diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,8687 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 5.0,
+  "eval_steps": 500,
+  "global_step": 1235,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.004048582995951417,
+      "grad_norm": 6.099686321364834,
+      "learning_rate": 3.2258064516129035e-07,
+      "loss": 0.8081,
+      "step": 1
+    },
+    {
+      "epoch": 0.008097165991902834,
+      "grad_norm": 6.118302739760215,
+      "learning_rate": 6.451612903225807e-07,
+      "loss": 0.8214,
+      "step": 2
+    },
+    {
+      "epoch": 0.012145748987854251,
+      "grad_norm": 6.145298613231104,
+      "learning_rate": 9.67741935483871e-07,
+      "loss": 0.8151,
+      "step": 3
+    },
+    {
+      "epoch": 0.016194331983805668,
+      "grad_norm": 5.643256767430969,
+      "learning_rate": 1.2903225806451614e-06,
+      "loss": 0.7681,
+      "step": 4
+    },
+    {
+      "epoch": 0.020242914979757085,
+      "grad_norm": 5.73156490181822,
+      "learning_rate": 1.6129032258064516e-06,
+      "loss": 0.7819,
+      "step": 5
+    },
+    {
+      "epoch": 0.024291497975708502,
+      "grad_norm": 5.4747720867021465,
+      "learning_rate": 1.935483870967742e-06,
+      "loss": 0.768,
+      "step": 6
+    },
+    {
+      "epoch": 0.02834008097165992,
+      "grad_norm": 4.470556007375453,
+      "learning_rate": 2.2580645161290324e-06,
+      "loss": 0.7746,
+      "step": 7
+    },
+    {
+      "epoch": 0.032388663967611336,
+      "grad_norm": 5.058379051900018,
+      "learning_rate": 2.580645161290323e-06,
+      "loss": 0.8296,
+      "step": 8
+    },
+    {
+      "epoch": 0.03643724696356275,
+      "grad_norm": 2.482173195760078,
+      "learning_rate": 2.903225806451613e-06,
+      "loss": 0.694,
+      "step": 9
+    },
+    {
+      "epoch": 0.04048582995951417,
+      "grad_norm": 2.3264249898931064,
+      "learning_rate": 3.225806451612903e-06,
+      "loss": 0.6891,
+      "step": 10
+    },
+    {
+      "epoch": 0.044534412955465584,
+      "grad_norm": 2.1279616751403325,
+      "learning_rate": 3.548387096774194e-06,
+      "loss": 0.6786,
+      "step": 11
+    },
+    {
+      "epoch": 0.048582995951417005,
+      "grad_norm": 3.361830164862254,
+      "learning_rate": 3.870967741935484e-06,
+      "loss": 0.684,
+      "step": 12
+    },
+    {
+      "epoch": 0.05263157894736842,
+      "grad_norm": 3.5385700810791505,
+      "learning_rate": 4.193548387096774e-06,
+      "loss": 0.6752,
+      "step": 13
+    },
+    {
+      "epoch": 0.05668016194331984,
+      "grad_norm": 3.646060489315829,
+      "learning_rate": 4.516129032258065e-06,
+      "loss": 0.6701,
+      "step": 14
+    },
+    {
+      "epoch": 0.06072874493927125,
+      "grad_norm": 3.1422825999108137,
+      "learning_rate": 4.838709677419355e-06,
+      "loss": 0.6318,
+      "step": 15
+    },
+    {
+      "epoch": 0.06477732793522267,
+      "grad_norm": 2.6534839933685013,
+      "learning_rate": 5.161290322580646e-06,
+      "loss": 0.5967,
+      "step": 16
+    },
+    {
+      "epoch": 0.06882591093117409,
+      "grad_norm": 2.742859439780988,
+      "learning_rate": 5.483870967741935e-06,
+      "loss": 0.6165,
+      "step": 17
+    },
+    {
+      "epoch": 0.0728744939271255,
+      "grad_norm": 2.0637507483661253,
+      "learning_rate": 5.806451612903226e-06,
+      "loss": 0.5977,
+      "step": 18
+    },
+    {
+      "epoch": 0.07692307692307693,
+      "grad_norm": 1.5893505525467324,
+      "learning_rate": 6.129032258064517e-06,
+      "loss": 0.5952,
+      "step": 19
+    },
+    {
+      "epoch": 0.08097165991902834,
+      "grad_norm": 1.38172307409863,
+      "learning_rate": 6.451612903225806e-06,
+      "loss": 0.5764,
+      "step": 20
+    },
+    {
+      "epoch": 0.08502024291497975,
+      "grad_norm": 1.6806823933847075,
+      "learning_rate": 6.774193548387097e-06,
+      "loss": 0.5695,
+      "step": 21
+    },
+    {
+      "epoch": 0.08906882591093117,
+      "grad_norm": 1.843224104424946,
+      "learning_rate": 7.096774193548388e-06,
+      "loss": 0.5419,
+      "step": 22
+    },
+    {
+      "epoch": 0.0931174089068826,
+      "grad_norm": 1.766776871199825,
+      "learning_rate": 7.4193548387096784e-06,
+      "loss": 0.5357,
+      "step": 23
+    },
+    {
+      "epoch": 0.09716599190283401,
+      "grad_norm": 1.8652221324780696,
+      "learning_rate": 7.741935483870968e-06,
+      "loss": 0.57,
+      "step": 24
+    },
+    {
+      "epoch": 0.10121457489878542,
+      "grad_norm": 1.1207692730516352,
+      "learning_rate": 8.064516129032258e-06,
+      "loss": 0.5242,
+      "step": 25
+    },
+    {
+      "epoch": 0.10526315789473684,
+      "grad_norm": 1.0379437863935217,
+      "learning_rate": 8.387096774193549e-06,
+      "loss": 0.5533,
+      "step": 26
+    },
+    {
+      "epoch": 0.10931174089068826,
+      "grad_norm": 1.1166984754294182,
+      "learning_rate": 8.70967741935484e-06,
+      "loss": 0.5278,
+      "step": 27
+    },
+    {
+      "epoch": 0.11336032388663968,
+      "grad_norm": 1.0925635024225429,
+      "learning_rate": 9.03225806451613e-06,
+      "loss": 0.5287,
+      "step": 28
+    },
+    {
+      "epoch": 0.11740890688259109,
+      "grad_norm": 0.7471053700092182,
+      "learning_rate": 9.35483870967742e-06,
+      "loss": 0.5231,
+      "step": 29
+    },
+    {
+      "epoch": 0.1214574898785425,
+      "grad_norm": 0.9145345098008165,
+      "learning_rate": 9.67741935483871e-06,
+      "loss": 0.5367,
+      "step": 30
+    },
+    {
+      "epoch": 0.12550607287449392,
+      "grad_norm": 1.2481570206426928,
+      "learning_rate": 1e-05,
+      "loss": 0.5309,
+      "step": 31
+    },
+    {
+      "epoch": 0.12955465587044535,
+      "grad_norm": 0.7986507724943911,
+      "learning_rate": 1.0322580645161291e-05,
+      "loss": 0.5108,
+      "step": 32
+    },
+    {
+      "epoch": 0.13360323886639677,
+      "grad_norm": 0.7430576920477759,
+      "learning_rate": 1.0645161290322582e-05,
+      "loss": 0.5127,
+      "step": 33
+    },
+    {
+      "epoch": 0.13765182186234817,
+      "grad_norm": 0.6412457255545513,
+      "learning_rate": 1.096774193548387e-05,
+      "loss": 0.4991,
+      "step": 34
+    },
+    {
+      "epoch": 0.1417004048582996,
+      "grad_norm": 0.7162392865092905,
+      "learning_rate": 1.1290322580645164e-05,
+      "loss": 0.4923,
+      "step": 35
+    },
+    {
+      "epoch": 0.145748987854251,
+      "grad_norm": 0.6152141590330319,
+      "learning_rate": 1.1612903225806453e-05,
+      "loss": 0.5089,
+      "step": 36
+    },
+    {
+      "epoch": 0.14979757085020243,
+      "grad_norm": 0.5251827231128092,
+      "learning_rate": 1.1935483870967743e-05,
+      "loss": 0.4786,
+      "step": 37
+    },
+    {
+      "epoch": 0.15384615384615385,
+      "grad_norm": 1.0229092071688777,
+      "learning_rate": 1.2258064516129034e-05,
+      "loss": 0.5129,
+      "step": 38
+    },
+    {
+      "epoch": 0.15789473684210525,
+      "grad_norm": 0.6245749396135634,
+      "learning_rate": 1.2580645161290324e-05,
+      "loss": 0.4817,
+      "step": 39
+    },
+    {
+      "epoch": 0.16194331983805668,
+      "grad_norm": 0.6257207123585126,
+      "learning_rate": 1.2903225806451613e-05,
+      "loss": 0.484,
+      "step": 40
+    },
+    {
+      "epoch": 0.1659919028340081,
+      "grad_norm": 0.5544075456509967,
+      "learning_rate": 1.3225806451612903e-05,
+      "loss": 0.4798,
+      "step": 41
+    },
+    {
+      "epoch": 0.1700404858299595,
+      "grad_norm": 0.527761031652804,
+      "learning_rate": 1.3548387096774194e-05,
+      "loss": 0.4876,
+      "step": 42
+    },
+    {
+      "epoch": 0.17408906882591094,
+      "grad_norm": 0.5878648738010059,
+      "learning_rate": 1.3870967741935486e-05,
+      "loss": 0.5056,
+      "step": 43
+    },
+    {
+      "epoch": 0.17813765182186234,
+      "grad_norm": 0.4732827580378767,
+      "learning_rate": 1.4193548387096776e-05,
+      "loss": 0.4864,
+      "step": 44
+    },
+    {
+      "epoch": 0.18218623481781376,
+      "grad_norm": 0.5052143754613014,
+      "learning_rate": 1.4516129032258066e-05,
+      "loss": 0.4825,
+      "step": 45
+    },
+    {
+      "epoch": 0.1862348178137652,
+      "grad_norm": 0.5109434243168585,
+      "learning_rate": 1.4838709677419357e-05,
+      "loss": 0.4709,
+      "step": 46
+    },
+    {
+      "epoch": 0.1902834008097166,
+      "grad_norm": 0.42657392647041853,
+      "learning_rate": 1.5161290322580646e-05,
+      "loss": 0.4744,
+      "step": 47
+    },
+    {
+      "epoch": 0.19433198380566802,
+      "grad_norm": 0.6367314954228658,
+      "learning_rate": 1.5483870967741936e-05,
+      "loss": 0.4843,
+      "step": 48
+    },
+    {
+      "epoch": 0.19838056680161945,
+      "grad_norm": 0.5296462717335911,
+      "learning_rate": 1.5806451612903226e-05,
+      "loss": 0.4727,
+      "step": 49
+    },
+    {
+      "epoch": 0.20242914979757085,
+      "grad_norm": 0.4734472305573063,
+      "learning_rate": 1.6129032258064517e-05,
+      "loss": 0.454,
+      "step": 50
+    },
+    {
+      "epoch": 0.20647773279352227,
+      "grad_norm": 0.543128655404803,
+      "learning_rate": 1.6451612903225807e-05,
+      "loss": 0.4844,
+      "step": 51
+    },
+    {
+      "epoch": 0.21052631578947367,
+      "grad_norm": 0.5006421620710721,
+      "learning_rate": 1.6774193548387098e-05,
+      "loss": 0.4699,
+      "step": 52
+    },
+    {
+      "epoch": 0.2145748987854251,
+      "grad_norm": 0.5010410493722355,
+      "learning_rate": 1.7096774193548388e-05,
+      "loss": 0.4712,
+      "step": 53
+    },
+    {
+      "epoch": 0.21862348178137653,
+      "grad_norm": 0.9259030283539433,
+      "learning_rate": 1.741935483870968e-05,
+      "loss": 0.4856,
+      "step": 54
+    },
+    {
+      "epoch": 0.22267206477732793,
+      "grad_norm": 0.5075437994152665,
+      "learning_rate": 1.774193548387097e-05,
+      "loss": 0.4771,
+      "step": 55
+    },
+    {
+      "epoch": 0.22672064777327935,
+      "grad_norm": 0.5498567344385248,
+      "learning_rate": 1.806451612903226e-05,
+      "loss": 0.4502,
+      "step": 56
+    },
+    {
+      "epoch": 0.23076923076923078,
+      "grad_norm": 0.5754916105589878,
+      "learning_rate": 1.838709677419355e-05,
+      "loss": 0.4549,
+      "step": 57
+    },
+    {
+      "epoch": 0.23481781376518218,
+      "grad_norm": 0.5160685050683056,
+      "learning_rate": 1.870967741935484e-05,
+      "loss": 0.4835,
+      "step": 58
+    },
+    {
+      "epoch": 0.2388663967611336,
+      "grad_norm": 0.6299502076987529,
+      "learning_rate": 1.903225806451613e-05,
+      "loss": 0.4944,
+      "step": 59
+    },
+    {
+      "epoch": 0.242914979757085,
+      "grad_norm": 0.6774077979685593,
+      "learning_rate": 1.935483870967742e-05,
+      "loss": 0.4641,
+      "step": 60
+    },
+    {
+      "epoch": 0.24696356275303644,
+      "grad_norm": 0.4620275580169436,
+      "learning_rate": 1.967741935483871e-05,
+      "loss": 0.4629,
+      "step": 61
+    },
+    {
+      "epoch": 0.25101214574898784,
+      "grad_norm": 0.5245369171883172,
+      "learning_rate": 2e-05,
+      "loss": 0.4616,
+      "step": 62
+    },
+    {
+      "epoch": 0.2550607287449393,
+      "grad_norm": 0.48266174613065954,
+      "learning_rate": 2.0322580645161292e-05,
+      "loss": 0.4806,
+      "step": 63
+    },
+    {
+      "epoch": 0.2591093117408907,
+      "grad_norm": 0.5162480009139384,
+      "learning_rate": 2.0645161290322582e-05,
+      "loss": 0.4703,
+      "step": 64
+    },
+    {
+      "epoch": 0.2631578947368421,
+      "grad_norm": 0.5240495817686905,
+      "learning_rate": 2.0967741935483873e-05,
+      "loss": 0.4608,
+      "step": 65
+    },
+    {
+      "epoch": 0.26720647773279355,
+      "grad_norm": 0.5082274969367158,
+      "learning_rate": 2.1290322580645163e-05,
+      "loss": 0.4483,
+      "step": 66
+    },
+    {
+      "epoch": 0.27125506072874495,
+      "grad_norm": 0.618171629686987,
+      "learning_rate": 2.161290322580645e-05,
+      "loss": 0.4791,
+      "step": 67
+    },
+    {
+      "epoch": 0.27530364372469635,
+      "grad_norm": 0.44451554183203557,
+      "learning_rate": 2.193548387096774e-05,
+      "loss": 0.4562,
+      "step": 68
+    },
+    {
+      "epoch": 0.2793522267206478,
+      "grad_norm": 0.6134574852553367,
+      "learning_rate": 2.225806451612903e-05,
+      "loss": 0.4719,
+      "step": 69
+    },
+    {
+      "epoch": 0.2834008097165992,
+      "grad_norm": 0.4884476702482166,
+      "learning_rate": 2.2580645161290328e-05,
+      "loss": 0.4386,
+      "step": 70
+    },
+    {
+      "epoch": 0.2874493927125506,
+      "grad_norm": 0.5279792562944238,
+      "learning_rate": 2.2903225806451615e-05,
+      "loss": 0.433,
+      "step": 71
+    },
+    {
+      "epoch": 0.291497975708502,
+      "grad_norm": 0.6004154658243289,
+      "learning_rate": 2.3225806451612906e-05,
+      "loss": 0.4406,
+      "step": 72
+    },
+    {
+      "epoch": 0.29554655870445345,
+      "grad_norm": 0.616072005991922,
+      "learning_rate": 2.3548387096774196e-05,
+      "loss": 0.4524,
+      "step": 73
+    },
+    {
+      "epoch": 0.29959514170040485,
+      "grad_norm": 0.7005105655830699,
+      "learning_rate": 2.3870967741935486e-05,
+      "loss": 0.4504,
+      "step": 74
+    },
+    {
+      "epoch": 0.30364372469635625,
+      "grad_norm": 0.48726321295192787,
+      "learning_rate": 2.4193548387096777e-05,
+      "loss": 0.4628,
+      "step": 75
+    },
+    {
+      "epoch": 0.3076923076923077,
+      "grad_norm": 0.5771595982954283,
+      "learning_rate": 2.4516129032258067e-05,
+      "loss": 0.4499,
+      "step": 76
+    },
+    {
+      "epoch": 0.3117408906882591,
+      "grad_norm": 0.4926720405161142,
+      "learning_rate": 2.4838709677419358e-05,
+      "loss": 0.4551,
+      "step": 77
+    },
+    {
+      "epoch": 0.3157894736842105,
+      "grad_norm": 0.5234566764300341,
+      "learning_rate": 2.5161290322580648e-05,
+      "loss": 0.4553,
+      "step": 78
+    },
+    {
+      "epoch": 0.31983805668016196,
+      "grad_norm": 0.49962852029274735,
+      "learning_rate": 2.548387096774194e-05,
+      "loss": 0.4536,
+      "step": 79
+    },
+    {
+      "epoch": 0.32388663967611336,
+      "grad_norm": 0.573980228248545,
+      "learning_rate": 2.5806451612903226e-05,
+      "loss": 0.4706,
+      "step": 80
+    },
+    {
+      "epoch": 0.32793522267206476,
+      "grad_norm": 0.64510998259783,
+      "learning_rate": 2.6129032258064516e-05,
+      "loss": 0.4598,
+      "step": 81
+    },
+    {
+      "epoch": 0.3319838056680162,
+      "grad_norm": 0.7721957768077252,
+      "learning_rate": 2.6451612903225806e-05,
+      "loss": 0.482,
+      "step": 82
+    },
+    {
+      "epoch": 0.3360323886639676,
+      "grad_norm": 0.6600203362638507,
+      "learning_rate": 2.6774193548387097e-05,
+      "loss": 0.4702,
+      "step": 83
+    },
+    {
+      "epoch": 0.340080971659919,
+      "grad_norm": 0.749187874980439,
+      "learning_rate": 2.7096774193548387e-05,
+      "loss": 0.4543,
+      "step": 84
+    },
+    {
+      "epoch": 0.3441295546558704,
+      "grad_norm": 0.6170225940509219,
+      "learning_rate": 2.7419354838709678e-05,
+      "loss": 0.4732,
+      "step": 85
+    },
+    {
+      "epoch": 0.3481781376518219,
+      "grad_norm": 0.5256666316848865,
+      "learning_rate": 2.774193548387097e-05,
+      "loss": 0.4624,
+      "step": 86
+    },
+    {
+      "epoch": 0.3522267206477733,
+      "grad_norm": 0.545852922719072,
+      "learning_rate": 2.8064516129032262e-05,
+      "loss": 0.4467,
+      "step": 87
+    },
+    {
+      "epoch": 0.3562753036437247,
+      "grad_norm": 0.48778614506372103,
+      "learning_rate": 2.8387096774193552e-05,
+      "loss": 0.4644,
+      "step": 88
+    },
+    {
+      "epoch": 0.3603238866396761,
+      "grad_norm": 0.48854976925357124,
+      "learning_rate": 2.8709677419354843e-05,
+      "loss": 0.4529,
+      "step": 89
+    },
+    {
+      "epoch": 0.3643724696356275,
+      "grad_norm": 0.5211673149088681,
+      "learning_rate": 2.9032258064516133e-05,
+      "loss": 0.4345,
+      "step": 90
+    },
+    {
+      "epoch": 0.3684210526315789,
+      "grad_norm": 0.5383447124039021,
+      "learning_rate": 2.9354838709677423e-05,
+      "loss": 0.4529,
+      "step": 91
+    },
+    {
+      "epoch": 0.3724696356275304,
+      "grad_norm": 0.5550219811598042,
+      "learning_rate": 2.9677419354838714e-05,
+      "loss": 0.4505,
+      "step": 92
+    },
+    {
+      "epoch": 0.3765182186234818,
+      "grad_norm": 0.5551876909705087,
+      "learning_rate": 3.0000000000000004e-05,
+      "loss": 0.4506,
+      "step": 93
+    },
+    {
+      "epoch": 0.3805668016194332,
+      "grad_norm": 0.6172980139542952,
+      "learning_rate": 3.032258064516129e-05,
+      "loss": 0.4438,
+      "step": 94
+    },
+    {
+      "epoch": 0.38461538461538464,
+      "grad_norm": 0.571984294229591,
+      "learning_rate": 3.0645161290322585e-05,
+      "loss": 0.4445,
+      "step": 95
+    },
+    {
+      "epoch": 0.38866396761133604,
+      "grad_norm": 0.5185510742675107,
+      "learning_rate": 3.096774193548387e-05,
+      "loss": 0.4179,
+      "step": 96
+    },
+    {
+      "epoch": 0.39271255060728744,
+      "grad_norm": 0.47893485845172906,
+      "learning_rate": 3.1290322580645166e-05,
+      "loss": 0.4436,
+      "step": 97
+    },
+    {
+      "epoch": 0.3967611336032389,
+      "grad_norm": 0.5520845598875853,
+      "learning_rate": 3.161290322580645e-05,
+      "loss": 0.449,
+      "step": 98
+    },
+    {
+      "epoch": 0.4008097165991903,
+      "grad_norm": 0.60748515241145,
+      "learning_rate": 3.193548387096774e-05,
+      "loss": 0.4633,
+      "step": 99
+    },
+    {
+      "epoch": 0.4048582995951417,
+      "grad_norm": 0.6215154899003764,
+      "learning_rate": 3.2258064516129034e-05,
+      "loss": 0.4441,
+      "step": 100
+    },
+    {
+      "epoch": 0.4089068825910931,
+      "grad_norm": 0.4850363511417887,
+      "learning_rate": 3.258064516129033e-05,
+      "loss": 0.4561,
+      "step": 101
+    },
+    {
+      "epoch": 0.41295546558704455,
+      "grad_norm": 0.6041798783314605,
+      "learning_rate": 3.2903225806451614e-05,
+      "loss": 0.4911,
+      "step": 102
+    },
+    {
+      "epoch": 0.41700404858299595,
+      "grad_norm": 0.6773972118690768,
+      "learning_rate": 3.322580645161291e-05,
+      "loss": 0.4502,
+      "step": 103
+    },
+    {
+      "epoch": 0.42105263157894735,
+      "grad_norm": 0.6692740154880154,
+      "learning_rate": 3.3548387096774195e-05,
+      "loss": 0.4435,
+      "step": 104
+    },
+    {
+      "epoch": 0.4251012145748988,
+      "grad_norm": 0.6408953516980191,
+      "learning_rate": 3.387096774193549e-05,
+      "loss": 0.4555,
+      "step": 105
+    },
+    {
+      "epoch": 0.4291497975708502,
+      "grad_norm": 0.48370919269102153,
+      "learning_rate": 3.4193548387096776e-05,
+      "loss": 0.4482,
+      "step": 106
+    },
+    {
+      "epoch": 0.4331983805668016,
+      "grad_norm": 0.5875068135599418,
+      "learning_rate": 3.451612903225807e-05,
+      "loss": 0.4374,
+      "step": 107
+    },
+    {
+      "epoch": 0.43724696356275305,
+      "grad_norm": 0.6820564104748379,
+      "learning_rate": 3.483870967741936e-05,
+      "loss": 0.4571,
+      "step": 108
+    },
+    {
+      "epoch": 0.44129554655870445,
+      "grad_norm": 0.49836787310804853,
+      "learning_rate": 3.516129032258065e-05,
+      "loss": 0.4456,
+      "step": 109
+    },
+    {
+      "epoch": 0.44534412955465585,
+      "grad_norm": 0.608646700595456,
+      "learning_rate": 3.548387096774194e-05,
+      "loss": 0.4398,
+      "step": 110
+    },
+    {
+      "epoch": 0.4493927125506073,
+      "grad_norm": 0.6169900650855518,
+      "learning_rate": 3.5806451612903225e-05,
+      "loss": 0.4518,
+      "step": 111
+    },
+    {
+      "epoch": 0.4534412955465587,
+      "grad_norm": 0.5769829348782364,
+      "learning_rate": 3.612903225806452e-05,
+      "loss": 0.4459,
+      "step": 112
+    },
+    {
+      "epoch": 0.4574898785425101,
+      "grad_norm": 0.6115225326652,
+      "learning_rate": 3.6451612903225805e-05,
+      "loss": 0.4355,
+      "step": 113
+    },
+    {
+      "epoch": 0.46153846153846156,
+      "grad_norm": 0.5658825793777497,
+      "learning_rate": 3.67741935483871e-05,
+      "loss": 0.4343,
+      "step": 114
+    },
+    {
+      "epoch": 0.46558704453441296,
+      "grad_norm": 0.5672103660321228,
+      "learning_rate": 3.7096774193548386e-05,
+      "loss": 0.4507,
+      "step": 115
+    },
+    {
+      "epoch": 0.46963562753036436,
+      "grad_norm": 0.7239628977414876,
+      "learning_rate": 3.741935483870968e-05,
+      "loss": 0.4453,
+      "step": 116
+    },
+    {
+      "epoch": 0.47368421052631576,
+      "grad_norm": 0.6397246378965932,
+      "learning_rate": 3.7741935483870974e-05,
+      "loss": 0.4478,
+      "step": 117
+    },
+    {
+      "epoch": 0.4777327935222672,
+      "grad_norm": 0.5929456807705266,
+      "learning_rate": 3.806451612903226e-05,
+      "loss": 0.4681,
+      "step": 118
+    },
+    {
+      "epoch": 0.4817813765182186,
+      "grad_norm": 0.6713724521779054,
+      "learning_rate": 3.8387096774193555e-05,
+      "loss": 0.4373,
+      "step": 119
+    },
+    {
+      "epoch": 0.48582995951417,
+      "grad_norm": 0.7161616558974805,
+      "learning_rate": 3.870967741935484e-05,
+      "loss": 0.4203,
+      "step": 120
+    },
+    {
+      "epoch": 0.4898785425101215,
+      "grad_norm": 0.54535023256009,
+      "learning_rate": 3.9032258064516135e-05,
+      "loss": 0.4456,
+      "step": 121
+    },
+    {
+      "epoch": 0.4939271255060729,
+      "grad_norm": 0.6064575144349783,
+      "learning_rate": 3.935483870967742e-05,
+      "loss": 0.4481,
+      "step": 122
+    },
+    {
+      "epoch": 0.4979757085020243,
+      "grad_norm": 0.5050364273519224,
+      "learning_rate": 3.9677419354838716e-05,
+      "loss": 0.4564,
+      "step": 123
+    },
+    {
+      "epoch": 0.5020242914979757,
+      "grad_norm": 0.6365430876033804,
+      "learning_rate": 4e-05,
+      "loss": 0.4721,
+      "step": 124
+    },
+    {
+      "epoch": 0.5060728744939271,
+      "grad_norm": 0.6583455124499552,
+      "learning_rate": 3.9999920040266474e-05,
+      "loss": 0.4512,
+      "step": 125
+    },
+    {
+      "epoch": 0.5101214574898786,
+      "grad_norm": 0.7835094640147059,
+      "learning_rate": 3.999968016170525e-05,
+      "loss": 0.4463,
+      "step": 126
+    },
+    {
+      "epoch": 0.5141700404858299,
+      "grad_norm": 0.6674057028281345,
+      "learning_rate": 3.9999280366234394e-05,
+      "loss": 0.4458,
+      "step": 127
+    },
+    {
+      "epoch": 0.5182186234817814,
+      "grad_norm": 0.5875314423850757,
+      "learning_rate": 3.999872065705065e-05,
+      "loss": 0.4381,
+      "step": 128
+    },
+    {
+      "epoch": 0.5222672064777328,
+      "grad_norm": 0.6799895922724815,
+      "learning_rate": 3.999800103862945e-05,
+      "loss": 0.4309,
+      "step": 129
+    },
+    {
+      "epoch": 0.5263157894736842,
+      "grad_norm": 0.6546747978824243,
+      "learning_rate": 3.9997121516724834e-05,
+      "loss": 0.4362,
+      "step": 130
+    },
+    {
+      "epoch": 0.5303643724696356,
+      "grad_norm": 0.7380610416131168,
+      "learning_rate": 3.999608209836944e-05,
+      "loss": 0.4419,
+      "step": 131
+    },
+    {
+      "epoch": 0.5344129554655871,
+      "grad_norm": 0.5663739428346416,
+      "learning_rate": 3.999488279187443e-05,
+      "loss": 0.4393,
+      "step": 132
+    },
+    {
+      "epoch": 0.5384615384615384,
+      "grad_norm": 0.7001916790062369,
+      "learning_rate": 3.9993523606829425e-05,
+      "loss": 0.4559,
+      "step": 133
+    },
+    {
+      "epoch": 0.5425101214574899,
+      "grad_norm": 0.5887560720933889,
+      "learning_rate": 3.9992004554102435e-05,
+      "loss": 0.4499,
+      "step": 134
+    },
+    {
+      "epoch": 0.5465587044534413,
+      "grad_norm": 0.5628918980527864,
+      "learning_rate": 3.999032564583976e-05,
+      "loss": 0.4556,
+      "step": 135
+    },
+    {
+      "epoch": 0.5506072874493927,
+      "grad_norm": 0.6098553617233414,
+      "learning_rate": 3.998848689546591e-05,
+      "loss": 0.4502,
+      "step": 136
+    },
+    {
+      "epoch": 0.5546558704453441,
+      "grad_norm": 0.5178603779309826,
+      "learning_rate": 3.9986488317683496e-05,
+      "loss": 0.4287,
+      "step": 137
+    },
+    {
+      "epoch": 0.5587044534412956,
+      "grad_norm": 0.6051951602893834,
+      "learning_rate": 3.9984329928473067e-05,
+      "loss": 0.4494,
+      "step": 138
+    },
+    {
+      "epoch": 0.562753036437247,
+      "grad_norm": 0.5754268655844345,
+      "learning_rate": 3.998201174509306e-05,
+      "loss": 0.4356,
+      "step": 139
+    },
+    {
+      "epoch": 0.5668016194331984,
+      "grad_norm": 0.5775472993080624,
+      "learning_rate": 3.997953378607961e-05,
+      "loss": 0.4317,
+      "step": 140
+    },
+    {
+      "epoch": 0.5708502024291497,
+      "grad_norm": 0.5384242968493611,
+      "learning_rate": 3.9976896071246403e-05,
+      "loss": 0.4526,
+      "step": 141
+    },
+    {
+      "epoch": 0.5748987854251012,
+      "grad_norm": 0.5320202248454518,
+      "learning_rate": 3.997409862168454e-05,
+      "loss": 0.442,
+      "step": 142
+    },
+    {
+      "epoch": 0.5789473684210527,
+      "grad_norm": 0.4949580611369673,
+      "learning_rate": 3.997114145976237e-05,
+      "loss": 0.422,
+      "step": 143
+    },
+    {
+      "epoch": 0.582995951417004,
+      "grad_norm": 0.5653419411694924,
+      "learning_rate": 3.9968024609125245e-05,
+      "loss": 0.4463,
+      "step": 144
+    },
+    {
+      "epoch": 0.5870445344129555,
+      "grad_norm": 0.441331963343144,
+      "learning_rate": 3.996474809469545e-05,
+      "loss": 0.445,
+      "step": 145
+    },
+    {
+      "epoch": 0.5910931174089069,
+      "grad_norm": 0.5704550288316456,
+      "learning_rate": 3.996131194267189e-05,
+      "loss": 0.4354,
+      "step": 146
+    },
+    {
+      "epoch": 0.5951417004048583,
+      "grad_norm": 0.48836475418131264,
+      "learning_rate": 3.995771618052995e-05,
+      "loss": 0.424,
+      "step": 147
+    },
+    {
+      "epoch": 0.5991902834008097,
+      "grad_norm": 0.5756160662026374,
+      "learning_rate": 3.9953960837021256e-05,
+      "loss": 0.4507,
+      "step": 148
+    },
+    {
+      "epoch": 0.6032388663967612,
+      "grad_norm": 0.6483801483956595,
+      "learning_rate": 3.995004594217343e-05,
+      "loss": 0.4503,
+      "step": 149
+    },
+    {
+      "epoch": 0.6072874493927125,
+      "grad_norm": 0.5019038686285493,
+      "learning_rate": 3.994597152728987e-05,
+      "loss": 0.435,
+      "step": 150
+    },
+    {
+      "epoch": 0.611336032388664,
+      "grad_norm": 0.5411933341813526,
+      "learning_rate": 3.994173762494947e-05,
+      "loss": 0.4349,
+      "step": 151
+    },
+    {
+      "epoch": 0.6153846153846154,
+      "grad_norm": 0.5303649941830172,
+      "learning_rate": 3.9937344269006424e-05,
+      "loss": 0.4372,
+      "step": 152
+    },
+    {
+      "epoch": 0.6194331983805668,
+      "grad_norm": 0.5430097929886917,
+      "learning_rate": 3.993279149458988e-05,
+      "loss": 0.4369,
+      "step": 153
+    },
+    {
+      "epoch": 0.6234817813765182,
+      "grad_norm": 0.5014829424214975,
+      "learning_rate": 3.9928079338103695e-05,
+      "loss": 0.4644,
+      "step": 154
+    },
+    {
+      "epoch": 0.6275303643724697,
