6172767281532288,
+      "learning_rate": 0.00011087294727744166,
+      "loss": 0.3713,
+      "step": 2583
+    },
+    {
+      "epoch": 13.366451612903226,
+      "grad_norm": 0.6329189538955688,
+      "learning_rate": 0.00011083837510803803,
+      "loss": 0.4064,
+      "step": 2584
+    },
+    {
+      "epoch": 13.371612903225806,
+      "grad_norm": 0.5739006400108337,
+      "learning_rate": 0.00011080380293863442,
+      "loss": 0.3769,
+      "step": 2585
+    },
+    {
+      "epoch": 13.376774193548387,
+      "grad_norm": 0.6020433902740479,
+      "learning_rate": 0.00011076923076923077,
+      "loss": 0.3962,
+      "step": 2586
+    },
+    {
+      "epoch": 13.381935483870969,
+      "grad_norm": 0.5600905418395996,
+      "learning_rate": 0.00011073465859982714,
+      "loss": 0.332,
+      "step": 2587
+    },
+    {
+      "epoch": 13.387096774193548,
+      "grad_norm": 0.6037775874137878,
+      "learning_rate": 0.00011070008643042351,
+      "loss": 0.3892,
+      "step": 2588
+    },
+    {
+      "epoch": 13.39225806451613,
+      "grad_norm": 0.5980088114738464,
+      "learning_rate": 0.00011066551426101988,
+      "loss": 0.3428,
+      "step": 2589
+    },
+    {
+      "epoch": 13.397419354838709,
+      "grad_norm": 0.5533635020256042,
+      "learning_rate": 0.00011063094209161627,
+      "loss": 0.3826,
+      "step": 2590
+    },
+    {
+      "epoch": 13.40258064516129,
+      "grad_norm": 0.5838978886604309,
+      "learning_rate": 0.00011059636992221262,
+      "loss": 0.3776,
+      "step": 2591
+    },
+    {
+      "epoch": 13.407741935483871,
+      "grad_norm": 0.539648175239563,
+      "learning_rate": 0.00011056179775280899,
+      "loss": 0.3199,
+      "step": 2592
+    },
+    {
+      "epoch": 13.412903225806451,
+      "grad_norm": 0.6010555624961853,
+      "learning_rate": 0.00011052722558340536,
+      "loss": 0.402,
+      "step": 2593
+    },
+    {
+      "epoch": 13.418064516129032,
+      "grad_norm": 0.6949271559715271,
+      "learning_rate": 0.00011049265341400172,
+      "loss": 0.4116,
+      "step": 2594
+    },
+    {
+      "epoch": 13.423225806451613,
+      "grad_norm": 0.5701996088027954,
+      "learning_rate": 0.0001104580812445981,
+      "loss": 0.4045,
+      "step": 2595
+    },
+    {
+      "epoch": 13.428387096774193,
+      "grad_norm": 0.543228268623352,
+      "learning_rate": 0.00011042350907519447,
+      "loss": 0.3873,
+      "step": 2596
+    },
+    {
+      "epoch": 13.433548387096774,
+      "grad_norm": 0.6302754878997803,
+      "learning_rate": 0.00011038893690579084,
+      "loss": 0.3655,
+      "step": 2597
+    },
+    {
+      "epoch": 13.438709677419356,
+      "grad_norm": 0.6277347803115845,
+      "learning_rate": 0.0001103543647363872,
+      "loss": 0.3977,
+      "step": 2598
+    },
+    {
+      "epoch": 13.443870967741935,
+      "grad_norm": 0.587343692779541,
+      "learning_rate": 0.0001103197925669836,
+      "loss": 0.368,
+      "step": 2599
+    },
+    {
+      "epoch": 13.449032258064516,
+      "grad_norm": 0.5482541918754578,
+      "learning_rate": 0.00011028522039757995,
+      "loss": 0.3798,
+      "step": 2600
+    },
+    {
+      "epoch": 13.454193548387098,
+      "grad_norm": 0.5151814818382263,
+      "learning_rate": 0.00011025064822817632,
+      "loss": 0.3191,
+      "step": 2601
+    },
+    {
+      "epoch": 13.459354838709677,
+      "grad_norm": 0.5933248400688171,
+      "learning_rate": 0.00011021607605877269,
+      "loss": 0.3889,
+      "step": 2602
+    },
+    {
+      "epoch": 13.464516129032258,
+      "grad_norm": 0.6394025087356567,
+      "learning_rate": 0.00011018150388936905,
+      "loss": 0.391,
+      "step": 2603
+    },
+    {
+      "epoch": 13.469677419354838,
+      "grad_norm": 0.5680641531944275,
+      "learning_rate": 0.00011014693171996543,
+      "loss": 0.3609,
+      "step": 2604
+    },
+    {
+      "epoch": 13.47483870967742,
+      "grad_norm": 0.6017460227012634,
+      "learning_rate": 0.0001101123595505618,
+      "loss": 0.3712,
+      "step": 2605
+    },
+    {
+      "epoch": 13.48,
+      "grad_norm": 0.5822615027427673,
+      "learning_rate": 0.00011007778738115817,
+      "loss": 0.3679,
+      "step": 2606
+    },
+    {
+      "epoch": 13.48516129032258,
+      "grad_norm": 0.5656458735466003,
+      "learning_rate": 0.00011004321521175454,
+      "loss": 0.3525,
+      "step": 2607
+    },
+    {
+      "epoch": 13.490322580645161,
+      "grad_norm": 0.6925727725028992,
+      "learning_rate": 0.00011000864304235093,
+      "loss": 0.3879,
+      "step": 2608
+    },
+    {
+      "epoch": 13.495483870967742,
+      "grad_norm": 0.5161901712417603,
+      "learning_rate": 0.00010997407087294728,
+      "loss": 0.3478,
+      "step": 2609
+    },
+    {
+      "epoch": 13.500645161290322,
+      "grad_norm": 0.6952807307243347,
+      "learning_rate": 0.00010993949870354365,
+      "loss": 0.3894,
+      "step": 2610
+    },
+    {
+      "epoch": 13.500645161290322,
+      "eval_loss": 2.497584104537964,
+      "eval_runtime": 21.3155,
+      "eval_samples_per_second": 3.706,
+      "eval_steps_per_second": 0.469,
+      "step": 2610
+    },
+    {
+      "epoch": 13.505806451612903,
+      "grad_norm": 0.569427490234375,
+      "learning_rate": 0.00010990492653414002,
+      "loss": 0.3745,
+      "step": 2611
+    },
+    {
+      "epoch": 13.510967741935485,
+      "grad_norm": 0.588301956653595,
+      "learning_rate": 0.00010987035436473638,
+      "loss": 0.3567,
+      "step": 2612
+    },
+    {
+      "epoch": 13.516129032258064,
+      "grad_norm": 0.5581251382827759,
+      "learning_rate": 0.00010983578219533276,
+      "loss": 0.3644,
+      "step": 2613
+    },
+    {
+      "epoch": 13.521290322580645,
+      "grad_norm": 0.6199673414230347,
+      "learning_rate": 0.00010980121002592913,
+      "loss": 0.3555,
+      "step": 2614
+    },
+    {
+      "epoch": 13.526451612903227,
+      "grad_norm": 0.6247883439064026,
+      "learning_rate": 0.0001097666378565255,
+      "loss": 0.3867,
+      "step": 2615
+    },
+    {
+      "epoch": 13.531612903225806,
+      "grad_norm": 0.70013028383255,
+      "learning_rate": 0.00010973206568712186,
+      "loss": 0.3345,
+      "step": 2616
+    },
+    {
+      "epoch": 13.536774193548387,
+      "grad_norm": 0.621616780757904,
+      "learning_rate": 0.00010969749351771826,
+      "loss": 0.3848,
+      "step": 2617
+    },
+    {
+      "epoch": 13.541935483870969,
+      "grad_norm": 0.6110267639160156,
+      "learning_rate": 0.00010966292134831461,
+      "loss": 0.3742,
+      "step": 2618
+    },
+    {
+      "epoch": 13.547096774193548,
+      "grad_norm": 0.7576095461845398,
+      "learning_rate": 0.00010962834917891098,
+      "loss": 0.4031,
+      "step": 2619
+    },
+    {
+      "epoch": 13.55225806451613,
+      "grad_norm": 0.6655965447425842,
+      "learning_rate": 0.00010959377700950735,
+      "loss": 0.374,
+      "step": 2620
+    },
+    {
+      "epoch": 13.557419354838709,
+      "grad_norm": 0.6183370351791382,
+      "learning_rate": 0.00010955920484010371,
+      "loss": 0.397,
+      "step": 2621
+    },
+    {
+      "epoch": 13.56258064516129,
+      "grad_norm": 0.6197425127029419,
+      "learning_rate": 0.0001095246326707001,
+      "loss": 0.4006,
+      "step": 2622
+    },
+    {
+      "epoch": 13.567741935483872,
+      "grad_norm": 0.6073870658874512,
+      "learning_rate": 0.00010949006050129646,
+      "loss": 0.3782,
+      "step": 2623
+    },
+    {
+      "epoch": 13.572903225806451,
+      "grad_norm": 0.6646258234977722,
+      "learning_rate": 0.00010945548833189283,
+      "loss": 0.3755,
+      "step": 2624
+    },
+    {
+      "epoch": 13.578064516129032,
+      "grad_norm": 0.6667753458023071,
+      "learning_rate": 0.0001094209161624892,
+      "loss": 0.3772,
+      "step": 2625
+    },
+    {
+      "epoch": 13.583225806451614,
+      "grad_norm": 0.6182884573936462,
+      "learning_rate": 0.00010938634399308556,
+      "loss": 0.3782,
+      "step": 2626
+    },
+    {
+      "epoch": 13.588387096774193,
+      "grad_norm": 0.6554269790649414,
+      "learning_rate": 0.00010935177182368194,
+      "loss": 0.3817,
+      "step": 2627
+    },
+    {
+      "epoch": 13.593548387096774,
+      "grad_norm": 0.681104838848114,
+      "learning_rate": 0.00010931719965427831,
+      "loss": 0.3785,
+      "step": 2628
+    },
+    {
+      "epoch": 13.598709677419356,
+      "grad_norm": 0.6747802495956421,
+      "learning_rate": 0.00010928262748487468,
+      "loss": 0.4091,
+      "step": 2629
+    },
+    {
+      "epoch": 13.603870967741935,
+      "grad_norm": 0.6579374670982361,
+      "learning_rate": 0.00010924805531547104,
+      "loss": 0.3445,
+      "step": 2630
+    },
+    {
+      "epoch": 13.609032258064516,
+      "grad_norm": 0.6795297861099243,
+      "learning_rate": 0.00010921348314606742,
+      "loss": 0.3898,
+      "step": 2631
+    },
+    {
+      "epoch": 13.614193548387096,
+      "grad_norm": 0.5717626214027405,
+      "learning_rate": 0.00010917891097666379,
+      "loss": 0.3663,
+      "step": 2632
+    },
+    {
+      "epoch": 13.619354838709677,
+      "grad_norm": 0.6309238076210022,
+      "learning_rate": 0.00010914433880726016,
+      "loss": 0.3401,
+      "step": 2633
+    },
+    {
+      "epoch": 13.624516129032259,
+      "grad_norm": 0.6644474267959595,
+      "learning_rate": 0.00010910976663785652,
+      "loss": 0.3991,
+      "step": 2634
+    },
+    {
+      "epoch": 13.629677419354838,
+      "grad_norm": 0.6357470750808716,
+      "learning_rate": 0.00010907519446845289,
+      "loss": 0.3814,
+      "step": 2635
+    },
+    {
+      "epoch": 13.63483870967742,
+      "grad_norm": 0.684786319732666,
+      "learning_rate": 0.00010904062229904927,
+      "loss": 0.3869,
+      "step": 2636
+    },
+    {
+      "epoch": 13.64,
+      "grad_norm": 0.5164991617202759,
+      "learning_rate": 0.00010900605012964564,
+      "loss": 0.3713,
+      "step": 2637
+    },
+    {
+      "epoch": 13.64516129032258,
+      "grad_norm": 0.6062116622924805,
+      "learning_rate": 0.000108971477960242,
+      "loss": 0.4063,
+      "step": 2638
+    },
+    {
+      "epoch": 13.650322580645161,
+      "grad_norm": 0.5571811199188232,
+      "learning_rate": 0.00010893690579083837,
+      "loss": 0.3628,
+      "step": 2639
+    },
+    {
+      "epoch": 13.655483870967743,
+      "grad_norm": 0.6700261235237122,
+      "learning_rate": 0.00010890233362143475,
+      "loss": 0.3873,
+      "step": 2640
+    },
+    {
+      "epoch": 13.660645161290322,
+      "grad_norm": 0.6224830150604248,
+      "learning_rate": 0.00010886776145203112,
+      "loss": 0.3459,
+      "step": 2641
+    },
+    {
+      "epoch": 13.665806451612903,
+      "grad_norm": 0.6466324925422668,
+      "learning_rate": 0.00010883318928262749,
+      "loss": 0.3943,
+      "step": 2642
+    },
+    {
+      "epoch": 13.670967741935485,
+      "grad_norm": 0.6888540387153625,
+      "learning_rate": 0.00010879861711322385,
+      "loss": 0.3776,
+      "step": 2643
+    },
+    {
+      "epoch": 13.676129032258064,
+      "grad_norm": 0.6216358542442322,
+      "learning_rate": 0.00010876404494382022,
+      "loss": 0.3518,
+      "step": 2644
+    },
+    {
+      "epoch": 13.681290322580645,
+      "grad_norm": 0.6676345467567444,
+      "learning_rate": 0.0001087294727744166,
+      "loss": 0.3825,
+      "step": 2645
+    },
+    {
+      "epoch": 13.686451612903225,
+      "grad_norm": 0.6463304162025452,
+      "learning_rate": 0.00010869490060501297,
+      "loss": 0.3553,
+      "step": 2646
+    },
+    {
+      "epoch": 13.691612903225806,
+      "grad_norm": 0.646493673324585,
+      "learning_rate": 0.00010866032843560934,
+      "loss": 0.3487,
+      "step": 2647
+    },
+    {
+      "epoch": 13.696774193548388,
+      "grad_norm": 0.5988717079162598,
+      "learning_rate": 0.0001086257562662057,
+      "loss": 0.3575,
+      "step": 2648
+    },
+    {
+      "epoch": 13.701935483870967,
+      "grad_norm": 0.5779381990432739,
+      "learning_rate": 0.00010859118409680208,
+      "loss": 0.3455,
+      "step": 2649
+    },
+    {
+      "epoch": 13.707096774193548,
+      "grad_norm": 0.5907453298568726,
+      "learning_rate": 0.00010855661192739845,
+      "loss": 0.3499,
+      "step": 2650
+    },
+    {
+      "epoch": 13.71225806451613,
+      "grad_norm": 0.5870479941368103,
+      "learning_rate": 0.00010852203975799482,
+      "loss": 0.298,
+      "step": 2651
+    },
+    {
+      "epoch": 13.717419354838709,
+      "grad_norm": 0.644700825214386,
+      "learning_rate": 0.00010848746758859118,
+      "loss": 0.4264,
+      "step": 2652
+    },
+    {
+      "epoch": 13.72258064516129,
+      "grad_norm": 0.6083597540855408,
+      "learning_rate": 0.00010845289541918755,
+      "loss": 0.3888,
+      "step": 2653
+    },
+    {
+      "epoch": 13.727741935483872,
+      "grad_norm": 0.6860308647155762,
+      "learning_rate": 0.00010841832324978393,
+      "loss": 0.3736,
+      "step": 2654
+    },
+    {
+      "epoch": 13.732903225806451,
+      "grad_norm": 0.5994934439659119,
+      "learning_rate": 0.0001083837510803803,
+      "loss": 0.3446,
+      "step": 2655
+    },
+    {
+      "epoch": 13.738064516129032,
+      "grad_norm": 0.6875132918357849,
+      "learning_rate": 0.00010834917891097667,
+      "loss": 0.3912,
+      "step": 2656
+    },
+    {
+      "epoch": 13.743225806451614,
+      "grad_norm": 0.6328716278076172,
+      "learning_rate": 0.00010831460674157303,
+      "loss": 0.3816,
+      "step": 2657
+    },
+    {
+      "epoch": 13.748387096774193,
+      "grad_norm": 0.6054090261459351,
+      "learning_rate": 0.00010828003457216941,
+      "loss": 0.3838,
+      "step": 2658
+    },
+    {
+      "epoch": 13.753548387096775,
+      "grad_norm": 0.6844907999038696,
+      "learning_rate": 0.00010824546240276578,
+      "loss": 0.4051,
+      "step": 2659
+    },
+    {
+      "epoch": 13.758709677419354,
+      "grad_norm": 0.6264023184776306,
+      "learning_rate": 0.00010821089023336215,
+      "loss": 0.4102,
+      "step": 2660
+    },
+    {
+      "epoch": 13.763870967741935,
+      "grad_norm": 0.685297966003418,
+      "learning_rate": 0.00010817631806395851,
+      "loss": 0.3662,
+      "step": 2661
+    },
+    {
+      "epoch": 13.769032258064517,
+      "grad_norm": 0.6513123512268066,
+      "learning_rate": 0.00010814174589455488,
+      "loss": 0.3852,
+      "step": 2662
+    },
+    {
+      "epoch": 13.774193548387096,
+      "grad_norm": 0.5953750014305115,
+      "learning_rate": 0.00010810717372515126,
+      "loss": 0.4003,
+      "step": 2663
+    },
+    {
+      "epoch": 13.779354838709677,
+      "grad_norm": 0.7236970663070679,
+      "learning_rate": 0.00010807260155574763,
+      "loss": 0.419,
+      "step": 2664
+    },
+    {
+      "epoch": 13.784516129032259,
+      "grad_norm": 0.6723933219909668,
+      "learning_rate": 0.000108038029386344,
+      "loss": 0.4083,
+      "step": 2665
+    },
+    {
+      "epoch": 13.789677419354838,
+      "grad_norm": 0.6353141665458679,
+      "learning_rate": 0.00010800345721694036,
+      "loss": 0.3811,
+      "step": 2666
+    },
+    {
+      "epoch": 13.79483870967742,
+      "grad_norm": 0.6286487579345703,
+      "learning_rate": 0.00010796888504753673,
+      "loss": 0.3936,
+      "step": 2667
+    },
+    {
+      "epoch": 13.8,
+      "grad_norm": 0.6646544337272644,
+      "learning_rate": 0.00010793431287813311,
+      "loss": 0.396,
+      "step": 2668
+    },
+    {
+      "epoch": 13.80516129032258,
+      "grad_norm": 0.692557156085968,
+      "learning_rate": 0.00010789974070872948,
+      "loss": 0.3968,
+      "step": 2669
+    },
+    {
+      "epoch": 13.810322580645161,
+      "grad_norm": 0.6538509726524353,
+      "learning_rate": 0.00010786516853932584,
+      "loss": 0.341,
+      "step": 2670
+    },
+    {
+      "epoch": 13.815483870967743,
+      "grad_norm": 0.664397120475769,
+      "learning_rate": 0.00010783059636992221,
+      "loss": 0.3671,
+      "step": 2671
+    },
+    {
+      "epoch": 13.820645161290322,
+      "grad_norm": 0.5754483342170715,
+      "learning_rate": 0.00010779602420051859,
+      "loss": 0.343,
+      "step": 2672
+    },
+    {
+      "epoch": 13.825806451612904,
+      "grad_norm": 0.6304504871368408,
+      "learning_rate": 0.00010776145203111496,
+      "loss": 0.3943,
+      "step": 2673
+    },
+    {
+      "epoch": 13.830967741935483,
+      "grad_norm": 0.6337558031082153,
+      "learning_rate": 0.00010772687986171133,
+      "loss": 0.378,
+      "step": 2674
+    },
+    {
+      "epoch": 13.836129032258064,
+      "grad_norm": 0.6387003660202026,
+      "learning_rate": 0.0001076923076923077,
+      "loss": 0.4202,
+      "step": 2675
+    },
+    {
+      "epoch": 13.841290322580646,
+      "grad_norm": 0.665871262550354,
+      "learning_rate": 0.00010765773552290406,
+      "loss": 0.411,
+      "step": 2676
+    },
+    {
+      "epoch": 13.846451612903225,
+      "grad_norm": 0.6927699446678162,
+      "learning_rate": 0.00010762316335350044,
+      "loss": 0.3701,
+      "step": 2677
+    },
+    {
+      "epoch": 13.851612903225806,
+      "grad_norm": 0.6897647380828857,
+      "learning_rate": 0.00010758859118409681,
+      "loss": 0.3698,
+      "step": 2678
+    },
+    {
+      "epoch": 13.856774193548388,
+      "grad_norm": 0.6687372326850891,
+      "learning_rate": 0.00010755401901469317,
+      "loss": 0.3794,
+      "step": 2679
+    },
+    {
+      "epoch": 13.861935483870967,
+      "grad_norm": 0.6605384945869446,
+      "learning_rate": 0.00010751944684528954,
+      "loss": 0.3928,
+      "step": 2680
+    },
+    {
+      "epoch": 13.867096774193548,
+      "grad_norm": 0.7405558228492737,
+      "learning_rate": 0.00010748487467588592,
+      "loss": 0.4189,
+      "step": 2681
+    },
+    {
+      "epoch": 13.87225806451613,
+      "grad_norm": 0.7174172401428223,
+      "learning_rate": 0.00010745030250648229,
+      "loss": 0.393,
+      "step": 2682
+    },
+    {
+      "epoch": 13.87741935483871,
+      "grad_norm": 0.7276833653450012,
+      "learning_rate": 0.00010741573033707866,
+      "loss": 0.3927,
+      "step": 2683
+    },
+    {
+      "epoch": 13.88258064516129,
+      "grad_norm": 0.6784871220588684,
+      "learning_rate": 0.00010738115816767502,
+      "loss": 0.3772,
+      "step": 2684
+    },
+    {
+      "epoch": 13.88774193548387,
+      "grad_norm": 0.6098483800888062,
+      "learning_rate": 0.00010734658599827139,
+      "loss": 0.3558,
+      "step": 2685
+    },
+    {
+      "epoch": 13.892903225806451,
+      "grad_norm": 0.6713758707046509,
+      "learning_rate": 0.00010731201382886777,
+      "loss": 0.3309,
+      "step": 2686
+    },
+    {
+      "epoch": 13.898064516129033,
+      "grad_norm": 0.6196441650390625,
+      "learning_rate": 0.00010727744165946414,
+      "loss": 0.3493,
+      "step": 2687
+    },
+    {
+      "epoch": 13.903225806451612,
+      "grad_norm": 0.6994657516479492,
+      "learning_rate": 0.0001072428694900605,
+      "loss": 0.4008,
+      "step": 2688
+    },
+    {
+      "epoch": 13.908387096774193,
+      "grad_norm": 0.6330907940864563,
+      "learning_rate": 0.00010720829732065687,
+      "loss": 0.3553,
+      "step": 2689
+    },
+    {
+      "epoch": 13.913548387096775,
+      "grad_norm": 0.7182058691978455,
+      "learning_rate": 0.00010717372515125325,
+      "loss": 0.3641,
+      "step": 2690
+    },
+    {
+      "epoch": 13.918709677419354,
+      "grad_norm": 0.7052255272865295,
+      "learning_rate": 0.00010713915298184962,
+      "loss": 0.3789,
+      "step": 2691
+    },
+    {
+      "epoch": 13.923870967741935,
+      "grad_norm": 0.5603846311569214,
+      "learning_rate": 0.00010710458081244599,
+      "loss": 0.3364,
+      "step": 2692
+    },
+    {
+      "epoch": 13.929032258064517,
+      "grad_norm": 0.6760348677635193,
+      "learning_rate": 0.00010707000864304235,
+      "loss": 0.3612,
+      "step": 2693
+    },
+    {
+      "epoch": 13.934193548387096,
+      "grad_norm": 0.6764511466026306,
+      "learning_rate": 0.00010703543647363872,
+      "loss": 0.39,
+      "step": 2694
+    },
+    {
+      "epoch": 13.939354838709678,
+      "grad_norm": 0.7094846367835999,
+      "learning_rate": 0.0001070008643042351,
+      "loss": 0.3953,
+      "step": 2695
+    },
+    {
+      "epoch": 13.944516129032259,
+      "grad_norm": 0.6594676971435547,
+      "learning_rate": 0.00010696629213483147,
+      "loss": 0.3959,
+      "step": 2696
+    },
+    {
+      "epoch": 13.949677419354838,
+      "grad_norm": 0.7209908962249756,
+      "learning_rate": 0.00010693171996542783,
+      "loss": 0.3513,
+      "step": 2697
+    },
+    {
+      "epoch": 13.95483870967742,
+      "grad_norm": 0.6495475769042969,
+      "learning_rate": 0.0001068971477960242,
+      "loss": 0.3783,
+      "step": 2698
+    },
+    {
+      "epoch": 13.96,
+      "grad_norm": 0.7266281247138977,
+      "learning_rate": 0.00010686257562662057,
+      "loss": 0.364,
+      "step": 2699
+    },
+    {
+      "epoch": 13.96516129032258,
+      "grad_norm": 0.6638822555541992,
+      "learning_rate": 0.00010682800345721695,
+      "loss": 0.3826,
+      "step": 2700
+    },
+    {
+      "epoch": 13.970322580645162,
+      "grad_norm": 0.759272038936615,
+      "learning_rate": 0.00010679343128781332,
+      "loss": 0.4074,
+      "step": 2701
+    },
+    {
+      "epoch": 13.975483870967741,
+      "grad_norm": 0.6814619898796082,
+      "learning_rate": 0.00010675885911840968,
+      "loss": 0.3533,
+      "step": 2702
+    },
+    {
+      "epoch": 13.980645161290322,
+      "grad_norm": 0.6862568855285645,
+      "learning_rate": 0.00010672428694900605,
+      "loss": 0.3975,
+      "step": 2703
+    },
+    {
+      "epoch": 13.985806451612904,
+      "grad_norm": 0.657184362411499,
+      "learning_rate": 0.00010668971477960243,
+      "loss": 0.3992,
+      "step": 2704
+    },
+    {
+      "epoch": 13.990967741935483,
+      "grad_norm": 0.6523164510726929,
+      "learning_rate": 0.0001066551426101988,
+      "loss": 0.3895,
+      "step": 2705
+    },
+    {
+      "epoch": 13.996129032258064,
+      "grad_norm": 0.6338108777999878,
+      "learning_rate": 0.00010662057044079516,
+      "loss": 0.3871,
+      "step": 2706
+    },
+    {
+      "epoch": 14.00258064516129,
+      "grad_norm": 1.0012285709381104,
+      "learning_rate": 0.00010658599827139153,
+      "loss": 0.5385,
+      "step": 2707
+    },
+    {
+      "epoch": 14.007741935483871,
+      "grad_norm": 0.47075605392456055,
+      "learning_rate": 0.0001065514261019879,
+      "loss": 0.3511,
+      "step": 2708
+    },
+    {
+      "epoch": 14.012903225806452,
+      "grad_norm": 0.5172664523124695,
+      "learning_rate": 0.00010651685393258428,
+      "loss": 0.3522,
+      "step": 2709
+    },
+    {
+      "epoch": 14.018064516129032,
+      "grad_norm": 0.5740194916725159,
+      "learning_rate": 0.00010648228176318065,
+      "loss": 0.3462,
+      "step": 2710
+    },
+    {
+      "epoch": 14.023225806451613,
+      "grad_norm": 0.5373719334602356,
+      "learning_rate": 0.00010644770959377701,
+      "loss": 0.3702,
+      "step": 2711
+    },
+    {
+      "epoch": 14.028387096774194,
+      "grad_norm": 0.6073417663574219,
+      "learning_rate": 0.00010641313742437338,
+      "loss": 0.3377,
+      "step": 2712
+    },
+    {
+      "epoch": 14.033548387096774,
+      "grad_norm": 0.6683434247970581,
+      "learning_rate": 0.00010637856525496976,
+      "loss": 0.3687,
+      "step": 2713
+    },
+    {
+      "epoch": 14.038709677419355,
+      "grad_norm": 0.6118896007537842,
+      "learning_rate": 0.00010634399308556613,
+      "loss": 0.335,
+      "step": 2714
+    },
+    {
+      "epoch": 14.043870967741935,
+      "grad_norm": 0.6505861282348633,
+      "learning_rate": 0.0001063094209161625,
+      "loss": 0.3456,
+      "step": 2715
+    },
+    {
+      "epoch": 14.049032258064516,
+      "grad_norm": 0.5730725526809692,
+      "learning_rate": 0.00010627484874675886,
+      "loss": 0.3311,
+      "step": 2716
+    },
+    {
+      "epoch": 14.054193548387097,
+      "grad_norm": 0.49507811665534973,
+      "learning_rate": 0.00010624027657735523,
+      "loss": 0.3631,
+      "step": 2717
+    },
+    {
+      "epoch": 14.059354838709677,
+      "grad_norm": 0.6088764071464539,
+      "learning_rate": 0.00010620570440795161,
+      "loss": 0.3721,
+      "step": 2718
+    },
+    {
+      "epoch": 14.064516129032258,
+      "grad_norm": 0.5416562557220459,
+      "learning_rate": 0.00010617113223854798,
+      "loss": 0.3348,
+      "step": 2719
+    },
+    {
+      "epoch": 14.06967741935484,
+      "grad_norm": 0.5651316046714783,
+      "learning_rate": 0.00010613656006914434,
+      "loss": 0.3603,
+      "step": 2720
+    },
+    {
+      "epoch": 14.074838709677419,
+      "grad_norm": 0.5948140621185303,
+      "learning_rate": 0.00010610198789974071,
+      "loss": 0.3213,
+      "step": 2721
+    },
+    {
+      "epoch": 14.08,
+      "grad_norm": 0.5520354509353638,
+      "learning_rate": 0.00010606741573033709,
+      "loss": 0.3196,
+      "step": 2722
+    },
+    {
+      "epoch": 14.085161290322581,
+      "grad_norm": 0.6144856810569763,
+      "learning_rate": 0.00010603284356093346,
+      "loss": 0.3734,
+      "step": 2723
+    },
+    {
+      "epoch": 14.09032258064516,
+      "grad_norm": 0.6550484299659729,
+      "learning_rate": 0.00010599827139152982,
+      "loss": 0.3338,
+      "step": 2724
+    },
+    {
+      "epoch": 14.095483870967742,
+      "grad_norm": 0.5698432326316833,
+      "learning_rate": 0.00010596369922212619,
+      "loss": 0.3697,
+      "step": 2725
+    },
+    {
+      "epoch": 14.100645161290323,
+      "grad_norm": 0.6020158529281616,
+      "learning_rate": 0.00010592912705272256,
+      "loss": 0.366,
+      "step": 2726
+    },
+    {
+      "epoch": 14.105806451612903,
+      "grad_norm": 0.6318191289901733,
+      "learning_rate": 0.00010589455488331894,
+      "loss": 0.3083,
+      "step": 2727
+    },
+    {
+      "epoch": 14.110967741935484,
+      "grad_norm": 0.5770072340965271,
+      "learning_rate": 0.0001058599827139153,
+      "loss": 0.3075,
+      "step": 2728
+    },
+    {
+      "epoch": 14.116129032258064,
+      "grad_norm": 0.5020620822906494,
+      "learning_rate": 0.00010582541054451167,
+      "loss": 0.3235,
+      "step": 2729
+    },
+    {
+      "epoch": 14.121290322580645,
+      "grad_norm": 0.5899618864059448,
+      "learning_rate": 0.00010579083837510804,
+      "loss": 0.3318,
+      "step": 2730
+    },
+    {
+      "epoch": 14.126451612903226,
+      "grad_norm": 0.5642794966697693,
+      "learning_rate": 0.00010575626620570442,
+      "loss": 0.321,
+      "step": 2731
+    },
+    {
+      "epoch": 14.131612903225806,
+      "grad_norm": 0.5339599847793579,
+      "learning_rate": 0.00010572169403630079,
+      "loss": 0.338,
+      "step": 2732
+    },
+    {
+      "epoch": 14.136774193548387,
+      "grad_norm": 0.596102774143219,
+      "learning_rate": 0.00010568712186689715,
+      "loss": 0.3258,
+      "step": 2733
+    },
+    {
+      "epoch": 14.141935483870968,
+      "grad_norm": 0.6053940057754517,
+      "learning_rate": 0.00010565254969749352,
+      "loss": 0.3669,
+      "step": 2734
+    },
+    {
+      "epoch": 14.147096774193548,
+      "grad_norm": 0.6011558771133423,
+      "learning_rate": 0.00010561797752808989,
+      "loss": 0.3885,
+      "step": 2735
+    },
+    {
+      "epoch": 14.152258064516129,
+      "grad_norm": 0.5816941857337952,
+      "learning_rate": 0.00010558340535868627,
+      "loss": 0.3599,
+      "step": 2736
+    },
+    {
+      "epoch": 14.15741935483871,
+      "grad_norm": 0.5658031702041626,
+      "learning_rate": 0.00010554883318928264,
+      "loss": 0.3244,
+      "step": 2737
+    },
+    {
+      "epoch": 14.16258064516129,
+      "grad_norm": 0.6641707420349121,
+      "learning_rate": 0.000105514261019879,
+      "loss": 0.3747,
+      "step": 2738
+    },
+    {
+      "epoch": 14.167741935483871,
+      "grad_norm": 0.568182647228241,
+      "learning_rate": 0.00010547968885047537,
+      "loss": 0.3108,
+      "step": 2739
+    },
+    {
+      "epoch": 14.172903225806452,
+      "grad_norm": 0.4996654987335205,
+      "learning_rate": 0.00010544511668107174,
+      "loss": 0.3433,
+      "step": 2740
+    },
+    {
+      "epoch": 14.178064516129032,
+      "grad_norm": 0.635292112827301,
+      "learning_rate": 0.00010541054451166812,
+      "loss": 0.392,
+      "step": 2741
+    },
+    {
+      "epoch": 14.183225806451613,
+      "grad_norm": 0.6963523626327515,
+      "learning_rate": 0.00010537597234226448,
+      "loss": 0.3543,
+      "step": 2742
+    },
+    {
+      "epoch": 14.188387096774193,
+      "grad_norm": 0.5611749291419983,
+      "learning_rate": 0.00010534140017286085,
+      "loss": 0.3341,
+      "step": 2743
+    },
+    {
+      "epoch": 14.193548387096774,
+      "grad_norm": 0.5659300088882446,
+      "learning_rate": 0.00010530682800345722,
+      "loss": 0.3652,
+      "step": 2744
+    },
+    {
+      "epoch": 14.198709677419355,
+      "grad_norm": 0.4616934657096863,
+      "learning_rate": 0.0001052722558340536,
+      "loss": 0.319,
+      "step": 2745
+    },
+    {
+      "epoch": 14.203870967741935,
+      "grad_norm": 0.6389197707176208,
+      "learning_rate": 0.00010523768366464997,
+      "loss": 0.3543,
+      "step": 2746
+    },
+    {
+      "epoch": 14.209032258064516,
+      "grad_norm": 0.5770089030265808,
+      "learning_rate": 0.00010520311149524633,
+      "loss": 0.3149,
+      "step": 2747
+    },
+    {
+      "epoch": 14.214193548387097,
+      "grad_norm": 0.6154471635818481,
+      "learning_rate": 0.0001051685393258427,
+      "loss": 0.3763,
+      "step": 2748
+    },
+    {
+      "epoch": 14.219354838709677,
+      "grad_norm": 0.6077872514724731,
+      "learning_rate": 0.00010513396715643907,
+      "loss": 0.3545,
+      "step": 2749
+    },
+    {
+      "epoch": 14.224516129032258,
+      "grad_norm": 0.651185154914856,
+      "learning_rate": 0.00010509939498703545,
+      "loss": 0.343,
+      "step": 2750
+    },
+    {
+      "epoch": 14.22967741935484,
+      "grad_norm": 0.6004589796066284,
+      "learning_rate": 0.00010506482281763181,
+      "loss": 0.3663,
+      "step": 2751
+    },
+    {
+      "epoch": 14.234838709677419,
+      "grad_norm": 0.593478262424469,
+      "learning_rate": 0.00010503025064822818,
+      "loss": 0.3388,
+      "step": 2752
+    },
+    {
+      "epoch": 14.24,
+      "grad_norm": 0.6067554950714111,
+      "learning_rate": 0.00010499567847882455,
+      "loss": 0.3431,
+      "step": 2753
+    },
+    {
+      "epoch": 14.245161290322581,
+      "grad_norm": 0.5950843691825867,
+      "learning_rate": 0.00010496110630942093,
+      "loss": 0.3602,
+      "step": 2754
+    },
+    {
+      "epoch": 14.250322580645161,
+      "grad_norm": 0.568414568901062,
+      "learning_rate": 0.0001049265341400173,
+      "loss": 0.344,
+      "step": 2755
+    },
+    {
+      "epoch": 14.255483870967742,
+      "grad_norm": 0.6429581046104431,
+      "learning_rate": 0.00010489196197061366,
+      "loss": 0.3168,
+      "step": 2756
+    },
+    {
+      "epoch": 14.260645161290322,
+      "grad_norm": 0.6145492196083069,
+      "learning_rate": 0.00010485738980121003,
+      "loss": 0.3433,
+      "step": 2757
+    },
+    {
+      "epoch": 14.265806451612903,
+      "grad_norm": 0.6260680556297302,
+      "learning_rate": 0.0001048228176318064,
+      "loss": 0.3503,
+      "step": 2758
+    },
+    {
+      "epoch": 14.270967741935484,
+      "grad_norm": 0.5943008661270142,
+      "learning_rate": 0.00010478824546240278,
+      "loss": 0.2933,
+      "step": 2759
+    },
+    {
+      "epoch": 14.276129032258064,
+      "grad_norm": 0.6391633749008179,
+      "learning_rate": 0.00010475367329299914,
+      "loss": 0.3802,
+      "step": 2760
+    },
+    {
+      "epoch": 14.281290322580645,
+      "grad_norm": 0.5643005967140198,
+      "learning_rate": 0.00010471910112359551,
+      "loss": 0.2959,
+      "step": 2761
+    },
+    {
+      "epoch": 14.286451612903226,
+      "grad_norm": 0.6687441468238831,
+      "learning_rate": 0.00010468452895419188,
+      "loss": 0.3632,
+      "step": 2762
+    },
+    {
+      "epoch": 14.291612903225806,
+      "grad_norm": 0.6497690081596375,
+      "learning_rate": 0.00010464995678478826,
+      "loss": 0.3501,
+      "step": 2763
+    },
+    {
+      "epoch": 14.296774193548387,
+      "grad_norm": 0.6333799958229065,
+      "learning_rate": 0.00010461538461538463,
+      "loss": 0.3557,
+      "step": 2764
+    },
+    {
+      "epoch": 14.301935483870968,
+      "grad_norm": 0.5643303394317627,
+      "learning_rate": 0.00010458081244598099,
+      "loss": 0.3421,
+      "step": 2765
+    },
+    {
+      "epoch": 14.307096774193548,
+      "grad_norm": 0.5864318013191223,
+      "learning_rate": 0.00010454624027657736,
+      "loss": 0.3272,
+      "step": 2766
+    },
+    {
+      "epoch": 14.31225806451613,
+      "grad_norm": 0.6277300119400024,
+      "learning_rate": 0.00010451166810717373,
+      "loss": 0.3797,
+      "step": 2767
+    },
+    {
+      "epoch": 14.31741935483871,
+      "grad_norm": 0.654905378818512,
+      "learning_rate": 0.00010447709593777011,
+      "loss": 0.3762,
+      "step": 2768
+    },
+    {
+      "epoch": 14.32258064516129,
+      "grad_norm": 0.5646421313285828,
+      "learning_rate": 0.00010444252376836647,
+      "loss": 0.3519,
+      "step": 2769
+    },
+    {
+      "epoch": 14.327741935483871,
+      "grad_norm": 0.6128076314926147,
+      "learning_rate": 0.00010440795159896284,
+      "loss": 0.3294,
+      "step": 2770
+    },
+    {
+      "epoch": 14.33290322580645,
+      "grad_norm": 0.5849538445472717,
+      "learning_rate": 0.00010437337942955921,
+      "loss": 0.3487,
+      "step": 2771
+    },
+    {
+      "epoch": 14.338064516129032,
+      "grad_norm": 0.6046317219734192,
+      "learning_rate": 0.00010433880726015558,
+      "loss": 0.364,
+      "step": 2772
+    },
+    {
+      "epoch": 14.343225806451613,
+      "grad_norm": 0.6184499859809875,
+      "learning_rate": 0.00010430423509075196,
+      "loss": 0.3213,
+      "step": 2773
+    },
+    {
+      "epoch": 14.348387096774193,
+      "grad_norm": 0.618805468082428,
+      "learning_rate": 0.00010426966292134832,
+      "loss": 0.3534,
+      "step": 2774
+    },
+    {
+      "epoch": 14.353548387096774,
+      "grad_norm": 0.5907148718833923,
+      "learning_rate": 0.00010423509075194469,
+      "loss": 0.3003,
+      "step": 2775
+    },
+    {
+      "epoch": 14.358709677419355,
+      "grad_norm": 0.6237258315086365,
+      "learning_rate": 0.00010420051858254106,
+      "loss": 0.3711,
+      "step": 2776
+    },
+    {
+      "epoch": 14.363870967741935,
+      "grad_norm": 0.6238266825675964,
+      "learning_rate": 0.00010416594641313744,
+      "loss": 0.3389,
+      "step": 2777
+    },
+    {
+      "epoch": 14.369032258064516,
+      "grad_norm": 0.665052056312561,
+      "learning_rate": 0.0001041313742437338,
+      "loss": 0.3497,
+      "step": 2778
+    },
+    {
+      "epoch": 14.374193548387098,
+      "grad_norm": 0.5933361649513245,
+      "learning_rate": 0.00010409680207433017,
+      "loss": 0.353,
+      "step": 2779
+    },
+    {
+      "epoch": 14.379354838709677,
+      "grad_norm": 0.573525071144104,
+      "learning_rate": 0.00010406222990492654,
+      "loss": 0.373,
+      "step": 2780
+    },
+    {
+      "epoch": 14.384516129032258,
+      "grad_norm": 0.6180545687675476,
+      "learning_rate": 0.0001040276577355229,
+      "loss": 0.3598,
+      "step": 2781
+    },
+    {
+      "epoch": 14.38967741935484,
+      "grad_norm": 0.619771420955658,
+      "learning_rate": 0.00010399308556611929,
+      "loss": 0.3521,
+      "step": 2782
+    },
+    {
+      "epoch": 14.394838709677419,
+      "grad_norm": 0.5874396562576294,
+      "learning_rate": 0.00010395851339671565,
+      "loss": 0.3884,
+      "step": 2783
+    },
+    {
+      "epoch": 14.4,
+      "grad_norm": 0.6538547873497009,
+      "learning_rate": 0.00010392394122731202,
+      "loss": 0.3617,
+      "step": 2784
+    },
+    {
+      "epoch": 14.40516129032258,
+      "grad_norm": 0.6547790765762329,
+      "learning_rate": 0.00010388936905790839,
+      "loss": 0.3193,
+      "step": 2785
+    },
+    {
+      "epoch": 14.410322580645161,
+      "grad_norm": 0.6283661127090454,
+      "learning_rate": 0.00010385479688850477,
+      "loss": 0.3475,
+      "step": 2786
+    },
+    {
+      "epoch": 14.415483870967742,
+      "grad_norm": 0.6837930083274841,
+      "learning_rate": 0.00010382022471910113,
+      "loss": 0.3535,
+      "step": 2787
+    },
+    {
+      "epoch": 14.420645161290322,
+      "grad_norm": 0.6627948880195618,
+      "learning_rate": 0.0001037856525496975,
+      "loss": 0.3794,
+      "step": 2788
+    },
+    {
+      "epoch": 14.425806451612903,
+      "grad_norm": 0.6747078895568848,
+      "learning_rate": 0.00010375108038029387,
+      "loss": 0.3615,
+      "step": 2789
+    },
+    {
+      "epoch": 14.430967741935484,
+      "grad_norm": 0.6608096957206726,
+      "learning_rate": 0.00010371650821089022,
+      "loss": 0.3279,
+      "step": 2790
+    },
+    {
+      "epoch": 14.436129032258064,
+      "grad_norm": 0.5525237321853638,
+      "learning_rate": 0.00010368193604148662,
+      "loss": 0.3525,
+      "step": 2791
+    },
+    {
+      "epoch": 14.441290322580645,
+      "grad_norm": 0.6812670826911926,
+      "learning_rate": 0.00010364736387208298,
+      "loss": 0.341,
+      "step": 2792
+    },
+    {
+      "epoch": 14.446451612903227,
+      "grad_norm": 0.5752320885658264,
+      "learning_rate": 0.00010361279170267935,
+      "loss": 0.3451,
+      "step": 2793
+    },
+    {
+      "epoch": 14.451612903225806,
+      "grad_norm": 0.7717772126197815,
+      "learning_rate": 0.00010357821953327572,
+      "loss": 0.3581,
+      "step": 2794
+    },
+    {
+      "epoch": 14.456774193548387,
+      "grad_norm": 0.5956462621688843,
+      "learning_rate": 0.0001035436473638721,
+      "loss": 0.3375,
+      "step": 2795
+    },
+    {
+      "epoch": 14.461935483870967,
+      "grad_norm": 0.554633378982544,
+      "learning_rate": 0.00010350907519446846,
+      "loss": 0.3453,
+      "step": 2796
+    },
+    {
+      "epoch": 14.467096774193548,
+      "grad_norm": 0.6363507509231567,
+      "learning_rate": 0.00010347450302506483,
+      "loss": 0.3574,
+      "step": 2797
+    },
+    {
+      "epoch": 14.47225806451613,
+      "grad_norm": 0.6431280374526978,
+      "learning_rate": 0.0001034399308556612,
+      "loss": 0.2919,
+      "step": 2798
+    },
+    {
+      "epoch": 14.477419354838709,
+      "grad_norm": 0.6716758608818054,
+      "learning_rate": 0.00010340535868625755,
+      "loss": 0.346,
+      "step": 2799
+    },
+    {
+      "epoch": 14.48258064516129,
+      "grad_norm": 0.6309583783149719,
+      "learning_rate": 0.00010337078651685395,
+      "loss": 0.3516,
+      "step": 2800
+    },
+    {
+      "epoch": 14.487741935483871,
+      "grad_norm": 0.6508819460868835,
+      "learning_rate": 0.00010333621434745031,
+      "loss": 0.3448,
+      "step": 2801
+    },
+    {
+      "epoch": 14.492903225806451,
+      "grad_norm": 0.6676908731460571,
+      "learning_rate": 0.00010330164217804668,
+      "loss": 0.3523,
+      "step": 2802
+    },
+    {
+      "epoch": 14.498064516129032,
+      "grad_norm": 0.6469940543174744,
+      "learning_rate": 0.00010326707000864305,
+      "loss": 0.3429,
+      "step": 2803
+    },
+    {
+      "epoch": 14.503225806451614,
+      "grad_norm": 0.6421766877174377,
+      "learning_rate": 0.00010323249783923943,
+      "loss": 0.3714,
+      "step": 2804
+    },
+    {
+      "epoch": 14.508387096774193,
+      "grad_norm": 0.6937445402145386,
+      "learning_rate": 0.0001031979256698358,
+      "loss": 0.4049,
+      "step": 2805
+    },
+    {
+      "epoch": 14.513548387096774,
+      "grad_norm": 0.6430542469024658,
+      "learning_rate": 0.00010316335350043216,
+      "loss": 0.3576,
+      "step": 2806
+    },
+    {
+      "epoch": 14.518709677419356,
+      "grad_norm": 0.6543039679527283,
+      "learning_rate": 0.00010312878133102853,
+      "loss": 0.3561,
+      "step": 2807
+    },
+    {
+      "epoch": 14.523870967741935,
+      "grad_norm": 0.6225786805152893,
+      "learning_rate": 0.00010309420916162488,
+      "loss": 0.3503,
+      "step": 2808
+    },
+    {
+      "epoch": 14.529032258064516,
+      "grad_norm": 0.707748532295227,
+      "learning_rate": 0.00010305963699222128,
+      "loss": 0.3677,
+      "step": 2809
+    },
+    {
+      "epoch": 14.534193548387098,
+      "grad_norm": 0.7335622906684875,
+      "learning_rate": 0.00010302506482281764,
+      "loss": 0.3832,
+      "step": 2810
+    },
+    {
+      "epoch": 14.539354838709677,
+      "grad_norm": 0.6188815832138062,
+      "learning_rate": 0.00010299049265341401,
+      "loss": 0.34,
+      "step": 2811
+    },
+    {
+      "epoch": 14.544516129032258,
+      "grad_norm": 0.6443189978599548,
+      "learning_rate": 0.00010295592048401038,
+      "loss": 0.3287,
+      "step": 2812
+    },
+    {
+      "epoch": 14.549677419354838,
+      "grad_norm": 0.6156713366508484,
+      "learning_rate": 0.00010292134831460673,
+      "loss": 0.3465,
+      "step": 2813
+    },
+    {
+      "epoch": 14.55483870967742,
+      "grad_norm": 0.6016891002655029,
+      "learning_rate": 0.00010288677614520312,
+      "loss": 0.3332,
+      "step": 2814
+    },
+    {
+      "epoch": 14.56,
+      "grad_norm": 0.7056647539138794,
+      "learning_rate": 0.00010285220397579949,
+      "loss": 0.3477,
+      "step": 2815
+    },
+    {
+      "epoch": 14.56516129032258,
+      "grad_norm": 0.6235531568527222,
+      "learning_rate": 0.00010281763180639586,
+      "loss": 0.3402,
+      "step": 2816
+    },
+    {
+      "epoch": 14.570322580645161,
+      "grad_norm": 0.6397790908813477,
+      "learning_rate": 0.00010278305963699221,
+      "loss": 0.364,
+      "step": 2817
+    },
+    {
+      "epoch": 14.575483870967743,
+      "grad_norm": 0.6796033978462219,
+      "learning_rate": 0.0001027484874675886,
+      "loss": 0.3649,
+      "step": 2818
+    },
+    {
+      "epoch": 14.580645161290322,
+      "grad_norm": 0.7013598084449768,
+      "learning_rate": 0.00010271391529818497,
+      "loss": 0.4044,
+      "step": 2819
+    },
+    {
+      "epoch": 14.585806451612903,
+      "grad_norm": 0.5599159598350525,
+      "learning_rate": 0.00010267934312878134,
+      "loss": 0.3759,
+      "step": 2820
+    },
+    {
+      "epoch": 14.590967741935485,
+      "grad_norm": 0.6569648385047913,
+      "learning_rate": 0.00010264477095937771,
+      "loss": 0.3663,
+      "step": 2821
+    },
+    {
+      "epoch": 14.596129032258064,
+      "grad_norm": 0.6489759683609009,
+      "learning_rate": 0.00010261019878997406,
+      "loss": 0.3555,
+      "step": 2822
+    },
+    {
+      "epoch": 14.601290322580645,
+      "grad_norm": 0.6897649168968201,
+      "learning_rate": 0.00010257562662057045,
+      "loss": 0.3489,
+      "step": 2823
+    },
+    {
+      "epoch": 14.606451612903225,
+      "grad_norm": 0.5831968188285828,
+      "learning_rate": 0.00010254105445116682,
+      "loss": 0.3169,
+      "step": 2824
+    },
+    {
+      "epoch": 14.611612903225806,
+      "grad_norm": 0.6416640281677246,
+      "learning_rate": 0.00010250648228176319,
+      "loss": 0.3502,
+      "step": 2825
+    },
+    {
+      "epoch": 14.616774193548387,
+      "grad_norm": 0.6002489328384399,
+      "learning_rate": 0.00010247191011235954,
+      "loss": 0.3548,
+      "step": 2826
+    },
+    {
+      "epoch": 14.621935483870967,
+      "grad_norm": 0.6205369830131531,
+      "learning_rate": 0.00010243733794295594,
+      "loss": 0.3511,
+      "step": 2827
+    },
+    {
+      "epoch": 14.627096774193548,
+      "grad_norm": 0.7159790992736816,
+      "learning_rate": 0.0001024027657735523,
+      "loss": 0.3527,
+      "step": 2828
+    },
+    {
+      "epoch": 14.63225806451613,
+      "grad_norm": 0.652669370174408,
+      "learning_rate": 0.00010236819360414867,
+      "loss": 0.349,
+      "step": 2829
+    },
+    {
+      "epoch": 14.637419354838709,
+      "grad_norm": 0.7023113965988159,
+      "learning_rate": 0.00010233362143474504,
+      "loss": 0.3831,
+      "step": 2830
+    },
+    {
+      "epoch": 14.64258064516129,
+      "grad_norm": 0.6868276000022888,
+      "learning_rate": 0.00010229904926534139,
+      "loss": 0.3321,
+      "step": 2831
+    },
+    {
+      "epoch": 14.647741935483872,
+      "grad_norm": 0.6512218713760376,
+      "learning_rate": 0.00010226447709593778,
+      "loss": 0.3387,
+      "step": 2832
+    },
+    {
+      "epoch": 14.652903225806451,
+      "grad_norm": 0.7037277817726135,
+      "learning_rate": 0.00010222990492653415,
+      "loss": 0.3805,
+      "step": 2833
+    },
+    {
+      "epoch": 14.658064516129032,
+      "grad_norm": 0.6521219611167908,
+      "learning_rate": 0.00010219533275713052,
+      "loss": 0.3403,
+      "step": 2834
+    },
+    {
+      "epoch": 14.663225806451614,
+      "grad_norm": 0.6446892023086548,
+      "learning_rate": 0.00010216076058772687,
+      "loss": 0.3517,
+      "step": 2835
+    },
+    {
+      "epoch": 14.668387096774193,
+      "grad_norm": 0.6616457104682922,
+      "learning_rate": 0.00010212618841832327,
+      "loss": 0.3419,
+      "step": 2836
+    },
+    {
+      "epoch": 14.673548387096774,
+      "grad_norm": 0.648313581943512,
+      "learning_rate": 0.00010209161624891963,
+      "loss": 0.3507,
+      "step": 2837
+    },
+    {
+      "epoch": 14.678709677419356,
+      "grad_norm": 0.6885864734649658,
+      "learning_rate": 0.000102057044079516,
+      "loss": 0.3535,
+      "step": 2838
+    },
+    {
+      "epoch": 14.683870967741935,
+      "grad_norm": 0.6264383792877197,
+      "learning_rate": 0.00010202247191011237,
+      "loss": 0.3497,
+      "step": 2839
+    },
+    {
+      "epoch": 14.689032258064517,
+      "grad_norm": 0.7062277793884277,
+      "learning_rate": 0.00010198789974070872,
+      "loss": 0.3345,
+      "step": 2840
+    },
+    {
+      "epoch": 14.694193548387096,
+      "grad_norm": 0.7072764039039612,
+      "learning_rate": 0.00010195332757130511,
+      "loss": 0.3549,
+      "step": 2841
+    },
+    {
+      "epoch": 14.699354838709677,
+      "grad_norm": 0.6753832101821899,
+      "learning_rate": 0.00010191875540190148,
+      "loss": 0.3427,
+      "step": 2842
+    },
+    {
+      "epoch": 14.704516129032259,
+      "grad_norm": 1.1632778644561768,
+      "learning_rate": 0.00010188418323249785,
+      "loss": 0.3912,
+      "step": 2843
+    },
+    {
+      "epoch": 14.709677419354838,
+      "grad_norm": 0.6571031808853149,
+      "learning_rate": 0.0001018496110630942,
+      "loss": 0.3494,
+      "step": 2844
+    },
+    {
+      "epoch": 14.71483870967742,
+      "grad_norm": 0.6987926363945007,
+      "learning_rate": 0.00010181503889369057,
+      "loss": 0.3325,
+      "step": 2845
+    },
+    {
+      "epoch": 14.72,
+      "grad_norm": 0.7170534133911133,
+      "learning_rate": 0.00010178046672428696,
+      "loss": 0.3497,
+      "step": 2846
+    },
+    {
+      "epoch": 14.72516129032258,
+      "grad_norm": 0.7260218262672424,
+      "learning_rate": 0.00010174589455488333,
+      "loss": 0.3767,
+      "step": 2847
+    },
+    {
+      "epoch": 14.730322580645161,
+      "grad_norm": 0.7352275848388672,
+      "learning_rate": 0.0001017113223854797,
+      "loss": 0.3787,
+      "step": 2848
+    },
+    {
+      "epoch": 14.735483870967743,
+      "grad_norm": 0.7478794455528259,
+      "learning_rate": 0.00010167675021607605,
+      "loss": 0.3824,
+      "step": 2849
+    },
+    {
+      "epoch": 14.740645161290322,
+      "grad_norm": 0.63609379529953,
+      "learning_rate": 0.00010164217804667244,
+      "loss": 0.3761,
+      "step": 2850
+    },
+    {
+      "epoch": 14.745806451612903,
+      "grad_norm": 0.7185914516448975,
+      "learning_rate": 0.00010160760587726881,
+      "loss": 0.3681,
+      "step": 2851
+    },
+    {
+      "epoch": 14.750967741935483,
+      "grad_norm": 0.6970106959342957,
+      "learning_rate": 0.00010157303370786518,
+      "loss": 0.3992,
+      "step": 2852
+    },
+    {
+      "epoch": 14.756129032258064,
+      "grad_norm": 0.6689693927764893,
+      "learning_rate": 0.00010153846153846153,
+      "loss": 0.348,
+      "step": 2853
+    },
+    {
+      "epoch": 14.761290322580646,
+      "grad_norm": 0.6985188126564026,
+      "learning_rate": 0.0001015038893690579,
+      "loss": 0.3796,
+      "step": 2854
+    },
+    {
+      "epoch": 14.766451612903225,
+      "grad_norm": 0.715948224067688,
+      "learning_rate": 0.00010146931719965429,
+      "loss": 0.3393,
+      "step": 2855
+    },
+    {
+      "epoch": 14.771612903225806,
+      "grad_norm": 0.7230039834976196,
+      "learning_rate": 0.00010143474503025066,
+      "loss": 0.3459,
+      "step": 2856
+    },
+    {
+      "epoch": 14.776774193548388,
+      "grad_norm": 0.6375747323036194,
+      "learning_rate": 0.00010140017286084703,
+      "loss": 0.3459,
+      "step": 2857
+    },
+    {
+      "epoch": 14.781935483870967,
+      "grad_norm": 0.6524990200996399,
+      "learning_rate": 0.00010136560069144338,
+      "loss": 0.3597,
+      "step": 2858
+    },
+    {
+      "epoch": 14.787096774193548,
+      "grad_norm": 0.7183625102043152,
+      "learning_rate": 0.00010133102852203977,
+      "loss": 0.4133,
+      "step": 2859
+    },
+    {
+      "epoch": 14.79225806451613,
+      "grad_norm": 0.667828381061554,
+      "learning_rate": 0.00010129645635263614,
+      "loss": 0.3589,
+      "step": 2860
+    },
+    {
+      "epoch": 14.79741935483871,
+      "grad_norm": 0.730376124382019,
+      "learning_rate": 0.00010126188418323251,
+      "loss": 0.3939,
+      "step": 2861
+    },
+    {
+      "epoch": 14.80258064516129,
+      "grad_norm": 0.6927716135978699,
+      "learning_rate": 0.00010122731201382886,
+      "loss": 0.3481,
+      "step": 2862
+    },
+    {
+      "epoch": 14.807741935483872,
+      "grad_norm": 0.7189210057258606,
+      "learning_rate": 0.00010119273984442523,
+      "loss": 0.3588,
+      "step": 2863
+    },
+    {
+      "epoch": 14.812903225806451,
+      "grad_norm": 0.697931170463562,
+      "learning_rate": 0.00010115816767502162,
+      "loss": 0.326,
+      "step": 2864
+    },
+    {
+      "epoch": 14.818064516129033,
+      "grad_norm": 0.6717986464500427,
+      "learning_rate": 0.00010112359550561799,
+      "loss": 0.3272,
+      "step": 2865
+    },
+    {
+      "epoch": 14.823225806451614,
+      "grad_norm": 0.6656721234321594,
+      "learning_rate": 0.00010108902333621436,
+      "loss": 0.3583,
+      "step": 2866
+    },
+    {
+      "epoch": 14.828387096774193,
+      "grad_norm": 0.7824127078056335,
+      "learning_rate": 0.00010105445116681071,
+      "loss": 0.3493,
+      "step": 2867
+    },
+    {
+      "epoch": 14.833548387096775,
+      "grad_norm": 0.7451285719871521,
+      "learning_rate": 0.0001010198789974071,
+      "loss": 0.3511,
+      "step": 2868
+    },
+    {
+      "epoch": 14.838709677419354,
+      "grad_norm": 0.7545776963233948,
+      "learning_rate": 0.00010098530682800347,
+      "loss": 0.3706,
+      "step": 2869
+    },
+    {
+      "epoch": 14.843870967741935,
+      "grad_norm": 0.6645585298538208,
+      "learning_rate": 0.00010095073465859984,
+      "loss": 0.3858,
+      "step": 2870
+    },
+    {
+      "epoch": 14.849032258064517,
+      "grad_norm": 0.739601194858551,
+      "learning_rate": 0.00010091616248919619,
+      "loss": 0.3946,
+      "step": 2871
+    },
+    {
+      "epoch": 14.854193548387096,
+      "grad_norm": 0.7024446129798889,
+      "learning_rate": 0.00010088159031979256,
+      "loss": 0.3636,
+      "step": 2872
+    },
+    {
+      "epoch": 14.859354838709677,
+      "grad_norm": 0.9212889671325684,
+      "learning_rate": 0.00010084701815038895,
+      "loss": 0.3396,
+      "step": 2873
+    },
+    {
+      "epoch": 14.864516129032259,
+      "grad_norm": 0.7592905163764954,
+      "learning_rate": 0.00010081244598098532,
+      "loss": 0.346,
+      "step": 2874
+    },
+    {
+      "epoch": 14.869677419354838,
+      "grad_norm": 0.6623084545135498,
+      "learning_rate": 0.00010077787381158169,
+      "loss": 0.3518,
+      "step": 2875
+    },
+    {
+      "epoch": 14.87483870967742,
+      "grad_norm": 0.734428346157074,
+      "learning_rate": 0.00010074330164217804,
+      "loss": 0.3462,
+      "step": 2876
+    },
+    {
+      "epoch": 14.88,
+      "grad_norm": 0.6849513053894043,
+      "learning_rate": 0.00010070872947277443,
+      "loss": 0.3321,
+      "step": 2877
+    },
+    {
+      "epoch": 14.88516129032258,
+      "grad_norm": 0.7301912903785706,
+      "learning_rate": 0.0001006741573033708,
+      "loss": 0.3459,
+      "step": 2878
+    },
+    {
+      "epoch": 14.890322580645162,
+      "grad_norm": 0.6769235134124756,
+      "learning_rate": 0.00010063958513396717,
+      "loss": 0.3675,
+      "step": 2879
+    },
+    {
+      "epoch": 14.895483870967741,
+      "grad_norm": 0.7474826574325562,
+      "learning_rate": 0.00010060501296456352,
+      "loss": 0.3753,
+      "step": 2880
+    },
+    {
+      "epoch": 14.900645161290322,
+      "grad_norm": 0.7444056272506714,
+      "learning_rate": 0.00010057044079515989,
+      "loss": 0.3858,
+      "step": 2881
+    },
+    {
+      "epoch": 14.905806451612904,
+      "grad_norm": 0.6257233023643494,
+      "learning_rate": 0.00010053586862575628,
+      "loss": 0.2887,
+      "step": 2882
+    },
+    {
+      "epoch": 14.910967741935483,
+      "grad_norm": 0.7656845450401306,
+      "learning_rate": 0.00010050129645635265,
+      "loss": 0.3236,
+      "step": 2883
+    },
+    {
+      "epoch": 14.916129032258064,
+      "grad_norm": 0.6776483058929443,
+      "learning_rate": 0.00010046672428694902,
+      "loss": 0.3639,
+      "step": 2884
+    },
+    {
+      "epoch": 14.921290322580646,
+      "grad_norm": 0.7685416340827942,
+      "learning_rate": 0.00010043215211754537,
+      "loss": 0.3803,
+      "step": 2885
+    },
+    {
+      "epoch": 14.926451612903225,
+      "grad_norm": 0.6683724522590637,
+      "learning_rate": 0.00010039757994814174,
+      "loss": 0.363,
+      "step": 2886
+    },
+    {
+      "epoch": 14.931612903225806,
+      "grad_norm": 0.6642737984657288,
+      "learning_rate": 0.00010036300777873813,
+      "loss": 0.388,
+      "step": 2887
+    },
+    {
+      "epoch": 14.936774193548388,
+      "grad_norm": 0.7816684246063232,
+      "learning_rate": 0.0001003284356093345,
+      "loss": 0.3512,
+      "step": 2888
+    },
+    {
+      "epoch": 14.941935483870967,
+      "grad_norm": 0.7637103796005249,
+      "learning_rate": 0.00010029386343993085,
+      "loss": 0.3393,
+      "step": 2889
+    },
+    {
+      "epoch": 14.947096774193549,
+      "grad_norm": 0.654755175113678,
+      "learning_rate": 0.00010025929127052722,
+      "loss": 0.3224,
+      "step": 2890
+    },
+    {
+      "epoch": 14.95225806451613,
+      "grad_norm": 0.7690180540084839,
+      "learning_rate": 0.00010022471910112361,
+      "loss": 0.3711,
+      "step": 2891
+    },
+    {
+      "epoch": 14.95741935483871,
+      "grad_norm": 0.7185976505279541,
+      "learning_rate": 0.00010019014693171998,
+      "loss": 0.3565,
+      "step": 2892
+    },
+    {
+      "epoch": 14.96258064516129,
+      "grad_norm": 0.7181047201156616,
+      "learning_rate": 0.00010015557476231635,
+      "loss": 0.3832,
+      "step": 2893
+    },
+    {
+      "epoch": 14.967741935483872,
+      "grad_norm": 0.7195354700088501,
+      "learning_rate": 0.0001001210025929127,
+      "loss": 0.3524,
+      "step": 2894
+    },
+    {
+      "epoch": 14.972903225806451,
+      "grad_norm": 0.6921024918556213,
+      "learning_rate": 0.00010008643042350907,
+      "loss": 0.3843,
+      "step": 2895
+    },
+    {
+      "epoch": 14.978064516129033,
+      "grad_norm": 0.7610807418823242,
+      "learning_rate": 0.00010005185825410546,
+      "loss": 0.3361,
+      "step": 2896
+    },
+    {
+      "epoch": 14.983225806451612,
+      "grad_norm": 0.7554695010185242,
+      "learning_rate": 0.00010001728608470183,
+      "loss": 0.3883,
+      "step": 2897
+    },
+    {
+      "epoch": 14.988387096774193,
+      "grad_norm": 0.7315402626991272,