+      "grad_norm": 0.4745245481385754,
+      "learning_rate": 3.992320783722616e-05,
+      "loss": 0.443,
+      "step": 155
+    },
+    {
+      "epoch": 0.631578947368421,
+      "grad_norm": 0.47777043942392294,
+      "learning_rate": 3.991817703090965e-05,
+      "loss": 0.4411,
+      "step": 156
+    },
+    {
+      "epoch": 0.6356275303643725,
+      "grad_norm": 0.47211125793245534,
+      "learning_rate": 3.991298695938038e-05,
+      "loss": 0.4272,
+      "step": 157
+    },
+    {
+      "epoch": 0.6396761133603239,
+      "grad_norm": 0.5162184555860596,
+      "learning_rate": 3.9907637664138e-05,
+      "loss": 0.4572,
+      "step": 158
+    },
+    {
+      "epoch": 0.6437246963562753,
+      "grad_norm": 0.44193353707312344,
+      "learning_rate": 3.990212918795534e-05,
+      "loss": 0.431,
+      "step": 159
+    },
+    {
+      "epoch": 0.6477732793522267,
+      "grad_norm": 0.7402608943922941,
+      "learning_rate": 3.9896461574878036e-05,
+      "loss": 0.4483,
+      "step": 160
+    },
+    {
+      "epoch": 0.6518218623481782,
+      "grad_norm": 0.46019659206322394,
+      "learning_rate": 3.989063487022416e-05,
+      "loss": 0.4346,
+      "step": 161
+    },
+    {
+      "epoch": 0.6558704453441295,
+      "grad_norm": 0.6199096142813506,
+      "learning_rate": 3.9884649120583903e-05,
+      "loss": 0.4531,
+      "step": 162
+    },
+    {
+      "epoch": 0.659919028340081,
+      "grad_norm": 0.9477594408593222,
+      "learning_rate": 3.987850437381915e-05,
+      "loss": 0.4779,
+      "step": 163
+    },
+    {
+      "epoch": 0.6639676113360324,
+      "grad_norm": 0.548742479868264,
+      "learning_rate": 3.987220067906313e-05,
+      "loss": 0.4318,
+      "step": 164
+    },
+    {
+      "epoch": 0.6680161943319838,
+      "grad_norm": 0.5752417327872278,
+      "learning_rate": 3.986573808672002e-05,
+      "loss": 0.4442,
+      "step": 165
+    },
+    {
+      "epoch": 0.6720647773279352,
+      "grad_norm": 0.6341409680259891,
+      "learning_rate": 3.9859116648464534e-05,
+      "loss": 0.4321,
+      "step": 166
+    },
+    {
+      "epoch": 0.6761133603238867,
+      "grad_norm": 0.5586235146142684,
+      "learning_rate": 3.985233641724152e-05,
+      "loss": 0.4422,
+      "step": 167
+    },
+    {
+      "epoch": 0.680161943319838,
+      "grad_norm": 0.44140086922960897,
+      "learning_rate": 3.984539744726554e-05,
+      "loss": 0.4416,
+      "step": 168
+    },
+    {
+      "epoch": 0.6842105263157895,
+      "grad_norm": 0.4922519018403677,
+      "learning_rate": 3.983829979402038e-05,
+      "loss": 0.4368,
+      "step": 169
+    },
+    {
+      "epoch": 0.6882591093117408,
+      "grad_norm": 0.4808220356163561,
+      "learning_rate": 3.983104351425871e-05,
+      "loss": 0.4387,
+      "step": 170
+    },
+    {
+      "epoch": 0.6923076923076923,
+      "grad_norm": 0.40290571382023,
+      "learning_rate": 3.982362866600154e-05,
+      "loss": 0.4406,
+      "step": 171
+    },
+    {
+      "epoch": 0.6963562753036437,
+      "grad_norm": 0.6606668411701291,
+      "learning_rate": 3.9816055308537806e-05,
+      "loss": 0.4321,
+      "step": 172
+    },
+    {
+      "epoch": 0.7004048582995951,
+      "grad_norm": 0.4860773365078243,
+      "learning_rate": 3.980832350242388e-05,
+      "loss": 0.4472,
+      "step": 173
+    },
+    {
+      "epoch": 0.7044534412955465,
+      "grad_norm": 0.39571195511091695,
+      "learning_rate": 3.980043330948306e-05,
+      "loss": 0.4459,
+      "step": 174
+    },
+    {
+      "epoch": 0.708502024291498,
+      "grad_norm": 0.47521042236759253,
+      "learning_rate": 3.9792384792805124e-05,
+      "loss": 0.4186,
+      "step": 175
+    },
+    {
+      "epoch": 0.7125506072874493,
+      "grad_norm": 0.5102015919872362,
+      "learning_rate": 3.978417801674579e-05,
+      "loss": 0.4473,
+      "step": 176
+    },
+    {
+      "epoch": 0.7165991902834008,
+      "grad_norm": 0.4815996767721079,
+      "learning_rate": 3.9775813046926244e-05,
+      "loss": 0.4362,
+      "step": 177
+    },
+    {
+      "epoch": 0.7206477732793523,
+      "grad_norm": 0.5733060259550246,
+      "learning_rate": 3.976728995023254e-05,
+      "loss": 0.437,
+      "step": 178
+    },
+    {
+      "epoch": 0.7246963562753036,
+      "grad_norm": 0.4785514977326472,
+      "learning_rate": 3.975860879481514e-05,
+      "loss": 0.4221,
+      "step": 179
+    },
+    {
+      "epoch": 0.728744939271255,
+      "grad_norm": 0.5270026038223529,
+      "learning_rate": 3.974976965008833e-05,
+      "loss": 0.4372,
+      "step": 180
+    },
+    {
+      "epoch": 0.7327935222672065,
+      "grad_norm": 0.459023536819358,
+      "learning_rate": 3.9740772586729674e-05,
+      "loss": 0.4153,
+      "step": 181
+    },
+    {
+      "epoch": 0.7368421052631579,
+      "grad_norm": 0.5886545291426545,
+      "learning_rate": 3.973161767667946e-05,
+      "loss": 0.4457,
+      "step": 182
+    },
+    {
+      "epoch": 0.7408906882591093,
+      "grad_norm": 0.4615527690682782,
+      "learning_rate": 3.972230499314009e-05,
+      "loss": 0.4192,
+      "step": 183
+    },
+    {
+      "epoch": 0.7449392712550608,
+      "grad_norm": 0.5738737080088512,
+      "learning_rate": 3.9712834610575545e-05,
+      "loss": 0.4298,
+      "step": 184
+    },
+    {
+      "epoch": 0.7489878542510121,
+      "grad_norm": 0.37798574583824546,
+      "learning_rate": 3.9703206604710746e-05,
+      "loss": 0.4395,
+      "step": 185
+    },
+    {
+      "epoch": 0.7530364372469636,
+      "grad_norm": 0.5136238477557756,
+      "learning_rate": 3.969342105253098e-05,
+      "loss": 0.458,
+      "step": 186
+    },
+    {
+      "epoch": 0.757085020242915,
+      "grad_norm": 0.4253922668306012,
+      "learning_rate": 3.9683478032281246e-05,
+      "loss": 0.4347,
+      "step": 187
+    },
+    {
+      "epoch": 0.7611336032388664,
+      "grad_norm": 0.522173086030332,
+      "learning_rate": 3.967337762346568e-05,
+      "loss": 0.4437,
+      "step": 188
+    },
+    {
+      "epoch": 0.7651821862348178,
+      "grad_norm": 0.47830491399626357,
+      "learning_rate": 3.9663119906846885e-05,
+      "loss": 0.4345,
+      "step": 189
+    },
+    {
+      "epoch": 0.7692307692307693,
+      "grad_norm": 0.5144776947877908,
+      "learning_rate": 3.965270496444528e-05,
+      "loss": 0.4363,
+      "step": 190
+    },
+    {
+      "epoch": 0.7732793522267206,
+      "grad_norm": 0.48497320416505607,
+      "learning_rate": 3.9642132879538465e-05,
+      "loss": 0.4258,
+      "step": 191
+    },
+    {
+      "epoch": 0.7773279352226721,
+      "grad_norm": 0.464068800736141,
+      "learning_rate": 3.9631403736660566e-05,
+      "loss": 0.4253,
+      "step": 192
+    },
+    {
+      "epoch": 0.7813765182186235,
+      "grad_norm": 0.5114604106757257,
+      "learning_rate": 3.962051762160151e-05,
+      "loss": 0.4533,
+      "step": 193
+    },
+    {
+      "epoch": 0.7854251012145749,
+      "grad_norm": 0.4754525300812924,
+      "learning_rate": 3.960947462140638e-05,
+      "loss": 0.4367,
+      "step": 194
+    },
+    {
+      "epoch": 0.7894736842105263,
+      "grad_norm": 0.43262409817471825,
+      "learning_rate": 3.959827482437472e-05,
+      "loss": 0.4218,
+      "step": 195
+    },
+    {
+      "epoch": 0.7935222672064778,
+      "grad_norm": 0.44511673102909816,
+      "learning_rate": 3.958691832005981e-05,
+      "loss": 0.4281,
+      "step": 196
+    },
+    {
+      "epoch": 0.7975708502024291,
+      "grad_norm": 0.44104169984733843,
+      "learning_rate": 3.9575405199267944e-05,
+      "loss": 0.4501,
+      "step": 197
+    },
+    {
+      "epoch": 0.8016194331983806,
+      "grad_norm": 0.4157250244324796,
+      "learning_rate": 3.956373555405774e-05,
+      "loss": 0.4636,
+      "step": 198
+    },
+    {
+      "epoch": 0.805668016194332,
+      "grad_norm": 0.5006503815083972,
+      "learning_rate": 3.955190947773937e-05,
+      "loss": 0.4419,
+      "step": 199
+    },
+    {
+      "epoch": 0.8097165991902834,
+      "grad_norm": 0.4333343340537449,
+      "learning_rate": 3.9539927064873824e-05,
+      "loss": 0.4276,
+      "step": 200
+    },
+    {
+      "epoch": 0.8137651821862348,
+      "grad_norm": 0.5031990334579663,
+      "learning_rate": 3.952778841127214e-05,
+      "loss": 0.4489,
+      "step": 201
+    },
+    {
+      "epoch": 0.8178137651821862,
+      "grad_norm": 0.4255425833177368,
+      "learning_rate": 3.95154936139947e-05,
+      "loss": 0.4407,
+      "step": 202
+    },
+    {
+      "epoch": 0.8218623481781376,
+      "grad_norm": 0.4521459628292438,
+      "learning_rate": 3.950304277135034e-05,
+      "loss": 0.4261,
+      "step": 203
+    },
+    {
+      "epoch": 0.8259109311740891,
+      "grad_norm": 0.40978324872963345,
+      "learning_rate": 3.94904359828957e-05,
+      "loss": 0.4177,
+      "step": 204
+    },
+    {
+      "epoch": 0.8299595141700404,
+      "grad_norm": 0.45209745088478415,
+      "learning_rate": 3.94776733494343e-05,
+      "loss": 0.4299,
+      "step": 205
+    },
+    {
+      "epoch": 0.8340080971659919,
+      "grad_norm": 0.473390610452477,
+      "learning_rate": 3.9464754973015826e-05,
+      "loss": 0.4303,
+      "step": 206
+    },
+    {
+      "epoch": 0.8380566801619433,
+      "grad_norm": 0.5405323453842239,
+      "learning_rate": 3.9451680956935275e-05,
+      "loss": 0.4255,
+      "step": 207
+    },
+    {
+      "epoch": 0.8421052631578947,
+      "grad_norm": 0.46808002184165,
+      "learning_rate": 3.943845140573213e-05,
+      "loss": 0.4441,
+      "step": 208
+    },
+    {
+      "epoch": 0.8461538461538461,
+      "grad_norm": 0.5536530366080852,
+      "learning_rate": 3.942506642518952e-05,
+      "loss": 0.4246,
+      "step": 209
+    },
+    {
+      "epoch": 0.8502024291497976,
+      "grad_norm": 0.46824336808797895,
+      "learning_rate": 3.9411526122333396e-05,
+      "loss": 0.4235,
+      "step": 210
+    },
+    {
+      "epoch": 0.854251012145749,
+      "grad_norm": 0.47517147282195643,
+      "learning_rate": 3.939783060543167e-05,
+      "loss": 0.4379,
+      "step": 211
+    },
+    {
+      "epoch": 0.8582995951417004,
+      "grad_norm": 0.4620866410250017,
+      "learning_rate": 3.938397998399333e-05,
+      "loss": 0.4397,
+      "step": 212
+    },
+    {
+      "epoch": 0.8623481781376519,
+      "grad_norm": 0.4358989341005772,
+      "learning_rate": 3.936997436876756e-05,
+      "loss": 0.4332,
+      "step": 213
+    },
+    {
+      "epoch": 0.8663967611336032,
+      "grad_norm": 0.38706608545842797,
+      "learning_rate": 3.935581387174289e-05,
+      "loss": 0.4271,
+      "step": 214
+    },
+    {
+      "epoch": 0.8704453441295547,
+      "grad_norm": 0.474858638657314,
+      "learning_rate": 3.9341498606146295e-05,
+      "loss": 0.4406,
+      "step": 215
+    },
+    {
+      "epoch": 0.8744939271255061,
+      "grad_norm": 0.417238629435867,
+      "learning_rate": 3.932702868644224e-05,
+      "loss": 0.43,
+      "step": 216
+    },
+    {
+      "epoch": 0.8785425101214575,
+      "grad_norm": 0.47862215905888517,
+      "learning_rate": 3.931240422833181e-05,
+      "loss": 0.4596,
+      "step": 217
+    },
+    {
+      "epoch": 0.8825910931174089,
+      "grad_norm": 0.41414663345014274,
+      "learning_rate": 3.9297625348751806e-05,
+      "loss": 0.4119,
+      "step": 218
+    },
+    {
+      "epoch": 0.8866396761133604,
+      "grad_norm": 0.5162048285008454,
+      "learning_rate": 3.9282692165873744e-05,
+      "loss": 0.4253,
+      "step": 219
+    },
+    {
+      "epoch": 0.8906882591093117,
+      "grad_norm": 0.4835748947078622,
+      "learning_rate": 3.9267604799102945e-05,
+      "loss": 0.4188,
+      "step": 220
+    },
+    {
+      "epoch": 0.8947368421052632,
+      "grad_norm": 0.4063083793931881,
+      "learning_rate": 3.9252363369077615e-05,
+      "loss": 0.4169,
+      "step": 221
+    },
+    {
+      "epoch": 0.8987854251012146,
+      "grad_norm": 0.453338445149202,
+      "learning_rate": 3.9236967997667796e-05,
+      "loss": 0.4067,
+      "step": 222
+    },
+    {
+      "epoch": 0.902834008097166,
+      "grad_norm": 0.4670470198016601,
+      "learning_rate": 3.9221418807974496e-05,
+      "loss": 0.4359,
+      "step": 223
+    },
+    {
+      "epoch": 0.9068825910931174,
+      "grad_norm": 0.4100710929835924,
+      "learning_rate": 3.92057159243286e-05,
+      "loss": 0.4303,
+      "step": 224
+    },
+    {
+      "epoch": 0.9109311740890689,
+      "grad_norm": 0.4004758844580647,
+      "learning_rate": 3.9189859472289956e-05,
+      "loss": 0.418,
+      "step": 225
+    },
+    {
+      "epoch": 0.9149797570850202,
+      "grad_norm": 0.43711897460772176,
+      "learning_rate": 3.9173849578646324e-05,
+      "loss": 0.4154,
+      "step": 226
+    },
+    {
+      "epoch": 0.9190283400809717,
+      "grad_norm": 0.44840944715146097,
+      "learning_rate": 3.9157686371412395e-05,
+      "loss": 0.4085,
+      "step": 227
+    },
+    {
+      "epoch": 0.9230769230769231,
+      "grad_norm": 0.42604764132434253,
+      "learning_rate": 3.914136997982874e-05,
+      "loss": 0.424,
+      "step": 228
+    },
+    {
+      "epoch": 0.9271255060728745,
+      "grad_norm": 0.5066812929857512,
+      "learning_rate": 3.912490053436079e-05,
+      "loss": 0.4462,
+      "step": 229
+    },
+    {
+      "epoch": 0.9311740890688259,
+      "grad_norm": 0.47712526612460393,
+      "learning_rate": 3.9108278166697805e-05,
+      "loss": 0.4296,
+      "step": 230
+    },
+    {
+      "epoch": 0.9352226720647774,
+      "grad_norm": 0.5421199244144147,
+      "learning_rate": 3.909150300975178e-05,
+      "loss": 0.417,
+      "step": 231
+    },
+    {
+      "epoch": 0.9392712550607287,
+      "grad_norm": 0.47741582090983253,
+      "learning_rate": 3.907457519765642e-05,
+      "loss": 0.4362,
+      "step": 232
+    },
+    {
+      "epoch": 0.9433198380566802,
+      "grad_norm": 0.4217663636638482,
+      "learning_rate": 3.905749486576607e-05,
+      "loss": 0.4327,
+      "step": 233
+    },
+    {
+      "epoch": 0.9473684210526315,
+      "grad_norm": 0.46168738028607764,
+      "learning_rate": 3.90402621506546e-05,
+      "loss": 0.4197,
+      "step": 234
+    },
+    {
+      "epoch": 0.951417004048583,
+      "grad_norm": 0.3727511411358983,
+      "learning_rate": 3.902287719011435e-05,
+      "loss": 0.4357,
+      "step": 235
+    },
+    {
+      "epoch": 0.9554655870445344,
+      "grad_norm": 0.45235235739309426,
+      "learning_rate": 3.900534012315499e-05,
+      "loss": 0.4493,
+      "step": 236
+    },
+    {
+      "epoch": 0.9595141700404858,
+      "grad_norm": 0.4330814026988108,
+      "learning_rate": 3.8987651090002454e-05,
+      "loss": 0.439,
+      "step": 237
+    },
+    {
+      "epoch": 0.9635627530364372,
+      "grad_norm": 0.45657594574265176,
+      "learning_rate": 3.896981023209777e-05,
+      "loss": 0.4253,
+      "step": 238
+    },
+    {
+      "epoch": 0.9676113360323887,
+      "grad_norm": 0.3891002043437717,
+      "learning_rate": 3.8951817692095966e-05,
+      "loss": 0.4215,
+      "step": 239
+    },
+    {
+      "epoch": 0.97165991902834,
+      "grad_norm": 0.4763702199117018,
+      "learning_rate": 3.893367361386491e-05,
+      "loss": 0.4429,
+      "step": 240
+    },
+    {
+      "epoch": 0.9757085020242915,
+      "grad_norm": 0.46027560851878757,
+      "learning_rate": 3.891537814248417e-05,
+      "loss": 0.4358,
+      "step": 241
+    },
+    {
+      "epoch": 0.979757085020243,
+      "grad_norm": 0.43448556182189685,
+      "learning_rate": 3.889693142424384e-05,
+      "loss": 0.399,
+      "step": 242
+    },
+    {
+      "epoch": 0.9838056680161943,
+      "grad_norm": 0.3784640166413325,
+      "learning_rate": 3.8878333606643405e-05,
+      "loss": 0.4473,
+      "step": 243
+    },
+    {
+      "epoch": 0.9878542510121457,
+      "grad_norm": 0.4051126590074385,
+      "learning_rate": 3.885958483839049e-05,
+      "loss": 0.417,
+      "step": 244
+    },
+    {
+      "epoch": 0.9919028340080972,
+      "grad_norm": 0.40805306121193385,
+      "learning_rate": 3.8840685269399786e-05,
+      "loss": 0.4299,
+      "step": 245
+    },
+    {
+      "epoch": 0.9959514170040485,
+      "grad_norm": 0.39978762273571783,
+      "learning_rate": 3.882163505079171e-05,
+      "loss": 0.4282,
+      "step": 246
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.4371260338430428,
+      "learning_rate": 3.880243433489132e-05,
+      "loss": 0.4328,
+      "step": 247
+    },
+    {
+      "epoch": 1.0040485829959513,
+      "grad_norm": 0.465305538388603,
+      "learning_rate": 3.878308327522702e-05,
+      "loss": 0.3584,
+      "step": 248
+    },
+    {
+      "epoch": 1.008097165991903,
+      "grad_norm": 0.4357137046420451,
+      "learning_rate": 3.876358202652936e-05,
+      "loss": 0.3593,
+      "step": 249
+    },
+    {
+      "epoch": 1.0121457489878543,
+      "grad_norm": 0.4426935894064553,
+      "learning_rate": 3.8743930744729835e-05,
+      "loss": 0.3313,
+      "step": 250
+    },
+    {
+      "epoch": 1.0161943319838056,
+      "grad_norm": 0.5520787725508367,
+      "learning_rate": 3.8724129586959535e-05,
+      "loss": 0.3729,
+      "step": 251
+    },
+    {
+      "epoch": 1.0202429149797572,
+      "grad_norm": 0.4989031801681423,
+      "learning_rate": 3.870417871154801e-05,
+      "loss": 0.3567,
+      "step": 252
+    },
+    {
+      "epoch": 1.0242914979757085,
+      "grad_norm": 0.4884392134896784,
+      "learning_rate": 3.868407827802192e-05,
+      "loss": 0.3496,
+      "step": 253
+    },
+    {
+      "epoch": 1.0283400809716599,
+      "grad_norm": 0.4457461825027633,
+      "learning_rate": 3.8663828447103804e-05,
+      "loss": 0.3671,
+      "step": 254
+    },
+    {
+      "epoch": 1.0323886639676114,
+      "grad_norm": 0.5384905133993667,
+      "learning_rate": 3.8643429380710765e-05,
+      "loss": 0.3597,
+      "step": 255
+    },
+    {
+      "epoch": 1.0364372469635628,
+      "grad_norm": 0.4874619857433414,
+      "learning_rate": 3.862288124195319e-05,
+      "loss": 0.3464,
+      "step": 256
+    },
+    {
+      "epoch": 1.040485829959514,
+      "grad_norm": 0.4595096469571877,
+      "learning_rate": 3.8602184195133464e-05,
+      "loss": 0.3572,
+      "step": 257
+    },
+    {
+      "epoch": 1.0445344129554657,
+      "grad_norm": 0.5219455313703582,
+      "learning_rate": 3.8581338405744604e-05,
+      "loss": 0.3506,
+      "step": 258
+    },
+    {
+      "epoch": 1.048582995951417,
+      "grad_norm": 0.4758200003923128,
+      "learning_rate": 3.8560344040468996e-05,
+      "loss": 0.348,
+      "step": 259
+    },
+    {
+      "epoch": 1.0526315789473684,
+      "grad_norm": 0.45794026315188446,
+      "learning_rate": 3.853920126717703e-05,
+      "loss": 0.3544,
+      "step": 260
+    },
+    {
+      "epoch": 1.05668016194332,
+      "grad_norm": 0.5716818875148567,
+      "learning_rate": 3.851791025492574e-05,
+      "loss": 0.3681,
+      "step": 261
+    },
+    {
+      "epoch": 1.0607287449392713,
+      "grad_norm": 0.46291562263864183,
+      "learning_rate": 3.8496471173957515e-05,
+      "loss": 0.3404,
+      "step": 262
+    },
+    {
+      "epoch": 1.0647773279352226,
+      "grad_norm": 0.6009911516057322,
+      "learning_rate": 3.847488419569866e-05,
+      "loss": 0.3524,
+      "step": 263
+    },
+    {
+      "epoch": 1.0688259109311742,
+      "grad_norm": 0.5406764296951811,
+      "learning_rate": 3.8453149492758077e-05,
+      "loss": 0.3731,
+      "step": 264
+    },
+    {
+      "epoch": 1.0728744939271255,
+      "grad_norm": 0.4433433119033292,
+      "learning_rate": 3.8431267238925886e-05,
+      "loss": 0.3538,
+      "step": 265
+    },
+    {
+      "epoch": 1.0769230769230769,
+      "grad_norm": 0.5259410248629625,
+      "learning_rate": 3.840923760917199e-05,
+      "loss": 0.3401,
+      "step": 266
+    },
+    {
+      "epoch": 1.0809716599190284,
+      "grad_norm": 0.4615510400803921,
+      "learning_rate": 3.838706077964473e-05,
+      "loss": 0.341,
+      "step": 267
+    },
+    {
+      "epoch": 1.0850202429149798,
+      "grad_norm": 0.4957590396095555,
+      "learning_rate": 3.836473692766944e-05,
+      "loss": 0.3474,
+      "step": 268
+    },
+    {
+      "epoch": 1.0890688259109311,
+      "grad_norm": 0.5025432223729408,
+      "learning_rate": 3.834226623174705e-05,
+      "loss": 0.3501,
+      "step": 269
+    },
+    {
+      "epoch": 1.0931174089068827,
+      "grad_norm": 0.5117239963409878,
+      "learning_rate": 3.831964887155264e-05,
+      "loss": 0.3671,
+      "step": 270
+    },
+    {
+      "epoch": 1.097165991902834,
+      "grad_norm": 0.604087925762895,
+      "learning_rate": 3.829688502793403e-05,
+      "loss": 0.3375,
+      "step": 271
+    },
+    {
+      "epoch": 1.1012145748987854,
+      "grad_norm": 0.5030330365120181,
+      "learning_rate": 3.827397488291029e-05,
+      "loss": 0.3506,
+      "step": 272
+    },
+    {
+      "epoch": 1.1052631578947367,
+      "grad_norm": 0.5040237634534994,
+      "learning_rate": 3.8250918619670343e-05,
+      "loss": 0.3537,
+      "step": 273
+    },
+    {
+      "epoch": 1.1093117408906883,
+      "grad_norm": 0.5104525279501217,
+      "learning_rate": 3.822771642257145e-05,
+      "loss": 0.3544,
+      "step": 274
+    },
+    {
+      "epoch": 1.1133603238866396,
+      "grad_norm": 0.4500586236572952,
+      "learning_rate": 3.820436847713776e-05,
+      "loss": 0.3329,
+      "step": 275
+    },
+    {
+      "epoch": 1.117408906882591,
+      "grad_norm": 0.46557607836027703,
+      "learning_rate": 3.8180874970058826e-05,
+      "loss": 0.3652,
+      "step": 276
+    },
+    {
+      "epoch": 1.1214574898785425,
+      "grad_norm": 0.4543524185938012,
+      "learning_rate": 3.8157236089188103e-05,
+      "loss": 0.3366,
+      "step": 277
+    },
+    {
+      "epoch": 1.125506072874494,
+      "grad_norm": 0.5114038221294194,
+      "learning_rate": 3.813345202354145e-05,
+      "loss": 0.3517,
+      "step": 278
+    },
+    {
+      "epoch": 1.1295546558704452,
+      "grad_norm": 0.47319845214252315,
+      "learning_rate": 3.810952296329563e-05,
+      "loss": 0.358,
+      "step": 279
+    },
+    {
+      "epoch": 1.1336032388663968,
+      "grad_norm": 0.45950963722972027,
+      "learning_rate": 3.808544909978676e-05,
+      "loss": 0.3548,
+      "step": 280
+    },
+    {
+      "epoch": 1.1376518218623481,
+      "grad_norm": 0.42372635496674194,
+      "learning_rate": 3.806123062550882e-05,
+      "loss": 0.366,
+      "step": 281
+    },
+    {
+      "epoch": 1.1417004048582995,
+      "grad_norm": 0.4045986677877708,
+      "learning_rate": 3.803686773411208e-05,
+      "loss": 0.3448,
+      "step": 282
+    },
+    {
+      "epoch": 1.145748987854251,
+      "grad_norm": 0.38783663514545,
+      "learning_rate": 3.801236062040158e-05,
+      "loss": 0.3559,
+      "step": 283
+    },
+    {
+      "epoch": 1.1497975708502024,
+      "grad_norm": 0.39339401583460504,
+      "learning_rate": 3.798770948033553e-05,
+      "loss": 0.3392,
+      "step": 284
+    },
+    {
+      "epoch": 1.1538461538461537,
+      "grad_norm": 0.4481352835515607,
+      "learning_rate": 3.796291451102381e-05,
+      "loss": 0.3746,
+      "step": 285
+    },
+    {
+      "epoch": 1.1578947368421053,
+      "grad_norm": 0.47414780496934505,
+      "learning_rate": 3.793797591072631e-05,
+      "loss": 0.3503,
+      "step": 286
+    },
+    {
+      "epoch": 1.1619433198380567,
+      "grad_norm": 0.4203594505941164,
+      "learning_rate": 3.791289387885144e-05,
+      "loss": 0.3578,
+      "step": 287
+    },
+    {
+      "epoch": 1.165991902834008,
+      "grad_norm": 0.4361791489637381,
+      "learning_rate": 3.788766861595443e-05,
+      "loss": 0.3621,
+      "step": 288
+    },
+    {
+      "epoch": 1.1700404858299596,
+      "grad_norm": 0.4620949503042661,
+      "learning_rate": 3.7862300323735835e-05,
+      "loss": 0.3546,
+      "step": 289
+    },
+    {
+      "epoch": 1.174089068825911,
+      "grad_norm": 0.39125228680413304,
+      "learning_rate": 3.783678920503983e-05,
+      "loss": 0.3573,
+      "step": 290
+    },
+    {
+      "epoch": 1.1781376518218623,
+      "grad_norm": 0.5185550223300976,
+      "learning_rate": 3.781113546385265e-05,
+      "loss": 0.3539,
+      "step": 291
+    },
+    {
+      "epoch": 1.1821862348178138,
+      "grad_norm": 0.4600822686191904,
+      "learning_rate": 3.7785339305300916e-05,
+      "loss": 0.3751,
+      "step": 292
+    },
+    {
+      "epoch": 1.1862348178137652,
+      "grad_norm": 0.4364600912966501,
+      "learning_rate": 3.7759400935650025e-05,
+      "loss": 0.3648,
+      "step": 293
+    },
+    {
+      "epoch": 1.1902834008097165,
+      "grad_norm": 0.4068506803826684,
+      "learning_rate": 3.77333205623025e-05,
+      "loss": 0.3591,
+      "step": 294
+    },
+    {
+      "epoch": 1.194331983805668,
+      "grad_norm": 0.41954478104916765,
+      "learning_rate": 3.7707098393796295e-05,
+      "loss": 0.3649,
+      "step": 295
+    },
+    {
+      "epoch": 1.1983805668016194,
+      "grad_norm": 0.4562013640500038,
+      "learning_rate": 3.7680734639803185e-05,
+      "loss": 0.3713,
+      "step": 296
+    },
+    {
+      "epoch": 1.2024291497975708,
+      "grad_norm": 0.45072198383738116,
+      "learning_rate": 3.765422951112704e-05,
+      "loss": 0.3658,
+      "step": 297
+    },
+    {
+      "epoch": 1.2064777327935223,
+      "grad_norm": 0.48078387838786174,
+      "learning_rate": 3.762758321970216e-05,
+      "loss": 0.3539,
+      "step": 298
+    },
+    {
+      "epoch": 1.2105263157894737,
+      "grad_norm": 0.42297461808755576,
+      "learning_rate": 3.7600795978591584e-05,
+      "loss": 0.3263,
+      "step": 299
+    },
+    {
+      "epoch": 1.214574898785425,
+      "grad_norm": 0.5154957998052971,
+      "learning_rate": 3.757386800198538e-05,
+      "loss": 0.3605,
+      "step": 300
+    },
+    {
+      "epoch": 1.2186234817813766,
+      "grad_norm": 0.40042893033683963,
+      "learning_rate": 3.7546799505198925e-05,
+      "loss": 0.3548,
+      "step": 301
+    },
+    {
+      "epoch": 1.222672064777328,
+      "grad_norm": 0.46014487999615655,
+      "learning_rate": 3.7519590704671197e-05,
+      "loss": 0.3382,
+      "step": 302
+    },
+    {
+      "epoch": 1.2267206477732793,
+      "grad_norm": 0.434626282342408,
+      "learning_rate": 3.749224181796305e-05,
+      "loss": 0.3736,
+      "step": 303
+    },
+    {
+      "epoch": 1.2307692307692308,
+      "grad_norm": 0.41173917054035764,
+      "learning_rate": 3.746475306375545e-05,
+      "loss": 0.3436,
+      "step": 304
+    },
+    {
+      "epoch": 1.2348178137651822,
+      "grad_norm": 0.44723998904633516,
+      "learning_rate": 3.743712466184774e-05,
+      "loss": 0.3672,
+      "step": 305
+    },
+    {
+      "epoch": 1.2388663967611335,
+      "grad_norm": 0.497023890288107,
+      "learning_rate": 3.7409356833155885e-05,
+      "loss": 0.3463,
+      "step": 306
+    },
+    {
+      "epoch": 1.242914979757085,
+      "grad_norm": 0.5179014564439103,
+      "learning_rate": 3.73814497997107e-05,
+      "loss": 0.3718,
+      "step": 307
+    },
+    {
+      "epoch": 1.2469635627530364,
+      "grad_norm": 0.5737346121071342,
+      "learning_rate": 3.7353403784656095e-05,
+      "loss": 0.3686,
+      "step": 308
+    },
+    {
+      "epoch": 1.2510121457489878,
+      "grad_norm": 0.4172583889321656,
+      "learning_rate": 3.732521901224724e-05,
+      "loss": 0.343,
+      "step": 309
+    },
+    {
+      "epoch": 1.2550607287449393,
+      "grad_norm": 0.5312453169984984,
+      "learning_rate": 3.7296895707848844e-05,
+      "loss": 0.346,
+      "step": 310
+    },
+    {
+      "epoch": 1.2591093117408907,
+      "grad_norm": 0.37372400845261516,
+      "learning_rate": 3.7268434097933275e-05,
+      "loss": 0.3734,
+      "step": 311
+    },
+    {
+      "epoch": 1.263157894736842,
+      "grad_norm": 0.4914672271555722,
+      "learning_rate": 3.723983441007882e-05,
+      "loss": 0.3527,
+      "step": 312
+    },
+    {
+      "epoch": 1.2672064777327936,
+      "grad_norm": 0.409318438935981,
+      "learning_rate": 3.721109687296781e-05,
+      "loss": 0.3684,
+      "step": 313
+    },
+    {
+      "epoch": 1.271255060728745,
+      "grad_norm": 0.5105431156518849,
+      "learning_rate": 3.718222171638484e-05,
+      "loss": 0.3531,
+      "step": 314
+    },
+    {
+      "epoch": 1.2753036437246963,
+      "grad_norm": 0.4152617305037566,
+      "learning_rate": 3.715320917121488e-05,
+      "loss": 0.3476,
+      "step": 315
+    },
+    {
+      "epoch": 1.2793522267206479,
+      "grad_norm": 0.41727113931401616,
+      "learning_rate": 3.712405946944149e-05,
+      "loss": 0.3727,
+      "step": 316
+    },
+    {
+      "epoch": 1.2834008097165992,
+      "grad_norm": 0.40683866204464725,
+      "learning_rate": 3.709477284414488e-05,
+      "loss": 0.3748,
+      "step": 317
+    },
+    {
+      "epoch": 1.2874493927125505,
+      "grad_norm": 0.4519393947530217,
+      "learning_rate": 3.7065349529500136e-05,
+      "loss": 0.3527,
+      "step": 318
+    },
+    {
+      "epoch": 1.291497975708502,
+      "grad_norm": 0.40241376490169906,
+      "learning_rate": 3.7035789760775305e-05,
+      "loss": 0.3639,
+      "step": 319
+    },
+    {
+      "epoch": 1.2955465587044535,
+      "grad_norm": 0.4232199138850649,
+      "learning_rate": 3.70060937743295e-05,
+      "loss": 0.3501,
+      "step": 320
+    },
+    {
+      "epoch": 1.2995951417004048,
+      "grad_norm": 0.36274624528096877,
+      "learning_rate": 3.6976261807611045e-05,
+      "loss": 0.3541,
+      "step": 321
+    },
+    {
+      "epoch": 1.3036437246963564,
+      "grad_norm": 0.42648907689404175,
+      "learning_rate": 3.694629409915555e-05,
+      "loss": 0.3489,
+      "step": 322
+    },
+    {
+      "epoch": 1.3076923076923077,
+      "grad_norm": 0.3701985609722712,
+      "learning_rate": 3.691619088858401e-05,
+      "loss": 0.3612,
+      "step": 323
+    },
+    {
+      "epoch": 1.311740890688259,
+      "grad_norm": 0.39675656981373797,
+      "learning_rate": 3.68859524166009e-05,
+      "loss": 0.3514,
+      "step": 324
+    },
+    {
+      "epoch": 1.3157894736842106,
+      "grad_norm": 0.3654222116628682,
+      "learning_rate": 3.6855578924992226e-05,
+      "loss": 0.3422,
+      "step": 325
+    },
+    {
+      "epoch": 1.319838056680162,
+      "grad_norm": 0.44176590174865576,
+      "learning_rate": 3.6825070656623626e-05,
+      "loss": 0.3544,
+      "step": 326
+    },
+    {
+      "epoch": 1.3238866396761133,
+      "grad_norm": 0.3754579375352588,
+      "learning_rate": 3.67944278554384e-05,
+      "loss": 0.3451,
+      "step": 327
+    },
+    {
+      "epoch": 1.3279352226720649,
+      "grad_norm": 0.46235223891852434,
+      "learning_rate": 3.676365076645557e-05,
+      "loss": 0.3537,
+      "step": 328
+    },
+    {
+      "epoch": 1.3319838056680162,
+      "grad_norm": 0.41176521783949305,
+      "learning_rate": 3.673273963576791e-05,
+      "loss": 0.3543,
+      "step": 329
+    },
+    {
+      "epoch": 1.3360323886639676,
+      "grad_norm": 0.44735165925935494,
+      "learning_rate": 3.670169471054001e-05,
+      "loss": 0.3652,
+      "step": 330
+    },
+    {
+      "epoch": 1.3400809716599191,
+      "grad_norm": 0.4010522512857303,
+      "learning_rate": 3.6670516239006254e-05,
+      "loss": 0.3566,
+      "step": 331
+    },
+    {
+      "epoch": 1.3441295546558705,
+      "grad_norm": 0.4718151694257999,
+      "learning_rate": 3.663920447046888e-05,
+      "loss": 0.3824,
+      "step": 332
+    },
+    {
+      "epoch": 1.3481781376518218,
+      "grad_norm": 0.3912094602223265,
+      "learning_rate": 3.6607759655295954e-05,
+      "loss": 0.3495,
+      "step": 333
+    },
+    {
+      "epoch": 1.3522267206477734,
+      "grad_norm": 0.40325629525497897,
+      "learning_rate": 3.657618204491937e-05,
+      "loss": 0.3579,
+      "step": 334
+    },
+    {
+      "epoch": 1.3562753036437247,
+      "grad_norm": 0.3833572344906781,
+      "learning_rate": 3.6544471891832875e-05,
+      "loss": 0.3569,
+      "step": 335
+    },
+    {
+      "epoch": 1.360323886639676,
+      "grad_norm": 0.42635950388921917,
+      "learning_rate": 3.6512629449589983e-05,
+      "loss": 0.3619,
+      "step": 336
+    },
+    {
+      "epoch": 1.3643724696356276,
+      "grad_norm": 0.4200106484070262,
+      "learning_rate": 3.6480654972802045e-05,
+      "loss": 0.3725,
+      "step": 337
+    },
+    {
+      "epoch": 1.368421052631579,
+      "grad_norm": 0.4452391088329314,
+      "learning_rate": 3.644854871713611e-05,
+      "loss": 0.343,
+      "step": 338
+    },
+    {
+      "epoch": 1.3724696356275303,
+      "grad_norm": 0.6466532123047348,
+      "learning_rate": 3.6416310939312924e-05,
+      "loss": 0.3757,
+      "step": 339
+    },
+    {
+      "epoch": 1.376518218623482,
+      "grad_norm": 0.42859053908753686,
+      "learning_rate": 3.638394189710493e-05,
+      "loss": 0.3692,
+      "step": 340
+    },
+    {
+      "epoch": 1.3805668016194332,
+      "grad_norm": 0.3775732164285291,
+      "learning_rate": 3.635144184933412e-05,
+      "loss": 0.3593,
+      "step": 341
+    },
+    {
+      "epoch": 1.3846153846153846,
+      "grad_norm": 0.42740264903608777,
+      "learning_rate": 3.6318811055869994e-05,
+      "loss": 0.3853,
+      "step": 342
+    },
+    {
+      "epoch": 1.3886639676113361,
+      "grad_norm": 0.3624414718027014,
+      "learning_rate": 3.628604977762752e-05,
+      "loss": 0.3606,
+      "step": 343
+    },
+    {
+      "epoch": 1.3927125506072875,
+      "grad_norm": 0.39155227268560355,
+      "learning_rate": 3.6253158276565004e-05,
+      "loss": 0.3515,
+      "step": 344
+    },
+    {
+      "epoch": 1.3967611336032388,
+      "grad_norm": 0.37164386504553587,
+      "learning_rate": 3.6220136815682016e-05,
+      "loss": 0.3664,
+      "step": 345
+    },
+    {
+      "epoch": 1.4008097165991904,
+      "grad_norm": 0.37640677663116495,
+      "learning_rate": 3.618698565901727e-05,
+      "loss": 0.3708,
+      "step": 346
+    },
+    {
+      "epoch": 1.4048582995951417,
+      "grad_norm": 0.371909827521948,
+      "learning_rate": 3.615370507164654e-05,
+      "loss": 0.3668,
+      "step": 347
+    },
+    {
+      "epoch": 1.408906882591093,
+      "grad_norm": 0.3860066926009176,
+      "learning_rate": 3.61202953196805e-05,
+      "loss": 0.3565,
+      "step": 348
+    },
+    {
+      "epoch": 1.4129554655870447,
+      "grad_norm": 0.37306468488984573,
+      "learning_rate": 3.6086756670262654e-05,
+      "loss": 0.3596,
+      "step": 349
+    },
+    {
+      "epoch": 1.417004048582996,
+      "grad_norm": 0.38490691961146434,
+      "learning_rate": 3.6053089391567135e-05,
+      "loss": 0.3641,
+      "step": 350
+    },
+    {
+      "epoch": 1.4210526315789473,
+      "grad_norm": 0.3802534111619299,
+      "learning_rate": 3.601929375279662e-05,
+      "loss": 0.3467,
+      "step": 351
+    },
+    {
+      "epoch": 1.425101214574899,
+      "grad_norm": 0.39936885814919976,
+      "learning_rate": 3.598537002418012e-05,
+      "loss": 0.3474,
+      "step": 352
+    },
+    {
+      "epoch": 1.4291497975708503,
+      "grad_norm": 0.3809378897845386,
+      "learning_rate": 3.5951318476970885e-05,
+      "loss": 0.3646,
+      "step": 353
+    },
+    {
+      "epoch": 1.4331983805668016,
+      "grad_norm": 1.0244313638811569,
+      "learning_rate": 3.591713938344416e-05,
+      "loss": 0.3546,
+      "step": 354
+    },
+    {
+      "epoch": 1.4372469635627532,
+      "grad_norm": 3.8730301657358086,
+      "learning_rate": 3.588283301689507e-05,
+      "loss": 0.355,
+      "step": 355
+    },
+    {
+      "epoch": 1.4412955465587045,
+      "grad_norm": 0.5404268516625832,
+      "learning_rate": 3.5848399651636424e-05,
+      "loss": 0.3681,
+      "step": 356
+    },
+    {
+      "epoch": 1.4453441295546559,
+      "grad_norm": 0.5690956688774892,
+      "learning_rate": 3.5813839562996473e-05,
+      "loss": 0.3583,
+      "step": 357
+    },
+    {
+      "epoch": 1.4493927125506074,
+      "grad_norm": 0.42003127431072895,
+      "learning_rate": 3.5779153027316775e-05,
+      "loss": 0.3549,
+      "step": 358
+    },
+    {
+      "epoch": 1.4534412955465588,
+      "grad_norm": 0.5933784818458554,
+      "learning_rate": 3.574434032194994e-05,
+      "loss": 0.362,
+      "step": 359
+    },
+    {
+      "epoch": 1.45748987854251,
+      "grad_norm": 1.810338254086514,
+      "learning_rate": 3.5709401725257436e-05,
+      "loss": 0.3507,
+      "step": 360
+    },
+    {
+      "epoch": 1.4615384615384617,
+      "grad_norm": 0.6582593918153665,
+      "learning_rate": 3.5674337516607346e-05,
+      "loss": 0.3643,
+      "step": 361
+    },
+    {
+      "epoch": 1.465587044534413,
+      "grad_norm": 1.472790613015716,
+      "learning_rate": 3.563914797637216e-05,
+      "loss": 0.3653,
+      "step": 362
+    },
+    {
+      "epoch": 1.4696356275303644,
+      "grad_norm": 0.6423062333880559,
+      "learning_rate": 3.560383338592649e-05,
+      "loss": 0.3638,
+      "step": 363
+    },
+    {
+      "epoch": 1.4736842105263157,
+      "grad_norm": 5.503726556121286,
+      "learning_rate": 3.556839402764487e-05,
+      "loss": 0.3611,
+      "step": 364
+    },
+    {
+      "epoch": 1.4777327935222673,
+      "grad_norm": 1.2101198553750068,
+      "learning_rate": 3.553283018489946e-05,
+      "loss": 0.3704,
+      "step": 365
+    },
+    {
+      "epoch": 1.4817813765182186,
+      "grad_norm": 0.7144164646499227,
+      "learning_rate": 3.54971421420578e-05,
+      "loss": 0.3767,
+      "step": 366
+    },
+    {
+      "epoch": 1.48582995951417,
+      "grad_norm": 0.8722699518003545,
+      "learning_rate": 3.5461330184480524e-05,
+      "loss": 0.3588,
+      "step": 367
+    },
+    {
+      "epoch": 1.4898785425101215,
+      "grad_norm": 0.636155044866059,
+      "learning_rate": 3.54253945985191e-05,
+      "loss": 0.35,
+      "step": 368
+    },
+    {
+      "epoch": 1.4939271255060729,
+      "grad_norm": 0.8650009495050177,
+      "learning_rate": 3.5389335671513524e-05,
+      "loss": 0.3624,
+      "step": 369
+    },
+    {
+      "epoch": 1.4979757085020242,
+      "grad_norm": 0.6364564653122696,
+      "learning_rate": 3.5353153691789986e-05,
+      "loss": 0.3587,
+      "step": 370
+    },
+    {
+      "epoch": 1.5020242914979756,
+      "grad_norm": 0.7387109854908066,
+      "learning_rate": 3.531684894865866e-05,
+      "loss": 0.3785,
+      "step": 371
+    },
+    {
+      "epoch": 1.5060728744939271,
+      "grad_norm": 0.8284117835255285,
+      "learning_rate": 3.528042173241128e-05,
+      "loss": 0.3682,
+      "step": 372
+    },
+    {
+      "epoch": 1.5101214574898787,
+      "grad_norm": 4.009831244711985,
+      "learning_rate": 3.524387233431892e-05,
+      "loss": 0.3705,
+      "step": 373
+    },
+    {
+      "epoch": 1.5141700404858298,
+      "grad_norm": 0.9340231115897835,
+      "learning_rate": 3.520720104662958e-05,
+      "loss": 0.3626,
+      "step": 374
+    },
+    {
+      "epoch": 1.5182186234817814,
+      "grad_norm": 0.5350841771437106,
+      "learning_rate": 3.5170408162565904e-05,
+      "loss": 0.3687,
+      "step": 375
+    },
+    {
+      "epoch": 1.522267206477733,
+      "grad_norm": 0.7384608706470113,
+      "learning_rate": 3.513349397632281e-05,
+      "loss": 0.3721,
+      "step": 376
+    },
+    {
+      "epoch": 1.526315789473684,
+      "grad_norm": 0.48939263262622673,
+      "learning_rate": 3.5096458783065145e-05,
+      "loss": 0.3555,
+      "step": 377
+    },
+    {
+      "epoch": 1.5303643724696356,
+      "grad_norm": 0.5470450901362519,
+      "learning_rate": 3.505930287892533e-05,
+      "loss": 0.365,
+      "step": 378
+    },
+    {
+      "epoch": 1.5344129554655872,
+      "grad_norm": 0.5412079349941012,
+      "learning_rate": 3.502202656100099e-05,
+      "loss": 0.3585,
+      "step": 379
+    },
+    {
+      "epoch": 1.5384615384615383,
+      "grad_norm": 0.5456794447913396,
+      "learning_rate": 3.498463012735256e-05,
+      "loss": 0.3671,
+      "step": 380
+    },
+    {
+      "epoch": 1.54251012145749,
+      "grad_norm": 0.519351701709218,
+      "learning_rate": 3.494711387700094e-05,
+      "loss": 0.3461,
+      "step": 381
+    },
+    {
+      "epoch": 1.5465587044534415,
+      "grad_norm": 0.5013699350239484,
+      "learning_rate": 3.490947810992505e-05,
+      "loss": 0.343,
+      "step": 382
+    },
+    {
+      "epoch": 1.5506072874493926,
+      "grad_norm": 0.4419993697055499,
+      "learning_rate": 3.48717231270595e-05,
+      "loss": 0.3556,
+      "step": 383
+    },
+    {
+      "epoch": 1.5546558704453441,
+      "grad_norm": 0.48974178786411354,
+      "learning_rate": 3.483384923029211e-05,
+      "loss": 0.3571,
+      "step": 384
+    },
+    {
+      "epoch": 1.5587044534412957,
+      "grad_norm": 0.469114846959984,
+      "learning_rate": 3.479585672246156e-05,
+      "loss": 0.3554,
+      "step": 385
+    },
+    {
+      "epoch": 1.5627530364372468,
+      "grad_norm": 0.4582780695685735,
+      "learning_rate": 3.4757745907354924e-05,
+      "loss": 0.3753,
+      "step": 386
+    },
+    {
+      "epoch": 1.5668016194331984,
+      "grad_norm": 0.4924140417601524,
+      "learning_rate": 3.471951708970527e-05,
+      "loss": 0.3334,
+      "step": 387
+    },
+    {
+      "epoch": 1.5708502024291497,
+      "grad_norm": 0.4155715613963137,
+      "learning_rate": 3.468117057518921e-05,
+      "loss": 0.3626,
+      "step": 388
+    },
+    {
+      "epoch": 1.574898785425101,
+      "grad_norm": 0.47671774773990183,
+      "learning_rate": 3.4642706670424426e-05,
+      "loss": 0.3607,
+      "step": 389
+    },
+    {
+      "epoch": 1.5789473684210527,
+      "grad_norm": 1.048748130364427,
+      "learning_rate": 3.460412568296731e-05,
+      "loss": 0.3588,
+      "step": 390
+    },
+    {
+      "epoch": 1.582995951417004,
+      "grad_norm": 0.5240424571819843,
+      "learning_rate": 3.456542792131039e-05,
+      "loss": 0.375,
+      "step": 391
+    },
+    {
+      "epoch": 1.5870445344129553,
+      "grad_norm": 0.41172639110449755,
+      "learning_rate": 3.4526613694879936e-05,
+      "loss": 0.3697,
+      "step": 392
+    },
+    {
+      "epoch": 1.591093117408907,
+      "grad_norm": 0.4678883898274471,
+      "learning_rate": 3.448768331403348e-05,
+      "loss": 0.3634,
+      "step": 393
+    },
+    {
+      "epoch": 1.5951417004048583,
+      "grad_norm": 1.024972530987665,
+      "learning_rate": 3.44486370900573e-05,
+      "loss": 0.3568,
+      "step": 394
+    },
+    {
+      "epoch": 1.5991902834008096,
+      "grad_norm": 0.4182316809572888,
+      "learning_rate": 3.440947533516396e-05,
+      "loss": 0.3697,
+      "step": 395
+    },
+    {
+      "epoch": 1.6032388663967612,
+      "grad_norm": 1.355095888447291,
+      "learning_rate": 3.4370198362489816e-05,
+      "loss": 0.3744,
+      "step": 396
+    },
+    {
+      "epoch": 1.6072874493927125,
+      "grad_norm": 0.40091682560297814,
+      "learning_rate": 3.43308064860925e-05,
+      "loss": 0.3597,
+      "step": 397
+    },
+    {
+      "epoch": 1.6113360323886639,
+      "grad_norm": 1.4053693889519092,
+      "learning_rate": 3.4291300020948393e-05,
+      "loss": 0.3921,
+      "step": 398
+    },
+    {
+      "epoch": 1.6153846153846154,
+      "grad_norm": 1.3567334959306179,
+      "learning_rate": 3.425167928295015e-05,
+      "loss": 0.4032,
+      "step": 399
+    },
+    {
+      "epoch": 1.6194331983805668,
+      "grad_norm": 0.4985817823859577,
+      "learning_rate": 3.421194458890411e-05,
+      "loss": 0.3557,
+      "step": 400
+    },
+    {
+      "epoch": 1.623481781376518,
+      "grad_norm": 0.6965847783244041,
+      "learning_rate": 3.4172096256527865e-05,
+      "loss": 0.3471,
+      "step": 401
+    },
+    {
+      "epoch": 1.6275303643724697,
+      "grad_norm": 0.5379423781581781,
+      "learning_rate": 3.4132134604447595e-05,
+      "loss": 0.3617,
+      "step": 402
+    },
+    {
+      "epoch": 1.631578947368421,
+      "grad_norm": 1.2324954639249224,
+      "learning_rate": 3.409205995219561e-05,
+      "loss": 0.3716,
+      "step": 403
+    },
+    {
+      "epoch": 1.6356275303643724,
+      "grad_norm": 0.5098955786036317,
+      "learning_rate": 3.4051872620207765e-05,
+      "loss": 0.3567,
+      "step": 404
+    },
+    {
+      "epoch": 1.639676113360324,
+      "grad_norm": 0.42185663437374343,
+      "learning_rate": 3.40115729298209e-05,
+      "loss": 0.3689,
+      "step": 405
+    },
+    {
+      "epoch": 1.6437246963562753,
+      "grad_norm": 0.6946152891186841,
+      "learning_rate": 3.397116120327025e-05,
+      "loss": 0.364,
+      "step": 406
+    },
+    {
+      "epoch": 1.6477732793522266,
+      "grad_norm": 0.4799485916263966,
+      "learning_rate": 3.3930637763686916e-05,
+      "loss": 0.3332,
+      "step": 407
+    },
+    {
+      "epoch": 1.6518218623481782,
+      "grad_norm": 0.40272401734953733,
+      "learning_rate": 3.389000293509524e-05,
+      "loss": 0.3624,
+      "step": 408
+    },
+    {
+      "epoch": 1.6558704453441295,
+      "grad_norm": 0.4476446859765025,
+      "learning_rate": 3.3849257042410225e-05,
+      "loss": 0.3481,
+      "step": 409
+    },
+    {
+      "epoch": 1.6599190283400809,
+      "grad_norm": 0.40862308468252284,
+      "learning_rate": 3.380840041143494e-05,
+      "loss": 0.3598,
+      "step": 410
+    },
+    {
+      "epoch": 1.6639676113360324,
+      "grad_norm": 0.48840827953219446,
+      "learning_rate": 3.376743336885792e-05,
+      "loss": 0.3595,
+      "step": 411
+    },
+    {
+      "epoch": 1.6680161943319838,
+      "grad_norm": 0.4692295462267613,
+      "learning_rate": 3.372635624225055e-05,
+      "loss": 0.3523,
+      "step": 412
+    },
+    {
+      "epoch": 1.6720647773279351,
+      "grad_norm": 0.37421034670791337,
+      "learning_rate": 3.3685169360064445e-05,
+      "loss": 0.3636,
+      "step": 413
+    },
+    {
+      "epoch": 1.6761133603238867,
+      "grad_norm": 0.4466430419621567,
+      "learning_rate": 3.3643873051628804e-05,
+      "loss": 0.359,
+      "step": 414
+    },
+    {
+      "epoch": 1.680161943319838,
+      "grad_norm": 0.3450755629566703,
+      "learning_rate": 3.3602467647147814e-05,
+      "loss": 0.3598,
+      "step": 415
+    },
+    {
+      "epoch": 1.6842105263157894,
+      "grad_norm": 0.41971440115016123,
+      "learning_rate": 3.356095347769799e-05,
+      "loss": 0.3518,
+      "step": 416
+    },
+    {
+      "epoch": 1.688259109311741,
+      "grad_norm": 0.3334645291567077,
+      "learning_rate": 3.351933087522552e-05,
+      "loss": 0.3664,
+      "step": 417
+    },
+    {
+      "epoch": 1.6923076923076923,
+      "grad_norm": 0.4246278893858576,
+      "learning_rate": 3.3477600172543626e-05,
+      "loss": 0.3686,
+      "step": 418
+    },
+    {
+      "epoch": 1.6963562753036436,
+      "grad_norm": 0.9216838183451351,
+      "learning_rate": 3.3435761703329894e-05,
+      "loss": 0.3916,
+      "step": 419
+    },
+    {
+      "epoch": 1.7004048582995952,
+      "grad_norm": 0.3569659034706086,
+      "learning_rate": 3.3393815802123606e-05,
+      "loss": 0.3459,
+      "step": 420
+    },
+    {
+      "epoch": 1.7044534412955465,
+      "grad_norm": 0.3773251566271473,
+      "learning_rate": 3.335176280432308e-05,
+      "loss": 0.3519,
+      "step": 421
+    },
+    {
+      "epoch": 1.708502024291498,
+      "grad_norm": 0.37516387337014606,
+      "learning_rate": 3.330960304618295e-05,
+      "loss": 0.3612,
+      "step": 422
+    },
+    {
+      "epoch": 1.7125506072874495,
+      "grad_norm": 0.38264000128175485,
+      "learning_rate": 3.326733686481153e-05,
+      "loss": 0.3622,
+      "step": 423
+    },
+    {
+      "epoch": 1.7165991902834008,
+      "grad_norm": 0.375881661660956,
+      "learning_rate": 3.3224964598168084e-05,
+      "loss": 0.3601,
+      "step": 424
+    },
+    {
+      "epoch": 1.7206477732793521,
+      "grad_norm": 0.533823802607281,
+      "learning_rate": 3.318248658506012e-05,
+      "loss": 0.3701,
+      "step": 425
+    },
+    {
+      "epoch": 1.7246963562753037,
+      "grad_norm": 0.40218862077524825,
+      "learning_rate": 3.313990316514069e-05,
+      "loss": 0.3468,
+      "step": 426
+    },
+    {
+      "epoch": 1.728744939271255,
+      "grad_norm": 0.3371563307430662,
+      "learning_rate": 3.309721467890571e-05,
+      "loss": 0.3584,
+      "step": 427
+    },
+    {
+      "epoch": 1.7327935222672064,
+      "grad_norm": 0.32184378419883336,
+      "learning_rate": 3.305442146769115e-05,
+      "loss": 0.354,
+      "step": 428
+    },
+    {
+      "epoch": 1.736842105263158,
+      "grad_norm": 1.2027293076637269,
+      "learning_rate": 3.30115238736704e-05,
+      "loss": 0.3684,
+      "step": 429
+    },
+    {
+      "epoch": 1.7408906882591093,
+      "grad_norm": 0.36435766735496145,
+      "learning_rate": 3.296852223985148e-05,
+      "loss": 0.3566,
+      "step": 430
+    },
+    {
+      "epoch": 1.7449392712550607,
+      "grad_norm": 0.3384960778020678,
+      "learning_rate": 3.2925416910074315e-05,
+      "loss": 0.3511,
+      "step": 431
+    },
+    {
+      "epoch": 1.7489878542510122,
+      "grad_norm": 0.33692134294122994,
+      "learning_rate": 3.288220822900796e-05,
+      "loss": 0.3466,
+      "step": 432
+    },
+    {
+      "epoch": 1.7530364372469636,
+      "grad_norm": 0.33738559872560964,
+      "learning_rate": 3.283889654214788e-05,
+      "loss": 0.3731,
+      "step": 433
+    },
+    {
+      "epoch": 1.757085020242915,
+      "grad_norm": 0.3511104525642406,
+      "learning_rate": 3.279548219581317e-05,
+      "loss": 0.3569,
+      "step": 434
+    },
+    {
+      "epoch": 1.7611336032388665,
+      "grad_norm": 0.41089831803843024,
+      "learning_rate": 3.275196553714379e-05,
+      "loss": 0.36,
+      "step": 435
+    },
+    {
+      "epoch": 1.7651821862348178,
+      "grad_norm": 0.36452323972075934,
+      "learning_rate": 3.270834691409778e-05,
+      "loss": 0.3542,
+      "step": 436
+    },
+    {
+      "epoch": 1.7692307692307692,
+      "grad_norm": 0.33946164991531,
+      "learning_rate": 3.266462667544849e-05,
+      "loss": 0.3466,
+      "step": 437
+    },
+    {
+      "epoch": 1.7732793522267207,
+      "grad_norm": 0.4108835526050675,
+      "learning_rate": 3.262080517078178e-05,
+      "loss": 0.3676,
+      "step": 438
+    },
+    {
+      "epoch": 1.777327935222672,
+      "grad_norm": 0.3571165953174601,
+      "learning_rate": 3.257688275049323e-05,
+      "loss": 0.3564,
+      "step": 439
+    },
+    {
+      "epoch": 1.7813765182186234,
+      "grad_norm": 0.3582643427350201,
+      "learning_rate": 3.253285976578535e-05,
+      "loss": 0.3512,
+      "step": 440
+    },
+    {
+      "epoch": 1.785425101214575,
+      "grad_norm": 0.5069524316215829,
+      "learning_rate": 3.2488736568664756e-05,
+      "loss": 0.3594,
+      "step": 441
+    },
+    {
+      "epoch": 1.7894736842105263,
+      "grad_norm": 0.4162006811967158,
+      "learning_rate": 3.244451351193935e-05,
+      "loss": 0.3634,
+      "step": 442
+    },
+    {
+      "epoch": 1.7935222672064777,
+      "grad_norm": 0.37907129440190784,
+      "learning_rate": 3.240019094921551e-05,
+      "loss": 0.3516,
+      "step": 443
+    },
+    {
+      "epoch": 1.7975708502024292,
+      "grad_norm": 0.3687870067326227,
+      "learning_rate": 3.235576923489527e-05,
+      "loss": 0.3546,
+      "step": 444
+    },
+    {
+      "epoch": 1.8016194331983806,
+      "grad_norm": 0.4052122133090806,
+      "learning_rate": 3.2311248724173476e-05,
+      "loss": 0.3726,
+      "step": 445
+    },
+    {
+      "epoch": 1.805668016194332,
+      "grad_norm": 0.342010372844965,
+      "learning_rate": 3.226662977303494e-05,
+      "loss": 0.3511,
+      "step": 446
+    },
+    {
+      "epoch": 1.8097165991902835,
+      "grad_norm": 0.3713407346250065,
+      "learning_rate": 3.2221912738251626e-05,
+      "loss": 0.3658,
+      "step": 447
+    },
+    {
+      "epoch": 1.8137651821862348,
+      "grad_norm": 0.37070728054893737,
+      "learning_rate": 3.2177097977379736e-05,
+      "loss": 0.3558,
+      "step": 448
+    },
+    {
+      "epoch": 1.8178137651821862,
+      "grad_norm": 0.3762300089526182,
+      "learning_rate": 3.21321858487569e-05,
+      "loss": 0.362,
+      "step": 449
+    },
+    {
+      "epoch": 1.8218623481781377,
+      "grad_norm": 0.34828960402944115,
+      "learning_rate": 3.208717671149931e-05,
+      "loss": 0.3471,
+      "step": 450
+    },
+    {
+      "epoch": 1.825910931174089,
+      "grad_norm": 0.3724360959500643,
+      "learning_rate": 3.204207092549882e-05,
+      "loss": 0.3557,
+      "step": 451
+    },
+    {
+      "epoch": 1.8299595141700404,
+      "grad_norm": 0.3510541656212395,
+      "learning_rate": 3.19968688514201e-05,
+      "loss": 0.3413,
+      "step": 452
+    },
+    {
+      "epoch": 1.834008097165992,
+      "grad_norm": 0.36833515094603747,
+      "learning_rate": 3.195157085069774e-05,
+      "loss": 0.3398,
+      "step": 453
+    },
+    {
+      "epoch": 1.8380566801619433,
+      "grad_norm": 0.3868497303848924,
+      "learning_rate": 3.190617728553332e-05,
+      "loss": 0.367,
+      "step": 454
+    },
+    {
+      "epoch": 1.8421052631578947,
+      "grad_norm": 0.3506573628188016,
+      "learning_rate": 3.1860688518892606e-05,
+      "loss": 0.3543,
+      "step": 455
+    },
+    {
+      "epoch": 1.8461538461538463,
+      "grad_norm": 0.3833643639646693,
+      "learning_rate": 3.181510491450254e-05,
+      "loss": 0.3507,
+      "step": 456
+    },
+    {
+      "epoch": 1.8502024291497976,
+      "grad_norm": 0.3805843040560928,
+      "learning_rate": 3.176942683684842e-05,
+      "loss": 0.369,
+      "step": 457
+    },
+    {
+      "epoch": 1.854251012145749,
+      "grad_norm": 0.3424090427926036,
+      "learning_rate": 3.1723654651170934e-05,
+      "loss": 0.3576,
+      "step": 458
+    },
+    {
+      "epoch": 1.8582995951417005,
+      "grad_norm": 0.40473028499629926,
+      "learning_rate": 3.167778872346327e-05,
+      "loss": 0.3554,
+      "step": 459
+    },
+    {
+      "epoch": 1.8623481781376519,
+      "grad_norm": 0.360175430833993,
+      "learning_rate": 3.163182942046815e-05,
+      "loss": 0.3762,
+      "step": 460
+    },
+    {
+      "epoch": 1.8663967611336032,
+      "grad_norm": 0.3960395647336505,
+      "learning_rate": 3.158577710967494e-05,
+      "loss": 0.3595,
+      "step": 461
+    },
+    {
+      "epoch": 1.8704453441295548,
+      "grad_norm": 0.3907242396311842,
+      "learning_rate": 3.153963215931669e-05,
+      "loss": 0.3571,
+      "step": 462
+    },
+    {
+      "epoch": 1.874493927125506,
+      "grad_norm": 0.38212809316132784,
+      "learning_rate": 3.14933949383672e-05,
+      "loss": 0.373,
+      "step": 463
+    },
+    {
+      "epoch": 1.8785425101214575,
+      "grad_norm": 0.36334069683954684,
+      "learning_rate": 3.144706581653805e-05,
+      "loss": 0.3695,
+      "step": 464
+    },
+    {
+      "epoch": 1.882591093117409,