+      "learning_rate": 9.998271391529818e-05,
+      "loss": 0.3507,
+      "step": 2898
+    },
+    {
+      "epoch": 14.993548387096775,
+      "grad_norm": 0.7120655179023743,
+      "learning_rate": 9.994814174589456e-05,
+      "loss": 0.3555,
+      "step": 2899
+    },
+    {
+      "epoch": 14.998709677419354,
+      "grad_norm": 1.1938225030899048,
+      "learning_rate": 9.991356957649093e-05,
+      "loss": 0.5132,
+      "step": 2900
+    },
+    {
+      "epoch": 14.998709677419354,
+      "eval_loss": 2.549619674682617,
+      "eval_runtime": 21.3407,
+      "eval_samples_per_second": 3.702,
+      "eval_steps_per_second": 0.469,
+      "step": 2900
+    },
+    {
+      "epoch": 15.005161290322581,
+      "grad_norm": 0.5872735381126404,
+      "learning_rate": 9.987899740708731e-05,
+      "loss": 0.3549,
+      "step": 2901
+    },
+    {
+      "epoch": 15.01032258064516,
+      "grad_norm": 0.5851104855537415,
+      "learning_rate": 9.984442523768366e-05,
+      "loss": 0.3526,
+      "step": 2902
+    },
+    {
+      "epoch": 15.015483870967742,
+      "grad_norm": 0.5479742288589478,
+      "learning_rate": 9.980985306828003e-05,
+      "loss": 0.3107,
+      "step": 2903
+    },
+    {
+      "epoch": 15.020645161290323,
+      "grad_norm": 0.5455775856971741,
+      "learning_rate": 9.977528089887641e-05,
+      "loss": 0.2762,
+      "step": 2904
+    },
+    {
+      "epoch": 15.025806451612903,
+      "grad_norm": 0.672476589679718,
+      "learning_rate": 9.974070872947278e-05,
+      "loss": 0.3147,
+      "step": 2905
+    },
+    {
+      "epoch": 15.030967741935484,
+      "grad_norm": 0.6025128364562988,
+      "learning_rate": 9.970613656006916e-05,
+      "loss": 0.304,
+      "step": 2906
+    },
+    {
+      "epoch": 15.036129032258064,
+      "grad_norm": 0.6719430685043335,
+      "learning_rate": 9.967156439066551e-05,
+      "loss": 0.2951,
+      "step": 2907
+    },
+    {
+      "epoch": 15.041290322580645,
+      "grad_norm": 0.6253131628036499,
+      "learning_rate": 9.963699222126189e-05,
+      "loss": 0.3171,
+      "step": 2908
+    },
+    {
+      "epoch": 15.046451612903226,
+      "grad_norm": 0.6282058358192444,
+      "learning_rate": 9.960242005185826e-05,
+      "loss": 0.3363,
+      "step": 2909
+    },
+    {
+      "epoch": 15.051612903225806,
+      "grad_norm": 0.5362942218780518,
+      "learning_rate": 9.956784788245463e-05,
+      "loss": 0.3092,
+      "step": 2910
+    },
+    {
+      "epoch": 15.056774193548387,
+      "grad_norm": 0.6121147274971008,
+      "learning_rate": 9.953327571305099e-05,
+      "loss": 0.3342,
+      "step": 2911
+    },
+    {
+      "epoch": 15.061935483870968,
+      "grad_norm": 0.593255341053009,
+      "learning_rate": 9.949870354364736e-05,
+      "loss": 0.2915,
+      "step": 2912
+    },
+    {
+      "epoch": 15.067096774193548,
+      "grad_norm": 0.6040060520172119,
+      "learning_rate": 9.946413137424374e-05,
+      "loss": 0.3528,
+      "step": 2913
+    },
+    {
+      "epoch": 15.072258064516129,
+      "grad_norm": 0.48336148262023926,
+      "learning_rate": 9.942955920484011e-05,
+      "loss": 0.3386,
+      "step": 2914
+    },
+    {
+      "epoch": 15.07741935483871,
+      "grad_norm": 0.5369147658348083,
+      "learning_rate": 9.939498703543649e-05,
+      "loss": 0.298,
+      "step": 2915
+    },
+    {
+      "epoch": 15.08258064516129,
+      "grad_norm": 0.5753533840179443,
+      "learning_rate": 9.936041486603284e-05,
+      "loss": 0.2993,
+      "step": 2916
+    },
+    {
+      "epoch": 15.087741935483871,
+      "grad_norm": 0.5671182870864868,
+      "learning_rate": 9.932584269662922e-05,
+      "loss": 0.3034,
+      "step": 2917
+    },
+    {
+      "epoch": 15.092903225806452,
+      "grad_norm": 0.6327847838401794,
+      "learning_rate": 9.929127052722559e-05,
+      "loss": 0.3292,
+      "step": 2918
+    },
+    {
+      "epoch": 15.098064516129032,
+      "grad_norm": 0.5765933394432068,
+      "learning_rate": 9.925669835782196e-05,
+      "loss": 0.2957,
+      "step": 2919
+    },
+    {
+      "epoch": 15.103225806451613,
+      "grad_norm": 0.6619073748588562,
+      "learning_rate": 9.922212618841832e-05,
+      "loss": 0.2912,
+      "step": 2920
+    },
+    {
+      "epoch": 15.108387096774193,
+      "grad_norm": 0.6458188891410828,
+      "learning_rate": 9.918755401901469e-05,
+      "loss": 0.3176,
+      "step": 2921
+    },
+    {
+      "epoch": 15.113548387096774,
+      "grad_norm": 0.5286791920661926,
+      "learning_rate": 9.915298184961107e-05,
+      "loss": 0.2837,
+      "step": 2922
+    },
+    {
+      "epoch": 15.118709677419355,
+      "grad_norm": 0.6508303284645081,
+      "learning_rate": 9.911840968020744e-05,
+      "loss": 0.2982,
+      "step": 2923
+    },
+    {
+      "epoch": 15.123870967741935,
+      "grad_norm": 0.5904936790466309,
+      "learning_rate": 9.908383751080382e-05,
+      "loss": 0.3342,
+      "step": 2924
+    },
+    {
+      "epoch": 15.129032258064516,
+      "grad_norm": 0.5999349355697632,
+      "learning_rate": 9.904926534140017e-05,
+      "loss": 0.3087,
+      "step": 2925
+    },
+    {
+      "epoch": 15.134193548387097,
+      "grad_norm": 0.7085767388343811,
+      "learning_rate": 9.901469317199654e-05,
+      "loss": 0.3231,
+      "step": 2926
+    },
+    {
+      "epoch": 15.139354838709677,
+      "grad_norm": 0.6735929250717163,
+      "learning_rate": 9.898012100259292e-05,
+      "loss": 0.2979,
+      "step": 2927
+    },
+    {
+      "epoch": 15.144516129032258,
+      "grad_norm": 0.6373269557952881,
+      "learning_rate": 9.894554883318929e-05,
+      "loss": 0.3249,
+      "step": 2928
+    },
+    {
+      "epoch": 15.14967741935484,
+      "grad_norm": 0.6741179823875427,
+      "learning_rate": 9.891097666378565e-05,
+      "loss": 0.3279,
+      "step": 2929
+    },
+    {
+      "epoch": 15.154838709677419,
+      "grad_norm": 0.6077371835708618,
+      "learning_rate": 9.887640449438202e-05,
+      "loss": 0.3003,
+      "step": 2930
+    },
+    {
+      "epoch": 15.16,
+      "grad_norm": 0.6023492217063904,
+      "learning_rate": 9.88418323249784e-05,
+      "loss": 0.3005,
+      "step": 2931
+    },
+    {
+      "epoch": 15.165161290322581,
+      "grad_norm": 0.651076078414917,
+      "learning_rate": 9.880726015557477e-05,
+      "loss": 0.3123,
+      "step": 2932
+    },
+    {
+      "epoch": 15.170322580645161,
+      "grad_norm": 0.7591175436973572,
+      "learning_rate": 9.877268798617115e-05,
+      "loss": 0.3287,
+      "step": 2933
+    },
+    {
+      "epoch": 15.175483870967742,
+      "grad_norm": 0.571537435054779,
+      "learning_rate": 9.87381158167675e-05,
+      "loss": 0.3058,
+      "step": 2934
+    },
+    {
+      "epoch": 15.180645161290322,
+      "grad_norm": 0.6434149146080017,
+      "learning_rate": 9.870354364736387e-05,
+      "loss": 0.3335,
+      "step": 2935
+    },
+    {
+      "epoch": 15.185806451612903,
+      "grad_norm": 0.625181257724762,
+      "learning_rate": 9.866897147796025e-05,
+      "loss": 0.3312,
+      "step": 2936
+    },
+    {
+      "epoch": 15.190967741935484,
+      "grad_norm": 0.6476885080337524,
+      "learning_rate": 9.863439930855662e-05,
+      "loss": 0.321,
+      "step": 2937
+    },
+    {
+      "epoch": 15.196129032258064,
+      "grad_norm": 0.6397370100021362,
+      "learning_rate": 9.859982713915298e-05,
+      "loss": 0.3295,
+      "step": 2938
+    },
+    {
+      "epoch": 15.201290322580645,
+      "grad_norm": 0.6444936990737915,
+      "learning_rate": 9.856525496974935e-05,
+      "loss": 0.3379,
+      "step": 2939
+    },
+    {
+      "epoch": 15.206451612903226,
+      "grad_norm": 0.6370543241500854,
+      "learning_rate": 9.853068280034573e-05,
+      "loss": 0.3312,
+      "step": 2940
+    },
+    {
+      "epoch": 15.211612903225806,
+      "grad_norm": 0.6462299823760986,
+      "learning_rate": 9.84961106309421e-05,
+      "loss": 0.3457,
+      "step": 2941
+    },
+    {
+      "epoch": 15.216774193548387,
+      "grad_norm": 0.609442949295044,
+      "learning_rate": 9.846153846153848e-05,
+      "loss": 0.3195,
+      "step": 2942
+    },
+    {
+      "epoch": 15.221935483870968,
+      "grad_norm": 0.567013144493103,
+      "learning_rate": 9.842696629213483e-05,
+      "loss": 0.3157,
+      "step": 2943
+    },
+    {
+      "epoch": 15.227096774193548,
+      "grad_norm": 0.6591845750808716,
+      "learning_rate": 9.83923941227312e-05,
+      "loss": 0.3199,
+      "step": 2944
+    },
+    {
+      "epoch": 15.23225806451613,
+      "grad_norm": 0.6111241579055786,
+      "learning_rate": 9.835782195332758e-05,
+      "loss": 0.3224,
+      "step": 2945
+    },
+    {
+      "epoch": 15.23741935483871,
+      "grad_norm": 0.6711726188659668,
+      "learning_rate": 9.832324978392395e-05,
+      "loss": 0.3516,
+      "step": 2946
+    },
+    {
+      "epoch": 15.24258064516129,
+      "grad_norm": 0.5939470529556274,
+      "learning_rate": 9.828867761452031e-05,
+      "loss": 0.3261,
+      "step": 2947
+    },
+    {
+      "epoch": 15.247741935483871,
+      "grad_norm": 0.6146358251571655,
+      "learning_rate": 9.825410544511668e-05,
+      "loss": 0.3302,
+      "step": 2948
+    },
+    {
+      "epoch": 15.252903225806453,
+      "grad_norm": 0.6353687047958374,
+      "learning_rate": 9.821953327571306e-05,
+      "loss": 0.3246,
+      "step": 2949
+    },
+    {
+      "epoch": 15.258064516129032,
+      "grad_norm": 0.7398808598518372,
+      "learning_rate": 9.818496110630943e-05,
+      "loss": 0.337,
+      "step": 2950
+    },
+    {
+      "epoch": 15.263225806451613,
+      "grad_norm": 0.6109195351600647,
+      "learning_rate": 9.81503889369058e-05,
+      "loss": 0.2991,
+      "step": 2951
+    },
+    {
+      "epoch": 15.268387096774193,
+      "grad_norm": 0.6817329525947571,
+      "learning_rate": 9.811581676750216e-05,
+      "loss": 0.3319,
+      "step": 2952
+    },
+    {
+      "epoch": 15.273548387096774,
+      "grad_norm": 0.6378356218338013,
+      "learning_rate": 9.808124459809853e-05,
+      "loss": 0.3429,
+      "step": 2953
+    },
+    {
+      "epoch": 15.278709677419355,
+      "grad_norm": 0.6305023431777954,
+      "learning_rate": 9.804667242869491e-05,
+      "loss": 0.3088,
+      "step": 2954
+    },
+    {
+      "epoch": 15.283870967741935,
+      "grad_norm": 0.6302935481071472,
+      "learning_rate": 9.801210025929128e-05,
+      "loss": 0.2904,
+      "step": 2955
+    },
+    {
+      "epoch": 15.289032258064516,
+      "grad_norm": 0.6229572296142578,
+      "learning_rate": 9.797752808988764e-05,
+      "loss": 0.3261,
+      "step": 2956
+    },
+    {
+      "epoch": 15.294193548387097,
+      "grad_norm": 0.660942018032074,
+      "learning_rate": 9.794295592048401e-05,
+      "loss": 0.3138,
+      "step": 2957
+    },
+    {
+      "epoch": 15.299354838709677,
+      "grad_norm": 0.5603232979774475,
+      "learning_rate": 9.790838375108039e-05,
+      "loss": 0.3387,
+      "step": 2958
+    },
+    {
+      "epoch": 15.304516129032258,
+      "grad_norm": 0.6988780498504639,
+      "learning_rate": 9.787381158167676e-05,
+      "loss": 0.315,
+      "step": 2959
+    },
+    {
+      "epoch": 15.30967741935484,
+      "grad_norm": 0.6317124962806702,
+      "learning_rate": 9.783923941227312e-05,
+      "loss": 0.2943,
+      "step": 2960
+    },
+    {
+      "epoch": 15.314838709677419,
+      "grad_norm": 0.5586963295936584,
+      "learning_rate": 9.780466724286949e-05,
+      "loss": 0.2868,
+      "step": 2961
+    },
+    {
+      "epoch": 15.32,
+      "grad_norm": 0.6731350421905518,
+      "learning_rate": 9.777009507346586e-05,
+      "loss": 0.3465,
+      "step": 2962
+    },
+    {
+      "epoch": 15.32516129032258,
+      "grad_norm": 0.6238142251968384,
+      "learning_rate": 9.773552290406224e-05,
+      "loss": 0.3206,
+      "step": 2963
+    },
+    {
+      "epoch": 15.330322580645161,
+      "grad_norm": 0.630856990814209,
+      "learning_rate": 9.77009507346586e-05,
+      "loss": 0.336,
+      "step": 2964
+    },
+    {
+      "epoch": 15.335483870967742,
+      "grad_norm": 0.6861684918403625,
+      "learning_rate": 9.766637856525497e-05,
+      "loss": 0.3378,
+      "step": 2965
+    },
+    {
+      "epoch": 15.340645161290322,
+      "grad_norm": 0.6771178841590881,
+      "learning_rate": 9.763180639585134e-05,
+      "loss": 0.3316,
+      "step": 2966
+    },
+    {
+      "epoch": 15.345806451612903,
+      "grad_norm": 0.7071759700775146,
+      "learning_rate": 9.75972342264477e-05,
+      "loss": 0.3169,
+      "step": 2967
+    },
+    {
+      "epoch": 15.350967741935484,
+      "grad_norm": 0.5784715414047241,
+      "learning_rate": 9.756266205704409e-05,
+      "loss": 0.236,
+      "step": 2968
+    },
+    {
+      "epoch": 15.356129032258064,
+      "grad_norm": 0.623200535774231,
+      "learning_rate": 9.752808988764045e-05,
+      "loss": 0.3237,
+      "step": 2969
+    },
+    {
+      "epoch": 15.361290322580645,
+      "grad_norm": 0.5639144778251648,
+      "learning_rate": 9.749351771823682e-05,
+      "loss": 0.3334,
+      "step": 2970
+    },
+    {
+      "epoch": 15.366451612903226,
+      "grad_norm": 0.6318480372428894,
+      "learning_rate": 9.745894554883319e-05,
+      "loss": 0.3548,
+      "step": 2971
+    },
+    {
+      "epoch": 15.371612903225806,
+      "grad_norm": 0.6293890476226807,
+      "learning_rate": 9.742437337942957e-05,
+      "loss": 0.3059,
+      "step": 2972
+    },
+    {
+      "epoch": 15.376774193548387,
+      "grad_norm": 0.6848973631858826,
+      "learning_rate": 9.738980121002594e-05,
+      "loss": 0.3549,
+      "step": 2973
+    },
+    {
+      "epoch": 15.381935483870969,
+      "grad_norm": 0.5913729667663574,
+      "learning_rate": 9.73552290406223e-05,
+      "loss": 0.3271,
+      "step": 2974
+    },
+    {
+      "epoch": 15.387096774193548,
+      "grad_norm": 0.5727496147155762,
+      "learning_rate": 9.732065687121867e-05,
+      "loss": 0.3032,
+      "step": 2975
+    },
+    {
+      "epoch": 15.39225806451613,
+      "grad_norm": 0.6385748982429504,
+      "learning_rate": 9.728608470181504e-05,
+      "loss": 0.3413,
+      "step": 2976
+    },
+    {
+      "epoch": 15.397419354838709,
+      "grad_norm": 0.6295161843299866,
+      "learning_rate": 9.725151253241142e-05,
+      "loss": 0.3089,
+      "step": 2977
+    },
+    {
+      "epoch": 15.40258064516129,
+      "grad_norm": 0.627684473991394,
+      "learning_rate": 9.721694036300778e-05,
+      "loss": 0.3168,
+      "step": 2978
+    },
+    {
+      "epoch": 15.407741935483871,
+      "grad_norm": 0.6628747582435608,
+      "learning_rate": 9.718236819360415e-05,
+      "loss": 0.336,
+      "step": 2979
+    },
+    {
+      "epoch": 15.412903225806451,
+      "grad_norm": 0.5424785614013672,
+      "learning_rate": 9.714779602420052e-05,
+      "loss": 0.3047,
+      "step": 2980
+    },
+    {
+      "epoch": 15.418064516129032,
+      "grad_norm": 0.6313603520393372,
+      "learning_rate": 9.71132238547969e-05,
+      "loss": 0.313,
+      "step": 2981
+    },
+    {
+      "epoch": 15.423225806451613,
+      "grad_norm": 0.7324481010437012,
+      "learning_rate": 9.707865168539327e-05,
+      "loss": 0.3236,
+      "step": 2982
+    },
+    {
+      "epoch": 15.428387096774193,
+      "grad_norm": 0.680216372013092,
+      "learning_rate": 9.704407951598963e-05,
+      "loss": 0.3206,
+      "step": 2983
+    },
+    {
+      "epoch": 15.433548387096774,
+      "grad_norm": 0.6375837326049805,
+      "learning_rate": 9.7009507346586e-05,
+      "loss": 0.3023,
+      "step": 2984
+    },
+    {
+      "epoch": 15.438709677419356,
+      "grad_norm": 0.64161616563797,
+      "learning_rate": 9.697493517718237e-05,
+      "loss": 0.3325,
+      "step": 2985
+    },
+    {
+      "epoch": 15.443870967741935,
+      "grad_norm": 0.6425021290779114,
+      "learning_rate": 9.694036300777875e-05,
+      "loss": 0.309,
+      "step": 2986
+    },
+    {
+      "epoch": 15.449032258064516,
+      "grad_norm": 0.6659666299819946,
+      "learning_rate": 9.690579083837511e-05,
+      "loss": 0.3124,
+      "step": 2987
+    },
+    {
+      "epoch": 15.454193548387098,
+      "grad_norm": 0.6483174562454224,
+      "learning_rate": 9.687121866897148e-05,
+      "loss": 0.3058,
+      "step": 2988
+    },
+    {
+      "epoch": 15.459354838709677,
+      "grad_norm": 0.7812500596046448,
+      "learning_rate": 9.683664649956785e-05,
+      "loss": 0.3578,
+      "step": 2989
+    },
+    {
+      "epoch": 15.464516129032258,
+      "grad_norm": 0.6302819848060608,
+      "learning_rate": 9.680207433016423e-05,
+      "loss": 0.2989,
+      "step": 2990
+    },
+    {
+      "epoch": 15.469677419354838,
+      "grad_norm": 0.7095418572425842,
+      "learning_rate": 9.67675021607606e-05,
+      "loss": 0.3452,
+      "step": 2991
+    },
+    {
+      "epoch": 15.47483870967742,
+      "grad_norm": 0.7122939825057983,
+      "learning_rate": 9.673292999135696e-05,
+      "loss": 0.3336,
+      "step": 2992
+    },
+    {
+      "epoch": 15.48,
+      "grad_norm": 0.6943867206573486,
+      "learning_rate": 9.669835782195333e-05,
+      "loss": 0.3658,
+      "step": 2993
+    },
+    {
+      "epoch": 15.48516129032258,
+      "grad_norm": 0.6084474325180054,
+      "learning_rate": 9.66637856525497e-05,
+      "loss": 0.3416,
+      "step": 2994
+    },
+    {
+      "epoch": 15.490322580645161,
+      "grad_norm": 0.67331463098526,
+      "learning_rate": 9.662921348314608e-05,
+      "loss": 0.3469,
+      "step": 2995
+    },
+    {
+      "epoch": 15.495483870967742,
+      "grad_norm": 0.6982548832893372,
+      "learning_rate": 9.659464131374244e-05,
+      "loss": 0.3515,
+      "step": 2996
+    },
+    {
+      "epoch": 15.500645161290322,
+      "grad_norm": 0.6256914138793945,
+      "learning_rate": 9.656006914433881e-05,
+      "loss": 0.307,
+      "step": 2997
+    },
+    {
+      "epoch": 15.505806451612903,
+      "grad_norm": 0.670316219329834,
+      "learning_rate": 9.652549697493518e-05,
+      "loss": 0.3077,
+      "step": 2998
+    },
+    {
+      "epoch": 15.510967741935485,
+      "grad_norm": 0.6786924600601196,
+      "learning_rate": 9.649092480553155e-05,
+      "loss": 0.3376,
+      "step": 2999
+    },
+    {
+      "epoch": 15.516129032258064,
+      "grad_norm": 0.6918968558311462,
+      "learning_rate": 9.645635263612793e-05,
+      "loss": 0.3158,
+      "step": 3000
+    },
+    {
+      "epoch": 15.547096774193548,
+      "grad_norm": 0.6595795154571533,
+      "learning_rate": 9.642178046672429e-05,
+      "loss": 0.2809,
+      "step": 3001
+    },
+    {
+      "epoch": 15.55225806451613,
+      "grad_norm": 0.6755903363227844,
+      "learning_rate": 9.638720829732066e-05,
+      "loss": 0.3083,
+      "step": 3002
+    },
+    {
+      "epoch": 15.557419354838709,
+      "grad_norm": 0.6701955199241638,
+      "learning_rate": 9.635263612791703e-05,
+      "loss": 0.3058,
+      "step": 3003
+    },
+    {
+      "epoch": 15.56258064516129,
+      "grad_norm": 0.6478583812713623,
+      "learning_rate": 9.631806395851341e-05,
+      "loss": 0.3431,
+      "step": 3004
+    },
+    {
+      "epoch": 15.567741935483872,
+      "grad_norm": 0.6554808616638184,
+      "learning_rate": 9.628349178910977e-05,
+      "loss": 0.338,
+      "step": 3005
+    },
+    {
+      "epoch": 15.572903225806451,
+      "grad_norm": 0.6322799324989319,
+      "learning_rate": 9.624891961970614e-05,
+      "loss": 0.3037,
+      "step": 3006
+    },
+    {
+      "epoch": 15.578064516129032,
+      "grad_norm": 0.7269136309623718,
+      "learning_rate": 9.621434745030251e-05,
+      "loss": 0.3283,
+      "step": 3007
+    },
+    {
+      "epoch": 15.583225806451614,
+      "grad_norm": 0.6845406293869019,
+      "learning_rate": 9.617977528089888e-05,
+      "loss": 0.3321,
+      "step": 3008
+    },
+    {
+      "epoch": 15.588387096774193,
+      "grad_norm": 0.7120646834373474,
+      "learning_rate": 9.614520311149526e-05,
+      "loss": 0.3155,
+      "step": 3009
+    },
+    {
+      "epoch": 15.593548387096774,
+      "grad_norm": 0.6661931872367859,
+      "learning_rate": 9.611063094209162e-05,
+      "loss": 0.334,
+      "step": 3010
+    },
+    {
+      "epoch": 15.598709677419356,
+      "grad_norm": 0.6541386842727661,
+      "learning_rate": 9.607605877268799e-05,
+      "loss": 0.3246,
+      "step": 3011
+    },
+    {
+      "epoch": 15.603870967741935,
+      "grad_norm": 0.7569315433502197,
+      "learning_rate": 9.604148660328436e-05,
+      "loss": 0.3404,
+      "step": 3012
+    },
+    {
+      "epoch": 15.609032258064516,
+      "grad_norm": 0.68463534116745,
+      "learning_rate": 9.600691443388074e-05,
+      "loss": 0.3389,
+      "step": 3013
+    },
+    {
+      "epoch": 15.614193548387096,
+      "grad_norm": 0.6939074993133545,
+      "learning_rate": 9.59723422644771e-05,
+      "loss": 0.3316,
+      "step": 3014
+    },
+    {
+      "epoch": 15.619354838709677,
+      "grad_norm": 0.76363605260849,
+      "learning_rate": 9.593777009507347e-05,
+      "loss": 0.355,
+      "step": 3015
+    },
+    {
+      "epoch": 15.624516129032259,
+      "grad_norm": 0.6316330432891846,
+      "learning_rate": 9.590319792566984e-05,
+      "loss": 0.3347,
+      "step": 3016
+    },
+    {
+      "epoch": 15.629677419354838,
+      "grad_norm": 0.7613214254379272,
+      "learning_rate": 9.58686257562662e-05,
+      "loss": 0.3315,
+      "step": 3017
+    },
+    {
+      "epoch": 15.63483870967742,
+      "grad_norm": 0.6633583307266235,
+      "learning_rate": 9.583405358686259e-05,
+      "loss": 0.3071,
+      "step": 3018
+    },
+    {
+      "epoch": 15.64,
+      "grad_norm": 0.7411945462226868,
+      "learning_rate": 9.579948141745895e-05,
+      "loss": 0.3033,
+      "step": 3019
+    },
+    {
+      "epoch": 15.64516129032258,
+      "grad_norm": 0.5899577736854553,
+      "learning_rate": 9.576490924805532e-05,
+      "loss": 0.2849,
+      "step": 3020
+    },
+    {
+      "epoch": 15.650322580645161,
+      "grad_norm": 0.6736947894096375,
+      "learning_rate": 9.573033707865169e-05,
+      "loss": 0.346,
+      "step": 3021
+    },
+    {
+      "epoch": 15.655483870967743,
+      "grad_norm": 0.7388281226158142,
+      "learning_rate": 9.569576490924807e-05,
+      "loss": 0.3498,
+      "step": 3022
+    },
+    {
+      "epoch": 15.660645161290322,
+      "grad_norm": 0.6317737102508545,
+      "learning_rate": 9.566119273984443e-05,
+      "loss": 0.2873,
+      "step": 3023
+    },
+    {
+      "epoch": 15.665806451612903,
+      "grad_norm": 0.6894678473472595,
+      "learning_rate": 9.56266205704408e-05,
+      "loss": 0.368,
+      "step": 3024
+    },
+    {
+      "epoch": 15.670967741935485,
+      "grad_norm": 0.6782621145248413,
+      "learning_rate": 9.559204840103717e-05,
+      "loss": 0.3355,
+      "step": 3025
+    },
+    {
+      "epoch": 15.676129032258064,
+      "grad_norm": 0.6311705112457275,
+      "learning_rate": 9.555747623163354e-05,
+      "loss": 0.3129,
+      "step": 3026
+    },
+    {
+      "epoch": 15.681290322580645,
+      "grad_norm": 0.7138367295265198,
+      "learning_rate": 9.552290406222992e-05,
+      "loss": 0.3309,
+      "step": 3027
+    },
+    {
+      "epoch": 15.686451612903225,
+      "grad_norm": 0.6593238115310669,
+      "learning_rate": 9.548833189282628e-05,
+      "loss": 0.3212,
+      "step": 3028
+    },
+    {
+      "epoch": 15.691612903225806,
+      "grad_norm": 0.6668590307235718,
+      "learning_rate": 9.545375972342265e-05,
+      "loss": 0.3396,
+      "step": 3029
+    },
+    {
+      "epoch": 15.696774193548388,
+      "grad_norm": 0.7480952739715576,
+      "learning_rate": 9.541918755401902e-05,
+      "loss": 0.3483,
+      "step": 3030
+    },
+    {
+      "epoch": 15.701935483870967,
+      "grad_norm": 0.6629341840744019,
+      "learning_rate": 9.53846153846154e-05,
+      "loss": 0.3427,
+      "step": 3031
+    },
+    {
+      "epoch": 15.707096774193548,
+      "grad_norm": 0.7577745318412781,
+      "learning_rate": 9.535004321521176e-05,
+      "loss": 0.3323,
+      "step": 3032
+    },
+    {
+      "epoch": 15.71225806451613,
+      "grad_norm": 0.6911799907684326,
+      "learning_rate": 9.531547104580813e-05,
+      "loss": 0.3051,
+      "step": 3033
+    },
+    {
+      "epoch": 15.717419354838709,
+      "grad_norm": 0.7227224707603455,
+      "learning_rate": 9.52808988764045e-05,
+      "loss": 0.3553,
+      "step": 3034
+    },
+    {
+      "epoch": 15.72258064516129,
+      "grad_norm": 0.596792459487915,
+      "learning_rate": 9.524632670700087e-05,
+      "loss": 0.32,
+      "step": 3035
+    },
+    {
+      "epoch": 15.727741935483872,
+      "grad_norm": 0.6729329824447632,
+      "learning_rate": 9.521175453759725e-05,
+      "loss": 0.3076,
+      "step": 3036
+    },
+    {
+      "epoch": 15.732903225806451,
+      "grad_norm": 0.6923701763153076,
+      "learning_rate": 9.517718236819361e-05,
+      "loss": 0.3528,
+      "step": 3037
+    },
+    {
+      "epoch": 15.738064516129032,
+      "grad_norm": 0.7372665405273438,
+      "learning_rate": 9.514261019878998e-05,
+      "loss": 0.346,
+      "step": 3038
+    },
+    {
+      "epoch": 15.743225806451614,
+      "grad_norm": 0.6655067205429077,
+      "learning_rate": 9.510803802938635e-05,
+      "loss": 0.3477,
+      "step": 3039
+    },
+    {
+      "epoch": 15.748387096774193,
+      "grad_norm": 0.6788662075996399,
+      "learning_rate": 9.507346585998271e-05,
+      "loss": 0.3364,
+      "step": 3040
+    },
+    {
+      "epoch": 15.753548387096775,
+      "grad_norm": 0.6486172676086426,
+      "learning_rate": 9.50388936905791e-05,
+      "loss": 0.3038,
+      "step": 3041
+    },
+    {
+      "epoch": 15.758709677419354,
+      "grad_norm": 0.7051050066947937,
+      "learning_rate": 9.500432152117546e-05,
+      "loss": 0.329,
+      "step": 3042
+    },
+    {
+      "epoch": 15.763870967741935,
+      "grad_norm": 0.680097758769989,
+      "learning_rate": 9.496974935177183e-05,
+      "loss": 0.3079,
+      "step": 3043
+    },
+    {
+      "epoch": 15.769032258064517,
+      "grad_norm": 0.7872160077095032,
+      "learning_rate": 9.49351771823682e-05,
+      "loss": 0.341,
+      "step": 3044
+    },
+    {
+      "epoch": 15.774193548387096,
+      "grad_norm": 0.7142850160598755,
+      "learning_rate": 9.490060501296458e-05,
+      "loss": 0.2971,
+      "step": 3045
+    },
+    {
+      "epoch": 15.779354838709677,
+      "grad_norm": 0.6300473213195801,
+      "learning_rate": 9.486603284356094e-05,
+      "loss": 0.3321,
+      "step": 3046
+    },
+    {
+      "epoch": 15.784516129032259,
+      "grad_norm": 0.6334571838378906,
+      "learning_rate": 9.483146067415731e-05,
+      "loss": 0.297,
+      "step": 3047
+    },
+    {
+      "epoch": 15.789677419354838,
+      "grad_norm": 3.087852954864502,
+      "learning_rate": 9.479688850475368e-05,
+      "loss": 0.3543,
+      "step": 3048
+    },
+    {
+      "epoch": 15.79483870967742,
+      "grad_norm": 0.6848283410072327,
+      "learning_rate": 9.476231633535004e-05,
+      "loss": 0.3354,
+      "step": 3049
+    },
+    {
+      "epoch": 15.8,
+      "grad_norm": 0.7506647109985352,
+      "learning_rate": 9.472774416594642e-05,
+      "loss": 0.325,
+      "step": 3050
+    },
+    {
+      "epoch": 15.80516129032258,
+      "grad_norm": 0.6936522126197815,
+      "learning_rate": 9.469317199654279e-05,
+      "loss": 0.3486,
+      "step": 3051
+    },
+    {
+      "epoch": 15.810322580645161,
+      "grad_norm": 0.7076968550682068,
+      "learning_rate": 9.465859982713916e-05,
+      "loss": 0.317,
+      "step": 3052
+    },
+    {
+      "epoch": 15.815483870967743,
+      "grad_norm": 0.7079557180404663,
+      "learning_rate": 9.462402765773552e-05,
+      "loss": 0.3406,
+      "step": 3053
+    },
+    {
+      "epoch": 15.820645161290322,
+      "grad_norm": 0.7786715030670166,
+      "learning_rate": 9.45894554883319e-05,
+      "loss": 0.3134,
+      "step": 3054
+    },
+    {
+      "epoch": 15.825806451612904,
+      "grad_norm": 0.7337389588356018,
+      "learning_rate": 9.455488331892827e-05,
+      "loss": 0.3445,
+      "step": 3055
+    },
+    {
+      "epoch": 15.830967741935483,
+      "grad_norm": 0.7195107936859131,
+      "learning_rate": 9.452031114952463e-05,
+      "loss": 0.3433,
+      "step": 3056
+    },
+    {
+      "epoch": 15.836129032258064,
+      "grad_norm": 0.7201449275016785,
+      "learning_rate": 9.4485738980121e-05,
+      "loss": 0.3377,
+      "step": 3057
+    },
+    {
+      "epoch": 15.841290322580646,
+      "grad_norm": 0.7555221319198608,
+      "learning_rate": 9.445116681071737e-05,
+      "loss": 0.3675,
+      "step": 3058
+    },
+    {
+      "epoch": 15.846451612903225,
+      "grad_norm": 0.8080562949180603,
+      "learning_rate": 9.441659464131375e-05,
+      "loss": 0.3676,
+      "step": 3059
+    },
+    {
+      "epoch": 15.851612903225806,
+      "grad_norm": 0.6877976059913635,
+      "learning_rate": 9.438202247191012e-05,
+      "loss": 0.3626,
+      "step": 3060
+    },
+    {
+      "epoch": 15.856774193548388,
+      "grad_norm": 0.7391126751899719,
+      "learning_rate": 9.434745030250649e-05,
+      "loss": 0.3204,
+      "step": 3061
+    },
+    {
+      "epoch": 15.861935483870967,
+      "grad_norm": 0.6473966240882874,
+      "learning_rate": 9.431287813310285e-05,
+      "loss": 0.33,
+      "step": 3062
+    },
+    {
+      "epoch": 15.867096774193548,
+      "grad_norm": 0.7835131287574768,
+      "learning_rate": 9.427830596369924e-05,
+      "loss": 0.3364,
+      "step": 3063
+    },
+    {
+      "epoch": 15.87225806451613,
+      "grad_norm": 0.7254303693771362,
+      "learning_rate": 9.42437337942956e-05,
+      "loss": 0.3237,
+      "step": 3064
+    },
+    {
+      "epoch": 15.87741935483871,
+      "grad_norm": 0.769818127155304,
+      "learning_rate": 9.420916162489196e-05,
+      "loss": 0.3565,
+      "step": 3065
+    },
+    {
+      "epoch": 15.88258064516129,
+      "grad_norm": 0.7043030261993408,
+      "learning_rate": 9.417458945548834e-05,
+      "loss": 0.3379,
+      "step": 3066
+    },
+    {
+      "epoch": 15.88774193548387,
+      "grad_norm": 0.7641374468803406,
+      "learning_rate": 9.41400172860847e-05,
+      "loss": 0.3278,
+      "step": 3067
+    },
+    {
+      "epoch": 15.892903225806451,
+      "grad_norm": 0.7636243104934692,
+      "learning_rate": 9.410544511668108e-05,
+      "loss": 0.282,
+      "step": 3068
+    },
+    {
+      "epoch": 15.898064516129033,
+      "grad_norm": 0.7542673349380493,
+      "learning_rate": 9.407087294727744e-05,
+      "loss": 0.3653,
+      "step": 3069
+    },
+    {
+      "epoch": 15.903225806451612,
+      "grad_norm": 0.6580871343612671,
+      "learning_rate": 9.403630077787382e-05,
+      "loss": 0.3058,
+      "step": 3070
+    },
+    {
+      "epoch": 15.908387096774193,
+      "grad_norm": 0.7982466816902161,
+      "learning_rate": 9.400172860847018e-05,
+      "loss": 0.3135,
+      "step": 3071
+    },
+    {
+      "epoch": 15.913548387096775,
+      "grad_norm": 0.7087135910987854,
+      "learning_rate": 9.396715643906655e-05,
+      "loss": 0.3454,
+      "step": 3072
+    },
+    {
+      "epoch": 15.918709677419354,
+      "grad_norm": 0.7346602082252502,
+      "learning_rate": 9.393258426966293e-05,
+      "loss": 0.3565,
+      "step": 3073
+    },
+    {
+      "epoch": 15.923870967741935,
+      "grad_norm": 0.7902827262878418,
+      "learning_rate": 9.389801210025929e-05,
+      "loss": 0.363,
+      "step": 3074
+    },
+    {
+      "epoch": 15.929032258064517,
+      "grad_norm": 0.7298442721366882,
+      "learning_rate": 9.386343993085567e-05,
+      "loss": 0.3242,
+      "step": 3075
+    },
+    {
+      "epoch": 15.934193548387096,
+      "grad_norm": 0.7000704407691956,
+      "learning_rate": 9.382886776145203e-05,
+      "loss": 0.2969,
+      "step": 3076
+    },
+    {
+      "epoch": 15.939354838709678,
+      "grad_norm": 0.7790756225585938,
+      "learning_rate": 9.379429559204841e-05,
+      "loss": 0.328,
+      "step": 3077
+    },
+    {
+      "epoch": 15.944516129032259,
+      "grad_norm": 0.7005168199539185,
+      "learning_rate": 9.375972342264477e-05,
+      "loss": 0.3342,
+      "step": 3078
+    },
+    {
+      "epoch": 15.949677419354838,
+      "grad_norm": 0.6787707805633545,
+      "learning_rate": 9.372515125324115e-05,
+      "loss": 0.3123,
+      "step": 3079
+    },
+    {
+      "epoch": 15.95483870967742,
+      "grad_norm": 0.8257725238800049,
+      "learning_rate": 9.369057908383751e-05,
+      "loss": 0.3217,
+      "step": 3080
+    },
+    {
+      "epoch": 15.96,
+      "grad_norm": 0.7420302629470825,
+      "learning_rate": 9.365600691443388e-05,
+      "loss": 0.3408,
+      "step": 3081
+    },
+    {
+      "epoch": 15.96516129032258,
+      "grad_norm": 0.7497138381004333,
+      "learning_rate": 9.362143474503026e-05,
+      "loss": 0.3296,
+      "step": 3082
+    },
+    {
+      "epoch": 15.970322580645162,
+      "grad_norm": 0.6898646354675293,
+      "learning_rate": 9.358686257562662e-05,
+      "loss": 0.3415,
+      "step": 3083
+    },
+    {
+      "epoch": 15.975483870967741,
+      "grad_norm": 0.6627447009086609,
+      "learning_rate": 9.3552290406223e-05,
+      "loss": 0.3142,
+      "step": 3084
+    },
+    {
+      "epoch": 15.980645161290322,
+      "grad_norm": 0.7508030533790588,
+      "learning_rate": 9.351771823681936e-05,
+      "loss": 0.3531,
+      "step": 3085
+    },
+    {
+      "epoch": 15.985806451612904,
+      "grad_norm": 0.7209461331367493,
+      "learning_rate": 9.348314606741574e-05,
+      "loss": 0.3406,
+      "step": 3086
+    },
+    {
+      "epoch": 15.990967741935483,
+      "grad_norm": 0.7673912048339844,
+      "learning_rate": 9.34485738980121e-05,
+      "loss": 0.3638,
+      "step": 3087
+    },
+    {
+      "epoch": 15.996129032258064,
+      "grad_norm": 0.7364399433135986,
+      "learning_rate": 9.341400172860848e-05,
+      "loss": 0.341,
+      "step": 3088
+    },
+    {
+      "epoch": 16.001290322580644,
+      "grad_norm": 0.8151412606239319,
+      "learning_rate": 9.337942955920484e-05,
+      "loss": 0.3724,
+      "step": 3089
+    },
+    {
+      "epoch": 16.006451612903227,
+      "grad_norm": 0.4582073986530304,
+      "learning_rate": 9.334485738980121e-05,
+      "loss": 0.2996,
+      "step": 3090
+    },
+    {
+      "epoch": 16.011612903225807,
+      "grad_norm": 0.49202093482017517,
+      "learning_rate": 9.331028522039759e-05,
+      "loss": 0.2933,
+      "step": 3091
+    },
+    {
+      "epoch": 16.016774193548386,
+      "grad_norm": 0.49170759320259094,
+      "learning_rate": 9.327571305099395e-05,
+      "loss": 0.3228,
+      "step": 3092
+    },
+    {
+      "epoch": 16.02193548387097,
+      "grad_norm": 0.4861731231212616,
+      "learning_rate": 9.324114088159033e-05,
+      "loss": 0.3178,
+      "step": 3093
+    },
+    {
+      "epoch": 16.02709677419355,
+      "grad_norm": 0.5125681757926941,
+      "learning_rate": 9.32065687121867e-05,
+      "loss": 0.2927,
+      "step": 3094
+    },
+    {
+      "epoch": 16.032258064516128,
+      "grad_norm": 0.6704266667366028,
+      "learning_rate": 9.317199654278307e-05,
+      "loss": 0.2977,
+      "step": 3095
+    },
+    {
+      "epoch": 16.03741935483871,
+      "grad_norm": 0.6088831424713135,
+      "learning_rate": 9.313742437337943e-05,
+      "loss": 0.3137,
+      "step": 3096
+    },
+    {
+      "epoch": 16.04258064516129,
+      "grad_norm": 0.5438926815986633,
+      "learning_rate": 9.31028522039758e-05,
+      "loss": 0.3051,
+      "step": 3097
+    },
+    {
+      "epoch": 16.04774193548387,
+      "grad_norm": 0.554445207118988,
+      "learning_rate": 9.306828003457217e-05,
+      "loss": 0.2866,
+      "step": 3098
+    },
+    {
+      "epoch": 16.052903225806453,
+      "grad_norm": 0.5734357237815857,
+      "learning_rate": 9.303370786516854e-05,
+      "loss": 0.2922,
+      "step": 3099
+    },
+    {
+      "epoch": 16.058064516129033,
+      "grad_norm": 0.5545299649238586,
+      "learning_rate": 9.299913569576492e-05,
+      "loss": 0.3161,
+      "step": 3100
+    },
+    {
+      "epoch": 16.063225806451612,
+      "grad_norm": 0.4806796908378601,
+      "learning_rate": 9.296456352636128e-05,
+      "loss": 0.2492,
+      "step": 3101
+    },
+    {
+      "epoch": 16.068387096774195,
+      "grad_norm": 0.5364171862602234,
+      "learning_rate": 9.292999135695766e-05,
+      "loss": 0.3288,
+      "step": 3102
+    },
+    {
+      "epoch": 16.073548387096775,
+      "grad_norm": 0.5335555076599121,
+      "learning_rate": 9.289541918755402e-05,
+      "loss": 0.2771,
+      "step": 3103
+    },
+    {
+      "epoch": 16.078709677419354,
+      "grad_norm": 0.6102363467216492,
+      "learning_rate": 9.28608470181504e-05,
+      "loss": 0.3108,
+      "step": 3104
+    },
+    {
+      "epoch": 16.083870967741934,
+      "grad_norm": 0.47745949029922485,
+      "learning_rate": 9.282627484874676e-05,
+      "loss": 0.2947,
+      "step": 3105
+    },
+    {
+      "epoch": 16.089032258064517,
+      "grad_norm": 0.5433423519134521,
+      "learning_rate": 9.279170267934312e-05,
+      "loss": 0.3007,
+      "step": 3106
+    },
+    {
+      "epoch": 16.094193548387096,
+      "grad_norm": 0.5847306847572327,
+      "learning_rate": 9.27571305099395e-05,
+      "loss": 0.266,
+      "step": 3107
+    },
+    {
+      "epoch": 16.099354838709676,
+      "grad_norm": 0.5138210654258728,
+      "learning_rate": 9.272255834053587e-05,
+      "loss": 0.2676,
+      "step": 3108
+    },
+    {
+      "epoch": 16.10451612903226,
+      "grad_norm": 0.5128393769264221,
+      "learning_rate": 9.268798617113225e-05,
+      "loss": 0.3042,
+      "step": 3109
+    },
+    {
+      "epoch": 16.10967741935484,
+      "grad_norm": 0.5895670056343079,
+      "learning_rate": 9.26534140017286e-05,
+      "loss": 0.3292,
+      "step": 3110
+    },
+    {
+      "epoch": 16.114838709677418,
+      "grad_norm": 0.5863261222839355,
+      "learning_rate": 9.261884183232499e-05,
+      "loss": 0.3263,
+      "step": 3111
+    },
+    {
+      "epoch": 16.12,
+      "grad_norm": 0.5026370882987976,
+      "learning_rate": 9.258426966292135e-05,
+      "loss": 0.2978,
+      "step": 3112
+    },
+    {
+      "epoch": 16.12516129032258,
+      "grad_norm": 0.5615347027778625,
+      "learning_rate": 9.254969749351772e-05,
+      "loss": 0.2857,
+      "step": 3113
+    },
+    {
+      "epoch": 16.13032258064516,
+      "grad_norm": 0.49115949869155884,
+      "learning_rate": 9.251512532411409e-05,
+      "loss": 0.2773,
+      "step": 3114
+    },
+    {
+      "epoch": 16.135483870967743,
+      "grad_norm": 0.5946159362792969,
+      "learning_rate": 9.248055315471045e-05,
+      "loss": 0.3191,
+      "step": 3115
+    },
+    {
+      "epoch": 16.140645161290323,
+      "grad_norm": 0.5648359656333923,
+      "learning_rate": 9.244598098530683e-05,
+      "loss": 0.3055,
+      "step": 3116
+    },
+    {
+      "epoch": 16.145806451612902,
+      "grad_norm": 0.5172491669654846,
+      "learning_rate": 9.24114088159032e-05,
+      "loss": 0.303,
+      "step": 3117
+    },
+    {
+      "epoch": 16.150967741935485,
+      "grad_norm": 0.6130844354629517,
+      "learning_rate": 9.237683664649958e-05,
+      "loss": 0.2886,
+      "step": 3118
+    },
+    {
+      "epoch": 16.156129032258065,
+      "grad_norm": 0.6303712129592896,
+      "learning_rate": 9.234226447709594e-05,
+      "loss": 0.3211,
+      "step": 3119
+    },
+    {
+      "epoch": 16.161290322580644,
+      "grad_norm": 0.5358259081840515,
+      "learning_rate": 9.230769230769232e-05,
+      "loss": 0.2912,
+      "step": 3120
+    },
+    {
+      "epoch": 16.166451612903227,
+      "grad_norm": 0.579641580581665,
+      "learning_rate": 9.227312013828868e-05,
+      "loss": 0.3017,
+      "step": 3121
+    },
+    {
+      "epoch": 16.171612903225807,
+      "grad_norm": 0.5237424373626709,
+      "learning_rate": 9.223854796888505e-05,
+      "loss": 0.3022,
+      "step": 3122
+    },
+    {
+      "epoch": 16.176774193548386,
+      "grad_norm": 0.5556448101997375,
+      "learning_rate": 9.220397579948142e-05,
+      "loss": 0.2886,
+      "step": 3123
+    },
+    {
+      "epoch": 16.18193548387097,
+      "grad_norm": 0.5810863971710205,
+      "learning_rate": 9.216940363007778e-05,
+      "loss": 0.3105,
+      "step": 3124
+    },
+    {
+      "epoch": 16.18709677419355,
+      "grad_norm": 0.5875245928764343,
+      "learning_rate": 9.213483146067416e-05,
+      "loss": 0.3133,
+      "step": 3125
+    },
+    {
+      "epoch": 16.19225806451613,
+      "grad_norm": 0.5115110278129578,
+      "learning_rate": 9.210025929127053e-05,
+      "loss": 0.2947,
+      "step": 3126
+    },
+    {
+      "epoch": 16.19741935483871,
+      "grad_norm": 0.5858906507492065,
+      "learning_rate": 9.206568712186691e-05,
+      "loss": 0.3158,
+      "step": 3127
+    },
+    {
+      "epoch": 16.20258064516129,
+      "grad_norm": 0.6641242504119873,
+      "learning_rate": 9.203111495246327e-05,
+      "loss": 0.2939,
+      "step": 3128
+    },
+    {
+      "epoch": 16.20774193548387,
+      "grad_norm": 0.530570387840271,
+      "learning_rate": 9.199654278305963e-05,
+      "loss": 0.3182,
+      "step": 3129
+    },
+    {
+      "epoch": 16.21290322580645,
+      "grad_norm": 0.6051984429359436,
+      "learning_rate": 9.196197061365601e-05,
+      "loss": 0.3414,
+      "step": 3130
+    },
+    {
+      "epoch": 16.218064516129033,
+      "grad_norm": 0.556510329246521,
+      "learning_rate": 9.192739844425238e-05,
+      "loss": 0.313,
+      "step": 3131
+    },
+    {
+      "epoch": 16.223225806451612,
+      "grad_norm": 0.5609940886497498,
+      "learning_rate": 9.189282627484875e-05,
+      "loss": 0.3104,
+      "step": 3132
+    },
+    {
+      "epoch": 16.228387096774192,
+      "grad_norm": 0.5449407696723938,
+      "learning_rate": 9.185825410544511e-05,
+      "loss": 0.2953,
+      "step": 3133
+    },
+    {
+      "epoch": 16.233548387096775,
+      "grad_norm": 0.499726265668869,
+      "learning_rate": 9.18236819360415e-05,
+      "loss": 0.2992,
+      "step": 3134
+    },
+    {
+      "epoch": 16.238709677419354,
+      "grad_norm": 0.6127626895904541,
+      "learning_rate": 9.178910976663786e-05,
+      "loss": 0.3334,
+      "step": 3135
+    },
+    {
+      "epoch": 16.243870967741934,
+      "grad_norm": 0.7047120332717896,
+      "learning_rate": 9.175453759723424e-05,
+      "loss": 0.3511,
+      "step": 3136
+    },
+    {
+      "epoch": 16.249032258064517,
+      "grad_norm": 0.5080937743186951,
+      "learning_rate": 9.17199654278306e-05,
+      "loss": 0.263,
+      "step": 3137
+    },
+    {
+      "epoch": 16.254193548387097,
+      "grad_norm": 0.5723878145217896,
+      "learning_rate": 9.168539325842696e-05,
+      "loss": 0.2953,
+      "step": 3138
+    },
+    {
+      "epoch": 16.259354838709676,
+      "grad_norm": 0.520988941192627,
+      "learning_rate": 9.165082108902334e-05,
+      "loss": 0.322,
+      "step": 3139
+    },
+    {
+      "epoch": 16.26451612903226,
+      "grad_norm": 0.6093537211418152,
+      "learning_rate": 9.161624891961971e-05,
+      "loss": 0.3253,
+      "step": 3140
+    },
+    {
+      "epoch": 16.26967741935484,
+      "grad_norm": 0.6203081011772156,
+      "learning_rate": 9.158167675021608e-05,
+      "loss": 0.3167,
+      "step": 3141
+    },
+    {
+      "epoch": 16.274838709677418,
+      "grad_norm": 0.45916426181793213,
+      "learning_rate": 9.154710458081244e-05,
+      "loss": 0.2875,
+      "step": 3142
+    },
+    {
+      "epoch": 16.28,
+      "grad_norm": 0.5836644172668457,
+      "learning_rate": 9.151253241140882e-05,
+      "loss": 0.2845,
+      "step": 3143
+    },
+    {
+      "epoch": 16.28516129032258,
+      "grad_norm": 0.5439537763595581,
+      "learning_rate": 9.147796024200519e-05,
+      "loss": 0.3218,
+      "step": 3144
+    },
+    {
+      "epoch": 16.29032258064516,
+      "grad_norm": 0.6037158370018005,
+      "learning_rate": 9.144338807260156e-05,
+      "loss": 0.3057,
+      "step": 3145
+    },
+    {
+      "epoch": 16.295483870967743,
+      "grad_norm": 0.4946170747280121,
+      "learning_rate": 9.140881590319793e-05,
+      "loss": 0.2812,
+      "step": 3146
+    },
+    {
+      "epoch": 16.300645161290323,
+      "grad_norm": 0.5521804690361023,
+      "learning_rate": 9.137424373379429e-05,
+      "loss": 0.2893,
+      "step": 3147
+    },
+    {
+      "epoch": 16.305806451612902,
+      "grad_norm": 0.5124611854553223,
+      "learning_rate": 9.133967156439067e-05,
+      "loss": 0.3291,
+      "step": 3148
+    },
+    {
+      "epoch": 16.310967741935485,
+      "grad_norm": 0.5372530817985535,
+      "learning_rate": 9.130509939498704e-05,
+      "loss": 0.3135,
+      "step": 3149
+    },
+    {
+      "epoch": 16.316129032258065,
+      "grad_norm": 1.0328184366226196,
+      "learning_rate": 9.127052722558341e-05,
+      "loss": 0.315,
+      "step": 3150
+    },
+    {
+      "epoch": 16.321290322580644,
+      "grad_norm": 0.6567739248275757,
+      "learning_rate": 9.123595505617977e-05,
+      "loss": 0.329,
+      "step": 3151
+    },
+    {
+      "epoch": 16.326451612903227,
+      "grad_norm": 0.5675033926963806,
+      "learning_rate": 9.120138288677615e-05,
+      "loss": 0.3066,
+      "step": 3152
+    },
+    {
+      "epoch": 16.331612903225807,
+      "grad_norm": 0.6278571486473083,
+      "learning_rate": 9.116681071737252e-05,
+      "loss": 0.318,
+      "step": 3153
+    },
+    {
+      "epoch": 16.336774193548386,
+      "grad_norm": 0.561926007270813,
+      "learning_rate": 9.113223854796889e-05,
+      "loss": 0.3045,
+      "step": 3154
+    },
+    {
+      "epoch": 16.34193548387097,
+      "grad_norm": 0.5829252004623413,
+      "learning_rate": 9.109766637856526e-05,
+      "loss": 0.3107,
+      "step": 3155
+    },
+    {
+      "epoch": 16.34709677419355,
+      "grad_norm": 0.6026630401611328,
+      "learning_rate": 9.106309420916162e-05,
+      "loss": 0.3149,
+      "step": 3156
+    },
+    {
+      "epoch": 16.35225806451613,
+      "grad_norm": 0.6100703477859497,
+      "learning_rate": 9.1028522039758e-05,
+      "loss": 0.3018,
+      "step": 3157
+    },
+    {
+      "epoch": 16.35741935483871,
+      "grad_norm": 0.6121968626976013,
+      "learning_rate": 9.099394987035437e-05,
+      "loss": 0.333,
+      "step": 3158
+    },
+    {
+      "epoch": 16.36258064516129,
+      "grad_norm": 0.5662276148796082,
+      "learning_rate": 9.095937770095074e-05,
+      "loss": 0.3033,
+      "step": 3159
+    },
+    {
+      "epoch": 16.36774193548387,
+      "grad_norm": 0.513518214225769,
+      "learning_rate": 9.09248055315471e-05,
+      "loss": 0.3074,
+      "step": 3160
+    },
+    {
+      "epoch": 16.37290322580645,
+      "grad_norm": 0.5865176916122437,
+      "learning_rate": 9.089023336214348e-05,
+      "loss": 0.292,
+      "step": 3161
+    },
+    {
+      "epoch": 16.378064516129033,
+      "grad_norm": 0.5219014883041382,
+      "learning_rate": 9.085566119273985e-05,
+      "loss": 0.2138,
+      "step": 3162
+    },
+    {
+      "epoch": 16.383225806451613,
+      "grad_norm": 0.5069121718406677,
+      "learning_rate": 9.082108902333622e-05,
+      "loss": 0.2789,
+      "step": 3163
+    },
+    {
+      "epoch": 16.388387096774192,
+      "grad_norm": 0.6222741007804871,
+      "learning_rate": 9.078651685393259e-05,
+      "loss": 0.2603,
+      "step": 3164
+    },
+    {
+      "epoch": 16.393548387096775,
+      "grad_norm": 0.5547738671302795,
+      "learning_rate": 9.075194468452895e-05,
+      "loss": 0.3149,
+      "step": 3165
+    },
+    {
+      "epoch": 16.398709677419355,
+      "grad_norm": 0.556571900844574,
+      "learning_rate": 9.071737251512533e-05,
+      "loss": 0.2991,
+      "step": 3166
+    },
+    {
+      "epoch": 16.403870967741934,
+      "grad_norm": 0.575568675994873,
+      "learning_rate": 9.06828003457217e-05,
+      "loss": 0.293,
+      "step": 3167
+    },
+    {
+      "epoch": 16.409032258064517,
+      "grad_norm": 0.5391756296157837,
+      "learning_rate": 9.064822817631807e-05,
+      "loss": 0.2829,
+      "step": 3168
+    },
+    {
+      "epoch": 16.414193548387097,
+      "grad_norm": 0.5292108058929443,
+      "learning_rate": 9.061365600691443e-05,
+      "loss": 0.3327,
+      "step": 3169
+    },
+    {
+      "epoch": 16.419354838709676,
+      "grad_norm": 0.5922752022743225,
+      "learning_rate": 9.05790838375108e-05,
+      "loss": 0.2827,
+      "step": 3170
+    },
+    {
+      "epoch": 16.42451612903226,
+      "grad_norm": 0.5588935613632202,
+      "learning_rate": 9.054451166810718e-05,
+      "loss": 0.3089,
+      "step": 3171
+    },
+    {
+      "epoch": 16.42967741935484,
+      "grad_norm": 0.6358153223991394,
+      "learning_rate": 9.050993949870355e-05,
+      "loss": 0.3035,
+      "step": 3172
+    },
+    {
+      "epoch": 16.434838709677418,
+      "grad_norm": 0.6708637475967407,
+      "learning_rate": 9.047536732929992e-05,
+      "loss": 0.3147,
+      "step": 3173
+    },
+    {
+      "epoch": 16.44,
+      "grad_norm": 0.6099734306335449,
+      "learning_rate": 9.044079515989628e-05,
+      "loss": 0.3128,
+      "step": 3174
+    },
+    {
+      "epoch": 16.44516129032258,
+      "grad_norm": 0.5912486910820007,
+      "learning_rate": 9.040622299049266e-05,
+      "loss": 0.2888,
+      "step": 3175
+    },
+    {
+      "epoch": 16.45032258064516,
+      "grad_norm": 0.566074013710022,
+      "learning_rate": 9.037165082108903e-05,
+      "loss": 0.2724,
+      "step": 3176
+    },
+    {
+      "epoch": 16.455483870967743,
+      "grad_norm": 0.6188801527023315,
+      "learning_rate": 9.03370786516854e-05,
+      "loss": 0.2876,
+      "step": 3177
+    },
+    {
+      "epoch": 16.460645161290323,
+      "grad_norm": 0.5821521282196045,
+      "learning_rate": 9.030250648228176e-05,
+      "loss": 0.3092,
+      "step": 3178
+    },
+    {
+      "epoch": 16.465806451612902,
+      "grad_norm": 0.6088340282440186,
+      "learning_rate": 9.026793431287813e-05,
+      "loss": 0.3101,
+      "step": 3179
+    },
+    {
+      "epoch": 16.470967741935485,
+      "grad_norm": 0.546067476272583,
+      "learning_rate": 9.023336214347451e-05,
+      "loss": 0.2846,
+      "step": 3180
+    },
+    {
+      "epoch": 16.476129032258065,
+      "grad_norm": 0.6576258540153503,
+      "learning_rate": 9.019878997407088e-05,
+      "loss": 0.286,
+      "step": 3181
+    },
+    {
+      "epoch": 16.481290322580644,
+      "grad_norm": 0.6244344711303711,
+      "learning_rate": 9.016421780466725e-05,
+      "loss": 0.312,
+      "step": 3182
+    },
+    {
+      "epoch": 16.486451612903227,
+      "grad_norm": 0.6041354537010193,
+      "learning_rate": 9.012964563526361e-05,
+      "loss": 0.2947,
+      "step": 3183
+    },
+    {
+      "epoch": 16.491612903225807,
+      "grad_norm": 0.573753297328949,
+      "learning_rate": 9.009507346585999e-05,
+      "loss": 0.2999,
+      "step": 3184
+    },
+    {
+      "epoch": 16.496774193548386,
+      "grad_norm": 0.5356807708740234,
+      "learning_rate": 9.006050129645636e-05,
+      "loss": 0.2941,
+      "step": 3185
+    },
+    {
+      "epoch": 16.501935483870966,
+      "grad_norm": 1.2119585275650024,
+      "learning_rate": 9.002592912705273e-05,
+      "loss": 0.2908,
+      "step": 3186
+    },
+    {
+      "epoch": 16.50709677419355,
+      "grad_norm": 0.628084659576416,
+      "learning_rate": 8.99913569576491e-05,
+      "loss": 0.3279,
+      "step": 3187
+    },
+    {
+      "epoch": 16.51225806451613,
+      "grad_norm": 0.5484738945960999,
+      "learning_rate": 8.995678478824546e-05,
+      "loss": 0.3096,
+      "step": 3188
+    },
+    {
+      "epoch": 16.517419354838708,
+      "grad_norm": 0.5581099987030029,
+      "learning_rate": 8.992221261884184e-05,
+      "loss": 0.2677,
+      "step": 3189
+    },
+    {
+      "epoch": 16.52258064516129,
+      "grad_norm": 0.563603937625885,
+      "learning_rate": 8.988764044943821e-05,
+      "loss": 0.3042,
+      "step": 3190
+    },
+    {
+      "epoch": 16.52258064516129,
+      "eval_loss": 2.697432518005371,
+      "eval_runtime": 21.5953,
+      "eval_samples_per_second": 3.658,
+      "eval_steps_per_second": 0.463,
+      "step": 3190
+    },
+    {
+      "epoch": 16.52774193548387,
+      "grad_norm": 0.5307852625846863,
+      "learning_rate": 8.985306828003458e-05,
+      "loss": 0.2863,
+      "step": 3191
+    },
+    {
+      "epoch": 16.53290322580645,
+      "grad_norm": 0.6390893459320068,
+      "learning_rate": 8.981849611063094e-05,
+      "loss": 0.3294,
+      "step": 3192
+    },
+    {
+      "epoch": 16.538064516129033,
+      "grad_norm": 0.7376899719238281,
+      "learning_rate": 8.978392394122732e-05,
+      "loss": 0.3457,
+      "step": 3193
+    },
+    {
+      "epoch": 16.543225806451613,
+      "grad_norm": 0.6546809673309326,
+      "learning_rate": 8.974935177182369e-05,
+      "loss": 0.3222,
+      "step": 3194
+    },
+    {
+      "epoch": 16.548387096774192,
+      "grad_norm": 0.5920929908752441,
+      "learning_rate": 8.971477960242006e-05,
+      "loss": 0.2749,
+      "step": 3195
+    },
+    {
+      "epoch": 16.553548387096775,
+      "grad_norm": 0.6130476593971252,
+      "learning_rate": 8.968020743301642e-05,
+      "loss": 0.3189,
+      "step": 3196
+    },
+    {
+      "epoch": 16.558709677419355,
+      "grad_norm": 0.703342080116272,
+      "learning_rate": 8.964563526361279e-05,
+      "loss": 0.3055,
+      "step": 3197
+    },
+    {
+      "epoch": 16.563870967741934,
+      "grad_norm": 0.7187470197677612,
+      "learning_rate": 8.961106309420917e-05,
+      "loss": 0.3305,
+      "step": 3198
+    },
+    {
+      "epoch": 16.569032258064517,
+      "grad_norm": 0.6494333148002625,
+      "learning_rate": 8.957649092480554e-05,
+      "loss": 0.2983,
+      "step": 3199
+    },
+    {
+      "epoch": 16.574193548387097,
+      "grad_norm": 0.5322182178497314,
+      "learning_rate": 8.95419187554019e-05,
+      "loss": 0.3247,
+      "step": 3200
+    },
+    {
+      "epoch": 16.579354838709676,
+      "grad_norm": 0.5971571207046509,
+      "learning_rate": 8.950734658599827e-05,
+      "loss": 0.3119,
+      "step": 3201
+    },
+    {
+      "epoch": 16.58451612903226,
+      "grad_norm": 0.6519023776054382,
+      "learning_rate": 8.947277441659464e-05,
+      "loss": 0.3006,
+      "step": 3202
+    },
+    {
+      "epoch": 16.58967741935484,
+      "grad_norm": 0.6013271808624268,
+      "learning_rate": 8.943820224719102e-05,
+      "loss": 0.2974,
+      "step": 3203
+    },
+    {
+      "epoch": 16.59483870967742,
+      "grad_norm": 0.636152982711792,
+      "learning_rate": 8.940363007778739e-05,
+      "loss": 0.3105,
+      "step": 3204
+    },
+    {
+      "epoch": 16.6,
+      "grad_norm": 0.6307245492935181,
+      "learning_rate": 8.936905790838375e-05,
+      "loss": 0.3033,
+      "step": 3205
+    },
+    {
+      "epoch": 16.60516129032258,
+      "grad_norm": 0.5954098701477051,
+      "learning_rate": 8.933448573898012e-05,
+      "loss": 0.3156,
+      "step": 3206
+    },
+    {
+      "epoch": 16.61032258064516,
+      "grad_norm": 0.6063530445098877,
+      "learning_rate": 8.92999135695765e-05,
+      "loss": 0.323,
+      "step": 3207
+    },
+    {
+      "epoch": 16.615483870967743,
+      "grad_norm": 0.5939659476280212,
+      "learning_rate": 8.926534140017287e-05,
+      "loss": 0.3031,
+      "step": 3208
+    },
+    {
+      "epoch": 16.620645161290323,
+      "grad_norm": 0.6220158934593201,
+      "learning_rate": 8.923076923076924e-05,
+      "loss": 0.3096,
+      "step": 3209
+    },
+    {
+      "epoch": 16.625806451612902,
+      "grad_norm": 0.550224781036377,
+      "learning_rate": 8.91961970613656e-05,
+      "loss": 0.2893,
+      "step": 3210
+    },
+    {
+      "epoch": 16.630967741935486,
+      "grad_norm": 0.5978041291236877,
+      "learning_rate": 8.916162489196197e-05,
+      "loss": 0.3082,
+      "step": 3211
+    },
+    {
+      "epoch": 16.636129032258065,
+      "grad_norm": 0.5865963101387024,
+      "learning_rate": 8.912705272255835e-05,
+      "loss": 0.3147,
+      "step": 3212
+    },
+    {
+      "epoch": 16.641290322580645,
+      "grad_norm": 0.6818796992301941,
+      "learning_rate": 8.909248055315472e-05,
+      "loss": 0.3122,
+      "step": 3213
+    },
+    {
+      "epoch": 16.646451612903228,