+      "grad_norm": 0.3600803088544394,
+      "learning_rate": 3.140064516427566e-05,
+      "loss": 0.3727,
+      "step": 465
+    },
+    {
+      "epoch": 1.8866396761133604,
+      "grad_norm": 0.3402007048112333,
+      "learning_rate": 3.135413335275833e-05,
+      "loss": 0.3616,
+      "step": 466
+    },
+    {
+      "epoch": 1.8906882591093117,
+      "grad_norm": 0.38021638379896455,
+      "learning_rate": 3.130753075389327e-05,
+      "loss": 0.3679,
+      "step": 467
+    },
+    {
+      "epoch": 1.8947368421052633,
+      "grad_norm": 0.3346323884894299,
+      "learning_rate": 3.1260837740313625e-05,
+      "loss": 0.3459,
+      "step": 468
+    },
+    {
+      "epoch": 1.8987854251012146,
+      "grad_norm": 0.34839291726910576,
+      "learning_rate": 3.121405468537547e-05,
+      "loss": 0.3438,
+      "step": 469
+    },
+    {
+      "epoch": 1.902834008097166,
+      "grad_norm": 0.36791604591811716,
+      "learning_rate": 3.116718196315488e-05,
+      "loss": 0.3546,
+      "step": 470
+    },
+    {
+      "epoch": 1.9068825910931175,
+      "grad_norm": 0.3503899161694769,
+      "learning_rate": 3.1120219948444894e-05,
+      "loss": 0.3559,
+      "step": 471
+    },
+    {
+      "epoch": 1.9109311740890689,
+      "grad_norm": 0.46065220243598215,
+      "learning_rate": 3.107316901675252e-05,
+      "loss": 0.3738,
+      "step": 472
+    },
+    {
+      "epoch": 1.9149797570850202,
+      "grad_norm": 0.37539376399355595,
+      "learning_rate": 3.1026029544295745e-05,
+      "loss": 0.3627,
+      "step": 473
+    },
+    {
+      "epoch": 1.9190283400809718,
+      "grad_norm": 0.3696343120186298,
+      "learning_rate": 3.097880190800057e-05,
+      "loss": 0.3561,
+      "step": 474
+    },
+    {
+      "epoch": 1.9230769230769231,
+      "grad_norm": 0.35623424200810905,
+      "learning_rate": 3.093148648549788e-05,
+      "loss": 0.3588,
+      "step": 475
+    },
+    {
+      "epoch": 1.9271255060728745,
+      "grad_norm": 0.32375529637003886,
+      "learning_rate": 3.088408365512055e-05,
+      "loss": 0.3478,
+      "step": 476
+    },
+    {
+      "epoch": 1.931174089068826,
+      "grad_norm": 0.3375257021023335,
+      "learning_rate": 3.083659379590034e-05,
+      "loss": 0.3564,
+      "step": 477
+    },
+    {
+      "epoch": 1.9352226720647774,
+      "grad_norm": 0.33772615530262934,
+      "learning_rate": 3.0789017287564913e-05,
+      "loss": 0.3435,
+      "step": 478
+    },
+    {
+      "epoch": 1.9392712550607287,
+      "grad_norm": 0.38444511902283435,
+      "learning_rate": 3.074135451053475e-05,
+      "loss": 0.3566,
+      "step": 479
+    },
+    {
+      "epoch": 1.9433198380566803,
+      "grad_norm": 0.3513822143180344,
+      "learning_rate": 3.069360584592015e-05,
+      "loss": 0.3566,
+      "step": 480
+    },
+    {
+      "epoch": 1.9473684210526314,
+      "grad_norm": 0.3892566606777769,
+      "learning_rate": 3.064577167551816e-05,
+      "loss": 0.3562,
+      "step": 481
+    },
+    {
+      "epoch": 1.951417004048583,
+      "grad_norm": 0.36512839793646334,
+      "learning_rate": 3.059785238180953e-05,
+      "loss": 0.3608,
+      "step": 482
+    },
+    {
+      "epoch": 1.9554655870445345,
+      "grad_norm": 0.3660770167666678,
+      "learning_rate": 3.054984834795566e-05,
+      "loss": 0.359,
+      "step": 483
+    },
+    {
+      "epoch": 1.9595141700404857,
+      "grad_norm": 0.3872935414670625,
+      "learning_rate": 3.0501759957795537e-05,
+      "loss": 0.3677,
+      "step": 484
+    },
+    {
+      "epoch": 1.9635627530364372,
+      "grad_norm": 0.33960072999501517,
+      "learning_rate": 3.0453587595842627e-05,
+      "loss": 0.3477,
+      "step": 485
+    },
+    {
+      "epoch": 1.9676113360323888,
+      "grad_norm": 0.40178921111516147,
+      "learning_rate": 3.0405331647281856e-05,
+      "loss": 0.3691,
+      "step": 486
+    },
+    {
+      "epoch": 1.97165991902834,
+      "grad_norm": 0.3296832000521388,
+      "learning_rate": 3.0356992497966508e-05,
+      "loss": 0.3561,
+      "step": 487
+    },
+    {
+      "epoch": 1.9757085020242915,
+      "grad_norm": 0.3407581775071839,
+      "learning_rate": 3.0308570534415137e-05,
+      "loss": 0.3673,
+      "step": 488
+    },
+    {
+      "epoch": 1.979757085020243,
+      "grad_norm": 0.3807842961486008,
+      "learning_rate": 3.0260066143808467e-05,
+      "loss": 0.3515,
+      "step": 489
+    },
+    {
+      "epoch": 1.9838056680161942,
+      "grad_norm": 0.3648732370021928,
+      "learning_rate": 3.021147971398632e-05,
+      "loss": 0.3391,
+      "step": 490
+    },
+    {
+      "epoch": 1.9878542510121457,
+      "grad_norm": 0.3249032075686843,
+      "learning_rate": 3.0162811633444478e-05,
+      "loss": 0.3522,
+      "step": 491
+    },
+    {
+      "epoch": 1.9919028340080973,
+      "grad_norm": 0.3507351749230337,
+      "learning_rate": 3.0114062291331635e-05,
+      "loss": 0.3646,
+      "step": 492
+    },
+    {
+      "epoch": 1.9959514170040484,
+      "grad_norm": 0.3131867705769306,
+      "learning_rate": 3.0065232077446217e-05,
+      "loss": 0.3502,
+      "step": 493
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.33609602189257154,
+      "learning_rate": 3.001632138223332e-05,
+      "loss": 0.3427,
+      "step": 494
+    },
+    {
+      "epoch": 2.0040485829959516,
+      "grad_norm": 0.5027324590043901,
+      "learning_rate": 2.9967330596781565e-05,
+      "loss": 0.2733,
+      "step": 495
+    },
+    {
+      "epoch": 2.0080971659919027,
+      "grad_norm": 0.6141702469822979,
+      "learning_rate": 2.9918260112819952e-05,
+      "loss": 0.2723,
+      "step": 496
+    },
+    {
+      "epoch": 2.0121457489878543,
+      "grad_norm": 0.3684266225191845,
+      "learning_rate": 2.9869110322714778e-05,
+      "loss": 0.2508,
+      "step": 497
+    },
+    {
+      "epoch": 2.016194331983806,
+      "grad_norm": 0.3706902585548604,
+      "learning_rate": 2.9819881619466447e-05,
+      "loss": 0.2631,
+      "step": 498
+    },
+    {
+      "epoch": 2.020242914979757,
+      "grad_norm": 0.3848075843488385,
+      "learning_rate": 2.9770574396706364e-05,
+      "loss": 0.2546,
+      "step": 499
+    },
+    {
+      "epoch": 2.0242914979757085,
+      "grad_norm": 0.3319148753920208,
+      "learning_rate": 2.972118904869376e-05,
+      "loss": 0.2498,
+      "step": 500
+    },
+    {
+      "epoch": 2.02834008097166,
+      "grad_norm": 0.3542141411722854,
+      "learning_rate": 2.9671725970312563e-05,
+      "loss": 0.2506,
+      "step": 501
+    },
+    {
+      "epoch": 2.032388663967611,
+      "grad_norm": 0.34600620243804725,
+      "learning_rate": 2.9622185557068238e-05,
+      "loss": 0.2592,
+      "step": 502
+    },
+    {
+      "epoch": 2.0364372469635628,
+      "grad_norm": 0.37619112045707387,
+      "learning_rate": 2.95725682050846e-05,
+      "loss": 0.2519,
+      "step": 503
+    },
+    {
+      "epoch": 2.0404858299595143,
+      "grad_norm": 0.3258516960350112,
+      "learning_rate": 2.9522874311100684e-05,
+      "loss": 0.2528,
+      "step": 504
+    },
+    {
+      "epoch": 2.0445344129554655,
+      "grad_norm": 0.37121892721893573,
+      "learning_rate": 2.9473104272467526e-05,
+      "loss": 0.2416,
+      "step": 505
+    },
+    {
+      "epoch": 2.048582995951417,
+      "grad_norm": 0.3702526272521282,
+      "learning_rate": 2.942325848714505e-05,
+      "loss": 0.2516,
+      "step": 506
+    },
+    {
+      "epoch": 2.0526315789473686,
+      "grad_norm": 0.35107797881395303,
+      "learning_rate": 2.9373337353698804e-05,
+      "loss": 0.2715,
+      "step": 507
+    },
+    {
+      "epoch": 2.0566801619433197,
+      "grad_norm": 0.4651640944131454,
+      "learning_rate": 2.932334127129686e-05,
+      "loss": 0.2645,
+      "step": 508
+    },
+    {
+      "epoch": 2.0607287449392713,
+      "grad_norm": 0.33175992672706,
+      "learning_rate": 2.9273270639706547e-05,
+      "loss": 0.2609,
+      "step": 509
+    },
+    {
+      "epoch": 2.064777327935223,
+      "grad_norm": 0.35208782494346996,
+      "learning_rate": 2.922312585929131e-05,
+      "loss": 0.2576,
+      "step": 510
+    },
+    {
+      "epoch": 2.068825910931174,
+      "grad_norm": 0.3437962718628704,
+      "learning_rate": 2.9172907331007478e-05,
+      "loss": 0.2401,
+      "step": 511
+    },
+    {
+      "epoch": 2.0728744939271255,
+      "grad_norm": 0.3830731916085603,
+      "learning_rate": 2.9122615456401053e-05,
+      "loss": 0.2572,
+      "step": 512
+    },
+    {
+      "epoch": 2.076923076923077,
+      "grad_norm": 0.32505467838987323,
+      "learning_rate": 2.9072250637604535e-05,
+      "loss": 0.2625,
+      "step": 513
+    },
+    {
+      "epoch": 2.080971659919028,
+      "grad_norm": 0.37248637075477137,
+      "learning_rate": 2.9021813277333667e-05,
+      "loss": 0.26,
+      "step": 514
+    },
+    {
+      "epoch": 2.08502024291498,
+      "grad_norm": 0.3415372262893222,
+      "learning_rate": 2.8971303778884244e-05,
+      "loss": 0.2647,
+      "step": 515
+    },
+    {
+      "epoch": 2.0890688259109313,
+      "grad_norm": 0.35172660671584594,
+      "learning_rate": 2.8920722546128863e-05,
+      "loss": 0.2555,
+      "step": 516
+    },
+    {
+      "epoch": 2.0931174089068825,
+      "grad_norm": 0.37161870255509405,
+      "learning_rate": 2.887006998351372e-05,
+      "loss": 0.2827,
+      "step": 517
+    },
+    {
+      "epoch": 2.097165991902834,
+      "grad_norm": 0.3448015819003686,
+      "learning_rate": 2.881934649605535e-05,
+      "loss": 0.2511,
+      "step": 518
+    },
+    {
+      "epoch": 2.1012145748987856,
+      "grad_norm": 0.32395856766478115,
+      "learning_rate": 2.8768552489337408e-05,
+      "loss": 0.2723,
+      "step": 519
+    },
+    {
+      "epoch": 2.1052631578947367,
+      "grad_norm": 0.34208949555026,
+      "learning_rate": 2.8717688369507425e-05,
+      "loss": 0.2578,
+      "step": 520
+    },
+    {
+      "epoch": 2.1093117408906883,
+      "grad_norm": 0.3405956263902834,
+      "learning_rate": 2.8666754543273535e-05,
+      "loss": 0.2586,
+      "step": 521
+    },
+    {
+      "epoch": 2.11336032388664,
+      "grad_norm": 0.3632039690245269,
+      "learning_rate": 2.8615751417901273e-05,
+      "loss": 0.2618,
+      "step": 522
+    },
+    {
+      "epoch": 2.117408906882591,
+      "grad_norm": 0.348421461178726,
+      "learning_rate": 2.856467940121025e-05,
+      "loss": 0.263,
+      "step": 523
+    },
+    {
+      "epoch": 2.1214574898785425,
+      "grad_norm": 0.34993004177264503,
+      "learning_rate": 2.8513538901570964e-05,
+      "loss": 0.2501,
+      "step": 524
+    },
+    {
+      "epoch": 2.125506072874494,
+      "grad_norm": 0.3448564892631117,
+      "learning_rate": 2.846233032790149e-05,
+      "loss": 0.2585,
+      "step": 525
+    },
+    {
+      "epoch": 2.1295546558704452,
+      "grad_norm": 0.331275851436689,
+      "learning_rate": 2.84110540896642e-05,
+      "loss": 0.2511,
+      "step": 526
+    },
+    {
+      "epoch": 2.133603238866397,
+      "grad_norm": 0.35485358461018823,
+      "learning_rate": 2.8359710596862555e-05,
+      "loss": 0.2785,
+      "step": 527
+    },
+    {
+      "epoch": 2.1376518218623484,
+      "grad_norm": 0.3461683699288589,
+      "learning_rate": 2.8308300260037734e-05,
+      "loss": 0.2631,
+      "step": 528
+    },
+    {
+      "epoch": 2.1417004048582995,
+      "grad_norm": 0.32317185050881897,
+      "learning_rate": 2.8256823490265424e-05,
+      "loss": 0.2616,
+      "step": 529
+    },
+    {
+      "epoch": 2.145748987854251,
+      "grad_norm": 0.33704646254026965,
+      "learning_rate": 2.8205280699152515e-05,
+      "loss": 0.2646,
+      "step": 530
+    },
+    {
+      "epoch": 2.1497975708502026,
+      "grad_norm": 0.3587924023698564,
+      "learning_rate": 2.8153672298833777e-05,
+      "loss": 0.2558,
+      "step": 531
+    },
+    {
+      "epoch": 2.1538461538461537,
+      "grad_norm": 0.34585749476060246,
+      "learning_rate": 2.8101998701968615e-05,
+      "loss": 0.2657,
+      "step": 532
+    },
+    {
+      "epoch": 2.1578947368421053,
+      "grad_norm": 0.3540439047994819,
+      "learning_rate": 2.805026032173773e-05,
+      "loss": 0.2584,
+      "step": 533
+    },
+    {
+      "epoch": 2.161943319838057,
+      "grad_norm": 0.3451876358706892,
+      "learning_rate": 2.7998457571839832e-05,
+      "loss": 0.2622,
+      "step": 534
+    },
+    {
+      "epoch": 2.165991902834008,
+      "grad_norm": 0.35758018055918933,
+      "learning_rate": 2.7946590866488322e-05,
+      "loss": 0.2603,
+      "step": 535
+    },
+    {
+      "epoch": 2.1700404858299596,
+      "grad_norm": 0.29097046409433336,
+      "learning_rate": 2.7894660620408e-05,
+      "loss": 0.2698,
+      "step": 536
+    },
+    {
+      "epoch": 2.174089068825911,
+      "grad_norm": 0.35268404654894747,
+      "learning_rate": 2.784266724883173e-05,
+      "loss": 0.2528,
+      "step": 537
+    },
+    {
+      "epoch": 2.1781376518218623,
+      "grad_norm": 0.32440791519465,
+      "learning_rate": 2.7790611167497128e-05,
+      "loss": 0.2638,
+      "step": 538
+    },
+    {
+      "epoch": 2.182186234817814,
+      "grad_norm": 0.37074771220051594,
+      "learning_rate": 2.7738492792643224e-05,
+      "loss": 0.2731,
+      "step": 539
+    },
+    {
+      "epoch": 2.1862348178137654,
+      "grad_norm": 0.33600627780189113,
+      "learning_rate": 2.768631254100716e-05,
+      "loss": 0.2642,
+      "step": 540
+    },
+    {
+      "epoch": 2.1902834008097165,
+      "grad_norm": 0.32144088320054165,
+      "learning_rate": 2.7634070829820836e-05,
+      "loss": 0.2671,
+      "step": 541
+    },
+    {
+      "epoch": 2.194331983805668,
+      "grad_norm": 0.3251280406164903,
+      "learning_rate": 2.7581768076807592e-05,
+      "loss": 0.2297,
+      "step": 542
+    },
+    {
+      "epoch": 2.1983805668016196,
+      "grad_norm": 0.3695959052879763,
+      "learning_rate": 2.7529404700178832e-05,
+      "loss": 0.2637,
+      "step": 543
+    },
+    {
+      "epoch": 2.2024291497975708,
+      "grad_norm": 0.32544822737409007,
+      "learning_rate": 2.7476981118630735e-05,
+      "loss": 0.277,
+      "step": 544
+    },
+    {
+      "epoch": 2.2064777327935223,
+      "grad_norm": 0.3509087327849691,
+      "learning_rate": 2.742449775134085e-05,
+      "loss": 0.2574,
+      "step": 545
+    },
+    {
+      "epoch": 2.2105263157894735,
+      "grad_norm": 0.33704580925157834,
+      "learning_rate": 2.7371955017964788e-05,
+      "loss": 0.2587,
+      "step": 546
+    },
+    {
+      "epoch": 2.214574898785425,
+      "grad_norm": 0.3417984956333491,
+      "learning_rate": 2.7319353338632845e-05,
+      "loss": 0.2579,
+      "step": 547
+    },
+    {
+      "epoch": 2.2186234817813766,
+      "grad_norm": 0.3331733327451225,
+      "learning_rate": 2.7266693133946645e-05,
+      "loss": 0.2636,
+      "step": 548
+    },
+    {
+      "epoch": 2.2226720647773277,
+      "grad_norm": 0.33797682215859753,
+      "learning_rate": 2.7213974824975785e-05,
+      "loss": 0.2599,
+      "step": 549
+    },
+    {
+      "epoch": 2.2267206477732793,
+      "grad_norm": 0.3378548249053523,
+      "learning_rate": 2.7161198833254464e-05,
+      "loss": 0.2618,
+      "step": 550
+    },
+    {
+      "epoch": 2.230769230769231,
+      "grad_norm": 0.32716990767752696,
+      "learning_rate": 2.7108365580778095e-05,
+      "loss": 0.2623,
+      "step": 551
+    },
+    {
+      "epoch": 2.234817813765182,
+      "grad_norm": 0.3578225169307475,
+      "learning_rate": 2.705547548999996e-05,
+      "loss": 0.264,
+      "step": 552
+    },
+    {
+      "epoch": 2.2388663967611335,
+      "grad_norm": 0.31989619082629517,
+      "learning_rate": 2.7002528983827817e-05,
+      "loss": 0.2603,
+      "step": 553
+    },
+    {
+      "epoch": 2.242914979757085,
+      "grad_norm": 0.3536224940157492,
+      "learning_rate": 2.6949526485620517e-05,
+      "loss": 0.2516,
+      "step": 554
+    },
+    {
+      "epoch": 2.246963562753036,
+      "grad_norm": 0.4469985110423576,
+      "learning_rate": 2.6896468419184623e-05,
+      "loss": 0.2835,
+      "step": 555
+    },
+    {
+      "epoch": 2.251012145748988,
+      "grad_norm": 0.3170830293262112,
+      "learning_rate": 2.6843355208771024e-05,
+      "loss": 0.2521,
+      "step": 556
+    },
+    {
+      "epoch": 2.2550607287449393,
+      "grad_norm": 0.3718178712059012,
+      "learning_rate": 2.6790187279071536e-05,
+      "loss": 0.2598,
+      "step": 557
+    },
+    {
+      "epoch": 2.2591093117408905,
+      "grad_norm": 0.36575255228431314,
+      "learning_rate": 2.67369650552155e-05,
+      "loss": 0.2692,
+      "step": 558
+    },
+    {
+      "epoch": 2.263157894736842,
+      "grad_norm": 0.32883515690199205,
+      "learning_rate": 2.6683688962766407e-05,
+      "loss": 0.2583,
+      "step": 559
+    },
+    {
+      "epoch": 2.2672064777327936,
+      "grad_norm": 0.3510816701922819,
+      "learning_rate": 2.6630359427718468e-05,
+      "loss": 0.2576,
+      "step": 560
+    },
+    {
+      "epoch": 2.2712550607287447,
+      "grad_norm": 0.3320532693377789,
+      "learning_rate": 2.6576976876493225e-05,
+      "loss": 0.2719,
+      "step": 561
+    },
+    {
+      "epoch": 2.2753036437246963,
+      "grad_norm": 0.34215232393738326,
+      "learning_rate": 2.652354173593614e-05,
+      "loss": 0.2597,
+      "step": 562
+    },
+    {
+      "epoch": 2.279352226720648,
+      "grad_norm": 0.32674770326153924,
+      "learning_rate": 2.647005443331317e-05,
+      "loss": 0.2626,
+      "step": 563
+    },
+    {
+      "epoch": 2.283400809716599,
+      "grad_norm": 0.3612546216022809,
+      "learning_rate": 2.6416515396307354e-05,
+      "loss": 0.2876,
+      "step": 564
+    },
+    {
+      "epoch": 2.2874493927125505,
+      "grad_norm": 0.3289410126997975,
+      "learning_rate": 2.636292505301542e-05,
+      "loss": 0.2722,
+      "step": 565
+    },
+    {
+      "epoch": 2.291497975708502,
+      "grad_norm": 0.32618252633330747,
+      "learning_rate": 2.630928383194431e-05,
+      "loss": 0.2581,
+      "step": 566
+    },
+    {
+      "epoch": 2.2955465587044532,
+      "grad_norm": 0.3402063019422622,
+      "learning_rate": 2.625559216200781e-05,
+      "loss": 0.2737,
+      "step": 567
+    },
+    {
+      "epoch": 2.299595141700405,
+      "grad_norm": 0.30361096579635555,
+      "learning_rate": 2.620185047252308e-05,
+      "loss": 0.2459,
+      "step": 568
+    },
+    {
+      "epoch": 2.3036437246963564,
+      "grad_norm": 0.330134372684027,
+      "learning_rate": 2.614805919320723e-05,
+      "loss": 0.2575,
+      "step": 569
+    },
+    {
+      "epoch": 2.3076923076923075,
+      "grad_norm": 0.35021105741945224,
+      "learning_rate": 2.6094218754173895e-05,
+      "loss": 0.2587,
+      "step": 570
+    },
+    {
+      "epoch": 2.311740890688259,
+      "grad_norm": 0.3103508549820915,
+      "learning_rate": 2.60403295859298e-05,
+      "loss": 0.2437,
+      "step": 571
+    },
+    {
+      "epoch": 2.3157894736842106,
+      "grad_norm": 0.350386657711586,
+      "learning_rate": 2.5986392119371297e-05,
+      "loss": 0.2485,
+      "step": 572
+    },
+    {
+      "epoch": 2.3198380566801617,
+      "grad_norm": 0.3166760300073881,
+      "learning_rate": 2.5932406785780928e-05,
+      "loss": 0.2579,
+      "step": 573
+    },
+    {
+      "epoch": 2.3238866396761133,
+      "grad_norm": 0.3830548243677527,
+      "learning_rate": 2.5878374016823987e-05,
+      "loss": 0.2543,
+      "step": 574
+    },
+    {
+      "epoch": 2.327935222672065,
+      "grad_norm": 0.3296438716012947,
+      "learning_rate": 2.5824294244545045e-05,
+      "loss": 0.2675,
+      "step": 575
+    },
+    {
+      "epoch": 2.331983805668016,
+      "grad_norm": 0.3581245261575585,
+      "learning_rate": 2.577016790136453e-05,
+      "loss": 0.2694,
+      "step": 576
+    },
+    {
+      "epoch": 2.3360323886639676,
+      "grad_norm": 0.323731171737227,
+      "learning_rate": 2.5715995420075225e-05,
+      "loss": 0.2692,
+      "step": 577
+    },
+    {
+      "epoch": 2.340080971659919,
+      "grad_norm": 0.36362426003001247,
+      "learning_rate": 2.5661777233838864e-05,
+      "loss": 0.2875,
+      "step": 578
+    },
+    {
+      "epoch": 2.3441295546558703,
+      "grad_norm": 0.35550843825650796,
+      "learning_rate": 2.5607513776182616e-05,
+      "loss": 0.2561,
+      "step": 579
+    },
+    {
+      "epoch": 2.348178137651822,
+      "grad_norm": 0.3716416224388596,
+      "learning_rate": 2.5553205480995633e-05,
+      "loss": 0.2455,
+      "step": 580
+    },
+    {
+      "epoch": 2.3522267206477734,
+      "grad_norm": 0.34829342409720804,
+      "learning_rate": 2.5498852782525604e-05,
+      "loss": 0.2788,
+      "step": 581
+    },
+    {
+      "epoch": 2.3562753036437245,
+      "grad_norm": 0.34177802760702214,
+      "learning_rate": 2.5444456115375252e-05,
+      "loss": 0.2623,
+      "step": 582
+    },
+    {
+      "epoch": 2.360323886639676,
+      "grad_norm": 0.3441377503880866,
+      "learning_rate": 2.539001591449889e-05,
+      "loss": 0.2469,
+      "step": 583
+    },
+    {
+      "epoch": 2.3643724696356276,
+      "grad_norm": 0.3510059906668155,
+      "learning_rate": 2.53355326151989e-05,
+      "loss": 0.24,
+      "step": 584
+    },
+    {
+      "epoch": 2.3684210526315788,
+      "grad_norm": 0.3395653419028503,
+      "learning_rate": 2.5281006653122288e-05,
+      "loss": 0.2551,
+      "step": 585
+    },
+    {
+      "epoch": 2.3724696356275303,
+      "grad_norm": 0.34808536904585075,
+      "learning_rate": 2.5226438464257217e-05,
+      "loss": 0.2706,
+      "step": 586
+    },
+    {
+      "epoch": 2.376518218623482,
+      "grad_norm": 0.3183472718636648,
+      "learning_rate": 2.5171828484929447e-05,
+      "loss": 0.2418,
+      "step": 587
+    },
+    {
+      "epoch": 2.380566801619433,
+      "grad_norm": 0.3290298819612162,
+      "learning_rate": 2.5117177151798934e-05,
+      "loss": 0.2809,
+      "step": 588
+    },
+    {
+      "epoch": 2.3846153846153846,
+      "grad_norm": 0.3215866055203612,
+      "learning_rate": 2.506248490185627e-05,
+      "loss": 0.2528,
+      "step": 589
+    },
+    {
+      "epoch": 2.388663967611336,
+      "grad_norm": 0.39835335968160146,
+      "learning_rate": 2.500775217241923e-05,
+      "loss": 0.2721,
+      "step": 590
+    },
+    {
+      "epoch": 2.3927125506072873,
+      "grad_norm": 0.3461420165394494,
+      "learning_rate": 2.4952979401129266e-05,
+      "loss": 0.2663,
+      "step": 591
+    },
+    {
+      "epoch": 2.396761133603239,
+      "grad_norm": 0.336794881991205,
+      "learning_rate": 2.4898167025948e-05,
+      "loss": 0.2592,
+      "step": 592
+    },
+    {
+      "epoch": 2.4008097165991904,
+      "grad_norm": 0.33341637546142305,
+      "learning_rate": 2.4843315485153703e-05,
+      "loss": 0.2574,
+      "step": 593
+    },
+    {
+      "epoch": 2.4048582995951415,
+      "grad_norm": 0.3414951254563967,
+      "learning_rate": 2.4788425217337856e-05,
+      "loss": 0.2964,
+      "step": 594
+    },
+    {
+      "epoch": 2.408906882591093,
+      "grad_norm": 0.3414879144578541,
+      "learning_rate": 2.4733496661401572e-05,
+      "loss": 0.2577,
+      "step": 595
+    },
+    {
+      "epoch": 2.4129554655870447,
+      "grad_norm": 0.3199787982347756,
+      "learning_rate": 2.4678530256552114e-05,
+      "loss": 0.2608,
+      "step": 596
+    },
+    {
+      "epoch": 2.417004048582996,
+      "grad_norm": 0.3423331031688544,
+      "learning_rate": 2.46235264422994e-05,
+      "loss": 0.273,
+      "step": 597
+    },
+    {
+      "epoch": 2.4210526315789473,
+      "grad_norm": 0.3189246741053864,
+      "learning_rate": 2.4568485658452452e-05,
+      "loss": 0.2792,
+      "step": 598
+    },
+    {
+      "epoch": 2.425101214574899,
+      "grad_norm": 0.3180393526742431,
+      "learning_rate": 2.451340834511592e-05,
+      "loss": 0.2802,
+      "step": 599
+    },
+    {
+      "epoch": 2.42914979757085,
+      "grad_norm": 0.34368112494674996,
+      "learning_rate": 2.4458294942686528e-05,
+      "loss": 0.2534,
+      "step": 600
+    },
+    {
+      "epoch": 2.4331983805668016,
+      "grad_norm": 0.3316184285458699,
+      "learning_rate": 2.440314589184958e-05,
+      "loss": 0.2661,
+      "step": 601
+    },
+    {
+      "epoch": 2.437246963562753,
+      "grad_norm": 0.34449885023447013,
+      "learning_rate": 2.434796163357541e-05,
+      "loss": 0.2907,
+      "step": 602
+    },
+    {
+      "epoch": 2.4412955465587043,
+      "grad_norm": 0.3202907147614065,
+      "learning_rate": 2.429274260911588e-05,
+      "loss": 0.2626,
+      "step": 603
+    },
+    {
+      "epoch": 2.445344129554656,
+      "grad_norm": 0.35059292976891543,
+      "learning_rate": 2.423748926000084e-05,
+      "loss": 0.2509,
+      "step": 604
+    },
+    {
+      "epoch": 2.4493927125506074,
+      "grad_norm": 0.33674165912942744,
+      "learning_rate": 2.418220202803459e-05,
+      "loss": 0.2864,
+      "step": 605
+    },
+    {
+      "epoch": 2.4534412955465585,
+      "grad_norm": 0.3465154859418166,
+      "learning_rate": 2.412688135529237e-05,
+      "loss": 0.2657,
+      "step": 606
+    },
+    {
+      "epoch": 2.45748987854251,
+      "grad_norm": 0.3180989025570217,
+      "learning_rate": 2.4071527684116807e-05,
+      "loss": 0.2495,
+      "step": 607
+    },
+    {
+      "epoch": 2.4615384615384617,
+      "grad_norm": 0.3221893907543127,
+      "learning_rate": 2.4016141457114372e-05,
+      "loss": 0.2534,
+      "step": 608
+    },
+    {
+      "epoch": 2.465587044534413,
+      "grad_norm": 0.29591880972741325,
+      "learning_rate": 2.396072311715187e-05,
+      "loss": 0.2638,
+      "step": 609
+    },
+    {
+      "epoch": 2.4696356275303644,
+      "grad_norm": 0.29609032039801075,
+      "learning_rate": 2.3905273107352863e-05,
+      "loss": 0.2693,
+      "step": 610
+    },
+    {
+      "epoch": 2.473684210526316,
+      "grad_norm": 0.3178457954669636,
+      "learning_rate": 2.3849791871094157e-05,
+      "loss": 0.2756,
+      "step": 611
+    },
+    {
+      "epoch": 2.477732793522267,
+      "grad_norm": 0.3142661731150715,
+      "learning_rate": 2.379427985200223e-05,
+      "loss": 0.2592,
+      "step": 612
+    },
+    {
+      "epoch": 2.4817813765182186,
+      "grad_norm": 0.326934980997563,
+      "learning_rate": 2.373873749394972e-05,
+      "loss": 0.2594,
+      "step": 613
+    },
+    {
+      "epoch": 2.48582995951417,
+      "grad_norm": 0.3089857943104338,
+      "learning_rate": 2.3683165241051836e-05,
+      "loss": 0.2713,
+      "step": 614
+    },
+    {
+      "epoch": 2.4898785425101213,
+      "grad_norm": 0.3138250319588033,
+      "learning_rate": 2.3627563537662823e-05,
+      "loss": 0.2568,
+      "step": 615
+    },
+    {
+      "epoch": 2.493927125506073,
+      "grad_norm": 0.3488912165213109,
+      "learning_rate": 2.3571932828372433e-05,
+      "loss": 0.275,
+      "step": 616
+    },
+    {
+      "epoch": 2.4979757085020244,
+      "grad_norm": 0.34801200657713577,
+      "learning_rate": 2.351627355800233e-05,
+      "loss": 0.2711,
+      "step": 617
+    },
+    {
+      "epoch": 2.5020242914979756,
+      "grad_norm": 0.34639459667094596,
+      "learning_rate": 2.346058617160256e-05,
+      "loss": 0.257,
+      "step": 618
+    },
+    {
+      "epoch": 2.506072874493927,
+      "grad_norm": 0.3182328175703955,
+      "learning_rate": 2.3404871114447976e-05,
+      "loss": 0.2679,
+      "step": 619