+      "grad_norm": 0.6676187515258789,
+      "learning_rate": 8.905790838375108e-05,
+      "loss": 0.311,
+      "step": 3214
+    },
+    {
+      "epoch": 16.651612903225807,
+      "grad_norm": 0.6419578790664673,
+      "learning_rate": 8.902333621434745e-05,
+      "loss": 0.3267,
+      "step": 3215
+    },
+    {
+      "epoch": 16.656774193548387,
+      "grad_norm": 0.6815788149833679,
+      "learning_rate": 8.898876404494383e-05,
+      "loss": 0.3139,
+      "step": 3216
+    },
+    {
+      "epoch": 16.661935483870966,
+      "grad_norm": 0.6365479230880737,
+      "learning_rate": 8.89541918755402e-05,
+      "loss": 0.3096,
+      "step": 3217
+    },
+    {
+      "epoch": 16.66709677419355,
+      "grad_norm": 0.5664595365524292,
+      "learning_rate": 8.891961970613657e-05,
+      "loss": 0.304,
+      "step": 3218
+    },
+    {
+      "epoch": 16.67225806451613,
+      "grad_norm": 0.5613018274307251,
+      "learning_rate": 8.888504753673293e-05,
+      "loss": 0.3183,
+      "step": 3219
+    },
+    {
+      "epoch": 16.677419354838708,
+      "grad_norm": 0.6023182272911072,
+      "learning_rate": 8.88504753673293e-05,
+      "loss": 0.3257,
+      "step": 3220
+    },
+    {
+      "epoch": 16.68258064516129,
+      "grad_norm": 0.5971197485923767,
+      "learning_rate": 8.881590319792568e-05,
+      "loss": 0.2823,
+      "step": 3221
+    },
+    {
+      "epoch": 16.68774193548387,
+      "grad_norm": 0.6491447687149048,
+      "learning_rate": 8.878133102852205e-05,
+      "loss": 0.2989,
+      "step": 3222
+    },
+    {
+      "epoch": 16.69290322580645,
+      "grad_norm": 0.5260685682296753,
+      "learning_rate": 8.874675885911841e-05,
+      "loss": 0.2886,
+      "step": 3223
+    },
+    {
+      "epoch": 16.698064516129033,
+      "grad_norm": 0.6450685262680054,
+      "learning_rate": 8.871218668971478e-05,
+      "loss": 0.318,
+      "step": 3224
+    },
+    {
+      "epoch": 16.703225806451613,
+      "grad_norm": 0.6343095898628235,
+      "learning_rate": 8.867761452031116e-05,
+      "loss": 0.2972,
+      "step": 3225
+    },
+    {
+      "epoch": 16.708387096774192,
+      "grad_norm": 0.6042504906654358,
+      "learning_rate": 8.864304235090753e-05,
+      "loss": 0.3181,
+      "step": 3226
+    },
+    {
+      "epoch": 16.713548387096775,
+      "grad_norm": 0.6649156212806702,
+      "learning_rate": 8.86084701815039e-05,
+      "loss": 0.3045,
+      "step": 3227
+    },
+    {
+      "epoch": 16.718709677419355,
+      "grad_norm": 0.6544021368026733,
+      "learning_rate": 8.857389801210026e-05,
+      "loss": 0.2969,
+      "step": 3228
+    },
+    {
+      "epoch": 16.723870967741934,
+      "grad_norm": 0.6604134440422058,
+      "learning_rate": 8.853932584269663e-05,
+      "loss": 0.3066,
+      "step": 3229
+    },
+    {
+      "epoch": 16.729032258064517,
+      "grad_norm": 0.6988194584846497,
+      "learning_rate": 8.850475367329301e-05,
+      "loss": 0.3237,
+      "step": 3230
+    },
+    {
+      "epoch": 16.734193548387097,
+      "grad_norm": 0.6062127351760864,
+      "learning_rate": 8.847018150388938e-05,
+      "loss": 0.3077,
+      "step": 3231
+    },
+    {
+      "epoch": 16.739354838709676,
+      "grad_norm": 0.6318645477294922,
+      "learning_rate": 8.843560933448574e-05,
+      "loss": 0.3171,
+      "step": 3232
+    },
+    {
+      "epoch": 16.74451612903226,
+      "grad_norm": 0.5959681272506714,
+      "learning_rate": 8.840103716508211e-05,
+      "loss": 0.2949,
+      "step": 3233
+    },
+    {
+      "epoch": 16.74967741935484,
+      "grad_norm": 0.6136818528175354,
+      "learning_rate": 8.836646499567849e-05,
+      "loss": 0.3114,
+      "step": 3234
+    },
+    {
+      "epoch": 16.75483870967742,
+      "grad_norm": 0.6902125477790833,
+      "learning_rate": 8.833189282627486e-05,
+      "loss": 0.3369,
+      "step": 3235
+    },
+    {
+      "epoch": 16.76,
+      "grad_norm": 0.7156970500946045,
+      "learning_rate": 8.829732065687121e-05,
+      "loss": 0.3148,
+      "step": 3236
+    },
+    {
+      "epoch": 16.76516129032258,
+      "grad_norm": 0.6319866180419922,
+      "learning_rate": 8.826274848746759e-05,
+      "loss": 0.3255,
+      "step": 3237
+    },
+    {
+      "epoch": 16.77032258064516,
+      "grad_norm": 0.5923711061477661,
+      "learning_rate": 8.822817631806396e-05,
+      "loss": 0.2758,
+      "step": 3238
+    },
+    {
+      "epoch": 16.775483870967744,
+      "grad_norm": 0.657859206199646,
+      "learning_rate": 8.819360414866034e-05,
+      "loss": 0.2765,
+      "step": 3239
+    },
+    {
+      "epoch": 16.780645161290323,
+      "grad_norm": 0.6484367251396179,
+      "learning_rate": 8.815903197925671e-05,
+      "loss": 0.2876,
+      "step": 3240
+    },
+    {
+      "epoch": 16.785806451612903,
+      "grad_norm": 0.6037227511405945,
+      "learning_rate": 8.812445980985307e-05,
+      "loss": 0.2975,
+      "step": 3241
+    },
+    {
+      "epoch": 16.790967741935482,
+      "grad_norm": 0.631151556968689,
+      "learning_rate": 8.808988764044944e-05,
+      "loss": 0.3443,
+      "step": 3242
+    },
+    {
+      "epoch": 16.796129032258065,
+      "grad_norm": 0.6684116721153259,
+      "learning_rate": 8.805531547104581e-05,
+      "loss": 0.3097,
+      "step": 3243
+    },
+    {
+      "epoch": 16.801290322580645,
+      "grad_norm": 0.7250446081161499,
+      "learning_rate": 8.802074330164219e-05,
+      "loss": 0.3545,
+      "step": 3244
+    },
+    {
+      "epoch": 16.806451612903224,
+      "grad_norm": 0.6721551418304443,
+      "learning_rate": 8.798617113223854e-05,
+      "loss": 0.3114,
+      "step": 3245
+    },
+    {
+      "epoch": 16.811612903225807,
+      "grad_norm": 0.653684139251709,
+      "learning_rate": 8.795159896283492e-05,
+      "loss": 0.3307,
+      "step": 3246
+    },
+    {
+      "epoch": 16.816774193548387,
+      "grad_norm": 0.6950149536132812,
+      "learning_rate": 8.791702679343129e-05,
+      "loss": 0.3224,
+      "step": 3247
+    },
+    {
+      "epoch": 16.821935483870966,
+      "grad_norm": 0.654804527759552,
+      "learning_rate": 8.788245462402767e-05,
+      "loss": 0.3074,
+      "step": 3248
+    },
+    {
+      "epoch": 16.82709677419355,
+      "grad_norm": 0.7211017608642578,
+      "learning_rate": 8.784788245462404e-05,
+      "loss": 0.3315,
+      "step": 3249
+    },
+    {
+      "epoch": 16.83225806451613,
+      "grad_norm": 0.6143552660942078,
+      "learning_rate": 8.78133102852204e-05,
+      "loss": 0.2746,
+      "step": 3250
+    },
+    {
+      "epoch": 16.83741935483871,
+      "grad_norm": 0.5979039669036865,
+      "learning_rate": 8.777873811581677e-05,
+      "loss": 0.3217,
+      "step": 3251
+    },
+    {
+      "epoch": 16.84258064516129,
+      "grad_norm": 0.7053089737892151,
+      "learning_rate": 8.774416594641314e-05,
+      "loss": 0.3056,
+      "step": 3252
+    },
+    {
+      "epoch": 16.84774193548387,
+      "grad_norm": 0.705548882484436,
+      "learning_rate": 8.770959377700952e-05,
+      "loss": 0.3132,
+      "step": 3253
+    },
+    {
+      "epoch": 16.85290322580645,
+      "grad_norm": 0.6782170534133911,
+      "learning_rate": 8.767502160760587e-05,
+      "loss": 0.33,
+      "step": 3254
+    },
+    {
+      "epoch": 16.858064516129033,
+      "grad_norm": 0.6166731715202332,
+      "learning_rate": 8.764044943820225e-05,
+      "loss": 0.2575,
+      "step": 3255
+    },
+    {
+      "epoch": 16.863225806451613,
+      "grad_norm": 0.6191650032997131,
+      "learning_rate": 8.760587726879862e-05,
+      "loss": 0.3236,
+      "step": 3256
+    },
+    {
+      "epoch": 16.868387096774192,
+      "grad_norm": 0.7264975905418396,
+      "learning_rate": 8.7571305099395e-05,
+      "loss": 0.3288,
+      "step": 3257
+    },
+    {
+      "epoch": 16.873548387096776,
+      "grad_norm": 0.6192386150360107,
+      "learning_rate": 8.753673292999137e-05,
+      "loss": 0.3027,
+      "step": 3258
+    },
+    {
+      "epoch": 16.878709677419355,
+      "grad_norm": 0.6431710720062256,
+      "learning_rate": 8.750216076058772e-05,
+      "loss": 0.3455,
+      "step": 3259
+    },
+    {
+      "epoch": 16.883870967741935,
+      "grad_norm": 0.671798050403595,
+      "learning_rate": 8.74675885911841e-05,
+      "loss": 0.3265,
+      "step": 3260
+    },
+    {
+      "epoch": 16.889032258064518,
+      "grad_norm": 0.6786014437675476,
+      "learning_rate": 8.743301642178047e-05,
+      "loss": 0.3117,
+      "step": 3261
+    },
+    {
+      "epoch": 16.894193548387097,
+      "grad_norm": 0.6199482083320618,
+      "learning_rate": 8.739844425237685e-05,
+      "loss": 0.2946,
+      "step": 3262
+    },
+    {
+      "epoch": 16.899354838709677,
+      "grad_norm": 0.6392812132835388,
+      "learning_rate": 8.73638720829732e-05,
+      "loss": 0.3077,
+      "step": 3263
+    },
+    {
+      "epoch": 16.90451612903226,
+      "grad_norm": 0.6316039562225342,
+      "learning_rate": 8.732929991356958e-05,
+      "loss": 0.3155,
+      "step": 3264
+    },
+    {
+      "epoch": 16.90967741935484,
+      "grad_norm": 0.6187916398048401,
+      "learning_rate": 8.729472774416595e-05,
+      "loss": 0.3179,
+      "step": 3265
+    },
+    {
+      "epoch": 16.91483870967742,
+      "grad_norm": 0.6222001910209656,
+      "learning_rate": 8.726015557476233e-05,
+      "loss": 0.2735,
+      "step": 3266
+    },
+    {
+      "epoch": 16.92,
+      "grad_norm": 0.7168742418289185,
+      "learning_rate": 8.72255834053587e-05,
+      "loss": 0.3038,
+      "step": 3267
+    },
+    {
+      "epoch": 16.92516129032258,
+      "grad_norm": 0.7255122661590576,
+      "learning_rate": 8.719101123595505e-05,
+      "loss": 0.3182,
+      "step": 3268
+    },
+    {
+      "epoch": 16.93032258064516,
+      "grad_norm": 0.588271975517273,
+      "learning_rate": 8.715643906655143e-05,
+      "loss": 0.288,
+      "step": 3269
+    },
+    {
+      "epoch": 16.93548387096774,
+      "grad_norm": 0.6159766912460327,
+      "learning_rate": 8.71218668971478e-05,
+      "loss": 0.3177,
+      "step": 3270
+    },
+    {
+      "epoch": 16.940645161290323,
+      "grad_norm": 0.6404950022697449,
+      "learning_rate": 8.708729472774418e-05,
+      "loss": 0.3047,
+      "step": 3271
+    },
+    {
+      "epoch": 16.945806451612903,
+      "grad_norm": 0.6357382535934448,
+      "learning_rate": 8.705272255834053e-05,
+      "loss": 0.3053,
+      "step": 3272
+    },
+    {
+      "epoch": 16.950967741935482,
+      "grad_norm": 0.641202986240387,
+      "learning_rate": 8.701815038893691e-05,
+      "loss": 0.3153,
+      "step": 3273
+    },
+    {
+      "epoch": 16.956129032258065,
+      "grad_norm": 0.6648682355880737,
+      "learning_rate": 8.698357821953328e-05,
+      "loss": 0.3136,
+      "step": 3274
+    },
+    {
+      "epoch": 16.961290322580645,
+      "grad_norm": 0.6851704120635986,
+      "learning_rate": 8.694900605012965e-05,
+      "loss": 0.2916,
+      "step": 3275
+    },
+    {
+      "epoch": 16.966451612903224,
+      "grad_norm": 0.6322888731956482,
+      "learning_rate": 8.691443388072603e-05,
+      "loss": 0.312,
+      "step": 3276
+    },
+    {
+      "epoch": 16.971612903225807,
+      "grad_norm": 0.6859917640686035,
+      "learning_rate": 8.687986171132238e-05,
+      "loss": 0.311,
+      "step": 3277
+    },
+    {
+      "epoch": 16.976774193548387,
+      "grad_norm": 0.6500895023345947,
+      "learning_rate": 8.684528954191876e-05,
+      "loss": 0.3145,
+      "step": 3278
+    },
+    {
+      "epoch": 16.981935483870966,
+      "grad_norm": 0.6599621176719666,
+      "learning_rate": 8.681071737251513e-05,
+      "loss": 0.3309,
+      "step": 3279
+    },
+    {
+      "epoch": 16.98709677419355,
+      "grad_norm": 0.645237922668457,
+      "learning_rate": 8.677614520311151e-05,
+      "loss": 0.3103,
+      "step": 3280
+    },
+    {
+      "epoch": 16.99225806451613,
+      "grad_norm": 0.6161308288574219,
+      "learning_rate": 8.674157303370786e-05,
+      "loss": 0.3076,
+      "step": 3281
+    },
+    {
+      "epoch": 16.99741935483871,
+      "grad_norm": 0.7892633080482483,
+      "learning_rate": 8.670700086430424e-05,
+      "loss": 0.2825,
+      "step": 3282
+    },
+    {
+      "epoch": 17.003870967741936,
+      "grad_norm": 0.9475956559181213,
+      "learning_rate": 8.667242869490061e-05,
+      "loss": 0.3909,
+      "step": 3283
+    },
+    {
+      "epoch": 17.009032258064515,
+      "grad_norm": 0.34752166271209717,
+      "learning_rate": 8.663785652549698e-05,
+      "loss": 0.277,
+      "step": 3284
+    },
+    {
+      "epoch": 17.014193548387098,
+      "grad_norm": 0.45308351516723633,
+      "learning_rate": 8.660328435609336e-05,
+      "loss": 0.2961,
+      "step": 3285
+    },
+    {
+      "epoch": 17.019354838709678,
+      "grad_norm": 0.4471478760242462,
+      "learning_rate": 8.656871218668971e-05,
+      "loss": 0.307,
+      "step": 3286
+    },
+    {
+      "epoch": 17.024516129032257,
+      "grad_norm": 0.5134677886962891,
+      "learning_rate": 8.653414001728609e-05,
+      "loss": 0.2885,
+      "step": 3287
+    },
+    {
+      "epoch": 17.02967741935484,
+      "grad_norm": 0.4869097173213959,
+      "learning_rate": 8.649956784788246e-05,
+      "loss": 0.2997,
+      "step": 3288
+    },
+    {
+      "epoch": 17.03483870967742,
+      "grad_norm": 0.3757888376712799,
+      "learning_rate": 8.646499567847884e-05,
+      "loss": 0.2741,
+      "step": 3289
+    },
+    {
+      "epoch": 17.04,
+      "grad_norm": 0.3819665014743805,
+      "learning_rate": 8.643042350907519e-05,
+      "loss": 0.2959,
+      "step": 3290
+    },
+    {
+      "epoch": 17.045161290322582,
+      "grad_norm": 0.5199093222618103,
+      "learning_rate": 8.639585133967157e-05,
+      "loss": 0.3143,
+      "step": 3291
+    },
+    {
+      "epoch": 17.05032258064516,
+      "grad_norm": 0.4499667286872864,
+      "learning_rate": 8.636127917026794e-05,
+      "loss": 0.3046,
+      "step": 3292
+    },
+    {
+      "epoch": 17.05548387096774,
+      "grad_norm": 0.5093558430671692,
+      "learning_rate": 8.63267070008643e-05,
+      "loss": 0.3169,
+      "step": 3293
+    },
+    {
+      "epoch": 17.060645161290324,
+      "grad_norm": 0.4989658296108246,
+      "learning_rate": 8.629213483146069e-05,
+      "loss": 0.3014,
+      "step": 3294
+    },
+    {
+      "epoch": 17.065806451612904,
+      "grad_norm": 0.4801577627658844,
+      "learning_rate": 8.625756266205704e-05,
+      "loss": 0.3225,
+      "step": 3295
+    },
+    {
+      "epoch": 17.070967741935483,
+      "grad_norm": 0.4111575186252594,
+      "learning_rate": 8.622299049265342e-05,
+      "loss": 0.3196,
+      "step": 3296
+    },
+    {
+      "epoch": 17.076129032258063,
+      "grad_norm": 0.4290415942668915,
+      "learning_rate": 8.618841832324979e-05,
+      "loss": 0.2933,
+      "step": 3297
+    },
+    {
+      "epoch": 17.081290322580646,
+      "grad_norm": 0.5352602601051331,
+      "learning_rate": 8.615384615384617e-05,
+      "loss": 0.2778,
+      "step": 3298
+    },
+    {
+      "epoch": 17.086451612903225,
+      "grad_norm": 0.4382925033569336,
+      "learning_rate": 8.611927398444252e-05,
+      "loss": 0.3111,
+      "step": 3299
+    },
+    {
+      "epoch": 17.091612903225805,
+      "grad_norm": 0.4457886815071106,
+      "learning_rate": 8.608470181503889e-05,
+      "loss": 0.3036,
+      "step": 3300
+    },
+    {
+      "epoch": 17.096774193548388,
+      "grad_norm": 0.40923193097114563,
+      "learning_rate": 8.605012964563527e-05,
+      "loss": 0.3114,
+      "step": 3301
+    },
+    {
+      "epoch": 17.101935483870967,
+      "grad_norm": 0.46263381838798523,
+      "learning_rate": 8.601555747623164e-05,
+      "loss": 0.2751,
+      "step": 3302
+    },
+    {
+      "epoch": 17.107096774193547,
+      "grad_norm": 0.433341920375824,
+      "learning_rate": 8.598098530682802e-05,
+      "loss": 0.2736,
+      "step": 3303
+    },
+    {
+      "epoch": 17.11225806451613,
+      "grad_norm": 0.5492580533027649,
+      "learning_rate": 8.594641313742437e-05,
+      "loss": 0.2991,
+      "step": 3304
+    },
+    {
+      "epoch": 17.11741935483871,
+      "grad_norm": 0.5195711851119995,
+      "learning_rate": 8.591184096802075e-05,
+      "loss": 0.2964,
+      "step": 3305
+    },
+    {
+      "epoch": 17.12258064516129,
+      "grad_norm": 0.4521338641643524,
+      "learning_rate": 8.587726879861712e-05,
+      "loss": 0.3036,
+      "step": 3306
+    },
+    {
+      "epoch": 17.127741935483872,
+      "grad_norm": 0.4749140739440918,
+      "learning_rate": 8.58426966292135e-05,
+      "loss": 0.2698,
+      "step": 3307
+    },
+    {
+      "epoch": 17.13290322580645,
+      "grad_norm": 0.4340582489967346,
+      "learning_rate": 8.580812445980985e-05,
+      "loss": 0.3064,
+      "step": 3308
+    },
+    {
+      "epoch": 17.13806451612903,
+      "grad_norm": 0.4568285644054413,
+      "learning_rate": 8.577355229040622e-05,
+      "loss": 0.2756,
+      "step": 3309
+    },
+    {
+      "epoch": 17.143225806451614,
+      "grad_norm": 0.4400692582130432,
+      "learning_rate": 8.57389801210026e-05,
+      "loss": 0.2759,
+      "step": 3310
+    },
+    {
+      "epoch": 17.148387096774194,
+      "grad_norm": 0.41769376397132874,
+      "learning_rate": 8.570440795159897e-05,
+      "loss": 0.3014,
+      "step": 3311
+    },
+    {
+      "epoch": 17.153548387096773,
+      "grad_norm": 0.36209264397621155,
+      "learning_rate": 8.566983578219535e-05,
+      "loss": 0.279,
+      "step": 3312
+    },
+    {
+      "epoch": 17.158709677419356,
+      "grad_norm": 0.4372861385345459,
+      "learning_rate": 8.56352636127917e-05,
+      "loss": 0.2989,
+      "step": 3313
+    },
+    {
+      "epoch": 17.163870967741936,
+      "grad_norm": 0.5037857294082642,
+      "learning_rate": 8.560069144338808e-05,
+      "loss": 0.3054,
+      "step": 3314
+    },
+    {
+      "epoch": 17.169032258064515,
+      "grad_norm": 0.446138471364975,
+      "learning_rate": 8.556611927398445e-05,
+      "loss": 0.3079,
+      "step": 3315
+    },
+    {
+      "epoch": 17.174193548387098,
+      "grad_norm": 0.527722954750061,
+      "learning_rate": 8.553154710458081e-05,
+      "loss": 0.3181,
+      "step": 3316
+    },
+    {
+      "epoch": 17.179354838709678,
+      "grad_norm": 0.5389798879623413,
+      "learning_rate": 8.549697493517718e-05,
+      "loss": 0.2854,
+      "step": 3317
+    },
+    {
+      "epoch": 17.184516129032257,
+      "grad_norm": 0.530697762966156,
+      "learning_rate": 8.546240276577355e-05,
+      "loss": 0.3007,
+      "step": 3318
+    },
+    {
+      "epoch": 17.18967741935484,
+      "grad_norm": 0.4504929780960083,
+      "learning_rate": 8.542783059636993e-05,
+      "loss": 0.3265,
+      "step": 3319
+    },
+    {
+      "epoch": 17.19483870967742,
+      "grad_norm": 0.4308592677116394,
+      "learning_rate": 8.53932584269663e-05,
+      "loss": 0.3042,
+      "step": 3320
+    },
+    {
+      "epoch": 17.2,
+      "grad_norm": 0.37667447328567505,
+      "learning_rate": 8.535868625756268e-05,
+      "loss": 0.2522,
+      "step": 3321
+    },
+    {
+      "epoch": 17.205161290322582,
+      "grad_norm": 0.4766097366809845,
+      "learning_rate": 8.532411408815903e-05,
+      "loss": 0.2916,
+      "step": 3322
+    },
+    {
+      "epoch": 17.210322580645162,
+      "grad_norm": 0.47263041138648987,
+      "learning_rate": 8.528954191875541e-05,
+      "loss": 0.3075,
+      "step": 3323
+    },
+    {
+      "epoch": 17.21548387096774,
+      "grad_norm": 0.3863697946071625,
+      "learning_rate": 8.525496974935178e-05,
+      "loss": 0.2655,
+      "step": 3324
+    },
+    {
+      "epoch": 17.22064516129032,
+      "grad_norm": 0.454364538192749,
+      "learning_rate": 8.522039757994814e-05,
+      "loss": 0.3113,
+      "step": 3325
+    },
+    {
+      "epoch": 17.225806451612904,
+      "grad_norm": 0.4540804922580719,
+      "learning_rate": 8.518582541054451e-05,
+      "loss": 0.2992,
+      "step": 3326
+    },
+    {
+      "epoch": 17.230967741935483,
+      "grad_norm": 0.4258403480052948,
+      "learning_rate": 8.515125324114088e-05,
+      "loss": 0.2869,
+      "step": 3327
+    },
+    {
+      "epoch": 17.236129032258063,
+      "grad_norm": 0.47260013222694397,
+      "learning_rate": 8.511668107173726e-05,
+      "loss": 0.3086,
+      "step": 3328
+    },
+    {
+      "epoch": 17.241290322580646,
+      "grad_norm": 0.4680251479148865,
+      "learning_rate": 8.508210890233363e-05,
+      "loss": 0.3131,
+      "step": 3329
+    },
+    {
+      "epoch": 17.246451612903225,
+      "grad_norm": 0.4642832577228546,
+      "learning_rate": 8.504753673293e-05,
+      "loss": 0.3051,
+      "step": 3330
+    },
+    {
+      "epoch": 17.251612903225805,
+      "grad_norm": 0.4760434925556183,
+      "learning_rate": 8.501296456352636e-05,
+      "loss": 0.2875,
+      "step": 3331
+    },
+    {
+      "epoch": 17.256774193548388,
+      "grad_norm": 0.4039584696292877,
+      "learning_rate": 8.497839239412273e-05,
+      "loss": 0.2732,
+      "step": 3332
+    },
+    {
+      "epoch": 17.261935483870968,
+      "grad_norm": 0.448319673538208,
+      "learning_rate": 8.494382022471911e-05,
+      "loss": 0.3091,
+      "step": 3333
+    },
+    {
+      "epoch": 17.267096774193547,
+      "grad_norm": 0.49899929761886597,
+      "learning_rate": 8.490924805531547e-05,
+      "loss": 0.3129,
+      "step": 3334
+    },
+    {
+      "epoch": 17.27225806451613,
+      "grad_norm": 0.4087042808532715,
+      "learning_rate": 8.487467588591184e-05,
+      "loss": 0.2669,
+      "step": 3335
+    },
+    {
+      "epoch": 17.27741935483871,
+      "grad_norm": 0.4301292300224304,
+      "learning_rate": 8.484010371650821e-05,
+      "loss": 0.2979,
+      "step": 3336
+    },
+    {
+      "epoch": 17.28258064516129,
+      "grad_norm": 0.3576381802558899,
+      "learning_rate": 8.480553154710459e-05,
+      "loss": 0.2684,
+      "step": 3337
+    },
+    {
+      "epoch": 17.287741935483872,
+      "grad_norm": 0.4135667383670807,
+      "learning_rate": 8.477095937770096e-05,
+      "loss": 0.306,
+      "step": 3338
+    },
+    {
+      "epoch": 17.29290322580645,
+      "grad_norm": 0.4208326041698456,
+      "learning_rate": 8.473638720829734e-05,
+      "loss": 0.3109,
+      "step": 3339
+    },
+    {
+      "epoch": 17.29806451612903,
+      "grad_norm": 0.46706411242485046,
+      "learning_rate": 8.470181503889369e-05,
+      "loss": 0.3153,
+      "step": 3340
+    },
+    {
+      "epoch": 17.303225806451614,
+      "grad_norm": 0.45957744121551514,
+      "learning_rate": 8.466724286949006e-05,
+      "loss": 0.3011,
+      "step": 3341
+    },
+    {
+      "epoch": 17.308387096774194,
+      "grad_norm": 0.5267328023910522,
+      "learning_rate": 8.463267070008644e-05,
+      "loss": 0.2949,
+      "step": 3342
+    },
+    {
+      "epoch": 17.313548387096773,
+      "grad_norm": 0.493944376707077,
+      "learning_rate": 8.45980985306828e-05,
+      "loss": 0.2924,
+      "step": 3343
+    },
+    {
+      "epoch": 17.318709677419356,
+      "grad_norm": 0.46279069781303406,
+      "learning_rate": 8.456352636127917e-05,
+      "loss": 0.305,
+      "step": 3344
+    },
+    {
+      "epoch": 17.323870967741936,
+      "grad_norm": 0.46898016333580017,
+      "learning_rate": 8.452895419187554e-05,
+      "loss": 0.3107,
+      "step": 3345
+    },
+    {
+      "epoch": 17.329032258064515,
+      "grad_norm": 0.4735638499259949,
+      "learning_rate": 8.449438202247192e-05,
+      "loss": 0.325,
+      "step": 3346
+    },
+    {
+      "epoch": 17.3341935483871,
+      "grad_norm": 0.47329092025756836,
+      "learning_rate": 8.445980985306829e-05,
+      "loss": 0.3029,
+      "step": 3347
+    },
+    {
+      "epoch": 17.339354838709678,
+      "grad_norm": 0.4749067425727844,
+      "learning_rate": 8.442523768366465e-05,
+      "loss": 0.3011,
+      "step": 3348
+    },
+    {
+      "epoch": 17.344516129032257,
+      "grad_norm": 0.4379393756389618,
+      "learning_rate": 8.439066551426102e-05,
+      "loss": 0.2948,
+      "step": 3349
+    },
+    {
+      "epoch": 17.34967741935484,
+      "grad_norm": 0.5110873579978943,
+      "learning_rate": 8.435609334485739e-05,
+      "loss": 0.3023,
+      "step": 3350
+    },
+    {
+      "epoch": 17.35483870967742,
+      "grad_norm": 0.6162368059158325,
+      "learning_rate": 8.432152117545377e-05,
+      "loss": 0.3073,
+      "step": 3351
+    },
+    {
+      "epoch": 17.36,
+      "grad_norm": 0.4562493860721588,
+      "learning_rate": 8.428694900605013e-05,
+      "loss": 0.2652,
+      "step": 3352
+    },
+    {
+      "epoch": 17.36516129032258,
+      "grad_norm": 0.453313946723938,
+      "learning_rate": 8.42523768366465e-05,
+      "loss": 0.2855,
+      "step": 3353
+    },
+    {
+      "epoch": 17.370322580645162,
+      "grad_norm": 0.38786619901657104,
+      "learning_rate": 8.421780466724287e-05,
+      "loss": 0.273,
+      "step": 3354
+    },
+    {
+      "epoch": 17.37548387096774,
+      "grad_norm": 0.43358564376831055,
+      "learning_rate": 8.418323249783925e-05,
+      "loss": 0.2867,
+      "step": 3355
+    },
+    {
+      "epoch": 17.38064516129032,
+      "grad_norm": 0.4569670557975769,
+      "learning_rate": 8.414866032843562e-05,
+      "loss": 0.2897,
+      "step": 3356
+    },
+    {
+      "epoch": 17.385806451612904,
+      "grad_norm": 0.44823533296585083,
+      "learning_rate": 8.411408815903198e-05,
+      "loss": 0.2894,
+      "step": 3357
+    },
+    {
+      "epoch": 17.390967741935484,
+      "grad_norm": 0.46879827976226807,
+      "learning_rate": 8.407951598962835e-05,
+      "loss": 0.3004,
+      "step": 3358
+    },
+    {
+      "epoch": 17.396129032258063,
+      "grad_norm": 0.5192408561706543,
+      "learning_rate": 8.404494382022472e-05,
+      "loss": 0.2965,
+      "step": 3359
+    },
+    {
+      "epoch": 17.401290322580646,
+      "grad_norm": 0.4678780436515808,
+      "learning_rate": 8.40103716508211e-05,
+      "loss": 0.3149,
+      "step": 3360
+    },
+    {
+      "epoch": 17.406451612903226,
+      "grad_norm": 0.48043930530548096,
+      "learning_rate": 8.397579948141746e-05,
+      "loss": 0.3069,
+      "step": 3361
+    },
+    {
+      "epoch": 17.411612903225805,
+      "grad_norm": 0.5066922903060913,
+      "learning_rate": 8.394122731201383e-05,
+      "loss": 0.3212,
+      "step": 3362
+    },
+    {
+      "epoch": 17.416774193548388,
+      "grad_norm": 0.35122090578079224,
+      "learning_rate": 8.39066551426102e-05,
+      "loss": 0.278,
+      "step": 3363
+    },
+    {
+      "epoch": 17.421935483870968,
+      "grad_norm": 0.5160029530525208,
+      "learning_rate": 8.387208297320658e-05,
+      "loss": 0.3046,
+      "step": 3364
+    },
+    {
+      "epoch": 17.427096774193547,
+      "grad_norm": 0.45094963908195496,
+      "learning_rate": 8.383751080380295e-05,
+      "loss": 0.3089,
+      "step": 3365
+    },
+    {
+      "epoch": 17.43225806451613,
+      "grad_norm": 0.4063982665538788,
+      "learning_rate": 8.380293863439931e-05,
+      "loss": 0.2842,
+      "step": 3366
+    },
+    {
+      "epoch": 17.43741935483871,
+      "grad_norm": 0.4433140158653259,
+      "learning_rate": 8.376836646499568e-05,
+      "loss": 0.2807,
+      "step": 3367
+    },
+    {
+      "epoch": 17.44258064516129,
+      "grad_norm": 0.5358518958091736,
+      "learning_rate": 8.373379429559205e-05,
+      "loss": 0.2826,
+      "step": 3368
+    },
+    {
+      "epoch": 17.447741935483872,
+      "grad_norm": 0.45153963565826416,
+      "learning_rate": 8.369922212618843e-05,
+      "loss": 0.2885,
+      "step": 3369
+    },
+    {
+      "epoch": 17.452903225806452,
+      "grad_norm": 0.4325893819332123,
+      "learning_rate": 8.36646499567848e-05,
+      "loss": 0.2612,
+      "step": 3370
+    },
+    {
+      "epoch": 17.45806451612903,
+      "grad_norm": 0.3848820924758911,
+      "learning_rate": 8.363007778738116e-05,
+      "loss": 0.2727,
+      "step": 3371
+    },
+    {
+      "epoch": 17.463225806451614,
+      "grad_norm": 0.47629714012145996,
+      "learning_rate": 8.359550561797753e-05,
+      "loss": 0.3029,
+      "step": 3372
+    },
+    {
+      "epoch": 17.468387096774194,
+      "grad_norm": 0.4321513772010803,
+      "learning_rate": 8.35609334485739e-05,
+      "loss": 0.2828,
+      "step": 3373
+    },
+    {
+      "epoch": 17.473548387096773,
+      "grad_norm": 0.42980268597602844,
+      "learning_rate": 8.352636127917028e-05,
+      "loss": 0.2839,
+      "step": 3374
+    },
+    {
+      "epoch": 17.478709677419356,
+      "grad_norm": 0.3601219058036804,
+      "learning_rate": 8.349178910976664e-05,
+      "loss": 0.3126,
+      "step": 3375
+    },
+    {
+      "epoch": 17.483870967741936,
+      "grad_norm": 0.3678518831729889,
+      "learning_rate": 8.345721694036301e-05,
+      "loss": 0.2581,
+      "step": 3376
+    },
+    {
+      "epoch": 17.489032258064515,
+      "grad_norm": 0.4044246971607208,
+      "learning_rate": 8.342264477095938e-05,
+      "loss": 0.2577,
+      "step": 3377
+    },
+    {
+      "epoch": 17.4941935483871,
+      "grad_norm": 0.45011138916015625,
+      "learning_rate": 8.338807260155576e-05,
+      "loss": 0.3115,
+      "step": 3378
+    },
+    {
+      "epoch": 17.499354838709678,
+      "grad_norm": 0.48584887385368347,
+      "learning_rate": 8.335350043215212e-05,
+      "loss": 0.2782,
+      "step": 3379
+    },
+    {
+      "epoch": 17.504516129032258,
+      "grad_norm": 0.4131127595901489,
+      "learning_rate": 8.331892826274849e-05,
+      "loss": 0.2932,
+      "step": 3380
+    },
+    {
+      "epoch": 17.509677419354837,
+      "grad_norm": 0.46984079480171204,
+      "learning_rate": 8.328435609334486e-05,
+      "loss": 0.3243,
+      "step": 3381
+    },
+    {
+      "epoch": 17.51483870967742,
+      "grad_norm": 0.47009003162384033,
+      "learning_rate": 8.324978392394123e-05,
+      "loss": 0.27,
+      "step": 3382
+    },
+    {
+      "epoch": 17.52,
+      "grad_norm": 0.47334417700767517,
+      "learning_rate": 8.32152117545376e-05,
+      "loss": 0.3005,
+      "step": 3383
+    },
+    {
+      "epoch": 17.52516129032258,
+      "grad_norm": 0.5193254351615906,
+      "learning_rate": 8.318063958513397e-05,
+      "loss": 0.3031,
+      "step": 3384
+    },
+    {
+      "epoch": 17.530322580645162,
+      "grad_norm": 0.43811875581741333,
+      "learning_rate": 8.314606741573034e-05,
+      "loss": 0.2771,
+      "step": 3385
+    },
+    {
+      "epoch": 17.53548387096774,
+      "grad_norm": 0.35361966490745544,
+      "learning_rate": 8.311149524632671e-05,
+      "loss": 0.2896,
+      "step": 3386
+    },
+    {
+      "epoch": 17.54064516129032,
+      "grad_norm": 0.4528488516807556,
+      "learning_rate": 8.307692307692309e-05,
+      "loss": 0.2915,
+      "step": 3387
+    },
+    {
+      "epoch": 17.545806451612904,
+      "grad_norm": 0.4351736903190613,
+      "learning_rate": 8.304235090751945e-05,
+      "loss": 0.2759,
+      "step": 3388
+    },
+    {
+      "epoch": 17.550967741935484,
+      "grad_norm": 0.5007736086845398,
+      "learning_rate": 8.300777873811582e-05,
+      "loss": 0.2927,
+      "step": 3389
+    },
+    {
+      "epoch": 17.556129032258063,
+      "grad_norm": 0.5140421986579895,
+      "learning_rate": 8.297320656871219e-05,
+      "loss": 0.3067,
+      "step": 3390
+    },
+    {
+      "epoch": 17.561290322580646,
+      "grad_norm": 0.4494225084781647,
+      "learning_rate": 8.293863439930856e-05,
+      "loss": 0.2837,
+      "step": 3391
+    },
+    {
+      "epoch": 17.566451612903226,
+      "grad_norm": 0.4277782738208771,
+      "learning_rate": 8.290406222990494e-05,
+      "loss": 0.2927,
+      "step": 3392
+    },
+    {
+      "epoch": 17.571612903225805,
+      "grad_norm": 0.48042312264442444,
+      "learning_rate": 8.28694900605013e-05,
+      "loss": 0.27,
+      "step": 3393
+    },
+    {
+      "epoch": 17.57677419354839,
+      "grad_norm": 0.4967063367366791,
+      "learning_rate": 8.283491789109767e-05,
+      "loss": 0.2859,
+      "step": 3394
+    },
+    {
+      "epoch": 17.581935483870968,
+      "grad_norm": 0.4789804220199585,
+      "learning_rate": 8.280034572169404e-05,
+      "loss": 0.3163,
+      "step": 3395
+    },
+    {
+      "epoch": 17.587096774193547,
+      "grad_norm": 0.430132657289505,
+      "learning_rate": 8.276577355229042e-05,
+      "loss": 0.2905,
+      "step": 3396
+    },
+    {
+      "epoch": 17.59225806451613,
+      "grad_norm": 0.4169827401638031,
+      "learning_rate": 8.273120138288678e-05,
+      "loss": 0.2733,
+      "step": 3397
+    },
+    {
+      "epoch": 17.59741935483871,
+      "grad_norm": 0.44731107354164124,
+      "learning_rate": 8.269662921348315e-05,
+      "loss": 0.3149,
+      "step": 3398
+    },
+    {
+      "epoch": 17.60258064516129,
+      "grad_norm": 0.46114876866340637,
+      "learning_rate": 8.266205704407952e-05,
+      "loss": 0.2961,
+      "step": 3399
+    },
+    {
+      "epoch": 17.607741935483872,
+      "grad_norm": 0.4405445158481598,
+      "learning_rate": 8.262748487467589e-05,
+      "loss": 0.2765,
+      "step": 3400
+    },
+    {
+      "epoch": 17.612903225806452,
+      "grad_norm": 0.42725634574890137,
+      "learning_rate": 8.259291270527227e-05,
+      "loss": 0.3193,
+      "step": 3401
+    },
+    {
+      "epoch": 17.61806451612903,
+      "grad_norm": 0.48483437299728394,
+      "learning_rate": 8.255834053586863e-05,
+      "loss": 0.2744,
+      "step": 3402
+    },
+    {
+      "epoch": 17.623225806451615,
+      "grad_norm": 0.5362881422042847,
+      "learning_rate": 8.2523768366465e-05,
+      "loss": 0.2897,
+      "step": 3403
+    },
+    {
+      "epoch": 17.628387096774194,
+      "grad_norm": 0.4925309121608734,
+      "learning_rate": 8.248919619706137e-05,
+      "loss": 0.3151,
+      "step": 3404
+    },
+    {
+      "epoch": 17.633548387096774,
+      "grad_norm": 0.386502206325531,
+      "learning_rate": 8.245462402765773e-05,
+      "loss": 0.302,
+      "step": 3405
+    },
+    {
+      "epoch": 17.638709677419357,
+      "grad_norm": 0.48424550890922546,
+      "learning_rate": 8.242005185825411e-05,
+      "loss": 0.3075,
+      "step": 3406
+    },
+    {
+      "epoch": 17.643870967741936,
+      "grad_norm": 0.43090105056762695,
+      "learning_rate": 8.238547968885048e-05,
+      "loss": 0.2908,
+      "step": 3407
+    },
+    {
+      "epoch": 17.649032258064516,
+      "grad_norm": 0.5209680199623108,
+      "learning_rate": 8.235090751944685e-05,
+      "loss": 0.3087,
+      "step": 3408
+    },
+    {
+      "epoch": 17.654193548387095,
+      "grad_norm": 0.5059174299240112,
+      "learning_rate": 8.231633535004322e-05,
+      "loss": 0.2764,
+      "step": 3409
+    },
+    {
+      "epoch": 17.659354838709678,
+      "grad_norm": 0.45120394229888916,
+      "learning_rate": 8.22817631806396e-05,
+      "loss": 0.2783,
+      "step": 3410
+    },
+    {
+      "epoch": 17.664516129032258,
+      "grad_norm": 0.4493536651134491,
+      "learning_rate": 8.224719101123596e-05,
+      "loss": 0.2824,
+      "step": 3411
+    },
+    {
+      "epoch": 17.669677419354837,
+      "grad_norm": 0.4473881721496582,
+      "learning_rate": 8.221261884183233e-05,
+      "loss": 0.2892,
+      "step": 3412
+    },
+    {
+      "epoch": 17.67483870967742,
+      "grad_norm": 0.48013707995414734,
+      "learning_rate": 8.21780466724287e-05,
+      "loss": 0.2968,
+      "step": 3413
+    },
+    {
+      "epoch": 17.68,
+      "grad_norm": 0.4927014112472534,
+      "learning_rate": 8.214347450302506e-05,
+      "loss": 0.3027,
+      "step": 3414
+    },
+    {
+      "epoch": 17.68516129032258,
+      "grad_norm": 0.4563363790512085,
+      "learning_rate": 8.210890233362144e-05,
+      "loss": 0.29,
+      "step": 3415
+    },
+    {
+      "epoch": 17.690322580645162,
+      "grad_norm": 0.44865772128105164,
+      "learning_rate": 8.207433016421781e-05,
+      "loss": 0.3178,
+      "step": 3416
+    },
+    {
+      "epoch": 17.695483870967742,
+      "grad_norm": 0.49999868869781494,
+      "learning_rate": 8.203975799481418e-05,
+      "loss": 0.3119,
+      "step": 3417
+    },
+    {
+      "epoch": 17.70064516129032,
+      "grad_norm": 0.4579893946647644,
+      "learning_rate": 8.200518582541055e-05,
+      "loss": 0.2883,
+      "step": 3418
+    },
+    {
+      "epoch": 17.705806451612904,
+      "grad_norm": 0.4025906026363373,
+      "learning_rate": 8.197061365600693e-05,
+      "loss": 0.2889,
+      "step": 3419
+    },
+    {
+      "epoch": 17.710967741935484,
+      "grad_norm": 0.36701250076293945,
+      "learning_rate": 8.193604148660329e-05,
+      "loss": 0.2951,
+      "step": 3420
+    },
+    {
+      "epoch": 17.716129032258063,
+      "grad_norm": 0.5069311261177063,
+      "learning_rate": 8.190146931719965e-05,
+      "loss": 0.3056,
+      "step": 3421
+    },
+    {
+      "epoch": 17.721290322580646,
+      "grad_norm": 0.489298552274704,
+      "learning_rate": 8.186689714779603e-05,
+      "loss": 0.311,
+      "step": 3422
+    },
+    {
+      "epoch": 17.726451612903226,
+      "grad_norm": 0.48009219765663147,
+      "learning_rate": 8.18323249783924e-05,
+      "loss": 0.3072,
+      "step": 3423
+    },
+    {
+      "epoch": 17.731612903225805,
+      "grad_norm": 0.51910001039505,
+      "learning_rate": 8.179775280898877e-05,
+      "loss": 0.3028,
+      "step": 3424
+    },
+    {
+      "epoch": 17.73677419354839,
+      "grad_norm": 0.5127443671226501,
+      "learning_rate": 8.176318063958514e-05,
+      "loss": 0.3129,
+      "step": 3425
+    },
+    {
+      "epoch": 17.741935483870968,
+      "grad_norm": 0.541519284248352,
+      "learning_rate": 8.172860847018151e-05,
+      "loss": 0.2566,
+      "step": 3426
+    },
+    {
+      "epoch": 17.747096774193547,
+      "grad_norm": 0.36631396412849426,
+      "learning_rate": 8.169403630077788e-05,
+      "loss": 0.2803,
+      "step": 3427
+    },
+    {
+      "epoch": 17.75225806451613,
+      "grad_norm": 0.4387405812740326,
+      "learning_rate": 8.165946413137426e-05,
+      "loss": 0.2963,
+      "step": 3428
+    },
+    {
+      "epoch": 17.75741935483871,
+      "grad_norm": 0.5485689640045166,
+      "learning_rate": 8.162489196197062e-05,
+      "loss": 0.2998,
+      "step": 3429
+    },
+    {
+      "epoch": 17.76258064516129,
+      "grad_norm": 0.4360097348690033,
+      "learning_rate": 8.159031979256698e-05,
+      "loss": 0.2884,
+      "step": 3430
+    },
+    {
+      "epoch": 17.767741935483873,
+      "grad_norm": 0.4602543115615845,
+      "learning_rate": 8.155574762316336e-05,
+      "loss": 0.2888,
+      "step": 3431
+    },
+    {
+      "epoch": 17.772903225806452,
+      "grad_norm": 0.5047771334648132,
+      "learning_rate": 8.152117545375972e-05,
+      "loss": 0.3084,
+      "step": 3432
+    },
+    {
+      "epoch": 17.77806451612903,
+      "grad_norm": 0.5580325722694397,
+      "learning_rate": 8.14866032843561e-05,
+      "loss": 0.2878,
+      "step": 3433
+    },
+    {
+      "epoch": 17.78322580645161,
+      "grad_norm": 0.5021994113922119,
+      "learning_rate": 8.145203111495247e-05,
+      "loss": 0.2831,
+      "step": 3434
+    },
+    {
+      "epoch": 17.788387096774194,
+      "grad_norm": 0.4647403955459595,
+      "learning_rate": 8.141745894554884e-05,
+      "loss": 0.2951,
+      "step": 3435
+    },
+    {
+      "epoch": 17.793548387096774,
+      "grad_norm": 0.45285171270370483,
+      "learning_rate": 8.13828867761452e-05,
+      "loss": 0.2802,
+      "step": 3436
+    },
+    {
+      "epoch": 17.798709677419353,
+      "grad_norm": 0.42979010939598083,
+      "learning_rate": 8.134831460674159e-05,
+      "loss": 0.2969,
+      "step": 3437
+    },
+    {
+      "epoch": 17.803870967741936,
+      "grad_norm": 0.4559570252895355,
+      "learning_rate": 8.131374243733795e-05,
+      "loss": 0.2715,
+      "step": 3438
+    },
+    {
+      "epoch": 17.809032258064516,
+      "grad_norm": 0.42129290103912354,
+      "learning_rate": 8.12791702679343e-05,
+      "loss": 0.303,
+      "step": 3439
+    },
+    {
+      "epoch": 17.814193548387095,
+      "grad_norm": 0.46711137890815735,
+      "learning_rate": 8.124459809853069e-05,
+      "loss": 0.3061,
+      "step": 3440
+    },
+    {
+      "epoch": 17.81935483870968,
+      "grad_norm": 0.4110909104347229,
+      "learning_rate": 8.121002592912705e-05,
+      "loss": 0.2651,
+      "step": 3441
+    },
+    {
+      "epoch": 17.824516129032258,
+      "grad_norm": 0.3656541109085083,
+      "learning_rate": 8.117545375972343e-05,
+      "loss": 0.2912,
+      "step": 3442
+    },
+    {
+      "epoch": 17.829677419354837,
+      "grad_norm": 0.5098711848258972,
+      "learning_rate": 8.11408815903198e-05,
+      "loss": 0.2913,
+      "step": 3443
+    },
+    {
+      "epoch": 17.83483870967742,
+      "grad_norm": 0.5041640400886536,
+      "learning_rate": 8.110630942091617e-05,
+      "loss": 0.3285,
+      "step": 3444
+    },
+    {
+      "epoch": 17.84,
+      "grad_norm": 0.47169792652130127,
+      "learning_rate": 8.107173725151254e-05,
+      "loss": 0.3103,
+      "step": 3445
+    },
+    {
+      "epoch": 17.84516129032258,
+      "grad_norm": 0.5252098441123962,
+      "learning_rate": 8.10371650821089e-05,
+      "loss": 0.2985,
+      "step": 3446
+    },
+    {
+      "epoch": 17.850322580645162,
+      "grad_norm": 0.45653441548347473,
+      "learning_rate": 8.100259291270528e-05,
+      "loss": 0.3026,
+      "step": 3447
+    },
+    {
+      "epoch": 17.855483870967742,
+      "grad_norm": 0.45006605982780457,
+      "learning_rate": 8.096802074330164e-05,
+      "loss": 0.2897,
+      "step": 3448
+    },
+    {
+      "epoch": 17.86064516129032,
+      "grad_norm": 0.5284290909767151,
+      "learning_rate": 8.093344857389802e-05,
+      "loss": 0.2892,
+      "step": 3449
+    },
+    {
+      "epoch": 17.865806451612904,
+      "grad_norm": 0.4529808759689331,
+      "learning_rate": 8.089887640449438e-05,
+      "loss": 0.2634,
+      "step": 3450
+    },
+    {
+      "epoch": 17.870967741935484,
+      "grad_norm": 0.47922345995903015,
+      "learning_rate": 8.086430423509076e-05,
+      "loss": 0.2819,
+      "step": 3451
+    },
+    {
+      "epoch": 17.876129032258063,
+      "grad_norm": 0.391666054725647,
+      "learning_rate": 8.082973206568713e-05,
+      "loss": 0.2504,
+      "step": 3452
+    },
+    {
+      "epoch": 17.881290322580647,
+      "grad_norm": 0.4000386893749237,
+      "learning_rate": 8.07951598962835e-05,
+      "loss": 0.2798,
+      "step": 3453
+    },
+    {
+      "epoch": 17.886451612903226,
+      "grad_norm": 0.46385136246681213,
+      "learning_rate": 8.076058772687987e-05,
+      "loss": 0.2938,
+      "step": 3454
+    },
+    {
+      "epoch": 17.891612903225806,
+      "grad_norm": 0.5311865210533142,
+      "learning_rate": 8.072601555747623e-05,
+      "loss": 0.3074,
+      "step": 3455
+    },
+    {
+      "epoch": 17.89677419354839,
+      "grad_norm": 0.5114882588386536,
+      "learning_rate": 8.069144338807261e-05,
+      "loss": 0.3138,
+      "step": 3456
+    },
+    {
+      "epoch": 17.901935483870968,
+      "grad_norm": 0.4975427985191345,
+      "learning_rate": 8.065687121866897e-05,
+      "loss": 0.2856,
+      "step": 3457
+    },
+    {
+      "epoch": 17.907096774193548,
+      "grad_norm": 0.5150017738342285,
+      "learning_rate": 8.062229904926535e-05,
+      "loss": 0.2525,
+      "step": 3458
+    },
+    {
+      "epoch": 17.91225806451613,
+      "grad_norm": 0.4497613310813904,
+      "learning_rate": 8.058772687986171e-05,
+      "loss": 0.2588,
+      "step": 3459
+    },
+    {
+      "epoch": 17.91741935483871,
+      "grad_norm": 0.42354074120521545,
+      "learning_rate": 8.05531547104581e-05,
+      "loss": 0.288,
+      "step": 3460
+    },
+    {
+      "epoch": 17.92258064516129,
+      "grad_norm": 0.40374526381492615,
+      "learning_rate": 8.051858254105446e-05,
+      "loss": 0.2739,
+      "step": 3461
+    },
+    {
+      "epoch": 17.927741935483873,
+      "grad_norm": 0.4595211446285248,
+      "learning_rate": 8.048401037165081e-05,
+      "loss": 0.3097,
+      "step": 3462
+    },
+    {
+      "epoch": 17.932903225806452,
+      "grad_norm": 0.5078491568565369,
+      "learning_rate": 8.04494382022472e-05,
+      "loss": 0.2823,
+      "step": 3463
+    },
+    {
+      "epoch": 17.93806451612903,
+      "grad_norm": 0.4694339334964752,
+      "learning_rate": 8.041486603284356e-05,
+      "loss": 0.2881,
+      "step": 3464
+    },
+    {
+      "epoch": 17.94322580645161,
+      "grad_norm": 0.4404694437980652,
+      "learning_rate": 8.038029386343994e-05,
+      "loss": 0.314,
+      "step": 3465
+    },
+    {
+      "epoch": 17.948387096774194,
+      "grad_norm": 0.5277571678161621,
+      "learning_rate": 8.03457216940363e-05,
+      "loss": 0.2777,
+      "step": 3466
+    },
+    {
+      "epoch": 17.953548387096774,
+      "grad_norm": 0.41521546244621277,
+      "learning_rate": 8.031114952463268e-05,
+      "loss": 0.281,
+      "step": 3467
+    },
+    {
+      "epoch": 17.958709677419353,
+      "grad_norm": 0.502161979675293,
+      "learning_rate": 8.027657735522904e-05,
+      "loss": 0.2895,
+      "step": 3468
+    },
+    {
+      "epoch": 17.963870967741936,
+      "grad_norm": 0.5763131380081177,
+      "learning_rate": 8.024200518582542e-05,
+      "loss": 0.305,
+      "step": 3469
+    },
+    {
+      "epoch": 17.969032258064516,
+      "grad_norm": 0.5498910546302795,
+      "learning_rate": 8.020743301642179e-05,
+      "loss": 0.2973,
+      "step": 3470
+    },
+    {
+      "epoch": 17.974193548387095,
+      "grad_norm": 0.4518023729324341,
+      "learning_rate": 8.017286084701814e-05,
+      "loss": 0.3023,
+      "step": 3471
+    },
+    {
+      "epoch": 17.97935483870968,
+      "grad_norm": 0.49084752798080444,
+      "learning_rate": 8.013828867761453e-05,
+      "loss": 0.2963,
+      "step": 3472
+    },
+    {
+      "epoch": 17.984516129032258,
+      "grad_norm": 0.5043189525604248,
+      "learning_rate": 8.010371650821089e-05,
+      "loss": 0.3333,
+      "step": 3473
+    },
+    {
+      "epoch": 17.989677419354837,
+      "grad_norm": 0.5007174015045166,
+      "learning_rate": 8.006914433880727e-05,
+      "loss": 0.2806,
+      "step": 3474
+    },
+    {
+      "epoch": 17.99483870967742,
+      "grad_norm": 0.5057551860809326,
+      "learning_rate": 8.003457216940363e-05,
+      "loss": 0.2983,
+      "step": 3475
+    },
+    {
+      "epoch": 18.001290322580644,
+      "grad_norm": 0.7423113584518433,
+      "learning_rate": 8e-05,
+      "loss": 0.4122,
+      "step": 3476
+    },
+    {
+      "epoch": 18.006451612903227,
+      "grad_norm": 0.2906516194343567,
+      "learning_rate": 7.996542783059637e-05,
+      "loss": 0.2983,
+      "step": 3477
+    },
+    {
+      "epoch": 18.011612903225807,
+      "grad_norm": 0.27401769161224365,
+      "learning_rate": 7.993085566119274e-05,
+      "loss": 0.3284,
+      "step": 3478
+    },
+    {
+      "epoch": 18.016774193548386,
+      "grad_norm": 0.25640857219696045,
+      "learning_rate": 7.989628349178912e-05,
+      "loss": 0.273,
+      "step": 3479
+    },
+    {
+      "epoch": 18.02193548387097,
+      "grad_norm": 0.22300757467746735,
+      "learning_rate": 7.986171132238547e-05,
+      "loss": 0.3016,
+      "step": 3480
+    },
+    {
+      "epoch": 18.02193548387097,
+      "eval_loss": 2.782594680786133,
+      "eval_runtime": 21.1968,
+      "eval_samples_per_second": 3.727,
+      "eval_steps_per_second": 0.472,
+      "step": 3480
+    },
+    {
+      "epoch": 18.02709677419355,
+      "grad_norm": 0.2770066559314728,
+      "learning_rate": 7.982713915298186e-05,
+      "loss": 0.3096,
+      "step": 3481
+    },
+    {
+      "epoch": 18.032258064516128,
+      "grad_norm": 0.32072097063064575,
+      "learning_rate": 7.979256698357822e-05,
+      "loss": 0.2807,
+      "step": 3482
+    },
+    {
+      "epoch": 18.03741935483871,
+      "grad_norm": 0.2557767927646637,
+      "learning_rate": 7.97579948141746e-05,
+      "loss": 0.2986,
+      "step": 3483
+    },
+    {
+      "epoch": 18.04258064516129,
+      "grad_norm": 0.3642880320549011,
+      "learning_rate": 7.972342264477096e-05,
+      "loss": 0.2678,
+      "step": 3484
+    },
+    {
+      "epoch": 18.04774193548387,
+      "grad_norm": 0.3627167344093323,
+      "learning_rate": 7.968885047536734e-05,
+      "loss": 0.3023,
+      "step": 3485
+    },
+    {
+      "epoch": 18.052903225806453,
+      "grad_norm": 0.31499749422073364,
+      "learning_rate": 7.96542783059637e-05,
+      "loss": 0.309,
+      "step": 3486
+    },
+    {
+      "epoch": 18.058064516129033,
+      "grad_norm": 0.3198189437389374,
+      "learning_rate": 7.961970613656007e-05,
+      "loss": 0.3083,
+      "step": 3487
+    },
+    {
+      "epoch": 18.063225806451612,
+      "grad_norm": 0.2809082567691803,
+      "learning_rate": 7.958513396715645e-05,
+      "loss": 0.2689,
+      "step": 3488
+    },
+    {
+      "epoch": 18.068387096774195,
+      "grad_norm": 0.34369632601737976,
+      "learning_rate": 7.95505617977528e-05,
+      "loss": 0.2885,
+      "step": 3489
+    },
+    {
+      "epoch": 18.073548387096775,
+      "grad_norm": 0.3165997564792633,
+      "learning_rate": 7.951598962834918e-05,
+      "loss": 0.2575,
+      "step": 3490
+    },
+    {
+      "epoch": 18.078709677419354,
+      "grad_norm": 0.3204419016838074,
+      "learning_rate": 7.948141745894555e-05,
+      "loss": 0.2948,
+      "step": 3491
+    },
+    {
+      "epoch": 18.083870967741934,
+      "grad_norm": 0.3258916437625885,
+      "learning_rate": 7.944684528954193e-05,
+      "loss": 0.2676,
+      "step": 3492
+    },
+    {
+      "epoch": 18.089032258064517,
+      "grad_norm": 0.2459242045879364,
+      "learning_rate": 7.941227312013829e-05,
+      "loss": 0.2685,
+      "step": 3493
+    },
+    {
+      "epoch": 18.094193548387096,
+      "grad_norm": 0.2741316258907318,
+      "learning_rate": 7.937770095073465e-05,
+      "loss": 0.2755,
+      "step": 3494
+    },
+    {
+      "epoch": 18.099354838709676,
+      "grad_norm": 0.3203302323818207,
+      "learning_rate": 7.934312878133103e-05,
+      "loss": 0.2643,
+      "step": 3495
+    },
+    {
+      "epoch": 18.10451612903226,
+      "grad_norm": 0.27996641397476196,
+      "learning_rate": 7.93085566119274e-05,
+      "loss": 0.3244,
+      "step": 3496
+    },
+    {
+      "epoch": 18.10967741935484,
+      "grad_norm": 0.38399893045425415,
+      "learning_rate": 7.927398444252378e-05,
+      "loss": 0.2984,
+      "step": 3497
+    },
+    {
+      "epoch": 18.114838709677418,
+      "grad_norm": 0.3236656188964844,
+      "learning_rate": 7.923941227312013e-05,
+      "loss": 0.2896,
+      "step": 3498
+    },
+    {
+      "epoch": 18.12,
+      "grad_norm": 0.3240182399749756,
+      "learning_rate": 7.920484010371651e-05,
+      "loss": 0.3054,
+      "step": 3499
+    },
+    {
+      "epoch": 18.12516129032258,
+      "grad_norm": 0.2689698338508606,
+      "learning_rate": 7.917026793431288e-05,
+      "loss": 0.3078,
+      "step": 3500
+    },
+    {
+      "epoch": 18.13032258064516,
+      "grad_norm": 0.2810169756412506,
+      "learning_rate": 7.913569576490926e-05,
+      "loss": 0.2883,
+      "step": 3501
+    },
+    {
+      "epoch": 18.135483870967743,
+      "grad_norm": 0.3821760416030884,
+      "learning_rate": 7.910112359550562e-05,
+      "loss": 0.2775,
+      "step": 3502
+    },
+    {
+      "epoch": 18.140645161290323,
+      "grad_norm": 0.24314919114112854,
+      "learning_rate": 7.906655142610198e-05,
+      "loss": 0.2744,
+      "step": 3503
+    },
+    {
+      "epoch": 18.145806451612902,
+      "grad_norm": 0.39408430457115173,
+      "learning_rate": 7.903197925669836e-05,
+      "loss": 0.3114,
+      "step": 3504
+    },
+    {
+      "epoch": 18.150967741935485,
+      "grad_norm": 0.2756858766078949,
+      "learning_rate": 7.899740708729473e-05,
+      "loss": 0.2814,
+      "step": 3505
+    },
+    {
+      "epoch": 18.156129032258065,
+      "grad_norm": 0.33478987216949463,
+      "learning_rate": 7.896283491789111e-05,
+      "loss": 0.2788,
+      "step": 3506
+    },
+    {
+      "epoch": 18.161290322580644,
+      "grad_norm": 0.36581605672836304,
+      "learning_rate": 7.892826274848746e-05,
+      "loss": 0.2762,
+      "step": 3507
+    },
+    {
+      "epoch": 18.166451612903227,
+      "grad_norm": 0.31646987795829773,
+      "learning_rate": 7.889369057908384e-05,
+      "loss": 0.284,
+      "step": 3508
+    },
+    {
+      "epoch": 18.171612903225807,
+      "grad_norm": 0.1945173740386963,
+      "learning_rate": 7.885911840968021e-05,
+      "loss": 0.2593,
+      "step": 3509
+    },
+    {
+      "epoch": 18.176774193548386,
+      "grad_norm": 0.2963591516017914,
+      "learning_rate": 7.882454624027659e-05,
+      "loss": 0.2913,
+      "step": 3510
+    },
+    {
+      "epoch": 18.18193548387097,
+      "grad_norm": 0.2943360209465027,
+      "learning_rate": 7.878997407087295e-05,
+      "loss": 0.2926,
+      "step": 3511
+    },
+    {
+      "epoch": 18.18709677419355,
+      "grad_norm": 0.2847323715686798,
+      "learning_rate": 7.875540190146931e-05,
+      "loss": 0.2674,
+      "step": 3512
+    },
+    {
+      "epoch": 18.19225806451613,
+      "grad_norm": 0.3193649649620056,
+      "learning_rate": 7.87208297320657e-05,
+      "loss": 0.2864,
+      "step": 3513
+    },
+    {
+      "epoch": 18.19741935483871,
+      "grad_norm": 0.2700697183609009,
+      "learning_rate": 7.868625756266206e-05,
+      "loss": 0.2773,
+      "step": 3514
+    },
+    {
+      "epoch": 18.20258064516129,
+      "grad_norm": 0.27984899282455444,
+      "learning_rate": 7.865168539325843e-05,
+      "loss": 0.3121,
+      "step": 3515
+    },
+    {
+      "epoch": 18.20774193548387,
+      "grad_norm": 0.31418922543525696,
+      "learning_rate": 7.86171132238548e-05,
+      "loss": 0.2903,
+      "step": 3516
+    },
+    {
+      "epoch": 18.21290322580645,
+      "grad_norm": 0.20915654301643372,
+      "learning_rate": 7.858254105445117e-05,
+      "loss": 0.2632,
+      "step": 3517
+    },
+    {
+      "epoch": 18.218064516129033,
+      "grad_norm": 0.3052136301994324,
+      "learning_rate": 7.854796888504754e-05,
+      "loss": 0.2744,
+      "step": 3518
+    },
+    {
+      "epoch": 18.223225806451612,
+      "grad_norm": 0.30958759784698486,
+      "learning_rate": 7.851339671564391e-05,
+      "loss": 0.275,
+      "step": 3519
+    },
+    {
+      "epoch": 18.228387096774192,
+      "grad_norm": 0.2932756841182709,
+      "learning_rate": 7.847882454624028e-05,
+      "loss": 0.2518,
+      "step": 3520
+    },
+    {
+      "epoch": 18.233548387096775,
+      "grad_norm": 0.2923828065395355,
+      "learning_rate": 7.844425237683664e-05,
+      "loss": 0.2374,
+      "step": 3521
+    },
+    {
+      "epoch": 18.238709677419354,
+      "grad_norm": 0.24791376292705536,
+      "learning_rate": 7.840968020743302e-05,
+      "loss": 0.2847,
+      "step": 3522
+    },
+    {
+      "epoch": 18.243870967741934,
+      "grad_norm": 0.21642200648784637,
+      "learning_rate": 7.837510803802939e-05,
+      "loss": 0.2696,
+      "step": 3523
+    },
+    {
+      "epoch": 18.249032258064517,
+      "grad_norm": 0.31038975715637207,
+      "learning_rate": 7.834053586862576e-05,
+      "loss": 0.2506,
+      "step": 3524
+    },
+    {
+      "epoch": 18.254193548387097,
+      "grad_norm": 0.26951247453689575,
+      "learning_rate": 7.830596369922212e-05,
+      "loss": 0.2921,
+      "step": 3525
+    },
+    {
+      "epoch": 18.259354838709676,
+      "grad_norm": 0.34260618686676025,
+      "learning_rate": 7.82713915298185e-05,
+      "loss": 0.2938,
+      "step": 3526
+    },
+    {
+      "epoch": 18.26451612903226,
+      "grad_norm": 0.32290157675743103,
+      "learning_rate": 7.823681936041487e-05,
+      "loss": 0.2821,
+      "step": 3527
+    },
+    {
+      "epoch": 18.26967741935484,
+      "grad_norm": 0.38260477781295776,
+      "learning_rate": 7.820224719101124e-05,
+      "loss": 0.287,
+      "step": 3528
+    },
+    {
+      "epoch": 18.274838709677418,
+      "grad_norm": 0.374955415725708,
+      "learning_rate": 7.81676750216076e-05,
+      "loss": 0.305,
+      "step": 3529
+    },
+    {
+      "epoch": 18.28,
+      "grad_norm": 0.30697882175445557,
+      "learning_rate": 7.813310285220397e-05,
+      "loss": 0.2869,
+      "step": 3530
+    },
+    {
+      "epoch": 18.28516129032258,
+      "grad_norm": 0.2533535957336426,
+      "learning_rate": 7.809853068280035e-05,
+      "loss": 0.2897,
+      "step": 3531
+    },
+    {
+      "epoch": 18.29032258064516,
+      "grad_norm": 0.24751999974250793,
+      "learning_rate": 7.806395851339672e-05,
+      "loss": 0.2928,
+      "step": 3532
+    },
+    {
+      "epoch": 18.295483870967743,
+      "grad_norm": 0.29046669602394104,
+      "learning_rate": 7.802938634399309e-05,
+      "loss": 0.2957,
+      "step": 3533
+    },
+    {
+      "epoch": 18.300645161290323,
+      "grad_norm": 0.2802209258079529,
+      "learning_rate": 7.799481417458945e-05,
+      "loss": 0.2889,
+      "step": 3534
+    },
+    {
+      "epoch": 18.305806451612902,
+      "grad_norm": 0.28976863622665405,
+      "learning_rate": 7.796024200518582e-05,
+      "loss": 0.2719,
+      "step": 3535
+    },
+    {
+      "epoch": 18.310967741935485,
+      "grad_norm": 0.3124150037765503,
+      "learning_rate": 7.79256698357822e-05,
+      "loss": 0.2921,
+      "step": 3536
+    },
+    {
+      "epoch": 18.316129032258065,
+      "grad_norm": 0.2912261188030243,
+      "learning_rate": 7.789109766637857e-05,
+      "loss": 0.2883,
+      "step": 3537
+    },
+    {
+      "epoch": 18.321290322580644,
+      "grad_norm": 0.2908060550689697,
+      "learning_rate": 7.785652549697494e-05,
+      "loss": 0.2974,
+      "step": 3538
+    },
+    {
+      "epoch": 18.326451612903227,
+      "grad_norm": 0.24188432097434998,
+      "learning_rate": 7.78219533275713e-05,
+      "loss": 0.2926,
+      "step": 3539
+    },
+    {
+      "epoch": 18.331612903225807,
+      "grad_norm": 0.26194629073143005,
+      "learning_rate": 7.778738115816768e-05,
+      "loss": 0.2972,
+      "step": 3540
+    },
+    {
+      "epoch": 18.336774193548386,
+      "grad_norm": 0.31210917234420776,
+      "learning_rate": 7.775280898876405e-05,
+      "loss": 0.3027,
+      "step": 3541
+    },
+    {
+      "epoch": 18.34193548387097,
+      "grad_norm": 0.3217468857765198,
+      "learning_rate": 7.771823681936042e-05,
+      "loss": 0.2955,
+      "step": 3542
+    },
+    {
+      "epoch": 18.34709677419355,
+      "grad_norm": 0.28317904472351074,
+      "learning_rate": 7.768366464995678e-05,
+      "loss": 0.2548,
+      "step": 3543
+    },
+    {
+      "epoch": 18.35225806451613,
+      "grad_norm": 0.40764299035072327,
+      "learning_rate": 7.764909248055315e-05,
+      "loss": 0.2956,
+      "step": 3544
+    },
+    {
+      "epoch": 18.35741935483871,
+      "grad_norm": 0.23810113966464996,
+      "learning_rate": 7.761452031114953e-05,
+      "loss": 0.2833,
+      "step": 3545
+    },
+    {
+      "epoch": 18.36258064516129,
+      "grad_norm": 0.3602663278579712,
+      "learning_rate": 7.75799481417459e-05,
+      "loss": 0.3013,
+      "step": 3546
+    },
+    {
+      "epoch": 18.36774193548387,
+      "grad_norm": 0.3370330333709717,
+      "learning_rate": 7.754537597234227e-05,
+      "loss": 0.3033,
+      "step": 3547
+    },
+    {
+      "epoch": 18.37290322580645,
+      "grad_norm": 0.27209189534187317,
+      "learning_rate": 7.751080380293863e-05,
+      "loss": 0.2963,
+      "step": 3548
+    },
+    {
+      "epoch": 18.378064516129033,
+      "grad_norm": 0.3720322549343109,
+      "learning_rate": 7.747623163353501e-05,
+      "loss": 0.3008,
+      "step": 3549
+    },
+    {
+      "epoch": 18.383225806451613,
+      "grad_norm": 0.3089645504951477,
+      "learning_rate": 7.744165946413138e-05,
+      "loss": 0.2949,
+      "step": 3550
+    },
+    {
+      "epoch": 18.388387096774192,
+      "grad_norm": 0.3061186969280243,
+      "learning_rate": 7.740708729472775e-05,
+      "loss": 0.3088,
+      "step": 3551
+    },
+    {
+      "epoch": 18.393548387096775,
+      "grad_norm": 0.34414634108543396,
+      "learning_rate": 7.737251512532411e-05,
+      "loss": 0.3026,
+      "step": 3552
+    },
+    {
+      "epoch": 18.398709677419355,
+      "grad_norm": 0.2578818202018738,
+      "learning_rate": 7.733794295592048e-05,
+      "loss": 0.2753,
+      "step": 3553
+    },
+    {
+      "epoch": 18.403870967741934,
+      "grad_norm": 0.21969980001449585,
+      "learning_rate": 7.730337078651686e-05,
+      "loss": 0.2849,
+      "step": 3554
+    },
+    {
+      "epoch": 18.409032258064517,
+      "grad_norm": 0.3408031761646271,
+      "learning_rate": 7.726879861711323e-05,
+      "loss": 0.3035,
+      "step": 3555
+    },
+    {
+      "epoch": 18.414193548387097,
+      "grad_norm": 0.21376477181911469,
+      "learning_rate": 7.72342264477096e-05,
+      "loss": 0.3123,
+      "step": 3556
+    },
+    {
+      "epoch": 18.419354838709676,
+      "grad_norm": 0.29017481207847595,
+      "learning_rate": 7.719965427830596e-05,
+      "loss": 0.2933,
+      "step": 3557
+    },
+    {
+      "epoch": 18.42451612903226,
+      "grad_norm": 0.3053000867366791,
+      "learning_rate": 7.716508210890234e-05,
+      "loss": 0.2902,
+      "step": 3558
+    },
+    {
+      "epoch": 18.42967741935484,
+      "grad_norm": 0.36737060546875,
+      "learning_rate": 7.713050993949871e-05,
+      "loss": 0.3069,
+      "step": 3559
+    },
+    {
+      "epoch": 18.434838709677418,
+      "grad_norm": 0.2365802526473999,
+      "learning_rate": 7.709593777009508e-05,
+      "loss": 0.277,
+      "step": 3560
+    },
+    {
+      "epoch": 18.44,
+      "grad_norm": 0.30874505639076233,
+      "learning_rate": 7.706136560069144e-05,
+      "loss": 0.2854,
+      "step": 3561
+    },
+    {
+      "epoch": 18.44516129032258,
+      "grad_norm": 0.2855931222438812,
+      "learning_rate": 7.702679343128781e-05,
+      "loss": 0.2894,
+      "step": 3562
+    },
+    {
+      "epoch": 18.45032258064516,
+      "grad_norm": 0.3279050588607788,
+      "learning_rate": 7.699222126188419e-05,
+      "loss": 0.2614,
+      "step": 3563
+    },
+    {
+      "epoch": 18.455483870967743,
+      "grad_norm": 0.34520426392555237,
+      "learning_rate": 7.695764909248056e-05,
+      "loss": 0.2892,
+      "step": 3564
+    },
+    {
+      "epoch": 18.460645161290323,
+      "grad_norm": 0.3375611901283264,
+      "learning_rate": 7.692307692307693e-05,
+      "loss": 0.2695,
+      "step": 3565
+    },
+    {
+      "epoch": 18.465806451612902,
+      "grad_norm": 0.3371983468532562,
+      "learning_rate": 7.688850475367329e-05,
+      "loss": 0.2793,
+      "step": 3566
+    },
+    {
+      "epoch": 18.470967741935485,
+      "grad_norm": 0.24558991193771362,
+      "learning_rate": 7.685393258426966e-05,
+      "loss": 0.2784,
+      "step": 3567
+    },
+    {
+      "epoch": 18.476129032258065,
+      "grad_norm": 0.30339670181274414,
+      "learning_rate": 7.681936041486604e-05,
+      "loss": 0.3034,
+      "step": 3568
+    },
+    {
+      "epoch": 18.481290322580644,
+      "grad_norm": 0.29995065927505493,
+      "learning_rate": 7.678478824546241e-05,
+      "loss": 0.2788,
+      "step": 3569
+    },
+    {
+      "epoch": 18.486451612903227,
+      "grad_norm": 0.34749937057495117,
+      "learning_rate": 7.675021607605877e-05,
+      "loss": 0.2957,
+      "step": 3570
+    },
+    {
+      "epoch": 18.491612903225807,
+      "grad_norm": 0.25475266575813293,
+      "learning_rate": 7.671564390665514e-05,
+      "loss": 0.2719,
+      "step": 3571
+    },
+    {
+      "epoch": 18.496774193548386,
+      "grad_norm": 0.3273221552371979,
+      "learning_rate": 7.668107173725152e-05,
+      "loss": 0.294,
+      "step": 3572
+    },
+    {
+      "epoch": 18.501935483870966,
+      "grad_norm": 0.32448798418045044,
+      "learning_rate": 7.664649956784789e-05,
+      "loss": 0.2817,
+      "step": 3573
+    },
+    {
+      "epoch": 18.50709677419355,
+      "grad_norm": 0.32100623846054077,
+      "learning_rate": 7.661192739844426e-05,
+      "loss": 0.275,
+      "step": 3574
+    },
+    {
+      "epoch": 18.51225806451613,
+      "grad_norm": 0.29360905289649963,
+      "learning_rate": 7.657735522904062e-05,
+      "loss": 0.2747,
+      "step": 3575
+    },
+    {
+      "epoch": 18.517419354838708,
+      "grad_norm": 0.34669604897499084,
+      "learning_rate": 7.654278305963699e-05,
+      "loss": 0.3235,
+      "step": 3576
+    },
+    {
+      "epoch": 18.52258064516129,
+      "grad_norm": 0.307190865278244,
+      "learning_rate": 7.650821089023337e-05,
+      "loss": 0.2735,
+      "step": 3577
+    },
+    {
+      "epoch": 18.52774193548387,
+      "grad_norm": 0.3140644133090973,
+      "learning_rate": 7.647363872082974e-05,
+      "loss": 0.2749,
+      "step": 3578
+    },
+    {
+      "epoch": 18.53290322580645,
+      "grad_norm": 0.3863638937473297,
+      "learning_rate": 7.64390665514261e-05,
+      "loss": 0.2946,
+      "step": 3579
+    },
+    {
+      "epoch": 18.538064516129033,
+      "grad_norm": 0.1831798106431961,
+      "learning_rate": 7.640449438202247e-05,
+      "loss": 0.3005,
+      "step": 3580
+    },
+    {
+      "epoch": 18.543225806451613,
+      "grad_norm": 0.2897786498069763,
+      "learning_rate": 7.636992221261885e-05,
+      "loss": 0.313,
+      "step": 3581
+    },
+    {
+      "epoch": 18.548387096774192,
+      "grad_norm": 0.19305111467838287,
+      "learning_rate": 7.633535004321522e-05,
+      "loss": 0.2691,
+      "step": 3582
+    },
+    {
+      "epoch": 18.553548387096775,
+      "grad_norm": 0.2920196056365967,
+      "learning_rate": 7.630077787381159e-05,
+      "loss": 0.2665,
+      "step": 3583
+    },
+    {
+      "epoch": 18.558709677419355,
+      "grad_norm": 0.24730150401592255,
+      "learning_rate": 7.626620570440795e-05,
+      "loss": 0.2657,
+      "step": 3584
+    },
+    {
+      "epoch": 18.563870967741934,
+      "grad_norm": 0.24002736806869507,
+      "learning_rate": 7.623163353500432e-05,
+      "loss": 0.277,
+      "step": 3585
+    },
+    {
+      "epoch": 18.569032258064517,
+      "grad_norm": 0.31698647141456604,
+      "learning_rate": 7.61970613656007e-05,
+      "loss": 0.274,
+      "step": 3586
+    },
+    {
+      "epoch": 18.574193548387097,
+      "grad_norm": 0.24972690641880035,
+      "learning_rate": 7.616248919619707e-05,
+      "loss": 0.2734,
+      "step": 3587
+    },
+    {
+      "epoch": 18.579354838709676,
+      "grad_norm": 0.2324492335319519,
+      "learning_rate": 7.612791702679343e-05,
+      "loss": 0.2578,
+      "step": 3588
+    },
+    {
+      "epoch": 18.58451612903226,
+      "grad_norm": 0.3612532615661621,
+      "learning_rate": 7.60933448573898e-05,
+      "loss": 0.3125,
+      "step": 3589
+    },
+    {
+      "epoch": 18.58967741935484,
+      "grad_norm": 0.4099826216697693,
+      "learning_rate": 7.605877268798618e-05,
+      "loss": 0.2973,
+      "step": 3590
+    },
+    {
+      "epoch": 18.59483870967742,
+      "grad_norm": 0.2697674334049225,
+      "learning_rate": 7.602420051858255e-05,
+      "loss": 0.2957,
+      "step": 3591
+    },
+    {
+      "epoch": 18.6,
+      "grad_norm": 0.313518762588501,
+      "learning_rate": 7.598962834917892e-05,
+      "loss": 0.3045,
+      "step": 3592
+    },
+    {
+      "epoch": 18.60516129032258,
+      "grad_norm": 0.2621823847293854,
+      "learning_rate": 7.595505617977528e-05,
+      "loss": 0.287,
+      "step": 3593
+    },
+    {
+      "epoch": 18.61032258064516,
+      "grad_norm": 0.27598562836647034,
+      "learning_rate": 7.592048401037165e-05,
+      "loss": 0.2579,
+      "step": 3594
+    },
+    {
+      "epoch": 18.615483870967743,
+      "grad_norm": 0.3887515664100647,
+      "learning_rate": 7.588591184096803e-05,
+      "loss": 0.3205,
+      "step": 3595
+    },
+    {
+      "epoch": 18.620645161290323,
+      "grad_norm": 0.3112069070339203,
+      "learning_rate": 7.58513396715644e-05,
+      "loss": 0.2721,
+      "step": 3596
+    },
+    {
+      "epoch": 18.625806451612902,
+      "grad_norm": 0.294627845287323,
+      "learning_rate": 7.581676750216076e-05,
+      "loss": 0.2853,
+      "step": 3597
+    },
+    {
+      "epoch": 18.630967741935486,
+      "grad_norm": 0.2922278940677643,
+      "learning_rate": 7.578219533275713e-05,
+      "loss": 0.2799,
+      "step": 3598
+    },
+    {
+      "epoch": 18.636129032258065,
+      "grad_norm": 0.3173793852329254,
+      "learning_rate": 7.574762316335351e-05,
+      "loss": 0.2943,
+      "step": 3599
+    },
+    {
+      "epoch": 18.641290322580645,
+      "grad_norm": 0.2220081239938736,
+      "learning_rate": 7.571305099394988e-05,
+      "loss": 0.2903,
+      "step": 3600
+    },
+    {
+      "epoch": 18.646451612903228,
+      "grad_norm": 0.2732108533382416,
+      "learning_rate": 7.567847882454625e-05,
+      "loss": 0.3033,
+      "step": 3601
+    },
+    {
+      "epoch": 18.651612903225807,
+      "grad_norm": 0.3762911260128021,
+      "learning_rate": 7.564390665514261e-05,
+      "loss": 0.3003,
+      "step": 3602
+    },
+    {
+      "epoch": 18.656774193548387,
+      "grad_norm": 0.24202701449394226,
+      "learning_rate": 7.560933448573898e-05,
+      "loss": 0.2647,
+      "step": 3603
+    },
+    {
+      "epoch": 18.661935483870966,
+      "grad_norm": 0.37915104627609253,
+      "learning_rate": 7.557476231633536e-05,
+      "loss": 0.28,
+      "step": 3604
+    },
+    {
+      "epoch": 18.66709677419355,
+      "grad_norm": 0.30647552013397217,
+      "learning_rate": 7.554019014693173e-05,
+      "loss": 0.2839,
+      "step": 3605
+    },
+    {
+      "epoch": 18.67225806451613,
+      "grad_norm": 0.3358681797981262,
+      "learning_rate": 7.55056179775281e-05,
+      "loss": 0.2962,
+      "step": 3606
+    },
+    {
+      "epoch": 18.677419354838708,
+      "grad_norm": 0.3105098605155945,
+      "learning_rate": 7.547104580812446e-05,
+      "loss": 0.3197,
+      "step": 3607
+    },
+    {
+      "epoch": 18.68258064516129,
+      "grad_norm": 0.3009048402309418,
+      "learning_rate": 7.543647363872083e-05,
+      "loss": 0.2861,
+      "step": 3608
+    },
+    {
+      "epoch": 18.68774193548387,
+      "grad_norm": 0.3516865372657776,
+      "learning_rate": 7.540190146931721e-05,
+      "loss": 0.3006,
+      "step": 3609
+    },
+    {
+      "epoch": 18.69290322580645,
+      "grad_norm": 0.3165079653263092,
+      "learning_rate": 7.536732929991358e-05,
+      "loss": 0.2958,
+      "step": 3610
+    },
+    {
+      "epoch": 18.698064516129033,
+      "grad_norm": 0.358366996049881,
+      "learning_rate": 7.533275713050994e-05,
+      "loss": 0.3099,
+      "step": 3611
+    },
+    {
+      "epoch": 18.703225806451613,
+      "grad_norm": 0.32362380623817444,
+      "learning_rate": 7.529818496110631e-05,
+      "loss": 0.2967,
+      "step": 3612
+    },
+    {
+      "epoch": 18.708387096774192,
+      "grad_norm": 0.24510978162288666,
+      "learning_rate": 7.526361279170269e-05,
+      "loss": 0.2626,
+      "step": 3613
+    },
+    {
+      "epoch": 18.713548387096775,
+      "grad_norm": 0.286342978477478,
+      "learning_rate": 7.522904062229906e-05,
+      "loss": 0.297,
+      "step": 3614
+    },
+    {
+      "epoch": 18.718709677419355,
+      "grad_norm": 0.30959978699684143,
+      "learning_rate": 7.519446845289542e-05,
+      "loss": 0.2899,
+      "step": 3615
+    },
+    {
+      "epoch": 18.723870967741934,
+      "grad_norm": 0.40931206941604614,
+      "learning_rate": 7.515989628349179e-05,
+      "loss": 0.2746,
+      "step": 3616
+    },
+    {
+      "epoch": 18.729032258064517,
+      "grad_norm": 0.34634852409362793,
+      "learning_rate": 7.512532411408816e-05,
+      "loss": 0.2897,
+      "step": 3617
+    },
+    {
+      "epoch": 18.734193548387097,
+      "grad_norm": 0.33392423391342163,
+      "learning_rate": 7.509075194468454e-05,
+      "loss": 0.3003,
+      "step": 3618
+    },
+    {
+      "epoch": 18.739354838709676,
+      "grad_norm": 0.23849691450595856,
+      "learning_rate": 7.50561797752809e-05,
+      "loss": 0.284,
+      "step": 3619
+    },
+    {
+      "epoch": 18.74451612903226,
+      "grad_norm": 0.32021257281303406,
+      "learning_rate": 7.502160760587727e-05,
+      "loss": 0.2701,
+      "step": 3620
+    },
+    {
+      "epoch": 18.74967741935484,
+      "grad_norm": 0.42715972661972046,
+      "learning_rate": 7.498703543647364e-05,
+      "loss": 0.2593,
+      "step": 3621
+    },
+    {
+      "epoch": 18.75483870967742,
+      "grad_norm": 0.29065489768981934,
+      "learning_rate": 7.495246326707002e-05,
+      "loss": 0.3127,
+      "step": 3622
+    },
+    {
+      "epoch": 18.76,
+      "grad_norm": 0.33101195096969604,
+      "learning_rate": 7.491789109766639e-05,
+      "loss": 0.3018,
+      "step": 3623
+    },
+    {
+      "epoch": 18.76516129032258,
+      "grad_norm": 0.2005617320537567,
+      "learning_rate": 7.488331892826274e-05,
+      "loss": 0.269,
+      "step": 3624
+    },
+    {
+      "epoch": 18.77032258064516,
+      "grad_norm": 0.38254567980766296,
+      "learning_rate": 7.484874675885912e-05,
+      "loss": 0.2904,
+      "step": 3625
+    },
+    {
+      "epoch": 18.775483870967744,
+      "grad_norm": 0.3352719843387604,
+      "learning_rate": 7.481417458945549e-05,
+      "loss": 0.2794,
+      "step": 3626
+    },
+    {
+      "epoch": 18.780645161290323,
+      "grad_norm": 0.2747732400894165,
+      "learning_rate": 7.477960242005187e-05,
+      "loss": 0.2923,
+      "step": 3627
+    },
+    {
+      "epoch": 18.785806451612903,
+      "grad_norm": 0.29041650891304016,
+      "learning_rate": 7.474503025064824e-05,
+      "loss": 0.2812,
+      "step": 3628
+    },
+    {
+      "epoch": 18.790967741935482,
+      "grad_norm": 0.3502669036388397,
+      "learning_rate": 7.47104580812446e-05,
+      "loss": 0.2572,
+      "step": 3629
+    },
+    {
+      "epoch": 18.796129032258065,
+      "grad_norm": 0.3039942681789398,
+      "learning_rate": 7.467588591184097e-05,
+      "loss": 0.2989,
+      "step": 3630
+    },
+    {
+      "epoch": 18.801290322580645,
+      "grad_norm": 0.26708725094795227,
+      "learning_rate": 7.464131374243735e-05,
+      "loss": 0.2958,
+      "step": 3631
+    },
+    {
+      "epoch": 18.806451612903224,
+      "grad_norm": 0.2600004971027374,
+      "learning_rate": 7.460674157303372e-05,
+      "loss": 0.2867,
+      "step": 3632
+    },
+    {
+      "epoch": 18.811612903225807,
+      "grad_norm": 0.28005778789520264,
+      "learning_rate": 7.457216940363007e-05,
+      "loss": 0.282,
+      "step": 3633
+    },
+    {
+      "epoch": 18.816774193548387,
+      "grad_norm": 0.2494955062866211,
+      "learning_rate": 7.453759723422645e-05,
+      "loss": 0.2561,
+      "step": 3634
+    },
+    {
+      "epoch": 18.821935483870966,
+      "grad_norm": 0.3290054500102997,
+      "learning_rate": 7.450302506482282e-05,
+      "loss": 0.2882,
+      "step": 3635
+    },
+    {
+      "epoch": 18.82709677419355,
+      "grad_norm": 0.3516405522823334,
+      "learning_rate": 7.44684528954192e-05,
+      "loss": 0.3016,
+      "step": 3636
+    },
+    {
+      "epoch": 18.83225806451613,
+      "grad_norm": 0.33000609278678894,
+      "learning_rate": 7.443388072601557e-05,
+      "loss": 0.2888,
+      "step": 3637
+    },
+    {
+      "epoch": 18.83741935483871,
+      "grad_norm": 0.33957284688949585,
+      "learning_rate": 7.439930855661193e-05,
+      "loss": 0.2731,
+      "step": 3638
+    },
+    {
+      "epoch": 18.84258064516129,
+      "grad_norm": 0.4123179018497467,
+      "learning_rate": 7.43647363872083e-05,
+      "loss": 0.3003,
+      "step": 3639
+    },
+    {
+      "epoch": 18.84774193548387,
+      "grad_norm": 0.2151627093553543,
+      "learning_rate": 7.433016421780467e-05,
+      "loss": 0.2401,
+      "step": 3640
+    },
+    {
+      "epoch": 18.85290322580645,
+      "grad_norm": 0.3162158131599426,
+      "learning_rate": 7.429559204840105e-05,
+      "loss": 0.2777,
+      "step": 3641
+    },
+    {
+      "epoch": 18.858064516129033,
+      "grad_norm": 0.23777632415294647,
+      "learning_rate": 7.42610198789974e-05,
+      "loss": 0.2867,
+      "step": 3642
+    },
+    {
+      "epoch": 18.863225806451613,
+      "grad_norm": 0.3373486399650574,
+      "learning_rate": 7.422644770959378e-05,
+      "loss": 0.2833,
+      "step": 3643
+    },
+    {
+      "epoch": 18.868387096774192,
+      "grad_norm": 0.337197482585907,
+      "learning_rate": 7.419187554019015e-05,
+      "loss": 0.2971,
+      "step": 3644
+    },
+    {
+      "epoch": 18.873548387096776,
+      "grad_norm": 0.41969072818756104,
+      "learning_rate": 7.415730337078653e-05,
+      "loss": 0.31,
+      "step": 3645
+    },
+    {
+      "epoch": 18.878709677419355,
+      "grad_norm": 0.323295921087265,
+      "learning_rate": 7.41227312013829e-05,
+      "loss": 0.3016,
+      "step": 3646
+    },
+    {
+      "epoch": 18.883870967741935,
+      "grad_norm": 0.29875582456588745,
+      "learning_rate": 7.408815903197926e-05,
+      "loss": 0.2778,
+      "step": 3647
+    },
+    {
+      "epoch": 18.889032258064518,
+      "grad_norm": 0.29150545597076416,
+      "learning_rate": 7.405358686257563e-05,
+      "loss": 0.2827,
+      "step": 3648
+    },
+    {
+      "epoch": 18.894193548387097,
+      "grad_norm": 0.30213436484336853,
+      "learning_rate": 7.4019014693172e-05,
+      "loss": 0.289,
+      "step": 3649
+    },
+    {
+      "epoch": 18.899354838709677,
+      "grad_norm": 0.37179046869277954,
+      "learning_rate": 7.398444252376838e-05,
+      "loss": 0.2922,
+      "step": 3650
+    },
+    {
+      "epoch": 18.90451612903226,
+      "grad_norm": 0.35115256905555725,
+      "learning_rate": 7.394987035436473e-05,
+      "loss": 0.3039,
+      "step": 3651
+    },
+    {
+      "epoch": 18.90967741935484,
+      "grad_norm": 0.3426741659641266,
+      "learning_rate": 7.391529818496111e-05,
+      "loss": 0.2848,
+      "step": 3652
+    },
+    {
+      "epoch": 18.91483870967742,
+      "grad_norm": 0.29893290996551514,
+      "learning_rate": 7.388072601555748e-05,
+      "loss": 0.2946,
+      "step": 3653
+    },
+    {
+      "epoch": 18.92,
+      "grad_norm": 0.418101966381073,
+      "learning_rate": 7.384615384615386e-05,
+      "loss": 0.3033,
+      "step": 3654
+    },
+    {
+      "epoch": 18.92516129032258,
+      "grad_norm": 0.2960748076438904,
+      "learning_rate": 7.381158167675023e-05,
+      "loss": 0.2955,
+      "step": 3655
+    },
+    {
+      "epoch": 18.93032258064516,
+      "grad_norm": 0.3259407877922058,
+      "learning_rate": 7.377700950734659e-05,
+      "loss": 0.3087,
+      "step": 3656
+    },
+    {
+      "epoch": 18.93548387096774,
+      "grad_norm": 0.33105579018592834,
+      "learning_rate": 7.374243733794296e-05,
+      "loss": 0.2927,
+      "step": 3657
+    },
+    {
+      "epoch": 18.940645161290323,
+      "grad_norm": 0.3670693039894104,
+      "learning_rate": 7.370786516853933e-05,
+      "loss": 0.3,
+      "step": 3658
+    },
+    {
+      "epoch": 18.945806451612903,
+      "grad_norm": 0.2657787799835205,
+      "learning_rate": 7.367329299913571e-05,
+      "loss": 0.2925,
+      "step": 3659
+    },
+    {
+      "epoch": 18.950967741935482,
+      "grad_norm": 0.335760235786438,
+      "learning_rate": 7.363872082973206e-05,
+      "loss": 0.2941,
+      "step": 3660
+    },
+    {
+      "epoch": 18.956129032258065,
+      "grad_norm": 0.35448819398880005,
+      "learning_rate": 7.360414866032844e-05,
+      "loss": 0.2888,
+      "step": 3661
+    },
+    {
+      "epoch": 18.961290322580645,
+      "grad_norm": 0.3398735821247101,
+      "learning_rate": 7.356957649092481e-05,
+      "loss": 0.2818,
+      "step": 3662
+    },
+    {
+      "epoch": 18.966451612903224,
+      "grad_norm": 0.2339407354593277,
+      "learning_rate": 7.353500432152119e-05,
+      "loss": 0.2765,
+      "step": 3663
+    },
+    {
+      "epoch": 18.971612903225807,
+      "grad_norm": 0.38045573234558105,
+      "learning_rate": 7.350043215211756e-05,
+      "loss": 0.2895,
+      "step": 3664
+    },
+    {
+      "epoch": 18.976774193548387,
+      "grad_norm": 0.3213560879230499,
+      "learning_rate": 7.346585998271391e-05,
+      "loss": 0.2679,
+      "step": 3665
+    },
+    {
+      "epoch": 18.981935483870966,
+      "grad_norm": 0.29758813977241516,
+      "learning_rate": 7.343128781331029e-05,
+      "loss": 0.2982,
+      "step": 3666
+    },
+    {
+      "epoch": 18.98709677419355,
+      "grad_norm": 0.29814115166664124,
+      "learning_rate": 7.339671564390666e-05,
+      "loss": 0.2933,
+      "step": 3667
+    },
+    {
+      "epoch": 18.99225806451613,
+      "grad_norm": 0.33674338459968567,
+      "learning_rate": 7.336214347450304e-05,
+      "loss": 0.284,
+      "step": 3668
+    },
+    {
+      "epoch": 18.99741935483871,
+      "grad_norm": 0.4020998179912567,
+      "learning_rate": 7.332757130509939e-05,
+      "loss": 0.3196,
+      "step": 3669
+    },
+    {
+      "epoch": 19.003870967741936,
+      "grad_norm": 0.34652256965637207,
+      "learning_rate": 7.329299913569577e-05,
+      "loss": 0.3508,
+      "step": 3670
+    },
+    {
+      "epoch": 19.009032258064515,
+      "grad_norm": 0.10217051953077316,
+      "learning_rate": 7.325842696629214e-05,
+      "loss": 0.2742,
+      "step": 3671
+    },
+    {
+      "epoch": 19.014193548387098,
+      "grad_norm": 0.12366849184036255,
+      "learning_rate": 7.322385479688852e-05,
+      "loss": 0.2939,
+      "step": 3672
+    },
+    {
+      "epoch": 19.019354838709678,
+      "grad_norm": 0.11130823194980621,
+      "learning_rate": 7.318928262748489e-05,
+      "loss": 0.2804,
+      "step": 3673
+    },
+    {
+      "epoch": 19.024516129032257,
+      "grad_norm": 0.20474757254123688,
+      "learning_rate": 7.315471045808124e-05,
+      "loss": 0.3039,
+      "step": 3674
+    },
+    {
+      "epoch": 19.02967741935484,
+      "grad_norm": 0.21021120250225067,
+      "learning_rate": 7.312013828867762e-05,
+      "loss": 0.2753,
+      "step": 3675
+    },
+    {
+      "epoch": 19.03483870967742,
+      "grad_norm": 0.24919083714485168,
+      "learning_rate": 7.308556611927399e-05,
+      "loss": 0.289,
+      "step": 3676
+    },
+    {
+      "epoch": 19.04,
+      "grad_norm": 0.14956878125667572,
+      "learning_rate": 7.305099394987037e-05,
+      "loss": 0.2564,
+      "step": 3677
+    },
+    {
+      "epoch": 19.045161290322582,
+      "grad_norm": 0.1347154825925827,
+      "learning_rate": 7.301642178046672e-05,
+      "loss": 0.2828,
+      "step": 3678
+    },
+    {
+      "epoch": 19.05032258064516,
+      "grad_norm": 0.2156568467617035,
+      "learning_rate": 7.29818496110631e-05,
+      "loss": 0.2977,
+      "step": 3679
+    },
+    {
+      "epoch": 19.05548387096774,
+      "grad_norm": 0.16839821636676788,
+      "learning_rate": 7.294727744165947e-05,
+      "loss": 0.2863,
+      "step": 3680
+    },
+    {
+      "epoch": 19.060645161290324,
+      "grad_norm": 0.18701505661010742,
+      "learning_rate": 7.291270527225583e-05,
+      "loss": 0.272,
+      "step": 3681
+    },
+    {
+      "epoch": 19.065806451612904,
+      "grad_norm": 0.214008629322052,
+      "learning_rate": 7.28781331028522e-05,
+      "loss": 0.29,
+      "step": 3682
+    },
+    {
+      "epoch": 19.070967741935483,
+      "grad_norm": 0.24483013153076172,
+      "learning_rate": 7.284356093344857e-05,
+      "loss": 0.3019,
+      "step": 3683
+    },
+    {
+      "epoch": 19.076129032258063,
+      "grad_norm": 0.15913604199886322,
+      "learning_rate": 7.280898876404495e-05,
+      "loss": 0.3022,
+      "step": 3684
+    },
+    {
+      "epoch": 19.081290322580646,
+      "grad_norm": 0.17925745248794556,
+      "learning_rate": 7.277441659464132e-05,
+      "loss": 0.2453,
+      "step": 3685
+    },
+    {
+      "epoch": 19.086451612903225,
+      "grad_norm": 0.2588502764701843,
+      "learning_rate": 7.27398444252377e-05,
+      "loss": 0.2857,
+      "step": 3686
+    },
+    {
+      "epoch": 19.091612903225805,
+      "grad_norm": 0.22445423901081085,
+      "learning_rate": 7.270527225583405e-05,
+      "loss": 0.297,
+      "step": 3687
+    },
+    {
+      "epoch": 19.096774193548388,
+      "grad_norm": 0.16800495982170105,
+      "learning_rate": 7.267070008643043e-05,
+      "loss": 0.2793,
+      "step": 3688
+    },
+    {
+      "epoch": 19.101935483870967,
+      "grad_norm": 0.16572265326976776,
+      "learning_rate": 7.26361279170268e-05,
+      "loss": 0.2576,
+      "step": 3689
+    },
+    {
+      "epoch": 19.107096774193547,
+      "grad_norm": 0.2097540944814682,
+      "learning_rate": 7.260155574762316e-05,
+      "loss": 0.2906,
+      "step": 3690
+    },
+    {
+      "epoch": 19.11225806451613,
+      "grad_norm": 0.15796098113059998,
+      "learning_rate": 7.256698357821953e-05,
+      "loss": 0.2872,
+      "step": 3691
+    },
+    {
+      "epoch": 19.11741935483871,
+      "grad_norm": 0.2291281670331955,
+      "learning_rate": 7.25324114088159e-05,
+      "loss": 0.2623,
+      "step": 3692
+    },
+    {
+      "epoch": 19.12258064516129,
+      "grad_norm": 0.23413367569446564,
+      "learning_rate": 7.249783923941228e-05,
+      "loss": 0.3193,
+      "step": 3693
+    },
+    {
+      "epoch": 19.127741935483872,
+      "grad_norm": 0.17914089560508728,
+      "learning_rate": 7.246326707000865e-05,
+      "loss": 0.2959,
+      "step": 3694
+    },
+    {
+      "epoch": 19.13290322580645,
+      "grad_norm": 0.16919897496700287,
+      "learning_rate": 7.242869490060503e-05,
+      "loss": 0.2755,
+      "step": 3695
+    },
+    {
+      "epoch": 19.13806451612903,
+      "grad_norm": 0.25541794300079346,
+      "learning_rate": 7.239412273120138e-05,
+      "loss": 0.2762,
+      "step": 3696
+    },
+    {
+      "epoch": 19.143225806451614,
+      "grad_norm": 0.17130370438098907,
+      "learning_rate": 7.235955056179775e-05,
+      "loss": 0.2679,
+      "step": 3697
+    },
+    {
+      "epoch": 19.148387096774194,
+      "grad_norm": 0.1854475885629654,
+      "learning_rate": 7.232497839239413e-05,
+      "loss": 0.3164,
+      "step": 3698
+    },
+    {
+      "epoch": 19.153548387096773,
+      "grad_norm": 0.1927035003900528,
+      "learning_rate": 7.22904062229905e-05,
+      "loss": 0.2931,
+      "step": 3699
+    },
+    {
+      "epoch": 19.158709677419356,
+      "grad_norm": 0.17612135410308838,
+      "learning_rate": 7.225583405358686e-05,
+      "loss": 0.2474,
+      "step": 3700
+    },
+    {
+      "epoch": 19.163870967741936,
+      "grad_norm": 0.21897238492965698,
+      "learning_rate": 7.222126188418323e-05,
+      "loss": 0.2805,
+      "step": 3701
+    },
+    {
+      "epoch": 19.169032258064515,
+      "grad_norm": 0.26504233479499817,
+      "learning_rate": 7.218668971477961e-05,
+      "loss": 0.311,
+      "step": 3702
+    },
+    {
+      "epoch": 19.174193548387098,
+      "grad_norm": 0.23090378940105438,
+      "learning_rate": 7.215211754537598e-05,
+      "loss": 0.2706,
+      "step": 3703
+    },
+    {
+      "epoch": 19.179354838709678,
+      "grad_norm": 0.19023926556110382,
+      "learning_rate": 7.211754537597236e-05,
+      "loss": 0.3034,
+      "step": 3704
+    },
+    {
+      "epoch": 19.184516129032257,
+      "grad_norm": 0.1844470202922821,
+      "learning_rate": 7.208297320656871e-05,
+      "loss": 0.2791,
+      "step": 3705
+    },
+    {
+      "epoch": 19.18967741935484,
+      "grad_norm": 0.1516748070716858,
+      "learning_rate": 7.204840103716508e-05,
+      "loss": 0.2808,
+      "step": 3706
+    },
+    {
+      "epoch": 19.19483870967742,
+      "grad_norm": 0.15470083057880402,
+      "learning_rate": 7.201382886776146e-05,
+      "loss": 0.3008,
+      "step": 3707
+    },
+    {
+      "epoch": 19.2,
+      "grad_norm": 0.14176692068576813,
+      "learning_rate": 7.197925669835782e-05,
+      "loss": 0.2866,
+      "step": 3708
+    },
+    {
+      "epoch": 19.205161290322582,
+      "grad_norm": 0.19849860668182373,
+      "learning_rate": 7.194468452895419e-05,
+      "loss": 0.2725,
+      "step": 3709
+    },
+    {
+      "epoch": 19.210322580645162,
+      "grad_norm": 0.5694109797477722,
+      "learning_rate": 7.191011235955056e-05,
+      "loss": 0.2729,
+      "step": 3710
+    },
+    {
+      "epoch": 19.21548387096774,
+      "grad_norm": 0.23671087622642517,
+      "learning_rate": 7.187554019014694e-05,
+      "loss": 0.2936,
+      "step": 3711
+    },
+    {
+      "epoch": 19.22064516129032,
+      "grad_norm": 0.17478233575820923,
+      "learning_rate": 7.18409680207433e-05,
+      "loss": 0.2683,
+      "step": 3712
+    },
+    {
+      "epoch": 19.225806451612904,
+      "grad_norm": 0.1353680044412613,
+      "learning_rate": 7.180639585133967e-05,
+      "loss": 0.2405,
+      "step": 3713
+    },
+    {
+      "epoch": 19.230967741935483,
+      "grad_norm": 0.1738402396440506,
+      "learning_rate": 7.177182368193604e-05,
+      "loss": 0.2937,
+      "step": 3714
+    },
+    {
+      "epoch": 19.236129032258063,
+      "grad_norm": 0.25448426604270935,
+      "learning_rate": 7.173725151253241e-05,
+      "loss": 0.2863,
+      "step": 3715
+    },
+    {
+      "epoch": 19.241290322580646,
+      "grad_norm": 0.16057902574539185,
+      "learning_rate": 7.170267934312879e-05,
+      "loss": 0.2838,
+      "step": 3716
+    },
+    {
+      "epoch": 19.246451612903225,
+      "grad_norm": 0.14119496941566467,
+      "learning_rate": 7.166810717372515e-05,
+      "loss": 0.2782,
+      "step": 3717
+    },
+    {
+      "epoch": 19.251612903225805,
+      "grad_norm": 0.17816932499408722,
+      "learning_rate": 7.163353500432152e-05,
+      "loss": 0.298,
+      "step": 3718
+    },
+    {
+      "epoch": 19.256774193548388,
+      "grad_norm": 0.1935812085866928,
+      "learning_rate": 7.159896283491789e-05,
+      "loss": 0.2997,
+      "step": 3719
+    },
+    {
+      "epoch": 19.261935483870968,
+      "grad_norm": 0.18799607455730438,
+      "learning_rate": 7.156439066551427e-05,
+      "loss": 0.2731,
+      "step": 3720
+    },
+    {
+      "epoch": 19.267096774193547,
+      "grad_norm": 0.183616042137146,
+      "learning_rate": 7.152981849611064e-05,
+      "loss": 0.255,
+      "step": 3721
+    },
+    {
+      "epoch": 19.27225806451613,
+      "grad_norm": 0.19768115878105164,
+      "learning_rate": 7.1495246326707e-05,
+      "loss": 0.299,
+      "step": 3722
+    },
+    {
+      "epoch": 19.27741935483871,
+      "grad_norm": 0.20255891978740692,
+      "learning_rate": 7.146067415730337e-05,
+      "loss": 0.2977,
+      "step": 3723
+    },
+    {
+      "epoch": 19.28258064516129,
+      "grad_norm": 0.1448439210653305,
+      "learning_rate": 7.142610198789974e-05,
+      "loss": 0.3007,
+      "step": 3724
+    },
+    {
+      "epoch": 19.287741935483872,
+      "grad_norm": 0.2151428461074829,
+      "learning_rate": 7.139152981849612e-05,
+      "loss": 0.2709,
+      "step": 3725
+    },
+    {
+      "epoch": 19.29290322580645,
+      "grad_norm": 0.2732800543308258,
+      "learning_rate": 7.135695764909248e-05,
+      "loss": 0.2841,
+      "step": 3726
+    },
+    {
+      "epoch": 19.29806451612903,
+      "grad_norm": 0.2554592490196228,
+      "learning_rate": 7.132238547968885e-05,
+      "loss": 0.2847,
+      "step": 3727
+    },
+    {
+      "epoch": 19.303225806451614,
+      "grad_norm": 0.1405654400587082,
+      "learning_rate": 7.128781331028522e-05,
+      "loss": 0.2913,
+      "step": 3728
+    },
+    {
+      "epoch": 19.308387096774194,
+      "grad_norm": 0.17488086223602295,
+      "learning_rate": 7.12532411408816e-05,
+      "loss": 0.2702,
+      "step": 3729
+    },
+    {
+      "epoch": 19.313548387096773,
+      "grad_norm": 0.2693886160850525,
+      "learning_rate": 7.121866897147797e-05,
+      "loss": 0.2827,
+      "step": 3730
+    },
+    {
+      "epoch": 19.318709677419356,
+      "grad_norm": 0.16780829429626465,
+      "learning_rate": 7.118409680207433e-05,
+      "loss": 0.2876,
+      "step": 3731
+    },
+    {
+      "epoch": 19.323870967741936,
+      "grad_norm": 0.15247713029384613,
+      "learning_rate": 7.11495246326707e-05,
+      "loss": 0.295,
+      "step": 3732
+    },
+    {
+      "epoch": 19.329032258064515,
+      "grad_norm": 0.16700220108032227,
+      "learning_rate": 7.111495246326707e-05,
+      "loss": 0.2631,
+      "step": 3733
+    },
+    {
+      "epoch": 19.3341935483871,
+      "grad_norm": 0.18959513306617737,
+      "learning_rate": 7.108038029386345e-05,
+      "loss": 0.2497,
+      "step": 3734
+    },
+    {
+      "epoch": 19.339354838709678,
+      "grad_norm": 0.2889779508113861,
+      "learning_rate": 7.104580812445981e-05,
+      "loss": 0.3074,
+      "step": 3735
+    },
+    {
+      "epoch": 19.344516129032257,
+      "grad_norm": 0.15483912825584412,
+      "learning_rate": 7.101123595505618e-05,
+      "loss": 0.3011,
+      "step": 3736
+    },
+    {
+      "epoch": 19.34967741935484,
+      "grad_norm": 0.07436084747314453,
+      "learning_rate": 7.097666378565255e-05,
+      "loss": 0.2876,
+      "step": 3737
+    },
+    {
+      "epoch": 19.35483870967742,
+      "grad_norm": 0.1538894921541214,
+      "learning_rate": 7.094209161624892e-05,
+      "loss": 0.2782,
+      "step": 3738
+    },
+    {
+      "epoch": 19.36,
+      "grad_norm": 0.20621225237846375,
+      "learning_rate": 7.09075194468453e-05,
+      "loss": 0.3077,
+      "step": 3739
+    },
+    {
+      "epoch": 19.36516129032258,
+      "grad_norm": 0.23089993000030518,
+      "learning_rate": 7.087294727744166e-05,
+      "loss": 0.3149,
+      "step": 3740
+    },
+    {
+      "epoch": 19.370322580645162,
+      "grad_norm": 0.15889911353588104,
+      "learning_rate": 7.083837510803803e-05,
+      "loss": 0.3018,
+      "step": 3741
+    },
+    {
+      "epoch": 19.37548387096774,
+      "grad_norm": 0.20269301533699036,
+      "learning_rate": 7.08038029386344e-05,
+      "loss": 0.2752,
+      "step": 3742
+    },
+    {
+      "epoch": 19.38064516129032,
+      "grad_norm": 0.10852048546075821,
+      "learning_rate": 7.076923076923078e-05,
+      "loss": 0.2733,
+      "step": 3743
+    },
+    {
+      "epoch": 19.385806451612904,
+      "grad_norm": 0.13975094258785248,
+      "learning_rate": 7.073465859982714e-05,
+      "loss": 0.3013,
+      "step": 3744
+    },
+    {
+      "epoch": 19.390967741935484,
+      "grad_norm": 0.2084692418575287,
+      "learning_rate": 7.070008643042351e-05,
+      "loss": 0.3037,
+      "step": 3745
+    },
+    {
+      "epoch": 19.396129032258063,
+      "grad_norm": 0.20040878653526306,
+      "learning_rate": 7.066551426101988e-05,
+      "loss": 0.2984,
+      "step": 3746
+    },
+    {
+      "epoch": 19.401290322580646,
+      "grad_norm": 0.21622706949710846,
+      "learning_rate": 7.063094209161625e-05,
+      "loss": 0.2681,
+      "step": 3747
+    },
+    {
+      "epoch": 19.406451612903226,
+      "grad_norm": 0.24168577790260315,
+      "learning_rate": 7.059636992221263e-05,
+      "loss": 0.3001,
+      "step": 3748
+    },
+    {
+      "epoch": 19.411612903225805,
+      "grad_norm": 0.21256987750530243,
+      "learning_rate": 7.056179775280899e-05,
+      "loss": 0.2843,
+      "step": 3749
+    },
+    {
+      "epoch": 19.416774193548388,
+      "grad_norm": 0.24000707268714905,
+      "learning_rate": 7.052722558340536e-05,
+      "loss": 0.3067,
+      "step": 3750
+    },
+    {
+      "epoch": 19.421935483870968,
+      "grad_norm": 0.2317504584789276,
+      "learning_rate": 7.049265341400173e-05,
+      "loss": 0.2707,
+      "step": 3751
+    },
+    {
+      "epoch": 19.427096774193547,
+      "grad_norm": 0.19242578744888306,
+      "learning_rate": 7.045808124459811e-05,
+      "loss": 0.2939,
+      "step": 3752
+    },
+    {
+      "epoch": 19.43225806451613,
+      "grad_norm": 0.2885685861110687,
+      "learning_rate": 7.042350907519447e-05,
+      "loss": 0.2879,
+      "step": 3753
+    },
+    {
+      "epoch": 19.43741935483871,
+      "grad_norm": 0.22074484825134277,
+      "learning_rate": 7.038893690579084e-05,
+      "loss": 0.2764,
+      "step": 3754
+    },
+    {
+      "epoch": 19.44258064516129,
+      "grad_norm": 0.29159894585609436,
+      "learning_rate": 7.035436473638721e-05,
+      "loss": 0.2906,
+      "step": 3755
+    },
+    {
+      "epoch": 19.447741935483872,
+      "grad_norm": 0.20829400420188904,
+      "learning_rate": 7.031979256698358e-05,
+      "loss": 0.2719,
+      "step": 3756
+    },
+    {
+      "epoch": 19.452903225806452,
+      "grad_norm": 0.20344607532024384,
+      "learning_rate": 7.028522039757996e-05,
+      "loss": 0.2801,
+      "step": 3757
+    },
+    {
+      "epoch": 19.45806451612903,
+      "grad_norm": 0.22039544582366943,
+      "learning_rate": 7.025064822817632e-05,
+      "loss": 0.2825,
+      "step": 3758
+    },
+    {
+      "epoch": 19.463225806451614,
+      "grad_norm": 0.17804743349552155,
+      "learning_rate": 7.021607605877269e-05,
+      "loss": 0.2726,
+      "step": 3759
+    },
+    {
+      "epoch": 19.468387096774194,
+      "grad_norm": 0.2350357174873352,
+      "learning_rate": 7.018150388936906e-05,
+      "loss": 0.2792,
+      "step": 3760
+    },
+    {
+      "epoch": 19.473548387096773,
+      "grad_norm": 0.1683480441570282,
+      "learning_rate": 7.014693171996544e-05,
+      "loss": 0.2807,
+      "step": 3761
+    },
+    {
+      "epoch": 19.478709677419356,
+      "grad_norm": 0.2569376230239868,
+      "learning_rate": 7.01123595505618e-05,
+      "loss": 0.2948,
+      "step": 3762
+    },
+    {
+      "epoch": 19.483870967741936,
+      "grad_norm": 0.17428235709667206,
+      "learning_rate": 7.007778738115817e-05,
+      "loss": 0.2979,
+      "step": 3763
+    },
+    {
+      "epoch": 19.489032258064515,
+      "grad_norm": 0.18843907117843628,
+      "learning_rate": 7.004321521175454e-05,
+      "loss": 0.2971,
+      "step": 3764
+    },
+    {
+      "epoch": 19.4941935483871,
+      "grad_norm": 0.21784169971942902,
+      "learning_rate": 7.00086430423509e-05,
+      "loss": 0.278,
+      "step": 3765
+    },
+    {
+      "epoch": 19.499354838709678,
+      "grad_norm": 0.1885073184967041,
+      "learning_rate": 6.997407087294729e-05,
+      "loss": 0.2514,
+      "step": 3766
+    },
+    {
+      "epoch": 19.504516129032258,
+      "grad_norm": 0.19468657672405243,
+      "learning_rate": 6.993949870354365e-05,
+      "loss": 0.2867,
+      "step": 3767
+    },
+    {
+      "epoch": 19.509677419354837,
+      "grad_norm": 0.1394231915473938,
+      "learning_rate": 6.990492653414002e-05,
+      "loss": 0.2689,
+      "step": 3768
+    },
+    {
+      "epoch": 19.51483870967742,
+      "grad_norm": 0.168666273355484,
+      "learning_rate": 6.987035436473639e-05,
+      "loss": 0.2934,
+      "step": 3769
+    },
+    {
+      "epoch": 19.52,
+      "grad_norm": 0.2498876005411148,
+      "learning_rate": 6.983578219533275e-05,
+      "loss": 0.2735,
+      "step": 3770
+    },
+    {
+      "epoch": 19.52,
+      "eval_loss": 2.961930751800537,
+      "eval_runtime": 21.1968,
+      "eval_samples_per_second": 3.727,
+      "eval_steps_per_second": 0.472,
+      "step": 3770
+    },
+    {
+      "epoch": 19.52516129032258,
+      "grad_norm": 0.12838207185268402,
+      "learning_rate": 6.980121002592913e-05,
+      "loss": 0.2535,
+      "step": 3771
+    },
+    {
+      "epoch": 19.530322580645162,
+      "grad_norm": 0.16358302533626556,
+      "learning_rate": 6.97666378565255e-05,
+      "loss": 0.29,
+      "step": 3772
+    },
+    {
+      "epoch": 19.53548387096774,
+      "grad_norm": 0.21424636244773865,
+      "learning_rate": 6.973206568712187e-05,
+      "loss": 0.2955,
+      "step": 3773
+    },
+    {
+      "epoch": 19.54064516129032,
+      "grad_norm": 0.13050352036952972,
+      "learning_rate": 6.969749351771824e-05,
+      "loss": 0.2595,
+      "step": 3774
+    },
+    {
+      "epoch": 19.545806451612904,
+      "grad_norm": 0.18172352015972137,
+      "learning_rate": 6.966292134831462e-05,
+      "loss": 0.2809,
+      "step": 3775
+    },
+    {
+      "epoch": 19.550967741935484,
+      "grad_norm": 0.1072440966963768,
+      "learning_rate": 6.962834917891098e-05,
+      "loss": 0.2961,
+      "step": 3776
+    },
+    {
+      "epoch": 19.556129032258063,
+      "grad_norm": 0.1729818731546402,
+      "learning_rate": 6.959377700950735e-05,
+      "loss": 0.2683,
+      "step": 3777
+    },
+    {
+      "epoch": 19.561290322580646,
+      "grad_norm": 0.38721874356269836,
+      "learning_rate": 6.955920484010372e-05,
+      "loss": 0.2749,
+      "step": 3778
+    },
+    {
+      "epoch": 19.566451612903226,
+      "grad_norm": 0.14268365502357483,
+      "learning_rate": 6.952463267070008e-05,
+      "loss": 0.2771,
+      "step": 3779
+    },
+    {
+      "epoch": 19.571612903225805,
+      "grad_norm": 0.1432485431432724,
+      "learning_rate": 6.949006050129646e-05,
+      "loss": 0.2896,
+      "step": 3780
+    },
+    {
+      "epoch": 19.57677419354839,
+      "grad_norm": 0.2510281801223755,
+      "learning_rate": 6.945548833189283e-05,
+      "loss": 0.2725,
+      "step": 3781
+    },
+    {
+      "epoch": 19.581935483870968,
+      "grad_norm": 0.31917905807495117,
+      "learning_rate": 6.94209161624892e-05,
+      "loss": 0.3112,
+      "step": 3782
+    },
+    {
+      "epoch": 19.587096774193547,
+      "grad_norm": 0.1656688004732132,
+      "learning_rate": 6.938634399308557e-05,
+      "loss": 0.289,
+      "step": 3783
+    },
+    {
+      "epoch": 19.59225806451613,
+      "grad_norm": 0.22717596590518951,
+      "learning_rate": 6.935177182368195e-05,
+      "loss": 0.287,
+      "step": 3784
+    },
+    {
+      "epoch": 19.59741935483871,
+      "grad_norm": 0.1314416080713272,
+      "learning_rate": 6.931719965427831e-05,
+      "loss": 0.2282,
+      "step": 3785
+    },
+    {
+      "epoch": 19.60258064516129,
+      "grad_norm": 0.18766683340072632,
+      "learning_rate": 6.928262748487468e-05,
+      "loss": 0.2857,
+      "step": 3786
+    },
+    {
+      "epoch": 19.607741935483872,
+      "grad_norm": 0.2467842400074005,
+      "learning_rate": 6.924805531547105e-05,
+      "loss": 0.293,
+      "step": 3787
+    },
+    {
+      "epoch": 19.612903225806452,
+      "grad_norm": 0.1715332716703415,
+      "learning_rate": 6.921348314606741e-05,
+      "loss": 0.2676,
+      "step": 3788
+    },
+    {
+      "epoch": 19.61806451612903,
+      "grad_norm": 0.2354295700788498,
+      "learning_rate": 6.91789109766638e-05,
+      "loss": 0.2954,
+      "step": 3789
+    },
+    {
+      "epoch": 19.623225806451615,
+      "grad_norm": 0.18113720417022705,
+      "learning_rate": 6.914433880726016e-05,
+      "loss": 0.299,
+      "step": 3790
+    },
+    {
+      "epoch": 19.628387096774194,
+      "grad_norm": 0.2612518072128296,
+      "learning_rate": 6.910976663785653e-05,
+      "loss": 0.2767,
+      "step": 3791
+    },
+    {
+      "epoch": 19.633548387096774,
+      "grad_norm": 0.19745570421218872,
+      "learning_rate": 6.90751944684529e-05,
+      "loss": 0.273,
+      "step": 3792
+    },
+    {
+      "epoch": 19.638709677419357,
+      "grad_norm": 0.1355098932981491,
+      "learning_rate": 6.904062229904928e-05,
+      "loss": 0.2938,
+      "step": 3793
+    },
+    {
+      "epoch": 19.643870967741936,
+      "grad_norm": 0.25549477338790894,
+      "learning_rate": 6.900605012964564e-05,
+      "loss": 0.2933,
+      "step": 3794
+    },
+    {
+      "epoch": 19.649032258064516,
+      "grad_norm": 0.1447935253381729,
+      "learning_rate": 6.897147796024201e-05,
+      "loss": 0.2695,
+      "step": 3795
+    },
+    {
+      "epoch": 19.654193548387095,
+      "grad_norm": 0.22543799877166748,
+      "learning_rate": 6.893690579083838e-05,
+      "loss": 0.2773,
+      "step": 3796
+    },
+    {
+      "epoch": 19.659354838709678,
+      "grad_norm": 0.2157682627439499,
+      "learning_rate": 6.890233362143474e-05,
+      "loss": 0.2927,
+      "step": 3797
+    },
+    {
+      "epoch": 19.664516129032258,
+      "grad_norm": 0.256872296333313,
+      "learning_rate": 6.886776145203112e-05,
+      "loss": 0.2992,
+      "step": 3798
+    },
+    {
+      "epoch": 19.669677419354837,
+      "grad_norm": 0.14030246436595917,
+      "learning_rate": 6.883318928262749e-05,
+      "loss": 0.2977,
+      "step": 3799
+    },
+    {
+      "epoch": 19.67483870967742,
+      "grad_norm": 0.19909608364105225,
+      "learning_rate": 6.879861711322386e-05,
+      "loss": 0.2727,
+      "step": 3800
+    },
+    {
+      "epoch": 19.68,
+      "grad_norm": 0.220820352435112,
+      "learning_rate": 6.876404494382023e-05,
+      "loss": 0.2899,
+      "step": 3801
+    },
+    {
+      "epoch": 19.68516129032258,
+      "grad_norm": 0.1583511233329773,
+      "learning_rate": 6.87294727744166e-05,
+      "loss": 0.2758,
+      "step": 3802
+    },
+    {
+      "epoch": 19.690322580645162,
+      "grad_norm": 0.18327906727790833,
+      "learning_rate": 6.869490060501297e-05,
+      "loss": 0.2786,
+      "step": 3803
+    },
+    {
+      "epoch": 19.695483870967742,
+      "grad_norm": 0.16285920143127441,
+      "learning_rate": 6.866032843560934e-05,
+      "loss": 0.3031,
+      "step": 3804
+    },
+    {
+      "epoch": 19.70064516129032,
+      "grad_norm": 0.2016686648130417,
+      "learning_rate": 6.862575626620571e-05,
+      "loss": 0.2888,
+      "step": 3805
+    },
+    {
+      "epoch": 19.705806451612904,
+      "grad_norm": 0.20480754971504211,
+      "learning_rate": 6.859118409680207e-05,
+      "loss": 0.2931,
+      "step": 3806
+    },
+    {
+      "epoch": 19.710967741935484,
+      "grad_norm": 0.2558112144470215,
+      "learning_rate": 6.855661192739845e-05,
+      "loss": 0.2576,
+      "step": 3807
+    },
+    {
+      "epoch": 19.716129032258063,
+      "grad_norm": 0.14005045592784882,
+      "learning_rate": 6.852203975799482e-05,
+      "loss": 0.2895,
+      "step": 3808
+    },
+    {
+      "epoch": 19.721290322580646,
+      "grad_norm": 0.21739013493061066,
+      "learning_rate": 6.848746758859119e-05,
+      "loss": 0.2869,
+      "step": 3809
+    },
+    {
+      "epoch": 19.726451612903226,
+      "grad_norm": 0.21150749921798706,
+      "learning_rate": 6.845289541918756e-05,
+      "loss": 0.2978,
+      "step": 3810
+    },
+    {
+      "epoch": 19.731612903225805,
+      "grad_norm": 0.2756323218345642,
+      "learning_rate": 6.841832324978392e-05,
+      "loss": 0.3067,
+      "step": 3811
+    },
+    {
+      "epoch": 19.73677419354839,
+      "grad_norm": 0.18218007683753967,
+      "learning_rate": 6.83837510803803e-05,
+      "loss": 0.2521,
+      "step": 3812
+    },
+    {
+      "epoch": 19.741935483870968,
+      "grad_norm": 0.15922443568706512,
+      "learning_rate": 6.834917891097667e-05,
+      "loss": 0.271,
+      "step": 3813
+    },
+    {
+      "epoch": 19.747096774193547,
+      "grad_norm": 0.16447433829307556,
+      "learning_rate": 6.831460674157304e-05,
+      "loss": 0.2759,
+      "step": 3814
+    },
+    {
+      "epoch": 19.75225806451613,
+      "grad_norm": 0.1795089840888977,
+      "learning_rate": 6.82800345721694e-05,
+      "loss": 0.2937,
+      "step": 3815
+    },
+    {
+      "epoch": 19.75741935483871,
+      "grad_norm": 0.2206021100282669,
+      "learning_rate": 6.824546240276578e-05,
+      "loss": 0.3058,
+      "step": 3816
+    },
+    {
+      "epoch": 19.76258064516129,
+      "grad_norm": 0.21043117344379425,
+      "learning_rate": 6.821089023336215e-05,
+      "loss": 0.3209,
+      "step": 3817
+    },
+    {
+      "epoch": 19.767741935483873,
+      "grad_norm": 0.14234735071659088,
+      "learning_rate": 6.817631806395852e-05,
+      "loss": 0.2627,
+      "step": 3818
+    },
+    {
+      "epoch": 19.772903225806452,
+      "grad_norm": 0.15274177491664886,
+      "learning_rate": 6.814174589455489e-05,
+      "loss": 0.2714,
+      "step": 3819
+    },
+    {
+      "epoch": 19.77806451612903,
+      "grad_norm": 0.17628660798072815,
+      "learning_rate": 6.810717372515125e-05,
+      "loss": 0.287,
+      "step": 3820
+    },
+    {
+      "epoch": 19.78322580645161,
+      "grad_norm": 0.14657282829284668,
+      "learning_rate": 6.807260155574763e-05,
+      "loss": 0.2675,
+      "step": 3821
+    },
+    {
+      "epoch": 19.788387096774194,
+      "grad_norm": 0.24634629487991333,
+      "learning_rate": 6.8038029386344e-05,
+      "loss": 0.3157,
+      "step": 3822
+    },
+    {
+      "epoch": 19.793548387096774,
+      "grad_norm": 0.18132056295871735,
+      "learning_rate": 6.800345721694037e-05,
+      "loss": 0.2992,
+      "step": 3823
+    },
+    {
+      "epoch": 19.798709677419353,
+      "grad_norm": 0.20292046666145325,
+      "learning_rate": 6.796888504753673e-05,
+      "loss": 0.2924,
+      "step": 3824
+    },
+    {
+      "epoch": 19.803870967741936,
+      "grad_norm": 0.13152319192886353,
+      "learning_rate": 6.793431287813311e-05,
+      "loss": 0.2571,
+      "step": 3825
+    },
+    {
+      "epoch": 19.809032258064516,
+      "grad_norm": 0.16888491809368134,
+      "learning_rate": 6.789974070872948e-05,
+      "loss": 0.2746,
+      "step": 3826
+    },
+    {
+      "epoch": 19.814193548387095,
+      "grad_norm": 0.2827218472957611,
+      "learning_rate": 6.786516853932583e-05,
+      "loss": 0.2822,
+      "step": 3827
+    },
+    {
+      "epoch": 19.81935483870968,
+      "grad_norm": 0.1586296409368515,
+      "learning_rate": 6.783059636992222e-05,
+      "loss": 0.2952,
+      "step": 3828
+    },
+    {
+      "epoch": 19.824516129032258,
+      "grad_norm": 0.28092968463897705,
+      "learning_rate": 6.779602420051858e-05,
+      "loss": 0.2558,
+      "step": 3829
+    },
+    {
+      "epoch": 19.829677419354837,
+      "grad_norm": 0.15690124034881592,
+      "learning_rate": 6.776145203111496e-05,
+      "loss": 0.2723,
+      "step": 3830
+    },
+    {
+      "epoch": 19.83483870967742,
+      "grad_norm": 0.1965457797050476,
+      "learning_rate": 6.772687986171133e-05,
+      "loss": 0.2291,
+      "step": 3831
+    },
+    {
+      "epoch": 19.84,
+      "grad_norm": 0.19047464430332184,
+      "learning_rate": 6.76923076923077e-05,
+      "loss": 0.2662,
+      "step": 3832
+    },
+    {
+      "epoch": 19.84516129032258,
+      "grad_norm": 0.20515355467796326,
+      "learning_rate": 6.765773552290406e-05,
+      "loss": 0.304,
+      "step": 3833
+    },
+    {
+      "epoch": 19.850322580645162,
+      "grad_norm": 0.18643818795681,
+      "learning_rate": 6.762316335350044e-05,
+      "loss": 0.3173,
+      "step": 3834
+    },
+    {
+      "epoch": 19.855483870967742,
+      "grad_norm": 0.14206133782863617,
+      "learning_rate": 6.758859118409681e-05,
+      "loss": 0.272,
+      "step": 3835
+    },
+    {
+      "epoch": 19.86064516129032,
+      "grad_norm": 0.20171724259853363,
+      "learning_rate": 6.755401901469316e-05,
+      "loss": 0.2939,
+      "step": 3836
+    },
+    {
+      "epoch": 19.865806451612904,
+      "grad_norm": 0.16068190336227417,
+      "learning_rate": 6.751944684528955e-05,
+      "loss": 0.2345,
+      "step": 3837
+    },
+    {
+      "epoch": 19.870967741935484,
+      "grad_norm": 0.21378830075263977,
+      "learning_rate": 6.748487467588591e-05,
+      "loss": 0.3017,
+      "step": 3838
+    },
+    {
+      "epoch": 19.876129032258063,
+      "grad_norm": 0.2554721534252167,
+      "learning_rate": 6.745030250648229e-05,
+      "loss": 0.3102,
+      "step": 3839
+    },
+    {
+      "epoch": 19.881290322580647,
+      "grad_norm": 0.1842278242111206,
+      "learning_rate": 6.741573033707866e-05,
+      "loss": 0.2622,
+      "step": 3840
+    },
+    {
+      "epoch": 19.886451612903226,
+      "grad_norm": 0.3018702268600464,
+      "learning_rate": 6.738115816767503e-05,
+      "loss": 0.2997,
+      "step": 3841
+    },
+    {
+      "epoch": 19.891612903225806,
+      "grad_norm": 0.18447451293468475,
+      "learning_rate": 6.73465859982714e-05,
+      "loss": 0.3119,
+      "step": 3842
+    },
+    {
+      "epoch": 19.89677419354839,
+      "grad_norm": 0.26525551080703735,
+      "learning_rate": 6.731201382886776e-05,
+      "loss": 0.2796,
+      "step": 3843
+    },
+    {
+      "epoch": 19.901935483870968,
+      "grad_norm": 0.18791264295578003,
+      "learning_rate": 6.727744165946414e-05,
+      "loss": 0.293,
+      "step": 3844
+    },
+    {
+      "epoch": 19.907096774193548,
+      "grad_norm": 0.21040455996990204,
+      "learning_rate": 6.72428694900605e-05,
+      "loss": 0.2572,
+      "step": 3845
+    },
+    {
+      "epoch": 19.91225806451613,
+      "grad_norm": 0.12429387122392654,
+      "learning_rate": 6.720829732065688e-05,
+      "loss": 0.2824,
+      "step": 3846