+    },
+    {
+      "epoch": 2.5101214574898787,
+      "grad_norm": 0.3142651453856532,
+      "learning_rate": 2.334912883203469e-05,
+      "loss": 0.2574,
+      "step": 620
+    },
+    {
+      "epoch": 2.51417004048583,
+      "grad_norm": 0.32565880516914364,
+      "learning_rate": 2.3293359770076505e-05,
+      "loss": 0.2525,
+      "step": 621
+    },
+    {
+      "epoch": 2.5182186234817814,
+      "grad_norm": 0.35080719008133743,
+      "learning_rate": 2.323756437450136e-05,
+      "loss": 0.2617,
+      "step": 622
+    },
+    {
+      "epoch": 2.522267206477733,
+      "grad_norm": 0.3120672526102726,
+      "learning_rate": 2.3181743091447754e-05,
+      "loss": 0.2631,
+      "step": 623
+    },
+    {
+      "epoch": 2.526315789473684,
+      "grad_norm": 0.32883478352173334,
+      "learning_rate": 2.3125896367261164e-05,
+      "loss": 0.2736,
+      "step": 624
+    },
+    {
+      "epoch": 2.5303643724696356,
+      "grad_norm": 0.2961422792558565,
+      "learning_rate": 2.307002464849052e-05,
+      "loss": 0.2446,
+      "step": 625
+    },
+    {
+      "epoch": 2.534412955465587,
+      "grad_norm": 0.3400966576499076,
+      "learning_rate": 2.30141283818846e-05,
+      "loss": 0.2509,
+      "step": 626
+    },
+    {
+      "epoch": 2.5384615384615383,
+      "grad_norm": 0.3033916929532129,
+      "learning_rate": 2.2958208014388462e-05,
+      "loss": 0.2456,
+      "step": 627
+    },
+    {
+      "epoch": 2.54251012145749,
+      "grad_norm": 0.3373907490668118,
+      "learning_rate": 2.2902263993139857e-05,
+      "loss": 0.2616,
+      "step": 628
+    },
+    {
+      "epoch": 2.5465587044534415,
+      "grad_norm": 0.3084667950327344,
+      "learning_rate": 2.2846296765465708e-05,
+      "loss": 0.2608,
+      "step": 629
+    },
+    {
+      "epoch": 2.5506072874493926,
+      "grad_norm": 0.30293256363210724,
+      "learning_rate": 2.2790306778878465e-05,
+      "loss": 0.274,
+      "step": 630
+    },
+    {
+      "epoch": 2.554655870445344,
+      "grad_norm": 0.31203928302614303,
+      "learning_rate": 2.273429448107257e-05,
+      "loss": 0.2592,
+      "step": 631
+    },
+    {
+      "epoch": 2.5587044534412957,
+      "grad_norm": 0.29621531793772526,
+      "learning_rate": 2.2678260319920868e-05,
+      "loss": 0.2599,
+      "step": 632
+    },
+    {
+      "epoch": 2.562753036437247,
+      "grad_norm": 0.3279478667629014,
+      "learning_rate": 2.262220474347102e-05,
+      "loss": 0.2539,
+      "step": 633
+    },
+    {
+      "epoch": 2.5668016194331984,
+      "grad_norm": 0.32358061430789364,
+      "learning_rate": 2.2566128199941918e-05,
+      "loss": 0.2649,
+      "step": 634
+    },
+    {
+      "epoch": 2.57085020242915,
+      "grad_norm": 0.3252648817198149,
+      "learning_rate": 2.2510031137720104e-05,
+      "loss": 0.2585,
+      "step": 635
+    },
+    {
+      "epoch": 2.574898785425101,
+      "grad_norm": 0.36264117839129306,
+      "learning_rate": 2.2453914005356206e-05,
+      "loss": 0.2741,
+      "step": 636
+    },
+    {
+      "epoch": 2.5789473684210527,
+      "grad_norm": 0.3448091936460968,
+      "learning_rate": 2.2397777251561302e-05,
+      "loss": 0.2608,
+      "step": 637
+    },
+    {
+      "epoch": 2.582995951417004,
+      "grad_norm": 0.34847209717714756,
+      "learning_rate": 2.2341621325203393e-05,
+      "loss": 0.2534,
+      "step": 638
+    },
+    {
+      "epoch": 2.5870445344129553,
+      "grad_norm": 0.36882892242814286,
+      "learning_rate": 2.2285446675303764e-05,
+      "loss": 0.2759,
+      "step": 639
+    },
+    {
+      "epoch": 2.591093117408907,
+      "grad_norm": 0.3061720051204777,
+      "learning_rate": 2.2229253751033423e-05,
+      "loss": 0.2511,
+      "step": 640
+    },
+    {
+      "epoch": 2.5951417004048585,
+      "grad_norm": 0.3450533999422898,
+      "learning_rate": 2.2173043001709492e-05,
+      "loss": 0.2608,
+      "step": 641
+    },
+    {
+      "epoch": 2.5991902834008096,
+      "grad_norm": 0.3564198882953695,
+      "learning_rate": 2.2116814876791627e-05,
+      "loss": 0.2763,
+      "step": 642
+    },
+    {
+      "epoch": 2.603238866396761,
+      "grad_norm": 0.30261800725321125,
+      "learning_rate": 2.2060569825878413e-05,
+      "loss": 0.2789,
+      "step": 643
+    },
+    {
+      "epoch": 2.6072874493927127,
+      "grad_norm": 0.31928826337068383,
+      "learning_rate": 2.2004308298703777e-05,
+      "loss": 0.2704,
+      "step": 644
+    },
+    {
+      "epoch": 2.611336032388664,
+      "grad_norm": 0.3026334909477034,
+      "learning_rate": 2.1948030745133396e-05,
+      "loss": 0.2682,
+      "step": 645
+    },
+    {
+      "epoch": 2.6153846153846154,
+      "grad_norm": 0.30907791746725083,
+      "learning_rate": 2.1891737615161083e-05,
+      "loss": 0.2692,
+      "step": 646
+    },
+    {
+      "epoch": 2.619433198380567,
+      "grad_norm": 0.32409065206571386,
+      "learning_rate": 2.1835429358905218e-05,
+      "loss": 0.2602,
+      "step": 647
+    },
+    {
+      "epoch": 2.623481781376518,
+      "grad_norm": 0.298470128876746,
+      "learning_rate": 2.17791064266051e-05,
+      "loss": 0.2674,
+      "step": 648
+    },
+    {
+      "epoch": 2.6275303643724697,
+      "grad_norm": 0.3476627036245367,
+      "learning_rate": 2.172276926861741e-05,
+      "loss": 0.2609,
+      "step": 649
+    },
+    {
+      "epoch": 2.6315789473684212,
+      "grad_norm": 0.30681945172412506,
+      "learning_rate": 2.166641833541255e-05,
+      "loss": 0.2707,
+      "step": 650
+    },
+    {
+      "epoch": 2.6356275303643724,
+      "grad_norm": 0.33023803271612023,
+      "learning_rate": 2.161005407757109e-05,
+      "loss": 0.245,
+      "step": 651
+    },
+    {
+      "epoch": 2.639676113360324,
+      "grad_norm": 0.3354710934455053,
+      "learning_rate": 2.155367694578013e-05,
+      "loss": 0.2691,
+      "step": 652
+    },
+    {
+      "epoch": 2.6437246963562755,
+      "grad_norm": 0.33875725178401694,
+      "learning_rate": 2.1497287390829712e-05,
+      "loss": 0.2607,
+      "step": 653
+    },
+    {
+      "epoch": 2.6477732793522266,
+      "grad_norm": 0.3287498997589712,
+      "learning_rate": 2.1440885863609208e-05,
+      "loss": 0.2655,
+      "step": 654
+    },
+    {
+      "epoch": 2.651821862348178,
+      "grad_norm": 0.3116732675370383,
+      "learning_rate": 2.1384472815103737e-05,
+      "loss": 0.2743,
+      "step": 655
+    },
+    {
+      "epoch": 2.6558704453441297,
+      "grad_norm": 0.3131797612938409,
+      "learning_rate": 2.1328048696390534e-05,
+      "loss": 0.2631,
+      "step": 656
+    },
+    {
+      "epoch": 2.659919028340081,
+      "grad_norm": 0.3160994812160718,
+      "learning_rate": 2.1271613958635342e-05,
+      "loss": 0.2615,
+      "step": 657
+    },
+    {
+      "epoch": 2.6639676113360324,
+      "grad_norm": 0.3002900528565736,
+      "learning_rate": 2.1215169053088825e-05,
+      "loss": 0.2539,
+      "step": 658
+    },
+    {
+      "epoch": 2.668016194331984,
+      "grad_norm": 0.344092995250021,
+      "learning_rate": 2.1158714431082937e-05,
+      "loss": 0.2538,
+      "step": 659
+    },
+    {
+      "epoch": 2.672064777327935,
+      "grad_norm": 0.31443244872585024,
+      "learning_rate": 2.110225054402733e-05,
+      "loss": 0.2649,
+      "step": 660
+    },
+    {
+      "epoch": 2.6761133603238867,
+      "grad_norm": 0.3507990144908591,
+      "learning_rate": 2.1045777843405747e-05,
+      "loss": 0.259,
+      "step": 661
+    },
+    {
+      "epoch": 2.6801619433198383,
+      "grad_norm": 0.32221832057995,
+      "learning_rate": 2.09892967807724e-05,
+      "loss": 0.2686,
+      "step": 662
+    },
+    {
+      "epoch": 2.6842105263157894,
+      "grad_norm": 0.30917829856042006,
+      "learning_rate": 2.093280780774835e-05,
+      "loss": 0.2613,
+      "step": 663
+    },
+    {
+      "epoch": 2.688259109311741,
+      "grad_norm": 0.35219160347718065,
+      "learning_rate": 2.087631137601793e-05,
+      "loss": 0.2707,
+      "step": 664
+    },
+    {
+      "epoch": 2.6923076923076925,
+      "grad_norm": 0.3239428074259248,
+      "learning_rate": 2.0819807937325104e-05,
+      "loss": 0.2623,
+      "step": 665
+    },
+    {
+      "epoch": 2.6963562753036436,
+      "grad_norm": 0.3352318585550522,
+      "learning_rate": 2.0763297943469847e-05,
+      "loss": 0.2746,
+      "step": 666
+    },
+    {
+      "epoch": 2.700404858299595,
+      "grad_norm": 0.3198295330773628,
+      "learning_rate": 2.070678184630458e-05,
+      "loss": 0.2633,
+      "step": 667
+    },
+    {
+      "epoch": 2.7044534412955468,
+      "grad_norm": 0.3389248820899613,
+      "learning_rate": 2.06502600977305e-05,
+      "loss": 0.2782,
+      "step": 668
+    },
+    {
+      "epoch": 2.708502024291498,
+      "grad_norm": 0.31893710338924636,
+      "learning_rate": 2.0593733149694008e-05,
+      "loss": 0.2575,
+      "step": 669
+    },
+    {
+      "epoch": 2.7125506072874495,
+      "grad_norm": 0.3275523367491945,
+      "learning_rate": 2.0537201454183074e-05,
+      "loss": 0.2926,
+      "step": 670
+    },
+    {
+      "epoch": 2.716599190283401,
+      "grad_norm": 0.29975733569351926,
+      "learning_rate": 2.0480665463223626e-05,
+      "loss": 0.2871,
+      "step": 671
+    },
+    {
+      "epoch": 2.720647773279352,
+      "grad_norm": 0.3347808028313618,
+      "learning_rate": 2.042412562887594e-05,
+      "loss": 0.2575,
+      "step": 672
+    },
+    {
+      "epoch": 2.7246963562753037,
+      "grad_norm": 0.31365545424083474,
+      "learning_rate": 2.0367582403231033e-05,
+      "loss": 0.2638,
+      "step": 673
+    },
+    {
+      "epoch": 2.7287449392712553,
+      "grad_norm": 0.2805385726508402,
+      "learning_rate": 2.0311036238407023e-05,
+      "loss": 0.2739,
+      "step": 674
+    },
+    {
+      "epoch": 2.7327935222672064,
+      "grad_norm": 0.3456977824182666,
+      "learning_rate": 2.0254487586545534e-05,
+      "loss": 0.2572,
+      "step": 675
+    },
+    {
+      "epoch": 2.736842105263158,
+      "grad_norm": 0.3232875762903484,
+      "learning_rate": 2.0197936899808088e-05,
+      "loss": 0.2837,
+      "step": 676
+    },
+    {
+      "epoch": 2.7408906882591095,
+      "grad_norm": 0.31013352901825025,
+      "learning_rate": 2.0141384630372466e-05,
+      "loss": 0.2702,
+      "step": 677
+    },
+    {
+      "epoch": 2.7449392712550607,
+      "grad_norm": 0.351770018852729,
+      "learning_rate": 2.0084831230429106e-05,
+      "loss": 0.2669,
+      "step": 678
+    },
+    {
+      "epoch": 2.748987854251012,
+      "grad_norm": 0.39670931375951074,
+      "learning_rate": 2.0028277152177492e-05,
+      "loss": 0.2727,
+      "step": 679
+    },
+    {
+      "epoch": 2.753036437246964,
+      "grad_norm": 0.33508885005072947,
+      "learning_rate": 1.9971722847822518e-05,
+      "loss": 0.2738,
+      "step": 680
+    },
+    {
+      "epoch": 2.757085020242915,
+      "grad_norm": 0.3157438008267563,
+      "learning_rate": 1.99151687695709e-05,
+      "loss": 0.2435,
+      "step": 681
+    },
+    {
+      "epoch": 2.7611336032388665,
+      "grad_norm": 0.31300873601867013,
+      "learning_rate": 1.9858615369627537e-05,
+      "loss": 0.2574,
+      "step": 682
+    },
+    {
+      "epoch": 2.765182186234818,
+      "grad_norm": 0.31370696474558013,
+      "learning_rate": 1.9802063100191916e-05,
+      "loss": 0.2718,
+      "step": 683
+    },
+    {
+      "epoch": 2.769230769230769,
+      "grad_norm": 0.301874531232632,
+      "learning_rate": 1.974551241345447e-05,
+      "loss": 0.2734,
+      "step": 684
+    },
+    {
+      "epoch": 2.7732793522267207,
+      "grad_norm": 0.30876894731839205,
+      "learning_rate": 1.9688963761592984e-05,
+      "loss": 0.2725,
+      "step": 685
+    },
+    {
+      "epoch": 2.7773279352226723,
+      "grad_norm": 0.29270690536695054,
+      "learning_rate": 1.963241759676897e-05,
+      "loss": 0.2683,
+      "step": 686
+    },
+    {
+      "epoch": 2.7813765182186234,
+      "grad_norm": 0.3176891932130674,
+      "learning_rate": 1.9575874371124066e-05,
+      "loss": 0.2671,
+      "step": 687
+    },
+    {
+      "epoch": 2.785425101214575,
+      "grad_norm": 0.3117928750264649,
+      "learning_rate": 1.951933453677638e-05,
+      "loss": 0.2536,
+      "step": 688
+    },
+    {
+      "epoch": 2.7894736842105265,
+      "grad_norm": 0.28082496214587704,
+      "learning_rate": 1.9462798545816932e-05,
+      "loss": 0.2561,
+      "step": 689
+    },
+    {
+      "epoch": 2.7935222672064777,
+      "grad_norm": 0.31742335143035477,
+      "learning_rate": 1.9406266850305996e-05,
+      "loss": 0.2736,
+      "step": 690
+    },
+    {
+      "epoch": 2.7975708502024292,
+      "grad_norm": 0.3174730771948474,
+      "learning_rate": 1.9349739902269503e-05,
+      "loss": 0.2609,
+      "step": 691
+    },
+    {
+      "epoch": 2.801619433198381,
+      "grad_norm": 0.32914737094181734,
+      "learning_rate": 1.9293218153695425e-05,
+      "loss": 0.259,
+      "step": 692
+    },
+    {
+      "epoch": 2.805668016194332,
+      "grad_norm": 0.3100627704960506,
+      "learning_rate": 1.9236702056530153e-05,
+      "loss": 0.2486,
+      "step": 693
+    },
+    {
+      "epoch": 2.8097165991902835,
+      "grad_norm": 0.32289085105290755,
+      "learning_rate": 1.9180192062674907e-05,
+      "loss": 0.2581,
+      "step": 694
+    },
+    {
+      "epoch": 2.813765182186235,
+      "grad_norm": 0.2957379794186454,
+      "learning_rate": 1.9123688623982076e-05,
+      "loss": 0.2676,
+      "step": 695
+    },
+    {
+      "epoch": 2.817813765182186,
+      "grad_norm": 0.29280069831201894,
+      "learning_rate": 1.906719219225165e-05,
+      "loss": 0.2556,
+      "step": 696
+    },
+    {
+      "epoch": 2.8218623481781377,
+      "grad_norm": 0.30218359162314634,
+      "learning_rate": 1.9010703219227603e-05,
+      "loss": 0.2734,
+      "step": 697
+    },
+    {
+      "epoch": 2.8259109311740893,
+      "grad_norm": 0.296520480443392,
+      "learning_rate": 1.8954222156594256e-05,
+      "loss": 0.2445,
+      "step": 698
+    },
+    {
+      "epoch": 2.8299595141700404,
+      "grad_norm": 0.3133529229954482,
+      "learning_rate": 1.8897749455972673e-05,
+      "loss": 0.2596,
+      "step": 699
+    },
+    {
+      "epoch": 2.834008097165992,
+      "grad_norm": 0.3162363331758247,
+      "learning_rate": 1.884128556891707e-05,
+      "loss": 0.2691,
+      "step": 700
+    },
+    {
+      "epoch": 2.8380566801619436,
+      "grad_norm": 0.2920231356694179,
+      "learning_rate": 1.878483094691119e-05,
+      "loss": 0.2558,
+      "step": 701
+    },
+    {
+      "epoch": 2.8421052631578947,
+      "grad_norm": 0.3085676701556866,
+      "learning_rate": 1.8728386041364664e-05,
+      "loss": 0.2589,
+      "step": 702
+    },
+    {
+      "epoch": 2.8461538461538463,
+      "grad_norm": 0.3137811478001395,
+      "learning_rate": 1.867195130360947e-05,
+      "loss": 0.2604,
+      "step": 703
+    },
+    {
+      "epoch": 2.850202429149798,
+      "grad_norm": 0.29659017764474843,
+      "learning_rate": 1.8615527184896263e-05,
+      "loss": 0.2607,
+      "step": 704
+    },
+    {
+      "epoch": 2.854251012145749,
+      "grad_norm": 0.33117466946852087,
+      "learning_rate": 1.8559114136390795e-05,
+      "loss": 0.2729,
+      "step": 705
+    },
+    {
+      "epoch": 2.8582995951417005,
+      "grad_norm": 0.3099035082225854,
+      "learning_rate": 1.8502712609170298e-05,
+      "loss": 0.2503,
+      "step": 706
+    },
+    {
+      "epoch": 2.862348178137652,
+      "grad_norm": 0.29180719542043165,
+      "learning_rate": 1.8446323054219876e-05,
+      "loss": 0.259,
+      "step": 707
+    },
+    {
+      "epoch": 2.866396761133603,
+      "grad_norm": 0.33957696081607913,
+      "learning_rate": 1.838994592242891e-05,
+      "loss": 0.2589,
+      "step": 708
+    },
+    {
+      "epoch": 2.8704453441295548,
+      "grad_norm": 0.3211038860812065,
+      "learning_rate": 1.8333581664587453e-05,
+      "loss": 0.2545,
+      "step": 709
+    },
+    {
+      "epoch": 2.8744939271255063,
+      "grad_norm": 0.32120423598878395,
+      "learning_rate": 1.8277230731382593e-05,
+      "loss": 0.2611,
+      "step": 710
+    },
+    {
+      "epoch": 2.8785425101214575,
+      "grad_norm": 0.3232955160212297,
+      "learning_rate": 1.82208935733949e-05,
+      "loss": 0.2431,
+      "step": 711
+    },
+    {
+      "epoch": 2.882591093117409,
+      "grad_norm": 0.3423941346565437,
+      "learning_rate": 1.8164570641094793e-05,
+      "loss": 0.2473,
+      "step": 712
+    },
+    {
+      "epoch": 2.8866396761133606,
+      "grad_norm": 0.3427088961205948,
+      "learning_rate": 1.810826238483892e-05,
+      "loss": 0.2626,
+      "step": 713
+    },
+    {
+      "epoch": 2.8906882591093117,
+      "grad_norm": 0.306614017311515,
+      "learning_rate": 1.8051969254866608e-05,
+      "loss": 0.2558,
+      "step": 714
+    },
+    {
+      "epoch": 2.8947368421052633,
+      "grad_norm": 0.3169993576173836,
+      "learning_rate": 1.7995691701296226e-05,
+      "loss": 0.257,
+      "step": 715
+    },
+    {
+      "epoch": 2.898785425101215,
+      "grad_norm": 0.317239895623503,
+      "learning_rate": 1.7939430174121593e-05,
+      "loss": 0.2779,
+      "step": 716
+    },
+    {
+      "epoch": 2.902834008097166,
+      "grad_norm": 0.30418038832102007,
+      "learning_rate": 1.788318512320838e-05,
+      "loss": 0.2731,
+      "step": 717
+    },
+    {
+      "epoch": 2.9068825910931175,
+      "grad_norm": 0.28939149793407537,
+      "learning_rate": 1.782695699829051e-05,
+      "loss": 0.2603,
+      "step": 718
+    },
+    {
+      "epoch": 2.910931174089069,
+      "grad_norm": 0.304158608284694,
+      "learning_rate": 1.7770746248966587e-05,
+      "loss": 0.239,
+      "step": 719
+    },
+    {
+      "epoch": 2.91497975708502,
+      "grad_norm": 0.33374734286742946,
+      "learning_rate": 1.7714553324696243e-05,
+      "loss": 0.2677,
+      "step": 720
+    },
+    {
+      "epoch": 2.919028340080972,
+      "grad_norm": 0.3185533378672342,
+      "learning_rate": 1.7658378674796614e-05,
+      "loss": 0.2598,
+      "step": 721
+    },
+    {
+      "epoch": 2.9230769230769234,
+      "grad_norm": 0.2973792170394329,
+      "learning_rate": 1.76022227484387e-05,
+      "loss": 0.2682,
+      "step": 722
+    },
+    {
+      "epoch": 2.9271255060728745,
+      "grad_norm": 0.3445520430009545,
+      "learning_rate": 1.7546085994643807e-05,
+      "loss": 0.2507,
+      "step": 723
+    },
+    {
+      "epoch": 2.931174089068826,
+      "grad_norm": 0.29465809629275064,
+      "learning_rate": 1.7489968862279902e-05,
+      "loss": 0.2548,
+      "step": 724
+    },
+    {
+      "epoch": 2.9352226720647776,
+      "grad_norm": 0.2960364934023133,
+      "learning_rate": 1.743387180005809e-05,
+      "loss": 0.2797,
+      "step": 725
+    },
+    {
+      "epoch": 2.9392712550607287,
+      "grad_norm": 0.30174314222333093,
+      "learning_rate": 1.737779525652899e-05,
+      "loss": 0.2581,
+      "step": 726
+    },
+    {
+      "epoch": 2.9433198380566803,
+      "grad_norm": 0.2943652117842969,
+      "learning_rate": 1.732173968007914e-05,
+      "loss": 0.2564,
+      "step": 727
+    },
+    {
+      "epoch": 2.9473684210526314,
+      "grad_norm": 0.2839099585587317,
+      "learning_rate": 1.7265705518927437e-05,
+      "loss": 0.2718,
+      "step": 728
+    },
+    {
+      "epoch": 2.951417004048583,
+      "grad_norm": 0.305420139870249,
+      "learning_rate": 1.7209693221121542e-05,
+      "loss": 0.261,
+      "step": 729
+    },
+    {
+      "epoch": 2.9554655870445345,
+      "grad_norm": 0.276503455147939,
+      "learning_rate": 1.7153703234534302e-05,
+      "loss": 0.2854,
+      "step": 730
+    },
+    {
+      "epoch": 2.9595141700404857,
+      "grad_norm": 0.31189574339691467,
+      "learning_rate": 1.709773600686015e-05,
+      "loss": 0.2602,
+      "step": 731
+    },
+    {
+      "epoch": 2.9635627530364372,
+      "grad_norm": 0.304286891313922,
+      "learning_rate": 1.7041791985611548e-05,
+      "loss": 0.2586,
+      "step": 732
+    },
+    {
+      "epoch": 2.967611336032389,
+      "grad_norm": 0.29670838231023267,
+      "learning_rate": 1.69858716181154e-05,
+      "loss": 0.2663,
+      "step": 733
+    },
+    {
+      "epoch": 2.97165991902834,
+      "grad_norm": 0.2849529171263787,
+      "learning_rate": 1.692997535150948e-05,
+      "loss": 0.2747,
+      "step": 734
+    },
+    {
+      "epoch": 2.9757085020242915,
+      "grad_norm": 0.29150884273141303,
+      "learning_rate": 1.687410363273884e-05,
+      "loss": 0.2731,
+      "step": 735
+    },
+    {
+      "epoch": 2.979757085020243,
+      "grad_norm": 0.28482852551914684,
+      "learning_rate": 1.6818256908552257e-05,
+      "loss": 0.2617,
+      "step": 736
+    },
+    {
+      "epoch": 2.983805668016194,
+      "grad_norm": 0.2819792264961017,
+      "learning_rate": 1.6762435625498646e-05,
+      "loss": 0.2843,
+      "step": 737
+    },
+    {
+      "epoch": 2.9878542510121457,
+      "grad_norm": 0.2791130659547564,
+      "learning_rate": 1.67066402299235e-05,
+      "loss": 0.2683,
+      "step": 738
+    },
+    {
+      "epoch": 2.9919028340080973,
+      "grad_norm": 0.28972452061315124,
+      "learning_rate": 1.6650871167965313e-05,
+      "loss": 0.2617,
+      "step": 739
+    },
+    {
+      "epoch": 2.9959514170040484,
+      "grad_norm": 0.2978440194644727,
+      "learning_rate": 1.6595128885552028e-05,
+      "loss": 0.2713,
+      "step": 740
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.29890000269671607,
+      "learning_rate": 1.653941382839745e-05,
+      "loss": 0.25,
+      "step": 741
+    },
+    {
+      "epoch": 3.0040485829959516,
+      "grad_norm": 0.5444838392007604,
+      "learning_rate": 1.6483726441997673e-05,
+      "loss": 0.2122,
+      "step": 742
+    },
+    {
+      "epoch": 3.0080971659919027,
+      "grad_norm": 0.3267989453036134,
+      "learning_rate": 1.642806717162757e-05,
+      "loss": 0.1801,
+      "step": 743
+    },
+    {
+      "epoch": 3.0121457489878543,
+      "grad_norm": 0.5258666099286505,
+      "learning_rate": 1.637243646233718e-05,
+      "loss": 0.1787,
+      "step": 744
+    },
+    {
+      "epoch": 3.016194331983806,
+      "grad_norm": 0.49497265347029895,
+      "learning_rate": 1.6316834758948174e-05,
+      "loss": 0.1743,
+      "step": 745
+    },
+    {
+      "epoch": 3.020242914979757,
+      "grad_norm": 0.39388098283563616,
+      "learning_rate": 1.6261262506050282e-05,
+      "loss": 0.197,
+      "step": 746
+    },
+    {
+      "epoch": 3.0242914979757085,
+      "grad_norm": 0.4722101496283096,
+      "learning_rate": 1.620572014799777e-05,
+      "loss": 0.1739,
+      "step": 747
+    },
+    {
+      "epoch": 3.02834008097166,
+      "grad_norm": 0.4051038385480473,
+      "learning_rate": 1.6150208128905857e-05,
+      "loss": 0.1652,
+      "step": 748
+    },
+    {
+      "epoch": 3.032388663967611,
+      "grad_norm": 0.3547678343770683,
+      "learning_rate": 1.6094726892647147e-05,
+      "loss": 0.1851,
+      "step": 749
+    },
+    {
+      "epoch": 3.0364372469635628,
+      "grad_norm": 0.3856233854786123,
+      "learning_rate": 1.6039276882848135e-05,
+      "loss": 0.201,
+      "step": 750
+    },
+    {
+      "epoch": 3.0404858299595143,
+      "grad_norm": 0.42264107566683734,
+      "learning_rate": 1.5983858542885635e-05,
+      "loss": 0.189,
+      "step": 751
+    },
+    {
+      "epoch": 3.0445344129554655,
+      "grad_norm": 0.329704010852696,
+      "learning_rate": 1.5928472315883203e-05,
+      "loss": 0.1806,
+      "step": 752
+    },
+    {
+      "epoch": 3.048582995951417,
+      "grad_norm": 0.32330355381392867,
+      "learning_rate": 1.5873118644707633e-05,
+      "loss": 0.1784,
+      "step": 753
+    },
+    {
+      "epoch": 3.0526315789473686,
+      "grad_norm": 0.3698325463415311,
+      "learning_rate": 1.5817797971965413e-05,
+      "loss": 0.1739,
+      "step": 754
+    },
+    {
+      "epoch": 3.0566801619433197,
+      "grad_norm": 0.30962515769454635,
+      "learning_rate": 1.576251073999917e-05,
+      "loss": 0.1796,
+      "step": 755
+    },
+    {
+      "epoch": 3.0607287449392713,
+      "grad_norm": 0.32150786954807764,
+      "learning_rate": 1.5707257390884126e-05,
+      "loss": 0.1933,
+      "step": 756
+    },
+    {
+      "epoch": 3.064777327935223,
+      "grad_norm": 0.3483940194393815,
+      "learning_rate": 1.5652038366424595e-05,
+      "loss": 0.171,
+      "step": 757
+    },
+    {
+      "epoch": 3.068825910931174,
+      "grad_norm": 0.3315806201908359,
+      "learning_rate": 1.5596854108150424e-05,
+      "loss": 0.183,
+      "step": 758
+    },
+    {
+      "epoch": 3.0728744939271255,
+      "grad_norm": 0.34818445848066915,
+      "learning_rate": 1.5541705057313476e-05,
+      "loss": 0.1812,
+      "step": 759
+    },
+    {
+      "epoch": 3.076923076923077,
+      "grad_norm": 0.3073946288446525,
+      "learning_rate": 1.5486591654884086e-05,
+      "loss": 0.1986,
+      "step": 760
+    },
+    {
+      "epoch": 3.080971659919028,
+      "grad_norm": 0.3332507418567952,
+      "learning_rate": 1.543151434154755e-05,
+      "loss": 0.1784,
+      "step": 761
+    },
+    {
+      "epoch": 3.08502024291498,
+      "grad_norm": 0.32087269371951227,
+      "learning_rate": 1.5376473557700615e-05,
+      "loss": 0.1772,
+      "step": 762
+    },
+    {
+      "epoch": 3.0890688259109313,
+      "grad_norm": 0.2957207870773861,
+      "learning_rate": 1.5321469743447893e-05,
+      "loss": 0.1854,
+      "step": 763
+    },
+    {
+      "epoch": 3.0931174089068825,
+      "grad_norm": 0.3596204762647272,
+      "learning_rate": 1.5266503338598434e-05,
+      "loss": 0.1659,
+      "step": 764
+    },
+    {
+      "epoch": 3.097165991902834,
+      "grad_norm": 0.31493111558866627,
+      "learning_rate": 1.5211574782662149e-05,
+      "loss": 0.1871,
+      "step": 765
+    },
+    {
+      "epoch": 3.1012145748987856,
+      "grad_norm": 0.3020623255612827,
+      "learning_rate": 1.5156684514846307e-05,
+      "loss": 0.1731,
+      "step": 766
+    },
+    {
+      "epoch": 3.1052631578947367,
+      "grad_norm": 0.41690512132267343,
+      "learning_rate": 1.5101832974052012e-05,
+      "loss": 0.1775,
+      "step": 767
+    },
+    {
+      "epoch": 3.1093117408906883,
+      "grad_norm": 0.3133884729500889,
+      "learning_rate": 1.5047020598870737e-05,
+      "loss": 0.2022,
+      "step": 768
+    },
+    {
+      "epoch": 3.11336032388664,
+      "grad_norm": 0.2872377382543154,
+      "learning_rate": 1.4992247827580778e-05,
+      "loss": 0.1839,
+      "step": 769
+    },
+    {
+      "epoch": 3.117408906882591,
+      "grad_norm": 0.33037852541905705,
+      "learning_rate": 1.493751509814374e-05,
+      "loss": 0.1797,
+      "step": 770
+    },
+    {
+      "epoch": 3.1214574898785425,
+      "grad_norm": 0.32636189136068156,
+      "learning_rate": 1.4882822848201073e-05,
+      "loss": 0.1741,
+      "step": 771
+    },
+    {
+      "epoch": 3.125506072874494,
+      "grad_norm": 0.30128564466439645,
+      "learning_rate": 1.4828171515070553e-05,
+      "loss": 0.1814,
+      "step": 772
+    },
+    {
+      "epoch": 3.1295546558704452,
+      "grad_norm": 0.3047777563278221,
+      "learning_rate": 1.4773561535742793e-05,
+      "loss": 0.1618,
+      "step": 773
+    },
+    {
+      "epoch": 3.133603238866397,
+      "grad_norm": 0.2842766334609874,
+      "learning_rate": 1.4718993346877715e-05,