+    },
+    {
+      "epoch": 19.91741935483871,
+      "grad_norm": 0.20797260105609894,
+      "learning_rate": 6.717372515125324e-05,
+      "loss": 0.2939,
+      "step": 3847
+    },
+    {
+      "epoch": 19.92258064516129,
+      "grad_norm": 0.18547578155994415,
+      "learning_rate": 6.713915298184962e-05,
+      "loss": 0.3252,
+      "step": 3848
+    },
+    {
+      "epoch": 19.927741935483873,
+      "grad_norm": 0.1721017211675644,
+      "learning_rate": 6.710458081244598e-05,
+      "loss": 0.2902,
+      "step": 3849
+    },
+    {
+      "epoch": 19.932903225806452,
+      "grad_norm": 0.20148113369941711,
+      "learning_rate": 6.707000864304236e-05,
+      "loss": 0.3067,
+      "step": 3850
+    },
+    {
+      "epoch": 19.93806451612903,
+      "grad_norm": 0.35284894704818726,
+      "learning_rate": 6.703543647363872e-05,
+      "loss": 0.3169,
+      "step": 3851
+    },
+    {
+      "epoch": 19.94322580645161,
+      "grad_norm": 0.29464128613471985,
+      "learning_rate": 6.700086430423509e-05,
+      "loss": 0.2796,
+      "step": 3852
+    },
+    {
+      "epoch": 19.948387096774194,
+      "grad_norm": 0.20987851917743683,
+      "learning_rate": 6.696629213483147e-05,
+      "loss": 0.269,
+      "step": 3853
+    },
+    {
+      "epoch": 19.953548387096774,
+      "grad_norm": 0.162649467587471,
+      "learning_rate": 6.693171996542782e-05,
+      "loss": 0.2876,
+      "step": 3854
+    },
+    {
+      "epoch": 19.958709677419353,
+      "grad_norm": 0.2108410745859146,
+      "learning_rate": 6.68971477960242e-05,
+      "loss": 0.2868,
+      "step": 3855
+    },
+    {
+      "epoch": 19.963870967741936,
+      "grad_norm": 0.1862463802099228,
+      "learning_rate": 6.686257562662057e-05,
+      "loss": 0.2905,
+      "step": 3856
+    },
+    {
+      "epoch": 19.969032258064516,
+      "grad_norm": 0.19103309512138367,
+      "learning_rate": 6.682800345721695e-05,
+      "loss": 0.2766,
+      "step": 3857
+    },
+    {
+      "epoch": 19.974193548387095,
+      "grad_norm": 0.1366184651851654,
+      "learning_rate": 6.67934312878133e-05,
+      "loss": 0.28,
+      "step": 3858
+    },
+    {
+      "epoch": 19.97935483870968,
+      "grad_norm": 0.1719592958688736,
+      "learning_rate": 6.675885911840967e-05,
+      "loss": 0.2681,
+      "step": 3859
+    },
+    {
+      "epoch": 19.984516129032258,
+      "grad_norm": 0.1863771229982376,
+      "learning_rate": 6.672428694900605e-05,
+      "loss": 0.2894,
+      "step": 3860
+    },
+    {
+      "epoch": 19.989677419354837,
+      "grad_norm": 0.14187979698181152,
+      "learning_rate": 6.668971477960242e-05,
+      "loss": 0.2853,
+      "step": 3861
+    },
+    {
+      "epoch": 19.99483870967742,
+      "grad_norm": 0.1258918195962906,
+      "learning_rate": 6.66551426101988e-05,
+      "loss": 0.2922,
+      "step": 3862
+    },
+    {
+      "epoch": 20.001290322580644,
+      "grad_norm": 0.25043925642967224,
+      "learning_rate": 6.662057044079515e-05,
+      "loss": 0.4783,
+      "step": 3863
+    },
+    {
+      "epoch": 20.006451612903227,
+      "grad_norm": 0.11905229091644287,
+      "learning_rate": 6.658599827139154e-05,
+      "loss": 0.3023,
+      "step": 3864
+    },
+    {
+      "epoch": 20.011612903225807,
+      "grad_norm": 0.03801474720239639,
+      "learning_rate": 6.65514261019879e-05,
+      "loss": 0.2713,
+      "step": 3865
+    },
+    {
+      "epoch": 20.016774193548386,
+      "grad_norm": 0.0792052373290062,
+      "learning_rate": 6.651685393258428e-05,
+      "loss": 0.2766,
+      "step": 3866
+    },
+    {
+      "epoch": 20.02193548387097,
+      "grad_norm": 0.03900390490889549,
+      "learning_rate": 6.648228176318064e-05,
+      "loss": 0.2905,
+      "step": 3867
+    },
+    {
+      "epoch": 20.02709677419355,
+      "grad_norm": 0.0819995105266571,
+      "learning_rate": 6.6447709593777e-05,
+      "loss": 0.3128,
+      "step": 3868
+    },
+    {
+      "epoch": 20.032258064516128,
+      "grad_norm": 0.10068647563457489,
+      "learning_rate": 6.641313742437338e-05,
+      "loss": 0.2646,
+      "step": 3869
+    },
+    {
+      "epoch": 20.03741935483871,
+      "grad_norm": 0.04621458053588867,
+      "learning_rate": 6.637856525496975e-05,
+      "loss": 0.2598,
+      "step": 3870
+    },
+    {
+      "epoch": 20.04258064516129,
+      "grad_norm": 0.11847042292356491,
+      "learning_rate": 6.634399308556613e-05,
+      "loss": 0.3012,
+      "step": 3871
+    },
+    {
+      "epoch": 20.04774193548387,
+      "grad_norm": 0.24337907135486603,
+      "learning_rate": 6.630942091616248e-05,
+      "loss": 0.2546,
+      "step": 3872
+    },
+    {
+      "epoch": 20.052903225806453,
+      "grad_norm": 0.0882343277335167,
+      "learning_rate": 6.627484874675887e-05,
+      "loss": 0.2963,
+      "step": 3873
+    },
+    {
+      "epoch": 20.058064516129033,
+      "grad_norm": 0.06064406409859657,
+      "learning_rate": 6.624027657735523e-05,
+      "loss": 0.3151,
+      "step": 3874
+    },
+    {
+      "epoch": 20.063225806451612,
+      "grad_norm": 0.1020735502243042,
+      "learning_rate": 6.620570440795161e-05,
+      "loss": 0.2744,
+      "step": 3875
+    },
+    {
+      "epoch": 20.068387096774195,
+      "grad_norm": 0.06819847971200943,
+      "learning_rate": 6.617113223854797e-05,
+      "loss": 0.2772,
+      "step": 3876
+    },
+    {
+      "epoch": 20.073548387096775,
+      "grad_norm": 0.16004006564617157,
+      "learning_rate": 6.613656006914433e-05,
+      "loss": 0.2885,
+      "step": 3877
+    },
+    {
+      "epoch": 20.078709677419354,
+      "grad_norm": 0.14454513788223267,
+      "learning_rate": 6.610198789974071e-05,
+      "loss": 0.2614,
+      "step": 3878
+    },
+    {
+      "epoch": 20.083870967741934,
+      "grad_norm": 0.05636129155755043,
+      "learning_rate": 6.606741573033708e-05,
+      "loss": 0.3089,
+      "step": 3879
+    },
+    {
+      "epoch": 20.089032258064517,
+      "grad_norm": 0.03750108554959297,
+      "learning_rate": 6.603284356093346e-05,
+      "loss": 0.275,
+      "step": 3880
+    },
+    {
+      "epoch": 20.094193548387096,
+      "grad_norm": 0.1638735681772232,
+      "learning_rate": 6.599827139152981e-05,
+      "loss": 0.3028,
+      "step": 3881
+    },
+    {
+      "epoch": 20.099354838709676,
+      "grad_norm": 0.08257674425840378,
+      "learning_rate": 6.59636992221262e-05,
+      "loss": 0.2838,
+      "step": 3882
+    },
+    {
+      "epoch": 20.10451612903226,
+      "grad_norm": 0.0834834948182106,
+      "learning_rate": 6.592912705272256e-05,
+      "loss": 0.2819,
+      "step": 3883
+    },
+    {
+      "epoch": 20.10967741935484,
+      "grad_norm": 0.19385063648223877,
+      "learning_rate": 6.589455488331893e-05,
+      "loss": 0.2697,
+      "step": 3884
+    },
+    {
+      "epoch": 20.114838709677418,
+      "grad_norm": 0.17644977569580078,
+      "learning_rate": 6.58599827139153e-05,
+      "loss": 0.3126,
+      "step": 3885
+    },
+    {
+      "epoch": 20.12,
+      "grad_norm": 0.0788203701376915,
+      "learning_rate": 6.582541054451166e-05,
+      "loss": 0.296,
+      "step": 3886
+    },
+    {
+      "epoch": 20.12516129032258,
+      "grad_norm": 0.07421212643384933,
+      "learning_rate": 6.579083837510804e-05,
+      "loss": 0.269,
+      "step": 3887
+    },
+    {
+      "epoch": 20.13032258064516,
+      "grad_norm": 0.15365798771381378,
+      "learning_rate": 6.575626620570441e-05,
+      "loss": 0.2748,
+      "step": 3888
+    },
+    {
+      "epoch": 20.135483870967743,
+      "grad_norm": 0.06854312121868134,
+      "learning_rate": 6.572169403630079e-05,
+      "loss": 0.2832,
+      "step": 3889
+    },
+    {
+      "epoch": 20.140645161290323,
+      "grad_norm": 0.19880817830562592,
+      "learning_rate": 6.568712186689714e-05,
+      "loss": 0.2989,
+      "step": 3890
+    },
+    {
+      "epoch": 20.145806451612902,
+      "grad_norm": 0.05790577828884125,
+      "learning_rate": 6.565254969749353e-05,
+      "loss": 0.2835,
+      "step": 3891
+    },
+    {
+      "epoch": 20.150967741935485,
+      "grad_norm": 0.14543989300727844,
+      "learning_rate": 6.561797752808989e-05,
+      "loss": 0.2955,
+      "step": 3892
+    },
+    {
+      "epoch": 20.156129032258065,
+      "grad_norm": 0.07881079614162445,
+      "learning_rate": 6.558340535868626e-05,
+      "loss": 0.2851,
+      "step": 3893
+    },
+    {
+      "epoch": 20.161290322580644,
+      "grad_norm": 0.13604287803173065,
+      "learning_rate": 6.554883318928263e-05,
+      "loss": 0.287,
+      "step": 3894
+    },
+    {
+      "epoch": 20.166451612903227,
+      "grad_norm": 0.08428359776735306,
+      "learning_rate": 6.551426101987899e-05,
+      "loss": 0.2551,
+      "step": 3895
+    },
+    {
+      "epoch": 20.171612903225807,
+      "grad_norm": 0.08471115678548813,
+      "learning_rate": 6.547968885047537e-05,
+      "loss": 0.2682,
+      "step": 3896
+    },
+    {
+      "epoch": 20.176774193548386,
+      "grad_norm": 0.13170702755451202,
+      "learning_rate": 6.544511668107174e-05,
+      "loss": 0.2856,
+      "step": 3897
+    },
+    {
+      "epoch": 20.18193548387097,
+      "grad_norm": 0.10338020324707031,
+      "learning_rate": 6.541054451166812e-05,
+      "loss": 0.2789,
+      "step": 3898
+    },
+    {
+      "epoch": 20.18709677419355,
+      "grad_norm": 0.09787094593048096,
+      "learning_rate": 6.537597234226447e-05,
+      "loss": 0.2751,
+      "step": 3899
+    },
+    {
+      "epoch": 20.19225806451613,
+      "grad_norm": 0.12093412131071091,
+      "learning_rate": 6.534140017286084e-05,
+      "loss": 0.2811,
+      "step": 3900
+    },
+    {
+      "epoch": 20.19741935483871,
+      "grad_norm": 0.08628757297992706,
+      "learning_rate": 6.530682800345722e-05,
+      "loss": 0.3056,
+      "step": 3901
+    },
+    {
+      "epoch": 20.20258064516129,
+      "grad_norm": 0.10647644847631454,
+      "learning_rate": 6.527225583405359e-05,
+      "loss": 0.2801,
+      "step": 3902
+    },
+    {
+      "epoch": 20.20774193548387,
+      "grad_norm": 0.19986313581466675,
+      "learning_rate": 6.523768366464996e-05,
+      "loss": 0.2743,
+      "step": 3903
+    },
+    {
+      "epoch": 20.21290322580645,
+      "grad_norm": 0.1564554125070572,
+      "learning_rate": 6.520311149524632e-05,
+      "loss": 0.2854,
+      "step": 3904
+    },
+    {
+      "epoch": 20.218064516129033,
+      "grad_norm": 0.07775875926017761,
+      "learning_rate": 6.51685393258427e-05,
+      "loss": 0.2759,
+      "step": 3905
+    },
+    {
+      "epoch": 20.223225806451612,
+      "grad_norm": 0.06307309120893478,
+      "learning_rate": 6.513396715643907e-05,
+      "loss": 0.2576,
+      "step": 3906
+    },
+    {
+      "epoch": 20.228387096774192,
+      "grad_norm": 0.06808029115200043,
+      "learning_rate": 6.509939498703545e-05,
+      "loss": 0.2916,
+      "step": 3907
+    },
+    {
+      "epoch": 20.233548387096775,
+      "grad_norm": 0.12024766206741333,
+      "learning_rate": 6.50648228176318e-05,
+      "loss": 0.2863,
+      "step": 3908
+    },
+    {
+      "epoch": 20.238709677419354,
+      "grad_norm": 0.1171506941318512,
+      "learning_rate": 6.503025064822817e-05,
+      "loss": 0.2949,
+      "step": 3909
+    },
+    {
+      "epoch": 20.243870967741934,
+      "grad_norm": 0.11385706067085266,
+      "learning_rate": 6.499567847882455e-05,
+      "loss": 0.2608,
+      "step": 3910
+    },
+    {
+      "epoch": 20.249032258064517,
+      "grad_norm": 0.13404999673366547,
+      "learning_rate": 6.496110630942092e-05,
+      "loss": 0.2795,
+      "step": 3911
+    },
+    {
+      "epoch": 20.254193548387097,
+      "grad_norm": 0.1291045844554901,
+      "learning_rate": 6.492653414001729e-05,
+      "loss": 0.3006,
+      "step": 3912
+    },
+    {
+      "epoch": 20.259354838709676,
+      "grad_norm": 0.09561709314584732,
+      "learning_rate": 6.489196197061365e-05,
+      "loss": 0.2653,
+      "step": 3913
+    },
+    {
+      "epoch": 20.26451612903226,
+      "grad_norm": 0.10968935489654541,
+      "learning_rate": 6.485738980121003e-05,
+      "loss": 0.2693,
+      "step": 3914
+    },
+    {
+      "epoch": 20.26967741935484,
+      "grad_norm": 0.14493335783481598,
+      "learning_rate": 6.48228176318064e-05,
+      "loss": 0.2963,
+      "step": 3915
+    },
+    {
+      "epoch": 20.274838709677418,
+      "grad_norm": 0.10407134890556335,
+      "learning_rate": 6.478824546240277e-05,
+      "loss": 0.2756,
+      "step": 3916
+    },
+    {
+      "epoch": 20.28,
+      "grad_norm": 0.04702259600162506,
+      "learning_rate": 6.475367329299913e-05,
+      "loss": 0.2718,
+      "step": 3917
+    },
+    {
+      "epoch": 20.28516129032258,
+      "grad_norm": 0.039372414350509644,
+      "learning_rate": 6.47191011235955e-05,
+      "loss": 0.28,
+      "step": 3918
+    },
+    {
+      "epoch": 20.29032258064516,
+      "grad_norm": 0.04738412797451019,
+      "learning_rate": 6.468452895419188e-05,
+      "loss": 0.3101,
+      "step": 3919
+    },
+    {
+      "epoch": 20.295483870967743,
+      "grad_norm": 0.09338925033807755,
+      "learning_rate": 6.464995678478825e-05,
+      "loss": 0.2952,
+      "step": 3920
+    },
+    {
+      "epoch": 20.300645161290323,
+      "grad_norm": 0.11271320283412933,
+      "learning_rate": 6.461538461538462e-05,
+      "loss": 0.2837,
+      "step": 3921
+    },
+    {
+      "epoch": 20.305806451612902,
+      "grad_norm": 0.07936187088489532,
+      "learning_rate": 6.458081244598098e-05,
+      "loss": 0.2632,
+      "step": 3922
+    },
+    {
+      "epoch": 20.310967741935485,
+      "grad_norm": 0.07116696238517761,
+      "learning_rate": 6.454624027657736e-05,
+      "loss": 0.2704,
+      "step": 3923
+    },
+    {
+      "epoch": 20.316129032258065,
+      "grad_norm": 0.13182641565799713,
+      "learning_rate": 6.451166810717373e-05,
+      "loss": 0.3046,
+      "step": 3924
+    },
+    {
+      "epoch": 20.321290322580644,
+      "grad_norm": 0.04023564234375954,
+      "learning_rate": 6.44770959377701e-05,
+      "loss": 0.2952,
+      "step": 3925
+    },
+    {
+      "epoch": 20.326451612903227,
+      "grad_norm": 0.10479599237442017,
+      "learning_rate": 6.444252376836646e-05,
+      "loss": 0.2767,
+      "step": 3926
+    },
+    {
+      "epoch": 20.331612903225807,
+      "grad_norm": 0.033570244908332825,
+      "learning_rate": 6.440795159896283e-05,
+      "loss": 0.2745,
+      "step": 3927
+    },
+    {
+      "epoch": 20.336774193548386,
+      "grad_norm": 0.15022867918014526,
+      "learning_rate": 6.437337942955921e-05,
+      "loss": 0.2564,
+      "step": 3928
+    },
+    {
+      "epoch": 20.34193548387097,
+      "grad_norm": 0.12927046418190002,
+      "learning_rate": 6.433880726015558e-05,
+      "loss": 0.2972,
+      "step": 3929
+    },
+    {
+      "epoch": 20.34709677419355,
+      "grad_norm": 0.1067814826965332,
+      "learning_rate": 6.430423509075195e-05,
+      "loss": 0.2966,
+      "step": 3930
+    },
+    {
+      "epoch": 20.35225806451613,
+      "grad_norm": 0.12175469100475311,
+      "learning_rate": 6.426966292134831e-05,
+      "loss": 0.2656,
+      "step": 3931
+    },
+    {
+      "epoch": 20.35741935483871,
+      "grad_norm": 0.11694155633449554,
+      "learning_rate": 6.423509075194468e-05,
+      "loss": 0.2848,
+      "step": 3932
+    },
+    {
+      "epoch": 20.36258064516129,
+      "grad_norm": 0.04336576163768768,
+      "learning_rate": 6.420051858254106e-05,
+      "loss": 0.2802,
+      "step": 3933
+    },
+    {
+      "epoch": 20.36774193548387,
+      "grad_norm": 0.19161789119243622,
+      "learning_rate": 6.416594641313743e-05,
+      "loss": 0.2975,
+      "step": 3934
+    },
+    {
+      "epoch": 20.37290322580645,
+      "grad_norm": 0.05343884974718094,
+      "learning_rate": 6.41313742437338e-05,
+      "loss": 0.2918,
+      "step": 3935
+    },
+    {
+      "epoch": 20.378064516129033,
+      "grad_norm": 0.15427429974079132,
+      "learning_rate": 6.409680207433016e-05,
+      "loss": 0.2941,
+      "step": 3936
+    },
+    {
+      "epoch": 20.383225806451613,
+      "grad_norm": 0.15056976675987244,
+      "learning_rate": 6.406222990492654e-05,
+      "loss": 0.3031,
+      "step": 3937
+    },
+    {
+      "epoch": 20.388387096774192,
+      "grad_norm": 0.052983302623033524,
+      "learning_rate": 6.402765773552291e-05,
+      "loss": 0.2323,
+      "step": 3938
+    },
+    {
+      "epoch": 20.393548387096775,
+      "grad_norm": 0.10543406009674072,
+      "learning_rate": 6.399308556611928e-05,
+      "loss": 0.2858,
+      "step": 3939
+    },
+    {
+      "epoch": 20.398709677419355,
+      "grad_norm": 0.18248103559017181,
+      "learning_rate": 6.395851339671564e-05,
+      "loss": 0.2759,
+      "step": 3940
+    },
+    {
+      "epoch": 20.403870967741934,
+      "grad_norm": 0.129746675491333,
+      "learning_rate": 6.392394122731201e-05,
+      "loss": 0.2729,
+      "step": 3941
+    },
+    {
+      "epoch": 20.409032258064517,
+      "grad_norm": 0.14153452217578888,
+      "learning_rate": 6.388936905790839e-05,
+      "loss": 0.2627,
+      "step": 3942
+    },
+    {
+      "epoch": 20.414193548387097,
+      "grad_norm": 0.16306671500205994,
+      "learning_rate": 6.385479688850476e-05,
+      "loss": 0.2994,
+      "step": 3943
+    },
+    {
+      "epoch": 20.419354838709676,
+      "grad_norm": 0.07368214428424835,
+      "learning_rate": 6.382022471910112e-05,
+      "loss": 0.2853,
+      "step": 3944
+    },
+    {
+      "epoch": 20.42451612903226,
+      "grad_norm": 0.0979842096567154,
+      "learning_rate": 6.378565254969749e-05,
+      "loss": 0.2614,
+      "step": 3945
+    },
+    {
+      "epoch": 20.42967741935484,
+      "grad_norm": 0.055947497487068176,
+      "learning_rate": 6.375108038029387e-05,
+      "loss": 0.2772,
+      "step": 3946
+    },
+    {
+      "epoch": 20.434838709677418,
+      "grad_norm": 0.06784556061029434,
+      "learning_rate": 6.371650821089024e-05,
+      "loss": 0.2579,
+      "step": 3947
+    },
+    {
+      "epoch": 20.44,
+      "grad_norm": 0.11327005922794342,
+      "learning_rate": 6.36819360414866e-05,
+      "loss": 0.2733,
+      "step": 3948
+    },
+    {
+      "epoch": 20.44516129032258,
+      "grad_norm": 0.08959171921014786,
+      "learning_rate": 6.364736387208297e-05,
+      "loss": 0.2965,
+      "step": 3949
+    },
+    {
+      "epoch": 20.45032258064516,
+      "grad_norm": 0.10187896341085434,
+      "learning_rate": 6.361279170267934e-05,
+      "loss": 0.2712,
+      "step": 3950
+    },
+    {
+      "epoch": 20.455483870967743,
+      "grad_norm": 0.10792164504528046,
+      "learning_rate": 6.357821953327572e-05,
+      "loss": 0.2808,
+      "step": 3951
+    },
+    {
+      "epoch": 20.460645161290323,
+      "grad_norm": 0.09317187964916229,
+      "learning_rate": 6.354364736387209e-05,
+      "loss": 0.2443,
+      "step": 3952
+    },
+    {
+      "epoch": 20.465806451612902,
+      "grad_norm": 0.03989115357398987,
+      "learning_rate": 6.350907519446845e-05,
+      "loss": 0.2565,
+      "step": 3953
+    },
+    {
+      "epoch": 20.470967741935485,
+      "grad_norm": 0.11695244908332825,
+      "learning_rate": 6.347450302506482e-05,
+      "loss": 0.3091,
+      "step": 3954
+    },
+    {
+      "epoch": 20.476129032258065,
+      "grad_norm": 0.09911570698022842,
+      "learning_rate": 6.34399308556612e-05,
+      "loss": 0.2781,
+      "step": 3955
+    },
+    {
+      "epoch": 20.481290322580644,
+      "grad_norm": 0.1197800487279892,
+      "learning_rate": 6.340535868625757e-05,
+      "loss": 0.291,
+      "step": 3956
+    },
+    {
+      "epoch": 20.486451612903227,
+      "grad_norm": 0.05094648152589798,
+      "learning_rate": 6.337078651685394e-05,
+      "loss": 0.3053,
+      "step": 3957
+    },
+    {
+      "epoch": 20.491612903225807,
+      "grad_norm": 0.13886436820030212,
+      "learning_rate": 6.33362143474503e-05,
+      "loss": 0.2991,
+      "step": 3958
+    },
+    {
+      "epoch": 20.496774193548386,
+      "grad_norm": 0.16888068616390228,
+      "learning_rate": 6.330164217804667e-05,
+      "loss": 0.2647,
+      "step": 3959
+    },
+    {
+      "epoch": 20.501935483870966,
+      "grad_norm": 0.1301189363002777,
+      "learning_rate": 6.326707000864305e-05,
+      "loss": 0.263,
+      "step": 3960
+    },
+    {
+      "epoch": 20.50709677419355,
+      "grad_norm": 0.1613759994506836,
+      "learning_rate": 6.323249783923942e-05,
+      "loss": 0.2963,
+      "step": 3961
+    },
+    {
+      "epoch": 20.51225806451613,
+      "grad_norm": 0.06363151967525482,
+      "learning_rate": 6.319792566983578e-05,
+      "loss": 0.2645,
+      "step": 3962
+    },
+    {
+      "epoch": 20.517419354838708,
+      "grad_norm": 0.07827261090278625,
+      "learning_rate": 6.316335350043215e-05,
+      "loss": 0.2797,
+      "step": 3963
+    },
+    {
+      "epoch": 20.52258064516129,
+      "grad_norm": 0.04080735892057419,
+      "learning_rate": 6.312878133102853e-05,
+      "loss": 0.2726,
+      "step": 3964
+    },
+    {
+      "epoch": 20.52774193548387,
+      "grad_norm": 0.13636164367198944,
+      "learning_rate": 6.30942091616249e-05,
+      "loss": 0.2894,
+      "step": 3965
+    },
+    {
+      "epoch": 20.53290322580645,
+      "grad_norm": 0.13814345002174377,
+      "learning_rate": 6.305963699222127e-05,
+      "loss": 0.2667,
+      "step": 3966
+    },
+    {
+      "epoch": 20.538064516129033,
+      "grad_norm": 0.14234992861747742,
+      "learning_rate": 6.302506482281763e-05,
+      "loss": 0.288,
+      "step": 3967
+    },
+    {
+      "epoch": 20.543225806451613,
+      "grad_norm": 0.1020820140838623,
+      "learning_rate": 6.2990492653414e-05,
+      "loss": 0.2818,
+      "step": 3968
+    },
+    {
+      "epoch": 20.548387096774192,
+      "grad_norm": 0.0857647955417633,
+      "learning_rate": 6.295592048401038e-05,
+      "loss": 0.2904,
+      "step": 3969
+    },
+    {
+      "epoch": 20.553548387096775,
+      "grad_norm": 0.07671070098876953,
+      "learning_rate": 6.292134831460675e-05,
+      "loss": 0.2933,
+      "step": 3970
+    },
+    {
+      "epoch": 20.558709677419355,
+      "grad_norm": 0.06399421393871307,
+      "learning_rate": 6.288677614520311e-05,
+      "loss": 0.3113,
+      "step": 3971
+    },
+    {
+      "epoch": 20.563870967741934,
+      "grad_norm": 0.10550291836261749,
+      "learning_rate": 6.285220397579948e-05,
+      "loss": 0.3071,
+      "step": 3972
+    },
+    {
+      "epoch": 20.569032258064517,
+      "grad_norm": 0.11791706085205078,
+      "learning_rate": 6.281763180639585e-05,
+      "loss": 0.2764,
+      "step": 3973
+    },
+    {
+      "epoch": 20.574193548387097,
+      "grad_norm": 0.08903222531080246,
+      "learning_rate": 6.278305963699223e-05,
+      "loss": 0.2769,
+      "step": 3974
+    },
+    {
+      "epoch": 20.579354838709676,
+      "grad_norm": 0.08697506785392761,
+      "learning_rate": 6.27484874675886e-05,
+      "loss": 0.2963,
+      "step": 3975
+    },
+    {
+      "epoch": 20.58451612903226,
+      "grad_norm": 0.13451740145683289,
+      "learning_rate": 6.271391529818496e-05,
+      "loss": 0.3024,
+      "step": 3976
+    },
+    {
+      "epoch": 20.58967741935484,
+      "grad_norm": 0.09535746276378632,
+      "learning_rate": 6.267934312878133e-05,
+      "loss": 0.2806,
+      "step": 3977
+    },
+    {
+      "epoch": 20.59483870967742,
+      "grad_norm": 0.42878589034080505,
+      "learning_rate": 6.264477095937771e-05,
+      "loss": 0.2849,
+      "step": 3978
+    },
+    {
+      "epoch": 20.6,
+      "grad_norm": 0.11779217422008514,
+      "learning_rate": 6.261019878997408e-05,
+      "loss": 0.2447,
+      "step": 3979
+    },
+    {
+      "epoch": 20.60516129032258,
+      "grad_norm": 0.07385122776031494,
+      "learning_rate": 6.257562662057044e-05,
+      "loss": 0.2845,
+      "step": 3980
+    },
+    {
+      "epoch": 20.61032258064516,
+      "grad_norm": 0.17103971540927887,
+      "learning_rate": 6.254105445116681e-05,
+      "loss": 0.2794,
+      "step": 3981
+    },
+    {
+      "epoch": 20.615483870967743,
+      "grad_norm": 0.11945139616727829,
+      "learning_rate": 6.250648228176318e-05,
+      "loss": 0.2608,
+      "step": 3982
+    },
+    {
+      "epoch": 20.620645161290323,
+      "grad_norm": 0.04338609054684639,
+      "learning_rate": 6.247191011235956e-05,
+      "loss": 0.2914,
+      "step": 3983
+    },
+    {
+      "epoch": 20.625806451612902,
+      "grad_norm": 0.1713802069425583,
+      "learning_rate": 6.243733794295593e-05,
+      "loss": 0.2788,
+      "step": 3984
+    },
+    {
+      "epoch": 20.630967741935486,
+      "grad_norm": 0.03561143949627876,
+      "learning_rate": 6.240276577355229e-05,
+      "loss": 0.271,
+      "step": 3985
+    },
+    {
+      "epoch": 20.636129032258065,
+      "grad_norm": 0.13885456323623657,
+      "learning_rate": 6.236819360414866e-05,
+      "loss": 0.2626,
+      "step": 3986
+    },
+    {
+      "epoch": 20.641290322580645,
+      "grad_norm": 0.2103729546070099,
+      "learning_rate": 6.233362143474504e-05,
+      "loss": 0.2691,
+      "step": 3987
+    },
+    {
+      "epoch": 20.646451612903228,
+      "grad_norm": 0.05242002382874489,
+      "learning_rate": 6.229904926534141e-05,
+      "loss": 0.2887,
+      "step": 3988
+    },
+    {
+      "epoch": 20.651612903225807,
+      "grad_norm": 0.10470898449420929,
+      "learning_rate": 6.226447709593777e-05,
+      "loss": 0.2844,
+      "step": 3989
+    },
+    {
+      "epoch": 20.656774193548387,
+      "grad_norm": 0.09964100271463394,
+      "learning_rate": 6.222990492653414e-05,
+      "loss": 0.2947,
+      "step": 3990
+    },
+    {
+      "epoch": 20.661935483870966,
+      "grad_norm": 0.0864102765917778,
+      "learning_rate": 6.219533275713051e-05,
+      "loss": 0.2862,
+      "step": 3991
+    },
+    {
+      "epoch": 20.66709677419355,
+      "grad_norm": 0.05474426969885826,
+      "learning_rate": 6.216076058772689e-05,
+      "loss": 0.2785,
+      "step": 3992
+    },
+    {
+      "epoch": 20.67225806451613,
+      "grad_norm": 0.08402929455041885,
+      "learning_rate": 6.212618841832326e-05,
+      "loss": 0.2743,
+      "step": 3993
+    },
+    {
+      "epoch": 20.677419354838708,
+      "grad_norm": 0.055753495544195175,
+      "learning_rate": 6.209161624891962e-05,
+      "loss": 0.2719,
+      "step": 3994
+    },
+    {
+      "epoch": 20.68258064516129,
+      "grad_norm": 0.04045002534985542,
+      "learning_rate": 6.205704407951599e-05,
+      "loss": 0.293,
+      "step": 3995
+    },
+    {
+      "epoch": 20.68774193548387,
+      "grad_norm": 0.11494702100753784,
+      "learning_rate": 6.202247191011237e-05,
+      "loss": 0.2696,
+      "step": 3996
+    },
+    {
+      "epoch": 20.69290322580645,
+      "grad_norm": 0.1509348601102829,
+      "learning_rate": 6.198789974070874e-05,
+      "loss": 0.2865,
+      "step": 3997
+    },
+    {
+      "epoch": 20.698064516129033,
+      "grad_norm": 0.1454014927148819,
+      "learning_rate": 6.19533275713051e-05,
+      "loss": 0.2959,
+      "step": 3998
+    },
+    {
+      "epoch": 20.703225806451613,
+      "grad_norm": 0.21300610899925232,
+      "learning_rate": 6.191875540190147e-05,
+      "loss": 0.2551,
+      "step": 3999
+    },
+    {
+      "epoch": 20.708387096774192,
+      "grad_norm": 0.07360894978046417,
+      "learning_rate": 6.188418323249784e-05,
+      "loss": 0.3077,
+      "step": 4000
+    },
+    {
+      "epoch": 20.713548387096775,
+      "grad_norm": 0.13883359730243683,
+      "learning_rate": 6.184961106309422e-05,
+      "loss": 0.2997,
+      "step": 4001
+    },
+    {
+      "epoch": 20.718709677419355,
+      "grad_norm": 0.09540779143571854,
+      "learning_rate": 6.181503889369059e-05,
+      "loss": 0.2947,
+      "step": 4002
+    },
+    {
+      "epoch": 20.723870967741934,
+      "grad_norm": 0.1312161237001419,
+      "learning_rate": 6.178046672428695e-05,
+      "loss": 0.2848,
+      "step": 4003
+    },
+    {
+      "epoch": 20.729032258064517,
+      "grad_norm": 0.09314605593681335,
+      "learning_rate": 6.174589455488332e-05,
+      "loss": 0.2823,
+      "step": 4004
+    },
+    {
+      "epoch": 20.734193548387097,
+      "grad_norm": 0.04202805832028389,
+      "learning_rate": 6.17113223854797e-05,
+      "loss": 0.249,
+      "step": 4005
+    },
+    {
+      "epoch": 20.739354838709676,
+      "grad_norm": 0.2304699718952179,
+      "learning_rate": 6.167675021607607e-05,
+      "loss": 0.3002,
+      "step": 4006
+    },
+    {
+      "epoch": 20.74451612903226,
+      "grad_norm": 0.09856641292572021,
+      "learning_rate": 6.164217804667243e-05,
+      "loss": 0.2957,
+      "step": 4007
+    },
+    {
+      "epoch": 20.74967741935484,
+      "grad_norm": 0.18351200222969055,
+      "learning_rate": 6.16076058772688e-05,
+      "loss": 0.2824,
+      "step": 4008
+    },
+    {
+      "epoch": 20.75483870967742,
+      "grad_norm": 0.16292767226696014,
+      "learning_rate": 6.157303370786517e-05,
+      "loss": 0.3111,
+      "step": 4009
+    },
+    {
+      "epoch": 20.76,
+      "grad_norm": 0.104148268699646,
+      "learning_rate": 6.153846153846155e-05,
+      "loss": 0.2747,
+      "step": 4010
+    },
+    {
+      "epoch": 20.76516129032258,
+      "grad_norm": 0.07792364805936813,
+      "learning_rate": 6.150388936905792e-05,
+      "loss": 0.2937,
+      "step": 4011
+    },
+    {
+      "epoch": 20.77032258064516,
+      "grad_norm": 0.058677151799201965,
+      "learning_rate": 6.146931719965428e-05,
+      "loss": 0.3016,
+      "step": 4012
+    },
+    {
+      "epoch": 20.775483870967744,
+      "grad_norm": 0.13691481947898865,
+      "learning_rate": 6.143474503025065e-05,
+      "loss": 0.3057,
+      "step": 4013
+    },
+    {
+      "epoch": 20.780645161290323,
+      "grad_norm": 0.11276078969240189,
+      "learning_rate": 6.140017286084702e-05,
+      "loss": 0.2639,
+      "step": 4014
+    },
+    {
+      "epoch": 20.785806451612903,
+      "grad_norm": 0.11298991739749908,
+      "learning_rate": 6.13656006914434e-05,
+      "loss": 0.2802,
+      "step": 4015
+    },
+    {
+      "epoch": 20.790967741935482,
+      "grad_norm": 0.11148647218942642,
+      "learning_rate": 6.133102852203975e-05,
+      "loss": 0.2917,
+      "step": 4016
+    },
+    {
+      "epoch": 20.796129032258065,
+      "grad_norm": 0.04417587071657181,
+      "learning_rate": 6.129645635263613e-05,
+      "loss": 0.2958,
+      "step": 4017
+    },
+    {
+      "epoch": 20.801290322580645,
+      "grad_norm": 0.05133761838078499,
+      "learning_rate": 6.12618841832325e-05,
+      "loss": 0.2938,
+      "step": 4018
+    },
+    {
+      "epoch": 20.806451612903224,
+      "grad_norm": 0.0870768204331398,
+      "learning_rate": 6.122731201382888e-05,
+      "loss": 0.241,
+      "step": 4019
+    },
+    {
+      "epoch": 20.811612903225807,
+      "grad_norm": 0.07032786309719086,
+      "learning_rate": 6.119273984442525e-05,
+      "loss": 0.2863,
+      "step": 4020
+    },
+    {
+      "epoch": 20.816774193548387,
+      "grad_norm": 0.0968097597360611,
+      "learning_rate": 6.115816767502161e-05,
+      "loss": 0.3038,
+      "step": 4021
+    },
+    {
+      "epoch": 20.821935483870966,
+      "grad_norm": 0.046811096370220184,
+      "learning_rate": 6.112359550561798e-05,
+      "loss": 0.3002,
+      "step": 4022
+    },
+    {
+      "epoch": 20.82709677419355,
+      "grad_norm": 0.08002229779958725,
+      "learning_rate": 6.108902333621435e-05,
+      "loss": 0.2733,
+      "step": 4023
+    },
+    {
+      "epoch": 20.83225806451613,
+      "grad_norm": 0.0731302872300148,
+      "learning_rate": 6.105445116681073e-05,
+      "loss": 0.2945,
+      "step": 4024
+    },
+    {
+      "epoch": 20.83741935483871,
+      "grad_norm": 0.12341710925102234,
+      "learning_rate": 6.101987899740709e-05,
+      "loss": 0.2485,
+      "step": 4025
+    },
+    {
+      "epoch": 20.84258064516129,
+      "grad_norm": 0.08110646158456802,
+      "learning_rate": 6.098530682800346e-05,
+      "loss": 0.2384,
+      "step": 4026
+    },
+    {
+      "epoch": 20.84774193548387,
+      "grad_norm": 0.11236345022916794,
+      "learning_rate": 6.095073465859983e-05,
+      "loss": 0.2651,
+      "step": 4027
+    },
+    {
+      "epoch": 20.85290322580645,
+      "grad_norm": 0.047020573168992996,
+      "learning_rate": 6.09161624891962e-05,
+      "loss": 0.2734,
+      "step": 4028
+    },
+    {
+      "epoch": 20.858064516129033,
+      "grad_norm": 0.19602610170841217,
+      "learning_rate": 6.088159031979257e-05,
+      "loss": 0.2822,
+      "step": 4029
+    },
+    {
+      "epoch": 20.863225806451613,
+      "grad_norm": 0.15256863832473755,
+      "learning_rate": 6.0847018150388936e-05,
+      "loss": 0.2854,
+      "step": 4030
+    },
+    {
+      "epoch": 20.868387096774192,
+      "grad_norm": 0.12896233797073364,
+      "learning_rate": 6.081244598098531e-05,
+      "loss": 0.3008,
+      "step": 4031
+    },
+    {
+      "epoch": 20.873548387096776,
+      "grad_norm": 0.06143242120742798,
+      "learning_rate": 6.0777873811581677e-05,
+      "loss": 0.2804,
+      "step": 4032
+    },
+    {
+      "epoch": 20.878709677419355,
+      "grad_norm": 0.09025882184505463,
+      "learning_rate": 6.074330164217805e-05,
+      "loss": 0.2977,
+      "step": 4033
+    },
+    {
+      "epoch": 20.883870967741935,
+      "grad_norm": 0.08006056398153305,
+      "learning_rate": 6.070872947277442e-05,
+      "loss": 0.2998,
+      "step": 4034
+    },
+    {
+      "epoch": 20.889032258064518,
+      "grad_norm": 0.06271643191576004,
+      "learning_rate": 6.067415730337079e-05,
+      "loss": 0.2808,
+      "step": 4035
+    },
+    {
+      "epoch": 20.894193548387097,
+      "grad_norm": 0.17057666182518005,
+      "learning_rate": 6.063958513396716e-05,
+      "loss": 0.2654,
+      "step": 4036
+    },
+    {
+      "epoch": 20.899354838709677,
+      "grad_norm": 0.0959225594997406,
+      "learning_rate": 6.060501296456353e-05,
+      "loss": 0.2994,
+      "step": 4037
+    },
+    {
+      "epoch": 20.90451612903226,
+      "grad_norm": 0.12191182374954224,
+      "learning_rate": 6.05704407951599e-05,
+      "loss": 0.2927,
+      "step": 4038
+    },
+    {
+      "epoch": 20.90967741935484,
+      "grad_norm": 0.22743819653987885,
+      "learning_rate": 6.0535868625756266e-05,
+      "loss": 0.2974,
+      "step": 4039
+    },
+    {
+      "epoch": 20.91483870967742,
+      "grad_norm": 0.2247859239578247,
+      "learning_rate": 6.050129645635264e-05,
+      "loss": 0.3035,
+      "step": 4040
+    },
+    {
+      "epoch": 20.92,
+      "grad_norm": 0.045981265604496,
+      "learning_rate": 6.0466724286949007e-05,
+      "loss": 0.2952,
+      "step": 4041
+    },
+    {
+      "epoch": 20.92516129032258,
+      "grad_norm": 0.17560067772865295,
+      "learning_rate": 6.043215211754538e-05,
+      "loss": 0.254,
+      "step": 4042
+    },
+    {
+      "epoch": 20.93032258064516,
+      "grad_norm": 0.0923171192407608,
+      "learning_rate": 6.039757994814175e-05,
+      "loss": 0.2921,
+      "step": 4043
+    },
+    {
+      "epoch": 20.93548387096774,
+      "grad_norm": 0.15123817324638367,
+      "learning_rate": 6.036300777873812e-05,
+      "loss": 0.2754,
+      "step": 4044
+    },
+    {
+      "epoch": 20.940645161290323,
+      "grad_norm": 0.03852042928338051,
+      "learning_rate": 6.032843560933449e-05,
+      "loss": 0.3001,
+      "step": 4045
+    },
+    {
+      "epoch": 20.945806451612903,
+      "grad_norm": 0.10583159327507019,
+      "learning_rate": 6.0293863439930855e-05,
+      "loss": 0.2943,
+      "step": 4046
+    },
+    {
+      "epoch": 20.950967741935482,
+      "grad_norm": 0.06219182536005974,
+      "learning_rate": 6.025929127052723e-05,
+      "loss": 0.3011,
+      "step": 4047
+    },
+    {
+      "epoch": 20.956129032258065,
+      "grad_norm": 0.19903752207756042,
+      "learning_rate": 6.0224719101123596e-05,
+      "loss": 0.2936,
+      "step": 4048
+    },
+    {
+      "epoch": 20.961290322580645,
+      "grad_norm": 0.06007864698767662,
+      "learning_rate": 6.019014693171997e-05,
+      "loss": 0.2506,
+      "step": 4049
+    },
+    {
+      "epoch": 20.966451612903224,
+      "grad_norm": 0.10773148387670517,
+      "learning_rate": 6.0155574762316337e-05,
+      "loss": 0.2947,
+      "step": 4050
+    },
+    {
+      "epoch": 20.971612903225807,
+      "grad_norm": 0.03945239633321762,
+      "learning_rate": 6.012100259291271e-05,
+      "loss": 0.2769,
+      "step": 4051
+    },
+    {
+      "epoch": 20.976774193548387,
+      "grad_norm": 0.15047156810760498,
+      "learning_rate": 6.008643042350908e-05,
+      "loss": 0.2678,
+      "step": 4052
+    },
+    {
+      "epoch": 20.981935483870966,
+      "grad_norm": 0.17246714234352112,
+      "learning_rate": 6.005185825410545e-05,
+      "loss": 0.2873,
+      "step": 4053
+    },
+    {
+      "epoch": 20.98709677419355,
+      "grad_norm": 0.07101119309663773,
+      "learning_rate": 6.001728608470182e-05,
+      "loss": 0.3021,
+      "step": 4054
+    },
+    {
+      "epoch": 20.99225806451613,
+      "grad_norm": 0.08763767033815384,
+      "learning_rate": 5.9982713915298185e-05,
+      "loss": 0.2827,
+      "step": 4055
+    },
+    {
+      "epoch": 20.99741935483871,
+      "grad_norm": 0.12635627388954163,
+      "learning_rate": 5.994814174589456e-05,
+      "loss": 0.3878,
+      "step": 4056
+    },
+    {
+      "epoch": 21.003870967741936,
+      "grad_norm": 0.06386898458003998,
+      "learning_rate": 5.9913569576490926e-05,
+      "loss": 0.3474,
+      "step": 4057
+    },
+    {
+      "epoch": 21.009032258064515,
+      "grad_norm": 0.0404454842209816,
+      "learning_rate": 5.98789974070873e-05,
+      "loss": 0.2704,
+      "step": 4058
+    },
+    {
+      "epoch": 21.014193548387098,
+      "grad_norm": 0.04145240783691406,
+      "learning_rate": 5.9844425237683667e-05,
+      "loss": 0.2978,
+      "step": 4059
+    },
+    {
+      "epoch": 21.019354838709678,
+      "grad_norm": 0.054107122123241425,
+      "learning_rate": 5.980985306828004e-05,
+      "loss": 0.2809,
+      "step": 4060
+    },
+    {
+      "epoch": 21.019354838709678,
+      "eval_loss": 3.075632333755493,
+      "eval_runtime": 21.2118,
+      "eval_samples_per_second": 3.724,
+      "eval_steps_per_second": 0.471,
+      "step": 4060
+    },
+    {
+      "epoch": 21.024516129032257,
+      "grad_norm": 0.060415517538785934,
+      "learning_rate": 5.977528089887641e-05,
+      "loss": 0.2694,
+      "step": 4061
+    },
+    {
+      "epoch": 21.02967741935484,
+      "grad_norm": 0.029110316187143326,
+      "learning_rate": 5.9740708729472774e-05,
+      "loss": 0.2805,
+      "step": 4062
+    },
+    {
+      "epoch": 21.03483870967742,
+      "grad_norm": 0.0249349232763052,
+      "learning_rate": 5.970613656006915e-05,
+      "loss": 0.2619,
+      "step": 4063
+    },
+    {
+      "epoch": 21.04,
+      "grad_norm": 0.09865783154964447,
+      "learning_rate": 5.9671564390665515e-05,
+      "loss": 0.3049,
+      "step": 4064
+    },
+    {
+      "epoch": 21.045161290322582,
+      "grad_norm": 0.07334308326244354,
+      "learning_rate": 5.963699222126189e-05,
+      "loss": 0.3041,
+      "step": 4065
+    },
+    {
+      "epoch": 21.05032258064516,
+      "grad_norm": 0.030111489817500114,
+      "learning_rate": 5.9602420051858256e-05,
+      "loss": 0.3029,
+      "step": 4066
+    },
+    {
+      "epoch": 21.05548387096774,
+      "grad_norm": 0.13574545085430145,
+      "learning_rate": 5.956784788245463e-05,
+      "loss": 0.2893,
+      "step": 4067
+    },
+    {
+      "epoch": 21.060645161290324,
+      "grad_norm": 0.02990131638944149,
+      "learning_rate": 5.9533275713050996e-05,
+      "loss": 0.2793,
+      "step": 4068
+    },
+    {
+      "epoch": 21.065806451612904,
+      "grad_norm": 0.04661112651228905,
+      "learning_rate": 5.949870354364737e-05,
+      "loss": 0.2931,
+      "step": 4069
+    },
+    {
+      "epoch": 21.070967741935483,
+      "grad_norm": 0.05501263961195946,
+      "learning_rate": 5.946413137424374e-05,
+      "loss": 0.2895,
+      "step": 4070
+    },
+    {
+      "epoch": 21.076129032258063,
+      "grad_norm": 0.04723319783806801,
+      "learning_rate": 5.9429559204840104e-05,
+      "loss": 0.2892,
+      "step": 4071
+    },
+    {
+      "epoch": 21.081290322580646,
+      "grad_norm": 0.028639281168580055,
+      "learning_rate": 5.939498703543648e-05,
+      "loss": 0.2594,
+      "step": 4072
+    },
+    {
+      "epoch": 21.086451612903225,
+      "grad_norm": 0.034408025443553925,
+      "learning_rate": 5.9360414866032845e-05,
+      "loss": 0.2732,
+      "step": 4073
+    },
+    {
+      "epoch": 21.091612903225805,
+      "grad_norm": 0.04973653703927994,
+      "learning_rate": 5.932584269662922e-05,
+      "loss": 0.2497,
+      "step": 4074
+    },
+    {
+      "epoch": 21.096774193548388,
+      "grad_norm": 0.03349469602108002,
+      "learning_rate": 5.9291270527225586e-05,
+      "loss": 0.2908,
+      "step": 4075
+    },
+    {
+      "epoch": 21.101935483870967,
+      "grad_norm": 0.15707750618457794,
+      "learning_rate": 5.925669835782196e-05,
+      "loss": 0.2835,
+      "step": 4076
+    },
+    {
+      "epoch": 21.107096774193547,
+      "grad_norm": 0.02164219506084919,
+      "learning_rate": 5.9222126188418326e-05,
+      "loss": 0.2928,
+      "step": 4077
+    },
+    {
+      "epoch": 21.11225806451613,
+      "grad_norm": 0.03439301997423172,
+      "learning_rate": 5.91875540190147e-05,
+      "loss": 0.293,
+      "step": 4078
+    },
+    {
+      "epoch": 21.11741935483871,
+      "grad_norm": 0.027028966695070267,
+      "learning_rate": 5.915298184961107e-05,
+      "loss": 0.2669,
+      "step": 4079
+    },
+    {
+      "epoch": 21.12258064516129,
+      "grad_norm": 0.036328986287117004,
+      "learning_rate": 5.9118409680207434e-05,
+      "loss": 0.2837,
+      "step": 4080
+    },
+    {
+      "epoch": 21.127741935483872,
+      "grad_norm": 0.11336719244718552,
+      "learning_rate": 5.908383751080381e-05,
+      "loss": 0.2863,
+      "step": 4081
+    },
+    {
+      "epoch": 21.13290322580645,
+      "grad_norm": 0.031607698649168015,
+      "learning_rate": 5.9049265341400175e-05,
+      "loss": 0.3086,
+      "step": 4082
+    },
+    {
+      "epoch": 21.13806451612903,
+      "grad_norm": 0.03750959038734436,
+      "learning_rate": 5.901469317199655e-05,
+      "loss": 0.2167,
+      "step": 4083
+    },
+    {
+      "epoch": 21.143225806451614,
+      "grad_norm": 0.06942273676395416,
+      "learning_rate": 5.8980121002592916e-05,
+      "loss": 0.3019,
+      "step": 4084
+    },
+    {
+      "epoch": 21.148387096774194,
+      "grad_norm": 0.09442391246557236,
+      "learning_rate": 5.894554883318929e-05,
+      "loss": 0.286,
+      "step": 4085
+    },
+    {
+      "epoch": 21.153548387096773,
+      "grad_norm": 0.03921661153435707,
+      "learning_rate": 5.8910976663785656e-05,
+      "loss": 0.3056,
+      "step": 4086
+    },
+    {
+      "epoch": 21.158709677419356,
+      "grad_norm": 0.1258828043937683,
+      "learning_rate": 5.8876404494382023e-05,
+      "loss": 0.2655,
+      "step": 4087
+    },
+    {
+      "epoch": 21.163870967741936,
+      "grad_norm": 0.06785678118467331,
+      "learning_rate": 5.88418323249784e-05,
+      "loss": 0.2696,
+      "step": 4088
+    },
+    {
+      "epoch": 21.169032258064515,
+      "grad_norm": 0.0449206717312336,
+      "learning_rate": 5.8807260155574764e-05,
+      "loss": 0.2841,
+      "step": 4089
+    },
+    {
+      "epoch": 21.174193548387098,
+      "grad_norm": 0.06908271461725235,
+      "learning_rate": 5.877268798617114e-05,
+      "loss": 0.2854,
+      "step": 4090
+    },
+    {
+      "epoch": 21.179354838709678,
+      "grad_norm": 0.03606576845049858,
+      "learning_rate": 5.8738115816767505e-05,
+      "loss": 0.2823,
+      "step": 4091
+    },
+    {
+      "epoch": 21.184516129032257,
+      "grad_norm": 0.03284405171871185,
+      "learning_rate": 5.870354364736388e-05,
+      "loss": 0.2586,
+      "step": 4092
+    },
+    {
+      "epoch": 21.18967741935484,
+      "grad_norm": 0.026450350880622864,
+      "learning_rate": 5.8668971477960246e-05,
+      "loss": 0.3008,
+      "step": 4093
+    },
+    {
+      "epoch": 21.19483870967742,
+      "grad_norm": 0.0715409517288208,
+      "learning_rate": 5.863439930855662e-05,
+      "loss": 0.2701,
+      "step": 4094
+    },
+    {
+      "epoch": 21.2,
+      "grad_norm": 0.0960545614361763,
+      "learning_rate": 5.8599827139152986e-05,
+      "loss": 0.2972,
+      "step": 4095
+    },
+    {
+      "epoch": 21.205161290322582,
+      "grad_norm": 0.027156401425600052,
+      "learning_rate": 5.856525496974935e-05,
+      "loss": 0.2913,
+      "step": 4096
+    },
+    {
+      "epoch": 21.210322580645162,
+      "grad_norm": 0.2087082713842392,
+      "learning_rate": 5.853068280034573e-05,
+      "loss": 0.2899,
+      "step": 4097
+    },
+    {
+      "epoch": 21.21548387096774,
+      "grad_norm": 0.12557059526443481,
+      "learning_rate": 5.8496110630942094e-05,
+      "loss": 0.2952,
+      "step": 4098
+    },
+    {
+      "epoch": 21.22064516129032,
+      "grad_norm": 0.048750657588243484,
+      "learning_rate": 5.846153846153847e-05,
+      "loss": 0.3101,
+      "step": 4099
+    },
+    {
+      "epoch": 21.225806451612904,
+      "grad_norm": 0.06964153051376343,
+      "learning_rate": 5.8426966292134835e-05,
+      "loss": 0.3065,
+      "step": 4100
+    },
+    {
+      "epoch": 21.230967741935483,
+      "grad_norm": 0.1326199471950531,
+      "learning_rate": 5.839239412273121e-05,
+      "loss": 0.2908,
+      "step": 4101
+    },
+    {
+      "epoch": 21.236129032258063,
+      "grad_norm": 0.03145857900381088,
+      "learning_rate": 5.8357821953327576e-05,
+      "loss": 0.2874,
+      "step": 4102
+    },
+    {
+      "epoch": 21.241290322580646,
+      "grad_norm": 0.0791567787528038,
+      "learning_rate": 5.8323249783923936e-05,
+      "loss": 0.2941,
+      "step": 4103
+    },
+    {
+      "epoch": 21.246451612903225,
+      "grad_norm": 0.028354141861200333,
+      "learning_rate": 5.8288677614520316e-05,
+      "loss": 0.273,
+      "step": 4104
+    },
+    {
+      "epoch": 21.251612903225805,
+      "grad_norm": 0.1903354674577713,
+      "learning_rate": 5.825410544511668e-05,
+      "loss": 0.2833,
+      "step": 4105
+    },
+    {
+      "epoch": 21.256774193548388,
+      "grad_norm": 0.030410494655370712,
+      "learning_rate": 5.821953327571306e-05,
+      "loss": 0.2658,
+      "step": 4106
+    },
+    {
+      "epoch": 21.261935483870968,
+      "grad_norm": 0.038786955177783966,
+      "learning_rate": 5.8184961106309424e-05,
+      "loss": 0.3016,
+      "step": 4107
+    },
+    {
+      "epoch": 21.267096774193547,
+      "grad_norm": 0.026577137410640717,
+      "learning_rate": 5.81503889369058e-05,
+      "loss": 0.304,
+      "step": 4108
+    },
+    {
+      "epoch": 21.27225806451613,
+      "grad_norm": 0.12804940342903137,
+      "learning_rate": 5.8115816767502165e-05,
+      "loss": 0.292,
+      "step": 4109
+    },
+    {
+      "epoch": 21.27741935483871,
+      "grad_norm": 0.12290894240140915,
+      "learning_rate": 5.808124459809854e-05,
+      "loss": 0.2818,
+      "step": 4110
+    },
+    {
+      "epoch": 21.28258064516129,
+      "grad_norm": 0.023023558780550957,
+      "learning_rate": 5.8046672428694906e-05,
+      "loss": 0.2894,
+      "step": 4111
+    },
+    {
+      "epoch": 21.287741935483872,
+      "grad_norm": 0.0317746102809906,
+      "learning_rate": 5.8012100259291266e-05,
+      "loss": 0.2428,
+      "step": 4112
+    },
+    {
+      "epoch": 21.29290322580645,
+      "grad_norm": 0.15513020753860474,
+      "learning_rate": 5.7977528089887646e-05,
+      "loss": 0.2932,
+      "step": 4113
+    },
+    {
+      "epoch": 21.29806451612903,
+      "grad_norm": 0.029827402904629707,
+      "learning_rate": 5.794295592048401e-05,
+      "loss": 0.2323,
+      "step": 4114
+    },
+    {
+      "epoch": 21.303225806451614,
+      "grad_norm": 0.0452035628259182,
+      "learning_rate": 5.790838375108039e-05,
+      "loss": 0.2536,
+      "step": 4115
+    },
+    {
+      "epoch": 21.308387096774194,
+      "grad_norm": 0.11305708438158035,
+      "learning_rate": 5.7873811581676754e-05,
+      "loss": 0.281,
+      "step": 4116
+    },
+    {
+      "epoch": 21.313548387096773,
+      "grad_norm": 0.11131157726049423,
+      "learning_rate": 5.783923941227313e-05,
+      "loss": 0.2612,
+      "step": 4117
+    },
+    {
+      "epoch": 21.318709677419356,
+      "grad_norm": 0.03678593412041664,
+      "learning_rate": 5.7804667242869495e-05,
+      "loss": 0.2688,
+      "step": 4118
+    },
+    {
+      "epoch": 21.323870967741936,
+      "grad_norm": 0.10723850876092911,
+      "learning_rate": 5.7770095073465855e-05,
+      "loss": 0.2713,
+      "step": 4119
+    },
+    {
+      "epoch": 21.329032258064515,
+      "grad_norm": 0.13832104206085205,
+      "learning_rate": 5.7735522904062236e-05,
+      "loss": 0.3051,
+      "step": 4120
+    },
+    {
+      "epoch": 21.3341935483871,
+      "grad_norm": 0.03026687167584896,
+      "learning_rate": 5.7700950734658596e-05,
+      "loss": 0.28,
+      "step": 4121
+    },
+    {
+      "epoch": 21.339354838709678,
+      "grad_norm": 0.04471990838646889,
+      "learning_rate": 5.7666378565254976e-05,
+      "loss": 0.2997,
+      "step": 4122
+    },
+    {
+      "epoch": 21.344516129032257,
+      "grad_norm": 0.02790313959121704,
+      "learning_rate": 5.7631806395851337e-05,
+      "loss": 0.2964,
+      "step": 4123
+    },
+    {
+      "epoch": 21.34967741935484,
+      "grad_norm": 0.08246771991252899,
+      "learning_rate": 5.759723422644772e-05,
+      "loss": 0.2569,
+      "step": 4124
+    },
+    {
+      "epoch": 21.35483870967742,
+      "grad_norm": 0.02107340842485428,
+      "learning_rate": 5.7562662057044084e-05,
+      "loss": 0.2925,
+      "step": 4125
+    },
+    {
+      "epoch": 21.36,
+      "grad_norm": 0.03764757141470909,
+      "learning_rate": 5.752808988764046e-05,
+      "loss": 0.2803,
+      "step": 4126
+    },
+    {
+      "epoch": 21.36516129032258,
+      "grad_norm": 0.04183781519532204,
+      "learning_rate": 5.7493517718236825e-05,
+      "loss": 0.3077,
+      "step": 4127
+    },
+    {
+      "epoch": 21.370322580645162,
+      "grad_norm": 0.027907099574804306,
+      "learning_rate": 5.7458945548833185e-05,
+      "loss": 0.2621,
+      "step": 4128
+    },
+    {
+      "epoch": 21.37548387096774,
+      "grad_norm": 0.05995682254433632,
+      "learning_rate": 5.7424373379429566e-05,
+      "loss": 0.3001,
+      "step": 4129
+    },
+    {
+      "epoch": 21.38064516129032,
+      "grad_norm": 0.04082728549838066,
+      "learning_rate": 5.7389801210025926e-05,
+      "loss": 0.2699,
+      "step": 4130
+    },
+    {
+      "epoch": 21.385806451612904,
+      "grad_norm": 0.04719356447458267,
+      "learning_rate": 5.7355229040622306e-05,
+      "loss": 0.2789,
+      "step": 4131
+    },
+    {
+      "epoch": 21.390967741935484,
+      "grad_norm": 0.05296272784471512,
+      "learning_rate": 5.7320656871218666e-05,
+      "loss": 0.2843,
+      "step": 4132
+    },
+    {
+      "epoch": 21.396129032258063,
+      "grad_norm": 0.033614497631788254,
+      "learning_rate": 5.728608470181505e-05,
+      "loss": 0.2982,
+      "step": 4133
+    },
+    {
+      "epoch": 21.401290322580646,
+      "grad_norm": 0.032487183809280396,
+      "learning_rate": 5.7251512532411414e-05,
+      "loss": 0.2719,
+      "step": 4134
+    },
+    {
+      "epoch": 21.406451612903226,
+      "grad_norm": 0.0312189981341362,
+      "learning_rate": 5.7216940363007774e-05,
+      "loss": 0.2719,
+      "step": 4135
+    },
+    {
+      "epoch": 21.411612903225805,
+      "grad_norm": 0.03727870061993599,
+      "learning_rate": 5.7182368193604155e-05,
+      "loss": 0.2815,
+      "step": 4136
+    },
+    {
+      "epoch": 21.416774193548388,
+      "grad_norm": 0.07301236689090729,
+      "learning_rate": 5.7147796024200515e-05,
+      "loss": 0.2869,
+      "step": 4137
+    },
+    {
+      "epoch": 21.421935483870968,
+      "grad_norm": 0.039962250739336014,
+      "learning_rate": 5.7113223854796896e-05,
+      "loss": 0.2955,
+      "step": 4138
+    },
+    {
+      "epoch": 21.427096774193547,
+      "grad_norm": 0.1801455318927765,
+      "learning_rate": 5.7078651685393256e-05,
+      "loss": 0.2937,
+      "step": 4139
+    },
+    {
+      "epoch": 21.43225806451613,
+      "grad_norm": 0.08347830921411514,
+      "learning_rate": 5.7044079515989636e-05,
+      "loss": 0.2809,
+      "step": 4140
+    },
+    {
+      "epoch": 21.43741935483871,
+      "grad_norm": 0.028520913794636726,
+      "learning_rate": 5.7009507346585996e-05,
+      "loss": 0.2958,
+      "step": 4141
+    },
+    {
+      "epoch": 21.44258064516129,
+      "grad_norm": 0.037673596292734146,
+      "learning_rate": 5.697493517718238e-05,
+      "loss": 0.2787,
+      "step": 4142
+    },
+    {
+      "epoch": 21.447741935483872,
+      "grad_norm": 0.05031321570277214,
+      "learning_rate": 5.6940363007778744e-05,
+      "loss": 0.2754,
+      "step": 4143
+    },
+    {
+      "epoch": 21.452903225806452,
+      "grad_norm": 0.060914162546396255,
+      "learning_rate": 5.6905790838375104e-05,
+      "loss": 0.3054,
+      "step": 4144
+    },
+    {
+      "epoch": 21.45806451612903,
+      "grad_norm": 0.08684848248958588,
+      "learning_rate": 5.6871218668971485e-05,
+      "loss": 0.2851,
+      "step": 4145
+    },
+    {
+      "epoch": 21.463225806451614,
+      "grad_norm": 0.03514571115374565,
+      "learning_rate": 5.6836646499567845e-05,
+      "loss": 0.2942,
+      "step": 4146
+    },
+    {
+      "epoch": 21.468387096774194,
+      "grad_norm": 0.058252446353435516,
+      "learning_rate": 5.6802074330164225e-05,
+      "loss": 0.2849,
+      "step": 4147
+    },
+    {
+      "epoch": 21.473548387096773,
+      "grad_norm": 0.1479421705007553,
+      "learning_rate": 5.6767502160760586e-05,
+      "loss": 0.2438,
+      "step": 4148
+    },
+    {
+      "epoch": 21.478709677419356,
+      "grad_norm": 0.03266794979572296,
+      "learning_rate": 5.6732929991356966e-05,
+      "loss": 0.2925,
+      "step": 4149
+    },
+    {
+      "epoch": 21.483870967741936,
+      "grad_norm": 0.07726222276687622,
+      "learning_rate": 5.6698357821953326e-05,
+      "loss": 0.3121,
+      "step": 4150
+    },
+    {
+      "epoch": 21.489032258064515,
+      "grad_norm": 0.031776171177625656,
+      "learning_rate": 5.666378565254971e-05,
+      "loss": 0.2616,
+      "step": 4151
+    },
+    {
+      "epoch": 21.4941935483871,
+      "grad_norm": 0.046746626496315,
+      "learning_rate": 5.6629213483146074e-05,
+      "loss": 0.2827,
+      "step": 4152
+    },
+    {
+      "epoch": 21.499354838709678,
+      "grad_norm": 0.029332991689443588,
+      "learning_rate": 5.6594641313742434e-05,
+      "loss": 0.294,
+      "step": 4153
+    },
+    {
+      "epoch": 21.504516129032258,
+      "grad_norm": 0.1220678985118866,
+      "learning_rate": 5.6560069144338815e-05,
+      "loss": 0.2537,
+      "step": 4154
+    },
+    {
+      "epoch": 21.509677419354837,
+      "grad_norm": 0.028324609622359276,
+      "learning_rate": 5.6525496974935175e-05,
+      "loss": 0.2852,
+      "step": 4155
+    },
+    {
+      "epoch": 21.51483870967742,
+      "grad_norm": 0.03936924785375595,
+      "learning_rate": 5.6490924805531555e-05,
+      "loss": 0.2767,
+      "step": 4156
+    },
+    {
+      "epoch": 21.52,
+      "grad_norm": 0.08785068988800049,
+      "learning_rate": 5.6456352636127916e-05,
+      "loss": 0.2724,
+      "step": 4157
+    },
+    {
+      "epoch": 21.52516129032258,
+      "grad_norm": 0.11234081536531448,
+      "learning_rate": 5.6421780466724296e-05,
+      "loss": 0.2852,
+      "step": 4158
+    },
+    {
+      "epoch": 21.530322580645162,
+      "grad_norm": 0.05394039675593376,
+      "learning_rate": 5.6387208297320656e-05,
+      "loss": 0.2989,
+      "step": 4159
+    },
+    {
+      "epoch": 21.53548387096774,
+      "grad_norm": 0.024116668850183487,
+      "learning_rate": 5.635263612791702e-05,
+      "loss": 0.2915,
+      "step": 4160
+    },
+    {
+      "epoch": 21.54064516129032,
+      "grad_norm": 0.030164502561092377,
+      "learning_rate": 5.6318063958513404e-05,