+      "loss": 0.1763,
+      "step": 774
+    },
+    {
+      "epoch": 3.1376518218623484,
+      "grad_norm": 0.30527664984492614,
+      "learning_rate": 1.466446738480111e-05,
+      "loss": 0.1918,
+      "step": 775
+    },
+    {
+      "epoch": 3.1417004048582995,
+      "grad_norm": 0.32585062438948065,
+      "learning_rate": 1.460998408550112e-05,
+      "loss": 0.1853,
+      "step": 776
+    },
+    {
+      "epoch": 3.145748987854251,
+      "grad_norm": 0.29090335064035977,
+      "learning_rate": 1.4555543884624751e-05,
+      "loss": 0.1788,
+      "step": 777
+    },
+    {
+      "epoch": 3.1497975708502026,
+      "grad_norm": 0.28465123652433905,
+      "learning_rate": 1.4501147217474402e-05,
+      "loss": 0.1827,
+      "step": 778
+    },
+    {
+      "epoch": 3.1538461538461537,
+      "grad_norm": 0.29038254824042997,
+      "learning_rate": 1.444679451900437e-05,
+      "loss": 0.1834,
+      "step": 779
+    },
+    {
+      "epoch": 3.1578947368421053,
+      "grad_norm": 0.305755660551029,
+      "learning_rate": 1.4392486223817397e-05,
+      "loss": 0.2046,
+      "step": 780
+    },
+    {
+      "epoch": 3.161943319838057,
+      "grad_norm": 0.309773434190766,
+      "learning_rate": 1.4338222766161143e-05,
+      "loss": 0.1571,
+      "step": 781
+    },
+    {
+      "epoch": 3.165991902834008,
+      "grad_norm": 0.3101415575219633,
+      "learning_rate": 1.4284004579924777e-05,
+      "loss": 0.1863,
+      "step": 782
+    },
+    {
+      "epoch": 3.1700404858299596,
+      "grad_norm": 0.3048705454847475,
+      "learning_rate": 1.4229832098635479e-05,
+      "loss": 0.1684,
+      "step": 783
+    },
+    {
+      "epoch": 3.174089068825911,
+      "grad_norm": 0.2911622613646154,
+      "learning_rate": 1.4175705755454963e-05,
+      "loss": 0.1849,
+      "step": 784
+    },
+    {
+      "epoch": 3.1781376518218623,
+      "grad_norm": 0.3081912025079141,
+      "learning_rate": 1.412162598317602e-05,
+      "loss": 0.175,
+      "step": 785
+    },
+    {
+      "epoch": 3.182186234817814,
+      "grad_norm": 0.29791910011534917,
+      "learning_rate": 1.406759321421907e-05,
+      "loss": 0.1862,
+      "step": 786
+    },
+    {
+      "epoch": 3.1862348178137654,
+      "grad_norm": 0.3089254899084175,
+      "learning_rate": 1.401360788062871e-05,
+      "loss": 0.1863,
+      "step": 787
+    },
+    {
+      "epoch": 3.1902834008097165,
+      "grad_norm": 0.2872144946976353,
+      "learning_rate": 1.3959670414070204e-05,
+      "loss": 0.1743,
+      "step": 788
+    },
+    {
+      "epoch": 3.194331983805668,
+      "grad_norm": 0.29112004462328916,
+      "learning_rate": 1.3905781245826108e-05,
+      "loss": 0.1699,
+      "step": 789
+    },
+    {
+      "epoch": 3.1983805668016196,
+      "grad_norm": 0.28725938106518845,
+      "learning_rate": 1.3851940806792778e-05,
+      "loss": 0.1851,
+      "step": 790
+    },
+    {
+      "epoch": 3.2024291497975708,
+      "grad_norm": 0.28829563268918673,
+      "learning_rate": 1.379814952747693e-05,
+      "loss": 0.1812,
+      "step": 791
+    },
+    {
+      "epoch": 3.2064777327935223,
+      "grad_norm": 0.29686477160668984,
+      "learning_rate": 1.3744407837992193e-05,
+      "loss": 0.1784,
+      "step": 792
+    },
+    {
+      "epoch": 3.2105263157894735,
+      "grad_norm": 0.2589719283277157,
+      "learning_rate": 1.3690716168055692e-05,
+      "loss": 0.1962,
+      "step": 793
+    },
+    {
+      "epoch": 3.214574898785425,
+      "grad_norm": 0.28490975449503525,
+      "learning_rate": 1.363707494698459e-05,
+      "loss": 0.1866,
+      "step": 794
+    },
+    {
+      "epoch": 3.2186234817813766,
+      "grad_norm": 0.2768462191296331,
+      "learning_rate": 1.358348460369265e-05,
+      "loss": 0.1782,
+      "step": 795
+    },
+    {
+      "epoch": 3.2226720647773277,
+      "grad_norm": 0.28011787291462437,
+      "learning_rate": 1.352994556668684e-05,
+      "loss": 0.1946,
+      "step": 796
+    },
+    {
+      "epoch": 3.2267206477732793,
+      "grad_norm": 0.28977880716080334,
+      "learning_rate": 1.3476458264063865e-05,
+      "loss": 0.1769,
+      "step": 797
+    },
+    {
+      "epoch": 3.230769230769231,
+      "grad_norm": 0.27168458493201775,
+      "learning_rate": 1.342302312350678e-05,
+      "loss": 0.1763,
+      "step": 798
+    },
+    {
+      "epoch": 3.234817813765182,
+      "grad_norm": 0.2916094722886731,
+      "learning_rate": 1.3369640572281537e-05,
+      "loss": 0.1721,
+      "step": 799
+    },
+    {
+      "epoch": 3.2388663967611335,
+      "grad_norm": 0.2767779932849625,
+      "learning_rate": 1.3316311037233596e-05,
+      "loss": 0.1991,
+      "step": 800
+    },
+    {
+      "epoch": 3.242914979757085,
+      "grad_norm": 0.285580488421904,
+      "learning_rate": 1.326303494478451e-05,
+      "loss": 0.1843,
+      "step": 801
+    },
+    {
+      "epoch": 3.246963562753036,
+      "grad_norm": 0.2751302043188186,
+      "learning_rate": 1.3209812720928472e-05,
+      "loss": 0.1789,
+      "step": 802
+    },
+    {
+      "epoch": 3.251012145748988,
+      "grad_norm": 0.28891772073638683,
+      "learning_rate": 1.315664479122898e-05,
+      "loss": 0.1896,
+      "step": 803
+    },
+    {
+      "epoch": 3.2550607287449393,
+      "grad_norm": 0.2793067679746219,
+      "learning_rate": 1.3103531580815378e-05,
+      "loss": 0.2067,
+      "step": 804
+    },
+    {
+      "epoch": 3.2591093117408905,
+      "grad_norm": 0.27693911324941073,
+      "learning_rate": 1.3050473514379488e-05,
+      "loss": 0.1801,
+      "step": 805
+    },
+    {
+      "epoch": 3.263157894736842,
+      "grad_norm": 0.26974231494601675,
+      "learning_rate": 1.2997471016172188e-05,
+      "loss": 0.1715,
+      "step": 806
+    },
+    {
+      "epoch": 3.2672064777327936,
+      "grad_norm": 0.27073016646708425,
+      "learning_rate": 1.2944524510000042e-05,
+      "loss": 0.1725,
+      "step": 807
+    },
+    {
+      "epoch": 3.2712550607287447,
+      "grad_norm": 0.2798983906450217,
+      "learning_rate": 1.2891634419221908e-05,
+      "loss": 0.1904,
+      "step": 808
+    },
+    {
+      "epoch": 3.2753036437246963,
+      "grad_norm": 0.2758385299019962,
+      "learning_rate": 1.2838801166745545e-05,
+      "loss": 0.1777,
+      "step": 809
+    },
+    {
+      "epoch": 3.279352226720648,
+      "grad_norm": 0.2671759783151897,
+      "learning_rate": 1.2786025175024216e-05,
+      "loss": 0.191,
+      "step": 810
+    },
+    {
+      "epoch": 3.283400809716599,
+      "grad_norm": 0.2830666721064141,
+      "learning_rate": 1.2733306866053357e-05,
+      "loss": 0.186,
+      "step": 811
+    },
+    {
+      "epoch": 3.2874493927125505,
+      "grad_norm": 0.2841315733426869,
+      "learning_rate": 1.2680646661367163e-05,
+      "loss": 0.1672,
+      "step": 812
+    },
+    {
+      "epoch": 3.291497975708502,
+      "grad_norm": 0.31439556649029937,
+      "learning_rate": 1.2628044982035219e-05,
+      "loss": 0.187,
+      "step": 813
+    },
+    {
+      "epoch": 3.2955465587044532,
+      "grad_norm": 0.28034813102337974,
+      "learning_rate": 1.2575502248659155e-05,
+      "loss": 0.186,
+      "step": 814
+    },
+    {
+      "epoch": 3.299595141700405,
+      "grad_norm": 0.2807395819009186,
+      "learning_rate": 1.2523018881369269e-05,
+      "loss": 0.1881,
+      "step": 815
+    },
+    {
+      "epoch": 3.3036437246963564,
+      "grad_norm": 0.2880411918614041,
+      "learning_rate": 1.2470595299821176e-05,
+      "loss": 0.1671,
+      "step": 816
+    },
+    {
+      "epoch": 3.3076923076923075,
+      "grad_norm": 0.30894873826064007,
+      "learning_rate": 1.2418231923192416e-05,
+      "loss": 0.1891,
+      "step": 817
+    },
+    {
+      "epoch": 3.311740890688259,
+      "grad_norm": 0.2767604590459352,
+      "learning_rate": 1.2365929170179165e-05,
+      "loss": 0.1853,
+      "step": 818
+    },
+    {
+      "epoch": 3.3157894736842106,
+      "grad_norm": 0.262244838614476,
+      "learning_rate": 1.2313687458992848e-05,
+      "loss": 0.196,
+      "step": 819
+    },
+    {
+      "epoch": 3.3198380566801617,
+      "grad_norm": 0.29121076221438225,
+      "learning_rate": 1.2261507207356785e-05,
+      "loss": 0.1653,
+      "step": 820
+    },
+    {
+      "epoch": 3.3238866396761133,
+      "grad_norm": 0.2811599652854115,
+      "learning_rate": 1.220938883250288e-05,
+      "loss": 0.1817,
+      "step": 821
+    },
+    {
+      "epoch": 3.327935222672065,
+      "grad_norm": 0.26451063514973144,
+      "learning_rate": 1.215733275116827e-05,
+      "loss": 0.1844,
+      "step": 822
+    },
+    {
+      "epoch": 3.331983805668016,
+      "grad_norm": 0.28211416875561807,
+      "learning_rate": 1.2105339379592005e-05,
+      "loss": 0.1955,
+      "step": 823
+    },
+    {
+      "epoch": 3.3360323886639676,
+      "grad_norm": 0.25840748031555133,
+      "learning_rate": 1.2053409133511681e-05,
+      "loss": 0.1877,
+      "step": 824
+    },
+    {
+      "epoch": 3.340080971659919,
+      "grad_norm": 1.016666717349693,
+      "learning_rate": 1.2001542428160173e-05,
+      "loss": 0.1961,
+      "step": 825
+    },
+    {
+      "epoch": 3.3441295546558703,
+      "grad_norm": 0.28045897217439364,
+      "learning_rate": 1.1949739678262278e-05,
+      "loss": 0.1825,
+      "step": 826
+    },
+    {
+      "epoch": 3.348178137651822,
+      "grad_norm": 0.2849812574258175,
+      "learning_rate": 1.189800129803139e-05,
+      "loss": 0.1642,
+      "step": 827
+    },
+    {
+      "epoch": 3.3522267206477734,
+      "grad_norm": 0.26989290239367847,
+      "learning_rate": 1.1846327701166228e-05,
+      "loss": 0.186,
+      "step": 828
+    },
+    {
+      "epoch": 3.3562753036437245,
+      "grad_norm": 0.25745431050081463,
+      "learning_rate": 1.1794719300847493e-05,
+      "loss": 0.1795,
+      "step": 829
+    },
+    {
+      "epoch": 3.360323886639676,
+      "grad_norm": 0.2710693675829293,
+      "learning_rate": 1.1743176509734582e-05,
+      "loss": 0.1897,
+      "step": 830
+    },
+    {
+      "epoch": 3.3643724696356276,
+      "grad_norm": 0.5620599681390822,
+      "learning_rate": 1.1691699739962275e-05,
+      "loss": 0.187,
+      "step": 831
+    },
+    {
+      "epoch": 3.3684210526315788,
+      "grad_norm": 0.28139918748710957,
+      "learning_rate": 1.1640289403137456e-05,
+      "loss": 0.1774,
+      "step": 832
+    },
+    {
+      "epoch": 3.3724696356275303,
+      "grad_norm": 0.2635477284901566,
+      "learning_rate": 1.1588945910335802e-05,
+      "loss": 0.1731,
+      "step": 833
+    },
+    {
+      "epoch": 3.376518218623482,
+      "grad_norm": 0.2692543348942793,
+      "learning_rate": 1.1537669672098517e-05,
+      "loss": 0.1983,
+      "step": 834
+    },
+    {
+      "epoch": 3.380566801619433,
+      "grad_norm": 0.27429695261972276,
+      "learning_rate": 1.1486461098429036e-05,
+      "loss": 0.1913,
+      "step": 835
+    },
+    {
+      "epoch": 3.3846153846153846,
+      "grad_norm": 0.2711941003484855,
+      "learning_rate": 1.1435320598789747e-05,
+      "loss": 0.1921,
+      "step": 836
+    },
+    {
+      "epoch": 3.388663967611336,
+      "grad_norm": 0.2679584427067952,
+      "learning_rate": 1.1384248582098736e-05,
+      "loss": 0.1727,
+      "step": 837
+    },
+    {
+      "epoch": 3.3927125506072873,
+      "grad_norm": 0.2927695708579257,
+      "learning_rate": 1.1333245456726467e-05,
+      "loss": 0.1989,
+      "step": 838
+    },
+    {
+      "epoch": 3.396761133603239,
+      "grad_norm": 0.28201518954618554,
+      "learning_rate": 1.128231163049258e-05,
+      "loss": 0.1766,
+      "step": 839
+    },
+    {
+      "epoch": 3.4008097165991904,
+      "grad_norm": 0.2916625138352681,
+      "learning_rate": 1.123144751066259e-05,
+      "loss": 0.198,
+      "step": 840
+    },
+    {
+      "epoch": 3.4048582995951415,
+      "grad_norm": 0.2605848138628549,
+      "learning_rate": 1.1180653503944657e-05,
+      "loss": 0.1922,
+      "step": 841
+    },
+    {
+      "epoch": 3.408906882591093,
+      "grad_norm": 0.2863394922716412,
+      "learning_rate": 1.1129930016486287e-05,
+      "loss": 0.1925,
+      "step": 842
+    },
+    {
+      "epoch": 3.4129554655870447,
+      "grad_norm": 0.27236150857115,
+      "learning_rate": 1.107927745387114e-05,
+      "loss": 0.196,
+      "step": 843
+    },
+    {
+      "epoch": 3.417004048582996,
+      "grad_norm": 0.28052755892478726,
+      "learning_rate": 1.1028696221115766e-05,
+      "loss": 0.1835,
+      "step": 844
+    },
+    {
+      "epoch": 3.4210526315789473,
+      "grad_norm": 0.269549232530123,
+      "learning_rate": 1.0978186722666341e-05,
+      "loss": 0.1811,
+      "step": 845
+    },
+    {
+      "epoch": 3.425101214574899,
+      "grad_norm": 0.26801439805143173,
+      "learning_rate": 1.0927749362395475e-05,
+      "loss": 0.1691,
+      "step": 846
+    },
+    {
+      "epoch": 3.42914979757085,
+      "grad_norm": 0.2814093325284829,
+      "learning_rate": 1.0877384543598954e-05,
+      "loss": 0.1839,
+      "step": 847
+    },
+    {
+      "epoch": 3.4331983805668016,
+      "grad_norm": 0.2796227989985453,
+      "learning_rate": 1.0827092668992529e-05,
+      "loss": 0.1865,
+      "step": 848
+    },
+    {
+      "epoch": 3.437246963562753,
+      "grad_norm": 0.2814777774392291,
+      "learning_rate": 1.077687414070869e-05,
+      "loss": 0.1751,
+      "step": 849
+    },
+    {
+      "epoch": 3.4412955465587043,
+      "grad_norm": 0.27230536176701914,
+      "learning_rate": 1.0726729360293451e-05,
+      "loss": 0.1863,
+      "step": 850
+    },
+    {
+      "epoch": 3.445344129554656,
+      "grad_norm": 0.27340048765219593,
+      "learning_rate": 1.0676658728703147e-05,
+      "loss": 0.1854,
+      "step": 851
+    },
+    {
+      "epoch": 3.4493927125506074,
+      "grad_norm": 0.28761954009468294,
+      "learning_rate": 1.0626662646301198e-05,
+      "loss": 0.1845,
+      "step": 852
+    },
+    {
+      "epoch": 3.4534412955465585,
+      "grad_norm": 0.28725341297093787,
+      "learning_rate": 1.0576741512854958e-05,
+      "loss": 0.1593,
+      "step": 853
+    },
+    {
+      "epoch": 3.45748987854251,
+      "grad_norm": 0.27622644936168644,
+      "learning_rate": 1.052689572753247e-05,
+      "loss": 0.1994,
+      "step": 854
+    },
+    {
+      "epoch": 3.4615384615384617,
+      "grad_norm": 0.28716250141772515,
+      "learning_rate": 1.0477125688899328e-05,
+      "loss": 0.1729,
+      "step": 855
+    },
+    {
+      "epoch": 3.465587044534413,
+      "grad_norm": 0.2762898445031864,
+      "learning_rate": 1.0427431794915406e-05,
+      "loss": 0.1851,
+      "step": 856
+    },
+    {
+      "epoch": 3.4696356275303644,
+      "grad_norm": 0.2781208968386213,
+      "learning_rate": 1.0377814442931769e-05,
+      "loss": 0.1716,
+      "step": 857
+    },
+    {
+      "epoch": 3.473684210526316,
+      "grad_norm": 0.2772546055084199,
+      "learning_rate": 1.0328274029687439e-05,
+      "loss": 0.1857,
+      "step": 858
+    },
+    {
+      "epoch": 3.477732793522267,
+      "grad_norm": 0.28281234731446325,
+      "learning_rate": 1.0278810951306253e-05,
+      "loss": 0.1918,
+      "step": 859
+    },
+    {
+      "epoch": 3.4817813765182186,
+      "grad_norm": 0.2707716691901727,
+      "learning_rate": 1.0229425603293647e-05,
+      "loss": 0.1893,
+      "step": 860
+    },
+    {
+      "epoch": 3.48582995951417,
+      "grad_norm": 0.27701773196238194,
+      "learning_rate": 1.0180118380533558e-05,
+      "loss": 0.1937,
+      "step": 861
+    },
+    {
+      "epoch": 3.4898785425101213,
+      "grad_norm": 0.2694938017555824,
+      "learning_rate": 1.0130889677285225e-05,
+      "loss": 0.19,
+      "step": 862
+    },
+    {
+      "epoch": 3.493927125506073,
+      "grad_norm": 0.2601322900232295,
+      "learning_rate": 1.008173988718005e-05,
+      "loss": 0.1824,
+      "step": 863
+    },
+    {
+      "epoch": 3.4979757085020244,
+      "grad_norm": 0.267168737877413,
+      "learning_rate": 1.0032669403218439e-05,
+      "loss": 0.1795,
+      "step": 864
+    },
+    {
+      "epoch": 3.5020242914979756,
+      "grad_norm": 0.2770857836645061,
+      "learning_rate": 9.983678617766677e-06,
+      "loss": 0.1863,
+      "step": 865
+    },
+    {
+      "epoch": 3.506072874493927,
+      "grad_norm": 0.27238934209500565,
+      "learning_rate": 9.934767922553788e-06,
+      "loss": 0.1876,
+      "step": 866
+    },
+    {
+      "epoch": 3.5101214574898787,
+      "grad_norm": 0.26231059942573937,
+      "learning_rate": 9.885937708668371e-06,
+      "loss": 0.1868,
+      "step": 867
+    },
+    {
+      "epoch": 3.51417004048583,
+      "grad_norm": 0.27794446878068296,
+      "learning_rate": 9.837188366555524e-06,
+      "loss": 0.1892,
+      "step": 868
+    },
+    {
+      "epoch": 3.5182186234817814,
+      "grad_norm": 0.2771948214989928,
+      "learning_rate": 9.788520286013691e-06,
+      "loss": 0.1649,
+      "step": 869
+    },
+    {
+      "epoch": 3.522267206477733,
+      "grad_norm": 0.27002204873289926,
+      "learning_rate": 9.739933856191537e-06,
+      "loss": 0.1843,
+      "step": 870
+    },
+    {
+      "epoch": 3.526315789473684,
+      "grad_norm": 0.2717443868683625,
+      "learning_rate": 9.691429465584866e-06,
+      "loss": 0.178,
+      "step": 871
+    },
+    {
+      "epoch": 3.5303643724696356,
+      "grad_norm": 0.2808923587621025,
+      "learning_rate": 9.643007502033492e-06,
+      "loss": 0.1732,
+      "step": 872
+    },
+    {
+      "epoch": 3.534412955465587,
+      "grad_norm": 0.27549994089898744,
+      "learning_rate": 9.594668352718152e-06,
+      "loss": 0.1759,
+      "step": 873
+    },
+    {
+      "epoch": 3.5384615384615383,
+      "grad_norm": 0.33183046976299696,
+      "learning_rate": 9.546412404157383e-06,
+      "loss": 0.182,
+      "step": 874
+    },
+    {
+      "epoch": 3.54251012145749,
+      "grad_norm": 0.28002077903364897,
+      "learning_rate": 9.498240042204471e-06,
+      "loss": 0.1654,
+      "step": 875
+    },
+    {
+      "epoch": 3.5465587044534415,
+      "grad_norm": 0.27712588959101564,
+      "learning_rate": 9.450151652044339e-06,
+      "loss": 0.1736,
+      "step": 876
+    },
+    {
+      "epoch": 3.5506072874493926,
+      "grad_norm": 0.26503910895879546,
+      "learning_rate": 9.402147618190472e-06,
+      "loss": 0.1852,
+      "step": 877
+    },
+    {
+      "epoch": 3.554655870445344,
+      "grad_norm": 0.2758230558414406,
+      "learning_rate": 9.354228324481841e-06,
+      "loss": 0.1788,
+      "step": 878
+    },
+    {
+      "epoch": 3.5587044534412957,
+      "grad_norm": 0.273161854687053,
+      "learning_rate": 9.306394154079852e-06,
+      "loss": 0.1732,
+      "step": 879
+    },
+    {
+      "epoch": 3.562753036437247,
+      "grad_norm": 0.2686725713289654,
+      "learning_rate": 9.258645489465254e-06,
+      "loss": 0.1909,
+      "step": 880
+    },
+    {
+      "epoch": 3.5668016194331984,
+      "grad_norm": 0.27173214986347616,
+      "learning_rate": 9.21098271243509e-06,
+      "loss": 0.1876,
+      "step": 881
+    },
+    {
+      "epoch": 3.57085020242915,
+      "grad_norm": 0.2564343660860784,
+      "learning_rate": 9.163406204099659e-06,
+      "loss": 0.1785,
+      "step": 882
+    },
+    {
+      "epoch": 3.574898785425101,
+      "grad_norm": 0.2715785596760097,
+      "learning_rate": 9.115916344879452e-06,
+      "loss": 0.1606,
+      "step": 883
+    },
+    {
+      "epoch": 3.5789473684210527,
+      "grad_norm": 0.26244695852326655,
+      "learning_rate": 9.068513514502126e-06,
+      "loss": 0.1799,
+      "step": 884
+    },
+    {
+      "epoch": 3.582995951417004,
+      "grad_norm": 0.2619616962168093,
+      "learning_rate": 9.021198091999439e-06,
+      "loss": 0.178,
+      "step": 885
+    },
+    {
+      "epoch": 3.5870445344129553,
+      "grad_norm": 0.2669015700312456,
+      "learning_rate": 8.973970455704251e-06,
+      "loss": 0.176,
+      "step": 886
+    },
+    {
+      "epoch": 3.591093117408907,
+      "grad_norm": 0.2612165255366564,
+      "learning_rate": 8.926830983247495e-06,
+      "loss": 0.1857,
+      "step": 887
+    },
+    {
+      "epoch": 3.5951417004048585,
+      "grad_norm": 0.2663758497585576,
+      "learning_rate": 8.879780051555116e-06,
+      "loss": 0.1603,
+      "step": 888
+    },
+    {
+      "epoch": 3.5991902834008096,
+      "grad_norm": 0.25347255483822395,
+      "learning_rate": 8.83281803684512e-06,
+      "loss": 0.2038,
+      "step": 889
+    },
+    {
+      "epoch": 3.603238866396761,
+      "grad_norm": 0.2698245313282202,
+      "learning_rate": 8.785945314624532e-06,
+      "loss": 0.1788,
+      "step": 890
+    },
+    {
+      "epoch": 3.6072874493927127,
+      "grad_norm": 0.26923681313334774,
+      "learning_rate": 8.739162259686385e-06,
+      "loss": 0.1772,
+      "step": 891
+    },
+    {
+      "epoch": 3.611336032388664,
+      "grad_norm": 0.2699340426309331,
+      "learning_rate": 8.692469246106734e-06,
+      "loss": 0.1952,
+      "step": 892
+    },
+    {
+      "epoch": 3.6153846153846154,
+      "grad_norm": 0.2556269648115576,
+      "learning_rate": 8.645866647241677e-06,
+      "loss": 0.1808,
+      "step": 893
+    },
+    {
+      "epoch": 3.619433198380567,
+      "grad_norm": 0.2719882989434687,
+      "learning_rate": 8.599354835724348e-06,
+      "loss": 0.176,
+      "step": 894
+    },
+    {
+      "epoch": 3.623481781376518,
+      "grad_norm": 0.2658438977407266,
+      "learning_rate": 8.552934183461958e-06,
+      "loss": 0.1568,
+      "step": 895
+    },
+    {
+      "epoch": 3.6275303643724697,
+      "grad_norm": 0.2734801808155281,
+      "learning_rate": 8.506605061632805e-06,
+      "loss": 0.1777,
+      "step": 896
+    },
+    {
+      "epoch": 3.6315789473684212,
+      "grad_norm": 0.2815370919095367,
+      "learning_rate": 8.460367840683308e-06,
+      "loss": 0.1836,
+      "step": 897
+    },
+    {
+      "epoch": 3.6356275303643724,
+      "grad_norm": 0.2705160969502997,
+      "learning_rate": 8.414222890325065e-06,
+      "loss": 0.1825,
+      "step": 898
+    },
+    {
+      "epoch": 3.639676113360324,
+      "grad_norm": 0.26304137908515485,
+      "learning_rate": 8.368170579531855e-06,
+      "loss": 0.1833,
+      "step": 899
+    },
+    {
+      "epoch": 3.6437246963562755,
+      "grad_norm": 0.27555026528073834,
+      "learning_rate": 8.322211276536733e-06,
+      "loss": 0.1707,
+      "step": 900
+    },
+    {
+      "epoch": 3.6477732793522266,
+      "grad_norm": 0.2766399290836623,
+      "learning_rate": 8.27634534882907e-06,
+      "loss": 0.1779,
+      "step": 901
+    },
+    {
+      "epoch": 3.651821862348178,
+      "grad_norm": 0.2708048223518182,
+      "learning_rate": 8.230573163151585e-06,
+      "loss": 0.1955,
+      "step": 902
+    },
+    {
+      "epoch": 3.6558704453441297,
+      "grad_norm": 0.26179040504252754,
+      "learning_rate": 8.184895085497466e-06,
+      "loss": 0.19,
+      "step": 903
+    },
+    {
+      "epoch": 3.659919028340081,
+      "grad_norm": 0.26146199551840743,
+      "learning_rate": 8.139311481107401e-06,
+      "loss": 0.1755,
+      "step": 904
+    },
+    {
+      "epoch": 3.6639676113360324,
+      "grad_norm": 0.2596925407959086,
+      "learning_rate": 8.093822714466686e-06,
+      "loss": 0.167,
+      "step": 905
+    },
+    {
+      "epoch": 3.668016194331984,
+      "grad_norm": 0.2580536085224053,
+      "learning_rate": 8.048429149302271e-06,
+      "loss": 0.1706,
+      "step": 906
+    },
+    {
+      "epoch": 3.672064777327935,
+      "grad_norm": 0.2587642655705478,
+      "learning_rate": 8.003131148579902e-06,
+      "loss": 0.1925,
+      "step": 907
+    },
+    {
+      "epoch": 3.6761133603238867,
+      "grad_norm": 0.25969859272587564,
+      "learning_rate": 7.957929074501182e-06,
+      "loss": 0.194,
+      "step": 908
+    },
+    {
+      "epoch": 3.6801619433198383,
+      "grad_norm": 0.260908369154634,
+      "learning_rate": 7.912823288500695e-06,
+      "loss": 0.1794,
+      "step": 909
+    },
+    {
+      "epoch": 3.6842105263157894,
+      "grad_norm": 0.2632203306665695,
+      "learning_rate": 7.867814151243102e-06,
+      "loss": 0.1896,
+      "step": 910
+    },
+    {
+      "epoch": 3.688259109311741,
+      "grad_norm": 0.2583422143480498,
+      "learning_rate": 7.822902022620267e-06,
+      "loss": 0.1641,
+      "step": 911
+    },
+    {
+      "epoch": 3.6923076923076925,
+      "grad_norm": 0.2560045314433935,
+      "learning_rate": 7.778087261748377e-06,
+      "loss": 0.1779,
+      "step": 912
+    },
+    {
+      "epoch": 3.6963562753036436,
+      "grad_norm": 0.25212445117527693,
+      "learning_rate": 7.733370226965056e-06,
+      "loss": 0.1824,
+      "step": 913
+    },
+    {
+      "epoch": 3.700404858299595,
+      "grad_norm": 0.27125471938004825,
+      "learning_rate": 7.688751275826527e-06,
+      "loss": 0.1767,
+      "step": 914
+    },
+    {
+      "epoch": 3.7044534412955468,
+      "grad_norm": 0.2511420348497763,
+      "learning_rate": 7.644230765104735e-06,
+      "loss": 0.1882,
+      "step": 915
+    },
+    {
+      "epoch": 3.708502024291498,
+      "grad_norm": 0.254414393125376,
+      "learning_rate": 7.599809050784502e-06,
+      "loss": 0.1852,
+      "step": 916
+    },
+    {
+      "epoch": 3.7125506072874495,
+      "grad_norm": 0.25947271064284366,
+      "learning_rate": 7.555486488060661e-06,
+      "loss": 0.1766,
+      "step": 917
+    },
+    {
+      "epoch": 3.716599190283401,
+      "grad_norm": 0.2677206855927295,
+      "learning_rate": 7.511263431335249e-06,
+      "loss": 0.1805,
+      "step": 918
+    },
+    {
+      "epoch": 3.720647773279352,
+      "grad_norm": 0.24717144007995553,
+      "learning_rate": 7.467140234214656e-06,
+      "loss": 0.1888,
+      "step": 919
+    },
+    {
+      "epoch": 3.7246963562753037,
+      "grad_norm": 0.26432358133684136,
+      "learning_rate": 7.423117249506777e-06,
+      "loss": 0.1793,
+      "step": 920
+    },
+    {
+      "epoch": 3.7287449392712553,
+      "grad_norm": 0.27273941031421234,
+      "learning_rate": 7.379194829218231e-06,
+      "loss": 0.1707,
+      "step": 921
+    },
+    {
+      "epoch": 3.7327935222672064,
+      "grad_norm": 0.27275161573439066,
+      "learning_rate": 7.335373324551518e-06,
+      "loss": 0.1604,
+      "step": 922
+    },
+    {
+      "epoch": 3.736842105263158,
+      "grad_norm": 0.26932205799409004,
+      "learning_rate": 7.291653085902224e-06,
+      "loss": 0.1777,
+      "step": 923
+    },
+    {
+      "epoch": 3.7408906882591095,
+      "grad_norm": 0.2644898525232193,
+      "learning_rate": 7.248034462856213e-06,
+      "loss": 0.175,
+      "step": 924
+    },
+    {
+      "epoch": 3.7449392712550607,
+      "grad_norm": 0.25139766308249234,
+      "learning_rate": 7.20451780418683e-06,
+      "loss": 0.1811,
+      "step": 925
+    },
+    {
+      "epoch": 3.748987854251012,
+      "grad_norm": 0.2579205921655789,
+      "learning_rate": 7.161103457852128e-06,
+      "loss": 0.1787,
+      "step": 926
+    },
+    {
+      "epoch": 3.753036437246964,
+      "grad_norm": 0.27189350771243637,
+      "learning_rate": 7.117791770992046e-06,
+      "loss": 0.1904,
+      "step": 927
+    },
+    {
+      "epoch": 3.757085020242915,
+      "grad_norm": 0.2595253490583985,
+      "learning_rate": 7.074583089925688e-06,
+      "loss": 0.1789,
+      "step": 928
+    },
+    {
+      "epoch": 3.7611336032388665,
+      "grad_norm": 0.26411541591686283,
+      "learning_rate": 7.0314777601485155e-06,
+      "loss": 0.1898,
+      "step": 929
+    },
+    {
+      "epoch": 3.765182186234818,
+      "grad_norm": 0.2903818240719531,
+      "learning_rate": 6.988476126329604e-06,
+      "loss": 0.164,
+      "step": 930
+    },
+    {
+      "epoch": 3.769230769230769,
+      "grad_norm": 0.2502495531500412,
+      "learning_rate": 6.945578532308857e-06,
+      "loss": 0.1744,
+      "step": 931
+    },
+    {
+      "epoch": 3.7732793522267207,
+      "grad_norm": 0.2689614421132996,
+      "learning_rate": 6.902785321094301e-06,
+      "loss": 0.1778,
+      "step": 932
+    },
+    {
+      "epoch": 3.7773279352226723,
+      "grad_norm": 0.27532304669263297,
+      "learning_rate": 6.8600968348593115e-06,
+      "loss": 0.1614,
+      "step": 933
+    },
+    {
+      "epoch": 3.7813765182186234,
+      "grad_norm": 0.2686258889715986,
+      "learning_rate": 6.817513414939892e-06,
+      "loss": 0.1837,
+      "step": 934
+    },
+    {
+      "epoch": 3.785425101214575,
+      "grad_norm": 0.27056769673156733,
+      "learning_rate": 6.775035401831924e-06,
+      "loss": 0.1849,
+      "step": 935
+    },
+    {
+      "epoch": 3.7894736842105265,
+      "grad_norm": 0.26672307944352214,
+      "learning_rate": 6.732663135188473e-06,
+      "loss": 0.1749,
+      "step": 936
+    },
+    {
+      "epoch": 3.7935222672064777,
+      "grad_norm": 0.2861759867061365,
+      "learning_rate": 6.690396953817055e-06,
+      "loss": 0.1889,
+      "step": 937
+    },
+    {
+      "epoch": 3.7975708502024292,
+      "grad_norm": 0.27296424051874063,
+      "learning_rate": 6.648237195676927e-06,
+      "loss": 0.1861,
+      "step": 938
+    },
+    {
+      "epoch": 3.801619433198381,
+      "grad_norm": 0.2543551677885275,
+      "learning_rate": 6.606184197876395e-06,
+      "loss": 0.1697,
+      "step": 939
+    },
+    {
+      "epoch": 3.805668016194332,
+      "grad_norm": 0.2540209833028275,
+      "learning_rate": 6.564238296670109e-06,
+      "loss": 0.1842,
+      "step": 940
+    },
+    {
+      "epoch": 3.8097165991902835,
+      "grad_norm": 0.2601821219388614,
+      "learning_rate": 6.52239982745638e-06,
+      "loss": 0.1737,
+      "step": 941
+    },
+    {
+      "epoch": 3.813765182186235,
+      "grad_norm": 0.25309309757961873,
+      "learning_rate": 6.480669124774484e-06,
+      "loss": 0.1819,
+      "step": 942
+    },
+    {
+      "epoch": 3.817813765182186,
+      "grad_norm": 0.2677062946524316,
+      "learning_rate": 6.439046522302013e-06,
+      "loss": 0.1656,
+      "step": 943
+    },
+    {
+      "epoch": 3.8218623481781377,
+      "grad_norm": 0.2631670182855105,
+      "learning_rate": 6.397532352852191e-06,
+      "loss": 0.1667,
+      "step": 944
+    },
+    {
+      "epoch": 3.8259109311740893,
+      "grad_norm": 0.28306828939552237,
+      "learning_rate": 6.356126948371202e-06,
+      "loss": 0.1835,
+      "step": 945
+    },
+    {
+      "epoch": 3.8299595141700404,
+      "grad_norm": 0.2520391638088552,
+      "learning_rate": 6.314830639935559e-06,
+      "loss": 0.1734,
+      "step": 946
+    },
+    {
+      "epoch": 3.834008097165992,
+      "grad_norm": 0.28447453133339384,
+      "learning_rate": 6.273643757749448e-06,
+      "loss": 0.1787,
+      "step": 947
+    },
+    {
+      "epoch": 3.8380566801619436,
+      "grad_norm": 0.26205113286559495,
+      "learning_rate": 6.2325666311420856e-06,
+      "loss": 0.1787,
+      "step": 948
+    },
+    {
+      "epoch": 3.8421052631578947,
+      "grad_norm": 0.2581141774226386,
+      "learning_rate": 6.1915995885650675e-06,
+      "loss": 0.1815,
+      "step": 949
+    },
+    {
+      "epoch": 3.8461538461538463,
+      "grad_norm": 0.2661161708127898,
+      "learning_rate": 6.150742957589784e-06,
+      "loss": 0.1636,
+      "step": 950
+    },
+    {
+      "epoch": 3.850202429149798,
+      "grad_norm": 0.2881519752789372,
+      "learning_rate": 6.109997064904774e-06,
+      "loss": 0.1686,
+      "step": 951
+    },
+    {
+      "epoch": 3.854251012145749,
+      "grad_norm": 0.2632534066042792,
+      "learning_rate": 6.069362236313092e-06,
+      "loss": 0.1735,
+      "step": 952
+    },
+    {
+      "epoch": 3.8582995951417005,
+      "grad_norm": 0.26557515993765546,
+      "learning_rate": 6.0288387967297565e-06,
+      "loss": 0.1908,
+      "step": 953
+    },
+    {
+      "epoch": 3.862348178137652,
+      "grad_norm": 0.26531347562584423,
+      "learning_rate": 5.988427070179108e-06,
+      "loss": 0.1625,
+      "step": 954
+    },
+    {
+      "epoch": 3.866396761133603,
+      "grad_norm": 0.28097268557236965,
+      "learning_rate": 5.948127379792234e-06,
+      "loss": 0.177,
+      "step": 955
+    },
+    {
+      "epoch": 3.8704453441295548,
+      "grad_norm": 0.24960090962312584,
+      "learning_rate": 5.90794004780439e-06,
+      "loss": 0.1899,
+      "step": 956
+    },
+    {
+      "epoch": 3.8744939271255063,
+      "grad_norm": 0.25533734603899966,
+      "learning_rate": 5.867865395552405e-06,
+      "loss": 0.1977,
+      "step": 957
+    },
+    {
+      "epoch": 3.8785425101214575,
+      "grad_norm": 0.2641165422726828,
+      "learning_rate": 5.8279037434721364e-06,
+      "loss": 0.1709,
+      "step": 958
+    },
+    {
+      "epoch": 3.882591093117409,
+      "grad_norm": 0.2672954926442514,
+      "learning_rate": 5.788055411095892e-06,
+      "loss": 0.1715,
+      "step": 959
+    },
+    {
+      "epoch": 3.8866396761133606,
+      "grad_norm": 0.29002527626513125,
+      "learning_rate": 5.7483207170498625e-06,
+      "loss": 0.1769,
+      "step": 960
+    },
+    {
+      "epoch": 3.8906882591093117,
+      "grad_norm": 0.26100483722473766,
+      "learning_rate": 5.708699979051613e-06,
+      "loss": 0.1634,
+      "step": 961
+    },
+    {
+      "epoch": 3.8947368421052633,
+      "grad_norm": 0.26168209559002187,
+      "learning_rate": 5.66919351390751e-06,
+      "loss": 0.1836,
+      "step": 962
+    },
+    {
+      "epoch": 3.898785425101215,
+      "grad_norm": 0.25716279349786203,
+      "learning_rate": 5.629801637510188e-06,
+      "loss": 0.1785,
+      "step": 963
+    },
+    {
+      "epoch": 3.902834008097166,
+      "grad_norm": 0.25838078028610384,
+      "learning_rate": 5.590524664836043e-06,
+      "loss": 0.1735,
+      "step": 964
+    },
+    {
+      "epoch": 3.9068825910931175,
+      "grad_norm": 0.27472877410951135,
+      "learning_rate": 5.551362909942708e-06,
+      "loss": 0.1791,
+      "step": 965
+    },
+    {
+      "epoch": 3.910931174089069,
+      "grad_norm": 0.2754694226396934,
+      "learning_rate": 5.51231668596653e-06,
+      "loss": 0.1833,
+      "step": 966
+    },
+    {
+      "epoch": 3.91497975708502,
+      "grad_norm": 0.2609404313805043,
+      "learning_rate": 5.47338630512007e-06,
+      "loss": 0.1837,
+      "step": 967
+    },
+    {
+      "epoch": 3.919028340080972,
+      "grad_norm": 0.2650241661039199,
+      "learning_rate": 5.434572078689617e-06,
+      "loss": 0.1797,
+      "step": 968
+    },
+    {
+      "epoch": 3.9230769230769234,
+      "grad_norm": 0.262989849503344,
+      "learning_rate": 5.395874317032694e-06,
+      "loss": 0.153,
+      "step": 969
+    },
+    {
+      "epoch": 3.9271255060728745,
+      "grad_norm": 0.26789308317867583,
+      "learning_rate": 5.357293329575573e-06,
+      "loss": 0.1901,
+      "step": 970
+    },
+    {
+      "epoch": 3.931174089068826,
+      "grad_norm": 0.25984517172572935,
+      "learning_rate": 5.318829424810797e-06,
+      "loss": 0.181,
+      "step": 971
+    },
+    {
+      "epoch": 3.9352226720647776,
+      "grad_norm": 0.26163879490290126,
+      "learning_rate": 5.280482910294728e-06,
+      "loss": 0.1877,
+      "step": 972
+    },
+    {
+      "epoch": 3.9392712550607287,
+      "grad_norm": 0.2651222756412902,
+      "learning_rate": 5.242254092645078e-06,
+      "loss": 0.1729,
+      "step": 973
+    },
+    {
+      "epoch": 3.9433198380566803,
+      "grad_norm": 0.2907194666199456,
+      "learning_rate": 5.2041432775384445e-06,
+      "loss": 0.1874,
+      "step": 974
+    },
+    {
+      "epoch": 3.9473684210526314,
+      "grad_norm": 0.3642309765197082,
+      "learning_rate": 5.166150769707892e-06,
+      "loss": 0.1802,
+      "step": 975
+    },
+    {
+      "epoch": 3.951417004048583,
+      "grad_norm": 0.26347606226194253,
+      "learning_rate": 5.128276872940503e-06,
+      "loss": 0.1664,
+      "step": 976
+    },
+    {
+      "epoch": 3.9554655870445345,
+      "grad_norm": 0.26341288549796044,
+      "learning_rate": 5.090521890074953e-06,
+      "loss": 0.1808,
+      "step": 977
+    },
+    {
+      "epoch": 3.9595141700404857,
+      "grad_norm": 0.2752897421772717,
+      "learning_rate": 5.0528861229990675e-06,
+      "loss": 0.1864,
+      "step": 978
+    },
+    {
+      "epoch": 3.9635627530364372,
+      "grad_norm": 0.2668035705782594,
+      "learning_rate": 5.015369872647442e-06,
+      "loss": 0.1803,
+      "step": 979
+    },
+    {
+      "epoch": 3.967611336032389,
+      "grad_norm": 0.26291177387104064,
+      "learning_rate": 4.9779734389990195e-06,
+      "loss": 0.1702,
+      "step": 980
+    },
+    {
+      "epoch": 3.97165991902834,
+      "grad_norm": 0.260973477352893,
+      "learning_rate": 4.940697121074675e-06,
+      "loss": 0.1749,
+      "step": 981
+    },
+    {
+      "epoch": 3.9757085020242915,
+      "grad_norm": 0.2448274266408971,
+      "learning_rate": 4.903541216934864e-06,
+      "loss": 0.1715,
+      "step": 982
+    },
+    {
+      "epoch": 3.979757085020243,
+      "grad_norm": 0.25223244853078564,
+      "learning_rate": 4.8665060236772e-06,
+      "loss": 0.1744,
+      "step": 983
+    },
+    {
+      "epoch": 3.983805668016194,
+      "grad_norm": 0.2587103678457052,
+      "learning_rate": 4.829591837434102e-06,
+      "loss": 0.1744,
+      "step": 984
+    },
+    {
+      "epoch": 3.9878542510121457,
+      "grad_norm": 0.25818262253714425,
+      "learning_rate": 4.792798953370424e-06,
+      "loss": 0.2031,
+      "step": 985
+    },
+    {
+      "epoch": 3.9919028340080973,
+      "grad_norm": 0.2663882839659393,
+      "learning_rate": 4.756127665681083e-06,
+      "loss": 0.1709,
+      "step": 986
+    },
+    {
+      "epoch": 3.9959514170040484,
+      "grad_norm": 0.26272326856739103,
+      "learning_rate": 4.719578267588725e-06,
+      "loss": 0.1874,
+      "step": 987
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.26716990188918005,
+      "learning_rate": 4.6831510513413525e-06,
+      "loss": 0.1698,
+      "step": 988
+    },
+    {
+      "epoch": 4.004048582995951,
+      "grad_norm": 0.35802781511044324,
+      "learning_rate": 4.6468463082100176e-06,
+      "loss": 0.1304,
+      "step": 989
+    },
+    {
+      "epoch": 4.008097165991903,
+      "grad_norm": 0.3008425916337663,
+      "learning_rate": 4.610664328486485e-06,
+      "loss": 0.1301,
+      "step": 990
+    },
+    {
+      "epoch": 4.012145748987854,
+      "grad_norm": 0.2405359479901709,
+      "learning_rate": 4.5746054014809e-06,
+      "loss": 0.1395,
+      "step": 991
+    },
+    {
+      "epoch": 4.016194331983805,
+      "grad_norm": 0.25890714919072405,
+      "learning_rate": 4.5386698155194766e-06,
+      "loss": 0.1377,
+      "step": 992
+    },
+    {
+      "epoch": 4.020242914979757,
+      "grad_norm": 0.3403612015449352,
+      "learning_rate": 4.502857857942204e-06,
+      "loss": 0.1376,
+      "step": 993
+    },
+    {
+      "epoch": 4.0242914979757085,
+      "grad_norm": 0.36450321752896137,
+      "learning_rate": 4.467169815100549e-06,
+      "loss": 0.1309,
+      "step": 994
+    },
+    {
+      "epoch": 4.02834008097166,
+      "grad_norm": 0.32878777338887044,
+      "learning_rate": 4.431605972355137e-06,
+      "loss": 0.1557,
+      "step": 995
+    },
+    {
+      "epoch": 4.032388663967612,
+      "grad_norm": 0.24249853364841722,
+      "learning_rate": 4.396166614073516e-06,
+      "loss": 0.1389,
+      "step": 996
+    },
+    {
+      "epoch": 4.036437246963563,
+      "grad_norm": 0.2884133159193406,
+      "learning_rate": 4.360852023627846e-06,
+      "loss": 0.1386,
+      "step": 997
+    },
+    {
+      "epoch": 4.040485829959514,
+      "grad_norm": 0.30170898586401723,
+      "learning_rate": 4.325662483392656e-06,
+      "loss": 0.1318,
+      "step": 998
+    },
+    {
+      "epoch": 4.044534412955466,
+      "grad_norm": 0.3112261284443041,
+      "learning_rate": 4.29059827474257e-06,
+      "loss": 0.1267,
+      "step": 999
+    },
+    {
+      "epoch": 4.048582995951417,
+      "grad_norm": 0.26456007005843984,
+      "learning_rate": 4.255659678050063e-06,
+      "loss": 0.1208,
+      "step": 1000
+    },
+    {
+      "epoch": 4.052631578947368,
+      "grad_norm": 0.24329510708089916,
+      "learning_rate": 4.22084697268323e-06,
+      "loss": 0.1492,
+      "step": 1001
+    },
+    {
+      "epoch": 4.05668016194332,
+      "grad_norm": 0.24669194121482227,
+      "learning_rate": 4.186160437003535e-06,
+      "loss": 0.1233,
+      "step": 1002
+    },
+    {
+      "epoch": 4.060728744939271,
+      "grad_norm": 0.26247424844876943,
+      "learning_rate": 4.151600348363582e-06,
+      "loss": 0.1236,
+      "step": 1003
+    },
+    {
+      "epoch": 4.064777327935222,
+      "grad_norm": 0.2941337015460765,
+      "learning_rate": 4.11716698310493e-06,
+      "loss": 0.1306,
+      "step": 1004
+    },
+    {
+      "epoch": 4.068825910931174,
+      "grad_norm": 0.25316483467548306,
+      "learning_rate": 4.082860616555848e-06,
+      "loss": 0.1126,
+      "step": 1005
+    },
+    {
+      "epoch": 4.0728744939271255,
+      "grad_norm": 0.23635964040831484,
+      "learning_rate": 4.048681523029125e-06,
+      "loss": 0.1401,
+      "step": 1006
+    },
+    {
+      "epoch": 4.076923076923077,
+      "grad_norm": 0.22614440909002856,
+      "learning_rate": 4.014629975819882e-06,
+      "loss": 0.1388,
+      "step": 1007
+    },
+    {
+      "epoch": 4.080971659919029,
+      "grad_norm": 0.44501270372221235,
+      "learning_rate": 3.9807062472033855e-06,
+      "loss": 0.1363,
+      "step": 1008
+    },
+    {
+      "epoch": 4.08502024291498,
+      "grad_norm": 0.24309135378543556,
+      "learning_rate": 3.94691060843287e-06,
+      "loss": 0.1355,
+      "step": 1009
+    },
+    {
+      "epoch": 4.089068825910931,
+      "grad_norm": 0.23851408511403718,
+      "learning_rate": 3.913243329737353e-06,
+      "loss": 0.1304,
+      "step": 1010
+    },
+    {
+      "epoch": 4.093117408906883,
+      "grad_norm": 0.2489787210685666,
+      "learning_rate": 3.879704680319503e-06,
+      "loss": 0.1227,
+      "step": 1011
+    },
+    {
+      "epoch": 4.097165991902834,
+      "grad_norm": 0.22583625813689714,
+      "learning_rate": 3.846294928353471e-06,
+      "loss": 0.1378,
+      "step": 1012
+    },
+    {
+      "epoch": 4.101214574898785,
+      "grad_norm": 0.2482087391603314,
+      "learning_rate": 3.8130143409827345e-06,
+      "loss": 0.1437,
+      "step": 1013
+    },
+    {
+      "epoch": 4.105263157894737,
+      "grad_norm": 0.23935341720149075,
+      "learning_rate": 3.7798631843179887e-06,
+      "loss": 0.1327,
+      "step": 1014
+    },
+    {
+      "epoch": 4.109311740890688,
+      "grad_norm": 0.24278779550058338,
+      "learning_rate": 3.7468417234349973e-06,
+      "loss": 0.1346,
+      "step": 1015
+    },
+    {
+      "epoch": 4.113360323886639,
+      "grad_norm": 0.24006619854293884,
+      "learning_rate": 3.7139502223724823e-06,
+      "loss": 0.1343,
+      "step": 1016
+    },
+    {
+      "epoch": 4.117408906882591,
+      "grad_norm": 0.22528182947502362,
+      "learning_rate": 3.6811889441300095e-06,
+      "loss": 0.1391,
+      "step": 1017
+    },
+    {
+      "epoch": 4.1214574898785425,
+      "grad_norm": 0.23738653626986445,
+      "learning_rate": 3.648558150665886e-06,
+      "loss": 0.1278,
+      "step": 1018
+    },
+    {
+      "epoch": 4.125506072874494,
+      "grad_norm": 0.23295266693898725,
+      "learning_rate": 3.616058102895075e-06,
+      "loss": 0.123,
+      "step": 1019
+    },
+    {
+      "epoch": 4.129554655870446,
+      "grad_norm": 0.2153195831543166,
+      "learning_rate": 3.583689060687079e-06,
+      "loss": 0.1327,
+      "step": 1020
+    },
+    {
+      "epoch": 4.133603238866397,
+      "grad_norm": 0.2290570577968874,
+      "learning_rate": 3.5514512828639023e-06,
+      "loss": 0.123,
+      "step": 1021
+    },
+    {
+      "epoch": 4.137651821862348,
+      "grad_norm": 0.23332155138021465,
+      "learning_rate": 3.51934502719796e-06,
+      "loss": 0.114,
+      "step": 1022
+    },
+    {
+      "epoch": 4.1417004048583,
+      "grad_norm": 0.24991759316300105,
+      "learning_rate": 3.4873705504100185e-06,
+      "loss": 0.1352,
+      "step": 1023
+    },
+    {
+      "epoch": 4.145748987854251,
+      "grad_norm": 0.23493477986479416,
+      "learning_rate": 3.4555281081671345e-06,
+      "loss": 0.1287,
+      "step": 1024
+    },
+    {
+      "epoch": 4.149797570850202,
+      "grad_norm": 0.23977693895444635,
+      "learning_rate": 3.4238179550806326e-06,
+      "loss": 0.1408,
+      "step": 1025
+    },
+    {
+      "epoch": 4.153846153846154,
+      "grad_norm": 0.2234507587562011,
+      "learning_rate": 3.3922403447040497e-06,
+      "loss": 0.1395,
+      "step": 1026
+    },
+    {
+      "epoch": 4.157894736842105,
+      "grad_norm": 0.22566472933864842,
+      "learning_rate": 3.3607955295311247e-06,
+      "loss": 0.1406,
+      "step": 1027
+    },
+    {
+      "epoch": 4.161943319838056,
+      "grad_norm": 0.23230447105705387,
+      "learning_rate": 3.32948376099375e-06,
+      "loss": 0.1394,
+      "step": 1028
+    },
+    {
+      "epoch": 4.165991902834008,
+      "grad_norm": 0.22979073413179538,
+      "learning_rate": 3.2983052894599976e-06,
+      "loss": 0.1395,
+      "step": 1029
+    },
+    {
+      "epoch": 4.17004048582996,
+      "grad_norm": 0.2179821529690815,
+      "learning_rate": 3.2672603642320945e-06,
+      "loss": 0.1624,
+      "step": 1030
+    },
+    {
+      "epoch": 4.174089068825911,
+      "grad_norm": 0.22195643349473082,
+      "learning_rate": 3.236349233544438e-06,
+      "loss": 0.126,
+      "step": 1031
+    },
+    {
+      "epoch": 4.178137651821863,
+      "grad_norm": 0.20907217298071387,
+      "learning_rate": 3.205572144561604e-06,
+      "loss": 0.1269,
+      "step": 1032
+    },
+    {
+      "epoch": 4.182186234817814,
+      "grad_norm": 0.21791445748868435,
+      "learning_rate": 3.174929343376374e-06,
+      "loss": 0.1263,
+      "step": 1033
+    },
+    {
+      "epoch": 4.186234817813765,
+      "grad_norm": 0.2282759300196614,
+      "learning_rate": 3.1444210750077777e-06,
+      "loss": 0.1524,
+      "step": 1034
+    },
+    {
+      "epoch": 4.190283400809717,
+      "grad_norm": 0.22621037247684836,
+      "learning_rate": 3.1140475833991046e-06,
+      "loss": 0.1131,
+      "step": 1035
+    },
+    {
+      "epoch": 4.194331983805668,
+      "grad_norm": 0.22122709030442766,
+      "learning_rate": 3.0838091114159895e-06,
+      "loss": 0.1303,
+      "step": 1036
+    },
+    {
+      "epoch": 4.198380566801619,
+      "grad_norm": 0.2220872351961223,
+      "learning_rate": 3.0537059008444524e-06,
+      "loss": 0.1214,
+      "step": 1037
+    },
+    {
+      "epoch": 4.202429149797571,
+      "grad_norm": 0.22518591285724054,
+      "learning_rate": 3.0237381923889563e-06,
+      "loss": 0.1474,
+      "step": 1038
+    },
+    {
+      "epoch": 4.206477732793522,
+      "grad_norm": 0.21763580739379007,
+      "learning_rate": 2.9939062256704997e-06,
+      "loss": 0.1468,
+      "step": 1039
+    },
+    {
+      "epoch": 4.2105263157894735,
+      "grad_norm": 0.2187802711396891,
+      "learning_rate": 2.964210239224696e-06,
+      "loss": 0.1278,
+      "step": 1040
+    },
+    {
+      "epoch": 4.2145748987854255,
+      "grad_norm": 0.22206668511459032,
+      "learning_rate": 2.934650470499869e-06,
+      "loss": 0.1326,
+      "step": 1041
+    },
+    {
+      "epoch": 4.218623481781377,
+      "grad_norm": 0.2353867209353303,
+      "learning_rate": 2.905227155855128e-06,
+      "loss": 0.1201,
+      "step": 1042
+    },
+    {
+      "epoch": 4.222672064777328,
+      "grad_norm": 0.22202042189028778,
+      "learning_rate": 2.875940530558519e-06,
+      "loss": 0.1153,
+      "step": 1043
+    },
+    {
+      "epoch": 4.22672064777328,
+      "grad_norm": 0.22008053633187186,
+      "learning_rate": 2.8467908287851154e-06,
+      "loss": 0.1229,
+      "step": 1044
+    },
+    {
+      "epoch": 4.230769230769231,
+      "grad_norm": 0.2138483815704101,
+      "learning_rate": 2.8177782836151603e-06,
+      "loss": 0.1341,
+      "step": 1045
+    },
+    {
+      "epoch": 4.234817813765182,
+      "grad_norm": 0.2278837122271396,
+      "learning_rate": 2.7889031270321874e-06,
+      "loss": 0.1239,
+      "step": 1046
+    },
+    {
+      "epoch": 4.238866396761134,
+      "grad_norm": 0.2225487448609366,
+      "learning_rate": 2.760165589921184e-06,
+      "loss": 0.1247,
+      "step": 1047
+    },
+    {
+      "epoch": 4.242914979757085,
+      "grad_norm": 0.21862869376457694,
+      "learning_rate": 2.7315659020667305e-06,
+      "loss": 0.1324,
+      "step": 1048
+    },
+    {
+      "epoch": 4.246963562753036,
+      "grad_norm": 0.21921263684749667,
+      "learning_rate": 2.703104292151164e-06,
+      "loss": 0.1359,
+      "step": 1049
+    },
+    {
+      "epoch": 4.251012145748988,
+      "grad_norm": 0.22218507506004528,
+      "learning_rate": 2.6747809877527607e-06,
+      "loss": 0.1315,
+      "step": 1050
+    },
+    {
+      "epoch": 4.255060728744939,
+      "grad_norm": 0.22438466054564438,
+      "learning_rate": 2.6465962153439082e-06,
+      "loss": 0.1224,
+      "step": 1051
+    },
+    {
+      "epoch": 4.2591093117408905,
+      "grad_norm": 0.21474216787257747,
+      "learning_rate": 2.618550200289303e-06,
+      "loss": 0.1455,
+      "step": 1052
+    },
+    {
+      "epoch": 4.2631578947368425,
+      "grad_norm": 0.2212317662031547,
+      "learning_rate": 2.59064316684412e-06,
+      "loss": 0.1422,
+      "step": 1053
+    },
+    {
+      "epoch": 4.267206477732794,
+      "grad_norm": 0.21784265973478584,
+      "learning_rate": 2.562875338152264e-06,
+      "loss": 0.1345,
+      "step": 1054
+    },
+    {
+      "epoch": 4.271255060728745,
+      "grad_norm": 0.21559088560103537,
+      "learning_rate": 2.5352469362445552e-06,
+      "loss": 0.1491,
+      "step": 1055
+    },
+    {
+      "epoch": 4.275303643724697,
+      "grad_norm": 0.2244081239414837,
+      "learning_rate": 2.5077581820369523e-06,
+      "loss": 0.1415,
+      "step": 1056
+    },
+    {
+      "epoch": 4.279352226720648,
+      "grad_norm": 0.2332814886653632,
+      "learning_rate": 2.480409295328803e-06,
+      "loss": 0.1125,
+      "step": 1057
+    },
+    {
+      "epoch": 4.283400809716599,
+      "grad_norm": 0.21403764307286932,
+      "learning_rate": 2.45320049480108e-06,
+      "loss": 0.1391,
+      "step": 1058
+    },
+    {
+      "epoch": 4.287449392712551,
+      "grad_norm": 0.22216683448232494,
+      "learning_rate": 2.4261319980146293e-06,
+      "loss": 0.1405,
+      "step": 1059
+    },
+    {
+      "epoch": 4.291497975708502,
+      "grad_norm": 0.23150845881086027,
+      "learning_rate": 2.399204021408421e-06,
+      "loss": 0.1453,
+      "step": 1060
+    },
+    {
+      "epoch": 4.295546558704453,
+      "grad_norm": 0.21485034380182708,
+      "learning_rate": 2.3724167802978395e-06,
+      "loss": 0.1371,
+      "step": 1061
+    },
+    {
+      "epoch": 4.299595141700405,
+      "grad_norm": 0.2259532482837255,
+      "learning_rate": 2.3457704888729625e-06,
+      "loss": 0.1216,
+      "step": 1062
+    },
+    {
+      "epoch": 4.303643724696356,
+      "grad_norm": 0.21251470445761753,
+      "learning_rate": 2.3192653601968162e-06,
+      "loss": 0.1222,
+      "step": 1063
+    },
+    {
+      "epoch": 4.3076923076923075,
+      "grad_norm": 0.22287538898068163,
+      "learning_rate": 2.2929016062037057e-06,
+      "loss": 0.1237,
+      "step": 1064
+    },
+    {
+      "epoch": 4.3117408906882595,
+      "grad_norm": 0.21797958202094297,
+      "learning_rate": 2.2666794376975055e-06,
+      "loss": 0.1362,
+      "step": 1065
+    },
+    {
+      "epoch": 4.315789473684211,
+      "grad_norm": 0.23972340207435605,
+      "learning_rate": 2.24059906434998e-06,
+      "loss": 0.1217,
+      "step": 1066
+    },
+    {
+      "epoch": 4.319838056680162,
+      "grad_norm": 0.2183078428812038,
+      "learning_rate": 2.214660694699091e-06,
+      "loss": 0.1388,
+      "step": 1067
+    },
+    {
+      "epoch": 4.323886639676114,
+      "grad_norm": 0.22096162009366974,
+      "learning_rate": 2.188864536147355e-06,
+      "loss": 0.1388,
+      "step": 1068
+    },
+    {
+      "epoch": 4.327935222672065,
+      "grad_norm": 0.25040168450001105,
+      "learning_rate": 2.1632107949601757e-06,
+      "loss": 0.1304,
+      "step": 1069
+    },
+    {
+      "epoch": 4.331983805668016,
+      "grad_norm": 0.2318531100896853,
+      "learning_rate": 2.137699676264171e-06,
+      "loss": 0.1399,
+      "step": 1070
+    },
+    {
+      "epoch": 4.336032388663968,
+      "grad_norm": 0.22553754083027508,
+      "learning_rate": 2.1123313840455738e-06,
+      "loss": 0.129,
+      "step": 1071
+    },
+    {
+      "epoch": 4.340080971659919,
+      "grad_norm": 0.22290415594159807,
+      "learning_rate": 2.08710612114857e-06,
+      "loss": 0.1311,
+      "step": 1072
+    },
+    {
+      "epoch": 4.34412955465587,
+      "grad_norm": 0.23637883917919367,
+      "learning_rate": 2.0620240892736953e-06,
+      "loss": 0.1308,
+      "step": 1073
+    },
+    {
+      "epoch": 4.348178137651822,
+      "grad_norm": 0.22972675369424034,
+      "learning_rate": 2.0370854889762e-06,
+      "loss": 0.1282,
+      "step": 1074
+    },
+    {
+      "epoch": 4.352226720647773,
+      "grad_norm": 0.21609493567983334,
+      "learning_rate": 2.0122905196644726e-06,
+      "loss": 0.1247,
+      "step": 1075
+    },
+    {
+      "epoch": 4.3562753036437245,
+      "grad_norm": 0.22087436077170336,
+      "learning_rate": 1.987639379598427e-06,
+      "loss": 0.1389,
+      "step": 1076
+    },
+    {
+      "epoch": 4.3603238866396765,
+      "grad_norm": 0.3550518532287784,
+      "learning_rate": 1.963132265887919e-06,
+      "loss": 0.123,
+      "step": 1077
+    },
+    {
+      "epoch": 4.364372469635628,
+      "grad_norm": 0.21117927034518175,
+      "learning_rate": 1.9387693744911784e-06,
+      "loss": 0.1377,
+      "step": 1078
+    },
+    {
+      "epoch": 4.368421052631579,
+      "grad_norm": 0.21408634574361451,
+      "learning_rate": 1.9145509002132368e-06,
+      "loss": 0.1195,
+      "step": 1079
+    },
+    {
+      "epoch": 4.372469635627531,
+      "grad_norm": 0.2164085378229301,
+      "learning_rate": 1.8904770367043723e-06,
+      "loss": 0.1341,
+      "step": 1080
+    },
+    {
+      "epoch": 4.376518218623482,
+      "grad_norm": 0.23293031246154733,
+      "learning_rate": 1.8665479764585504e-06,
+      "loss": 0.1327,
+      "step": 1081
+    },
+    {
+      "epoch": 4.380566801619433,
+      "grad_norm": 0.2293070078616296,
+      "learning_rate": 1.8427639108118977e-06,
+      "loss": 0.1319,
+      "step": 1082
+    },
+    {
+      "epoch": 4.384615384615385,
+      "grad_norm": 0.22141247553872767,
+      "learning_rate": 1.8191250299411755e-06,
+      "loss": 0.1311,
+      "step": 1083
+    },
+    {
+      "epoch": 4.388663967611336,