+      "loss": 0.294,
+      "step": 4161
+    },
+    {
+      "epoch": 21.545806451612904,
+      "grad_norm": 0.02736390382051468,
+      "learning_rate": 5.6283491789109764e-05,
+      "loss": 0.2857,
+      "step": 4162
+    },
+    {
+      "epoch": 21.550967741935484,
+      "grad_norm": 0.11054631322622299,
+      "learning_rate": 5.6248919619706145e-05,
+      "loss": 0.2754,
+      "step": 4163
+    },
+    {
+      "epoch": 21.556129032258063,
+      "grad_norm": 0.059282172471284866,
+      "learning_rate": 5.6214347450302505e-05,
+      "loss": 0.2909,
+      "step": 4164
+    },
+    {
+      "epoch": 21.561290322580646,
+      "grad_norm": 0.06375820189714432,
+      "learning_rate": 5.6179775280898885e-05,
+      "loss": 0.3005,
+      "step": 4165
+    },
+    {
+      "epoch": 21.566451612903226,
+      "grad_norm": 0.07598790526390076,
+      "learning_rate": 5.6145203111495246e-05,
+      "loss": 0.2894,
+      "step": 4166
+    },
+    {
+      "epoch": 21.571612903225805,
+      "grad_norm": 0.029939178377389908,
+      "learning_rate": 5.6110630942091626e-05,
+      "loss": 0.2885,
+      "step": 4167
+    },
+    {
+      "epoch": 21.57677419354839,
+      "grad_norm": 0.032854560762643814,
+      "learning_rate": 5.6076058772687986e-05,
+      "loss": 0.2483,
+      "step": 4168
+    },
+    {
+      "epoch": 21.581935483870968,
+      "grad_norm": 0.023499110713601112,
+      "learning_rate": 5.604148660328435e-05,
+      "loss": 0.293,
+      "step": 4169
+    },
+    {
+      "epoch": 21.587096774193547,
+      "grad_norm": 0.032280899584293365,
+      "learning_rate": 5.6006914433880734e-05,
+      "loss": 0.2675,
+      "step": 4170
+    },
+    {
+      "epoch": 21.59225806451613,
+      "grad_norm": 0.2969343066215515,
+      "learning_rate": 5.5972342264477094e-05,
+      "loss": 0.2801,
+      "step": 4171
+    },
+    {
+      "epoch": 21.59741935483871,
+      "grad_norm": 0.13323049247264862,
+      "learning_rate": 5.5937770095073475e-05,
+      "loss": 0.2798,
+      "step": 4172
+    },
+    {
+      "epoch": 21.60258064516129,
+      "grad_norm": 0.10714597254991531,
+      "learning_rate": 5.5903197925669835e-05,
+      "loss": 0.2728,
+      "step": 4173
+    },
+    {
+      "epoch": 21.607741935483872,
+      "grad_norm": 0.03778146579861641,
+      "learning_rate": 5.5868625756266215e-05,
+      "loss": 0.2935,
+      "step": 4174
+    },
+    {
+      "epoch": 21.612903225806452,
+      "grad_norm": 0.11497175693511963,
+      "learning_rate": 5.5834053586862576e-05,
+      "loss": 0.3134,
+      "step": 4175
+    },
+    {
+      "epoch": 21.61806451612903,
+      "grad_norm": 0.027278726920485497,
+      "learning_rate": 5.579948141745894e-05,
+      "loss": 0.2965,
+      "step": 4176
+    },
+    {
+      "epoch": 21.623225806451615,
+      "grad_norm": 0.15559861063957214,
+      "learning_rate": 5.5764909248055316e-05,
+      "loss": 0.2988,
+      "step": 4177
+    },
+    {
+      "epoch": 21.628387096774194,
+      "grad_norm": 0.07007851451635361,
+      "learning_rate": 5.573033707865168e-05,
+      "loss": 0.2839,
+      "step": 4178
+    },
+    {
+      "epoch": 21.633548387096774,
+      "grad_norm": 0.03244547173380852,
+      "learning_rate": 5.569576490924806e-05,
+      "loss": 0.2873,
+      "step": 4179
+    },
+    {
+      "epoch": 21.638709677419357,
+      "grad_norm": 0.0363052636384964,
+      "learning_rate": 5.5661192739844424e-05,
+      "loss": 0.2935,
+      "step": 4180
+    },
+    {
+      "epoch": 21.643870967741936,
+      "grad_norm": 0.050539255142211914,
+      "learning_rate": 5.5626620570440805e-05,
+      "loss": 0.2667,
+      "step": 4181
+    },
+    {
+      "epoch": 21.649032258064516,
+      "grad_norm": 0.037443432956933975,
+      "learning_rate": 5.5592048401037165e-05,
+      "loss": 0.2537,
+      "step": 4182
+    },
+    {
+      "epoch": 21.654193548387095,
+      "grad_norm": 0.03377171978354454,
+      "learning_rate": 5.5557476231633545e-05,
+      "loss": 0.2721,
+      "step": 4183
+    },
+    {
+      "epoch": 21.659354838709678,
+      "grad_norm": 0.05607049539685249,
+      "learning_rate": 5.5522904062229906e-05,
+      "loss": 0.2827,
+      "step": 4184
+    },
+    {
+      "epoch": 21.664516129032258,
+      "grad_norm": 0.03358304873108864,
+      "learning_rate": 5.548833189282627e-05,
+      "loss": 0.3147,
+      "step": 4185
+    },
+    {
+      "epoch": 21.669677419354837,
+      "grad_norm": 0.09935665875673294,
+      "learning_rate": 5.5453759723422646e-05,
+      "loss": 0.3008,
+      "step": 4186
+    },
+    {
+      "epoch": 21.67483870967742,
+      "grad_norm": 0.030461855232715607,
+      "learning_rate": 5.541918755401901e-05,
+      "loss": 0.2656,
+      "step": 4187
+    },
+    {
+      "epoch": 21.68,
+      "grad_norm": 0.04648221284151077,
+      "learning_rate": 5.538461538461539e-05,
+      "loss": 0.2823,
+      "step": 4188
+    },
+    {
+      "epoch": 21.68516129032258,
+      "grad_norm": 0.10592552274465561,
+      "learning_rate": 5.5350043215211754e-05,
+      "loss": 0.2837,
+      "step": 4189
+    },
+    {
+      "epoch": 21.690322580645162,
+      "grad_norm": 0.06887082010507584,
+      "learning_rate": 5.5315471045808135e-05,
+      "loss": 0.2926,
+      "step": 4190
+    },
+    {
+      "epoch": 21.695483870967742,
+      "grad_norm": 0.027295559644699097,
+      "learning_rate": 5.5280898876404495e-05,
+      "loss": 0.2949,
+      "step": 4191
+    },
+    {
+      "epoch": 21.70064516129032,
+      "grad_norm": 0.03161584213376045,
+      "learning_rate": 5.524632670700086e-05,
+      "loss": 0.3119,
+      "step": 4192
+    },
+    {
+      "epoch": 21.705806451612904,
+      "grad_norm": 0.11891087144613266,
+      "learning_rate": 5.5211754537597236e-05,
+      "loss": 0.3018,
+      "step": 4193
+    },
+    {
+      "epoch": 21.710967741935484,
+      "grad_norm": 0.14185966551303864,
+      "learning_rate": 5.51771823681936e-05,
+      "loss": 0.3034,
+      "step": 4194
+    },
+    {
+      "epoch": 21.716129032258063,
+      "grad_norm": 0.05896966531872749,
+      "learning_rate": 5.5142610198789976e-05,
+      "loss": 0.2882,
+      "step": 4195
+    },
+    {
+      "epoch": 21.721290322580646,
+      "grad_norm": 0.025536222383379936,
+      "learning_rate": 5.510803802938634e-05,
+      "loss": 0.2955,
+      "step": 4196
+    },
+    {
+      "epoch": 21.726451612903226,
+      "grad_norm": 0.023713937029242516,
+      "learning_rate": 5.507346585998272e-05,
+      "loss": 0.2627,
+      "step": 4197
+    },
+    {
+      "epoch": 21.731612903225805,
+      "grad_norm": 0.07392848283052444,
+      "learning_rate": 5.5038893690579084e-05,
+      "loss": 0.3096,
+      "step": 4198
+    },
+    {
+      "epoch": 21.73677419354839,
+      "grad_norm": 0.08494075387716293,
+      "learning_rate": 5.5004321521175465e-05,
+      "loss": 0.2844,
+      "step": 4199
+    },
+    {
+      "epoch": 21.741935483870968,
+      "grad_norm": 0.10398995131254196,
+      "learning_rate": 5.4969749351771825e-05,
+      "loss": 0.25,
+      "step": 4200
+    },
+    {
+      "epoch": 21.747096774193547,
+      "grad_norm": 0.12182550132274628,
+      "learning_rate": 5.493517718236819e-05,
+      "loss": 0.2882,
+      "step": 4201
+    },
+    {
+      "epoch": 21.75225806451613,
+      "grad_norm": 0.040169958025217056,
+      "learning_rate": 5.4900605012964566e-05,
+      "loss": 0.2666,
+      "step": 4202
+    },
+    {
+      "epoch": 21.75741935483871,
+      "grad_norm": 0.029618071392178535,
+      "learning_rate": 5.486603284356093e-05,
+      "loss": 0.3067,
+      "step": 4203
+    },
+    {
+      "epoch": 21.76258064516129,
+      "grad_norm": 0.036087751388549805,
+      "learning_rate": 5.4831460674157306e-05,
+      "loss": 0.2728,
+      "step": 4204
+    },
+    {
+      "epoch": 21.767741935483873,
+      "grad_norm": 0.02884778007864952,
+      "learning_rate": 5.479688850475367e-05,
+      "loss": 0.2815,
+      "step": 4205
+    },
+    {
+      "epoch": 21.772903225806452,
+      "grad_norm": 0.06687039136886597,
+      "learning_rate": 5.476231633535005e-05,
+      "loss": 0.3067,
+      "step": 4206
+    },
+    {
+      "epoch": 21.77806451612903,
+      "grad_norm": 0.08954150229692459,
+      "learning_rate": 5.4727744165946414e-05,
+      "loss": 0.264,
+      "step": 4207
+    },
+    {
+      "epoch": 21.78322580645161,
+      "grad_norm": 0.03196097910404205,
+      "learning_rate": 5.469317199654278e-05,
+      "loss": 0.3097,
+      "step": 4208
+    },
+    {
+      "epoch": 21.788387096774194,
+      "grad_norm": 0.02816467359662056,
+      "learning_rate": 5.4658599827139155e-05,
+      "loss": 0.2865,
+      "step": 4209
+    },
+    {
+      "epoch": 21.793548387096774,
+      "grad_norm": 0.038676030933856964,
+      "learning_rate": 5.462402765773552e-05,
+      "loss": 0.2707,
+      "step": 4210
+    },
+    {
+      "epoch": 21.798709677419353,
+      "grad_norm": 0.05813990533351898,
+      "learning_rate": 5.4589455488331895e-05,
+      "loss": 0.278,
+      "step": 4211
+    },
+    {
+      "epoch": 21.803870967741936,
+      "grad_norm": 0.11664920300245285,
+      "learning_rate": 5.455488331892826e-05,
+      "loss": 0.2665,
+      "step": 4212
+    },
+    {
+      "epoch": 21.809032258064516,
+      "grad_norm": 0.03278597071766853,
+      "learning_rate": 5.4520311149524636e-05,
+      "loss": 0.2862,
+      "step": 4213
+    },
+    {
+      "epoch": 21.814193548387095,
+      "grad_norm": 0.03193863481283188,
+      "learning_rate": 5.4485738980121e-05,
+      "loss": 0.3051,
+      "step": 4214
+    },
+    {
+      "epoch": 21.81935483870968,
+      "grad_norm": 0.0346556082367897,
+      "learning_rate": 5.445116681071738e-05,
+      "loss": 0.2776,
+      "step": 4215
+    },
+    {
+      "epoch": 21.824516129032258,
+      "grad_norm": 0.12117064744234085,
+      "learning_rate": 5.4416594641313744e-05,
+      "loss": 0.2716,
+      "step": 4216
+    },
+    {
+      "epoch": 21.829677419354837,
+      "grad_norm": 0.030956542119383812,
+      "learning_rate": 5.438202247191011e-05,
+      "loss": 0.3013,
+      "step": 4217
+    },
+    {
+      "epoch": 21.83483870967742,
+      "grad_norm": 0.030325327068567276,
+      "learning_rate": 5.4347450302506485e-05,
+      "loss": 0.262,
+      "step": 4218
+    },
+    {
+      "epoch": 21.84,
+      "grad_norm": 0.11939515918493271,
+      "learning_rate": 5.431287813310285e-05,
+      "loss": 0.2812,
+      "step": 4219
+    },
+    {
+      "epoch": 21.84516129032258,
+      "grad_norm": 0.03034338541328907,
+      "learning_rate": 5.4278305963699225e-05,
+      "loss": 0.2838,
+      "step": 4220
+    },
+    {
+      "epoch": 21.850322580645162,
+      "grad_norm": 0.16348332166671753,
+      "learning_rate": 5.424373379429559e-05,
+      "loss": 0.3058,
+      "step": 4221
+    },
+    {
+      "epoch": 21.855483870967742,
+      "grad_norm": 0.05401937663555145,
+      "learning_rate": 5.4209161624891966e-05,
+      "loss": 0.3117,
+      "step": 4222
+    },
+    {
+      "epoch": 21.86064516129032,
+      "grad_norm": 0.043188273906707764,
+      "learning_rate": 5.417458945548833e-05,
+      "loss": 0.2726,
+      "step": 4223
+    },
+    {
+      "epoch": 21.865806451612904,
+      "grad_norm": 0.0450209304690361,
+      "learning_rate": 5.414001728608471e-05,
+      "loss": 0.2686,
+      "step": 4224
+    },
+    {
+      "epoch": 21.870967741935484,
+      "grad_norm": 0.07262181490659714,
+      "learning_rate": 5.4105445116681074e-05,
+      "loss": 0.2773,
+      "step": 4225
+    },
+    {
+      "epoch": 21.876129032258063,
+      "grad_norm": 0.03629288822412491,
+      "learning_rate": 5.407087294727744e-05,
+      "loss": 0.2156,
+      "step": 4226
+    },
+    {
+      "epoch": 21.881290322580647,
+      "grad_norm": 0.18222935497760773,
+      "learning_rate": 5.4036300777873815e-05,
+      "loss": 0.3102,
+      "step": 4227
+    },
+    {
+      "epoch": 21.886451612903226,
+      "grad_norm": 0.02765553817152977,
+      "learning_rate": 5.400172860847018e-05,
+      "loss": 0.2748,
+      "step": 4228
+    },
+    {
+      "epoch": 21.891612903225806,
+      "grad_norm": 0.07979393750429153,
+      "learning_rate": 5.3967156439066555e-05,
+      "loss": 0.2935,
+      "step": 4229
+    },
+    {
+      "epoch": 21.89677419354839,
+      "grad_norm": 0.07759953290224075,
+      "learning_rate": 5.393258426966292e-05,
+      "loss": 0.2876,
+      "step": 4230
+    },
+    {
+      "epoch": 21.901935483870968,
+      "grad_norm": 0.0871763750910759,
+      "learning_rate": 5.3898012100259296e-05,
+      "loss": 0.2857,
+      "step": 4231
+    },
+    {
+      "epoch": 21.907096774193548,
+      "grad_norm": 0.02132992446422577,
+      "learning_rate": 5.386343993085566e-05,
+      "loss": 0.2571,
+      "step": 4232
+    },
+    {
+      "epoch": 21.91225806451613,
+      "grad_norm": 0.03740430623292923,
+      "learning_rate": 5.382886776145203e-05,
+      "loss": 0.2338,
+      "step": 4233
+    },
+    {
+      "epoch": 21.91741935483871,
+      "grad_norm": 0.024644866585731506,
+      "learning_rate": 5.3794295592048404e-05,
+      "loss": 0.2601,
+      "step": 4234
+    },
+    {
+      "epoch": 21.92258064516129,
+      "grad_norm": 0.05976215377449989,
+      "learning_rate": 5.375972342264477e-05,
+      "loss": 0.2953,
+      "step": 4235
+    },
+    {
+      "epoch": 21.927741935483873,
+      "grad_norm": 0.07059676945209503,
+      "learning_rate": 5.3725151253241145e-05,
+      "loss": 0.2741,
+      "step": 4236
+    },
+    {
+      "epoch": 21.932903225806452,
+      "grad_norm": 0.02502642385661602,
+      "learning_rate": 5.369057908383751e-05,
+      "loss": 0.2637,
+      "step": 4237
+    },
+    {
+      "epoch": 21.93806451612903,
+      "grad_norm": 0.04003347083926201,
+      "learning_rate": 5.3656006914433885e-05,
+      "loss": 0.2743,
+      "step": 4238
+    },
+    {
+      "epoch": 21.94322580645161,
+      "grad_norm": 0.037743523716926575,
+      "learning_rate": 5.362143474503025e-05,
+      "loss": 0.2837,
+      "step": 4239
+    },
+    {
+      "epoch": 21.948387096774194,
+      "grad_norm": 0.09548471868038177,
+      "learning_rate": 5.3586862575626626e-05,
+      "loss": 0.2998,
+      "step": 4240
+    },
+    {
+      "epoch": 21.953548387096774,
+      "grad_norm": 0.10069025307893753,
+      "learning_rate": 5.355229040622299e-05,
+      "loss": 0.2851,
+      "step": 4241
+    },
+    {
+      "epoch": 21.958709677419353,
+      "grad_norm": 0.040851082652807236,
+      "learning_rate": 5.351771823681936e-05,
+      "loss": 0.3005,
+      "step": 4242
+    },
+    {
+      "epoch": 21.963870967741936,
+      "grad_norm": 0.08858054876327515,
+      "learning_rate": 5.3483146067415734e-05,
+      "loss": 0.2729,
+      "step": 4243
+    },
+    {
+      "epoch": 21.969032258064516,
+      "grad_norm": 0.08921169489622116,
+      "learning_rate": 5.34485738980121e-05,
+      "loss": 0.2529,
+      "step": 4244
+    },
+    {
+      "epoch": 21.974193548387095,
+      "grad_norm": 0.04330304265022278,
+      "learning_rate": 5.3414001728608475e-05,
+      "loss": 0.2992,
+      "step": 4245
+    },
+    {
+      "epoch": 21.97935483870968,
+      "grad_norm": 0.07263077050447464,
+      "learning_rate": 5.337942955920484e-05,
+      "loss": 0.2752,
+      "step": 4246
+    },
+    {
+      "epoch": 21.984516129032258,
+      "grad_norm": 0.06787117570638657,
+      "learning_rate": 5.3344857389801215e-05,
+      "loss": 0.2279,
+      "step": 4247
+    },
+    {
+      "epoch": 21.989677419354837,
+      "grad_norm": 0.060145244002342224,
+      "learning_rate": 5.331028522039758e-05,
+      "loss": 0.2486,
+      "step": 4248
+    },
+    {
+      "epoch": 21.99483870967742,
+      "grad_norm": 0.035893961787223816,
+      "learning_rate": 5.327571305099395e-05,
+      "loss": 0.2714,
+      "step": 4249
+    },
+    {
+      "epoch": 22.001290322580644,
+      "grad_norm": 0.04512147605419159,
+      "learning_rate": 5.324114088159032e-05,
+      "loss": 0.3734,
+      "step": 4250
+    },
+    {
+      "epoch": 22.006451612903227,
+      "grad_norm": 0.025695808231830597,
+      "learning_rate": 5.320656871218669e-05,
+      "loss": 0.2995,
+      "step": 4251
+    },
+    {
+      "epoch": 22.011612903225807,
+      "grad_norm": 0.027173655107617378,
+      "learning_rate": 5.3171996542783064e-05,
+      "loss": 0.2634,
+      "step": 4252
+    },
+    {
+      "epoch": 22.016774193548386,
+      "grad_norm": 0.07203159481287003,
+      "learning_rate": 5.313742437337943e-05,
+      "loss": 0.2546,
+      "step": 4253
+    },
+    {
+      "epoch": 22.02193548387097,
+      "grad_norm": 0.025898169726133347,
+      "learning_rate": 5.3102852203975805e-05,
+      "loss": 0.2835,
+      "step": 4254
+    },
+    {
+      "epoch": 22.02709677419355,
+      "grad_norm": 0.04824727401137352,
+      "learning_rate": 5.306828003457217e-05,
+      "loss": 0.2781,
+      "step": 4255
+    },
+    {
+      "epoch": 22.032258064516128,
+      "grad_norm": 0.024010464549064636,
+      "learning_rate": 5.3033707865168545e-05,
+      "loss": 0.2572,
+      "step": 4256
+    },
+    {
+      "epoch": 22.03741935483871,
+      "grad_norm": 0.03480759263038635,
+      "learning_rate": 5.299913569576491e-05,
+      "loss": 0.3097,
+      "step": 4257
+    },
+    {
+      "epoch": 22.04258064516129,
+      "grad_norm": 0.022228162735700607,
+      "learning_rate": 5.296456352636128e-05,
+      "loss": 0.2634,
+      "step": 4258
+    },
+    {
+      "epoch": 22.04774193548387,
+      "grad_norm": 0.028917403891682625,
+      "learning_rate": 5.292999135695765e-05,
+      "loss": 0.2737,
+      "step": 4259
+    },
+    {
+      "epoch": 22.052903225806453,
+      "grad_norm": 0.0182275902479887,
+      "learning_rate": 5.289541918755402e-05,
+      "loss": 0.2636,
+      "step": 4260
+    },
+    {
+      "epoch": 22.058064516129033,
+      "grad_norm": 0.10751822590827942,
+      "learning_rate": 5.2860847018150394e-05,
+      "loss": 0.2828,
+      "step": 4261
+    },
+    {
+      "epoch": 22.063225806451612,
+      "grad_norm": 0.13468705117702484,
+      "learning_rate": 5.282627484874676e-05,
+      "loss": 0.2826,
+      "step": 4262
+    },
+    {
+      "epoch": 22.068387096774195,
+      "grad_norm": 0.15612323582172394,
+      "learning_rate": 5.2791702679343135e-05,
+      "loss": 0.3092,
+      "step": 4263
+    },
+    {
+      "epoch": 22.073548387096775,
+      "grad_norm": 0.03704652190208435,
+      "learning_rate": 5.27571305099395e-05,
+      "loss": 0.2979,
+      "step": 4264
+    },
+    {
+      "epoch": 22.078709677419354,
+      "grad_norm": 0.034283824265003204,
+      "learning_rate": 5.272255834053587e-05,
+      "loss": 0.2462,
+      "step": 4265
+    },
+    {
+      "epoch": 22.083870967741934,
+      "grad_norm": 0.017200633883476257,
+      "learning_rate": 5.268798617113224e-05,
+      "loss": 0.2922,
+      "step": 4266
+    },
+    {
+      "epoch": 22.089032258064517,
+      "grad_norm": 0.01821410469710827,
+      "learning_rate": 5.265341400172861e-05,
+      "loss": 0.2912,
+      "step": 4267
+    },
+    {
+      "epoch": 22.094193548387096,
+      "grad_norm": 0.026952140033245087,
+      "learning_rate": 5.261884183232498e-05,
+      "loss": 0.2946,
+      "step": 4268
+    },
+    {
+      "epoch": 22.099354838709676,
+      "grad_norm": 0.05561650171875954,
+      "learning_rate": 5.258426966292135e-05,
+      "loss": 0.2926,
+      "step": 4269
+    },
+    {
+      "epoch": 22.10451612903226,
+      "grad_norm": 0.03197106719017029,
+      "learning_rate": 5.2549697493517724e-05,
+      "loss": 0.2709,
+      "step": 4270
+    },
+    {
+      "epoch": 22.10967741935484,
+      "grad_norm": 0.031340569257736206,
+      "learning_rate": 5.251512532411409e-05,
+      "loss": 0.2991,
+      "step": 4271
+    },
+    {
+      "epoch": 22.114838709677418,
+      "grad_norm": 0.030056919902563095,
+      "learning_rate": 5.2480553154710465e-05,
+      "loss": 0.2574,
+      "step": 4272
+    },
+    {
+      "epoch": 22.12,
+      "grad_norm": 0.026764174923300743,
+      "learning_rate": 5.244598098530683e-05,
+      "loss": 0.2904,
+      "step": 4273
+    },
+    {
+      "epoch": 22.12516129032258,
+      "grad_norm": 0.019760994240641594,
+      "learning_rate": 5.24114088159032e-05,
+      "loss": 0.2965,
+      "step": 4274
+    },
+    {
+      "epoch": 22.13032258064516,
+      "grad_norm": 0.03369047865271568,
+      "learning_rate": 5.237683664649957e-05,
+      "loss": 0.3061,
+      "step": 4275
+    },
+    {
+      "epoch": 22.135483870967743,
+      "grad_norm": 0.03741384670138359,
+      "learning_rate": 5.234226447709594e-05,
+      "loss": 0.2723,
+      "step": 4276
+    },
+    {
+      "epoch": 22.140645161290323,
+      "grad_norm": 0.020829223096370697,
+      "learning_rate": 5.230769230769231e-05,
+      "loss": 0.2837,
+      "step": 4277
+    },
+    {
+      "epoch": 22.145806451612902,
+      "grad_norm": 0.03411373123526573,
+      "learning_rate": 5.227312013828868e-05,
+      "loss": 0.2563,
+      "step": 4278
+    },
+    {
+      "epoch": 22.150967741935485,
+      "grad_norm": 0.10742532461881638,
+      "learning_rate": 5.2238547968885054e-05,
+      "loss": 0.2935,
+      "step": 4279
+    },
+    {
+      "epoch": 22.156129032258065,
+      "grad_norm": 0.07373546808958054,
+      "learning_rate": 5.220397579948142e-05,
+      "loss": 0.2915,
+      "step": 4280
+    },
+    {
+      "epoch": 22.161290322580644,
+      "grad_norm": 0.033313948661088943,
+      "learning_rate": 5.216940363007779e-05,
+      "loss": 0.2361,
+      "step": 4281
+    },
+    {
+      "epoch": 22.166451612903227,
+      "grad_norm": 0.039179727435112,
+      "learning_rate": 5.213483146067416e-05,
+      "loss": 0.2796,
+      "step": 4282
+    },
+    {
+      "epoch": 22.171612903225807,
+      "grad_norm": 0.029311498627066612,
+      "learning_rate": 5.210025929127053e-05,
+      "loss": 0.2874,
+      "step": 4283
+    },
+    {
+      "epoch": 22.176774193548386,
+      "grad_norm": 0.04471776261925697,
+      "learning_rate": 5.20656871218669e-05,
+      "loss": 0.2736,
+      "step": 4284
+    },
+    {
+      "epoch": 22.18193548387097,
+      "grad_norm": 0.10114475339651108,
+      "learning_rate": 5.203111495246327e-05,
+      "loss": 0.2451,
+      "step": 4285
+    },
+    {
+      "epoch": 22.18709677419355,
+      "grad_norm": 0.02495030127465725,
+      "learning_rate": 5.199654278305964e-05,
+      "loss": 0.2501,
+      "step": 4286
+    },
+    {
+      "epoch": 22.19225806451613,
+      "grad_norm": 0.031571246683597565,
+      "learning_rate": 5.196197061365601e-05,
+      "loss": 0.2688,
+      "step": 4287
+    },
+    {
+      "epoch": 22.19741935483871,
+      "grad_norm": 0.023481128737330437,
+      "learning_rate": 5.1927398444252384e-05,
+      "loss": 0.2832,
+      "step": 4288
+    },
+    {
+      "epoch": 22.20258064516129,
+      "grad_norm": 0.02130170352756977,
+      "learning_rate": 5.189282627484875e-05,
+      "loss": 0.2539,
+      "step": 4289
+    },
+    {
+      "epoch": 22.20774193548387,
+      "grad_norm": 0.024031946435570717,
+      "learning_rate": 5.185825410544511e-05,
+      "loss": 0.2988,
+      "step": 4290
+    },
+    {
+      "epoch": 22.21290322580645,
+      "grad_norm": 0.040037769824266434,
+      "learning_rate": 5.182368193604149e-05,
+      "loss": 0.2942,
+      "step": 4291
+    },
+    {
+      "epoch": 22.218064516129033,
+      "grad_norm": 0.04943759739398956,
+      "learning_rate": 5.178910976663786e-05,
+      "loss": 0.2457,
+      "step": 4292
+    },
+    {
+      "epoch": 22.223225806451612,
+      "grad_norm": 0.029437171295285225,
+      "learning_rate": 5.175453759723423e-05,
+      "loss": 0.2427,
+      "step": 4293
+    },
+    {
+      "epoch": 22.228387096774192,
+      "grad_norm": 0.050126709043979645,
+      "learning_rate": 5.17199654278306e-05,
+      "loss": 0.2764,
+      "step": 4294
+    },
+    {
+      "epoch": 22.233548387096775,
+      "grad_norm": 0.025746893137693405,
+      "learning_rate": 5.168539325842697e-05,
+      "loss": 0.2779,
+      "step": 4295
+    },
+    {
+      "epoch": 22.238709677419354,
+      "grad_norm": 0.024472882971167564,
+      "learning_rate": 5.165082108902334e-05,
+      "loss": 0.2645,
+      "step": 4296
+    },
+    {
+      "epoch": 22.243870967741934,
+      "grad_norm": 0.023123372346162796,
+      "learning_rate": 5.1616248919619714e-05,
+      "loss": 0.2963,
+      "step": 4297
+    },
+    {
+      "epoch": 22.249032258064517,
+      "grad_norm": 0.020349685102701187,
+      "learning_rate": 5.158167675021608e-05,
+      "loss": 0.2848,
+      "step": 4298
+    },
+    {
+      "epoch": 22.254193548387097,
+      "grad_norm": 0.0484347864985466,
+      "learning_rate": 5.154710458081244e-05,
+      "loss": 0.2799,
+      "step": 4299
+    },
+    {
+      "epoch": 22.259354838709676,
+      "grad_norm": 0.032545071095228195,
+      "learning_rate": 5.151253241140882e-05,
+      "loss": 0.2527,
+      "step": 4300
+    },
+    {
+      "epoch": 22.26451612903226,
+      "grad_norm": 0.022796887904405594,
+      "learning_rate": 5.147796024200519e-05,
+      "loss": 0.3055,
+      "step": 4301
+    },
+    {
+      "epoch": 22.26967741935484,
+      "grad_norm": 0.027361690998077393,
+      "learning_rate": 5.144338807260156e-05,
+      "loss": 0.3117,
+      "step": 4302
+    },
+    {
+      "epoch": 22.274838709677418,
+      "grad_norm": 0.022246241569519043,
+      "learning_rate": 5.140881590319793e-05,
+      "loss": 0.2832,
+      "step": 4303
+    },
+    {
+      "epoch": 22.28,
+      "grad_norm": 0.021735643967986107,
+      "learning_rate": 5.13742437337943e-05,
+      "loss": 0.2619,
+      "step": 4304
+    },
+    {
+      "epoch": 22.28516129032258,
+      "grad_norm": 0.030167274177074432,
+      "learning_rate": 5.133967156439067e-05,
+      "loss": 0.2564,
+      "step": 4305
+    },
+    {
+      "epoch": 22.29032258064516,
+      "grad_norm": 0.02640429325401783,
+      "learning_rate": 5.130509939498703e-05,
+      "loss": 0.2773,
+      "step": 4306
+    },
+    {
+      "epoch": 22.295483870967743,
+      "grad_norm": 0.03045840375125408,
+      "learning_rate": 5.127052722558341e-05,
+      "loss": 0.244,
+      "step": 4307
+    },
+    {
+      "epoch": 22.300645161290323,
+      "grad_norm": 0.04198964685201645,
+      "learning_rate": 5.123595505617977e-05,
+      "loss": 0.2752,
+      "step": 4308
+    },
+    {
+      "epoch": 22.305806451612902,
+      "grad_norm": 0.10023930668830872,
+      "learning_rate": 5.120138288677615e-05,
+      "loss": 0.2706,
+      "step": 4309
+    },
+    {
+      "epoch": 22.310967741935485,
+      "grad_norm": 0.10821788012981415,
+      "learning_rate": 5.116681071737252e-05,
+      "loss": 0.2807,
+      "step": 4310
+    },
+    {
+      "epoch": 22.316129032258065,
+      "grad_norm": 0.03186626732349396,
+      "learning_rate": 5.113223854796889e-05,
+      "loss": 0.2687,
+      "step": 4311
+    },
+    {
+      "epoch": 22.321290322580644,
+      "grad_norm": 0.032801348716020584,
+      "learning_rate": 5.109766637856526e-05,
+      "loss": 0.2815,
+      "step": 4312
+    },
+    {
+      "epoch": 22.326451612903227,
+      "grad_norm": 0.021244438365101814,
+      "learning_rate": 5.106309420916163e-05,
+      "loss": 0.2926,
+      "step": 4313
+    },
+    {
+      "epoch": 22.331612903225807,
+      "grad_norm": 0.022751549258828163,
+      "learning_rate": 5.1028522039758e-05,
+      "loss": 0.2902,
+      "step": 4314
+    },
+    {
+      "epoch": 22.336774193548386,
+      "grad_norm": 0.03625796362757683,
+      "learning_rate": 5.099394987035436e-05,
+      "loss": 0.289,
+      "step": 4315
+    },
+    {
+      "epoch": 22.34193548387097,
+      "grad_norm": 0.028205417096614838,
+      "learning_rate": 5.095937770095074e-05,
+      "loss": 0.3109,
+      "step": 4316
+    },
+    {
+      "epoch": 22.34709677419355,
+      "grad_norm": 0.03380962833762169,
+      "learning_rate": 5.09248055315471e-05,
+      "loss": 0.2942,
+      "step": 4317
+    },
+    {
+      "epoch": 22.35225806451613,
+      "grad_norm": 0.07254226505756378,
+      "learning_rate": 5.089023336214348e-05,
+      "loss": 0.2916,
+      "step": 4318
+    },
+    {
+      "epoch": 22.35741935483871,
+      "grad_norm": 0.019410861656069756,
+      "learning_rate": 5.085566119273985e-05,
+      "loss": 0.2522,
+      "step": 4319
+    },
+    {
+      "epoch": 22.36258064516129,
+      "grad_norm": 0.0357050746679306,
+      "learning_rate": 5.082108902333622e-05,
+      "loss": 0.2963,
+      "step": 4320
+    },
+    {
+      "epoch": 22.36774193548387,
+      "grad_norm": 0.02619711309671402,
+      "learning_rate": 5.078651685393259e-05,
+      "loss": 0.3013,
+      "step": 4321
+    },
+    {
+      "epoch": 22.37290322580645,
+      "grad_norm": 0.03796404227614403,
+      "learning_rate": 5.075194468452895e-05,
+      "loss": 0.2921,
+      "step": 4322
+    },
+    {
+      "epoch": 22.378064516129033,
+      "grad_norm": 0.025019211694598198,
+      "learning_rate": 5.071737251512533e-05,
+      "loss": 0.2673,
+      "step": 4323
+    },
+    {
+      "epoch": 22.383225806451613,
+      "grad_norm": 0.02760924957692623,
+      "learning_rate": 5.068280034572169e-05,
+      "loss": 0.2648,
+      "step": 4324
+    },
+    {
+      "epoch": 22.388387096774192,
+      "grad_norm": 0.09126710146665573,
+      "learning_rate": 5.064822817631807e-05,
+      "loss": 0.316,
+      "step": 4325
+    },
+    {
+      "epoch": 22.393548387096775,
+      "grad_norm": 0.024025749415159225,
+      "learning_rate": 5.061365600691443e-05,
+      "loss": 0.2915,
+      "step": 4326
+    },
+    {
+      "epoch": 22.398709677419355,
+      "grad_norm": 0.02887902222573757,
+      "learning_rate": 5.057908383751081e-05,
+      "loss": 0.295,
+      "step": 4327
+    },
+    {
+      "epoch": 22.403870967741934,
+      "grad_norm": 0.04702477157115936,
+      "learning_rate": 5.054451166810718e-05,
+      "loss": 0.2781,
+      "step": 4328
+    },
+    {
+      "epoch": 22.409032258064517,
+      "grad_norm": 0.03215954080224037,
+      "learning_rate": 5.050993949870355e-05,
+      "loss": 0.2911,
+      "step": 4329
+    },
+    {
+      "epoch": 22.414193548387097,
+      "grad_norm": 0.13051266968250275,
+      "learning_rate": 5.047536732929992e-05,
+      "loss": 0.2764,
+      "step": 4330
+    },
+    {
+      "epoch": 22.419354838709676,
+      "grad_norm": 0.026147279888391495,
+      "learning_rate": 5.044079515989628e-05,
+      "loss": 0.2625,
+      "step": 4331
+    },
+    {
+      "epoch": 22.42451612903226,
+      "grad_norm": 0.03915189951658249,
+      "learning_rate": 5.040622299049266e-05,
+      "loss": 0.2965,
+      "step": 4332
+    },
+    {
+      "epoch": 22.42967741935484,
+      "grad_norm": 0.04475945606827736,
+      "learning_rate": 5.037165082108902e-05,
+      "loss": 0.2709,
+      "step": 4333
+    },
+    {
+      "epoch": 22.434838709677418,
+      "grad_norm": 0.03477980196475983,
+      "learning_rate": 5.03370786516854e-05,
+      "loss": 0.3022,
+      "step": 4334
+    },
+    {
+      "epoch": 22.44,
+      "grad_norm": 0.03761838376522064,
+      "learning_rate": 5.030250648228176e-05,
+      "loss": 0.3013,
+      "step": 4335
+    },
+    {
+      "epoch": 22.44516129032258,
+      "grad_norm": 0.02715521678328514,
+      "learning_rate": 5.026793431287814e-05,
+      "loss": 0.2843,
+      "step": 4336
+    },
+    {
+      "epoch": 22.45032258064516,
+      "grad_norm": 0.13360649347305298,
+      "learning_rate": 5.023336214347451e-05,
+      "loss": 0.274,
+      "step": 4337
+    },
+    {
+      "epoch": 22.455483870967743,
+      "grad_norm": 0.06870749592781067,
+      "learning_rate": 5.019878997407087e-05,
+      "loss": 0.316,
+      "step": 4338
+    },
+    {
+      "epoch": 22.460645161290323,
+      "grad_norm": 0.02148730307817459,
+      "learning_rate": 5.016421780466725e-05,
+      "loss": 0.313,
+      "step": 4339
+    },
+    {
+      "epoch": 22.465806451612902,
+      "grad_norm": 0.04800047725439072,
+      "learning_rate": 5.012964563526361e-05,
+      "loss": 0.2857,
+      "step": 4340
+    },
+    {
+      "epoch": 22.470967741935485,
+      "grad_norm": 0.02527809701859951,
+      "learning_rate": 5.009507346585999e-05,
+      "loss": 0.3026,
+      "step": 4341
+    },
+    {
+      "epoch": 22.476129032258065,
+      "grad_norm": 0.028669394552707672,
+      "learning_rate": 5.006050129645635e-05,
+      "loss": 0.2375,
+      "step": 4342
+    },
+    {
+      "epoch": 22.481290322580644,
+      "grad_norm": 0.023236067965626717,
+      "learning_rate": 5.002592912705273e-05,
+      "loss": 0.3063,
+      "step": 4343
+    },
+    {
+      "epoch": 22.486451612903227,
+      "grad_norm": 0.020622270181775093,
+      "learning_rate": 4.999135695764909e-05,
+      "loss": 0.2939,
+      "step": 4344
+    },
+    {
+      "epoch": 22.491612903225807,
+      "grad_norm": 0.02843880094587803,
+      "learning_rate": 4.9956784788245465e-05,
+      "loss": 0.2763,
+      "step": 4345
+    },
+    {
+      "epoch": 22.496774193548386,
+      "grad_norm": 0.028301237151026726,
+      "learning_rate": 4.992221261884183e-05,
+      "loss": 0.2707,
+      "step": 4346
+    },
+    {
+      "epoch": 22.501935483870966,
+      "grad_norm": 0.09172875434160233,
+      "learning_rate": 4.9887640449438205e-05,
+      "loss": 0.2812,
+      "step": 4347
+    },
+    {
+      "epoch": 22.50709677419355,
+      "grad_norm": 0.029398003593087196,
+      "learning_rate": 4.985306828003458e-05,
+      "loss": 0.3217,
+      "step": 4348
+    },
+    {
+      "epoch": 22.51225806451613,
+      "grad_norm": 0.08838598430156708,
+      "learning_rate": 4.9818496110630946e-05,
+      "loss": 0.2832,
+      "step": 4349
+    },
+    {
+      "epoch": 22.517419354838708,
+      "grad_norm": 0.020292574539780617,
+      "learning_rate": 4.978392394122731e-05,
+      "loss": 0.2593,
+      "step": 4350
+    },
+    {
+      "epoch": 22.517419354838708,
+      "eval_loss": 3.238849639892578,
+      "eval_runtime": 21.1577,
+      "eval_samples_per_second": 3.734,
+      "eval_steps_per_second": 0.473,
+      "step": 4350
+    },
+    {
+      "epoch": 22.52258064516129,
+      "grad_norm": 0.029798032715916634,
+      "learning_rate": 4.974935177182368e-05,
+      "loss": 0.3037,
+      "step": 4351
+    },
+    {
+      "epoch": 22.52774193548387,
+      "grad_norm": 0.028746144846081734,
+      "learning_rate": 4.9714779602420054e-05,
+      "loss": 0.2978,
+      "step": 4352
+    },
+    {
+      "epoch": 22.53290322580645,
+      "grad_norm": 0.09172552824020386,
+      "learning_rate": 4.968020743301642e-05,
+      "loss": 0.2618,
+      "step": 4353
+    },
+    {
+      "epoch": 22.538064516129033,
+      "grad_norm": 0.027887294068932533,
+      "learning_rate": 4.9645635263612794e-05,
+      "loss": 0.2665,
+      "step": 4354
+    },
+    {
+      "epoch": 22.543225806451613,
+      "grad_norm": 0.031142232939600945,
+      "learning_rate": 4.961106309420916e-05,
+      "loss": 0.2869,
+      "step": 4355
+    },
+    {
+      "epoch": 22.548387096774192,
+      "grad_norm": 0.02766416221857071,
+      "learning_rate": 4.9576490924805535e-05,
+      "loss": 0.2957,
+      "step": 4356
+    },
+    {
+      "epoch": 22.553548387096775,
+      "grad_norm": 0.023209067061543465,
+      "learning_rate": 4.954191875540191e-05,
+      "loss": 0.2822,
+      "step": 4357
+    },
+    {
+      "epoch": 22.558709677419355,
+      "grad_norm": 0.025978531688451767,
+      "learning_rate": 4.950734658599827e-05,
+      "loss": 0.2936,
+      "step": 4358
+    },
+    {
+      "epoch": 22.563870967741934,
+      "grad_norm": 0.03850207477807999,
+      "learning_rate": 4.947277441659464e-05,
+      "loss": 0.294,
+      "step": 4359
+    },
+    {
+      "epoch": 22.569032258064517,
+      "grad_norm": 0.033415861427783966,
+      "learning_rate": 4.943820224719101e-05,
+      "loss": 0.2981,
+      "step": 4360
+    },
+    {
+      "epoch": 22.574193548387097,
+      "grad_norm": 0.08876105397939682,
+      "learning_rate": 4.9403630077787384e-05,
+      "loss": 0.2777,
+      "step": 4361
+    },
+    {
+      "epoch": 22.579354838709676,
+      "grad_norm": 0.026671038940548897,
+      "learning_rate": 4.936905790838375e-05,
+      "loss": 0.2827,
+      "step": 4362
+    },
+    {
+      "epoch": 22.58451612903226,
+      "grad_norm": 0.08743712306022644,
+      "learning_rate": 4.9334485738980124e-05,
+      "loss": 0.272,
+      "step": 4363
+    },
+    {
+      "epoch": 22.58967741935484,
+      "grad_norm": 0.018266616389155388,
+      "learning_rate": 4.929991356957649e-05,
+      "loss": 0.2601,
+      "step": 4364
+    },
+    {
+      "epoch": 22.59483870967742,
+      "grad_norm": 0.07181750982999802,
+      "learning_rate": 4.9265341400172865e-05,
+      "loss": 0.2891,
+      "step": 4365
+    },
+    {
+      "epoch": 22.6,
+      "grad_norm": 0.028820691630244255,
+      "learning_rate": 4.923076923076924e-05,
+      "loss": 0.252,
+      "step": 4366
+    },
+    {
+      "epoch": 22.60516129032258,
+      "grad_norm": 0.030098695307970047,
+      "learning_rate": 4.91961970613656e-05,
+      "loss": 0.2812,
+      "step": 4367
+    },
+    {
+      "epoch": 22.61032258064516,
+      "grad_norm": 0.02215169370174408,
+      "learning_rate": 4.916162489196197e-05,
+      "loss": 0.2982,
+      "step": 4368
+    },
+    {
+      "epoch": 22.615483870967743,
+      "grad_norm": 0.027879446744918823,
+      "learning_rate": 4.912705272255834e-05,
+      "loss": 0.2835,
+      "step": 4369
+    },
+    {
+      "epoch": 22.620645161290323,
+      "grad_norm": 0.022842910140752792,
+      "learning_rate": 4.9092480553154714e-05,
+      "loss": 0.3018,
+      "step": 4370
+    },
+    {
+      "epoch": 22.625806451612902,
+      "grad_norm": 0.0220413226634264,
+      "learning_rate": 4.905790838375108e-05,
+      "loss": 0.2969,
+      "step": 4371
+    },
+    {
+      "epoch": 22.630967741935486,
+      "grad_norm": 0.08707819879055023,
+      "learning_rate": 4.9023336214347454e-05,
+      "loss": 0.3034,
+      "step": 4372
+    },
+    {
+      "epoch": 22.636129032258065,
+      "grad_norm": 0.02661856822669506,
+      "learning_rate": 4.898876404494382e-05,
+      "loss": 0.2622,
+      "step": 4373
+    },
+    {
+      "epoch": 22.641290322580645,
+      "grad_norm": 0.0377453938126564,
+      "learning_rate": 4.8954191875540195e-05,
+      "loss": 0.2568,
+      "step": 4374
+    },
+    {
+      "epoch": 22.646451612903228,
+      "grad_norm": 0.04136831685900688,
+      "learning_rate": 4.891961970613656e-05,
+      "loss": 0.2783,
+      "step": 4375
+    },
+    {
+      "epoch": 22.651612903225807,
+      "grad_norm": 0.038190994411706924,
+      "learning_rate": 4.888504753673293e-05,
+      "loss": 0.2895,
+      "step": 4376
+    },
+    {
+      "epoch": 22.656774193548387,
+      "grad_norm": 0.03999673202633858,
+      "learning_rate": 4.88504753673293e-05,
+      "loss": 0.2788,
+      "step": 4377
+    },
+    {
+      "epoch": 22.661935483870966,
+      "grad_norm": 0.026538575068116188,
+      "learning_rate": 4.881590319792567e-05,
+      "loss": 0.2847,
+      "step": 4378
+    },
+    {
+      "epoch": 22.66709677419355,
+      "grad_norm": 0.12291552871465683,
+      "learning_rate": 4.8781331028522044e-05,
+      "loss": 0.3057,
+      "step": 4379
+    },
+    {
+      "epoch": 22.67225806451613,
+      "grad_norm": 0.061864182353019714,
+      "learning_rate": 4.874675885911841e-05,
+      "loss": 0.2634,
+      "step": 4380
+    },
+    {
+      "epoch": 22.677419354838708,
+      "grad_norm": 0.03748684376478195,
+      "learning_rate": 4.8712186689714784e-05,
+      "loss": 0.2953,
+      "step": 4381
+    },
+    {
+      "epoch": 22.68258064516129,
+      "grad_norm": 0.0921582579612732,
+      "learning_rate": 4.867761452031115e-05,
+      "loss": 0.2818,
+      "step": 4382
+    },
+    {
+      "epoch": 22.68774193548387,
+      "grad_norm": 0.02660153992474079,
+      "learning_rate": 4.864304235090752e-05,
+      "loss": 0.3118,
+      "step": 4383
+    },
+    {
+      "epoch": 22.69290322580645,
+      "grad_norm": 0.07685314118862152,
+      "learning_rate": 4.860847018150389e-05,
+      "loss": 0.2792,
+      "step": 4384
+    },
+    {
+      "epoch": 22.698064516129033,
+      "grad_norm": 0.02766243740916252,
+      "learning_rate": 4.857389801210026e-05,
+      "loss": 0.2562,
+      "step": 4385
+    },
+    {
+      "epoch": 22.703225806451613,
+      "grad_norm": 0.03175685927271843,
+      "learning_rate": 4.853932584269663e-05,
+      "loss": 0.2884,
+      "step": 4386
+    },
+    {
+      "epoch": 22.708387096774192,
+      "grad_norm": 0.017997197806835175,
+      "learning_rate": 4.8504753673293e-05,
+      "loss": 0.281,
+      "step": 4387
+    },
+    {
+      "epoch": 22.713548387096775,
+      "grad_norm": 0.03474852070212364,
+      "learning_rate": 4.8470181503889374e-05,
+      "loss": 0.3057,
+      "step": 4388
+    },
+    {
+      "epoch": 22.718709677419355,
+      "grad_norm": 0.0602448545396328,
+      "learning_rate": 4.843560933448574e-05,
+      "loss": 0.2992,
+      "step": 4389
+    },
+    {
+      "epoch": 22.723870967741934,
+      "grad_norm": 0.029222218319773674,
+      "learning_rate": 4.8401037165082114e-05,
+      "loss": 0.2852,
+      "step": 4390
+    },
+    {
+      "epoch": 22.729032258064517,
+      "grad_norm": 0.025412922725081444,
+      "learning_rate": 4.836646499567848e-05,
+      "loss": 0.2922,
+      "step": 4391
+    },
+    {
+      "epoch": 22.734193548387097,
+      "grad_norm": 0.03837407007813454,
+      "learning_rate": 4.833189282627485e-05,
+      "loss": 0.2725,
+      "step": 4392
+    },
+    {
+      "epoch": 22.739354838709676,
+      "grad_norm": 0.03703325241804123,
+      "learning_rate": 4.829732065687122e-05,
+      "loss": 0.2573,
+      "step": 4393
+    },
+    {
+      "epoch": 22.74451612903226,
+      "grad_norm": 0.03456931933760643,
+      "learning_rate": 4.826274848746759e-05,
+      "loss": 0.2542,
+      "step": 4394
+    },
+    {
+      "epoch": 22.74967741935484,
+      "grad_norm": 0.03185340389609337,
+      "learning_rate": 4.822817631806396e-05,
+      "loss": 0.264,
+      "step": 4395
+    },
+    {
+      "epoch": 22.75483870967742,
+      "grad_norm": 0.02060624025762081,
+      "learning_rate": 4.819360414866033e-05,
+      "loss": 0.2898,
+      "step": 4396
+    },
+    {
+      "epoch": 22.76,
+      "grad_norm": 0.04136863723397255,
+      "learning_rate": 4.8159031979256704e-05,
+      "loss": 0.2765,
+      "step": 4397
+    },
+    {
+      "epoch": 22.76516129032258,
+      "grad_norm": 0.044871874153614044,
+      "learning_rate": 4.812445980985307e-05,
+      "loss": 0.3004,
+      "step": 4398
+    },
+    {
+      "epoch": 22.77032258064516,
+      "grad_norm": 0.07586289197206497,
+      "learning_rate": 4.808988764044944e-05,
+      "loss": 0.286,
+      "step": 4399
+    },
+    {
+      "epoch": 22.775483870967744,
+      "grad_norm": 0.034491654485464096,
+      "learning_rate": 4.805531547104581e-05,
+      "loss": 0.2809,
+      "step": 4400
+    },
+    {
+      "epoch": 22.780645161290323,
+      "grad_norm": 0.03732624650001526,
+      "learning_rate": 4.802074330164218e-05,
+      "loss": 0.2967,
+      "step": 4401
+    },
+    {
+      "epoch": 22.785806451612903,
+      "grad_norm": 0.029282864183187485,
+      "learning_rate": 4.798617113223855e-05,
+      "loss": 0.2729,
+      "step": 4402
+    },
+    {
+      "epoch": 22.790967741935482,
+      "grad_norm": 0.02887118235230446,
+      "learning_rate": 4.795159896283492e-05,
+      "loss": 0.3031,
+      "step": 4403
+    },
+    {
+      "epoch": 22.796129032258065,
+      "grad_norm": 0.0906093493103981,
+      "learning_rate": 4.791702679343129e-05,
+      "loss": 0.2526,
+      "step": 4404
+    },
+    {
+      "epoch": 22.801290322580645,
+      "grad_norm": 0.022870320826768875,
+      "learning_rate": 4.788245462402766e-05,
+      "loss": 0.2978,
+      "step": 4405
+    },
+    {
+      "epoch": 22.806451612903224,
+      "grad_norm": 0.054514799267053604,
+      "learning_rate": 4.7847882454624034e-05,
+      "loss": 0.3001,
+      "step": 4406
+    },
+    {
+      "epoch": 22.811612903225807,
+      "grad_norm": 0.08115226775407791,
+      "learning_rate": 4.78133102852204e-05,
+      "loss": 0.262,
+      "step": 4407
+    },
+    {
+      "epoch": 22.816774193548387,
+      "grad_norm": 0.03023412637412548,
+      "learning_rate": 4.777873811581677e-05,
+      "loss": 0.2777,
+      "step": 4408
+    },
+    {
+      "epoch": 22.821935483870966,
+      "grad_norm": 0.031936876475811005,
+      "learning_rate": 4.774416594641314e-05,
+      "loss": 0.2827,
+      "step": 4409
+    },
+    {
+      "epoch": 22.82709677419355,
+      "grad_norm": 0.039646223187446594,
+      "learning_rate": 4.770959377700951e-05,
+      "loss": 0.2803,
+      "step": 4410
+    },
+    {
+      "epoch": 22.83225806451613,
+      "grad_norm": 0.029585188254714012,
+      "learning_rate": 4.767502160760588e-05,
+      "loss": 0.2861,
+      "step": 4411
+    },
+    {
+      "epoch": 22.83741935483871,
+      "grad_norm": 0.02701611816883087,
+      "learning_rate": 4.764044943820225e-05,
+      "loss": 0.297,
+      "step": 4412
+    },
+    {
+      "epoch": 22.84258064516129,
+      "grad_norm": 0.024209266528487206,
+      "learning_rate": 4.760587726879862e-05,
+      "loss": 0.2791,
+      "step": 4413
+    },
+    {
+      "epoch": 22.84774193548387,
+      "grad_norm": 0.023834964260458946,
+      "learning_rate": 4.757130509939499e-05,
+      "loss": 0.2579,
+      "step": 4414
+    },
+    {
+      "epoch": 22.85290322580645,
+      "grad_norm": 0.07242608815431595,
+      "learning_rate": 4.753673292999136e-05,
+      "loss": 0.2813,
+      "step": 4415
+    },
+    {
+      "epoch": 22.858064516129033,
+      "grad_norm": 0.025466622784733772,
+      "learning_rate": 4.750216076058773e-05,
+      "loss": 0.2881,
+      "step": 4416
+    },
+    {
+      "epoch": 22.863225806451613,
+      "grad_norm": 0.01929425075650215,
+      "learning_rate": 4.74675885911841e-05,
+      "loss": 0.3129,
+      "step": 4417
+    },
+    {
+      "epoch": 22.868387096774192,
+      "grad_norm": 0.02349851280450821,
+      "learning_rate": 4.743301642178047e-05,
+      "loss": 0.3027,
+      "step": 4418
+    },
+    {
+      "epoch": 22.873548387096776,
+      "grad_norm": 0.027905091643333435,
+      "learning_rate": 4.739844425237684e-05,
+      "loss": 0.2907,
+      "step": 4419
+    },
+    {
+      "epoch": 22.878709677419355,
+      "grad_norm": 0.03666729852557182,
+      "learning_rate": 4.736387208297321e-05,
+      "loss": 0.28,
+      "step": 4420
+    },
+    {
+      "epoch": 22.883870967741935,
+      "grad_norm": 0.18977682292461395,
+      "learning_rate": 4.732929991356958e-05,
+      "loss": 0.2652,
+      "step": 4421
+    },
+    {
+      "epoch": 22.889032258064518,
+      "grad_norm": 0.04586656391620636,
+      "learning_rate": 4.729472774416595e-05,
+      "loss": 0.3039,
+      "step": 4422
+    },
+    {
+      "epoch": 22.894193548387097,
+      "grad_norm": 0.03590463474392891,
+      "learning_rate": 4.726015557476231e-05,
+      "loss": 0.2878,
+      "step": 4423
+    },
+    {
+      "epoch": 22.899354838709677,
+      "grad_norm": 0.029037198051810265,
+      "learning_rate": 4.722558340535869e-05,
+      "loss": 0.2613,
+      "step": 4424
+    },
+    {
+      "epoch": 22.90451612903226,
+      "grad_norm": 0.030352793633937836,
+      "learning_rate": 4.719101123595506e-05,
+      "loss": 0.2827,
+      "step": 4425
+    },
+    {
+      "epoch": 22.90967741935484,
+      "grad_norm": 0.03891788050532341,
+      "learning_rate": 4.715643906655143e-05,
+      "loss": 0.2601,
+      "step": 4426
+    },
+    {
+      "epoch": 22.91483870967742,
+      "grad_norm": 0.025335263460874557,
+      "learning_rate": 4.71218668971478e-05,
+      "loss": 0.2703,
+      "step": 4427
+    },
+    {
+      "epoch": 22.92,
+      "grad_norm": 0.03957659378647804,
+      "learning_rate": 4.708729472774417e-05,
+      "loss": 0.2576,
+      "step": 4428
+    },
+    {
+      "epoch": 22.92516129032258,
+      "grad_norm": 0.02497333474457264,
+      "learning_rate": 4.705272255834054e-05,
+      "loss": 0.2875,
+      "step": 4429
+    },
+    {
+      "epoch": 22.93032258064516,
+      "grad_norm": 0.028858091682195663,
+      "learning_rate": 4.701815038893691e-05,
+      "loss": 0.3001,
+      "step": 4430
+    },
+    {
+      "epoch": 22.93548387096774,
+      "grad_norm": 0.015040344558656216,
+      "learning_rate": 4.6983578219533276e-05,
+      "loss": 0.2659,
+      "step": 4431
+    },
+    {
+      "epoch": 22.940645161290323,
+      "grad_norm": 0.03293311968445778,
+      "learning_rate": 4.694900605012964e-05,
+      "loss": 0.2973,
+      "step": 4432
+    },
+    {
+      "epoch": 22.945806451612903,
+      "grad_norm": 0.035227589309215546,
+      "learning_rate": 4.691443388072602e-05,
+      "loss": 0.3031,
+      "step": 4433
+    },
+    {
+      "epoch": 22.950967741935482,
+      "grad_norm": 0.03252521902322769,
+      "learning_rate": 4.6879861711322384e-05,
+      "loss": 0.2997,
+      "step": 4434
+    },
+    {
+      "epoch": 22.956129032258065,
+      "grad_norm": 0.03301423043012619,
+      "learning_rate": 4.684528954191876e-05,
+      "loss": 0.2785,
+      "step": 4435
+    },
+    {
+      "epoch": 22.961290322580645,
+      "grad_norm": 0.033733464777469635,
+      "learning_rate": 4.681071737251513e-05,
+      "loss": 0.2931,
+      "step": 4436
+    },
+    {
+      "epoch": 22.966451612903224,
+      "grad_norm": 0.07159372419118881,
+      "learning_rate": 4.67761452031115e-05,
+      "loss": 0.2813,
+      "step": 4437
+    },
+    {
+      "epoch": 22.971612903225807,
+      "grad_norm": 0.02947429195046425,
+      "learning_rate": 4.674157303370787e-05,
+      "loss": 0.276,
+      "step": 4438
+    },
+    {
+      "epoch": 22.976774193548387,
+      "grad_norm": 0.03292400389909744,
+      "learning_rate": 4.670700086430424e-05,
+      "loss": 0.2815,
+      "step": 4439
+    },
+    {
+      "epoch": 22.981935483870966,
+      "grad_norm": 0.06614965200424194,
+      "learning_rate": 4.6672428694900606e-05,
+      "loss": 0.3064,
+      "step": 4440
+    },
+    {
+      "epoch": 22.98709677419355,
+      "grad_norm": 0.03046196699142456,
+      "learning_rate": 4.663785652549697e-05,
+      "loss": 0.2459,
+      "step": 4441
+    },
+    {
+      "epoch": 22.99225806451613,
+      "grad_norm": 0.03662056103348732,
+      "learning_rate": 4.660328435609335e-05,
+      "loss": 0.2898,
+      "step": 4442
+    },
+    {
+      "epoch": 22.99741935483871,
+      "grad_norm": 0.04582582041621208,
+      "learning_rate": 4.6568712186689714e-05,
+      "loss": 0.3735,
+      "step": 4443
+    },
+    {
+      "epoch": 23.003870967741936,
+      "grad_norm": 0.04604649543762207,
+      "learning_rate": 4.653414001728609e-05,
+      "loss": 0.3682,
+      "step": 4444
+    },
+    {
+      "epoch": 23.009032258064515,
+      "grad_norm": 0.02649128995835781,
+      "learning_rate": 4.649956784788246e-05,
+      "loss": 0.2824,
+      "step": 4445
+    },
+    {
+      "epoch": 23.014193548387098,
+      "grad_norm": 0.026875579729676247,
+      "learning_rate": 4.646499567847883e-05,
+      "loss": 0.3155,
+      "step": 4446
+    },
+    {
+      "epoch": 23.019354838709678,
+      "grad_norm": 0.030555635690689087,
+      "learning_rate": 4.64304235090752e-05,
+      "loss": 0.2738,
+      "step": 4447
+    },
+    {
+      "epoch": 23.024516129032257,
+      "grad_norm": 0.0222158282995224,
+      "learning_rate": 4.639585133967156e-05,
+      "loss": 0.3116,
+      "step": 4448
+    },
+    {
+      "epoch": 23.02967741935484,
+      "grad_norm": 0.026817036792635918,
+      "learning_rate": 4.6361279170267936e-05,
+      "loss": 0.2864,
+      "step": 4449
+    },
+    {
+      "epoch": 23.03483870967742,
+      "grad_norm": 0.017044581472873688,
+      "learning_rate": 4.63267070008643e-05,
+      "loss": 0.2786,
+      "step": 4450
+    },
+    {
+      "epoch": 23.04,
+      "grad_norm": 0.01927134208381176,
+      "learning_rate": 4.629213483146068e-05,
+      "loss": 0.312,
+      "step": 4451
+    },
+    {
+      "epoch": 23.045161290322582,
+      "grad_norm": 0.028131628409028053,
+      "learning_rate": 4.6257562662057044e-05,
+      "loss": 0.2649,
+      "step": 4452
+    },
+    {
+      "epoch": 23.05032258064516,
+      "grad_norm": 0.10648788511753082,
+      "learning_rate": 4.622299049265342e-05,
+      "loss": 0.2639,
+      "step": 4453
+    },
+    {
+      "epoch": 23.05548387096774,
+      "grad_norm": 0.021223848685622215,
+      "learning_rate": 4.618841832324979e-05,
+      "loss": 0.2922,
+      "step": 4454
+    },
+    {
+      "epoch": 23.060645161290324,
+      "grad_norm": 0.0321904681622982,
+      "learning_rate": 4.615384615384616e-05,
+      "loss": 0.2551,
+      "step": 4455
+    },
+    {
+      "epoch": 23.065806451612904,
+      "grad_norm": 0.026661261916160583,
+      "learning_rate": 4.6119273984442525e-05,
+      "loss": 0.286,
+      "step": 4456
+    },
+    {
+      "epoch": 23.070967741935483,
+      "grad_norm": 0.0487893670797348,
+      "learning_rate": 4.608470181503889e-05,
+      "loss": 0.2889,
+      "step": 4457
+    },
+    {
+      "epoch": 23.076129032258063,
+      "grad_norm": 0.01985863409936428,
+      "learning_rate": 4.6050129645635266e-05,
+      "loss": 0.2954,
+      "step": 4458
+    },
+    {
+      "epoch": 23.081290322580646,
+      "grad_norm": 0.024131163954734802,
+      "learning_rate": 4.601555747623163e-05,
+      "loss": 0.3035,
+      "step": 4459
+    },
+    {
+      "epoch": 23.086451612903225,
+      "grad_norm": 0.01707417704164982,
+      "learning_rate": 4.598098530682801e-05,
+      "loss": 0.2998,
+      "step": 4460
+    },
+    {
+      "epoch": 23.091612903225805,
+      "grad_norm": 0.02618245966732502,
+      "learning_rate": 4.5946413137424374e-05,
+      "loss": 0.2806,
+      "step": 4461
+    },
+    {
+      "epoch": 23.096774193548388,
+      "grad_norm": 0.03623175621032715,
+      "learning_rate": 4.591184096802075e-05,
+      "loss": 0.2748,
+      "step": 4462
+    },
+    {
+      "epoch": 23.101935483870967,
+      "grad_norm": 0.024031182751059532,
+      "learning_rate": 4.587726879861712e-05,
+      "loss": 0.29,
+      "step": 4463
+    },
+    {
+      "epoch": 23.107096774193547,
+      "grad_norm": 0.026866108179092407,
+      "learning_rate": 4.584269662921348e-05,
+      "loss": 0.2705,
+      "step": 4464
+    },
+    {
+      "epoch": 23.11225806451613,
+      "grad_norm": 0.025357285514473915,
+      "learning_rate": 4.5808124459809855e-05,
+      "loss": 0.2772,
+      "step": 4465
+    },
+    {
+      "epoch": 23.11741935483871,
+      "grad_norm": 0.022798912599682808,
+      "learning_rate": 4.577355229040622e-05,
+      "loss": 0.2837,
+      "step": 4466
+    },
+    {
+      "epoch": 23.12258064516129,
+      "grad_norm": 0.022715039551258087,
+      "learning_rate": 4.5738980121002596e-05,
+      "loss": 0.2857,
+      "step": 4467
+    },
+    {
+      "epoch": 23.127741935483872,
+      "grad_norm": 0.020643992349505424,
+      "learning_rate": 4.570440795159896e-05,
+      "loss": 0.266,
+      "step": 4468
+    },
+    {
+      "epoch": 23.13290322580645,
+      "grad_norm": 0.028402915224432945,
+      "learning_rate": 4.5669835782195337e-05,
+      "loss": 0.2964,
+      "step": 4469
+    },
+    {
+      "epoch": 23.13806451612903,
+      "grad_norm": 0.03131416067481041,
+      "learning_rate": 4.5635263612791704e-05,
+      "loss": 0.2574,
+      "step": 4470
+    },
+    {
+      "epoch": 23.143225806451614,
+      "grad_norm": 0.029550423845648766,
+      "learning_rate": 4.560069144338808e-05,
+      "loss": 0.3038,
+      "step": 4471
+    },
+    {
+      "epoch": 23.148387096774194,
+      "grad_norm": 0.026933692395687103,
+      "learning_rate": 4.5566119273984444e-05,
+      "loss": 0.2376,
+      "step": 4472
+    },
+    {
+      "epoch": 23.153548387096773,
+      "grad_norm": 0.029639961197972298,
+      "learning_rate": 4.553154710458081e-05,
+      "loss": 0.2907,
+      "step": 4473
+    },
+    {
+      "epoch": 23.158709677419356,
+      "grad_norm": 0.04075729101896286,
+      "learning_rate": 4.5496974935177185e-05,
+      "loss": 0.2895,
+      "step": 4474
+    },
+    {
+      "epoch": 23.163870967741936,
+      "grad_norm": 0.02657892554998398,
+      "learning_rate": 4.546240276577355e-05,
+      "loss": 0.2758,
+      "step": 4475
+    },
+    {
+      "epoch": 23.169032258064515,
+      "grad_norm": 0.025655878707766533,
+      "learning_rate": 4.5427830596369926e-05,
+      "loss": 0.2821,
+      "step": 4476
+    },
+    {
+      "epoch": 23.174193548387098,
+      "grad_norm": 0.026334214955568314,
+      "learning_rate": 4.539325842696629e-05,
+      "loss": 0.2974,
+      "step": 4477
+    },
+    {
+      "epoch": 23.179354838709678,
+      "grad_norm": 0.026016488671302795,
+      "learning_rate": 4.5358686257562667e-05,
+      "loss": 0.2922,
+      "step": 4478
+    },
+    {
+      "epoch": 23.184516129032257,
+      "grad_norm": 0.030983591452240944,
+      "learning_rate": 4.5324114088159034e-05,
+      "loss": 0.2733,
+      "step": 4479
+    },
+    {
+      "epoch": 23.18967741935484,
+      "grad_norm": 0.027822257950901985,
+      "learning_rate": 4.52895419187554e-05,
+      "loss": 0.295,
+      "step": 4480
+    },
+    {
+      "epoch": 23.19483870967742,
+      "grad_norm": 0.024885350838303566,
+      "learning_rate": 4.5254969749351774e-05,
+      "loss": 0.2815,
+      "step": 4481
+    },
+    {
+      "epoch": 23.2,
+      "grad_norm": 0.017586059868335724,
+      "learning_rate": 4.522039757994814e-05,
+      "loss": 0.3027,
+      "step": 4482
+    },
+    {
+      "epoch": 23.205161290322582,
+      "grad_norm": 0.01863211579620838,
+      "learning_rate": 4.5185825410544515e-05,
+      "loss": 0.2851,
+      "step": 4483
+    },
+    {
+      "epoch": 23.210322580645162,
+      "grad_norm": 0.03200831636786461,
+      "learning_rate": 4.515125324114088e-05,
+      "loss": 0.267,
+      "step": 4484
+    },
+    {
+      "epoch": 23.21548387096774,
+      "grad_norm": 0.01868685893714428,
+      "learning_rate": 4.5116681071737256e-05,
+      "loss": 0.2867,
+      "step": 4485
+    },
+    {
+      "epoch": 23.22064516129032,
+      "grad_norm": 0.026152106001973152,
+      "learning_rate": 4.508210890233362e-05,
+      "loss": 0.2654,
+      "step": 4486
+    },
+    {
+      "epoch": 23.225806451612904,
+      "grad_norm": 0.026241319254040718,
+      "learning_rate": 4.5047536732929997e-05,
+      "loss": 0.297,
+      "step": 4487
+    },
+    {
+      "epoch": 23.230967741935483,
+      "grad_norm": 0.03507125377655029,
+      "learning_rate": 4.5012964563526364e-05,
+      "loss": 0.3065,
+      "step": 4488
+    },
+    {
+      "epoch": 23.236129032258063,
+      "grad_norm": 0.036124248057603836,
+      "learning_rate": 4.497839239412273e-05,
+      "loss": 0.2826,
+      "step": 4489
+    },
+    {
+      "epoch": 23.241290322580646,
+      "grad_norm": 0.029703443869948387,
+      "learning_rate": 4.4943820224719104e-05,
+      "loss": 0.2912,
+      "step": 4490
+    },
+    {
+      "epoch": 23.246451612903225,
+      "grad_norm": 0.022915560752153397,
+      "learning_rate": 4.490924805531547e-05,
+      "loss": 0.2411,
+      "step": 4491
+    },
+    {
+      "epoch": 23.251612903225805,
+      "grad_norm": 0.01708982326090336,
+      "learning_rate": 4.4874675885911845e-05,
+      "loss": 0.3093,
+      "step": 4492
+    },
+    {
+      "epoch": 23.256774193548388,
+      "grad_norm": 0.03020479716360569,
+      "learning_rate": 4.484010371650821e-05,
+      "loss": 0.2952,
+      "step": 4493
+    },
+    {
+      "epoch": 23.261935483870968,
+      "grad_norm": 0.03103744238615036,
+      "learning_rate": 4.4805531547104586e-05,
+      "loss": 0.2509,
+      "step": 4494
+    },
+    {
+      "epoch": 23.267096774193547,
+      "grad_norm": 0.02467467449605465,
+      "learning_rate": 4.477095937770095e-05,
+      "loss": 0.2472,
+      "step": 4495
+    },
+    {
+      "epoch": 23.27225806451613,
+      "grad_norm": 0.026541024446487427,
+      "learning_rate": 4.473638720829732e-05,
+      "loss": 0.2647,
+      "step": 4496
+    },
+    {
+      "epoch": 23.27741935483871,
+      "grad_norm": 0.025952385738492012,
+      "learning_rate": 4.4701815038893693e-05,
+      "loss": 0.2603,
+      "step": 4497
+    },
+    {
+      "epoch": 23.28258064516129,
+      "grad_norm": 0.026181470602750778,
+      "learning_rate": 4.466724286949006e-05,
+      "loss": 0.2455,
+      "step": 4498
+    },
+    {
+      "epoch": 23.287741935483872,
+      "grad_norm": 0.026191117241978645,
+      "learning_rate": 4.4632670700086434e-05,
+      "loss": 0.2767,
+      "step": 4499
+    },
+    {
+      "epoch": 23.29290322580645,
+      "grad_norm": 0.024103214964270592,
+      "learning_rate": 4.45980985306828e-05,
+      "loss": 0.3082,
+      "step": 4500
+    },
+    {
+      "epoch": 23.29806451612903,
+      "grad_norm": 0.035395219922065735,
+      "learning_rate": 4.4563526361279175e-05,
+      "loss": 0.2891,
+      "step": 4501
+    },
+    {
+      "epoch": 23.303225806451614,
+      "grad_norm": 0.02255386672914028,
+      "learning_rate": 4.452895419187554e-05,
+      "loss": 0.3153,
+      "step": 4502
+    },
+    {
+      "epoch": 23.308387096774194,
+      "grad_norm": 0.028303589671850204,
+      "learning_rate": 4.4494382022471916e-05,
+      "loss": 0.3009,
+      "step": 4503
+    },
+    {
+      "epoch": 23.313548387096773,
+      "grad_norm": 0.026795249432325363,
+      "learning_rate": 4.445980985306828e-05,
+      "loss": 0.3027,
+      "step": 4504
+    },
+    {
+      "epoch": 23.318709677419356,
+      "grad_norm": 0.02896464616060257,
+      "learning_rate": 4.442523768366465e-05,
+      "loss": 0.2998,
+      "step": 4505
+    },
+    {
+      "epoch": 23.323870967741936,
+      "grad_norm": 0.03650444373488426,
+      "learning_rate": 4.4390665514261023e-05,
+      "loss": 0.271,
+      "step": 4506
+    },
+    {
+      "epoch": 23.329032258064515,
+      "grad_norm": 0.036284055560827255,
+      "learning_rate": 4.435609334485739e-05,
+      "loss": 0.2723,
+      "step": 4507
+    },
+    {
+      "epoch": 23.3341935483871,
+      "grad_norm": 0.022155331447720528,
+      "learning_rate": 4.4321521175453764e-05,
+      "loss": 0.2718,
+      "step": 4508
+    },
+    {
+      "epoch": 23.339354838709678,
+      "grad_norm": 0.023087259382009506,
+      "learning_rate": 4.428694900605013e-05,
+      "loss": 0.2905,
+      "step": 4509
+    },
+    {
+      "epoch": 23.344516129032257,
+      "grad_norm": 0.028332030400633812,
+      "learning_rate": 4.4252376836646505e-05,
+      "loss": 0.2904,
+      "step": 4510
+    },
+    {
+      "epoch": 23.34967741935484,
+      "grad_norm": 0.035183053463697433,
+      "learning_rate": 4.421780466724287e-05,
+      "loss": 0.2687,
+      "step": 4511
+    },
+    {
+      "epoch": 23.35483870967742,
+      "grad_norm": 0.02922380343079567,
+      "learning_rate": 4.4183232497839246e-05,
+      "loss": 0.289,
+      "step": 4512
+    },
+    {
+      "epoch": 23.36,
+      "grad_norm": 0.047262128442525864,
+      "learning_rate": 4.4148660328435606e-05,
+      "loss": 0.2895,
+      "step": 4513
+    },
+    {
+      "epoch": 23.36516129032258,
+      "grad_norm": 0.022557778283953667,
+      "learning_rate": 4.411408815903198e-05,
+      "loss": 0.3061,
+      "step": 4514
+    },
+    {
+      "epoch": 23.370322580645162,
+      "grad_norm": 0.030774420127272606,
+      "learning_rate": 4.4079515989628353e-05,
+      "loss": 0.2837,
+      "step": 4515
+    },
+    {
+      "epoch": 23.37548387096774,
+      "grad_norm": 0.029270000755786896,
+      "learning_rate": 4.404494382022472e-05,
+      "loss": 0.2732,
+      "step": 4516
+    },
+    {
+      "epoch": 23.38064516129032,
+      "grad_norm": 0.03570728376507759,
+      "learning_rate": 4.4010371650821094e-05,
+      "loss": 0.2822,
+      "step": 4517
+    },
+    {
+      "epoch": 23.385806451612904,
+      "grad_norm": 0.08316502720117569,
+      "learning_rate": 4.397579948141746e-05,
+      "loss": 0.3019,
+      "step": 4518
+    },
+    {
+      "epoch": 23.390967741935484,
+      "grad_norm": 0.0197465680539608,
+      "learning_rate": 4.3941227312013835e-05,
+      "loss": 0.2671,
+      "step": 4519
+    },
+    {
+      "epoch": 23.396129032258063,
+      "grad_norm": 0.026749471202492714,
+      "learning_rate": 4.39066551426102e-05,
+      "loss": 0.2575,
+      "step": 4520
+    },
+    {
+      "epoch": 23.401290322580646,
+      "grad_norm": 0.021129177883267403,
+      "learning_rate": 4.387208297320657e-05,
+      "loss": 0.2789,
+      "step": 4521
+    },
+    {
+      "epoch": 23.406451612903226,
+      "grad_norm": 0.029019273817539215,
+      "learning_rate": 4.3837510803802936e-05,
+      "loss": 0.2832,
+      "step": 4522
+    },
+    {
+      "epoch": 23.411612903225805,
+      "grad_norm": 0.018306231126189232,
+      "learning_rate": 4.380293863439931e-05,
+      "loss": 0.3025,
+      "step": 4523
+    },
+    {
+      "epoch": 23.416774193548388,
+      "grad_norm": 0.036178458482027054,
+      "learning_rate": 4.3768366464995683e-05,
+      "loss": 0.26,
+      "step": 4524
+    },
+    {
+      "epoch": 23.421935483870968,
+      "grad_norm": 0.018910247832536697,
+      "learning_rate": 4.373379429559205e-05,
+      "loss": 0.2947,
+      "step": 4525
+    },
+    {
+      "epoch": 23.427096774193547,
+      "grad_norm": 0.025569172576069832,
+      "learning_rate": 4.3699222126188424e-05,
+      "loss": 0.2748,
+      "step": 4526
+    },
+    {
+      "epoch": 23.43225806451613,
+      "grad_norm": 0.02849627658724785,
+      "learning_rate": 4.366464995678479e-05,
+      "loss": 0.3076,
+      "step": 4527
+    },
+    {
+      "epoch": 23.43741935483871,
+      "grad_norm": 0.028210703283548355,
+      "learning_rate": 4.3630077787381165e-05,
+      "loss": 0.255,
+      "step": 4528
+    },
+    {
+      "epoch": 23.44258064516129,
+      "grad_norm": 0.032352909445762634,
+      "learning_rate": 4.3595505617977525e-05,
+      "loss": 0.2942,
+      "step": 4529
+    },
+    {
+      "epoch": 23.447741935483872,
+      "grad_norm": 0.025657789781689644,
+      "learning_rate": 4.35609334485739e-05,
+      "loss": 0.2847,
+      "step": 4530
+    },
+    {
+      "epoch": 23.452903225806452,
+      "grad_norm": 0.0313183069229126,
+      "learning_rate": 4.3526361279170266e-05,
+      "loss": 0.2871,
+      "step": 4531
+    },
+    {
+      "epoch": 23.45806451612903,
+      "grad_norm": 0.026510998606681824,
+      "learning_rate": 4.349178910976664e-05,
+      "loss": 0.2967,
+      "step": 4532
+    },
+    {
+      "epoch": 23.463225806451614,
+      "grad_norm": 0.02029336988925934,
+      "learning_rate": 4.345721694036301e-05,
+      "loss": 0.2791,
+      "step": 4533
+    },
+    {
+      "epoch": 23.468387096774194,
+      "grad_norm": 0.03399164602160454,
+      "learning_rate": 4.342264477095938e-05,
+      "loss": 0.272,
+      "step": 4534
+    },
+    {
+      "epoch": 23.473548387096773,
+      "grad_norm": 0.027786124497652054,
+      "learning_rate": 4.3388072601555754e-05,
+      "loss": 0.2695,
+      "step": 4535
+    },
+    {
+      "epoch": 23.478709677419356,
+      "grad_norm": 0.03950401023030281,
+      "learning_rate": 4.335350043215212e-05,
+      "loss": 0.2595,
+      "step": 4536
+    },
+    {
+      "epoch": 23.483870967741936,
+      "grad_norm": 0.017367541790008545,
+      "learning_rate": 4.331892826274849e-05,
+      "loss": 0.2861,
+      "step": 4537
+    },
+    {
+      "epoch": 23.489032258064515,
+      "grad_norm": 0.023790795356035233,
+      "learning_rate": 4.3284356093344855e-05,
+      "loss": 0.2677,
+      "step": 4538
+    },
+    {
+      "epoch": 23.4941935483871,
+      "grad_norm": 0.03633527457714081,
+      "learning_rate": 4.324978392394123e-05,
+      "loss": 0.2977,
+      "step": 4539
+    },
+    {
+      "epoch": 23.499354838709678,
+      "grad_norm": 0.02475871331989765,
+      "learning_rate": 4.3215211754537596e-05,
+      "loss": 0.2765,
+      "step": 4540
+    },
+    {
+      "epoch": 23.504516129032258,
+      "grad_norm": 0.022029077634215355,
+      "learning_rate": 4.318063958513397e-05,
+      "loss": 0.2915,
+      "step": 4541
+    },
+    {
+      "epoch": 23.509677419354837,
+      "grad_norm": 0.037139371037483215,
+      "learning_rate": 4.314606741573034e-05,
+      "loss": 0.2518,
+      "step": 4542
+    },
+    {
+      "epoch": 23.51483870967742,
+      "grad_norm": 0.02594936266541481,
+      "learning_rate": 4.311149524632671e-05,
+      "loss": 0.2925,
+      "step": 4543
+    },
+    {
+      "epoch": 23.52,
+      "grad_norm": 0.02703866735100746,
+      "learning_rate": 4.3076923076923084e-05,
+      "loss": 0.2906,
+      "step": 4544
+    },
+    {
+      "epoch": 23.52516129032258,
+      "grad_norm": 0.01931145414710045,
+      "learning_rate": 4.3042350907519444e-05,
+      "loss": 0.2699,
+      "step": 4545
+    },
+    {
+      "epoch": 23.530322580645162,
+      "grad_norm": 0.025654751807451248,
+      "learning_rate": 4.300777873811582e-05,
+      "loss": 0.3049,
+      "step": 4546
+    },
+    {
+      "epoch": 23.53548387096774,
+      "grad_norm": 0.02864878810942173,
+      "learning_rate": 4.2973206568712185e-05,
+      "loss": 0.2948,
+      "step": 4547
+    },
+    {
+      "epoch": 23.54064516129032,
+      "grad_norm": 0.048496633768081665,
+      "learning_rate": 4.293863439930856e-05,
+      "loss": 0.3077,
+      "step": 4548
+    },
+    {
+      "epoch": 23.545806451612904,
+      "grad_norm": 0.050801388919353485,
+      "learning_rate": 4.2904062229904926e-05,
+      "loss": 0.2329,
+      "step": 4549
+    },
+    {
+      "epoch": 23.550967741935484,
+      "grad_norm": 0.03256988152861595,
+      "learning_rate": 4.28694900605013e-05,
+      "loss": 0.2959,
+      "step": 4550
+    },
+    {
+      "epoch": 23.556129032258063,
+      "grad_norm": 0.031233780086040497,
+      "learning_rate": 4.283491789109767e-05,
+      "loss": 0.2876,
+      "step": 4551
+    },
+    {
+      "epoch": 23.561290322580646,
+      "grad_norm": 0.0289958156645298,
+      "learning_rate": 4.280034572169404e-05,
+      "loss": 0.2846,
+      "step": 4552
+    },
+    {
+      "epoch": 23.566451612903226,
+      "grad_norm": 0.02849949710071087,
+      "learning_rate": 4.276577355229041e-05,
+      "loss": 0.2821,
+      "step": 4553
+    },
+    {
+      "epoch": 23.571612903225805,
+      "grad_norm": 0.027441425248980522,
+      "learning_rate": 4.2731201382886774e-05,
+      "loss": 0.2223,
+      "step": 4554
+    },
+    {
+      "epoch": 23.57677419354839,
+      "grad_norm": 0.02399890124797821,
+      "learning_rate": 4.269662921348315e-05,
+      "loss": 0.3071,
+      "step": 4555
+    },
+    {
+      "epoch": 23.581935483870968,
+      "grad_norm": 0.03607773408293724,
+      "learning_rate": 4.2662057044079515e-05,
+      "loss": 0.2631,
+      "step": 4556
+    },
+    {
+      "epoch": 23.587096774193547,
+      "grad_norm": 0.017658231779932976,
+      "learning_rate": 4.262748487467589e-05,
+      "loss": 0.2763,
+      "step": 4557
+    },
+    {
+      "epoch": 23.59225806451613,
+      "grad_norm": 0.04924013838171959,
+      "learning_rate": 4.2592912705272256e-05,
+      "loss": 0.2661,
+      "step": 4558
+    },
+    {
+      "epoch": 23.59741935483871,
+      "grad_norm": 0.020522287115454674,
+      "learning_rate": 4.255834053586863e-05,
+      "loss": 0.2743,
+      "step": 4559
+    },
+    {
+      "epoch": 23.60258064516129,
+      "grad_norm": 0.026417868211865425,
+      "learning_rate": 4.2523768366465e-05,
+      "loss": 0.2588,
+      "step": 4560
+    },
+    {
+      "epoch": 23.607741935483872,
+      "grad_norm": 0.045558493584394455,
+      "learning_rate": 4.2489196197061364e-05,
+      "loss": 0.2881,
+      "step": 4561
+    },
+    {
+      "epoch": 23.612903225806452,
+      "grad_norm": 0.054303739219903946,
+      "learning_rate": 4.245462402765774e-05,
+      "loss": 0.2935,
+      "step": 4562
+    },
+    {
+      "epoch": 23.61806451612903,
+      "grad_norm": 0.03542930260300636,
+      "learning_rate": 4.2420051858254104e-05,
+      "loss": 0.276,
+      "step": 4563
+    },
+    {
+      "epoch": 23.623225806451615,
+      "grad_norm": 0.04111995920538902,
+      "learning_rate": 4.238547968885048e-05,
+      "loss": 0.2585,
+      "step": 4564
+    },
+    {
+      "epoch": 23.628387096774194,
+      "grad_norm": 0.04764649271965027,
+      "learning_rate": 4.2350907519446845e-05,
+      "loss": 0.2695,
+      "step": 4565
+    },
+    {
+      "epoch": 23.633548387096774,
+      "grad_norm": 0.02613864652812481,
+      "learning_rate": 4.231633535004322e-05,
+      "loss": 0.2704,
+      "step": 4566
+    },
+    {
+      "epoch": 23.638709677419357,
+      "grad_norm": 0.021038616076111794,
+      "learning_rate": 4.2281763180639586e-05,
+      "loss": 0.3018,
+      "step": 4567
+    },
+    {
+      "epoch": 23.643870967741936,
+      "grad_norm": 0.027142874896526337,
+      "learning_rate": 4.224719101123596e-05,
+      "loss": 0.2895,
+      "step": 4568
+    },
+    {
+      "epoch": 23.649032258064516,
+      "grad_norm": 0.028704458847641945,
+      "learning_rate": 4.2212618841832326e-05,
+      "loss": 0.3016,
+      "step": 4569
+    },
+    {
+      "epoch": 23.654193548387095,
+      "grad_norm": 0.03075498528778553,
+      "learning_rate": 4.2178046672428693e-05,
+      "loss": 0.2821,
+      "step": 4570
+    },
+    {
+      "epoch": 23.659354838709678,
+      "grad_norm": 0.03029526211321354,
+      "learning_rate": 4.214347450302507e-05,
+      "loss": 0.257,
+      "step": 4571
+    },
+    {
+      "epoch": 23.664516129032258,
+      "grad_norm": 0.04832185432314873,
+      "learning_rate": 4.2108902333621434e-05,
+      "loss": 0.3128,
+      "step": 4572
+    },
+    {
+      "epoch": 23.669677419354837,
+      "grad_norm": 0.05157089978456497,
+      "learning_rate": 4.207433016421781e-05,
+      "loss": 0.3101,
+      "step": 4573
+    },
+    {
+      "epoch": 23.67483870967742,
+      "grad_norm": 0.03300579637289047,
+      "learning_rate": 4.2039757994814175e-05,
+      "loss": 0.3033,
+      "step": 4574
+    },
+    {
+      "epoch": 23.68,
+      "grad_norm": 0.025897029787302017,
+      "learning_rate": 4.200518582541055e-05,
+      "loss": 0.2988,
+      "step": 4575
+    },
+    {
+      "epoch": 23.68516129032258,
+      "grad_norm": 0.026845164597034454,
+      "learning_rate": 4.1970613656006916e-05,
+      "loss": 0.273,
+      "step": 4576
+    },
+    {
+      "epoch": 23.690322580645162,
+      "grad_norm": 0.01715543493628502,
+      "learning_rate": 4.193604148660329e-05,
+      "loss": 0.2819,
+      "step": 4577
+    },
+    {
+      "epoch": 23.695483870967742,
+      "grad_norm": 0.02313525602221489,
+      "learning_rate": 4.1901469317199656e-05,
+      "loss": 0.2482,
+      "step": 4578
+    },
+    {
+      "epoch": 23.70064516129032,
+      "grad_norm": 0.03727136179804802,
+      "learning_rate": 4.1866897147796023e-05,
+      "loss": 0.3079,
+      "step": 4579
+    },
+    {
+      "epoch": 23.705806451612904,
+      "grad_norm": 0.02697007544338703,
+      "learning_rate": 4.18323249783924e-05,
+      "loss": 0.2938,
+      "step": 4580
+    },
+    {
+      "epoch": 23.710967741935484,
+      "grad_norm": 0.01870059035718441,
+      "learning_rate": 4.1797752808988764e-05,
+      "loss": 0.2294,
+      "step": 4581
+    },
+    {
+      "epoch": 23.716129032258063,
+      "grad_norm": 0.02977440133690834,
+      "learning_rate": 4.176318063958514e-05,
+      "loss": 0.3084,
+      "step": 4582
+    },
+    {
+      "epoch": 23.721290322580646,
+      "grad_norm": 0.02491171285510063,
+      "learning_rate": 4.1728608470181505e-05,
+      "loss": 0.2692,
+      "step": 4583
+    },
+    {
+      "epoch": 23.726451612903226,
+      "grad_norm": 0.03325429558753967,
+      "learning_rate": 4.169403630077788e-05,
+      "loss": 0.2646,
+      "step": 4584
+    },
+    {
+      "epoch": 23.731612903225805,
+      "grad_norm": 0.03841523826122284,
+      "learning_rate": 4.1659464131374246e-05,
+      "loss": 0.2958,
+      "step": 4585
+    },
+    {
+      "epoch": 23.73677419354839,
+      "grad_norm": 0.03515361621975899,
+      "learning_rate": 4.162489196197061e-05,
+      "loss": 0.2966,
+      "step": 4586
+    },
+    {
+      "epoch": 23.741935483870968,
+      "grad_norm": 0.024720823392271996,
+      "learning_rate": 4.1590319792566986e-05,
+      "loss": 0.2921,
+      "step": 4587
+    },
+    {
+      "epoch": 23.747096774193547,
+      "grad_norm": 0.03258386626839638,
+      "learning_rate": 4.1555747623163353e-05,
+      "loss": 0.2709,
+      "step": 4588
+    },
+    {
+      "epoch": 23.75225806451613,
+      "grad_norm": 0.03188784047961235,
+      "learning_rate": 4.152117545375973e-05,
+      "loss": 0.2955,
+      "step": 4589
+    },
+    {
+      "epoch": 23.75741935483871,
+      "grad_norm": 0.024496952071785927,
+      "learning_rate": 4.1486603284356094e-05,
+      "loss": 0.2986,
+      "step": 4590
+    },
+    {
+      "epoch": 23.76258064516129,
+      "grad_norm": 0.03138714283704758,
+      "learning_rate": 4.145203111495247e-05,
+      "loss": 0.2867,
+      "step": 4591
+    },
+    {
+      "epoch": 23.767741935483873,
+      "grad_norm": 0.035272929817438126,
+      "learning_rate": 4.1417458945548835e-05,
+      "loss": 0.2748,
+      "step": 4592
+    },
+    {
+      "epoch": 23.772903225806452,
+      "grad_norm": 0.03560638055205345,
+      "learning_rate": 4.138288677614521e-05,
+      "loss": 0.2884,
+      "step": 4593
+    },
+    {
+      "epoch": 23.77806451612903,
+      "grad_norm": 0.027641797438263893,
+      "learning_rate": 4.1348314606741576e-05,
+      "loss": 0.2523,
+      "step": 4594
+    },
+    {
+      "epoch": 23.78322580645161,
+      "grad_norm": 0.034015215933322906,
+      "learning_rate": 4.131374243733794e-05,
+      "loss": 0.3004,
+      "step": 4595
+    },
+    {
+      "epoch": 23.788387096774194,
+      "grad_norm": 0.03625859320163727,
+      "learning_rate": 4.1279170267934316e-05,
+      "loss": 0.281,
+      "step": 4596
+    },
+    {
+      "epoch": 23.793548387096774,
+      "grad_norm": 0.014648028649389744,
+      "learning_rate": 4.124459809853068e-05,
+      "loss": 0.2917,
+      "step": 4597
+    },
+    {
+      "epoch": 23.798709677419353,
+      "grad_norm": 0.02682194672524929,
+      "learning_rate": 4.121002592912706e-05,
+      "loss": 0.2821,
+      "step": 4598
+    },
+    {
+      "epoch": 23.803870967741936,
+      "grad_norm": 0.03149648755788803,
+      "learning_rate": 4.1175453759723424e-05,
+      "loss": 0.2786,
+      "step": 4599
+    },
+    {
+      "epoch": 23.809032258064516,
+      "grad_norm": 0.022352274507284164,
+      "learning_rate": 4.11408815903198e-05,
+      "loss": 0.2868,
+      "step": 4600
+    },
+    {
+      "epoch": 23.814193548387095,
+      "grad_norm": 0.017300507053732872,
+      "learning_rate": 4.1106309420916165e-05,
+      "loss": 0.2521,
+      "step": 4601
+    },
+    {
+      "epoch": 23.81935483870968,
+      "grad_norm": 0.03211958333849907,
+      "learning_rate": 4.107173725151253e-05,
+      "loss": 0.2719,
+      "step": 4602
+    },
+    {
+      "epoch": 23.824516129032258,
+      "grad_norm": 0.03076106496155262,
+      "learning_rate": 4.1037165082108906e-05,
+      "loss": 0.2752,
+      "step": 4603
+    },
+    {
+      "epoch": 23.829677419354837,
+      "grad_norm": 0.023528829216957092,
+      "learning_rate": 4.100259291270527e-05,
+      "loss": 0.277,
+      "step": 4604
+    },
+    {
+      "epoch": 23.83483870967742,
+      "grad_norm": 0.031862374395132065,
+      "learning_rate": 4.0968020743301646e-05,
+      "loss": 0.2746,
+      "step": 4605
+    },
+    {
+      "epoch": 23.84,
+      "grad_norm": 0.028995363041758537,
+      "learning_rate": 4.093344857389801e-05,
+      "loss": 0.2316,
+      "step": 4606
+    },
+    {
+      "epoch": 23.84516129032258,
+      "grad_norm": 0.026143595576286316,
+      "learning_rate": 4.089887640449439e-05,
+      "loss": 0.2861,
+      "step": 4607
+    },
+    {
+      "epoch": 23.850322580645162,
+      "grad_norm": 0.02908845618367195,
+      "learning_rate": 4.0864304235090754e-05,
+      "loss": 0.2781,
+      "step": 4608
+    },
+    {
+      "epoch": 23.855483870967742,
+      "grad_norm": 0.0290109533816576,
+      "learning_rate": 4.082973206568713e-05,
+      "loss": 0.2955,
+      "step": 4609
+    },
+    {
+      "epoch": 23.86064516129032,
+      "grad_norm": 0.037289783358573914,
+      "learning_rate": 4.079515989628349e-05,
+      "loss": 0.2925,
+      "step": 4610
+    },
+    {
+      "epoch": 23.865806451612904,
+      "grad_norm": 0.027937376871705055,
+      "learning_rate": 4.076058772687986e-05,
+      "loss": 0.3013,
+      "step": 4611
+    },
+    {
+      "epoch": 23.870967741935484,
+      "grad_norm": 0.026025138795375824,
+      "learning_rate": 4.0726015557476236e-05,
+      "loss": 0.2883,
+      "step": 4612
+    },
+    {
+      "epoch": 23.876129032258063,
+      "grad_norm": 0.02295156940817833,
+      "learning_rate": 4.06914433880726e-05,
+      "loss": 0.2316,
+      "step": 4613
+    },
+    {
+      "epoch": 23.881290322580647,
+      "grad_norm": 0.07883531600236893,
+      "learning_rate": 4.0656871218668976e-05,
+      "loss": 0.2798,
+      "step": 4614
+    },
+    {
+      "epoch": 23.886451612903226,
+      "grad_norm": 0.03291665390133858,
+      "learning_rate": 4.062229904926534e-05,
+      "loss": 0.2852,
+      "step": 4615
+    },
+    {
+      "epoch": 23.891612903225806,
+      "grad_norm": 0.02256193198263645,
+      "learning_rate": 4.058772687986172e-05,
+      "loss": 0.2617,
+      "step": 4616
+    },
+    {
+      "epoch": 23.89677419354839,
+      "grad_norm": 0.026634113863110542,
+      "learning_rate": 4.0553154710458084e-05,
+      "loss": 0.2729,
+      "step": 4617
+    },
+    {
+      "epoch": 23.901935483870968,
+      "grad_norm": 0.03301660344004631,
+      "learning_rate": 4.051858254105445e-05,
+      "loss": 0.299,
+      "step": 4618
+    },
+    {
+      "epoch": 23.907096774193548,
+      "grad_norm": 0.03278056159615517,
+      "learning_rate": 4.048401037165082e-05,
+      "loss": 0.2826,
+      "step": 4619
+    },
+    {
+      "epoch": 23.91225806451613,
+      "grad_norm": 0.02332090586423874,
+      "learning_rate": 4.044943820224719e-05,
+      "loss": 0.297,
+      "step": 4620
+    },
+    {
+      "epoch": 23.91741935483871,
+      "grad_norm": 0.03341955319046974,
+      "learning_rate": 4.0414866032843566e-05,
+      "loss": 0.2574,
+      "step": 4621
+    },
+    {
+      "epoch": 23.92258064516129,
+      "grad_norm": 0.027706757187843323,
+      "learning_rate": 4.038029386343993e-05,
+      "loss": 0.3041,
+      "step": 4622
+    },
+    {
+      "epoch": 23.927741935483873,
+      "grad_norm": 0.030081667006015778,
+      "learning_rate": 4.0345721694036306e-05,
+      "loss": 0.3026,
+      "step": 4623
+    },
+    {
+      "epoch": 23.932903225806452,
+      "grad_norm": 0.0270834993571043,
+      "learning_rate": 4.031114952463267e-05,
+      "loss": 0.2786,
+      "step": 4624
+    },
+    {
+      "epoch": 23.93806451612903,
+      "grad_norm": 0.03246445208787918,
+      "learning_rate": 4.027657735522905e-05,
+      "loss": 0.3011,
+      "step": 4625
+    },
+    {
+      "epoch": 23.94322580645161,
+      "grad_norm": 0.0371626652777195,
+      "learning_rate": 4.024200518582541e-05,
+      "loss": 0.2692,
+      "step": 4626
+    },
+    {
+      "epoch": 23.948387096774194,
+      "grad_norm": 0.03453659266233444,
+      "learning_rate": 4.020743301642178e-05,
+      "loss": 0.2817,
+      "step": 4627
+    },
+    {
+      "epoch": 23.953548387096774,
+      "grad_norm": 0.02738458849489689,
+      "learning_rate": 4.017286084701815e-05,
+      "loss": 0.2684,
+      "step": 4628
+    },
+    {
+      "epoch": 23.958709677419353,
+      "grad_norm": 0.03504529967904091,
+      "learning_rate": 4.013828867761452e-05,
+      "loss": 0.2816,
+      "step": 4629
+    },
+    {
+      "epoch": 23.963870967741936,
+      "grad_norm": 0.045265864580869675,
+      "learning_rate": 4.0103716508210896e-05,
+      "loss": 0.3103,
+      "step": 4630
+    },
+    {
+      "epoch": 23.969032258064516,
+      "grad_norm": 0.027424948289990425,
+      "learning_rate": 4.006914433880726e-05,
+      "loss": 0.2777,
+      "step": 4631
+    },
+    {
+      "epoch": 23.974193548387095,
+      "grad_norm": 0.039359722286462784,
+      "learning_rate": 4.0034572169403636e-05,
+      "loss": 0.3008,
+      "step": 4632
+    },
+    {
+      "epoch": 23.97935483870968,
+      "grad_norm": 0.026595979928970337,
+      "learning_rate": 4e-05,
+      "loss": 0.271,
+      "step": 4633
+    },
+    {
+      "epoch": 23.984516129032258,
+      "grad_norm": 0.019551947712898254,
+      "learning_rate": 3.996542783059637e-05,
+      "loss": 0.2997,
+      "step": 4634
+    },
+    {
+      "epoch": 23.989677419354837,
+      "grad_norm": 0.06167061626911163,
+      "learning_rate": 3.993085566119274e-05,
+      "loss": 0.2678,
+      "step": 4635
+    },
+    {
+      "epoch": 23.99483870967742,
+      "grad_norm": 0.02485267072916031,
+      "learning_rate": 3.989628349178911e-05,
+      "loss": 0.2786,
+      "step": 4636
+    },
+    {
+      "epoch": 24.001290322580644,
+      "grad_norm": 0.06756655871868134,
+      "learning_rate": 3.986171132238548e-05,
+      "loss": 0.4289,
+      "step": 4637
+    },
+    {
+      "epoch": 24.006451612903227,
+      "grad_norm": 0.02605280838906765,
+      "learning_rate": 3.982713915298185e-05,
+      "loss": 0.2997,
+      "step": 4638
+    },
+    {
+      "epoch": 24.011612903225807,
+      "grad_norm": 0.02208350971341133,
+      "learning_rate": 3.9792566983578226e-05,
+      "loss": 0.2449,
+      "step": 4639
+    },
+    {
+      "epoch": 24.016774193548386,
+      "grad_norm": 0.016653843224048615,
+      "learning_rate": 3.975799481417459e-05,
+      "loss": 0.2939,
+      "step": 4640
+    },
+    {
+      "epoch": 24.016774193548386,
+      "eval_loss": 3.2900354862213135,
+      "eval_runtime": 21.1921,
+      "eval_samples_per_second": 3.728,
+      "eval_steps_per_second": 0.472,
+      "step": 4640
+    },
+    {
+      "epoch": 24.02193548387097,
+      "grad_norm": 0.02609771117568016,
+      "learning_rate": 3.9723422644770966e-05,
+      "loss": 0.2665,
+      "step": 4641
+    },
+    {
+      "epoch": 24.02709677419355,
+      "grad_norm": 0.022001883015036583,
+      "learning_rate": 3.9688850475367326e-05,
+      "loss": 0.2979,
+      "step": 4642
+    },
+    {
+      "epoch": 24.032258064516128,
+      "grad_norm": 0.022489281371235847,
+      "learning_rate": 3.96542783059637e-05,
+      "loss": 0.2817,
+      "step": 4643
+    },
+    {
+      "epoch": 24.03741935483871,
+      "grad_norm": 0.01841900497674942,
+      "learning_rate": 3.961970613656007e-05,
+      "loss": 0.2657,
+      "step": 4644
+    },
+    {
+      "epoch": 24.04258064516129,
+      "grad_norm": 0.020545780658721924,
+      "learning_rate": 3.958513396715644e-05,
+      "loss": 0.2919,
+      "step": 4645
+    },
+    {
+      "epoch": 24.04774193548387,
+      "grad_norm": 0.024092670530080795,
+      "learning_rate": 3.955056179775281e-05,
+      "loss": 0.2656,
+      "step": 4646
+    },
+    {
+      "epoch": 24.052903225806453,
+      "grad_norm": 0.02349112369120121,
+      "learning_rate": 3.951598962834918e-05,
+      "loss": 0.3082,
+      "step": 4647
+    },
+    {
+      "epoch": 24.058064516129033,
+      "grad_norm": 0.020418843254446983,
+      "learning_rate": 3.9481417458945555e-05,
+      "loss": 0.2846,
+      "step": 4648
+    },
+    {
+      "epoch": 24.063225806451612,
+      "grad_norm": 0.03126358613371849,
+      "learning_rate": 3.944684528954192e-05,
+      "loss": 0.2949,
+      "step": 4649
+    },
+    {
+      "epoch": 24.068387096774195,
+      "grad_norm": 0.03459442779421806,
+      "learning_rate": 3.9412273120138296e-05,
+      "loss": 0.2833,
+      "step": 4650
+    },
+    {
+      "epoch": 24.073548387096775,
+      "grad_norm": 0.022487018257379532,
+      "learning_rate": 3.9377700950734656e-05,
+      "loss": 0.3031,
+      "step": 4651
+    },
+    {
+      "epoch": 24.078709677419354,
+      "grad_norm": 0.019192297011613846,
+      "learning_rate": 3.934312878133103e-05,
+      "loss": 0.2589,
+      "step": 4652
+    },
+    {
+      "epoch": 24.083870967741934,
+      "grad_norm": 0.024340255185961723,
+      "learning_rate": 3.93085566119274e-05,
+      "loss": 0.3106,
+      "step": 4653
+    },
+    {
+      "epoch": 24.089032258064517,
+      "grad_norm": 0.023865967988967896,
+      "learning_rate": 3.927398444252377e-05,
+      "loss": 0.2863,
+      "step": 4654
+    },
+    {
+      "epoch": 24.094193548387096,
+      "grad_norm": 0.025856509804725647,
+      "learning_rate": 3.923941227312014e-05,
+      "loss": 0.284,
+      "step": 4655
+    },
+    {
+      "epoch": 24.099354838709676,
+      "grad_norm": 0.018780669197440147,
+      "learning_rate": 3.920484010371651e-05,
+      "loss": 0.2759,
+      "step": 4656
+    },
+    {
+      "epoch": 24.10451612903226,
+      "grad_norm": 0.025720687583088875,
+      "learning_rate": 3.917026793431288e-05,
+      "loss": 0.2685,
+      "step": 4657
+    },
+    {
+      "epoch": 24.10967741935484,
+      "grad_norm": 0.04230986163020134,
+      "learning_rate": 3.913569576490925e-05,
+      "loss": 0.2918,
+      "step": 4658
+    },
+    {
+      "epoch": 24.114838709677418,
+      "grad_norm": 0.022553395479917526,
+      "learning_rate": 3.910112359550562e-05,
+      "loss": 0.2554,
+      "step": 4659
+    },
+    {
+      "epoch": 24.12,
+      "grad_norm": 0.027823330834507942,
+      "learning_rate": 3.9066551426101986e-05,
+      "loss": 0.2607,
+      "step": 4660
+    },
+    {
+      "epoch": 24.12516129032258,
+      "grad_norm": 0.02621481753885746,
+      "learning_rate": 3.903197925669836e-05,
+      "loss": 0.3003,
+      "step": 4661
+    },
+    {
+      "epoch": 24.13032258064516,
+      "grad_norm": 0.02598949894309044,
+      "learning_rate": 3.899740708729473e-05,
+      "loss": 0.2633,
+      "step": 4662
+    },
+    {
+      "epoch": 24.135483870967743,
+      "grad_norm": 0.028216052800416946,
+      "learning_rate": 3.89628349178911e-05,
+      "loss": 0.2868,
+      "step": 4663
+    },
+    {
+      "epoch": 24.140645161290323,
+      "grad_norm": 0.02465706132352352,
+      "learning_rate": 3.892826274848747e-05,
+      "loss": 0.2589,
+      "step": 4664
+    },
+    {
+      "epoch": 24.145806451612902,
+      "grad_norm": 0.030888205394148827,
+      "learning_rate": 3.889369057908384e-05,
+      "loss": 0.2914,
+      "step": 4665
+    },
+    {
+      "epoch": 24.150967741935485,
+      "grad_norm": 0.02769562229514122,
+      "learning_rate": 3.885911840968021e-05,
+      "loss": 0.2674,
+      "step": 4666
+    },
+    {
+      "epoch": 24.156129032258065,
+      "grad_norm": 0.02157648652791977,
+      "learning_rate": 3.8824546240276576e-05,
+      "loss": 0.2648,
+      "step": 4667
+    },
+    {
+      "epoch": 24.161290322580644,
+      "grad_norm": 0.021434837952256203,
+      "learning_rate": 3.878997407087295e-05,
+      "loss": 0.2833,
+      "step": 4668
+    },
+    {
+      "epoch": 24.166451612903227,
+      "grad_norm": 0.020740414038300514,
+      "learning_rate": 3.8755401901469316e-05,
+      "loss": 0.2914,
+      "step": 4669
+    },
+    {
+      "epoch": 24.171612903225807,
+      "grad_norm": 0.03375081345438957,
+      "learning_rate": 3.872082973206569e-05,
+      "loss": 0.2837,
+      "step": 4670
+    },
+    {
+      "epoch": 24.176774193548386,
+      "grad_norm": 0.01912795938551426,
+      "learning_rate": 3.868625756266206e-05,
+      "loss": 0.2779,
+      "step": 4671
+    },
+    {
+      "epoch": 24.18193548387097,
+      "grad_norm": 0.02308102697134018,
+      "learning_rate": 3.865168539325843e-05,
+      "loss": 0.2701,
+      "step": 4672
+    },
+    {
+      "epoch": 24.18709677419355,
+      "grad_norm": 0.02449401654303074,
+      "learning_rate": 3.86171132238548e-05,
+      "loss": 0.2945,
+      "step": 4673
+    },
+    {
+      "epoch": 24.19225806451613,
+      "grad_norm": 0.03184036538004875,
+      "learning_rate": 3.858254105445117e-05,
+      "loss": 0.2654,
+      "step": 4674
+    },
+    {
+      "epoch": 24.19741935483871,
+      "grad_norm": 0.026937218382954597,
+      "learning_rate": 3.854796888504754e-05,
+      "loss": 0.2558,
+      "step": 4675
+    },
+    {
+      "epoch": 24.20258064516129,
+      "grad_norm": 0.02631545066833496,
+      "learning_rate": 3.8513396715643906e-05,
+      "loss": 0.2838,
+      "step": 4676
+    },
+    {
+      "epoch": 24.20774193548387,
+      "grad_norm": 0.022608008235692978,
+      "learning_rate": 3.847882454624028e-05,
+      "loss": 0.2892,
+      "step": 4677
+    },
+    {
+      "epoch": 24.21290322580645,
+      "grad_norm": 0.023954281583428383,
+      "learning_rate": 3.8444252376836646e-05,
+      "loss": 0.2965,
+      "step": 4678
+    },
+    {
+      "epoch": 24.218064516129033,
+      "grad_norm": 0.078207828104496,
+      "learning_rate": 3.840968020743302e-05,
+      "loss": 0.3003,
+      "step": 4679
+    },
+    {
+      "epoch": 24.223225806451612,
+      "grad_norm": 0.016731860116124153,
+      "learning_rate": 3.837510803802939e-05,
+      "loss": 0.2894,
+      "step": 4680
+    },
+    {
+      "epoch": 24.228387096774192,
+      "grad_norm": 0.02804030105471611,
+      "learning_rate": 3.834053586862576e-05,
+      "loss": 0.2493,
+      "step": 4681
+    },
+    {
+      "epoch": 24.233548387096775,
+      "grad_norm": 0.028191395103931427,
+      "learning_rate": 3.830596369922213e-05,
+      "loss": 0.2671,
+      "step": 4682
+    },
+    {
+      "epoch": 24.238709677419354,
+      "grad_norm": 0.03099088929593563,
+      "learning_rate": 3.8271391529818495e-05,
+      "loss": 0.2846,
+      "step": 4683
+    },
+    {
+      "epoch": 24.243870967741934,
+      "grad_norm": 0.030788784846663475,
+      "learning_rate": 3.823681936041487e-05,
+      "loss": 0.3043,
+      "step": 4684
+    },
+    {
+      "epoch": 24.249032258064517,
+      "grad_norm": 0.025580085813999176,
+      "learning_rate": 3.8202247191011236e-05,
+      "loss": 0.2925,
+      "step": 4685
+    },
+    {
+      "epoch": 24.254193548387097,
+      "grad_norm": 0.019689228385686874,
+      "learning_rate": 3.816767502160761e-05,
+      "loss": 0.2848,
+      "step": 4686
+    },
+    {
+      "epoch": 24.259354838709676,
+      "grad_norm": 0.026808228343725204,
+      "learning_rate": 3.8133102852203976e-05,
+      "loss": 0.2921,
+      "step": 4687
+    },
+    {
+      "epoch": 24.26451612903226,
+      "grad_norm": 0.025861376896500587,
+      "learning_rate": 3.809853068280035e-05,
+      "loss": 0.2539,
+      "step": 4688
+    },
+    {
+      "epoch": 24.26967741935484,
+      "grad_norm": 0.016368286684155464,
+      "learning_rate": 3.806395851339672e-05,
+      "loss": 0.2629,
+      "step": 4689
+    },
+    {
+      "epoch": 24.274838709677418,
+      "grad_norm": 0.07432130724191666,
+      "learning_rate": 3.802938634399309e-05,
+      "loss": 0.2819,
+      "step": 4690
+    },
+    {
+      "epoch": 24.28,
+      "grad_norm": 0.031635183840990067,
+      "learning_rate": 3.799481417458946e-05,
+      "loss": 0.2852,
+      "step": 4691
+    },
+    {
+      "epoch": 24.28516129032258,
+      "grad_norm": 0.036796748638153076,
+      "learning_rate": 3.7960242005185825e-05,
+      "loss": 0.2597,
+      "step": 4692
+    },
+    {
+      "epoch": 24.29032258064516,
+      "grad_norm": 0.020966149866580963,
+      "learning_rate": 3.79256698357822e-05,
+      "loss": 0.279,
+      "step": 4693
+    },
+    {
+      "epoch": 24.295483870967743,
+      "grad_norm": 0.0303361713886261,
+      "learning_rate": 3.7891097666378566e-05,
+      "loss": 0.2797,
+      "step": 4694
+    },
+    {
+      "epoch": 24.300645161290323,
+      "grad_norm": 0.01608606055378914,
+      "learning_rate": 3.785652549697494e-05,
+      "loss": 0.2889,
+      "step": 4695
+    },
+    {
+      "epoch": 24.305806451612902,
+      "grad_norm": 0.02812582440674305,
+      "learning_rate": 3.7821953327571306e-05,
+      "loss": 0.2913,
+      "step": 4696
+    },
+    {
+      "epoch": 24.310967741935485,
+      "grad_norm": 0.031856827437877655,
+      "learning_rate": 3.778738115816768e-05,
+      "loss": 0.2909,
+      "step": 4697
+    },
+    {
+      "epoch": 24.316129032258065,
+      "grad_norm": 0.035230763256549835,
+      "learning_rate": 3.775280898876405e-05,
+      "loss": 0.293,
+      "step": 4698
+    },
+    {
+      "epoch": 24.321290322580644,
+      "grad_norm": 0.029683005064725876,
+      "learning_rate": 3.7718236819360414e-05,
+      "loss": 0.2496,
+      "step": 4699
+    },
+    {
+      "epoch": 24.326451612903227,
+      "grad_norm": 0.02654772438108921,
+      "learning_rate": 3.768366464995679e-05,
+      "loss": 0.2724,
+      "step": 4700
+    },
+    {
+      "epoch": 24.331612903225807,
+      "grad_norm": 0.020341627299785614,
+      "learning_rate": 3.7649092480553155e-05,
+      "loss": 0.3046,
+      "step": 4701
+    },
+    {
+      "epoch": 24.336774193548386,
+      "grad_norm": 0.03199196234345436,
+      "learning_rate": 3.761452031114953e-05,
+      "loss": 0.2915,
+      "step": 4702
+    },
+    {
+      "epoch": 24.34193548387097,
+      "grad_norm": 0.027304632589221,
+      "learning_rate": 3.7579948141745896e-05,
+      "loss": 0.2372,
+      "step": 4703
+    },
+    {
+      "epoch": 24.34709677419355,
+      "grad_norm": 0.02666746824979782,
+      "learning_rate": 3.754537597234227e-05,
+      "loss": 0.2951,
+      "step": 4704
+    },
+    {
+      "epoch": 24.35225806451613,
+      "grad_norm": 0.015121554024517536,
+      "learning_rate": 3.7510803802938636e-05,
+      "loss": 0.2636,
+      "step": 4705
+    },
+    {
+      "epoch": 24.35741935483871,
+      "grad_norm": 0.028118696063756943,
+      "learning_rate": 3.747623163353501e-05,
+      "loss": 0.2949,
+      "step": 4706
+    },
+    {
+      "epoch": 24.36258064516129,
+      "grad_norm": 0.022632723674178123,
+      "learning_rate": 3.744165946413137e-05,
+      "loss": 0.2909,
+      "step": 4707
+    },
+    {
+      "epoch": 24.36774193548387,
+      "grad_norm": 0.013088705018162727,
+      "learning_rate": 3.7407087294727744e-05,
+      "loss": 0.279,
+      "step": 4708
+    },
+    {
+      "epoch": 24.37290322580645,
+      "grad_norm": 0.019534530118107796,
+      "learning_rate": 3.737251512532412e-05,
+      "loss": 0.2839,
+      "step": 4709
+    },
+    {
+      "epoch": 24.378064516129033,
+      "grad_norm": 0.02182270959019661,
+      "learning_rate": 3.7337942955920485e-05,
+      "loss": 0.2981,
+      "step": 4710
+    },
+    {
+      "epoch": 24.383225806451613,
+      "grad_norm": 0.026150712743401527,
+      "learning_rate": 3.730337078651686e-05,
+      "loss": 0.2747,
+      "step": 4711
+    },
+    {
+      "epoch": 24.388387096774192,
+      "grad_norm": 0.027071913704276085,
+      "learning_rate": 3.7268798617113225e-05,
+      "loss": 0.2875,
+      "step": 4712
+    },
+    {
+      "epoch": 24.393548387096775,
+      "grad_norm": 0.029957285150885582,
+      "learning_rate": 3.72342264477096e-05,
+      "loss": 0.2756,
+      "step": 4713
+    },
+    {
+      "epoch": 24.398709677419355,
+      "grad_norm": 0.02137676440179348,
+      "learning_rate": 3.7199654278305966e-05,
+      "loss": 0.3047,
+      "step": 4714
+    },
+    {
+      "epoch": 24.403870967741934,
+      "grad_norm": 0.02996964380145073,
+      "learning_rate": 3.716508210890233e-05,
+      "loss": 0.2589,
+      "step": 4715
+    },
+    {
+      "epoch": 24.409032258064517,
+      "grad_norm": 0.023129479959607124,
+      "learning_rate": 3.71305099394987e-05,
+      "loss": 0.3063,
+      "step": 4716
+    },
+    {
+      "epoch": 24.414193548387097,
+      "grad_norm": 0.03332192450761795,
+      "learning_rate": 3.7095937770095074e-05,
+      "loss": 0.3054,
+      "step": 4717
+    },
+    {
+      "epoch": 24.419354838709676,
+      "grad_norm": 0.0266598928719759,
+      "learning_rate": 3.706136560069145e-05,
+      "loss": 0.2775,
+      "step": 4718
+    },
+    {
+      "epoch": 24.42451612903226,
+      "grad_norm": 0.04147280752658844,
+      "learning_rate": 3.7026793431287815e-05,
+      "loss": 0.286,
+      "step": 4719
+    },
+    {
+      "epoch": 24.42967741935484,
+      "grad_norm": 0.027389733120799065,
+      "learning_rate": 3.699222126188419e-05,
+      "loss": 0.274,
+      "step": 4720
+    },
+    {
+      "epoch": 24.434838709677418,
+      "grad_norm": 0.028576146811246872,
+      "learning_rate": 3.6957649092480555e-05,
+      "loss": 0.2922,
+      "step": 4721
+    },
+    {
+      "epoch": 24.44,
+      "grad_norm": 0.02376684732735157,
+      "learning_rate": 3.692307692307693e-05,
+      "loss": 0.2538,
+      "step": 4722
+    },
+    {
+      "epoch": 24.44516129032258,
+      "grad_norm": 0.024726567789912224,
+      "learning_rate": 3.6888504753673296e-05,
+      "loss": 0.2823,
+      "step": 4723
+    },
+    {
+      "epoch": 24.45032258064516,
+      "grad_norm": 0.029869601130485535,
+      "learning_rate": 3.685393258426966e-05,
+      "loss": 0.2907,
+      "step": 4724
+    },
+    {
+      "epoch": 24.455483870967743,
+      "grad_norm": 0.04106883704662323,
+      "learning_rate": 3.681936041486603e-05,
+      "loss": 0.2996,
+      "step": 4725
+    },
+    {
+      "epoch": 24.460645161290323,
+      "grad_norm": 0.028706392273306847,
+      "learning_rate": 3.6784788245462404e-05,
+      "loss": 0.2875,
+      "step": 4726
+    },
+    {
+      "epoch": 24.465806451612902,
+      "grad_norm": 0.02702278643846512,
+      "learning_rate": 3.675021607605878e-05,
+      "loss": 0.2863,
+      "step": 4727
+    },
+    {
+      "epoch": 24.470967741935485,
+      "grad_norm": 0.03315016254782677,
+      "learning_rate": 3.6715643906655145e-05,
+      "loss": 0.2854,
+      "step": 4728
+    },
+    {
+      "epoch": 24.476129032258065,
+      "grad_norm": 0.0262935608625412,
+      "learning_rate": 3.668107173725152e-05,
+      "loss": 0.2885,
+      "step": 4729
+    },
+    {
+      "epoch": 24.481290322580644,
+      "grad_norm": 0.03377668559551239,
+      "learning_rate": 3.6646499567847885e-05,
+      "loss": 0.2681,
+      "step": 4730
+    },
+    {
+      "epoch": 24.486451612903227,
+      "grad_norm": 0.023020094260573387,
+      "learning_rate": 3.661192739844426e-05,
+      "loss": 0.2896,
+      "step": 4731
+    },
+    {
+      "epoch": 24.491612903225807,
+      "grad_norm": 0.02667117491364479,
+      "learning_rate": 3.657735522904062e-05,
+      "loss": 0.2988,
+      "step": 4732
+    },
+    {
+      "epoch": 24.496774193548386,
+      "grad_norm": 0.019025728106498718,
+      "learning_rate": 3.654278305963699e-05,
+      "loss": 0.3182,
+      "step": 4733
+    },
+    {
+      "epoch": 24.501935483870966,
+      "grad_norm": 0.02615918219089508,
+      "learning_rate": 3.650821089023336e-05,
+      "loss": 0.2787,
+      "step": 4734
+    },
+    {
+      "epoch": 24.50709677419355,
+      "grad_norm": 0.024409102275967598,
+      "learning_rate": 3.6473638720829734e-05,
+      "loss": 0.2776,
+      "step": 4735
+    },
+    {
+      "epoch": 24.51225806451613,
+      "grad_norm": 0.03601079806685448,
+      "learning_rate": 3.64390665514261e-05,
+      "loss": 0.2897,
+      "step": 4736
+    },
+    {
+      "epoch": 24.517419354838708,
+      "grad_norm": 0.02868579514324665,
+      "learning_rate": 3.6404494382022475e-05,
+      "loss": 0.2879,
+      "step": 4737
+    },
+    {
+      "epoch": 24.52258064516129,
+      "grad_norm": 0.028838055208325386,
+      "learning_rate": 3.636992221261885e-05,
+      "loss": 0.2715,
+      "step": 4738
+    },
+    {
+      "epoch": 24.52774193548387,
+      "grad_norm": 0.02248472347855568,
+      "learning_rate": 3.6335350043215215e-05,
+      "loss": 0.2933,
+      "step": 4739
+    },
+    {
+      "epoch": 24.53290322580645,
+      "grad_norm": 0.03303847089409828,
+      "learning_rate": 3.630077787381158e-05,
+      "loss": 0.2686,
+      "step": 4740
+    },
+    {
+      "epoch": 24.538064516129033,
+      "grad_norm": 0.020006190985441208,
+      "learning_rate": 3.626620570440795e-05,
+      "loss": 0.3019,
+      "step": 4741
+    },
+    {
+      "epoch": 24.543225806451613,
+      "grad_norm": 0.029418526217341423,
+      "learning_rate": 3.623163353500432e-05,
+      "loss": 0.2859,
+      "step": 4742
+    },
+    {
+      "epoch": 24.548387096774192,
+      "grad_norm": 0.021255088970065117,
+      "learning_rate": 3.619706136560069e-05,
+      "loss": 0.272,
+      "step": 4743
+    },
+    {
+      "epoch": 24.553548387096775,
+      "grad_norm": 0.027109112590551376,
+      "learning_rate": 3.6162489196197064e-05,
+      "loss": 0.2993,
+      "step": 4744
+    },
+    {
+      "epoch": 24.558709677419355,
+      "grad_norm": 0.021172592416405678,
+      "learning_rate": 3.612791702679343e-05,
+      "loss": 0.2948,
+      "step": 4745
+    },
+    {
+      "epoch": 24.563870967741934,
+      "grad_norm": 0.025365089997649193,
+      "learning_rate": 3.6093344857389805e-05,
+      "loss": 0.2776,
+      "step": 4746
+    },
+    {
+      "epoch": 24.569032258064517,
+      "grad_norm": 0.033188942819833755,
+      "learning_rate": 3.605877268798618e-05,
+      "loss": 0.2849,
+      "step": 4747
+    },
+    {
+      "epoch": 24.574193548387097,
+      "grad_norm": 0.026937179267406464,
+      "learning_rate": 3.602420051858254e-05,
+      "loss": 0.2983,
+      "step": 4748
+    },
+    {
+      "epoch": 24.579354838709676,
+      "grad_norm": 0.038281362503767014,
+      "learning_rate": 3.598962834917891e-05,
+      "loss": 0.2663,
+      "step": 4749
+    },
+    {
+      "epoch": 24.58451612903226,
+      "grad_norm": 0.02381383627653122,
+      "learning_rate": 3.595505617977528e-05,
+      "loss": 0.3021,
+      "step": 4750
+    },
+    {
+      "epoch": 24.58967741935484,
+      "grad_norm": 0.019825566560029984,
+      "learning_rate": 3.592048401037165e-05,
+      "loss": 0.3016,
+      "step": 4751
+    },
+    {
+      "epoch": 24.59483870967742,
+      "grad_norm": 0.028184104710817337,
+      "learning_rate": 3.588591184096802e-05,
+      "loss": 0.2659,
+      "step": 4752
+    },
+    {
+      "epoch": 24.6,
+      "grad_norm": 0.01890227571129799,
+      "learning_rate": 3.5851339671564394e-05,
+      "loss": 0.2955,
+      "step": 4753
+    },
+    {
+      "epoch": 24.60516129032258,
+      "grad_norm": 0.024583544582128525,
+      "learning_rate": 3.581676750216076e-05,
+      "loss": 0.2565,
+      "step": 4754
+    },
+    {
+      "epoch": 24.61032258064516,
+      "grad_norm": 0.018444258719682693,
+      "learning_rate": 3.5782195332757135e-05,
+      "loss": 0.2712,
+      "step": 4755
+    },
+    {
+      "epoch": 24.615483870967743,
+      "grad_norm": 0.028081731870770454,
+      "learning_rate": 3.57476231633535e-05,
+      "loss": 0.2703,
+      "step": 4756
+    },
+    {
+      "epoch": 24.620645161290323,
+      "grad_norm": 0.01912839151918888,
+      "learning_rate": 3.571305099394987e-05,
+      "loss": 0.2914,
+      "step": 4757
+    },
+    {
+      "epoch": 24.625806451612902,
+      "grad_norm": 0.03434016555547714,
+      "learning_rate": 3.567847882454624e-05,
+      "loss": 0.2939,
+      "step": 4758
+    },
+    {
+      "epoch": 24.630967741935486,
+      "grad_norm": 0.024894466623663902,
+      "learning_rate": 3.564390665514261e-05,
+      "loss": 0.2775,
+      "step": 4759
+    },
+    {
+      "epoch": 24.636129032258065,
+      "grad_norm": 0.023237096145749092,
+      "learning_rate": 3.560933448573898e-05,
+      "loss": 0.3028,
+      "step": 4760
+    },
+    {
+      "epoch": 24.641290322580645,
+      "grad_norm": 0.039763763546943665,
+      "learning_rate": 3.557476231633535e-05,
+      "loss": 0.2661,
+      "step": 4761
+    },
+    {
+      "epoch": 24.646451612903228,
+      "grad_norm": 0.02068629302084446,
+      "learning_rate": 3.5540190146931724e-05,
+      "loss": 0.2879,
+      "step": 4762
+    },
+    {
+      "epoch": 24.651612903225807,
+      "grad_norm": 0.02837016060948372,
+      "learning_rate": 3.550561797752809e-05,
+      "loss": 0.2794,
+      "step": 4763
+    },
+    {
+      "epoch": 24.656774193548387,
+      "grad_norm": 0.02135073021054268,
+      "learning_rate": 3.547104580812446e-05,
+      "loss": 0.2949,
+      "step": 4764
+    },
+    {
+      "epoch": 24.661935483870966,
+      "grad_norm": 0.02839534729719162,
+      "learning_rate": 3.543647363872083e-05,
+      "loss": 0.2777,
+      "step": 4765
+    },
+    {
+      "epoch": 24.66709677419355,
+      "grad_norm": 0.020558951422572136,
+      "learning_rate": 3.54019014693172e-05,
+      "loss": 0.268,
+      "step": 4766
+    },
+    {
+      "epoch": 24.67225806451613,
+      "grad_norm": 0.0263463594019413,
+      "learning_rate": 3.536732929991357e-05,
+      "loss": 0.2654,
+      "step": 4767
+    },
+    {
+      "epoch": 24.677419354838708,
+      "grad_norm": 0.0185295008122921,
+      "learning_rate": 3.533275713050994e-05,
+      "loss": 0.2444,
+      "step": 4768
+    },
+    {
+      "epoch": 24.68258064516129,
+      "grad_norm": 0.032571982592344284,
+      "learning_rate": 3.529818496110631e-05,
+      "loss": 0.2627,
+      "step": 4769
+    },
+    {
+      "epoch": 24.68774193548387,
+      "grad_norm": 0.02861125022172928,
+      "learning_rate": 3.526361279170268e-05,
+      "loss": 0.2916,
+      "step": 4770
+    },
+    {
+      "epoch": 24.69290322580645,
+      "grad_norm": 0.016636928543448448,
+      "learning_rate": 3.5229040622299054e-05,
+      "loss": 0.2765,
+      "step": 4771
+    },
+    {
+      "epoch": 24.698064516129033,
+      "grad_norm": 0.02067486383020878,
+      "learning_rate": 3.519446845289542e-05,
+      "loss": 0.2988,
+      "step": 4772
+    },
+    {
+      "epoch": 24.703225806451613,
+      "grad_norm": 0.031121861189603806,
+      "learning_rate": 3.515989628349179e-05,
+      "loss": 0.2842,
+      "step": 4773
+    },
+    {
+      "epoch": 24.708387096774192,
+      "grad_norm": 0.02822178602218628,
+      "learning_rate": 3.512532411408816e-05,
+      "loss": 0.3131,
+      "step": 4774
+    },
+    {
+      "epoch": 24.713548387096775,
+      "grad_norm": 0.03461252152919769,
+      "learning_rate": 3.509075194468453e-05,
+      "loss": 0.269,
+      "step": 4775
+    },
+    {
+      "epoch": 24.718709677419355,
+      "grad_norm": 0.032957471907138824,
+      "learning_rate": 3.50561797752809e-05,
+      "loss": 0.2711,
+      "step": 4776
+    },
+    {
+      "epoch": 24.723870967741934,
+      "grad_norm": 0.014953196980059147,
+      "learning_rate": 3.502160760587727e-05,
+      "loss": 0.2898,
+      "step": 4777
+    },
+    {
+      "epoch": 24.729032258064517,
+      "grad_norm": 0.022889407351613045,
+      "learning_rate": 3.498703543647364e-05,
+      "loss": 0.2726,
+      "step": 4778
+    },
+    {
+      "epoch": 24.734193548387097,
+      "grad_norm": 0.026240196079015732,
+      "learning_rate": 3.495246326707001e-05,
+      "loss": 0.2604,
+      "step": 4779
+    },
+    {
+      "epoch": 24.739354838709676,
+      "grad_norm": 0.03163983300328255,
+      "learning_rate": 3.491789109766638e-05,
+      "loss": 0.2754,
+      "step": 4780
+    },
+    {
+      "epoch": 24.74451612903226,
+      "grad_norm": 0.026734840124845505,
+      "learning_rate": 3.488331892826275e-05,
+      "loss": 0.2425,
+      "step": 4781
+    },
+    {
+      "epoch": 24.74967741935484,
+      "grad_norm": 0.023920204490423203,
+      "learning_rate": 3.484874675885912e-05,
+      "loss": 0.2937,
+      "step": 4782
+    },
+    {
+      "epoch": 24.75483870967742,
+      "grad_norm": 0.029201235622167587,
+      "learning_rate": 3.481417458945549e-05,
+      "loss": 0.2995,
+      "step": 4783
+    },
+    {
+      "epoch": 24.76,
+      "grad_norm": 0.017785780131816864,
+      "learning_rate": 3.477960242005186e-05,
+      "loss": 0.2904,
+      "step": 4784
+    },
+    {
+      "epoch": 24.76516129032258,
+      "grad_norm": 0.02908969111740589,
+      "learning_rate": 3.474503025064823e-05,
+      "loss": 0.2904,
+      "step": 4785
+    },
+    {
+      "epoch": 24.77032258064516,
+      "grad_norm": 0.030694156885147095,
+      "learning_rate": 3.47104580812446e-05,
+      "loss": 0.2768,
+      "step": 4786
+    },
+    {
+      "epoch": 24.775483870967744,
+      "grad_norm": 0.032455720007419586,
+      "learning_rate": 3.467588591184097e-05,
+      "loss": 0.2423,
+      "step": 4787
+    },
+    {
+      "epoch": 24.780645161290323,
+      "grad_norm": 0.03394845128059387,
+      "learning_rate": 3.464131374243734e-05,
+      "loss": 0.2792,
+      "step": 4788
+    },
+    {
+      "epoch": 24.785806451612903,