+      "grad_norm": 0.2294056211601881,
+      "learning_rate": 1.7956315228622446e-06,
+      "loss": 0.1376,
+      "step": 1084
+    },
+    {
+      "epoch": 4.392712550607287,
+      "grad_norm": 0.2430377748157182,
+      "learning_rate": 1.7722835774285553e-06,
+      "loss": 0.1297,
+      "step": 1085
+    },
+    {
+      "epoch": 4.396761133603239,
+      "grad_norm": 0.21932700306076366,
+      "learning_rate": 1.7490813803296624e-06,
+      "loss": 0.1184,
+      "step": 1086
+    },
+    {
+      "epoch": 4.40080971659919,
+      "grad_norm": 0.22699512204061234,
+      "learning_rate": 1.726025117089718e-06,
+      "loss": 0.1464,
+      "step": 1087
+    },
+    {
+      "epoch": 4.4048582995951415,
+      "grad_norm": 0.22649909782588623,
+      "learning_rate": 1.70311497206598e-06,
+      "loss": 0.1344,
+      "step": 1088
+    },
+    {
+      "epoch": 4.4089068825910935,
+      "grad_norm": 0.23505712354782882,
+      "learning_rate": 1.6803511284473638e-06,
+      "loss": 0.1157,
+      "step": 1089
+    },
+    {
+      "epoch": 4.412955465587045,
+      "grad_norm": 0.23778661476877955,
+      "learning_rate": 1.6577337682529538e-06,
+      "loss": 0.1217,
+      "step": 1090
+    },
+    {
+      "epoch": 4.417004048582996,
+      "grad_norm": 0.20970729592011816,
+      "learning_rate": 1.6352630723305617e-06,
+      "loss": 0.1529,
+      "step": 1091
+    },
+    {
+      "epoch": 4.421052631578947,
+      "grad_norm": 0.22643048138722371,
+      "learning_rate": 1.612939220355272e-06,
+      "loss": 0.1297,
+      "step": 1092
+    },
+    {
+      "epoch": 4.425101214574899,
+      "grad_norm": 0.21653578187732236,
+      "learning_rate": 1.5907623908280112e-06,
+      "loss": 0.1549,
+      "step": 1093
+    },
+    {
+      "epoch": 4.42914979757085,
+      "grad_norm": 0.21731859611450025,
+      "learning_rate": 1.568732761074121e-06,
+      "loss": 0.1315,
+      "step": 1094
+    },
+    {
+      "epoch": 4.433198380566802,
+      "grad_norm": 0.21111054090588663,
+      "learning_rate": 1.5468505072419237e-06,
+      "loss": 0.1396,
+      "step": 1095
+    },
+    {
+      "epoch": 4.437246963562753,
+      "grad_norm": 0.21924290905059698,
+      "learning_rate": 1.5251158043013448e-06,
+      "loss": 0.1325,
+      "step": 1096
+    },
+    {
+      "epoch": 4.441295546558704,
+      "grad_norm": 0.24856842710262847,
+      "learning_rate": 1.5035288260424885e-06,
+      "loss": 0.1318,
+      "step": 1097
+    },
+    {
+      "epoch": 4.445344129554655,
+      "grad_norm": 0.23030727121224248,
+      "learning_rate": 1.4820897450742623e-06,
+      "loss": 0.1262,
+      "step": 1098
+    },
+    {
+      "epoch": 4.449392712550607,
+      "grad_norm": 0.2233816463900621,
+      "learning_rate": 1.4607987328229767e-06,
+      "loss": 0.1119,
+      "step": 1099
+    },
+    {
+      "epoch": 4.4534412955465585,
+      "grad_norm": 0.25959343842960625,
+      "learning_rate": 1.4396559595310056e-06,
+      "loss": 0.1253,
+      "step": 1100
+    },
+    {
+      "epoch": 4.4574898785425106,
+      "grad_norm": 0.22117737709512583,
+      "learning_rate": 1.4186615942553973e-06,
+      "loss": 0.1299,
+      "step": 1101
+    },
+    {
+      "epoch": 4.461538461538462,
+      "grad_norm": 0.22581813550495544,
+      "learning_rate": 1.3978158048665424e-06,
+      "loss": 0.1173,
+      "step": 1102
+    },
+    {
+      "epoch": 4.465587044534413,
+      "grad_norm": 0.22062615883372425,
+      "learning_rate": 1.3771187580468115e-06,
+      "loss": 0.1354,
+      "step": 1103
+    },
+    {
+      "epoch": 4.469635627530364,
+      "grad_norm": 0.2454506238385963,
+      "learning_rate": 1.3565706192892392e-06,
+      "loss": 0.1289,
+      "step": 1104
+    },
+    {
+      "epoch": 4.473684210526316,
+      "grad_norm": 0.22405723276456538,
+      "learning_rate": 1.3361715528961993e-06,
+      "loss": 0.1274,
+      "step": 1105
+    },
+    {
+      "epoch": 4.477732793522267,
+      "grad_norm": 0.22478103885882467,
+      "learning_rate": 1.3159217219780817e-06,
+      "loss": 0.1414,
+      "step": 1106
+    },
+    {
+      "epoch": 4.481781376518219,
+      "grad_norm": 0.2183678231996747,
+      "learning_rate": 1.2958212884519949e-06,
+      "loss": 0.1252,
+      "step": 1107
+    },
+    {
+      "epoch": 4.48582995951417,
+      "grad_norm": 0.21921552576221762,
+      "learning_rate": 1.2758704130404675e-06,
+      "loss": 0.1395,
+      "step": 1108
+    },
+    {
+      "epoch": 4.489878542510121,
+      "grad_norm": 0.22382605518346596,
+      "learning_rate": 1.2560692552701738e-06,
+      "loss": 0.1266,
+      "step": 1109
+    },
+    {
+      "epoch": 4.493927125506072,
+      "grad_norm": 0.3279208777732047,
+      "learning_rate": 1.2364179734706382e-06,
+      "loss": 0.1311,
+      "step": 1110
+    },
+    {
+      "epoch": 4.497975708502024,
+      "grad_norm": 0.21153561444287355,
+      "learning_rate": 1.2169167247729852e-06,
+      "loss": 0.1179,
+      "step": 1111
+    },
+    {
+      "epoch": 4.502024291497976,
+      "grad_norm": 0.22348925475295944,
+      "learning_rate": 1.1975656651086864e-06,
+      "loss": 0.1361,
+      "step": 1112
+    },
+    {
+      "epoch": 4.506072874493928,
+      "grad_norm": 0.21104458051834424,
+      "learning_rate": 1.1783649492082926e-06,
+      "loss": 0.1214,
+      "step": 1113
+    },
+    {
+      "epoch": 4.510121457489879,
+      "grad_norm": 0.21976628682592694,
+      "learning_rate": 1.1593147306002183e-06,
+      "loss": 0.1367,
+      "step": 1114
+    },
+    {
+      "epoch": 4.51417004048583,
+      "grad_norm": 0.2282728889099151,
+      "learning_rate": 1.140415161609505e-06,
+      "loss": 0.1373,
+      "step": 1115
+    },
+    {
+      "epoch": 4.518218623481781,
+      "grad_norm": 0.21820563393527886,
+      "learning_rate": 1.121666393356604e-06,
+      "loss": 0.1144,
+      "step": 1116
+    },
+    {
+      "epoch": 4.522267206477733,
+      "grad_norm": 0.2282313041174654,
+      "learning_rate": 1.1030685757561632e-06,
+      "loss": 0.1407,
+      "step": 1117
+    },
+    {
+      "epoch": 4.526315789473684,
+      "grad_norm": 0.2106450788297929,
+      "learning_rate": 1.0846218575158373e-06,
+      "loss": 0.1323,
+      "step": 1118
+    },
+    {
+      "epoch": 4.530364372469636,
+      "grad_norm": 0.26114538802396314,
+      "learning_rate": 1.066326386135097e-06,
+      "loss": 0.1254,
+      "step": 1119
+    },
+    {
+      "epoch": 4.534412955465587,
+      "grad_norm": 0.22276686608610285,
+      "learning_rate": 1.0481823079040398e-06,
+      "loss": 0.1301,
+      "step": 1120
+    },
+    {
+      "epoch": 4.538461538461538,
+      "grad_norm": 0.22957154789542555,
+      "learning_rate": 1.0301897679022344e-06,
+      "loss": 0.1353,
+      "step": 1121
+    },
+    {
+      "epoch": 4.5425101214574894,
+      "grad_norm": 0.2567402560352487,
+      "learning_rate": 1.0123489099975491e-06,
+      "loss": 0.1206,
+      "step": 1122
+    },
+    {
+      "epoch": 4.5465587044534415,
+      "grad_norm": 0.21940271730481775,
+      "learning_rate": 9.9465987684501e-07,
+      "loss": 0.1281,
+      "step": 1123
+    },
+    {
+      "epoch": 4.550607287449393,
+      "grad_norm": 0.2182380103050568,
+      "learning_rate": 9.771228098856534e-07,
+      "loss": 0.1264,
+      "step": 1124
+    },
+    {
+      "epoch": 4.554655870445345,
+      "grad_norm": 0.23749175766344582,
+      "learning_rate": 9.597378493454013e-07,
+      "loss": 0.13,
+      "step": 1125
+    },
+    {
+      "epoch": 4.558704453441296,
+      "grad_norm": 0.21576628143248858,
+      "learning_rate": 9.425051342339353e-07,
+      "loss": 0.1329,
+      "step": 1126
+    },
+    {
+      "epoch": 4.562753036437247,
+      "grad_norm": 0.22050001850513923,
+      "learning_rate": 9.254248023435864e-07,
+      "loss": 0.1217,
+      "step": 1127
+    },
+    {
+      "epoch": 4.566801619433198,
+      "grad_norm": 0.21272703097509488,
+      "learning_rate": 9.084969902482288e-07,
+      "loss": 0.1423,
+      "step": 1128
+    },
+    {
+      "epoch": 4.57085020242915,
+      "grad_norm": 0.2203917560163949,
+      "learning_rate": 8.917218333021993e-07,
+      "loss": 0.1284,
+      "step": 1129
+    },
+    {
+      "epoch": 4.574898785425101,
+      "grad_norm": 0.21467524080112974,
+      "learning_rate": 8.750994656392109e-07,
+      "loss": 0.1218,
+      "step": 1130
+    },
+    {
+      "epoch": 4.578947368421053,
+      "grad_norm": 0.21271295855941413,
+      "learning_rate": 8.58630020171265e-07,
+      "loss": 0.1031,
+      "step": 1131
+    },
+    {
+      "epoch": 4.582995951417004,
+      "grad_norm": 0.21454579889123787,
+      "learning_rate": 8.423136285876099e-07,
+      "loss": 0.122,
+      "step": 1132
+    },
+    {
+      "epoch": 4.587044534412955,
+      "grad_norm": 0.21555804842481563,
+      "learning_rate": 8.26150421353682e-07,
+      "loss": 0.148,
+      "step": 1133
+    },
+    {
+      "epoch": 4.5910931174089065,
+      "grad_norm": 0.20884365995893658,
+      "learning_rate": 8.101405277100549e-07,
+      "loss": 0.1396,
+      "step": 1134
+    },
+    {
+      "epoch": 4.5951417004048585,
+      "grad_norm": 0.22205234245948144,
+      "learning_rate": 7.942840756714077e-07,
+      "loss": 0.1413,
+      "step": 1135
+    },
+    {
+      "epoch": 4.59919028340081,
+      "grad_norm": 0.22488568955536528,
+      "learning_rate": 7.785811920255093e-07,
+      "loss": 0.1276,
+      "step": 1136
+    },
+    {
+      "epoch": 4.603238866396762,
+      "grad_norm": 0.22376294847406575,
+      "learning_rate": 7.630320023322024e-07,
+      "loss": 0.1422,
+      "step": 1137
+    },
+    {
+      "epoch": 4.607287449392713,
+      "grad_norm": 0.22534972555943386,
+      "learning_rate": 7.476366309223903e-07,
+      "loss": 0.1341,
+      "step": 1138
+    },
+    {
+      "epoch": 4.611336032388664,
+      "grad_norm": 0.21955807408318964,
+      "learning_rate": 7.323952008970537e-07,
+      "loss": 0.1368,
+      "step": 1139
+    },
+    {
+      "epoch": 4.615384615384615,
+      "grad_norm": 0.21299155944085763,
+      "learning_rate": 7.173078341262618e-07,
+      "loss": 0.1283,
+      "step": 1140
+    },
+    {
+      "epoch": 4.619433198380567,
+      "grad_norm": 0.21050944700712976,
+      "learning_rate": 7.023746512481989e-07,
+      "loss": 0.1213,
+      "step": 1141
+    },
+    {
+      "epoch": 4.623481781376518,
+      "grad_norm": 0.21839335174682029,
+      "learning_rate": 6.875957716681902e-07,
+      "loss": 0.106,
+      "step": 1142
+    },
+    {
+      "epoch": 4.62753036437247,
+      "grad_norm": 0.2148969449883673,
+      "learning_rate": 6.729713135577687e-07,
+      "loss": 0.1388,
+      "step": 1143
+    },
+    {
+      "epoch": 4.631578947368421,
+      "grad_norm": 0.24226053179815102,
+      "learning_rate": 6.585013938537143e-07,
+      "loss": 0.1447,
+      "step": 1144
+    },
+    {
+      "epoch": 4.635627530364372,
+      "grad_norm": 0.21968266196281017,
+      "learning_rate": 6.441861282571115e-07,
+      "loss": 0.1445,
+      "step": 1145
+    },
+    {
+      "epoch": 4.6396761133603235,
+      "grad_norm": 0.23028694588986198,
+      "learning_rate": 6.300256312324471e-07,
+      "loss": 0.1312,
+      "step": 1146
+    },
+    {
+      "epoch": 4.6437246963562755,
+      "grad_norm": 0.2215623895669462,
+      "learning_rate": 6.160200160066776e-07,
+      "loss": 0.1466,
+      "step": 1147
+    },
+    {
+      "epoch": 4.647773279352227,
+      "grad_norm": 0.2200443143176215,
+      "learning_rate": 6.021693945683327e-07,
+      "loss": 0.1265,
+      "step": 1148
+    },
+    {
+      "epoch": 4.651821862348179,
+      "grad_norm": 0.21723952801239013,
+      "learning_rate": 5.884738776666044e-07,
+      "loss": 0.1432,
+      "step": 1149
+    },
+    {
+      "epoch": 4.65587044534413,
+      "grad_norm": 0.2172795470683769,
+      "learning_rate": 5.749335748104878e-07,
+      "loss": 0.1251,
+      "step": 1150
+    },
+    {
+      "epoch": 4.659919028340081,
+      "grad_norm": 0.2114366061574977,
+      "learning_rate": 5.615485942678778e-07,
+      "loss": 0.133,
+      "step": 1151
+    },
+    {
+      "epoch": 4.663967611336032,
+      "grad_norm": 0.21573059659298854,
+      "learning_rate": 5.483190430647289e-07,
+      "loss": 0.1339,
+      "step": 1152
+    },
+    {
+      "epoch": 4.668016194331984,
+      "grad_norm": 0.21561654396974295,
+      "learning_rate": 5.352450269841747e-07,
+      "loss": 0.125,
+      "step": 1153
+    },
+    {
+      "epoch": 4.672064777327935,
+      "grad_norm": 0.21272714409043944,
+      "learning_rate": 5.22326650565701e-07,
+      "loss": 0.1231,
+      "step": 1154
+    },
+    {
+      "epoch": 4.676113360323887,
+      "grad_norm": 0.21747312373848565,
+      "learning_rate": 5.095640171043048e-07,
+      "loss": 0.1337,
+      "step": 1155
+    },
+    {
+      "epoch": 4.680161943319838,
+      "grad_norm": 0.2134803019876513,
+      "learning_rate": 4.969572286496594e-07,
+      "loss": 0.1395,
+      "step": 1156
+    },
+    {
+      "epoch": 4.684210526315789,
+      "grad_norm": 0.21647989457399483,
+      "learning_rate": 4.845063860053079e-07,
+      "loss": 0.1421,
+      "step": 1157
+    },
+    {
+      "epoch": 4.6882591093117405,
+      "grad_norm": 0.22025979943764332,
+      "learning_rate": 4.7221158872785997e-07,
+      "loss": 0.1325,
+      "step": 1158
+    },
+    {
+      "epoch": 4.6923076923076925,
+      "grad_norm": 0.25165562528532365,
+      "learning_rate": 4.600729351261857e-07,
+      "loss": 0.1267,
+      "step": 1159
+    },
+    {
+      "epoch": 4.696356275303644,
+      "grad_norm": 0.2301533849683246,
+      "learning_rate": 4.4809052226063355e-07,
+      "loss": 0.1358,
+      "step": 1160
+    },
+    {
+      "epoch": 4.700404858299595,
+      "grad_norm": 0.22442519490810506,
+      "learning_rate": 4.3626444594226046e-07,
+      "loss": 0.1181,
+      "step": 1161
+    },
+    {
+      "epoch": 4.704453441295547,
+      "grad_norm": 0.22209609344191622,
+      "learning_rate": 4.245948007320588e-07,
+      "loss": 0.1293,
+      "step": 1162
+    },
+    {
+      "epoch": 4.708502024291498,
+      "grad_norm": 0.22157452600622465,
+      "learning_rate": 4.130816799401971e-07,
+      "loss": 0.13,
+      "step": 1163
+    },
+    {
+      "epoch": 4.712550607287449,
+      "grad_norm": 0.22158848732039443,
+      "learning_rate": 4.017251756252827e-07,
+      "loss": 0.1375,
+      "step": 1164
+    },
+    {
+      "epoch": 4.716599190283401,
+      "grad_norm": 0.23782584802525744,
+      "learning_rate": 3.9052537859362473e-07,
+      "loss": 0.125,
+      "step": 1165
+    },
+    {
+      "epoch": 4.720647773279352,
+      "grad_norm": 0.2175076505703528,
+      "learning_rate": 3.7948237839849687e-07,
+      "loss": 0.1248,
+      "step": 1166
+    },
+    {
+      "epoch": 4.724696356275303,
+      "grad_norm": 0.21388722542185118,
+      "learning_rate": 3.6859626333943797e-07,
+      "loss": 0.122,
+      "step": 1167
+    },
+    {
+      "epoch": 4.728744939271255,
+      "grad_norm": 0.22840628372259705,
+      "learning_rate": 3.578671204615347e-07,
+      "loss": 0.137,
+      "step": 1168
+    },
+    {
+      "epoch": 4.732793522267206,
+      "grad_norm": 0.21962648098463586,
+      "learning_rate": 3.472950355547267e-07,
+      "loss": 0.1253,
+      "step": 1169
+    },
+    {
+      "epoch": 4.7368421052631575,
+      "grad_norm": 0.22411114843905394,
+      "learning_rate": 3.3688009315312244e-07,
+      "loss": 0.1303,
+      "step": 1170
+    },
+    {
+      "epoch": 4.7408906882591095,
+      "grad_norm": 0.22233064385137516,
+      "learning_rate": 3.266223765343224e-07,
+      "loss": 0.1207,
+      "step": 1171
+    },
+    {
+      "epoch": 4.744939271255061,
+      "grad_norm": 0.2075067726294773,
+      "learning_rate": 3.16521967718757e-07,
+      "loss": 0.1304,
+      "step": 1172
+    },
+    {
+      "epoch": 4.748987854251012,
+      "grad_norm": 0.24018429100377972,
+      "learning_rate": 3.0657894746902726e-07,
+      "loss": 0.1443,
+      "step": 1173
+    },
+    {
+      "epoch": 4.753036437246964,
+      "grad_norm": 0.22259716087859413,
+      "learning_rate": 2.9679339528925654e-07,
+      "loss": 0.1352,
+      "step": 1174
+    },
+    {
+      "epoch": 4.757085020242915,
+      "grad_norm": 0.21634001475831335,
+      "learning_rate": 2.871653894244597e-07,
+      "loss": 0.1115,
+      "step": 1175
+    },
+    {
+      "epoch": 4.761133603238866,
+      "grad_norm": 0.22961259514913332,
+      "learning_rate": 2.7769500685991266e-07,
+      "loss": 0.1184,
+      "step": 1176
+    },
+    {
+      "epoch": 4.765182186234818,
+      "grad_norm": 0.23469787914606766,
+      "learning_rate": 2.683823233205485e-07,
+      "loss": 0.119,
+      "step": 1177
+    },
+    {
+      "epoch": 4.769230769230769,
+      "grad_norm": 0.23142888759984326,
+      "learning_rate": 2.592274132703287e-07,
+      "loss": 0.1311,
+      "step": 1178
+    },
+    {
+      "epoch": 4.77327935222672,
+      "grad_norm": 0.22898141545168968,
+      "learning_rate": 2.502303499116754e-07,
+      "loss": 0.1258,
+      "step": 1179
+    },
+    {
+      "epoch": 4.777327935222672,
+      "grad_norm": 0.23608009371747352,
+      "learning_rate": 2.413912051848666e-07,
+      "loss": 0.1411,
+      "step": 1180
+    },
+    {
+      "epoch": 4.781376518218623,
+      "grad_norm": 0.2051817811987931,
+      "learning_rate": 2.3271004976746613e-07,
+      "loss": 0.1268,
+      "step": 1181
+    },
+    {
+      "epoch": 4.7854251012145745,
+      "grad_norm": 0.20635071935970495,
+      "learning_rate": 2.2418695307376393e-07,
+      "loss": 0.1306,
+      "step": 1182
+    },
+    {
+      "epoch": 4.7894736842105265,
+      "grad_norm": 0.22653905056221343,
+      "learning_rate": 2.1582198325420968e-07,
+      "loss": 0.1361,
+      "step": 1183
+    },
+    {
+      "epoch": 4.793522267206478,
+      "grad_norm": 0.22007322366970078,
+      "learning_rate": 2.0761520719488447e-07,
+      "loss": 0.1187,
+      "step": 1184
+    },
+    {
+      "epoch": 4.797570850202429,
+      "grad_norm": 0.2096832090095806,
+      "learning_rate": 1.9956669051694798e-07,
+      "loss": 0.1348,
+      "step": 1185
+    },
+    {
+      "epoch": 4.801619433198381,
+      "grad_norm": 0.21286257479892748,
+      "learning_rate": 1.916764975761254e-07,
+      "loss": 0.144,
+      "step": 1186
+    },
+    {
+      "epoch": 4.805668016194332,
+      "grad_norm": 0.2148444733807671,
+      "learning_rate": 1.839446914621923e-07,
+      "loss": 0.1203,
+      "step": 1187
+    },
+    {
+      "epoch": 4.809716599190283,
+      "grad_norm": 0.2197086198028541,
+      "learning_rate": 1.7637133399845962e-07,
+      "loss": 0.1411,
+      "step": 1188
+    },
+    {
+      "epoch": 4.813765182186235,
+      "grad_norm": 0.2251044707981646,
+      "learning_rate": 1.6895648574129397e-07,
+      "loss": 0.147,
+      "step": 1189
+    },
+    {
+      "epoch": 4.817813765182186,
+      "grad_norm": 0.21649068104847416,
+      "learning_rate": 1.6170020597962465e-07,
+      "loss": 0.1634,
+      "step": 1190
+    },
+    {
+      "epoch": 4.821862348178137,
+      "grad_norm": 0.21846080486512326,
+      "learning_rate": 1.5460255273447078e-07,
+      "loss": 0.1168,
+      "step": 1191
+    },
+    {
+      "epoch": 4.825910931174089,
+      "grad_norm": 0.22714078150242792,
+      "learning_rate": 1.4766358275847715e-07,
+      "loss": 0.1391,
+      "step": 1192
+    },
+    {
+      "epoch": 4.82995951417004,
+      "grad_norm": 0.2133980764054063,
+      "learning_rate": 1.4088335153546573e-07,
+      "loss": 0.1475,
+      "step": 1193
+    },
+    {
+      "epoch": 4.834008097165992,
+      "grad_norm": 0.21692882481593956,
+      "learning_rate": 1.3426191327998496e-07,
+      "loss": 0.132,
+      "step": 1194
+    },
+    {
+      "epoch": 4.838056680161944,
+      "grad_norm": 0.2284480310298406,
+      "learning_rate": 1.277993209368744e-07,
+      "loss": 0.1453,
+      "step": 1195
+    },
+    {
+      "epoch": 4.842105263157895,
+      "grad_norm": 0.22033775995773255,
+      "learning_rate": 1.214956261808542e-07,
+      "loss": 0.1261,
+      "step": 1196
+    },
+    {
+      "epoch": 4.846153846153846,
+      "grad_norm": 0.21601872004577408,
+      "learning_rate": 1.1535087941609846e-07,
+      "loss": 0.1405,
+      "step": 1197
+    },
+    {
+      "epoch": 4.850202429149798,
+      "grad_norm": 0.21081685136776243,
+      "learning_rate": 1.0936512977583802e-07,
+      "loss": 0.1285,
+      "step": 1198
+    },
+    {
+      "epoch": 4.854251012145749,
+      "grad_norm": 0.2310794716811077,
+      "learning_rate": 1.0353842512196732e-07,
+      "loss": 0.1361,
+      "step": 1199
+    },
+    {
+      "epoch": 4.8582995951417,
+      "grad_norm": 0.2193232220794139,
+      "learning_rate": 9.787081204466253e-08,
+      "loss": 0.1075,
+      "step": 1200
+    },
+    {
+      "epoch": 4.862348178137652,
+      "grad_norm": 0.2672262259496977,
+      "learning_rate": 9.236233586200627e-08,
+      "loss": 0.1386,
+      "step": 1201
+    },
+    {
+      "epoch": 4.866396761133603,
+      "grad_norm": 0.21402552995441212,
+      "learning_rate": 8.70130406196279e-08,
+      "loss": 0.1309,
+      "step": 1202
+    },
+    {
+      "epoch": 4.870445344129554,
+      "grad_norm": 0.22448580701757084,
+      "learning_rate": 8.182296909035047e-08,
+      "loss": 0.1257,
+      "step": 1203
+    },
+    {
+      "epoch": 4.874493927125506,
+      "grad_norm": 0.2071647787098745,
+      "learning_rate": 7.679216277384439e-08,
+      "loss": 0.1403,
+      "step": 1204
+    },
+    {
+      "epoch": 4.8785425101214575,
+      "grad_norm": 0.214181361646021,
+      "learning_rate": 7.19206618963053e-08,
+      "loss": 0.1297,
+      "step": 1205
+    },
+    {
+      "epoch": 4.882591093117409,
+      "grad_norm": 0.22000194035374696,
+      "learning_rate": 6.720850541012347e-08,
+      "loss": 0.1219,
+      "step": 1206
+    },
+    {
+      "epoch": 4.886639676113361,
+      "grad_norm": 0.2232135635936407,
+      "learning_rate": 6.265573099357714e-08,
+      "loss": 0.1212,
+      "step": 1207
+    },
+    {
+      "epoch": 4.890688259109312,
+      "grad_norm": 0.2153913698784519,
+      "learning_rate": 5.826237505053067e-08,
+      "loss": 0.1224,
+      "step": 1208
+    },
+    {
+      "epoch": 4.894736842105263,
+      "grad_norm": 0.22508144889537254,
+      "learning_rate": 5.4028472710137e-08,
+      "loss": 0.1295,
+      "step": 1209
+    },
+    {
+      "epoch": 4.898785425101215,
+      "grad_norm": 0.2156942651187748,
+      "learning_rate": 4.9954057826571146e-08,
+      "loss": 0.1295,
+      "step": 1210
+    },
+    {
+      "epoch": 4.902834008097166,
+      "grad_norm": 0.21641687961673758,
+      "learning_rate": 4.6039162978743776e-08,
+      "loss": 0.1331,
+      "step": 1211
+    },
+    {
+      "epoch": 4.906882591093117,
+      "grad_norm": 0.2060463421429199,
+      "learning_rate": 4.228381947004812e-08,
+      "loss": 0.124,
+      "step": 1212
+    },
+    {
+      "epoch": 4.910931174089069,
+      "grad_norm": 0.2120878666701048,
+      "learning_rate": 3.868805732811343e-08,
+      "loss": 0.1278,
+      "step": 1213
+    },
+    {
+      "epoch": 4.91497975708502,
+      "grad_norm": 0.21162872671627805,
+      "learning_rate": 3.525190530455636e-08,
+      "loss": 0.1234,
+      "step": 1214
+    },
+    {
+      "epoch": 4.919028340080971,
+      "grad_norm": 0.22509160543389453,
+      "learning_rate": 3.197539087475887e-08,
+      "loss": 0.1355,
+      "step": 1215
+    },
+    {
+      "epoch": 4.923076923076923,
+      "grad_norm": 0.22371668363760638,
+      "learning_rate": 2.8858540237641786e-08,
+      "loss": 0.1398,
+      "step": 1216
+    },
+    {
+      "epoch": 4.9271255060728745,
+      "grad_norm": 0.21371720754643492,
+      "learning_rate": 2.5901378315458247e-08,
+      "loss": 0.1232,
+      "step": 1217
+    },
+    {
+      "epoch": 4.931174089068826,
+      "grad_norm": 0.20812120840462855,
+      "learning_rate": 2.3103928753598347e-08,
+      "loss": 0.1243,
+      "step": 1218
+    },
+    {
+      "epoch": 4.935222672064778,
+      "grad_norm": 0.22315275646078012,
+      "learning_rate": 2.0466213920393718e-08,
+      "loss": 0.1308,
+      "step": 1219
+    },
+    {
+      "epoch": 4.939271255060729,
+      "grad_norm": 0.21497812439014624,
+      "learning_rate": 1.7988254906942115e-08,
+      "loss": 0.1177,
+      "step": 1220
+    },
+    {
+      "epoch": 4.94331983805668,
+      "grad_norm": 0.2137278555033474,
+      "learning_rate": 1.5670071526936447e-08,
+      "loss": 0.1169,
+      "step": 1221
+    },
+    {
+      "epoch": 4.947368421052632,
+      "grad_norm": 0.22001606442702004,
+      "learning_rate": 1.3511682316509344e-08,
+      "loss": 0.1262,
+      "step": 1222
+    },
+    {
+      "epoch": 4.951417004048583,
+      "grad_norm": 0.2214899968369224,
+      "learning_rate": 1.1513104534086605e-08,
+      "loss": 0.131,
+      "step": 1223
+    },
+    {
+      "epoch": 4.955465587044534,
+      "grad_norm": 0.2068048748256585,
+      "learning_rate": 9.67435416023843e-09,
+      "loss": 0.1385,
+      "step": 1224
+    },
+    {
+      "epoch": 4.959514170040486,
+      "grad_norm": 0.2133702489102633,
+      "learning_rate": 7.995445897566178e-09,
+      "loss": 0.1254,
+      "step": 1225
+    },
+    {
+      "epoch": 4.963562753036437,
+      "grad_norm": 0.21295495719626967,
+      "learning_rate": 6.476393170578021e-09,
+      "loss": 0.132,
+      "step": 1226
+    },
+    {
+      "epoch": 4.967611336032388,
+      "grad_norm": 0.22013185074497565,
+      "learning_rate": 5.11720812557348e-09,
+      "loss": 0.1241,
+      "step": 1227
+    },
+    {
+      "epoch": 4.97165991902834,
+      "grad_norm": 0.22153399481438754,
+      "learning_rate": 3.917901630563492e-09,
+      "loss": 0.1403,
+      "step": 1228
+    },
+    {
+      "epoch": 4.9757085020242915,
+      "grad_norm": 0.21462465750254087,
+      "learning_rate": 2.878483275170485e-09,
+      "loss": 0.1438,
+      "step": 1229
+    },
+    {
+      "epoch": 4.979757085020243,
+      "grad_norm": 0.25424698456036654,
+      "learning_rate": 1.998961370555108e-09,
+      "loss": 0.1347,
+      "step": 1230
+    },
+    {
+      "epoch": 4.983805668016195,
+      "grad_norm": 0.22408847040532623,
+      "learning_rate": 1.2793429493518361e-09,
+      "loss": 0.1365,
+      "step": 1231
+    },
+    {
+      "epoch": 4.987854251012146,
+      "grad_norm": 0.25648141696729915,
+      "learning_rate": 7.196337656112384e-10,
+      "loss": 0.1308,
+      "step": 1232
+    },
+    {
+      "epoch": 4.991902834008097,
+      "grad_norm": 0.21330364435803859,
+      "learning_rate": 3.198382947511292e-10,
+      "loss": 0.1314,
+      "step": 1233
+    },
+    {
+      "epoch": 4.995951417004049,
+      "grad_norm": 0.21169224674987108,
+      "learning_rate": 7.995973352770137e-11,
+      "loss": 0.13,
+      "step": 1234
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 0.21086930282181018,
+      "learning_rate": 0.0,
+      "loss": 0.13,
+      "step": 1235
+    },
+    {
+      "epoch": 5.0,
+      "step": 1235,
+      "total_flos": 2.0745947931067023e+18,
+      "train_loss": 0.28116225621960905,
+      "train_runtime": 21824.7333,
+      "train_samples_per_second": 7.239,
+      "train_steps_per_second": 0.057
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1235,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 5,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.0745947931067023e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}