+      "grad_norm": 0.03602224960923195,
+      "learning_rate": 3.460674157303371e-05,
+      "loss": 0.2913,
+      "step": 4789
+    },
+    {
+      "epoch": 24.790967741935482,
+      "grad_norm": 0.030233416706323624,
+      "learning_rate": 3.457216940363008e-05,
+      "loss": 0.2856,
+      "step": 4790
+    },
+    {
+      "epoch": 24.796129032258065,
+      "grad_norm": 0.04525449872016907,
+      "learning_rate": 3.453759723422645e-05,
+      "loss": 0.2845,
+      "step": 4791
+    },
+    {
+      "epoch": 24.801290322580645,
+      "grad_norm": 0.023684484884142876,
+      "learning_rate": 3.450302506482282e-05,
+      "loss": 0.292,
+      "step": 4792
+    },
+    {
+      "epoch": 24.806451612903224,
+      "grad_norm": 0.02849588729441166,
+      "learning_rate": 3.446845289541919e-05,
+      "loss": 0.2961,
+      "step": 4793
+    },
+    {
+      "epoch": 24.811612903225807,
+      "grad_norm": 0.031291764229536057,
+      "learning_rate": 3.443388072601556e-05,
+      "loss": 0.2973,
+      "step": 4794
+    },
+    {
+      "epoch": 24.816774193548387,
+      "grad_norm": 0.030267586931586266,
+      "learning_rate": 3.439930855661193e-05,
+      "loss": 0.288,
+      "step": 4795
+    },
+    {
+      "epoch": 24.821935483870966,
+      "grad_norm": 0.029514864087104797,
+      "learning_rate": 3.43647363872083e-05,
+      "loss": 0.2708,
+      "step": 4796
+    },
+    {
+      "epoch": 24.82709677419355,
+      "grad_norm": 0.03276633098721504,
+      "learning_rate": 3.433016421780467e-05,
+      "loss": 0.2979,
+      "step": 4797
+    },
+    {
+      "epoch": 24.83225806451613,
+      "grad_norm": 0.04009691998362541,
+      "learning_rate": 3.429559204840104e-05,
+      "loss": 0.2732,
+      "step": 4798
+    },
+    {
+      "epoch": 24.83741935483871,
+      "grad_norm": 0.03161231055855751,
+      "learning_rate": 3.426101987899741e-05,
+      "loss": 0.2442,
+      "step": 4799
+    },
+    {
+      "epoch": 24.84258064516129,
+      "grad_norm": 0.024575125426054,
+      "learning_rate": 3.422644770959378e-05,
+      "loss": 0.2877,
+      "step": 4800
+    },
+    {
+      "epoch": 24.84774193548387,
+      "grad_norm": 0.02975618839263916,
+      "learning_rate": 3.419187554019015e-05,
+      "loss": 0.2796,
+      "step": 4801
+    },
+    {
+      "epoch": 24.85290322580645,
+      "grad_norm": 0.025476573035120964,
+      "learning_rate": 3.415730337078652e-05,
+      "loss": 0.3138,
+      "step": 4802
+    },
+    {
+      "epoch": 24.858064516129033,
+      "grad_norm": 0.04720594733953476,
+      "learning_rate": 3.412273120138289e-05,
+      "loss": 0.2946,
+      "step": 4803
+    },
+    {
+      "epoch": 24.863225806451613,
+      "grad_norm": 0.03387271612882614,
+      "learning_rate": 3.408815903197926e-05,
+      "loss": 0.282,
+      "step": 4804
+    },
+    {
+      "epoch": 24.868387096774192,
+      "grad_norm": 0.025648804381489754,
+      "learning_rate": 3.4053586862575626e-05,
+      "loss": 0.2794,
+      "step": 4805
+    },
+    {
+      "epoch": 24.873548387096776,
+      "grad_norm": 0.041306331753730774,
+      "learning_rate": 3.4019014693172e-05,
+      "loss": 0.2685,
+      "step": 4806
+    },
+    {
+      "epoch": 24.878709677419355,
+      "grad_norm": 0.014421621337532997,
+      "learning_rate": 3.398444252376837e-05,
+      "loss": 0.2845,
+      "step": 4807
+    },
+    {
+      "epoch": 24.883870967741935,
+      "grad_norm": 0.030996158719062805,
+      "learning_rate": 3.394987035436474e-05,
+      "loss": 0.2576,
+      "step": 4808
+    },
+    {
+      "epoch": 24.889032258064518,
+      "grad_norm": 0.024918200448155403,
+      "learning_rate": 3.391529818496111e-05,
+      "loss": 0.2904,
+      "step": 4809
+    },
+    {
+      "epoch": 24.894193548387097,
+      "grad_norm": 0.023690849542617798,
+      "learning_rate": 3.388072601555748e-05,
+      "loss": 0.2601,
+      "step": 4810
+    },
+    {
+      "epoch": 24.899354838709677,
+      "grad_norm": 0.028375620022416115,
+      "learning_rate": 3.384615384615385e-05,
+      "loss": 0.2981,
+      "step": 4811
+    },
+    {
+      "epoch": 24.90451612903226,
+      "grad_norm": 0.03208836913108826,
+      "learning_rate": 3.381158167675022e-05,
+      "loss": 0.2906,
+      "step": 4812
+    },
+    {
+      "epoch": 24.90967741935484,
+      "grad_norm": 0.03496381267905235,
+      "learning_rate": 3.377700950734658e-05,
+      "loss": 0.2977,
+      "step": 4813
+    },
+    {
+      "epoch": 24.91483870967742,
+      "grad_norm": 0.022008977830410004,
+      "learning_rate": 3.3742437337942956e-05,
+      "loss": 0.2748,
+      "step": 4814
+    },
+    {
+      "epoch": 24.92,
+      "grad_norm": 0.02474852278828621,
+      "learning_rate": 3.370786516853933e-05,
+      "loss": 0.2865,
+      "step": 4815
+    },
+    {
+      "epoch": 24.92516129032258,
+      "grad_norm": 0.026334669440984726,
+      "learning_rate": 3.36732929991357e-05,
+      "loss": 0.2482,
+      "step": 4816
+    },
+    {
+      "epoch": 24.93032258064516,
+      "grad_norm": 0.027599075809121132,
+      "learning_rate": 3.363872082973207e-05,
+      "loss": 0.2775,
+      "step": 4817
+    },
+    {
+      "epoch": 24.93548387096774,
+      "grad_norm": 0.03356274962425232,
+      "learning_rate": 3.360414866032844e-05,
+      "loss": 0.2614,
+      "step": 4818
+    },
+    {
+      "epoch": 24.940645161290323,
+      "grad_norm": 0.027799921110272408,
+      "learning_rate": 3.356957649092481e-05,
+      "loss": 0.2992,
+      "step": 4819
+    },
+    {
+      "epoch": 24.945806451612903,
+      "grad_norm": 0.026381291449069977,
+      "learning_rate": 3.353500432152118e-05,
+      "loss": 0.2963,
+      "step": 4820
+    },
+    {
+      "epoch": 24.950967741935482,
+      "grad_norm": 0.031160477548837662,
+      "learning_rate": 3.3500432152117545e-05,
+      "loss": 0.2974,
+      "step": 4821
+    },
+    {
+      "epoch": 24.956129032258065,
+      "grad_norm": 0.03411593288183212,
+      "learning_rate": 3.346585998271391e-05,
+      "loss": 0.2856,
+      "step": 4822
+    },
+    {
+      "epoch": 24.961290322580645,
+      "grad_norm": 0.034743521362543106,
+      "learning_rate": 3.3431287813310286e-05,
+      "loss": 0.288,
+      "step": 4823
+    },
+    {
+      "epoch": 24.966451612903224,
+      "grad_norm": 0.030682075768709183,
+      "learning_rate": 3.339671564390665e-05,
+      "loss": 0.2721,
+      "step": 4824
+    },
+    {
+      "epoch": 24.971612903225807,
+      "grad_norm": 0.025759510695934296,
+      "learning_rate": 3.336214347450303e-05,
+      "loss": 0.2914,
+      "step": 4825
+    },
+    {
+      "epoch": 24.976774193548387,
+      "grad_norm": 0.024776676669716835,
+      "learning_rate": 3.33275713050994e-05,
+      "loss": 0.2881,
+      "step": 4826
+    },
+    {
+      "epoch": 24.981935483870966,
+      "grad_norm": 0.03283006697893143,
+      "learning_rate": 3.329299913569577e-05,
+      "loss": 0.2944,
+      "step": 4827
+    },
+    {
+      "epoch": 24.98709677419355,
+      "grad_norm": 0.031208185479044914,
+      "learning_rate": 3.325842696629214e-05,
+      "loss": 0.2677,
+      "step": 4828
+    },
+    {
+      "epoch": 24.99225806451613,
+      "grad_norm": 0.02366589568555355,
+      "learning_rate": 3.32238547968885e-05,
+      "loss": 0.2854,
+      "step": 4829
+    },
+    {
+      "epoch": 24.99741935483871,
+      "grad_norm": 0.04011349752545357,
+      "learning_rate": 3.3189282627484875e-05,
+      "loss": 0.3451,
+      "step": 4830
+    },
+    {
+      "epoch": 25.003870967741936,
+      "grad_norm": 0.04097389057278633,
+      "learning_rate": 3.315471045808124e-05,
+      "loss": 0.354,
+      "step": 4831
+    },
+    {
+      "epoch": 25.009032258064515,
+      "grad_norm": 0.017793450504541397,
+      "learning_rate": 3.3120138288677616e-05,
+      "loss": 0.2948,
+      "step": 4832
+    },
+    {
+      "epoch": 25.014193548387098,
+      "grad_norm": 0.02741445228457451,
+      "learning_rate": 3.308556611927398e-05,
+      "loss": 0.2777,
+      "step": 4833
+    },
+    {
+      "epoch": 25.019354838709678,
+      "grad_norm": 0.022029457613825798,
+      "learning_rate": 3.305099394987036e-05,
+      "loss": 0.2762,
+      "step": 4834
+    },
+    {
+      "epoch": 25.024516129032257,
+      "grad_norm": 0.027535149827599525,
+      "learning_rate": 3.301642178046673e-05,
+      "loss": 0.2631,
+      "step": 4835
+    },
+    {
+      "epoch": 25.02967741935484,
+      "grad_norm": 0.018552957102656364,
+      "learning_rate": 3.29818496110631e-05,
+      "loss": 0.2716,
+      "step": 4836
+    },
+    {
+      "epoch": 25.03483870967742,
+      "grad_norm": 0.024942126125097275,
+      "learning_rate": 3.2947277441659465e-05,
+      "loss": 0.3032,
+      "step": 4837
+    },
+    {
+      "epoch": 25.04,
+      "grad_norm": 0.02588256262242794,
+      "learning_rate": 3.291270527225583e-05,
+      "loss": 0.2751,
+      "step": 4838
+    },
+    {
+      "epoch": 25.045161290322582,
+      "grad_norm": 0.019401656463742256,
+      "learning_rate": 3.2878133102852205e-05,
+      "loss": 0.3096,
+      "step": 4839
+    },
+    {
+      "epoch": 25.05032258064516,
+      "grad_norm": 0.023466547951102257,
+      "learning_rate": 3.284356093344857e-05,
+      "loss": 0.2888,
+      "step": 4840
+    },
+    {
+      "epoch": 25.05548387096774,
+      "grad_norm": 0.02681884728372097,
+      "learning_rate": 3.2808988764044946e-05,
+      "loss": 0.291,
+      "step": 4841
+    },
+    {
+      "epoch": 25.060645161290324,
+      "grad_norm": 0.025864025577902794,
+      "learning_rate": 3.277441659464131e-05,
+      "loss": 0.2979,
+      "step": 4842
+    },
+    {
+      "epoch": 25.065806451612904,
+      "grad_norm": 0.016466213390231133,
+      "learning_rate": 3.273984442523769e-05,
+      "loss": 0.2927,
+      "step": 4843
+    },
+    {
+      "epoch": 25.070967741935483,
+      "grad_norm": 0.023751206696033478,
+      "learning_rate": 3.270527225583406e-05,
+      "loss": 0.2775,
+      "step": 4844
+    },
+    {
+      "epoch": 25.076129032258063,
+      "grad_norm": 0.01959926076233387,
+      "learning_rate": 3.267070008643042e-05,
+      "loss": 0.2543,
+      "step": 4845
+    },
+    {
+      "epoch": 25.081290322580646,
+      "grad_norm": 0.026109516620635986,
+      "learning_rate": 3.2636127917026795e-05,
+      "loss": 0.2824,
+      "step": 4846
+    },
+    {
+      "epoch": 25.086451612903225,
+      "grad_norm": 0.021967578679323196,
+      "learning_rate": 3.260155574762316e-05,
+      "loss": 0.2865,
+      "step": 4847
+    },
+    {
+      "epoch": 25.091612903225805,
+      "grad_norm": 0.026360806077718735,
+      "learning_rate": 3.2566983578219535e-05,
+      "loss": 0.2897,
+      "step": 4848
+    },
+    {
+      "epoch": 25.096774193548388,
+      "grad_norm": 0.02068096026778221,
+      "learning_rate": 3.25324114088159e-05,
+      "loss": 0.252,
+      "step": 4849
+    },
+    {
+      "epoch": 25.101935483870967,
+      "grad_norm": 0.00979693979024887,
+      "learning_rate": 3.2497839239412276e-05,
+      "loss": 0.2921,
+      "step": 4850
+    },
+    {
+      "epoch": 25.107096774193547,
+      "grad_norm": 0.01860295981168747,
+      "learning_rate": 3.246326707000864e-05,
+      "loss": 0.2784,
+      "step": 4851
+    },
+    {
+      "epoch": 25.11225806451613,
+      "grad_norm": 0.016835566610097885,
+      "learning_rate": 3.242869490060502e-05,
+      "loss": 0.2889,
+      "step": 4852
+    },
+    {
+      "epoch": 25.11741935483871,
+      "grad_norm": 0.019058626145124435,
+      "learning_rate": 3.2394122731201384e-05,
+      "loss": 0.2488,
+      "step": 4853
+    },
+    {
+      "epoch": 25.12258064516129,
+      "grad_norm": 0.0267679151147604,
+      "learning_rate": 3.235955056179775e-05,
+      "loss": 0.299,
+      "step": 4854
+    },
+    {
+      "epoch": 25.127741935483872,
+      "grad_norm": 0.020871764048933983,
+      "learning_rate": 3.2324978392394124e-05,
+      "loss": 0.2746,
+      "step": 4855
+    },
+    {
+      "epoch": 25.13290322580645,
+      "grad_norm": 0.021220631897449493,
+      "learning_rate": 3.229040622299049e-05,
+      "loss": 0.2766,
+      "step": 4856
+    },
+    {
+      "epoch": 25.13806451612903,
+      "grad_norm": 0.022922759875655174,
+      "learning_rate": 3.2255834053586865e-05,
+      "loss": 0.2522,
+      "step": 4857
+    },
+    {
+      "epoch": 25.143225806451614,
+      "grad_norm": 0.019960233941674232,
+      "learning_rate": 3.222126188418323e-05,
+      "loss": 0.2866,
+      "step": 4858
+    },
+    {
+      "epoch": 25.148387096774194,
+      "grad_norm": 0.027706388384103775,
+      "learning_rate": 3.2186689714779606e-05,
+      "loss": 0.284,
+      "step": 4859
+    },
+    {
+      "epoch": 25.153548387096773,
+      "grad_norm": 0.02100892923772335,
+      "learning_rate": 3.215211754537597e-05,
+      "loss": 0.2746,
+      "step": 4860
+    },
+    {
+      "epoch": 25.158709677419356,
+      "grad_norm": 0.030665097758173943,
+      "learning_rate": 3.211754537597234e-05,
+      "loss": 0.2551,
+      "step": 4861
+    },
+    {
+      "epoch": 25.163870967741936,
+      "grad_norm": 0.025826474651694298,
+      "learning_rate": 3.2082973206568714e-05,
+      "loss": 0.2519,
+      "step": 4862
+    },
+    {
+      "epoch": 25.169032258064515,
+      "grad_norm": 0.027872029691934586,
+      "learning_rate": 3.204840103716508e-05,
+      "loss": 0.2767,
+      "step": 4863
+    },
+    {
+      "epoch": 25.174193548387098,
+      "grad_norm": 0.02551429532468319,
+      "learning_rate": 3.2013828867761454e-05,
+      "loss": 0.2664,
+      "step": 4864
+    },
+    {
+      "epoch": 25.179354838709678,
+      "grad_norm": 0.03503783419728279,
+      "learning_rate": 3.197925669835782e-05,
+      "loss": 0.2887,
+      "step": 4865
+    },
+    {
+      "epoch": 25.184516129032257,
+      "grad_norm": 0.01989118568599224,
+      "learning_rate": 3.1944684528954195e-05,
+      "loss": 0.2689,
+      "step": 4866
+    },
+    {
+      "epoch": 25.18967741935484,
+      "grad_norm": 0.024500668048858643,
+      "learning_rate": 3.191011235955056e-05,
+      "loss": 0.2671,
+      "step": 4867
+    },
+    {
+      "epoch": 25.19483870967742,
+      "grad_norm": 0.02544206939637661,
+      "learning_rate": 3.1875540190146936e-05,
+      "loss": 0.3069,
+      "step": 4868
+    },
+    {
+      "epoch": 25.2,
+      "grad_norm": 0.0162148829549551,
+      "learning_rate": 3.18409680207433e-05,
+      "loss": 0.2735,
+      "step": 4869
+    },
+    {
+      "epoch": 25.205161290322582,
+      "grad_norm": 0.020251674577593803,
+      "learning_rate": 3.180639585133967e-05,
+      "loss": 0.3073,
+      "step": 4870
+    },
+    {
+      "epoch": 25.210322580645162,
+      "grad_norm": 0.023475006222724915,
+      "learning_rate": 3.1771823681936044e-05,
+      "loss": 0.292,
+      "step": 4871
+    },
+    {
+      "epoch": 25.21548387096774,
+      "grad_norm": 0.02954765409231186,
+      "learning_rate": 3.173725151253241e-05,
+      "loss": 0.3029,
+      "step": 4872
+    },
+    {
+      "epoch": 25.22064516129032,
+      "grad_norm": 0.028549276292324066,
+      "learning_rate": 3.1702679343128784e-05,
+      "loss": 0.2877,
+      "step": 4873
+    },
+    {
+      "epoch": 25.225806451612904,
+      "grad_norm": 0.027352729812264442,
+      "learning_rate": 3.166810717372515e-05,
+      "loss": 0.2834,
+      "step": 4874
+    },
+    {
+      "epoch": 25.230967741935483,
+      "grad_norm": 0.03407920524477959,
+      "learning_rate": 3.1633535004321525e-05,
+      "loss": 0.2593,
+      "step": 4875
+    },
+    {
+      "epoch": 25.236129032258063,
+      "grad_norm": 0.02326350472867489,
+      "learning_rate": 3.159896283491789e-05,
+      "loss": 0.291,
+      "step": 4876
+    },
+    {
+      "epoch": 25.241290322580646,
+      "grad_norm": 0.026500411331653595,
+      "learning_rate": 3.1564390665514266e-05,
+      "loss": 0.26,
+      "step": 4877
+    },
+    {
+      "epoch": 25.246451612903225,
+      "grad_norm": 0.026065438985824585,
+      "learning_rate": 3.152981849611063e-05,
+      "loss": 0.2692,
+      "step": 4878
+    },
+    {
+      "epoch": 25.251612903225805,
+      "grad_norm": 0.0231636930257082,
+      "learning_rate": 3.1495246326707e-05,
+      "loss": 0.2922,
+      "step": 4879
+    },
+    {
+      "epoch": 25.256774193548388,
+      "grad_norm": 0.03608477860689163,
+      "learning_rate": 3.1460674157303374e-05,
+      "loss": 0.2813,
+      "step": 4880
+    },
+    {
+      "epoch": 25.261935483870968,
+      "grad_norm": 0.033223431557416916,
+      "learning_rate": 3.142610198789974e-05,
+      "loss": 0.3033,
+      "step": 4881
+    },
+    {
+      "epoch": 25.267096774193547,
+      "grad_norm": 0.036997996270656586,
+      "learning_rate": 3.1391529818496114e-05,
+      "loss": 0.2849,
+      "step": 4882
+    },
+    {
+      "epoch": 25.27225806451613,
+      "grad_norm": 0.027209848165512085,
+      "learning_rate": 3.135695764909248e-05,
+      "loss": 0.2537,
+      "step": 4883
+    },
+    {
+      "epoch": 25.27741935483871,
+      "grad_norm": 0.024684887379407883,
+      "learning_rate": 3.1322385479688855e-05,
+      "loss": 0.2815,
+      "step": 4884
+    },
+    {
+      "epoch": 25.28258064516129,
+      "grad_norm": 0.03034946322441101,
+      "learning_rate": 3.128781331028522e-05,
+      "loss": 0.2919,
+      "step": 4885
+    },
+    {
+      "epoch": 25.287741935483872,
+      "grad_norm": 0.02621709555387497,
+      "learning_rate": 3.125324114088159e-05,
+      "loss": 0.3032,
+      "step": 4886
+    },
+    {
+      "epoch": 25.29290322580645,
+      "grad_norm": 0.026160109788179398,
+      "learning_rate": 3.121866897147796e-05,
+      "loss": 0.3039,
+      "step": 4887
+    },
+    {
+      "epoch": 25.29806451612903,
+      "grad_norm": 0.02128629758954048,
+      "learning_rate": 3.118409680207433e-05,
+      "loss": 0.2984,
+      "step": 4888
+    },
+    {
+      "epoch": 25.303225806451614,
+      "grad_norm": 0.026484211906790733,
+      "learning_rate": 3.1149524632670704e-05,
+      "loss": 0.2652,
+      "step": 4889
+    },
+    {
+      "epoch": 25.308387096774194,
+      "grad_norm": 0.02388433925807476,
+      "learning_rate": 3.111495246326707e-05,
+      "loss": 0.28,
+      "step": 4890
+    },
+    {
+      "epoch": 25.313548387096773,
+      "grad_norm": 0.02174467407166958,
+      "learning_rate": 3.1080380293863444e-05,
+      "loss": 0.2743,
+      "step": 4891
+    },
+    {
+      "epoch": 25.318709677419356,
+      "grad_norm": 0.024606136605143547,
+      "learning_rate": 3.104580812445981e-05,
+      "loss": 0.3054,
+      "step": 4892
+    },
+    {
+      "epoch": 25.323870967741936,
+      "grad_norm": 0.039182551205158234,
+      "learning_rate": 3.1011235955056185e-05,
+      "loss": 0.2944,
+      "step": 4893
+    },
+    {
+      "epoch": 25.329032258064515,
+      "grad_norm": 0.033759258687496185,
+      "learning_rate": 3.097666378565255e-05,
+      "loss": 0.2734,
+      "step": 4894
+    },
+    {
+      "epoch": 25.3341935483871,
+      "grad_norm": 0.0212418083101511,
+      "learning_rate": 3.094209161624892e-05,
+      "loss": 0.2858,
+      "step": 4895
+    },
+    {
+      "epoch": 25.339354838709678,
+      "grad_norm": 0.02912987396121025,
+      "learning_rate": 3.090751944684529e-05,
+      "loss": 0.2894,
+      "step": 4896
+    },
+    {
+      "epoch": 25.344516129032257,
+      "grad_norm": 0.03438873961567879,
+      "learning_rate": 3.087294727744166e-05,
+      "loss": 0.2845,
+      "step": 4897
+    },
+    {
+      "epoch": 25.34967741935484,
+      "grad_norm": 0.02792998217046261,
+      "learning_rate": 3.0838375108038034e-05,
+      "loss": 0.2471,
+      "step": 4898
+    },
+    {
+      "epoch": 25.35483870967742,
+      "grad_norm": 0.03374675661325455,
+      "learning_rate": 3.08038029386344e-05,
+      "loss": 0.2791,
+      "step": 4899
+    },
+    {
+      "epoch": 25.36,
+      "grad_norm": 0.027046913281083107,
+      "learning_rate": 3.0769230769230774e-05,
+      "loss": 0.2469,
+      "step": 4900
+    },
+    {
+      "epoch": 25.36516129032258,
+      "grad_norm": 0.024601057171821594,
+      "learning_rate": 3.073465859982714e-05,
+      "loss": 0.2829,
+      "step": 4901
+    },
+    {
+      "epoch": 25.370322580645162,
+      "grad_norm": 0.020092222839593887,
+      "learning_rate": 3.070008643042351e-05,
+      "loss": 0.2627,
+      "step": 4902
+    },
+    {
+      "epoch": 25.37548387096774,
+      "grad_norm": 0.0169574823230505,
+      "learning_rate": 3.0665514261019875e-05,
+      "loss": 0.2999,
+      "step": 4903
+    },
+    {
+      "epoch": 25.38064516129032,
+      "grad_norm": 0.025548312813043594,
+      "learning_rate": 3.063094209161625e-05,
+      "loss": 0.2277,
+      "step": 4904
+    },
+    {
+      "epoch": 25.385806451612904,
+      "grad_norm": 0.0333445742726326,
+      "learning_rate": 3.059636992221262e-05,
+      "loss": 0.2985,
+      "step": 4905
+    },
+    {
+      "epoch": 25.390967741935484,
+      "grad_norm": 0.03033309057354927,
+      "learning_rate": 3.056179775280899e-05,
+      "loss": 0.3013,
+      "step": 4906
+    },
+    {
+      "epoch": 25.396129032258063,
+      "grad_norm": 0.02710852399468422,
+      "learning_rate": 3.0527225583405364e-05,
+      "loss": 0.2801,
+      "step": 4907
+    },
+    {
+      "epoch": 25.401290322580646,
+      "grad_norm": 0.027628682553768158,
+      "learning_rate": 3.049265341400173e-05,
+      "loss": 0.2699,
+      "step": 4908
+    },
+    {
+      "epoch": 25.406451612903226,
+      "grad_norm": 0.019500454887747765,
+      "learning_rate": 3.04580812445981e-05,
+      "loss": 0.2859,
+      "step": 4909
+    },
+    {
+      "epoch": 25.411612903225805,
+      "grad_norm": 0.03231529891490936,
+      "learning_rate": 3.0423509075194468e-05,
+      "loss": 0.2678,
+      "step": 4910
+    },
+    {
+      "epoch": 25.416774193548388,
+      "grad_norm": 0.021519023925065994,
+      "learning_rate": 3.0388936905790838e-05,
+      "loss": 0.2777,
+      "step": 4911
+    },
+    {
+      "epoch": 25.421935483870968,
+      "grad_norm": 0.030250662937760353,
+      "learning_rate": 3.035436473638721e-05,
+      "loss": 0.308,
+      "step": 4912
+    },
+    {
+      "epoch": 25.427096774193547,
+      "grad_norm": 0.02260669507086277,
+      "learning_rate": 3.031979256698358e-05,
+      "loss": 0.2961,
+      "step": 4913
+    },
+    {
+      "epoch": 25.43225806451613,
+      "grad_norm": 0.028791967779397964,
+      "learning_rate": 3.028522039757995e-05,
+      "loss": 0.2991,
+      "step": 4914
+    },
+    {
+      "epoch": 25.43741935483871,
+      "grad_norm": 0.038297880440950394,
+      "learning_rate": 3.025064822817632e-05,
+      "loss": 0.2575,
+      "step": 4915
+    },
+    {
+      "epoch": 25.44258064516129,
+      "grad_norm": 0.04047620669007301,
+      "learning_rate": 3.021607605877269e-05,
+      "loss": 0.2628,
+      "step": 4916
+    },
+    {
+      "epoch": 25.447741935483872,
+      "grad_norm": 0.03187387064099312,
+      "learning_rate": 3.018150388936906e-05,
+      "loss": 0.2572,
+      "step": 4917
+    },
+    {
+      "epoch": 25.452903225806452,
+      "grad_norm": 0.029919113963842392,
+      "learning_rate": 3.0146931719965428e-05,
+      "loss": 0.2631,
+      "step": 4918
+    },
+    {
+      "epoch": 25.45806451612903,
+      "grad_norm": 0.035078808665275574,
+      "learning_rate": 3.0112359550561798e-05,
+      "loss": 0.2819,
+      "step": 4919
+    },
+    {
+      "epoch": 25.463225806451614,
+      "grad_norm": 0.024055929854512215,
+      "learning_rate": 3.0077787381158168e-05,
+      "loss": 0.3049,
+      "step": 4920
+    },
+    {
+      "epoch": 25.468387096774194,
+      "grad_norm": 0.0273646991699934,
+      "learning_rate": 3.004321521175454e-05,
+      "loss": 0.2887,
+      "step": 4921
+    },
+    {
+      "epoch": 25.473548387096773,
+      "grad_norm": 0.02604985423386097,
+      "learning_rate": 3.000864304235091e-05,
+      "loss": 0.2862,
+      "step": 4922
+    },
+    {
+      "epoch": 25.478709677419356,
+      "grad_norm": 0.023823456838726997,
+      "learning_rate": 2.997407087294728e-05,
+      "loss": 0.2826,
+      "step": 4923
+    },
+    {
+      "epoch": 25.483870967741936,
+      "grad_norm": 0.03063439577817917,
+      "learning_rate": 2.993949870354365e-05,
+      "loss": 0.3005,
+      "step": 4924
+    },
+    {
+      "epoch": 25.489032258064515,
+      "grad_norm": 0.02144439145922661,
+      "learning_rate": 2.990492653414002e-05,
+      "loss": 0.2669,
+      "step": 4925
+    },
+    {
+      "epoch": 25.4941935483871,
+      "grad_norm": 0.02770889736711979,
+      "learning_rate": 2.9870354364736387e-05,
+      "loss": 0.2912,
+      "step": 4926
+    },
+    {
+      "epoch": 25.499354838709678,
+      "grad_norm": 0.026472626253962517,
+      "learning_rate": 2.9835782195332757e-05,
+      "loss": 0.2775,
+      "step": 4927
+    },
+    {
+      "epoch": 25.504516129032258,
+      "grad_norm": 0.025066006928682327,
+      "learning_rate": 2.9801210025929128e-05,
+      "loss": 0.3013,
+      "step": 4928
+    },
+    {
+      "epoch": 25.509677419354837,
+      "grad_norm": 0.024162862449884415,
+      "learning_rate": 2.9766637856525498e-05,
+      "loss": 0.2579,
+      "step": 4929
+    },
+    {
+      "epoch": 25.51483870967742,
+      "grad_norm": 0.01867320016026497,
+      "learning_rate": 2.973206568712187e-05,
+      "loss": 0.2836,
+      "step": 4930
+    },
+    {
+      "epoch": 25.51483870967742,
+      "eval_loss": 3.3416199684143066,
+      "eval_runtime": 21.1749,
+      "eval_samples_per_second": 3.731,
+      "eval_steps_per_second": 0.472,
+      "step": 4930
+    },
+    {
+      "epoch": 25.52,
+      "grad_norm": 0.02196194790303707,
+      "learning_rate": 2.969749351771824e-05,
+      "loss": 0.2983,
+      "step": 4931
+    },
+    {
+      "epoch": 25.52516129032258,
+      "grad_norm": 0.017390912398695946,
+      "learning_rate": 2.966292134831461e-05,
+      "loss": 0.2949,
+      "step": 4932
+    },
+    {
+      "epoch": 25.530322580645162,
+      "grad_norm": 0.02737581357359886,
+      "learning_rate": 2.962834917891098e-05,
+      "loss": 0.291,
+      "step": 4933
+    },
+    {
+      "epoch": 25.53548387096774,
+      "grad_norm": 0.017877381294965744,
+      "learning_rate": 2.959377700950735e-05,
+      "loss": 0.2956,
+      "step": 4934
+    },
+    {
+      "epoch": 25.54064516129032,
+      "grad_norm": 0.03136872872710228,
+      "learning_rate": 2.9559204840103717e-05,
+      "loss": 0.2742,
+      "step": 4935
+    },
+    {
+      "epoch": 25.545806451612904,
+      "grad_norm": 0.03966499865055084,
+      "learning_rate": 2.9524632670700087e-05,
+      "loss": 0.261,
+      "step": 4936
+    },
+    {
+      "epoch": 25.550967741935484,
+      "grad_norm": 0.031122516840696335,
+      "learning_rate": 2.9490060501296458e-05,
+      "loss": 0.2956,
+      "step": 4937
+    },
+    {
+      "epoch": 25.556129032258063,
+      "grad_norm": 0.03811769559979439,
+      "learning_rate": 2.9455488331892828e-05,
+      "loss": 0.2749,
+      "step": 4938
+    },
+    {
+      "epoch": 25.561290322580646,
+      "grad_norm": 0.03248748928308487,
+      "learning_rate": 2.94209161624892e-05,
+      "loss": 0.2631,
+      "step": 4939
+    },
+    {
+      "epoch": 25.566451612903226,
+      "grad_norm": 0.02457324042916298,
+      "learning_rate": 2.938634399308557e-05,
+      "loss": 0.2828,
+      "step": 4940
+    },
+    {
+      "epoch": 25.571612903225805,
+      "grad_norm": 0.038903553038835526,
+      "learning_rate": 2.935177182368194e-05,
+      "loss": 0.2792,
+      "step": 4941
+    },
+    {
+      "epoch": 25.57677419354839,
+      "grad_norm": 0.022060327231884003,
+      "learning_rate": 2.931719965427831e-05,
+      "loss": 0.2198,
+      "step": 4942
+    },
+    {
+      "epoch": 25.581935483870968,
+      "grad_norm": 0.012824085541069508,
+      "learning_rate": 2.9282627484874677e-05,
+      "loss": 0.2628,
+      "step": 4943
+    },
+    {
+      "epoch": 25.587096774193547,
+      "grad_norm": 0.025639697909355164,
+      "learning_rate": 2.9248055315471047e-05,
+      "loss": 0.2929,
+      "step": 4944
+    },
+    {
+      "epoch": 25.59225806451613,
+      "grad_norm": 0.02547692321240902,
+      "learning_rate": 2.9213483146067417e-05,
+      "loss": 0.2945,
+      "step": 4945
+    },
+    {
+      "epoch": 25.59741935483871,
+      "grad_norm": 0.022831594571471214,
+      "learning_rate": 2.9178910976663788e-05,
+      "loss": 0.2703,
+      "step": 4946
+    },
+    {
+      "epoch": 25.60258064516129,
+      "grad_norm": 0.0323130302131176,
+      "learning_rate": 2.9144338807260158e-05,
+      "loss": 0.2856,
+      "step": 4947
+    },
+    {
+      "epoch": 25.607741935483872,
+      "grad_norm": 0.024827508255839348,
+      "learning_rate": 2.910976663785653e-05,
+      "loss": 0.3042,
+      "step": 4948
+    },
+    {
+      "epoch": 25.612903225806452,
+      "grad_norm": 0.020206045359373093,
+      "learning_rate": 2.90751944684529e-05,
+      "loss": 0.2709,
+      "step": 4949
+    },
+    {
+      "epoch": 25.61806451612903,
+      "grad_norm": 0.021752173081040382,
+      "learning_rate": 2.904062229904927e-05,
+      "loss": 0.2743,
+      "step": 4950
+    },
+    {
+      "epoch": 25.623225806451615,
+      "grad_norm": 0.03036498837172985,
+      "learning_rate": 2.9006050129645633e-05,
+      "loss": 0.3079,
+      "step": 4951
+    },
+    {
+      "epoch": 25.628387096774194,
+      "grad_norm": 0.030204709619283676,
+      "learning_rate": 2.8971477960242007e-05,
+      "loss": 0.2776,
+      "step": 4952
+    },
+    {
+      "epoch": 25.633548387096774,
+      "grad_norm": 0.02790232188999653,
+      "learning_rate": 2.8936905790838377e-05,
+      "loss": 0.2664,
+      "step": 4953
+    },
+    {
+      "epoch": 25.638709677419357,
+      "grad_norm": 0.020262716338038445,
+      "learning_rate": 2.8902333621434747e-05,
+      "loss": 0.2926,
+      "step": 4954
+    },
+    {
+      "epoch": 25.643870967741936,
+      "grad_norm": 0.026114294305443764,
+      "learning_rate": 2.8867761452031118e-05,
+      "loss": 0.2906,
+      "step": 4955
+    },
+    {
+      "epoch": 25.649032258064516,
+      "grad_norm": 0.03378100320696831,
+      "learning_rate": 2.8833189282627488e-05,
+      "loss": 0.2824,
+      "step": 4956
+    },
+    {
+      "epoch": 25.654193548387095,
+      "grad_norm": 0.02722490206360817,
+      "learning_rate": 2.879861711322386e-05,
+      "loss": 0.2718,
+      "step": 4957
+    },
+    {
+      "epoch": 25.659354838709678,
+      "grad_norm": 0.02869894541800022,
+      "learning_rate": 2.876404494382023e-05,
+      "loss": 0.2911,
+      "step": 4958
+    },
+    {
+      "epoch": 25.664516129032258,
+      "grad_norm": 0.03287602961063385,
+      "learning_rate": 2.8729472774416592e-05,
+      "loss": 0.2852,
+      "step": 4959
+    },
+    {
+      "epoch": 25.669677419354837,
+      "grad_norm": 0.02635756880044937,
+      "learning_rate": 2.8694900605012963e-05,
+      "loss": 0.2902,
+      "step": 4960
+    },
+    {
+      "epoch": 25.67483870967742,
+      "grad_norm": 0.0168833676725626,
+      "learning_rate": 2.8660328435609333e-05,
+      "loss": 0.2915,
+      "step": 4961
+    },
+    {
+      "epoch": 25.68,
+      "grad_norm": 0.02571439929306507,
+      "learning_rate": 2.8625756266205707e-05,
+      "loss": 0.2785,
+      "step": 4962
+    },
+    {
+      "epoch": 25.68516129032258,
+      "grad_norm": 0.026331644505262375,
+      "learning_rate": 2.8591184096802077e-05,
+      "loss": 0.2933,
+      "step": 4963
+    },
+    {
+      "epoch": 25.690322580645162,
+      "grad_norm": 0.02835226058959961,
+      "learning_rate": 2.8556611927398448e-05,
+      "loss": 0.2693,
+      "step": 4964
+    },
+    {
+      "epoch": 25.695483870967742,
+      "grad_norm": 0.02747434191405773,
+      "learning_rate": 2.8522039757994818e-05,
+      "loss": 0.2775,
+      "step": 4965
+    },
+    {
+      "epoch": 25.70064516129032,
+      "grad_norm": 0.031617216765880585,
+      "learning_rate": 2.848746758859119e-05,
+      "loss": 0.2906,
+      "step": 4966
+    },
+    {
+      "epoch": 25.705806451612904,
+      "grad_norm": 0.03253605589270592,
+      "learning_rate": 2.8452895419187552e-05,
+      "loss": 0.2833,
+      "step": 4967
+    },
+    {
+      "epoch": 25.710967741935484,
+      "grad_norm": 0.03054051287472248,
+      "learning_rate": 2.8418323249783922e-05,
+      "loss": 0.2904,
+      "step": 4968
+    },
+    {
+      "epoch": 25.716129032258063,
+      "grad_norm": 0.03341417759656906,
+      "learning_rate": 2.8383751080380293e-05,
+      "loss": 0.2907,
+      "step": 4969
+    },
+    {
+      "epoch": 25.721290322580646,
+      "grad_norm": 0.02559356763958931,
+      "learning_rate": 2.8349178910976663e-05,
+      "loss": 0.278,
+      "step": 4970
+    },
+    {
+      "epoch": 25.726451612903226,
+      "grad_norm": 0.018665654584765434,
+      "learning_rate": 2.8314606741573037e-05,
+      "loss": 0.26,
+      "step": 4971
+    },
+    {
+      "epoch": 25.731612903225805,
+      "grad_norm": 0.03677509352564812,
+      "learning_rate": 2.8280034572169407e-05,
+      "loss": 0.2901,
+      "step": 4972
+    },
+    {
+      "epoch": 25.73677419354839,
+      "grad_norm": 0.025261640548706055,
+      "learning_rate": 2.8245462402765778e-05,
+      "loss": 0.296,
+      "step": 4973
+    },
+    {
+      "epoch": 25.741935483870968,
+      "grad_norm": 0.0219507347792387,
+      "learning_rate": 2.8210890233362148e-05,
+      "loss": 0.2868,
+      "step": 4974
+    },
+    {
+      "epoch": 25.747096774193547,
+      "grad_norm": 0.025778932496905327,
+      "learning_rate": 2.817631806395851e-05,
+      "loss": 0.2945,
+      "step": 4975
+    },
+    {
+      "epoch": 25.75225806451613,
+      "grad_norm": 0.02132287807762623,
+      "learning_rate": 2.8141745894554882e-05,
+      "loss": 0.2991,
+      "step": 4976
+    },
+    {
+      "epoch": 25.75741935483871,
+      "grad_norm": 0.02804333157837391,
+      "learning_rate": 2.8107173725151252e-05,
+      "loss": 0.2952,
+      "step": 4977
+    },
+    {
+      "epoch": 25.76258064516129,
+      "grad_norm": 0.029567280784249306,
+      "learning_rate": 2.8072601555747623e-05,
+      "loss": 0.2989,
+      "step": 4978
+    },
+    {
+      "epoch": 25.767741935483873,
+      "grad_norm": 0.02279028482735157,
+      "learning_rate": 2.8038029386343993e-05,
+      "loss": 0.2909,
+      "step": 4979
+    },
+    {
+      "epoch": 25.772903225806452,
+      "grad_norm": 0.02194342575967312,
+      "learning_rate": 2.8003457216940367e-05,
+      "loss": 0.2924,
+      "step": 4980
+    },
+    {
+      "epoch": 25.77806451612903,
+      "grad_norm": 0.042160678654909134,
+      "learning_rate": 2.7968885047536737e-05,
+      "loss": 0.2952,
+      "step": 4981
+    },
+    {
+      "epoch": 25.78322580645161,
+      "grad_norm": 0.018985047936439514,
+      "learning_rate": 2.7934312878133108e-05,
+      "loss": 0.264,
+      "step": 4982
+    },
+    {
+      "epoch": 25.788387096774194,
+      "grad_norm": 0.030838189646601677,
+      "learning_rate": 2.789974070872947e-05,
+      "loss": 0.2962,
+      "step": 4983
+    },
+    {
+      "epoch": 25.793548387096774,
+      "grad_norm": 0.018628466874361038,
+      "learning_rate": 2.786516853932584e-05,
+      "loss": 0.3039,
+      "step": 4984
+    },
+    {
+      "epoch": 25.798709677419353,
+      "grad_norm": 0.023005064576864243,
+      "learning_rate": 2.7830596369922212e-05,
+      "loss": 0.248,
+      "step": 4985
+    },
+    {
+      "epoch": 25.803870967741936,
+      "grad_norm": 0.019832709804177284,
+      "learning_rate": 2.7796024200518582e-05,
+      "loss": 0.2898,
+      "step": 4986
+    },
+    {
+      "epoch": 25.809032258064516,
+      "grad_norm": 0.03335195034742355,
+      "learning_rate": 2.7761452031114953e-05,
+      "loss": 0.2598,
+      "step": 4987
+    },
+    {
+      "epoch": 25.814193548387095,
+      "grad_norm": 0.03431616723537445,
+      "learning_rate": 2.7726879861711323e-05,
+      "loss": 0.2698,
+      "step": 4988
+    },
+    {
+      "epoch": 25.81935483870968,
+      "grad_norm": 0.028159720823168755,
+      "learning_rate": 2.7692307692307694e-05,
+      "loss": 0.3018,
+      "step": 4989
+    },
+    {
+      "epoch": 25.824516129032258,
+      "grad_norm": 0.01753964275121689,
+      "learning_rate": 2.7657735522904067e-05,
+      "loss": 0.3141,
+      "step": 4990
+    },
+    {
+      "epoch": 25.829677419354837,
+      "grad_norm": 0.027113554999232292,
+      "learning_rate": 2.762316335350043e-05,
+      "loss": 0.2771,
+      "step": 4991
+    },
+    {
+      "epoch": 25.83483870967742,
+      "grad_norm": 0.02955162711441517,
+      "learning_rate": 2.75885911840968e-05,
+      "loss": 0.3057,
+      "step": 4992
+    },
+    {
+      "epoch": 25.84,
+      "grad_norm": 0.03287277743220329,
+      "learning_rate": 2.755401901469317e-05,
+      "loss": 0.3016,
+      "step": 4993
+    },
+    {
+      "epoch": 25.84516129032258,
+      "grad_norm": 0.027672912925481796,
+      "learning_rate": 2.7519446845289542e-05,
+      "loss": 0.298,
+      "step": 4994
+    },
+    {
+      "epoch": 25.850322580645162,
+      "grad_norm": 0.027836183086037636,
+      "learning_rate": 2.7484874675885912e-05,
+      "loss": 0.2979,
+      "step": 4995
+    },
+    {
+      "epoch": 25.855483870967742,
+      "grad_norm": 0.021570321172475815,
+      "learning_rate": 2.7450302506482283e-05,
+      "loss": 0.2936,
+      "step": 4996
+    },
+    {
+      "epoch": 25.86064516129032,
+      "grad_norm": 0.026424121111631393,
+      "learning_rate": 2.7415730337078653e-05,
+      "loss": 0.2775,
+      "step": 4997
+    },
+    {
+      "epoch": 25.865806451612904,
+      "grad_norm": 0.034337833523750305,
+      "learning_rate": 2.7381158167675024e-05,
+      "loss": 0.2555,
+      "step": 4998
+    },
+    {
+      "epoch": 25.870967741935484,
+      "grad_norm": 0.03136303275823593,
+      "learning_rate": 2.734658599827139e-05,
+      "loss": 0.2835,
+      "step": 4999
+    },
+    {
+      "epoch": 25.876129032258063,
+      "grad_norm": 0.02129243314266205,
+      "learning_rate": 2.731201382886776e-05,
+      "loss": 0.2484,
+      "step": 5000
+    },
+    {
+      "epoch": 25.881290322580647,
+      "grad_norm": 0.026196694001555443,
+      "learning_rate": 2.727744165946413e-05,
+      "loss": 0.2809,
+      "step": 5001
+    },
+    {
+      "epoch": 25.886451612903226,
+      "grad_norm": 0.025988955050706863,
+      "learning_rate": 2.72428694900605e-05,
+      "loss": 0.2814,
+      "step": 5002
+    },
+    {
+      "epoch": 25.891612903225806,
+      "grad_norm": 0.025400839745998383,
+      "learning_rate": 2.7208297320656872e-05,
+      "loss": 0.2417,
+      "step": 5003
+    },
+    {
+      "epoch": 25.89677419354839,
+      "grad_norm": 0.03339563310146332,
+      "learning_rate": 2.7173725151253242e-05,
+      "loss": 0.2937,
+      "step": 5004
+    },
+    {
+      "epoch": 25.901935483870968,
+      "grad_norm": 0.01674491912126541,
+      "learning_rate": 2.7139152981849613e-05,
+      "loss": 0.2993,
+      "step": 5005
+    },
+    {
+      "epoch": 25.907096774193548,
+      "grad_norm": 0.028927097097039223,
+      "learning_rate": 2.7104580812445983e-05,
+      "loss": 0.2542,
+      "step": 5006
+    },
+    {
+      "epoch": 25.91225806451613,
+      "grad_norm": 0.028809677809476852,
+      "learning_rate": 2.7070008643042353e-05,
+      "loss": 0.314,
+      "step": 5007
+    },
+    {
+      "epoch": 25.91741935483871,
+      "grad_norm": 0.03750497102737427,
+      "learning_rate": 2.703543647363872e-05,
+      "loss": 0.2871,
+      "step": 5008
+    },
+    {
+      "epoch": 25.92258064516129,
+      "grad_norm": 0.01993640325963497,
+      "learning_rate": 2.700086430423509e-05,
+      "loss": 0.2964,
+      "step": 5009
+    },
+    {
+      "epoch": 25.927741935483873,
+      "grad_norm": 0.03030923567712307,
+      "learning_rate": 2.696629213483146e-05,
+      "loss": 0.2807,
+      "step": 5010
+    },
+    {
+      "epoch": 25.932903225806452,
+      "grad_norm": 0.02740771882236004,
+      "learning_rate": 2.693171996542783e-05,
+      "loss": 0.2622,
+      "step": 5011
+    },
+    {
+      "epoch": 25.93806451612903,
+      "grad_norm": 0.02459333837032318,
+      "learning_rate": 2.6897147796024202e-05,
+      "loss": 0.2938,
+      "step": 5012
+    },
+    {
+      "epoch": 25.94322580645161,
+      "grad_norm": 0.01254823338240385,
+      "learning_rate": 2.6862575626620572e-05,
+      "loss": 0.277,
+      "step": 5013
+    },
+    {
+      "epoch": 25.948387096774194,
+      "grad_norm": 0.03390992805361748,
+      "learning_rate": 2.6828003457216943e-05,
+      "loss": 0.3032,
+      "step": 5014
+    },
+    {
+      "epoch": 25.953548387096774,
+      "grad_norm": 0.024940960109233856,
+      "learning_rate": 2.6793431287813313e-05,
+      "loss": 0.2886,
+      "step": 5015
+    },
+    {
+      "epoch": 25.958709677419353,
+      "grad_norm": 0.02526767924427986,
+      "learning_rate": 2.675885911840968e-05,
+      "loss": 0.289,
+      "step": 5016
+    },
+    {
+      "epoch": 25.963870967741936,
+      "grad_norm": 0.017741575837135315,
+      "learning_rate": 2.672428694900605e-05,
+      "loss": 0.2573,
+      "step": 5017
+    },
+    {
+      "epoch": 25.969032258064516,
+      "grad_norm": 0.03575950488448143,
+      "learning_rate": 2.668971477960242e-05,
+      "loss": 0.2529,
+      "step": 5018
+    },
+    {
+      "epoch": 25.974193548387095,
+      "grad_norm": 0.024482127279043198,
+      "learning_rate": 2.665514261019879e-05,
+      "loss": 0.2337,
+      "step": 5019
+    },
+    {
+      "epoch": 25.97935483870968,
+      "grad_norm": 0.02081799879670143,
+      "learning_rate": 2.662057044079516e-05,
+      "loss": 0.2992,
+      "step": 5020
+    },
+    {
+      "epoch": 25.984516129032258,
+      "grad_norm": 0.032717857509851456,
+      "learning_rate": 2.6585998271391532e-05,
+      "loss": 0.2883,
+      "step": 5021
+    },
+    {
+      "epoch": 25.989677419354837,
+      "grad_norm": 0.03491813689470291,
+      "learning_rate": 2.6551426101987902e-05,
+      "loss": 0.3048,
+      "step": 5022
+    },
+    {
+      "epoch": 25.99483870967742,
+      "grad_norm": 0.03236105293035507,
+      "learning_rate": 2.6516853932584273e-05,
+      "loss": 0.2802,
+      "step": 5023
+    },
+    {
+      "epoch": 26.001290322580644,
+      "grad_norm": 0.05422203987836838,
+      "learning_rate": 2.648228176318064e-05,
+      "loss": 0.4219,
+      "step": 5024
+    },
+    {
+      "epoch": 26.006451612903227,
+      "grad_norm": 0.021458743140101433,
+      "learning_rate": 2.644770959377701e-05,
+      "loss": 0.3022,
+      "step": 5025
+    },
+    {
+      "epoch": 26.011612903225807,
+      "grad_norm": 0.029737919569015503,
+      "learning_rate": 2.641313742437338e-05,
+      "loss": 0.2594,
+      "step": 5026
+    },
+    {
+      "epoch": 26.016774193548386,
+      "grad_norm": 0.028481924906373024,
+      "learning_rate": 2.637856525496975e-05,
+      "loss": 0.3025,
+      "step": 5027
+    },
+    {
+      "epoch": 26.02193548387097,
+      "grad_norm": 0.019640548154711723,
+      "learning_rate": 2.634399308556612e-05,
+      "loss": 0.2704,
+      "step": 5028
+    },
+    {
+      "epoch": 26.02709677419355,
+      "grad_norm": 0.020789561793208122,
+      "learning_rate": 2.630942091616249e-05,
+      "loss": 0.2629,
+      "step": 5029
+    },
+    {
+      "epoch": 26.032258064516128,
+      "grad_norm": 0.020942620933055878,
+      "learning_rate": 2.6274848746758862e-05,
+      "loss": 0.2752,
+      "step": 5030
+    },
+    {
+      "epoch": 26.03741935483871,
+      "grad_norm": 0.02292122319340706,
+      "learning_rate": 2.6240276577355232e-05,
+      "loss": 0.2983,
+      "step": 5031
+    },
+    {
+      "epoch": 26.04258064516129,
+      "grad_norm": 0.025462854653596878,
+      "learning_rate": 2.62057044079516e-05,
+      "loss": 0.2804,
+      "step": 5032
+    },
+    {
+      "epoch": 26.04774193548387,
+      "grad_norm": 0.019412336871027946,
+      "learning_rate": 2.617113223854797e-05,
+      "loss": 0.223,
+      "step": 5033
+    },
+    {
+      "epoch": 26.052903225806453,
+      "grad_norm": 0.03124912828207016,
+      "learning_rate": 2.613656006914434e-05,
+      "loss": 0.2864,
+      "step": 5034
+    },
+    {
+      "epoch": 26.058064516129033,
+      "grad_norm": 0.03574112802743912,
+      "learning_rate": 2.610198789974071e-05,
+      "loss": 0.2601,
+      "step": 5035
+    },
+    {
+      "epoch": 26.063225806451612,
+      "grad_norm": 0.031705696135759354,
+      "learning_rate": 2.606741573033708e-05,
+      "loss": 0.3039,
+      "step": 5036
+    },
+    {
+      "epoch": 26.068387096774195,
+      "grad_norm": 0.031786803156137466,
+      "learning_rate": 2.603284356093345e-05,
+      "loss": 0.2856,
+      "step": 5037
+    },
+    {
+      "epoch": 26.073548387096775,
+      "grad_norm": 0.0239912960678339,
+      "learning_rate": 2.599827139152982e-05,
+      "loss": 0.2792,
+      "step": 5038
+    },
+    {
+      "epoch": 26.078709677419354,
+      "grad_norm": 0.03086407296359539,
+      "learning_rate": 2.5963699222126192e-05,
+      "loss": 0.2937,
+      "step": 5039
+    },
+    {
+      "epoch": 26.083870967741934,
+      "grad_norm": 0.027276402339339256,
+      "learning_rate": 2.5929127052722555e-05,
+      "loss": 0.2916,
+      "step": 5040
+    },
+    {
+      "epoch": 26.089032258064517,
+      "grad_norm": 0.01986819878220558,
+      "learning_rate": 2.589455488331893e-05,
+      "loss": 0.275,
+      "step": 5041
+    },
+    {
+      "epoch": 26.094193548387096,
+      "grad_norm": 0.016831379383802414,
+      "learning_rate": 2.58599827139153e-05,
+      "loss": 0.2914,
+      "step": 5042
+    },
+    {
+      "epoch": 26.099354838709676,
+      "grad_norm": 0.03260750696063042,
+      "learning_rate": 2.582541054451167e-05,
+      "loss": 0.3025,
+      "step": 5043
+    },
+    {
+      "epoch": 26.10451612903226,
+      "grad_norm": 0.02231096662580967,
+      "learning_rate": 2.579083837510804e-05,
+      "loss": 0.2797,
+      "step": 5044
+    },
+    {
+      "epoch": 26.10967741935484,
+      "grad_norm": 0.028718847781419754,
+      "learning_rate": 2.575626620570441e-05,
+      "loss": 0.2954,
+      "step": 5045
+    },
+    {
+      "epoch": 26.114838709677418,
+      "grad_norm": 0.018761124461889267,
+      "learning_rate": 2.572169403630078e-05,
+      "loss": 0.2509,
+      "step": 5046
+    },
+    {
+      "epoch": 26.12,
+      "grad_norm": 0.039033155888319016,
+      "learning_rate": 2.568712186689715e-05,
+      "loss": 0.2335,
+      "step": 5047
+    },
+    {
+      "epoch": 26.12516129032258,
+      "grad_norm": 0.03331764414906502,
+      "learning_rate": 2.5652549697493515e-05,
+      "loss": 0.2818,
+      "step": 5048
+    },
+    {
+      "epoch": 26.13032258064516,
+      "grad_norm": 0.02080392651259899,
+      "learning_rate": 2.5617977528089885e-05,
+      "loss": 0.296,
+      "step": 5049
+    },
+    {
+      "epoch": 26.135483870967743,
+      "grad_norm": 0.025441154837608337,
+      "learning_rate": 2.558340535868626e-05,
+      "loss": 0.2724,
+      "step": 5050
+    },
+    {
+      "epoch": 26.140645161290323,
+      "grad_norm": 0.023995021358132362,
+      "learning_rate": 2.554883318928263e-05,
+      "loss": 0.2873,
+      "step": 5051
+    },
+    {
+      "epoch": 26.145806451612902,
+      "grad_norm": 0.027870889753103256,
+      "learning_rate": 2.5514261019879e-05,
+      "loss": 0.2887,
+      "step": 5052
+    },
+    {
+      "epoch": 26.150967741935485,
+      "grad_norm": 0.026800477877259254,
+      "learning_rate": 2.547968885047537e-05,
+      "loss": 0.3125,
+      "step": 5053
+    },
+    {
+      "epoch": 26.156129032258065,
+      "grad_norm": 0.020607564598321915,
+      "learning_rate": 2.544511668107174e-05,
+      "loss": 0.2991,
+      "step": 5054
+    },
+    {
+      "epoch": 26.161290322580644,
+      "grad_norm": 0.023993466049432755,
+      "learning_rate": 2.541054451166811e-05,
+      "loss": 0.2773,
+      "step": 5055
+    },
+    {
+      "epoch": 26.166451612903227,
+      "grad_norm": 0.026607539504766464,
+      "learning_rate": 2.5375972342264475e-05,
+      "loss": 0.2415,
+      "step": 5056
+    },
+    {
+      "epoch": 26.171612903225807,
+      "grad_norm": 0.02137015014886856,
+      "learning_rate": 2.5341400172860845e-05,
+      "loss": 0.2646,
+      "step": 5057
+    },
+    {
+      "epoch": 26.176774193548386,
+      "grad_norm": 0.021952934563159943,
+      "learning_rate": 2.5306828003457215e-05,
+      "loss": 0.2413,
+      "step": 5058
+    },
+    {
+      "epoch": 26.18193548387097,
+      "grad_norm": 0.024296097457408905,
+      "learning_rate": 2.527225583405359e-05,
+      "loss": 0.2864,
+      "step": 5059
+    },
+    {
+      "epoch": 26.18709677419355,
+      "grad_norm": 0.022220319136977196,
+      "learning_rate": 2.523768366464996e-05,
+      "loss": 0.3032,
+      "step": 5060
+    },
+    {
+      "epoch": 26.19225806451613,
+      "grad_norm": 0.02732829749584198,
+      "learning_rate": 2.520311149524633e-05,
+      "loss": 0.2758,
+      "step": 5061
+    },
+    {
+      "epoch": 26.19741935483871,
+      "grad_norm": 0.026985349133610725,
+      "learning_rate": 2.51685393258427e-05,
+      "loss": 0.272,
+      "step": 5062
+    },
+    {
+      "epoch": 26.20258064516129,
+      "grad_norm": 0.026289423927664757,
+      "learning_rate": 2.513396715643907e-05,
+      "loss": 0.2773,
+      "step": 5063
+    },
+    {
+      "epoch": 26.20774193548387,
+      "grad_norm": 0.017776185646653175,
+      "learning_rate": 2.5099394987035434e-05,
+      "loss": 0.2802,
+      "step": 5064
+    },
+    {
+      "epoch": 26.21290322580645,
+      "grad_norm": 0.020172422751784325,
+      "learning_rate": 2.5064822817631805e-05,
+      "loss": 0.2831,
+      "step": 5065
+    },
+    {
+      "epoch": 26.218064516129033,
+      "grad_norm": 0.03073960356414318,
+      "learning_rate": 2.5030250648228175e-05,
+      "loss": 0.2985,
+      "step": 5066
+    },
+    {
+      "epoch": 26.223225806451612,
+      "grad_norm": 0.012851755134761333,
+      "learning_rate": 2.4995678478824545e-05,
+      "loss": 0.2925,
+      "step": 5067
+    },
+    {
+      "epoch": 26.228387096774192,
+      "grad_norm": 0.024941062554717064,
+      "learning_rate": 2.4961106309420916e-05,
+      "loss": 0.2748,
+      "step": 5068
+    },
+    {
+      "epoch": 26.233548387096775,
+      "grad_norm": 0.02666398510336876,
+      "learning_rate": 2.492653414001729e-05,
+      "loss": 0.2698,
+      "step": 5069
+    },
+    {
+      "epoch": 26.238709677419354,
+      "grad_norm": 0.02476966381072998,
+      "learning_rate": 2.4891961970613656e-05,
+      "loss": 0.2869,
+      "step": 5070
+    },
+    {
+      "epoch": 26.243870967741934,
+      "grad_norm": 0.03069852478802204,
+      "learning_rate": 2.4857389801210027e-05,
+      "loss": 0.2608,
+      "step": 5071
+    },
+    {
+      "epoch": 26.249032258064517,
+      "grad_norm": 0.030030153691768646,
+      "learning_rate": 2.4822817631806397e-05,
+      "loss": 0.282,
+      "step": 5072
+    },
+    {
+      "epoch": 26.254193548387097,
+      "grad_norm": 0.030903391540050507,
+      "learning_rate": 2.4788245462402768e-05,
+      "loss": 0.2903,
+      "step": 5073
+    },
+    {
+      "epoch": 26.259354838709676,
+      "grad_norm": 0.027770640328526497,
+      "learning_rate": 2.4753673292999135e-05,
+      "loss": 0.286,
+      "step": 5074
+    },
+    {
+      "epoch": 26.26451612903226,
+      "grad_norm": 0.019711460918188095,
+      "learning_rate": 2.4719101123595505e-05,
+      "loss": 0.2822,
+      "step": 5075
+    },
+    {
+      "epoch": 26.26967741935484,
+      "grad_norm": 0.02307111769914627,
+      "learning_rate": 2.4684528954191875e-05,
+      "loss": 0.2798,
+      "step": 5076
+    },
+    {
+      "epoch": 26.274838709677418,
+      "grad_norm": 0.01801329292356968,
+      "learning_rate": 2.4649956784788246e-05,
+      "loss": 0.2719,
+      "step": 5077
+    },
+    {
+      "epoch": 26.28,
+      "grad_norm": 0.01758679375052452,
+      "learning_rate": 2.461538461538462e-05,
+      "loss": 0.2882,
+      "step": 5078
+    },
+    {
+      "epoch": 26.28516129032258,
+      "grad_norm": 0.028662247583270073,
+      "learning_rate": 2.4580812445980986e-05,
+      "loss": 0.2894,
+      "step": 5079
+    },
+    {
+      "epoch": 26.29032258064516,
+      "grad_norm": 0.013499297201633453,
+      "learning_rate": 2.4546240276577357e-05,
+      "loss": 0.2918,
+      "step": 5080
+    },
+    {
+      "epoch": 26.295483870967743,
+      "grad_norm": 0.02025272697210312,
+      "learning_rate": 2.4511668107173727e-05,
+      "loss": 0.2697,
+      "step": 5081
+    },
+    {
+      "epoch": 26.300645161290323,
+      "grad_norm": 0.018671996891498566,
+      "learning_rate": 2.4477095937770098e-05,
+      "loss": 0.2753,
+      "step": 5082
+    },
+    {
+      "epoch": 26.305806451612902,
+      "grad_norm": 0.02253464236855507,
+      "learning_rate": 2.4442523768366465e-05,
+      "loss": 0.2905,
+      "step": 5083
+    },
+    {
+      "epoch": 26.310967741935485,
+      "grad_norm": 0.025629157200455666,
+      "learning_rate": 2.4407951598962835e-05,
+      "loss": 0.2812,
+      "step": 5084
+    },
+    {
+      "epoch": 26.316129032258065,
+      "grad_norm": 0.034222427755594254,
+      "learning_rate": 2.4373379429559205e-05,
+      "loss": 0.2822,
+      "step": 5085
+    },
+    {
+      "epoch": 26.321290322580644,
+      "grad_norm": 0.03047211840748787,
+      "learning_rate": 2.4338807260155576e-05,
+      "loss": 0.2612,
+      "step": 5086
+    },
+    {
+      "epoch": 26.326451612903227,
+      "grad_norm": 0.025568487122654915,
+      "learning_rate": 2.4304235090751946e-05,
+      "loss": 0.2861,
+      "step": 5087
+    },
+    {
+      "epoch": 26.331612903225807,
+      "grad_norm": 0.02075038105249405,
+      "learning_rate": 2.4269662921348316e-05,
+      "loss": 0.2847,
+      "step": 5088
+    },
+    {
+      "epoch": 26.336774193548386,
+      "grad_norm": 0.021189220249652863,
+      "learning_rate": 2.4235090751944687e-05,
+      "loss": 0.3076,
+      "step": 5089
+    },
+    {
+      "epoch": 26.34193548387097,
+      "grad_norm": 0.030519669875502586,
+      "learning_rate": 2.4200518582541057e-05,
+      "loss": 0.2862,
+      "step": 5090
+    },
+    {
+      "epoch": 26.34709677419355,
+      "grad_norm": 0.024235054850578308,
+      "learning_rate": 2.4165946413137424e-05,
+      "loss": 0.2626,
+      "step": 5091
+    },
+    {
+      "epoch": 26.35225806451613,
+      "grad_norm": 0.018903762102127075,
+      "learning_rate": 2.4131374243733795e-05,
+      "loss": 0.2868,
+      "step": 5092
+    },
+    {
+      "epoch": 26.35741935483871,
+      "grad_norm": 0.023301344364881516,
+      "learning_rate": 2.4096802074330165e-05,
+      "loss": 0.2982,
+      "step": 5093
+    },
+    {
+      "epoch": 26.36258064516129,
+      "grad_norm": 0.02418130822479725,
+      "learning_rate": 2.4062229904926535e-05,
+      "loss": 0.2988,
+      "step": 5094
+    },
+    {
+      "epoch": 26.36774193548387,
+      "grad_norm": 0.02879425883293152,
+      "learning_rate": 2.4027657735522906e-05,
+      "loss": 0.2911,
+      "step": 5095
+    },
+    {
+      "epoch": 26.37290322580645,
+      "grad_norm": 0.01806686259806156,
+      "learning_rate": 2.3993085566119276e-05,
+      "loss": 0.2469,
+      "step": 5096
+    },
+    {
+      "epoch": 26.378064516129033,
+      "grad_norm": 0.033989399671554565,
+      "learning_rate": 2.3958513396715646e-05,
+      "loss": 0.2497,
+      "step": 5097
+    },
+    {
+      "epoch": 26.383225806451613,
+      "grad_norm": 0.03214331343770027,
+      "learning_rate": 2.3923941227312017e-05,
+      "loss": 0.2841,
+      "step": 5098
+    },
+    {
+      "epoch": 26.388387096774192,
+      "grad_norm": 0.017315732315182686,
+      "learning_rate": 2.3889369057908384e-05,
+      "loss": 0.2658,
+      "step": 5099
+    },
+    {
+      "epoch": 26.393548387096775,
+      "grad_norm": 0.03736332431435585,
+      "learning_rate": 2.3854796888504754e-05,
+      "loss": 0.2698,
+      "step": 5100
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 5790,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 30,
+  "save_steps": 300,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.1100587970841805e+19,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}